Pagei

AdvancesinGiardiaResearch
editedby:
PeterM.Wallis
KananaskisCentreforEnvironmentalResearch
UniversityofCalgary,Calgary,ABCANADAT2N1N4
and
BrianR.Hammond
AlbertaEnvironment,10405JasperAve.,
Edmonton,ABCANADAT5K3N4

TheUniversityofCalgaryPress

Pageii

Disclaimer:
Thisbookcontainscharacterswithdiacritics.WhenthecharacterscanberepresentedusingtheISO88591characterset(http://www.w3.org/TR/images/latin1.gif),
netLibrarywillrepresentthemastheyappearintheoriginaltext,andmostcomputerswillbeabletoshowthefullcharacterscorrectly.Inordertokeepthetext
searchableandreadableonmostcomputers,characterswithdiacriticsthatarenotpartoftheISO88591listwillberepresentedwithouttheirdiacriticalmarks.
1988KananaskisCentreforEnvironmentalResearch,UniversityofCalgary.Allrightsreserved.
ISBN0919813860
TheUniversityofCalgaryPress,
2500UniversityDriveNW,
Calgary,ABCANADAT2N1N4
CanadianCataloguinginPublicationData
Mainentryundertitle:
AdvancesinGiardiaresearch
PapersfromtheCalgaryGiardiaConference
heldFeb.2325,1987.
Includesindex.
ISBN0919813860
1.GiardialambliaCongresses.2.
GiardiasisCongresses.I.Wallis,Peter
Malcolm.II.Hammond,BrianR.,1934
III.CalgaryGiardiaConference(1987:
Calgary,Alta.)
QR201.G45A381989616'.016C880916370
Nopartofthisbookmaybereproduced,storedinaretrievalsystem,ortransmittedinanyformorbyanymeans,electronic,mechanical,photocopying,microfilming,
recording,orotherwise,withoutwrittenpermissionfromthePublisher.
PrintedinCanada

Pageiii

PREFACE
ThepapersinthisbookwerepreparedfromtheCalgaryGiardiaConferencewhichwasintendedtoprovideaforumforthereportingandsummarizingoftheresults
ofrecentresearchanddevelopmentinthestudyofthisimportant,worldwideparasite.AllofthepaperswerereviewedbytheEditorsandtheChairpersonofthe
sessioninwhichtheywerepresented.Theeditorshavemadetheformatofthepapersasuniformaspossiblebuthavenotattemptedtostandardizespellingin
recognitionoftheinternationalnatureofthepaperscontainedinthisvolume.
TheConferencewasattendedby150scientists,engineers,andpublichealthofficialsfromCanada,theUnitedStates,CentralAmerica,theUnitedKingdom,
Australia,andEurope.TheirenthusiasticparticipationwasdirectlyresponsibleformakingtheConferenceasuccess.
Theeditorswouldliketoexpresstheirdeepappreciationtothefollowingagenciesandindividualsfortheireffortsandsupport.
TheCalgaryGiardiaConferencewasfundedbytheAlbertaEnvironmentalResearchTrust,AlbertaEnvironment,HealthandWelfareCanada,theAlbertaHeritage
FoundationforMedicalResearch,theAlbertaEnvironmentalCentreandtheUniversityofCalgary.Withouttheirsupport,theConferenceandthisvolumewouldnot
havebeenpossible.
Theeditorswishtothankalloftheauthorsofscientificpapersfortheirexcellentpresentationsandpatiencethroughoutthelengthyprocessofpublication.
TheConferenceandthisvolumewouldnothavebeenpossiblewithouttheorganizationalandwordprocessingeffortsofGraceLebelandthepublicrelationswork
ofJaniceCrowther.
TheeditorsareespeciallyindebtedtoJaneLancasterwhomastereddesktoppublishinginrecordtime,toAnneHannanforheraccurateandpatientformattingof
manypapersandtoDaveSavageforhisimaginativeprogrammingwhichsavedusallanenormousamountofwork.
WealsowishtoacknowledgetheassistanceofTerryZenith,SectionHeadoftheAlbertaEnvironmentDraftingPoolforhishelpinoverhaulingsomeofour
graphics.
PETERWALLIS
BRIANHAMMOND

Pageiv

ThisbookisdedicatedtomywifeMarcia
withoutwhoseunderstandingandsupport
thisprojectwouldneverhavebeencompleted

Pagev

ADVANCESINGIARDIARESEARCH
iii

Preface
Epidemiology,PathogenesisandDrugSensitivity

DrugResistanceandtheTreatmentofGiardiasis
P.F.L.Boreham,N.C.SmithandR.W.Shepherd
CellInjuryinGiardialambliaDetectedbyForwardLightScatter
B.Kinosian,R.H.Gilman,J.Ordonez,J.O'Hare,S.Wahl,F.Kosterand
W.Spira

913

TheImportanceofNonwaterborneModesofTransmissionforGiardiasis,A
CaseStudy
S.Harley

1519

ANewMinicultureTechniquefordeterminingInVitroAntimicrobialAgent
SensitivityofAxenicallyCultivatedStrainsofGiardialamblia
S.M.Wahl,R.H.Gilman,J.P.O'Hare,D.B.KeisterandW.M.Spira

2124

UltrastructuralStudyofaBacterialSymbiontofGiardialamblia
S.Radulescu,E.A.Meyer,B.BurgheleaandT.Meitert

2528

MorphologyofGiardiaEncystationInVitro
D.G.Schupp,M.M.JanuschkaandS.L.Erlandsen

2932

CytopathogenicityofGiardialambliainHeLaandVeroCellMonolayers
A.JyothisriandU.K.Baveja

3337

StudiesonthePrevalenceofGiardiasisinCzechoslovakia
M.Giboda

3941

Immunology

ImmunologyofGiardiaInfections
M.F.Heyworth

4548

TheSecretoryImmuneResponseinRatsInfectedwithRodentGiardiaduodenalis
IsolatesandEvidenceforPassiveProtectionwithImmuneBile
G.MayrhoferandA.WaightSharma

4954

BiologicalDifferencesinGiardialamblia
T.E.NashandA.Aggarwal

5558

AnimalModelsandCrossInfection

37

PrevalenceofGiardiasp.inDogsfromAlberta
P.D.Lewis,Jr.

6164

LocationofGiardiaTrophozoitesintheSmallIntestineofNaturallyInfectedDogs
inSanDiego
H.Douglas,D.S.Reiner,M.J.GaultandF.D.Gillin

6569

SeasonalIncreaseintheIncidenceofGiardialambliainArkansas
J.J.Daly,M.A.Gross,D.McCullough,T.McChesney,S.K.Tank,E.B.
DalyandC.L.Puskarich

7174

Pagevi

InfectionofMongolianGerbils(Merionesunguiculatus)withGiardiafromHuman
andAnimalSources
K.D.Swabby,C.P.HiblerandJ.G.Wegrzyn

7577

TransmissionofGiardiaduodenalisfromHumanandAnimalSourcesinWild
Mice
P.D.RoachandP.M.Wallis

7982

WaterTreatment

WaterTreatmentandtheGiardiaCyst
A.vanRoodselaar

8586

RemovalofGiardiaThroughSlowSandFiltration100MileHouse,British
Columbia
J.M.G.Bryck,B.L.WalkerandD.W.Hendricks

8793

ComparisonofSomeFiltrationProcessesAppropriateforGiardiaCyst
Removal
G.S.Logson
MonitoringasaToolinWaterborneGiardiasisPrevention
J.L.Sykora,W.D.Bancroft,A.H.Brunwasser,S.J.States,M.A.Shapiro,
S.N.BoutrosandL.F.Conley

103
106

TheEfficiencyofPointofUseDevicesfortheExclusionofGiardiamuriscysts
fromaModelWaterSupplySystem
D.R.CullimoreandH.Jacobsen

107
112

DiatomiteFiltration:WhyitRemovesGiardiafromWater
H.G.Walton

113
116

SmallWaterSystemImprovementsforGiardiaRemovalACaseStudy
M.R.Alberi,S.J.QuailandR.A.Kruse

117
124

InactivationofGiardialambliaCystsfromaSurfaceWaterbyOxidationwith
Ozone
C.Nebel,A.Lally,T.Bosher,J.W.Hmurciak,L.HmurciakandD.A.Breen

125
128

ARegulatoryAgency'sExperiencewithGiardia
S.McClureandI.B.Mackenzie

129
131

EffectsofChlorineontheUltrastructureofGiardiaCysts
M.Neuwirth,P.D.Roach,J.M.BuchananMappinandP.M.Wallis

133
135

RemovalandInactivationofGiardiaCystsinaMobileWaterTreatmentPlant
UnderFieldConditions:PreliminaryResults
P.M.Wallis,J.S.Davies,R.Nutbrown,J.M.BuchananMappin,P.D.
RoachandA.vanRoodselaar

137
144

DifferentiationofGiardiaIsolates

95102

TheGenomeofGiardiaintestinalis
P.Upcroft,P.F.L.BorehamandJ.A.Upcroft

147
152

ThePartialCharacterizationofanImmunodominantAntigenofGiardia
intestinalis
J.A.Upcroft,A.G.Capon,A.DharmkrongAt,P.Upcroft,andP.F.L.
Boreham

153
157

ImmunofluorescenceDifferentiationBetweenVariousAnimalandHumanSource
GiardiaCystsUsingMonoclonalAntibodies
H.H.Stibbs,E.T.Riley,J.Stockard,J.L.Riggs,P.M.WallisandJ.Issac
Renton

159
163

ComparisonofGiardiaIsolatesbyDNADNAHybridization
A.Uji,P.M.WallisandW.M.Wenman

165
167

Pagevii

DifferentiationofGiardiaduodenalisfromGiardiamurisbyImmobilizationin
VariousSera
D.L.LehmannandP.M.Wallis

169
172

ConservedSequencesoftheHSPGeneFamilyinGiardialamblia
A.Aggarwal,P.Romans,V.F.delaCruzandT.E.Nash

173
175

TheResponseofHumanstoAntigensofGiardialamblia
M.G.OrtegaPierres,R.Lascurain,R.A.Garcia,R.C.Vazquez,G.Acosta
andJ.I.Santos

177
180

PropertiesofGiardialambliaRNAs.
C.Montanez,L.Cervantes,C.OvandoandM.G.OrtegaPierres

EnzymeActivitesofGiardialambliaandGiardiamurisTrophozoitesandCysts
D.G.Lindmark,andJ.J.Miller

187
189

StudiesonGiardialambliaTrophozoiteAntigensUsingSephacrylS300Column
Chromatography,PolyacrylamideGelElectrophoresisandEnzymelinked
ImmunosorbentAssay
P.P.Chaudhuri,S.Pal,S.C.Pal,andP.Das

191
194

DetectionofGiardiaCysts

181
185

AnOverviewoftheTechniquesUsedforDetectionofGiardiaCystsinSurface
Water
C.P.Hibler

197
204

MethodsfortheRecoveryofGiardiaandCryptosporidiumfromEnvironmental
WatersandtheirComparativeOccurrence
J.B.Rose,D.Kayed,M.S.Madore,C.P.Gerba,M.J.Arrowood,C.R.
SterlingandJ.L.Riggs

205
209

ComparisonofFiveProceduresfortheSedimentationofGiardialambliaand
OtherProtozoanCysts
D.R.Pennell,J.F.Stoebig,D.E.Sampson,andR.F.Schell

211
213

ComparisonoftheModified"ReferenceMethod"andtheIndirectFluorescent
AntibodyTechniqueforDetectionofGiardiaCystsinWater
B.E.Quinones,C.P.HiblerandC.M.Hancock.

215
217

GiardiaDetectionusingMonoclonalAntibodiesRecognizingDeterminantsofIn
VitroDerivedCysts
C.R.Sterling,R.M.Kutob,M.J.Gizinski,M.Verastegui,andL.
Stetzenbach

219
222

RoutineMonitoringofWatershedsforGiardiaCystsinNortheastern
Pennsylvania
S.A.M.McFarlane

223
225

WaterborneGiardiasis:SourcesofGiardiaCystsandEvidencePertainingtotheir
ImplicationinHumanInfection
S.L.ErlandsenandW.J.Bemrick

227
236

AnalysisofMunicipalWaterSamplesforCystsofGiardia
C.P.Hibler

237
245

Pageviii

ViabilityTesting
AReviewofMethodsthatareusedtoDetermineGiardiaCystsViability
F.W.Schaefer,III

249
254

FluorescentDyeExclusionasaMethodforDeterminingGiardiaCystViability
S.J.Hudson,J.F.SauchandD.G.Lindmark

255
259

ANewMethodforExcystationofGiardia
J.F.Sauch

261
264

AssessingGiardiaCystsViabilitywithFluorogenicDyes:ComparisonstoAnimal
InfectivityandCystMorphologybyLightandElectronMicroscopy
D.G.Schupp,M.M.JanuschkaandS.L.Erlandsen

265
269

PanelDiscussions
ExcystationandEncystation
F.D.Gillin,E.A.Meyer,S.Erlandsen,C.Sterling

273

TheImplicationsofRegulatoryChangesforWaterTreatmentintheUnited
States
S.Regli,A.Amirtharajah,B.Borup,C.Hibler,J.Hoff,andR.Tobin

275
286

TaxonomyoftheGenusGiardia
S.L.Erlandsen,E.A.Meyer,T.E.Nash

287
289

MethodsofHandlingGiardiaintheLaboratory
W.Jakubowski,E.A.Meyer,T.E.Nash,C.P.Hibler

291
294

Index

295
302

Page1

EPIDEMIOLOGY,PATHOGENESISANDDRUGSENSITIVITY

Page3

DrugResistanceandtheTreatmentofGiardiasis
P.F.L.Boreham*,N.C.SmithandR.W.Shepherd
QueenslandInstituteofMedicalResearch,BramstonTerrace,Herston,Brisbane,Qld.4006,Australia
ThepossibleexistenceofdrugresistantGiardiaintestinalishasbeeninvestigated,asapossibleexplanationfortreatmentfailuresinpatients.Analysisof15isolatesofG.intestinalis
hasdemonstratedmajordifferencesinsensitivitiestometronidazole,tinidazole,furazolidoneandquinacrine.Eachisolateisheterogeneousandiscomposedofpopulationsof
parasiteswithdifferingdrugsensitivitiesanddoublingtimes.Crossresistancebetweenthe5nitroimidazoleshasbeendemonstratedinaninvitrotest.Clinicalandlaboratorydata
providestrongevidencefordrugresistanceinG.intestinalis.Investigationsofthemolecularbasisofdrugresistancesuggestthatdifferentmechanismsoccurwiththenitrofurans
andthe5nitroimidazoles,withtheformerbeingrelatedtotheglutathionecyclingenzymes,glutathioneperioxidaseandgluthathionereductaseandthelattertopyruvate:
ferredoxinoxidoreductaseactivity.

Introduction
Managementofpatientswithgiardiasisoftenprovestobedifficultduetoproblemsrelatedtoaccuratediagnosis,lackofknowledgeconcerningpathophysiologyanda
lackoffullyeffectivechemotherapeuticagents(8,10,11).Researchintothetreatmentofgiardiasishasbeenlimitedbythelackofsuitablelaboratorymodels.Allthe
existingdrugsweredevelopedforotherinfectiousdiseasesandsubsequentlyfoundempiricallytobeactiveagainstGiardiaintestinalis.Inthispaperwebriefly
reviewtheexistingdrugsandtheirdeficienciesanddiscusssomeofourcurrentresearchwhichisdesignedtoeffectimprovementsinthetherapyofinfectedhumans,
particularlychildren.
TheCurrentArmamentarium
Fourdrugs,metronidazole,tinidazole,furazolidoneandquinacrinearecommonlyusedtotreatgiardiasis,butthechoiceislargelydependentuponthepersonal
preferenceoftheprescribingphysician,drugavailabilityandtoadegree,theoccurrenceofuntowardeffects.Thetwo5nitroimidazolescommonlyusedare
metronidazoleandtinidazole.Thesedrugsmaycausenausea,gastrointestinaldiscomfort,lassitude,skinrashes,drowsiness,disulfiramlikereactionswithalcoholand
occasionallytransientleucopeniaandperipheralneuropathy.Nitroimidazoleshavebeenshowntobecarcinogenicinrodentsandmutagenicinbacteria.Singleand
multidoseregimenshavebeenevaluatedbutthereisnotgeneralconcensusonthemostappropriatecourseoftreatment(10,11).Athirdnitroimidazole,ornidazole,
hasbeenevaluatedandappearstobeequipotenttotinidazole(22).Anewmemberofthisgroup,satranidazole,iscurrentlyundergoingclinicaltrialsinIndiaandthe
preliminarydatalookmostpromising(20).
Manyphysiciansconsiderfurazolidone,a2nitrofuran,tobethedrugofchoiceforthetreatmentofgiardiasisinyoungchildren.However,awiderangeofmildside
effectscanoccur,includingheadache,nausea,vomiting,skinrashes,diarrheaandmalaise.Moreseveresideeffects,suchasagranulocytosisandhemolyticanemia,in
patientswithglucose6phosphatedehydrogenasedeficiency,mayoccur.QuinacrineiscommonlyusedinNorthAmerica.Thisantimalarialcompoundmaycause
dizziness,headacheandgastrointestinaldisturbances.Thefactthatquinacrinecausestoxicpsychosesin12%ofpatients,togetherwithoccasionalcasesofexfoliative
dermatitisandaplasticanemia,hasresultedinthisdrugnotbeingusedbysomephysicians.Paromomycinsulfatehasbeenrecommendedforthetreatmentofgiardiasis
inpregnancy,mainlybecause,asanaminoglycoside,itisnotabsorbedfromthegut(14).However,controlledtrialshavenotyetbeenconductedanditshouldbe
usedwithcaution.Otherdrugswhichhavebeenrecommendedforthetreatmentofgiardiasisincludeamodiaquine(21),berberinesulfate(13),sulfasalazine(1)and
erythromycin(19)butagainnoneofthesedrugshavebeenexposedtocontrolledclinicaltrials.
Amajorproblemwiththecurrentdrugsisthattreatmentfailuresareknowntooccurwithallofthem.Itisverydifficulttoassignaccuratefigurestothesefailurerates
sinceeverystudyusesdifferentassessmentcriteriaforcure,rangingfromasinglestoolexaminationtomultipleexaminationsoverseveralmonthstogetherwithasmall
intestinalbiopsy.Basedon11publishedreportsmetronidazolehasacurerateof4695%,tinidazole88100%,furazolidone5892%andquinacrine60100%.Itis
generallyacceptedthattreatmentfailuresdooccurwithallfourdrugsandthatthisposesaseriousproblemtophysicians.
Manyreasonscanbepostulatedtoexplainthesetreatmentfailuresincluding:
patientnoncompliancewiththeprescribeddrugregimen.Thisiscertainlyanimportantconsiderationandrecentstudieshavedemonstratedmajorproblemsinthis
area(3)
thepossiblereinfectionofthepatient.Atpresentthereisnowaytoeffectivelymonitorthisbytypingisolates
changesinthepharmacokineticsofthedrug
*Correspondingauthor.

Page4

possibleescapeoforganismstopriviligedsiteswhereantigiardialdrugsareunabletoreach
deficienciesinthehost'simmunesystem
theinactivationofthedrugbyconcommitantbacterialinfections(12)
theexistenceofdrugresistantstrains.ResistancetometronidazolehasbeenwelldocumentedintheTrichomonadsandalsoinsomeanaerobicbacteria.
ScreeningTestsAgainstGiardiaintestinalis
1.Invitro
Researchondrugsforthetreatmentofgiardiasishasbeenseverelyhinderedbythelackofappropriatescreeningtests.Mostworkhaseitherinvolvedtestingdrugs
directlyonman,wherestandardizationandassessmentofcurehavebeenproblems,orbyusingGiardiamurisinthemouseasamodel(2,11).Inordertoinvestigate
thepossibleexistenceofresistantstrainsofG.intestinaliswefirstdevelopedappropriatestandardizeddrugscreeningtests.Cultureinmicrotitretrayswasachieved
byincubatingtheplatesinanatmosphereofnitrogeninsealedcontainersandatesttomeasurereproductiveviability,basedontheuptakeof[3H]thymidineintothe
nucleioftheorganisms,wasdeveloped(4).Thisassayprovedtobeconsiderablymoresensitivethanusingeitherflagellarmovementordyeexclusionasanindexof
viability.DevelopmentofthisinvitrotesthasallowedthedrugsensitivitiesofdifferentisolatesofhumanG.intestinalistobecomparedandcompoundstobe
screenedfortheiractivityagainstG.intestinalis(5).
Analysisof15isolatesculturedfrompatientsattendingtheRoyalChildren'sHospital,Brisbane,hasshownthatthereisatenfolddifferenceinsensitivitybetween
theseisolatesformetronidazoleandfurazolidone,athreefolddifferencefortinidazoleandatwentyfolddifferenceforquinacrine(8,15andunpublisheddata).In
addition,byexaminingthedrugsensitivitiesofclonedlinesderivedfromasinglestockithasbeenshownthateachisolateofG.intestinalisisnothomogeneous,butis
composedofdifferentpopulationsoforganismshavingdifferingdrugsensitivities(7).Doublingtimesofthestocksalsovaryconsiderably,rangingfrom12.5to44.5
hourswhengrowninaxeniccultureintheabsenceofbilefromthemedium(15).
2.Invivo
Aninvivotestfordrugsensitivityhasalsobeendevelopedusinganeonatalmousemodel(6).Littersofmicelessthan5daysofageareinfectedwith3104
trophozoitesviaanintragastrictubeandafter6dayshalfofeachlitteraretreatedwithanappropriateconcentrationofthedrugunderstudyin50Lofvehiclealsoby
theintragastricroute.Theotherhalfofthelitteractascontrolsandaretreatedwiththevehiclealone.Afurther2dayslaterthemicearekilled,thesmallintestine
removed,openedlongitudinallyandplacedincoldbuffertoallowthetrophozoitestodetach.Thetotalnumberofparasitespresentcanthenbecountedand
expressedasapercentageoftheuntreatedcontrols.Usingthistechniqueanumberofcompoundshavebeenassayedandithasbeenshownthatthereisadirect
correlationbetweeninvitroandinvivoactivityfortwelve5nitroimidazoles(P<0.05)(6).Experimentalresultshavealsoshownthattheobservedactivityinmanfor
thecommonlyuseddrugsisreflectedinthismodelsystem.Forexample,tinidazoleismoreactivethanmetronidazolebutroughlyequipotentwithornidazole.
Furazolidoneisslightlylessactivethantinidazolebutmoreactivethanquinacrine.Itisinterestingtonotethattheseinvivostudieshaveindicatedthatberberinesulfate,
sulfasalazineanderythromycinareinactive.Paromomycinsulfateisactivewhengivenbytheoralroutetoneonatalmicebutisinactivewheninjectedsubcutaneously.
TheseresultsindicatethatisolatesofG.intestinalisexistwhichhavedifferingsensitivitiesforthefourdrugscommonlyusedtotreatgiardiasisandthatasinglestockis
heterogeneous,beingcomposedofpopulationsoforganismshavingdifferentdrugsensitivities.
Fromstudiessuchastheseitisnotpossibletodefinewhatconstitutesresistantorsensitiveorganisms.Suchinformationcanonlybeobtainedfromappropriatestudies
onpatients.Byanalogywiththeearlyworkonmalariadrugresistance,itisnecessarytocompareinvitroandinvivoresponsesofalargenumberofisolateswiththe
clinicalresponseandeffectonparasitenumbersfollowingtreatment.Onlythenwillitbepossibletonominatedrugconcentrationswhere,iftheorganismsarenotkilled
invitro,drugresistancecanbesuspected.Preliminaryresultstoinvestigatethisquestionaregivenbelow.
EvidenceforDrugResistanceinHumanPatients
Todatewehavestudiedindetailtenchildrentreatedwithfurazolidone(8mg/kg/day)alone(15).Clinicaldatawerecollectedbeforeandaftertherapyandstudies
conductedontheorganismstakenfromthesepatientsandestablishedinaxenicculture.Sevenofthetenpatientshadpromptresolutionofsymptomsbutinthree
patientsthesymptomspersisted.Clinicalcureonlywasusedinthesestudiedbecauseofthewellknowndifficultiesassociatedwithaccuratelydocumenting
parasitologicalcure.Twoofthethreeisolatesfrompatientswithpersistentsymptomsweretheleastsensitiveofthetentestedtofurazolidoneinvitro(Figure1).
Thesetwopatientsdidnotrespondclinically

Figure1.
Furazolidonesensitivityoftrophozoitesisolated
frompatientsshowingeitheraclinicalresponse
ornoclinicalresponsetotherapy.ID50theconcentration
ofdrugrequiredtoinhibittheuptakeof[3H]thymidine
invitroby50%.*patienttreatedfor5daysonly.

Page5

toasecondtendaycourseoftreatment.Theisolatesfromthesetwopatientsfellwithinthenormalrangeformetronidazoleandtinidazolesensitivity.Bothpatients
respondedsatisfactorilywhentreatedwithanitroimidazole.Thethirdpatientinwhomfurazolidonetherapyfailed,hadorganismswithinvitrosensitivityintherange
wheretherapywouldbeexpectedtobesatisfactory.Thispatienthowever,receivedonly5daystreatmentinsteadoftherecommended10days(18)andwhengiven
acomplete10daycourseoffurazolidonerespondedsatisfactorily.
Thisstudyindicatesthatintwopatientseithertheorganismswereresistanttofurazolidoneorthatdifferencesinthepharmacokineticsofthedrugoccurred.Sincelittle
absorptionoffurazolidonetakesplacefromthegutthesecondexplanationwouldseemunlikely.Inaddition,theseresultsillustratethatitispossibletodesignspecific
therapyregimensforthosepatientswhoprovedtoberefractorytostandardtreatments.
SelectionofDrugResistanceintheLaboratory
ExperimentalselectionofadrugresistantlineofG.intestinalishasbeenachievedinthelaboratorybygrowingstockBRIS/83/HEPU/106inmediumcontainingtwice
thedoseofmetronidazolerequiredtokill10%oftheorganisms(2ID10).Overaperiodoftwelvemonthstheconcentrationofdrugrequiredtokill50%oftrophozoites
(ID50)increasedapproximatelytenfoldwhencomparedtothestockgrownintheabsenceofthedrug.Thisselectedlinehasbeenusedtoexaminethepossiblityof
theexistenceofcrossresistancebetweenthenitroimidazolesandalsotoinvestigatethemechanismsofresistance.

Figure2.
Thedecreaseinsensitivityofeleven5nitroimidazoles
whentestedagainstalineofstockBRIS/83/HEPU/106
selectedforresistancebygrowingtheorganisms
inasublethalconcentrationofmetronidazolefor
oneyear,comparedwiththestockgrowninthe
absenceofdrug:NIMnimorazoleS75S750400A
[1methyl2(1dimethylaminomethyleniminophenoxymethyl)
5nitroimidazolehydrochloride]FEXfexinidazole
ORNornidazolePANpanidazoleSATsatranidazole
FLUflunidazoleRONronidazoleTINtinidazole
SECsecnidazoleMETmetronidazole.

Figure3.
Diagramaticrepresentationofthemodeof
actionofmetronidazolebasedonMuller(17).

CrossResistanceBetweentheNitroimidazoles
Atotalofeleven5nitroimidazolecompoundshavebeentestedagainstthelaboratoryselectedlineofG.intestinalisandallshowedlesssusceptibilitytothedrug
whencomparedwiththecontrolstock(Figure2).Thereductioninsensitivityvariedbetweentwoandsixfold.Littleornoresistancewasfoundwithunrelateddrugs,
includingfurazolidoneandquinacrine.Inviewoftheconsiderableinterestshownrecentlybypharmaceuticalcompaniesincompoundsofthe5nitroimidazoleseriesfor
thetreatmentofprotozoalinfections,thisfindingisofimportancefordesigningtreatmentregimensforpatientswhofailtorespondeffectivelytometronidazoleor
tinidazole.
MechanismsofDrugResistance
AlthoughnostudiesonthemodeofactionofmetronidazolehavebeenconductedwithGiardiathereismuchinformationonitsmodeofactionandonpossible
mechanismsofresistancewithothermicroorganisms.ThemodeofactionofmetronidazolehasbeenreviewedbyMuller(17)andisshowndiagramaticallyinFigure
3.Itisknownthatmetronidazoleistakenupintocellsandreducedtoreactiveintermediateswiththeformationofassociatedfreeradicalswhichareresponsiblefor
thedrug'sactivitybycausingstrandbreakageandcrosslinkingofDNA.Drugresistancecouldresultinthreeways:
decreaseduptakeofmetronidazolebytheresistantcells
accumulationoffreemetronidazolewithinthecellresultingfromthefailureoftheformationofthereactiveintermediatesbytheenzymepyruvate:ferredoxin
oxidoreductase
failureofthereactivemetabolitesofmetronidazoletoactonDNA.

Page6

Themodeofactionoffurazolidonehasnotbeenstudiedinthesamedepthbutitseemsprobablethatitalsoproducesfreeradicalswithinthecellwhichcausedamage
toDNA.Thereisevidencethatthenitroimidazolesandnitrofuransactviaslightlydifferentmechanisms(16).Anaerobicreductionofmetronidazoleisdependentupon
thehydrogenosomalenzymepyruvate:ferredoxinoxidoreductasebutthenitrofuransmaybereducedbycytosolicenzymessuchasNADHandNADPHoxidaseas
wellaspyruvate:ferredoxinoxidoreductase.Thisisbecausenitrofuranshaveahigherpositivereductionpotentialthanthe5nitroimidazoles.Todatewehave
investigatedaspectsofthefirsttwomechanismsofresistance.
1.UptakeofMetronidazoleintotheCell
Theamountof[14C]metronidazoletakenupintothecellislessinthelineselectedforresistancethaninthesamestockgrownintheabsenceofmetronidazole.This
suggestsaccumulationofafreeintracellularpoolofmetronidazoleresultinginadecreaseduptakeacrosstheconcentrationgradientoralternatively,adefectiveuptake
mechanism.Sincetheuptakeofthenitroimidazolesisbelievedtobebyapassiveprocesstheformerexplanationwouldseemtobemostlikely.
2.EnzymeStudies
Theglutathioneredoxcyclingenzymes,glutathioneperoxidaseandglutathionereductasearepresentinmostcellsandareresponsibleforscavengingfreeradicals,
especiallyperoxides.Theythusactasaprotectivemechanismforthecell.Glutathionereductaseensuresthesupplyofreducedglutathioneforthisreactionaswellas
convertingNADPHtoNADP,thefirststepinthepentosephosphatepathway.Wehavemeasuredtheconcentrationofglutathioneperoxidaseandglutathione
reductasein15stocksofG.intestinalisandcorrelatedtheresultswithdrugsensitivitiestometronidazole,tinidazole,furazolidoneandquinacrinedeterminedbythein
vitro[3H]thymidineincorporationassay.Furazolidonesensitivityofthesestockscorrelatedwiththeactivityoftheglutathioneredoxcyclingenzymesbutthiswasnot
trueforthe5nitroimidazoles.Thisindicatesthatfurazolidoneresistanceisprobablyrelatedtofreeradicalscavengingbyglutathioneperoxidase,butthesamedoesnot
appeartobetrueformetronidazoleortinidazole.Asignificantnegativecorrelationbetweenscavengingenzymeconcentrationsandquinacrineresistancewasfoundin
thesestocksbuttheinterpretationofthisresultisunclear.
Pyruvate:ferredoxinoxidoreductaseappearstobeprimarilyresponsibleforthemetabolismandsubsequenttoxicityof5nitroimidazoles(16)anddeficienciesinthis
enzymehavebeenreportedtoaccountformetronidazoleresistanceinTritrichomonasfoetus(9).Preliminarydataindicatesthatthesamemechanismmaybetruefor
G.intestinalissincethemetronidazoledrugselectedlinehasonlyonesixththeamountofpyruvate:ferrodoxinoxidoreductaseactivitycomparedtotheparentisolate.
Thustheprimarydefectinthelineresistanttometronidazoleappearstobeadecreaseinthepyruvate:ferredoxinoxidoreductaseenzymeresultinginadecreasedrate
ofreductionofthedrug.Thiswilllowerthedrugconcentrationgradientacrossthecellandhencereducedruguptake.Theabsenceofsuperoxidedismutaseactivityin
anyoftheisolatesofG.intestinalissofarexaminedwouldfavourananaerobicinterpretationofdrugresistance.Thusthehydrogenosomalferredoxinsappeartoplay
acriticalroleinnitroimidazoletoxicity,whereasflavinesmayberesponsibleforthereductionofnitrofurans(16).
Conclusions
Thereisanurgentneedtodevelopnewdrugsandimprovedstrategiesfortreatmentofgiardiasistoovercomethedifficultiesofdrugfailuresandadversereactions
inherentintheexistingdrugs.Arationalapproachtothedevelopmentofnewchemotherapeuticagentsshouldincludeastudyofparasitespecificmetabolicpathways
anddevelopmentofanappropriateinhibitor.CurrentlyweknowverylittleofthemetabolicpathwaysofG.intestinalis.Anynewpotentialcompoundswouldneedto
betestedinthelaboratorybothinvitroandinaninvivomodelpriortoclinicaltrials.Untilrecentlythisapproachhasnotbeenpossiblebecauseofthelackof
suitablemodels,buttheworkdescribedhereisthefirststeptowardssucharationaldesignofantigiardialcompounds.
Therecognitionofthelikelihoodthatdrugresistantorganismsexistisanimportantconsideration.Althoughdrugresistanceappearstobeasignificantclinicalproblem
withonlyafewoftheparasiticdiseases,itsexistencewithatleasttwoflagellates,Trypanosomasp.andTrichomonassp.,meansthatcarefulsurveillanceofthe
situationmustbemaintained.Muchmoreinformationontheextentandsignificanceofresistanceingiardiasisisrequiredasourknowledgeofthisareaisverylimited.
Oneimportantimplicationofthesestudiesisthatitisnowpossibletodevelopspecifictherapyregimensforthosepatientsrefractorytostandardtreatments,provided
thatitispossibletoestablishaxenicculturesoftheirparasiticorganisms.
Acknowledgements
TheoriginalresearchdescribedinthispaperhasbeensupportedbygenerousgrantsfromtheNationalHealthandMedicalResearchCouncilofAustralia.Wethank
ProfessorC.BryantforhelpfuldiscussionsandMr.R.E.PhillipsandMrs.K.Andersonforexcellenttechnicalassistance.WealsothankG.D.SearleandCo.for
theirgenerousgiftfor[14C]metronidazoleandthefollowingcompaniesforgiftsofdrugs:BoehringerIngelheimPtyLtd.CibaGeigyResearchCentre,India
FarmitaliaCarloErbaLtd.HoechstAGMay&BakerLtd.Merck,Sharpe&Dohme(Aust)PtyLtd.NorwichEatonPtyLtd.ParkeDavisPtyLtd.PfizerPty
Ltd.RocheProductsPtyLtd.andSpecia.
LiteratureCited
1.Azizi,E.,J.Karpuchas,andT.Rosenberg.1983.Salicylazosulfapyridinetherapyinachildwithchroniclambliasisduetoacquiredhypogammaglobulinemia.Helv.
Paediatr.Acta38:8790.

Page7

2.Bemrick,W.J.1963.AcomparisonofsevencompoundsforgiardiacidalactivityinMusmusculus.J.Parasitol.49:819823.
3.Boreham,P.F.L.,S.Benrimoj,M.Ong,M.Craig,R.W.Sheperd,D.Hill,andB.Singleton.1986.Acompliancestudyinpediatricpatientsreceivingtreatmentfor
giardiasis.Aust.J.Hosp.Pharm.16:138142.
4.Boreham,P.F.L.,R.E.Phillips,andR.W.Shepherd.1984.ThesensitivityofGiardiaintestinalistodrugsinvitro.J.Antimicrob.Chemother.14:449461.
5.Boreham,P.F.L.,R.E.Phillips,andR.W.Shepherd.1985.Acomparisonoftheinvitroactivityofsome5nitroimidazolesandothercompoundsagainstGiardia
intestinalis.J.Antimicrob.Chemother.16:589595.
6.Boreham,P.F.L.,R.E.Phillips,andR.W.Shepherd.1986.TheactivityofdrugsagainstGiardiaintestinalisinneonatalmice.J.Antimicrob.Chemother.18:393
398.
7.Boreham,P.F.L.,R.E.Phillips,andR.W.Shepherd.1987.HeterogeneityintheresponsesofclonesofGiardiaintestinalistoantigiardialdrugs.Trans.R.Soc.
Trop.Med.Hyg.81:406407.
8.Boreham,P.F.L.,andR.W.Shepherd.1985.Thetreatmentofgiardiasis,pp.320326,InS.Tzipori(ed.),Infectiousdiarrhoeaintheyoung.Elsevier,Amsterdam.
9.Cerkasovova,A.,J.Cerkosov,andJ.Kulda.1984.MetabolicdifferencesbetweenmetronidazoleresistantandsusceptiblestrainsofTritrichomonasfoetus.
Mol.Biochem.Parasitol.11:105118.
10.Craft,J.C.1982.Giardiaandgiardiasisinchildhood.Pediatr.Infect.Dis.1:196211.
11.Davidson,R.A.1984.Issuesinclinicalparasitology:thetreatmentofgiardiasis.Am.J.Gastroenterol.79:256261.
12.Edwards,D.I.,E.J.Thompson,J.Tomusange,andD.Shanson.1979.Inactivationofmetronidazolebyaerobicorganisms.J.Antimicrob.Chemother.5:315
316.
13.Gupte,S.1975.Useofberberineintreatmentofgiardiasis.Am.J.Dis.Child.129:866.
14.Kreutner,A.K.,V.E.DelBene,andM.S.Amstey.1981.Giardiasisinpregnancy.Am.J.Obstet.Gynecol.140:895901.
15.McIntyre,P.,P.F.L.Boreham,R.E.Phillips,andR.W.Shepherd.1986.Chemotherapyofgiardiasis:clinicalresponsesandinvitrodrugsensitivityofhuman
isolatesinaxenicculture.J.Pediatr.108:10051010.
16.Moreno,S.N.J.,R.P.Mason,andR.Docampo.1984.DistinctreductionofnitrofuransandmetronidazoletofreeradicalmetabolitesbyTritrichomonasfoetus
hydrogenosomalandcytosolicenzymes.J.Biol.Chem.259:82528259.
17.Muller,M.1983.Modeofactionofmetronidazoleonanaerobicbacteriaandprotozoa.Surgery93:165171.
18.Murphy,T.V.,andJ.D.Nelson.1983.Fivevstendays'therapywithfurazolidoneforgiardiasis.Am.J.Dis.Child.137:267270.
19.Nash,P.H.1976.Giardiasis.Can.Med.Assoc.J.115:1819.
20.Poltera,A.A.,R.Figueroa,H.Reyes,T.Koura,andR.Degen.1986.SuccessfullowdosetherapyinGiardialambliapatientswithCibemideRanew5nitro
imidazole.Abstr.516,Handbook,SixthInt.Congr.Parasitol.M.J.Howell(ed.).AustralianAcademyofSciences,Canberra.
21.Rosenberg,J.,andE.Neumann.1957.Efficacyofamodiaquinhydrochloride(Camoquinhydrochloride)againstgiardiasis.Am.J.Trop.Med.Hyg.6:679680.
22.Sabchareon,A.,T.Chongsuphajaisiddhi,andP.Attanath.1980.Treatmentofgiardiasisinchildrenwithquinacrine,metronidazole,tinidazoleandornidazole.
SoutheastAsianJ.Trop.Med.PublicHealth11:280284.

Page9

CellInjuryinGiardialambliaDetectedbyForwardLightScatter
B.Kinosian*,R.H.Gilman,J.Ordonez,J.O'Hare,S.Wahl,F.Koster,andW.Spira
Dept.ofMedicine,FrancisScottKeyMedicalCenter,
TheJohnHopkinsUniversitySchoolofMedicine,Baltimore,Maryland,21224
CurrentmethodsfordeterminingantimicrobialsusceptibilityofGiardialambliarelyontediousmicroscopicobservationoftrophozoitemotilityortimeconsumingassaysofclonal
growth.WeaxenicallyculturedGiardiatrophozoitesinthepresenceofvariousconcentrationsofmetronidazole(.1610g/mL)andfurazolidone(0.82.5g/mL).Viabilitywas
assessedat24hourswith1)fluoresceindiacetate(FDA),ametabolismdependentdye,2)microscopicobservationofmotilityand3)meanforwardlightscatter(FLS),measuredby
passingthecellsthroughafluorescenceactivatedcellsorter.Cellinjury,indicatedbyasignificantincreaseinFLS,occurredat0.64g/mLofmetronidazole(MTZ)and0.08
g/mLoffurazolidone(FZD).Lossoffluorescence(0.64g/mLMTZ)andlossofmotility(1.25g/mLMTZand0.32g/mLFZD)occurredatthesameorhigherdrug
concentrations.IncreaseinFLScorrelatedwithdecreaseinFDAfluorescence,amarkerofcellviability(r=0.94,p<.001).Wheninjurywasassessedaftervariousdurationsof
drugexposure,FLSincreased8hoursafterexposureto1.25g/mLMTZand12hoursafterexposureto0.64g/mLMTZ.Ateffectivedoses(>0.64mcg/mL),increaseinFLSwas
correlatedwithincreasingtimeofexposure,upto24hours(r>0.94foralldrugconcentrations).Analysisoflightscatterprovidesarapid,sensitive,andaccuratemethodof
detectingearlyinjuryinG.lamblia.

Introduction
CurrentmethodsfordeterminingtheantimicrobialsusceptibilityofGiardialambliarelyonmicroscopicobservationoftrophozoitemotility(6),orassayofclonal
growth(2).Theformermethodistedious,andhighlyobserverdependent.Thelattermethodistimeconsuming,requiringupto3daysforfinalcultureresults.
Flowcytometryprovidesameansofexaminingstructuralandfunctionalpropertiesoflargenumbersofindividualcellsrapidly,withhighprecision.Applicationsof
cytometryinmicrobialstudieshaveincludedrapididentificationoforganismsinenvironmentalandhumansamples(9,12,15),determinationofcellcycleparameters
(15),growthcharacteristics(18)andantibioticsensitivitiesofbacteria(5,8,19).Theprimarymethodusedhasbeenquantitativefluorescence.Thenonfluorescent
parameteroflightscatterhasbeenrelativelyneglected(3,4,10,11).Thisisthefirststudythatuseslightscattertodetectcellinjuryinviable,nonfixedorganisms.
Ourhypothesisisthatearlycellinjurywillbereflectedbyalterationsinsurfacestructure,suchasswelling,whichwillincreasetheamountoflightscatter(3).We
examinedwhetherantimicrobialswouldaffecttheamountoflightscatteringbyGiardiatoadetectableandreproducibledegree.First,wecomparedtheparameterof
lightscattertotwoindependentmeasuresofcellviability,microscopicobservationofmotility(6),andfluorescenceofamethabolismdependentdye,fluorescein
diacetate(FDA)(13).Organismswereexposedtovaryingconcentrationsoftwodifferentantimicrobials,andhadviabilityassessedbyallthreemethods.Ourcriteria
toacceptlightscatterasamarkerforearlycellinjurywasdemonstrationofbothathresholdandadoseresponserelationship.Next,wecomparedchangesinlight
scattertoincreasingdosesofantimicrobialby1)increasingantimicrobialconcentrationand2)increasingthedurationofexposure.Wereportthatflowcytometric
analysisoflightscatterprovidesarapid,sensitive,andaccuratemethodofdeterminingcellinjuryinG.lamblia.
MaterialsandMethods
ClinicalSpecimens
PreviouslydescribedG.lambliastrainsWB,LT,andRSwereobtainedfromO.Keister(LaboratoryofParasiticDiseases,NationalInstituteofAllergyand
InfectiousDiseases)(16).
Medium
OrganismswereaxenicallyculturedinfiltersterilizedTY1S33mediummodifiedbytheadditionofbilesalts,asdescribedbyKeister(7).
AntimicrobialAgents
Metronidazole(Searle)andfurazolidone(NorwichEaton)dissolvedinTYIS33wereusedatthespecifiedconcentrations.
FluorescingAgent
Fluoresceindiacetate(FDA)stockwaspreparedbydissolving5mgFDA(USBiochemicalCorporation)in1mLofacetone.Theworkingsolutionwasprepared
immediatelypriortoexperimentationbyadding10mLPBSto0.05mLofstocksolution.Theworkingsolutionwaskeptonice,inafoilwrappedcontainerto
preservestability.Intheindicatedexperiments,FDAwasaddedprecisely3minutespriortoanalysis,astimewasacriticalfactorinquantitatingfluorescence(13).
FDAmustbemetabolizedbythecellsinordertofluoresce.Therefore,theintensityofthefluorescenceisproportionaltotheuptakeofthedyeandthemetabolic
activityofthecell(13).
*Correspondingauthor.
Presentaddress:DepartmentofMedicine,UniversityofMaryland,22GreeneSt.,Baltimore,MD21201,U.S.A..

Page10
TABLE1.Forwardlightscatter
24hourmetronidazoleexposure
dose(g/mL)
Organism

0.16

0.32

0.64

1.25

2.5

G.lamblia
WB

Meanscatter50

11911.9

1257.4

1357.2

1575a

Fluorescence
(%control)

100

952

8410

623a

1829a
a
499

Motility(%)

833

804

736

584

237

0.32

0.64

1809.2
2416
166

24hourfurazolidoneexposure
dose(g/mL)
Organism

G.lamblia

0.08

0.16

2.5

Meanscatter

752.5

9514.5b

12510

13114.5

12611

9712

Motility(%)

844

883

828

643

367

142

ap<.01
b p<.05

Twotailedttest

SusceptibilityTests
A0.1mLinoculumof105/mLlogphaseorganismspertube,with15mLofmediumatthespecifieddrugconcentration,wasused.Thetubeswereincubatedat37
C,for24hours.Inthetimeseriesstudy,however,thespecifiedtimesarethehourspriortoanalysisthatthetubeswereinoculated.Thecellswereharvestedby
coolingtheculturetubesat4Cfor20minutestoreleaseadherenttrophozoitesfromtheglass.Immediatelypriortoanalysis,cellswerefilteredthrougha40mwire
meshtoremoveclumps.
Microscopy
Inselectedexperiments,eachtubewasinspectedatamagnificationof450Xatthetimeofcytometry,toassesscellmotilityqualitatively.Atleastfourfieldsof50cells
wereobservedpertube.
WequantitativelyassessedviabilityforG.lambliaat24hoursbycalculating''percentmotility".Inthismethod,twovolumesofcellsculturedunderthespecified
conditions,weremixedwithonevolumeofa0.1%solutionoftrypanblueandtheresultingsolutionwasloadedinahemacytometer.Onehundredcellswere
examinedat450Xmagnification.Thecellsthatmaintainedmotilityoftheirventralflagellaandexcludedtrypanbluewerereportedas"percentmotile"(Wahl,etal.,
ConferenceProceedings).
FlowCytometry
Analysisofthepreparedcellswasperformedwithafluorescenceactivatedcellsorter(BectonDicksonFACSIV).Inthisinstrument,theorganisms,carriedbya
flowofPBS,passsinglythroughaviewingchamberwheretheyareilluminatedbyanargonlaserbeam.ThelaserlightbothexcitesthemetabolizedFDAtofluoresce,
andisscatteredbythecell.
TheFDAwasexcitedwith300mwofthe488nmlinefromaSpectraPhysics16445argonlaser.Thefluorescentpulsesemittedfromtheorganismswere
transformedintoelectricalimpulses,andstoredbyamultichannelpulseheightanalyzer(MCA).Fluorescenceintensityisexpressedasthepercentageoffluorescence
fromexposedcellstofluorescencefromcellswithoutexposuretotheindicatedantimicrobial.
TheamountofexcitationlightscatteredbyeachcellisdetectedthroughasecondfilterandstoredinthesameMCA.Weexaminedforwardlightscatter,thatis,light
collectedfrom8aboveandbelowthehorizontalplane(3).Lightscatterresultsareexpressedasmeanscatter,whichareobtainedbyintegratingcellsoverchannels
(rangingfromthelowestvalueinchannel1tothehighestvalueinchannel255)andtakingthemean.Cellsoverrangearestoredinchannel255,resultinginaslight
underestimateofthetruepopulationmean.
StatisticalAnalysis
Populationmeanswerecomparedwithatstatistic,acceptingasignificancelevelof0.05.Thesignificancelevelwascorrectedformultiplecomparisonswhen
appropriate.
Results
G.lambliacellsexposedtometronidazoleandharvestedat24hoursshowedasignificantincreaseinmeanforwardlightscatteratametronidazoleconcentrationof
0.64g/mL.Atthisconcentration,thereweresignificant

Figure1.
MeanforwardlightscatterandmeanfluorescenceofG.lamblia
trophozoitesexposedtovaryingconcentrationsofmetronidazole.
Meanfluorescenceformetronidazoleexposedtrophozoites
isfluorescence(trophozoites)/fluorescence(control).Meanfluorescence
forcontrolisthefluorescenceofunexposedtrophozoites
oftheithrun/lowestfluorescenceofunexposedtrophozoites(N=9).

Page11

decreasesinfluorescence(623%)andmotility(584%),comparedtocontrol(100%and833%,respectively)(Table1).
Thereisacontinuedincreaseinmeanlightscatterwithincreasingdrugconcentrationtoamaximumof2.5g/mL,afterwhichmeanscatterdeclines.Withincreasing
drugconcentration,upto2.5g/mLofmetronidazole,thereisacontinueddeclineinfluorescence(to24%)andmotility(to16%)parallelingincreasingforwardlight
scatter.Overthedoserangezeroto2.5g/mLofmetronidazole,fluorescenceandmeanlightscatterweresignificantly,negativelycorrelated(r=0.94,p<.001Figure
1).
AsimilarcorrelationbetweenmeanlightscatterandmotilitywasdemonstratedforG.lambliaexposedtofurazolidone.Whilethemaximaldegreeoflightscatterand
initiallossofmotilityoccurredat0.32g/mLoffurazolidone,theinitialincreaseinforwardlightscatteroccurredataconcentrationof0.08g/mL(Table1).
TheinitialincreaseinmeanforwardlightscatterisdepictedinserialhistogramsofG.lambliaexposedtoincreasingconcentrationsoffurazolidone(Figure2).These
histogramsarederivedbyplottingthenumberof

Figure2.
HistogramsofG.lambliatrophozoitesexposedtoincreasing
concentrationsoffurazolidone.Theverticalaxisrepresentsthe
numberofcellcounts,whilethehorizontalaxisrepresentsthe
channelnumberfrom1(atorigin)to255(ontheright).Cells
withgreaterforwardlightscatterareplacedinthehighernumbered
channels.Thetopcurveshowsthescatterhistogramofthe
controlculture(a).Subsequentcurvesshowculturesexposed
to0.1g/mL(b),1g/mL(c),and10g/mL(d)offurazolidone,
respectively.Increasedcountsintheverylowscatter
channelsreflectcellulardebris(c&d).

Figure3.
MeanforwardlightscatterofG.lambliatrophozoitesafter
incubationwithdifferentconcentrationsofmetronidazole,
forvarioustimeperiods.Differentsymbolsareusedtoindicatethe
differenceconcentrationsofmetronidazole.Time(ofexposure)is
thenumberofhourspriortoanalysisthattheculturewasinoculated.

cells(verticalaxis)againsttheirchannelnumber(horizontalaxis).Withincreasingconcentrationsofantimicrobial,thereisaninitialshiftinthehistogramtohigher
channels,signifyingincreasingforwardlightscatter.Athigherconcentrations(1g/mL),thecurvebecomesbimodal,reflectinglowscatterdebrisandswollen,intact
cells.Themeanscatterdeclines,duetothedebrisinlowscatterchannels.Atveryhighconcentrations(>10g/mL)onlylowscatterdebrisispresent.
Visualinspectionofcellscorroboratestherelationshipbetweenlightscatterandantimicrobialeffect.At1.25g/mL,themetronidazoleconcentrationdemonstrating
themaximaleffectonlightscatter,237%ofthetestcellsweremotile.Atthisconcentration,thereasmarkedclumpingandnotableswelling.At10g/mL,onlydebris
andrareghostswereseen.
AntimicrobialsusceptibilityofG.lambliaisdependentuponbothconcentrationofdrugandtimeofdrugexposure(6).Weassessedinjuryaftervariousdurationsof
exposuretoconcentrationsofmetronidazolerangingfrom0.16g/mLto2.25g/mL(Figure3).At8hours,thereisasignificantincreaseinmeanlightscatterata
concentrationof1.25g/mLorgreater.Meanlightscatterissignificantlyincreasedafter12hoursofexposureto0.64g/mLofmetronidazole.Forthose
concentrationsofmetronidazolehavingasignificanteffectonfluorescenceandmotilityat24hours(>0.64g/mL),meanforwardlightscatterissignificantlycorrelated
withincreasingtimeofexposure.Ataconcentrationof2.25g/mL,thecorrelationoflightscatterandtimeofexposureis0.975.Inthecontrolcells,ata
concentrationof0.16g/mL,thecorrelationis0.002.

Page12

Discussion
Thecloseassociationofchangesinforwardlightscatterwithothermarkersofinjuryandviabilitysuggestthatforwardlightscattercanbeusedtodetectcellinjuryin
G.lamblia.Comparedtotheotherwidelyusedmethodsofmotilityandclonalgrowth,itisrapid,sensitive,andstandardizedbetweenobservers.
Forwardlightscattercorrelatedwellwithtwostandardmeasuresofviability:metabolismdependentfluorescenceandmotility.Theeffectiveconcentrationsof
metronidazoledefinedbyforwardlightscatterarealsoinaccordwithpreviouswork.JokipiiandJokipii(6)reported1060%motilityinG.lambliaafter24hoursof
exposuretoametronidazoleconcentrationof0.08mcg/mLGillinandDiamond(2),usingaclonalgrowthassay,reportedaminimumlethalconcentrationof2g/mL
after24hours.Inbothofthosestudies,qualitativechangesinmotility,reflectingalteredcellfunction,occurredpriortothecompletecessationofmotility(2,6).
Forwardlightscatterprovidesasimplemethodtoquantifytheseeffectsofantimicrobials,earlyintheprocessofcellinjury.
Theconcentrationsofantimicrobialsatwhichasignificantchangeinlightscatterwasdetectedafter24hoursofexposurecorrespondtotheminimumlethal
concentrationafter72hoursofexposure(2).Thissuggeststhattheminimumconcentrationtoincreasethemeanforwardlightscattermightperformwellasameasure
ofantimicrobialsusceptibility.Totestthiswouldrequirecomparingforwardlightscatterwiththegoldstandardofclonalgrowth.
PreviousstudieshaveexaminedchangedinDNAasmeasuredbyfluorescentdyes(ethidiumbromide,mithramycin,propidiumiodide)toevaluatesusceptibilityof
bacteriaandfungi(5,8,19).Noneofthesestudieshaveexaminedprotozoa.However,fluorescentdyesareexpensive,requirestrictcontroloflightingconditions,
frequentlyrequirefixation,andmayrequireprecisetimingofthedyeexposure.
Otherinvestigatorshavenotedchangesinlightscatter,thoughonlywithcellsofhigherorganisms,orwithfixed,nonlivingmicrobes.Nash,etal.,Usedlightscatterto
separateviable,rodshapedratventricularmusclecellsfromdamaged,roundcells(10).HutternandEipel,inreportingondoublefluorescentstainingtoassess
viabilityofethanolfixedyeast(Saccharomycescerevisiae)presenteddatademonstratingastrongcorrelationbetweenfluorescencedefinedviabilityandlight
scatterpatternwithoutcomment(5).However,ethanolfixationhasbeenshowntoalterthelightscatterpattern(1).
Underappropriateconditions,forwardlightscatterperformswellasasingleparametertodetectcellinjury.Itisasimple,sensitive,andrapidmethodtodeterminethe
effectofantimicrobialsandotheragentsonprotozoa.
Forwardlightscattercanbeusedasamarkerforcellinjuryinnonfixedcellsinconjunctionwithfluorescentprobesofcellularfunction.Useoflightscatteranda
singlefluorescentprobeislesscomplexthandoublefluorescence.Further,unlikemostfluorescentmarkersofviability,fixationisnotrequired,permittingexamination
ofcellularphysiology.Forexample,calciumfluxinearlycellinjuryandsubsequentdeathcouldbeevaluatedwithQuinII(14).Asquantitativefluorescenceisperhaps
themostpowertechniqueindualparametercytometry,theuseofforwardlightscatterasaviabilityparameterwouldfacilitatemoreelegant,invivoevaluationof
Giardiaphysiology.
LiteratureCited
1.Braylan,R.C.,N.A.Benson,U.Nourse,andH.S.Kroth.1982.CorrelatedanalysisofcellularDNA,membraneantigens,andlightscatterofhumanlymphoid
cells.Cytometry2:337343.
2.Gillin,F.,andL.Diamond.1981.InhibitionofclonalgrowthofGiardialambliaandEntamoebahistolyticabymetronidazole,quinacrine,andotherantimicrobial
agents.Antimicrob.Chemother.8:k305316.
3.Hansen,W.P.,R.A.Hoffman,S.H.Ip,andK.W.Healey.1982.Lightscatterasanadjuncttocellularimmunofluorescenceinflowcytometricsystems.J.Clin.
Immunol.2(Suppl.):32s41s.
4.Hoffman,R.A.,P.C.Kung,W.P.Hansen,andG.Goldstein.1980.SimpleandrapidmeasurementofhumanTlymphocytesandtheirsubclassesinperipheral
blood.Proc.Natl.Acad.Sci.(USA)77:49144917.
5.Huttern,J.J.,andH.E.Eipel.1978.Advancesindeterminationofcellviability.J.Gen.Microbiol.107:165167.
6.Jokipii,K.,andA.M.M.Jokipii.1980.InvitrosusceptibilityofGiardialambliatrophozoitestometronidazoleandtinidazole.J.Infect.Dis.1412:317325.
7.Keister,D.1983.AxeniccultureofGiardialambliainTY1S33mediumsupplementedwithbile.Trans.Roy.Soc.Trop.Med.Hyg.77:487488.
8.Martinez,O.,H.G.Gratzner,T.I.O.Malinin,andM.Ingram.1982.TheeffectofsomebetalactamantibioticsonEscherichiacolistudiedbyflowcytometry.
Cytometry3:129133.
9.Muldrow,L.L.,R.L.Tyndall,andC.B.Fliermans.1982.Aapplicationofflowcytometrytostudiesofpathogenicfreelivingamoebae.Appl.Environ.Microbiol.
44:12581269.
10.Nash,G.,P.E.R.Tatham,T.Powell,V.W.Twist,R.D.Speller,andL.T.Loverock.1979.SizemeasurementsonisolatedratheartcellsusingCoulteranalysis
andlightscatterflowcytometry.Biochimica.etBiophysica.Acta587:99111.
11.Salzmann,G.C.,P.F.Mullaney,andB.J.Price.1979.Lightscatteringapproachestocellcharacterization,pp.105124.In:Melamed,M.R.,R.F.Mullaney,and
M.L.Medelsohn(eds.),FlowCytometryandSorting,JohnWileyandSons,NewYork.
12.Saul,A.,P.Myler,T.Mangan,andC.Kidson.1982.Plasmodiumfalciparum:automatedassayoferythrocyteinvasionusingflowcytofluoromety.Exp.Parasitol.
54:6471.
13.Serentz,M.1973.Fluoresceindiacetate.In:ThaerA.A.andM.Serentz(eds.),FluorescentTechniquesinCellBiology,SpringerVerlag,Heidelberg.243.
14.Shev,S.S.,V.K.Sharma,andS.P.Banergee.1984.MeasurementofcytosolicfreecalciumconcentrationinisolatedratventricularmyocyteswithQuinnII.
CirculationRes.55:830834.
15.Skarstad,K.,H.B.Steen,andE.Boyle.1983.CellcycleparametersofslowlygrowingEscherichiacoliB/rstudiedbyflowcytometry.J.Bacteriol.154:656
662.
16.Smith,P.,F.Gillin,N.Kawhal,andT.Nash.1982.AntigenicanalysisofGiardialambliafromAfghanistan,PuertoRico,Ecuador,andOregon.Infect.Immun.
36:714719.

Page13

17.Steen,H.B.1980.Furtherdevelopmentsofamicroscopebasedflowcytometer:lightscatterdetectionandexcitationintensitycompensation.Cytometry1:2631.
18.Steen,H.B.,andE.Boyle.1980.Escherichiacoligrowthstudiedbydualparameterflowcytometry.J.Bacteriol.145:10911094.
19.Steen,H.B.,E.Boyle,K.Skarstad,B.Bloom,T.Godal,andS.Mustafa.1982.Applicationsofflowcytometryonbacteria:cellcyclekinetics,drugeffects,and
antibodybinding.Cytometry2:249257.
20.vanDilla,M.A.,R.G.Langlois,D.Pinkel,D.Yajko,andW.K.Hadley.1983.Bacterialcharacterizationbyflowcytometry.Science220:620622.

Page15

TheImportanceofNonwaterborneModesofTransmissionforGiardiasisACaseStudy
V.SusanHarley
SturgeonHealthUnit,Box174,23SirWinstonChurchillAvenue,
St.Albert,AlbertaT8N1N3,Canada
Itisacommonassumptionthatoutbreaksofgiardiasisaremostfrequentlyassociatedwiththewaterbornemodeoftransmission.Recentstudiesonoutbreaksofgiardiasishave
shownthatothermodesoftransmissionareequallyasimportant.Fecaloraltransmissionwhetherfoodborneorpersontopersonisasfrequentlyimplicatedasthewaterroute.
ThiswasshownduringarecentinvestigationofagiardiasisoutbreakatanAlbertaHutteritecolony.Theinfectionrateatthecolonywas37%ofthepopulation(76members).The
fieldinvestigationresultedinwaterbornetransmissionbeingeliminatedasthemostprobableroute.Fecaloraltransmissionviapersontopersoncontactappearedtobethemost
probablemodeoftransmission.Inaddition,deficienciesinfoodserviceandlaundryfacilitiesatthecolonymayhavealsocontributedtotheoutbreak.Controlproceduresincluded
antiprotozoaltreatmentofinfectedindividuals,aswellaseducationinfoodhandlingpracticesandentericdiseasetransmission.Withinathreemonthperioda90%recoveryrate
wasachieved.

Introduction
ThefieldinvestigationofanoutbreakofgiardiasisonanAlbertaHutteritecolonyrevealedthatwhiletheinitialsourceoftheinfectioncouldnotbeisolated,several
modesoftransmissionwerefoundwhichfacilitatedthespreadoftheinfection.Thenatureofthewatersupplyprecludedwaterbornetransmission.Thefecaloralroute
viapersontopersonandfoodbornetransmissionisbelievedtohavebeenmostimportant.
Thepurposeofthispaperistoreviewthemodesoftransmissionofgiardiasisandtoexaminetherelativeimportanceofnonwaterbornemodeswithinthecontextof
thisfieldinvestigation.Inparticular,parallelsbetweenthecommunallifestyleofaHutteritecolonyandinstitutionalsettingswithinourNorthAmericansocietyare
examinedtohighlighttheneedforadditionalresearchintopreventionandcontrolofthisincreasinglyimportantentericinfection.
ModesofTransmission
Benenson(1)describesthreemodesoftransmissionforgiardiasis,namelywaterborne,foodborneandpersontoperson.
1.WaterborneTransmissionoccurswhenwatercontaminatedbyGiardiacystsisconsumed.Tworiskfactorsassociatedwithwaterbornetransmissionarethe
sourceofthewaterandthemethodoftreatmentused.Illnesseshaveoccurredwhereunfilteredwaterfromsurfacesources(3,10)hasbeenconsumed.InBritish
Columbia,sixtyninecasesofgiardiasisweretracedtoconsumptionofamunicipalwatersupplywhichderiveditsrawwaterfromacreek(3).Thewatersupplywas
unfiltered.InAlberta,twolocalizedoutbreakshavebeenlinkedtomunicipalwatersupplies.InBanff,thelinkbetweenthewaterreservoirandthesurfacewater
sourceisbelievedtohavebeenresponsiblefortheoutbreakin1982(EpidemiologicNotesandReports,AlbertaSocialServicesandCommunityHealth,1983).In
Edmonton,deficienciesinthemunicipalwatertreatmentfacilitiesarebelievedtobetheprobablecauseofanoutbreakin1983(Logsdon,G.S.Evaluationof
MonitoringandTreatmentforProtozoanPathogens,EdmontonWaterSupplyEvaluation,Edmonton,1986).
2.FoodborneTransmissionoccurswhenfecallycontaminatedfoodisconsumed.WhileGiardiacystsdonotreplicateinfood,theymaysurviveonmoistfoods
whicharenotcookedpriortoeating.Osterholm(5)tracestwentyninecasesofgiardiasisresultingfromtheingestionofcannedsalmon.Hehypothesizesthat
conditionswithinthefood'senvironmentmayenhancetheinfectivityofGiardiacysts.
3.PersontoPersonTransmissionoccurswhenthecystsfromthefecesofaninfectedindividualaredirectlyorindirectlytransferredtoanotherperson.Sartwell
andLast(7)statethatthethreemostobviousriskfactorsinpersontopersontransmissionarethepopulationdensity,hygienicconditionsandtheproportionof
susceptibleindividualsinthepopulation.
Theseriskfactorsinteracttofacilitatetherapidspreadofinfectionthroughoutinstitutionalandhouseholdsettings.Inchildcareandresidentialinstitutions,the
frequencyofinterpersonalcontactwithinasusceptiblepopulationcoupledwithpoorenvironmentalandhygienicconditionscontributetothespreadofdisease(9).
Keystone's(4)reportonpersontopersontransmissionofgiardiasisindaycarenurserieshighlightsthesusceptiblepopulationasyoungchildrenwhoaremobilebut
notyettoilettrainedoreducatedinpersonalhygiene.InAlbertain1984onethirdofthecasesofgiardiasiswereunderfiveyearsofagewith31%betweenoneand
fouryears(EpidemiologicNotesandReports,AlbertaSocialServicesandCommunityHealth,1985:271273).Householdcontactsarea

Page16

highlysusceptibleportionofthepopulationforpersontopersontransmission.Polis(6)suggestsinhisreportofthespreadofgiardiasisfromadaycaretothe
communitythat"47%ofthechildreninthestudygrouptransmittedthediseasetoatleastonehouseholdcontact".InAlbertain1984,almost30%ofgiardiasiscases
hadatleastoneothercasewithinthefamilyandhouseholdcontacts.
CaseStudy
Introduction
TheHutteritecolonyislocatedinaruralfarmingareanorthwestofEdmonton.Attheperiodoftheinvestigation(JunetoSeptember,1985)therewerefifteenfamilies
livingonthecolonywithatotalpopulationofseventysixpeopleranginginagesfromfourmonthstoeightytwoyears.Thefacilitiesofthecolonywillbediscussedas
theyexistedatthetimeoftheinvestigation.Changesmadesincethattimewillbediscussedinthepostscript.Whileeachfamilymaintainsseparatelivingquarters,many
activitiesareconductedcommunally.Thefoodandlaundryfacilitiesarelocatedinonecentralbuilding.Akindergartenwhichisrunbythemothersofthecolonycares
forchildrenthreeyearstoschoolage.Aschoolsupportedbythelocalschooldistricthasprogramsforchildrenuptoagefifteen.
Thecolonylivelihoodisbaseduponamixedfarmingoperationwithhogs,cattle,poultry,vegetablesandgrain.Salesofpoultryfromanonsitepoultryabbatoirand
vegetablesalesatlocalfarmer'smarketsprovideadditionalincome.
TheearliestdocumentedcaseofgiardiasisatthecolonyoccurredinJune,1983.Asecondcase,thewifeofthefirstcase,wasdiagnosedinOctober,1984.The
secondcasewastheheadcookforthecolony.Sherequiredtwocoursesoftreatmentwithmetronidazolebeforenegativestoolsampleswereobtained.Although
stoolsamplesfromfamilymemberswererequested,thereisnorecordofsamplesbeingsubmittedtotheProvincialLaboratoryofPublicHealth.
OnJune24,1985theSturgeonHealthUnitreceivednotificationofacaseofgiardiasisinatenmontholdmale.Duringthecommunicablediseaseinterviewwiththe
child'smotheronJune25,1985itbecameapparentthatanumberofindividualsonthecolonywereexperiencingsymptomsofabdominalcramps,bloatingand
diarrhea.
Methods
Toassesstheextentoftheoutbreak,itwasdecidedtocollectstoolspecimensfromallcolonymembers,bothasymptomaticandsymptomatic.Stoolkitscontaining
formalinsolutionweredistributedtoeachfamily.Afamilytreesuppliedbythepreacherwasusedtodevelopacaserecordandallmemberswerelistedaccordingto
familygroupandbirthdate.BecauseofsomeduplicationinChristiannamesplustheuseofacommonlastname,everyonewasrequestedtosupplybirthdatesonthe
stoolrequisitionform.Althoughitisnormalproceduretorequestthreestoolsamplesperperson,itwasdecidedbytheProvincialEpidemiologistthatonesampleper
personwouldfacilitateboththecollectiononthecolonyandidentificationbytheProvincialLaboratoryofPublicHealth.Initially,considerableresistancetothe
submissionofsampleswasencountered.Asthescopeoftheoutbreakbecameapparent,mostpeoplecompliedwiththerequest.Oftheseventysixmembers,
seventypeoplesuppliedatleastonesample.
Theonsiteinvestigationfocussedonthewatersupply,sewagedisposalsystem,foodandlaundryserviceandpersontopersontransmissionasthepossiblesourcesof
infectionandmodesoftransmission.
Results
Thecolonywaterissuppliedfromanunfiltered,untreateddeepwelllocatednorthofthebuildingsite.Routinewatersampleshavebeensubmittedfromvarious
locationsthroughoutthecolonysince1984.AlltwelvesamplessubmittedintheseventeenmonthperiodtotheendofAugust,1985hadnilfecalandtotalcoliform
counts.Waterfromdeepwellsisnotconsideredtobeaslikelyasourceofgiardiasisaswaterfromsurfacewatersupplies.Thiswouldsuggestthatthepossibilityof
contaminationofthewaterbyGiardiacystswaslow,however,laboratoryverificationofGiardiacystswasnotfeasibleatthattime.
Everydwellingunitplusthekindergartenandschoolisequippedwithwaterclosetsandlavatories.Thepoultryabbatoirandkitchenhavefullhotandcoldwater
service.Thefinaldisposalofalleffluentisintoalargenaturallagoonlocatedwestofthebuildingsiteandapproximatelyseventyfivefeetbelowtheabbatoir.
Apparently,therehavenotbeenanyproblemswithsewageblockagesorbackupsinthedwellingunits.Therehavebeenrecurringproblemswiththegreywaterfrom
thekitchenbackingupintothekitchensinks.Torectifythisproblem,anewseptictankwasinstalledinearlysummer,1985tohandlekitcheneffluentpriortofinal
dischargeintothelagoon.Theeffluentdisposalfromtheabbatoirwasupgradedduringthewinterof1985.Previously,theeffluentfromtheslaughteringoperation,
includingblood,feathersandvisceradraineddirectlydowntheslopeintothelagoon.Theupgradingincludedtheinstallationofcoveredfloorguttersandasettling
tank.Becauseofthemixedfarmingoperation,thereisconstantcontactwithanimalmanure,especiallybythemen.Therearetwoduckponds,onetothesouthofthe
animalbarnsandonetothenorthatalowerelevationfromthebuildingsite.Additionally,severalsheepgrazeamongthedwellingunits.Thesewagedisposaldoesnot
differgreatlyfrommanysimilarlargemixedfarmingenterprises.Thepossibilityofcontactwitheitherhumanoranimalmanureishigh.
Thefoodandlaundryfacilitiessharetheareaknownasthekitchen.Whileeachdwellingunithasasmallkitchenareawithsinks,waterandasmallhotplate,most
foodserviceisconductedcommunallyinthekitchen.Thefoodispreparedandservedbythewomenofthecolonyaccordingtoaweeklydutyroster.Threedaily
mealsareserved"familystyle"intheadjoiningroom.Theyoungchildrenareservedintheirhomespriortotheadultmealtime.Thekitchenisprimarilythefood
preparationareawithalargecommercialgrillandpressurecooker,aswellaspreparationtables.Thefoodstorageareaisinaseparatebuildinglocatednorthofthe
kitchen.Thisbuildingcontainsalargefreezerroomwheremostproductsare

Page17

stored.Milkanddairyproductsarekeptinalargecommercialrefrigeratorinthedininghall.
Duringtheinvestigationofthefoodservice,deficiencieswerenotedincoldfoodstorage,dishwashingmethodandhygiene.Perishableproductswereoftenstoredat
roomtemperature.Dishwashingwasbeingperformedmanuallyintwosinkslocatedbacktoback.Nosanitizerswereusedintherinsewater.Therewerenoseparate
handwashingfacilitiesavailableeitherforfoodhandlers,orthecolonymemberswhoarrivedformealsdirectlyfromotherareasofthecolony.
Thesecondcaseofgiardiasisaspreviouslystatedwastheheadcook.Atthetimeofthatinvestigation,itwassuggestedthatsherefrainfromfoodhandling.Shewas
notabletoconformwiththissuggestion.Duringtheoutbreakinvestigation,itwassuggestedagainthatallsymptomaticorpositiveasymptomaticpersonsbeexcluded
fromfoodhandling.Thereisaculturalbarrierwhichdoesnotpermitcompliancewiththissuggestion.
Thelaundryfacilitiesarelocatedatthewestendofthekitchen.Therearetwolargecommercialfrontloadingwashingmachinesandawaterextractor.Dryingis
accomplishedbyoutdoorclotheslines.Duringobservationsmadeduringseverallaundrysessions,twoproblemsbecameobvious.Thewashingmachinesleakedor
wereinterruptedbyopeningthedoorduringoperation.Thiscausedwashwatertorunalloverthekitchenfloor.Thesecondpotentiallymoreseriousproblemwasthe
useofthelargedishwashingsinksaslaundrysinks.Clothingwasrinsedinthesinkpriortobeingplacedintheextractor.
Thesuggestionwasmadetoceaseusageofthedishwashingsinkforlaundryrinsinguntilimprovedarrangementscouldbeinstalled.Thepotentialfortransmissionof
thediseasethroughcontaminationofthefoodserviceareafromsoiledwaterandclothingwouldappeartobeveryhigh.Severalrecommendationsweredirectedat
theseproblems.
Interpersonalcontactisveryfrequentamongthemembersofthecolonyandwithpeoplefromoutsideofthecolony.Withinthecolony,thecommunalfoodservice,
laundryfacilities,kindergartenandgeneralfarmingoperationsprovidefrequentdailycontactwithallmembers.Themenworkandeattogether,whilethewomenand
childrenspendaconsiderableportionoftheirdayinshareddutiesorcommunalcare.Thecolonymembersfrequentlyvisittheirrelativesandfriendsinothercolonies
throughoutAlbertaandMontana.Socialeventssuchasweddingsareoccasionsforlongvisits.DuringJune,1985oneweddingwasheldonthecolonyandthree
otherswereattendedbymanycolonymembersoncoloniesthroughoutAlberta.Oftenonlytheadultsattendtheseweddingswhiletheirchildrenarecaredforby
anotherfamily.Duringconversationwithvariousindividuals,theymentionedthatsimilarsymptomswerebeingexperiencedonothercoloniesthattheyfrequentlyvisit,
orwhosemembersfrequentlyvisit.TherewerereportsofgiardiasisonotherAlbertaHutteritecoloniesduringthistime.Thepossibilityofthediseasebeingintroduced
intothecolonybyvisitorsorbeingspreadtoothercoloniesbymembersofthiscolonyappearhigh.
TABLE1.Casedistributionbyageandsex
Age

Male

Female

011months

Total
3

14years

10

59years

1014years

1519years

2024years

2529years

3039years

4059years

60+years

TOTAL

19

28

Ofthestoolspecimenssubmittedbyseventycolonymembers,twentyeightwerepositiveforGiardialamblia.ThecasedistributionisshowninTable1.Sixtyeight
percentofcasesoccurredintheunderfifteenagecategorywith37%intheunderfiveagegroup.Malesaccountedfor68%ofcasesintheunderfifteenagegroup.
Thisisunderstandableinlightoftheageandsexdistributionofthecolonypopulation.Intheunderfifteenagecategory,malesaccountfor70%percentofthegroup.
Overall,43%ofthepopulationisunderfifteenyearswithonly10%overfortyfiveyears.Theagefifteenwaschosenbecausethatistheageatwhichchildrenleave
schoolandbegintoparticipateinthecolonyduties.Theclusteringofcasesintheloweragegroupscorrespondstopreviouslyreportedinfectionpatterns.The
percentageofsymptomaticcaseswas61%overallwith39%reportingnosignificantsymptoms.Becauseofthedifficultiesinterviewingthemen,whoarenormally
workingawayfromthemainarea,aswellasthemildnessofsymptomsexperiencedbymanypeople,theinformationrelatingtosymptomsandtherateof
symptomatic/asymptomaticcasesisunreliable.Ofthepeoplereportingsymptoms,themostcommonwerebloating,diarrhea,abdominalcrampsandfatigue.Themost
severesymptomswereexperiencedbythechildren.
Thecontrolproceduresincludedtreatmentofallcaseswithmetronidazoleandeducationinseveralfacetsofcommunicablediseasecontrol.Nosideeffectstothe
medicationwerereportedalthoughseveralchildrenencounteredproblemsswallowingthepills.Noneofthefemalecaseswerepregnant,thereforetreatmentproblems
intheadultpopulationwerenotencountered.
Stoolsampleswererequestedafterthecompletionofthemedication.Threecasesdidnotrespondtotheinitialtreatmentandwereprescribedasecondcourseof
medication.Twoofthesethreeinvolvedchildrenwhohaddifficultyinswallowingthepillsandmaynothavereceivedthefullamountrequired.
Themedicationwasdistributedbythecolonyrepresentativewhenshereceivedpositiveresultsfromthepublichealthinspector.Therewasatimelagofapproximately
10daysbetweenthesubmissionofstoolsamplesandreceiptofresults.

Page18

Discussion
TheoutbreakofgiardiasisattheAlbertaHutteritecolonywasidentifiedinlateJune,1985.Theinvestigation,controlandtreatmentprocedureswereconductedover
thesummermonthswithalltwentyeightcasestreatedbymidSeptember,1985.
Themostprobablesourceoftheoutbreakwasnotidentified.Theextendedtimeperiodwhichhadelapsedsincethepreviouscasesonthecolonycoupledwiththe
factthatnooneinthatfamilywasidentifiedasacaseorcarrierinthisinvestigation,wouldsuggestthattheywerenotinvolvedasasourceofinfection.
Themostprobablemodeoftransmissionwouldappeartobepersontopersoncontactbetweenbothcolonymembersandvisitorsfromothercolonies.The
deficienciesnotedinthefoodserviceareamayalsohavebeenimplicatedinthediseasetransmission.
Theattackrateof37%issignificantlyhigherthanwouldbeexpected.Becausemostpersonsinfectedwereidentifiedasapartoftheinvestigation,theinformation
relatingtotheonsetandintensityofsymptomswasnotreportedreliably.Thesymptomsreportedweresimilartothosenormallyexperienced.
Theattackrateandseverityofsymptomswerehigherinthebelowfifteenagecategory.Thisfindingisconsistentwiththeresultsofotheroutbreakswherechildren
caredforinadaycareorextendedfamilyenvironmentappeartobeatriskinacquiringtheinfection.
Theoutbreakrequiredapproximatelythreemonthstobringundercontrolwithallpositivecasesbeingidentifiedandtreatedwithinthattimeframe.Concernwas
expressedthatthecolonymembers'selfadministrationofthemedicationmayhaveprolongedtheperiodofcommunicability.Theresultsofthefollowupstool
samplesshoweda90%recoveryratewithtwoofthethreepositivecaseshavingencounteredtreatmentdifficulties.
Becauseoftheriskofareoccurrenceofcommunicablediseaseatthecolony,recommendationsweremadeinthoseareaswheredeficiencieswerenotedandwhereit
wasfeltthatimprovementswouldbemosteffective.Itwasrecommendedthatthelaundryfacilitiesberelocatedawayfromthekitchenandthatthemachinesbe
repairedintheinterim.Foodhandlingrecommendationsincludedtheuseofasanitizerintherinsewaterandtheadoptionofapolicytoexcuseillindividualsfromfood
handlingduties.Itwasalsorecommendedthataseparatehandwashareabeestablishedinthekitchenfortheuseofbothfoodhandlersandcolonymembers.An
educationalseminaronthepreventionandcontrolofcommunicablediseasewasproposedforallmembersofthecolony.
Atthehealthunitlevel,therecommendationwasmadetodevelopaneducationalprogramsuitableforadultsonthepreventionandcontrolofcommunicabledisease.
ContinuedsurveillanceoftheprevalenceofgiardiasisonAlbertaHutteritecolonieswasalsorecommended.
Postscript
Oneyearlater,itisencouragingtoreportthatseveralrecommendationshavebeenimplemented.Anewlaundryfacilityhasbeenconstructedwellawayfromthe
kitchen.Atthiswriting,onefurthercaseofgiardiasishasbeenreportedfromthecolonynofamilycontactswerepositiveduringscreening.
Conclusions
CantheexperiencegainedfromtheinvestigationofanoutbreakofanentericinfectionintheuniqueenvironmentofanAlbertaHutteritecolonybegeneralizedto
mainstreamNorthAmericansociety?Ifweconsideracolonytobealargeextendedfamily,wecanappreciatetheincreasedriskofdiseasetoitsmembers.Thereare
numerousopportunitiesfordiseasetransmissionthroughthecloseinterpersonalcontactandcommunalfood,laundryandchildcareservice.Inthiscontext,weshould
consideragaintheriskfactorspreviouslydiscussedforpersontopersontransmission.Thefrequencyofinterpersonalcontact,theproportionofsusceptiblepeoplein
apopulationwhere24%percentareundertheageoffiveyears,andthecombinationoflowerlevelsofenvironmentalsanitationandunderstandingofpersonalhygiene
allinteracttocreateahighriskenvironmentforthespreadofentericillnesses.
Theseriskfactorsarepresentinchildcareandresidentialinstitutions.Toappreciatethemagnitudeoftheproblem,recentestimatesbyAlbertaSocialServicessuggest
that10%ofpreschoolchildrenarebeingcaredforinformalchildcarecarefacilities(AlbertaSocialServicesDayCareBranch,PersonalCommunication,Edmonton,
1986).Thispercentagerisessharplywhenthosechildrenbeingcaredforininformalbabysittingarrangementsareincluded.Controllingtheincidenceandtransmission
ofentericaswellasothercommunicablediseasesinthesefacilitieshasemergedasapublichealthpriority.
Inrecentreportsofoutbreaksofentericinfectionsindaycarecentres,severalmethodsofcontrolhavebeenexamined.Black's(2)studyonhandwashingsuggests
thatasupervisedhandwashingprogramcouldreducetheincidenceofentericinfectionsindaycarecentres.Astudyonthecontrolofshigellosisindaycaresstrongly
supportstheisolationofcaseswithinthefacilityasopposedtoexclusionorfacilityclosure(8).Ifafacilityisclosedorachildexcluded,parentsmaybeforcedinto
makingalternatedaycarearrangementsthroughwhichthediseasemaybeintroducedandspreadtothecommunity.
Doestheincreaseinincidenceofcommunicablediseaseinchildcareinstitutionswarrantraisingtheirstandardstothoseofasmallhospitalwithtrainednursingstaff,
isolationroomsandaninfectioncontrolcommittee?Whilesuchmeauresmaybeuneconomical,thetrendtowardspreschoolchildcare,particularlyinfantandtoddler
care,coupledwiththeeaseofdiseasetransmissionfromtheinstitutionalenvironmenttothehouseholdandcommunity,demandspublichealthintervention.The
deliveryofeducationalprogramsdesignedtoteachthenature,preventionandcontrolofcommunicablediseasesaswellascontingencyplansforhandlingsuch
situationsarewithinthefinancialandpersonnelresourcesofexistingpublichealthprograms.Whatisrequiredisacompleteunderstandingofthemechanismsof
disease

Page19

transmission.ThecasestudyoftheAlbertaHutteritecolonysuggestsseveralpossiblemodesoftransmissionofwhichpersontopersontransmissionwasthemost
important.Researchintogiardiasiswhichfocusesontheimportanceofnonwaterbornemodesoftransmissionwouldbeofgreatassistancetofieldepidemiologists
andhealtheducationprogramdesigners.
LiteratureCited
1.Benenson,AbramS.1985.In:ControlofCommunicableDiseasesinMan,14thed.TheAmericanPublicHealthAssociation,Washington,D.C.
2.Black,R.E.,A.C.Dykes,K.E.Anderson,J.G.Wells,S.P.Sinclair,G.W.Gary,M.H.Hatch,andE.J.Gangarosa.1982.Handwashingtopreventdiarrheainday
carecentres.Am.J.Epidemiol.113:445451.
3.Fogel,D.A.1982.Waterbornegiardiasisandtheroleofthehealthinspector:Acasehistory.Environ.HealthReview,2629.
4.Keystone,J.S.,S.Krajden,andM.R.Warrne.1978.PersontopersontransmissionofGiardialambliaindaycarenurseries.Can.Med.J.119:241248.
5.Osterholm,M.T.,J.C.Forfang,T.L.Ristinen,A.G.Dean,J.W.Washburn,J.R.Godes,R.A.Rude,andJ.G.McCullough.1981.Anoutbreakoffoodborne
giardiasis.TheNewEnglandJ.Med.304:2428.
6.Polis,M.A.,C.U.Tuazon,D.W.Alling,andE.Talmanis.1986.TransmissionofGiardialambliafromadaycarecentretothecommunity.Am.J.PublicHealth
76(9):11421144.
7.Sartwell,P.E.,J.M.Last.1980.Epidemiology,p.70.In:LastJ.M.,(ed.),MaxcyRosenauPublicHealthandPreventativeMedicine,11thed.Appleton
CenturyCrofts,NewYork.
8.Tauxe,R.V.,K.E.Johnson,J.C.Boase,S.D.Helgerson,andP.A.Blake1986.Controlofdaycareshigellosis:Atrialofconvalescentdaycareinisolation.Am.J.
PublicHealth76(6):627630.
9.Thacker,S.B.,S.Simpson,T.J.Gordon,M.Wolfe,andA.M.Kimball.1979.Parasiticdiseasecontrolinaresidentalfacilityfortheretarded.Am.J.PublicHealth
69(12):12791281.
10.Weniger,B.G.,M.J.Blaser,J.Gedrose,E.C.Lippy,andD.D.Juranek.1983.Anoutbreakofwaterbornegiardiasisassociatedwithheavywaterrunoffdueto
warmweatherandvolcanicashfall.Am.J.PublicHealth73(8):868871.

Page21

ANewMinicultureTechniquefordeterminingInVitroAntimicrobialAgentSensitivityofAxenicallyCultivatedStrainsof
Giardialamblia
StephenM.Wahl,RobertH.Gilman*,JaneP.O'Hare,DavidB.Keister,andWilliamM.Spira
DivisionofGeographicMedicine,TheJohnsHopkinsUniversitySchoolofMedicine,
Baltimore,Maryland
WedevelopedanassayinwhichtheinvitrosensitivitiesofGiardialambliatoantimicrobialagentscanbedetermined.Weusedaminiculturetechnique,inwhichadherenceof
trophozoitestothecultureplatewasthecriterionforviability.MeasurementofadherencepermitsonetodeterminetheadhesiveactivityofaggregatesofG.lambliaratherthan
confiningdeterminationofviabilitytoindividualtrophozoites.Theadherenceassaywasvalidatedbycomparingitsresultstothoseofthestandardassayforinwhichventral
flagellamotilityofG.lambliaisthecriterionforviability.FivedifferencestrainsofG.lambliatestedhadthesamesensitivitytometronidazole,whileone(theCATstrain)wasslightly
moreresistanttoquinacrinethantheotherfour.Anevaluationofsensitivitytovariousantimicrobialagentsdemonstratedthat80Sribosomalinhibitors(suchascycloheximide
andanisomycin)andfurazolidoneandnalidixicacid(whichlikemetronidazoleandquinacrineaffectDNAsynthesis)areeffectiveinvitroagainstG.lamblia.

Introduction
RecentdevelopmentsintechniquesfortheinvitrocultivationofGiardialamblia(1)havestudiedtheparasite'ssusceptibilitytoantimicrobialagents.Earliermethods
forinvestigatingG.lambliasusceptibilityweretediousortimeconsuming.IntheclonalmethodofGillin(2),susceptibilityismeasuredbyexposingcellstoan
antimicrobialagentfor72hours,incubatingthemforanadditional72hoursinnutrientagar,thencountingthenumberofvisiblecolonies.Inthestandardmotilityassay
byJokipiiandJokipii(3),theeffectsofantimicrobialagentsonG.lambliatrophozoitesaredeterminedbydirectmicroscopicexaminationofindividualtrophozoites:
aftereach24hoursofexposure,100randomlyselectedG.lambliaareexaminedforventralflagellamotility,thecriterionforviability.
Indevelopinganewinvitrominiculturetechnique,ourgoalwastoavoidthetediumoftheJokipiiandJokipiimethodandthelongwaitingperiodoftheGillinmethod
whilekeepingthesensitivityoftheseprocedures.
Inthecurrentstudy,wehavevalidatedamicrocultureadherenceassayforviability(amodifiedversionoftheJokipiiandJokipiimethod),andusedittoexaminethe
susceptibilityofG.lambliatodifferentantimicrobialagentsandtoexaminevariationsbetweendifferentG.lambliastrainstothesameantimicrobialagents.
MaterialsandMethods
GiardialambliaTrophozoites
Fivepreviouslycharacterizedstrains(4,5)ofG.lamblia(WB,PO,LT,RS,andCAT)weremaintainedinaxenicculture.
Media
AllculturesweregrowninamodifiedDiamond'sTYIS33mediumdescribedbyKeister(4).Afterthemedium,includingserum,washeatedinat56Cwaterbath
for30minutes,penicillinandstreptomycinwereadded.Themediumwasthenfiltersterilized.Allmediumwaspreparedonedaypriortousetoavoiderraticresults
duetodifferencesintheoxidationofcysteine.Gillinhasdemonstratedthattheinclusionofpenicillinandstreptomycininthemediumdoesnoteffectantimicrobialagent
susceptibilitytesting(2).
AntimicrobialAgents
TheantimicrobialagentstestedarelistedinTable1.Alldrugswerereconstitutedinsterilemedium.
CultureofG.lambliaTrophozoitesandPreparationofTrophozoitesforAssays
Thetrophozoiteculturesweregrownstandinguprightfor72hoursat37Cin15mL(11150mm)screwtopborosilicatetubeswhichwerefilledwith14mLof
medium(typementionedabove).After72to96hoursofincubation,theculturewasexaminedmicroscopicallyat100xmagnificationtoverifythatthetrophozoites
hadgrownsuccessfully.Successfulgrowthwasdefinedasalayeroftrophozoitescompletelyoralmostcompletelycoveringtheinnerwalloftheculturetube.
Trophozoiteswerereleasedfromthewallsofthetubebyincubatingthetubeina4Cwaterbathfor15minutesandthengentlyagitatingthetubetoevenlysuspend
thetrophozoites.
AdherenceAssay
Eachwellofa24well(1.716cm)flatbottomtissuecultureplate(LinbroT M)wasfilledwith2mLofmediumcontainingvaryingconcentrationsofantimicrobial
agentandinoculatedatthesametimewith0.1mLofmediumcontainingbetween70,000and80,000logphaseG.lambliatrophozoites.Eachplateincludeda
negativecontrolofG.lambliatrophozoitesandmediumaloneandapositivecontrolcontainingtrophozoitesincubatedinmediumcontaining30g/mLof
metronidazole.
Plateswereincubatedanaerobically(GasPakRSystemBBL)at37Candexaminedat24,48,and72hours.Priortoexamination,theplateswerecoveredwith
Parafilmtoavoidleakagefromwelltowell.Theplatesweregentlyagitatedinordertosuspendnonadherent(i.e.dead)trophozoitesintothemedium.Thepercentage
ofadherence,anestimateofthepercentofthetotalsurfaceareaofthebottomofeachwellthatwascoveredwithadheringG.lambliatrophozoites,wasthen
recorded.Wellswerecodedandreadblindlyforpercentageofadheranceat100xmagnification.Eachtestwasrunwithatleastsixreplicatesofeachdrug
concentration.
*Correspondingauthor.

Page22

Areplicatesetofplateswasmadetobereadateachof24,48,and72hours.Aftereachreading,thesetofreplicateplateswasdiscarded,andanewsetreadatthe
nextperiod.
StandardViabilityAssay
Inordertovalidatetheadherenceassay,wecompareditsresultswiththoseobtainedusingamodifiedversionofJokipiiandJokipii'smotilityassay(3).Trophozoites
wereconsideredviableiftheyfulfilledthefollowingtwocriteria:theventralflagellaweremotileandTrypanBluewasexcludedfromenteringthetrophozoite.
Validationstudieswereperformedusingthesameplatesasusedfortheadherenceassay.Afterthepercentageofadherencewasrecordedforeachwell,theplates
werecooledat4Cfor3040minutes(oruntilthetrophozoitesreleasedthemselvesfromtheplatewalls)andthengentlyagitatedtoproducehomogeneous
suspensionoftrophozoites.Analiquotofsuspensionwastakenfromeachwellandmixedinaratioof2volumesofcellsuspensionto1volumeof0.1%solutionof
TrypanBlue.Aportionofthismixturewasloadedonahemacytometerinordertocountthetotalnumberoftrophozoitesandthetotalnumberofviabletrophozoites
at450xmagnification.Resultsoftheadherenceandviabilityassayswerecomparedafter24,48,and72hoursofincubationat37C.
Preliminarystudiesweredonetoprovethattrophozoites(nonadherant)floatinginwellscontainingantibioticmediawerenotabletoreproduceagainunderstandard
cultureconditions.Conversely,wealsodemonstratedthatadherentcellswereabletoreproducenormally.Platesincubatedat72hourswithvaryingconcentrationsof
eithermetronidazoleorquinicrine,orwithmediumalonewereused.Theplatesweregentlyagitatedand0.3mLofmediumwasaspiratedfromeachwell(avoiding
contactwiththebottomorsideofthewells)andinoculatedinto3.5mLoffreshmedium.Theconcentrationofanyantibioticremaininginthemediumwasthusdiluted
atleasttenfold.Thevialswerethenincubatedintheuprightpositionfor72hoursat37Cafterwhichtheywereexaminedmicroscopicallyforevidenceofregrowth.
Regrowthwasconsideredpresentifover75%oftheinnerwallofthevialwascoveredwithacompleteoralmostcompletelayeroftrophozoites.
Theviabilityofadherenttrophozoiteswasdeterminedafterthemediumcontainingloosecellswasaspiratedfromtheplatesusedintheabovetest.Freshmediumwas
introducedandtheplateswerecooledto4Candheldfor3045minutesinordertoreleasetheadherenttrophozoites.Themediumwasthenaspiratedandcultured
asdescribedabove.
Inallcases,adherenttrophozoitesprovedviablebycultureandfloatingtrophozoitesinmediumcontaininggreaterthananIC50ofanantimicrobialwerenonviable.
Floatingtrophozoitesinmediumcontrolwellsorinwellswithlowconcentrationofantimicrobialagentswereoftenculturableandprobablyreflectedsloughingdueto
crowdingoftheadherentpopulation.
Results
ValidationoftheAdherenceAssay
TheplotsofthemetronidazoledoseresponsecurvedeterminedbytheadherenceassayandbythestandardviabilityassayareshowninFigures1and2,respectively.
Inordertocomparethetwoassays,theresultsofeachtimeperiodwerefittedtoalinearregression.Nosignificantdifferenceswereseen(byanalysisofvariance)
betweenthedoseresponsecurvesdetermined:1)atdifferenttimesforthesameassayor2)betweenthedifferentassaysforthesametimeperiod.Theresultsofthe
twoassayswerehighlycorrelated(r=0.94,p<.002at48hours)asshowninFigure3.Unlessnoted,allfurtherresultswillbeexpressedas48hourreadings.The
reproducibilityoftheadherenceassaywasverygoodthevariationbetweenreplicatewellswaslessthan7.3%andthevariationbetweenendpointsbetweentests
waslessthan12.6%.

Figure1.
SusceptibilityofG.lamblia(WB)trophozoitestovarying
concentrationsofmetronidazoleat24,48,72hours
usingthestandardviabilityassay.

Figure2.
SusceptibilityofG.lamblia(WB)trophozoitestovarying
concentrationsofmetronidazoleat24,48,72hours
usingtheadherenceassay.

Figure3.
SusceptibilityofG.lamblia(WB)trophozoitestovarying
concentrationsofmetronidazoleat48hours:Comparisonofresults
inadherenceassaytotheresultsinthestandardviabilityassay.

Page23
TABLE1.ComparisonofantimicrobialagentsastotheireffectivenessonG.lambliatrophozoites.Theparameterusedtocompare
differenceantimicrobialagentsastotheireffectivenessonG.lambliawasthe50%InhibitoryConcentrationorIC50.(g/mL).
HighSusceptibility
(<5g/mL)

IntermediateSusceptibility
(5to50g/mL)

LowSusceptibility
(50to500g/mL)

NegligableSusceptibility
(>500g/mL)

cycloheximide<0.0625

actinomycinD810

minocycline5060

erythromycin>800

quinacrineHCl0.090.12

nalidixicacid1520

paromomycin90

azosulphamide>800

anisomycin0.1250.25

tetracycline90100

bacitracin>800

metronidazole0.30.75

chlorotetracycline90100

neomycin>1000

furazolizone0.751.5

doxycyclinehyclate100150

bycozomycin>1000

rifampicin*>125

clindamycin>100

chloramphenicol137250

TMP/SMX**75/375

polymyxinBsulfate400500

streptozoticin400

*Thehighestconcentrationtested.
**Trimethoprim/sulphamethoxazolewerecombinedintheratio1/5.

EffectivenessofAntimicrobialAgents
TheparameterusedtocomparedifferentanitmicrobialagentsastotheireffectivenessonG.lambliawasthe50%inhibitoryconcentration,alsocalledIC50.IC50is
definedastheconcentrationofantimicrobialagentatwhichthepercentageofadherencereadingis50%ofthepercentageofadherencereadingofthenegative
control.AllIC50swerecalculatedfromthepercentageofadherenceseeninG.lambliaafter48hoursofexposure.AsshowninTable1,G.lambliawashighly
susceptibletoquinacrineHCl,metronidazole,andfurazolidone.Inadditiontothesecommonantigiardialdrugs,cycloheximideandanisomycinshowedinhibitory
effectsatthesameorevenlowerconcentrations.G.lambliaalsoshowedintermediatesusceptibilitytoactinomycinDandnalidixicacid.
StrainDifferences
FiveG.lambliastrainsweretestedfordifferencesinsensitivitytometronidazoleandquinacrineHCl(Figure4).Formetronidazole,nomeaningfuldifferencesinIC50.
wereobserved,buttheCATstrainwassignificantlylesssusceptibletoquinacrine(IC50.=0.185g/mL)comparedtotheothers(IC50.=0.10g/mL).

Figure4.
ComparisonofsusceptibilitiesofG.lambliastrainsto
metronidazoleandquinacrineat24,48,72hours:50%Inhibitory
Concentration(IC50.)intheadherenceassay.Samplevariances
werenegligibleforallstrains(n=18).

Discussion
TheinvitrominicultureadherenceassayfordeterminingG.lambliasusceptibilitytoantimicrobialagentsshowedahighcorrelationwiththestandardmethodovera
numberofagents,strainsandtestconditions.TheadherenceassayappearstobeassensitiveasGillin'sclonalmethodorJokipii'smotilityassayyetoffersresultsin
onethirdofthetimeoftheformer(48hoursto144hours)respectively,withmuchlesstediumthanthelatter.
AprominentcharacteristicoflivingG.lambliatrophozoitesisadhesiontosurfaces.Adhesionisprobablycausedbytheflowgeneratedbythecontinuingactivityof
theventralflagellawhichproducesasuctionpressureundertheventraldisc3(7,8).Adherencethusissynonymouswithventralflagellamotility,thecriteriawhich
Jokipiiusedtojudgeviabilityinhisassay.
This,ofcourse,meansitisnotverycommonfordeadtrophozoitestoadherepassivelytocultureplates.Theuseofadherancepermitsonetoexamineviabilityby
measuringtheadhesiveactivityofaggregatesofG.lambliaratherthanconfiningdeterminationofviabilitytoindividualtrophozoites.
WeusedtheadherenceassaytodetermineG.lambliasusceptibilitytoseveraldifferentantimicrobialagents.Weexaminedgeographicallydistinctaxenicallycultured
strainsfortheirsensitivitytoantimicrobialagents.Nodifferenceswereseeninthesensitivitytometronidazolebetweenthefivestrains.After48hoursexposure,there
wasseenasignificantdecreaseinthesensitivityoftheCATstrainstoquinacrineHClcomparedtotheotherfourstrains,butthisdifferencewaslessthan2fold.In
contrast,Jokipiiusingnonaxenicallyculturedstrainsfounduptoa4foldvariationinantimicrobialagentsensitivitypatternsat48hours.Thisdifferencemaybecaused
bydifferencesinourtechniquesfortestingantimicrobialagentsensitivityaswellasthepossibledifferencesinstrainstested.Weusedwellestablishedaxenicculture
strainswhereastheFinnishgroupusednewlyisolatedstrains.

Page24

G.lambliawassusceptibletoanisomycinandcycloheximide,bothofwhicharethoughttoactbybindingtothe80Sribosomeandinterferingwiththetranslationof
peptidyl+RNA(9,10).Emetine,whichisalsothoughttobindtothe80Sribosome,hasalsobeenreportedtobeaneffectiveantigiardialagentwhentestedbythe
clonalmethod(2).G.lambliawasalsosusceptibletoantimicrobialagentswhicharebelievedtoaffectthesynthesisorstructureofDNA.Theseincludequinacrine
HCl,metronidazole,furazolidone,actinomycinD,andnalidixicacid.Ofthese,metronidazole,quinacrineHCl,andfurazolidonearethecurrentagentsforuseagainst
giardiasissincetheyarecomparativelynontoxic.Nalidixicacidwasshowninthisstudytobeapotentiallyusefulantigiardialagent.Duetoitspharmokinetics,
however,itsclinicaluseisunlikely.Drugssuchasminocycline,tetracycline,chlorotetracycline,anddoxycycline(allofwhicharebelievedtoinhibitproteinsynthesison
the70Sribosomealongwithsomeinteractiononthe80S)showedonlysmalleffectsonG.lamblia.WewerenotabletoconfirmGillin'sfindings(2)thatbacitracin
wasaneffectiveantigiardialagentsinceitdemonstratednoinhibitionofgrowthinourexperimentsevenatconcentrationsashighas1000g/mL.Ourdatasuggest
that,inadditiontothetraditionalantigiardialdrugs(theDNAinhibitors:metronidazoleandquinacrine),G.lambliaisespeciallysensitivetodrugswhichbindsolelyto
the80Sribosome,andisnotaffectedbydrugswhichbindsolelytothe70Sribosome.
CurrentdrugsusedforthetherapyofG.lambliaareexpensiveandhavesignificanteffects.Bettertoleratedagentswouldindeedbewelcome.Oneapproachtothis
problemmaybetosearchforantimicrobialagentswhichbindtothe80Sribosomethatwouldaffectonlyprotozoaleucaryoticcells.
LiteratureCited
1.Keister,D.B.1983.AxeniccultureofGiardialambliainTYIS33mediumsupplementedwithbile.Trans.Roy.Soc.Trop.Med.andHyg.77:487488.
2.Gillin,F.D.,andL.S.Diamond.1982.InhibitionofclonalgrowthofGiardialambliaandEntamoebahistolyticabymetronidazole,quinacrineandother
antimicrobialagents.J.Antimicrob.Chemother.8:305316.
3.Jokipii,L.andA.M.Jokipii.1980.InvitrosusceptibilityofGiardialambliatrophozoitestometronidazoleandtinidazole.J.Inf.Dis.141:317325.
4.Meyer,E.A.1970.IsolationandaxeniccultivationofGiardiatrophozoitesfromtherabbit,chinchilla,andcat.Exp.Parasitol.27:179183.
5.Smith,P.D.,F.D.Gillin,N.A.Kaushal,andT.E.Nash.1982.AntigenicanalysisofGiardialambliafromAfghanistan,PuertoRico,Ecuador,andOregon.Infect.
Immun.36:714719.
6.Zar,J.1974.BiostatisticalAnalysis.PrenticeHall,EngelwoodCliffs,N.J.228pp.
7.Holberton,D.V.1973.FinestructureoftheventraldiskapparatusandthemechanismofattachmentintheflagellateGiardiamuris.J.Cell.Sci.13:1141.
8.Holberton,D.V.1974.AttachmentofGiardiaAhydrodynamicmodelbasedonflagellaractivity.J.Exp.Bio.60:207221.
9.Franklin,T.J.,andG.A.Snow.1978.BiochemistryofAntimicrobialAction.2ndEd.ChapmanandHall,London.
10.Gilman,A.G.,L.S.Goodman,andA.Gilman.1980.ThePharmacologicalBasisofTherapeutics.6thEd.MacmillanPublishingCo.,Inc.NewYork,N.Y.

Page25

UltrastructuralStudyofaBacterialSymbiontofGiardialamblia
S.Radulescu,E.A.Meyer*,B.Burghelea,andT.Meitert
CantacuzinoInstitute,Bucharest,Romania
WereportherethepresenceofbacterialendosymbiontsinGiardiacystsfromhumans.Subjectsofthisstudywere10preschoolchildrenwithGiardiainfections.Fiveofthesechildren
receivedaliveShigellavaccine,producedinRomania.Theother5childrenwerenotvaccinated.TEMexaminationsofthecystsofthecontrolgroupfailedtorevealendosymbionts.
Cystsfromthevaccinatedchildrenrevealedbacterialikeendosymbionts.TheseendosymbiontsweresimilartothosedescribedbyNemanicetal.(12)inG.muris,andare
apparentlyboundedbytwounitmembranes.ThepossiblesignificanceofthisassociationbetweenGiardiaandShigellaarediscussed.Webelievethatthisisthefirstreportofa
bacterialsymbiontinGiardialamblia.

Introduction
Giardia,perhapstheearliestdescribedintestinalprotozoan,hasrecentlybeenthesubjectofrenewedresearchinterestwiththerecognitionof(a)itspathogenicityand
(b)itsinvolvementinepidemicsofwaterbornediarrhealdisease.
OneareaofinteresthasbeenthestudyoftheultrastructureoftheseprotozoaitistobehopedthatanunderstandingofthecellbiologyofGiardiamayincludean
understandingofaneffectivedefensestrategyagainsttheseparasites.
Duringtheseultrastructuralstudiesofparasiteorganelles,thepresenceofsymbioticorganismswasnotedtheroleofthesesymbiontsremainsunclear.
MostofthereportstodatedescribeendosymbiontsinGiardiafrommice.Thus,Nemanicetal.(12),describedendosymbiontsintrophozoitesandcystsfromtheG.
murisemployedinthemousemodeloriginallyestablishedbyRobertsThomsonetal.(17)theoriginalsourceoftheseorganismswasthegoldenhamster.
Otherworkersearlierhadreportedsimilarstructures.Boeck,forexample,notedin1917(5)rodlikebacteriacoveringthetrophozoitesofGiardiafrommeadow
micesomeoccurredasinclusionswithinthetrophozoiteaswell.Wenrichin1940(23)observedintracytoplasmicrodsinGiardiatrophozoitesandcystsfromrats
andmice.Brug(6)describedcytoplasmicinclusionsfromtheGiardiaofwhitemice,whichBall(2)laterconcludedcouldhavebeenchytridfungi.Solovievand
Chentsovin1970(21)illustratedbytransmissionelectronmicroscopythepresenceofroundedstructures,ofunknownnature,surroundedbyadoublemembranein
thecytoplasmofbinucleatecystsoftheG.muristype.Inanunpublishedobservationin1978,RadulescufoundendosymbiontsinGiardiaisolatedfromnaturally
infectedhamsters.
Radulescuetal.(15)notedin1982thepresenceofrodshapedstructuresinGiardiacystsfromchildrenwhohadreceivedaliveShigellavaccinesimilarrod
shapedinclusionswerenotedintracytoplasmicallyinthetrophozoitesofG.murisfrommicewhichhadbeenintubatedwithS.flexneri.Thesebodieswerenot
observedintheprotozoacollectedbeforetheadministrationofthevaccine.
SogayarandGregorio(20)reportedfindingtwotypesofcytoplasmicinclusions,introphozoitesofG.murisfromhamsters,andintrophozoitesoftheG.duodenalis
typefromdomesticrats.Bothwerelimitedbyadoublemembrane.
Morerecently,Feely(Abstr.Annu.Meet.Am.Soc.Parasitol.1986,97,p.56)describedtheuseofaDNAspecificfluorescentdye,Hoechst33258,todetect
symbiontsinsideG.microtitrophozoites.Morethanthreequartersoftheparasitesexaminedhadsymbionts,somecontainingmorethan150inthecytoplasm.
ItseemslikelythattheoccurrenceofsymbiontswithinGiardiaisnotarareevent,butarelativelycommononethathasnotbeenmorefrequentlyreportedbecauseit
wasnotsearchedfor.
Intheexperimentsdescribedhere,wedecidedtoexperimentallyreproducetheconditionspresentinchronicgiardiasisassociatedwithShigellainfection.Todothis,
weadministeredaliveoralShigellaflexnerivaccinetoagroupofchildrennaturallyinfectedwithGiardia.Thevaccineregimen,whichinvolvesrepeateddosesof
largenumbersofbacteria,providedtheopportunityforthebacterialprotozoalassociation.
MaterialandMethods
Theseexperimentswereconductedonagroupof10preschoolchildrenwithGiardiainfection,fromachildcarefacilitywithahistoryofintestinaldisease.The
childrenwerevaccinatedorallywithlivedysenteryvaccine(VADIZEN)producedbytheCantacuzinoInstitute,Bucharest.Thevaccinewasadministeredin5graded
dosesrangingfrom0.3to1.2mLofliveShigellaflexneri2aT32Instratesuspension(1011
*Correspondingauthor.
Address:Dept.ofMicrobiology&Immunology,OregonHealthSciencesUniversity,3181SamJacksonParkRd.,Portland,Oregon,97201,U.S.A..

Page26

organisms/mL),atthreedayintervals.
Beforevaccination,andthreedaysafterthelastvaccineadministration,feceswerecollected,andGiardiacystswereconcentratedaccordingtothefiltrationand
sucroseflotationmethodofBinghametal.(3).
PreparationoftheMaterialforElectronMicroscopeStudy
Thepelletofcystswasfixedin2.5%glutaraldehydein0.1Mphosphatebuffer(pH7.2)forahourat4C.Aftersuccessivewashingin0.15Mphosphatebuffer(pH
7.2)thepelletwaspostfixedin1%osmiumtetroxidein0.1Mphosphatebufferfor1hourat4C,thenthepelletwasembeddedinagar.Theresultingfragmentswere
dehydratedinincreasingacetoneconcentrationsandfinallyembeddedinVestosol.ThinsectionswerecutonaPorterBlumMT1microtomeandstainedwith
saturateduranylacetatein50%ethanol,followedbyleadcitrate.ThesectionswereexaminedinaHitachiHU11electronmicroscope.Thecontrolcysts,obtained
fromthefecesofthefivechildrenwithgiardiasiswhohadnotreceivedthevaccine,wereconcentratedasdescribedabove.
Results
TheelectronmicroscopestudyoftheGiardiacystsobtainedfromthechildrenwhohadreceivedtheShigellavaccinerevealedthefollowing:
Thecystsareellipticallyshapedandrangeinlengthfrom6to10m.
Thecystwallisoftencontiguouswiththecellmembrane,butinsomeplacesaspace,containingflagellaraxonemes,canbeobserved.Anetworkofvacuolesand
mictrotubulesexistsintheperipheralcytoplasm.Otheridentifiableorganellesinthecytoplasmincludenuclei,flagellaraxonemes,roughendoplasmicreticulum,median
bodies,ventraldiscfragments,andribosomes.

Figure1.
TEMofaG.lambliacystillustratingthefibrillarcystwall(CW),
profilesoffragmentedportionsoftheventraldisc(VD),
flagellaraxonemes(FA),andendosymbiont(E).x27,000.

Figure2.
TEMofpartofG.lambliacyst.Microtubulesofthemedian
body(MB)endosymbiontsuggestingbinaryfission(E).x37,000.

Inadditiontoparasiteorganelles,weobservedotherstructures,apparentbacilluslikeendosymbionts,whosesizerangedfrom250350nminwidthand5001200
nminlength(Figure1).Theendosymbiontswereboundedbytwounitmembranes,oftenwithaclearspacebetweenthemembranes.Theseendosymbiontshavethe
appearanceofbacteria,withdenseperipheralcytoplasmandcentralbrightareascontainingstrandsoffibrillarmaterialresemblingDNA.Someendosymbionts
resembledbacteriaundergoingbinaryfission.Noneoftheendosymbiontsappearedtobeintheprocessofdegenerationorintracellulardigestion(Figures2and3).
TheGiardiacystsfromthecontrolgroupofchildrenpresentedallofthesamecharacteristicsasthoseofthevaccinatedgroup,exceptthatinnoneofthemwere
endosymbiontsobserved.
Discussion
ThisstudyindicatesthatbacteriacanassociateendosymbioticallywithGiardiafromhumansaswellasloweranimals.Webelievethatthisisthefirstreportofsuch
endosymbiontsinGiardiaisolatedfromhumans.
TheendosymbiontsinGiardiafromhumanswereessentiallythesameasthosedescribedinG.murisbyNemanicetal.(12),exceptthatinthelatterstudylarger

Page27

Figure3.
TEMofpartofaG.lambliacystillustratingthe
endosymbiont(E),nucleus(N),andventral(VD).x37,000.

numbersofsymbiontswereseen.
Thedetectionoftheseendosymbiontsraisesanumberofquestionswhichrequirefurtherstudy.Onequestionunaddressedthusfarconcernsthemeansbywhichthe
endosymbiontbecomesinternalizedwithintheparasite.Giardiaarenotknowntoconsumeparticlesthesizeofbacteria.BockmanandWinborn(4)have
demonstratedthatferritincanbetakenupfromthesmallintestinallumenandlocalizedinvacuolesbeneaththeplasmalemmaofGiardiatrophozoites.Thereisno
evidencetosuggestthatthesevacuolesareinvolvedintheuptakeofparticlesaslargeasbacteria.
Wesuggestthepossibilitythatthebacteriamayactivelyinvadethetrophozoiteinamannersimilartothatdescribedinsomeamoebae(2).Onceinsidethetrophozoite,
thebacteriumissurroundedbymembraneswhichfusewithlysosomesasinthecaseofphagolysosomes.Thestrategy,intheevolutionofthebacteriaprotozoan
relationshiptowardendosymbiosisandawayfromdigestion,consistsinthecapacitytohamperthisfusionwithlysosomes.Inthecaseofamoeba,themembranesof
symbiontcontainingvesicleshavea200kDpolypeptideonthecytoplasmicside,whichissuspectedtobeinvolvedinthepreventionoflysosomalfusion(16).These
membranesplayakeyroleintheinteraction.Theyarespecificallyformedaroundthesymbiontasaresultofcellcellrecognitionevents,theyhaveprotective
properties,andtheycanprobablycontrolmetabolicchanges.Nonsymbioticbacteriaaredigestedin48hours,whilesymbioticonessurviveandestablishthe
association.Weconsideritlikelythattheintracellularbacteriaweobservedinthisstudyaresymbioticbecauseoftheirintactmorphologicalappearanceandthe
absenceofanysignofintracellulardigestion.
Innature,agivenprotozoanparasiteisexposedtoalimitedrangeofpotentialbacterialpartnersofthese,evenfewerapparentlybecomeassociatedinaspecific
symbioticrelationship.Inthepresentsituation,bacterialendosymbiontswereobservedinGiardiacystsinchildrenwhohadreceivedliveShigellabacteria
endosymbiontswereabsentinthosenotexposedtothebacteria.Itseemslikelythatthisrelationshipwastheresultoftheparasitebeingoverwhelmedbylarge
numbersofbacteriacapableofparticipatingintherelationship.
TheaboveobservationsdonotpermitconclusionsregardingthesignificanceoftherelationshipbetweenGiardiaandendosymbioticbacteria.Itisgenerally
consideredthatasymbioticassociationresultsintheacquisitionofnewqualitativelysuperiorpropertiesbyeachparticipant.Thesepropertiesbecomeofmedical
significancewhentheyinvolvesuchcharacteristicsaspathogenicity,antigenicity,drugresistance,andenvironmentalsurvival.
GiardiaandShigellaoftenoccurtogetherinachronicdiarrhealsyndrome.Anumberofworkershavereportedthatthefrequencyanddurationofdiarrhealepisodes
ishigherinthedoubleinfectionthanininfectioncausedbyeitherorganism(8,22).Noone,however,hassuggestedthepresenceofShigellaendosymbiontsas
contributingtoprolongationofdiarrhea.
AnotheraspectofthisrelationshipyettobestudiedisthedeterminationwhetheraplasmidisinvolvedintheabilityofShigellatoinvadeGiardiaanS.flexneri
plasmidispresentlyknowntobeinvolvedinitsinvasionofHeLacells(10).
Otherexamplesareknownofprotozoanbacterialsymbiosis,inwhichonepartnerortheotherisinvolvedinhumandisease.Theyhavebeendescribedin
Acanthamoebasp.,forexample(7,14),Naegleriafowleri(13),andDientamoebafragilis(19).Rowbotham(18)hassuggestedthatAcanthamoebaand
NaegleriaarepossiblenaturalhostsforLegionellahesuggestsevenmoreimportantlythatinfectedamoebaeorvesiclescontainingbacteriacouldbetheinfective
particleforLegionellainfectionofhumans.
Trypanosomatidsalsoharborbacterialikeendosymbionts.Whilethisgroupofflagellatedprotozoaincludesmemberswhichareintracellularparasitesofavarietyof
hosts,somespeciesareinturnhoststointracellularsymbionts.OneoftheseorganismswasthesubjectofarecentstudyreportedbyKrylovetal.(9).Theorganism
studied,Crithidiaoncopelti,wasconsideredacompositespecies,madeupofcellscontainingbacterialendosymbiontsandsymbiontfreecells.Whenseparated,the
twodifferedsignificantlyinanumberofrespectsincludingsize,flagellumlength,colonymorphology,velocityofmovement,malatedehydrogenaseisoenzymes,growth
at35C,

Page28

oxygenrequirement,andsensitivitytoantibiotics.Theauthorsproposethatthetwoisolatesbegivenseparatespeciesnames.
ThepresenceofbacterialendosymbiontsinGiardiaclearlycouldhaveanyofavarietyoframifications.Ontheonehand,itmayresultinprotecting,andfacilitatingthe
transferof,abacterialendosymbiontpathogenictohumans.Ontheotherhand,endosymbiontsrepresentextrachromosomalcarriersofgeneticinformation,the
consequencesofwhichmayresultinmorphologic,physiologic,andpathogenicchangesintheprotozoan.Bacterialendosymbiontscanaffecttheresultsof
endonucleaserestriction,isozyme,andantigenanalysesthatpresentlyarebeingconductedineffortstoclarifythetaxonomyofthiscomplexgroupofprotozoa.One
approachtodeterminingthesignificanceoftheGiardiaShigellarelationshipisitsstudyinvitro.
LiteratureCited
1.Alfieri,S.C.,andE.P.Camargo.1982.Trypanosomatidae:Isoleucinerequirementandthreoninedeaminaseinspecieswithandwithoutendosymbionts.Exp.
Parasitol.53:371380.
2.Ball,G.H.1969.Organismslivingonandinprotozoa.In:T.T.Chen(ed.),Researchinprotozoology,Vol.3,PergamonPress,Oxford,p565718.
3.Bingham,A.K.,E.L.Jarroll,E.A.Meyer,andS.Radulescu.1979.IntroductionofGiardiaexcystationandtheeffectoftemperatureoncystviabilityascompared
byeosinexclusionandinvitroexcystation.In:Jakubowski,W.andJ.C.Hoff(eds.),WaterborneTransmissionofGiardiasis,EnvironmentalProtectionAgency
600/979001,p217229.
4.Bockman,A.K,.andW.B.Winborn.1968.ElectronmicroscopiclocalizationofexogenousferritinwithinvacuolesofGiardiamuris.J.Protozool.15:2630.
5.Boeck,W.C.1917.MitosisinGiardiamicroti.Univ.Calif.Publ.Zool.18:126.
6.Brug,S.I.1942.EigentumlicheEinschlussinLambliamurisZentralbl.Bakteriol.(Orig.A)148:166168.
7.Hall,J.,andH.Voelz.1985.BacterialendosymbiontsofAcanthamoebasp.J.Parasitol.71:8995.
8.Ingram,V.,F.L.Rights,K.Hashimi,andK.Asari.1966.DiarrheainchildrenofWestPakistan:occurrenceofbacterialandparasiticagents.Am.J.Trop.Med.
Hyg.15:743750.
9.Krylov,M.V.,S.A.Pidlipaev,A.S.Khaetskii,L.M.Belove,A.O.Frolov,andB.Y.Niyazbekova.1985.Isonly1speciespresentinacultureofCrithidia
oncopelti(Kinetoplastmonada,Trypanosomatidae)Zool.Zh.64:165171.
10.Maurelli,A.T.,B.Baudry,H.D'Hauteville,T.L.Hale,andP.M.Sansonetti.1985.CloningofplasmidDNAsequencesinvolvedininvasionofHeLacellsby
Shigellaflexneri.Infect.Immun.49:164171.
11.McGhee,R.B.,andW.B.Cosgrove.1980.BiologyandphysiologyofthelowerTrypanosomatida.Microbiol.Rev.44:140173.
12.Nemanic,P.C.,R.L.Owen,D.P.Stevens,andJ.C.Mueller.1979.Ultrastructuralobservationsongiardiasisinamousemodel.II.Endosymbiosisandorganelle
distributioninGiardiamurisandGiardialamblia.J.Infect.Dis.140:222228.
13.Phillips,B.P.1974.Naegleria:Anotherpathogenicamoeba.Studiesingermfreeguineapigs.Am.J.Trop.Med.Hyg.23:850855.
14.ProcaCiobanu,M.,G.H.Lupascu,A.L.Petrovici,andM.D.Ionescu.1975.ElectronmicroscopicstudyofpathogenicAcanthamoebacastellanistrain:the
presenceofbacterialendosymbionts.Intern.J.forParasitol.549556.
15.Radulescu,S.,M.Smolinschi,C.Rau,andE.A.Meyer.1982.Giardiaspp.cysts:possiblevectorsofenterobacteria.Mol.Biochem.Parasitol.Suppl.p578.
16.Reisser,W.,R.Meier,H.D.Gortz,andK.W.Jeon.1985.Establishment,maintenance,andintegrationmechanismsofendosymbiontsinprotozoa.J.Protozool.
32:383390.
17.RobertsThomson,I.C.,D.P.Stevens,A.A.F.Mahmoud,andK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gastroenterology71:5761.
18.Rowbotham,R.J.1980.PreliminaryreportonthepathogenicityofLegionellapneumophilaforfreshwaterandsoilamoebae.J.Clin.Pathol.33:11791183.
19.Silard,R.,andB.Burghelea.1986.EndosymbiontsinDientamoebafragilistrophozoitesresistanttoantiprotozoaldrugs.Arch.Roum.Pathol.Expl.Microbiol.
45:6574.
20.Sogayar,M.L.,andE.A.Gregorio.1966.ElectronmicroscopiclocalizationofexogenousferritinwithinvacuolesofGiardiamuris.J.Protozool.80:4952.
21.Soloviev,M.M.,andJ.S.Chentsov.1970.UltrastructureofcystsofLambliamuris.Parasitologia4:510514.
22.Wanner,R.G.,F.O.Atchley,andM.A.Wasley.1963.AssociationofdiarrheawithGiardialambliainfamiliesobservedweeklyforoccurrenceofenteric
infections.Am.J.Trop.Med.Hyg.12:851853.
23.Wenrich,D.H.1940.Observationsonparasitesandinclusionbodiesincertainintestinalprotozoa.Science92:416417.

Page29

MorphologyofGiardiaEncystationInVitro
DanielG.Schupp,MaryM.Januschka,andStanleyL.Erlandsen*
DepartmentofCellBiologyandNeuroanatomy,SchoolofMedicine,UniversityofMinnesota,Minneapolis,Minnesota55455,U.S.A.
DevelopmentofaninvitroencystationmodelforGiardiatogetherwithcurrentexcystationmodelswouldpermitinvitroregulationoftheentirelifecycleofthisprotozoan.Topursue
thisgoalwemonitoredseveralaxenicallygrownstrainsofGiardiaanddetectedtheproductionofcystsinvitro.Thesecystsdisplayedalightmicroscopicmorphologyidenticalto
controlGiardiacystsisolatedfromfeces.ThismorphologywasbasedonthecharacteristicsizeandshapeofGiardiacysts,andthepresenceof24nuclei.Giardiacystsformedinvitro
wereshowntohavethesamepositiveimmunoreactivityforthecystwallascontrolGiardiacysts.Theviabilityofcystsformedinvitrowasmeasuredbyincorporationofthe
fluorogenicdye,fluoresceindiacetate,andwasvariable,rangingupto50%.Ultrastructuralanalysis,usingtransmissionelectronmicroscopy,revealedanoutercystwallwitha
fibrillarorganization,thepresenceofcystwallmembranes,aperitrophicspace,andnuclei.Otherintracellularorganellesobservedincludedelementsofthedisassembledventral
adhesivedisc,peripheralvacuoles,flagellaraxonemes,andbasalbodies.Examinationofcultureswiththescanningelectronmicroscopeshowedthatnewlyformedcystshadthe
characteristicsizeandshapeofcontrolcystsandwereinterspersedamongtrophozoites.TheseresultsdemonstratedthatGiardiaencystationoccurredinaxenicculturesof
trophozoites.CystsformedinvitrowerebothmorphologicallyandimmunologicallysimilartoGiardiacystsformedinvivoandalsowereviable,asdemonstratedbytheuptakeof
fluorogenicdyes.

Introduction
GiardiaencystationinvivohasbeendescribedbyPerroncitoin1888(13)andbyLavierin1942(8),whomicroscopicallystudiedtrophozoitesflushedfromthe
intestinesofamphibians,rats,andthediarrheicstoolsofhumans.NodescriptionofGiardiaencystationinvitrohasbeenreportedeventhoughaxenicculturesof
Giardiaweredevelopedalmostadecadeago(11).TheinabilityofGiardiatrophozoitestoundergoencystationinvitrohasraisedquestionsastowhetherhost
relatedfactors,suchasgutassociatedmicroorganismsornutritivefactorsandstimulifromthesmallintestine,mightplayaroleinthisprocess.
ThedevelopmentofaninvitromodelforencystationofGiardiawouldenablethecompletelifecycleofthisprotozoantobemanipulatedoutsideofthehost.This
modelwouldbebeneficialinmanyaspectsofGiardiaresearch,includingbiochemicalanalysisofthecystandthecystwall,developmentofgiardicidalagentsinvolved
inblockingcystwallformation,thetestingofKochs'postulatesinregardtoinfectivityofcysts,andtheinvestigationofquestionsrelatedtonucleardivisionandcyst
wallproduction.
HerewereportmorphologicalevidencefortheformationofGiardiacystsinaxenicculturesoftrophozoitesandpresentpreliminaryfindingsontheirviability,as
determinedbytheincorporationoffluorogenicdyes(15).
MaterialsandMethods
AxenicculturesofGiardiaderivedfromman(WBstrain,AmericanTypeCollection,Bethesda,MD),beaver(IP0482:1andIP0583:1fromDr.LouisDiamond
PB1andB5fromDrs.PeterWallisandHenryStibbs),andmuskrat(MR4fromDrs.PeterWallisandHenryStibbs)weregrownonTYI33mediumwithadult
bovineserumasdescribedbyKeister(5).Encystationinvitrowasstimulatedbyadditionofbiletothemedium(17).Thecultureswereexamineddailybyinverted
phasemicroscopyand,afterthreedaysofgrowth,thecontentsoftheculturewerepelletedbycentrifugationfor10minutesat600g.Aslidewasmadedirectly
fromthispelletforlightmicroscopicobservation,andexaminedusingeitheraZeissphotomicroscopeIIoranOlympusBH2lightmicroscopeequippedwithphase,
DICandUVepiillumination.Theviabilityofcystsformedinvitrowasdeterminedbytheincorporationofthefluorogenicdyes,fluoresceindiacetate(FDA)and
propidiumiodide(PI)(15).Forscanningelectronmicroscopy(SEM),thepelletedcultureswerewashedtwicein0.9%saline,centrifugedasabovebetweenwashes,
thenresuspendedinsalineandplacedonglasschipsthathadbeencoatedwithpoly1lysine(SigmaChemicalCo.,St.Louis,Mo.).Afterallowingtheculturedcells
toadheretotheglasschipsfor1hour,thesampleswerefixedforthreetofourhourswith2.5%glutaraldehydebufferedwith0.1MsodiumcacodylateHCl,pH7.4.
Thiswasfollowedbypostfixationin1%osmiumtetroxideinthesamebuffer,foronetotwohoursat4C.TheglasschipswithattachedGiardiawerethen
dehydratedinanascendingethanolseriesandcriticalpointdriedusingthetechniqueofAnderson(1).Thespecimensweresputtercoatedwithgoldpaladiumand
mountedonstubswithsilverpaintandcoppertape.Allsampleswereexaminedat20kVinaHitachiS450scanningelectronmicroscopeandmicrographswere
recordedonPolaroidtype55P/Nfilm.Fortransmissionelectronmicroscopy(TEM)thepelletedcultureswerefixedasdescribedabovewiththeexceptionofadding
afewdropsofalbumintotheculturedmaterialbetweenthefixativesteps.Thefixedalbuminblockcontainingthecellsuspensionwasthendehydratedandembedded
inepoxyasdescribedbyLuft(10).SectionswerecutonaLKBHuxley
*Correspondingauthor.

Page30

ultramicrotome,stainedwithuranylacetateandleadcitrate,thenexaminedwithaJEOL100CXelectronmicroscope.
Results
AxenicculturesofGiardiawerestimulatedtoundergoencystationinvitrobyadditionofbiletothemedium(17).Afteroneday,examinationoftheculturesby

Figures1and2.
Differentialinterferencecontrast(DIC)micrographsofa
Giardiaculturethreedaysafterinductionofencystationwithbile.
AclusterofviablelookingGiardiacysts(arrowheads)formedinvitro
areseenbetweentrophozoites(T).InFigure2,theviablecysts
possessawelldefinedcystwall(CW)andperitrophicspace(PS).
Thecytoplasmhasahyalineappearance,butfaintoutlinesofsome
organellesincludingnuclei(N),axonemesofflagella(AX),andportions
oftheadhesivedisc(AD)canbeseen.Barequals5microns.

Figures3and4.
DICandfluorescencemicrographoftwoGiardiacystsafterincorporation
ofthefluorogenicdyes,FDAandPI.Thecystformedinvitro
(arrowhead)showninFigure3hadthemorphologicalcharacteristics
usedtodetermineviabilityinG.muriscysts(16).Thiscyst
incorporatedthefluorogenicdye,FDA,andemittedagreenfluorescence
whenviewedwithanexcitationwavelengthof450490nm.Thecystwith
nonviablemorphology(N)seeninFigure3wasstainedwiththefluorogenic
dye,PI,asshowninFigure4,andfluorescedorange.
Barequals5micronsforFigures3and4.

Figure5.
TEMofGiardiacystformedinvitro.Thiscysthasthemorphological
characteristicsofviablecystsbyDIC,namelyaclearlydelineatedcyst
wall(CW)ofuniformthicknessandaperitrophicspace(PS)justinterior
tothecystwall.Otherorganelleswithinthecystincludeflagellar
axonemes(AX),peripheralvacuoles(V),andportionsofthe
disassembledadhesivedisc(AD).Barequals1micron.

Figures6and7.
SEMsofaGiardiacultureafterexposuretobileforthreedays.
Giardiacysts(arrowheads)formedinvitrowereseenscattered
betweenflagellatedtrophozoites,whichwereattachedtothesubstratum.
Athighermagnification,seeninFigure7,thecontrastbetweentheflagellated
trophozoiteandtheinvitroformedcystwasreadilyapparent.
Thecystswereovaltoroundinshapeandpossessedasmoothsurface,
whereasthepearshapedtrophozoiteswereflattenedinadorsalventral
planeandattachedtothesubstratumbytheirventraladhesivedisc.
Barequals5micronsforFigure6and1micronforFigure7.

invertedmicroscopyrevealedsmallfociorclustersoftrophozoites.Bydaythree,theseclustershadincreasedinsizeandincludedtrophozoitesaswellascysts.
ExaminationoftheculturesondaythreebyDICmicroscopyrevealedthepresenceofGiardiacystsinterspersedamongsttrophozoites(Figures1and2).Thecysts
formedinvitroweredeterminedtobeoftwotypes,usingmorphologicalcriteria.Onetypeappearedviablebasedonpreviouslypublishedcriteria(16)and
possessedthetypicalmorphologicalfeaturesofcystsformedinvivo.Theseviablecysts,showninFigure2,wererecognizedbytheircharacteristicsize,shape,and
thepresenceofdistinguishingfeatures,suchascystwall,peritrophicspace,nuclei,flagellaraxonemes,andcurvedportionsoftheadhesivedisc.Theothertypeofcyst
(notshown)wasofsimilarsize,shape,andhadacystwall,butappearednonviablesincethecystusuallylackedthepresenceoftypicalorganellesorcontainedtwoto
fournucleiwithinashrunkencytoplasmicmass.
TheviabilityofGiardiacystsformedinvitrowastestedusingtheincorporationoffluorogenicdyes.Morphologicallyintactcystswereoftenseentoincorporate

Page31

FDA,whereasthosecystslackingrecognizablecontentsorhavingashrunkencytoplasmicmassstainedwithPI(Figures3and4).Awidevariationincystviability
wasdetectedusingtheincorporationofFDA,rangingfrom4050%tolessthan1%,inGiardiacystsformedinvitro.
AGiardiacystformedinvitroisillustratedbyTEMinFigure5.Thiscyst,basedonits'morphologicalappearance,issimilartotheviablecystsseenbyDIC(Figure
2).TypicalorganellescharacteristicforGiardiaareseenwithinthecytoplasm,includingacystwall,peritrophicspace,flagellaraxonemes,andcurvedportionsofthe
adhesivedisc,allofwhichareindistinguishablefromthoseofcontrolcystsformedinvivo.
Examinationofaxenicculturesproducingcysts,usingSEM,alsorevealedthepresenceofcystsinterspersedamongtrophozoites(Figures6and7).Theflattened
trophozoiteswerepearshaped,oftenattachedtothesubstratumbytheiradhesivediscs,andwereeasilyrecognizedbythepresenceofflagella.Thecystsformedin
vitrowereroundtoovalinshape,rangingfrom711micronsinwidthand1114micronsinlength,andhadacystwallthat,examinedbySEM,appearedtobe
smooth.
Discussion
Ourresults,usingbothlightandelectronmicroscopy,havedemonstratedthatGiardiacystsformedinvitrohaveamorphologicalappearanceindistinguishablefrom
thatofmurineorhumancystsformedinvivo(3,4,7,9,12,18).Themorphologicalappearanceofthecystsformedinvivo,asseenbyDIC,closelyresembledthatof
G.muriscyststhathadbeenshowntobeviableduebothtotheirabilitytoincorporatefluorogenicdyesandinfectanimals(16).Theabilitytodifferentiateviablefrom
nonviablecystswasbasedonthedistinctappearanceofthecystwall,thepresenceofaperitrophicspacebetweenthecystwallandcytoplasm,andthedistinct
definitionoftheintracellularorganellesobserved.Thesemorphologicalfeatureswerealsoprominentincystsformedinvitro(Figure2)andsuggestedthattheytoo
wereviable.ThestrikingultrastructuralsimilaritybetweencystsformedinvitroandviableGiardiacystscorroboratedourpriorassumptionthatthecystsformedin
culturewereindeedviable.
TheviabilityofGiardiacystscanbemeasuredbyeithertheincorporationoffluorogenicdyes(15),excystation(2)orbyproducinganinfectioninananimalmodel
(14).AbriefreporthasindicatedthatsmallintestinalstimulicaninduceGiardiacystformationinvitro,butnoinformationwasprovidedoneithertheultrastructural
appearanceofthesecysts,ortheirviability(Gillinetal.,35thAnnualMeetingoftheAm.Soc.Trop.Med.Hyg.,abstractno.37,1986).Theuseoffluorogenicdyes
todeterminetheviabilityofG.muriscystshasbeencorrelatedwithboththeirmorphologyandtheirabilitytoproduceinfectioninananimalmodel(15).Aspresented
here,theincorporationofFDAbyGiardiacystsformedinvitrohasprovidedthefirstphysiologicalevidencefortheviabilityofcystsproducedinculture.Wehave
alsorecentlydemonstratedthatGiardiacysts,formedinvitro,werecapableofundergoingexcystationandwereabletoproduceinfectioninananimalmodel(17).
Therefore,Giardiacystsformedinvitrohavefulfilledallofthecriteriacurrentlybeingusedtomeasurecystviability.
OurstudieshavebeenthefirsttodemonstratetheviabilityofGiardiacystsformedinvitro.Thephysiologicalevidenceobtainedusingtheincorporationofthe
fluorogenicdye,FDA,togetherwiththeexcellentmorphologicalappearanceofthecystsproduced,invitro,hasdemonstratedthatthelifecycleofGiardiacould
nowbecompletedoutsideoftheanimalhost.ThedevelopmentofthisinvitromodelofthelifecycleofGiardiashouldfacilitatethetestingofKochs'postulates,
sinceitwillnowbepossiblenotonlytoisolatetrophozoitesfromananimalandgrowtheminculture,butalsotoproducecystsinvitro,whichcouldbeusedto
reinfectthesame(orpossiblyadifferent)typeofhost.
Acknowledgements
TheauthorswishtothankDr.W.J.BemrickforhisreviewofthemanuscriptandMs.LeeAnnSherlockforherexcellenttechnicalassistance.Althoughtheresearch
describedinthisarticlehasbeenfundedinpartbytheU.S.EnvironmentalProtectionAgencythroughcooperativeagreement#CR811834totheUniversityof
Minnesota,itnecessarilydoesnotreflecttheviewoftheAgency,andsoofficialendorsementshouldnotbeinferred.Mentionoftradenamesormaterialproducts
doesnotconstituteendorsementorrecommendationforuse.
LiteratureCited
1.Anderson,T.F.1951.Techniquesforthepreservationofthreedimensionalstructureinpreparingspecimensfortheelectronmicroscope.Trans.N.Y.Acad.Sci.
13:130134.3.
2.Bingham,A.K.,Jarroll,E.L.,Meyer,E.A,andS.Radulescu.1979.Giardiasp.:physicalfactorsofexcystationinvitro,andexcystationvs.eosinexclusionas
determinantsofviability.Exp.Parasitol.47:281291.
3.Coggins,J.R.andF.W.Schaefer.1986.Giardiamuris:ultrastructuralanalysisoftheinvitroexcystation.Exper.Parasit.61:219228.
4.Feely,D.E.,Erlandsen,S.L.,andD.G.Chase.1984Structureofthetrophozoiteandcyst.In:GiardiaandGiardiasis:Biology,Pathogenesis,and
Epidemiology(Erlandsen,S.L.,andMeyer,E.A.,eds.)PlenumPress,NewYork,NY,pp.331.
5.Keister,D.B.1983.AxeniccultureofGiardialambliainTYIS33mediumsupplementedwithbile.Trans.R.Soc.Trop.Med.Hyg.77:487488.
6.Knight,D.P.1977.Cytologicalstainingmethodsinelectronmicroscopy.In:Lewis,P.R.,andKnight,D.P.,eds.,StainingMethodsforSectionedMaterials
NorthHolland,Amsterdam,pp.2976.
7.Kulda,J.,andE.Nohynkova.1978Flagellatesofthehumanintestineandofintestinesofotherspecies.In:ParasiticProtozoa(Kreier,J.P.,ed.)AcademicPress,
NewYork,NY,pp.69104.
8.Lavier,G.1942.LesModalitesdel'enkystementchezlesflagellesdugenreGiardia.Soc.Biol.Paris136:6770.
9.Luchtel,D.L.,Lawrence,W.P.,andF.B.DeWalle.1980.ElectronmicroscopyofGiardialambliacysts.Appl.Environ.Micro.40:821832.

Page32

10.Luft,J.H.,1961.Improvementsinepoxyresinembeddingmethods.J.BiophysicalandBiochemicalCytology9:409413.
11.Meyer,E.A.1976.Giardialamblia:isolationandaxenicculture.Exp.Parasitol.2:187196.
12.Owen,R.L.1980TheultrastructuralbasisofGiardiafunction.Trans.R.Soc.Trop.Med.Hyg.74:429433.
13.Perroncito,E.1888.Notesurl'enkystementduMegastomaintestinale.Bull.Soc.Zool.Fr.13:1618.
14.RobertsThomson,I.C.,Stevens,D.P.,Mahmoud,A.A.F.,andK.S.Warren.1976Giardiasisinthemouse:ananimalmodel.Gastroenterology71:5761.
15.Schupp,D.G.andS.L.Erlandsen.1987.AnewmethodtodetermineGiardiacystviability:correlationbetweenfluoresceindiacetate/propidiumiodidestaining
andanimalinfectivity.Appl.Envir.Micro.53:704707.
16.Schupp,D.G.andS.L.Erlandsen.1987.DeterminationofGiardiamuriscystviabilitybydifferentialinterferencecontrast,phase,orbrightfieldmicroscopy.J.
Parasit.73:723729.
17.Schupp,D.G.,Januschka,M.M.,Sherlock,L.A.F.,Stibbs,H.H.,Meyer,E.A.,Bemrick,W.J.andS.L.Erlandsen.1988.ProductionofviableGiardiacystsin
vitro:Determinationbyfluorogenicdyestaining,excystation,andanimalinfectivityinthemouseandMongoliangerbil.Gastroenterology,Inpress.
18.Sheffield,H.1979.TheultrastructuralaspectsofGiardia.In:Jakubowski,W.andHoff,J.C.,eds,WaterborneTransmissionofGiardia.U.S.Environmental
ProtectionAgency,Cincinnati,OH,EPA600/979001pp.921.

Page33

CytopathogenicityofGiardialambliainHeLaandVeroCellMonolayers
A.Jyothisri*andUshaK.Baveja
DepartmentofMicrobiology,MaulanaAzadMedicalCollege,
NewDelhi110001,India
ThebehaviourandpathogeniceffectsoffourhumanstrainsofGiardialamblia(threelocallydesignatedND1,ND2,ND3andPortland1),werestudiedinHeLaandVerocellculture
systemsusinglightandscanningelectronmicroscopy.Allthestrainstestedhaveshownthesamedegreeofinitialattachmenttothetargetcellandhaveproducedcytopathogenic
effectsinbothcellculturesystems.Afterinoculationofparasitesmorphologicalchangesappeared4hourslaterinHeLacellsand6hourslaterinVerocells.Changesin
cytoplasmicgranulation,vacuolationandnuclearmarginationwereapparentby24hoursinHeLaandVerocells.Thenumberofcellsshowingthesechangesincreased
progressivelywithtime.By96hours,themonolayerofHeLacellsandby87hours,themonolayerofVerocellswascompletelydestroyedandreplacedbyG.lambliatrophozoites.
Significantchangesinthesurfacemorphologyofcellsafterinteractionwithtrophozoiteswereexaminedbyscanningelectronmicroscopy.Cellsthatwerenotindirectcontactwith
thetrophozoitesshowedlittledamagecomparedtothosecellsindirectcontactwithtrophozoites.CellfreeextractsandculturesupernatantsofG.lambliaproducedno
morphologicalchangesineitherHeLaorVerocellsinvitro.Theseobservationssuggestthatdirectcelltoparasitecontactmayplayanimportantroleindamagingthehostcells.

Introduction
Giardialambliaisanestablishedhumanpathogencausingawidespectrumofclinicalillnessvaryingfromanasymptomaticcarrierstatetoacutediarrheaand
malabsorption(1,15).Anumberofdifferentmechanismshavebeenproposedtoexplaintheintestinaldysfunctioningiardiasis.Theseincludethepresenceofabarrier
toabsorption,epithelialcelldamage,inflammatoryreactiontoparasites,bilesaltdeconjugationandmucosalinvasionbytrophozoites(2,10,13).Noneofthese
mechanismspreciselyexplainshowG.lambliaproducessymptomsinonlyafractionofinfectedindividuals.TheaxeniccultivationofG.lambliahasmadeitpossible
tocarryoutinvitrostudiestounderstandthemechanismbywhichtheparasitemaybeproducingitsdeleteriouseffectsanddisease.Sofar,twosuchstudieshave
beenconducted.Onereportedthesignificanceofbilesaltdeconjugationbytheparasite(12)andtheotherthecytopathogeniceffectsofGiardiaoncellmonolayers,
invitro(11).
ThepresentstudywasundertakentofindouttheprobablemechanismbywhichthedifferentstrainsofG.lambliadamagethecells(epithelialcellsandfibroblasts)in
vitroandtoassesswhethersimilarmechanismsoperateinhumandiarrheaandmalabsorption.Thisstudyreportsthemorphologicalchangesinducedbyfourdifferent
humanstrainsofG.lambliaonHeLaandVerocells.
MaterialsandMethods
GiardialambliaCulture
FourhumanstrainsofG.lambliaND1,ND2,ND3(3)andPortland1ACTCNo.30888(9)weremaintainedinDiamond'sTPS1(4)completefiltersterilized
mediumaccordingtothemethodofVisvesvara(14).
PreparationofDifferentInocula
1).EachofthefourstrainsofG.lambliatrophozoiteswerepreparedbychillingtubeswith48holdculturesinanicewaterbathandcentrifugingat600gfor10
minutesat4C.ThetrophozoiteswereresuspendedinTPS1medium,soastogetcountsof1104to1107trophozoites/mL.
2).Cellfreeculturesupernatantwascollectedfrom48holdculturesafterchillingandcentrifugationasdescribedabove.
3).Cellfreeextractswerepreparedfromactivelymultiplyingtrophozoitesatconcentrationsrangingfrom1104to1107trophozoitesaccordingtothemethodof
Lushbaughetal.(8).Afterwashingwithphosphatebufferedsaline,trophozoiteswererupturedbyafreeze/thawcycleandthencentrifuged(105gfor1h)at4C.
Thesupernatants(cellfreeextract)wereusedtoassaythecytotoxicity.
4).FreshTPS1completemediumwithouttheparasiteandminimalessentialmedium(MEM)with5%ofcalfserumwereusedascontrols.
CultivationandPreparationofHeLaandVeroCellMonolayers
HeLaandVerocellsweremaintainedinDulbecco'sMEMsupplementedwith5%adultcalfserumandantibiotics(PenicillinG100units/mLandStreptomycin100
g/mL).Trypsinizedmonolayers(0.01%Trypsinand0.05%EDTA)wereharvestedandadjustedtoaconcentrationof1.5105cells/mLtoinitiatetheculturesfor
cellparasiteinteractions.
SelectionofInoculumSize
Beforestartingthemainexperimentalseries,itwasnecessarytodeterminetheoptimalnumberofG.lambliatrophozoitesthatwouldproduceoptimalresultsfora
properandmeaningfulanalysisafterinteractionsbetweentheparasiteandcellculture.Forthispurpose,effectsofinoculaofvaryingnumberoftrophozoitesranging
from1104to1107onHeLaandVerocellsatdifferentintervals(0,5,18,24,48,72and96h)afterincubationat37Cwerestudied.
MethodsofStudyofCellParasiteInteractions
1).PreparationofCellMonolayersonCoverslips
Tissueculturecells(HeLaandVero)atadensityof1.5105cells/mL
Correspondingauthor.

Page34

werepipettedintoLeightontubescontainingcoverslipscloselyappliedtotheflatsurfaceofthetubes(7).Thecellsformedmonolayersoverthecoverslipafter
incubationat37Cfor48h.TheMEMwasgentlydecantedbeforethestartoftheexperimentandreplacedwithoneoftheexperimentalorcontrolculturesystems
describedaboveaccordingtothemethodofRadulescu(11).Alltheinoculatedtubeswereincubatedat37Cinaslantedposition.Thesewereexamineddailyfor
attachmentanddetachmentofparasitestothesurfaceofcells,andothereventsoccurringuptoaperiodof96hwerenoticed.Inalltheseries,theexperimentswere
terminatedatdifferentintervals(05,18,24,48,72and96h)afterincubationandthecoverslips(2foreachintervalineachseriesforthetwocellsystemsstudied),
wereremovedfromthetubesandprocessedforlightandelectronmicroscopicstudiesasdescribedbelow.
2).LightMicroscopy
Aftertherespectiveinteractionperiodtheusedmediumwasdecantedoffandcoverslipswerewashedgentlywithwarm(37C)PBSinsitu.Afterfixingwith
methanolandstainingwithGiemsa,coverslipswereremovedfromthetubes,airdried,placedmonolayersidedownonadropofmountingmediumDPXonaglass
microslide.ThesecoverslipswereexaminedwithaZeissmicroscope(W.GermanyHBO100w/2)equippedwitha35mmcamera(winderM,470799901
model).Atleast10representativefieldsoneachcoverslipwerestudiedtocomputetheresults.Representativephotographsweretakenatdifferentintervalsafter
interaction.
3).ScanningElectronMicroscopy(SEM)
TopreparethespecimensforSEM,coverslipswiththecellmonolayerafterinteractionswerefixedwith0.2Msodiumcacodylatebuffer.Themonolayerswere
dehydratedthroughagradedseriesofethanol.Theexposuretimeforeachconcentrationofethanolwas5minutes.Thespecimensweredriedatthecriticalpointina
PolaronapparatuswithliquidCO2.Finally,thecoverslipsweremountedonaluminumstubsandshadowedwithsilverinanargonatmosphereusingPolarondriode
sputtercoaterE5000.ThespecimenscoatedwithsilverwereobservedinaScanningElectronMicroscope(Phillips500B)operatingat10to20kVand
photographsweretakenwith120mmcamera(Superlollex,Holland).
Results
SelectionofInoculum
Aninoculumconsistingof1105trophozoiteswasfoundtobeidealforaproperevaluationofprogressionofmorphologicalchangeswithconvenienttimelimits.The
changesoccurredtoorapidlywithahigherinoculum(1107trophozoites)andtooslowlywithalowerinoculum(1104trophozoites).Hence,aninoculumsizeof1
105trophozoiteswasusedinallthesubsequentexperiments.
LightMicroscopicStudies
1).AttachmentofG.lambliatotheMonolayers
Manytrophozoiteswerefoundtobecloselyattachedtotheexternalsurfacesoftheculturedcellsasearlyas15minutesafterincubationat37C.Thenumberof
attachedtrophozoitesincreasedwithtimeduringthefirstthreehoursandbythistimealmosthalftheculturecellswereindirectcontactwiththetrophozoites.Allthe
fourstrainsofG.lambliashowedasimilartendencytoattachtothesurfacesofHeLaandVerocells.Attachmentdidnotseemtobeprolongedorpermanent,asthe
trophozoitesattachedanddetachedfromthetargetcellinafractionofasecond.Thiswasobservedinthewetpreparationsduringexaminationbyphasecontrast
microscope.Sometrophozoiteswereswimmingfreelyinthemediumandshowedactivemovementanddivision.

Figure1.
Lightmicrographsshowingepithelialcells(HeLacellline)after
incubationwithG.lamblia.A).Control:24hincubationofHeLa
cellmonolayerwithoutG.lambliainTPS1medium
showingnormalmorphology.Giemsa400.Thetermh(hours)
indicatesthelengthofexposureofculturecellswithorwithout
G.lambliatrophozoites.B).24hepithelialcellsshowingvacuolation
inthecytoplasmandattachedparasites.Giemsa400.C).48hcytoplasm
ofcellsshowingextensivevacuolationandtheparasitesfilledin
thecelldenudedareas.Giemsa1000.D).72hcytoplasmofepithelial
cellsshowingvacuolationandnuclearabnormalitiesafter
progressiveincreaseintheparasitenumber.Giemsa400.

2).InteractionsBetweenParasitesandHeLaCells
MorphologicalchangesinHeLacellsduringtheinitialhoursofinteraction(4to18h)consistedofcytoplasmicvacuolation.Thesewereapparentby24hafter
incubation(Figure1B).Withfurtherincubation,cellfreeareasappearedinthemonolayersasaresultofdetachmentoftheparasitedamagedcells.At48hafter
interactionthemajorityofcellshadahighlyvacuolatedcytoplasm.Thenumberoftrophozoitesincreasedsteadilyastheinteractionperiodproceeded.Trophozoites
filledthecellfreeareas(Figure1C).By72hafterincubationalmostallthecellsshowedcytoplasmicvacuolation.Fewcellsshowedpyknoticnucleiandfewhad
multiplenuclei(Figure1D).Attheendoftheinteractionperiod(96h)thecellmonolayerwasreplacedcompletelybytheparasitemonolayer.
Identicalmorphologicalchangesofthesameintensitywereinducedbyallthe4strainsviz.ND1,ND2,ND3andPortland1oninteractedHeLacells.
ThemorphologicalchangesexhibitedbyVerocellsuponinteractionwithG.lambliatrophozoitesareshown(Figure2AF).Thesechangesconsistedofretractionof
cytoplasminsomecells(6h,Figure2B)lossofcelltocellcontactwithslightvacuolationofcytoplasm(24h,

Page35

Figure2.
Photomicrographsoffibroblastcells(Vero)afterincubationwith
G.lamblia.A).6hcontrolVerocellmonolayershowingthe
normalmorphology.Giemsa400.B).6hshowingearlysignsof
changesinappearanceofcellfreeareas.AttachedG.lambliaare
seen.Giemsa400.C).24hfibroblastcellsshowingslightvacuolation,
increaseincellfreeareas.Giemsa400.D).48hcontrolfibroblast
monolayer.Giemsa100.E).48hfibroblastcellsshowingloss
ofintercellularcontacts,multiplenucleiandabnormalcytoplasm.
Giemsa100.F).72hfibroblastcellsshowingexcessivevaculoation
inthecytoplasm,twonuclear,andattachedG.lamblia.Giemsa400.

Figure2C)whichgraduallyincreasedsothatby48hthefibroblastswerenarrow,elongatedandbranched(Figure2E).Extensivevacuolationofcytoplasm,nuclear
pyknosisandmultiplenucleiwereobserved72hafterinteraction(Figure2F).Thetrophozoitesmultipliedandincreasedinnumberprogressivelyuntilthewholecell
monolayerwasreplacedbyaG.lambliamonolayer(87h).
MorphologicalchangesinVerocellmonolayersofthefourhumanstrainsofG.lambliaweresimilar.
3).InteractionBetweenVeroandHeLaCellsandtheCultureSupernatant
NomorphologicalchangeswereseenineitherHeLaorVerocellsfromeachofthefourstrainsuponincubationwiththeculturesupernatantuptoaperiodof96h
comparedtothematchedcontrols.
4).InteractionBetweenVeroandHeLaCellsandtheCellFreeExtract
CellfreeG.lambliaextractsapparentlyhadnodeleteriouseffectsanddidnotinduceanymorphologicalchangescomparedwithTPS1andMEMcontrolculture
cellsineitherHeLaorVerocellstill96hafterinteraction.
ScanningElectronMicroscopy(SEM)
Thesurfacechangesseeninfibroblastcells(Verocells)atdifferentintervals(0,6,24,48,72h)afterinteractionwithG.lambliaareshowninFigure3(AE).
Thefirstinteractionwastheadhesionoftrophozoitestothecellsurfacefollowedbyprogressivedestructivechangesinthecell.
Anumberoftrophozoiteswereattachedtothecellsurfacesby6hafterinteraction.Inthesetrophozoitesthedorsalsurfaceandfourpairsofflagellawerenoticeable
(Figure3B).Theventralsurfacewasinclosecontactwiththecellsurface.Onfurtherinteractionfibroblastslostcelltocellcontactandrufflingappearedonthesurface
(24h,Figure3C).Someofthetrophozoiteswereseentobeattachedtothecoverslip.Thelateralshieldalongtheperipheryofthetrophozoitesandtheventrolateral
flangecouldbeseenindirectcontactwithboththecellsurfaceinsomeandthecoverglassinothers.
Filamentousprocessesextendingfromthecellbordersandincreasedrufflingofsurfaceswasseen48hafterinteraction(Figure3D).Trophozoiteswereorientedat
randomandwereattachedtoanypartofthecellsurface.Thesechangesprogressivelyincreasedastheinteractionsprogressed(Figure3E)untilthedegeneratedand
damagedcellsdetachedfromthesurfaceofthecoverslip.
Discussion
ThepredominanceofmorphologicalchangesinthoseVeroandHeLacellsthatwereincontactwiththetrophozoitesemphasizesthedirectdamagingeffectsofthe
parasite.CellfreeextractsandculturesupernatantsfromGiardiaculturehadnodeleteriouseffectoneitherVeroorHeLacells.Similarfindingshavebeenreported
earlieralso(11).AllthefourstrainsofG.lambliainducedsimilarmorphologicalchangesinVeroandHeLacells.Thismaybebecauseallfourstrainswereisolated
fromcasesofsymptomaticgiardiasis(3,9).Hence,thecapacityofastraintoproduceCPEintissueculturesystemmayrelatetoitspathogenicityinvitro.
ThemorphologicalchangesobservedinparasiteinteractedcellsweresimilartothosereportedbyRadulescuetal.(11).Scanningelectronmicroscopyrevealedthe
ventralsurface,lateralshieldandventrolateralflangeoftheparasitetobeinclosecontactwiththesurfaceofcells.Erlandsenetal.(5)reportedasimilarpatternof
attachmentofG.lambliatothemicrovillousborderoftheepithelialcells.Fibroblastsshowedrufflingofthesurfaceuponinteractionwithtrophozoites.However,the
cellmembraneofthecellssurroundedbyparasiteswascontinousindicatingthattheparasitedidnotpenetratethemembraneandproducedthecelldamagejustby
attachingtothesurfaceofthecells.
G.lambliamaybedamagingtheculturedcellseitherbymechanicalorchemicalinjuryduringclosecontactwiththecellorbyactivationandreleaseoffactors/toxins
aftercontactwiththecellresultingincelldamage.TheresultsofthepresentstudyindicatethatG.lambliamostprobablydoesnotproducetoxinsasreportedinother
protozoanslikeE.histolytica(8)andT.vaginalis(6).
Webelievethatparasitetocellcontactaloneisresponsibleformorphologicalchangesanddamageinepithelialandfibroblastcells.However,thesedataarenot
sufficienttodifferentiatebetweenthedamagecausedbychemicalandmechanicaleffectsoftheparasite.Futurestudiesonthelocalizationoftheparasiteenzymeson
theplasmamembraneandthemolecularmechanismsofthe

Page36

Figure3.
ScanningelectronmicrographsoffibroblastcellsafteradditionofG.lambliatrophozoites.A).6hscanningelectronmicrographofcontrolfibroblastcellmonolayer(Verocellline)having
wellspreadnormalmorphology(originalmagnification1250).B).6hadherenceofG.lambliatrophozoitestothefibroblasts(slightlyoutoffocus).Initialcontactbetweenventralsurface
oftheparasiteandexternalsurfaceofthefibroblastcell(originalmagnification2500).C).24hcellsshowingmorerufflesonthecellsurfacethanthenormalcellsandlossofcontactwith
adjacentcells.Thetrophozoitesattachedtothefibroblastcellsaswellasthecoverglassesareseen(originalmagnification1250).D).48hfibroblastcellsshowingextensiverufflingon
thesurfaceandcytoplasmicprocessesthatareindicativeoffurtherdamagecausedtofibroblastsbyG.lambliatrophozoites.Trophozoitesareindirectcontactwiththesurfaceof
fibroblasts(originalmagnification1250).E).72hfibroblastsshowingfurtherdamageduetoG.lambliatrophozoites.Branchedfilamentousprocessesandrufflingsonthesurfacecanbe
seenclearly(originalmagnification1250)withthesurfaceoffibroblasts(originalmagnification1250).

specificadherenceoftheventraldiscoftheparasitetothecellmembranewillbeimportantinpinpointingtheexactmechanismsofpathogenesis.Studieswithstrains
isolatedfromasymptomaticpatientswillalsobevaluableindistinguishingthedifferencesbetweenvirulentandavirulentstrains.OurresultsindicatedthatbothHeLa
andVerocelllinesaresusceptibletoG.lambliatrophozoites.Thisleadstothefactthatonecanuseavarietyofcellculturesystemstostudythehostparasite
interactionsincaseofG.lamblia.Thissimpleinvitroassaysystemisrapidandsensitivecomparedwithinvivosystemsforthestudyofpathogenesis.
Acknowledgements
Thisworkwassupportedbyaresearchgrant(70/288/85ECDII)fromtheIndianCouncilofMedicalResearch.
Literaturecited
1.Ament,M.E.,andC.E.Rubin.1972.Relationofgiardiasistoabnormalintestinalstructureandfunctioninimmunodeficiencysyndromes.Gastroenterol.62:216
226.
2.Anand,B.S.,R.Chaudhary,A.S.Jyothi,R.S.Yadav,andU.K.Baveja.1985.ExperimentalexaminationofthedirectdamagingeffectsofGiardialambliaon
intestinalmucosalscrapingsofmice.Trans.R.Soc.Trop.Med.Hyg.79:613617.
3.Baveja,U.K.,A.S.Jyothi,M.Kaur,D.S.Agarwal,B.S.Anand,andR.Nanda.1986.IsoenzymestudiesofGiardialambliaisolatedfromsymptomaticcases.
Aust.J.Exp.Biol.Med.Sci.64:119126.
4.Diamond,L.S.1968.TechniquesofaxeniccultivationofEntamoebahistolyticaSchaudinn,1903andE.histolyticalikeamoeba.J.Parasitol.54:10471056.
5.Erlandsen,S.L.,andD.E.Feely.1984.Trophozoitemotilityandthemechanismsofattachment,pp.3361,In:Erlandsen,S.L.andE.A.Meyer(eds.),Giardia
andgiardiasis:Biology,pathogenesisandepidemiology.Plenum,NewYork.
6.Farris,V.,andB.M.Honigberg.1970.BehaviourandcytopathogenicityofTrichomonasvaginalisDonnlivercultures.J.Parasitol.56:849882.

Page37

7.Koester,S.K.,andP.G.Engelkirk.1984.GlasscoversliptechniqueforstudyinginvitrointeractionsbetweenGiardiatrophozoitesandhostleukocytesbyTEM,
SEM,andlightmicroscopy.70:443445.
8.Lushbaugh,W.B.,A.F.Hofbauer,A.B.Kairalla,J.R.Cantey,andF.E.Pittman.1984.RelationshipofcytotoxinsofaxenicallycultivatedEntamoebahistolytica
tovirulence.Gastroenterol.86:14881495.
9.Meyer,E.A.1976.Giardialamblia:isolationandaxeniccultivation.Exp.Parasitol.39:101105.
10.Meyer,E.A.,andS.Radulescu.1979.Giardiaandgiardiasis.Adv.Parasitol.17:147.
11.Radulescu,S.,C.Rau,D.Petrasincy,N.Gaicu,andE.A.Meyer.1980.BehaviourandcytopathogenicityofGiardialambliaincellcultures.Arch.Roum.Path.
Exp.Microbiol.39:163170.
12.Smith,P.O.,C.R.Horsburgh,Jr.,andW.R.Brown.1981.invitrostudiesonbileaciddeconjugationandlipolysisinhibitionbyGiardialamblia.Dig.Dis.Sci.
26:700704.
13.Smith,P.D.1985.Pathophysiologyandimmunologyofgiardiasis.Ann.Rev.Med.36:295307.
14.Visvesvara,G.S.1980.AxenicgrowthofGiardialambliainDiamond'sTPS1medium.Trans.R.Soc.Trop.Med.Hyg.74:213215.
15.Wolfe,M.S.1978.Giardiasis.N.Engl.J.Med.298:319321.

Page39

StudiesonthePrevalenceofGiardiasisinCzechoslovakia
MichalGiboda
InstituteofParasitology,CzechoslovakAcademyofSciences,37005CeskeBudejoviceBraniovska31,Czechoslovakia.
Theprevalencerateofgiardiasiswithinaparticularagegroupisnotuniform.Amongchildrenofpreschoolage,thehighestprevalenceofgiardiasiswasfoundinchildren'shomes,
followedbychildrenwhoattenddaycarecentres.Giardiasisisleastcommonlyfoundinchildrenwhoremainathome.Thepermanentcontactbetweenthechildreninthe
children'shomecreatesbetterconditionsforGiardiatransmissionthanindaycarecentersandkindergartenswhereclosecontactoccursduringtheperiodoftheparents'working
day.ThevalidityofthetheoryoffamilialoccurrenceofgiardiasiswastestedbycarryingoutstudiesindifferentgeographicalregionsinCzechoslovakia.Theresultsshowedthatin
familieswithonereportedcaseofgiardiasis,theriskofotherfamilymembersacquiringtheinfectionwasbetween5.7and8.3timeshigherthanthatexperiencedbytherestofthe
generalpopulation.Theauthorsuggeststhatthetheoryoffamilialoccurrenceofgiardiasisisvaliduniversally.

Introduction
GiardialambliaishistoricallyimportantinCzechoslovakia.In1859theCzechphysicianDuanLambldescribedforthefirsttimeaparasitefoundinthestoolsofa
childsufferingfromdiarrhea.HenameditCercomonasintestinalis.Bytheendofthe1960'sandduringthe1970'sCzechoslovakparasitologistspaidmoreattention
totheprevalenceandepidemiologyofgiardiasis.TheprevalenceofgiardiasisinCzechoslovakiahasbeenreportedfromanumberofsources
(1,2,3,7,8,9,11,12,13,14,15,16)andthedataarereportedinTable1.
Theprevalenceratewithinaparticularagegroupisnotalwaysuniform.Theprevalencerateofgiardiasisislowamongchildrenofpreschoolagewhoremainathome.
Itisgreateramongchildrenwhoattenddaycarecentresandthehighestprevalenceofgiardiasisisusuallyfoundinchildren'shomes(institutions).Itisinterestingto
notethataspecialsectionoftheCzechoslovakpopulationisrepresentedbygypsies.Manyauthorshavestudiedgiardiasisingypsies(3,4,5,9)andtheyhavefound
thattherateofgiardiasisis4to6timeshigheramonggypsiesthaninsimilaragegroupswithinthesameregion.
Discussion
Whydoesthefrequencyofgiardiasisvarywithinthesameagegroupsofthepopulationwholiveindifferentsocialandepidemiologicalconditions?Pazdioraand
Palicka(11)statethatindaycarecenters(forchildrenupto3years)thedistributioncurveofgiardiasisreachesitspeakbythethirdyearoflife.Fortyninepercentof
childrenwhoattenddaycarecentersforlessthan1yeararereportedtobeinfectedcomparedwithonly12.2%ofchildrenwhohaveattendedformorethantwo
years.TheredoesnotappeartobesuchastrikingdifferenceintheprevalenceofGiardiainfectionbetweenchildren3to5yearsofagewhoattendpreschool
facilitiesandthoseattendingkindergarten(7.3%vs.10.1%onaverage).
TABLE1.PrevalenceofGiardialambliainCzechoslovakiafromsurveys19551984(%).

02years

Authors
Family
Cerva(1962)

Ditrichetal.
(1984)

6.6

Giboda(1978)

11.3

Kvasz(1979,
1980)

Moravec
(1980)

Daycare
Center Children'sHomes
20.0

Giboda(1971)

7.3

35years

43.0

Family

55.2

21.9

Pazdiora
(1972)

7.6

Vota(1955)
Zitek,Palicka
(1979)
Average

17.8

3.0

6.7

13.2

32.8

4.5

3.2

7.3

10.1

4.0
23.7

19.0

7.3

4.1

10.0

3.2

3.5

2.0

6.6

15.2

2.7

14.3

11.2

12.2

>15
years

Children's
homes

8.2

6.8

28.1

5.4

Family
school

12.7

1015years

Orphan
school

12.1
29.1

11.3
1.8

7.5

24.1

5.2

6.5
16.7

15.7

35.6

9.1

5.5

Volna,
Amera(1968)

11.6

5.2

8.3

Pazdiora,
Palicka(1971)

kracikovaet
al.(1981)

Family
Kindergartens Children'shomes school

27.3
15.0

69years

3.7

6.8

2.5
10.0

3.2

Page40
TABLE2.EffectivenessofthesearchforfurtherinfectionwithGiardiaamongfamilymembersinthemicrofociofgiardiasis
(Palicka1973).
Frequencyin%
Giardia
positivefamilies
examinedmicro
foci

Examined
membersof
families

107

394

Familymembers
withsecondary withoutprimary
Giardia
giardiasis
infections
59

withprimary
Giardia
infections
(n=107)

15

prevalenceof
giardiasisin
population
(%)

Index
prevalencein
microfoci
prevalencein
population

42

1.8

8.3

Childrenininstitutionsrepresentaspecialcase.EveryauthorwhostudiedtheprevalenceofGiardiaintheseinstitutionsfoundthatchildrenofeveryagecategory
sufferedfromgiardiasismorefrequentlythanthoseatdaycarecentres(Table1).Thepermanentcontactbetweenthechildreninthesehomescreatesbetter
conditionsforGiardiatransmissionthanindaycarecentersandkindergartenswhereclosecontactisrestrictedtotheperiodoftheparents'workingday.Thisisalso
thecasewithotherintestinalparasitessuchPentatrichomonashominis,ChilomastixmesniliandHymenolepisnanawhicharecommonlytransmittedbythefecal
oralmodeoftransmission(3).IntheirstudyofGiardiatransmissioninpreschoolfacilities,PazdioraandPalicka(1971)examinedthestaffoftheseestablishmentsfor
intestinalparasites.Theprevalencerateofgiardiasiscorrespondedtotheaveragerateamongtheadultpopulationintheregionstudied.Thissuggeststhatthestaffof
preschoolfacilitiesdonotplayanactivepartinspreadingGiardiaamongthechildrenalthoughtheycaninfluencetransmissionthroughfoodhandling,changing
diapersetc.
Kvasz(6)examinedallthemembersoffamiliesofchildrenwhowereinfectedwithGiardia,bothfrompreschoolfacilitiesandhospitalizedpatients.Heexamined790
membersof142familiesandfoundthattheaverageprevalencerateinthosefamilieswas23.07%.InanotherregionPazdiora(12)reportedaGiardiafrequencyof
9.77%infamiliesofGiardiapositivechildren.InfamilieswithnoninfectedchildrenthefrequencyofGiardiastoodatonly3.21%.
TheeffectivenessofantiepidemicmeasuresamonginfectedfamilieswasstudiedbyPalicka(10)whoseresultsarereproducedinTable2.Heexamined394
membersof107families(microfoci)whowerefoundtohaveatleastoneinfectedmember.Additionalcasesofgiardiasiswerediscoveredin15%oftheirfamily
members.When107originalinfectionswereadded,thefrequencyofgiardiasisincreasedinsuchfamiliesto42%.Sincetheaveragefrequencyofgiardiasisintherest
ofthepopulationwas1.8%,thisrepresenteda23timeshigherfrequency.TheriskofbecominginfectedwithGiardiainfamilieswithevenasinglecaseofgiardiasis
wascomputedtobe8.3timeshigher(15/1.8%)thanthatexperiencedbythegeneralpopulation.
Giboda(4)testedthevalidityofthetheoryoffamilialoccurrenceofgiardiasis.Thestudywascarriedoutindifferentgeographicalandepidemiologicalconditionsfrom
thoseoftheauthorsmentionedabove.ThedataarereportedinTable3.Of157membersof44familiesinwhichonechildwasinfectedwithgiardiasis,newgiardiasis
wasdiscoveredin31personsfrom25families(56.8%).Thismeansthattheprevalencerateamongthemembersofaninfectedfamilywas19.74%.Amongnewly
discoveredinfectionsadultsweremorecommonlyinfectedthanchildren(20.8%vs.17.6%).Inacontrolgroup(familieswithoutprimarygiardiasis)of151members
of35families,newgiardiasiswasdetectedinthreeindividualsonlywhobelongedtothreefamilies.Theprevalencerateinthenoninfectedfamilieswasthereforeonly
1.91%.AcomparisonoftheaverageprevalencerateofGiardiainchildrenupto15yearsofageinthestudyregion(9.8%)withtheprevalenceofGiardiain
childrenfrominfectedfamilies(primaryplussecondaryinfections55.8%)resultedinanindexof5.7(55.8/9.8=5.7).Thisindexwaslowerthantheonefoundin
Palicka'sstudy(8.3).Thefrequencyofgiardiasisinbothstudiesamonginfectedfamilieswassimilar(Palicka33.1%Giboda37.3%)asweretheoverallprevalence
ratesofgiardiasisinbothregionsofCzechoslovakia.
TheseresultshavealreadyhadaninfluenceonthepracticalactivityofthePublicHealthServiceinCzechoslovakia.IntheeventofGiardiaandgeohelminths
diagnosis,allmembersofthefamilyareparasitologicallyexaminedaswell.Thisantiepidemicmeasurementishighlyeffective,especiallyinregionswithlowprevalence
ofintestinalparasites.
TABLE3.GiardiainfectionsinfamilymembersofGiardiapositiveandGiardianegativechildren.
Infectedchild
inhouse

Giardia()
adult

children

Giardia(+)
total

adult

children

totalexamined
total

adult

total

84

42

126

22

31

106

51

157*

no

77

71

148

79

72

151**

*Represents44households25(57%)ofhouseholdshadasecondcaseattributedtoindexcase
**Represents35households

children

yes

Page41

LiteratureCited
1.Cerva,L.1962.OccurrenceofintestinalparasitesinthepopulationofCentralBohemia.(InCzech,Englishsummary).Cs.Parasitol.IX:135141.
2.Ditrich,O.,J.terba,J.Prokopic,K.Kadlcik,andI.Maleckova.1984.IntestinalparasitosesinSouthBohemiafarmer.(InCzech).IV.Prowazkovydny,
Komarno,4.5.10.1984.
3.Giboda,M.1971.TheproblemofintestinalparasitesespecialProtozoainEastSlovakia.(InSlovak).ThesistoB.Sc.
4.Giboda,M.1978.ConditionsofoccurenceofGiardiasisandascariasisinchildrenpopulationofEastSlovakia.(InSlovak).ThesistoPh.D.
5.Jecny,V.1965.TheresultsofexaminationonintestinalparasitesinsomegroupsofpopulationindistrictMost.(InCzech,Englishsummary).Cs.Parasitol.
XII:185195.
6.Kvasz,L.1972.ContributiontolambliasisinSlovakia.(InSlovak).ThesistoPh.D.
7.Kvasz,L.1979.AccumulationofGiardiasisinfamiliesandclosedcollectives.(InSlovak,Englishsummary).Bratisl.Lek.Listy72:597600.
8.Kvasz,L.,B.Petranska,M.Pavlina,A.Halasova,andJ.Vodrazka.1986.Screeningofparasitesofgastrointestinaltractinemploysoflargescaleanimalfarms
andthemeatconcernintheNitradistrict.(InSlovak,Englishsummary).Cs.Epidem.35:5054.
9.Moravec,P.1980.PrevalenceofintestinalparasitesinpopulationofdistrictOpava.(InCzech).Cas.Slez.muz.Opava(A)29:5764.
10.Palicka,P.1973.Effectivityofepidemiologicalworkinthefociofintestinalparasitosis.(InCzech,Englishsummary).Cs.Epidem.22:3944.
11.Pazdiora,E.,andP.Palicka.1971.Notestoepidemiologyofsomeintestinalparasitoses.(InCzech,Englishsummary).Cs.Epidem.20:216220.
12.Pazdiora,E.1972.Someepidemiologicalaspectsofoccurrenceoflambliasisincreche.(InCzech,Englishsummary).Cs.Epidem.21:271276.
13.kracikova,J.,S.Straka,E.Galikova,andG.Klimentova.1981.FamilialincidenceofGiardiasis.Bratisl.Lek.Listy76:369373.
14.Volna,L.,andJ.Amera.1968.TheoccurrenceofparasitesatthepopulationoftheOstravaregion.(InCzech,Englishsummary).Prirodoved.Sborn.
(Ostrava):179183.
15.Vota,J.1955.IntestinalparasitesofchildreninthesurroundingsofTabor(InCzech).Cs.Parasitol.II:177180.
16.Zitek,K.,andP.Palicka.1979.Incidenceofintestinalparasitesinthepopulationofacommunityandscopeforinfluencingit.(InCzech,Englishsummary).Cas.
Lek.ces.118:447450.

Page43

IMMUNOLOGY

Page45

ImmunologyofGiardiaInfections
MartinF.Heyworth
IntestinalImmunologyResearchCenter,CellBiologySection,VeteransAdministrationMedicalCenter,SanFrancisco,California94121andDepartmentof
Medicine,UniversityofCalifornia,SanFrancisco.
ManystudieshaveshownthathumansubjectsinfectedwithGiardialamblia,andmiceinfectedwithGiardiamuris,developantibodyresponsestoGiardiatrophozoites.Thepresent
authorhasshownthatimmunocompetentBALB/cmiceproduceintestinalIgAandIgGantitrophozoiteantibodiesduringG.murisinfection.Suchmiceeliminatetheinfection.In
contrast,athymic(nude)micedonotclearG.murisinfection,andshowlittleevidenceofanintestinalantibodyresponsetoGiardiatrophozoites.Theseobservationssuggestthat
antibodiesplayanimportantpartinclearanceofG.murisinfection.ExperimentsinwhichBALB/cmicewereselectivelydepletedofeitherhelper/inducer(Th/i)or
cytotoxic/suppressor(Tc/s)Tlymphocytesbytreatmentwithmonoclonalantibodies,haveshownthatTh/ilymphocytesarenecessaryforclearanceofG.murisbyinfectedmice.Tc/s
lymphocytesandnaturalkillercellsarenotrequiredforeliminationofthisparasitefromthemouseintestine.Importantareasoffuturestudyincludethefollowing:(i)todetermine
whetherintestinalantibodiesarecytotoxictoGiardiatrophozoites,and(ii)toidentifyandcharacterizetrophozoiteantigenswhicharemajortargetsformouseandhuman
intestinalantiGiardiaantibodies.

Introduction
HumansubjectsbecomeinfectedwiththeintestinalprotozoanparasiteGiardialambliabyingestingGiardiacysts.Thesecanbeacquiredbydrinkingcyst
contaminatedwater(11,42,49),orbyfecal/oralcontact,asininfantdaycarecenters(4,38).G.lambliaisanimportantcauseofdiarrheainimmunologicallynormal
individuals(51),andpatientswithimmunodeficiencydiseases(particularlycommonvariablehypogammaglobulinemiaandXlinkedimmunoglobulindeficiency)show
increasedsusceptibilitytogiardiasis(18,31).Suchimmunodeficiencydiseasespredisposetochronicgiardiasis,whichcanleadtosevere,persistentdiarrheaand
malabsorption(18,31).TheassociationofchronicgiardiasiswithimmunodeficiencydiseasessuggeststhatimmunologicalprocessesareresponsibleforclearingG.
lambliainfectionsinimmunologicallynormalindividuals.ThissuggestionisstrengthenedbythedemonstrationofantiGiardiaantibodiesinimmunologicallynormal
humansubjects(16,48).TheseantibodiesincludeIgGantitrophozoiteantibodieswhichoccurinhumansera(43),IgMantitrophozoiteantibodieswhicharepresent
inpatients'seraduringG.lambliainfection(17),andIgAantitrophozoiteantibodiesfoundinhumanmilk(33).Thefunctionalsignificanceoftheseantibodiesis,
however,unknown.Furthermore,verylittleisknownabouthumanimmunologicalresponsestoG.lambliaatthesiteoftheinfection,namelyinthegastrointestinal
tract,althoughIgAhasbeendemonstratedonGiardiatrophozoitespresentontheepithelialsurfaceofhumanjejunalbiopsyspecimens(6).
Studyofthepathophysiologyofgiardiasis,andoftheimmunologicalresponsetoGiardiatrophozoites,hasbeenfacilitatedbythedevelopmentofamousemodelof
giardiasis.Inthismodelsystem,miceareinfectedwiththeintestinalparasiteGiardiamuris(5,40).ByanalogywithG.lambliainfectioninhumansubjects,G.muris
trophozoitescolonizethemousesmallintestine(5,15,37).Inimmunocompetentmice,G.murisinfectionlastsforseveralweeksandtheparasitesarethencleared
fromthegastrointestinaltract(5,40).MicewithvarioustypesofimmunodeficiencyhaveanimpairedabilitytoclearG.murisinfection.Theinfectionischronicin
athymic(nude)mice(39,46),inmicetreatedfrombirthwithanantiserumdirectedagainstmouseIgM(44suchmicearedeficientinIgM,IgA,andIgG),andinmice
depletedofhelper/inducerTlymphocytes(20).SuchobservationsindicatethatimmunologicaleventsplayanimportantpartintheclearanceofG.murisinfection.
ProductionandRoleofAntiGiardiaAntibodies
ImmunocompetentmicewhichareinfectedwithG.murisproduceantibodiesdirectedagainstGiardiatrophozoites.Suchantibodieshavebeendemonstratedinthe
serum,milk,andintestinalsecretionsofG.murisinfectedmice(1,2,24,44,45).ThereisevidencethatantibodiesdirectedagainstG.muristrophozoitesplayapartin
theclearanceofG.murisinfection.Thus,inmicetreatedwithrabbitantiserumdirectedagainstmouseIgM,G.murisinfectionispersistent,andantibodyproduction
againstG.muristrophozoitesisimpairedasjudgedbytitersoftrophozoitespecificantibodyintheserumandintestinalsecretionsoftreatedmice(44).Thepresent
authorhasshownthatIgAandIgGbecomeboundtoG.muristrophozoitesintheintestinallumenofGiardiainfectedimmunocompetentBALB/cmice,fromday10
ofG.murisinfectiononwards(19).ThereislittleevidenceofIgAorIgGontrophozoitesharvestedfromimmunocompetentmicelessthan10daysafterthestartof
Giardiainfection,suggestingthatimmunoglobulinsdetected

Page46

ontrophozoiteslaterintheinfectionareGiardiaspecificantibodymolecules,ratherthanimmunoglobulinsthatarenonspecificallyadsorbedtothetrophozoites.There
islittleifanyIgAorIgGontrophozoitesharvestedfromtheintestineofnudemice,atanytimeafterthestartofGiardiainfection,suggestingthatthesemicehavean
impairedantibodyresponsetoGiardiatrophozoitesintheintestinallumen(19).Thisimpairmentofantibodyproductionmayexplainwhynudemiceareunableto
clearG.murisinfection.
IthasbeenshownthatintroductionofG.lambliatrophozoitesintotheduodenallumenofratsleadstotheappearanceoftrophozoitespecificIgAinratbile(30).
AlthoughthesourceofthisIgAisunknown,itmayincludeIgAtransportedfromserumtobileviatheratliver(35).
ItislikelythatmuchoftheintestinaltrophozoitespecificIgAproducedinrodentswithGiardiainfectionarisesfromplasmacellsintheintestinalmucosa.Numerous
plasmacellsarepresentinthelaminapropriaoftheintestineinvariousmammalianspecies(9,10),andthesecellsarebelievedtooriginatefromBlymphocytesin
Peyer'spatches(3).Carlsonetal(7)haveshownthatthenumberofIgA+cellsinPeyer'spatchesofimmunocompetentmicewithG.murisinfectionincreasesbefore
theinfectioniscleared.ItisjustifiabletospeculatethatthePeyer'spatchIgA+cellswhichincreaseinnumberduringG.murisinfectionareprecursorsofintestinal
mucosalplasmacellsthatsecretetrophozoitespecificIgA.
AlthoughitisprobablethattrophozoitespecificantibodycontributestotheclearanceofGiardiainfections,littleisknownaboutthemechanismsbywhichantibodies
mayeliminatetrophozoitesfromtheintestine.OnetheoreticalpossibilityisthatantibodiesmightinhibitadherenceofGiardiatrophozoitestotheluminalsurfaceof
intestinalepithelialcells.ThereisrecentevidencethatGiardiatrophozoiteshavealectinlikesurfacemoleculebywhichtheybindtocarbohydrateresidueson
mammaliancellmembranes(14,29).Thislectinmayfacilitatetrophozoiteattachmenttotheluminalsurfaceofintestinalepithelialcellsinvivo.Ifantibodiesaredirected
againstthelectin,theymightinhibitthisattachment.Similarly,antibodiesdirectedagainstcomponentsofthetrophozoiteadhesivediskmightimpairattachmentofthe
parasitestotheintestinalepithelium(12,30,47).IthasbeenshownthatrabbitserumandmousemilkwhichcontainantibodiesdirectedagainstG.muristrophozoites
inhibitadherenceoftrophozoitestomouseintestinalvilli(23).
Anotherpossibility,thatwarrantsinvestigation,isthatantitrophozoiteantibodymayactuallykillGiardiatrophozoitesintheintestinallumen.Ithasbeenshownthat
monoclonalantibodiesdirectedagainstatrophozoitesurfaceantigenareabletokillG.lambliatrophozoitesinvitro(34).Importantareasoffuturestudyincludethe
following:(i)todeterminewhetherintestinalsecretionsandserum,fromimmunocompetentmicewhichhaverecentlyclearedG.murisinfection,containantibodiesthat
arecytotoxictoG.muristrophozoitesinvitro,and(ii)todeterminewhetherantibodycoatedG.muristrophozoitesharvestedfromtheintestinallumenof
immunocompetentmice(19)areviableornonviable.
RolesofTLymphocytes,NaturalKillerCells,andMacrophages
Asnotedabove,nudemicelacktheabilitytoclearG.murisinfection,andbecomechronicallyinfectedwithGiardiatrophozoites(39,46).Thisobservationindicates
thatTlymphocytesplayanimportantpartinclearanceofG.murisinfectionfromthegastrointestinaltractofimmunocompetentmice,butdoesnotidentifytheTcell
subsetthatisinvolvedinclearance.Eitherhelper/inducer(Th/i)orcytotoxic/suppressor(Tc/s)Tlymphocytes,orconceivablybothofthesesubsets,mightbeimportant.
ToidentifytheTcellsubpopulationthatplaysamajorpartineliminationofG.murisinfection,thepresentauthortreatedimmunocompetentBALB/cmicewith
monoclonalantibodydirectedagainsteitherthemousehelper/inducerTcellantigenL3T4(13)orthecytotoxic/suppressorTcellantigenLy2(27).Thismaneuver
depletesL3T4+orLy2+lymphocytesrespectively(28,50).TheTh/idepletedandTc/sdepletedmiceweretheninfectedwithG.muriscysts,andthetimecourseof
theinfectionwascomparedinthesetwogroupsofanimals.ItwasfoundthatTc/sdepletedmiceclearedtheinfectionatthesamerateasimmunologicallynormalmice
thatweretreatedwithphosphatebufferedsaline.Incontrast,micedepletedofTh/ilymphocytesbecamechronicallyinfected,andcontinuedtoexcretelargenumbers
ofG.muriscystsforthedurationofthestudy(20).Thesedataindicatethathelper/inducerTlymphocytesplayamajorroleintheclearanceofG.murisinfection,
andthatcytotoxic/suppressorTcellsareoflittleimportancefortheeliminationofthisinfection.Itislikelythattheimpairedabilityofnudemicetomountanintestinal
antibodyresponseagainstG.muristrophozoitesistheresultofhelper/inducer(L3T4+)Tcelldeficiency.Nudemiceareknowntohaveamoreprofounddeficiency
ofhelper/inducerTlymphocytesthanofcytotoxic/suppressor(Ly2+)Tlymphocytes(8,32).Preliminarystudiesbythepresentauthorsuggestthatimmunocompetent
BALB/cmicewhichhavebeendepletedofTh/ilymphocytes,bytreatmentwithantiL3T4monoclonalantibody,haveanimpairedabilitytomountanintestinal
antibodyresponseagainstG.muristrophozoites(M.F.Heyworth,unpublisheddata).
Naturalkiller(NK)cellsarepresentintheintestinalmucosaofimmunocompetentmice.However,littleisknownabouttheroleofNKcellsinthegastrointestinaltract.
TodeterminewhetherNKcellsplayapartintheeliminationofG.murisinfection,thetimecourseofthisinfectionhasbeenstudiedinbeigemice,whicharedeficient
inNKcells(41).TheresultsofthisworkshowthatbeigemiceeliminateG.murisinfectionatthesamerateasmicewithnormalNKcellactivity(21).Thisfinding
stronglysuggeststhatNKcellsarenotinvolvedintheclearanceofG.murisinfectionfromthemouseintestine.

Page47

InanattempttodeterminewhethermacrophagesareimportantfortheclearanceofG.murisinfection,thepresentauthorharvestedleukocytesfromtheintestinal
lumenofGiardiainfectedimmunocompetentBALB/cmiceandnudemice.Leukocytesubsetswerethenidentifiedbyimmunofluorescentstainingwithmonoclonal
antibodiesdirectedagainstleukocytesurfaceantigens,includingthemacrophagesurfaceantigenMac1(22).Leukocytesbearingthisantigenwerequantifiedby
fluorescencemicroscopy.ThisworkshowedthattherewasnoappreciabledifferencebetweenthenumberofMac1+cellsharvestedfromtheintestinallumenof
BALB/cmiceornudemice,andthatonlysmallnumbersofcellsbearingtheMac1antigenwerepresentincellsuspensionsharvestedfromthemouseintestinallumen
(25103Mac1+cellspermouse22).TheseobservationssuggestthatintraluminalmacrophagesdonotplayanimportanteffectorroleintheeliminationofG.
murisinfection.TransmissionelectronmicroscopyofPeyer'spatchsectionsfromGiardiainfectedmicehasshownthatmacrophagesinthepatchesareableto
phagocytoseGiardiatrophozoites(36).BecausePeyer'spatchesareknowntobeimportantsitesfortheinitiationofintestinalimmuneresponses(25,26),itislikely
thatingestionoftrophozoitesbyPeyer'spatchmacrophagesisfollowedbypresentationoftrophozoiteantigenstolocalhelper/inducerTcellsandBcells,with
subsequentproductionofGiardiaspecificantibodiesinnormalimmunocompetentmice.
Conclusions
ThereisextensiveevidencethathumansubjectsinfectedwithGiardialamblia,andmiceinfectedwithG.muris,developantibodyresponsestoGiardia
trophozoites.TheabilityofmicetomountanantibodyresponsetoGiardiatrophozoitescorrelateswiththeabilityoftheseanimalstoeliminateG.murisinfection.The
observationthatnudemiceareunabletoclearG.murisinfectionindicatesthatTlymphocytesplayanimportantpartinantiGiardiaimmunity.Studiesinwhich
immunocompetentBALB/cmicewereselectivelydepletedofeitherhelper/inducerorcytotoxic/suppressorTlymphocytes,bytreatmentwithmonoclonalantibody,
haveshownthatclearanceofG.murisinfectionisdependentonhelper/inducerTcells.Cytotoxic/suppressorTcellsandnaturalkillercellsplaylittle,ifany,partin
eliminationofG.murisinfectionfromthemouseintestine.Importantareasoffuturestudyinclude:(a)todeterminewhetherGiardiatrophozoitesarekilledby
trophozoitespecificantibodiespresentinmouseorhumanintestinalsecretions,and(b)tocharacterizetrophozoiteantigenswhicharemajortargetsforintestinalanti
Giardiaantibodies.
Acknowledgements
GrantsupportfromtheNationalInstitutesofHealth(grantsAM33930andAM33004)andfromtheAcademicSenateCommitteeonResearchoftheUniversityof
California,SanFrancisco,isgratefullyacknowledged.
LiteratureCited
1.Anders,R.F.,I.C.RobertsThomson,andG.F.Mitchell.1982.Giardiasisinmice:analysisofhumoralandcellularimmuneresponsestoGiardiamuris.Parasite
Immunol.4:4757.
2.Andrews,J.S.,Jr.,andE.L.Hewlett.1981.ProtectionagainstinfectionwithGiardiamurisbymilkcontainingantibodytoGiardia.J.Infect.Dis.143:242246.
3.Bienenstock,J.,andA.D.Befus.1980.Mucosalimmunology.Immunology41:249270.
4.Black,R.E.,A.C.Dykes,S.P.Sinclair,andJ.G.Wells.1977.Giardiasisindaycarecenters:evidenceofpersontopersontransmission.Pediatrics60:486491.
5.Brett,S.J.,andF.E.G.Cox.1982.ImmunologicalaspectsofGiardiamurisandSpironucleusmurisinfectionsininbredandoutbredstrainsoflaboratorymice:a
comparativestudy.Parasitology85:8599.
6.Briaud,M.,M.MorichauBeauchant,C.Matuchansky,G.Touchard,andP.Babin.1981.Intestinalimmuneresponseingiardiasis.Lancetii:358.
7.Carlson,J.R.,M.F.Heyworth,andR.L.Owen.1986.ResponseofPeyer'spatchlymphocytesubsetstoGiardiamurisinfectioninBALB/cmice.II.Bcell
subsets:entericantigenexposureisassociatedwithimmunoglobulinisotypeswitchingbyPeyer'spatchBcells.Cell.Immunol.97:5158.
8.Carlson,J.R.,M.F.Heyworth,andR.L.Owen.1987.TlymphocytesubsetsinnudemicewithGiardiamurisinfection.Thymus9:189196.
9.Crabb'e,P.A.,D.R.Nash,H.Bazin,H.Eyssen,andJ.F.Heremans.1969.AntibodiesoftheIgAtypeinintestinalplasmacellsofgermfreemiceafteroralor
parenteralimmunizationwithferritin.J.Exp.Med.130:723744.
10.Crago,S.S.,W.H.Kutteh,I.Moro,M.R.Allansmith,J.Radl,J.J.Haaijman,andJ.Mestecky.1984.DistributionofIgA1,IgA2,andJchaincontainingcellsin
humantissues.J.Immunol.132:1618.
11.Craun,G.F.1979.WaterbornegiardiasisintheUnitedStates:areview.Am.J.PublicHealth69:817819.
12.Crossley,R.,andD.Holberton.1985.Assemblyof2.5nmfilamentsfromgiardin,aproteinassociatedwithcytoskeletalmicrotubulesinGiardia.J.CellSci.78:
205231.
13.Dialynas,D.P.,Z.S.Quan,K.A.Wall,A.Pierres,J.Quintans,M.R.Loken,M.Pierres,andF.W.Fitch.1983.CharacterizationofthemurineTcellsurface
molecule,designatedL3T4,identifiedbymonoclonalantibodyGK1.5:similarityofL3T4tothehumanLeu3/T4molecule.J.Immunol.131:24452451.
14.Farthing,M.J.G.,M.E.A.Pereira,andG.T.Keusch.1986.DescriptionandcharacterizationofasurfacelectinfromGiardialamblia.Infect.Immun.51:661
667.
15.Fleck,S.L.,S.E.Hames,andD.C.Warhurst.1985.DetectionofGiardiainhumanjejunumbytheimmunoperoxidasemethod.Specificandnonspecificresults.
Trans.R.Soc.Trop.Med.Hyg.79:110113.
16.Gilman,R.H.,K.H.Brown,G.S.Visvesvara,G.Mondal,B.Greenberg,R.B.Sack,F.Brandt,andM.U.Khan.1985.EpidemiologyandserologyofGiardia
lambliainadevelopingcountry:Bangladesh.Trans.R.Soc.Trop.Med.Hyg.79:469473.
17.Goka,A.K.J.,D.D.K.Rolston,V.I.Mathan,andM.J.G.Farthing.1986.DiagnosisofgiardiasisbyspecificIgMantibodyenzymelinkedimmunosorbentassay.
Lancetii:184186.
18.Hermans,P.E.,J.A.DiazBuxo,andJ.D.Stobo.1976.Idiopathiclateonsetimmunoglobulindeficiency:clinicalobservationsin50patients.Am.J.Med.61:
221237.
19.Heyworth,M.F.1986.AntibodyresponsetoGiardiamuristrophozoitesinmouseintestine.Infect.Immun.52:568571.

Page48

20.Heyworth,M.F.,J.R.Carlson,andT.H.Ermak.1987.ClearanceofGiardiamurisinfectionrequireshelper/inducerTlymphocytes.J.Exp.Med.165:1743
1748.
21.Heyworth,M.F.,J.E.Kung,andE.C.Eriksson.1986.ClearanceofGiardiamurisinfectioninmicedeficientinnaturalkillercells.Infect.Immun.54:903904.
22.Heyworth,M.F.,R.L.Owen,andA.L.Jones.1985.ComparisonofleukocytesobtainedfromtheintestinallumenofGiardiainfectedimmunocompetentmice
andnudemice.Gastroenterology89:13601365.
23.Kaplan,B.,andD.Altmanshofer.1985.GiardiamurisadherencetointestinalepitheliumtheroleofspecificantiGiardiaantibodies.Microecologyand
Therapy15:133140.
24.Kaplan,B.S.,S.Uni,M.Aikawa,andA.A.F.Mahmoud.1985.Effectormechanismofhostresistanceinmurinegiardiasis:specificIgGandIgAcellmediated
toxicity.J.Immunol.134:19751981.
25.Keren,D.F.,P.S.Holt,H.H.Collins,P.Gemski,andS.B.Formal.1978.TheroleofPeyer'spatchesinthelocalimmuneresponseofrabbitileumtolive
bacteria.J.Immunol.120:18921896.
26.Kiyono,H.,J.R.McGhee,M.J.Wannemuehler,M.V.Frangakis,D.M.Spalding,S.M.Michalek,andW.J.Koopman.1982.InvitroimmuneresponsestoaT
celldependentantigenbyculturesofdisassociatedmurinePeyer'spatch.Proc.Natl.Acad.Sci.USA79:596600.
27.Ledbetter,J.A.,andL.A.Herzenberg.1979.Xenogeneicmonoclonalantibodiestomouselymphoiddifferentiationantigens.Immunol.Rev.47:6390.
28.Ledbetter,J.A.,andW.E.Seaman.1982.TheLyt2,Lyt3macromolecules:structuralandfunctionalstudies.Immunol.Rev.68:197218.
29.Lev,B.,H.Ward,G.T.Keusch,andM.E.A.Pereira.1986.LectinactivationinGiardialambliabyhostprotease:anovelhostparasiteinteraction.Science
232:7173.
30.Loftness,T.J.,S.L.Erlandsen,I.D.Wilson,andE.A.Meyer.1984.OccurrenceofspecificsecretoryimmunoglobulinAinbileafterinoculationofGiardia
lambliatrophozoitesintoratduodenum.Gastroenterology87:10221029.
31.LoGalbo,P.R.,H.A.Sampson,andR.H.Buckley.1982.SymptomaticgiardiasisinthreepatientswithXlinkedagammaglobulinemia.J.Pediatr.101:7880.
32.MacDonald,H.R.,C.Blanc,R.K.Lees,andB.Sordat.1986.AbnormaldistributionofTcellsubsetsinathymicmice.J.Immunol.136:43374339.
33.Miotti,P.G.,R.H.Gilman,L.K.Pickering,G.RuizPalacios,H.S.Park,andR.H.Yolken.1985.PrevalenceofserumandmilkantibodiestoGiardialambliain
differentpopulationsoflactatingwomen.J.Infect.Dis.152:10251031.
34.Nash,T.E.,andA.Aggarwal.1986.CytotoxicityofmonoclonalantibodiestoasubsetofGiardiaisolates.J.Immunol.136:26282632.
35.Orlans,E.,J.V.Peppard,A.W.R.Payne,B.M.Fitzharris,B.M.Mullock,R.H.Hinton,andJ.G.Hall.1983.Comparativeaspectsofthehepatobiliarytransport
ofIgA.Ann.N.Y.Acad.Sci.409:411427.
36.Owen,R.L.,C.L.Allen,andD.P.Stevens.1981.PhagocytosisofGiardiamurisbymacrophagesinPeyer'spatchepitheliuminmice.Infect.Immun.33:591
601.
37.Owen,R.L.,P.C.Nemanic,andD.P.Stevens.1979.Ultrastructuralobservationsongiardiasisinamurinemodel.I.Intestinaldistribution,attachment,and
relationshiptotheimmunesystemofGiardiamuris.Gastroenterology76:757769.
38.Pickering,L.K.,W.E.Woodward,H.L.DuPont,andP.Sullivan.1984.OccurrenceofGiardialambliainchildrenindaycarecenters.J.Pediatr.104:522
526.
39.RobertsThomson,I.C.,andG.F.Mitchell.1978.Giardiasisinmice.I.Prolongedinfectionsincertainmousestrainsandhypothymic(nude)mice.
Gastroenterology75:4246.
40.RobertsThomson,I.C.,D.P.Stevens,A.A.F.Mahmoud,andK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gastroenterology71:5761.
41.Roder,J.C.1979.Thebeigemutationinthemouse.I.Astemcellpredeterminedimpairmentinnaturalkillercellfunction.J.Immunol.123:21682173.
42.Shaw,P.K.,R.E.Brodsky,D.O.Lyman,B.T.Wood,C.P.Hibler,G.R.Healy,K.I.E.MacLeod,W.Stahl,andM.G.Schultz.1977.Acommunitywide
outbreakofgiardiasiswithevidenceoftransmissionbyamunicipalwatersupply.Ann.Intern.Med.87:426432.
43.Smith,P.D.,F.D.Gillin,W.R.Brown,andT.E.Nash.1981.IgGantibodytoGiardialambliadetectedbyenzymelinkedimmunosorbentassay.
Gastroenterology80:14761480.
44.Snider,D.P.,J.Gordon,M.R.McDermott,andB.J.Underdown.1985.ChronicGiardiamurisinfectioninantiIgMtreatedmice.I.Analysisofimmunoglobulin
andparasitespecificantibodyinnormalandimmunoglobulindeficientanimals.J.Immunol.134:41534162.
45.Snider,D.P.,andB.J.Underdown.1986.QuantitativeandtemporalanalysesofmurineantibodyresponseinserumandgutsecretionstoinfectionwithGiardia
muris.Infect.Immun.52:271278.
46.Stevens,D.P.,D.M.Frank,andA.A.F.Mahmoud.1978.ThymusdependencyofhostresistancetoGiardiamurisinfection:studiesinnudemice.J.Immunol.
120:680682.
47.Torian,B.E.,R.C.Barnes,R.S.Stephens,andH.H.Stibbs.1984.TubulinandhighmolecularweightpolypeptidesasGiardialambliaantigens.Infect.Immun.
46:152158.
48.Visvesvara,G.S.,P.D.Smith,G.R.Healy,andW.R.Brown.1980.AnimmunofluorescencetesttodetectserumantibodiestoGiardialamblia.Ann.Intern.
Med.93:802805.
49.Vogt,R.L.,A.A.Little,K.C.Spitalny,andG.Visvesvara.1984.InvestigationofawaterborneoutbreakofgiardiasisusingserologictestingbyIFA.Am.J.
PublicHealth74:272.
50.Wofsy,D.,D.C.Mayes,J.Woodcock,andW.E.Seaman.1985.InhibitionofhumoralimmunityinvivobymonoclonalantibodytoL3T4:studieswithsoluble
antigensinintactmice.J.Immunol.135:16981701.
51.Wolfe,M.S.1978.Currentconceptsinparasitology.Giardiasis.N.Engl.J.Med.298:319321.

Page49

TheSecretoryImmuneResponseinRatsInfectedwithRodentGiardiaduodenalisIsolatesandEvidenceforPassiveProtection
withImmuneBile
GrahamMayrhofer*andAgnesWaightSharma
DepartmentofMicrobiologyandImmunology,TheUniversityofAdelaide,Box498,G.P.O.,Adelaide,SouthAustralia,5001,Australia.
TwoisolatesofGiardia,onefrommiceandonefromrats,havebeenidentifiedasGiardiaduodenalisbymorphologicalcriteria.Theyareidenticalwitheachotherandwithahuman
andafelineisolatebyisoenzymeanalysis.Nevertheless,theratisolateproducedachronicinfectioninallseveninbredratstrainstested,whilethemouseisolateineachcase
producedanacuteinfection.Infectionswithbothisolateswerechronicincongenitallyhypothymicnuderats.Naturalormetronidazoleinducedterminationofprimaryinfections
witheitherorganismwasfollowedbyahighlevelofimmunitytoreinfection.IgMandIgAantibodyresponsestohomologoustrophozoiteantigenshavebeenmeasuredinserumby
enzymelinkedimmunoabsorbentassaysduringprimaryandsecondaryinfectionswithbothisolates.However,onlyIgAantibodiesweredetectedinbilefrominfectedrats.Immune
bile,butnotnormalbile,ledtoasubstantialfallinfaecalcystexcretionwheninfusedintotheduodenaofconsciousratsinfectedwiththemouseisolate.Thefindingssuggestthat
secretoryantibodiesareprotectiveandthatcomparisonsbetweentheimmuneresponsesagainstthesecloselyrelatedrodentisolatesmayhelpdefineprotectiveantigens.

Introduction
MostoftheGiardiathatinfectmammalshavesimilarmorphologyandtheyhavebeengroupedintoasinglespecies,Giardiaduodenalis(5).Bythisclassification,
theorganismsresponsibleforgiardiasisinmanarereferredtoasG.duodenalis(lamblia).G.duodenalishasbeendividedintoracesonthebasisofsupposedhost
specificityandbymorphometry(20).However,therelationshipsbetweenthetentativespeciesareonlynowbeingexploredbymoresophisticatedmethodssuchas
isoenzymeanalysis(3,6),restrictionendonucleaseanalysisofDNAandgenomeprobingwithclonedfragmentsofGiardiaDNA(9).Evenwithinisolatesfrom
humans,thesemethodshaverevealedevidenceofgeneticdiversity,whiletheusefulnessofhostspecificityisnowrecognizedtobelimited.
Mostoftheexperimentalworkontheimmunologyofgiardiasishasbeencarriedoutinmice,usingisolatespresumedtobeGiardiamuris.G.muriscanbe
separatedeasilyfromG.duodenalisonmorphologicalgroundsandisthereforeadistinctspecies.Thismayaffectitsvalueasanexperimentalmodelforhuman
giardiasis,especiallyinworkaimedatdefiningantigensresponsibleforinducingprotectiveimmunity.AlthoughantigensareknowntobesharedbetweenG.murisand
G.duodenalis(lamblia),theextentofsharingisnotknown(13).
TheremaythereforebevalueinusinganimalisolatesofG.duodenalisasmodelsforhumandisease.Thiswouldallowstudiesofinfectioninthenaturalhostwith
organismsthatarepotentiallymorecloselyrelatedtoG.duodenalis(lamblia)thanisG.muris.ThemostreadilyavailableparasitehostcombinationisG.
duodenalis(simoni)intherat.Thechoiceoftherathastheaddedadvantagethatsecretoryimmunityiswellunderstoodinthisspecies.Inparticular,secretory
antibodyresponsestointestinalinfectionscanbemeasuredinbilebecausemostoftheIgAthatenterstheintestineinratsistransportedtherefromthebloodviathe
liver(8).ThereisevidencethattherodentisolatesofG.duodenalisusedinthisstudyareinfactverysimilartoG.duodenalis(lamblia).Thisraisesthepossibility
thatprotectiveantigensidentifiedinthismodelmaybemorecloselysimilartothoseofG.duodenalis(lamblia)thanwouldbethecaseforG.muris.
MaterialsandMethods
Animals
Thespecificpathogenfree(SPF)ratsusedtostudythekineticsofcystexcretionafterinfectionwithGiardiaisolateswere810weekoldfemalesobtainedfromthe
AnimalResourcesCentre,WesternAustralia.Noevidenceofanyparasiticinfectionwasfoundintheseanimals.ThestrainsusedwereFischer344(F344,RT1lvl),
WAG(RT1u),Lou/M(RT1u),WistarFurth(WF,RT1u),BrownNorway,(BN,RT1n),PVG/c(RT1c),andDA(RT1avl).Hypothymicnuderats(CBHmu/mu)came
fromthesamesource.Female810weekoldratsusedtostudyantibodyresponseswereobtainedfromtheGillesPlainsAnimalResourceCentre,SouthAustralia.
ThesewerebelievedtobeSPFbuttheyweresubseqentlyfoundtoexcretecystsofEntamoebaspp.TheywerefreefromallotherparasitesincludingGiardiaspp.
Duringexperiments,animalswerehousedinacleanconventionalroomonsterilizedlitterandhadfreeaccesstosterilizedwaterandfood.
GiardiaIsolatesandParasitologicalTechniques
TwoisolatesofG.duodenalishavebeenstudied.ThemouseGiardiawasisolatedfromrandombredalbinomiceinanonlaboratorycolonyandtheratGiardia
wasisolatedfromaconventionallaboratorycolonyofinbredGingerHoodedrats.Theisolateshavebeenmaintainedbyserialpassagethroughnuderatsat34month
intervals.Forcystcounts,faecescollectedfromindividual

Page50

ratsovera2hourperiodwereemulsifiedin0.01%Tween20indistilledwaterandthecystsconcentratedbycentrifugationover1Msucrose(14).Thecystswere
thencountedinahaemocytometerbyphasecontrastmicroscopyandthecountsareexpressedasthelog10meancystspergramoffaecesfortheanimalsineach
group.Toinfectanimals,cystswereconcentratedfromfaecalsuspensionsasdescribedabove.Aftercounting,theconcentrationwasadjustedtoallowintragastric
intubationof5000cystsin0.05mLofdistilledwater.Tocompletelyeradicatetheprimaryinfection,animalsweretreatedwith50mgofmetronidazolebyintragastric
intubationonthreeconsecutivedays.
MorphologicalStudies
Trophozoiteswereobtainedbyexcystationfrompurifiedcysts,asdescribedbelow.Smallamountsofsuspensionwerepartiallyairdriedonslidesandfixedin
Schaudin'sfixative.ThesmearswerethenstainedwithTrichrome(19)toidentifythemedianbodies.
IsoenzymeStudies
Trophozoiteswereexcystedfromcystspurifiedfromratfaecesbyinitialconcentrationover1Msucrose(seeabove),followedbyrepeatedsedimentationatunit
gravitythroughPercollgradients(15).Thecystswereheldindistilledwatercontainingpenicillin(200g/ml),gentamicin(200g/ml)andamphotericinB(2g/ml)for
3daysandwereshowntobebacteriologicallysterile.ExcystationwasperformedessentiallyasdescribedbySchaefer,RiceandHoff(16).Aliquotscontaining5
107trophozoitesweresnapfrozenindryiceacetoneandtransportedondryicetotheEvolutionaryBiologyUnitoftheSouthAustralianMuseum.Enzymeanalysis
wasperformedbyelectrophoresisoncelluloseacetategelsusingasonicateoftheorganisms,essentiallyasdescribedelsewhere(11).
SurgicalProceduresandSpecimenCollection
Atthetimeofbileductcannulation,bloodwascollectedfromthetailtipunderetheranaesthesia.Thebileductwasthenapproachedbyamidlineincisionand
cannulatedasneartheportahepatisaspossiblewithapolythenecannula.Approximately2mLofbilewascollectedimmediatelyintoicechilledtubesfromthe
consciousanimalsheldinBollmanmetaboliccages.Bileandserumsampleswerefrozenindryiceacetoneandheldat100Cuntilassayedforantibodies.Further
bilewasthencollectedintoicechilledcontainersfor24days.Thismaterialwasalsofrozenandstoredforuseinpassivetransferexperiments.
Intraduodenalinfusionofbilewasachievedthroughacannulaconnectedtoaperistalticpumpdelivering0.5mLperhour.Thecannulaconsistedofmedicalgrade
polyethylenetubing(0.4mminternaldiameter,0.8mmexternaldiameter),tippedwith2.5cmofsoftsiliconerubbertubing,(Silastic,DowCorning0.012in.internal
diameter,0.025in.externaldiameter).Ananteriormidlineabdominalincisionwasmadeunderetheranaesthesia.Thecannulawaspassedthroughtheposterior
abdominalwallontheleftsideandintothestomachthroughapunctureintheantralregion.Thecannulawasfedthroughthepylorusandanchoredwithapursestring
sutureasitenteredthestomach.Theendofthesiliconetubingwasadjustedtolieinproximitytotheentryofthecommonbileduct.Theanimalswereheld
unanaesthetizedandwithfreeaccesstofoodandwaterinBollmanmetaboliccages.Attheconclusionoftheperiodofbileinfusion,cannulaewereremovedbyquick
tractionandtheanimalsreturnedtoindividualholdingcages.
Antigens
AntigensforcoatingELISAplateswerepreparedfromcystspurifiedbyunitgravitysedimentationthroughPercollgradients(seeabove).Thepurifiedcystswere
storedat4Cindistilledwatercontainingantibiotics(seeabove)andusedwithin7days.Trophozoiteswereexcysted,suspendedinphosphatebufferedsaline(PBS,
400mOsm/L),adjustedto2106organismspermLandsonicated.PreliminaryexperimentsshowedthatsonicatespreparedinthiswayoptimallysensitizedELISA
traysforantiGiardiaantibodyestimations.
ImmunologicalReagents
RabbitantiratIgAwasraisedagainstIgApurifiedfromratthoracicductlymphanditwasabsorbedbypassagethrough2Sepharose4Bcolumns,onecoatedwith
normalratserumproteinsandtheotherwithpurifiedratIgG(allclasses).PureantiIgAantibodywaspreparedbyadsorptiontoandelutionfromanIgASepharose
4Bcolumn.Theantibodywasconjugatedwithalkalinephosphatase(CalfIntestineVIIS,SigmaChemicalCo.,Missouri,USA)bytheonestepglutaraldehyde
procedure(1).TheconjugateofrabbitantiratIgAwithalkalinephosphatasewaspreparedasabove,usingimmunopurifiedantibodykindlyprovidedbyDr.D.W.
Mason(Oxford).
TheantiIgAandantiIgMconjugateswereisotypespecificwhentestedbyELISAinwellscoatedwithoptimalamountsofpurifiedratIgA,IgG,IgMorIgE.
EnzymeLinkedImmunoabsorbentAssay(ELISA)
Assayswerecarriedoutusing96wellroundbottomedvinylmicrotitreplates(Costar,DataPackagingCorporation,Cambridge,Mass.)coatedwithantigenby
incubationwith100LofGiardiasonicatefor1hourat37Candthenovernightat4C.AfterwashingwithnormalPBScontaining0.05%Tween20and0.05%
sodiumazide(PBSTween20buffer),freebindingsitesontheplateswereblockedbyincubationwith1%bovineserumalbumen(BSAFractionV,Flow
Laboratories,NSW,Australia)inPBSTween20for68hoursat4C.Toassayantibodiesinserumorbile,2foldserialdilutionsinPBSTween20containing1%
BSAwereincubatedat4Covernightinwellscoatedwiththehomologoustrophozoiteantigenpreparation.Afterwashing,boundantibodywasdetectedbyafurther
incubationovernightat4Cwithpredetermineddilutionsofalkalinephosphataseantibodyconjugates.Substrate(pnitrophenylphosphatedisodium,SigmaChemical
Co.,in10%diethanolaminebuffer)wasaddedin100Laliquotstothewashedwellsandfurtherincubatedfor4hoursat37C.Opticaldensitiesofwellswereread
at405nmusingaTitertekMultiscanautomatedspectrophotometeradjustedtozeroonasubstrateblank.Apositiveantibodytitreinanysamplewasdefinedasthe
reciprocalofthedilutionwhichproducedameanopticaldensityof0.150,representingtwicethemeanOD405producedbyconjugatesreactinginantigencoatedwells
withoutaddedratantibodies.
Results
CharacterizationoftheTwoGiardiaIsolates
TrophozoitesfromthetwoisolatesareshowninFigure1.Morphometryhasnotbeenperformed,buttheyappearverysimilarinshapeandsize.Inparticular,the
medianbodiesinbothisolateshavethe''clawhammer"appearanceandthisplacesthemintheG.duodenalisgroup.
Evidencesupportingacloserelationshipbetweentheisolatescamefromcomparisonoftheelectrophoreticmobilitiesofthe27enzymesshowninTable1.Differences
intheelectrophoreticmobilitiesofenzymesbetweenindividualsorbetweenspecies(i.e.isoenzymes)

Figure1.
Examplesoftrophozoitesexcystedinvitrofromtherodentisolates.
a)Mouseisolate.b)Ratisolate.Bothhavemedianbodiescharacteristic
ofG.duodenalisandtheyhavesimilargeneralmorphologicalfeatures.

Page51
TABLE1.EnzymesidentifiedintheG.duodenalisisolates.
Enzyme

E.C.No.

Enzyme

E.C.No.

Aconitase

4.2.1.3

GlutathioneReductase

AcidPhosphatase

3.1.3.2

Hexokinase

2.7.1.1

AdenosineDeaminase

3.5.4.4

MalateDehydrogenase

1.1.1.27

AlcoholDehydrogenase

1.1.1.1

MalicEnzyme

1.1.1.40

Aldolase

4.1.2.13

MannosePhosphateIsomerase

5.3.1.8

Enolase

4.2.1.11

NucleosidePhosphorylase

2.4.2.1

FructoseDiphosphatase

3.1.3.11

Peptidase(Valineleucine)

3.4.11or13

Glyceraldehyde3Phosphate
Dehydrogenase

1.2.1.12

PhosphoglycerateMutase

2.7.5.3

GlutamateDehydrogenase

1.4.1.3

6PhosphogluconateDehydrogenase

1.1.1.44

GlutamateOxaloacetateTransaminase

2.6.1.1

PhosphoglycerateKinase

2.7.2.3

Glucose6PhosphateDehydrogenase

1.1.1.49

Phosphoglucomutase

2.7.5.1

GlycerophosphateDehydrogenase

1.1.1.8

SorbitolDehydrogenase

1.1.1.14

5.3.1.9

TriosePhosphateIsomerase

5.3.1.1

UridineMonophosphateKinase

2.7.4.?

GlucosePhosphateIsomerase

1.6.4.2

reflectcorrespondingstructuraldifferencesbetweenthegenesthatencodethoseenzymes.Nodifferenceswerenotedbetweentheratandmouseisolatesforanyof
the27enzymesexaminedandidentityatthisnumberoflocisuggestsstronglythatbothorganismsbelongtothesamespecies.
Furthermore,therewerenodifferencesatanyoftheselociwhenthetworodentisolateswerecomparedwithahumanisolate(Adelaide1)orwiththefelineisolate
Portland1(E.A.Meyer,Portland).TherodentisolatesofG.duodenalisthereforeappeartobecloselyrelatedtoG.duodenalis(lamblia)andtoaG.duodenalis
fromcats.
InfectionsinNormalRats
ThedurationofinfectionwithG.murisvaries,dependingonwhichstrainsofmicearestudied(2,10,12).Primaryinfectionswiththemouse(Figure2)andrat(Figure
3)isolatesofG.duodenalishavebeenstudiedinthestrainsofinbredratsthatwereavailable.Thestrainhadlittleinfluenceoneitherthemagnitudeordurationofcyst
excretionbyratsduringinfectionswitheitherisolate.
However,whentheinfectionsproducedbythetwoisolateswerecompared,therewerestrikingdifferences.Withthemouseisolate,cystexcretionreachedapeak
quicklyandwithinaweekcommencedtodecline,untilfinallyitceasedafterapproximately56weeks.Incontrast,allratstrainsbecameinfectedchronicallywiththe
ratisolateandshowednoevidenceofexpellingtheorganisms.Theinfectionhasbeenobservedtopersistatasimilarlevelforatleast6monthsinF344strainrats
(datanotshown).
Infectionswithbothisolatesweretreatedwithmetronidazoleafter10weeks.Whenchallengedwiththehomologousorganism2weekslater,theanimalswerefound
tobehighlyresistanttoreinfection.Someanimalsinmoststrainsfailedtoexcretecystsatdetectablelevelsandwhereexcretiondidoccur,itwasinsmallnumbersand
foronlyafewdays.Inmorerecentexperiments(notshown),where4weekshavebeenallowedbetweenmetronidazoletreatmentandreinfection,resistanceaftera
primaryinfectionhasbeencomplete.

Figure2.
ThecourseofinfectionwiththemouseisolateofG.duodenalis
in7normalinbredratstrainsandincongenitallyhypothymic
nudemu/murats.Primaryandsecondaryinfectionswereboth
initiatedwith5000cysts.Ratsweretreatedwithmetronidazole10
weeksaftercommencementoftheprimaryinfectionandwerechallenged
withaseconddoseofcysts2weekslater.()courseofprimary
infection.
courseofsecondaryinfection.Limitsofdetection
ofcystsinfaecesareindicatedbyhorizontaldottedlines.
Pointsaremeanfor5animals,standarderror.

Page52

InfectionsinHypothymicNudeRats
NuderatsoftheCBHmu/mustrainhavebeenusedtomaintainstocksofthetwoG.duodenalisisolates.AsshowninFigures2and3,theseratscontinuedto
excretehighlevelsofcystsforlongperiodsoftimeafterinfectionwitheitherorganism.Inparticular,theinfectionproducedbythemouseisolatewasdramatically
differentwhencomparedinthymusdeficientandimmunocompetentratstrains.
SerumandBiliaryAntibodyResponses
TitresofIgMandIgAantibodieswereestimatedbyisotypespecificELISAintheserumandbileofratsinfectedwiththemouseandratisolates(Figures4and5).
DuringprimaryinfectionswitheitherorganismtherewasanearlyriseinIgMantiGiardiaantibodiesinserum,whichdeclinedbetween10and15daysafterinfection.
ThiswasfollowedbyariseinserumIgAantibodies,whichwassustaineduntilthetimeofmetronidazoletreatment6weeksafterinfection.
SecondaryinfectionsproducedserumIgMantibodyresponsessimilartothosefollowingprimaryinfections.However,theIgAantibodyresponsesinserumwere
secondaryincharacter(i.e.largeandmorerapid).
IgMantibodieswerenotdetectedinbileatanytimeduringprimaryorsecondaryinfectionswitheitherisolate.IgAantibodylevelsinbileroseinparallelwiththelevels
inserum.AsecondaryIgAantibodyresponsewasdetectedinthebileofratsundergoingsecondaryinfectionswitheitherisolate.
Priortoinfection,theratsusedintheseexperimentshadsignificantlevelsofIgMantibodyinserumandofIgAantibodyinbile.Theselowlevelsofspecificorcross
reactiveantibodyappearrelatedtothesourceofanimals.SPFratsofthesamestrainfromadifferentsource(AnimalResourcesCentre,WesternAustralia)were
foundtohavelevelsofserumandbiliaryantibodiesneartothebackgroundsoftheELISAassays(datanotshown).
PassiveProtectionWithImmuneBile
Inthisexperiment,8ratswereinfectedwiththemouseisolateofG.duodenalis.Cannulaewereinsertedintotheduodenaoftheanimalsonday8afterinfection.Four
ratsreceivedapproximately12mLofbileperdayfromapoolcollectedfromtheanimalswiththeprimaryinfectionsinthepreviousexperiment.TheELISAtitreof
IgAantibodyinthispoolwas2048.Theremaining4ratsreceivedthesamevolumeofbile,butfromapoolofuninfecteddonors.Therateofbileinfusion
approximatedthenormaldailybileoutput.Infusionwascontinuedforaperiodof84hours,afterwhichthecannulaewereremovedandtheanimalswerereturnedto
holdingcages.
FaecalcystexcretionduringthecourseofinfectioninthetwogroupsisshowninFigure6.Infusionofcontrolbilefromuninfecteddonorshadnoeffectontherateof
cystexcretion.Incontrast,inanimalsreceivingimmunebiletherewasaverysignificantdeclineincystexcretion,commencingtowardstheendoftheinfusionperiod.
Animalsthathadreceivedimmunebilecontinuedtoexcretefewercystsuntiltheendoftheinfection,whencystexcretiondeclinedinbothgroups.

Figure3.
ThecourseofinfectionwiththeratisolateofG.duodenalisin6
normalinbredratstrainsandincongenitallyhypothymicnude
CBHmu/murats.Alldetailsareasdescribedinthelegend
toFigure2.()courseofprimaryinfection.
courseofsecondaryinfection.

Figure4.
Specificantibodytitres,againsthomologoustrophozoiteantigens,
intheseraandbileoffemaleDAratsinfectedwiththemouse
isolateofG.duodenalis.a)Primaryinfection.b)Secondaryinfection.
Theprimaryinfectionwith5000cystswasterminatedafter
6weekswithmetronidazoleandtheanimalswerereinfected
withthesamedoseofcysts4weekslater.Antibodytitreswere
measuredusingisotypespecificenzymelinkedimmunoabsorbentassays.
(Dottedlines)measurementsonsera,(Solidlines)measurements
onbile,( )IgAantibodies,()IgMantibodies.
Pointsaremeanstandarderrorfor5animals.

Page53

Figure5.
Specificantibodytitres,againsthomologoustrophozoiteantigens,
intheseraandbileoffemaleDAratsinfectedwiththeratisolate
ofG.duodenalis.a)Primaryinfection.b)Secondaryinfection.
AllotherdetailsasdescribedinlegendtoFigure4.(Dottedlines)
measurementsonsera,(Solidlines)measurementsonbile,( )IgAantibodies,
()IgMantibodies.Pointsaremeanstandarderrorfor5animals.

Figure6.
Theeffectofpassiveintraintestinalimmunizationwithimmune
bileonthecourseofinfectionwiththemouseisolateinfemale
DArats.Theratsineachgroupwereinfectedwith5000cystsand
7dayslateracannulawasintroducedintotheduodenum.Immunebile
(Dottedline),orcontrolnormalbile(Solidline),wasinfusedat
therateof12mLperdayfortheperiodindicatedbythehorizontalbar.
Thecannulaewerethenwithdrawnandcystexcretionwasmonitored
fortheperiodindicated.Horizontaldottedlineindicateslimits
ofcystdetection.Pointsaremeanstandarderrorfor4animals.

Discussion
ImportantfindingswithregardtospeciationofG.duodenalishaveemergedfromthisstudy.Althoughrelatedbymorphologyandatallenzymelocitested,thetwo
rodentisolatescauseverydifferentinfectionsinrats.Theinfectionproducedbythemouseisolatewasacute.ThisisolatealsoproducesanacuteinfectioninBALB/c
mice(unpublishedresults),similartothatdescribedforG.muris(12).AfurthersimilaritytoG.muriswasthatthemouseisolateproducedachronicinfectionin
C3H/HeJmicethatwasonlyclearedafterapproximately15weeks(unpublishedresults).Incontrast,theratisolateproducedachronicinfectioninrats,withno
evidenceofresolutionaftermanyweeks.InBALB/cmice,thisisolateproducedanacuteinfectionsimilartothatcausedbythemouseisolate,althoughinC3H/HeJ
micetheinfectionwiththeratisolatewaslesschronic,resolvinginapproximately12weeks(unpublishedresults).Ingeneralterms,eachorganismappearstoproduce
amorechronicinfectioninitshostoforigin,butthisismorepronouncedwiththeratisolate.Noevidencewasobtainedforlargedifferencesinsusceptibilityto
infectionbetweenratstrains,incontrasttothefindingsinmice(2,10,12).
ComparativestudiesbetweentheseorganismsmaythereforeprovideimportantcluestothenatureofvirulencedeterminantsinGiardia.Thefindingssuggestthat
parasitefactors,aswellashostfactors,influencethechronicityofgiardiasis.Theyraisethepossibilitythatsimilarparasiterelatedfactorsmayberesponsibleforthe
chronicityofinfectioninsomecasesofhumangiardiasis.Furthermore,itisclearthatdifferencesbetweenisolatesproducingacuteorchronicinfectionsinmancould
bequitesubtleandtheycouldbemissedbytechniquessuchasisoenzymeanalysis.
Apotentialadvantageofthismodelofgiardiasisisthatcomparisonscanbemadeoftheimmuneresponsestoinfectionwiththetwocloselyrelatedorganismsinthe
samestrainofrat.Thismayallowidentificationoftheantigensagainstwhichtheprotectiveimmuneresponseisdirected.AsimilarstrategyhasbeenattemptedwithG.
muris,whentheimmuneresponsewascomparedbetweenstrainsofmicethatdifferedintheirsusceptibilitiestoinfectionwiththeoneorganism(4).
PreviousisoenzymestudieshavesuggestedacloserelationshipbetweensomehumanisolatesandthefelinePortland1strain(3).Inthepresentstudy,examininga
greaternumberofenzymes,Portland1wasfoundtobeidenticaltoahumanisolate(Adelaide1)andtothetworodentisolatesofG.duodenalis.Itisthereforelikely
thatsomehumanisolatesarecloselyrelatedtoGiardiathatinfectothermammalianspecies.ThesimilarityoftherodentG.duodenalisisolatestoG.duodenalis
(lamblia)maythereforeassistindefiningtheprotectiveantigensinhumangiardiasis.Themechanismofprotectiveimmunityingiardiasisisstillpoorlyunderstood.The
chronicinfectionsinhypothymicratssuggestaroleforTlymphocytes,ashasbeensuggestedfromstudiesonmurinegiardiasis(12,18).However,furtherstudies,
includingantibodymeasurementsinseraandsecretions,willbenecessarytoidentifythepreciseroleofTcells.Evidenceoftheroleofantibodyinprotectionhasbeen
strengthenedbyrecentstudiesinimmunoglobulindeficientmice(17).LevelsofIgMandIgAantibodieshavethereforebeenmeasuredintheseraandbilefromrats
infectedwitheachrodentisolate.Theserologicalfindingsindicatethatbothisolatesarecomparablyimmunogenicandthatbothinducesecretoryantibodyresponses
whichcouldbeimportantinimmunitytothisessentiallylumendwellingorganism.Studiesareinprogresstoexaminewhetherqualitativedifferencesexistbetweenthe
antibodyresponsestothetwoorganisms.Thesestudiesmayrevealwhytheimmuneresponseiseffectiveagainstthemouseisolatebutnotagainsttherat

Page54

isolateinprimaryinfections.However,theprocessesthatfrustratetheimmuneresponsetotheratisolatemaybesubtle,becauseanimalsprimedagainstthisorganism
andcuredbytreatmentwithmetronidazolearesubsequentlyhighlyresistanttoreinfection.Thelatterphenomenonwillprovideaninterestingareaforfuture
investigation.
Finally,directevidencehasbeensoughtfortheroleofsecretoryantibodyinimmunityagainstGiardiabypassivetransferofbile,whichcanbeobtainedeasilyandisa
majorsourceofintestinalIgAinrats(7,8).Immunebile,containingIgAagainstthehomologousisolate,wasinfusedintotheduodenaofanimalsinfectedwiththe
mouseisolate.Infusionofimmunebile,butnotofbilefromuninfectedrats,ledtoamarkeddecreaseincystexcretion.Thissortofstudyislimitedbytheamountsof
bileavailableandthetimeforwhichanimalscanberestrained.Nevertheless,theevidenceiscompellingthatimmunebiledeliveredinphysiologicalamountsreduced
theparasiteload.ThiseffectisattributedtothecontentofIgAantibodyinthebile.Furtherstudiesareinprogresstoexaminetheeffectsofimmunebileinpreventing
establishmentofinfectionswiththehomologousisolatesandinaffectingthecourseofinfectionwiththeratisolate.
ThesestudiessuggestthatinfectionsintheratwithrodentG.duodenalisisolatesmaybeusefulasamodelforinvestigatingimmunityingiardiasis.Althoughthe
importanceofbileasarouteofsecretionofIgAantibodiesintotheintestinediffersbetweenratsandman,theconvenienceofbileinratsformeasurementand
collectionofIgAantibodiesisobvious.Themodeofactionofsecretoryantibodiesinreducingthetrophozoitepopulationintheduodenumisunknown,buttheyare
likelytointerferewithattachmenttothemucosa.Invitrostudiessuggestthatserumantibodiescanpreventattachmentoftrophozoitestoartificialsubstratesby
immobilizationofflagellae(Mayrhofer,unpublishedresults).
Acknowledgements
ThisworkwassupportedbygrantsfromtheNationalHealthandMedicalResearchCouncilofAustraliaandtheChannel10Children'sMedicalResearch
Foundation,Adelaide.
TheauthorsthankDr.P.EyforhisadviceonimmunoassaysandMrs.GlenysKingandMrs.RosieThomasfortheirhelpinpreparingthemanuscript.Theisoenzyme
analysiswaskindlyperformedbyDr.R.AndrewsattheEvolutionaryBiologyUnitoftheSouthAustralianMuseum.
LiteratureCited
1.Avrameas,S.1969.Couplingofenzymestoproteinswithglutaraldehyde.Useoftheconjugatesforthedetectionofantigensandantibodies.Immunochemistry
6:4352.
2.Belosevic,M.,G.M.Faubert,E.Skamene,andJ.D.MacLean.1984.SusceptibilityandresistanceofinbredmicetoGiardiamuris.Infect.Immun.44:282286.
3.Bertram,M.A.,E.A.Meyer,J.D.Lile,andS.A.Morse.1983.AcomparisonofisozymesoffiveaxenicGiardiaisolates.J.Parasitol.69:793801.
4.Erlich,J.H.,R.F.Anders,I.C.RobertsThomson,J.W.Schrader,andG.F.Mitchell.1983.Anexaminationofdifferencesinserumantibodyspecificitiesand
hypersensitivityreactionsascontributingfactorstochronicinfectionwiththeintestinalprotozoanparasite,Giardiamuris,inmice.Aust.J.Exp.Biol.Med.Sci.
61:599615.
5.Filice,F.P.1952.StudiesonthecytologyandlifehistoryofGiardiafromthelaboratoryrat.Univ.Calif.Publ.Zool.57:53146.
6.Korman,S.H.,S.M.LeBlancq,D.T.Spira,J.ElOn,R.M.Reifen,andR.J.Deckelbaum.1986.Giardialamblia:identificationofdifferentstrainsfromman.Z.
Parasitenkd.72:173180.
7.LematreCoelho,I.,G.D.F.Jackson,andJ.P.Vaerman.1977.RatbileasaconvenientsourceofsecretoryIgAandfreesecretorycomponent.Eur.J.Immunol.
7:588590.
8.LematreCoelho,I.,G.D.F.Jackson,andJ.P.Vaerman.1978.RelevanceofbiliaryIgAantibodiesinratintestinalimmunity.Scand.J.Immunol.8:459463.
9.Nash,T.E.,J.McCutchan,D.Keister,J.D.Dame,J.D.Conrad,andF.D.Gillin.1985.RestrictionendonucleaseanalysisofDNAfrom15Giardiaisolates
obtainedfromhumansandanimals.J.Infect.Dis.152:6473.
10.Olveda,R.,J.S.Andrews,andE.L.Hewlett.1982.Murinegiardiasis:localizationoftrophozoitesandsmallbowelhistopathologyduringthecourseofinfection.
Am.J.Trop.Med.Hyg.31:6066.
11.Richardson,B.J.,P.R.Baverstock,andM.Adams.1986.Allozymeelectrophoresis:ahandbookforsystematicandpopulationstudies.AcademicPress,Sydney.
12.RobertsThomson,I.C.,andG.F.Mitchell.1978.Giardiasisinmice.I.Prolongedinfectionsincertainmousestrainsandhypothymic(nude)mice.
Gastroenterology75:4246.
13.RobertsThomson,I.C.,andR.E.Anders.1981.Serumantibodiesinadultswithgiardiasis.Gastroenterology80:1262.
14.RobertsThomson,I.C.,D.P.Stevens,A.A.F.Mahmoud,andK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gastroenterology71:5761.
15.Sauch,J.F.1984.PurificationofGiardiamuriscystsbyvelocitysedimentation.Appl.Environ.Microbiol.48:454455.
16.SchaeferIII,F.W.,E.W.Rice,andJ.C.Hoff.1984.FactorspromotinginvitroexcystationofGiardiamuriscysts.Trans.Roy.Soc.Trop.Med.Hyg.78:795
800.
17.Snider,D.P.,J.Gordon,M.R.McDermottandB.J.Underdown.1985.ChronicGiardiamurisinfectioninantiIgMtreatedmice.I.Analysisofimmunoglobulin
andparasitespecificantibodyinnormalandimmunoglobulindeficientanimals.J.Immunol.134:41534163.
18.Stevens,D.P.,D.M.Frank,andA.A.F.Mahmoud.1978.ThymusdependencyofhostresistancetoGiardiamurisinfection:studiesinnudemice.J.Immunol.
120:680682.
19.Wheatley,W.B.1951.Arapidstainingprocedureforintestinalamoebaeandflagellates.Am.J.Clin.Pathol.21:990991.
20.Woo,P.K.1984.EvidenceforanimalreservoirsandtransmissionofGiardiainfectionbetweenanimalspecies.p.341364.In:S.L.ErlandsenandE.A.Meyer
(eds.),GiardiaandGiardiasis.PlenumPress,N.Y.

Page55

BiologicalDifferencesinGiardialamblia
T.E.Nash*andA.Aggarwal
NationalInstitutesofHealth,NIAIB,LPD,Building5,Rm118Bethesda,Maryland20205,U.S.A..
AlthoughisolatesofGiardiadifferbiochemicallyitisnotknowniftheseorotherdifferencesresultinalteredbiologicalbehavior.Gerbilswereinfectedwithtwouniquehuman
isolates,GS/EandWB.AllWBinoculatedgerbilsbecameinfectedandwereabletoselfcurebyday28.IncontrastGS/Einfectedgerbilstendedtoremaininfected.Aftertreatment,
previouslyWBinfectedgerbilsresistedinfectionafterrechallengewithWBandGS/E,whileGS/EinfectedgerbilspartiallyresistedrechallengewithGS/E.Allwereinfectedwith
WB.ResistancetoinfectioncorrelatedwiththedevelopmentofcytotoxicantibodieswhichreactedwiththesurfaceoftheGiardia.Experimentalinfectionsinhumansconfirmedthat
Giardiaisolatesdifferedbiologically.Inthefirststudy,fivevolunteerswereenterallyinoculatedwith50,000trophozoitesofisolatesGS/MorIsr,followedseriallyandtreatedon
day15.Inthesecondstudy,twoofthepreviouslyinoculated,infectedandtreatedvolunteerswererechallengedalongwithfivenewvolunteersascontrols.All10oftheGS/M
inoculatedvolunteersbecameinfectedcomparedtononeofthe5inoculatedwithIsr.Bothrechallengedindividualsbecameinfectedalthoughoneonlytransientlyshedcysts.Of
the10infected,5or50%becameill,4withdiarrheaandtypicalsymptomsofgiardiasis.HumoralimmuneresponsestoGiardiaoccurredinallinfectedvolunteers.Therefore,these
experimentsgivecredencetotheideathatGiardiaarenotonlybiochemicallydifferent,butalsodifferintheirbiologicalbehavior.

Introduction
Morphologically,Giardialambliaisolatesappeartobesimilar(4)however,bothcasualobservationsanddetailedanalysisrevealdifferencesamongisolates.Some
isolates,afteraxenization,adheremostlytothesurfaceofthetubeswhileothersswimvigorouslyinthemediumandtendnottoadheretosurfaces(personal
observation).Otherisolatesarelongandslenderwhilesomeareplumperandfullerinshape(personalobservation).Someareeasilyaxenizedwhileothersgrow
axenicallywithdifficultyornotatall(personalobservation).Biochemicaldifferenceshavealsobeennoted(6,8,9,10).WhentheDNAsofvarioushumanGiardia
isolatesweredigestedwithendonucleaserestrictionenzymesandhybridizedtoGiardiaspecificprobes,thenumberandpositionofthebandsdifferedinamajorityof
theisolates(9).Inaddition,thesurfaceantigensvariedinmostisolatesasdemonstratedbysurfacelabeling(8)andthebindingofmonoclonalantibodies(McAb)to
thesurfaceofisolatespossessinga170kdantigen(5).Otherstudiessuggestedmostoftheheterogeneityresidedonthesurface(12).
Sincemanyinteractionsamongcellsoccuratthesurface,itisnotunreasonabletoexpectthatvaryingsurfaceantigensandotherdifferenceswouldresultinalterations
inbehavior.OthershavenoteddifferencesincystsheddinginmiceinfectedwithhumaninfectiveGiardiacysts(1),andwehavenotedmarkeddifferencesinthe
abilityofcystsobtainedfrominfectedhumanstoinfectinfantmice(personalobservation).
Tomorefullystudywhetherisolatesvariedintheirbiologicalbehavior,gerbils(3)andhumanswereinfectedwithcharacterizedisolatesandthecourseofinfection
followed(2).
Thegoalsofthegerbilinfectionsweretoanswerthefollowingquestions:
(i)Dogerbilsinfectedwithdifferentisolatesbecomeinfectedandundergoasimilarcourseofinfection?
(ii)Doesresistancetoreinfectiondevelop?
(iii)Ifresistancedevelops,isitthesametohomologousandheterologousisolates?
(iv)Whataretheimmuneresponsestoinfectionanddoanycorrelatewiththecourseofinfectionorthedevelopmentofresistance?
Results
SixweekoldMongoliangerbils(Merionesunguiculatus)wereinoculatedwith2milliontrophozoitesbygavageandthenumberofGiardiaintheintestines
determinedovertime.Thetwocomparedisolatesdiffereddramatically.Bothoriginatedfromsymptomaticinfectedhumans.WBwasisolatedfromapatientinfected
inAfghanistan(9,11)andGS/EfromascientistinfectedwhilefishingandcampinginAlaska(9).Theisolatesdifferedinallparameterstestedincludingendonuclease
restrictionpatterns(9),surfaceantigens(8),andexcretorysecretoryproducts(8).SomegerbilswerealsoinfectedwithisolateIsr.ThisisolatecloselyresemblesWB
infact,WBandIsrappearindistinguishablebiochemically(8,9).
Boththecourseofinfectionandabilitytoinduceresistancetoreinfectiondiffered(2).AlthoughallthegerbilsinoculatedwitheitherWBorGS/Ebecameinfected,
*Correspondingauthor.

Page56
TABLE1.PercentGiardiatrophozoiteskilledbycytotoxicantibodiesinducedbyWBorGSEinfectioningerbils.
Infection
WB

GSE

Daypostinoculation

IsolateUsed
asTarget

14

21

28

WB

12.50.5*

27.55.7

48.34.3

48.50.5

GSE

10.35.5

25.03.0

47.05.0

48.05.0

WB

3.62.3

3.13.1

10.55.0

18.72.0

GSE

20.44.0

29.05.0

32.33.0

38.515.0

Cytotoxicitywasdeterminedaspreviouslydescribed(2).Briefly,survivingGiardiaweresubculturedinTYIS33
after12hexposuretogerbilserum.ThenumberofviableGiardiaafter24hwasproportionaltotheoriginal
inoculum.ControlsconsistedofGiardiaexposedtomediumalone.
*EachtimepointrepresentsthemeanS.D.percentcytotoxicityofthreeexperiments.Pooledserafrom35animals
wereusedforeachtimepointperexperiment.

thoseinfectedwithWBselfcured(Figure1,PanelsAandD)byday35.Incontrast,mostGS/Einfectedgerbilswerestillinfectedevenonday42althoughthe
numberoftrophozoitesintheintestineshaddecreased.Gerbilsinfectedwitheitherisolateweretreatedwithmetronidazoleonday28andchallengedwitheitherthe
homologousorheterologousisolate7dayslater.GerbilspreviouslyinfectedwithWB(Figure1,PanelsBandE),resistedinfectionwitheitherisolatehowever,
gerbilspreviouslyinfectedwithGS/EwerepartiallyresistanttochallengewiththehomologousisolatebutallanimalschallengedwiththeWBisolatebecameinfected
althoughtherewerefewerorganismsintheintestines(Figure1,PanelsC.andF).ThepatternofinfectionwithIsrwassimilartothatwithWB.
ComplementindependentcytotoxicantibodiesdevelopedintheseraofgerbilsduringinfectionandthedegreeofcytotoxicitywasdependentontheinfectingGiardia
isolates,thetestisolateemployed,andthedurationoftheinfection(Table1).Thedevelopmentofcytotoxicantibodiescorrelatedwiththeabilitytoresistinfection.
WBinfectedgerbilsdevelopedappreciablecytotoxicityforbothisolatesandresistedinfectiontobothisolates.Ontheotherhand,GS/Einfectedgerbilsdeveloped
higherlevelsofantibodiestoGS/EthantoWB.TheseanimalswerepartiallyresistanttobothisolatesbutmoresotothehomologousGS/EGiardia.
ThesestudiesconclusivelyshowthatGiardiaisolatesnotonlydifferbiochemicallybutintheirbehaviourinvitro.Inaddition,theabilityofeachisolatetoinduce
varyingamountsofcytotoxicantibodiesagainstthesetwoisolatessuggestsurfaceantigensmaybeimportantastargetantigensinprotectinggerbilsfromreinfection.
HumanswerealsoinfectedwithtwodifferentGiardiaisolates(7).Thegoalsoftheseexperimentswerenecessarilydifferentthanthoseusinggerbilsmainlybecause
muchlessisknownaboutexperimentalGiardiainfectionsinhumans.Thegoalsoftheseexperimentsweretoanswersomeofthefollowingquestions:
(i)Canaxenizedtrophozoitesinfecthumans?
(ii)Aredifferentisolatesequallycapableofinfectinghumans?
(iii)Dotheseisolatescausedisease?

Figure1.
Thelowerhalfofthegraphshowsthenumberoftrophozoites
inthesmallintestinesofgerbils.Theupperhalfshowsthepercentof
animalsinfectedondifferentdaysafterinoculationwithWB
orGSEisolates.Eachpointrepresentsthelogof
meantrophozoitecountsfrom1015gerbils.
WB,GSEPanelsA&D,infectionwiththeWBorGSEisolate.
WB/WBPanelsB&E,primaryWBinfectionchallengedwithWB.
WB/GSEPanelsB&E,primaryWBinfectionchallengedwithGSE.
GSE/WBPanelsC&F,primaryGSEinfectionchallengedwithWB.
GSE/GSEPanelsC&F.,primaryGSEinfectionchallengedwithGSE.

(iv)Whatisthecourseofinfectionanddisease?
(v)Doesresistancetoinfectiondevelop,andisitequalinhomologousandheterologousisolates?
(vi)Whataretheimmunologicalresponsestoinfectioninhumans?
Twoisolateswereused,GS/MandIsr.GS/MisthesameisolateasGS/E,exceptthetrophozoitesusedforaxenizationwereobtainedfromneonatallyinfectedmice
andnotafterinvitroexcystmentofpurifiedcysts.Isrwasisolatedfromaninfant(fromBethesda,MD)withdiarrhea(8,9).Asmentionedabove,thisisolate
resembledWB.WBcouldnotbeusedbecausetheisolateoriginatedfromapersonwhowasresistanttostandardcoursesofchemotherapy.

Page57

ExtensivestudiesweredonetoexcludeinfectionwithGiardiaorotheragentsandtoexcludeotherunderlyingorcomplicatingconditions.Onday0volunteerswere
inoculatedwith50,000viabletrophozoitesenterallyviaapolyvinylcatheterpositionedinthejejunum.Eachdaytheywereaskedaboutthepresenceorabsenceof
particularsymptomsandexaminedifnecessary.Stoolswerecollecteddaily,andtheconsistencyandpresenceofGiardianoted.Jejunalaspirateswereobtainedon
days0,14,and19.Stoolswereinitiallyexaminedwithoutconcentration,andifnegative,werereexaminedfollowingconcentration.Volunteersweretreatedonday
15.Twelveweeksafterinoculationandinfection,3GS/Minfectedandtreatedpatientswerechallengedwiththesameisolateasbefore.Anadditional5volunteers
wereinoculatedatthesametime,ascontrols.
Inthefirstexperiment,all5oftheGS/MinoculatedvolunteersbecameinfectedandnoneoftheIsrinoculatedvolunteers.Inthesecondexperimentall5ofthe
controlsbecameinfected.Therefore,10of10GS/Minoculatedvolunteersbecameinfectedwhile0of5Isrinoculatedwereinfected(p<.004,twotailedFishers
exacttest).
Ofthe10GS/Minfectedvolunteers,4developeddiarrheawhichwasdefinedasstoolswhichtooktheshapeofthecontainer(Grade3).Onepersondevelopedfever
andheadachesbutwithoutdiarrhea,andalthoughevaluatedextensively,noothercauseforthefeverandheadachewasfound.Altogethertherewere17episodesof
diarrheaintheinfectedgroupand1inthe5nonIsrinfectedgroup.Diarrheaoccurredsignificantlymoreoftenintheinfectedvolunteersthaninthenoninfected
volunteers(p<.035,Yatesmeanscore).Althoughthesesymptomsweremorefrequentintheinfectedgroup,theywerenotsignificantlymorefrequentatthep<.05
level.
ThecourseofinfectionintwosymptomaticvolunteersinthefirstgroupisshowninFigure2.Cystswerefirstnotedintheirstoolsonday6.Inpatient2,asingle
episodeofdiarrheaoccurredonday3followedbynumerousepisodesbeginningonday7,1dayaftertheonsetofcystshedding.Othersymptomsincludingvomiting,
cramping,anorexia,andflatulencealsobeganataboutthistime.Patient1showedalesssymptomaticcourse.Diarrheabeganonday8.Crampsandflatuswerehis
maincomplaints.Themeanprepatentperiodforall10infectvolunteerswas7.5.97(meanSD)daysanddiarrheabegan7.252.99daysafterinoculation.
Ofthechallengedvolunteers,onewasexcretingcystsatthetimeofchallenge,andtheothertwobecamereinfected.Noneweresymptomatic.GiardiaspecificIgM
antibodiesdevelopedinallinfectedpatientsandinnoneoftheIsrinoculatedpatients.SpecificIgGandIgAresponsesoccurredin70%and60%oftheinfected
volunteers,respectively.IntestinalIgAresponsesoccurredin60%ofpatients.Cytotoxicantibodieswerenotdetectedfailuretodetectthisresponsemaybedueto
theshortdurationofinfectionpriortotreatment.
Thesehumanstudiesareimportantinanumberofwaysalthoughnotalwayscomparabletotheexperimentalgerbilinfections.First,asafeandreliablemethodof
infectingandstudyinghumanswasdevelopedwhichwill

Figure2.
Horizontalbarsdenotethepresenceofthesymptomorfinding
notedontheleftofthegraph.Thebottommostbarrefersto
stoolgrade:grade1,formedgrade2,softandgrade3,liquid.
Thenumberofgrade3stoolsinaparticulardayis
denotedbelowthehorizontalbar.

allowfurtherexperiments.Second,forthefirsttimeKoch'spostulateswerefulfilledandGiardiaformallyshownasacauseofdiarrhea.Third,isolatesdifferedintheir
abilitytoinitiateinfections,confirmingtheimportanceofdifferencesamongisolates.Fourth,theimmuneresponseswerebetterdefined,particularlytheuniversalIgM
responseinpresumablyfirsttimeinfections.
Manyofthequestionsansweredandraisedbyexperimentalinfectionsingerbilswerenotansweredinhumaninfections.Thereweremanyobviousdifferences
betweenthehostsandbetweenexperimentalprotocolshowever,biologicaldifferencesamongisolateswereclearlydemonstratedinbothseriesofexperiments.
LiteratureCited
1.Aggarwal,A.,A.Bhatia,S.R.Naik,andV.K.Vinayak.1981.VariablevirulenceofisolatesofGiardialambliainmice.Ann.Trop.Med.Parasitol.77:163167.
2.Aggarwal,A.,andT.W.Nash.1987.ComparisonoftwoantigenicallydistinctGiardialambliaisolatesingerbils.Am.J.Trop.Med.Hyg.36:325332.
3.Belosevic,M.,G.M.Faubert,J.D.MacLean,C.Law,andN.A.Croll.1983.GiardialambliainfectionsinMongoliangerbils:Ananimalmodel.J.Inf.Dis.
147:222226.
4.Bertram,M.A.,MeyerE.A.,AndersonD.L.,andC.T.Jones.1984.AmorphometriccomparisonoffiveaxenicGiardiaisolates.J.Parasit.70:530535.
5.Nash,T.E.,andA.Aggarwal.1986.CytotoxicityofmonoclonalantibodiestoasubsetofGiardiaisolates.J.Immunol.136:26282632.
6.Nash,T.E.,F.D.Gillin,andP.D.Smith.1983.ExcretorysecretoryproductsofGiardialamblia.J.Immunol.131:20042010.
7.Nash,T.E.,D.A.Herrington,G.A.Losonsky,andM.M.Levine.ExperimentalhumaninfectionswithGiardialamblia.J.Inf.Dis.156:974984.
8.Nash,T.E.,andD.B.Keister.1985.Differencesinexcretorysecretoryproductsandsurfaceantigensamong19isolatesofGiardia.J.Inf.Dis.152:11661171.

Page58

9.Nash,T.E.,T.McCutchan,D.Keister,J.B.Dame,andF.D.Gillin.1985.EndonucleaserestrictionanalysisofDNAfrom15Giardiaisolatesobtainedfromman
andanimals.J.Inf.Dis.152:6473.
10.Smith,P.D.,F.D.Gillin,N.A.Kaushal,andT.E.Nash.1982.AntigenicanalysisofGiardialamblia.J.Immunol.6:714719.
11.Smith,P.D.,F.D.Gillin,W.M.Spira,andT.E.Nash.1982.Chronicgiardiasis:studiesondrugsensitivity,toxinproduction,andhostimmuneresponse.
Gastroenterology3:797803.
12.Ungar,B.L.P.,andT.E.Nash.1987.CrossreactivityamongdifferentGiardialambliaisolatesusingimmunofluorescentantibodyandenzymeimmunoassay
techniques.Am.J.Trop.Hyg.37:283289.

Page59

ANIMALMODELSANDCROSSINFECTION

Page61

PrevalenceofGiardiasp.inDogsfromAlberta
PaulD.Lewis,Jr.
DepartmentofBiologicalSciences,UniversityofLethbridge,Lethbridge,Alberta,T1K3M4,Canada.
Atotalof101of1,005straydogsfromfourcitiesinAlberta(10%)wereinfectedwithGiardiasp.asdetectedbysinglestoolexaminations.Prevalencesreportedare33.3%of81
dogsinLethbridge,11.9%of310dogsinCalgary,6.9%of246dogsinRedDeer,and5.4%of368dogsinEdmonton.InCalgaryandEdmonton,whereseparateanimalshelters
areoperatedbythemunicipalitiesandbytheSocietyforthePreventionofCrueltytoAnimals(SPCA),prevalencesofinfectionwerehigherindogsfromSPCAsheltersthan
municipalshelters(8.7%vs3.2%inEdmonton,14.3%vs10.7%inCalgary).AtthepeakofanoutbreakofhumangiardiasisinEdmonton,duringFebruary1983,dogsfromthe
municipalanimalsheltershowedaprevalence5timeshigherthandogsfromthesamesheltertheprevioussummer(17.2%vs3.2%),suggestingthatbothdogsandhumanswere
exposedtothesamesourceofinfection.SurveillanceofprevalenceofGiardiasp.indogsmayserveasamonitorforincipientwaterborneoutbreaksinthehumanpopulation.

Introduction
EpidemicoutbreaksofgiardiasisarewellknownincommunitiesalongtheeasternslopesoftheRockyMountains.AlthoughreportsfirstcentredintheUnitedStates,
particularlyinColorado(22,23),outbreakshavebeenreportedmorerecentlyinCanadafromBritishColumbia,AlbertaandSaskatchewanaswell(7,8,9,28).
Giardialambliafromhumanscandevelopinavarietyofnonhumanhostspecies(14),andbeavershavebeenimplicatedasreservoirhosts(15).Specificpathogen
freebeagleshavebeenusedsuccessfullytotestforinfectivecystsofGiardiasp.frompublicwatersupplies(25),andmongreldogshavebeeninfectedwithboth
humansourcecystsandculturedtrophozoites(16).Aswell,theMongoliangerbilnowseemsestablishedasareliablelaboratorymodelforhumangiardiasis
(4,19,20).Thus,theprevalenceofGiardiaindogsand/orotherpotentialreservoirhostsmayconstituteusefulepidemiologicalinformationwithrespecttohuman
outbreaks.
ThispaperreportstheresultsofacoproanalyticsurveytodeterminetheprevalenceofGiardiasp.instraydogsinLethbridge,Calgary,RedDeerandEdmonton
(Alberta)duringthesummerof1982.Additionally,theoccurrenceofanepidemicoutbreakofhumangiardiasisinEdmonton,peakinginFebruary,1983(9),provided
anunusualopportunitytosamplestraydogsatthattime.
MethodsandMaterials
CitiesSurveyed
StoolsampleswereobtainedbetweenmidMayandmidAugust,1982,fromstrayandunwanteddogsin(2)municipallyoperatedanimalsheltersinthefourlargest
citiesinAlberta:Edmonton(5333'N,11328'W,population560,000),RedDeer(5216'N,11348'W,population50,000),Calgary(5103'N,11405'W,
population620,000),andLethbridge(4942'N,11249'W,population58,000)and(2)sheltersoperatedbytheSocietyforthePreventionofCrueltytoAnimals
(SPCA)inCalgaryandEdmonton(Table1).
CollectionMethodology
AteachparticipatingshelterexceptforLethbridge,wheresampleswerecollectedweeklybytheauthor,shelterpersonnelweretrainedtocollectsinglestoolsamples
fromincarcerateddogs.Eachsamplewasenclosedinanumberedbag,anditsbagnumber,theshelter'sanimalIDnumber,andthelocationofcapturerecordedona
datasheet.ExceptforLethbridge,collectedsamplesandtheirdatasheetsweretransportedtothelaboratoryweeklybycourierforanalysis(usually24to36htransit
time)sampleswererefrigerated(4C)uponarrivaluntiltheywereprocessed(upto96h).
Coproanalyses
Inthelaboratory,sampleswereexaminedoncebyeachof(1)MIF(merthiolateiodineformalin)centrifugation,and(2)ZnSO4flotationprocedures(27).Reference
materialforeachstoolwasretainedaspermanentsmearsstainedwithWheatley'sStain(29)andasaliquotspreservedinvialsofSAF(sodiumacetate,aceticacid,
formalin)preservative(30).
AnalysisofTechniquesEmployed
(1)TechniqueSensitivity:AcomparisonoftheMIFcentrifugationandZnSO4procedureswasmadebydeterminingthepercentofallpositiverecordsobtainedby
eachmethod.
(2)ExaminationofFreshvs.SAFPreservedStoolsbyZnSO4Flotation:SixfreshstoolsfromLethbridgedogsfoundtobeheavilyinfectedwithGiardiasp.were
examinedasfollows:foreachofthe6stools,twoequalportionsweighing0.5g0.02gweretakenthefirstwasexaminedimmediatelybyZnSO4flotation,andthe
secondwasexaminedbythesameprocedureafterpreservationfor18hinSAF.Foreach,6scansat400Xweremadeof22mmcoverglasspreparations,andthe
numbersandappearancesoftrophozoitesandcystswererecorded.
AnalysisofCollectionsDuringtheEdmontonOutbreak
AnoutbreakofhumangiardiasisoccurredinEdmontonbetweenDecember,1982,andApril,1983itreachedapeakinFebruary(9).Stoolsamplesfromatotalof
100dogsfromtheEdmontonmunicipalanimalshelterwerecollectedinvialsofSAFbyshelterpersonnelwhohadbeentrainedtoensurethatpropercollection
methodologywasused.Upontransporttothelaboratory,sampleswereexaminedusingtheZnSO4flotationtechnique.
TABLE1.Summaryofsourcesofcaninestoolsamples

Locality

NumberofSamplesCollected
MunicipalShelters

S.P.C.A.Shelters

Edmonton

218

150

RedDeer

246

Calgary

205

105

Lethbridge

81

AllCentres

750

255

Page62
TABLE2.PrevalenceofGiardiasp.inAlbertadogs*

CITY

MAYJUNE

JULYAUGUST

TOTAL

Positive/Examined

%Prevalence

Positive/Examined

%Prevalence

Positive/Examined

%Prevalence

Edmonton

12/196

6.1

8/172

4.7

20/368

5.4

RedDeer

8/117

6.8

9/129

7.0

17/246

6.9

Calgary

10/114

8.8

27/196

13.8

37/310

11.9

Lethbridge

17/51

33.3

10/30

33.3

27/81

33.3

AllCities

47/478

9.8

54/527

10.2

101/1005

10.0

*Basedupondualanalyses(MIFcentrifugationplusZnSO4flotation)ofsinglestoolsamplesfromstraydogsincarceratedinanimal
shelters.

TABLE3.PrevalenceofGiardiasp.indogsfrommunicipalvs.S.P.C.A.animalsheltersinCalgaryandEdmonton,1982*

CITY

MUNICIPALSHELTERS

SPCASHELTERS

TOTAL

Positive/Examined

%Prevalence

Positive/Examined

%Prevalence

Positive/Examined

Edmonton

7/218

3.2

13/150

8.7

20/368

%Prevalence
5.4

Calgary

22/205

10.7

15/105

14.3

37/310

11.9

BothCities

29/423

6.9

28/255

11.0

57/678

8.4

*SinglesampleswerecollectedfromincarceratedstraydogsfrommidMaythroughmidAugustandwereexaminedbybothMIF
centrifugationandZnSO4flotationtechniques.

Results
Coproanalyses
Inall,10%ofthestoolsofstraydogsfromAlbertacitieswerepositiveforGiardiasp.duringthesummerof1982(Table2).Morphologically,theparasitesappeared
tobecongruentwiththeduodenalistypeasevidencedbycystappearanceanddimensions,andbythepresenceofaclawhammermedianbodyinthosetrophozoites
whichwereobserved.
PrevalencesvariedonlyslightlybetweentheMayJuneandJulyAugustperiodsfromRedDeerandLethbridge.However,forCalgarytheJulyAugustprevalence
increasedby57%comparedtoMayJune,andforEdmontontheJulyAugustprevalencedecreasedby23%comparedtoMayJune.Fortheentireperiod,aswell
asforJulyAugust,prevalencestendedtodecreasefromnorth(Edmonton)tosouth(Lethbridge),butthiswaslessapparentforMayJune.
WhentheresultsforEdmonton,Calgary,andbothcitiescombinedarecomparedbysheltersource(municipalvsSPCA),prevalencesofinfectionwithGiardiasp.in
dogswerehigheratsheltersoperatedbytheSPCAthanatsheltersoperatedbythemunicipalityineachcityandforbothcitiescombined(Table3).Prevalenceof
infectionfordogsattheSPCAsheltersexceededthatfordogsfromthecorrespondingmunicipalsheltersbynearlyonehalfinCalgary,bynearly3foldinEdmonton,
andbyjustoveronehalfforbothcitiescombined.
AnalysisofTechniques
Thesensitivitiesofthetwocoproanalytictechniquesemployedinthisstudywerecompared.Ofthe101positiverecordsobtained,67(66%)weredetectedusingthe
MIFcentrifugationtechnique,and94(93%)usingtheZnSO4flotationtechnique.
AstudywasmadeoftheeffectivenessoftheZnSO4flotationprocedureindetectingGiardiasp.inpaired(samestool)freshandSAFpreservedsamplesfromdogs
knowntobeheavilyinfected.Theflotationprocedurerevealed1to89(average34)cystsperhorizontalscanforfreshstools,whereasitrevealed4to47(average
21)cystsperhorizontalscanforthepairedSAFpreservedsamples.IntheZnSO4coverslipmountsofsomeSAFpreservedsamples,upto1/3oftheparasites
observedweretrophozoites,whereastrophozoitesrarelyweredetectedusingthistechniquewiththepairedfreshstools.Moreover,thecharacteristicmorphologyof
Giardiasp.cystsinstoolspreservedinSAFremainedobservableforuptoseveralhoursinZnSO4flotationcoverslippreparations,whereascystsfromfreshstools
quicklybecameunrecognizableinsuchpreparationsbecauseofcollapseoftheorganismagainsttheinnersurfaceofthecystwall.
PrevalenceofGiardiasp.inDogsDuringaHumanOutbreak
SAFpreservedstoolsamplesfrom100dogsinthemunicipalanimalshelter,Edmonton,wereobtainedduringthepeakoftheoutbreakofhumangiardiasisin
February,1983.ThesewereexaminedbytheZnSO4flotationmethod.Onesamplewasfoundtobeaduplicate,andthereforewasnotcounted.Theprevalenceof
Giardiasp.inEdmontondogsinFebruary,1983,was17.2%(Table4),avalue5timesgreaterthanfordogsfromthesame(municipal)shelter,and3timesgreater
thanforallEdmontondogs,theprevioussummer.

Page63
TABLE4.PrevalenceofGiardiasp.infecesofEdmontondogsduringandpriortoanoutbreakofhumangiardiasis*
Stool
Sources:

Municipal
Shelter

Municipal
Shelter

SPCA
Shelter

Municipal+SPCA
Shelters,1982

Dateof
Samples:

February
1983

Summer
1982#

Summer
1982#

Summer
1982#

Pos/Exam.

17/99

7/218

13/159

20/368

%Prevalence:

17.2

3.2

8.7

5.4

*ThepeakofthehumanoutbreakoccurredinFebruary,1983thepreoutbreakdataarefrom1982.
PrevalenceofinfectionindogsinFebruary,1983,wasdeterminedusingsingleZnSO4flotationexaminationsofsingle
stoolsamplespreservedinSAF.
#PreoutbreakprevalencesweredeterminedusingdualMIFcentrifugationplusZnSO4flotationtechniquesonfreshsingle
stoolsamplesfromstraydogsheldinthemunicipalandSPCAsheltersinEdmonton.

Discussion
TheprevalencesreportedhereforGiardiasp.inAlbertadogsfallwithintherangeofprevalencesreportedelsewhereintheworld(Table5).However,comparisons
betweenstoolanalysissurveysaredifficulttomakebecausethereportedresultscanvaryforavarietyofreasonswhicharequiteadditionaltodifferencesin
endemicitywhichmayoccurthroughtimeorfromareatoarea.
TheincreaseinprevalencesfromnorthtosouthfortheentiresamplingperiodappearstobeanartifactcreatedbythehigherprevalenceofGiardiasp.inCalgary
dogsduringJulyAugust,andbytheconsistentlyhighprevalenceinLethbridgedogsthroughoutthesummer.TheJulyAugustprevalenceforCalgarydogsisnearly
50%higherthanforMayJune(Table2),andwasobservedinanimalsfromboththemunicipalandtheSPCAshelters.Thereasonforthehighprevalencein
Lethbridgedogsisnotknown.TheLethbridgeshelterappearedtobesomewhatlesssanitarythansheltersintheothercitiesbutthefactthatstoolsampleswere
obtainedfromLethbridgedogswithin1to6daysfollowingcapture,atimespanlessthantheusualprepatentperiodforGiardiasp.infectionsindogs(16),suggests
thattheLethbridgedogsalreadywereinfectedatcapture,andthattheprevalenceinLethbridgedogsindeedishigherthanfordogselsewhereintheProvince.
ThehigherprevalencesofGiardiasp.observedindogsfromSPCAsheltersovermunicipalsheltersinCalgaryandEdmonton(Table3)ispuzzling.Bothmunicipal
andSPCAshelterswerecleanandwellrun.ItispossiblethattheSPCAshelterstendedtoholddogsforlongerperiodsoftime,thusincreasingtheopportunityfor
dogtodogtransmission.
Atleastinthislaboratory,theZnSO4flotationtechniqueappearstobemoreefficientthantheMIFcentrifugationtechniqueindetectinginfectionswithGiardiasp.
TABLE5.PrevalencesofGiardiasp.fromdogs.
Reference

Pos./Exam.

%Prevalence

Locality

Hewlettetal.1982

25/37

68

Ohio

Burrows&Lillis1967

99/273

36

NewJersey

Cotteleer&Fameree1980

25/94

27

Belgium

Swan&Thompson1986

70/333

21

Australia

Yang&Scholten1977

96/495

19

Ontario

Pfeiffer&Supperer1976

13/70

19

Austria

Catcott1946

20/113

18

Ohio

deCarnierietal.1964

3/31

10

Italy

14/160

California

158/2063

Minnesota

Levine&Ivens1965

7/175

Illinois

Agrestietal.1977

11/300

Italy

Jungmannetal.1986

5/141

GDR

Hoskinsetal.1982

34/4752

Louisiana

PRESENTSTUDY

101/1005

10

Alberta

Craige1948
Bemrick1961

indogs.ThevalueofSAFasapreservativehasbeennotedpreviously(30),butitsvalueinenablingbothcystsandtrophozoitesofGiardiaspp.toresistthe
distortingeffectsofZnSO4isanewobservation.Itmaybethattheapparentabsenceoftrophozoitesinunpreservedstoolsisexplainedbytheirfragility,whichis
reducedbythepreservative.
AnintermittentpatternofcystreleaseisobservedinnaturalinfectionswithGiardiasp.indogs(3,16),andinexperimentalinfectionswithGiardialambliaindogs
(16),andgerbils(4,20).Thissuggeststhatsurveyssuchasthepresentone,whicharebaseduponsinglestoolcollections,willfailtoidentifyallactiveinfections.
Multiplestoolexaminationswillrevealadditionalinfectionstheuseofaseriesof6consecutivestoolexaminationsresultedinaprevalencerateof68%forGiardiasp.
inapopulationof37straydogsinOhio(16).Furthermore,thecortisoneinducedrecrudescenceofinfectionwithGiardialambliainapparentlyselfcuredgerbilsas
muchas7monthsaftertheirinitialinfectionsuggeststhattheactualnumbersofinfectedindividualsinapopulationmaybehigherthanthatdeterminedbystool
examinationsforcysts(20).Thus,prevalencesreportedinthepresentstudyundoubtedlyareunderstated.
ThereisnopublishedstudyofthehumanoutbreakofgiardiasisinEdmonton,butsummariesindicatethattheoutbreakextendedfromlate1982toApril,1983witha
peakinFebruary,1983(9,28).Atotalof895humancaseswerereported(28),buttheactualnumberofinfectionsundoubtedlywasmuchhigher(9).Thecauseof
theoutbreakcouldnotbedeterminedbytheinvestigators(9).However,anumberofcommunitywideoutbreaksinNorthAmericahavebeenwellstudied,and,where
ithasbeendemonstrated,themodeofinfectioninvariablyhasbeenwaterborne(15,23,25).
Nopreviousstudieshavecomparedtheprevalenceofinfectionindogsorotherpotentialreservoirhostsbeforeandduringacommunitywideoutbreakofhuman
giardiasis.AtthepeakofthehumanoutbreakinEdmonton(February,1983)theaveragenumberofcasesreportedper

Page64

week(about60)was4timeshigherthantheaveragenumberofcasesreportedperweek(about15)forthepreoutbreakmonthofOctober,1982(9).This
compareswellwiththe5foldincreaseintheprevalenceofGiardiasp.indogsfromtheEdmontonmunicipalshelterbetweenthesummerof1982andFebruary,
1983.Inviewofthefailureofinvestigatorstoimplicatepetsasacauseofhumanillness(9),itthusappearsthatbothhumansanddogsinEdmontonwereexposedto
thesamesource,probablythemunicipalwatersupply.
Acknowledgements
ThisresearchwasfundedbyagrantfromtheAlbertaEnvironmentalResearchTrust.Igratefullyacknowledgethecooperationandassistanceoftheparticipating
animalshelterpersonnel:Mr.LorneCreor(Lethbridge),Ms.TrudyDeBecker(Calgary,municipal),Ms.R.Falco(Calgary,SPCA),Ms.K.McLaren(RedDeer),
Mr.R.Wilson(Edmonton,municipal),andMr.BillHartandMs.LoriStingley(Edmonton,SPCA).IndefatigableassistanceinthelaboratorywasprovidedbyMs.
LesleyCurthoys,Ms.PaulaHarperandMr.JoeHarrison.
LiteratureCited
1.Agresti,A.,G.d'Ambrosio,andE.Gravino.1977.Lagiardiasineicane:indagineepizoologicaedosservazionicliniche.ActaMedicaVeterinaria.Napoli23:175.
2.Allison,D.J.1984.Giardiasisarecentinvestigation.Can.J.PublicHealth75:318320.
3.Barlough,J.E.1979.Caninegiardiasis:areview.J.SmallAnimalPractice20:613623.
4.Belosevic,M.,G.M.Faubert,J.D.MacLean,C.Law,andN.A.Croll.1983.GiardialambliainfectionsinMongoliangerbils:Ananimalmodel.J.Infect.Dis.
147:222226.
5.Bemrick,W.J.1961.AnoteontheincidenceofthreespeciesofGiardiainMinnesota.J.Parasit.47:8789.
6.Burrows,R.B.,andW.G.Lillis.1967.Intestinalprotozoaninfectionsindogs.J.Am.Vet.Med.Assoc.150:880883.
7.CanadaDiseasesWeeklyReport.1982a.WaterbornegiardiasisoutbreakAlberta.8:9798.
8.CanadaDiseasesWeeklyReport.1982b.GiardialambliainacommunitywatersupplyBritishColumbia.8:98100.
9.CanadaDiseasesWeeklyReport.1983.GiardiasisEdmonton,Alberta.9:189192.
10.deCarnieri,I.,andS.Castellino.1964.Entamoebacanibuccalis,Trichomonascanistomae,GiardiacanisneicaniaMilano.LaClinicaVeterinaria,Milano
87:193196.
11.Catcott,E.F.1946.Theincidenceofintestinalprotozoainthedog.J.Am.Vet.Med.Assoc.108:3436.
12.Cotteleer,C.,andL.Famere'e.1980.HelminthesetprotozoairesintestinauxparasitesduchienenBelgique.CasparticulierdesEucoccidia.SchweizerArchivfur
Tierheilkunde122:519526.
13.Craige,J.E.1948.Differentialdiagnosisandspecifictherapyofdysenteriesindogs.J.Am.Vet.Med.Assoc.113:343347.
14.Davies,R.B.,andC.P.Hibler.1979.AnimalreservoirsandcrossspeciestransmissionofGiardia,pp.104126.In:W.Jakubowski,andJ.C.Hoff(eds.).
WaterborneTransmissionofGiardiasis.EPAPubl.No.600/979001.
15.Dykes,A.C.,D.D.Juranek,R.A.Lorenz,S.Sinclair,W.Jakubowski,andR.Davies.1980.Municipalwaterbornegiardiasis:anepidemiologicinvestigation.
Ann.Int.Med.92:165170.
16.Hewlett,E.L.,J.S.AndrewsJr.,J.Ruffier,andF.W.SchaeferIII.1982.ExperimentalinfectionofmongreldogswithGiardialambliacystsandcultured
trophozoites.J.Inf.Dis.145:8993.
17.Hoskins,J.D.,J.B.Malone,P.H.Smith,andS.A.Uhl.1982.PrevalenceofparasitismdiagnosedbyfecalexaminationinLouisianadogs.Am.J.Vet.Res.
43:11061109.
18.Jungmann,R.,T.Hiepe,andC.Scheffler.1986.ZurparasitarenIntestinalfaunabeiHundundKatzemiteinemspeziellenBeitragzurGiardiaInfektion.
MonatsheftefurVeterinarmedizin41:309311.
19.Kirkpatrick,C.E.,andJ.P.Farrell.1984.Felinegiardiasis:observationsonnaturalandexperimentalinfections.Am.J.Vet.Res.45:21822188.
20.Lewis,P.D.Jr.,M.Belosevic,G.M.Faubert,L.C.Curthoys,andJ.D.MacLean.1987.CortisoneinducedrecrudescenceofGiardialambliainfectionsin
gerbils.Am.J.Trop.Med.andHyg.36:3542.
21.Levine,N.D.,andV.Ivens.1965.Prevalenceofnematodes,GiardiaandDemodexinIllinoisdogs.IllinoisVet.8:19,2123.
22.Meyer,W.T.1973.Epidemicgiardiasis.Acontinuedelusiveentity.RockyMtn.Med.J.70:4849.
23.Moore,G.T.,W.M.Cross,D.McGuire,C.S.Mollohan,N.N.Gleason,G.R.Healy,andL.H.Newton.1969.Epidemicgiardiasisataskiresort.NewEngland
J.Med.281:402407.
24.Pfeiffer,H.,andR.Supperer.1976.UberdenGiardiabefallderHundeundseinAuftretunginOsterreich.WienerTierarztlicheMonatsschrift63:16.
25.Shaw,P.K.,R.E.Brodsky,D.O.Lyman,B.T.Wood,C.P.Hibler,G.R.Healy,K.I.E.MacLeod,W.Stahl,andM.G.Schultz.1977.Acommunitywide
outbreakofgiardiasiswithevidenceoftransmissionbyamunicipalwatersupply.Ann.Int.Med.87:426432.
26.Swan,J.M.,andR.C.A.Thompson.1980.TheprevalenceofGiardiaindogsandcatsinPerth,WesternAustralia.Aust.Vet.J.63:110112.
27.UnitedStatesNavalMedicalSchool.1965.MedicalProtozoologyandHelminthology.U.S.GovernmentPrintingOffice,Washington.238pp.
28.WeeklyEpidemiologicalRecord.1984.Giardiasissurveillance.No.15:114115.
29.Wheatley,W.B.1951.Arapidstainingprocedureforintestinalamoebaeandflagellates.Am.J.Clin.Path.21:990991.
30.Yang,J.,andT.Scholten.1977.Afixativeforintestinalparasitespermittingtheuseofconcentrationandpermanentstainingprocedures.Am.J.Clin.Path.
67:300304.

Page65

LocationofGiardiaTrophozoitesintheSmallIntestineofNaturallyInfectedDogsinSanDiego
HerndonDouglas,DavidS.Reiner,MichaelJ.GaultandFrancesD.Gillin*
DepartmentofPathologyH811F,UniversityofCalifornia,SanDiegoMedicalCenter,225DickinsonStreet,SanDiego,CA92103,U.S.A..
Sincelittleisknownoftheanatomicallocusoftrophozoitesinnaturallyinfectedhumansorcanines,weascertainedthepresenceanddistributionofGiardiainthesmallintestine
(SI)of29apparentlyhealthyadultdogswhichwereeuthanizedforotherstudies.Of13dogssampledbyaspiratingduodenaljejunalfluid,only3(23%)werepositive.Incontrast,
Giardiawasfoundin12of16dogs(75%)whentheentireSIwasexamined.ThetrophozoitedistributionwasdeterminedbyligatingtheSIat12inchintervalsandenumerating
Giardiainbothsalinewashesandmucosalscrapings.Threepatternswereobserved:(1)Broaddistribution,withtrophozoitesin>90%oftheintestinalsegments(2)Intermediate,
withtrophozoitesin~50%ofthesegments(3)Restricted,withtrophozoitesinonly~20%oftheSI.Inthelatterclass,Giardiawasfoundintheupperfirstandsecondsegmentsor
thelowesttwosegments.Thus,innaturallyinfecteddogs,GiardiacancolonizeanywhereintheSI.Thetotaltrophozoiteburdenwasalsoextremelyvariable:3105to6.9108
perdog.Fewcystswereobservedinthececalcontents.Thisresemblesaformofhumangiardiasisinwhichcystsarenotreadilydetectableinfeces,buttrophozoitesmaybefound
byaspiration.

Introduction
DespitetheprevalenceofgiardiasisintheUnitedStatesandlessdevelopedcountries,(5,19)littleisknownaboutthemechanismorfrequencyofcolonizationofthe
humansmallintestinebyGiardialamblia.Duodenalandupperjejunalfluidaspiration,''stringtests",andbiopsiesarefrequentlyusedindiagnosis(reviewedin19)
whenfecalcystsarenotdetected.Thesemethodsarebasedontheassumptionthattrophozoitescolonizetheuppersmallintestine,butthereisnodetailedinformation
onthedistributionoftrophozoitesinthehumansmallintestine.ThepresentstudywasundertakentoelucidatethelocationofGiardiatrophozoitesintheintestinal
tractsofnaturallyinfectedadultdogs,asamodelofthedistributioninpeople.
Caninegiardiasismaybeausefulmodelofhumaninfectionforseveralreasons:
(i)ExperimentalinfectionofdogswithGiardiafromhumanshasbeenreported(6,12).
(ii)AlthoughthehostspecificitiesandspeciesdefinitionsofGiardiafromhumansanddogsarenotclear(11,13,seealso14),bothareoftheduodenalis,or
intestinalismorphologictype,with"clawhammer"medianbodies(13).
(iii)Bothsymptomaticandasymptomaticgiardiasisarewelldocumentedindogs(15)andhumans(19).
(iv)Boththenumbersandfrequencyofcystsexcretedinfecescanbeextremelyvariableindogs,asinhumans.Oftentrophozoitescanberecoveredfromthesmall
intestinewhencystsarenotdetectedinthestool(3,15).
(v)Ithasbeenproposed(12)thatdogsmaybeareservoirfordisseminationofGiardiacyststohumanssincefecalcontaminationiscommonbothnearthehomeand
inwildernessareas(Seealso4,20,21forcriticaldiscussionsofcrossspeciestransmission).Moreover,recentstudieshavedemonstratedhighincidences(38%and
68%)ofgiardiasisinapparentlyhealthydogsfrompetshops(17)andananimalshelter(12).
Ininitialstudies,weattemptedtoidentifyinfecteddogsbyexaminationoffreshstools.Veryfewcystswereobservedeitherinsmearsorpurifiedpreparations(see
methods).Therefore,weascertainedthepresenceanddistributionofGiardiatrophozoitesinthesmallintestinesof29unselected,consecutive,apparentlyhealthy
adultdogssacrificedforotherstudieswhichinvolvedneithertheintestinenordrugsthatmightaffecttheparasites.Boththenumberandpatternofcolonizationof
trophozoiteswereextremelyvariable.Giardiawereobservedinfluidwithdrawnbyneedleaspirationfrom23%ofanesthetizeddogs.Incontrast,Giardiawere
foundin75%ofanimalsinwhichtheentiresmallintestinewasexamined.Therefore,thediagnosismaybemissedbysamplingonlytheuppersmallintestine.
MaterialsandMethods
AllanimalsusedwereacquiredfromtheSanDiegoPoundforotherstudiesandkeptinthevivariumforthreetofivedaysbeforeuse.Duringthisisolationperiod,no
medicationorspecialprecautionswereusedtoaltertheintestinalflora.DogswerefedIamsChunksdogfoodandwateredaccordingtoUniversitypolicy.Foodwas
withheldthenightbeforeexperiments.Weobtainedalldogsimmediatelyaftertheotherexperimentswerecompleted.
*Correspondingauthor.

Page66

Thirteendogswereacquiredfromamedicalstudentteachinglab(UCSD)duringastudyofcardiacdrugreactions.Giardiaweresoughtbyinsertingan18gauge
needleintotheduodenumofananesthetizeddog,withdrawingthesmallvolumeoffluid,andexaminingitmicroscopicallyinahemacytometerchamberforGiardia.
Trophozoiteswereobservedinfluidfromonlythreedogs.Theentireintestinesoftwodogswereprocessedbymethod1below(dogs1and2,Table1).An
additionalsixteendogswereeitherdonorsforlivertransplantsbytheSurgeryDepartmentorusedbythePulmonaryDivisionforcardiacoutputandbloodgasstudies.
Immediatelyaftersacrifice,theanimalsweresampledbytwomethods.Methodoneconsistedofligatingthesmallintestinewith"O"silk(Ethicon)intoapproximately
twelveinchsections,startingimmediatelybelowthepylorusandendingjustbelowthececum.Afterligationwascompleted,thegutwasremovedenbloc.Each
sectionwasfilledwith30mLofsterilepyrogenfreesaline(Travenol),rinsedthoroughly,andthefluidemptiedintoa50mLtube.Thesectionswerethenopened
lengthwiseandscrapedgentlywithwoodentonguedepressorstorecoverthemucus.TenmLofcoldnormalsalinewasaddedtoallmucuspreparations.Giardiain
thefluidandmucuspreparationswerecountedinhemacytometerchambers(fivecountswereaveraged)andalsoanalyzedbyindirectimmunofluorescence(IIF).This
analysiswasusedfor2dogsidentifiedaspositivebyaspirationand8dogswhichwerenotprescreened(dogs110,Table1).
Methodtwoconsistedofseparatingtheentiresmallintestineintofour(dogs1117)orfive(dog18)equalsegmentswhichwerethenprocessedandquantifiedas
above,againisolatingthececumasaunitbeforeremoval.Insomecases,Giardiacystsandtrophozoiteswerepurifiedfromsamplesofstoolorcecalcontentsby
suspendingupto20goffecesincoldsalinesolution,filteringthroughcheesecloth,passingoveraSephadexG50column(7)andwashing3timesatlowspeed.Cysts
andtrophozoiteswerequantifiedbyhemacytometercountsasaboveorIIFasbelow.
IndirectImmunofluorescence
IIFwasusedtodetectandcountGiardiaintheviscousmucuslayersincetheywerenotreadilyvisibleotherwise.Thelumenalphasewasanalyzedinthesameway,
sothatthenumbersofparasitesineachcompartmentcouldbecompareddirectly.Samples(20L)ofintestinalfluidormucuswerespottedon8wellslides,airdried
andfixedforIIFwith1%Formalinindistilledwaterfor10minutesatroomtemperature.After3distilledwaterwashes,thecellsweretreatedwithacetonefor10
minutesatroomtemperatureanddried.Rabbitantiserumtopurifiedhumancysts(10)orsonicatedculturedstrainWBtrophozoites,dilutedinPBScontaining1%
bovineserumalbumin(BSA)and1%(v/v)Tween20,wasaddedtoeachwell(20L)andtheslideswereincubatedat37Cinahumidifiedchamber(20minutes).
Afterextensivewashingwithdistilledwaterandairdrying,20LofFITClabelledgoatantirabbitIgG(Sigma)wereaddedtoeachwell.Thisconjugatewasdiluted
1:800inPBSBSATween20withethidiumbromide(2g/mL)asacounterstain.Afterfurtherextensivewashingwithdistilledwaterandairdrying,theslideswere
mountedinpH9glycerol.Thenumberofparasites/fieldwasthemeanoffivefields.
ResultsandDiscussion
IncidenceofGiardiasisinDogs
Inpreviousstudies(seereference1forreviewofliteratureupto1979),theincidenceofgiardiasisindogsrangedfrom0.6%to67%(12).Thesedifferencesmay
reflectactualvariationinfrequencyofinfectionbutourdatasuggestthattheymaybedue,atleastinpart,tothemethodsofdeterminingthepresenceofinfection.We
foundveryfewcystsinfecesofhealthyadultdogs,butdidnotrelyonthisasascreeningmethod.Theincidencesofinfectionweobservedweregreatlyaffectedby
thesamplingmethod.
Wefirstassayedforthepresenceofgiardiasisin13dogsbyexaminingupperintestinalfluidaspiratesandfoundonlythreepositive.Thedistributionoftrophozoites
alongthesmallintestinewasdeterminedintwoofthesedogsandfoundtobequitebroad,withtrophozoitespresentbothintheloweranduppersmallintestine.Since
weobservedtrophozoitesinthelowerintestine,weproposedthatinfectionsmightnotalwaysbemanifestintheuppersmallintestineandwouldnotbedetectedby
aspirationatthatsite.Therefore,weexaminedtheentiresmallintestineinthenext16unselecteddogsandfound12positive(p<0.01,Fisherexactprobabilitytest).
Wedidnotobserveloosestoolinthececumofanydog.
Twomethodsofenumerationwereused,asdescribedinMethods.Directhemacytometercountsoftrophozoitesinthefluidphasewashyieldanestimateofthe
numberofparasitespersegment,sincethefluidvolumeisknown.However,thefluidphaseisactivelybeingmoved"downstream"andmaynotrepresentthelong
termpatternofcolonization.Anypostmortemchangesindistributionoftrophozoiteswouldalsoaffectthefluidphase.Incontrast,Giardiatrophozoitesassociated
withthemucusblanketprobablyreflectstableassociations.Sinceitisdifficulttocountthetrophozoitesintheviscousmucuspreparation,weusedantibodiesagainst
trophozoitestovisualizethembyIIF.Thismethoddoesnotyieldabsolutenumbersofparasitesandcannotbecompareddirectlywiththehemacytometercounts.
Accordingly,wealsousedIIFanalysisofthetrophozoitesinthefluidphaseforcomparison(meancells/field)betweenfluidandmucusphasesandfromanimalto
animal.
TheNumbersandDistributionofGiardiaTrophozoitesintheSmallIntestine
Thesmallintestinesoftendogswerecutintosegmentsofapproximately12inchesandthenumbersofparasitesdeterminedasdescribedinMethods.Sincethe
numberofsegmentsvariedfromdogtodog,theparasitedensitieswereplottedintermsofthepercentofdistancealongthesmallintestinewiththepylorusas0%and
thececumas100%forcomparisonsofthedistributionsfromanimaltoanimal.
Boththenumberofparasitesperdogandtheirdistributionwereextremelyvariable.Thenumberoftrophozoitesinthefluidphaserangedfrom0.3106to692.0
106perdog.Theproportionoftheintestinecolonizedvariedfrom24%to100%(Table1,lastcolumn).Interestingly,ofthetwodogswithtrophozoitesineach
segment(Table1),dog#1hadthegreatesttotalnumberoftrophozoites(>6.9108),whiledog#17hadrelativelyfew(<2106).Thesestudiesindicatethatthe
parasiteburdendoesnotcorrelatewiththedistributionofGiardiatrophozoites.Inaddition,thelocationofthesectionwiththegreatestnumberofparasitesvaried
greatly.The"peak"sectionwasintheupperthirdofthesmallintestineinthreedogs(dogs1,2,3)whileitwasinthelowestsegmentinfourdogs(dogs7,15,17,18).
IllustrativedistributionsofGiardiatrophozoitesareshowninFigures1to4.Dog#1hadthehighestparasiteburdeninthisstudyandtrophozoiteswererecovered
fromeachsegmentofthesmallintestine(Figure1).Thedistributionwassomewhatbimodalwithpeaksintheupperand

Page67
TABLE1.Totalandpeakparasiteburdensinintestinalsegmentsa

Dog
#

Total
Trophozoites
(106)

NumberandPercentageof
TrophozoitesinPeakSegment

106(%)

TerminusofPeak
Segment(%distance
#PositiveSegments/
frompylorusto
Totald
secum)

692.0

273.8

(40)c

33

9/9

38.9

9.4

(24)

30

8/10

7.6

5.6

(74)

4/12

33.8

10.4

(31)

50

10/14

15.7

4.6

negativeb

9.1

5.5

2.6

(29)

62

12/13

negative

0/9

(60)

100

3/8

0.9

(35)

56

4/9

19.8

10.6

(53)

45

7/11

10

19.2

8.4

78

5/9

11

negative

negative

negative

0/4

12

negative

negative

negative

0/4

13

negative

negative

negative

0/4

14

0.3

0.3

(100)

75

1/4

15

36.9

27.0

(73)

100

3/4

16

1.3

0.9

(69)

100

2/4

17

1.5

0.6

(40)

100

4/4

18

137.6

78.6

(57)

60

3/5

negative

(44)

#aAlldataarebasedonthemeanof5hemacytometercountsofparasitesinthefluidphaseofsegmentsofthesmall
intestine(pylorustocecum).
#bAnegativesegmenthas<3.3104trophozoites(thelimitofdetectionofthisassay).Anegativedoghasnopositive
segment.
#cNumbersinparenthesesarethenumbersoftrophozoitesinthepeaksegmentdividedbythetotalnumberoftrophozoites
(100).
dApositivesegmenthas 105trophozoites.

lowerjejunum.
Dogs#3and#7illustratetheotherextreme,averynarrowdistributionofaGiardiatrophozoites.Indog#3(Figure2),thepeakparasiteburdenwasinthe
uppermostsegment(duodenum).Incontrast,indog#7,thepeaksegmentwasthelowest(ileumcecum)andnoparasiteswereobservedintheupper60%ofthe
smallintestine(Figure3).Thepatternoftrophozoitedistributionwasintermediateindog#9(Figure4).
ThedistributionpatternsdeterminedbythethreeanalyticmethodswereroughlyparallelexceptthatnotrophozoitesweredetectedbyIIFinthelumenalcontentsof
dog

Figure1.
BroaddistributionofGiardiatrophozoites:dog#1.

#7(Figure3).Giardiawereobserved,however,inthemucuslayerbyIIF,aswellasinthelumenalphasebyhemacytometercounts.Thisdiscrepancywasalso
observedinthelowestsegmentofdog#1.Hemacytometercountsofthissegmentwerepositive.ApossiblereasonforthenegativeIIF,whentrophozoitesare
detectablebyhemacytometer,isthatbacteriaortheirproductsinthelowersmallintestinemaydigestantigensdetectedbyIIF.Ingeneral,thedistributionofGiardia
inthemucusphasewassimilartothatinthelumen,suggestingthattrophozoitesinthemoremobilefluidphaseusuallyreflect

Figure2.
RestricteddistributionofGiardiatrophozoites,mainlyanterior:dog#3.

Page68

Figure3.
RestricteddistributionofGiardiatrophozoites,posterior:dog#7.

Figure4.
Intermediatedistributionoftrophozoites:dog#9.

colonizationofthemucuslayer.Wehavereportedthatintestinalmucuspromotesattachment(21)andgrowth(9)ofG.lambliatrophozoitesinvitro.
Manyfactorsmayinfluenceboththeincidenceofgiardiasisandtheintestinaldistributionoftrophozoites.Inasurveyof>2,000dogs,7.66%ofzincsulfate
concentratesofstoolsampleshadGiardiacysts(2).Itwasstrikingthat70%ofthepositivedogswereyoungerthan6monthsand~90%lessthanoneyear(2).
Similarly,inendemicareas,giardiasisismoreprevalentinhumanchildrenthaninadults.
DifferinglociofGiardiacolonizationindogshavebeenreportedpreviously.Faust(8)reportedthatinanunspecifiednumberofnaturallyinfecteddogs,"theprimary
seatoftheorganismisthececumandappendix"andoccasionallyintheterminalileum.Similartrophozoitedistributionswerefoundin11dogsinfectedrectallywith
trophozoites.Onlythreedogshadtrophozoitesintheanteriorileumorabove(8).Tsuchiya(18)infectedfourpuppiesorallywithGiardiacystsfromdogs.Most
trophozoiteswerefound10to20inchesbelowthepylorusinthetwopuppiesfedahighcarbohydratediet.Trophozoiteconcentrationswereverymuchlowerinthe
twopuppiesfedahighproteindietandthepeakincidencewas30to35inches(intestinelength~55inchestocecum).Cystswereobservedinthelowerhalfofthe
lowerintestineaswellasinthececumandcolon(18).
Theseobservationsarecongruentwithourstudiesofalargergroupofunselecteddogsofunknownage(apparentlyfullgrown),diet(atleastpriortoarrivalatthe
pound),anddurationofinfection,inwhichtrophozoitescolonizedanywhereinthesmallintestine.
OurdetailedstudiesofgiardiasisinapparentlyhealthyadultpounddogsinSanDiegoreveal:(i)ahighincidenceofinfection,(ii)fewcysts,(iii)highlyvariablepatterns
oftrophozoitecolonization.Infectioninthesedogsresemblesgiardiasisintwogroupsofhumansinwhomcystsarenotfoundinthefeces.Inthefirstgroup,
trophozoitesarefoundbyintestinalaspiration,biopsy,orstringtest(19).Inthesecondgroup,intestinalaspirationisnegative,butsymptomsdisappearinresponseto
antigiardialtherapy(19).Therapeuticresponsetobroadspectrumantiprotozoancompoundsdoesnotprovethediagnosisofgiardiasishowever,andthetrueetiology
ofinfectioninthesepatientsusuallyremainsproblematic.OurobservationthatGiardiatrophozoiteswererestrictedtothelowersmallintestineincertaindogs
suggeststhatsomeofthesepeoplemaybeinfectedwithGiardiatrophozoitesonlyinthelowerintestine.
Acknowledgements
ThisworkwassupportedbyNIHAI19863andEPA#811950010.WearegratefultoC.Davis,L.Corbeil,J.Sauch,andW.Jakubowskiforcriticalreadingof
themanuscript,toS.McFarlinforexcellenttyping,andtokeyUCSDpersonnelforhelpinobtainingdogs.
ThisdocumenthasbeenreviewedinaccordancewiththeU.S.EnvironmentalProtectionAgencypolicythroughassistanceagreementnumber811950010to
UniversityofCaliforniaatSanDiegoandapprovedforpublication.Mentionoftradenamesorcommercialproductsdoesnotconstituteendorsementor
recommendationforuse.
LiteratureCited
1.Barlough,J.E..1979.Caninegiardiasis:Areview.J.SmallAnim.Pract.20:613623.
2.Bemrick,W.J..1961.AnoteontheincidenceofthreespeciesofGiardiainMinnesota.J.Parasitol.47:8789.
3.Bemrick,W.J..1963.ObservationsondogsinfectedwithGiardia.J.Parasitol.49:10311032.
4.Bemrick,W.J..1984.Someperspectivesonthetransmissionofgiardiasis.In:GiardiaandGiardiasis:Biology,Pathogenesis,andEpidemiology.Erlandsen
S.L.andE.A.Meyer,eds.PlenumPress,NewYork.pp.379400.
5.Craun,G.F..1984.Waterborneoutbreaksofgiardiasis.In:GiardiaandGiardiasis.ErlandsenS.L.andE.A.Meyer.PlenumPress,NewYork.pp.243261.
6.Davies,R.B.andC.P.Hibler.1979.AnimalreservoirsandcrossspeciestransmissionofGiardia.In:WaterborneTransmissionofGiardiasis.JakubowskiW.
andJ.C.Hoff,eds..U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio.pp.104126.

Page69

7.Douglas,H.,ReinerD.S.andF.D.Gillin.1987.AnewmethodforpurificationofGiardiacysts.Trans.Roy.Soc.Trop.Med.Hyg..81:315316.
8.Faust,E.C..1931.HabitatofGiardiaintheintestine.Proc.Soc.Exp.Biol.Med.528:621623.
9.Gault,M.J.,Gillin,F.D.andA.J.Zenian.1987.Giardialamblia:Stimulationofgrowthbyhumanintestinalmucusandepithelialcellsinserumfreemedium.Exp.
Parasitol..64:2937.
10.Gillin,F.D.,Reiner,D.S.,Gault,M.J.,Douglas,H.,Das,S.,Wunderlich,A.andJ.Sauch.1987.EncystationandexpressionofcystantigensbyGiardialamblia
invitro.Science.235:10401043.
11.Hegner,R.W..1922.AcomparativestudyoftheGiardiaslivinginman,rabbitanddog.Am.J.Hyg.2:422454.
12.Hewlett,E.L.,Andrews,J.S.,Ruffier,J.andF.W.SchaeferIII.1981.ExperimentalinfectionofmongreldogswithGiardialambliacystsandcultured
trophozoites.J.Infec.Dis.145:8993.
13.Levine,N.D..1979.Giardialamblia:Classification,structure,identification.In:WaterborneTransmissionofGiardiasis.JakubowskiW.andJ.C.Hoff,eds..
U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio.pp.28.
14.Nash,T.E.,McCutchan,T.,Keister,D.,Dame,J.B.,ConradJ.D.andF.D.Gillin.1985.RestrictionendonucleaseanalysisofDNAfromfifteenGiardiaisolates
obtainedfromhumansandanimals.J.Infec.Dis.152:6473.
15.Pitts,R.P.,Twedt,D.C.andK.A.Mallie.1983.Comparisonofduodenalaspirationwithfecalflotationfordiagnosisofgiardiasisindogs.J.Am.Vet.Med.
Assoc.182:12101211.
16.Sogayar,M.I.L.,Curi,P.R.,andE.F.daSilva.1987.GiardiacanisHegner,1922.Localizacaonotubodigestivadecaesnaturalmenteinfectados.Arq.Bras.
Med.Vet.Zool.39:265272.
17.StehrGreen,J.K.,Murray,G.,Schantz,P.M.andE.RuizTiben.1985.Intestinalparasitesinpetstorepuppies.Abstract#90,Amer.Soc.Trop.Med.Hyg.
34thMeeting,Miami,Florida,U.S.A.Nov.3,1985.
18.Tsuchiya,H..1931.ThelocalizationofGiardiacanis(Hegner,1922)asaffectedbydiet.Am.J.Hyg.15:232246.
19.Wolfe,M.S..1984.Symptomatology,diagnosisandtreatment.In:GiardiaandGiardiasis:Biology,Pathogenesis,andEpidemiology.ErlandsenS.L.and
E.A.Meyer,eds..PlenumPress,NewYork.pp.147162.
20.WooP.K.,1984.EvidenceforanimalreservoirsandtransmissionofGiardiainfectionbetweenanimalspecies.In:GiardiaandGiardiasis:Biology,
Pathogenesis,andEpidemiology.ErlandsenS.L.andE.A.Meyer,eds..PlenumPress,NewYork.pp.341364.
21.Woo,P.K.andW.B.Patterson.1986.GiardialambliaindaycarecentresinsouthernOntario,Canada,andsusceptibilityofanimalstoGiardialamblia.
Trans.Roy.Soc.Trop.Med.Hyg.80:5659.
22.Zenian,A.andF.D.Gillin.1985.InteractionsofGiardialambliawithhumanintestinalmucus:Enhancementoftrophozoiteattachmenttoglass.J.Protozool.
32:664668.

Page71

SeasonalIncreaseintheIncidenceofGiardialambliainArkansas
JamesJ.Daly,MarkA.Gross*,DavidMcCullough,ThomasMcChesney,SuzanneK.Tank,EleanoraB.Daly,andCherylL.Puskarich.
DepartmentofElectronicsandInstrumentation,UniversityofArkansasatLittleRock,GraduateInstituteofTechnology,LittleRock,AR72204,U.S.A..
AseasonalincreaseforthepresenceofGiardialambliainstoolexaminationswasfoundbythreeclinicallaboratoriesincentralArkansas.DatafromtherecordsoftheArkansas
StateHealthDepartment(19831985),ArkansasChildren'sHospital(19831985),andSt.VincentInfirmary(19801986)showthatstoolspecimenspositiveforG.lambliabeginto
increaseinmidsummer,peakinSeptember,anddecreasethereafter.Geographicoriginsofthepatientswhosestoolswereexaminedattheseinstitutionsareprimarilyfromthe
LittleRockNorthLittleRockmetropolitanareasandthemountainousregionsofthestate.Theseasonalepidemictypeincreaseis,inpart,dependentonthetotalnumberofstools
examinedpermonthbutthereasonfortheincreaseinthepercentageofpositivestoolsisnotknown.AgesofpatientswerenotavailablefromtheArkansasStateHealthDepartment
andSt.VincentInfirmarybutdatafromArkansasChildren'sHospitalshowthattheseasonalincreaseoccursinthepediatricpopulation.

Introduction
AreviewoftherecordsoftheparasitologyserviceoftheArkansasStateHealthDepartmentin1985revealedanoticeableincreaseinthenumberofstoolspositive
forGiardialambliaduringthemonthsofAugust,September,andOctober.Furtherexaminationoftheserecordsshowedthatthisseasonalincreaseoccurredevery
yearfrom19831985.Sincesuchaseasonalincreasecouldhavebothepidemiologicandhealthcaresignificance,itwasdecidedtoinvestigatethisfindingfurtherby
examiningtherecordsoftwootherlaboratoriesincentralArkansas.Selectionoftheselaboratories,ArkansasChildren'sHospitalandSt.VincentInfirmary,was
basedonthedifferentsocioeconomiccharacteristicsofthepatientsseenateachinstitutionandtheavailabilityoflaboratorydata.
EndemicityofGiardialambliainArkansas
GiardialambliaiscurrentlythemostcommonlyreportedentericprotozoanparasiteinArkansas.Itspresenceinthestatehasbeenknownformanyyears.An
averageof37positivestoolsperyearwerefoundbytheoldStateHygieneLaboratorybetweentheyears1954and1963(2).In1978,67G.lambliapositivestools
werefoundbytheStateHealthDepartmentLaboratory,whichrepresented4.1%ofthetotalstoolsexaminedbytheState.Themidsouthregionalpercentfigurefor
G.lambliapositivestoolsatthattimewas5.7andthenationalfigurewas3.8%(8).Sincethenanaverageofabout5%hasbeenfoundforallstoolsexaminedbythe
StateLaboratory.GiardiahadalsobeenfoundinapotentialanimalreservoirinArkansas.In1985aprevalencerateof11.5%wasreportedin78beaverscollected
fromdiverseareasofArkansas.AnimalsthatwerefoundpositiveforGiardiahadbeentrappedfromsevendifferentcounties(6).
InstitutionalProfiles
TheparasitologyserviceoftheArkansasStateHealthDepartment(AHSD)performsapproximately2200fecalexaminationsayear.Thestoolsexamineddonot
representafairsamplingofthestate'spopulationandthetestresultscannotbeusedasanaccurateassessmentofaparasite'sincidenceinArkansas.Becauseofthe
centrallocationofASHD,manyofthestoolssubmittedarefromtheLittleRockNorthLittleRockarea.Until1986theUniversityHospitaloftheUniversityof
ArkansasforMedicalSciencesutilizedtheparasitologyserviceoftheASHDandthisrepresentedaconsiderableproportionofthestoolsreceivedfromthecentral
Arkansasarea.OutsideofcentralArkansastheserviceisprimarilyusedbyphysicianswhodonothavereadyaccesstoagoodparasitologylaboratory.Stoolsare
infrequentlyseenfrompatientslivinginornearmanyofthelargercitiesinArkansas.ThepercentdistributionbygeographicregionofstoolsdonebyASHDina
typicalyearare:LittleRockNorthLittleRock,48%OzarkMountains,25%MississippiPlains,16%OuachitaMountains,7%andGulfCoastPlains,4%.
ArkansasChildren'sHospital(ACH)isanonprofitpediatrichospitalassociatedwiththeUniversityofArkansasforMedicalSciences.Theparasitologyserviceof
ACHperformsfrom1000to1200stoolexaminationsperyear.Becauseofitspediatricemphasistherelativenumberofstoolspositiveforparasitesishigherthanat
theothertwoinstitutionsinthisstudy.GeographicdataonpatientswassimilartoASHDandSt.VincentInfirmarywithheavyusagebycentralArkansasresidents
(55%).
St.VincentInfirmaryisoneofthelargest(600bed)privatehospitalslocatedincentralArkansas.Patientstendtobehigherineconomicstatusthanpatientsutilizing
theothertwofacilities.Theparasitologyserviceoftheclinicallaboratoryperformsfrom1500to1800examinationsperyear.DuetothetypeofpatientsatSt.
VincentInfirmarythepercentpositivestoolsforparasitesisverylow,relativetotheothertwoinstitutions.
*Correspondingauthor.

Page72

Figure1.
MonthlyincidenceofGiardiapositivestoolspecimensfoundby
theArkansasStateHealthDepartmentandArkansas
Children'sHospitalLaboratoriesfortheyears19831985.

Methods
DatawereobtainedfromtherecordsoftheparasitologyservicesoftheclinicallaboratoriesofASHD,ACH,andSt.VincentInfirmary.Allthreelaboratoriesutilize
thePVAfixationtrichromestainingandformalinethylacetateconcentrationiodinestainingproceduresforexaminationofstoolspecimens.CasespositiveforG.
lambliawererecorded,withrepeatorduplicatepositivesonthesameindividualbeingignored,ifoccurringwithinasixweekperiod.Patient'sclinicalhistorieswere
notconsulted,thereforepositivestoolsinthisstudyonlyreflectincidenceoftheparasiteandnottheclinicalstatusofthegiardiasis.Statisticalmethodsforcorrelation
andtheuseoftheWilcoxonTwoSampleTest(anonparametricanalogofanalysisofvariance)arefromSokolandRohlf(12).
Results
AnalysisofdatafromtherecordsofASHDandACHforathreeyearperiodshowayearlycycleinwhichanincreaseinthenumberofG.lambliapositivestools
beganinmidsummer,peakedinornearSeptember,anddecreasedthereafter(Figure1).Thiseffectisevenmoreevidentifthedataforeachmonthforthethreeyear
periodareaveraged(Figure2).InFigure3aredatafromSt.VincentInfirmaryinwhichthedataforeachmontharetotaledforasixyearperiod(19801986).The
relativelyfewernumbersofpositivestoolsseenatSt.VincentprecludedanindividualmonthlyaverageaswasdoneforASHDandACH.Datafromallthree
institutionsclearlyshowthatthecyclicalincreaseisnotpeculiartoonelaboratorynortoonesetofpatients.

Figure2.
AveragemonthlypercentofGiardiapositivestoolspecimensfoundattheArkansas
StateHealthDepartmentandArkansasChildren'sHospitalfortheyears19831985.

Page73
TABLE1.MonthlyincidenceofG.lambliainstoolsexaminedbytheArkansasStateHealthDepartmentfortheyears
1983to1985
AverageNo.of
SpecimensSubmitted

AverageNo.ofPositive
Specimens

PercentPositive

January

190.037.8*

6.35.5*

3.3

February

175.353.g

6.74.9

3.8

March

212.775.1

6.06.0

3.1

April

151.317.6

6.03.0

4.0

May

175.329.3

6.03.6

3.4

June

171.341.0

4.34.1

2.5

July

20835.4

11.33.2

5.4

August

245.771.5

13.72.9

5.6

September

215.360.g

17.72.1

8.3

October

232.364.8

13.03.5

5.6

November

186.738.1

5.02.0

2.7

December

116.714.2

3.01.0

2.6

Month

*MeanS.D.

Table1showstheaveragenumberofstoolexaminationsdonepermonthandtheaveragenumberofpositivestoolsforeachmonthforathreeyearperiodatthe
ASHD.TheWilcoxonTwoSampleTestwasusedtocomparethemeannumberofpositivesinthemonthscomprisingtheseasonalincrease(July,August,
September,andOctober,n=12)withthemeansoftheothermonths(n=36).Asignificantdifferenceisfoundbetweenthenumberofpositivestoolsseenduringthe
peakmonthsandtheothermonthsoftheyear(p=<0.001).Itwasconsideredthattheincreaseinthetotalnumberofstoolsduringthepeakofthecyclicalperiod
mightbeafactorintheincreaseinthenumberofpositivestoolsfound.SucharelationshipwouldindicateanenhancementintheprobabilityoffindingG.lamblia.The
numberofstoolexaminationsdonepermonthwashighlyvariable,asindicatedbythelargestandarddeviations,buttherewasasignificantcorrelationbetweenthe
numberofpositivetestsandthetotalnumberofstoolssubmitted(r=0.61,p=<0.01).

Figure3.
Thetotalnumberofpositivestoolspecimensfound
atSt.VincentInfirmaryforeachmonthfrom19801986.

ThisindicatesthattheincreasednumberofpositiveGiardiastoolspecimensmay,inpart,bearesultoftheincreaseinthetotalnumberofstoolssubmittedduringthe
seasonalperiodofpeakincidence.Thecorrelation,however,cannotaddressthequestionofwhymorepositivestoolsclusteraroundthemonthofSeptember.
EvaluationofpatientdatafromASHDbasedonregionandthepresenceofG.lambliarevealedthefollowingpercentageofpositivestoolsbyphysiographicregion
forathreeyearperiod(19831985):LittleRockNorthLittleRock,5.8%OzarkMountain,5.3%Ouachita,4.3%MississippiPlain,4.0%andGulfCoastPlain,
6.8%.FromthesevaluesthereappearstobenoobviousregionalconcentrationofG.lamblia.Thisdataalsoconfirmsanearlier,lessspecificreport(9)showing
statewidedistributionoftheparasite.
Discussion
EpidemicsofgiardiasisintheUnitedStateshavebeenprimarilyassociatedwithcontaminatedwatersupplies(4),anddaycarecenters(13).Inthepresentstudy
epidemiclikecurvesarereportedforthepresenceofG.lambliathatshowayearlycyclicaloccurrence.RecentlytheDivisionofPublicHealth,StateofDelaware
(1),reportedasimilarseasonaloccurrenceinDelaware,whichpeakedduringthelatesummerandearlyfallperiodforboth1985and1986.Thisincreasecouldnot
beexplainedbydaycarecenterinvolvement,agedistribution,orgeographiclocation.InArkansas,specimensfromdaycarecenterswerenotnumerousenoughto
altertherefineddataand,althoughthedatawasskewedtowardcentralArkansas,geographiclocationsdidnotseemtobeafactor.Agedistributionwasunavailable
fortwooftheinstitutionsinthepresentstudybutdatafromACHshowedthattheseasonalincreaseisfoundinthepediatricpopulation.ThedatafromArkansasand
fromDelawaresuggestthattheremaybea"Giardiaseason"incertainareasoftheUnitedStates.
AmajorfactorintheincreaseofpositivestoolsduringmidandlatesummerinArkansasappearstobepartlydue

Page74

totheincreaseinthetotalnumberofstoolssubmitted.However,otherfactorsareprobablyinvolvedasindicatedbythegreaterthanexpectedpercentincreaseinthe
numberofpositivestoolsduringcertainmonthsoftheseasonalincrease.Thereasonsforthisincreasearenotknown.ThecyclicaloccurrenceinArkansascoincides
withthegreatestperiodofoutdooractivityandwaterresourceusage.Ithasbeennotedthatoutbreaksofwaterbornediseasesassociatedwithwatersupplieshave
peakedduringthesummermonthsintheU.S.(5).ThemostobviouspossibilitytoexplainthisyearlyseasonalincreaseinArkansaswouldbetheaccidentalor
purposefuluseofuntreatedsurfacewaterfordrinking.Thisissupportedbytheknownendemichumanandanimalreservoirsinthestatewhichwouldprovidea
sourceofwatercontaminationaspreviouslynotedbyepidemiologicstudiesdoneinthewesternUnitedStates(12,11).Themountainous,westernsectionsof
Arkansasresemblethewesternstatesinhavingfastmovingstreamswithdeceptivelyclearwaterandheavyrecreationaluse.However,therearenosubstantivedata
fromArkansastosupportthisconclusion.Otherfactorsmaybeinvolved,includingtheexoticpossibilityofacircannualrhythm(10)suchasreportedwithgastric
ulcers(7).Becauseofthepublichealthimplications,furtherepidemiologicstudiesshouldbedonetoclarifytheseasonalincreaseofGiardialambliainstateswhere
thisphenomenonisfound.
Acknowledgements
TheauthorswanttothankMr.JohnClarke,DepartmentofPhysiology,UniversityofArkansasforMedicalSciencesandPaulT.Archer,UniversityofArkansasat
LittleRock,GraduateInstituteofTechnology,forproductionofthefigures.
LiteratureCited
1.Anonymous.1986.Frequentlyreporteddiseaseseries:Giardiasis.DivisionofPublicHealth.StateofDelaware.Delawaremonthlysurveillancereport86:No.12.
2.Anonymous.1964.PublichealthataglanceIntestinalparasites.J.Ark.Med.Soc.61:8485.
3.Barbour,A.G.,Nichols,C.R.andT.Fukaskima.1976.Anoutbreakofgiardiasisinagroupofcampers.Am.J.Trop.Med.Hyg.25:384389.
4.Craun,G.F.1979.WaterbornegiardiasisintheUnitedStates:AReview.A.J.P.H.69:817819.
5.Dykes,A.C.,JuranekD.D.,Lorenz,R.A.,Sinclair,S.,Jakubowski,W.andR.Davies.1980.Municipalwaterbornegiardiasis:Anepidemiologicinvestigation.
Ann.Intern.Med.92:165170.
6.Heidt,G.A.,Nichols,A.H.andJ.J.Daly.1985.IncidenceofGiardiainArkansasbeaver.Ark.Acad.Sci.Proc.34:137.
7.Horwitz,M.A.,Hughes,J.M.andG.F.Craun.1974.OutbreaksofwaterbornediseaseintheUnitedStates.J.Infect.Dis.133:588592.
8.Howell,R.T.andB.S.Waldron.1978.IntestinalparasitesinArkansas.J.Ark.Med.Soc.75:212214.
9.Ivy,C.S.andJ.E.Steed.1971.IntestinalparasitesofArkansas,J.Ark.Med.Soc.67:329331.
10.Markiewicz,A.,Koszyk,T.andJ.Reising.1986.Seasonalvariationofinflammationandulcerincidenceasassessedbyupperdigestivetractendoscopy.
Chronobiologia1986:117121.
11.Pasley,J.N..1987.UniversityofArkansasForMedicalSciences,PersonalCommunication.
12.Rohlf,J.F.andR.R.Sokol.1981.StatisticalTables,1sted.,W.H.FreemanandCo.,SanFrancisco.
13.Sealy,D.P.andS.H.Schuman.1983.Endemicgiardiasisanddaycare.Pediatrics72:154158.

Page75

InfectionofMongoliangerbils(Merionesunguiculatus)withGiardiafromHumanandAnimalSources
K.DianeSwabby*,CharlesP.HiblerandJohnG.Wegrzyn
DepartmentofPathology,ColoradoStateUniversity,FortCollins,Colorado80523,U.S.A..
ForthepastfouryearsaspecificpathogenfreebreedingcolonyofMongolianGerbils,(Merionesunguiculatus),hasbeenmaintainedforthepurposeofprovidinghumansource
cystsofGiardiaduodenalisforourresearchandservicecommitments.Whencasesofgiardiasisinhumansandotheranimalsarepresented,andgerbilsarenotcommittedtoother
projects,crosstransmissionstudieshavebeenattempted.Thestandardprocedureistouse5to7weekoldgerbilsandexposeeachto5000cystsbygavage.Fiveuninoculated
gerbilsserveasnegativecontrols.Twentyfourhumansourcetransmissionshavebeenattemptedand14(58%)weresuccessful.Theprepatentperiodvariesbetween5and8days.
Forotheranimalstheresultsare:4/4beaver,prepatentperiod5to8days2/2chinchilla,prepatentperiod5to6days2/2domesticcats,prepatentperiod5to7days6/6
muskrats,prepatentperiod5to7days1/2domesticcattle,prepatentperiod7daysand1/1horses,prepatentperiod6days.Numerousattemptstocrosstransmitfromdogshave
failed.Onedomesticpigand5gerbilswereinfectedwithonehumansource,butcystsfromthispigfailedtoinfecttwodifferentgroupsofgerbils.

Introduction
Researchongiardiasisinhumansandotheranimalshasincreasedconsiderablyoverthepastdecade,initiallybecauseofthediscoverythatGiardiaisresponsiblefor
mostoftheepidemicsofwaterbornedisease(3)andsubsequentlybecauseofthefrequencythediseaseisdiagnosedinchildrenindaycarecenters,kindergartensand
nurseriesandinothersituations(travelers,hikers,etc.).Muchoftheresearchnecessitatesuseofcystsand/ortrophozoitesfromGiardiaduodenalisofhumanorigin.
Often,asintheevaluationofpilotfiltersdesignedtotestefficiencyoffiltrationtechniquesortodevelopC.tvaluesforinactivationbyozone,chlorine,etc.,millionsof
viablecystsarenecessary.Whileeffectiveinvitroexcystationprocedureshavebeendevelopedforsomeoftheseneeds,invitroencystationhasnotbeenperfected.
Humandonorsrarelyareavailableandhumanstoolsobtainedfromhospitals,etc.frequentlycontaincyststhataretooold,moreover,theyareseldomavailablewhen
neededbytheinvestigator.
WhenDaviesandHibler(1979)(4)performedsomeoftheearlycrosstransmissionexperimentsandnoticedthatthegerbil(Gerbillusgerbillus)wasareasonably
goodhost,thispromptedFaubert(1,3)toevaluatetheMongoliangerbil(Merionesunguiculatus)asananimalmodelforGiardia.Theirsuccesspromptedustouse
thisanimalforourmanyresearchandservicecommitments.OurlaboratoryhassuppliedhumansourceGiardiacystsforuseinengineeringstudies,asqualitycontrols
forwaterbornediagnosticprocedures,andmanyotherresearcheffortsoverthepasttwoyears.Currentlywesupplyfrom20to100106cysts/weektovarious
researchers.Thepurposeofthisreportistosharewithyousomeoftheproblems,failuresandsuccessesassociatedwithinfectingandmaintaininginfectionin
MongoliangerbilswithGiardiacystsfromvarioussources.
MethodsandMaterials
TheGerbils
TheoriginalbreedingpairswerepurchasedfromTumblebrookFarms,Inc.,WestBrookfield,Massachusetts.Theywereisolatedandtreateddailyfor5dayswith0.6
mgofFlagyl(metronidazole)bygavage.Iftheoffspring(weanlings)fromanyofthesepairswereinfectedwithTrichomonassp.orEndamoebasp.,thebreeding
pairswereagaintreated.Ifadultswereinfectedafterasecondtreatmenttheywereeliminatedfromthecolony.Presenceorabsenceofinfectionintheoffspringwas
determinedatpostmortembyscrapingsand/orwashingsfromtheintestineandcecum.Whenthecolonywasdeterminedtobefreeofthesecommensals,F2
generationoffspringwereusedforexperimentalstudies.Thoroughrecordswerekeptforage,breedingsuccess,andsubsequentmatings.Eventually,gerbilswere
selectedforbreedingpairsbasedonthesuccessofinfectingtheirsiblingswithcystsofGiardiafromhumansources.Whilethesuccessofthisapproachwouldbe
difficulttoanalyze,wecontinuetoselectbreedingpairsonthisbasis.
InfectionofGerbils
FecalmaterialcontainingGiardiacystswasdilutedwithdistilledwater,filteredthroughseverallayersofgauze(ifnecessary),andthenthrougha40mesh(40
openings/inch)screenandthecysts/mLofsuspensiondeterminedbydirectcountwithacalibratedmicropipette.Ifnotusedimmediatelythesuspensionwasstoredat
5C.Ifsamplesweretobestoredforseveraldaysthewatertothesedimentwasreplaceddailyandthesamplegentlyagitated.Theroutinedosage/gerbilusedinan
attempttoestablishinfectionwas5103cystsadministeredbygavage.Five5to7weekoldgerbilswereusedinallexperimentalcrosstransmissiontrialsand5
uninoculatedgerbilsservedasnegativecontrols.
CollectionofCystsfromGerbils
Beginning4dayspostinoculationandcontinuingthrough810dayspostinoculation,gerbilswereplacedasagroupinastandardclearplasticmousecagecontaining
awiremeshfloorraisedabout25mmabovethebottom.Thebottomofthecagewasfloodedwithdistilledwater.Gerbilsremainedinthecagefor1hourandwere
thenreturnedtotheiroriginalcagetobefedsunflowerseedsbeforerepeating
*Correspondingauthor.

Page76

theprocedure.Oftenitwasnecessarytocollectgerbilsindividuallyand,sometimes,euthanasiawasnecessarydependingontheexperiment.Pelletscollectedwere
immediatelymaceratedthroughan80meshscreenanddistilledwateraddedatarationof1/50suspension/distilledwaterandstoredat5C.
Initialexaminationofthesuspensionwasbydirectmicroscopy.Ifnocystswereobserved,thesamplewasevaluatedbyZnSO4centrifugalfloatation.Ifcystswere
beingproduced,anarbitraryscaleofcystnumbers/100bydirectmicroscopywasused:rare(15)occasional~(550)+(50100)++(100500)+++(500
1000)and++++(1000totoonumeroustocount).
ResultsandDiscussion
General
EarlyattemptstoinfectMongoliangerbilswithcystsofGiardiafromvarioussourcesoftenproducedvariableresults.Frequentlytheseweredogsourcesand/or
humansourcesobtainedfromchildreninfectedindaycarecenters.Sourcesfrombeaverand/ormuskratwerealwayssuccessful,butourinterestwasprimarilyto
establishhumansources.Unfortunatelyduringtheseearlyattempts(almostayear)goodrecordsofsuccessversusfailurewerenotmaintained.Eventually,through
experience,weimprovedonourabilitytoevaluatethequalityofthecystsfromthesourceanimals,improvefecalcollectionandevaluationprocedures(postmortem)
forthegerbils,andkeepbetterrecordsregardingthesourceandqualityofthecystsusedforinoculation.Theonlyattemptsatcrosstransmissionincludedinthis
reportarethoseattemptswherethecystswerefreshandmicroscopicevaluationindicatedtheywereviable.
TheresultsaresummarizedinTable1.The14/24(58%)humansourcesthatestablishedingerbilsoriginatedfromindividualsranginginagefrom4monthsto39
years.Generallycystsobtainedfromindividuals(supposedly)infectedwhilebackpacking,hiking,etc.wouldestablishaninfectionwhilecystsobtainedfromyoung
childreneitherdidnotestablishinfectionorwaspoorlyadaptedtogerbilsandcouldnotbefurtherpassaged.Inmostoftheinfectionscystpassagebeganbetween
days5and7.Onesourcebegancystpassageonday8.Thoseanimalsnotpassingcystsbyday8wereeuthanisedandexaminedpostmortemfortrophozoites.
Trophozoiteswerenotfoundinanimalsinoculatedwiththe10sourcesthatdidnotinfectgerbils.Fourofthe14successfulattemptsbeganaheavy(++++)cyst
passageonday5postinoculationand9598%ofthecystsweremorphologicallyexcellent.Mostoftheinfectionsresultedinlowtomoderatecystpassagebydays6
to7.Subsequentpassageofthesecystsintogerbilsoften,butnotalways,resultedingreatercystproductionandbetterqualitycyststhantheinitialinfectionhowever,
itwasnotunusualforthesourcetofailaftertwoorthreepassagesifcystproductionwasinitiallypoor.Generallyifcystscouldbepassaged5to6timesintogerbilsat
weeklyintervals,thesourcecouldbemaintainedforatleast15passagesbeforethesourcefailed.Onesourcewaspassaged27timesbeforefailingandatthattime
thecystsappearedviable.Atattempttoreestablishtheinfectionwithcystsfrompassage25wassuccessfultopassage27whentheyagainfailed.Currentlywe
maintaintwosourcesthathavebeen
TABLE1.SusceptibilityofMongoliangerbilstohumanandanimalGiardiacystsources
CystSource

Month

Prepatent
Period(Days)

HUMAN

January

++++

January

January

notinfectious

February

February

notinfectious

February

notinfectious

February

February

+++

February

+++

March

++

April

notinfectious

April

notinfectious

May

notinfectious

June

++

July

notinfectious

July

++

July

notinfectious

July

notinfectious

July

~to+

August

++++

September

++

November

++

December

~to+

December

notinfectious

CystProductioninGerbils

ANIMAL

Beaver4

Chinchilla2

Dom.Cat2

57

2/2++to+++

Muskrat6

6/6++to++++

Dom.Cattle2

1/2+

Horse1

1/1++++

Pig1

notinfectious

Dognumerous

notinfectious

58

4/4++to++++
2/2+++

passagedweekly21to50times,respectively.Somehumansourcesthatcanbepassagedregularlythroughgerbilsneverseemtoadaptsufficientlysothattheycanbe
usedforotherexperimentalpurposes(e.g.chlorinestudies).Cystproductionisoftenvariable,beginningaslateasday6or7,andcystqualityrangesbetween85
90%viableasdeterminedbymicroscopicevaluation.Cystproductionbygerbilsinfectedwithahumansourcethatquicklyadaptsandresultsinanabundanceof
viablecystswillusuallycontinuecystproductionthroughdays8to10postinoculationandthencystproductioneitherceasesorbecomesrareandsporadic.Generally
wecollectcystsfromdays5through8andthenterminatethetrial.Onehumansourcewecurrentlyuseoftenproducesconsiderablenumbersofcysts15dayspost
inoculation.
IfhumansourceGiardiaistobemaintainedinMongoliangerbilsasasourceofcystsforexperimentalpurposes,thehumansourcemustbewelladaptedtogerbils
andproduce5105to8105cysts/gerbil/day(8hourcollection)duringpeakproduction(days5through8)andthecystsmustbe9598%viable.Ifcystsfroma

Page77

sourcethatwillnotreadilyadapttogerbilsareused,theresultsofexperimentsusingthecysts(e.g.establishingC.tvaluesforinactivationofcystsbychlorine,etc.)can
bemisleadingandoftendisastrous.Evenwhenusingsourceswelladaptedtogerbilsthecystsmustbecollectedproperly,maintainedproperly,andpassagedregularly
throughtheanimals.Continuousmonitoringofthesourceinuseisnecessarytoachievereliableresults.
Mostoftheothercrosstransmissionexperiments(Table1)wereneverpassagedmorethan2or3timesingerbilsbeforethesourcewasdiscarded,primarilydueto
thecostandafearofcontaminationofthehumansourcesmaintainedforotherexperiments.Wedidnotethatwhilebeaver,muskratandchinchillasourcesreadily
establishedingerbils,someofthebeaverand/ormuskratsourceswerenobetteradapted,basedoncystproductionandqualityofthecysts,thansomeofthehuman
sources.Thehorsesourceandoneofthetwocatsourcesapparentlywerewelladaptedforgerbils,producinglargenumbersofviablecystsbetweendays5and7.
Thecattlesourcewasonlypassagedtwotimesandcystproductionbythegerbilswasalwayspoor.
InoneexperimentaweanlingdomesticpigwasmaintainedandexaminedforGiardiacystproductionfortwoweeks.Itwasthenexposedbyintubationtoahuman
sourceand5gerbilsalsowereexposed.Bothspeciesbeganpassingcystsonday6.Cystsfromthepigwereofexcellentquality,butinfectioncouldnotbe
establishedintwodifferentgroupsof5gerbils/groupwithcystscollectedondays7and8fromthepig.
LiteratureCited
1.Belosevic,M.,G.M.Faubert,J.D.MacLean,C.Law,andN.A.Croll.1983.GiardialambliainfectionsinMongoliangerbils:animalmodel.J.Inf.Dis.147:222
226.
2.Belosevic,M.,G.M.Faubert,T.S.Walker,E.Meerovitch.1983.ComparativestudiesonthepatternofinfectionwithGiardiaspp.inMongoliangerbils.J.
Parasitol.802805.
3.Craun,S.F.1986.WaterbornediseasesintheUnitedStates.CRCPress,Inc.,2000CorporateBlvd.,N.W.,BocaRaton,FL.295pp.
4.Davies,R.B.,andC.P.Hibler.1979.AnimalreservoirsandcrossspeciestransmissionofGiardiaIn:WaterborneTransmissionofGiardiasis.Jakubowski,W.
andJ.C.Hoff,(eds.).UnitedStatesEnvironmentalProtectionAgency.pp104126.
5.Wegrzyn,JohnG.1987.GiardiainColoradomuskratsPhD.dissertationColoradoStateUniversity,FortCollins,Colorado.

Page79

TransmissionofGiardiaDuodenalisfromHumanandAnimalSourcesinWildMice
P.D.RoachandP.M.Wallis*
KananaskisCentreforEnvironmentalResearch,UniversityofCalgary,2500UniversityDriveN.W.,Calgary,Alberta,Canada,T2N1N4
MatureDeerMice(Peromyscusmaniculatus)weretrappedatrandominthevicinityoftheKananaskisFieldStation.AbsenceofGiardiainfectionwasconfirmedbythreefaecal
analysesbothbeforeandaftertreatmentwiththreeconsecutive,dailydosesof10mgofmetronidazole.CystsofsevenstrainsofG.duodenaliswereproducedbyinfectinggerbils
(Merionesunguiculatus)withtrophozoitesfromculturedstocks.CystsoftwoadditionalstrainsthathadbeenmaintainedbyserialpassagethroughSwissWebstermicewerealso
usedtoinfectgerbils.LargenumbersofviablecystswereusedtoinfectwildDeerMicebygavageandtheresultsweredeterminedbyduodenalbiopsy.WildDeerMicebecame
infectedwithcystsofG.duodenalisstrainsthatoriginallycamefromsheep,muskrat,cattle,andsomehumanstrains.NoinfectionresultedfromG.duodenalisfrombeaverandadog.
Aninverserelationshipwasfoundbetweenlengthoftimeastrainhadbeenkeptincultureandsuccessofinfection.

Introduction
Historically,parasitologistshaveassumedthatdifferenthostsharbouredseparatespeciesofGiardiaandtheliteratureisrepletewithspeciesnamesderivedfromhost
animals.Filice(4)simplifiedthetaxonomyofthisgenusbyreorganizingitintothreespeciesG.agilis(amorphologicallydistinctparasiteofamphibians),G.muris
(foundinmiceandrats),andG.duodenalis(foundinmanandothermammalsincludingsomerodents).Thisschemeisnowgenerallyacceptedbymostworkers.G.
murisisverysimilarinmorphologytoG.duodenalisbutnoevidenceisavailabletosuggestthatitisinfectivetohumans.ItissometimespossibletodistinguishG.
murisfromG.duodenalis(4,5)bymorphologyalthoughthedegreeofoverlapbetweentherangesnormallyencounteredfortherelevantmorphometric
measurementsoftenmakeidentificationofindividualspecimensimpossible(1).
GiardiasisisanotifiablediseaseinAlbertaandapproximately150casesarereportedeachmonth.VeterinarianscommonlyseeGiardiainanimalstools.The
frequencyofthesereportssuggestthatGiardiasisisendemicamongbothanimalsandhumans.Transmissionbetweenhumansprobablyaccountsformanyofthe
reportedcasesbutinfectionfromanimalsmayalsooccur.Beavers(Castorcanadensis)havebeenimplicatedinzoonoticinfections(3,7,8,9,16)buttheroleof
otherwildanddomesticanimalsinthetransmissionofGiardiaisunknown.Giardiaspp.havebeendetectedinanumberofanimalsinsouthernAlbertaincluding
voles(Clethrionomysgapperi,Microtuspennsylvanicus),mice(Peromyscusmaniculatus),dogs(Canisfamiliaris),woodrats(Neotomacinerea),muskrat
(Ondatrazibethicus),cattle,(Bosbovis),sheep(Ovisaries)andbeaver(13,15).
EvidencesupportingcrossinfectionofstrainsofG.duodenalisfromanimalsotherthanbeaversisaccumulating.DaviesandHibler(2)haveshownthathuman
Giardiacystsareinfectivetoavarietyofanimalsincludingrats(Rattusnorvegicus),gerbils(Gerbillusgerbillus)guineapigs(Caviaporcellus),beavers,racoons
(Procyonlotor),dogs,andbighornXmouflonsheep(OviscanadensisXO.musimon).Hewlettetal.(6)successfullyinfectedSPFbeagleswithGiardiacystsfrom
ahumandonor.Woo(17)criticizedtheresultsofmanyofthesecrossinfectionexperimentsbutconcludedthatdogs,andprobablybeaver,couldactasareservoir
forhumaninfectiveGiardia.Webelievetherefore,thatGiardiafoundinhumansandsomeanimalsarenonspecificandzoonotictransmissionisprobable.
Thepurposeofthisstudywastodetermineifdeermice(Peromyscusmaniculatus)couldharbourG.duodenalisfromdifferentanimalsources.Thedeermousewas
chosenasananimalmodelbecauseitiscommonthroughoutNorthAmerica.ItsnaturalrateofinfectioninsouthwesternAlbertaisapproximately7%(15),a
prevalenceratesimilartothatinhumanpopulations(12).ThisstudywasnotdesignedtodemonstratethatPeromyscusmaniculatusisanimportantreservoirof
humaninfectiveGiardiabutthattransmissionofG.duodenalistoanabnormalhostcanoccurundertherightconditions.
MaterialsandMethods
AnimalModel
MatureP.maniculatusweretrappedatrandominthevicinityoftheKananaskisFieldStationinsouthwesternAlberta,Canada.Severalofthesewerefoundtobe
pregnantwhencapturedandtheyoungwererearedinthelaboratoryandusedforinfectionexperiments.Allanimalstrappedinthefieldwereassumedtobeadults
andtheageofyoungeranimalsbornandrearedinthelaboratorywasrecorded.
AlldeermicewerehousedinshoeboxcagesintheanimalroomsattheFieldStation.Threeconsecutive,daily,faecalsampleswerecollectedusingwiremeshgrids
supportedoverasmall
*Correspondingauthor.

Page80
TABLE1.SourcesofstrainsofG.duodenalis.
Strain

OriginalHost

Date

Location

H8

Human

850826

Canmore

H7

Human

850803

Calgary

CH3

Human

860715

FortCollins,Colorado(from
C.Hibler,hisstrain#H3)

Sheep(domestic)

850702

farmnearStrathmore45km
eastofCalgary

MR4

Muskrat

850920

SibbaldMeadowsPond
KananaskisCountry65km
westofCalgary

MR7

Muskrat

860529

farmnearStrathmore45km
eastofCalgary

D3

Dog

840816

CalgaryAnimalShelter

B5

Beaver

850619

SibbaldMeadowsPond
KananaskisCountry65km
westofCalgary

Cow

860529

farmnearStrathmore45km
eastofCalgary

S1

CW2

amountofwater.FaecalsampleswereexaminedforcystsusingthesucrosegradientmethodofRobertsThomsonetal.(11).Alldeermicewerethentreatedwith3
consecutive,dailydosesof7mgofmetronidazolebygavage.Higherdoseswereabandonedbecauseofunacceptablemortalityrates.Aftercompletingmetronidazole
treatment,themicewereagainplacedongridsforthreeconsecutivedaysandthecollectedfaeceswereexaminedforthepresenceofcysts.Iftheywerefoundtobe
freeofGiardia,themicewereplacedinfreshshoeboxcagesandsuppliedwithautoclavedbedding,foodandwater(adlib).
StrainsofG.Duodenalis
DeermicewerechallengedwithcystsofninestrainsofG.duodenalis(Table1).Theseincluded3fromhumans,1fromadomesticsheep,2frommuskrat,1froma
dog,1fromabeaver,and1fromacow.ProceduresfortheisolationandculturingofstrainsweregiveninWallisandWallis(14).
CystProduction
CystsofallstrainsexceptCW2andMR7wereproducedbyinfectinggerbils(Merionesunguiculatus)withtrophozoitesfromculturedstocks.Thiswas
accomplishedbycoolingamatureculturetubeoftrophozoitesto5Cfor30minutes,centrifuging,andresuspendingthetrophozoitesin2mLofphosphatebuffered
saline.HalfamLcontainingabout106suspendedtrophozoiteswasusedtoinoculategerbilstreatedasdescribedinWallisandWallis(14).
WehavenotbeenabletocultureCW2andMR7andtheyaremaintainedinvivoinoutbredCrl:CFW(SW)BRSwissWebstermice.Cystswereconcentratedfrom
thefaecesoftheseanimalsandapproximately105wereusedtoinfectgerbils.Inordertoenhancecystproduction,thegerbilswereimmunocompromisedbyadding
dexamethasone(Schering)totheirdrinkingwaterataconcentrationof40g/mL.Dexamethasonetreatmentbeganonthedayofinoculationandwasmaintaineduntil
theanimalwassacrificed.Cystsrecoveredfromgerbilfaeceswerecountedusingahaemocytometerandtheirviabilitywasdeterminedbyinvitroexcystationafterthe
methodofRiceandSchaefer(10).
InfectionofPeromyscus
CystsrecoveredfromgerbilswerewashedintapwatercontainingTritonX100andusedtoinoculatefrom6to11maturePeromyscusforeachstrainofthe9
strainsofGiardiatested.Atotalof81deermicewereinoculated.Infectedanimalswereheldfor5daystoallowanypossibleinfectiontoestablishitself,andthen
sacrificedbycervicaldislocationafterbeinganaesthetizedwithdiethylether.Uponnecropsy,biopsysamplesoftheduodenum,lumen,andcaecumweretakenand
examinedforprotozoausingphasemicroscopy.Experimentaldeermicewerenottreatedwithdexamethasoneordeliberatelyimmunocompromisedinanyway.
TheexperimentwascomplicatedbythepresenceofTrichomonassp.indeermicetrappedinthefieldandinthegerbilsusedforproductionofcysts.Thepresenceof
Trichomonasisdifficulttodetectbyfaecalanalysisandtheorganismisnotveryresponsivetometronidazoleatthedosagesused(higherdosageswereoftenlethalto
themouse).Forthisreason,thepresenceofTrichomonassp.wasrecordedwheneveritwasdetected.
Results
Theresultsofallfaecalanalysesfordeermicebeforeandaftertreatmentwithmetronidazolewerenegative.Wethereforeassumedthatthepreviousexposureofthese
deermicetoGiardiawasminimalornonexistent.
TheaveragenumbersandviabilitiesofcystsinoculatedaregiveninTable2alongwiththepercentageofpositiveGiardiainfectionsthatresulted.Thenumbersof
cystsinoculatedrangedfrom16,000to640,000andviabilitiesrangedfrom12to97%.StrainsCH3(9/11),S1(1/6),MR4(4/11)MR7(6/10)andCW2(5/11)all
resultedininfection.Inallcaseslargenumbersoftrophozoiteswerefound.NoinfectionswereobservedwithstrainsH7,H8,D3,andB5despitetheinoculationof
largenumbersofviablecysts.
ThecompleteresultsofgutnecropsiesarelistedinTable3.Atotalof25deermicebecameinfectedwithGiardiaoriginallyobtainedfromhuman,sheep,muskratand
bovinesources.Trichomonassp.wasfoundin57ofthedeermicebutitspresencewasnotwellcorrelatedwithGiardiainfection.Ofthe25positiveGiardia
infections,16(64%)werealsopositiveforTrichomonassp.
Discussion
TheresultsinTable3showthatitispossiblefordeermicetoactasareservoirforG.duodenalisfromavarietyofhosts.Althoughsomeofthemiceusedwere
rearedinthelabandwerethereforelessthan5weeksoldwhenused,theseanimalsdidnotshowahigherinfectionratethanolderadults.
Noattemptwasmadetofollowthecourseoftheinfectioninpositivecasesbutthenumbersoftrophozoitesobservedwerealwayshighwheninfectionoccurred.Our
experiencewithmiceingeneralisthattheyeitherbecomeinfectedwithlargenumbersoftrophozoitesthatpersistfor
TABLE2.Numbersofcysts,%viabilityandresultsofinfectionindeermice.

Strain

Average
Inoculum

%
Viability

%Giardia
Positive

H8

50,000

12

H7

50,000

27

CH3

288,000

75

82

S1

150,000

35

17

MR4

500,000

25

36

MR7

250,000

95

60

D3

17,000

16

B5

640,000

37

CW2

16,000

97

45

Page81
TABLE3.Giardia(G)andTrichomonas(T)detectedinDeerMiceinfectedwithvariousstrainsofG.duodenalis.

Total

T+

T+

Strain

G+

G+

G+

H8

H7

CH3

11

S1

MR4

11

MR7

10

D3

B5

10

10

CW2

11

Total

81

24

16

48

atleast2weeksortheydonotbecomeinfectedatall.FurtherworkwillberequiredtofullydescribethecourseofinfectionwithG.duodenalisindeermicebutitis
probablethatitwillbeanalogouswithG.duodenalisingerbils.
ItispossiblethattheprevioushistoryofeachstrainofGiardiahadaninfluenceonthesuccessofinfection.Thestrainsavailabletoushadbeenkeptinvitroorin
vivoforvaryingperiodsoftime.Forexample,CH3hadbeenpassagedthroughgerbils20timesbeforeitwasculturedandthetrophozoitesusedforinfectinggerbils
hadonlybeenheldinvitrofor2months.CH3wasthereforearelativelynewstrainatthetimeofuse.Theothertwohumanstrains(H7andH8)hadbeen
propagatedinvitroforoverayearbeforetheywereusedintheseexperiments.ExaminationofthedatainTables1and2showsthatnoneofthestrainsthathad
beenheldinculturefrombeforeJuly,1985werecapableofinfectingdeermice(althoughalloftheminfectedgerbils).Furtherinspectionofthedatarevealsthatthe%
positiveinfectionisinverselycorrelatedwiththelengthoftimebetweenisolationanduse,regardlessofthestrain.Weconcludethatalthoughitmaybepossibleto
infectyounggerbilswitholderstrains,maturedeermicetrappedinthewildaremoreresistanttochallenge.
Itisalsointerestingtonotethattwoofthestrainsused,MR7andCW2,provedrefractorytoculturedespiterepeatedattempts.Bothwereobtainedfromthesame
farmonthesamedateanditispossiblethattheyareidentical.Apreviousisolateobtainedfromasheepheldatadifferentlocationonthesamefarmoneyear
previouslywassuccessfullyculturedandwasalsoinfectioustodeermice.
TheabilityofdeermicetoactasahostforpotentiallyhumaninfectiveGiardiacouldnothavebeenpredictedfromsomepreviousstudies.Itisapparentfromthese
experiments,however,thattheseanimalsarecapableofharbouringGiardiaduodenalisforatleastashorttime.Thepurposeoftheseexperimentswasnottoshow
thatdeermiceareanimportantreservoirforwhatmustsurelybeanunnaturalparasitebutrathertodemonstratethatcrosstransmissioncanoccurundertheright
circumstancesifstrainsaresufficientlyvirulent.InotherwordstheabilityofG.duodenalistocrossinfectdependsasmuchonthevirulenceofthestrain,thequalityof
theinoculum,andtheimmunologicalstatusofthehostasitdoesontheidentityofthesourcehost.
Acknowledgements
ThisresearchwassupportedbygrantsandcontractsfromtheAlbertaEnvironmentalResearchTrustandtheResearchManagementDivisionofAlbertaEnvironment.
LiteratureCited
1.Bertram,M.A.,Meyer,E.A.,Anderson,D.L.andC.T.Jones.1984.AmorphometriccomparisonoffiveaxenicGiardiaisolates.J.Parasit.70:530535.
2.Davies,R.B.andC.P.Hibler.1979.AnimalreservoirsandcrossspeciestransmissionofGiardia.In:WaterborneTransmissionofGiardiasis.Jakubowski,W.
andJ.C.Hoff(eds.).U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio,EPA600/979001,pp.104126.
3.Dykes,A.C.,Juranek,D.D.,Lorenz,R.A.,Sinclair,S.,Jakubowski,W.,andR.Davies.1980.MunicipalwaterborneGiardiasis:anepidemiologicalinvestigation.
Ann.Int.Med.92:165170.
4.Filice,F.P.1952.StudiesonthecytologyandlifehistoryofaGiardiafromthelaboratoryrat.Univ.Calif.Pub.Zool.57:53146.
5.Grant,D.R.andP.T.K.Woo.1978.ComparativestudiesofGiardiaspp.insmallmammalsinsouthernOntario.I.Prevalenceandidentityoftheparasiteswitha
taxonomicdiscussionofthegenus.Can.J.Zool.56:13481359.
6.Hewlett,E.L.,Andrews,J.S.,Ruffier,J.,andF.W.SchaeferIII.1982.ExperimentalinfectionofmongreldogswithGiardialambliacystsandcultured
trophozoites.J.Infect.Dis.145:8993.
7.Lippy,E.C.1981.Waterbornedisease:occurrenceisontheupswing.J.Am.Wat.Wks.Assoc.73:5762.
8.Lopez,C.E.,Dykes,A.C.,Juranek,D.D.etal.1980.Waterbornegiardiasis:Acommunitywideoutbreakofdiseaseandahighrateofasymptomaticinfection.
Am.J.Epidemiol.112:495507.
9.Morbid.Mortal.Wkly.Rpt.1977.Waterbornegiardiasisoutbreaks.26:269275.
10.Rice,E.W.andF.W.SchaeferIII.1981.ImprovedinvitroexcystationprocedureforGiardialambliacysts.J.Clin.Microbiol.14:709710.
11.RobertsThomson,I.C.,D.P.Stevens,A.A.F.MahmoudandK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gastroenterol.71:5761.
12.Schmidt,G.D.andL.S.Roberts.1981.FoundationsofParasitology.2nded.,p.89,C.V.MosbyCo.,St.Louis,p.795.
13.Wallis,P.M.,BuchananMappin,J.M.,Faubert,G.M.andM.Belosevic.1984.ReservoirsofGiardiaspp.insouthwesternAlberta.J.WildlifeDis.20:279283.
14.Wallis,P.M.andH.M.Wallis.1986.ExcystationandculturingofhumanandanimalGiardiaspp.byusinggerbilsandTYIS33medium.Appl.Environ.
Microbiol.51:647651.
15.Wallis,P.M.,Zammuto,R.M.andJ.M.BuchananMappin.1986.CystsofGiardiaspp.inmammalsandsurfacewatersinsouthwesternAlberta.J.WildlifeDis.
22:115118.
16.Wilson,H.P.S.,Stauffer,S.J.andT.S.Walker.1982.WaterbornegiardiasisoutbreakAlberta.Can.Dis.Wkly.Rpt.820:9798.
17.Woo,P.T.K.1984.Evidenceforanimalreservoirsand

Page82

transmissionofGiardiainfectionbetweenanimalspecies.In:GiardiaandGiardiasis.Erlandsen,S.L.andE.A.Meyer(eds.),Plenum,NewYork,pp.341364.

Page83

WATERTREATMENT

Page85

WaterTreatmentandtheGiardiaCyst
AlbertvanRoodselaar
AlbertaEnvironmentalCentre,Bag4000,Vegreville,Alberta,CanadaTOB4L0.
TheconcernofthissessioniswithGiardiaasitpertainstoPotableWaterTreatment.ThehistoryofPotableWaterTreatmentreflectsdifferentconcernsatvarious
stagesofsocietaldevelopment.AtthetimeoftheRomans,asdescribedinthepreservedwritingsoftheengineer,SextusJuliusFrontinus,theprimaryemphasiswas
ontransportoftheoptimumattainableaestheticqualitywaterfrompointofavailabilitytothepointofneed.ForthispurposetheRomansbuiltanimpressivesystemof
aqueducts.Evenatthistime,thespecialneedsofdrinkingwaterwererecognizedandthewatersofhighestpuritywerereservedforthisapplication,''Marcia,so
charminginitspurityandandcoldness,shouldservewhollyfordrinkingpurposes".Thedifferencebetweenthenandnowisthepracticeoftreatmenttoimprove
qualityandthedevelopmentofcomprehensivecriteriabywhichtojudge"purity".Thisisimportantinourassessmentofcurrentwatertreatmentprocessessuchas
clarification,filtrationanddisinfectionsincetheseareevaluatedbytheireffectonvariousphysical,chemicalandmicrobiologicalparameters.
Consequently,indeterminingtherelevanceofagiventreatmentprocesswithrespecttoit'seffectivenessonGiardiacysts,wemaysimplybeconsideringanother
parameterwhichbehavesinparallelwiththosealreadyavailableintheliteratureorwemaybeconsideringaparameterwhichmustbedealtwithinanindependent
mannerfromthosewhichhavetraditionallybeenofconcern.Hereinliesanimportantquestion,whetherasimple,easilymeasurableparameterwillserveasan
adequeatesurrogateofGiardiaremovaland/orinactivation.
Giardiacystshavebeendetectedinrawwatersvaryingfromcontaminatedlakesandhighlyturbid,colouredriverstomountainstreamswhicharecrystalclear.The
widerangingpresenceisachievedbythevarietyofhostsavailableforthetransmissionofthisorganism.Traditionally,communitieslocatedonwatersourcesofhigh
quality,asdeterminedbyturbidity,colourandcoliformcount,havehadnominaltreatmentfacilities,oftenrelyingsolelyonchlorinedisinfectiontoensure
microbiologicalqualitysuitableforpotableuse.ThisisinkeepingwiththeGuidelinesforCanadianDrinkingWaterQuality,1978,whichstates,"Allsuppliesderived
fromsurfacewatersourcesshouldreceivedisinfectionasaminimumtreatment."(p.33)and"Thetotalcoliformgroupisthereforepreferredasanindicatoroftreatment
adequacyindrinkingwatersupplysystems.(p.25).However,outbreaksofgiardiasisincommunitiesservedbysystemswhichhaveinthepastbeenconsidered
adequateformicrobiologicalprotectionhavethrownintoquestionthebasisofthisconclusion.ComparingtheresponseofGiardiacystsandothermicrobiological
organismstodisinfectantsandotherwatertreatmentprocessesbecamenecessarytounderstandhowthegoalofmicrobiologicallysafedrinkingwatercouldbe
attainedinclusiveoftherecognitionofthisadditionalrisk.
Concernsrelevanttodrinkingwatertreatmenthavethuscomefullcircle.Itisimportanttoreiteratethattheprimeconsiderationfordrinkingwaterhasbeenandshould
continuetobetheproductionofawaterofsuperiormicrobiologicalquality.Itisofinteresttoexaminetheaspectofpublichealthasitpertainstomicrobiological
qualityofdrinkingwater.Table1demonstratesthehighprevalenceofgastroentericdifficultiesamongwaterbornediseases.Waterbornediseasesareeitherincreasing
orareincreasinglyreportedinAmerica(Craun1977).Sharpenedpublicawarenesswillplaceaneverincreasingpressureonthewatertreatmentindustrytoimprove
productqualityandsafety.Microbiologicalqualityhasbeenthrownintodoubtandjoinstheotherparameterscurrentlyinthepublicconciousnesssuchasorganics
(carcinogens),lead,mercury,andtasteandodour.
Table2showstheresponseofafewselectedorganismstochlorinedisinfection(ashypochlorousacid).Thisdataindicatesthatevenpriortothegenerationofspecific
informationonGiardiacysts,theresistanceofanothercyst,E.histolytica,relativetoE.coliandpoliovirus,showedthatthecyststructurewascapableofbeingtwo
orthreeordersoffmagnitudemoreresistanttochlorinethancertainbacteriaandviruses.Table3,usingdataavailablein1967,indicatesthattheconditionunderwhich
chlorine
TABLE1.Distributionofetiologybywaterbornediseaseoutbreaksfrom1946to1974*
Outbreaksforwatersystems
Community
Disease

Other

Number

Percentage

Number

Percentage

Gastroenteritis
(unknownetiology)

71

52.2

153

47.6

Infectioushepatitis

22

16.2

44

13.7

Shigellosis

13

9.6

33

10.3

Chemicalpoisoning

5.9

13

4.0

Giardiasis

5.1

2.5

Typhoid

4.4

51

15.9

Salmonellosis

4.4

2.8

Amebiasis

0.7

1.2

Poliomyelitis

0.7

EnteropathogenicE.
coli

1.2

Tularemia

0.6

Leptospirosis

0.7

136

321

Totaloutbreaks

*Sources:CraunandMcCabe(1973)andCraunetal.(1976).

Page86
TABLE2.Dosagesofchlorinerequiredtoachieve99%inactivation.
Concentration Contacttime c.ta
(mg/L)
(min)

Test
Microorganism

ChlorineSpecies

pH

TempC References

E.coli

HypochlorousAcid
(HOCl)

0.1

0.4

0.04

6.0

Scarpino,etal.,
1974

Poliovirous1

Hypochlorousacid
(HOCl)

1.0

1.0

1.0

6.0

Weidenkopf,1958

0.5

2.1

1.05

6.0

Englebrecht,etal.,
1978

1.0

2.1

2.1

6.0

Scarpino,etal.,
1974

E.histolytica
Cystsb

Hypochlorousacid
(HOCl)

5.0

18.0

90

6.0

Snow,1956

aConcentrationofcompoundmultipliedbycontacttime(mg/L)(min).
bExtrapolateddata.

TABLE3.Lethalitycoefficientsfordifferentmicroorganismsbasedontreatmentwithfreeandcombinedchlorineat5C.
Oxidant

Enteric
bacteria

Protozoan
cysts

Virus

Bacterial
spores

HOCl

20

0.05

1.0

0.05

OCL

0.2

0.0005

0.02

0.0005

NH2Cl

0.1

0.02

0.005

0.001

FromMorris,1967.

disinfectantisapplied(pHandconcentrationofNH4inthewaterbeingtreated)hasalargeimpactontheeffectivenessofthedisinfectantdose.Itshouldthereforenot
comeasamajorsuprisethatGiardiaisplayinghavocwithdisinfectantbasedtreatmentsystems.Theemphasisshouldbeonestablishingthoseconditionswhich
provideadequatelevelsofprotectionagainstGiardiacysts,determiningthecurrentstatusofoperatingplantsandimplementingimprovementsinthoseplantswhich
areatanunacceptablehighlevelofrisk.
Todothis,anunderstandingoftreatmentprocesses,bothconventionalandotherwiseisrequired.EffectivenessofmorethanoneprocessinreducingGiardiacyst
riskinthetreatedwater,eitherthroughremovalorinactivation,resultsinamultibarrierscreen.Someplantsmaydependononlyonesuchbarrierwitharesultant
Giardiapenetrationifthisbarrierfailsoriftheeffectivenessofthisbarrierisoverestimated.Consequently,detailedinformationoneachbarrierandthenetimpactof
multiplebarrieroperationisnecessaryforthetreatmentengineertobeabletoassessifhisplantisatrisk.

Page87

RemovalofGiardiathroughSlowSandFiltration100MileHouse,BritishColumbia
JackM.G.Bryck*,BrianL.WalkerandDavidW.Hendricks
Dayton&KnightLtd.,P.O.Box91247,WestVancouver,B.C.V7V3N9,Canada.
Inthefallof1981,about60casesofwaterbornegiardiasiswereconfirmedatthevillageof100MileHouse,Canada.Contaminationofthevillage'swatersystemwassuspected.
Theserviceareapopulationwasabout2,000persons.Beaversandmuskrats,subsequentlyconfirmedpositivefortheGiardiacyst,weretrappedupstreamofthevillage'ssurface
waterintake.Atthetimeoftheincidentthesurfacewatertreatmentincludedonlychlorinationwithminimalcontacttime.Afterapilotfiltrationprogramtomonitortheeffort
requiredtooperateseveralfiltertypesinamanneridentifiedbytheU.S.EPAthatwouldremovetheGiardiacyst,theslowsandfiltrationprocesswaschosenforincorporationinto
thevillage'snewwatertreatmentplant.ConstructionofthewatertreatmentplantwascompletedDecember15,1984andoperationcommencedNovember15,1985.Thetotal
constructioncostwas$780,000CDNor$0.11perlitreofinstalledpeakcapacity.WithfundingfromHealthandWelfareCanada,anevaluationofthetreatmentplantwas
undertakenbetweenOctober1985andOctober1986.IncludedwasmonthlyhighvolumesamplingintherawandfilteredwaterfortheGiardiacystandothermicroorganismsand
anevaluationofthemicroorganismsandbiologicalprocessespresentonthesandmedia.TheremovalthroughtheslowsandfiltersofGiardiaandothermicroorganismsintheraw
waterwascompleted.

Introduction
Inthefallof1981atleastsixtycasesofgiardiasiswereconfirmedat100MileHouse,BritishColumbia.Theoutbreakaffectedpeoplelivinginawidegeographical
areaaround100MileHouseandnoothersourceexplainedthefindingsaswellaswaterbornecontaminationofthemunicipalwatersystem.Thesourcewas
suspectedtobebeaversandmuskratslocatedupstreamofthevillage'ssurfacewaterintake,whichweresubsequentlyconfirmedpositiveforGiardiacysts.Atthe
timeoftheincidentthewatersupplywassurfacewaterfromBridgeCreek,locatedwithintheuncontrolledHorseLakewatershed,andtoaminorextentfroma
groundwaterwell.Theadditionofchlorinesolution,withminimalcontacttimeprovidedinthedistributionsystem,wastheonlytreatmentpractised.Thesurfacewater
qualityatthetimeoftheincidentisnotknown.However,subsequenttestingoveracalendaryeardeterminedthatwaterqualitygenerallyexceededtherecommended
objectivelevelssetbytheBritishColumbiaMinistryofHealth.Thisincludedturbiditywhichwaslessthat2NTUoveracalendaryearandwasgenerallylessthan1
NTUoverseveralmonths.
ThefurtherdevelopmentoftheBridgeCreeksurfacewatersource,includingatreatmentplantincorporatingfiltration,waschosenbyvillagecouncilasthepreferred
methodofmeetingthevillage'swaterdemands.AfteranonsitepilottestingoffourtypesoffiltersbetweenJuneandOctober1983,includinggravityandpressurized
rapidratemultimedia,slowsandanddiatomaceousearthfiltration,slowsandfiltrationwaschosen.TheplantwentintoproductionNovember15,1985.Between
November15,1985andNovember15,1986researchwasconductedtoascertainremovalefficienciesofGiardiacysts,ofcoliformbacteriaandofanyother
measurableindicatorparticlesbytheoperatingandpilotslowsandfilters.Also,determinationofthebiologicalactivityandenumerationofinvertebrateswithinthefilter
mediawasmade.
SlowSandFiltration
Themaincomponentofaslowsandfilteristhefiltersandsupportedbygradedgravelthroughwhichdrainsareplacedtocollectthefilteredwater.Theeffectivesand
sizetypicallyrangesfrom0.15to0.35mmwithauniformitycoefficientlessthat2.Thedepthofthesandistypically.75to1.2mm.Thesystemiscontainedbyafilter
boxofsufficientheighttopermitabout1.8mofheadwaterabovethesandbed.Thesurfaceareaofaslowsandfilterisquitelargewhencomparedwithrapidrate
filtration,asthehydraulicloadingrateforslowsandfiltrationis0.04to0.4m/hcomparedto5.0to2.4m/hforrapidratefiltration,afactorof50to100timesslower.
Theoperationofthefilteris"passive",whichmeansthatthephysical/biologicalprocesseswithinthefilterbedrequirenoactiveinterventionbyanoperatorotherthan
adjustmentofthefilterproductionrateandscrapingofthefiltersurfaceatterminalheadloss.Theinitialheadlossofthefilteristypically0.2mincreasingtoabout1.3
mattheendoftherun,dependingontheheadwaterdepth.Atterminalheadlossthefilterbedispartiallydewateredandthesurfacedeposit,termedthe
schmutzdecke,isremovedbyscraping.Usuallyabout5to10mmofsandisremoved.Afterrepeatedscrapingsover5to10yearsthedepthofthebedisreducedto
alowerlimitofabout.30m,atwhichtimeitmustberebuilttoitsoriginaldepthof1.0to1.3m.
*Correspondingauthor.

Page88

Operation
Asimplifiedprocessoftransport,attachmentandpurificationhasbeendevelopedtoexplainremovalofparticulates(4).
Thescreeningprocesstrapsandretainsparticlestoolargetopassthroughthevoidsinthesandmedia.Thistypicallyoccursatthefiltersurface.Withaneffectivegrain
sizeof150mthoseparticlesgreaterthan20mshouldberemovedbyscreening.Theschmutzdeckeorbuildupofparticulatesatthefilterwillincreasethe
effectivenessofthescreeningmechanism.Thesedimentationprocessinvolvessettlingofparticulateswithinmediavoids.Forawatertemperatureof10Candorganic
matterwithadensityslightlygreaterthanwaterthecompleteremovalofparticlesgreaterthan4mbysedimentationcanbeexpected.Particleslessthan1min
diameterwillnotberemovedbysedimentation.
Theprincipalmechanismthatholdstheparticlesintherawwaterinplaceoncecontacthasbeenmadewiththemediaisadhesion.Theadhesionmechanismoccursat
thesurfaceoftheslowsandfilter,wherebacteriaandothermicroorganismsusingtheorganicmattertrappedonthefiltersurfaceasanenergysource,produceaslimy
masstermedzoogloea.Thezoogloea,activebacteria,theirwastesanddeadcellsandorganicmatter,formsagelatinousfilmtowhichparticulatesintherawwater
adhere.Thetrappedorganicparticlesareassimilatedintothefilmwhiletheinertmatterishelduntilthefilterisscraped.Bacteriaandmicroorganismsmultiplyusingthe
depositedorganicmatterasanenergysourcefortheirmetabolismandgrowth.Asthebacterialpopulationisdependentonthefoodsourceavailablethebacterial
growthisaccompaniedbyanequivalentbacterialdieoffwhichprovidesasimplerorganicfoodsourcetolowerdepthsinthesandfilter.Throughanumberofsteps
theorganicmatterisgraduallybrokendownintocarbondioxide,water,sulfates,nitratesandphosphates.
Microorganismandbacterialactivityismostpronouncedintheupperpartoftheslowsandfilter,graduallydecreasingwithfilterdepthasfoodlevelsdecrease.As
theirfoodsourcedecreasestheystarve,particularlyduringwarmwatertemperatures,whenmetabolicratesarehigh.Theslowsandfiltrationefficiencywillbereduced
atlowertemperaturesastherateofmetabolismofbacteriaandothermicroorganismssuchasbacteriaconsumingprotozoaandnematodeswilldecrease.Theresultis
anincreaseinthesurvivalrateofintestinalbacteriacarriedthroughthebed.
RecentworkbyBellamyetal.(1)hasshownthatabsorptionandthenmetabolismbymicrobialfilmsonthesandgrainsconstituteamajorremovalmechanism.Pilot
plantworkhasshown99.9%removalsofcoliformbacteriaspikesand100%removalsofGiardiacystswhenthebiofilmissufficientlydeveloped.
RemovalEfficiency
Anevaluation(4,5)oftheeffectivenessofslowsandfiltrationandtheroleofoperatingconditionslookedattheeffectofcystremovalonhydraulicloadingrate,
temperature,cystconcentration,ageofschmutzdecke,andageoffilterbed(biologicalmaturity).Thehydraulicloadingratesincluded0.04m/h,0.12m/hand0.40
m/h,therawwatercystconcentrationsrangedfrom50to5,075cysts/litre,andtherawwatertemperatureswere15Cand5C.Conclusionswerethatthecyst
removalcanbeexpectedtobegreaterthan98%whenthefilterisestablishingthebiopopulationandisvirtuallycompleteonceestablished.Withinthelimitationsofthe
research,temperature,cystconcentrationandhydraulicloadingratedidnotaffectcystremoval.Themostimportantvariableintheperformanceofthefiltermediawas
establishedtobethebiopopulation.
Logsdon(6)reportedcoliformremovalof99%forthemonthsofMarchtoNovemberand94%to97%forNovembertoJanuaryatafullscaleslowsandfilter.The
appliedcystremovalwas99.9%atavarietyofoperatingconditionsbuttheauthornotedthatatlowtemperaturesGiardiacystremovalefficiencymaydecrease.
TheUnitedstatesPublicHealthServicereportedvirusremovalsat22to96%.TheMetropolitanBoardofHealthinLondon,England,foundreductionof99.9%in
virusesthroughalaboratorysizeslowsandfilteratatemperatureof1112Cand99.8%atatemperatureof6C(7).
McConnel(8)demonstratedexcellentremovalofreovirusthroughslowsandfiltrationevenatfilterstartupintheabsenceoffiltermaturation.E.coliremovalwas
99.97%orgreateratanaverageloadingof7.5107E.coliperpilotfilter,butwhentheloadingwasincreasedby3logunitsbreakthroughoccurred.Biological
activitywaspresentthroughtheentirefilterdepth.
TreatmentPlantDescription
Thewatertreatmentplant,withanominalandpeakcapacityof3.63ML/dand7.26ML/drespectively,includesasurfacewaterintake,rawwaterpumpstation,3
slowsandfilters,chlorinationequipmentandchlorinecontacttank,aclearwell,treatedwaterpumpsandacontrolbuilding.The3slowsandfiltersareeach43mlong
by6mwideforafilterareaof258m2.Thetotalfilterareais774m2.Thenominalhydraulicloadingrateis0.19m3/m2/hwithapeakhydraulicloadingrateof0.40
m3/m2/h.Thefiltermediahasadepthof1,050mm,aneffectivesizeof0.20to0.30mmandcoefficientofuniformityof3.30to3.80.Thefiltersarecoveredto
minimizetheimpactoflowwintertemperatures(20to40Cfor24weeks)andtopreventalgaegrowthinthefilterheadwaters.Totalconstructioncostwas
$780,000,$0.11/Latpeakcapacity.
ThetendersforconstructionofthewatertreatmentplantclosedMay29,1984,thecontractawardedJune4,1984andthesubstantialcompletionwasDecember11,
1984.DuetoproblemswithdeliveryofmediatheplantstartupusingonlytwofilterswasNovember1,1985withtreatedwaterdeliverytothevillagedistribution
systemstartingNovember15,1985.ThethirdslowsandfilterwasputintoproductionMay29,1986.
Threepilotslowsandfilterswereconstructedof300mmdiameterClass200PVCpipeandincludedawrappingof50mmthickfoamglasscellularglassinsulationto
minimizethewatertemperaturechangethroughthefilter.Theverticaldimensionsandthemediaandgravelsupportofeachpilotfilterwereidenticaltothevillage's
operatingfilters.A25mmdiameterPVCpilotheaderpipewasconnectedtotheplant'srawwaterheaderandtoeachpilotfilter.Fromthebaseofeachfilterwasa
17mmPVCpipe

Page89

toatreatmentplantdrain.OnthepilotdischargepipingwasaGilmontrotometer(101,000mL/min)toregulatethepilotplanttreatmentrateandafilterhousingto
holdthesamplingcartridge.
Methods
WaterQuality
Therawandfilteredwatersampleswerefromtherespectivetreatmentplantpipeheaders.
Giardiasamplingwasaccomplishedusingonemicronporesize,polypropylenecartridgefilterhousedinclearplexiglassCunoFilterModelIMI4459101.From
3,600to4,500Lwassampledovera3to5hourperiodforasamplingrateof15to20L/m.Thecartridgefilterswerewrappedinplasticbags,placedoniceinan
icecooler,transportedviaovernightcouriertoDepartmentofPathology,ColoradoStateUniversity,FortCollins,Colorado,U.S.A.Theelapsedtimebetween
completionofsamplingat100MileHouseandarrivalatColoradoStateUniversitywastypically3660hours.
AnalysisofthegrabsamplesforturbiditywasdoneusingaHACH2100Aturbidimeter.Watersamplesfortotalcoliform(MultipleTubeFermentationTechnique)and
standardplatecountweretakenasgrabsamplesinsterile250mLplasticbottleswiththeresultsreportedasMPN/100mLandcolonies/mLfortotalcoliformand
totalplatecount,respectively.Watersampleswereanalyzedforsolids(total,suspended,dissolved),nitrogen(ammonia,nitrate,TKN),phosphorus(total,ortho),
totalorganiccarbon(TOC),alkalinity,hardnessandpH,withproceduresasperStandardMethods(9).Thesewatersamplesweretakenasgrabsamplesinsterile1
Lplasticbottleswithduplicatesamplestakenateachlocation.Oneoftheduplicatesampleswaspreservedwithsulfuricacid.Thesampleswereplacedoniceand
analyzedbyCanTestLtd.,Vancouver.
FilterMedia
Filtermediasamplesat0to5mm,300mmand600mmweretakenfromfilter2immediatelypriortoscraping.Eachsample,weighingabout0.5kg(1lb),was
removedbyshovel,placedinaplasticbag,storedoniceinanicecoolerandtransportedviaovernightcouriertoEVSConsultantsLtd.ofNorthVancouver.
SandsamplesforinvertebrateswerepreservedinisopropanolcontainingphloxineB(ahistologicalstainusedtofacilitatesorting),returnedtothelaboratory,washed
througha250mmeshsieve,andsortedintomajortaxausingadissectingmicroscope.Sandandwatersamplesforphytoplanktonanalysiswerepreservedwith
isopropanolandphloxineBorLugol'siodine,respectively.Utermohlchambercountsof50mLsubsampleswereundertakenwith20fieldcountedforeach
subsampleasperStandardMethods(9).Theactivityofheterotrophicbacteriaatvariousdepthsinthesandfilterwasassessedbyamodificationoftheresazurin
reductiontechnique(9)tomeasuremicrobialdehydrogenaseenzymes.Awetweightsample(100g)ofsandwasincubatedin100mLofplatecountbroth(PCB)
supplementedwithresazurin.Theactivityofheterotrophs(bacteriautilizingorganicmatterasasourceofenergy)wasassessedbyincubatingthesandbrothmixtureat
20Candmonitoringthereductionofresazurinat610nmusingaNovaSpecspectrophotometer.Duplicatecontrolsamplescontained5mg/LHgClasabacteriocide
toindicatethecontributiontoresazurinreductionbychemicalconstituents.Thedifferencebetweentotalandchemicalreductiontestswasexpressedasmole/g/hour
ofresazurinreductiontoprovidearelativeestimateofbiologicalactivityamongthedifferentsamples.
Results
Temperature
AtemperatureprofileisshowninFigure1.FromNovember1985toFebruary28,1986thewatertemperatureintherawwaterwasbetween1Cand3C.From
March1,1986toOctober31,1986thewatertemperatureintherawwaterrangedfrom3Ctoapeakof19CinAugustanddecreasedtoaminimumof3C.
Turbidity
AturbidityprofileisshownonFigure2.FortheoneyearperiodNovember15,1985toNovember15,1986themaximumrawwaterturbidityvaluewas1.8NTUin
May1986whiletheminimumvaluewas0.5NTUinDecember1985.Therawwaterturbiditywasreduced20to50%betweenNovember1985andApril1986
andwasreduced40to80%intheperiodMay1986toNovember1986.Thefilteredwaterturbidityrangedbetween0.15NTUand0.8NTUandtheprofilewith
timetrackedtherawwaterprofilewithtime.
GiardiaCystsandOtherMicroorganisms
BetweenNovember1985andNovember1986therawwatertothetreatmentplantwassampled21times.In52%ofthesamplestaken(11samples),cystswere
detectedinnumbersthatvariedfrom14to>300cysts(Figure3).Inmostcasesthecystswereidentifiedas"excellentandprobablyinfectious".Thepresenceofthe
cystsvariednotonlythroughouttheyearbutalsobetweentwosamplestaken1624hoursapart.Forexample,onAugust26,1986asampletakenbetween0900
and1329hoursfound110cystsbutnocystsweredetectedinasampletakenthenextdaybetween0850and1400hours.

Figure1.
Airandfilteredwatertemperatures.

Figure2.
Averagerawandfilteredwater
turbidity, =rawwater,+=filteredwater.

Page90

Nocystsweredetectedinthe22samplesoftreatedwaterfromtheslowsandfilters.ThiswaseventrueofsamplestakeninNovemberandDecemberwhenthe
filtershadbeeninoperationatotalofonly21and40days,respectively.InNovemberover300cystsweredetectedintherawwaterwhileinDecember14cysts
weredetected.Atthetimetherawwatertemperaturewas1to2Candatthiswatertemperatureitisunlikelyanybiologicalactivitywaspresentinthemedia.
PilotfilterBwasspikedwithGiardiacystsandcoliformonNovember27,1985andMarch20,1986.Thepilotfilterhadbeenincontinuousoperationfor12daysat
thetimeofthespikinginNovemberand125daysatthetimeofthespikinginMarch.ThewatertemperaturetoNovemberwascontinuouslyat1Cwhileitvaried
between1Cand4CfromNovember27,1985toMarch20,1986.
OnNovember27,1985anestimated2,000,000cystsweresuppliedbutonly675,000cystsweredetectedintwo1litregrabsamplesfromthefilterheadwater.Itis
unknowniftherewasadieoffofcystsbetweenFortCollinsand100MileHousethatresultedinasmalleramountrecovered,orifthecystdistributionwasnotequal
throughouttheheadwater.Tencystswererecoveredinthefilteredwaterforaremovalefficiencyofatleast99.99%.Itisunknownifcystsmovedthroughthemedia
aftercompletionofthe31hoursofsampling.OnMarch29,1986theestimatednumberofcystssuppliedwas4,400,000.Thetwo1litrefilterheadwatergrab
sampleswerespoiledintransit.Therewerenocystsdetectedinthefilteredwater.
ThevariationinalgaelevelsisidentifiedinFigure4.ThealgaelevelintherawwaterpeakedintheJulyandAugustperiodwhenthetotalcountontwooccasionswas
59and460cells/mL.Thereweregenerallylowalgaelevelsinthefilteredwaterwiththecountontwooccasionsbeing21,22and43cells/mL.Thedominantgenera
ofalgaeintherawwaterinMarch1986wereSynedrasp.,Tabellariasp.andFragellariasp.,whileFragellariacrotonensis,F.virescensandAchnanthessp.
werefoundintheJune1986rawwatersample.Fortherawwater

Figure3.
NumberofGiardiacystsinrawwater
tovillage'swatertreatmentplant

Figure4.
Relativealgaelevelsinrawandfilteredwater

sampletakeninAugust1986thedominantgenerawereCocconeisplacentulaandAchnanthesminutissima.
SynedraandCeratiumcontributetotasteandodour.Dinobryon,Synedra,andFragilariacontributetofilterclogging.Thevillagehashadcommentsfromusersof
thesystemthatthetasteassociatedwiththerawwaterhasdisappearedwithstartupofthewatertreatmentplant.ItissuspectedthattheremovalofSynedraand
Ceratiumalgaewasthereason.
Coccidiaisamammalianparasitenotinfectioustoman.Coccidiaweredetectedintherawwaterfourteentimesoutof21samplestaken.Thetypewasfrombeaver,
rodentormammal.Theoccurrencewasthroughouttheyear,andwheneverGiardiacystsweredetectedCoccidiawasalsodetected.TheCoccidiawasnot
detectedinthefilteredwater.
Plantdebrisinthecontextofthisresearchreferstorodent(beaver,muskrat,mouse)fecaldebris.Plantdebriswasdetectedineverysampleofrawwatertakenin
amountsthatvariedbetweenoccasionaltoasmallamount,butwasinmoderateamountsinonesample.Plantdebriswasnotdetectedinthefilteredwaterexceptina
rareamountinNovember1985andanextremelyrareamountonJanuary20,1986whenthefilterswerestillintheripeningstage.Completeremovalswerealso
foundforcrustaceans/eggs,pollen,ciliatesandflagellates.
GeneralWaterQualityParameters
Amajorpurificationmechanismintheslowsandfilterisbiologicaloxidationoforganicmatter.Thenitrogen,phosphorusandtotalorganiccarbonconcentrationstaken
onfiveoccasionsarepresentedonTable1.
Themainbiodegradablesubstances,ammoniumandorganicmatter,havethepotentialtobeoxidizedbybacteriainthesandmedia.Organicremovalisachievedwith
ahighgrowthrateofheterotrophicbacteriawhilenitrificationisachievedbyautotrophicbacteria.Importantparametersincludemediasize,oxygenlevels,solids
retentiontimeandshearstress.BiodegradationoforganicmatterasmeasuredbythereductionofTOCseemedtohaveoccurredinJunebutnotSeptemberwhile
nitrificationoftheammoniaoccurredinJunebutnottheMarchorNovember/85samples.

Page91
TABLE1.Nitrogenandphosphorusconcentrations(mg/L)

Ammonia

Nitrate

TKN

Total

Ortho

Nov.21/85

Raw

.014

.018

.32

<.02

<.02

<1.0

Filtered

.014

.016

.34

<.02

<.02

<1.0

March21/86

Raw

.029

<.01

.44

.10

.031

<1.0

Filtered

.040

<.01

.46

.10

.044

<1.0

June6/86

Raw

.026

.019

.31

.10

<.02

14

Filtered

<.01,

.010

.35,

.075,

<.02,

4.6,

<.01

.031

.32

.068

<.02

5.8

0.19

.036

.27

.035

<.01

4.8

Sept.29,30/86

Raw

<.01

<.01

.14

.12

<.02

3.5,3.0

Filtered

<.01

<.01

.089

.14

<.01

4.0,3.0

Nov.4/86

Raw

<.01

.044

.21

.021

<.02

Filtered

<.01

.033

.12

<.02

<.02

Nitrogen

Phosphorus

TOC

TABLE2.Summaryofcoliformandstandardplatecount

OPERATINGFILTERS

Coliform(MPN/100mL)

StandardPlateCount(MPN/100mL)

Raw

Filtered

Raw

Filtered

13,7.8

2.0

2400,<1.0

30,000

January21/86

<2.0

49,22

March20/86

<2.0,33

<2.0

1,110,67

40

April28/86

<2.9,240

<2.0,<2.0

750,10,000

100,320

<2.0

<2.0,<2.0

500

70,30,100

November15/85

June5/86

September29/86

<2.0,<2.0
<2.0,<2.0

<2.0,<2.0

16,000
<1,5200

5,<1

Coliform
AsummaryofthetotalcoliformbacteriaandstandardplatecountisshownonTable2.Thetotalcoliformwastypicallybelowthedetectablelimitinthefilteredwater
exceptinNovember/85andJanuary/86samples.Inthelattercasesmediabiologicalactivitywaslikelyminimalduetolowwatertemperaturesandloworganicmatter.
ThestandardplatecountinthefilteredwaterwasalsoveryhighintheNovember/85samplebutwasgenerallyreducedovertherawwaterlevelexceptononesample
takeninJune/86.Theschmutzdeckewasonlyabout12hoursoldatthetimeofsampling.
PilotFilterBwasspikedwithtotalcoliformonNovember27andMarch20,1986withwatersamplestakenapproximatelyevery1to2hoursforupto36hours
afterspiking.Thewatersampleswereanalysedfortotalcoliform.
OnNovember22,1986thenumberofcoliformequivalentunitssuppliedwas1108orabout75,100MPN/100mL.Table3isasummaryofthecoliformcountin
therawandfilteredwatersampleswithtime.Thereisnoexplanationfortheresultspresented.Notonlywasthepeakcoliformcountintherawwaterdetected11
hoursafterspikingbutthepeaktreatedwatercoliformcountsoccurredat6,13and30hoursafterspikingandwerehigherthantherawwatersample.
OnMarch20,1986thenumberofcoliformequivalentunitssuppliedwas1108or75,000MPN/100mL.Table4isasummaryofthetotalcoliformcountinthe
rawandtreatedwatersampleswithtime.Thepeakrawwatercoliformcountwas1,600,000MPN/100mLintheinterval0to1:15hafterstartofspiking,and9,200
MPN/100mLinthewatersampletaken4:45hoursafterspiking.Thereductionwasintheorderofabout99.5%.
FilterMedia
OnApril3,June4andAugust21,1986samplesweretakenoftheschmutzdeckeandofthefiltermediaat300mmand600mmdepths.Onthelattertwodates
samplesweretakenimmediatelypriortofilterscraping.Theanalysesincludedbacterialactivityandenumerationofphytoplanktonandzooplankton.
TheheterotrophicbacterialactivityofsandsamplesfromthreedepthsarepresentedonTable5.Thehighestlevelsofbacterialactivitywereatthefiltersurfaceorthe
schmutzdecke.OnJune3,1986andAugust21,1986theschmutzdeckeswereabouttoberemovedasthefiltersurfacewassealedoff.Thebacterialactivitywas
lowintheMarchsampleprobablyduetothecontinuouslowrawwatertemperatureof1to3Csincefilterstartup.BacterialactivityincreasedintheJuneandAugust
samples,likelyreflectingahigherwatertemperatureandmorefavourableclimateforbacterialactivity.

Page92
TABLE3.ColiformspikingNovember27,1986
ElapsedTimeFrom
StartofSpiking
(h:mins)

TotalColiformCount
(MPN/100mL)
RawWater

FilteredWater

49,00079,000

0:40

23

1:40

540,000

170

3:00

13,000

540,000

6:00

3,500

1,600,000

8:00

13,000

540,000

11:00

1,600,000

49,000

13:00

23,000

1,600,000

15:00

110,000

920,000

18:00

23,000

540,000

19:30

3,300

350,000

30:00

1,600,000

30:30

79,000

2,300

TABLE4.ColiformspikingMarch20,1986
ElapsedTimeFrom
StartofSpiking
(h:mins)

TotalColiformCount
(MPN/100mL)
RawWater

FilteredWater

3,300

4.5

1:15

1,600,000

2.0

2:45

350,000

3,500

4:45

350,000

9,200

8:45

17,000

2,400

11:45

11,000

540

14:15

9,200

540

17:15

3,500

920

23:55

17,000

540

25:55

16,000

540

34:10

79,000

920

TABLE5.Heterotrophicbacterialactivityfiltermedia
RateofResazurinReduction(mole/g/hour)
Depth
(mm)

March17

June3

August21

17

0.32

0.060

300

0.12

0.016

600

0.016

TABLE6.Summaryofinvertebrates
TotalNumberofSpecimens
per400gofSample
Depth
(mm)

TotalNumberofTaxa
per400gofSample

March

June

August

March

June

August

17

13

61

2264

300

852

600

337

AsummaryofinvertebratenumbersfoundinthemediasamplesissummarizedonTable6.Countsofinvertebrateswithinthesandfilterdemonstratedanincreasing
trendfromMarchtoAugustatalldepths,withthemajorityoftheorganismsbeinglocatedinthetoplayer.InMarch,organismswerefoundonlyinthetoplayer.They
werefewinnumberandwerelikelyentrappedfromtheincomingwater.Overtime,colonizationbyorganismsmoresuitedtotheenvironmentoccurred,whereinJune
thetaxaChironomidae(midgefly)predominated.InAugustthetaxaOligochaeta(worm)wasdominant.Oligochaetesaretrueaquaticanimalswhichfeedonbacteria,
andtheirincreasingpresencereflectsthematurationofthebiologicalecosystemwithinthefilter(i.e.thedevelopmentofindigenousspecies)andindicatesaneffective
filtrationprocess.
SummaryandConclusions
TheremovalofGiardiacyststhroughtheslowsandfilterwasessentiallycompleteevenintheabsenceofschumutzdecke.Asimilarremovalwasfoundoverfifteen
categoriesofothermicroorganisms.Whenthepilotfilterwasspikedwith5,000cysts/L,anextremelyhighconcentrationcomparedtocystsdetectedintherawwater,
theremovalefficiencywas99.99%.
Theremovalwasinspiteofthefactthattheheterotrophicbacterialactivityandinvertebratepopulationwerenotmeasurableuntilthesummerperiod.Thiswasdueto
thecoldwatertemperaturesandthesmallamountofbiodegradableorganicmatterintherawwaterduringthewinterof1985andspringof1986.Amaturebiological
ecosystemwasdetectedonlyintheAugustsamplingperiodwhenthewatertemperaturepeakedat19Candthealgaewasheavy.
TheincreaseinheterotrophicbacterialactivitywasalsolikelyresponsibleforthereductionintheammoniaandTOCconcentrationsinthesummerperiodandthe
absenceofsameinthewinterof1985andspringof1986.Theimpactoftheslowerformingnitrateswasnotedinthesummerof1986whenthenitratelevelsinthe
filteredwaterwerehigherthanintherawwater.TheTKNwasalsoobservedtodecreasebetweenrawandfilteredwaterbutonlyinthefallof1986.Total
phosphoruswasreducedinthesummerof1986,possiblybecauseofanincreaseintheactivityofheterotrophs.
Thereductioninturbiditywasabout2050%betweenNovember/85andApril/86whileitwas4080%intheperiodMay/86toNovember/86.Thefilteredwater,
with

Page93

oneexception,waslessthantheB.C.MinistryofHealthobjectiveof1NTU.Theincreasedefficiencyinthesummerandfallof1986waslikelyduetothe
effectivenessoftheschmutzdeckeintrappingandholdingtheparticles.Thisactivitywasnegligibleinthewinterof1985andspringof1986.
Thetotalcoliformandstandardplatecountreductionincreasedwiththeincreaseintheheterotrophicactivityandinvertebratepopulation.Ononeoccasioninthe
summerof1986thefilteredwaterstandardplatecountwashigh.Thiswassome12hoursafterscrapingwhenduetovillagewaterdemand,thefilterwasoperatingat
ahighfilterrate.Thetrendtohigherremovalefficiencywithincreasingfilterageandbiologicalactivitywasnotedwiththespikedpilotfilters.
Acknowledgements
ThisresearchwascompletelyfundedbyHealth&WelfareCanadathroughSupplyandServicesCanada.TheauthorsthankDr.R.S.TobinoftheDepartmentof
NationalHealthandWelfareCanada,Ottawa.Thanksalsotothevillagecouncilandstaff,especiallyMr.G.MillsandMr.R.Hume.
LiteratureCited
1.Bellamy,W.D.,Silverman,G.P.,Hendricks,D.W.andG.S.Logsdon.1985.RemovingGiardiacystswithslowsandfiltration.J.AWWA77(1):52.
2.Bellamy,W.D.,Silverman,G.P.,andD.W.Hendricks.1984.FiltrationofGiardiacystsandothersubstances.Volume2:Slowsandfiltration.ColoradoState
UniversityEnvironmentalEngineeringTechnicalReport5847843forU.S.EPAundercontractCR880865002.
3.Bellamy,W.D.,Silverman,G.P.andD.W.Hendricks.1983.Giardialambliaremovalbyslowsandfiltration.In:ProceedingsofSundaySeminaron
InnovativeFiltrationTechniques.AmericanWaterWorksAssociationAnnualConference.LasVegas,Nevada.
4.HuismanL.andW.E.Wood.1974.Slowsandfiltration.WorldHealthOrganization,Geneva,Switzerland.122p..
5.Liu,D.andW.M.J.Strachan.1980.Characterizationofmicrobialactivityinsedimentbyresasurinreduction.Arch.Hydrobiol.Berh.12:2431.
6.Logsdon,G.S.,Hendricks,D.W.andG.R.Pyper.1983.ControlofGiardiacystsbyfiltration.Thelaboratory'srole.In:ProceedingsofAmericanWaterWorks
AssociationWaterQualityTechnologyConference,Norfolk,Virginia.AWWA.
7.Logsdon,G.S.andE.C.Lippy.1983.Theroleoffiltrationinpreventingwaterbornedisease.J.AWWA74(12):649.
8.McConnel,L.K.,Sims,R.C.andB.B.Barnett.1984.Reovirusremovalandinactivationbyslowratesandfiltration.J.Appl.Env.Microbiol.48(4):818.
9.StandardMethodsfortheExaminationofWaterandWastewater.1985.16thEdition,APHA,AWWAandWPCF,Washington,D.C..

Page95

ComparisonofSomeFiltrationProcessesAppropriateforGiardiaCystRemoval
GaryS.Logsdon
DrinkingWaterResearchDivision,WaterEngineeringResearchLaboratory,U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio45268,U.S.A..
Slowsandfiltration,diatomaceousearth(DE)filtration,andcoagulationfiltration(includingconventionaltreatment,directfiltrationandinlinefiltration),havebeenevaluated
forGiardiacystremovalatpilotplantand/orfieldscale.Properlydesignedandoperated,theaboveprocessescanattain99%cystreductions,orhigher.Thispaperdiscusses
relativeadvantagesanddisadvantagesoftheprocesses,andfactorsthatmayresultinsuccessorfailureoftreatment.Slowsandfiltrationistheleastcomplicatedfromthe
operator'sperspective.Itmaybethemostappropriateforsmallsystemsiftherawwateristreatable.Itveryeffectivelyremovesviruses,bacteriaandcystsbutisnotveryeffectivefor
removalofTHMprecursorororganicchemicals.Itgivestheoperatortheleastabilitytochangetreatmentinresponsetochangesinrawwater.DEfiltrationisveryeffectiveforcyst
removal,butremovalofverysmallparticlesrequiresuseoffinegradesofDEorchemicalpreconditioningofDE.Processmodificationscanyieldironandmanganeseremoval.THM
precursorremovalissmall.Operatorskillsrequiredaremostlyofamechanicalnature.Coagulationfiltrationhasthegreatestflexibility,andcanremove30to50%ofTHM
precursoralsoturbidity,microorganisms,andmetalsthatcanbeprecipitatedbeforefiltration.Manyfactorsinfluenceprocessperformancesoagoodunderstandingof
coagulationchemistryisneededformosteffectiveoperationregardlessofplantsize.Thisrequiresthegreatestlevelofoperatorabilityforcontinueddependableperformance.
Processvariationsincludeconventionaltreatment,directfiltrationandinlinefiltration.

Introduction
WaterbornegiardiasisoutbreakshavebeenoccurringintheUSAforthepasttwodecades,andcontinuetooccur.Thissuggestsaneedforbetterwatertreatment.
DisinfectionprovidesabarrierforwaterbornetransmissionofGiardiacysts.Craun(8)reportedthat19,770casesofwaterbornegiardiasiswererelatedto
deficienciesintreatmentofsurfacewatersourcesbycommunitywatersystemsfrom1965through1984.Ofthese,61%wererelatedtofailurestoadequately
disinfectinsystemshavingdisinfectionastheonlytreatment.Anotherbarrieriseffectivefiltration.Thispaperreviewsfiltrationstudiesatpilotscaleorfullscale,or
both,andcomparesperformancecapabilitiesandadvantagesofslowsandfiltration,diatomaceousearth(DE)filtration,andcoagulationfiltration.Thelattercategory
includesconventionalfiltration(coagulantfeedandrapidmix,flocculation,sedimentation,andfiltration),directfiltration(coagulantfeedandrapidmix,flocculation,and
filtration),andinlinefiltration(coagulantfeedandrapidmix,followedbyfiltration).
Alloftheabovefiltrationprocesses,iftheyareproperlydesignedandoperated,andiftheyaretreatingasourcewaterofsuitablequality,canreducetheconcentration
ofGiardiacystsby99%ormore.Filtrationfailurescanoccurbecauseofimproperdesignoroperation,orbecauseagivenprocessisnotappropriatefortheraw
waterbeingtreated.Ofthe19,770casesofgiardiasismentionedabove,38%occurredbecauseoffailuresinfiltration.Aspectsoffilterplantdesignandoperationare
discussedinsubsequentsectionsofthispaper.Therelativecostsoftheprocessesarenotdiscussedbecausethesewouldbeinfluencedbyconditionsthataresite
specificthusgeneralcomparisonswouldbeoflimitedusefulness.
SlowSandFiltration
SlowsandfiltrationstudieshavebeensupportedinrecentyearsbytheU.S.EnvironmentalProtectionAgency,theAmericanWaterWorksAssociationResearch
Foundation,andtheStateofUtah,amongothers.SomeparametersinEPAfundedstudiesaregiveninTable1.Filtershavebeenevaluatedforabilitytoremove
Giardiacysts,bacteria,turbidity,particles,andtrihalomethane(THM)precursor.
Slowsandfiltershavebeenshowncapableofremoving99to99.99%oftheGiardiacystsinrawwater(3,4,21).UsingpilotfiltersBellamyetal.(3)foundthatcyst
removaldidnotdeteriorateafterfilterscraping.Pyper(21)observedthatat7.5Cto21C,cystremovalwas99.98%to99.99%.At0.5Cto0.75C,removal
rangedfrom99.36%to99.91%however,at0.5C,cystremovaldeterioratedto93.7%whenbothGiardiacystsandprimaryunchlorinatedsewageeffluentwere
addedtotherawwatersimultaneously.Inthissituation,theloadingoforganismsintheinfluentwatermayhavebeengreaterthantheestablishedbiologicalpopulation
oftheslowsandfiltercouldcopewith.
Totalcoliformremovalwasfoundtobeadverselyinfluencedbyincreasesinfiltrationratefrom0.04to0.4m/h(3),bydecreasesinfilterbeddepthfrom0.97mto

Page96
TABLE1.Parametersinslowsandfilterresearch
FilterDesign
Sand
Size
mm

RawWaterQuality

Reference
Other

Uniformity
Coefficient

Filtration
Rate
m/h

Bed
Depth
m

Temperature
C

Turbidity
NTU

TotalColiform
per100mL

0.17

2.1

0.12

0.76

~25roomtemp.

<110

1010,000

(11)

0.32

1.4

0.12

0.94

228

<1to>30

4010,000

0.2143mg/m3
chlorophylla

(7)

0.33

2.8

0.08

1.07

025

0.259

1to8,700

(2.126)106
Giardiacystsspiked
[35425cysts/Lif
dilutedoverfilter
uniformly]

(21)

0.13to
0.62

1.5to1.6

0.04to
0.40

0.48to
0.97

217

2.711

0209,000

505,075Giardia
cysts/Lspiked

(3,4)

0.48m(4),byincreasesinsandsizefrom0.13mmto0.61mm(4),andbydecreasesintemperaturefrom17Cto2C(4).Oftheseparameters,the0.61mmsand
sizewouldbegreaterthansizestypicallyusedandmighthaveaccentuatedtheadverseimpactofthatvariable.Theuseof0.61mmsandresultedinaveragetotal
coliformremovalof96%vs.99.4%for0.13mmsand.Temperaturedecreasesfrom17Cto5Cor2Cresultedindeteriorationincoliformremovalfromthe99%
leveltoabout90%forthecolderwaters.Cleasbyetal.(7)foundthattotalcoliformremovalwaslowerduringthefirsttwodaysafterscrapingthanduringthe
remainderoftherun.Insomeinstances,differencesinthetwotimeperiodswereslight,but5of9runsexhibitedcoliformremovalsrangingfrom82%to95%during
thefirsttwodays.Duringtheremainderoftheruns,removalsrangedfrom97%to100%.LettermanandCullen(14)inmostcasesdidnotobserveanyeffectsof
scraping(aripeningperiod)intotalcoliformdatacollectedinastudyofsevenoperatingslowsandfilterplantsintheStateofNewYork.
Virusremovalhasbeenreported(Taylor,NoDate)tobeinfluencedbytemperatureandfiltrationrate.At0.20m/hand11to12C,removalwas99.9999%vs.
99.8%for0.40m/hand6C.Inanothersetofexperiments,Taylorreported99.8%removalat0.20m/hbutonly91%at0.40m/h.
Researchershaveobservedvariationintheabilityofslowsandfilterstoreduceturbiditytothe1NephelometricTurbidityUnit(NTU)MaximumContaminantLevel
(MCL)specifiedintheU.S.EnvironmentalProtectionAgency'sDrinkingWaterRegulations.Foxetal.(11)foundthatwhenwaterfromagravelpitinsouthwestern
Ohiowasfilteredat0.12m/h,afteraninitialripeningperiodhadallowedthebiopopulationtobecomeestablishedonnewsand,the1NTUMCLwasalwaysmet.
Rawwaterturbidityrangedfrom0.2to10NTU.Cleasbyetal.(6)reportedthatafterthefirsttworuns,typicaleffluentturbiditywas0.1NTUexceptduringthefirst
twodaysafterscraping.WaterforthatresearchcamefromagravelpitincentralIowa,withturbidityrangingfrom<1to30NTU.Pyperobservedslowsandfiltered
waterturbidityof0.1NTUorlessfor50%ofthetime,and1.0NTUorlessfor99%ofthetimeinMcIndoeFalls,Vt.ThesourceofwaterwasCoburnPond,a
bodyofwaterwithanopenwatersurfaceareaofabout4ha,plusabout20haofwetland.Rawwaterturbidityrangedfrom0.4to4.6NTUandcoloraveraged24
C.U.Incontrastwiththeseresults,whenHorsetoothReservoirwastreated(3,4),thefilteredwaterturbidityrangedfrom3NTUto5NTU,andthe1NTUMCL
wasnotmet.RawwaterturbidityofHorsetoothReservoirgenerallywas6NTUto8NTU.SlezakandSims(23)reportedthatabout15%of27plantssurveyed
producedfilteredwaterwithanaverageturbidityof1.0NTUorhigher,whereasturbidityaveraged0.4NTUorlowerathalfoftheplants.
Thedifferentdegreesofturbidityreductioninsomecasesmaybeattributedtothenutrientconditionofthefilters.WatercollectedhighintheRockyMountainsand
transportedtoHorsetoothReservoirwouldnotbeexpectedtobehighinnutrientsforgrowthofbiopopulationinfilters.Bellamyetal.(4)reportedaddingsterile
nutrient(BODabout4mg/L)toonetestfilter,whichshouldhaveincreasedthebiopopulationinthefilter.Underparalleloperation,turbidityreductionaveraged52%
fromthisfiltervs.15%fromthefiltertreatingunalteredHorsetoothReservoirwater.Pavonietal.(20)reportedthatexocellularpolymersproducedbybacteriainan
activatedsludgeculturewerecapableofflocculatingkaolinitesuspensionsandpromotingsettling.Itappearspossiblethatthebiologicalpopulationofaslowsandfilter
mayproduceexocellularpolymersthatenhancethe''stickiness"offiltermediaandinorganicparticlesintheslowsandfilter,thusimprovingthefilter'scapabilityto
removesuchparticles.ThesurfacewaterstestedinIowaandOhiocontainedsufficientnutrienttosupportalgaeduringthesummer,andthewaterinVermontwould
beexpectedtobehighin

Page97

nutrientsresultingfromdecayingvegetationinthewetlands.Thus,wewouldinferthatthosewatershadhighernutrientlevelsthantheHorsetoothReservoirwater.
SlowsandfiltersshouldnotbeexpectedtoremovelargeamountsofTHMprecursorunlesssomethinghasbeendonetochemicallyaltertheprecursorbefore
filtration.Humicmaterials,althoughincontactwithmicroorganismsinnature,seemtopersistintheenvironment.ThebiopopulationintheVermontslowsandfilter
removedabout10%ofthetrihalomethaneformationpotential(THMFP)thatwasbetween100g/Land200g/Linrawwater.Foxetal.(11)reportedTOC
removalof19%andTHMFPremovalof18%whentreatingsouthwesternOhiogravelpitwater.
IntheresearchatIowaState(7),algaewereencounteredandevaluatedforremovalandinfluenceonfilterefficiency.Chlorophyllameasurementswerelessthan5
g/Lduringthewinterandspringof19811982,untilmidApril,increasingtonearly60g/LinlateApril.ChlorophylladeclinedinMayandJunebutappearedto
peaknear140g/LinJuly.Algalbloomsoccurred,andtheseinfluencedrunlength.Fourrunsrangedinlengthfrom10to22dayswhenmeanchlorophyllavalues
were8to138g/L.Runsof34to123dayswereassociatedwithchlorophyllavaluesof1to4g/L.Algaeremoval,asmeasuredbychlorophyllareductions,was
quitehighandsimilartoremovalofotherparticulatematter(generallyapproaching99%).
Rawwaterqualitylimitsforslowsandfiltersarestringentbecauseparticulatemattertendstoberemovedatthetopofthefilterandbecauseslowsandfiltershave
limitedcapabilitytoremoveinorganiccontaminantsandsyntheticorganicchemicals.Cleasbyetal.(6)reportedthatenumerationofalgaeorperformingasurrogate
measureofalgalpopulationwasnecessarytojudgethesuitabilityofrawwaterforslowsandfiltration.Foxetal.(11)reportedthattreatmentofOhioRiverwater
(0.423NTU)resultedinprogressivelypoorerfilteredwaterqualityover250daysofoperation,witheffluentturbidityexceeding1NTUduringthelast20daysof
operation,andtimetoterminalheadloss(0.4m)decreasingfrom98daysto6days.Duringthefirst230days,meaninfluentturbidityrangedfrom2.4to7.6NTU,
levelsthatdonotseemexcessivelyhigh.Averagerawwaterturbiditywas10NTUorlowerat90%oftheoperatingplantssurveyedbySlezakandSims(1984).
Experiencethusfarsuggeststhatthemostreliablewaytodeterminetreatabilityofwaterbyslowsandfiltrationistoconductanextendedpilotplantstudy.
Slowsandfiltersaresimpletooperateandmaintain,whenrawwaterqualityisappropriateandwhentheplantsaresmallenoughthatcomplicatedequipmentisnot
neededforfilterscraping.Dailydutiesatasmallinstallation(10,000to1,000,000L/day)wouldincludereadingandrecordingheadloss,flowratesortotals,chlorine
residual,rawandfilteredwaterturbidity,andadjustingflow.
LettermanandCullen(14)studiedfilterscrapingatsevenslowsandfiltrationplantsinNewYork.Averageflowsrangedfrom1to23millionL/day.Scraping,or
removalofathinlayerofsandwhenterminalheadlossisreached,requiredanaverageof5hoursper100m2offiltersurface.Thethicknessofthelayerremovedwas
typically2to3cm.Thefrequencyofscrapingwouldbedeterminedbyrunlength,whichwouldbeinfluencedbytheturbidityandalgaeintherawwater.Afterasand
filterhasbeenscrapedanumberoftimes,thefulldepthofthebedisrestoredinanoperationcalledresanding.LettermanandCullenestimatedthatresandingadepth
of15to30cmwouldrequire4859hoursoflaborper100m2.
Theadvantagesofslowsandfiltersarerelatedmainlytothesimplicityinherentintheprocess.Smallplantsaresimplytoconstruct.Simple,manuallycontrolledvalves
canservetocontrolflow.Headlosscanbemeasuredbyapiezometer.Becausechangesinheadlossoccurslowly,recordingequipmentisnotneeded.Coagulant
chemicalsarenotusedinslowsandfiltration,sooperatorsdonotneedtounderstandcoagulationchemistry.Chemicalfeedpumpswouldnotbeneededforcoagulant
chemicals,sofewerpumpswouldbeused,loweringmechanicalmaintenancework.Operatorskillsdonotneedtobeashighasforplantsusingcoagulation.Another
advantageassociatedwithabsenceofcoagulationisaminimumofwastedisposalproblems.Scrapedsandisessentiallytheonlywaste,andoftenitiswashedand
reused.
Manyofthedisadvantagesofslowsandfiltrationarealsorelatedtotheabsenceofcoagulation.Withoutpretreatment,limitationsexistonthequalityofwaterthatis
suitableforslowsandfiltration.Thesewereexplainedearlier.Becausemodifyingaslowsandfilterplanttotreatadifficultwatermightbecostly,ornotpossible,pilot
studiesshouldbeperformedtoverifytreatability.Inaddition,astudyshouldbeconductedtoestablishthattherawwatersourceisnotlikelytochangeordeteriorate
inqualitytosuchadegreethatthewaterwouldbecomeuntreatableinthefuture.Thismaynotalwaysbepossibletoascertain,butaneffortshouldbemadetopredict
whatsortofhumanactivitiesordevelopmentmighthappenintheforeseeablefuture.Thiswouldatleastalertauthoritiestopossibleneedforchangesintreatmentif
rawwaterqualitydeteriorated.Becausepretreatmentisminimalornonexistentatslowsandfilterplants,littlecapabilitygenerallyexiststoremovesyntheticorganic
chemicals,trihalomethaneprecursors,anddissolvedinorganicsubstancessuchasheavymetals.Inaddition,veryfineclaysorglacialflourmaynotbereadilyremoved.
Finally,slowsandfiltersmaynotbeappropriateformediumtolargeinstallationsintheUSA,becauseofoperatinglaborcostsandlandcosts.Thetrendforlarge
systemsistoautomateandusemechanicalequipmentwherepossible,butcleaningenclosedslowsandfiltersbymechanicalmeansisverydifficult.Thus,theyseem
mostappropriateforsmallsystemslocatedonveryhighqualitysourcewaters.
DiatomaceousEarthFiltration
Diatomaceousearth(DE)filtershavebeenstudiedforremovalofavarietyofcontaminants.TheyhavebeenshowntoattainexcellentremovalofGiardiacystsovera
broadrangeofoperatingconditions.Cystremovals

Page98

exceeding99%,andoften99.9%,werereportedbyLangeetal.(12)forfiltrationratesof2.4to9.6m/h,fortemperaturesfrom3.5to15C,andforfourdifferent
gradesofdiatomaceousearth(Celite545T M,Celite535T M,Celite503T M,andHyfloSuperCelT M).
Pyper(21)reported99.97%foroneDEfilterruninwhichGiardiacystswereadded.Logsdonetal.(15)reportedthatwhensufficientDEprecoatandbodyfeed
wereused,removalof9mradioactivebeadswasnearlyalways99.9%orhigher.Useofaprecoatofatleast1.0kg/m2wasshowntobeappropriateforobtaining
mosteffectiveremovalofthe9umparticles.TheyalsoreportedthatelevenfilterrunsweremadewithG.muriscystsatfiltrationratesof2.2to3.5m/h,withCelite
535T Mprecoatandbodyfeed.Cystremovalexceeded99.0%inallruns,andexceeded99.9%infiveoftheruns.DeWalleetal.(9)reportedonfourDEfilterruns
conductedforGiardiacystremoval.Cystremovalexceeded99%ineachofthefourruns.TheoverallresultsofallresearchforGiardiacystremovalindicatethat
DEfiltrationisveryeffectiveforcontrollingGiardiacysts.Factorsimportanttocontinuedeffectiveperformanceareusingadequateprecoatandbodyfeed,and
keepingtheseptumveryclean(goodcleaningattheendofeachrun).
RemovaloftotalcoliformbacteriabyDEfiltrationwasstudiedextensivelyatColoradoStateUniversitybyLangeetal.(12).Coliformremovalswerestrongly
influencedbythegradeofdiatomaceousearthused.Coarsergradesattainedremovalsrangingfrom30%to50%forCelite545T Mandfrom50%to70%withCelite
503T M.Thefinegrades,withsmallerpores,wereconsiderablymoreeffective.RemovalwithCelite512T Mwas92%to96%,andtotalcoliformremovalwithSuper
CelT Mwas99.92%togreaterthan99.98%.
Malinaetal.(18)reportedthatahighpercentageofremovalcouldbeattainedforpolioviruswhencoatedDEfilteraidwasusedorwhencationicpolymerwasadded
totherawwater.Inone12hourfilterrun,diatomaceousearthcoatedwith1mgofcationicpolymerpergramofDEproducedfilteredwaterinwhichnoviruseswere
recoveredfrom11samples(removal>99.95%).Oneof12sampleswaspositive,andinthisinstance,virusremovalwas99%.Ina12hourruninwhichuncoated
DEwasusedand0.14mg/Lofcationicpolymerwasaddedtotherawwater,noviruseswererecoveredfromanyofthe12samplesanalyzed.
TurbidityremovalwhentreatingHorsetoothReservoirwater,asreportedbyLangeetal.(12),waslessthan20%forthegradesofdiatomaceousearthcommonly
usedforwatertreatment(Celite545T M,Celite535T M,Celite503T M,andHyfloSuperCelT M).TurbidityoftheHorsetoothReservoirrawwaterrangedfrom4.5to
5.4NTU.Thefinestgradetested,FilterCelT M,couldreducetheturbiditybyover95%.Incontrasttotheseresults,Logsdonetal.(15)reportedthatturbidity
reductionsof56%to78%wereattainedwithCelite535T Mwhenrawwaterturbidityrangedfrom0.95to2.5NTU,butlittlechangewasobservedwhenrawwater
turbidityrangedfrom0.24to0.45NTU.Pyper(21)reportedanaverageturbidityreductionof71%,withaneffluentqualityof0.5NTU.
PyperevaluatedDEfiltrationforremovalofTHMprecursorinVermont.Resultsshowednodifferencebetweentherawwaterandthefilteredwater,suggestingthat
theTHMprecursormaterialpresentinCoburnPondwasdissolved.Thewaterwascolored(24CU,average),andthismayexplainthelackofchangeduring
filtration,becauseDEfiltrationalonedoesnotremovecolor,aknownprecursor.
BecauseturbidityremovalwiththegradesofdiatomaceousearthcommonlyusedforwatertreatmentwassolowwhenHorsetoothReservoirwaterwasfiltered,the
ColoradoStateUniversityresearchersinvestigatedthenatureoftheturbidity(2).When5.6NTUrawwaterwasfiltered,turbiditywasreduced2%bya5mpore
sizemembrane,36%bya1.2mmembrane,73%bya0.45mmembrane,and91%bya0.22mmembrane.Mostofthelightscatteringmatterinthewater(the
causeoftheturbidity)wasmadeupofparticlesthatcouldpassthrough1.2mpores,andthusfineenoughtopassthroughtypicalpotablewatergradesofDE.
AdditionalworkwasdoneatColoradoStateUniversitytoimprovethecapabilitiesofDEfiltration.Inordertoalterthesurfacepropertiesofdiatomaceousearth,
aluminumhydroxidewasprecipitatedtothesurfaceofaDEslurry.With0.05gramsofalumpergramofCelite545T M,totalcoliformremovalwas99.86%,as
comparedto30%to50%removalforuncoatedCelite545T M.ForthesamegradeofDE,turbidityremovalwas98%,forcoatedDEvs.under20%foruncoated
DE(12).TheseresultsshowthatthestrainingmechanismofremovalcanbeaugmentedbyasurfaceattachmentremovalmechanismifDEisgivenanelectropositive
coating.
LimitsonthequalityofrawwaterthatwouldbeappropriateforDEfiltrationarenoteasytoset.Theprocessremovesparticulatematterbytrappingitwithinthefilter
cake.Astheconcentrationofparticulatematterinrawwaterincreasestheloadappliedtothefiltercakeincreases.Tomaintainhighpermeabilityofthefiltercakeand
goodheadlosscharacteristics,bodyfeeddiatomaceousearthisaddedtotherawwater.Aruleofthumbisthathigherrawwaterparticleconcentrationsrequiremore
bodyfeed,ifthenatureoftheparticlesdoesnotchange.Thenatureoftheparticlesbeingremovedisquiteimportantthough,especiallythecompressibility.Rigid
turbiditycausingparticles,suchasveryfinesand,wouldnotblockorblindthefiltercake,butcompressibleparticles,suchasalgae,coagulationfloc,precipitatediron,
orbiologicalmattercouldblindthefiltercake.PilotfiltrationstudiesareadvisableifthewaterinquestionisnotalreadybeingtreatedbyDEfiltration.Suchstudies
wouldestablishtheappropriategradeofDEtousetoobtainthedesiredeffluentturbidity,theamountofbodyfeedtoaddunderconditionsofthetestruns,andthe
approximatelengthoffilterruntoexpect.LettermanandLogsdon(13)surveyed13DEfiltrationplantsandreportedthatfilteredwaterturbiditiesabove1NTUor
filterrunsof6orfewerhourswereobservedatDEplantshavingmaximumrawwaterturbiditiesof20NTUorgreater(Figure1).Thisfigureshowsthepercentageof
plantsexceedingspecifiedvaluesforminimum,average,andmaximumraw

Page99

waterturbidity.Symbolsshowninthelegendidentifyplantproblemswithhighfilteredwaterturbidityorshortrunsorboth.
Operationandmaintenanceofdiatomaceousearthfiltersaresomewhatmorecomplexthanforslowsandfilters,butlesscomplicatedthancoagulationfiltration.Daily
monitoringwouldincludeturbidity,disinfectionresidual,rateofwaterproductionwithadjustmentsifneeded,filterheadloss,andrateofuseofbodyfeed.Periodic
choreswouldincludepreparationofbodyfeedslurryandprecoatslurryandmaintenancechecksonbodyfeedandprecoatpumps.Also,filterswouldneedtobe
backwashedperiodically,butdisposalofspentfiltercakeshouldpresentfewproblems,becauseitisnotgelatinousanddriesreadily.Filterelements(septa)needto
bekeptveryclean.Thecleanlinessoftheseptacanbereadilycheckedifvacuumfiltersorquickopeningpressurefiltersareused.Becauseofthenumberofpumps,
valves,andothermechanicalitemsinuseataDEfiltrationplant,operatorsshouldpossessgoodmechanicalskills.Knowledgeofcoagulationchemistrywouldnotbe
neededunlessthediatomaceousearthwasconditionedbythealumcoatingtechnique.
Diatomaceousearthfiltrationhasseveralimportantadvantages,especiallywithrespecttotreatingwatersthatmaycontainGiardiacysts.Theprocesshasbeenshown

Figure1.
Influenceofrawwaterturbidityondiatomaceousearthplantperformance

infourstudiestobeveryeffectiveforcystremoval,andtheremovalefficiencyisnotaffectedbyverylowtemperatures.Differentgradesofdiatomaceousearthcanbe
keptonhand,givingtheoperatorsomeflexibilityifthegradeinusepassestoomanyturbiditycausingparticles.Ifnecessary,thesurfaceattachmentpropertiesofthe
coarsegradesofdiatomaceousearthcanbemarkedlyenhancedbythealumcoatingprocedure.Diatomaceousearthfilterplantsdonotrequirelargelandarea,and
areinuseforcapacitiesupto50or60millionL/day.
Amongthedisadvantagesofdiatomaceouseartharetheneedforhighqualityrawwater,theinabilitytoremovedissolvedsubstances,andtheinabilitytoremovevery
fineparticleswithplaindiatomaceousearth.Excessivesuspendedsolids(turbidity,algae)inrawwatercancauseshortfilterruns.Bubblesmayformandcollapsein
thefiltercakeifthevacuumDEfiltersareusedtotreatcold,highlyoxygenatedwater.IfpressureDEfiltersareusedandoperatedtohighheadlosstoobtainlong
runsandeconomicaluseofDEprecoatmaterial,highenergycostsmayresult.
CoagulationFiltration
TheprocesstrainusedmostoftenintheUnitedStatesforfiltrationinvolveschemicalpretreatment(coagulation,andfrequentlyflocculationandsedimentation)
followedbydeepbedgranularmediafiltration.MostU.S.coagulationfiltrationresearchforGiardiacystremovalhasfocusedonthecoagulationfiltration(inline)or
coagulationflocculationfiltration(directfiltration)variationsoftheprocess,becausewaterbornegiardiasisoutbreakstendedtobeobservedinregionsofthecountry
thathadlowturbiditywaterswhichwerethoughttobesuitableforsuchtreatment.ResearchbyLogsdonetal.(15),DeWalleetal.(9),andAlAnietal.(1)involved
coagulationwithalum,oralumplusapolymerfiltrationthroughsandordualmediaat5to14m/handtemperaturerangingfrom3to20C.Laterresearch(17)was
conductedonconventionaltreatment,withalumoralumandpolymer,dualmediaandthreemonomediatypes(sand,anthracite,GAC),filtrationat7m/h,androom
temperatures(about25C).
ResultsofthethreeciteddirectfiltrationstudiesindicatethatGiardiacystremovalcanexceed99.0%oreven99.9%whentherawwateriscoagulatedproperlyand
filtered.ResultsofLogsdonetal.(15)andDeWalleetal.(9)indicatedthatwithproperpretreatment,cystremovalexceeded99.0%whenfilteredwaterturbiditywas
below0.30NTU.AlAnietal.(1)showedthatcystremovalof99%ormorewaslikelytooccurifturbidityremovalwas70%ormore,whenrawwatersinthe0.2
to1NTUrangeweretreated.Thiswouldproducefilteredwatersinthe0.06to0.30NTUrange.
Alloftheaboveresearchersshowedthatdependablecystremovalresultscannotbeattainedifaclearwater(about1NTU)isfilteredwithoutbeingproperly
coagulated.Useofnocoagulant,orofanimproperdose,resultedinerraticcystremovalresults.Inaddition,DeWalleetal.(9)showedthatforalumcoagulation,

Page100

usingtheproperpHisnecessarywhensoft,lowalkalinitywateristreated.TheyobservedeffectivetreatmentatpH5.6and6.2,butatpH6.8withalumcoagulation,
cystremovalwasreducedfrom99%to95%.
Thecoagulationfiltrationprocesscanremoveavarietyofcontaminants.Robecketal.(22)showedthatdirectfiltrationcouldremove90%to99%ofviruses,while
conventionaltreatmentremovalsconsistentlywere99%.McCormickandKing(19)statedthatcoliformremovalbydirectfiltrationwaspractically100%when
filteredwaterturbiditywas0.10NTUorless.Cleasbyetal.(6)reportedthatinlinefiltrationremovedmorethan86%ofthetotalcoliformbacteriainrawwater,after
thefirsthourofthefilterrunhadpassed,in10testruns.Edzwald(10)showedthatdirectfiltrationcouldremovenonpurgeabletotalorganiccarbon(NPTOC)and
organicprecursormaterialsthatformtrihalomethanes(TTHMFP,ortotaltrihalomethaneformationpotential).Withcationicpolymerastheprimarycoagulant,both
NPTOCremovalandTTHMFPremovalwereabout40%whereaswithalumastheprimarycoagulantremovalsofNPTOCandTTHMFPwerenearly60%.With
thesamewaters,whenconventionaltreatmentwasemployedwithalumastheprimarycoagulant,removalsofNPTOCandTHMFPwereabout70%.Cleasbyetal.
(6)reportedthatwaterswithlowtomoderatealgalpopulations,watercouldbetreatedbydirectfiltration.Waterwithfewalgaehadachlorophyllaconcentrationof
lessthan5g/L(7).Waterwithanalgalpopulationsufficienttoresultinachlorophyllaconcentrationof130g/Lcouldnotbeeffectivelytreatedbydirectfiltration
withoutprechlorination.
Suggestedlimitsonrawwaterqualityforsourcesreceivingcompleteconventionaltreatment(includingpredisinfection,coagulation,sedimentation,rapidgranular
filtration,andpostdisinfection)weregiveninthe"ManualForEvaluatingPublicDrinkingWaterSupplies"asamonthlygeometricmeanofnotmorethan2,000fecal
coliformper100mLoramonthlygeometricmeanofnotmorethan20,000totalcoliformbacteriaper100mL,colornottoexceed75units,odornottoexceeda
thresholdodornumberof5,andturbiditynottobesohighastooverloadthewatertreatmentworks(26).
Suggestedlimitsonrawwaterqualityfordirectfiltrationandinlinefiltrationaremuchmorestringent.Cleasbyetal.(6)suggestedthataveragerawwaterturbidity
shoulddependonwhethertheprimarycoagulantisalumoracationicpolymer,andonwhetheralgalpopulationislowormoderate.Suggestedvaluesrangedfrom7
NTUformoderatealgaeandalumcoagulationto16NTUforlowalgaeandcationicpolymercoagulation.TheDirectFiltrationSubcommitteeoftheAWWA
FiltrationCommittee(5)reportedthatwaterswithlessthan40unitsofcolor,turbiditybelow5NTU,ironlessthan0.3mg/L,manganeselessthan0.05mg/L,and
algaecountsupto2000ASU/mLappearedtobe"perfectcandidatesfordirectfiltration."Inasurveyof17directfiltrationplants(13),shortfilterruns(6orfewer
hours)wereoccasionallyobservedwhenmaximumrawwaterturbiditywas8NTUorhigher,andbothshortrunsandfilteredwaterturbidityabove1NTUwere
sometimesobservedwhenrawwaterturbiditywas20NTUorhigher(Figure2).Thisfigureshowsthepercentageofplantsexceedingspecifiedvaluesforminimum,
average,andmaximumrawwaterturbidity.Symbolsshowninthelegendidentifyplantproblemswithhighfilteredwaterturbidityorshortrunsorboth.Fromthework
ofEdzwald(10),itcanbeinferredthatiftheTHMformationpotentialofawaterexceeds0.20mg/L,directfiltrationmaynotbeabletoproduceawaterthatwill
meetthe0.10mg/LMCLfortrihalomethanes.
OperationandmaintenanceforcoagulationfiltrationplantscanbemoredemandingthanthatforDEplantsorslowsandfilterplants.Bothconventionalplantsand
directfiltrationplantsshouldbemonitoredcarefully,becausefailuretoobtainoptimumcoagulationcanresultinpoorfilterperformance.Conventionalplantsare
generallyconsideredtohavea"marginofsafety"withrespecttocoagulationcontrol.Ifcoagulationcontrolislostatthechemicalfeedandrapidmixpoint,becauseof
thehoursofdetentiontimeaffordedbysettlingbasins,andifthisgoesunnoticeduntilthepoorlycoagulatedwaterreachesthefilters,plantoperatorscouldfind
themselvesinthedilemmaofhavingsettlingbasinsfullofwaterthatcouldnotbefilteredsuccessfully.
Coagulationmonitoringandcontrolareveryimportant,whetherornottheplantemployssedimentation.Onetraditionalapproachtocontrolisjartesting.For

Figure2.
Influenceofrawwaterturbidityongranularmediaplantperformance

Page101

watersofperhaps10NTUorhigher,jartestingcombinedwithcontinuousmonitoringoftheturbidityofthefilteredwateratindividualfiltersisanapproachfrequently
used.Ifrawwaterqualitycanchangerapidly,oriftherawwaterturbidityislow(below10NTU),jartestsmaynotbeveryeffective,becauseofthetimerequiredfor
testing,orbecauseofthesmallerdifferencesinrawandsettledwaterturbidities.Insuchinstances,coagulantdosecontrolbyzetapotentialinstrumentation,a
streamingcurrentdetector,orapilotfiltermaybeappropriate.WagnerandHudson(25)suggestedthatfilterpaperfiltrationusingWhatmanNo.40papercouldgive
informationonthetreatmentlevelsthatproduceacceptablewaterquality.OtherappropriatemonitoringwouldincludepH,headloss,chemicalfeed,andrawand
filteredwaterturbidity.
Maintenanceoperationswouldincludecareofchemicalmixersandfeeders,perhapsflocculationbasinmixersandsludgeremovalequipmentinsettlingbasins.Filter
backwashingisnecessary,andbackwashwaterandsettlingbasinsludgemayrequiretreatmentandultimatedisposal.Ifsludgeremovalfromsettlingbasinsisnotdone
mechanically,periodicmanualbasincleaningwouldbeneeded.
Thelevelofoperatingskillneededatcoagulationfiltrationplantsissubstantial.Inordertoeffectivelyandefficientlycontrolthecoagulationfiltrationprocessandattain
lowfilteredwaterturbidity,operatorsneedtounderstandthechemicalaspectsofcoagulation.Largeandmediumsizedplantsareabletohireandkeeptrained
operatorswhocaneffectivelyoperatecoagulationfiltrationplants.Ontheotherhand,smallplantsmaynothavetheresourcestohireortrainoperatorswhohavea
solidunderstandingofcoagulation.Thiscanleadtoproblemsofpoortreatedwaterquality,ifoperatorsareunabletoadjusttreatmentwhenrawwaterquality
changes.
Ofthethreeprocessesdiscussedinthispaper,coagulationfiltrationhasthegreatestflexibilityinthekindandconcentrationofcontaminantsthatcanberemovedinthe
process,especiallywhensedimentationisemployed.Conventionaltreatmentcanhandlethewidestrangeinrawwaterquality,andhasbeeninuseforseveral
decades.Coagulationfiltrationplants,becausetheyemploymoretreatmentprocesses,canbedesignedwiththemostflexibilityintermsofthenumberofprocesses
used.Forexample,settlingmightbeusedformuddywaterbutbypassedwhenrawwaterturbidityislow.Recentdevelopments,suchasuseofmediainthe1to2
mmsizerange,bedsabout2mindepth,andfiltrationratesof25m/horhigherprovideevenmoretreatmentcapabilityforthecoagulationfiltrationprocess,butuntil
experiencewithsuchplantsisgained,theveryhighratesoffiltrationprobablyshouldbeconsideredonlyatlargewaterutilitieswithwelltrained,fulltimeoperatorsand
laboratorypersonnel.
Inspiteofthemanyadvantagesthatcanbelistedforcoagulationfiltration,anumberofdrawbacksexist.Themostimportantpotentialproblemisthis:forrapidrate
granularmediafiltrationtobeaneffectiveprocessforremovalofparticulatematter,thechemistryofthewatermustbemanipulatedsothatcoagulationiseffective.
ThiscanbedonethroughadjustmentofpHandadditionofaninorganiccoagulantorpolymerorboth.Atutilitiesthatserve50,000to100,000personsormore,
hiringoneormorescientiststoworkinawaterqualitycontrollaboratorycanbeconsideredfeasible,asitispresentlybeingdone.Atwaterutilitiestoosmallto
employachemist,operationofthecoagulantfiltrationprocessmaybelessthanoptimum.Testingbypersonswhounderstandtheprocesscanestablishtheproper
chemicaltreatmentundertherawwaterqualityconditionsexistingduringthetestperiod.Theoperator'sfailuretounderstandtheimplicationsofchangingrawwater
qualityandmakeproperadjustmentscouldresultinlowerprocessefficiency,though.Afundamentalconceptisthatcoagulationchemistryisnotinfluencedbythe
magnitudeoftheflowinaplant.FactorssuchaspH,alkalinity,andtemperaturemustbeconsidered,regardlessofthesizeoftheplant.
Aparticularconcerninnorthernlatitudesormountainousareaswheregiardiasisoutbreaksmayhaveoccurredisthedifficultyofeffectivelycoagulatingandfiltering
cold,clearwaters.Whentherawwaterturbidityiscloseto1NTU,someplantoperatorsmayquestionthevalueofaddingacoagulant.Othersmaybediscouraged
bytheapparentdifficultyintreatingaclearwaterattemperaturescloseto0C,andinbothinstances,operatorsmayshutoffthechemicalfeeders.Coagulantfeed
shouldneverbeinterruptednorshutoff.Techniquesareavailablefortreatingcoldwatersandlowturbiditywaters.Performingjartestswiththejarsinanicewater
bathisappropriate.Useofpaperfiltersorsmall(2.5cm)minifilterswithbeds30cmdeep,orshallower,couldbeusedtoevaluatefilterabilityofclearwaters.Useof
streamingcurrentdetectorsasanonlinecoagulantdosecontroldeviceappearstoworkwellinwinter.Experienceindicatesthatcoagulationfiltrationplantscan
producehighqualitywaterevenwhentemperatureandturbidityintherawwaterarelow.TheDuluth,Minnesotafiltrationplantconsistentlyproducedfilteredwater
below0.10NTUandattained99%to99.99%reductionsofasbestosfibersevenwhentemperatureswereinthe3to5Crangeandrawwaterturbiditywas1NTU
(16).
Conclusions
(i)Eachofthethreefiltrationprocessesreviewedisdifferent,andnosingleprocessisidealineverycircumstance.
(ii)Asprocesscomplexityincreases,fromslowsandfiltration,toDEfiltration,tocoagulationfiltration,theskilllevelneededforeffectiveoperationincreases.
(iii)Asprocesscomplexityincreases,producinghighqualityfilteredwaterincreasinglybecomesdependentonoperatorskillandability.
(iv)AvarietyoffiltrationprocesseshavebeenusedsuccessfullyeitheronapilotplantscaleoratfullscaletoremoveGiardiacystsfromwater.

Page102

(v)Slowsandfiltration,DEfiltration,andthecoagulationfiltration(inlineordirectfiltration)processesusedwithoutsedimentationareallaffectedbyrawwater
quality,withrespecttobothfilteredwaterqualityandplantperformancecharacteristics,suchasfilterrunlength.Therefore,ifuseofanyoftheseprocessesis
contemplatedwithawatersourcethatisnotpresentlybeingtreatedsuccessfullybytheprocess,performingapilotplantstudybeforedesignandconstructionofthe
treatmentplantishighlyadvisable.
(vi)Eventhoughimportantlimitationsexistandmustbetakenintoaccount,filtrationtechnologycapableofremoving99%ormoreoftheGiardiacystsfromdrinking
waterexistsandisinuseinmanylocations.
Disclaimer
MentionoruseofcommercialproductsdoesnotconstituteendorsementbytheU.S.EnvironmentalProtectionAgency.
LiteratureCited
1.AlAni,M.Y.,Hendricks,D.W.,Logsdon,G.S.andC.P.Hibler,1986.RemovingGiardiacystsfromlowturbiditywatersbyrapidratefiltration.J.American
WaterWorksAssoc.78:5:6673.
2.Bellamy,W.D.,Lange,K.P.andD.W.Hendricks.1984.FiltrationofGiardiacystsandothersubstances:Volume1.diatomaceousearthfiltration.EPA600/2
84114,U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio,1967.
3.Bellamy,W.D.,Hendricks,D.W.andG.S.Logsdon.1985a.Slowsandfiltration:Influencesofselectedprocessvariables.J.AmericanWaterWorksAssoc.
77:12:6266.
4.Bellamy,W.D.,Silverman,G.P.andD.W.Hendricks.1985b.FiltrationofGiardiacystsandothersubstances:Volume2.Slowsandfiltration.EPA600/2
85/026,U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio.
5.Bishop,S.,Craft,T.F.,Fisher,D.R.,Ghosh,M.,Prendiville,P.W.,Roberts,K.J.,Steimle,S.andJ.Thompson.1980.Thestatusofdirectfiltration,Committee
report.J.AmericanWaterWorksAssoc.72:7:405411.
6.Cleasby,J.L.,Hilmoe,D.J.andC.J.Dimitracopoulos.1984a.Slowsandanddirectinlinefiltrationofasurfacewater.J.AmericanWaterWorksAssoc.
76:12:4455.
7.Cleasby,J.L.,Hilmoe,D.J.,Dimitracopoulos,C.andL.M.DiazBossio.1984b.Effectivefiltrationofsmallwatersupplies.EPA600/284083,U.S.
EnvironmentalProtectionAgency,Cincinnati,Ohio.
8.Craun,G.F.,1986.WaterbornegiardiasisintheUnitedStates19651984.LancetII:8505:513514.
9.DeWalle,F.B.,Engeset,J.andW.Lawrence.1984.RemovalofGiardialambliacystsbydrinkingwatertreatmentplants.EPA600/284069,U.S.
EnvironmentalProtectionAgency,Cincinnati,Ohio.
10.Edzwald,J.K.,1986.Conventionalwatertreatmentanddirectfiltration:Treatmentandremovaloftotalorganiccarbonandtrihalomethaneprecursors,p.199
236.In:OrganicCarcinogensinDrinkingWater:Detection,TreatmentandRiskAssessment,Ram,N.M.,Calabrese,E.J.andR.F.Christmaneds.,JohnWiley
&Sons,N.Y.
11.Fox,K.R.,Miltner,R.J.,Logsdon,G.S.,Dicks,D.L.andL.F.Drolet.1984.Pilotplantstudiesofslowratefiltration.J.AmericanWaterWorksAssoc.
76:12:6268.
12.Lange,K.P.,Bellamy,W.D.,Hendricks,D.W.andG.S.Logsdon.1986.DiatomaceousearthfiltrationofGiardiacystsandothersubstances.J.American
WaterWorksAssoc.78:1:7684.
13.Letterman,R.D.andG.S.Logsdon.1976.SurveyofdirectfiltrationpracticePreliminaryreport.PresentedatAmericanWaterWorksAssociationAnnual
Conference,NewOrleans,Louisiana.June,1976.
14.Letterman,R.D.andT.R.Cullen,Jr.,1985.SlowSandFilterMaintenance.EPA/600/285/056,U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio.
15.Logsdon,G.S.,Symons,J.M.,Hoye,R.L.,Jr.andM.M.Arozarena.1981.AlternativefiltrationmethodsforremovalofGiardiacystsandcystmodels.J.
AmericanWaterWorksAssoc.73:2:111118.
16.Logsdon,G.S.,Evavold,G.L.,Patton,J.L.andJ.Watkins,Jr.1983.Filterplantdesignforasbestosfiberremoval.J.ofEnvironmentalEngineering.109:4:900
914.
17.Logsdon,G.S.,Thurman,V.C.,Frindt,E.S.andJ.G.Stoecker.1985.EvaluatingsedimentationandvariousfiltermediaforremovalofGiardiacysts.J.
AmericanWaterWorksAssoc.77:2:6166.
18.Malina,J.F.,Jr.,Moore,B.D.andJ.L.Marshall.1972.Poliovirusremovalbydiatomaceousearthfiltration.Centerforresearchinwaterresources,The
UniversityofTexas,Austin,Texas.
19.McCormick,R.F.andP.H.King.1982.Factorsthataffectuseofdirectfiltrationintreatingsurfacewaters.J.AmericanWaterWorksAssoc.74:5:234242.
20.Pavoni,J.L.,Tenney,M.W.andW.F.Echelberger,Jr.,1972.Bacterialexocellularpolymersandbiologicalflocculation.J.WaterPollutionControlFederation
44:3:414431.
21.Pyper,G.R.1985.Slowsandfilterandpackagetreatmentplantevaluation:Operatingcostsandremovalofbacteria,Giardia,andtrihalomethanes.EPA/600/2
85/052,U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio.
22.Robeck,G.G.,Clarke,N.A.andK.A.Dostal.1962.Effectivenessofwatertreatmentprocessesinvirusremoval.J.AmericanWaterWorksAssoc.
54:10:12751290.
23.Slezak,L.A.,andR.C.Sims.1984.TheapplicationandeffectivenessofslowsandfiltrationintheUnitedStates.J.AmericanWaterWorksAssoc.76:12:3843.
24.Taylor,E.W.NoDate.Fortyfifthreportontheresultsofthebacteriological,chemicalandbiologicalexaminationoftheLondonwatersfortheyears19711973.
MetropolitanWaterBoard,London,England.
25.Wagner,E.G.andH.E.Hudson,Jr.1982.Lowdosagehighratedirectfiltration.J.AmericanWaterWorksAssoc.74:5:256261.
26.WaterSupplyDivision,U.S.EnvironmentalProtectionAgency.1980.Manualforevaluatingpublicdrinkingwatersupplies.EPA430/975011.Washington,
D.C.

Page103

MonitoringasaToolinWaterborneGiardiasisPrevention
JanL.Sykora*,WilderD.Bancroft,AlbertH.Brunwasser,StanleyJ.States,MauriceA.Shapiro,SusanN.BoutrosandLouisF.Conley
GraduateSchoolofPublicHealth,DepartmentofIndustrialEnvironmentalHealthSciences,UniversityofPittsburgh,PittsburghPA15261
InDecember1983awaterborneepidemicoccurredinMcKeesport,PAwhere347peopledevelopedgiardiasis.Aseriesofunusualevents,suchasexceedinglylowtemperatures
andwaterlinebreakswereblamedforthisoutbreak.Aftertheoutbreakwascontrolled,amonitoringprogramforGiardiacystswasintroducedandtheexistingturbidity
monitoringprogramwasexpanded.UsingtheEPAtechniquesampleswerecollectedfromrawwater(YoughioghenyRiver)andMcKeesportfinishedwateronafortnightlybasis.
ThemajorsourceofGiardiacystswastreatedanduntreatedsewagecontaminatingtheYoughioghenyRiver.OnJanuary12,1986acystwasagainisolatedfromafinishedwater
sample.Subsequentanalysisrecoveredcystsfromareservoir,astandpipeandtapsamples.A"boilwateradvisory"wasissuedonJanuary17,1986.Avigorousprocessofremedial
actions,whichincludedrepairoffiltersandincreasedchlorineadditiontoachieveaCTvalueof240,wereinstituted.Nogiardiasiscasesabovethebackgroundlevelwere
reportedandthe"boilwateradvisory"wasliftedonFebruary21,1986.Themonitoringresultsshowedthatturbiditycannotbesolelyrelieduponwhenevaluatingthe
effectivenessoffiltrationplantsforGiardiacystremoval.TheGiardiamonitoringofrawwaterindicatedannualaswellasseasonalvariationsincystcountswiththegreatestnumber
ofcystsdetectedincolderseasons.BasedonthisstudyitisrecommendedthatpublicwatersystemsdeterminedtobeatriskofGiardiacontaminationconsideraroutinemonitoring
program.Itisfurthersuggestedthatmonitoringeffortsbeintensifiedduringthecolderseasonoftheyear.

Introduction
OnFebruary22,1984theAlleghenyCountyHealthDepartment(ACHD)declaredthatawaterbornegiardiasisepidemicwasoccurringintheMcKeesport
MunicipalWaterAuthority's(MMWA)serviceareaandissueda"boilwater"advisory.Anepidemiologicalinvestigationconfirmedthehypothesisthatdrinkingwater
wasthevehicletransmittingGiardiacyststothe48,700peopleintheservicearea.ByApril20,1984therewere347medicallyconfirmedcasesofgiardiasisinthe
affectedpopulation.
TheMMWAislocatedsouthofPittsburghattheconfluenceoftheMonongahelaandYoughioghenyRivers.Itisa9MGD(34.1ML/d)treatmentfacilityutilizing
chemicalcoagulation,flocculation,sedimentation,filtration,andchlorination.Theplant'srawwatersupplyisderivedfromtheYoughioghenyRiver.Thetreatmentplant
wasconstructedin1907toserveresidentsinthecommunitiesofMcKeesport,Versailles,PortVue,WhiteOak,andtheirmajorindustry,theNationalTubeCo..
Theinvestigationintothecauseofthegiardiasisoutbreakhasdeterminedthatitoccurredasaresultofacombinationofseveraleventsandcircumstances.Inlate
December1983unusuallylowambienttemperatureswereexperiencedoveranextendedperiodoftime,resultinginnumerouswaterlinebreaksthatwerenot
promptlyidentifiedorrepaired.OnebreakoccurredinawaterlineinthebedoftheYoughioghenyRiver.Anotheroccurredinanindustrialplantthatwasshutdown.
Stillotherbreaksoccurredinthewaterdistributionnetwork.Monitoringcapabilitiesofthewatersystemhadnotbeenmodernizedtoenableexpeditiousidentification
ofthelocationandmagnitudeofsuchproblems.Theresultinghighwaterdemandexceededthetreatmentcapacityoftheplant'scoagulation,flocculation,
sedimentationandfiltrationsystems.Additionally,becauseanelevatedbackwashtankwasoutofserviceatthetime,allbackwashwaterwasbeingsuppliedfromthe
distributionsystem.Consequently,filterswererunforseveraldayswithoutbackwashingandalargescalebreakthroughledtoasignificantincreaseinturbidity(5).
Thesecondmajorfactorcontributingtotheoutbreakwasthatoverseveralprecedingdecades,thewatertreatmentplanthadnotbeenmaintainedproperlyandthere
werenumeroustreatmentdeficiencies.Insummary,theoutbreakoccurredasaresultofthepresenceofGiardiacystsintherawwatersupplyandexcessivewater
demandswhichcompromisedthetreatmentcapabilitiesoftheplant.
AtthetimeofthegiardiasisoutbreaktheexperienceofwaterindustryregulatorsinPennsylvaniadealingwiththisdiseasewasinitsinfancy.InalloftheUnitedStates,
therewereonlyafewlaboratoriescapableofanalyzingwatersamplesforGiardiacysts.TheAlleghenyCountyHealthDepartment,incooperationwiththeFederal
EnvironmentalProtectionAgencyandthePennsylvaniaDepartmentofEnvironmentalResourcesissuedtwoorderstotheMMWA.Thefirst,issuedonMarch16,
1984,requiredremedialmaintenanceofthe
*Correspondingauthor.

Page104

plantandthedisinfectionofreservoirsandthedistributionsystemtoallowtheliftingofthe"boilwater"advisory.Thesecondorder,basedonanengineeringanalysis
bytheMMWA'sconsultantandissuedonSeptember17,1984,directedtheconstructionofanewwatertreatmentplant.
Theordersalsoestablishedathreephasemonitoringprogramasabarrieragainstfuturewaterbornegiardiasisoutbreaks.Inthefirstphase,theAlleghenyCounty
HealthDepartmentexpandedtheFederalSafeDrinkingWaterAct's(7)turbiditymonitoringprotocoltorequiretheMMWAtosamplefinishedwatereveryfour
hours.TheFederalSafeDrinkingWaterActrequiresthatturbiditiesbelessthan1NTUfora30dayaverageandnotexceed5NTU'sfora2dayaverage.The
AmericanWaterWorksAssociation(1)hasrecommended0.1NTUasaqualitygoalforpotablewater.Thesecondphaseoftheprogramestablishedmonitoringof
thetreatmentplant'srawandfinishedwatersforGiardiacystsonafortnightlybasis.Thefinishedwatersamplingprogramusuallyincludedaclearwellsampleandan
openreservoirsample.ThethirdphaseinvolvedidentificationofsourcesofGiardiacystsoccurringintheMcKeesportrawwatersupply,theYoughioghenyRiver.
Theresultsofthislastphasewerepublishedelsewhere(9,10).
MaterialsandMethods
GiardiaSamplingProcedures
SamplescollectedatthewatertreatmentplantwereobtainedusingstandardGiardiafiltrationequipment(3).Inthecaseoffinishedwater,500600gallonsofwater
werefilteredatarateof12gal/min.Inthecaseofrawwaterlessthan200gallonswerefilteredduetoclogging.Followingsamplecollectionfilterswererefrigerated
andmaintainedat5Cuntilanalyzed.
GiardiaAnalysisProcedure
ThesampleswereanalyzedusingtheEnvironmentalProtectionAgencytechniquebySchaeferandRice(8)asmodifiedbySchaefer(personalcommunication,1984).
Inthelaboratory,thefiltersweredividedintofourthsbyunwindingtheyarnintoskeins.Eachskeinwasrinsedinsuccessive1literaliquotsof0.01%Tween20T M.
OnepercentTweensolutionwasaddedasneededtomaintainsuds.Eachsamplewasconcentratedbysedimentationunderrefrigerationfor24hours.Thesediment
wasthencentrifugedat600Xgfor3minutes,andtheresultingpelletscombined.12mLofsediment,resuspendedin7090mLof0.01%Tween20T M,werethen
layeredontopof70mLof1.5Msucrosesolution(sp.gr.1.18)ina250mLcentrifugebottle(8).Aftercentrifugationat800Xgfor5minutes,thesupernatantabove
theinterfaceandthepelletwerediscarded.Theinterfaceandallunderlyingsucrosesolutionwerediluted5timeswith0.01%Tween20T Mandfurtherprocessedby
centrifugationat600Xgfor2minutes.TheresultingsedimentwaswashedtwicewithTween20T Min50mLcentrifugebottles,transferredto15mLcentrifugetubes,
concentratedbycentrifugationandstainedbyLugol'ssolutionusingtheJakubowskiandEricksenprocedure(3).Thesedimentobtainedbythisflotationprocesswas
examinedmicroscopicallyusingaPalmerMaloneynannoplanktoncountingchamber(0.1mL)anda45xobjectiveonamicroscopeequippedwithphasecontrast.
PositiveidentificationoftheGiardiacystswasbasedondimension,shape,anddetectionofatleasttwointernalmorphologicalfeatures.
TurbidityMeasurements
TurbiditymeasurementswereperformedusingaHach18900continuousflowturbidimeterequippedwithachartrecordingdevice.Theturbidimeterwascalibrated
usingthemanufacturer'soperatinginstructionsandEPAqualityassurancestandards.Theturbiditymonitoring
procedure,establishedaftertheoutbreak,consistedofmeasurementseveryfourhoursresultinginsixdeterminationsperday.Anaveragevaluewascalculatedfor
eachday.Theturbiditywasmeasuredatthepressuresideofthepumpattheendoftheclearwellpriortoenteringthedistributionsystem.
Results
FromNovember1984throughthebeginningofDecember1986,atotalof43sampleswerecollectedfromtheYoughioghenyRiverattheMMWAplant.Thecyst
levelsinthesesamplesaveraged73.6cysts/100gal(Table1).Thestandarddeviation(100.9)exceededthemeanindicatinghighvariabilityincystsconcentration
duringthesamplingperiod.Theresultsshowthatmostofthesamples[20]containedbetween11and100Giardiacysts/100gal,elevensamplescontained110
cysts/100gal,andtwelvesamples101438cysts/100gal.Itisofinterestthatverylowcystcountsweredetectedinthesummerwithaveragesrangingfrom12.5
cysts/100galin1986to25.4cysts/100galin1985(Table1).Ontheotherhand,thegreatestnumberofcystswasdetectedincolderseasons.Theresultsalso
indicateannualvariationincystcounts.Thecystcountsweremuchlowerin1986thanin1985withaveragesof50.6cysts/100galand103.6cysts/100gal
respectively.Similardecreaseswerenotedwhenaveragesforindividualseasonsduring1985and1986werecompared(Table1).
Methodqualitycontrolstudiesperformedonsewageanddescribedinapreviouspublicationindicatethatthefiltrationandsampleprocessingproceduresinvolvedin
thestandardtechniquerecoveronly314%ofthecystspresent(3,10).Theseresultsagreewiththoseofotherauthorsandsuggestthattheactualconcentrationsof
Giardiacystsenteringthetreatmentplantwereprobablysubstantiallyhigherthandetectedbythestandardtechnique.
Aspreviouslyreported,samplescollectedfromfinishedwaterwerenegativeuntilacystwasisolatedfromaclearwellsamplecollectedonJanuary8,1986(10).
Subsequentsampleanalysesbylaboratoriesotherthanourownrecoveredcystsfromareservoir,astandpipeandtapsamples.Table2summarizestheresultsof
TABLE1.GiardiaCystCountsintheYoughioghenyRiver

Timeperiod

Range

Mean

StandardDeviation

#ofSamples

Nov84Dec86
JanDec85

1438

73.6

100.9

43

1438

103.6

132.1

21

JanDec86

3201

50.6

54.9

21

Winter85

2416

133.0

194.6

Spring85

35438

147.6

165.8

Summer85

1114

25.4

49.6

Fall85

12262

112.5

103.8

Winter8586

32201

84.3

79.4

Spring86

6148

46.7

53.4

Summer86

325

12.5

9.0

15141

74.0

51.6

Fall86

Cysts/100gal

Page105
TABLE2.McKeesportWaterAuthorityGiardiacysts/100gallons(December4,1985April25,1986)*
SamplingPeriod
RawWater

December
418

January
823

February
1127

March
1327

April
1025

36103

37**

32201

67148

721

Clearwell

01

Filters

02

Standpipe

04

Reservoir

01

DistributionSystem

01

TotalSamples

15

13

%ContaminatedSamples***

66

*FromSykoraetal.(10)
**Onesampleonly
***Rawwaterexcluded
Samplesnotcollected

ourGiardiacystanalysesperformedduringandshortlyafterdetectionofGiardiaintheclearwell.ThistableindicatesthatduringthecriticalperiodofJanuary823,
1986,66%ofdrinkingwatersamplescontainedcysts.ThefirstsamplesobtainedfromthedistributionsystemonJanuary13,1986(3samples)andonesample
collectedfromclearwellNo.4onJanuary15,1986,werenegative.However,threesamplescollectedfromthedistributionsystemonJanuary23,1986were
contaminated.TheAlleghenyCountyHealthDepartmentissueda"boilwateradvisory"onJanuary17,1986.Avigorousprogramofactionswhichincludedrepairof
filters,additionoffiltermediaandanincreaseinchlorineconcentrationtoachieveachlorineconcentrationxcontacttime(CT)of240wasalsoinstituted.NoGiardia
casesabovethebackgroundlevelwerereportedandthe"boilwateradvisory"wasliftedonFebruary21,1986.
SincetherevisedmonitoringprogramwasestablishedthedailyaverageturbiditymeasurementshavenotexceededtheFederalSafeDrinkingWaterStandards(7).
However,thequalitygoalof0.1NTUasrecommendedbyAWWAhasbeenexceededmostofthetime(1).Table3andFigure1describethevariationinturbidity
values,expressedasdailyaveragesduringthe1986incident.
TABLE3.Resultsofturbiditymeasurements(NTU)McKeesportClearwell(December1985February1986)
Date

Range

Mean

S.D.

Dec110

0.140.25

0.19

0.03

10

Dec1120

0.140.70

0.28

0.17

10

Dec2130

0.180.45

0.26

0.09

10

Jan110

0.250.75

0.41

0.15

10

Jan1120

0.190.52

0.30

0.11

10

Jan2131

0.150.75

0.40

0.19

10

Feb111

0.120.50

0.27

0.12

11

Feb1222

0.060.24

0.15

0.06

11

S.D.=Standarddeviation
N=Numberofdailyaveragesamples

Discussion
TheresultsshowthatturbiditymonitoringattheclearwellcannotbesolelyrelieduponwhenevaluatingtreatmenteffectivenessoffiltrationplantsforGiardiacyst
removal.TheMcKeesportexperiencedemonstratesthatafiltercouldbedefectiveinoneareaallowingthepassageofGiardiacysts,whilethecombinedeffluentof
allfiltersmasksthebreakthroughwhenusingturbidityasanindicator.ThisisconsistentwithearlierlaboratorystudieswithgranularfiltrationperformedbyLogsdonet
al.(4)whichshowedthattheconcentrationofcystspassingthroughafiltercouldbehigh,whilefinishedwaterturbiditiesremainedbelow1.0NTU.Logsdonetal.(4)
havealsoindicatedthattheremovalofG.muriscystsbydiatomaceousearth(DE)filtersdoesnotseemtobecloselyrelatedtoturbidityremoval.Inadifferent
publicationdirectlyrelatedtotheMcKeesportoutbreak,Logsdonetal.(5)showedsubstantialvariationsinGiardiacystpassageratesthroughtreatmentplantfilters
whilefinishedwaterturbiditieswerebetween0.3and1.2NTU.Substantiallyfewercystspassedthroughthefilterswhenturbiditywas0.20NTUorlowerhowever.
Thus,theMcKeesportfieldexperienceindicatesthatifturbidityismonitoredattheclearwell,ratherthanatindividualfilters,operatorsmaybeunawareofthe
conditionsofindividualfiltersthatcouldpermitpassageofGiardiacystsintofinishedwater.
OthercontributingfactorstothewaterbornegiardiasisoutbreakinMcKeesportwerehighconcentrationsandpersistentpresenceofcystsintheYoughioghenyRiver
withdistinctmaximaincolderseasons.Marroccoetal.(6)showedthatmostoftheGiardialambliacontaminationofunfilteredwatersystemsinPennsylvaniawas
detectedduringthetimeframeofDecemberthroughJune.Craun(2),whoevaluatedthecurrentstatusofwaterbornegiardiasisintheUnitedStates,showedthatthe
outbreaksassociatedwithcommunitywatersystemsoccurredmostfrequentlyinthespringandfall/earlywinter,whileoutbreaksaffectingvisitorsorcampersoccurred
mostfrequentlyduringthesummermonths.BothMcKeesportincidentshadonsetsinearlywinterandallrecent

Page106
TABLE4.RecentgiardiasisoutbreaksinPennsylvania
Location
Bradford
Pittston

BoilWaterAdvisoryDate
October26,1979
December23,1983

WilkesBarre/Scranton
Houtzdale
McKeesport

March9,1984
November14,1984
February22,1984

giardiasisoutbreaksassociatedwithcommunitywatersystemsinPennsylvaniaoccurredduringcoldweather(Table4).Thus,theresultsfromthisstudysupport
Craun'ssuggestionthattheoutbreaksaffectingpermanentresidentsmayreflectnotonlylesseffectivetreatmentduringfall/earlywinter,butalsoincreased
contaminationofrawwatersupplies.
ThehighconcentrationsofGiardiacystsduringcolderseasonsmaybetheresultofseveralfactorssuchasflow,temperature,waterdensityandlesseffectivesewage
treatment,aswellasincreasedsurvivalofthecystsincoldwater.Resultsfromourpreviousstudiesshowedthateffluentsfromsewagetreatmentplantslocatedonthe
YoughioghenyRivercontainedbetween50and5.1103Giardiacysts/100galwhilerawwastewaterconcentrationsrangedbetween6.6103and1.5105
cysts/100gal(9,10).Thus,thedischargeofraworpoorlytreatedsewageduringwintercanmakeasubstantialdifferenceinGiardiacystconcentrationsinrawwater.
Inconclusion,sincethemonitoringoffinishedwaterforthepresenceofGiardiacystsislimitedbyalaboratorytechniquethatresultsinalowrecoveryrate,thereisa
continuingneedforresearchanddevelopmentintheareaofsamplingandlaboratoryanalysistoimprovetheanalyticalprocedureforGiardiacystisolation.Inspiteof
theselimitations,theMcKeesportexperiencesuggeststhatmonitoringofrawandfinishedwatersuppliescanbeusefulandcancomplementturbiditymeasurements.
Therefore,waterutilitiessubjecttoGiardiacontaminationoftheirrawwatersource(s),maychoosetoadoptroutinemonitoringforGiardiawithemphasisoncold
seasonsampling.

Figure1.
McKeesportdailyeffluentturbiditiesfrom
December1,1985toFebruary22,1986

Acknowledgements
ThisworkwassupportedbythePennsylvaniaDepartmentofEnvironmentalResourcesandtheBureauofEnvironmentalHealth,AlleghenyCountyHealth
Department,Pittsburgh,Pennsylvania.
References
1.AmericanWaterworksAssociation.QualitygoalsforpotablewateradoptedbyAmericanWaterWorksAssociationBoardofDirectors,January28,1968.In:
198586OfficersandCommiteeDirectory,AWWA,Denver,Colorado,1986.
2.Craun,G.F..1984.Waterborneoutbreaksofgiardiasis:currentstatus.In:GiardiaandGiardiasis.S.L.ErlandsenandE.A.Meyer,(eds.).PlenumPress.New
YorkandLondon.pp.243259.
3.Jakubowski,W.,andT.H.Ericksen.1979.MethodsfordetectionofGiardiacystsinwatersupplies.In:WaterborneTransmissionofGiardiasis.W.
JakubowskiandJ.C.Hoff(eds.),EnvironmentalProtectionAgency600/979001.pp.193210.
4.Logsdon,G.S.,DeWalle,F.B.,andD.W.Hendricks.1984.Filtrationsasabarriertopassageofcystsindrinkingwater,In:GiardiaandGiardiasis.S.L.
ErlandsenandE.A.Meyer,(eds.).PlenumPress.NewYorkandLondon.pp.287309.
5.Logsdon,G.S.,Thurman,V.C.,Frindt,E.S.,andJ.G.Stoecker.1985.EvaluatingsedimentationandvariousfiltermediaforremovalofGiardiacysts.Journalof
AWWA77:6166.
6.Marrocco,F.A.,Lengel,L.L.,andD.N.Greenfield.1987.GiardiamonitoringandregulationinPennsylvaniasurfaceWaters.In:Proceedingsof1986AWWA
WaterQualityTechnologyConference,PortlandOR.,November1620.pp.10551066.
7.SafeDrinkingWaterAct.PublicLaw93523,December16,1974,WashingtonDC.,U.S.A..
8.Schaefer,F.W.,andE.W.Rice.1981.Giardiamethodologyforwatersupplyanalysis,In:WaterQualityTechnologyConferenceProceedings,Seattle,
December69,1981.AmericanWaterWorksAssociation,DenverCO.pp.143146.
9.Sykora,J.L.,Bancroft,W.D.,States,S.J.,Shapiro,M.A.,Boutros,S.N.,Keleti,G.,Turzai,M.,andL.F.Conley.1986.Giardiacystsinrawandtreated
sewage.In:ControllingWaterborneGiardiasis.G.S.Logsdon(ed),EnvironmentalEngineeringDivisionASCE(InPress).
10.Sykora,J.L.,States,S.J.,Bancroft,W.D.,Boutros,S.N.,Shapiro,M.A.,andL.F.Conley.1987.MonitoringofWaterandWastewaterforGiardia.In:
Proceedingsof1986AWWAWaterQualityTechnologyConference,PortlandOR.,November1620.pp.10431054.

Page107

TheEfficiencyofPointofUseDevicesfortheExclusionofGiardiamuriscystsfromaModelWaterSupplySystem
D.RoyCullimore*andHelenJacobsen
ReginaWaterResearchInstitute,UniversityofRegina,Regina,SaskatchewanS4S0A2,Canada.
Arecyclingwatersystemwasdevelopedwhichcouldcirculatewaterfreeofparticleslargerthan0.22mthroughvariouspointofusewatertreatmentdevicesincludingultraviolet
light,reverseosmosis,ozone,granulatedactivatedcarbonandproprietaryfaucetattachmentsystems.Onceoperating,therecyclingwaterwaschargedwithviableGiardiamuris
cystsandtheabilityofthevariousdevicestokilland/orexcludethecystswasmonitoredundervariousconditions.Viabilitywasdeterminedbyexcystationandanovelstaining
techniquereportedelsewhere.Thepresenceofthecystsinthewaterwasmonitoredusinga1.2mcelluloseacetatemembranefilter.Wherecystsbecameentrappedonthefilter,
thiswasmonitoredbyeitherpressuredifferentialshifts,dischargeflowratesand/orthedetectionofcystsonthefilter.Underabnormaloperatingconditionssuchasultrahigh
loadingsandprolongedoperationaltimes,allofthedevicesfailedtoeitherexcludeG.muriscystsfromtheproductwaterortokillallofthepassagedcysts.Therelativeefficiencies
ofthedevicestestedforeliminationofviablecystsfromtheproductwateraregroupedasfollows:(1)highefficiencyultravioletirradiation,ozone(2)moderateefficiency
reverseosmosisand(3)poorefficiencygranulatedactivatedcarbon,faucetattachments.

Introduction
WaterbornegiardiasisoutbreakshaverecentlybeenmorefrequentlyreportedontheNorthAmericancontinent.Giardiasisisadiarrhealdiseasewhichiscommonly
causedbythedigestionofGiardialambliacysts(1).Followingcystingestion,excystmentoccursinthesmallintestinewiththeemergenceofbilaterallysymmetrical,
flagellatedtrophozoiteswhichattachthemselvesviaasuckingdisktothewallsofduodenumandupperjejunum.Thiscausesanacuteorchronicdiarrhealillnesswhich
sometimesprogressestosteatorrheaand/orthemalabsorptionsyndrome(2,16).
In1976,theStateofWashingtonrecordeditsfirstconfirmedoutbreakofwaterbornegiardiasiswhichoccurredintheCityofCamas.Approximately600people
wereaffectedrepresenting1015%ofthepopulation.Thecitywatersupplywasablendedmixtureofsurfaceandwellwatersourcesinvolvingtwomountaincreeks
andsevendeepwells.Whilethewellwaterswerenotimplicated,G.lambliacystswererecoveredfromboththerawsurfacewatersupplyandthedistribution
system(15).Clearly,thewatertreatmentprocedureswhichincludedmixedmediafiltersandchlorinationhadfailedtoexcludethecystsfromthedistributionsystem.
Anothermajoroutbreakinanearbyareaoccurredinthewinterof198182attheBanffNationalParkinAlberta,Canada.Atotalof121confirmedcasesof
giardiasiswerereported.Theseandothersimilaroutbreakshighlightedtheneedtoreevaluatethecapabilitiesofthevariousoptionspresentlyusedinexistingwater
treatmentfacilitiesspecificallytoremoveGiardiacystsfromrawwatersupplies.Ithasbeenproposedthatoutbreaksofgiardiasiswhicharewaterborneinoriginmay
betheresultofprocessdeficienciesinthewatertreatmentplants,excessivecontaminationofwatersourcesforwhichthetreatmentprocessbecomesinadequate,or
thatthecontaminationofthewateroccurredaftertreatment(8,14).Ingeneral,ithasbeenobservedthattheGiardiacystsaregenerallymoreresistanttodisinfection
thanarethenormal''hygieneindicator"organismssuchascoliforms(4,12).Unfortunately,thestandarddisinfection(chlorination)practicesrecommendedforwater
treatmenthavebeenfoundtobeineffectivebecauseofinsufficientdosagesand/ortooshortacontacttime(6).Treatmentbyultravioletradiationhasbeenconsidered
asmajoralternativetochemicaldisinfectionproceduresforsmallwatersystemsbecauseofitssimplicityandeconomyinoperation.However,RiceandHoff
suggestedin1981thatinareaswherethewatersuppliescontainG.lamblia,theuseofultravioletradiationatconventionaldosagesprovedtobeinadequateforthe
satisfactorydisinfectionofpotablewatersupplies(11).
Anotherdisinfectionprocedurereceivingincreasedattentionistheuseofozone.Mostresearchhasfocusedonthecontrolofaquaticbacterialandviralpopulations
withverylittleemphasisonthecontrolofprotozoalcysts.ResearchattheUniversityofWashington(3)concludedthat,ingeneral,protozoalcystsaremoreresistant
toozonethanarebacteriaandviruses.Inaddition,thedisinfectivecapabilityoftheozonewasaffectedverysignificantlybybothpHandtemperature.Atlow
temperatures(1C)andhighpH(79),Giardiacystswerefoundtopossessagreaterresistancetoozonethanatroomtemperaturesandtheslightlyacidicwaters
withapHrangeof5to7.
*Correspondingauthor.

Page108

Oneoptionavailabletotheconsumerofapotablewatersupplyinwhichtheuser,forwhateverreason,considersthatthereisaninherentriskinconsumingwateristo
installapointofusedevicetoremovetheperceivedrisk.Therecentpublicityonoutbreaksofgiardiasishasgeneratedclaimsthatsomeofthesedeviceswill
specificallyexcludeGiardiacystsfromtheproductwater.Apointofusedevicecanbedefinedasasingleormultipleseriesofwatertreatmentdeviceswhichwhen
interfacedwiththesupplywillcauseadesignedexclusiononsuchselectedagentsaswillrenderthefinalproductacceptableasapotablewatersupply.Ingeneral,
theseunitsareeitherfreestandingbatchsystems,undersinkrestrictedcontinuousflowondemandorunrestrictedcontinuousflowsetonthewholewatersystem.
ThisstudywasdesignedtodeterminetheeffectivenessofalimitedrangeofpointofusedevicesinexcludingGiardiamuriscystsartificiallysupplementedintoa
particulatefreewater.Thespecificselectedunitsfunctionedbyeitherultravioletlight,activatedcarbon,reverseosmosis,filtrationorozonation.
Methods
SelectionofthePointofUseSystems
Avarietyofunitswereevaluatedinordertodeterminewhichsinglerepresentativewouldbetestedforeachsystem.Inallcases,emphasiswasplaceduponunitsthat
werelikelytobeinstalledinadomestichomeinvolvinganaveragethroughputintotalof1,000L/day.Selectionalsowasbasedupontheclaimsofthevarious
manufacturersforeachstyleofunit.Theseclaimsaresummarized(inbrackets)foreachsystem:ultraviolettreatment(killswaterbornebacteriaandviruses)reverse
osmosis(willremoveupto95%oftotaldissolvedsolidsincludingsalt,sulfates,sediments,bacteria,algae,phosphates,mercury,arsenic,nitrates,chlorine,detergents,
rust)ozonation(killsbacteria,removesodour,taste,colourandoxidesmanyorganicandinorganicimpurities)granulatedactivatedcharcoal(removestasteand
odor,reduceslevelsofpotentiallytoxicorcarcinogeniccompounds,producesawaterwithalowerviableorganismcount)faucetfilters(aidsintheremovalof
chemicals,contaminants,chlorine,toxicsubstancessuchastrichloroethyleneandtrihalomethane,iron,rust,silt,sediment).Fromtheserecommendationsitisevident
thatnonerelatespecificallytotheexclusionofprotozoalcystsandscientificsupportingevidenceforsomeoftheotherclaimsappearstenuous.
Toevaluatethesedevicesundercomparableconditionstosmalldomesticsystems,atestsystemwasdesignedwhichcouldutilizearecyclingwatersystemwitha
controllableflowintowhichGiardiamuriscystscouldbeinoculated.TheconfigurationofthesystemisgiveninFigure1.Waterwasretainedinastoragetank
holdingapproximately80Lofwaterandwaspumpedfromtheretoapressuretankandonthrough1/2inchP.V.C.pipingthroughtotwoseparaterecyclable
systems.Theoperatingrangefortherecyclingunitincludedfluctuatingwaterpressureoperatingover30to80p.s.i.withflowratesadjustableuptoamaximumof3
gal/min(11L/min).Inmostexperiments,singledistilledwaterwasusedunlessotherwisestated.
Oneloopedsystemwasspecificallydesignedtotestareverseosmosisfiltrationsystem(UnbottledWaterSystemsmanufacturedbyWetcoLtd.,LasVegas,
Nevada).Waterwaspumpedthroughthedeviceandthe"brine"waterdischargerecycledbackintothemainsystemflow.Waterwhichhadpassedthroughthe
membranesofthereverseosmosisunitwascollectedusingaseparatereturnlineintowhichamembranefiltrationhousingcouldbeinterjectedtoallowsamplingfor
passagedcysts.
Totesttheotherinlinecontinuousdevices,thesecondloopwasspecificallydesignedtoallowtheultravioletlight,activatedcarbonandthefaucetfiltertobeinstalled
atdifferenttimes.Beforeanydevicewastestedoneitherloop,thewaterwasfiltereddowntoremoveallparticlesgreaterthanoneminsize.Thiswasdoneby
recyclingthewaterthroughtheloopandfilteringthewaterthrough50,20and10mspunOrlonfiltersfollowedbymembranefiltrationat2and1m.Continuous
runningoftheloopwithnoplugging(i.e.,pressuredifferentialbuildup)wastakentoindicatethatthewaterwasparticlefree.Problemsdidarisefromsomesteel
fixturesgeneratingrustwithinthesystembutthiswascontrolledbytheapplicationofpolyurethanevarnishtoallexposedsurfaces.
Variousunitswereinstalledatdifferenttimesonthissecondloop.Thefirstunittestedwasanultraviolettreatmentsystem(UltravioletTechnologyInc.,SanMarcos,
California,U.S.A.).Thedischargedwaterwaseithercollectedseparatelyorpassedoncontinuouslythroughtherecyclinghead.Twooptionalfilterswereinstalledat
theconfluencepointforthereturnofthewatertothereservoir.Onefilterwasusedfortheevaluationofgranulatedactivatedcharcoalfilterswhilethesecondunitwas
usedasageneralparticulatefilterwhenthesystemneededtobescrubbed.Pressuregaugeswereinstalledatpointsaroundthecycliccircuitinordertoassureno
significantpressurelossesthroughoutthesystem.Theflowratethrougheachlinewasmeasuredusingthedrumtestmethod(usinga1Lgraduatedcylinder),thefilling
timemeasuredwithastopwatch.
Alowflow(0.2to3.5mL/min.)peristalticpumpwasusedtodirectlyinoculatecystsuspensionsintothereservoirwhichwasconstantlystirredusingarotarypropeller
inordertoensureanevendistributionofthecystswithoutanysettlingofcyststothebottomofthereservoir.
AftertestingoftheU.V.systemwascompleted,theunitwasexchangedinpositionforafaucetfilter.
Todeterminetheeffectivenessofozonation,aportablebatchtreatment(5Lcapacity)countertopozonator(OzonatorSystemsInc.,Mississauga,Ontario,Canada)
wasemployedinwhichthemanufacturer'sspecificationcalledfora15minutepowerupinordertoallowtheozonationgeneratedwithinthedevicetotreatthewater.
Asmallgranulatedcarbonfilterwasemployedinthechanneltothefaucetasafurthertreatmentdevice.Allstudiesconductedinvolvedsamplingfromthemidpointof
thereactionvesseltoavoidtheevaluationofthissecondaryprocess.
SamplingTechniques
Threedifferentsamplingtechniqueswereemployedinthisstudy.Thesewere:(1)cystentrapmentofthetotalflowusingamembranefilter(U.V.)(2)cystentrapment
ofasampleflowbymembranefiltration(reverseosmosis)and(3)directsamplingfromthetreatedwater(ozonator).Otherevaluationsusedmorethanoneofthese
techniques.

Figure1.
Recyclingunitformatfortestingtheabilityofvariouspointofuse
devicestoexcludeviableGiardiamurisfromproductwater.
ALloopABLloopBBPbipassline
FMflowmeterCFchemicalfeedertodispensecystsuspensions
FSfilteringsystemPMpressuremetersPSpressureswitch
PUpumpRTreturnlinetoSTSTsol'nmixingtank
XTXtrolpressuretank

Page109

Inordertoefficientlyusethemembranefilterstoentrapanysuspendedcystsinthewater,therecyclingsystemhadtobeparticulatefree.Prematurepluggingofthe
membranefilterswithotherparticulatematerialreducedtheefficiencyoftherecoverysystemsintwoways:(1)thecystscouldnolongerbeseenamongsttheforeign
particulatematerialpresentonthemembraneand(2)therapidpluggingofthemembranefiltermaysignificantlyreducetheefficiencyofentrapmentallowingsmall
numbersofcyststhroughthetreatmentsystem.
Intheroutineevaluationofallofthepointofusedevices,G.murisratherthanG.lambliacystswereusedasthemarkercells.Thehumaninfectivespecies(G.
lamblia)wasconsideredtopresentahealthrisk,aprobleminobtainingexcystationandunreliableforobtainingaconsistentsupplyofcysts.G.murisontheother
handismoreconvenienttoreproduceinSwissAlbinomiceandthecystsaremoreamenabletoinvitroexcystation(1,7).Cystswereoriginallyobtainedfromthe
collectionattheUniversityofWashington,Seattle,Washington.ThesecystswereusedtoinfectSwissAlbinomicesuppliedbytheAnimalResourcesCenter,
UniversityofSaskatchewan,Saskatoon.ProblemswithcrossinfectionscausederraticcystproductionwhichwascorrectedbytheuseofC3H/HEmicewhich
resistedanytransientlaboratoryinfectionsandalsoremainedinfectedforalongerperiodoftime(13).ThesemicewereobtainedfromCharlesRiver,Montreal,
Quebec,Canada.
CystisolationfromthemousefeceswasbyamodificationofthesucrosegradientmethoddescribedbyDeWalleandJanssonatWashingtonUniversity(3).One
majormodificationofthistechniquewasthedevelopmentofascreencustomizedtofitinthebottomofthemicecages.Themiceweremaintainedona10mmwire
screensothatthefeceswoulddropautomaticallythroughintoa5mmdeepdistilledwatertroughbelow.Thismethoddecreasedcystisolationtimetoaboutonehalf.
Cystswereproducedinaroutinemannerwithproductiongraduallyrisingduringtheprojectfrom100,000to40,000,000cystsperweek.Themousecolonyvariedin
sizefrom10to48membersdependingondemand.Thepurifiedcystswereresuspendedindistilledwaterandstoredat8C(5).
DeterminationofCystViability
Giardiacystsurvivalafterthevarioustreatmentswasmonitoredasthepercentageofcystswhichcouldbeinducedtoundergoinvitroexcystation.Theprocedure
usedwasamodificationoftheprocedureusedattheUniversityofWashington(3).Inthismethod,thecystswereexposedtoHanksbalancedsaltsolutionenriched
withcarbondioxidegeneratedbysodiumbicarbonate.Viabilitywasdeterminedbytheratioofcystswhichexcystedwhenexposedtoatyrodetrypsinsolution,
observedusinga0.5mLcountingchamberoveraninvertedmicroscopeatamagnificationof200.
ResultsandDiscussion
ReverseOsmosisEfficiency
Thereverseosmosisdrinkingwatersystemused18cellulosetriacetatecoatedperforatedplastictubestoprovideasemipermeablefilethroughwhichtheultrafiltered
waterwouldpass.Thesetubeswereinstalledverticallyinaconcentricringwiththetreatedwaterpassingverticallytoacommoncollectionsystematarateofupto
12Lperday.Theremainingwaterreturnedtothedownstreammainwaterflow.Lowvolumesofwaterunderregularhouseholdpressureswouldthereforebeforced
throughtheperforatedtubesandbefedintothe"pure"waterline(productwater).Theadditionalbrinecreatedwasretainedwithinthemajorwaterstockandflushed
throughthesystemwiththenormalutilizationofthewaterfornondrinkingpurposes.
Duringexperimentswiththereverseosmosisunit,themonitoringforcystpassageoccurredinthetotalvolumeoftheproductwater.Thenatureoftheunitinactingas
anultrafiltrationsystemshouldensurethatnocystswouldpassthroughtotheproductwater.Tomonitorthis,thereservoirwaterwasinoculatedwith10cysts/mLof
Giardiamurisandsubjectedtoconstantrecyclingthroughthesystemoperatingat30to80p.s.i.andtheproductwaterwasallpassedthrougha1.2mmembrane
filter.Thesemonitoringfilterswereremovedonaregularbasis.Todeterminewhethercystswerepresent,themembraneswerethoroughlywashedbyagitationona
wristactionshaker(50mLsterileparticlefreewaterfor5minutes)followedbyconcentrationusingcentrifugationat2,500r.p.m.for10minutes.ANeubauer
hemocytometerwasusedtodeterminewhetheranycystswerepresent.Inpractice,thereverseosmosisuniteffectivelyexcludedthecystsbothwithnewandrecoated
tubesforaperiodof80and100daysrespectively,butafterthissporadicreleasesofcystsoccurredinacyclicmanner(Figure2).Theseperiodicreleasesofcysts
suggestthatsomemicrosheeringhadoccurredinthemembranefabrictoasufficientextenttoallowthedirectpassageofthecystsintotheproductwater.Atthe
terminationofeachexperiment,theunitwasdisassembled.Eachtubewasremovedandtheincumbentwaterinsideeachwasexaminedforthepresenceofcysts.
Directmicroscopicexaminationrevealedhighnumbersofcystsinonetubeandlownumbersinsevenothersafterthecompletionofthelastexperimentonrecoated
tubes.Excystationwasperformedonthecystsand30%ofthemwerefoundtobestillviable.ParallelresearchworkbeingconductedconcurrentlybyMorrellin1985
(16)revealedthatG.muriscyststendedtocollectatlowseepage(rollover)pointswithinarecirculatingbiofoulinggravelpack/tubesystem.Asimilarphenomenon
mayhavehappenedduetotheverticaltubepositioninginthetestunitwheretheslowinfiltrationofthewaterthroughthe

Figure2.
Releaseofcyststhroughanewlyinstalled(continuousline)and
arecoated(discontinuousline)reverseosmosisunitwheretimeis
recordedindaysofoperation(xaxis)andthenumberofentrapped
recordedcystspermembranefilterisgiven(yaxis).

Page110

membranecouldleadtothecystsactuallyprecipitatingandconcentratingatthebottomofeachofthefiltertubes.Inthetubereleasingveryhighconcentrationsof
cysts,passageofcystsintotheproductwaterprobablyoccurredmainly,ifnotentirely,througharuptureinthemembranefabricofthistubewithmorecasual
contaminationoccurringthroughsecondarypassagestotheothertubes.ThiserraticpresenceofGiardiacystsinthedischargedproductwatermayalsobe,ineffect,
areflectionofthedegreeofbiofoulingoccurringaroundthetearswithinthecellulosetriacetatemembraneandthedegreeofturbulencecausedbytearsbreakingopen.
Ifenoughcystshavecollectedinthetubewhenanewtearoccurs,sufficientpulsemaybecreatedtoagitatethewatercystsandejectsomefromthetubewiththe
productwater.Assubsequentmicrobialbiofoulingwouldatleastpartiallycloseoffthetear,thenumbersofcystsintheproductwaterwoulddecrease.Wherethe
tearsbecamecompletelysealedoff,nomorespuriousturbulencewouldoccurandthecystswouldagaintendtosettleoutinthetubes.Inallprobability,theefficiency
ofthemembranesatexcludingGiardiaaremuchshorterindurationthantheevidencewouldsuggestintheexpelledproductwatertubes.
Intherepeatstudiesusingrecoatedtubes,productwaterwasfrequentlypositiveformacronidiauponmicroscopicexaminationat200.Thiswouldimplythatamold
wasnowactivelygrowingintheproductwatersystem.
Insummary,thereverseosmosisunitappearedtobeinitiallyeffectiveatexcludingGiardiacystsfromtheproduct.Unfortunately,thereisnosimplefieldtestwhich
couldrapidlydeterminewhethertheproductwaterwasfreefrompassagedcystsorthatthemembranesare,infact,stillintactandfunctioningsatisfactorily.Because
ofthis,restraintsshouldbeplacedupontheeffectivelifespanofthesemembraneswherepromotedfortheeffectiveexclusionofGiardiacystsfromdrinkingwater.
UltravioletLightTreatment
Theinstalledultraviolettreatmentunitgeneratedamaximumof40,000watts/cm2of253.7nmwavelength.Themethodfortestingthissysteminvolvedloadingthe
waterwithdifferentconcentrationsofcystsatthereservoirpointinthesystemandrecyclingthiswatercontinuouslypasttheU.V.unit.Allofthewaterthathadpassed
throughtheunitwasfilteredthrougha1.2mmembranefiltertoentrapthecysts.Thecystconcentrationsusedwere800,8,000and80,000cysts/L.Controlswere
runbyrecyclingthesameconcentrationsofcyststhroughwithaninactivatedU.V.lightsource.Thenumbersofexcystedcystspresentinthecontrolswastakento
representthenulleffectwitha100%viabilityinthepassagedcysts.Todeterminetheefficiencyofthetreatment,thenumbersofviablecystsfoundinthetreated
sampleswereadjustedaccordingly.
WhiletheU.V.lighttreatmentdidfunctiononrecyclingwater,thecystspassagedthroughtheunitweresubjectedtoonlyasingleexposureandthenfilteredout.Inno
experimentswerethecystssubjectedtophasedorcontinuousreexposure.Fromtheseexperiments,itwouldappearthattheU.V.lightunittestedwascompletely
effectiveincontrolling800cysts/Linthewaterbutthatefficiencydeclinedasthecystloadingsincreased,particularlybeyond8,000cysts/L.Unfortunately,itwasnot
possibletodeterminewhetherthehighernumbersofcystscreatedashieldingeffectwhichincreasedviability.Alternateparticulateloadingsusingpassivematerials
couldnotbeconductedsincethiswouldseriouslyincreasetherateofpluggingattheinterdictingfilters.Theloadingsandretentiontimesintheseexperimentswere
belowthosenormallyfoundinwatersystemsduetodecreasedflowratesgeneratedbythegradualpluggingofthemonitoringmembranefilterwithcysts.
At800cysts/L,theinputwaterwasalmosttotallyfreeofparticulateswithlittleshieldingeffect,sothatwithanadequateretentiontime,itwassufficienttocausethe
totalinactivationoftheGiardiacysts.Sincewaterwouldnormallybeexpectedtocontainfarlessthat800cysts/L,itisprobablethatwheretheparticulatesarelow
andretentiontimeisadequate,thisunitcouldbeusedtocontrolGiardiacystspotentiallypresentinadrinkingwatersupplyusedfordomesticpurposes.
GranularActivatedCarbonFiltrationSystem
StandardcartridgetypefilterswereobtainedfromWaterConditioningCanadaLtd.,Regina,Saskatchewan,Canada.Immediatelyafterinstallationofthefirstfilter
andduringthefirstthreedaysoftesting,somecystswerelocatedintheproductwater.Uponprolongedtestingovera21dayperiod,nocystswererecoveredfrom
thewaterwithtwoexceptionswhentwoandfourcystswererecovered.ThisabsenceofGiardiawaspostulatedtobeduetoretentionandpossibleeliminationofthe
cystsfromthewaterthroughthepresenceofanactivelygrowingbiofilmonthecarbonparticulatesurfaces.Toexaminethishypothesis,whenthenextnewfilterwas
installed,amoredetailedbacterialtestingwasconductedinconjunctionwiththeevaluationforthecystpassage.Thisnewfilterexcludedcyststhroughuntilday13
whenlargeclumpsofcystsweresporadicallyreleasedintothedischargedtreatedwater.Veryhighbacterialnumbersoccurredintheeffluentwatertoday9and
subsequentlydeclined.Atthetimewhencystclumpingwasoccurring,thebacterialnumbershadreducedfromahighof640,000c.f.u./mLdowntoalowpointof
320c.f.u./mL.Thisreductioninthedetached(planktonic)bacterialpopulationinthewaterappearedtocoincidewiththeoccurrencesofclumpedcystsintheeffluent
water.Thesesporadicappearancesofcystsintheproductwateratdifferentpopulationdensitiesshowedsignsofcystwalldestructionandlysis.Theseinteractive
effectswarrantadetailedstudyinordertocatagorizethedynamicsofthispotentiallyimportantreactiveinterface.Granulatedactivatedcharcoalfiltersystemswould
fromthislimitedstudyappeartoallowthepassageofGiardiacystseitherdirectlyorthroughinteractionwiththebacterialbiofilmcoatingthecarbongranulated
particles.

Page111

Ozone
TheefficiencyofozonetoinactivateGiardiamuriscystswastestedusinga5Lbenchtopozonator(OzonatorSystemInc.,Mississauga,Ontario).Thisdevice
operatesonabatchbybatchbasisinwhich5Laliquotsofwateraresubjectedtoanozonedischargeonatimedbasis(recommendedtimeisnormally15minutes).
Themanufacturersclaimisthatthisexposureprovidesa"completedisinfectionofthedrinkingwater"renderingitsafetodrink.Theseexperimentsthereforeexamined
therelationshipsbetweenthetimerequiredforthecontrolofGiardiacystsandthetimerequiredtocontrolbacteria.Toachievethis,exposure/inactivationtestswere
conductedonthebacteriaPseudamonasflourescensandEscherichiacoli,bothisolatedandidentifiedtospeciesattheReginaWaterResearchInstitute,aswellas
onG.muris.Forthesetrials,steriledistilledwaterwasusedforthebacterialtrialsandregulartapwaterfortheG.murisstudiesinordertoduplicatemorenormal
conditionsthatwouldbeexpectedinregularuse.
ParallelstudiesdoneonP.fluorescensandE.colirevealedamuchfasterrateofkillforthelatterorganism.Theinitialpopulationof56,000c.f.u.E.coli/mLwas
reducedto0in60seconds(seeFigure3).ForP.fluorescens,prolongedsurvivalofozonationoccurredwhen60,000c.f.u.bacteria/mLwereintroducedwith
survivalstillbeingrecordedafterfifteenminutesofexposure.
ViabilitytestsconductedonGiardiaincludedthedirectadditionof5,000,000cyststo5Loftapwaterintheozonatortank.Aftersupplementation,theozonatorwas
turnedonandmidpointsamplesweretakenat0,1,2.5,5,10,and15minutes.Thesetestswererunat8C,22C,and37C.At8Cand37Cthecystswere
inactivatedwithinoneminute.At22C,10%ofthecystsinthesampletakenat1minuteremainedviable.Asaresultofthis,theexperimentalrunsat22Cwere
repeatedandmidpointsamplesdrawnat0,5,10,20,30,60,70,80,and90seconds(Figure3).Ozonationfor90secondsrenderedallofthecystsunabletoexcyst
withaparalleledreductioninviability.
FaucetFilter
Simplefiltrationdevicesdirectlyattachedtothefaucetarebeingwidelypromotedasameanstoensurea"safe"potablewatersupply.Thesystemchosenfortesting
wasthePureWater"99"(AssociatedMillsInc.,Chicago,Illinois).Thefilterwasattachedtoafaucetspecificallyinstalledonthesecondrecyclingloop.Giardia
muriscystswereinoculatedintotherecyclingwaterwithafinalcellconcentrationof10cysts/mL.Waterwaspumpedthroughthedevicecontinuouslyandthe
passageofcystsdeterminedusinganinterdictingmembranefilter(1.2mporesize).For3days,membraneswereremovedandexaminedforcystsat24hour
intervals.After3daystheinterdictionintervaltimewaschangedto7daystoevaluatetheeffectofprolongedstoragebetweenuse.Cystswerewashedoffthefilter,
concentratedbycentrifugationandcountedusingahemocytometer.

Figure3.
Thepercentageofresidualviableorganisms(yaxis)whenexposedto
ozonationinabatchpointofusedeviceforperiodsofupto20
minutes(xaxis)forthethreetestorganisms:Escherichiacoli(thin
continuousline),Giardiamuris(thickcontinuousline),
andPseudomonasfluorescens(discontinuousline).

Cystswererecoveredfromthefilteronday1butbyday3nocystpassagewasevident,indicatingthatthefilterhadretainedthecysts.However,afterthefilterswere
examinedagainatday10,25cystswereentrappedintheproductwater.Afterafurther7daysoftesting,thenumberofcystsrecoveredincreasedto40.Thefaucet
filterwasthereforineffectiveattotallyexcludingcystsfromthepostdiluvianwater.Concurrentbacteriologicalexaminationsoftheproduct(premembranefiltration)
waterperformeddailyrevealedarapidriseinthebacterialpopulationtoaplateauof200,000c.f.u./mLonday5.Inthisdeviceitcouldbethatacomplexinteraction
wasoccurringbetweenthebiofoulingbacteriawithinthefilterandthetransientcystsentrappedwithinthedevice.
GeneralDiscussion
AllofthedevicesfailedtoeitherexcludeordestroytheGiardiamuriscystsundersomecircumstances.Avarietyofpotentialinteractivefactorswereobservedwhich
couldeffectthisincludinghighcystloadings,biofoulingofthesystem,screeningeffectsfromothersuspendedparticlesinthewater(beitothercysts,bacteria,
chemicalprecipitatesetc.),temperature,mechanicaldamageorpluggingwithinthedeviceandthemethodforobservingcystnumbers(i.e.,recoveryefficiency).Most
ofthestudiesinvolvedrelativelyshortintensiveexposurescenarioswhichwoulddifferfromtheactualpracticesinvolvedinanormalinstallationonawatersystem.
AnyrecommendationsfortheabilityofagivendevicetosuccessfullyandcontinuouslyremoveordestroyGiardiacystsfromapotablewatersupplyhastobe
temperedbythefactthatundersomeconditionsthemechanismscanfailandnosystemexistswhichwilladequatelymonitorfortheseoccurrencesinaeconomicaland
efficient

Page112

manner.Furtherresearchisneededinordertomorepreciselyidentifytherelationshipsofthevariousparticularinteractivefactorstotheefficiencyofexclusionof
viableGiardiacystsfromagiventreatmentsystem.
Acknowledgements
TheauthorswishtothanktheCanadianFederalDepartmentofHealthandWelfarefortheprovisionofaresearchcontractwithinwhichthisresearchwasperformed.
Inparticular,wewishtothankDr.RichardTobinoftheDepartmentforhissupportandcriticalencouragementastheprojectdeveloped,ErlJanssonforhisexpertise
andadviceastheprojectdevelopedandMehdiKaramchiwhoassistedinthemaintenanceoftheculturesanddevelopedalternativemethodsforthestainingofthe
cystswhichwillbereportedelsewhere.
LiteratureCited
1.Bingham,A.R.,andE.A.Meyer.1979.Giardiaexcystationcanbeinducedinvitroinacidicsolutions.Nature(London)277:301302.
2.Craun,G.F..1978.Waterborneoutbreaksofgiardiasis.In:WaterbornetransmissionofGiardiasis.Jakubowski,W.andJ.C.Hoff,(eds.).EPA600/979
001.
3.DeWalle,F.B.,Jansson,E.,andD.A.Carlson.1983.InactivationofGiardiabychlorine,U.V.andozone.DepartmentofEnvironmentalHealth,Universityof
Washington,Seattle,Washington.
4.Jarroll,E.L.,Bingham,A.K.,andE.A.Meyer.1980.Giardiacystdestruction:effectivenessofsixsmallquantitywaterdisinfectionmethods.Am.J.Trop.Med.
Hyg.29:811.
5.Jarroll,E.L.,Bingham,A.K.,andE.A.Meyer.1981.EffectofchlorineonGiardialambliacystviability.Appl.andEnviron.Microbiol.41:483.
6.Lippy,E.C..1978.Watersupplyproblemsassociatedwithawaterborneoutbreakofgiardiasis.In:WaterborneTransmissionofGiardiasis.Jakubowski,W.
andJ.C.Hoff,(eds.).EPA600/979001.
7.Meyer,E.A.,Erlandsen,S.L.,andW.S.Radulescu.1984.Animalmodelsforgiardiasis.In:GiardiaandGiardiasis.Erlandsen,S.L.,andE.A.Meyer,(eds.).
PlenumPress.NewYork.
8.Moore,G.T.,Cross,W.M.,McGuire,D.,Mollohan,C.S.,Gleason,N.N.,Healy,G.R.,andL.H.Newton.1969.Epidemicgiardiasisataskiresort.New
EnglandJ.Med.281(8):402.
9.Morrell,R..1985.ImpedenceeffectsofbiofilmformationonthepassageofGiardiamuriscysts.SubmittedB.Sc.Honoursthesis,BiologyDepartment,University
ofRegina.
10.Pluntze,J.C..1983.ThesignificanceofgiardiasisonwaterqualitystandardsandwaterutilitypracticeinWashingtonstate.PresentedattheBritishColumbia
WaterandWastewaterAssociation,November22.
11.Rice,E.W.,andJ.C.Hoff.1981.InactivationofGiardialambliacystsbyultravioletirradiation.Appl.andEnviron.Microbiol.42(3):546.
12.Rice,E.W.,Hoff,J.C.,andF.W.Schaefer,III.1982.InactivationofGiardiacystsbychlorine.Appl.andEnviron.Microbiol.43:250251.
13.RobertsThompson,I.C.,andG.C.Mitchel.1978.Giardiasisinmice.I.Prolongedinfectionsincertainmousestrainsandhypothymic(nude)mice.
Gastroenterology75:4246.
14.Shaw,P.K.,Brodsky,R.E.,Lyman,D.O.,Wood,B.T.,Hibler,C.P.,Healy,G.R.,MacLeod,K.I.,Stahl,W.,andN.G.Schultz.1977.Acommunitywide
outbreakofgiardiasiswithdocumentedtransmissionbymunicipalwater.AnnalsInt.Med.87(4):426.
15.Shun,D.L..1985.Giardialambliaandwatersupply.JournalAWWAFeb.:4047.
16.Visvesvara,G.S..1983.Giardiasis:anoverview.IllinoisMed.J.July:3439.

Page113

DiatomiteFiltration:WhyItRemovesGiardiafromWater
HarrisG.Walton
ManvilleResearchandDevelopmentCenter,2500MiguletoCanyonRd.,Lompoc,California93436,U.S.A..
Inthispresentation,theauthorutilizesscanningelectronmicroscopy(S.E.M.)todocumentthepositivemechanicalremovalofthecystsofGiardialamblia.Dataarealsopresentedto
compareelectronicandlaserparticlesizedistributionsofthecystswithmercuryintrusionporosimetryofatypicaldiatomitefiltercake.Finally,atheoreticalporeofasandfilteris
showninjuxtapositionwiththesamemagnificationofcystsanddiatomite.Thesemagnifiedandgraphedvisualcomparisonsareintendedtodemonstratetheefficacyofdiatomite
waterfiltrationandshouldprovidemeaningfulreferencesforfutureresearchers.

Introduction
Becausediatomitefiltrationusuallytakesplaceinanenclosedvessel,andthemechanismsofparticleretentionaremicroscopic,itisnotpossibletoobservetheactual
separationofGiardiacystsfromawaterstreamwhileitishappening.Thispaperpresentsmeasuredandpictorialdatathatenablesthereadertovisualizetheactual
physicalentrapmentofthesecystsbydiatomaceousearthasitwouldoccurinapotablewaterfilter
Discussion
Figure1isascanningelectronmicrograph(SEM)ofGiardialambliacystsat2000magnification.Unfortunately,thesecystspartiallycollapsedduringsample

Figure1.
ScanningelectronmicrographofGiardiacysts,2000x.

preparation,butthishasnotdramaticallychangedtheirphysicaldimensions.Figure2isaplotofcystparticlesizemadeonaCilasmodel715granulometer.This
instrumentmeasuresparticlesizedistributionbylightscatterofalaserbeam.Thecystswerepreservedinformalinjust20minutespriortomeasurement,sotheir
morphologyisunchangedfromtheviablecyst.Thedifferentialcurveshowsthemtobemonodispersedbetween6and15mwithamediansizeof8m.

Figure2.
Giardiacystparticlesizedistributionmeasuredby
lightscatterwithaCilasgranulometer.

Figure3.
Giardiacystparticlesizedistributionas
measuredby(Coulter)electrolyticdisplacement.

Page114

CystsfromthesamesamplewerethenmeasuredonaCoultercountermodelTawhichmeasuresparticlesizeelectronicallybythedisplacementofelectrolyteinan
electricfield.Figure3showsagainthatthecystsaremonodispersedbetween5and10m.AnotherobservationevidentinFigures2and3isthetotallackof
materialbelowthesizerangeofthecysts.Thisiscreditedtothetechniqueofseparating"superclean"specimensatColoradoStateUniversity,whowerethesuppliers
ofthecystsinthisstudy.
TheprinciplesofdiatomitefiltrationasshowninFigure4arewelldocumentedintheliterature(1,2,6,7)andarebeyondthescopeofthispaper.Thisisshownhereto
orientthereadertotherelationshipoftheprecoatandthefiltercake.
Figure5isascanningelectronmicrograph(SEM)ofasectionofanactualdiatomiteprecoatmagnified2000.Thegradeshownhere(CeliteT M545)isatthecoarse
endofarangeof11gradesthatareavailableforwaterfiltration.Thismaterialhasapermeabilityof4.8darcyswhilethefinergradesrangedownastightas0.06
darcys(1).Notethattheintersticialspaceshaveporesrangingfromseveralmthrough50m.Notealsothatindividualdiatomshaveporeswithintheirownstructure
thataresubmicronindiameter(1).TheentirefieldofdiatomsshowninthisSEMisonly45macross.Actualprecoatsofdiatomiteusedinwaterfiltersareabout6
mmor600mthick.Whenoneenvisionsthatwaterflowingthroughtheprecoatlayeralone,mustpassthroughamazeapproximately12,000timesthickerthanthat
displayedinFigure5,itcanbereadilyrealizedthatacystparticleof8mwillrapidlyberetainedbymechanicalentrapmentinthediatomite.

Figure4.
Schematicofdiatomitefiltration.

Figure5.
Scanningelectronmicrographofa
diatomiteprecoat(CeliteT M545),2000.

Theporesizedistribution,asmeasuredbymercuryintrusionporosimetry,isshowninFigure6.Whilethisshowsthat85%oftheporesaregreaterthan8m,their
randomdistributionandinterspersionwiththefinerporesproducesafiltermediawithretentionpotentialofevensubmicronparticles.
InFigure7,therelationshipofthecyststoaprecoatsurfaceareshowninasidebysidecomparisonat2000.Itisagainnotedthatsomeporesformedbetween
diatomiteparticleswillallowthepenetrationofacystintothesurface,butafewmdeeperintotheprecoat,thecystwillberemovedbyalabyrinthoffinerpores.
Ithasbeendocumented(3,4,5)thatdiatomaceousearthwaterfilters,usingthissamecoarsegradeofdiatomite,effectivelyremove99.9+%ofGiardiacystsfrom
watercontainingupto3.36104cystsperliter.ThereasonforthisisdemonstratedinFigure8,whichisamix

Figure6.
PoresizedistributionofthediatomiteprecoatCeliteT M545.

Page115

Figure7.
TherelationshipofGiardiacyststoa
TM
Celite 545precoat,2000.

Figure8.
AGiardiacystprecoatfiltercake,2000.

tureofcystsanddiatomitedisplayedat2000.Thiscouldbetakenasatheoreticalrepresentationofafiltercakehowever,inthe"realworld"theparticleswouldbe
moretightlycompactedtogether.Thefrictionaldragofthewaterpassingthroughandaroundtheparticlescausesadifferentialpressurewhichcausesthecompressible
cysttodeformandpartiallyextrudeintothediatomitematrix.Thecystispreventedfromextrudingthroughthefilter,againbythemultilayereffectofthefinepore
labyrinth.Tofurtherguaranteethiseffect,waterfiltersforGiardiaremovalarerecommendedtousetwicethethicknessofanormalindustrialprecoat(3,7),i.e.20
lbs/100ft2insteadof10lbs/100ft2aswouldnormallybeapplied.
Theliterature(4)reportsthatmanysurfacewatersourcesinmountainshaverawwaterturbiditieswellbelowtherequired1.0NTUlimit.Forthisreason,andbecause
thelowtemperatureofthewatermakesflocculationdifficult,somesandfiltersareoperatedwithouttheuseofalum.Inordertoputthisinperspectivewithadiatomite
filter,whichusesnoflocculatingagents,thefollowingcomparisonismadewithatheoreticalporeformedbyidealizedsandgrainshavingameandiameterof0.5mm.
Inordertogainthebestperspective,thisporeisdemonstratedintwosteps.Figure9istheporeimagedat200showinghowthe0.5mmdiametersformaminimum
contacttrianglehavingthreeequalsidesof130m.
InFigure10,theporeisnowimagedat2000withGiardiacystssuperimposedatthesamemagnification.Fromthisillustration,itcanbeunderstoodwhyrandom
entrapment,eddycurrentandsedimentationareprobablytheonlyfiltrationfunctionsthatwillremovecystsinthisfilterifitisoperatedwithoutaschmutzdeckelayer,
andpassageofsomecystsintothefinishedwaterisalmostacertainty.

Figure9.
Theoreticalintersticeof0.5mmsandgrain.

Page116

Figure10.
Giardiacystssuperimposedontheoretical
intersticeof0.5mmsandgrain,2000.

Inthisfinalcomparison(Figure11),theSEMofthediatomitesurfaceisshowntothesamemagnificationasthesandpore.Oncemoreitisdramaticallydemonstrated
whyadiatomitefilterremovesfineturbidityandGiardiacystssocompletely.
Conclusion
Byvisualcomparisonsandmeasuredporeversusparticlesizes,ithasbeenvisuallydemonstratedthataproperlyoperateddiatomitefilterwillprovideanefficient
barrierfortheremovalofGiardiacystsfrompotablewater.
Acknowledgments
TheauthorgratefullyacknowledgestheassistancerenderedbyKathySmith,ElectronMicroscopy,andHubertAttaway,EnzymeandMicrobeImmobilization
SectionsofManvilleR&D.

Figure11.
Diatomitesurfacessuperimposedontheoretical
intersticeof0.5mmsandgrain,2000.

LiteratureCited
1.Cain,C.W.,Jr..1984.Filteraid,useinfiltration.EncyclopediaofChemicalProcessingandDesign21:348372.
2.Cummins,A.B..1942.Clarifyingefficiencyofdiatomaceousfilteraids.IndustrialandEngineeringChemistry34:403414.
3.Hendricks,D.W.,etal..1982.RemovalofGiardialambliafromwatersupplies.EnvironmentalEngineeringTechnicalReportNo.5836824.
4.Hendricks,D.W.,etal.1984.FiltrationofGiardiacystsandothersubstances.Volume1:Diatomaceousearthfiltration.E.P.A.projectsummaryNo.EPA
660/S284114.
5.Logsdon,G.S.,etal.1983.ControlofGiardiacystsbyfiltration:Thelaboratory'srole.A.W.W.A.waterqualityconference,Norfolk,Virginia.
6.Purchase,D.B.,1967.IndustrialFiltrationofLiquids.LeonardHillbooks,London,England.pp.9098.
7.Svarovsky,L.,etal.1977.SolidliquidSeparation.ButterworthPublishing,LondonBoston.pp.196198.

Page117

SmallWaterSystemImprovementsForGiardiaRemoval
ACaseStudy
MichaelR.Alberi,StevenJ.Quail*,andRobertA.Kruse
WoodwardandCurranInc.,41HutchinsDr.,Portland,Maine04102,U.S.A..
Acasestudyofatreatmentsysteminstalledtocorrectaknown,persistentGiardiaprobleminthewatersupplyofasouthcentralresorttownofpopulation2000ispresented.The
resultsoffieldpilotstudiesformingthebasisforsystemconfigurationandselectionofdesigncriteriaforamodifieddirectfiltrationprocessarepresented.Summariesofdaily
operatingdataarepresentedforthepastthreeyearsshowingplantperformanceduringvariousrawwaterqualityconditions.Planteffluentturbidityhasaveraged0.054NTU
duringthisperiodwithalumandanionicpolymerdosagesaveraging2.3mg/Land0.48mg/Lrespectivelyinthewinterand11.2mg/Land0.56mg/Lrespectivelyinthesummer.
Thewinterandsummeroperatingconditionsrepresentingtwoseparatecoagulationregimesadsorbtiondestabilizationduringthesummer,andcombinationsweepfloc
adsorbtioninthewinterarepresentedonpHvslogalumdosediagrams.DataonGiardiacystcountsontherawwaterandfinishedwaterarepresentedwhichconfirmtheir
continuedpresenceinthesourceandabsenceinthefinishedwater.Operatingstrategies,andoperatingcostsforlabor,energyandchemicalsarealsopresented.

Introduction
Inthespringandsummerof1980,anoutbreakofgiardiasisoccurredinthecityofRedLodge,Montana,asmallresortcommunitylocatedatthebaseoftheBear
toothRangeoftheRockiesinsouthcentralMontana.Overthefollowingyear,about860casesofgiardiasiswereconfirmedbylocalhealthprofessionals.
AnonsiteinvestigationoftheoutbreakbytheEmergencyResponseTeamoftheU.S.EPAandtheCenterforDiseaseControlledtoimplicatingtheantiquatedwater
systemanduntreatedsurfacewatersourcesasthecauseoftheproblemalthoughsamplingoftherawwaterdidnotidentifycysts.
Therateofconfirmationofnewcasescorrelatedwellwithspikesontheturbidityofthemountainstreamservingasthesourceofsupply.Figure1showsthatthe
confirmationofnewcaseslaggedeachriseinturbiditybythenormallyexpected6to22daysincubationperiod.
Theexistingwatersystemconsistedofagravitysupply,coarsescreenedsurfacewaterintakewithahydraulicelevationabout300feetabovethecity,andaconstant
ratechlorinatorfeedingwaterdirectlyintothedistributionsystemthroughtwomilesoftransmissionline.Neitherdistributionsystemstoragenorindividualservicewater
metersexisted.Thesecircumstancesresultedingrosslyexcessivepercapitausehabits,highpeakingfactors,widelyvaryingchlorineresidualsinthesystem,excessive
servicepressuresinthelowerendofthesingleservicezone,andtasteandodorproblemsinthefallcausedbytheactionofchlorineonleavestakenintothesystem
anddepositedinthepipes.Thissystemlackedthenecessarybarriersagainstwaterborne,diseasecausingagents.
Remedialactionstakenforaninterimperioduntilimplementationofpermanentcomprehensivecorrectivemeasurescouldbeundertakenincludedtheadditionofflow
pacedchlorinationontheintakeandtheimpositionofaboilorderforallenterprisesservicingthepublic.
Oneofthechallengeswastoimplementaneconomicallysizedwatersystemtoservepresentwaterusersandareasonableamountofgrowth.Sincethesystemwas
currentlyunmeteredandcustomerscoulduseallthewatertheywanted,thecityhadtocriticallyreviewhowthesystemwasmanagedandfinanced.
Amasterplanwasquicklypreparedasabasisfordesigning,constructing,andfinancingsystemimprovementsthatwouldprovidetreatedwaterfreefromGiardia
cysts,reducedistributionsystempressurestoreasonable

Figure1.
Correlationofspikesinrawwaterturbiditywith
developmentofconfirmedcasesofgiardiasis.
*Correspondingauthor.

Page118

values,providedistributionsystemstorage,andprovideservicemetersasanequitablemeansofapportioningthecostofsystemconstructionandoperation.These
objectiveswereaccomplishedbyaddinganewwatertreatmentplant,changingthedistributionsystemto2pressurezoneseachregulatedbyconcretestorage
reservoirswithapressurereducingvalve(PRV)stationseparatingthetwo.Watermeterswereinstalledonallservices,andundersizeddistributionmainsinthelower
pressurezonewereupgraded.
PermanentfinancingforthesystemimprovementswasarrangedthroughgrantsandloansfromtheFarmHomeAdministration,CommunityDevelopmentBlockGrant
programs,andinterimfinancingthroughthesaleofbondanticipationnotes.
Theseimprovementswereimplementedintwophases.Inthefirstphasethedistributionsystemmodificationsweremadetoseparatethe2pressurezones,anda
lowerzone0.75MG(milliongallon)storagetankwasconstructedandconnectedtoanexistingshallow1MGDwelllocatedinthelowerdistributionzone.Thiswell
hadbeendevelopedintheearly1960'stoaugmentlowpressureinthecentralbusinessdistrictduringfireflowconditions.Theupperzonecreatedbythisseparation,
whichatthattimeservedlessthan50residentialusers,continuedtobeservedfromthesurfacewatersourcewithoutfiltrationuntilthetreatmentfacilitycouldbeput
intooperation.Duringtheconstructionofthefirstphaseimprovements,pilotstudieswereconductedtodevelopdesigncriteriaforthetreatmentplant.
Thedecisiontoimplementasurfacewatertreatmentplantinsteadofdevelopingwellsasthesolesourceofsupplywasbasedonthefollowingconsiderations:1)The
cityhadthefirstwaterrightonthisstreamwhichwouldhavebeenrelinquishedwithoutcompensationifitsusewasdiscontinued,2)Thesourcefeedstheserviceby
gravity,3)Thegroundwateraquiferavailableforwellsisshallowanddifficulttoprotectduetothepresenceofexistingdevelopmentwithonsitesewagedisposal
systems,and4)Thesurfacesourcewaterhasalowturbiditywhichcouldbetreatedeconomicallybydirectfiltration.
TreatmentPlantDesignCriteriaDevelopment
AreviewoftheliteratureatthetimethatthepilotstudywasconductedrevealedthatthetreatmentobjectivesforGiardiacystremovalbyrapidsandfiltrationshould
provideafinishedwaterturbidityof0.10NTUorlessatalltimesandprovide2hoursofcontacttimeforchlorinedisinfection(1).
A0.5gal/minpilotplantwithamixedmediafilterobtainedfromNeptuneMicrofocInc.wassetupatthestreamintake.Thepilotplantoperatedoveran8month
periodfromNovember1981toJuly1982toevaluatevariationsinstreamconditions.TheaveragerawwaterconditionsencounteredarepresentedinTable1.
JartestswereconductedusingthetraditionalPhippsandBirdStirrerprotocoltoestablishchemicaldosagesforthepilotunit,whichwasoperatedforabriefinitial
periodinacompletetreatmentmodeandthenasadirectfiltrationunit.Thejartestswereusedasaqualitativemeasureofflocdevelopmentbyaspecifiedamountand
typeof
TABLE1.Sourcecharacteristics
Turbidity(NTU)

0.22.5

Temperature(F)

3243

Alkalinity(mg/LCaCO3)
pH

30
77.3

conditioner,coagulantandflocculantaidadded,theirsequenceofaddition,andtheapparentstrengthofthefloc.Thiswasdeterminedvisuallyberesuspendingthe
settledflocbyrapidmixingfor15seconds.Ifthefloccompletelyshearedorifnosignificantflocredevelopedafterrapidmix,theflocwasconsideredfragileand
unsuitablefordirectapplicationtothefilter.Thisprocedurerevealedtwoproblemswiththepilotplanthardwarethatwerecorrectedbyphysicalmodificationsshown
inFigure2.Jartestingrevealedthatthecoldlowturbiditywaterwasbestcoagulatedbyaddingthechemicalssequentiallyto1)provideanucleationsiteforfloc
development,2)provideabuffertomaintainaconstantpH,3)providethecoagulant,aluminthiscase,and4)addanionicpolymerasaflocculantaid.Thiswas
accommodatedbyconstructinga4cellseriesrapidmixunitfortheadditionofthefourreagents.Theearlyfilterrunsusingthepaddleflocculatorwithandwithoutthe
sedimentationunitwereunabletoproduceafilteredwaterwithaturbiditylessthanabout0.3NTU.Sincethefilterwasfedwithacentrifugalpumplocatedbetween
theflocculationunitandthefilter,itwassurmisedfromthejartestflocstrengthteststhattheflocsweredestroyedbythepump.Thiswasovercomebyeliminatingthe
paddleflocculationandsedimentationunits,andpumpingdirectlyfromtherapidmixtrainthrougha1inchcoiledpolyethylenetubetothefiltersoastoachievea
velocitygradientof30sec1andadetentiontimeof10minutes.Thisarrangementprovedtobeanadequateflocculator.Thesetwomodificationsprovidedthecontrol
andflexibilityrequiredforthetreatmenttraintoproduceafilteredwateroflessthan0.1NTU.
Duringthefiltertrials,twooperatingregions,asdefinedonthelogalumpHdiagrampreviouslypublishedbyAmirtharajah(2),wereevaluated.Thesweepflocregion
forthiswatercoverstheregionboundedbypH6.8

Figure2.
Pilotplantschematic.

Page119

to7.3andalumdose12mg/Lto25mg/L.Operatinginthisregionwouldresultinrelativelyshortfilterrunsofabout8hoursandthuswouldrequiresedimentation
priortothefilter.TheprocessalsooperatedverywellintheabsorptiondestabilizationregioncharacterizedbytheregionboundedbypH6.8to7.3andalumdose12
mg/Lto7.0mg/L.PlotsofturbidityremovalandheadlossfortypicalpilotfilterrunsarepresentedinFigure3whichshowtheneedforfilteringtowasteafter
backwash,orafterfiltershutdownandrestart.
Thefollowingdesignrecommendationsresultedfromthepilotstudy:
1.Chemicaladditionduringrapidmixinthefollowingorder:
(i)bentonitetoincreasetherawwaterturbidityto1.0NTU
(ii)sodaashtomaintainapHrangeofpH6.8to7.2
(iii)alumat12mg/Lforadsorbtiondestabilization,25mg/Lforcombinedsweepcoagulantoperatingregimes
(iv)anionicpolymerasaflocculantaidat0.5to1.0mg/L.
2.RapidmixGtof33,000.
3.FlocculationGof25to35sec1.
4.FlocculationHRTof10minutes.
5.Mixedmediafilters,constantrate,withsurfacewash.
6.Continuousturbiditymonitoringforprocesscontrolandtoinitiateandstopfiltertowaste.
7.DonotreclaimbackwashwaterduetoconcernaboutGiardiacystcontamination.
Thedesignofthefullscaleplantwasbasedonoperatingintheadsorbtiondestabilizationregionbecauseitismoreeconomicalthanoperationinthesweepregion.It
wasalsodesiredthattheplantbecapableofoperatinginthecombination(sweepandadsorbtion)regioniffuturerawwaterconditionswarranted.Thiswould
necessitateasedimentationdeviceinadditiontoaflocculator.Duringequipmentselection,acontactflocculatorwasselectedas

Figure3.
Watertreatmentplantgeneralarrangement.
TABLE2.Watertreatmentplantdesigncriteria
DESIGNCRITERIA

1)

Plantdesigncapacity

1.4MGD

2)

Finishedwaterturbidity

0.1NTU

3)

Freechlorineresidual,insystemextremes

PLANTSYSTEMDESCRIPTION

1)

Plantinfluent3lowservicepump

Verticalturbinetype

LSP1

1000GPM@35FTTDH

LSP2

1000GPM@35FTTDH

LSP3

550GPM@35FTTDH

2)

Chemicalfeed

3dryfeedunits

1dryfeedunit,standby

1liquidfeedunit

1liquidfeedunit,standby

FeedratesBentonite

Sodaash

6mg/L

Alum

12mg/L

Anionicpolymer

1.0mg/L

3)

Rapidmix

4chamberstaticmixer,12''

4)

Filtration

Filterloadingrate,designmedia

Contactflocculator

140SF

Filter

280SF

Washwatertroughs,perfilter

Filterbackwash

Rate,design

240HPvert.turb.pumps

Filtermediasurfacewash

Rate

90GPM@230FT

2HPsubm.turb.pumps

90GPM@230FT

Flocculatorbackwash

Rate

Flocculatorairscour

Rate

2positivedisp.blowers

5)

Disinfection

2chlorinators

Rate,max.ea.

20lbs/day

6)

ClearwellCapacity

137,500gal

7)

ReservoirCapacity

253,000gal

8)

Emergencygenerator

0.5mg/L

1.0mg/L

3.6gal/min/ft2

15gal/min/ft2
2100GPM@42FTTDH

7.2gal/min/ft2

2SCFM/SF
140SCFM@9PSIG

80KW

acompactdevicethatcouldperformbothfunctions.ThefullscaleplantdesigncriteriaispresentedinTable2.ThegeneralarrangementoftheplantisshowninFigure
3.Theplantisfullyautomatedandiscontrolledbyaprogrammablecontroller.Plantstartandstopfunctionsareinitiatedbylowandhighlevelsignalsrespectively
fromtheupperservicezonestoragereservoirbuildintegraltothefilterplant.
PlantPerformance
TheplantstartedupinMarch1984,andhasperformedwellwithoutmechanicalproblems.Plantwaterproductionintermsofasevendaymovingaveragepresented
inFigure4showsthedecreaseindailyproductionrateovertimeduetochangingusehabitsdrivenbymeteredservice.Italsoshowsthedecreaseinthedifference
betweeninfluentwaterpumpedtothe

Page120

Figure4.
Plantwaterproductionasa7daymovingaverage.

Figure5.
Dailyrawwaterandfinishedwaterturbidity.

Figure6.
Rawwaterturbidityprobabilitydistribution.

Figure7.
Finishedwaterturbidityprobabilitydistribution.
TABLE3.Frequencydistributionofrawandfinishedwaterturbidityvalues.
RawWaterTurbidity
<

FinishedWaterTurbidity

Frequency

Probability

<

Frequency

Probability

0.00

0.00%

0.00

0.00%

0.25

273

33.96%

0.02

0.68%

0.50

354

77.96%

0.03

49

6.68%

0.75

44

83.99%

0.04

166

27.17%

1.00

22

86.19%

0.05

206

52.66%

1.25

21

88.81%

0.06

201

77.48%

1.50

19

91.17%

0.07

96

89.36%

1.75

91.67%

0.08

47

95.17%

2.00

15

93.53%

0.09

13

96.72%

2.25

94.65%

0.10

12

98.14%

2.50

10

95.90%

0.11

98.21%

2.75

96.14%

0.12

98.39%

3.00

97.01%

0.13

98.51%

3.25

97.01%

0.14

98.95%

3.50

97.26%

0.15

99.01%

3.75

97.26%

0.16

99.20%

4.00

98.13%

0.17

99.50%

4.25

98.38%

0.18

99.63%

4.50

98.63%

0.19

99.69%

4.75

98.63%

0.20

99.75%

5.00

98.88%

0.21

99.81%

5.25

99.00%

0.22

99.81%

5.50

99.50%

0.23

99.88%

5.75

99.50%

0.24

99.88%

6.00

99.75%

0.25

99.88%

6.25

99.88%

0.26

99.88%

6.50

99.88%

0.27

99.88%

6.75

99.88%

0.28

99.88%

7.00

99.88%

0.29

99.88%

7.25

99.88%

0.30

99.94%

7.50

99.88%

7.75

99.88%

8.00

99.88%

8.25

99.88%

8.50

100.00%

0.31

100.00%

treatmentplantandflowdeliveredtothesystem.Thisdifferencerepresentsbothbackwashwaterandfiltertowasteuponfilterstartup.Thereductioninthis
differenceovertimeisduetoageneralreductioninwaterthroughput,andoptimizationoftheoperationthroughoperatorfamiliaritywiththetreatmentsystemand
changesinthesourceonadailyandseasonalbasis.
DatacharacterizingrawwaterandfinishedwaterturbidityarepresentedonFigures5,6and7,andTable3.Thesedatashowthattheplanthasmetthe0.10NTU
criteriaforfinishedwaterturbidity98.14%ofthetimeoveraperiodcomprisedof809daysofdata.Withtheexceptionof7daysinthespringof1985duringa
periodofhighinfluentturbidity,a7dayperiodinthefallof1985whenthechiefplantoperatorwasaway,and2daysinthespringof1986,thefilterturbidityhasbeen
equaltoorlessthan0.10NTU.RecentresearchbyAlAnietal.(3)indicatesthatachievingafinishwaterturbidityof0.10NTUorless,orachievingatleasta70%
turbidity

Page121

removalareadequateguidelinesforjudgingtheeffectivenessofGiardiacystremoval.Table4characterizestheturbidityremovalsforthosedaysthatafinishedwater
turbidityof0.10NTUwasnotachieved.Forthedatapresentedandthetwocriterialistedbyreference2forcystremoval,thedaysApril1,1985,September25,
26,28and29,1985,andMarch8,1986representsituationsinwhichlessthanoptimumcystremovalcouldhaveoccurredifcystswerepresentintherawwater.
Theturbidityremovalspresentedinthistablearebasedonrawwaterturbidity.On4ofthedaysthatthe70%removalcriteriawasnotmet,bentonitewasbeingfed
whichwouldhavemadetheeffectiverawwaterturbidityabout1NTU.Thustheeffectiveremovalforthese4dayscouldbeconsideredtobegreaterthan70%
leavingonlytwodaysoutoftheperiodofrecordwhenlessthanoptimumcystremovalmayhaveoccurred.Itissomewhatconjecturalthattheremovalofartificially
createdturbidityappliestothiscasesinceresearchhashotbeenconductedtodatethatwoulddirectlyconfirmthis.
Duringan18dayperiodinJulyandAugust1984,theplantwasoperatedat150%ofthedesignsurfaceloading.Thefinishedwaterturbidityduringthisperiodranged
between0.04and0.08NTU.
DuringthecourseofatypicalyeartheplantoperatespredominantlyintheadsorbtiondestabilizationregionofthelogalumpHdiagrams,withsomeshiftingintothe
combinationregion.Figures8,9,10,11and12showtherangeofoperationforthevariousseasons.Atransitionoccursinthespringandfallintermsofsodaashand
bentoniteaddition.Sodaash,butnobentoniteisfedfromAprilthroughSeptember.Thebalanceoftheyearbentonite,butnosodaash,isfed.Thiscanbeexplained
byachangeinalkalinityintherawwaterfromabout30mg/Linthewintertoabout15to20mg/Linthespringandsummer,andthefactthatthewinterwaterhas
fewer
TABLE4.Turbidityremovalefficiencyfordateswhenfinishedwaterturbidityexceeded0.10NTU

TurbidityNTU

Date

Raw

Filtered

%Removal

Bentonite
Dosagemg/L

April10,1985

0.53

0.13

74

0.9

April11,1985

0.40

0.14

65

0.9

May18,1985

2.20

0.12

95

June15,1985

5.80

0.15

97.5

June17,1985

6.20

0.18

97

July8,1985

1.50

0.15

90

July9,1985

1.10

0.16

85

Sept23,1985

0.65

0.23

65

Sept24,1985

0.47

0.14

70

Sept25,1985

0.51

0.21

59

Sept26,1985

0.25

0.14

44

1.1

Sept27,1985

0.60

0.17

72

1.1

Sept28,1985

0.25

0.17

32

1.1

Sept29,1985

0.34

0.17

50

0.8

March8,1986

0.26

0.12

54

1.1

June18,1986

5.00

0.12

98

Figure8.
OperatingregionforJanuary.

Figure9.
OperatingregionforMarch.

Page122

Figure10.
OperatingregionforJune.

Figure11.
OperatingregionforAugust.

Figure12.
OperatingregionforOctober.

andthedifferencebetweentherawandfinishedwatertotalsolidsvalues.Inthewinterthedifferencebetweenrawandfinishedwatertotalsolidsisintheorderof70
mg/Landinthespringandsummeritisabout200mg/L.Theoperatingregionofthetworawwatertemperatureextremesencounteredforthiswatersourceare
presentedinFigures13and14.ThedatapointswereplottedonthelogalumpHdiagramasoriginallypublishedbyAmirtharajah.Sincesomeoperatingpoints
plottedonthisdiagramfalloutsideoftheadsorbtiondestabilizationregionintherestabilizationregion,acorrespondingadjustmentintheboundarybetweenthese2
regionsshouldbemadeforthiswatersource.
Datacharacterizingtheoveralloperationoftheplantwithrespecttowaterproduction,chemicaldosages,turbidity,andtotalsolidsarepresentedinTable5.
TheStateofMontanaDepartmentofHealthandEnvironmentalScienceshasconductedasamplingandanalysisprogramontwoseparateoccasionstomeasurethe
presenceofGiardiacystsintherawandfinishedwater.ThisdatapresentedinTable6showsthatcystsandlargeamountsofGiardiasizematerialarepresentinthe
rawwaterwhilethefinishedwatercontainedlittlesuspendedmatterandnocystswerefound.Thecontactflocculatorclarifierandfilterbackwashwassampled,but
containedtoomuchflocculatedmattertovisuallyidentifycysts.
Backwash
Thecontactflocculatorisbackwashedwithairandcoagulatedinfluentonthebasisofheadlossacrossthemedia.Thesandfiltercallsforabackwashwith

Page123
TABLE5.Statisticaldatafortreatmentplanoperation
FLOWDATA(gal/day)

Influent

Backwash

Clearwell/Res

802

800

800

Maximum

2,006,000

560,000

1,699,000

Minimum

472

49,000

Average

788,297

22,831

669,703

Std.Dev.

349,325

33,200

303,781

Number

CHEMICALDOSAGES(mg/L)

Alum

SodaAsh

Bentonite

Number

792

775

733

Polymer
801

Maximum

17.4

19.0

1.8

27.0

Minimum

0.5

0.0

0.0

0.3

Average

6.1

3.5

0.7

0.6

Std.Dev.

3.4

4.8

0.5

0.9

TURBIDITY(NTU)

Raw

Floc1

Floc2

Eff1

Number

804

804

801

809

807

Maximum

8.50

30.00

21.00

0.23

0.31

Minimum

0.10

0.13

0.10

0.01

0.02

Average

0.61

0.40

0.35

0.056

0.057

Std.Dev.

0.91

1.28

0.74

0.02

0.02

Eff2

TOTALSOLIDS(mg/L)

Raw

Floc1

Floc2

Eff1

Eff2

Number

48

49

49

49

49

Maximum

450

55

70

41

50

Minimum

19

Average

121

28

29

16

16

Std.Dev.

97

11

12

finishedwaterautomaticallyuponreading8feetofheadloss.Asapracticalmatter,theoperatorinitiatesbackwashonceperdayorevery24hoursofoperationto
keepthefilterbedfresh.Theaveragefilterbackwashuseof22,381gal/dayequatestobackwashingasinglefilterbasineachdayontheaverage.Thedifference
betweentheamountofrawwaterpumpedtotheplantandthesumofthebackwashwater(forfilters)andwaterproducedisabout95,800gal/dayontheaveragefor
theperiodofrecord.Thisdifferenceiswaterfilteredtowaste,andunfilteredwaterusedtoflushthecontactflocculatorunits.Thisamountofwaterhasbeenreduced
to77,000gal/dayin1986.
CostData
Theannualcostsforchemicals,power,labor,andmaintenancearepresentedinTable7.Theprojectcostofconstructionofthewaterplant,upperzone,storage,and
lowservicepumpstationof$1,200,000U.S.inMay1983hasbeenupdatedtofirstquarter1987andamortizedat10%interestover20years.Basedonacurrent
annualdeliveryof242MGtothedistribution

Figure13.
Operatingregionfortemperaturerangeof33to34F.

Figure14.
Operatingregionfortemperaturerangeof44to51F.

Page124
3

system,thecostofproductionis$0.77/1,000galor$0.58/100ft .ThesecostsaresummarizedinTable7.
TABLE6.Giardiacystsamplinganalysisandresults

1291986

Turbidity

12111985

Raw

Filtered

Raw

Filtered
0.02

0.2

0.03

0.2

#cystsfound

Gallonsfiltered

400

570

900

1390

Smallparticledebris

+++

++

+++

++

Largeamorphous
debris

++

++

Cellularplantmaterial

+++

+++

Diatomsandalgae

++

Protozoa

++

Insects

Nematodes

Chemicals:

Nalco8184

0.75MG/L

Alum

6.0MG/L

Bentonite

1.1ppm

Turbidityafter

bentoniteaddition

1.1NTU

0nonefound
+rare
++moderate
+++heavy

TABLE7.Water
productioncosts

Operation&Maintenance

$/year

$/1000gal

Power

10,623

0.044

Chemicals

7,168

0.029

Labor

16,425

0.068

Total

34,216

0.141

Constructioncosts1stQTR1987
Amortized20yearsat10%

1,304,000

153,205

Costper1000galproduced

0.630

Currentcostofwaterproduced

0.771

LiteratureCited
1.Logsdon,G.S.,andF.B.DeWalle.Filtrationasabarriertopassageofcystsindrinkingwater.JointpublicationofU.S.E.P.A.,Breidenback,Environmental
ResearchCenterandtheUniversityofWashington.
2.Amirtharajah,A.,andK.J.Mills.1982.Rapidmixdesignformechanismsofalumcoagulation.JAWWA774(4):210.
3.AlAni,M.Y.,etal.1986.RemovingGiardiacystsfromlowturbiditywaterbyrapidratefiltration.JAWWA78(5):66.

Page125

InactivationofGiardialambliaCystsfromaSurfaceWaterbyOxidationwithOzone
CarlNebel*,AnthonyLally,ThomasBosher,J.WilliamHmurciak,LindaHmurciakandDorothyA.Breen
PCIOzone&ControlSystemsInc.,1FairfieldCrescent,WestCaldwell,NewJersey07006,U.S.A..
Giardialambliafoundinsurfacewatercanbeinactivatedbyoxidationwithozone.Ozoneismanufacturedonthesitefromelectricpower,airandcoolingwater.Theozoneair
mixtureisdispersedinthewaterviafinebubblediffusers.OzonedissolvedinthewaterinactivatedGiardiacystsviaoxidation.Concurrentlyozonealsooxidizescolour,taste,and
odourfoundinthewater.TherateofoxidationofGiardialambliabyozoneisfasterthanchlorineandchlorinedioxide.Atheoreticaldiscussionofmicrobialinactivationwillbe
presented.Theengineeringdesign,startupandoperationofafullscalefacilitywillalsobedocumented.

Introduction
TheProblem
TheTownofNorthAndoverMassachusettshasfordecadesobtaineditswaterfromnearbyLakeCochichewick.Twowaterpumpingstationssuppliedwaterintothe
distributionsystemafterchlorination.Althoughthewaterwasplaguedwithtaste,colourandodourproblems,itwasgenerallyacceptedasbiologicallysafeafter
chlorination.Thetasteandodourproblemsweremainlyduetochlorinationofhumicmaterialsfoundinthiswatersource.Thesehumicacidsimpartedatancolourto
thewaterandalsoaretheresultofaratherhighchemicaloxygendemand(COD)whichisintherangeof1220mg/L.Thehumicmaterialsservedasafoodsourceto
abiofilmcomposedofnonpathogenicmicrobesinthewaterdistributionsystem.Attempttocontrolthegrowthofthisbiofilmduringthesummermonthswith
chlorineand/orchlorinedioxidewereunsuccessful.Duringthewintermonths,thebiofilmwascompletelycontrolledbyemployingadditionalchlorinationpointsinthe
distributionsystem.Inpart,thesuccessofthiseffortcanbeattributedtolowerchlorinedemandduetotheuseofozoneandtothefactthatcoolerwaterwillholdthe
chlorineresiduallonger.
TheMassachusettsDepartmentforEnvironmentalQualityEngineeringfoundthatahigherthannormalnumberofthetown'scitizensweresubjecttoGiardiasisand
thereforeorderedthattheTownofNorthAndoverrequirethatallwaterforhumanconsumptionbeboiledforfiveminutespriortouse.Thewatersource,Lake
Cochichewick,wastestedand3to4cystswerefoundina300gallonsampleofwater.Thetotalcoliformcountinthelakewaterwasintherangeof10to700/100
mL.
SolutiontotheProblem
Initialconsiderationforthecontrolofgiardiasiswasgiventochlorination.Increasingthechlorinedosagelevelshouldresultinadequatecontrol.Asearchofthe
literatureindicatedthatchlorinationwasnotsuccessfulininactivatingGiardialambliacysts(1).ItwasalsofoundthatG.lambliacystsbecomemoreresistantto
chlorineasthewatertemperaturedecreases(2).NorthAndoverMassachusettsislocatedinageographicareawherethewatertemperaturesdropsubstantially,
thereforeconsiderationfortheuseofchlorinewasterminated.
Sproulandcoworkers(3)foundthatozoneisveryeffectiveinoxidizingG.lambliacystsatboth25Cand5C.Asexpected,theinactivationofthecystrequires
moreozoneatalowertemperaturethanathighertemperatures.(Figure1and2).
Intheabovestudies,Sprouletal.,oxidizedG.lambliacystshavingaconcentrationof10,000cysts/mLwith

Figure1.
InactivationofG.lambliacystsbyozoneatpH7and5C
*Correspondingauthor.

Page126

Figure2.
InactivationofG.lambliacystsbyozoneatpH7and25C.

ozoneattwodifferenttemperaturesandthreedifferentozoneconcentrationseach.Theexcystationwasdeterminedbythemicroscopicmethod.Theproductofozone
concentrationtimesthecontacttimeinminutes(Ct)for99%inactivationis0.17and0.53mgmin/Lat25Cand5Crespectively.
Table1showsthatG.lambliacystsareveryresistanttochlorinewhencomparedtopoliovirusandE.Coli.G.lambliaisthreetimesmoreresistanttoozoneat5C
thanitisat25C.InthesametemperaturerangeG.lambliabecomeseighttimesmoreresistanttochlorine.Itisforthisreasonthatozonewaschosenoverchlorine
tocontrolG.lambliaforNorthAndover.
Theuseofhigherconcentrationsofchlorineduringthecoldmonthsinthistypeofwaterwouldhavecreatedmoretasteandodourproblemsandwouldhaveincreased
thetrihalomethaneconcentration.
ItwasdesiredthatthetimeforimplementinganysolutiontotheGiardiaproblembekeptataminimumbecausethetownwantedthestateimposedwaterboilorder
tobeliftedassoonaspossible.Beforewecanaddresstheengineeringaspectsofutilizingthisoxidant,anunderstandingofozoneanditsgenerationisinorder.
Ozone
Ozoneisalowmolecularweightmolecule(M.W.48)whichiscomposedofthreeoxygenatomsthatarechemicallyarrangedinachain.Thebondanglebetweenthe
oxygenatomsis116,hencethechainisintheshapeofatriangle.Ozone,(O3)anallotropeofoxygen(O2),iscomposedofthesameatomscombinedinadifferent
form.Thepropertyofozoneforoxidizingmicroorganismsisattributedtothefactthatitisthesecondmostpowerfuloxidantknown(2.08V).Onlyfluorineexceeds
ozoneinitsoxidationpotential(2.87V)andchlorineis(1.34V).Thehighchemicalreactivityofozoneisrelatedtothefactthatitpossessesanunstableelectron
configurationwhichrequiresittoseekelectronsfromothermolecules.Duringitsreactionwithothermolecules,ozoneisdestroyed.Theendresultofozoneoxidation
oforganicmolecules,suchasthosefoundinmicroorganisms,iscarbondioxideandwater.
Ozone,likeoxygen,isacolorlessgas.Unlikeoxygen,ozoneisdifficulttoobtaininpureform.Commercialozoneisformedinconcentrationsoftwopercentbyweight
fromair.Beforetheoxidationofmicrobesinwatertakesplace,theozonedispersedinairmustbetransferredfromairintowater.Thisimpliesthatameansof
transferringozoneintowatermustbeincorporatedintothissystem.Theozonecontactingdevicewillbediscussedlater.
OzoneGeneration
Ozoneisanunstablemoleculewhichslowlydecomposesbacktooxygenfromwhichitwasmade.Atambienttemperatures,thehalflifeofozoneinairis
approximatelyfourteenhours.Becauseozoneisnotcompletelystable,itcannotbepurchasedasacompressedgas.Itmustbegeneratedatthesiteoftheapplication
andusedshortlyafteritsregeneration.Thisonsitegenerationcircumventsthenecessityofhavingtotransport,storeandhandleacompressedgas.
Commercialozoneisgeneratedbyacceleratingelectronsbetweentwoelectricallychargedelectrodes.Whenanelectronispropelledtoahighvelocitywhoseenergy
isintherangeof67eV,aninteractionbetweentheelectronandtheoxygenmoleculetakesplacetodisassociatetheoxygenmoleculeintotwooxygenatoms(4).
O2+HighEnergyElectron>
2O+LowEnergyElectron
Theoxygenatomsformedareaveryreactivespeciesandreactalmostimmediatelywithoxygenmoleculestoformozone.
O+O2>O3
Thenetreactionis:
3O2>2O3
H=34.61kcal/mole
TABLE1.Comparativeconcentrationtimedatafor99%inactivation.

G.lambliaCysts

E.Coli

ConcentrationTime

Ozone

25

0.17

0.53

25

15.00

125.00

20

0.08

0.22

Chlorine

2.00

Ozone

0.02

Chlorine

0.04

Chlorine

PoliovirusI

TemperatureC

Disinfectant

Ozone

Page127

Figure3.
OzoneGeneratorElectrodeAssemblydoublefluidcooled.

MaterialsandMethods
TheOzoneGenerator
Theozonegeneratorisanelectronicdevicewhichaccelerateselectronsinthepresenceofverydryair.Asilentcoronadischargeisproducedbetweentwocharged
electrodes.Thecoronadischargeaccelerateselectronswhichinturndisassociatesoxygenfoundinairmoleculesintooxygenatoms.Alternatingcurrentmustbe
employedwhenozoneisgeneratedinthecoronadischarge.Ifadirectcurrentwereemployed,theelectronwouldenterthecoronadischargeandimmediately
proceedtothegroundedelectrodewhereitwouldbecomeunavailableforfurtherinteractionwithoxygenmolecules.Whenalternatingcurrentisused,theelectron
vibratesbetweenelectrodesinaccordancewiththefrequencyofthealternatingcurrent.Thehigherthefrequency,thegreaterthetimetheelectronwillexistinthe
dischargearea.Ozoneproductionshould,therefore,beafunctionofthefrequencyappliedtothehighvoltageelectrode.Generally,ozoneproductionisdoubled
whenevertheelectricalfrequencyisdoubled.Withtheuseofproperlydesignedhighfrequencyinverters,modernozonegeneratorscanoperateat2,500cyclesper
second.Therateofozoneproductioncanbereadilychangedthroughatentooneturndownbyalteringthepowerappliedtotheelectrodes.
Figure3showsadiagramoftheozonegeneratingelectrodeswhicharecontainedinanozonegenerator.Thegrounded321stainlesssteel1/4"thickelectrode(A)is
placedinthecenteroftheozonegeneratingmodule.Theelectrodeisfixedinaverticalpositionandiscooledbypassingpotablewaterthroughit.Thetotalcooling
waterrequirementforanozonesystemcapableofproducing1150poundsofozoneperdayis32gpmandthetemperatureriseofthewaterisfrom70Fto76F.
Thecoolingwaterisdiscardedintotheozonecontactor.Surroundingthegroundedinnerstainlesssteelelectrode(A)isaglasselectrode(B).Theoutersurfaceofthe
glass(C)isplatedwithsilverwhichservesasthehighpotentialelectrode.Topreventdestructionofozoneontheinnersurfaceoftheglasselectrode,theoutersurface
oftheelectrodeiscooledwithanonelectricalconductingfluidwhichiscontinuouslyrecirculatedinaclosedloopsystem.Theheatwhichisremovedbythecooling
fluidistransferredtowaterinashellandtubeheatexchanger.Attentionispaidtothatremovedfromtheozonegeneratorbecauseheatinducesdecompositionof
ozonebacktooxygen.
2O3>3O2
Dryairispassedthroughtheannularspace(D)betweenelectrodes(A)and(B).Itisinthisareawherethecoronadischargetakesplaceandwheretheozoneis
produced.Thesilverglasselectrode(C)ischargedwith10,000voltsatafrequencyof2,500cyclespersecond.Theglasselectrode,whenoperatedat10,000volts,
hasaninfinitelife.
Theairwhichispassedthroughtheozonegeneratormustbeoilless,particlefreeandmusthavemoistureremovedtoa40Fdewpointorlower.Therateofozone
productioncanbevariedfromanexternal420mADCcontrolsignal.Thestartuptimeofairpreparationunitandtheozonegeneratorisapproximatelyoneminute.
OzoneSystemSizing
ThesinglemostimportantconcerninsystemsizingistosupplyanadequatecontacttimeandozoneconcentrationtokillGiardialambliacysts.TheworkofSproul
(3)hasshownthatthisshouldbeatleast0.53mgmin/L.ThewateratNorthAndoverexhibitsaveryhighinstantaneousozonedemandwhichwilldecreasethe
amountofozoneavailableformicrobialcontrol.ThisozonedemandexistsbecausethewaterhasbeenshowntohaveCODvaluesashighas20mg/L.Considering
theozonedemandofwaterandtheresidualrequiredforG.lambliainactivation,anapplieddosagelevelof5mg/Lwaschosen.Atawaterflowrateof2400gpm,
thecorrespondingquantityofozoneis150poundsperday.Afurtherconsiderationincontactorsizingisthegastoliquidvolumeratio.Ozoneisgeneratedata
concentrationof2%inair,henceozoneand98%percentairmustbediffusedthroughthewater.Thequantityofgasappliedtothewaterfrom150poundsofozone
ata2%concentrationis70scfm.Ifthecontactoristoosmall,thenthesmallairbubbleswhichareformedwillcoalesceintolargerbubbles.Thiswouldresultina
poortransferofozonefromairintowater.Factorsaffectingozonecontactingdesignaredocumentedintheliterature(5)aswellastheapplicationofozonetopotable
water(6).
Consideringtheabovethreefactors,acontactorwasconstructedwhichhasthefollowingdimensions:10feetwideby20feetlongwithawaterdepthof16feet.Ata
waterflowrateof2400gpm(3.5mgd)thecontacttimewas10minutes.Thecontactorconsistedoffourportsthroughwhichthewaterflowedonaverticalplane
(Figure4).Atthebottomofeachport,porous

Figure4.
OzoneContactor.

Page128
TABLE2.PublicHealthimpactofstudybeforeandafterozone.

Totalreportedcases
ofgiardiasis

Beforemandatemadeitreportablein1985

Beforetheinstallationofozone(1/1/8610/1/86)

23

Aftertheinstallationofozonefrom10/1/86

Totalreportedcasesin1987

stonediffuserswereplacedbywhichtheozoneairmixturewasspargedthroughthewater.
Inpractice,itwasfoundthattheozoneresidualattheendofthecontactorwasintherangeof0.91.0mg/L.Ifweassumethatthisresidualcontinuestobuild
throughthecontactorinalinearmode,theneffectiveresidualisonehalfofthefinalresidualor0.5mg/L.Theproductoftime(10minutes)andresidual(0.5mg/L)
nowbecomes5mgmin/L.Thiscomparesfavorablywiththe0.53mgmin/LestablishedbySproul(3)for5Cwater.Althoughthisisnearlyanorderof
magnitudegreaterthantheminimumrequirement,itshouldbenotedthatthisresidualwilldecreaseslightlyduringthesummermonthswhenthewateriswarmer.
Frominceptiontosystemstartup,thisprojecttookapproximatelytwomonths.Thistimeincludestheengineering,equipmentmanufacturingandsiteconstruction.
Constructiontimeofthecontactorwasdecreasedbytheuseofmildsteelsheetpiling,woodbafflesandaconcretetop.Theuseofthemildsteelandwoodinan
ozonecontactoriscertainlyinnovative,butwaswarrantedonaneconomicandtimingbasis.
ResultsandDiscussion
ThepresenceofviableGiardialambliacystshasbeeneliminatedviaozonation.Totalcoliformcountenteringtheozonecontactorisintherangeof10700mg,
whereastheplatecountleavingthecontactoriszero.Whenthepumpingstationemployedonlychlorine(~4mg/L),thetrihalomethanelevelswereintherangeof8
120mg/L.Theinstallationofozonepriortochlorinationloweredthetrihalomethaneconcentrationtotherangeof1.1to2.0mg/L.Thetruecolorlevelsdropped65to
95%.Thetasteandodourlevelsalsoshowedasubstantialdecrease.ThepublichealthimpactofthisstudybeforeandafterozonearelistedinTable2.
LiteratureCited
1.Craun,G.G..1979.WaterborneGiardiasisintheUnitedStates:areview.Am.J.PublicHealth69:817.
2.Jarroll,E.A.etal..1981.EffectofchlorineonGiardialambliacystsviability.Appl.andEnviron.Microbiol.41:483.
3.Wickramanayake,G.B.,Rubin,A.J.,andO.J.Sproul.1984.InactivationofGiardialambliacystswithozone.Appl.andEnviron.Microbiol.48:671.
4.Nebel,C..1981.Ozone.EncyclopediaofChemicalTech.16:683.
5.Nebel,C..1981.Ozonewatertreatmentsystems.WaterEng.&Manage.ReferenceHandbookR77.
6.Nebel,C..1981.Ozonetreatmentofpotablewater.PublicWorks112(1):86,112(2):68.

Page129

ARegulatoryAgency'sExperiencewithGiardia
S.McClure*andI.B.Mackenzie
AlbertaDepartmentofEnvironment,9820106thStreet,Edmonton,Alberta,Canada.T5K2J6.
TheprovinceofAlbertahasexperiencedtwomajoroutbreaksofgiardiasiswhichcouldbeattributedtopublicwatersupplytransmission.Followingtheseincidences,theprovince
initiatedanextensiveGiardiamonitoringprogramspanningthreeyearsandinvolvingover700watersamplesfrommorethan40Albertacommunities.Onlyinthreeofthese
sampleswereGiardiacystsdetected.Itisthoughtthattheexistingmonitoringmethodhasmanyinherentlimitationsandthuscannotbeeffectiveforpredictingorcontrolling
giardiasisoutbreaks.UntilsuchtimeassignificantimprovementsinGiardiacystdetectionandrecoveryareavailaable,itisfeltthatmoretraditionalindicatorsofplant
performancewillbemoreeffectiveinensuringcystfreewater.

Introduction
TheProvinceofAlbertahasexperiencedtwomajoroutbreaksofgiardiasiswhichcouldbeattributedtopublicwatersupplies.Thisresultedinthedevelopmentofa
Giardiamonitoringprogramandareviewofthewatertreatmentpractices.ThispaperreviewstheeffortsmadebytheMunicipalEngineeringBranch,Pollution
ControlDivision,AlbertaEnvironment,indealingwithpotentialwaterbornegiardiasisoutbreaks.
Thefirstknownoutbreaksoccurredlateinthewinterof1982intheresorttownofBanff,locatedintheCanadianRockies.Thetownhasapermanentpopulationof
approximately4,000,withpeakseasonfluxesreaching20,000.Over150casesofgiardiasiswerediagnosed.
Thetown'swatersupplycamefromareservoironFortyMileCreekwhichwasuntreatedexceptforchlorination.TheCreekwatershediswellprotectedfromhuman
activity,however,thePublicHealthInspectorinvestigatingtheoutbreakfoundthatbeaverhadcolonizedthereservoir,andGiardialambliacystsweresubsequently
detectedinsamplestakenfromthereservoir.Thetownhassinceconvertedtogroundwatersourcesandthereappearstohavebeennoreccurrenceofthedisease.
ThesecondmajoroutbreakoccurredintheCityofEdmontonlateinthefallof1982andthespringof1983.Thelocalhealthunitdidafollowupsurveytodetermine
theextentoftheinfection.Theirepidemiologicalreport(CollierandMacdonald1983)indicatedtherehadbeen895laboratoryconfirmedcases.Thedatesonwhich
thenumberofinfectionswerereportedformsaclassicepidemiccurvegenerallyassociatedwithasingleinfectivesource.Dataplottedonacitymapindicatedthatthe
majorityofpeopleaffectedlivedinthedowntownorUniversityareaanareaservicedbytheRossdaleWaterTreatmentPlant.
TheRossdaleWaterTreatmentPlantisaconventionalwatertreatmentplantutilizingchlorinedioxideasapredisinfectant,chloraminesasapostdisinfectant,alum
coagulation,sedimentation,filtration,andlimesoftening.Duringtheincident,theplantwasneitherpredisinfectingwithchlorinedioxidenorcarryingoutsoftening
procedures(Lippy1984).Althoughtherewasnoconclusiveevidencetoimplicatethewatersupply,noothercommonsourcefortheinfectionwasidentified.
Inthesummerof1983,itbecamemandatorytoreportgiardiasisinAlberta.Sincethattimeithasbeenoneofthemostcommondiseasesintheprovincewith
approximately1300to1500casesperyear.However,thedatabasedoesnotsuggestanyepidemicsattributabletowatersupplies.
Becauseofthehighincidenceofgiardiasisandtheaforementionedepidemics,itwasdecidedthattheprovinceshoulddevelopawatertreatmentplantmonitoring
program.
TheoverallobjectivesoftheprovincewideGiardiamonitoringprograminitiatedinthespringof1983,wereto:
1.Developthecapacitytopredictandcontrolgiardiasisoutbreaks
2.ProvideawidecoverageofthewatersuppliesforthemajorityoftheAlbertapopulationtoevaluateiftheorganismswerepresent
3.Developandrefinesamplinganddetectiontechniquesand
4.Monitorwatertreatmentplantmethodsandperformance.
MaterialsandMethods
TheGiardiasampleconcentrationmethodasoutlinedinthe15theditionofStandardsMethods(APHA1980)whichutilizesa7mwoundorlonfilterwasinitially
investigated.AnalysiswereconductedbytheProvinciallaboratoryofPublicHealthassociatedwiththeUniversityofAlbertainEdmonton.Thislaboratoryhadhad
considerableexperienceinperformingprotozoananalysisfromstoolsamplesfordiagnosticpurposes.
ThelaboratorysubsequentlyreportedthatthereweredifficultiesinutilizingthefloatationmethodasoutlinedinStandardMethodsandconcludedthattheyweremore
successfulincentrifugingtoconcentratethewaterfromthefilter,andthenmicroscopicallyexaminingtheentirepellet(thisresultedinapproximately60slidesper
samples).Becauseofthedifficultyandtimerequiredtoprocesstheorlonfilters,andthepotentialcystloss,othermethodswereinvestigated.
*Correspondingauthor.

Page130

Figure1.
SchematicofGiardiasamplingapparatus.

TheBranchnextlookedata5mpolycarbonatemembrane.Ahousinggenerallyusedingeologicalworkandknownasageofilterwasused,howeverthemembrane
quicklycloggedandincreasedpressureonthefilterhousing.Thehousingwouldthenopenorseparatecausingawashoutofthemembranesurface.Thelaboratory
wasalsoexperiencingdifficultiesinconductingtheirexaminationsbecauseoftheamountofdebris,particularlyalumsludgethatwasbeingfilteredout.
Duringthesummerof1984,anotherconcentratingdevice(Figure1)wasdesigned.Itbasicallycondistedofa20mwoundorlonroughingfilterfollowedbythree
5mpolycarbonatefilters.Initialtrialswerecarriedoutinthefallof1984todeterminesuitableflowandpressureconditions.ThedevicewascalibratedinJuneof
1985usingGiardiacystsobtainedfromdogfecesandthencomparedtoamembranefiltrationdeviceusedbytheKananaskisCentre.TheCentre'ssystemconsisted
ofthesamemembrancewithnoprefiltrationdevice.
Thefiltrationdeviceswerefoundtobegenerallycomparable.However,theKananaskislaboratoryobtainedhigherrecoveryratesusingthezincsulphateflotation
methodcomparedtotheProvincialLaboratory'smethodofexaminingthecentrifugeplug.Forthe1985/86program,samplingwasconductedusingtheAlberta
EnvironmentdiviceandtheKananaskisLaboratorymethodology.
Discussion
Allsamplingwasconductedontreatedmunicipalsurfacewatersupplies.Largermunicipalitiesweresampledonaweeklybasis,andsmallermunicipalitiesona
monthlybasis.Duringthe1983/84and1984/85seasons,thesampleswerecollectedbyEnvironmentstaffandanalysiswereperformedbytheProvincialLaboratory
ofPublicHealth.Duringthe1985/86season,someofthemunicipalitieswereprovidedwithsamplingdevicesandrequiredtotaketheirownsamples.Thesewere
thensenttotheKananaskisLaboratoryforanalysis.
AmajorproblemencounteredinthesamplingforGiardiacystswaspluggingofthemembranebydebris.Notonlydidthisaffecttheconcentrationprocedurebutit
alsogreatlyhamperedcystidentification.Alumsludgewasthemajorinterferingagentbutotherorganicandinorganicdebrisincludingalgaeandnematodeswere
frequentlyencountered.Occasionally,airwasentrainedintheinfluentfilterwaterandcausedittobindoff.
Themostsignificantproblemwasthatoflowcystrecoveryrate.Theorlonfiltersgenerallyhadrecoveriesoflessthan10%whilerecoveriesforthepolycarbonate
membranesrangedfromlessthan10%tomorethan40%dependingontheinfluentcystconcentration.Therecoveryrateswereverifiedseveraltimesthrough
calibrationsusingbothliveorganismandpolystyrenebeads.Althoughthelowcystrecoveryisconsideredaseriousshortcoming,thetechniquesusedwereconsidered
comparabletothosepractisedelsewhere.
Table1presentstheresultsobtainedfromthethreeyearprogram.Forthisperiodover700sampleswereexaminedfromconcentratingmorethan450,000Lof
water.Fromthis,threepossibleGiardiacystsweredetected.Followupintensesamplingatthemunicipalitieswheretheyweredetectedrevealednofurthercysts,
norwasthereanyincreaseingiardiasiscasesreportedbythelocalhealthunits.
Conclusions
1.Theorganismcouldnotbedetectedinsignificantnumbersinthetreatedwatertested.
2.Usingcurrentprocedures,theGiardiamonitoringprogramwouldnotbeeffectiveinpredictingorcontrollingtheoutbreakofgiardiasis.Thetestsaretime
consumingandexpensiveforthetypeofresultstheyproduce.
3.Thesamplingandanalysistechniquesdevelopedarethoughttobeatleastcomparabletothosepractisedelsewhere.
4.Theprocessofmicroscreeningwaterisalsoausefulindicatoroftheefficiencyofoperationofwatertreatmentplants.Thisisindicatedbytheamountoforganisms,
sludge,anddebristhatwasaccumulatedonthefilters.
AlbertaEnvironmentwillcontinuetomaintainamonitoringcapabilityforGiardia,butwillnotconductasamplingprogramofthemagnitudeofthe1983/84and
19844/85season.
TABLE1.AlbertaEnvironmentGiardiaMonitoringProgram.
NumberCommunities
Monitored

Total*
Water
Filtered(L)

Number**
Samples

Potential***
NumberofCysts

1983/84

29

316.0103

321

1984/85

40

79.4103

337

1985/86

12

66.7103

66

Year

*Usingvariousfilter
**Varyingsamplesizes
***Unconfirmedbeyondvisualidentification

Page131

TheDepartmentwillembarkonaprogramtoencourageoperatorstoimprovetheirwatertreatmentpractices.Initiallythiswillinvolveconcentratingontheproperuse
ofcoagulantsusingturbidityastheindicator.
LiteratureCited
1.APHAAWWAWPCF.1980.Standardmethodsfortheexaminationofwaterandwastewater.15thEdition,Washington,D.C.
2.Collier,M.K.andP.Macdonald.1983.GiardiasisinEdmonton.EdmontonLocalBoardofHealth,Edmonton,Alberta.
3.Lippy,E.1984.Reviewoftreatmentpractices.RossdaleWaterTreatmentPlant,AlbertaEnvironment,Edmonton,Alberta.

Page133

EffectsofChlorineontheUltrastructureofGiardiaCysts
M.Neuwirth*,P.D.Roach,J.M.BuchananMappinandP.M.Wallis
AlbertaEnvironmentalCentre,Vegreville,Alberta,Canada,TOB4L0.
Giardiamuriscystswereexposedtochlorineconcentrationsof0,1.3,and4.3mg/Lfor10,20,and30minutesandat10.5mg/Lfor90minutes.Cystswererecoveredfromthe
experimentalbeakersbyconcentrationon5mNucleporemembranesandcentrifugation.Samplesweresplithalfwerepreservedin5%glutaraldehydeforTEMexaminationand
theotherhalfweresubjectedtoviabilityassaysbyinvitroexcystation.Afterabufferrinse,aliquotswerefixedandembeddedforTEM.Controlcystswerefixedwithouttreatment.
TransmissionEMshowedthatcyststreatedwith0mg/Lchlorinefor10,20and30minutesappeared''normal".Theycontained2or4nuclei,ribosomes,roughER,axonemes,
microtubules,remnantsofventraldiscs,andvacuoles.Thecystsexposedtochlorineexhibitedvaryingdegreesofmorphologicalchanges.Theseincludesomecystwall
deterioration,rufflingoftheplasmamembraneandprogressivegranularityofthecytoplasm.Nucleiandflagellaappearedunaffectedinmostcystsstudied.Theseultrastructural
changesarecorrelatedwithviabilityasdeterminedbyexcystation.

Introduction
Waterbornegiardiasishasbeenwellestablishedasanimportantworldwidehealthproblem.TransmissionofGiardiacystsisfrequentlyviapublicwatersupplies.The
efficacyofchlorinationfortheinactivationofhumaninfectiveGiardiaduodenaliscystsisamatterofgreatconcerntothoseresponsibleforwatertreatmentplants.
Inspiteofthenumberofstudiesoneffectsofhalogensoncystviability(1,4,5,7,10),thereisnoinformationdealingwithultrastructuralchangesincystsinactivatedby
chlorine.
Inthisstudy,theeffectsofchlorineonthemorphologyofGiardiamuriscystswereobserved.G.muriscystswereusedsincetheyaresimilarinmorphologytothe
humanG.duodenaliscystsbutarenoninfectivetohumans.
MaterialsandMethods
Cystswereseparatedfromfeccalmaterialcollectedfrommicebyflotationon1Msucrose(14).Chlorinestocksolutionswerepreparedbyaddingsodium
hypochloritetodistilledwaterin20Lquantitiesandwereallowedtoequilibratefor24hours.ExposuremediumconsistedofnonchlorinatedtapwateratpH8.2to
whichsufficientstockwasaddedtoprovideafreeresidualchlorineconcentrationof0,1.3,4.3,and10.5mg/L.Alltestswereperformedat6Conastirring
apparatuswhichstirredthecontentsof6beakerssimultaneously.
Approximately106cystswereaddedtothechlorinesolutiontomakeupafinalvolumeof800mL.Chlorineconcentrationsweredeterminedbyamperometric
titrationusingaFishermodel397Cltitrimeter.Thecontentsofeachbeakerwerefilteredthrougha5mNucleoporefilterafterstirringfor10,20,30,or90minutes.
OnemLof10%sodiumthiosulphatewasaddedduringfiltrationtoneutralizethechlorine.CystswerewashedfromthefilterswithTritonX100solutioninto15mL
tubesandconcentratedbycentrifugation.
Eachpelletwasdividedintotwoaliquots.OnewasexaminedforexcystationbythemethodofRiceandSchaefer(11).Theotherwasresuspendedin5%
glutaraldehydein0.1Mcacodylatebuffer,pH7.2,rinsedin0.1McacodylatebufferpH7.2,postfixedin1%osmiumtetroxidein0.1McacodylatebufferpH7.2,
dehydratedinethanolseriesandembeddedinSpurrsepoxyresin(13).Thepelletobtainedfollowing10mg/Lchlorineand90minexposurewasalsotreatedfor
excystation.ThinsecctionswerecutonanLKBUltrotome,stainedinuranylacetateandleadcitrate(9)andviewedandphotographedwithaHitachiH600electron
microscope.
Results
TheresultsofexcystationofGiardiacystsafterchlorinetreatmentaresummarizedinTable1.Theseresultsshowthatasconcentrationofchlorineandexposuretime
increase,theinactivationofcystsincreases.Aconcentrationof10.5mg/Lfor90minutesensurestotalinactivationofcysts.
Morphologically,thecystsexposedto0mg/Lchlorinehadanormalappearance(Figure1).Thecystwallis0.30.5mthickandiscloselyappliedtowhatappears
asanarrowbandofcytoplasm.Aspaceorlacunaispresentbetweenthisnarrowbandofcytoplasmandtheorganism.Thecystmembraneenclosedtheorganism
whichconsistsof2or4nuclei,axonemes,ventraldiscfragments,ribosomesandoccasionalroughendoplasmic
TABLE1.ViabilityofGiardiamuriscystsfollowingchlorinetreatmentmeasuredas%excystation.

Chlorineconcentration(mg/L)

0
Exposuretime
(min)

4.3

10.5

100

100

100

10

100

97

86

20

100

100

73

30

100

93

63

90

*Correspondingauthor

1.3

Page134

Figure1.
Cystexposedto0mg/Lchlorinefor10minutesandvesicles
withelectrondenseandelectronlucentmaterial(x9000).

Figure2.
Cystexposedto0mg/Lchlorinefor10minutes,
showingflagellainlacuna(x13,500).

Figure3.
Cystexposedto1.3mg/Lchlorinefor30minutes
showinga"normal"appearance(x10,000).

Figure4.
Cystexposedto4.3mg/Lchlorinefor20minutes
showingmoderategranulation(x15,000).

Figure5.
Cystexposedto4.3mg/Lchlorinefor20minutesshowing
granulationofcytoplasmandshrinkageofcyst(x17,000).

Figure6.
Cystexposedto10.5mg/Lchlorinefor90minutesshowing
"normal"appearanceofcytoplasm,rufflingof
narrowbandofcytoplasm(17,000).

Figure7.
Cystexposedto10.5mg/Lchlorinefor90minutes
showinganecroticcell(x8,000).
A,axonemesCW,cystwallEr,roughendoplasmicreticulum
F,flagellum<granulation/flocculationL,lacunan,
nucleusv,vesiclevd,ventraldiscfragment.

Page135

reticulum.Aprominentfeatureisalayerofvacuolesjustinsidethecellmembrane,someofwhichcontainelectrondensematerialbutmostofwhichcontainsome
flocculentbutnotveryelectrondensematerial.Electrondensevesiclesalsoappearinthelacunae(Figure2).
Someofthecystsexposedtochlorinealsoappear"normal"inmorphology(Figure3).However,incaseswherechlorinehasobviouslyenteredthecyststhereis
damagetothecytoplasmwhichisevidencedbygranulationofthecytoplasmwithclumpingofribosomesandspacesrepresentingleachedoutmaterial(Figure4).The
lacunarspaceisenlargedduetotheshrinkageoftheorganism(Figure5).Thenucleiandflagella/axonemesarestillrecognizable.Thereappearstobealossofthecyst
wallinmanycystsexposedtochlorinesincethewalldoesnotstainwithuranylacetateandleadcitrate.Insamplesexposedfor90minutestomg/Lchlorine,afew
cystsappear"normal",however,thebandofcyloplasmappearsruffled(Figure6),butinmostcyststhecytoplasmisflocculatedandthecystsappearnecrotic(Figure
7).
Discussion
Controlcystsobservedinthisstudyagreewiththedescriptionsofcystsintheliterature(2,6,8,12).Thecystwallis0.30.5mthickasreportedearlierandisclosely
appliedtotheoutermembraneofthenarrowbrandofcytoplasmsurroundingtheorganism.Theinnerandoutermembranesofthenarrowbandofcytoplasmare
identicalinthicknesstotheplasmamembrane.Thedensevesicles,whichappeartobebuddingofffromtheorganism,maybeinvolvedinsecretionofelementsofthe
cystswall(3).
Chlorineappearstoattackthecystwall,slowlydegradingit,sincethewallchangesitsstainingpropertieswithheavymetalsafterlongerexposuretochlorine.
However,whenthecystwalldoesstain,itisnotalwaystotallyorevenlydegradedand/ordissolvedevenwhenthecystsappearnecrotic.Therefore,chlorinemust
alsopenetratethecystwallandattackcellularcomponents.Thisisevidencedbytheflocculation(granulation)ofthecytoplasminterspersedwithclearareas,shrinkage
ofthecystwithinthecystwallandappearanceofcellulardebrisinformsofmembranousvesiclesintheenlargedlacunarspace.Theseclearareasareprobablydueto
chlorinedegradationofcytoplasmsinceleachingdidnotoccurincontrolcystsandallsampleswerepreparedidenticallyforelectronmicroscopy.Thedegreeof
flocculationvariesfromsmallareastothewholecytoplasm.Morphologically,nucleiandflagellaappearmostresistanttochlorine.Incystswhichweretreatedwith10
mg/Lchlorinewithviabilityof0%therearesomethatappear"normal",thatistheydonotexhibitcytoplasmicflocculation,butthecellmembraneappearsruffled,and
vacuolespersist.Thiscellmembranerufflingisprobablyduetotheacidtreatmentduringtheinvitroexcystationprocedureandnottochlorineexposure.However,
sincenoneofthecystsinthislattertreatmentareviable,a"normal"appearanceofthecytoplasmdoesnotnecessarilymeanthatacellisviable.Chlorinedamagemust
thereforeoccuratthemolecularlevelevenbeforethisismanifestedinmorphologicalchanges.
Acknowledgements
WethankR.HarrisandA.Oatwayfortheirtechnicalassistanceinelectronmicroscopy.
LiteratureCited
1.DeWalle,F.P.andC.R.ErlandJansson.1983.InactivationofGiardiabychlorineandUV.UnpublishedreporttoParksCanada.p.24.
2.Feely,D.E.,Erlandsen,S.L.andD.G.Chase.1984.Structureofthetrophozoiteandcyst.In:GiardiaandGiardiasis,Biology,Pathogenesisand
Epidemiology.S.L.ErlandsenandE.A.Meyer(eds).Plenum,NewYork.pp.330.
3.Friend,D.S.1966.ThefinestructureofGiardiamuris.J.CellBiol.29:317332.
4.Jarroll,E.L.,Bingham,A.K.andE.A.Meyer.1980.Giardiacystdestruction:effectivenessofsixsmallquantitywaterdisinfectionmethods.Am.J.Trop.Med.
Hyg.29:811.
5.Jarroll,E.L.,Bingham,A.K.andE.A.Meyer.1981.EffectofchlorineonGiardiacystviability.Appl.Environ.Microbiol.41:483487.
6.Luchtel,D.L.,Lawrence,W.P.andF.B.DeWalle.1980.ElectronmicroscopyofGiardialambliacysts.Appl.Environ.Microbiol.40:821832.
7.Meyer,E.A.1981.EffectofhalogensonGiardiacystviabilityUSEPAreport.600/281174.
8.Nemanic,P.S.,Owen,R.L.,Stevens,D.P.andJ.C.Mueller.1979.Ultrastructuralobservationsongiardiasisinamousemodel.II.Endosymbiosisandorganelle
distributioninGiardiamurisandGiardialamblia.J.Infect.Dis.140:222228.
9.Reynolds,E.S.1963.TheuseofleadcitrateathighpHasanelectronopaquestaininelectronmiroscopy.J.CellBiol.17:208211.
10.Rice,E.W.,Hoff,J.C.andF.W.SchaeferIII.1982.InactivationofGiardiacystsbychlorine.Appl.Environ.Mirobiol.43:250251.
11.Rice,E.W.andF.W.SchaeferIII.1981.ImprovedinvitroexcystationprocedureforGiardialambliacysts.J.Clin.Microbiol.14:709710.
12.Sheffield,H.G.andB.Bjorvatn.1977.UltrastructureofthecystofGiardialamblia.AmJ.Trop.med.Hyg.26:2330.
13.Spurr,A.R.1969.Alowviscosityepoxyresinembeddingmediumforelectronmicroscopy.J.Ultrastruct.Res.26:3143.
14.Wallis,P.M.andH.M.Wallis.1986.ExcystationandculturingofhumanandanimalGiardiasppbyusinggerbilsandTYIS33medium.Appl.Environ.
Microbiol.51:647651.

Page137

RemovalandInactivationofGiardiaCystsinaMobileWaterTreatmentPlantunderFieldConditions:PreliminaryResults
P.M.Wallis*,J.S.Davies,R.Nutbrown,J.M.BuchananMappin,P.D.RoachandA.vanRoodselaar
KananaskisCentreforEnvironmentalResearch,UniversityofCalgary,2500UniversityDriveN.W.,Calgary,Alberta,CanadaT2N1N4.
ThispaperreportsthepreliminaryresultsfromacooperativeprojectbetweentheAlbertaEnvironmentalCentre(Engineering)andtheUniversityofCalgary(Microbiology).The
majorobjectiveofthisphaseoftheprojectwastoevaluatetheeffectivenessofvariousoxidantsintheinactivationofGiardiacystsandtheefficiencyofcystremovalusingthe
filtrationprocess.AmobilewatertreatmentplantwasusedtoconductpreliminaryexperimentsatVegrevilleusingmunicipalwaterandsubsequentworktookplaceunderfield
conditionsattheKananaskisFieldStationintheEastSlopesoftheRockyMountains.Thispaperreportsi)thedevelopmentofthemethodologynecessarytoproduce,recover,and
evaluatetheviabilityofcystsinquantitieslargeenoughforexperimentationinapilotplantataflowrateof150L/min,andii)preliminaryresultsoffiltrationtrialswithand
withoutfilteraidsandoftheinactivationofastrainofGiardiamurisusingchlorine.Inthefirstexperiment,Giardiamuriscystswerecontinuouslyinjectedintothefeedwaterandand
recoveredafterpassagethrougheachofthetwoclarifiersandthefilter.Theresultsindicatedthatcystsdecreasedinbothnumbersandviabilityovertheinjectionperiod(6h).
Reductionsincystnumbersandviabilitywerecausedbyexcessiveagitationofthestocksolution,thepresenceofchloramineinthefeedwater,andadherencetosurfacesinsidethe
clarifiersandpipes.TwosetsofexperimentswerecarriedoutunderfieldconditionsusingBarrierLakewater(averagetemperature3C,pH8.2)tosupplythemobilewater
treatmentplant.Thesand/anthracitefiltrationstudiesshowedthatcystremovalefficiencywasafunctionofmaturationtimeuntilthefilterhadbeenoperatedfor30h.Afterthis
time,98%ofseededcystswereconsistentlyremoveduntilbackflushingwasrequired.Cystsweredetectedinthebackflushwater.Afinalexperimentusingpolymericfilteraid
showedthatcystremovalefficiencycouldberaisedto99%evenbeforethematurationtimehadelapsed.Thedisinfectionstudiesshowedthat99%ofGiardiamuriscystscouldbe
inactivatedusingatotalchlorineconcentrationof0.5mg/Landacontacttimeof30minutes.AsingletrialusingG.duodenaliscystsofhumanorigingavesimilarresults.Ata
concentrationof0.9mg/Lofchlorine,94.76.1%and995.3%ofthesehumaninfectivecystswereinactivatedafter30and90minutesofcontacttime(CT=26and77)
respectively.AnimalinfectiondataindicatedthatCTvaluesof35and77+weremoreappropriateforG.murisandG.duodenalisrespectively.Theresultsoffurtherexperimentswill
bereportedinseparatepublications(14,15).

Introduction
WaterbornediseaseisnowbecomingmorecommonintheUnitedStates(2,7),reversingadownwardtrendthatpersistedformanyyears.Thesituationissimilarin
Canadaarecentreport(13)identifiedsixwaterborneoutbreaksofgiardiasisthathaveoccurredinAlbertaandBritishColumbiasince1982.Inresponsetothese
problemsinWesternCanada,theUniversityofCalgaryandtheAlbertaEnvironmentalCentrehavejointlyestablishedaresearchprogramtoinvestigatethetreatment
ofdrinkingwaterfortheinactivationandremovalofGiardiacysts.TheCityofEdmonton,whichexperiencedanoutbreakofgiardiasisin1983,hasbeenactively
involvedwiththisresearchthroughtheprovisionoffiltermediaandoperationalinformationdesignedtosimulatetheirwatertreatmentprocesses.
Theoccurrenceofwaterborneoutbreaksofgiardiasisinmanycommunitiesisslowlyrevolutionizingthewatertreatmentindustry.Watertreatmentengineersare
questioningtheeffectivenessoftheirexistingproceduresandnewtechnologiesfortheinactivationandremovalofGiardiacystsarebeingexamined.Thereisa
demandforoperationaldatathatwillpermittreatmentplantstofunctionatmaximumefficiency.Theresultsofthisstudyprovidedatarelevanttothetreatmentofcold,
alkalinewaterthatistypicallyfoundinCanada.
Threesetsofexperimentsaredescribedinthispaper.Inthefirst,cystswereinjectedintothepilotplantwhenitwasoperatingwithmunicipalwaterinVegreville
containingchloramine.Theobjectofthispreliminaryexperimentwastotestthemethodologyforcystinjection,recovery,andviabilitytesting.Thesecondseriesof
experiments
*Correspondingauthor.

Page138
TABLE1.Operatingspecificationsforeachclarifier.
PotableOutput

RiseRateUnder
Troughs

RetentionTime
toOverflow

L/min

IGPM

L/min/m2

GPM/ft2

(min)

45.50

10.0

14.70

0.30

250.0

91.00

20.0

29.40

0.60

125.0

113.75

25.0

36.75

0.75

100.0

136.50

30.0

44.10

0.90

83.3

182.00

40.0

58.80

1.20

62.5

227.50

50.0

73.50

1.56

50.0

testedtheeffectoffiltermaturationandconditioningontheremovalofcystsandthethirdserieswasconcernedwiththeinactivationofG.muriscystsusingchlorine.
MaterialsandMethods
DevelopmentoftheMobileWaterTreatmentPlant
Toencompassvariousapplications,thepilotplantwasdesignedtoprovideoptimumflexibility.Theincorporationofalternativetechnologiesintoafacilitybasedon
conventionaltreatment(Class4)providedtheabilitytosubjectsourcewaterstoawidevarietyoftreatmentoptions.Thisapproachlendsthecapabilityofrapid
configurationofthepilotplanttomeetspecificrequirements.
Acomparisonofinitialcapitalcostsshowedonlyasmalldifferencebetween136L/min(30gpm)andthe227L/min(50gpm)capacitieswiththeoverallweightand
dimensionsincreasingslightly.A341L/min(75gpm)plant,however,wassignificantlymoreexpensiveandtheoverallweightofthecomponentsnearlydoubled.A
capacitysmallerthan100L/minwouldhaverequiredcustombuiltcomponents,makingitscostssimilartothatofa227L/min(50gpm)plant.Also,thissmallaplant
wouldhaveaconsiderablescaleupfactortofullmunicipalsizing.Consequently,thesystemwastailoredaroundanominalcapacityof227L/min(50gpm).
Withtheselectionofthe227L/min(50gpm)plantcapacity,theprospectiveshelterforthecomponentswasdesignedtoprovidetheoptimumaccommodationforthe
selectedwatertreatmentplant.Thisplantconsistsof2reactorclarifiers,2filtersandacollectionreservoir,whoseheightsnormallyrunabout2.44m(8')to3.05m
(10').Thesizeoftheshelterwhichprovedmostfeasiblewas3.66m(12')wideby15.85m(52')longwithanoverallmaximumouterheightof4.75m(15'7").There
isa0.305m(1')overhangonbothsidesofthelowboy,butthisdoesnotcauseanyinstabilitysincethesignificantloadingswerelocatedalongthetrailer'scentreline.
Thestaticloadoftheplantwhenfullofwateris45.4tonnesandthedynamicloadis15.0tonnes.Toensurecompatibilitywithmunicipalplants,specificationswere
preparedforarangeofflowconditionsforbothclarifiers(Table1)andfilters(Table2).ThefinallayoutoftheplantisgiveninFigure1.
TABLE2.Operatingspecificationsforeachfilter.
PotableOutput

OperationRate
2

L/min

IGPM

L/min/m

GPM/ft2

45.50

10.0

32.81

0.67

91.00

20.0

65.13

1.33

113.75

25.0

97.94

2.00

136.50

30.0

130.75

2.67

182.00

40.0

163.07

3.33

225.50

50.0

195.88

4.00

Figure1.
Layoutofmobilewatertreatmentplant.

Versatilityinwatertreatmentwasincreasedbytheinsertionofextensiveinterconnectingpipingandflowcontrolvalves.Theplantcontainstworeactorclarifiersand
twofilters,sointerconnectingplumbingallowstheclarifiersandfilterstooperateinseries,inparallel,orasindependentclarifierfiltersystemsoperatingsidebysideat
differentchemicaldosages/flowrates(Figure2).Clarificationtimesandchemicaldosagescanthereforebemodifiedtosuitdifferentwaterinletsources.Thishighly
adaptableunitactuallysimulatesmanydifferentwatertreatmentplantsbutatafractionofthecost.
Easeofoperationwasincreasedwiththeadditionofamicroprocessorwhichremotelycontrolsandmonitorstheoutputparameters.Thispermittedtheoptionof
runningtheplantunmannedforextendedperiodsoftimeprovidingthatthechemicaladditivesareprovidedfor.A36to48channelinputmicroprocessorwhichwould
monitorandrecordsignalsfrompumps,thermocouples,turbidimeters,colourimeters,hardnesstesters,pHmeters,flowmeters,chlorinemeters,chlorinedioxide,
conductivitymeters,andchemicalfeedequipmentwasthereforebuiltin.
FieldSiteandWaterConditions
ThemobilewatertreatmentplantwasmovedtotheKananaskisFieldStationatBarrierLakeinKananaskisCountry(80kmwestofCalgary,AB,Canada).Thepilot
plantwassituated2kmfromthemicrobiologicallaboratoryattheFieldStation.Thesiteaffordedeasyaccesstoanabundantsupplyofrawwater(BarrierLake,
Table3)andsewagetreatmentlagoonsfordisposalofwatercontaminatedwithGiardiacysts.Whennocystswerepresentintheplanteffluent,waterwas
dischargedtotheKananaskisRiver.Allcontaminatedeffluentwassuperchlorinatedatachlorineconcentrationof15mg/L.
BarrierLakeisashallowreservoirfedbytheKananaskisRiver.Duringourexperiments,thewaterwashard(139165mg/LasCaCO3)andtheaverageturbidity
duringtheseexperimentswas3.60.8NTU.Thetemperaturewas7.0atthebeginningoftheexperimentsinSeptemberanddroppedto2.5CinlateNovember
(mean4.90.7).ThepHofBarrierLakerangedfrom8.1to8.5inthesametimeperiod.
SourcesofGiardiaCysts
WeobtainedastrainofGiardiamuris(GM1)fromG.FaubertattheInstituteofParasitologyatMcGillUniversityandmaintaineditinvivousingoutbredSwiss
Webstermice[Crl:CFW(SW)BR].Ahumanstrain(H8)wasobtainedfromthestoolofasymptomaticmalechildinCanmore,ABandmaintainedbothinculture(12)
andinvivousingthemongoliangerbil(Merionesunguiculatus)modelofBelosevicetal.(1).
CystProduction
Cystswereproducedbyinfecting30ormoremice(forGM1)orgerbils(forH8)bygavage.Sixtoeightdaysafterinfection,animalswereplacedonawiremesh
suspendedoverashallowlayerofwaterinalargecageandfaeceswerecollectedovernight.Cystproductionwasenhancedbyaddingdexamethasone(Schering)ata
concentrationof40g/mLtotheirdrinkingwater.Thisresultedinuptoa10fold

Page139

Figure2.
Flowschematicsformobilewatertreatmentplant.

increaseincystproduction.Micewerefoundtobemoretolerantofdexamethasonethangerbilsbutbothanimalswerefoundtobeverysensitivetothedrugandthe
useofdexamthasoneathigherconcentrationsisnotrecommended.FaeceswerehomogenizedusingaBraunkitchenhandblenderandcarefullylayeredover1.0M
sucroseincentrifugetubes(50mLcapacity).Thetubeswerecentrifugedfor10minutesat2500rpminaSorvallRC5refrigeratedcentrifuge.Aftercentrifugation,the
interfacebetweenthesucroseandtheremainingliquidwassiphonedoffusinganOxfordMacroSetpipettewhosetiphadbeencutofftoprovidealargerorifice.This
crudefaecalisolatewasdilutedto1500mLandstirredfor30minutesusingamagneticstirrerafterwhichthecystswerecountedusingahaemocytometer.Atpeak
production,itwaspossibletoisolateupto109cystsfrom30miceusingthesemethods.
ViabilityTesting
InitialexperimentationwithtrypanbluedyeexclusionandtheGomorriacidphosphatasemethodsresultedintheconclusionthatthesemethodswereunreliable.Thein
vitroexcystationmethodofRiceandSchaefer(9)wasthereforeusedforallofourviabilitytests.Theprocedureusedwasessentiallythesameasthatdescribedby
RiceandSchaeferexceptthatlyophilizedtrypsin(Gibco)wassubstituted.Wecountedanycystthatshowedinternalmovement,protoplasmextrusion,orflagellar
motionasa"partiallyexcystedtrophozoite"veryfewfullyexcystedtrophozoiteswereobserved.Stockcystsolutionswerealwayspreparedonthesamedayasthe
experimentandwerealways95+%viable.
Viabilitywasalsotestedbyanimalinfection.SwissWebstermicewereusedtotestG.muriscystsandgerbilswereusedtotestG.duodenaliscysts.Animalswere
infectedbygavagewithapproximately2500cystsrecoveredusingmetrizamideasdescribedbelow.Thepresenceofinfectionwastestedbyfaecalexamination.
Animalswhichdidnotpasscysts8daysafterinfectionweresacrificedandthegutwasexaminedforthepresenceoftrophozoitesbythreebiopsysnipstakenfromthe
duodenum,jejunum,andileumrespectively.
CystRecoverybyMembraneFiltration
Cystswererecoveredbytappingoffwaterfromthetreatmenttrainandpumpingitthrough5mNucleporemembranefiltersin142mmplexiglasshousings.Peristaltic
pumpssettopumpattherateof1L/minwereused.Attheendofthecollectionperiod,thefiltersweredisassembledandwashedoffusingafinestreamoftapwater
containingTritonX100and1%sodiumthiosulphateintoasamplebottle.Thewashingsweretakenimmediatelytothelaboratory,concentratedbycentrifugation,and
countedusingahaemocytometer.InordertotesttheviabilityofcyststhatwererecoveredfromwaterusingNucleporefilters(describedbelow),itwasnecessaryto
removebackgroundsiltandalgae.Thiswasaccomplishedbyclarifyingthecentrateovermetrizamide,ahighdensitylowviscosityinertmediumthatcanbemade
isotonicwithhumanserum(8).Theconcentratedwatersuspensionwaslayeredoverthemetrizamidesolutionina15mLcentrifugetubeandcentrifugedfor5minutes
at500g.Thecystswerethenaspiratedfromtheinterfaceusingaglasspipetteandwashed.
Inordertoquantifytheeffectsofmetrizamideoncystrecoveryefficiencyandviability,thefollowingexperimentwasperformed.Giardiamuriscystswereaddedto
100Lofuntreatedstreamwateratconcentrationsof50,100,500,and1000cystsperL.Usingcontinuousagitation,two50Lbatcheswerefilteredthrough5m
Nucleporemembranes.Therecoveredmaterialwasthenclarifiedbycentrifugationovermetrizamideandthepurifiedcystswereexcystedaccordingtothemethodof
RiceandSchaefer(9).Theresultswerepooledandrecoveryefficiencywasfoundtoaverage31.311%.Theviabilityoftherecoveredcystswas8310%.Ina
separateexperimentwhereknownnumbersofcystswerelayeredovermetrizamidewithoutanyfiltration,therecoveryefficiencyoftheclarificationstepwasfoundto
be679%.Previouswork(11)demonstratedthattherecoveryefficiencyofthefiltrationprocedurealonewas7113%.
TABLE3.Barrierlakewaterchemistry.
Parameter

mg/L(unlessotherwisenoted)

pH

7.888.65*

Conductivity

292346*

Alkalinity

127133*

Hardness

139165*

Calcium

40.645.0

Magnesium

9.113.1

Sodium

<11.2

Potassium

0.330.40

Iron

<0.010.04

Chloride

<11.5

Sulphate

27.930.0

Fluoride
Nitrate&Nitrite
Silica
Totaldissolvedsolids(calc)
Totalphosphorus
TotalKjeldahlNitrogen

0.060.16
0.0240.055*
3.103.27
156170
0.0040.006
0.180.30

*pHisinpHunitsconductivitynotedinmicrosiemens/cmAlkalinityandhardnessexpressedascalciumcarbonateNitrate
andNitriteexpressedasN.

Page140
TABLE4.Resultsfromthefirstpreliminarytrial.
Sample

ElapsedTime
(min)

Volume
Filt.(L)

Recovery
(%ofcystsadded)

Viability
(%)

A1

150

54.2

58

89

A2

225

54.2

84

95

B1

330

55.0

36

93

B2

410

53.8

44

90

C1

355

53.2

N.D.

C2

440

36.5

N.D.

StockCystSolution:278cysts/mL
InjectionRate:225mL/min
InitialViability:98%
NotDeterminedbecauseofinsufficientnumbersofcysts.

PreliminaryInjectionandRecoveryTrial
Theobjectofthisdemonstrationwastoshowthatcystscouldbeproducedinadequatenumbers,addedtotherawwaterfeed,quantitativelyrecovered,andtested
forviability.Twotrialswererunin1985usingG.muris(GM1)cysts.Inbothexperiments,aconcentratedsolutionofcystswasslowlyaddedtotherapidmix
chamberupstreamofClarifier1over6hoursonthefirstoccasionand8hoursonthesecond.ThefeedwaterwasVegrevillemunicipalwaterdrawnfromahydrant
onthegroundsoftheAlbertaEnvironmentalCentre.Thiswatercontainedapproximately1.5mg/Lchloramine.Theflowrateofwaterthroughtheplantwas150
L/minforthefirsttrialand101L/minforthesecond.
Theconcentratedcystsolutionwascontinuouslyagitatedbyamagneticallydrivenimpellerpumpthroughouttheexperimentandinjectedintotherapidmixchamber
usingaperistalticpump.Astrontiumchloridesolutionwassimultaneouslyinjectedfromaseparatecontainerbyperistalticpumpinordertoprovideatracer.Cysts
wererecoveredseveraltimesdownstreamofClarifier1(samplesiteA),downstreamofClarifier2(samplesiteB),anddownstreamofthesand/anthracitefilter
(samplesiteC).Therecoveredcystswerecountedandviabilitywastestedbyexcystation.Samplesweretakenateachofthethreepointsusing5mNuclepore
filters.Countsweremadeonsamplesthatwerecentrifugeconcentratedonly.Viabilitytestswereperformedonsamplesafterclarificationwithmetrizamideas
describedabove.
FilterTrials
Thesand/anthracitefilterwasinitiallybackflushedandthenoperatedat150L/minintermittentlyoveraperiodof17daysuntilthefilterwastoocloggedtomaintainthis
flowrate.Atotalofapproximately760,000LofBarrierLakewaterwasfiltered.Cystremovalefficiencywastestedafter7,17.5,29,31,51,and85hoursof
operation.Attheendoftherun,thefilterwasbackflushedandsampleswerecollectedandanalyzedforcysts.Afinaltrialwasconductedwhentherejuvenatedfilter
hadbeenoperatingfor18hours.Thisexperimentwasconductedinthesamemannerastheprecedingtrialsbutpolymerfilteraidwasaddedtotherapidmixchamber
ofClarifier1.
StrontiumtracerandvaryingnumbersofGM1cysts(6to140106)weresimultaneouslyinjectedintothewatertrainupstreamofthefilterinaspikelasting60sec.
Sequentialsampleswerecollecteddownstreamofthefilterfor30minutesinfiveseparatesequentialfractions.Thestrontiumconcentrationwasmeasuredinthefeed
stock,thefiltrate,andinsamplestakencontinuouslyfromthepostfiltereffluent.Thisallowedapredictionofthenumberofcyststhatshouldhavebeenpresentineach
6minuteportionoftheinjectedspikeifnofiltrationhadtakenplace.Thecystsrecoveredwerecountedineachfraction,summedandcomparedwiththepredicted
numbersusingthemethodofWallisandBuchananMappin(11).Alloftherecoverieswerecorrectedbydividingthenumberofcystsby0.7toallowforcystloss
duringthefiltrationprocedure(seeaboveand11).Whenthefilterwasbackflushed,samplesweretakenofthesurfacescumandofthebackflushwater.
DisinfectionExperiments
DisinfectiontrialswerecarriedoutwithGiardiamuris(GM1)cystsandprechlorinationwithgaseouschlorine.Totalandfreechlorineconcentrationswerecontinually
monitoredwhilethetreatmentplantwasinoperation.Thelossoffreechlorineaswaterflowedthroughtheclarifierswascalculatedinordertoquantifythechlorine
demandofthepilotplant.Chlorinewasaddedtothetreatmenttrainbetweenexperimentstoensurethatthechlorinedemandoftheclarifiersandpipingwaskepttoa
minimum.
InitialstudiesshowedthattheconcentrationoftracerpeakeddownstreamofClarifier130minutesaftertheadditionofaspike.Attenuationofthetracer
concentrationoccurreddownstreamofClarifier2butaplateauconcentrationwasreachedafter90minutes.Cystswerethereforeinjectedover15minutesintothe
rapidmixchamberaboveClarifier1andrecoveredbypumpingwaterthroughfourNucleporefiltersatoncefromrecoverypointspostClarifier1(samplesiteA)and
postClarifier2(samplesiteB),after30and90minutesequivalenttimerespectively.Inthisway,fourreplicatesampleswerecollectedforviabilitytestingatsampling
pointsAandB.Viabilitywasassayedusingbothinvitroexcystationandanimalinfection.
Results
PreliminaryInjectionandRecoveryTrials
TheresultsofthefirstexperimentarereportedinTable4.Thenumberofcystsinjectedturnedouttoberatherlowandtherecoveryrateswerevariable.Viability,
howeverremainedhighthroughouttheexperiment.VeryfewcystswererecoveredatsiteCbecausemostofthecystswereremovedbythefilter.Itwasdecidedto
repeatthetrialwithmorereplicatesamplesateachsiteandtodropsamplingsiteC.
TheresultsofthesecondexperimentarereportedinTable5andFigure3.Ahigherconcentrationofcystswasinjectedbutitwasfoundthatcystrecoverydropped
steadilythroughouttheexperiment.Thestocksolutionwascheckedattheendoftheexperimentanditwas
TABLE5.Resultsfromthesecondpreliminarytrial.
ElapsedTime
(min)

Volume
Filt.(L)

Recovery
(%ofcystsadded)

Viability
(%)

A1

149

52.3

58

94

A2

230

52.1

22

88

A3

305

52.1

18

70

A4

373

52.5

13

59

A5

450

53.1

12

N.D.

B1

175

53.5

44

94

B2

234

52.1

24

88

B3

305

54.4

25

71

B4

364

53.8

22

52

B5

435

52.3

11

N.D.

Sample

StockCystSolution:1302cysts/mL
InjectionRate:92mL/min
InitialViability:96%
FinalViability:54%
NotDeterminedbecauseofinsufficientnumbersofcysts.

Page141

Figure3.
RecoveryandviabilityofGM1cysts
fromthesecondVegrevilletrial.

foundthattheconcentrationofcystshaddroppedfrom1302to810cysts/mL.Atthesametimeviabilitydroppedfrom94toapproximately55%.Examinationofthe
systemdisclosedthatthepumpusedtokeeptheconcentratedcystsolutionagitatedwasresponsibleforcystlossprobablybecausethepumpingactionwastooharsh.
Thisproblemwascompoundedbythepresenceofchlorinecompoundsinthefeedwaterandthelackoftemperaturecontrolintheconcentratedstocksolution.It
wasconcludedthatimpellorpumpsshouldnotbeusedforagitationandapaddlestirrerwasusedforallsubsequentexperiments.Theresultsoftheseexperimentsalso
suggestedthatonlyunchlorinatedwatershouldbeusedtosuspendcystsdespitethehighchlorinedemandofthefaecalmaterialandthattheinjectiontimeshouldbe
keptasshortaspossible.
FiltrationTrials
TheresultsofthefiltrationtrialsarereportedinFigure4.Cystremovalbythefilterwasverylow(33%)initially(7hours),butreachedconsistentlyhighlevelsafter30
hoursofoperation.Themaximum
removalefficiencyobservedwas98.6%.Theexperimentwithpolymerfilteraidtookplaceafteronly18hoursofoperationinasubsequentrunandresultedina
removalefficiencyof99%.Cystswerefoundinthesurfacescumontopofthefilterbeforebackflushingandinthefirsttwosamplestakenofthebackflushwater.
Eightfurthersampleswerenegative.
DisinfectionTrials
ThecombinedchlorineleveloftreatedBarrierLakewaterwasfoundtoaverage0.05mg/Landthelossofchlorineinwaterflowingthroughtheclarifierswaslow
(Figure5).
TheexcystationresultsofthedisinfectionexperimentsareplottedinFigure6.VeryhighreductionsinGM1cystviabilitywereobservedatconcentrationsabove0.15
mg/Lwithcontacttimesofboth30and90minutes.Whentheexcystationdatawereplottedagainsttheproductofchlorineconcentrationandcontacttime(CT),a
similarpatternemerged(Figure6).Mortalityratesof99%orgreaterwereobservedatCTvaluesabove13basedonexcystationdata.Micebecameinfected,
however,withG.murisatCTlevelsupto35mg*min/L.
ThesingletrialemployingG.duodenaliscystsresultedinslightlyhigherviabilitylevels.Atafreechlorineconcentrationof0.85mg/L,5.3%ofH8cystswereviable
after30minutesand1.0%wereviableafter90minutesofcontacttime.Most(3/4)ofthegerbilsthathadbeeninoculatedwithcystsrecoveredpostclarifier1
(CT=26)

Figure4.
Filterremovalefficiency.

Page142

Figure5.
Chlorinelossinthemobilewatertreatmentplant
duringexperimentsatdifferentchlorineconcentrations.

becameinfectedbutonly1/4gerbilsbecameinfectedwithcystsrecoveredpostclarifier2(CT=77).
Discussion
FilterExperiments
Mostofthefiltrationexperimentswerecarriedoutwithoutusinganyformofcoagulant.Theuseofcoagulatingagentsisknowntoincreasecystremovalefficiencysoit
isofinterestthatremovalefficienciesashighas98%couldbeachievedwithoutit.However,itwasconcludedthatthecriticaltimeforfiltersintheabsenceoffilteraid
wasthefirst30hoursafterbackflushing,typicallythatperiodofnormalfilteroperation.Thistimeperiodisundoubtablyflexibledependingupontheturbidityandthe
compositionofthefiltermedia.Thisseriesofexperimentsclearlyshowsthevariabilityincystremovalthatcanbeencountereddependingonthematurationofthefilter
andtheneedforcarefuldisposaloffilterbackwashwaterwasdemonstrated.
Oneofthedrawbackstoourexperimentswasthatthefilterwasoperatedintermittentlyoveraperiodof17days.Thisdoesnotcorrespondtonormaltreatmentplant
procedureanditispossiblethattherewassomedevelopmentofbiologicalmaterialinthefilterthatenhancedtheremovalofGiardiacysts.Futureexperimentswillbe
carriedoutunderconditionsofcontinuousoperation.
DisinfectionExperiments
TheresultsreportedabovesuggestthatasignificantproportionofGM1cysts(G.muris)canbekilledinawatertreatmentplantatchlorineconcentrationsof
approximately0.5mg/Landcontacttimesaslowas30minutes.TheresultsofoneadditionaltrialusinghumanGiardiacystssuggestedthatthisstrainwasmore
resistant.
Jarrolletal.(5)reportedthat99.8%ofG.lambliacystscouldbekilledat3and20Cincloudyriverwaterwithfreechlorineconcentrationsof6to8mg/L.Ina
laterpaperJarrolletal.(6)foundthat99.9%ofG.lambliacystscouldbekilledataconcentrationof2mg/Lafter60minutes(pH8,5C)butthatconcentrations
upto8mg/Lwererequiredtoachievethesameeffectafter30minutes.DeWalleandJansson(3)testedtheviabilityofG.muriscystsinBanffmunicipalwater(pH
7.8,1C)andreportedthatmortalitylevelsinexcessof99%werepossibleatconcentrationsabove2mg/Landcontacttimesof90min.Highmortalitylevelscould
beachievedatconcentrationsoflessthan1.0mg/Lbutonlyaftercontacttimesof15h.Riceetal.(10)foundthatG.muriswasslightlymoreresistantthanG.
lambliabutontheaverageonlyabout95%ofcystscouldbekilledafter60minutesat2.5mg/Lchlorine(pH8).Faubertetal.(4)testedtheviabilityofG.muris
(thesamestrainusedinthisstudy,GM1)byinfectingCD1micewithcyststhathadbeenexposedtoafreechlorineconcentrationof0.59mg/Landfoundthat
animalscouldstillbeinfectedafter60minutesofcontacttime(CT=35).Theseresultsshowedthatatleastsomeofthecystswerestillviable.Theseauthorsattributed
thisviabilityleveltothehighpHofthetapwater(pH=8.24,temp.=6C,turbidity=3.1NTU).ThewaterchemistryofBarrierLakewaterissimilar.

Figure6.
Viability(%excystation)asafunctionoftheproductof
chlorineconcentrationandtime(CT)forGM1cysts.

Page143

ResultsreportedintheliteratureforbenchscaleexperimentsindicatethatfortheconditionsexistinginBarrierLake(pH8,23C),concentrationsofapproximately4
to5mg/Lwouldberequiredtokill99%ofG.murisorG.lamblia(duodenalis)cystswith30to90minutesofcontacttime(CT=50to360)basedonexcystation.
Basedonexcystationdata,ourresultsshowthat99%mortalitycouldbeobtainedataCTofapproximately13forG.murisand77forG.duodenalisundersimilar
conditions.AnimalinfectiondatasuggestthatCTvaluesof35and77+aremorereasonableforG.murisandG.duodenalisrespectively.Itmustberemembered,
however,thatanimalswereinfectedwithaminimumof2500cystswhichisarelativelylargeinoculum.Weconcludethatchlorinewasmoreeffectiveinthepilotplant
thandatafromtheliteraturewouldhavepredicted.ItisalsopossiblethatthestrainsofGiardiaweusedwerelessresistanttochlorinethanthoseusedbyother
investigators.Thispossiblilitywillbeinvestigatedinfutureexperiments.
Theresultsfromtheliteraturesummarizedabovewereperformedonthebenchtopinvolumesofwaterrangingfrom200to1000mL.Inmostcases,thecystswere
suspendedinphosphatebuffersolutionwithlittleornoagitation.Thecystswerecollectedwithaminimumofdisturbanceandtestedforviabilityimmediately.These
conditionscouldthereforebesaidtobeoptimumforcystsurvival.Experimentsperformedinthemobilewatertreatmentplant,however,werequitedifferent.The
volumesofwaterusedwere2to3ordersofmagnitudegreater,werecontinuallymixedandcirculated,andwererecoveredbymembranefiltrationbefore
transportationtothelabforviabilitytesting.Thewaterusedwasnotbufferedandcontainedallofthebackgroundcontaminationtypicallyfoundinraw(butgood
quality)feedstockwater.Itispossible,therefore,thatthesefactorsmadecystsmoresusceptibletochlorine.Inthecontrolexperiment(nochlorine),cystviability
droppedfrom88%inthestocksolutionto78%after30min.and75%after90min.(rawdata)suggestingthatthetreatmentplantitselfwascontributingtothe
mortalityofcystsevenintheabsenceofchlorine.Theequipmentandproceduresusedinthemobilewatertreatmentplantwereverysimilartothoseusedroutinelyin
watertreatmentplantsthroughoutAlberta.
Conclusions
1.Methodologywasdevelopedpermittingtheproduction,quantitativerecovery,andviabilitytestingoflargenumbersofdifferentstrainsofGiardiacystsforwater
treatmentexperiments.
2.Filtrationefficiencyexperimentsdemonstratedthatfilterconditionisveryimportanttocystremoval.Thesand/anthracitefilterwasabletoremoveupto98%of
seededGiardiamuris(GM1)cystsbutonlyafterapproximately30hofoperation.
3.Basedonasinglefiltrationtrial,theuseofapolymerfilteraidincreasedthe(GM1)cystremovalefficiencyto99%eventhoughthefilterhadnotfullymatured.
4.FilterbackwashwatercontainedGiardia(GM1)cystsindicatingthatcautionshouldbeemployedwhendisposingofsuchwastewater.
5.Mortalityratesof99%wereachievedwiththeGiardiamuris(GM1)cystsusedintheseexperimentsatCTvaluesaslowas13basedonexcystationdataand35
basedonanimalinfection.
6.Basedonasingletrial,humanGiardiacystsrespondedtochlorineinamannersimilartoGiardiamuriscysts.Ataconcentrationof0.85mg/Lfreechlorine,94.7
6.1%and995.3%ofH8cystswerekilledatcontacttimesof30and90minutesrespectively.ThiscorrespondstoaCTof77for99%mortalitybasedon
excystationdatabut1/4gerbilsstillbecameinfected.
7.Theresultsfromtheseexperimentshaveshownthatthemobilewatertreatmentplantisausefultoolfordevelopingwatertreatmentprotocolsforremovingand
inactivatingGiardiacystsatthepilotplantscale.Someoftheexperimentshaveyieldedresultsthatwerenotpredictedbypreviousbenchscalestudiesbutthe
investigatorsbelievethattheresultspresentedinthisreportaremoreapplicabletofullscalewatertreatmentplantsinAlberta.
LiteratureCited
1.Belosevic,M.,Faubert,G.M.,MacLean,J.D.,Law,C.,andN.A.Croll.1983.G.lambliainfectionsinMongoliangerbils:ananimalmodel.J.Inf.Dis.147:222
226.
2.Craun,G.F.1984.Waterborneoutbreaksofgiardiasis:CurrentStatus.in:Erlandsen,S.L.andE.A.Meyer(eds.).GiardiaandGiardiasis,PlenumPress,NY.
3.DeWalle,F.P.andC.R.ErlandJansson.1983.InactivationoGiardiabychlorineandUV.UnpublishedreporttoParksCanada.24p.
4.Faubert,G.M.,Leziy,S.S.,Bourassa,A.andJ.D.MacLean.1986.Effectsofenvironmentalconditionsandstandardchlorinationpracticesontheinfectivityof
Giardiacysts.Dis.Aq.Org.2:15.
5.Jarroll,E.L.,A.K.Bingham,andE.A.Meyer.1980.Giardiacystdestruction:effectivenessofsixsmallquantitywaterdisinfectionmethods.Am.J.Trop.Med.
Hyg.29:811.
6.Jarroll,E.L.,A.K.Bingham,andE.A.Meyer.1981.EffectofchlorineonGiardialambliacystviability.Appl.Environ.Microbiol.41:483487.
7.Lippy,E.C.1981.Waterbornedisease:occurrenceisontheupswing.J.Am.Wat.Wks.Assoc.73:5762.
8.Loos,J.A.andD.Roos.1976.Densityanalysisasatoolforbloodseparation.in:Rickwood,D.(ed.)BiologicalSeparationsinIodinatedDensityGradient
Media,InformationRetrievalLtd.,London.p.100.
9.Rice,E.W.andF.W.SchaeferIII.1981.ImprovedinvitroexcystationprocedureforGiardialambliacysts.J.Clin.Microbiol.14:709710.
10.Rice,E.W.,J.C.Hoff,andF.W.SchaeferIII.1982.InactivationofGiardiacystsbychlorine.Appl.Environ.Microbiol.43:250251.

Page144

11.Wallis,P.M.andJ.M.BuchananMappin.1985.DetectionofGiardiacystsatlowconcentrationsinwaterusingNucleporemembranes.Wat.Res.19:331334.
12.Wallis,P.M.andH.M.Wallis.1986.ExcystationandculturingofhumanandanimalGiardiaspp.byusinggerbilsandTYIS33medium.Appl.Environ.
Microbiol.51:647651.
13.Wallis,P.M.1987.CriteriaforGiardiaindrinkingwater.unpublishedreporttoHealthandWelfareCanada.54pp.
14.Wallis,P.M.,Davies,J.S.,Nutbrown,J.M.,BuchananMappin,J.M.,Roach,P.D.,andA.vanRoodselaar.1988.RemovalandinactivationofGiardiacystsin
amobilewatertreatmentplantunderfieldconditions:preliminaryresults.inprep.
15.Wallis,P.M.,andA.vanRoodselaar.1988.RemovalofGiardiacystsbyfiltrationinapilotwatertreatmentplantunderfieldconditions.inprep.

Page145

DIFFERENTIATIONOFGIARDIAISOLATES

Page147

TheGenomeofGiardiaIntestinalis
PeterUpcroft*,PeterF.L.BorehamandJaquelineA.Upcroft
QueenslandInstituteofMedicalResearch,BramstonTerrace,HerstonBrisbane,Qld.4006,Australia
WehaveusedorthogonalfieldalternationandfieldinversiongelelectrophoresistoeffectseparationandidentificationofchromosomesfromGiardiaintestinalis.Mostchromosomes
areverylargecomparedtothoseofPlasmodiumfalciparum.However,thereisadiscretepopulationofminichromosomesapproximately100kb,similarinsizetothosepresentin
Trypanosomabrucei.Thesehavebeenisolated,analyzedbyrestrictionendonucleasecleavage,clonedinE.coliandprobedwithuniqueandrepetitiveDNAsegmentsfromcDNA
andgenomiclibraries.Thisstudyisanattempttoestablishabaseforgeneticanalysis,genemappingandstrainidentificationofG.intestinalis.

Introduction
AlthoughmembersofthegenusGiardiaaredistributedwidelythroughouttheworldinmanandotheranimals,thereisstillconsiderablecontroversyconcerning
speciationwithintheduodenalisgroup,whichincludesG.intestinalis.Morphometricanalysisdoesnotassistwiththeelucidationofthisproblem,eventhoughithas
provenusefulindistinguishingtheduodenalisgroupfromthemurisandagilisgroups(27).AtageneticlevelverylittleisknownabouttheGiardiagenusingeneral.
Lessisknownaboutgeneticdifferenceswhichmayconferphenotypicdistinctionbetweenisolatese.g.pathogenicity,althoughdifferencesinrestrictionenzyme
cleavagepatternsofDNAfromisolateshavebeenobserved(17).
Forallparasitegroupsofmedicalsignificance,theprincipledrawbackingeneticanalysishasbeenthelackofaconsistentmethodofchromosomeidentification.This
hasbeencircumventedinthecaseoftheVariantSurfaceGlycoproteinsoftrypanosomes,tosomeextent,bycloningofthegenesbyrecominantDNAtechniques(2).
However,thedevelopmentoftechniquestoseparate,andhenceidentify,chromosomesbygelelectrophoresis,hashadawiderimpactonparasitegenetics.This
involveslysingwholeparasiteswhichareembeddedinanagaroseblock,topreventshearingofthelargeDNAmolecules(19).ChromosomesizeDNAmolecules
thataresmallerthan3Mbpthencanbeseparatedbythreebasictypesofelectrodeandelectrophoreticconfigurations,pulsedfieldgel(PFG)eletrophoresis(19),
orthogonalfieldalternationgelelectrotrophoresis(OFAGE)(7)orfieldinversiongelelectrophoresis(FIGE)(8).Usingthesetechniqueschromosomesofyeast(8,19),
Trypanosomaspp.(11,12,23,24),Plasmodiumfalciparum(15,25),Leishmaniaspp.(10,23),Crithidiafasciculata(23),Herpetomonasmuscarum(23)and
Leptomonasctenocephali(23)havebeenseparatedeffectively.Thechromosomesvarydramaticallyinsizefrom50kbto>4Mbpandfromafew,tohundredsin
number.Evenwithinagenustherearegreatdifferences.ForexamplewithinthegenusTrypanosoma,T.cruzihasatleast20chromosomeslargerthan300kband
probablynonelargerthan4Mbp.T.bruceihassomeverylargechromosomeswhichdonotenterthegelcomplementedbyapproximately100minichromosomes50
150kbinlength.Moreover,differencesareobservedbetweenandwithinstrains(12,23).Thisvariationwithinstrainshasbeenobservedinmalariaalso(15,25).
Thesebasicchromosomeanalysesthereforehavebeenveryinformativebothfromageneticandataxonomicperspective.
WehaveusedOFAGEandFIGEtoattemptseparationandidentificationofchromosomesinanumberofG.intestinalisisolates
(i)tocomparethemwiththeotherdescribedprotozoa
(ii)todetermineiftherearesignificantstraindifferencesthatmaybeusedforspeciesandstrainidentification
(iii)toestablishabaseforgeneticanalysisandmapping.
ThelatterhasbeencoupledtostudieswithcDNAandgenomicDNAlibrarieswhichwehaveclonedinE.colifromG.intestinalisisolates.
MaterialsandMethods
G.intestinalisisolatesweremaintainedinaxeniccultureinmodifiedTYIS33mediumsupplementedwith1mg/mLbile(3).
DNAwasextractedfromwholeparasitesbasicallyasdescribedbyKetnerandKellyformammaliancells(16),followedbybandinginCsCl(1.60gm/mLCsCl,
Beckman75Tirotor,45Krpm,36h)toremovecontaminatingcarbohydrate.
ForOFAGEandFIGEwholeparasiteswerelysedinSeaplaqueagaroseblocks.Usually5108trophozoiteswerewashedtwiceinphosphatebufferedsaline(PBS)
afterculturing,andwereresuspendedin0.5mlPBS.0.5mLof2%Seaplaqueagarose(MarineColloids)insalinewasaddedandtheblockswereallowedtosetin
LabTek8wellchamberslides(MilesLaboratories).SliceswerecuttofitthewellsofaPharmaciaGNA200electrophoresisapparatus.InthecaseofP.
falciparumchromosomeblocks,5108parasiteswerecollectedfromsaponinlysedinfectedredbloodcells.TheywerewashedinPBSandsetinSeaplaqueas
above.Theparasiteswerelysedintheblocksaspreviouslydescribed(19,24).
Tolyseparasitesdirectlyintheloadingslotscastintheagarose,5107trophozoiteswerewashedtwiceinPBScontaining10%glycerol.Theywereloaded(30L)
over15Loflysisbuffer(5Leachof10%sarkosyl100mMTrisHCl/10mM
*Correspondingauthor.

Page148

EDTA,pH7.5PBSin30%glycerol).5LofproteinaseKataconcentrationof20mg/mLwasaddedalso.Theparasiteswereallowedtolyseatroomtemperature
for3hours.
cDNAandgenomiclibrarieswillbedescribedindetailelsewhere.Briefly,RNAwasextractedfromwholeparasites(28).cDNAwaspreparedbythemethodof
GublerandHoffman(14),theendsmadebluntwithmungbeannuclease(13)andligatedintoPUC19(28)cleavedwithSmaI.GenomicDNAwascleavedto
appropriatesizesbyDNaseIinthepresenceofMn2+(1,20)andclonedintopUC19.ThebacterialhostwasE.coliJM109(28).Restrictionendonucleaseswere
purchasedfromNewEnglandBiolabsandusedaccordingtothemanufacturer'sinstructions.
DNAsegmentswereextractedfromagarose(Seaplaque,MarineColloids)asdescribed(6)orbyelectroelection(InternationalBiotechnologies,Inc.,ModelUEA
electroelutor)accordingtothemanufacturer'sinstructions.
Results
Figure1showsanexampleofOFAGEusedtoseparateP.falciparumchromosomes,aftergentlelysisoftheparasitesinSeaplaqueagarose.InparallellanesG.
intestinalistrophozoiteswerelysedandsubjectedtothesameOFAGE.Onecanseeclearresolutionofmalariachromosomes.MostoftheGiardiaDNAisstillin
theloadingslot.SomestructurecanbeseenasfaintbandsinthelightsmearofDNAmigratingintothegel,butisnotparticularlyconvincingaschromosometype
structureswhencomparedwiththemalariachromosomes.Thechromosomesofthemalariastrainused(FC27)havebeen

Figure1.
G.intestinalisandP.falciparumchromosomessetinagarose
blocksandseparatedbyOFAGE.ParasiteDNAsetinblocks
wasseparatedina1%agarosegelinTBE(0.5x)recirculating
buffer.Pulsetimeswere45sec.Electrophoresiswascarried
outat300V,150mAfor12hat4C.Lane1containsP.falciparumK
Lane2P.falciparumFC27Lane3P.falciparumHB3
Lane4G.intestinalisBRIS/3/HEPU106/1/3.

wellcharacterizedpreviously(15)andtheOFAGEseparationhereisconsistentwiththepublishedPFGanalysis(15).Theconditionsofpreparationofbothmalaria
andGiardiachromosomesintheagaroseblockswereidentical,sothattheslightsmearofDNAintheGiardialanesisapropertyofthatparasite.Changingthe
conditionsofelectrophoresis,forexamplelongerrunningtimes,orlongerorshorterpulsetimes,didnotdemonstrablychangetheelectrophoresispatternsforGiardia
overmanyOFAGEruns.However,theexpectedchangesinmalariachromosomeprofile,suchasfurthermigrationoflargechromosomeswithlongerpulsetimes,
wereconsistentlyobserved.Changesinlysisconditions,suchasincreasedEDTA,proteinaseKordetergenthadnoobservableeffecteither.
TheOFAGEseparationwastransferredtonitrocellulosebytheSouthernprocedure(21)afternickingofthelargemoleculesbyacidtreatment(26)andhybridized
withsinglecopycDNAprobesandrepetitivesequencegenomicprobes.Nopatternwasdetectedwhichwasconsistentwithasinglebandofchromosomalmaterial
(datanotshown)thathadmigratedasignificantdistancefromtheloadingslotintothegel.
Figure2showsseparationofmalariachromosomesbyFIGE.ResolutionissuperiortoOFAGEandPFGandseparationisverygoodalso.However,theGiardia
chromosomepreparationsanalyzedonthesamegelaresimilartothoseseenbyOFAGE,withperhapsalittlemoredetailorstructureseeninthesmearofDNAthat
hasmigratedintothegel.Somebandingpatterncanalsobeseenunderlyingthefaintsmear.MostoftheDNAagainremainedintheloadingslot.However,thereisa
verydistinct,butbroad,bandofmaterialmigratingslightlyslowerthanthelambdaphagemarkerat50kb.
InthepreviousexamplesthechromosomeswerereleasedfromwholeparasitesaftertheywereentrappedinSeaplaqueagarose.Todeterminewhetherthisprocedure
hadanyeffectonthereleaseofthelargechromosomes,welysedparasitesdirectlyinthegelslotwithoutagaroseentrapment,priortoFIGE,inamannersimilartothe
analysisoflargeplasmidsinRhizobia(9).Inthiscase(Figure3)thebandmigratingslightlyslowerthanlambdaat50kbwasnotasintenseasseeninFigure2.
However,thereweredistinctspeciesseenat0.8Mbpand2Mbp.Thesmearofethidiumbromidestainingmaterialwasalmostabsentintheregionbetween2Mbpand
theslot.Againmostofthechromosomalmaterialdidnotenterthegelandremainsintheslot.
SincetherewerebroadbandsofDNAseenintheaboveexamplesmigratingatdefinedmolecularweightsandinareproduciblefashion,itwasofinteresttodetermine
iftheseweredistinctentities,orarangeofspeciesineachband.Wethereforesetthelysedparasites,whichweremountedinSeaplaque,intothesameagarosefor
FIGEseparation,fromwhichtheappropriatebandcouldbeexcisedandtheDNAextracted(6).Thegelwasrunfor16hat250V/150mAwithapulsetimeof22
secondsforwardand7secondsinthereversedirection.Thebroadbandmigratingslightlyslowerthan50kb(Figure3)wasexcisedfromtheSeaplaqueagarose,
dilutedsixfoldinto

Page149

10mMTrisHCl,pH7.5,1mMEDTAandextractedtwicewithanequalvolumeofbufferedphenol.Theresidualphenolwasremovedbythreeextractionswithan
equalvolumeofether.AlthoughthesmearednatureofthebandobviatesaccuratedeterminationofDNAconcentration,theyieldsofsuchlargemoleculeswerevery
good,estimatedatgreaterthan50%.
ThenatureoftheextractedDNAwasanalysedbyfurthergelelectrophoresisundertheusualuniformfieldconditions.Figure4showsthemigrationofasampleofthis
DNAin0.8%agarose.IncontrasttotheappearanceofthebandmigratingunderOFAGEorFIGEconditions,

Figure2.
G.intestinalisandP.falciparumchromosomessetinagarose
blocksandseparatedbyFIGE.ParasiteDNAwasseparatedina
1%agarosegelinTBEbufferat300V,150mAfor20h.
Forwardandreversepulseswere22and7sec.,respectively.Lane1
containsP.falciparum7G8lane2P.falciparumFC27
lane3G.intestinalisBRIS/3/HEPU1061/3.

thebandwasverytightanduniform.MostifnotalltheDNAwasofasingle,discretesizeclass,againmigratingslightlyslowerthanlambdaat50kb.Althoughitis
difficulttoestimateaccuratelyDNAlengthsinthisregion,evenusingconcatamersoflambdaasmarkers,weestimatethatthesizeoftheextractedDNAwas
approximately100kb.MalariachromosomesmigratingunderPFG,OFAGE,orFIGEalsoshowedabandintheregionof50kb.Whentheseareextractedasabove
andreanalyzedbygelelectrophoresis,theDNAmigratedasabroadsmearofdifferentsizeclasses(datanotshown):nodistinct,singlebandwasobserved.The
discretespeciesat100kbappearstobeapropertyoftheGiardiagenome,therefore,andnotoftheextractionconditionsused.
The100kbbandofDNAseenbygelelectrophoresisisconsistentwitheitherasingle'minichromosome'oracollectionof'minichromosomes'ofthesamesize.
Cleavagewithrestrictionenzymesindicatedthatthe100kbDNAisunlikelytobeasinglespecies(unpublisheddata)becauseadiscreteseriesofbandswhichhasa
summationof100kbinsizewasnotobserved.However,the'minichromosomes'didcontainrepeatedsequences,whichweredetectedasdistinctbandsaftergel
electrophoresisandbyhybridizationtoclonedrepetitivesequencesafterSoutherntransfer.Theserepeatedsequencesinthe'minichromosome'arenotarrangedin
anorderwhichisconsistentwiththeirtandemrepetition,asmightbeexpectedforrDNA,forexample.Theirrelativemassinthe'minichromosomes'isalsonot
consistentwiththetotalmassoftheserepeatsintheGiardiagenome.

Figure3.
LysisofG.intestinaliswithoutentrapmentinagaroseblocks
andseparationofchromosomalmaterialbyFIGE.Lanes1and2
containDNAfromG.intestinalisparasiteswhichwerelysed
inblocksasdescribedforFigures1and2.Lanes3and4contain
DNAfromwholeparasiteswhichwerelyseddirectlyintheloadingslots.
Theconditionsforelectrophoresiswere250V,150mAfor16h.
Pulsetimeswere22secintheforwarddirection
and7secinthereversedirection.

Page150

Figure4.
Electrophoreticanalysisof100kb'minichromosomes'.
Thebroadbandofapproximately100kbseeninFigure2was
extractedfromSeaplaqueagaroseandanalyzedon0.8%Seakemagarose.
Lane1contains'minichromosome'DNAfromG.intestinalislane2
bacteriophagelambda(48.5kb)lane31kbladderfromBRL.
Thetopbandis12kbinlengthlane4islambdaDNAcleaved
withEcoRIandHindIII.Thetopbandis21kbinlength.

Discussion
WehavedescribedaninitialapproachtoestablishingformalgeneticstudiesinG.intestinalis.SinceitisnotpossibletoobservethechromosomesofG.intestinalis
reliablybyclassicalmeans,thatismetaphasespreadingandstaining,whichseemstobeafrequentproblemamongtheprotozoa,wehaveattemptedtoseparatethe
chromosomesbygelelectrophoretictechniques.ThesehaveincludedtherecentlydevisedtechniquesofOFAGEandFIGE.Electrophoresisofmalariachromosomes
yieldedwellseparatedspecieswhichcanbecomparedfavourablywithpublishedPFGseparations.Incontrast,mostoftheGiardiachomosomalmaterialremainedin
theloadingslotofthegel,evenaftertwodaysofelectrophoresis.Themajorityofthechromosomesarethereforeverylarge(>4Mbp)orareentrappedinanetwork
intheparasiteanddonotmigratethroughtheagarose.Sincethelatterconclusionisprobablyincompatiblewiththenecessaryseparationofchromosomesduring
binaryfissionoftheparasiteduringreproduction,weconcludethatthechromosomesareverylarge.SimilarlargechromosomeshavebeenseeninT.brucei(23,24).
Evenifseparationwerepossible,saybylongerelectrophoretictimes,orbyimprovedelectrophoretictechniquesunavailableatpresent,minorchangeswouldnotbe
informativeforgeneticanalysis(therewouldhavetobemajordifferencesinsizeofchromosomeswhicharealreadylarge,todetectanygeneticchanges).Thisisin
contrasttothesituationwithmalariaandthetrypanosomes(11,12,15,23,24,25),wherelargedifferencesinsmallchromosomes,bothbetweenspeciesandwithin
strainsareobserved.Alternatively,thelargechromosomesofmammalscanbeidentifiedbymetaphasespreading,andregionsmappedbyconsistentstainingofbands
(18)relativetothecentromeresandtelomeres.Furthermore,theidentificationofgeneticchanges,suchasrestrictionfragmentlengthpolymorphisms(RFLPs),which
mayormaynotberelatedtoaknowngene,allowmappingatboththegeneticandDNAlevels(4,5),andultimatelytothespreadchromosomesobservedby
microscopy.SincenoneofthesechoicesisavailablepresentlyforthewholeGiardiagenome,itmaybepossibletodissect'artificialchromosomes',whichareshorter
than4MbpandcanbeseparatedbyFIGE.OneshouldbeabletomapgenesandRFLPstothesemoleculesandobserveclassicalgeneticchanges,suchasdeletions,
translocationsandinversions,bothbydirectobservationafterelectrophoresisandbyhybridizationstudies.Weareexploringthispossibilitycurrently,usingenzymes
suchasNaeIandSacI,whichtheoreticallywouldcleavetheGiardiagenomeonceevery62.5kband250kb,respectively(seeref.5forcalculation).Theseenzymes
recognizeGCrichsequencesandcontaintheCGdoubletalso.Wearealsousingenzymeswhichhave8bprecognitionsequences,suchasSfiIandNotI,and
theoreticallywouldcleavetheGiardiagenomeonceevery400kb.Sufficientoverlapofthese'artificialchromosomes'shouldbuilduptothecompletegenome
structure.
AlthoughthemajorityoftheGiardiagenomeiscontainedinverylargechromosomes,asassessedbyelectrophoreticseparation,DNAbandswereobservedwhich
enteredthegelandmigratedreproducibly.SomestructurewasseenalsointhesmearofDNAmigratingbetween50kbandtheslotunderOFAGEconditions,but
wasnotanalyzedhere.WhentheGiardiachromosomesweresubjectedtoFIGE,aconsistentbandofDNAwasseenmigratingatapproximately100kb.Whenthis
broadbandwasextractedfromSeaplaqueandelectrophoresedagain,itmigratedasatight,discreteband,whereasmalariaDNAextractedfromasimilarregionof
thegel,migratedasadiffuse,polydispersesmearofspecies.ThisbandisnotseenwhenDNAisdirectlyextractedfromparasitesbecausethebandmigratesinthe
bulkofgenomicDNAduringelectrophoresis.Sincetheagaroseblockingofparasiteshasbeenaveryreliablemethodofpreparingchromosomesforgelanalysis,the
blockscanbestoredformonthswithoutDNAdegradationandthemalariachromosomeswhichwepreparedmigratedasexpectedfrompublishedresults,we
concludethatthe100kbbandisa

Page151

propertyoftheGiardiagenome.Thediffusionofthebandseeninpulsedfieldconditionsisprobablyduetodifferentialretardationofthe'minichromosomes'asthey
leavetheirsiteoflysis,andtothediffusionoflargemoleculesunderchangingelectricalfields.
Astothenatureofthe'minichromosomes',weconcludethattheyareofaverydiscretesizeclass,theycontainrepetitivesequences,whichwehavedetectedafter
restrictionenzymecleavageandbyhybridizationtoclonedrepetitivesequences,butthatthemajorityoftherepeatsarenottandemlyarrayed,asmightbeexpectedfor
amplifiedrDNA,forexample(unpublisheddata).Cleavagepatternsarenotconsistentwiththe'minichromosome'beingauniquespecies,butratheracollectionof
predominantlysinglecopysequencesinmoleculesofthesamelength.Theyhavebeenobservedinanumberofisolates(unpublisheddata),sotheyappeartobeapart
ofthestructureoftheG.intestinalisgenome.TheiruniquesizedifferentiatesthemfromtheminichromosomesofT.brucei,wheretheyareapproximately100in
number,varyingfrom50kbto150kbinlength.ThesearethoughttoplayapredominantroleintheactivationofVSGsinthetelomeres(2,11,24).Theroleofantigenic
variationinGiardiaisyettobeassessed.TheseparationofGiardia'minichromosomes'leavesthemajorityoftheDNAintheslot.The'minichromosomes'are
unlikelythereforetobegeneratedbya'systematic'disruptionofthelargeGiardiachromosomes,otherwisetheremainingsegmentsshouldenterthegelasreadily.
Furthermore,theiruniquesizesuggeststhattheyareanewlydescribedclassofminichromosomes.Thelargerspeciesat0.8Mbpand2Mbp,asseenbyFIGEafter
directlysisintheloadingwell,haveyettobeanalyzedfurther.
Inconclusion,wehavedescribedverylargechromosomesofG.intestinalisandsmallerminichromosomecounterpartsofauniquesizeclass.Wehavesuggested
howthelargechromosomescanbeconvertedinto'artificialchromosomes'whichcanbeseparatedbyfieldinversionelectrophoresis.CoupledwithclonedcDNAand
genomiclibrarieswhichwehaveconstructed,containingbothuniqueandinterpersedrepetitivesequences,abasehasbeendevelopedfromwhichasystematicgenetic
analysisforGiardiacanbeapproached.
Acknowledgements
WethanktheNationalHealthandMedicalResearchCouncilofAustraliaforsupport.
LiteratureCited
1.Anderson,S.1981.ShotgunDNAsequencingusingclonedDNaseIgeneratedfragments.Nucl.Acids.Res.9:30153027.
2.Bernards,A.,L.H.T.VanderPloeg,W.C.Gibson,P.Leegwater,F.Eijgenraam,T.DeLange,P.Weijers,J.Calafat,andP.Borst.1986.Rapidchangesofthe
repertoireofvariantsurfaceglycoproteingenesintrypanosomesbygeneduplicationanddeletion.J.Mol.Biol.190:110.
3.Boreham,P.F.L.,R.E.Phillips,andR.W.Shepherd.1986.TheactivityofdrugsagainstGiardiaintestinalisinneonatalmice.J.Antimicrob.Chemother.18:393
398.
4.Bostein,D.,R.L.White,M.Skolnick,andR.W.Davis.1980.Constructionofageneticlinkagemapinmanusingrestrictionfragmentlengthpolymorphisms.Am.
J.Hum.Genet.32:314331.
5.Brown,W.R.A.,andA.P.Bird.1986.LongrangerestrictionsitemappingofmammaliangenomicDNA.Nature322:477481.
6.Burns,D.M.,andI.R.Beacham.1983.AmethodfortheligationofDNAfollowingisolationfromlowmeltingtemperatureagarose.Anal.Biochem.135:4851.
7.Carle,G.F.,andM.V.Olson.1984.SeparationofchromosomalDNAmoleculesfromyeastbyorthogonalfieldalternationgelelectrophoresis.Nucl.AcidsRes.
12:56475664.
8.Carle,G.F.,M.Frank,andM.V.Olson.1986.ElectrophoreticseparationsoflargeDNAmoleculesbyperiodicinversionoftheelectricfield.Science232:6568.
9.Djordjeciv,M.A.,W.Zurkowski,andB.G.Rolfe.1982.PlasmidsandstabilityofsymbioticpropertiesofRhizobiumtrifolii.J.Bacteriol.151:560568.
10.Giannini,S.H.,M.Schittini,J.S.Keithly,P.W.Warburton,C.R.Cantor,andL.H.T.VanderPloeg.1986.KaryotypeanalysisofLeishmaniaspeciesanditsuse
inclassificationandclinicaldiagnosis.Science232:762765.
11.Gibson,W.C.,andP.Borst.1986.SizefractionationofthesmallchromosomesofTrypanozoonandNannomonastrypanosomesbypulsedfieldgradientgel
electrophoresis.Mol.Biochem.Parasitol.18:127140.
12.Gibson,W.C.,andM.A.Miles.1986.ThekaryotypeandploidyofTrypanosomacruzi.EMBOJ.5:12991305.
13.Green,M.R.,andR.G.Roeder.1980.DefinitionofanovelpromoterforthemajoradenovirusassociatedvirusmRNA.Cell22:231242.
14.Gubler,U.,andB.J.Hoffman.1983.AsimpleandveryefficientmethodforgeneratingcDNAlibraries.Gene.25:263269.
15.Kemp,D.J.,L.M.Corcoran,R.L.Coppel,H.D.Stahl,A.E.Bianco,G.V.Brown,andR.F.Anders.1985.Sizevariationinchromosomesfromindependent
culturedisolatesofPlasmodiumfalciparum.Nature315:347350.
16.Ketner,G.,andT.J.Kelly,Jr..1976.Integratedsimianvirus40sequencesintransformedcellDNA:analysisusingrestrictionendonucleases.Proc.Natl.Acad.
Sci.U.S.A.73:11021106.
17.Nash,T.E.,T.McCutchan,D.Keister,J.B.Dame,J.D.Conrad,andF.D.Gillin.1985.RestrictionendonucleaseanalysisofDNAfrom15Giardiaisolates
obtainedfromhumansandanimals.J.Infect.Dis.152:6473.
18.Priest,J.H.1977.Medicalcytogeneticsandcellculture.2nded.Lea&Febiger,Philadelphia.
19.Schwartz,D.C.,andC.R.Cantor.1984.SeparationofyeastchromosomesizedDNAsbypulsedfieldgradientgelelectrophoresis.Cell37:6775.
20.Shenk,T.E.,J.Carbon,andP.Berg.1976.Constructionandanalysisofviabledeletionmutantsofsimianvirus40.J.Virol.18:664671.
21.Southern,E.M.1975.DetectionofspecificsequencesamongDNAfragmentsseparatedbygelelectrophoresis.J.Mol.Biol.98:503517.
22.Upcroft,J.,S.Chiu,andC.Kidson.1984.InvitrotranslationofPlasmodiumfalciparumproteins.Aust.J.Exp.Biol.Med.Sci.62:125135.
23.VanderPloeg,L.H.T.,A.W.C.A.Cornellisen,J.D.Barry,andP.Borst.1984.Chromosomesofkinetoplastida.EMBOJ.3:31093115.
24.VanderPloeg,L.H.T.,D.C.Schwartz,C.R.Cantor,andP.Borst.1984.AntigenicvariationinTrypanosomabruceianalyzedbyelectrophoreticseparationof
chromosomesizedDNAmolecules.Cell37:7784.

Page152

25.VanderPloeg,L.H.T.,M.Smits.,T.Ponnadurai,A.Vermeulen,J.H.E.Th.Meuwissen,andG.Langsley.1985.ChromosomesizedDNAmoleculesof
Plasmodiumfalciparum.Science229:658661.
26.Wahl,G.M.,M.Stern,andG.R.Stark.1979.EfficienttransferoflargeDNAfragmentsfromagarosegelstodiazobenzyloxymethylpaperandrapidhybridization
byusingdextransulfate.Proc.Natl.Acad.Sci.U.S.A.76:36833687.
27.Woo,P.K.1984.EvidenceforanimalreservoirsandtransmissionofGiardiainfectionbetweenanimalspecies.pp.341364In:GiardiaandGiardiasisBiology,
Erlandsen,S.L.,andE.A.MeyerEds.PlenumPress,NewYork.
28.YanischPerron,C.,J.Vieira,andJ.Messing.1985.ImprovedM13phagecloningvectorsandhoststrains:nucleotidesequencesoftheM13mp18andpUC19
vectors.Gene33:103119.

Page153

ThePartialCharacterizationofanImmunodominantAntigenofGiardiaintestinalis
JacquelineA.Upcroft*,AnthonyG.Capon,ArayaDharmkrongAt,PeterUpcroftandPeterF.L.Boreham
QueenslandInstituteofMedicalResearch,BramstonTerrace,Herston,Brisbane,Qld,4006,Australia.
Immunoblotsofantigensfrom13isolatesofGiardiaintestinalishaverevealeda32kDimmunodominantantigenwhichisrecognizedbyantiseraraisedagainstclonedisolates.This
antigenhasbeenpartiallypurifiedbyelectroelutionfrompolyacrylamidegelsandhasbeenusedtoraiseapolyclonalantiseruminrabbits.Whenthisantiserumwastestedby
immunofluorescenceassay,flagellaeandsomesurfacecomponentswererecognizedprimarilyandonimmunoblotsthe32kDantigenwastheonlyreactingproduct.Several
proteinsproducedbyinvitrotranslationofG.intestinalismRNAwereimmunoprecipitatedwithanantiserumraisedagainstwholetrophozoites.However,onlyoneofthese,amajor
translationproductof32kDwasimmunoprecipitatedwiththeantiserumraisedagainstthepartiallypurified32kDantigen.cDNApreparedfromGiardiamRNAandgenomicDNA
havebeenclonedintoexpressionvectorsofE.coliandsevencloneswhichexpressGiardiaantigenshavebeenanalyzed.

Introduction
TheidentificationofoneormoreantigensantigenswhichareuniquetoGiardiaintestinalisisaprerequisiteforthedevelopmentofanimproveddiagnostictestof
highsensitivityandspecificity.Thereisanurgentneedforsuchatesttoidentifybothsymptomaticandasymptomaticcasesforgiardiasis,ascurrentmethods,
dependingoneitherparasitologicalexaminationsorinvasiveprocedures,aretimeconsumingandmissmanypositivecases(5).Itisclearfromseveralreportsthat
differencesamongGiardiaisolatescanbedetectedattheantigenic(17),enzymic(1)andgenomic(14)levels.Anynewdiagnostictestmusttakeintoaccountsuch
variationandbeabletodetectparasitesfromallgeographiclocations.
QualitativeandquantitativeantigenicdifferencesamongfourisolatesofG.intestinalisfromAfghanistan,Oregon(USA),EcuadorandPuertoRicohavebeen
demonstratedbycrossedimmunoelectrophoresis(17).Thisstudyemployedantiseraraisedagainstwholeparasitesandpartiallypurifiedantigens.Amajorsurface
antigenof82kD,commontotheabovefourisolates,hasbeendetectedbysurfacelabellingtechniques(8).Otherminorsurfacelabelledproteinsof180,105,63,55,
37,30and24kDwerealsoidentified(8).Someoftheseproteinsweresecretory/excretoryproductsanddifferencesbetweenthesecretory/excretoryproteinsoftwo
isolateshavebeenreported(13).Itisunclearonwhatbasisanyofthesedifferencesoccuranditappearstheyareneitherrelatedtohostspecificitynortothe
geographiclocation(14).
WehavebeenstudyingtheantigensofG.intestinalisinordertoidentifythosewhicharecommontodifferentisolatesandwhicharepotentiallyusefulasdiagnostic
reagents.Wehaveisolatedonesuchantigenofmolecularweight32kDandshownitisasurfaceantigen.Wereportonpartialpurification,characterizationanda
preliminarysearchforarecombinantDNAclonewhichexpressesit.
Results
DetectionandCharacterisationofanImmunodominantAntigen
1.DetectionofaCommonImmunodominantAntigen
Ourinitialobservationsconfirmedotherstudies(17)thatanalysisofCoomassieBluestainedG.intestinalisproteinsseparatedbypolyacrylamidegelelectrophoresis
(PAGE)revealnoobviousdifferencesamongisolates.However,immunologicaldetectionofG.intestinalisantigenswhichhavebeenelectroblottedfrom
polyacrylamidegelsontonitrocellulose(immunoorWesternblot)(19)revealedmanydifferences(Figure1).ElevenisolatesfromBrisbaneandenvirons,onefrom
WesternAustralia,onefromPapuaNewGuineaandthePortland1isolatewereusedinthesestudiesandwerereactedonimmunoblotswithantiseraraisedinrabbits
(2)againstsixclonedlinesofG.intestinalis(3)(BRIS/82/HEPU/41/1/6BRIS/83/HEPU/99/1/4BRIS/83/HEPU/106/1/1BRIS/83/HEPU/106/1/5
BRIS/83/106/1/7BRIS/83/HEPU/120/1/13).Alloftheisolatesexaminedhaveamajor,immunodominantantigenat32kDwhichisrecognizedbyalloftheantisera
used.Thisantigenorgroupofantigensappearsasawidebandintheimmunoblot(Figure1).
2.PartialPurificationofthe32kDAntigen
Preparative12.5%polyacrylamidegelswereusedtoseparatethe32kDprotein.2108trophozoiteswereloadedontothegelsaftersolubilizationinSDSbuffer.
Followingseparationoftheproteins,the32kDbandwasexcisedfromthegelandtheproteinelectroelutedintoa7.5Mammonium
*Correspondingauthor.

Page154

acetatecushionusinganInternationalBiotechnologiesInc.modelUEAelectroeluterfollowingthemethodgivenbythemanufacturers.Theeluatewasthendialyzed
overnightagainstphosphatebufferedsaline,pH7.2,andconcentratedbylyophilization.
3.ImmunofluorescenceStudiesWithAntiserumRaisedAgainstthe32kDAntigen
Rabbitantiserumraisedagainstthe32kDbandwasusedinimmunofluorescenceassays(IFA)tovisualizethelocationofthe32kDantigenintheparasite.Rabbits
wereinoculateddirectlyintolymphnodeswiththepartiallypurifiedproteinfrom2108trophozoites,mixedwithFreundscompleteadjuvantinthreesuccessive
inoculations(2).Thisantiserumhadatitreof1in4000byIFAagainstfixedparasites.Twomethodswereusedintheimmunofluorescencetests.Firstly,livewashed
trophozoiteswereincubatedwithantibodyagainstthe32kDantigen,gentlywashedtoremove

Figure1.
Immunoblotofproteinsoftwoisolatesof
G.intestinalisBRIS/82/HEPU/41
andBRIS/85/HEPU/449separatedbyP.A.G.E..
Lanes1,2and3werereactedwithantiseraraisedagainst
clonedlinesofG.intestinalis:BRIS/83/HEPU/99/1/4,
BRIS/83/HEPU/106/1/1andBRIS/83/HEPU/106/1/5,respectively.

unboundantibody,allowedtoadheretoglassslidesandthenacetonefixed.Secondly,antiserumwasreactedwithparasitesafteracetonefixation.Thebindingof
antibodytoparasiteswasdetectedbyuseofasecondfluoresceinlabelledantirabbitantibody.Theantibodytothe32kDproteinreactedprimarilywithflagellaeand
toalesserextentthesurfaceofthetrophozoite(Figure2).Bothtechniquesrevealedsimilarfluorescencepatterns.However,therewasagreaterintensityof
fluorescenceinthoseorganismswhichhavebeenacetonefixedpriortoexposuretotheantibody.
4.ImmunoblotofAntigensReactedwiththeAnti32kDSerum
Immunoblotsofparasitesreactedwiththeanti32kDantiserum(Figure3)consistentlyshowedonlythemajor32kDband.Thisresultsuggeststhatthe32kDprotein
orgroupofproteinsisnotadegradationproductofalargermoiety.
Wehaveshownthatthereisa32kDimmunodominantantigen(s)presentinG.intestinaliswhichiscommontoalltheisolatessofarexamined.Thisproteinappears
tobeassociatedwiththesurfaceofthetrophozoiteandinparticularwithflagellaeandrelatedstructures.HolbertonandWard(11)whohavebeenstudying
microtubulesofGiardiahaveidentifiedagroupofproteinsofapproximately30kDwhichareassociatedwiththeflagellarmembranesandthecytoskeletonofthe
parasite.Thisgroupofproteins,togetherwithtubulin(52.5kD)arederivedfromtheventraldiscandaxonemesofGiardia.Theymigrateasabroadbandwhen
separatedbyPAGEandrepresentagroupofpossiblyeightrelatedpolypeptides,calledgiardins.Inaddition,furthercharacterizationofthis30kDgroupofproteins
hasrevealedthatsomeofthemareunrelatedtothegiardinsandderivedfromtheflagellarmembraneortheparaxialrodsbeneaththemembrane.Ifthecommon32kD
proteinwehaveidentifiedisindeed

Figure2.
IndirectIFAofG.intestinalistrophozoites(x1000).Isolate
BRIS/85/HEPU/449reactedwithrabbitantiserumraised
againstthepartiallypurified32kDprotein.

Page155

Figure3.
ImmunoblotofproteinsofG.intestinalisisolateBRIS/85/HEPU/449
separatedbyP.A.G.E..Lane1:reactedwithantiserumprepared
againstaclonedlineofG.intestinalisBRIS/83/HEPU/106/1/5.
Lane2:reactedwithanantiserumraisedagainstthepartially
purified32kDproteinfromisolateBRIS/83/HEPU/449.

associatedwiththeflagellarmembranes,andisthesameasthatdescribedbyCrossleyetal.(7),itmightbeexpectedtobeamajorsurfaceantigenandtherefore
detectedasaprominentlyiodinatedproteininstudiesonsurfacelabellingofparasites(8,13).InterestinglybothEinfieldsandStibbs(8)andNashetal.(13)have
identified,inadditiontoamajor82kDsurfaceantigen,aminorsurfaceproteinofapproximately32kD.However,iftheproteinhasalowcontentoftyrosineandless
importantlyislowinhistidine,tryptophanandsulphydrylgroups,itmaynotbeiodinatedandidentifiedbysurfacelabellingtechniquesbutmaystillbeanabundant
surfaceantigen.
CloningofGiardiaantigens
1.Identificationofthe32kDAntigenininvitroTranslationsofmRNA
Agroupofproteinsrecognizedbyantibodyasabroadbandinimmunoblotsmayindicatethataparticularproteinisglycosylatedvariablysuchthatthesizeofthe
glycosylatedproteinisnotconstantorthatitrepresentsagrouporfamilyofproteinswhicharemodifiedposttranslationally.Manyantigensofprotozoanparasitesare
knowntobeglycosylated(6,12,15)andG.intestinalismaywellbesimilar.
InvitrotranslationofmRNAresultsintheproductionofproteinswhicharesynthesizeddirectlyfromthemRNAandwhichhavenotundergoneanyintracellularpost
translationalmodifications,forinstance,glycosylationorremovalofpeptidesequences.
G.intestinalisRNApreparedbylysisofparasitesinSDSbuffercontaining10mMribonucleosidevanadylcomplexes(BRL)asRNaseinhibitorand
phenol/chloroformextraction(20),wasallowedtoadsorbontoHybondmessengeraffinitypaper(Amersham)andthemRNAwhichwasspecificallyadsorbedwas
subsequentlyeluted(23).ThemRNAwastranslatedinvitroinarabbitreticulocytecellfreesystem(16)andanalyzedbyPAGE(Figure4).Thetranslationproducts
rangedinsizefromlowtohighmolecularweightproteins.Whenthesetranslationproductswerereactedwithantiseraraisedagainstwholeparasitesusingthemethod
ofTaylorandcolleagues(18),severalmajortranslationproductswereimmunoprecipitated.Theseincludeda32kDtranslationproduct.Theantiserumraisedagainst
thepartiallypurified32kDbandimmunoprecipitatedonlythe32kDtranslation

Figure4.
ImmunoprecipitationoftranslationproductsofG.intestinalis
mRNA.Lane1:wholemRNAtranslationproducts.Lanes23:
immunoprecipitationofwholetranslationproductswith
antiseraraisedagainstwholeparasites.Lane4:controlofimmunoprecipitation,
minusantibody.Lane5:immunoprecipitationwithnormal
rabbitserum.Lane1wasexposedforashorterperiodthan25.
Lane3hadtwiceasmanycountsaslane2.Thearrowindicatesthetranslation
productimmunoprecipitatedwithantiserumraisedagainstthe32kDantigen.

Page156

product(arrowedinFigure4).
2.CloningofGiardiaintestinalisDNA
FollowingthepositiveidentificationofimmunoprecipitabletranslationproductswesynthesizedcDNA(9)fromthemRNAwhichcodesfortheinvitrotranslationof
this32kDproteinandclonedthe.IxcDNAcDNAintotheE.colivectorpUC19(21).GenomicDNAfromGiardiawaspartiallycleavedwithRsa1,Alu1and
DNase1andclonedbythesamemethodintopUC13andpUC19(21).Antiseraraisedagainstwholeparasitesfrommixedisolatesandantiseraraisedagainstthe
32kDbandhavebeenusedasthefirstantibodyinacolonyassaytodetectcolonieswhichexpressGiardiaantigens(10).Thesecondantibodywasbiotinylatedand
thedetectionsystemusedstreptavidinandbiotinylatedhorseradishperoxidase(22).ToconfirmthatthecoloniesdoexpressGiardiaantigens,bacteriallysateswere
inoculatedintomice(4)andtheresultingantiserumtestedbyIFAtodeterminewhichantigenthebacterialcloneswereexpressing.Atpresentsevenbacterialclones
whichreactwithantiseraraisedagainstwhole,mixedparasiteshavebeentestedagainstantiseraraisedagainstclonedisolates,butnoneappeartoreactwithantisera
raisedagainstthe32kDband.Noneoftheclonesreactedwithalloftheantiserawhichindicatesthatthecommon32kDproteinisnotrepresentedintheseclones.The
rabbitwhichwasusedtoraisetheanti32kDantiserahadaveryhighlevelofE.coliantibodiesconsequentlythebackgroundinthecolonyassayswashighdespite
repeatedadsorptionswithsonicatedE.colicellscarryingtheplasmidvectorwithoutaninsert.Morerecentlywehaveraisedantiserainspecificpathogenfree(SPF)
rabbitsandthebackgroundreactionagainstE.coliisgreatlyreduced.
The32kDantigenappearstobeamajorcomponentoftheinvitrotranslationproductsofmRNAfromGiardiasincetheantiserumraisedagainstthe32kDband
immunoprecipitatedamajortranslationproductofthesamesize.Thesedatashowthatsome,ifnotall,oftheantibodyintheantiserumraisedagainstthe32kDband
aredirectedagainstpolypeptideandnotglycosylatedepitopes.Thusthisantiserumislikelytorecognizethe32kDantigenexpressedasaforeignpolypeptidewhen
clonedintoanexpressionvectorofE.coli,suchasthepUCseriesofplasmids.OurinitialsearchforGiardiaantigensinourlibrarieshasprovedsuccessfulandwe
haveidentifiedcloneswhichexpressGiardiaantigensinbothcDNAandgenomicDNAlibraries.Wearepresentlyscreeningtheselibrarieswithantiseraraisedin
SPFrabbits,againstthe32kDantigen.
Conclusion
WehaveidentifiedasurfaceantigenofG.intestinaliswhichappearstobeassociatedprimarilywiththeflagellaeandsurfacecomponentsofthetrophozoite.This
antigenhasamolecularweightasdeterminedbyPAGEof32kDanditmaybethesameproteinthatHolbertanandcoworkershaveidentifiedasassociatedwith
flagellarmembranes.Acommonsurfaceantigen,suchasthe32kDantigen,associatedwithflagellaemayproveusefulinthedevelopmentofadiagnosticreagentfor
G.intestinalissinceflagellarproteinsarelikelytobepresentinstoolsevenifintactorganismscannotbeseen.Workiscurrentlybeingundertakentodetermine
whethertheseproteinscanbedetectedinstoolsbyELISA.
The32kDantigenisamajorcomponentofinvitrotranslationproductsofmRNAfromG.intestinalisandthereisnoevidenceofanygrossposttranslational
modification.WehaveproducedlibrariesofG.intestinalisDNAclonedintoexpressionvectorsofE.coliandhavescreenedtheselibrariesforexpressionofG.
intestinalisantigens.PreliminarysearchesinourclonedlibrarieshaverevealedtheexpressionofsomeGiardiaantigensbutasyetnonethatcanbeconfirmedasthe
32kDcommonantigen.
Acknowledgements
TheoriginalresearchdescribedinthispaperhasbeensupportedbyagrantfromtheNationalHealthandMedicalResearchCouncilofAustralia.A.G.C.isaNational
HealthandMedicalResearchCouncilMedicalPostgraduateResearchScholar.
LiteratureCited
1.Bertram,M.A.,E.A.Meyer,J.D.Lile,andS.A.Morse.1983.AcomparisionofisozymesoffiveaxenicGiardiaisolates.J.Parasitol.69:793801.
2.Boreham,P.F.L.,andG.S.Gill.1973.SerologicalidentificationofreptilefeedsofGlossina.ActaTrop.30:356365.
3.Boreham,P.F.L.,R.E.Phillips,andR.W.Shepherd.1987.HeterogeneityintheresponseofclonesofGiardiaintestinalistoantigiardialdrugs.Trans.R.Soc.
Trop.Med.Hyg.81:406407.
4.Coppel,R.L.,G.V.Brown,G.F.Mitchell,R.F.Anders,andD.J.Kemp.1984.IdentificationofacDNAcloneencodingamaturebloodstageantigenof
Plasmodiumfalciparumbyimmunizationofmicewithbacteriallysates.EmboJ.3:403407.
5.Craft,J.C.1982.Giardiaandgiardiasisinchildhood.Pediatr.Infect.Dis.1:196211.
6.Cross,G.A.M.1978.AntigenicvariationinTrypanosomes.Proc.R.Soc.Lond.B.202:5572.
7.Crossley,R.,J.Marshall,J.T.Clark,andD.V.Holberton.1986.ImmunocytochemicaldifferentiationofmicrotubulesinthecytoskeletonofGiardialambliausing
monoclonal antibodiestotubulinandpolyclonalantibodiestoassociatedlowmolecularweightproteins.J.CellSci.80:233252.
8.Einfeld,D.A.,andH.H.Stibbs.1984.IdentificationandcharacterizationofamajorsurfaceantigenofGiardialamblia.Infect.Immun.46:377383.
9.Gubler,U.,andB.J.Hoffman.1983.AsimpleandveryefficientmethodforgeneratingcDNAlibraries.Gene.25:263269.
10.Helfman,D.M.,J.R.Feramisco,J.C.Fiddes,G.P.Thomas,andS.H.Hughes.1983.Identificationofclonesthatencodechickentropomysinbydirect
immunologicalscreeningofacNDAexpressionlibrary.Proc.Nat.Acad.Sci.80:3135.
11.Holberton,D.V.,andA.P.Ward.1981.IsolationofthecytoskeletonfromGiardia.Tubulinandalowmolecularweightproteinassociatedwithmicroribbon
structures.J.Cell.Sci.47:139166.

Page157

12.Howard,R.F.,andR.T.Reese.1984.SynthesisofmerozoiteproteinsandglycoproteinsduringtheschizogonyofPlasmodiumfalciparum.Mol.Biochem.
Parasitol.10:319334.
13.Nash,T.E.,F.D.Gillin,andP.D.Smith.1983.ExcretorysecretoryproductsofGiardialamblia.J.Immunol.131:20042010.
14.Nash,T.E.,T.McCuthcan,D.Keister,J.B.Dame,J.D.Conrad,F.D.Gillin.1985.RestrictionendonducleaseanalysisofDNAfrom15Giardiaisolates
obtainedfromhumansandanimals.J.Infect.Dis.152:6473.
15.Newbold,C.I.,D.B.Boyle,C.C.Smith,andK.N.Brown.1982.Identificationofaschizontandspeciesspecificsurfaceglycoproteinonerythrocytesinfected
withrodentmalarias.Mol.Biochem.Parasitol.5:45554.
16.Pelham,R.B.,andR.J.Jackson.1976.AnefficientmRNAdependenttranslationsystemfromreticulocytelysates.Eur.J.Biochem.67:247256.
17.Smith,P.D.,F.D.Gillin,N.A.Kaushal,andT.E.Nash.1982.AntigenicanalysisofGiardialambliafromAfghanistan,PuertoRico,EcuadorandOregon.Infect.
Immun.36:714719.
18.Taylor,D.W,J.S.Cordingley,andA.E.Butterworth.1983.ImmunoprecipitationofsurfaceantigenprecursorsfromSchistosomamansonimessengerRNAin
vitrotranslationproducts.Mol.Biochem.Parasitol.10:305318.
19.Towbin,H.,T.Staehelin,andJ.Gordon.1979.Electrophoretictransferofproteinsfrompolyacrylamidegelstonitrocellulosesheets:Procedureandsome
applications.Proc.Natl.Acad.Sci.76:43504354.
20.Upcroft,J.A.,S.Chiu,andC.Kidson.1984.InvitrotranslationofPlasmodiumfalciparumproteins.Aust.J.Exp.Biol.Med.Sci.62:125135.
21.Upcroft,P.,andA.Healey.1987.ArapidandefficientmethodforcloningofbluntendedDNAfragments.Gene51:6975.
22.Wilchek,M,andE.A.Bayer.1984.Theavidinbiotincomplexinimmunology.Immunol.Today5:3943.
23.Wreschner,D.H.,andM.Herzberg.1984.AnewblottingmediumforthesimpleisolationandidentificationofhighlyresolvedmessengerRNA.NucleicAcids
Res.12:13491359.

Page159

ImmunofluorescenceDifferentiationbetweenVariousAnimalandHumanSourceGiardiaCystsUsingMonoclonalAntibodies
HenryH.Stibbs,*EdwardT.Riley,JosephStockard,JohnL.Riggs,PeterM.WallisandJudyIsaacRenton
U.S.JapanBiomedicalResearchLaboratory,DepartmentofMedicine,TulaneUniversity,HebertCentre,Bldg30,BelChase,LosAngelos,California70037,
U.S.A..
MousemonoclonalantibodieshavebeendevelopedagainstcystsofGiardiamurisandofGiardiasimoniisolatedfromawildNorwayratcapturedontheUniversityofWashington
campus.ThefourantiG.murisantibodiesreactedpositivelyinindirectimmunofluorescencewiththeratsourcecystsinadditiontothehomologousG.muris,butnotwithcysts
isolatedfrombeaver(2),dog(4),human(8),muskrat(3),orRichardson'svole(1).Theoneantiratcystmonoclonalantibodyreactedonlywithratandcow(CW1Alberta)
sourcecysts.TheantiG.lambliacystmonoclonalantibodyprovidedbyJ.Riggswasfoundtoreactwithallhuman,beaver,anddogsourcecysts,andwithratGiardia,butnotwithG.
murisorcystsofmuskratorRichardson'svoleorigin.TheresultssuggestthatsystematicdifferencesoccurinthecystsurfacemembraneantigensofvariousGiardiastrains,andthat
monoclonalantibodiesmayproveusefulindevelopinganantibodytypingsystemforGiardiastrainandanimalsourceidentification.

Introduction
WaterborneGiardiacystshavebeenimplicatedasthecauseofover80outbreaksofgiardiasisintheUnitedStatesoverthepast30years(4,11,15,22).Giardia
infectedmammals,mostnotablythebeaver,havebeenimplicatedastheprobablesourceofGiardiacystsinfectinghumansinseveraloftheseoutbreaks
(1,6,15,28,30).Ithasnever,however,beenpossibletoprovewhethertheinfectingGiardiastrainsintheseoutbreakswereofanimalorhumanoriginorbothorto
identifywithabsolutecertaintywhichspeciesofanimal(s)(ifany)contributedthehumaninfectivecysts.Giardiainfectionsarecommoninawidevarietyofwildand
domesticmammalsandbirds(1,3,5,8,14,20,27,28,30)therefore,itisreasonabletoassumethatinmostNorthAmericanwatershedssomeanimalspecieswillbe
GiardiainfectedandexcretingGiardiacystsatanypointintime.Aportionoftheexcretedcystswillfindtheirwayintothesurfacerunoff.Thus,Giardiasp.cysts
canprobablybeconsideredomnipresentandubiquitousinnaturalsurfacewatersinNorthAmerica.
TheinfectivityformanofthemanyanimalstrainsorspeciesofGiardiafoundinNorthAmericanwatershedsisanalmostcompletelyunexploredsubject,aboutwhich
manyassumptionsandguesseshavebeenmadeovertheyears.Controlledexperimentsinvolvinghumansubjectshaveneverbeenperformed.Studiesinvolvingcross
infectivitybetweenanimalhostspecieshaverevealedgreatvariabilityinhostspeciesspecificity,sothatonemaynotconfidentlymakepredictionsabouttheinfectivity
formanofanyanimalisolates(1,14,30).
EffectivemethodshavebeendevelopedforfilteringsurfacewaterforthepurposeofconcentratingGiardiacystsandforidentifyingthecystsintherecoveredparticles
(10,12).GiardiacystshavebeenidentifiedinsurfacewatersinmanyareasofNorthAmerica,andinsomecasesquantitativedataontheconcentrationofcystsin
surfacewaterhavebeenobtained(19,28).However,inusinglightmicroscopyorimmunofluorescencetoexaminecystsrecoveredfromfilteredsurfacewater,one
cannotreliablyidentifytheanimalorigin,strainorspeciesofthecysts.ThisisduetothefactthatGiardiacystsofallanimalandhumanoriginsareanatomicallyvery
similar(14)andalsothefactthatGiardiacystsofmost,ifnotall,mammaliansourcesareequallywellvisualizedbyimmunofluorescencemethodsperformedwith
polyclonalanticystserum(21,22,26).
Methodscapableofdistinguishingantigenically,biochemically,orgeneticallybetweentheGiardiacystsproducedbydifferentanimalsandbymanwouldenableone
toidentifybyanimalsourcetheGiardiacyststhathavecausedahumanoutbreakofgiardiasis(bytestingcystsrecoveredfrompatients),andalsotoidentifythe
animalorhumanoriginofcystsrecoveredfromsurfacewater.Identifyingthesource(s)ofthecystsinwaterwouldallowonetoestimatethedegreetowhichhuman
fecalpollutionofawatershediscontributinghumansourcecyststothesurfacewater,andalsotoestimatetherelativecontributionsofvariousanimalspecies.At
present,however,almostnothingisknownofthepossibleantigenic,biochemical,orgeneticdifferencesthatmayexistbetweencystsofvariousanimalandhuman
sourceGiardia.Riggs,however,hasalreadyreportedtheuseofamousemonoclonalantibodyagainsthumansourceG.lambliacystsindifferentiatingbetween
humanandanimalGiardiacystsbyimmunofluorescence(21).Thisantibodyhasbeenfoundtobindtoallhumansourcecystisolatestestedaswellastocystsfrom
beaversanddogs
*Correspondingauthor.

Page160

italsohasreactedpositivelywithGiardiacystsfromonecattlespecimenfromYosemiteNationalParkandwithcystsfromacoyoteandachipmunk.Wehavealso
found(seeresultsbelow)thatthisantibodyreactsstronglywithGiardiacystsoftheduodenalismorphologicaltyperecoveredfromawildNorwayratinWashington
state.TheRiggsantibodydoesnot,however,bindtoGiardiacystsproducedbyawidevarietyofotheranimals,includingdeer,muskrats,anddifferenttypesofvoles
andmice(seeresults).Smithetal.(25)alsoreportedproductionofamousemonoclonalantibodythatwouldreactinimmunofluorescencewithG.lambliacystsbut
notcystsfrommice(G.muris)ordogs(G.canis).Themiceusedasthesourceofspleencellsforthehybridomafusionhadinthiscasebeenimmunizedwithinvitro
trophozoitesofG.lamblia.
TheexistenceofantigenicdiversityamongcystsofhumanandanimalGiardiaisolatesseemslikelyinlightofpreviousdiscoveriesofantigenicdiversityamong
trophozoitesofhumanandanimalisolates(13,16,18,24)andofdifferencesinagarosegelelectrophoresisbandingpatternsofrestrictionendonucleasefragmentsof
DNAfromhumanandanimalisolates(17).DiversityinisozymepatternsamongGiardiatrophozoitesofhumanandanimaloriginshasalsobeenreported(2).
WehavenowproducedmousemonoclonalantibodiesagainstGiardiamuriscystsisolatedfrommiceandagainstGiardiacystsisolatedfromwild,livetrapped
Norwayratsinanefforttoseeifantibodiescanbeproducedwhichbindtosomeoralloftheanimalsourcecystsbutnottohumansourcecysts.Theseantibodieswill
alsobeusedtoidentifytheanimaloriginofGiardiacyststhroughthecreationanduseofanantibodytypingsystemforcystsofthisparasite.
Methods
GiardiaIsolatesUsedinImmunizingMice
GiardiamuriswasacquiredfromDr.MartinHeyworthoftheDivisionofCellBiology,VAMedicalCenter,SanFrancisco,andmaintainedinBALB/c,Swiss
Webster,andnude(nu/nu)mice.ThisisolatehadoriginallybeenprocuredfromaninfectedgoldenhamsteratCaseWesternReserveUniversity(23).
TwowildNorwayratswerelivetrappedalongadrainagecanallocatedontheUniversityofWashingtoncampusinSeattleandwerefoundtobeinfectedwitha
Giardiastrainoftheduodenalismorphologicaltype(presumablyG.simoni).Theratsweremaintainedinthelaboratoryforseveralmonthsonadietofcommercial
ratchow,apples,carrots,andwater.Cystexcretionseemedtoberelativelyconstantoverthemonthsthattheratsweremaintainedincaptivity.Gerbils(Tumblebrook
FarmsCo.,Massachusetts)wereinoculatedwithcystsfromtherats,andcystsisolatedfromsuccessfullyinfectedgerbilsusedtosupplementthecystsobtained
directlyfromtheratsforthepurposeofimmunizingmice.WewerenotabletoinfectweanlingSpragueDawleyorLongEvansratswiththiswildratGiardiaisolate.
PurificationofCysts
MouseorratsourceGiardiacystswerepurifiedfromfecesbycentrifugationover1Msucroseinwater(10min,500g)followedbystepgradientcentrifugation
overtwolayersofPercoll(SigmaChemicalCo.,St.Louis,MO.)ofspecificgravities1.05and1.09(15min,500g).Occasionally,cystswerepurifiedbyflotation
onzincsulfate(specificgravity,1.18)followedbycentrifugationover1Msucrose.
ProductionofMonoclonalAntibodies
FemaleBALB/cmicewereimmunizedagainstfreshlyisolatedcyststhroughaseriesoffourtofiveintraperitonealinjectionsof13106cysts/animal/injectionover6
weeksfollowedby1or2intravenoustailinjections,4daysapart,of23106cystsinsterilenormalsaline.Fouror5daysafterthelastintravenousinjection,mice
weresacrificedandspleencell:myelomacellfusionscarriedout.SpleencellsfromtheimmunizedmicewerefusedwithNS1(P3NS1Ag4.1)mousemyeloma
cellsgrowninRPMI1640mediumwith15%fetalbovineserum,using40%polyethyleneglycol1500,usingconventionalmethods(9).Cellswereplatedonto96
wellcultureplatesandhybridomagrowthselectedforusingRPMI1640medium(15%fetalbovineserum)supplementedwithHAT
(hypoxanthine/aminopterine/thymidine).Growthwasenhancedbyusingmousethymuscellsasfeedercells.
Hybridomassecretingantibodiestocystsweredetectedbyindirectimmunofluorescenceusingthehomologouscystsairdriedontothebottomsofflatbottom,96well
ELISAplates(Falcon3915)at5,000to10,000cystsperwell,andalso(withG.muriscysts)byELISA,againwithcystsdrieddownontothewellsof96well
plates,using25,000cystsperwell.Stablehybridomaswereclonedbylimitingdilutionthreetimes.AscitesfluidwassometimesproducedinBALB/cmicepretreated
withPristane(SigmaChemicals).Bothculturesupernatantandascitesfluidwereusedintheimmunofluorescencecrosstestingstudiesdescribedbelow.
ImmunofluorescenceCrosstesting
TheGiardiaisolatesusedintheimmunofluorescencecrosstestingexperimentswiththemonoclonalantibodieshaddiverseanimalorigins(seeTables1and2).
MuskratswerelivetrappedontheUniversityofWashingtoncampus(samelocationwheretheNorwayratsweretrapped)andineasternWashingtonoutsideof
Ellensburg(theNA,orNaneum,isolates),andweremaintainedinrabbitcagesonadietofapples,carrots,celery,lettuce,andalfalfahay.TheMR6muskratisolate,
theC3vole(Clethrionomysgapperi)isolate,andtheCW1cowisolatewereacquiredfromPeterWallisinCalgary,Alberta,andweremaintainedinSwiss
Webstermice,or,inthecaseoftheCW1,ingerbils.GiardiainfectedMicrotusochrogasterwereprovidedbyStanleyErlandsenoftheUniversityofMinnesota
theanimalshadoriginatedinMissouri.TheManastashbeaverisolatewasfromtheEllensburgareaofeasternWashingtonandwasprovidedbyGlenClarkofCentral
WashingtonUniversity.Dr.ClarkalsoprovidedthefecesofMicrotusrichardsoniwhichhadbeencollectedatParadiseCreekinMr.RainierNationalPark.TheM
beaverisolateoriginatedinBritishColumbiaandwasprovidedasculturedtrophozoitesbyJudyIsaacRenton.TheD3dogisolatewasadaptedtocultureand
providedbyPeterWallis.TheH2,H3,andH4humanisolateswereprovidedbyCharlesHiblerofColoradoStateUniversityininfectedgerbils,andwereadapted
tocultureinourlaboratory.TheTBhumanisolatecamefromapatientinSeattleandwasadaptedtocultureononeoccasionitwasusedtoinfectaLongEvansrat,
fromwhichcystswereobtained.ThedogGiardiaspecimenslistedinTable1camefromdogshousedintheUniversityofWashingtonvivariumandusedforother
researchpurposes.Inanumberofcasescystsforcrosstestingwereobtainedfromgerbilsthathadbeeninoculatedwitheitherculturedtrophozoitesorwithcystsand
immunosuppressedwithDexamethazone(IntensolRoxaneLaboratories,Columbus,OH.1.5mginto100mldrinkingwater)thesegerbilderivedcystsareindicated
assuchintheTables.
Cystsforimmunofluorescencetestingwereusuallypurifiedbytheproceduresdescribedabovesometimes,however,cystswerespottedontothetestwellsasan
aqueousfecalslurry.CystswerespottedontoeightspotTefloncoatedslides(Bellco,Vineland,N.J.),airdried,andfixedinacetoneatroomtemperature.Theslides
wereusuallystoredat75Cwithdessicantintightlysealedboxesuntiltesting.Hybridomasupernatantorasciteswasallowedtoreactwiththecystsforaboutone
hourat37C(orovernightat4C)followedbytwobriefrinsesin0.0175Mphosphatebufferedsaline,pH7.4(PBS),andafurtheronehourincubationwithFITC
labeledgoatantimouse

Page161

immunoglobulinantibody(CappelLaboratories,WestChester,PA.)at1:80dilutioninPBSwith2%normalgoatserum.Afterrinsing,theslidesweremountedwith
90%glycerol/PBScontaining0.5mg/mLpphenylenediamine,andviewedusingepifluorescence.
Results
Fourstable,clonedhybridomalinessecretingimmunofluorescencepositiveandELISApositivemonoclonalantibodieswereproducedagainstG.muriscysts.All
wereoftheIgG1class.OnestablehybridomalinesecretingimmunofluorescencepositiveIgG1antibodyagainsttheratsourceGiardiacystswasalsoproduced.
TheresultsofcrosstestingtheseantibodiesbyindirectimmunofluorescencewithcystsfromthemanyanimalandhumanoriginsareshowninTables1and2.Theanti
G.lambliacystmonoclonalantibodyofRiggs,directlylabeledwithfluorescein,wasalsotestedagainstallofthecystisolatestheresultsareincludedinTable1.The
fourG.murismonoclonalsallshowedanidenticalpatternofreactivitywiththevariouscystisolates.Allfourreactedstronglywithcystsofthehomologousisolateand
alsowiththewildNorwayratsourceGiardia.Nobindingtocystsofhuman,beaver,dog,muskrat,orvoleoriginswasobserved.Riggsmonoclonalreactedstrongly
withcystsofhuman,beaver,dog,andratorigins.TheratGiardia
TABLE1.ImmunofluorescencereactivitypatternsobtainedwithGiardiacystsofvariousanimalsandhumansources,after
incubationwithfourantiG.murismonoclonalantibodiesandwiththeantiG.lambliamonoclonalofRiggs.

SourceofGiardia

Antibodies
47B

16C

1010B

211B

Riggs

Beaver(MB.C.gerb.)

++

Beaver(Manast.Wa.beav.)

++

Dog1(U.W.vivarium)

++

Dog2''

++

Dog3"

++

Dog4"

++

Human1(patient)

ND

++

Human2(TBrat)

++

Human3(H2Co.gerb.)

++

Human4(H3Co.gerb.)

++

Human5(H4Co.gerb.)

++

Human6(patient)

++

Human7(patient)

++

Human8(patient)

++

Norwayrat(U.W.1,2)

++

++

++

++

++

Mouse(G.murisorig.hamster)

++

++

++

++

Muskrat(U.W.1,2)

Muskrat(E.Wa.NA2)

Muskrat(MR6Alb.mice)

Vole(M.richardsoniMt.Rainier)

Vole(M.ochrogasterMo.)

Vole(C.gapperiAlb.mice)

++Strongreactionbetweenantibodyandcysts
Negativereaction
NDNodataavailable

TABLE2.ImmunofluorescencereactivityofMAb6E10(antiNorwayratGiardiacyst)withcystsofotheranimaland
humansources.
SourceofGiardia
Norwayrat(G.simoniUW2infeces)

IFAreaction(,+,++)
++

Mouse(G.murisorig.hamster)

Vole(M.richardsoniMt.Rainier)

Vole(C.gapperiC3Alb.mice)

Vole(M.ochrogasterMo.)

Muskrats(UW5,UW2)

Muskrat(MR6Alb.mice)

Beaver(MB.C.gerb.)

Beaver(ManastashE.Wa.gerb.)

Dog(D3Alb.gerb.)

Cattle(CW1Alb.gerb.)

++

Human(2patientsWa.)

Human(H2Coloradogerb.)

monoclonalboundonlytocystsofthehomologousratsourceisolateand,oddly,tocystsofthecattle(CW1)isolatefromAlberta.Wehaverecentlyfoundthat
neithertheantiratGiardiamonoclonalnortheRiggsmonoclonalwillbindtomuristypecyststhathavebeenrecoveredfromotherNorwayratstrappedfromthe
samelocationasthefirsttwo.
Alloftheantibodies,includingtheoneofRiggs,appearedtobindstronglytothecystwallofthoseisolatesthatreactedpositively,thepatternofimmunofluorescence
beingevenlydistributedaroundtheentiresurfaceofthecysts.Nointernalstructuresofthecystscouldbeseenfluorescing.Wedidnotobserveinanycasepartial
reactionsofcystpopulationswithanyoftheseantibodies.However,becausesomeofthetestcystpreparationswerecystscontainedinfecalslurries,itwas
impossibletoprecludebyimmunofluorescencethepossibilitythatafractionoftheGiardiacystsinsomepreparationswereunreactive.
Onseveraloccasions,cystsofeithermouseorratorigin,storedinadrystateontestslidesforsixweeksormoreat4Corthawedandrefrozenanumberoftimes
from75C,failedtoreactwiththeantibodies.Thesecystswouldcontinuetoreactpositively,althoughweakly,withpolyclonalrabbitantiseratoG.lambliacysts.
Also,cystsoftheH2isolate,harvestedfromgerbils,didnotreactwiththeRiggsantibodyafterstorageforseveralweeksin5%formalininPBSat4C.The
monoclonalantibodiesalsotendedtolosereactivitywiththeirhomologousantigenafterrepeatedfreezingandthawing.
Discussion
Thecrosstestingresultsshowthatantigenicdifferencesexistbetweencystsofvariousanimaloriginsandofhumanorigin,andsuggestthatitmayeventuallybefeasible
toidentifycystsinenvironmentalsamples(waterorfeces)accordingtotheiranimalsourcebytestingtheirimmunofluorescencereactivitywithabatteryofselected
monoclonalantibodiespreparedagainstcysts.Theresultsalsoshowthatvoleandmuskrat(i.e.

Page162

microtine)Giardiaisolatesseemtosegregateantigenicallyfromtheothersinthattheydidnotreactwithanyofthemonoclonalsused.Thefactthattheratandthe
mousesourceGiardiaisolatesbothbindtheantimurismonoclonalsindicatesantigenicrelatednesshowever,themuriscystsdidnotbindtheantiratGiardia
monoclonal.Incidentally,wehavealsofoundthatallofthecystisolatesshownintheTablesreactstronglyinindirectimmunofluorescencewithpolyclonalrabbit
antiserumagainsthumanG.lambliacysts(datanotincludedinTables),anobservationthatcorroboratespreviousresultsreportedbyourselves(26)andbySauch
(22).Therefore,whileitappearsthatallofthecystisolatessharesomeantigensoratleastsomeepitopes(determinants),crosstestingwithmonoclonalantibodieshas
revealedthatdifferencesexistinthedistributionofcertainepitopesbetweenisolates.
Westilldonotknowanythingofthephysicochemicalnatureoftheantigensrecognizedbythesemonoclonalantibodies,nordoweknowatthistimewhetherthefour
antimurismonoclonalsrecognizethesameordifferentepitopes.Thepreciseultrastructurallocationoftheantigensonorinthecystsalsoremainstobedetermined.
Theantigensmaybecomponentsofthethin(0.150.2mthick)filamentousoutercoatofthecystoroftheunderlyingoutercystmembrane,structuresdescribedby
Erlandsenetal.(7).Inaddition,theantigensmayhavesomebiochemicalorultrastructuralrelationshiptothechitinpresentinthecystwalls(29).Finally,thepossibility
thatsomeoftherecognizedantigens,particularlyiflocatedontheouterfilamentouscoatofthecyst,maybehostderivedcannotyetbediscounted.
TwootherimportantconsiderationsthatrelatetothepossiblepracticalutilityoftheseantibodiesindifferentiatingbetweenGiardiacystisolatesinnatureare(1)the
possiblecystantigenicvariabilitythatmayexistintheGiardiapopulationsfoundwithinonehostspecies,especiallyamonghostpopulationsfromgeographically
diverseoriginsand(2)thepossiblelossofimmunofluorescencereactivityofcystsafterstorageinvariousfixatives(e.g.formalin)orinwater,andatvarious
temperaturesandotherconditionsofstorageandshipping.
Inconclusion,itappearsthatantigenicdifferencesexistbetweenGiardiacystsofisolatesfromvariousanimalandhumansources.Theuseofmonoclonalantibodiesin
differentiatingbetweencystisolatesandthusinidentifyingtheanimalorhumansourceofanunknowntestcystisolatemayprovetobefeasible,althoughmuchmore
needstobedonetostudythevariablesmentionedaboveandtodevelopstandardpracticalmethodsfortesting.Wearecurrentlycontinuingthisworkbytryingto
developmonoclonalantibodiesagainstthemicrotinetypesofGiardiacysts(inparticular,cystsfrommuskratsinWashington)aswellasagainsthumanandbeaver
isolates,sothatamoreextensivebatteryofantibodiescanbeusedindevelopinganantibodytypingschemeforGiardiacystidentification.Whiletheidentification
schemeweareproposingwouldidentifytheGiardiaisolatesastohostspeciesorigin,itispossible(providedextensiveantigenicvariationdoesnotoccurwithin
isolates)thatthisinformationonantigenorepitopedistributionamongcystisolatesmaybehelpfulinestablishingspeciesdesignationswithinthegenusGiardia.
Acknowledgements
ThisresearchwasfundedwhollybytheUnitedStatesEnvironmentalProtectionAgency,OfficeofExploratoryResearch,undercooperativeagreement#R81197001
toH.H.S.thisreportdoesnothowever,necessarilyreflecttheviewsofthatagencyandnoofficialendorsementshouldbeinferred.Theauthorsgratefully
acknowledgetheassistanceoftheDepartmentofGameofthestateofWashingtoninprovidingacollectingpermitforcapturingsomeoftheanimalsusedinthiswork.
LiteratureCited
1.Bemrick,W.J.1984.Perspectivesonthetransmissionofgiardiasis,In:Giardiaandgiardiasis:biology,pathogenesis,andepidemiology.S.L.Erlandsenand
E.A.Meyer(eds).PlenumPress.NewYork.pp.379400.
2.Bertram,M.A.,Meyer,E.A.,Lile,J.D.,andS.A.Morse.1983.AcomparisonofisozymesoffiveaxenicGiardiaisolates.J.Parasitol.69:793801.
3.Clark,G.W.,andR.E.Pacha.1984.AnimalsassociatedwithmountainstreamsandmeadowsasreservoirsofGiardia.Abstract#30,59thAnnualmeeting,
AmericanSocietyofParasitologists,Snowbird,Utah.
4.Craun,G.F.1984.Waterborneoutbreaksofgiardiasis:currentstatus,In:Giardiaandgiardiasis:biology,pathogenesis,andepidemiology.S.L.Erlandsenand
E.A.Meyer(eds.).PlenumPress.NewYork.pp.243261.
5.Davies,R.B.,andC.P.Hibler.1979.AnimalreservoirsandcrossspeciestransmissionofGiardia,In:Waterbornetransmissionofgiardiasis,proceedingsofa
symposium.U.S.HealthEffectsResearchLaboratory,U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio.pp.104126.
6.Dykes,A.C.,D.D.Juranek,R.A.Lorenz,S.Sinclair,W.Jakubowski,andR.Davies.1980.Municipalwaterbornegiardiasis:anepidemiologicinvestigation.Ann.
Int.Med.92:165170.
7.Erlandsen,S.L.,D.G.Schupp,M.M.Januschka,andW.J.Bemrick.1986.UltrastructureofGiardiaspp.cysts:developmentandformationofmembranesinthe
cystwallandtheirlossduringexcystation.Abstract#95,61stannualmeeting,AmericanSocietyofParasitologists.Denver,CO.
8.Frost,F.,B.Plan,andB.Leichty.1980.Giardiaprevalenceincommerciallytrappedmammals.J.Environ.Hlth.42:245249.
9.Galfre,G.,andC.Milstein.1981.Preparationofmonoclonalantibodies:strategiesandprocedures.Meth.Enzymol.73:346.
10.Hausler,W.J.,Jr.,W.E.Davis,andN.P.Moyer.1984.DevelopmentandtestingofafiltersystemforisolationofGiardialambliacystsfromwater.Appl.
Environ.Microbiol.47:13461347.
11.Hopkins,R.S.,P.Shillam,B.Gaspard,L.Eisnach,andR.J.Karlin.1985.WaterbornediseaseinColorado:threeyears'surveillanceand18outbreaks.Am.J.
Pub.Hlth.75:254257.
12.Jakubowski,W..1984.DetectionofGiardiacystsindrinkingwater:stateoftheart.In:Giardiaandgiardiasis:biology,pathogenesis,andepidemiology.
S.L.ErlandsenandE.A.Meyer(eds).PlenumPress.NewYork.pp.286363.

Page163

13.Korman,S.H.,S.M.LeBlancq,D.T.Spira,J.ElOn,R.M.Reifen,andR.J.Deckelbaum.1986.Giardialamblia:identificationofdifferentstrainsfromman.Z.
Parasitenkd.72:173180.
14.Kulda,J.,andE.Nohynkova.1978.Flagellatesofthehumanintestineandofintestinesofotherspecies.In:Parasiticprotozoa,Vol.2.J.P.Kreiser(ed.).
AcademicPress.NewYork.pp.1138.
15.Lopez,C.E.,A.C.Dykes,D.D.Juranek,S.P.Sinclair,J.M.Conn,R.W.Christie,E.C.Lippy,M.G.Schultz,andM.H.Mires.1980.Waterbornegiardiasis:a
communitywideoutbreakofdisease,andahighrateofasymptomaticinfection.Am.J.Epidemiol.112:495507.
16.Nash,T.E.,andD.B.Keister.1985.Differencesinexcretorysecretoryproductsandsurfaceantigensamong19isolatesofGiardia.J.Inf.Dis.152:11661171.
17.Nash,T.E.,T.McCutchan,D.Keister,J.B.Dame,J.D.Conrad,andF.D.Gillin.1985.RestrictionendonucleaseanalysisofDNAfrom15Giardiaisolates
obtainedfromhumansandanimals.J.Inf.Dis.152:6473.
18.Nash,T.E.,andA.Aggarwal.1986.CytotoxicityofmonoclonalantibodiestoasubsetofGiardiaisolates.J.Immunol.136:26282632.
19.NationalParkService.1985.FieldsurveyofGiardiainstreamsandwildlifeoftheGlacierGorgeandLochValebasins,RockyMountainNationalPark.Natural
ResourcesReportSeries#853.WaterResourcesDivision,AppliedResearchBranch,NationalParkService,U.S.,FortCollins,CO.
20.Pacha,R.E.,G.W.Clark,andE.A.Williams.1985.OccurrenceofCampylobacterjejuniandGiardiaspeciesinmuskrat(Ondatrazibethica).Appl.Environ.
Microbiol.50:177178.
21.Riggs,J.L.1984.GiardiaMethodsWorkshopWaterQualityTechnologyConference.AmericanWaterWorksAssoc..Denver,CO.pp.496.
22.Sauch,J.F.1985.UseofimmunofluorescenceandphasecontrastmicroscopyfordetectionandidentificationofGiardiacystsinwatersamples.Appl.Environ.
Microbiol.50:14341438.
23.Schaefer,F.W.,III,E.W.Rice,andJ.C.Hoff.1984.FactorspromotinginvitroexcystationofGiardiamuriscysts.Trans.Roy.Soc.Trop.Med.Hyg.
78:795800.
24.Smith,P.D.,F.G.Gillin,N.A.Kaushal,andT.E.Nash.1982.AntigenicanalysisofGiardialambliafromAfghanistan,PuertoRico,Ecuador,andOregon.Inf.
Immun.36:714719.
25.Smith,R.D.,N.S.Stimler,J.A.Sauch,andF.W.Schaefer,III.1983.MonoclonalantibodiestoGiardialamblia:potentialforidentificationandspeciation.
Abstract#173,58thannualmeeting,AmericanSocietyofParasitologists,SanAntonio,TX.
26.Stibbs,H.H.1985.MorphologicalandantigeniccomparisonsofGiardiaisolatedfrommuskrat,beaver,mice,andman.Abstract,TenthInternationalSymposium
onIntestinalMicroecology.SocietyforIntestinalMicrobialEcologyandDisease.UniversityofMinnesota,Minneapolis,MN.
27.Wallis,P.M.,J.M.BuchananMappin,G.M.Faubert,andM.Belosevic.1984.ReservoirsofGiardiaspp.insouthwesternAlberta.J.Wildl.Dis.20:279283.
28.Wallis,P.M.,R.M.Zammuto,andJ.M.BuchananMappin.1986.CystsofGiardiaspp.inmammalsandsurfacewatersinsouthwesternAlberta.J.Wildl.Dis.
22:115118.
29.Ward,H.D.,J.Alroy,B.I.Lev,G.T.Keusch,andM.E.A.Pereira.1985.IdentificationofchitinasastructuralcomponentofGiardiacysts.Inf.Immun.49:629
634.
30.Woo,P.K.1984.EvidenceforanimalreservoirsandtransmissionofGiardiainfectionbetweenanimalspecies.In:Giardiaandgiardiasis:biology,
pathogenesis,andepidemiology.S.L.ErlandsenandE.A.Meyer(eds).PlenumPress,NewYork.pp.341364.

Page165

ComparisonofGiardiaIsolatesbyDNADNAHybridization
A.Uji*,P.M.WallisandW.M.Wenman
DepartmentsofPediatricsandMicrobiology,UniversityofAlberta,Edmonton,Alberta,Canada.
WehaveanalyzedtheDNArestrictionendonucleaseprofilesofseveralAlbertaGiardiaduodenalisisolatesbySouthernblothybridizations.WiththeexceptionofstrainWB,these
GiardiastrainswereisolatedfromasinglegeographicareainsoutheasternAlberta.Thefollowingstrainswereanalyzed(originalhostsareshowninparentheses):H7(human),H8
(human),WB(human),B5(beaver),PB1(beaver),D3(dog),MR4(muskrat)andS1(sheep).Trophozoiteswereharvestedfrom1litreculturesandlysedusing0.5%SDSand500
g/mLproteinaseK.DNAwasextractedwithchloroformisoamylalcoholandphenolandthentreatedwithRNAase.TheDNAfromall8G.duodenalisstrainswasdigestedwith
BamH1andelectrophoresedin1%agarosegels.TheseparatedDNAfragmentswerethenSouthernblottedontonitrocellulosefiltersandhybridizedwith32PlabelledDNAprobes
fromstrainsD3orH7.TheSouthernblotpatternsofallstrainsanalyzed,whetherprobedunderstringentorlessstringentconditions,weremarkedlysimilar.Likewise,little
differencewasdetectedbetweenthepatternusingthedogisolateprobe(D3)orthehumanGiardiaprobe(H7).ThesedatasuggestthatG.duodenalisstrainsisolatedfromaspecific
geographicarea,regardlessoftheiroriginalhost,arecloselyrelatedgenetically.

Introduction
Giardiaduodenalis,anintestinalprotozoan,isconsideredtobeamajorcauseofdiarrheainmanycountries.IntheUnitedStates,G.duodenaliswasthemost
commonlyidentifiedpathogeninwaterborneoutbreaksduring1972and1981(5).InAlberta,Canada,giardiasisisthemostfrequentlyreportedendemicintestinal
infection(1).
Thisinfectioncanbetransmitteddirectlyfrominfectedindividuals(11),orthroughcontaminatedfoods(2)orwater(5).Largecommunityoutbreaksofgiardiasishave
beencausedbywaterbornetransmission(5).Beavershavebeenimplicatedinwaterbornetransmissionasareservoirhostforhumaninfection(3),butthemechanism
ofGiardiatransmissionbetweenanimalspeciesisnotclear.Inordertoinvestigatethepossibilityofcrossinfectionamongdifferentanimalspecies,itisimportantto
definehostspecificityandtoclassifyspecieswithinGiardia.
Earlierworkershaveinvestigatedbiochemicaldifferencesinisolatesobtainedfromvariousmammalianspeciesinanefforttoclassifythem.Nashetal.differentiated15
Giardiaisolatesinto9groupsbySouthernblotanalysisofchromosomalDNAusingrecombinantplasmidscontainingGiardiaDNAasprobes(10).Bertrametal.
differentiated6isolatesinto3groupsbycomparingtheelectrophoreticpatternsofGiardiaenzymes(4).ThesestudieswereperformedonGiardiaisolatesobtained
fromgeographicallydifferentareas.
Inthisreport,wecomparedthegeneticsimilaritybetweenGiardiaisolatesobtainedfromvariousmammalianspeciesfoundinonespecificgeographicarea.Giardia
DNAwasanalyzedbyrestrictionendonucleasedigestionandbySouthernblottechniqueusing32PlabelledwholeDNAasprobes.Thesedatamayhaverelevancein
theclassifictionofGiardiastrainsandalsoforunderstandingthepossiblecrosstransmissionofgiardiasiswithingeographicareas.
MaterialsandMethods
GiardiaIsolates
TheoriginalhostsandlocationoftheGiardiastrainsanalyzedinthisstudyareshowninTable1.WB,ahumanisolate(ATCC30957)originallyestablishedinculture
attheNationalInstitutesofHealth,wasobtainedfromtheAmericanTypeCultureCollection.Otherisolateswereestablishedinculturefromcystsextractedfrom
faecesof5differentmammals:beaver(B5,PB1),muskrat(MR4),dog(D3),sheep(S1),andhuman(H7,H8),foundinasinglegeographicarea.Thecystsofall
isolatesexceptPB1wereusedtoinfectmongoliangerbils(Merionesunguiculatus).Trophozoiteswereextractedfromtheirduodenumsandcultureswere
establishedbymethodsreportedpreviously(13).Adeermouse(Peromyscusmaniculatus),livetrappedandclearedofintestinalprotozoawithmetronidazole,was
usedfortheextractionofPB1trophozoites.
TABLE1.SourcesofG.duodenalisisolates.
Strain

Original
Host

Date

WB

Human

H7

Human

850803

Calgary

H8

Human

Calgary

B5

Beaver

850619

PB1

Beaver

MR4

Location
Afghanistan

SibbaldMeadowsPond
KananaskisCountry,65km
westofCalgary

841123

LuskCr.Pond
KananaskisCountry,80km
westofCalgary

Muskrat

850920

SameasB5

S1

Sheep(domestic)

850702

FarmnearStrathmore,45km
eastofCalgary

D3

Dog

840816

Calgaryanimalshelter

*Correspondingauthor.

Page166

Figure1.
AgaroseelectrophoresisofPstIrestrictedDNAfrom8different
G.duodenalisisolatesstainedwithethidiumbromide.Lane1,
Molecularsizestandards(HindIIIdigestedlambdaphageDNA)
lane2,WBlane3,H7lane4H8lane5,B5
lane6,PB1lane7,MR4lane8,S1lane9,D3.

IsolateswereculturedinmodifiedTYIS33medium(6)usingcaseindigestpeptoneinsteadoftrypticasepeptone,supplementedwithpenicillin(200g/mL),
gentamicin(200g/mL)andticarcillinclavulanicacid(500g/mL).Theorganismsweregrownat37Cfor7296h.
DNAPreparation
DNAwasisolatedfromoneLcultureofeachGiardiaisolate.Thecellswereharvestedbychillingtheculturebottlesinicefor10minandthencentrifugingthe
mediumat1000gfor10min.AfterwashingtwicebycentrifugationincoldPBS(pH7.5),theharvestedcellsweresuspendedin2to4mLof50mMTrisHCl
(pH7.9)containing50mMNaCland10mMEDTA,andstoredat20C.TwelvemLoflysissolution[0.1MTrisHCl(pH7.9),0.1MNaCl,0.05MEDTA,0.5%
SDS,and500g/mLproteinaseK)wasaddedtothethawedcellsuspension,andthemixturewasincubatedovernightat37C.Thelysatewasmixedwithonehalf
volumeofredistilledphenolsaturatedwith10mMTrisHCl(pH7.9)andonehalfvolumeofchloroformisoamylalcohol(24:1),androtatedgentlyfor30minatroom
temperature.Toseparatephases,thismixturewascentrifugedfor10minat1000g,andtheaqueousphasewastransferedintoanewtube.Anequalvolumeof
chloroformisoamylalcohol(24:1)wasaddedtotheaqueousphaseandrotatedgentlyfor15minatroomtemperature.Aftercentrifuging,heatinactivatedRNase
(100g/mL)wasaddedtotheaqueousphaseandincubatedovernightat37C.ThenproteinaseK(100g/mL)wasaddedandincubatedfor2hat37C.The
mixturewasextractedwithanequalvolumeofchloroformisoamylalcohol(24:1)andDNAwasethanolprecipitated.Theprecipitatewasdissolvedin1mLof10mM
TrisHCl(pH7.5)containing10mMNaCland1mMEDTA,anddialysedovernightagainstthesamebuffer.TheDNAwasethanolprecipitatedagainand
resuspendedin0.5mLto1mLofthesamebuffer.
DNAAnalysis
Thefollowing4restrictionendonucleaseswereusedtodigestDNA:BamHI(N.EnglandBiolabs,Inc.),EcoRI(N.EnglandBiolabs,Inc.),HindIII(Boehringer
Mannheim),PstI(BoehringerMannheim).Approximately1gofDNApersamplewasdigestedwith2to5unitsofenzyme(7)underconditionssuggestedbythe
manufacturer,andelectrophoresedovernightat30Vina1%agarosegel(9).Afterethidiumbromidestainingofthegel,DNAwastransferredtonitrocellulose
accordingtothemethodofSouthern(12).
32

PlabelledwholeDNAprobeswerepreparedfromD3,H7,orWBDNA(8).Specificactivityofeachprobewas4.69107CPM/mLfortheD3probe,1.76
107CPM/mLfortheH7probe,and1.14107fortheWBprobe.
TheDNArestrictedwitheachof4endonucleaseswashybridizedwithD3,H7,andWBDNAprobesusingstringentconditionsandtheDNAdigestedwithBamHI
orEcoRIwashybridizedwiththeD3DNAprobeusinglessstringentconditions.Understringentconditions,DNAfragmentswerehybridizedwith106107CPMof
probeat37Covernightin50%formamide.Underlessstringentconditions,DNAfragmentswerehybridizedin25%formamide.Hybridizedfragmentsweredetected
byautoradiography.
ResultsandDiscussion
TheethidiumbromidestainingofDNAfragmentsrestrictedwithPstI(Figure1),BamHI(Figure2),EcoRIorHindIIIfailedtodifferentiatetheGiardiastrainswe
studied.Nashetal(10)alsonotedasimilarrestrictionendonucleasedigestionprofileamongmostoftheisolateswhichtheyanalyzed.TheGiardiagenomeis
sufficientlycomplextorenderdifferentiationonthebasisofrestrictionendonucleaseanalysisaloneimpractical.Therefore,ourinabilitytodistinguishisolatesbyagarose
gelelectrophoresismayreflecttechnicallimitationsasmuchasit

Figure2.
AgaroseelectrophoresisofBamHIrestrictedDNAfrom8
differentG.duodenalisisolatesstainedwithethidiumbromide.
LanesarethesameasinFigure1.

Page167

Figure3.
SouthernblothybridizationofHindIIIrestrictedDNAfrom8
differentG.duodenalisisolateshybridizedto32Plabelled
WBDNAprobe.LanesarethesameasinFigure1.

mayindicatestrainsimilarity.
Southernblotanalysiswasperformedusingthesamestrainsandrestrictionenzymes.WeusedwholenicktranslatedDNAfromtwohumanstrains(WB,H7)andone
dogisolate(D3)asprobes.TheDNAfromdifferentisolatesdisplayedverysimilarhybridizationpatterns,regardlesswhichprobewasemployed.Likwise,the
fragmentsgeneratedbythe4restrictionendonucleasesutilizedwereseldomuniquewithrespecttooriginalhost.Thiswasthecasewhetherstringentorlessstringent
hybridizationconditionswereemployed.HindIIIappearedtobethemostusefulenzymetodifferentiateGiardiaDNAsamples(Figure3).Whilethisenzyme
producedatleast10commonbandsonSouthernblot,aminorfragment(~5kb)appearedtobepresentinthebeaver,muskratandsheepstrains,butabsentinDNA
fromthedogand3humanstrains(H7,H8,andWB).ThemajorhybridizationbandcomontoallDNAsamplesanalyzedwasapproximately4.0kbintheHindIII
digest.Theother3restrictionsendonucleasesutilizeddidnotclearlydifferentiatetheDNAsamples,regardlesswhethertheH7,WBorD3probeswerereactedwith
theseblots.TheseresultssuggestthatthereisverylittlegeneticdisparityamongtheGiardiastrainswhichweanalyzed.
Nashetal(10)reportedtheSouthernblotanalysisof15Giardiaisolatesfromavarietyofgeographiclocations.Whilemostrestrictionenzymehybridizationpatterns
didnotexhibitanydifferences,theywereabletogrouptheirGiardiastrainsinto9groups.Thesedifferencesfromourresultsmayreflectthenumberofstrains
analyzed,thevariedgeographicoriginofstrainsorthemuchlargernumberofrestrictionenzymes(sixteen)whichtheseworkersemployedintheirstudy.However,our
DNAhybridizationresultsareconsistentwiththehighdegreeofantigenicrelatednessshownbytheGiardiastrainswhichweanalyzed(14).
TheseresultssuggestthatG.duodenalisisolatesobtainedfromdifferentmammalsfoundinonespecificgeographicareaaregeneticallyveryclosetoeachother
regardlessofthespeciesoftheiroriginalhosts,andthedifferenceinDNAsequenceissosmallthatwholeDNAprobescannotdistinguishit.ItislikelythatsomeG.
duodenalisstrainshaveabroadhostspecificityandcancauseoutbreaksofgiardiasisbytransmissionamongdifferentanimalspecies.
LiteratureCited
1.AlbertaSocialServicesandCommunityHealthEpidemiol.Rpt.,1981.9:14
2.Barnard,R.J.andG.J.Jackson.1984.Giardialamblia:Thetransferofhumaninfectionsbyfoods.In:Giardiaandgiardiasis.S.L.ErlandsenandE.A.Meyer
(eds).PlenumPress,NewYork.pp.365378.
3.Bemrick,W.J.1984.Someperspectivesonthetransmissionofgiardiasis.In:Giardiaandgiardiasis.S.L.ErlandsenandE.A.Meyer(eds).PlenumPress,New
York.pp.379400.
4.Bertram,M.A.,Meyer,E.A.,Lile,J.D.,andS.A.Morse.1983.AcomparisonofisozymesoffiveaxenicGiardiaisolates.J.Parasitol.69:793801.
5.Craun,G.F.1984.Waterborneoutbreaksofgiardiasis:currentstatus.In:Giardiaandgiardiasis.S.L.ErlandsenandE.A.Meyer(eds).PlenumPress,New
York.pp.243261.
6.Keister,D.B.1983.AxeniccultureofGiardialambliainTYIS33mediumsupplementedwithbile.Trans.Roy.Ssoc.Trop.Med.Hyg.77:487488.
7.Maniatis,T.,Fritsch,E.F.,andJ.Sambrook.1982.Enzymesusedinmolecularcloning.In:Molecularcloningalaboratorymanual.ColdSpringHarbor
Laboratory,NewYork.pp.104106.
8.Maniatis,T.,Fritsch,E.F.,andJ.Sambrook.1982.Enzymesusedinmolecularcloning.In:Molecularcloningalaboratorymanual.ColdSpringHarbor
Laboratory,NewYork.pp.109112.
9.Maniatis,T.,Fritsch,E.F.,andJ.Sambrook.1982.Gelelectrophoresis.In:Molecularcloningalaboratorymanual.ColdSpringHarborLaboratory,newYork.
pp.150163.
10.Nash,T.E.,McCutchan,T.,Keister,D.,Dame,J.B.,Conrad,J.D.,andF.D.Gillin.1985.RestrictionendonucleaseanalysisofDNAfrom15Giardiaisolates
obtainedfromhumansandanimals.J.Infect.Dis.152:6473.
11.Owen,R.L.1984.Directfecaloraltransmissionofgiardiasis.In:Giardiaandgiardiasis.S.L.ErlandsenandE.A.Meyer(eds).PlenumPress,NewYork.pp.
329339.
12.Southern,E.M.1975.DetectionofspecificsequencesamongDNAfragmentsseparatedbygelelectrophoresis.J.Mol.Biol.98:503517.
13.Wallis,P.M.andH.M.Wallis.1986.ExcystationandculturingofhumanandanimalGiardiaspp.byusinggerbilsandTYIS33medium.Appl.Environ.
Microbiol.51:647:651.
14.Wenman,W.M.,Meuser,R.U.andP.M.Wallis.1986.AntigenicanalysisofGiardiaduodenalisstrainsisolatedinAlberta.Can.J.Microbiol.32:926929.

Page169

DifferentiationofGiardiaduodenalisfromGiardiamurisbyImmobilizationinVariousSera
D.L.LehmannandP.M.Wallis*
KananaskisCentre,BioSciences042,UniversityofCalgary,
2500UniversityDriveN.W.,Calgary,Alberta,CanadaT2N1N4.
ImmobilizationoftrophozoitesbyantiserumwasusedasacriteriontodistinguishGiardiamuriscollectedfromnaturallyinfectedvolesfromGiardiaduodenalistrophozoites
obtainedfromculturedstocks.G.muristrophozoiteswereobtainedfromtworedbackedvoles(Cleithrionomysgapperi)andthreemeadowvoles(Microtuspennsylvanicus).
CulturedG.duodenalisstrainswereobtainedfromhumans(WB,H7),beaver(PB1,B5),adog(D3),andasheep(S1).AntiserawasraisedinrabbitsagainstG.muristrophozoites
fromaredbackedvoleandG.duodenalistrophozoitesfromculturedhuman(WB)trophozoites.Serumfromanuninfectedhumanandpoolednormalrabbitserawereusedas
controls.G.muristrophozoiteswereimmobilizedbyrabbitantiWBserum,antiWBserumadsorbedonWBandG.muris,rabbitantiG.murisserum,antiG.murisadsorbedonG.
murisandnormalrabbitandnormalhumansera.G.duodenalistrophozoiteswereimmobilizedonlybyantiWBserumandantiWBserumadsorbedonG.muris.Immobilizationwas
showntobecomplementmediated.ThemaincontributortoG.murisimmobilizationwasfoundtobeactivatedserum,rabbitorhuman.Suchasimpletestcouldbeusedto
distinguishhumaninfectivestrainswithaconcurrentidentificationofpotentialreservoirhosts.

Introduction
ThedifferentiationofvariousspeciesofGiardiahaslongbeenaproblemandofspecialinteresthavebeenthedifferencesbetweentheGiardiaofrodentsandthose
ofothermammals.Filice(7)recognizedtwospeciesofGiardiathatwerefoundinmammalsG.duodenalisandG.muriswhichhedistinguishedprimarilyonthe
morphologyofthemedianbody.ThismethodhasbeenusedbyGrantandWoo(9)andotherworkerstoidentifyindividualGiardiaisolatesbutthemethodisdifficult
toreproduceandhasnotproventobepractical(4).Anadditionalpointofdifferentiationisthatthemembersoftheduodenalisgroupcanbeculturedwhilethemuris
groupflagellateshavesofarproventoberefractivetocultivation(13).
Severalworkershaveshownthatbothmononuclearleukocytes(11,16,17)andserum(10)canbelethaltoGiardiatrophozoites.PearsonandSteigbigel(15)
demonstratedsimilarresultsusingLeishmaniadonovani.Vinayaketal.(19)showedthatserumantibodieswillreactwithGiardiacystsinimmunodiffusiontests.
ItisthecontentionofStevens(18)thatnoaccurateauthenticationofGiardiaspeciesispossibleuntilappropriateimmunologicaltechniquesareestablished.
Immunologicalstudiesusingsodiumdodecylsulphatepolyacrylamidegelelectrophoresisanalysishavebeencarriedoutbyvariousworkers(6,12,16,21).Thesedata
haveshownthatbothdifferencesandsimilaritiescanbefoundbetweenmajorantigensofvariousstrainsofG.duodenalis.Detailedantigenanalysisrequiresvery
cleanpreparationsthatarefreeofcontaminatingprotein.Sofarthesehaveonlybeenavailablefromculturedtrophozoitesand,becausenoonehasyetsucceededin
culturingG.muris,thesestudieshavebeenrestrictedtoG.duodenalis.ThispaperisdevotedtothedemonstrationofamethodtodistinguishbetweenG.
dudodenalisandG.murisbasedonthedegreeofimmobilizationfollowingexposuretovariousserainvitro.
TABLE1.SourcesofstrainsofG.duodenalis
Strain

Original
Host

Date

B5

Beaver

850619

SibbaldMeadowsPond
KananaskisCountry65km
westofCalgary

PB1

Beaver

841123

LuskCr.PondKananaskis
Country80kmwestof
Calgary

D3

Dog

840816

CalgaryAnimalShelter

S1

Sheep(domestic)

850702

FarmnearStrathmore45km
eastofCalgary

H7

Human

850803

Calgary

WB

Human

840615

ObtainedfromG.Faubert
ATCCNo.30957

*Correspondingauthor.

Location

Page170
TABLE2.%ImmobilizationofG.duodenalistrophozoitesafter30minutes.
Antibody

WB

H7

PB1

B5

D3

S1

NormalAntiWB,1:1dilution

30

18

17

28

11

47

AntiWBadsorbedonG.muris,1:1
dilution

30

15

13

31

44

AntiWBadsorbedonWB1:1dilution

AntiWBinactivated

NormalRabbitserum

Rabbitseruminactivated

NormalHumanserum

Humanseruminactivated

Methods
Strains
SixstainsofduodenalisgroupGiardiaand25strainsofmurisgroupflagellateswereemployedduringthisstudy.Giardiafromthefollowingsourceswereassumed
tobelongtotheduodenalisgroupbasedontheiroriginandtheirculturability:HumanWB,HumanH7,BeaverPB1,BeaverB5,DogD3,andSheepS1(Table1).
Attemptstocharacterizethesetrophozoitesbasedontheshapeofthemedianbodywereinconclusiveandnoconfidencecouldbeplacedinthismethod(4).Giardia
muriswereobtainedfromacolonyofSwissWebstermiceattheUniversityofCalgary(GM2)andfromvolescollected(Microtispennsylvanicusand
Cleithrionomysgapperi)inthevicinityoftheKananaskisFieldStation(80kmwestofCalgary,AB,intheEastSlopesoftheRockyMountains).
TrophozoitePreparation
ForimmobilizationstudiesG.duodenaliswereconcentratedfromculturesinTYIS33culturemedium(23dayculturesin1012mLmedium)bycentrifugationat
500Xgfor5minutesafterwhichtheywereresuspendedin2mLofRPMI1640medium(Gibco).MurineGiardiawererinsedfromtheintestineofthehostwith
RPMI1640,centrifugeconcentratedandresuspendedinthesamefluid.Controlswererunforeachportionoftheexperimentalworkbymixingtrophozoiteswith
RPMIandmeasuringimmobilizationovera30minuteperiod.
AntibodyProduction
AntibodiestoG.duodenalisandG.muris(fromC.gapperi)trophozoiteswereproducedinadult,malealbinolaboratoryrabbitsby4intraperitonealinjectionsof
intact,livingtrophozoitesatweeklyintervals.Approximately5105trophozoiteswereusedforeachinjectionandthefirstthreeinoculationswerediluted1:1with
Freund'scompleteadjuvant.Oneweekafterthefinalinjection,therabbitwasanaesthetizedwithetherandexsanguinated.Thebloodwasallowedtoclotincentrifuge
tubesfor30minutesandthencentrifugedat15,000Xgfor10minutesinanIECultracentrifuge.Theserumwasstoredat60C.Normalhumanserumwasobtained
fromoneoftheauthors(PMW)whowasnotinfectedwithG.duodenalisasshownby5negativestoolexaminations.
AntibodyInactivationandAdsorption
Serumwasinactivatedbyplacingitinawaterbathat56Cfor30minutes.TheadsorptionofGiardiaantibodywasaccomplishedbymixingglassbeadfragmented
Giardiatrophozoites(concentratedfromculturemediumorhost)withequalpartsofantiserum.Themixturewasplacedonashakerfor30minutes,afterwhichthe
suspensionwascentrifugedat2000rpmfor2minutes,complement(Difco)replaced,andthesupernatentretainedforexperimentalwork.
ImmobilizationStudies
ImmobilizationstudieswereperformedbymixingonedropofGiardiasuspensionwithanequalsizeddropofserumandcoveringwithaslip.Normalantibodywas
diluted1:1sothattheconcentrationofallserumfactorswouldbesimilartothatfoundinadsorbedantibodywhichwasdilutedequallywithasuspensionoffragmented
G.muris.Preparationswereobservedinitiallyandafter30minutesunderphasecontrastat400XusingaZeissphotomicroscope.Immobilizationwasdeterminedby
movement,orlackofmovement,oftheflagellae.
Observationoftheparasiteserumpreparationbeganatonecorneroftheslideandprogresseduntil50organismshadbeenseenandtheproportionoflivingforms
determined.Countingrequired24minutes.Tworeplicatesofa50parasitecountwerecarriedoutwithatotalof100flagellatesobserved.Thismethodwassimilarto
theimmobilizationtest(TPI)usedtodetectserumantibodiesagainstTreponemapallidum(8).Hilletal.(10)usedacloselyrelatedmethodtostudytheeffectsof
humanserumontrophozoitesandotherauthors(11,16)havealsousedinvitromethodstoinvestigatetheinteractionsbetweenleukocytesandGiardiatrophozoites.
Results
Controls
Inallcases,95100%oftrophozoitessuspendedinRPMIalonewereactiveafter30minutes(5to0%immobilization).Additionally,parasitesmixedwithequal
volumesof1:10complementsalineresultedin95100%trophozoiteactivityafter30minutes.
ImmobilizationofG.duodenalisTrophozoites
VerylittledifferencewasobservedbetweenG.duodenalis(WB)antibodyadsorbedonfragmentedG.muris(fromMicrotispennsylvanicus)andnormalantiWB
(1:1dilution).Bothcausedbetween8and47%immobilizationofthe6strainsofG.duodenalisandagreementwithinstrainswasclose(Table2).AntiWBantibody
adsorbedonhomologousWBtrophozoitescausedalmostnoimmobilizationasdidinactivatedantiWBantiserum.Normalandinactivatedrabbitserumwereboth
inactiveagainstG.duodenalistrophozoitesaswerenormalandinactivatedhumanserum.
ImmobilizationofG.murisTrophozoites
Mostmurisgroupparasiteswerecompletelyimmobilized(Table3)bynormalantiWBrabbitserum,antiWBadsorbedonG.muristrophozoitesandantiWB
adsorbed
TABLE3.%ImmobilizationofG.muristrophozoitesafter30minutes.
C.gapperi

Antibody

M.pennsylvanicus
2

NormalAntiWB1:1dilution

100

87

100

100

100

AntiWBadsorbedonG.muris,1:1dilution

85

82

100

100

100

AntiWBadsorbedonWB1:1dilution

85

82

AntiWBinactivated

Normalrabbitserum

85

90

Rabbitseruminactivated

Normalhumanserum

100

71

92

Humanseruminactivated

Page171
TABLE4.%ImmobilizationofGiardiaspp.trophozoitesafter30minutesusingantibodiesraisedagainstG.murisfromC.
gapperi.
C.gapperi

M.pennsylvanicus

H7

B5

D3

AntiG.muris,1:1dilution

100

100

AntiG.muris,inactivated

AntiG.murisadsorbedwithG.muris1:1
dilution

95

91

Antibody

onWBtrophozoites.Highlevelsofinactivationwerealsoobservedwithnormalrabbitandnormalhumanserumwhileinactivatedhumanandrabbitserumproduced
noeffect.Immobilizationwassometimesaccompaniedbydistortionofthemainbodyoftheparasitewhichdidnotappeartobereversible.
ImmobilizationofG.murisandG.duodenalisTrophozoitesbyAntiG.murisSerum
AntibodyinrabbitserumtoG.muris(fromC.gapperi)whenexposedtoG.muristrophozoitescompletelyimmobilizedtheorganismswithin30minutes,and,
reductionsofmotilityby95and91%werenotedafteradsorptionoftheantibodyonfragmentedG.muris(Table4).Afterheatinactivationofantiserumno
immobilizationeffectwasnotedonG.muris.Normal,adsorbed,or,inactivatedG.murisantibodydidnotimmobilizeanyoftheduodenalisgroupstrains(Table4).
Discussion
Thedataindicatethat,forthestrainsofGiardiainvestigated,therearedistinctdifferencesbetweentheduodenalisgroupandthemurisgroupflagellates.These
differencesare:
1.G.duodenalisantibody(inrabbitserum)isnotadsorbedontoG.muristrophozoites.
2.G.murisantibody(inrabbitserum)willimmobilizemurineGiardiabuthasnoeffectuponduodenalisgroupflagellates.
3.G.murisantibodyisresponsibleforasmallpercentageofimmobilizedG.muristhemaincontributortoimmobilizationofG.murisistherabbitserumvehicle.
4.G.duodenalisantibodyproducesvaryingdegreesofimmobilizationof6strainsofduodenalisgroupparasitesbutdoesnotappreciablyinfluencemurisgroup
GiardiatheobservedimmobilizationofG.murisbyG.duodenalisantibodyis,inthemain,attributedtotherabbitserumbearingtheantibodyasimmobilizationwas
foundaftertheG.duodenalisantibodyhasbeenadsorbed.
5.Normalserafromtherabbitanduninfectedhumanimmobilizesfrom71100%ofmurineparasiteswhileduodenalisgroupflagellatesarealmostcompletelymotile.
Rabbitserum,fromwhichG.murisorG.duodenalisantibodyhasbeenadsorbedbyhomologousantigen,reactsasnormalrabbitserum.
HeatinactivationnegatestheimmobilizingeffectofvariousserauponG.muriswiththeimplicationthattheimmobilizingactionoftheseraiscomplementmediated.
ThisisinagreementwiththeresultsofBelosevicandFaubert(3)whofoundthatimmuneserumfrommicewouldlyseG.murisinvitroandthatthiseffect
disappearedwithinactivation.Becausecomplementisnecessaryforimmobilization,itmustbepresumedthatcertainimmunoglobulinsareinvolved.Inasmuchasthere
isnoadsorptionofhumanGiardiaantibodytomurineGiardiaantigen(Table3),itisspeculatedthatcrossreactingantibodiesmaybepresentinnormalhumanand
rabbitsera(2,10,15).
ThesixstrainsofduodenalisgroupGiardiainvestigatedareimmobilizedtovariousdegreesbyG.duodenalisantibody(inrabbitserum)andtheactionis
complementmediated.TheduodenalisgroupstrainsarenotadverselyeffectedbyG.murisantibody,adsorbedserum,inactivatedhumanorrabbitsera,normal
rabbitserumornormaluninfectedhumanserum.
ImmobilizationofmurineGiardiabyhumanandrabbitseramaypreventthosehostsfrombecominginfectedwithmurisgroupparasitesifexcystationdoesoccur.
Cohen(5)indicatesthatallclassesofimmunoglobulinscanbedetectedinthegutand,therefore,theintraluminallocationoftheparasitesdoesnotprotectthemfrom
humoralfactorsinthisinstance,thehumoralfactorswouldimmobilizetheparasites.AggarwalandNash(1)showedthatGiardiapossessingdifferentsurfaceantigens
havedifferentpatternsofinfectionandinducedifferentimmuneresponses.ThesedifferencesarepresumblysimilartothosethatpermitthedistinctionofG.duodenalis
fromG.muriswithknownantisera.
Ifflagellatesareimmobilizedinthegut,theywouldbeunabletoattachandmultiplyandwould,instead,beevacuated.Itisalsoconceivablethattheimmobilized
GiardiamaybeacteduponbythecytotoxicactionsofintestinalepitheliummacrophagesasreportedbyOwenetal.(14)andSmithetal.(16).Itislikewisepossible
thatcertainantibacterialsubstancessecretedbythemucousmembranesmayactuponGiardia(20).
Anadditionalfeaturewhichcanbeattributedtothisinvestigationisthatbyusinguninfectedhumanserum,immobilizationstudiescanbemadeupontheGiardiaof
otherspeciesofmammalstheimplicationherebeingthatpotentiallyhumaninfectivestrainsofGiardiamaybedistinguishedfromthosespecies/strainswhich
presumablywillfailtoinfectman,andpotentialreservoirhostsmaybeidentified.
Acknowledgements
ThisworkwasfundedbycontractsfromAlbertaEnvironmentandagrantfromtheAlbertaEnvironmentalResearchTrust.
LiteratureCited
1.AggarwalA.andT.E.Nash.1987.ComparisonoftwoantigenicallydistinctGiardialambliaisolatesingerbils.Am.J.Trop.Med.Hyg.36:325332
2.Allison,A.C.1978.Macrophageactivityandnonspecificimmunity.Int.Rev.Path.18:3038
3.Belosevic,M.andG.M.Faubert.1987.Lysisand

Page172

immobilizationofGiardiamuristrophozoitesinvitrobyimmuneserumfromsusceptibleandresistantmice.ParasiteImmunol.9:1119
4.Bertram,M.A.,E.A.Meyer,D.L.Anderson,andC.T.Jones.1984.AmorphometriccomparisonoffiveaxenicGiardiaisolates.J.Parasitol.70:530535
5.Cohen,S.1976.Survivalofparasitesintheimmunocompetenthost.In:ImmunologyofParasiticInfections.Cohen,S.andK.S.Warren(eds.),pp.144145,
BlackwellSci.Pub.,Oxford,848p.
6.Einfeld,D.A.andH.H.Stibbs.1984.IdentificationandcharacterizationofamajorsurfaceantigenofGiardialamblia.Infect.Immun.46:377383
7.Filice,F.P.1952.StudiesonthecytologyandlifehistoryofaGiardiafromthelaboratoryrat.Univ.Calif.Pub.Zool.57:53146.
8.Finegold,S.M.andJ.W.Martin.1982.DiagnosticMicrobiology.6thed.,p.581,Mosby,St.Louis,Mo.,705p.
9.Grant,D.R.andP.T.K.Woo.1978.ComparativestudiesofGiardiaspp.insmallmammalsinsouthernOntarioI.Prevalenceandidentityoftheorganismswitha
taxonomicdiscussionofthegenus.Can.J.Zool.56:13481359
10.Hill,D.R.,J.J.Burge,andR.D.Pearson.1984.SusceptibilityofGiardiaduodenalistrophozoitestothelethaleffectofhumanserum.J.Immunol.,132:2046
2052.
11.Koester,S.K.andP.G.Engelkirk.1984.GlasscoversliptechniqueforstudyinginvitrointeractionsbetweenGiardiatrophozoitesandhostleukocytesby
TEM,SEM,andlightmicroscopy.J.Parasitol.70:443445.
12.Nash,T.E.andJ.Kiester.1985.Differencesinexcretoryproductsandsurfaceantigensamong19isolatesofGiardia.J.Infect.Dis.152:11661171
13.Meyer,E.A.1976.Giardiaduodenalis:isolationandaxeniccultivation.Expt.Parasitol.39:101106
14.Owen,R.L.,C.L.Allen,andD.P.Stevens.1981.PhagocytosisofGiardiamurisbymacrophagesinPeyer'sPatchepitheliuminmice.Infect.Immun.,33:591
601
15.Pearson,R.D.andP.T.Steinbigel.1980.MechanismsofthelethaleffectofhumanserumuponLeishmaniadonovani.J.Immunol.125:21952201.
16.Smith,P.D.C.C.Elson,D.B.Keister,andT.E.Nash.1982.HumanhostresponsetoGiardiaduodenalis.I.Spontaneouskillingbymononuclearleukocytesin
vitro.J.Immunol.,128:13731376.
17.Srivastava,R.K.,Agarwal,A.K.andS.R.Das.1986.Asimpletechniqueforassessmentofimmuneresistanceofhostsagainstgiardiasis.Cur.Sci.55:1141
1143.
18.Stevens,D.P.1976.Giardiasis:Immunity,ImmunopathologyandImmunodiagnosis.In:ImmunologyofParasiticInfections,Cohen,S.andK.S.Warren(eds.),
BlackwellSci.Pub.,Oxford,848p.
19.Vinayak,V.K.,Jain,P.,andS.R.Naik.1978.DemonstrationofantibodiesingiardiasisusingtheimmunodiffusiontechniquewithGiardiacystsasantigen.Ann.
Trop.Med.Parasitol.72:5812
20.Weir,K.P.1983.Immunology.ChurchillLivingstone.p.14,Edinburgh,London,MelbourneandNewYork.178p.
21.Wenman,W.M.,Meuser,R.U.andP.M.Wallis.1986.AntigenicanalysisofGiardiaduodenalisstrainsisolatedinAlberta.Can.J.Microbiol.32:926929.

Page173

ConservedSequencesoftheHSPGeneFamilyinGiardialamblia
AnitaAggarwal*,P.Romans,VidalF.delaCruzandT.E.Nash
NIAD,NationalInstitutesofHealth,Bethesda,Maryland20892,U.S.A..
TransformationofGiardialambliafromtrophicformtothecysticformmaybeinducedbyaheatshockresponse.AfactorwasfoundinnuclearextractofGiardiatrophozoitesboundto
heatshockelementofDrosophilahsp70.Thesequenceanalysisofthe5'flankingareashowedaTATAAboxatposition28to33upstreamtotheinitiationsiteandconsensus
promoterregionat56to69.Thelevelofthisfactorwassignificantlyincreasedafterheatshock.

Introduction
Alllivingorganismsexaminedtodaterespondtosignificanttemperatureincreasesbyactivatingaspecificsetofgenescalledheatshockgenes(1).Thesamegenesare
alsoactivatedbystresssignalssuchasanoxia,celltransformation,glucosedeprivationandchemicalwhichinterferewithoxidativephosphorylation(2).Heatshock
genesalsoplayaroleinnormaldevelopment.
TheDNAsequencesinvolvedintheheatshockresponsearestronglyconservednotonlyintheirproteincodingsequences,butalsointheirregulatorysequences.The
genesfortheDrosophilaheatshockproteinswereamongthefirsteukaryoticgenestobeidentifiedandcloned(3,4).Threefamiliesofheatshockgeneshavebeen
identifiedinDrosophilaspp.,genesencodingheatshockproteinsofabout83and70kDaandgenesencodingforsmallproteinsofabout2330kDa.Conserved14
basepair(bp)DNAsequences(heatshockelements[HSE])werefirstfoundupstreamfromtheTATAboxoftheDrosophila70kDaheatshockprotein(hsp70)
encodinggenes(5).AlleukaryoticspeciesfromplantstohumanshaveHSE'swithatleastfiveoftheeightconsensusnucleotidesconserved.Heatshocktranscription
factorsbindtoHSEandconfertemperaturesensitivetranscriptiontothegene(6).ThepresenceofasingleHSEissufficienttoregulatetemperatureinducible
transcription.
AGiardialambliaheatshockgenewasidentifiedanditspromoterregiondetermined.ThereasonsforidentifyingthepromoterareaofGiardiahsp70likeproteins
arethefollowing:
a.Todetermineanalogytootherhspgenes.
b.Todeterminetheroleofthepromoterregionintranscriptionofhsps.
c.Toemploythepromotersequenceforotherexperiments.
MaterialsandMethods
IdentificationofhspGeneinGiardia
GiardiatrophozoitesweregrownaxenicallyandDNAisolatedasdescribedelsewhere(7).OnegofDNAwasdigestedwithHindIII,electrophoresedina1%
agarosegel,transferredtonitrocelluloseandhybridizedwith32PlabeledPpw229(3).ThisvectorhasaDrosophilahsp70insert(3.4kb).Afterhybridizationand
washingunderlowstringencyconditions(2XSSC,0.1%SDSat50C),abroadbandofpositivesignalwasfoundataround4.2kb.Thepositiveareawascutfrom
asimilarlypreparedgelandligatedintoHindIIIsitesoftheplasmidvectorpuc18.E.coliDH5CwastransformedwiththeDNAplasmidligationmixtureandthe
transformantsscreenedwitha32Plabeled3.4kbfragmentofDrosophilahsp70(clone229).Onestronglypositiveclone(pucG.2c)waspurifiedandfurther
characterized.
AnalysisofGiardiahspGene(pucG.2c)
TherestrictionmapofpucG.2c4.2kbwasconstructedusingcombinationofdifferentrestrictionenzymes(Figure1).The5'terminalpositionoftheGiardiahspgene
wasdeterminedbyhybridizingthedoublydigestedGiardiahspgenewiththe32Plabeled5'fragmentofDhsp70(theXhoIandHincIIfragmentofPpw229).A
positivesignalwasobtainedwiththefragmentnumbers24(BglPst,PstPst,PstSal)(Figure1)indicatingthelocationofthe5'endofGiardiahsp70genewithin
thesefragments.
DNASequenceAnalysis
Threefragments,Bg1IIPst,PstPstandPstSaldoubledigestsweresequencedbystandardtechniquesusingthesinglestrandedvectorsM13mp18andM13mp19
(8).
ResultsandDiscussion
Analysisofthenucleotidesequenceof1470basepairsinallsixreadingframesrevealedonlyonecontinuousopenreadingframefrom142to1470bp.Thefirst
ATG,theinitiationsite,waslocatedat142(Figure2)andthereafter,atotalof442aminoacidswerepresent.

Figure1.
RestrictionmapofhspGiardia(pucG.2c).Thehatchedboxes
2,3,and4hybridizedwith0.9kb5'fragmentofDrosophila
hsp70indicatingthe5'terminiofGiardiahspgene.Doubledigested
BglIIPst(Box2),PstPst(Box3)andPstSal(Box4)werecloned
intovectorM13mp18andM13mp19whichhadbeen
digestedwithappropriaterestrictionenzymes.
*Correspondingauthor.

Page174

Figure2.
TheaminoacidsequenceoftheGiardiahspfragment
in5'3'orientation.Thefirstmethionine*ispresentat
position142andatotalofcontinuous442aminoacidsareshown.

SequencehomologiesweredeterminedusingGenebanksearchDec10computerprograms.NosignificanthomologywaspresentbetweenthehspgeneofGiardia
andDrosophilahsps.Sequenceanalysisof5'flankingareashowedaTATAAboxatposition28to33upstreamtotheinitiationsite(Figure3).
ConservedSequenceElementofthehsp
Welookedfortheputativeregulatorysignalsinthe5'flankingregionoftheGiardiahspgenebycomparisonofthe

Figure3.
Nucleotidesequenceanalysisof5'flankingregionofGiardiahsp.
Regionscorrespondingtotheconsensusheatshockpromoter(B)andTATAAbox(A)arenoted.

Figure4.
ComparisonoftheGiardiaHSEconsensussequences
withsequencesderivedfromhuman,chicken,mouse,T.brucei
andDrosophilahsp70genes.Theunderlinednucleotidesconstitute
thebasesofpotentialhomology.Altogether,8ofthebases
areconserved.SixofeightfallundertheconsensusinGiardia.

sequenceofthisregionwithHSEconsensussequences(9,10).AnHSErelated14nucleotidesequence(CTTGAGGGCTCCGG)wasfoundatpositions56to69
(Figure3).Itsrelatednesstotheconsensuspromoterregion(CNNGAANNTTCNNG)was6/8(Figure4).SincedivergenceintheHSEsequencesexistsamongst
differentDrosophilahspsandamongotherspecies,itismostlikelythattheconsensusregionintheGiardiahsphasafunctionintranscriptionalregulation.
RoleofthePromoterRegioninTranscriptionalRegulation
Todeterminethetemperaturedependentregulatoryroleofthisconsensuspromoterregion,Giardiawereheatshockedfrom5minto5hrsandtheamountof
transcribedRNAtoGiardiahspdetermined.RNAwasisolatedandspottedontonitrocelluloseandhybridizedwitheither32Plabeled4.2kbGiardiahspgeneor
anactingenefromDictyostelium.NorthernblotanalysisisshowninFigure5.At42Ccomparedto37C,RNAwas10foldmoreabundantafter5hrsofheat
shockwhereastheamountofGiardiaRNAtoactinremainedunaffected.Therefore,enhancedtranscriptionofGiardiahspoccurredafterheatshockshowing
temperaturedependentregulationofthegene.

Figure5.
AnalysisofheatshockedRNAs.Giardiaweregrownaxenicallyand
heatshockedfor0min,5min,10min,30min,1hr,3hrsand
5hrsat42C,andtheRNAisolatedbyhotphenolnethod(11).
100ngwasspottedontonitrocelluloseinduplicates,hybridizedwith
32Plabeled4.2kbGiardiahspgene,puc18,theactingenefrom
Dictyostelium,andribosomalDNAofPlasmodiumfalciparum.

Page175

LiteratureCited
1.Lindquist,S.1986.Theheatshockresponse.Ann.Rev.Biochem.55:11511191.
2.Atkinson,B.G.andD.B.Walden.1985.Changesineukaryoticgeneexpressioninresponsetoenvironmentalstress.D.B.Walden(ed.)AcademicPressInc.,
N.Y.
3.Livak,K.J.,Freund,R.,Schweber,M.,Wensink,P.C.,andM.Meselson.1978.SequenceorganizationandtransciptionattwoheatshocklociinDrosophila.
Proc.Nat.Acad.Sci.U.S.A..75:5613.
4.Schedl,P.,ArtavanisTsakonas,S.,Steward,R.,Gehring,W.J.,andM.E.Mirault.1978.TwohybridplasmidswithD.melanogasterDNAsequence
complementarytomRNAcodingforthemajorheatshockprotein.Cell14:921929.
5.Pelham,H.R.B.1982.AregulatoryupstreampromoterelementinDrosophilaHsp70heatshockgene.Cell30:517528.
6.Topol,J.,Ruden,D.M.andC.S.Parker.1985.SequencesrequiredforinvitrotranscriptionalactivationofaDrosophilahsp70gene.Cell42:527537.
7.Nash,T.E.,McCutchan,T.,Keister,D.,Dame,J.B.,Conrad,J.andF.D.Gillin.1985.EndocucleaserestrictionanalysisofDNAfrom15Giardialamblia
isolatesobtainedfrommanandanimals.J.Inf.Dis.152:6473.
8.Sanger,F.,Coulsen,A.R.,Barrell,B.G.,Smith,A.J.J.andB.Roe.1980.CloninginsinglestrandedbacteriophageasanaidtorapidDNAsequencing.J.Mol.
Biol.143:161178.
9.Corces,V.,Pellicer,A.,Axel,R.,andM.Meselson.1981.Integration,transciptionandcontrolofaDrosophilaheatshockgeneinmouse.Proc.Natl.Acad.Sci.
USA78:70387042.
10.Holmgren,R.,Corces,V.,Morimoto,R.,Blackman,R.andM.Meselson.1981.Sequencehomologiesinthe5'regionsoffourDrosophilaheatshockgenes.
Proc.Nat.Acad.Sci.USA78:37753778.
11.Maniatis,T.,Fritsch,E.F.andJ.Sambrook.1982.Molecularcloning:alaboratorymanual.ColdSpringHarborLaboratory,NewYork.

Page177

TheResponseofHumanstoAntigensofGiardialamblia
M.G.OrtegaPierres*,R.Lascurain,R.ArguelloGarcia,R.CoralVazquez,G.AcostaandJ.I.Santos
CentrodeInvestigationydeEstudiosAvanzadosDelIPN,ApartadoPostal14740,Mexico14,D.F.
SurfacelabelledantigensfromGiardialambliaaswellasimmunoblottingtechniqueswereusedtoanalysethehumoralresponseofasymptomaticandsymptomaticMexicanpediatric
patientstocomponentsofG.lambliatrophozoites.Theresultsdemonstratedthatfewradiolabelledsurfacecomponentsareprecipitatedbyallsera.Westernblotanalysisrevealed
thatseveralparasitecomponentsareconstantlyrecognizedbyserafrominfectedpatients.Howeverthelevelofrecognitionvarieswithindividualsera.Theisolationanduseof
commonlyrecognizedantigensmightpermitarationalapproachtothedevelopmentofimprovedimmunodiagnosticmethodsforgiardiasis.

Introduction
HumangiardiasisisaninfectioncausedbytheintestinalprotozoanGiardialamblia.Thisdiseasehasaworldwidedistribution(25)andisasignificanthealthproblem
indevelopingcountries(24).Infectionwiththisparasitecanbetransmittedfrompersontoperson(9,18)orbyingestionoffoodorwatercontaminatedwithcysts(3).
TheclinicalspectrumofinfectionwithG.lambliarangesfromasymptomaticpassageofcyststopersistentandseverediarrhoeawithmalabsorption(19).
SeveralstudieshaveindicatedthatG.lambliaelicitsanimmuneresponseinitshost.ThiswassuggestedfirstbythefactthatindividualsexposedrepeatedlytoG.
lambliadevelopedresistancetoinfection(11).AroleforhumoralresponsesincontrollingthediseasehasbeensuggestedsinceIgdeficientpatientsaremore
susceptibletoinfectionwithG.lamblia(1).Subsequently,variousreportshaveindicatedthepresenceofcirculatingantibodiesinpatientswithgiardiasis
(15,16,20,23).
Ofparticularimportanceinthestudyofimmuneresponsestothisparasiteistheidentificationofparasitecomponentswhichactivateimmunemechanismsduring
infection.Recently,twogroupshavereportedtheirfindingsonthereactivityofhumanserafrompatientswithgiardiasistoG.lambliacomponents.Onestudyshowed
thataG.lambliasurfaceantigenof88kDawasprecipitatedbytwohumansera(5)whileanotherreportedthataproteinwithamolecularweightof31kDawas
mainlyrecognizedinhumaninfections(21).
Here,weanalysedthereactivityof29serafromasymptomaticandsymptomaticMexicanpediatricpatientstoG.lambliaantigens.Weusedradiolabelledparasite
surfaceantigensandWesternblotstoanalyseawiderangeofhumansera.Thesestrategiespermitfinedissectionofhumoralresponsesfollowinginfectionby
providingquantitativeandqualitativeimmunoprecipitationdata.OurresultsshowedthatfewsurfaceG.lambliacomponentswereimmunoprecipitatedbyallsera
tested.Thepatternsofreactivity,asdeterminedbyWesternblotanalysisvariedforeachindividualserum.However,thereweresomeantigensfromthisparasite
whichwereconstantlyrecognizedbyallseratested.
MaterialsandMethods
ParasiteCultures
TrophozoitesofG.lambliaPortland1(P1)strainweregrownaxenicallyat37CinDiamond'smodifiedTYIS33medium(8)supplementedwith10%heat
inactivatedcalfserumand250g/mLeachofstreptomcinandampicillin.Organismswereharvestedinlatelogphasebychillingcultureflasksonicefor30min,
invertingtheflasksseveraltimesandcentrifugingtheircontentsat250gfor10min.Theorganismswerewashedthreetimesinphosphatebufferedsaline(PBS)and
thecellpelletfromthefinalwashwassuspendedinasmallvolumeofPBS.Theconcentrationoftrophozoiteswasdeterminedwithahemacytometer.
AntigenPreparation
Trophozoitesculturedandharvestedasdescribedabovewereusedtoobtainparasiteantigens.Theparasitepelletwasresuspendedin10mMtrisHClpH8.3
containing1mMphenylmethylsulfonylfluoride,25mMNethylmaleimideand0.5%TritonX100.Thissuspensionwassonicatedwithsix15secburstsinanice
bathandcentrifugedat15600gfor30mintoremovedebris.ProteinconcentrationofthesupernatantswasdeterminedbyamodifiedLowryassay(4).
SurfaceLabellingofParasites
G.lambliatrophozoitesgrownandharvestedasdescribedaboveweresurfacelabelledwith125Ibythelactoperoxidasemethod(12).Afterlabelling,trophozoites
weresonicatedunderthesameconditionsasforantigenpreparationandtheradiolabelledantigenswereusedforimmunoprecipitationassayswithhumansera.
PatientsSera
Serumsampleswereobtainedfromchildren(2to14yearsold)attheHospitalInfantildeMexico.Theseserawerefrom5patientswithasymptomaticandfrom24
patientswithvarioussymptomsofgiardiasiswhichincludedmainlydiarrhoeaandabdominalpain.Analysisoffecesfromallpatientsatthetimeofbloodcollection
demonstratedonlythepresenceofG.lambliacysts.Controlserawereobtainedfrom3children(5to6yearsold)withnodetectablecystsintheirfecesandno
historyofgiardiasisorsymptomsofgastrointestinaldisease.
Immunoprecipitation
RadiolabelledG.lambliaantigens(100,000cpm)wereprecipitatedwith58Lofhumanseraovernightat4C.AnexcessofproteinASepharosethenwasadded
andthetubeswereplacedat4Cfor2h.TheproteinASepharosebeadswerewashedthreetimeswithNET(150mMNaCl,5mMEDTA,50mMTrisHCl)
containing0.5%TritonX100andthenwereboiledinsamplesolutionfor5minbeforeanalysisbySDSPAGE.
*Correspondingauthor.

Page178

Electrophoresis
AnalysesofprecipitatesandseparationofG.lambliaantigensforimmunoblottingassayswereperformedbypolyacrylamidegelelectrophoresisinthepresenceof
sodiumdodecylsulfate(SDSPAGE).ThiswascarriedoutbythemethodofLaemmli(10)using515%gradientslabsgels.Afterelectrophoresis125Icontaininggels
weredriedandexposedat70CtoKodakXOmatARfilms.
Immunoblotting
AntigensseparatedbySDSPAGE(100gproteinpercmofgelwidth)weretransferredtonitrocellulosepaperandanalysedbythemethodofTowbinetal.(22)as
modifiedbyHanffetal.(7)andPekkalaandRuoslathi(14).Afterthetransfer,nitrocellulosesheetsweretransferredtodilutingbuffer[PBSpH7.2,containing4%
(wt/vol)BSAand0.5%(vol/vol)TritonX100]containing1:50dilutionofhumanantiG.lambliaseraandincubatedwithslowshakingfor1hfollowedbythree15
minwashesinPBS(pH7.2)containing1%(vol/vol)TritonX100(PBST).ThesheetswereplacedinPBScontaining10%(vol/vol)heatinactivatedbovinefetal
calfserumandhorseradishperoxidaseconjugatedgoatantihumanwholeIgsfollowedbythree15minwashesinPBST.Nitrocellulosesheetswerethendeveloped
bytheadditionofafreshlypreparedsolutionof0.05%(wt/vol)4chloro1naptholand0.01%(v/v)hydrogenperoxideinPBS.
Results
TheabilityofserafromasymptomaticandsymptomaticpatientstoprecipitatetheradiolabelledantigensofG.lambliawasdeterminedbyimmunoprecipitationand
qualitativeanalysisbySDSPAGE.Theanalysisoftotallabelledparasitecomponentsrevealedthepresenceofapproximatelytwelvebands(Figure1).Ofthesethere
wasasinglemajorpolypeptideof82kDaandlessprominentbandsof63,55,53,49,43,40,35,32,27,and24kDa.Othersurfacelabelledcomponentsbetween
190kDaand144kDaweresometimesdetected.Figure1showstheprecipitationpatternsobtainedwhen5serafromeitherasymptomaticorsymptomaticpatients
wereused.Intheseprecipitations,radiolabelledantigenswithmolecularweightsof85,63kDaand55kDawereobserved.ThesameradiolabelledG.lamblia
componentswereprecipitatedbyalltheseratestedinthisassay.

Figure1.
Autoradiographof125IlabelledproteinsofG.lambliaprecipitated
byantiG.lambliahumansera.Solublesurfaceiodinatedcomponents
wereimmunoprecipitatedasdescribedabovewithcontrolhuman
sera(trackN),serafromasymptomaticpatients(tracksA1A 5)orsera
fromsymptomaticpatients(tracksS1S5).TrackP1Tisthetotal
surfacelabelledproteinsofG.lambliaP1strain.Molecularweight:(103).

Figure2.
Westernblotanalysisofantigenantibodyreactionsdetected
whenG.lambliasolubleextractswereoverlaidwithcontrol
humansera(tracksmarkedN),serafromasymptomaticpatients
(tracks14)orserafromsymptomaticpatients(tracks528)and
developedwithperoxidaseconjugatedto
antihumanIg.Molecularweight:103.

Althoughhighmolecularweightcomponentswerefoundtobeiodinated,noprecipitationofthesecomponentswasobservedwiththeseratested.
SerafromasymptomaticandsymptomaticpatientsweretestedforthereactivityagainsttotalsolubleG.lambliaantigensbyimmunoblotting.Controlserafromnormal
individualswithnohistoryofgiardiasiswerealsotestedwiththesameparasiteextracts.Thespectrumofantigen/antibodyreactionsdetectedbyWesternblottingcan
beseeninFigure2.AlltheseratestedreactedtoG.lambliasolublecomponentswithmolecularweightsbetween35kDaand200kDa.Thisanalysisrevealed
variationsintherecognitionofG.lambliacomponentsbothquantitativelyandqualitatively.Amongtheserafromasymptomaticpatients(tracks14)therewere
antigensofapproximately121,90,46,44and35kDawhichwerepreferentiallyrecognizedbyallsera.Oneoftheseserareactedmorestronglytoa31kDa
component(track1).Similarproteinstothoserecognizedbyasymptomaticserawerealsodetectedwithserafromsymptomaticpatients(tracks528).However
differencesinreactivitywerealsoobservedamongthissera.Someofthemreactedmorestronglytoantigensof23kDa(track7),62kDaand57kDa(track20),53
kDa(track22),44kDa(track23),152kDaand57kDa(track27).Thehumanserausedascontrolinthisassay(tracksmarkedN)gavealmostnoreactivitywith
theG.lambliaextract.
Discussion
Severalstudiesofhumanresponsivenesstoantigensfromparasitesingeneralareaimedtowardsthedevelopmentofmoresensitiveandspecificimmunodiagnostic
tests.Suchstudiesinvolvethedetectionofspecificantibodiescirculatinginthebloodaswellasinotherhostsamplematerialsuchasfaeces.Theuseofpreparations
withmorerestrictedantigeniccompositionpermitsafinerdissectionofhostparasiteinteractions.Thiscanbeachievedbytheisolationofradiolabelledproteinsfrom
specificparasitecompartments.Wehaveusediodinated

Page179

surfacecomponentsfromG.lambliatoanalysethehumanantibodyresponsetosuchparasitemolecules.
InoursurfacelabellingexperimentswefoundthatG.lambliacomponentsof82,63,55,53,49,43,40,35,32,27and24kDawereaccessibletoradioiodination.
Inadditionweobservedinsomeoccasionslessdefinedbandsbetween190kDaand144kDa.Someofthebandsobservedinthisstudyhavealsobeendetectedin
studiesreportedbyEinfeldandStibbs(6)andClarkandHolberton(2).Nashetal.(13)havefoundamajorpolydispersebandofmaterialfrom94kDato225kDa
presentasasmearandsometimesasadiscreteladderofbands.Thesevariableiodinationpatternscouldbeduetominortechnicalvariations(e.g.labellingconditions,
cultureconditions)orparasitevariation(e.g.stablesubpopulationsorspontaneousoccurringphenotypicvariants).Thesepossibilitiesmightexplainthefactthathigh
molecularweightcomponentswerenotalwaysdetectedinouriodinationexperiments.Furtherstudiesregardingtheuseofclonedorganismsaswellasamoredefined
mediumandbettercontrolofthegrowthandmultiplicationoftheparasitemightallowforabetterunderstandingofsurfacemoleculesofG.lamblia.
Inthestudyreportedhereanumberofthesurfacelabelledproteinswereinvariablyantigenicinbothasymptomaticorsymptomaticpatients.Thusinallcaseswefound
precipitationof82,63and55kDacomponents.Therewashowever,amoremarkedresponsetothe82kDaproteinregardlessofclinicalfeaturesandtimecourseof
infection.Recently,Edsonetal.(5)havereportedan88kDaproteinfromG.lambliawhichwasprecipitatedbytwohumanserafrompatientswithgiardiasis.This
proteinmaybesimilartothe82kDacomponentrecognizedbyallMexicansera.Hence,theseradiolabelledproteinsmightbepotentiallyusefulfordiagnosisof
giardiasis.AlthoughwedetectradiolabelledG.lambliasurfacecomponentsofhighmolecularweight(190to116kDa)inouriodinationexperimentsthesewerenot
precipitatedbyanyofthehumanseratested.
Recently,TaylorandWenman(21)reportedamajor31kDaG.lambliaantigenrecognizedduringhumaninfections.Inthestudyreportedherewedidnotdetecta
majorreactivityoftheseratestedtowardsthiscomponent.TheanalysisofthereactivityofthehumanserabyimmunoblottingshowedthatseveralotherproteinsofG.
lambliawithmolecularweightsof121,9046,44and35kDawerepreferentiallyrecognizedbymostoftheseratested.Therewere,however,differencesbetween
theindividualseraused.Thismightbeduetothefactthathumansareanoutbredpopulationandthismightsuggestamorewidelyvaryingresponsetothecomplex
arrayofantigenicdeterminantspresentedduringthecourseofparasiticinfection.Asyetwehavefailedtonoteconsistentdifferencesinthepatternofreactivityof
asymptomaticandsymptomaticsera.Ontheotherhand,theobservedpatternofrecognitionofG.lambliaantigensmightalsobeinfluencedbythebalanceof
differentIgisotypesaswellasbythetitreofantibodiespresentatthetimeofsamplecollection.Inspiteofthesepotentialsforvariability,thepresentstudyhasallowed
theidentificationofG.lambliacomponentswhichmightpotentiallybeusefulfordiagnosisofthedisease.Aninterestingaspectwillbetocarryoutlongitudinalstudies
inhumanswithgiardiasiswhichwillgiveadditionalinformationregardingtherecognitionofG.lambliaantigensduringthecourseofinfection.
Analternativeapproachtothediagnosisofgiardiasisandparasiticinfectionsingeneralisnottodetectspecificcirculationantibodiesbutspecificparasiteantigensin
suitablehostsamplessuchasfeces.ArecentstudybyRosoffandStibbs(17)usedthisapproachisolatinga65kDaG.lambliaantigeninstoolsofparasitepositive
patients.
Thus,theidentificationandisolationofG.lambliaspecificcomponentswhichareimmunogenicinhumansorreleasedbytheparasiteduringinfectionshouldhopefully
providethebasisforarationalimprovementofdiagnostictestforgiardiasis.
Acknowledgements
WewishtoacknowledgeDr.D.PeattieandDr.R.M.E.ParkhouseforcriticallyreadingthismanuscriptandMrs.MariadeLourdesVazquezforsecretarial
assistance.
R.LascurainandR.coralVazquezarerecipientsofaCONACYTstudentship.ThisworkwassupportedinpartbyCONACYT(Mexico)andtheMACARTHUR
FOUNDATION.
LiteratureCited
1.Ament,M.E.andC.E.Rubin.1972.Relationofgiardiasistoabnormalintestinalstructuresandfunctioningastointestinalimmunodeficiencysyndromes.
Gastroenterology.62:216266.
2.Clark,J.T.andD.V.Holberton.1986.PlasmamembraneisolatedfromGiardialamblia:Identificationofmembraneproteins.J.CellBiol.42:200206.
3.Craun,G.F.1979.Waterborneoutbreaksofgiardiasis.In:WaterborneTransmissionofGiardiasis.JakubowskiandHoff(eds).U.S.EnvironmentalProtection
Agency.pp.127147.
4.Dulley,J.R.andP.A.Greive.1975.AsimpletechniqueforeliminatinginterferencebydetergentintheLowrymethodofproteindetermination.Anal.Biochem.
64:136141.
5.Edson,C.M.,Farthing,M.J.G.,ThorleyLawson,D.A.,andG.T.Keusch.1986.An88,000MrGiardialambliasurfaceproteinwhichisimmunogenicinhumans.
Infect.Immun.34:621625.
6.Einfield,D.A.,andE.A.Stibbs.1984.IdentificationandcharacterizationofamajorsurfaceantigenofGiardialamblia.Infect.Immun.46:377383.
7.Hanff,P.A.,Feihniger,T.E.,Miller,J.N.,andM.A.Lovett.1982.HumoralimmuneresponseinhumanssyphilistopolypeptidesofTreponemapallidum.J.
Immunol.129:12871291.
8.Keister,D.B.1983.AxeniccultureofGiardialambliainTYIS33mediumsupplementedwithbile.Trans.R.Soc.Trop.Med.Hyg.77:487488.
9.Keystone,J.S.,Krajden,S.,andM.R.Warren.1978.PersontopersontransmissionofGiardialambliaindaycarenurseries.Can.Med.Assoc.J.119:241
258.
10.Laemmli,U.K.1970.CleavageofstructuralproteinsduringtheassemblyoftheheadofbacteriophageT4.Nature227:680685.

Page180

11.Moore,G.T.,Cross,W.M.,McGuire,D.,Mollohan,C.S.,Gleason,N.N.,Healy,G.R.,andL.H.Newton.1969.Epidemicgiardiasisataskiresort.N.Engl.J.
Med.281:402407.
12.Morrison,M.1980.Lactoperoxidasecatalyzedoxidationasatoolforinvestigationofproteins.Meths.Enzymol.70:214220.
13.Nash,T.E.,Guillin,F.D.andP.D.Smith.1983.ExcretorySecretoryproductsofGiardialamblia.J.Immunol.131:20042010.
14.PekkataFlagan,A.andE.Ruoslathi.1982.Unfoldedtransferringpolypeptidechainisimmunologicallycrossreactivewithsimilarderivatesofserumalbuminand
alphafetoprotein.J.Immunol.128:11631167.
15.Radulescu,S.,Iancu,L.,Simionescu,D.andE.A.Meyer.1976.Serumantibodiesingiardiasis.J.Clin.Pathol.29:863.
16.Ridley,J.J.andD.S.Ridley.1976.Serumantibodiesandjejunalhistologyingiardiasisassociatedwithmalabsorption.J.Clin.Pathol.29:3034.
17.Rosoff,J.D.andH.H.Stibbs.1986.IsolationandidentificationofGiardialambliaspecificstoolantigen(GS65)usefulincoprodiagnosisofgiardiasis.J.Clin.
Microbiol.23:905910.
18.Schmerin,M.J.,Jones,T.C.,andH.Klein.1978.Giardiasis:Associationwithhomosexuality.Ann.Intern.Med.88:801803.
19.Smith,J.W.,andM.S.Wolfe.1980.Giardiasis.Ann.Rev.Med.31:373383.
20.Smith,P.D.,Gillin,F.D.,Brown,W.R.andT.E.Nash.1981.IgGantibodytoGiardialambliadetectedbyenzymelinkedimmunosorbentassay.
Gastroenterology.80:14761480.
21.Taylor,G.D.andW.M.Wenman.1987.HumanimmuneresponsetoGiardialambliainfection.J.Infect.Dis.155:137140.
22.Towbin,H.T.,Staehelin,T.,andJ.Gordon.1979.Electrophoretictransferofproteinsfrompolyacrylamidegelstonitrocellulosesheets:Procedureandsome
applications.Proc.Nat.Acad.Sci.U.S.A.76:43504354.
23.Visvesvara,G.S.,Smith,P.D.,Healy,G.R.andW.R.Brown.1980.AnimmunofluorescencetesttodetectserumantibodiestoGiardialamblia.Ann.Intern,
Med.93:802805.
24.Walsh,D.J.andK.S.Warren.1979.Selectiveprimaryhealthcare:Aninterimstrategyfordiseasecontrolindevelopingcountries.N.Engl.J.Med.301:967
974.
25.Wolfe,M.S.1978.Currentconceptsinparasitology.Giardiasis.N.Engl.J.Med.298:319321.

Page181

PropertiesofGiardiaLambliaRNAs
CeciliaMontanez*,LourdesCervantes,CesarOvandoandGuadalupeOrtegaPierres
DepartmentofGeneticsandMolecularBiology,CentrodeInvestigacionydeEstudiosAvanzadosdelIPN,ApartadoPostal14740,MexicoCity07000,Mexico.
RNAfromtheparasiticprotozoanGiardialambliawasobtainedbythreedifferentmethodsandanalysedbyagarosegelelectrophoresisunderdenaturingconditions.Inallcasestwo
prominentpopulationswereobserved.Ourresultsshowedthatthelarge(LSrRNA)andsmall(SSrRNA)subunitrRNAsareapproximately2390and1420nucleotideslong.
AnalysisofsucrosegradientprofilesofG.lambliarRNAsrevealedsedimentationvaluesof21Sand15SfortheLSrRNAandSSrRNAspeciesrespectively.SmallRNAswere
characterizedonpolyacrylamidedenaturinggelsinwhichtwobandsofunusuallysmallsize,130and118nucleotideslong,werefound.OthersmallerRNAswerealsoobserved.
TotalRNAwastranslatedinvitrousingarabbitreticulocytelysate.Inthiscaseabroadspectrumoftranslationproductswasobtained.Whentheseproductswere
immunoprecipitatedusingimmuneserafromhumansinfectedwithG.lamblia,wefoundthatmostoftheinvitrotranslatedcomponentswereimmunogenic.Nodifferencesinthe
immunoprecipitationpatternswereobservedwhenserafromasymptomaticorsymptomaticpatientswereused.

Introduction
Giardialambliaisaparasiticprotozoawhichinfectstheintestinaltractofhumans.Thisflagellatedandbinucleatedprotozoanisdistributedworldwideandcausesthe
diarrhealdiseaseknownasgiardiasis(13).Despitethehighprevalenceofthisinfection,thegeneticsandmolecularbiologyofG.lambliaremainpoorlyunderstood.
Specifically,thestructuralorganizationandexpressionofthenucleicacidsfromthisparasitehavenotbeenwellcharacterized.
Variousstudiesconcerningtheisolationandcharacterizationofribonucleicacid(RNA),especiallyribosomalRNA(rRNA)inprotozoahavebeenreported
(1,6,8,15,16).Thesestudieshaverevealedconsiderablediversitybetweenhomologousnucleicacidsspeciesoftheseorganisms.Theyhavealsocontributedtoa
betterunderstandingoftheorganizationandexpressionoftheirnucleicacidsaswellasinitiatingtheestablishmentofanevolutionaryrelationshipbetweenthem.Inthis
context,westartedstudiesonthecharacterizationofG.lambliarRNAsaswellastheanalysisofmessengerRNAs(mRNAs)whichcodeforantigensthatactivate
immuneresponsesinthehost.WehavedeterminedthattherRNAsofthisparasiteareapproximately2390and1420nucleotideslong,whichareparticularlysmallfor
eukaryoticcells.TheisolatedtotalRNAtranslatedinvitroencodepolypeptidescontainingantigenicdeterminantsthatarerecognizedbyserafrompatientswith
giardiasis.
MaterialsandMethods
GrowthofParasites
G.lambliaPortland1(P1)andWBtrophozoites(obtainedfromE.WeinbachandT.Nash,NationalInstitutesofHealth,respectively)wereculturedaxenicallyin
vitrounderanaerobicconditions.Thetrophozoitesweregrownat37CinDiamond'smodifiedTYIS33medium(18)supplementedwith10%heatinactivatedcalf
serumandantibiotics(penicillinat250/mLandstreptomycinat250g/mL).Organismswereharvestedinlatelogphasebycentrifugationat250gfor10min,and
thecellswerewashedtwicewithphosphatebufferedsalinepH7.2(PBS).ThecellpelletfromthefinalwashwassuspendedinasmallvolumeoficecoldPBSand
theconcentrationoftrophozoitesdeterminedbycountinginahematocytometer.CellsusedforRNAextractionwerestoredat70Cuntiluse.
ExtractionofTotalRNA
TotalRNAfromG.lambliatrophozoiteswaspurifiedusingthefollowingthreemethods:1)G.lambliatrophozoitesweresuspendedinasolutionof4Mguanidine
isothiocyanatein50mMTrisHClpH7.6,10mMEDTA(Ethylendiaminetetraaceticacid)2%SDS(sodiumdodecylsulphate)and0.14MBmercaptoethanoland
centrifugedat8000gfor10minutes(7).Thesupernatantwaslayeredover1.2mLof5.9Mcesiumchloride(CsCl)andcentrifuged17hrsat35,000rpmat16C.
Thepelletwasresuspended,extractedwithchloroformbutanol(4:1v/v)andtreatedwithLiClandethanoltoobtainpureRNA(28).2)G.lambliatrophozoites
werewashedincoldPBSandlysedin0.1MTrisHCl,pH9.0,0.1MLiCl,1mMEDTA,1%SDS.Sampleswereextractedwithphenolchloroformisoamyl
alcohol(25:24:1v/v)andRNAwasprecipitatedwithethanoland5MLiCl(3).3)G.lambliatrophozoitesweredissolvedin10volumesof8Murea,0.15M
sodiumphosphatepH6.8,0.01MEDTA,1%SDSandextractedwithphenolchloroformisoamylalcohol(25:24:1v/v)(11).Theaqueouslayerwastreatedwith
ether,andloadedonahydroxylapatitecolumn(23).RNAwascollectedbyelutionwith0.19MsodiumphosphatepH6.8,dialyzedandprecipitatedwithethanolin
thepresenceof0.3MsodiumacetatepH5.0
AgaroseGelElectrophoresis
TheG.lambliatotalRNAobtainedbythethreemethodsdescribedabovewasanalyzedondenaturingslabgels:2to4gofRNAwereheatedfor5minin25%
deionizedformamide,20%glycerol,0.25%bromophenolblue(BPB)and0.04%xylenecyanol(XCFF)(w/v)at65Cand
*Correspondingauthor.

Page182

Figure1.
AnalysisofG.lamblia(P1strain)RNAbydenaturingagarose
gelelectrophoresis.TotalG.lambliaRNAobtainedbyphenol
extractionswererunonadenaturing1.5%agarosegelinthepresence
offormaldehydeandformamideandstainedwithethidiumbromide.
Track1:ratbraintotalRNAtrack2:G.lambliaRNAtrack3:E.colitotalRNA.

thenchilledonicebeforeloading.Thesampleswereappliedtoa1.5%agarosegelcontaining6%formaldehydein2mMNaH2PO4,18mMNa2PO4.7H2Oandthegel
wasrunat50Vfor4.55hrs.TheRNAwasvisualizedusingashortwaveultravioletlightafterthegelswerestainedwith0.5g/mLethidiumbromidefor30minat
roomtemperature.
PolyacrylamideGelElectrophoresis
Verticalpolyacrylamidegelscontaining8.8%acrylamide,2%bisacrylamideand0.2Mureawerepreparedin1XTBEbuffer.Thegelswereprerunat200Vfor30
min.RNA(58g)wasdissolvedin80%formamide,0.25%BPB,0.04%XCFFandheatedat85Cfor5minpriortoloading.Electrophoresiswasperformedat
250Vfor1.5hr,andthegelswerestainedwithethidiumbromideasdescribedabove.
DeterminationofRNASizes
ApproximatesizesofG.lambliaRNAweredeterminedondenaturinggelsbycomparisonwithRNAmarkersofknownsize.TheRNAsusedasstandardsinthese
determinationswereratbrain28S(4802bases)(14),18S(1869bases)(30),E.coli23S(2904bases)(4),and16S(1541bases)(5)andSaccharomyces
cerevisiae5.8S(158bases)and5S(121bases)(12).
SucroseGradientCentrifugation
RNAsamples(80g)suspendedin300Lof100mMsodiumchloride,10mMsodiumacetatepH5.2,and1mMEDTAwerelayeredon1035%linearsucrose
gradientscontaining10mMsodiumacetatepH5.2.Thesampleswerecentrifugedat23,000rpmfor18hrsat4C.Followingcentrifugationthegradientswere
collectedandtheiropticaldensitiesat260nmdetermined.Svedberg(S)valueswereestimatedusingratbrainRNAsasstandardmarkers(25).
InvitroTranslations
AllinvitrotranslationswereperformedusingrabbitreticulocytelysatesfromAmershamRadiochemicals,Ltdaccordingtoinstructionsprovidedbythevendorinthe
presenceof1.2mMMg(OAc)2.TotalRNA(10g),wasaddedto20mLofrabbitreticulocytelysatecontaining30Ci(35S)methionine(1190Ci/mmol.Amersham
Radiochemicals).Thesampleswereincubatedat30Cfor90min.Sampleswereusedtodeterminetheamountof(35S)methionineincorporatedintotrichloroacetic
acidprecipitablematerial.Theremainderofthemixturewasanalyzedon10%SDSpolyacrylamidegelsunderreducingconditions(SDSPAGE)(19).Protein
detectionwasachievedbyCoomasiebluestainingandautoradiography.
Sera
HumanserawereobtainedfromasymptomaticandsymptomaticMexicanpatientswithG.lambliainfectionsconfirmedbystoolexamination.Controlhumansera
wereobtainedfromadultswithnohistoryofgiardiasis.Titrationofallserasampleswasperformedbyenzymelinkedimmunosorbentassay(11).
ImmunoprecipitationofTranslationProducts
Humanserasamplescollectedfrompatientswithgiardiasiswereusedtoprecipitateinvitrotranslationproducts.Forthis,50LoftranslationmixtureinNETT(150
mMNaCl,5mMEDTA,50mMTrisHCLpH8,0.5%TritonX100)containing200,000cpmwereincubatedovernightat4Cwithsuitabledilutionsofsera
obtainedfromhumansasdescribedabove.AfterincubationimmunecomplexeswereabsorbedonproteinAsepharoseandunboundpolypeptideswereremovedby
washinginNETTbuffersupplementedwith1%bovineserumalbumin(9).Radiolabelledantigenswereeluted,andthesampleswereanalyzedbySDSPAGE(19)on
1.5mmslabgels.ProteindetectionwasachievedbyCoomasiebluestainingandautoradiography.
Results
ExtractionandCharacterizationofTotalG.lambliaRNA
TotalRNAfromG.lambliatrophozoitesofP1andWBstrainswasisolatedbythreedifferentmethodsinordertodetectallRNApopulationspresentinthese
organisms.Theseincludeguanidineisothiocyanateextraction(7),phenolchloroformextraction(3),andhydroxylapatitechromatography(seematerialsandmethods).
SizeDeterminationofrRNAs
RNAsamplesobtainedasdescribedabovewereanalyzedbyelectrophoresisunderstronglydenaturingconditions.ElectrophoreticanalysisoftheRNAobtained
fromtheP1strainbythethreedifferentmethodsrevealedthepresenceoftwoprominentrRNAbandswhichcorrespondtothelarge(LSrRNA)andsmall(SSrRNA)
rRNAspeciesofG.lamblia(Figure1).ThemolecularweightsoftheserRNAspecieswereapproximatedusingratbrainandEscherichiacolirRNAsasstandards.
ThesizesofthelargeandsmallrRNAscorrespondto2390and1420nucleotides,respectively.SimilarsizesweredeterminedforLSrRNAandSSrRNAfromWB
strain(datanotshown).InterestinglytheseRNAspeciesfrombothstrainsofG.lambliaaresmallerthanmostoftheprokaryoticandeukaryoticrRNAsdescribed
(24).
SedimentationVelocityMeasurements
RNApurifiedbyphenolextractionorbyhydroxylapatitechromatographywascentrifugedthroughsucrosegradientsundernondenaturingconditions.Theprofilesof
theO.D.measurementsfromthecollectedfractionsrevealedthepresenceoftwopeakswhichcorrespondtothetworRNAspeciesofG.lamblia(P1strain)(Figure
2).SedimentationvaluesoflargeandsmallrRNAsascalculatedusingratbrainrRNAsasstandardswere21Sand15S,respectively.Thesedimentationvaluesof
thesetwoG.lambliarRNAsare

Page183

Figure2.
SucrosegradientsedimentationanalysisofRNAfrom
G.lamblia(P1strain).TheG.lambliaRNAsamplesisolated
byphenolextractionwerelayeredontosucrosegradients,centrifuged,
andfractionated.Theopticaldensityofeachsamplewasdetermined
at260nm.Thearrowsindicatethesedimentationpositionsofthe
twomainrRNAspeciesofratbrainusedasmarkers,whichwereruninparallel.

significantlysmallerascomparedtovaluesreportedforothereukaryoticrRNAs.ThesamesedimentationvalueswereobtainedforrRNAsfromtheWBstrain(data
notshown).
AnalysisofSmallRNAs
InordertocharacterizethesmallRNAspeciesfromG.lamblia,RNApreparationsfromP1strainwerefractionatedbydenaturingpolyacrylamidegel
electrophoresis.TwomajorRNApopulationsandseveralotherlessabundantspeciesoflowermolecularweightswereobserved(Figure3).Thesizesofthetwo
prominentRNAmoleculesweredeterminedrelativetodenaturedratbrainandyeast5.8Sand5SRNAstandards,andcorrespondtoapproximately130and118
bases.ThesesizeswerealsocalculatedfortheWBstrainsmallRNAswhenanalyzedundersimilarconditions.Thus,thesesmallribosomalRNAs,liketheirlarge21S
and15Scounterparts,arealsosignificantlysmallerthanthe5.8Sand5SRNAsfoundineukaryoticcells(12).
AnalysisofinvitroTranslationProductsPrecipitatedByImmuneSera
TotalRNAobtainedfromP1andWBstrainswastranslatedinvitroinmRNAdependentrabbitreticulocytelysates.Supplementationofreticulocytelysateswith10
goftotalcellularRNAstimulatedincorporationof35Smethioninebyapproximately3to10fold,relativetoacontrolwithoutRNA.Analysisofinvitrotranslation
productsfromtotalRNArevealednumerousproteinswithmolecularweightsbetween20and150kDbySDSPAGE(Figure4).
Inordertoprecipitateantigensrecognizedbyserafrominfectedhumans,totalG.lambliaRNAwastranslatedinvitroandtheproductswereimmunoprecipitated
withhumanserasamples.Figure5showsthatarangeofinvitrotranslationproductsarerecognizedbyserafrombothasymptomatic(tracks1and2)and
symptomaticpatients(tracks3and4).Allserareactedwithmostoftheinvitroproducedpolypeptides.TheseresultsdemonstratethattheG.lambliaRNA
obtainedencodesawidevarietyofantigenswhichareimmunogenicinthehumanhost.
Discussion
ThereareseveralreportsconcerningtheisolationandcharacterizationofrRNAinprotozoa.InsomeprotozoathelargerRNAisapparentlyanintactpolynucleotide
chain(16).Inotherprotozoa,however,thelargerRNAislabile,anddissociatesunderdenaturingconditionsintotwofragmentssimilarinsizetotherRNAfound
associatedwiththesmallersubunit(1,6).
InthisstudywereporttheisolationofrRNAfromtwostrainsofG.lamblia:P1andWB.InordertodetectallrRNApopulationswehaveusedthreedifferent
methodstoisolatethetotalRNA.Theseincludeguanidineisothiocyanate,phenolextractionandhydroxylapatitechromatography.TheanalysisoftheRNAisolated
withthesemethodsrevealedonlytworRNApopulations.ThesizesoftheserRNAs,asdeterminedbygelelectrophoresisunderdenaturingconditions,wereof
approximately2390and1420nucleotides.ThesecorrespondtotheLSrRNAandSSrRNAspeciesrespectively.AsfarasthesmallRNApopulationsare
concerned,therearefivespecies.Twoofthem,ofapproximately130and118nucleotides,aremoreabundantandprobablycorrespondtothe5.8Sand5SRNAs
ofothereukaryotes.OuranalysisrevealedthattherRNAsofG.lambliaarethereforethesmallest

Figure3.
ElectrohoreticanalysisofsmallG.lambliaRNAs.Total
G.lambliaRNAwaselectrophoresedinan8.8%acrylamide,
0.2Mureagelunderdenaturingconditionsandstainedwithethidium
bromideasdescribedinMaterialsandMethods.Track1:totalRNAfrom
S.cerevisiaepolysomestrack2:totalRNAfromratbrain
track3:totalRNAfromG.lamblia.

Page184

Figure4.
AnalysisofcellfreeproductssynthesizedbytotalG.lambliaRNA.
Proteinssynthesizedinarabbitreticulocytesystemfromtotal
G.lambliaRNAsobtainedbyphenolextractionwereanalyzedona
10%polyaccrylamidegel.Track1:G.lambliatotalRNA
track2:TobaccoMosaicVirusRNAtrack3:NoaddedRNA.
Molecularweightmarkersareasindicated.

rRNAsreportedforeukaryoticcells,andinthisrespectdifferfromotherprotozoaaswell(12,24).Theseresultsarefurthersupportedbyotherstudiesfromour
groupinwhichtherRNAgeneswerefoundtobelocatedonasmallrepetitiveDNAunitof5.4kb(manuscriptsubmittedforpublication).Recentstudiesby
Boothroydetal.(2),andEdlindandChakraborty,(10)showedsimilarresultsregardingtherRNAsizesoftheseprotozoaaswellasthesizeoftherepeatedunit
encodingforrRNAs.
Theseresults,togetherwiththedataobtainedforotherprotozoashowaconsiderablediversityineukaryoticrRNAs,theirgenes,andtheprocessingmechanismsin
theseorganisms.Forinstance,thenuclearribosomalrepeatunitofmosteukaryotesincludesthreematurerRNAspecies:the18S,5.8S,and28SrRNAs.Theseare
processedfromasinglelargeprimarytranscript(forreviewsee22).However,insomeorganisms,otherprocessingstepsoccurwhichresultinunusualspeciesof
thesethreerRNAs(1,6,17,21,31).Together,allthesedatamayreflectdifferencesinthetranslationapparatusoftheseorganisms.
FurthercharacterizationofG.lambliarRNAsandtheirgeneswillcertainlycontributetoabetterunderstandingofthemeaningoftheseparticulardifferences.Onthe
otherhand,theribosomalRNAgenesareamongsomeofthemostconserved,universallydistributed,andfunctionallyequivalentgenesofallorganisms.These
characteristicsmakethesegeneswellsuitedfordefiningevolutionary

Figure5.
Immunoprecipitationof35Smethionineinvitrolabelled
translationproductsbyhumansera.Serafrompatientsinfected
withG.lambliawereusedtoprecipitateinvitrotranslationproducts
asdescribedinMaterialsandMethods.Tracks1and2:invitro
translationproductsprecipitatedbyserafromtwoasymptomaticpatients.
Track3and4:invitrotranslationproductsprecipitatedbyserafrom
twosymptomaticpatients.Track5:invitrotranslationproductsprecipitated
bycontrolhumanserum.Molecularweightmarkersareasindicated.

relationshipsamongeukaryoticorganisms(20,26,29)andwillhelptoplaceG.lambliainthecorrectevolutionaryposition.
InthepresentstudytotalRNAsfromtheG.lambliaP1strainwereusedtoconductinvitrotranslationofpolypeptides.Inthiscase,afullspectrumofparasite
componentswasobtained.Mostoftheinvitrotranslatedproductswereprecipitatedwithhumanimmunesera,suggestingthattotalRNApreparationscontainmost
oftheRNAmessangerswhichencodeforimmunogeniccomponentsoftheparasite.Nomajordifferencesweredetectedintheimmunoprecipitationpatternsobtained
whenserafromasymptomaticandsymptomaticpatientsincludedinthisstudyweretested.Finally,wehavepurifiedpolyA+RNAfromtotalRNAwhichhasbeen
usedtoprepareacDNAlibrary.Thiswillallowustoidentifyimportantproteinsinvolvedintheimmuneresponseofhumansinfectedwiththeparasiteaswellasto
studytheinductionandcontroloftheseandothercomponentsofthisimportantpathogen.
Acknowledgements
WewouldliketothankDr.A.Ratray,Dr.D.PeattieandDr.I.MezaforcriticallyreadingthismanuscriptandMrs.R.Barreraforsecretarialassistance.C.Ovando
isa

Page185

recipientofaCONACyTstudentship.ThisworkwassupportedinpartbyCONACyT(Mexico),COSNET(SEPMexico)andMacArthurFoundation(U.S.A.).
LiteratureCited
1.Albach,R.,Prachayasittiko,V.andG.Heebner.1984.Mol.Biochem.Parsitol.12:261272.
2.Boothroyd,J.C.,Wang,A.,Campbell,D.A.andC.C.Wang.1987.Nucl.Acids.Res.15:40654084.
3.Brawerman,G.1974.In:MethodsinEnzymologyXXX.F.K.MoldaveandL.Grossman(eds).AcademicPress,NewYork.pp.605612.
4.Brosius,J.,Doll,T.,andH.F.Noller.1980.Proc.Natl.Acad.Sci.77:201204.
5.Brosius,J.,Palmer,M.,Kennedy,P.andH.F.Noller.1978.Proc.Natl.Acad.Sci.75:48014805.
6.Castro,C.,Hernandez,R.andM.Castaneda.1981.Mol.Biochem.Parasitol.2:219233.
7.Chirgwin,J.,Przybyla,A.,MacDonald,R.andW.Rutter.1979.Biochem.18:52945299.
8.Dame,J.B.andT.F.McCutchan.1983.Mol.Biochem.Parasitol.8:263297.
9.David,P.H.,Hadley,T.J.,Aikawa,M.andL.H.Miller.1984.Mol.Biochem.Parasitol.11:267282.
10.Edlind,T.D.andR.R.Chakraborty.1987.Nucl.Acids.Res.15:78897901.
11.Engvall,E.andP.Perlmann.1972.J.Immunol.109:129135.
12.Erdmann,U.A.1982.Nucl.Acids.Res.10:93115.
13.Faust,E.C.,Russell,P.F.,andR.C.Jung.1970.In:CraigandFaustClinicalParasitology.8thed.LeaandFebiger(eds).Philadelphia.pp5974.
14.Hadjiolov,A.A.,Georgiev,O.J.,Nosikov,V.V.,andL.P.Yavachev.1984.Nucl.Acids.Res.12:36773693.
15.Hernandez,R.,Nava,G.andM.Castaneda.1983.Mol.Biochem.Parasitol.8:297304.
16.Hyde,J.E.,Zolg,J.W.andJ.S.Scaife.1981.Mol.Biochem.Parasitol.4:283290.
17.Jordan,B.R.,LatilDamotte,M.,andR.Jourdan.1980.Nucl.Acids.Res.8:35653573.
18.Keister,D.B.1983.Trans.R.Soc.Trop.Med.Hyg.77:487488.
19.Laemmli,U.K.1970.Nature.London.277:680687.
20.Lane,D.J.,Pace,B.,Olsen,G.S.,Stahl,D.A.,Sogin,M.L.andN.R.Pace.1985.Proc.Natl.Acad.Sci.82:69556959.
21.Leipoldt,M.andJ.Schmidth.1982.In:GenomeEvolution.G.A.DoverandR.B.Florell(eds).AcademicPress,NewYork,N.Y.pp.219236.
22.Mandal,R.1984.ProgressinNucleicAcidResearchandMolecularBiology.31:115159.
23.Markov,G.G.andI.G.Ivanov.1974.Anal.Biochem.59:555563.
24.Noller,H.F.1984.Ann.Rev.Biochem.53:119162.
25.Osterman,L.A.1984.In:MethodsofProteinandNucleicAcidResearch.BerlinSpringerVerlag,BerlinHeidelberg.pp.241274.
26.Pace,N.R.,Olsen,G.J.andC.R.Woese.1986.Cell45:325326.
27.Panasci,L.D.,Green,D.C.,Fox,P.A.andP.S.Schein.1977.Anal.Biochem.83:678688.
28.Rhoades,R.1975.J.Biol.Chem.250:80888097.
29.Sogin,M.L.,Elwood,H.J.andJ.H.Gunderson.1986.Proc.Natl.Acad.Sci.83:13831387.
30.Torezynski,R.,Bollon,A.P.andM.Fuke.1983.Nucl.Acids.Res.11:48794890.
31.Ware,V.C.,Renakawitz,R.andS.Gerbi.1985.Nucl.Acids.Res.13:35813597.

Page187

EnzymeActivitiesofGiardialambliaandGiardiamurisTrophozoitesandCysts
DonaldG.Lindmark*andJamesJ.Miller
ClevelandStateUniversity,Cleveland,Ohio44115,U.S.A.
Giardialambliaisthemostcommonintestinalprotozoanparasiteintheworld.Itexistsintwoforms,theactivelygrowingtrophozoiteandtheinfectiveresistantcyst.Theinformation
availableontheparasite'scarbohydrateandenergymetabolismandhydrolyticabilitiesarelimitedtodataonthetrophozoite.Wereportheredataobtainedonthecarbohydrate
andenergymetabolismandhydrolyticabilitiesofthecyststageofG.lamblia.Inaddition,comparativedataarepresentedonthecarbohydrateandenergymetabolismand
hydrolyticabilitiesofthetrophozoiteandcyststagesofGiardiamuris.ThefollowingenzymeactivitiesweredetectedinhomogenatesofthetrophozoitesandcystsofbothG.lamblia
andG.muris:hexokinase,pyruvatekinase,phosphoenolpyruvatecarboxykinase,pruvate:ferredoxinoxidoreductase,alcoholdehydrogenase,NADPHoxidoreductase,malate
dehydrogenase,malatedehydrogenase(decarboxylating),acidphosphatase,DNaseandRNase.Theseenzymesshowedsimilarlevelsofactivityand,theenzymesofcarbohydrate
andenergymetabolism,similarcharacteristics,betweenspeciesandstages(trophozoiteandcyst).Thehydrolyticenzymeswerealsosimilarinspecificactivitiesamongthespecies
andstages.Nocarbohydratesplittinghydrolasescouldbedetectedinanyspecies.

Introduction
Giardialambliaisthemostcommonintestinalprotozoanparasiteintheworld(2).Itexistsintwoforms,theactivelygrowingtrophozoiteandtheinfectiveresistant
cyst.Theinformationavailableontheparasite'scarbohydrateandenergymetabolismandhydrolyticabilitiesislimitedtodataonthetrophozoite(2).Usingthegerbil
(aproposedmodelforhumangiardiasis)asasourceofG.lambliacysts(1),weinvestigatedtheenzymesofcarbohydrateandenergymetabolisminthisstagein
ordertobetterunderstandthepotentialthatthecysthasforcarbohydrateandenergymetabolism.Sincehydrolyticprocessesmaybeinvolvedintheexitof
trophozoitesfromthecyst(excystation),weinvestigatedthehydrolyticabilitiesofbothstagestogaininsightintotheexcystationandpossibilytheencystation
processes.
SinceG.muriscysts(andtrophozoitesobtainedbyinvitroexcystation)canbeobtainedinlargequantitiesusingthemousemodel(6),G.murishasbeenusedin
manystudiesasasubstituteforG.lamblia.HencewebelievedthatastudyofthemetabolicpotentialofG.muris(trophozoiteandcyst)wasofimportanceand
thereforehaveobtainedcomparativeresultsonthecarbohydrateandenergymetabolismandhydrolyticenzymesfromboththetrophozoiteandcyststagesofG.
muris.
Ingeneral,therearefewdifferencesinthespecificactivitiesoftheenzymesofcarbohydrateandenergymetabolismandthehydrolyticenzymesbetweenG.lamblia
andG.murisandthetrophozoiteandcyststage.Thecharacteristicsoftheimportantenzymesofcarbohydrateandenergymetabolismsareverysimilarifnot
identificalinbothparasites.
MaterialsandMethods
Organisms
TrophozoitesofG.lamblia(PortlandIstrain)weregrownandharvestedasdescribedbyLindmark(2).CystsofG.lamblia(1106per20animals)wereobtained
fromthegerbilmodelasdescribedbyBelosevicetal.(1).CystsofG.muris(1107per30animals)wereobtainedfromthemousemodelasdescribedby
RobertsThomsonetal.(6).GiardiacystswerepurifiedbysucrosegradientcentrifugationandvelocitysedimentationasdescribedbySauch(7).TrophozoitesofG.
muriswereobtainedbyexcystationofcystsasdescribedbyRiceandSchaefer(5)andpurifiedbysucrosegradientcentrifugation(1105per30animalsafter
excystationandpurification).Thecystandtrophozoitepreparationswerestoredaspelletsafterwashing2timesin0.25Msucroseat70Cuntilenoughwere
accumulatedforenzymeanalysis.ThelowquantitiesofcystsandG.muristrophozoitesavailableforenzymeanalysisandcharacterizationmadesomeexperiments,
suchasKm determinationsofsubstratesandcofactors,impossible.Forexample,asingleharvestofcystsofG.murisproducedenoughhomogenatetoconduct12
assayswithcontrolsformalatedehydrogenase(theenzymewiththehighestspecificactivityinGiardiapreparations).
EnzymeAssays
HomogenatestobeusedforenzymeanalysiswerepreparedfromtrophozoiteswithaPotterElvehjamhomogenizerasdescribedbyLindmark(2).Homogenatesof
cystswerepreparedinthesamemanneronpreparationsthatwerefrozenandthawed5Xinthepresenceof0.2%TritonX100(thiswasdonetorupturethe
resistantcystwall).Forassaysandcharacterizationofoxygensensitiveenzymes,homogenateswerepreparedasdescribedbyLindmarkunderArgoninthepresence
ofmercaptoethanolandstoredunderArgon.Publishedassayswereusedformalatedehydrogenase(EC1.1.1.37),malatedehydrogenase(decarbonylating)(EC
1.1.1.39),fumaratehydratase(EC4.2.1.2),lactatedehydrogenase(EC1.1.1.27),catalase(EC1.11.1.6)(3),hydrogenase(EC1.18.3.1),pyruvate:ferredoxin
oxidoreductase(EC1.2.7.1)(2),succinatedehydrogenase(EC1.3.99.1),acidphosphatase(EC3.1.3.2),citratesynthase(EC4.1.3.7),isocitratedehydrogenase
(EC1.1.1.42),protein(4),hexokinase(EC2.7.1.1),pyruvatekinase(EC2.7.1.40),
*Correspondingauthor.

Page188

phosphoenolpyruvatecarboxykinase(EC4.1.1.32),alcoholdehydrogenase(EC1.1.1.2),NADPHoxidoreductase(EC1.6.99.1),acetatekinase(EC2.7.2.1)(2),
BNacetylglucosaminidase(EC3.2.3.30),Bgalactosidase(EC3.2.1.23),Bglucuronidase(EC3.2.1.3),DNase(EC3.1.4.5),andRNase(EC2.7.7.16)(2).
ThepHdependenceoftheactivitiesoftheenzymeswasdeterminedin100mMtrisphosphatebuffer.Enzymeunitsweredefinedastheamountofenzymenecessary
toform1molofproductortodegrade1molofsubstrateperminuteundertheassayconditionsstated.Allenzymeswereassayedat30Cunlessotherwisestated.
Results
EnzymeActivities
ThespecificactivitiesofenzymesassayedinhomogenatesaregiveninTable1.Thefollowingenzymeswerebelowthelimitofdetection:citratesynthase,isocitrate
dehydrogenase,succinatedehydrogenase,fumaratehydratase,lactatedehydrogenases,acetatekinase,hydrogenase,catalase,Bglucuronidase,Bgalactosidase,and
BNacetylglucosaminidase.
EnzymeProperties
Asshownbelow,comparisonsofthecharacteristicsofthemajorenzymesofenergymetabolismamongthehomogenatespreparedfromthedifferentspeciesand
stagesrevealedmanysimilarities.UnlessotherwisestatedthefollowingholdstrueforthetrophozoiteandcystformsofG.lambliaandG.muris.Theresultsobtained
withtrophozoitesofG.lambliaagreewiththoseofLindmark(2).
Pyruvate:FerredoxinOxidoReductase
TheactivitiesrequireCoenzymeA(0.1mM)andthiolcompounds(dithiolthreitolormercaptoethanol)forfullactivity.Theenzymesareoxygensensitivewith60%loss
inactivityinthepresenceofairin2h.FMN(0.5mM),FAD(0.05mM),andferredoxin(0.5mg/mL)canbeusedaselectronacceptors.NAD(5mM)andNADP
(5mM)areineffectiveaselectronacceptors(2).TheseexperimentswerenotdonewithG.muristrophozoitesbecauseofthedifficultyinobtainingsufficient
quantities.
PyruvateKinase
TheactivitieshaveapHoptimumofapproximately7.2.TheenzymesrequireADPandMg2+IDPandGDPcannotsubstituteforADP.Ca2+,Mn2+,andCo2+cannot
substituteforMg2+.
MalateDehydrogenase
TheactivitieshaveapHoptimumofapproximately7.0.NADHisthemainelectrondonor.NADPHis25%aseffective.
MalateDehydrogenase(Decarboxylating)
TheactivitieshaveapHoptimumofapproximately7.3.NADPisrequired,NADwillnotsubstitute.Theenzymesarecompletelyinhibitedby1mMEDTAandshow
arequirementforadivalentcation(Mn2+,Co2+Fe2+).Mg2+andCa2+areineffective.
NADPHOxidoreductase
TheactivitieshaveapHoptimumofapproximately7.3.Likepyruvate:ferredoxinoxidoreductasetheactivitiesareoxygensensitivelosing50%activityinthepresence
ofairin2h.NoactivitycouldbedetectedwithNADHasanelectrondonor.ExperimentswerenotperformedwithG.muristrophozoites.
AlcoholDehydrogenase
NADPHisrequired.NADHisineffective.Theenzymesareirreversible,onlyutilizingacetaldehydeasasubtrate.Ethanol,isopropanolandproanolcannotbeusedas
substrates.
Insummarywithinthelimitsofthisstudy,theenzymeactivitiesfoundintrophozoitesofG.lambliabyLindmark(2)arealsofoundinthecystsofG.lambliaandthe
trophozoitesandcystsofG.muris.Enzymeactivitiesbelowthelevelofdetection(2)inG.lambliatrophozoites,suchascarbohydratesplittinghydrolasesand
enzymesoftheKrebscycle,arebelowthelevelofdetectionincystsofG.lambliaandtrophozoitesandcystsofG.muris.Thespecificactivitiesoftheenzymes
detectedaresimilar.Thecharacteristicsoftheenzymesofcarbohydrateandenergymetabolismaresimilarbetweenspeciesandstages.
Discussion
EarlierstudiesbyLindmark(2)havedemonstratedtheoccurrenceofmanyenzymeactivitiesintrophozoitesofG.lamblia.Ourresultspresentedhereconfirmthe
above
TABLE1.SpecificactivitiesofenzymesinGiardia.
Activity
mU/mgproteinS.D.(No.ofdeterminations)
G.lamblia
Enzyme

G.muris
Cyst

Trophozoite

Cyst

254(4)

206(5)

153(4)

104(4)

14015(3)

10120(3)

19039(3)

1105(2)

202(5)

1814(6)

224(3)

102(3)

Malatedehydrogenase

85032(6)

70115(5)

60040(5)

51018(5)

Malatedehydrogenase
(decarbonylating)

12015(8)

8510(3)

10030(3)

6010(2)

Pyruvate:ferredoxinoxidoreductase

35060(10)

21054(11)

20041(2)

24033(3)

Alcoholdehydrogenase

31010(5)

20035(4)

39026(2)

21040(4)

NADPHoxidoreductase

39030(6)

30020(3)

30038(2)

20051(3)

Acidphosphatase

804(10)

6010(12)

8510(3)

4210(4)

DNase

5815(8)

4018(3)

8417(2)

3018(2)

RNase

4211(4)

345(2)

6020(2)

283(2)

Hexokinase
Pyruvatekinase
Phosphoenolpyruvatecarboxykinase

Trophozoite

Allassayswereconductedat30Cexceptalcoholdehydrogenase(19C).

Page189

mentionedresearchanddemonstratethattremendoussimilaritiesexistamongtheenzymesofcarbohydrateandenergymetabolismoftrophozoitesofG.lambliaand
G.muris.Thespecificactivitiesoftheenzymesaresimilarandthecharacteristicsofpyruvatekinase,malatedehydrogenase,malatedehydrogenase(decarboxylating),
pyruvate:ferredoxinoxidoreductase,alcoholdehydrogenaseandNADPHoxidoreductaseareidentical.Otherenzymeactivitiespossiblyinvolvedincarbohydrateand
energymetabolismsuchastheenzymesoftheKrebscycle,catalase,hydrogenase,lactatedehydrogenaseandacetatekinasewerebelowthelevelofdetectionin
homogenates.ThesesimilaritiessuggestthatG.lambliaandG.murishavesimilarpathwaysofcarbohydrateandenergymetabolism.Inaddition,thecystsofboth
specieshavethesameenzymeticpotentialofthetrophozoite,suggestingthattheenzymesofcarbohydrateandenergymetabolismarepresentinthecyst,tocarryout
metabolismofthetrophozoiteuponexcystation.
Thecystsandtrophozoitesofeachspecieshavethesamecomplementofhydrolyticenzymes.Bothformslacktheenzymesneededtosplitcomplexcarboydrates(B
galactosidase,Bglucuroinadse,BNacetylglucosaminidase).Thisfindingshouldbegivenconsiderationinfuturestudiesontheexcystationprocessandthechemical
componentsofthecystwall.
OurresultsdemonstratealargedegreeofsimilaritybetweenG.lambliaandG.murisintheenzymesofcarbohydrateandenergymetabolismandhydrolyticenzyme
activities,andalsoshowthatthetrophozoitesandcystsofeachspeciesexhibitthesamemetabolicpotential.
Acknowledgements
TheresearchwassupportedbyanAcademicChallengeGrantinParasitologyfromtheOhioBoardofRegents,Columbus,OhioTheThrasherResearchFundand
WHO(PDP)(P2/181/20).
LiteratureCited
1.Belosevic,M.,Faubert,G.M.,MacLean,J.D.,Law,C.,andN.A.Croll.1983.GiardialambliainfectionsinMongoliangerbils:ananimalmodel.J.Inf.Diseases
147(2):222226.
2.Lindmark,D.G.1980.EnergymetabolismofGiardialambliatrophozoites.Mol.Biochem.Parasitol.1:112.
3.Muller,M.1973.Biochemicalcytologyoftrichomonadflagellates.I.Subcellularlocalizationofhydrolases,dehydrogenases,andcatalaseinTritrichomonas
foetus.J.CellBiol.57:453474.
4.Muller,M.,Hogg,J.F.,andC.deDuve.1968.DistributionoftricarboxylicacidcycleandglyozylatecycleenzymesinTetrahymenapyriformis.J.Biol.Chem.
243:53855395.
5.Rice,E.,andF.W.Schaefer,III.1981.ImprovedinvitroexcystationprocedureforGiardialambliacysts.J.Clin.Micro.14(6):709710.
6.RobertsThomson,I.C.,Stevens,D.P.,Mahmoud,A.A.F.andK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gastroentoerology.71:5761.
7.Sauch,J.F.1984.PurificationofGiardiamuriscystsbyvelocitysedimentation.Appl.Environ.Micro.48:454455.

Page191

StudiesonGiardiaLambliaTrophozoiteAntigensUsingSephacrylS300ColumnChromatography,PolyacrylamideGel
ElectrophoresisandEnzymelinkedImmunosorbentAssay
P.P.Chaudhuri,S.Pal,S.C.Pal,andP.Das*
DepartmentofParasitology,NationalInstituteofCholeraandEntericDiseases,P33,C.I.T.Road,SchemeXM,Beliaghata,Calcutta700010India
AntigenpreparedfromGiardialambliatrophozoitesculturedinvitroinDiamond'sTYIS33mediumwasanalyzedusingsodiumdodecylsulfatepolyacrylamidegelelectrophoresis,
SephacrylS300columnchromatography,counterimmunoelectrophoresisandenzymelinkedimmunosorbentassay.AfterelutionthroughS300columnfourdistinctfractionswere
obtained.Molecularweightsofthesefractionswere150,00065,00050,000and10,000daltonsforFI,FII,FIII,andFIVrespectively.TheSDSPAGEanalysisrevealedaminimum
of28distinctbandswithcrudeantigenand13,22,26and30bandswithFI,FII,FIIIandFIVfractionsrespectively.Themolecularweightofthesebandsrangedfrom125,000to
14,000daltons.AntigenicactivitywasobservedinallfourfractionsintheCIEPtest.However,whenassayedbytheELISAtestthemaximumantigenicactivitywaslinkedtothe
highermolecularweightfraction.

Introduction
Giardialamblia,aflagellatedprotozoanparasitethatthrivesintheupperintestineofhumansandcausesaspectrumofdiseases,includingasymptomaticcarriage,
acutefulminatingdiarrheaandchronicdiarrheawithmalabsorption.Thediseasemostcommonlyoccursininfantsandchildren(18),particularlythoseattendingday
carecenters(8),travelers(3),homosexuals(13),hypogammaglobulinanaemicsandbackpackers(3).Inaddition,theorganismhasbeenestablishedastheetiologic
agentofnumerousoutbreaksofdiarrhealdiseaseinvariouspartsoftheworld(4,7).
DespitetheconsiderablemorbiditycausedbyG.lambliaverylittleisknownabouttheantigenicconfigurationofthisprotozoan.Earlierworkonthisparasiteby
variousworkers(11,12,15,16)suggeststhattheorganismisantigenicallyacomplexmoiety.
TodefinetheantigenicnatureofG.lambliatrophozoite(strainPortland1)further,SephacrylS300columnchromatographyforfractionation,sodiumdodecylsulfate
polyacrylamidegelelectrophoresis(SDSPAGE)forthecomparisonofproteinconstituentsinCSAanditsfractions,andcounterimmunoelectrophoresis(CIEP)and
enzymelinkedimmunosorbentassay(ELISA)forimmunologicalactivitywereused.Thecrudesolubleantigen(CSA)anditsfractionswerepolydisperseintheir
molecularweightwhenanalyzedbothinSephacrylS300columnchromatographyandSDSPAGEwithabroadspectrumofimmunogenicity.
MaterialsandMethods
ParasiteCulture
TheP1strainofG.lambliawassubculturedtwiceperweekat37CinfiltersterilizedTYIS33medium(6),supplementedwithvitaminsand10%heatinactivated
adultbovineserum,penicillin(50g/mL)andstreptomycin(50g/mL)asantibiotics.
AntigenPreparation
ActivelygrowingG.lambliatrophozoitesshowingexponentialgrowth(7296h)weredislodgedfromthewallsofculturetubesbyimmersioninanicebathfor10
minutesfollowedbycentrifugationat800gfor5minutes.Pooledviabletrophozoiteswerewashed5timesinsterilephosphatebufferedsaline(PBSpH7.4,0.05
M)andfinallyresuspendedinnormalsaline.Thissuspensionwasthensonicatedinanicebathwitheight30secbursts(MSESonicator,U.K.).Thesonicatedmaterial
wascentrifugedat10,000gfor20minutesat4C.Thesupernatantwascollectedascrudesolubleantigen(CSA)andusedforantigenicanalysisaftertheprotein
contentswereestimatedbythemethodofLowryetal.(10).
PreparationofAntisera
Albinorabbitsweighing23kgwereimmunizedwithG.lambliaCSA.About2mgofantigenicproteinin0.5mLwereemulsifiedwithanequalvolumeofFreund's
completeadjuvant(Difco)andinjectedsubcutaneouslyintothehindlegsofeachrabbit.Atotalofthreesuchinjectionsweregiventoeachrabbitatweeklyintervals.
ThiswasfollowedbythreeintravenousinjectionswithCSAalone(approximately34mgofantigenicproteinperrabbit)at2dayintervals.Animalswerebledaweek
afterthelastinjectionandtheprecipitatingantibodyintheimmunerabbitserumwasdetectedbyaCIEPtestagainstthehomologousantigen.
FractionationofCSAbySephacrylS300(gelfiltration)ColumnChromatography
CSAofG.lambliawassubjectedtogelfiltrationthroughSephacrylS300columnsinordertoseparateitsantigenicfractions.About80mLofpreswollenSephacryl
S300(PharmaciaFineChemicals,Sweden)wetbeaddiameter40105mwaspouredinaglasscolumn(1.640cm,PharmaciaFineChemicals,Sweden).A
flowrateof20mLperhourwasmaintainedwithaperistalticpump(Pharmacia)throughouttheexperiment.Thevoidvolumeofthecolumnwasdeterminedby
applyingDextranBlue2000.
About1.5mL(i.e.12%ofbedvolumeofcolumn)ofCSAwhichwaspreviouslydialyzedwitheluentbuffercontainingabout24mgofprotein,wasappliedand3.5
mLfractionswerecollectedineachtubewiththehelpofanautomaticfractioncollector(Frac100,PharmaciaFineChemicals,Sweden).Anelutionprofilewas
obtainedbymeasuringtheopticaldensity(O.D.)
*Correspondingauthor.

Page192

at280nmandthesevalueswereplottedagainsttheelutionnumbers(tubes).AccordingtotheO.D.valueseachpeakandtrailingeluates(Figure1)werepooled
separatelyandlistedbyfractions.Thesefractionswereconcentratedbylyophilizationanddialyzedagainstphysiologicalsaline.Finally1mgprotein/mLofsolutionwas
preparedandstoredat4Cuntilusedforfurtherstudies.ThemolecularweightofeachfractionwasdeterminedbycomparingtheKavvalueofeachfractionwiththe
knownmol.wt.MarkerofgelfiltrationproteinssuppliedbyPharmaciaFineChemicals,Sweden.
SDSPAGEAnalysis
TheproteinconstituentsofCSAaswellasthefourfractionsseparatedbyS300columnchromatographywerecomparedbySDSPAGEusing10%separatinggels,
in0.5trisHClbufferpH6.8.Theelectrodebufferandsamplepreparationwasmadeafterfollowingstandardprocedures(9).
SDSPAGEanalysiswasperformedinaverticalslabgelelectrophoresischamber(LKB,Sweden)withaconstanttemperatureof15C.Aconstant120Vwas
appliedoncetheGiardiaantigenhadenteredtheseparatinggel.Knownmolecularweightproteinstandards(PharmaciaFineChemicals,Sweden)wererun
simultaneously.Thegelswerefixed,washedandstainedforproteinwith0.125%CoomassieBrilliantBluedye.
CIEPTest
TheantigenicactivityofCSAanditsfourfractionswerecomparedincounterimmunoelectrophoresistestsagainstthedifferentdilutionsofrabbitantiGiardia
antibodies.TheCIEPtestprocedurewasessentiallythesameasthatdescribedbySharmaetal.(14).Briefly,approximately3mLof1%agarose(SigmaChemicals,
U.S.A.)inbarbitalbuffer(pH8.8,0.05M)waslayeredonmicroslidesandwellsforantigenandantibodywerepunchedaccordingtothestandardsize.The
antibodywellswereplacedtotheanodalsideandtheantigenwellstothecathodalend.Theexperimentwasruninaelectrophoresischamber(Shandon,U.S.A.)ata
constant160Vfor30minutes.Readingsweretakenjustaftertheexperimentandalsoafter24hofincubationat4C.
ELISATest
ThemicroELISAtestperformedwasthesameasdescribedearlierintheserodiagnosisofamoebiasis(5),withslightmodification.Inthisexperimentaknownamount
ofantibody(i.e.1:200dilution)wasusedtoreactwithvariableamountsofantigen.FourpreimmunerabbitseraandfourimmunizedagainstGiardiaCSArabbitsera
werepooledinseparatebatchesandusedasreferencepositiveandnegativeserafortheexperiment.Beforetheexperimentwasconductedachequerboardtitration
wascarriedouttodeterminetheoptimalconcentrationsofantigenandantibodyrequired.A20gpermLG.lambliaproteinfromCSAanda1:200dilutionof
pooledpositiveandnegativeserawerefoundoptimalandspecificforobtainingaclearcutdistinctionbetweenpositiveandnegativeresults.
ForcomparisonofantigenicactivityamongCSAanditsfractions,differentconcentrationsofantigenicproteinviz.20g,10g,5g,2.5gand1.2gpermLwere
usedinthistest.A1:1000dilutionofantirabbitIgGlabelledwithhorseradishperoxidase(SigmaChemicals,U.S.A.)wasusedasaconjugate.Resultswereread
photometricallyat490nm(O.D.)inanautomaticELISAreader(DynatechLabs,U.S.A.).
Results
TheSephacrylS300gelfiltrationpatternofGiardiaCSAispresentedinFigure1.Twomajorlightabsorbingpeakswereobserved,oneofwhichappearedinthe
voidvolumeandtheotherattheendoftotalcolumnvolume.Thelastpeakmostlyconsistedofyellowishcolouringmaterial(originallyseenintheCSA).Eluted
materialswereappropriatelypooledasindicatedInFigure1togivefourdifferentfractionsFItoFIVmolecularweightsofFItoFIVwere150K,65K,50K,and
10Kdaltons,respectively.

Figure1.
ChromatographicpatternofaxenicGiardialambliacrude
solubleantigen(CSA)onSephacrylS300gel.Columndimensions
were1.640cm,samplesize24mg/1.5mL0.05
MtrisHClbufferpH7.6,flowratewas20mL/h.

IntheCIEPtesttheFIfractionshowedapositiveprecipitinreactionupto1:64titreofimmunizedrabbitserumtheotherfractionsaswellasCSAitselffailedtoshow
reactivitybeyond1:8titreofimmunizedrabbitserum,althoughthesameconcentration(1mg/mL)of

Figure2.
Bardiagrammaticrepresentationsofthedistributionof
precipitinactivitiesofcrudesolubleantigen(CSA)anditsdifferent
fractions(FIFIV).Antigenicactivitywasdeterminedbyacounter
immunoelectrophoresistestingofdifferentdilutionsofpooled
positiveimmunizedrabbitseratoGiardiaCSA.

Page193
TABLE1.Comparisionofopticaldensityvaluesfordifferentconcentrationsofcrudesolubleantigen(CSA)andits
fractions(FIFIV).
Antigenicprotein
concentrations
(g/mL)

MeanO.D.valuesa
FI

FII

FIII

FIV

20

0.243

CSA

0.556

0.410

0.256

0.156

10

0.203

0.500

0.318

0.196

0.079

0.138

0.426

0.235

0.110

0.051

2.5

0.120

0.250

0.135

0.128

0.023

1.2

0.039

0.165

0.092

0.059

O.D.valuesweretakenat490nm.

antigenicproteinwasusedineachcase.
IntheELISAtestaclearcutdistinctioninO.D.valuewasobservedwhen20gofcrudesolubleantigenwasusedagainsta1:200dilutionofpositiveandnegative
serum.ThefourpreimmunenegativeserashowedanO.D.valuerangingfrom0.05to0.09withameanO.D.of0.07at490nm.Thefourimmunizedrabbitserawith
thesamedilutionshowedO.D.valuesrangingbetween0.12and0.325withameanO.D.of0.234.Thecutoffvalueindicatinganegativeorapositiveresultwas
takenas0.100(meanO.D.ofcontrols).TheresultsofCSAanditsfractionswhencomparedusingtheELISAtestforantigenicactivityagainstthepooledimmunized
rabbitseraareshowninTable1.Afteranalysis,fractionIwasfoundtobemostantigenicascomparedtotheparent(CSA)andtheotherthreefractions.inother
words,the2.5g/mLproteinofthisfraction(FI)showedalmost

Figure3.
SDSPAGEprofileofGiardialambliacrudesolubleantigen(CSA),
anditsdifferentfractions(FIFIV).Lowmolecularweightmarker
proteinwasalsousedintheexperiment.

thesameO.D.valueasthoseobtainedwith20g/mLofCSA,5g/mLofFII,and20g/mLofFIII,respectively.FractionIVwasfoundtobetheleastsensitivein
comparisontoCSAandotherfractions.
ThepatternofSDSPAGEwithCSAanditsfractionsforcomparisonoftheirconstituentproteinsubunitsareshowninFigure3.TheCSAshowedseveral(about28)
discreteproteinbandsinthemolecularweightregionof~12.5104to~1.4104daltons.FractionIshowedasimilarbandingpatternwithlessnumberofsubunits
(about13bands).FractionIIshowedproteinbandsinthemolecularweightrangeof~9.4to~4.3104daltons.ThebandsofFIIIweremainlyconfinedtothe
molecularweightregionof<9.4104to2.1104daltons.However,theFIVcontainedlowmolecularweightproteinsubunitsrangingfrom~8.2104to~1.4
104daltons.
Discussion
ThepresentinvestigationdemonstratesthecomplexnatureofCSAofG.lamblia(strainP1)trophozoites.Theycontainabout28polypeptidesaccordingtoSDS
PAGEanalysisandthesefindingscorrespondtotheobservationsofSmithetal.(15).However,theresultsdifferslightlyfromthoseofMooreetal.(12)who
demonstratedabout20distinctproteindeterminantswiththesameexperiment.Asimilarstudy(17)usinganimmunoelectroprecipitiontestshowed24precipitinarcs.
Thesedifferencesintheproteinpolymersmaybeduetodifferentstrainsandmethodsusedbydifferentworkers.TheresultsoffractionationinHPLC(12)and
SephacrylS300columnchromatographyofCSAshowedalmostsimilarobservations.Inbothcasesthemaximumantigenicactivitywasrecordedinthehighmol.wt.
fraction(FI).FractionI,whichshowedmaximumantigenicity,containedonly13.5%ofthetotalproteinpresentinthewholeGiardiaextractisserologicallyactive.
FractionsII,IIIandIV,althoughtheyhadproteincontentsof22%,10%and2.5%respectivelyofthetotalextract,showedsignificantlylowerserologicalvalues.
Furthermore,theprecipitinreactionswhichhavebeendemonstratedinwholeCSAanditsvariousfractionscanbeexplainedonthebasisthatfractionsII,IIIandIV
arenotpureandarecontaminatedwiththeprecedingfraction.FractionsIIIandIVwhichwerefoundtoberelativelylessactivethantheothertwofractionsby
ELISAtestcouldperhapsthrowsomelightonthedifferencebetweentheELISAandprecipitintestforthedetectionofclinicalGiardiacasesasdescribedin
amoebiasis(5)where,thehaemaglutinationandIFAtestshowednegativeobservations.However,ELISAwasfoundpositivewiththesamesera.Itispossiblethat
thetwotestsdetectdifferentsubclassesofIgGantibodies.
Thedemonstrationthattheimmunologicactivitywasassociatedwithaparticularfraction(s)makesitpossibletoisolatetheseparticularfractionsbySephacrylS300
gelchromatographyandtousethesefractionsfortheproductionofamorespecificantiG.lambliaantiserum.Theproductionofmorespecificantigensand
antibodieswill

Page194

allowformoresensitiveassaystostudyandelucidatetheroletheseantigensplayduringG.lambliainfection.
LiteratureCited
1.Ament,M.E.,H.D.Ochs.,andS.D.Davis.1973.Structureandfunctionofthegastrointestinaltractinprimaryimmunodeficiencysyndromes:astudyof39
patients.Med.52:227248.
2.Barbour,A.G.,C.R.Nichols,andT.Fakushima.1976.Anoutbreakofgiardiasisinagroupofcampers.Am.J.Trop.Med.HyG.25:384389.
3.Brodsky,R.E.,H.C.Spencer,Jr.,andM.G.Schultz.1974.GiardiasisinAmericantravellerstotheSovietUnion.J.Infect.Dis.130:319323.
4.Craun,G.F.1979.Waterborneoutbreaksofgiardiasis,pp.127149.In:Jakubowski,W.andJ.C.Hoff(ed.),WaterborneTransmissionofGiardiasis.U.S.
Environ.Protect.Agency,Cincinnati,Ohio,EPA600/979001.
5.Das,P.,S.Pal,andS.C.Pal.1984.Evaluationofthemicroenzymelinkedimmunosorbentassay,indirecthaemagglutinationandindirectfluorescenceantibody
techniquesforserodiagnosisofamoebiasis.J.Diar.Dis.Res.2:238242.
6.Diamond,L.S.,D.R.Harlow,andC.C.Cunnick.1978.AnewmediumfortheaxeniccultivationofEntamoebahistolyticaandotherEntamoeba.Trans.R.Soc.
Trop.Med.Hyg.72:431432.
7.Juranek,D.1979.Waterbornegiardiasis,pp.150153.In:Jakubowski,W.andJ.C.Hoff(ed.),WaterborneTransmissionofGiardiasis.U.S.Environ.Protect.
Agency,Cincinnati,Ohio.
8.Keystone,J.S.,S.Krajden,andM.R.Warren.1978.PersontopersontransmissionofGiardialambliaindaycarenurseries.Can.Med.Assoc.J.119:241248.
9.Laemmli,U.K.1970.CleavageofstructuralproteinsduringtheassemblyoftheheadofbacteriophageT4.Nature(London)227:680685.
10.Lowry,O.H.,N.J.Rosebrough,A.L.Farr,andR.J.Randall.1951.Proteinmeasurementwithfolinphenolreagent.J.Biol.Chem.193:265275.
11.Meyer,E.A.,andS.Radulescu.1979.Giardiaandgiardiasis.Adv.Parasitol.17:147.
12.Moore,G.W.,F.S.Bernal,andM.V.Dennis.1982.CharacterizationofGiardialambliatrophozoiteantigensusingpolyacrylamidegelelectrophoresis,high
performanceliquidchromatographyandenzymelabelledimmunosorbentassay.Vet.Parasitol.10:229237.
13.Schmerin,M.J.,T.C.Jones,andH.Klein.1978.Giardiasis:associationwithhomosexuality.Ann.Intern.Med.88:801804.
14.Sharma,P.,P.Das,andG.P.Dutta.1981.RapiddiagnosisofamoebicliverabscessusingEntamoebahistolyticaantigen.Arch.Invest.Med.(Mex.)12:215
218.
15.Smith,P.D.,F.D.Gillin,N.A.Kaushal,andT.E.Nash.1982.AntigenicanalysisofGiardialambliafromAfghanistan,PuertoRico,EcuadorandOregon.Infect.
Immun.36:714719.
16.Visvesvara,G.S.,E.A.Meyer,andG.R.Healy.1976.AntigenicanalysisofGiardialamblia.AmericanSocietyofParasitologists,AbstractNo.68,51stAnnual
Meeting,SanAntonio,Texas,p.40.
17.Visvesvara,G.S.1981.Giardialamblia:America'snumber1intestinalparasite.Diag.Med.4:2429.
18.Visvesvara,G.S.1982.Giardiasisinchildren.J.Pediatr.Gastroenterol.Nutr.1:463465.

Page195

DETECTIONOFGIARDIACYSTS

Page197

AnOverviewoftheTechniquesUsedforDetectionofGiardiaCystsinSurfaceWater
CharlesP.Hibler
DepartmentofPathology,ColoradoStateUniversity,
FortCollins,Colorado80523,U.S.A..
AnalysisofwatersamplesfordetectionofGiardiacystsinsurfacewaternecessitatestrappingparticulatematerialonafiltercartridge,andselectiveseparationofparticulatesfrom
thecysts.Thetechniquesusedbyanalyststoisolatecystsareessentiallysimilarandtheresultsdonotdiffersignificantly.Someanalystsusethe''reference"technique,othersusean
overlay/underlaymodificationofthistechnique,andsomeuseimmunofluorescencetovisualizethecysts.Allofthetechniquesareequallysubjecttoandmaybeadverselyaffected
bywaterquality.Efficiencyisinverselyproportionaltotheturbidityofthesourcewater.Highturbidity,especiallyorganicturbidity,algae,freelivingprotozoans,andother
material(alumand/orpolymers)inthewatersourceseverelylimiteffectiverecoveryand/orvisualizationofthecysts.

Introduction
AttheSymposiumonWaterborneGiardiasisin1978,JakubowskiandEricksen(2)reviewedthetechniquespreviouslyemployedbyinvestigatorsattemptingto
recovercystsofGiardiafromsurfacewaterandpresentedtheEPAmethoddevelopedforrecovery.In1980theEPAbroughtagroupofexperiencedinvestigators
totheEPAheadquatersinCincinnati,Ohiotodiscusstheproblemswithdiagnosisandtosuggestamoreefficientprocedure.Theresultofthismeetingwasthe
"consensus"or"reference"method(3),themethodrecommendedinVolume16ofStandardMethods(1).
ThesamplingdevicedevelopedbytheEPA(2)hasbecomestandardequipmentforsamplingwater,butthefiltercartridges,theporosityofthesecartridges,andthe
methodsforisolatingcystsofGiardiafromtheparticulatestrappedbythecartridgeshaveundergoneconsiderablemodificationsincethe1978and1980meetings,
primarilythroughtrialanderroreffortsbyanumberofinvestigators.
ThebasicprinciplerequirestrappingGiardiacystsonafilter,recoveringthesecystsfromthefilter,andselectivelyconcentratingandidentifyingthecysts.Thetypeof
filtercartridgepreferred(glassfiber/epoxy,orlon,cottonorpolypropylene),theselectiveconcentrationmediaemployed(potassiumcitrate,sucrose,zincsulfateor
percoll)andthespecificproceduresdevelopedforanalysis(directlightmicroscopyorimmunofluorescence)usuallyvaryamonglaboratoriesbecauseexperienced
investigatorsdeveloppersonalpreferences.However,investigatorsattemptingtoimprovethecurrentstateoftheartforisolatingGiardiacystsfromwatergenerally
areinfrequentcommunication,exchangingideasandtechniquesconsequently,thebasictechniquesemployedbydifferentlaboratoriesandbyindividualstrainedat
oneoftheselaboratorieshavemanysimilarities.
Thefiltercartridgespreferred,theselectiveconcentrationmediaemployed,andthespecificproceduresdevelopedforanalysisareallequallysubjecttoand
compromisedbyvariationsinwaterquality.Thetypeandamountofmaterialsuspendedinwatervarieswiththesource(river,lake,spring,etc.),season,or
geographiclocationandaresubjecttospecificcircumstances(springrunoff,flood,construction,orthunderstorm).Anyamountofsuspendedmaterialinterfereswith
therecoveryofGiardiacystsandthehighertheturbiditythegreaterthelossofcysts,especiallyiftheturbidityisprimarilyorganic.Ifamunicipalfiltrationsystemuses
alumorpolymers,andsmallamountsofthesechemicalspassthesystem,coagulationofthesuspendedmaterialtrappedonthefiltercartridgemayeffectivelyprevent
recoveryofanycysts.
Healthywatersourcessupportanumberofaquaticorganisms(crustaceans,insects,algae,diatoms,protozoa,etc.)andlargenumbersoftheseorganisms,especially
someprotozoaandalgae,interferewithvisualizationofthecyst,eitherbecauseofsimilarityinsizeandshapeor,sometimes,duetosheernumbers,resultinginfatigue
atthemicroscope.
Giardiacystsarenotevenlydistributedinwaterandtheirdispersioninstreamsorriversisdependentuponvolumeandrateofflow(4).Despitetheirrelativelylight
weight(s.g.1.0451.050)cystssettleratherquicklyinslowlymovingorstandingwater.
ThemorphologicqualityofGiardiacystsvariesfromsampletosample,withinasample,betweenseasons,andduringspecialcircumstances(springrunoff,thunder
storms),generallyduetoamixingoffreshcystswitholdercysts.Insomesamplesallofthecystsmaybeexcellentandmorphologiccriterianecessaryforidentification
(twotofournuclei,axonemesandmedianbodies)arereadilyvisualizedat450Xwhileinothersamplesthesefeaturescanbevisualizedonlywithconsiderable
difficulty,oftennecessitatingamagnificationof1000X.
SomeGiardiacystsarequiteresistanttothetraumaassociatedwithsampling,extractionandconcentrationandwillcontinuetobemorphologicallyexcellentfor
severalweeks,evenwithoutpreservationhoweverinothersamples,evenfromthesamesitebutatadifferenttime,onlyasmallpercentageofthecystsfoundinitially
inthe

Page198

samplecanberecoveredafewdayslater.Sometimesthecystsaredeadandshrivelledtotheextenttheyarealmostunrecognizable,inothersamplescystsare
presentandappeartobeingoodconditionbutcannotbeselectivelyconcentrated(irrespectiveoftheconcentrationchemicalused,orthespecificgravityofthat
chemical)suggestingthattheintegrityofthecystmembranehasbeencompromisedandthespecificgravityofthecystaltered.Live,infectiouscystspossessa
cytoplasmthatisessentiallyclear(hyaline)whenviewedbyphasecontrastand/orbrightfieldmicroscopywhereasthosethataredeadand/ordyinghaveacoagulated
appearancetothecytoplasmandintracellularorganellesareeasilydetected.Visualassessmentcorrelatesextremelywellwiththefluorogenicdyes,fluorescein
diacetateandpropidiumiodide,usedtodeterminecystviability(7,8).
InterpretationofmorphologicqualityofGiardiacystsforviabilityissubjectiveatbest.Cyststhathavebeendeadformonths,orthoseinactivatedbychlorine,ozone,
etc.oftenappearmorphologicallyexcellentwhereasthosethatareconsideredtobemorphologicallypoorinquality(andthereforedead)couldhavebeenexcellent
(butnotnecessarilyalive)atthetimeofsampling.Timelagbetweensamplingandanalysis,mishandlingofthesample,activityofchlorineandthepropertiesofthe
cystsinthatparticularsamplecanallaffectcystmorphologyatthetimeofanalysis.
Allofthefactorsinterferingwithrecoveryand/orvisualizationmustbeconsideredwhenanalyzingasampleofwaterforGiardiacysts.Recoveryofaspecificnumber
ofcystsindicatesonlythatthisisapercentageofthetotalnumberpresent,apercentagethatisexpectedtobeinverselyproportionaltotheamountandtypeof
suspendedmaterialpresentandthemorphologicqualityofthecysts.
ThebasictechniquestobedescribedhereforrecoveryofGiardiacystsfromsurfaceand/orgroundwater,theproblemsassociatedwiththesetechniques,andthe
modificationsappliedinanefforttoovercometheproblemshaveevolvedfrom12yearsofexperiencewithover6500watersamplesfromgeographicareas
throughoutNorthAmerica,and6yearsofexperiencewithrecoveryofGiardiacystsfrompilotfiltrationsystemsseededwithcyststoevaluateefficiencyofthe
systems.
Discussion
SamplingDevice
Thedescriptionsandcommentsgivenbelowonthesamplingtechniquesarederivedprimarilyfromexperienceverylittleinformationontheprecision,sensitivityand
efficiencyofthesetechniquesappearsinthepublishedliterature.ThedevicedevelopedbytheEPAforrecoveryofGiardiacysts(2,1)hasbecomestandard
equipmentforsamplingwater.Thefiltercartridgesgenerallyusedare1m(nominal)porositydepthtypecartridges.Twoor3m(nominal)porositydepthtype
cartridgeswouldprobablybesuitable,butnonehavebeentested.Fivem(nominal)porositycartridgeswilltrapabout98%ofthecysts,buttoofewhavebeen
evaluatedforanytoberecommended.Sevenand10m(nominal)porosityfilterswillpassmanyofthecysts.Cartridgesshouldbehandledwithrubberglovesto
preventcontaminationandcareshouldbetakensotheirexposureisonlytothewaterandpackagingmaterials.
LaboratoriesperformingGiardiaanalysisoftenpreferdifferentcartridges,butthereislittledifferencebetweenorlon,polypropylene,cottonoracetateforrecoveryor
easeofprocessing.Thesurfacetypefilters,suchastheBalstonepoxyfiberglasstubeorthemembranefiltershavedefinitedisadvantagesbecausetheyarerapidly
pluggedbymaterialsuspendedinthewater,severelylimitingtheamountofturbidwaterthatcanbesampled.
Thefilterhousingshouldbehighimpactclearplastictopermitvisualizationoftheselectedfiltercartridge.Theinfluenthosegenerallyisadoublefemalecouplinghigh
pressurehose(dishwasherorwashingmachinetype),fourtoeightfeetinlength,whereastheeffluenthoseoftenisagardenhosemeasuring25feetorlonger,alength
sufficienttoreachafloordraininthefacilityortobedirectedaconsiderabledistanceawayfromtheinfluenthosewhensamplingfromastreamoralake.Awater
metergenerallyisattacheddirectlytotheeffluentsideofthefilterhousingandtheeffluenthoseattachedtothemeter.Rateofflowcanbesatisfactorilyregulatedwith
thefaucettowhichthedeviceisattachedor,ifnecessary,alimitingorificeflowcontrol.
Samplingdirectlyfromastreamorlakecanbedonewithanelectricorgasolinepoweredportablepump,orwithalightweight12voltmarinebilgepump(about3to
3.5gallons/minute)powereddirectlyfrombatteriesoranautomotivevehicle.Ifanelectricorgasolinepoweredpumpisusedthefilterhousingshouldbeplaced
betweentheinfluenthousingandthepumptopreventpossibledestructionofthecystsbythepump.Theinfluenthosemustbesuspendedinthewater,usuallywitha
flotationcollarconstructedfromstyrofoamtopreventthesuctionofsedimentintothedevice.Ifuseofthe12voltmarinebilgepumpisnecessary,thepumpis
suspendedinthewaterbyastyrofoamflotationcollarandanextremelylightweighthoseattachedbetweenthepumpandthefilterhousing.Oftenananchoris
necessaryinrapidlyflowingwater.MonzingoandHibler(4)havedescribedtheapplicationsofthisequipmentforsamplinginremoteareas.
Thesamplingdevicemustbethoroughlycleansed,preferablywithsoapandhotwatercontainingchlorineatanapproximateconcentrationof100mLsofhousehold
bleach(5.25%sodiumhypochloritebyweight)to20Lofwaterandthismustbedonebeforetheequipmenthasbeenallowedtodry.Sauch(personal
communication)hasobservedthatairdriedcystscanbereconstitutedwithwaterandoftenretainmostofthefeaturesnecessaryforidentification.Ifthesampling
equipmentisassembledasindicatedabove,thehousingandinfluentequipmentonlyneedtobewashedbecausethe1mcartridgeisessentially100%effectivein
removingparticulatesgreaterthan1minsizeorlargerwhenproperlyseated.Ifthetypeofcleansingpreferredisnotpossible,theinfluenthoseandhousingshouldbe
thoroughlyrinsedwiththewaterfromthenextsourcetobesampled.Prudentplanningcanpreventproblemswhensamplingequipmentislimited.

Page199

Forexample,ifonlyonedeviceisavailableandamunicipalfiltrationsystemisbeingsampled,theplanteffluentmustbesampledbeforetheinfluent.
Ifatallpossible,samplesshouldbetakenbeforechlorinehasbeenappliedtothesystem.Ifchlorinatedwatermustbesampled,itshouldbetakenasnearthe
chlorinatoraspossibleand0.5%sodiumthiosulfateintroducedbyafluidproportionerpumpinaccordancewithVolume16ofStandardMethods(1).Analternative
method,althoughnotassatisfactory,istoaddabout100mLof1%sodiumthiosulfatetothecartridge(inthebagwiththecartridge)aftersampling.Unfortunately,
withthelattermethodanycyststrappedonthecartridgewillhavebeenexposedtochlorinatedwaterforseveralhoursandtheGiardiacystsmaybedeadand
unrecognizablebeforeanalysisoccurs.
VolumeofWaterSamplesandRateofFlow
Theamountofsuspendedmaterial,thetypeofmaterial(organic,inorganic),andtheparticlesizeofthatsuspendedmaterialwilldeterminehowmuchwatercouldor
shouldbesampled,andtherearenoclearguidelines.Jakubowski(3)suggestedaminimumof380L(100U.S.gallons).Highlyturbidwater,especiallywaterwith
considerableorganicmaterial,severelylimitstheanalyst'schancesofrecoveringGiardiacystsand,asindicatedearlier,recoveryisinverselyproportionaltothe
amountofthismaterial.SinceGiardiacystsarenotevenlydistributedinwateranddonotmovedownstreamcontinuously,theamountofwatersampledisatrade
off.However,samplesof3800L(1000gallons)areeasilyobtainediftheturbidityislessthan1NTUandtheturbidityisprimarilyinorganicincomposition.Generally
asampleof380to1520L(100to400gallons)isthebestsuggestionthatcanbeofferedatpresentunlessthewatertemperatureislessthan5Candtheturbidityis
lessthan1NTU,then3800Lwouldincreasethechancesofrecoveringcysts.Somesamplerscloselymonitortherateofflowandwhentheratesharplydecreases
theyterminatesampling:thisapproachisacceptable.
SinceGiardiacystsarenotevenlydistributedinwatertherateofflow,andeventhetimeofday,areprobablyascriticalasthevolumesampled.Jakubowski(3)
recommendedarateof1gallon/minuteandthisisanacceptableapproach.Unfortunatelya1gpmrateisnotalwayspossibleduetoothersamplingcommitments,
weatherconditions,etc.Samplestakenat1gpmversusthosetakenat3to4gpmdonotappeartodiffersignificantlyinthenumberofcystsrecovered.Thetimeof
dayisimportant.Animalpopulations,actingasreservoirs,frequentlyaremoreactiveatnightandthereforesamplingispreferablydoneovernightwhenanimalsare
present.Ifthesourceofcontaminationissewage,peakperiodsofsewageeffluentshouldinfluencethesamplingeffort.
Packaging,LabellingandShipment
Filtercartridgesshouldberemovedfromthefilterholder,placedintoaplasticbagandsecuredtopreventleakageand,asanaddedinsurance,doublebagged(zip
loctypebagsareexcellent).Eachinnercartridgebagshouldbeclearlylabelledwithwaterproofink.Theinformationshouldincludethename,addressandtelephone
numberoftheindividualresponsibleforthesample,dateandtimesamplingwasstarted,dateandtimesamplingwascompleted,thenumberofgallonsorlitres
sampled,andthesourceofthewaterandthetypeofsample(rawortreatedwaterandtypeoftreatment).Turbidity,watertemperatureandpHshouldbeincluded.If
thesamplewasfromachlorinatedsource,theamountofchlorineused(mg/Lorppm),thecontacttime(inminutes)beforesamplingoccurredandtheprocedureused
todechlorinatethesample(inlinedechlorinationordechlorinatedpostsampling)shouldbeindicated.Allofthisinformationisimportantfortherecordsofthe
municipalitymoreover,thisinformationfacilitatesanalysisandinterpretationbythelaboratory.
Filtercartridgesmustberefrigeratedwithweticeandshippedinastrong,secure,leakproofcontainer.Samplesmustnotbeexposedtofreezingconditions.Samples
shouldarriveatthelaboratorynolaterthan48hoursaftersamplingiscompleted.Thetypeofcarrieremployedforshipmentisdependentuponthelocationofthe
laboratoryandthesamplingsite.Buslinesarereliableandefficientforsitesnearthelaboratory,andalloftheaircarriersareequallyefficientandreliablewhenthe
distanceisconsiderable.Shipmentshouldbeplannedforsamplestoarriveatthelaboratoryduringaweekday,primarilybecausereceiptofpackagesonweekends,
irrespectiveofthecarrier,oftenisdifficult.Asindicatedearlier,cystsfromthatsourcemaynotsurviveandberecognizableunlessthesampleisprocessed
immediately.
ProcessingtheFilterCartridgeforAnalysis
Extractionofsuspendedmaterialfromthefiltercartridgeisdonebyhand,usingdistilledwater.Extractionofmembranesortheepoxyfiberglasstubeisperformedby
abackwashprocedure.Somelaboratoriesusingtheyarnwound(orlon)filtercartridgesunwindtheyarn,separatetheyarnintosections,andhandwash.This
laborious,timeconsumingproceduresimplycannotbeperformedbybusylaboratoriesprocessing120filtercartridgesdailymoreover,theprocedurerequiresan
inordinateandunnecessaryamountofhandlingwithnogreatercystrecovery.Afastandefficientprocedureistoslicethecartridgetothecorewitharazorknife.
Washingshouldbeginwiththeinnerfibersandallofthefibersshouldbewashedinsixtotwelvesectionsdependingontheamountofdebris.Cartridgefibersections
shouldberepeatedlywashedinfreshdistilledwateruntilthefibersappearclean.Thenumberofwashingsnecessaryisdependentupontheamountandtypeof
suspendedmaterialpresent.Thewashwateristhencombinedforsedimentation.
Thehandwashingprocedureisnecessaryandalllaboratories(exceptthosesamplingwiththeepoxyfiberglasstubeorthemembranes)usethisprocedure.Some
laboratoriesuseawettingagent(0.1%Tween20orTween80indistilledwater)tofacilitateremovalofGiardiacystsfromthefibers(especiallyorlon)althoughthere
isnoproofcystsare"sticky".Irrespectiveofthefiltercartridgeused,orthespecificprotocolfollowedbydifferentlaboratories,thefirst,andpotentiallythegreatest,
lossofcystsoccursduringthiswashingprocedure.AnalysisofthisstepbyJakubowskiandEricksen(2)indicatesthatrecoveryvariesfrom58139%.Inour

Page200

laboratorywehavediscoveredthatrecoveryfromorlon,polypropyleneorcottonfiltercartridgesvariesfrom85109%.Nodifferencescouldbedetectedamong
filtercartridgesordifferentmaterials.However,intheanalysesforbothoftheaboveevaluationstheonlysuspendedmaterialinthefiltercartridgeswasGiardiacysts:
turbiditywasnotafactor.
Afterthewashingprocedurehasbeenperformed,samplescontainingahighorganiccontent,especiallyinthesummermonths,shouldbepreservedwithsufficient
formalintomakea2%(v/v)solution.Thispreventsfurthergrowthofalgaeandprotozoa.Afterpreservation,thesamplescanberefrigeratedovernightat46Cor
concentrationofthewashingscanbeinitiatedimmediatelywithacentrifuge.Allcartridgesgenerallyarewashedinclearglass,widemouthgallonjars(3.78litres)to
facilitatevisualizationofthenextprocedure.Ifthesamplehasbeenrefrigeratedandallowedtosettle,thesupernatantshouldbesiphonedcarefullytothesediment.
Theamountofsupernatanttoberemovedisdependentupontheamountofsediment.Agoodruleistoleaveanamountofsupernatantequivalenttotheamountof
sedimentunlessthesampleisfiltrationplanteffluentcontainingonlyatraceofsediment.Siphoningthenshouldbeterminatedsooner,leavingabout200300mLof
supernatant,otherwisesedimentcouldbeinadvertentlysiphoned.
SelectiveConcentrationoftheGiardiaCysts.TheConsensusorReferenceMethod
AsaresultofthemeetinginCincinnati,Ohio,andatthepublicationofthelatesteditionofStandardMethods(1),therecommendedtechniquewasthezincsulfate
centrifugalflotationtechnique,aprocedurethatwashighlymodifiedandimproveduponbyseverallaboratoriesduringthelagtimebeforepublicationofthetechnique.
Thetechniqueworksverywellifthewaterisnothighlyturbidhowever,aprimarylimitationistheamountofcentrifugatefromsamplealiquots.Ifapproximately0.25
mLofsediment(centrifugate)ina25mLaliquotispresentcystrecoveryisabout85%,providingthecystsareofexcellentmorphologicquality.Cystrecovery
decreasesproportionatelywithanincreaseinsediment(centrifugate).Ifthevolumeofcentrifugateisapproximately0.75mL,cystrecoveryisabout20%andonlyan
occasionalcystcanbefoundwhenthecentrifugateisapproximately1mL.Moreover,vortexing,orothermixingofthesampletomixthecentrifugatewiththeLugol's
iodineandzincsulfateresultsinaninordinateamountofinorganicandorganicmaterialadheringtothemeniscusofthefluidinthetubeandthemajorityofthismaterial
cannotbeeffectivelyremovedfromthemeniscusthereforethecystsfloatingtothetoparethencombinedwiththematerialinthemeniscusofthezinc.Thisinterferes
withvisualizationofthecystandincreasesthefatiguefactor.Ifcystsareofpoormorphologicqualityandthecystmembraneiscompromised,lossesareevengreater.
Regardlessofthechemicalusedforselectiveconcentration,lossesarecomparableallareequallycompromisedbytheamountofsedimentand/orthequalityofthe
cysts.
TheOverlay/UnderlayandtheMembraneFilterSelectiveConcentrationTechniques
Amodificationoftheconsensusmethodthatconsiderablyimprovesrecoveryofcystsandotherlivingmaterialinhighlyturbidwateristheoverlayorunderlay
technique.Fortheoverlaytechnique2025mLofselectiveconcentrationmedia(zincsulfate,surose,percollorpotassiumcitrate)ispoured(orinjected)intoaclear
50mLconicalcentrifugetube(glassornalgene)andanequalamountoftheconcentratefromthefiltercartridgewashingsintroduced(layered)ontothesurfaceofthe
chemicalwithoutdisturbingtheinterfacebetweenchemicalandconcentrate.Anotherapproach,theunderlaytechnique,istopour2025mLofthefilterconcentrate
intothecentrifugetubeandthenintroduce2025mLoftheselectiveconcentrationchemicalbeneaththefilterwashingconcentratewithoutdisturbingtheinterface.No
doubtanumberofapproachescanbeusedtoaccomplishthisstep,butthislaboratoryusesa30mLdisposablesyringetowhichisattacheda16gaugespinalneedle.
TheStylexsyringe(ParmasealAmericalPharmaseal,Valencia,CA91355)ispreferredforasmootherinsertionofthechemical.Althoughwehaveusedbothoverlay
andunderlay,andlaboratorypersonneltrainedbyushavedevelopedpreferencesforoneortheother,weprefertheunderlaytechnique,primarilybecausefilter
washingconcentratecansometimesplugthebarrelofthesyringe,resultinginamixingwiththechemicalconcentratewhenthefilterwashingsaredischarged
moreover,washingandcleaningasyringeandneedlethathasbeenfilledwithchemicalismucheasierandmuchmoreeffectivethanwashingequipmentthathasbeen
filledwithfilterwashings.
Aftertheoverlay/underlaystep,centrifugationmustbeginimmediately.Anydelaywillallowthematerialsuspendedinthefilterwashingtosettleattheinterface
betweenthechemical/filterwashing,resultinginaconsiderablenumberofGiardiacystsbeingpulledthroughtheinterface.Thetubesshouldbecentrifugedfor5to8
minutesinacentrifugethatwillprovideabout380g.Asindicatedpreviously,ifthefilterwashingscontainaminimumofsediment,thereferencemethodisbetterthan
theoverlay/underlaytechnique.Thepurposeoftheoverlay/underlaytechniqueistotrapmaterialsuspendedinthefilterwashinghavingaspecificgravityoflessthan
thechemicalattheinterfacebetweenthefilterwashingsandthechemical.ThisincludesGiardiacysts,otheranimalparasites,algae,diatoms,protozoa,plantdebris,
crustaceans,freelivingnematodes,etc.
Aftercentrifugation,thefilterwashingsupernatanttotheinterface,andthesurfaceoftheinterfaceissiphonedthrougha5mNucleporemembrane.Weusea47mm
membranesecuredinaMilliporechamber.SlightvacuumisappliedwithafaucetsiphonattachedtotheMilliporecup.Severalapproachescanbeusedtorecover
materialatthesurface.Thefirstistouseasyringeandneedle,beginningatthemeniscusandremovingmaterialtotheinterfaceandabout5mmbelowtheinterface
(intothechemical)oranotherapproachistoinserttheneedleattheinterfaceandwithdrawtheliquid.Witheithertechniquetheliquidleveldecreasesfromthe
meniscustotheinterface.Theinnerperimeterofthetubeshouldbecircledasthefluidiswithdrawn.Inthislaboratoryathird

Page201

approachisused.Webreaktheinterfacewithanapplicatorstickandthenquicklypourthewashingsandchemicalsolutionthroughthemembrane.
Aftersiphoningthroughthemembrane,themembraneisremovedwithforcepsandheldagainsttheinnerwallofabeakertobewashedwithafine,strongstreamof
distilledwaterfromasquirtbottle.Thesuspendedmaterialinthebeakerispouredintoa50mLcentrifugetube,thebeakerrinsedandtherinseaddedtothetube.
Thistube(firstwashing)isthencentrifugedat380gfor3to5minutesandtheliquidsiphonedtothe610mLlevel.Thisfirstwashingmaterialismixedwitha
vortexerandtransferred,withrinsing,toa15mLclearnalgenetube(preferablywithaliparoundthetop)forasecondwashingandcentrifugationat380gfor3to5
minutes.Followingthiscentrifugation,thewaterissiphonedtoabovethepelletandtwodropsofLugol'siodineadded.ThereareseveralformulasinuseforLugol's
iodine.Weuse10gpotassiumiodideand5giodineto100mLwater.TwotothreemLofchemicalsolution(zincsulfate,percoll,potassiumcitrate,sucrose)is
addedimmediatelyandthetubevortexedtosuspendthematerial.Chemicalsolutionisthenaddedtofillthetubeuntilthemeniscusbulgesveryslightly.Acleanglass
coverslipisaddedandthetubecentrifugedfor3to5minutesat380g.Thepurposeinusingatubewithalipistoinsureabettersealbetweentubeandcoverslip,
preventinglossofthecoverslipduringcentrifugationatthislast,criticalstep.Aftercentrifugationthecoverslipisremoved,placedonaglassslideandthematerial
adheringtothecoverslipexaminedsystematicallywithamicroscopeat100%.
Thefirstlossofcystsoccursintheprocessingofthefiltercartridge,andthesecondlossoccursintheoverlay/underlaytechnique,irrespectiveofthechemical
employed.Thethirdlossoccursinthefinalstep.Notallofthecystswilladheretothecoverslip,andwhenliftedoffthecentrifugetubesomeremainbehindonthe
meniscus.Theamountandtypeofsuspendedmaterialandthequalityofthecystswillcausecystlosses.Whencystsaresuspendedinwater,orinasmallamountof
inorganicororganicmaterial,recoveryapproaches9095%however,aswiththeconsensusmethod,largeamountsofsuspendedmaterialresultinconsiderableloss
ofcysts.Averyeffectiveproceduretomitigatetheselossesistodilutetheconcentratetoatleasta1:2ratioandprocessmorereplicatesamplesthroughthe
membrane.Thisdoesnotnecessitatemicroscopicexaminationofaninordinatenumberofreplicatestheycanbecombinedandtrappedononemembrane.
Eachstepinthisproceduremustbeperformedcarefullyandthoroughly.Thetoplipofthe15mLcentrifugetubemustbeperfectlyflattoformaneffectiveseal.Glass
coverslipsoftenaregreasyiftheyaregreasytheglassmustbethoroughlydegreased,rinsedwithdistilledwateranddried.TomTrok(personalcommunication)
preparesaneffectivecleanserbyaddingAjaxorasimilarcleansertodistilledwater,centrifugingthesuspensionandusingthesupernatantasacleanser.Squirting
alcoholononeortwocoverslipsfromeachnewboxwillquicklydetermineifgreaseispresent.
Theconsensusmethod(StandardMethods,1985)offerstwoalternatives,addingthecoversliptothetubebeforecentrifugation,oraddingthecoverslipafterwards
(touchingthemeniscuswiththecoverslip).Thelatterprocedureisnotaseffective.SomeGiardiacystsbegintosettleintothechemicalimmediatelyafter
centrifugation.
Twomajorconsiderationsforthesetechniquesarethequalityofthecentrifugeandthemicroscope.Thecentrifugeshouldhaveanaccuratetimerandrpmmeter
moreover,itshouldcoastslowlytoastop,notsuddenly.WefindthatacentrifugewhichstopstooquicklydisturbstheGiardiacystsonthecoversliporatthe
meniscus.Themicroscopemustpossessexcellentqualitylenses,excellentobjectives,andastrong,bright,wellbalancedlightsource.Whilethequalityofthe
microscopedoesnotinterferewithrecoveryofcysts,itcancertainlylimiteffectivevisualizationandquicklyinitiatemicroscopefatigue.
SelectiveConcentrationMedia
SeveralchemicalshavebeenusedtoselectivelyconcentratewatersamplespossiblycontainingcystsofGiardia.MostexperiencedparasitologistspreferZnSO4,but
investigatorsindifferentlaboratoriesusesucrose,percoll,ficollhypaque,orpotassiumcitrate.Allhavegoodandbadpoints.Thechemicalpreferredbydiagnosticians
inthedifferentlaboratoriesisgenerallythechemicalwithwhichthatindividualhasthemostexperience,andwhenusedbythatindividualwillprobablyprovideresults
comparabletoachemicalpreferredbyanindividualinanotherlaboratory.
Zincsulfate,potassiumcitrateandsucroseareallextremelyhygroscopicandwillshrinkthecystshowever,thisdoesnotinterferewithdiagnosis.Percollandficoll
hypaquedonotshrinkthecysts,butthesechemicalsarecostprohibitiveforroutineanalysis.Thislaboratory'sexperiencewiththesevariouschemicalsevaluatedata
s.g.of1.13indicatesthatpercolland/orficollhypaquearethebestchemicalsforrecovery,andareabout10%betterthanZnSO4,whichisabout10%betterthan
sucrosepotassiumcitratewastheworstofthechemicalsexamined.However,zincsulfate,potassiumcitrate,andsucrosewhenusedatspecificgravitiesof1.2to1.3
werebetterthanpercollofficollhypaque.
SpecificGravityoftheChemical
InvestigatorsmustbepreparedtousechemicalswithdifferentspecificgravitiestoselectivelyremoveGiardiacystsandotherlivingmaterialfromthefilterwashings.A
hydrometermustbeusedtodeterminethespecificgravity,becausepreparationofanexactsolutioncannotbedoneeffectivelybyweight/volume.WeuseZnSO4at
specificgravitiesof1.1,1.2and1.3.Somelaboratoriesuse40%potassiumcitrate(about1.3sp.g.).Weprefertounderlaywithachemicalof1.3s.g.because
comparisonofresultswith1.1,1.2or1.3ZnSO4indicatesthehigherdensitywillresultinmorecystsrecovered(theshrinkageisalsomorepronouncedbutstillnotan
interference).Unfortunatelyhighorganiccontent,alumorpolymersinthefilterconcentrateirrespectiveoftheamountofdilution,oftennessitatesuseofchemicalsat
1.2s.g.oreven1.1s.g.otherwisematerialfromthefilterconcentratepacksatthe

Page202

interfaceofconcentrate/chemical.Asthespecificgravityofthechemicalisdecreasedtoselectivelyremovethelowdensitymaterialpresentintheconcentrate,cyst
losses(trappedinthematerialandpulledthroughtheinterface)increaseproportionately.Ifaconsiderableamountofalumorpolymersispresentintheconcentrate,
selectiveremovalofcystsisalmostimpossible."Noconfidenceintheresults"isreportedwhenorganicturbidityorthesechemicalcoagulantsinterferewithcyst
detection.Analystsmayuseoneortwoapproachesinanattempttoselectivelyremovecystsfromthematerialcontainingexcessivesediment:1)dilutetheconcentrate
andoverlay/underlaymoretubesand/or2)useachemicalwithlowerspecificgravity.Usingachemicalwithalowerspecificgravityshouldbethelastchoicebecause
ofthelossofefficiency.
MicroscopicExamination
Iftheglasscoverslipmethodisemployedpriortocentrifugationinthelaststep,itshouldbeorientedwithonecornerofthecoverslipinthedirectionofthecentrifugal
forceand,whenremoved,orientedontheglassslidewithamarkontheslideindicatingthisquadrant.Examinationshouldbegininthisquadrantbecauseexperience
hasshownthatduetothecentrifugalactionandthespecificgravityofthecystsversusthechemical,themajorityofcystswillbepresentinthisquadrant.Before
fatiquebecomesaproblemthisquadrantshouldbethoroughlyexamined,thentheremainderofthecoverslipexamined,fieldbyfield.Duringtheexaminationthe
miroscopistshouldnotetherelativeamountsofmaterial(inorganicandorganicdebris)andorganismspresenttoassesstheriskfactor,especiallywhencomparingfilter
plantinfluent(rawwater)andfilterplanteffluent(finishedwater).WhenexamininginfluentandeffluenttoassesstheefficiencyofaplanttoremoveGiardiacysts,the
presenceorabsenceofcystsintherawwaterisirrelevanttheabilityofthesystemtoremoveparticulatesthesizeoforlargerthanGiardiacystsisextremely
important,andtherelativepercentagesremovedareanassessmentofrisk.Manyplantsremovemostofthesmallparticulates(clay,etc.)butallowpassageofplant
debris,anicenamefortheundigestedfecaldebrisfromherbivorousrodents(muskratandbeaver).Plantdebrisisverypliableandlightinweightandisthebest
indicatoroffilterplantefficiency.Ifbeaverand/ormuskratarepresentonthesourcethengenerallyGiardiacystswillbefoundintherawwater,ifnotinthatsample
certainlyinasubsequentsample.Iftheanimalsarenotsheddingcystsorcystsarenotfound,theplantdebriswillbepresentandthefilterplant'sabilitytoremovethis
materialisimportant.
ModificationstotheAnalysisTechniques
Mostinvestigatorsnowusetheoverlay/underlaytechniquewithoneoftheabovechemicalstoseparatethecystsandotherlivingmaterialfromtheinorganic
particulates.Some,asinourlaboratory,attachthecoverslipbeforecentrifugation,some"touch"thecoversliptothemeniscusofthetubeaftercentrifugation,and
othersusethebacteriologiclooptosecureasampleofthemeniscus.Withthesetechniquesarelatively"clean"sample(about8001000gallonsofwaterat0.5to1.0
NTU)requiresabout2hoursofeffortforanalysis,whilea"dirty''samplemayrequireabout4hours.Otherinvestigatorshavechosentowashthematerialfromthe
membraneandthenconcentratebycentrifugation,ultimatelyobtainingabout0.5to1mLofmaterialtobeexamineddropbydrop,eitherafterstainingwithLugol's
iodineorwiththeaidofaphasecontrastmicroscope.Thisprocedurerequires824hoursofmicroscopetime,dependingontheamountofmaterialtrappedatthe
interfacebetweenconcentrateandchemical.Eyefatigueisaveryrealproblemandeffectivelycompromisestheresults.Busylaboratoriescannotaffordthetimefor
thistechniqueandmunicipalitiescannotaffordthecost.
Immunofluorescence
Riggs(5)developedadirectfluorescentantibody(FA)techniqueandSauch(6)developedanindirectfluorescentantibody(IFA)techniqueforthedetectionof
Giardiacystsinwater.Recently,Riggsdevelopedamonoclonalantibodyfromhumansourcecystsforthispurpose.Sauchselectivelycleanstheconcentratewith
percollwhichisthebestchemicalforthispurpose,Riggsuses40%potassiumcitrate,andthislaboratoryuseszincsulfatebecauseitiscosteffectiveforroutine
monitoring.Afterselectiveconcentrationwiththechemical,thematerialtrappedattheinterfacebetweenthetwochemicalsisevaluatedmicroscopicallyandthe
volumeofconcentrateadjustedtoprovideamonolayerofparticulateswhen1mLisappliedtoacellulosetriacetate25mmdiameterfilter(GelmanMetrical).Sauch
usesa0.2mabsoluteporosity,whileweusea5mabsoluteporosity.Dilutedantiserumisappliedtoeachmembranefor15minutes,rinsed5timeswithphosphate
bufferedsaline(PBS)anddilutedconjugage(goatantirabbitIgGconjugatedtoFITC,MilesScientific,Naperville,IL)isappliedfor15minutesandrinsedfivetimes
withPBS.SauchusesPBSsupplementedwith2%bovineserumalbuminand0.05%polyoxyethylenesorbitanmonolaurate(Tween20)whileweuseonlythePBS.
SauchstainsthematerialwithEvansblue(0.003%Evansbluein0.15MKCL,SigmaChemicalCompany)for10minutes,dehydrateswithanaqueousethanolseries
(v/v)containing5%(v/v)glycerol(10,20,40and80%ethanol),thencleansthemembranewithglycerolcontaining0.12%(w/v)propylgallate(SigmaChemical
Company),coversthemembranewitha25mmcoverslipandsealswithclearfingernailpolish.Generallyweforegothestaininganddehydrationstepsandonlyclear
themembrane,coverwithacoverslipandreaddirectly.Sauchpreferstoclearthemembranebecauseshedoublechecksherfindingsbyswitchingfromfluorescence
tophasecontrastmicroscopytoverifytheresults.TheproceduresdevelopedbyRiggsandSauchiffollowedclosely,willworkverywellprovidingtheanalysthasa
goodantibodyproperlytitratedformaximumcystfluorescenceandminimalbackgroundnonspecificfluorescence.
ProblemswithDirectMicroscopyandFluorescentMicroscopy
Aspreviouslymentioned,thegreatestlossesofcystsforanyofthesetechniquesoccursinthefirsttwosteps:washingthefibers(orbackflushingtheepoxyfiberglass
cartridges)andintheselectiveconcentrationstep.Lossesaregoingtooccurandatthecurrentstate

Page203

oftheartthereislittlethatcanbedoneexceptattempttominimizetheselosses.
Aproblemwithzincsulfate,sucrose,orpotassiumcitrateisthedehydrationfactor.Thecystsaresmaller(shrunken),necessitatingconsiderableexperiencefor
recognition.Moreover,ifthecystwalliscompromisedtheorganismmayfillwithchemicalandfailtofloatproperlyinthelastflotationstepbecauseitthenhasthe
samespecificgravityasthechemical.Lossesmayoccurbecauseofgreasycoverslips,etc.Oneparticularlydifficultproblemisthatmanydifferentorganismswilltake
theiodinestain,allappearingbrownincolor.Whilelive,freshGiardiacystsarehighlyrefractileandbrownincolor,olderordeadcystsmaybelightbrownoreven
greenincolor.This,togetherwiththedehydration,oftennecessitatescarefulexaminationat1000xtolocatetheinternalstructuresnecessarytoconfirmthatthe
organismisaGiardiacyst.Healthy,wellbalancedsourcesofwatermaybeheavilyladenwithsmallflagellatesand/oralgaesimilarinsizeandstainingpropertiesof
Giardiacyststhiswillcontributequicklytofatigueandafailuretofindorseethefewcystspresent.
Individualsplanningtousethedirectmicroscopyprocedure,irrespectiveofwhethertheyusedropbydropdirectexaminationoraflotationprocedureshouldbe
welltrainedinprotozoology,streambiologyandfreshwateralgae.
Oneofthemainproblemswithimmunofluorescenceisthatitisonlyavailablefromonecommercialsource.Thisantibodyhasnotbeenextensivelyevaluated.Another
problemisthecost:goodqualitymicroscopeswithbothphasecontrastandepifluorescentcapabilitiesareexpensive.Someinvestigatorshavesuggested
immunofluorescencemaybeabetteroptionfortheinexperiencedmicroscopistthanthedirectmicroscopytechniquehowever,othersexperiencedinbothtechniques
feelthatfluorescenceofotherorganismsmayresultinfalsepositivediagnoses.Thereisnoadvantagetoimmunofluorescenceforexperienced,welltrained
microscopists.Immunofluorescenceismoretimeconsumingthandirectmicroscopyandfatiguebecomesafactormorequicklythereforebusylaboratoriesexamining
120samples/daycannotaffordthetime.Bothtechniquesaremuchquickerthanthedropbydroptechniqueandaremoreeffectivebecauseoftheevengreater
fatiguefactorassociatedwiththelatter.
Adistinctadvantageofthedirectmicroscopetechniqueistheabilitytoevaluatefiltrationplantperformance.Themajorityofoursamplescomefrommunicipalitieswith
filtrationsystemsandweareaskedtoassessefficiencybasedonremovalofriskmaterial,organismsinthewaterthesizeoforlargerthanGiardia.Asindicated
earlierthepresenceorabsenceofGiardiaisofinterest,buttheperformanceoftheplantismoreimportant.
AproblemformicroscopistsusingeithertechniqueisthedetectionandidentificationofGiardiacystsfromsourcesthatprobablyarenotinfectiousforhumans.We
findthat100%oftheblackcrownednightheronsareinfectedwithGiardiaandshedacystthatisverysimilartotheGiardiaduodenalistypefoundinmammals.
Nodoubtsomeoftheotherwaterfowland/orshorebirdsareinfectedwithGiardia.Wehavefoundthebirdtypeofcystinwaterfromreservoirsinboththeeastern
andwesternUnitedStates.Conversationwithmunicipalauthoritieshasrevealedthatwaterfowlwerepresentonthereservoiratthetimeofsampling.Thiscyst
fluorescesextremelywellwiththepolyclonalantibodyweuse,butmaynotfluorescewithamonoclonalantibody.Currentlytheriskofmisidentificationisveryreal,
irrespectiveofthetechnique,andcouldresultinanunnecessary"boilwater"orderforamunicipality.
QualityAssurance
Ifmicroscopistsdonotregularlyexaminesourcesofwatercontaminatedwithcysts,theyshould"seed"waterwithcyststocheckontheirabilitytovisualizethe
organisms.Seedingsampleswithcystsoftenisfrustrating.Somebatches(fromthesamesource)willbeexcellentandcystscanbeeffectivelyrecoveredfordaysor
weeksinotherbatchescystscanberecoveredforonlyafewdaysbecausethecystnumbershavedecreasedorthecystqualitydeteriorates.Weexamine120
filters/dayand2040%arecontaminatedwithcystshowever,wecontinuetousequalityassurancesamplesasacontrolonourtechnique.Weregularlysupply
qualityassurancesamples,actualconcentratesfromwatersamplescontaminatedwithcysts,topersonstrainedinourlaboratoryandthisprocedureworksreasonable
wellhowever,asindicatedearlier,thenumberofcystspresentinasampleonthedayexaminedmayormaynotbethesameaswhatcanberecoveredthenextday.
Generally,fewercystswillbepresent.Thisnecessitatessendingsamplescontainingahigherconcentrationofcyststoinsurethattheindividualwillfindatleastafew.
Ananimalmodelisusedasasourceforcysts.Thisistimeconsumingandcostlymoreover,theU.S.DepartmentofAgriculturehasestablishedextremelystrict
guidelinesforuseoflaboratoryanimals.Mostsmalllaboratoriescouldnotcomplywiththestandardwithoutcostlyremodelingandtrainingofpersonnel.
Conclusions
DiagnosisofwaterborneGiardiaisadifficultandtimeconsumingtaskofselectivelyconcentratingandfindingafewcystsintheconcentrateof100to1000gallonsof
water.Nospecifictechniquecurrentlyavailableisanybetterorworsethantheothertechniquesavailable:theyareallrelativelyinefficient.Wefindcystsin
approximately20%ofthesamplesexaminedlikelycloserto30or40%ofthesesampleshavecystspresentbutduetothereasonsdiscussedtheyarenotrecovered
orvisualizedifrecovered.Repeatedsamplingofanegativesourceusuallyprovidespositiveresults.Wehaveameansbywhichwecandetermineifthesecystsare
alivebutnotiftheyaredead(thegerbilsmaynotbesusceptibletothecyststraininthesample).IfthecystisoftheGiardiaduodenalistypewemustassume
potentialinfectivityforhumansandinitiatethebarriersnecessarytopreventinfection.Wemustassumethatallsurfacewatersourcesofwaterareeithercontaminated
withorwillbecontaminatedwiththecystsofGiardia.

Page204

LiteratureCited
1.A.P.H.A.,A.W.W.A.,W.P.C.F.1985.Standardmethodsfortheexaminationofwaterandwastewater.16thedition,A.P.H.A.Washington,D.C.pp.1268.
2.Jakubowski,W.andT.H.Ericksen.1979.MethodsfordetectionofGiardiacystsinwatersupplies.In:WaterborneTransmissionofGiardiasis.W.
JakubowskiandJ.C.Hoff(eds).USEPA,OfficeofResearchandDevelopment,EnvironmentalResearchCenter.Nat.Tech.Info.Service,Springfield,VA.EPA.
3.Jakubowski,W.1985.DetectionofGiardiacystsindrinkingwater:stateoftheart.In:GiardiaandGiardiasis.S.L.ErlandsenandE.A.Meyer(eds).Plenum
Press,NewYork,NY,pp.263285.
4.Monzingo,D.L.andC.P.Hibler.1987.TheprevalenceofGiardiainabeavercolonyandtheresultingenvironmentalcontamination.J.Wildl.Dis.Vol.23.pp.
576585.
5.Riggs,J.L.,Nakamura,K.andJ.Crook.1984.IdentifyingGiardialambliabyimmunofluorescence.In:Proc.ofthe1984SpecialtyConf.,Environmental
Engineering.M.PirbazariandJ.S.Devinny(eds).
6.Sauch,J.F.1985.UseofimmunofluorescenceandphasecontrastmicroscopyfordetectionandidentificationofGiardiacystsinwatersamples.Appl.andEnviron.
Microbiol.50(6):14341438.
7.Schupp,D.C.andS.L.Erlandsen.1987.AnewmethodtodetermineGiardiacystviability:correlationoffluoresceindiacetateandpropidiumiodidestainingwith
animalinfectivity.Appl.Environ.Microbiol.53:704707.
8.Schupp,D.C.,Januschka,M.M.andS.L.Erlandsen.1987.AssessmentofGiardiacystviabilitywithfluorogenicdyes:comparisonstoanimalinfectivityandcyst
morphologybylightandelectronmicroscopy.AdvancesinGiardiaResearch.UniversityofCalgaryPress.

Page205

MethodsfortheRecoveryofGiardiaandCryptosporidiumfromEnvironmentalWatersandtheirComparativeOccurrence
JoanB.Rose*,DimaKayed,MaryS.Madore,CharlesP.Gerba,MichaelJ.Arrowood,CharlesR.SterlingandJohnL.Riggs
TheUniversityofArizona,CollegeofAgriculture,DepartmentofNutritionandFoodScience,309ShantzBuilding,Tucson,Arizona85721,U.S.A..
CryptosporidiumhasrecentlybeenassociatedwithawaterbornediseaseoutbreakintheUnitedStates.Informationontheoccurrenceofthisparasiteindomesticsewageand
water,however,isalmostnonexistant.Recently,ourgroupdevelopedmethodsforthedetectionofCryptosporidiuminlargevolumesofsewageandwater.Thisstudyappliedthese
methodstodeterminetherelativeoccurrenceofbothGiardiaandCryptosporidiuminsurfacewatersandsewageeffluentsinthewesternUnitedStates.Bothparasiteswere
concentratedfromwaterusingpolypropylenespunfiberfilters,andidentifiedwiththeaidofmonoclonalantibodies.Giardiacystlevelsinrawsewageaveraged51/L.Giardiawas
onlyoccasionallyobservedintreatedsewageeffluents(1.3cysts/L)andcontaminatedsurfacewaters(0.351.2cysts/L),whileCryptosporidiumwasobservedinalmostallsamples
examined.Averageconcentrationsofoocystsinrawsewagewere5,200/Land1,400/Lintreatedsewageeffluents.Insurfacewater,oocystconcentrationsaveragedfrom0.08to
28.5/Lfrompollutedtopristinewatersrespectively.Insummary,watersourcesfreeofGiardiacannotbeassumedtobefreeofCryptosporidium.

Introduction
Theabilitytorecoverpathogenicmicroorganismsfromenvironmentalsamplescanaidthemicrobiologist,epidemiologistandwaterengineerindefiningwaterborne
outbreaksofdisease(11).Becausebacterialindicatorsystemsusedtojudgewaterqualitydonotaccuratelypredictthepresenceorabsenceofsomespecific
pathogens,suchasGiardia,itmaybecomeparticularlyimportanttohavemethodswhichcandetecttheseorganisms(12).
RoutinesamplingproceduresusedforbacteriaarenotapplicabletoGiardia.Specializedtechniquesarerequiredforthesampling,recovery,anddetectionof
Giardiacystsfromwater.Earlyeffortsemployedtheuseofmembranefiltrationforcollectionofcystsfromsewagewithlimitedsuccess(13).Thisprocedure,
however,wasnotapplicabletoothertypesofwater(11).Apoolsandfilterwasalsousedtocollectlargervolumesofwaterandcystswererecoveredbyfilter
backflushing,andalumcoagulation(19).Forroutineuse,however,thismethodprovedtoocomplicatedandlaborious.Theuseoftheyarnwoundcartridgefilterfor
cystconcentrationhasnowproventobeausablesystem(10).Largevolumesofwatercanbeprocessed,yet,overallrecoveryefficiencieswerestillpoor(6.3%).
AlthoughmuchattentionhasbeengiventothedevelopmentandmodificationofmethodsfortheisolationofGiardiacystsfromenvironmentalsamples(11),
improvementsincystdetectionbymicroscopicmethodshavelagged.Undermostcircumstances,theaccurateidentificationoftheGiardiacystsrequiresan
experiencedindividual.Recognizingcystsinenvironmentalsamples,however,mayprovedifficulttoeventheexperiencedparasitologistduetopresenceofartifacts,
algaecells,orotherdebrisresemblingGiardiacysts.Cystquantificationmayalsobedifficult.Spauldingetal.(20)foundthatmembranefiltrationwasanaccurate
meansfordeterminingcystconcentrations.Sauch(18)usedimmunofluorescenceinconjunctionwithmembranefiltrationtodetectGiardiafromenvironmental
samples.Thiscombinationoftechniqueshasprovenreliableforcystdetectionandquantitation.
Giardiaisolationmethodshavebeenadoptedforuseintherecoveryofanotherentericprotozoan,Cryptosporidium(14,15).Polypropylenewoundfilterswere
usedtocollectoocystsandasucrosegradientcentrifugationmethodwasusedtopurifythem.Directimmunofluorescencewasusedtoidentifyandquantitateoocysts.
Althoughinitialoocystrecoverieswerelow(14.5%),improvementshaveincreasedrecoveriestoa59%average(16).Asimilarsystemhasbeenusedtodetect
Giardiainwatersamplesusingapotassiumcitratemediumduringthepurificationstep(17).
ItisnowpossibletoexaminewatersforthepresenceofentericprotozoasuchasCryptosporidiumandGiardia.Improvementsandthoroughevaluationsofthe
methodsareneeded,however,beforethesetechniquescanhavewidespreadapplicability.
MaterialsandMethods
SampleCollection
Watersampleswerecollectedusingaportablegasolinedrivenwaterpumpandfilteredthroughteninchspunpolypropylenecartridgefilters(MicroWyndII,
*Correspondingauthor.

Page206

AMF/CUNODivision,Meriden,CT)withanominalporosityof1m.Flowrateswereadjustedtofourtofivegallonsperminuteand100to400gallonsofwater
werefiltered.
FilterElution
Filterswereinitiallyprocessedbybackflushingthefilterwith2700mLofeluent[deionized(DI)watercontaining0.1%Tween80].Thefilterwascutlongitudinally,
separatedfromthecore,teasedapart,andwashedthreetimeseachtimeinonethird(900mL)oftheeluent.Thewashingwasdoneonanautomaticshakerfor10
minutesinaonegalloncontainer.Thesamplewasconcentratedandcombinedintoasinglepelletbycentrifugation(1200gfor10minutes).Thefinalpelletwas
dividedinhalfandresuspendedin10%formalinor2.5%potassiumdichromate.
PelletProcessing
Pelletssuspendedinformalinwerewashedandresuspendedwithadetergentsolution[DIwaterwith1%Tween80and1%sodiumdodecylsulfate(SDS)]andthen
homogenized.Onedropofantifoamwasaddedtofacilitatetotalsamplerecovery.Next,thesamplewaswashedandresuspendedwiththedetergentsolution,orDI
water,andfinallysonicatedimmediatelypriortothelayeringontoaclarificationgradient.
CentrifugationonSheather's(500gsucrose,320mLDIwater,and9.7mLliquidphenol,1.29g/mL)wasusedtoseparateoocystsfromsediments.FiveortenmL
ofsamplewaslayeredonto10or30mLofa3/5or4/5dilutionofSheather'ssolution(1.17g/mLand1.24g/mL,respectively).Thetubeswerecentrifugedat1200
gfor10minutesandthesupernatantswererecovered.Potassiumcitrate(40%solution,1.16g/mL)wasusedtoseparateGiardiacystsfromsediments.The
samplessuspendedinDIwaterwereusedforthiscentrifugation.A1:3ratioofsampletomedia(10mLto30mL)wasused.Thetubeswerecentrifugedat800g
fortwominutesandsupernatantswererecovered.
OocystandCystDetection
Recoveredsupernatantswerewashed.Sampleswerefiltered,eitherdirectlyorafter1:10dilutionsthrough13mmcellulosenitratemembranefilters(poresize,1.2
m,5.0mforoocystsandcysts,respectively).Monoclonalantibodies,directlyconjugatedtofluoresceinisothiocyanate(FITC),wereusedtodetectbothGiardia
(2)andCryptosporidium(21).AnindirectimmunofluorescenttechniquewasalsousedforCryptosporidiumidentification(9).TheGiardiafilterwascounterstained
with0.003%Evans'Blue.Thefilterswererinsedandmountedonglassslidesinaglycerolsolution.Theorganismswereenumeratedonthefilterusingepifluorescent
microscopy.Numberoforganismsperliterofwaterwasthencalculated.
EfficiencyStudies
TheefficiencyoftheseparationprocedureswasevaluatedforbothCryptosporidiumandGiardia.OnetofivemLofsedimentedpelletfromwaterwereseededwith
Cryptosporidiumoocystspreviouslypurifiedonpercollgradients(3).Afterconcentrationsweredetermined,eachwassubjectedtothefullstrengthsucrose
centrifugationandvariousdilutions,aspreviouslydescribed.
AstoolsamplecontainingGiardiacystswassuspendedinthedetergentsolution,andaftercountsweremade,sixcentrifugationsolutions,(fullstrengthSheather's,4/5
and3/5dilutionsofSheather's,zincsulfate,40%potassiumcitrate,andpercollsucrose(18)wereevaluated.Anumberofvariablesweretestedincludingsonication
anddetergents.Thetotalsupernatantwascollectedandrecoverieswerecalculated.
TheefficiencyofCryptosporidiumoocystrecoveryfromtapwaterandsewagewasevaluated.Onehundredgallonsofdechlorinatedtapwaterorsecondaryeffluent
wereseededwithknownlevelsofCryptosporidiumoocysts.Thiswaterwasfilteredandthesampleprocessedaspreviouslydescribed.Theinitialnumberofoocysts
wasdeterminedbyadditionofoocystsfromastocksuspension(indeionizedwaterwith1%Tween80)to50100mLofthetapwater.Afterasamplewastakenfor
assay,thetapwatersuspensionwasthenaddedtothe100gallonvolume.
ResultsandDiscussion
Toimproveprotozoarecoveryfromwatersamples,theentireprocedurewasevaluatedatfourdistinctsteps.Theseinclude:1)samplecollection2)filterelution3)
samplereconcentrationandclarificationand4)parasitedetection.Eachstepwasoptimizedtoenhancetheoverallmethodefficiency.
Duringthecollectionstep,twofilterswereevaluatedfortheirabilitytorecoverCryptosporidiumoocystsfromsecondaryeffluent.Bothwereteninchcartridge,
polypropylene,yarnwoundfilterswitha1mnominalporosity.Filter1wasmadeasacontinuousspiralofasinglestrand,wovenbackandforth.Incontrast,the
polypropyleneforFilter2wasappliedinablanketform.Filter2wassuperiorwithanaverageoocystrecoveryof36.6%ascomparedtoa3%recoverywithFilter1.
ThesuperiorperformanceofFilter2maybeduetoincreasedoocystsentrapmentorduetoincreasedrecoveryofoocystsfromthefilterduringbackflushingand
washing.ThelatterissuspectedsinceFilter2washedcleanerduringelutionandalmosttwicethevolumeofpelletwasrecovered.
Thefilterelutionstepascomparedtotheotherproceduresisthemostcumbersomeandtimeconsumingandpossiblymaybewherethegreatestlossinoocystorcyst
recoveryoccurs.Previouslyithadbeenshownthatbackflushingofthefilterenhancedelution(14).Duringseededtapwaterstudies,itwasfoundthatinitial
backflushingrecovered16%oftheoocysts.Furtherprocessingofthefilterwas,therefore,necessary.Aftercuttingthefilterapart,aninitialwashingrecovered20%
moreoocysts.Secondandthirdwashingsachieveda58%recovery(datanotshown).Additionalwashingswithgreatervolumesdidnotappreciablyincreasethe
efficiencyofthemethod.Currently,threereplicatewashingsofthefilter,afterbackflushing,withvolumesofapproximately900mLeacharerecommendedforoptimal
elution.Thetotaleluentvolumeof2700mLcanbeconcentratedtoasinglepelletinlargevolumecapacitycentrifuge,makingsampleprocessingmoreconvenient.
Thethirdstepintheprocedure,clarificationofthesample,isnecessarytoremoveinterferingdebriswithoutconcomitantlossoftheorganism.Musialetal.(14)found
thatCryptosporidiumoocystrecoveriesusingaSheather'smediaweredramaticallyimprovedbytheadditionofTween80andSDStothesample.Itwasspeculated
thatthedetergentsmayactbydisruptinghydrophobicandelectrostaticinteractionsbetweenoocystsandsediment.AlthoughSheather'sgradientswithdetergents
wereefficient(82%recoveries),detectionoflowoocystnumberswasdifficultsincesampleswerenevertotallyclearedofdebris.DilutionoffullstrengthSheather's
solutionby4/5,3/5,2/5,and1/5(1.24g/mL,1.17g/mL,1.11g/mLand1.06g/mL)decreasedrecoveriesto72,76,67and18%,respectively(Table1).The
samplewassufficientlycleared,however,usingthe4/5and3/5dilutionstodetectaslowas0.06oocysts/L.Inaddition,an

Page207
TABLE1.RecoveriesofCryptosporidiumOocystswithSheather's.
Sheather's

Specific
Gravity(g/mL)

Numberof
Trials

Average%
Recovery

Fullstrength1

1.29

82

4/5*

1.24

72

3/5*

1.17

76

2/5*

1.11

67

1/5*

1.06

18

1500gmsucrose/320mLdeionizedwater,9.7mLphenol
*Dilutionsoffullstrength

equivalentof378Lcouldbefilteredthroughasinglemembranefilter.
TheclarificationstepwasalsoevaluatedforGiardiarecovery.Sixgradientsolutionsweretested(Table2)andtheentiresupernatantwascollectedandexaminedin
eachcase.Recoveriesaveraged76,77,70,68,66and40%forpotassiumcitrate,percollsucrose,Sheather's,4/5Sheather's,3/5Sheather's,andzincsulfate,
respectively.UsingANOVA,nostatisticalsignificantdifferencewasobservedbetweenthevariousmedia.Potassiumcitrateand4/5Sheather'sresultedincleaner
preparations,however,whenusedforenvironmentalsamples.ThesetwogradientsolutionswerealsoevaluatedforGiardiarecoverywithouttheuseofdetergentsor
sonication.Asignificantdecreaseinrecoveryatthe95%confidencelimitwasfoundforbothpotassiumcitrateand4/5Sheather'swhenusedwithoutdetergents,41
and35.6%ascomparedto76and68%withdetergents,respectively(Table3).Sonicationwasnotfoundtostatisticallyincreaserecoveries.
Thefinalstepintheprocedureisthedetectionandenumerationofoocystsandcysts.Theuseofmonoclonalantibodieshasincreasedthesensitivityofparasite
detectionandtheabilitytoaccuratelydetectthetargetedmicroorganisms(2,9,14,17,21).Inaddition,theaccurateenumerationofGiardiaonmembranefiltershad
beenpreviouslyreported(20).Variousmembranefilters,includingcellulosenitrate,cellulosetriacetateandpolycarbonate,wereexaminedfortheirabilitytoretain
oocystsandcysts.Polycarbonatefilterswerenotsatisfactoryforthisapplicationbecausemanyorganismswere
TABLE2.RecoveryofGiardiacystsfromstoolsusingvarioussolutions.
CentrifugationSolutions

Specific
Gravity(g/mL)

Numberof
Trials

Average%
Recovery

40%PotassiumCitrate

1.16

76

Percollsucrose

1.09

77

FullStrengthSheather's

1.29

70

4/5Sheather's

1.24

68

3/5Sheather's

1.17

66

ZincSulfate

1.18

40

1500gmsucrose/320mLdeionizedwater,9.7mLphenol

TABLE3.EffectofDetergentsontheRecoveryofGiardiacystsduringclarification.

AverageRecoveries(%)

Solution
PotassiumCitrate(40%)
4/5Sheather's

WithTween80
SDS1

WithoutTween80
SDS

76.2

41.0

68.0

35.6

SodiumDodecylSulfate

500gmsucrose/320mLdeionizedwater9.7mLphenol

lostduringtheantibodywashingpartoftheprotocol.Cellulosenitrateandtriacetatefiltersperformedwell.Filterporesizewasalsoimportant.Membranefilterswitha
5.0mporositywere100%efficientintheretentionofGiardiacysts,while98.2%oftheCryptosporidiumoocystswerelost.Oocystlossesof57.5%werefound
usinga3.0mporosity,filterthusfilterswithaporesizeof1.2m,giving100%retentionoftheoocysts,wereusedforCryptosporidium(Table4).
TheforegoingimprovementsinmethodologywereputtousetostudyGiardiaandCryptosporidiumoccurrenceinenvironmentalwaters.Wastewatersandsurface
watersweresampledandcystandoocystconcentrationsweredetermined.Rawsewagesampleswerepositiveforbothprotozoa(Table4).Cryptosporidium
oocystsaveraged5191/LwhileGiardiacystsaveraged51/L.Cryptosporidiumwasdetectedinalltreatedeffluents(average1374oocysts/L)whileGiardiawas
detectedinonly40%ofthesamples(average1.3cysts/L).Basedonthispreliminarydata,itappearsthatCryptosporidiummaybelessefficientlyremovedby
secondarysewagetreatmentprocessesthanGiardia.Cryptosporidiumconcentrationswere100timeshigherthanGiardiainrawsewage.Theremaybeanumber
ofpossibleexplanationsforthis:1)theremaybemoreindividualsinfectedwithCryptosporidiumor2)thoseinfectedareexcretinglargenumbersofoocystsor3)
perhapstheoverallefficiencyofrecoveryanddetectionislowerforGiardia.
Threedifferentsurfacewatersourceshavebeenexaminedfortheparasites(Table6).Of21samples,
TABLE4.Enumerationofcystsandoocystsonmembranefilters.

Organism

Number
OfTrials

Percent
Recovered

Type

PoreSize

Cryptosporidium

1001

CN2

1.2

Cryptosporidium

42.5

CN

3.0

Cryptosporidium

1.9

CN

5.0

Giardia

100

CT

5.0

Giardia

46

PC4

5.0

1Countson0.2filterschosenas100%forcomparison.
2Cellulosenitrate.
3Cellulosetriacetate.
4Polycarbonate.

Filter

Page208
TABLE5.ConcentrationsofCryptosporidiumandGiardiainSewageinArizona.
RangeofVolumesSampled
(L)

Oocysts/L
Range
(Average)

Cysts/L
Range
(Average)

Rawsewaged

34170

84513,738(5191)

0.7198(51)

ChlorinatedSecondary
Effluent

121757

1433699(1374)

02.6(1.3)

TypeofSample

CryptosporidiumandGiardiaweredetectedin19,and12ofthesamplesrespectively,andoocystconcentrationswere10to20timeshigherthancystlevels.
Source1,ariver(originatingfromsource2)whichranthroughanareaconcentratedwithcattlepastureshadthegreaternumbersofbothparasites.Source2,alake
thatwasreceivingdomesticeffluentshadlowernumbers(7.1oocystsand0.35cysts/L)ofbothparasites.Source3wasariverinaprotectedwatershed.Thelow
numbersofoocystsandcystsmayhavecomefromindigenousanimalspeciessincebothCryptosporidiumandGiardiacanbefoundinothermammals(5,8,22).
WaterbornetransmissionofGiardiaiswellestablishedandrecently(6)hasbeendocumentedforCryptosporidium(7).Sewagecontaminationofdrinkingwaterhas
beenresponsibleformanyoftheseoutbreaks.Inafewcases,outbreakshaveoccurredinchlorinatedwatersystems,freeofcoliformbacteria.Inaddition,Giardia
outbreakshaveoccurredwhereanimalsotherthanmanhavebeenincriminatedasthecontaminatingsourceofawatersupply.Ithasbeensuggested,therefore,thatall
surfacewatersusedforpotablepurposesbefilteredanddisinfectedtopreventthetransmissionofGiardia(1).Similarconcernsandrecommendationsmaybestated
inthefutureforCryptosporidium.WatersfreeofGiardiacannotbeassumedtobefreeofCryptosporidiumsincethiscoccidianprotozoanwasalwaysfound
frequentlyandingreaternumbers.DuringoneinvestigationofanoutbreakofGiardiaincampers,thesuspectedstreamsampleyieldednoGiardia,however,
coccidianoocystsinsmallnumberswereobserved(4).Furtherresearchwithimprovedmethodsisnecessarytodocumenttheenvironmentaloccurrenceof
CryptosporidiumandGiardia.
TABLE6.ConcentrationsofCryptosporidiumandGiardiainsurfacewaters.
Source
Description

Numberof
Samples

1
Riverwithdomesticeffluentdischargesandcattlepasturerunoff

2
Lakewithdomesticeffluentdischarges

10

1.2

7.1

Cysts/L

28.5

3
Riverinaprotectedwatershed

Oocysts/L

0.35

0.08

0.009

LiteratureCited
1.Akin,E.W.andW.Jakubowski.1986.DrinkingwatertransmissionofGiardiasisintheUnitedStates.WaterSci.&Tech.18:219226.
2.AmericanWaterWorksAssociation.1985.GiardiaMethodsWorkshopIn:WaterSuppliesDetection,OccurrenceandRemoval.AWWA.p.49.
3.Arrowood,M.J.andC.R.Sterling.1987.IsolationofCryptosporidiumoocystsandsporozoitesusingdiscontinuoussucroseandisopycnicpercollgradients.J.
Parasit.73:314319.
4.BarbourA.G.,Nichols,C.R.andT.Fukushima.1976.AnoutbreakofGiardiasisinagroupofcampers.Am.J.Trop.Med.Hyg.25:384389.
5.CentersforDiseaseControl.1982.Humancryptosporidiosis:Alabama.MMWR31:252254.
6.Craun,G.F.1986.WaterbornediseasesintheUnitedStates.CRCPress,BocaRaton,FL.
7.D'Antonio,R.G.,Winn,R.E.,Taylor,J.P.,Gustafson,T.L.,Current,W.L.,Rhodes,M.W.,Gary,G.W.andR.A.Zayac.1985.Awaterborneoutbreakof
cryptosporidiosisinnormalhosts.Ann.Inter.Med.103:886888.
8.Davies,R.B.andC.P.Hibler.1979.AnimalreservoirsandcrossspeciestransmissionofGiardia.In:WaterborneTransmissionofGiardiasis.W.Jakubowski
andJ.C.Hoff.(eds).ReportEPA600/979001.Cincinnati,OhioU.S.EnvironmentalProtectionAgency.pp.104126.
9.Garcia,L.S.,Brewer,T.C.andD.A.Bruckner.1987.FluorescentdetectionofCryptosporidiumoocystsinhumanfecalspecimensusingmonoclonalantibodies.
J.Clin.Microbiol.25:119121.
10.Jakubowski,W.andT.H.Ericksen.1979.MethodsfordetectionofGiardiacystsinwatersupplies.In:WaterborneTransmissionofGiardiasis.W.
JakubowskiandH.C.Hoff.(eds).ReportEPA600/979001.Cincinnati,OhioU.S.EnvironmentalProtectionAgency.pp.193210.
11.Jakubowski,W.1984.DetectionofGiardiacystsindrinkingwater:StateoftheArt.In:GiardiaandGiardiasis.S.L.ErlandsenandE.A.Meyer(eds).Plenum
Press,NewYork.pp.263285.
12.Lippy,E.C.andS.C.Waltrip.1984.Waterbornediseaseoutbreaks19461980:Athirtyfiveyearperspective.J.Amer.WaterWorksAssoc..76:6067.
13.Moore,G.T.,Cross,W.M.,McGuire,D.,Mollohan,C.S.,Gleason,N.W.,Healy,G.R.,andL.H.Newton.1969.EpidemicGiardiasisataskiresort.N.Engl.J.
Med.281:402407.
14.Musial,C.E.,Arrowood,M.J.,Sterling,C.R.,andC.P.Gerba.1986.DevelopmentofamethodforthedetectionofCryptosporidiuminwater.Appl.Environ.
Microb.53:687692.
15.Rose,J.B.,Musial,C.E.,Arrowood,M.J.,Sterling,C.R.,andC.P.Gerba.1985.DevelopmentofamethodforthedetectionofCryptosporidiumindrinking
water.In:AdvancesinWaterAnalysisandTreatment.WaterQualityTechnologyConference,Houston,TX,Amer.WaterWorksAssoc.,Denver,CO.Dec.8
11.p.117.
16.Rose,J.B.,Cifrino,A.,Madore,M.S.,Gerba,C.P.,Sterling,C.R.andArrowood,M.J.(1986).DetectionofCryptosporidiumfromwastewaterandfresh
waterenvironments.WaterSci.Techn.18:233239.
17.Rose,J.B.,Madore,M.S.,Riggs,J.L.,andC.P.Gerba.1986.DetectionofCryptosporidiumandGiardiainenvironmentalwaters.WaterQualityTechnology
Conference,Amer.WaterWorksAssoc.Denver,CONov.1619,Portland,OR.
18.Sauch,J.F.1985.UseofimmunofluorescenceandphasecontrastmicroscopyfordetectionandidentificationofGiardiacystsinwatersamples.Appl.Environ.
Microbiol.50(6):14341438.

Page209

19.Shaw,P.K.,Brodsky,R.E.,Lyman,D.O.,Wood,B.T.,Hibler,C.P.,Healy,G.R.,MacLeod,K.I.E.,Stahl,W.andM.G.Schultz.1977.Acommunitywide
outbreakofGiardiasiswithevidenceoftransmissionbyamuniciplewatersupply.Ann.Intern.Med.87:426432.
20.Spaulding,J.J.,Pacha,R.E.,andG.W.Clark.1983.QuantitationofGiardiacystsbymembranefiltration.J.Clin.Microbiol.18(3):713715.
21.Sterling,C.R.andM.J.Arrowood.1986.DetectionofCryptosporidiumsp.infectionsusingadirectimmunofluorescentassay.Ped.Infect.Dis.5:51395142.
22.Woo,P.K.1984.EvidenceforanimalreservoirsandtransmissionofGiardiainfectionbetweenanimalspecies.In:GiardiaandGiardiasis.S.L.Erlandsenand
E.A.Meyer(eds).PlenumPress,NewYork,pp.341364.

Page211

ComparisonofFiveProceduresfortheSedimentationofGiardiaLambliaandOtherProtozoanCysts
D.R.Pennell*,J.F.Stoebig,D.E.Sampson,andR.F.Schell.
StateLaboratoryofHygiene,UniversityofWisconsin,465HenryMall,Madison,Wisconsin53706,U.S.A..
ThreecommercialsedimentationproductsFeKalCONTrate(TrendScientific),FecalParasiteConcentrator(EvergreenScientific),andParaPakMacroCon(Meridian
Diagnostics)wereevaluatedfortheirabilitytoconcentrateGiardialambliaandotherprotozoancystsfromFormalinpreservedspecimens.Allthreeproducts,utilizingethylacetate
asadigestionagentandthe(manufacturersupplied)detergent,werecomparedwiththestandardFormalinacetate(FA)sedimentationprocedurewithandwithoutadditionofa
detergent,TritionX100.Initialandfinalcystcountsallowedforcalculationoftheconcentrationefficacyofeachprocedure.FAwassuperiortoallthreecommercialproductsfor
theconcentrationofG.lambliacystsandforcystsofEntamoebacoli,Entamoebahistolytica,Endolimaxnana,andChilamastixmesnili.AdditionofdetergenttothestandardFA
procedureloweredthecystyields.Althoughthecommercialproductssimplifytheprocedure,FAsedimentationresultsinoptimalrecoveryofG.lambliaandotherprotozoancysts.

Introduction
TheclassicalFormalinether(FE)sedimentationprocedureforconcentrationofGiardialamblia,otherprotozoancysts,andhelmintheggsinfecalspecimenswas
describedbyRitchiein1948(5).AnimportantmodificationbyYoungetal.in1979(8)replaceddiethyletherwithalessvolatilesolvent,ethylacetate.Theresulting
Formalinacetate(FA)procedureissaferforlaboratoryuseandproducescomparableresults(1,7).FAisthestandardsedimentationprocedurefortheconcentration
oftheseparasitestoday.
Recentlythreecommercialsedimentationproductshavebeenintroduced.FecalParasiteConcentrator(FPC),originallydescribedandevaluatedagainstFE
sedimentationin1978(9),hassincebeenmodifiedandisnowmarketedbyEvergreenScientific.FeKalCONTrate(FCTTrendScientific)hasbeenevaluated
againstFA(3).ParaPakMacroCon(PPMCMeridianDiagnostics)hasnotbeenpreviouslyevaluated.
Themanufacturersofthesecommercialproductseitherrecommendorstipulatetheuseofacetateasalipiddigestionagent.Eachproductisalsosuppliedwitha
secondreagent,describedasonewhichreducesadhesiveforcesand/orhelpstobreakdownfecalaggregates.Thisreagentwasidentifiedas20%TritonX100for
FPC.ItisnotidentifiedforFCTorPPMC.TheuseofTritonX100,inasedimentationprocedure,hasbeenpreviouslydescribed(9).
InthisstudythethreecommercialsedimentationprocedureswerecomparedwithFAsedimentation,aloneandwiththedetergentTritonX100(FAD),fortheirability
toconcentrateprotozoancysts.
MaterialsandMethods
Specimens
Humanfecalspecimensweretransportedtoourlaboratoryin10%bufferedFormalinforroutineparasitologicexamination.Specimensestablishedaspositivefor
protozoancystsbyconventionalFAorFEwereusedforthisstudywithin1weekofreceipt.Onlyspecimenswithenoughcyststopermitaninitialcountwereused.
Ofthe43specimens,19containedcystsofG.lamblia,11Entamoebacoli,4Entamoebahistolytica,8Endolimaxnana,and1Chilomastixmesnili.
FecalHomogenates
Eachfecalhomogenatewaspreparedbysuspendingthespecimenin10%bufferedFormalinandadjustingthesuspensiontoninepartsliquid,onepartsediment.
SedimentvolumesweredeterminedwithaWassermanngraduatedcentrifugetube.
SedimentationProcedures
Withcontinuousmixingonamagneticstirrer,aspecifiedvolumeofeachfecalhomogenatewastransferredforeachsedimentationprocedure(Figure1).FAwas
performedinaccordancewiththerecommendationsoftheCentersforDiseaseControl,UnitedStatesDepartmentofHealthandHumanServices(4).FAwasalso
performedwiththeincorporationof20%TritonX100(FAD).Thecommercialprocedureswereperformedaccordingtothemanufacturers'instructions.Detailsof
allfiveproceduresareshowninTable1.Finalsedimentvolumesweredeterminedandadjustedtoninepartsliquid,onepartsediment.
CystCounts
Cystcountswereperformedonalloriginalfecalhomogenatesandfinalsedimentationsuspensionsbymakingduplicate1:4or1:16dilutionsofeachsuspensionand
transferring10Lofeachdilutiontoahemocytometer.PhysiologicalsalinecontainingDobellandO'Conner'siodinewasusedasadiluent.Allcountswereadjusted
forthedilution.Countswereperformedbyatrainedmicrobiologistwithoutknowledgeofthesedimentationprocedureutilized.
Calculations
Foreachsedimentationprocedureperformedwitheachspecimen,aconcentrationcoefficient(CC)wascalculatedastheratiooftheconcentrationinthefinal
sedimentationsuspension(FC)totheconcentrationintheintialhomogenate(IC):CC=FC/IC.ThusCC>1showsahigherconcentrationofcystsinthefinal
suspension.ThehighertheCC,thegreatertheconcentrationefficacy.ThemeanCCforeachprocedurewascalculated,andthemeanswerecomparedbythetwo
tailedTtestforpairedsamples(6).P<0.05wasdefinedassignificant.Precisionwascalculatedasthecoefficientofvariation(CV)ofthefinalcystcounts(n=4for
eachspecimen).
*Correspondingauthor.

Page212
TABLE1.Proceduralstepsandconditionsforthefivesedimentationprocedures.

Proceduralsteps

Formalin(FA)

Formalin
Trend,with
AcetateAcetate detergentTrate
(FAD)
(FCT)

Evergreen,FeKal
CON
Meridian,Fecal
Concentrator
ParasiteMacro
(FPC)
Con(PPMC)

7.5mL

7.5mL

7.5mL

7.5mL

12.5mL1

2.Detergentadded(amount,
type)

3drops,20%
TritonX100

2drops
supplied2

3drops,20%
TritonX100

10drops
supplied2

3.Acetateadded(mL):

3.0

4.Shaking(s)

30sec.

30sec.

30sec.

60sec.

2layersgauze

2layersgauze

metalscreen3

plasticmesh3

plasticmesh3

6.Firstwashsedimentation

650g,2min.

650g,2min.

500g2min.

7.Secondwashsedimentation

650g,2min.

650g,2min.

3.0

3.0

3.0

5.0

1.Fecalhomogenateused(mL)

5.Filtration(type)

8.Acetateadded(mL):
9.Shaking(s)
10.Finalsedimentation

30sec.

30sec.

30sec.

60sec.

500g,2min.

500g,2min.

500g,2min.

500g,2min.

500g,2min.

InitialvolumegreaterforPPMCthanforotherprocedures,butthehomogenate/acetateratioswereconstant.

Manufacturersupplied,formulationnotgiven.

Manufacturersupplied.

Results
TheprecisionofthereferenceFAprocedurewithfourreplicateswasCV7.2and10.6%forfinalconcentrationsof229,000and841,000cysts/mLrespectively
(Table2).ForG.lambliaFAwithoutTritonX100hadgreaterconcentrationefficiencythanFAD(CC7.5vs.6.7Table3).Thedifferencebetweenthesemeans
wasstatisticallysignificant(P<0.03).Similarresultswereobtainedforcystsofotherprotozoans(P<0.02).FAwasmoreeffectivethananyofthecommercial
procedures(P<0.001forFAvs.eachprocedurewithG.lambliaandP<0.003withotherprotozoancysts).Forbothtypesofspecimenstherelativeefficacyofthe
commercialproductswasFCT>FPC>PPMC(P<0.02).EachprocedureproducedawiderangeofCC(Table3).ForPPMCoccasionalspecimensproduced
CC<1.

Figure1.
MethodofAnalysisofSedimentationProcedures.

Discussion
Manyrecentevaluationsofparasitesedimentationproceduresutilizedamethodofevaluationwherebythesedimentationsandslidepreparationswereperformedas
theywouldbeinaroutineclinicalparasitologylaboratory(13,7,8).Thepresumedadvantageofthisapproachisthatthefindingsareapplicabletoroutinepractice.
However,theroutinepracticefortransferofaportionofthefinalsedimentandforslidepreparationentailsariskofvariableresults.Ifcystconcentationsvary
throughoutthesedimentpellet,theportiontransferredtotheslidemaynotberepresentative.Theratioofliquidtoparticulatematerialmayalsovary,andthedepthof
materialunderastandardslidecoverslipcanvarybothfromslidetoslideandfromoneareatoanotherwithinaslide.Inanattempttoeliminatethesefactorsof
potentialimprecision,themethodofanalysisusedinthisstudyincorporatedastandardizedfinalsedimentsuspensionwithtransferofaconstantvolumeofwellmixed
suspensiontoahemocytometerforcystcountdetermination.Inaddition,foreachsedimentation,twosuspensionsoffinalsedimentwerepreparedandcounted.A
precisionanalysiswhenappliedtotheFAprocedureproducedCV7.2and10.6%atfinalconcentrationsof229,000and841,000cysts/mL,respectively(Table2).
AllfivesedimentationproceduresproducedawiderangeofCC,yettherelativeefficacyoftheseproceduresremainedveryconsistentfromonespecimentoanother.
ThusthewideCCrangeforeachprocedureappearstoreflectvariationsintheinherentpropertiesofthefecalspecimens.ForPPMC,occasionalspecimensyieldeda
TABLE2.Precisionevaluationforthemethodofanalysisusedtocomparethesedimentationprocedures.
Sedimentation
Procedure

MeanCystCountinFinal
SedimentSuspension(cysts/mL)

CV

FA

229,000

7.2%

FA

841,000

10.6%

Replications

Page213
TABLE3.Efficacyofsedimentationprocedures.

G.lambliaCysts(n=19)

FA

FAD

FCT

FPC

PPMC

MeanCC

7.5

6.7

5.6

3.5

2.7

CCrangehigh

16.3

15.7

12.2

8.6

7.3

low

1.8

1.7

1.8

1.1

0.6

OtherProtozoanCysts(n=24)

MeanCC

6.8

6.0

5.7

4.0

2.5

CCrangehigh

13.0

10.9

9.1

8.0

4.7

low

2.4

1.9

2.3

1.6

0.8

CC<1i.e.,thecystcountswerehigherintheinitialfecalhomogenatethaninthefinalsedimentsuspension.
IncorporationofTritonX100intotheFAproceduredecreasedFA'sconcentrationefficacyforcysts.Thisiscontrarytothereportedefficacyforuseofdetergentin
theconcentrationofparasiteeggs(9).Thisdiscrepancyposesaproblemfortheclinicalparasitologylaboratory,whereoptimalrecoveryofbothcystsandeggsis
ideal,yetperformanceoftwosedimentationsforeachspecimeniscostly.Confirmationofthesefindingswouldbeuseful,aswouldaninvestigationofalternative
detergentsand/ordetergentconcentrations.
FCTwasthemostefficaciousofthecommercialproductstested,followedbyFPCandthenPPMC.However,FAwassuperiortoallthreecommercialproducts.
ThisfindingisinconsistentwithapreviousreportcomparingFCTandFA(3).Oneapparentdifferencebetweenthetwostudiesisinthemethodsusedtoanalyzethe
procedures.
Allthreecommercialproductsmadesedimentationeasier,partlythroughuniqueproductdesign,especiallyatthelevelofspecimenfiltration,andpartlybyomissionof
stepsthatarerequiredinthestandardFA(Table1).Inaddition,bothFPCandPPMCfunctionasanenclosedprocessingunitduringmuchoftheprocedure,
improvinglaboratorysafetybyreducingthechancesofexposuretothespecimen,Formalin,andacetate.Theseconveniences,however,aregainedattheexpenseof
sensitivityincystdetection.
LiteratureCited
1.Erdman,D.D.1981.ClinicalcomparisonofethylacetateanddiethyletherintheFormalinethersedimentationtechnique.J.Clin.Microbiol.14:483485.
2.Garcia,L.S.,andR.Shimizu.1981.ComparisonofclinicalresultsfortheuseofethylacetateanddiethyletherintheFormalinethersedimentationtechnique
performedonpolyvinylalcoholpreservedspecimens.J.Clin.Microbiol.13:709713.
3.Long,E.G.,TsinA.T.,andB.A.Robertson.1985.ComparisonoftheFekalCONTrateSystemwiththeFormalinethylacetatetechniquefordetectionof
intestinalparasites.J.Clin.Microbiol.22:210211.
4.Melvin,D.M.,andM.M.Brooke.1974.Laboratoryproceduresforthediagnosisofintestinalparasites.U.S.DepartmentofHealthandHumanServices
publicationno.(CDC)758282.U.S.GovernmentPrintingOffice,Washington,D.C.p.104106.
5.Ritchie,L.S.1948.Anethersedimentationtechniqueforroutinestoolexaminations.Bull.U.S.ArmyMed.Dept.8:326.
6.Snedecor,G.W.,andW.G.Cochran.1967.Statisticalmethods.6thed.IowaStateUniversityPress,Ames,Iowa.p.6264,9197.
7.Truant,A.L.,Elliott,S.H.,Kelly,M.Tm.,andJ.H.Smith.1981.ComparisonofFormalinethylethersedimentation,Formalinethylacetatesedimentation,andzinc
sulfateflotationtechniquesfordetectionofintestinalparasites.J.Clin.Microbiol.13:882884.
8.Young,K.H.,Bullock,S.L.,Melvin,D.M.,andC.L.Spruill.1979.EthylacetateasasubstitutefordiethyletherintheFormalinethersedimentationtechnique.J.
Clin.Microbiol.10:852853.
9.Zierdt,W.S.1978.Asimpledeviceforconcentrationofparasiteeggs,larvae,andprotozoa.Am.J.Clin.Pathol.70:8993.

Page215

ComparisonoftheModified''ReferenceMethod"andtheIndirectFluorescentAntibodyTechniqueforDetectionofGiardia
CystsinWater
BennyE.Quinones*,CharlesP.HiblerandCarrieM.Hancock.
DepartmentofPathology,ColoradoStateUniversity,FortCollins,Colorado80523,U.S.A..
Sixtymunicipalwatersamplesfromvariousgeographicareaswereanalysedbyboththemodifiedreferencemethod(ZincSulfate)andtheIndirectFluorescentAntibody(IFA)
techniquetomonitorsurfacewatersourcesforcystsofGiardia.StatisticalanalysiswiththeStudent'sttestindicatesnostatisticallysignificantdifferencebetweentechniquesif
analystswerehighlyskilledandexperiencedwiththetechniquetheywereusing.Differencesnotedwereoftenpersonalpreferences.Inexperiencedanalystsusuallyfoundmorecysts
withIFA.Bothtechniqueswereaffectedbywaterqualityrecoveryofcystswasinverselyproportionaltotheturbidityofthesourceandhighorganicturbiditysamplescontaining
considerablealgaeseverelylimitedtheefficiencyofbothtechniques.IFArequiredmorepreparationtime/sampleandthemicroscopefatiquefactorwasgreater,thereforeIFAcould
notbeusedeffectivelybylaboratoriesmonitoring10ormoresamples/day.Thecostfactorforaqualityepifluorescentmicroscopewithphasecontrastcapabilitiesandthecostof
antibodywouldbeaconsiderationforlaboratoriesplanningtouseIFA.Confidenceintheresultswouldbecontingentupontheavailabilityandreliabilityoftheantibody.

Introduction
Duringthepast11yearswehaveexaminedover6000watersamplesforcystsofGiardia.Mostofthesearefrommunicipalities,CountyandStateHealth
Departmentsrequestingthisservice,somearefromresearchonfiltrationsystemsandsomearefromforeigncountries.Initiallyweusedthe"referencemethod"
describedbyJakubowskiinVolume16,StandardMethods(1)andthenswitchedtotheoverlay/underlaymodificationofthistechnique.
Sauch(1985)(3)developedanindirectfluorescentantibodytechnique(IFA)forthedetectionofGiardiacystsinsurfacewatersources,andwithherassistancewe
incorporatedthetechniqueandofferedthistypeofanalysisalongwiththemodifiedreferencemethodwegenerallyuseforanalysisofwatersamples.Ourlaboratory
routinelymonitors120samples/dayor80100samples/monthformunicipalitiesrequestingthisservice.SamplesaresentfromgeographicareasacrosstheUnited
States.SinceimmunofluorescencehasnotbeenusedtoanygreatextentformonitoringsurfacewatersuppliesforcystsofGiardia,wefeltthatanevaluationofthe
advantagesanddisadvantagesofbothtechniquesformonitoringpurposeswasnecessary.RecoveryofGiardiacystsfromcontaminatedsourcesduringanepidemic
ofwaterbornegiardiasisisrelativelysimplebecausecystnumbersoftenreach1015/gallonofwaterhowever,routinemonitoringfortheirpresencefrequentlymeans
attemptingtofindonecyst/100gallonsofwaterandthiscanbeatimeconsumingprocedurenecessitatingnumeroushoursofmicroscopetime.
MaterialsandMethods
Samplesofconcentratefrom60differentmunicipalsourceswereanalyzedbythemodifiedreferencemethod,thetechniquewegenerallyusewithZnSO4for
monitoring,withacoverslipusedinthefinalstep,andwithSauch'stechnique(3).Theonlychangesmadeinhertechniquewastheeliminationofthedehydrationsteps
andthedye.TheresultswerecomparedbythePairedDifferenceStudent'sttest.Samplesforcomparisonwereselectedatrandom.Wheneverpossible,thesame
amountofconcentratewasusedandthechemicalusedforselectiveconcentrationduringthefirststepinthecleanupprocedurewasatthesamespecificgravity.On
occasionfurtherdilutionoftheconcentratewasnecessarywiththeimmunofluorescencetechniquebecause25mmdiametermembraneswouldplugwhereasthe47
mmmembranesusedinthemodifiedreferencemethoddidnotplug.
Moreover,alowerspecificgravitysometimeswasnecessarytoovercomethepluggingencounteredwiththesmallermembranesizeusedintheIFAtechnique.Lower
specificgravitiesofthechemicalresultinlowercystrecovery(2).
Results
Thesamplesanalyzedvariedfromwesternwatersourcescontainingprimarilyinorganicturbiditytoeasternwatersourcescontaininghighlevelsoforganicturbidityand
numerousalgae,flagellates,etc.inthesample.Irrespectiveofthetechniqueemployed,increasingturbidityinterferedwiththeselectiveconcentrationofcystsandother
organisms.Iftheturbiditywasprimarilyorganic,orthesamplecontainedalumand/orpolymers,considerabledilutionwasnecessary.Forafewofthesamples,useof
chemicalswithalowerspecificgravity(e.g.1.1)wasnecessary.Apparentlybothtechniqueswere
*Correspondingauthor.

Page216

equallycompromisedbythewaterqualitybecauseresultsoftheanalysiswerestatisticallyinsignificantforthe60samplescompared.Atotalof446cysts,rangingfrom
0to50cysts/samples,averaging7.4cysts/samplewasfoundbythemodifiedreferencemethodand383cystsrangingfrom0to48cysts/sample,andaveraging6.4
cysts/samplewasfoundbyimmunofluorescence.Themeanofthedifferencesbetweentechniquesforeachpairedsamplewas1.07,range+21.Thisvalueisnot
statisticallydifferentfrom0(p<0.01).Thedifferencewaslessthan+21in17(28%)ofthesamples.
Discussion
Theresultsofthiscomparisonindicatedthatifanalystswerehighlyskilledandexperiencedwiththetechniqueusedforanalysisresultswerenotsignificantlydifferent.
However,comparisonofthetwotechniquesusinglessexperiencedanalystsindicatedthatinexperiencedindividualssometimeswouldfindmorecystsby
immunofluorescence,butthedifferencesobservedthroughoutthetrialwerenotsignificant.
UseofdirectmicroscopyorthemodifiedreferencemethodfordiagnosisofGiardiacystsinwaternecessitatesconsiderabletraininginanumberofacademic
disciplinesandexperiencewiththemicroscope.Preferablyindividualshavetrainingandexperienceinprotozoology,parasitology,streamorfreshwaterbiologyand
botany.Ourexperiencefromtrainingindividualsindicatesthatstudentswithabroadbackgroundinbiologyormicrobiologybecomeproficientinashortperiodoftime
andgainconfidenceastheyanalyzemoresamplesprovidingtheyfrequentlyseepositivesamplesorusequalityassurancesamples.Ifanindividualisnotacademically
trainedinthesedisciplinesimmunofluorescencemaybeabetteroption,butcurrentlywedoubtthateffectiveanalysiscanbeperformedbyeithertechniquewithout
considerableexperienceandacademictraining.
Fortheaveragesample,preparationtimeandanalysisbyIFArequiredapproximatelytwicethetimenecessarytoprepareandanalyzeasamplebythemodified
referencemethod.Iflaboratoriesareanalyzing10ormoresamples/daythetimefactorwouldbeprohibitive,especiallyiftheanalysthadotherduties.Analysts
experiencedwithbothtechniquesfeltthatthemicroscopefatiguefactorwasconsiderablygreaterwithIFAthanwiththemodifiedreferencemethod.
Laboratoriesplanningtouseeithertechniquemustrealizethatinvestmentinanexcellent,highqualitymicroscopewithastrong,wellbalancedlightsourceisamust.
Thecostofanepifluorescentmicroscopewithphasecontractcapabilitiesismuchgreaterthanaqualitybrightfieldmicroscope,afinancialconsiderationforprivate
ormunicipallaboratoriescontemplatinganalysiswithimmunofluorescence.
Currentlypolyclonalormonoclonalantibodyisnotcommerciallyavailabletolaboratoriesplanningtouseimmunofluorescenceforanalysisofwatersamples.Nodoubt
commerciallaboratorieswillsoonmakeantibodyavailable.ThepolyclonalantibodiesinusebyseverallaboratoriesarecertainlyspecificforGiardiacysts,however,
ourantibodywillnotdistinguishbetweencystsfrommammalsandthosefrombirds.Futureresearchmayshowthatseveral"strains"ofGiardiaarepresentinhuman
and/oranimalpopulations,andmorethanonestrainmaybepresentinanygivenhumanoranimal.Whatcriteriawillbeusedfortheselectionof"strains"toproduce
theseantibodies?WillmonoclonalantibodydistinguishbetweenGiardiacystsinfectiousorpotentiallyinfectiousforhumans,asopposedtothosefromanimalsor
birdsthatarenotinfectiousforhumans?Thesearequestionsneedinganswersbeforeimmunofluorescencecanberecommendedasbetterthananyothertechnique.
Anotherkeyproblembesidesthequalityofantibodyistheavailabilityofareliableantibody.
AnimportantconcernwiththeIFAtechniqueisthatparticulateanalysiscannotbeperformedandthisanalysisisessentialtoassessingfilterplantefficiency.Plant
efficiencyinremovingparticulatesthesizeofandlargerthanGiardiaismoreimportantthandetectingGiardia.
AproblemwiththemodifiedreferencemethodisthatZnSO4,potassiumcitrateandsucrose,thechemicalsgenerallyusedtoselectivelyseparatecysts,are
hygroscopicandcauseshrinkage.Thisnecessitatesexperiencetorecognizethecysts,however,cystswerereadilyvisualizedandidentifiedbyexperiencedindividuals.
Ifpercollwasusedshrinkagedidnotoccur,unfortunatelythischemicalisnotcosteffectiveforroutinemonitoringofsurfacewatersupplies.Anotherproblemwiththe
referencemethod,especiallywhenusingZnSO4,isthenecessityofstainingwithLugol'siodineifabrightfieldmicroscopeisusedforanalysis.Theiodinealso
effectivelystainsalgaeandflagellates.Visualizingacystamongmyriadsofsimilarshapedandsimilarstainedorganismsquicklyinitiatesmicroscopefatigue,however,
thebackgroundfluorescencefromalgaealsointerferedwithanalysisbytheimmunofluorescenttechniqueandwithcomparableresults.Occasionallyalgaecompletely
prohibitsanalysisbyIFA.Aswasshownbythestatisticalanalysis,bothtechniquesapparentlywereeffectivelycompromisedifsampleshadtheseorganismspresent.
Personalpreferencesandbiaseswasaproblemforoneoftheanalysts.Sincethisindividual(CPH)hadexperiencewithmorethanonetechniquehedislikedsittingina
darkenedroomandswitchingfromepifluorescencetophasecontrastmicroscopytoconfirmidentificationofthecystsvisualizationandconfirmationbyLugol's
stainedsamplesusingZnSO4wasquickerandeasier.Moreover,evaluationofthequalityofthecyst(possiblyalive,definitelydead,etc.)waseasierwiththeZnSO4
technique.Oftentheimmunofluorescencetechniquewoulddetectorganismsthatmayhavebeenacystatonetimebutwasnotrecognizableandhadtobelistedasa
"cystlikestructure".
LiteratureCited
1.A.P.H.A.,A.W.A.A.,W.P.C.F.1985.StandardMethodsfortheExaminationofWaterandWasteWater.16thEdition,A.P.H.A.WashingtonD.C..pp.1268.

Page217

2.Hibler,C.P.1987.AnoverviewofthetechniquesusedfordetectionofGiardiacystsinsurfacewater.(CalgaryGiardiaConference).
3.Sauch,J.F.1985.UseofimmunofluorescenceandphasecontrastmicroscopyfordetectionandidentificationofGiardiacystsinwatersamples.Appl.Environ.
Microbiol.50(6):14341438.

Page219

GiardiaDetectionusingMonoclonalAntibodiesRecognizingDeterminantsofInVitroDerivedCysts
C.R.Sterling*,R.M.Kutob,M.J.Gizinski,M.Verastegui,andL.Stetzenbach
DepartmentofVeterinaryScience,UniversityofArizona,Tucson,Arizona85721,U.S.A..
Monoclonalantibodies(MAbs)weremadetoinvitroderivedcystsofanaxenicallyculturedPeruvianGiardiaisolate.Encystmentofthetrophozoiteswasaccomplishedusing
Keister'sModifiedTYIS33mediumwithoutbileincubatedundera90:10nitrogen:carbondioxideatmosphere.BALB/cmicewereimmunizedwithsonicatedinvitroderivedcysts.
HybridomafusionswereperformedusingspleencellsofsensitizedmiceandP3x63Ag8.653myelomacells.Antibodytocystswasdetectedin11of144wellsbyindirect
immunofluorescence.PositivewellswereclonedandMAbswererecoveredinascitesfluid.OneMAb(2B3G6)anIgG,gaveanIIFtiterof100,000.ThisMAbandculture
supernatantfromunclonedpositivewells(6C1and2B3)werehighlyspecificforGiardiacystsanddidnotcrossreactwithEntamoebahistolytica,E.coli,E.hartmanni,Eimeriasp.,
Cryptosporidium,Candidasp.,Rhodotorulasp.,oralgae.TheMAbandtheculturesupernatantswereusedtodetectGiardiacystsinfecalsmearsfromsymptomaticand
asymptomaticpatients,primary,secondary,andtertiarysewageeffluentsfromPeru,andsurfacewatersfromArizonausingindirectimmunofluorescence.Thespecificityofthese
MAbsgreatlyenhancedGiardiadetectionandallowedfortherapidscreeningoffecalandenvironmentalsamples.

Introduction
NewandimproveddiagnosticmethodologieshaverecentlybeendevelopedtodetectGiardiainsymptomaticandasymptomaticindividuals.EnhancedGiardia
detectionhaslargelyresultedfromthedevelopmentofseveralhighlysensitiveimmunodiagnosticproceduresincludingindirectimmunofluorescentdetectionofserum
antibodies(20,21),counterimmunoelectrophoresis(CIE)offeces(5),andenzymelinkedimmunosorbentassays(ELISA)fordetectionofserumantibodies(21)or
thepresenceofGiardiaantigensinfecalspecimens(8,19).
TheprincipletechniqueforGiardiadetection,however,remainsconventionalmicroscopy(4).Microscopicexaminationofmultiplestoolsamples,flotation
concentrates,duodenalcontents,andintestinalbiopsiesforcharacteristictrophozoitesandcystsofGiardiamayberequiredforconfirmationofinfection
(4,13,15,22).Positiveidentificationoforganismsfromthesesamplesisdefinitive.Unfortunately,notalllaboratoriesareskilledinperformingmanyofthesetechniques,
resultinginvariableratesofaccuratediagnosis.DetectionofGiardiainenvironmentalwatershasbeenevenmoredifficult.First,itisusuallynecessarytoprocess
largevolumesofwatertodetectthefewcystslikelytobepresent.Second,detectionandreportingerrorsmayresultfromthepresenceofalgalcells,artifactsorother
celltypesresemblingGiardiacystsinsizeandshape.
Immunofluorescentstainingprocedureshavebeenusedrathersparinglytodetectparasitesdespitethefactthattheyhavebeenavailableforover40years(3).
Recently,severalworkershavedevelopedpolyclonalantibodiestoGiardiaandCryptosporidiumanddescribedtheiruseineitherindirect(IIF)ordirect(DIF)
immunofluorescentassaystodetecttheseorganismsinfecalandenvironmentalsamples(14,16,18).Inaddition,monoclonalantibodies(MAbs)havebeenproduced
todeterminantsoftheoocystwallofCryptosporidiumandsimilarlyusedinIIFandDIFassays(7).TheuseoftheMAbswasconsideredtobeatleasttentimes
moresensitivethanotherspecializedstainingtechniques.
ThesuccessfuldevelopmentandapplicationofMAbsforthediagnosisofCryptosporidiumwithinourownlaboratoryledustoattemptasimilarstrategyat
developingMAbreagentswhichcouldbeusedtoidentifyGiardia.Inthepresentstudy,wedescribethedevelopmentoftheseMAbsusinginvitroderivedGiardia
cystsformousesensitizationpriortohybridfusions.Inaddition,wedescribetheuseoftheseMAbstodetectGiardiainbothfecalandenvironmentalsamplesusing
anindirectimmunofluorescentassay.
Methods
InvitroEncystment
AnaxenicallyculturedhumanGiardiaisolateofPeruvianoriginwasinducedtoencystinvitro.FourdayoldtrophozoiteculturesgrowninKeister'sTYIS33
medium(10)wereplacedonicefor10minutesandharvestedbycentrifugationfor10minat250g..Thepelletsoftwoculturetubeswerecombinedandincubated
undera90:10nitrogen:carbondioxideatmospherefor48hrs.in1.5mLofKeister'sModifiedTYIS33mediumwithoutbile.Cultureswerethenplacedat23Cfor
24hrs.afterwhichtheywerecentrifugedandthepelletsresuspendedin1.0mLphosphatebufferedsaline(PBS,pH7.4)forexaminationbyphasecontrast
microscopy.
*Correspondingauthor.

Page220

Figure1.
Phasecontrastimageofinvitroderivedcysts.Bar,10m.

ScanningElectronMicroscopy
Invitroderivedcystswerepreservedwith4%formalin,1%glutaraldehydein0.1M.phosphatebuffer,pH7.2(12)andfilteredonto0.1mpolycarbonatefilters.
Filterswerepostfixedwith2%osmiumtetroxide(OsO4),dehydratedthroughagradedethanolseries,andcriticalpointdriedusingethanol:FreonTF(50:50)(2).
Processedfiltersweremountedonstubsandcoatedwithgold:paladium.SampleswereviewedwithanISIDS130scanningelectronmicroscope.
GiardiaCystViability
Theinvitroderivedcystsweretestedforviabilityusinginvivoandinvitroassays.Twelve45dayoldCF1micewereselectedfortheinvivoviabilitystudy.Nine
ofthemicewerefed8104cysts/mouseandtheremainingthreeservedascontrols.Onecontrolmouseandthreetestmiceweresacrificedonday4,6,and8.
Sectionsofintestinewereexcised,placedinPBS,andexaminedforGiardiatrophozoitesbyphasemicroscopy.
Invitroexcystation,asameasureofviability,wasattemptedusingthetechniqueofBhatiaandWarhurst(1).Invivoderivedcystswereaddedtohydrochloricacid
(pH2)for15minutes,centrifuged(250g,10min.),andwashedtwiceinKeister'sModifiedMediumSolutionA.CystswerethensuspendedinKeister'scomplete
mediumandincubatedat37C.Thesuspensionwasexaminedfortrophozoitesattimedintervals(0,10,20,60,90min.,2hrs.,and24hrs.)usingphasemicroscopy.
MonoclonalAntibodies
MonoclonalantibodiesrecognizingGiardiacystwalldeterminantswereproducedusingtheinvitroderivedcystsasasourceofantigenformouseimmunizations.
BALB/cmicewereimmunizedwithsonicatedsuspensionsof106cysts/mouseatday0(Freund'scompleteadjuvantI.M.),14(Freund'sincompleteadjuvantI.M.),
and27(I.V.).Themicewerebledfromtheretroorbitalplexusonday30andtheserumtestedforIIFreactivitytoGiardiacystsobtainedfromahumanfecalsample.
Spleencellsofseropositiveimmunizedmicewerefusedonday31withP363Ag8.653myelomacellsusingpolyethyleneglycol4000(6).Antibodytothehuman
sourceGiardiacystswasdetectedin11of144fusionwellsbyIIF(11).Culturesupernatantsfromtwounclonedexpandedcelllines(6C1and2B3)wereusedfor
initialGiardiascreening.Cellsfromfusionwell2B3wereclonedbylimitingdilutionandputintopristaneprimedBALB/cmiceforascitesproduction(6).Ascitesfluid
fromthisclonedcelllineandculturesupernatantsfromunclonedcelllineswerescreenedforcrossreactivitywithvariousprotozoa,yeast,andalgaeinanIIFassay.
ImmunoglobulinisotypeandsubclasswasdeterminedbyIIFusingbiotinylatedsubclassspecificantiseraandstrepavidinlabelledfluoresceinisothiocyanate(FITC).
DetectionofGiardiainFecalSamples
FormalinfixedhumanfecalsamplessubmittedbyMaricopaCountyHealthagencieswerescreenedforGiardiabyIIFusingtheMAbs.Sampleswereobtainedfrom
routinesubmissionsbynewly

Figure2.
Scanningelectronmicrographofinvitroderivedcysts.
Notetrophozoiteontheleft.Bar,10m.

immigratedAsiannationalsandfromhospitalandclinicpatientspresentingwithdiarrhea.Sampleswereidentifiedonlybyaccessionnumber.Fecalsmearswereair
dried,heatfixedandstoredat23CuntilfurtherprocessedbyIIF.CaninefecalsamplessubmittedbylocalveterinariansassuspectGiardiacasesweresimilarly
tested.
SurfaceWaterandSewageEffluentSampling
SurfacewatersusedforrecreationandasdrinkingwatersourcesweresampledusingamodifiedEnvironmentalProtectionAgency(EPA)filtrationtechnique(8).
FilterprocessingforGiardiacystdetectionwascarriedoutusingmodificationsoftechniquesdesignedtoisolateCryptosporidiumoocysts(C.E.Musial,Ph.D.
Dissertation,UniversityofArizona,Tucson,1985).Pelletsfromprocessedsamplesweresmearedonglassslides,heatfixed,andstainedforIIFusingMAbs.
Primary,secondary,andtertiarysewageeffluentscollectedinLima,Peruweresimilarlytreatedwiththeexceptionthatfilteredeffluentswereformalinfixedonsitefor
transportbacktotheUnitedStatesforprocessing.
Results
RatesofGiardiaencystmentrangingfrom1025%wereobservedduringtheconductofourexperiments.Invitrocystformationwasobservedattheendofthe48
hr.incubationat37Cundera90:10N2:CO2atmosphereinmediumwithoutbile.Maximumencystmentwasobservedfollowingafurther24hr.incubationat23C.
Prolongedincubationbeyond24hrs.at23Cdidnotresultinanincreasednumberofcysts.Phasecontrastmicroscopyshowedcystlikestructureswithdiscernible
walls(Figure1).Characteristicinternalcyststructures,i.e.,axonemesandmedianbody,werenotobserved.Scanningelectronmicroscopyofinvitrocultures
showedsmoothwalledcystformsdistinctfromtrophozoites(Figure2).Viabilityoftheinvitroderivedcystsasassessedbyinvivoandinvitrotechniqueswasnot
demonstrated.
Culturesupernatantsfromunclonedcelllines(6C1and2B3)andMAbfromascitesofaclonedlineinmice(2B3G6)werespecificforGiardiaanddidnotcross
reactwithEntamoebahistolytica,E.coli,E.hartmanni,Eimeriasp.,Cryptosporidium,Candida,Rhodotorula,oralgae.Peripheralcystfluorescencewas
observedusingboththeunclonedculturesupernatantsandtheclonedascitesfluidinIIFassays(Figure3).Inidenticalmicroscopicfields

Page221

Figure3.
Epifluorescentilluminationofcystsstained
withMAb2B3G6.Bar,10m.

viewedbyphasecontrastmicroscopyitwasoftendifficulttodiscernthecysts(Figure4).Monoclonalantibody2B3G6gaveanIIFtiterof100,000andwas
determinedtobeanIgG1subclass.Culturesupernatant6C1andMAb2B3G6wereusedtodetectGiardiain6%ofover900humanfecalsamplessubmittedto
ourlaboratoryduringthemonthofOctober,1986.AportionofthesamplespositivebyIIFforGiardia(16%)hadbeenreportedasnegativefollowingroutine
microscopicexamination(personnelcommunication,MaricopaCountyHealthDepartment).Allofthesepositiveswerereportedas1+infectionsbyIIF.Twocanine
fecalsampleswerealsopositiveforGiardiacystsusingIIF.Giardiacystsweredetectedinfourof72surfacewatersitessampledwithinArizona(5.5%)andthey
wereobservedinhighnumbersinsewageeffluentsfromPeruusingIIF.
Discussion
ExaminationoffecesbyconventionalmicroscopyiscurrentlythemostavailableandwidelyusedclinicaldiagnostictestfordetectingGiardiainfections.The
preparationrequiredforsuchanexaminationmaybequitetimeconsumingandinvolvethecollectionofmultiplefecalsampleswithsubsequentprocessingbyvarious
flotationtechniquestoconcentratecysts.Eventhen,cystdetectionmaynotreadilybeaccomplishedbecauseof

Figure4.
PhasecontrastimageofcystsinFigure3(arrows).Bar,10m.

technicianinexperienceinrecognizingGiardia,orinnotbeingabletodistinguishthemfromotherprotozoancystsorartifacts.Permanentstains,whichenhance
internalcystfeatures,arenotwidelyusedbecauseoftheadditionalexpenseandtimeinvolvedinspecimenpreparationandtechniciantraining.Examinationof
specimensobtainedbyproceduressuchastheEnteroTestandintestinalbiopsiesarelikelytoincreasethechancesofdetectinginfection,yetsufferfrombeing
invasive(13).Serologictesting,eitherbyimmunofluorescence(20,21)ortheELISA(21)hasthedisadvantageofnotalwaysbeingabletodetectactiveinfection
becauseofvariableantibodypersistance.TheCIEassaytodetectGiardiaantigensinfeces(5)dependsonsubjectiveanalysisandisnotreadilyavailableforclinical
use.OnlytheELISAforGiardiaantigendetection(8,19)appearstoovercomemanyoftheproblemsassociatedwiththeotherdetectionmethods.It,however,may
stillbecumbersomeinunskilledhandsbecauseoftheneedforcontinualstandardizationofthenumerousreagentsrequiredtoconductthetest.
ImmunofluorescentvisualizationofcystsusingMAbs,asreportedherein,offersahighlyspecificandsensitivemethodfordetectingGiardiainsymptomaticand
asymptomaticindividualsandinenvironmentalsamples.TheidentificationoflowlevelinfectionsusingIIFinindividualsreportedasnegativebyconventional
microscopicexaminationunderscoresthesensitivityofthistechnique.

Page222

Inaddition,fecalorenvironmentalslidespreparedforIIFinitiallycouldbescreenedusingalowpowerobjective(10)toidentifyfluorescingcysts.Thereafter,a
highermagnification(40)couldbeusedtoconfirmcystsshapeandsizeandcharacteristicinternalstructuresafterswitchingtophasecontrastmicroscopy.The
absenceofcrossreactivityoftheMAbswithothercommonlyencounteredfecalprotozoancysts,yeasts,oralgae,however,makesphasecontrastconfirmationof
internalstructuresunnecessary.PolyclonalantiserapreparedagainstGiardiacysts(14,16)andCryptosporidiumoocysts(18)havesimilarlybeenusedtoidentify
theseinfectiousagentsfromindividualsandenvironmentalsamples.Suchsera,however,arenoteasilyproducedinlargequantityandoftenlackthespecificityof
MAbs.
TheuseofinvitroderivedcyststogenerateMAbsrecognizingepitopesofhumanandanimalGiardiacystsindicatesthatthetwoformssharecommondeterminants.
FurtherconfirmationofidentitybetweenthesetwocystformscomesfromtheobservationofidenticalimmunofluorescentstainingpatternsfollowinguseoftheIIF
assay.Unfortunately,itappearsthattheinvitroderivedcystswerenotviable.Thefactorsresponsibleforinducingencystationremainunknowndespiteoursuccess
atgetting1025%oftrophozoitesgrowinginvitrotoencyst.
Overall,weconsidertheimmunofluorescentapproachtotheidentificationofGiardiacystsfromfecalorenvironmentalsamplestobesuperiortoothermicroscopic
methodsofdetection.ThesuperiorityofthistechniqueemployingtheuseofMAbsasprimaryantibodiesintheIIFassayhaspreviouslybeendemonstratedina
clinicalstudy(7).Becausecystsfluorescesobrightly,evenwhenviewedwitha10objective,onlyafewneedbepresentonaslidefordetection.Undermost
circumstancesthiswouldvirtuallyeliminatetheneedfortheuseoftimeconsumingflotationtechniquesrequiredtoconcentratecysts.Enhanceddetectionwouldalso
makeiteasiertoidentifyasymptomaticcystexcretorsorpatientswithvariablecystexcretionpatterns.Likewise,useofthistechniquewillmakeiteasiertodetectthe
fewcyststhatarelikelytobefoundinpositiveenvironmentalsamples.Finally,becausetheMAbweareusingisanIgG1,wemayassume,basedonpastexperience
(17),thattheimmunofluorescentassaycaneasilybeconvertedtoadirectprocedure.Thiswillservetoshortenproceduraltimeconsiderably.
Acknowledgements
ThisinvestigationwassupportedinpartbyArizonaDiseaseControlResearchCommissioncontract827700000011AQ6622,USDAAnimalHealthand
DiseaseResearchAwardARZT360458A0204andThrasherResearchFund27985.TheauthorsaregratefultoShanAnneEdwards,HumbertoMena,Staci
MatlockMena,ShannathMerbs,andLisaShubitzfortheirtechnicalassistance.TheauthorsarealsogratefultoMarilynMarshallforprojectenvironmentalsampling
coordinationandJimToppingoftheMaricopaCountyHealthDepartmentforpatientfecalsamplesanddata.
LiteratureCited
1.Bhatia,V.N.,andD.C.Warhurst.1981.HatchingandsubsequentcultivationofGiardiaintestinalsinDiamond'smedium.J.Trop.Med.Hyg.84:45.
2.Cohen,A.L..1974.Criticalpointdrying.In:PrinciplesandTechniquesofScanningElectronMicroscopy,Vol.I.M.A.Hayat(ed.),VanNostrandReinhold
Co.,NewYork.
3.Coons,A.H.,Creech,H.J.,andR.N.Jones.1941.Immunologicalpropertiesofanantibodycontainingafluorescentgroup.Proc.Soc.Exp.Biol.47:200202
4.Craft,J.C..1982.Giardiaandgiardiasisinchildhood.Ped.Inf.Dis.1:196211.
5.Craft,J.C.,andJ.D.Nelson.1982.Diagnosisofgiardiasisbycounterimmunoelectrophoresisoffeces.J.Inf.Dis.145:499504.
6.FazekasDeSt.Groth,S.,andD.Scheidegger.1980.Productionofmonoclonalantibodies:strategyandtactics.J.Immunol.Methods.35:121.
7.Garcia,L.S.,Brewer,T.C.andD.A.Brnckner.1987.FluorescentdetectionofCryptosporidiumoocystsinhumanfecalspecimensusingmonoclonalantibodies.
J.Clin.Microbiol.25:119121.
8.Green,E.L.,Miles,M.A.,andD.C.Warhurst.1985.ImmunodiagnosticdetectionofGiardiaantigeninfaecesbyarapidvisualenzymelinkedimmunosorbent
assay.TheLancetii:691693.
9.Jakubowski,W.1984.DetectionofGiardiacystsindrinkingwater:Stateoftheart.In:GiardiaandGiardiasis.S.L.ErlandsenandE.A.Meyers(eds).Plenum
Press,NewYork.pp.263286.
10.Keister,D.B..1983.AxeniccultureofGiardialambliainTYIS33mediumsupplementedwithbile.Trans.Roy.Soc.Trop.Med.Hyg.77:487488.
11.Kuvin,S.F.,Tobie,J.E.,Evans,C.B.,Coatney,G.R.,andP.G.Contacos.1962.Antibodyproductioninhumanmalariaasdeterminedbythefluorescence
antibodytechnique.Science.135:11301131.
12.McDowell,E.M.1978.Fixationandprocessing.In:DiagnosticElectronMicroscopy.B.F.TrumpandR.J.Jones(eds.).JohnWileyandSons.NewYork.
13.Pickering,L.K..1985.Problemsindiagnosingandmanaginggiardiasis.Ped.Inf.Dis.4:s6s10.
14.Riggs,J.L.,Dupuis,K.W.,Nakamura,K.,andD.P.Spath.1983.DetectionofGiardialambliabyimmunofluorescence.Appl.Environ.Microbiol.45:698700.
15.Rosenthal,P.&W.M.Liebman.1980.Comparativestudyofstoolexaminations,duodenalaspiration,&pediatricEnterotestforgiardiasisinchildren.J.Pediatr.
96:27879.
16.Sorenson,S.K.,Riggs,J.L.,Dileanis,P.D.,andT.J.Suk.1986.Isolation&detectionofGiardiacystsfromwaterusingdirectimmunofluorescence.WaterRes.
Bull.22:843845.
17.Sterling,C.R.,andM.J.Arrowood.1986.DetectionofCryptosporidiumsp.infectionsusingadirectimmunofluorescentassay.Ped.Inf.Dis.5:s139s142.
18.Stibbs,H.H.,andJ.E.Ongerth.1986.ImmunofluorescentdetectionofCryptosporidiumoocystsinfecalsmears.J.Clin.Microbiol.24:517521.
19.Unger,B.L.B.,Yolken,R.H.,Nash,T.E.,andT.C.Quinn.1984.EnzymelinkedimmunosorbentassayforthedetectionofGiardialambliainfecalspecimens.
J.Inf.Dis.149:9097.
20.Visvesvara,G.S.,Smith,P.D.,Healy,G.R.,&W.R.Brown.1980.AnimmunofluorescencetesttodetectserumantibodiestoGiardialamblia.Ann.Intern.
Med.93:802805.
21.Wittner,M.,Maayan,S.,Farrer,W.,andH.B.Tanowitz.1983.Diagnosisofgiardiasisbytwomethods:Immunofluorescenceandenzymelinkedimmunosorbent
assay.Arch.Pathol.Lab.Med.107:524527.
22.Wolfe,M.S.1979.Giardiasis.Pediatr.Clin.NorthAm.26:295303.

Page223

RoutineMonitoringofWatershedsforGiardiaCystsinNortheasternPennsylvania
SallyA.M.McFarlane
WaterQualityLaboratory(No.35114),PennsylvaniaGasandWaterCompany,135JeffersonAvenue,Scranton,Pennsylvania,185031799U.S.A.
PennsylvaniaGasandWaterCompany,alargeinvestorownedutilitycompanylocatedinNortheasternPennsylvania,wasfacedwithanoutbreakofGiardiasisinDecember,1983.
Atthattime,tworeservoirsweredeemedcontaminatedandboiladvisorieswereissuedthataffected250,000customers.Sincethattime,aninhouseGiardiatestinglaboratorywas
incorporated.RoutinemonitoringoftwentysevenreservoirsintheCompany'sdistributionsystemwasinitiatedinJanuaryof1985.Laboratorypersonnelobtainedtrainingfrom
Dr.CharlesP.Hibler,ColoradoStateUniversity,FortCollins,Colorado.TheCompany'sGiardialaboratorypresentlyperformsboththezincflotationorreferencemethod(1)and
theindirectfluorescentantibodytechnique.Todate,approximately1300filtershavebeenanalyzedonbothrawanddistributionwatersforourreservoirsaswellascontractwork.
WearecurrentlyrecognizedbythePennsylvaniaStateDepartmentofEnvironmentResourcesasoneofthefew"accepted"laboratoriesfordetectionofwaterborneGiardia.We
believeourexperiencesanddatawillaidotherlaboratoriesandresearchersinunderstandingcomplexitiesfacinglaboratoriesthatmayberequiredtoroutinelymonitorfor
Giardia.

Introduction
MonitoringtheWatershedEnvironment:BacteriologicalandChemicalIndicatorsofPossibleSewageContamination
AftertheGiardiasisoutbreaks,safeguardswereimplementedbyTheCompanytoassuretheothertwentyfivereservoirsservicingthe1700milesofpipeline
distributionwouldnotbeaffected.However,onlytworeservoirshavecompletefiltrationandutilizepostchlorinationasthemeansofdisinfection.Theremainderrely
solelyonchlorinationtoremoveorinactivatepotentiallyharmfulmicroorganisms.Withwildlifeabundantthroughoutthe265squaremilesofwatershed,cyst
multiplicationthroughcrosstransmission(2)becameaconstantthreat.ElaboratemeasurestakenbytheCompanyaftertheoutbreakstoprotectandmonitorthe
watershedincludedthefollowing:
1.Wildlifecontrolwasincreasedthroughpatrols,establishmentofbeaverandmuskrattrappingprograms,andremovaloffeedstocktodiscouragebeaverhabitation.
2.Hiredadditionallaboratorypersonneltomonitorforbacteriologicalandchemicalindicatorsofpossiblesewagecontaminationupstreamaswellasroutine
monitoringofallreservoirsforGiardiacysts(4).
3.Constructeda2.4milepipelinefromNesbittReservoirtobypassSpringBrookReservoiraftercontaminationwasfound.Thisincreasedthechlorinecontacttime
tolevelsthatwouldinactivatecysts(6).
4.Selectiveuseofreservoirs,wherepossible,toreducetheriskofGiardiasis.Newtreatmentfacilitieswerebuilttoincreasecontacttimesatreservoirsfoundtobe
contaminated.
5.Beganconstructionofthreenewwaterfiltrationplantsatanestimatedcostof52milliondollars.
InJanuaryof1984,anextensivemonitoringprogramof60reservoirinlets(performedmonthly)and48tributaries(performedquarterly)forvariousbacteriological
andchemicalparameterscommenced.
Ifthetestsandconsensusofthedataaccumulatedfromanyofthe108monitoredpointsarefoundtobesuspicious,itisreportedtothePennsylvaniaState
DepartmentofEnvironmentalResources.(Table1).
Afteranalyzingtheaccumulateddataforthepastthreeyears,weconcludedthatoutof108sites,only8tributariesindicatedhighreadingsofbacterialcountsand
somechemicalparametersanalyzed.However,ofthe27reservoirsmonitoredmonthlyforGiardiacysts,22haveatvariousinstancesbeenfoundharboringcysts.
ThosefoundtobecontaminatedarelistedinTable2.
TABLE1.Monitoringfrom1984,1985,and1986.

TestsPerformed

Ranges

HeterotrophicPlateCount

20to2,000,000colonies/1.0mL

MFProcedureTotalColiform

20to2,000,000colonies/100mL

MFProcedureFecalColiform

<1to190,000colonies/100mL

MFProcedureStreptococcal

<1to20,000colonies/100mL

Color

<5to165

Turbidity

0.25to60.0NTU

pH

4.1to7.7

Temperature

1Cto24C(seasonal)

NitrogenasNitrate(NO3N)

<0.01to3.00mg/L

NitrogenasAmmonia(NH3N)

<0.10to0.50mg/L

BiochemicalOxygenDemand

<1.0to13.00mg/L

TotalPhosphate

<0.01to3.00mg/L

TotalSuspendedSolids

<1.0to150mg/L

Detergents(MBS)

<0.01to0.02mg/L

Iron(Soluble)

<0.01to1.00mg/L

Manganese(Soluble)

<0.01to0.50mg/L

Page224
TABLE2.ContaminatedSites.
BraceBrook(+)

LakeScranton

Olyphant

Ceasetown(*)

LaRue(+)

PineRun

Edgerton

LaurelRun(+)

PlymouthRelief

Elmhurst(*)(*)

LaurelRunNo.2(+)

SpringBrook(*)(*)(+)

Fallbrook

MillCreek

Stillwater(F)

GardnerCreek

Nesbitt(*)

Watres(*)

Huntsville(*)(F)

No.5Dam(+)

WhiteOak(+)

(*)Reservoirwithtributaryshowinghigherthannormalreadings.
(F)Reservoirwithfiltration.
(+)Reservoirtakenoutofservice.
ImportantPoints:
1.DataindicatesNOthreatofpossiblesewagecontaminationat15ofthe22reservoirsknowntoharborcysts.
2.Monitoringintwocaseswasindicativeofpossiblecontamination.ElmhurstandSpringBrookarethetworeservoirs
whichwereinvolvedintheGiardiasisoutbreaks.

ProceduresUsedinMonitoringIndirectFluorescentAntibodyvs.ZincFlotation
Asyouarewellaware,detectionofwaterborneGiardiaisquiteinvolvedandmuchanart(5).OurlaboratorywasinitiallytrainedbyDr.CharlesP.HiblerinJuly,
1984.Aftersixmonthsofqualitycontrolsamples,weacquiredenoughexperienceandweredeemed"accepted"bythePennsylvaniaStateDepartmentof
EnvironmentalResources.WiththeCompanyconsciousofGiardiainsomanyofitsreservoirs,aswellastheupgradingoftheSafeDrinkingWaterAct(8),training
stateoftheartproceduresisessential.AfteradditionaltrainingatColoradoStateUniversityinJune1986,theindirectfluorescentantibodytechniquewas
incorporatedinthemonitoringlaboratory(7).Followingareoursamplingprocedures,zincflotation,andindirectfluorescencetechniques.Wewillelaborateon
problemsassociatedwiththemboth.
Methods
SamplingProcedure
Apparatus:Hose,hoseadapter,filterhousing,flowmeter,filtertype1micronpolypropyleneCunoDPPY(1).Location:Waterpressure3035psiregulator15
psi,creek/reservoirusegasolinewaterpump.Approximately400gallonsoverafewhoursdependentonalgaegrowth.Maximum1000gallons.
ZincFlotationProcedure
1.Processfiltersin4Lofdeionizedwater,refrigeratefor24hours.
2.Aspiratetopelletnotevolumeoffiltrate.
3.Underlaywith20mL1.28(sp.g.)ZnSO4with20mLs.offiltrate.
4.Centrifuge58minutes.1500RPM.
5.Takeapplicatorstick,breakinterface,aspirateonto5mNucleporemembrane.
6.Rinsewithdeionizedwater,pourinto50mLtube.
7.Repeatstep4.
8.Aspirateto6mLs.,transferinto15mLtube.
9.Repeatstep4.
10.Aspiratetopellet,add3dropsLugol'siodine,filltubewith1.20(sp.g.)ZnSO4,addglasscoverslip.
11.Repeatstep4.
12.Placecoversliponcleanglassslide,scanat10X.Verifycystswith40Xor100X.
Noteciliates,plantdebris,totalnumberofcysts.Aminimumof2slidesperfiltershouldbeanalyzed.Ifnocystsarepresent,reportasNONEDETECTABLE.If
cystsarepresent,polaroidpicturesaretaken.WecategorizecystsbyappearanceperorderofthePennsylvaniaStateDepartmentofEnvironmentalResources.When
consideringissuingboiladvisories,theyarbitrarilychoseaminimumcontacttimeof1hourat4.0mg/Lfreechlorineasaworkingdosageforcystinactivation.
Understandably,categorizingcystqualityissolelydependentontheanalysts'experienceandjudgement(Table3).
IndirectFluorescentAntibodyTechnique
1.AcquireoptimumdilutionsofrabbitantiserumandgoatantirabbitFITCconjugatepurchasedfromColoradoStateuniversityandMilesScientific,Inc.,respectively.
2.Underlay10mLoffiltratewith20mL1.28(sp.g.)ZnSO4.
3.Centrifuge58minutesat1500RPM.
4.Aspirateonto25mm5mNucleporemembraneinGelmanfilterholder.
5.Aspirate200lofultracleanGiardiacystsontoapositivecontrolmembrane.
6.Apply2incrementsof125mdilutedantiserumontomembranes.
7.Incubateat37Cfor30minutes.
8.Cleanmembraneswithapproximately600mLofPBS(7.07.2).
9.Repeatsteps6,7,and8usingdilutedconjugate.
10.Oncleanglassslide,placeonedrop4%glycerolsolution.Placemembraneonglassslide,adddropofglycerolsolution.Placeglasscoverslipontop,removeall
airbubbles.Sealwithclearfingernailpolish.Examineslideimmediatelyat20Xwithmicroscopewithfluorescentapparatus.Giardiacystsappeargreenapplein
color,rangingfrom815m.Insomeinstances,detectionofsomeinternalstructuresispossible.
Discussion
AsofAugustof1986,ourlaboratoryperformsbothtechniquesonsamplesthatarefoundtobepositiveorsuspiciouswhenroutinelymonitoringwithzincsulfate
(Table4).Eachtechniquehasbothadvantagesanddisadvantages(Table5).
TABLE3.CategorizationofCysts.
Excellent

cystclearlyrevealsthreeinnerstructures
cystswallwellintact
colorisbrightandveryrefractive

2.

Good

cystrevealstwoorthreeinnerstructures
cystwallpulledslightlyfromoutermembrane
colorisrefractive

3.

Fair

cystrevealstwoorthreeinnerstructures
cystwallpulledawayfromoutermembrane
colorislessrefractiveindicativeofanoldercyst

4.

Poor

cystrevealsoneortwoinnerstructures
cystwalliswrinkled
coloriseitherdarkbrownorgreen
refractivenessisminor

5.

HardlyRecognizable

cystrevealsoneinnerstructure
cystwalliscollapsed
coloriseitherbrownorgreen
refractivenessisslight

1.

Page225
TABLE4.

Sample
Date

NumberCystsZnSO4
20mLCondition

NumberCystsF.A.
10mLFluorescence

LaurelRunRaw
81286

NoneDetectable

1/10(yes)

HuntsvilleRaw
81386

3/20
(FairExcellent)

8/10(yes)

LaurelRunRaw
91686

1/20(Poor)

BackgroundFluorescence
MonitoringImpossible

FallbrookRaw
92086

NoneDetectable

1/10(yes)

Q.C.SampleC.S.U.
92386

4/20(Poor)

9/10(yes)

EdgertonRaw
10986

8/20(Excellent)

4/10(yes)

HuntsvilleRaw
101486

3/20(Excellent)

11/10(yes)

WatresDistribution
101786

1/20(Fair)

3/10(yes)

WatresDistribution
102186

NoneDetectable

2/10(yes)

WatresRaw
102886

2/20(Poor)

8/10(yes)

FallbrookRaw
102986

NoneDetectable

13/10(yes)

EdgertonRaw
111086

2/20(Good)

4/10(yes)

ZincSulfate

IndirectFluorescence

12Hours

45Hours

TABLE5.AdvantagesandDisadvantages.

1.AnalysisTime
2.DetectionofCystQuality
3.Expense
4.CystVerification
5.Detectability
6.MicroscopeFatigue
7.BackgroundAlgaeGrowth

Yes

No

$50.00/Filter

$175.00/Filter

Yes

NotAlways

<IndirectFluorescence

>ZincSulfate

Minimal

Excessive

CanbeEliminated

CannotbeEliminated

Whenroutinelymonitoringreservoirs,thezincsulfateprocedureismoretimeandcostefficient.Thefluorescentprocedureiseffectiveinassuringagainstcontamination
andsafeguardingoutbreaksituations.
Acknowledgements
ParticularthanksisgiventoDr.CharlesP.HiblerandhisstaffatColoradoStateUniversityfortheirexpertiseanddedicationintrainingpersonnelandextendingtheir
generosityandhospitalityoverthepastfewyears.Additionalthankstothelaboratorypersonnelwhoallplayedanimportantroleinpreparingthisreport:Karen
Caparo,RichardUngvarsky,RobertNotartomaso,LynneRogers,SuzanneSwartz.
LiteratureCited
1.APHAAWWAWPCF.1985.StandardMethodsfortheExaminationofWaterandWastewater,16thEd.AmericanPublicHealthAssociation,Washington
D.C..
2.Davies,R.B.,Fukutaki,K.,andC.P.Hibler.1983.CrosstransmissionofGiardia.EPA600/S182013.
3.Jakubowski,W..1981.DetectionofGiardiacystsindrinkingwater:stateoftheart.In:GiardiaandGiardiasis.S.L.ErlandsenandE.A.Meyer(eds).Plenum
Press,NewYork.pp.263286.
4.Jakubowski,W..1984.AreferencemethodfordetectingGiardiacystsinwater.In:AdvancesinWaterAnalysisandTreatment.(12thAnnualAWWA
WQTC1985),AWWA.pp.7378.
5.Jarroll,E.L.,Bingham,A.K.,andE.A.Meyer.1981.EffectofchlorineonGiardialambliacystviability.Appl.Environ.Microbiol.41:483487.
6.Sauch,J.A.1984.DetectionandidentificationofGiardiacystsusingimmunofluorescenceandphasecontrastmicroscopy.In:AdvancesinWaterAnalysisand
Treatment.(12thAnnualAWWAWQTC1985),AWWA.pp.7986.
7.Thompson,J.C..1986.Updatingthesafedrinkingwateractandthedrinkingwaterregulations.WaterEngineeringandManagement.pp.2124.

Page227

WaterborneGiardiasis:SourcesofGiardiaCystsandEvidencePertainingtotheirImplicationinHumanInfection
StanleyL.Erlandsen*andWilliamJ.Bemrick
DepartmentofCellBiologyandNeuroanatomy,4135JacksonHall,UniversityofMinnesotaSchoolofMedicine,Minneapolis,Minnesota55455,U.S.A..
Insomewaterborneoutbreaksofgiardiasis,thecontaminationofwaterwithGiardiacystshasbeenattributedtothepresenceofbeaverslivinginthewatershedandinreservoirs.
Ourstudiesofaquaticmammals,atsiteshavingexperiencedoutbreaksofwaterbornegiardiasisin5NewEnglandstates,revealedthatabout17%oflivetrappedbeaversand
100%oflivetrappedmuskratswereinfectedwithGiardia.CrossspeciestransmissionstudiesofhumanoriginGiardia,intoeitherbeaversormuskrats,weresuccessfulin
establishingGiardiainfections.DespitethehighincidenceofnaturalinfectionintheseaquaticmammalsandtheirsusceptibilitytoinfectionwithhumanGiardia,itispossiblethat
manyoutbreaksofgiardiasismayhavebeencausedbyGiardiacystsfromothersources,because1)Giardiacystsfrommuskratsandotherrodents,suchasvoles,hada
morphologicallydistinctbinaryappearance(contained2maturetrophozoites)entirelydifferentfromhumancystsandwerealsoimmunologicallydissimilar.2)Trophozoites
isolatedfromlivetrappedbeaverspossessedamorphologicalcharacteristic,ashortpairofcaudalflagella,distinctlydifferentfromhumanGiardia.3)Wadingmigratorybirds
sharingthesamehabitatofaquaticmammalswerefoundtocontainGiardia.4)Thepotentialforhumanwastecontaminationexistedatallwaterborneoutbreaksites,butinvarying
degrees.Finally,noapparentalterationinaquaticmammalpopulationshasoccurredatthesesites,yetnositehasexperiencedasecondoutbreakofgiardiasis.Thus,Giardiacysts
inwatermaybederivedfrommultiplesources.

Introduction
TheproliferationandinteractionoftheintestinalprotozoanGiardiawithinthegastrointestinaltractmucosaofbothmanandanimalsleadstosignsandsymptomsof
intestinaldysfunctionknownasgiardiasis.Theincidenceofgiardiasishasreachedendemicproportionsthroughouttheworldandtodaythisdiseaseisoneofthe
leadingcausesofwaterborneepidemicdiseaseintheUnitedStates(67).
Giardiacystspassedbyanimals,includingman,havebeenshowntoremainviableformonthsat410Cinwater(38).Watercontaminatedbyanimalorhumanfecal
wastemayserveasasourceforinfectingcampers,hikers,orevenlargesegmentsofthepopulationthatderivedrinkingwaterfromacontaminatedwatershed(9).The
sourcesofwaterborneGiardiacystswhichinfectmanmaybeofhumanorigin,butithasbeensuggestedthatsomeanimalsmayactasreservoirs(13,22)andover40
differentanimalspecieshavebeenreportedtoharborthisprotozoan(31).AnimalssuggestedasbeingreservoirsforGiardialambliafrommanincludedrodents,
dogs,cats,cattle,muskrats,andbeavers,withthelatterbeingconsideredthemostfrequentsuspectinwaterborneoutbreaksofgiardiasis(13,22,37,38).
Discussion
AnimalsImplicatedinWaterborneGiardiasis
InwaterborneoutbreaksofgiardiasisthemostlogicalanimalstobeconsideredaspotentialreservoirsforGiardiacyststhatproducehumaninfectionwouldinclude
notonlyaquaticmammalssuchasbeavers,butalsomuskrats,domesticanimals,birdsandman.Eachofthesepotentialanimalsourceswillbediscussedintermsof
theevidencefortheirimplication.
Beavers
Withinthelastdecadeseveralsuggestionsintheliteraturehaveimplicatedthebeaverinoutbreaksofwaterbornegiardiasis.First,epidemiologicalstudiesindicatedthat
individualsconsumingmountainwaterinUtahandColoradowereathighriskforinfectionwithGiardia(3,69).Secondly,anoutbreakofgiardiasisinCamas,WA
waslinkedtothepresenceofbeaversinthewatershedanditwasreportedthatcystsisolatedfrombeaverfeceswerecapableofinducinggiardiasisinspecific
pathogenfreebeagles(15).Also,itwasreportedbyDaviesandHibler(13),thattwoofthreehumansingestingGiardiacystsisolatedfrombeaversbecameinfected
anddisplayedGiardiacystsintheirstools.Subsequently,inotherwaterborneoutbreaksofgiardiasis,thecystsofGiardiawereeitherdetectedwithinsamplesof
municipalwaterobtainedfrombeaverinhabitedwatershedsorwithinsamplesofbeaverfecesortissues(Table1).Despitetheattentionbeingfocusedonthebeaver
asa''possible"reservoirforinfectingmanwithgiardiasis,itwasnotatallclearthatthebeavershouldbeconsideredguilty(4).ExaminationofTable1hasrevealed
thatonlyalimitednumberofbeaversinwatershedsthatsuppliedwaterto
*Correspondingauthor.

Page228
TABLE1.RelationshipbetweenWaterborneOutbreaksofGiardiasisandNaturallyInfectedAnimals
Waterborne
Giardiasis
OutbreakSite*
City,State

GroundFecal
Collection
Beaver

Muskrat

LivetrappedAnimals

KilltrappedAnimals
CarcassFecal
Sample
Beaver

Muskrat

ColonicFecal
Sample
Beaver

Muskrat

SmallIntestine
Smear/Histology
Beaver

Muskrat

GiardiaCysts
inMunicip:
WaterSupply

Camas,WA

2/3+

1/1+

EstesPark,CO

0/10+

0/7+

Berlin,NH

1/4+(infected,not
shedding)

ZigZag,OR

+
(#
unknown)

Government
Camp,OR

+(#
unknown)

Graeagle,CA

2/3+

Bradford,PA

1/1+

Reno,NV

1/1+

DoverFoxcroft,
ME

1/2+

1/2+

Pittsfield,MA

1/9+

2/7+

7/7+

*basedondatafromDykesetal.(1980),Lippy(1979),Keifer(1980),andunpublisheddata.

humanpopulationshavingexperiencedgiardiasisoutbreaksweretestedforthepresenceofGiardia.Otheranimalspecies,includingmuskrats,birdsandmanwere
almosttotallyignoredaspotentialsourcesofcysts.AlthoughthedetectionofGiardiacystsinthemunicipalwatersuppliesattheoutbreaksites(Table1)appearedto
correlatewiththepresenceofGiardiainthefewbeaversexamined,inretrospect,thelackofasystematicsurveyofotheranimalshasmadethiscorrelation
inconclusive.
AssessmentofthenaturalprevalenceofGiardiainbeavershasbeencarriedoutinfivedistinctlydifferentgeographicallocalesasshowninTable2.Aclose
correlationwasseenbetweentheprevalenceofGiardiainfecalsamplesfromkilltrappedbeaversfromWashingtonandNewEnglandwhereapproximately11%of
thesampleswerepositiveforcysts.Ontheotherhand,theprevalenceofGiardiawasseentobesomewhathigherinnecropsystudies(17%)basedonthedetection
oftrophozoiteswithinintestinalscrapingsfromlivetrappedbeavers.Analysisoffecalsamplesfromthesesamebeavers(notshown)resultedinaslightlylower
prevalence(14%),indicatingthattheexaminationofintestinalscrapingsatnecropsywasamoresensitivemethodfordeterminingGiardiaprevalenceinabeaver
population.
TABLE2.NaturalPrevalenceofGiardiainBeaversandMuskrats.
Intestinal**Trophozoites

CystsinFecalSamples

GeographicalLocation

Beavers

Muskrats

Beavers

Muskrats

1.NewEngland*
(ME,NH,NY,VT)

17%
(N=171)

100%
(N=97)

11.7%
(N=369)

33.1%
(N=432)

2.Minnesota*

10.1%
(N=89)

100%
(N=49)

3.Colorado(13)

18%
(N=244)

0%
(N=21)

4.Washington(22)

10.7%
(N=529)

42%
(N=133)

5.BritishColumbia(26)

14.7%
(N=299)

40%
(N=20)

*ErlandsenandBemrick,unpublishedobservations.
**Detectedbylightmicroscopicexaminationofintestinalscrapings.

Page229

Figures12.
ScanningelectronmicrographsofaGiardiatrophozoiteisolatedfromtheintestineofabeaverlivetrappedinVermont(figure1)andaGiardiatrophozoiteseenadherenttotheintestinal
mucosaobtainedfromamuskratlivetrappedinNewYork(figure2).Infigure1thetrophozoiteobtainedfromanindigenousinfectionhadashortpairofcaudalflagella(arrowheads),which
standoutinsharpcontrasttothelongcaudalflagella(arrowheads)seeninGiardiatrophozoitesobtainedfromeithermuskratintestine(figure2)orhumanorigin(notillustrated).Bar
equals2microns,figures1and2.

(i)GeographicDistributionofBeaversversusSitesofGiardiasisOutbreaks
ThebeaverhasbeenshowntoinhabitmostoftheNorthAmericancontinentfromAlaskaandCanadatotheMexicanborder,exceptinWesternCaliforniaandthe
SoutheasternUnitedStates,includingtheGulfcoast(44).Duetoitsextensivegeographicaldistribution,thephysicalpresenceofbeaverswouldappeartocoincide
closelywithknownsitesofgiardiasisoutbreaks,however,waterborneoutbreaksofhumangiardiasishaveoccurredinsiteswherebeaverswerenonexistent
(14,28,36)andGiardiacystshavebeendetectedinwateruninhabitedbybeaversusedfordrinkingpurposes(C.P.Hibler,personalcommunication).Thus,themere
physicalpresenceofbeaversinareservoirorwatershedexperiencingagiardiasisoutbreakmaynotnecessarilyhaveanyepidemiologicalsignificanceregarding
zoonotictransmission.
(ii)AretheGiardiainBeaverstheSameAsThoseinMan?
TheassumptionhasbeenmadeinpreviousstudiesimplicatingbeaversinwaterborneoutbreaksofgiardiasisthattheGiardiaorganismsfoundwithinthebeaverwere
identicaltothoseproducingdiseaseinman.StatementshavebeenmadethatbeaverGiardia,presumablycysts,weremorphologicallyidenticaltoGiardiafrom
infectedhumans(15,35)butnoevidencewaspresentedtosubstantiatetheseclaims,andthesefindingshavebeenquestioned(4).However,evidencehasbeen
obtainedsuggestingmorphologicaldissimilaritybetweenGiardiatrophozoitesofbeaverandhumanorigin.Ourlightandscanningelectronmicroscopicstudiesof
GiardiatrophozoitesrecoveredfromnaturallyinfectedbeaversinNewEnglandandMinnesotahaverevealedatypeofGiardiatrophozoitethatwasmorphologically
distinctfromGiardiaofhumanorigin.Giardiatrophozoitesfromnaturallyinfectedbeaverspossessedashortpairofcaudalflagella(3.32.5mFigure1)andwas
thepredominantform,beingfoundinnineofeleven(81%)infectedanimalsstudied(Table3).ThelengthofcaudalflagellafoundinGiardiaofhumanoriginwas9.1
2.9m.AcomparisonofthelengthofcaudalflagellainbeaverGiardiawiththosefromothersourcesclearlydemonstratedthatGiardiafromindigenousinfectionsin
beavers,andoneofthetwoaxenicallyculturedGiardiastrainsfrombeavers(IP0482:1),weremorphologicallydistinctinthattheircaudalflagellaweresignificantly
shorter(p=<0.05)thanthoseofotherGiardiaderivedfrom1)man,2)waterbornecystsofunknownoriginfromagiardiasisoutbreakinPittsfield,MA,3)indigenous
infectionsinmuskrats,and4)fromanaxenicbeavercultureobtainedfromanotherlaboratory(IP0583:1).Althoughitmayseeminconsistentthatthelatterbeaver
culture(IP0583:1)resembledtheflagellarlengthofhumanorigin,thisculturedbeaverisolatehasbeenshowninendonucleaserestrictionexperimentstohaveaDNA
bandingpatternidenticaltothatofhuman

Page230
TABLE3.DistributionofTrophozoitesfromBeaver,Muskrats,andMan,BasedonLengthofTailFlagella(um)
TROPHOZOITESOURCE

LENGTHOFTAILFLAGELLA(um)

(n)* <12345678910>
Human**(56)

xsd

12%

11%

21%

9%

20%

11%

16%

7.82.2

13%

7%

21%

9%

35%

9.12.3

Human**(60)

11%

Muskrat
Livetrapped(98)

1%

3%

12%

10%

12%

14%

11%

36%

8.62.4

Beaver
Livetrapped(56)

9%

9%

21%

14%

10%

13%

5%

6%

3.32.5

Beaveraxenic(50)
Culture#IP0482:1

2%

14%

16%

12%

32%

14%

4%

4%

2%

4.21.6

Beaveraxenic(50)
Culture#IP0583:1

6%

12%

6%

4%

14%

4%

6%

48%

10.54.5

*Numberofflagellameasured.
**TrophozoitesobtainedfromgerbilsinfectedwithGiardiacystsofhumanorigin(courtesyofDr.F.W.SchaefferIIIandMr.WaltJakubowski,U.S.Environmental
ProtectionAgency,Cincinnati,OH).
***TrophozoitesobtainedfrombeaversinfectedwithGiardiacystsofhumanoriginatUniversityofMinnesota.

originGiardia,whereasthebeaverculture(IP0482:1)possessingtheshortflagellarlength,similartothatfoundinbeaverswithindigenousinfections,hadaDNA
bandingpatternthatwasdistinctlydifferent(42).TheinterestingcorrelationbetweenDNAbandingpatternsandflagellarlengthsuggestedthattwodifferenttypesof
Giardiatrophozoitesmayhavebeenisolatedfrombeavers,oneresemblingtheGiardiaseeninmanwhiletheothertypeoftrophozoiteinbeaverspossessedshort
caudalflagellaandwasthepredominanttypefoundinindigenousinfections.
Inavarietyofprotozoa,includingChlamydomonas,Euglena,Ochromonas,Polytomella,andTetrahymena,thecontrolofthelengthofflagellaandciliamaybe
regulatedbyavarietyoffactors,particularlygeneticones,ratherthanbeingdeterminedsolelybytheavailabilityofassemblycompetentproteins(33).Thepresenceof
twodistinctpopulationsofGiardiatrophozoitesinbeavers,basedonflagellarlength,raisedinterestingquestionsastowhetherornotthisreflectedgeneticcontroland
therefore,representeddifferenttypes,perhapsspecies,ofGiardia.
ThepresenceofmultiplespeciesofGiardiawithinasinglehostisnotwithoutprecedencesinceratsappeartoharbornotonlyG.muris,butalsoasecondspecies,
G.simoni,thatcloselyresemblesthespeciesinman(20,32,58).Also,therecentdevelopmentofthemongoliangerbilmodelforGiardiainfectiondemonstratesthat
Giardiacystsderivedfromavarietyofhostsincludingmice,muskrats,beaversandman,cansuccessfullyinfecttheseanimals,andthattheGiardiatrophozoitesor
cystsrecoveredhavethesamemorphologicalfeaturesastheoriginalinoculum.However,itremainstobedeterminedwhetherornotflagellarlengthcanserveasa
markerfordifferenttypes(species)ofGiardiaandfuturedevelopmentsinthisfieldawaitthediscoveryofimmunologicprobesthatcandifferentiatethedifferent
subtypesofGiardiabelongingwithintheG.duodenalisclassificationofFilice(20).
(iii)InfectionofBeaverswithGiardiaCystsofHumanOrigin
BeaverscanbesuccessfullyinfectedwithGiardiacystsderivedfromhumanfecalsamples.IntheirreviewofanimalreservoirsforGiardia,DaviesandHibler(13)
reportedthattheyinfectedtwobeaversusinganinoculumof10,000humancystsderivedfromanunknowndonor(s),butnoexperimentaldetailswereprovided.
Recently,ourlaboratoryhasdescribedcrossspeciestransmissionexperimentsinwhich38beaverswereusedtoinvestigatetheinfectivityofG.lambliacysts
(Bemrick,

Page231

W.J.etal.1986).Crossspeciestransmissionofgiardiasis:infectionofbeaverswithhumanGiardialamblia,61stAnnualMeetingoftheAmericanSocietyof
Parasitologists,abstract96).Viablehumancystsfrommultipledonorswereobtainedbyfluorescencecellsortingusingfluorogenicdyes(55).SuccessfulGiardia
infectionswereobtainedintwelveoftwentybeaverschallengedwithoraldosesrangingfrom5102,5103,and5105cystsperanimal.
ThesuccessfulinfectionofbeaverswithhumanoriginGiardiaindicatethatitistheoreticallypossiblethatbeaversmayactasareservoirofhumancystsandpotentially
couldserveasavehicleforthespreadofwaterbornecysts.Todate,inourstudiesonlyaminorityofbeavers(17%ofthelivetrappedanimals)areknowntocontain
Giardia,andamajorityoftheseanimals(81%)appeartohaveGiardiatrophozoitesmorphologicallydifferentfromthoseseeninhumans.Ifthehumantypeof
Giardiacanbedetectedwithinbeavers,thepresenceofsuchaninfectionshouldserveasanimpetusfortheinvestigationofotherspeciesofwaterbirdsormammals
ascontributingsourcesofGiardiacysts,sincethepresenceofcystsinthewaterisprobablyrequiredfortheestablishmentofaninfectionwithinbeaversinthefirst
place.However,sincetheinfectionofbeaversrequirestheirexposuretohumantypecysts(endogenousbeaverGiardiamaybedifferent),thepresenceofthehuman
typeofGiardiaintheseanimalsmaybeanimportantindicatorofcontaminatedwater,derivedfromothersources,suchasrawwastewaterfromsewagespills,septic
systems,humanrecreationalusageofthewatershed,andpossibly,fromotherinfectedanimalsorbirds.
Muskrats.(i)AnimalDistributionandPrevalenceofGiardia
Themuskratsharesthesameaquaticenvironmentasthebeaver,buthasreceivedlittleattentionasapotentialsourceofGiardiacysts,eventhoughourdataindicates
theprevalenceofGiardiainfectioninthemuskrattobe100%inlivetrappedanimals(Table2).Comparedtothebeaver,populationestimatesofmuskratdensityin
awatershedaredifficulttodetermine(19).Signsofhabitationorfeedingactivityofmuskratsarenotasobviousasthatofbeavers.Manywatershedsappeartobe
capableofsupportingamuchlargerpopulationofmuskratsthanbeavers,becauseoftheformers'lesserimpactontheenvironmentintermsoffooddepletionandalso
duetotheirgreaterreproductivecapacitysinceonepairofmuskratscanyieldasmanyas100ormoreprogenyinoneyearversus46forbeavers.
AnalysisofGiardiacystsonapergrambasisrevealedestimatesrangingfrom7.4101to1.2105inmuskratfeces,whileupto1.4106[Erlandsenand
Bemrick,unpublishedobservations(60)]cystspergramweredetectedinbeaverfeces.ItwouldappearthatbeaverfecesmayhavecontainedmoreGiardiacysts
thanthatofmuskrats,however,noneofthesestudiestookintoaccountthedailyfecaloutputofeachanimal,norwashomogeneityoffecalcystdistribution
determined.
(ii)MorphologyandImmunoreactivity
MuskratsareinfectedwiththeG.duodenalisspecies(20)andthetrophozoitesaresimilar,ifnotidentical,toGiardiatrophozoitesofhumanorigin.Themedianbody
doeshavethetypicalclawhammershapethatcharacterizesthisspecies.However,thecystsofGiardiaisolatedfromnaturallyinfectedmuskratsarecompletely
differentfromhumancystsinthateachmaturecystcontainstwofullyformedtrophozoites,eachpossessingaformedadhesivediscandatotalcomplementofflagella.
Boeck(7)namedthesebinarycystsandtheyhavebeendetectedonlyinmicrotinerodents(volesandmuskrats).ThebinaryappearanceoftheGiardiacystfrom
muskratsisnothostdependentsincecrossspeciestransmissionofthesecyststoeithermongoliangerbilsormice,performedintwoseparatelaboratories,resultsin
excretionofGiardiacystswiththesamebinarymorphology(Erlandsen,S.L.,Bemrick,W.J.,Schaefer,F.W.III,andW.Jakubowski,unpublishedobservations).
Theconsistentfindingofthesetypesofcystsinmicrotinerodentsandtheirabsenceinotherspecies,includingman,stronglysuggeststhattheymaybeaseparatetype
orevenadifferentspeciesofGiardia.
ImmunologicdifferenceshavebeenobservedinGiardiacystsfrommuskrats(andothermicrotinerodents)whencomparedtoG.duodenaliscystsobtainedfrom
dogs,beavers,andman.Immunofluorescenttestingofantisera,raisedinourlaboratoryandobtainedfromothers(52,53)directedagainstGiardiacystsrevealed
immunostainingofallGiardiacyststestedincludingthosefrommouse,beaver,muskrat,dogandman.Oneantiserum(52)recognizedG.duodenaliscystsfromdog,
beaver,andman,butdidnotreactwithG.duodenaliscystsfrommicrotinerodents,includingthemuskrat,orwithG.muriscystsfrommice.Thisimmunologic
differenceclearlyindicatedthatnotallcystswithintheG.duodenalisspecies,basedontrophozoitemorphology,shouldbeconsideredidentical.Italsoprovided
additionalsupportfortheideathatbinarycystsmayhavebeenderivedfromaseparatespeciesofGiardia.
(iii)CrossSpeciesTransmissionofHumanorBeaverGiardiaCyststoMuskrats
StudiesinourlaboratoryonthecrossspeciestransmissionofGiardiacystsisolatedfromfecesobtainedfromeitherbeaversorhumansdemonstratethatbothtypes
ofcystscansuccessfullyestablishGiardiainfectionsinmuskrats.Theseresultssuggestthatitmaybefeasibleformuskratstoserveasavectorforthespreadof
human(orbeaver)cysts.Beforeanyconclusionscanbedrawnregardingthemuskratsinvolvementinhumangiardiasis,moredetailedinvestigationsofthe
morphologicalandimmunologicalpropertiesofGiardiacystscollectedfromnaturallyinfectedmuskratsneedtobeperformedtodeterminewhetherornotmuskrats
(orotheranimals)cansimultaneouslysupportmultipleinfectionsincludingGiardiaindigenoustomuskratsandthosederivedfromotherspecies.
DomesticAnimalsandBirds
Inthepast,anumberofstudieshaveindicatedthatGiardiaderivedfromavarietyofdomesticanimalhostswereimplicatedaspossiblezoonoticsourcesofhuman
giardiasis.TheexperimentsofPadchenkoandStolyarchuck(47),Shawetal.(56),andHewlettetal.(24)utilizedcanineGiardiacystsintheircrossspecies
transmissionexperiments.Their

Page232

experimentalmethodologyhasbeencriticized(4)andthesuggestionthatdogswereinvolvedinthetransmissionofGiardiatohumans,baseduponexistingdata,was
consideredinvalid.
CatshavebeenpostulatedtoplayaroleinzoonotictransmissionofGiardiabasedontheabilityofbothcatandhumanGiardiatoinfectgerbils(6).Thisinference,
derivedfromtheextrapolationofinfectivityinformationfromoneanimalspeciestoothers,mayhaveledtoerroneousconclusions.
Similarly,otherinferencesthatinvolvedthecatinthetransmissionofhumangiardiasishavebeenmade(5)butwerebasedoninconclusiveevidence.Wooand
Paterson(68)recentlyreportedaseriesofdetailedexperimentsoncrossspeciestransmissionthatinvolvedattemptstoinfectpuppies,kittens,andmicewithhuman
Giardiafromanaxenicculture.TheywereunabletoinfectanyoftheseanimalspecieswiththishumanGiardia.
Otheranimalsincludingcattle,sheep,andelk(13)havebeensuggestedasplayingsomeroleinthetransmissionofhumangiardiasis,butwithoutanydatatosupport
thiscontentionotherthanthedetectionofGiardiacystsobservedinfecalsamples.Thelackofanyinformationregardingthepresence/absenceofmorphologicor
immunologicsimilaritiesofGiardiaderivedfromthesehoofedanimalswithotherknownspeciesofGiardia,hasmadeitimpossibletoassesstheirroleinhuman
giardiasis.Itshouldbenotedthatnoneoftheseanimalshavebeenshowntohaveageographicaldistributionthatcanbecorrelatedwithamajorityofknowngiardiasis
outbreaksites,therefore,itseemedunlikelythattheirrolewasofgeneralsignificance,althoughthepossibilityofacontributoryroleatindividualsitescouldnotberuled
out.
SeveralspeciesofbirdshavebeenshowntoharborspeciesofG.duodenalis(2,31).Giardiasisoutbreakshavebeenreportedinbirds(48,54)andBox(8)has
suggestedthatthebudgerigarmaybeinvolvedinhumangiardiasis.Thelatterseemsunlikelysinceascanningelectronmicroscopystudy(18)ofGiardiapsittaci
trophozoitesfrombudgerigarshasshownthatthiswasaseparatespecies,morphologicallydistinctfromanyothertypeofGiardia,andthatthisspecieshasnotbeen
reportedinultrastructuralstudiesofGiardiainanimals(16,17,21,25,45,46)orman(41,51,57,63)Arecentreport(23)ofGiardiainagreatblueheron,suggested
thatthisspeciesofbirdshouldbeconsideredasapotentialsourceofGiardiacystsinwaterbornegiardiasis.Wehavecorroboratedthisfinding,andalsohavefound
Giardiaingreenheronsandegrets.Allofthesewadingbirdshavetrophozoiteswithaclawhammershapemedianbodyacharacteristicsharedwiththetypeof
Giardiafoundinmanandclassifiedasthespecies,G.duodenalis(20).Wadingbirds,suchasheronsandegrets,sharethesamehabitatasbeaversandmuskrats,
therefore,theyshouldbegiventhesameconsiderationasbeavers,muskrats,andman,aspotentialvectorsforthespreadofwaterbornegiardiasis.Thisisespecially
true,sincetheyhavethecapabilityofmovingfromonewatershedtoanotherovergreatdistancesandthuscaneasilyestablishnewfociofinfectionswithoutthe
physicallimitationsoftravelimposedonmammalsbytheterrain.However,innorthernstateswadingbirdsmayonlybetemporaryorseasonalinhabitantsofthe
watershed,andnoinformationhasbeenreportedontheprevalencewithinanyavianspecies.
Man.(i)PrevalenceofGiardiainManandSewage
GiardialambliahasbeenconsideredthemostcommonhumanintestinalparasiteintheUnitedStates,beingreportedinonesurveyin3.8%offecalsamples
examined(40).Aprevalencerateof26%hasbeenreportedthroughouttheworld,althoughinsomeareasithasapproached30%(49).Ahigherprevalencerate,
rangingfrom2050%wasoftenencounteredamonginfantsandchildren,manyofthembeingasymptomatic(50).Transmissionbetweenchildrenhasbeenconsidered
tobeofafecaloralrouteduetoinadequatehygiene.Personsillwithgiardiasishavebeenreportedtoshedasmanyas7.1108cystsperday(64).
Basedonaprevalencerateofaslittleas1%orasmuchas25%ofthepopulation,Jakubowski(27)hasestimatedthatforanaveragecity,sheddingofcystsby
infectedpersonswouldleadtolevelsofGiardiacystsrangingfrom3.6104to9.0105pergallonofrawsewage.Valuesclosetothelowerestimatehavebeen
confirmedrecentlybySykoraetal.(61)whoanalyzedrawandtreatedsewageinMcKeesport,PA,anddemonstratedthatrawwastewatercontained5.0103to
1.5105Giardiacystspergallon,whiletreatedsewagereleasedfromtheplantcontainedasmanyas5.0102cystspergallon.Theviabilityofthecystsinrawor
treatedsewagewasnotdetermined.
(ii)WaterborneOutbreaksofGiardiasisHumanversusAnimalOriginforCysts
IntheUnitedStates,90outbreaksofwaterbornegiardiasisweredetectedfrom19651984,and73%wererelatedtocommunitywatersystemsusingsurfacewater
asthesourcefordrinkingwater(11,12).Waterborneoutbreaksofgiardiasishaveoccurredmainlyatsitesthathavetraditionallydependeduponsurfacewater,where
watertreatmentconsistedofdisinfectionwithchlorinewithoutfiltration.Thesourcesofcystsinthesewaterborneoutbreaksappearedtohavebeenderivedfromeither
contaminationwithhumanwaste,orpossibly,asjustdiscussed,fromanimalslivinginthewatershed.
Humansewagehasbeenshowntocontainfrom104to105Giardiacystspergallon(27,61)andwasshowntoberesponsibleintheUnitedStatesfor15%ofthe
casesofwaterbornegiardiasisduring19651984(11,12).In1946,awaterborneoutbreakofgiardiasiswasattributedtosewagecontaminationofthewatersupplyin
aTokyoapartmentbuilding(18).WaterborneoutbreaksintheUnitedStatessuggestedasbeingduetohumansewagecontaminationofsurfacewatersupplies
includedoutbreaksduring196465inAspen,CO(39),197475inRome,NY(56),1979inBradford,PA(29),andduring1984inMcKeesport,PA(62).
Waterborneoutbreaksinvolvingfilteredsurfacewatersuppliesoccurredduring1977inBerlin,NH(35),and1978inVail,CO(10).Thelatteroutbreakwasthe
largestinvolvingafilteredwatersystem,havingeffectedanestimated5000cases.

Page233

ThedirectinvolvementofanimalsinwaterbornegiardiasisoccurredinCamas,WA,in1976,whichwasreportedasthe"firstsubstantiatedcaseofwildanimals
(beavers)contaminatingahumanpopulationwithGiardia.ThirtyGiardiacystswererecoveredfromtherawwaterintworeservoirs,waterinfluenttothewater
treatmentplant,andinBoulderCreek,oneoftwofeederstreamstothewatertreatmentplant.BeaversweretrappedonJonesCreek,theotherfeederstream,
however,noGiardiacystsweredetectedinthewateratthissite.AlthoughthreebeaversinfectedwithGiardiaweretrappedonJonesCreek,theyweretrapped
furtherdownstreamfromthelocationofthecitywaterintake,thanwerethreebeaversthatwerenegativefortheparasite(15).Thewatershedwassubjecttosome
loggingactivitybutotherwisehadextremelylimitedhumanusage,althoughithadbeenpostulatedthatthebeaversmayhavebeeninfectedwithGiardiathrough
exposuretowatercontaminatedwithhumanfeces,obtainedfurtherdownstream.GiardiacystsisolatedfrombeaverfeceswerereportedtobeinfectiveinfourSPF
beaglepuppies(15).However,crossspeciestransmissionofGiardiacystsbetweenanimalsshouldnotbeconsideredasevidencethatitwillnecessarilyoccurfrom
animalstoman(4).Also,aclaim(30,35)wasmadethattheGiardiacystsrecoveredfrombeaversandthecitywatersupplyappearedmorphologicallyidenticalto
G.lambliacystsfrominfectedhumans,butthisshouldnotbesurprisingsincemostcystsdescribed(atthattime)inwarmbloodedanimalshadastriking,ifnot
identical,morphologicalappearance.Thus,uponreevaluation,theevidenceforimplicationofbeaversinthewaterbornegiardiasisoutbreakwasnottotallyconvincing,
especiallyinlightoftheacknowledgementofpossiblehumancontaminationofwaterdownstream.
SubsequenttothewaterbornegiardiasisoutbreaksinCamas,WAandBerlin,NH[anoutbreaklaterattributedtosewagecontamination(34)],waterborneoutbreaks
wereinsomeinstances,attributedtothepresenceofbeaversinthewatersheds.Thisoccurredeventhoughthepotentialforhumancontaminationwaspresent,andin
certaincases,amorelikelyexplanation.InRedLodge,MT,agiardiasisoutbreakin1983wasrelatedtoincreasedmeltrunoffduetotheMt.St.Helensvolcanic
eruption(66).Theinvestigationrevealedbeayer,dogs,andcattlewithinthewatershedandtheywerepostulatedaspotentialsourcesofGiardiacysts.However,the
presenceoffecalcoliformsinsamplesoftapwatertogetherwithhumanusageofthewatershed,includingpotentialoverflowfromsepticsystemsupstreamfromthe
citywaterintake,suggestedthatthewaterwascontaminatedwithhumanwaste,theoreticallyamorelogicalsourceofGiardiacystsinfectivetoman.InReno,NV,a
giardiasisoutbreakoccurredin1982andbothGiardiacystsandaninfectedbeaverweredetectedinoneoftwowatersupplyreservoirs(43).Giardiacystswere
detectedin22/27samplesofrawwaterfromtheTruckeeRiverdiversionsthatsuppliedbothreservoirs.Asstatedbytheauthors(43)thepresenceofthebeaverin
thereservoirmayhavebeencoincidentalsincetheriseingiardiasiscasesoccuredseveralmonthspriortothearrivalofthebeaverinthereservoir.Also,Giardiacysts
weredetectedinwaterfromHunterCreekreservoir,wherenobeaverswerefound,butwerenotdetectedinHighlandsreservoirwhichwasinhabitedbytheinfected
beaver.NosystematicsearchwasundertakenforotheranimalsinfectedwithGiardia,includingthelargepopulationsofbirdsknowntofrequentthesereservoirs.
Morerecently,alargewaterbornegiardiasisoutbreakoccuredinPittsfield,MA,inDecember,1985andbeaverswereimmediatelyimplicatedasthesourceof
waterbornecysts,resultinginheadlinesinoneofthelocalpapersreading,"Beaversarenolaughingmatter"(65).Theanalysisoffecalsamplesfromkilltrapped
beaversrevealedthatoneofnineanimals(11%)wasinfectedwithGiardia.Thisinfectedanimalwasoneofthreetrappedinthevicinity(downstream)oftheAshley
reservoir,whichwasindicatedasthesourceofcystcontaminatedwater.ThreemuskratslivetrappedintheAshleyreservoirwereinfectedwithGiardia.However,
humanscouldalsohavebeenasourceofGiardiacystsatthissitebecauseofhumanactivityinthewatershed,includingcanoeingandfishing.Therewerealsoobvious
signsofyoungadultrecreationalusealongtheshoreline,eventhoughtheareawaspostedasarestrictedaccessarea.
TheSourcesofGiardiaCystsinWater
ThehostsourceofGiardiacystsinwatercannotbeattributedtoasinglespecificspeciesofanimalbasedontheuseofthecurrentcriteriaoflightmicroscopic
morphologyandexistingserologicalmethods.MostwatershedsandreservoirshaveanimalsthatarecapableofbeingnaturallyinfectedwithGiardia.Theseanimals
includebeavers,muskrats,variousspeciesofvolesandrodents,domesticanimalsandpets.Also,itincludeshumansandavarietyofavianspeciesincludingwading
birds.Thecystsfrommostoftheseanimalsources,ashasbeenpreviouslystated,areessentiallymorphologicallyindistinguishablefromoneanother,exceptforthe
characteristicbinarycystmorphologyseeninmicrotinerodentcysts.Serologicaltests,basedonimmunofluorescence(52,53,59)havebeenusedfortheidentification
ofGiardiacystsinconcentratedwatersamples,butnoneoftheseoranyotherimmunologicalmethodsreportedtodate,havebeenshowntobecapableof
distinguishingvariousspeciesofGiardiabasedonhostorigins.TheyarealsoincapableofadequatelydifferentiatingsubpopulationswithintheG.duodenalis
classification(20),thepredominantforminmammalsandbirds.Therefore,sinceGiardiacystscannotbeeasilydistinguishedfromoneanotheronthebasisoftheir
morphology,themeredetectionofGiardiacystswithinwatersamplesandtheircorrelationwiththepresenceofinfectedanimals(andsometimesuninfectedones)ina
watershedcanleadtoerroneousconclusionspertainingtotheoriginofthecysts.Thisisparticularlyimportantbecausemanyanimalsmayhaveendogenousinfections
ofGiardia,insomecaseswithaprevalencerateofgreaterthan90%,thatmaybeentirelyunrelatedtotheGiardiaspeciesproducingdiseaseinman.Insupportof
thisidea,areevaluationoftheuseofanimalstoolsurveysinwatershedshasalsoindicatedthatsuch

Page234

surveysmaybeoflittlevalueinassessingtheriskforwaterbornetransmissionofgiardiasis(12).
Evidencehasaccumulated,overthepasttenyears,thatsupportedtheconceptthat,withinlimitations,crossspeciestransmissionofgiardiasiscanoccurbetween
animals(4).ThesuccessfulinfectionofgerbilswithGiardiacystsderivedfromavarietyofhostsincludingmouse,beaver,muskrat,andman,andalsothesuccessful
infectionofbeaversandmuskratswithhumancystsindicatedthatcrossspeciestransmissioncouldoccur.Itwastemptingtoextrapolatefromsuccessfulcrossspecies
transmissionofgiardiasiswithinanimalmodelsandassumethattransmissioncouldoccurfromanimals,suchasthebeaverormuskrat,toman.However,withoutany
supportiveevidencethatthecystsintheseanimalswerethesamespeciesasthatfoundwithinthewaterandwithinman,itwouldseempresumptuoustostatethatany
particularanimalisthesourceofthecystsinfectingman.Ifitcanbeshownthatinfectedanimalswithinthewatershedwereharboringorganismscapableofproducing
diseaseinman,thenitmightbefeasiblethattheymayhaveactedashostsofazoonoticinfection.However,sinceitwaslikelythattheseanimalswouldobtaintheir
infectionfromhumancystspresentinthewater,itmightbepossiblethattheirinfectionwasonlyasecondarysignreflectingthecontaminationofthewaterfroman
unidentifiedprimarysource.
Giardiacystsinwaterwouldseemtohavebeenderivedeitherfromhumansorfromanimalsfrequentingorinhabitingthewatershed.Thepotentialforthepresenceof
largenumbersofGiardiacystsinrawhumanwaste[3.69.0105gallon(27)]andtheoccurrenceofhumanrecreationorhabitationnearorwithinwatershedshas
mademanalikelysuspectastheprimarysourceofGiardiacystsinwater.Ifhowever,weassumedthatanimalsotherthanman,suchasthebeaver,werethemajor
sourceofcystsinwater,thenwewouldbefacedwiththefollowingparadox.Waterborneoutbreaksofgiardiasishaveneveroccuredmorethanonceatanyspecific
site,despitethefactthatthepopulationsofanimals,beaverorotherwise,haveessentiallyremainedunchangedoverthepastdecade(4).Theoretically,theanimals
shouldhavebeenactingasaconstantsourceofcysts.If,subsequenttotheinitialGiardiaoutbreak,thesamesituationregardingthequalityofwatertreatment
(chlorinelevel,contacttime,turbidity)weretorecur,thenwhyhadasecondoutbreakofgiardiasisnotoccuredatanyofthesesites?Ithasbeenpostulated(1)that
thelackofrepetitionofgiardiasisoutbreaksatanysitemayhavebeenduetoeithera)theuseofcorrectivemeasures,suchasinstallingfiltrationoreffective
disinfection,orb)someimmunityacquiredbythepopulationaffected.Theroleofimmunityinpreventingreoccurrencesofgiardiasisoutbreaksmaybequestionable,
sinceifanimalshavealwaysbeenacontinuingsourceofcysts,thenimmunitytoGiardiashouldhavebeenongoing,andthus,preventedtheinitialoutbreak.Another
plausibleexplanationcouldbethattherepeateddetectionofGiardiacystsinwater,togetherwiththelackofasecondoutbreak,wouldsuggestthatnotallcysts
presentwithinwaterwerepossiblyinfectioustoman.
DeterminingtheSourceofGiardiaCystsinWaterborneOutbreaksofGiardiasis
ToanswerthequestionastowhichanimalsmaybethesourceofGiardiacystsinfectioustoman,wewouldrecommendthefollowingapproach:1)thereshouldbea
systematicsamplingofallsuspectedanimals,mammalsandbirds,foundwithinthewatershed,2)trophozoitesandcystscollectedfromnecropsiedanimals,cysts
concentratedfromwatersamples,andcystsfromgiardiasispatients,shouldbeusedtoestablishinvitrocultures,3)thecontinueddevelopmentofthetechniqueforin
vitroencystationshouldpermitthecollectionofcystsfreeofbacterialcontamination,4)usingantigensisolatedfrominvitroinducedcystsandeachaxeniccultureof
Giardia,attemptsshouldbemadetoproduceimmunologicandmolecularprobesselectiveorspecificforspeciesofGiardia,and5)Giardiaisolatedfromall
sourcesshouldbeexaminedandthespeciesdeterminedbymeansofmorphological,biochemical,immunological,andmetabolictechniques.Theinformationderived
shouldbecomparedtoresultsobtainedfromcrossspeciestransmissionexperimentsdonewheneverfeasible.Thecomplexityandneedforfurtherdevelopmentof
severalofthemethodsdescribedabovewouldprecludetheiraccomplishmentwithinanyonelaboratory,butacooperativeeffortutilizingtheexpertiseavailablein
variousspecializedlaboratoriesmaybesuccessful.
Acknowledgements
TheauthorswishtoexpresstheirappreciationtoMs.LeeAnnSherlock,Ms.MaryJanuschka,andMs.LisaKampfortheirexcellenttechnicalassistance.Wewould
liketothankDr.LouisDiamond,LaboratoryofParasiticDiseases,NIH,forprovidingtheaxenicGiardiacultures,IP0581:1andIP0482:1.Wealsowishto
acknowledgethecooperationandassistanceprovidedbyMr.EdBogges,MinnesotaDepartmentofNaturalResources,Mr.HenryHilton,MaineInlandFisheries
andWildlife,Mr.EricOrff,NewHampshireDepartmentofFishandGame,Mr.PaulBishop,NewYorkDepartmentofEnvironmentalConservation,Mr.Jim
Distefano,VermontDepartmentofFishandGame,Mr.JackKorlath,MinnesotaDepartmentofPublicHealth,andMr.RichardBuech,U.S.D.A.ForestServiceat
theUniversityofMinnesota.WewouldalsoliketothankMr.WarrenMathis(NY),Mr.WarrenAnderson(NH),Mr.HenryLaramie(NH),andMr.Clarence
Rademacher(MN)fortheirskilledassistanceintrappinganimals,aswellasotherstoonumeroustomention.Theresearchdescribedinthisarticlehasbeenfundedby
theU.S.EnvironmentalProtectionAgencythroughcooperativeagreementNo.CR811834,butithasnotbeensubjectedtotheAgencys'review,andtherefore,does
notnecessarilyreflecttheviewsoftheAgencyandnoofficialendorsementshouldbeinferred.

Page235

LiteratureCited
1.Amirtharajah,A.1986.Varianceanalysesandcriteriafortreatmentregulations.J.Am.WaterworksAssoc.78:3449.
2.Ansari,M.A.R.1952.ContributionaletudedugenreGiardiaKunstler,1882(Mastigophophora:Octimitidae).Am.Parasitol.Hum.Comp.27:421479.
3.Barbour,A.G.,Nichols,C.R.,andT.Fukushima.1976.Onoutbreaksofgiardiasis.Am.J.Trop.Med.Hyg.25:384389.
4.Bemrick,W.J.1984.Someperspectivesofthetransmissionofgiardiasis.InGiardiaandGiardiasis:Biology,Pathogenesis,andEpidemiology.Erlandsen,
S.L.andE.A.Meyer.(eds).PlenumPress,NewYork,NY.pp.379400.
5.Belosevic,M.,Faubert,G.M.,Guy,R.,andJ.D.MacLean.1984.ObservationsonnaturalandexperimentalinfectionswithGiardiaisolatedfromcats.Can.J.
Comp.Med.48:241244.
6.Belosevic,M.,Faubert,G.M.,Maclean,J.D.,Law,C.,andN.A.Croll.1983.Giardialambliainfectionsinmongoliangerbils:ananimalmodel.J.Infect.Dis.
147:222224.
7.Boeck,W.L.1919StudiesonGiardiamicroti.Univ.Calif.Pub.Zool.19:85134.
8.Box,E.D.1981.ObservationsonGiardiaofbudgerigars.J.Protozool.28:491494.
9.CraunG.F.1979.Waterborneoutbreaksofgiardiasis.In:WaterborneTransmissionofGiardiasis.Jakubowski,W.andJ.C.Hoff.(eds).U.S.Environmental
ProtectionAgency,Cincinnati,OH.pp.127149.
10.Craun,G.F.1984.Waterborneoutbreaksofgiardiasis:currentstatus.In:GiardiaandGiardiasis:Biology,Pathogenesis,andEpidemiology.Erlandsen,S.L.,
andE.A.Meyer.(eds).PlenumPress,NewYork,NY.pp.243261.
11.Craun,G.F.1986.WaterbornegiardiasisintheUnitedStates,19651985.Lancetii(8505):513514.
12.Craun,G.F.,andW.Jakubowski.1987.Statusofwaterbornegiardiasisoutbreaksandmonitoringmethods.WaterResourcesBull.,Inpress.
13.Davies,R.R.,andC.P.Hibler.1979.AnimalreservoirsandcrossspeciestransmissionofGiardia.In:WaterborneTransmissionofGiardiasis.Jakubowski,
W.andJ.C.Hoff(eds).U.S.EnvironmentalProtectionAgency,Cincinnati,OH.pp.104126.
14.Davis,C.andL.S.Ritchie.1948.ClinicalmanifestationsandtreatmentofepidemicamebiasisocurringinoccupantsoftheMantestsaapartmentbuilding,Tokyo,
Japan.Am.J.Trop.Med.28:817822.
15.Dykes,A.C.,Juranek,D.D.,Lorenz,R.A.,Sinclair,S.,Jakubowski,W.,andR.Davis.1980.Municipalwaterbornegiardiasis:anepidemiologicalinvestigation.
AnnInt.Med.92:165170.
16.Erlandsen,S.L.1974.Scanningelectronmicroscopyofintestinalgiardiasis:Lesionsofthemicrovillousborderofthevillusepithelialcellsproducedby
trophozoitesofGiardia.In:ScanningElectronMicroscopy.1973.O.Johari(ed).IITResearchInstitute,Chicago.pp.775782.
17.Erlandsen,S.L.,andD.G.Chase.1974.Morphologicalalterationsinthemicrovillousborderofvillousepithelialcellsproducedbyintestinalmicroorganisms.Am.
J.Clin.Nutr.27:12771286.
18.Erlandsen,S.L.,andW.J.Bemrick.1987.SEMevidenceforanewspecies,Giardiapsittaci.J.Parasitology.73:623629.
19.Errington,P.L.1963.MuskratPopulations.IowaStateUniversityPress,Ames,IA.pp.664.
20.Filice,F.P.1952.StudiesonthecytologyandlifehistoryofaGiardiafromthelaboratoryrat.Univ.Calif.Publ.Zool.57:53143.
21.Friend,D.S..1966.ThefinestructureofGiardiamuris.J.CellBiol.29:317332.
22.Frost,F.,Plan,B.andB.Leichty.1980.Giardiaprevalenceincommerciallytrappedanimals.J.Environ.Health42:245249.
23.Georgi,M.E.,Carlisle,M.S.,andL.E.Smiley.1986.Giardiasisinagreatblueheron(Ardeaherodiasis)inNewYorkState:anotherpotentialsourceof
waterbornegiardiasis.Am.J.Epidemiology123:916917.
24.Hewlett,E.L.,Andrews,J.S.,Ruffin,J.,andF.W.SchaeferIII.1982.ExperimentalinfectionofmongreldogswithwithGiardialambliacystsandcultured
trophozoites.J.Infect.Dis.145:8993.
25.Holberton,D.V..1974.AttachmentofGiardiaAhydrodynamicmodelbasedonflagellaractivity.J.Exp.Biol.60:207221.
26.IsaacRenton,J.L.,Moriez,M.M.,andE.M.Proctor.1987.AGiardiasurveyoffurbearingmammalsinBritishColumbia.J.Environ.Health.50:8083.
27.Jakubowski,W..1984.DetectionofGiardiacystsindrinkingwater:stateoftheart.In:GiardiaandGiardiasis:Biology,Pathogenesis,andEpidemiology.
Erlandsen,S.L.andE.A.Meyer.(eds).PlenumPress,NewYork,NY.pp.263286.
28.Jephcott,A.E.,Begg,N.T.,andI.A.Baker.1986.OutbreakofgiardiasisassociatedwithmainswaterintheUnitedKingdom.TheLancet29:730732.
29.Keifer,A.,Roberto,R.R.,Powers,L.,Gaston,J.,Sherwood,R.W.,Johnston,L.,Blair,J.,Early,B.,Hopkins,R.S.,Scott,E.,Smede,R.,Osterud,H.,Slifman,
N.,Shilke,J.,Hills,L.,Goodins,J.A.,Burkhart,T.J.,Jenkins,M.T.,Witte,E.J.,McCarthy,M.,deMelfi,T.,Erb,J.,andE.C.Lippy.1980.Waterbornegiardiasis
inCalifornia,Colorado,Oregon,Pennsylvania.Morbid.Mortal.WeeklyRep.29:121123.
30.Kirner,J.C.,Littler,J.D.,andL.A.Angelo.1978.AwaterborneoutbreakofgiardiasisinCamas,Wash.J.Am.WaterworksAssoc.70:3540.
31.Kulda,J.,andE.Nohynkova.1978.Flagellatesofthehumanintestineandofintestinesofotherspecies.In:ParasiticProtozoa.J.P.Kreier(ed).Academic
Press,NewYork,NY.pp.69104.
32.Lavier,G.1924.DeuxespecesdeGiardiaduratd'egoutParisien.(Epimysnorvegicus).Ann.Parasitol.Hum.Comp.11:161168.
33.Lefebrve,P.A.,andJ.L.Rosenbaum.1986.Regulationofthesynthesisandassemblyofciliaryandflagellarproteinsduringregeneration.Ann.Rev.CellBiol.
2:515544.
34.Lin,S.D.1985.Giardialambliaandwatersupply.J.Am.WaterworksAssoc.77:4047.
35.Lippy,E.C..1978.TracingagiardiasisoutbreakatBerlin,NewHampshire.J.Am.WaterworksAssoc.70:512520.
36.Lopez,C.E.,Juranek,D.D.,Sinclair,S.P.,andM.G.Schultz.1978.GiardiasisinamericantravelerstoMadeiraIsland,Portugal.Am.J.Trop.Med.Hyg.
27:11281132.
37.Meyer,E.A.,andM.Radulescu.1979.Giardiaandgiardiasis.Adv.Parasitol.17:147.
38.Meyer,E.A.,andE.L.Jarroll.1980.Giardiasis:reviewandcommentary.J.Epidemiology111:112.
39.Moore,G.T.,Cross,W.M.,McGuire,D.,Mollohan,C.S.,Gleason,N.,Healy,G.R.,andL.H.Newton.1969.Epidemicgiardiasisataskiresort.NewEng.J.
Med.281:402407.
40.MorbidityandMortalityWeeklyReport,CenterforDiseaseControl.1978.IntestinalparasitesurveillanceUnitedStates1976.27:167168.
41.Mueller,J.C.,Jones,A.L.,andL.L.Brandborg.1974.Scanningelectronmicroscopeobservationsinhumangiardiasis.In:ScanningElectronMicroscopy.O.
Johari(ed).IITResearchInstitute,Chicago.pp.557564.
42.Nash,T.E.,McCutchan,T.,Keister,D.,Dame,J.B.,Conrad,J.D.,andF.D.Gillin.1985.RestrictionendonucleaseanalysisofDNAfrom15Giardiaisolates
obtainedfromhumansandanimals.J.Infect.Dis.152:6473.
43.Navin,T.R.,Juranek,D.D.,Ford,M.,Minedew,D.J.,Lippy,E.C.andR.A.Pollard.1985.CasecontrolstudyofwaterborneGiardiasisinReno,Nevada.Am.
J.Epidemiol.122:269275.

Page236

44.Nentyl,J.A..1983.TheBeaver.CrestwoodHouse,Mankato,MN.pp.147.
45.Owen,R.L.,Nemanic,P.C.,andD.P.Stevens.1979.Ultrastructuralobservationsongiardiasisinamurinemodel.I.Intestinaldistribution,attachment,and
relationshiptotheimmunesystemofGiardiamuris.Gastroenterology.76:757769.
46.Owen,R.L.1980.TheultrastructuralbasisofGiardiafunction.Trans.R.Soc.Trop.Med.Hyg.74:429433.
47.Padchenko,J.K.,andN.G.Stolyarchuk.1970.DogsasaspontaneouscarrierofLambliaandprobablesourceandvectoroflambliasisinnature.Vestnik
Zoologiya4:5661.
48.Panigraphy,B.,Mathewson,J.J.,Hall,C.F.,andL.C.Grumbles.1981.Unusualdiseaseconditionsinpetandaviarybirds.J.Am.Vet.Med.Assoc.178:394
395.
49.Peterson,H.1972.Giardiasis(lambliasis).Scand.J.Gastroenterology7(supplement14):144.
50.Pickering,L.K,Woodward,W.E.,DuPont,H.L.,andP.Sullivan.1984.OccurrenceofGiardialambliainchildrenindaycarecenters.J.Pediatrics.
104:522526.
51.Poley,J.R.,andS.Rosenfield.1982.Malabsorptioningiardiasis:Presenceofaluminalbarrier(mucoidpseudomembrane).Ascanningandtransmissionelectron
microscopestudy.J.Pediatr.Gastroenterol.1:6380.
52.Riggs,J.L.,Dupuis,K.W.,Nakamura,K.,andD.P.Spath.1983.DetectionofGiardialambliabyimmunofluorescence.Appl.Environ.Micro.45:698700.
53.Sauch,J.F.1985.UseofimmunofluorescenceandphasecontrastmicroscopyfordetectionandidentificationofGiardiacystsinwatersamples.Appl.Environ.
Micro.50:14341438.
54.Scholtens,R.G.,New,J.C.,andS.Johnson.1982.Thenatureandtreatmentofgiardiasisinparakeets.J.Am.Vet.Med.Assoc.180:170173.
55.Schupp,D.G.,andS.L.Erlandsen.1987.Anewmethodtodeterminecystviability:correlationbetweenfluoresceindiacetate/propidiumiodidestainingand
animalinfectivity.J.Parasitol.53:704707.
56.Shaw,P.K.,Brodsky,R.Z.,Lyman,D.O.,Wood,B.T.,Hibler,C.P.,Healy,G.R.,McLeod,K.I.E.,Stahl,W.,andM.G.Schultz.1977.Acommunitywide
outbreakofgiardiasiswithevidenceoftransmissionbyamunicipalwatersupply.Ann.Int.Med.87:426432.
57.Sheffield,H.1979.TheultrastructuralaspectsofGiardia.In:WaterborneTransmissionofgiardiasis.Jakubowski,W.andJ.C.Hoff.(eds).U.S.Environmental
ProtectionAgency,Cincinnati,OH.pp.921.
58.Simon,C.E.1922.Acritiqueofthesupposedrodentoriginofhumangiardiasis.Am.J.Hyg.1:406434.
59.Sorenson,S.K.,Riggs,J.L.,Dileanis,P.D.,andT.J.Suk.1986.IsolationanddetectionofGiardiacystsfromwaterusingdirectimmunofluorescence.Am.
WaterRes.Bull.22:843845.
60.Spaulding,J.J.,Pacha,R.E.,andG.W.Clark.1983.QuantitationofGiardiacystsbymembranefiltration.J.Clin.Microbiol.18:713715.
61.Sykora,J.L.,Bancroft,W.D.,States,S.J.,Shapiro,M.A.,Boutros,S.N.,Keleti,G.,Turzai,M.,andL.F.Conley.1987.Giardiacystsinrawandtreated
sewage.In:ControlingWaterbornegiardiasis.Chapter3.G.S.Logsdon(ed).EnvironmentalEngineeringDivision,AmericanSocietyofCivilEngineers.pp.2233.
62.Sykora,J.L.,States,S.J.,Bancroft,W.D.,Boutros,S.N.,Shapiro,M.A.,andL.F.Conley.1986.MonitoringofwaterandwastewaterforGiardia.Proc.of
1986AmericanWaterworksAssociationWaterQualityTechnologyConference,Portland,OR.pp.10431054.
63.Takano,J.,andJ.H.Yardley.1965.Jejunallesionsinpatientswithgiardiasisandmalabsorption:Anelectronmicroscopicstudy.Bull.JohnsHopkinsHosp.
116:413429.
64.Tsuchiya,H.1931.AstudyonthevariabilitiesindimensionalandnumbersofdischargedcystsinGiardialamblia(Stiles,1915)fromdaytodayundernormal
conditions.Am.J.Hyg.13:544567.
65.Tynan,T.1985.BeaversarenolaughingmatterinMass.NewHampshireSundayNews,Sunday,Dec.29.p.4a.
66.Weniger,B.G.,Blaser,M.J.,Gedrose,J.,Lippy,E.C.,andD.D.Juranek.1983.Anoutbreakofwaterbornegiardiasisassociatedwithheavywaterrunoffdueto
warmweatherandvolcanicashfall.Am.J.Publ.Health73:868872.
67.Wolfe,M.S.1978.Currentconceptsingiardiasis.NewEng.J.Med.298:319321.
68.Woo,P.T.K.,andW.B.Paterson.1986.GiardialambliainchildrenindaycarecentresinsouthernOntario,Canada,andsusceptibilityofanimalstoG.
lamblia.Roy.Soc.Trop.Med.Hyg.80:5659.
69.Wright,R.A.,Spencer,H.C.,Brodsky,R.E.,andT.M.Vernon.1977.GiardiasisinColorado.Anepidemiologystudy.Am.J.Epidemiology.105:330336.

Page237

AnalysisofMunicipalWaterSamplesforCystsofGiardia
CharlesP.Hibler
CHDiagnostic&ConsultingServiceInc.,2012DerbyCourt,FortCollins,Colorado80526,U.S.A..
Analysisof4423watersamplesfrom301municipalsitesin28statesbetween1979and1986hasshownthat102/301(34%)sitessometimeswerepositive.The102sitesaccounted
for3633/4423(82%)ofthesesamplesthe199negativesitessent651samples,or3.27samples/site.Theremaining139sampleswerefromunknownsites,unknownsources,orof
unknowntype.Rawsamplesinclude512/1968(26%)positive,finishedsamplesinclude267/2372(11%)positive,andunknowntype15/83(18%)positive.Positivesamples(both
rawandfinished)weredistributed:creeks346/1218(28%),rivers212/828(26%),lakes193/1983(10%),springs16/84(19%),andwells2/63(3%)(mixedsourcesexcludedfrom
total).Positivefinishedsamplesweredistributed:unfiltered80/1214(6.6%),directfiltration148/615(24%),conventionaltreatment12/357(3.4%),commercialmanufactured
4/33(12%),slowsand0/11,diatomaceousearth0/7,mechanical11/51(21.6%)andinfiltrationgallery7/37(18.9%)(unknownsexcludedfromtotal).

Introduction
Morethan6500watersampleshavebeenexaminedforcystsofGiardiathepast12yearsbytheauthor'slaboratory.Sampleshavebeenanalyzedfrom28ofthe
contiguous48UnitedStates.Atotalof4423watersamplesfrom301sites(municipalities)areincludedinthisreport.DatafromSeptember,1979through
September,1986havebeenincludeddataobtainedpriorto1979arenotcompletebecausethoroughrecordswerenotkeptduringtheseyears.Over1000water
samplesfromforeigncountriesandresearchsamplesonpilotfiltrationsystemshavenotbeenincluded.
MaterialsandMethods
SamplingandDiagnosticTechniques
Themodifiedreferencemethodwasusedprimarilyfortheseanalysesandtheindirectimmunofluorescenttechniquewasoccasionallyusedasacomparison.A
descriptionofthetechniques,themodificationsorimprovementsinthosetechniquesandtheassociatedproblems,pastandpresent,havebeenpresentedinanother
articleinthisbook.Recoveryofcystswashighlydependentupontheturbidity,thetypeofturbidity(organicorinorganic)andthequalityofthecysts.Recoveryof
cystswasinverselyproportionaltotheturbidityofthesource.Cystswereeffectivelyrecoverediftheturbiditywasprimarilyinorganicandtheconcentratefromthe
filtercartridgeswasdilutedtotheextentthatthecentrifugatefroma25mLaliquotwasequaltoorlessthan0.25mLinvolume.Iftheturbiditywasprimarilyorganic,
dilutionoftheconcentratetogetherwithuseofalowerthanoptimumspecificgravityofthechemicalwasusuallynecessary.Ifcystswerealiveandingoodcondition,
recoveryfromlowinorganicturbiditysourceswasexcellent.Ifcystswerealiveandingoodconditionandtheturbiditywasprimarilyorganic,recoverywaspoor.If
cystsweredyingand/ordead,recoverywasverypoor,irrespectiveofthetypeandamountofturbidity.Ifthesourcecontainedlargenumbersofalgae,diatomsand
freelivingflagellates,theirpresenceinterferedwithvisualizationandquicklyledtoeyefatigue.
Ifthesamplehadbeenchlorinated,eventhoughinlinedechlorinationmayhavebeenappliedtothesample,thecystswereusuallyinactivated.Thetimelagbetween
sampling,shipping,processingandanalysisofsamplesfromchlorinatedsourcesinvariablyresultedindecreasedrecovery.Ifanunfilteredsourceused
presedimentationwithalum,orifafiltrationsystemusedalumorpolymers,andthesechemicalswerepresentinthesample,thesampleswereessentiallyimpossibleto
analyzeeffectively.Alumandpolymerswouldveryeffectivelycoagulatedebris,cystsandchemicalintoaninseparablemassintheconcentrateobtainedfromwashing
thefiltercartridges.
AssessmentofRisk
Ifawatersamplewasobtainedfromamunicipalityusingsometypeoffiltration,botharawsampleandafinished(filterplanteffluent)samplewerenecessaryto
assessfilterplantefficiency.ThepresenceorabsenceofGiardiacystswassecondarytotheabilityofthefiltrationplanttoremovematerialthesizeoforlargerthan
Giardiacysts.Ariskassessmentof0to100%removalwasusedtoevaluateplantefficiency.Plantdebris(theundigestedfecalmaterialfromherbivorousanimals),
somespeciesofalgaeanddiatoms,coccidia(bothfishandmammal),parasiticnematodeeggs(usuallybeaver)andcrustaceansorcrustaceaneggswereallexcellent
indicators.Plantdebriswasthebestindicatorbecauseitisextremelylightinweightandfiltrationplantshavingproblemsremovingriskmaterialinvariablypassedplant
debris.Themostcommonsourceoftheplantdebriswasbeaverormuskrat(muskratand/orsmallerrodents).
TypesofGiardiaCysts
ThemajorityoftheGiardiacystsfoundinsurfacewatersourcesweretheG.duodenalistype.Afewsources,especiallylargeriversreceivingsewageeffluentfrom
largemunicipalities,hadbothG.duodenalisandG.muris.
Necropsyof28blackcrownedherons,amigratorywaterfowl,revealed100%infectionwithaspeciesofGiardiaverysimilarmorphologicallytoG.duodenalis.
Thecyst,however,isnearsphericalinshape.Likelyotherwaterfowlalsoareinfected.The''birdtype"Giardiacysthasbeenfoundinlakes,pondsandlarge,open
springsthroughouttheUnitedStates.
ResultsandDiscussion
AllSamples
The4423watersamplesoriginatedfromsprings,creeks,rivers,lakesandwells.Thisincludes1968rawwatersamplesand2372finishedwatersamples.Atotalof
83wereunknowntypesamples(Table1,Figures1,1A).Eighteenpercent(794/4423)ofthesamplesfromallsourceswerecontaminatedwithcystsofGiardia,and
15/83(18%)oftheunknownswerecontaminated.
Ofthe301sitesfromwhichsampleswerereceived,102/301(34%)sitessentsamplesoftencontaminatedwithGiardiacysts3633/4423(82%)ofthetotalsamples

Page238
TABLE1.DetectionofGiardiacystsinwatersamples:classificationbytypeofwater*.
Watertype

#ofsamplesexamined

%oftotalsamples

#ofsamplespositive

%ofsamplespositive

Raw

1968

44.5

512

26

Finished

2372

53.6

267

11

Unknown

83

1.9

15

18

Totals

4423

100

794

18

*Samplesoriginatedfrom301municipalsitesin28states102(34%)siteswerepositiveforGiardiacysts.Negative
sampleswereobtainedfrom199sites,butony651samples(average3.27samples/site)wereexamined.

originatedfromthese102sitesand794/3633(22%)werepositive.
Rawwatersamplesconsistedof1968/4423(44.5%)ofthesamplesexaminedand512/1968(26%)werecontaminatedwithcysts512/794(64%)ofthepositive
sampleswererawwater(Table1).The102sitesthathadsamplesoftencontaminatedwithcystssent1655/1968(84%)oftheserawwatersamplesforanalysisand
512/1655(31%)werecontamined.
Finishedwatersamplesconsistedof2372/4423(54%)ofthesamplesexaminedand267/2372(11%)werecontaminatedwithcysts,267/794(34%)ofthepositive
sampleswerefinishedwater(Table1).The102sitesthathadsamplesoftencontaminatedwithcystssent1978/2372(83%)ofthesefinishedsamplesforanalysisand
267/1978(13.5%)werecontaminated.
The199negativesitesfromwhichsampleswereobtainedsent651samplesforanalysis,anaverageof3.27samples/site.Whilepositivesamplesfrequentlywere
obtainedatasinglesampling,oftenrepeatedsamplingwasnecessary.Withfewexceptions,repeatedsamplingprovidedpositiveresults.
Somesitesconsistentlyhadcystspresentinthewaterwhileinothersthecontaminationwassporadic.The

Figure1.
Compositeofsamplesfromallsources:794/4423(18%)
positive,19791986. Negative

numberofcystsrecoveredvariedconsiderablybetweensamplesfromthesamesiteandbetweendifferentsitesatthesamesource.Onoccasion,sampleswouldbe
comparedfromthesamesitebuttakenatdifferenttimesovera24hourperiod.Ifonlyafewcystswerepresent,recoverywouldvaryfromnonetoanumbersimilar
tothenumberfoundpreviously.Ifconsiderablenumbersconsistentlywerepresent,littledifferencewasnotedinrecovery.Numbersvariedfromaslowas0.07
cysts/100gallonsofwatersampledtoashighas1472cysts/100gallonsofwatersampledduringepidemicsofwaterbornegiardiasis.Inthosefewinstanceswhere
sampleswereobtainedduringthecourseofanepidemic,10to15cysts/100gallonsofwatersampledwasnotunusual.Sourcescontaminatedwithrawortreated
sewageoftenhadcyststoonumeroustocounteffectively.
ThedatafromallsitesindicatethatGiardiacystsaremorenumerousinthelatewinterthroughearlyspringmonthsthantheothermonthsoftheyear(Figures1,1A).
If,indeed,cystsaremorenumerousduringthesemonths,systemswouldbeatgreaterriskbecausewintertemperatureswouldnecessitateuseofmorechlorineand/or
timeofcontacttoinactivatethecysts.However,anumberoffactorscaninfluencethesedata:1)turbidityisgenerallylowerinallsourcesduringthewinter,increasing
theefficiencyofthetechniques2)numbersofalgae,diatoms,flagellates,pollen,etc.areconsiderablyreduced,facilitatingvisualizationofthecysts3)volumeofwater
isoftenconsiderablyreducedincreeksandrivers,essentiallyconcentratingthecystsand4)cystslivelongerincoldwater.
Asinanalysisofrawwater,manyfactorsinfluencecystrecoveryinfinishedwater.Manyofthefinishedwatersamples(filteredorunfiltered)andallofthedistribution
sampleshadbeenexposedtochlorineandeventhoughmunicipalitiesusuallydechlorinatedtheirsampleswith1%sodiumthiosulfate,eithermeteredinlineor
subsequenttodisconnectingthesamplingcartridge,theresultsoftheseanalysesobviouslywerecompromisedbythechlorineandthetimelagbetweensampling,
shipping,processingandanalysis.Someofthefinishedwatersamplesoriginatedfromsourcesusingonlypresedimentation

Figure1A.
Compositesamplesfromallsources:794/4423(18%)positive,19791986.

Page239
TABLE2.DetectionsofGiardiacystsinwatersamples.Classificationbysourceandtypeofwater.
Numberof
sites

Numberof
sitespositive

Percentof
sitespositive

Numberof
samples

Numberof
samples
positive

Percentof
samples
positive

Creeks(total)

75

38

51

1218

346

28

Raw

444

181

41

Finished

774

165

21

Rivers(total)

74

38

51

828

212

26

Raw

449

163

36

Finished

379

49

13

Lakes(total)

49

19

39

1983

193

10

Raw

829

138

17

Finished

1154

55

Springs(total)
**

36

14

84

16

19

Wells(total)**

40

63

Source*

*Municipalitiesusingacombinationofsources(i.e.creekandspring)havebeenexcludedfromtotals.
**Mostwaterfromspringsandwellsisunfilteredandmay/maynotbedisinfectedbeforeconsumption.

andthealumpresentmadeeffectiveanalysisessentiallyimpossible.Filteredsourcesusingchemicalpretreatmentwithalumand/orpolymersoftenpassedthese
chemicalsthroughthefiltrationsystemandthesesamplesalsowereoftenimpossibletoanalyze.
Somemunicipalitiesinitiatedanimalcontrol(beaverandmuskrat)onthesourceifaconsiderablenumberofcystswereconsistentlypresent.Frequentlyanimalcontrol
wasinitiatedevenwhenthemunicipalityhadasuperbfiltrationsystem.Generallytheauthorrecommendedanimalcontrolaroundalakeorwithinthechannelleading
fromacreekorrivertotheplantifcystswerenumerous.Moreoftenthannottheanimalcontrolprogramwastheninitiatedwithoutinformingthelaboratory.This
resultedinalossofpotentiallyimportantdataregardingtheefficiencyofthecontrolprogram.
Creeks
Thedefinition"creek"variesbetweengeographicregionsintheUnitedStates.A"river"inthewestmaybea"creek"intheeast.Atotalof1218creeksampleswere
analyzedandGiardiacystswerefoundin346/1218(28%)oftherawandfinishedsamples(Figures

Figure2.
Compositeofcreeeksamples:346/1218(28%)positive,
19791986. Negative

2,2A)(Table2).Creeksaresubjecttorunoffand/orthunderstormconditionsmuchthesameasrivers,butonasmallerscale.Ingeneral,creeksprovideanexcellent
habitatforaquatictypemammalsbecauseofthesmallervolumeofwaterandtheproximityofvegetationforfoodandcover.ThepresenceofGiardiacystsincreek
samplesgenerallywastheresultofanimalcontaminationbecausemanycreeksoriginatedinpristineenvironmentsrarely,ifever,visitedbyhumans,especiallywhen
theywereunder810feetofsnowinthewinter.Considerablenumbersofbeaver(51%infected)andmuskrat(100%infected)livingoncreeksinColoradohave
beenexamined.Smallrodents(voles,mice)havenotbeenfoundinfectedinColorado,buttheyhavebeenfoundinfectedinthestateofWashingtonandinwestern
Canada.About20%ofthecattleoncreeksareinfected,andabout5%ofthedomesticsheep(notaccurateforsheepbecausesheepfecalmatterdriesquickly),but
infectionhasnotbeenfoundinfreeranginghorses,muledeer,bighornsheep,etc.Captivemoose,captivemuledeerandpasturedhorseshavebeenfoundinfected.
InfectionhasbeenfoundinelkinthestateofWashington.Cattle,because

Figure2A.
Compositeofsamplesfromcreeks:
346/1218(28%)positive,19791986.

Page240

oftheirhabitoffrequentingstreamscanbealimitedthreat,generallyfunctioningtospreadinfectionbetweenwatersheds.Moose,becauseoftheiraquatichabitscould
alsoposeapotentialthreat.
Creeksamples,fromtheiralpineoriginstotheplains,havebeenexaminedforcontamination.Cystshavebeenfoundinsamplesfromalpinehabitats,butre
examinationusuallyshowedatransientbeavertryingtosurvive.Samplestakenbelowthealpine,intimberedareas,frequentlyarecontaminated.Increeks,asinrivers,
lakes,etc.younganimalpopulationsbeginmovingafterweaningactivitiesinthesummertoestablishahomeaftertherunoff.Theyalsobeginwinterpreparationsin
midsummer.AslateasNovembermanyyoungarestilltryingtofindahomeandoftenwillmovedirectlyintomunicipalintakes.Animalpopulationsstabilizeinthe
winterandthenbeginmovingagaininthespring,whentheystabilizeuntilweaninginthesummer.Samplestakeninbeaverpondsorbelowbeaverdamsinvariably
resultsinrecoveryofcysts,butsamplestakeninthenextpondbelowabeaverdam(stillwater)areusuallynegative.Samplestakenatorbelowcampsitesoncreeks
frequentlyarecontaminatedwithGiardia,however,thesecystscouldbeeitheranimalorhumanorigin.
Ifcreeks(orrivers)traverseurbanareastoreachmunicipalintakes,parasitesofcatsanddogs(especiallyascaridsandcoccidia)arecommonlyfound.Residents
alongthecourseofthiswateroftentossfecalrefuseintothewater,effectivelyfoulingthesource.
Somecreeksourceshavebeenadilemmaformunicipalitiesbecausecarefulexaminationhasnotrevealedtheanimal(s)responsibleforcontamination.These
watershedsareusuallyunder810feetofsnowmostofthewinter,yetcystsarealwayspresentintherawwater.
Smallermunicipalitiesgenerallyusedcreeksasasourceformunicipalsupply,asdidsummercamps,guestranchesorlodges,etc.Manytypesoffiltrationsystems
wereemployedbythesecampsandsmallercommunities,frommechanicalpointofusetoconventionaltreatment.Financesavailableusuallydictatethetypeof
system,andthefunctionalefficiencyofthesystem.

Figure3.
Compositeofriversamples:212/828(26%)positive,
19791986. Negative

Figure3A.
Compositeofriversamples:
212/828(26%)positive,19791986.

Filtrationsystemsfoundoncreekswereusuallymoreatriskforcontaminationwithcysts,primarilybecausethemunicipalitiescouldnotaffordtobuildandmaintain
systemsthatwouldeffectivelyremovethecysts.Creeksgenerallyhadanabundantaquaticanimalpopulationand,moreoftenthannot,ahighqualitysourceofwater
thatmetorexceededexistingstateorfederalregulations.
Rivers
Atotalof828riversampleswereexaminedand212(26%)oftherawandfinishedwatersampleshadcystsofGiardia(Table2,Figures3,3A).Riversinthewest,
northwestandsouthwestmaybecalledcreeksintheeastmoreover,frequentlyanopenchannelorditchfromthemainriversource,oftenseveralmileslong,supplies
theplant.Forallpracticalpurposesthisditchorchannelisacreekandnotariver.Theditchorchannelprovidesahabitatforaquaticanimalsmuchlikeacreekand,if
ittraversesurbanareas,issubjectedtothesamefecalrefusefromdomesticanimalsasacreekthereforedistinguishingbetweencreeksandriversisdifficultatbest.
Riversinthewest,northwestandsouthwestweresubjecttomoresevererunoffconditionsthancreeksandtheeffectsofthunderstormswascompounded.Riversin
theabovegeographicareasgenerally(butnotalways)werecontaminatedwithcystsfromanimalsourceswhereasinthelargerriversintheeastoftenthis
contaminationobviouslywasfromsewageplants.Municipalitiesusingriversourcesforwaterwereusuallylarger,metropolitantypeareaswiththefinancialresources
toconstructplantscapableofremovingthecysts,especiallyintheeastandalongtheplainsareasinthewest.
Thecommentsregardinganimalpopulations,campsites,etc.relatingtocreeksalsoappliestorivers,especiallyinthewesternpartoftheUnitedStates.Unlikecreeks,
theturbidityofriverwater,duringrunofforthunderstormswasmuchhigher.Whilethepercentageofcontaminationwithcystswasaboutthesame,likelytheturbidity
inriverscompromisedtheseresults.
Lakes
Atotalof1983lakesampleswereanalyzedand193(10%)rawandfinishedwatersampleshadGiardiacystspresent(Table2,Figures4,4A).Lakesvariedfrom
manmadetonatural,somenolargerthanponds

Page241

Figure4.
Compositeoflakesamples:193/1983(10%)positive,
19791986. Negative

(alsosubjecttointerpretationindifferentgeographicareas)tothosecoveringseveralsquaremiles.Lakesoftenweremunicipallyownedandprotected.Inthewestern
UnitedStates,lakesweresituatednearorabovethelimitofaquaticanimalactivity.Whilethesetimberlinelakessometimeshavecystspresent,contaminationisusually
sporadic.Highmountain,coldwaterlakesgenerallywereveryclear,withlittleturbidityevenafterthunderstormactivity.Theselakesdidnotsupportverymuchalgae
growth(oranythingelse).Lakesintheeastandnortheastvariedfromthosethatwereveryclear,withlittleorganicturbiditytothosewithhighlevelsoforganicmaterial
andalgae,flagellates,crustaceans,ciliates,amoeba,etc.ThelatterisaveryhealthywatersourcebutdiagnosisofGiardiacystsinsamplesfromtheselakes,especially
inthesummerandearlyfallwasessentiallyimpossible,irrespectiveofthetechiqueused.ConsequentlyFigures4and4A,indicatingthatcystsaremorecommoninthe
winter,areprobablyareflectionoftheturbidityandotherorganismsinterferingwitheffectivediagnosisduringthesummer.Sometimessummersamplesfromlakesare
almostimpossibletoprocesswhenanalgalbloomisunderwaythealgaeandmyriadsofflagellatesquicklycausedeyefatigue.
Somelakeswereverydeep,otherswereshallow.Shallowlakes,whenmunicipaldemandisconsiderable,frequentlydevelopedachannelthatfunctionedmuchlikea
creekorrivermoreover,shallowlakesseemedtobemoreattractivetoaquaticanimals.Largelakes,withtheinfluentatoneendandtheeffluentattheother,and
withoutany"shortcircuiting",seldomhadcystsunlessaquaticanimalswerelivingclosetoorfrequentingtheeffluentarea.
Municipalitiesusingalakeasaprimarysourceofteninitiatedanimalcontrolonthelakebutnotonthesourcesupplyingthelake.Generallythiswasamorecommon
practicewhenthefinishedwaterwasunfiltered.Ifthelakewasdeepthiswaseffectiveifthelakewasshallowitwasnoteffective.Giardiacystssettleveryquicklyin
undisturbedwater.Inlakeswitharichorganicbottom,cystsquitelikelyareconsumedbyfaunalivingonthebottomorbecomelockedintothethickdetritusand

Figure4A.
Compositeoflakesamples:193/1983(10%)positive,19791986.

eventuallydecompose.Monzingo(personalcommunication)foundthatcystssettlingtothebottomofbeaverponds,arichorganicsituation,cannotbefoundashort
periodoftimeaftersettling.
Springs
Atotalof84springsourceswereexaminedand16(19%)rawandfinishedsampleswerecontaminatedwithGiardiaotherswereatriskeventhoughcystswerenot
found(Table2,Figures5,5A).Somewerenottruespringsbutnaturalinfiltrationgalleriescontainingorganismsthatnormallyliveonthesurfaceandneedsunlightto
survive.Somespringsourceswereefficientlyprotected,otherswerehaphazardlyprotectedandsomewereunprotected.Ifthespringwasopen,Giardiacystsoften
werepresent.Whileusersoftenchlorinatedspringwater,veryseldomwasanyfiltrationused.
Wells
Atotalof63wellsampleswereexaminedand2(3%)hadcystsofGiardiapresent(Table2).Onewellwasessentiallyaninfiltrationgalleryofthecreekabout25
feetfromthewell.ThecreeknearthiswellalwayshasconsiderableGiardia,bothfromtheanimalpopulationsandthesewageplantupstream.Theotherwasatrue
well,250feetdeep,thathadbeencontaminatedbyprimingwithriverwaterfromasourcethatwasheavilycontaminedwithGiardia.Sincecystswerefoundinthe
samplesentfromthetruewell,thishadtoberecorded.Someofthewellsampleshadsurfacetypeorganismspresent,suchascrustaceansthereforethesesystems
werenottruewells.Thewellsystemsatriskarenotincludedbecausecystswerenotfound.
FinishedWater
Atotalof2372/4423(54%)ofthesampleswerefinishedwaterand267/2372(11%)ofthesewerecontaminatedwithcysts267/794(34%)ofthepositivesamples
werefinishedwater(Table1).Finishedwatersamplesfromthe102sitesthatoftenhadpositivesamplesaccountedfor1978/2372(83%)ofthefinishedwater
samplesand267/1978(13.5%)ofthesewerecontaminated.
Unfiltered
Unfiltered,finishedwatersamplescamefrom94sitesand16/94(17%)siteshadsamplesoftencontaminatedwithcysts.Atotalof1214/2372(51%)ofthefinished
watersampleswereunfilteredand1046/1214(86%)ofthesamplescamefromthe16sites80/1046

Page242

Figure5.
Compositeofspringsamples:16/84(19%)positive,19791986.
Negative

(7.7%)ofthesesampleshadcystspresent(Table3).Manyoftheunfilteredsampleshadbeentreatedwithchlorinemoreover,aconsiderablenumberofthese
finished,unfilteredsamplesweretakenwithinthedistributionsystem.Thedatapresentedwerecompromisedbecauseofthechlorineandthetimelag.
Mostoftheunfilteredsamplesoriginatedfromsitesusinglakesasasourceformunicipalsupply,butafewsamplescamefromsitesusingcreeks,riversandspringsas
sources.Thecreekandriversitesoftenusedpresedimentationwithchemicalsastheonlytreatmentpriortochlorination.
Fewerpositiverawwatersampleswereobtainedfromlakes,especiallydeeplakes,thanfromcreekorriversourcesandfewercystswerefoundinthesepositiveraw
samplesfromlakes,nodoubtaresultofthe"settlingeffect"instillwater.Irrespectiveofthechlorineandthetimelagthatprobablycompromisedtheresults,80/1046
(7.7%)oreven80/1214(6.6%)positiveindicatesthatalakeorareservoirisagoodbarrierforunfilteredsystemsprovidingadequatechlorineand/ortimeisavailable
orcanbebuiltintothesystem.

Figure5A.
Compositeofspringsamples:16/84(19%)positive,19791986.

DirectFiltration
Ofthe2372finishedwatersamplesanalyzed,615/2372(26%)weredirectfiltration(Table3).Atotalof92sitesuseddirectfiltration,eitherthroughouttheyearorin
thewintermonthsand17/92(18.5%)ofthesesitesoftenhadcystsinthefilteredwater.Atotalof336/615(55%)ofthesamplescamefromthese17sitesand
148/336(44%)hadcystspresent.Usuallycystswerepresentinthosesystemsthatdidnotusechemicalpretreatment(Figure7)ratherthaninsystemsusingchemical
pretreatment(Figure6).
Manysitesswitchfromconventionaltreatmenttodirectfiltrationwhenturbiditiesdecreaseandstabilizenear1NTUinthewintermonthsothersusedirectfiltrationall
year,especiallysmallersites.
Somesitesusedchemicalpretreatmentwithalumorpolymersbeforefiltrationinothersthesystemfunctionedasascreen,generallybecausethewatermet(and
frequentlyexceeded)existingregulations.Severalofthosesitesnotusingchemicalpretreatmentbegantousechemicalswhenanalysisrevealedpassageofcystsand/or
riskmaterial.
TABLE3.DetectionsofGiardiacystsinfinisheddrinkingwatersupplies.
Classification

Numberof
sites

Numberof
sitespositive

Percentof
sites
positive

Numberof
samples

Unfiltered,chlorinated

94

Directfiltration*

92

Conventionaltreatment

86

Commercialand/orPressure
filters

12

SlowsandandDiatomaceous
earth

Numberof
samples
positive

Percentof
samples
positive

16

17

1214

80

6.6

17

18.5

615

148

24.0

5.8

357

12

3.4

16.7

33

12.1

18

MechanicalorCartridgeFilters

13

53.8

51

11

21.6

Infiltrationgallery

16

31.3

37

18.9

Filtertypeunknown**

24

25.0

132

15

11.4

*Mostofthepositivesamplesoriginatedfromsystemsthatwerenotusingcoagulationpriortofiltration.
**Manyofthesesamplesarealsoincludedinthe83samples(Table1)wherethetypeofsample(rawoffinished)isalso
unknown.

Page243

Figure6.
Compositeofdirectfiltrationsystems,19791986(Chemicalpretreatment).

DatausedinFigures6and7wastakenfromdirectfiltrationsystems,generallyinthefallandwintermonths.Mostoftheseoriginatedfromsiteswithsurfacewater
temperaturesoflessthan1tonear5Candturbiditiesoflessthan0.5tonear1.0NTU.EachdatapointofFigures6and7representsaminimumof4anda
maximumof8rawandfinishedsamples.Dataareexpressedasthepercentageofturbidityreductionandthepercentageofcystsremovedinthefiltrationprocess.
Figure7revealstheproblemsassociatedwithremovalofGiardiacystsfromcoldwater,lowturbiditysourcesinactualfiltrationplantsituations.Figure6isan
excellentdemonstrationoftheincreasedefficiencywhenalumand/orpolymersisadded.Surprisingly,removalof8085%oftheincomingturbidityeffectively
removedthecystsifchemicalswereaddedwhileremovalof8095%oftheturbiditydidnotremovecystswhennochemicalswereadded.
DatainFigures6and7wasdevelopedfromfiltrationplantswithcystsintherawwateranumberofplantsusingdirectfiltrationwithoutchemicalpretreatmentarenot
includedbecausenocystswerefoundintherawwaterhowever,analysisfortheriskmaterialindicatesmostofthesesystemswereatriskGiardiacystssimplywere
notpresentatthetimeofsampling.
Additionofchemicalsineffectivecombinationsand/oramountswasdifficultwhenwaterconditionswerefluctuating.Forexample,municipalitiesswitchingtodirect
filtrationwhentemperaturesdecreasedinthewinter,andbeforeswitchingtoconventionaltreatmentinthespringwhentemperaturesincreased,wereoftenatriskfor
severaldaysorevenweeksbecausetemperatureandpHwerewidelyfluctuating.
ConventionalTreatment
Ofthe2372finishedwatersamplesexamined,357/2372(15%)camefromsitesusingconventionaltreatment.Atotalof86sitesusedconventionaltreatmentatleast
partoftheyearand5/86(5.8%)ofthesesitessometimeshadcystspresentinthefinishedwater(Table3).Atotalof75/357(21%)ofthesamples

Figure7.
Compositeofdirectfiltrationsystems,
19791986(NoChemicalpretreatment).

camefromthese5sitesand12/75(16%)hadcystsinthesample.Samplesobtainedfrom3ofthese5sitesappearedtobefromdirectfiltrationwithoutchemical
pretreatmentbecausenothingwasbeingremovedfromtherawwaterthereforethesethreesampleswerenotincludedinFigure8.
ConventionaltreatmentdataindicatethatthisprocedureissuperbforremovingGiardiacystsandallparticulatesthesizeoforlargerthanGiardia(Figure8).
Generallyweattemptedtodiscouragemunicipalitiesfromsendingsamplesbecausetheygenerallywerewastingtheirmoneyandourtime.
Commercialand/orPressureFilters
Ofthe2372finishedwatersamplesexamined,33/2372(1.5%)camefromcommerciallymanufacturedfiltrationsystems.Oftenitwasdifficulttodetermineifthe
systemwasapressurefilterorasmallcommercialfilterdesignedsimilartoadirectfiltrationsystembecauseitwasusuallylistedbycompanynameorsimplyas
"commercial","readybuilt","metalbox",etc.Atotalof4/33(12%)ofthesefilterswerepassingcysts(Table3).Obviously33samplesisaninadequatenumberto
useforanyconclusionsregardingtheirefficiencyhowevermostofthesamplesanalyzedfromthesesystemsindicatedthatthesystemswereatriskbecausetheywere
notremovingparticulatesthesizeoforlargerthanthecysts.
SlowSandandDiatomaceousEarth
Only11samplesfromslowsandfilters(onesite)and7samplesfromdiatomaceousearthfilters(twosites)wereexaminedforGiardiacystsandwhilethisistoo
smallanumbertoprovideahighdegreeofconfidence,bothtypesofsystemsperformedsuperbly(Table3).Oneofthediatomaceousearthsystemswasinplaceona
guestranchusingriverwaterfromasourcethatwasalwayscontaminatedwithcysts.Theslowsandsamplesalsocamefromasourcethatusuallyhascystsintheraw
water.
CartridgeorMechanicalFilters
Severaltypesofmechanicalfilterswereemployedand11ofthe51

Page244

Figure8.
Compositeofconventionaltreatmentsamples,19791986.

(21.6%)samplesexaminedhadGiardiacysts(Table3).SomeofthesefiltersremovedallparticulatesthesizeoforlargerthanGiardia,othersdidnotfunctionas
intendedbythemanufacturer.Moreoftenthannot,theproblemwasnotwiththefilter,orthecartridges,buttheoperatorsandthesalespersonnel.Sometimesthe
cartridgeswerenot1micron,materialintheeffluentindicatedabouta10micronporosity.Othertimestheoperator"stacked"9.75inchwoundfiltersindrumorkettle
typehousingswithoutusingspacerstosealindividualcartridges(spacersoftenarenotneededforpleatedtypecartridges,onlythewound).Sometimesthelidorcover
wasnotsufficientlytighttomakeagoodseal,orgasketsneededreplacementand/oralignment.
InfiltrationGalleries
Someoftheinfiltrationgalleriesweresuperb,otherswerenotand7ofthe37(18.9%)sampleshadGiardiacysts(Table3).Mostofthesewereusedastheonly
meansoffiltration,otherswereinplaceaheadofeithermechanicalfilters,pressure,orcommercialfiltersandonewastheinitialscreeningmechanismbefore
conventionaltreatment.Mostofthemappearedtohavehadsimilarconstructionandusedasimilardistributionofsand,gravelandrocksforfiltration.
FilterTypeUnknown
Only15ofthe132(11.4%)sampleswherethetypeoffiltrationwasnotgivenwereinfectedwithcystsofGiardia(Table3).Samplesofthisnaturearenotvaluable
foranalysiswhenimportantinformationhasnotbeenincluded.However,theyrepresentonly3%ofthetotalsamplesexamined.
Conclusions
Anumberoflimitingfactorsmustbeconsideredintheinterpretationoftheseresults:1)efficiencyofthediagnostictechniquesusedoverthelast12yearsprobablyis
themostimportantofthesefactorsbecauseresultsareonlyasgoodasthetechniqueemployedandthetechniques,despiteimprovementsareonlyabout50%
effective2)turbidityandtypeofturbidity3)timeofyear4)timeofdaysamplesweretaken(someanimalsareactiveonlyatnight)5)handling,shippingthe
samples,thetimefactor6)eyefatigue,adirectresultoftheabovefactorsand7)qualityofcystsinthesample.
GiardiacystsprobablyarepresentinmostifnotallofthesurfacewatersourcesintheUnitedStates.Repeatedsamplingofnegativesourcesusuallyprovidedpositive
results.Ahighpercentageofthenegativesourcessentonly12samples.Giardiacystsarenotuniformlydistributedinsurfacewatersand1or2samplesfromasite
isinadequate.
Thesourceofsurfacewatercontaminationsometimesisdifficulttodetermine,atothertimesthesourceisclearlyhumansewageorclearlywildordomesticanimal
unfortunately,inseveraloutbreakswildanimals(beaver)wereincriminatedandfoundguiltywithoutatrial.Investigatorsseemreluctanttoplaceanyblameonsewage.
ManandmanywildanddomesticanimalsfrequentlyareinfectedwiththeGiardiaduodenalistypeparasite.Theanimalsresponsibleformostofthecystsfoundin
surfacewaterarehumans,especiallyindirectlythroughsewage,beaver,muskrat,otherlargeandsmallrodents,cattle,andmoose.Thislaboratoryandother
investigators(Erlandson,personalcommunication)havefoundinfectioninahighpercentageofherons.Therisktohumansisunknownbuttotheanalyst,itisacyst
anditisinwater.Whileithasbeenfoundindeer,elk,horses,dogs,cats,coyotes,wolves,etc.,theylikelydonotconstituteasignificantthreatbecauseoftheirhabits
andhabitats.
Watershedmanagementgenerallyispracticedbymunicipalitieswhentheyownorcontrolthereservoirshowever,manysitesobtainwaterfrompublicownedsources
(e.g.,U.S.ForestService)thatpracticethemultipleuseconcept(agriculture,forestryandrecreation).Managementofthesewatershedsformunicipalwatersupplies
isnotalwaysconsistentwiththepolicyoftheagencyincontrol.Grazingpermits,timbermanagement,fishandwildlifemanagementandrecreationaluse(campsites,
waterskiing,etc.)mayindirectlyaffectandprohibiteffectivemanagementforpotentialgiardiasisproblems.Ifthecreekand/orrivertraversesprivatelyowned
(agricultural)land,andtheownerhasrightstoacertainpercentageofthatwater,managementisattheowner'sdiscretion.
Thenumberofcasesofgiardiasisdocumentedinbackpackers,hikers,campers,etc.whodrinkrawsurfacewaterdoesindeedpointoutthatwildand/ordomestic
animalscanbeasourceofinfection,althoughthelimitedfacilitiesavailableforpersonalhygieneshouldmakethecampcooksuspectinsomeoftheseinstances!
Wildanddomesticanimalsaresusceptibletoinfectionwithhumansourcecysts.Likelythereverseistrueinmanybutnotallinstances.Toomanycrosstransmission
trialshavefailedwhenattemptshavebeenmadetofurthercrosstransmittheorganism.Crosstransmissionresultsarenotalwaysreliableandshouldbeinterpreted
withcaution.Insomeoutbreakswildanimalsdefinitelyaretoblame,inothersthereissufficientreasonfordoubt.Wehavebeendirectlyorindirectlyinvolvedinmost
ofthe

Page245

outbreaksintheUnitedStates,andafewofthoseinCanada.Someofthesehavebeenpublicizedanddocumented,othershavenot.Theauthorisnotatlibertyto
providedetailsoftheunreportedoutbreaksonlytopointouttheresultsofhisinvestigations.Inonecommunityoutbreak,thesurfacewatersourceoriginatedinawell
protected,pristineenvironment.Thisoutbreakwaswrittenandpublishedbuttheepidemiologistsdidnotmakeanefforttoinvestigatethewatershed.Thislaboratory
didinvestigateaftertheoutbreaktheonlyanimalspresentonthewatershedwerebeaversandmuskratsandmostofthosewereinfectedandsheddingcysts.No
septicsystemorotherpossiblehumancontaminationwasobserved.Cystsinthefinished(filtered)waterwerealmosttoonumeroustocountoveraperiodofseveral
weeks.AratherlargeguestranchinthewesternUnitedStateshadapproximately100casesoveraperiodoftwoweeks.Acreekoriginatinginthealpinewasthe
sourceofwater.Thewaterwentthroughaninfiltrationgalleryandacommercialfilter(neitherofwhichremovedanyofthecysts).Thiscreekwasthoroughly
investigated:theonlyinhabitantswerebeaversandallwereinfected.Beaverswereremoved,thecystcountdroppedtozero,andtheoutbreaksubsided.The
infiltrationgalleryandthefilterarenomoreeffectivenowthanduringtheoutbreak.Whilethesetwooutbreaksclearlywereofanimalorigin,inmostoftheoutbreaksin
whichwehavebeendirectlyorindirectlyinvolvedthesource,humanoranimal,wasneverdetermined.Sewagewasalwaysagoodpossibility.
ThepresenceofGiardiacystsinafinishedwatersampledoesnotnecessarilyindicatethatanoutbreakisimminent.Manyofthesitesfromwhichthislaboratoryhas
receivedsamplesthepast12yearshadasmanyifnotmorecystsinthefinishedwaterthansitesinthemidstofanoutbreakmoreover,thechlorineand/orcontact
timewasminimal(oftennone!)yetnocasesofgiardiasisotherthantheusualbackgroundcaseswerepresent.Thesecystswerealive:frequentlytheywereintroduced
intomongoliangerbils(Merionesunguiculatus)andcausedinfection.Insomeofthesmallercommunitiesallofthenativesprobablywereimmune,butforother
communitiestheindicationisthatthesecystswerenotinfectiousforhumans.Severalofthesesiteshavebeeninvestigatedandtheanimalspresentinvariablywere
beaversandmuskrats,usuallylivingpeacefullyatthemunicipalintakes.Obviouslytherearestrains(?)ofG.duodenalisthatdonotreadilycrossinfectbetween
animalsandhumans.Unfortunately,thereisnomeansofdeterminingwhichsourceswillandwhichsourceswillnotinfecthumans.
Acknowledgements
IshouldliketoacknowledgeMr.WalterJakubowski,USEPA,Cincinnati,OhioandDr.MartinJ.Allen,AmericanWaterWorksAssociationResearchFoundation,
Denver,Coloradofortheunrelentingbutalwaysdiplomaticpressuretheyappliedtoconvincemetocompilethesedatawithoutcompromisingthemunicipalsites,our
customerswhosentthesesamplesforanalysis.ThisstudywasfundedbyMr.StigRegli,USEPA,OfficeofDrinkingWater,Washington,D.C.Tomymanygraduate
students,whooftenspenteveningsandweekendsatthemicroscopeanalyzingthesesamples,youknowhowmuchIappreciateyourefforts.Overtheyearsthese
sampleshavebeenprocessedandanalyzedby:Dr.RobertDavies,Mr.SteveHenry,Dr.DonMonzingo,Dr.JohnWegrzynandMs.DonnaHowell.Ms.Carrie
Hancock,ResearchAssociate,andMs.DianeSwabby,LaboratoryCoordinator,havebeenandcontinuetobeanintegralpartofthestaff.

Page247

VIABILITYTESTING

Page249

AReviewofMethodsthatAreUsedtoDetermineGiardiaCystViability
FrankW.Schaefer,III.
HealthEffectsResearchLaboratory,UnitedStatesEnvironmentalProtectionAgency,26WestM.L.KingDrive,U.S.A..
Overthepast55yearsanumberofmethodshavebeendescribedforexcystingGiardiacystsasameansofdeterminingviability.TheexcystationmethodsforG.muriscystsare
reliableandreproducible.However,methodspublishedtodatefortheexcystationofG.lambliacysts,thehumanpathogen,havenotyieldedreliable,reproducibleresults.Forboth
GiardiacysttypesthestimulatoryfactorspromotingexcystationarelowpH,carbondioxide,atemperaturearound37C,andafinalneutralizingstepatpH7.0.Theseexcystation
promotersinducenumerousectoplasmicvacuolestodumptheircontentsbetweenthetrophozoiteandthecystwall.Unfortunately,largenumbersofcystsarerequiredforallthe
excystationmethods.Recently,othermethodshavebeendevelopedfordeterminingGiardiacystviability.Theseincludeseveralfluorescentvitaldyes,differentialinterference
contrastmicroscopy,andMongoliangerbilinfectivitytests.Thefluorescentvitaldyeanddifferentialinterferencecontrastmicroscopymethodsarenewandmayrequirevalidation.
Theanimalinfectivitymethodhasseveraldisadvantages.NotallG.lambliacystisolatesfromhumanswillinfectgerbils.Cystisolateswhichdoinfectdonotnecessarilyproduce
cysts.Thisrequiresanimalnecropsytoverifytheinfection.Also,thelowertheG.lambliacystdosage,thelowertheprobabilityisofinfectionbeingproduced.Clearly,thereisa
greatneedforfurtherstudyinthisareaofGiardiaresearch.

Introduction
Giardiaspeciesareflagellatedprotozoanparasitesfoundintheintestinaltractofmanyvertebrates.Twoformsoftheparasiteareusuallyfoundwithinthehost's
intestinaltract.Thetrophozoite,whichisanactive,vegetativeform,livesintheupperthirdofthesmallintestine.Astrophozoitesgetcaughtintheluminalflowof
nutrients,theyaresweptdowntheintestinaltractandareinducedtotransformintocysts.Thecystisadormant,transmissionformoftheparasite.Transmissionis
directbythefecaloralroute.Thematurecystisaroundedtrophozoite,whichissurroundedbyacystwall,andhascompletedkaryokinesis,butusuallynot
cytokinesis.Thetransformationfromtrophozoitetocystisusuallycompletebythetimetheorganismreachesthedistalportionofthesmallintestine.Besides
transmissionbycontaminatedfood(2,28)andpersontopersoncontact(6,22,27),transmissioncanoccurbyingestionofcontaminatedwaterandhasbeen
responsiblefornumerouswaterborneoutbreaksintheUnitedStates(9,24).
Determiningtrophozoiteviabilitygenerallyisnotaproblem,becauseflagellarandcaudalmovementcanbedetectedbyconventionalmicroscopy.Cysts,however,do
notexhibitrapidmovementunlesstheyareinducedtoexcyst.Sincecystsaretheusualinfectiveform,itisespeciallyimportanttoknowwhethercystsareviableor
not.Viabilityinformationonacystpopulationisnecessaryforcystdisinfectionstudiesandindeterminingriskassessmentsforcontaminateddrinkingwatersupplies.
ExcystationprocedureshavebeeninstrumentalinthesuccessfulcompletionofGiardiacystdisinfectionstudies(16,20,37).
TodatetheproceduresfordeterminingGiardiacystviabilityhaveincludedinvivoexcystation,invivoinfectivity,invitroexcystation,vitalstains,anddifferential
interferencecontrastmicroscopy.Thefollowingdiscussionwilladdressvariousaspectsoftheseviabilitytestingprocedures.Amoredetaileddiscussionofexcystation
maybefoundinthereviewbyMeyerandSchaefer(26).
Discussion
Excystation
Hegner(13,14,15)describedinvivoexcystationofhumansourceGiardiacyststhathadbeeninjectedintothestomachof,andlaterremovedfrom,rats.Hegner
notedthatexcystationoccursinthejejunumandileumandspeculatedthatthefactorsresponsibleforexcystationincludeatemperatureof37Candmoisture.Healso
suggested(15)thathostdigestivejuicesareunnecessaryforexcystation.
Armaghan(1)attemptedtodeterminethesiteofexcystationbyplacingG.muriscystsuspensionsinvariouspartsoftheratgastrointestinaltract(stomach,duodenum,
ileum,andcecum)bylaparotomyfollowedbyinjectiontodeterminethesiteofexcystation.Excystationwasconsideredpositiveifcystscouldbeisolatedfromthe
fecalmaterialand/ortrophozoitescouldbeisolatedfromthegastrointestinaltractpostexposure.Infectionsweredetectedonlyinthoseratsinjectedeitherinthe
stomachorduodenum.
ThedevelopmentofinvitroexcystationmadetheapplicationofquantitativeproceduresforassessingsurvivalofGiardiacystsexposedtochemicalfactorseasier.
ExcystationhasbeenquantifiedtwowaysdependingprimarilyonwhetherG.lambliaorG.muriscystswerebeingexcysted.Binghametal.(4)andRiceand
Schaefer(30)determinedthepercentageexcystationofG.lamblia

Page250

cystsbycountingthenumberofintactcysts(IC),partiallyexcystedtrophozoites(PET),andtotallyexcystedtrophozoites(TET)andapplyingthefollowingformula:
%excystation=(TET/2+PET)+(TET/2+PET+IC)100.
Thenumberoftotallyexcystedtrophozoiteswasdividedby2inthisformula,becauseeachexcystedorganismyieldedapairoftrophozoites.Emptycystwallswere
notcounted,becausetheyweredifficulttodetect.EachexcystationpercentagederivedbyBinghametal.(4)involved3to5separatecountsinwhichanaverageof
708cystswerecountedtoobtaineachvalue.Feely(11)hasusedthisprocedureforG.murisexcystation.Ontheotherhand,Schaeferetal.(34)countedfulland
emptyG.muriscystwallsratherthantotallyexcystedtrophozoites,becauseemptycystwallswereeasiertocountthanswimmingtrophozoites.Theycalculated
percentageexcystationusingthefollowingequation:
%excystation=(ECW+PET)(ECW+PET+IC)100,
whereECWisthenumberofemptycystwalls,PETisthenumberofpartiallyexcystedtrophozoites,andICisthenumberofintactcysts.Eachexcystation
percentagedeterminationinvolvedatotalof100cysttypes.TheseworkersdeterminedthethermaldeathpointofG.muriscystsusingthisprocedure.Atexposure
temperaturesof50Cand52Cnoemptycystwallswerefoundoncounting100cysts.If,however,theentireslidewhichheldapproximately100,000cystswas
scannedformobiletrophozoites,somewerefoundatbothofthesetemperatures.Thisindicatedthatcountingonly100cystsdidnotsamplealargeenoughportionof
thepopulationtodistinguishpercentagesoflessthan1.Indeterminingpercentageexcystationbyeitherquantitativeprocedure,thereisnouniversallyaccepted
minimumnumberofcyststhatmustbecountedtoinsureprecisionandaccuracyoftheresult.Furtherresearchonthispointisneeded.
BinghamandMeyer(5)reportedthefirstsuccessfulinvitroexcystationofGiardiacystsfromhumans,monkeys,anddogs,andG.muriscystsfromratsandmice.
Theirprocedure,whichwascarriedoutat37C,consistsofthreesteps:alowpHinductionstepfor1hour,awashsteptoremovetheacid,andanincubationstep
for1hourinanutrientmediumtocompletetheexcystationprocess.Syntheticgastricjuice(pepsinandhydrochloricacid),aqueousacids,andwaterweretriedas
inductionsolutions.Forincubationmedia,theyusedwaterandHSP3(Hanks'balancedsaltsolution,phytone,serum,andNCTC13525).Excystationoccurred
whentheinductionsolutionwassyntheticgastricjuiceoraqueousacid,andtheincubationmediumwasHSP3.Saltsandpepsininthesyntheticgastricjuicedidnot
significantlyaltertheexcystationlevelindicatingthattheacidalonewascriticaltotheprocess.Althoughanumberofinorganicacidsweretestedininductionsolutions
andwerefoundeffective,hydrochloricacidwasselectedforroutineuse.Peakexcystation(2226%)occurredwhenthepHwasbetween1.3and2.7intheinduction
solutionandtheincubationmediumwasHSP3.
Binghametal.(4)studiedotherfactorsintheexcystationprocessofGiardiacystsisolatedfromasinglehumandonor.Ofthesefactors,time,temperature,pH,and
incubationmediumwereshowntoaffectexcystationlevels.TheinductionsolutionexposuretimedecreasedasthepHdecreased.Ahydrochloricacidinduction
solutionatpH2.0onlyrequiredanexposuretimebetween10and30minutes.Thehighestexcystationlevelsalwaysoccurredwhenboththeinductionsolutionand
theincubationmediumwereat37C.ThepHoftheHSP3incubationmediumwascritical,too.NoexcystationoccuredwhentheHSP3wasinthe0.5to4.0
rangeexcystationstartedatpH6.2andwasbestatpH6.8.TheexcystationofacidinducedcystsinvariouscomponentsofHSP3mediumandcompleteHSP3
mediumwascompared(A.K.Bingham,M.S.thesis,OregonHealthSciencesUniversity,Portland,1979).SignificantlylowerexcystationlevelsoccurredinHanks'
phytonebroththaninHanks'phytonebrothplusserumorcompleteHSP3medium.ThefactorsBinghamandMeyer(5)foundfavoringinvitroexcystationclosely
approximatedthehost'sinvivoenvironmentinthestomach(inductionstep)andtheduodenum(incubationstep).Theyalsoreportedthataperiodofcystmaturation
varyingfrom2to7dayspostisolationwasrequiredforG.lambliacystsbeforemaximalexcystationcouldbeachievedwiththeirprocedure.
RiceandSchaefer(30)reportedtheirexperiencewiththeBinghamMeyerprocedureforexcystingG.lambliacystsisolatedfromhumandonors.Theyattemptedto
usetheprocedureforG.lambliaandG.muriscystsbutusuallyobtainedonly23%excystation.TheRiceSchaeferprocedureforG.lambliacystsincludes
induction,wash,andincubationstepsliketheBinghamMeyerprocedure.Theinductionstepinvolvestheadditionofhydrochloricacid(pH2),Hanks'balancedsalt
solutionsupplementedwithLcysteinehydrochlorideandglutathione,andsodiumbicarbonatetothecystpreparationinthatorder.Immediatelythereafter,thetube
containingthecystsiscappedtightly,mixed,andthenincubatedfor30minutesat37C.ThecystsarewashedbycentrifugationintrypsinTyrode'sincubationmedium
andresuspendedintrypsinTyrode'sincubationmediumwhichhasthepHadjustedto8.0withsodiumbicarbonate.Theresultantsuspensionisincubated1hourat
37C.
In1981,RiceandSchaefer(30)reportedtheresultsofthisprocedureusedin28G.lambliaexcystationtrials.Tentrialsusedcystsfromtwosymptomaticdonors,
while18trialsemployedcystsfromthreeasymptomaticdonors.Thelowestobservedpercentexcystation(40%)occurredwithcystsfromanasymptomaticdonor
thehighestpercentage(95%)occurredwithcystsfrombothsymptomaticandasymptomaticdonors.Themeanpercentageexcystationfromthesymptomaticdonors
(87%)washigherthanthatfromtheasymptomaticdonors(70%).ThelowestpercentageexcystationRiceandSchaeferobservedwas10%higherthanthemaximum
reportedbyJarrolletal.(19)usingtheBinghamMeyerprocedure.

Page251

ThesedataindicatethattheRiceSchaefermethodmaybesuccessfullyusedtoexcystG.lambliacystsfrombothasymptomaticandsymptomaticdonors,andthat
higherpercentagesofexcystationmaybeexpectedfromsymptomaticdonors.TheyalsoconfirmedBinghametal.'s(4)observationthatG.lambliacystsexhibited
higherexcystationratesafter7daysmaturation.
Schaeferetal.(34)reportedtheresultsofexcystationexperimentsemployingG.muriscystsisolatedfrommousefeces.Theirprocedureresultedinexcystationlevels
consistentlygreaterthan90%ascomparedtolevelsoflessthan5%withtheBinghamMeyerprocedure.AlloftheG.murisexcystationproceduresconsistofthe
samesteps:induction,wash,andincubation.ThemaindifferenceintheSchaeferetal.procedurewasthatinductionwaspromotedbyexposureofG.muriscyststo
aninductionsolutionconsistingof1partreducingsolution(Hanks'balancedsaltsolutionsupplementedwithglutathioneand1cysteinehydrochloride25)and1part
0.1Msodiumbicarbonatefor30minutesat37Cinasealedtesttube.Inthiscase,noinorganicacidisused.Whentheinductioncomponentsaremixed,carbon
dioxideisevolved,thepHis2,andtheoxidationreductionpotentialis120mV.Agradualdeclineinexcystationto2%orlessoccurredasthepHandtheoxidation
reductionpotentialwerechangedto7and57mV,respectively.TheinductionstepwasfollowedbyawashandincubationstepbothcarriedoutintrypsinTyrode's
solution(30).ThetrypsinTyrode'ssolutionisreportedbytheseinvestigatorstobecrudeandtedioustomake.Thetrypsintheyusedwasanimpurelyophilized
extractofhogpancreaswhichdoesnotcompletelydissolveaftervigorousmixinginTyrode'ssolutionfor30minutesatambienttemperature.Theundissolvedtrypsin
wasalwaysremovedbyhighspeedcentrifugationfollowedbypositivepressurefiltration.WhenthetrypsinTyrode'ssolutionwasboiled,G.murisexcystations
declinedtothe70thpercentile,indicatingtheenzymaticactivityofthesolutionisnotcrucialandconfirmingHegner's(15)andBinghamandMeyer's(5)observations
thatdigestiveenzymesarenotobligatory.AttemptsbySchaeferetal.tomaketrypsinTyrode'ssolutionwithpurifiedpancreatinresultedinlittleifanyG.muris
excystation.SimilarresultswereobtainedbyMarchin(personalcommunication,1982)whenheusedpurifiedtrypsininhistrypsinTyrode'ssolution.Onlywhenhe
substitutedcrudetrypsinforpurifiedtrypsinwasheabletoobtainG.murisexcystationsinthe90thpercentile.Thesedataimplynutrientsweresuppliedbythe
complextrypsinTyrode'sincubationmediumthatwerenotsuppliedbythehighlypurifiedsolutions.
Feely(11)recentlypublishedaprocedurewhichheusedtoexcystG.muriscysts.InductionofthecystswasdoneinHanks'balancedsaltsolutionatpH2for30
minutesat37C.ThisprocedurespecifiesthattheHanks'balancedsaltsolutionmustbepreparedfreshdailyandthepHadjustedtoneutralitywithsodium
bicarbonatebeforebeingacidifiedtopH2with2Nhydrochloricacid.Thisallowsslowliberationofcarbondioxidefromtheinductionmediumduringtheincubation
steplikeintheRiceandSchaefer(30)andSchaeferetal.(34)methods.ThewashisinHanks'balancedsaltsolutionatpH7.2followedbyincubationinTYImedium
(salts,trypticase,yeastextract,glucose,cysteinehydrochloride,ascorbicacid,ferricammoniumcitrate,bovinebile,andserum21)at37Cfor30minutes.No
exogenousenzymesarerequiredinthisprocedure.Theexcystationratesreportedforthisprocedurewerealwaysgreaterthan90%.Theadvantagesreportedforthis
procedurearethatiteliminatesthetediouslypreparedtrypsinTyrode'sincubationmediumpreparationandrequiresnoreducingagents.
Intheseproceedings,Sauch(33)reportsproceduresforexcystingG.lambliaandG.muriscystsinproteinfreemedia.HerinductionsolutionforG.lambliacysts
wasthesameasRiceandSchaefer's(30)andforG.muriscystswasthesameasSchaeferetal.'s(34)however,exposuretotheinductionsolutionat37Cwas
increasedfrom30to45minutesinthecaseofG.lamblia.Inherprocedures,trypsinTyrode'ssolutionwasreplacedwithHanks'balancedsaltsolution
supplementedwithcysteine,sodiumbicarbonate,andeitherproteosepeptoneforG.muriscystsorphytonepeptoneforG.lambliacysts.Comparisonofher
methodswiththatofRiceandSchaeferforG.lambliacystsandthatofSchaeferetal.forG.muriscystsshowednosignificantdifferencesinpercentexcystation.
Theadvantagereportedforthisprocedureisthatexcystationcanbecompletedintheabsenceofenzymeswhichmaydestroyantigensoncysts.
ExcystationofG.lambliaandG.muriscystsisanactiveprocessonthepartoftheparasite(5,7,8,34).Thecaudalflagellaanddistalendsoftheotherflagella
protrudefromoneendofthecyst.Flagellarmovementstartsslowlybutrapidlyincreaseswithinthefirstfivetotenminutesoftheiremergence.Therapidflagellar
movementseemstohelppulland/orbreakthetrophozoiteoutofthecystwall.ScanningelectronmicroscopyofthisprocesshasshownatearinthecystwallofG.
muriscystsfromthepolaropeningtowardtheoppositecystpole(7).Thisopeningappearedtobesubsequentlyenlarged,presumablybyflagellaraction.
TransmissionelectronmicroscopyofinducedG.muriscystsbyCogginsandSchaefer(8)hasshownacytoplasmdevoidofendoplasmicreticulum,golgibodies,and
mitochondria.Inuninducedcysts,largemembraneboundectoplasmicvacuolesareseen.Afterinductiontheseectoplasmicvacuolesappeartodumptheircontents
intotheperitrophicspace.Speculation(8)thatthecontentsofthesevacuolesareenzymaticissupportedindirectlybycurrentinformationconcerningthepresenceof
lysosomalenzymes(10,23)alongwithevidenceformucusmaterialsecretedduringexcystation(5,7).MoreconclusiveevidencewasrecentlyreportedbyFeelyand
Dyer(12)whodemonstratedacidphosphataseactivityintheectoplasmic(peripheral)vacuolesofG.lambliaandG.muristrophozoitesbycytochemicallightand
electronmicroscopy.Newlyemergedtrophozoitesappeardisorganized,ovalinshape,andassociatedwithmucoidlikematerial.Quadrinucleatetrophozoitesquickly
begintoflatten,elongate,andundergocytokinesiswithin30minutesofemergencetoformbinucleatetrophozoiteswithcompletelyorganizedadhesivedisks.The
characteristicectoplasmicvacuoles

Page252

documentedinmatureGiardiacystsandtrophozoitesstillarerareinexcystedtrophozoitesundergoingcytokinesisandindaughtertrophozoites30minutesintothe
incubationperiod.Theseobservationsareconsistentwiththesuggestionthattheseectoplasmicvacuolesplayaroleinexcystation.
AnumberofconclusionsmaybedrawnfromtheexcystationstudiesdonetodateonG.lambliaandG.muriscysts:1)thetemperatureoftheinductionand
incubationreagentsshouldbe37C2)theinductionmediumshouldhaveapHaround2andevolvecarbondioxide3)thewashstepshouldhelpneutralizetheacidity
fromtheinductionstep4)theincubationmediumshouldcontainnutrientsbutneednotcontainserum,bile,ortrypsin5)presently,noobligatoryneedforexogenous
enzymesintheexcystationprocesshasbeendemonstrated6)acystmaturationperiodisnotneededforG.muriscysts7)acystmaturationperiodforG.lamblia
excystationisneeded8)G.muriscystscanberoutinelyexcystedwithefficienciesinthe90thpercentile9)G.lambliaexcystationefficienciesareerraticandusually
muchlessthan90%and10)excystationisaprocessrequiringactiveparticipationbytheparasite.
OtherMethods
Giardiacystdensitiesinsurfacewatershavenotbeenreportedtoanygreatextent.OnestudywhichgathereddatafromthestatesofOregon,Idaho,Wyoming,and
Pennsylvaniareporteddensitiesrangingfrom0.05to680Giardiacystsper380liters(Craun,G.F.andW.Jakubowski,WaterResour.Bull.,inpress).Thisindicates
thatGiardiacystdensitiesareusuallylowerthanathousandcystsper380litersofsample.EvenifalltheGiardiacystscouldberecoveredefficientlyfromthe
sample,whichtheycannot,therewouldbeinsufficientcystnumberstodetermineviabilitybycurrentexcystationprocedureswhichrequirebetween1and5105
cysts.Inaddition,theproceduresforexcystingGiardiacystsrequireseveralhourstocompleteandarenotequallysuccessfullyineverylaboratory.Theseproblems
haveforcedthedevelopmentofotherviabilitymethodswhichcanbeusedwithsmallnumbersofcysts.Amongthesealternatemethodsarevitaldyes,fluorogenic
dyes,differentialinterferencecontrastmicroscopy,andanimalinfectivityprocedures.
Binghametal(4)comparedtheBinghamMeyer(5)procedureforexcystingG.lambliacystswitheosindyeexclusionandfoundthattherewasnocorrelation
betweenthetwotechniques.Theeosindyeexclusiontechniquealwaysindicatedgreaterviabilityinthecystpopulationthanexcystationdid.Thiscouldhavebeenfor
variousreasons.Theexcystationtechniquemaynothavebeenoptimal.Itisalsopossiblethattherewerecystsinthepopulationthatwerealivebutcouldnotexcyst.
Furthermore,thecystsuspensioncouldhavecontaineddeadcystswhichexcludedthedye.
Intheseproceedings,Hudsonetal.(18)reportfluorescentdyeexclusionasamethodfordeterminingGiardiacystviability.Inthisprocedure,asintheeosin
procedure,livecystsexcludethedye.TheseworkersutilizedFluoraBoraI(3(dansylamido)phenylboronicacid)inG.murisandG.lambliacystsuspensionsand
comparedtheresultswithexcystation.TheyfoundahighdegreeofcorrelationbetweenthedyeexclusionandexcystationmethodsinthecaseofG.muriscysts.In
thecaseofG.lambliacysts,however,thecorrelationwaslow.ThisagainsuggeststhatviabilityasdeterminedbyexcystationofG.lambliacystsisbeingeither
overestimatedorunderestimated.
SchuppandErlandsen(Abstr.Annu.Meet.Am.Soc.Parasitol.,1986,16,p.35)havedevelopedamethodofdeterminingG.muriscystviabilitywhichemploys
fluoresceindiacetateandpropidiumiodidestains.Viablecystsfluorescegreenduetouptakeandmetabolicconversionofthestaintofluorescein.Deadcystsfluoresce
orangetobrightreddependingontheexcitationwavelength,duetostainingbypropidiumiodide.Theirresultswithfreshlyisolatedcystsrevealedgreaterthan85to
90%ofthecystpopulationstainedgreenandlessthan10%ofthecystsdidnotstainatall.Freshlyisolatedfluoresceindiacetatepositivecystsandcyststhatdidnot
stainwitheitherdyewereinoculatedintoseparategroupsofmice.Similarly,heatkilledpropidiumiodidestainedcystswereinoculatedintoanothergroupofmice.
Infectionsoccurredinthosemicereceivingeitherfluoresceindiacetatestainedorunstainedcystsbutdidnotoccurinthosemicereceivingthepropidiumiodidestained
cysts.Necropsiesofthemiceexposedtoeitherfluoresceindiacetateorpropidiumiodidestainedcystsalwaysrevealedtrophozoitesinthefluoresceindiacetate
exposedmicebutneverinthemicereceivingpropidiumiodidestainedcysts.Theseinvestigatorshavesaidthatfluoresceindiacetatepositivestainedcystsarecapable
ofexcystation,buttheydidnotprovidedetails.Thesedatademonstratethatfluoresceinstainedcystswereviableasdeterminedbothbystainingandinfectivityand
possiblybyexcystationaswell.RotmanandPapermaster(32)reportedthatenzymatichydrolysisoffluorogenicestersintissueculturecellsvariedasmuchas80
fluoresceinfluorimeterunits.SomevariabilityinthefluorescenceoffluoresceindiacetatestainedG.muriscystshasbeennotedbySchuppandErlandsen.
Intheseproceedings,Schuppetal.(35)comparedthelightmicroscopicmorphologyoffluoresceindiacetatestainedG.muriscyststopropidiumiodidestainedG.
muriscysts.Viable,fluoresceinpositivecystshadclearlydefinedcystwalls,aperitrophicspace,andflagellaatonepole.Occasionallyaxonemeswerealsoobserved,
butnoothercytoplasmicorganelleswereseenintheseviablecysts.Ontheotherhand,innonviable,propidiumiodidestainedcyststhecytoplasmicorganelles
(axonemes,medianbodies,portionsoftheadhesivedisk,andnuclei)wereseen.Theperitrophicspaceinnonviablecystswasnotobserved,becausethecytoplasm
appearedtobecloselyadherenttothecystwall.Thesedifferenceswereobservedwithbothdifferentialinterferencecontrastandbrightfieldmicroscopy.These
observationsimplythatstructuraldifferencesexistbetweenthesetwotypesofG.muriscystsandmaybeausefulwaytoassessviability.
BoththefluoresceindiacetatepropidiumiodideandtheFluoroBoraIviabilitystainingmethodsarenot

Page253

selectivestainingmethods.ThesemethodsareeasilyusedwithpureorhighlyconcentratedGiardiacystsuspensions.However,incystsuspensionscontainingother
macroandmicroorganisms,thereactionoftheGiardiacystscouldeasilybemaskedbytheotherlivingorganisms.Thereisaneedtolinkselectiveidentification
methodswiththeseviabilitymethods.
Animalmodelsofgiardiasishavebeendevelopedinthemouse(31)andtheMongoliangerbil,Merionesunguiculatus(3).Bothmodelshavebeensuccessfullyused
todeterminecystviabilityandinfectivityhowever,bothhaveseveralunsatisfactoryaspects.Neithermodelmanifestsovertpathologylikethatseeninsomehumans.
ThemousemodelproducesG.muriscystscontinuouslyforseveralweekspostexposure,unlikethehumancondition,whichproducescystsintermittently.Although
theMongoliangerbilintermittentlyexcretescystsforseveralweekspostexposure,likehumansdo,notallhumanderivedG.lambliacystswillinfectgerbils
(Visvesvara,G.S.,J.Dickerson,andG.R.Healy,Abstr.Ann.Meet.Am.Soc.Trop.Med.Hyg.,1983,K139,p.117).SomehumanderivedG.lambliacystsinfect
thegerbilasevidencedbytheproductionoftrophozoitesbutnotcysts.Between1and15cystsarenecessarytoinfectamousewithG.muriscysts(17)andasfew
as10cystsarerequiredtoinfectmanwithG.lambliacysts(29).InthecaseofMongoliangerbils,theprobabilityofinfectionisdirectlyrelatedtothedosage
(Visvesvara,G.S.,J.Dickerson,andG.R.Healy,Abstr.Annu.Meet.Am.Soc.Trop.Med.Hyg.,1983,K139,p.117)cystdosagesshouldbegreaterthan100
cystspergerbiltoinsure70%ofthegerbilsbecomeinfected,whenG.lambliacystsareused.SimilarresultshaverecentlybeenobtainedwhenG.lambliacysts
fromhumansweregiventobeavers(Bemrick,W.J.,S.L.Erlandsen,L.A.Kemp,L.F.Sherlock,andD.G.Schupp,Abstr.Ann.Meet.Am.Soc.Parasitol.,1986,
96,p.55).WallisandWallis(36)recentlysuccessfullyinfectedgerbilswithGiardiacystisolatesfrommeadowvoles,humans,dogs,andbeavers.Bothcystsand
trophozoiteswerecapableofinfectinggerbils.Overallratesofinfectionwere89%formeadowvoles,46%forhumanisolates,50%fordogisolates,and91%for
beaverisolates.Cystdosagesrangedbetween4000and1.5millioncystsperanimal.TodatetheMongoliangerbilisthebestexperimentalhostforG.lamblia,but
thesystemissubjecttoerrors.Justbecauseaninfectiondidnottakeinthegerbildoesnotmeanthatthecystsusedwerenotaliveorinfectivetohumans.
ThemethodsusedtodetermineGiardiacystviabilityincludeddyeexclusion,fluorogenicdyes,excystation,andanimalinfectivity.Therearedisadvantagestoeachof
thesemethods.Excystationrequireslargenumbersofcystsand2to3hourstocomplete.G.lambliacystsdonotexcystroutinelyathighlevels.Animalinfectivity
studiesrequireweekstocomplete.Furthermore,animalsareexpensivetobuy,feed,andhouse.Also,onlypositiveresultscanbeconclusivelyusedinanimal
infectivitystudies.Eosindyeexclusiondoesnotcorrelatewellwithexcystation.FluorogencdyesseemtocorrelatewellwithG.murisexcystationbutnotwithG.
lambliaexcystation.MorphologicaldisparitybetweenlivinganddeadcystscorrelatewellwithinfectivityinG.muriscysts.Todate,however,thistechniquehasnot
beenreportedforG.lambliacysts.Ifdifferentialinterferencecontrastmicroscopyisusedtodifferentiatelivingfromdeadcysts,alargeinvestmentisrequiredforthe
optics.Whilefluorogenicdyesanddifferentialinterferencecontrastmicroscopyarepromisingtechniquesfordeterminingcystviability,theyrequirefurtherevaluation.
Clearly,thereisagreatneedforfurtherstudyinthisinterestingareaofGiardiaresearch.
Notice
ThisdocumenthasbeenreviewedinaccordancewithU.S.EnvironmentalProtectionAgencypolicyandapprovedforpublication.Mentionoftradenamesor
commercialproductsdoesnotconstituteendorsementorrecommendationforuse.
LiteratureCited
1.Armaghan,V.1937.BiologicalstudiesontheGiardiaofrats.Am.J.Hyg.26:236258.
2.Barnard,R.J.andG.J.Jackson.1984.Giardialamblia:thetransferofhumaninfectionsbyfoods.In:GiardiaandGiardiasis.S.L.ErlandsenandE.A.Meyer
(eds).PlenumPress,NewYork.pp.365378.
3.Belosevic,M.,Faubert,G.M.,MacLean,J.D.,Lau,C.andN.A.Croll.1983.GiardialambliainfectionsinMongoliangerbils:ananimalmodel.J.Inf.Dis.
147:222226.
4.Bingham,A.K.,Jarroll,Jr.,E.L.,Meyer,E.A.andS.Radulescu.1979.Giardia,sp:physicalfactorsofexcystationinvitro,andexcystationvs.eosinexclusionas
determinantsofviability.Exp.Parasitol.47:284291.
5.Bingham,A.K.andE.A.Meyer.1979.Giardiaexcystationcanbeinducedinvitroinacidicsolutions.Nature(Lond.)277:301302.
6.Black,R.E.,Dykes,A.C.,Sinclair,S.P.andJ.G.Wells.1977.Giardiasisindaycarecenters:evidenceofpersontopersontransmission.Pediatrics60:486491.
7.Coggins,J.R.andF.W.Schaefer,III.1984.Giardiamuris:scanningelectronmicroscopyofinvitroexcystation.Exp.Parasitol.57:6267.
8.Coggins,J.R.andF.W.Schaefer,III.1986.Giardiamuris:ultrastructuralanalysisofinvitroexcystation.Exp.Parasitol.61:219228.
9.Craun,G.F.1986.WaterbornegiardiasisintheUnitedStates,19651984.Lancet2:513514.
10.Feely,D.E.1982.HistochemicallocalizationofacidphosphataseinGiardia.Anat.Rec.202:54A.
11.Feely,D.E.1986.AsimplifiedmethodforinvitroexcystationofGiardiamuris.J.Parasitol.72:474475.
12.Feely,D.E.andJ.D.Dyer.1987.LocalizationofacidphosphataseactivityinGiardialambliaandGiardiamuristropozoites.J.Protozool.34:8083.
13.Hegner,R.1927.ExcystationandinfectionintheratwithGiardialambliafromman.Am.J.Hyg.7:433447.
14.Hegner,R.1927.TheviabilityofcystsofGiardialambliafrommaninthestomachoftherat.Am.J.Hyg.7:782785.
15.Hegner,R.1927.Excystationinvitroofhumanintestinalprotozoa.Science(Wash.D.C.)65:577578.
16.Hoff,J.C.,Rice,E.W.andF.W.Schaefer,III.1984.Disinfectionandthecontrolofwaterbornegiardiasis.In:EnvironmentalEngineering.M.Pirbazariand
J.S.Devinny(eds).AmericanSocietyofCivilEngineers,NewYork.pp.239244.
17.Hoff,J.C.,Rice,E.W.andF.W.Schaefer,III.1985.ComparisonofanimalinfectivityandexcystationasmeasuresofGiardiamuriscystinactivationby
chlorine.Appl.Environ.

Page254

Microbiol.50:11151117.
18.Hudson,S.J.,Sauch,J.S.andD.G.Lindmark.1987.FluorescentdyeexclusionasamethodfordeterminingGiardiacystviability.In:Proceedingsofthe
CalgaryGiardiaConference.P.WallisandB.R.Hammond(eds).UniversityofCalgary,Calgary.pp.265269.
19.Jarroll,E.L.,Bingham,A.K.andE.A.Meyer.1981.Effectofchlorineoncystviability.Appl.Environ.Microbiol.41:483487.
20.Jarroll,E.L.,Hoff,J.C.andE.A.Meyer.1984.Resistanceofcyststodisinfectionagents.In:GiardiaandGiardiasis.S.L.ErlandsenandE.A.Meyer(eds).
PlenumPress,NewYork.pp.311329.
21.Keister,D.B.1983.AxeniccultureofGiardialambliainTYIS33mediumsupplementedwithbile.Trans.Roy.Soc.Trop.Med.Hyg.77:487488.
22.Keystone,J.S.,Krajden,S.andM.R.Warren.1978.PersontopersontransmissionofGiardialambliaindaycarenurseries.Can.Med.Assoc.J.119:241
248.
23.Lindmark,D.G.1980.EnergymetabolismoftheanaerobicprotozoanGiardialamblia.Mol.Biochem.Parasitol.1:112.
24.Lippy,E.C.1981.Waterbornegiardiasis.In:TheProceedingsoftheInternationalConferrenceonLakeManagementandRestoration.Environmental
ProtectionAgency.EPA440/581010.
25.Meyer,E.A.1976.Giardialamblia:isolationandaxeniccultivation.Exp.Parasitol.39:101105.
26.Meyer,E.A.andF.W.Schaefer,III.1984.Modelsforexcystation.In:GiardiaandGiardiasis.S.L.ErlandsenandE.A.Meyer(eds).PlenumPress,New
York.pp.131144.
27.Meyers,J.D,Kuharic,H.A.andK.K.Holmes.1977.Giardialambliainfectioninhomosexualmen.Br.J.Vener.Dis.53:5455.
28.Osterholm,M.T.,Forfang,J.C.,Ristinen,T.L.,Dean,A.G.,Washburn,J.W.,Godes,J.R.,Rude,R.A.andJ.G.McCullough.1981.Anoutbreakoffoodborne
giardiasis.N.Engl.J.Med.304:2428.
29.Rendtorff,R.C.1954.Theexperimentaltransmissionofhumanintestinalprotozoanparasites.II.Giardialambliacystsgivenincapsules.Am.J.Hyg.59:209
220.
30.Rice,E.W.andF.W.Schaefer,III.1981.ImprovedinvitroexcystationprocedureforGiardialambliacysts.J.Clin.Microbiol.14:709710.
31.RobertsThomson,I.C.,Stevens,D.P.,Mahmoud,A.A.F.andK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gastroenterol.71:5761.
32.Rotman,B.andB.W.Papermaster.1966.Membranepropertiesoflivingmammaliancellsasstudiedbyenzymatichydrolysisoffluorogenicesters.Proc.Natl.
Acad.Sci.U.S.A.55:134141.
33.Sauch,J.F.1987.AnewmethodforexcystationofGiardia.In:ProceedingsoftheCalgaryGiardiaConference.P.WallisandB.R.Hammond(eds).
UniversityofCalgary,Calgary,Alberta.pp.271274.
34.Schaefer,III,F.W.,Rice,E.W.andJ.C.Hoff.1984.FactorspromotinginvitroexcystationofGiardiamuriscysts.Trans.Roy.Soc.Trop.Med.Hyg.
78:795800.
35.Schupp,D.G.,Januschka,M.M.andS.L.Erlandsen.1987.AssessingGiardiacystviabilitywithfluorogenicdyes:comparisonstoanimalinfectivityandcyst
morphologybylightandelectronmicroscopy.In:ProceedingsoftheCalgaryGiardiaConference.P.WallisandB.R.Hammond(eds).UniversityofCalgary,
Calgary,Alberta.pp.275279.
36.Wallis,P.M.andH.M.Wallis.1986.ExcystationandculturingofhumanandanimalGiardiaspp.byusinggerbilsandTYIS33medium.Appl.Environ.
Microbiol.41:483487.
37.Wickramanayake,G.B.,Ruben,A.J.andO.J.Sproul.1985.EffectofozoneandstoragetemperatureonGiardiacysts.J.Am.WaterWorksAssoc.77:7477.

Page255

FluorescentDyeExclusionasaMethodforDeterminingGiardiaCystViability
SusanJ.Hudson,JudithF.Sauch,andDonaldG.Lindmark*.
ClevelandStateUniversity,Cleveland,Ohio44115,U.S.A..
Exclusionofthefluorescentdye,3[dansylamidophenylboronicacid](FluoroBoraIFBI)fromGiardiamuriscystswascomparedwithexcystationasameasureofcystviability.
NonviableGiardiamuriscystsaccumulatedthedyeandfluorescedwhereasviablecystsexcludedthedyeandshowednofluorescence.TheexclusionofFBIfromviableGiardiacysts
iscontrarytothereactionwithviablefibroblasts,ChinesehamsterovarycellsandGiardialambliatrophozoites.Theeffectof22differentchemicalson%excystationwasdetermined
andcomparedwithFBIexclusionunderidenticalconditions.Correlationandregressionanalysesindicateahighdegreeofassociationbetweenthetwomethods.Ourdata
indicatethatthedyeexclusionmethodhaspotentialforuseasanalternativemethodtoexcystationasameasureofcystviability.Inthedyeexclusionmethod,FBIwasadded
directlytoacystpreparation.ObservationsweremadeimmediatelyusingaNikonEpiscopicFluorescentUnitwithawidebluefluorescentcube(excitingfilter410485nm,
dichroicmirror505nm,andabarrierfilter515nm).Routinely,200300cystswerescoredwithin4minaseitherfluorescingornotfluorescing.Thismethodwasperformedinonly
5mincomparedtothe23hrequiredforexcystation.Sinceonlyfluorescenceornonfluorescencewereobserved,themethodwaslesssubjectivethanexcystation.Wepresentdata
whichshowthattheFBImethodismorereproducibleandmoreprecisethanexcystationatlevelsofviabilityabove60%.TheFBImethodhaspotentialforuseasarapidscreenof
cystviabilityandfortherapidassessmentoftheeffectivenessofchemicalandphysicalagentsonviability.PreliminaryevidencesuggeststhattheFBImethodhasthepotentialfor
useinassessingtheviabilityofGiardialambliacysts.

Introduction
GiardialambliaisthemostcommonhumanintestinalprotozoanparasitereportedintheUnitedStatesandEngland(1,2,13).Theorganismexistsintwo
morphologicallydistinctforms,thetrophozoiteandthecyst.Theinfectivecystformistransmittedviathefecaloralroute.
Thehighincidence,symptomologyandwaterbornedisseminationofGiardiahasresultedinimprovedmethodsforcystdetectioninwatersystems(11).Howevera
rapid,simple,reliablemethodfordeterminingtheviabilityofcysts,detectedinwatersystems,andexposedtowatertreatmentproceduresisneeded.
TheviabilityofGiardiacystshasbeenassessedbyeosinexclusion,invitroexcystation,andanimalinfectivity(3,7).Whileeachofthesemethodshasitsownvalue,
nonearerapid,simple,andreliable.Althougheosinexclusionisarapidmethod,itindicateshighercystviabilitythancanbedemonstratedbyexcystationandshows
littlecorrelationwithexcystation(3).Excystationisthemostfrequentlyusedmethodfordeterminingcystviability.Itistedious(23h),andsubjective.Itcannotbe
usedfordeterminingtheviabilityofindividualorsmallnumbersofcysts(3).Thecost,timeandlargenumbersofcystsrequiredtoperformanimalinfectivitystudies
makethemethodimpracticalforroutineuse(7).
Theobjectiveofthisstudywastodeveloparapid,simpleandreliablemicroscopicmethodfordeterminingtheviabilityofGiardiacystswhich,whencomparedto
excystation,wouldprovideamorepracticalmethodforestimatingcystviability.Todothis,aprocesscalledboronicaciddependentphasetransfer,or''boradeption,"
asdescribedbyGallopetal.(5,6)wasused.Thisfluorescentdyeexclusionmethodwascomparedwithexcystation.Inthismethod,specificfluorescentwater
insolubleboronatesareexcludedbynonviablecellsandtakenupbyviablecellsinvitro(6).Theviablecellsthenexhibitedfluorescence.Ourresultsusing
FluoroBoraIwithGiardialambliatrophozoites,weresimilartothoseobtainedbyGallopetal.(5,6)withmammaliancells,i.e.viablecellsaccumulatedthedyeand
exhibitedfluorescence.Incontrasttotheresultsobtainedwithviabletrophozoites,viableGiardiacystsexcludedthedye(nonviablecyststookupthedyeand
fluoresced).TheFBIexclusionmethodwithGiardiacystsshowedahighdegreeofassociationwithexcystationandahigherdegreeofreproducibilityandprecision
whencomparedtoexcystation.
MaterialsandMethods
Organisms
GiardialambliacystswerecollectedfromfreshlyexcretedfecesoftheMongoliangerbil(Merionesunguiculatus)(4).Giardiamuriscystswereobtainedfrom
freshlyexcretedfecesoffemaleCF1mice(10).GiardialambliatrophozoitesweregrownasdescribedbyLindmark(8).
*Correspondingauthor.

Page256

CystPurification
GiardiacystswerepurifiedaccordingtoamodificationofthemethodofRobertsThomsonetal.(10).Fecalsamplesweredilutedwithdistilledwater,filtered
throughadoublelayerofcheesecloth,andcentrifugedat450gfor5min.Thepelletwaswashedwithdistilledwaterbycentrifugingfor2min.untilthesupernatant
solutionwasfreeofparticulatematter.One50mLconicalcentrifugetubewaspreparedforeachmLoffinalpelletvolume.Thepelletwasthenresuspendedin
distilledwateranddividedequallyamongthetubesanddilutedto25mLineachtubeEachsuspensionwasthenunderlaidwith25mLof1.0Msucroseand
centrifugedfor10minat800g.
CystsatthesucrosewaterinterfaceswereaspiratedwithaPasteurpipetintoclean50mLcentrifugetubesanddiluted1to5withdistilledwaterandmixedby
vortexing.Finally,thesecystswerewashedbycentrifugationat1500gfor5minandsuspendedin10to20mLdistilledwater.Cystcountsweremadeusingan
improvedNeubauerhemocytometer.Thecystswerestoredat4Cin20,000IUofpenicillinand20mgstreptomycinpermL.androutinelyusedwithinoneweekof
isolation.Cystviabilityandsensitivitytochlorineasmeasuredbyexcystation,werenotaffectedduringthistime(Jarroll,E.L.pers.comm.).
PreparationofFluorescentDye
Theboronicacidderivatives(FluoroBorasT M,FB)werepurchasedfromPolysciences,Inc.ofWarrington,Pennsylvania.DyeswerepreparedasdescribedbyGallop,
etal.(6).FBI,(3dansylamidophenylboronicacid)wasusedinourstudiesbecauseitproducedthemostintensefluorescence(unpublishedobservations,Lindmark).
FBIwaspreparedbydissolving2.0mgin200Lofdimethylsulfoxide(DMSO).Thiswasthenmixedwith600Lof25mMMOPSO[(3morpholino)2
hydroxypropanesulfonicacid]buffer,pH7.2.FiveLoftheFBIreagentweremixedwith25LofGiardiacystpreparation.Observationsofthepresence
(indicatingnonviablecysts)orabsence(indicatingviablecysts)offluorescenceweremadeimmediately(upto4min)withaNikonEpiscopicFluorescenceUnit
equippedwithawidebluefluorescencecube(excitingfilter410485,dichroicmirror505nm,andabarrierfilter515nm).Between200and300cystswerecounted
perassay.
Excystation
ExperimentalandcontrolGiardiacystswereexcystedasdescribedbyRiceandSchaefer(9).Cysts(106)werecentrifugedinaBeckmanmicrofuge12at1500g
for5min.Thesedimentedcystswerenextresuspendedin500LHClsaline,300Lfreshlyprepared100mMNaHCO3,and300Lofareducingsolution
consistingof67mMglutathioneand29mMLcysteineHClin1XHanks'balancedsaltsolution.Cystswerethenincubatedfor30min.ina37Cwaterbath.
Afterthisincubationstep,cystswerewashed3Xin5%trypsininTyrode'ssalts(9).Thefinalpelletwasresuspendedin3050LtrypsinTyrode'ssalts,mixed,and
placedonadepressionslidesealedwithvaseline.Slideswerethenincubated3060minat37Cinawarmairincubator.Thenumberofintactcysts(IC),partially
excystedtrophozoites(PET)andtotallyexcystedtrophozoites(TET)at400Xmagnificationweredetermined.Percentexcystationwascalculatedasfollows(3):
(#TET/2+#PET)+(#TET/2+#PET+#IC)100.
MethodComparison
CystswereexposedtochemicaldisinfectantsothatviabilitycouldbeassessedbyexcystationandFBIexclusion.Disinfectantswereusedwhichshouldexhibit
differentmodesofactioninordertoqualitativelyobservechemicallydependentdiscrepanciesthatmightoccurbetweenthetwomethods.Examplesofthedisinfectant
groupsusedare:phenolderivatives(orthobenzylparachlorophenol,orthophenylphenol)quaternaryammoniumcompounds(alkyldimethylbenzylammonium
chloride,octyldecyldimethylammoniumchloride,dioctyldimethylammoniumchloride,didecyldimethylammoniumchloride,tetradecyltrimethylammoniumbromide,
benzyldimethylhexadecylammoniumchloride,hexadexylpryidiniumbromide,benzyldimethyldodecylammoniumbromide)chlorinecontainingcompounds(chlor
hexidinediacetate,sodiumhypochlorite,Alcide)glutaraldehydesandmixturesoftheabove.Thechemicalswereusedatdifferentconcentrationsandcontacttimes
(proprietaryrecommendationsofmanufacturers,manuscriptinpreparationLindmark,MillerandZimmer).
Approximately1106cystswereplacedina1.8mLmicrofugetubeandcentrifugedfor5minat1500ginaBeckmanmicrofuge(12).Thesupernatantsolution
wasaspiratedanddiscarded.Controlcystsweresuspendedin1mLofwatercyststobeexposedtodisinfectantsweresuspendedin1mLofthechemical
disinfectantatthespecifiedconcentrations(proprietaryrecommendationsofmanufacturers).Cysts(controlsandchemicallytreated)wereincubatedat7,21and24
Cfor10and20min(determinedaccordingtomanufacturer'srecommendations).Aftertreatment,controlandchemicallytreatedcystswerewashed3Xindistilled
waterfor5minat1500g,resuspendedin100Ldistilledwaterandmixedwell.FiftyLwereremovedandusedforexcystation25Loftheremaining50L
wereexposedtoFBI.Enumerationofbothmethodswasdonesimultaneously.
Thecomparisonsofthetwomethodsbetweeninvestigatorswasdoneutilizingtheproceduregivenabovewiththefollowingalterations.Theinitialamountofcysts(0
16daysold)usedwas1107.Theseweresuspendedin1mLwateraftertreatment.TenaliquotswerethenremovedandusedforreplicatecountingusingtheFBI
method.Excystationforthreereplicateswasdeterminedsimultaneously.
StatisticalAnalysis
Statistically,thetwomethodswerecomparedusingaleastsquareslinearregressionanalysisandacorrelationanalysis.Thelinearrelationshipbetweenthetwo
methodswasestimatedbyregressingthepercentageofviablecystsfromtheexcystationmethodonthepercentageofviablecystsfromtheFBImethod.The
estimatedslopewastestedforaonetoonecorrespondencebetweenthetwomethods(H0:B=1)andforagenerallineartrendbetweenthetwomethods(H0:B=0).
TheexistenceanddegreeofassociationbetweenthetwomethodswasexaminedbycalculatingPearson'scorrelationcoefficientandSpearman'sRankcorrelation
coefficient(12).Therankcorrelationanalysishastheadvantagethatitdoesnotrequirethedistributionalassumptionsnecessaryforthevalidityoflinearregressionand
thesignificancetestofPearson'scorrelationcoefficient.Thestatisticsindicateageneralrelationshipbutnottheabilitytopredicttheresultsofonemethodaccurately
fromtheother.
Results
PreliminaryStudieswithFluoroBoraIF
Ninetyfivepercent(4%,5determinations)ofGiardialambliatrophozoites(72hrold)fluoresced(indicatingviability)whenmixedwithFBIasdescribedin
materialsandmethods.Whenthesamecellpreparationswereexposedtoheatat50Cfor2min[time(0to5mintested)andtemperature(50to60Ctested)
neededtodestroymotilityasdeterminedmicroscopically]theviabilityasdeterminedbyfluorescencedecreasedto6%(5%,5determinations).Theseresults
indicatethattheprocedureforstainingviablecellsisapplicabletoGiardialambliatrophozoitesascouldbepredictedaccordingtotheresultspresentedbyGallopet.
al.(6).However,viable,freshlyharvestedcystsofGiardialambliaandGiardiamuris,whenheattreated(50Cfor5min,determinedempiricallyasabove)and
showntobekilledasmeasuredbyexcystation(Jarroll,pers.communications),reactedwithFBIinanoppositemanner.Viablecontrolcysts(unheated)showed5%
(3%,3determinations)and8%(4%,3determinations)fluorescenceforGiardialambliaandGiardiamurisrespectively.Heattreatedcystsexhibited91%(
6%,3determinations)and87%(10%,3determinations)

Page257

Figure1.
RegressionanalysisofG.murisPEX%excystation).onPNF
(%nonfluorescing).[Thesolidlineistheestimatedlineand
thedashedlinerepresentsalinewithaslope=1(i.e.,11correlation)].
DatapointswereobtainedusingthechemicalsgiveninMaterials
andMethods.Theconcentrations,timeofcontact,temperatureand
chemicalforeachdatapointarenotprovidedbecauseofproprietary
restrictionsmandatedbythemanufacturers.

fluorescenceforGiardialambliaandGiardiamurisrespectively.DespitethedifferenceinstainingbetweenGiardiacystsandtrophozoitesculturedinvitro(6),our
resultsindicatethatFBIexclusionfromcystscouldpotentiallybeusedasamethodfordeterminingcystviability.
ComparisonsBetweenExcystationandtheFBIMethod
Figure1displaysthedatafromthecontrolandchemicallytreatedGiardiamurissamples.Somechemicaldisinfectantsusedresultedinclumpingand/ordestruction
ofcysts.Datapointsproducedwhencystclumpingwasnoted(causingproblemswithexcystationcounts),andwhencystdestructioncausedlowcystnumberswere
omitted(N=8).Figure2excludesthepointsinFigure1with0%excystationand0%nonfluorescence(N=19).Aseparateanalysiswasperformedomittingthese
pointsbecauseofthepossiblebiastheymighthaveontheresults.Theplotsshowtherawdata,thefittedregressionline,andalinewithaslopeofone.Bothfigures
showagoodrelationshipbetweenpercentageexcystation(PEX)andpercentagenonfluorescing(PNF).
TheresultsofthestatisticalanalysesarepresentedinTable1.Whilethedatadidnotfullymeettheassumptionsrequiredfortheparametricanalysis,theresultsfromall
analyseswereconsistentinthatahighlysignificantassociationwasfoundbetweenthetwomethodsinestimatingthepercentofviablecysts.
PrecisionandReproducibilityoftheFBIandExcystationMethods
TheviabilityofthesamepreparationsofGiardiamuriscysts(0,1,2,3,and16daysold,withreplicatesamples)wasassessedwithFBIandexcystationbytwo
investigators.Figure3graphicallyrepresentstheprecisionofthemethods.Intherangetested,thedataindicateahighdegreeofprecision(StandardDeviationSD=
0.017)andreproducibility(SD=0.006)fortheFBImethodcontrarytotheprecision(SD=0.061)and

Figure2.
RegressionanalysisofG.murisPEX(%excystation)onPNF
(%nonfluorescing)excluding19(0,0)points.[Thesolidlineisthe
estimatedlineandthedashedlinerepresentsalinewithaslope=1
(i.e.,11correlation)].Thesameproprietaryrestrictions
givenforFigure1applytoFigure2.

reproducibility(SD=0.030)oftheexcystationmethod.
PreliminaryStudiesWithGiardiaLambliaCysts
FreshlyharvestedGiardialambliacystsexhibitedlowerexcystationrates(2050%)thancystsofGiardiamuris(9099%)andtheexcystationratesshowedno
correlationwithFBIexclusion(Table2).Themeanexcystationofthesecysts(35%)was42%oftheviabilityasdeterminedbytheFBImethod(82%).However,it
wasnoticedthatafterinduction,whenthecystswereplacedinexcystationmedium,thetrophozoitesinsidethecystexhibitedmotilitywithin2min.Ifmotilityinsidethe
cystisusedasameasureofviability,insteadofexcystation,amuchbettercorrelationwiththeFBImethodisobserved(Table2).Thesepreliminarydata(Table2)
withGiardialamblia,correlatingmotilityinsidethecystduringtheexcystationprocedure,asameasureofviability,withFBIexclusion,indicatethattherelationship
betweenthesetwomethodsismoresimilartothatfoundusingexcystationandFBIexclusionwithGiardiamuriscysts.
TABLE1.ResultsofdataanalysisforG.murissamples.

ExcludingObservations
PNF=0andPEX=0

AllData
RegressionAnalysis:

N
2

19

0.97

0.86

(slope)

1.02

1.00

95%C.I.*forB

(0.96,1.08)

(0.80,1.21)

CorrelationAnalysis:

Pearson'sR(p)

0.98(<0.001)

0.93(<0.001)

Spearman'sR(p)

0.97(<0.001)

0.89(<0.001)

*ConfidenceInterval
PNF=%nonfluorescing,PEX=%excystation

44

Page258

Figure3.
PrecisionofFBIandexcystationmethodsfordeterminingG.muris
cystviability:InvestigatorA[PNF(%nonfluorescing)=0,
PEX(%excystation)=X)andB(PNF=,PEX=).Replicateanalysis
wasdoneon8replicatesamples(from116daysold).

Discussion
Fluorescentdyeexclusion(FBImethod)showsgoodcorrelation(atcystviabilities>60%)withexcystationasamethodfordeterminingGiardiamuriscystviability.
FBIisexcludedfromviablecysts,incontrasttotheresultsobtainedwithviablefibroblasts,Chinesehamsterovarycells(6)andGiardialambliatrophozoitesin
whichviablecellstakeupFBIandfluoresce.Thismaybeexplainedinpartbythechemicalnature(unknownatthepresenttime)ofthecystwallandthepermeability
barriercausedbythecystwall.Thecystwallmaylackthelipoidalcomponents,foundinplasmamembranes,whichareneededforthe"boradeption"reaction
(permeationofFBIintoviablecells).Howeverifthecystwallpermeabilityisalteredbychemicalandheattreatment,theFBIcanthenenterandformcomplexeswith
hydroxylandaminogroupswithinthecellresultinginfluorescence.
TheFBImethodwassimpletoperform,rapid(5min),andcanbemicroscopicallyread.Themethodshowsahigherdegreeofprecisionthanexcystationatviability
levelsabove60%forcontrolcysts.TheFBImethodisalsolesssubjectivethanexcystation.Themajorreasonsforthislieinthefactsthat1)theFBImethodrelieson
anobservationoffluorescenceornonfluorescenceasameasureofviabilityexcystationrequirestheperformanceofaseriesofcountsofintactcysts,totallyexcysted
trophozoitesandpartiallyexcystedtrophozoitesanditisdifficultforallbutthewelltrainedinvestigatortodifferentiateamongthedifferentstagesrequiredinthe
countingprocedureand2)theexcystationprocedureislongandtedious(23h)requiringcontrolofanumberofchemicalandphysicalvariables(temperature,pH,
reducingconditions,etc.)notencounteredintheFBImethod.
WhentheFBImethodwasusedtodeterminetheeffectofvariouschemicalsoncystviabilityclumpingand/orcystdestructionoccurredwithseveralchemical
disinfectants(phenolandchlorinecontainingcompounds).Samplesthatshowclumpingcannotbeevaluatedforviabilitybyexcystationbecauseofthedifficultyin
identifyingthevariousstagesneededforcounting.Clumping,ontheotherhand,haslittleeffectoncountsbytheFBImethod.Cystdestructionisaprobleminboth
methodsinthatitisdifficulttoobtainstatisticallyaccuratecounts.Chemicaldisinfectantconcentrationsshouldbechosenthatcausecystdeathwithoutdestroyingthe
cystssothatenoughcystsremaintomakethecountssignificant.
ThedatapresentedinFigure3indicatesthatFBImeasurementsofGiardiamurisviabilityareconsistantlyhigherthanbyexcystation,suggestingthepossibilitythat
theFBImethodismoresensitivefordeterminingcystviabilitythanexcystation.SimilarresultswereshownbyBinghametal.(3)witheosindyeexclusionusing
Giardialambliacysts.
PreliminaryevidencesuggeststhattheFBIcanbeusedtodeterminetheviabilityofGiardialambliacysts.ThedatapresentedinTable2showsagoodcorrelation
betweenmotilityinsidethecystandFBIexclusionbutnotbetweenexcystationandFBIexclusion.Thesuggestioncanalsobemade(basedonourpreliminary
information)thattheexcystationprocedureweroutinelyusedforGiardia(9)maynotbeappropriateforGiardialambliainthatitgivesalowestimateofGiardia
lambliacystviabilitycomparedtotheFBImethod.ThissuggestionisinagreementwiththeresultsfoundbyBinghametal.(3)witheosinexclusioninwhich
excystationgavealowerestimateofviabilitythaneosinexclusion.Thesuggestionshouldalsobemadethateosinexclusionmightbefurtherinvestigatedasamethodof
assessingcystviability.Giardialambliatrophozoitesinsidethecystsexhibitmotility(suggestingviability)duringtheexcystationprocedurebutahighpercentfailto
exitthecystwall.Itcanbehypothesizedthattheenvironmentintheexcystationmediumisnotcondusivetothefinalstepsofexcystation
TABLE2.ViabilityofGiardialambliacystsmeasuredbyexcystation,motilityinsidethecystandFBIexclusion.

CystPreparation

Age*(days)

Excystation

Motility
inCyst

FBI
Exclusion

35.2

72.4

77.9

40.1

71.6

80.0

20.0

75.5

75.6

39.6

85.4

89.6

41.5

91.0

90.2

25.6

72.3

71.7

27.4

71.2

78.1

51.6

90.3

95.6

29.8

75.2

79.2

10

42.8

84.1

85.3

*Relativetotimeofcollection.

PerCentofCystsShowing

Page259

becauseoflackingnutrients,cofactors,etc.orthepresenceofinhibitors.Furtherstudiesshouldinvestigatethisphenomenon.
Inconclusion,wesuggestthattheFBImethodhasthepotentialforuseintherapidassessmentofGiardiamuriscystviability(atlevelsofviabilityabove60%).Itis
lesssubjectiveandmoreprecisethanexcystation.Inordertostatethat"theFBImethodcanbeusedasasubstituteforexcystation"furtherworkneedstobedone.
ThisresearchwouldinvolveanindepthcomparisonofthetwomethodswithGiardialambliacystsandacomparisonofthetwomethodswithcystpreparationsof
lessthan50%viability.Presentlyonlyminimalobservations(unpublished)havebeenmadebetween1%and50%viabilityandthedatahavebeenhighlyvariable.
Acknowledgements
TheauthorswouldliketothankDr.JudyStoberandTammyMills(USEPA,Cincinnati)fortheirstatisticalanalysisofthedata.Thisdocumenthasbeenreviewedin
accordancewithU.S.EnvironmentalProtectionAgencypolicythroughcooperativeagreement#CR811949010toClevelandStateUniversityandapprovedfor
publication.Mentionoftradenamesorcommercialproductsdoesnotconstituteendorsementorrecommendationforuse.
LiteratureCited
1.IntestinalparasitesrangingfarafieldintheUnitedStates.1978.MedicalNews.J.Amer.Med.Assoc.239:2756.
2.CommunicableDiseaseWeeklyReports.1977.PublicHealthLaboratoryService,London.
3.Bingham,A.K.,Jarroll,E.L.,Meyer,E.A.,andS.Radulescu.1979.Giardiasp.:physicalfactorsofexcystationinvitro,andexcystationvs.eosinexclusionas
determinantsofviability.Exp.Parasitol.47:284291.
4.Belosevic,M.,Faubert,G.M.,MacLean,J.D.,Law,C.,andN.A.Croll.1983.GiardialambliainfectionsinMongoliangerbils:ananimalmodel.J.Infect.
Diseases.147:222226.
5.Gallop,P.M.,Paz,M.A.,andE.Henson.1982.Boradeption:anewprocedurefortransferringwaterinsolubleagentacrosscellmembranes.Science217:166
169.
6.Gallop,P.M.,Paz,M.A.,Henson,E.,andS.Latt.1984.Dynamicapproachestothedeliveryofreporteragentsintolivingcells.Biotechniques2:3236.
7.Kasprzak,W.andA.C.Majewska.1983.InfectivityofGiardiasp.inrelationtoeosinexclusionandexcystationinvitro.Tropenmed.Parasit.34:7072.
8.Lindmark,D.G.1980.EnergymetabolismoftheanaerobicprotozoanGiardialambliatrophozoites.Mol.Biochem.Parasitol.1:112.
9.Rice,E.W.andF.W.Schaefer,III.1981.ImprovedinvitroexcystationprocedureforGiardialambliacysts.J.Clin.Microbiol.14:709710.
10.RobertsThomson,I.C.,Stevens,D.P.,MahmoudA.A.F.,andK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gastroenterology71:5761.
11.Sauch,J.F.1985.UseofimmunofluorescenceandphasecontrastmicroscopyfordetectionandidentificationofGiardiacystsinwatersamples.Appl.Environ.
Microbiol.50:14341438.
12.Wardlaw,A.C.1985.PracticalStatisticsforExperimentalBiologists.JohnWileyandSons,NewYork.
13.Wolfe,M.S.1975.Giardiasis.J.Amer.Med.Assoc.233:13621365.

Page261

ANewMethodforExcystationofGiardia
JudithF.Sauch
U.S.EnvironmentalProtectionAgency,HealthEffectsResearchLaboratory,26WestMartinLutherKing,Cincinnati,Ohio45268,U.S.A..
InvitroexcystationofGiardiaisusedtoevaluatecystviabilityandmayalsobeusedtoobtaintrophozoitesandcystwallsforanalysis.Recentlypublishedexcystationprocedures
includetheuseoftrypsin,serumorbilesaltsinthe"excystation"step.BecauseGiardiamurisandGiardialambliacystwallsandtrophozoitesweretobeusedforantigenicanalysis,
analternativemethodwasdevisedtoavoidexposuretotrypsin,serumandbilesalts.AninitialacidorlowpH"induction"stepwasretained,butcystexposuretothemediumwas
increasedto45min.Trypsin,serumorbilesaltsinthe"excystation"mediumwasreplacedbyproteosepeptoneforG.murisandphytonepeptoneforG.lamblia.Thesereagentswere
addedtoHanks'BalancedSaltSolution(HBSS)supplementedwithcysteineandsodiumbicarbonate.IncreasedlevelsofcysteineandglucoseinHBSSdecreasedpercent
excystationforG.lambliacysts(fromhumandonors).Comparisonbetweenthismethodutilizingpeptonesandamethodutilizingtrypsinshowednosignificantdifferencesinpercent
excystationforbothG.murisandG.lamblia.VariationinpercentexcystationamongsevenG.lambliacystsamplesfromhumandonorsoccurredbutthetwomethodsfollowedsimilar
trendsastheyvaried,withnosignificantdifferencesbetweenthetwo.Thus,neithertrypsin,serumorbilesaltsisrequiredforexcystationofG.murisorG.lamblia.

Introduction
CurrentlyavailableexcystationproceduresexposeGiardiacystwallsandtrophozoitestocomplexmediacontainingenzymes,serum,orbile,whichmightinterfere
withsubsequentelectrophoreticanalyses.Therefore,alternativereagentsweredevisedinordertoavoidexposuretotheseagentsandstillobtainhighlevelsof
excystation.
AlthoughHegnerconcludedin1927(6),afterobservingexcystationinvivo,thatmoisture,temperature,anddigestivejuicesweremajorfactorsintheprocess,itwas
notuntilthe1970'sthatexcystationrequirementsforGiardiaweredefinedmoreprecisely.BinghamandMeyer(2)werethefirsttodemonstratethatalthough
hydrogenionsinitiatetheexcystationprocess,transfertoasecondfavorablemediumiscriticaltocompleteexcystation.The"excystation"mediumtheyused(HSP
39)containshumanserum,phytone,vitamins,aminoacids,andsalts.Subsequenttothesepublications,anumberofexcystationproceduresweredevelopedas
modificationsoftheoriginal.Inallofthem,theacidincubationstepisretainedforinduction,butsupplementationoftheexcystationmediumvaries.Serum(1,2,4,7),
bile(4,7),andenzymessuchastrypsin(10,14),pepsin(7,14),andpancreatin(8)havebeenusedwithavarietyofexcystationlevelsreported.Schaeferetal.(14),
however,notedthattrypsin,whileappearingnottobeessentialforhighlevelsofexcystationofG.muriscysts,seemstoaccentuatetheprocess.
Analternativeexcystationmediumconsistingofbalancedsaltsolutionsupplementedwithaminimumofcomplexreagentswasdeveloped.Sincephytoneisincludedin
Bingham'smediumandappearedtoaffectexcystationfavorably(E.W.Rice,personalcommunication),itwasusedfirsttoreplaceserumenzymesandbile.
Experimentsinwhichphytonepeptone,proteosepeptone,glutathione,cysteine,andglucosewerevariedoreliminatedledtotheformulationofasimplifiedexcystation
mediumconsistingofabalancedsaltssolutionsupplementedwithproteoseorphytonepeptone,cysteine,andglucose.HighexcystationlevelsforG.muris,
comparabletopublishedresults,andvariable,butsometimeshigh,excystationlevelsforG.lambliawereobtained.Thissuggeststhatcomplexagentssuchas
enzymes,wholeserum,andbilearenotessentialforGiardiaexcystation.
MaterialsandMethods
Organisms
G.muriscystswereobtainedfromfemaleSwissalbinomice(CF1)infectedwithcystsperos.Cystswereharvestedbymodificationsoftheproceduredeveloped
byRobertsThomsonetal.(11).Mousefecesfromapproximately60animalswerepooledafter1621hourscollection,mixedintoaslurry,anddilutedtoabout
2500mLwith0.01%Tween20indistilledwater(TDW).Aftermixingthoroughly,theslurrysatatroomtemperatureforonehalfhourtosedimentlargedebris.After
thissedimentationwasrepeatedtwice,theresultingsupernatantfluidswerecentrifugedat800gfor2minutes.Thepelletswererepeatedlymixedbyvortexinginto
0.01%TDWandcentrifugeduntilthesupernatantfluidswerefreedoffineparticles.Finally,thepelletswerepooled,suspendedinabout600mLTDW,mixedwell,
dispersedintosixroundbottomplastic250mLcentrifugebottles,underlayedwith80mL1.0Msucrose,andcentrifugedinaswingingbucketrotorfor10minutesat
800g.Thebandsofcystsfloatingbetweenthesucroseandwaterlayerswerecollectedbyaspirationintoaflask,diluted1:5withTDW,andwashedthreetimesin
TDWbycentrifugationfor5minutesat800g.Cystswerefurtherpurifiedbysedimentationatunitgravity(12)andstoredindistilledwaterat25C.G.lamblia
cystswereharvestedinasimilarmanner,withafewmodifications,fromhumandonorstoolspecimens.Largefloatingparticleswerefilteredoutwith23layersof
cheesecloth.Afterwashing,pelletsweresuspendedinTDW25mLofthepelletsuspension(representing1mLofpackedpelletvolume)wasoverlayedonto25mL
Percollsucrose,sp.gravity1.09(13).CystswerefoundinabandbetweenthewaterandPercollsucroselayers.Thesebandswerecollected,pooled,diluted1:5
withTDWandwashedthreetimes

Page262
7

bycentrifugationfor5minutesat800g.Cystswerepurifiedbysedimentationatunitgravityifmorethan10 werecollected.Theywerethenstoredindistilledwater
at25Cuntiluse.
ExcystationofG.lamblia
Theacidincubationstep(10)wasretained.Briefly,0.5mLofupto107cystsindistilledwaterwastransferredtoa15mLconicalcentrifugetubetowhichwasadded
5mLHClsaline(0.7mLconc.HCl,100mL0.85%NaCl,pH1.5),3mLreducingsolution(HBSSsupplementedwith32mMglutathioneand57mMLcysteine
HCl)and3mL0.1MNaHCO3.Thesuspensionwasmixedbyvortexing,incubatedina37Cwaterbathfor45minutes,andcentrifugedfor2minutesat650g.
Thecystsinthepelletswerewashedonceinprewarmed0.85%NaClbycentrifugationfor2minutesat650g.Thepelletwasfinallysuspendedin15mL
prewarmedexcystationmediumandincubatedfor30minutesina37Cwaterbath.Theexcystationmedium,preparedfresh,waspreparedbydissolving0.015g
NaHCO3and0.01gLcysteineHClmonohydratein4mLof5%phytonepeptoneand1mL10XHBSSthefinalvolumewasbroughtto10mLwithdistilledwater
(finalpH7.1andfinalphytonepeptone2%).Stockphytonepeptone(5%W/V)waspreparedbyaddingphytonepeptone(BBL#11906)todistilledwaterand
boilinggentlywithstirringfor10minutes.Afterfiltersterilization,thestocksolutionwasstoredat25Cinsmallvolumes.Sampleswereremovedfromtheexcystation
tubeandcountedinahemocytometer.Atleast200totallyexcystedtrophozoites(TET),partiallyexcystedtrophozoites(PET),andintactcysts(IC)werecounted.
ExcystationwasquantitatedaccordingtotheformulareportedbyBinghametal.(3)asfollows:[TET+2+PET]+[TET+2+PET+IC]100.IncontrasttoG.
muris,emptycystwallsofG.lambliaaredifficulttoobservehowever,G.lambliatrophozoitessettlequicklyandcanbecounted(10).ThenumberofTET'sis
halvedbecauseeveryexcystedcystyieldsapairoftrophozoites(3).ThismethodwascomparedtothatofRiceandSchaefer(10)usingG.lambliacystsisolated
fromseveralasymptomaticandsymptomatichumandonors.
ExcystationofG.muris
TheinductionstepwascarriedoutasreportedbySchaeferetal.(14)withminormodifications.Between106and3107cystsweresuspendedin1mLdistilled
waterinaplasticconicalbottomcentrifugetubetowhichwasadded20mLofreducingsolution[Hanks'BalancedSaltSolution(HBSS)supplementedwith32mM
glutathioneand57mMLcysteineHCl]and20mLof0.1MNaHCO3(finalpH4.7).Thesuspensionwasmixedbyvortexing,incubatedina37Cwaterbathfor30
minutes,centrifugedat650gfor2minutes,andthenwashedonceintheexcystationmediumbycentrifugationat650gfor2minutes.Cystswerefinallysuspened
in15mL(dependinguponthenumberofcysts)prewarmed0.5%proteosepeptoneinPBS,pH7.2,andincubatedforonehalfhourina37Cwaterbath.Astock
proteosepeptone(5%W/V)solutionwaspreparedindistilledwater,gentlyboiledfor10minutestodestroyanyremainingenzymes,filtersterilized,andstoredat2
5Cinsmallvolumes.Excystationmedium,preparedfresh,wasmadebyadding10mLstockproteosepeptoneand10mLof10XPhosphateBufferedSaline(PBS,
80gNaCl,2gKH2PO4,29gNa2HPO412H2O,2gKCl,1000mLfinalvolume)to80mLdistilledwater.ThismethodwascomparedtothatofSchaeferetal.
(14).Bothexcystationmethodswereperformedincentrifugetubessampleswereremovedandcountedinahemocytometer,usingphasecontrastopticsat200400
magnification.Between100and200emptycystwalls(ECW),partiallyexcystedtrophozoites(PET),andintactcysts(IC)werecounted.Thepercentexcystation
wascalculatedaccordingtotheformula[ECW+PET]+[IC+ECW+PET]100(14).
ResultsandDiscussion
DevelopmentofExcystationMethodforG.lamblia
Asinglesamplefromanasymptomatichumandonorwasusedtodevelopserumandenzymefreereagents.Theacidinductionstep(10)wasretained,butimproved
excystationoccurredwhenthisstepwasincreasedto45minutes(unpublishedobservations).Aftercystswereexposedtovariousexcystationmediaconsistingof
0.5%phytoneorproteosepeptoneineitherHBSSorTyrode'smedium,excystationwashigher(11and8%,respectively)thanthatobserved(<6%)whencystswere
exposedtotrypsinTyrode'ssolution,phytoneTRIS,orproteoseTRISbuffer,pH7.0.
Tyrode'smediumorHBSSwasnextsupplementedwithvariousconcentrationsofphytone(0.252.0%).HBSSplus2.0%phytonepeptoneyielded57%excystation,
whereastheotherconcentrationsofphytoneinHBSSyielded5055%.Atphytoneconcentrationslowerthan2%,however,theTETwerenonrefractileandjudged
tobedead.Therefore,theywerenotamenabletosuccessfulseparationfromcystwallsbyisopycniccentrifugation.Inthesameexperiment,Tyrode'smedium
containingtrypsinorphytoneresultedinlessthan21%excystation.
FurthersupplementationofHBSSphytone(2.0%)withglucose(0.05%abovethatroutinelyusedforHBSS)and1.0%cysteineincreasedexcystationto73%.
However,eliminationoftheextraglucose,butnotcysteineorphytone,yielded90%excystation.ControlsconsistingofcystsexposedtotrypsinTyrode'smedium
yieldedonly27%excystation.Inthesameexperiment,lowerlevelsofexcystationresultedwhencystswereexposedtoHBSS(14%),HBSSglucose(15%),HBSS
plus2.0%phytone(17%),andHBSSplusglucoseandphytone(22%).
WhenHBSSwassupplementedwithonly1.0%cysteine,butnotphytone,excystationwasashighas83%.Thissuggestedthatcysteineplaysanimportantroleinthe
excystationprocess.Adecreaseintheconcentrationofcysteineto0.1%resultedin96%excystation,whereas,inthesameexperiment,eliminationofcysteineby
incubationofcystsintrypsinTyrode'smediumalonereducedexcystationto57%.Inalloftheseexperiments,HBSSwassupplementedwith0.15%NaHCO3.Thisis
consistentwiththeresultsofGillinandReiner(5)whichdemonstratethatG.lambliatrophozoitesdependuponthiolreducingagents,suchascysteine,forsurvival
andattachmenttoglasssurfaces.
Theseresultsdemonstratedthatacomplexmediumcontainingenzymesisunnecessaryandcanbereplacedwithasimplemediumconsistingofabalancedsalts
solution,cysteineandphytonepeptone.Theconcentrationofphytonefoundtobemosteffectivewasslightlyhigherthanthatreportedpreviouslyforthefirstsuccessful
invitroexcystations(2).TheseresultssuggestaminorroleofsomephytoneconstituentandamajorroleofLcysteine.
DevelopmentofExcystationMethodforG.muris
G.muriscystswereabletoexcystwhentheabovemethoddevelopedforG.lambliawasused.Unfortunately,theTETdiedquicklyandcouldnotbeadequately
separatedfromtheemptycystwallsbyisopycniccentrifugation.Thissituationimprovedwhenproteosepeptonewasusedintheexcystationstep.Optimalappearance
oftheTEToccurredwith0.5%proteosepeptone,eventhoughexcystationwas>98%withallconcentrationsofproteosepeptoneused(0.22.0%).Cysteineand

Page263
TABLE1.ComparisonofmethodsforexcystingGiardialamblia.

Mean%ExcystationofMethod(range)

Samplea

Cystage
(days)b

TrypsinTyrodec

CysteinePhytoned

15(1217)

15(1417)

82(8084)

93(9096)

12

84(8385)

93(8996)

21

38(3343)

49(4354)

13

<1(01)

2(2)

14

21(1824)

21(1725)

15

30(2732)

19(1820)

Samples14werefromthesameasymptomaticdonor,collectedatdifferenttimes.Samples57werefromdifferent
asymptomaticdonors.
b

Timefromcystpurificationtodayofexcystation.

MethodofRiceandSchaefer(10).Resultsweretheaverageoftwotrials.

MethoddescribedintextforG.lamblia.Resultsweretheaverageoftwotrials.Allsampleswereexcystedonthesame
day.

NaHCO3wereeliminatedfromtheexcystationmediumafteritwasdeterminedthatexcystationremainedat>99%intheabsenceofthesereagents.Incontrast,G.
lambliacystswereincapableofexcystationwhenproteosepeptonewassubstitutedforphytonepeptone,andshowedreducedlevelsofexcystationwhencysteine
wasnotpresent.
G.murisexcysted(>99%)whenasimpleinductionmediumconsistingofHBSS,cysteine,glutathione,andNaHCO3wasused.EventhoughG.muriscystswere
capableof>99%inthepresenceorabsenceofHClintheinductionstep,thatreagentwasomittedtoavoidunnecessaryexposureofcyststoit.Recently,Feely(4)
reportedthatreducingagentswerenotnecessaryforinductionofG.muris.Althoughreducingagentshadbeenretainedinourinductionstep,anexperimentwas
performedinwhichthestandardinductionmedium(HBSS,cysteine,glutathione,andNaHCO3)wascomparedtoacidifiedHBSSminuscysteineandglutathione(4)
andHBSStowhichonlyNaHCO3wasadded.Whiletheuseofthestandardinductionsolutionyielded>99%excystation,useofthelattertwomediaresultedin
lowerexcystation(20%and5%,respectively).TheTYIexcystationmediumusedbyFeely(4)issupplementedwithserumandbile.Thedifferencein
supplementationoftheexcystationmediacouldaccountforthedecreaseinexcystationwhenreducingagentswereomittedfrommyinductionstep.
MethodComparisonandUse
OncetheexcystationmethodsforG.lambliaandG.muriswerestandardized,theywerecomparedtothemethodsofRiceandSchaefer(10,14).Foreitherspecies
ofcyst,bothexcystationprocedureswereperformedonthesamedaywithallsamplesused.Ascanbeseenfromtheresults(Tables1and2),noapparent
differencesexistedbetweenthetwoexcystationmethodsforbothspeciesofGiardia.Bothsampletosamplevariationandextremesofpercentexcystationoccurred,
butthetwomethodsfollowedsimilartrendsastheyvaried.Ithasbeenpreviouslyreported(10)thatasymptomaticdonorsofGiardiacystsyieldlower
TABLE2.ComparisonofmethodsforexcystingGiardiamuris.

Mean%excystationofmethod(range)
Cystage
(days)a

TrypsinTyrodeb

Proteosepeptonecc

84(8089)

91(8994)

91(9097)

89(8890)

91(9092)

89(8890)

Timefromcystpurificationtoexcystation.Thesamebatchofcystswasusedforeachday.

MethodofSchaeferetal.(14).Resultsweretheaverageofthreetrials.

MethoddescribedintextforG.muris.Resultsweretheaverageofthreetrials.

excystationpercentagesthansymptomaticcarriers.However,datareportedhere(Table3)showthatthemeanexcystationofcystsfromasymptomaticcarriers(56%
11%standarderror)wassimilartothatofcystsfromsymptomaticdonors(50%13%SE).Thisdiscrepancybetweenthepreviouslyreporteddataandthat
presentedhereisnotsurprisingwhendatafromthesamedonorareexamined.CystsfromdonorsAandB(Table3)showedextremesofpercentexcystationwhen
differentsamplesfromthesamedonorandcystsofvariousagesfromthesamespecimenwerecompared.Also,samples14
TABLE3.Giardialambliacystsfromsymptomaticandasymptomatichumandonorsexcystedutilizingacysteine&phytone
containingmediuma.
Donorb

Sample

Cystage
(days)c

Mean%
excystationd

No.oftrials(n)range

93

n=1

90

n=1

11

17

n=31519

13

n=312

20

96

n=39598

53

n=34163

75

n=274&76

n=335

21

n=1

89

n=1

78

n=1

95

n=1

93

n=1

61

n=1

10

n=27&10
n=1

12

79

15

75

n=1

15

31

n=82540

15

n=26&10

16

27

n=1

17

93

n=39293

MethoddescribedintextforG.lamblia.

AandBwereasymptomaticcarriersandCthroughOweresymptomaticdonors.

Timefromcystpurificationtoexcystation.

Atleast200objects(IC,TET,PET)werecountedpertrial.

Page264

(Table1)fromthesamedonorexhibitedvariableexcystationlevels.Cystsfromseveralsymptomaticcarriers(Table3)showedwideexcystationvariability.
ThemethodsreportedherewhichincludedtheincorporationofphytoneorproteosepeptoneandcysteineintoHBSSwerefoundtobesimplerthanthosepreviously
published.ThelatterproceduresusecomplicatedmediasuchasHSP3(2)orTYI(4)andincorporatecrudetrypsin(10,14),bile(4,7)orwholeserum(1,2,4,7)into
theexcystationmedium.Thepreparationofthesemediaistimeconsumingandexposedthetrophozoitesandcyststosuchundesirablereagentsastrypsin,whole
serum,orbile.ForG.lambliaandG.muriscysts,itwasshownthattheexcystationprocessdidnotrequiretheuseofserum,specificenzymes,orbile.Aheat
inactivatedproteinorproteindigestsuchasphytonepeptone,butnotproteosepeptone,plusareducingagent,suchasLcysteine,werenecessaryintheexcystation
mediumforG.lamblia.G.murisdoesnotrequireareducingagentsuchasLcysteineinitsexcystationmedium.
Acknowledgements
IwishtothankFrankW.Schaefer,IIIfortheG.muriscystsandFloydFrostfortheG.lambliacystsfromhumandonors.Ialsowishtoacknowledgetheveryable
technicalassistanceofMelanieMormileandDebbieFlanigan.
Notice
ThisdocumenthasbeenreviewedinaccordancewithU.S.EnvironmentalProtectionAgencypolicyandapprovedforpublication.Mentionoftradenamesor
commercialproductsdoesnotconstituteendorsementorrecommendationforuse.
LiteratureCited
1.Bhatia,V.N.,andD.C.Warhurst.1981.HatchingandsubsequentcultivationofcystsofGiardiaintestinalisinDiamond'smedium.J.Trop.Med.Hyg.84:45.
2.Bingham,A.K.,andE.A.Meyer.1979.Giardiaexcystationcanbeinducedinvitroinacidicsolutions.Nature(Lond.)277:301302.
3.Bingham,A.K.,JarrollJr.,E.L.,andE.A.Meyer.1979.Giardiasp.:physicalfactorsofexcystationinvitro,andexcystationvs.eosinexclusionasdeterminants
ofviability.Exp.Parasitol.47:284291.
4.Feely,D.E.1986.AsimplifiedmethodforinvitroexcystationofGiardiamuris.J.Parasitol.72:474475.
5.Gillin,F.D.,andD.S.Reiner.1982.AttachmentoftheflagellateGiardialamblia:roleofreducingagents,serum,temperature,andioniccomposition.Mol.Cell.
Biol.2:369377.
6.Hegner,R.1927.ExcystationandinfectionintheratwithGiardialambliafromman.Am.J.Hyg.7:433447.
7.Kasprezak,W.,andA.C.Majewska.1983.InfectivityofGiardiasp.cystsinrelationtoeosinexclusionandexcystationinvitro.Tropenmed.Parasitol.34:70
72.
8.Marchin,G.L.,Fina,L.R.,Lambert,J.L.,andG.T.Fina.1983.EffectofresindisinfectantsI3andI5onGiardiamurisandGiardialamblia.Appl.Environ.
Microbiol.46:965969.
9.Meyer,E.A.1976.Giardialamblia:isolationandaxeniccultivation.Exp.Parasitol.39:101105.
10.Rice,E.W.,andF.W.Schaefer,III.1981.ImprovedinvitroexcystationprocedureforGiardialambliacysts.J.Clin.Microbiol.14:709710.
11.RobertsThomson,I.C.,Stevens,D.P.,Mahmound,A.A.F.,andK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gastroenterol.71:5761.
12.Sauch,J.F.1984.PurificationofGiardiamuriscystsbyvelocitysedimentation.Appl.Environ.Microbiol.48:454455.
13.Sauch,J.F.1985.UseofimmunofluorescenceandphasecontrastmicroscopyfordetectionandidentificationofGiardiacystsinwatersamples.Appl.Environ.
Microbiol.50:14341438.
14.Schaefer,F.W.,III,Rice,E.W.,andJ.C.Hoff.1984.FactorspromotinginvitroexcystationofGiardiamuriscysts.Trans.Roy.Soc.Trop.Med.Hyg.
78:795800.

Page265

AssessingGiardiaCystViabilitywithFluorogenicDyes:ComparisonstoAnimalInfectivityandCystMorphologybyLightand
ElectronMicroscopy
DanielG.Schupp*,MaryM.Januschka,andStanleyL.Erlandsen
DepartmentofCellBiologyandNeuroanatomy,4135JaccksonHall,UniversityofMinnesota,Minneapolis,Minnesota55455,U.S.A..
Thefluorogenicdyesfluoresceindiacetate(FDA)andpropidiumiodide(PI)wereusedtoassessGiardiacystviability.ViableGiardiacysts,possessinganintactcellmembrane,
retainedthedyeFDAandwereseentofluorescegreenat450490nm.NonviablecystsincorporatedthedyePIandfluorescedredat545546nm.NeonatalmicegivenFDA
positivecystinoculabecameinfectedasevidencedbyfecalcystsheddingandpositivepostmortemexaminationfortrophozoites.Noneofthemice,givencystinoculathatwerePI
positive,becameinfected.Usingdifferentialinterferencecontrastmicroscopy,comparisonsbetweenGiardiamuriscystslabeledwithFDAandthoselabeledwithPIshowedunique
morphologicaldifferencesincludingaprominentcystwallandperitrophicspaceintheFDApositivecysts,butnotinthePIpositivecysts.FDAcystspossessedahyaline
cytoplasmicappearancewhereasPIcystslookedgranularandtheintracellularorganelleswereeasilydetected.Preliminaryultrastructuralobservationshaveshownthat
nonviablecystspossessedacystwalllackingstructuralcontinuityandthatthetrophozoitemembraneadjacenttotheinneraspectofthecystwallwasdifficulttodelineate.Based
onourstudies,fluorogenicdyesprovideasimpleanddirectmethodofmeasuringGiardiacystviabilitythatcorrelateswellwithbothmorphologicaldifferencesandanimal
infectivity.

Introduction
Fluorogenicdyeshavebeenacceptedasasensitivemeansfordeterminationofcellviabilitysincetheirdescriptionin1966byRotmanandPapermaster(16).
Recently,thesedyeswereemployedascytochemicaltoolsformakingmeasurementsonmanycellularparameters.Theserangedfromcytosolicviscosity(5)andpH
(21)tonuclearcytoskeletalinteractions(13).Cellmigrationstudies(4)andviabilitydeterminationsinbothstaticsystems[countingabsolutenumbersofdifferentially
stainedcells(10)]anddynamicsystems[measuringantibodydependentcellcytotoxicityorintracellularkillingofbacteria(20,9)]havebeenmade.
MethodsthathavebeenusedtodetermineGiardiacystviabilityhaveincluded:dyeexclusion(2,11),excystationinvitro(2,3,11,14)andtheuseofanimalmodelsfor
infectivity(1,15).Drawbackstothesetechniqueshaverecentlybeendiscussed(17)andincludesubjectivity,time,expenseandlossofthetestcystpopulationfor
furtherexperimentalmanipulation.
Recently,ourlaboratoryhasreportedanexcellentcorrelationbetweenfluorogenicdyestainingofGiardiacystswithbothanimalinfectivityandcystmorphology
(17,18).Theincorporationofthefluorogenicdyes,fluoresceindiacetate(FDA)andpropidiumiodide(PI),hasbeenshowntorepresenteitherviableornonviableG.
muriscysts,respectively.Thishasbeenverifiedusingtheindirectanimalinfectivitytest.Furthermore,ithasbeendemonstratedthatdistinctmorphologicalfeatures
enabledviabilitydeterminationsofG.muriscyststobemadebydifferentialinterferencecontrast(DIC)microscopy.
ThemorphologicalbasisforstructuraldifferencesobservedbyDICinviableandnonviableG.muriscystswasinvestigatedbyelectronmicroscopy.Basedonthese
observations,wehaveproposedanexplanationfortheuptakeoffluorogenicdyebyviableandnonviablecystsandhavealsopresenteddatashowingthatcyst
morphologybyDICwascloselycoordinatedwithchangesinfluorogenicdyeuptakewhenviablecystsbecomenonviableasdefinedbyPIincorporation.
MaterialsandMethods
FluorogenicstainingsolutionsweremadeaccordingtothetechniquedescribedbySchuppandErlandsen(17).AllcystswereviewedoneitheraZeiss
photomicroscopeIIIoranOlympusBH2lightmicroscopeequippedwithepiillumination,DICandphaseoptics.CystswereisolatedfromthefecesofinfectedCF1
micebythetechniqueofRobertsThomsonet.al.(15).StainedcystswereseparatedusingaBectonDickinsonfluorescenceactivatedcellsorter(FACSIV).Sorted
groupsofeitherFDAorPIpositivecystswereverifiedbyfluorescencemicroscopy.AllaspectsoftheanimalinfectivitywereperformedasdescribedbySchuppand
Erlandsen(17).
Totestwhetherchangeshadoccuredincystmorphologyasafunctionoffluorogenicdyeuptake,cystswerestainedwithFDA/PI.Theywerethenpelletedand
placedonamicroscopeslide,andcoverslipped.AfieldcontainingcystsstainedwithFDAwasselectedandphotographedunderbothepiilluminationandDIC
optics.Thesamefieldwasphotographedatvaryingtimeintervalsforupto12hoursafterstaining.CystswerestainedwithFDAandPI,thensortedonaFACSIV,
andpreparedforelectronmicroscopyasdescribedbyShands(19).
*Correspondingauthor.

Page266
TABLE1.
Inoculum
Cysts/Animal

FluorogenicDye
Uptake

Controls:

Experimental:

GiardiaCystAppearanceDaysPostInoculation

ByDIC

0Cysts(N=8)

TypeofCyst
Morphology

None

0/8

0/8

11

0/8

0/8

1,000Cysts
(N=10)

FDA
Postive

Viable

0/10

10/10

10/10

10/10

5,000Cysts
(N=10)

PI
Positive

NonViable

0/10

0/10

0/10

0/10

ThesectionswerecutonaLKBHuxleyultramicrotome,stainedwithuranylacetateandleadcitrate(12)andexaminedwithaJEOL100CXelectronmicroscope.
Results
FluorogenicDyeUptakeandAnimalInfectivity
TheviabilityofGiardiamuriscystswasinvestigatedbyadministrationoffluorogenicdyeexposedcyststoneonatalmice.TheresultshavebeenshowninTable1.
ControlneonatalmicethatdidnotreceiveaninoculumofGiardiamuriscystsalsodidnotpasscystsintheirfecesduringthecourseoftheexperiment(day3,5,8,
11,postinoculation)norwereanytrophozoitespresentwithintheirsmallintestineatnecropsy.Noneofthemice(N=10)experimentallyinoculatedwithdosesof
1,000FDApositiveGiardiamuriscysts,isolatedbyFACS,shedcystsondaythreepostinoculation.However,100%ofthesemicewerepositiveforfecalcystson
days5,8,and11.NecropsyoftheanimalsinfectedwithFDApositivecystsrevealedGiardiatrophozoiteswithinthesmallintestine.Tenmicewereeachinoculated
with5,000PIpositivecysts.InoculationofPIpositiveGiardiacystsfailedtoestablishaninfectionasevidencedby(1)thelackofGiardiacystsinthefecesand(2)
theabsenceofanytrophozoiteswithinthesmallintestineatnecropsy.
CystViabilitybyLightMicroscopy
ThemorphologicalappearanceofGiardiamuriscystswasstudiedbyDICmicroscopyandcorrelatedwithincorporationoffluorogenicdyesasseeninFigures1a
1c.Thistechniqueshowedstrikingdifferencesbetweenthenonviableandtheviablecysts.TheviableFDAcystsrevealedbyDICaclearlydelineatedcystwall,a
spacebetweencystwallandcytoplasm[referredtoastheperitrophicspace(6)]andahyalineappearancetothecystwhichmadetheviewingofintracellulardetail
difficult(Figure1a).ThenonviablePIstainedcystsdisplayedeasilyrecognizablecytoplasmicorganellesincludingtwotofourpolarnuclei,intracytoplasmicflagellar
axonemesandcurvedportionsofthedisassembledadhesivedisc.ByDIC,thenonviablePIcystcytoplasmhadafinegranularappearance,asopposedtothehyaline
appearanceoftheviableFDAcysts.
ToinvestigatehowcloselyintimeGiardiacystmorphologybyDICwascoupledtochangesinthefluorogenicstainingpattern,asampleoffluorogenicdyeexposed
G.muriscystswassealedunderacoverslip.AmicroscopicfieldwasselectedthatcontainedviableFDAcysts,viablenonstainingcysts,(17)andnonviablePI
cysts.ArepresentativefieldcontainingbothviableFDAstainingandnonstainingviablecystswasphotographed5hoursafterstainingwithFDA/PIbyDIC(Figure
2a),at

Figure13.
ComparisonbetweenviableandnonviableG.muriscystsusing
DIC(1a,2a,and3a)andfluorescencemicroscopy(1b,c2b,cand3b,c).
InFigure1a,notethestructuraldifferences(nuclei,flagellaraxonemes,
andcytoplasmicappearance)betweenviable(V)andnonviable(N)cysts.
n,nucleips,peritrophicspace.InFigure1b,viableG.muriscysts
incorporatingFDAfluorescegreenatanexcitationwavelengthof450490nm,
whilethenonviablecyststainingwithPIfluorescesorange.Thestaining
forFDAappearsaroundtheperipheryofthecystandintheperitrophic
space(ps).InFigure1c,onlythenonviablecyststainingwithPIfluoresces
redatanexcitationwavelengthof545546nm.CystsincorporatingFDA
arenotexcitedtofluoresceatthiswavelength.Figure2aillustratesby
DICagroupofG.muriscystsafter5hoursexposuretoFDA/PI.
Allcystsareviablebymorphologiccriteriaexceptforonenonviablecyst(N).
ExaminationofthissamegroupofcystsforFDAuptake(450490nm
excitation)isshowninFigure2bandforPIstaining(545546nmexcitation)
inFigure2c.ThetwoviablecystsindicatedwithdoublearrowsinFigure2a
areseenonehourlaterinFigure3abyDIC.Thesecystsarenownonviable
(N)bymorphologiccriteriaandalsofluorescepositivelyforPIasshown
at450490nmexcitationinFigure3band545546nmexcitationinFigure3c.
Inaddition,comparisonofFigures2aand2btoFigures3aand3b,
showsthatwithinthesameonehourtimeframe,twononstainingcysts
arenowaccumulatingFDA.BarequalsfivemicronsforFigure1ac
barequalsfivemicronsforFigures2acand3ac.

Page267

Figures45.
IllustratedinFigure4isanexampleofaviableG.muriscyst
isolatedbyFACSusingthefluorogenicdyeFDA.Thecystwall
(arrowheads)andperitrophicspace(ps)areeasilydistinguished.Within
thecytoplasm,variousorganellesareseenincludingnuclei(n),flagellar
axonemes(f),andportionsoftheadhesivedisc(ad).Smallvacuoles
arepresentattheperipheryoftheorganism.InFigure5,anexample
ofanonviablecystofG.murisisolatedbyFACSusingthefluorogenic
dyePI.IncomparisontotheviablecystseeninFigure4,thecyst
wall(betweenarrowheads)appearstobeincreasedinwidthandalsois
partiallydisorganized.Theperitrophicspace(ps)ispresent,butisless
obviousthanthatseenintheviablecystinFigure4.Inthecytoplasm,
profilesoftheadhesivedisc(ad)andflagellaraxonemes(f)areseen,
withthelatteralsobeingpresentintheperitrophicspace.
BarequalsonemicronforFigures4and5.

450490nmforFDAuptake(Figure2b),andat545546nmforPIstaining(Figure2c).TheidenticalfieldwasagainphotographedbyDICandforfluorescenceone
hourlater,sixhoursafterexposuretoFDA/PI(Figure3).ThecomparisonofFDAstaining(compareFigures2bto3b)andPIstaining(compareFigures2cto3c)
indicatedthat,withinonehour,thesamecyststhatunderwentmorphologicalchangesandappearednonviablebyDICcriteria,alsohadincorporatedthefluorogenic
dyePI.Also,twoofthenonstainingviablecystsbecameFDApositiveduringthissametimeperiod.
ElectronMicroscopyofViableandNonviableCysts
ThebasisforstructuraldifferencesinGiardiacystmorphologyseenbyDICwasexaminedattheultrastructurallevelusingFDAandPIstainedGiardiacystssorted
byFACS.ThemorphologyofFDAstainedGiardiacysts,asillustratedinFigure4,wascharacterizedbythepresenceoftypicalcystorganelles,includingnuclei,
axonemesofflagella,dissembledportionsoftheadhesivedisc,peripheralvacuoles,theperitrophicspace,andacystwallcomposedofafinefibrillarlayer.Thecyst
wallwasapproximately0.200.25mthick,andincludedtwocystmembranesseparatedbyathinlayerofcytoplasm.Cytoplasmicstructuresthatcorrespondedto
theDICobservationsmadeofcystmorphology,werethewelldefinedcystwallandtheperitrophicspacebetweencystwallandcytoplasm.ThePIsortedGiardia
cysts,asrepresentedinFigure5,werequitedistinctfromFDAstainedcystsinthatthecystwallappearedtobedisorganizedandtohaveincreasedinwidth.The
ultrastructuralchangesincystwallmorphologyandtheextremelydiminishedperitrophicspacecorrespondedtotheDICimageofPIstainedGiardiacysts.The
cytoplasminPIstainedcystsappearedlessintenselystained(contrastoforganellestocytoplasm)thantheFDAstainedcysts,butotherwise,noobviousstructural
counterpartwasobservedthatcouldexplainthedifferencesincytoplasmicappearanceseenbyDIC,includingthehyalinenatureofFDAcystsversusthegranular
appearanceofPIcysts.
Discussion
Thefluorogenicdyesofferarapid,directmeasureofGiardiacystviabilitybasedontheselectivepermeabilityofanintactcellmembranetothedyesFDAandPI.
Figure6displaysadiagramrepresentingtheproposedmechanismoffluorogenicdyestaininginGiardiacysts.Thefluorogenicdye,FDA,isanonpolarmoleculethat
iscapableoftraversing,bydiffusion,anintactbilipidmembrane.Uponentryintoacell,intracellularenzymescleavetheacetategroupsoffofFDA,leavingthepolar
fluoresceinmoleculewithintheintracellularcompartment.TheoriginalFDAparentmoleculeisnonfluorescent,whereas,thehydrolyzedproduct,fluorescein,ishighly
fluorescent.Sincetheproductinthisreactionispolartheaccumulationoffluorescencewithinthecellisfavored.However,cellsthatdonotpossessanintactplasma
membrane(orcystwallmembrane)arenotabletoretainfluorescein,andthuswillnotstainpositivelywithFDA.Inaddition,itmaybeproposedthatcystswithoutan
intactmembranemayhavelostmostoftheiroriginalenzymaticactivityandconsequentlywillnotbeabletoconvertthefluorogenicsubstratetothefluorescent
product.ThemechanismofstainingforPIisquitedifferent.ThelargeplanarphenanthridiumringstructureofPIisunabletotraversetheintactbilipidmembraneby
diffusion.Therefore,onlycyststhathavelosttheirmembraneintegritywillallowPItoenterthecytoplasm.Onceinsidethecyst,thisfluorochromemayeither
intercalateintonucleicacids,orduetoitshydrophilicnature,interactwithothercytoplasmicsites(7).PIexhibitsaslightfluorescencepriortointeractionwithnucleic
acids,butuponintercalationthefluorescenceintensityincreasesbyapproximately100fold(10).

Page268

Figure6.
Schematicdiagramofthemechanismforuptakeoffluorogenic
dyesbyG.muriscysts.Seetextforexplanation.

LysosomalenzymesthathavebeenproposedtocausehydrolysisoftheesterlinkageinFDAincludenonspecificesterasesaswellasproteinases(16).Thestaining
patternofFDAinG.muriscyststypicallyshowedinitialstainingintheregionofperipheralvacuolesanditwasnotuncommontoobserveintensestainingfor
fluoresceinintheperitrophicspace.Thelysosomalenzymes,acidphosphatase(8)andarylsulfatase(DennisFeely,personalcommunication)havebeendemonstrated
intheperipheralvacuolesofGiardiaandthereforetheyareconsideredtobelysosomalorganelles.This,togetherwithrecentevidencethattheperitrophicspacewas
derivedfromlysosomallikeperipheralvacuolesduringcystformation(Erlandsenetal.,61stAnnualMeetingoftheAmericanSocietyofParasitologists,abstract95,
1986)haslentsupporttoourproposalthatthesecellularsitesmaybethelocationofthehydrolysisofFDA.
UltrastructuralexaminationofFDAandPIstainedG.muriscystssortedbyFACSrevealedthatPIstainedcysts(nonviable)didnotpossessanextensiveperitrophic
spaceandthecystwallappearedtobepartiallydegraded.Theappearanceofthiscystwallmorphologicallyresembledacystwallafterithasundergonedegradative
changesduringexcystation.Therefore,itsappearanceinPIstainedcystscouldreflectproteolysisresultingfromdamagetolysosomalcompartmentswithinthecyst.
ThesechangeswouldbecompatiblewiththeuptakeofPIsincemembraneintegritywasrequiredfortheexclusionofthisfluorogenicdyefromviablecysts.No
ultrastructuralcorrelationwasfoundtoexplaintheintracytoplasmicdetailseenbyDICinPIstainedcysts.Presumably,theuniquemorphologymayhavebeenrelated
todifferencesincytoplasmicrefractiveindexcausedbythelossofmembranepermeabilitythatcouldnotbeseenbyelectronmicroscopy.Previously,ourlaboratory
(17,18)hasshownthatFDAandPIstainedG.muriscystshadadistinctmorphologicalappearancewhenviewedbybrightfield,phase,orDICmicroscopy.Here,
wehavedemonstratedthat,whenviablecystshavelosttheirmembraneintegrityandincorporatedPI,themorphologyofthecystwasconvertedfromviableto
nonviablewithintheshortesttimeintervalstudied(lessthanonehour).ThisseemedtoindicatethatthemorphologicalappearancebyDICwascloselycoupledintime
withdyeuptake,therefore,eithermethodwouldseemtobeanaccuratereflectanceofcystviabilitysincethelagtimebetweenPIuptakeandmorphologicalchange
wassmall.
Acknowledgements
TheauthorswishtothankMs.LeeAnnSherlockandMs.AndreaToedterfortheirtechnicalassistance,andDr.W.J.Bemrickforhisadviceandcriticalreviewofthe
manuscript.AlthoughtheresearchdescribedinthisarticlehasbeenfundedinpartbytheU.S.EnvironmentalProtectionAgencythroughcooperativeagreementNo.
CR811834totheUniversityofMinnesota,itdoesnotnecessarilyreflecttheviewsoftheAgencyandnoofficialendorsementshouldbeinferred.
LiteratureCited
1.Belosevic,M.,Faubert,G.M.,MacLean,J.D.,Law,C.,andN.A.Croll.1983.Giardialambliainfectionsinmongoliangerbils:ananimalmodel.J.Infect.Dis.
147(2):222226.
2.Bingham,A.K.,Jarroll,E.L.,Meyer,E.A.,andS.Radulescu.1979.Giardiasp.:physicalfactorsofexcystationinvitro,andexcystationvs.eosinexclusionas
determinantsofviability.Exp.Parasitol.47:281291.
3.Bingham,A.K.,andMeyer,E.A.1979.Giardiaexcystationcanbeinducedinvitroinacidicsolutions.Nature(London)277:301302.
4.Brenan,M.,andC.R.Parish.1984.Intracellularfluorescentlabellingofcellsforanalysisoflymphocytemigration.J.Immunol.Methods74(1):3138.
5.Cercek,L.,andB.Cercek.1972.Studiesonthestructurednessofcytoplasmandratesofenzymatichydrolysisingrowingyeastcells.I.Changesinducedby
ionizingradiation.Int.J.Radiat.Biol.21:445453.
6.Coggins,J.R.,andF.W.Schaefer,III.1986.Giardiamuris:ultrastructuralanalysisoftheinvitroexcystation.ExperimentalParasitology61:219228.
7.Cox,B.A.,Yielding,L.W.,andK.L.Yielding.1984.SubcellularlocalizationofphotoreactiveethidiumanalogsinTrypanosomabruceibyfluoresence
microscopy.J.Parasit.70(5):694702.
8.Feely,D.E.,andJ.K.Dyer.1987.LocalizationofacidphosphataseactivityinGiardialambliaandGiardiamuristrophozoites.J.Protozool.34:8083.
9.Jackson,P.R.,Pappas,M.G.,andB.D.Hansen.1985.Fluorogenicsubstratedetectionofviableintracellularandextracellularpathogenicprotozoa.Science
227:435438.
10.JonesK.W.,andJ.A.Sneft.1985.Animprovedmethodtodetermineviabilitybysimultaneousstainingwithfluoresceindiacetatepropidiumiodide.J.Histochem.
Cytochem.33:7779.

Page269

11.Kasprzak,W.,andA.C.Majewska.1983.InfectivityofGiardiasp.cystsinrelationtoeosinexclusionandexcystationinvitro.Tropenmed.Parasit.34:7072.
12.Knight,D.P.1977.Cytologicalstainingmethodsinelectronmicroscopy.In:StainingMethodsforSectionedMaterial.P.R.LewisandD.P.Knight(eds).
NorthHolland,Amsterdam.pp.2976.
13.Ockleford,C.D.,Hsi,B.L.,Wakely,J.,Badley,R.A.,Whyte,A.,andW.P.Faulk.1981.Propidiumiodideasanuclearmarkerinimmunofluorescence.I.Use
withtissueandcytoskeletonstudies.J.Immunol.Methods43(3):261267.
14.Rice,E.W.,andF.W.Schaefer.1982.AnimprovedinvitroexcystationprocedureforGiardialambliacysts.J.Clin.Microbiol.14:709710.
15.RobertsThomsom,I.C.,Stevens,D.P.,Mahmoud,A.A.F.,andK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gasteroenterology71:5761.
16.Rotman,B.,andB.W.Papermaster.1966.Membranepropertiesoflivingmamaliancellsasstudiedbyenzymatichydrolysisoffluorogenicesters.Proc.Natl.
Acad.Sci.55:134141.
17.Schupp,D.G.,andS.L.Erlandsen.1987.AnewmethodtodetermineGiardiacystviability:correlationbetweenfluoresceindiacetate/propidiumiodidestaining
andanimalinfectivity.Appl.Envir.Micro.53:704707.
18.Schupp,D.G.,andS.L.Erlandsen.1987.DeterminationofGiardiamuriscystviabilitybydifferentialinterferencecontrast,phase,orbrightfieldmicroscopy.J.
Parasitol.73:723729.
19.Shands,J.W.1968.Embeddingfreefloatingcellsandmicroscopicparticles:serumalbumincoagulumepoxyresin.Staintechnol.43.15.
20.Szollosi,J.,Kertai,P.,Somogyi,B.,andS.Damjanovich.1981.Characterizationoflivingnormalandleukemicmouselymphocyteswithfluoresceindiacetate.J.
Histochem.Cytochem.29:503510.
21.Thomas,J.A.,Buschbaum,R.N.,Zimmick,A.,andE.Racker.1979.IntracellularpHmeasurementsinEhrlichascitestumorcellsutilizingspectroscopicprobes
generatedinsitu.Biochemistry18:22102218.

Page271

PANELDISCUSSIONS

Page273

PanelDiscussiononExcystationandEncystation
Chairperson:FrancesD.Gillin*
PanelMembers:E.A.Meyer,S.ErlandsenandC.Sterling
DepartmentofPathologyH811F,UniversityofCalifornia,SanDiegoMedicalCenter,225DickinsonStreet,SanDiego,California92103,U.S.A..
TheCystFormandtheImportanceofEncystationandExcystation
IfGiardiatrophozoitesmove''downstream",theymustcompletetheirlifecyclebyencystingortheywilldie,sincetrophozoitesdonotnaturallysurviveoutsidethe
host.Incontrast,Giardialambliacystsarewelladaptedtosurvivalandremainviableformonthsincoldwater.Despitetheimportanceofcysts,relativelylittleis
knownaboutthisform,largelybecauseithadnotbeeninducedinvitrountilrecently.Instead,investigatorshaverelieduponcystsisolatedfromfecesofinfected
humansoranimals.Purificationoffecalcystsisalengthyandunpleasantprocessandsomecontaminantsalwaysremain.Moreover,boththenumbersandqualityof
cystsseemtovarygreatlyfrompatienttopatientandevenfrommultipleisolationsfromasingledonor.Therefore,thegoalofthisworkshopwastosummarize
researchtodateoncompletingtheGiardialifecycleinvitro.
Excystation
InfectionwithGiardiaisinitiatedbyingestionofcystsfromfecallycontaminatedwaterorfood.Cystspassthroughthestomachandintothesmallintestinewherethe
newlyemergedtrophozoiteimmediatelydivides,givingrisetotwodaughterswhichcanmultiply,colonizethesmallintestineandcausesymptoms.Itiscrucialthatthe
trophozoitenotemergeuntilthecystpassesintotheduodenum,sinceitwouldbekilledbygastricacid.Understandingexcystationisimportantbecauseinterferingwith
thisprocesswouldpreventinfection.
ThefirstsuccessininducingG.lambliacystspurifiedfromfecestoexcystinvitrowasachievedinthelaboratoryofDr.E.A.Meyer.Thekeyobservationwasthat
excystationwastriggeredbyexposureofcyststolowpH(~102MHCl).TrophozoiteemergenceoccursaftertransfertoanutrientmediumatneutralpH.This
mimicsexposurefirsttogastricacid,thenpassageintotheduodenum.
Dr.Meyersummarizedpublishedmethodsforachievingexcystationinvitro.Studiesfromhislaboratoryfocussedontheoptimaltimeandtemperatureofcyststorage
priortoexcystation,aswellasthepHandtemperaturesofthetriggeringandemergencesteps.
Modificationsoftheinitialprocedure(mainlyattheEPA)identifiedotherphysiologicfactorssuchasbicarbonateandreducingagents(glutathione,cysteine)which
tendtoimprovetheefficiencyofexcystationinvitro.
Muchremainstobelearnedaboutexcystation,sincetheefficienciesobservedwithG.lambliacystsfromhumanfecesarefrequentlylowandhighlyvariable,in
contrasttocystsofG.muris.Moreover,G.muriscystscanexcystimmediatelyaftershedding,whileexcystationofG.lambliacystincreasesaftera"maturation"
periodofseveraldaysat4C.
Encystation
Understandingoftheprocessofencystationislesscompleteasthisisamuchmorecomplexprocessofcellulardifferentiationinwhichtheactive,mobiletrophozoite
roundsupandsecretesanewcystwallaroundit.Moreover,neithertheanatomicsitenorthefactorswhichinduceencystationinthehosthadbeenclearlyidentified.
RecentpublishedandunpublishedstudiesoninducingencystationinvitroweresummarizedbyDr.'sErlandsen,Sterling,andGillin.Theapproachesofeach
laboratorydifferedgreatly,althoughacommonmotifwasmodificationofthebilecontentofthemedium.
Dr.Erlandsen'sgroupinducedencystationofGiardiatrophozoitesfromthemuskrat.Thecystsobtainedweresimilartothoseisolatedfrommuskratfecesby
differentialinterferenceaswellaselectronmicroscopyanduptakeorexclusionoffluorogenicdyes(thisvolume).
Dr.SterlingreportedthatthemajorgoalofhisgroupwastoprepareG.lambliacystsfreeoffecalcontaminationwhichcouldbeutilizedforisolationofcystspecific
monoclonalantibodies(thisvolume).Theirmethodwasbasedonthehypothesisthatlargeintestinalconditions,suchasanaerobiasis,cellularcrowding,andremovalof
waterwouldbeimportantinencystation.
Dr.GillinsummarizedquantitativestudiesfromherlaboratorywhichshowedthatlargenumbersofwaterresistantcystswereinducedbyexposingculturedG.lamblia
trophozoitestofactorswhicharepresentinthehumansmallintestine,suchasprimarybilesaltsandfreefattyacids.Cystspecificantigensweredemonstratedby
Westernblottingandincreasedactivityofchitinsynthetasewasobservedinencystingcultures.
Dr.GillinreportedlowlevelsofexcystationandsuccessfulinfectionofsucklingmicewithinvitroderivedhumansourceG.lambliacysts.Dr.Erlandsenreported
excystationofinvitroderivedcystsfromhismuskratmodel,alsoatalowfrequency.Therefore,muchworkremainsinunderstandingtheregulationofencystation,as
wellastheproductionofbiologicallyactivecysts.
*CorrespondingAuthor

Page275

PanelDiscussionontheImplicationsofRegulatoryChangesforWaterTreatmentintheUnitedStates
StigRegli*,A.Amirtharajah,B.Borup,CHibler,J.Hoff,andR.Tobin.
OfficeofDrinkingWater,U.S.EnvironmentalProtectionAgency,401MStreet,S.W.,Washington,D.C.20460,U.S.A..
Introduction
EPAintendstoproposeandpromulgateSurfaceWaterTreatmentRequirements(SWTR)in1987whichwill:(a)providecriteriabywhichstateregulatoryagencies
willdeterminewhichsystemsusingsurfacewatersourceswillberequiredtofilter,(b)setdisinfectiontreatmentrequirements,and(c)regulateforGiardialamblia,
viruses,heterotrophicplatecountbacteria,Legionella,andturbidity.Theproposedcriteriawillspecifyminimumtreatmentrequirementsforachievingatleast99.9%
removaland/orinactivationofGiardialambliaand99.99%removaland/orinactivationofvirusesbyallsurfacewatersystems.Criteriafordeterminingwhether
systemswillberequiredtofilterincluderawwaterqualitylimits(turbidity,andtotalcoliformsorfecalcoliforms),maintenanceofaneffectivewatershedcontrol
program,disinfectionwhichachievesatheoretical99.9%inactivationofGiardialambliaand99.99%inactivationofviruses(asdeterminedbytheproductof
concentration[C,mg/L]multipliedbydisinfectantcontacttime[T,minutes],i.e.,CTvalues),compliancewiththetotaltrihalomethaneMaximumContaminantLevel
(MCL)forsystemsservinggreaterthan10,000people,andcompliancewithalongtermtotalcoliformstandardfordistributionsystemmonitoring,alsotobe
proposedatthesametimeastheSWTR.NorequirementsforGiardiamonitoringarespecifiedinthedraftrule.
EPAconsideredtheuseofariskmodel(18,6)inthedevelopmentoftheproposedSWTR.Applicationoftheriskmodelintherulewouldallowutilitiestoprovide
lessstringentdisinfection(i.e.,lowerCTvalues)thanthatnecessarytoachievetheminimum99.9and99.99percentinactivationofGiardialamblia,andviruses,
respectively,ifitcouldbedemonstratedthatthepopulationservedwouldnotbeexposedaboveanacceptablelevelofhealthrisk.Intheriskmodelitisassumedthat
anannualinfectionratecouldbeestimatedasafunctionofthenumberofrawwatersamplestaken,totalsamplevolume,percentrecoverybytheanalyticalmethod,
numberofGiardiacystsandvirusesdetected,percenttheoreticalinactivationbydisinfection,amountofwaterconsumedperpersonperday,andtheviabilityand
infectivityoftheorganism.
EPAdecidednottoincludetheriskmodelintheproposedregulationbecause:precision,efficiency,andsensitivityofanyanalyticalmethodformeasuringGiardia
lambliacystshavenotyetbeenadequatelydefinednostandardmethods,validationprocedures,orlaboratorycertificationproceduresareavailableforassuming
confidenceintheanalyticalmethodologyverylargenumbersofsampleswouldbeneededtoascertainthatasystemwasprovidingwaterbelowareasonablelevelof
healthriskanddisinfectiondataonwhichtobasetheoreticalinactivationisverylimited.
ThepanelmembersdiscussedsomeoftheissuesthatwouldneedtobeaddressedbeforeariskmodelcouldbeappliedtoregulateGiardialamblia.
Discussion
APossibleSamplingApproachforEvaluatingtheLevelofRisk(M.B.Borup)
Todevelopasamplingplanthatwillbeusedtoevaluatethequalityofaproductitisfrequentlystandardproceduretospecifytwovaluesofthemeannumberof
defectsperunitofproduct,saym0andm1wherem0<m1,andtwosmallprobabilities,saya0andb1.Ifthemeannumberofdefectsperunitofproductisequaltom0,
theproductisregardedassatisfactoryanditisdesirabletoconcludeitissatisfactorywithsomehighprobability,sayatleast1a0.Ifthemeannumberofdefectsper
unitofproductisequaltom1,theproductisunsatisfactoryanditisdesirablethatweconcludetheproductissatisfactorywithsomeverylowprobability,saynomore
thanb1.Thesamplingplanshouldthenbedesignedsothatthesecriteriamaybemet.Thenumberofsamplesrequiredwilldependonthevaluesofm0,m1,a0andb1.
Inthecaseofwaterbornepathogens,m0andm1arethenumberofpathogenicorganismsfoundpersamplevolumeofwater.Thenumberofpathogensfoundinthe
rawwatermayberelatedtothenumberofinfectionsper10,000peopleperyearinthefollowingmanner:

where:
mirepresentseitherm0orm1andisequaltotheaveragenumberoforganismsperrawwatersamplevolume.
Ei=thenumberofinfectionsper10,000peryearcorrespondingtomi.
Vs=samplevolume.
R=thefractionoforganismsthatarerecoveredbytheanalyticaltechniquesused.
V=volumeofwaterconsumed,10,000people/year
G=thefractionsofpathogensremovedorinactivatedinthewatertreatmentprocess.
*Correspondingauthor

Page276

Thisequationisbasedonseveralsimplifyingassumptions.First,itisassumedthatoneorganismcancauseinfection.Second,itisassumedthateachorganismwillbe
consumedbyadifferentindividual.Theseassumptionswillpresentaworstcaseandproduceaconservativeestimate,protectingpublichealth.
Ifyrepresentsthetotalnumberofpathogenicorganismsfoundinaseriesofsamplestakenfromtherawwatertheobjectiveofasamplingplanmaybeexpressedin
thefollowingrequirements:
P(y c) 1a0whenmi=m0
P(y c) b1whenmi=m1
where:
P(y c)representstheprobabilitythatthenumberofpathogensfoundinaseriesofsamplesislessthanorequaltoc.
c=themaximumallowablenumberofpathogens.
Todeterminethenumberofsamplesrequireditisnecessarytodetermineastatisticaldistributionthatwilldescribetheoccurrenceofpathogenicorganisms.Verylittle
informationisavailablethatcanbeusedtodetermineasuitablestatisticalmodelforGiardiaoccurrence.Amodelwhichisoftenusedtocomputeprobabilities
associatedwiththenumberofdefectspersampleisthePoissonModel.ThelargestbodyofdataonGiardiaoccurrence(9)doesnotprompttherejectionofthis
model.
UsingarelationshipbetweenthePoissondistributionandtheChisquaredistributionitcanbeshownthatthefollowingequationmustbesatisfiedtomeetthecriteria
describedabove:

where:
2

=chisquareprobabilities.

n=numberofsamplesrequired.
Considerthisexampleoftheapplicationofthisequation.Autilitywouldliketodetermineiftheirtreatmentprocessisadequatelyprotectingthepublichealthfrom
giardiasis.Ithasbeendeterminedthatwaterqualitywhichcauseslessthanorequaltooneinfectionper10,000peopleperyearisacceptableandthatitisdesirableto
acceptthiswaterassatisfactoryatleast90%ofthetime.Ithasalsobeendeterminedthatwaterqualitywhichcausesfiveinfectionsper10,000peopleperyearis
unacceptableandthiswatershouldnotbedeterminedtobeacceptablemorethan5%ofthetime.Testshaveshownthatthetreatmentprocessusedbythis
hypotheticalutilityremovesorinactivates99%oftheGiardiacystsintherawwater,buttheanalyticalmethodusedtoidentifythecystswillonlyidentify50%ofthe
cystsactuallypresentinthesample.
Fromthisinformation:
a0=0.10b1=0.05G=0.99R=0.50
assume:
1)Vs=500L
2)Eachpersondrinks2Lwaterperday
3)Ingestionofonecystwillcauseinfection
4)Eachcystpresentinthetreatedwaterwillcauseinfectioninadifferentperson
then:
V=10,000peoplex(2L/(person.day))x365days/year
=7,300,000L
m0=(1x500x0.50)/(7,300,000(10.99))
=0.003425
m1=(5x500x0.50)/(7,300,000(10.99))
=0.017125
Nowthefollowingequationmustbesatisfiedbytheselectionofnandc:

Thisequationissatisfiedwhenc=3andn 453.Thismeansiftheutilitytakes453,500Lsamplesoftherawwaterandtheyfindlessthan4cysts,theycanbe95%
surethewaterwillproducelessthan5giardiasisinfectionsper10,000peopleperyear.Usingthissametechniquethenumberofsamplesoftherawwaterrequired
undergivenconditionsarelistedinTable1.
Forexample,ifasystemachieved99.5%inactivationbydisinfection,anda50%recoveryintheanalysisforGiardiacysts,itwouldneedtocollect1,132rawwater
samples,eachof500liters,peryearanddeterminethattherewerenomorethan3Giardiacystsdetectedforthetreatedwatertobeofacceptablequality.
Thenumberofsamplesthatwouldberequiredtodemonstratethatareasonablelevelofriskwasbeingavoided,accordingtotheprecedinganalysis,makes
applicationoftheriskmodelcostprohibitive.
Questions/Comments
Regli:Ifdifferentassumptionsweremadethanthosediscussed,thenumberofsamplesthatwouldbeneededtodemonstratethatanacceptablerisklevelwasbeing
metcouldbesignificantlyreduced.Forexample,iftheaveragesamplevolumewere5,000Lversus500L,asappearstobepossibleinverylowturbidwaters,the
numberofsamplesthatwouldbeneededcouldbereducedbyanorderofmagnitude.Also,theassumptionthatallcystsareviableandwouldcauseinfectionif
ingestedisveryconservative.Underalessconservativeassumption,fewersampleswouldberequired.
Borup:Yes,thisistrue.TheassumptionsthatweremadewererequestedbyEPAtoallowforaconservativeanalysis.

Page277
TABLE1.Numberof500Lsamples(n)requiredtodeterminetheacceptabilityofatreatedwater.*

m0=m1/5

m0=m1/10

Recovery
%

Removal
%

m1

a0=0.1
c

b 1=0.05
n

a0=0.1
c

b 1=0.1
n

a0=0.10
c

b 1=0.05
n

a0=0.10
c

b 1=0.1
n

75

90

0.0005137

15,094

22,590

50

90

0.0003425

22,638

33,882

9,235

16,364

13,851

30

90

0.0002055

37,730

24,543

56,470

23,086

75

99

0.005137

1,510

40,906

2,260

924

50

99

0.0034247

1,637

2,264

3,389

1,386

30

99

0.0020548

2,455

3,773

5,647

2,309

75

99.5

4,091

0.010274

755

1,130

462

50

819

99.5

0.0068493

1,132

1,695

693

1,228

30

99.5

0.0041096

1,887

2,824

1,155

1,046

m1=averagenumberofcystsper500Lsampleofrawwater.
E1=numberofinfectionsper10,000peopleperyearcorrespondingtom1basedonthefollowing:
1)Vs=500L
2)onecystcausesinfection
3)eachcystwillbeingestedbyadifferentindividual
4)eachpersondrinks2Lwater/day
*Calculationsofm1arebasedonanE1valueofoneinfectionper10,000peopleperyear.

Regli:Onemorepointforclarification.Inconsideringapplicationoftheriskmodelintherule,itwasnotEPA'sintentiontorequireallsystems,wishingtoavoid
filtration,toconductGiardiamonitoring.Rather,thepurposeoftheriskmodelwastoofferutilitiesanalternativetomeetingtheminimumtreatmentperformance
requirementof99.9percentinactivationofGiardiacystsinordertoavoidfiltration.Ifalargeutilityhadinhousemonitoringcapabilityitmightbeabletodemonstrate,
throughmonitoringtherawwater,thata99%levelofinactivationwouldachieveafinishedwaterqualitybelowtheacceptablerisklevel,saylessthan1infectionper
10,000peopleperyear.Alowerlevelofdisinfectioncouldsavecostsandminimizedisinfectionbyproductformation,therebyreducinghealthriskfromcarcinogens.
Tobin:Inyouranalysis,Dr.Borup,youmadetheassumptionthatGiardiacystoccurrencecouldbecharacterizedasaPoissondistribution.Couldyouelaborateon
thebasisforthis?
Borup:Therewereverylimiteddataonwhichtobasethisassumption.However,aPoissondistributionappearsreasonablebasedonhistogramsofrawwaterdata
gatheredbyDr.Hiblerwhichincludedover600samplesfrom20sites,includinglakes,rivers,andcreeks.Consideringthesedataasawhole,onecannotrejectthat
thedatafitaPoissondistributionbasedontheChiSquarehypothesistest.
DeterminationofanAllowableMorbidityRate(R.S.Tobin)
Riskassessmenthasbecomeanintegralpartofdrinkingwaterregulationandguidelineestablishment.AccordingtotheschemeproposedbytheWorldHealth
Organization(28),therearefourmaincomponentsofriskassessment:hazardidentification,riskestimation,riskevaluationandriskmanagement.Theestablishmentof
apolicyforan"acceptable"levelofriskofinfectionbyGiardiafallswithintheareaofriskmanagement.
Previousstudieshaveindicatedthatcertainrisksaremoreacceptabletothepublicthanothers(21).Someofthefactorsinvolvedintheriskofgiardiasisfrompotable
watermitigateagainstahighdegreeofriskacceptancebythepublic.Forexample,therisksduetoGiardiamaybeconsideredasinvoluntaryrisks,immediaterisks,
newrisks,andrisksthatindividualshavetopayforthemselves(tosomeextent)inordertoremedy.Ontheotherhand,theriskswhichresultfromnonsecret(open)
activities,areratherdiffuserisksandresultfromknownnaturalcauses.
Experienceandcommonsenseinformsusthattherearesomeupperandlowerlimitstorisksthatcanbetermed"acceptable"(Table2).Fortheriskofdeathithas
beensuggestedthat,undercertainconditions,individualswouldacceptanupperlimitofriskof103,whereasthenegligibleriskofdeathisconsideredbetween106
and107(21).
TABLE2.Suggestedupperandlowerlimitsorrisktoanindividual(21)
UpperLimitofRisk
Imposingacontinualannualriskofdeathof1in100shouldbedescribedasunacceptable(1,000per100,000).
Animposedriskof1in1,000(100per100,000)isnottotallyunacceptable,providedtheindividualknewofthesituation,
felthehadsomebenefitasaresult,andknewthateverythingpossiblehadbeendonetoreducetherisk.
NegligibleLevelofRisk
Fewpeoplewouldcommittheirownresourcestoreduceanannuallevelofriskthatwasalreadyaslowas105,andeven
fewerat106
Thefigureof106isprobablyanappropriatenegligibleriskexceptforclearcausalrelationshipswithconsumerproducts.

Page278
TABLE3.WaterborneoutbreaksofgiardiasisintheUnitedStates.
Year

Outbreaks

Cases

Cases/100,000**

1972

124

0.06

1973

73

0.04

1974

4,930

2.47

1975

0.005

1976

639

0.32

1977

1,012

0.51

1978

5,171

2.59

1979

5,864

2.93

1980

1,730

0.87

1981

11

311

0.16

Mean:

1,986

0.99

*ModifiedfromCraun(4)
**Approximation,basedonU.S.populationof200106.

AreviewofsomeoftheGiardiastatisticsinNorthAmericarevealsthattheknownwaterborneoutbreaksintheU.S.overa10yearperiodresultsinariskofabout
105(Table3).Thetotal(waterborneplusnonwaterborne)diseaseintheU.S.isnotknown.Ontheotherhand,dataareavailablefortotalgiardiasisinCanada
(Table4).Summarydatafromtheyears19831986revealedanaverageCanadawideriskofabout30105.Ifweassumethatwaterbornegiardiasiscauses
approximately10%ofthisfigure,theriskbecomes3105,relativelyclosetotheU.S.value.
ItisalsoinstructivetocomparethecurrentlevelofriskposedbyagentsotherthanGiardia.Gerba(7)hascompletedariskassessmentofvirusesinpotablewater,
basedonthestringencyofproposedviralstandards(Table5).Ifalimitforvirusesinpotablewaterwassetat1plaqueformingunitper1,000L,theannualriskof
infectionwouldbe1.5103,or150casesper100,000personsperyear.Basedonthesameviruslimit,thepredictedannualriskofdeathfromhepatitisAviruswas
estimatedas1.1105Obviously,theseriskswouldbeshiftedupwardordownward,dependingonthevirusstandardestablished.
Anotherareawheremicrobialriskhasbeenquantifiedisforbathingbeaches.EpidemiologicalstudieshavebeenusedbyU.S.EPAtodeveloprecommendationsfor
bathingwaterquality(Table6).Thecriteriaforfreshwaterpredictanillnessrateof80per105bathersandformarinewaters,theypredict190per105bathers.(8)
predictbetween120and1,500illnessesper105bathers,basedonthesameepidemiologicalevidence.
TABLE4.TotalgiardiasisinCanada.
PEI

NS

NB

QUE

ONT

MAN

SASK

ALB

BC

YT

NWT

CDA

rate/105*

Year

NFLD

1983

82

21

67

22

2,942

112

456

812

43

4,562

22.8

1984

51

11

107

275

3,692

258

1,362

895

16

36

6,710

30.5

1985

47

14

111

349

3,713

220

1,526

1,244

22

17

7,268

33.0

1986**

54

23

97

388

3,382

222

1,244

1,980

29

7,431

33.8

*Basedonapopulationof2210 .
**EstimatebasedondatatoAugust30,1986.

TABLE5.Estimatedrisksofinfectionduetolowlevelsofvirusindrinkingwater.
Standardof
1PFUper

Annual
Risk

PredictedCases
per100,000/yr.

10L

1.5101

15,000

100L

1.5102

1,500

1,000L

1.5103

150

10,000L

1.5104

15

BasedonGerba(7)

Whenallofthesemicrobiologicalriskdataaresummarized(Table6),theexampleof10allowablewaterborneillnessesper100,000peopleperyearduetoGiardia
(18)appearstobewithinthe"expected"riskrange.ThisvaluewouldbeaboutonethirdthecurrenttotalGiardiaillnessrateinCanada(oraboutthreetimesthe
currentbestestimateofillnessduetowaterborneGiardiainCanada)andabouttentimestheannualaveragewaterbornediseasecasesintheU.S.Itis,perhaps,
closetothecurrentrisklevelsinareasinNorthAmericawhereGiardiaareprevalentinthewatersupplies.Thisrisklevelis,however,moreprotectivethancurrently
acceptedrecreationalwaterqualityguidelines.Itisacknowledgedthatrecreationalbathinginvolvesavoluntarychoicewheretheindividualmayacceptacertainrisk
whichisbalancedbytheperceivedbenefits.
Inconclusion,arisklevelof10per105personsperyearisnotoutofproportionwithothercommonmicrobiologicalrisks.Whetherornotitwouldbethemost
appropriateor"acceptable"levelofriskstillrequirescarefulconsideration.
Questions/Comments
Borup:Couldyouelaborateontheacceptablevirusrisklevelof150infectionsper100,000peopleperyearwhichyoumentionedhadbeenproposed.Shouldn'tan
acceptableGiardiarisklevelbeinlinewithanacceptablevirusrisklevel?
Tobin:Iwasnotsuggestingthattherisklevelof150/100,000forviruseshadbeenproposed,onlywhattheriskwouldberelativetoafinishedwaterconcentrationof
1virus/1,000liters.Ifastandardweresetat1virus/10,000literstheriskwouldbe15/100,000peopleperyear.
Audience:Howwouldyoucomparethesameriskleveldemonstratedbyapublicwatersystemwhichfilteredanddisinfectedversusasystemwithaprotected
watershedwhichonlydisinfected?

Page279
TABLE6.Currentandproposedmicrobiologicalrisksfromwater.
Predictedillness
per100,000persons

Criteria
Recreationalwater:

126E.colior35enterococcus/100mLfreshwater

80a

35enterococcus/100mL,marinewater

190a

200fecalcoliforms

1201,500b

Drinkingwater:

10coliforms/100mL

unknown

1virus/1,000L

150(infections)c

1.1(deaths,HAV)c

Giardiaindrinkingwater

10?

(26)

(8)

(7)

Tobin:Goodquestion.Ihaven'taddressedthatissue.DoyousettheriskforagivensystembasedontheaverageriskexpectedacrosstheU.S.ordoyouconsider
differentsystemsbeingatmuchgreaterriskthanothersandregulateaccordingly?
AllowableConcentrationsofCystsinRawandFinishedWater(A.Amirtharajah)
Itisevidentfromrecentstudiesthatthereexistsatleasttwoavenuestowardsestablishingrelationshipsbetweencystlevelsintherawandfinishedwaterfrom
treatmentplants.Onepossibleapproachistoestablishthecystconcentrationinthefinishedwateronthebasisofagivenrisklevelofinfectionsper10,000peopleper
yearandbackcalculatetheconcentrationsintherawwaterbyusingtheexpectedremovalsandinactivationsofcyststhroughawatertreatmentplant.Thiswouldthen
leadtoestimatesofrawwatersamplesizeandnumberofsamplestomeettherisklevelspecified.Thevalidityofthisapproachneedstobeconfirmedbyempirical
results.Analternateapproachistoestablishasignificantmonitoringprogramfortherawandfinishedwatercystlevelsinavarietyofwatertreatmentplants,which
wouldcoversomeepisodesofgiardiasisandhencedeterminetheallowablecystlevelsinthefinishedwater.Theunderlyingpremiseofthisapproachisthatasystem
whichhasnohistoryofdiseaseoutbreak,orrecordofincreasedGiardiaincidencehasanacceptablefinishedwaterGiardiacystconcentration.Thelatterapproach
isinfactanembryonicstageforthefuturedevelopmentofamaximumcontaminantlevel(MCL)forGiardia.
Inbothapproachestheestimatedremovalsandinactivationsofcyststhroughthesequenceofprocessesinawatertreatmentplantistheessentiallinkbetweenraw
waterconcentrationsandfinishedwaterlevels.Amirtharajah(1)haspresentedareviewoftheliteraturewhichsummarizestheexpectedtransformationsofGiardia
cystsinwatertreatmentprocesses.Figure1illustratestheexpectedtransformations(removalsandinactivations)ofcontaminantsinaconventionalplantandTable7
summarizesextensivedatafrompilotplantstudiesfortheremovalefficienciesofGiardiabywatertreatmentprocesses.Table7indicatesthatfiltrationwithgood
coagulationwillremove95to99.9%ofthecysts,whilepoorcoagulationornocoagulationpriortofiltrationwilldroptheremovalefficiencytolowlevelsof10to
70%.Includingtheeffectofdisinfection,theanticipatedtotalremovalandinactivationis90to99.9%orgreater,dependinguponthelevelofdisinfectionprovided.
Thesetransformationefficienciesmaybeusedtocouplethefinishedwatercystconcentrationstotherawwatercystlevels.
Hibler(9)hascollectedextensivedataontheoccurrenceofGiardialambliacystsinrawandfinishedwatersoftreatmentplants.Hisdatasetincludesapproximately
6,000samples,analyzedfor302citiesoveraperiodof10years.Thefrequenciesofsamplingrangedfrom12samplespermonthto510samplesperweek.Hibler
(9)hasrecentlyanalyzedthesedata.
AsampleofhisdatawasextractedandrearrangedasshowninTable8.Somegeneralconclusionsoncystconcentrationsintherawandfinishedwatersareclearly
evidentinthedatashowninTable8.
1.Thefractionofsampleswithnodetectablecystsintherawwatersourcesrangefrom0%to90.5%withatypicalvalueofapproximately70%i.e.,typically30%of
thesamplesanalyzedwerepositive.Therefore,cystsseemtobeubiquitousinsurfacewatersourcesinthecountry.
2.Thelevelsofcystsinrawwatersarecommonly010per100gallons.
3.Inwelloperatedconventionalplantsthecystconcentrationsinthefinishedwaterswerezeroinallofthesamplestested.

Figure1.
Estimatesofcontaminanttransformationsina"standardsystem".

Page280
TABLE7.RemovalefficienciesofGiardialambliabywatertreatmentprocesses
Rawwater
concentration

Unitproces

Percentremoval

Operatingparameters

Reference**

1.RapidFiltrationwith
coagulationsedimentation

231,200/L

96.999.9

min.alum=10mg/L
opt.pH=6.5
filt.rate=4.99.8m/hr

DeWalleetal.
(1984)

2.DirectFiltration

20106

95.999.9

min.alum=10mg/L
pHrange=5.66.8

DeWalleetal.
(1984)

withcoagulation

(asslug)

48

filt.rate=4.99.8m/hr

nocoagulation

eff.NTU/inf.NTU=(0.020.5)/(0.71.9)

eff.poorduringripening

withflocculation

9599

alum=25mg/L
polymer(Magnifloc572CR)=12
mg/L
temp=5C,18Ceff.NTU/inf.NTU=
0.05/1.0
Filt.rate=4.818.8m/hr

nocoagulation

1070

AlAnietal.
(1984)

3.DiatomaceousEarth

1.510 9.010 /L 9999.99

filteraid=20mg/Lbodyfeed

DeWalleetal.
(1984)

Filtration

102104

>99.9

filt.rate=2.49.8m/hr
temp=513C
eff.NTU/inf.NTU=(0.130.16)/(1.0
2.0)

Bellamyetal.
(1984a)

4.SlowSand

505103/L

100

filt.rate=0.04
temp0.4m/hr=0C,5C,17C
eff.NTU/inf.NTU=(37)/(410)

Bellamyetal.
(1984b)

*Amirtharajah(1)
**ThesereferencesaregiveninAmirtharajah.

4.Evenconventionalplantsarenotfailsafeandwouldpasscysts,underpooroperatingconditionsand/orwhencystlevelsincreasedtonumbersintherangeof20to
hundredsper100gallonsintherawwater.
5.Rapidgravityfiltrationwithoutcoagulationinvariablypassedcystsintothefinishedwater.
6.Unfilteredsupplieswithchlorinationondnofiltrationoftenpassedcystsatlevelsof010cystsper100gallonstothefinalwater.
AnimportantconclusionfromtheresultspresentedinTable8,i.e.,fromplantscaledata,isthatitgenerallycorroboratesthedatainTable7derivedprincipallyfrom
pilotplantstudies,andsomeoftheassumedremovalsinFigure1.Someverytentativegeneralizationsonallowablecystlevelsintherawandfinishedwatersmaybe
madeonthebasisoftheavailabledata.
Awelloperatedconventionalsystem,withasamplingprogramofonesampleperweekonrawwater,shouldcommonlydetect2030percentofsamplespositive.If
thecystlevelsinthesesamplesarelessthan10per100gallonsthentheriskfromsuchasystemisverysmall.Ifthecystlevelsaregreaterthan10for100gallonsin
therawwater,animmediateanalysisforpossibleweaknessesinthetreatmenttrainshouldbeinitiated.
Mostunfilteredsystemshavefinishedwatercystlevelsof010per100gallonsinthesamplesthatwerepositiveandhigherlevelsintherawwater.Oftenonly510%
ofthefinishedwatersampleswerepositive.Theseremovalratesmaybeattributedtodisinfectionandsedimentationsincemanysamplesweredrawnfromlakes.
However,ifthepercentageofpositivesampleswasonly2.5%(i.e.,97.5%negativeasinTable8)ofthesamplestested,thesecitiesseemtohavebeenhistorically
safefromanoutbreakofgiardiasis.Ifthecystlevelsreachintothetensorhundredsper100gallons,thentheriskissignificantlyincreased.
Ifweassumethatcontaminatedrawwatersupplies,withcystconcentrationsashighas100cysts/100gallons(or100cysts/378L)wouldreceiveadequatetreatment
ifremovalremoval/disinfectionefficiencieswere99.999%(sincesystemswithcleanwatersourcesmustachieve99.9%removalitwouldbereasonabletoexpect
systemswithcontaminatedwatertoachievesignificantlygreaterremovals),thenacceptablefinishedwaterconcentrationswouldbe100105cysts/378Lor2.6
106cysts/L.Then,ifasystemwereabletoachieve99%inactivationbydisinfectionalone,anacceptablerawwaterconcentrationforsuchasystemwouldbe2.6
104cysts/L.Thisconcentrationlevelishigherthanthatdeterminedbytheriskmodelbutfallswithinthesameorderofmagnitude.Accordingtotheanalysispresented
inTable1,ifasystemachieved99%removalofGiardiacystsbytreatment(e.g.,bydisinfectionalone)itwouldproducean"acceptable"treatedwateriftheaverage
concentrationofGiardiacystsinthesourcewaterwaslessthan1.36105cysts/L.i.e:

Page281

Theempiricaldatapresentsometentativepossibilitiesforestimatingallowablecystconcentrationlevels,buttheseneedtobeconfirmedbyananalysisofallofthedata
collectedbyHibler(9)andadditionaldataastheybecomeavailable.Inaddition,theassumptionsofthemodelneedtobereassessedonthebasisoftheavailable
data.
InfectivityandViability(J.C.Hoff)
IndeterminingtheeffectsofvariousenvironmentalconditionsandexposuretodisinfectantsonGiardiacysts,itisimportanttoconsiderthemethodsusedtodetermine
whetherornotthecystshavebeenkilled.Itisimportanttonotethatthetermsinfectivityandviabilityhavedifferentmeanings.Infectivityreferstotheabilitytoproduce
aninfectioninahost,withorwithouttheproductionofdisease.Viabilityisabroadertermthatincludesinfectivitybutalsoincludesmeasurementofotherparameters
relatedtobeingalivesuchasphysiologicalactivity,respiration,reproduction,excystationanddyeexclusion.
Thelattertwomethods,invitroexcystationanddyeexclusion,havebeenthepredominantmethodsusedtodetermineGiardiacystviability.Earlystudies,usingdye
exclusion,suchasthoseofMillsetal.(17),indicatedthatcystswereextremelyresistanttoinactivationbydisinfectants.Followingthedevelopmentofinvitro
excystationmethodsforG.lamblia(2,20)andG.muris(22)andresearchindicatingthatdyeexclusiondidnotcorrelatewithexcystation(2),invitroexcystation
becamethepredominantmethodusedformeasuringcystviability.Newermethodsbasedontheuseoffluorogenicdyessuchasfluoresceindiacetateandpropidium
iodide(23)andboronicacidderivatives(13)arebeingdevelopedbutlittlequantitativetestingoftheircorrelationwithexcystationoranimalinfectivityhasbeendone.
RelativelylittleresearchontheminimumnumberofGiardiacystsrequiredtoinfectanimalshasbeendone.Inanearlystudy,Rendtorff(19)studiedtheminimum
numberofG.lambliacystsrequiredtoinfecthumans.Basedonthesestudies,heconcludedthatonecystwascapableofcausinginfectioninasusceptiblehost.
Hibler(10)determinedthat<8G.lambliacystsproducedinfectionin80to100%ofMongoliangerbilsfedcysts.UsingG.muriscysts,Hoffetal.(11)reportedthat
ID50formicerangedfrom<1to16cysts.Whilethedataarelimited,itisevidentthatveryfewcystsarerequiredtoproduceinfectioninmanorotheranimals.Inthe
samestudyHoffetal.(11)showedthatexcystationandinfectivityassaysgavesimilarresultsinindicatingthedegreeofinactivationcausedbyexposureofG.muris
cyststochlorine.
InfectivityandViability(C.P.Hibler)
InterpretationofinfectivityandviabilityofGiardiacystsmustbedonewithcaution.OurlaboratoryhasmadeanumberofindirectanddirectobservationsonGiardia
cystsfromsurfacewatersources.Examinationof4,423samplesofwaterfrom301municipalsites,hasrevealedthat102sitesoftenhadcystspresentinfinished
watersamples.Only5/102(5%)ofthesesiteshadanepidemicofwaterbornegiardiasisatthetimeofsampling.Severalofthe97sitesoftenhadasmanycysts/gallon
asthosesitesinthemidstofanepidemic.Mostofthe97siteshadadequate
TABLE8.Cystlevelsinrawandfinishedwatera

No.ofSamplesPositive
%SamplesNegative

Treatment

City
No.

Areaof
Country

Source

TotalNo.
of
Samples
Raw

Finished

010

Raw

1020

(Cystlevels#/100gal.)
>200101020

Finished

>20

Conventional

20

Lake

112

83

100

32

River/Lake

43

86.4

100

93

NE

River

139

76.6

93.5*

16

109

NE

Lake/Creek

139

70.8

100

20

NE

River

87

66*

11

10

22

10

RapidGravity

Creek

95

52.4

70.3*

17

Filtration

74

Creek

149

27.3

44.7*

20

13

27

17

Without

152

River

100

52

90

13

10

Coagulation

Unfiltered

40

SW

River

134

50.8

97.3

29

41

SW

Creek

124

70.7

92.8

89

NE

Lake

206

78.9

87*

20

90

NE

Lake

449

90.5

96.4

16

91

NE

Lake

428

89.9

97.5

120

NE

Lake/Creek

102

96.3

89.6*

152

River

142

52.0

90

13

10

aAlldatapresentedinthetablearefromHibler(7),exceptforcityMfromSykora's(22)studies.Onlydatawithbothrawandfinishedwatercyst
levelsareincludedinthetable.
*Thesecitieshavehadoutbreaksofgiardiasis.

Page282

disinfectionafewdidnot,yettherewasnoreportedincreaseofcasesabovethenormal"backgroundlevel"inthesemunicipalities.Microscopicevaluationofcysts
fromsomeofthesemunicipalitiesindicatedcystspossiblywerealiveconsequentlyon6occasions,theywereintroducedinto10Mongoliangerbils.Thenumberof
cystsintroducedrangedfrom1to10cysts/mL/gerbil.In4/6(67%)attempts,25to50%ofthegerbilsbecameinfectedwithGiardia.Ontwooccasionstheamount
ofconcentrateavailablenecessitatedanattemptwith1to2cysts/mL/gerbilandpreviousexperiencehadshownthatthisnumberofcystswouldonlyestablishinfection
inabout20%ofthegerbils.Noneoftheanimalsintrialswith1to2cystsbecameinfected.Theonlysuccessfultrialsinvolveduseof5to10cysts/mL/gerbil.
TheresultsshowedthatcystswerealiveandinfectiousforMongoliangerbilsfrom4/6municipalitiesthatdidnothaveadequatedisinfection,indicatingthatthecysts
wereinfectiousforhumans.However,therearetwofactorstobeconsidered:(1)Mongoliangerbilsaresusceptibletoonly50to60%ofthehumansourcestowhich
theyareexposed,suggestingthecystsmayhavebeeninfectiousforhumansbutnotforgerbilsand(2)thecystsmayhavebeenresponsibleforthe"background
level",especiallyinsmallmunicipalities(<400)wheremostofthepopulationwasalreadyimmune.
InfectionofMongoliangerbils,oranyanimalmodelwithcystsfromasurfacewatersource,meansonlythatthosecystwerealiveandinfectiousfortheanimalmodel
nootherinterpretationsarepossible.Ifinfectionisnotestablished,theresultssuggest:(1)cystsweredead(2)cystswerenotinfectiousfortheanimalmodelor(3)
thenumberofcystswasinadequatetoestablishaninfection.Theresultsfromuseofanimalmodelsdoesnotprovideanyusefulinformationregardingthesusceptible
humanpopulation.
Afewmunicipalitiesfromwhichwehavereceivedsampleshadenoughcystsinthefinishedwatertohavecausedamajorepidemic.Thesecystswerefoundinthefirst
sampleseverreceivedfromthesesitesundoubtedlytheyhadbeenpresentforsometime.Whilenoeffortsweremadetoinfectanimalsanddetermineifcystswere
aliveordead,microscopicevaluationindicatedtheywerealive.Despiteourlackofproof,andthecautionsand/orprecautionswemustexercisebeforeconcluding
thatthesecystswerenotinfectiousforhumans,theevidencecertainlysuggeststheywerenotinfectiousbecausethechlorinelevelsandthecontacttimebeforethefirst
customerswasadequate.
Testsforviabilityusingananimalmodelmustincorporatetheconstraintsgivenforinfectivityhumansourcesarenotuniformlyinfectiousforanimalmodels.
Microscopicevaluationrequiresconsiderableskillandexperienceandmostmicroscopistsdonotseeenoughcystsfromdifferentanimalsand/orsourcestodevelop
thisskill.Evenexperiencedmicroscopistsgenerallycannotrecognizedyingorrecentlydeadcysts,especiallyiftheyhavebeendeadonlyashortperiodoftime.The
eosinexclusiontestandexcystationarenotadequate:eosinisunreliableandtherearenotenoughcystsforexcystation.Currentlyeffortsareunderwaytodevelopan
immunofluorescencetestforviability.Likelysuchatestwillhavethesameproblemsfacedbytheexperiencedmicroscopistsdyingorrecentlydeadcystswillappear
tobealive.
InfectivityandviabilityofGiardiacystsfromsurfacewatersourceswillcontinuetobeaproblemforanalystsuntiltheentirehostrangeofGiardiafromhumanand
animalsources,hasbeenelucidatedandclarified.Itisunlikelythatmonoclonalantibodieswilldistinguishbetweenhumanandanimalsources.Therecertainlyare
"strains"presentinhumansthatwillnotinfectanimals,andthereare"strains"inanimalsthatwillnotinfecthumans.Theremaybestrainsinhumansthatwillnotinfect
otherhumans.Untilmoreinformationisavailable,wemustassumethatGiardiacystspresentinsurfacewaterpossessingmorphologicfeaturescompatiblewithG.
duodenalisarepotentiallyinfectiousforhumans.
UseofCTValuesinEstimatingGiardiaCystInactivation(J.C.Hoff)
CTvaluesrepresenttheproductofdisinfectantconcentration(C)andcontacttime(T)requiredtocauseinactivationofacertainpercentage(e.g.,99%)ofaspecific
targetpathogenbyaspecificdisinfectantunderspecifiedpHandtemperatureconditions.TheuseofCTvaluesisbasedonanalogiesbetweenchemicalreactionrates
andmicrobialinactivationrates.IndeterminingCTvaluestobeusedinestablishingdisinfectionrequirementsforsurfacewaters,thetargetpathogenisG.lamblia
cysts,becauseofthehighdegreeofresistanceshownbythecysts.Basedonthenumberslikelytobefoundinwaterandtheirlowinfectiousdose,alevelof99.9%
inactivationhasbeendeterminedtobeadequateforcontrolofwaterbornegiardiasis.
ThebasisfortheuseofCTvaluestoestablishdisinfectionrequirements,CTvaluesfordifferenttypesofpathogens,andsomeoftheproblemsassociatedwith
applicationofthisapproachtoestablishingdisinfectionrequirements,werereviewedbyHoff(12).Amajordifficultyisthelimiteddatabaseofdisinfectiondata
availableanduncertaintiesassociatedwithextrapolationsfromtheavailabledatabasetootherpHs,temperatures,anddisinfectantconcentrations.
Anotherconsideration,especiallywithregardtocystdata,isthatbecauseoftheassaymethodusedtodetermineviablevs.nonviablecysts(microscopy)and
limitationsonthenumbersofcyststhatcanbeobserved,thesedataarelimitedtodeterminingvaluesatthe99%level.Extrapolationisrequiredtoestimatetimes
neededfor99.9%inactivationandbecausethecurvesdonotfollowfirstorderkinetics,thecalculatedCTvaluesatthislevelarelessreliable.Hibler(10)has
developeddatashowingCTvaluesfor99.99%inactivationofG.lambliacystsusingMongoliangerbils.ThesedatashouldaidinestablishingtheneededCTvalues.
MeasurementofCTValuesonSite(A.Amirtharajah)
TheextrapolationoflaboratoryCTvaluestofieldconditionsrequiresacriticalassessmentofconcentrationofdisinfectant,Cinmg/Landcontacttime,Tinminutes.
Theeffectiveconcentrationisdefinedonthebasisofitsvalue,shortlybeforeorafteritreachesthefirstcustomer

Page283

orshortlybeforeaseconddisinfectantdoseisapplied.Contacttimeinpipelinescanbecalculatedassumingplugflowconditionsanddividingthelengthofthepipeline
bythemeanvelocityofflow.Thiscomputationmustbedeterminedforpeakflowconditions.
Thecontacttimeinchlorinecontacttanksisbestestimatedonthebasisoftracerstudies.Twotracersthatareconvenienttouse,presentnoadversehealtheffectsand
areconvenientforchemicalanalysesarelithiumchlorideorsodiumfluoride.Fluorideisaconvenienttracersinceitisalreadyusedinwatertreatmentplantsandmost
plantlaboratorieshaveequipmentforitsanalysis.Ininstanceswherefluorideisused,backgroundcorrectionsforfluorideinthewaterpriortodisinfectionneedtobe
made.Severalstudies(15,25)haveshownthatdesignandevaluationofchlorinecontactchambersareusefullyevaluatedonthebasisofdispersionnumbers.The
tracerstudywillyieldadimensionlessdispersionnumber,dforthechlorinecontacttank.Theidealreactorclassifiedasaplugflowreactorhasadispersionnumberof
zero.Theflowinpipelinesapproximatesareactorwithzerodispersionnumber.Theoppositeextremeofidealreactors,thecomplexmixreactor,hasadispersion
numberofinfinity.TheeffluenttoinfluentratioCi/Cioisplottedasafunctionofdimensionlesstime, inFigure2aandillustratesthecharacteristictracercurvesfor
differentdispersionnumbers.Thevalueof iscalculatedbydividingthedifferenttimesofsampling,tinthetracerstudybythevolumetricdisplacementtimetofthe
contacttank[t=(volume)/(rateofflow)=V/Q].Thetracer(lithiumchlorideorsodiumfluoride)isnormallyaddedasaslugattheinletendofthecontacttankatthe
dimensionlesstime =0asshowninFigure2a,sothatitproducesaconcentrationCio,ifdispersedinstantaneouslyinthevolumeofthetank.Theeffluentconcentration
CiatvarioustimestismeasuredandadispersioncurveasshowninFigure2bisdrawn.Thedispersioncurvethatisgeneratedisthencomparedwiththetypical
curvesshowninFigure2bandanestimateismadeofthedispersionnumber.Thedispersionnumbercanalsobecalculatedfromequationsgiveninseveralreferences
(27,25).TrussellandChao(25)suggestthatdispersionnumberscanbealternativelyestimatedfromlengthtowidth(L/W)ratiosofthechlorinecontactchamberby
usingthefollowingequation:

inwhichK'=coefficientofanonidealitywithvaluesfrom1to15,L=lengthofcontacttank,W=widthofcontacttank.AconservativeapproachwillassumeK=
15,whilechamberswithgoodhydraulicdesignandapproximatingplugflowconfigurations,haveKvaluesof3to5.
ByusingthecalculatedormeasureddispersionnumberandthecurvesinFigure2aitispossibletomakereasonableestimationsofcontacttimeTthatcanthenbe

Figure2.
Tracerresponsesincontacttankwithapulseinput(afterWeber,(27).

usedtoevaluateCTvalues.Figure2ashowsthatcontacttankswithdispersionnumberslessthan0.025areclosertoplugflowconditionsthantocompletemix
conditions.ThemidpointoftheincreasinglimbofthetracercurvecouldbeusedasareasonablysafefirstapproximationforthedetentiontimeTtobeusedinCT
calculations.Ifcontacttankshavedispersionnumbersgreaterthan0.1,thenitimpliessignificantshortcircuitinginthetankandtheeffectivenessofdisinfectionmaybe
improvedbyrearrangementoftheflowconditionswithinthetankwithadditionalbaffles,etc.
Itispossibletocombinetheplugflowdispersioncharacteristicsofacontacttankwiththeinactivationsatvariousconcentrationsandcontacttimesandhencemakea
verygoodestimateoftheoveralldisinfectionefficacyforvirusesandGiardiacysts.Thisapproachisveryvalidifdisinfectionkineticscanbemodeledbywelldefined
reactionkinetics.Forafirstorderreactionwithrateconstantk,Figure3developedbyLevenspielandBischoff(14,15)maybeusedforanalysis.Asanexample,for
akTvalueof10andadispersionnumberof0.25,Figure3showsthat(Ci)/Cio=0.0065.
Therefore,theinactivationrateoforganismsis(10.0065)100=99.65%.Additionaldetailsofthemethodologyforcompletingthesecomputationsarediscussedin
thereferencesnoted(14,15,27,25)ThesecomputationsarenotroutinelymadeandforpurposesofaconservativeapproachtoassessmentofCTvaluesand
disinfectionefficaciesasinglelowerboundvalueforcontacttimeasrecommendedmaybeused.

Page284

TABLE3.
Comparisonofrealandplugflowreactorsforfirst
orderreactions(afterLevenspielandBiscchoff,14,15).

MinimumSampling(C.P.Hibler)
Samplingeffortstodateindicatethat25to30%ofthesamplesfromopensprings,creeksandriversand10%ofthelakesandreservoirsampleswillbecontaminated
withcystsofGiardia.Likelytheactualpercentageismuchhigherbecausethetechniquesforanalysisareoftenseverelycompromisedbywaterquality.
Somesurfacewatersourcesarealwayscontaminatedwithcystsfromanimalorhumansources.Othersourcesarecontaminatedonlywhenanimalpopulationsare
moving,asaresultofsewagetreatmentplantbreakdownorotherunusualcircumstances(runoff,flood).Everymajorsourceofsurfacewaterformunicipalities
probablyiscontaminatedwithcystsofGiardia:repeatedsamplinghasusuallyprovidedpositiveresults.Thecatchphrasehereis''repeatedsampling."Threetofour
samplesfromasitegenerallyareinadequate,especiallyifthesourceofcontaminationisintermittent,orthesampleshavebeentakenduringhighturbiditysituations.Six
to12samplesfromasitegenerallyareadequate,providingthesamplesaretakenduringthelatefallandwintermonths,thetimeofyearwhenturbidityandinterfering
organisms(algae,flagellates)areataminimum.IfGiardiacystsarepresentduringthesecircumstances,wecanassumetheyarepresentduringhigherturbidity
situations.
Asurveyofthewatershedforpossiblesourcesofcontaminationshouldprovidemunicipalitieswithanassessmentoftherisk.Ifthepossibilityorprobabilityfor
sewagecontaminationispresent,orthewatershedprovidessuitablehabitatforaquaticanimals(beaver,muskratandsmallerrodents),thencontaminationofthe
sourceispredictable,ifnottoday,certainlyintheforseeablefuture.
Ifcystsarefoundinasurfacewatersupplyfromasite,repeatedsamplingwillnotchangethisresult.Municipalitiesmustassumethatcontaminationisafactandtake
appropriateactiontopreventanepidemic.Unfortunately,manymunicipalitiescontinuetosample,oftenatafreneticrate,hopingthatsubsequentsampleswillnotbe
contaminated.Whilecystsmaynotbefoundateverysampling,itdoesnotchangethefactthattheywerepresentandwillbethereagain.
Inmyexperience,municipalitiesthatdonothavepotentialproblemswithGiardiaarethosethatassumethesourceiscontaminatedandtakeappropriateactionto
preventanypossibilityofwaterbornegiardiasis.
Questions/Comments
(Regardingentirepaneldiscussion)Audience:Thereasonsforwhichyoudroppedapplicationoftheriskmodelfromtheproposedruledonotseeminsurmountable
withadditionalresearch.Thecostsforbuildingoneadditionaltreatmentplantasaresultoftheproposedrulecouldprobablypayfortheadditionalresearchrequired
tosatisfactorilyaddresstheissuesthathavebeendiscussed.Ultimately,youarenotgoingtoreallyknowifyouhaveeffectivetreatmentunlessyoucanmonitorthe
water.Regardingviruses,youcandoalotofmonitoring.Onpaper,sometreatmentplantslookgreat,butwhenyouactuallygooutandmonitorforvirusesyoufind
problems.Whenallthesetreatmentplantsarebuilt,withoutsomemodelofassessment,youarenevergoingtoreallyknowhowwelltheyareperforming.Couldyou
commentonthis?
Regli:Itwouldbegreattohaveamodelfromwhichtoevaluatetreatmentplantsonacasebycasebasis,andthentomodifyandupgradetreatmenttosatisfysome
acceptablerisklevelendpoint.Wearejustnotthereyetand,therefore,areproposingtreatmentrequirements.Webelievethatifasystemweretocomplywiththe
proposedcriteria,therewouldbeahighprobabilityofasafewaterbeingprovided.Regardingthequestiononfillinginthegapstotheriskmodelwithadditional
research,theproblemistimenothavingenough.WearemandatedbyCongresstohavetreatmentrequirementspromulgatedbyDecemberofthisyear.
Audience:Couldyoucommentontheturbiditycriteriaintheproposedrule?
Regli:Ourcurrentthinkingisthattherewillbedifferentturbiditycriteriafordifferentsituations.Forfilteredwatersystems,theturbiditylimitwoulddependonthe
filtrationtechnologyinplace.Systemsusingdirectfiltrationorconventionaltreatment,wheregoodcorrelationhasbeenshownbetweenturbidityremovalandGiardia
cystremoval,wouldberequiredtohavefilteredwaterlessthan0.5NTUinatleast95percentofthemeasurementstakeneachmonth.Forslowsandand
diatomaceousearthfiltration,wherethereisnocorrelationbetweenturbidityremovalandGiardiacystremoval,thelimitwouldbelessthan1NTUinatleast95
percentofthesamplesthelimitof1NTUistoensureahighprobabilitythatthereisnointerferencewithdisinfectionofviruses.Dr.Logsdon,wouldyouliketoadd
anythingtothis?
Logsdon:Onlytostressthepointthatitisessentialforeveryfilterwithintheplanttomeetthecriteria.
Regli:Yes,thisisastrongrecommendationthatwearemakinginguidance.Wedidnotfeel,however,thatsucharequirementwouldbeappropriateinarule
becauseofthedifficultyinenforcingsucharequirement.

Page285

Audience:Arethereanyspecialcriteriaforsystemswithlowturbiditysourcewaters?
Regli:Inapreviousdraftrule,systemswithsourcewaterturbiditiesoflessthan1NTUwererequiredtoachieveeither70percentturbidityremovalorafiltered
waterturbidityoflessthan0.1NTU.Thisisnowarecommendationinguidanceratherthanintherule.
Audience:Howwouldtheruleapplytosystemswhichpurchasetheirwaterfromanothersystem?
Regli:Thetreatmentrequirementswouldapplytothesystemwiththetreatmenttechnologyinplace.Forexample,asystemwhichpurchaseditswaterfromafiltered
watersupplywouldnotberequiredtomonitorforturbidity.Ifthefirstsystemmetthetreatmentrequirementssuchastheoverallminimumperformanceofatleast99.9
and99.99percentremovaland/orinactivationofGiardiacystsandviruses,thesystempurchasingitswaterfromthatsystemwouldnotbesubjecttoalsomeetingthis
requirement,i.e.,providingfiltrationanddisinfectionasnecessary.Thepurchasedwatersystemwouldberequiredtomaintainadisinfectantresidualinthedistribution
system.
Audience:Dr.Borup,istheriskmodelsensitivetothenumberofserviceconnectionsorpeopleserved?
Borup:No,themodelisbasedonfinishedwaterqualitymeetingsomeacceptablerisklevel.Thenumberofsamplesrequiredtodemonstratethatarisklevelwasnot
beingexceededwouldnotdependonsystemsize.
Audience:Theriskmodeldoesnotseemappropriateinthatitonlyaddressesriskrelativetothefirstcustomer.Customersfurtheroutinthedistributionsystemwould
havelongerdisinfectantcontacttimesandbesubjecttolessriskthanatthefirstcustomer.Also,itseemsthatifthemodelwereaccurate,basedontheassumptions
made,indefiningwhenasystemwouldbesubjecttosignificantrisk,wewouldseemorepeoplegettingsick.Themodelseemstobeveryconservativeinestimating
risk.
Borup:Muchmoreresearchisneededtodeterminetheappropriatenessofthedifferentassumptionsthatareneededforamodeltobeapplicableinarule.The
purposeoftheanalysispresentedwastodeterminethefeasibilityofdemonstratingthroughmonitoringwhetheradequateprotectionwasbeingprovidedtoavoida
specificlevelofrisk.
Atthispointthereisnotenoughscientificdatatosupportmostoftheassumptionsnecessaryforapplyingthemodelinaregulatorycontext.
Audience:CouldyouelaborateonhowCTwouldbecalculatedinanunfilteredsystemfordeterminingthepercentinactivation?
Regli:CandTwouldbedeterminedduringpeakhourlyflow,priortothefirstcustomer.TheproductofCandT,CT,woulddependonpHandwatertemperature.
CTtableswouldbeprovidedintherule.ThesystemwouldcomparethecalculatedCTvaluewiththoseintheruletoseeifthepercentinactivationwasbeingmet.
LiteratureCited
1.Amirtharajah,A.1986.Varianceanalysesandcriteriafortreatmentregulations.JournaloftheAmer.WaterWorksAssoc.,78:3449.
2.Bingham,A.K.,Jarroll,E.L.,Jr.,andE.A.Meyer.1979.Giardiasp:Physicalfactorsofexcystationinvitro,andexcystationvs.eosinexclusionasdeterminants
ofviability.Exp.Parasitol.,47:284.
3.Bingham,A.K.,andE.A.Meyer.1979.Giardiaexcystationcanbeinducedinvitroinacidicsolutions.Nature(Lond),277:301.
4.Craun,G.F.1984.Waterborneoutbreaksofgiardiasis:Currentstatus.In:GiardiaandGiardiasis.S.L.ErlandsenandE.A.Meyer(eds).PlenumPress,New
York,NewYork,pp.243261.
5.Craun,G.F.1987.HealthaspectsofSurfacewatersupplies.PresentedatPreconferenceSeminar,PacificNorthwestSection,AmericanWaterWorks
Association,Bellevue,WA,May6,1987.
6.Borup,M.B.1986.Thedeterminationofwaterbornepathogensamplingrequirementsusingstatisticalqualitycontroltechniques.ReportpreparedforOfficeof
DrinkingWater,U.S.EPA.
7.Gerba,C.P.1984.Strategiesforthecontrolofvirusesindrinkingwater.UnpublisheddocumentpreparedundertenureofAAAS/EPAEnvironmentalScienceand
EngineeringFellowship.
8.HealthandWelfareCanada.1983.GuidelinesforCanadianRecreationalWaterQuality.SupplyandServicesCanada,Ottawa.
9.Hibler,C.P.1987a.AnalysisofmunicipalwatersamplesforcystsofGiardia.ReportsubmittedtoOfficeofDrinkingWater,U.S.EnvironmentalProtection
Agency.January.
10.Hibler,C.P.,Hancock,C.M.,Perger,L.M.,Wegrzyn,J.G.,andK.D.Swabby.1987b.InactivationofGiardiacystswithchlorineat0.5Cto5.0C.American
WaterWorksAssociationResearchFoundationReport.
11.Hoff,J.C.,Rice,E.W.,andF.W.SchaeferIII.1985.ComparisonofanimalinfectivityandexcystationasmeasuresofGiardiamuriscystinactivationby
chlorine.Appl.Env.Microbiol.50:1115.
12.Hoff,J.C.1986.Inactivationofmicrobialagentsbychemicaldisinfectants.EPA/600/286/067.
13.Hudson,S.J.,Sauch,J.F.,andLindmark,D.G.1988.FluorescentdyeexclusionasamethodfordeterminingGiardiacystviability.AdvancesinGiardia
Research.P.M.WallisandB.R.Hammond(eds),pp.265269,UniversityofCalgaryPress,Calgary,AB.
14.Levenspiel,O.andK.B.Bischoff.1959.Backmixinginthedesignofchemicalreactors.Ind.Engr.Chem.,51:1431.
15.Levenspiel,O.andK.B.Bischoff.1961.Reactionrateconstantmaymodifytheeffectsofbackmixing.Ind.Engr.Chem.,53:313.
16.Marske,D.M.andJ.D.Boyle.1973.ChlorinecontactchamberdesignAfieldevaluation.WaterandSewageWorks,120:7077.
17.Mills,R.G.,Barlett,C.L.,andJ.F.Kessel.1925.Thepenetrationoffruitsandvegetablesbybacteriaandotherparticulatematterandtheresistanceofbacteria,
protozoancystsandhelminthovatocommondisinfectionmethods.Am.J.Hyg.,5:559.
18.Regli,S.,Amirtharajah,A.,Hoff,J.,andP.Berger.1986.Treatmentforcontrolofwaterbornepathogens:Howsafeissafeenough?Proc.ofthe3rdConference
inChemicalDisinfection,Binghamton,NY,April,1986.

Page286

19.Rendtorff,R.C.1954.Theexperimentaltransmissionofhumanintestinalprotozoanparasites.II.Giardialambliacystsgivenincapsules.Am.J.Hyg.59:209.
20.Rice,E.A.,andF.W.Schaefer,III.1981.ImprovedinvitroexcystationprocedureforGiardialambliacysts.J.Clin.Microbiol.14:709.
21.RoyalSocietyStudyGroup.1983.Riskassessment.RoyalSociety,London.
22.Schaefer,F.W.III,Rice,E.W.,andJ.C.Hoff.1984.FactorspromotinginvitroexcystationofGiardiamuriscysts.Trans.RoyalSoc.Trop.Med.andHyg.
78:795.
23.Schupp,D.G.,andS.L.Erlandsen.1987.AnewmethodtodetermineGiardiacystviability:correlationoffluoresceindiacetateandpropidiumiodidestaining
withanimalinfectivity.Appl.Env.Microbl.53:704.
24.Sykora,J.L.,etal.1988.Monitoringasatoolinwaterbornegiardiasisprevention.AdvancesinGiardiaResearch.P.M.WallisandB.R.Hammond(eds),pp.
103106,UniversityofCalgaryPress,Calgary,AB.
25.Trussell,R.R.,andChao,J.L.1977.Rationaldesignofchlorinecontactfacilities.Jour.WaterPoll.ControlFed.49:659667.
26.U.S.EnvironmentalProtectionAgency.1986.Ambientwaterqualitycriteriaforbacteria1986.EPA440/584002.
27.Weber,W.J.,Jr.1972.Physicochemicalprocessesforwaterqualitycontrol.WileyInterscience,NewYork,NY.pp.4955.
28.WorldHealthOrganization.1985.Riskmanagementinchemicalsafety.EuropeanRegionalProgramonChemicalSafety.Geneva.ICPCEH506MOI56881.

Page287

PanelDiscussiononTaxonomyoftheGenusGiardia.
StanleyL.Erlandsen*,E.A.Meyer,T.E.Nash.
DepartmentofCellBiologyandNeuroanatomy,UniversityofMinnesota,Minneapolis,Minnesota55455,U.S.A..
Nomenclature
TheintestinalprotozoanGiardiawasfirstseenbyLeeuwenhoekin1681whenheexaminedhisowndiarrheicstools.However,asreviewedbyErlandsenetal(10),
thefirstmorphologicaldescriptionandillustrationoftheorganismwasperformedbyLamblin185960.Laterin1881,adetaileddescriptionofGiardiatrophozoites
andcystswaspresentedbyGrassi.
ThetaxonomicpositionoftheflagellateGiardiahasbeenrecentlyreviewed(14,15)andwasclassifiedasfollows:
PhylumSarcomastigophoraHonigberg&Balamuth1963.
ClassZoomastigophoreaCalkins,1909
OrderDiplomonadidaWenyon,1926emend.Brugerolle,1975
GenusGiardiaKunstler,1882
Thebasisforusingvariousspecificnames(ie.lamblia,intestinalis,duodenalis,enterica)hasbeendiscussedindetailbyFilice(12),Kulda&Nohynkova(14),
Levine(15),andMeyer(17).ThemostcommonlyusedspecificnamesforGiardiafrommanincludeGiardialambliaStiles,1915GiardiaintestinalisLambl,
1859andGiardiaduodenalisFilice,1952.Inspiteofalloftheexistingargumentsregardingwhichspecificnameismostappropriate,noagreeablesolutionappears
tobeinsight.However,thedevelopmentofmoresophisticatedtechnologyandadvancesinbiochemicalmethodology,andtheirapplicationtothisproblemmay
eventuallyprovideananswertothequestionofspeciesdesignation.
CriteriaforSpeciation
Threemajorcriteria(hostspecificity,morphology,andchemotaxonomy)havebeenusedforthedesignationofspeciesinthegenusGiardia.Noneofthesecriteria
hasprovidedanirrefutablesolutiontotheproblemofspeciation,buteachhascontributedinterestinginformation,whichcollectivelymayeventuallyleadtoabetter
understandingofthisorganismandasolutiontoitsspeciesdesignation.
HostSpecificity
EarlyinvestigatorsheldarigidviewthateachhostwasinfectedbyasinglespeciesofGiardia,andasdiscussedbyKulda&Nohynkova(14)andMeyer(17),this
ledtothedescriptionofmorethan40speciesofGiardiawithveryfewdetectablemorphologicaldifferences.Thedevelopmentofgerbilandmousemodelsfor
giardiasis(2,20)hasprovideddirectevidenceforthecrossspeciestransmissionofGiardiabetweenanimalhosts.Thishasledtoamoreliberalviewthatonlyone
speciesofGiardiamayexistandthatgiardiasismaybeviewedasazoonoticdisease(11)whileothersbelievedthateventhoughcrossspeciestransmissioncould
occur,multiplespeciesofGiardiaexistedinmammalsandmayrepresenttheendogenousstrainsofGiardiadetectedincertainmammalsandbirds(3,9).
Morphology
ThemedianbodyclassificationofFilice(12)hasbeen,perhaps,themostwidelyacceptedmorphologicalcharacteristicthathasbeenacceptedforuseinthe
determinationofGiardiaspecies.Bylightmicroscopy,theshapeofthemedianbody,amicrotubularcontainingcytoplasmicorganelle,maybeusedtodistinquish
threedifferentmorphologicalgroupsoftrophozoites(acceptedbysomeasspecies):G.agiliswithmedianbodiesthatresembledlongclubshapedrodsG.muris
whichhavesmallroundedpairofmedianbodiesseeninthemidlineandG.duodenalisshowntopossessapair(thatsometimeshaveappearedtobesingle)of
medianbodiesresemblingaclawhammerandwhichlietransverselyacrossthebody.Althoughthemedianbodyclassificationhasbeenofvalueindistinguishingmajor
groupsofGiardia,particularilyrodentstrainsfromthoseinfectingman,thissystemhasproventobeinadequatesincemultiplespeciesofGiardiahavebeenidentified
withinthismorphologicalgroup.Forexample,G.psittaciandG.ondatraehavebothbeenshowntopossessaclawhammermedianbodymorphology,yetthey
havedistinctmorphologicalcharacteristicsthathaveseparatedthemfromotherknowntypesofGiardia.EvenGiardiatrophozoitesisolatedfrombeaverintestines
havebeensuggestedasnotbeingidenticaltothosefoundinman(9).Morphometricanalysisofthemajordimensionsoftrophozoiteshasbeenproposedasameans
ofdifferentiatingGiardiaspp.[seereviewbyWoo(24)Bertrametal(5)],however,itsapplicationwasquestionedbyFilice(12)andthebiologicalvariationin
trophozoitesizesencounteredinculturesraisedquestionsabouttheusefulnessofthismethod.
Chemotaxonomy
Anumberofbiochemicalandmolecularbiologicalmethods,includingthedetectionofisoenzymes,immunologicalanalysisofsurfaceantigens,andcharacterizationof
chromosomesandDNA,havejustbeguntobeutilizedfortheinvestigationanddeterminationofGiardiaspp.Ithasbeenanticipatedthatthesenewapproaches,
togetherwithinformationobtainedfromcrossspeciestransmissionandmorphologicalstudies,mayprovidesomeanswerstohelpclarifytheexistingcontroversiesof
determiningGiardiaspp.However,asnotedbyBorehametal.(6)eventhisevidencemayneedtobeviewedwithsomecautionsincethesemethodshaveusually
requiredlargenumbersofcellsthatcurrentlymay
*Correspondingauthor

Page288

onlybeobtainedfromcultures.Notallstrains(species)ofGiardiacanbegrownincultureandevidencehasbeenpresentedthatcultureconditionsmightbeselective
forspecificgenotypes(1).InvestigationofisoenzymesinGiardiahavebeencarriedoutbyBertram&Meyer(5),Kormanetal.(13),andMelonietal(16).These
resultshaveshownthatculturedGiardiaisolatescanbegroupedintomultiplezymodemes,butmoreimportantly,havedemonstratedthatisolatesfromhumanscan
haveconsiderableheterogeneity.TheseresultshavebeensuggestedtosupporttheviewthatgeneticallydifferentstrainsofG.lambliacanoccurinthehumanhost
(13).However,othershaveinterpretedthefindingsofsimilaritybetweenisolatesfromdifferentanimalsasevidencethathumanscouldcontractGiardiainfectionsbya
zoonoticmechanism(22).ImmunologicalinvestigationsoftheantigeniccompositionofGiardiaspp.havedetecteddifferencesinsurfacelabelingandantigenicprofiles
betweenGiardiastrainsobtainedfromdifferentcountries(7,18,19,21)whileothershavenotedsimilaritiesbetweenisolatesfromthesamegeographicallocale
(23,16).RestrictionendonucleaseanalysisofDNAfrom15differentGiardiaisolatesobtainedfromhumanandanimalsourcesalsoshowedconsiderable
heterogeneityinbandingpatterns,withsixisolates(twoanimalandfourhuman)formingonegroup,whilethepatternforanothergroupconsistedoftwohumanand
onebeaver(19).Interestingly,morphologicalinvestigationofthesetwobeaverstrainsofGiardiasuggestedthatoneofthem,inadditiontosharingthehumanDNA
pattern,wasmorphologicallysimilartohumanorganisms,whiletheotherstrainwhichhadaDNAbandingpatternunlikethatofhumanGiardia,resembledthe
phenotypeofGiardiaisolatedfromlivetrappedbeavers(9).Thiscorrelationbetweentwodifferentmethodologies,providesaninterestingapproachforinvestigating
theproblemofGiardiaspp.,whichneedsfurtherconfirmation,butifcorrect,itshouldprovidetheimpetusforusingcombinedapproachesforanalysingwhathas
proventobeacomplexproblem.
Insummary,thecriteriaforGiardiaspeciesproposedbyFilice(12)wouldseemtobeacceptableforidentificationofG.agilisandG.muris.ThetermG.
duodenaliswouldnot,however,representasinglespecies,butratheragroup,sincetheoriginaldefinitionisbasedsolelyonamorphologicalentity(medianbody
shape),andseveraldistinctspecies(G.psittaci,G.ondatrae)sharethissamecharacteristic.ThesubstitutionofthetermG.intestinalisforG.lambliadoesnot
seemappropriateatthistime,especiallysincethereissomuchconfusioninregardtowhetheroneormorespecies(strain)ofGiardiacaninfectman.Tominimizethe
confusion,itmightbemorehelpfultoconfirmwhetherornotanisolatedGiardiaorganismfitswithintheG.duodenalisclassification,andthenprovideasmuch
informationaspossibleaboutitsanimalhost,geographicallocation,andhowitcomparestootherknownstrains.Untildefinitiveresultscandemonstratetheexistence
ofdistinctspecies,itmightalsomightbeprudenttoviewbiochemicalormorphologicalsimilaritiesforindividualcharacteristicswithsomeskepticism,andnotassume
thatitmeansalackofhostspecificity.Thedevelopmentofbetterbiochemicalormolecularbiologicalapproachesthatcanbeappliedtosmallernumbersoforganisms
inadirectisolatefrominfectedanimals,soastoavoidthepotentialproblemofchanginggenotypesinculture,mayeventuallyprovidevaluableinformationthatwill
assistinthedeterminationofspeciesintheG.duodenalisgroup.
LiteratureCited
1.Aggarwal,A.andT.E.Nash.1988.AntigenicvariationofGiardialambliainvivo.Infect.Immun.56:14201423.
2.Belosevic,M.,Faubert,G.M.,MacLean,J.D.,Law,C.,andN.A.Croll.1983.GiardialambliainfectionsinMongoliangerbils:ananimalmodel.J.Infect.Dis.
147:222226.
3.Bemrick,W.J.andS.L.Erlandsen.1988.Giardiasisisitreallyazoonosis?Parasit.Today4:6971.
4.Bertram,M.A.,Meyer,E.A.,Lile,J.andS.A.Morse.1983.AcomparisonofisozymesoffiveaxenicGiardiaisolates.J.Parasit.69:793801.
5.Bertram,M.A.,Meyer,E.A.,Anderson,D.L.andC.T.Jones.1984.AmorphometriccomparisonoffiveaxenicGiardiaisolates.J.Parasit.70:530535.
6.Boreham,P.F.L.,Upcroft,J.,Upcroft,P.,andR.Andrews.1988.Importanceofgiardiasis.Parasit.Today,inpress.
7.Einfield,D.A.andH.H.Stibbs.1984.IdentificationandcharacterisationofamajorsurfaceantigenofGiardialamblia.Infect.Immun.46:377383.
8.Erlandsen,S.L.andW.J.Bemrick.1987.SEMevidenceforanewspecies,Giardiapsittaci.J.Parasitol.73:623629.
9.Erlandsen,S.L.andW.J.Bemrick.1988.Waterbornegiardiasis:sourcesofGiardiacystsandevidencepertainingtotheirimplicationinhumaninfection.In:
AdvancesinGiardiaResearch.P.M.WallisandB.R.Hammond(eds).pp.237246,UniversityofCalgaryPress,Calgary,Canada.
10.Erlandsen,S.L.andD.G.Feely.1984.Trophozoitemotilityandthemechanismofattachment.In:GiardiaandGiardiasis:Biology,Pathogenesis,and
Epidemiology.S.L.ErlandsenandE.A.Meyer.(eds).PlenumPress,NewYork,NY,pp.3363.
11.Faubert,G.M.1988.Evidencethatgiardiasisisazoonosis.Parasit.Today4:6668.
12.Filice,F.P.1952.StudiesonthecytologyandlifehistoryofaGiardiafromthelaboratoryrat.Univ.Calif.Publ.Zool.57:53143.
13.Korman,S.H.,LeBlancq,S.M.,Spira,D.T.,ElOn,J.,Reifen,R.M.andR.J.Deckelbaum.1986.Giardialamblia:identificationofdifferentstrainsfromman.
Z.Parasitenkd.72:173180.
14.Kulda,J.andE.Nohynkova.1978.Flagellatesofthehumanintestineandofintestinesofotherspecies.In:ParasiticProtozoa.J.P.Kreier(ed).Academic
Press,NewYork,NY,pp.69104.
15.Levine,N.D.1978.Giardialamblia:classification,structure,identification.In:WaterborneTransmissionofGiardiasis.W.JakubowskiandJ.C.Hoff(eds).
U.S.EnvironmentalProtectionAgency,Cincinnati,OH,pp.28.
16.Meloni,B.P.,Lymbery,A.J.andR.C.Thompson.1988.Isoenzymeelectrophoresisof30isolatesofGiardiafromhumansandfelines.Am.J.Trop.Med.Hyg.
38:6573.
17.Meyer,E.A.1985.Theepidemiologyofgiardiasis.Parasit.Today1:101105.
18.Nash,T.E.andD.B.Keister.1985.Differencesinexcretorysecretoryproductsandsurfaceantigensamong19isolatesGiardia.J.Infect.Dis.152:11661171.
19.Nash,T.E.,McCutchan,T.,Keister,D.,Dame,J.B.,Conrad,J.D.,andF.D.Gillin.1985.RestrictionendonucleaseanalysisofDNAfrom15Giardiaisolates
obtainedfromhumansandanimals.J.Infect.Dis.152:6473.
20.RobertsThomson,I.C.,Stevens,D.P.,Mahmoud,A.A.F.

Page289

andK.S.Warren.1976.Giardiasisinthemouse:ananimalmodel.Gastroenterology71:5761.
21.Smith,P.D.,Gillin,F.D.,Kaushavl,N.A.andT.E.Nash.1982.AntigenicanalysisofGiardialambliafromAfghanistan,PuertoRico,Equador,andOregon.
Infect.Immun.36:714719.
22.Thompson,R.C.A.,Lymbery,A.J.andB.P.Meloni.1988.GiardiasisazoonosisinAustralia?Parasit.Today4:201.
23.Wenman,W.M.,Meuser,R.U.andP.M.Wallis.1986.AntigenicanalysisofGiardiaduodenalisstrainsisolatedinAlberta.Can.J.Microbiol.32:926929.
24.Woo,P.T.K.1984.EvidenceforanimalreservoirsandtransmissionofGiardiainfectionbetweenanimalspecies.In:GiardiaandGiardiasis:Biology,
Pathogenesis,andEpidemiology.S.L.ErlandsenandE.A.Meyer(eds).PlenumPress,NewYork,NY.pp.341364.

Page291

PanelDiscussiononMethodsofHandlingGiardiaintheLaboratory
WalterJakubowski*,ErnestA.Meyer,TheodoreE.Nash,CharlesP.Hibler
U.S.EnvironmentalProtectionAgency,Cincinnati,Ohio45268,U.S.A.
ObjectivesandFormat
Theobjectiveofthispanelwastodiscussparticularlytroublesomeareasinworkingwithtrophozoitesandcystsinthelaboratory.Theintentwastoidentifypossible
solutionsthroughasharingofexperienceswhileunderscoringsignificantproblemsrequiringfurtherresearch.Theformatwasinformalandthesessionwasroughly
dividedintotwosectionsconcerningproblemswithculturingandmaintainingtrophozoitesandproblemswithsourcesandqualityofcysts.Eachpanelmemberbriefly
introducedatopicandquestionsandcommentsweretakenfromthefloor.Thesummariesanddiscussionbelowweredevelopedbythechairmanfromwritten
materialsubmittedbythepanelmembersandfromnotesandtapesofthesession.Thechairmanacceptsfullresponsibilityfortheaccuracyoftheircontent.
Summaries
TheodoreE.Nash
MethodsfortheaxenizationofGiardialambliawerereviewed.Trophozoitesusedforaxenizationareobtainedinthreeways.Themostdirectwayisbysamplingthe
smallintestinalcontentseitherbyduodenalintubationorbyuseofthestringtest.Immediatelyafterverifyingthepresenceoftheparasite,thesamplesareinoculated
intoTYIS33mediumcontainingantibioticsandviableGiardiaareallowedtoadheretotheglasstubesat37C.Themediumisthendecantedandreplacedwith
freshmedium.Inthisway,theintestinalcontentsandtheoldmedium,whicharetoxictotheorganism,areseparatedfromtheadheredGiardia.Theothertwo
methodsderivetrophozoitesfromcystspurifiedfromfeces.Stoolsfrompatientsexcretinglargenumbersofcystsareusedinordertoincreasethenumbersofcysts
obtainedsothatseveralattemptsandprocedurescanbeemployed.Thereareanumberofmethodsusedforpurifyingcystsfromstoolsbutweusesucrosegradients.
Perhapsthemostsuccessfulmethodforobtainingtrophozoitesfromcystsinourlaboratoryistoinject50,000cystsin0.1mLintothestomachof1to2dayold
sucklingmice(KeisterandMattern,unpublishedinformation).Micearesacrificedatspacedintervalsbeginning7daysafterinoculation.Portionsoftheintestinesare
examinedandplacedintotubescontainingmediumandantibioticsasabove.Intestinalcontentsandcontaminatedmediumarequicklyremovedfromadhered
trophozoites.
Inthethirdmethod,trophozoitesareobtainedafterinvitroexcystment.Atleast100,000cystsin100Lofwaterareaddedto1mLof1%pepsinin0.85%NaCl,
pH2.0.Tubesareslowlymixedat37Cforatleast30minutes.Differencesinnumbersoftrophozoitesobtainedafterincubationfrom30to90minutesarenotfound,
sowewilloftenuselongerincubationtimestokillsomeoftheaccompanyingbacteria.Thepelletedcystsarethenaddedtotubescontainingmediumandantibiotics.
Trophozoitesarenotimmediatelyfoundbutappearafter6to24hours.
MostbacterialantibioticshavenoeffectoncultivationofGiardiasothatalargeassortmentcanbeused.Incontaminatedspecimens,someorallofthefollowing
antibioticsareused:penicillin(200g/mL),gentamycin(40g/mL),ticarcillin(1000g/mL),clindamycin(50g/mL),moxalactam(1000g/mL)andamphotericin(1
g/mL).Problemorganismstendtoberesistantgramnegativerods,mostlyPseudomonasspecies,yeastandanoccasionalanaerobe.
Noneofthesemethodsworkallofthetime.However,themouseinoculationmethodyieldsenormousnumbersoftrophozoiteswhichincreasesthelikelihoodof
axenization.SuccessfulaxenizationofGiardiaisobtainedinourlaboratoryfromabout70%ofstoolsfromdifferentpatients.
ErnestA.Meyer
AfterpresentinghistoricalinformationontheestablishmentofaxenicculturesofGiardia,severalproblemsrelatingtohandlingtheorganisminthelaboratorywere
discussed.Problemareasmeritingadditionalattentionincludedtheneedfor:
1.Morereliablemeansforestablishingorganismsinculture.Successfuladaptationtoculturein10to50%ofattemptswouldseemtoreflecttheaudienceexperience.
Thenecessitytoincludeintheculturemediumsuchundefinedandvariablecomponentsastrypticase,yeastextract,phytoneandserummostlikelyisresponsiblefor
someofourfailures.ThedevelopmentofaculturemediumthatwouldmorereliablyyieldGiardiaculturesfromasmallinoculumcouldbewelcome.Effortsto
developsuchanimprovedmediummightwellstartbysystematicallyvaryingexistingGiardiaculturemedia.
2.Morestrainsestablishedinculture,andpromptlypreservedbyfreezing,inordertominimizethepossibilityofchangesoccurringwithcontinuedcultivation.Our
abilitytodevelopasatisfactorysystemofclassifyingorganismsinthisgenusdependsonourabilitytocharacterizeGiardia.Clearly,themoreisolatesthatare
availableforstudy,thebetterweareabletounderstandtheirhostrange,pathogenicityandchemotaxonomy,andtoclassifythem.
*Correspondingauthor.

Page292

3.DevelopmentofmethodologyforthecultureofG.muristypeorganisms.ItwaspointedoutthattheRobertsThomsonmousemodelisnowusedworldwide,and
hasyieldedagreatdealofinformationrelativetoGiardiabehaviorinthemouse.Atthesametime,wearenotyetabletocultureandstudythisoranyotherG.muris
straininvitro.
4.Acentralrepositorywhereallavailableculturescanbekept,andfromwhichtheycanbesuppliedtointerestedinvestigatorsasneeded.Thefindingsofanyworker
canonlybegenerallyacceptedwhentheycanberepeated,andtheycanonlyberepeatedwhentheparticularGiardiaorganismsusedinanexperimentareavailable
toothers.
5.Animprovednomenclature.Threeaspectsofthisproblemwerediscussed.Firstiswhathasbecomethe"traditional"speciesproblem,andwhetherspeciesshould
bebasedonthosefewknownmorphologicaldifferencesthatcanbeobservedincludingmedianbodymorphology,presenceorabsenceinthetrophozoiteofaventral
flange,theexistenceof"binary"cysts,andlengthofcaudalflagella,oronthebasisoftheanimalhost.AsecondaspectoftheperceivedproblemofGiardia
nomenclaturehasbeenthetendencyofvariousauthorstousedifferentnamesforwhatmaywellbethesameorganism.Thus,itispossibletofindGiardiafrom
humansasignedbyvariousauthorsthespeciesnamesoflamblia,duodenalisandintestinalis.Athirdemergingaspectoftheproblemhasbeentheabsenceofa
uniformmethodofnamingGiardiacultureisolates.Itwouldbemosthelpfulifaninterested,knowledgeablegroupcouldstudytheproblemandmake
recommendationsregardinga)thenomenclatureofthisgroup,andb)arecommendedstandardmethodofidentifyingGiardiaisolates.Thenamesbywhichthese
organismsareidentifiedisoffarlessimportancethanthefactthatauniform,agreeduponsystembeemployedthateveryonecanuse.
CharlesP.Hibler
Proceduresandproblemsinobtainingandpurifyingcystswerediscussed.Cystsmaybeobtainedfromhumanoranimalsources.HumansourceGiardiacysts
generallyareobtainedfrommedicalhospitals,medicalcarecentersorstudenthealthcentersonuniversitycampuses.Onlargeuniversitycampuses,studentsfrequently
areagoodsource.AnumberofproblemsfaceinvestigatorsplanningtousehumansourceGiardia.Onepotentialproblemisthatthesourceisinfectedwith
pathogensotherthanGiardia,causingarisktolaboratorypersonnelprocessingthesample.Amajorproblemforinvestigatorsisavailabilityofasamplewhen
needed.Theageandsexofthepatient,thepossiblesourceoftheinfection,thedurationofthesymptoms,andthetreatmenthistorygenerallyareunknown.
Frequently,thesamplesareinadequate(quantityofcystsand/orstool),andrangefromafewhourstoseveraldaysold.Samplesfromsomehumansources(even
whenfresh)oftencontainmoreobviouslydeadthanlivecysts,indicatingeithertreatmentand/oranimmuneresponse.
Mostinvestigatorsprefertoworkwithhumanratherthananimalsourcecysts.However,iftheGiardiaduodenalistypecystisallthatisrequired,agoodsourceis
humaneshelters.Generally,80to90%ofthedogshousedinthesesituationsareinfectedandhighpercentage(20to30%)aresheddingcystsatanygivensampling
period.OtheranimalsourcesofG.duodenalisaremoredifficulttoobtain.Veterinaryhospitals,clinicsorstatediagnosticlaboratoriesthatroutinelyexamineanumber
oflargeanimals(cattle,sheep,horses,etc.)areagoodsource.Otheranimals,suchasbeaver,muskrat,moose,etc.,necessitateacertainamountoffieldexperience
withwildlifetobesuccessful.AswiththehumansourceGiardia,theseanimals(exceptdogsinahumaneshelter)arerarelyifeveravailablewhenneededandmany
ofthesameproblemsapply(durationofinfection,historyoftreatment,etc.).
ThoseinvestigatorsusingG.murisasamodelforG.duodenalismustexerciseconsiderablecautionbecauseitisadifferentspeciesandinformationobtainedmaynot
beapplicabletoG.duodenalis.If,forexample,inactivationbychlorine,ozoneorotherdisinfectantsisdifferentforG.murisandG.duodenalis,thendatagenerated
usingG.murismaynotbeacceptablewheninactivationdataforG.duodenalis(=G.lamblia)arerequiredinthedevelopmentofdrinkingwaterregulations.
Severalanimals,fromneonatalratstoneonatalmiceandotherlaboratoryrodents,havebeenmoreorlessevaluatedfortheirsusceptibilitytohumansourcecystsof
G.duodenalisandfortheirsuitablityasanimalmodels.TheMongoliangerbil,Merionesunguiculatus,appearstobethebestmodelforhumanaswellasother
animalsources.Certainly,thesegerbilspossessconsiderabledesirableattributes:theyaregentle,odorless,andwithproperhusbandry,areprolific.However,mostof
ushavenotmadethesameeffortwithotherlaboratoryanimalsthatwehaveputintoMongoliangerbils.DevelopingtheMongoliangerbilasamodelisnotaneasy
task.TheinitialcolonymustbetreatedtoeliminatepossibleGiardiainfection,andthecommensalsTrichomonasandEndamoeba.Thisbreedingcolonymustthen
bemaintainedunderstrictgovernmentalregulationsregardingcareandmaintenanceoflaboratoryanimals.Thisinvolvesconsiderabletime,expenseandfacilities
generallyavailableonlyatlargeruniversitiesorgovernmentallaboratories.
Separationofcystsfromthestoolsofhumans,dogsandotheranimalsourcesthatareomnivorousorcarnivorous,andproducefineparticlewaste,isverydifficult,
oftennecessitatingconsiderabledilutionofthesampleandpassagethroughgauzeand/orscreensseveraltimes.Separationofcystsfromstoolsofmostherbivorous
animals(exceptcattle)isnotasdifficult.Recoveryofcystsismuchmoreeffectivebecausethewasteconsistsoflargerparticles.Iflaboratoryrodents,suchasthe
Mongoliangerbil,arethesource,fecalpelletsmustbecollectedinwaterbecausetheremovalofexcesswaterintherectumofrodentswillresultindehydrationofthe
cystsshortlyafterthefecalpelletisexpelled.Thesepelletscanbemaceratedand,withwashing,thecystspassedthrough80to100meshscreens.Thisprocedurecan
beusedfor

Page293

stoolsfromhumansandanimalsandwillprovidea"semiclean"suspensionthatisreadyforfurthercleaningorrefrigeratorstorage.Cystsfromhumanordogsources
generallywillnotremainviablemorethanafewdays,evenwithantibiotics(Penicillinandstreptomycin)addedhowever,laboratoryorother(wild)rodentsource
cystswillremainviableforaconsiderableperiodprovidinginvestigatorseitheraddantibioticsorreplacethewatertothesedimentdailyandagitate.Dailywater
replacementmaybeaseffectiveasantibiotics.Ifthesuspensioniscarefullycleaned,thewaterreplaceddaily,and/orantibioticsadded,someofthecystswillremain
viablefor6weekstotwomonths.
Thefinalpurificationstepgenerallyinvolvesoverlay/underlayofhighlydilutesuspensionswithchemicalssuchassucrose,zincsulfate,percoll,ficollhypaque,etc.The
lowerthespecificgravity(about1.07to1.1)ofthechemicalflotationsolution,thebetter.Cystlossesaregreater,buttheeffectofthechemicalsislessdetrimental.
Recoveryofviablecystsisbetterwithpercollthanwithanyotherchemical,butpercolliscostprohibitiveformanylaboratories.Aftertheunderlay/overlaystep,the
suspensioniscentrifugedfor3to8minutesandthematerialtotheinterface,orallofthematerialtothepellet,issiphonedwithslightvacuumthrougha3or5m
nucleporetypemembrane.Themembraneisthenrinsedintoabeakerwithastrongjetofdistilledwater,andthecystsarerefrigerated.Aswiththesemiclean
preparation,dailyreplacementofwaterwillhelplongevityofthecysts.However,mostifnotallchemicalsaredetrimentaltocystsurvivaltherefore,ifcritical
experimentsareplanned,cystsshouldbeusedimmediately.OurexperienceswithGiardiacystsfromvarioushumanandanimalsourcesindicatesthata"maturation
period"ofseveraldaystoaweekisactuallydetrimental.Fewerwillexcystandmorearerequiredtoinfectanimalsthanthosecollectedandusedwithintwoorthree
days.
WalterJakubowski
Theneedforstandardization,bothindefinitionsandinproceduresforhandlingGiardia,wasdiscussed.Healthandregulatoryagenciesrequirequantitative
informationontheoccurrence,distribution,survivalandinactivationofthisorganism,andontheefficacyofwatertreatmentprocessesforitsremoval.Theinformation
maybeusedinassessingrisksandinsettingstandards,andfordeterminingtheappropriatenessofinterventiontointerrupttransmission.Protocolsanddefinitions
developedinonelaboratoryforagiventypeofexperimentmaydiffersignificantlyfromthoseusedinotherlaboratoriesforthesametypeofexperiment.Forexample,
inevaluatingtheefficiencyofwatersamplingtechniques,onelaboratorymayuse"fresh"cystsandanothermayuse"preserved"cysts.Bothtermsareimpreciseand
havedifferentmeaningstodifferentinvestigators.Thecystsusedbydifferentlaboratoriesmaybefromdifferentsourcesandofdifferentagesandspecies.Assay
proceduresmayalsovaryandmaybeofunknownorundeterminedprecision.Differencesinanyofthesefactorscouldaffecttheresultsandinterpretationof
experiments.
Standardizationofcystsource,age,purity,preservation,storageconditions,quantitationandidentificationcouldimprovereproducibilityofresultswithinalaboratory
andsimplifycomparisonandinterpretationofresultsacrosslaboratories.Thesesamefactorshaverelevancetoalltypesofstudieswherecystsareusedincluding
crossspeciesinfectivity,survivalandinactivationstudies,speciationandevaluationofdetectionmethodologies.WorkingwithGiardiaisverymuchan"art"those
factorsthatareimportantinreproducibilityofresultsneedtobeidentifiedandacceptablecriteriaforthemmustbeestablishedinordertoadvancethe"science".
Discussion
Commentsandquestionsfromthefloorreinforcedsomepointsmadebythepanelmembers,challengedothersandraisedadditionalquestionsforconsideration.With
respecttoaxenization,otherinvestigatorsaswellhavefoundinoculationintosucklingmicetobetheprocedureofchoiceforestablishingtrophozoitecultures.Gram
negativeorganismsappeartobethemostdifficultcontaminanttodealwithandoneinvestigatorreportednosuccessinfindingasatisfactoryantibioticagainst
Aerobacter.Itwasalsomentionedthatcontaminatedculturesmightberescuedifthecultureisinoculatedintoananimal.TheanimalmayallowgrowthoftheGiardia
andnotofthecontaminant.
Itwassuggestedthatclonalisolationandcryopreservationofisolatesbeperformedearlyinculturesincetheorganismscanbesubjecttoselectionandchangeupon
subculture.Itwasagreedthatthisisadesirableprocedure.However,existingcloningmethodsareverydifficulttouseandbettermethodsareneeded.
Thecommentwasmadethatcystsmayloseinfectivityifleftincontactwithfecesfor24hours.Anotherinvestigatorhasobtainedinfectivityofcystsleftinstoolsfor
morethanaweek.Theconflictingresultsmayreflectthenumbersofcystsoriginallypresentinthesample,themakeupofthestoolortheanimalmodelused.
However,thereappearedtobegeneralagreementthatcystsshouldbeseparatedassoonaspossiblefromfecalmaterialinordertoincreasethechancesofcyst
survival.
TheproblemofconfusioninGiardianomenclaturewasreiterated.InanAustralianlaboratory,theyhaveadoptedtheWHOsystemusedfordesignatingtrypanosome
isolates.AdescriptionofthissystemmaybefoundintheJournalofAntimicrobialChemotherapy14:449461,1984.
ThequestionofamaturationperiodforGiardiacystsstillappearstobeaquestion.WhiletherewasagreementthatG.murisdoesnotrequireamaturationperiod,
someinvestigatorsbelieveitisnecessaryforG.duodenaliswhileothersdonot.Again,differences,intheexperimentalprotocolsmayaccountforthecontraryresults
ortheremaybetruedifferencesamongspecies.Furtherinvestigationofthisquestionisneeded.

Page294

CryopreservationofGiardiawasdiscussed.Thepreservationofactivelygrowingculturesoftrophozoitesisnowreadilyachievedbyfreezingtheseorganismsinthe
presenceofglycerolordimethylsulfoxide.However,cryopreservationofcystsrequirestremendousnumbersofcystsandtheresultsareerratic.Thecommentwas
offeredthatCryptosporidiumoocystsdonotsurvivefreezingbutthatthesporozoitescanbefrozenandrecovered.
Withrespecttocystsusedtoevaluatedrinkingwatertreatmentprocesses,suggestionsweremadethatonlycystsnotmorethan7daysoldbeusedandthattheynot
bepreserved.Formalinpreservedcystsmaybemorefragilethanunpreservedcystsandtheirusecanleadtofalselyhighefficienciesforwatertreatmentremoval
processes.Itwassuggestedthatthecystsbeofgoodquality,i.e.,havingcytoplasmthatisalmosttranslucentandnotgranular.
Onelaboratoryindicatedtheyfindmorecystsinthesupernatantofextractsfromstoolspecimensandwaterfiltersthantheyfindinthesedimentafterusingestablished
protocolsforcystrecovery.Thiswascontrarytoresultsobtainedinatleasttwootherlaboratoriesbutitisaquestiondeservingfurtherstudyandagainreflectsour
lackofknowledgeonwhattoconsiderassignificantfactorsandhowtostandardizeprocedures.

Page295

INDEX
A
acidphosphatase139,187188,252,254,268269
actin175
activatedserumrabbitorhuman169
acute1,33,49,5253,107,191
adhesivedisc2930,231,266
adsorbed46,155,169171
agilis79,287288
Alberta1519,6164,7982,107,129131,133,137,140,143144,159,161,164165,167169,172,255,289
amodiaquine3
animalinfectivity32,205,237,249,252255,265,269,281,286
animalmodel28,3132,49,55,64,75,77,79,82,144,189,204,234235,254255,260,265,268269,282,288289,293294
animalreservoir55,69,71,74,82,152,164,210,231,289
animalshelter61,6365,165
antiepidemicmeasurement40
antiGiardiaantibody2,4748,50,52,192
antiGiardiaimmunity47
antibodies2,4,4550,52,5458,66,156,159164,169172,177,179180,192,194,205207,209,216,219220,222223,273,283
antigen14,28,4648,50,55,153157,162,169,171172,177180,191194,220,222,288
antisera1,153156,162,169,171,191,220,222,231
aquaticmammals227
Arkansas7172,74
arylsulfatase268
assessment34,71,103,172,198,202,205,228,237,255,259,277278,283286
asymptomatic1,1618,33,36,65,82,153,163,177179,181184,191,219,222,232,251,262263
attachment24,3334,3637,46,49,54,6869,88,99,107,236,262,265,288
axenic2,4,7,12,21,24,2931,33,37,54,58,81,147,156,162,167168,172,180,192,194,223,230,232,235,254,265,288,291
axonemes26,2930,133134,154,197,221,253,266
B
Bcell47
bacteria34,7,9,12,14,2528,48,67,8789,9192,95,9798,100,103,107108,111112,205,208,265,275,286,291
basalbodies29
beaver29,7477,7980,9091,129,159162,164165,167,169170,202,204,223,227234,236237,239241,245,253,285,287288,293
beigemice47
berberinesulfate34
bile4,9,12,24,2931,33,37,46,4850,52,5455,147,168,180,219221,223,251252,254,261,263264,273
bileductcannulation50
bilesalt9,33,261,273
biliaryantibodyresponses52
binarycysts227,231232
BinghamMeyer250252
biochemical9,29,31,55,160,162,165,189,223,235,287288
biopsy3,45,68,7980,139
brightfieldmicroscopy32,253,269
C
Calgary3,6164,7980,137138,160,165,169170,205,217,254255,286,288
canine61,6465,69,220221,232
carbohydrate46,68,147,187189
carbondioxide88,109,126,219220,249,251252
cartridgefilters89,206
categorization224
cats51,65,75,227,232,235,240,245
cattle16,7577,79,160161,208,227,232233,240,245,293
cDNA147148,151153,155157,184
cell6,914,22,2427,2930,3337,4549,111,118,133135,152,155157,160,166,173,175,177,180181,183,185,189,219221,227,231,236,
256,258,260,265,267,287
cellfree33,35,155,183
cellviability9,12,265
characterization4,1314,48,94,153154,157,172,180184,187,194,269,287
chemical29,36,50,89,95,97,100103,107108,112,116,118119,121,125126,128,138,160,162,189,198,200203,215216,223,237,239,243
244,250,255258,283,

Page296

286287,293
chemotaxonomy287,292
children'shomes39
Chilomastixmesnili40,211
chlorine7577,8788,97,103,105,108,112113,117119,125126,128129,133135,137138,140144,198199,223225,233234,238,242,246,
254,256,258,282283,286,293
chromosomes147152,287
chronic1,7,25,27,45,49,5355,58,107,191
clinicalfeatures179
coagulation95,97,99103,117,124,129,197,205,242,280281
coagulationfiltration95,100102
complementmediated169,171
contact1518,22,3336,39,8788,97,105,107,115,118119,121,123,125,127,137,141144,199,223224,234,238,246,249,256257,275,282
286,294
contamination1617,65,74,77,87,103,106107,110,119,143,198199,205,208,223,225,227,233235,238241,245,273,284285
counterimmunoelectrophoresis12,191192,219,222
coyotes245
criteria3,9,2223,3031,49,117121,144,197,216,234235,266,275276,278,285288,294
CTvalue103,137,141,143144,275,283284,286
culture1,4,7,912,2124,2831,3336,7981,97,139,147,152,160161,165166,168,170,177,179180,191,219221,223,230,232,235,253254,
288,291292,294
cysteine21,250251,256,261264,273
cytokinesis249,252
cytometry910,1214
cytopathogenic33
cytotoxic33,4548,5558,163,171,265
Czechoslovakia3940
D
daycare1,4,1516,1819,49,69,7374,191,194,236
daycarecentres39,237
detachment3435
detection4,14,27,48,5152,54,67,103107,129,144,152153,155,163164,168,179,182,188189,194,197,202210,213,215,217,219225,228,
232,234,236238,255,260,265,269,287,294
detergent148,180,206,211213
diagnosis34,40,48,65,6869,151,179,194,197,202,204,213,216,219,222223,241
diagnostictest153,179,221
diatomaceousearth87,95,9899,102103,113114,237,244,285
diet6869,160
diethylether80,211,213
differentialinterferencecontrast(DIC)2932,249,252253,265269
disinfection95,99100,104,107,111112,118119,135,137,140141,144,223,233234,249,254,275,277,280286
distribution2,17,28,39,4749,6568,71,7374,87,8990,103105,107108,113114,117118,120,124125,135,162,177,189,192,223224,228,
231232,236,238,242,244,275277,285,293
DNA0,6,12,21,2426,28,49,55,58,69,147153,155157,160,163,165168,173,175176,183,230,236,287289
dog6169,7576,7980,82,130,159161,165,167,169170,227,231233,236,240,245,250,253,293
drug37,910,12,14,2124,2728,58,6566,139,156
duodenalis25,4954,75,7981,133,137,139,141,143,165172,204,231232,234,237,245246,283,287289,293294
DusanLambl39
dyeexclusion4,139,252255,258259,265,281,286
E
ectoplasmicvacuoles249,252
Edmonton1516,19,6164,129,131,137,165
effector4748
electronmicroscopy25,29,31,3336,47,113,135,205,220221,232,236,252,254255,265266,268269,273
electrophoresis12,4,50,55,147153,160,165166,168169,178,181183,191192,194,289
ELISA14,50,52,156,160161,191194,219,222
elution12,50,181,191192,206207
encystation2930,69,75,187,222,235,273274
encystment219221

Page297

endonuclease28,49,55,69,147148,152,165168,230,236,289
endosymbiont2628
energymetabolism187189
Entamoebahistolytica4,12,24,37,194,211,219
enteric1,15,1819,28,47,71,191,205
enzyme1,4,6,4852,58,116,147,151,166167,180,182,187189,191,194,219,222223,262
eosin28,31,252255,259260,265,268269,283,286
EPA4,32,65,68,81,87,9495,102104,112113,116117,135,194,197198,204,209,220,225,254,273,275,277278,286
epidemiological3941,61,65,81,103,129,227,229,236,278
epidemiology19,31,37,39,41,48,69,135,162164,235237,288289
erythromycin34,22
ethylacetate72,211,213
excretion49,5152,54,160,222,231
excystation2829,3132,50,55,75,80,82,107,109,112,125,133,135,139141,143144,163164,168,171,187,189,220,249265,268269,273
274,281283,286
F
familialoccurrence3940
fattyacid273
FDA910,2931,265268
FecalParasiteConcentrator211
fecaloral15,40,168,232,255
FeKalCONTrate211
fibroblastcells3536
fieldinversiongelelectrophoresis147
filter1,910,21,33,75,8793,95110,112,114116,118121,123,129131,133,137138,140143,163,191,197203,205208,216,220,224,237,
242,244245,255256,262,275,280,285,295
filtration12,26,75,8788,92,94105,108119,124,129130,133,137,139143,145,191192,197200,203,205,209,212213,215,220,223,233
234,237,239245,251,277,280281,285
flagella10,2123,3031,35,133135,227231,251,253,266,292
Flagyl(seemetronidazole)75
flange3536,292
FluoraBoraI(3(dansylamido)phenylboronicacid252
fluoresceindiacetate9,29,32,198,205,237,252253,265,269,281,286
fluorescence4,910,12,30,47,154,194,203,217,221,223224,231,253,255258,265268
fluorescenceactivatedcellsorter(FACSIV)265
fluorescent10,12,25,203,206,209,215216,222225,231,249,252,254256,258,267,269,286
fluorogenic2932,198,205,231,252253,255,265269,273,281
food1518,40,50,66,80,88,125,177,205,231,239,249,273
Formalinether(FE)sedimentation211,213
furazolidone37,911,21,2324
G
gastrointestinal34,41,4547,177,194,227,249
gelelectrophoresis12,4,147,149,151153,160,166,168169,178,181183,191192,194
genes51,147,150151,173,175176,183184
genetic28,49,147,150151,160,165,167,230
genome49,147,149151,166
genotype288
gerbil55,75,7980,139,165,246,253,256,293
giardicidal29
giardins154
glutathione3,6,250251,256,261263,273
glutathioneperoxidase6
glutathionereductase6
golgi252
granularmedia100101
growth1,9,12,14,21,24,2829,37,6869,88,91,96,117,125,160,179,191,200,224,241,294
gypsies39
H
Hanks'balancedsaltsolution250251,256
HeLa2728,3336
helper/inducer4548
homosexual254
host4,29,31,33,3637,4849,53,55,61,65,75,7981,148,152153,159,162,165,167,169170,172,177,179,181,183,

Page298

230231,234,249250,253,273,281283,287288,292
hspgene173175
HSP3250,261,264
human4,7,9,12,17,27,31,33,35,45,4749,51,53,55,5758,61,6366,6869,7477,7982,97,109,125,129,133,135,137,139,142,144,159
163,165,167172,175,177180,182184,187,203,209,211,213,216,219223,227234,236237,240,245,249250,253255,261264,273274,
282284,286,288289,292293
humoral47,49,55,171,177,180
Hutterite1519
hydrolyticabilities187
hypersensitivity54
hypothymic49,5152,5455,113
I
IgA4550,52,5455,57
IgE50
IgG2,4,4546,4850,57,66,180,192,194,203,219
IgM45,4850,52,5455,57
immobilization54,116,169172
immune12,4,4749,5255,5758,171172,177,180184,191192,236,246,282,293
immunity47,49,5355,172,234
immunodeficiency45,180
immunodiffusion169,172
immunoelectrophoresis12,153,191192
immunofluorescence12,47,49,59,66,153154,159162,164,180,197,203,205206,209210,215217,219,222223,225,234,236237,260,265,269,
283
immunoglobulin45,4749,55,161,220
immunological1,45,47,57,81,153,157,169,191,232,234235,287288
immunology25,37,45,47,49,157
immunoreactivity29,231
invitro1,37,21,23,2833,3537,46,51,5557,6869,75,8082,109,112,133,135,139140,144,153,155157,160,164,169172,175,181184,
187,189,191,219222,235,249250,254255,257,260,263,265,268269,273274,281,286,291292
invivo4,6,12,2931,36,46,49,8081,139,220221,249250,261,288
inactivation4,7,75,77,94,110113,125128,133,135,137138,144145,170171,224,254,275,277,280286,293294
incidence19,41,64,66,6869,7174,129,227,255,279
incorporation6,2931,87,138,183,211,213,264266
incubation2,11,2122,3435,50,117,161,182,192,221,250252,256,261262,291
indicator87,105,107,130131,202,205,231,237
indigenous92,208,228,230,232
indirect4,66,155,159162,194,203,206,215,219,223224,237,265,282
induction30,184,250252,258,261263
infection4,1519,2528,31,39,4557,6166,6869,7577,7982,129,137,139140,143144,152,163165,171,177,179181,194,204,210,219,
222,227228,231,234,236237,240,245246,249,253254,265266,273278,281282,288289,292293
infectivity15,29,32,144,159,204205,231232,237,249,252255,260,265266,269,275,281283,286,294
infusion50,52,54
inhibitor6,155
inoculation33,48,5657,7677,80,266,291,294
inoculum10,34,56,81,143,231,266,291
institution71
intestinal34,2528,31,33,3637,4041,4549,52,5455,57,6466,6869,74,88,164165,171,177,180181,187,194,213,219,222,227228,232,
236,249,254255,260,273,286287,291
ioniccomposition265
isoenzyme37,49,5354,289
isolation19,24,31,37,65,8081,106,109,151,157,163,172,177,179181,183,205,209,223,237,250,254,256,265,273,294
K
karyokinesis249
kD183
killercells45,4748
Kochs'postulates29,31
L
Lcysteinehydrochloride250
lamblia1,813,17,19,2126,28,31,3334,39,45,54,61,6869,7174,77,82,94,

Page299

107,109,125127,129,135,142144,156157,159164,167168,172173,176184,187189,191,205,211213,223,225,227,231233,235237,
249265,268269,273275,280283,286,288289,293
lateralshield3536
lectin46,48
length26,28,31,34,68,79,81,97,99,102,147,149151,198,229231,283284,292
lesions236237
Lethbridge6164
libraries147148,151152,156157,184
lifecycle29,31,273
lightscatter914,98,113
lipid211
lymphocyte47,269
lysosomalenzymes27,252,268
M
macrophages4647,49,171172
malabsorption1,33,45,107,177,180,191,236237
maturation88,92,137138,142,250252,273,293294
McKeesport103106,233
mechanism6,24,33,48,53,65,88,91,99,165,244,267268,288
media22,8792,97,100101,103,105107,114,118119,121,137,142,144,197,200,202,206207,250251,261264,291
medianbody26,62,169170,221,231232,287288,292
membrane12,2526,36,98,107112,130,133135,139,143,154,159,162,180,198,200203,205209,215,224,237,252,255,265,267269,293
membranefilter107111,139,198,200,206208
membraneintegrity268
membranepermeability268
Merionesunguiculatus55,75,7980,139,165,246,253,256,293
metabolic67,10,27,50,88,187,189,235,252
metabolism6,9,12,88,187189,254,260
methods3,9,12,19,21,31,49,6567,72,89,94,103,106,112,129131,135,139,144,153154,159160,162163,165,170,177,181184,193,197,
199201,204205,208209,213,215,217,222223,225,234235,249,251253,255259,261265,269,273,275,281,286288,291,294
metronidazole(seeFlagyl)312,16,18,2124,4952,54,56,75,7980,165
microfoci3940
microscope14,26,2931,34,89,104,109,133,197,201203,215217,220,224,236,246,265266
microscopy25,2937,47,50,76,80,113,116,135,150,159,164,172,197198,203,205206,209,216217,219223,225,232,236,249,252255,260,
265266,268269,273,283,287
microtine162,231232,234
microtubules26,48,133,154,156
microvillousborder36,236
MIFcentrifugation6163
milk17,4548
minichromosomes147,149,151
mitochondria252
models3,6,29,49,103,113,234,253254,265,282,287,293
Mongoliangerbil(seeMerionesunguiculatus)32,55,58,61,64,7577,144,189,249,253254,256,260,282283,288,293
monitoring15,77,89,99,101,103107,109110,119,129131,203,215216,223225,235,237,275,277,279,285286
monoclonal4549,55,58,156,159164,203207,209,216,219222,273,283
monolayer3336,203
mononuclearleukocytes169,172
morbidityrate277
morphology2831,3336,49,52,62,79,113,133134,169,198,205,231232,234,253,255,265268,287,292
motility910,12,2123,37,171,257259,288
mouse4,25,28,32,4555,76,7980,82,91,109,113,135,159162,165,175176,187,189,219220,231,234,251,253,255,260261,265,269,287,
289,291292
mRNA152153,155157,175,183
mucosa4647,54,227228
murine31,4849,5455,170171,236
muris4,2425,2728,3032,45,49,5155,79,98,105,107,109,111,133,135,137144,159161,169172,187189,230,232,236237,249259,261
266,268269,273,281282,286288,292294
muskrat29,7677,7980,91,159165,167,202,223,228,231232,234,236237,239240,245,273274,285,293

Page300

N
NADPHoxidase6
neonatal4,7,151,265266,293
nonviable3031,265266,268,283
nucleicacid181,268
nudemice4649,55
O
ondatrae287288
ornidazole34,8
orthogonalfieldalternation147
outbreak4,1519,49,61,6365,74,82,87,103107,113,117,129130,137,159,163,194,205,208209,223,225,227230,232234,236237,245,
255,279,281
outercystwall29
oxygen28,91,125127,187188,223
P
PAGE(seepolyacrylamidegelelectrophoresis)13,153156,178,182183,191193
pancreatin251,261
ParaPakMacroCon211
particulate97,99,101,108,110111,197,212,216,256,286
passive6,4950,52,54,87,110
passivetransferofbile54
pathogenesis31,3637,69,135,162164,235236,288289
pathway6
Pentatrichomonashominis40
peptone166,251,261264
peripheralvacuoles2930,266,268
peritrophicspace2931,252253,265268
persontoperson1516,1819,47,177,249,254
perspectives69,162,167,235
pets64,234
Peyer'spatch4649
pH12,26,29,66,89,100102,108,117119,121,125,133,135,137139,142143,149,154,161,177178,181182,188,191192,199,220,223,243,
249252,256,258,261262,265,269,273,280,283,286,291
phagocytosis49,172
phenotype288
physiological2,31,54,192,211,281
phytone250251,261264,291
PI(seepropidiumiodide)2930,265268
pig7577
plasmacell4647
polyacrylamidegelelectrophoresis(seePAGE)1,4,153,169,178,183,191,194
pregnancy3,7
prepatent57,63,75
prevalence18,3941,48,6165,71,79,81,163,172,181,204,228,231232,234,236
primary4,6,9,3940,4952,54,56,68,95,100,180,184,194,200,219220,222,234,241,273
production4,29,32,4547,58,74,7677,80,87,89,99,109,120121,124,127,139,143,155,160,170,194,220,222223,253,274,281
propidiumiodide(seePI)12,32,198,205,237,252253,265,269,281,286
proteolysis268
proteosepeptone251,261264
protozoa7,12,25,28,31,41,64,80,88,147,150,163,165,181,183184,197,200201,205206,208,213,220,230,236,254,269,289
protozoan1,15,25,2729,45,54,64,68,71,130,155,165,177,181,187,191,205,208,211213,222,227,249,254255,260,286287
psittaci236,287288
pyruvate:ferredoxinoxidoreductase6,188189
Q
quinacrine36,8,12,2124
R
rabbits1,153154,156,169170,191
rat1213,46,4855,81,159162,172,182183,236,249,254,265,288
rawwastewater231,233
reducingagents251,262263,265,273
removal66,8788,90103,105106,108,113114,116119,121,137138,140143,145,155,200,202203,209,223,237,243,273,275,277,280281,
285,293294
Rendtorff255,282,286
reservoirs55,65,69,74,77,8182,104,117,152,162164,199,204,209210,223,225,227,231,233235,245,289
resistance37,16,27,4849,51,5457,108,172,177,254,283,286

Page301

respiration281
risk15,18,3940,103,108109,202204,212,223,227,234,237238,241245,249,275,277279,281,285287,292
riskmodel275,277,281,285
RNA0,24,148,155,157,175,181184
S
Saccharomycescerevisiae12,182
SAF6163
sample2,16,23,50,61,63,72,7576,8992,103105,108,111,113,125,129,131,139141,149,166,178179,183,192,197200,202,204,206208,
215216,220,224,237,242,244245,250,252,262263,266,275277,279280,285,292294
satranidazole3
scanningelectronmicroscope2930,220
secondaryinfection40,49,5152
secretory4849,5456,58,153,157,163,180,289
sedimentation50,55,88,95,100104,106,115,118119,129,181,183,187,189,200,211213,261262,265,280281
SEM2931,3435,37,113115,172,236,288
sensitivity47,21,2324,28,58,61,153,198,207,213,222,256,275
serum12,21,29,33,4546,4850,52,5456,66,69,139,154155,159161,169172,177178,180182,184,191192,203,219220,223,250252,
261265,269,291
serumfactor170
slowsand8789,9192,94100,102103,237,244,285
smallbowel55
sodiumbicarbonate109,250251,261
sodiumdodecylsulfate1,178,191,206
SPCA6164
speciation52,147,164,287,294
specificity49,153,159,165,167,219,222,287288
stage91,156,187,279
staining12,3132,34,47,55,65,72,107,112,134,149150,166,182,202203,205,217,219,222,237,252253,257,265269,286
stimulation69
stoolanalysis63
stools29,39,57,6163,65,7175,79,156,179,207,227,287,291,293294
strainidentification147
straydogs6164
sucklingmice274,291,294
sucrosegradient80,109,181183,187,205,291
sulfasalazine34
suppressor4547
surfacechanges35
surfacewater1516,74,82,8788,95,97,102103,107,115,117118,125,130,159,164,197,204205,208,215217,219,221,233,237,243,245,252,
275,280,282285
survival27,88,106,109,111,143,172,250,262,273,293294
symbiont25,2728
symptomatic1618,35,37,48,55,57,65,139,153,177179,181184,219,222,251,262264
symptomaticdonor251,263
symptomatology69
symptoms4,1618,33,55,57,68,177,227,273,292
T
Tlymphocyte4548,54
Tcell46
taxonomic81,147,172,287
TEM2526,30,37,133,172
temperature2,17,28,66,8891,93,96,100,102,106,108,112,115,118,121,125127,137,139,141,148,151,161,166,173,175,182,192,199,223,
243,249252,255,257258,261,265,273,283,286
tetracycline22,24
therapy35,78,24,48,65,68
tinidazole36,8,12,24
titer219,221
toxin58
TPS13335,37
translation24,152153,155157,181184
transmission4,1519,25,2830,32,3940,47,49,55,63,65,69,7577,79,8182,95,106,112113,117,129,133,152,162165,167168,180,194,
204,208210,223,225,227,229,231237,245,249,252,254255,286289,293
transmissionelectronmicroscopy25,29
trichomonads4
Trichomonasvaginalis3
TritonX10080,133,139,177178,182,187,211213
trophozoites1,35,912,2125,2731,3337,4548,5051,5457,6163,6569,7576,7982,107,139,147148,153155,160

Page302

161,165,169173,177,181182,187189,191,193,219222,227232,235236,249250,252253,255262,264266,269,273,287,291,294
trypanosomatids28
trypanosome294
trypsin33,109,139,251252,256,261264
trypsinTyrode's251,256,262
turbidity87,89,93,95106,116121,124,131,139,142143,197,199200,202,215216,223,234,237238,241,243,245,275,284285
TYIS1,9,21,24,31,56,82,135,144,147,166,168,170,177,180181,191,219220,223,254255,291
U
ultrastructural25,2829,3132,49,133,135,162163,232,236237,254,265269
ultravioletradiation107
V
ventraldisc23,26,29,36,133,154
ventrolateralflange3536
Vero3336
viability4,910,12,2124,2832,8081,107,109111,113,128,133,135,137144,198,205,220221,225,233,237,249,252261,265269,275,281
283,286
virulencedeterminants53
virus88,96,98,103,107,152,183,278279
vitalstain249
W
wall2122,26,2931,48,50,63,111,133134,161,163,187,189,201,203,219220,224,249,252253,258259,265268,273
watertreatmentplant8790,102104,107,117118,129130,133,137139,142145,233,279,283
waterborneoutbreak15,49,61,95,113,165,205,209,227229,233234,236,249,278
watershed87,129,159,208,223,227,229,231235,245,275,279,285
Westernblot153,177178,273
Y
YoughioghenyRiver103106
Z
Zibethecusondatrae287288
zincsulfateflotation213,223224
ZnSO46163,76,202,215217,224
zoonotic79,229,232,234,287288
zymodemes288