Gene flow with wild relatives in cultivated pearl millet

Poster opresented at the Jacques-Monod Conference on "Evolutionary Genomics and Systems Biology: Bringing together Theoretical and Experimental Approaches" in Roscoff (2016)

Gene flow can greatly impact the evolutionary trajectories of population and species. For instance, it can be a source of novel adaptations that allows organisms to face environmental changes or to colonise novel environments. In crop-wild study systems, spontaneous and intentional hybridization and introgression between wild populations and locally adapted landraces is an important source of variation for crop improvement in formal breeding and in traditionally managed agroecosystems. A striking example has been recent uncovered in maize, whose cultivated forms are able to face lower temperatures and precipitation regime in Mexican highlands thanks to the introgression of alleles from the locally adapted wild maize ecotypes. Pearl millet (Pennisetum glaucum) is one of the main cultivated cereals in arid and low fertility soil areas of Africa and India. Cultivated and wild pearl millet distributions overlap in the Sahelian region, where intermediate morphologies and genotypes have been documented. Given that wild populations occupy drier environments than cultivated forms, wild alleles' introgression into the cultivated gene pool may lead to an increase in adaptability to harsher conditions.
To assess the frequency of wild-cultivated gene flow and to identify potentially adaptive introgressions in cultivated pearl millet, we used full-genome data for 31 wild and 190 cultivated accessions. A genome wide assignation analysis allowed us to estimate up to 9% and 16% of wild-crop gene flow in the Eastern and Western part of Sahel respectively. Three hundred genomic regions showed a site frequency spectrum departing from neutral expectations suggesting the action of selection in cultivated populations of pearl millet. The most frequent signal is found in Western Sahel. Functional annotation of these regions revealed that they include several genes involved in the response to stress (i.e. disease and drought), in plant development and reproduction (e.g. auxine responsive proteins and seed development) and in photosynthesis (e.g. phytocrome related proteins), supporting their implications in adaptation. According to different tests, a large fraction of the potentially selected regions seem to be of introgressed origin, which provides a list of candidate adaptive introgressions for further validation.