No abstract

No abstract

Background Bacterial vaginosis (BV) is a proinflammatory genital condition associated with adverse reproductive health outcomes, including increased HIV incidence. However, BV recurrence rates are high following standard antibiotic treatment. While the composition of the vaginal microbiota prior to BV treatment may be linked to BV recurrence, it is unclear whether the preceding genital immune milieu is predictive of treatment success. Methods Here we assessed whether baseline vaginal soluble immune factors or the composition of the vaginal microbiota predicted treatment success one month after metronidazole treatment in two separate cohorts of women with BV, one in the US and one in Kenya; samples within 48 hours of BV treatment were also available for the US cohort. Results Neither soluble immune factors nor the composition of the vaginal microbiota prior to BV treatment were associated with treatment response in either cohort. In the US cohort, although the absolute abundances of key vaginal bacterial taxa pre-treatment were not associated with treatment response, participants with sustained BV clearance had a more pronounced reduction in the absolute abundance of Gardnerella vaginalis immediately following treatment. Conclusions Pre-treatment immune and microbial parameters were not predictive of BV treatment success in these clinical cohorts.

HIV incidence continues to be unacceptably high in Eastern and Southern Africa, with women disproportionately affected. An increased per-contact risk of HIV acquisition among African, Caribbean, and other Black (ACB) women has been associated with the higher prevalence of bacterial vaginosis (BV) in these communities, wherein the vaginal microbiota is predominated by diverse pro-inflammatory anaerobic bacteria. However, while the vaginal microbiota in BV-free women is typically predominated by one of several different Lactobacillus spp., the degree of HIV protection afforded by a Lactobacillus-predominant vaginal microbiota also varies considerably. Specifically, L. crispatus is associated with an immunoregulatory genital immune environment, exclusion of BV-associated bacteria, and reduced HIV risk. In contrast, less HIV protection or exclusion of BV-associated bacteria and fewer immune benefits have been associated with L. iners—which is unfortunately the most common Lactobacillus species among ACB women. These species-specific clinical differences are underpinned by substantial genomic differences between Lactobacillus species: for instance, the much smaller genome of L. iners lacks the coding sequence for D-lactic acid dehydrogenase and cannot produce the D-lactate isomer that enhances HIV trapping in mucus but encodes for epithelial cell toxins and stress resistance proteins that may enhance bacterial survival in the context of microbiota and environmental fluctuations. While more studies are needed to elucidate whether differences in HIV protection between Lactobacillus species are due to direct genital immune effects or the exclusion of proinflammatory BV-associated bacteria, the current body of work suggests that for BV treatment to succeed as an HIV prevention strategy, it may be necessary to induce a vaginal microbiota that is predominated by specific (non-iners) Lactobacillus species.

The genital epithelial barrier is a crucial rst line of defence against HIV, and epithelial disruption may enhance HIV susceptibility. Assessment of genital epithelial integrity require biopsies, but their collection is not practical in many research settings. A validated biomarker of genital epithelial barrier integrity would therefore be useful. E-cadherin, an adhesion component of the cell-cell junction in epithelial tissues, may be released as soluble E-cadherin (sE-cad) upon epithelial trauma, and so we evaluated sEcad as a marker of genital epithelial disruption. First, using an in vitro monolayer of immortalised endocervical epithelial cells, we demonstrate that sE-cad, IL-1β, and IL-1α levels are immediately increased after physical disruption, followed by a delayed increase in IL-6 levels compared to undisrupted controls. In vivo studies con rmed that sE-cad levels in cervicovaginal secretions were signi cantly elevated 6 hours after endocervical cytobrush sampling. Furthermore, the level of sE-cad in coronal sulcus swabs from Ugandan men was inversely correlated with the amount of membrane-bound Ecadherin within the overlying foreskin tissues. Our results validate the use of soluble E-cadherin as a marker of epithelial disruption and demonstrate that the processes of physical disruption and in ammation in the genital tract are strongly intertwined.