Abstract
BACKGROUND: Cronobacter species are associated with severe foodborne infections in neonates and infants, with particular pathovars associated with specific clinical presentations. However, before 2008 the genus was regarded as a single species named Enterobacter sakazakii which was subdivided into 8 phenotypes. This study re-analyzed, using multi-locus sequence typing (MLST) and whole genome sequence with single nucleotide polymorphism analysis (WGS-SNP), 52 strains which had been identified as Enterobacter sakazakii as according to the convention at the time of isolation. These strains had been isolated from dairy product imports into China from 9 countries between 2005 and 6. Bioinformatic analysis was then used to analyze the relatedness and global dissemination of these strains. RESULT: FusA allele sequencing revealed that 49/52 strains were Cronobacter sakazakii, while the remaining 3 strains were Escherichia coli, Enterobacter cloacae, and Franconibacter helveticus. The C. sakazakii strains comprised of 8 sequence types (STs) which included the neonatal pathovars ST1, ST4 and ST12. The predominant sequence type was ST13 (65.3%, 32/49) which had been isolated from dairy products imported from 6 countries. WGS-SNP analysis of the 32 C. sakazakii ST13 strains revealed 5 clusters and 5 unique strains which did not correlate with the country of product origin. CONCLUSION: The mis-identification of E. coli, E. cloacae and F. helveticus as Cronobacter spp. reinforces the need to apply reliable methods to reduce the incidence of false positive and false negative results which may be of clinical significance. The WGS-SNP analysis demonstrated that indistinguishable Cronobacter strains within a sequence type can be unrelated, and may originate from multiple sources. The use of WGS-SNP analysis to distinguishing of strains within a sequence type has important relevance for tracing the source of outbreaks due to Cronobacter spp.