Antibodies Produced by Rabbits Immunized

INFECTION AND IMMUNITY, Dec. 1971, p. 715-719 Copyright 1971 American Society for Microbiology Vol. 4, No. 6 Printed in U.S.A. Antibodies Produced by Rabbits Immunized ith Visna Virus SEUNG C. KARL AND HALLDOR THORMAR Ne York State Institute for Research in Mental Retardation, Staten Island, Ne York 1314 Received for publication 21 June 1971 Immunization of Ne Zealand White rabbits ith purified visna virus elicited antibody activity demonstrated by passive hemagglutination (PHA), complement fixation (CF), and indirect immunofluorescent tests. The antibody activities of hyperimmune sera and Sephadex G-2 fractions of the sera ere studied. It as found that the PHA test as 1 to 1 times more sensitive than the CF test in detecting visna antibodies in rabbits. It as also found that the immunoglobulin M fractions from Sephadex G-2 filtration displayed greater PHA activity than did the immunoglobulin G fractions. Although neutralizing antibody as demonstrated in the serum of the natural host (sheep), our attempts to demonstrate neutralizing antibody in the sera from hyperimmunized rabbits (non-natural host) so far have failed. In its natural host, the sheep, visna virus causes the formation of neutralizing (9) and complement-fixing antibodies (1) and of antibodies detectable by fluorescein labeling (2, 11). On the other hand, there have been no reports of successful hyperimmunization of other animal species ith visna virus. In fact, several attempts to inoculate small laboratory animals ith the virus have given negative results, since proliferation of virus or antibody response could not be detected (H. Thormar, unpublished data). In the present study, rabbits ere hyperimmunized ith visna virus by using repeated injections of high concentrations of purified virus. Complement fixation (CF) test and immunofluorescent techniques ere employed for demonstration of an antibody response. In addition, the passive hemagglutination (PHA) test as shon to be a sensitive method for assay of antibodies against visna virus. MATERIALS AND METHODS Virus. Visna virus strain K796 in its 17th tissue culture passage as used for immunization of rabbits and as antigen in serological tests. The virus as gron in monolayer cultures of sheep choroid plexus cells, as previously described (4). Infectious tissue culture fluid as harvested, and the virus as concentrated and purified by the method described by Lin and Thormar (5). Before use, the purified virus suspension as dialyzed against.15 M phosphate-buffered saline (PBS) at 4 C overnight. The titer of the final virus preparation used for immunization as 19 5% tissue culture infectious doses (TCID5o) per milliliter. Antisera. Ne Zealand White rabbits (5. to 6. lb.) ere given 1 ml of primary intravenous injection of virus. On days 7, 14, 22, 36, 5, 64, 78, and 92, they received intramuscularly 2 ml of a mixture of virus and complete Freund's adjuvant. Ten days after the last inoculation, blood as collected via heart puncture. The sera ere separated from the hole blood, placed in vials, and stored at -2 C. Antisera against a homogenate of normal choroid plexus cells and against bovine serum albumin ere prepared in the same manner. Unless otherise stated, all sera used in PHA and CF tests ere inactivated at 56 C for 3 min and absorbed ith ashed sheep red blood cells (RBC) until no hemagglutination (HA) reactions ere observed. The immune rabbit sera ere then absorbed ith sheep choroid plexus cells. A visna antiserum from a sheep as used for comparison ith the immune rabbit sera. The sheep had been inoculated intracerebrally ith a strain of visna virus about 4 months before bleeding. The serum as stored at -2 C for about 3 years and had thaed out occasionally during this period. HA test. PHA tests ere carried out according to Stavitsky (1). Briefly, sheep RBC ere ashed three times and suspended to 2.5% in PBS at ph 7.2. Equal volumes of cell suspension and 1:2, dilution of tannic acid in ph 7.2 PBS ere incubated for 1 min at 37 C. The cells ere ashed once ith ph 7.2 PBS and suspended to original volume in (ph 6.4) PBS. The resulting tanned cells ere mixed ith viral antigen and ith (ph 6.4) PBS in the proportion of 1:1:4, respectively. In the initial experiments, it as found that a 1:2 dilution of viral antigen gave satisfactory titers hen coated on sheep RBC. Therefore, a 1:2 dilution of virus as used in all PHA tests (Tables 1 and 2). The virus-cell mixture as incubated for 1 min at room temperature. After incubation, the cells ere ashed three times and suspended in 1% normal 715

716 KARL AND THORMAR INFEC. IMMUN. rabbit serum in ph 6.4 PBS. Serum dilutions ere made in tofold steps in.5-ml volumes, and.5 ml of the 2.5% sensitized sheep RBC prepared as described above as added to each tube. HA patterns ere read after 3 hr of incubation at room temperature. CF test. CF tests ere carried out by the microtiter technique described by Sever (7). All dilutions ere made ith Veronal-buffered saline, ph 7.2, containing 1 ml of 1% magnesium sulfate per liter. The folloing components ere mixed and placed at 4 C overnight:.25 ml of viral antigen,.25 ml of serum, and.5 ml (2 full units) of complement. After incubation,.5 ml of sensitized sheep RBC as added, and the mixture as incubated at 37 C for 3 min. CF titers ere determined under conditions hereby no anticomplementary activity could be detected. Immunofluorescence test. The indirect staining method as used throughout. Visna-infected cell monolayers on cover slips ere ashed three times ith PBS, air dried, and fixed in cold acetone for 1 min. The cover slips ere then covered ith rabbit immune sera or ith visna sheep serum and incubated in a moist chamber at room temperature for 6 min. After thorough ashing ith PBS, the monolayers ere stained for 3 min at room temperature ith a TABLE 1. Hemagglhtiniationi (HA), complementt fixation (CF), immunofluorescent (IF), and neutralizing (N) activities in the sera of rabbits immunized ith visna virus Serumā HA CF IF N AVS-6 1, 28b 64 NT NT AVS-9 2,48 > 256 64 <4 AVS-1 5,12 128 32 <4 AVS-11 64 64 NT < 4 AVS-32 32 NTc NT NT ACS-18 <4 < 8 NT NT ACS-19 <4 < 8 NT NT NRS-31 <4 <8 <8 <4 a AVS, anti-visna serum; ACS-18, anti-choroid plexus cell serum; ACS-19, anti-bovine serum albumin; NRS-31, normal rabbit serum. b Values expressed as reciprocal of the antibody titer in serum collected 1 days after the last injection. c Not tested. TABLE 2. Hemagglutiniationz (HA), complement fixationi (CF), immunofluorescenit (IF), and neutralizing (N) activities int sheep sera Serum HA CF IF N Anti-visna 8a 16 128 > 512 Control <4 <4 <8 <4 a Values expressed as reciprocal of the antibody titer. fluorescein-labeled goat anti-rabbit globulin globulin (Bell Products Co., South Orange, N.J.) or ith a fluorescein-labeled rabbit anti-sheep globulin globulin (Hyland Division of Travenol Laboratories, Inc., Costa Mesa, Calif.). They ere then ashed, air-dried and mounted in buffered glycerol. The stained cell layers ere examined under a Leitz fluorescence microscope ith an HBO 2 mercury burner as the ultraviolet light source. Photographs ere taken ith a Leica 35-mm camera ith Kodak Ektachrome highspeed film. For titration of antibody activities by immunofluorescent staining, infected cell monolayers ere first incubated ith serial tofold dilutions of the visna immune sera and then stained ith the fluoresceinlabeled anti-rabbit or anti-sheep globulins. The immunofluorescence titer of the sera is the highest dilution hich shoed specific staining ith fluorescein-labeled globulins. All sera used in the immunofluorescence tests ere absorbed ith sheep brain poder. Neutralization test. The test as carried out as previously described (12). Serial tofold dilutions of the sera ere mixed ith an equal volume of a virus dilution containing approximately 2 TCID5 per.1 ml. The mixtures ere incubated at room temperature overnight and ere then inoculated in.1 -ml amounts into tissue culture tubes. The tubes ere incubated for 14 days before final reading. The neutralizing titer is the highest dilution of serum hich shoed complete neutralization of virus in 5% of the inoculated tubes. Sephadex G-2 chromatography. Hyperimmune rabbit sera and the visna sheep serum ere fractionated on Sephadex G-2 (Pharmacia, Uppsala, Seden) in.2 M phosphate buffer, ph 7.3, at room temperature. Fractions ere collected ith an automatic device and ere assayed for protein content by electrophotometry at 28 nm. Antibody activities of the fractions ere determined by the various serological tests. RESULTS Hyperimmunization ith purified virus. To determine if hyperimmunization ith purified visna virus ould result in an antibody response, five rabbits ere immunized by the method previously described, and their sera ere assayed for antibody activities by PHA and CF tests. In the tests, three types of control sera ere used: serum from rabbits immunized ith bovine serum albumin, serum from rabbits immunized ith sheep choroid plexus cells, and normal serum obtained from uninoculated rabbits. Table 1 summarizes the results of these tests carried out for each of the various rabbit sera. The hyperimmune sera ere all strongly positive in the PHA and CF tests against visna virus, although the CF antibody titers ere considerably loer than the PHA titers. None of the control and normal rabbit sera reacted ith the virus antigen, and no reactions ere observed beteen control antigen

VOL. 4, 1971 ANTIBODIES AGAINST VISNA VIRUS 717 FIG. 1. Visna-infected sheep choroid plexus cell monolayer treated ith immune rabbit serum and stained ith fluorescein-labeled goat anti-rabbit globulin globulin. Specificfluorescent staining in the cytoplasm ofmultinucleated cells. X 25. (choroid plexus cells) and positive hyperimmunized rabbit sera. Antibody activity detectable by immunofluorescence. To of the rabbit hyperimmune sera, AVS-9 and AVS-1, ere tested for immunofluorescent activity against visna antigen in monolayers of sheep choroid plexus cells shoing cytopathic effect of visna virus. A specific, yellogreen cytoplasmic fluorescence as observed in the visna-infected cells (Fig. 1). In contrast to N- - this characteristic fluorescence obtained ith CM immune rabbit sera in visna-infected cells, no specific fluorescence as seen in controls. The controls included infected sheep choroid plexus cells treated ith normal rabbit serum and noninfected sheep choroid plexus cells treated ith immune rabbit sera. A titration of the indirect immunofluorescent activities in the to hyperimmune rabbit sera shoed that they ere active in dilutions up to 1:64 and 1:32, respectively (Table 1). Neutralizing antibodies. In contrast to the observation of PHA, CF, and immunofluorescent activities against visna virus in hyperimmunized rabbit sera, no neutralizing activities could be detected (Table 1). Antibody activities of Sephadex G-2 fractions. The gamma globulin fractions of rabbit immune sera ere tested by PHA and CF tests. Figure 2 shos the PHA activities in the fractions. The HA antibody activities ere located in the first [immunoglobulin M (IgM)J and second E co 6 FRACTI ON FIG. 2. Gel filtration on Sephaex G-2 of immune rabbit serum and hemagglutination (HA) activities of individualfraction tubes. HA activities in gamma globulin fractions of rabbit hyperimmune sera fractionated by gel filtration on Sephadex G-2..D *cn I

718 KARL AND THORMAR INFEC. IMMUN. peaks of the fractionated sera. The second peak ill hereafter be referred to as the immunoglobulin G (IgG) fraction, although it probably contains a mixture of IgG and IgA. The IgM fraction gave a higher titer than the IgG fraction. Fractions containing the IgM and IgG gamma globulins, respectively, ere pooled, concentrated by means of pervaporation, and tested for CF activities. Figure 3 shos the CF titers of the pooled fractions. In contrast to the finding for the HA activity, the CF activity as found in the pooled IgG fractions; hereas no CF activity as observed against visna antigen in the IgM fractions. The pooled and concentrated IgM fraction and IgG fractions from immune rabbit sera ere also tested by immunofluorescent test. While there as some staining of cells incubated ith the IgM pool, most of the activity as found in the IgG pool. Antibody activities in visna sheep serum. The visna antiserum from sheep as tested in PHA, CF, immunofluorescent. and neutralization tests. A serum from a normal sheep served as a control. In contrast to the immune rabbit sera, the visna me OD N c 3 5 7 F R A C T I ON FIG. 3. Gel filtration on Sephadex G-2 of immune rabbit serum and complement fixation (CF) activities of fraction pools. CF activities in pooled gamma globulin fractions of rabbit hyperimmune sera fractionated by gel filtration on Sephadex G-2. H- H- L E N C\ 14988 POOLS 2 3 4 5 6 7 FRACTION FIG. 4. Gel filtration on Sephadex G-2 of visna sheep serum and hemagglutination (HA) activities of fraction pools. HA activities in pooled gamma globulin fractions of visna sheep serum fractionated by gel filtration on Sephadex G-2. sheep serum shoed a high neutralizing activity against visna virus. It also shoed a higher immunofluorescent activity than the rabbit sera. On the other hand, the CF activity and particularly the PHA activity ere much loer in the visna sheep serum than in the immune rabbit sera (Table 2). Pooleu and concentrated IgM and IgG fractions from the visna sheep serum ere titrated in PHA, CF, and neutralization tests against visna virus. The results of the PHA test ere similar to those obtained ith rabbit immune globulin fractions, except that the HA activity as higher in the IgG than in the IgM fraction (Fig. 4). As in the rabbit sera, the CF activity and most of the immunofluorescent activity ere found in the IgG fraction. Most of the neutralizing activity as recovered in the pooled IgG fraction. DISCUSSION In the present study it has been shon for the first time that antibodies against visna virus can be formed in a nonnatural host, namely the rabbit. A specific antibody formation as demonstrated by CF and immunofluorescent tests, hich have previously been used in studies of antibody formation in sheep (1, 11). Furthermore, very high antibody titers against visna virus antigen ere demonstrated by means of the PHA test, hich has not previously been applied to visna antisera. The PHA test is a highly sensitive serological procedure for measuring antibodies to protein antigens (3, 8). It is a very effective

VOL. 4, 1971 ANTIBODIES AGAINST VISNA VIRUS 719 assay method for detecting the presence of visna virus antibodies in rabbits, since titers ranging from 1:32 to 1:2,48 (1 to 1 times greater than titers obtained in the CF test) ere obtained in the PHA test. It as found that crude visna virus preparations produced nonspecific PHA reactions. Therefore, to ascertain the specificity of the antigen-antibody reactions, only purified, concentrated virus as used in the present experiments. It is noteorthy that despite the significant antibody titers obtained by the CF, immunofluorescent, and particularly the PiIA tests, no neutralizing activity against visna virus as detectable in the immune rabbit sera by our standard neutralization test. This is in contrast to the visna sheep serum, hich shoed a very high neutralizing activity but a rather lo PHA activity. The cause of this difference in the antibody pattern in the to hosts is presently unknon. Possibly, visna-neutralizing antibodies reside in IgA, and rabbits do not produce IgA antibodies against visna virus. It is also unknon hether the virus propagates in or causes any lesions in the injected rabbits. The animals are still alive and ithout clinical signs 11 months after initiation of the experiment. Fractionation of the rabbit immune sera by Sephadex G-2 chromatography shoed that both IgM and IgG types of antibody ere present in the sera as a response to injection ith purified visna virus. The experiments depicted in Fig. 2 and 3 indicate that most of the HA activity of the rabbit immune sera as present in the IgM fraction rather than in the IgG fraction, hereas the CF activity as present in the IgG fraction. The pattern as somehat different from that observed in fractionated sheep sera, here the larger part of the HA activity as found in the IgG fraction. Most of the neutralizing activity of the visna sheep serum as also found in the IgG fraction. Hoever, because the IgG fraction of the sheep serum might possibly include an IgA component, not distinguishable by Sephadex G-2 filtration, the present results are not in conflict ith those of Petursson (6), ho concluded that visna-neutralizing antibodies of sheep sera apparently reside in the IgA fraction. A further study of the antibody formation in sheep infected ith visna is needed. ACKNOWLEDGMENTS These experiments ere carried out ith the excellent technical assistance of Arlene Tetley and Hannah Bron. The help of F. H. Lin in purification of the virus is gratefully acknoledged. LITERATURE CITED 1. Gudnadottir, M., and K. Kristinsdottir. 1967. Complementfixing antibodies in sera of sheep affected ith visna and maedi. J. Immunol. 98:663-667. 2. Harter, D. H., K. C. Hsu, and H. M. Rose. 1967. Immunofluorescence and cytochemical studies of visna virus in cell culture. J. Virol. 1:1265-127. 3. Lefkoitz, S. S. 1968. Application of passive hemagglutination methods for detection of antibody to viral antigen in hamsters bearing adenovirus type 12 tumors. J. Immunol. 11:611-616. 4. Lin, F. H., and H. Thormar. 197. Ribonucleic acid-dependent deoxyribonucleic acid polymerase in visna virus. J. Virol. 6:72-74. 5. Lin, F. H., and H. Thormar. 1971. Characterization of ribonucleic acid from visna virus. J. Virol. 7:582-587. 6. Petursson, G. 197. Studies on viral antibodies in visna. Proc. Sixth Int. Congr. Neuropathol. Paris, p. 831-832. 7. Sever, J. L. 1962. Application of a microtechnique to viral serological investigations. J. Immunol. 88:32-329. 8. Sibal, L. R., M. A. Fink, J. L. Vice, B. L. Brandt, and T. E. O'Connor. 1966. Hemagglutination studies of the viral antigen in a murine leukemia (Rauscher). Proc. Soc. Exp. Biol. Med. 122:591-596. 9. Sigurdsson, B., H. Thormar, and P. A. Palsson. 196. Cultivation of visna virus in tissue culture. Arch. Gesamte Virusforsch. 1:368-381. 1. Stavitsky, A. B. 1954. Micromethods for the study of proteins and antibodies. J. Immunol. 72:36-367. 11. Thormar, H. 1969. Visna and maedi virus antigen in infected cell cultures studies by the fluorescent antibody technique. Acta Pathol. Microbiol. Scand. 73:296-32. 12. Thormar, H., and H. Helgadottir. 1965. A comparison of visna and maedi virus. I1. Serological relationship. Res. Vet. Sci. 6:456-465.