Giovanni Tassistro

Summary: Antibiotic resistance is a rising threat for human health. Although in clinical settings and terrestrial environments the rise of antibiotic resistant bacteria is well documented, their dissemination and spread in the marine environment, covering almost two‐thirds of the Earth's surface, is still poorly understood. In this study, the presence and abundance of sulphonamide resistance gene ( sul 2) and class 1 integron‐integrase gene ( intI 1), used as markers for the occurrence and spread of antibiotic resistance genes since the beginning of the antibiotic era, were investigated. Twenty‐nine archived formalin‐fixed samples, collected by the Continuous Plankton Recorder (CPR) survey in the Atlantic Ocean and North Sea from 1970 to 2011, were analysed using Droplet Digital PCR (ddPCR) applied for the first time on CPR samples. The two marker genes were present in a large fraction of the samples (48% for sul 2 and 76% for intI 1). In contrast, results from Real‐Time PCR performed on the same samples greatly underestimate their occurrence (21% for sul 2 and 52% for intI 1). Overall, besides providing successful use of ddPCR for the molecular analysis of CPR samples, this study reveals long‐term occurrence and spread of sul 2 gene and class 1 integrons in the plankton‐associated bacterial communities in the ocean.

The pacific oyster Crassostrea gigas and the Mediterranean mussel Mytilus galloprovincialis are two widely farmed bivalve species which show contrasting behaviour in relation to microbial diseases, with C. gigas being more susceptible and M. galloprovincialis being generally resistant. In a recent study, we showed that different susceptibility to infection exhibited by these two bivalve species may depend on their different capability to kill invading pathogens (e.g., Vibrio spp.) through the action of haemolymph components. Specific microbial-host interactions may also impact bivalve microbiome structure and further influence susceptibility/resistance to microbial diseases. To further investigate this concept, a comparative study of haemolymph and digestive gland 16SrDNA gene-based bacterial microbiota profiles in C. gigas and M. galloprovincialis co-cultivated at the same aquaculture site was carried out using pyrosequencing. Bacterial communities associated with bivalve tissues (hemolymph and digestive gland) were significantly different from those of seawater, and were dominated by relatively few genera such as Vibrio and Pseudoalteromonas. In general, Vibrio accounted for a larger fraction of the microbiota in C. gigas (on average 1.7-fold in the haemolymph) compared to M. galloprovincialis, suggesting that C. gigas may provide better conditions for survival for these bacteria, including potential pathogenic species such as V. aestuarianus. Vibrios appeared to be important members of C. gigas and M. galloprovincialis microbiota and might play a contrasting role in health and disease of bivalve species. Accordingly, microbiome analyses performed on bivalve specimens subjected to commercial depuration highlighted the ineffectiveness of such practice in removing Vibrio species from bivalve tissues.

The detection and typing of Vibrio cholerae in natural aquatic environments encounter major methodological challenges related to the fact that the bacterium is often present in environmental matrices at very low abundance in nonculturable state. This study applied, for the first time to our knowledge, a whole-genome enrichment (WGE) and next-generation sequencing (NGS) approach for direct genotyping and metagenomic analysis of low abundant V. cholerae DNA (V. cholerae metagenomic DNA via hybridization. An enriched V. cholerae metagenome library was generated and sequenced on an Illumina MiSeq platform. Up to 1.8 × 10⁷ bp (4.5× mean read depth) were found to map against V. cholerae reference genome sequences representing an increase of about 2500 times in target DNA coverage compared to theoretical calculations of performance for shotgun metagenomics. Analysis of metagenomic data revealed the presence of several V. cholerae virulence and virulence associated genes in river water including major virulence regions (e.g. CTX prophage and Vibrio pathogenicity island-1) and genetic markers of epidemic strains (e.g. O1-antigen biosynthesis gene cluster) that were not detectable by standard culture and molecular techniques. Overall, besides providing a powerful tool for direct genotyping of V. cholerae in complex environmental matrices, this study provides a ‘proof of concept’ on the methodological gap that might currently preclude a more comprehensive understanding of toxigenic V. cholerae emergence from natural aquatic environments.