High-throughput quantitative PCR combined with Illumina sequencing and network analysis were used to characterize the antibiotic resistance gene (ARG) profiles in air-conditioning filters from different environments. In total, 177 ARGs comprising 10 ARG types were determined. The detectable numbers and the relative abundance of ARGs in hospitals and farms were significantly higher than those in city and village residences. Compared to hospitals, farms had a higher level of tetracycline, multidrug, integrase, and macrolide⁻lincosamide⁻streptogramin (MLS) B resistance genes but a lower level of beta-lactam resistance genes. The bl3_cpha gene was the most abundant resistance gene subtype in hospital samples with an abundance of 2.01 × 10-4 copies/16S rRNA, while a level of only 5.08 × 10-12 copies/16S rRNA was observed in farm samples. There was no significant difference in bacterial diversity among the hospitals, farms, and residences, and Proteobacteria was the most abundant phylum. Network analysis revealed that Proteobacteria and Actinobacteria were possible hosts of the beta-lactam, MLSB, aminoglycoside, multidrug, sulfonamide, and tetracycline resistance genes. The results demonstrate that ARGs exist in indoor environments and that farms and hospitals are important sources. This study provides a useful reference for understanding the distribution patterns and risk management of ARGs in indoor environments.
Keywords: air-conditioning system; antibiotic resistance genes; human health; indoor environment.