Аңдатпа
Introduction Chronic heavy metal contamination can restructure soil microbiomes and may co-select for antibiotic resistance, yet genome-resolved evidence from industrial soils remains limited. Methods In this study, we applied Oxford Nanopore long-read metagenomic sequencing to soil samples collected across industrially influenced sites in East Kazakhstan to characterize strain-level community composition, profile antibiotic resistance genes and metal resistance genes, and relate these patterns to soil physicochemical properties. Results Across all samples, we identified 3,053 strains, with Actinobacteria and Proteobacteria together accounting for 94.1% of the total community. Heavy metal concentrations varied markedly among sites. The resistome comprised antibiotic resistance genes from several drug classes and 238 distinct metal resistant genes, with aminoglycoside, glycopeptide, and multidrug resistance dominating the antibiotic resistance gene profile, while czcA, ruvB, arsM , and arsT were among the most abundant Metal resistant genes. Multivariate analyses showed that heavy metals, particularly Zn, significantly shaped microbial community structure as well as antibiotic resistance gene and metal resistance gene composition, and redundancy analysis identified Zn and soil pH as the principal environmental drivers. Network analyses further revealed that Bradyrhizobium icense and Conexibacter woesei acted as key super-hosts linking ARGs and MRGs, supporting heavy metal-driven co-selection within the soil microbiome. Discussion Together, these findings show that long-read genome-resolved metagenomics can uncover how chronic industrial pollution maintains metal-adapted microbial communities while promoting the persistence and potential dissemination of antibiotic resistance in soil ecosystems.