| Summary: | <p>Abstract</p> <p>Background</p> <p><it>Salmonella </it>Enteritidis is currently the world's leading cause of salmonellosis, in part because of its ability to contaminate the internal contents of eggs. Previous analyses have shown that it is an exceptionally clonal serotype, which nonetheless generates considerable phenotypic heterogeneity. Due to its clonality, whole genome analysis is required to find genetic determinants that contribute to strain heterogeneity of <it>Salmonella </it>Enteritidis. Comparative whole genome mutational mapping of two PT13a strains that varied in the ability to contaminate eggs and to form biofilm was achieved using a high-density tiling platform with primers designed from a PT4 reference genome. Confirmatory Sanger sequencing was used on each putative SNP identified by mutational mapping to confirm its presence and location as compared to the reference sequence. High coverage pyrosequencing was used as a supporting technology to review results.</p> <p>Results</p> <p>A total of 250 confirmed SNPs were detected that differentiated the PT13a strains. From these 250 SNPS, 247 were in the chromosome and 3 were in the large virulence plasmid. SNPs ranged from single base pair substitutions to a deletion of 215 bp. A total of 15 SNPs (3 in egg-contaminating PT13a 21046 and 12 in biofilm forming PT13a 21027) altered coding sequences of 16 genes. Pyrosequencing of the two PT13a subpopulations detected 8.9% fewer SNPs than were detected by high-density tiling. Deletions and ribosomal gene differences were classes of SNPs not efficiently detected by pyrosequencing.</p> <p>Conclusions</p> <p>These results increase knowledge of evolutionary trends within <it>Salmonella enterica </it>that impact the safety of the food supply. Results may also facilitate designing 2<sup>nd </sup>generation vaccines, because gene targets were identified that differentiate subpopulations with variant phenotypes. High-throughput genome sequencing platforms should be assessed for the ability to detect classes of SNPs equivalently, because each platform has different advantages and limits of detection.</p>
|
|---|
