A microbiota–host axis mediates prostaglandin sensitivity: Lactobacillus crispatus as a biomarker and regulator of human labor induction

Labor induction failure increases the risk of unplanned cesarean delivery and maternal-neonatal complications. However, the determinants of induction sensitivity remain poorly understood. In this prospective cohort of 85 term pregnant women undergoing labor induction with Propess® (prostaglandin E2), we combined full-length 16S rRNA sequencing of the vaginal microbiota, untargeted metabolomics of vaginal secretions, and transcriptomic analysis of cervical stromal cells exposed to Lactobacillus crispatus supernatant to identify predictive factors and underlying mechanisms. We found that women with a poor induction response exhibited higher vaginal microbiota α-diversity and a significant reduction in L. crispatus abundance. The relative abundance of L. crispatus predicted induction success with an area under the curve (AUC) of 0.80 (95% CI: 0.70-0.90). Metabolomic analysis revealed distinct vaginal metabolic alterations in induction-insensitive women. Importantly, in vitro experiments showed that L. crispatus supernatant directly modulates the transcriptome of cervical stromal cells, upregulating genes involved in uterine contraction, tissue remodeling, and immune regulation. Our results identify vaginal L. crispatus as a key biomarker for labor induction sensitivity and elucidate a potential mechanism by which it primes the cervix for prostaglandin response. These findings provide a novel microbiota-host interaction framework for personalizing induction strategies in precision obstetrics.