A large scale investigation of the vaginal ecosystem in preeclampsia

ABSTRACT Preeclampsia is a pregnancy complication characterized by high blood pressure and signs of systemic organ damage, often the liver and kidneys. It poses significant clinical challenges in terms of accurate early prediction and effective prevention and treatment, with grievous maternal and neonatal consequences. One promising yet relatively unexplored avenue is the vaginal ecosystem: its microbiome, metabolome, and host immune components. The vaginal microbiome is capable of influencing systemic and local inflammation, producing a wide range of metabolites, and is the likely source for ascending infections with the ability to translocate to other reproductive organs. It may therefore provide insights that will advance our understanding and management of preeclampsia. This project therefore proposes an in-depth study of the vaginal ecosystem in severe preeclampsia, aiming to uncover the microbial and molecular properties of the vaginal ecosystem that indicate, predispose to, or contribute to preeclampsia. We hypothesize that the early pregnancy vaginal microbiome is strongly associated with subsequent development of preeclampsia, facilitated by specific microbes, host immune states, or microbial metabolites. We will capitalize on one of the largest and most well- phenotyped pregnancy cohorts collected to date, the nuMoM2b study. This was a diverse, national, multi- center cohort that collected detailed clinical data in addition to vaginal swabs at three time points along pregnancy. We will conduct longitudinal metagenomic and metabolomic profiling of ~4,200 vaginal samples from 389 nulliparous pregnancies with severe preeclampsia and 389 matched controls, in addition to profiling host immune factors using immunoassays. This dataset would contribute significantly to the microbiome community, drive discovery in the field, and enable us and other researchers to study host-microbiome interactions, in general and specifically in the context of preeclampsia, with an unprecedented depth. Using this data, we will identify microbial associations with severe preeclampsia, longitudinally and cross-sectionally, and their interactions with host immune factors. We will further use a metabolomics approach focused on broad structural annotation and identifying metabolite sources to detect vaginal metabolites associated with preeclampsia, aiming to elucidate their roles as mediators of host-microbiome interactions. We will combine all collected molecular data along with clinical covariates to devise supervised machine learning models that predict the risk for preeclampsia early in pregnancy. Findings from this project would illuminate the potential processes by which the vaginal microbiome influences preeclampsia, providing specific mechanistic hypotheses for future validation and possibly leading to new therapeutic targets and diagnostics.