Few genetic risk factors for melanoma have been examined and confirmed. We examined five important candidate genes (MGMT, ICAM-5, VEGF, MPO, and TNF-α) involved in direct reversal DNA repair, cell growth, oxidative stress, and inflammatory response, in relation to melanoma susceptibility. Eight candidate SNPs in the five genes were selected according to previous studies (MGMT 84Leu>Phe (rs12917), MGMT 143Ile>Val (rs2308321) [1], VEGF 398G>C (rs2010963), VEGF 1967C>T (rs3025039) [2], MPO −764T>C (rs2243828) [3], TNF-α−487A>G (rs1800629) [4] and ICAM5 301Val>Ile (rs1056538), ICAM5 −542C>G (rs281439) [5]). Due to genotyping difficulty, the MPO −463G>A (rs2333227) was replaced by the MPO −764 T>C (rs2243828). Genotyping concordance between the two SNPs was 100%. (http://snp500cancer.nci.nih.gov). We investigated the associations between these genetic variants and melanoma susceptibility in a nested case-control study of 219 incident melanoma cases and 219 matched controls within the Nurses’ Health Study (NHS) Cohort. The detailed information about the study population and laboratory assays were described elsewhere [6].
The detailed statistical analyses have been published previously [6]. We used unconditional logistic regression to evaluate the association between genotype and melanoma risk, firstly adjusting for matching factors (age and race (Caucasians and missing)), and then additionally adjusting for other melanoma risk factors. We evaluated the interactions between SNPs and melanoma risk factors (constitutional susceptibility score and cumulative sun exposure with a bathing suit) on melanoma susceptibility. We used the expectation maximization algorithm to estimate haplotype frequencies in cases and controls for the SNPs MGMT 84Leu>Phe and 143Ile>Val. We also explored the potential modification by antioxidants intake on the association between the MPO polymorphism and melanoma risk (See previous report for data collection and analysis [7]).
No departure from the Hardy-Weinberg Equilibrium among controls was observed for all the 8 SNPs. We observed that the MGMT 84Leu>Phe polymorphism was associated with melanoma susceptibility. The Phe carriers had a higher risk of melanoma than non-carriers (84Leu/Leu) with OR (95%CI) 1.55 (0.99–2.44). The association was attenuated in the multivariate model. The MGMT gene encodes a direct reversal DNA repair protein that removes alkyl or methyl adducts from the O6 position of guanine to an internal cysteine residue at codon 145 of the protein [2]. An animal study suggested that MGMT protein prevents N-nitroso-N-methylurea-initiated skin carcinogenesis through avoiding tumor initiation triggered by alkylation damage [8]. There are three common non-synonymous polymorphisms in the MGMT gene, 84Leu>Phe, 143Ile>Val, and 178 Lys>Arg, and the latter two are in linkage disequilibrium. The Leu84 and Ile143 are close to the reactive cysteine in the 3D structure and may have subtle functional consequences [1]. The minor alleles in the 84Leu>Phe and 143Ile>Val polymorphisms were reported to be inversely associated with the capacity to repair tobacco-induced DNA damage [9]. This evidence suggests that these two minor alleles may confer a higher cancer susceptibility. Consistently, we observed a higher risk of melanoma for the 84Phe carriers and the 143Val carriers compared to their wildtypes. The 84Phe carriers had about a 50% increased risk of melanoma compared to non-carriers (84Leu/Leu). The 143Val carriers had an OR of 1.34(0.86–2.09) for melanoma risk compared to non-carriers (143Ile/Ile). Comparing to the common haplotype 84Leu/143Ile, the haplotype 84Phe/143Ile and 84Leu/143Val had an OR (95%CI) of 1.75 (1.11–2.76) and 2.00 (1.21–3.29), respectively. These associations remained significant after adjusting for other risk factors. Our results combined with the functional results of these two SNPs suggest that alkylation-related DNA damage (or other damage repaired by MGMT) may be involved in melanoma carcinogenesis.
Because the ICAM expression in melanocytes was shown to be inhibited by alpha-melanocyte-stimulating hormone (α-MSH), a sun exposure–induced ligand for skin pigmentation [10], we evaluated the interactions between the ICAM5 variants and the constitutional susceptibility score and cumulative sun exposure with a bathing suit on melanoma risk. Comparing to the first tertiles, the ORs (95%CIs) of the third tertiles were 3.25(1.44–7.36) for constitutional susceptibility score and 3.03(1.71–5.39) for cumulative sun exposure with a bathing suit. In the interaction analysis, the highest risk of melanoma was observed among women with the ICAM5 301 Val/Val genotype and highest susceptibility (OR, 5.73; 95%CI, 1.62–20.31) or greatest sun exposure (OR, 4.83; 95%CI, 1.79–13.06). The interactions did not approach statistical significance (Table 2).
Table 2.
Interaction between ICAM5 301 Val>Ile and risk factors on melanoma risk
| Val/Val | Val/Ile | Ile/Ile | ||||
|---|---|---|---|---|---|---|
| Cases/Controls | OR (95%CI) | Cases/Controls | OR (95%CI) | Cases/Controls | OR (95%CI) | |
| Constitutional susceptibility score 1 | ||||||
| Low | 15/12 | 1.00 | 21/28 | 0.83 (0.29, 2.35) | 6/9 | 1.02 (0.25, 4.23) |
| Medium | 13/35 | 0.73 (0.20, 2.68) | 26/34 | 2.15 (0.60, 7.67) | 5/12 | 0.92 (0.19, 4.50) |
| High | 57/21 | 5.73 (1.62, 20.31) | 50/46 | 1.95 (0.58, 6.63) | 20/11 | 3.10 (0.76, 12.62) |
| p for interaction = 0.12 | ||||||
| Cumulative sun exposure with a bathing suit 2 | ||||||
| Low | 14/24 | 1.00 | 16/33 | 1.00 (0.38, 2.66) | 7/9 | 0.98 (0.26, 3.65) |
| Medium | 17/18 | 1.57 (0.57, 4.34) | 22/33 | 1.45 (0.57, 3.68) | 8/10 | 1.96 (0.55, 6.94) |
| High | 38/13 | 4.83 (1.79, 13.06) | 41/29 | 2.30 (0.94, 5.64) | 13/9 | 2.59 (0.76, 8.80) |
| p for interaction = 0.34 | ||||||
Unconditional logistic regression adjusted for the matching variables, family history of melanoma, number of lifetime severe sunburns, sunlamp use or tanning salon attendance, cumulative sun exposure while wearing a bathing suit, and geographic region.
Unconditional logistic regression adjusted for the matching variables, constitutional susceptibility score, family history of melanoma, number of lifetime severe sunburns, sunlamp use or tanning salon attendance, and geographic region.
We did not observe significant results for SNPs in other genes (Table 1), or the interaction between the MPO polymorphism and dietary intake of antioxidants on melanoma risk.
Table 1.
Polymorphisms in selected candidate genes and melanoma risk
| Polymorphism | Cases (%) | Controls (%) | OR (95%CI)1 | OR (95%CI)2 |
|---|---|---|---|---|
| MGMT 84 Leu>Phe (rs12917) | ||||
| Wt | 152 (71.0) | 168 (79.2) | 1.00 | 1.00 |
| Het | 60 (28.0) | 43 (20.3) | 1.55 (0.99, 2.44) | 1.44 (0.87, 2.37) |
| Var | 2 (0.9) | 1 (0.5) | ||
| P value | 0.06 | 0.16 | ||
| MGMT 143 Ile>Val (rs2308321) | ||||
| Wt | 154 (72.0) | 164 (77.4) | 1.00 | 1.00 |
| Het | 54 (25.2) | 46 (21.7) | 1.34 (0.86, 2.09) | 1.25 (0.76, 2.05) |
| Var | 6 (2.8) | 2 (0.9) | ||
| P value | 0.19 | 0.39 | ||
| ICAM5 301 Val>Ile (rs1056538) | ||||
| Wt | 85 (39.9) | 68 (32.7) | 1.00 | 1.00 |
| Het | 97 (45.5) | 108 (51.9) | 0.70 (0.46, 1.08) | 0.78 (0.49, 1.26) |
| Var | 31 (14.6) | 32 (15.4) | 0.75 (0.42, 1.36) | 0.80 (0.41, 1.55) |
| P value | 0.25 | 0.58 | ||
| ICAM5 −542 C>G (rs281439) | ||||
| Wt | 126 (59.4) | 130 (61.6) | 1.00 | 1.00 |
| Het | 71 (33.5) | 76 (36.0) | 1.10 (0.75, 1.63) | 1.04 (0.68, 1.62) |
| Var | 15 (7.1) | 5 (2.4) | ||
| P value | 0.63 | 0.85 | ||
| VEGF 398 G>C (rs2010963) | ||||
| Wt | 88 (42.3) | 93 (46.7) | 1.00 | 1.00 |
| Het | 100 (48.1) | 85 (42.7) | 1.25 (0.83, 1.89) | 1.14 (0.72, 1.81) |
| Var | 20 (9.6) | 21 (10.6) | 1.02 (0.51, 2.02) | 1.38 (0.64, 2.96) |
| P value | 0.55 | 0.68 | ||
| VEGF 1967 C>T (rs3025039) | ||||
| Wt | 157 (75.1) | 163 (79.9) | 1.00 | 1.00 |
| Het | 49 (23.4) | 39 (19.1) | 1.35 (0.84, 2.15) | 1.47 (0.86, 2.49) |
| Var | 3 (1.4) | 2 (1.0) | ||
| P value | 0.21 | 0.16 | ||
| MPO −764 T>C (rs2243828) | ||||
| Wt | 131 (63.0) | 112 (56.9) | 1.00 | 1.00 |
| Het | 68 (32.7) | 76 (38.6) | 0.78 (0.51, 1.19) | 0.90 (0.57, 1.44) |
| Var | 9 (4.3) | 9 (4.6) | 0.85 (0.32, 2.26) | 0.87 (0.29, 2.65) |
| P value | 0.50 | 0.89 | ||
| TNF-α −487 A>G (rs1800629) | ||||
| Wt | 156 (73.6) | 140 (66.4) | 1.00 | 1.00 |
| Het | 46 (21.7) | 61 (28.9) | 0.67 (0.43, 1.05) | 0.63 (0.38, 1.04) |
| Var | 10 (4.7) | 10 (4.7) | 0.89 (0.36, 2.22) | 0.91 (0.33, 2.48) |
| P value | 0.21 | 0.20 | ||
Wt: wild type; Het: heterozygote; Var: homozygous variant.
The number of participants does not sum to total women because of missing data on genotype.
Some percentages among cases or controls do not sum to 100% due to rounding.
The p values were calculated based upon 2-degree-of-freedom (df) test for ICAM5 301 Val>Ile, VEGF 398 G>C, TNF-α −487 A>G, and MPO −764 T>C; 1 df test for the other SNPs.
Unconditional logistic regression adjusted for the matching variables (age and race (Caucasians and missing)).
Unconditional logistic regression adjusted for matching variables, constitutional susceptibility score, family history of melanoma, number of lifetime severe sunburns that blistered (0, 1–5, 6–11, or >11), sunlamp use or tanning salon attendance (yes/no), cumulative sun exposure while wearing a bathing suit, and geographic region.
In conclusion, this is the first study of the MGMT, ICAM5, and MPO genetic polymorphisms in relation to melanoma susceptibility. We found a borderline-significant association between the MGMT84 Leu>Phe and melanoma risk, and significant associations between the MGMT haplotypes and melanoma risk. A higher risk of melanoma was observed among 84Phe or 143Val carriers who have lower alkylation-damage-repair capacity. The statistical power of our study to detect a modest association is limited, and larger studies are warranted to confirm these findings.
Acknowledgements
The authors thank Ms. Pati Soule and Drs. Hardeep Ranu and David Cox for their assistance in genotyping melanoma samples. We are indebted to the participants in the Nurses’ Health Study. We used SAS v9.0 (SAS Institute, Cary, NC) for all statistical analyses.
Grant sponsor: NIH; Grant numbers: CA132175 and CA122838.
References
- 1.Bugni JM, Han J, Tsai MS, Hunter DJ, Samson LD. Genetic association and functional studies of major polymorphic variants of MGMT. DNA Repair (Amst) 2007 Aug 1;6(8):1116–1126. doi: 10.1016/j.dnarep.2007.03.023. [DOI] [PubMed] [Google Scholar]
- 2.Howell WM, Bateman AC, Turner SJ, Collins A, Theaker JM. Influence of vascular endothelial growth factor single nucleotide polymorphisms on tumour development in cutaneous malignant melanoma. Genes Immun. 2002 Jun;3(4):229–232. doi: 10.1038/sj.gene.6363851. [DOI] [PubMed] [Google Scholar]
- 3.Ahn J, Gammon MD, Santella RM, Gaudet MM, Britton JA, Teitelbaum SL, et al. Myeloperoxidase genotype, fruit and vegetable consumption, and breast cancer risk. Cancer Res. 2004 Oct 15;64(20):7634–7639. doi: 10.1158/0008-5472.CAN-04-1843. [DOI] [PubMed] [Google Scholar]
- 4.Howell WM, Turner SJ, Collins A, Bateman AC, Theaker JM. Influence of TNFalpha and LTalpha single nucleotide polymorphisms on susceptibility to and prognosis in cutaneous malignant melanoma in the British population. Eur J Immunogenet. 2002 Feb;29(1):17–23. doi: 10.1046/j.1365-2370.2002.00269.x. [DOI] [PubMed] [Google Scholar]
- 5.Kammerer S, Roth RB, Reneland R, Marnellos G, Hoyal CR, Markward NJ, et al. Large-scale association study identifies ICAM gene region as breast and prostate cancer susceptibility locus. Cancer Res. 2004 Dec 15;64(24):8906–8910. doi: 10.1158/0008-5472.CAN-04-1788. [DOI] [PubMed] [Google Scholar]
- 6.Gu F, Qureshi AA, Niu T, Kraft P, Guo Q, Hunter DJ, et al. Interleukin and interleukin receptor gene polymorphisms and susceptibility to melanoma. Melanoma Res. 2008 Oct;18(5):330–335. doi: 10.1097/CMR.0b013e32830658b2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Han J, Colditz GA, Hunter DJ. Manganese superoxide dismutase polymorphism and risk of skin cancer (United States) Cancer Causes Control. 2007 Feb;18(1):79–89. doi: 10.1007/s10552-006-0079-6. [DOI] [PubMed] [Google Scholar]
- 8.Becker K, Dosch J, Gregel CM, Martin BA, Kaina B. Targeted expression of human O(6)-methylguanine-DNA methyltransferase (MGMT) in transgenic mice protects against tumor initiation in two-stage skin carcinogenesis. Cancer Res. 1996 Jul 15;56(14):3244–3249. [PubMed] [Google Scholar]
- 9.Hill CE, Wickliffe JK, Wolfe KJ, Kinslow CJ, Lopez MS, Abdel-Rahman SZ. The L84F and the I143V polymorphisms in the O6-methylguanine-DNA-methyltransferase (MGMT) gene increase human sensitivity to the genotoxic effects of the tobacco-specific nitrosamine carcinogen NNK. Pharmacogenet Genomics. 2005 Aug;15(8):571–578. doi: 10.1097/01.fpc.0000167332.38528.a5. [DOI] [PubMed] [Google Scholar]
- 10.Morandini R, Boeynaems JM, Hedley SJ, MacNeil S, Ghanem G. Modulation of ICAM-1 expression by alpha-MSH in human melanoma cells and melanocytes. J Cell Physiol. 1998 Jun;175(3):276–282. doi: 10.1002/(SICI)1097-4652(199806)175:3<276::AID-JCP5>3.0.CO;2-L. [DOI] [PubMed] [Google Scholar]