Clinically prevalent transposons contribute to erm gene dissemination in the field soil under pseudo-persistent erythromycin contamination

Clinically relevant antibiotic resistance genes (ARGs) or their ancestral genes are widespread in natural soil microbiome at ultralow abundance. Whether and how long-term antibiotic pressure in soil accelerate dissemination of these ARGs remain unclear. Here, annual cycle of erythromycin exposure at levels around 5–20 μg∙kg⁻¹ was conducted in previously undisturbed field soil for consecutive five years, to simulate the pseudo-persistent characteristic of antibiotic contamination in soil environment. The primary clinically relevant macrolide resistance genes, rRNA methyltransferase genes (erm genes), were initially rare but gradually enriched, exhibiting a 37.8-fold increase after five years, which was greatly higher than macrolide efflux pump genes and inactivation genes (less than 2.3-fold). Among diverse mobile genetic elements, transposase gene tnpA exhibited potential association with the horizontal transfer of erm genes during long-term erythromycin exposure. From genetic and statistical evidence, enriched erm genes were presumed to locate on Bacilli with mobile transposable elements Tn554 and Tn551, which were clinically prevalent gene clusters in pathogens-Enterococcus and Staphylococcus. Thus, there may be a historical contribution of long-term erythromycin contamination to erm-carrying clinical transposable elements in soil microbiome. Our findings also demonstrated soil erythromycin exposure at levels much lower than laboratory-determined minimal selective concentrations (MSCs) still exhibits long-term effects on erm genes. Taking pseudo-persistent characteristic of antibiotic contamination, we further proposed long-term in-situ assessment with endpoint of clinically relevant ARGs to obtain a real-world MSC in the future studies.