Inhibitory effect of passive antibody on active immunity induced against rabies by vaccination*

Bull. Org. mond. Sante 1971, 45, 747-753Bull. Wld Hlth Org. Inhibitory Effect of Passive Antibody on Active Immunity Induced Against Rabies by Vaccination* T. J. WIKTOR,1 R. A. LERNER 2 & H. KOPROWSKI3 The administration of rabies-immune serum in addition to regular vaccination has been recommended as routine treatment in man after severe exposure to rabies virus. The inhibitory effect ofpassive antibody on active immunity induced against rabies by vaccina- tion was investigated in rabbits. It was demonstrated that a dose of rabies-immune serum could completely block the neutralizing antibody response engendered by a very potent rabies vaccine. The inhibitory effect could be partially overcome if the number of vaccine doses was increased, if vaccination was started 15 days after serum administration, or if immune serum was given 12-24 hours after vaccination. Even under these circumstances, the antibody level was lower than that observed after administration of vaccine alone. Both 19S and 7S antibody exerted similar effects. Field trials conducted in Iran in 1954 indicated that rabies-immune serum, when combined with 21 doses of Fermi-type (phenolized brain tissue) vaccine, markedly improved the survival rate of subjects severely bitten by rabid wolves, in com- parison with the results of treatment with vaccine alone (Baltazard & Bahmanyar, 1955). Since then, immune serum has been recommended as routine treatment in man after severe exposure to rabies virus (WHO Expert Committee on Rabies, 1957). Experi- mental justification for including immune serum in post-exposure treatment has been obtained from laboratory studies on the protection of mice, ham- sters, and guinea-pigs (Habel, 1945; Koprowski & Black, 1954; Koprowski, Van der Scheer & Black, 1950), by treatment with serum alone or in combina- tion with vaccine. Observations made during the past few years have indicated, however, that heterologous immune serum may inhibit the antibody response induced by active immunization in man and that an increase in the * From the Wistar Institute of Anatomy and Biology, Philadelphia, Pa., USA. This study was supported in part by US Public Health Service grants AI-02954 and AI-07988 from the Institute for Allergy and Infectious Disease, FR-05540 from the General Research Support Branch, and CA-10028 from the National Cancer Institute, and by funds from the World Health Organization. I Associate Member, Wistar Institute. 2 Present address: Scripps Clinic and Research Founda- tion, La Jolla, Calif., USA. 3Director, Wistar Institute. dosage of vaccine, combined with booster doses of vaccines administered at long-term intervals after the initial treatment, may overcome this effect (Atanasiu et al., 1957). These findings were supported by the results of experimental studies in guinea-pigs, which showed that the administration of homologous rabies- immune serum inhibited active immunization fol- lowing vaccination much more effectively than did heterologous serum (Archer & Dierks, 1968). Recent evidence indicates that the number of fail- ures of human post-exposure treatment after therapy with both immune serum and vaccine, followed by booster doses of vaccine, is increasing (Anderson, Daly & Kidd, 1966; Dehner, 1970). In view of this increase, and of the availability on the one hand of a potent preparation of homologous immune serum for the treatment of man, and, on the other hand, of a potent experimental vaccine that necessitates only one vaccinating injection (Wiktor et al., 1969; Sikes et al., 1971), we decided to evaluate, on a more quantitative basis, the effect in experimental animals of passive immunity on active immunity induced by vaccination. MATERIALS AND METHODS Animals Young adult New Zealand rabbits (2-3 kg body weight) of both sexes were used in all experiments, as well as for the preparation of immune serum. 2766 - 747 - 748 T. J. WIKTOR AND OTHERS Virus Tissue-culture-adapted PM, CVS, ERA, or HEP strains of rabies virus (Wiktor, Fernandes & Koprowski, 1964) were purified by cloning, using a plague technique employing agarose-suspended BHK-S13 cells (Sedwick & Wiktor, 1967). The cloned virus stocks were propagated in monolayer cultures of BHK-Sl3 cells in a medium containing 0.1% bovine serum albumin according to methods described previously (Sokol et al., 1968). A roller- type culture was used for virus propagation. Immune serum The immunizing antigens were PM, CVS, ERA, and HEP rabies virus prepared by the concentration of infected tissue-culture fluid by ultrafiltration (Strohmaier, 1967) followed by 2 cycles of high- speed centrifugation at 12 500 rev/min in a No. 30 rotor, Spinco model L centrifuge. The final sedi- ment was resuspended in NT buffer (0.13 M NaCl and 0.05 M 2-amino-2-[hydroxymethyl]-1,3-propane- diol [" Tris "] hydrochloride, pH 7.8) containing 0.1 % human serum albumin. Rabbits immunized with HEP virus were inocu- lated intravenously with 3 x 109 plaque-forming units (PFU) of infectious virus once a week for 4 weeks and bled 7 days after the last inoculation. For the preparation of immune serum against the PM, CVS, and ERA strains, animals were inoculated intra- muscularly once a week for 2 weeks with 4 x 109 PFU of virus inactivated by ultraviolet irradiation (Wiktor et al., 1969) and emulsified in equal parts of incom- plete Freund's adjuvant. The animals then received 2 intravenous inoculations of I x 109 PFU of live virus, and were bled 7 days after the last inoculation. Sera obtained from rabbits immunized against each of the four virus strains were pooled to form a polyvalent immune rabies serum for use in this experiment. The specific activity of this serum was 400 international units (IU)/ml as measured by the standard technique (Atanasiu et al., 1966), using the reference serum preparation supplied by the Division of Biologics Standards, National Institutes of Health, Bethesda, Md., USA. 19S antibody Rabbits were inoculated intramuscularly with a single dose of inactivated rabies vaccine (see below) and were bled 5 days later. The euglobulin fraction of this serum was separated by precipitation over- night at 4°C (pH 5.5) after the addition of 1O volumes of distilled water. The precipitate was washed twice and resuspended in saline to approximately 1/3 of the original volume. Vaccine Virus from infected tissue-culture fluid (PM strain) was concentrated by ultrafiltration (Strohmaier, 1967) and sedimented by centrifugation at 12 500 rev/min for 60 min. The resulting sediment was resuspended in NT buffer containing 0.1 % bovine serum albumin. Under these conditions, rabies virions were nearly quantitatively sedimented (Schlumberger et al., 1970). The final preparation (concentrated one hundred times and showing infectivity of 4-6 x 109/ml) was inactivated with ,B-propiolactone, as described pre- viously (Wiktor et al., 1969). The absence of infec- tive virus in the vaccine preparation was tested by intracerebral inoculation of the undiluted product into mice aged 4-5 weeks. The protective activity (National Institutes of Health (NIH) mouse potency test) was determined as described by Wiktor et al. (1969). The antigenic value of the experimental vac- cine was obtained by dividing the effective dose (ED5O) dilution endpoint of the test preparation by that of the reference vaccine. The two different vaccine preparations used in this experiment had antigenic values of 30 and 45, respectively. Serum neutralization test The level of serum neutralizing antibody in immu- nized animals was determined in vitro by a plaque assay technique (Sedwick & Wiktor, 1967). Two- fold dilutions of inactivated (heated at 56°C for 60 min) serum were tested by incubation with a virus preparation (PM strain) diluted to contain 500 PFU/ml. The neutralizing antibody titre was expressed as the reciprocal of the highest dilution of immune serum causing a 50% reduction in plaque numbers. Virulent challenge Ninety days after the initial vaccination, rabbits, together with a number of unvaccinated control animals, were challenged by intramuscular inocula- tion with 105-7 mouse LD50 of a DR street rabies virus of vampire bat origin.' I Obtained through the courtesy of Dr Boris Szyfres, Director, Centro Panamericano de Zoonosis, Buenos Aires, Argentina. INHIBITORY EFFECT OF PASSIVE ANTIBODY ON ACTIVE IMMUNITY TO RABIES RESULTS Combinied treatment of rabbits with immune serum anid vaccine In the experiments summarized in Fig. 1, one group of rabbits received 1 dose of vaccine, one group received 1 dose of immune serum (100 IU/kg), and one group received simultaneously 1 dose of vaccine and I dose of immune serum. The results indicate that whereas the passively acquired immu- nity could be detected on the first day after admin- istration of serum, antibodies engendered by 1 dose of vaccine could be demonstrated only on the third day after injection. From then on, however, the antibody levels in the group receiving 1 injection of vaccine rose rapidly, reaching a peak on the 7th day after vaccination, with an antibody titre of 4600. This antibody level was maintained for 26 days and had declined only slightly by the 50th day of obser- vation. In marked contrast, the antibody level observed following administration of immune serum alone remained stationary for 7 days, then declined rapidly, reaching undetectable levels on the 50th day after administration. 10 000- a 0 co z 4 (D z -J 4 mDI` W 000 - 00 l0 20 30 DAYS AFTER VACCINATION 40 50 Fig. 1. Dynamics of virus-neutralizing antibody forma- tion in rabbits inoculated with 1 dose of rabies vaccine alone, 100 IU of homologous rabies antibody alone, and a combination of both. When a combination of vaccine and serum was administered, the antibody level paralleled that ob- served after the administration of serum alone for 10 days, and then declined even more rapidly. On the 50th day after immunization, with the serum and vaccine combined, however, a titre of 30 was still detectable. In the next set of experiments, attempts were made to overcome the inhibitory effect of passive immu- nity on the induction of antibodies by vaccine through the use of a reduced dose of 50 IU/kg of immune serum, corresponding to a dose recom- mended for human post-exposure treatment, and also through administration of a larger number of vaccine doses than in the first experiment. The results (Fig. 2) indicate that although, in all instances, passive immunity lowered the antibody level induced by vaccination, the daily administra- tion of 5 doses of vaccine following the administra- & VACCINE THREE DOSES A VACCINE THREE DOSES AND IMMUNE SERUM 501 U / KG O VACCINE FIVE DOSES * VACCINE FIVE DOSES AND IMMUNE SERUM 501U /KG p,..-..O0 .0',. .......................0~~~~~~~~~~~~~~~~~~ 0 -: 4 St, (2 000 251 4 6 8 '2 14 i I'18 2 22 24' 26 DAYS AFTER VACCINATION Fig. 2. Dynamics of virus-neutralizing antibody forma- tion in rabbits inoculated with 3 or 5 doses of rabies vaccine alone and in combination with 50 IU of homo- logous rabies antibody given as a single dose together with the first dose of vaccine. 749 O VACCINE ONE DOSE O IMMUNE SERUM 1001I U KO * IMMUNE SERUM AND VACCINE ONE DOSE 0--.-0-.... 0. I"--v-_ T. J. WIKTOR AND OTHERS DAYS AFTER VACCINATION Fig. 3. Dynamics of virus-neutralizing antibody forma- tion in rabbits inoculated with 1 dose of rabies vaccine alone, and in combination with 20, 4, and 0.8 IU of homologous rabies antibody. tion of serum seemed to overcome, to a certain extent, the inhibitory effect of passive immunity. Conversely, the administration of 3 doses of vac- cine, by comparison with 1 dose, seemed to make little difference in overcoming the inhibitory effect of passive immunity. In the next set of experiments, we tried to demon- strate on a more quantitative basis the inhibitory effect of immune serum. Three groups of rabbits received 3 different doses of immune serum-namely, 20, 4, and 0.8 IU/kg, respectively-simultaneously with 1 dose of vaccine, as used in the previous experiments. The results (Fig. 3) indicate that the amount of serum administered determined the level of passive immunity observed on the first day after administra- tion; the smallest dose of serum (0.8 IU/kg) had the smallest inhibitory effect on the development of anti- bodies following vaccination, whereas the largest dose (20 IU/kg) had the greatest effect. It should 10 000oo 0~~~~~~ L 00 z 4 -I 00/ z 10 O VACCINE ONE DOSE * VACCINE ONE DOSE AND 19S ANTIBODY 5 U /KG * VACCINE ONE DOSE AND 19S ANTIBODY U / KG A VACCINE ONE DOSE AND 19S ANTIBODY 0.2 U /KG 2 4 6 N 10 12 14 16 l8 20 22 24 26 28 DAYS AFTER VACCINATION Fig. 4. Dynamics of virus-neutralizing antibody forma- tion in rabbits inoculated with 1 dose of rabies vaccine alone, and in combination with 5, 1, and 0.2 IU of homologous rabies 1 9S antibody. be emphasized that the dosage of 20 IU/kg is less than that recommended at present for post-exposure treatment of man. Effect of 19S antibody in the combined immuniue serum- vaccine treatment To explore the possibility that 19S rabies antibody administered to rabbits may stimulate rather than inhibit antibodies induced by active immunization (Henry & Jerne, 1968), we immunized rabbits with rabies vaccine and obtained serum 5 days after immunization. Euglobulin was precipitated from the serum twice by the method already described. After the neutralizing antibody content had been determined by the plaque reduction method, euglobu- lin was given to groups of rabbits in three dosages- namely, 5, 1, and 0.2 IU/kg-together with vaccine. The results (Fig. 4) indicate that no antibody level could be detected on the first day following the administration of the 0.2 IU/kg dose. The antibody 750 INHIBITORY EFFECT OF PASSIVE ANTIBODY ON ACTIVE IMMUNITY TO RABIES level determined on the same day following adminis- tration of 1 and 5 IU/kg corresponds to that observed after administration of the 7S antibody (Fig. 3). There was no clear indication, however, that pas- sive immunity engendered by 19S antibody is less inhibitory to active immunization induced by vaccine than that engendered by 7S antibody. Following the administration of either 5 IU/kg or 1 IU/kg of 19S antibody and vaccine, the antibody level seemed to be the same as that after administra- tion of 4 IU/kg or 0.8 IU/kg, repectively, of 7S anti- body and vaccine (Fig. 3). Effect of delayed administration of vaccine on the combined immune serum-vaccine treatment Because the half-life of passively acquired anti- body (see Fig. 1) is approximately 14 days, immune serum (7S) was administered to rabbits at the rate of 50 IU/kg and delayed injections of vaccine were begun 8 and 15 days later; 3 doses of vaccine were administered in each case. The results (Fig. 5) indi- IC a 0 D z 4 (2 z 4 4 O VACCINE THREE DOSES * IMMUNE SERUM 501U /KG ANO THREE DOSES OF VACCINE STARTING ON OAYS OR 15 O IMMUNE SERUM ON DAY O 000- ,.~~ 000 100 --- -~~~~~~~~~~~~C 2 4 6 8 10 12 14 16 IS 20 22 24 26 25 DAYS AFTER VACCINATION Fig. 5. Dynamics of virus-neutralizing antibody forma- tion in rabbits inoculated with 3 doses of rabies vaccine alone and with 50 IU of rabies antibody, followed by 3 doses of rabies vaccine starting on days 8 or 15 after the administration of antibody. cate that whereas initiation of vaccine treatment on the 8th day after the injection of serum did not overcome the inhibitory effect of passive immunity, initiation of vaccine treatment on the 15th day seemed to induce much higher antibody levels; these levels, however, were still lower than those observed after administration of 3 doses of vaccine alone. Effect of delayed administration of immune serum on the combined immune serum-vaccine treatment In another series of experiments, groups of rabbits were vaccinated 3, 12, or 24 hours before the admin- istration of immune serum. Control groups received vaccine or serum alone. The results (Fig. 6) indicate that in all instances the inoculation of immune serum lowered the antibody level induced by vaccination. When, however, immune serum was administered 12 or 24 hours after vaccination, the interference was partially overcome. .'0 0.' 1000- 0 0 z 4 z ,' VACCINE ONE DOSE ,' a VACCINE ON E DOSE AND ,' IMMiUNE SERUM 24 H LATER 0 VACCINE ONE DOSE AND ,' IMMUNE SERUM 12 H LATER ,' A VACCINE ONE DOSE AND ,' IMMUNE SERUM 3 H LATER - IMMUNE SERUM 401 U / KG 2AJ 1006 12 I4 __Ii 20 22 i4 DAYS AFTERD CCINATION Fig. 6. Dynamics of virus-neutralizing antibody forma- tion in rabbits inoculated with 1 dose of rabies vaccine alone, with 40 IU of rabies antibody alone, and with 1 dose of vaccine followed 3, 12, and 24 hours later by inoculation of 40 IU of rabies antibody. 751 T. J. WIKTOR AND OTHERS DISCUSSION It is well known that the presence of passively acquired antibodies may interfere with the actual formation of antibodies against an antigen (Uhr & Baumann, 1961). It was somewhat surprising, however, to find that the rabies-immune serum so effectively blocked the neutralizing antibody response engendered by an otherwise very potent rabies vac- cine. The inhibitory effect could be partially over- come if the number of vaccine doses was increased from I to 5, when vaccination was started on the 15th day after administration of immune serum, or when immune serum was given 12-24 hours after vaccine inoculation. Even under these circumstances, however, the antibody level was lower than that observed after the administration of vaccine alone. Blocking of theantibody response by immune serum made the animals susceptible to challenge with street virus, even though rabbits in one group received a smaller dose-namely, 50 IU/kg-of immune serum follow;ed by 3 injections of the vaccine (see accom- panying table). Animals that received vaccine alone always withstood challenge with virulent virus. Although all the experiments reported here were performed on rabbits, the blocking action of immune serum has also been observed in other species of animal and against a different type of vaccine. Winkler et al. (1969) showed that the treatment of dogs with 83 IU/kg of rabies-immune serum of horse origin effectively blocked the protective effect of 14 injections of duck-embryo vaccine. Only when the dose of immune serum was decreased to an undetectable level in the serum of the inoculated animals was the blocking effect eliminated. This dosage of serum (Fig. 4) would be ineffective per se in preventing the spread of rabies virus if admin- istered after exposure to virulent virus. The blockingeffect ofpassive immunity could not be related to a particular fraction of immune serum, since both 19S and 7S antibody exerted a similar effect. The rationale of post-exposure treatment was based on the fact that rabies is probably the only infection in which the time of actual exposure to the virus can be determined exactly and in which the spread of the virus from the site of infection to the central nervous system occurs relatively slowly. Within this incubation time, the virus may become bound to an antibody engendered by either active or passive immunization. The latter treatment would be par- ticularly important in cases of severe exposure, in which arrest of the spread of the virus would have to take place almost immediately. Since in virus- antibody complexes the virus may not be completely inactivated (Wiktor & Koprowski, unpublished data), it was thought that an assurance of constant surveil- lance by antibody produced as a response to active immunization had to be secured by administering vaccine following treatment with immune serum. If, however, administration of immune serum in doses that convey a certain level of passive immunity to the animal results in blocking, or decreasing, the antibody response engendered by vaccine, the pat- tern of post-exposure treatment with immune serum and vaccine may have to be more carefully scruti- nized and a scheme of treatment devised to provide Protection of rabbits treated with va serum, or a combination of both. Va trated rabies vaccine of tissue culture ( value = 30; immune serum, homc immune serum, 400 IU/mi; challenge, virus, 105 .7 mouse LD5o, administered b inoculation, 90 days after vac( Treatment Vaccine (doses) Serum (doses) 100 iu 3 _ 3 50 IU o100 IU iccine, immune an exposed person or animal with an adequate level iccine, concen- of rabies antibodies. An attempt to face this prob- rigin, antigenic lem directly was made by the WHO Expert Commit-)logous rabies- tee on Rabies (1957), which recommended booster D)R street rabies doses of vaccine after the completion of the stan-iy intramuscular cination dard course of treatment, if the patient had also received immune serum. Since the vaccine currently used in northern-hemisphere countries has a lower Mortality ratio immunogenicity than the vaccine used previously, combined treatment with serum and vaccine may have to be followed with great care in order to 0/4 determine how many injections of the vaccine should 4/4 be given, and at what dosage level, before a patient 0/4 can be assured of adequate immunological surveil- 3/4 lance over a long period. Once the new and potent rabies vaccine becomes 4/4 available, it is not inconceivable that the use of the 4/4 immune serum may become obsolete. 752 1 INHIBITORY EFFECT OF PASSIVE ANTIBODY ON ACTIVE IMMUNITY TO RABIES 753 ACKNOWLEDGEMENT The authors gratefully acknowledge the excellent technical assistance of Mrs Doris Grella. RESUME EFFET INHIBITEUR DES ANTICORPS PASSIFS SUR L'IMMUNITE ANTIRABIQUE ACTIVE INDUITE PAR LA VACCINATION Des essais pratiques menes en Iran en 1954 ont montre qu'en associant I'administration de serum antirabique A l'injection de 21 doses d'un vaccin type Fermi on augmentait notablement les chances de survie des sujets mordus par des loups enrages. Depuis lors, l'emploi d'immunserum a ete recommande chez l'homme comme traitement de routine apres une exposition grave. Cepen- dant, on a signale recemment un nombre croissant d'echecs du traitement mixte serum-vaccin suivi de doses de rappel de vaccin chez les sujets exposes A l'infection. Le fait a incite les auteurs A etudier chez le lapin 1'effet suppressif des anticorps passifs homologues sur la reponse immunitaire active suscitee par la vaccination. On a utilise comme vaccin la souche PM de virus rabique propagee sur culture tissulaire, concentree par ultrafiltration, partiellement purifiee par centrifugation a vitesse elev&e et inactivee par la /-propiolactone. La valeur antigenique de cette preparation etait 30-45 fois superieure A celle du vaccin de reference NIH. Les serums antirabiques ont e prepares chez des lapins inmmunises par plusieurs inoculations de virus rabique de culture tissulaire, vivant ou inactive. Quant aux anticorps 19S, ils ont e obtenus par inoculation A des lapins d'une dose unique de vaccin antirabique inactive 5 jours avant la saignee, la fraction euglobulinique du serumn tant isolee par precipitation a pH 5,5. La determination des titres d'anticorps a ete faite par l'epreuve de neutralisation sur plaque. Au cours d'une serie d'experiences, il a e demontre que la reponse immunitaire a la vaccination est complete- ment abolie chez les animaux traites par administration simultanee d'une dose d'immunserum et d'une dose d'un vaccin de culture tissulaire ah forte activite immunogene. L'effet suppressif n'est que partiellement surmonte par l'injection quotidienne de vaccin pendant les 5 jours suivant l'injection du serum. L'injection de 3 doses de vaccin se revele pratiquement impuissante a contrecarrer l'action inhibitrice de l'immunite passive. L'effet inhi- biteur est fonction de la dose de serum et la dose de 40 Ul/kg actuellement recommandee pour le traitement de l'homme expose a l'infection rabique a un pouvoir d'inhibition considerable. Les fractions globuliniques 19S et 7S ont une action analogue. Si le traitement vaccinal est instaure 8 jours apres l'administration de serum, il ne reussit pas a vaincre l'inhibition due a l'immunite passive; s'il est commence apres 15 jours, il suscite une reponse immunitaire active beaucoup plus forte, mais qui reste cependant inferieure a celle qu'entraine le traitement vaccinal seul. Le blocage de la reponse immunitaire active peut Wtre partiellement empeche si le serum est administre 12 a 24 heures apres la vaccination. REFERENCES Anderson, J. A., Daly, F. T. & Kidd, J. C. (1966) Ann. intern. Med., 64, 1297 Archer, B. G. & Dierks, R. E. (1968) Bull. Wld Hlth Org., 39, 407 Atanasiu, P. et al. (1957) Bull. Wld Hith Org., 17, 911 Atanasiu, P. et al. (1966) Laboratory techniques in rabies, Geneva, World Health Organization Baltazard, M. & Bahmanyar, M. (1955) Bull. Wld Hith Org., 13, 747 Dehner, L. P. (1970) Ann. intern. Med., 72, 375 Habel, K. (1945) Publ. Hlth Rep. (Wash.), 60, 545 Henry, C. & Jerne, N. K. (1968) J. exp. Med., 128, 133 Koprowski, H. & Black, J. (1954) J. Immunol., 72, 85 Koprowski, H., Van der Scheer, C. E. & Black, J. (1950) Amer. J. Med., 8, 412 Schlumberger, H. D., Wiktor, T. J. & Koprowski,H. (1970) J. Immunol., 105, 291 Sedwick, W. D. & Wiktor, T. J. (1967) J. Virol., 1, 1224 Sikes, R. K., Cleary, W. F., Koprowski, H., Wiktor, T. J., & Kaplan, M. (1971) Bull. Wld Hlth Org., 45, 1-11 Sokol, F., Kuwert, E., Wiktor, T. J., Hummeler, K. & Koprowski, H. (1968) J. Virol., 2, 836 Strohmaier, K. (1967) In: Methods in virology, New York, Academic Press, New York, vol. 1, pp. 245-274 Uhr, J. W. & Baumann, J. B. (1961) J. exp. Med., 113, 935 Wiktor, T. J., Fernandes, M. V. & Koprowski, H. (1964) J. Inmunol., 93, 353 Wiktor, T. J., Sokol, F., Kuwert, E. & Koprowski, H. (1969) Proc. Soc. exp. Biol. (N.Y.), 131, 799 Winkler, W. G., Schmidt, R. C. & Sikes, R. K. (1969) J. Immunol., 102, 1314 WHO Expert Committee on Rabies (1957) Wld Hlth Org. tech. Rep. Ser., No. 121 5