Antimicrobial profiles and resistance genes in Salmonella isolates in diarrheal patients in Hainan province in 2020

  Objective  To characterize the drug resistance of Salmonella in Hainan province, and to evaluate the accuracy of predicting drug resistant phenotype by drug resistance genes.

  Methods  Total 63 Salmonella strains were collected from human in Hainan Province and were subjected to serological identification, antimicrobial susceptibility testins and whole genome sequencing. The association between the drug resistant phenotype and the corresponding resistance genes were analyzed.

  Results  By measuring the sensitivity of all tested strains to 11 kinds of antibiotics, we found that the resistance rates of 63 strains to ampicillin and cotrimoxazole were high, more than 50.00%, and the resistance rate to the third generation cephalosporin was more than 15.00%. There was high level multiple drug resistance with the rate reaching 44.44%. We also found the strains carried a variety of drug resistance genes, especially aminoglycoside resistance genes which were harbored in all strains (100.00%). The association analysis showed that the prediction results of ampicillin and cefotaxime using their encoding resistance genes were in good agreement with the phenotype, and the prediction accuracy rates reached 88.89% and 98.41%, respectively. When the single parC gene with mutation in the strain was excluded, the prediction results of quinolone resistance genes were well consistent with the phenotype, and the Kappa value is beyond 0.50.

  Conclusion  Salmonella in Hainan province is highly resistant to traditional antibiotics and has high level of multiple drug resistance. The drug resistance gene information obtained by second-generation sequencing has high accuracy in predicting β-lactams resistance phenotype, while the prediction accuracy of quinolone resistance genes is relatively lower. This provides a sight of replacing antimicrobial susceptibility test by detecting resistance genes for these two types of antibiotics. However, it is still needed to expand the number of strains for more accurate and detailed correlation studies.