Horizontal Plasmid Transfer Drives Antibiotic Resistance in Intensive Aquaculture Farms

Genetic transfer, including horizontal plasmid transfer (HPT), drives bacterial evolution by mediating the sharing of adaptive traits such as antibiotic resistance (AR). However, HPT mechanisms of diverse plasmids in natural environments in complex bacterial populations remain unexplored. Here we assessed the role of HPT in the emergence and spread of antibiotic-resistant genes (ARGs) in intensive frog farms. Metagenomic analyses of frog farm sediments, combined with high throughput antibiotic susceptibility examination of bacterial isolates, revealed Escherichia coli and Edwardsiella tarda as the core antibiotic resistant bacteria (ARB). Upon sequencing of the genomes of 95 multidrug-resistant bacteria (mainly E. coli, E. tarda, Citrobacter, and Klebsiella), we identified 250 plasmids (average size ≥110 kb) that harbored various ARGs flanked by mobile genetic elements (MGEs). AR against commonly used antibiotics strongly correlated with plasmid borne ARGs. Many plasmids (57%) were conjugative, with 20% being multi-replicons. Sixty-two percent of these plasmids belonged to 32 distinct groups that tracked ARG dissemination via inter- and intra-species HPT. Multiple plasmids in various combinations (2-3 per host), were shown to preferentially co-habit to achieve the desirable AR phenotype for adaptation. Same mobile ARGs (flanked by MGEs) were found in different locations on similar or different plasmids from the same or different bacterial hosts. Blasting public datasets also identified plasmids from other environmental niches globally, that were highly identical to those from the frog farms. Our data provide a real time record of the spread of ARGs in intensive animal farms and the mechanisms involved in resistome development.
Since the majority of isolated plasmids were conjugative, we then investigated interactions between two high-efficient conjugative plasmids from E. coli (pET82-3 and pET979-1) that had different type IV secretion system machineries. We address two questions in these studies: (1) How does one conjugative plasmid affect the transferability of another conjugative plasmid in the same host? (2) How does the presence of a conjugative plasmid in one bacterial host affect the conjugation of an incoming plasmid? The two plasmids were labelled with either gfp or mScarlet genes to track conjugation using florescence microscopy. Mutations in the ATPase genes (virB4 /virD4 in pET82-3 or traC /traD in pET979-1) completely blocked the transfer of the respective plasmid. Remarkably, the mutated plasmid could be transferred in the presence of a different native plasmid, although at relatively low conjugation frequency. Additionally, the bacterial host with the presence of a plasmid prevented accepting the same plasmid but not a different plasmid. This indicates that an exclusion system is present on each unique conjugative plasmid. Understanding conjugation at the molecular level can help us develop strategies to combat the spread of ARGs and prevent the emergence of superbugs. Our pioneering work will also shed light on the mechanisms involved in resistome development in natural communities.