DNA-Sequence Based Typing of the Cronobacter Genus Using MLST, CRISPR-cas Array and Capsular Profiling

Abstract

The <i>Cronobacter</i> genus is composed of seven species, within which a number of pathovars have been described. The most notable infections by <i>Cronobacter</i> spp. are of infants through the consumption of contaminated infant formula. The description of the genus has greatly improved in recent years through DNA sequencing techniques, and this has led to a robust means of identification. However some species are highly clonal and this limits the ability to discriminate between unrelated strains by some methods of genotyping. This article updates the application of three genotyping methods across the <i>Cronobacter</i> genus. The three genotyping methods were multilocus sequence typing (MLST), capsular profiling of the <i>K</i>-antigen and colanic acid (CA) biosynthesis regions, and CRISPR-<i>cas</i> array profiling. A total of 1654 MLST profiled and 286 whole genome sequenced strains, available by open access at the PubMLST <i>Cronobacter</i> database, were used this analysis. The predominance of <i>C. sakazakii</i> and <i>C. malonaticus</i> in clinical infections was confirmed. The majority of clinical strains being in the <i>C. sakazakii</i> clonal complexes (CC) 1 and 4, sequence types (ST) 8 and 12 and <i>C. malonaticus</i> ST7. The capsular profile K2:CA2, previously proposed as being strongly associated with <i>C. sakazakii</i> and <i>C. malonaticus</i> isolates from severe neonatal infections, was also found in <i>C. turicensis</i>, <i>C. dublinensis</i> and <i>C. universalis</i>. The majority of CRISPR-<i>cas</i> types across the genus was the I-E (Ecoli) type. Some strains of <i>C. dublinensis</i> and <i>C. muytjensii</i> encoded the I-F (Ypseudo) type, and others lacked the <i>cas</i> gene loci. The significance of the expanding profiling will be of benefit to researchers as well as governmental and industrial risk assessors.