- Record: found
- Abstract: found
- Article: found
Association of the OCA2 Polymorphism His615Arg with Melanin Content in East Asian Populations: Further Evidence of Convergent Evolution of Skin Pigmentation
research-article
Author(s):
Melissa Edwards
1 ,
Abigail Bigham
2 ,
Jinze Tan
3 ,
Shilin Li
3 ,
Agnes Gozdzik
1 ,
Kendra Ross
1 ,
Li Jin
3 ,
Esteban J. Parra
1
,
*
Publication date (Electronic):
5 March 2010
Journal:
PLoS Genetics
Publisher:
Public Library of Science
Read this article at
There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.
Abstract
The last decade has witnessed important advances in our understanding of the genetics of pigmentation in European populations, but very little is known about the genes involved in skin pigmentation variation in East Asian populations. Here, we present the results of a study evaluating the association of 10 Single Nucleotide Polymorphisms (SNPs) located within 5 pigmentation candidate genes ( OCA2, DCT, ADAM17, ADAMTS20, and TYRP1) with skin pigmentation measured quantitatively in a sample of individuals of East Asian ancestry living in Canada. We show that the non-synonymous polymorphism rs1800414 (His615Arg) located within the OCA2 gene is significantly associated with skin pigmentation in this sample. We replicated this result in an independent sample of Chinese individuals of Han ancestry. This polymorphism is characterized by a derived allele that is present at a high frequency in East Asian populations, but is absent in other population groups. In both samples, individuals with the derived G allele, which codes for the amino acid arginine, show lower melanin levels than those with the ancestral A allele, which codes for the amino acid histidine. An analysis of this non-synonymous polymorphism using several programs to predict potential functional effects provides additional support for the role of this SNP in skin pigmentation variation in East Asian populations. Our results are consistent with previous research indicating that evolution to lightly-pigmented skin occurred, at least in part, independently in Europe and East Asia.
Author Summary
Our knowledge of the genetic basis of normal pigmentation variation in human populations is quite incomplete. Recent studies have identified some of the genes responsible for the reduction in melanin content in European populations, but this is not the case for other population groups, such as East Asians. Here, we report that a genetic variant located within the gene OCA2 (rs1800414) is associated with skin pigmentation in two samples of East Asian ancestry. The allele associated with lower melanin levels is found at high frequencies in East Asian populations, but is absent or at very low frequencies in other population groups. This is one of the first reports of association of genetic markers with quantitative measures of pigmentation in East Asian populations and it confirms previous evidence indicating that evolution towards light skin occurred, at least in part, independently in Europe and East Asia. The OCA2 gene has been under positive selection in Europe and East Asia, but different alleles have been selected in each region.
Related collections
Most cited references30
- Record: found
- Abstract: found
- Article: not found
The evolution of human skin coloration.
N Jablonski, G Chaplin (2000)
Skin color is one of the most conspicuous ways in which humans vary and has been widely used to define human races. Here we present new evidence indicating that variations in skin color are adaptive, and are related to the regulation of ultraviolet (UV) radiation penetration in the integument and its direct and indirect effects on fitness. Using remotely sensed data on UV radiation levels, hypotheses concerning the distribution of the skin colors of indigenous peoples relative to UV levels were tested quantitatively in this study for the first time. The major results of this study are: (1) skin reflectance is strongly correlated with absolute latitude and UV radiation levels. The highest correlation between skin reflectance and UV levels was observed at 545 nm, near the absorption maximum for oxyhemoglobin, suggesting that the main role of melanin pigmentation in humans is regulation of the effects of UV radiation on the contents of cutaneous blood vessels located in the dermis. (2) Predicted skin reflectances deviated little from observed values. (3) In all populations for which skin reflectance data were available for males and females, females were found to be lighter skinned than males. (4) The clinal gradation of skin coloration observed among indigenous peoples is correlated with UV radiation levels and represents a compromise solution to the conflicting physiological requirements of photoprotection and vitamin D synthesis. The earliest members of the hominid lineage probably had a mostly unpigmented or lightly pigmented integument covered with dark black hair, similar to that of the modern chimpanzee. The evolution of a naked, darkly pigmented integument occurred early in the evolution of the genus Homo. A dark epidermis protected sweat glands from UV-induced injury, thus insuring the integrity of somatic thermoregulation. Of greater significance to individual reproductive success was that highly melanized skin protected against UV-induced photolysis of folate (Branda & Eaton, 1978, Science201, 625-626; Jablonski, 1992, Proc. Australas. Soc. Hum. Biol.5, 455-462, 1999, Med. Hypotheses52, 581-582), a metabolite essential for normal development of the embryonic neural tube (Bower & Stanley, 1989, The Medical Journal of Australia150, 613-619; Medical Research Council Vitamin Research Group, 1991, The Lancet338, 31-37) and spermatogenesis (Cosentino et al., 1990, Proc. Natn. Acad. Sci. U.S.A.87, 1431-1435; Mathur et al., 1977, Fertility Sterility28, 1356-1360).As hominids migrated outside of the tropics, varying degrees of depigmentation evolved in order to permit UVB-induced synthesis of previtamin D(3). The lighter color of female skin may be required to permit synthesis of the relatively higher amounts of vitamin D(3)necessary during pregnancy and lactation. Skin coloration in humans is adaptive and labile. Skin pigmentation levels have changed more than once in human evolution. Because of this, skin coloration is of no value in determining phylogenetic relationships among modern human groups. Copyright 2000 Academic Press.
- Record: found
- Abstract: found
- Article: not found
A Genome-Wide Association Study Identifies Novel Alleles Associated with Hair Color and Skin Pigmentation
Jiali Han, Peter Kraft, Hongmei Nan … (2008)
Introduction There is substantial variation in human pigmentation within and across populations. Ultraviolet radiation (UV) exposure is the most important environmental factor influencing evolutionary selection pressure on pigmentation. In addition to UV-induced DNA damage, UVA can break down folic acid, and the major source of circulating vitamin D is synthesized in UVB-exposed skin. Because both nutrients are essential for human reproduction, it has been proposed that human pigmentation is selected, at least in part, to optimize levels of these two UV-related nutrients [1]. UV light is also the major environmental risk factor for skin cancer in humans. Red and blonde hair color, light skin pigmentation, and blue eye color are major host susceptibility factors for skin cancer [2]. Human pigmentation is a polygenic quantitative trait with high heritability [3]–[5]. A handful of genes underlying rare, extreme pigmentation phenotypes have been discovered [6], although until recently, only six genes were known to contain common genetic variants associated with human pigmentation in the normal range (MC1R, TYR, OCA2, SLC24A5, MATP, and ASIP). The proteins that these genes encode contribute to the control of melanin production and the maturation of melanosomes in melanogenesis, which determines human pigmentation. With new technologies that enable genotyping of hundreds of thousands of single nucleotide polymorphisms (SNPs), together with new insights into the structure of variation in the human genome [7], it is now possible to scan the genome in an agnostic manner in search of common genetic variants associated with human pigmentation. To identify common genetic variants associated with variation in natural diversity of human pigmentation, we performed a genome-wide association study (GWAS) of natural hair color in 2,287 U.S. women of European ancestry using data on 528,173 SNPs genotyped as part of the Cancer Genetic Markers of Susceptibility breast cancer GWAS [8]. Promising SNPs were examined in four additional studies with data on hair color and other pigmentation phenotypes: 870 U.S. women controls free of diagnosed skin cancer from a skin-cancer case-control study; 3,750 U.S. women from a diabetes case-control study; 2,405 U.S. men from a diabetes case-control study; and 1,440 parents of twins from an ongoing Australian family-based study of genetic and environmental factors contributing to the development of pigmented nevi [9]. Results The frequencies of pigmentary phenotypes collected in the 5 component studies are presented in Table 1. The samples were broadly similar. 10.1371/journal.pgen.1000074.t001 Table 1 The distributions of human pigmentary phenotypes in the five component studies Initial GWAS women skin cancer study (controls) women NHS diabetes women HPFS diabetes men Australian study both genders hair color n (%) n (%) n (%) n (%) n (%) black 57 (2.6) 27 (3.3) 160 (4.3) 256 (10.7) 134 (9.3) dark brown 911 (41.7) 358 (43.0) 1581 (42.5) 1046 (43.7) 474 (32.9) light brown 849 (38.8) 333 (40.0) 1379 (37.1) 755 (31.5) 520 (36.1) blonde 277 (12.7) 89 (10.7) 451 (12.1) 274 (11.4) 238 (16.5) red 92 (4.2) 25 (3.0) 151 (4.1) 64 (2.7) 73 (5.1) skin color fair - 348 (49.0) - - 812 (56.4) medium - 320 (45.1) - - 493 (34.2) olive or black - 42 (5.9) - - 135 (9.4) tanning ability Practically none 188 (8.7) 68 (8.3) 375 (10.2) - - light tan 454 (21.0) 173 (21.1) 850 (23.1) 621 (24.9) - average tan 1011 (46.7) 387 (47.3) 1664 (45.1) 1089 (43.7) - deep tan 513 (23.7) 191 (23.3) 797 (21.6) 784 (31.4) - eye color Brown/dark - - - 776 (32.3) 328 (22.8) Hazel/green/medium - - - 781 (32.5) 529 (36.8) Blue/light - - - 844 (35.2) 582 (40.4) The percentages may not sum to 100 due to rounding. We compared the distribution of observed p-values from each of the 528,173 SNPs in the GWAS with those expected under the global null hypothesis that none of the tested SNPs is associated with natural hair color (Figure 1). The distribution of the observed p-values for the crude analyses that restricted analysis to women of self-reported European ancestry but did not further adjust for potential population stratification shows evidence for systematic bias: the genomic control inflation factor for the crude analyses (the ratio of the median observed test statistic to the theoretical median) is λGC = 1.24. This systematic bias is most likely due to confounding by latent population stratification. Hair color varies along a light-dark gradient from northern to southern Europe, so it will be associated with any SNP marker whose minor allele frequency also varies along a North-South gradient, even if that marker is not in linkage disequilibrium (LD) with a causal hair-color locus [10]. Adjusting for the top four principal components of genetic variance [11] eliminated most of the apparent residual confounding due to population stratification (λGC = 1.02 for the adjusted analyses); further control for up to 50 principal components did not alter the λGC. All of the association results from the initial GWAS reported below are from analyses that adjusted for the top four principal components of genetic variation. 10.1371/journal.pgen.1000074.g001 Figure 1 Quantile-quantile plot of the -log10 p-values from an analysis of the initial GWAS that did not adjust for principal components of genetic variation (black dots) and an analysis that did adjust for the four largest principal components (red dots). p-values smaller than 10−8 are not plotted. The GWAS identified several genomic locations as potentially associated with hair color (Figure 2). Of 528,173 SNPs tested, the 38 SNPs with the most extreme p-values associated with hair color are listed in Table 2. 10.1371/journal.pgen.1000074.g002 Figure 2 -log10 p-values from the primary test of association with hair color in the initial GWAS, by position along chromosome. Only p-values smaller than 0.05 are plotted. 10.1371/journal.pgen.1000074.t002 Table 2 Thirty-eight SNPs with the smallest p-values of the 528,173 tested for association of hair color in the initial GWAS of in 2,287 women of European ancestry hair color (black to red) non-red hair color (black to blonde) red vs. non-red color SNP Chromosome Location GeneNeighborhood WT/VT MAF beta s.e. p value Selected for replication beta s.e. p value beta s.e. p value rs12913832 15 26039213 HERC2 G/A 0.25 −0.37 0.03 4.3E-39 Y −0.38 0.02 2.7E-56 −0.20 0.18 0.25 rs1667394 15 26203777 HERC2 T/C 0.17 −0.33 0.03 1.5E-22 LD with rs8039195 −0.33 0.03 1.2E-29 −0.30 0.22 0.18 rs12203592 6 341321 IRF4 C/T 0.17 −0.31 0.03 8.2E-20 Y −0.36 0.03 6.6E-36 0.42 0.18 0.02 rs258322 16 88283404 MC1R C/T 0.09 0.36 0.04 8.0E-18 Y 0.07 0.04 0.08 1.89 0.18 5.0E-27 rs4785763 16 88594437 MC1R C/A 0.33 0.23 0.03 1.7E-17 Y 0.07 0.02 4.8E-03 1.75 0.17 5.7E-24 rs6497268 15 26012308 OCA2 C/A 0.18 −0.25 0.03 8.5E-15 Y −0.24 0.03 9.5E-18 −0.35 0.22 0.10 rs8039195 15 26189679 HERC2 C/T 0.15 −0.28 0.04 1.9E-14 Y −0.33 0.03 7.4E-30 −0.24 0.23 0.30 rs11855019 15 26009415 OCA2 A/G 0.14 −0.26 0.04 2.1E-12 Y −0.26 0.03 4.4E-16 −0.33 0.24 0.18 rs11636232 15 26060221 HERC2 C/T 0.40 0.18 0.03 3.2E-12 Y 0.16 0.02 2.1E-12 0.38 0.15 0.01 rs8049897 16 88551703 MC1R G/A 0.15 0.24 0.04 4.8E-12 Y 0.08 0.03 0.01 1.36 0.16 1.4E-16 rs4238833 16 88578190 MC1R T/G 0.37 0.18 0.03 5.6E-12 Y 0.04 0.02 0.09 1.62 0.17 1.7E-20 rs4408545 16 88571529 MC1R T/C 0.50 0.16 0.03 3.0E-10 Y 0.04 0.02 0.06 1.54 0.19 1.9E-15 rs7204478 16 88322986 MC1R C/T 0.44 0.15 0.03 1.0E-08 Y 0.02 0.02 0.31 1.52 0.18 4.7E-17 rs4904866 14 91838256 C/T 0.43 0.14 0.03 3.0E-08 LD with rs12896399 0.17 0.02 1.2E-13 −0.10 0.15 0.53 rs12896399 14 91843416 SLC24A4 T/G 0.43 0.14 0.03 3.0E-08 Y 0.16 0.02 1.7E-13 −0.10 0.15 0.51 rs7174027 15 26002360 OCA2 G/A 0.11 −0.22 0.04 5.0E-08 Y −0.22 0.04 2.6E-10 −0.31 0.27 0.25 rs7183877 15 26039328 HERC2 C/A 0.08 −0.26 0.05 6.0E-08 Y −0.24 0.04 3.5E-09 −0.27 0.32 0.39 rs7196459 16 88668978 MC1R G/T 0.08 0.25 0.05 1.0E-07 Y 0.05 0.04 0.26 1.40 0.19 3.1E-13 rs164741 16 88219799 MC1R C/T 0.30 0.14 0.03 1.2E-07 Y 0.01 0.02 0.62 1.30 0.16 4.6E-17 rs7188458 16 88253985 MC1R G/A 0.43 0.13 0.03 3.1E-07 Y 0.01 0.02 0.74 1.49 0.18 5.8E-17 rs8033165 15 26805134 C/T 0.46 0.12 0.02 4.3E-07 Y 0.10 0.02 1.9E-06 0.36 0.14 0.01 rs35391 5 33991430 MATP C/T 0.03 −0.41 0.08 4.6E-07 LD with rs28777 −0.43 0.07 7.9E-10 −0.17 0.51 0.74 rs7495174 15 26017833 OCA2 A/G 0.08 −0.24 0.05 7.1E-07 Y −0.25 0.04 2.6E-09 −0.24 0.31 0.44 rs1635168 15 26208861 HERC2 C/A 0.08 −0.25 0.05 9.0E-07 LD with rs8028689 −0.27 0.04 6.6E-10 −0.18 0.32 0.57 rs8007923 14 103726412 KIF26A T/C 0.47 −0.12 0.03 1.2E-06 Y −0.10 0.02 6.0E-06 −0.28 0.15 0.07 rs10861741 12 106353647 C/T 0.15 0.17 0.04 1.8E-06 Y 0.12 0.03 1.6E-04 0.55 0.19 3.1E-03 rs28777 5 33994716 MATP A/C 0.03 −0.36 0.08 1.9E-06 Y −0.37 0.06 1.1E-08 −0.31 0.51 0.54 rs9806558 15 23458783 C/T 0.33 −0.13 0.03 3.4E-06 Y −0.09 0.02 1.1E-04 −0.40 0.17 0.02 rs9392056 6 463078 EXOC2 A/G 0.38 0.12 0.03 4.1E-06 LD with rs6918152 0.10 0.02 4.5E-06 0.19 0.15 0.22 rs4778211 15 25872900 OCA2 C/A 0.16 −0.16 0.03 4.1E-06 Y −0.15 0.03 3.5E-07 −0.17 0.22 0.44 rs2493040 6 480839 EXOC2 G/A 0.32 0.13 0.03 4.3E-06 LD with rs6918153 0.11 0.02 3.3E-06 0.16 0.16 0.30 rs6918152 6 487159 EXOC2 G/A 0.37 0.12 0.03 5.1E-06 Y 0.11 0.02 2.5E-06 0.14 0.15 0.34 rs2353033 16 87913062 MC1R T/C 0.43 0.12 0.03 5.2E-06 Y 0.04 0.02 0.09 0.92 0.16 6.1E-09 rs12142165 1 224689937 OBSCN T/C 0.35 −0.12 0.03 6.3E-06 Y −0.09 0.02 1.2E-04 −0.41 0.17 0.02 rs7195066 16 88363824 MC1R C/T 0.31 −0.12 0.03 9.5E-06 Y −0.01 0.02 0.64 −1.83 0.29 3.9E-10 rs2241039 16 88615938 MC1R C/T 0.38 −0.11 0.03 1.1E-05 Y −0.05 0.02 0.04 −0.90 0.19 1.1E-06 rs8028689 15 26162483 HERC2 T/C 0.06 −0.24 0.05 1.1E-05 Y −0.25 0.05 1.5E-07 −0.24 0.35 0.49 rs16950987 15 26199823 HERC2 G/A 0.06 −0.24 0.05 1.2E-05 LD with rs8028689 −0.24 0.05 1.7E-07 −0.24 0.35 0.49 The p-values are based on primary association test (including women with red hair) adjusted for top four principal components of genetic variance. The regression parameter beta refers to the mean change in pigmentation scoring (or change in log odds of red hair for red hair analyses) per copy of the SNP minor allele. We selected 31 of these 38 SNPs for further study in an independent sample. The remaining seven SNPs were in strong LD (r2>0.8) with one of these 31 SNPs (Table 2). The sample consisted of 870 controls of European ancestry from a nested case-control study of skin cancer within the Nurses' Health Study (NHS). Thirty of the 31 attempted SNPs were genotyped successfully. Twenty-two of these 30 SNPs showed very strong evidence for association with natural hair color (p 0.05) after adjustment for the other SNPs, and association between rs6918152 and hair color was much weaker in the GWAS and no longer significant in the Australian samples after adjustment. While the IRF4 SNP rs12203592 was also associated with skin color, eye color and tanning ability, the EXOC2 SNP rs6918152 was not associated with these phenotypes. These results suggest that the IRF4 SNP rs12203592 is most likely to be in strong LD with the causal variant in this region. 10.1371/journal.pgen.1000074.t006 Table 6 Association between SNPs in EXOC2 and IRF4 and hair color (black to blonde) marginal analysis multivariable analysis beta s.e. p value beta s.e. p value Initial GWAS rs12203592 −0.36 0.03 6.6E-36 −0.35 0.03 8.0E-29 rs6918152 0.11 0.02 2.5E-06 0.07 0.02 3.6E-03 rs1540771 0.06 0.02 7.2E-03 −0.03 0.02 0.16 Australian Study rs12203592 −0.38 0.04 4.6E-23 −0.36 0.04 2.7E-17 rs6918152 0.14 0.03 2.0E-05 0.05 0.03 0.12 rs1540771 0.12 0.03 3.3E-04 0.02 0.04 0.57 Pooled rs12203592 −0.35 0.02 3.1E-52 −0.35 0.03 2.7E-44 rs6918152 0.12 0.02 2.7E-09 0.06 0.02 1.4E-03 rs1540771 0.09 0.02 3.4E-06 −0.01 0.02 0.52 The multivariable analysis mutually adjusted for the three SNPs. The rs12896399 SNP 15.5 kb upstream of the SLC24A4 gene was highly associated with light hair color, and relatively weakly associated with less tanning ability in the pooled analysis of four studies (p = 6.0×10−62 for hair color, and p = 0.01 for tanning ability). The percentage of residual variation in hair color from black to blonde explained by this SNP after controlling for the top four principal components of genetic variation was 2.6%. This variant was also associated with blue/light eye color (p = 2.9×10−6 in the HPFS set). The SLC24A4 gene belongs to a family of potassium-dependent sodium/calcium exchangers. At least two other members of this family are associated with skin pigmentation. The SLC24A5 gene was recently shown to be involved in skin pigmentation in both zebrafish and humans [13]. Another member of this family, MATP (SLC45A2), is a pigmentation gene transcriptionally regulated by MITF [14],[15]. We identified the SNP rs28777 in the MATP gene from the GWAS, and the association with hair color was replicated in the controls of the skin cancer study (pooled P value = 8.9×10−14). This SNP was also associated with skin color (pooled P value = 9.5×10−4) and tanning ability (pooled P value = 2.2×10−10). Three SNPs in the MATP gene have been associated with human pigmentation: rs16891982 (Phe374Leu), rs26722 (Glu272Lys), and rs13289 C/G (-1721 in the promoter region) [16],[17]. We genotyped these three SNPs in the controls of the skin cancer study. None of the three previously reported SNPs were in LD with rs28777 (r2≤0.01), which is an intronic SNP. A multivariable analysis mutually adjusting for rs28777, rs16891982, rs26722, and rs13289 simultaneously showed that only rs16891982 remained significant in the model (P = 0.036 for hair color (black to blonde), P = 0.016 for tanning ability, and P = 0.0009 for skin color) and other SNPs became non-significant (p>0.05). These data suggested that rs16891982 is most likely to be the causal variant or in strong LD with the causal variant in the MATP gene. Eleven SNPs spanning 1 Mb on chromosome 15 were strongly associated with hair color in the initial GWAS. These SNPs were located in the OCA2 5′ regulatory region and the HERC2 gene region and included the 3 SNPs reported previously with eye color: rs7495174, p = 7.1×10−7; rs6497268, p = 8.5×10−15; and rs11855019, p = 2.1×10−12) [18]. In an analysis mutually adjusting for all 11 SNPs simultaneously, only the HERC2 SNP rs12913832 (not on the HumanHap 300 version used in Sulem et al. [12]) remained significantly associated with hair color (p = 2.73×10−32) and tanning ability (p = 3.03×10−9). The associations between all other SNPs and hair color became non-significant (p>0.05). This suggested that the SNP rs12913832 was most likely to be in strong linkage disequilibrium with the causal variant in this region. The percentage of residual variation in hair color from black to blonde explained by this SNP after controlling for the top four principal components of genetic variation was 10.7%. We observed 12 SNPs on chromosome 16 associated with hair color in the GWAS, spanning >756 kb. The MC1R gene, well established to be associated with red hair color, is located within this region. We had previously genotyped 7 common MC1R variants among the NHS skin cancer controls [19]. The analysis mutually adjusting for all 19 SNPs in the controls of the skin cancer study indicates that the signals that we detected in this region were mainly due to the three MC1R red hair color alleles (Arg151Cys, Arg160Trp, and Asp294His) (Table S3). The pairwise LD among these 19 SNPs was very low (the pattern of LD across these 19 SNPs is shown in Figure S1). Discussion It has been a longstanding hypothesis that human pigmentation is tightly regulated by genetic variation. However, very few genes have been identified that contain common genetic variants associated with human pigmentation. We conducted a genome-wide association study of hair color and identified several new variants associated with variation in hair color, skin color, eye color, and tanning ability among individuals of European ancestry. Among the loci identified from our GWAS, the IRF4 and SLC24A4 loci had not been linked to human pigmentation before we began our study. Recently, Sulem et al. [12] reported a pigmentation GWAS using 316,515 SNPs in the Icelandic population. The associations of hair color in our GWAS with the 60 SNPs reported by Sulem et al. are listed in Table S4. These authors identified two SNPs (rs4959270 and rs1540771) between the EXOC2 and IRF4 genes in relation to freckles, hair color, and skin sensitivity to sun [12]. We identified a SNP in intron 4 of the IRF4 gene, not genotyped by Sulem et al. [12], with a much stronger association than they observed with hair color and tanning ability. The IRF4 gene product is a member of the interferon regulatory factor family of transcription factors [20]–[23], which are involved in the regulation of gene expression in response to interferon and other cytokines. The IRF4 gene encodes a B-cell proliferation/differentiation protein, which has been proposed as a sensitive and specific marker for conventional primary and metastatic melanomas and benign melanocytic nevi [24]. The SNP rs12896399 upstream of the SLC24A4 gene showed strong association with hair color in our study and that of Sulem et al. [12]. In addition, we identified three chromosomal regions adjacent to the previously known pigmentation genes: MC1R, OCA2, and MATP. The MC1R gene encodes a 317-amino acid 7-pass-transmembrane G protein coupled receptor and has been shown as the rate-limiting step in the activation of the cAMP pathway in terms of melanin production. Although the LD between the MC1R variants and surrounding highly significant SNPs was relatively low, the multivariable models mutually adjusting for all surrounding SNPs suggests that the signals that we identified on chromosome 16 were explained by the functional variants in the MC1R gene. A previous report showed that three SNPs in intron 1 of the OCA2 gene were associated with eye, skin, and hair color [18]. We identified a SNP (rs12913832) in the upstream HERC2 gene with a much stronger association with hair color and tanning ability. Because the HERC2 gene has not been linked to human pigmentation to our knowledge, the SNP may be involved in the regulation of the expression or the function of the OCA2 gene. Similarly, Sulem et al. identified rs1667394 (∼165 kb upstream from rs12913832, r2 = 0.58) as the strongest hit in this region in their study [12]. However, we found a stronger association of rs12913832 with hair color than that of rs1667394 in our study. Similar to our findings, Sulem et al. reported associations in the regions encompassing MC1R, OCA2, and SLC24A4. In addition, they reported loci in two established pigmentation genes, TYR and KITLG, which were not among the 38 SNPs with the strongest associations with hair color (black to red) that we sought to genotype in additional samples. The p-value for the association between KITLG rs12821256 and hair color (black to red) in the initial GWAS was 0.0002; the p-value for the association with hair color excluding women with red hair (black to blonde) was 1.28×10−8. The p-values for association between TYR rs1393350 and hair color coded as black to red or black to blonde were 0.05 and 0.02, respectively. We additionally identified another previously reported pigmentation gene from our GWAS, the MATP gene that was not reported by Sulem et al. [12]. In our analysis of testing the trend across hair color from black to blonde, there were no other loci reaching genome-wide significance level. Two of the four regions we found to be associated with variation in hair color among Europeans without red hair (MATP and HERC2/OCA2) show strong evidence of recent positive selection, based on a comparison of allele frequencies across samples from three continental populations (Africa, Asia, and Europe) [25]. Both of the markers we identified in the two remaining regions showed significant differences in allele frequency across the HapMap CEU, CHB, JPT and YRI panels: the IRF4 SNP rs12203592 was monomorphic in CHB, JPT and YRI panels (the minor allele in Europeans was absent from these samples); the minor allele among Europeans for SLC24A4 SNP rs12896399 (G) was the major allele for the CHB and YRI panels, with the G allele frequency in the YRI sample being above 99%. Moreover, all of the markers with strongest association with hair color in these four regions were significantly associated with one or more of the top four principal components of genetic variation (Table S5), suggesting that allele frequencies for these markers also vary among European populations. Because we adjusted for latent population structure using these four principal components—and there are multiple lines of evidence suggesting these regions influence hair color among Europeans—we believe it unlikely that the strong associations we see between these markers and hair color are solely due to population stratification bias. Rather it is likely that differences in the distribution of hair color across Europe are due in part to differences in allele frequencies at these loci and other as-yet-unknown loci. Taken together, these four regions explain approximately 21.9% of the residual variation in hair color (black-blond) after adjusting for the top four principal components of genetic variation. (Conversely, after adjusting for these four regions, the top four principal components of genetic variation explain 2.6% of the residual variation in hair color.) In our study of men and women of European ancestry we focused on the most statistically significant associations from our GWAS among women, identifying the IRF4 variant as reproducibly associated with human pigmentation. Further work is needed to identify the causal variant at this locus. Because a subset of true associations would be weakly associated with outcome in any given GWAS, large-scale replication is necessary for confirmation, and some true associations may be missed if they are not carried forward into replication studies. In this regard, the precomputed rankings and P values for all the SNPs included in the GWAS conducted in the NHS are freely available (http://www.channing.harvard.edu/nhs/publications/index.shtml) for others to use in subsequent studies. Materials and Methods Nurses' Health Study (NHS) The NHS was established in 1976, when 121,700 female U.S. registered nurses between the ages of 30 and 55, residing in 11 larger U.S. states, completed and returned the initial self-administered questionnaire on their medical histories and baseline health-related exposures, forming the basis for the NHS cohort. Biennial questionnaires with collection of exposure information on risk factors and (every 4 years since 1980) nutritional assessments have been collected prospectively. Along with exposures every 2 years, outcome data with appropriate follow-up of reported disease events, including melanoma and non-melanoma skin cancers, are collected. Overall, follow-up has been very high; after more than 20 years approximately 90% of participants continue to complete questionnaires. From May 1989 through September 1990, we collected blood samples from 32,826 participants in the NHS cohort. Subsequent follow-up has been greater than 99% for this subcohort. The information on natural hair color at age of 20 and childhood and adolescence tanning ability were collected in the 1982 questionnaire. Initial GWAS We initially performed genotyping in a nested case-control study of postmenopausal invasive breast cancer within the Nurses' Health Study (NHS) cohort [26] using the Illumina HumanHap550 array, as part of the National Cancer Institute's Cancer Genetic Markers of Susceptibility (CGEMS) Project [8]. We performed our initial genome-wide analysis on 528,173 SNPs in 2,287 women [8]. All cases and controls were self-described as being of European ancestry. Four samples were excluded because of evidence of intercontinental admixture. Controlling for breast cancer case-control status made no material difference to the GWAS results. Information on natural hair color at age 20 was collected in the NHS main questionnaire and grouped into five categories (black, dark brown, light brown, blonde, and red). Detailed methods related to the initial GWAS were published previously [8], including genotyping and quality control, initial assessment of sample completion rates, assessment of SNP call rates, concordance rate, deviation from Hardy–Weinberg proportions in control DNA, and final sample selection and exclusion for association analysis. The Controls in the Skin Cancer Nested Case-Control Study within the NHS The promising SNPs from the initial GWAS were further genotyped among 870 controls in the skin cancer nested case-control study within the NHS. The distribution of risk factors for skin cancer in the subcohort of those who donated blood samples was very similar to that in the overall cohort [2]. A common control series was randomly selected from participants who gave a blood sample and were free of diagnosed skin cancer up to and including the questionnaire cycle in which the corresponding case was diagnosed. Health Professionals Followup Study (HPFS) In 1986, 51,529 men from all 50 U.S. States in health professions (dentists, pharmacists, optometrists, osteopath physicians, podiatrists, and veterinarians) aged 40–75 answered a detailed mailed questionnaire, forming the basis of the study. Between 1993 and 1994, 18,159 study participants provided blood samples by overnight courier. The information on natural hair color and eye color was collected in the 1988 questionnaire, and the information on tanning ability was asked in the 1992 questionnaire. The Diabetes Nested Case-Control Studies within the NHS and HPFS Two additional studies were used to genotype novel pigmentation loci: 3,750 samples from the nested case-control study of diabetes in the NHS and 2,405 samples from the nested case-control study of diabetes in the HPFS [27]. All samples that we used were cases and controls from these two studies. Cases were incident cases of diabetes after blood collection, and controls were matched on age. Controlling for case-control status made no material difference to the results. There was no sample overlap among the initial GWAS, the skin cancer case-control study, and the two diabetes case-control studies. The study protocol was approved by the Institutional Review Board of Brigham and Women's Hospital and Harvard School of Public Health. Informed consent was obtained from all patients. Australian Study from the Queensland Institute of Medical Research (QIMR) The Australian sample comprised 1,442 parents of twins taking part in a long-running study of melanoma risk factors [9],[18],[28],[29]. Participants rated their own hair color (at age 20 years) on a five-point classification (blonde, light brown, dark brown, black, red), eye color (blue/grey, green/hazel, brown), and skin color (light, medium, or dark). Statistical Analysis For the primary analysis of hair color we regressed an ordinal coding for hair color (1 = black; 2 = dark brown; 3 = light brown; 4 = blonde; and 5 = red) on an ordinal coding for genotype (0, 1 or 2 copies of SNP minor allele) separately for each SNP that passed quality control filters [8]. Crude analyses that did not adjust for any other variables showed evidence of systematic bias (see results); as this bias was greatly reduced by adjusting for the four largest principal components of genetic variation, all subsequent association analyses in the initial GWAS included these four components in the regression model. These principal components were calculated for all individuals on the basis of ca. 10,000 unlinked markers using the EIGENSTRAT software [8],[11]. The top four eigenvectors were chosen on the basis of significant (p<0.05) Tracy-Wisdom tests [30]. Adjusting for up to the top 50 principal components did not further reduce the genomic control inflation factor λGC. We chose markers for genotyping in subsequent validation studies based on the p-values for association from the primary analysis. Partial correlation coefficients (i.e., adjusted r2 or the percent of residual variance explained by the SNP marker) were calculated from the linear regressions adjusted for the top four principal components of genetic variation. There is some evidence that determinants of hair color may act along two phenotypic axes: red hair color versus non-red color and light-dark variation among those without red hair. For example, alleles at the MC1R locus primarily determine presence or absence of red hair [31]. Hence, we conducted further analyses among individuals without red hair and comparing those with red hair to those without to evaluate whether discovered loci act on one or both phenotypic axes. We regressed the ordinal coding for hair color on minor allele counts excluding the individuals with red hair and used logistic regression to test the association between the ordinal genotype coding and a binary red-hair phenotype (red vs. non-red hair color). The regression parameter beta refers to the mean change in hair color scoring (or change in log odds of red hair for red hair analyses) per copy of the SNP minor allele. We also used linear regression to test association between minor allele counts and self-reported tanning in response to sunlight (1 = deep tan, 2 = average tan, 3 = light tan, 4 = no tan), eye color (1 = brown/dark, 2 = hazel/green/medium, and 3 = blue/light), and skin color (1 = black, 2 = medium, and 3 = fair). Pooled analyses of multiple studies were conducted by merging data sets and including separate baseline parameters for each study. Genotyping in Followup Studies The TaqMan/BioTrove assays on the 31 SNPs in the skin cancer controls were performed at the Dana Farber/Harvard Cancer Center Polymorphism Detection Core (primers and probe sequences are available on request). Two loci (IRF4 rs12203592 and SLC24A4 rs12896399) were further genotyped in diabetes samples in the NHS and HPFS studies using the Taqman assay. Laboratory personnel were blinded to the case-control status, and 10% blinded quality control samples were inserted to validate genotyping procedures; concordance for the blinded samples was 100%. Primers, probes, and conditions for genotyping assays are available upon request. For the Australian study, genotyping was performed as a single multiplex reaction on the Sequenom high-throughput genotyping platform on IRF4 SNP rs12203592 and EXOC2 rs6918152 and the best pigmentation-associated SNPs in the region of 6p25 (rs1540771) reported by Sulem et al. [12]. Supporting Information Figure S1 The LD pattern of SNPs around the MC1R locus on Chromosome 16. (0.06 MB PDF) Click here for additional data file. Table S1 Associations between the 30 most significant SNPs, which were identified in the GWAS of hair color, and tanning ability in the GWAS, the skin cancer controls, and the pooled data. (0.02 MB XLS) Click here for additional data file. Table S2 Associations between the 30 most significant SNPs, which were identified in the GWAS of hair color, and skin color in the skin cancer controls. (0.02 MB XLS) Click here for additional data file. Table S3 Associations between SNPs around the MC1R locus on chromosome 16 and hair color and tanning ability. (0.02 MB XLS) Click here for additional data file. Table S4 The associations between hair color and the 60 SNPs in Supplementary Table 1 of Sulem et al. [12]. (0.03 MB XLS) Click here for additional data file. Table S5 The associations between SNPs in the MATP, IRF4, SLC24A4, and HERC2 genes and EVs (from 1 to 10). (0.02 MB XLS) Click here for additional data file.
- Record: found
- Abstract: found
- Article: not found
How reliable are empirical genomic scans for selective sweeps?
Kosuke Teshima, Graham Coop, Molly Przeworski (2006)
The beneficial substitution of an allele shapes patterns of genetic variation at linked sites. Thus, in principle, adaptations can be mapped by looking for the signature of directional selection in polymorphism data. In practice, such efforts are hampered by the need for an accurate characterization of the demographic history of the species and of the effects of positive selection. In an attempt to circumvent these difficulties, researchers are increasingly taking a purely empirical approach, in which a large number of genomic regions are ordered by summaries of the polymorphism data, and loci with extreme values are considered to be likely targets of positive selection. We evaluated the reliability of the "empirical" approach, focusing on applications to human data and to maize. To do so, we considered a coalescent model of directional selection in a sensible demographic setting, allowing for selection on standing variation as well as on a new mutation. Our simulations suggest that while empirical approaches will identify several interesting candidates, they will also miss many--in some cases, most--loci of interest. The extent of the trade-off depends on the mode of positive selection and the demographic history of the population. Specifically, the false-discovery rate is higher when directional selection involves a recessive rather than a co-dominant allele, when it acts on a previously neutral rather than a new allele, and when the population has experienced a population bottleneck rather than maintained a constant size. One implication of these results is that, insofar as attributes of the beneficial mutation (e.g., the dominance coefficient) affect the power to detect targets of selection, genomic scans will yield an unrepresentative subset of loci that contribute to adaptations.
Author and article information
Contributors
: Role: Editor
Journal
Journal ID (nlm-ta): PLoS Genet
Journal ID (publisher-id): plos
Journal ID (pmc): plosgen
Title:
PLoS Genetics
Publisher:
Public Library of Science
(San Francisco, USA
)
ISSN
(Print):
1553-7390
ISSN
(Electronic):
1553-7404
Publication date Collection:
March
2010
Publication date
(Print):
March
2010
Publication date
(Electronic):
5
March
2010
Volume: 6
Issue: 3
Electronic Location Identifier: e1000867
Affiliations
Author notes
* E-mail:
esteban.parra@
123456utoronto.ca
Article
Publisher ID: 09-PLGE-RA-1957R2DOI: 10.1371/journal.pgen.1000867
PMC ID: 2832666
PubMed ID: 20221248
SO-VID: 5bb00f72-2e17-45bb-b415-342785fc571d
Copyright © Edwards et al. This is an open-access article distributed under the terms of the Creative
Commons Attribution License, which permits unrestricted use, distribution, and reproduction
in any medium, provided the original author and source are credited.
History
Date
received
: 9
November
2009
Date
accepted
: 30
January
2010
Page count
Pages: 8
Categories
Subject:
Research Article
Subject:
Evolutionary Biology/Human Evolution
Subject:
Genetics and Genomics/Population Genetics
ScienceOpen disciplines: Genetics
Data availability:
ScienceOpen disciplines: Genetics