ABSTRACT: Bacterial vaginosis (BV), the most common cause of vaginal discharge, is associated with multiple adverse outcomes. The rate of recurrence after therapy is >60% yet BV etiology remains unknown. BV is characterized by loss of vaginal lactobacilli and increases in facultative (Gardnerella vaginalis) and strict anaerobes. G. vaginalis, present in 95-100% of BV cases, is more virulent than other BV-associated bacteria in vitro. However, it is also found in women with normal flora and colonization is not sufficient for BV development. A notable feature of BV is the appearance of a multi-species biofilm on vaginal epithelial cells containing abundant G. vaginalis, fewer Atopobium vaginae, and other undefined bacterial species. We hypothesize that G. vaginalis initiates BV biofilm formation, but incident BV (iBV) requires incorporation of other key bacteria into the biofilm that alter the transcriptome of the polymicrobial consortium. This is consistent with our finding that, among women who have sex with women, the mean relative abundance of Prevotella bivia, G. vaginalis, and A. vaginae became sequentially higher prior to iBV. We propose that a similar distribution of these bacterial species will increase prior to iBV in women who have sex with men (WSM). Our specific aims are: Aim 1. Investigate changes in the vaginal microbiota preceding iBV in a longitudinal study of WSM. We will obtain twice daily vaginal specimens from 150 women with normal vaginal flora (Nugent score 0-3) and follow them for iBV (Nugent score 7-10 on ≥4 consecutive vaginal specimens) for 60 days. 16S rRNA gene sequencing targeting V4 [and broad range 16S rRNA gene qPCR] will be done in women for the 14 days prior to iBV as well as age-comparable women maintaining normal vaginal microbiota. Aim 2: Determine the contribution of P. bivia to the multi-species BV biofilm in vivo. We will use peptide nucleic acid fluorescence in situ hybridization (PNA-FISH) probes previously optimized for G. vaginalis and A. vaginae to analyze BV biofilm formation over time in clue cells from vaginal specimens from women with iBV in Aim 1, focusing on the 14 days prior to iBV. We will design, validate, and optimize a new P. bivia PNA-FISH probe for multiplex use (Aim 2A). Light microscopy PNA-FISH will be performed on specimens from women with iBV and negative controls (Aim 2B). Confocal laser scanning microscopy FISH will be used to determine the interspatial localization of the 3 bacteria of interest in the BV biofilm on clue cells from women with iBV (Aim 2C). Aim 3. Identify molecular markers associated with iBV by using RNA sequencing to analyze the transcriptome of G. vaginalis, P. bivia, and A. vaginae. We will probe specific bacterial interactions during the 14 days prior to iBV, first by RNA sequencing using in vitro models in a chemically defined medium simulating vaginal secretions (mGTS) (Aim 3A). Venn-diagram analysis will select the most commonly up-regulated genes shared in the in vitro assays. The in vivo relevance of potential molecular markers for development of iBV will be assessed by qPCR, using vaginal samples from women with iBV and negative controls in Aim 1 (Aim 3B).

PROJECT NARRATIVE: The primary objective of this R01 is to investigate the pathogenesis of incident bacterial vaginosis (iBV) using a combination of cultivation independent and cultivation dependent studies. A notable feature of BV is the appearance of a multi-species biofilm on vaginal epithelial cells. We hypothesize that Gardnerella vaginalis initiates BV biofilm formation, but that iBV requires incorporation of other key bacteria (Prevotella bivia and Atopobium vaginae) into the biofilm that alter the transcriptome of the polymicrobial consortium. [By further investigating the etiology of BV, this study has great potential to improve diagnosis, treatment, and prevention.]