Studies on Japanese B Encephalitis Virus Vaccines from Tissue Culture

APPLI1F MICROBIoLoGY, Apr. 1971, p. 743-748 Copyright 1971 American Society for Microbiology Vol. 21, No. 4 Printed in U.S.A. Studies on Japanese B Encephalitis Virus Vaccines from Tissue Culture XI. Immune Mechanism and Evaluation of the Mouse Challenge Potency Test BALWANT SINGH AmN W. McD. HAMMON Department of Epidemiology and Microbiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania 15213 Received for publication 28 January 1971 A study was carried out to evaluate the reliability of and to determine the mechanism involved in an antigen extinction mouse intraperitoneal (ip) challenge test for potency of a cell culture vaccine for Japanese B encephalitis, a modification of a test originated by Sabin for a mouse brain vaccine. Some comparisons were made with the official Japanese test using an intracerebral (ic) challenge after a more prolonged immunization procedure. The Japanese method of using a lyophilized reference vaccine with each test was also employed. It was found that the ip and the ic test appeared to show similar relative differences between lots. The ip test was more quickly and readily performed, gave reasonably consistent results on repetition, and, when used with a suitable reference vaccine, gave promise of being an entirely suitable and reliable test. Immunization by the intramuscular route rather than by the regular ip route appeared to offer no advantage and was less consistent in responses shown. Neutralizing antibody responses of the mice in the standard procedure were very quick to appear, about 4 days after the first dose of vaccine and had a peak titer about the seventh day, the time of challenge. This titer fell quickly unless challenge occurred. The antibody was heat stable, but it was readily inactivated by 2-mercaptoethanol (2-ME). Not until the 11th or 15th day did a small amount of immunoglobulin G appear. Challenge on day 7 significantly increased titers, but this antibody was also mostly inactivated by 2-ME. Interferon did not appear to play any significant role in the protection shown by the mice. The mouse protection potency test for Japanese B encephalitis (JBE) vaccine was first described by Sabin et al. (12) and was later modified and accepted as a standard test by the Division of Biological Standards (DBS), National Institutes of Health (NIH, reference 10). Koprowski and Cox (8) criticized this test, questioning its significance for determining the potency of their chick embryo product. However, Smadel et al. (15) pointed out that the variations in range of minimal immunizing dose (MID) between 0.005 and 0.02 ml (fourfold) were of little significance in respect to potency as measured by antibody response in man. Darwish and Hammon (3) used this test with very slight modification and reported an inverse correlation between the MID and the preinactivated titers of different lots of vaccines prepared in a variety of ways. The higher the infectivity and HA titers the lower the MID of 743 the inactivated vaccine. They found a rather wide range of experimental variability (far less than fourfold), but in most instances results were repeatable. The study reported here is based on work of the last several years during which further extensive testing of vaccine lots by the mouse potency test was done using, for much of the time, a lyophilized "standard" or "reference" vaccine with each test as a reference preparation not included in earlier requirements (10). The repeatability of this test and the profile and correlation of antibody response of the mouse were studied. The level and type of antibodies developed in mice and their time relationships are discussed in this paper. Comparisons were also made with the official potency test used in Japan and a variation of the American test by using the intramuscular (im) route of immunization instead of the intraperitoneal (ip) route.

744 SINGH AND HAMMON APPL. MICROBIOL. MATERIALS AND METHODS Preparation of inactivated vaccine. An attenuated strain of JBE virus (OCT-541,35-24 ) was grown in primary hamster kidney tissue culture. The method of virus harvest and vaccine preparation have been previously described (1, 2) as well as modifications in production (B. Singh, M. A. Darwish, and W. M. Hammon, Proc. Working Conf. Japanese Encephalitis Virus Vaccine, Igaku Shoin Ltd., Tokyo, in press). Usually the Formalin-inactivated liquid vaccine was shell frozen in ampoules containing about 1.5-ml quantity. These were stored in a Revco at about -65 C, and at the time of use the ampoule was thawed by using a 37 C water bath. Vials of lyophilized vaccine were stored at -20 C. At the time of use, the latter preparation was rehydrated to its original volume by adding 2 ml of doubledistilled water. Mouse potency test. The method of immunization and challenge of mice, both performed by the ip route, were as described by Darwish and Hammon (3). Serial fourfold dilutions of the vaccine from 1/20 to 1/1,280 were prepared and 12 weanling mice weighing 18 to 22 g were inoculated twice at an interval of 3 days with each dilution. For intramuscular (im) immunization the vaccine was diluted from 1/2 to 1/128, thus reducing the volume of each dose in order to have the same quantity of antigen irrespective of the route used for immunization. The number and sequence of injections were the same as for the ip method. The challenge was always with the Peking strain of JBE virus, and an intracerebral (ic) titration in weanling mice was made at the time of each test. The inoculum for ip challenge was 0.3 ml of the 10-2 dilution of a frozen 20%, suspension of suckling mouse brain containing about 107 log LD-o ic weanling mouse doses per 0.3 ml. The mice were observed for 14 days after ip challenge. When the Japanese NIH ic mouse challenge method for potency test (7) was used, fourfold dilutions of vaccine from 1/4 to 1/256 were prepared in buffer saline solution, ph 7.0. Again 12 mice, weighing 18 to 22 g, were used per dilution of vaccine, but four doses of vaccine (0.5 ml each) were given ip on alternate days. After 7 days from the last dose of vaccine, the mice were challenged ic with the Pekin strain virus (0.03 ml containing 150 LD50 as shown by simultaneous ic titration in controls of the same age). Mouse antibody assay. Serum samples from six mice receiving identical injections were pooled and were stored at -20 C. The antibody titers were determined by a plaque reduction neutralization test with chick embryo tissue culture and a single agar overlay technique. The technique was essentially the same as described by Nagai and Hammon (9). Plaques were counted after 4 days, and reciprocals of serum dilutions inhibiting 50% of the plaques were determined graphically. To inactivate IgM antibody the serum samples were treated with 2-mercaptoethanol (2-ME). Interferon assay. The interferon titers of serum pooled from six mice were determined by the technique of plaque inhibition assay as described previously (14). The duplicate L-cell culture plates were inoculated with 2 ml of serum dilutions. After overnight incubation at 37 C under 5% CO2, the plates were drained and then challenged with about 200 plaque-forming units (PFU) of vesicular stomatitis virus (13). The serum dilutions inhibiting 50% of the plaques were calculated by plotting reciprocals of the serum dilutions against the percentage of plaques reduced. RESULTS Repeatability of mouse challenge test. Several vaccine lots prepared and stored at -65 C were tested more than once. Results of mouse potency tests on six such lots are shown in Table 1. Vaccine lots 1-4 were tested at different intervals ranging from 1 to 8 months. In case of vaccine lots 2, 5, and 6, two sets of mice were immunized with the same vaccine on the same day and then challenged with a similar dose by the standard method. Vaccine lot 3 was tested three times. The results of second and third tests were the same and differed only by 0.001 ml from the first test. The repeated tests on other vaccine lots also gave very similar results. As shown in Table 1, the maximum variation in the MID on these lots when tested on the same day or at various intervals was 0.003 ml. The total data on the lyophilized vaccine which was used as a reference is shown in Table 2. In 22 tests done at different times during a 3-yr period, the mean MID on this vaccine was 0.018 ml with a range of 0.011 to 0.027 ml and a standard deviation of 0.0049 ml. TABLE 1. Repeatability of the mouse potenicy testa oii cell culture, Formalin-inactivated vaccine lots Vaccine lot 2 3 4 5 6 Date of test 9/6/66 10/11/66 4/4/67 4/18/67 4/18/67 2/3/69 4/25/69 10/22/69 3/11/69 6/3/69 6/23/69 6/23/69 10/29/68 10/29/68 MIDb (ml) 0.028 0.027 0.019 0.006 0.007 0.007 0.011 0.009 0.010 0.012 0.019 0.016 a Immunization and challenge by ip route. b Minimal immunizing dose.

VOL. 21 1971 JBE VIRUS VACCINES 745 TABLE 2. Minimal immunogenic dose (MID) of a lyophilized inactivated Japanese B encephalitis virus vaccine, as determined by repeated tests Test no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Date of test 4/4/67 9/30/68 10/22/68 2/3/69 2/25/69 3/25/69 4/15/69 4/25/69 4/25/69 5/19/69 6/3/69 6/23/69 7/11/69 7/18/69 7/25/69 9/29/69 10/20/69 10/22/69 12/8/69 2/6/70 4/17/70 5/15/70 0.025 0.027 0.022 0.024 0.013 0.020 0.021 0.012 0.015 0.019 0.011 0.012 0.013 0.025 0.021 0.018 0.014 0.026 Lyophilized vaccine stored at -20 C. b Determined by Reed and Muench formula from results obtained in standard, intraperitoneal mouse challenge test. Mean = 0.018 ml; range = 0.011-0.027 ml; standard deviation = 0.0049 ml. Effect of route of immunization and method of challenge. Groups of mice were immunized by the ip route with vaccines A and B. One half of the mice in each group were challenged by the ip and the other half by the ic route, the latter as required by the Japanese NIH method. The results of these two different methods are shown in Table 3. The MID on vaccine A were 0.014 and 0.19 ml by ip and ic challenge, respectively, as compared to 0.031 and 0.26 ml of vaccine B. By both methods of challenge, vaccine A gave the lesser MID value. By using only one lot of vaccine (C, Table 3), the ip method of immunization was compared with the im route, the standard ip route of challenge employed for both. As shown in Table 3, the MID value was significantly less when mice were immunized by the ip route. The MID were 0.014 and 0.018 ml, when immunization was by the usual ip route. However, when the mice were immunized by the im route, the MID were not only higher (0.023 and 0.062 ml) but the results of the duplicate tests were not nearly as consistent, one result being almost three times greater than the other. This difference is greater than ever TABLE 3. Effect of immunization route and method of challenge on minimal immunizing dose (MID) in mouse potency test Vaccine Immuniza- No. of Challenge lot tion route doses route6 MID (m) A ipc 2 ip 0.014 ipd 4 ic 0.19 B jpc 2 ip 0.031 ipd 4 ic 0.26 C ipc 2 ip 0.014 ime 2 ip 0.023 jpc 2 ip 0.018 imc 2 ip 0.062 Abbreviations: ip, intraperitoneal; ic, intracerebral; im, intramuscular. b Peking strain of Japanese B encephalitis virus as a 20% frozen suckling mouse brain. For ip challenge, 0.3 ml of 10-2 dilution of the frozen suspension which killed more than 90% of unvaccinated mice. For ic challenge, 0.03 ml contained 150 LD50. c Two equal doses on days 0 and 3. d Four equal doses at weekly intervals, Japanese National Institutes of Health method. obtained in our experience with the same vaccine when immunization was by the ip route (Tables 1 and 2). Neutralizing antibody levels. Serum neutralizing antibody titers were determined at different times on pooled sera of mice after receiving two doses of the 1:20 dilution of a vaccine, some challenged and others not. Animals were sacrificed to obtain blood on days 4, 7, 11, and 15. Results are shown in Table 4. The pooled serum sample showed hardly any neutralizing effect on day 4, but a significant level of 1:30 was found on day 7. This titer appeared to drop rapidly to 1:7 on day 11 and to 1:5 on day 15 in those mice held without challenge. Those challenged at the normal interval of 7 days after the first injection of vaccine experienced a rapid boosting effect to 1:90 on day 11, then the titer appeared to drop to 1: 34 on day 15. The heat stability of this neutralizing antibody was demonstrated by retest after heating the serum for 30 min at 56 C. Immunoglobulin M and immunoglobuin G activity. The early development of neutralizing antibodies and their short duration led us to study the type of antibody present. Antibody titers were determined before and after heating and after treatment with 2-ME on the 7th, 11th, and 15th day of bleedings without challenge after two doses of undiluted vaccine. After challenge, bleeding was on day 11 only. Results are presented

746 SINGH AND HAMMON APPL. MICROBIOL. in Table 5. Without challenge, the titer 7 days postvaccination and prior to 2-ME treatment was 1:100. It was not decreased by heating, but after 2-ME treatment it failed to neutralize even undiluted. Eleven days postvaccination, titers were 1:44 and 1:42, respectively, before and after heat inactivation, and after 2-ME treatment was only 1:10. The 15-day serum had the most 2-ME resistant antibody (titer 1:19) and the lowest untreated (1:36) titer. On day 11 in the challenged group, titers were greatly elevated until treated with 2-ME which left them at essentially the same level as those similarly treated that had not been challenged. TABLE 4. Serum antibody levels in groups of mice at different times during a vaccine potency test Group Treatment Serum dilution end points' (days after first vaccine) 4 7 11 15 1 Vaccineb 2 30 7 5 2 Vaccine + chal- 90 34 lengec 3 Unvaccinated and <2 <2 2 <2 not challenged areciprocal dilutions of serum inhibiting 50% of plaques of Nakayama strain on chick embryo tissue. b Two 1-cc doses of a 1:20 dilution of Japanese B encephalitis killed vaccine on day 0 and 3 by ip route. c A 0.3-ml amount of virulent Japanese B encephalitis virus on day 7 by intraperitoneal route. TABLE 5. Effect of heat and 2-mercaptoethanol treatment on serum neutralizing antibody titers Serum neutralizing antibody titersb Days- Group Challenge post- 2-Mervaciona tion' No captoe- treatment Heated thanold treated 1 No 7 100 79 2 2 No 11 44 42 10 3 No 15 36 32 19 4 Yes 11 > 125 > 125 14 a Mice were vaccinated on days 0 and 3 with undiluted vaccine. breciprocal dilutions of serum inhibiting 50% of plaques of Nakayama strain on chick embryo tissue. c For 30 min at 56 C. d Equal volume of serum and 0.1 M 2-mercaptoethanol for 2 hr at room temperature. TABLE 6. Serum interferon levels in mice receiving vaccine with or without challenge Treatment Interferon titers' (hr after injection) 2 6 24 48 Vaccine, 1st doseb <40 <40 <40 Vaccine, 2nd dose <40 <40 <40 Vaccine + challenge <40 <40 650 Challenge without 2,800 3, 300 <40 vaccine Measured in L cells; titers represent highest dilutions inhibiting 50% of vesicular stomatitis virus plaques. b Each vaccine dose was 1 ml of 1:20 dilution of vaccine. Effect of vaccine with and without challenge on levels of serum interferon. In an attempt to analyze more completely the mechanism of protection in the mouse challenge test, the levels of serum interferon before and after challenge were determined. Serum samples were obtained 2, 6, 24, and 48 hr after the first and second doses of vaccine, as well as after the challenge. The results are shown in Table 6. The mice receiving killed vaccine alone did not develop circulating interferon. But, when challenged on day 7, circulating interferon was detected 24 hr after challenge. Unvaccinated mice showed high levels of interferon by the sixth hour after challenge, persisting or rising by 24 hr but not detectable at 48 hr. DISCUSSION The mouse challenge test for determining the potency of JBE vaccine is basically an antigen extinction or titration type test. The mice are vaccinated with serially decreasing amounts of vaccine and then challenged equally. The repeatability of this test under similar conditions as performed here seems reasonably good. Apparently there is no reason why the test should not be equally sensitive and appropriate for tests on vaccines of any level of potency above some minimal level. The use of a stable standard reference vaccine, particularly if diluted constantly to represent an acceptable potency, should make this test quite reliable. The Japanese have chosen to do the test in a different way since they found that with their mice and virus strains tested, deaths by ip injection were irregular, and even with low dilutions of virus it was not possible to kill most of the mice. They therefore avoided the Sabin type test and challenged by the ic route. This much more severe challenge required a larger series of immunizing injections and more time for the imunity to develop. Furthermore, the titers of

VOL. 21, 1971 JBE VIRUS VACCINES 747 the challenge inoculum had to be minimal and very precise. To circumvent this latter difficulty, they produced a reference vaccine of an acceptable potency and depended greatly on the comparison run simultaneously with each lot or lots of vaccine to be released. In comparative tests some Japanese workers have used a mouse antibody test for the determination of vaccine potency. Results of these have closely paralleled the mouse challenge tests, but questions have been raised as to whether antibody is a true measure of protection. Furthermore, this latter test requires even more time and animals for its performance than the ic challenge test but may lend itself to greater quantitative precision. Since Smadel et al. found such poor correlation between the antibody responses of groups of men and the ip mouse challenge tests as then performed, when the MID were in what was considered an acceptable range (0.005 to 0.02 ml), and considered vaccines manifesting such differences as essentially identical, it appeared that the test needed further study and precision in connection with the newer tissue culture vaccines. Our modification of the formerly approved U.S., DBS test was principally the use of a more peripherally invasive, low passage isolate, the Pekin strain. Instead of only being certain of about 80% deaths in control mice of the suitable age by using a fresh 10% mouse brain suspension of the Nakayama strain, a 10-2 dilution of the newer strain would usually kill more than 90%, frequently all mice, and the 10-1 and 103 dilution usually gave similar results. The challenge dose was therefore not in a critical range. We had found furthermore that there was good correlation between the infectivity titration of the virus suspension prior to inactivation and the potency thus determined (3). In the experiments reported here we found quite satisfactory consistency when the same lot of vaccine was tested on more than one occasion or when duplicate tests were run the same day. The lyophilized lot, tested 22 times during 4 yr, had more variability than one would like to see, with a range from 0.011 to 0.027 ml and a standard deviation of 0.004. However, through an error in communication the lyophilization time of this very large lot was cut short, and the product was obviously not thoroughly dried; different degrees of moisture could be observed in different ampoules. We therefore suspect that the stability of the vaccine in different ampoules held at only -20 C probably varied as much or more than the test procedures. However, we feel reasonably certain that a suitably dried product can serve, as it does in Japan, as a standard reference of potency, at a level that every lot of vaccine must meet when tested in a simultaneous test. A comparison of ip challenge with ic challenge with only two lots of vaccine provided evidence that the two methods detected a difference in the same direction, but no more exact comparison was attempted. Theoretically, the ic test would appear to be less exact and perhaps not as meaningful since the virus bypasses most circulating antibody and other body defenses. One of our misgivings about the ip test has been the possible role of nonspecific resistance to the challenge produced by giving the two doses of vaccine ip prior to challenge by the same route. In experiments not reported here, we found that other peripheral routes of challenge, subcutaneous and im, were significantly less effective and consistent. Thus, we gave the vaccine by the im route and used ip challenge. Since no other vaccine was used as a control, the degree of nonspecific resistance afforded by the ip injections of vaccine was not measured, but it was apparent that im injections of the same amount of antigen as given ip were less effective in the test employed and, of greatest importance, gave less concistent results. Since the amount of vaccine that can be given with accurate quantitation by this route in a mouse is restricted, it did not seem practical to pursue this aspect further. In an attempt to understand the mechanism of protection in the ip mouse potency test, the vaccinated mice were examined for neutralizing antibody and interferon. The quantitative and qualitative antibody response of the mice appears quite clear, as might be expected. Antibody of unchallenged mice began to appear by the fourth day, peaked somewhere between the fourth and the 11th day as shown by a level of 1:30 to 1:100 on the 7th day, depending on the concentration of vaccine used, and was lower on the 11th day. Apparently this was immunoglobulin (Ig) M on the 7th day since it was completely 2-ME sensitive. A small amount of IgG began to appear on the 11th and 15th days, but the amount was minimal. After the stimulus of challenge in those bled on day 11, the titer rose significantly, but this was still an IgM response and thus not a true anamnestic type response. We are not surprised with the rapidity of the IgM titer fall shown by the unchallenged mice, since the half-life of mouse antibody is very short as compared to other animal species. Dixon et al. (5) reported the half-life of mouse immune gamma globulin to be 1.9 days only, when injected in an homologous system. Recent studies of Takeya and Nomoto (16) also suggest similar characteristics of

748 SINGH AND HAMMON APPL. MICROBIOL. IgM antibody response in mice. In their experiments with 90-day-old mice, they observed that peak antibody titers developed on about the 7th day followed by a continuous drop till the 13th day. With 10-day-old mice, the pattern of total antibody formation and the peak were somewhat similar, but these young mice did not develop 2- mercaptoethanol-resistant antibodies until 13 days. In view of these findings, the interesting question posed here with the mouse potency test, and as yet unanswered, is whether the rapid elevation of titer after challenge played a role in giving increased protection to the animal or whether protection was adequate from the vaccine alone. A graded challenge might help to clarify this. Interferon apparently played no role in the protection of the mice, under the given experimental conditions (14). The immunization of the mice may have a suppressing effect on the production of circulating interferon as induced by the challenge virus. Whether the minor interferon response detected 24 hr after the virus challenge played any role at all cannot be stated, but obviously the much greater response detected at both 6 and 24 hr in the unvaccinated mice failed to protect them. Thus, any effect it might have had in the vaccinated animals must have been extremely minimal. Apparently this test is a valid measure of immunizing potential, even though the interval before the challenge is short, but it reduces the period required for testing the vaccine. Seroconversion alone may or may not be a sufficient criterion of effectiveness, but this point has not been established. ACKNOWLEDGMENTS We thank Bosko Postic of our faculty for carrying out the interferon assays and Mary White for her excellent technical assistance. The work was carried out under the sponsorship of the Commission on Viral Infections, Armed Forces Epidemiological Board, and was supported by the U.S. Army Medical Research and Development Command, Department of the Army, under contract no. DA-49-193-MD-2042, and by Public Health Service research grant Al-02953 from the National Institute of Allergy and Infectious Diseases. LITERATURE CITED 1. Darwish, M. A., and W. McD. Hammon. 1966. Studies on VI. Developmnent of a hamster kidney tissue culture inactivated vaccine for man. (1) Obtaining maximum titers of virus using an attenuated strain of OCT-541. J. Immunol. 96:691-698. 2. Darwish, M. A., and W. McD. Hammon. 1966. Studies on VI. Development of a hamster kidney tissue culture inactivated vaccinie for man. (2) The characteristics of inactivation of an attenuated strain of OCT-541. J. Immunol. 96:806-813. 3. Darwish, M. A., and W. McD. Hammon. 1966. Studies on VI. Development of a hamster kidney tissue culture inactivated vaccine for man. (3) Potency of an attenuated strain of OCT-541. J. Immunol. 96:878-884. 4. Darwish, M. A., and W. McD. Hammon. 1966. Studies on VI. Development of a hamster kidney tissue culture inactivated vaccine for man. (4) Preparation and characterization of vaccine lot for human trial. Amer. J. Trop. Med. Hyg. 15:765-768. 5. Dixon, F. J., D. W. Talmage, P. H. Maurer, and M. Deichmiller. 1952. The half-life of homologous gamma globulin (antibody) in several species. J. Exp. Med. 96:313-318. 6. Hammon, W. McD., M. A. Darwish, G. E. Sather, and B. Singh. 1970. Response in adults to an inactivated cell culture vaccine employing a highly attenuated strain, OCT-541-35 -24 P1.4-5. Proc. Working Conf. Japanese encephalitis virus vaccine. Igaku Shoin, Ltd, Tokyo. 7. Japanese Ministry of Health and Welfare. 1967. Minimum requirements of biologic products. 8. Koprowski, H., and H. R. Cox. 1946. Studies on chick embryo vaccines against Japanese B encephalitis. J. Immunol. 54:357-370. 9. Nagai, K., and W. McD. Hammon. 1964. Plaque studies with certain group B arboviruses. I. Japanese B encephalitis virus strains on hamster kidney and chick embryo tissue culture. Proc. Soc. Exp. Biol. Med. 117:154-159. 10. National Institutes of Health. 1947. Minimum requirements: Japanese encephalitis vaccine, chick embryo type, dried, p. 1-4. August, 7. 11. Postic, B., C. J. Schleupner, J. A. Armstrong, and M. Ho. 1969. Two variants of sindbis virus which differ in interferon induction and serum clearance. I. The phenomenon. J. Infect. Dis. 120:339-347. 12. Sabin, A. B., C. E. Duffy, J. Waffen, R. Ward, J. L. Peck, Jr., and I. Ruchman. 1943. The St. Louis and Japanese B types of epidemic and encephalitis. Development of noninfective vaccines: report of basic data, J. Amer. Med. Ass. 112:477-486. 13. Schleupner, C. J., B. Postic, J. A. Armstrong, R. W. Atchison, and M. Ho. 1969. Two variants of sindbis virus which differ in interferon induction and serum clearance. II. Virological characterizations. J. Infect. Dis. 120:348-355. 14. Singh, B., and B. Postic. 1970. Enhanced resistance of mice to virulent Japanese B encephalitis virus following inactivated vaccine and poly I: C. J. Infec. Dis. 122:339-342. 15. Smadel, J. E., R. Randall, and J. Warren. 1947. Preparation of Japanese encephalitis vaccine. Bull. U.S. Army Med. Dep. 7:963-972. 16. Takeya, K. and K. Nomoto. 1967. Characteristics of antibody response in young or thymectomized mice. J. Immunol. 11:831-836.