Antimicrobial resistance (AMR) is a growing global health threat. Environmental compartments such as wastewater treatment plants (WWTPs) and drinking water treatment plants (DWTPs) collect a mixture of domestic, industrial, agricultural, and hospital effluents containing antibiotics, heavy metals and biocides. These chemical loads can exert selective pressure on microbial communities, promoting survival of antimicrobial-resistant bacteria (ARBs) and facilitating horizontal transfer of antimicrobial resistance genes (ARGs) via mobile elements such as plasmids, transposons and integrons. Biofilms within treatment systems may act as reservoirs that increase genetic exchange relative to planktonic populations. Although WWTPs and DWTPs reduce overall microbial loads, they are not specifically designed to eliminate ARBs or ARGs, potentially releasing them into downstream ecosystems.
This study aimed to provide an integrated genomic and phenotypic characterization of antibiotic-resistant Escherichia coli isolated across treatment stages in two WWTPs and one DWTP in metropolitan Barcelona, combining whole-genome sequencing with phenotypic testing to examine persistence mechanisms and co-selection patterns.
Sampling was performed in four campaigns during 2023 (two winter, two summer) at two WWTPs (Baix Llobregat and Gavà-Viladecans) and one DWTP (Sant Joan Despí) in the Barcelona metropolitan area. At Baix Llobregat WWTP, samples were collected from the primary inlet, secondary outlet, tertiary outlet and advanced tertiary outlet. Gavà-Viladecans WWTP sampling points included primary inlet, secondary outlet from an IFAS system, secondary outlet from an MBR system, and tertiary outlet. At Sant Joan Despí DWTP, inlet and outlet samples were collected.
Volume and concentration approaches varied by sampling point and matrix complexity. Inlet and early treatment-stage samples were processed directly from 1 L volumes. For secondary MBR, tertiary and advanced tertiary WWTP stages, 100 L were concentrated using a Rexeed™ 25-A ultrafiltration system. The DWTP outlet was concentrated from 1,000 L using the same ultrafiltration system. Ultrafiltrated samples were reduced to a final volume of 2–6 mL before microbiological analysis.
A total of 152 antibiotic-resistant E. coli strains were isolated across treatment stages. Characterization included antimicrobial susceptibility testing to define multidrug-resistant (MDR) and extensively drug-resistant phenotypes and to identify carbapenemase producers. Isolates underwent whole-genome sequencing and multilocus sequence typing to determine sequence types (STs) and to identify high-risk clones. The genomes were screened for ARGs, virulence factor genes (VFGs), integrase genes, and genes associated with biocide and heavy-metal tolerance (HMTGs). Biofilm assays were performed to assess phenotypic biofilm formation as a potential persistence mechanism.
Although bacterial loads decreased along treatment processes, AMR remained highly prevalent among the isolated strains. Reported findings indicate that 85.5% of isolates were multidrug-resistant (MDR), 5.3% were classified as extensively drug-resistant, and 11.2% were carbapenemase-producers. One MDR strain was detected at the DWTP inlet, and several MDR and high-risk clones persisted in reclaimed water sampled after treatment.
Genomic analysis revealed associations between mobile genetic elements and ARG carriage. Strains harboring integrase genes were substantially more likely (reported ranges 2.4–11.8 fold) to carry ARGs conferring resistance to sulfonamides, aminoglycosides, phenicols, trimethoprim, mercury and quaternary-ammonium compounds. Presence of bla CTX-M genes was linked to a higher likelihood (3.0–20.3 fold) of carrying VFGs. High‑risk clones were also more frequently associated with VFGs (3.2–7.0 fold), suggesting co-selection of resistance and virulence determinants.
The study detected persistence of some MDR and high‑risk E. coli clones in reclaimed water following treatment stages. The detection of an MDR strain at the DWTP inlet highlights potential routes for environmental or human exposure. These observations indicate that standard treatment steps reduced bacterial counts but did not fully eliminate clinically relevant resistant clones or mobile genetic elements associated with ARG dissemination.
Reported statistical modeling quantified increased likelihoods of ARG and VFG co-occurrence in the presence of integrases, bla CTX-M genes, and high‑risk clone backgrounds. Integrase presence correlated with multiple non–beta-lactam resistance determinants and tolerance genes for mercury and quaternary ammonium compounds, supporting the role of mobile elements and co‑selective pressures (e.g., heavy metals, biocides) in maintaining AMR in treatment systems.
Findings underscore WWTPs and DWTPs as environmental AMR reservoirs that may facilitate persistence and dissemination of MDR and high‑risk E. coli clones into downstream environments and potentially into human exposure pathways. The results support integrated One Health strategies that combine antibiotic stewardship, surveillance of environmental compartments, and targeted interventions at treatment facilities to mitigate AMR spread.
Ethics approval was not required because only environmental samples were analyzed. The paper reports that reads were deposited in the European Nucleotide Archive (Accession PRJEB102514) and that all data are publicly available at the cited Dataverse repository. Additional methodological or numerical details beyond those summarized here are present in the full article; where specifics were not reported in the provided source excerpt, they are not reproduced.
In this genomic and phenotypic surveillance of 152 antibiotic-resistant E. coli strains from two WWTPs and one DWTP in Barcelona, high rates of MDR and presence of carbapenemase producers were observed despite reductions in bacterial load through treatment. Associations between integrases, bla CTX-M, VFGs and HMTGs indicate possible co‑selection mechanisms. Persistence of MDR and high‑risk clones in reclaimed water and detection at a DWTP inlet highlight environmental and public health concerns and reinforce the need for One Health–oriented monitoring and interventions.