Bacterial vaginosis (BV) is highly prevalent amongst women residing in sub-Saharan Africa, where HIV is also widespread. BV, characterised by the overgrowth of non-optimal vaginal anaerobic bacteria, most commonly Gardnerella vaginalis, is associated with genital inflammation which facilitates HIV acquisition. However, G. vaginalis is also found in healthy women with low levels of inflammation, suggesting strain-level differences that may influence inflammatory responses. Hence, this study aims to characterize vaginal G. vaginalis isolates from South African women to elucidate their role in inflammation.
Cervicovaginal fluid samples from 10 BV-negative and 10 BV-positive South African women (aged 16 – 25) were cultured on Columbia Blood agar to isolate single strains of G. vaginalis, followed by species-level identification via 16S rRNA Sanger sequencing. Isolates were then co-cultured with vaginal epithelial VK2/E6E7 cells and the concentrations of inflammatory cytokines previously associated with HIV risk were measured via Luminex assay. The protein profiles of the isolates were also analysed via liquid chromatography tandem mass spectrometry.
Thirty-nine isolates were acquired from all women, including n = 15 G. vaginalis and other bacterial taxa. Three and four G. vaginalis isolates from different BV-negative and BV-positive women, respectively, were selected for inflammatory profile assessment. Significant increases in interleukin (IL)-1β, IL-6, IL-8 and chemokine ligand (CCL)2, CCL4 and CCL5 were induced by isolates from BV-positive women compared to BV-negative women (p < 0.05). Proteomics analyses detected 2,139 proteins and although 125 proteins were significantly differentially abundant between isolates from BV-negative versus BV-positive women, only two remained significant after adjusting for multiple comparisons. This included chaperone protein ClpB and peptidase.
G. vaginalis isolates from BV-positive women elicited higher levels of inflammatory cytokines previously associated with increased HIV acquisition risk compared to those from BV-negative women. This suggests that strain-level differences may play an important role in genital inflammation and resultant HIV acquisition risk. Few proteins differed significantly between isolates from BV-negative versus BV-positive women, largely due to the small sample size and high protein variance between individual strains. Hence, the proteomic and genomic profiles of additional G. vaginalis isolates are being investigated to determine differences that may influence inflammation.