P2-125 Evaluation of Whole Genome Sequencing (WGS) to Molecularly Characterize, Serotype, and Predict Antibiograms of Salmonella spp. Isolated from Raw Chicken Products in Singapore

Tuesday, July 11, 2017
Exhibit Hall (Tampa Convention Center)
Ye Htut Zwe , Food Science and Technology Programme, National University of Singapore , Singapore , Singapore
Seow Fong Chin , Singapore Centre for Environmental Life Sciences Engineering, School of Biological Sciences, Nanyang Technological University , Singapore , Singapore
Liang Yang , Singapore Centre for Environmental Life Sciences Engineering, School of Biological Sciences, Nanyang Technological University , Singapore , Singapore
Hyun-Gyun Yuk , Korea National University of Transportation , Chungju-si , South Korea
Introduction: Salmonellosis is the most common foodborne disease in Singapore, comprising up to 75.6% of all reported foodborne diseases in 2014. It is important to be able to rapidly characterize, perform source attribution, and predict the antibiotic resistance for tracking and clinical decision making in cases of outbreaks.

Purpose: This study aimed to compare the serotype, genotype, and predicted antibiograms obtained through WGS to those of traditional methods such as multilocus sequence typing (MLST), serotyping, and disk diffusion antibiotic susceptibility testing; thereby, assessing the potential of WGS to replace these traditional methods as a faster, more reliable characterization system for an outbreak investigation.

Methods: Using the ISO method, a total of 42 Salmonella spp. strains were isolated from retail raw chicken products being sold in Singapore. The sequence types and serotypes of the isolates were determined by MLST. Their antibiotic resistances were examined by the disk diffusion method. Subsequently, the whole genome of each isolate was sequenced using an Illumina system. Using the whole genome sequence of each isolate, the presence of the antibiotic resistance genes was determined using the ResFinder 3.0.

Results: WGS pipelines were able to accurately determine the MLST sequence types and serotypes of the 42 isolates with 100% accuracy. Overall, the antibiotic resistance profile from WGS could predict the phenotype with 98.5% sensitivity and 56.3% specificity. The specificity was the lowest for aminoglycosides (11.8%) and quinolones (33.3%). Single nucleotide polymorphism (SNP) trees obtained from WGS data could further genotypically characterize S. Saintpaul (n = 17) and S. Brancaster (n = 11) isolates, which were not distinguishable by MLST.

Significance: This study indicates that WGS could be an effective method to obtain accurate sequence types and serotypes of Salmonella spp. without the hassles of laborious MLST and serotyping procedures. Molecular typing by SNP trees through WGS could offer superior resolution as compared to the traditional MLST.