Title: Spatial Dynamics of the Evolution and Spread of Antibiotic-Resistant Escherichia coli in the United States
Abstract:
Antibiotics are essential for treating bacterial infections in both humans and animals. However, growing resistance to antibiotics among bacteria has become a one of the leading global public health concerns in the 21st century. The spread of antibiotic resistance is the result of complex interactions between human, animal, and microbial ecologies. To account for these dynamics, this dissertation proposed an extended disease ecology framework for zoonotic diseases. This framework incorporates a One Health perspective recognizing the important role of animal population given the substantial consumption of antibiotics in livestock and companion animals. It also explicitly considers the adaptation of pathogen populations to the changing environment.
Among the various types of microbes developing antibiotic resistance, Escherichia coli (E. coli) is of particular concern because of its ubiquity and potential to act as a large pool for various resistance genes. Following the proposed theoretical framework, this dissertation investigated spatial dynamics of the evolution and spread of antibiotic resistance among E. coli populations in the US.
Using genotypic resistance data, we compared the composition and co-occurrence patterns of resistance genes in E. coli isolates from humans, companion animals, and livestock. While the isolates from different host groups often showed genotypic resistance to the same classes of antibiotics, the underlying gene composition differed substantially, indicating potential species barriers to gene flow. In particular, resistance gene profiles in companion animals overlapped with both human and livestock isolates, indicating the potential role of companion animals as reservoirs and intermediaries in the cross-species transmission of resistance. The study also revealed widespread but spatially diverse co-occurrence patterns of resistance genes, particularly among human and companion animal isolates.
Building on this, I employed whole genome sequences and phylogenetic analysis of the ST131 lineage, a widespread highly pathogenic E. coli strain frequently associated with multi-drug resistance. The results substantiated that direct transmissions between human and livestock were limited, whereas companion animals could serve as a potential reservoir that bridges the two. Pagel’s λ statistics confirmed that the co-occurrence of certain resistance genes was due to shared evolutionary histories, rather than parallel evolution. Finally, the study also highlighted the different regions’ particular roles in the circulation of the strain in different host groups.
Complementing the examination of genotypic resistance at the national and regional scales, we employed a large sample population of Veteran’s Health Administration outpatients in seven Midwest states to examine the county-level spatial patterns of and individual- and county-level risk factors for phenotypic resistance found in E. coli isolates. Certain urban regions in the southern parts of Illinois, Indiana and Ohio with levels of resistance exceeding clinically significant thresholds may reflect local prescribing practices. While being male, having diabetes and having recent antibiotic exposure were significant risk factors for all four classes of antibiotics, variations in other factors among the classes of antibiotics likely reflected differences in prescribing scenarios of different antibiotics.
Together, these studies demonstrated the value of an extended disease ecology framework that integrates a One Health perspective and considers the adaptation of pathogen population in studying antibiotic resistance. The finding highlighted that an effective antibiotic stewardship effort must involve multisectoral collaborations that engage human and veterinary medicine, livestock husbandry industry, governmental oversight as well as the general public in reducing infections and unnecessary antibiotic usage.