with monoclonal antibodies. encephalitis virus by passive ...

0022-1767/88/14110-3606$02.00/0

THEJOURNAL OF IMMUNOLOGY

Vol. 141.3606-3610. No. 10. November 15. 1988 Prlnted In U . S . A .

Copyright Q 1988 by The American Associationof Immunologists

PROTECTION OF MICE AGAINST JAPANESE ENCEPHALITIS VIRUS BY PASSIVE ADMINISTRATION WITHMONOCLONAL ANTIBODIES J U N K 0 KIMURA-KURODA’

AND

KOTARO YASUI

We have identified and characterized nine antireaction in a N test. It has been proposed that by crossgenic epitopes on the E envelope of Japanese en- reactive antibodies ADE might be responsible for severe cephalitisvirus (JEV) by using mAb. Passive admin- DEN diseases, hemorrhagic fever, and shock (2).There istration of most of theanti-JEVmAbprotected has not been distinct evidence that severe diseases are mice fromi.v. challenge with 1.5 x lo3plaque-form- caused by ADE, although increased viremia was seen in in$ units of JEV, JaGAr-01 strain. Some mAb, whichDEN-infected monkeys that had been pretreated with possess high neutralization activity in vitro, showed anti-DEN serum (3). Barrett and Gould reported early high protection, and JEV-specific N mAb 503 was death of yellow fever virus-infected mice was caused by found the most protective. Evenan injection of 2.5 specific antibody (4). It has also been reported that the &mouse of mAb 503 protected all micefrom JEV infection. Furthermore, an injection of about 200pg early death seen in rabies-infected mice may be caused and a similar phenomenonhas been observed of mAb 503 on day 5 postinfection protected 82%of by ADE (5), for feline infectious peritonitis virus(6). the mice, even when JEV was detected in more than Analysis of the antigenic epitopes of major structural 85% of theinfectedmousebrains.Synergism of envelope glycoprotein E of flaviviruses provides the most protection was observed with mixtures of several mAb directed against different epitopes. Although fundamental information, not only for development of a in a murine macrophage cell line, all of the mAb new vaccine, but also for elucidation of the mechanism of infection. The important biologic functions of hemaggroupsshowedantibody-dependentenhancement (ADE) of JEV infectivity in vitro, and only two fla- glutination and N are attributableto this protein, as well vivirus cross-reactive mAb groups showed ADE of as antigenic specificities that range from specific to dengue virus type 2. The ADE of JEV by mAbseems cross-reactive. The antigenic analysis of E proteins of not to be harmful for in vivo protection experiments, several flaviviruses by mAb has brought to light several exceptfortwo mAb groups:mAb 302 and201 important areas of concern: 1) a critical protective or N showed little or no protective activity againstJEV epitope (7)or severalepitopes (8);2) synergistic antibodies infection and, rather, caused early death in infected (9, 10); 3) protection in vivo with non-neutralizing antimice. body (11); and 4) effects of ADE of virus infectivity (12). We have previously determined and characterized antigenic epitopes of the E protein of JEV by mAb (10, 13). JEV3 is one of about 60 distinct viruses inthe flavivirThe existence of at least eight epitopes on the E protein idae (1) and has high pathogenicity for humans. Forwas suggested, one of which was the critical N site and malin-inactivated JEV vaccine,which is made from showed JEV specificity. suckling mouse brain homogenate, is used now in epiIn this report, adding another mAb group, we deterdemic areas, butas a vaccine it is accompanied by several mined, first, ADE of JEV or DEN virus infectivity by antiproblems: high cost, no guarantee of availability to all JEV mAb and polyclonal anti-virus sera on a murine strains of JEV, and possibility of allergic effects. Furmacrophage cell line, P388D1.and, second, the protective thermore, in Southeast Asia, overlapping epidemics of activity of mAb and polyclonal antibodies from JEV inJEV and DEN viruses,which belongs to a separate fection in mice. The protective and nonprotective episubgroup of flaviviruses, have posed a serious hazard to topes on Eprotein of JEV are discussed. public health. This has complicated the aspects of both vaccination and hostimmunologic response. Flaviviruses MATERIALS AND METHODS show high cross-reactions in a HI test but low crossReceived for publication April 21. 1988. Accepted for publication August 15, 1988. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked adoertisernent in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‘This research was supported by Grants 61570233 and 63570217 from the Ministry of Education, Science and Culture of Japan. Please address all correspondence and reprint requests to: Dr. Junko Kimura-Kuroda, Department of Microbiology, Tokyo Metropolitan Institute for Neurosciences, 2-6Musashi-dai, Fuchu-shi. Tokyo. Japan. 183. Abbreviations used in this paper: JEV. Japanese encephalitis virus; ADE, antibody-dependentenhancement; N, neutralization: HI. hemagglutination inhibition; DEN, dengue: CBA. competitive binding assay.

Viruses and mice. JEV. JaCAr-01 strain, was propagated in suckling mouse brains, and supernatantof brain homogenates was used as the stock virussolution for the mice protection experiment. JEV, JaGAr-01 strain andDEN virus type 2 (DEN-2).Tr. 1751 strain were propagatedin C6/36 cells cultivatedin Eagle’s MEM containing 10% heat-inactivated fetal bovine serum and nonessential amino acid (GIBCO. Grand Island, NY). The infected culture fluids were for the ADE exueriment. BALB/C/AnNCrLBr used as the stock virus mice (males, 3 wk old) were obtained from Shizuoka Laboratories (Shizuoka. Japan). Antibodies. Anti-JEV mAb were prepared as described previously (10, 13). The specificity of mAb was determined by radioimmunoprecipitation of JEV-infected cell lysates, and the lg subclass was determined by Ouchterlony gel diffusion test (13).Anti-JEV polyclonal hyperimmune sera was obtained from BALB/C mice (8wk

3606

3607

PROTECTION OF MICE AGAINST J E V BY mAb old) given two or three i.p. injections of formalin-inactivated JEV vaccine (Nakayama-NIH strain). Purified lg fractions were obtained from ascitic fluid of each mAb and anti-JEV polyclonal serum by ammonium sulfate precipitation, followed by chromatography over Sepharose-protein A. The antibodyprotein concentration was determined by using the Bio-Rad protein assay (Bio-Rad Laboratories, Richmond, CA). Mouse anti-DEN-2 sera wasprovided by Dr. S. Inoue of the Institute of Public Health of Japan. Mice were immunized by i.p. inoculation of supernatant of DEN-2-infected mouse brain homogenate emulsified with complete Freund adjuvant four times a t 1-wk intervals. Antibody characterization. N test, HI test, ELISA and CBA were performed as previously described in detail [ 10, 13). N titers were measured inVero cells by using the50%plaque reduction end point (14). The mAb and virus mixture was incubated in Eagle's MEM containing 20% Haemaccel [Hoechst-Russell Pharmaceuticals, Somenrille. N J ) and 25 mM HEPES, pH 7.3 (dilution buffer). a t 37°C for 1 h. HI test was done by using pigeon red blood cells. Purified live virus from infected-C6/36 cells was used as Ag in ELISA and CBA. Flavivirus cross-reactivities were examined using six other flaviviruses. there of the same subgroup(Murray Valley, West Nile, and St. Louis, encephalitis) and three more of different subgroups (DEN-1, DEN-2,and yellow fever viruses). ADE experiment. A murine macrophage cell line, P388D1, was provided by Dr. T. Masuda at Kyoto University, cultured in Dulbecco's modified Eagle MEM (GIBCO) with 10% fetalbovine serum and antibiotics (growth medium). About 5 x lo5 cells were distributed into wells of 24-well tissue culture plates (Nunc Laboratories, Roskilde. Denmark). The next day, the purified mAb and polyclonal antibodies wereseriallydiluted in the dilution buffer described above. Each diluted antibody was incubated witha n equal volume of JEV suspension for 30 min 37°C. at A mixture of 100 pg of normal mouse IgG2a and IgG2b (Cappel Laboratories, Malvern,PA) or diluted normal mouse serum (lo-') was used as control. The antibody and virus mixtures were added to the P388D1 cell monolayer a t a final multiplicity of infection of 0.1 to 0.2. After incubation for 1 h a t 37°C. the mixtures were removed, and growth medium containing the same concentration of anti-JEV antibodies was added to the cells and cultured at 37°C in a humidified 10% COz incubator. On days 2, 4. and 6.culture fluids were removed and stored a t -70°C for the virus titration. Tenfold dilutions of culture fluid wereassayed for plaque formation onVero cells (10). The same method was used for ADE of DEN-2 virus and JEV. The virus titration DEN-2 for virus was performed by a modified focus assayof the peroxidase-staining technique(13). ADE titers weredefined as thehighest dilution exhibiting a more than fivefold increase in virus titer compared with controls (normalmouse IgGZa and IgGZb, or normal mouse serum). All values are mediansof two or three replicate experiments. Peak multifold enhancement was obtainedby dividing virus titers in test mAb or serumby virus titers incontrols. Passlue anttbody admintstration. Mice were inoculated i.p. with purified anti-JEV antibodiesdiluted with PBS, pH 7.3. The next day the animals were challenged i.v. with 3 X lo3 plaque-forming units of JEV JaGAr-01 strain diluted in dilution buffer as described above. Control mice received i.p. >lo0 pg of normal mouse IgG2a and IgG2b. or normal mouse 1gG (Cappel), or PBS, and the next daywere challenged i.v. by a comparable dose of JEV. Survival rates were calculated from two or more independent observations, in which 8 of 10 mice were used as one experimentalgroup. For the virusabrogation experiments, various concentrations of antibody were injected i.p. on days 1 to 5 after i.v. inoculation of 3 X lo3 plaque-forming units of JEV. and the mice were observed for 4 wk. On day 5 after JEV inoculation. the brainsof infected mice were dissectedto detect virus growth. The supernatant of brain homogenates was assayed for virus titration.

TABLE I Anti-JEV mAb andooluclonal antibodies used" ~~

Antibodies

mAbb IgG Group Subclass

~

Crossreactivity

Biologic Activities'

N

Monoclonal 30 1 109 112 503 N.04 20 1 204 302 504

1 2 3 8 7

6 5 4 9

Flavivirus 2a 2a Subgroup 2a JEV specific 2a specific JEV Subgroup 2b Subgroup 2a Subgroup 2a Subgroup 2b 2a Flavivirus

200

ND ND' 100 ND 67 82

'

1

'

In Brain"

0 ND

Virus positive/total. Parenthesisshowsmean brain homogenates. Percentage of survivors. Not done.

L

Virus Growth

ND 25 17

ND ND 7 / 8 [ 103.6)

virus-titer (PFU) of

mAb N.04(group 7) andJEV-specific mAb 112 (group 3). In mouse protection experiments, JEV-specific high N mAb 503 was found the most protective among all mAb Figure 1 . Earlier death of JEV-infected-mice by passive administratested and polyclonal anti-JEV IgG. MAb 112 and N.04 tion with certain mAb. The mice were administered i.p. with 2 0 0 p a mouse of each mAb 1 day before JEV inoculation. Symbols: 0, control were the next strongest in effectiveness. The sites rec(normal mouse IgC):0, mAb 201: A, rnAb 302: A, mAb 109:0. mAb 503. ognized by mAb503, 112, andN.04 are surely important Similar results were obtained from two or more independent observations. for protection against JEV infection. The resultsof day 5 injection indicate that mAb 503 protected mice in which JEV replication had already begun within the central A nervous system.This mAb may prove useful a s a medical lool--"-treatment for Japanese encephalitis as well as for its prevention. A s described in several reports(7, 15). the in vivo protection activity is likely to parallel in vitro N activity, and thisis confirmed by the following evidence. The synergistic effect on the protection activity of mAb was also observed by using mixtures of several mAb groups, a s expected from the synergistic N activity against JEV as previously described (10, 13). Similarly, the rather higher protective activity of polyclonal antiJEV might be caused by this synergistic protection. This phenomenon may be important for protection from JEV infection in vivo. 0 Some non-N mAb, 301, 109, and 504, showed partial protective activity. In these cases, theN activity does not always parallel the protective activity. The mechanism of partial protection by a non-N mAb is not clear, but similar phenomena were observed in several viruses(1 1, 16, 17). C-dependent N or cytolysis of JEV-infected cells may be mediated by these non-N mAb. Several reports haverevealed that not all mAb against E protein of other flaviviruses always cause ADE of virus infection. In some cases, ADE was seen by only virusspecific mAb (9, 12, 18). whereas in one caseit was only by a cross-reactive mAb (19). In JEV, all mAb groups tested, which included both JEV specific and cross-reactive, showed ADE activities in P388D1 cells (Table 11). 3 0 1 109 1 I2 503 N O 4 201 204 302 504 There may be some epitopes recognized by mAb on E Anti -JEV MAbs protein of JEV thatdo not result in ADE. Figure 2. Survival rates of JEV infected-mice administered with antiIt is unclear whether ADE influences in vivo infection JEV mAb. MAb (100 pg/mouse) were given 1 day before JEV inoculation of JEV. In JEV, macrophages or monocytes seem not to ( A ) .or 1 day after inoculation (B). be important targets of virus infection. In fact, severe the specific anti-JEV activity in ELISA (200 to 800)and disease or early death was not observed by passive adN activity (20to 40) was firstdetected on day 7, but was ministration witha just diluted concentration of antibodprotective activity except mAb 302 not foundon day 5 in ELISA ((50) and inN activity (