Prospects and Perspectives for Development of a Vaccine Against Herpes Simplex Virus Infections

Abstract

Herpes simplex viruses 1 and -2 are human pathogens that lead to significant morbidity and mortality in certain clinical settings. The development of effective antiviral medications, however, has had little discernible impact on the epidemiology of these pathogens, largely because the majority of infections are clinically silent. Decades of work have gone into various candidate HSV vaccines, but to date none has demonstrated sufficient efficacy to warrant licensure. This review examines developments in HSV immunology and vaccine development published since 2010, and assesses the prospects for improved immunization strategies that may result in an effective, licensed vaccine in the near future.

Introduction

Herpesviruses all share a general virion structure that includes a large, linear, double-stranded DNA genome that is densely packed within an icosahedral capsid [1]. The capsid is surrounded by an amorphous tegument that is made up of more than 30 viral proteins. Finally, the virion is surrounded by a lipid envelope that is studded with multiple viral glycoproteins as well as some cellular proteins.

The eight known human Herpesviruses are divided into three subfamilies: Alphaherpesvirinae (Herpes simplex-1 and -2, Varicella-zoster virus); Betaherpesvirinae (Cytomegalovirus, Human Herpesvirus-6 and -7); and Gammaherpesvirinae (Epstein-Barr virus, Kaposi sarcoma Herpesvirus). Members of the three subfamilies differ in genomic content, routes of transmission, sites of latency, clinical manifestations of viral reactivation, and susceptibility to current anti-viral medications [1].

Herpes simplex virus- (HSV-) 1 and 2 have considerable medical impact across all geographic and socioeconomic divisions. HSV infections are common and ubiquitous. Approximately two-thirds of the world's population is infected with herpes simplex virus type 1 and/or type 2 (HSV-1 & HSV-2) that cause ocular, oro-facial (cold sores) and genital herpetic diseases [2]. Primary infection with HSV-1 & HSV-2 in immunocompetent individuals is usually asymptomatic and results lifelong viral latency in sensory ganglia and both the ubiquitous nature of infection and well as the propensity for the virus to cause latency are complicating variables in vaccine development [3,4]. HSV also encodes a plethora of immune evasion genes that complicate the concept of vaccine development [5-7]. HSV-1 is classically associated with oculo-facial infections and HSV-2 with genitourinary infections. However, this anatomical distinction is no longer considered to be as strict [8]. Ocular infections with HSV-1 are the leading cause of infectious blindness in the developed world, with recurrent episodes of viral reactivation leading to progressive scaring and opacity of the cornea [9]. HSV-2 is sexually transmitted and estimated to infect 500 million people worldwide including 23 million new infections per year [10]. Seroprevalence is 16% among 14 to 49 year olds in the United States, with rates in women nearly twice that of men [11,12]. Seroprevalence tends to be higher in developing countries with rates of 80% found in some Sub-Saharan African nations [13].

Acquisition of HSV by infants during parturition can result in serious morbidity and mortality [14-16]. Mothers who acquire infection during pregnancy are at highest risk of transmitting HSV to infants because of higher titers of virus in vaginal fluid and lack of transplacental transfer of HSV-specific IgG to the infant [17]. Numerous studies have evaluated screening protocols for pregnant women, anti-viral prophylaxis of HSV-positive pregnant women, and caesarian delivery in preventing transmission to neonates [17-21]. None of these measures, however, has shown significant protection.

Of grave concern is the fact that genital infection with HSV-2 increases the risk of acquiring HIV infection by three-fold [22]. In countries with high HSV-2 seroprevalence, as much as 50% of HIV infections are attributable to antecedent genital herpes [23]. This increased risk is not due to disruption of the genital mucosa by reactivation as asymptomatic HSV-2 infections have the same rate of transmission and HSV-2 suppression with anti-viral medications does not mitigate risk [24-26]. Latent HSV-2 infection is associated with a robust influx of antigen-specific CD4+ T cells, thus increasing the local concentration of cells permissive for HIV infection [27-29].

One reason an effective vaccine is urgently needed is that asymptomatic shedding of transmissible virus occurs in the absence of clinical symptoms in the majority of cases [30,31]. In fact, following primary infection the majority of infected individuals never has clinically evident reactivations, and therein lies the failure of antiviral medications to dramatically alter the epidemiology of HSV infection [12]. Serological studies have established that only 10% to 25% of serpositive subjects in the United States is aware that they are infected [11,12,31]. Recent studies that have employed extensive ano-genital sampling of infected, asymptomatic individuals have demonstrated that clinically-silent viral shedding occurs frequently and in short bursts [29,32-34]. Thus, anti-viral medication prescribed only to individuals with clinically apparent reactivations has no bearing on transmission resulting from asymptomatic viral shedding. The finding that the majority of new infections passed within sero-discordant couples in the absence of clinical disease underscores the importance of silent reactivation. Furthermore, the reliance of a history of genital herpes lesions in pregnant women is insufficient to accurately determine the risk of perinatal transmission of HSV infection [35]. The challenges of vaccination of seropositive individuals are formidable and beyond the scope of this review, although they have been reviewed in several excellent papers [4,36]. For purposes of this analysis, HSV-seronegative women of child-bearing age are considered the major target population for vaccine development, given the devastating consequences of neonatal HSV infection (reviewed below) [37,38]. Vaccine development targeting at decreasing asymptomatic viral shedding, or limiting recurrent infection rather than preventing primary infection, may be an appropriate endpoint in clinical trial design for some populations.

Primary Infection

Primary infection of multiple cell types, including epithelial cells, fibroblasts, lymphocytes, and neurons, with HSV-1 and -2 proceeds by in a stepwise fashion beginning with the interaction between infectious viral particles and the glycocalyx of permissive cells [39]. The subsequent fusion of the viral envelope with a host cell lipid membrane then delivers the viral capsid to the cytoplasm. Unlike other enveloped viruses such as vesicular stomatitis virus and varicella zoster virus, where attachment and fusion events are mediated by a single viral glycoprotein, the HSVs encode for at least a dozen glycoproteins that mediate viral entry [39]. Some of these glycoproteins are essential for infection while others increase the efficiency of infection but are not necessary.

The classic view of HSV entry is that viral particles fused directly with the host cell plasma membrane at neutral pH. However, electron micrographs of infected cells have demonstrated the presence of enveloped viral particles within the host cell cytoplasm [40]. Subsequent work has demonstrated that HSV infection can be mediated by the endocytic pathway, in a pH-dependent or –independent manner [41-43]. Both the classic and endocytic entry pathways require the same fourglycoproteins: gB, gH/gL, and gD [44,45]. Subtle differences in these pathways, however, may in part explain the failure of glycoprotein-based subunit vaccine candidates (discussed below).

Irrespective of the route of entry, viral particles must de-envelope in order to release the viral capsid into the host cell cytosol. De-enveloped capsids are then transported to the nuclear membrane along microtubules [46] and interact with nuclear pore complexes leading to extrusion of viral DNA into the nucleus [47]. Upon delivery of the viral genome to the nucleus two types of infection may proceed: latent or lytic.

Latent and Lytic Replication

A distinguishing feature of Herpesviruses is the establishment of latency following primary infection [48]. Latency is typified by expression of few or no viral proteins and RNA transcripts, and maintenance of the viral genome as circularized episome that is maintained in the host cell nucleus by tethering to histone proteins. Following initial productive infection primarily in epithelial cells, retrograde axonal transport delivers viral DNA to sensory neurons within dorsal root ganglia, the sight of latency for all the alpha Herpesviruses [49]. Oculo-facial infections resolve with establishment of latency in the trigeminal ganglia, whereas genital infections lead to latency in the pelvic splanchnic neurons.

Periodically lytic replication may be induced. Extensive remodeling of the nucleus accompanies productive viral infection [50,51]. Expression of viral genes proceeds in a coordinated cascade beginning with immediate early genes, the expression of which does not require translation of any viral proteins [52]. Delayed early genes follow prior to the onset of viral DNA replication. Both immediate and delayed early genes have regulatory functions. Following viral DNA replication, expression of late genes, which code for predominantly structural proteins, allows for packaging of viral DNA copies into new capsids. Acquisition of tegument proteins and a host cell-derived lipid envelop studded with viral glycoproteins results in release of infectious viral progeny [53].

Innate and Adaptive Immunity to HSV

The ability to induce expression of neutralizing antibodies has been the gold standard of assessing vaccine efficacy in pre-clinical and clinical trials. However, the role of humoral immunity in preventing primary infection with HSV and limiting the extent of reactivation is less clear. For example, primary immunodeficiencies typified by reduced humoral immunity, such as common variable immunodeficiency or agammaglobulinemia, generally are not associated with increased risk of severe viral infections [54]. One reason humoral responses do not fully eliminate viral shedding is that many infectious virions produced during lytic replication spread laterally at the junctions between epithelial cells, rather than at apical membranes where antibody would have access to viral particles. Lateral spread of viral particles is mediated by viral glycoproteins gE and gI, but the inability of antibodies to gain access to the paracellular compartment makes these glycoproteins poor vaccine candidate antigens [55,56]. Neutralizing antibodies undoubtedly have a role, however, in mitigating the risk of perinatal transmission. Transplacental transfer of IgG specific for HSV to the developing fetus likely contributes to the reduced risk of neonatal infection in infants born to mothers whose primary HSV infection preceded pregnancy [17].

The essential role of cell mediated immunity (CMI) in control of HSV infection can be inferred from the fact that reduced CMI responses due to iatrogenic immune suppression or HIV infection lead to a significantly increased risk of developing severe HSV infections [57,58]. Recruitment of antigen-specific CD8+ T cells to lesional tissue is a necessary step in clearing infectious virus and infected cells [59]. Biopsy specimens from genital infections have furthermore demonstrated an inverse correlation between lesional CD8+ T cell content and the severity of viral reactivation [29]. Recruitment of CD4+ T cells is also essential for controlling infection; however, the influx of cells that are susceptible to infection with HIV has been proposed as one mechanism by which genital HSV infections increase the risk of HIV transmission [27-29]. A detailed discussion of the range of epitopes recognized in the context of HSV infection is beyond the scope of this review, which is intended to provide an update of vaccines currently in the clinic, but the immunology of the anti-HSV response has been the subject of several recent papers [36,60,61]. Of particular note is the importance of CD8+ T cell epitopes gB342-350 and gB561-569 in protection against reactivation, both in seropositive asymptomatic individuals and in protection of HLA transgenic mice against ocular HSV infection [62]. Another elegant recent study, employing cell-type-specific laser capture microdissection, transcriptional profiling and T-cell antigen receptor beta-chain (TCR-beta) genotyping on sequential genital skin biopsies, showed that CD8alphaalpha(+) T cells are the dominant resident population of DEJ CD8(+) T cells that persist at the site of previous HSV-2 reactivation [63]. Hopefully, elicitation of these key cells for immune surveillance and initial containment of HSV-2 reactivation in human peripheral tissue can be the basis for future innovative vaccine approaches.

In addition to adaptive immune responses, recent evidence implicates the innate immune system in protection from severe HSV infection. The intracellular toll like receptors (TLR) 3 and 9 have been shown to be necessary for induction of interferon responses to HSV infection. Interferons are essential for inducing an anti-viral state in neighboring cells, inducing recruitment of CD8+ T cells to sites of HSV infection, and stimulating maturation of dendritic cells, which are indispensable in priming adaptive CTL responses. Likewise, natural killer cell recruitment and activation is in part dependent on interferon signaling. Augmentation of immune responses to HSV has been achieved in murine models with co-administration of the TLR9 agonist cPG [64]. Adjuvant development focused on maximizing stimulation of innate immunity may have a large impact on future candidate HSV vaccines [65,66].

Lessons from Animal Models of HSV Vaccines

Animal models for HSV infection have been important in furthering our understanding of key pathophysiological mechanisms [67]. However, there are important differences in the natural history of HSV infection in humans versus mice. Most importantly, primary HSV infection of mice does not lead to a persistent, predominantly latent infection as it does in humans. Accordingly, murine studies rely on lethality as an endpoint, which is not necessarily relevant to the natural history of HSV infection in humans. Guinea pigs, in contrast, do establish latency and undergo spontaneous reactivation in the genital track. Though more physiologically relevant to humans, the guinea pig model suffers from a paucity of reagents making it more difficult to conduct in depth immunological analyses. The use of mouse and guinea pig models for studying interventions against genital HSV infection is the subject of several recent reviews [67,68]. Some of the more instructive preclinical vaccine studies from these animal models are reviewed below.

Some of the first animal model studies of HSV vaccines were performed in the 1970s. Experimental evaluation of vaccines for genital HSV infection has largely performed in the guinea pig model, because of its unique relevance to human HSV-2 disease - in particular, the propensity for animals infected by vaginal route to undergo spontaneous genital tract recurrences [69-71]. Early studies mostly consisted of evaluation of immunogenicity of crude glycoprotein extracts, or evaluation of immune response following animal inoculation with virus [72-76]. The first suggestion that these proteins might be effective in controlling acute and recurrent genital HSV disease came in 1981, when Thornton et al. reported that vesicles of plasma membrane prepared by treatment of HSV2-infected Vero cells with dithiothreitol and formaldehyde were antigenic in guinea-pigs and also protected these animals against intravaginal challenge with HSV-2 [69]. The advent of cloning technologies for expression of viral glycoproteins enabled investigation of adjuvanted, recombinant HSV-1 gD as a candidate vaccine in the guinea pig model. Studies reported by Berman et al. indicated that vaccinated animals formed antibodies that neutralized both HSV-1 and HSV-2, and were protected against disease (although not against recurrent infection) following challenge with HSV-2 by intravaginal route [77,78]. Studies using lectin-purified gD and gB subunit vaccine demonstrated similar protection, not only against primary infection, but recurrent disease [79]. Adjuvant choice plays a crucial role in protection in adjuvanted glycoprotein studies in the guinea pig model, with adjuvants capable of eliciting robust cellular immune responses in additional to antibody responses appearing to enhance protection against disease [64,80-86]. Other novel routes of delivery of adjuvanted HSV glycoproteins have been evaluated, including intranasal administration [87,88]. Variations of this approach have included recombinant adjuvanted HSV gD vaccine co-administered with the ICP4 protein [89], or in combination with another HSV glycoprotein, gC [90]. Another interesting study examined the proprietary adjuvant IC31 in combination with HSV-2 glycoprotein gD administered via intranasal, intradermal, or subcutaneous routes. Immunization conferred 80-100% protection against an otherwise lethal vaginal HSV-2 challenge with amelioration of viral replication and disease severity in the vagina [91]. Adjuvanted combinations of three glycoproteins, gB, gD and gE, have been evaluated in the guinea pig model, using aluminum hydroxide-based adjuvants [92]. In this study, guinea pigs were vaccinated after establishment of recurrent genital HSV disease. Immunization results in a reduction in recurrences, supporting a potential role for an HSV vaccine as an immunotherapeutic approach for individuals with recurrent genital HSV infection.

In addition to recombinant protein vaccines, a number of preclinical studies have been performed in the guinea pig genital HSV model with DNA vaccines. Initial studies focused on gD and gB plasmid vaccines [93-96], either alone, in combination, or in a “prime-boost” sequence of DNA vaccine priming followed by recombinant protein boosting [97]. As with recombinant glycoprotein, addition of adjuvant to HSV DNA vaccines appears to be important in maximizing immune response. One study demonstrated efficacy of a gD-based DNA vaccine when administered with bupivacaine [98]. A DNA-based vaccine developed by Vical corporation that includes the coding region for full-length gD2 plus a proprietary cationic lipid-based adjuvant, Vaxfectin, demonstrated excellent protection in a murine model of lethal intravaginal challenge [99]. Compared with unadjuvanted immunized animals, mice immunized with full the full length gD2-encoding plasmid in conjunction with Vaxfectin had significantly improved survival at 20 days following HSV challenge with 500 times the LD₅₀ (0% v. 80%, unadjuvanted v. adjuvanted, respectively). The only correlate of protection besides survival tested was antibody production. Vaxfectin-adjuvanted HSV-2 DNA vaccines have also been studied in the guinea pig model of genital HSV disease [100]. This study examined a gD2 plasmid vaccine combined with HSV-2 UL46 and UL47. Vaccination significantly reduced viral replication in the genital tract, provided complete protection against primary and recurrent genital skin disease following intravaginal HSV-2 challenge, and significantly reduced latent HSV-2 DNA in the dorsal root ganglia compared to controls. The vaccine also reduced the frequency of recurrences in animals already latently infected. Another recent DNA vaccine study in a mouse HSV-2 challenge model indicated that the addition of ICP27 to gD- and gB-based vaccines regimens improved efficacy, presumably by an enhancement of the T cell response [101]. Finally, HSV-2 gD2 DNA vaccines were recently examined in combination with two conserved HSV-2 genes necessary for virus replication, UL5 (DNA helicase) and UL30 (DNA polymerase), in a “prime-boost” approach with formalin-inactivated HSV-2 (FI-HSV2) formulated with monophosphoryl lipid A (MPL) and alum adjuvants, in a mouse model of genital HSV-2 challenge [102]. Overall, DNA HSV-2 vaccines continue to be an area of intensely active research, and human trials are warranted. Indeed, two DNA vaccine plasmid constructs encoding HSV-2 gD have moved forward in phase 1 clinical trials [103], and are considered in the next section of this review.

Other recent pre-clinical developments in live-attenuated HSV vaccine methods deserve mention. An ICP0 “knock-out” virus showed significantly greater protection in lethal dose experiments in mice compared with a subunit vaccine preparation of gD2 [104]. Viral shedding was also significantly reduced following live-attenuated vaccine strain administration over subunit vaccination. The only immune parameter assessed in initial studies was production of anti-gD neutralizing antibodies, although a follow up study measured the pan-HSV-2 antibody response following vaccination with ICP0 deletion vaccines, and found in mice and guinea pigs quantitatively correlation between both reductions in viral shedding and increased survival frequency following HSV-2 challenge [105]. Further analyses of novel correlates of protective antibody responses, as well as the induction of CMI in this model, would be of great interest and may help move this interesting candidate vaccine forward into clinical testing.

Another interesting and novel vaccine strategy dubbed “prime and pull” was recently published [106]. Vaccination of mice with an attenuated HSV-2 strain followed by a single topical application of the T cell-tropic chemokines CXCL9 and CXCL10 to the vaginal mucosa of vaccinated mice resulted in significant recruitment and retention of virus-specific effector memory T cells. This recruitment resulted in significantly improved survival in subsequent lethal challenge studies. Though in the present two-step form this strategy may be difficult to implement in human trials, the underlying validation of the prime importance of local CMI in HSV immunity is important. Less invasive methods of improving recruitment of effector memory T cells to mucosal compartments may be essential to developing HSV vaccines effective in humans.

Recent and Ongoing Clinical Trials of HSV Vaccines

A summary of ongoing clinical trials is provided in Table 1. The viral glycoproteins gD and gB elicit the strongest neutralizing antibody titers to HSV-2 and have demonstrated varying degrees of efficacy in animal models, as reviewed above, and have therefore received the most attention from vaccine developers. However, the results with glycoprotein subunit vaccines have been generally disappointing. Recently, a subunit glycoprotein vaccine by GlaxoSmithKline failed to reduce acquisition of primary HSV-2 infection in seronegative women of reproductive age [107,108]. The recombinant truncated gD2 vaccine administered with alum and 3-O-deacylated monophosphoryl lipid A adjuvants initially showed promise in preventing HSV-2 transmission to women but not men within sero-discordant couples [109]. However, in the follow up trial that enrolled over 8000 HSV-1 and HSV-2 uninfected women ages 18 to 30, no benefit was found in preventing acquisition of HSV-2 infection after up to three doses of vaccine despite the ability of the vaccine to generate neutralizing antibody. Furthermore, the vaccine did not reduce the level of viral shedding in those women who acquired HSV-2 infection during the study period. Surprisingly, there was a small but significant reduction in acquisition of HSV-1, despite the inclusion of only the HSV-2 gD protein in the vaccine. The ∼80% sequence identity between the gD proteins of HSV-1 and -2 may have led to cross reactivity. It is also known that antecedent HSV-2 infection is partially protective against subsequent HSV-1 infection, though not vice versa REF.

Table 1. Recent and Ongoing Clinical Trials.

Vaccine	Population	Adjuvant	Vaccination Schedule	Outcomes	Results
Recombinant subunit
HERPEVAC (GSK) truncated HSV-2 gD	Sero-discordant monogamous heterosexual couples HSV-1 and -2 negative women	Alum and MPL	0, 1, 6 months	Prevention of primary infection	Moderate efficacy against genital herpes in women (seronegative for HSV-1 and HSV-2) but not men [109] Slight protection from HSV-1 infection and disease; no protection from HSV-2 nor reduction in viral shedding [107]
GEN-003 (Genocea) truncated HSV-2 gD and ICP4	Subjects with moderate to severe HSV-2 infection	Matrix M™	0, 3, 6 weeks	Safety, tolerability, reduced viral shedding	Ongoing [Clinical Trials.gov identifier NCT01667341]
HerpV (Agenus) peptides from 22 HSV-2 proteins non-covalently linked to HSC70	HSV-2 seropositives	Stimulon®	1 to 3 doses	Safety, tolerability, immunogenicity	Broad induction of humoral and cell mediated immunity Transient humoral responses to HSC70 Ongoing [ClinicalTrials.gov identifier NCT01687595]
VCL-HM01 (Vical; Phase 2; monovalent) and HB01 (Phase 1; bivalent) DNA vaccines	HSV-2 seropositives 2-9 recurrences/year 18-50 years of age	Vaxfectin®	3 doses Dose escalation	Safety, tolerability, viral shedding, recurrence rate, T-cell response	Ongoing [ClinicalTrials.gov identifier NCT02030301]
Live, replication deficient
HSV529 (Sanofi Pasteur)	HSV-1 and HSV-2 infected and uninfected subjects		0, 1, 6 months	Safety, tolerability, immunogenicity	Ongoing [ClinicalTrials.gov identifier NCT01915212]

GSK, GlaxoSmithKline; MPL, 3-O-deacylated-monophosphoryl lipid A

A subunit vaccine by Genocea, GEN-003, that contains truncated recombinant gD2 and ICP4 has shown promise in preclinical studies [89]. Using a novel and proprietary screening method to identify immune-dominant epitopes, Genocea demonstrated that CD8+ clones reactive to ICP4 distinguished asymptomatic HSV-2-positive or HSV-2-exposed but seronegative individuals from HSV-2-positive patients with frequent, clinically-evident reactivations. The bivalent vaccine adjuvanted with a proprietary, saponin-based compound Matrix M™ stimulated production of neutralizing antibody in immunized mice, as well as CD4+ and CD8+ responses that demonstrated anamnestic responses up to 44 days after vaccination. Vaccination of HSV-2-infected guinea pigs, which develop a clinically-relevant latent/lytic chronic phase of infection, reduced the incidence and severity of reactivation as well as reducing the level of asymptomatic shedding [89]. These finding suggested that the Genocea vaccine is a potentially efficacious therapeutic vaccine candidate. Accordingly, a phase 1/2a double-blind, placebo-controlled, dose-escalation trial is now enrolling with a goal of recruiting approximately 150 subjects with moderate to severe HSV-2 infection (ClinicalTrials.gov identifier NCT01667341). The primary endpoints of this trial are safety, tolerability, and reduced viral shedding.

Promising results were published following a phase I trial of another novel HSV-2 vaccine formulation, HerpV, manufactured by Agenus [110]. Recombinant human heat shock protein 70 (rhHsc70) non-covalently linked to synthetic HSV-2 peptides adjuvanted with the proprietary saponin-based compound QS-21 (Stimulon^®) was administered to 35 HSV-2-infected subjects. Notably, the vaccine includes 32 peptides derived from 22 viral proteins that cover all phases of the viral life cycle. The high valency of the study vaccine induced broad CD4+ and CD8+ responses as measured by cell proliferation assays and ELISPOT. Of concern, transient antibody responses to HSC70 were detectable after up to three doses of the study vaccine, though no evaluation of CMI to HSC70 was reported. The potential for eliciting an auto-immune response needs to be closely evaluated as this otherwise promising vaccine candidate proceeds through clinical testing. Based on these promising results, a phase II double-blind, placebo-controlled trial (ClinicalTrials.gov identifier NCT01687595) is now enrolling with a recruitment goal of 75 adults with recurrent HSV-2 infection. The primary endpoint of this trial is reduced viral shedding during the treatment period versus baseline period.

A phase I study of a live, replication-defective HSV-2 vaccine, HSV529, is planned by Sanofi Pasteur (ClinicalTrials.gov identifier NCT01915212). This virus, dl5-29, fails to form plaques or to give any detectable single cycle yields in normal monkey or human cells, but expresses nearly the same pattern of gene products as wild-type virus, and induces antibody titers in mice that are equivalent to those induced by single deletion mutant viruses [111]. Further improvements of immunogenicity for this virus have been reported in mice after deletion of the UL41 gene [112], although the triple deletion virus did not appear to provide better protection against HSV-2 challenge in mouse or guinea pig studies [113,114]. In addition, a comparison of this vaccine to adjuvanted gD2 demonstrated its superiority in the guinea pig model of genital HSV-2 infection [115]. A phase I study of the double-deletion virus is currently underway to (a) determine the safety of HSV529, in persons with or without HSV infection; and (b) examine the ability of HSV529 to elicit immune responses to HSV-2 including virus-specific antibodies and T cell responses. The approach has similarities to that of disabled, infectious, single-cycle (DISC) vaccines developed against HSV-1 and HSV-2 by Cantab Pharmaceuticals in the 1990s. These DISC vaccines were based on deletion of the gH gene, which was then supplied by a complementing cell line: hence, the virus could bind, attach, and initiate gene expression, but only undergo a single cycle infection. Although these vaccine showed promise in preclinical models of HSV infection [116-118], when studied in a clinical trial in patients with recurrent genital infection, no clinical or virological benefit in the amelioration of genital HSV-2 disease could be identified [119], and the vaccine did not undergo any further clinical trial testing.

Expert Commentary

The recent high-profile failure of the GlaxoSmithKline vaccine to prevent transmission of HSV infection is likely multifactorial, involving the inherent challenges posed by a highly adapted pathogen (discussed below) as well as issues related to trial design. The initial promising findings with the GlaxoSmithKline vaccine may have been the result of the stringent selection criteria for subject enrollment. Sero-discordant, monogamous couples may not accurately represent the immunological characteristics of the wider population that acquire a primary HSV infection. Others have shown that up to a third of uninfected partners in sero-discordant couples have HSV-specific cell mediated immunity despite lacking evidence of an antibody response [120]. Such “immune-seronegative” subjects that remain uninfected despite repeated exposures from an infected partner may not be immunologically similar in heretofore unknown ways; findings in such subjects may therefore have limited generalizability.

The in vitro neutralizing activity of antibodies induced by experimental vaccines may not reflect actual neutralizing ability in vivo. Notably, plaque reduction assays used in laboratory analysis to detect neutralizing antibodies employ cell lines, usually of epithelial origin, that may bear little or no resemblance to the permissive cell types infected in vivo. As mentioned, there are many permissive cell types involved in primary infection, each with potentially different glycoprotein requirements. Furthermore, plaque assays are performed at neutral pH but normal vaginal pH is significantly acidic (pH 3.5 – 4.5). Though viral glycoproteins are thought to function during viral entry via the classic pathway irrespective of pH, the endocytic pathway has been found to be pH dependent in certain cell types [41,42]. Subtle differences in conformation induced by acidic pH have been demonstrated in gB [121,122], a viral antigen included in numerous subunit vaccines. These differences may be important in escaping neutralization in some settings but not others such that despite blockade of the classic entry pathway infection can still be established by the alternative endocytic pathway. Additionally, it is unclear if “neutralizing” concentrations of HSV-specific IgG accumulate at mucosal sites following intramuscular injection of vaccine candidates.

An important challenge for vaccine development is that assays to evaluate CMI following experimental vaccine challenge are expensive, cumbersome, and not widely standardized or agreed upon. Assessment of a proliferative response to viral antigen, for example, gives no insight into functionality of the activated clones. However, while assessment of functionality (production of cytokines, granzyme, perforin, etc.) provides insight into the robustness of the antigen, the lack of validated correlates of CMI protection makes using such data to guide vaccine development difficult.

Choice of antigens for inclusion in future vaccine candidates should be informed by new immunological data. As mentioned, up to a third of seronegative partners of HSV-positive subjects are immune-seronegative with robust adaptive cell mediated immune responses to HSV antigens. In the largest study of immune-seronegative subjects published to date, the most common antigen-specific responses were to ICP6, ICP4, ICP0, UL19, and UL29 [120]. The skewing of reactivity towards immediate early proteins was in sharp contrast to the CMI responses of the sero-positive partners where late genes, including gD and gB, were more common. Another recent study evaluated the humoral responses to HSV in ten asymptomatic (defined as up to one reactivation per year) and ten symptomatic subjects (defined one to five reactivations per year) [123]. Reactivity to gD and gB did not distinguish these groups but similar to the immune-seronegative reactivity, the asymptomatic group was significantly more likely to have responses to early viral proteins (in this case UL30 and UL42, two proteins necessary for viral DNA replication. Another novel approaches to discovery of new T cell epitopes should provide insight into the components of HSV-2 vaccines that are likely to perform successfully in the clinic [63,124-126].

To date the only successful vaccine developed for a human Herpesvirus infection is the live attenuated Oka strain varicella-zoster vaccine [127]. Several differences between VZV and HSV are likely to contribute to this, including sustained viremia following primary VZV but not HSV infection [128], allowing prolonged access by neutralizing antibody prior to establishment of latency; expression of multiple latency genes by VZV but none by HSV, [128], thus providing ongoing stimulation of the adaptive arms of the immune system following primary VZV; and the relative paucity of immune evasion strategies employed by VZV, which is in sharp contrast to the multiple known immune evasion strategies of HSV [129,130]. Lastly, given the near constant shedding of HSV but not VZV by asymptomatic individuals it has been proposed that natural immunity to HSV is not as robust as that for VZV [130]. In the face of a more well-adapted pathogen, we may be attempting to exploit immune mechanisms that are inherently insufficient.

Five-Year Commentary

Development of glycoprotein subunit vaccines needs to take into account possible pH-dependent conformational epitopes. Additionally, optimization of epitope inclusion into future candidate vaccines will be necessary. Clues from clinical observations that point to the importance of early antigens in protecting against transmission between sero-discordant couples and that also distinguish asymptomatic from symptomatic HSV carriers should guide antigen selection.

Further understanding of aspects of CTL functionality will help define and validate correlates of protection. The evidence supporting the need to elicit robust CTL responses is overwhelming, but our tools to design a perfect vaccine able to do so is limited. Improvements in eliciting therapeutic CTL responses may be aided by advances in adjuvant science.

Key Issues.

• Antivirals for HSV infection have not significantly altered the epidemiology of HSV infections; therefore, vaccination is required to reduce morbidity and mortality.

• Significant differences in the natural history of Varicella-zoster virus infection compared with HSV, particularly the lack of viral protein expression during HSV latency, may inform the development of future candidate HSV vaccines. The short incubation period for HSV as well as the limited viremic stage of primary infection further complicate vaccine development; mucosal immunity may be more important than systemic IgG response.

• Protective adaptive responses in immune-seronegative and asymptomatic subjects seem to be skewed towards early HSV proteins. Future candidate vaccines should include some of these “protective” and “asymptomatic” antigens.

• Antibodies induced by experimental vaccines that are found to have neutralizing activity in vitro may have little or no effect on viral neutralization in vivo, possibly because of inadequate tissue-specific concentrations, pH-dependency of certain antigens, and the inherent inability of antibody to prevent cell-to-cell spread of progeny virus during lytic replication.

• Cell mediated immunity is essential for control of HSV infection and may in some instances develop in the absence of a humoral response. However, tools for assessing the effectiveness of CMI responses are infrequently employed in vaccine research and little standardization and consensus opinion on correlates of infection has been established.

• Further development in adjuvant technology may increase our ability to induce innate immune mechanisms, CMI, and responses targeted to key mucosal sites.

• Innovative work dissecting and characterizing local resident T cell populations in individuals with excellent disease control may lead to novel future vaccine approaches.