INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Mucosal surfaces are a favored entry site for numerous pathogenic microorganisms. Infections with some of these organisms remain localized to the mucosal epithelium, while others spread systemically. The mucosal entry points are thought to be guarded by local mucosal immune responses, but systemic immune protection also can extend into the mucosa. This is particularly true of humoral immunity; antibody bathes interstitial spaces and can pass through the mucosa as a transudate from serum. Herpes simplex virus type 2 (HSV-2) is a common human pathogen that enters the body primarily via the genital mucosa. HSV-2 replicates in the genital epithelium and spreads to lumbosacral sensory ganglia, where latent infection is maintained for the life of the individual. Periodic reactivation results in reinfection of the genital epithelium innervated by the infected dorsal root ganglia (DRG). Prophylactic immunization ideally would reduce infection of the genital epithelium and prevent latent infection of the ganglia, thereby eliminating the recurrent HSV-2 infections that provide opportunities for transmission to sex partners and newborns (60), as well as provide portals of entry for other pathogens such as human immunodeficiency virus (6, 11, 49). An understanding of how the response to immunization protects mucosally and systemically against subsequent HSV-2 genital infection would further the development of vaccines against sexually transmitted diseases, and HSV in particular.

HSV-2 infection of the genital mucosa elicits HSV-specific immunoglobulin G (IgG) and IgA in the genital tracts of both humans (1) and mice (25, 27, 35, 44). HSV-specific IgG, but not IgA, can also be detected in genital secretions after parenteral immunization of mice (36, 56). Using a mouse model of genital infection (27), numerous investigators have demonstrated an inability of passively transferred immune serum to reduce infection of the genital mucosa by HSV-2 (25, 45, 51) or HSV-1 (14, 15). Only Parr and Parr (45) have observed that serum IgG collected from mice immunized intravaginally (i.vag.) with attenuated HSV-2, purified, and injected into naive mice can decrease HSV-2 replication in the genital mucosa. Some studies have demonstrated that development of genital disease after vaginal challenge can be retarded by transfer of immune serum (14, 15, 45), though the mechanism mediating this form of protection is not known. Most controversial is the role of HSV-specific serum IgG in protection of the nervous system. Studies using corneal and footpad routes of challenge with HSV have indicated no decrease in latent infection in mice receiving immune serum (41, 61). Using the genital route of challenge, Schneweis et al. demonstrated a decrease in the number of acutely and latently infected DRG upon transfer of immune serum to naive recipients (51), an observation that was recently confirmed in HSV-immune, B-cell-deficient mice to which immune serum was passively transferred (13). Mortality has been influenced by passively transferred immune serum in some studies (29, 41), while others have suggested that immune serum does not influence survival rate (25). These studies suggest that immune serum antibody generated by infection of mice with wild-type virus or thymidine kinase (TK) mutants of HSV can influence protection, but they provide little information about the protective capacity of vaccine-generated antibody. Second, the capacity to reduce latent genome loads after vaginal challenge with HSV has not been quantitatively assessed, whether after vaccination with replication-competent HSV or with a form of vaccine.

Several approaches to live virus vaccination against HSV have been developed, including vector-encoded HSV glycoproteins (5, 8, 26) and attenuated viruses that contain additional copies of glycoprotein genes (30). These replication-competent viruses elicit antibody responses and protect mice against HSV-2 challenge, but the role of antibody in vaccine-mediated protection has not been elucidated. Replication-defective mutants of HSV-2 have been developed as viable prototypes of a safe form of vaccine that elicits both humoral and cellular responses (4, 9, 36). Using the replication-defective, ICP8 mutant 5BlacZ (9) as a vaccination paradigm with potential for human application, we have demonstrated that mice immunized subcutaneously (s.c.) are efficiently protected from genital and systemic disease upon i.vag. challenge with virulent HSV-2 and that replication of the challenge virus in the genital epithelium is dramatically reduced (36). Systemic immunization with replication-defective HSV-2 elicits an HSV-specific serum IgG response with low levels of IgG detectable in the genital tract; HSV-specific IgA cannot be detected (36) (unpublished data). With a combination of serum passive transfer to SCID mice, and to naive and immune B-cell-deficient mice, we thoroughly assess the role of serum IgG in protecting against HSV-2 genital infection in isolation or in combination with immune T cells. Our study is novel in three ways. First, we investigate the role of vaccine-induced serum antibody in reducing mucosal and systemic infection. Second, we use an extremely sensitive PCR assay to quantitate viral DNA in latently infected DRG when latency is established in the presence or absence of immune serum. Third, we identify an important interaction between the humoral and cellular arms of immune defense in the genital tract.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cells and viruses. The HSV-2 replication-defective mutant strain 5BlacZ (9) was propagated in S-2, a Vero cell line that stably expresses ICP8 (19). Virus used for inoculation into mice and for antibody neutralization assays was partially purified by harvest of extracellular virus from infected cell monolayers as previously described (38). Supernatant from uninfected cell cultures was prepared as a control. HSV-2, strain G-6, was isolated by plaque purification of strain G and was propagated in Vero cells. Virus titers were determined by standard plaque assay (38).

Animals, immunization, and challenge. C57BL/6 (B6) and BALB/c mice were purchased from the National Cancer Institute. CB.17-SCID mice were purchased from the National Cancer Institute or were kindly provided by Dan Hoft, Saint Louis University. Igh-6, B-cell-deficient (µMT) mice (23), on a B6 genetic background, were used with permission of Werner Muller (Institute for Genetics, University of Cologne, Cologne, Germany) and were kindly provided by Skip Virgin, Washington University, St. Louis, Mo. SCID and µMT mice were bred at the Saint Louis University animal facility and were housed under specific-pathogen-free conditions in sterile microisolator cages in accordance with institutional and federal guidelines. All mice were used beginning at 6 weeks of age. Experimental procedures were approved by the institutional Animal Care and Use Committee.

Virus neutralization assay and ELISA. Blood was collected from the tail veins of immunized mice 5 days prior to challenge. Complement-independent neutralizing antibody titers in the sera were determined by 50% plaque reduction assay as previously described (37). HSV-2-specific IgG titers in sera were determined by enzyme-linked immunosorbent assay (ELISA) on plates coated with lectin-purified HSV-2 G-6 glycoproteins (42) as previously described (36).

Intracellular cytokine staining. Genital lymph node cells were prepared as described above. Cell surface staining with fluorescein isothiocyanate-conjugated anti-CD4 (Caltag) and anti-CD8 (Caltag) and intracellular staining with phycoerythrin-conjugated anti-IFN- (PharMingen) were carried out using a CytoStain kit with GolgiStop (PharMingen) according to the manufacturer's directions.

Virus replication. Acute replication of challenge virus in the genital mucosa was assessed as previously described (38). To determine virus titer in neural tissues, samples were placed in microcentrifuge tubes with no. 1 glass beads and filled with phosphate-buffered saline. Samples were frozen, thawed, disrupted in a Mini-Bead Beater (BioSpec, Inc.), and diluted for standard plaque assay.

Clinical disease. Signs of inflammation and disease of the external genitalia and signs of neurologic disease were monitored daily and were scored in a masked manner to avoid bias. Disease was recorded on a scale of 0 to 4: 0, no apparent disease; 1, slight swelling and erythema of the genitals; 2, marked inflammation of the genitals; 3, purulent lesions on the genitals; 4, bilateral hind limb paralysis; 5, death. Mice were weighed daily postchallenge, and mean weight change ± standard error of the mean (SEM) compared with initial body weight was calculated daily for each group.

Serum passive transfer. BALB/c and B6 mice were immunized three times s.c. with 5BlacZ. Blood was collected 8 to 13 days after the final immunization, and serum obtained from individual mice was pooled. Titer of HSV-specific IgG was determined by ELISA, and 0.15 to 0.3 ml (depending on titer) was transferred by intraperitoneal injection into recipient mice 3 h before and 3 days or 3 and 6 days after challenge. Two mice per group were bled posttransfer to determine the level of HSV-specific antibody in recipients compared to mice immunized once with 5BlacZ.

Detection of HSV-2 DNA by nested PCR. A limiting-dilution, nested PCR assay was developed to detect the TK gene of HSV-2 at single-copy sensitivity. The sequences of the outer PCR primers (GIBCO BRL) used were 5'-TGGATTACGATCAGTCGCC-3' and 5'-ACACCACACGACAACAATGC-3', which amplify a 235-bp product. The sequences of the inner PCR primers used were 5'-ATGATCCCAACCCGCGTCACAA-3' and 5'-TTTATTGCCGTCATCGCCGGGA-3', which amplify a 180-bp product. Tenfold dilutions of ganglionic DNA samples were added to PCR mixtures containing 50 mM KCl, 10 mM Tris-HCl (pH 9.0), 0.1% Triton X-100, 2.5 mM MgCl₂, 0.2 mM deoxyribonucleoside triphosphates, 10 pmol of each primer, and 1 U of Taq DNA polymerase. The first round of PCR was performed in a 25-µl volume and subjected to a hot start at 94°C for 2 min, followed by denaturation at 94°C for 45 s, annealing for 20 s at 60°C, and extension at 72°C for 30 s. Thirty-five cycles of amplification were followed by a final extension at 72°C for 7 min. For the second round, 1 µl of the first-round PCR product was amplified in a total volume of 20 µl under conditions otherwise identical to those for the first round. Second-round PCR products were visualized by electrophoresis on 2% agarose gels stained with ethidium bromide. Plasmid pEH48 (David Knipe, Harvard Medical School) containing the HSV-2 TK gene was used as a standard to determine the sensitivity of the nested PCR for detection of HSV-2 DNA. This limiting-dilution, nested PCR method was shown to have a sensitivity of one copy of HSV-2 DNA in a background of 0.5 µg of herring sperm DNA by adding known concentrations of plasmid pEH48. pEH48 was quantitated by spectrophotometry, and 12 replicates were serially diluted into herring sperm DNA. The plasmid dilutions were subjected to two rounds of PCR, and the dilution in which 63% of replicates were positive was considered to contain one copy.

PCR amplification of HSV-2 genome in latently infected DRG. Twenty-eight days after challenge, the DRG (L2-S4) of surviving mice were dissected and pooled for individual mice. DNA was extracted as previously described (22). To determine the frequency of HSV-2 genome occurring in the DRG of infected mice, nested PCR was performed on serial dilutions of ganglionic DNA prepared from mice at 28 days postchallenge as follows. DRG were dissected, pooled for each mouse, and digested overnight at 37°C with proteinase K (100 µg/ml) in digestion buffer (10 mM Tris-HCl [pH 7.4], 20 mM EDTA, 0.5% sodium dodecyl sulfate). The samples were phenol-chloroform extracted, ethanol precipitated, and resuspended in a total volume of 20 µl of Tris-EDTA. Each sample was assayed in replicates of 12. Nested PCR was performed as described above. The controls for one-copy sensitivity, as well as negative controls of water and irrelevant plasmids, were performed for each set of PCRs. To avoid contamination, sterile instruments were used, dissections were performed under a biosafety cabinet, and all pipettors and reagents including phenol and chloroform were dedicated exclusively to the nested PCR.

Statistical analyses. Differences in viral titers between groups on individual days were determined by t test. The number of latent genomes between groups of mice was compared by chi-square analysis. The nonparametric Kruskal-Wallis test was used to determine the statistical significance of differences in disease scores between multiple groups.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Passive transfer of immune serum to SCID mice. To evaluate the individual role of HSV-specific antibody induced by replication-defective HSV-2 vaccine strain in protection of the genital tract and nervous system in the absence of other effector mechanisms, serum from wild-type mice immunized with 5BlacZ was transferred to CB.17-SCID mice. Groups of mice received two or three injections of immune serum 3 h prior to i.vag. challenge and once or twice at 3-day intervals after challenge with virulent HSV-2 G-6. Replication of challenge virus in the genital tract epithelia of immune serum recipients was not altered compared to SCID mice receiving control serum (Fig. 1). Development of genital disease (scores of 0.5 to 3.0) and neurologic disease (scores of 3.5 to 5.0) was delayed a statistically significant amount, however, in mice receiving two or three injections of immune serum (Fig. 2A; P  0.021 for days 5 through 11). Weight loss also occurred more slowly in mice receiving immune serum (data not shown), and survival was prolonged in mice receiving two or three injections of immune serum (Fig. 2B). The delay in severe disease and death seen in mice receiving two or three injections of immune serum was not due to increasing concentration of HSV-immune antibody in the blood because comparison of ELISA titers on days 1 and 7 posttransfer revealed that the serum concentration of specific antibody in recipients remained relatively steady over this period (data not shown; P = 0.34 for day 1 compared with day 7). The half-life of antibody in serum has been estimated at 4 days (24). Interestingly, the titers of challenge virus in the lumbosacral spinal cords of control SCID mice or mice receiving immune serum did not differ at days 3, 4, and 5 postchallenge (data not shown). By day 6, however, virus titers in thoracic spinal cord, brainstem, and brain were all lower in mice receiving immune serum than in control mice (Fig. 3), suggesting that antibody delayed development of neurologic disease by reducing spread to, or replication within, spinal cord and central nervous system (CNS) tissues. By day 6 postchallenge, virus could be detected only rarely in spleen, liver, or lungs (data not shown). Thus, in SCID mice, replication-defective virus-immune serum antibody did not affect primary replication but did prolong time to development of genital and neurologic disease.

View larger version (23K): [in this window] [in a new window]   FIG. 1.   Replication of challenge virus in the genital mucosae of SCID mice after transfer of immune serum. Groups of three to four CB.17-SCID mice that received two or three injections (at 0 and 3 days or 0, 3, and 6 days, respectively) of 5BlacZ-immune serum were challenged i.vag. with HSV-2 G-6. Titers of virus collected in vaginal swabs were determined by standard plaque assay. Data are from one of two experiments and represent the geometric mean titer ± SEM for all mice per group.

View larger version (22K): [in this window] [in a new window]   FIG. 2.   Delay of disease and death in SCID mice receiving HSV-immune serum. Groups of SCID mice, as described in the legend to Fig. 1, were observed daily for signs of genital and neurologic disease (A) and survival (B). Disease data are expressed as mean ± SEM for all mice per group.

View larger version (16K): [in this window] [in a new window]   FIG. 3.   Virus titer in the nervous system after genital challenge. Groups of four to five SCID mice were injected twice with 5BlacZ-immune or control serum as described for Fig. 1. Six days after challenge, the spinal cord, brainstem, and brain were removed, and virus titer in the tissues was determined by standard plaque assay.

Passive transfer of HSV-immune serum to B-cell-deficient mice. We next examined whether replication-defective virus-specific antibody had an effect on protection from genital challenge in mice with functional cellular immunity, but whose T cells were not primed by vaccination. Sera were collected from wild-type (B6) mice immunized with 5BlacZ or control supernatant, pooled, and transferred into naive recipient mice genetically deficient in mature B cells (Igh-6 [µMT]) 3 h before and again 3 days after challenge. As a positive control, B6 mice were immunized once with 5BlacZ 4 weeks prior to challenge. Negligible amounts of HSV-specific and total IgG could be found in vaginal wash samples of µMT mice after transfer, indicating that IgG transudate from serum was below the level of detection. HSV-specific IgG in the genital tracts of B6 mice was also below the level of detection prior to challenge. HSV-specific ELISA and neutralizing antibody titers in the sera of recipient mice 4 h after transfer were comparable to the titers in B6 control mice (Fig. 4), indicating that the concentration of immune serum in the recipients was physiologic and represented the amount present after a single immunization of B6 mice with replication-defective virus. Replication of HSV-2 in the genital mucosae of µMT mice receiving HSV-immune serum was not significantly different from that of µMT mice receiving control serum at any time postchallenge, but unlike in SCID mice, mucosal replication was curtailed within 4 days (Fig. 5A). In contrast, B6 mice immunized once with 5BlacZ efficiently controlled virus replication. A difference between recipients of control and immune serum became evident, however, when changes in body weight and disease were assessed. Control serum recipients rapidly lost weight (Fig. 6A) and developed severe disease (Fig. 7A), resulting in 50% mortality. Mice receiving immune serum initially lost weight but then began to recover. Signs of disease in these mice were less marked, with no evidence of neurologic disease, and all mice survived the challenge infection. These results reinforce the conclusion that antibody alone helps reduce the severity of genital and systemic disease but has no effect on acute replication in the genital mucosa.

View larger version (23K): [in this window] [in a new window]   FIG. 4.   Concentration of HSV-specific antibody in immune serum recipients. Blood was collected from two mice 4 h after transfer of 5BlacZ-immune serum. Concentration of HSV-specific IgG in these sera and sera from mice immunized once with 5BlacZ was determined by ELISA as described in Materials and Methods. Data represent the geometric mean titer ± SEM. Complement-independent HSV-2-neutralizing antibody titers are shown above the bar graphs.

View larger version (21K): [in this window] [in a new window]   FIG. 5.   Replication of challenge virus in the genital tracts of µMT B-cell-deficient mice or B6 control mice. (A) Groups of four µMT mice were injected with control serum or 5BlacZ-immune serum. B6 mice were immunized once with 5BlacZ. (B) Groups of four to six B6 and µMT mice were immunized with control supernatant (naive) or 5BlacZ (immune). (C) Groups of five to six immunized µMT mice were injected with control serum or 5BlacZ-immune serum 4 weeks after immunization. B6 mice were immunized once with 5BlacZ. All mice were challenged i.vag. with HSV-2 G-6 4 weeks after immunization or 3 h after serum transfer. Titers of virus shed from the genital mucosae were determined from vaginal swabs at the indicated times. Each data set is from one of two experiments and represents the geometric mean titer ± SEM for all mice per group.

View larger version (21K): [in this window] [in a new window]   FIG. 6.   Change in body weight after genital challenge. The mice described in the legend to Fig. 5 were monitored daily for change in body weight following challenge. (A) µMT mice injected with control serum or 5BlacZ-immune serum prior to challenge; (B) B6 and µMT mice immunized with control supernatant (naive) or 5BlacZ (immune) prior to challenge; (C) immunized µMT mice injected with control serum or 5BlacZ-immune serum prior to challenge. Mean weight change of all mice per group was determined daily. Weight determination for a group was discontinued after one or more of its members succumbed to infection. Controls were unmanipulated, age- and sex-matched mice.

Susceptibility of HSV-immunized, B-cell-deficient mice. To define the effect of a lack of HSV-specific antibody on vaccine-induced protection against genital infection with HSV-2, B6 and µMT mice were immunized s.c. with the replication-defective virus 5BlacZ. Mice immunized with uninfected cell supernatant served as controls. Four weeks later, all mice were challenged i.vag. with HSV-2. We observed virtually no difference in mucosal replication of challenge virus between B6 and µMT mice immunized with control supernatant (Fig. 5B). µMT mice immunized with 5BlacZ were able to control primary replication, but not until 3 days postchallenge. Acute replication of challenge virus in the genital mucosae of 5BlacZ-immunized B6 mice was considerably less over the first 2 days postchallenge than in immunized µMT mice (Fig. 5B). This result suggested that two phases exist in the immune response to genital infection: an initial phase that is affected by the lack of antibody in µMT mice, and a second phase (>48 h postchallenge) that is independent of HSV-specific antibody. B6 and µMT mice immunized with control supernatant concomitantly lost weight and developed severe genital and systemic disease (Fig. 6B and 7B), whereas B6 mice immunized with 5BlacZ gained weight and showed no signs of disease after challenge. Weight loss and disease were observed in µMT mice immunized with 5BlacZ but were more variable and mild in most mice than in the control group (for disease, P = 0.003 to 0.044 for day 4 and days 6 through 8; P = 0.06 for days 10 through 12). None of the mice immunized with control supernatant survived infection, but signs of neurologic disease and eventual death occurred in only 43% of the immunized µMT mice. These data suggested a contribution of HSV-specific antibody to protection since immune T cells in immunized µMT mice could not control mucosal infection as rapidly and prevent genital and systemic disease as completely as T cells in wild-type mice that also had immune antibody.

View larger version (21K): [in this window] [in a new window]   FIG. 7.   Signs of genital and neurologic disease after genital challenge. The mice described in the legend to Fig. 5 were monitored daily for development of genital and systemic disease following challenge. (A) µMT mice injected with control serum or 5BlacZ-immune serum prior to challenge; (B) B6 and µMT mice immunized with control supernatant (naive) or 5BlacZ (immune) prior to challenge; (C) immunized µMT mice injected with control serum or 5BlacZ-immune serum prior to challenge. Mean disease scores ± SEM for all mice per group are shown.

View larger version (18K): [in this window] [in a new window]   FIG. 8.   IFN- production by memory B6 and µMT T cells. Genital lymph node T cells from groups of B6 and µMT mice immunized s.c. with 5BlacZ and challenged were placed in culture with UV-inactivated 5BlacZ antigen (A) or PMA and calcium ionophore (B) for 24 or 6 h, respectively, prior to harvest of supernatant. Cytokine concentrations in the supernatants were determined by standard ELISA. Data are from one representative experiment of three performed.

Reconstitution of protection against acute replication and disease in B-cell-deficient mice. To determine whether HSV-specific antibody could affect the response of replication-defective virus-immune, B-cell-deficient mice to HSV-2 challenge, we transferred an amount of 5BlacZ-immune serum that achieved a physiologic concentration, i.e., a concentration equivalent to that seen in B6 mice after a single immunization with 5BlacZ (data not shown). Immunized µMT mice receiving control serum had high titers of challenge virus shed from their genital mucosae until 3 days postchallenge. In contrast, immunized µMT mice to which HSV-immune serum was transferred were able to control challenge virus replication at levels comparable to those for immunized B6 mice (Fig. 5C). In addition, no weight loss or genital disease was observed in immunized, immune serum recipients, although HSV-immune µMT mice receiving control serum developed genital lesions (Fig. 7C; P = 0.028 for day 6, P  0.007 for days 7 through 12) and lost weight (Fig. 6C). These results indicate that serum antibody induced by immunization with replication-defective HSV-2 contributes to protection of the genital mucosa only in the context of HSV-immune cellular components of the response. Nonetheless, serum antibody in isolation can help to protect against disease in the genital tract and nervous system.

Protection against latent infection. We observed that serum antibody was apparently able to restore wild-type levels of protection against mucosal replication and disease, but latent infection of the sensory ganglia can be established by HSV in the absence of disease or even in the absence of peripheral replication (10, 22, 52, 55). We therefore investigated whether passive transfer of immune serum antibody to immunized µMT mice also would restore wild-type levels of protection against establishment of latent infection. 5BlacZ-immunized B6 mice and 5BlacZ-immunized µMT mice receiving 5BlacZ-immune or control serum were sacrificed 4 weeks after challenge. Lumbosacral (L2-S4) DRG were dissected and pooled from each animal, and DNA was prepared from each ganglionic pool. Samples were aliquoted and serially diluted, and replicates of 12 were subjected to nested PCR using primer pairs that amplify the HSV-2 TK gene. The number of copies of viral DNA in each sample was estimated based on the frequency of positive second-round PCR results at each dilution. The mean viral genome copy number in the DRG of immunized µMT mice receiving immune serum was 8, whereas a mean of 3,388 genomes was found in DRG of immunized µMT mice receiving control serum (P < 0.002) (Fig. 9). In fact, transfer of immune serum to immunized µMT mice at the time of challenge resulted in a genome copy number that was similar to, and indeed below, that found in the immunized B6 mice. Our current PCR primers do not distinguish between the HSV-2 strains used for immunization and challenge; however, replication-defective viruses establish latency with extremely low frequency (10, 22, 52, 55). We have been able to detect a mean of only six molecules of HSV DNA per ganglion pool in mice immunized with 5BlacZ but not challenged (data not shown), suggesting that most of the viral genomes detected in immunized µMT mice receiving immune serum may have derived from the immunizing virus. Thus, immune serum antibody apparently aids in preventing latent as well as acute infection of the nervous system.

View larger version (14K): [in this window] [in a new window]   FIG. 9.   (A) Numbers of HSV genomes detected in DRG during latent infection. Twenty-eight days after challenge of the mice used for Fig. 5C, lumbosacral DRG were dissected and DNA was prepared for limiting-dilution, nested PCR. The number of copies of genome detected per sample using primers specific for the HSV-2 TK gene is shown. Each symbol represents the pooled DRG of one mouse. Bars represent the mean of each group. (B) Assay sensitivity, determined using a TK-expressing plasmid as described in Materials and Methods. The lower limit of detection is two molecules of HSV-2 DNA per ganglion pool.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We have observed that immune serum antibody, generated by immunization of mice with a replication-defective HSV-2 vaccine strain and present at physiologic levels, (i) is not effective in reducing challenge virus replication in the genital tract epithelium except in the presence of immune T cells; (ii) partially reduces the severity of genital disease; (iii) has the capacity to control virus spread to and replication in the nervous system and delays or prevents disease of the CNS; and (iv) reduces the establishment of latent infection by HSV, as defined by number of viral genomes in the lumbosacral DRG. Virus-specific antibody induced by parenteral vaccination could in theory affect a subsequent HSV infection via the genital tract at any point in its route from the mucosal epithelium to the CNS. The genital tract is unique among mucosal surfaces in that the quantity of antigen-specific IgG transcends that of IgA (18, 25, 35, 44), although antigen-specific IgA may predominate after mucosal immunization (36) (unpublished data). IgA is produced primarily by plasma cells in the mucosa and is actively secreted, whereas IgG appears in genital secretions as a transudate from mucosal production sites and/or serum (3, 31), in which IgG is the predominant Ig component. HSV-specific IgA exhibits far inferior binding and neutralizing activity compared with IgG at physiologic concentrations and also on a molar basis (45). In addition, HSV-immune IgA knockout mice are no more susceptible to HSV-2 challenge than wild-type mice (46). The apparent lack of a protective effect of HSV-specific IgA has focused more attention on the role of virus-specific IgG in protecting the host from infection via the genital tract. It follows, then, that IgG in immune serum should play a role in mucosal as well as systemic protection.

Immune serum antibody has been shown to play a role in clearance of several mucosal pathogens, and passively transferred immune serum alone is often sufficient to control infection. For example, passively transferred immune serum can resolve reovirus (2) or influenza virus (43) infection in mice. In contrast to the intestinal and lung mucosae, serum IgG, except in large quantities, appears to be insufficient for prevention or clearance from the genital tract of pathogenic organisms such as chlamydiae (58). In the case of genital infection with HSV-2, passive transfer of immune serum does not affect mucosal replication (25, 45, 51) except when large amounts of IgG are transferred (45). Nevertheless, immune serum antibody does make a discernible contribution to protection of the genital mucosa from HSV-2 in the context of immune T cells.

Development of genital and neurologic disease is also affected by the absence of immune serum antibody. Immunized µMT mice, while uniformly able to control acute replication of HSV-2 in the genital mucosa from 3 days postchallenge onward, nonetheless develop genital inflammation and in some cases fatal neurologic disease. Immune serum transfer to immunized µMT mice completely restores protection against genital tract disease. Immune serum antibody also influences systemic disease progression after challenge of naive mice either by delaying the onset of severe symptoms (SCID) or by reducing the incidence and severity of neurologic disease (µMT). In light of the effect of immune serum on apparent neurologic disease, we were surprised to find that no significant difference in titer could be detected in the inferior spinal cords of SCID mice receiving immune serum versus control serum until 6 days post challenge. The immune serum antibody did, however, reduce replication in the CNS, which likely prevented or delayed onset of encephalitis and/or dampened inflammation in the spinal cord to prevent paralysis. In contrast to the genital tract, these affects are achieved independently of T cells since they were observed in SCID as well as µMT mice.

Assuming that most of the viral TK gene sequences detected in latently infected DRG represent intact, latent genomes, we have demonstrated for the first time, using a sensitive DNA PCR assay, the capacity of immune serum antibody to quantitatively reduce establishment of latent infection. The number of challenge virus genomes in the DRG 4 weeks after challenge of immunized, µMT mice receiving immune serum was reduced to levels comparable to those in immunized and challenged B6 mice. These results extend previous observations on the establishment of latency in the lumbosacral DRG after challenge of HSV-immune mice by permitting quantitation of genome load rather than by relying on assays of reactivation from explanted ganglia. They also complement the previous findings of Dudley et al. (13) that immune serum reconstitution of µMT mice reduces acute infection of the DRG after vaginal challenge. Notably, reduced establishment of latency following virulent HSV-2 infection of the genital tract was accomplished by immunization with a replication-defective HSV vaccine strain.

Antiviral effector functions of antibody could include viral neutralization, complement fixation, antibody-dependent cellular cytotoxicity (ADCC), and opsonization for elimination by macrophages or neutrophils, and we can only speculate on the effector mechanisms operating in our experiments. We found that HSV-specific IgG in the genital tract prior to challenge was below the level of detection in immunized B6 mice or in µMT mice to which immune serum had been transferred, suggesting that serum transudate across the genital mucosa and thus intralumenal neutralizing activity is normally quite low in parenterally immunized mice. Within 1 week after challenge, HSV-specific IgG reached 1 to 3 ng/ml in vaginal secretions of each (data not shown). Since µMT mice cannot synthesize endogenous IgG, this result suggests that permeability to serum transudate increases with inflammation and/or disruption of the mucosal epithelium induced by virus infection, which in turn could increase the neutralization potential of immune serum in the mucosa. Noteworthy in this regard, the polyclonal immune serum used had complement-independent neutralizing activity in vitro, and serum antibodies from 5BlacZ-immunized mice bind HSV proteins with electrophoretic mobilities of gB and gD, as detected by Western blotting (L. A. Morrison, unpublished observation). Our observation, however, that immune antibody alone was ineffectual in reducing virus replication in the mucosal epithelium of naive mice indeed suggests that antibody does not neutralize a significant amount of virus there. Although the SCID and µMT mice used in our experiments were not of the same genetic background, viral replication in the genital mucosa of both mouse strains was unabated by the presence of immune serum antibody, indicating that neutralizing capacity was not host strain dependent and did not increase over time postchallenge. Thus, it seems unlikely that neutralization was a primary effector mechanism because, despite neutralizing capability and glycoprotein specificity, immune serum antibody required the presence of immune T cells for full expression of its protective capacity.

Nor did immune serum antibody in the genital mucosa appear to act merely by fixing complement or by arming killer cells to perform ADCC. Complement and innate cell types expressing Fc receptor (FcR) for IgG such as NK cells, neutrophils, and macrophages would be equally present in SCID, µMT, and wild-type mice and would be expected to act rapidly to curtail infection. However, naive SCID and µMT mice to which serum antibody is transferred do not control early mucosal infection. Conversely, cellular immune responses generated by immunization of B-cell-deficient mice do not exert control over mucosal replication until 3 days postchallenge. Passive transfer of a physiologic amount of HSV-immune serum restores to immunized, B-cell-deficient mice the capacity to limit early virus replication, suggesting a central role for antigen-specific T cells. The role of T cells, particularly CD4⁺ T cells, in the B-cell response to antigen has been envisioned primarily as providing help in the form of cytokines for B-cell differentiation and antibody production. Our experiments highlight another aspect of T-B-cell collaboration: a dependence on immune T cells for full function of preformed, antigen-specific antibody. In support of our observation that efficient clearance of challenge virus replication in the genital mucosa during the first 2 days of a memory response to HSV-2 requires the presence of immune T cells and immune serum antibody, Dudley et al. (13) and Parr and Parr (48) observed higher levels of challenge virus replication in µMT mice immunized i. vag. with TK HSV-2 than in immunized wild-type mice. Dudley et al. also restored wild-type capacity to reduce replication by passive transfer of immune serum, but the serum transferred was likely in excess of a physiologic amount since the level of replication in the µMT serum recipients was 1 log₁₀ less than that of wild-type controls. Thus, we have definitively shown that full expression of the capacity of antibody to resist HSV-2 infection in the mucosa manifests itself only in the presence of immune T cells. It is interesting that depletion of naive CD4⁺ T cells prior to antibody transfer abrogated the capacity of a monoclonal antibody to exert any control over acute replication (14). Work with influenza virus infection of the lung has also suggested that control of replication by memory CD4⁺ T cells is relatively inefficient in the absence of immune antibody (57).

How do performed antibody molecules and immune T cells work in concert to control replication in the genital mucosa? A likely scenario also involves FcR⁺ cells of the innate arm of the response that bind immune antibody and mediate ADCC. NK cells and neutrophils mediating ADCC would be present in both HSV-immune mice and nonimmune mice provided with serum antibody (40, 54); however, cytokine-induced activation greatly enhances their killing capacity (7, 17, 50, 53). IFN- and tumor necrosis factor alpha are produced rapidly and in high concentration by immune T cells responding to HSV infection but not by naive T cells (20, 47). These cytokines may activate ADCC function of NK cells and/or neutrophils to which immune antibody has bound, thus exerting early control over virus infection in the genital epithelium. In this scenario, antibody taking part in ADCC would be relatively ineffective in the absence of cytokine-producing, HSV-immune T cells. Further experiments will be required to validate this hypothesis. Of interest, depletion of neutrophils or NK cells prior to i.vag. challenge of HSV-immune mice or guinea pigs, respectively, results in greater virus replication in the mucosal epithelium over the first few days of infection (34, 59). A three-way collaboration between innate, humoral, and immune cellular components of the response to HSV infection could explain in part why depletion of NK cells or IFN- results in a decreased capacity to control replication in the mucosa (33, 47). Similarly, opsonization of virus by HSV-specific antibody leading to uptake by FcR⁺ cells of the innate arm of the response and subsequent intracellular destruction via nitric oxide production would be enhanced by T-cell-derived cytokines, principally IFN-.

In contrast to the genital mucosa, immune serum in isolation did impede progress of infection in the CNS. Antibody intercepts virus at axon termini and synapses and can also be taken up by the neuron for intra-axonal neutralization (16). Whereas some form of ADCC would be an effective means of virus control in the genital mucosa, the nervous system may have evolved a preference for neutralization or other noncytolytic mechanisms as a means of self-preservation . Indeed, removal of the Fc portion of HSV-specific antibody prior to transfer to mice has been shown to only partially reduce its capacity to prevent virus infection of the nervous system (28). Neutralizing antibodies have also been observed to inhibit spread of HSV between infected dorsal root ganglion neurons and epidermal cells in vitro (32). The antibodies were not taken up into the neurons, suggesting that neutralization occurs at the axon termini.

The HSV-1 virion itself is armed with an FcR for IgG composed of the gE-gI glycoprotein complex that can protect cells from ADCC in vitro (12) and has been shown to increase the replication and pathogenic potential of the virus in mice (39). Clearly, we determined a role for physiologic concentrations of HSV-specific serum antibody in reducing viral replication in the mucosa and nervous system despite the HSV FcR. Thus, the viral FcR may not be completely effective in assisting the virus to evade immune antibody function in vivo.

In summary, purified HSV-specific serum IgG or monoclonal antibody, if given in sufficient quantities, may be capable of reducing viral infection of the vaginal mucosa independently of immune T cells. Our results, however, suggest that under conditions of more modest, vaccine-induced immunity, HSV-specific serum antibodies play a correspondingly more modest role in protection of the genital tract and a greater role in reducing genital and systemic disease. This does not preclude an important role for vaccine-induced T cells in protection of the genital tract. Indeed, our results suggest that antibody contributes only to limiting virus replication and reducing disease, and other effectors must function in addition. A goal of vaccine-induced immunity against HSV therefore should include the induction of serum IgG responses. Eliciting HSV-specific serum IgG through vaccination in addition to T-cell responses would allow both effectors to work in concert to reduce infection in the mucosa and thereby limit the amount of virus capable of escaping the genital tract. Serum antibody would also then be available as a second line of defense against systemic spread and establishment of latency.