Possible mechanistic drivers of rapid shifts within the vaginal microbiome

Identification: Schiffer, Joshua


Possible mechanistic drivers of rapid shifts within the vaginal microbiome
Florencia A. T. Boshier1, Sujatha Srinivasan1, Amalia Magaret1,2, Hyunju Son2, Anthony Lopez1, Sean Proll1, David N. Fredricks1,2, Joshua T. Schiffer1,2
1University of Washington, Seattle
2Fred Hutchinson Cancer Research Center, Seattle
Dramatic transitions from Lactobacillus predominance to a complex polymicrobial state correspond with development of bacterial vaginosis (BV). We sought to understand how individual species interact to establish this dysbiotic state. We enrolled 20 women with recurrent BV who self-collected vaginal swabs every 8 hours for 60 days. Swabs were processed for quantitative PCR (qPCR) of seven key species, with daily assessment of bacterial diversity using broad-range PCR and deep sequencing. We generated univariate Poisson mixed models: independent variables were menses, vaginal sex and oral sex; outcome was proportion of days with an increase in bacterial quantity of >0.5 log; time windows between exposure and outcome varied between 0 and 5 days. We mapped bacterial interaction networks using statistical and dynamical systems approaches including Granger causality, LIMITS (Learning Interactions from Microbial Time Series) and SSMAP (Equation Free Modeling Linear Regression Analysis). Three women had stable high levels of Lactobacillus species with periodic introduction but rapid elimination of BV-associated species; five women had high levels of Lactobacillus species with moderate levels of fluctuating BV-associated species; two women demonstrated BV-associated species predominance, high diversity and periodic introduction and elimination of Lactobacillus species; ten women had occasional profound shifts between Lactobacillus predominance and a diverse polymicrobial profile. Shifts of 2-3 logs in individual species occurred in 8-24 hour-windows resulting in complete reorganization of the vaginal microbiome. Menses and sexual activity did not predict shifts in any species. The two dynamical approaches (LIMITS and SSMAP) detected interactions between species that were relatively sparse with evidence of intra- and inter-species competition which varied between individuals, but no growth synergy between pairs.  While the vaginal microbiome is often volatile over intervals of hours in women with BV, the key mechanistic drivers of abrupt shifts in vaginal microbial diversity remain unknown. 
This work was funded by the NIH STI CRC (U19 AI 113173)


Credits: None available.

You must be logged in and own this product in order to post comments.