The Diversity, Ecology, and Clinical Implications of Antibiotic Resistance Genes in Uncultured Soil Microbial Communities

Antibiotic resistance genes from soil are ancient, diverse, and represent an evolutionary origin for resistance seen in the clinic. In my thesis, I demonstrate that the lateral exchange of antibiotic resistance genes has occurred between non-pathogenic soil bacteria and diverse human pathogens on recent timescales, indicating that the vast soil resistome may serve as a reservoir of resistance available to the clinic. Despite the apparent overlap between soil and clinical resistomes, factors that influence resistance gene composition in soil and their movement between genomes and habitats are largely unknown. To investigate these relationships, I performed functional metagenomic selections for antibiotic resistance genes from 18 grassland and agricultural soils, identifying 2895 resistance genes with sequences predominantly unlike those in public databases. I was able to identify factors that drive resistome composition across these soils, detecting strong correlations between resistance genes and (i) soil type, (ii) the amount of added nitrogenous fertilizer, and (iii) bacterial community composition. I also describe a novel family of tetracycline-inactivating enzymes, widespread in soil and with ties to pathogenic Legionella, but previously unrecognizable as antibiotic resistance genes. Because tetracycline inactivation is scarcely observed in hospitals, these enzymes may fill an empty niche in pathogenic organisms and should therefore be monitored for their dissemination potential into the clinic. Taken together, my data indicate that soil resistomes are largely structured by bacterial community composition, but are capable of exchanging resistance genes with clinical pathogens, including previously unseen resistance phenotypes. Further work to identify the ecological, taxonomic, and genetic risk factors associated with this resistance gene flow is warranted