Gut microbiota signatures are associated with prostate cancer risk

Background: Prostate cancer (PC) is the most common cancer in men worldwide. The incidence of PC varies significantly geographically, which might result e.g. from genetic factors and local discrepancies in screening policies, but also from differences in lifestyle such as diet. Novel environmental factor, namely gut microbiota (GM) has been recently associated with many pathological processes including tumor progression within human body but its role in PC is disputable. Methods: Within a clinical prospective single center trial, the GM profiles were assessed from 181 men with clinical suspicion of PC utilizing 16S rRNA gene sequencing (Illumina). Sequences were assigned to operational taxonomic units (OTUs) after which differential abundance analysis, - and {beta}-diversities, and predictive functional (PICRUST) analysis were performed. Further, plasma steroid hormone levels correlated to predicted microbiota functions. Results: PC was diagnosed in 60% (108/181). Apart for less smoking among subjects with PC, there were no life-style differences between the groups. The GM profiles of men with PC differed significantly from those without cancer, e.g. Prevotella 9, members of family Erysipelotrichaceae and potentially pathogenic Escherichia-Shigella were increased, and e.g. Jonquetella, Moryella, Anaeroglobus, Corynebacterium and CAG-352 were reduced in PC cases. Predictive functional analyses revealed increased 5--reductase activity (5-AR), copper absorption and retinal metabolism as functional results of different microbiota. Plasma testosterone negatively correlated with predicted microbial 5-AR activity (Wilcoxon rank sum p=0.057) and in a subgroup of men taking 5-AR inhibitors (n=17), plasma estrone 3 (p=0.027), and estradiol (p=0.054) levels were higher in men with predicted increased microbial 5-AR function. Conclusions: Certain bacterial and functional features within GM composition are associated with PC risk and altered androgen, copper and retinol metabolism are potential mechanisms. Findings could explain the previously reported association of life-style effects and geographical differences observed in PC.