Metagenomic profiling and temporal dynamics of antimicrobial resistance genes across hospital wastewaters in Dhaka, Bangladesh

Hospital wastewater (HWW) in Bangladesh is often discharged into municipal sewage and water bodies without adequate treatment. It raises concern that hospitals may act as reservoirs and amplification points for clinically important antimicrobial resistance genes (ARGs). However, culture-independent data on the composition, mobility, and seasonal behavior of these genes in Bangladeshi hospital wastewater remain limited. Here, we combined single monsoon time-point baseline shotgun metagenomic sequencing with longitudinal qPCR to characterize wastewater from three tertiary hospitals in Dhaka and a nearby community wastewater line across three seasonal sampling points. Across sites based on baseline metagenomic profiling, the putative resistome was dominated by aminoglycoside, sulfonamide, macrolide–lincosamide–streptogramin (MLS), tetracycline, and β-lactam resistance, primarily associated with Actinomycetota and Pseudomonadota. Hospital effluents contain putative clinically important genes, including the vanA cluster, blaVIM , blaOXA-232 , blaIMP , and blaKPC . Metagenomic analysis linked many putative ARGs to plasmid-predicted contigs, transposition, or integration modules, while virulence profiling showed enrichment of putative genes related to adhesion, motility, iron acquisition, and secretion, particularly in MHWW-2. Exploratory network co-occurrence from baseline metagenomics primarily suggested fecal-derived gut commensals, alongside pathogens, as reservoirs and putative vectors for ARG input into the wastewater system. Longitudinal qPCR showed higher absolute loads of selected ARGs at the pre-winter sampling point, and the class 1 integron gene ( intI1LC ) tracked total ARG burden across seasonal transitions. These findings identify hospital wastewater in Dhaka primarily as a dynamic reservoir of clinically important and potentially mobile resistance determinants, and support its use as a target for wastewater-based AMR surveillance.

IMPORTANCE

Wastewater surveillance can provide early warning of AMR, but in Bangladesh, it is rarely applied beyond culture or single-gene PCR, limiting detection of emerging, mobile, and facility-specific threats. This study integrates metagenomics with longitudinal qPCR to connect which resistance genes are present with who carries them and how they change seasonally across multiple hospitals. We show that hospital effluents contain a shared background resistome yet disproportionately concentrate last-line resistance in taxa that include clinically important lineages, with signals rising toward pre-winter. By placing high-risk putative ARGs in putative plasmid, integron, and transposase contexts and alongside virulence repertoires from baseline metagenomic profiling, our results identify plausible genomic mechanisms that may contribute to spread beyond hospitals. These findings position hospital effluents in Dhaka preliminarily as dynamic reservoirs of clinically important, mobile AMR and support wastewater-based epidemiology for sustained surveillance and risk mitigation.