Quantitative analysis of initial adhesion of bacterial vaginosis anaerobes in ME-180 cells

Abstract

Bacterial vaginosis (BV) is the most common vaginal disorder in women of reproductive age [1]. Despite decades of research, the etiology of BV still remains elusive. It is well established that adhesion to host cells or tissues is a necessary early step in the establishment of infection [2]. Since it is often considered a polymicrobial condition, it has been proposed that some bacteria have a preponderant role as early colonizers, while others have an impact later in the development of a multi-species biofilm infection [1]. So, we tested this hypothesis by first quantifying the potential of known anaerobic bacterial species found in BV (Gardnerella vaginalis, Atopobium vaginae, Mobiluncus mulieris, Prevotella bivia and Fusobacteria nucleatum) to compete with a protective lactobacillus (L. crispatus) for initial adhesion to ME-180 cervical epithelial cells. Then, we tested the potential of the two vaginal lactobacilli; L. crispatus, which appears to be protective against BV, and L. iners, which does not protect as effectively against BV, to block these BV-associated anaerobes. And finally, we tested the ability of the BV-associated anaerobes to displace pre-adhered lactobacilli. This work allowed quantification, of the initial adhesion of these BV-associated anaerobes to ME-180 epithelial cells and competition and displament/blockage assays. In competition assays, G. vaginalis exhibited the greatest capacity for adherence to ME-180 cells, confirming our previous report [3]. Interesting, G. vaginalis also maintained its ability to adhere in the presence of L. crispatus better than the other species, and there was only a 10% reduction in adherence with respect to the control. This was statically different from the others BV anaerobes (ANOVA Tukey statistical test values, P < 0.05), as it adhered approximately 4-fold better than A. vaginae or M. mulieris and approximately 2-fold better than P. bivia. Adherence of L. crispatus was not statistically inhibited by any of the BV anaerobes tested. In order to simulate the introduction of BV-associated bacteria into a healthy vagina colonized by lactobacilli, we performed displacement/blockage assays in the presence of ME-180 cell line. To this end, we first allowed L. crispatus or L. iners to adhere to the epithelial monolayers and subsequently added a BV-associated species to quantify the inhibitory effect of the lactobacilli on secondary colonization. In these displacement/blockage assays, L. crispatus inhibited adherence of G. vaginalis by approximately 43%. Addition of G. vaginalis appeared to cause a slight displacement of adherent L. crispatus but this did not reach statistical significance. L. crispatus also reduced adherence of A. vaginae and M. mulieris by approximately 50%. P. bivia and F. nucleatum appeared to be less susceptible to inhibition by L. crispatus. Interestingly, L. iners, which has been shown in previous studies to be less protective against BV relative to other vaginal lactobacilli [4], had a similar inhibitory effect on adherence by all of the BV-associated species except G. vaginalis. Adherence of G. vaginalis actually increased somewhat in the presence of L. iners, although this increase did not reach statistical significance. None of the BV anaerobes displaced L. iners. To conclude, L. crispatus and L. iners inhibited adherence of all BV-associated species tested with the exception that L. iners actually enhanced adherence of G. vaginalis. This study supports the possibility that G. vaginalis could be an initial colonizer in BV etiology.