Plasmid Replicon Typing of Antibiotic-Resistant Escherichia coli From Clams and Marine Sediments

Unlike human isolates, environmental Escherichia coli isolates have not been thoroughly investigated for the diversity and transferability of antibiotic-resistant plasmids. In this study, antibiotic-resistant strains from marine sediment (n = 50) and clams (n = 53) were analyzed (i) for their plasmid content using a PCR-based plasmid replicon typing (PBRT) kit and (ii) for the transferability of plasmid-associated antibiotic resistance (AR) traits by mating experiments. Fifteen of the thirty replicons targeted by the PBRT kit were detected in the isolates; 8/15 were identified in both sediment and clam isolates, although at different frequencies. The most frequent replicons in sediment (74%) and in clam strains (66%) alike, were FIA, FIB, or FII, which are associated with the IncF group, followed by the I1α replicon, which was more frequent in clam (24.5%) than in sediment (10%) strains. More than 50% of the strains contained multiple replicons; although 15 were untypable, S1-PFGE analysis demonstrated that 14/15 carried no plasmids. All cryptic strains were successfully typed and were positive for IncF or IncI replicons. Antibiotic-resistant strains accounted for 63% of all isolates and were significantly (p < 0.05) more frequent in phylogroup A. Most (35%) multidrug-resistant (MDR) strains belonged to phylogroup A, too. Although 25/26 MDR strains were positive for IncF plasmids (the exception being a clam strain), the FII-FIB rep combination was predominant (63%) among the sediment isolates, whereas most clam isolates (40%) carried the FII replicon alone. In mating experiments, selected MDR strains carrying FIB, FII, and I1α replicons, used as the donors, transferred multiple ARs together with the IncF or IncI plasmids at high frequency. Since IncI plasmids are common in E. coli and Salmonella enterica isolates from poultry, our findings suggest an animal origin to the E. coli clam strains carrying IncI plasmids. They also suggest a role for IncI plasmids in the spread of ARs among environmental Enterobacteriaceae and, through the food chain, to human isolates. In conclusion, the PBRT kit proved to be a useful tool to identify plasmids carrying antibiotic-resistant genes and to shed light on the factors underpinning their diffusion.