Cultivated Human Vaginal Microbiome Communities Impact Zika and Herpes Simplex Virus Replication in ex vivo Vaginal Mucosal Cultures

Abstract

The human vaginal microbiome (VMB) is a complex bacterial community that interacts closely with vaginal epithelial cells (VECs) impacting the mucosal phenotype and its responses to pathogenic insults. The VMB and VEC relationship includes nutrient exchange and regulation of signaling molecules that controls numerous host functions and defends against invading pathogens. To better understand infection and replication of sexually transmitted viral pathogens in the human vaginal mucosa we used our ex vivo VEC multilayer culture system. We tested the hypothesis that selected VMB communities could be identified that alter the replication of sexually transmitted viruses consistent with reported clinical associations. Sterile VEC multilayer cultures or those colonized with VMB dominated by specific Lactobacillus spp., or VMB lacking lactobacilli, were infected with Zika virus, (ZIKV) a single stranded RNA virus, or Herpes Simplex Virus type 2 (HSV-2), a double stranded DNA virus. The virus was added to the apical surface of the cultured VEC multilayer to model transmission during vaginal intercourse. Viral replication was measured 48 h later by qPCR. The results indicated that VEC cultures colonized by VMB containing Staphylococcus spp., previously reported as inflammatory, significantly reduced the quantity of viral genomes produced by ZIKV. HSV-2 titers were decreased by nearly every VMB tested relative to the sterile control, although Lactobacillus spp.-dominated VMBs caused the greatest reduction in HSV-2 titer consistent with clinical observations. To explore the mechanism for reduced ZIKV titers, we investigated inflammation created by ZIKV infection, VMB colonization or pre-exposure to selected TLR agonists. Finally, expression levels of human beta defensins 1-3 were quantified in cultures infected by ZIKV and those colonized by VMBs that impacted ZIKV titers. Human beta defensins 1-3 produced by the VEC showed no association with ZIKV titers. The data presented expands the utility of this ex vivo model system providing controlled and reproducible methods to study the VMB impact on STIs and indicated an association between viral replication and specific bacterial species within the VMB.

Keywords: Zika virus; herpes simplex virus type 2; vaginal microbiome; vaginal mucosa; women’s health.

FIGURE 1

ZIKV replicated in VEC. Immunofluorescent labeling showed the presence of ZIKV (Panamanian) in VEC monolayer cell culture at 48 h post infection (32×). Red = cytokeratin, Blue = cell nucleus (DAPI), Green = ZIKV. (A) ZIKV infected cells. (B) Negative control. (C) Replication kinetics of ZIKV (Panamanian and Dakar) in the mature multilayer VEC culture system (N = 6) (average genomes are shown as log₁₀ values ±SD). ZIKV in the basal chamber modeled systemic release of the virus from the vaginal mucosa.

FIGURE 2

ZIKV infection did not impact the abundance of selected vaginal bacteria. The graph shows the average log₁₀ difference (±SD) of the absolute quantity of individual species detected with and without ZIKV infection in all microbiomes tested. There were no significant changes in the quantity of the indicated organism after ZIKV infection. Lactobacillus spp., N = 28; L. crispatus, N = 11; L. gasseri, N = 7; L. jensenii, N = 10; L. iners, N = 10; G. vaginalis, N = 25; G. vaginalis sialidase, N = 16; Aerococcus spp., N = 11; A. vaginae, N = 8; Enterococcus spp., N = 24; Streptococcus spp., N = 26. Unpaired Student’s t-test analyses were used to compare the VMB colonized control well to the ZIKV infected VMB colonized well.

FIGURE 3

VMB bacterial profiles. The community profiles (average of six replicate cultures) are shown as a relative abundance proportional bar chart compared to the total 16S detected. Each profile was developed using a customized qPCR array. Each target is represented as a different color as shown in the legend.

FIGURE 4

VMBs differently impacted replication of ZIKV and HSV-2. Average ZIKV (A) and HSV-2 (B) titers (±SD) 48 h post infection in parallel cultures colonized with one of 11 different VMBs (1–11) or a monoculture of S. epidermidis (S. epi). Sterile controls were not colonized with a VMB. The horizontal line shows the average ZIKV titer in sterile (no microbiome) VEC. A one-way ANOVA was performed comparing each condition to the sterile control (^∗p ≤ 0.05, ^∗∗p ≤ 0.005).

FIGURE 5

Changes in cytokine production by VEC when colonized with specific VMBs that significantly impacted ZIKV replication. Cytokines were quantitated using a multiplex bead array assay. (A) GAPDH levels (established by qPCR) indicated there were no significant impacts on cell counts for the conditions tested. Average levels of selected cytokines (N = 6 replicates; ±SD) are illustrated for IL-8 (B); IL-22 (C); PDGF-BB (D); RANTES (E) or IL-1alpha (F). Two-way ANOVA analyses between conditions determined significance; ^∗p-value < 0.05, ^∗∗p-value < 0.01, ^∗∗∗p-value < 0.001, ^∗∗∗∗p-value < 0.0001. As a reminder, VMB 6 significantly increased ZIKV replication while VMBs 10 and 11 significantly decreased ZIKV replication.