New antibiotic resistance genes and their diversity
Antibiotic resistance is increasing worldwide and is considered a severe threat to public health. Often, antibiotic resistance is caused by antibiotic resistance genes, of which many are hypothesized to have been transferred into human pathogens from environmental bacteria. It is, therefore, of great importance to explore bacterial communities to identify new antibiotic resistance genes before they reach clinical settings. The six papers presented in this thesis aim to identify new antibiotic resistance genes in large genomic and metagenomic datasets and to place them in an evolutionary context. In Paper I, a new method for the identification and reconstruction of new antibiotic resistance genes directly from fragmented metagenomic data was developed and was shown to outperform other methods significantly. In Papers II and III, novel genes of the clinically important class metallo-β-lactamases were identified. By analyzing metagenomes and bacterial genomes, 96 novel putative metallo-β-lactamase genes were predicted. In Paper IV, the diversity and phylogeny of the metallo-β-lactamases were further investigated. The results showed that the genes mainly clustered based on the taxonomy of the host species and that many of the mobile metallo-β-lactamases potentially were mobilized from species of the phylum Proteobacteria. In Paper V, the aim was to identify new genes providing resistance to the antibiotic class tetracyclines. A total of 195 gene families were predicted, of which 164 were new putative tetracycline resistance genes. Finally, in Paper VI, we searched for and predicted 20 novel putative quinolone resistance (qnr) genes from a large amount of metagenomic data. Throughout the thesis, a total of 54 novel genes have been functionally verified in Escherichia coli, of which 37 expressed the predicted phenotype. The results of this thesis provide deeper insights into the diversity and evolutionary history of three major classes of antibiotic resistance genes. It also provides new methodologies for efficient and reliable identification of new resistance genes in genomic and metagenomic data.
hidden Markov model