The occurrence and removal patterns of 24 antimicrobial agents and antimicrobial resistant determinants namely 6 antibiotic resistance genes (ARGs) and 2 mobile genetic elements (MGEs), and the fecal indicator E. coli were investigated in three full-scale wastewater treatment plants. Their waterlines and biosolids lines (including secondary treatment based on both granular and activated sludge) were sampled monthly throughout one year. Samples were analyzed by means of LC-MS/MS, qPCR and cell enumeration, respectively. The influence of rainfall, temperature, and turbidity on the occurrence and removal of the aforementioned agents was assessed through statistical linear mixed models. Ten of the antimicrobial agents (macrolides, fluoroquinolones, tetracyclines, and sulfonamides) were commonly found in influent in concentrations of 0.1-2 µg L⁻¹, and the predominant ARGs were ermB and sul1 (6.4 and 5.9 log₁₀ mL⁻¹ respectively). Warmer temperatures slightly reduced gene concentrations in influent whilst increasing that of E. coli and produced an uneven effect on the antimicrobial concentrations across plants. Rainfall diluted both E. coli (-0.25 logs, p < 0.001) and antimicrobials but not genes. The wastewater treatment reduced the absolute abundance of both genes (1.86 logs on average) and E. coli (2.31 logs on average). The antimicrobials agents were also partly removed, but 8 of them were still detectable after treatment, and 6 accumulated in the biosolids. ARGs were also found in biosolids with patterns resembling those of influent. No significant differences in the removal of antimicrobials, genes and E. coli were observed when comparing conventional activated sludge with aerobic granular sludge. Irrespective of the type of sludge treatment, the removal of genes was significantly reduced with increasing hydraulic loads caused by rainfall (-0.35 logs per ∆ average daily flow p < 0.01), and slightly decreased with increasing turbidity (-0.02 logs per ∆1 nephelometric turbidy unit p < 0.05).

Plasmid-mediated dissemination of antibiotic resistance among fecal Enterobacteriaceae in natural ecosystems may contribute to the persistence of antibiotic resistance genes in anthropogenically impacted environments. Plasmid transfer frequencies measured under laboratory conditions might lead to overestimation of plasmid transfer potential in natural ecosystems. This study assessed differences in the conjugative transfer of an IncP-1 (pKJK5) plasmid to three natural Escherichia coli strains carrying extended-spectrum beta-lactamases, by filter mating. Matings were performed under optimal laboratory conditions (rich LB medium and 37°C) and environmentally relevant temperatures (25, 15 and 9°C) or nutrient regimes mimicking environmental conditions and limitations (synthetic wastewater and soil extract). Under optimal nutrient conditions and temperature, two recipients yielded high transfer frequencies (5 × 10^–1) while the conjugation frequency of the third strain was 1000-fold lower. Decreasing mating temperatures to psychrophilic ranges led to lower transfer frequencies, albeit all three strains conjugated under all the tested temperatures. Low nutritive media caused significant decreases in transconjugants (−3 logs for synthetic wastewater; −6 logs for soil extract), where only one of the strains was able to produce detectable transconjugants. Collectively, this study highlights that despite less-than-optimal conditions, fecal organisms may transfer plasmids in the environment, but the transfer of pKJK5 between microorganisms is limited mainly by low nutrient conditions.

Wastewater treatment plants (WWTPs), linking human fecal residues and the environment, are considered as hotspots for the spread of antimicrobial resistance (AMR). In order to evaluate the role of WWTPs and underlying operational parameters for the removal of AMR, the presence and removal efficiency of a selected set of 6 antimicrobial resistance genes (ARGs) and 2 mobile genetic elements (MGEs) was evaluated by means of qPCR in influent and effluent samples from 62 Dutch WWTPs. The role of possible factors impacting the concentrations of ARGs and MGEs in the influent and their removal was identified through statistical analysis. ARGs and the class I integron-integrase gene (intI1) were, on average, removed to a similar extent (1.76 log reduction) or better (+0.30–1.90 logs) than the total bacteria (measured as 16S rRNA gene). In contrast, broad-host-range plasmids (IncP-1) had a significantly increased (p < 0.001) relative abundance after treatment. The presence of healthcare institutions in the area served did only slightly increase the concentrations of ARGs or MGEs in influent. From the extended panel of operational parameters, rainfall, increasing the hydraulic load of the plant, most significantly (p < 0.05) affected the treatment efficiency by decreasing it on average −0.38 logs per time the flow exceeded the average daily flow. Our results suggest that overall, WWTP treatments do not favor the proliferation of the assessed resistance genes but might increase the relative abundance of broad-host-range plasmids of the IncP-1 type.