Riverine antibiotic resistome along an anthropogenic gradient

Abstract

The introduction of antibiotic-resistant bacteria into riverine systems through the discharge of wastewater treatment plant (WWTP) effluent and agricultural waste poses significant health risks. Even when not pathogenic, these bacteria can act as reservoirs for antibiotic resistance genes (ARGs), transferring them to pathogens that infect humans and animals. In this study, we used fluorescence in situ hybridization, qPCR, and metagenomics to investigate how anthropogenic activities affect microbial abundance and the resistome along the Holtemme River, a small river in Germany, from near-pristine to human-impacted sites. Our results showed higher bacterial abundance, a greater absolute and relative abundance of ARGs, and a more diverse ARG profile at the impacted sites. Overall, the ARG profiles at these sites reflected antibiotic usage in Germany, with genes conferring resistance to drug classes such as beta-lactams, aminoglycosides, folate biosynthesis inhibitors, and tetracyclines. There were also variations in the ARG profiles of the impacted sites. Notably, there was a high abundance of the oxacillin resistance gene OXA-4 at the downstream site in the river. In the metagenome assembly, this gene was associated with a contig homologous to small plasmids previously identified in members of the Thiotrichaceae. The likely in-situ host of the putative plasmid was a close relative of Thiolinea (also known as Thiothrix) eikelboomii, a prominent member of WWTP microbiomes worldwide. Our results show that the effluent from WWTPs can introduce bacteria into the environment that act as shuttle systems for clinically relevant ARG.

Keywords: anthropogenic activities; antibiotic resistome; fluorescence in situ hybridization; metagenomic sequencing; riverine system.

Figure 1

Representative fluorescence images and cell counts from CARD-FISH. (A–C) Overlay images of DAPI staining and the 16S rRNA-targeting EUBI-III probe mix at Sites 1, 2, and 3, respectively; (D) Bacterial and total cell abundances at the sampling sites. Total cell abundances are shown in blue, while bacterial cells appear in pink, resulting from the overlay of the Alexa594 tyramide dye (red) and DAPI (blue) signals.

Figure 2

Relative abundance of various antimicrobial resistance gene types and subtypes detected at three sampling sites along the Holtemme River (site 1 is near-pristine, while sites 2 and 3 are influenced by effluent from wastewater treatment plants). The drug class referred to as “others” includes aminocoumarin, antibacterial free fatty acids, disinfectants, intercalating dyes, elfamycin, fosfomycin, mupirocin, nitroimidazole, nucleoside, phenicol, pleuromutilin, rifamycin, and triclosan.

Figure 3

Relative abundance of the 30 most prevalent antimicrobial resistance genes from three sampling sites along the Holtemme River (site 1 is nearly pristine; sites 2 and 3 are both affected by the effluent from wastewater treatment plants).

Figure 4

Contigs and antibiotic resistance genes (ARGs) associated with each contig. (A) Reads per kilobase per million reads (RPKM) of contigs and ARGs indicate that the ARGs in each row of the “ARG on contigs” heatmap are found on the corresponding contigs in the heatmap on the left. The suffixes “_1”, “_2”, and “_3” in some contig names denote those contigs that contain multiple ARGs; (B) gene arrangement on contig Site 3_267832_length_7668; (C) alignment of contig Site 3_267832_length_7668 with GenBank sequences (the size of pTunz5 in the alignment is 7.9 kb, which is half the size of the plasmid in GenBank).