WP 115 Whole Genome Sequencing as a Tool for Multilocus Sequence Typing of Neisseria gonorrhoeae Isolates

Tuesday, June 10, 2014
International Ballroom
Carolyn Caron, BSc1, Anthony Kusalik, PhD2, Timothy D Read, PhD3, Sinisa Vidovic, PhD1 and Jo-Anne R Dillon, PhD4, 1University of Saskatchewan, 2Computer Science, University of Saskatchewan, 3Department of Human Genetics, Emory University, 4Microbiology and Immunology, College of Medicine, University of Saskatchewan, Saskatoon, SK, Canada

Background: Multi-locus sequence typing (MLST) of the pathogen Neisseria gonorrhoeae typically includes amplifying gene sequence fragments of housekeeping genes obtained by PCR followed by DNA sequence analysis.  However, whole genome sequencing (WGS) of organisms is becoming increasingly common when working with prokaryotes.  This study compares the results of using sequence information from WGS assembly for MLST versus a previous analysis using the traditional PCR approach.

Methods: The genomes of 25 isolates of N. gonorrhoeae were sequenced by Illumina Hi-Seq and contigs were assembled using the CLC Genomics Workbench.  MLST analysis for fumC, gdh, glnA, gnd, pilA, pyrD and serC was conducted using contigs obtained for three predominant clones, comprising 23 gonococcal isolates from Saskatchewan, as well as two other strains from different geographic areas and collected several decades previously.  The strains were analyzed for population structure and the evolution of antibiotic resistance using the START2 software.  Visualization and analysis of clonal complexes from allelic profile data was performed using eBURST. 

Results: The distribution of strain types using WGS data was generally similar to that of the original study, though clonal expansion of a ciprofloxacin resistant strain type was evident in the new results. The two unrelated strains were not distinguished as outliers, suggesting possible lineage of the Saskatchewan strains from these older strains.  Use of WGS data presented various unique problems, but also showed the potential for enhancing MLST analysis.  For example, additional variable regions can be utilized for inferring evolutionary relationships when WGS data is used.  The study also identifies pitfalls to be avoided when assembling WGS reads destined for MLST analysis.

Conclusions: This study is timely given the continued improvements to sequencing technology and assembly software, and the prospect of WGS data becoming the standard for MLST analysis.