Welcome international healthcare professionals

This site is no longer supported and will not be updated with new content. You are welcome to browse and download all content already included in the site. Please note you will have to register your email address to access the site.

You are here

MC1R, SLC45A2 and TYR genetic variants involved in melanoma susceptibility in Southern European populations: Results from a Meta-analysis

European Journal of Cancer, 14, 48, pages 2183 - 2191


Background and methods

Seven genetic biomarkers previously associated with melanoma were analysed in a meta-analysis conducted in three South European populations: five red hair colour (RHC) MC1R alleles, one SLC45A2 variant (p.Phe374Leu) and one thermosensitive TYR variant (p.Arg402Gln). The study included 1639 melanoma patients and 1342 control subjects.


The estimated odds ratio (OR) associated with carrying at least one MC1R RHC variant was 2.18 (95% confidence interval (CI): 1.86–2.55; p-value = 1.02 × 10−21), with an additive effect for carrying two RHC variants (OR: 5.02, 95% CI: 2.88–8.94, p-value = 3.91 × 10−8). The SLC45A2 variant, p.Phe374Leu, was significantly and strongly protective for melanoma in the three South European populations studied, with an overall OR value of 0.41 (95% CI: 0.33–0.50; p-value = 3.50 × 10−17). The association with melanoma of the TYR variant p.Arg402Gln was also statistically significant (OR: 1.50; 95% CI: 1.11–2.04; p-value = 0.0089). Adjustment for all clinical potential confounders showed that melanoma risks attributable to MC1R and SLC45A2 variants strongly persisted (OR: 2.01 95% CI: 1.49–2.72 and OR: 0.50, 95% CI: 0.31–0.80, respectively), while the association of TYR p.Arg402Gln was no longer significant. In addition, stratification of clinical melanoma risk factors showed that the risk of melanoma was strong in those individuals who did not have clinical risk factors.


In conclusion, our results show without ambiguity that in South European populations, MC1R RHC and SCL45A2 p.Phe374Leu variants are strong melanoma risk predictors, notably in those individuals who would not be identified as high risk based on their phenotypes or exposures alone. The use of these biomarkers in clinical practice could be promising and warrants further discussion.

Keywords: MC1R, SLC45A2, TYR, Polymorphism, Meta-analysis.

1. Introduction

Malignant melanoma (MM) incidence is currently increasing faster than that of any other malignancy in almost all Western countries. 1 This incidence is influenced by geographical parameters, with areas closer to the equator and higher in altitude generally having higher rates, indicating that ultraviolet radiation (UVR) plays an important role in the development of the disease. 1 Incidence is also higher in individuals with fair skin than in those with dark skin, suggesting that skin colour, which is known to be related to the degree of protection against UVR, is also important.2 and 3 Therefore, genes involved in the determination of skin colour and tanning response are potentially implicated in MM predisposition, and may be useful predictors of MM risk in the general population. 4

More than 120 genes involved in pigmentation processes, such as maturation, transport and distribution of melanosomes, have been identified through animal models. 5 However, only several genes have been identified as containing common genetic variants associated with human pigmentation in the normal range.3, 6, 7, and 8 Recent genome-wide association studies (GWAS) have unveiled single nucleotide polymorphisms (SNPs) or genetic variants in MC1R, TPCN2, ASIP, KITLG, NCKX5, TYR, IRF4, OCA2, SLC45A2 and TYRP1 pigmentation genes. These findings emphasise the contribution of pigmentation pathways to melanoma predisposition and tumourigenesis through gene-environment interactions.6, 7, 8, and 9

The melanocortin-1 receptor gene (MC1R, [MIM# 155555]), located in chromosome 16q, encodes the melanocyte stimulating hormone receptor, a membrane bound protein central to pathways that signal the production of melanin. Inherited variation in MC1R is a robust genetic marker for increased risk of melanoma. The frequency of MC1R variants in the general population suggests that a considerable proportion of melanoma risk may be attributable to these genetic variants. 10

Oculocutaneous albinism syndromes (OCA) consist essentially of a failure to synthesise melanin, the main contributor to human skin colour.11 and 12 Mutations in each of the genes OCA2 (MIM# 611409), TYR (MIM# 606933), TYRP1 (MIM# 115501) and SLC45A2 (also known as MATP (MIM# 606202)) (melanosome enzymes) are responsible for the four distinct OCA subtypes. 12

The gene encoding tyrosinase (TYR, MIM# 606933) is located on chromosome 11q and consists of five exons coding for a 529 amino acids protein. Tyrosinase is one of the important enzymes that play a key role in pigmentation processes, catalysing the first two steps, and at least one subsequent step, in the conversion of tyrosine to melanin. 13 Mutations in this gene result in oculocutaneous albinism and skin pigmentation variation. One TYR polymorphism, p. Arg402Gln (rs1126809), has been described to be associated with eye colour and skin type. 14

The SLC45A2 gene is located on chromosome 5p, comprised of seven exons spanning 40 kb, and encodes a 530 amino acid protein presumably located in the melanosome membrane. 15 SLC45A2 exhibits structural homology to plant sucrose-proton symporters and probably directs the traffic of melanosomal proteins and other substances to the melanosomes. 16 SLC45A2 mutations cause pigmentation variation in the medaka, 17 mouse,15 and 18 horse, 19 chicken and Japanese quail. 20 In humans, pathogenic mutations in SLC45A2 lead to type IV oculocutaneous albinism. SLC45A2 mutations disrupt tyrosinase processing and trafficking at the post-Golgi level.21 and 22 Two human SLC45A2 polymorphisms, p.Phe374Leu (rs16891982) and p.Glu272Lys (rs26722), display differing population frequency distributions. Interestingly, these variants have been shown to be significantly associated with dark hair, skin and eye pigmentation in a Caucasian population and recently, p.Phe374Leu has been associated with melanoma.11, 23, and 24 Thus, the SLC45A2 gene has been proposed as a melanoma susceptibility gene in light-skinned populations.23 and 24

As pigmentation characteristics and melanoma risk are closely correlated, we investigated the impact of these three pigmentation genes (MC1R, TYR and SLC45A2) on melanoma risk by performing a meta-analysis comprising three Mediterranean populations from France, Italy and Spain. We point out the potential importance of pigmentation genes in MM susceptibility, and specifically, we confirmed the SLC45A2 gene as a novel protective melanoma low penetrance susceptibility gene.

2. Patients and methods

2.1. Study subjects and data collection

An overall number of 1639 melanoma patients and 1342 control subjects were included in the study23, 24, 25, and 26 comprising three different Mediterranean populations. One thousand thirty-seven French melanoma patients of which 238 familial melanoma patients and 117 multiple primary melanoma patients were recruited from the MELAN-COHORT including all melanoma patients from Departments of Dermatology of six different hospitals in Paris (Bichat, Percy, Ambroise Pare´, Henri Mondor, Cochin and Saint-Louis hospitals). Two hundred and seventy-one Italian melanoma patients were enrolled at the Department of Dermatology of the University of L’Aquila, L’Aquila, Italy. Finally, 131 Spanish melanoma patients were recruited from the Department of Dermatology of Gregorio Marañón hospital in Madrid.

The control groups, matched by ethnicity and sex to the case group, were recruited among patients affected by diseases unrelated to melanoma attending the same hospitals and were composed of 925 French, 171 Italian and 245 Spanish control subjects with no personal or family history of skin cancer.

A standardised questionnaire was used to collect information on pigmentation characteristics such as eye, hair and skin colour, number of nevi, presence of solar lentigines, sun exposure habits and presence of childhood sunburns. Fitzpatrick’s classification of skin type, tumour location, Breslow thickness and personal or family history of cancer was also included in the questionnaire. Fitzpatrick’s classification of skin type was extracted from the medical record of cases only ( Supplementary Table 1) . The study was approved by the Local Ethical Committees and The Declaration of Helsinki Principles was followed. Informed consent was obtained from all the patients and control subjects enrolled in the study.

Genomic DNA from cases and controls was isolated from peripheral blood leukocytes using QIAamp Blood Mini Kit (QIAgen GmbH, Hilden, Germany) and MagNA Pure LC Instrument (according to the manufacturer’s protocol [Roche Molecular Biochemicals AQ2, Mannheim, Germany]). Some samples were extracted using the traditional saline method.

2.2. Genotyping MC1R, SLC45A2 and TYR variants

MC1R, SLC45A2 and TYR cDNA reference sequences with GeneBank accession numbers NM_002386.2 , NM_016180.3 and NM_000372.4 were used.

The MC1R polymorphisms retained for genetic analysis were those associated with the red hair colour phenotype (RHC alleles), and included c.252 C>A p.Asp84Glu, c.425 G>A p.Arg142His, c.451 T>C p.Arg151Cys, c.478 C>T p.Arg160Trp and c.880 G>C p.Asp294His.3, 27, and 28 Details of the genes and SNPs studied, including MIM code, location, encoded protein, gene function, nucleotide changes and the context sequence are provided in Supplementary Table 2 . French and Italian samples were genotyped for these SNPs using the KBiosciences PCR SNP genotyping system (KASPAR SNP Genotyping System KBiosciences, Hoddesdon Herts, United Kingdom (UK)), which is a homogeneous fluorescent genotyping system, using a unique form of allele-specific polymerase chain reaction (PCR) allowing allelic discrimination.24 and 25 The MC1R coding region was amplified in the Spanish samples by polymerase chain reaction PCR using two overlapping pairs of primers which have been described previously 29 and the coding region was completely sequenced. 26

We selected two non-synonymous SNPs described as being associated with melanoma risk in two low penetrance genes: c.1122 C>G p.Phe374Leu in SLC45A2 and c.1707 G>A p.Arg402Gln in TYR (NCBI dbSNP rs16891982 and rs1126809, respectively)23 and 24 to perform a meta-analysis. Details of the genes and SNPs studied, including MIM code, location, encoded protein, gene function, nucleotide changes and the context sequence are provided in Supplementary Table 2 . For these two SNPs, French and Italian samples were genotyped by Kaspar technology 24 while Spanish samples were genotyped by Taqman technology 23 (Genotyping of the TYR variant, p.Arg402Gln, was not available for the Italian set). The PCR primers and probes were designated by Applied Biosystems (Foster City, CA) using their Custom Taqman SNP genotyping assays. All genotyping reactions were performed using TAQMAN SNP Genotyping Assay-allelic discrimination method (Applied Biosystems, Foster City, CA) (assay primers, probes and PCR conditions will be supplied upon request).

MC1R variants genotyping were successfully completed in 1397 patients and 1254 controls. SLC45A2 variant was genotyped efficiently in 1449 patients and 1327 controls. Finally, TYR variant was genotyped in 1172 patients and 1158 control subjects.

2.3. Statistical analysis

Statistical analyses were performed using the R software (version 2.12.2). Significance was declared at the 5% level. Hardy–Weinberg equilibrium has been checked for each genetic marker by applying of chi2 test for adjustment. Association was tested according to four modes of inheritance (additive, recessive, dominant and genotypic) with Logistic Regression (glm() R procedure) adjusted in both gender and age. All odds ratios (ORs) are reported with their 95% confidence interval (CI). Considering the number of tests performed, and the high levels of associations, most of them would remain significant if a correction for multiple testing was applied. We conducted a meta-analysis in the French, the Italian and the Spanish populations by applying a Cochran–Mantel–Haenszel test, 30 as previously described. 31 Impact on pigmentation characteristics was further analysed by conducting analyses stratified on the various clinical melanoma risk factors. A Woolf test for homogeneity of the populations analysed was performed. Impact on pigmentation characteristics was further analysed by conducting multivariate analyses on the various clinical melanoma risk factors.

In order to specify the impact of each genetic variant on melanoma risk, we also calculate their population attributable fraction (PAF), defined as the proportion of disease cases in a population that is attributable to a particular exposure or cause. Using the relevant variables from the Poisson regression models, individual PAFs for each of the variables were calculated on the basis of RRs generated by new Poisson regression models with the variables classified dichotomously (exposed/unexposed). The population belonging to the lowest level of the variables in the initial Poisson regression models was defined as unexposed and the other levels were aggregated into the exposed group. The population-attributable fraction was calculated according to the formula ((RR − 1)/RR) × the proportion of cases in the exposed population, where RR was the risk in the exposed population. 32 Interactions between risk factors were assessed by fitting a Logistic Regression with a multiplicative interaction term as implemented in Plink. 33

3. Results

3.1. Association between SNPs genotypes and melanoma risk: meta-analysis

For this meta-analysis, we have used previously published results from the French and the Spanish data set.23 and 24 Additional genotyping of MC1R and SLC45A2 was performed in Italian patients and controls from l’Aquila, and genotyping of the TYR variant was conducted in all the French patients and controls. The meta-analysis comprised 1639 melanoma patients and 1343 cancer-free control subjects. Results are shown in Table 1 .

Table 1 Meta-analysis of genetic risk factors with Malignant Melanoma.

Gene SNP Genotypes Cases n (%) Controls n (%) OR (95% CI) p-Value c
A. Genotypic frequencies across MC1R, SLC45A2 and TYR genes and assessment of individual associations with Malignant Melanoma
MC1R RHC a 0/0 882 (63.1) 1011 (80.6) 2.18 (1.86–2.55) 1.02×10−21
  variants lowast 1/0 440 (31.5) 229 (18.3)    
    1/1 75 (5.4) 14 (1.1)    
    MAF b 0.211 0.102    
SLC45A2 rs16891982 lowastlowast CC 1329 (91.7) 1052 (79.3) 0.41 (0.33–0.50) 3.50×10−17
  Phe374Leu CG 110 (7.6) 243 (18.3)    
    GG 10 (0.7) 32 (2.4)    
    MAF b 0.045 0.116    
TYR rs1126809 lowastlowastlowast GG 584 (49.8) 599 (51.7) 1.50 (1.11–2.04) 0.0089
  Arg402Gln AG 475 (40.5) 482 (41.6)    
    AA 113 (9.7) 77 (6.7)    
    MAF b 0.299 0.275    
      pHWE MAF Controls MAF Cases  
B. Allele frequencies in cases and controls across the three Mediterranean populations
MC1R RHC a France 0.68 0.11 0.22 1.32×10−19
  variants lowast Italy 0.57 0.10 0.20 0.00068
    Spain 0.22 0.07 0.13 0.005
SLC45A2 rs16891982 lowastlowast France 0.0002 0.10 0.04 5.07×10−17
  Phe374Leu Italy 0.89 0.11 0.06 0.011
    Spain 0.60 0.16 0.07 0.0006
TYR rs1126809 lowastlowast France 0.07 0.28 0.30 0.08
  Arg402Gln Spain 0.85 0.26 0.26 0.96

lowast Sequencing was successfully performed in 1397 patients and 1254 controls.

lowastlowast Genotyping was successfully performed in 1449 patients and 1327 controls.

lowastlowastlowast Genotyping was successfully performed in 1172 patients and 1158 controls.

a RHC variants are rs1805006 (c.252 C>A), rs11547464 (c.425 G>A), rs1805007 (c.451 C>G), rs1805008 (c.478 C>T) and rs1805009 (c.880 G>C).

b MAF, minor allele frequency.

c p-Value was calculated by Fisher’s exact test.

OR, odds ratio per minor allele; CI, confidence interval; SNP, single nucleotide polymorphisms.

Bold indicates statistically significant results.

The MC1R gene was only evaluated for the presence of RHC variants. These variants were significantly and strongly associated with melanoma risk, increasing seriously with the number of RHC alleles carried. The estimated OR associated with carrying at least one RHC variant was 2.18 (95% CI: 1.86–2.55, p-value = 1.02 × 10−21); however, OR for carrying two RHC variants was 5.02 (95% CI: 2.82–8.94, p-value = 3.91 × 10−8) (Data are shown in Fig. 1 ).


Fig. 1 Forest plot showing the association of carriers of one or two MC1R red hair colour (RHC) variants with melanoma risk. RHC variants are p.Asp84Glu, p.Arg142His, p.Arg151Cys, p.Arg160Trp, p.Asp294His. Dots represent odds ratio in the three South European populations. Diamond shapes represent pooled results for both stages. The size of the dot is proportional to the number of individuals, and error bars represent 95% confidence intervals. Eight Spanish melanoma cases have two RHC variants; however, the odds ratio is not available due to the absence of controls with two RHC variants.

The p.Phe374Leu variant analysed in the SLC45A2 gene was significantly and strongly protective for melanoma in the three South European populations studied, with an overall value of 0.41 (95% CI: 0.33–0.50, p-value = 3.46 × 10−17) (Results are shown in Fig. 2 ). Minor allele frequency (MAF) was 0.11 for France and Italian populations while MAF was 0.16 for the Spanish individuals.


Fig. 2 Forest plot showing SLC45A2 polymorphism, p.Phe374Leu, and melanoma protection. Dots represent odds ratio in the three South European populations. Diamond shapes represent the overall odds ratio. The size of the dot is proportional to the number of individuals, and error bars represent 95% confidence intervals.

Finally, association with melanoma and the TYR variant studied, p.Arg402Gln, was also statistically significant (OR: 1.50, 95% CI: 1.11–2.04, p-value = 0.0089) although the significance was not as high as for the previous two genes. (OR for each population studied are shown in Fig. 3 ).


Fig. 3 Forest plot showing TYR polymorphism, p.Arg402Gln, and melanoma risk predisposition. Dots represent odds ratio in the three South European populations. Diamond shapes represent the overall odds ratio. The size of the dot is proportional to the number of individuals, and error bars represent 95% confidence intervals.

3.2. Population attributable fraction (PAF)

The genetic risk factors considered for the calculation of the PAF were the presence of MC1R RHC variants, the presence of the C allele of the SLC45A2 polymorphism (374Phe) and the presence of the A risk allele at the TYR variant (402Gln). We obtained a PAF of 21.4% for MC1R RHC variants, 29% for the SLC45A2 variant and 1.89% for the TYR polymorphism.

3.3. Association between SNPs genotypes and phenotypic characteristics

We assessed whether MC1R, SLC45A2 and TYR polymorphisms were associated with various phenotypic characteristics, after adjustment by age, sex and case–control status (Results are shown in Table 2 ). MC1R RHC variants were strongly associated with skin phototypes I–II, presence of ephelides, light hair colour and presence of solar lentigines, and moderately associated with light eye colour. The variant allele p.Phe374Leu of the SLC45A2 gene was strongly associated with skin phototypes III–IV, and dark eye and hair colour, while it was moderately associated with the absence of ephelides and lentigines and with a low number of nevi. Finally, the variant allele p.Arg402Gln of the TYR gene was strongly associated with skin phototypes I–II, and moderately associated with light hair colour and the number of nevi.

Table 2 Association between single nucleotide polymorphisms (SNPs) and various phenotypic characteristics after adjustment on age, sex and case control status.

Characteristic MC1R SLC45A2 TYR
RHC variants lowast rs16891982 rs1126809
OR (95% CI) p-Value OR (95% CI) p-Value OR (95% CI) p-Value
Eye colour 1.48 (1.25–1.75) 5.0×10−6 0.36 (0.29–0.46) 4.6×10−17 1.09 (0.78–1.53) 0.60
Hair colour 2.24 (1.89–2.64) 2.7×10−21 0.34 (0.27–0.42) 1.5×10−21 1.65 (1.20–2.26) 2.1×10−3
Number of Nevi 1.14 (0.95–1.36) 0.16 0.56 (0.42–0.75) 8.2×10−5 1.42 (1.01–2.00) 0.04
Presence/Absence of lentigines 1.70 (1.38–2.10) 9.6×10−7 0.73 (0.55–0.96) 0.03 0.96 (0.64–1.43) 0.84
Presence/Absence of ephelides 3.94 (3.23–4.80) 5.41×10−42 0.44 (0.31–0.61) 1.06×10−6 1.18 (0.82–1.69) 0.37
Fitzpatrick’s phototype 2.80 (2.36–3.33) 8.6×10−32 0.28 (0.21–0.36) 2.2×10−20 2.11 (1.51–2.94) 1.0×10−5

lowast RHC variants are rs1805006 (c.252 C>A), rs11547464 (c.425 G>A), rs1805007 (c.451 C>G), rs1805008 (c.478 C>T) and rs1805009 (c.880 G>C).

OR, odds ratio per minor allele; CI, confidence interval.

Bold indicates statistically significant results.

3.4. Multivariate analyses

We considered hair and eye colour, skin phototype and the number of nevi as potential confounders in a multivariate model. Adjustment for all these potential confounders, plus age and sex, gave ORs shown in Supplementary Table 3 . SLC45A2 p.Phe374Leu polymorphism (OR = 0.50, 95% CI: 0.31–0.80, p-value = 0.004) and MC1R RHC variants (OR = 2.01, 95% CI: 1.49–2.72, p-value = 5 × 10−6) retained statistically significant results even when adjusted for all potential confounders.

3.5. Stratification on phenotypic risk factors

SLC45A2 and MC1R RHC variants were further analysed stratifying for phenotypic risk factors. The p.Phe374Leu SLC45A2 change remained protective in fair phenotypic traits while MC1R RHC variants conferred risk even for those olive/dark complexion characteristics ( Supplementary Table 4 ).

3.6. Interactions

We observed strong interaction with sunburns and eye colour with both MC1R (p = 1.8 × 10−3) and SLC45A2 (p = 0.036) genes on melanoma risk. Furthermore, phototype and sunburns showed statistically significant interaction with the SLC45A2 gene (p = 7.5 × 10−3) and hair colour with the SLC45A2 and TYR genes (p = 0.019). Data is shown in Supplementary Table 5 .

We further characterised the genetic interaction between risk MC1R RHC variants and protective SLC45A2 allele. A dose effect reduction of MC1R RHC variants risk effect is observed when summing up the protective G allele of the SLC45A2 p.Phe374Leu variant. Results are shown in Table 3 .

Table 3 Interactions between protective (SLC45A2) and risk (MC1R) variants and their effect on Melanoma susceptibility.

Joint genotype Cases Controls N OR (95% CI) p-Value
MC1R RHC lowast variants SLC45A2 rs16891982
0/0 CC 740 794 1534 1.00 REF
  G/– 70 206 276 0.36 (0.27–0.49) <1×10−4
0/1 CC 364 185 549 2.11 (1.72–2.59) <1×10−4
  G/– 34 42 76 0.87 (0.55–1.38) 0.56
1/1 CC 59 11 70 5.75 (3.03–10.92) <1×10−4
  G/– 3 3 6 1.07 (0.25–4.66) 1.00

lowast RHC variants are rs1805006 (c.252 C>A), rs11547464 (c.425 G>A), rs1805007 (c.451 C>G), rs1805008 (c.478 C>T) and rs1805009 (c.880 G>C).

OR, odds ratio per minor allele; CI, confidence interval; REF, reference value.

Bold indicates statistically significant results.

4. Discussion

Human pigmentation pathways have been shown to play a crucial role in the pathogenesis of MM. Up to date, only the MC1R low penetrance gene was known to unequivocally account for a substantial variation in the incidence of MM. Nowadays, different studies have suggested that other pigmentation genes such as SLC45A2, TYR, TYRP1, ASIP and OCA2 are also important in MM susceptibility.23, 24, 34, and 35 Genetic variation across populations of different ethnic background, however, has yielded conflicting results concerning the role of these genes on melanoma susceptibility.

To clarify the role of three pigmentation genes, MC1R, SLC45A2, and TYR in melanoma susceptibility in a specific region, we conducted a meta-analysis concerning the role of several melanoma predisposing alleles (MC1R RHC alleles, SLC45A2 p.Leu374Phe and TYR p.Arg402Gln) with melanoma risk in three South European regions (France, Italy and Spain). Meta-analysis represents a useful approach that leads to more conclusive results due to the pooled dataset, having greater power than each of the studies taken individually. Thus, 1639 melanoma cases and 1342 cancer-free control subjects were analysed. Furthermore, questions on whether the association of melanoma with genetic biomarkers may depend on the composition of the population under study, the country or the methodological features of the studies could be addressed by this meta-analysis.

MC1R was already known to have a role in melanoma susceptibility across Mediterranean populations26, 29, 36, 37, 38, and 39 and its implication is strongly confirmed in the current study. The PAF of melanoma associated with MC1R RHC variants has been shown to be important in Mediterranean Europe (16%) compared to Northern Europe (9%), 37 which is also confirmed in this meta-analysis, where the PAF was calculated to be 21%, close to the one calculated by Williams et al. (16%). 37 In addition, our results clearly indicate that MC1R RHC alleles are strong melanoma risk factors that appear to be independent of the presence of clinical melanoma risk factors. This is in concordance with previous studies from France, 29 Greece, 36 Italy 25 and Spain.26 and 38 Melanoma risk attributable to MC1R may arise through the determination of the tanning response of skin to UV light, which can then either ameliorate or exacerbate the genotoxic effects of sunlight. Nevertheless, the relationship between some MC1R variants and melanoma in darkly-pigmented Caucasian populations suggests the MC1R signalling pathway may have an additional role in skin carcinogenesis beyond the UV-filtering differences between dark and fair skin.

We observed a strong association of MC1R with MM, even after stratification on fair pigmentation characteristics (i.e. pale skin and skin type I-II, see Supplementary Table 4 ). By opposite, in the Australian population, Palmer et al. observed an association of MC1R variants with melanoma only in darkly pigmented people, 40 suggesting different role or mechanisms for MC1R variants in these populations regarding melanoma risk. Recently, Kanetsky et al. suggested that MC1R genotyping provides information about melanoma risk in those individuals that would not be identified as high risk based on their phenotype. 41 In our study, RHC alleles were associated with melanoma risk after stratification either for the presence of melanoma clinical risk factors (i.e. fair pigmentation characteristics), or in their absence (i.e. dark skin pigmentation, skin type 3–4, dark hair and eye colour and absence of ephelides). Consequently, given the relatively high PAF (21%) and their independent effect on MM risk, discussions should be opened on whether MC1R RHC variants could be used as melanoma predictive biomarkers in populations from the south of Europe.

In 2008, a novel genetic biomarker in the SLC45A2 gene, rs16891982, was identified as being associated with a genetic predisposition to melanoma.23 and 24 This variant encodes a non-synonymous amino acid change and has previously been reported to be strongly associated with ethnic ancestry and normal human pigmentation variation. 11 In this meta-analysis, we state the role of this variant for melanoma susceptibility, and as shown for MC1R, the effect persists after the adjustment for pigmentation characteristics, protecting even those individuals with a fair phenotype. Leu374 variant confers a protective effect against melanoma in Caucasian populations. The PAF to the p.Leu374Phe variant was particularly high (28%), emphasising the role of this variant in the multifactorial susceptibility to melanoma. There was a decreasing gradient of Leu374 allele frequencies from the Northern Africa to Europe. In our study, the Leu374 allelic MAF was 0.16 in Spanish controls while a slight decrease is observed in French and Italian controls (0.11). The Leu374 MAF in cases was lower, ranging from 0.04 in French melanoma patients to 0.07 in Spanish ones. However, the role of the variant Leu374 in the predisposition to melanoma seems to vary across different countries. For example, this variant was not associated to melanoma in Iceland, 42 where the MAF is 0.02 in the general population. However, it has been consistently associated with MM in other countries such as the Netherlands, Sweden, Austria and Australia.42 and 43

Finally, the TYR p.Arg402Gln variant has previously been associated with melanoma in GWAS studies performed in Caucasian populations. 9 Its association with melanoma is confirmed in the current meta-analysis, although the Gln402 variant seems not to be as strongly associated with melanoma as the other two genes studied. The lower impact of this variant on melanoma susceptibility was confirmed by its low PAF (<2%) compared to those for SLC45A2 and MC1R.

In conclusion, our results show without ambiguity that in South Europe, MC1R RHC and SCL45A2 p.Phe374Leu variants are strong melanoma risk predictors independent of clinical characteristics. Whether or not these biomarkers could be used in clinical practice warrants further discussion.

Conflict of interest statement

None declared.


This study was supported by Grants from the Ministerio de Salud Carlos III (ISCIII) (FI10-00405) and grants from La Societe Francaise de Dermatologie (SFD). M.I-V is funded by the Spanish Ministerio de Educación y Ciencia under a Grant FPI (BES-2008-009234). G.R. is funded by the Ministerio de Salud Carlos III under a ‘Miquel Servet’ contract (CP08-00069). We thank the Madrid College of Lawyers and the patients from the Gregorio Marañon Hospital.

Appendix A. Supplementary data


Download file

Supplementary data 1 Supplementary Table 1.

Download file

Supplementary data 2 Supplementary Table 2.

Download file

Supplementary data 3 Supplementary Table 3.

Download file

Supplementary data 4 Supplementary Table 4.

Download file

Supplementary data 5 Supplementary Table 5.


  • 1 M.A. Tucker, A.M. Goldstein. Melanoma etiology: where are we?. Oncogene. 2003;22(20):3042-3052 Crossref
  • 2 J.N. Bishop, M. Harland, D.T. Bishop. The genetics of melanoma. Br J Hosp Med (Lond). 2006;67(6):299-304
  • 3 J.L. Rees. The genetics of sun sensitivity in humans. Am J Hum Genet. 2004;75(5):739-751 Crossref
  • 4 J.A. Newton Bishop, D.T. Bishop. The genetics of susceptibility to cutaneous melanoma. Drugs Today (Barc). 2005;41(3):193-203
  • 5 I.J. Jackson. Molecular and developmental genetics of mouse coat color. Annu Rev Genet. 1994;28:189-217 Crossref
  • 6 P. Sulem, D.F. Gudbjartsson, S.N. Stacey, et al. Genetic determinants of hair, eye and skin pigmentation in Europeans. Nat Genet. 2007;39(12):1443-1452 Crossref
  • 7 P. Sulem, D.F. Gudbjartsson, S.N. Stacey, et al. Two newly identified genetic determinants of pigmentation in Europeans. Nat Genet. 2008;40(7):835-837 Crossref
  • 8 J. Han, P. Kraft, H. Nan, et al. A genome-wide association study identifies novel alleles associated with hair color and skin pigmentation. PLoS Genet. 2008;4(5):e1000074 Crossref
  • 9 D.F. Gudbjartsson, P. Sulem, S.N. Stacey, et al. ASIP and TYR pigmentation variants associate with cutaneous melanoma and basal cell carcinoma. Nat Genet. 2008;40(7):886-891 Crossref
  • 10 M.R. Gerstenblith, A.M. Goldstein, M.C. Fargnoli, K. Peris, M.T. Landi. Comprehensive evaluation of allele frequency differences of MC1R variants across populations. Hum Mutat. 2007;28(5):495-505 Crossref
  • 11 J. Graf, R. Hodgson, A. van Daal. Single nucleotide polymorphisms in the MATP gene are associated with normal human pigmentation variation. Hum Mutat. 2005;25(3):278-284 Crossref
  • 12 R.A. Sturm, R.D. Teasdale, N.F. Box. Human pigmentation genes: identification, structure and consequences of polymorphic variation. Gene. 2001;277(1–2):49-62 Crossref
  • 13 R.A. Spritz. Molecular genetics of oculocutaneous albinism. Hum Mol Genet. 1994;3 Spec No:1469-1475
  • 14 H. Nan, P. Kraft, D.J. Hunter, J. Han. Genetic variants in pigmentation genes, pigmentary phenotypes, and risk of skin cancer in Caucasians. Int J Cancer. 2009;125(4):909-917 Crossref
  • 15 J.M. Newton, O. Cohen-Barak, N. Hagiwara, et al. Mutations in the human orthologue of the mouse underwhite gene (uw) underlie a new form of oculocutaneous albinism, OCA4. Am J Hum Genet. 2001;69(5):981-988 Crossref
  • 16 G.E. Costin, J.C. Valencia, W.D. Vieira, M.L. Lamoreux, V.J. Hearing. Tyrosinase processing and intracellular trafficking is disrupted in mouse primary melanocytes carrying the underwhite (uw) mutation. A model for oculocutaneous albinism (OCA) type 4. J Cell Sci. 2003;116(Pt. 15):3203-3212 Crossref
  • 17 S. Fukamachi, A. Shimada, A. Shima. Mutations in the gene encoding B, a novel transporter protein, reduce melanin content in medaka. Nat Genet. 2001;28(4):381-385 Crossref
  • 18 J. Du, D.E. Fisher. Identification of Aim-1 as the underwhite mouse mutant and its transcriptional regulation by MITF. J Biol Chem. 2002;277(1):402-406
  • 19 D. Mariat, S. Taourit, G. Guerin. A mutation in the MATP gene causes the cream coat colour in the horse. Genet Sel Evol. 2003;35(1):119-133
  • 20 U. Gunnarsson, A.R. Hellstrom, M. Tixier-Boichard, et al. Mutations in SLC45A2 cause plumage color variation in chicken and Japanese quail. Genetics. 2007;175(2):867-877 Crossref
  • 21 V.J. Hearing. Biogenesis of pigment granules: a sensitive way to regulate melanocyte function. J Dermatol Sci. 2005;37(1):3-14 Crossref
  • 22 T. Kushimoto, J.C. Valencia, G.E. Costin, et al. The Seiji memorial lecture: the melanosome: an ideal model to study cellular differentiation. Pigment Cell Res. 2003;16(3):237-244 Crossref
  • 23 L.P. Fernandez, R.L. Milne, G. Pita, et al. SLC45A2: a novel malignant melanoma-associated gene. Hum Mutat. 2008;29(9):1161-1167 Crossref
  • 24 M. Guedj, A. Bourillon, C. Combadieres, et al. Variants of the MATP/SLC45A2 gene are protective for melanoma in the French population. Hum Mutat. 2008;29(9):1154-1160 Crossref
  • 25 M.C. Fargnoli, E. Altobelli, G. Keller, et al. Contribution of melanocortin-1 receptor gene variants to sporadic cutaneous melanoma risk in a population in central Italy: a case-control study. Melanoma Res. 2006;16(2):175-182
  • 26 L. Fernandez, R. Milne, J. Bravo, et al. MC1R: three novel variants identified in a malignant melanoma association study in the Spanish population. Carcinogenesis. 2007;28(8):1659-1664 Crossref
  • 27 D.L. Duffy, N.F. Box, W. Chen, et al. Interactive effects of MC1R and OCA2 on melanoma risk phenotypes. Hum Mol Genet. 2004;13(4):447-461
  • 28 N. Flanagan, E. Healy, A. Ray, et al. Pleiotropic effects of the melanocortin 1 receptor (MC1R) gene on human pigmentation. Hum Mol Genet. 2000;9(17):2531-2537 Crossref
  • 29 E. Matichard, P. Verpillat, R. Meziani, et al. Melanocortin 1 receptor (MC1R) gene variants may increase the risk of melanoma in France independently of clinical risk factors and UV exposure. J Med Genet. 2004;41(2):e13
  • 30 N. Mantel, W. Haenszel. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst. 1959;22(4):719-748
  • 31 J. Di Lucca, M. Guedj, J.J. Lacapere, et al. Variants of the xeroderma pigmentosum variant gene (POLH) are associated with melanoma risk. Eur J Cancer. 2009;45(18):3228-3236 Crossref
  • 32 O.S. Miettinen. Proportion of disease caused or prevented by a given exposure, trait or intervention. Am J Epidemiol. 1974;99(5):325-332
  • 33 S. Purcell, B. Neale, K. Todd-Brown, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81(3):559-575 Crossref
  • 34 M.R. Gerstenblith, J. Shi, M.T. Landi. Genome-wide association studies of pigmentation and skin cancer: a review and meta-analysis. Pigment Cell Melanoma Res. 2010;23(5):587-606 Crossref
  • 35 M. Ibarrola-Villava, L.P. Fernandez, S. Alonso, et al. A customized pigmentation SNP array identifies a novel SNP associated with melanoma predisposition in the SLC45A2 gene. PLoS One. 2011;6(4):e19271 Crossref
  • 36 A.J. Stratigos, G. Dimisianos, V. Nikolaou, et al. Melanocortin receptor-1 gene polymorphisms and the risk of cutaneous melanoma in a low-risk southern European population. J Invest Dermatol. 2006;126(8):1842-1849 Crossref
  • 37 P.F. Williams, C.M. Olsen, N.K. Hayward, D.C. Whiteman. Melanocortin- 1-receptor and risk of cutaneous melanoma: a meta-analysis and estimates of population burden. Int J Cancer. 2011;
  • 38 M. Ibarrola-Villava, L.P. Fernandez, G. Pita, et al. Genetic analysis of three important genes in pigmentation and melanoma susceptibility: CDKN2A, MC1R and HERC2/OCA2. Exp Dermatol. 2010;19(9):836-844 Crossref
  • 39 D. Scherer, E. Nagore, J.L. Bermejo, et al. Melanocortin receptor 1 variants and melanoma risk: a study of 2 European populations. Int J Cancer. 2009;125(8):1868-1875 Crossref
  • 40 J.S. Palmer, D.L. Duffy, N.F. Box, et al. Melanocortin-1 receptor polymorphisms and risk of melanoma: is the association explained solely by pigmentation phenotype?. Am J Hum Genet. 2000;66(1):176-186 Crossref
  • 41 P.A. Kanetsky, S. Panossian, D.E. Elder, et al. Does MC1R genotype convey information about melanoma risk beyond risk phenotypes?. Cancer. 2010;116(10):2416-2428
  • 42 S.N. Stacey, P. Sulem, G. Masson, et al. New common variants affecting susceptibility to basal cell carcinoma. Nat Genet. 2009;41(8):909-914 Crossref
  • 43 D.L. Duffy, Z.Z. Zhao, R.A. Sturm, et al. Multiple pigmentation gene polymorphisms account for a substantial proportion of risk of cutaneous malignant melanoma. J Invest Dermatol. 2010;130(2):520-528 Crossref


a Department of Haematology and Medical Oncology, Fundacion Investigacion Hospital Clinico-INCLIVA, Valencia, Spain

b Genetic and Biochemical laboratory, Bichat Claude Bernard Hospital, APHP, Paris 7 University, Paris, France

c Inserm U976, Cutaneous Research Center, Unit EA “Genetic Biomarkers of Skin Cancers”, Saint Louis Hospital, Paris 7 University, Paris, France

d Institute of Biomedical Research, Alberto Sols, CSIC, University Autonoma of Madrid, Madrid, Spain

e Department of Dermatology, Bichat Claude Bernard Hospital, APHP, Paris, France

f Department of Dermatology, Saint Louis Hospital, APHP, Paris 7 University, Paris, France

g Department of Dermatology and Research Unit EA 4339 “Skin, environment, and cancer”, Ambroise Paré University Hospital, Boulogne-Billancourt, APHP, University of Versailles-Saint Quentin en Yvelines, France

h Department of Dermatology, University of L’Aquila, L’Aquila, Italy

i Department of Dermatology, Gregorio Marañon Hospital, Madrid, Spain

lowast Corresponding authors: Address: Department of Haematology and Medical Oncology, Fundacion Investigacion Hospital Clinico, Av. Blasco Ibañez 17, 46010 Valencia, Spain. Tel.: +34 963862894; fax: +34 963987860 (G. Ribas), Laboratoire de Biochimie Hormonale et Génétique, Hôpital Bichat Claude Bernard, 46 rue Henri Huchard, 75018 Paris, France. Tel.: +33 140258551; fax: +33 140258785 (N. Soufir).

j Both authors contributed equally.