Metagenomic profiling of antibiotic resistance genes and their associations with the bacterial community along the Kanda River, an urban river in Japan

Antibiotic resistance genes (ARGs) present in urban rivers have the potential to disseminate antibiotic-resistant bacteria into other environments, posing significant threats to both ecological and public health. Although metagenomic analyses have been widely employed to detect ARGs in rivers, our understanding of their dynamics across different seasons in diverse watersheds remains limited. In this study, we performed a comprehensive genomic analysis of the Kanda River in Japan at 11 sites from upstream to estuary throughout the year to assess the spread of ARGs and their associations with bacterial communities. Analysis of 110 water samples using the 16S rRNA gene revealed variations in bacterial composition corresponding to seasonal changes in environmental parameters along the river. Shotgun metagenomics-based profiling of ARGs in 44 water samples indicated higher ARG abundance downstream, particularly during the summer. Weighted gene co-expression network analysis (WGCNA) linking bacterial lineages and ARGs revealed that 12 ARG subtypes co-occurred with 128 amplicon sequence variants (ASVs). WGCNA suggested potential hosts for ErmB, ErmF, ErmG, tetQ, tet (W/N/W), aadA2, and adeF, including gut-associated bacteria (e.g., Prevotella, Bacteroides, Arcobacter) and indigenous aquatic microbes (e.g., Limnohabitans and C39). In addition, Pseudarcobacter (a later synonym of Arcobater) was identified as a host for adeF, which was also confirmed by single cell genomics. This study shows that ARG distribution in urban rivers is affected by seasonal and geographical factors and demonstrates the importance of monitoring rivers using multiple types of genome sequencing, including 16S rRNA gene sequencing, metagenomics, and single cell genomics.