To determine the role of neutrophils in our model of vaginal HSV-2 infection, wild-type (WT) female C57BL/6 mice were treated with Depo-Provera (depot medroxyprogestrone, DMPA) to hold mice at the diestrus phase of the estrus cycle and normalize susceptibility to infection (Kaushic et al., 2003). Neutrophils were depleted in DMPA-treated mice by intraperitoneal (i.p.) injection of an antibody against Ly6G, a neutrophil-specific marker, or an isotype control. One day later, mice were inoculated intravaginally with 5000 plaque forming units (PFU) of WT HSV-2 strain 186 syn+ (WT HSV-2). Neutrophils were effectively reduced up to 6 days post-infection (d.p.i.) in the vagina (Figure 1A) and the blood (Figure 1—figure supplement 1). In order to focus on genital inflammation, mice were monitored for 1 week after infection, as progression of disease within the second week of our infection model is largely indicative of viral dissemination into the central nervous system. In both cohorts, mild genital inflammation was apparent starting at 4 d.p.i. in a small fraction of mice (Figure 1B). Over time, progression of disease in the neutrophil-depleted mice was significantly slower compared to the controls. Remarkably, as late as 7 d.p.i., a proportion of the neutrophil-depleted group remained uninflamed, in contrast to the isotype control group in which all mice displayed signs of inflammation (Figure 1B). To confirm the disparity in disease severity, we examined the vagina and genital skin by histology. At 6 d.p.i., epithelial denuding and damage was apparent in the isotype control-treated mice (Figure 1C). In contrast, only a limited amount of epithelial destruction was observed in neutrophil-depleted mice, with less cellular infiltrates at sites of damage and in the lumen (Figure 1C). Furthermore, the epithelial layer proximal to areas of damage was morphologically distinct in isotype control-treated animals compared to neutrophil-depleted animals, suggesting diverse epithelial responses after infection in the presence or absence of neutrophils (Figure 1C). Similarly, destruction of the epidermis and separation of the epidermis from the dermis were widespread in the genital skin of isotype control-treated mice but not in neutrophil-depleted mice (Figure 1C). Unexpectedly, differences in genital inflammation and mucosal damage were largely independent of changes in viral control in the absence of neutrophils, as viral shedding into the vaginal lumen (Figure 1D) and viral control in the tissue parenchyma (Figure 1E) were similar between the two groups. Indeed, disease severity was decreased in neutrophil-depleted mice despite a slight delay in the resolution of viral replication at 5 d.p.i. (Figure 1D).

We next evaluated whether the decreased inflammation after neutrophil depletion was due to changes in the cellular response against HSV-2 infection. We examined the recruitment of Ly6C+ monocytes, NK cells, and CD4 and CD8 T cells (Figure 1—figure supplement 2A), all of which have been implicated in either the control of HSV or modulation of disease severity (Lee and Ashkar, 2012; Shin and Iwasaki, 2013; Truong et al., 2019). To remove intravascular cells and to limit our analysis to cells within the vagina, tissues were thoroughly perfused prior to collection (Scott et al., 2018). Unexpectedly, there was no significant difference in the number of Ly6C + CD11b + cells (Figure 1—figure supplement 2B), NK cells (Figure 1—figure supplement 2C), total CD4 (Figure 1—figure supplement 2D), or CD8 T cells (Figure 1—figure supplement 2E) that were recruited to the vagina over the first 6 days after infection regardless of whether neutrophils were present or not. Thus, our data demonstrate that neutrophils do not play a significant antiviral role in our model of vaginal HSV-2 infection, and rather promote genital inflammation with minimal impact viral burden and recruitment of other immune cells to the vagina.

We next wanted to determine whether neutrophil-specific effector functions were promoting disease after HSV-2 infection. NETs have been associated with tissue damage in the context of both infectious (Jenne and Kubes, 2015) and non-infectious disease (Granger et al., 2019). To test whether NETs play a role in genital disease after HSV-2 infection, we first examined the ability of neutrophils to form NETs when exposed to HSV-2. In vitro stimulation of neutrophils with HSV-2 resulted in the enlargement of cell nuclei and the characteristic expulsion of DNA coated in citrullinated histones, which is a key characteristic of NETs (Figure 1—figure supplement 3A). The formation of NETs requires input from multiple pathways, including histone citrullination by enzymes such as PAD4, which leads to chromatin de-condensation and the eventual release of DNA (Li et al., 2010). To generate animals that were specifically lacking PAD4 in neutrophils, we bred Padi4^fl/fl x S100a8-Cre mice (PAD4 CKO). HSV-2 infection of these mice and their littermate controls demonstrated minimal impact on genital inflammation (Figure 1—figure supplement 3B) or viral replication (Figure 1—figure supplement 3C) in the genital mucosa. Thus, our data show that PAD4 expression in neutrophils and likely NET formation are not the mechanisms by which these cells mediate disease after HSV-2 infection.

We next tested whether ROS production by neutrophils mediated inflammation after HSV-2 infection. While production of ROS in neutrophils supports antimicrobial activity against a variety of pathogens (Dinauer, 2019), excessive oxidative stress can be associated with tissue injury (Mittal et al., 2014). We found that in vitro stimulation of neutrophils with HSV-2 led to an increase in ROS production compared to unstimulated cells (Figure 1—figure supplement 4A). To determine whether respiratory burst in neutrophils promoted genital inflammation after HSV-2 infection in vivo, we infected mice with germline deficiency in Ncf2 (Ncf2 KO), which encodes p67^phox, a key component of the NADPH oxidase complex (Jacob et al., 2017). HSV-2 infection of Ncf2 KO and heterozygous controls resulted in similar progression of disease (Figure 1—figure supplement 4B) and did not alter viral titer (Figure 1—figure supplement 4C). To confirm that tested neutrophil effector functions, including ROS production, had little impact on genital inflammation, we infected mice in which the calcium-sensing molecules STIM1 and STIM2 were deleted from neutrophils, as these calcium-sensing molecules cooperatively regulate neutrophil activation and select effector functions (Clemens et al., 2017). Stim1^fl/fl x Stim2^fl/fl x S100a8-Cre (STIM1/2 DKO) mice were infected with HSV-2 and monitored for disease. As expected, there was little difference in genital inflammation severity between the STIM1/2 DKO and Cre- controls (Figure 1—figure supplement 4D) or viral titers (Figure 1—figure supplement 4E). Together, our data show that ROS production from neutrophils and other cell types play little role in driving genital inflammation after HSV-2 infection.

To identify the factors that drove pathogenic neutrophil responses after HSV-2 infection, we turned to a complementary model of HSV-1 genital infection that we had previously described (Lee et al., 2020). Inoculation with the same dose of HSV-1 and HSV-2 led to profound differences in genital inflammation (Figure 2A) despite comparable levels of mucosal viral shedding throughout most days after infection (Figure 2—figure supplement 1A; Lee et al., 2020), although resolution of HSV-2 infection at 6 and 7 d.p.i. was delayed (Figure 2—figure supplement 1A,B). Importantly, magnitude of the neutrophil response in the vagina was similar between HSV-1- and HSV-2-infected mice during the course of acute mucosal infection (Figure 2B), and neutrophils could be found infiltrating sites of both HSV-1- and HSV-2-infected epithelium (Figure 2C). In contrast to HSV-2 infection, antibody-mediated depletion of neutrophils with anti-Ly6G antibody prior to inoculation with HSV-1 did not reduce the development of genital inflammation during the first 7 days after infection (Figure 2D) and had minimal impact on acute viral control (Figure 2D - Figure 2—figure supplement 1C). Together, our data suggests that the regulation of the neutrophil response after HSV-1 or HSV-2 infection was distinct, which may contribute to the differences in disease outcomes between these infections.

To better understand the differences between pathogenic neutrophil responses after HSV-2 infection and the non-pathogenic neutrophil responses after HSV-1, we performed single-cell RNA sequencing (scRNA-seq) on sorted live vaginal cells from a mock-infected mouse or mice infected with HSV-1 or HSV-2 using the 10x Genomics platform (Zheng et al., 2017). Each sample was composed of cells from a single animal to better delineate potential subsets within cell populations, particularly neutrophils. Analysis across 21,633 cells in all samples revealed 17 unique clusters in the vagina during HSV infection after filtering, including myeloid cells, lymphocytes, and epithelial cells (Figure 3A). Neutrophils were identified by expression of known cell markers such as S100a8 and Csf3r (Figure 3B). In mock-infected animals, the vaginal neutrophil population was dominated by cluster 0, and upon infection, at least two additional neutrophil subsets, cluster 2 and cluster 5, were clearly present (Figure 3C). While HSV-1-infected mice retained all three subpopulations of neutrophils in the vagina at 5 d.p.i., in HSV-2-infected mice, the presence of cluster 0 was greatly reduced, and the bulk of the neutrophils was composed of cluster 2 and 5 (Figure 3C). One major distinguishing characteristic between ‘homeostatic’ cluster 0 and ‘infection’ clusters 2 and 5 was the extent of ISG expression, in which cluster 0 expressed low levels of genes associated with a type I IFN response, even in infected animals, while clusters 2 and 5 expressed high levels of these genes (Figure 3D; Liberzon et al., 2015). Furthermore, the gene expression profile of clusters 2 and 5 was different between HSV-2 and HSV-1 infection at 5 d.p.i. (Figure 3—figure supplement 1), including the expression of ISGs (Figure 3C, Figure 3—figure supplement 1). Differential expression of select ISGs was confirmed by quantitative reverse transcription PCR (qRT-PCR) analysis in the vagina at 5 days after HSV-1 or HSV-2 infection (Figure 3—figure supplement 2). qRT-PCR shows that expression of CXCL10 (Figure 3—figure supplement 2A,B) and Gbp2 (Figure 3—figure supplement 2C,D; Glennie et al., 2015) is increased in HSV-2-infected vaginas compared to HSV-1, while IL-15 is not (Figure 3—figure supplement 2E,F), which supports the accompanying scRNA-seq analysis. While type I IFN was robustly produced early during acute infection after both HSV-1 and HSV-2 infection (Figure 3—figure supplement 3), IFNβ levels were higher in the vaginal lumen after HSV-2 infection compared to HSV-1 at time points corresponding to the onset of genital inflammation (Figure 3E) despite similar viral burden between the two infection models (Figure 2—figure supplement 1A). IFNβ was undetectable in both the vaginal lumen and the parenchyma by 7 days after both HSV-1 and HSV-2 infection (Figure 3—figure supplement 3B,C). Thus, during viral infection, distinct neutrophil subsets can be classified by transcriptional profiling, and expression of ISGs suggests that a key difference between a pathogenic and non-pathogenic neutrophil response during viral infection may be sustained IFN production and signaling.

We next wanted to test whether type I IFN signaling promoted immunopathology during genital HSV-2 infection. IFNAR1-deficient mice are highly susceptible to HSV, regardless of the route of inoculation (Gill et al., 2006; Iversen et al., 2010; Iversen et al., 2016; Reinert et al., 2012; Royer et al., 2019; Svensson et al., 2007; Wilcox et al., 2016), and rapidly succumb to infection, mainly due to a loss of viral control. To investigate the temporal effects of type I IFNs in HSV-2 genital disease, we used an antibody against IFNAR1 to block the receptor at different time points after infection (Scott et al., 2018). When mice were injected i.p. with anti-IFNAR1 antibody on the day of HSV-2 inoculation, disease progression was more rapid compared to isotype control-treated animals (Figure 4—figure supplement 1A), and the mice succumbed to infection at a faster rate (Figure 4—figure supplement 1B), in a manner similar to IFNAR1-deficient mice (Iversen et al., 2010; Iversen et al., 2016; Lee et al., 2017; Reinert et al., 2012; Wang et al., 2012). Inflammation and rapid disease progression were likely due to significantly elevated viral burden in the anti-IFNAR1 antibody-treated mice compared to isotype controls (Figure 4—figure supplement 1C), as HSV is a highly lytic virus that is capable of independently inducing epithelial tissue damage (Horbul et al., 2011). To focus on the effects of persistent IFN signaling in the vagina after HSV-2 infection, we also treated mice with a single injection of anti-IFNAR1 antibody or an isotype control at 4 d.p.i. In stark contrast to early anti-IFNAR1 antibody treatment, one treatment with therapeutic IFNAR1 blockade led to a significant reduction in the severity of inflammation compared to isotype controls (Figure 4A). Histology of vaginal tissues from isotype-treated controls at 6 d.p.i. showed widespread epithelial denuding and immune cell infiltrates within the epithelial layer of the vagina (Figure 4B). In contrast, damage to the vaginal epithelium in anti-IFNAR1 antibody-treated mice appeared to be localized (Figure 4B), similar to neutrophil-depleted mice (Figure 1C). Similarly, the genital skin of isotype control-treated mice displayed signs of severe inflammation and destruction of the epidermis, while the skin structure of anti-IFNAR1 antibody-treated mice was largely intact (Figure 4B). Furthermore, IFNAR1 blockade at 4 d.p.i. had little impact on mucosal viral shedding (Figure 4C). Collectively, these data show that the protective effect of type I IFN on control of genital HSV infection is limited to the early stages of acute infection, and that sustained IFN signaling in the later stages of acute HSV-2 genital infection drives inflammation with minimal effect on viral replication.

Single-cell transcriptional profiling data suggested that type I IFN signaling was robust in vaginal neutrophils after HSV-2 infection (Figure 3D). To determine whether intrinsic IFN signaling in neutrophils promoted immunopathology, we deleted IFNAR1 from granulocytes by breeding Ifnar1^fl/fl x S100a8-Cre mice (IFNAR1 CKO). After confirming that IFNAR1 ablation was limited to the neutrophil population (Figure 5A), IFNAR1 CKO mice and littermate Cre- controls were vaginally infected with HSV-2. Despite differences in IFNAR1 expression, the number of neutrophils recovered from the vaginal lumen was similar between the IFNAR1 CKO mice and their Cre- control littermates (Figure 5B). Strikingly, although the magnitude of the vaginal neutrophil response was similar, we found that the severity of genital inflammation presented by the IFNAR1 CKO mice was significantly reduced compared to the Cre- controls (Figure 5C). As observed after neutrophil depletion, a subset of the IFNAR1 CKO cohort did not develop any signs of inflammation as late as 7 d.p.i. (Figure 5C). Similar to our observations with therapeutic IFNAR1 blockade, IFNAR1 CKO mice exhibited less pathology in both the vagina and genital skin compared to Cre- controls (Figure 5D). Distinct disease outcomes between the Cre- controls and IFNAR1 CKO mice occurred independently of viral control, as viral loads in the mucosa were similar between the two groups (Figure 5E). Together, our data demonstrates that tissue inflammation during HSV-2 infection is largely driven by prolonged type I IFN production, which acts directly upon neutrophils to drive disease.

Type I IFN stimulation of neutrophils can upregulate ISGs as well as several pro-inflammatory cytokines (Galani et al., 2017). To determine whether type I IFN was driving disease by shaping the cytokine milieu within the vagina, we first measured several pro-inflammatory cytokines in the vagina at 5 d.p.i., in the presence or absence of neutrophils. The production of inflammatory cytokines such as IL-6 (Figure 6—figure supplement 1A), IL-1β (Figure 6—figure supplement 1B), or TNF (Figure 6—figure supplement 1C), all of which have been associated with genital inflammation and HSV-2 infection in humans (Gosmann et al., 2017; Masson et al., 2014; Murphy and Mitchell, 2016), was similar between both neutrophil-depleted and control groups. Production of IFNγ (Figure 6—figure supplement 1D) as well as IL-12p70 (Figure 6—figure supplement 1E), both cytokines associated with a type I immune response and important for HSV control, was similar between the neutrophil-depleted and control groups. However, when we measured IL-18, an IL-1 family cytokine that is primarily known for mediating innate defense (Harandi et al., 2001) and for promoting IFNγ production from NK cells during genital HSV-2 infection (Lee et al., 2017), we detected a notable difference between neutrophil-depleted and control mice (Figure 6A), suggesting an unexpected role for this cytokine in driving disease during HSV-2 infection.

To determine whether type I IFN signaling regulated IL-18 production in the vagina, we assessed IL-18 levels in the vaginal lumen after therapeutic antibody-mediated IFNAR1 blockade. At 5 d.p.i., similarly to neutrophil-depleted mice, we found that IL-18 levels were markedly reduced (Figure 6B). Importantly, measurement of IL-18 in the vagina of IFNAR1 CKO at 5 d.p.i. also revealed a significant decrease in cytokine levels compared to littermate controls (Figure 6C).

To determine whether IL-18 was playing a key role in driving immunopathology during genital HSV-2 infection, we therapeutically administered an IL-18-neutralizing antibody directly at the site of infection to HSV-2-infected animals starting at 3 d.p.i. in order to promote sufficient antibody concentration and activity in the relevant tissue. Remarkably, neutralization of IL-18 led to a considerable reduction in disease severity (Figure 6D), without any impact on viral control (Figure 6E). To determine the source of pathogenic IL-18 in the vagina, we probed vaginal tissues for the neutrophil marker S100A8 and IL-18 at 6 d.p.i. (Figure 6F). Detection of IL-18 and S100A8 around a single nucleus demonstrated that neutrophils could be a source of IL-18 during vaginal HSV-2 infection (Figure 6F). However, we also identified IL-18-reactive cells that were negative for S100A8 but in close proximity to neutrophils (Figure 6F), suggesting the potential for multiple cellular sources of IL-18. Thus, our data demonstrate that sustained type I IFN signaling in neutrophils leads to the production of vaginal IL-18 and reveal IL-18 to be a novel regulator of disease after HSV-2 infection.

Six-week-old female C57BL/6J mice were purchased from Jackson Laboratories and rested for at least 1 week and infected at a minimum of 7 weeks of age. Ncf2 KO mice and controls were provided by M.C. Dinauer (Washington University, St Louis) and generated as previously described (Jacob et al., 2017). Stim1^fl/fl x Stim2^fl/fl x S100a8-Cre mice were provided by G.A. Clemens (Washington University, St Louis) and were generated as previously described (Clemens et al., 2017). Ifnar1^fl/fl mice (Ifnar1^tm1Uka) were a gift from H.W. Virgin (Kamphuis et al., 2006; Nice et al., 2016). Padi4^fl/fl mice (B6(Cg)-Padi4^tm1.2Kmow/J) and S100a8-Cre (B6.Cg-Tg(S100a8-cre,-EGFP)1Ilw/J) were obtained from Jackson Laboratories and bred at Washington University School of Medicine. Cre- littermates generated from breeding pairs were used as controls. All mice were maintained on a 12 hr light/dark cycle with unlimited access to food and water. This study was carried out in accordance with the recommendations in the Guide for the Car and Use of Laboratory Animals of the National Institutes of Health.

Vero Cells (African green monkey kidney epithelial cells, ATCC) were cultured in Dulbeco’s Modified Eagle Medium (Gibco) containing 1% fetal bovine serum (FBS, Corning) and maintained at 37°C with 5% CO₂. Cells were regularly tested for mycoplasma contamination, and all cells used for this study were mycoplasma-free. Primary neutrophils were isolated from the bone marrow (BM) of naive female C57BL/6J mice. A Histopaque gradient was used to isolate primary neutrophils for ROS assays, while a Percoll gradient was used for NET assays. For Histopaque isolation: 3 ml of Histopaque 1119 (Sigma-Aldrich) was overlaid with 3 ml of Histopaque 1077 (Sigma-Aldrich). A single-cell suspension of isolated BM cells in 1 ml of PBS was layered over the Histopaque gradient. Cells were centrifuged for 30 min at room temperature (RT), and neutrophils were collected from the bottom interface. For Percoll isolation: BM cells were resuspended in HBSS (Gibco) with 20 mM HEPES (Gibco) and layered over 6 ml of 62% Percoll solution (GE Healthcare). Cells were centrifuged for 30 min at RT, and neutrophils were collected from the bottom of the tube. All tissue culture experiments were performed under BSL2 containment.

WT HSV-2 186 syn+ (Spang et al., 1983) and HSV-1 McKrae (Williams et al., 1965) were propagated and titered on Vero cells as previously described (Lee et al., 2020). Briefly, for propagation of virus stocks, Vero cells were plated in T150 tissue culture flasks, inoculated at 0.01 MOI at 80% confluence, and incubated at 37°C. Infected cells were harvested 2–3 days after infection, resuspended in equal volumes of virus supernatant and twice-autoclaved milk, and sonicated. Lysed cells were aliquoted and used as viral stock. To titer, Vero cells were plated in six-well plates and inoculated with 10-fold serial dilutions of stock virus. After inoculation, overlay media with 20 μg/ml human IgG was added to each well and plates were incubated at 37°C for 2–3 days. To count, Vero cells were stained with 0.1% crystal violet. All tissue culture experiments were performed under BSL2 containment. For titration of virus in the vaginal lumen, 50 ul washes with sterile PBS were collected using a pipette and a sterile calginate swab, and diluted in 950 ul of ABC buffer (0.5 mM CaCl₂, 0.5 mM MgCl₂, 1% glucose, and 1% FBS in sterile PBS). For titration of virus from tissue, vaginas were harvested into pre-weighed tubes and flash frozen on dry ice. ABC buffer was added to weighed tissues, which were bead-homogenized and clarified by centrifugation. 10-fold serial dilutions of vaginal washes or tissue homgenate were titered by plaque assays on Vero cells (Lee et al., 2020).

All mice were injected subcutaneously in the neck ruff once with 2 mg of DMPA (Depo-Provera, Pfizer) 5–7 days prior to virus inoculation. For experiments in which neutrophils were depleted, mice were i.p. injected once with 500 μg of anti-Ly6G (clone 1A8) or rat IgG2a isotype control (anti-trinitrophenol +KLH) (Leinco Technologies) diluted in sterile PBS (Sigma-Aldrich) 1 day prior to inoculation. For experiments in which IFNAR blockade was conducted, mice were i.p. injected once with 1 mg of anti-IFNAR1 (clone MAR1-5A3) or mouse IgG1 isotype control (clone HKSP) (Leinco Technologies) on either the day of inoculation (‘early’) or at 4 d.p.i. (‘late’). For ‘late’ treatments, only mice without overt signs of genital inflammation were chosen for antibody injection in both anti-IFNAR and isotype control groups to avoid biasing of results. For experiments in which IL-18 was neutralized, mice were treated intravaginally with 100 μg of anti-IL-18 antibody (clone YIGIF74-1G7) or rat IgG2a isotype control (clone 2A3) (BioXCell) on days 3–5 after infection. Selection of mice for isotype control or experimental antibody treatment was random. For intravaginal inoculation, a sterile calginate swab (McKesson) moistened with sterile PBS was used to gently disrupt mucous from the vaginal cavity. Stock virus was diluted in sterile PBS and either 5000 PFU or 10⁴ PFU virus was delivered into the vaginal cavity via pipette tip in a 10 μl volume. Mice were weighed and monitored for signs of disease for 1 week following infection in an unblinded manner and monitored for survival for 2 weeks. Genital inflammation was scored as follows: 0 – no inflammation, 1 – mild redness and swelling around the vaginal opening, 2 – fur loss and visible ulceration, 3 – severe ulceration and mild signs of sickness behavior (lack of grooming), 4 –hindlimb paralysis, and 5 – moribund.

All tissues were harvested from animals sedated with ketamine and xylazine and thoroughly perfused with a minimum of 15 ml of PBS. Vaginas were processed as follows: tissue was cut into pieces and digested for 15 min in a shaking water bath held at 37°C in a 0.5 mg/ml solution of Dispase II (Roche) in PBS. Tissues were then transferred to a solution of 0.5 mg/ml Collagenase D (Roche) and 15 μg/ml DNase I (Roche) in RPMI media (Gibco) supplemented with 10% FBS (Corning) and 1% pen/strep (Gibco) and digested for 25 min in a shaking water bath held at 37°C. 50 μl of sterile EDTA was added to each sample and incubated at 37°C for another 5 min. Tissues were then mechanically disrupted through a 70 um cell strainer into a single-cell suspension using a 3 ml syringe plunger. Tissues were washed with RPMI media with 1% FBS, centrifuged, and resuspended in 200 μl RPRM with 1% FBS and 1% pen/strep.

Single-cell suspensions from vaginal tissues, or luminal cells collected in vaginal washes were plated in 96-well plates and incubated with Live/Dead Fixable Aqua Dead Cell Stain kit (Molecular Probes) for 15 min at room temperature (RT) in the dark. Cells were then incubated with Fc block (anti-CD16/32, Biolegend) for 15 min at RT in the dark. Surface staining was performed in FACS buffer (1% FBS and 0.02% sodium azide in PBS) on ice and in the dark using the following antibodies: CD3 (clone 145–2 C11), CD4 (clone GK1.5), CD8a (clone 53–6.7), CD11b (clone M1/70), CD45 (clone 30-F11), Gr-1 (clone RB6-8C5), Ly6C (clone HK1.4), Ly6G (clone 1A8), and NK1.1 (clone PK136). All antibodies were purchased from Biolegend. For surface staining of IFNAR1, cells were incubated with an anti-IFNAR1 antibody or a mouse IgG1 isotype control (Leinco Technologies) for 20 min at 37°C. Cells were washed and then surface staining of other markers proceeded as described above. Cell counts were performed by adding Precision Count Beads (Biolegend) to samples prior to flow cytometric acquisition. Samples were acquired on an LSR Fortessa (BD Biosciences) and analyzed by FlowJo (Treestar).

All tissues were harvested from animals sedated with ketamine and xylazine and thoroughly perfused with a minimum of 15 ml of PBS, followed by 15 ml of PLP fixative (0.01 M NaIO₄, 0.075 M lysine, 0.0375 M sodium phosphate, and 2% paraformaldehyde [PFA]) for immunofluorescent (IF) staining or 4% PFA for immunohistochemistry (IHC). Tissues were cryoprotected in 30% sucrose, frozen in OCT medium (Sakura), and cut into 7 um sections. Cryosections were blocked 5% bovine serum albumin (BSA), 5% goat serum (Jackson Immunoresearch), and 0.1% Triton-X in PBS for 1 hr at RT. HSV antigens were detected with a rabbit anti-HSV primary antibody (Dako), incubated overnight at 4°C, washed in PBS, and incubated for 1 hr at RT with a goat anti-rabbit IgG conjugated to AlexaFluor 488 (Life Technologies). S100A8 was detected with a rat anti-mouse S100A8 primary antibody (clone 63N13G5, Novus Biologicals) and a goat anti-rat IgG conjugated to AlexaFluor 568 (Life Technologies) in a similar manner. IL-18 was detected using a biotinylated rat anti-mouse IL-18 primary antibody (clone 93–10C, MBL International). Cryosections were blocked as described above and then treated with the Avidin/Biotin Blocking Kit (Vector Laboratories) according to manufacturer's protocol. Endogenous peroxidases were quenched with a 2% hydrogen peroxide solution. Anti-mouse IL-18 or a rat IgG1 isotype control was incubated overnight at 4°C. The AlexaFluor 647 Tyramide Signal Amplification kit (Invitrogen) was used to visualize IL-18 and used according to manufacturer's protocol. DNA was visualized with 4′,6-diamidino-2-phenylindole (DAPI) (Life Technologies). Sections were imaged with a Zeiss Cell Observer inverted microscope using a 40x objective, acquired with Zen software, and image brightness was adjusted using Photoshop (Adobe). For IHC, sections were probed with an anti-HSV antibody incubated overnight at 4°C (Dako), a donkey anti-rabbit IgG-HRP antibody (Jackson Immunoresearch) for 1 hr at RT and then enzymatically visualized by 3,3’-diaminobenzidine (DAB) enzyme reaction (Sigma-Aldrich). Sections were counterstained with hematoxylin and eosin, and images were captured using Zeiss ZEN software on a Zeiss Cell Observer inverted microscope witβh an Axiocam dual B/W and color camera with a 20x objective. Image brightness was adjusted using Photoshop (Adobe) and merged with Image J64 (NIH).

Harvested tissues were homogenized in RLT buffer (RNeasy Kit, Qiagen) with approximately 100 μl of sterile 1.0 mm zirconia/silica beads (Biospec Products) in a bead beater. Homogenized tissue samples were processed according to manufacturer's protocol using the RNeasy Mini Kit (Qiagen) and RNA quality and quantity was assessed on a Nanodrop (ThermoFisher). qRT-PCR was performed in 10 μl reactions using the iTaq Universal SYBR Green One-Step kit (Biorad) according to manufacturer's protocol.

Single-cell suspensions from digested vaginas were stained with Live/Dead Fixable Aqua Dead Cell Stain kit (Molecular Probes) for 15 min at RT in the dark. Live cells were sorted on BD FacsAria II housed in a BSL2 biosafety cabinet. A minimum of 16,000 cells were resuspended in PBS with 2% FBS and 0.2 U/μl RNase inhibitor at a concentration of 800–1400 cells/μl, submitted to McDonnell Genome Institute, and prepared for droplet-based 3' end scRNA-seq using the Chromium 3' v3 single-cell reagent kit as per manufacturer's protocol (10x Genomics). Library sequencing was performed on a NovaSeq S4 (Illumina).

For cytokine analysis by Bio-Plex Pro Mouse Cytokine 23-Plex Immunoassay (Bio-rad): 2 × 50 μl washes with sterile PBS were collected from the vaginal lumen using a pipette. Samples were centrifuged for 3 min at 13000*g to remove mucous and cells, and supernatants were added to 200 μl of ABC buffer. The assay was performed according to manufacturer instructions, and plates were read on a Luminex Bioplex 100 system (Biorad). For measurement of IL-18 or IFNβ, 2 × 50 μl washes with sterile PBS were collected from the vaginal lumen and centrifuged to remove mucous and cells. IL-18 was measured using the mouse IL-18 ELISA kit (MBL International) according to manufacturer's instructions at half-volumes, while IFNβ was measured using the LEGEND MAX Mouse IFNβ ELISA kit (Biolegend) according to manufacturer's instructions. For the measurement of IFNβ in tissue homogenates, vaginal tissues were collected in pre-weighed tubes and flash frozen on dry ice. NP40 lysis buffer (150 mM NaCl, 50 mM Tris, pH 8.0, 1% NP40 alternative) with protease inhibitor cocktail (Sigma Aldrich) was added to weighed tissue and bead-homogenized. Lysates were clarified by centrifugation and supernatants used for ELISA.

To measure ROS production, isolated neutrophils were stimulated with heat-killed HSV-2 (56°C for 30 min) at an MOI of 5 for 16 hr at 37°C. ROS levels were quantified using DCFDA Cellular ROS Detection Assay kit (Abcam) according to manufacturer's protocol. Fluorescence levels were measured by flow cytometry. To induce NET formation, neutrophils were stimulated with heat-killed HSV-2 at an MOI of 1 for 4 hr at 37°C. Cells were fixed with 8% PFA overnight and probed with a polyclonal rabbit antibody against mouse citrullinated histone H3 (Abcam) for 1 hr at RT in 1% BSA and 0.1% Triton-X for 1 hr at RT, a goat anti-rabbit antibody conjugated to AlexaFluor 488 (Life Technologies) for 1 hr at RT and DAPI diluted in PBS for 6 min at RT. Cells were imaged with a Zeiss Cell Observer inverted microscope using a 63x objective and image brightness was adjusted using Photoshop (Adobe).

All numerical data analyses except for scRNA-seq data analysis were performed on Graphpad Prism8 software. Values were log-transformed to normalize distribution and variances where necessary. Immune cell numbers and cytokine measurement were analyzed by two-way ANOVA with Bonferroni's multiple comparisons test. Log-transformed viral titers were analyzed by repeated measures two-way ANOVA with Bonferroni's multiple comparisons test. Inflammation scores were analyzed by repeated-measures two-way ANOVA or mixed-effects analysis with Geisser-Greenhouse correction and Bonferroni's multiple comparisons test. The Geisser-Greenhouse correction was used for inflammation scores to correct any violations of sphericity and to provide a more restrictive, stringent calculation of p values. ROS MFI was measured by unpaired two-tailed Student's t-test. qPCR results were analyzed by one-way ANOVA with Tukey's multiple comparisons test. A p<0.05 was considered statistically significant. No experimental data points were excluded from statistical analysis, including potential outliers. Mouse and sample numbers per group and experimental repeat information is provided in the figure legends. All data points represent individual biological replicates, and the 'n' for each group refers to biological replicates. No power calculations were performed to determine sample size; rather sample sizes were determined based on historical data.