162. Use of Suppressive Subtractive Hybridization to Identify Genomic Differences among Enteropathogenic Strains of Yersinia enterocolitica and Yersinia pseudotuberculosis 

Lyndsay Radnedge, Peter Agron, Lisa Glover, and Gary Andersen 
Biology and Biotechnology Research Program, Lawrence Livermore National Laboratory, Livermore, California 
andersen2@llnl.gov 

A comparison of genomic sequences among closely related species is likely to reveal unique DNA regions that define the genetic basis for the underlying differences in their phenotypic variation. An example of two closely related human pathogens that differ in their ability to colonize animal hosts as well as their persistence in the environment are the enteropathogenic bacteria, Yersinia enterocolitica and Y. pseudotuberculosis. Of the two pathogens, Y. enterocolitica is more often associated with human infection, especially in day-care centers where the disease is transmitted through infected food, water or soil. Although less frequently diagnosed, infection with Y. pseudotuberculosis is most commonly transmitted through contact with infected birds or mammals. A large percentage of Y. pseudotuberculosis infections are subclinical with no observable symptoms in the exposed individuals. Unlike Y. enterocolitica, Y. pseudotuberculosis may enter the bloodstream of predisposed individuals, causing a lethal septicemia. 

We used a PCR-based subtractive hybridization method termed suppressive subtractive hybridization (SSH) (Diatchenko et al., 1996. Proc. Natl. Acad. Sci. 93:6025-6030) to define the differences between the genomes of Y. enterocolitica and Y. pseudotuberculosis. This technique uses PCR amplification to enrich for unique segments of restricted DNA and simultaneously limits non-target amplification by suppression PCR. The bacterial genome of interest in this comparison is called the tester DNA and the comparison genome is called driver DNA. Using pair-wise comparisons among four strains of Y. enterocolitica and four strains of Y. pseudotuberculosis our initial aim was to identify tester-specific sequences in the type-strains of both species. Control subtractions yielded no PCR products, indicating that the protocol effectively subtracts identical DNA sequences. We have optimized the reaction conditions for the subtraction experiments. Subtracted DNAs were successfully cloned into pGEMT-Easy plasmid vector (Promega); almost 100% of resulting white colonies contained an insert. The clones so far characterized contain inserts that range in size from 200 bp to 1700 bp. The band size distribution of the cloned products represents the distribution of the amplified subtractive hybridization products. Plasmids containing an insert were sequenced on an ABI 377 using dye terminator chemistry. 

BLAST searches of tester-specific DNA sequences reveal homologies to known bacterial genes (including genes involved in pathogenicity) and eleven novel DNA sequences. Included in the regions unique to the Y. enterocolitica type-strain is a difference product with homology to the response-regulator phoP, which has been associated with virulence in Salmonella typhimurium. 92% of oligonucleotide probes designed using these tester-specific DNA sequences distinguished genomic DNA isolated from Y. enterocolitica, while 100% of such probes were specific for Y. pseudotuberculosis. Furthermore, SSH has proved to be sensitive enough to design probes against tester-specific DNA sequence that have been shown to discriminate between genomic DNA isolated from two strains of Y. enterocolitica (58% of the probes successfully discriminate tester DNA). 

Streamlining, automating the steps, and increasing the throughput of this technique should enable large-scale genomic comparison among closely related strains and the generation of strain-specific oligonucleotide probes for molecular epidemiology studies.