Distinct mucosal and systemic immunological characteristics in transgender women potentially relating to HIV acquisition

Abstract

Transgender women (TGW) are disproportionally affected by HIV infection, with a global estimated prevalence of 19.9%, often attributed to behavioral risk factors, with less known about biological factors. We evaluated potential biological risk factors for HIV acquisition in TGW at the sites of viral entry by assessing immune parameters of the neovaginal surface and gut mucosa. The neovagina in TGW, compared with the vagina in cisgender women (CW), shows distinct cell composition and may pose a more inflammatory environment, evidenced by increased CD4+ T cell activation and higher levels of soluble markers of inflammation (C-reactive protein, soluble CD30). Increased inflammation may be driven by microbiome composition, as shown by a greater abundance of Prevotella and a higher Shannon Diversity Index. In addition, we have observed higher frequency of CD4+CCR5+ target cells and decreased DNA methylation of the CCR5 gene in the gut mucosa of TGW compared with CW and men who have sex with men, which was inversely correlated with testosterone levels. The rectal microbiome composition in TGW appears to favor a proinflammatory milieu as well as mucosal barrier disruption. Thus, it is possible that increased inflammation and higher frequencies of CCR5-expressing target cells at sites of mucosal viral entry may contribute to increased risk of HIV acquisition in TGW, with further validation in larger studies warranted.

Keywords: AIDS/HIV; Cellular immune response; Immunology; Sex hormones; T cells.

Figure 1. Frequency of CD4⁺ and CD8⁺ T cells differs between the vagina and the neovagina.

(A) Example of parent gating strategy of freshly isolated neovaginal mononuclear cells. The staining profile of vaginal mononuclear cells from a representative volunteer is shown, indicating the frequency of activated CD4⁺ and CD8⁺ T cells by the expression of Ki67 and coexpression of CD38 and HLA-DR. (B) Frequency of vaginal (n = 8) and neovaginal (n = 7) CD4⁺ T cells and (C) CD8⁺ T cells, (D) CD4/CD8 T cell ratio, and (E) frequency of CD4⁺CCR5⁺ T cells are shown depending on sample availability. Difference between groups were analyzed using unpaired t test.

Figure 2. Cellular and soluble inflammation profiles differ between the vagina and the neovagina.

Frequencies of cycling CD4⁺ (A) and CD8⁺ (E) T cells are increased in the neovagina (n = 7) compared with the vagina (n = 8), while there is no difference in the frequency of cycling CD4⁺CCR5⁺ (C) T cells observed between the vagina and the neovagina. In contrast, CD4⁺ (B), CD4⁺CCR5⁺ (D), and CD8⁺ (F) T cells in the vagina had a higher activation status compared with the neovagina indicated by the coexpression of HLA-DR and CD38. However, there was also a significant increase in PD-1–expressing CD8⁺ T cells observed in the neovagina (G), which was not seen in CD4⁺ T cells (data not shown). The only soluble biomarkers out of 18 tested that were significantly different between the vagina and the neovagina were the inflammation makers CRP (median: neovaginal 548 pg/mL vs. vaginal 480 pg/mL) (H) and sCD30 (median: neovaginal 2.45 pg/mL vs. vaginal 0.55 pg/mL) (I) that were significantly increased in neovaginal secretions (H). Differences between groups were analyzed using unpaired t tests.

Figure 3. Microbial profiles determined by 16s rRNA sequencing revealed distinct microbial community structures between the vagina and neovagina.

(A) Bar graph of phylum rank of vaginal microbiome in cisgender women (CW; n = 10) and neovaginal microbiome in transgender women (TGW; n = 10), including Shannon Alpha Diversity displayed below bar graphs, suggesting a different microbiome composition between the vaginal and the neovaginal compartment. (B) Neovaginal samples have a unique microbiome profile compared with vaginal samples based on principal component analyses. Brown circles, vaginal; purple circles, neovaginal. (C) Heatmap of relative abundance plot of species in neovaginal/vaginal compartment highlighting differences in compositions between vaginal and neovaginal microbiome.

Figure 4. Distinct microbial community structures in the neovagina correlate with markers of local inflammation.

Neovaginal (n = 10) microbial communities are characterized by a higher Shannon Diversity Index (A), a higher Prevotella abundance (B), and a lower Lactobacillus abundance (C) compared with vaginal (n = 10) microbial communities. The higher Shannon Diversity Index in the neovagina was linked to an increased inflammation profile, indicated by elevated neovaginal CRP and sCD30 levels (D and E). In addition, a direct correlation between the Shannon Diversity Index and the cycling of Ki67-expressing CD4⁺ T cells was observed (F). Differences between groups were analyzed using unpaired t tests. Spearman’s correlation was used to analyze associations between 2 variables. Brown circles, vaginal; purple squares, neovaginal.

Figure 5. Frequency of CD4⁺CCR5⁺ T cells in the sigmoid colon in MSM (n = 10), CW (n = 9), and TGW (n = 3).

The frequency (A) and absolute number (B) of colonic CD4⁺CCR5⁺ T cells was higher in CW and TGW compared with MSM. (C) Altered DNA methylation at loci cg22066626 at regulatory region of CCR5 gene in the gut mucosa in TGW compared with CW and MSM. (D) DNA methylation levels related to CCR5 gene were inversely associated with the frequency of gut mucosa CD4⁺CCR5⁺ T cells. (E) The frequency of colonic CD4⁺CCR5⁺ T cells was indirectly correlated with the plasma testosterone levels. Differences between groups were adjusted for multiple comparisons using Dunn’s test with Benjamini-Hochberg multiple comparison adjustment. Spearman correlation was used to analyze association between 2 variables. Orange triangles, MSM; brown triangles, CW; purple triangles, TGW.

Figure 6. Differences in levels of soluble biomarkers in plasma/rectal secretions and rectal microbial profiles in MSM (plasma/rectal n = 9), CW (plasma n = 9, rectal n = 7), and TGW (plasma n = 9, rectal n = 10).

The levels of I-FABP in (A) plasma and in rectal secretions (B) were increased in TGW compared with MSM and/or CW, indicating increased enterocyte damage. In addition, an increase in plasma IL-1RA (C) and MCP-1 (D) was observed in TGW compared with MSM and/or CW. (E) Bar graph of phylum rank of rectal microbiome in MSM (n = 9), CW (n = 7), and TGW (n = 9), including Shannon Diversity Index, suggesting differences in the level of Fusobacteria and Actinobacteria, with the abundance of Actinobacteria inversely correlating with plasma I-FABP levels (F) and the abundance of Fusobacteria directly correlating with rectal I-FABP (G) and plasma MCP-1 levels (H). Spearman correlation was used to analyze associations between 2 variables. Orange triangles, MSM; brown triangles, CW; purple triangles, TGW.