In this study, we sought to identify immune correlates of genital herpes severity, defined both clinically and virologically. However, no association between expression of IFN-γ, IL-2 or TNF-α by HSV-2-stimulated circulating CD4 T cells and genital HSV shedding or lesion rates was noted. We also did not observe any correlation between either the number or function of pDC in the blood and genital herpes severity. Other immune cell subsets or functions, cell populations resident at sites of HSV-2 reactivation or latency, or other host or viral factors may be more closely related to the spectrum of recurrent genital severity in the general population. To our knowledge, the only prior report of an immune correlate of genital herpes severity was that circulating HSV-2-specific CD8 T cells were found to have an inverse relationship correlate with genital herpes severity in a cross-sectional study of HIV-coinfected persons (42
). More recently, tissue resident-memory HSV-specific CD8 T cells have been convincingly demonstrated using in situ
analyses in both animal models and human genital HSV-2 specimens (16
). The ability of HSV-specific CD4 memory T cells to persist at sites of infection may differ between species, with human studies also indicating that these T cells can have prolonged residence in healed skin and could participate in modulating HSV recurrences (17
). Inclusion of CD8 T-cell and local T-cell measures in future studies is a technically challenging but compelling goal for future research on the correlation of immunity and HSV severity.
Previous studies of PBMC from HSV-infected persons focused mostly on orolabial herpes and suggested a correlation between higher IFN secretion and increased time to next HSV recurrence (11
). Similar work with PBMC from patients with symptomatic orolabial HSV infection suggested that lower levels of CD4 T-cell-secreted IFN-γ were associated with more frequent recurrences (19
). Spruance et al. found that secretion of IFN-γ and IL-2 by cultured PBMC, in a format emphasizing CD4 T cells, tended to be lower in symptomatic than in asymptomatic HSV-1-infected persons (52
). Chentoufi et al. observed an association between fine antigenic specificity of circulating HSV-1-specific CD4 T cells and participant-reported orolabial HSV severity (8
). In contrast, in the setting of genital herpes we did not observe a correlation between circulating CD4 T-cell responses to HSV-2, as measured by cytokine-based enumeration, and HSV severity. This agrees with a previous investigation of HSV-2/HIV-1-coinfected persons in whom anogenital HSV-2 clinical severity did not correlate with HSV-2-specific CD4 T-cell responses (42
An important implication of our findings is that immune manipulations that simply increase the total number of Th1-like circulating CD4 T cells without also inducing other effector cells or cells at sites of infection are unlikely to have a significant impact on genital herpes severity or on the intensity of HSV shedding. CD4 T cells alone do not seem to be biologically active in the setting of preventative vaccination for HSV-2. In a recent phase III study of a subunit HSV-2 vaccine, CD4 T-cell responses analyzed for the cytokines investigated in this study and also for CD40L did not correlate with vaccine efficacy in a case-control analysis (4
). Interestingly, the vaccine may have protected against HSV-1 acquisition, and the subunit antigen used was very sequence-similar between HSV-1 and HSV-2. At the level of individual CD4 T-cell epitopes, a quite clear distinction can be drawn between epitopes recognized by HSV-2 only or both HSV-2 and HSV-1 (30
). Because of this difference in protection afforded by an antibody and CD4-targeted vaccine, it is possible that studies of acquired CD4 responses and HSV-1 severity could reveal associations not noted in the current work. Vaccine formats that can induce strong antigen-specific CD8 T-cell responses to HSV-2 have already been tested in humans (62
). It may be important to combine these types of approaches with methods to optimize the homing of these effector cells to sites of HSV exposure and recurrence.
Our data on viral coinfections revealed two surprising findings. First, HSV-1 coinfection was not associated with an increase in CD4 T-cell responses to HSV-2. As our recent comprehensive discovery effort for HSV-1 T-cell epitopes showed that about 50% are sequence-identical with HSV-2 (25
), we expected that total HSV-2 CD4 responses might be higher in coinfected persons if the intensity of local antigenic exposures led to increased systemic T-cell levels. We actually observed a trend toward lower HSV-2-specific CD4 T-cell levels in persons coinfected with HSV-1 and HSV-2 that did not reach statistical significance. HSV-2 infection, typically acquired later in life than HSV-1, does not seem to add an increment to cross-reactive T-cell responses. Our data are in agreement with the findings of Schmid et al. (47
), who used a PBMC enzyme-linked immunosorbent spot assay (ELISPOT) and observed the same inverse trend of lower responses to HSV-2 in HSV-1-coinfected persons.
The second surprising finding was the direct correlation between the magnitudes of the HSV-2 and CMV-specific CD4 T-cell immune responses among CMV- and HSV-2-coinfected persons. We did confirm the observations of others that CMV infection can lead to large accumulations of CMV-specific CD4 T cells (54
) and found that CMV-specific CD4 T-cell responses were usually but not invariably larger than the corresponding HSV-2-specific responses. The magnitude of integrated CD4 responses to HSV-2 detected in the present study is also generally similar to that noted by two independent labs (2
). We initially hypothesized that the magnitude of the HSV-2-specific CD4 T-cell response would be lower in persons coinfected with CMV and HSV-2 than in those not infected with CMV, but in fact our findings were the opposite. The possible immune suppressive effects of CMV (26
), including its association, if any, with the severity of HSV infections, require additional study. These effects have been most prominently investigated in the setting of human aging (28
), and our study population contained few people in their seventh decade of life or older. The predilection for HSV-2-specific CD4 responses to be more polyfunctional than CMV-specific CD4 T-cell responses has been previously noted, albeit without within-subject pairwise analyses (20
). Based on our finding that HSV-2-specific CD4 T cells are mostly polyfunctional, we predict that relatively few will be CD57+
effector-memory cells expressing lytic molecules as are seen for CMV (7
), but further research into the phenotype of HSV-2-specific CD4 T cells is required.
Of note, despite the collection of PBMC for CD4 T-cell tests at various time intervals from the collection of genital herpes severity data, our results correlating CD4 responses and genital herpes severity remained robust and did not change after we adjusted for this time interval. Longitudinal CD4 T-cell studies in HSV-2-infected subjects showed that the variable acute changes associated with a clinical recurrence were followed by relative stability. The fact that our specimens for CD4 T-cell analyses were collected during these stable periods argues against a strong influence by time gaps between the collection blood for CD4 T-cell measures and the collection of virologic severity data. Higher-temporal-resolution studies surrounding genital herpes recurrences will be required to determine if recurrences lead to predictable acute fluctuations in circulating CD4 T cells.
pDC were defined by Siegal et al. (49
) as the natural interferon-producing cells in PBMC responding to HSV (15
). Several genetic defects in TLRs and their signaling pathways are associated with both severe primary HSV-1 infection and deficient responses to HSV and synthetic TLR agonists (41
). The precise viral and host molecules involved in HSV detection by pDC are still controversial. Our data are consistent with the sensing of HSV-2 by TLR9, although it is also possible that an underlying variable in the pDC activation threshold could influence reactivity to both HSV-2 and synthetic TLR9 agonists. Animal data indicate that TLR9 can sense HSV and modulate HSV pathogenesis (35
). Thus far, the one study of human TLR9 genetic variants and genital HSV severity did not detect any relationship (5
). Persons with defects in molecules downstream of TLR7, TLR8, and TLR9 do not exhibit a phenotype of severe HSV infections (39
). A recent study (64
) correlated TLR3 genotypes and circulating NK-cell responses to TLR3 agonists with HSV-1 severity, indicating that these cells should be included in future work. While we did detect a considerable dynamic range in both circulating pDC number and reactivity to HSV-2, we did not observe any correlation of these pDC measures with either genital herpes lesion severity or viral shedding. This does not rule out a local role for pDC in recurrent genital herpes or the possibility that these mobile cells could shift in or out of the blood or tissues during or after recurrences. Future studies, requiring timely processing of fresh blood on multiple days, will be required to determine if pDC number or function shift is dynamically related to symptomatic recurrent HSV-2.
We chose to study pDC reactivity to HSV-2 for 5 h, based on literature reports (1
) and the fact that pDC are labile and start to die upon prolonged culture. It is possible that use of other durations of incubation would lead to contrasting results. pDC can display dynamic differences in maturation status (48
) and studies of maturation markers would also be of interest. As the molecular and structural details of pDC triggering by HSV are not completely understood, it is possible that HSV-1 and HSV-2 differ in this regard and that studies of HSV-1 severity and pDC biology would reveal associations that we did not observe in this study for HSV-2. The causes and consequences of the relatively stable interindividual differences in pDC number and function between persons remain undefined. pDC do express chemokine receptors suitable for homing to sites of inflammation and infiltrate the skin in HSV-2 lesions (14
) and autoimmune conditions (56
). It is not known if pDC are retained at mucosal sites after HSV clearance.
A strength of our study is that it included prospective, objective ascertainment of disease severity as defined by both clinical and virologic outcomes that provide precise data with which to explore the relationship between cellular immunity and the natural history of genital herpes infection. In addition, in our investigation we enumerated antigen-specific CD4 T cells and studied three cytokines. Prior studies used self-reported recurrence frequency for severity, did not enumerate the number of antigen-reactive cells, and focused on IFN-γ alone. Our study focused on anogenital rather than orolabial herpes. It is possible that the immune parameters controlling HSV-1 and HSV-2 infection differ. Because most recurrent orolabial herpes is due to HSV-1 infection and most recurrent genital herpes is caused by HSV-2 (10
), the conclusions of previous studies focusing on orolabial herpes and HSV-1 (8
) may not apply to recurrent genital herpes.
Perhaps the most important limitation of our study is that it was limited to PBMC. Circulating cells may not reflect mucosal immunity, in which a complex milieu including compartmentalized virus-reactive T cells may influence HSV disease course (68
). Antigen-specific CD4 T cells can persist at the site of healed herpes (68
) or be recruited from the periphery during recurrences (17
). Little is known concerning the comparative phenotype or specificity of tissue-resident memory versus circulating HSV-specific CD4 T cells. HSV-2-specific CD4 T cells have additional effector functions, including cytotoxicity and CD40L expression (18
), that were not measured in this study. The limited sample size and sampling period for our HSV-shedding studies could have also led to misclassification of true shedding rates, such that larger cohorts or longer severity measurement periods could have allowed biological signals to emerge. The HSV antigen we used was previously shown to include nonstructural as well as structural viral proteins (25
) but it is possible that PBMC could contain T lymphocytes nonresponsive to the HSV-2 preparations used in this report. Another possible limitation of our study was the use of ICC rather than of ELISA to measure cytokine reactivity to HSV-2. ICC has the advantage of positively identifying and enumerating the number of reactive cells but the potential disadvantage of not capturing responses by cell populations not specifically gated for during analysis. Cytokine ELISAs can reflect the net, integrated response of all cells but do not provide information concerning the number or identity of reactive cells. The inclusion of brefeldin A during the stimulation for ICC assays prevents ELISA from being performed on the same set of stimulated cell specimens, because this compound inhibits protein egress from cells. Advances in multiplex cytokine ELISA methods suggest that such assays to identify biomarkers of HSV severity may be rational in the future. Multiplex reverse transcription-PCR (RT-PCR) to measure RNA levels is another attractive assay to measure multiple immune parameters in innate or acquired cell populations responding to microbial antigens.
In summary, we have completed a detailed study of the number and function of circulating HSV-2-specific CD4 T cells and of pDC in immunocompetent persons with recurrent genital herpes infection. No correlations were noted between several parameters concerning these cell types and either objectively measured genital HSV shedding or prospectively noted genital lesions. Our data are consistent with neutral influences of HSV-1 on the CD4 T-cell immune response to HSV-2 and with pDC innate sensing of HSV-2 via TLR9. Our data do not support the simple use of PBMC Th1 CD4 T-cell parameters as either a goalpost for vaccine-induced immune responses or as the sole surrogate immune measure of the efficacy of HSV vaccines. Future research will optimally include detailed analyses of defined subsets of these circulating cell types, of CD8 T cells and NK cells, and also of their trafficking and localization to sites of infection.