Metagenomics Analysis Reveal Host-Microbe Interactions in Cervical Cancer

Access restricted, at author's request, from June 07, 2017 - June 07, 2018.Cervical cancer is the fourth most common cancer in women worldwide. Recently it has been shown that microbiota play a role in some types of cancer. Therefore, to understand whether microbes contribute in any way to the development and progress of cervical cancer, we investigated the cervical microbiome via analysis of 16S rRNA sequencing data. Using qPCR we detected bacterial DNA in 121 out of 123 samples of cervical carcinoma. However, 16S library preparation and sequencing was possible for only the 58 samples with higher amounts of bacterial DNA. After matching our results with the Human Microbiome Project, principle coordinate analysis showed that the closest neighbors to cancer samples were healthy vaginal samples, representing compositional similarity. Overall, the alpha diversity of cervical cancer communities was quite high (Shannon index 3.88) compared to the healthy microbiome of other body sites. Furthermore, the most abundant genera in these samples were Prevotella (17.0%), Fusobacterium (9.4%), and Porphyromonas (7.4%), all Gram-negative, anaerobic bacteria known to be associated with bacterial vaginosis. To explore host-microbe interactions, we reconstructed a transkingdom network between tumor gene expression and microbial relative abundances. Prevotella spp. was revealed as a key node with high bipartite betweenness centrality, and was found to induce the expression of key cancer driver genes in vitro. This study represents the first comprehensive integrative analysis of the microbiome and transcriptome in cervical cancer, showing the potential role of key bacteria in driving carcinogenesis