Substrate-labeled fluorescent

immunoassay for measuring

dibekacin concentrations in
serum and plasma.
J D Place and S G Thompson
Antimicrob. Agents Chemother. 1983,
24(2):240. DOI: 10.1128/AAC.24.2.240.

Downloaded from http://aac.asm.org/ on May 20, 2014 by guest

Updated information and services can be
found at:
http://aac.asm.org/content/24/2/240

These include:
CONTENT ALERTS Receive: RSS Feeds, eTOCs, free email
alerts (when new articles cite this article),
more

Information about commercial reprint orders:

http://journals.asm.org/site/misc/reprints.xhtml
To subscribe to to another ASM Journal go to: http://journals.asm.org/site/subscriptions/
ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Aug. 1983, p. 240-245 Vol. 24, No. 2
0066-4804/83/080240-06$02.00/0
Copyright 0 1983, American Society for Microbiology

Substrate-Labeled Fluorescent Immunoassay for Measuring

Dibekacin Concentrations in Serum and Plasma
JANET DOBBINS PLACE* AND STEPHAN G. THOMPSON
Immunochemistry Laboratory, Ames Research and Development, Division of Miles Laboratories, Inc.,
Elkhart, Indiana 46515
Received 7 February 1983/Accepted 2 June 1983

We have developed a homogeneous substrate-labeled fluorescent immunoassay

for the measurement of dibekacin concentrations in serum and plasma. The

Downloaded from http://aac.asm.org/ on May 20, 2014 by guest

fluorogenic enzyme substrate ,B-galactosyl-umbelUiferone was covalently attached
to tobramycin, an aminoglycoside structurally similar to dibekacin, to prepare a
fluorogenic drug reagent (FDR). The FDR is nonfluorescent under assay condi-
tions, but fluoresces upon hydrolysis by P-galactosidase. However, binding of the
FDR by antiserum to the cross-reactive drug kanamycin prevents enzyme
hydrolysis. The fixed level of FDR competes with dibekacin within the sample for
the limiting number of antibody-binding sites in the reaction mixture. Unbound
FDR is hydrolyzed by ,B-galactosidase to release a fluorescent product that is
proportional to the dibekacin concentration in the sample. The assay exhibits
good precision, standard curve reproducibility, recovery, sensitivity, and correla-
tion with a comparative method. Additionally, the substrate-labeled fluorescent
immunoassay is rapid and easy to perform.

Dibekacin (3',4'-dideoxykanamycin B) is a is generated as a function of the dibekacin

semisynthetic aminoglycoside antibiotic useful
in the treatment of severe gram-negative bacteri-
al infections (6, 9, 10). Since dibekacin has a
rather narrow therapeutic range and may be
ototoxic or nephrotoxic at high concentrations
in the blood, therapeutic drug monitoring is
indicated (3, 4, 7, 8). We have developed a
homogeneous substrate-labeled fluorescent im-
munoassay (SLFIA) for the determination of
dibekacin levels in serum or plasma (J. D. Place
and S. G. Thompson, Abstr. Annu. Meet. Am.
Soc. Microbiol. 1982, C230, p. 309). The princi-
ple of the dibekacin SLFIA is identical to that of
the SLFIA for the related aminoglycoside antibi-
otic tobramycin, which was described previous-
ly (1, 2). Indeed, except for the calibrators, the
ingredients of the reagents are the same. A
fluorogenic drug reagent (FDR) is nonfluores-
cent under the assay conditions but yields a
fluorescent product when hydrolyzed by ,-ga-
lactosidase (Escherichia coli ,B-D-galactoside ga-
lactohydrolase; EC 3.2.1.23). Antiserum to the
cross-reactive drug kanamycin A binds to the
FDR, prevents enzyme hydrolysis, and thus
inhibits the production of fluorescence. Dibeka-
cin in the sample competes with a constant
amount of FDR for a limiting number of anti-
body-binding sites so that a fluorescent response
240
concentration.
The performance of the dibekacin SLFIA was
evaluated in terms of accuracy and precision,
standard curve reproducibility, sensitivity, and
antiserum cross-reactivity. The validity of the
assay was further established by comparison of
assay results with a high-pressure liquid chroma-
tography (HPLC) analysis of clinical serum sam-
ples.
MATERIALS AND METHODS
Indruments. Fluorescence was measured with an
Aminco-Bowman fluorescence spectrophotometer
(SLM Instruments, Inc., Urbana, Ill.) or with an Ames
Fluorostat (Ames Co., Elkhart, Ind.), with excitation
and emission wavelengths set at 400 and 450 nm,
respectively. All fluorescence determinations were
performed at room temperature in disposable polysty-
rene cuvettes (Evergreen Scientific, Los Angeles,
Calif.). Absorbance was measured with a model 250
spectrophotometer (Gilford Instrument Laboratories,
Inc., Oberlin, Ohio).
Enzyme. 3-Galactosidase from E. coli (Sigma Chem-
ical Co., St. Louis, Mo.) was assayed at 25C in 50 mM
bicine-0.1% azide, pH 8.3, containing 3 mM o-nitro-
phenyl-P13-Dgalactoside. Under these conditions, the
millimolar extinction coefficient for the product of this
reaction, o-nitrophenol, is 4.27 at 415 nm. One unit of
VOL. 24, 1983 ASSAY FOR DIBEKACIN IN SERUM AND PLASMA 241

Downloaded from http://aac.asm.org/ on May 20, 2014 by guest

Dlbkacin Tobramycin Kanamycin
FIG. 1. Structures of the aminoglycoside antibiotics dibekacin, tobramycin, and kanamycin.

enzyme activity hydrolyzes 1.0 ,umol of substrate per
min.
Chemicals. Bicine buffer, N,N-bis(2-hydroxyeth-
yl)glycine (50 mM; grade A; Calbiochem, La Jolla,
Calif.), was used at pH 8.3. Sodium azide was pur-
chased from Fisher Scientific Co., Fairlawn N.J.
Gentamicin, amikacin, and tobramycin were ob-
tained from U.S. Pharmacopeial Convention, Inc.,
Rockville, Md., dibekacin was obtained from Meiji
Seika Kaisha, Ltd., Tokyo, Japan, and kanamycin was
purchased from Bristol Laboratories, Syracuse, N.Y.
The structural similarity of the related aminoglyco-
sides dibekacin, kanamycin, and tobramycin is shown
in Fig. 1.
Dibekacin clinical serum samples were obtained
from USV Laboratories, Revlon Health Care Group,
Tuckahoe, N.Y. and Miles Sankyo Limited, Tokyo,
Japan.
HPLC. HPLC of the clinical serum specimens was
carried out by the method of Patel et al. (manuscript in
50 mM Bicine-0.1% azide, pH 8.3, plus 500 ,ul of the
same buffer were added to the antibody-enzyme rea-
gent. (iii) The reaction was initiated by the addition of
50 ,ul of FDR (absorbance per ml at 343 nm, 0.0165) in
5 mM sodium formate buffer, pH 3.5, plus 500 ,ul of 50
mM Bicine-0.1% azide, pH 8.3. The FDR was deliv-
ered to each cuvette at 15-s intervals until all of the
reactions in the run were initiated. The final reaction
1004
90
80
70
10
X60 -

preparation).
FDR. The FDR 3-galactosyl-umbelliferone-tobra- ~50-
mycin was synthesized as described previously (1). U.40-
Antiserum. Antiserum to kanamycin was produced
in goats with a kanamycin-bovine serum albumin 30-
conjugate (5).
SLFIA for the determination of dibekacin levels in 20
serum. The SLFIA was performed using a semiauto-
mated procedure. The following additions were made 10
sequentially with an Ames diluter to a series of reac-
tion cuvettes. (i) A 50-,ul portion of an antibody-
enzyme reagent prepared in 50 mM Bicine-0.1% 1 2 3 4 5 6 7 8 9 10
azide, pH 8.3, plus 500 ,ul of the same buffer was ML Antiserum
dispensed into the cuvette. The antibody-enzyme rea-
gent contained 30 times the concentration of antiserum FIG. 2. Effect of antiserum to kanamycin on the
required for the assay and 1.5 U of 3-galactosidase per hydrolysis of the dibekacin FDR by 3-galactosidase.
ml. (ii) Samples (50 ,ul each) of dibekacin standards, Reactions were conducted in the presence (0) and
controls, and unknowns previously diluted 1:51 with absence (0) of a diluted 16-p.g/ml dibekacin standard.
242 PLACE AND THOMPSON ANTIMICROB. AGENTS CHEMOTHER.
A Time (min.) B
15 60 15

14 40
14-

13 30 13

12 12

l ' l l '20
. ' '
10 lo10

Downloaded from http://aac.asm.org/ on May 20, 2014 by guest

0

08 10 03

0
7Li
66
0
5 4 8 12 164
5Z -
8 2 1

4 4-

3 3

2 2

1 1

0 I0
4 8 12 16 4 8 12 16

Ag Dibekacin/mL
FIG. 3. Effect of incubation time on the dibekacin SLFIA standard curve. Standard curves generated after
incubation periods of 5 to 60 min are presented. The curves in (A) show fluorescence values, whereas those in (B)
are normalized to the fluorescence of the cuvette containing the highest drug standard.

mixture contained sufficient antiserum to inhibit hy- mined from a standard curve of fluorescence versus
drolysis of the FDR by 80%o, 0.05 U of 0-galactosidase the dibekacin concentration.
per ml, and FDR (absorbance per ml at 343 nm,
0.0005) in a final volume of 1.65 ml. After the first RESULTS
reaction mixture had incubated at room temperature
for 20 min, the fluorescence intensity for this and Antibody-binding reactons. The inhibition of
subsequent cuvettes was measured at 15-s intervals. enzymatic hydrolysis of FDR by antiserum to
The unknown dibekacin concentrations were deter- kanamycin is shown in Fig. 2. The addition of

TABLE 1. Dibekacin SLFIA intraassay accuracy and precision

Control Mean (Lg/ml) SD (pg/ml) % CVa
value 1 Dilution, 20 Dilutions, 1 Dilution, 20 Dilutions, 1 Dilution, 20 Dilutions,
(gLg/ml) 20 replicates 1 replicate of each 20 replicates 1 replicate of each 20 replicates 1 replicate of each
2 1.79 1.76 0.09 0.08 5.03 4.55
6 5.97 6.05 0.10 0.08 1.68 1.32
14 14.50 14.29 0.26 0.34 1.79 2.38
a CV, Coefficient of variation.
VOL. 24, 1983 ASSAY FOR DIBEKACIN IN SERUM AND PLASMA 243
TABLE 2. Dibekacin SLFIA interassay accuracy 11 _
and precisiona
10
Control value Mean SD % CVb
(Lg/mL) (,ug/mL) (pLg/mL) 9
2 1.88 0.123 6.53 08
6 6.05 0.064 1.06
14 14.02 0.243 1.73 07
a n = 50 assays; 10
runs over 8 days. 0
6
b CV, Coefficient of variation.
5
4-
the diluted 16-,ug/ml dibekacin standard to the 3

Downloaded from http://aac.asm.org/ on May 20, 2014 by guest

reaction mixtures yielded increased fluores- 0
cence due to competition of the drug with the 4 8 12 16
FDR for the antibody-binding sites. The fluores- 9g Dibekacin/mL
cence difference in the presence and absence of FIG. 4. Reproducibility of the dibekacin standard
the standard at any one level of antiserum ap- curve. The brackets indicate 1 standard deviation
proximates the dose response at that level. The over 10 runs (8 days).
maximum difference usually indicates what level
of antiserum will provide both a good dose
response and a reasonably shaped standard
curve. The dibekacin concentrations of clinical sam-
Competitive binding reactions. The effect of ples determined by SLFIA were compared with
various incubation times on the dibekacin stan- those measured by HPLC (Fig. 5). Comparison
dard curve was followed (Fig. 3). The curves in of the two methods revealed both good correla-
Fig. 3A show the measured fluorescence, tion (r = 0.987) and regression parameters.
whereas the curves in Fig. 3B are normalized to Specificity of the dibekacin SLFIA. The cross-
the fluorescence of the high standard. Although reactivity of the antiserum to kanamycin for
the dibekacin assay is normally run with a 20- dibekacin and other aminoglycosides was exam-
min incubation, it is possible to incubate the ined by measuring the dose response of various
reactions for alternative lengths of time, depend- aminoglycosides in the dibekacin assay (Fig. 6).
ing on the requirements of the user. Shorter The cross-reactivity of the drugs was calculated
incubation times enable one to obtain results at the concentration of drug that elicits a fluores-
quickly but limit the number of samples that can cent response equivalent to 50% of the maxi-
be processed per run. Longer incubation times mum response due to dibekacin. Even though
enable one to assay a larger number of samples
in a run. Acceptable standard curves were gen-
erated with incubation times offrom 5 to 60 min. 9-
The fluorescent response of the assay (the differ-
ence in fluorescence between the 0- and 16- 8- y =0.94 x +0.24
r = 0.987
p,g/ml standards) increased from 3.3 relative 7
n =151
fluorescence units at 5 min to 7.4 at 60 min. S.E.E -0.36Ag/mL
Performance characteristics of the dibekacin .0 6-6..*
SLFIA. The intraassay precision of the dibeka-
cin SLFIA was evaluated by assaying 1 dilution
of each of the dibekacin controls (2, 6, and 14 Xj ....
5 -.
p,g/ml) 20 times and also 20 dilutions of each , 4-.
control once (Table 1). Each standard was as- AC
sayed in duplicate. Interassay precision was Q8
0 3. g. .
determined from data generated in 10 runs over
a period of 8 days. In each run, five dilutions of
each control were assayed once (Table 2). Preci-
sion over the 10 runs was good. The assay was
very reproducible over the 8-day period (Fig. 4).
Recovery experiments were performed by di-
luting a 20-,ug/ml contrived dibekacin serum 0 1 2 3 4 5 6 7 8 9
sample 1:1 with 50 mM Bicine-0.1% azide, pH 9g Dibekacin/mL by HPLC
8.3. The recovery of the diluted sample was FIG. 5. Comparison of the dibekacin SLFIA with
100.6% of the expected value. HPLC. S.E.E., Standard error of estimate.
244 PLACE AND THOMPSON
100 -
so[
W 70 I.
W 50
so
E
0 40
-a
I
10
.1 0.2 1 2
FIG. 6. Cross-reactivity of other aminoglycosides in the dibekacin assay.
the antiserum was produced in response to a
kanamycin immunogen, dibekacin was the most
reactive, whereas tobramycin and kanamycin,
both structurally very similar to dibekacin,
cross-reacted 94 and 69o, respectively. Amika-
cm cross-reacted 1%, whereas the unrelated
aminoglycoside gentamicin did not cross-react
at all. Other antibiotics that may be coadminis-
tered with dibekacin, such as carbenicillin,
cephalothin, chloramphenicol, erythromycin,
methicillin, neomycin, and tetracycline, did not
cross-react or interfere in the assay when thera-
peutic levels were assayed in the presence of
dibekacin.
Sensitivity of the dibekacin SLFIA. The sensi-
tivity of the dibekacin SLFIA was evaluated by
assaying five replicates each of the -,ug/ml
standard and 0.8-, 1.6-, 2.4-, and 3.2-ixg/ml stan-
dards which were contrived by diluting the 4-
,g/ml standard with the 0-,ug/ml standard. This
study indicated that 0.5 ,ug of dibekacin per ml
could be distinguished from the 0-u.g/ml stan-
dard at a 95% confidence level.
DISCUSSION
The dibekacin SLFIA takes advantage of the
cross-reactivity of dibekacin, tobramycin, and
kanamycin for binding to an antiserum to kana-
mycin. The assay described herein is similar to
the tobramycin SLFIA (1) in that both assays
employ antiserum to kanamycin and 3-galacto-
syl-umbelliferone-tobramycin as the FDR. They
differ, however, in the optimal concentrations of
these reagents.
ANTIMICROB. AGENTS CHEMOTHER.
Downloaded from http://aac.asm.org/ on May 20, 2014 by guest
6 1012142030 50 50 IS0 2u000
Various studies were carried out to establish
the validity of the dibekacin SLFIA. This assay
exhibited very good intra- and interassay preci-
sion, reproducibility of the standard curve, and
recovery of the drug. Comparison of the SLFIA
with HPLC values yielded good correlation, and
the assay is sensitive to 0.5 ,ug/ml. The evalua-
tion demonstrated that the dibekacin SLFIA is a
rapid, Precise, and accurate assay.
ACKNOWLEDGMENTS
We acknowledge Joan K. Clayborn for her technical assist-
ance and thank John C. Godfrey, USV Laboratories, Revlon
Health Care Group, for providing clinical samples and helpful
information.
LITERATURE CITED
1. Bord, J. F., R. J. Carrico, H. M. Kramer, ad C. E.
D_nIng. 1978. Homogeneous substrate-labeled fluores-
cent immunoassay for determining tobramycin concentra-
tions in human serun, p. 387-403. In S. B. Pal (ed.),
Enzyme labeled immunoassay of hormones and drugs.
Walter de Gruyter, Inc., New York.
2. Burd, J. F., S. G. Tbomps, and C. A. Mlfler. 1980.
Substrate-labeled fluorescent immunoassays for measur-
ing levels of the aminoglycoside antibiotics gentamicin,
sisomicin, netilmicin, tobramycin, kanamycin, and amika-
cin, p. 517-519. In J. D. Nelson and C. Grassi (ed.),
Current chemotherapy and infectious disease. Proceed-
ings of the 11th International Congress of Chemotherapy
and the 19th Interscience Conference on Antimicrobial
Agents and Chemotherapy, vol. 1. American Society for
Microbiology, Washington, D.C.
3. Camf, J., J. Segura, J. E. Baflos, I. Garcia, and L.
Drobaic. 1982. Evaluation of urinary elimination of N-
acetyl-p-glucosaminidase in healthy volunteers treated
with dibekacin or gentamicin. Antimicrob. Agents Che-
mother. 21:1013-1015.
4. Fujff, R., and S. Hahira. 1978. Dibekacin in children:
clinical evaluation and pharmacokinetics, p. 933-935. In
VOL. 24, 1983 ASSAY FOR DIBEKACIN IN SERUM AND PLASMA
W. Siegenthaler and R. Luthy (ed.), Current chemothera-
py. Proceedings of the 10th International Congress of
Chemotherapy, vol. 2. American Society for Microbiolo-
gy, Washington, D.C.
5. Lewis, J. E., J. C. Nebon, and H. A. Elder. 1972. Radio-
immunoassay of an antibiotic: gentamicin. Nature (Lon-
don) New Biol. 239:214-216.
6. PullIam, L., W. K. Hadley, and J. Mlls. 1981. In vitro
comparison of third-generation cephalosporins, piperacil-
lin, dibekacin, and other aminoglycosides against aerobic
bacteria. Antimicrob. Agents Chemother. 19:490492.
7. Rimoldi, R., L. Curcio, and A. SanfiHppo. 1978. Pharma-
cokinetic study of dibekacin in humans, p. 928-930. In W.
Siegenthaler and R. Luthy (ed.), Current chemotherapy.
Proceedings of the 10th International Congress of Chemo-
245
therapy, vol. 2. American Society for Microbiology,
Washington, D.C.
8. Ueda, Y., A. Salto, F. Matsumoto, M. Omori, K. Shiba, T.
Yamaji, and H. Jhara. 1978. Clinical studies on dibekacin,
p. 931-933. In W. Siegenthaler and R. Ltithy (ed.),
Current chemotherapy. Proceedings of the 10th Interna-
tional Congress of Chemotherapy, vol. 2. American Soci-
ety for Microbiology, Washington, D.C.
9. Umezawa, H., S. Umezawa, T. Tsuchlya, and Y. OkazakL.
1971. 3',4'-Dideoxykanamycin B active against kanamy-
cin-resistant Escherichia coli and Pseudomonas aerugino-
sa. J. Antibiot. 24:485-487.
10. Umezawa, S., J. Unzawa, Y. Okazald, and T. Tsuchlya.
1972. Studies on aminosugars. Synthesis of 3',4'-Dide-
oxykanamycin B. Buli. Chem. Soc. Jpn. 45:3624-3628.
Downloaded from http://aac.asm.org/ on May 20, 2014 by guest