Meta-analysis of metagenomics reveals the signatures of vaginal microbiome in preterm birth

Abstract

Preterm birth (PTB) is the main cause of perinatal incidence rate and mortality worldwide and seriously threatens lives of newborns. In order to improve the understanding of PTB etiology, this study aimed to investigate associations between vaginal microbiome and PTB. We conducted an in-depth meta-analysis on publicly available shotgun metagenomics datasets of vaginal microbiome, including comparisons of microbial composition, metabolic pathways, biosynthetic gene clusters (BGCs), and virulence factors (VFs) between PTB and Healthy groups. Our results showed that 113 species existed in PTB group and 161 species in Healthy group, and their species compositions were significantly different. PTB group was associated with six species, namely Lactobacillus crispatus, Atopobium vaginae, Prevotella bivia, Neisseria subflava, Corynebacterium sp. HMSC078H07 and Capnocytophaga leadbetteri. Between the two groups exhibited 314 significantly different KEGG orthologys. The distribution of BGCs in PTB group were significantly different from that in Healthy group. The total amount of BGCs in PTB group was 95 and they were divided into 12 types, in Healthy group 300 BGCs into 16 types. We also obtained 7080 types of VF genes in PTB group, and 10,748 in Healthy group. The virulence gene with the highest proportion in both groups was ssrA. To conclude, this meta-analysis indicated that significant differences of microbial relative abundance were observed between PTB and Healthy groups. PTB group carried less total amount and types of BGCs, and less types of VFs than those in Healthy group, and PTB group showed significantly different metabolic pathways from Healthy group. We also provided new hypotheses related to vaginal microbiome and PTB.