Response of viable bacteria to antibiotics in aerobic granular sludge: Resistance mechanisms and behaviors, bacterial communities, and driving factors

The assessment of antimicrobial resistance (AMR) risk by DNA-based techniques mainly relies on total bacterial DNA. In this case, AMR risk recognition is restricted to the genotype level, lacking crucial phenotypic infor-mation, such as the distribution of antibiotic resistance genes (ARGs) and mobile genetic elements (MGEs) in dead and viable bacteria. This limitation hinders the recognition of AMR behavior. Herein, based on propidium monoazide (PMA) shielding method, this work firstly quantified the intracellular ARGs/MGEs in viable and dead bacteria, and the impact of viable bacteria composition on the formation of intracellular/extracellular polymeric substance-related /cell-free ARGs (i/e/cARGs) and MGEs (i/e/cMGEs) in aerobic granular sludge (AGS). The shielding efficiency of PMA against dead bacteria was optimized to be as high as 97.5% when the MLSS of AGS was 2.0 g/L. Under antibiotic stimulation, 29.0% similar to 49.0% of iARGs/iMGEs were carried by viable bacteria, and the remaining proportion were carried by dead bacteria. 18 out of the top 20 dominant genera showed a change in abundance by more than 1% after PMA treatment. 29 viable hosts were identified to associate with 52 iARGs, of which 28 and 15 hosts were also linked to 40 eARGs and 26 cARGs. Also, partial least-squares path model and variance partitioning analysis disclosed that viable bacteria and i/e/cMGEs had a positive effect on i/e/cARGs, with both contributing as much as 64.5% to the total ARGs enrichment. These results better visualized the AMR risk carried by viable bacteria and the categories of viable hosts. This work provides a novel insight into analyzing the actual AMR risk and viable hosts, helping to the reduction and control of AMR in wastewater treatment plants.