|dc.description.abstract||Gardnerella spp. are a hallmark of bacterial vaginosis – a vaginal dysbiosis characterized by the shift from a lactobacilli dominated vaginal microbiome to a diverse microbiome composed of many aerobic and anerobic bacteria. Gardnerella spp. are diverse both phenotypically and genotypically and likely differ in their roles in the pathogenesis of vaginosis. Four cpn60-defined subgroups: A, B, C, and D, corresponding to thirteen “genome species” have been identified. Multiple subgroups of Gardnerella usually colonize vagina simultaneously. Subgroups A and C are the most common and abundant, while subgroup D is rarely abundant. The factors determining Gardnerella community composition and dynamics are unknown.
Since the vaginal environment has a limited supply of host-derived nutrients, it is intuitive that species of Gardnerella may interact to secure resources, especially when they dominate the microbiome. The first objective of our study, therefore, was to determine the types of interactions between Gardnerella spp.. While no evidence of contact-independent interaction was identified, evaluation of the outcomes of co-cultures of 2, 3 or 4 isolates indicated that resource-based, scramble competition was prevalent among Gardnerella spp.. Although competition was common, not all Gardnerella subgroups were affected equally by the competition. While A, B, and C were negatively affected, the growth rate of subgroup D isolates increased with increasing numbers of competitors.
The success of this rare subgroup and its maintenance in the vaginal microbiome at low abundance might be explained by negative frequency-dependent selection in which the fitness of a species is highest when its relative abundance in a community is low. Species that are nutritional generalists and that occupy distinct niches from their more abundant competitors are candidates for negative frequency-dependent selection. To test this hypothesis, we evaluated the amount of niche overlap between Gardnerella species and investigated if subgroup D is a nutritional generalist using comparative genomics and a phenotypic assay. Comparison of predicted proteomes of 40 isolates revealed minimal overlap between Gardnerella subgroup D and the three other subgroups. Subgroup D isolates utilized more carbon sources than those other subgroups and had distinct patterns of utilization.
If subgroup D is favoured by negative frequency-dependent selection, why is it so rarely abundant in the real world? We speculated that the dynamics of the vaginal ecosystem could periodically reset the bacterial population density providing abundant resources for the growth of common species, and checking the success of rare subgroup D. We generated contrived communities of subgroup C and D isolates in vitro at varying initial population densities and measured relative abundance of each subgroup by deep sequencing of rpoB amplicons at 0h and 72h. Density did affect the outcome of competition but unexpectedly, both positive and negative effects of initial population density on subgroup D isolates were observed. The findings suggest that although population density may affect the outcome of subgroup D in competition, other factors including growth rate of competing species and biofilm forming potential may also impact their success.
Biofilm formation is considered a major cause of chronic and recurrent infection and is also often implicated in treatment failure of bacterial vaginosis. It has also been suggested that biofilm formation is enhanced by sub-therapeutic levels of antibiotics. Using a broth microdilution assay, we demonstrated that minimum inhibitory concentrations of metronidazole for both planktonic and biofilm cultures were similar for isolates in all subgroups, suggesting that growth as biofilm does not provide enhanced protection against metronidazole. Stimulation of biofilm production by sub-inhibitory concentrations of metronidazole was observed in a small minority of cases.
The findings presented in this thesis are a way forward to unveil the complexity of interactions at play in the vaginal ecosystem, which will enable us to better understand the dynamics of the vaginal ecosystem, and with any luck, the reason(s) for occurrence of bacterial vaginosis.||