Abstract
Antimicrobial resistance in Shigella spp. is a global public health concern. In this study, the AMR phenotypic profiles of 10 kinds of antibiotics were compared with the genotypic profiles using genomic analysis of 218 Shigella isolates from Taiyuan City, Shanxi Province, China, 2005 to 2016. Core genome Multilocus Sequence Typing (cgMLST) based on the EnteroBase Escherichia/Shigella scheme was used to obtain the genetic relatedness of Shigella isolates. Multiple-drug resistance was observed in 96.79% Shigella spp., and the resistance to antimicrobial agents varied between S. flexneri and S. sonnei. The genotypic results correlated well with the phenotypic profiles with concordance rates of 96.42% and 94.50% in S. flexneri and S. sonnei isolates, respectively, from Taiyuan City, Shanxi Province. The sensitivity and specificity of the genotypic antimicrobial susceptibility testing (AST) were 97.56% and 95.34% for S. flexneri, and 95.65% and 93.31% for S. sonnei isolates, respectively. A discrepancy of genotypic and phenotypic AST results existed in some cephalosporin- and azithromycin-resistant Shigella isolates; there were no clear resistance patterns to predict ciprofloxacin resistance. There were major discrepancies between genotypic and phenotypic AST in the genotypically resistant but phenotypically susceptible isolates. The drug-resistance patterns and essential drug-resistance genes to predict the phenotypic drug-resistant profiles were the discrepancies between S. flexneri and S. sonnei isolates. Phylogenetic analysis showed that isolates of the same cluster but with different antibiotic-resistance gene patterns occurred because of the loss or gain of antibiotic-resistance genes located in the plasmids and multidrug-resistance islands. IMPORTANCE Antimicrobial resistance in Shigella spp. has become a global public health concern. In this study, we identified the antimicrobial susceptibility testing (AST) characteristics based on genomic sequences of 218 Shigella isolates and analyzed the correlation between genotypic and phenotypic antibiotic resistance profiles of Shigella spp., especially for fluoroquinolone, macrolides, and third-generation cephalosporins. Our results show that the genotypic results correlated with the phenotypic profiles with concordance rates of 96.42% and 94.50% in S. flexneri and S. sonnei isolates, respectively. The drug-resistance patterns and essential drug-resistance genes to predict the phenotypic drug-resistant profiles of S. flexneri and S. sonnei isolates in Taiyuan city were distinct. The discrepancy between genotypic and phenotypic AST was considerable in the genotypically resistant but phenotypically susceptible isolates. The information on drug resistance and resistance genes in this study can offer more details on the prevalence of drug resistance of Shigella spp.