Research Article: Genome-Wide Association Studies of Quantitatively Measured Skin, Hair, and Eye Pigmentation in Four European Populations

Date Published: October 31, 2012

Publisher: Public Library of Science

Author(s): Sophie I. Candille, Devin M. Absher, Sandra Beleza, Marc Bauchet, Brian McEvoy, Nanibaa’ A. Garrison, Jun Z. Li, Richard M. Myers, Gregory S. Barsh, Hua Tang, Mark D. Shriver, Nicholas John Timpson. http://doi.org/10.1371/journal.pone.0048294

Abstract

Pigmentation of the skin, hair, and eyes varies both within and between human populations. Identifying the genes and alleles underlying this variation has been the goal of many candidate gene and several genome-wide association studies (GWAS). Most GWAS for pigmentary traits to date have been based on subjective phenotypes using categorical scales. But skin, hair, and eye pigmentation vary continuously. Here, we seek to characterize quantitative variation in these traits objectively and accurately and to determine their genetic basis. Objective and quantitative measures of skin, hair, and eye color were made using reflectance or digital spectroscopy in Europeans from Ireland, Poland, Italy, and Portugal. A GWAS was conducted for the three quantitative pigmentation phenotypes in 176 women across 313,763 SNP loci, and replication of the most significant associations was attempted in a sample of 294 European men and women from the same countries. We find that the pigmentation phenotypes are highly stratified along axes of European genetic differentiation. The country of sampling explains approximately 35% of the variation in skin pigmentation, 31% of the variation in hair pigmentation, and 40% of the variation in eye pigmentation. All three quantitative phenotypes are correlated with each other. In our two-stage association study, we reproduce the association of rs1667394 at the OCA2/HERC2 locus with eye color but we do not identify new genetic determinants of skin and hair pigmentation supporting the lack of major genes affecting skin and hair color variation within Europe and suggesting that not only careful phenotyping but also larger cohorts are required to understand the genetic architecture of these complex quantitative traits. Interestingly, we also see that in each of these four populations, men are more lightly pigmented in the unexposed skin of the inner arm than women, a fact that is underappreciated and may vary across the world.

Partial Text

Human pigmentation of the skin, hair, and eyes varies worldwide. Skin pigmentation forms a gradient correlated with latitude, and variation in hair and eye color is particularly extensive in Europe (reviewed in [1]). For skin color, global-level variation is likely driven primarily by natural (ecological) selection in response to UV radiation levels. The strong linear relationship between global skin pigmentation variation and latitude/UV radiation levels is thought to result from UV-mediated vitamin D synthesis, folate photolysis, sunburn, and skin cancer [2]–[5]. Unlike for global skin pigmentation, there is no obvious ecological selective pressure to explain the variation of hair and eye color or local variation in skin pigmentation. Given the conspicuousness of hair, eye, and skin color, a reasonable hypothesis is that variation in these traits has been shaped by sexual selection [6], 7. Because the geographical differentiation of these traits correlates with the demographic history of human populations, a strong and intrinsic confounding with population structure presents a significant challenge for genetic association studies of human pigmentary traits.

In this study, we measured skin, hair, and eye pigmentation in four European countries and conducted a GWAS for each pigmentary trait. We used the quantitative measurements to analyze correlations between these pigmentary phenotypes and to test for stratification of the phenotypes along axes of European genetic variation. Objective and quantitative phenotyping has the potential to improve the power to detect a genetic effect compared to GWAS based on subjective categorical phenotypes, but due to small sample size our study was only sufficiently powered to identify the major genetic effect of the HERC2/OCA2 locus on eye color. These results are consistent with the currently known genetic architecture of hair and eye color and allow us to make a prediction regarding the genetic architecture of skin pigmentation variation in Europe.

Source:

http://doi.org/10.1371/journal.pone.0048294