Anda di halaman 1dari 11

G Model

VETMIC-6881; No. of Pages 11

Veterinary Microbiology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Veterinary Microbiology
journal homepage: www.elsevier.com/locate/vetmic

The anti-canine distemper virus activities of


ex vivo-expanded canine natural killer cells
Ji-Yun Park a,1, Dong-Jun Shin a,b,1, Soo-Hyeon Lee a,b, Je-Jung Lee b,
Guk-Hyun Suh d, Duck Cho b,c,**, Sang-Ki Kim a,b,*
a
Department of Companion & Laboratory Animal Science, Kongju National University, Yesan, Chungnam 340-702, Republic of Korea
b
Research Center for Cancer Immunotherapy, Chonnam National University Hwasun Hospital, Jeollanamdo 519-809, Republic of Korea
c
Department of Laboratory Medicine, Chonnam National University Medical School, Gwangju, Republic of Korea
d
Department of Veterinary Internal Medicine, College of Veterinary Medicine, Chonnam National University, Gwangju, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Article history: Natural killer (NK) cells play critical roles in induction of antiviral effects against various
Received 10 October 2014 viruses of humans and animals. However, few data on NK cell activities during canine
Received in revised form 22 January 2015 distemper virus (CDV) infections are available. Recently, we established a culture system
Accepted 23 January 2015
allowing activation and expansion of canine non-B, non-T, large granular NK lymphocytes
from PBMCs of normal dogs. In the present study, we explored the ability of such expanded
Keywords:
NK cells to inhibit CDV infection in vitro. Cultured CD3CD5CD21 NK cells produced
Canine
large amounts of IFN-g, exhibited highly upregulated expression of mRNAs encoding NK-
NK cells
Canine distemper virus cell-associated receptors, and demonstrated strong natural killing activity against canine
Antiviral effects tumor cells. Although the expanded NK cells were dose-dependently cytotoxic to both
IFN-g normal and CDV-infected Vero cells, CDV infection rendered Vero cells more susceptible
to NK cells. Pretreatment with anti-CDV serum from hyperimmunized dogs enhanced
the antibody-dependent cellular cytotoxicity (ADCC) of NK cells against CDV-infected
Vero cells. The culture supernatants of NK cells, added before or after infection, dose-
dependently inhibited both CDV replication and development of CDV-induced cytopathic
effects (CPEs) in Vero cells. Anti-IFN-g antibody neutralized the inhibitory effects of NK
cell culture supernatants on CDV replication and CPE induction in Vero cells. Such results
emphasize the potential significance of NK cells in controlling CDV infection, and indicate
that NK cells may play roles both during CDV infection and in combating such infections,
under certain conditions.
ß 2015 Elsevier B.V. All rights reserved.

1. Introduction

Canine distemper virus (CDV), a negative-strand RNA


* Corresponding author at: Department of Companion and Laboratory
Animal Science, College of Industrial Science, Kongju National University, morbillivirus closely related to measles virus, is globally
Yesan-eup, Yesan-gun, Chungnam 340-702, Republic of Korea. distributed and is the agent of highly contagious disease
Tel.: +82 41 330 1525; fax: +82 41 330 1529. (Martella et al., 2008). CDV causes systemic infections
** Corresponding author at: Research Center for Cancer Immunothera- associated with high mortality rates. Essential clinical signs
py, Chonnam National University Hwasun Hospital, Jeollanamdo 519-
809, Republic of Korea. Tel.: +82 61 379 7951; fax: +82 61 379 7984.
may be noted in the respiratory, gastrointestinal, and central
E-mail address: sangki@kongju.ac.kr (S.-K. Kim). nervous systems (Leisewitz et al., 2001). The mounting of a
1
These authors contributed equally to this work. good host immune response is crucial if virus spread is to be

http://dx.doi.org/10.1016/j.vetmic.2015.01.021
0378-1135/ß 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

2 J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx

contained. Recent studies have revealed that the host Hospital, Memphis, TN). Canine thyroid adenocarcinoma
immune defense against CDV infections is very complex. (CTAC) cells were purchased from the European Collection
Such complexity arises from the multicomponent nature of of Cell Cultures (Porton Down, UK). Cells were maintained in
the defense and interactions among various components RPMI-1640 medium supplemented with 10% (v/v) fetal
thereof (Appel et al., 1982; Ho and Babiuk, 1980; Rima et al., bovine serum (FBS) and antibiotics. Vero cells were obtained
1991). CDV replicates in the cytoplasm of host cells, and CDV from the American Type Culture Collection (Manassas, VA)
surface antigen proteins including hemagglutinin (H) and and were cultured in Dulbecco’s Modified Eagle Medium
fusion protein (F) are expressed on the plasma membrane of (DMEM) supplemented with 10% (v/v) FBS and antibiotics.
the host cells during viral replication. CDV-infected cells are
thus excellent targets for cytolysis (Sarute et al., 2013; Shek 2.2. Expansion of canine non-B, non-T, large granular NK
et al., 1980). Some reports have explored features of canine lymphocytes
lymphocytes, including killer cells that effectively destroyed
CDV-infected target cells via antibody-dependent cellular Blood was collected from nine healthy dogs (beagles,
cytotoxicity (ADCC), under circumstances where polymor- miniature schnauzers, and English cocker spaniels) for
phonuclear cells and monocytes were ineffective in this selective expansion of non-B, non-T, large granular NK
context (Ho and Babiuk, 1979; Ho et al., 1978). lymphocytes (NK cells). All animal protocols used con-
Natural killer (NK) cells are a subset of lymphocytes that formed to the guidelines of the Institutional Animal Care
can directly kill both virus-infected and tumor cells, thus and Use Committee of Kongju National University. In order
in the absence of prior activation, and that produce anti- to expand NK cells, isolated canine PBMCs were incubated
viral cytokines (including IFN-g) that limit viral infection with K562 cells expressing membrane-bound IL-15, and
(Brunetta et al., 2010; Rager-Zisman and Bloom, 1982; Sun the 41BB ligand (gift of Dr. D. Campana, National
and Lanier, 2009). Recently, it was found that NK cells play University of Singapore), at 37 8C, under 5% (v/v) CO2,
critical roles in induction of antiviral effects in humans and for 21 days in 24-well tissue culture plates. Cells were
animals (Lin et al., 2012; Pyzik et al., 2011; Rossini et al., cultured in RPMI 1640 medium (GIBCO, Grand Island. NY)
2012; Sentsui et al., 2010; Zhang et al., 2010). CDV is supplemented with 100 IU/ml human IL-2 (PeproTech,
lymphotropic, and causes a systemic (and often fatal) Rocky Hill, NJ), 5 ng/ml canine IL-21 (PeproTech), 10 IU/ml
disease commonly associated with severe immunosuppres- human IL-15 (PeproTech), and 10% (v/v) FBS. The medium
sion (Kumagai et al., 2004; Moro et al., 2003; Tipold et al., was changed every other day.
2001). However, NK cell activities during CDV infection
remain poorly understood. It is not clear whether canine NK 2.3. Flow cytometric analysis
cell activities significantly inhibit CDV replication in vivo,
although one early report evaluating pre- and post-infection Cells were stained as described previously (Shin et al.,
CDV-specific NK activity found no infection-associated 2013). Briefly, fluorescence-activated cell sorting (FACS)
increase in the cytolysis of CDV-infected or -noninfected analysis was performed with the aid of an FITC-labeled
Vero target cells (Ringler and Krakowka, 1985). mAb against CD3, RPE-labeled mAbs against CD21, and an
Few data on the antiviral activities of canine NK cells APC-labeled mAb against CD5 (Serotec, Oxford, UK).
against CDV or other viruses are available, because canine Surface expression of MHC-class I molecules and NKG2D
NK cells are poorly characterized. No specific markers of ligands (MIC-A, MIC-B, ULBP-1, ULBP-2, and ULBP-3) on
such cells are known, and NK cells are not abundant in Vero cells was evaluated with the aid of appropriate FITC-
peripheral blood (Bonkobara et al., 2007; Huang et al., conjugated antibodies (Becton Dickinson, Franklin Lakes,
2008; Lin et al., 2010). Thus, canine NK cells have been NJ). Isotype controls were run in parallel. Flow cytometric
simply defined as non-B, non-T, large granular lympho- analyses were performed using a FACSCalibur flow
cytes (Michael et al., 2013). In our present study, we for the cytometer (Becton Dickinson).
first time expanded such lymphocytes, exhibiting mor-
phological, genetic, and functional characteristics of NK 2.4. Propagation and infection of canine distemper virus
cells, from peripheral blood mononuclear cells (PBMCs) of
normal dogs, and explored the ability of such lymphocytes The Lederle strain of canine distemper virus (CDV; Animal,
to inhibit CDV infection in vitro. We found that expanded Plant and Fisheries Quarantine and Inspection Agency,
NK cells efficiently inhibited CDV infection of Vero cells, via Anyang, Korea) was used throughout the study. CDV was
ADCC, direct cytotoxicity, and IFN-g synthesis. The data propagated in Vero cells at 37 8C under 5% (v/v) CO2, in a
suggest that activated canine NK cells potentially play an humidified atmosphere, and cytopathic effects (CPEs) were
important role in control of CDV infection, particularly noted daily. When the CPE was maximal (at 4.5 days post
when viral maturation and release is prolonged. infection), cell-free culture medium was harvested by
centrifugation and filtered through membranes 0.2 mm in
2. Materials and methods pore diameter. Viral titer (4.58  105 PFU/ml) was deter-
mined via a plaque assay, and the virus stock stored at 80 8C.
2.1. Cell lines Vero cells were infected with CDV at a multiplicity of infection
(MOI) of 0.5. The proportions of CDV-infected cells were
The K562 cell line, genetically modified to express determined by flow cytometry, with the aid of an anti-CDV
membrane-bound IL-15 and the 41BB ligand, was kindly antibody (Santa Cruz Biotechnology, Dallas, TX) and FITC-
provided by Dr. D. Campana (St. Jude Children’s Research conjugated anti-mouse IgG (Serotec), 3 days after infection.

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx 3

2.5. Enzyme-linked immunosorbent assay (ELISA) washed with culture medium, and PBMCs or expanded NK
cells were resuspended at effector-to-target (E:T) ratios of
The levels of IFN-g and TNF-a production by PBMCs, 2.5:1, 5:1, 10:1, and 20:1, and incubation continued at
and expanded NK cells, in response to CTAC cells, were 37 8C for a further 3 h. After addition of 10-ml WST-1
analyzed by ELISA. Briefly, CTAC cells (5  104) were added, solution (ItsBio) to each well, the plates were incubated for
in triplicate, to 96-well microplates, and cultured over- a further 1 h and then placed on ice for 10 min to stop the
night at 37 8C; the plates were then washed with medium. reactions. Absorptions at 450 nm were measured using an
PBMCs (5  105) or cultured NK cells (5  105) were co- Infinite M200 PRO (Salzburg Umgebung, Salzburg,
cultured with CTAC cells at a 10:1 effector-to-target (E:T) Austria). To measure ADCC, CDV-infected or -uninfected
ratio, without cytokine addition. Twenty-four hours later, Vero cells were pretreated with heat-inactivated CDV
cell-free culture supernatants were harvested and the antiserum collected from dogs exhibiting high antibody
IFN-g and TNF-a levels therein assayed using OptEIATM titers, at 37 8C for 1 h, and washed with medium prior to
Set Canine Kits (BD Biosciences, Franklin Lakes, NJ), co-culture with effector cells at a 5:1 E:T ratio. Cytotoxicity
according to the manufacturer’s instructions. Cell-free proportions were calculated using the following equation:

A450 of effector cell-treated target cells  A450 of effector cells ðeffector cell backgroundÞ
100%  100
A450 of target cells  A450 of target cells lacking WST-1 ðtarget cell backgroundÞ

culture supernatants from CTAC, PBMC, and NK cells 2.8. Inhibition of CDV replication by NK cell culture
cultured in medium alone, for 24 h, served as controls. supernatants

2.6. Real-time RT-PCR To explore whether NK cells and NK cell culture


supernatants could inhibit CDV replication, 1 ml amounts
Real-time RT-PCR was used to measure the expression of NK cell culture supernatants at 1:2, 1:4, and 1:8 dilutions
levels of mRNAs encoding NK-associated genes in both were added to Vero cells (1.2  105) in 12-well plates before,
expanded NK cells and PBMCs, and to determine viral copy or 2 h after, CDV infection. To determine whether pretreat-
numbers, using a QuantiTect SYBR Green PCR Kit and a ment of Vero cells with NK cell supernatants inhibited
Rotor-Gene Q kit (QIAGEN, Hilden, Germany). For NK- CDV infection, Vero cells were treated with NK cell culture
related genes, extraction of total RNAs, synthesis of first- supernatants for 24 h, followed by washing and infection
strand cDNAs, and real-time PCR, were performed as with CDV. Control supernatants were those of lethally
described previously (Shin et al., 2013). To determine viral irradiated (125 Gy) K562 cell cultures grown without or
copy numbers, viral RNAs were extracted from cell-free with addition of IFN-g (100 mg/ml). Vero cell cultures were
culture supernatants of CDV-infected Vero cells using a observed daily for development of CPEs, and 600-ml
viral RNA extraction kit (Bioneer, Daejeon, Korea), accord- amounts of culture supernatants were collected on day
ing to the manufacturer’s instructions. CDV copy numbers 5 for CDV viral RNA extraction and real-time RT-PCR.
were calculated by reference to a standard curve generated
using 10-fold serial dilutions of RNA complementary to 2.9. Assay of IFN-g neutralizing antibody levels
the CDV N gene. This cRNA was synthesized via in vitro
transcription of a linearized pGEM T-Easy vector (Promega, An anti-IFN-g antibody (PeproTech) was used to
Madison, WI) carrying the full-length CDV N gene, with the neutralize the anti-CDV activities of IFN-g-containing
aid of T7 RNA Polymerase (TaKaRa, Tokyo, Japan). The and cell-free culture supernatants of NK cells. Culture
thermal cycling conditions for quantitative PCR were 95 8C supernatants were diluted 1:1 with culture medium
for 15 min, followed by 45 cycles of 94 8C for 30 s; 53 8C for containing anti-IFN-g antibody at a dilution of 1:1000,
30 s; and 72 8C for 30 s. The sequences of the oligonucleo- and incubated at 37 8C for 1 h prior to addition of Vero cells
tide primers used to amplify NK cell-associated genes previously infected with CDV (for 2 h); all tests were
have been described previously (Shin et al., 2013). The performed in triplicate. CDV-infected Vero cells were
sequences of oligonucleotide primers for the CDV N gene observed daily for development of CPEs, and culture
amplified will be made available upon request. supernatants were collected on day 5 for viral RNA
extraction and real-time RT-PCR analysis. Control super-
2.7. Cytotoxicity assays natants were those of lethally irradiated K562 cell cultures,
diluted 1:1 with normal culture medium, and next added
An EZ-Cytox Cell Viability Assay kit (ItsBio, Seoul, to CDV-uninfected or -infected Vero cell cultures growing
Korea) was used to measure the 4-h cytotoxic activities of with or without IFN-g (100 mg/ml).
expanded NK cells or PBMCs against both tumor cells and
CDV-infected cells (Park et al., 2012). Totals of 10,000 2.10. Statistical analysis
target cells (CTAC cells, or CDV-infected or -uninfected
Vero cells) were cultured in 96-well flat-bottomed plates, The statistical significances of observed differences
in triplicate, at 37 8C overnight. The target cells were next were determined using the Mann–Whitney U test for

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

4 J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx

nonparametric comparisons. The minimal level of signifi- cells at a 5:1 effector-to-target ratio (p < 0.05) (Fig. 3B).
cance was set at p < 0.05. Control serum (thus lacking anti-CDV antibodies) did not
influence the cytotoxicity of NK cells against CDV-infected
3. Results cells.

3.1. Characteristics of expanded non-B, non-T, large granular 3.3. CDV infection was significantly suppressed by NK cell
NK lymphocytes culture supernatants, and anti-IFN-g antibody neutralized
the inhibitory effects of NK cell supernatants on CDV
NK cells were selectively proliferated over a 4-week replication and CPE induction in Vero cells
period in the culture system. Expanded NK cells increased
in size and became very granular after 21 days in culture, We next evaluated the inhibitory effects of NK cell
as revealed by flow cytometric analysis of front scatter culture supernatants on CDV-induced CPEs and viral
(FSC)/side scatter (SSC) features (data not shown). The replication in Vero cells. NK cell culture supernatants at
phenotype of most expanded cells was CD3CD21CD5 different dilutions (1:2, 1:4, and 1:8) were added to CDV-
(Fig. 1A). To explore the antiviral effects of such NK cells infected Vero cells in 12-well plates, and CPEs were
against CDV, we used NK cell preparations that were over recorded daily. NK cell culture supernatants dose-depen-
85% pure after 21 days of culture. The quantitative real- dently suppressed CDV-induced CPEs in Vero cells (Fig. 4).
time RT-PCR data of Fig. 1B show that the levels of mRNAs The inhibitory effects of NK cell culture supernatants on
encoding NK-associated genes were significantly higher in CDV replication in Vero cells were further confirmed via
NK cells cultured for 21 days than in freshly isolated real-time RT-PCR of viral RNAs extracted from culture
PBMCs; this was true of cells from all donors. Remarkable supernatants collected on day 5 post-infection (Fig. 5).
increases in the levels of mRNAs encoding NKG2D (32.4- Addition of NK cell culture supernatants 1 h post-infection
fold), NKp30 (25.9-fold), NKp44 (70.1-fold), NKp46 (160.2- significantly reduced CDV viral load, in a dose-dependent
fold), perforin (37.7-fold), granzyme B (14.5-fold), and manner (p < 0.01) (Fig. 5). In control experiments, treat-
Ly49 (295-fold) were observed in cultured NK cells ment with culture supernatants of K562 cells alone, at a
compared to freshly isolated PBMCs (all p values <0.05). 1:1 dilution, did not suppress CDV infection, and addition
Cultured NK cells produced significantly higher levels of of 100 mg/ml amounts of INF-g to control supernatants
both IFN-g and TNF-a, important cytokines produced by completely inhibited the development of CDV-induced
activated NK cells, compared to the levels synthesized CPEs (Fig. 4), and viral replication in Vero cells, to the time
by PBMCs or IL-2-activated LAK cells (p < 0.01) (Fig. 1C). at which cell culture was terminated (Fig. 5). Anti-IFN-g
The IFN-g concentrations in supernatants from NK antibody (10 mg/ml) significantly neutralized the abilities
cells, LAK cells, and PBMCs were 8999.6  81.2 pg/ml, of NK cell culture supernatants and INF-g to inhibit
284.0  11.0 pg/ml, and 364.1  7.1 pg/ml, respectively. The induction of CPEs (Fig. 6A) and CDV replication (Fig. 6B) in
TNF-a levels in supernatants from NK cells, LAK cells, and Vero cells (p < 0.05).
PBMCs were 82.8  6.32 pg/ml, 9.15 0.9 pg/ml, and
13.2  0.1 pg/ml, respectively. The natural cytotoxicity of 3.4. Pretreatment of Vero cells with NK cell culture
cultured NK cells against CTAC cells (the standard target supernatants reduced CDV replication
cell line used in assays of NK cell cytotoxicity) was significant,
and increased in a dose-dependent manner when NK cells To determine whether pretreatment of Vero cells with
from all donors were tested (Fig. 1D) (p < 0.01). NK cell supernatants could inhibit CDV infection, Vero cells
were treated with NK cell culture supernatants for 24 h,
3.2. Cultured NK cells kill CDV-infected cells via both natural followed by washing and infection with CDV. As shown in
killing and ADCC Fig. 4C, development of CDV-induced CPEs in Vero cells
were inhibited upon pretreatment with NK cell culture
To assess the direct cytolytic activities of cultured NK supernatants at a 1:2 dilution. The inhibitory effects of
cells against CDV-infected cells, both CDV-infected and - such pretreatment on CDV replication in Vero cells was
uninfected Vero cells were used as targets. The median confirmed via real-time RT-PCR of viral RNAs extracted
infection rate of CDV was 44% (data not shown). Infection from culture media collected on day 5 post-infection
with CDV downregulated the surface expression levels of (Fig. 7). Pretreatment with NK cell culture supernatants
MHC-class I molecules on Vero cells. However, the prior to infection significantly reduced the CDV viral load,
expression levels of NKG2D ligands including MIC-A and in a dose-dependent manner (p < 0.05). Pretreatment with
-B; and ULBP-1, -2, and -3 on such cells did not change culture supernatants of K562 cells exerted no inhibitory
(Fig. 2). As shown in Fig. 3A, after 21 days of culture, NK effect on CDV infection, and addition of 100 mg/ml of INF-g
cells were dose-dependently cytotoxic to both normal and to control supernatants completely inhibited CDV replica-
CDV-infected Vero cells, and CDV infection rendered Vero tion in Vero cells (p < 0.01) at all tested culture times
cells more susceptible to killing by cultured NK cells (Fig. 7).
(p < 0.05). The cytotoxicity exhibited against CDV-infected
Vero cells was 54.7  7.0%, and that against uninfected Vero 4. Discussion
cells 42.5  5.1%, at a 20:1 effector-to-target ratio. Hyper-
immunized dog anti-CDV serum significantly enhanced the NK cells mediate innate defenses against infections, and
cytotoxic activities of NK cells against CDV-infected Vero many studies have described NK-cell-dependent protective

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx 5

actions during viral infections of the human and mouse.


Important antiviral effects include direct killing of infected
cells and production of cytokines exhibiting antiviral
activity (Lee and Biron, 2010; Vidal et al., 2011; Zhang
et al., 2010). However, little has been published on the NK
cell defenses active against severe virus infections, such as
CDV infection of dogs. Furthermore, studies seeking to
correlate NK activities with resistance or susceptibility to
CDV infections have yielded conflicting results (Ho and
Babiuk, 1979; Ho et al., 1978; Ringler and Krakowka, 1985;
Shek et al., 1980). Canine NK cells are not well-characterized,
and this may explain the limited data gathered to date on
correlations between the severity of CDV infection and NK
cell activities, and among-study inconsistencies in results
(Bonkobara et al., 2005, 2007; Huang et al., 2008; Shin et al.,
2013). No NK cell-restricted marker, or receptor, allowing
canine NK cells to be distinguished from B and T cells, has yet
been identified, and canine NK cells are simply defined as
non-B, non-T, large granular lymphocytes (Michael et al.,
2013). The present study is the first to selectively expand
canine non-B, non-T, large granular NK lymphocytes ex vivo
from PBMCs of normal dogs and determine their anti-CDV
antiviral activities.
The expanded non-B, non-T (CD3CD5CD21) large
granular lymphocytes exhibited the morphological, genet-
ic, and functional characteristics of NK cells (Fig. 1). The
levels of mRNAs encoding NK-associated genes, such as
NKG2D, natural cytotoxicity receptors (NCRs), perforin,
and granzyme B, were markedly higher in cultured non-B,
non-T, large granular lymphocytes than in freshly isolated
PBMCs (Fig. 1B). The cultured test cells produced markedly
high levels of both IFN-g and TNF-a (Fig. 1C) and showed
significantly increased natural cytotoxicity against canine
NK-sensitive CTAC cells (the standard target cell used in
NK cell cytotoxicity assays) without specific antigen
recognition (p < 0.01) (Fig. 1D). Additionally, pretreatment
with hyperimmunized canine anti-CDV serum significant-
ly enhanced the ADCC-mediated cytotoxic activity of NK
cells against CDV-infected Vero cells (p < 0.05) (Fig. 3B).
One potent effector mechanism of NK cells is ADCC
mediated by antiviral antibodies binding to the FcgRIII on
NK cells (Johansson et al., 2011). Because CDV replicates
via budding from the cell membrane, and because CDV
surface antigen proteins such as hemagglutinin (H) and
fusion protein (F) are expressed on the plasma membrane
of the host cells during viral replication, CDV-infected cells
are excellent targets for destruction via ADCC (Sarute et al.,
2013; Shek et al., 1980). Taken together, these results
strongly suggest that selectively expanded non-B, non-T
large granular lymphocytes represent populations of
Fig. 1. Phenotypic and functional characteristics of expanded non-B, non-
T, large granular NK lymphocytes. (A) PBMCs (day 0) and expanded NK
canine NK cells.
cells (day 21) were stained with anti-CD3, -CD5, and -CD21 antibodies CDV infection rendered Vero cells significantly more
(markers of T and B cells), and analyzed via flow cytometry. Expanded NK susceptible to the cytotoxic activities of cultured NK cells
cell (CD3CD21CD5) populations exhibiting over 85% purity after (p < 0.05) in a dose-dependent manner (Fig. 3A). Very little
21 days of culture were used to evaluate the antiviral effects of such cells
on CDV (n = 6). (B) The levels of mRNAs encoding NK cell-associated
molecules, including CD16, CD56, NKG2D, NKp30, NKp44, NKp46, Ly49, expanded NK cells, in response to CTAC cells, as analyzed via ELISA. The
perforin, and granzyme B, in expanded NK cells from six different donors, IFN-g and TNF-a levels are shown as means  standard deviations (SDs)
as determined by real-time RT-PCR. All mRNA levels are expressed (n = 5). (D) NK cell cytotoxic activities against CTAC cells (a canine NK cell-
relative to those of b-actin, and the median, the first (Q1) and third (Q3) sensitive line). The 4-h cytotoxicities of NK cells were measured at various
quartiles, and the minimum and maximum, are shown. (C) The levels of effector-to-target (E:T) ratios. The results are shown as means  SDs of the
IFN-g and TNF-a produced by PBMCs, IL-2-activated (LAK) cells, and NK cell cytotoxicity percentages. N = 6 independent experiments.

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

6 J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx

Fig. 2. Expression profiles of MHC class I molecules and NKG2D ligands (MIC-A, MIC-B, ULBP-1, ULBP-2, and ULBP-3) on CDV-infected Vero cells (gray-filled)
compared to normal Vero cells (solid line) and the isotype controls of normal (dotted line) and CDV-infected cells (gray-thick line). Representative flow
cytometry histograms derived from data of three independent experiments are shown. Expression of MHC class I molecules was downregulated upon CDV
infection.

information is available regarding the mechanism by the NCRs NKp30, NKp44, and NKp46. However, whether
which NK cells recognize CDV-infected target cells. the hemagglutinin protein of CDV can bind to NCR to
Infection with CDV downregulated the surface expression activate canine NK cells has not been reported. Further
of MHC-class I molecules on Vero cells, while, CDV studies are required to explore the roles of the NCRs
infection did not alter expression of ligands for the NKp30, NKp44, and NKp46 in the recognition of CDV-
activating NKG2D receptor (Fig. 2). The effector functions infected cells by NK cells.
of NK cells are balanced by signals transmitted via Recent studies have shown that NK cells play important
activating and inhibitory receptors. The lack of KIR- or roles in eradication of viral infections by secreting
Ly49-MHC class I interactions is associated with the inflammatory cytokines, especially INF-g, as well as by
potency of NK-mediated target cell killing (Pyzik et al., directly lysing infected cells (Shabani et al., 2014; Zhang
2011; Rossini et al., 2012; Shabani et al., 2014; Vidal et al., et al., 2010). INF-g and TNF-a levels were markedly
2011; Zhang et al., 2010). An additional hypothesis to elevated in NK cell culture supernatants (Fig. 1C); however,
explain the increased susceptibility of CDV-infected cells GM-CSF was not detected in such supernatants (data not
to NK cell lysis is that the CDV H protein, which is the major shown). Thus, we evaluated the inhibitory abilities of NK
factor determining cell tropism, functions as a ligand for cell culture supernatants against CDV replication and

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx 7

Fig. 3. The natural cytotoxicity of expanded NK cells against CDV-infected Vero cells (A). The 4-h cytotoxicities of NK cells were measured at various effector-
to-target (E:T) ratios. The cytotoxicities of NK cells from six donors were tested; all reactions were performed in triplicate. (B) The antibody-dependent
cellular cytotoxicities (ADCCs) of expanded NK cells against CDV-infected Vero cells. In this assay, CDV-infected or -uninfected Vero cells were pretreated
with heat-inactivated CDV antiserum collected from dogs exhibiting high antibody titers, at 37 8C for 1 h, and the cells washed with medium prior to co-
culture with effector cells at an E:T ratio of 5:1. The ADCCs of expanded NK cells from six donors were assessed. The median infection rate of CDV was 44%.
The results are shown as means  SDs of the percentages of NK cell cytotoxicities. *p < 0.05.

induction of CPEs in Vero cells. The results showed that by inducing the synthesis of several proteins that interfere
treatment with INF-g or NK cell culture supernatants, either with viral replication (Goodbourn et al., 2000). Zhang et al.
before or after CDV infection, significantly suppressed both (2010) reported that the culture supernatant of human NK
CDV-induced CPEs (Fig. 4) and CDV replication in Vero cells cells inhibited replication of West Nile virus (WNV) and that
(Figs. 5 and 7), and anti-IFN-g antibody (10 mg/ml) anti-IFN-g neutralizing antibody blocked the NK superna-
significantly blocked the abilities of NK cell culture super- tant-mediated anti-WNV effect.
natants and INF-g to inhibit CDV replication and induction Numerous studies have found that human NK cells are
of CPEs in Vero cells (p < 0.05) (Fig. 6). Addition of anti-TNF- important in the immune response to a number of viruses,
a antibody did not affect the ability of NK cell culture including influenza virus, vaccinia virus, hepatitis B virus,
supernatants to inhibit CDV replication or induction of CPEs WNV, HIV-1, and measles virus (Griffin et al., 1990;
in Vero cells (data not shown). These results suggest (at Johansson et al., 2011; Kottilil et al., 2003; Lin et al., 2012;
least) that IFN-g produced by canine NK cells exerts a Shabani et al., 2014; Vidal et al., 2011; Zhang et al., 2010).
significant anti-CDV effect when present in NK cell culture Additionally, NK cell activation levels are negatively
supernatants, although we did not measure the levels of correlated with virus DNA levels in infected patients
type I interferons or other cytokines in such supernatants. (Kottilil et al., 2003; Shabani et al., 2014). It has also been
IFN-g is an important component of the host immune shown that the number of activated NK cells is markedly
response to viral infections and achieves its antiviral effects increased during the acute phase of measles virus infection

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

8 J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx

Fig. 4. The inhibitory effects of NK cell culture supernatants on CDV-induced cytopathic effects (CPEs) in Vero cells. NK cell culture supernatants at different
dilutions (NK sup 1:2, 1:4, and 1:8) were added to CDV-infected Vero cells growing in 12-well plates, in triplicate, and observed daily for development of
CPEs. Supernatants from lethally irradiated K562 cell cultures alone (CDV control) or added IFN-g (100 mg/ml) served as controls. To determine whether
pretreatment of Vero cells with NK cell supernatants inhibited CDV infection, Vero cells were treated with culture supernatants for 24 h, followed by
washing and CDV infection. (A) CDV-uninfected control cells. (B) Each experimental supernatant was added to Vero cells after 1 h of CDV infection. (C) Vero
cells were pretreated with culture supernatants for 24 h, followed by washing and CDV infection. On day 5 after infection, CPEs were microscopically
documented (100).

Fig. 5. NK cell culture supernatants inhibit CDV replication in Vero cells. NK cell culture supernatants at different dilutions (NK sup 1:2, 1:4, and 1:8) were
added to CDV-infected or control Vero cell cultures 2 h after infection. Culture supernatants were collected on day 5 for CDV viral RNA extraction and real-
time RT-PCR. Control supernatants were those of lethally irradiated K562 cell cultures, and were added to CDV-uninfected (normal control) or -infected
Vero cell cultures without (CDV control) or with IFN-g (100 mg/ml). The CDV viral RNA copy numbers were determined by reference to a standard curve
constructed using data on the CDV N gene. The median; first (Q1) and third (Q3) quartiles; and minimum and maximum, are shown. **p < 0.01.

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx 9

Fig. 6. Anti-IFN-g antibody blocks the inhibitory effect of NK cell culture supernatants on CDV-induced development of CPEs (A) and viral replication (B) in
Vero cells. NK cell culture supernatants were diluted 1:1 with culture medium (NK sup). Anti-IFN-g antibody was added both to the diluted medium (NK
sup+a-IFN-g Ab) and IFN-g supplemented control medium (IFN-g+a-IFN-g Ab) at a dilution of 1:1000, and incubated at 37 8C for 1 h prior to addition to
Vero cells preinfected with CDV for 2 h; all tests were performed in triplicate. Control supernatants were collected from lethally irradiated K562 cell
cultures, diluted 1:1 with normal culture medium, and added to CDV-uninfected (normal control) or -infected Vero cell cultures without (CDV control) or
with IFN-g (IFN-g). (A) On day 5 after infection, CPEs were observed microscopically (100). a, normal control; b, CDV control; c, IFN-g; d, NK sup; e, IFN-
g+a-IFN-g Ab; f, NK sup+a-IFN-g Ab. (B) Culture supernatants were collected for viral RNA extraction and real-time RT-PCR on day 5 after infection. CDV
viral RNA copy numbers were determined by reference to a standard curve constructed using data on the CDV N gene. The median; first (Q1) and third (Q3)
quartiles; and minimum and maximum, are shown. *p < 0.05.

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

10 J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx

Science and Technology (grant no. 2010-0010213), and by


the Bio-industry Technology Development Program
(112016-3), Ministry for Food, Agriculture, Forestry and
Fisheries, Republic of Korea. The authors would like to
thank Dr. D. Campana (National University of Singapore)
for providing cell lines.

References

Appel, M.J., Shek, W.R., Summers, B.A., 1982. Lymphocyte-mediated


immune cytotoxicity in dogs infected with virulent canine distemper
virus. Infect. Immun. 37, 592–600.
Bonkobara, M., Saito, T., Yamashita, M., Tamura, K., Yagihara, H., Isotani,
M., Sato, T., Washizu, T., 2007. Blastic natural killer cell leukaemia in a
dog – a case report. Vet. J. 174, 659–662.
Bonkobara, M., Sato, T., Takahashi, N., Kasahara, Y., Yagihara, H., Tamura,
K., Isotani, M., Azakami, D., Ono, K., Washizu, T., 2005. Characteriza-
tion of cDNA and the genomic sequence encoding canine neural-cell
Fig. 7. Pretreatment with NK cell culture supernatants rendered Vero cells adhesion molecule, CD56/N-CAM. Vet. Immunol. Immunopathol. 107,
resistant to CDV infection. Vero cells were treated with culture 171–176.
supernatants for 24 h, followed by washing and infection with CDV. Brunetta, E., Hudspeth, K.L., Mavilio, D., 2010. Pathologic natural killer cell
Culture supernatants were collected for CDV viral RNA extraction and real- subset redistribution in HIV-1 infection: new insights in pathophysi-
time RT-PCR on day 5. Control supernatants were collected from lethally ology and clinical outcomes. J. Leukoc. Biol. 88, 1119–1130.
irradiated K562 cell cultures, diluted 1:1 with normal culture medium, and Goodbourn, S., Didcock, L., Randall, R.E., 2000. Interferons: cell signalling,
immune modulation, antiviral response and virus countermeasures. J.
added to CDV-uninfected (normal control) or -infected Vero cell cultures
Gen. Virol. 81, 2341–2364.
without (CDV control) or with IFN-g (IFN-g). The CDV viral RNA copy
Griffin, D.E., Ward, B.J., Jauregui, E., Johnson, R.T., Vaisberg, A., 1990.
numbers were determined by reference to a standard curve constructed Natural killer cell activity during measles. Clin. Exp. Immunol. 81,
using data on the CDV N gene. The median; first (Q1) and third (Q3) 218–224.
quartiles; and minimum and maximum, are shown. **p < 0.01; *p < 0.05. Ho, C.K., Babiuk, L.A., 1979. Immune mechanisms against canine distem-
per. I. Identification of K cell against canine distemper virus infected
target cells in vitro. Immunology 37, 231–239.
(Okada et al., 2000) and that loss of NK cell activity is Ho, C.K., Babiuk, L.A., 1980. Immune mechanisms against canine distem-
correlated with measles disease progression (Griffin et al., per. III. Role of complement lysis in the immunity and persistent
1990). Taken together, these previous results imply that infection of canine distemper virus. Immunology 39, 231–237.
Ho, C.K., Babiuk, L.A., Rouse, B.T., 1978. Immune effector cell activity in
NK cells play an important role in protection against both canines: failure to demonstrate genetic restriction in direct antiviral
certain viral infections and progression of these viral cytotoxicity. Infect. Immun. 19, 18–25.
diseases. However, few studies have investigated the Huang, Y.C., Hung, S.W., Jan, T.R., Liao, K.W., Cheng, C.H., Wang, Y.S., Chu,
R.M., 2008. CD5-low expression lymphocytes in canine peripheral
antiviral activities of canine NK cells in veterinary
blood show characteristics of natural killer cells. J. Leukoc. Biol. 84,
medicine. Additional studies on NK cell activities against 1501–1510.
various viruses would clarify the importance of NK cells in Johansson, S.E., Rollman, E., Chung, A.W., Center, R.J., Hejdeman, B.,
protection against viral infections in dogs. Stratov, I., Hinkula, J., Wahren, B., Karre, K., Kent, S.J., Berg, L., 2011.
NK cell function and antibodies mediating ADCC in HIV-1-infected
Taken together, the results of this study demonstrate viremic and controller patients. Viral Immunol. 24, 359–368.
that activated canine NK cells have the ability to suppress Kottilil, S., Chun, T.W., Moir, S., Liu, S., McLaughlin, M., Hallahan, C.W.,
CDV replication and that such suppression by NK cells is Maldarelli, F., Corey, L., Fauci, A.S., 2003. Innate immunity in human
immunodeficiency virus infection: effect of viremia on natural killer
mediated by direct killing of infected cells and production cell function. J. Infect. Dis. 187, 1038–1045.
of IFN-g. However, a direct in vivo effect of NK cells on CDV Kumagai, K., Yamaguchi, R., Uchida, K., Tateyama, S., 2004. Lymphoid
infection in dogs remains to be shown. apoptosis in acute canine distemper. J. Vet. Med. Sci. 66, 175–181.
Lee, S.H., Biron, C.A., 2010. Here today – not gone tomorrow: roles for
In conclusion, our results suggest that NK cells play activating receptors in sustaining NK cells during viral infections. Eur.
crucial roles in the immune response against CDV infection J. Immunol. 40, 923–932.
in dogs, and correctly orchestrated NK cell responses can Leisewitz, A.L., Carter, A., van Vuuren, M., van Blerk, L., 2001. Canine
distemper infections, with special reference to South Africa, with a
completely remove CDV. Future research should focus on review of the literature. J. S. Afr. Vet. Assoc. 72, 127–136.
the use of novel therapies to enhance NK cell function Lin, S.J., Cheng, P.J., Lin, T.Y., Lee, P.T., Hsiao, H.S., Kuo, M.L., 2012. Effect of
during CDV infection. influenza A infection on umbilical cord blood natural killer function
regulation with interleukin-15. J. Infect. Dis. 205, 745–756.
Lin, Y.C., Huang, Y.C., Wang, Y.S., Juang, R.H., Liao, K.W., Chu, R.M., 2010.
Canine CD8T cells showing NK cytotoxic activity express mRNAs for
Conflicts of interest NK cell-associated surface molecules. Vet. Immunol. Immunopathol.
133, 144–153.
Martella, V., Elia, G., Buonavoglia, C., 2008. Canine distemper virus. The
No author has any possible financial or commercial
Veterinary clinics of North America. Small Anim. Pract. 38, 787–797,
conflict of interest relevant in the context of this article. vii–viii.
Michael, H.T., Ito, D., McCullar, V., Zhang, B., Miller, J.S., Modiano, J.F.,
2013. Isolation and characterization of canine natural killer cells.
Acknowledgments Vet. Immunol. Immunopathol. 155, 211–217.
Moro, L., de Sousa Martins, A., de Moraes Alves, C., de Araujo Santos, F.G.,
dos Santos Nunes, J.E., Carneiro, R.A., Carvalho, R., Vasconcelos, A.C.,
This research was supported by the Basic Science
2003. Apoptosis in canine distemper. Arch. Virol. 148, 153–164.
Research Program through the National Research Founda- Okada, H., Kobune, F., Sato, T.A., Kohama, T., Takeuchi, Y., Abe, T.,
tion of Korea (NRF) funded by the Ministry of Education, Takayama, N., Tsuchiya, T., Tashiro, M., 2000. Extensive lymphopenia

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021
G Model
VETMIC-6881; No. of Pages 11

J.-Y. Park et al. / Veterinary Microbiology xxx (2015) xxx–xxx 11

due to apoptosis of uninfected lymphocytes in acute measles patients. Sentsui, H., Wu, D., Murakami, K., Kondo, T., Matsumura, T., 2010. Antivi-
Arch. Virol. 145, 905–920. ral effect of recombinant equine interferon-gamma on several equine
Park, Y.K., Shin, D.J., Cho, D., Kim, S.K., Lee, J.J., Shin, M.G., Ryang, D.W., Lee, viruses. Vet. Immunol. Immunopathol. 135, 93–99.
J.S., Park, M.H., Yoon, J.H., Jegal, Y.J., 2012. Interleukin-21 Shabani, Z., Bagheri, M., Zare-Bidaki, M., Hassanshahi, G., Arababadi, M.K.,
increases direct cytotoxicity and IFN-gamma production of ex vivo Mohammadi Nejad, M., Kennedy, D., 2014. NK cells in hepatitis
expanded NK cells towards breast cancer cells. Anticancer Res. 32, B virus infection: a potent target for immunotherapy. Arch. Virol.
839–846. 159, 1555–1565.
Pyzik, M., Gendron-Pontbriand, E.M., Vidal, S.M., 2011. The impact of Shek, W.R., Schultz, R.D., Appel, M.J., 1980. Natural and immune cytolysis
Ly49-NK cell-dependent recognition of MCMV infection on innate of canine distemper virus-infected target cells. Infect. Immun. 28,
and adaptive immune responses. J. Biomed. Biotechnol. 2011, 724–734.
641702. Shin, D.J., Park, J.Y., Jang, Y.Y., Lee, J.J., Lee, Y.K., Shin, M.G., Jung, J.Y., Carson
Rager-Zisman, B., Bloom, B.R., 1982. Natural killer cells in resistance to 3rd, W.E., Cho, D., Kim, S.K., 2013. Ex vivo expansion of canine
virus-infected cells. Springer Semin. Immunopathol. 4, 397–414. cytotoxic large granular lymphocytes exhibiting characteristics of
Rima, B.K., Duffy, N., Mitchell, W.J., Summers, B.A., Appel, M.J., 1991. natural killer cells. Vet. Immunol. Immunopathol. 153, 249–259.
Correlation between humoral immune responses and presence of Sun, J.C., Lanier, L.L., 2009. Natural killer cells remember: an evolutionary
virus in the CNS in dogs experimentally infected with canine distem- bridge between innate and adaptive immunity? Eur. J. Immunol. 39,
per virus. Arch. Virol. 121, 1–8. 2059–2064.
Ringler, S.S., Krakowka, S., 1985. Effects of canine distemper virus Tipold, A., Vandevelde, M., Wittek, R., Moore, P., Summerfield, A., Zur-
on natural killer cell activity in dogs. Am. J. Vet. Res. 46, 1781– briggen, A., 2001. Partial protection and intrathecal invasion of CD8(+)
1786. T cells in acute canine distemper virus infection. Vet. Microbiol. 83,
Rossini, G., Cerboni, C., Santoni, A., Landini, M.P., Landolfo, S., Gatti, D., 189–203.
Gribaudo, G., Varani, S., 2012. Interplay between human cytomega- Vidal, S.M., Khakoo, S.I., Biron, C.A., 2011. Natural killer cell responses
lovirus and intrinsic/innate host responses: a complex bidirectional during viral infections: flexibility and conditioning of innate immu-
relationship. Mediators Inflamm. 2012, 607276. nity by experience. Curr. Opin. Virol. 1, 497–512.
Sarute, N., Calderon, M.G., Perez, R., La Torre, J., Hernandez, M., Francia, L., Zhang, M., Daniel, S., Huang, Y., Chancey, C., Huang, Q., Lei, Y.F., Grinev, A.,
Panzera, Y., 2013. The fusion protein signal-peptide-coding region of Mostowski, H., Rios, M., Dayton, A., 2010. Anti-West Nile virus activity
canine distemper virus: a useful tool for phylogenetic reconstruction of in vitro expanded human primary natural killer cells. BMC Immu-
and lineage identification. PLOS ONE 8, e63595. nol. 11, 3.

Please cite this article in press as: Park, J.-Y., et al., The anti-canine distemper virus activities of ex vivo-expanded canine
natural killer cells. Vet. Microbiol. (2015), http://dx.doi.org/10.1016/j.vetmic.2015.01.021