Retail Food as a Primary Driver for the Transmission of Antibiotic-Resistant Escherichia coli in Food-Water-Human Systems

Abstract

Antibiotic resistance (AR) dissemination is shaped by interconnected food, water, human-associated, and environmental systems that influence the movement, persistence, and redistribution of resistant bacteria and mobile genetic elements. This dissertation examined how multidrug-resistant (MDR) and potentially pathogenic Escherichia coli are introduced, enriched, and shared across retail food, river water, and human-associated compartments in the Chobe region of northern Botswana. It also characterized how environmental and geochemical gradients structure bacterial communities in Chobe River floodplain soils. Together, these studies provide insight into how food-associated inputs, cross-source bacterial connectivity, and environmental heterogeneity influence the distribution and persistence of resistance and microbial communities within a complex regional landscape. Chapter 3 evaluated retail beef and poultry imports (n = 109) as potential reservoirs of AR and virulent E. coli. E. coli was isolated from 54.7% of samples, with higher multidrug resistance observed in poultry (45.5%) than beef (13.3%). Resistance was detected to 11 of 12 tested antibiotics, with tetracycline, trimethoprim/sulfamethoxazole, and doxycycline most frequently observed. Whole-genome sequencing of MDR isolates revealed diverse sequence types, including diarrheagenic and extraintestinal pathogenic lineages. Plasmid-borne resistance genes co-occurring with conjugation-associated genes were common, and phylogenetic analyses demonstrated clustering with globally distributed human and poultry isolates. These findings indicate that imported meat can function as a reservoir of MDR and pathogenic E. coli and serve as a pathway for introducing mobile resistance determinants into new ecological contexts. Chapter 4 integrated phenotypic and genomic analyses of E. coli from retail food (n = 164; 109 meat and 55 produce), river water (n = 315), and human fecal samples (n = 146) collected in 2022, with comparison to historical genomes from 2011. Phylogenetic reconstruction revealed interspersed clustering of food, water, and human isolates, indicating a shared lineage pool rather than source-specific segregation. Water isolates from human-dense areas clustered closely with human isolates, suggesting spatially structured ecological connectivity. IncF-family plasmid replicons were shared across compartments, and identical plasmid-associated resistance genes were observed in food and human isolates. Integrons were detected exclusively in food isolates, and food-associated isolates exhibited the greatest plasmid diversity and mobile AR gene burden, supporting evidence of food-driven mobile resistance enrichment in 2022. Historical genomes were phylogenetically interspersed with 2022 isolates, with overlapping sequence types across years, indicating temporal continuity of lineages. However, water and human isolates from 2011 exhibited higher AMR gene prevalence despite the absence of food sampling, suggesting that resistance drivers may shift across time. Collectively, these results support a model of food-associated resistance enrichment embedded within a temporally and spatially connected ecological network. Chapter 5 characterized bacterial community composition across ridge-swale floodplain soils (n = 30) and three depth intervals. While alpha diversity was similar between landforms at equivalent depths, beta diversity was strongly structured by depth, consistent with redox and substrate gradients. Swale soils exhibited higher carbon, nitrogen, and electrical conductivity, and genus-level distributions were most strongly associated with pH and nitrogen. These findings demonstrate that hydrological processes homogenize richness laterally while depth-dependent chemical gradients filter community composition. Together, this dissertation demonstrates that food can act as a significant enrichment node for mobile resistance within interconnected ecological systems, that lineage connectivity across food, water, and humans occurs across spatial and temporal scales, and that environmental structuring processes further shape microbial community distribution. These findings underscore the need for integrated genomic surveillance and longitudinal monitoring to understand shifting resistance drivers within complex ecological networks.