Defining and Evaluating a Core Genome Multilocus Sequence Typing Scheme for Whole-Genome Sequence-Based Typing of Klebsiella pneumoniae

Abstract

At present, the most used methods for <i>Klebsiella pneumoniae</i> subtyping are multilocus sequence typing (MLST) and pulsed-field gel electrophoresis (PFGE). However, the discriminatory power of MLST could not meet the need for distinguishing outbreak and non-outbreak isolates and the PFGE is time-consuming and labor-intensive. A core genome multilocus sequence typing (cgMLST) scheme for whole-genome sequence-based typing of <i>K. pneumoniae</i> was developed for solving the disadvantages of these traditional molecular subtyping methods. Firstly, we used the complete genome of <i>K. pneumoniae</i> strain HKUOPLC as the reference genome and 907 genomes of <i>K. pneumoniae</i> download from NCBI database as original genome dataset to determine cgMLST target genes. A total of 1,143 genes were retained as cgMLST target genes. Secondly, we used 26 <i>K. pneumoniae</i> strains from a nosocomial infection outbreak to evaluate the cgMLST scheme. cgMLST enabled clustering of outbreak strains with <10 alleles difference and unambiguous separation from unrelated outgroup strains. Moreover, cgMLST revealed that there may be several sub-clones of epidemic ST11 clone. In conclusion, the novel cgMLST scheme not only showed higher discriminatory power compared with PFGE and MLST in outbreak investigations but also showed ability to reveal more population structure characteristics than MLST.