Evidence for a selective sweep on chromosome 1 of cultivated sorghum

Recently, a simple sequence repeat (SSR)-based genome-wide diversity scan of Sorghum bicolor (L.) Moench identified several candidate loci with patterns of variation consistent with directional selection in cultivated lines. Data were insufficient, however, to determine if selection had actually occurred at or near candidate SSR loci or if the unusual diversity patterns observed were due to the effects of demographic factors such as population bottlenecks or mating system. In the present study, we collected DNA sequences from 10 segments within a 99 kb region flanking one of the previously identified candidates, SSR locus Xcup 15, located near the distal end of chromosome 1. We performed statistical tests both to address alternative hypotheses to selection and to aid in localizing the selection target. Analyses of genomic DNA sequences from a panel of 17 cultivated and 13 wild accessions indicated that cultivated lines had reduced diversity in this region (about one-third of the diversity present in wild sorghums) and a moderate degree of differentiation was observed between cultivated and wild groups ([F.sub.st]- 0.15). Several features of the data support the hypothesis that recent directional selection shaped diversity patterns around Xcup 15, including overall low levels of variation and extensive haplotype structure (a predominant haplotype occurred over the 99 kb region) in cultivated sorghum, and a derived fixed difference at the 5' untranslated region (UTR) of a protein phosphatase 2C (PP2C) gene between cultivated and wild sorghums. Moreover, two of the four tests employed to detect deviations from the neutral, equilibrium model, the Hudson Kreitman Aguade (HKA), and the composite likelihood ratio (CLR) tests indicated that patterns of diversity in the Xcup15 region were consistent with a selective sweep. Although we were unable to rule out demography as a possible explanation for the diversity patterns observed along this region, this study supported previous findings based on SSR diversity and identified candidates far the target of selection; the confirmation of which will require functional and association studies.