Bridget Barker PhD Candidate Genetics-GIDP 24th Annual Fungal Genetics Meeting Asilomar Conference Center, Pacific Grove, CA March 20-25, 2007

ABSTRACT
The genus Coccidioides was recently divided into two putatively allopatric species using molecular markers: Coccidioides immitis, proposed to be restricted to California, and Coccidioides posadasii in the remainder of the fungus’s range: Arizona, Texas, Mexico, Central and South America (Fisher et al. 2002). The sharpness of the geographical separation is ambiguous, as some isolates obtained from patients in California were infected with C. posadasii and some patients from Texas, Arizona and Mexico were infected with C. immitis, the reverse of the usual observed pattern (Fisher et al. 2002). This effect is also observed at the population level, as both species are divided into two populations: northern and southern California for C. immitis, and Arizona/Mexico and Texas/South America for C. posadasii. Again, patients may be infected with a strain of an unexpected origin. These atypical infections may be due to patients traveling to and being exposed in other regions than where they were diagnosed, or due to long distance wind dispersal of infective arthrospores that occur in soil. Alternatively, the ranges of C. immitis and C. posadasii and their constituent populations may overlap.

To test these alternative hypotheses, we determined the genotypes of 65 isolates of Coccidioides obtained from soil samples in the Tucson basin. We also determined the genotypes of seven Coccidioides isolates from domesticated and wild animals from the Tucson basin, which may be less likely to be infected with an atypical isolate, by factoring out frequent travel, but not long distance spore dispersal. These new genotypes were added to the database of all previously characterized isolates, and submitted to phylogenetic analysis to place all isolates in clades that represent previously determined species and populations.

Without exception, soil isolates and animal isolates from Tucson are all grouped within the putative southern Arizona population of C. posadasii. Therefore, soil and animal isolates delineate population structure more accurately than human isolates, which are frequently found to be members of the other populations and species of Coccidioides. Tucson soil and animal isolates are scattered throughout the portion of the tree that represents the Arizona population of C. posadasii. This distribution suggests that gene flow in Arizona is sufficient to maintain diversity within the Tucson subpopulation, and thus is reflective of the Arizona population. The distribution also suggests that there is no population substructuring among isolates that infect humans, domesticated and wild animals, and those found in the soil. However, we note that the fact that Arizona isolates represent one population and this analysis does not preclude the existence of genotypes that vary in their virulence against human and animal host.