Impact of Herpes Simplex Virus Type 2 and Human Immunodeficiency Virus Dual Infection on Female Genital Tract Mucosal Immunity and the Vaginal Microbiome

Abstract

Background: Mechanisms linking herpes simplex virus type 2 (HSV-2) with human immunodeficiency virus (HIV) are not fully defined. We tested the hypothesis that HSV-2 and HIV dual infection is associated with cervicovaginal inflammation and/or vaginal dysbiosis.

Methods: Genital tract samples were obtained weekly over a 12-week period from 30 women seropositive (+) for HIV and HSV-2 and 15 women each who were seropositive for one or seronegative (-) for both viruses. Immune mediators, antimicrobial activity, and microbial composition and diversity were compared.

Results: Significant differences in the concentrations of interferon-γ (P = .002), tumor necrosis factor-α (P = .03), human beta defensin 1 (P = .001), secretory leukocyte protease inhibitor (P = .01), and lysozyme (P = .03) were observed across the 4 groups (Kruskal-Wallis). There were also significant differences in vaginal microbial alpha diversity (Simpson index) (P = .0046). Specifically, when comparing HIV-1+/HSV-2+ to HIV-1-/HSV-2- women, a decrease in Lactobacillus crispatus and increase in diverse anaerobes was observed. The number of genital HSV outbreaks was greater in HIV+ versus HIV- women (39 versus 12) (P = .04), but there were no significant differences when comparing outbreak to non-outbreak visits.

Conclusions: Increased microbial diversity and cervicovaginal inflammation in HIV and HSV-2 dually infected women may adversely impact genital health and, in the absence of antiretroviral therapy, facilitate HIV shedding.

Keywords: herpes simplex virus; human immunodeficiency virus; microbiome; mucosal immunity; vaginal dysbiosis.

Figure 1.

Differences in antimicrobial activity of cervicovaginal lavage (CVL) in human immunodeficiency virus (HIV) and/or herpes simplex virus type 2 (HSV-2) seropositive women. Scatter dot plots showing the percentage inhibition of (A) HIV infection of TZM-bl cells, (B) HSV-2 infection of Vero cells, and (C) Escherichia coli growth in the presence of cervicovaginal lavage fluid relative to cells treated with control buffer. Values greater than 100% are indicative of enhancement of infection. Each individual point indicates results from duplicate wells for a single participant; the bar indicates the mean value for the group. The asterisk indicates P = .0049 comparing anti-HIV activity of CVL in HIV⁺/HSV-2⁺ versus HIV⁻/HSV-2⁻ controls (Kruskal-Wallis with Dunn’s multiple comparisons).

Figure 2.

Vaginal microbiome differs in dually infected women. Stacked bar plots showing vaginal microbial community profiles from samples within the 4 groups human immunodeficiency virus (HIV)⁻/herpes simplex virus type 2⁻ (HSV-2⁻), HIV⁻/HSV-2⁺, HIV⁺/HSV-2⁻, and HIV⁺/HSV-2⁺ in the cross-sectional cohort. The color legend only shows selected taxons with a total of 197 species plotted. Samples are grouped by the most abundant species, and those within each community group are clustered on bray distances using Ward’s method. The distances were calculated using vegan R package and the figure was prepared using ggplot2.

Figure 3.

Alpha diversity is significantly higher in human immunodeficiency virus (HIV)⁺/herpes simplex virus type 2⁺ (HSV-2⁺) women. Boxplot showing differences in alpha diversity measures within each group of the cross-sectional cohort. Alpha diversity was measured as the inverse Simpson index, which considers the richness as well as the abundance of species within the sample. The diversity is significantly different among the groups (Kruskal-Wallis, P = .0046), with post hoc pairwise comparisons indicating significantly higher diversity in the HIV⁺/HSV-2⁺ group compared with HIV⁻/HSV-2⁻ and HIV⁻/HSV-2⁺ groups (P = .02 and P = .03, respectively, false discovery rate adjustments).

Figure 4.

Statistical association analysis using linear discriminant analysis effect size (LefSe). Bacterial species with significantly differential abundance between dually infected human immunodeficiency virus (HIV)⁺/herpes simplex virus type 2⁺ (HSV-2⁺) and HIV⁻/HSV-2⁻ women (A), and HIV⁻/HSV-2⁺ women (B) were identified using LEfSe. Features with linear discriminant analysis (LDA) score greater than 3.0 are shown. The inner boxplots show relative abundance of select species within groups compared by LEfSe.