A large scale investigation of the vaginal metagenome and metabolome and their role in spontaneous preterm birth

点击翻译按钮获取中文摘要

ABSTRACT Spontaneous preterm birth (sPTB) is a leading cause of neonatal morbidity and mortality. Despite extensive research and medical efforts, we lack both the means to identify it early and the therapeutic options to prevent it when identified. The vaginal microbiome is believed to have a causal role in a substantial fraction of sPTBs, and multiple studies have shown that even when measured early in pregnancy, the microbiome is associated with adverse pregnancy outcomes. Despite this great promise, robust microbiome-based predictors for sPTB have not been developed yet, and we lack a detailed understanding of the mechanisms underlying its involvement in sPTB. We attribute this to a small sample size in current studies; the aggregative definition of sPTB as a binary variable, ignoring its heterogeneous presentations and etiologies; and a focus on microbial taxonomy instead of data that is more functionally and mechanistically oriented. We propose to capitalize on one of the largest and best-phenotyped pregnancy cohorts collected to date, the nuMoM2b study. This was a national, multi-center trial that collected vaginal swabs at three time-points from women with diverse geographic and demographic backgrounds. This study profiled nearly 500 women with sPTB, almost an order of magnitude more than any previous microbiome study. We have compiled an interdisciplinary team with intimate knowledge of the nuMoM2b study, ample experience in compiling, profiling, and analyzing large-scale microbiome and metabolomic cohorts, and a track record of innovation in microbiome analysis and its clinical applications. We will perform deep metagenomic sequencing, longitudinal sample sequencing, and matched metabolomic analysis of a total of almost 6,000 samples from nearly 1,500 pregnant women, generating the largest such dataset to date. This dataset would advance 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 sPTB, with an unprecedented depth. Our proposed data analysis offers a nuanced and high-resolution view of both clinical and biological data. We will study sPTB while accounting for its various clinical presentations, etiologies, and important maternal covariates. We will study the vaginal ecosystem from multiple angles, investigating microbial genomic adaptations, levels of metabolites in the ecosystem, and dynamic changes to the microbiome profile. We will devise an aggregative computational framework, based on state-of-the-art methods and algorithms, that provides early prediction of sPTB based on microbiome-related features. We will employ a combination of parametric and non-parametric methods to obtain mechanistic insights into host- microbiome interactions that potentially contribute to sPTB and other adverse pregnancy outcomes. Altogether, this study will be transformative to our understanding of the vaginal microbiome in adverse pregnancy outcomes and of host-microbiome interactions in general.