12: Alterations in the placental microbiome among spontaneous preterm births

We have previously demonstrated that the placenta harbors a low biomass microbiome which varies in association with preterm birth (PTB). However, with regard to examination of the placental microbiome, there are inherent limitations to longitudinal placenta collection in any given pregnancy, making it problematic to delineate causation from association. Here, we aimed to examine associations with the microbiome-encoded metabolic pathways and preterm birth using a large cohort with both longitudinal and cross-sectional sampling. We reasoned that robust causal inference analysis across multiple body sites in a prospective longitudinal cohort inclusive of both spontaneous and indicated PTB could potentially overcome such obstacles. Subjects were enrolled (n=331) in the early third trimester or at delivery (196 term, 135 preterm). Extensive clinical metadata (such as comorbidities and indications for inductions) enabled covariate analytics and gestational age (GA) comparison. Oral, vaginal, stool, and placental swabs and tissue were uniformly collected from subjects and their infants. DNA was extracted and subjected to 16S and whole genome shotgun (WGS) metagenomics. Quality filtered sequences were analyzed (QIIME and MG-RAST) and causal inference approaches (hierarchical clustering by Manhattan distance and regression modeling). Upon examination of term and preterm subjects, we saw minimal differences by virtue of GA and type of labor (spontaneous versus indicated) within the posterior fornix (p=0.053), the maternal oral cavity (p=0.534), and stool (p=0.585). However, with close examination of the preterm placental microbiome, we found differences in taxa abundance manifest as increases in Ureaplasma and Mycoplasmatales in subjects with sPTB. We found significant increases in Streptophyta in subjects with iPTB (p<0.05). Inferred metabolic pathways revealed that lipid metabolism was significantly increased in subjects with sPTB (p<0.05) while lipoic and linolenic acid metabolism were increased in subjects with iPTB (p<0.05). Causal inference analysis in a longitudinal cohort enabled us to delineate an association of the placental microbiome with sPTB, which appears distinct from indicated births at akin gestational ages.