An exploration of Prevotella-rich microbiomes in HIV and men who have sex with men

Abstract

Background: Gut microbiome characteristics associated with HIV infection are of intense research interest but a deep understanding has been challenged by confounding factors across studied populations. Notably, a Prevotella-rich microbiome described in HIV-infected populations is now understood to be common in men who have sex with men (MSM) regardless of HIV status, but driving factors and potential health implications are unknown.

Results: Here, we further define the MSM-associated gut microbiome and describe compositional differences between the fecal microbiomes of Prevotella-rich MSM and non-MSM that may underlie observed pro-inflammatory properties. Furthermore, we show relatively subtle gut microbiome changes in HIV infection in MSM and women that include an increase in potential pathogens that is ameliorated with antiretroviral therapy (ART). Lastly, using a longitudinal cohort, we describe microbiome changes that happen after ART initiation.

Conclusions: This study provides an in-depth characterization of microbiome differences that occur in a US population infected with HIV and demonstrates the degree to which these differences may be driven by lifestyle factors, ART, and HIV infection itself. Understanding microbiome compositions that occur with sexual behaviors that are high risk for acquiring HIV and untreated and ART-treated HIV infection will guide the investigation of immune and metabolic functional implications to ultimately target the microbiome therapeutically.

Keywords: Antiretroviral therapy (ART); Human immunodeficiency virus (HIV); Men who have sex with men (MSM); Microbiome.

Fig. 1

MSM is more strongly associated with Prevotella richness than HIV infection. a Weighted UniFrac PCoA and b unweighted UniFrac with points colored by orientation and shaped by HIV status. c Genus-level taxonomic summary plot. (Bottom) Genera with mean relative less than 2% abundance are binned together into the category “Other”. Each column represents one individual. (Top) Samples are marked with HIV status and orientation. Each column corresponds to the genus plot below. Samples are ordered by the coordinate on the weighted UniFrac principle coordinate 1

Fig. 2

HIV-negative MSM have significantly different microbiome composition compared to HIV-negative MSW. a MSM have significantly higher alpha diversity than MSW with and without controlling for HIV status (Kruskal-Wallis test; *p< 0.05, **p< 0.01, ***p< 0.001, NS not significant). b Several OTUs are significantly different between HIV-negative Prevotella-rich MSM and HIV-negative Prevotella-rich non-MSM. The significant (Kruskal-Wallis; FDR-corrected p < 0.1) OTUs are displayed as the log2 abundance fold change in MSM compared to non-MSM. More detailed information on the significant OTUs is given in Additional file 9: Table S6

Fig. 3

FFQ diet analysis of MSM compared to non-MSM shows MSM eat significantly few servings of red meat and fruits and less fiber. Reported values are normalized per 1000 kcals. More detail on significant diet components is given in Additional files 14 and 15: Tables S9 and S10 (Kruskal-Wallis test; *p< 0.05)

Fig. 4

HIV-specific changes in women in MSM. Weighted UniFrac PCoA of a MSM only and b women only shows that only half of HIV-positive women cluster apart from HIV-negative women whereas there is no distinctive clustering of HIV in MSM. c Alpha diversity differences in women only (Kruskal-Wallis test; *p< 0.05, NS not significant)

Fig. 5

Microbiome immune correlations. Correlations in HIV-positive, ART-treated (top), and ART-naive (bottom) MSM between OTUs and serum measures of CD4+ T cell count, CD4+/CD8+ CD38+ HLA-DR+ cell percent, and viral load (All OTUs with one or more correlation r > |0.4| are displayed) (Spearman rank test; FDR p value: ^{^}p < 0.2, ⁺p < 0.1, *p < 0.05)

Fig. 6

Longitudinal analysis of ART. a Beta diversity and b change in alpha diversity of HIV-negative individuals compared to HIV-positive pre- and post-ART initiation. c Two taxa observed in the cross-sectional cohort as significantly reduced with HIV and ART have significantly reduced relative abundance longitudinally (Wilcoxon signed-rank test; *p value < 0.05, NS not significant)