Evidence for a selective sweep in the wapl region of Drosophila melanogaster

Abstract

A scan of the X chromosome of a European Drosophila melanogaster population revealed evidence for the recent action of positive directional selection at individual loci. In this study we analyze one such region that showed no polymorphism in the genome scan (located in cytological division 2C10-2E1). We detect a 60.5-kb stretch of DNA encompassing the genes ph-d, ph-p, CG3835, bcn92, Pgd, wapl, and Cyp4d1, which almost completely lacks variation in the European sample. Loci flanking this region show a skewed frequency spectrum at segregating sites, strong haplotype structure, and high levels of linkage disequilibrium. Neutrality tests reveal that these data are unlikely under both the neutral equilibrium model and the simple bottleneck scenarios. In contrast, newly developed maximum-likelihood ratio tests suggest that strong selection has acted recently on the region under investigation, causing a selective sweep. Evidence that this sweep may have originated in an ancestral population in Africa is presented.

Figure 1.

Intraspecific variation around the wapl fragment. (a) The European sample. (b) The African sample. Solid lines and triangles correspond to θ, dashed lines and squares indicate π. Diamonds indicate Tajima's D, and the shaded dashed line represents chromosome-wide average heterozygosity as reported by Glinka et al. (2003). Arrows indicate estimated positions of the target of selection following tests of Kim and Stephan (2002) and Kim and Nielsen (2004); Asterisk indicates the position of the wapl fragment. Absolute genomic positions of fragments are given in megabases, according to release 3 of the annotated D. melanogaster genome. For fragment 3 of the African lines, sequences could not be obtained.

Figure 1.

Intraspecific variation around the wapl fragment. (a) The European sample. (b) The African sample. Solid lines and triangles correspond to θ, dashed lines and squares indicate π. Diamonds indicate Tajima's D, and the shaded dashed line represents chromosome-wide average heterozygosity as reported by Glinka et al. (2003). Arrows indicate estimated positions of the target of selection following tests of Kim and Stephan (2002) and Kim and Nielsen (2004); Asterisk indicates the position of the wapl fragment. Absolute genomic positions of fragments are given in megabases, according to release 3 of the annotated D. melanogaster genome. For fragment 3 of the African lines, sequences could not be obtained.

Figure 2.

Polymorphism data for the 2C10–2E1 region. (a) The European sample. (b) The African sample. D. melanogaster lines 01 (European sample) and 82 (African sample) are taken as references. For the African sample (b), we present polymorphism data only for fragments 4–8 due to space limitations. sim, D. simulans; yak, D. yakuba; (−) no sequence data available.

Figure 2.

Polymorphism data for the 2C10–2E1 region. (a) The European sample. (b) The African sample. D. melanogaster lines 01 (European sample) and 82 (African sample) are taken as references. For the African sample (b), we present polymorphism data only for fragments 4–8 due to space limitations. sim, D. simulans; yak, D. yakuba; (−) no sequence data available.

Figure 3.

Haplotype diversity and linkage disequilibrium in the European sample. Solid lines and squares denote Hd. Dashed lines and triangles indicate levels of linkage disequilibrium (Z_nS). Z_nS was estimated using parsimony informative sites only. Therefore, no value for fragment 2 was obtained. Asterisk indicates the position of the wapl fragment.