Prediction of antimicrobial resistance in clinical Campylobacter jejuni isolates from whole-genome sequencing data

• Published in: European journal of clinical microbiology & infectious diseases 2021-04, Vol.40 (4), p.673-682
• Main Authors: Dahl, Louise Gade, Joensen, Katrine Grimstrup, Østerlund, Mark Thomas, Kiil, Kristoffer, Nielsen, Eva Møller
• Format: Article
• Language: English
• Subjects: Antimicrobial agents; Antimicrobial resistance; Bacterial diseases; Biomedical and Life Sciences; Biomedicine; Campylobacter; Campylobacter jejuni; Ciprofloxacin; Correlation; Dilution; Disease control; Drug resistance; Epidemiology; Erythromycin; Gastroenteritis; Gene sequencing; Genes; Genomes; Genotypes; Gentamicin; Internal Medicine; Medical Microbiology; Mutation; Nalidixic acid; Original; Original Article; Phenotypes; Public health; rRNA; rRNA 23S; Streptomycin; Surveillance; Whole genome sequencing
• ISSN: 0934-9723; 1435-4373
• DOI: 10.1007/s10096-020-04043-y

Campylobacter jejuni is recognised as the leading cause of bacterial gastroenteritis in industrialised countries. Although the majority of Campylobacter infections are self-limiting, antimicrobial treatment is necessary in severe cases. Therefore, the development of antimicrobial resistance (AMR) in Campylobacter is a growing public health challenge and surveillance of AMR is important for bacterial disease control. The aim of this study was to predict antimicrobial resistance in C. jejuni from whole-genome sequencing data. A total of 516 clinical C. jejuni isolates collected between 2014 and 2017 were subjected to WGS. Resistance phenotypes were determined by standard broth dilution, categorising isolates as either susceptible or resistant based on epidemiological cutoffs for six antimicrobials: ciprofloxacin, nalidixic acid, erythromycin, gentamicin, streptomycin, and tetracycline. Resistance genotypes were identified using an in-house database containing reference genes with known point mutations and the presence of resistance genes was determined using the ResFinder database and four bioinformatical methods (modified KMA, ABRicate, ARIBA, and ResFinder Batch Upload). We identified seven resistance genes including tet(O) , tet(O/32/O) , ant(6)-Ia , aph(2″)-If , blaOXA , aph(3′)-III , and cat as well as mutations in three genes: gyrA , 23S rRNA , and rpsL . There was a high correlation between phenotypic resistance and the presence of known resistance genes and/or point mutations. A correlation above 98% was seen for all antimicrobials except streptomycin with a correlation of 92%. In conclusion, we found that WGS can predict antimicrobial resistance with a high degree of accuracy and have the potential to be a powerful tool for AMR surveillance.