Complex signatures of selection for the melanogenic loci TYR, TYRP1 and DCT in humans

Abstract

Background: The observed correlation between ultraviolet light incidence and skin color, together with the geographical apportionment of skin reflectance among human populations, suggests an adaptive value for the pigmentation of the human skin. We have used Affymetrix U133a v2.0 gene expression microarrays to investigate the expression profiles of a total of 9 melanocyte cell lines (5 from lightly pigmented donors and 4 from darkly pigmented donors) plus their respective unirradiated controls. In order to reveal signatures of selection in loci with a bearing on skin pigmentation in humans, we have resequenced between 4 to 5 kb of the proximal regulatory regions of three of the most differently expressed genes, in the expectation that variation at regulatory regions might account for intraespecific morphological diversity, as suggested elsewhere.

Results: Contrary to our expectations, expression profiles did not cluster the cells into unirradiated versus irradiated melanocytes, or into lightly pigmented versus darkly pigmented melanocytes. Instead, expression profiles correlated with the presence of Bovine Pituitary Extract (known to contain alpha-MSH) in the media. This allowed us to differentiate between melanocytes that are synthesizing melanin and those that are not. TYR, TYRP1 and DCT were among the five most differently expressed genes between these two groups. Population genetic analyses of sequence haplotypes of the proximal regulatory flanking-regions included Tajima's D, HEW and DHEW neutrality tests analysis. These were complemented with EHH tests (among others) in which the significance was obtained by a novel approach using extensive simulations under the coalescent model with recombination. We observe strong evidence for positive selection for TYRP1 alleles in Africans and for DCT and TYRP1 in Asians. However, the overall picture reflects a complex pattern of selection, which might include overdominance for DCT in Europeans.

Conclusion: Diversity patterns clearly evidence adaptive selection in pigmentation genes in Africans and Asians. In Europeans, the evidence is more complex, and both directional and balancing selection may be involved in light skin. As a result, different non-African populations may have acquired light skin by alternative ways, and so light skin, and perhaps dark skin too, may be the result of convergent evolution.

Figure 1

Median-Joining network describing the genealogical relationships among the haplotypes found in the resequenced 5'-flanking regions of TYR, TYRP1 and DCT. Areas are proportional to absolute frequency. Blue circles: Pygmies; white circles: Europeans; red circles: European melanoma patients; gray circles: Chinese; green circles: Senegalese; and black circles: Australian Aborigines. The ancestral allele for DCT and TYR is indicated in red. The ancestral allele for TYRP1 could not be unequivocally identified, but most likely falls two mutational steps away from AU1A.

Figure 2

Extended Haplotype Homozygosity (EHH) tests. After scanning the HapMap data with Sweep (see Methods) candidate core haplotypes were tested for the significance of their extended haplotype homozygosity (EHH). Circles represent the observed EHH values extending about 50 kb from the core haplotype in both directions, white for 3' and gray for 5'. Dots represent the corresponding 95% upper percentiles of the simulated EHH distributions. The approximate location of the lociwith respect to the core haplotypes is indicated on top: coding exons in black, UTRs in white, and the intervening lines represent the introns. The arrows indicate the direction of transcription. a) EHH for DCT in the HapMap Chinese+Japanese population; b) EHH for TYRP1 in the HapMap Chinese+Japanese population and c) EHH for TYRP1 in the HapMap Yoruba population.