Frequency of Antimicrobial Resistance Genes in Salmonella From Brazil by in silico Whole-Genome Sequencing Analysis: An Overview of the Last Four Decades

Salmonella is a leading human pathogen and a significant public health concern worldwide. Massive food production and distribution have contributed to this pathogen dissemination, which, combined with antimicrobial resistance (AMR), creates new control challenges in food safety. The development of AMR is a natural phenomenon and can occur in the bacterial evolutionary process. However, the overuse and the misuse of antimicrobial drugs in humans and in animals have increased AMR selective pressure. In Brazil, there is an accuracy lack in AMR frequency in Salmonella because too many isolates are under-investigated for genetic and phenotypic AMR by the Brazilian health authorities and the research community. This underreporting situation makes the comprehension of the real level of Salmonella AMR in the country difficult. The present study aimed to use bioinformatics tools for a rapid in silico screening of the genetic antimicrobial resistance profile of Salmonella through whole-genome sequences (WGS). A total of 930 whole-genome sequences of Salmonella were retrieved from the public database of the National Biotechnology Information Center (NCBI). A total of 65 distinct resistance genes were detected, and the most frequent ones were tet(A), sul2, and fosA7. Nine point mutations were detected in total, and parC at the 57 position (threonine → serine) was the highest frequent substitution (26.7%, 249/930), followed by gyrA at the 83 position (serine → phenylalanine) (20.0%, 186/930) and at the 87 position (aspartic acid → asparagine) (15.7%, 146/930). The in silico prediction of resistance phenotype showed that 58.0% (540/930) of the strains can display a multidrug resistance (MDR) profile. Ciprofloxacin and nalidixic acid were the antimicrobial drugs with the highest frequency rates of the predicted phenotype resistance among the strains. The temporal analysis through the last four decades showed increased frequency rates of antimicrobial resistance genes and predicted resistance phenotypes in the 2000s and the 2010s when compared with the 1980s and 1990s. The results presented herein contributed significantly to the understanding of the strategic use of WGS associated with in silico analysis and the predictions for the determination of AMR in Salmonella from Brazil.