Metagenomics-based source attribution of antimicrobial resistance in wastewater for improved epidemiological risk assessment

Wastewater-based epidemiology (WBE) offers a powerful approach for monitoring antimicrobial resistance (AMR) at the population level. However, distinguishing between human gut-derived and sewer-derived AMR-carrying organisms remains a key challenge for accurate surveillance and risk assessment. In this study, we used genome-resolved metagenomics to distinguish human gut-derived organisms, and their associated antimicrobial resistance genes (ARGs), mobile genetic elements (MGEs) and virulence-associated determinants (VFs), from taxa endemic to the sewer network. We applied this approach to wastewater samples collected from three hospital outflows (near-source healthcare sites), as well as from untreated influent and final treated effluent at the corresponding municipal wastewater treatment plants serving the surrounding communities. Along the wastewater pathway, microbial communities progressively shifted from human gut-associated to sewer adapted taxa; consequently, the final treated effluent was dominated by sewer-adapted taxa. Human gut-derived taxa were further examined in detail: 84% carried ARGs and VFs, predominantly within Bacillota and Bacteroidota; all gut-associated Pseudomonadota also harboured multiple ARGs, VFs and MGEs. Opportunistic-pathogen taxa of gut origin (Escherichia coli, Klebsiella spp., E. faecium) accounted for a substantial fraction of ARGs in hospital wastewater. Combined sewer overflow (CSO) events may allow these carriers to bypass wastewater treatment and reach receiving waters, posing public health risks. This genome-resolved framework strengthens WBE by resolving human-derived contributions for surveillance and risk assessment.