Integrated cross-sectional study and functional validation indicate the association of lactobacillus crispatus-derived D-lactic acid with cervical gene expression and precancerous cervical lesions

BACKGROUND: Human papillomavirus (HPV) persistence and progression to high-grade squamous intraepithelial lesions (HSIL) involves complex interactions between the cervicovaginal microbiome, metabolome, and host immune responses. While Lactobacillus crispatus has been associated with the risk of cervical lesion progression, the underlying molecular mechanisms remain poorly understood. METHODS: We conducted a cross-sectional study of 74 HPV-positive Chinese women diagnosed with low-grade SIL (LSIL) or HSIL. Cervicovaginal swabs were analyzed using 16S rRNA sequencing and untargeted LC-MS metabolomics, while cervical exfoliated cells underwent RNA-sequencing. Community state types (CSTs) were determined by hierarchical clustering, and differentially expressed genes were identified across CSTs and SIL grades. Bayesian network analysis was employed to model molecular interactions between CVM compositions, metabolites, host gene expression, and HSIL status. Key findings were validated through in vitro experiments using HPV16-infected cervical epithelial cells (Ect1/E6E7) co-cultured with Lactobacillus supernatants and D-lactic acid. RESULTS: L. crispatus-dominated communities (CST_I) were significantly enriched in LSIL compared to HSIL patients (16.22% vs 2.70%, p < 0.05). CST_I exhibited 2.2–2.8-fold higher D-lactic acid levels compared to L. iners-dominated (CST_III) and non-Lactobacillus-dominated (CST_IV) communities. Transcriptomic analysis identified 116 DEGs varying across both CSTs and SIL grades, with functional enrichment revealing HSIL-associated genes involved in axonogenesis and pattern specification, while LSIL-associated genes were linked to immune responses. Bayesian network modeling suggested that CST_I-derived D-lactic acid inhibits FOXD1 expression, which directly correlates with HSIL status. In vitro validation confirmed that L. crispatus supernatant and D-lactic acid significantly reduced FOXD1 expression (p < 0.05), decreased HPV16 viral load, and inhibited Ect1/E6E7 cell growth. CONCLUSIONS: This study elucidates a novel protective mechanism whereby L. crispatus-derived D-lactic acid modulates cervical gene expression, particularly through FOXD1 downregulation, conferring negative links to precancerous cervical lesions. These findings provide mechanistic insights into microbiome-mediated cervical cancer prevention and suggest potential therapeutic targets for intervention strategies aimed at reducing HSIL progression.