Bacterial vaginosis associates with dysfunctional T cells and altered soluble immune factors in the cervicovaginal tract

Abstract

BACKGROUNDBacterial vaginosis (BV) is a dysbiosis of the vaginal microbiome that is prevalent among reproductive-age females worldwide. Adverse health outcomes associated with BV include an increased risk of sexually acquired HIV, yet the immunological mechanisms underlying this association are not well understood.METHODSTo investigate BV-driven changes to cervicovaginal tract (CVT) and circulating T cell phenotypes, Kinga Study participants with or without BV provided vaginal tract (VT) and ectocervical (CX) tissue biopsies and PBMC samples.RESULTSHigh-parameter flow cytometry revealed an increased frequency of cervical CD4+ conventional T (Tconv) cells expressing CCR5 in BR+ versus BR- women. However, we found no difference in the number of CD3+CD4+CCR5+ cells in the CX or VT of BV+ versus BV- individuals, suggesting that BV-driven increased HIV susceptibility may not be solely attributed to increased CVT HIV target cell abundance. Flow cytometry also revealed that individuals with BV had an increased frequency of dysfunctional CX and VT CD39+ Tconv and CX tissue-resident CD69+CD103+ Tconv cells, reported to be implicated in HIV acquisition risk and replication. Many soluble immune factor differences in the CVT further support that BV elicits diverse and complex CVT immune alterations.CONCLUSIONOur comprehensive analysis expands on potential immunological mechanisms that may underlie the adverse health outcomes associated with BV, including increased HIV susceptibility.TRIAL REGISTRATIONClinicalTrials.gov NCT03701802.FUNDINGThis work was supported by National Institutes of Health grants R01AI131914, R01AI141435, and R01AI129715.

Keywords: AIDS/HIV; Immunology; Infectious disease; T cells.

Figure 1. BV did not alter the balance of T cell frequency in the CVT tissues.

Flow cytometry was used to quantify the proportions of T cells within different tissue sites as indicated. (A) The frequency of CD3⁺ (CX BV^– N = 147, BV⁺ N = 37; VT BV^– N = 144, BV⁺ N = 40; PBMC BV^– N = 143, BV⁺ N = 42) among total CD45⁺ cells. (B and C) The frequency of Tconv (CX BV^– N = 129, BV⁺ N = 29; VT BV^– N = 122, BV⁺ N = 32; PBMC BV^– N = 143, BV⁺ N = 42) (B) and CD3⁺CD8⁺ (CX BV^– N = 141, BV⁺ N = 33; VT BV^– N = 128, BV⁺ N = 36; PBMC BV^– N = 143, BV⁺ N = 42) (C) among total CD45⁺ cells. (D and E) The frequency of Tconv (CX BV^– N = 129, BV⁺ N = 29; VT BV^– N = 122, BV⁺ N = 32; PBMC BV^– N = 143, BV⁺ N = 42) (D) and CD8⁺ (CX BV^– N = 141, BV⁺ N = 33; VT BV^– N = 130, BV⁺ N = 36; PBMC BV^– N = 143, BV⁺ N = 42) (E) among total CD3⁺ cells. Adjusted rank regression analysis was performed to compare frequencies in each tissue between BV^– and BV⁺ individuals. PBMC comparisons were a priori adjusted for hormonal contraceptive use, and CX and VT comparisons were a priori adjusted for hormonal contraceptive use, HSV-2 serology, HIV exposure, and semen exposure, to reduce the effects of potential confounding variables on the analysis of BV-driven T cell alterations. Comparisons with adjusted P value > 0.10 labeled as not significant. Each dot represents a measurement from an individual sample. Each horizontal bar indicates the median for its respective group. Additional statistical information for each comparison is provided in Supplemental Table 1.

Figure 2. Conventional CD4⁺ T cells displayed increased markers of activation and tissue residency in the cervix of individuals with BV.

Flow cytometry was used to examine Tconv cell phenotypes within different tissue sites as indicated. (A–C) The frequency of CCR5⁺ (A), HLA-DR⁺ (B), and CD69⁺CD103⁺ (C) among total Tconv cells (CX BV^– N = 128, BV⁺ N = 29; VT BV^– N = 122, BV⁺ N = 30; PBMC BV^– N = 143, BV⁺ N = 42). Adjusted rank regression analysis was performed to compare frequencies in each tissue between BV^– and BV⁺ individuals. PBMC comparisons were a priori adjusted for hormonal contraceptive use, and CX and VT comparisons were a priori adjusted for hormonal contraceptive use, HSV-2 serology, HIV exposure, and semen exposure, to reduce the effects of potential confounding variables on the analysis of BV-driven T cell alterations. Adjusted P value displayed for all P_adj ≤ 0.10; all other comparisons labeled as not significant. Each dot represents a measurement from an individual sample. Each horizontal bar indicates the median for its respective group. Additional statistical information for each comparison is provided in Supplemental Table 1.

Figure 3. The total density of T cells and HIV target cells in the cervix and vagina was not altered by BV.

(A) Representative H&E-stained VT tissue section imaged with bright-field microscopy at ×20 magnification. (B) Representative immunofluorescently stained tissue section from the same VT biopsy as A. DAPI stain is shown in blue, CD3 stain in red, CD4 stain in cyan, and CCR5 stain in green. All fluorescent signals are overlaid. Scale bar: 100 μm. (C and D) Comparison of the density of CD3⁺ T cells, CD3⁺CD4⁺ T cells, or CD3⁺CD4⁺CD5⁺ HIV target cells in the CX (BV^– N = 44, BV⁺ N = 11) (C) and VT (BV^– N = 55, BV⁺ N = 12) (D). Wilcoxon’s rank sum test was performed for each comparison shown. Comparisons with P > 0.05 labeled as not significant. Each dot represents a measurement from an individual sample. Each horizontal bar indicates the median for its respective group. Additional statistical information for each comparison is provided in Supplemental Table 2.

Figure 4. Conventional CD4⁺ T cells in the cervix showed signs of dysfunction in individuals with BV.

Flow cytometry was used to examine Tconv cell phenotypes within different tissue sites as indicated. (A–C) The frequency of CD39⁺ (A), CD101⁺ (B), and T cell factor-1⁺ (TCF-1⁺) (C) among total Tconv cells (CX BV^– N = 128, BV⁺ N = 29; VT BV^– N = 122, BV⁺ N = 30; PBMC BV^– N = 143, BV⁺ N = 42). Adjusted rank regression analysis was performed to compare frequencies in each tissue between BV^– and BV⁺ individuals. PBMC comparisons were a priori adjusted for hormonal contraceptive use, and CX and VT comparisons were a priori adjusted for hormonal contraceptive use, HSV-2 serology, HIV exposure, and semen exposure, to reduce the effects of potential confounding variables on the analysis of BV-driven T cell alterations. Adjusted P value displayed for all P_adj ≤ 0.10; all other comparisons labeled as not significant. Each dot represents a measurement from an individual sample. Each horizontal bar indicates the median for its respective group. Additional statistical information for each comparison is provided in Supplemental Table 1.

Figure 5. Th17 cells exhibited increased markers of activation and tissue residency in cervical samples from individuals with BV.

Flow cytometry was used to examine Th17 phenotypes within different tissue sites as indicated. (A) The frequency of Th17 cells, defined as CD161⁺CCR6⁺ Tconv, among total Tconv cells (CX BV^– N = 128, BV⁺ N = 29; VT BV^– N = 122, BV⁺ N = 30; PBMC BV^– N = 143, BV⁺ N = 42). (B–E) HLA-DR⁺ (B), CD69⁺CD103⁺ (C), CD101⁺ (D), and CCR5⁺ (E) frequencies among total Th17 cells (CX BV^– N = 42, BV⁺ N = 9; VT BV^– N = 32, BV⁺ N = 9; PBMC BV^– N = 143, BV⁺ N = 42). Adjusted rank regression analysis was performed to compare frequencies in each tissue between BV^– and BV⁺ individuals. PBMC comparisons were a priori adjusted for hormonal contraceptive use, and CX and VT comparisons were a priori adjusted for hormonal contraceptive use, HSV-2 serology, HIV exposure, and semen exposure, to reduce the effects of potential confounding variables on the analysis of BV-driven T cell alterations. Adjusted P value displayed for all P_adj ≤ 0.10; all other comparisons labeled as not significant. Each dot represents a measurement from an individual sample. Each horizontal bar indicates the median for its respective group. Additional statistical information for each comparison is provided in Supplemental Table 1.

Figure 6. Cervical CD8⁺ T cells displayed a dysfunctional phenotype in individuals with BV.

Flow cytometry was used to examine CD8⁺ T cell phenotypes within different tissue sites as indicated. (A–C) The frequency of CD39⁺ (A), granzyme B⁺ (B), and T-bet⁺ (C) among total CD8⁺ T cells (CX BV^– N = 127, BV⁺ N = 29; VT BV^– N = 116, BV⁺ N = 31; PBMC BV^– N = 143, BV⁺ N = 42). Adjusted rank regression analysis was performed to compare frequencies in each tissue between BV^– and BV⁺ individuals. PBMC comparisons were a priori adjusted for hormonal contraceptive use, and CX and VT comparisons were a priori adjusted for hormonal contraceptive use, HSV-2 serology, HIV exposure, and semen exposure, to reduce the effects of potential confounding variables on the analysis of BV-driven T cell alterations. Adjusted P value displayed for all P_adj ≤ 0.10; all other comparisons labeled as not significant. Each dot represents a measurement from an individual sample. Each horizontal bar indicates the median for its respective group. Additional statistical information for each comparison is provided in Supplemental Table 1.

Figure 7. BV was associated with reduced chemokine concentrations and increased inflammatory cytokine concentrations in CVT fluid.

Luminex was used to quantify cytokines and chemokines from CVT fluid (BV^– N = 134, BV⁺ N = 36 for IL-4; BV^– N = 135, BV⁺ N = 37 for IL-2; BV^– N = 136, BV⁺ N = 37 for LIF and TARC; BV^– N = 136, BV⁺ N = 38 for all other cytokine/chemokine comparisons) (A) and serum (BV^– N = 149, BV⁺ N = 44) (B). Adjusted estimated mean difference (BV⁺ – BV^–) for CVT fluid cytokines/chemokines and serum cytokines/chemokines that met the criteria for quantification of cytokine/chemokine concentrations (≥80% of samples were detectable). The adjusted 95% confidence interval is shown for all comparisons. Significant results when P_adj ≤ 0.05 comparing BV^– versus BV⁺ are colored in orange, and non-significant differences (P > 0.05) comparing BV^– versus BV⁺ are purple. Vertical dashed line at x = 0 for reference. Serum comparisons were a priori adjusted for hormonal contraceptive use, and CVT fluid comparisons were a priori adjusted for hormonal contraceptive use, HSV-2 serology, HIV exposure, and semen exposure, to reduce the effects of potential confounding variables on the analysis of BV-driven T cell alterations. Additional statistical information for each comparison is provided in Supplemental Table 3.