Role of Specific Cell-Mediated Cytotoxicity in Protecting Fish from Viral Infections

Virology 297, 120–127 (2002) doi:10.1006/viro.2002.1486

Tomonori Somamoto,* Teruyuki Nakanishi,† and Nobuaki Okamoto*,1

*Department of Aquatic Biosciences, Tokyo University of Fisheries, Konan 4, Minato, Tokyo 108-8477, Japan; and †Laboratory of Fish Pathology, Department of Veterinary Medicine, College of Bioresource Sciences, Nihon University, 1866 Kameino, Fujisawa, Kanagawa 252-8570, Japan Received November 27, 2001; returned to author for revision March 5, 2002; accepted March 28, 2002

We report the in vivo role of specific cytotoxic cells in protecting fish from acute viral infections. Specifically,we found that (1) there is an inverse relationship between cytotoxic activities and viral load and (2) adoptive transfer of immune leukocytes prevented viral infection. Crucian carp hematopoietic necrosis virus (CHNV),which has a virulence to ginbuna crucian carp, was recently found and identified in Japan. Specific cell-mediated cytotoxicity of ginbuna leukocytes against CHNV-infected syngeneic cells was induced as a result of intraperitoneal inoculation with CHNV. This cytotoxicity was not induced against either virus-infected allogeneic cells or eel rhabidovirus- (EVA) infected syngeneic cells. In these respects,the cytotoxic activity was similar to that of mammalian cytotoxic T-lymphocyte (CTL) activity. Viral titers of tissues from infected fish were remarkably reduced 8 days after infections,when specific cytotoxic activity reached a peak. This result suggested that specific cytotoxic cells were responsible for the early control of CHNV replication. On the other hand,CHNV-specific antibody was greatly increased when the virus was eliminated by cytotoxic activities. The effectiveness of the virus-specific cytotoxicity was evaluated using adoptive cell transfer. Recipients that received leukocytes from immune syngeneic donors escaped CHNV infection. These findings suggest that virus-specific cytotoxic cells have a role in controlling viral infections in a fish. © 2002 Elsevier Science (USA) Key Words: fish; specific cell-mediated cytotoxicity; protection from viral infection; adoptive transfer.

INTRODUCTION are present in fish and that they have functional similar- ities with those of higher vertebrates (Fischer et al., 1998; Although viral infections are the most serious disease Hasegawa et al., 1998; Manning and Nakanishi, 1996; in many fish species (Wolf, 1988), the interaction between Nakanishi et al., 2002; Stuge et al., 2000). Cytotoxicity the host defense and viruses is not clearly understood. against virus-infected autologous cells has been already Specific cellular and humoral immunities to viruses are reported in catfish, and it is shown that the cytotoxic cells believed to be present in fish, but the only response that are distinct from nonspecific cytotoxic cells (NCC) or has been characterized in detail so far has been the alloreactive cytotoxic cells (Hogan et al., 1997). Mckinney antibody response (Lorenzen and LaPatra, 1999; Loren- and Schmale demonstrated that damselfish with neuro- zen et al., 1999; Olsen and Jørgensen, 1986; Olsen et al., fibromatosis exhibit cytotoxicity against autologous cells 1991). In rainbow trout infected with salmonid rhabdovi- infected with retrovirus (Mckinney and Schmale, 1997). In ruses [infectious hematopoietic necrosis virus (IHNV) addition, we recently reported that specific cytotoxicity and viral hemorrhagic septicaemia virus (VHSV)], neu- against virus-infected cells was induced by in vivo im- tralizing serum activities of rainbow trout were detect- munization using clonal ginbuna crucian carp and syn- able after the appearance of clinical infection (Hatten- berger-Baudouy et al., 1989; Lorenzen et al., 1999). This geneic cell lines, suggesting that virus-specific CTL is suggests that antibodies to viruses have a protective inherent in fish (Somamoto et al., 2000). These findings effect during the latter stages of viral infections. implied that fish, as well as mammals, might have spe- Mammalian cytotoxic T-lymphocyte (CTL) responses cific cell-mediated cytotoxicities that provide an impor- provide a major defense mechanism for elimination of tant role in antiviral defence systems. virus-infected cells (Oldstone, 1987; Zinkernagel and Isogeneic ginbuna crucian carp is a naturally occur- Doherty, 1979) and in some cases CTL have been able to ring gynogenetic fish that has potential as a model for confer complete protection, even in the absence of an studying fish immunology (Nakanishi and Okamoto, antibody response (Bevan, 1989; Luckacher et al., 1984). 1999). Recently crucian carp hematopoietic necrosis vi- The results of some studies strongly suggest that CTL rus (CHNV), which is a negative-strand RNA virus belonging to the rhabdovirus family, was identified in Japan (H. Fukuda, unpublished data). This virus has a virulence 1 To whom correspondence and reprint requests should be ad- to ginbuna crucian carp and can propagate in two cell dressed. Fax: ϩ81-3-5463-0552. E-mail address: nokamoto@tokyo-u- lines (CFS and CFK) derived from two strains of ginbuna fish.ac.jp. crucian carp. Thus, a system consisting of this virus,

6 FIG. 1. Cell-mediated cytotoxicity of leukocytes from ginbuna crucian carp against CHNV-infected CFS cells after infection with 10 TCID50/ml of CHNV. Secondary infection was carried out 3 weeks after the primary infection. Peripheral blood leukocytes (PBL) and kidney leukocytes (KL) collected on the selected days were incubated with target cells (CHNV-infected CFS cells) at an effector:target ratio of 100:1. The results are shown as means Ϯ SD of three fish. Asterisks indicate significant difference between cytotoxicities against uninfected and CHNV-infected CFS cells at *P Ͻ 0.05 and **P Ͻ 0.01. (E and ᮀ) uninfected CFS cells; (F and ■) CHNV-infected CFS cells.

isogeneic ginbuna crucian carp, and their cell lines will cantly higher than those in the primary infection (PBLs, provide a good model to analyze host–virus interactions P Ͻ 0.01; KLs, P Ͻ 0.05). in fish, such as those that occur in in vivo and in vitro cell-mediated immunity. Cytotoxicity of immune leukocytes against CHNV- In this study, we examined the in vivo role of specific infected allogeneic cells and eel rhabidovirus-infected cytotoxic cells in protecting fish from acute viral infec- syngeneic cells tions through examining the association of cytotoxic ac- Effector cells from fish infected twice with CHNV lysed tivity with elimination of virus and the effect of adoptive CHNV-infected syngeneic cells (CFS cells) more strongly transfer of immune leukocytes on CHNV infection. than uninfected CFS cells (P Ͻ 0.01) (Fig. 2A). On the other hand, no difference in the cytotoxic activity was RESULTS observed between the uninfected allogeneic cells (CFK cells) and the CHNV-infected ones. Specific cell-mediated cytotoxicity against CHNV- Although the cytotoxic activity of CHNV-sensitized ef- infected syngeneic cells fector cells against CHNV-infected CFS cells was signif- Leukocytes obtained from the fish with primary CHNV icantly higher than those of effector cells from the fish Ͻ infection were capable of killing CHNV-infected CFS injected without viruses (P 0.01), no difference in the cells (Fig. 1A). Secondary infection was more efficient in activity against EVA-infected CFS cells was observed inducing specific cytotoxicity against CHNV-infected CFS between both effector cells (Fig. 2B). These results sug- cells (Fig. 1B). Spontaneous cytotoxicity of nonimmune gest that sensitized effector cells specifically recognized effector cells against CHNV-infected CFS cells as a con- viral antigens on syngeneic cells. The same results were trol was not detected. In primary infection, cytotoxic ac- obtained in effector cells from the fish infected once with tivities of effector cells against uninfected target cells CHNV (data not shown). and virus-infected ones were significantly different 4 and The neutrophil-rich fraction (1.08–1.09 g/ml) from kid- 8 days after infection in kidney leukocytes (KLs) and ney of immune fish did not kill virus-infected syngeneic peripheral blood leukocytes (PBLs) respectively, and in cells (data not shown). secondary infection, these were recognized 4 days after Time course of the cytotoxic activity,antibody infection in KLs and PBLs. The cytotoxicity of KLs from response,and viral replication in fish infected with primary infected fish reached a peak 8 days after infec- CHNV tion, whereas the cytotoxicity of KLs from secondary infected fish reached a peak 4–6 days after infection. The Time course of the cytotoxic activity, antibody re- highest activity was detected in the PBLs from infected sponse and viral replication in fish infected with CHNV is 7 fish sampled 8 days after infection in both the primary shown in Fig. 3. Viral titers of more than 10 TCID50/ml and secondary infections. And the highest activities of were observed in kidney and liver of fish sampled 2 days PBLs and KLs in the secondary infection were signifi- after infection, showing that CHNV propagated in vivo. 122 SOMAMOTO, NAKANISHI, AND OKAMOTO

FIG. 2. The cytotoxicity of immune leukocytes against CHNV-infected allogeneic cells and EVA-infected syngenic cells. (A) Cytotoxic activity against CHNV-infected CFS cells (syngeneic) or CHNV-infected CFK cells (allogeneic). (B) Specificity of cytotoxic activity against CHNV-infected (homologous) or EVA- (heterologous) infected CFS cells. Peripheral blood leukocytes were collected 8 days after secondary infection with CHNV. Secondary injection was carried out 3 weeks after the primary infection. Effector cells were incubated with target cells at an effector:target ratio of 100:1. The results are shown as means Ϯ SD of three fish. Asterisks indicate significant difference at **P Ͻ 0.01.

Virus titers gradually decreased until 8 days after infec- increased and reached a peak 8 days after infection. tion and were remarkably reduced 12 days after infec- This suggested that the remarkable reduction of the virus tion. On the other hand, the cytotoxic activity gradually in the period between 8 and 12 days after infection was

FIG. 3. Time course of cell-mediated cytotoxic response, antibody (IgM) response, and viral replication in fish infected with CHNV. Effector cells were collected from blood or kidney on the selected days and incubated with CHNV-infected syngeneic cells at an effector:target ratio of 100:1. Virus titers are shown as means Ϯ SD of four fish. Data of cytotoxic and antibody responses are shown as means of three fish (SD in cytotoxicity Ͻ 4.5%, in antibody titer Ͻ 0.048). (F) Lysis of target cells by PBL. (■) Lysis of target cells by kidney leukocytes. (ᮀ) Anti-CHNV antibody IgM quantified by ELISA. ROLE OF VIRUS-SPECIFIC CYTOTOXICITY IN FISH 123

TABLE 1 Effect of Adoptive Transfer of Syngeneic Leukocytes on Acute Illness 3 Days after Infection with CHNV

a b c Leukocytes transferred Ϫ ϩ ϩϩ ϩϩϩ Virus titer (Log10TCID50/ml)

From nonimmunized fish 0 0 2 2 8.2 Ϯ 0.4 From immunized fish Collected 8 days after viral immunization 2 2 0 0 6.6 Ϯ 0.6* Collected 16 days after viral immunization 0 0 3 1 7.8 Ϯ 0.6 Without leukocytes (physiological saline) 0 0 2 2 8.1 Ϯ 0.2

a Syngeneic leukocytes. b Number of fish with illness. Scored as follows: Ϫ, no clinical sign; ϩ, presence of hemorrhage in fins and muscles; ϩϩ, plus ascitic fluid in abdominal cavity; ϩϩϩ, plus severe anemia. (Hematocrit values of fish with anemia were less than 7%, compared with more than 25% in normal.) c Virus titres of a mixed sample with kidney and liver from CHNV-infected fish. An asterisk indicates significant difference at P Ͻ 0.01 compared to the virus titers of the other recipients. The results are shown as means Ϯ SD. caused by enhancement of cytotoxicity. In contrast, spe- fish sampled 8 days after both of the primary and sec- cific IgM antibody titers were increased from 12 days ondary infection. These kinetics of CHNV-specific cyto- after infection, when cytotoxic activity started decreas- toxicity are similar to the CTL response described for ing. The highest antibody titer was observed 30 days avian and murine viruses (Seo and Collisson, 1997; after infection, when no viral titers were detected. Tashiro et al., 1988; Zinkernagel and Doherty, 1974). It is interesting that the kinetics between blood and kidney Effect of adoptive transfer of immune leukocytes on leukocytes differed. This tendency has also been ob- CHNV infection served in alloantigen-specific cytotoxicity of ginbuna cru- All fish that received nonimmune leukocytes showed cian carp (Fischer et al., 1998). Effector cells from blood severe clinical signs and had virus titers that were the consisted almost completely of lymphocytes and throm- same as those of fish that received only physiological saline (Table 1). Immune leukocytes from fish sampled 8 days after infection showed protective effects against illness with significant reduction of virus titers (P Ͻ 0.01) in the recipients. In contrast, immune leukocytes from donors sampled 16 days after infection did not affect the course of illness or viral load in the recipients. CHNV-specific antibodies from the plasma of all recipients with donor cells were the same as those from the nonsensitized fish (Fig. 4).

DISCUSSION The preceding results show that specific cytotoxic cells have a role of in controlling viral infections in a fish. This finding was based on the association of cytotoxic responses with elimination of virus and the adoptive transfer of immune leukocytes. Specific cell-mediated cytotoxicity of ginbuna leukocytes against CHNV-infected syngeneic cells was induced as a result of intraperitoneal inoculation of a lethal dose of CHNV. Our results show that the cytotoxicity was not induced against either virus-infected allogeneic cells FIG. 4. Detection of CHNV-specific IgM of recipients on ELISA ab- or eel rhabidovirus- (EVA) infected syngeneic cells. This sorbance. The results are shown as means of four fish (SD Ͻ 0.15). suggests that specific cell-mediated cytotoxicity of gin- Specific antibody titers of recipients were measured 3 days after buna may be restricted to certain allotypes, in agreement infection with CHNV. Donor cells that recipients received are indicated F E with CTL activity in higher vertebrates. Further studies as follows: ( ) immune leukocytes (8 days after infection); ( ) immune leukocytes (16 days after infection); (Œ) None (physiological saline); (‚) are needed to clarify the allelic diversity of histocompat- nonimmune leukocytes; (■) positive control (the plasma was obtained ability antigens and genetic restriction of cytotoxicity. The from fish sampled 16 days after the secondary infection); and (ᮀ) highest activity was detected in the PBL from infected negative control. The plasma was obtained from uninfected fish. 124 SOMAMOTO, NAKANISHI, AND OKAMOTO bocytes (Ͼ95%), and those from kidney leukocytes con- cells was demonstrated in catfish and damselfish sisted of 80% lymphocytes. Moreover, we tested the (Hogan et al., 1997; Mckinney and Schmale et al., 1997), specific cytotoxicity of sIg-positive (B-cell) or sIg-nega- and they described that cytotoxic cells, which kill virus- tive PBLs obtained from immune fish by MAC magnetic infected target, are distinct from nonspecific cytotoxic cell sorting (data not shown). We found that the cytotox- cells (NCC) or alloreactive cytotoxic cells. Our data icity of sIg-negative PBLs against CHNV-infected CFS showed the cytotoxic activity against not only virus-in- cells was higher than that of sIg-positive PBLs. This fected syngeneic cells but also allogeneic cells, suggest- suggests that the effector cells that possessed specific ing that the effector cells may involve alloreactive CTL or cytotoxicity were sIg-negative lymphocytes. Therefore, NK-like cells. The previous report also showed that non- these results show the cytotoxic activity obtained in this sensitized lymphocytes from catfish possess the cytotox- study was similar to the mammalian CTL activity. icity against allogeneic cells (Yoshida et al., 1995). Mono- The viral titer of infected fish dropped dramatically clonal antibodies against lymphocyte subset in ginbuna after the peak of cytotoxic responses, suggesting that the crucian carp are probably necessary to better under- cytotoxic activity was responsible for controlling CHNV stand specific cell-mediated cytotoxicity. And also, since replication between 8 and 12 days after infection. This T-cell receptor (TCR) and CD8 genes have already been result indicates that the fish immune system may utilize found in several fish species (Hansen and Strassburger, specific cytotoxic cells to eliminate CHNV at the early 2000; Rast and Litman, 1998), these genes from ginbuna stage of acute infections. In contrast, the IgM antibody crucian carp should be identified in the future. titer was greatly increased after decreasing to viral titers Only a few vaccines against fish viral diseases are less than 102 tissue culture infection dose, 50% endpoint available at present (Gudding et al., 1999). Because our assay (TCID50)/ml. Antibody neutralization in fish immune results suggest that virus-specific cytotoxic cells have a systems may control the later stage of viral infections. role of in controlling viral infections in a fish, a vaccine Together, these observations suggest that virus-specific able to stimulate specific cytotoxic cells may confer pro- cytotoxic cells in fish plays an important role in protect- tection against virus infections. In addition, testing ing fish from acute viral infections. This finding probably whether viral proteins act as inducers of cytotoxic activ- can be extrapolated to other fishes because some stud- ities should aid in the development of effective vaccines. ies have reported that neutralizing serum activity from rainbow trout appeared too late to play any role in pro- MATERIALS AND METHODS tection against acute salmonid rhabdovirus diseases Fish and cell lines (Hattenberger-Baudouy et al., 1989; Lorenzen et al., 1999). Clonal triploid ginbuna crucian carp (Carassius aura- The production of interferon (IFN) provided an early tus langsdorfii), an isogeneic strain from Lake Suwa in and effective protection against viral infection. Mx pro- the Nagano prefecture (S3n clone), were obtained from teins, which are IFN-inducible proteins, have been found the National Research Institute of Aquaculture, Japan. in trout, where they are synthesized in response to viral The fish, weighing 40–60 g for preparation of effector infection (Trobridge and Leong, 1995; Trobridge and cells, were maintained with running water at a temper- Leong, 1997). Furthermore, a recent report showed that ature of 25 Ϯ 1°C. The fish were fed daily with commer- DNA vaccines encoding viral glycoproteins induce Mx cial pellets. protein synthesis in fish (Kim et al., 2000). This suggests CFS and CFK cells, derived from the S3n and K1 that there is more to learn about the relationship be- strains of ginbuna crucian carp, respectably, were estab- tween cell-mediated cytotoxicity and IFN response in the lished by Hasegawa et al. (1997a,b) and maintained as early stage of virus infections in fish. described by the authors. The origin of these cell lines Adoptive cell transfer of immune lymphocytes has was confirmed with DNA fingerprinting (Hasegawa et al., been shown to be effective in protecting mammals and 1998). These cells were highly sensitive to the viruses birds from viral infections (Cannon et al., 1987; Klavinskis used in this study. et al., 1989; Offit and Dudzik, 1990; Seo et al., 2000). Our Virus findings that transferred immune leukocytes had the ca- pacity to protect fish from viral infections thus extend this Crucian carp hematopoietic necrosis virus, belonging phenomenon to fish. Our data showed that only neutral- to rhabdoviridae, was obtained from the Laboratory of izing serum activity resulting from the transfer was ex- Fish Pathology, Tokyo University of Fisheries. CHNV was cluded as a mechanism for protection because CHNV- inoculated into the CFS cells grown in Eagle’s minimal specific IgM was not detected in the plasma of any of the essential medium (MEM, Nissui) with 2% fetal bovine recipients. serum (FBS, Commonwealth). An eel rhabdovirus, EVA The characteristics of the cell types that are respon- (Sano, 1976), was used as an antigen unrelated to the sible for controlling viral infections remain to be resolved immunogen. The titers of each virus were determined by in our study. Cytotoxicity against virus-infected target TCID50. ROLE OF VIRUS-SPECIFIC CYTOTOXICITY IN FISH 125

CHNV antigens for enzyme linked immunosorbent as- Cytotoxic assays say (ELISA) were purified by the method of Fukuda et al. Cytotoxicity was assayed by the method of Somamoto (1989) with slight modifications. CHNV was suspended in et al. (2000). Target cells were seeded in 96-well, flat- 6.5% polyethylene glycol (PEG-6000) and 2.3% NaCl and bottom microtiter plates (Corning) at 104 cells/well and agitated at 4°C for 3 h. After centrifugation at 4,000 g for allowed to settle down in the wells for 6 h. Then 20 ␮lof 20 min, the pellet was suspended in TNE [0.01 M Tris, 0.1 ␮ 51 MEM-10 containing 2 Ci of Na2 CrO4 (NEN) was added M NaCl, and 1 mM EDTA (pH 7.6)]. The suspension was to each well and the plates were incubated overnight at centrifuged at 20,000 g for 150 min. The pellet was 25°C. After washing three times with OPTI-MEM I, the resuspended in TNE, overlaid on 20, 40, and 55% su- cells were infected with a virus at 25°C for 5–6 h (m.o.i. crose steps and centrifuged for 1.5 h at 80,000 g at 4°C. ϭ 10). Then, the cells were washed three times with Virus bands were harvested and diluted in PBS. OPTI-MEM. Effector cells were added to each well in a final volume of 200 ␮l. After incubation at 25°C for 6 h, supernatants (200 ␮l) were harvested with an automated Viral replication in the infected fish supernatant collection device (Skatron) and radioactivity S3n ginbuna crucian carp were infected with 106 was measured using the COBRA II autogamma counting

TCID50/ml of CHNV by intraperitoneal (i.p.) injection. All system (Packard Instruments). Effector cells from each fish survived during this test. Tissues (kidney, liver, ovary, individual were divided and tested against uninfected and spleen) from the fish were collected at various days and virus-infected target cells. Radioactivity of the super- after CHNV infection, weighed, and homogenized. A 10% natants from target cells without effector cells and from weight/volume tissue suspension was prepared in MEM target cells lysed with detergent (10% 7ϫ, ICN Biomedi- containing 2% FBS. Tissue suspensions were centri- cals) served as the spontaneous and maximum release fuged at 4000 g for 20 min and the supernatants were controls, respectively. Spontaneous release was 8–16% filtered through 0.45-␮m membrane filters. Virus titers of of maximum release. Percentage of cytotoxicity was cal- culated as follows: the supernatants were determined by TCID50. Test release Ϫ Spontaneous release ϫ 100. Preparation of effector cells Maximum release Ϫ Spontaneous release S3n ginbuna crucian carp were infected with 106 ELISA to detect the antibody against CHNV TCID50/ml of CHNV by i.p. injection. The secondary infection was carried out 3 weeks after the primary infec- CHNV antigens were coated onto 96-well Polysorp tion. Effector cells were prepared from blood or kidney immunoplates (Nunc) by adding 50 ␮l of the purified on the days shown in the figure legends. Fish were bled CHNV diluted in carbonate–bicarbonate coating buffer from the caudal vein into heparinized syringes. Blood [(15mMNa2CO, 35 mM NaHCO3,3mMNaN3 (pH 9.6)] at samples from clonal ginbuna were overlaid onto a Per- a concentration of 10 ng/well, and the plates were incu- coll (Pharmacia) density gradient of 1.085 g/ml and cen- bated overnight at 4°C before being washed four times trifuged at 350 g for 20 min at 4°C to separate peripheral with phosphate-buffered saline (PBS)–0.05% Tween 20. ␮ blood leukocytes. The kidney cell suspension was pre- Each well was blocked with 100 l of 1.5% bovine serum pared by gently pressing the tissues through a 150- albumin (Sigma) for2hatroom temperature. After wash- gauge mesh stainless steel sieve in OPTI-MEM I ing four times with PBS–0.05 Tween 20, the plasma prepared from tested fish and diluted in PBS (pH 7.2) was (GIBCO). The cell suspension was washed with OPTI- added to each well. The plates were incubated for2hat MEM I. Washed cells were then applied to a Percoll room temperature and washed four times with PBS–0.05 density gradient of 1.08 g/ml and centrifuged at 350 g for Tween 20. One hundred microliters of mouse antigin- 20 min at 4°C. The collected leukocytes were stained buna IgM monoclonal antibody diluted in PBS (1:10,000) with Giemsa solution. The composition of the leukocyte was added, and the plates were incubated for 1.5 h at populations was determined by examining a total of 1000 room temperature and were again washed four times cells. Consequently, PBL as effector cells consisted al- with PBS–0.05 Tween 20. One hundred microliters of most completely of lymphocytes (75–86%) and thrombo- alkaline phosphate-conjugated goat antimouse IgM and cytes (9–23%), and kidney leukocytes consisted of ap- IgG (Biosource) was added. The plates were incubated proximately 80% lymphocytes. These collected leuko- for1hatroom temperature and washed four times with cytes were washed twice with OPTI-MEM I and finally PBS–0.05 Tween 20. One hundred microliters of sub- suspended in OPTI-MEM I and supplemented with 10% strate buffer in p-nitrophenyl phosphate (pNPP; GIBCO) FBS. Viability of the cells, which thereafter served as was added to each well, and the plates were incubated effector cells, was greater than 95% by the trypan blue for 20 min in the dark. The absorbance was read at 405 dye exclusion method. nm using a microplate reader (LabSystems). 126 SOMAMOTO, NAKANISHI, AND OKAMOTO

Adoptive transfer of PBLs sensitized with CHNV cytotoxicity against allogeneic erythrocytes in ginbuna crucian carp and goldfish using a non-radioactive assay. Dev. Comp. Immunol. 22, Donor fish to confirm antiviral activity, weighing 40–60 195–206. 6 g, were infected with 10 TCID50/ml of CHNV, and the Fukuda, H., Shuu, S., Ban, F., Nishimura, T., and Sano, Y. (1989). Con- PBLs were collected from CHNV-infected fish 8 days and centration purification of IHN virus. Fish Pathol. 24, 57–58. 16 days after the secondary infection, according to the Gudding, R., Lilehaug, A., and Evensen, Ø. (1999). Recent development in fish vaccinology. Vet. Immunol. Immunopathol. 72, 203–212. method of separation of effector cells. The reason we Hansen, J. D., and Strassburger, P. (2000). Description of an ectothermic selected these days is based on the data shown in Fig. TCR coreceptor, CD8 alpha, in rainbow trout. J. Immunol. 15, 3132– 1. These donor cells from infected fish did not contain the 3139. virus because no cytopathogenic effect (CPE) appeared Hasegawa, S., Nakayasu, C., Okamoto, N., Nakanishi, T., and Ikeda, Y. on the cell culture (data not shown). As a control, donor (1997a). Fin cell line from isogeneic ginbuna crucian carp. In Vitro cells from uninfected fish were prepared. Cell Dev. Biol. (Animal) 33, 232–233. Hasegawa, S., Somamoto, T., Nakayasu, C., Nakanishi, T., and Oka- Each group of naive fish, weighing 20–28 g, consisted moto, N. (1997b). A cell line (CFK) from fin of isogeneic ginbuna of 4 fish. Peripheral blood leukocytes pooled from 10 crucian carp. Fish. Pathol. 32, 127–128. virus-infected or 10 uninfected donors were equally di- Hasegawa, S., Nakayasu, C., Yoshitomi, T., Nakanishi, T., and Okamoto, vided and transferred to 4 recipients by i.p. injection (4 ϫ N. (1998). Specific cell-mediated cytotoxicity against an allogeneic 10 7 cells/0.5 ml/fish). Donor cells consisted of lympho- target cell line in isogeneic ginbuna crucian carp. Fish Shellfish Immunol. 8, 303–313. cytes (83–86%) and thrombocytes (6–8%). One group re- Hattenberger-Baudouy, A. M., Danton, M., Merle, G., Torchy, C., and de ceived the same amount of 0.85% NaCl in the absence of Kinkelin, P. (1989). Serological evidence of infectious hematopoietic leukocytes. One day after receiving the donor cells, the necrosis in rainbow trout from French outbreak of disease. J. Aquat. 8 recipients were challenged with 10 TCID50/ml of CHNV Animal Health 1, 126–134. and subsequently observed for clinical signs of illness Hogan, R. J., Stuge, T. B., Clem, L. W., and Miller, N. W. (1996). Anti-viral such as petechial hemorrhage in the fins and muscles, cytotoxic cells in the channel catfish (Ictalurus punctatus). Dev. Comp. Immunol. 20, 115–127. ascitic fluid in the abdominal cavity, and severe anemia. Kim, C. H., Johnson, M. C., Drennan, J. D., Simon, B. E., Thomann, E., and Kidney and liver from the recipients were collected and Leong, J. C. (2000). DNA vaccine encoding viral glycoprotein induce mixed at 3 days after CHNV infection, weighed, and nonspecific immunity and Mx protein synthesis in fish. J. Virol. 74, homogenized. A 10% weight/volume tissue suspension 7048–7054. was prepared in MEM containing 2% FBS. Tissue sus- Klavinskis, L. S., Tishon, A., and Oldstone, M. B. (1989). Efficiency and effectiveness of cloned virus-specific cytotoxic T lymphocytes in vivo. pensions were centrifuged at 4000 g for 20 min and the ␮ J. Immunol. 15, 2013–2016. supernatants were filtered through 0.45- m membrane Lorenzen, N., and LaPatra, S. E. (1999). Immunity to rhabdoviruses in filters. Virus titers of the supernatants were determined rainbow trout: The antibody response. Fish Shellfish Immunol. 9, by TCID50. Plasmas from recipients were collected 3 345–360. days after infection, and specific antibody titers were Lorenzen, N., Olsen, N. J., and Koch, C. (1999). Immunity to VHS virus in detected by ELISA. rainbow trout. Aquaculture 172, 41–61. Lukacher, A. E., Braciale, V. L., and Braciale, T. J. (1984). In vivo effector function of influenza virus-specific cytotoxic T lymphocyte clones is Statistical analysis highly specific. J. Exp. Med. 160, 814–826. The results are expressed as means Ϯ standard de- Manning, M. J., and Nakanishi, T. (1996). The specific immune system: Cellular defenses. In “Fish Physiology XV: The Fish Immune System viation of the mean (SD). Differences in the level of (G. K. Iwama and T. Nakanishi, Ed.), pp. 159–205. Academic Press, cytotoxic activities were determined by a paired or an London. unpaired t test. Comparison between virus titers of the Mckinney, E. C., and Schmale, M. C. (1997). Damselfish with neurofi- recipients was analyzed using one-way analysis of vari- bromatosis exhibit cytotoxicity toward retrovirus infected cells. Dev. ance (ANOVA). Difference was considered significant Comp. Immunol. 21, 287–289. Ͻ Nakanishi, T., and Okamoto, N. (1999). Cell-mediated cytotoxicity in when P 0.05. isogeneic ginbuna crucian carp. Fish Shellfish Immunol. 9, 259– 267. ACKNOWLEDGMENTS Nakanishi, T., Fischer, U., Dijkstra, J. M., Hasegawa, S., Somamoto, T., Okamoto, N., and Ototake, M. (2002). Cytotoxic T cell function in fish. This study supported in part by Grants-in-aid for Scientific Research Dev. Comp. Immunol. 26, 131–139. from the Ministry of Education, Science, Sports and Culture, Japan (No. Offit, P. A., and Dudzik, K. L. (1990). Rotavirus-specific cytotoxic T 12460088) and the Ministry of Agriculture, Forestry of Japan. lymphocytes passively protect against gastroenteritis in suckling mice. J. Virol. 64, 6325–6328. REFERENCES Oldstone, M. B. (1987). Immunotherapy for virus infection. Curr. Top. Microbiol. Immunol. 134, 211–229. Bevan, M. J. (1989). Stimulating killer cells. Nature 30, 478–479. Olsen, N. J., and Jørgensen, P. E. V. (1986). Detection of rainbow trout Cannon, M. J., Stott, E. J., Taylor, G., and Askonas, B. A. (1987). Clear- antibody to Egtved virus in rainbow trout, Salmo gairdneri, by plaque ance of persistent respiratory syncytial virus infections in immuno- neutralisation test with component additions. J. Applied Ichthyol. 2, deficient mice following transfer of primed T cells. Immunology 62, 33–41. 133–138. Olsen, N. J., Lorenzen, N., and Jørgensen, P. E. V. (1991). Detection of Fischer, U., Ototake, M., and Nakanishi, T. (1998). In vitro cell-mediated rainbow trout antibody to Egtved virus by enzyme-linked immunosor- ROLE OF VIRUS-SPECIFIC CYTOTOXICITY IN FISH 127

bent assay (ELISA), immunofluorescence (IF) and plaque neutralisa- ated immunity induced in mice after vaccination with a protease tion test. Dis. Aquat. Organism 10, 31–38. activation mutant, TR-2, of Sendai virus. J. Virol. 62, 2490–2497. Rast, J. P., and Litman, G. W. (1998). Towards understanding the evolu- Trobridge, G. D., and Leong, J. C. (1995). Characterization of a rainbow tionary origins and early diversification of rearranging antigen recep- trout Mx gene. J. Interferon Cytokine Res. 15, 691–702. tors. Immunol. Rev. 166, 79–86. Trobridge, G. D., and Leong, J. C. (1997). Induction of the Mx protein of Sano, T. (1976). Viral diseases of cultured fish in Japan. Fish. Pathol. 10, rainbow trout, Oncorhynchus mykiss, in vitro and in vivo with poly IC 221–226. and infectious hematopoietic necrosis virus. Dis. Aquat. Org. 30, Seo, S. H., and Collisson, E. W. (1997). Specific cytotoxic T lymphocytes 91–98. are involved in vivo clearance of infectious bronchitis virus. J. Virol. Wolf, K. (1988). “Fish Viruses and Fish Viral Diseases.” Cornell Univ. 71, 5173–5177. Press, New York. Seo, S. H., Pei, J., Briles, W. E., Dzielawa, J., and Collisson, E. W. (2000). Yoshida, S. H., Stuge, T. B., Miller, N. W., and Clem, L. W. (1995). Adoptive transfer of infectious bronchitis virus primed alphabeta T Phylogeny of lymphocyte heterogeneity: Cytotoxic activity of channel cells bearing CD8 antigen protects chicks from acute infection. catfish peripheral blood leukocytes directed against allogeneic tar- Virology 30, 183–189. Stuge, T. B., Wilson, M. R., Zhou, H., Barker, K. S., Bengten, E., Chinchar, gets. Dev. Comp. Immunol. 19, 71–77. G., Miller, N. W., and Clem, L. W. (2000). Development and analysis of Zinkernagel, R. M., and Doherty, P. C. (1974). Restricted of in vitro T various clonal alloantigen-dependent cytotoxic cell lines from chan- cell-mediated in lymphocytic choriomeningitis within a syngeneic or nel catfish. J. Immunol. 164, 2971–2977. semiallogeneic system. Nature 248, 701–702. Somamoto, T., Nakanishi, T., and Okamoto, N. (2000). Specific cell- Zinkernagel, R. M., and Doherty, P. C. (1979). MHC-restricted cytotoxic mediated cytotoxicity against a virus-infected syngeneic cell line in T cells: Studies on the biological role of polymorphic major trans- ginbuna crucian carp. Dev. Comp. Immunol. 24, 633–640. plantation antigens determining T-cell restriction-specificity, function. Tashiro, M., Fujii, Y., Nakamura, K., and Homma, M. (1988). Cell-medi- Adv. Immunol. 27, 51–177.