Prophylactic Herpes Simplex Virus 2 (HSV-2) Vaccines Adjuvanted with Stable Emulsion and Toll-Like Receptor 9 Agonist Induce a Robust HSV-2-Specific Cell-Mediated Immune Response, Protect against Symptomatic Disease, and Reduce the Latent Viral Reservoir

ABSTRACT

Several prophylactic vaccines targeting herpes simplex virus 2 (HSV-2) have failed in the clinic to demonstrate sustained depression of viral shedding or protection from recurrences. Although these vaccines have generated high titers of neutralizing antibodies (NAbs), their induction of robust CD8 T cells has largely been unreported, even though evidence for the importance of HSV-2 antigen-specific CD8 T cells is mounting in animal models and in translational studies involving subjects with active HSV-2-specific immune responses. We developed a subunit vaccine composed of the NAb targets gD and gB and the novel T cell antigen and tegument protein UL40, and we compared this vaccine to a whole-inactivated-virus vaccine (formaldehyde-inactivated HSV-2 [FI-HSV-2]). We evaluated different formulations in combination with several Th1-inducing Toll-like receptor (TLR) agonists in vivo. In mice, the TLR9 agonist cytosine-phosphate-guanine (CpG) oligodeoxynucleotide formulated in a squalene-based oil-in-water emulsion promoted most robust, functional HSV-2 antigen-specific CD8 T cell responses and high titers of neutralizing antibodies, demonstrating its superiority to vaccines adjuvanted by monophosphoryl lipid A (MPL)-alum. We further established that FI-HSV-2 alone or in combination with adjuvants as well as adjuvanted subunit vaccines were successful in the induction of NAbs and T cell responses in guinea pigs. These immunological responses were coincident with a suppression of vaginal HSV-2 shedding, low lesion scores, and a reduction in latent HSV-2 DNA in dorsal root ganglia to undetectable levels. These data support the further preclinical and clinical development of prophylactic HSV-2 vaccines that contain appropriate antigen and adjuvant components responsible for programming elevated CD8 T cell responses.

IMPORTANCE Millions of people worldwide are infected with herpes simplex virus 2 (HSV-2), and to date, an efficacious prophylactic vaccine has not met the rigors of clinical trials. Attempts to develop a vaccine have focused primarily on glycoproteins necessary for HSV-2 entry as target antigens and to which the dominant neutralizing antibody response is directed during natural infection. Individuals with asymptomatic infection have exhibited T cell responses against specific HSV-2 antigens not observed in symptomatic individuals. We describe for the first time the immunogenicity profile in animal models of UL40, a novel HSV-2 T cell antigen that has been correlated with asymptomatic HSV-2 disease. Additionally, vaccine candidates adjuvanted by a robust formulation of the CpG oligonucleotide delivered in emulsion were superior to unadjuvanted or MPL-alum-adjuvanted formulations at eliciting a robust cell-mediated immune response and blocking the establishment of a latent viral reservoir in the guinea pig challenge model of HSV-2 infection.

INTRODUCTION

Herpes simplex virus 2 (HSV-2) infections are a major global public health burden. An estimated 417 million people worldwide aged 15 to 49 years live with HSV-2 infection, and roughly 20 million incident infections occurred in 2012 (1). HSV-2 infection also increases the risk of HIV acquisition and transmission. Infection with HSV-2 is lifelong and is characterized by periods of symptomatic disease or asymptomatic viral shedding (2), and roughly 80% of HSV-2-seropositive people are unaware of their infection (3). While antiviral therapies are available, daily use cannot eliminate the viral reservoir or completely prevent transmission in serodiscordant couples (4). With the lack of additional therapeutic options, an efficacious vaccine is necessary to control the continued spread of HSV-2 infections around the world.

For an HSV-2-targeted vaccine administered either prophylactically or therapeutically, robust antibody (Ab) and T cell responses will likely be required. Much effort has been directed at identifying the most appropriate neutralizing antibody (NAb) targets in both animal models and clinical trials. Most studies have identified glycoprotein D (gD2) and glycoprotein B (gB2) as promiscuous targets of both binding and neutralizing Abs in naturally infected humans (5–7). Several prophylactic vaccines based on gD2, alone or combined with gB2, and adjuvanted with strong B cell-directed adjuvants such as aluminum salts and 3-de-O-acylated monophosphoryl lipid A (MPL) have generated robust neutralizing Ab responses in vaccine recipients, but primary clinical endpoints have not been achieved (8–10). While antibody responses directed against HSV-2 glycoproteins are thought to play an important role in the control of HSV-2 infection after vaccination, accumulating evidence suggests that robust HSV-2-specific T cell responses are critically important. Among HIV–HSV-2-coinfected patients, a low HSV-specific CD8 T cell frequency is correlated with the formation of severe genital herpes lesions (11). Infiltrating CD8 T cells can be detected for many weeks after lesion healing near sensory nerve endings, where it is believed that they can quickly mount an effector immune response during times of viral reactivation (12). Thus, an effective HSV vaccine should promote long-lasting CD8 T cell responses, although the identity of ideal HSV-2 T cell targets is unclear.

In this report, we evaluated multiple HSV-2 antigens (Ags) combined with several Toll-like receptor (TLR) agonists and formulations using both mouse and guinea pig models. Biological effects of adjuvants can be categorized into two major modes of action: improved delivery of antigen and acceleration of antiviral humoral and cellular immune responses. Adjuvant comparison studies in mice led us to identify the cytosine-phosphate-guanine (CpG) oligodeoxynucleotide formulated with an oil-in-water squalene-based emulsion as being optimal for the delivery of the subunit antigens HSV-2 gDt and HSV-2 UL40 so as to retain high neutralizing Ab induction while maximizing the induction of robust CD8 T cell responses. While mice allow more detailed immunological analyses, guinea pigs are a more biologically relevant model and represent a higher bar to demonstrate efficacy. Subunit combination and inactivated whole-virus vaccines adjuvanted with CpG-emulsion were tested in the HSV-2 challenge guinea pig model and were found to promote robust virus-specific cell-mediated immune (CMI) responses while preventing HSV-2 infection and disease. These results support the further development of CpG-emulsion-adjuvanted HSV-2 vaccines.

RESULTS

Synthesis of HSV-2 subunit antigens gD and UL40.

HSV-2 gD has undergone extensive investigation as a vaccine antigen, and gD2-based vaccines have demonstrated the ability to induce high titers of NAbs and protection from disease in mice and guinea pigs (13, 14). However, substantial gD2-specific CD4/CD8 T cell responses have not been reported, and we sought an additional antigen to drive a robust CD8 T cell response. Among other candidates, the UL40 tegument protein was selected because of its ease of expression and purification and because UL40 has been identified as the T cell target of HSV-2-specific CD8 T cell responses detected in symptomatic patients (15, 16).

HSV-2 gD is a 393-residue glycoprotein containing a single transmembrane segment (17). The majority of the protein mass is displayed on the viral surface, with a short segment residing in the virion interior (17, 18). To reduce the technical challenges of working with a membrane protein, the HSV-2 gD expression construct used in our studies contained only the extracellular portion of the gD sequence (residues 25 to 281). This expression construct produced high levels of secreted glycoprotein from Chinese hamster ovary (CHO) cells, and glycosylation was confirmed by incubation with the deglycosylating enzyme peptide-N-glycosidase F (PNGase F) (Fig. 1A). A Western blot analysis was performed with polyclonal anti-gD antibody to confirm the identity of HSV-2 gD. We also determined the mass of the glycans on the molecule using size exclusion chromatography with multiangle light scattering (SEC-MALS) (Fig. 1B). The measured mass of the protein was 34.2 kDa, in good agreement with the predicted mass of the expressed construct (32.5 kDa), and we observed 7.1 kDa of glycan mass on the molecule.

FIG 1.

Purification of HSV-2 gDt and UL40 antigens. (A, left and middle) SDS-PAGE showing purified HSV-2 gD material from CHO cells and treatment of this material with the deglycosylating enzyme PNGase F. (Right) Western blotting of the purified gD protein using an anti-gD antibody. (B) SEC-MALS data for the purified gD protein, with a UV chromatogram and a light scattering (LS) chromatogram shown (the refractive index trace was omitted for clarity). The data points extending across the elution peak are the total particle mass, protein mass, and glycan (conjugate) mass. The measured mass of the gD protein (34.2 kDa) is in good agreement with the theoretical mass of 33.4 kDa. (C, left) SDS-PAGE showing purified HSV-2 UL40 protein (left). (Middle and right) The identity of the protein was confirmed by Western blot analysis using both anti-His (middle) and anti-Strep (right) antibodies. (D) HSV-2 UL40 SEC-MALS data showing the light scattering chromatogram and the UV chromatogram (the refractive index data have been omitted for clarity). The data points extending across the peaks indicate the measured mass of the protein species in that peak. The theoretical mass of UL40 is 40.0 kDa, which is in good agreement with the smallest-eluting peak of 41.5 kDa. Based upon the masses of the additional peaks, UL40 can oligomerize and is predominantly dimeric in solution.

HSV-2 UL40 is expressed during the replicative phase of the HSV-2 life cycle. UL40 is a member of the small ribonucleotide reductase family of proteins (19), which homodimerizes and then in turn dimerizes again to form an active tetrameric enzyme (20, 21). UL40 was expressed at high levels in Escherichia coli, purified from cell lysates, and identified by Western blot detection of the N-terminal His₆ tag and the C-terminal Strep tag, confirming full-length expression (Fig. 1C). Western blot analysis revealed two UL40 protein species consistent with both monomers and dimers. SEC-MALS indicated that the protein was approximately 40 kDa and that it readily formed dimers with small proportions of monomers, trimers, and tetramers (Fig. 1D).

Optimal HSV-2 antigen-specific CMI induced by vaccine formulations containing TLR agonists formulated in an oil-in-water emulsion.

To generate strong CD4 and CD8 responses specific for HSV-2 gD and HSV-2 UL40, we combined several Toll-like receptor ligands (TLRLs) with AddaVax, a commercially available squalene-based oil-in-water emulsion that shares many physicochemical characteristics with MF59, which has been approved for human use (22). Well-studied representatives of several TLRL families were utilized, including the TLR3L poly(I·C), the TLR4L MPL, the TLR7L gardiquimod (GDQ), and the TLR9L CpG-2395 (2395). C57BL/6 mice were immunized intramuscularly (i.m.) twice at biweekly intervals with both HSV-2 antigens formulated with TLRL-emulsion, and spleens were harvested 7 days after the second immunization and analyzed for gamma interferon (IFN-γ) production (Fig. 2). Immunization with antigens delivered with AddaVax alone resulted in a substantial UL40-specific IFN-γ response but a very poor response to gD. When the adjuvant formulation included 2395, poly(I·C), or MPL, a significant enhancement of the UL40 CMI response was observed. In addition, CMI responses to gD were then detectable with the addition of these TLR agonists to AddaVax. The addition of gardiquimod to AddaVax did not improve CMI induction above that with AddaVax alone. Furthermore, the MPL-alum adjuvant formulation was extremely poor at the induction of CMI responses, falling to or below the levels achieved with the emulsion alone. Thus, the combination of 2395, poly(I·C), or MPL and the oil-in-water emulsion AddaVax resulted in the highest levels of HSV-2 antigen-specific CMI responses.

FIG 2.

Vaccination with HSV-2 antigens gD and UL40 with a TLR agonists plus an emulsion results in robust T cell responses. C57BL/6 mice were immunized on days 0 and 14 with 10 μg each of the gD and UL40 antigens, which were adjuvanted as described in Materials and Methods. Spleens were harvested on day 21 and processed, and the resultant splenocytes were then stimulated with antigen-specific overlapping peptide pools for 48 h according to ELISPOT protocols. Data represent results for individual animals with group log₁₀ geometric means ± 95% confidence intervals. Statistical comparisons are between the antigen-alone group and each adjuvanted group unless otherwise indicated. ns, not significant (P > 0.05); *, P < 0.05; **, P < 0.01; ***, P < 0.001.

The potency of these adjuvant formulations was further supported by the secreted-cytokine profile of splenocytes from immunized mice after 48 h of stimulation with HSV-2 antigens. The formulation of UL40 with AddaVax and 2395, poly(I·C), or MPL resulted in a robust cytokine profile characterized by elevated IFN-γ, interferon gamma-induced protein 10 (IP-10), monokine induced by gamma interferon (MIG), and interleukin-6 (IL-6) levels, accompanied by moderate levels of production of IL-2 and tumor necrosis factor alpha (TNF-α) and suppression of IL-5, which was induced at high levels by the emulsion alone (data not shown). MPL-alum poorly induced the production of nearly all cytokines measured. The same adjuvant formulations were also responsible for HSV-2 gD-specific cytokine profiles similar to those observed with UL40 but with generally lower levels of cytokines being expressed.

Closer analysis of UL40-specific T cells via flow cytometry identified the presence of multifunctional CD4 T cell subsets, with substantial populations of double- and triple-cytokine expressers (of IFN-γ, TNF-α, or IL-2) being promoted by vaccination with the antigen plus 2395, poly(I·C), or MPL formulated with AddaVax (Fig. 3). Such subsets have been demonstrated to be preferential for protection from viral disease (23, 24). The 2395-AddaVax formulation also promoted the differentiation of a multifunctional CD8 T cell subset and was the only adjuvant to do so. MPL-alum and AddaVax alone largely induced poorly differentiated single-cytokine-expressing CD4/CD8 T cells. CD4/CD8 T cell subsets specific for HSV-2 gD were largely single-cytokine expressers, regardless of the choice of adjuvant, underlining the low propensity for gD to induce T cell responses (data not shown).

FIG 3.

Highly differentiated multifunctional CD4 and CD8 T cell responses achieved after vaccination with TLR9L plus an emulsion. C57BL/6 mice were immunized on days 0 and 14 with 10 μg the gD and UL40 antigens, which were adjuvanted as described in Materials and Methods. Spleens were harvested on day 21 and processed, and the resultant splenocytes were then stimulated with antigen-specific overlapping UL40 peptide pools for 5 h according to ICS protocols. Data are presented as pie charts, with sections representing subsets of UL40-specific cytokine production expressed as averages of data from 8 individual mice. The total is X, where X is the percentage of the UL40-specific cytokine-expressing population within the total CD4 or CD8 T cell population.

Substantial HSV-2 UL40-specific CTL responses induced by 2395-AddaVax and poly(I·C)-AddaVax.

A potent functional cytotoxic T lymphocyte (CTL) response is generally thought to be a requisite for a successful HSV-2 vaccine, so we further characterized the ability of these adjuvants to induce T cells capable of exerting cytotoxicity in vivo. Carboxyfluorescein succinimidyl ester (CFSE)-labeled naive splenocytes were loaded with a UL40 peptide pool and mixed 1:1 with a control splenocyte population loaded with a peptide pool for an irrelevant antigen. This target population mix was injected intravenously into vaccinated mice, and 24 h later, splenocytes from these mice were analyzed by flow cytometry for the presence of target cells, distinguishable by CFSE staining. Results were normalized for background killing of gp350 peptide-loaded cells. 2395-AddaVax induced an extremely robust UL40-specific cytotoxic response, with a nearly 95% reduction in the frequency of UL40 target cells (Fig. 4A), which was significantly higher than those for all other adjuvants tested. The poly(I·C)-AddaVax-adjuvanted group also demonstrated a robust in vivo killing response (68%), while other adjuvanted groups, including the MPL-alum-adjuvanted group, induced the killing of only ≤26% of UL40 targets.

FIG 4.

Promotion of in vivo CTL and neutralizing antibody development after vaccination with a TLR agonist plus an emulsion adjuvant. (A) C57BL/6 mice (3/group) were immunized on days 0 and 14 with 10 μg each of the gD and UL40 antigens, which were adjuvanted as described in Materials and Methods. On day 21, target cells were incubated with 1 μM CFSE and an irrelevant peptide pool or 10 μM CFSE and an HSV-2 UL40 overlapping peptide pool. Loaded target cells were reinjected into vaccinated mice by the intravenous route via the tail vein. Splenocytes were harvested at 18 to 24 h postinjection, and the CFSE signal was detected by flow cytometry. UL40-specific target cell killing was normalized to control target cell killing by using the following formula: % specific killing = (1 − % CFSE^hi cells/% CFSE^lo cells) × 100. Data are presented as individual points per mouse with means and standard errors of the means. Statistical comparisons are between the antigen-alone group and each adjuvanted group. *, P < 0.05; **, P < 0.01, ***, P < 0.001. (B) C57BL/6 mice were immunized on days 0 and 14 with 10 μg the gD and UL40 antigens, which were adjuvanted as described above. Serum from peripheral blood drawn at day 21 was analyzed for the presence of HSV-2-neutralizing Ab titers. Presented are individual results, with bars representing the group means with standard error of the means. Statistical comparisons are between the antigen-alone group and each adjuvanted group (n = 8). **, P < 0.01; ***, P < 0.001.

Optimal HSV NAb induction by emulsion formulations.

To confirm whether 2395-AddaVax and poly(I·C)-AddaVax adjuvant formulations were also able to promote robust functional Ab responses in addition to a strong CMI response, we tested serum samples from immunized mice for HSV NAb titers. HSV-2 strain MS was incubated with serum samples and then used to infect Vero cells at a multiplicity of infection (MOI) of 0.01 PFU, and CPE (cytopathic effects) were scored by microscopy. Serum from mice adjuvanted with poly(I·C)-AddaVax and 2395-AddaVax induced higher titers of HSV-2 NAbs than did serum from mice adjuvanted with AddaVax alone or MPL-alum (Fig. 4B).

Vaccination with FI-HSV-2 or subunit antigens elicits robust antibody responses in guinea pigs when combined with 2395-AddaVax or MPL-alum.

Since HSV-2 infection in guinea pigs results in disease symptoms that mimic those that are observed with natural infection in humans (13, 25), we continued the evaluation of vaccines adjuvanted with 2395-AddaVax or MPL-alum in HSV-2-challenged guinea pigs and compared them against a calibrator vaccine representing Simplirix, gD plus MPL-alum (26). The HSV-2 glycoprotein subunit gB was added to the subunit vaccine of gD and UL40 to optimize NAb target selection. Additionally, we investigated the format of formaldehyde-inactivated virus (formaldehyde-inactivated HSV [FI-HSV]), a process of vaccine generation which allows a broadened T cell response by using whole, inactivated virus and which was previously examined in guinea pigs (14, 27). Accordingly, we next compared the abilities of the adjuvant formulations to boost either the subunit vaccine gD/gB/UL40 or FI-HSV-2 in immunogenicity and challenge studies using the guinea pig model of genital HSV-2 infection. Guinea pigs were vaccinated twice by i.m. injection at 3-week intervals, and sera were drawn 2 weeks after the second immunization to analyze the humoral response. Vaccination with FI-HSV-2 induced IgG titers to both gD and gB without further amplification by either 2395-AddaVax or MPL-alum, the adjuvant formulation found in Simplirix. In contrast, anti-gD/gB IgG titers induced with the subunit vaccine gD/gB/UL40 were substantially enhanced when adjuvanted with 2395-AddaVax and to a significantly greater extent than that observed with MPL-alum (Fig. 5A). The antigen alone, whether FI-HSV-2 or gD/gB/UL40, was insufficient to induce high NAb titers (Fig. 5B) and required 2395-AddaVax in the case of FI-HSV-2, or either adjuvant in the case of gD/gB/UL40, to promote NAb titers comparable to those induced by the calibrator vaccine comprising gD plus MPL-alum. Due to limitations in the number of animals for in vivo studies, vaccines adjuvanted with 2395 alone versus 2395-AddaVax were not investigated. These results suggest that 2395-AddaVax is a more suitable adjuvant than MPL-alum for vaccines containing several antigen components.

FIG 5.

Active vaccination with FI-HSV-2 or subunit antigens elicits robust antibody responses in guinea pigs. (A) Anti-HSV-2 gD and gB binding antibody titers in peripheral blood serum as measured by an ELISA. Bar graphs represent the geometric mean titers (GMT) ± 95% confidence intervals (CI). ** indicates statistical significance for gD plus MPL-alum (P < 0.01), § indicates statistical significance for both gD2 and gB2 binding antibody titers between the indicated groups (P < 0.01), and # indicates statistical significance for gB2 binding antibody titers between the indicated groups (P < 0.01). (B) Neutralizing antibody levels in serum collected prior to HSV-2 challenge. Vaccination groups are listed on the x axis. Results represent the means ± 95% confidence intervals. ** and *** indicate statistical significance between groups (P < 0.01 and P < 0.001, respectively). ns, not significant. Dashed lines indicate the limit of detection. Serum levels in the PBS group and the 2395-AddaVax group are below the limit of detection for all assays (data not shown).

2395-AddaVax induces robust IFN-γ CMI responses in vaccinated guinea pigs.

To determine the effects of adjuvanting FI-HSV-2 or gD/gB/UL40 on T cell responses in vivo, we developed a modified version of the murine IFN-γ enzyme-linked immunosorbent spot (ELISPOT) assay using monoclonal antibodies (MAbs) specific for guinea pig IFN-γ (see Materials and Methods). Peripheral blood mononuclear cells (PBMCs) were isolated from immunized guinea pigs 1 week after the second vaccination and stimulated in vitro for 48 h with overlapping peptide pools spanning the sequences of HSV-2 gD2, gB2, and UL40. Robust IFN-γ responses specific for gD, gB, and UL40 were observed when FI-HSV-2 was adjuvanted with 2395-AddaVax (Fig. 6) and were significantly higher than those with adjuvanting with MPL-alum in the case of gD and gB. The combination of 2395-AddaVax and the multiple-subunit vaccine gD/gB/UL40 also resulted in strong IFN-γ responses to all 3 antigens comparable to those with FI-HSV-2 plus 2395-AddVax. In contrast, the comparator vaccine comprising gD2 plus MPL-alum promoted poor CMI responses against gD2, with undetectable responses in 4/7 animals. Individually, the subunit antigens did not elicit robust CMI responses when adjuvanted with MPL-alum in guinea pigs (data not shown). Therefore, we did not test the combination of antigens and MPL-alum for these studies, nor did we test PBMCs from unadjuvanted vaccine groups, as we did not expect that T cell responses from these groups would be of a significant magnitude. These data underscore that 2395-AddaVax is an effective adjuvant for the elicitation of HSV-2 Ag-specific CMI responses in guinea pigs and has an efficacy superior to that of MPL-alum.

FIG 6.

Animals vaccinated with formulations containing 2395-AddaVax elicit more robust T cell responses than those elicited in groups vaccinated with MPL-alum. PBMCs from vaccinated guinea pigs were collected 7 days after boost vaccination and prior to HSV-2 challenge. IFN-γ levels were measured by an ELISPOT assay after stimulation with overlapping peptide pools for HSV-2 gD, gB, and UL40 sequences. Results represent the log₁₀ geometric means ± 95% confidence intervals. *, **, and *** indicate statistical significance (P < 0.05, P < 0.01, and P < 0.001, respectively). ns, not significant. Samples for which there were no detectable IFN-γ spots were assigned a value of 0.5 for graphing purposes and statistical analysis.

Adjuvanted subunit vaccines and FI-HSV-2 with or without an adjuvant reduce acute viral shedding and protect guinea pigs from symptomatic disease after HSV-2 challenge.

To measure the ability of the vaccine to control HSV-2 infection, we intravaginally challenged guinea pigs with 5 × 10⁵ PFU of strain MS 3 weeks after the second vaccination. Vaginal swabs were collected on days 1, 2, 5, and 9 postchallenge, and titers of infectious virus were determined by a plaque assay. Day 1 viral titers in all vaccinated groups were significantly lower than those in the phosphate-buffered saline (PBS)-vaccinated group, except for gD plus MPL-alum and gD/gB/UL40 alone (Fig. 7A). In particular, FI-HSV-2 alone or in combination with either adjuvant had lower log₁₀ geometric mean titers than did the comparator vaccine comprising gD plus MPL-alum. By day 2 after challenge, viral titers began to decrease across all groups, with FI-HSV-2 (with or without an adjuvant)-vaccinated groups approaching the lower limit of detection (LLOD) for the assay. On day 5, viral titers approached the LLOD for all groups except the groups administered PBS, gD/gB/UL40, and gD/gB/UL40 plus MPL-alum, and by day 9, values for all groups had fallen below the LLOD, with the exception of the groups administered PBS and gD/gB/UL40 alone (data not shown). Daily clinical observations were also recorded for 2 weeks postchallenge, and we observed that most vaccine treatments were able to control clinical symptoms of disease. In the PBS placebo group, all but one guinea pig exhibited at least one disease symptom at some point during 2 weeks postchallenge. Vaccination with FI-HSV-2 with or without an adjuvant greatly diminished clinical disease scores in challenged guinea pigs (Fig. 7B). Adjuvanting of the gD/gB/UL40 subunit with MPL-alum was less effective in the prevention of clinical disease (5/10 [50%] animals with clinical disease signs) than was 2395-AddaVax (2/18; 11%). HSV-2 infection in guinea pigs encompasses sequelae of infection other than lesions, including hind-limb paralysis, hematuria, urine retention, and mortality. Only two treatment groups, the PBS-alone and subunit gD/gB/UL40 groups, demonstrated signs of these severe sequelae after challenge (Table 1). These data highlight that unadjuvanted subunit vaccines are insufficient to confer protection in the guinea pig model and must be paired with a robust adjuvant such as 2395-AddaVax.

FIG 7.

HSV-2 vaccines adjuvanted with MPL-alum and 2395-AddaVax partially control acute viral replication and prevent clinical disease in HSV-2 MS-infected guinea pigs. (A) Guinea pigs were swabbed intravaginally on the indicated days postchallenge. Swabs were placed into RPMI supplemented with 10% FBS plus penicillin-streptomycin and were frozen at −80°C until tested by a standard Vero cell plaque assay. Levels of virus are shown as log₁₀ PFU per milliliter. The dashed line indicates the limit of detection. Bars indicate the geometric means ± 95% confidence intervals. § indicates statistical significance for gD plus MPL-alum (P < 0.05). *, **, and *** indicate statistical significance for PBS (P < 0.05, P < 0.01, and P < 0.001, respectively). (B) Mean cumulative disease scores of clinical HSV-2 disease for each group. Scores were assigned on the following scale for individual guinea pigs: 0 for no disease, 1 for erythema only, 2 for a single vesicle or a few vesicles, 3 for large or fused vesicles, and 4 for ulcerated lesions. A score of 0.5 was given for mild erythema or edema. Guinea pigs that were euthanized or died from HSV-2 infection did not contribute to cumulative disease scoring measures on measurement days after death.

TABLE 1.

Clinical disease manifestations and reduction in latent HSV-2 infection in DRG in vaccinated guinea pigs^d

Vaccine antigen(s)	Adjuvant	No. of guinea pigs	% protection (no. of protected animals/total no. of animals)a	% of guinea pigs positive for sequelab
				Urine retention	Hematuria	Hind-limb paralysis	Death/euthanasia
NA (PBS)	NA	23	10.5 (2/19)c	73.9	37.8	13.0	30.4
FI-HSV-2	NA	11	72.7 (8/11)	0	0	0	0
FI-HSV-2	MPL-alum	18	83.3 (15/18)	0	0	0	0
FI-HSV-2	2395-AddaVax	16	100 (16/16)	0	0	0	0
gD	MPL-alum	7	28.6 (2/7)	0	0	0	0
gD/gB/UL40	NA	11	18.2 (2/11)	36.4	27.3	0	18.2
gD/gB/UL40	MPL-alum	10	10 (1/10)	0	0	0	0
gD/gB/UL40	2395-AddaVax	18	100 (18/18)	0	0	0	0

^aPercent protection is defined as the percentage of guinea pigs where latent HSV-2 genome copy numbers, as measured by qPCR, fell below the lower limit of detection for the assay.

^bPercentage of guinea pigs positive for each HSV-2 sequela if detected any time over the course of observation after challenge (2 weeks).

^cDorsal root ganglia (DRG) were harvested from only 19 of 23 PBS-vaccinated animals. DRG were not collected if the guinea pig died between observations.

^dNA, not applicable.

Adjuvanting with 2395-AddaVax blocks establishment of latency in guinea pig DRG.

To determine if both the reduction in viral shedding and the lack of clinical symptomatic disease additionally resulted in a reduced burden of latent HSV-2 in the lumbrosacral ganglia, we assayed for HSV-2 genome copy numbers by quantitative real-time PCR. Vaccination with the gD/gB/UL40 subunit vaccine alone resulted in no decrease in the levels of HSV-2 DNA, but adjuvanting with 2395-AddaVax, and not MPL-alum, resulted in undetectable levels of DRG (dorsal root ganglia) HSV-2 DNA for all 18 guinea pigs in the group (Fig. 8 and Table 1). FI-HSV-2 vaccination also dramatically decreased latent HSV-2 copy numbers compared to those in the PBS control group, but complete protection in all guinea pigs was achieved only when FI-HSV-2 was adjuvanted with 2395-AddaVax (Fig. 8 and Table 1). Both 2395-AddaVax-adjuvanted vaccines outperformed the comparator vaccine composed of gD plus MPL-alum in achieving the highest percentage of guinea pigs with undetectable levels of HSV-2 DNA, indicating that the choice of adjuvant is critically important in maximally driving down the nervous system reservoir of latent HSV-2.

FIG 8.

Vaccination with FI-HSV-2 plus 2395-AddaVax or gD/gB/UL40 plus 2395-AddaVax prevents establishment of HSV-2 latency within guinea pig dorsal root ganglia. Latent HSV-2 genome copy numbers in DNA isolated from dorsal root ganglia were measured by quantitative real-time PCR. Geometric mean titers ± 95% confidence intervals for individual guinea pigs are shown. ** and *** indicate a significant difference (P < 0.01 and P < 0.001, respectively). # indicates a significant difference for gD plus MPL-alum (P < 0.05). ns, not significant. The dashed line indicates the lower limit of detection for the assay (9 copies). Samples that fell below the limit of detection were assigned a value of 4.5 for graphing and statistical purposes.

DISCUSSION

Many investigators have defined a link between robust cytolytic T cell responses and HSV-2 clearance in animal models (28, 29). We designed vaccines that included antigen and adjuvant choices suitable for the induction of vigorous antigen-specific T cell responses that would demonstrate efficacy in an in vivo model of HSV-2 infection. We determined that the internal tegument antigen UL40 provided an efficient target for high-quality CD8 T cell responses in mice. Additionally, we determined in the mouse model that the optimal adjuvant formulation to achieve this response was a combination of CpG and oil-in-water squalene-based emulsion. Antigen-adjuvant vaccines, including UL40 and CpG-emulsion, achieved highly active UL40-specific CD4/CD8 T cell responses characterized by high-level IFN-γ production to peptide, multifunctional CD4 and CD8 subsets, and in vivo cytolytic activity in mice. Furthermore, when either antigen subunit or FI-HSV-2 was combined with CpG-emulsion in guinea pigs, we observed elevated HSV-2 antigen-specific T cell responses, high titers of antigen-specific Abs and HSV-2-neutralizing Abs, decreased vaginal viral shedding, suppression of disease symptoms, and elimination of detectable HSV-2 in the DRG.

Although HSV-2 antigenic targets for NAb responses, such as gD and gB, have long been identified (30, 31), clinical trials of vaccines based on these antigens have not resulted in broad protection across the cohort population. Indeed, T cell responses to gD and gB are relatively poor in comparison to those to other HSV-2 antigens (16, 32, 33), possibly because these glycoproteins have evolved to resist degradation by proteolytic enzymes (34, 35). Thus, broadening the T cell response by including at least one T cell-targeting antigen should improve vaccine potency. The choice of antigen targets to which CD4/CD8 T cell responses should be directed during HSV-2 infection has been less clear, in part due to a panoply of viral proteins that have been identified in preclinical and clinical studies as having potential for T cell induction (36–38). We identified the HSV-2 ribonucleotide reductase small subunit (UL40) (15, 16) as an effective target antigen for robust CD4 and CD8 T cell responses in immunized mice. This subunit, also known as RR2, forms a holoenzyme with the subunit RR1 that is responsible for providing precursors necessary for viral DNA synthesis (39). We observed that UL40-specific responses were induced in both mice and guinea pigs by adjuvant-optimized vaccines.

A growing body of evidence has been amassed to indicate that the choice of adjuvant to combine with subunit vaccines can greatly influence the magnitude and quality of subsequent T cell responses. Most notably, agonists of the endosomally expressed TLRs (TLR3, TLR7, TLR8, and TLR9) have often demonstrated a facility for promoting robust T cell responses with the activation of both CD4 and CD8 compartments. This is generally thought to be due to the mimicking of innate immune responses programmed for viral pathogens, which often display agonists to these TLRs. Indeed, the adoptive transfer of ovalbumin (OVA)-specific TCR-transgenic OT-I cells into C57BL/6 mice followed by a boost vaccination with OVA peptides adjuvanted with either the TLR3L poly(I·C) or the TLR9L CpG resulted in expansion of functional effector memory CTLs at a higher frequency than what could be achieved with the TLR4L lipopolysaccharide (LPS) (40). Such observations have been extended to HSV-2 gD- and gB-specific responses as well. IC31 is a TLR9-targeted bicomponent complex between a cationic peptide and an oligodeoxynucleotide, and this adjuvant, when combined with gD and administered subcutaneously to mice, was able to provide complete protection against lethal HSV-2 challenge and a robust high IFN-γ/IL-5 ratio of splenocyte cytokine production (41). Adjuvantation of a gB peptide with CpG led to an increase in the frequency of peptide-specific CD8 T cells in the draining lymph node and spleen (42), and CpG-alum also proved to be a far-superior adjuvant when combined with gD and gB in BALB/c mice to elicit antigen-specific IFN-γ responses compared to MPL-alum (43). Furthermore, HSV-2 replication in cultures of human epithelial cells experienced significant inhibition when cells were stimulated with the TLR3L poly(I·C) or the TLR9L CpG but not the TLR4L LPS (44), indicating the possibility of additional direct antiviral effects with this class of TLR-based adjuvants.

Our data confirm and extend the utility of a CpG-based adjuvant system for HSV-2 subunit vaccines. We observed that the superior adjuvant qualities of CpG-emulsion extended to multiple vaccine platforms, including both recombinant antigen subunit vaccines and formaldehyde-inactivated whole-virus preparations. FI-HSV-2 has the advantage over recombinant subunit platforms due to the inherent adjuvanting properties of the particle size and the presence of viral DNA. Additionally, FI-HSV-2 provides an advantage of a wider breadth of antigen diversity, and adjuvantation of FI-HSV with MPL-alum indeed achieves superiority over gD2 plus MPL-alum in the reduction of HSV-2 vaginal shedding and protection from lethal challenge after two intramuscular immunizations in mice (14). Although this vaccine resulted in elevated HSV-2 NAb titers, the level of induction of HSV-2-specific CD4 T cells in response to whole virus was very low, and antigen-specific CD8 T cells were undetectable. Furthermore, FI-HSV-2 plus MPL-alum resulted in a higher IgG1/IgG2a ratio of IgG titers, possibly due to the influence of alum, which is well known for its ability to induce Th2 responses, characterized by high IgG1 levels (45). However, higher levels of protection against HSV-2 disease in mice also correlated with high-level IgG2a induction, which is a by-product of robust Th1 responses, and IgG2a has also been identified as exerting more HSV-2-neutralizing activity than IgG1 (46, 47). We observed that CpG-emulsion was far superior to MPL-alum in the induction of UL40-specific CD8 T cell responses in mice (Fig. 2 and 3) and in preventing the establishment of latent HSV-2 genomes in guinea pig dorsal root ganglia (Fig. 8 and Table 1). CpG has also frequently been demonstrated to facilitate the induction of an IgG2a-dominant Ab response (48, 49), underlining several weaknesses in the capacity of MPL-alum to elicit an appropriate protective response in mice.

There have been several HSV-2 subunit and whole-virus vaccines evaluated in murine challenge models that have demonstrated various levels of disease protection. Guinea pigs are regarded as the more relevant animal model for HSV-2 infection because viral replication occurs in the genital mucosa, causing lesions that can reoccur after initial resolution, similar to the infection cycle in humans (13, 25). Thus, this model can test for protection against mucosal infection as well as the establishment of latency in the DRG compartment. However, characterization of the responses of guinea pigs to TLR agonists has been limited, and consequently, it is not clear how well guinea pigs can respond to mouse-optimized adjuvants, partially due to the limited availability of guinea pig-specific immunological reagents. Bourne et al. determined that the inclusion of MPL plus alum (AS04) with gD2 in guinea pigs provided greater protection against lethal infection and recurrent disease than did alum alone but failed to reduce the number of guinea pigs that shed virus vaginally or the length of time during which shedding was detectable (13, 25). Another HSV-2 vaccine that combined gD2 and gC2 with CpG-alum induced elevated NAb titers and protected guinea pigs from acute vaginal disease and recurrent shedding (50), but this vaccine was not compared to unadjuvanted vaccines. While the guinea pig ELISPOT assay has been employed by other groups to measure herpesvirus-specific CMI responses (51, 52), our data represent the first indication of HSV responses specific to a particular viral antigen (gD, gB, and UL40) (Fig. 6) and demonstrate the superiority of CpG-emulsion over MPL-alum in the induction of T cell responses to all three HSV-2 antigens employed: gD, gB, and UL40.

Despite the potency of gD as a NAb target, previous HSV-2 prophylactic and therapeutic gD2-based vaccines have not demonstrated a substantial capacity to induce robust T cell responses in the clinic. Two trials were conducted with HSV-2-seronegative individuals who were vaccinated three times over a 6-month period with gD2 plus gB2 adjuvanted with the squalene-based oil-in-water emulsion MF59 (8). Elevated NAb titers to both gD2 and gB2 were detected in the range of levels observed during natural infection, but only partial and transient protection from subsequent HSV-2 infection was achieved, and no T cell responses were reported (8). In one study, HSV-1/2-double-seronegative individuals were vaccinated with the gDt subunit adjuvanted with AS04, in which the TLR4 agonist MPL is complexed with aluminum salts, a vaccine often called Simplirix. The primary endpoint of disease prevention in seropositive individuals was not achieved, and no T cell responses were reported (26). A follow-up study with the same vaccine tested on HSV-1/2-double-seronegative women elicited enzyme-linked immunosorbent assay (ELISA) and NAb titers but found an overall vaccine efficacy of only 20% (10, 53). Modest T cell responses were observed after in vitro gD stimulation of PBMCs, but only CD4 and not CD8 T cell responses were detected. Correspondingly, many clinical studies have indicated that alum is generally a poor inducer of T cell responses and that the addition of MPL to alum as AS04 does not impart signals for the robust amplification of CD8 T cell responses (reviewed in reference 54). Our studies highlight that the induction of robust T cell responses to HSV-2 antigens is achievable in both mice and guinea pigs with an appropriate adjuvant selection, such as a CpG-based formulation.

Recent clinical success has been reported for a novel therapeutic antigen-adjuvant in a phase II dose-escalation trial of HSV-2-infected individuals (55). This vaccine combines the traditional NAb target gD2 with a novel T cell target antigen, ICP4, and adjuvants them with matrix-M2, a preparation of stable, immunogenic complexes composed of saponin, cholesterol, and phospholipid (56). ICP4 was identified by using a high-throughput proteomic screen of T cells from HSV-2-infected but asymptomatic individuals and HSV-2-exposed but seronegative individuals (57). Subjects who received three intramuscular doses of this vaccine at 3-week intervals experienced a 50% decrease in viral shedding and a 69% reduction in the rate of genital lesions compared to placebo, which continued for 6 months postvaccination (55). T cell responses have not been reported for this trial, but adjuvants with compositions similar to that of matrix-M2 have been shown in animal models to promote major histocompatibility complex (MHC) class I cross-presentation and the promotion of CTL responses (58). It will be interesting to learn if CD8 T cell responses specific for ICP4 are detectable in these vaccinated individuals.

Although several vaccine formats have achieved success in mouse and guinea pig animal models, this success has not historically translated to a substantial or sustained reduction in the rate of lesion occurrences in patients or the rate of infection in uninfected individuals. As an understanding of the host mechanisms involved in the prevention of HSV-2 latency has become clearer, it is apparent that the development of a robust HSV-2-specific CD8 T cell response is critical (59, 60). In the mouse model, there is a positive correlation between the number of latently infected neurons and the number of reactivations (61), and equivalent data have been established in the guinea pig model of HSV (62). Therefore, any vaccine that can effectively nullify or reduce the establishment of latent HSV genomes in guinea pigs is more likely to translate into a more effective human prophylactic vaccine. Our data identify a candidate vaccine antigen for T cell targeting in the form of UL40, to which robust T cell responses were observed in mice and guinea pigs. We additionally identified a CpG-emulsion adjuvant preparation as the superior choice for driving these T cell responses. It should be noted that we did not test the individual components of the CpG-emulsion adjuvant in our guinea pig studies, and therefore, we cannot discern the contribution of the individual adjuvant components compared to the combination in regard to the immunogenicity profile or vaccine efficacy. Additionally, we did not examine T cell responses in unadjuvanted vaccine groups of guinea pigs, and therefore, we can compare T cell responses only between the adjuvanted groups tested. Nevertheless, our data demonstrate that a vaccine composed of both B cell and T cell antigenic targets, whether subunit or whole inactivated virus, delivered with an appropriate adjuvant preparation can drive HSV-2-specific CD8 T cell immunity and prevent disease establishment in guinea pigs and therefore provides the basis for further development of this vaccine format for the prevention of HSV-2 infection as a stand-alone strategy or as a component of a heterologous vaccine approach.

MATERIALS AND METHODS

Ethics statement.

All mouse and guinea pig procedures were performed in accordance with federal, state, and institutional guidelines in an AAALAC-accredited facility under specified protocols (ACF 086-11, ACF 12-0009-12, and ACF 10-0001-12) and were approved by the MedImmune Institutional Animal Care and Use Committee. MedImmune maintains a class R research license with the U.S. Department of Agriculture (USDA) and applies the standards for the institutional animal care and use program as outlined in the Guide for the Care and Use of Laboratory Animals (63).

HSV-2 UL40 expression and purification.

The HSV-2 UL40 gene encodes the small subunit of the ribonucleotide reductase (GenBank accession no. YP_009137192.1), and the nucleotide sequence was codon optimized by DNA 2.0 (Menlo Park, CA) and cloned into the pJexpress 401 vector. The final construct contained an N-terminal His₆ tag and an SSG spacer, followed by UL40 residues 2 to 337. A Strep tag was added to the C terminus of the UL40 sequence for Western blot detection. For the expression and purification of UL40, the plasmid was transformed into E. coli BL21(DE3)/pLysS cells and grown in LB medium. Cells were grown to an A₆₀₀ of 0.8 and induced with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) for 4 h at 37°C. The cells were harvested and resuspended in buffer with EDTA-free complete protease inhibitors (Roche) and then lysed on ice by sonication. The resulting lysate was clarified by centrifugation for 20 min at 20,000 × g at 4°C, and immobilized-metal affinity chromatography (IMAC) was performed by using Ni-nitrilotriacetic acid (NTA) resin (Qiagen). The Ni-NTA resin was preequilibrated in binding buffer (50 mM Tris [pH 8.0], 200 mM NaCl) and then loaded with the clarified protein sample. The resin was washed with a solution containing 50 mM Tris (pH 8.0), 200 mM NaCl, and 30 mM imidazole and then eluted with a solution containing 50 mM Tris (pH 8.0), 200 mM NaCl, and 300 mM imidazole. The Ni-NTA-eluted protein was pooled and run over a Superdex S-200 SEC column (GE Healthcare) with a mobile phase of 20 mM KPO₄ (pH 6.5) and 150 mM NaCl. The fractions containing UL40 were pooled, diluted 10-fold into 20 mM KPO₄ (pH 6.5), bound to a Mono-Q column (GE Healthcare) preequilibrated with 20 mM KPO₄ (pH 6.5), and eluted with a linear gradient of a solution containing 20 mM KPO₄ (pH 6.5) and 1 M NaCl. The eluted UL40 protein was pooled, concentrated to 2.5 mg/ml, dialyzed against storage buffer (20 mM histidine [pH 6.5], 150 mM NaCl, 10% sucrose), and stored at −80°C.

HSV-2 gD expression and purification.

HSV-2 gD amino acids 25 to 281 were cloned in frame behind the honey bee melittin signal peptide in the pCLD mammalian expression plasmid. A Strep tag was added to the C terminus of the protein for purification and Western blotting purposes. The gD pCLD expression plasmid was stably transfected into CHO cells adapted for suspension growth (existing cell line developed by MedImmune, Cambridge, UK) by using standard techniques. Stable clones were tested for protein expression and growth characteristics, and ideal clones were frozen for future protein production. gD protein was produced by growing stable CHO clone cells in CD-CHO medium (Invitrogen) supplemented with 65 μM methionine sulfoximine (MSX) until they reached a density of 1.5 × 10⁶ cells/ml, at which point they were split 1:3. The cells were grown in baffled shaking flasks at 140 rpm in a 37°C incubator with 5% CO₂ until the desired expression volume was reached. At this point, a proprietary glucose feed was used to dilute the cells 1:2 to reach a final cell density of 1 × 10⁶ cells/ml, and the cells were grown for 7 days, with feeding every other day, to maximize protein expression and cell viability (upon harvest, the cells were >90% viable). After protein expression was completed, the cell culture supernatant (SN) was harvested for protein purification. The medium was clarified by using a 10-min centrifugation step at 2,500 × g and then filtered through a 0.22-μm filter. The filtered medium was concentrated 25- to 30-fold, and 10 ml of the supernatant was loaded onto a 1-ml Strep-Tactin column (IBA). Standard protein purification was performed according to the manufacturer's guidelines.

HSV-2 gB expression and purification.

Truncated gB2 (727t) was prepared from Sf9 (Spodoptera frugiperda) cells infected with a recombinant baculovirus expressing this protein in a manner modified from the one previously described for gB1 (730t) (64). Briefly, soluble gB2 (727t) was purified from the supernatant fluids of baculovirus-infected Sf9 cells. The supernatant containing the glycoprotein was clarified by centrifugation, dialyzed against 0.1 M Tris–0.15 M saline (pH 7) (TS), and passed over a column of anti-gB monoclonal antibody A22 (34) that was coupled to Sepharose 4B. The column was washed with TS, and gB2 (727t) was eluted with 0.1 M ethanolamine, concentrated by using a YM3 membrane (Amicon), and dialyzed against PBS. The source of gB2 (727t) was HSV-2 strain 333 (6).

Size exclusion chromatography with multiangle light scattering.

The SEC-MALS experiments were performed by using an Agilent 1260 high-performance liquid chromatography (HPLC) system connected in series with a Wyatt Heleos light scattering detector followed by a Wyatt Optilab T-rEX differential refractive index detector. An Agilent SEC3, 300-Å column was used for protein separation. The UL40 experiments were performed with a mobile phase of 20 mM histidine (pH 6.0) and 150 mM NaCl, while the gD experiments were performed with a mobile phase of 25 mM Tris (pH 8.0) and 150 mM NaCl. The data were collected and analyzed by using ASTRA V software according to the manufacturer's guidelines. The UV extinction coefficients used for UL40 and gD2 were 985 ml/g · cm and 1,503 ml/g · cm, respectively, while the protein dn/dc value used was 0.185, and the glycan dn/dc value used was 0.135.

Western blotting.

Standard Western blotting procedures were used to confirm the identity of the purified UL40 and gD2 proteins by using penta-His antibody (Qiagen), anti-Strep antibody, or a rabbit polyclonal anti-gD2 antibody, followed by a goat anti-mouse horseradish peroxidase (HRP)- or anti-rabbit HRP-conjugated secondary antibody (Jackson ImmunoResearch). To visualize the bands, the SuperSignal West Dura Extended Duration substrate (Pierce) was used according to the manufacturer's guidelines.

HSV-2 strain MS propagation.

Vero cells and HSV-2 MS were obtained from the American Type Culture Collection (Manassas, VA). Vero cells were propagated in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin G, and 100 mg/ml streptomycin, here referred to as “Vero cell medium.” Wild-type HSV-2 MS for guinea pig challenges was propagated and titers were determined on Vero cell monolayers. HSV-2 MS used for formalin-inactivated HSV-2 (FI-HSV-2) was grown in Vero cell medium supplemented with 2% FBS in T-225 flasks. Infection of the Vero cell monolayer was allowed to proceed for 2 to 3 days until nearly 100% CPE was observed by using microscopy. Medium was aspirated from the flasks, replaced with 10 ml of Hanks' balanced salt solution (HBSS), and stored at −70°C until frozen. Frozen flasks of virus were left at room temperature just until thawed, and the contents were transferred to 50-ml conical tubes. Virus preparations were sonicated on ice by using a Misonix XL-2000 probe sonicator (Qsonica, LLC, Newtown, CT) three times for 30 s followed by 30 s of rest. Following sonication, the virus was centrifuged at 4°C for 5 min at 900 × g to pellet the remaining cell debris. To concentrate the virus, the supernatant containing HSV-2 MS was added to precleaned ultraclear tubes (30 ml; Beckman Coulter), and 7 ml of 25% (vol/vol) sucrose in HBSS was carefully pipetted into the bottom of the tube as described previously (27, 65). The virus preparation was centrifuged for 4 h at 25,000 rpm at 4°C with a Beckman SW28 rotor. The supernatant was removed from the tubes, and the resulting pellet was resuspended in HBSS by sonication and low-speed centrifugation as described above. Viral PFU were quantitated, and the remaining virus was frozen at −70°C until formalin inactivation.

Formalin inactivation of HSV-2.

To inactivate concentrated HSV-2, an equal volume of a 1:2,000 dilution of a 37% (wt/vol) formaldehyde solution (Sigma) in HBSS was added to the virus as described previously by Morello et al. (27), with slight variations. Following the addition of formalin, the preparation was incubated for 3 days on an end-over-end rotator at 37°C. After incubation, an equimolar amount of sodium bisulfite was added to quench the formalin. The inactivated-virus preparation was dialyzed against HBSS by using a Slide-a-Lyzer dialysis cartridge (3,500-molecular-weight cutoff [MWCO]; Thermo Scientific) for 1 h overnight and again for 1 h. For long-term storage and stability, one part 50% sucrose-HBSS solution was added to four parts inactivated virus (1:4) for a final sucrose content of 10%. The inactivated virus was aliquoted and frozen at −70°C. To ensure that inactivation was complete, inactivated HSV-2 was tested by a standard plaque assay on Vero cell monolayers. PFU-equivalent titers in the inactivated virus preparation were calculated to be 1.8 × 10⁹ PFU/ml based on the titer of the input virus prior to formalin inactivation.

Mouse immunogenicity studies. (i) Vaccination for in vivo mouse studies.

For immunization studies, C57BL/6 mice (aged 6 to 8 weeks; Charles River) were immunized at days 0 and 14 intramuscularly with 10 μg HSV-2 gD and 10 μg HSV-2 UL40 antigens formulated in a volume of 100 μl with AddaVax (50%, vol/vol) with or without 20 μg CpG ODN 2395, 50 μg poly(I·C), 20 μg GDQ, or 20 μg MPL (all from InvivoGen) or with 100 μg alum (Alhydrogel; Brenntag) plus 20 μg MPL. Antigens were added last to the adjuvant mixture and then vortexed for 10 s before administration. Mice underwent terminal bleeds via cardiac puncture on day 28, followed by splenectomy.

(ii) Cytokine detection by ELISPOT and Luminex assays.

Mouse spleens were processed by forcing spleens through 70-μm-pore-size mesh screens (BD Biosciences), rinsing with 5% FBS–RPMI 1640, and centrifugation at 300 × g for 10 min at 4°C before red blood cell (RBC) lysis was performed with ammonium chloride-potassium (ACK) lysing buffer (Invitrogen) for 10 min at room temperature (RT). For the mouse IFN-γ ELISPOT assay (BD Biosciences), 96-well plates were coated with anti-mouse IFN-γ antibody overnight at 4°C according to the manufacturer's instruction. Plates were washed with mouse culture medium (RPMI 1640, 10% FBS, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and a 1:1,000 dilution of beta-mercaptoethanol) and then blocked for 2 h with medium at RT. Splenocytes at 5 × 10⁵ cells/well were stimulated with 1 μg/ml HSV-2 UL40 and HSV-2 gD peptide pools (15-mers overlapping by 11 amino acids covering the entire protein sequence of UL40 and residues 25 to 281 of gD) for 48 h. Plates were washed and incubated with a 1:250 dilution of biotinylated anti-mouse IFN-γ detection antibody (BD Biosciences) for 2 h, followed by streptavidin-HRP for 1 h at RT. The development of a signal occurred by exposure to the 3-amino-9-ethylcarbazole (AEC) substrate (BD Biosciences), and then scanning and analysis were performed on the CTL ImmunoSpot analyzer (Cellular Technologies Limited). For measurement of levels of secreted cytokines by a multiplex Luminex assay, 5 × 10⁵ splenocytes/well from immunized mice were plated onto 96-well round-bottom plates and stimulated with HSV-2 UL40 and HSV-2 gD antigens at 5 μg/ml for 48 h. Cell-free SNs were harvested and quick-frozen for future analysis. Mouse cytokines were assayed by using the mouse MilliPlex kit (Millipore) according to the manufacturer's instructions.

(iii) Flow cytometric techniques for an intracellular staining (ICS) assay and in vivo CTL detection.

Splenocytes were plated at 1 × 10⁶ cells/well in 96rb plates and stimulated with 1 μg/ml HSV-2 gD and HSV-2 UL40 peptide pools for 1 h, followed by 5 h in the presence of GolgiPlug (BD Biosciences). Cells were transferred to a 96Vb plate, centrifuged, and washed with fluorescence-activated cell sorter (FACS) staining buffer (BD Biosciences). Staining for cell surface markers was performed with rat anti-mouse CD3e-BV421, rat anti-mouse CD4 peridinin chlorophyll protein (PerCP)-Cy5.5, rat anti-mouse CD8a allophycocyanin (APC)-H7 (BD Biosciences), and Live/Dead stain-Blue reagent (Invitrogen), followed by fixing with 2% Cytofix (BD Biosciences). Cells were resuspended in BD Perm Wash permeabilization buffer (BD Biosciences); incubated for 30 min at RT; and then stained with a cocktail containing IFN-γ–APC, IL-2–phycoerythrin (PE), and TNF-α–Alexa 488 (BD Biosciences). Acquisition (100,000 events/sample) was conducted on a BD Biosciences LSR2 cytometer utilizing FACS DIVA software. Subset and Boolean gating analyses (for multifunctional T cells) were performed by using FLOWJO software (TreeStar, Menlo Park, CA). For the detection of in vivo CTL killing, two populations of naive C57BL/6 splenocytes were labeled by incubation with 10 μM and 1 μM CFSE for 15 min at 37°C and then loaded with 1 μg/ml HSV-2 UL40 peptide pool (10 μM CFSE targets) or an irrelevant Epstein-Barr virus (EBV) gp350 peptide pool (1 μM CFSE targets) during 1 h of incubation at 37°C. Afterwards, target cells were washed and analyzed for CFSE signals by flow cytometry. Target cells were mixed 1:1, and recipient immunized mice were injected via the tail vein with 10 × 10⁶ cells of each target population in 200 μl. Recipient mice were sacrificed 18 to 24 h later, and splenocytes were isolated and analyzed for CFSE signals via flow cytometry. UL40-specific target killing was normalized to control gp350-specific target killing by using the following formula: % specific killing = (1 − % CFSE^hi cells/% CFSE^lo cells) × 100.

Guinea pig immunogenicity and challenge studies. (i) Guinea pig immunization and challenge.

For immunization studies, female Hartley guinea pigs (aged 5 to 6 weeks; Charles River, Wilmington, MA) were immunized intramuscularly on days 0 and 21. Subunit vaccines were formulated with 10 μg HSV-2 gD, 10 μg HSV-2 UL40, and 10 μg HSV-2 gB antigens in a volume of 200 μl with or without AddaVax (50%, vol/vol) and 20 μg CpG ODN 2395 or 275 μg alum (Alhydrogel; Brenntag) plus 12.5 μg MPL. TLRL and AddaVax were purchased from InvivoGen (San Diego, CA). FI-HSV-2 vaccines were formulated as described above for the subunit antigen vaccines, and only 2e7 PFU-equivalent particles of FI-HSV-2 were added in lieu of subunit antigens. Inoculations were split equally between the hind limbs (100 μl/hind limb). gD and gB contents in 2e7 PFU of FI-HSV-2 were based on data reported in the literature and were estimated to be between 1.98 μg and 2.20 μg and between 0.34 μg and 0.50 μg, respectively (66). On day 28, 3 to 4 ml of blood was drawn by venipuncture to investigate vaccine-induced CMI responses in peripheral blood. On day 35, 0.5 to 1 ml of blood was drawn by venipuncture to measure anti-HSV-2 neutralizing antibody titers and anti-gD2/anti-gB2 binding antibody titers in response to vaccination by an ELISA (see below). On day 41, isoflurane gas-anesthetized guinea pigs were vaginally challenged. Briefly, prior to challenge, the vaginal closure membrane was disrupted by using a moistened calcium alginate-tipped swab, and 100 μl of an HSV-2 MS inoculum containing 5 × 10⁵ PFU of HSV-2 MS was administered by a 20-gauge feeding needle into the vaginal canal. Guinea pig hindquarters were elevated to prevent leakage of the challenge inoculum until the effects of anesthesia wore off. Guinea pig clinical scoring and documentation of HSV-2-associated sequelae were performed for 14 days postchallenge (days 42 to 55) (Fig. 9). All experimental data from guinea pig immunogenicity and challenge experiments in this study were combined into one data set and represent results from two independent experiments, with the exception of guinea pig ELISPOT data, which are from one experimental data set (see below).

FIG 9.

Study design for guinea pig vaccination and challenge. Guinea pigs were vaccinated on day 0 (D0) and day 21. Blood draws were collected 1 and 2 weeks after boost vaccination for measurement of antigen-specific CMI responses (day 28) and measurement of binding and neutralizing antibody responses (day 35). Vaccinated guinea pigs were challenged 3 weeks following the administration of the boost vaccination. Vaginal swabs were collected at 1, 2, 5, and 9 days postchallenge for enumeration of HSV-2 shedding. Daily observation and scoring were done for 2 weeks after challenge. Guinea pigs were euthanized on days 67 to 71, and DRG were isolated to enumerate the latent viral copy numbers of HSV-2.

(ii) Guinea pig IFN-γ ELISPOT assay.

One week after the second vaccination, 3 to 4 ml of peripheral blood was drawn into EDTA collection tubes, layered onto Lympholyte-Mammal cell separation medium (Cedarlane Labs), and centrifuged at 800 × g for 20 min. Interface PBMCs were extracted, washed, and plated at 5e5 cells/well in an ELISPOT plate (Millipore) coated with anti-guinea pig IFN-γ capture antibody (clone 5036; H. Schäfer, Robert Koch Institut). PBMCs were stimulated with 1 μg/ml HSV-2 UL40 and gD2 overlapping peptide pools for 48 h. Plates were washed and incubated with biotinylated anti-guinea pig IFN-γ detection antibody (clone 10595; H. Schäfer, Robert Koch Institut) for 2 h, followed by streptavidin-alkaline phosphatase (Mabtech) for 1 h. After washing, plates were developed by using the BCIP (5-bromo-4-chloro-3-indolylphosphate)-NBT (nitroblue tetrazolium) substrate (Mabtech), and spots were scanned and counted by using the CTL ImmunoSpot analyzer (Cellular Technologies Limited).

(iii) HSV-2-neutralizing Ab assay.

Two weeks after the second vaccination, 0.5 to 1 ml of peripheral blood was drawn into serum separation tubes (Becton Dickinson) and centrifuged for the collection of serum according to the manufacturer's recommendations. Collected sera were heat inactivated at 56°C for 45 min and then stored at −70°C until used in the assay. Heat-inactivated serum was serially diluted 2-fold in Vero cell medium in 96-well round bottom-plates, and an equal volume of medium containing HSV-2 MS was added to the serum. Twofold serial dilutions of pooled human gamma globulin (Seracare, Milford, MA) was used as a positive control and to monitor assay-to-assay variability. The virus-guinea pig serum mixture was shaken on an orbital plate shaker for 2 min at 450 rpm and then placed at 37°C in 5% CO₂ for 1 h. After incubation, the virus-serum mixture (containing HSV-2 at a multiplicity of infection [MOI] of 0.1) was plated onto Vero cell monolayers that were ∼90 to 100% confluent and allowed to incubate for 3 days at 37°C in 5% CO₂. Wells were scored for CPE by light microscopy, and the neutralizing titer was defined as the highest reciprocal dilution of serum that protected >50% of the Vero cell monolayer from infection and CPE.

(iv) gD2- and gB2-specific antibody responses.

Vaccine-induced binding antibody titers against gD2 and gB2 were measured by an ELISA. Recombinant HSV-2 proteins were diluted in PBS to a concentration of 1 μg/ml, and 96-well MaxiSorp plates (Nunc) were coated with 100 μl of the mixture. Plates were incubated overnight at 4°C. Plates coated with the antigen were washed four times with PBS–0.05% Tween 20 (PBST) and blocked at room temperature for 1 h with blocking buffer (PBST, 2% bovine serum albumin [BSA], and 2% nonfat dry milk). Postvaccination guinea pig sera were diluted in blocking buffer starting at a 1:60 dilution. Prebleed sera from each treatment group were pooled and diluted as described above for postvaccination sera. One hundred microliters of diluted serum samples was transferred onto 96-well plates and incubated for 1 h at 37°C. Plates were washed five times with PBST. Guinea pig antibodies were bound with 100 μl of a 1:5,000 dilution of donkey anti-guinea pig IgG-HRP (Jackson ImmunoResearch) for 1 h at 37°C and washed as described above. One hundred microliters of a 3,3′,5,5′-tetramethylbenzidine (TMB)-ELISA substrate solution (Thermo Fisher Scientific) was added to the plate, and reactions were allowed to proceed at room temperature for 30 min in the dark. Reactions were stopped by adding 100 μl of phosphoric acid to the mixture. A₄₅₀ values were read by using a spectrophotometer. Fit spline/LOWESS was used to interpolate the endpoint titers, which were defined as the dilution of serum where the absorbance values were >3 times the prebleed absorbance values within each treatment group (GraphPad Prism 5.0; GraphPad, La Jolla, CA).

(v) Quantitation of viral shedding postchallenge.

Quantitation of viral shedding was done on days 1, 2, 5, and 9 postchallenge. Briefly, a PBS-premoistened calcium alginate-tipped swab was inserted into the vaginal canal, rotated gently three times, and then placed into a 15-ml conical tube containing 1 ml of Vero cell medium. The tubes containing the swabs were frozen on dry ice and then transferred to −80°C for storage. On the day of the assay, swabs were removed from the freezer, quickly thawed in a 37°C water bath just until thawed, and quickly placed on ice. Swabs were vortexed for 30 s, and swabs were removed from the tubes. The resulting volume was then centrifuged for 5 min at 2,000 rpm, and the supernatant was moved to new tubes. Serial 1:10 dilutions of virus swabs were diluted in Vero cell medium. A 1-ml volume of each dilution was transferred onto Vero cell monolayers in 6-well tissue culture (TC) plates (Corning) and incubated at 37°C for 2 h. Medium was aspirated from the wells, replaced with Vero cell medium containing 0.1% human gamma globulin, and allowed to incubate for 3 days at 37°C. After 3 days, medium was aspirated from the wells, and cell monolayers were fixed with a 10% formaldehyde solution for 15 min. Following fixation, monolayers were stained with a solution containing 1% crystal violet in 20% ethanol. Plates were gently rocked for 30 min and then washed with deionized water (diH₂O). Plates were dried, and plaques were counted by visual inspection.

(vi) Clinical scoring of HSV-2 disease.

Following HSV-2 challenge, clinical disease was monitored for 14 days. Clinical signs were recorded and given a score from 0 to 4 based on the severity of disease, as previously described (67), where a score of 0 indicating no disease, 1 indicating erythema only, 2 indicating a single vesicle or a few vesicles, 3 indicating large or fused vesicles, and 4 indicating ulcerated lesions. A score of 0.5 was given for slight erythema or swelling, and intermediate scores (1.5, 2.5, and 3.5) were used when the severity fell between the scores described above. HSV-2-associated sequelae that were recorded but not scored were hind-limb paralysis, hematuria, and urine retention. Guinea pigs were euthanized if hind-limb paralysis was observed and the guinea pig did not have the mobility to eat and drink, if there was profound hematuria in two subsequent bladder expressions, or if the animal suffered >20% weight loss. Scores for guinea pigs that were euthanized or died from disease were not included in cumulative disease scores.

(vii) Quantitation of latent HSV-2 genome copy numbers in dorsal root ganglia.

For the quantitation of latent HSV-2 genomes, guinea pigs were euthanized on days 67 to 71 (26 to 30 days postchallenge), and DRG (L3 through S1/S2) were harvested and placed into 15-ml conical tubes that were immediately placed on dry ice. Separate instruments were used for each guinea pig, or alternatively, instruments were cleaned with a 10% bleach solution for 15 min, followed by rinsing with diH₂O. DRG were stored at −80°C until they were processed. DRG DNA isolation was carried out by using the DNeasy blood and tissue kit (Qiagen) according to the manufacturer's protocol, as described previously (27). DRG tissue homogenization was done with a Qiagen TissueLyser at 15 Hz for 20 s in ATL buffer. Following homogenization, tissues were incubated overnight in the presence of proteinase K, and RNase A was added prior to column loading. AE elution buffer (Qiagen) was heated to 56°C, and the mixture was allowed to incubate on the column for 1 min prior to centrifugation. DNA concentrations were determined by A₂₆₀ measurements (NanoDrop; Thermo Scientific).

TaqMan real-time quantitative PCR (qPCR) was utilized for the quantification of HSV-2 genomes in DRG DNA by using an ABI 7900HT thermocycler. All primers and probes were ordered from Integrated DNA Technologies (San Diego, CA). The primers and probe were based on HSV-2 gB (UL27), as previously described (68), and are as follows (underlined nucleotides were changed to ensure homology to the HSV-2 strain MS gB sequence): primer gBfor (TTTCGGTACGAAGACCAG), primer gBrev (AGCAGGCCGCTGTCCTTG), and probe gBprobe (6-carboxyfluorescein [FAM]-TGGTCCTCCAGCATGGTGATGTTCAGGTCG-black hole quencher 1). qPCR was run in triplicate reactions in 30-μl reaction mixture volumes containing 15 μl TaqMan master mix (1×), 900 nM (each) forward and reverse primers, 250 nM probe, and 300 ng DRG DNA. Latent HSV-2 copy numbers in DRG DNA were generated from standard-curve reaction mixtures containing commercially quantitated HSV-2 genomes (Advanced Biotechnologies, Inc., Columbia, MD) in the presence of 300 ng DRG DNA from naive guinea pigs. Separate reactions to measure guinea pig beta-actin in DRG DNA were used to ensure sample quality. Primers and probes to measure beta-actin levels were described previously (27). A total of 300 nM (each) the forward and reverse primers was used, while 250 nM the probe was used in each reaction mixture. Analysis of quantitative real-time PCRs was done by using SDS software (Applied Biosystems, Foster City, CA). Samples were scored as being positive if the threshold cycle (C_T) in at least two of the three triplicate reactions was <40 cycles. The limit of detection for HSV-2 genome copy numbers in our assay was 9 copies. Samples that did not have a measureable signal by qPCR were given a score of 4.5 copies for graphing and statistical purposes.

Statistical analysis.

The one-way analysis of variance (ANOVA) model with heterogeneous within-group variance was applied for data analysis and comparisons of continuous measurements. For the control group with all values below the detection limit, the nonparametric Kruskal-Wallis method was applied. One-sided P values for comparisons against negative controls and two-sided P values for comparisons between noncontrol treatment groups are reported. SAS 9.3 (SAS Institute, Inc., Cary, NC) and Prism 6.05 (GraphPad Software, Inc., La Jolla, CA) were used for statistical analysis.