Unleashing the potential of mRNA-seq to uncover the microbiome structure and their crosstalk with host cells: the vulvar ecosystem.

Abstract

Background: To describe both host gene expression and microbiome composition in a single sample, parallel experimental and computational workflows (mRNA-sequencing and either 16S rRNA gene or metagenomics) have been traditionally applied. The vulvar milieu represents an area of emerging research for its role in health and disease. Located at the interface between the vagina and the perineum, the vulvar microbiome displays an intermediate signature, with influx from both ecosystems.

Results: Following validation of the reliability of poly(A)-enriched mRNA-sequencing in reconstructing the microbiota composition using both a quantitative microbial standard (mock) and metagenomic analysis, we analyze a full cohort of 30 healthy vulvar samples. Crucially, the analysis of the entire cohort relies solely on mRNA-sequencing without the use of parallel DNA metagenomics. This unified approach allows us to analyze not only the vulvar cell transcriptome, but also the composition and dynamics of microbial communities, including the microbial gene expression signatures. This three-level analysis (host-mRNA, individual bacterial species, bacterial gene pathways) on the very same specimens further enables a gene-level exploration of host-microbe molecular crosstalk. Using this unified framework, we reveal marked heterogeneity and high inter-individual variability in the vulvar microbiota, identifying community state types that mirror those described in the vagina. Importantly, we show that distinct microbial configurations are associated with specific host transcriptional programs: Lactobacillus crispatus correlates with epithelial differentiation and barrier integrity, whereas communities enriched in Gardnerella vaginalis, or other taxa associated with dysbiosis, exhibit transcriptional signatures linked to inflammation. Interestingly, Lactobacillus gasseri, which has been associated with lower protection, shows an intermediate effect on vulvar cells.

Conclusions: Beyond providing new biological insights into an understudied anatomical niche, our study introduces a broadly applicable strategy with substantial impact for the field. With tens of thousands of human RNA-seq datasets already available in public repositories, our approach enables retrospective extraction of microbiome information and host-microbe interaction signals from existing transcriptomic data, without the need for additional sequencing or specialized microbiome protocols. This unlocks a powerful and cost-effective opportunity to revisit archived RNA-seq studies across tissues, diseases, and low-biomass environments, revealing previously inaccessible layers of host-microbiome crosstalk and maximizing the scientific value of published data. Video Abstract.