Evaluation of the use of CRISPR loci for discrimination of Salmonella enterica subsp. enterica serovar Enteritidis strains recovered in Canada and comparison with other subtyping methods

Abstract

Salmonella enterica subsp. enterica serovar Enteritidis remains one of the most important foodborne pathogens worldwide. To minimise its public health impact when outbreaks of the disease occur, timely investigation to identify and recall the contaminated food source is necessary. Central to this approach is the need for rapid and accurate identification of the bacterial subtype epidemiologically linked to the outbreak. While traditional methods of S. Enteritidis subtyping, such as pulsed field gel electrophoresis (PFGE) and phage typing (PT), have played an important role, the clonal nature of this organism has spurred efforts to improve subtyping resolution and timeliness through molecular based approaches. This study uses a cohort of 92 samples, recovered from a variety of sources, to compare these two traditional methods for S. Enteritidis subtyping with recently developed molecular techniques. These latter methods include the characterisation of two clustered regularly interspaced short palindromic repeats (CRISPR) loci, either in isolation or together with sequence analysis of virulence genes such as fimH. For comparison, another molecular technique developed in this laboratory involved the scoring of 60 informative single nucleotide polymorphisms (SNPs) distributed throughout the genome. Based on both the number of subtypes identified and Simpson's index of diversity, the CRISPR method was the least discriminatory and not significantly improved with the inclusion of fimH gene sequencing. While PT analysis identified the most subtypes, the SNP-PCR process generated the greatest index of diversity value. Combining methods consistently improved the number of subtypes identified, with the SNP/CRISPR typing scheme generating a level of diversity comparable with that of PT/PFGE. While these molecular methods, when combined, may have significant utility in real-world situations, this study suggests that CRISPR analysis alone lacks the discriminatory capability required to support investigations of foodborne disease outbreaks.

Keywords: CRISPR; PFGE; S. Enteritidis; SNP-PCR; Salmonella enterica; phage; serovar; subtyping.

Figure 1.. Graphical representation of the spacer patterns identified for the two CRISPR loci in this study and their corresponding two-letter binary codes. Black rectangles represent the direct repeat sequences, while color coded ovals represent the spacer sequences; the latter are also identified numerically according to their 5′ to 3′ orders. Two significant variations in the spacer sequences are identified by rectangles in place of ovals in the CRISPR 2 graphic; these are identified as 3′ and 4′, in place of 3 and 4, respectively. Spacers 9, 10 and 11 of CRISPR 2 were identical in sequence. The two-letter binary code is indicated to the right of each combined CRISPR profile.

Figure 2.. Phylogeny of the 92 S. Enteritidis isolates based on their SNP-PCR profiles. For each sample, a concatenated set of specific nucleotides at 60 genomic positions was compiled. The combined fasta file for all samples was used to generate a phylogenetic tree using the UPGMA method, with 1000 bootstrap replicates in MEGA X. The cladal subtype is indicated for each isolate after its sample designation and bootstrap values are indicated at all nodes. Separations of clades 1 and 3 and clades 15 and 16 were not achieved in this illustration due to the software's pairwise deletion of positions that include gaps; such gaps are the sole means of differentiating these subtypes.

Figure 3.. Phylogenetic tree of the 16 CRISPR subtypes identified in this cohort of 92 samples as generated using the UPGMA method. CRISPR and SNP-PCR subtypes as well as sample designations are shown.