Sequence differentiation in regions identified by a genome scan for local adaptation

TLDR

Genome scans have identified a small number of loci that deviate from neutral expectations, marking genomic regions potentially involved in local adaptation, but measuring sequence differentiation and gene identity in these regions is challenging, especially in species with limited genetic resources. We followed up a genome scan in Littorina saxatilis by screening a bacterial artificial chromosome library with differentiated and undifferentiated markers, sequencing four BACs, and analyzing sequence variation in population samples for fragments at or near the original marker polymorphisms. Sequence differentiation matched the original marker frequencies, revealing independent, highly localized sites outside coding regions, including two loci with recent transposable element insertions that may affect downstream gene expression, thereby providing strong candidate loci for local adaptation and demonstrating that genome scans can rapidly identify candidate genes in nonmodel organisms.

Abstract

Genome scans using large numbers of randomly selected markers have revealed a small proportion of loci that deviate from neutral expectations and so may mark genomic regions that contribute to local adaptation. Measurements of sequence differentiation and identification of genes in these regions is important but difficult, especially in organisms with limited genetic information available. We have followed up a genome scan in the marine gastropod, Littorina saxatilis, by searching a bacterial artificial chromosome library with differentiated and undifferentiated markers, sequencing four bacterial artificial chromosomes and then analysing sequence variation in population samples for fragments at, and close to the original marker polymorphisms. We show that sequence differentiation follows the patterns expected from the original marker frequencies, that differentiated markers identify independent and highly localized sites and that these sites fall outside coding regions. Two differentiated loci are characterized by insertions of putative transposable elements that appear to have increased in frequency recently and which might influence expression of downstream genes. These results provide strong candidate loci for the study of local adaptation in Littorina. They demonstrate an approach that can be applied to follow up genome scans in other taxa and they show that the genome scan approach can lead rapidly to candidate genes in nonmodel organisms.