Bioprospecção de bacteriófagos para o combate a bactérias resistentes a antibióticos a partir de dados metagenômicos

Abstract

The growing problem of bacterial resistance to antibiotics, aggravated by the indiscriminate use of these drugs in human and veterinary medicine, has demanded the development of new therapeutic strategies. Among these strategies, bacteriophages (phages) stand out as a promising alternative due to their specificity and effectiveness in infecting and eliminating pathogenic bacteria, including multidrug-resistant strains. The use of metagenomic sequencing to study microbial communities can also aid in the detection of bacteriophage sequences. In this sense, taking advantage of the high biodiversity present in the Amazon biome, a shotgun metagenomic study was conducted in the city of Bragança, Pará, with water samples collected at five different points along the Cereja River (BP1 – BP5), in triplicate, from the source to a more contaminated area. After metagenomic DNA extraction, the samples were sequenced using the Illumina NovaSeq 6000 platform, and the data were preprocessed with fastp followed by assembly with metaSpades. Kraken2, Krakentools, and Microeco were used for taxonomic characterization and diversity analysis of viruses and bacteria. Correlations with resistance profiles and taxonomic data were normalized and correlated using redundancy analysis (RDA) to identify the point of greatest environmental impact, which showed BP4 as the most contaminated. Contigs from the 5 points were analyzed by VIBRANT, VirSorter2, and DeepVirFinder, and the identified viral sequences were selected based on completeness (> 90%), contamination (< 10%), and viral quality criteria, with provirus removal based on CheckV and ViralComplete results, resulting in 19, 14, 11, 11, and 6 viral sequences selected for points BP1 to BP5, respectively. In parallel, metagenomic "metagenomic assembled genomes" (MAGs) were generated and filtered, and taxon identification was performed using the kMetaShot program. An inferential comparison was made to detect MAGs that would be likely hosts of the selected viral sequences. Point BP4, being the most contaminated and meeting these criteria, was selected and its phage-host relationship was analyzed. They were evaluated for the presence of antimicrobial resistance genes (AMRs) and defense systems. Phages with likely hosts of identified MAGs from Citrobacter, Acinetobacter, and Enterobacter were found. No AMR genes were identified in the viral sequences, but AMRs were identified in 9 of 10 MAGs selected as likely bacterial hosts, indicating potentially effective phages against antibiotic-resistant pathogenic bacteria. Functional analyses also showed that selected MAGs from the genus Acinetobacter exhibited a greater diversity of phage defense systems and identification of Cas genes, with lytic phages identified for this genus; and MAGs with a higher presence of AMR genes showed identification of the MazEF toxin-antitoxin system, and greater identification of the restriction-modification (RM) system in Enterobacter, with temperate phages identified for this genus. Although further studies are needed to validate and strengthen the findings, the results point to a promising approach to guide phage bioprospecting in contaminated environments and predict possible phage-host relationships so that more precise experimental analyses can be performed.