Functional genomics approach to identify antibiotic resistance genes

The risk of antibiotic resistance development in the clinic is a well-known public health problem. Big data and large-scale sequencing projects have revolutionized the field, providing a greater understanding of the genomic and molecular aspects of antibiotic resistance genes (ARGs) and how they are distributed. However, most studies that use genomic tools to detect the presence of ARGs do not perform the functional validation of these findings. Therefore, the hypothesis was raised that in silico methodologies are not able to identify all ARGs active in vivo. To test this hypothesis, an approach based on functional screenings was developed. Strains of clinical bacteria highly resistant to various antibiotics (Escherichia coli, Klebsiella pneumoniae and Morganella morganii) were selected, and libraries of the genomic material of these bacteria were built. Initially, it was proposed that this approach would be able to confirm the function of the ARGs identified through bioinformatics approaches, as well as assist in the discovery of ARGs whose sequences are different from the known ARGs. Three genomic libraries were tested against three beta-lactam antibiotics (amoxicillin, oxacillin and penicillin G) and it was possible to functionally confirm the activity of 6 of the 15 beta-lactam ARGs identified in silico for these strains, with at least two of them present in plasmids - blaKPC-2 and blaLAP-2. Also, the genomes of these bacteria were sequenced by new generation DNA sequencing, resulting in the first genomic description of a strain of Morganella morganii from clinical isolates from South America. The results presented here contribute to the understanding of the genomic aspects of the determinants of antibiotic resistance and its dispersion in pathogenic bacteria underrepresented in Brazil and in the world. (AU)