Most studies of Salmonella use strains which have been isolated from infected endotherms. This sampling strategy could bias our conclusions regarding the population biology of Salmonella. In this project, I used a set of presumptive Salmonella isolated from turtles to ask three questions: are they genetically Salmonella, what do they look like phenotypically, and do they possess SPIs (Salmonella Pathogenicity Islands)? These islands are groups of genes which aid in infection. Many of the strains studied, which would have been diagnosed phenotypically in clinics as Salmonella did not appear to be Salmonella genetically as determined by using species-specific primers for the mdh and hisJ genes. They also had non-standard phenotypes when tested for fermentation and use of certain carbon sources. SPI1, which is needed for infection, was not present as determined by the absense of spvC and invA genes. This suggests that what traditionally has been diagnosed in turtles as Salmonella may not be, or they may represent a new lineage of Salmonella.
This work was funded NIH grant GM060731.