Noncognate Signals Drive Enhanced Effector CD8⁺ T Cell Responses through an IFNAR1-Dependent Pathway after Infection with the Prototypic Vaccine, 0ΔNLS, against Herpes Simplex Virus 1

ABSTRACT

Previous studies by our group identified a highly efficacious vaccine 0ΔNLS (deficient in the nuclear localization signal of infected cell protein 0) against herpes simplex virus 1 (HSV-1) in an experimental ocular mouse model. However, details regarding fundamental differences in the initial innate and adaptive host immune response were not explored. Here, we present a side-by-side analysis of the primary infection characterizing differences of the host immune response in mice infected with 0ΔNLS versus the parental, GFP105. The results show that local viral infection and replication are controlled more efficiently in mice exposed to 0ΔNLS versus GFP105 but that the clearance of infectious virus is equivalent when the two groups are compared. Moreover, the 0ΔNLS-infected mice displayed enhanced effector CD8⁺ but not CD4⁺ T cell responses from the draining lymph nodes at day 7 postinfection measured by gamma interferon (IFN-γ) and tumor necrosis factor alpha production along with changes in cell metabolism. The increased effector function of CD8⁺ T cells from 0ΔNLS-infected mice was not driven by changes in antigen presentation but lost in the absence of a functional type I IFN pathway. These results are further supported by enhanced local expression of type I IFN and IFN-inducible genes along with increased IL-12 production by CD8α⁺ dendritic cells in the draining lymph nodes of 0ΔNLS-infected mice compared to the GFP105-infected animals. It was also noted the recall to HSV-1 antigen by CD8⁺ T cells was elevated in mice infected with HSV-1 0ΔNLS compared to GFP105. Collectively, the results underscore the favorable qualities of HSV-1 0ΔNLS as a candidate vaccine against HSV-1 infection.

IMPORTANCE Cytotoxic T lymphocytes (CTLs) play a critical role in the clearance for many viral pathogens including herpes simplex virus 1 (HSV-1). Here, we compared the cellular innate and adaptive immune response in mice infected with an attenuated HSV-1 (0ΔNLS) found to be a highly successful experimental prophylactic vaccine to parental HSV-1 virus. We found that CD8⁺ T cell effector function is elevated in 0ΔNLS-infected mice through noncognate signals, including interleukin-12 and type I interferon pathways along with changes in CD8⁺ T cell metabolism, whereas other factors, including cell proliferation, costimulatory molecule expression, and antigen presentation, were dispensable. Thus, an increase in CTL activity established by exposure to HSV-1 0ΔNLS in comparison to parental HSV-1 likely contributes to the efficacy of the vaccine and underscores the nature of the attenuated virus as a vaccine candidate for HSV-1 infection.

INTRODUCTION

Herpes simplex virus 1 (HSV-1) is a successful human pathogen that can result in asymptomatic infection to more severe clinical manifestations such as herpes stromal keratitis or encephalitis (1, 2). Given that humans are the natural host, there have been many attempts undertaken experimentally to control HSV-1 infection/reactivation, including the development of subunit, recombinant DNA, replication incompetent virus, and live-attenuated virus vaccines. However, to date, no approved vaccine for human application exists despite the numerous clinical trials conducted in the last 3 decades. One of the major challenges with vaccine development is understanding the complexity of the interactions between the virus and the host innate and adaptive immune responses. Regarding the design of vaccines against pathogens that infect the eye, an efficacious vaccine would need to generate a robust T cell and/or antibody response and simultaneously, protect the visual axis from collateral damage driven, in large part, by the host immune response (3, 4).

Cutaneous inoculation with HSV-1 is a model that has be used to characterize the innate and adaptive immune response during acute infection (5–8). It has been shown that viral replication is not detected in the draining lymph nodes but rather, is limited to the skin or epithelial cells (5, 6). The virus is either cleared at the site of entry or gains access to peripheral sensory nerve endings and travels by retrograde transport to the sensory ganglia where latency is established (5). HSV-1-specific T cells are rapidly activated and armed within the draining lymph nodes by antigen-presenting cells bearing viral antigens after cutaneous virus inoculation (8). This process is spatiotemporally regulated (9) and depends on numerous variables, such as the amount and the nature of antigen, its presentation efficiency, and the types of dendritic cells (DC) that interact with T cells (10–12). While these studies clarify the kinetics of T cell activation, highlight the importance of CD4⁺ and CD8⁺ T cell responses in viral clearance, and the contribution of DC in the regulation of these processes, the importance of other factors that influence T cell activation and expansion during acute HSV-1 infection remain to be defined.

Our previous studies reported that a prophylactic live-attenuated prototype HSV-1 vaccine, termed 0ΔNLS, showed appreciable efficacy following ocular challenge of vaccinated mice without unwarranted effects. Specifically, mice inoculated with 0ΔNLS displayed no corneal pathology and retained corneal function when infected with a highly neurovirulent HSV-1 strain (13). While the correlate of protection was identified to be antibody-driven (14), the vaccine has also been found to elicit a protective CD8⁺ T cell response in the absence of neutralizing antibodies (15). However, the caveat of these studies is that either the comparison of 0ΔNLS-driven protection was compared to another subunit vaccine (13) or the 0ΔNLS vaccine was used in a transgenic mouse model which has limited physiological significance (15).

In the present study, we employed a cutaneous mouse footpad infection model that follows our vaccine immunization route (13–15) in a side-by-side comparison of the primary T cell responses after inoculation with either parental or 0ΔNLS HSV-1. The HSV-1 0ΔNLS mutant contains an insertion of the green fluorescent protein (GFP) coding sequence between codons 104 and 105 of the ICP0 gene, as well as a deletion of amino acids 501 to 508 of ICP0 removing the nuclear localization sequence of ICP0, RPRKRR, thus generating a virus that is highly sensitive to type I interferon (IFN) (13). Previous data showed that vaccination with 0ΔNLS changed neither the overall T cell nor antigen-specific T cell numbers (16). However, the functional characterization of T cell responses was not assessed and remained unanswered. In this study, we found enhanced IFN-γ production by antigen-specific CD8⁺ but not CD4⁺ T cells in 0ΔNLS-infected animals’ draining lymph nodes compared to those infected with parental HSV-1. The enhanced primary CD8⁺ T cell response was not due to a change in the kinetics of antigen-specific CD8⁺ T cell clonal expansion or efficiency of antigen presentation by DC. Rather, the effect correlates with an enhanced type I IFN response, along with an increase in the production of interleukin-12 (IL-12), resulting in an elevated number of polyfunctional effector cells as well as changes in cell metabolism. Thus, this study underscores the critical role cytokines (third signal) play in driving efficient CD8⁺ T cell activation/function within the confines of primary HSV-1 infection. Moreover, the data also demonstrate that 0ΔNLS, while attenuated, elicits a powerful stimulus that drives a sustainable CD8⁺ T cell response superior to the parental virus under an experimental setting.

RESULTS

HSV-1 0ΔNLS is cleared similar to parental (GFP105) virus.

The HSV-1 0ΔNLS mutant virus was constructed by homologous recombination of the parent strain HSV-1 KOS and a plasmid containing an ICP0 gene with a deletion of the nuclear localization signal (NLS) between codons 500 and 507 and an in-frame insertion of green fluorescent protein (GFP) between codons 104 and 105. Next-generation sequencing was performed on the HSV-1 0ΔNLS virus. In the absence of the parental lab strain, the sequence was compared to the GenBank KOS genome accession number KT899744.1. Aside from the ICP0 gene, the sequencing revealed only a small number of variations none of which would be expected to account for the mutant phenotype. The variations include two five-base deletions and 7 single-base deletions, none of which occur in protein coding sequences. As well, 16 substitutions were identified, 9 of which occur within the coding regions of 5 genes (UL27, UL36, UL39, US6, and RS1_1). Of these 9 substitutions, 4 result in amino acid changes, 2 of which are conservative (RS1_1 A2V and UL27 R515H) and 2 of which are mutations to proline residues which occur in proline-rich regions (UL36 A1894P and UL39 L393P). Based on these results and previously published data demonstrating the restoration of resistance to type I IFN when an ICP0 complementing cell line was infected with HSV-1 0ΔNLS (17), we were confident the phenotype observed by the HSV-1 0ΔNLS mutant was due to the targeted deletion within the ICP0 gene. Therefore, we sought to compare our prototype HSV-1 vaccine, the live attenuated HSV-1 0ΔNLS virus, to the parental virus (termed GFP105) measuring infectious virus, lytic gene expression, and virus genome copy number over a 7-day period in the footpad and/or popliteal lymph nodes (PLN). Previous studies utilizing a mouse footpad infection model to study effector T cell immune responses reported that HSV-1 is detectable exclusively in the footpad but virtually absent in the draining lymph nodes (5, 6). Using this same approach, we infected mice with GFP105 or HSV-1 0ΔNLS and found equivalent infectious virus levels recovered from the footpad following HSV-1 0ΔNLS and GFP105 inoculation at days 1 and 3 postinfection (p.i.) (Fig. 1A). However, by day 5 p.i. through day 7 p.i., infectious virus levels were significantly reduced in the footpad in the 0ΔNLS-inoculated group compared to GFP105 samples (Fig. 1A). By day 9 p.i., there was no detectable infectious virus in the footpad of either GFP105- or 0ΔNLS-infected mice. By comparison, infectious virus was not detected in the PLN at any time point. However, the HSV-1 lytic gene transcript, thymidine kinase (TK), was detected in the PLN in both groups at day 1 p.i. with significantly higher expression in the tissue of GFP105-infected animals. However, at later time points, TK was no longer detectable in the organized lymphoid tissue (Fig. 1B). The HSV-1 DNA copy number was not significantly different comparing PLN content from GFP105- versus 0ΔNLS-infected mice at any time point evaluated with the eventual clearance of virus DNA by day 7 p.i. (Fig. 1C). Taken together, the results show the parental virus replicates more efficiently than the 0ΔNLS mutant, but the viruses are ultimately cleared from the original site of infection at an equivalent time point.

FIG 1.

Mice infected with HSV-1 0ΔNLS control virus more efficiently compared to mice infected with GFP105. Male and female C57BL/6 mice were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. At the indicated time postinfection (p.i.), viral titers were determined in the footpad by plaque assay (A), and lytic gene expression of thymidine kinase (TK) in the popliteal lymph nodes was determined by RT-PCR (B). (C) The HSV-1 DNA copy number was determined at the indicated times p.i. The experiment was repeated twice, encompassing a total 6 to 9 animals for each time point, except in panel C, which was a repeat experiment with two animals/time point/experiment. The values are presented either as log PFU/homogenized footpad (A), the mean value for relative expression for TK at the indicated time point (B), or the HSV-1 copy number/μg total DNA in the PLN at the indicated time point (C). *, P < 0.05 (comparing the GFP105- to 0ΔNLS-infected groups using Student t test with Bonferroni correction).

HSV-1 0ΔNLS enhances local type I IFN response compared to mice infected with HSV-1 GFP105.

Type I IFNs control acute HSV-1 infection and modulate the innate and adaptive immune responses during infection (18). Since infected cell protein 0 (ICP0) is a potent antagonist of type I IFNs (19, 20), and HSV-1 0ΔNLS is mutated in the nuclear localization signal of ICP0, making it highly susceptible to type I IFNs (13), we infected mice with GFP105 or 0ΔNLS and analyzed changes in type I IFN and type I IFN-related gene transcripts in the footpad in comparison to mock-infected animals after 12 h p.i. Infected mice displayed elevated levels of all genes analyzed compared to mock-infected mice. Comparable analysis showed higher expression of most of the type I IFN or type I IFN-inducible genes (21 of 24) except for Ifna12, Ifna13, and Socs1 in mice infected with 0ΔNLS in comparison to GFP105 (Fig. 2A). Notably, the expression of Socs1, which is recognized as a negative regulator of type I IFN responses (21), was elevated in mice infected with GFP105 compared to mice infected with 0ΔNLS (Fig. 2B). In contrast, interferon regulatory factor 1 (Irf1), Irf9, and interferon stimulatory gene 15 (ISG15) were all significantly elevated in the footpad of 0ΔNLS-infected mice compared to mice infected with the parental virus, GFP105 (Fig. 2B). These results are consistent with an elevation of IFN-α7 and IFN-β, which were also significantly different compared to 0ΔNLS- to GFP105-infected footpads (Fig. 2B). Other type I IFNs analyzed, including IFN-α1, -α2, -α9, and -α12 expression, were increased in the 0ΔNLS-infected footpad compared to the GFP105-infected tissue but the differences did not reach significance. With the exception of IFN-α7, IFN-β, Irf1, Irf9, ISG15, Janus kinase 1 (JAK1), protein tyrosine phosphatase, nonreceptor 11 (Ptpn11), and signal transducer and activator of transcription 2 (STAT2), all other IFN-inducible genes and type I IFNs were elevated in 0ΔNLS- and GFP105-infected footpads compared to the mock (PBS)-infected group. Collectively, these results show a robust type I IFN response in 0ΔNLS-infected mouse footpads compared to that found in the footpad after GFP105 infection within the first 12 h p.i.

FIG 2.

Infection with HSV-1 0ΔNLS induces increased expression of type I IFNs and IFN-inducible genes in the footpad compared to HSV-1 parental (GFP105). Male and female C57BL/6 mice were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. Control mice were mock infected (PBS only). (A) At 12 h p.i., RNA was isolated from the footpad and processed for detection of type I IFNs and IFN-inducible genes by real-time PCR. A heat map with color-coded gene expression was generated. Host gene expression was normalized relative to GAPDH using tissue from mock-infected animals. (B) Relative expression of select genes from panel A. The data are displayed as the means ± the SEM (n = 3/group). **, P < 0.01; *, P < 0.05 [comparing the indicated group(s) to the PBS control]; ΔΔ, P < 0.01 (comparing 0ΔNLS to GFP105). (C) Mice were infected as described above. At 24 h p.i., footpads were harvested and processed for analyte expression by multiplex suspension array analysis. Only cytokines/chemokines that were significantly different between groups are shown. The data are displayed as means ± the SEM (n = 5/group). **, P < 0.01; *, P < 0.05 (comparing the indicated groups). (D) Mice were infected as described above. At 24 h p.i., footpads were harvested and processed for flow cytometry targeting myeloid cells (CD45⁺ CD11b⁺) cells. Granulocytes were defined as Ly6G⁺ Ly6C^low, and monocytes were defined as Ly6G-Ly6C^high. **ΔΔ, P < 0.01 (comparing the indicated groups). Analysis for significance was determined by one-way ANOVA and Tukey’s post hoc t test.

Type I IFN signals through a number of intracellular pathways (22) drive the expression of cytokines and chemokines. Since there were noted differences in the transcript expression of signaling proteins, including Jak1, Stat1, and tyrosine kinase 2 (Tyk2), comparing GFP105- to 0ΔNLS-infected tissues (Fig. 2A and B), we assessed 31 different cytokine and chemokine protein levels from the footpads of infected mice at 24 h p.i. The results show differences in the expression of seven of the soluble factors measured. Specifically, CCL5, CXCL1, G-CSF, and TNF-α were all significantly elevated in the footpad following 0ΔNLS infection compared to the mock (PBS)- or GFP105-infected groups (Fig. 2C). CCL2, CXCL9, and CXCL10 were elevated in the 0ΔNLS- and GFP105-infected footpads compared to the PBS group (Fig. 2C). As a consequence of changes of predominantly chemokine expression, we next investigated the infiltrate of leukocytes into the inflamed tissue. The results show an increase in the influx of myeloid cells including granulocytes (CD45⁺ CD11b⁺ Ly6G⁺ Ly6C^lo) and monocyte/macrophage groups (CD45⁺ CD11b⁺ Ly6G-Ly6C^high) at 24 h p.i. in the infected groups compared to the PBS control (Fig. 2D). Specifically, the granulocyte populations were significantly elevated in GFP105- and 0ΔNLS-infected mouse footpads in which a more pronounced difference was observed in the 0ΔNLS-infected group that was significantly elevated above that of the GFP105-infected group (Fig. 2D). A similar pattern was reflected in the monocyte/macrophage population as well. However, the number of myeloid cell types declined in both infected groups over the next 6 days, with no significant differences between them (data not shown). NK cells were also investigated in which it was found that there was a temporal increase in the number of NK cells residing in the footpad of 0ΔNLS-infected mice compared to the GFP105-infected group at day 3 p.i. (Fig. 3). However, other time points were not found to be significantly different comparing the footpad population of the two groups of infected mice (Fig. 3). Likewise, CD3⁺ T cell numbers isolated from the footpad of HSV-1-infected were not found to be different at any time point investigated (Fig. 3). Therefore, the 0ΔNLS infection elicits an elevated type I IFN response in the local environment and promotes a more robust innate immune response in the footpad reflected by higher expression of select chemokines and myeloid and NK cell infiltration compared to the GFP105-infected mice.

FIG 3.

Footpad infection with HSV-1 0ΔNLS results in a temporal increase in NK but not CD3⁺ T cells compared to HSV-1 GFP105-infected footpad. Male and female C57BL/6 mice (n = 6/group) were injected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL if PBS. At day 1, 3, 5, and 7 p.i., skin explants were harvested, minced on small pieces, and digested with collagenase I (1 mg/mL) for 30 min at 37°C to generate single-cell suspensions. Next, the cells were washed with PBS supplemented with 2% FBS and stained Zombie Aqua discriminating live and dead cells, followed by staining with an antibody cocktail containing anti-mouse CD45 conjugated with PerCP-Cy5.5, anti-mouse CD3 conjugated with PE-Cy7, and anti-mouse NK1.1 conjugated with PE. Finally, the cells were incubated 30 min on ice, washed with PBS supplemented with 2% FBS, and fixed with 2% PFA until flow cytometry acquisition. The samples were acquired on an Aurora spectral flow cytometer (Cytek Biosciences), and the data were analyzed using FlowJo software. (A) Gating strategy for identifying NK1.1⁺ and CD3⁺ T cell populations. (B) NK1.1 cell number at the indicated time points p.i. in the footpad. (C) Numbers of CD3⁺ T cells in the footpad at days 1, 3, 5, and 7 p.i. The values are presented as the means ± the SEM. *, P < 0.05 (comparing the indicated groups as determined by Student t test).

HSV-1 0ΔNLS infection induces a robust type I IFN response in the draining lymph nodes and accumulation of CD8α DC producing IL-12.

Dendritic cells (DC) are major antigen-presenting cells shaping T cell responses (23). Exposure of DC to type I IFN has previously been reported to modulate their phenotype as well as maturation status (24). Here, we hypothesized that initial exposure of DC in the footpad to type I IFN levels prompted by GFP105 versus 0ΔNLS infection would significantly impact DC number, phenotype, and/or maturation, as well as contribute to the magnitude of CD8⁺ T cell clonal expansion in the PLN. Moreover, since there was an increase in TK lytic gene expression in the PLN of the GFP105-infected mice at 24 h p.i. (Fig. 1B), initial assessment of IFN-α and IFN-β transcript expression was conducted in the PLN. IFN-α and IFN-β transcript levels were significantly higher in the PLN from 0ΔNLS-infected mice compared to GFP105- or mock-infected animals (Fig. 4A and B). Next, DC subsets from the PLN of infected mice were examined by flow cytometry in which it was found that 0ΔNLS-infected mice possessed significantly more CD205⁺ CD8α⁺ DC but not plasmacytoid DC, dermal CD205⁺ CD103⁺ DC, or CD205⁺ Langerhans cells (LC) compared to those infected with GFP105 (Fig. 4C). However, there were no differences in the frequency of CD205⁺ CD8α DC expressing major histocompatibility complex (MHC) class I, MHC class II, CD80, or CD86, which indicates that the number but not the activation status of these cells is altered by 0ΔNLS infection (Fig. 5).

FIG 4.

Type I IFN expression and CD8α dendritic cell IL-12 production is elevated in the PLN of HSV-1 0ΔNLS-infected mice. C57BL/6 male and female mice were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. The PLN were removed at 24 h p.i. and processed for the detection of IFN-α (A) and IFN-β (B) by real-time RT-PCR. Bars represent the means ± the SEM (n = 5/group). **ΔΔ, P < 0.01; *, P < 0.05 (comparing the indicated groups as determined by one-way ANOVA and Tukey’s post hoc t test). (C) The PLN were removed at 48 h p.i., and DC subsets were determined by flow cytometry within the CD45⁺ CD11c⁺ gate with antibody cocktails targeting the following DC subsets: CD11b⁺ (CD45⁺ CD11c⁺ CD11b⁺), CD8α (CD45⁺ CD11c⁺ CD205⁺ CD8α⁺ CD11b⁻ CD103^– CD326⁻), plasmacytoid DC (PDCA-1) (CD45⁺ CD11c⁺ CD11b⁻ CD317⁺), dermal DC CD103⁺ (CD11c⁺ CD103⁺ CD205⁺ CD8α^– CD11b⁻ CD326⁻), and Langerhans cells (EpCAM) (CD45⁺ CD11c⁺ CD8α^– CD11b⁻ CD103^– CD326⁺). (D) Example of contour plots showing intracellular staining in CD8α⁺ DC from mice infected with either HSV-1 GFP105 or 0ΔNLS. (E) Cell frequency of CD8α⁺ CD11c⁺ producing IL-12p40 after stimulation in vitro with 1 μM CpG ODN 1826. ΔΔ, P < 0.01; **, P < 0.01; *, P < 0.05 (comparing the indicated groups, as determined by ANOVA and Sidak multiple-comparison test for results shown in panels A and B and the Student t test for panels C and E).

FIG 5.

HSV-1 GFP105 and 0ΔNLS Infection show similar levels of select activation markers of CD8α⁺ DC. C57BL/6 (n = 6 to 7/group) mice were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 at 10⁵ PFU in 10 μL of PBS. At day 2 p.i., PLN were harvested and processed for flow cytometry for the expression of CD80, CD86, MHC-I, and MHC-II on the surfaces of CD8α⁺ DC. Panel A includes representative contour plots. Panel B includes the cell frequencies of CD80, CD86, MHC-I, and MHC-II on CD8α⁺ DC. The data are summarized, where the bars represent the means ± the SEM.

Among DC subsets, CD8α⁺ DC are efficient in priming CD8⁺ T cells during virus infection (25, 26). CD8α⁺ DC are also believed to be a rich source of IL-12 during inflammatory conditions (27). Therefore, we initially sought to determine whether this DC subset obtained from the PLN of 0ΔNLS- versus GFP105-infected mice displayed changes in IL-12 production comparing cells from the two HSV-1-infected groups. Ex vivo stimulated CD8α DC from 0ΔNLS-infected mice expressed significantly more IL-12 compared to cells from mice infected with GFP105 (Fig. 4D and E). Since IL-12 expression favors the development of a T cell response that promotes clearance of viral pathogens, including HSV-1 (10, 28), we next investigated whether initial infection of mice with GFP105 or 0ΔNLS altered HSV-1 glycoprotein B (gB)-specific CD8⁺ T cell clonal expansion in vivo. CFSE (carboxyfluorescein succinimidyl ester)-labeled HSV-1 gB-specific CD8⁺ T cells were adoptively transferred into mice that were infected with GFP105 or 0ΔNLS. At 5 days p.i., CD8⁺ T cell proliferation between mice infected with GFP105 versus 0ΔNLS was evaluated by flow cytometry and found not to be different (Fig. 6A and B). In a similar fashion, ex vivo experiments using an equal ratio of sorted DC from PLN of infected mice and HSV-1 gB-specific CD8⁺ T cells cocultured ex vivo for 3 days resulted in no detectable differences in antigen-driven CD8⁺ T cell clonal expansion (Fig. 6C and D). To exclude any intrinsic effect in T cells by infection, the T cells from the PLN of infected mice were labeled with CFSE and stimulated for 3 days with anti-CD3 along with anti-CD28 to measure T cell receptor (TCR)-driven proliferation. The results show no difference in CD4⁺ and CD8⁺ T cell expansion measured by CFSE dilution comparing T cells from GFP105- versus 0ΔNLS-infected mice suggesting that gross TCR signaling is not modified in T cells taken from mice infected with 0ΔNLS versus GFP105 (Fig. 6E and F). Collectively, these results show that although the CD8α DC populations from 0ΔNLS-infected mice express more IL-12 in comparison to CD8α DC populations from GFP105-infected mice, this difference is not reflected in clonal expansion of HSV-1 gB-specific CD8⁺ T cells.

FIG 6.

HSV-1 GFP105 or 0ΔNLS infection does not alter antigen presentation to HSV-1 gB-specific T cell receptor transgenic CD8⁺ T cells. Male and female C57BL/6 mice (n = 6 to 7/group) were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 at 10⁵ PFU in 10 μL of PBS, followed by adoptive transfer of 5 × 10⁶ HSV-1 gB-specific transgenic T cell receptor (TCR) CD8⁺ T cells labeled with CFSE. (A) At day 5 p.i., popliteal lymph nodes were harvested and CFSE dilution was determined within CD8⁺ T cells by flow cytometry. (B) The frequency of each single peak is shown for CSFE-labeled HSV-1 gB-specific transgenic TCR CD8⁺ T cells. (C and D) C57BL/6 mice were infected as described above. At day 2 p.i., PLN were harvested, and CD11c-positive cells were enriched by positive selection with magnetically labeled cells and columns under an intense magnetic field. The DC were cocultured with enriched HSV-1 gB-specific transgenic TCR CD8⁺ T cells labeled with CFSE and stimulated with gB peptide (10 μg/mL). (C) Cell proliferation measured by CFSE dilution was determined 3 days after stimulation by flow cytometry. (D) The frequency of each single peak is shown for CSFE-labeled HSV-1 gB-specific transgenic TCR CD8⁺ T cells. (E and F) C57BL/6 mice were infected as described above. At day 7 p.i., the PLN were harvested, and 5 × 10⁶ cells were labeled with CFSE. Labeled cells were stimulated with different concentration of soluble anti-mouse CD3e antibody and a constant concentration of anti-mouse CD28 (1 μg/mL) for 72 h. CFSE dilution was determined with flow cytometry for CD4⁺ (E) and CD8⁺ (F) T cells.

HSV-1 0ΔNLS infection of mice results in a local enhanced IFN-γ and TNF-α response by antigen-specific effector CD8⁺ T cells.

Type I IFN has been shown to prevent CD8⁺ T cell exhaustion resulting in a more robust effector response (29), whereas IL-12 is involved in the stimulation and maintenance of the Th1 immune response (30). The results above suggested there was no discernible change in HSV-1-specific T cell in vivo clonal proliferation during the early phase of T cell stimulation. However, the antigen-driven effector function of T cells was not evaluated in that assay. Therefore, antigen-stimulated effector CD8⁺ T cells isolated from the PLN of HSV-1 0ΔNLS- and GFP105-infected mice 7 days p.i. were assessed for IFN-γ production after stimulation ex vivo with the immunodominant Kb-restricted HSV-1 peptide (SSIEFARL, gB_498–505). The results show significantly more IFN-γ levels produced by CD8⁺ T cells isolated from 0ΔNLS-infected mice as measured by ELISPOT assay (Fig. 7A) and ELISA (Fig. 7B). The difference in IFN-γ production was not due to changes in the number of T cells as there were no significant differences in the absolute number of HSV-1 gB-specific CD8⁺ T cells comparing the two groups of infected mice (Fig. 7C and D). To further characterize antigen-specific CD8⁺ T cell responses, intracellular staining for TNF-α and IFN-γ was conducted on CD8⁺ T cells isolated from the PLN of HSV-1 0ΔNLS- and GFP105-infected mice. Consistent with the production of IFN-γ by HSV-1 gB-specific effector CD8⁺ T cells, mice infected with HSV-1 0ΔNLS displayed a higher frequency of TNF-α- and IFN-γ-expressing cells compared to antigen-specific CD8⁺ T cells from GFP105-infected mice (Fig. 7E). By comparison, there were no significant differences in CD4⁺ T cell production of IFN-γ after stimulation with the immunodominant HSV-1 class II-specific glycoprotein D peptide (Fig. 7A and B). While the effects on CD8⁺ T cell cytokine expression correlate with results assessing IL-12 expression and type I IFN-associated pathways, other events elicited by HSV-1 may also contribute to these changes including but not limited to regulatory T cells or expression of proteins associated with immunosuppression, including CD39 and CD79, which can suppress effector T cell responses (31). Consequently, CD39 and CD73 expression were evaluated on T cells isolated from PLN at day 7 p.i. from infected mice. The results show no difference in the expression of these proteins on the surface of CD4⁺ or CD8⁺ T cells (Fig. 8A and B). Analysis of regulatory CD4⁺ T cells (Treg) that control the magnitude of the T cell response (32) showed no significant difference in the number of cells in the PLN of infected mice (Fig. 8C and D).

FIG 7.

CD8⁺ T cells from the PLN of HSV-1 0ΔNLS-infected mice secrete more IFN-γ and express more TNF-α and IFN-γ compared to CD8⁺ T cells from GFP105 parental HSV-1. C57BL/6 male and female mice (n = 7 to 9/group) were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. At day 7 p.i., the PLN were harvested, processed to a single-cell suspension (5 × 10⁵ cells/well), and stimulated overnight with either gB (10 μg/mL) or gD peptide (10 μg/mL) to evaluate IFN-γ production by ELISPOT assay (A) or ELISA (B). (C) Gating strategy used to evaluate intracellular IFN-γ and TNF-α staining in gB-specific CD8⁺ T cells. Briefly, PLN cells (2 × 10⁶/mL) were stimulated with PMA and ionomycin in the presence of brefeldin A for 6 h. Next, the cells were stained with antibodies targeting HSV-1 gB-specific CD8⁺ T cells (CD45⁺ CD3⁺ CD8⁺ gB⁺), followed by fixation and permeabilization and intracellular staining for IFN-γ and TNF-α. The samples were acquired by flow cytometry and the expression of IFN-γ and TNF-α was determined within HSV-1 gB-specific CD8⁺ T cells by flow cytometry. (D) Number of HSV-1 gB-specific CD8⁺ T cells (CD45⁺ CD3⁺ CD8⁺ gB⁺) from the PLN of HSV-1 0ΔNLS- and GFP105-infected mice was determined by flow cytometry. (E) Summary of intracellular staining for IFN-γ and TNF-α within HSV-1 gB-specific CD8⁺ T cells. The values are presented as means ± the SEM. **, P < 0.01 (as determined by Student t test).

FIG 8.

Infection of mice does not alter the Treg cell number or the expression of CD39 and CD73 on CD4⁺ or CD8⁺ T cells. C57BL/6 mice (n = 6 to 7/group) were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 at 10⁵ PFU in 10 μL of PBS. At day 7 p.i., PLN were harvested and single cell suspensions (1 × 10⁶ cells) were stained for CD39 and CD73 expression in panels A and B or the number of CD4⁺ Treg cells in panels C and D.

Since we found significantly more CD8α⁺ DC express IL-12 obtained from the PLN of 0ΔNLS footpad-infected mice, we hypothesized these cells may play a critical role in the increase in HSV-1 gB-specific CD8⁺ T cell IFN-γ response observed in the 0ΔNLS-infected mice. To test this idea, we infected WT and basic leucine zipper transcription factor, ATF-like 3 (Batf3)-deficient (Batf3 KO) mice that have a deficiency in the development of CD8α DC (33), and assessed the generation of IFN-γ⁺ gB⁺ CD8⁺ T cells in the PLN at day 7 p.i. following footpad administration. The results show that even though the 0ΔNLS-infected mice still possessed more IFN-γ⁺ gB⁺ CD8⁺ T cells in the PLN at day 7 p.i. compared to GFP105-infected Batf3 KO mice (Fig. 9A), the difference was considerably lower in frequency (Fig. 9B) and number (Fig. 9C) compared to that found in WT-infected mice, reinforcing the observation that CD8α DC are a principal APC in the generation of antiviral CD8⁺ T cells (33), in this case against HSV-1 infection.

FIG 9.

BATF3 KO mice display a loss in the production of HSV-1 gB⁺ IFN-γ⁺ CD8⁺ T cells in the PLN after virus infection compared to WT mice. C57BL/6 (WT) and Batf3^−/− male and female mice (n = 4 to 6/group) were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. At day 7 p.i., the PLN were harvested, processed to a single-cell suspension (2 × 10⁶ cells/mL) and stimulated with PMA (100 nM) and ionomycin (1 μM) in the presence of brefeldin A (GolgiPlug) for 6 h. Next, the cells were stained with antibodies targeting CD45, CD3, and CD8, as well as the gB tetramer (CD45⁺ CD3⁺ CD8⁺ gB⁺), followed by fixation, permeabilization, and intracellular staining for IFN-γ. The samples were acquired by flow cytometry, and the expression of IFN-γ was determined within HSV-1 gB-specific CD8⁺ T cells by flow cytometry. (A) Representative examples of flow cytometry plots showing IFN-γ staining within HSV-1 gB-specific CD8⁺ T cells in WT and Batf3^−/− mice. (A and B) Frequency (B) and number (C) of gB⁺ CD8⁺ T cells stained for IFN-γ. The values are presented as means ± the SEM; **, P < 0.01; *, P < 0.05 (comparing the two groups as determined by Student t test).

In addition to changes in the level of IL-12 expressed by CD8α⁺ DC from GFP105- versus 0ΔNLS-infected mice, there were noted changes in type I IFN gene expression that could also contribute to the increase in IFN-γ-expressing HSV-1 gB-specific CD8⁺ T cells observed in the 0ΔNLS-infected mice. To further pursue this possibility, WT and type I IFN receptor (IFNAR1) deficient (IFNAR KO) mice were infected in the footpad with GFP105 or 0ΔNLS, and the PLN were removed at day 7 p.i. and evaluated for IFN-γ⁺ gB⁺ CD8⁺ T cells. The results show the absence of IFNAR1 eliminates the difference in IFN-γ⁺ gB⁺ CD8⁺ T cell frequency and number comparing GFP105- to 0ΔNLS-infected mice compared to infected WT animals (Fig. 10). Collectively, the data show that CD8⁺ T cells from the PLN of 0ΔNLS-infected mice produce significantly more IFN-γ and TNF-α after activation than CD8⁺ T cells from GFP105-infected animals, and this response correlates with an increase in the IL-12 production by CD8α DC and type I IFN levels in the PLN of the 0ΔNLS-infected mice. Furthermore, the absence of a functional type I IFN pathway through the loss of IFNAR1 negates the effect of 0ΔNLS increasing functional CD8⁺ T cells, as measured by IFN-γ expression.

FIG 10.

Deficiency in type I IFN signaling results in diminished IFN-γ production by PLN gB⁺ CD8⁺ T cells of vaccinated mice. C57BL/6 (B6) and IFNAR KO female mice were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. At day 7 p.i., the PLN were harvested, processed to a single-cell suspension (2 × 10⁶ cells/mL), and stimulated with PMA (100 nM) and ionomycin (1 μM) in the presence of brefeldin A (GolgiPlug) for 6 h. Next, the cells were stained with antibodies targeting HSV-1 gB-specific CD8⁺ T cells (CD45⁺ CD3⁺ CD8⁺ gB⁺), followed by fixation, permeabilization, and intracellular staining for IFN-γ. The samples were acquired by flow cytometry, and the expression of IFN-γ was determined within the HSV-1 gB-specific CD8⁺ T cell population. (A) Representative example of flow cytometry plots showing IFN-γ staining for cells from infected WT and IFNAR KO mice. (B and C) Frequency (B) and number (C) of gB⁺ CD8⁺ T cells stained for IFN-γ. The values are presented as means ± the SEM. *, P < 0.05; **, P < 0.01 (as determined by Student t test; n = 4).

GFP105 infection results in more dead/necrotic HSV gB-specific effector CD8⁺ T cells compared to 0ΔNLS infection.

HSV-1 infection of cells can lead to programmed cell death, including apoptosis, necroptosis, and pyroptosis through DNA-sensing receptors, including STING, that collectively are thought to restrict the replication and spread of the pathogen (34, 35). Apoptosis and cell death associated with HSV-1 infection are elicited by a variety of factors that include ICP0 gene expression (36), IFN-induced antiviral pathways (37, 38), and TNF-α (39). Since the HSV-1 0ΔNLS-infected mouse footpad tissue displayed an increase in select type I IFN-inducible genes and an increase in TNF-α expression within the first 24 h p.i., we further investigated whether these events would impact changes to the program cell death pathways measuring apoptosis, necrosis, and death within the PLN population comparing GFP105- to 0ΔNLS-infected mice. The overall PLN cell profile displayed 25% more dead cells recovered from PLN-stimulated cell cultures from GFP105-infected mice compared to 0ΔNLS-infected group (Fig. 11). However, there was no difference in the degree of apoptotic or necrotic cells comparing the two groups of infected mouse PLN cultured cells (Fig. 11). Within the HSV-1 antigen-specific CD8⁺ T cells, it was noted there were significantly more dead and necrotic effector cells in the stimulated PLN cultures from GFP105-infected mice compared to those from 0ΔNLS-infected animals (Fig. 12). Thus, even though greater cell death was found in the HSV-1 gB-specific CD8⁺ T cells from GFP105-infected mouse PLN, there was no apparent difference in the viable number of HSV-1 antigen-specific CD8⁺ T cells from either group of HSV-1-infected mice over the initial 7-day period postinfection.

FIG 11.

Infection of mice with 0ΔNLS induces resistance to cell death after recall stimulation with HSV-1 gB peptide. C57BL/6 mice were injected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. At day 7 p.i., PLN were harvested, processed to single-cell suspensions, plated at a concentration of 5 × 10⁶ cells/mL, and stimulated overnight (O/N) with HSV-1 class I-specific peptide (gB; 10 μg/mL). After O/N cell stimulation, the cells were washed with PBS supplemented with 2% FBS, and 2 × 10⁶ cells were stained with Apotracker Green 15 min at room temperature. Next, the cells were washed with PBS and stained with Zombie Aqua (final dilution, 1:500) for 15 min at room temperature. After a wash with PBS supplemented with 2% FBS, the cells were fixed with 2% PFA until flow cytometry acquisition. (A) Gating strategy for Zombie Aqua and Apotracker Green expression within total acquired cells. (B) Percentages of cells expressing either Zombie Aqua (dead), Apotracker Green (apoptotic), or double-positive results (necrotic) based on the summary data containing n = 5/group. The values are presented as means ± the SEM. **, P < 0.01 (as determined by two-way ANOVA, followed by Sidak’s multiple-comparison test).

FIG 12.

PLN HSV-1 gB peptide-stimulated cultures from 0ΔNLS-infected mice display a lower frequency of necrotic and dead cells compared to cultures from GFP105-infected animals. C57BL/6 mice were injected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. At day 7 p.i., PLN were harvested, processed to a single-cell suspension, plated at a concentration of 5 × 10⁶ cells/mL, and stimulated overnight (O/N) with HSV-1 class I-specific peptide (gB; 10 μg/mL). After O/N cell stimulation, the cells were washed with PBS supplemented with 2% FBS, and 2 × 10⁶ cells were stained with Apotracker Green for 15 min at room temperature. Next, the cells were washed with PBS and stained with Zombie Aqua (final dilution, 1:500) for 15 min at room temperature. After a wash with PBS supplemented with 2% FBS, the cells were stained with an antibody cocktail containing anti-mouse CD45 conjugated to Pacific Blue, anti-mouse CD3 conjugated to PE-Cy7, anti-mouse CD8a conjugated to APC, and HSV-1 gB tetramer conjugated to PE. The cells were incubated for 30 min on ice, washed with PBS supplemented with 2% FBS, and finally fixed with 2% PFA until flow cytometry acquisition. (A) Gating strategy for Zombie Aqua and Apotracker Green expression within total acquired cells. (B) Percentages of cells expressing either Zombie Aqua (dead), Apotracker Green (apoptotic), or double-positive results (necrotic) based on the summary data containing n = 5/group. The values are presented as means ± the SEM (**, P < 0.01; *, P < 0.05 [as determined by two-way ANOVA, followed by Sidak’s multiple-comparison test]).

HSV-1 0ΔNLS infection induces a potent polyfunctional effector CD8⁺ T cell response.

We previously reported CD8⁺ T cells that infiltrate the trigeminal ganglia of 0ΔNLS vaccinated mice displayed a high degree of polyfunctional characteristics compared to mock-vaccinated mice after HSV-1 challenge (15). To determine whether the initial exposure to HSV-1 0ΔNLS provided a similar effect in the local organized lymphoid tissue (i.e., PLN) and compare the response to the parental HSV-1 GFP105 infection, the HSV-1 gB-specific CD8⁺ T cells from PLN of infected mice were analyzed for their polyfunctional profile using IFN-γ, CD107a, and granzyme B as the prototypic markers. The profile of the HSV-1 gB-specific CD8⁺ T cell response from GFP105- and 0ΔNLS-infected mice is shown in Fig. 13A and B. Each pie chart contains color-coded segments with the average frequencies of antigen-specific CD8⁺ T cells producing different combinations of markers analyzed by SPICE software. Each color-coded arc represents the targeted molecule expressed by the gB-specific CD8⁺ T cells. Specifically, there was a significant increase in the frequency of cells expressing a combination of CD107a, IFN-γ, and granzyme B, as well as CD107a and IFN-γ, from the PLN of 0ΔNLS-infected mice compared to those infected with GFP105 (Fig. 13C). It should also be noted the nonresponsive gB-specific CD8⁺ T cells was >60% of the cells from the GFP105-infected mice compared to <40% of those mice infected with 0ΔNLS (Fig. 13C). These results indicate during acute exposure to HSV-1 0ΔNLS, there is a more favorable development of functionally active gB-specific CD8⁺ T cells within the local organized lymphoid tissue compared to animals infected with the parental virus suggesting the attenuation of the immediate early lytic protein, ICP0, plays a significant role in the influence of effector CD8⁺ T cell function most likely through changes in type I IFN expression and activation of downstream pathways that ultimately contribute to the adaptive immune response.

FIG 13.

HSV-1 0ΔNLS infection induces a favorable polyfunctional CD8⁺ T cell response during acute infection. Male and female C57BL/6 mice were infected with 0ΔNLS or GFP105. PLN were harvested at day 7 p.i. processed to create single-cell suspensions. Two million cells were stimulated with PMA and ionomycin for 6 h in the presence of brefeldin A (GolgiPlug), followed by gB⁺ CD8⁺ T cells staining with CD107a, IFN-γ, and granzyme B, as described in Materials and Methods. Stained cells were acquired by flow cytometry, and the frequency of gB⁺ CD8⁺ T cells expressing CD107a, IFN-γ, and granzyme B was determined using FlowJo software. Finally, the cell polyfunctionality was analyzed with SPICE software, and the results are presented as pie charts. (A and B) Average frequencies of gB⁺ CD8⁺ T cells expressing CD107a, IFN-γ, and granzyme B from mice infected with either GFP105 (A) or 0ΔNLS (B). The inner segments show cells expressing different combinations of molecules (listed in panel C, below the graph). The outer color-coded arcs around the circle indicate frequency of cells expressing each molecule (CD107a, IFN-γ, or granzyme B). Polyfunctionality is displayed by extension of the arcs surrounding the pie chart. (C) Summary data presenting frequency of gB⁺ CD8⁺ T cells producing different combinations of CD107a, IFN-γ, and granzyme B. The values are presented as means ± the SEM (n = 6/group). *, P < 0.05 (comparing the two groups, as determined by Student t test).

HSV-1 0ΔNLS-infected mice show a pronounced recall to HSV-1 antigen by CD8⁺ T cells.

The expression of the killer cell lectin-like receptor subfamily G member (KLRG1) is induced on cytotoxic CD8⁺ T cells upon exposure to strong inflammatory signals and along with IL-7Rα, is used as a marker to determine CD8⁺ effector T cell fate. Specifically, the KLRG1⁺ IL-7Rα-CD8⁺ T cell phenotype defines short-lived effector cells (SLEC), whereas KLRG1-IL-7Rα⁺ CD8⁺ T cell expression identifies memory precursor effector cells (MPEC) (Fig. 14A) (40). Thus, the frequency between SLEC and MPEC characterizes CD8⁺ T cell survival and lineage potential. To examine these characteristics in the HSV-1 0ΔNLS- and GFP105-infected mice, a comparison of KLRG1 and IL-7Rα expression on the HSV-1 gB-specific CD8⁺ T cells in the PLN population 7 days p.i. was conducted. The results show an increase in expression of KLRG1 on HSV-1 gB-specific CD8⁺ T cells from 0ΔNLS-infected mice compared to cells from GFP105-infected animals (Fig. 14B and C). However, the frequency of double-positive as well as HSV-1 gB-specific CD8⁺ T cells expressing only IL-7Rα was not significantly altered between the two groups of infected animals (Fig. 14B and C). Since the spleen becomes a reservoir for antigen-specific T cells long after viral clearance in the footpad and lymph node (41), we analyzed recall response of HSV-1 gB-specific CD8⁺ T cell isolated from spleens of primary infected animals at day 30 p.i. to gB peptide. The response of splenic CD8⁺ T cells to antigen (HSV-1 gB peptide) recall, as measured by IFN-γ production, was found to be significantly elevated by the CD8⁺ T cells isolated from the 0ΔNLS-infected mice compared to those isolated from GFP105-infected animals (Fig. 14D). The elevated expression of IFN-γ was not reflected by an increase in the number of HSV gB-specific CD8⁺ T cells from the spleen of 0ΔNLS-infected mice prior to stimulation. Specifically, we monitored the levels of the HSV-1 gB-specific CD8⁺ T cells over time in the PLN and spleen after footpad infection with HSV-1 GFP105 and 0ΔNLS (Fig. 15). As expected, the number of HSV-1 gB-specific CD8⁺ T cells in the PLN peaked at day 7 p.i. and subsequently contracted through 21 days p.i. (Fig. 15A). The peak in the spleen was found to occur at day 14 p.i. and contracted similar to the PLN population by day 28 p.i. (Fig. 15B). These results suggest previous exposure to HSV-1 0ΔNLS improves the CD8⁺ T cell response to the major MHC class I viral epitope in the absence of an excess memory pool in comparison to the parental GFP105-infected mouse memory HSV-1 gB-specific CD8⁺ T cell response.

FIG 14.

0ΔNLS infection induces an elevated HSV-1 gB-specific CD8⁺ T cell secondary response compared to mice infected with GFP105. Male and female C57BL/6 mice were infected in the footpad with either HSV-1 parental (GFP105) or live-attenuated (0ΔNLS) virus at 10⁵ PFU in 10 μL of PBS. At day 7 p.i., PLN were harvested, processed for a single-cell suspension, and stained for expression of KLRG1 and IL-7R within an HSV-1 gB-specific CD8⁺ T cell population. The frequency of SLEC and MPEC was determined based on the graph presented in panel A. (B) Representative contour plots showing KLRG1 and IL-7R expression within HSV-1 gB-specific CD8⁺ T cells. (C) Cell frequency for SLEC and MPEC phenotypes within the HSV-1 gB-specific CD8⁺ T cells. (D) C57BL/6 mice were infected in the footpad with either GFP105 or 0ΔNLS at 10⁵ PFU in 10 μL of PBS. At day 30 p.i., the spleens were harvested, and single-cell suspensions were stimulated with gB peptide for 24 h. CD8⁺ T cells expressing IFN-γ were determined by an ELISPOT assay. The values are presented as means ± the SEM (n = 7/group). P < 0.05 (comparing groups as determined by the Student t test).

FIG 15.

The number of HSV-1 gB-specific CD8⁺ T cells residing in the draining lymph node and spleen remains the same between HSV-1 GFP105- and 0ΔNLS-infected mice over time. Male and female C57BL/6 mice (n = 6 to 7 mice/group/time point) were injected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. At days 7, 14, and 21 p.i., PLN and spleens were harvested and processed to single-cell suspensions to a concentration of 5 × 10⁶ to 10 × 10⁶ cells/mL in PBS supplemented with 2% FBS. One million cells were taken for labeling using an antibody cocktail containing anti-mouse CD45 conjugated to Pacific Blue, anti-mouse CD3 conjugated to PE-Cy7, anti-mouse CD8a conjugated to APC-Cy7, and HSV-1 gB tetramer conjugated to PE. Cells were incubated for 30 min on ice, washed with PBS supplemented with 2% FBS, and fixed with 2% PFA until flow cytometry acquisition. The samples were acquired on either MACSQuant (Miltenyi Biotec) flow cytometer or an Aurora spectral flow cytometer (Cytek Biosciences), and the data were analyzed using FlowJo software. (A) Enumeration of HSV-1 gB-specific CD8⁺ T cells at days 7, 14, 21, and 28 p.i. in the PLN. (B) Enumeration of HSV-1 gB-specific CD8⁺ T cells at day 7, 14, and 21 p.i. in the PLN. The values are presented as means ± the SEM.

HSV-1 0ΔNLS-induced effector CD8⁺ T cells possess a decreased mitochondrial spare respiratory capacity.

CD8⁺ effector T cells require distinct metabolic attributes to support their energy needs. Specifically, upon activation these cells switch from oxidative phosphorylation to aerobic glycolysis resulting in a loss of reserve energy-generating capacity, i.e., mitochondrial spare respiratory capacity (SRC) (42). To determine whether a correlation exists between HSV-1 gB-specific CD8⁺ T cell effector function and SRC in the infected mice, PLN were harvested at day 7 p.i. After prestimulation ex vivo with gB peptide along with mouse rIL-2 for 3 days, cells were analyzed for changes in the oxygen consumption rate (OCR) in real time (Fig. 16A). Whereas basal respiration and ATP levels were not changed comparing HSV-1 antigen-specific CD8⁺ T cells from infected mice, the SRC was significantly lower in the cells from the 0ΔNLS-infected animals compared to cells from mice infected with parental GFP105 virus (Fig. 16B). To test whether the mitochondrial mass correlates with lower SRC, cultured antigen-specific CD8⁺ T cells were stained with MitoTracker Green after overnight stimulation with gB peptide. The results show the mitochondrial mass was smaller in the effector cells from the PLN of 0ΔNLS-infected animals compared to GFP105-infected mice (Fig. 16C). In addition, HSV-1 antigen-specific CD8⁺ T cells from the PLN of 0ΔNLS-infected mice displayed an increase in glucose uptake compared to cells from mice infected with GFP105 (Fig. 16D). Taken together, the metabolic profile of the HSV-1 antigen-specific CD8⁺ effector T cells from 0ΔNLS-infected mice is consistent with the increase in function of these effector cells in comparison to those cells from HSV-1 GFP105-infected mice.

FIG 16.

HSV-1 antigen-specific CD8⁺ T cell metabolism. C57BL/6 mice were infected in the footpad with either parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 virus at 10⁵ PFU in 10 μL of PBS. At day 7 p.i., PLN were harvested, processed for single-cell suspension, stimulated with 10 μg/mL of gB peptide along with 50 ng/mL of rIL-2 for 72 h, and magnetically enriched for HSV-1 gB-specific CD8⁺ T cells. The cells were then plated at a concentration 3 × 10⁵ cells per well, and the analysis of key parameters of mitochondrial function in terms of oxygen consumption rate (OCR) was performed using a Seahorse apparatus. (A) Metabolic flux measured at the baseline and after injection of oligomycin, FCCP, and rotenone/antimycin (Rot/AA). (B) The data gathered in panel A was used to determine select metabolic measurements, including basal respiration, ATP production, and spare respiratory capacity (SRC). (C and D) C57BL/6 mice were infected in the footpad with parental (GFP105) or live-attenuated (0ΔNLS) HSV-1 at 10⁵ PFU in 10 μL of PBS. At day 7 p.i., PLN were harvested, processed for single-cell suspension, and stimulated with gB peptide to evaluate either mitochondrial mass with MitoTracker Green FM (C) or glucose uptake using the substrate 2-NDBG (D). The values on the panels A to D are presented as means ± the SEM (n = 7 to 8/group). **, P < 0.01; *, P < 0.05 (comparing the indicated groups as determined by Student t test).

DISCUSSION

Cytotoxic T lymphocytes (CTLs) are historically the major T cells that target virally infected cells. After encountering pathogen-derived antigens presented by DC, CTLs integrate TCR and costimulatory molecule signaling from DC, along with cytokines, to drive their clonal expansion and function resulting in pathogen clearance by cytolytic mechanisms or cytokine-mediated responses (43). In our experimental model, it was found CD8⁺ but not CD4⁺ effector T cells taken from 0ΔNLS-infected mice show an elevated response to viral antigen compared to T cells from parental GFP105-infected animals. The dichotomy in the response of the antigen-specific CD4⁺ and CD8⁺ T cells is not suppression, since CD8⁺ and CD4⁺ T cells expressed IFN-γ poststimulation and have been shown to be anatomically distinct and regulated by different antigen-presenting cells (22). Thus, the role of CD4⁺ T cells remains to be elucidated in our infection model but is predicted to play a greater role in the development of the antibody response as we previously found adoptive transfer of CD4⁺ T cells into TCR-deficient mice restored the HSV-1 antibody neutralization titer and modestly reduced mortality of recipient mice following HSV-1 challenge (14).

CD8⁺ T cell cytolytic activity and/or production of IFN-γ have been associated with efficient regulation of HSV-1 infection (15, 44). In the present study, we found CD8⁺ effector T cells from 0ΔNLS-infected mice displayed a polyfunctional profile consisting of CD107a, granzyme B and IFN-γ expression above that found by CD8⁺ effector T cells from the parental GFP105-infected animals. This outcome was associated with an increase in IL-12 production by CD8α⁺ DC, an environment that is conducive to enhanced IFN-γ production following antigenic stimulation (45). The differences in IFN-γ production seen in our study does not only depend on CD8α⁺ DC production of IL-12 but also includes type I IFN. Type I IFN responses have been shown to be critical regulators of viral infection by inducing sets of genes that limit viral spread, as well as promote CD8⁺ T cell priming (46, 47). In addition, two prominent antiviral pathways activated by IFN-α and IFN-β include the oligoadenylate synthetases and double-stranded RNA-dependent protein kinase that have been shown to be critical in suppressing HSV-1 replication in mice (48, 49). Here, we show that 0ΔNLS infection elicited a pronounced expression of genes associated with type I IFN receptor activation, including Stat1, Stat2, Jak2, Irf1, Irf9, ISG15, and Tyk2, within the footpad at the site of initial infection compared to mice infected with the parental GFP105 virus. These downstream pathways have been associated with resistance to virus infection. For example, the antiviral effect of ISG15 has been linked to autophagic clusters and ubiquitin-like modifier properties elicited by IFN-β that antagonize HSV-1 (50, 51). In addition to the augmentation of expression of select antiviral molecules, there was also an elevation in select type I IFNs, including IFN-α7 and IFN-β. Previous studies have found various degrees of HSV-1 inhibition by specific type I IFN-α species and IFN-β with IFN-β showing the strongest inhibitory capacity (19). Furthermore, we also found the absence of a type I IFN pathway eliminated the elevated IFN-γ response by HSV-1 gB-specific CD8⁺ T cells from the PLN of 0ΔNLS-infected mice. Previous results have reported the absence of a functional type I IFN response significantly impacts the T cell response to ocular HSV-1 infection due to a loss of T cells associated with remodeling of the reticular fiber network in the paracortical-medullary T cell zones of the draining lymph nodes (52, 53). These results, along with the significant accumulation of IL-12-producing CD8α⁺ DC following 0ΔNLS infection, underscores the value of 0ΔNLS in priming CD8⁺ effector T cells and reaffirms the attributes of this attenuated virus as a strong HSV-1 vaccine candidate. Mechanistically, HSV-1 0ΔNLS contains an in-frame deletion of the ICP0 nuclear localization signal that prevents translocation of ICP0 to the nucleus, resulting in an increase in sensitivity to type I IFN and a deficiency to establish latency (13). ICP0 has been shown to inhibit IRF3- and IRF7-dependent activation (54, 55), reduce the levels of Toll-like receptor 2 adaptors, myeloid differentiation factor 88 (MyD88), and MyD88-like adaptor protein (Mal), and modulate NF-κβ activity (56). Moreover, ICP0 limits STING activation and inhibits STAT1 phosphorylation resulting in the reduction of type I IFN expression (57). Thus, 0ΔNLS induction of type I IFNs and type I IFN-inducible pathways, augmentation in CD8⁺ T cell priming, and the capacity to generate a potent neutralizing antibody (13) are encouraging traits of a prototypic HSV-1 vaccine.

In addition to the generation of a powerful adaptive immune response, a successful vaccine should also establish a memory T cell pool that can efficiently respond to subsequent pathogen exposure (58). We found here that infection with 0ΔNLS does generate a significantly higher frequency of antigen-specific CD8⁺ T cells expressing KLRG1, which may protect them from exhaustion (59). However, cells that are positive for KLRG1 but negative for IL-7R are short-lived effector cells and will die once the pathogen is cleared. Instead, the presence of double-positive and/or memory precursor effector cells (MPEC) determines the formation of long-term memory T cells. It has been reported that KLRG1⁺ IL-7R⁺ cells downregulate KLRG1 expression but mount an effective antiviral response (40). In the present study, we found no significant difference in the frequency of this population within the antigen specific CD8⁺ T cells and no difference in the frequency of MPEC formation. In addition, we found no difference in the number of HSV-1 gB-specific CD8⁺ T cells in the PLN or spleen well past the clearance of the virus. However, the function of the memory HSV-1 gB-specific CD8⁺ T cells from the 0ΔNLS-infected mice was elevated in terms of antigen-driven IFN-γ production compared to GFP105-infected mice, which we attribute to either differences in the status of “exhaustion” or anergy or to an increase in necroptosis/cell death of the memory cells after antigen stimulation, as observed during acute infection in the present study (Fig. 12).

Cell metabolism is a critical component central to the regulation of immune cell fate (60). The characteristic feature associated with effector T cells is a switch from oxidative phosphorylation to aerobic glycolysis (61). As a result, effector T cells possess a high glucose uptake profile, undergo mitochondrial fission presenting punctuate mitochondrial morphology, and adopt glycolysis for energy production (62). Furthermore, effector T cells show a reduction in the ratio between mitochondrial mass and overall cell mass, as well as a loss of reserve energy-generating capacity, SRC (42). Importantly, aerobic glycolysis is not required for cell proliferation, but strongly associated with IFN-γ production (61). We found that 0ΔNLS infection induced a lower OCR rate by antigen-specific CD8⁺ T cells resulting in a significant change in the SRC. Moreover, the cells were found to possess reduced mitochondrial mass but an increase in glucose uptake. Collectively, the data correlate the polyfunctional response of the 0ΔNLS-induced CD8⁺ effector T cells to changes in their metabolism.

In summary, we have found specific differences in the performance of HSV-1 specific CD8⁺ T cells from 0ΔNLS- versus parental GFP105-infected mice that underscores attributes of the 0ΔNLS mutant as a promising vaccine candidate against HSV-1 infection. CD8⁺ T cells play a critical role in suppressing HSV-1 reactivation that is the principal driver of corneal pathology in the human patient. Whether vaccination with 0ΔNLS can alter the immune response in the latent host is unknown but actively under investigation. Furthermore, the data generated to date characterizing the HSV-1 0ΔNLS live-attenuated virus as a vaccine has been conducted using an experimental animal model. Whether these results translate to the human host have yet to be ascertained but, thus far, the results provide a rational basis to pursue studies in the human patient.

MATERIALS AND METHODS

Viruses.

An ICP0^GFP gene was constructed in which GFP was inserted between amino acids 104 and 105 of ICP0. The chimeric ICP0^GFP was then introduced into the LAT-ICP0 locus of HSV-1 strain KOS by homologous recombination to yield HSV-1 GFP105. Northern and Western blot analysis confirmed the expression of the construct as previously described (63). HSV-1 GPF105 was propagated in Vero cells, and titers were determined, along with HSV-1 0ΔNLS, on ICP0-complementing L7 cells (64). HSV-1 0ΔNLS was 100-fold more sensitive to IFN-β (200 U/mL) compared to HSV-1 GFP105 in vitro in a standard growth assay using Vero cells at an MOI of 0.1 (17).

Mice and infection.

Male and female C57BL/6 (WT) (stock 000664) and type I IFN receptor (IFNAR1) deficient (stock 28288) mice were obtained from the Jackson Laboratory (Bar Harbor, ME). gBT-I.I (gBT) male and female transgenic mice (65) were bred in-house from the original stock and genotypically validated before use. All animals were housed in specific-pathogen-free conditions in the vivarium at the Dean McGee Eye Institute. This study was conducted according to approved protocol (19-060) by the University of Oklahoma Health Sciences Center animal care and use committee. WT mice (8 to 12 weeks old) were subcutaneously infected with the parental HSV-1 (GFP105) or mutated live-attenuated HSV-1 (0ΔNLS) in the footpad at 1 × 10⁵ PFU in a 10-μL volume of PBS, as described previously (13), and tissue was collected at the indicated times p.i.

DNA sequencing.

DNA sequencing of the 0ΔNLS mutant virus was performed at the Microbial Genome Sequencing Center (Pittsburgh, PA) on the NextSeq 200 platform with paired-end reads (2*151 bp). The DNA was extracted using DNeasy blood and tissue kit (Qiagen, Germantown, MD) performed according to the manufacturer’s procedure except for the final DNA elution, which was done in 20 μL of Tris-EDTA.

The paired-end sequence was processed on the T-BioInfo platform (https://server.t-bio.info), using Bowtie 2 to align sequencing reads to the reference HSV-1 KOS genome (GenBank accession number KT899744.1) and analyzed using the PernuclBasic Mutations, ConfInterval Binom95, and Mutability tools. Insertions, deletions, and substitutions were assigned based on a read prevalence cutoff of ≥0.5. The sequencing is based on 7,209,422 filtered reads and a coverage of 7,116×.

Viral titer.

Mice were anesthetized by intraperitoneal injection of ketamine (100 mg/kg) plus xylazine (6.6 mg/kg), followed by cardiac perfusion with 10 mL of PBS. Harvested skin tissue from euthanized mice at day 1, 3, 5, 7, and 9 p.i. was collected and homogenized in serum-free cell culture medium. The supernatant was clarified by centrifugation (300 × g, 5 min), and infectious virus was quantified by a plaque assay using ICP0-complementing L7 cells, as described previously (13).

Immunophenotyping cells from popliteal lymph nodes and footpads.

For immunophenotyping, the PLN were harvested from infected mice at the indicated time points. PLNs were homogenized into single-cell suspensions using a 40-μm-pore-size mesh filter and resuspended in staining buffer (PBS supplemented with 2% fetal bovine serum [FBS]). Footpad skin explants were digested in 1 mg/mL of type 1 collagenase (Sigma) for 30 min at 37°C, with trituration every 5 to 10 min. Next, cell suspensions were passed through 40-μm-pore-size mesh filters and resuspended in staining buffer. Spleens were harvested and processed as described for PLN but with an extra step to lyse red blood cells using Red Blood Cell Lysis Buffer (multi-species; Thermo Fisher Scientific). Antibody staining cocktails were used to target monocytes (CD45⁺ CD11b⁺ Ly6G-Ly6C⁺), neutrophils (CD45⁺ CD11b⁺ Ly6G⁺ Ly6C^–), HSV-1-specific CD4⁺ T cells (CD45⁺ CD3⁺ CD4⁺CD8^– gD-tetramer⁺) and HSV-1-specific CD8⁺ T cells (CD45⁺ CD3⁺ CD4^– CD8⁺ gB-tetramer⁺). Immunophenotyping of DC subsets in the lymph nodes, including epidermal Langerhans cells, dermal DC (dDC), CD103^– dDC, and lymph node-resident CD8a⁺ DC, has been previously described (10). DC activation was determined using the following antibody cocktail: CD86-Pacific Blue, MHC-I–Alexa Fluor 700, MHC-II-BV605, CD11b–FITC, CD45–Spark Blue, CD8a–PerCP-Cy5.5, CD103–PE, CD205–PE-Cy7, CD80–APC, and CD11c–APC-Fire 750. Regulatory T cells were phenotyped using antibodies to CD3, CD4, and CD25, followed by cell fixation and permeabilization and staining for FoxP3 (all from Thermo Fisher Scientific). The HSV-1 gD and gB tetramers were provided by the NIH tetramer core facility for the identification of HSV-1-specific CD4⁺ and CD8⁺ T cells, respectively.

Samples were analyzed using a three-laser flow cytometer MACSQuant (Miltenyi Biotec) or four-laser spectral flow cytometer Aurora (Cytek Biosciences, Fremont, CA) containing 16 violet, 14 blue, 10 yellow-green, and 8 red channels (4L-16V-14B-10YG-8R). The settings for flow cytometry acquisitions were established according to the manufacturer’s guidelines based on the instrument prior to the acquisition time. For acquisition, the MACSQuant compensation was established using the instrument’s software and, if necessary, modified post hoc. For acquisition using Aurora, compensation was performed using reference controls integrated in the SpectroFlo software. Each fluorochrome peaked in separate channels, and the spectral pattern was validated based on the references provided by Cytek Biosciences (Cytek Full Spectrum Viewer). Acquired samples were exported as FCS files and further analyzed using FlowJo software version 10.7.1 (BD Biosciences, Ashland, OR).

For SLEC and MPEC assessment, 3 × 10⁶ cells were stained with anti-mouse CD45 conjugated with PerCP-Cy5.5, anti-mouse CD3 conjugated with PE-Cy7, anti-mouse CD8 conjugated with APC-Cy7, HSV-1 gB tetramer conjugated with PE, anti-mouse KLRG1 (killer cell lectin-like receptor 1) conjugated with BV421, and anti-mouse IL-7R (IL-7 receptor) conjugated with APC (all antibodies from BioLegend, San Diego, CA). Cells were stained for 30 min on ice in the dark and finally washed with PBS supplemented with 2% FBS, fixed with 2% paraformaldehyde (PFA), and stored in a fridge until flow cytometry acquisition. Samples were acquired on MACSQuant flow cytometer (Miltenyi Biotec, Bergisch Gladbach, Germany) and next analyzed with FlowJo software (BD Biosciences, Ashland, OR). The expression of KLRG1 and IL-7R was determined within CD45⁺ CD3⁺ CD8⁺ gB⁺ T cells, and naming for the population was as follows: the SLEC (short-lived effector cells) population had the cell phonotype CD45⁺ CD3⁺ CD8⁺ gB⁺ KLRG1⁺ IL-7R^–, whereas the MPEC (memory precursor effector cells) population had the cell phenotype CD45⁺ CD3⁺ CD8⁺ gB⁺ KLRG1-IL-7R⁺.

Cell sorting.

Cell sorting was performed by using negative or positive selection with commercially available cell isolation kits for enrichment of naive CD8⁺ T cells, or isolation of gB-specific CD8 T cells (CD8⁺ gB⁺) using anti-PE microbeads (Miltenyi Biotec). Alternatively, the cells were labeled with biotin-conjugated antibodies (BioLegend), along with anti-biotin beads (Miltenyi Biotec), and enrichment was accomplished by positive selection using LS or MS columns under a strong magnetic field (Miltenyi Biotec) according to the manufacturer’s guidelines.

Semiquantitative RT-PCR.

For analysis of virus lytic gene expression, total RNA was isolated from PLN or footpad samples using the TRIzol reagent (Thermo Fisher Scientific, Waltham, MA) at the indicated time points after footpad administration of 1 × 10⁵ PFU GFP105 or 0ΔNLS. Total RNA was converted to cDNA with iScript reverse transcription (RT) supermix (Bio-Rad Laboratories). Real-time PCR analysis of thymidine kinase (TK) expression was performed using a forward (5′-ATACCGACGATCTGCGACCT-3′) and the reverse (5′-TTATTGCCGTCATAGCGCGG-3′) TK primer set as described previously (13). (For IFN-α and IFN-β expression, the total RNA was isolated from PLN after footpad administration of 1 × 10⁵ PFU GFP105 or 0ΔNLS and converted into cDNA using iScript RT Supermix [Bio-Rad].) The reaction was set using iTaq Universal SYBR green Supermix, along with the recommended cDNA and primer concentrations. To determine the expression of IFN-inducible genes, total RNA was isolated from the footpad at 12 h p.i. and converted into cDNA using iScript RT Supermix, and primePCR technology was utilized with commercially validated primer sequences (Bio-Rad Laboratories) and performed according to the manufacturer’s guidelines. The relative expression of targeted genes was normalized to naive controls using the 2^–ΔΔCT method with glyceraldehyde 3-phosphage dehydrogenase (GAPDH) as the internal reference gene. The heat map for the visual comparison of gene expression was built by using the open-source software package Morpheus (available from the Broad Institute website [https://software.broadinstitute.org/morpheus]).

Cytokine/chemokine evaluation.

ELISPOT assay for IFN-γ expression was described previously (15). Briefly, 96-plate with Immobilon-P membrane were coated overnight with primary IFN-γ antibody (BD Biosciences). Cells isolated from PLN were plated at a concentration of 5 × 10⁵ cells per well, followed by overnight restimulation with HSV-1-derived gB or gD peptides (10 μg/mL). After the cells were removed, the plates were developed as described previously (15). To evaluate the IFN-γ in the cell culture supernatants, an ELISA kit for IFN-γ was used according to the manufacturer’s instructions (R&D Systems). The cell culture supernatants were collected after overnight cell stimulation using HSV-1-derived gB or gD peptides (10 μg/mL). IFN-γ and TNF-α intracellular staining was conducted according to a previously published methodology (15). Briefly, at day 7 p.i., the PLN were harvested and processed to a single-cell suspension, and 2 × 10⁶ cells/mL were cultured and stimulated with phorbol-12-myristate-13-acetate (PMA) and ionomycin (both from Millipore) for 6 h at 37°C and 5% CO₂ in the presence of brefeldin A (GolgiPlug from BD Biosciences). Next, the cells were washed with PBS (supplemented with 5% FBS and NaN₃) and stained with Abs targeting HSV-1 gB-specific CD8⁺ T cells (CD45⁺ CD3⁺ CD8⁺ gB⁺) followed by fixation and permeabilization using the commercially available eBioscience FoxP3/transcription factor staining buffer set accordingly to the manufacturer’s guidelines. At the time of permeabilization, the cells were stained using anti-IFN-γ and TNF-α antibodies (both from BioLegend) for 30 min at room temperature. The cells were then washed in 2 mL of permeabilization buffer, centrifuged for 5 min at 300 × g, 4°C, and resuspended in PBS supplemented with 2% FBS. The samples were then analyzed using a MACSQuant flow cytometer (Miltenyi Biotec). IL-12 production was induced using CpG ODN 1826 as described previously (54). Briefly, the PLN were harvested at day2 p.i., processed to single cell suspensions, and 2 × 10⁶ cells/mL were cultured and stimulated with 1 μM CpG ODN 1826 (InVivoGen) for 6 h in the presence of brefeldin A (GolgiPlug from BD Biosciences). After stimulation, the cells were washed with PBS supplemented with 2% FBS and stained for cell surface antigens including CD45-Pacific Blue, CD11c-APC, and CD8α-FITC (all from BioLegend). To exclude the cells of lymphoid lineage, the cells were costained with anti-mouse antibodies conjugated with PE, including CD3, CD4, CD19, B220, TCR γ/δ, NK1.1 (all from BioLegend). The intracellular staining for IL-12 was performed similar like it was for intracellular staining for IFN-γ and TNF-α (described above). After cell surface staining, the cells were fixed and permeabilized using a FoxP3/transcription factor staining buffer set (Thermo Fisher Scientific). At the permeabilization step, the anti-IL-12/IL-23 p40 (monomer, dimer, heterodimer) antibody conjugated with APC (clone: C15.6 from BioLegend) was added and the samples were incubated for 30 min at room temperature. Finally, the samples were washed with PBS supplemented with 2% FBS (2 mL/tube) and acquired with MACSQuant flow cytometer (Miltenyi Biotec).

Cytokine and chemokine content in the footpad was determined as previously described (66). Briefly, footpads were dissected from exsanguinated PBS-treated, GFP105-, and 0ΔNLS-infected mice 24 h p.i. and weighed. The samples were then placed in Next Advance Green bead lysis tubes containing a protease inhibitor mixture (Santa Cruz) in PBS. Following homogenization for two 5 min pulses using a storm 24 bullet blender (Next Advance), the samples were sonicated in a water bath for 10 min, centrifuged (10,000 × g, 2 min at 4°C), and the clarified supernatant was analyzed for cytokine/chemokine content using a customized kit targeting the following analytes: G-CSF, GM-CSF, IFN-γ, IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-9, IL-10, IL-12p40, IL-12p70, IL-13, IL-15, IL-17, CCL2, CCL3, CCL4, CCL5, CCL11, CXCL1, CXCL2, CXCL9, CXCL10, LIF, M-CSF, TNF-α, and VEGF-A and a Luminex-based, 32 plex assay (Millipore Sigma, Billerica, MA) analyzed on a Bio-Plex 100 system (Bio-Rad Laboratories) as described previously (66).

Assessment of polyclonal T cell proliferation.

PLN were harvested from GFP105- or 0ΔNLS-infected mice, and the cells were processed to single cell suspensions. Cells were then cultured in the presence of 2 μM CFSE (CellTrace CFSE cell proliferation kit from Thermo Fisher Scientific) and incubated for 72 h in wells of 24-well plates at a concentration of 0.5 × 10⁶ cells/mL with soluble polyclonal stimuli using mouse anti-CD3 (concentration range, 0.001 to 1.0 μg/mL) and anti-CD28 (1 μg/mL). After incubation, CFSE-labeled cells were stained with anti-mouse CD4 conjugated with APC, and anti-mouse CD8 conjugated with APC-Cy7 (both from BioLegend) and analyzed using a MACSQuant flow cytometer. The CFSE expression within CD8⁺ and CD4⁺ T cell populations was analyzed using FlowJo software.

Assessment of cell proliferation in vivo and ex vivo.

For evaluation of clonal CD8 T cell expansion in vivo, 5 × 10⁶ cells were retro-orbitally transferred into recipient mice administered 1 × 10⁵ PFU of GFP105 or 0ΔNLS/footpad. The cells were first isolated from pooled spleen and lymph nodes of gBT-I.1 transgenic mice using commercially available magnet cell sorting kit (CD8a T cell isolation kit) from Miltenyi Biotec. The isolated cells were labeled with a CellTrace CFSE cell proliferation kit (Thermo Fisher Scientific) at a concentration of 2 μM for 15 min at room temperature. Next, the cells were washed with 4°C complete cell culture media (containing 10% FBS) and resuspended in PBS followed by adoptive cell transfer via retro-orbital route. Five days later, PLN of infected animals were harvested, and the cells were stained with CD45-Pacific Blue, CD3-PECy7, and CD8-APC-Cy7. CFSE dilution was determined within CD45⁺ CD3⁺ CFSE⁺ CD8⁺ cells. The analysis was performed with regard of analysis the cell frequency in single cell division peaks.

For evaluation of CD8⁺ T cell expansion ex vivo, mice were infected with 1 × 10⁵ PFU of GFP105 or 0ΔNLS/footpad and 2 days after infection PLN were harvested, processed to single-cell suspensions and stained for CD11c-biotin. Next, the cells were incubated with anti-biotin beads (Miltenyi Biotec) and enriched by positive selection using MS columns (Miltenyi Biotec). Isolated CD11c⁺ cells were mixed with enriched naive gBT-I.1 transgenic CD8⁺ T cells stained with CFSE at the ratio 1:20 and cocultured for 3 days. CFSE dilution was determined within CD45⁺ CD3⁺ CFSE⁺ CD8⁺ T cells. The analysis was performed with regard to the frequency in single cell division peaks.

Evaluation of polyfunctional CD8 T cell response.

The methodology for evaluation of polyfunctional CD8 T cell response has been described elsewhere (15). Briefly, polyfunctional T cell responses were evaluated by simultaneously staining of CD8⁺ T cells with CD107-BV421, granzyme B–FITC, and IFN-γ-APC antibodies (all from BioLegend) using single-cell suspensions from PLN of mice infected with either GFP105 or 0ΔNLS. At day 7 p.i., the PLN were harvested, processed to single-cell suspensions and 2 × 10⁶ cells/mL were stimulated with PMA and ionomycin (both from Millipore) for 6 h at 37°C and 5% CO₂ in the presence of brefeldin A (GolgiPlug from BD Biosciences). CD107a-BV421 Ab was added at beginning of cell stimulation. At the end of the 6-h stimulation period, the cells were washed with PBS (supplemented with 5% PBS and 0.09% NaN₃) and stained with antibodies targeting HSV-1-specific CD8⁺ T cells (CD45⁺ CD3⁺ CD8⁺ gB⁺), followed by fixation and permeabilization using a commercially available FoxP3/transcription factor staining buffer set accordingly to the manufacturer’s guidelines (eBioscience). At the permeabilization step, the cells were intracellularly stained with granzyme B-FITC and IFN-γ-APC. Stained cells were acquired with a flow cytometer (MACSQuant), and the flow cytometry files were exported and analyzed using FlowJo software. Polyfunctional CD8⁺ T cell responses were evaluated using simplified presentation of incredibly complex evaluation (SPICE) software (version 5.3; National Institute of Allergy and Infectious Diseases, National Institutes of Health). Prior to SPICE analysis, the data were processed using FlowJo software by manual gating of cell populations of interest, and the output data with all possible combinations of marker expression were saved in Excel file format (converted to .csv format) and uploaded and processed using SPICE software.

Metabolic flux assay.

Oxygen consumption rates (OCR) were determined using XFe96 Analyzers (Agilent), as previously described (15). Magnet-enriched HSV-1 antigen-specific CD8⁺ T cells, previously stimulated in vitro with congenic gB peptide, along with rIL-2, were resuspended in Seahorse XF RPMI medium containing 2 mM l-glutamine, 1 mM sodium pyruvate, and 10 mM glucose and then seeded on Cell-Tak-coated 96-well plates (Agilent) at 3 × 10⁵ cells per well. OCR were measured before and after injection with 1.5 μM oligomycin, 1 μM carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP), and 0.5 μM rotenone/antimycin A (Rot/AA). All changes in metabolic flux were recorded in real-time with Wave software (v2.6.1). Basal OCR levels were determined as OCR values before the injection of oligomycin. Baseline ATP production was determined by subtracting the average OCR values obtained after oligomycin administration from the average basal OCR values. The spare respiratory capacity was defined as the average OCR values after FCCP administration minus the average baseline OCR values.

Detection of mitochondrial mass and assessment of glucose uptake.

The detection of mitochondrial mass has been described in detail elsewhere (67). Briefly, the cells from PLN were stimulated overnight with gB peptide (10 μg/mL). Thirty minutes before ending the cell culture, MitoTracker Green FM (Thermo Fisher Scientific) was added at a concentration of 150 nM, followed by incubation at 37°C and 5% CO₂. Next, the cells were washed and stained with cell surface markers (CD45⁺ CD3⁺ CD8⁺), along with gB tetramer, for 30 min on ice. After ending the incubation, the cells were washed with PBS supplemented with 2% FBS and kept on ice under cover until the flow cytometry acquisition time was initiated. To measure glucose uptake, a fluorescent d-glucose analog, 2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]-2-deoxy-d-glucose (2-NBDG), with flow cytometry detection of fluorescence produced by cells was used. Overnight-cultured cells in the presence of gB peptide (10 μg/mL) were treated at the end of cell culture with 2-NDBG for 30 min at 37°C and 5% CO₂. The cells were washed with staining buffer (PBS supplemented with 2% FBS), stained with cell surface antigens (CD45⁺ CD3⁺ CD8a⁺) along with gB tetramer, and incubated on ice for 30 min. The cells were then washed, resuspended in staining buffer, and kept on ice in the dark until initiating the flow cytometry acquisition time. Samples were acquired on spectral flow cytometer (Aurora; Cytek Biosciences) with presettings using single cell-stained controls. Recorded FCS files were exported and analyzed using FlowJo software.

Statistical analysis.

GraphPad Prism, version 9.1.0, was used for statistical analysis. All data are expressed as means ± the standard errors of the mean (SEM). Data were analyzed as indicated for each panel of each figure. A P value of <0.05 was considered significant.