Water as a Driver of Antibiotic-Resistant and Pathogenic <i>Escherichia coli</i> Transmission in Areas with Mixed Land Use

Abstract

Introduction: Aquatic environments are critical reservoirs and transmission pathways for Escherichia coli, linking human populations with surrounding ecosystems. The Chobe River in northern Botswana is a vital water source for people, livestock, and wildlife, creating opportunities for microbial exchange at the human-environment interface. Areas upstream of Kasane are protected parkland with limited human inputs. We hypothesized that anthropogenic drivers increase the persistence of antibiotic resistance (ABR) and pathogenic E. coli and that environmental isolates will be more similar to human isolates. Methods: Water (n = 324) was collected along a transect passing through pristine, mixed-land use and town environments. Human fecal samples (n = 507) were collected in Kasane, Botswana. E. coli isolates were obtained by selective enrichment and plating on EMB, passed for purity on MacConkey agar, and confirmed through phoA gene amplification by PCR. ABR profiles were determined against 12 antibiotics using CLSI disk diffusion. Whole-genome sequencing (WGS) was performed on 12 multidrug-resistant isolates (six per source) on Illumina NextSeq platform, and genomes were analyzed in Galaxy and the BV-BRC platform. Results: E. coli was recovered from 96.60% (313/324) of water samples and 45.4% (230/507) of human fecal samples. Resistance to at least one antibiotic was detected in 27.16% (85/313) of water isolates and 43.0% (99/230) of human isolates, and multidrug resistance was observed in 12.78% (40/313) and 11.7% (27/230) of isolates, respectively. By land-use, ABR prevalence among water isolates was 33.33% (25/75) in mixed-use areas, 34.31% (35/102) in town areas, and 18.38% (25/136) in park areas. ABR patterns were similar across sources, with ampicillin, tetracycline, streptomycin, and trimethoprim/sulfamethoxazole being the most common. WGS revealed distinct sequence type (ST) distributions, with water isolates dominated by ST11 whereas human isolates spanned six different STs. Although no STs were shared between sources, one water isolate (ST2852) collected near town clustered with human isolates in the phylogeny. Human genomes carried multiple acquired ABR genes including blaCTX-M-15, blaTEM-1, qnrS1, tetA/B, and sul2, whereas water genomes lacked frequently acquired resistance genes despite exhibiting multidrug-resistant phenotypes. Virulence profiling identified shared adhesins and secretion systems across sources, while water isolates carried LEE, stx1, and hemolysin genes, and human isolates encoded siderophore systems, toxins, and serum resistance factors. All isolates from water and human were predicted to be human pathogens with probability > 0.9. Significance: This study highlights the circulation of antibiotic-resistant and potentially pathogenic E. coli across water and human sources in the Chobe region. Differences across land-use contexts indicate an anthropogenic signal shaping resistance ecology along the river. The detection of multidrug-resistant phenotypes in both sources, distinct sequence type distributions, and phylogenetic proximity of a town-adjacent water isolate to human genomes underscores the importance of genomic surveillance at the human-environment interface. Water serves as a major driver of E. coli transmission, facilitating the spread of ABR and pathogenic strains between humans and the environment. These findings provide a framework for monitoring resistance emergence in shared water systems, including the Chobe River, and for guiding strategies to mitigate public health risks.