Micro-organisms which dwell in sediments have a crucial role in biogeochemical cycles and are thus expected to have a strong influence on the cycle of arsenic, a metalloid responsible for severe water pollution and presenting a major health risk for the populations. Nowadays, environmental genomics techniques allow to study the adaptive and cooperative strategies of microbial communities thriving in polluted environments. We sequenced the sediment metagenomes from l'Estaque marina (Marseille, France), close to a former metallurgic industrial site highly polluted with arsenic, and from St Mandrier marina, near Toulon (France). The functional and taxonomic profiles of both metagenomes, along with those from four publicly available metagenomes used as control data sets, were obtained using the RAMMCAP workflow. The biodiversity was higher in both harbour sediment communities in comparison to the controls. The order of Desulfobacterales represented 54.7% of the observed orders in l'Estaque and 31.7% in St Mandrier. Other orders were evenly distributed though the diversity was lower in the highly polluted site of l'Estaque than in St Mandrier.
The Gene Ontology (GO) entries describing the functional profiles were compared with the controls in order to highlight the over-represented categories in the metagenomes of interest. We observed that categories related to arsenic resistance and oxidative stress were over-represented in l'Estaque. More importantly, the integration of metagenomic and physico-chemical data showed that in l'Estaque, dissimilatory sulphate reduction in the upper layer of sediment could increase arsenic diffusion to the water column and deeper layers, resulting in local detoxification of the sediment.