Preinfection Prophylaxis with Herpes Simplex Virus Glycoprotein Immunogens: Factors Influencing Efficacy

Transcription

1 J. gen. Virol. (1989), 70, Printed in Great Britain 3177 Key words: HSV/vaccines/guinea-pig Preinfection Prophylaxis with Herpes Simplex Virus Glycoprotein Immunogens: Factors Influencing Efficacy By LAWRENCE R. STANBERRY, 1. MARTIN G. MYERS, 1 DIMITRIOS E. STEPHANOPOULOS ~ AND RAE LYN BURKE 2 ~ The Division of Infectious Diseases, Childrens Hospital Research Foundation, Elland and Bethesda Avenues, Cincinnati, Ohio and 2Chiron Corporation, Emeryville, California 94608, U.S.A. (Accepted 14 August 1989) SUMMARY Using a guinea-pig model of genital herpes simplex virus (HSV) infection we explored the protection afforded by preinfection immunization with HSV glycoproteins. Glycoprotein immunogens prepared by recombinant DNA technology were found to be as effective as immunogens purified from HSV-infected cell cultures. Immunized animals developed less severe primary disease and also experienced less frequent recurrent infections. Protection was influenced by both adjuvant and route of administration. These studies suggest that recombinant HSV glycoproteins may be effective immunogens for human clinical trials, but that the development of an effective vaccine will require identification of new potent adjuvants that are safe for human use. INTRODUCTION We have been exploring the development of herpes simplex virus (HSV) subunit vaccines designed to modify primary and recurrent genital herpes. In initial studies we reported that immunization with a mixture of HSV glycoproteins provided protection against both primary and spontaneous recurrent disease when administered to guinea-pigs prior to intravaginal HSV-2 inoculation (Stanberry et al., 1987). Using this animal model we also showed that the administration of these glycoproteins, after the establishment of a latent infection, reduced both the frequency and severity of recurrent disease, as well as the frequency of cervicovaginal viral shedding (Myers et al., 1988; Stanberry et al., 1988). In these studies immunogens were delivered with complete Freund's adjuvant in the hindlimb footpad, an adjuvant and route of vaccine delivery not compatible with use in humans. As a prelude to human clinical trials, we explored how the efficacy of glycoprotein immunogens was affected by adjuvant, by route of administration and by immunogen composition. METHODS Animal model. The animal model and the clinical course of HSV disease in the female guinea-pig have been described previously (Stanberry et al., 1982, 1985 a). Female Hartley guinea-pigs weighing between 300 and 400 g were obtained from Charles River Breeding Laboratories. At the time of viral challenge the animals weighed between 550 and 700 g. Challenge on day 0 was with 5.7 loglo p.f.u, of HSV-2 strain MS by intravaginal instillation. To quantify the extent of local viral replication, vaginal swab samples were collected during the acute infection and the virus titre was determined by plaque assay. The clinical course of the acute infection was monitored by daily observation on days 1 to 14. Animals were examined daily for evidence of recurrent herpetic lesions from day 15 to 92, by an observer uninformed as to treatment group. Nineteen animals were killed 98 to 103 days post-inoculation and their lumbosacral dorsal root ganglia were removed and cocultivated on primary rabbit kidney ceils for the detection of latent virus (Stanberry et al., 1985a). Selected animals were reinoculated on day 100 by intravaginal instillation of 5.7 log10 p.f.u, of HSV-2 strain 333 and evaluated for evidence of reinfection (Stanberry et al., 1986) SGM

2 3178 L.R. STANBERRY AND OTHERS Immunogens. The HSV immunogen preparations were similar to those employed in our previous vaccine studies (Pachl et al., 1987; Stanberry et al., 1987; Myers et al., 1988; Sanchez-Pescador et al., 1988). An HSV-2 glycoprotein mixture (gp2) was prepared by detergent lysis of strain 333-infected Veto cells, followed by chromatography of the soluble fraction on lentil lectin-sepharose (Pachl et al., 1987). As determined by radioimmunoprecipitation followed by SDS-PAGE or ELISA, this mixture contained approximately 15 ~ gb, 5 ~o gd, detectable amounts of glycoproteins ge and gg and many unidentified HSV and Vero cell proteins. The recombinant glycoproteins gb and gd were produced by cloning truncated genes obtained from HSV-1 strain Patton into the mammalian cell expression vector psv7d (Burke et al., 1986) and cotransfecting dihydrofolate reductase-deficient (dhfr-) Chinese hamster ovary (CHO) cells with the expression plasmid and a second plasmid, paddhfr, encoding dhfr as a selectable and amplifiable marker (Stuve et al., 1987). The plasmid-encoded glycoprotein derivatives lack the hydrophobic membrane anchor domain and are therefore secreted as glycosylated proteins into the medium. The secreted proteins are 696 and 290 amino acids in length for gb1 and gd1, respectively. Neither protein is produced as a fusion protein. These proteins were purified from the conditioned medium as described previously for gb (Pachl et al., 1987) and gd (Sanchez-Pescador et al., 1988), and were judged to be > 90~ homogeneous by SDS-PAGE. As a control for the recombinant antigen, dhfr- CHO cells were grown to confluence, lysed with detergent and the soluble extract (control extract) was passed over a lentil lectin-sepharose column using the same protocol as used for the gp2 preparation. Complete Freund's adjuvant (CFA), obtained from Sigma, was emulsified with an equal volume of antigen in phosphate-buffered saline. Aluminium hydroxide (alum), obtained from Aldrich, was prepared as described previously by Sanchez-Pescador et al. (1988). Proteins were adsorbed to reach a final 10~ (w/v) ratio. The glycoprotein mixture from virus-infected cells was used at 50 ktg per dose, recombinant gb1 at 20 ttg, gd1 at 10 ~tg and the CHO cell glycoprotein extract at 8 ktg per dose. Animals that received vaccine formulated with CFA were immunized twice on days - 63 and - 28 (with respect to viral challenge on day 0). The remainder of the animals received three immunizations on days -63, -42 and -21. Serology. Antibody to HSV was measured by ELISA using the gp2 mixture as a coating antigen as described previously (Stanberry et al., 1987, 1988). Statistics. Data were analysed by multivariate analysis of variance. Multiple comparisons between groups were assessed by Duncan's multiple-range test and the Kruskal-Wallis one-way analysis of variance (Stanberry et al., 1988). RESULTS Effect of adjuvant and immunogen preparation To determine the impact of adjuvant on the protective efficacy of a subunit vaccine, we immunized 15 Hartley guinea-pigs with gp2. This immunogen was either emulsified with CFA (group I, n = 6) or was adsorbed to alum (group II, n = 9). Animals in group I were immunized twice and those in group II were immunized three times in the hindlimb footpad prior to viral challenge. To assess the effect of route of administration, we also compared the alum-adsorbed gp2 mixture given in the footpad (group II) with subcutaneous (s.c.) administration (group III, n = 8) of the same vaccine. To determine whether a specific viral glycoprotein or mixture of glycoproteins could replace the semi-defined gp2 mixture, additional sets of animals were immunized with gb (group IV, n = 8), or gd (group V, n = 8) or a mixture of gb and gd (group VI, n = 8), all adsorbed to alum. Control animals were immunized s.c. with a CHO cell glycoprotein mixture adsorbed to alum (group VII, n = 11) or were untreated (group VIII, n = 19). Table 1 and Fig. 1 summarize the clinical course of the acute disease after immunization and viral challenge. All animals in the control groups experienced severe primary infection, prolonged urinary retention and a high incidence of mortality (32 ~ for the untreated group and 279/00 for the control extract-immunized group). In contrast, animals that received the gp2 mixture administered with CFA in the footpad (group I) showed no evidence of herpetic disease. Further, this regimen had the greatest effect on peak (day 1) and total viral replication within the vagina (Fig. 2, P < 0.001). All other immunization strategies reduced peak viral replication in the vagina by 37 to 92~ whereas in the CFA-immunized group the peak virus titre was only ~0"003~ (mean = 2.7 log10 p.f.u./ml) that of the combined control groups (mean = 6.3 loglo p.f.u./ml). While the gp2 mixture with CFA provided complete protection from clinical disease, the alum-adsorbed gp2 preparations were less effective. Although immunogens with alum had variable effects on the incidence of skin disease and urinary retention all combinations reduced

3 ,0 o0.5 = 0 Q i = 2.5 ~2-0 1, ,5 Factors influencing HSV subunit vaccines,.~ I I I i ,,/ / \,,,, ~ i i I llolll ll2, i Time (days) i p i i i/.~ i i (c),~ \ -0 / "~, ii)~,, "~ ~ J ~ 1~01~11~21~31' Time (days) Fig. 1. The effect of immunization with HSV glycoproteins on the course of skin disease in HSV-2- infected guinea-pigs. Time refers to days after intravaginal HSV-2 inoculation. (a) Untreated animals (O) and CHO extract controls (11). (b) CHO extract controls (O) and gp2 (11), gb (N), gd (A) or gbgd (O) s.c. administered with alum adjuvant 9, 6 and 3 weeks before intravaginal HSV-2 challenge. (c) CHO extract controls and gp2-immunized animals: gp2/cfa/footpad (11) given 9 and 4 weeks before ItSV challenge; CHO extract (O), gp2/alum/footpad ([]) and gp2/alum/s.c. (A) given 9, 6 and 3 weeks prior HSV challenge. Table 1. Effect of HSV glycoprotein immunogens on primary HSV-2 genital infection in guinea-pigs Dose Group* Immunogen (~g) Adjuvant I gp2 50 CFA II gp2 50 Alum III gp2 50 Alum IV gb 20 Alum V gd 10 Alum VI gd + gb Alum VII CHO extract 8 Alum VIII None - - Skin disease Urinary retention A k Mortality Route Incidence Severity'~ Incidence Duration:~ (~) FP 0/6 0 0/6 0 0 FP 4/ / S.c. 5/ i tl 25 S.c. 7/ '111 7/ "711 0 S.c. 7/8 2.1 _ /8 2-5 _ S.c. 8/ / S.c. 11/ /11 5.3± i ±_ i ± * Vaccines were administered 9, 6 and 3 weeks prior to intravaginal inoculation with 5.7 log10 p.f.u. HSV-2 (MS strain) except group I which was immunized 9 and 4 weeks prior to viral challenge. The immunogens were lectinpurified HSV-2-infected cell lysates (gp2), truncated HSV-1 gb and/or gd produced by recombinant DNA technology or uninfected cell lysates (see text). t Area under the skin lesion score-day curve for days 1 to 14, mean + s.e. Mean days + S.E. FP, Hindlimb footpad. II Significantly different from CHO extract control group, P < both the severity of skin disease and duration of urinary retention (Tables 1 and 2, P < 0-01). For example, in group II (gp2 with alum via footpad), four of nine animals developed skin lesions, although the infection was mild compared to the untreated controls (lesion score compared to , P < 0.001). The recombinant glycoproteins, gb and gd, individually or combined, when administered with alum afforded a level of protection similar to that generated by the gp2-alum preparation (group III). Effect of site of immunization Subcutaneous administration of immunogens was less effective than footpad administration (Table 1). Animals immunized via the s.c. route with gp2 adsorbed to alum (group III) experienced more severe disease (lesion score ) and a longer duration of urinary retention (3.0 +_ 1-0 days) than footpad recipients ( and _ 0-6, respectively) (P <

4 ~4 =3 ~2 L. R. STANBERRY AND OTHERS (a) - i i i i i i i (c) O ~0: "3 ~7 ~6 N5 g4 N3 2 Route* 1 0 Hind footpad Front footpad Intradermal Intradermal No vaccine I I I I I I I (b) [ I I I I L i Time (days) I t I i I"i""~ i Time (days) Fig. 2. The effect of immunization with HSV glycoproteins on HSV-2 replication in the genital tract. Time refers to days after intravaginal HSV-2 inoculation. (a) Untreated animals (O) and CHO extract controls (i). (b) CHO extract controls (O) and gp2 (11), gb (D), gd (z~) or gbgd (0) s.c. administered with alum adjuvant 9, 6 and 3 weeks before HSV challenge. (c) CHO extract controls and gp2-vaccinated animals: gp2/efa/footpad (i) given 9 and 4 weeks before HSV challenge; CHO extract (O), gp2/alum/footpad (V]) and gp2/ alum/s.c. (ZX) given 9, 6 and 3 weeks before HSV challenge. Table 2. Effect of route of administration on gp2 immunogen efficacy Skin disease Urinary retention r ~ ~ -~ Mortality Viral titre Adjuvant Incidence Severityt Incidence Duration;t (~) (day 1)11 Alum 2/ / _ 0.1 Alum 2/ / Alum 3/ / None 9/ ~ 9/ None 4/ / * Lectin-purified HSV-2-infected cell lysate (gp2) (50 ~tg) administered 9, 6 and 3 weeks prior to intravaginal inoculation with 5.7 log10 p.f.u. HSV-2 (MS strain). t Area under the lesion score-day curve for days 1 to 14, mean + s.e. :~ Mean days + s.e. Deaths within 14 days after HSV-2 inoculation. II log10 p.f.u./ml, Mean + s.e. Significantly different from no-vaccine control, P < "001), although the incidence of clinical disease and the magnitude of viral replication were similar for the two groups. Both mortality and urinary retention also occurred less frequently in animals immunized in the footpad. In a second experiment guinea-pigs were immunized with the gp2-alum preparation in the hindlimb footpad, the forelimb footpad or intradermally in the left hip (Table 2). After three immunizations at 3 week intervals, animals were intravaginally challenged with HSV-2 and monitored to determine the course of primary infection. Equivalent protection from skin disease was obtained upon immunization in the hindlimb or forelimb footpad, both routes being slightly superior to intradermal immunization. Alum enhanced the efficacy of intradermal immunogen administration compared to gp2 administered intradermally without adjuvant (P < 0.05). The gp2-alum mixture administered in the hind footpad had a variable effect on the incidence of urinary retention in experiments 1 (Table 1) and 2 (Table 2), but in both experiments this preparation significantly reduced the duration of urinary retention (P < 0.01). Effect of immunization on recurrent disease All immunogen preparations significantly reduced the frequency of recrudescence (Table 3). The greatest reduction in the total number of lesion days (85~) was obtained by footpad

5 Factors influencing HSV subunit vaccines 3181 Table 3. Effect of HSV glycoprotein immunogens on the pattern of recurrent genital HSV-2 infection in guinea-pigs Animals experiencing Days lesions Group* Treatment recurrent disease observedt Recurrent episodes~ Days/episode I gp2/cfa/fp 5/6 3.2 _ II gp2/alum/fp 7/ III gp2/alum/s.c. 5/ IV gb/alum/s.c. 6/ V gd/alum/s.c. 8/ VI gbgd/alum/s.c. 8/ _ VII Control 9/ * Animals immunized as described in Table 1. Animals were examined for recurrent lesions 15 to 92 days after intravaginal HSV-2 challenge. t Total days animals experienced a recurrent herpetic lesion, mean + s.e. :~ Total number of episodes of recurrent disease, mean + s.e. An episode was defined as being preceded and followed by a day without any lesions. Significantly different from control, P < Table 4. Recovery of latent HSV-2 from dorsal root ganglia Virus Total Latent titre Severity of lesion HSV-2 Group* Treatment Animal day It skin disease~ days recoveredll I gp2/cfa/fp a b c d II gp2/alum/s.c, a b c d e VI gbgd/alum/s.c, a b c d e VII No vaccine a b 6-I c d "0 r,rd + e "5 ND -- * Animals immunized as described in Table 1. t logt0 P.f.u./ml recovered from vaginal swab sample on day 1 post-inoculation. :~ Area under skin lesion score-day curve. Recurrences scored on days 15 to 92. II Ganglia were removed 98 to 103 days after HSV-2 inoculation and examined for latent virus by cocultivation on primary rabbit kidney cells; +, virus was recovered; -, no virus was recovered. r~d, Not done because scarring of the genital skin precluded scoring for recurrent disease. administration of gp2 mixture with either the CFA or alum adjuvants. Subcutaneous delivery of immunogens with alum was less efficacious, these animals exhibiting a 43 ~ reduction in lesion days. Effect on latent infection Latent virus was recovered from five of 19 animals as shown in Table 4. All animals from which latent virus was recovered showed both high titre HSV replication in the genital tract during acute infection and frequent recurrent infections. However, as observed previously (Stanberry et al., 1985 a), latent virus was not recovered from all animals that exhibited recurrent

6 3182 L. R. STANBERRY AND OTHERS Table 5. Mean ELISA titres* t Bleedl" A Group Treatment n I gp2/cfa/fp II gp2/alum/fp III gp2/alum/s.c IV gb/alum/s.c V gd/alum/s.c VI gbgd/alum/s.c VII Control/CHO extract 6 < VIII Control/none 7 < _ * Animals immunized as described in Table 1. Geometric mean titres + S.E. of ELISA performed using the gp2 mixture as coating antigen at 10 lag/ml. t Sera were collected 3 days prior to viral challenge on day 18 after the final immunization (1 st) and on days 34 and 95 post-challenge (2nd and 3rd bleeds respectively). All preimmunization sera had titres of < 30. disease. Protection against clinically apparent primary infection did not guarantee protection against latent HSV infection. Effect on antibody response Antibody titres measured by ELISA using the glycoprotein mixture as a coating antigen are shown for all groups of animals in Table 5. Sera were collected 18 days after the final immunization, 3 days before viral challenge (lst bleed) and on days 34 and 95 after viral challenge (2nd and 3rd bleeds, respectively). Immunization in the footpad with immunogen emulsified with CFA induced the highest antibody titres and the magnitude of the response to immunogens given with alum was affected by the site of immunization. Antibody titres achieved by immunization were greater than those acquired by viral infection, ranging from 26-fold for CFA recipients to fivefold and twofold for animals that received the glycoprotein immunogens adsorbed to alum in the footpad and s.c. tissue, respectively. For groups I, II and III, animals immunized with gp2, there was an apparent threshold antibody titre which predicted protection, i.e. 13 of 16 animals with 1st bleed ELISA titres of > 1500 exhibited complete protection from acute disease. In contrast, six of seven animals with 1st ELISA titres of < 1500 developed clinical disease. For the treatment groups IV, V and VI, there was no correlation between disease severity and prechallenge antibody titre, with 22 of 24 animals manifesting herpetic skin disease. Intravaginal HSV-2 rechallenge study We sought to determine whether prior immunization with a glycoprotein immunogen altered the natural history of subsequent HSV-2 genital reinfection. Animals reinoculated with HSV-2 strain 333, 100 days after the initial HSV-2 strain MS challenge exhibited asymptomatic reinfection as characterized by vaginal viral replication without evidence of genital skin disease or urinary retention. Pre-exposure modification of initial and recurrent genital herpes by immunization with either the gp2-cfa preparation administered via the footpad or the gbgdalum mixture given s.c. did not render these animals susceptible to more severe reinfection. DISCUSSION Preinfection immunization with subunit glycoprotein immunogens can reduce symptomatic clinical disease in animals challenged with HSV, although the effect is adjuvant-dependent (Chan, 1983; Schrier et al., 1983; Long et al., 1984; Dix & Mills, 1985; Roberts et al., 1985; Berman et al., 1985, 1988; Dix, 1987; Meignier et al., 1987; Stanberry et al., 1987). We report here that alum enhances the efficacy of HSV glycoprotein immunogens but is a less potent adjuvant than CFA in providing protection against primary disease. However, alum- and CFAcontaining immunogen preparations were approximately equipotent in protecting animals from

7 Factors influencing HSV subunit vaccines 3183 recurrent infections. Using the alum-adsorbed immunogens we observed that animals immunized in the footpad prior to viral challenge have both a lower incidence of disease, as well as less severe disease compared to s.c. immunized animals. Immunogens administered in the forepaw were as effective as in the hind footpad. We have previously shown that other adjuvants including CFA and muramyl dipeptide derivatives show a similar variation of protection with route of administration (Burke et al., 1989). Other investigators have also reported partial protection using alum-adsorbed HSV vaccines (Thomson et al., 1983; Meignier et al., 1987; Phillpots et al., 1987; Berman et al., 1988). Since alum is the only adjuvant for human use currently approved in many countries, these results suggest that alternative adjuvants must be developed in order to achieve significant protection with subunit vaccines. We observed that animals immunized in the footpad exhibited less severe skin disease and urinary retention of shorter duration than animals immunized s.c. This effect of route may be due to the density of antigen-presenting cells, such as Langerhans and dendritic cells present at the site of delivery and/or the residence of the immunogen at the site of deposition. If the enhanced effectiveness of footpad immunization is a consequence of the persistence of the immunogen in a depot, it may be possible to duplicate the efficacy by intramuscular administration. A mixture of recombinant glycoproteins gb and gd was shown to be as immunogenic and protective as a mixture of glycoproteins purified from virus-infected cells. The use of specific immunogens produced by recombinant DNA technology may be a preferable alternative to the use of purified infected cell lysates which are subject to batch-to-batch variability of glycoprotein representation and the potential for viral DNA contamination. All immunization regimens produced antibody responses greater than those resulting from natural infection. Interestingly, no secondary antibody response after viral challenge was observed in immunized animals. This may suggest that the anamnestic response resulting from natural infection is too small to detect in the presence of high titre circulating antibody or, alternatively, the failure to detect secondary responses may simply be a consequence of sampling times with collection of sera 34 days after viral challenge too late to observe anamnestic responses. In animals immunized with the gp2 preparation there was a relationship between the magnitude of the ELISA antibody response and protection from primary herpetic disease. This may be analogous to the inverse correlation of maternal HSV antibody levels and risk of acquiring neonatal herpes (Yeager et al., 1980; Prober et al., 1988). These results suggest it may be possible to use serological data as a predictor of protection. Although immunization protected animals against severe primary disease, the experimental vaccines did not prevent the establishment of persistent viral infection as evidenced by recurrent episodes and the recovery of latent virus from the dorsal root ganglia of animals with frequent recurrences. Detection of latent virus from additional animals might have been enhanced by incorporation of demethylating agents in the culture medium (Stephanopoulos et al., 1988) or by hybridization techniques (Stanberry et al., 1985 b). We and others have previously reported that HSV may also establish a latent infection in guinea-pig genital skin (Stanberry et al., 1985b; Bernstein & Kappes, 1988). If the source of HSV responsible for recurrent infections is virus latent in the periphery then it will be important to determine whether immunization with viral glycoproteins reduces latent infection in skin. Additional studies will be required to determine whether the observed reduction in recurrences is a consequence of a lower level of latent infection in dorsal root ganglia and/or genital skin or rather the suppression of reactivation of latent virus from one of these sites. Immunization with HSV glycoprotein, whether consisting of defined components produced by recombinant technology or of crude viral glycoprotein mixtures, favourably altered the course of initial and subsequent recurrent genital herpes. Efficacy of immunization was influenced by the type of adjuvant and the route of administration. In unimmunized animals recovered from primary HSV-2 genital infection, genital reinoculation resulted in asymptomatic reinfection with lowered viral replication (Stanberry et al., 1986). In this study modification of initial and recurrent genital herpes disease by immunization with viral glycoproteins did not appear to enhance the susceptibility to genital reinfection. These studies suggest that HSV

8 3184 L.R. STANBERRY AND OTHERS glycoproteins may be effective immunogens for human clinical trials, but their efficacy will be dependent upon the identification of more potent adjuvants which are safe for human use. We thank Lisa Sanchez-Pescador, Laura Stuve, Philip Ng, Alisa Reece and Frances Childs for technical assistance and Pauline Montes and Sandra Orlando for manuscript preparation. This work was supported in part by grant AI from the NIH and by the Chiron Corporation. Animals used in this study were maintained in accordance with the guidelines of the Institutional Animal Care and Use Committee of the Children's Hospital Research Foundation. REFERENCES BERMAN, P. W., GREGORY, T., CRASE, D. & LASKY, L. A. (1985). Protection from genital herpes simplex virus type 2 infection by vaccination with cloned type 1 glycoprotein D. Science 278, BE~N, P. W., VOLT, P. E., GREGORY, T., LASKV, L. A. & KERN, E. R. (1988). Efficacy of recombinant glycoprotein D subunit vaccines on the development of primary, recurrent and latent genital infections with herpes simplex virus type 2 in guinea pigs. Journal of Infectious Diseases 157, BERNSTEIN, D. L & KAPPES, J. C. (1988). Enhanced in vitro reactivation of latent herpes simplex virus from neural and peripheral tissues with hexamethylenebisacetamide. Archives of Virology 99, BURKE, R. L., PACHL, C., QUIROGA, M., ROSEN'BERG, S., HAIGWOOD, N., NORDFANG, O. & EZBAN, M. (1986). The functional domains of coagulation factor VIII. Journal of Biological Chemistry 261, BURKE, R. L., VAN NEST, G., CARLSON, J., GERVASE, B., GOLDBECK, C., NG, P., SANCHEZ-PESCADOR, L., STANBERRY, L. R. & OTT, G. (1989). Development of a herpes simplex virus subunit vaccine. In Vaccines 89. Edited by R. M. Chanock & R. A. Lerner. New York: Cold Spring Harbor Laboratory (in press). CHAN, W. L. (1983). Protective immunization of mice with specific HSV-1 glycoproteins. Immunology 49, DIX, S. D. (1987). Prospects for a vaccine against herpes simplex virus types 1 and 2. Progress in Medical Virology 34, DIX, a. D. MILLS, J. (1985). Acute and latent herpes simplex virus neurological disease in mice immunized with purified virus-specific glycoproteins gb and gd. Journal of Medical Virology 17, LONG, D., MADARA, T. S., PONCE DE LEON, M., COHEN, G. H., MONTGOMERY, P. C. & EISENBERG, R. J. (1984). Glycoprotein D protects mice against lethal challenge with herpes simplex virus types 1 and 2. Infection and Immunity 43, MEIGNIER, B., JOURDIER, T. M., NORRILD, B., PEREIRA, L. & ROIZMAN, B. (1987). Immunization of experimental animals with reconstituted glycoprotein mixtures of herpes simplex virus 1 and 2: protection against challenge with virulent virus. Journal of Infectious Diseases 155, MYERS, M. G., BERNSTEIN, D. I., HARRISON, C. J. & STANBERRY, L. R. (1988). Herpes simplex virus glycoprotein treatment of recurrent genital herpes reduces cervicovaginal virus shedding in guineas pigs. Antiviral Research 10, PACItL, C., BURKE, R. L., SANCHEZ-PESCADOR, L., VAN NEST, G., MASIARS, F. & DINA, D. (1987). Expression of cellassociated and secreted forms of herpes simplex virus type 1 glycoprotein B in mammalian cells. Journal of Virology 61, PHILLPOTS, g. J., WELCH, M. G., RIDGEWAY, P. H., WALKLAND, A. C. & MELLING, J. (1988). A test for the relative potency of herpes simplex virus vaccines based upon the female guinea-pig model of HSV-2 genital infection. Journal of Biological Standardization 16, PROBER, G. G., SULLENDER, W. M., YASUKAWA, L. L., AU, D. S., YEAGER, A. S. & ARVIN, A. M. (1988). Use of routine viral cultures at delivery to identify neonates exposed to herpes simplex virus. New England Journal of Medicine 316, ROBERTS, P. L., DUNCAN, B. E., RAYBOULD, T. J. G. & WATSON, D. H. (1985). Purification of herpes simplex virus glycoproteins B and C using monoclonal antibodies and their ability to protect mice against lethal challenge. Journal of General Virology 66, SANCHEZ-PESCADOR, L., BURKE, R. L., OTT, G. & VAN NEST, G. (1988). The effect of adjuvants on the efficacy of a recombinant HSV glycoprotein vaccine. Journal of Immunology 141, SCHRIER, R. D., PIZER, L. I. & MOORHEAD, J. W. (1983). Type-specific delayed hypersensitivity and protection immunity induced by isolated herpes simplex virus glycoprotein. Journal of Immunology 130, STANBERRY, L. R., KERN, E. R., RICHARD, J. T., ABBOTT, T. M. & OVERALL, J. C., JR (1982). Genital herpes in guinea pigs: Pathogenesis of the primary infection and description of recurrent disease. Journal of Infectious Diseases 146, STANBERRY, L. R., KERN, E. R., RICHARD, J. T. & OVERALL, J. C., JR (1985a). Recurrent genital herpes simplex virus infection in guinea pigs. Intervirology 24, STANBERRY, L. R, KIT, S. & MYERS, M. G. (1985b). Thymidine kinase-deficient herpes simplex virus type 2 genital infection in guinea pigs. Journal of Virology 55, STANBERgY, L. R., BERNSTEIN, D. I., KIT, S. MYERS, M. G. (1986). Genital reinfection after recovery from initial genital infection with HSV-2 in guinea pigs. Journal of Infectious Diseases 153, STANBERRY, L. R., BERNSTEIN, D. I., BURKE, R. L., PACHL, C. & MYERS, M. G. (1987). Vaccination with recombinant herpes simplex virus glycoproteins: protection against initial and recurrent genital herpes. Journal of Infectious Diseases 155,

9 Factors influencing HSV subunit vaccines 3185 STArOERRY, L. R., BURKE, R. L. & MYra, S, M. t~. (1988). Herpes simplex virus glycoprotein treatment of recurrent genital herpes. Journal of Infectious Diseases 157, STEPHANOPOULOS, D. E., KAPPF.S, J. C. & BERNSTEIN, D. I. (1988). Enhanced in vitro reaction of herpes simplex virus type 2 from latently infected guinea-pig neural tissues by 5-azacytidine. Journal of General Virology 69, STUVE, L. L., BROWN*SHIMER, S., PACHI,, C., NAJARIAN, R., DINA, D. & BURKE, R. L. (1987). Structure and expression of herpes simplex type 2 glycoprotein gb gene. Journal of Virology 61, THOMSON, T. A., HILFENHAUS, J., MOSER, H. & MORAHAN, P. S. (1983). Comparison of effects of adjuvants on efficacy of virion envelope herpes simplex virus vaccine against labial infection of BALB/c mice. Infection and Immunity 41, YEAGER, A. S., ARVIN, A. M., URBANI, L. J. & KEMP, J. A. (1980). Relationship of antibody to outcome in neonatal herpes simplex virus infections. Infection and Immunity 29, (Received 3 April 1989)