Comparative Metagenomics of Vaginal Microbiomes Reveal Strain-Level Differences with Relevance to Colonization and Host-Microbe Interactions

Abstract The vaginal microbiome, a relatively simple, low diversity ecosystem crucial for female health, is often dominated by Lactobacillus spp.. Studies of this community have often relied on short read amplicon sequencing, limiting the taxonomic resolution to genus or species level. More detailed strain-level data, facilitated by shotgun sequencing, can provide a greater understanding of the mechanisms of colonization and host-microbe interactions. In this study, we analysed 354 vaginal metagenomes from pregnant women in Ireland to investigate the metagenomic community state types and strain-level variations, particularly those involved in cell surface interface with the host. Our analysis showed the presence of multiple metagenomic subspecies across major vaginal bacteria with Lactobacillus crispatus and Lactobacillus iners being the most dominant. Detailed examination revealed distinct genes, including mucin-binding genes, present in Lactobacillus crispatus subspecies but absent in Lactobacillus iners. Within the dataset, we recovered 337 high-quality metagenome-assembled genomes and determined that L. crispatus harbours a greater number of strain-specific genes related to cell wall biogenesis, carbohydrate and amino acid metabolism compared to L. iners, with genes predicted to be involved in mucin binding and cell surface structure under positive selection in L. crispatus. Supplementing our dataset with publicly available genomes, we revealed a cell surface glycan gene cluster predominantly present in L. crispatus that is absent from L. iners and Gardnerella vaginalis. Overall, our comprehensive analysis offers new insights into the intra-species diversity of the vaginal microbiome, emphasising the importance of strain-specific variations in factors associated with colonisation and host-microbe interactions, which are essential for maintaining vaginal health.