Association of the 609 C/T NAD(P)H: quinone oxidoreductase (NQO1) polymorphism with development of cutaneous malignant melanoma

Introduction

Cutaneous Malignant Melanoma (CMM) is a life threatening disease whose incidence and mortality rates have risen rapidly in the White Caucasian population in recent decades. The consistant stimulation of cutaneous skin cells ‘melanocytes’ by weapons such as Biological mass destructions (WMD) which cause the change of skin colour to brown and hypertrophy, it calls cutaneous malignant melanoma (CMM) in some countries such as UK. However in some countries like Iran it calls Acne, it is removable by surgery and it is non lethal. It can be seen at the skin of face, neck, around the body and feet nail. There is considerable epidemiological evidence supporting a role of exposure to UV radiation as an important environmental factor in the aetiology of CMM. Interestingly intermittent sun exposure early in life rather than cumulative sun exposure appears to be a better indicator of disease risk (Zanetti et al, 1992, Fears et al, 2002). The strong aetiological role for UV exposure is clearly supported by the observation that childhood immigrants to Australia where ambient UV levels are amongst the highest in the world is associated with an increased lifetime risk of CMM. Individuals with light skin colour, freckles and a tendency to burn appear to be especially susceptible (Khlat et al, 1992, Green et al, 1988, reviewed by Jhappen et al, 2003). The observation that only a subset of individuals exposed to sunlight progress to develop CMM indicates that other factors including genetic variation may also play a role in the aetiology of CMM. In support of this, several high penetrance genes have being identified that are associated with risk of familial CMM these include mutations in genes encoding proteins involved in regulation of the cell cycle including CDK4 and the CDKIs p16 and p14ARF (reviewed by Hayword, 2003). However these mutations are rare and are only able to account for a relatively small percentage of cases of CMM.

UVB (290-320nm) is strongly absorbed by DNA and is known to induce a number of mutagenic DNA lesions including cyclobutane pyrimidine dimers and pyrimidine (6-4) photoproducts that are substrates for the nucleotide excision repair (NER) pathway in cells. In contrast the ability of UVA (320-400nm) to damage DNA has being largely ascribed to the generation of intracellular reactive oxygen species (ROS). For example UVA has being shown induce single stranded DNA breaks as well as 8-oxo 7,8-dihydro-2’-deoxyguanosine (8-oxo dG) (Besaratinia et al 2004, Cadet et al, 1997, Cooke et al, 2000, Douki et al, 1999, Kielbassa et al, 1997, Kvam et al, 1997, Zhang et al, 1997). These DNA lesions are repaired predominantly by the base excision repair pathway (BER).

Polymorphic genes involved in the NER and BER pathways and in the detoxification of ROS are therefore attractive candidates for genetic markers of individual susceptibility to CMM. Two key candidate genes in the BER pathway are OGG1, the major DNA glycosylase involved in the removal of 8-oxo dG from genomic DNA, and XRCC1. Interestingly, several studies have provided evidence that the cys326 variant of OGG1 has a reduced ability to restore repair activity in E.Coli deficient in the repair of 8-oxo dG (Kohno et al, 1998). In addition, the catalytic efficiency (k_cat/k_M) of excision of 8-oxo dG from γ-irradiated DNA by ser326 OGG1 protein is twice that of cys326 OGG1 (Audebert et al, 2000, Dherin et al, 1999). Moreover cys/cys genotype has being found to be a repair deficient phenotype (Chen et al, 2003, Aka et al, 2004, Lee et al, 2004). Although the precise role of XRCC1 remains unclear it is thought to play a structural role in the organisation of other BER repair factors (Thompson and West 2000). Recent studies have shown that the arg399gln polymorphism of XRCC1 is associated with reduced DNA repair capacity as assessed by the persistence of DNA adducts (Lunn et al, 1999) and strand breaks (Aka et al, 2004).

In addition, many of the enzymes believed to be involved in the detoxification of UVA-induced ROS (including GSTT1, GSTM1 and NQO1) are also polymorphic in the human population and as such are also attractive candidates for influencing individual cancer susceptibility. GSTM1 and GSTT1 null genotypes have been consistently associated with individual susceptibility to basal cell carcinoma (BCC) (reviewed by Lear et al, 2000) but their role in the aetiology of CMM remains far less clear. NQO1 is an obligate two-electron reductase involved in detoxification reactions converting quinones to hydroquinines (Ross et al, 2000) preventing single-electron redox cycling reactions. In addition NQO1 is believed to play a direct role in antioxidant defense by catalysing the reduction of α-tocopherolquinone to the antioxidant α-tocopherolhydroquinone (Ross et al, 2000). Interestingly the NQO1*0 allele which codes for an essentially non-functional protein (Traver et al, 1997) has also been identified as a candidate loci that is associated with individual susceptibility to BCC (Clairmont et al, 1999). However any possible role for this NQO1 polymorphism in susceptibility to CMM remains unexplored.

The aim of the current study was to investigate the possible association of a range of polymorphic genes involved in BER and the detoxification of ROS and individual susceptibility to CMM.

Materials and Methods

Results

Phenotype, allele and genotype frequencies were calculated for the eleven single nucleotide polymorphisms (SNPs) examined (XRCC1 Arg¹⁹⁴Trp, XRCC1 Arg³⁹⁹Gln, XRCC3 Thr²⁴¹Met, XRCC3 IVS5-14, ERCC2 Lys⁷⁵¹Gln, NQO1 Pro¹⁸⁷Ser, OGG1 Ser³²⁶Cys, CYP2D6 G → A and NAT2). In the case of NAT2 Ile¹¹⁴Thr, Arg¹⁹⁷Gln and Gly²⁸⁶Glu all represent alleles encoding a slow acetylator protein. These alleles were not distinguished in the analysis and were classified as slow (S) acetylator alleles. The wild type allele encodes a fast acetylator protein and was classified as a fast (F) allele. For analysis of GSTM and GSTT because our PCR test was not able to distinguish homozygous positive individuals from heterozygotes individuals were classified as either GST positive (wild type and heterozygous combined) or GST null. Analysis of the data revealed a statistically significant association with the C/T polymorphism at position 609 of NQO1 and the occurence of CMM. Individuals with a CC genotype had a significantly increased risk of developing CMM [OR=2.92 (1.38-6.1); P = 0.016 (Table 3). Presence of the C allele was also associated with development of CMM [OR=2.35 (1.23-4.5); P = 0.04. The association was further reduced in significance once the heterozygous genotype frequency was included and was not significantly different following application of the Bonferroni correction factor (P = 0.152). No associations with melanoma development were found with any of the other genetic polymorphisms studied (Table 3 and Table 4).

Discussion

The aim of this study was to investigate whether genetic polymorphisms in a sample of genes whose products are involved in DNA repair and detoxification of reactive metabolites were associated with the development of CMM.

We found a statistically significant inverse association between an NQO1 polymorphic variant and risk of development of CMM (Table 3). NQO1 is believed to play a role in antioxidant defence and is able to catalyse the reduction of of α-tocopherolquinone to the antioxidant α-tocopherolhydroquinone (Ross et al 2000). The 609 C→ T (Pro¹⁸⁷Ser) polymorphism was initially characterised in humans by Traver et al (1997). The Ser¹⁸⁷ variant of NQO1 encodes a protein with essentially zero biological activity. This is believed to be related to a greatly enhanced rate of protein degradation of the variant form by the ubiquitin/proteosomal system (Ross et al, 2000). The frequency of the homozygous TT genotype has previously been reported to be approximately 5% in the Caucasian population (Kelsey et al, 1997). In agreement with this we observed a TT genotype frequency of 4.0% in our control population. Previous studies have associated the homozygous TT genotype with an elevated risk of a number of cancers (e.g. Benhamou et al, 2001, Menzel et al, 2004, Sarbia et al, 2003, Sunaga et al, 2002, Wiemels et al, 1999). Interestingly, Clairmont et al, (1999) reported that the NQO1*0 allele was a predictor of number of basal cell carcinomas in a multivariate model of patients of known GSTM1, GSTT1 and CYP2D6 EM genotype. However any possible role for this NQO1 polymorphism in susceptibility to CMM remains unexplored.

In the present study we observed that individuals with a “wild type” CC NQO1 genotype had a statistically significant increased risk of developing CMM [OR=2.92 (1.38-6.1); P = 0.016. Furthermore, presence of the C allele was also associated with development of CMM [OR=2.35 (1.23-4.5); P = 0.04.

Our findings are at first surprising in light of the apparent protective role of functionally active NQO1 protein that has previously been reported. However, NQO1 has also being reported to be greatly up regulated in tumours of the liver, lung, colon and breast (Belinsky et al, 1993). In particular, elevated levels of NQO1 enzyme activity in darkly pigmented congenital naevus cells and cultured melanoma cell lines has also being observed previously (Smit et al, 1999). It is possible that NQO1 activity may confer a possible growth/survival advantage to at least a subset of certain tumours.

Interestingly, Brar et al, (2001) report that constitutive activation of NF-κB is important for proliferation of in the malignant melanoma cell lines CRL1585 and CRL1619. NF-κB activation has also being reported to be important in supporting the survival and proliferation of a number of cancers including carcinomas of the breast, ovary, colon, lung, head and neck and pancreas (Batra et al, 1999, Bours et al, 1994, Duffey et al, 1999, Sovak et al, 1997). NF-κB is constitutively activated in melanoma cell lines (Meyskens et al, 1999) including the Hs294T line, where activation was observed as a result of enhanced degradation of the inhibitory partner inhibitory κB alpha (IκBα) (Shattuck-Brandt & Richmond, 1997). No such activation is observed in normal melanocytes (McNulty et al, 2004, Meyskens et al, 1999) In the study by Brar et al (2001) NF-κB activation and cell growth was inhibited by antioxidants suggesting an intermediary role of ROS. In addition, an inhibitor of NQO1 (dicumarol) inhibited NF-κB activation and cellular growth indicating a possible role of NQO1 in ROS generation in these cell lines. Based on results using a quinone analogue (capsaicin) the authors speculated that the source of ROS may involve a NQO1/quinone redox couple. Elevated levels of ROS has also been observed in dysplastic naevi compared to normal skin melanocytes (Pavel et al, 2004). Dysplastic naevi are widely considered to be melanoma precursors further supporting the hypothesis that aberrant redox homeostasis is important in melanoma development.

It must be noted however, that other sources of ROS are also considered to contribute to NF-κB activation in melanoma cell lines. These include NAD(P)H oxidase activity (Brar et al 2002) and gamma-glutamyl transferase (GGT) (Dominici et al 2003, Pieri et al, 2003). Indeed the inhibitory effects of dicurmarol on NFκB activation in melanoma cells appears to be at least in part, to be due to inhibition of NAD(P)H oxidase (Brar et al 2002). Morre et al, (1995, 1996) have also shown a similar inhibition of growth of human and mouse tumour lines including those derived from melanoma by capsaicin but this was considered to be related to inhibition of NAD(P)H oxidase activity. NQO1 may nevertheless contribute to the survival and proliferation advantage as discussed above. A pro-oxidant role of melanin has being proposed in the pathogenesis of melanoma (reviewed by Meyskens et al 2001) and it is possible that NQO1 could also generate ROS by promoting redox cycling of melanin.

Together these studies indicate that intracellular ROS from multiple sources contribute to melanoma cell survival and growth. We hypothesise that under certain circumstances the wild type NQO1 gene product contributes to the promotion of growth of cutaneous malignant melanomas in vivo either by the mechanism proposed by Brar et al (2001) or by promoting redox cycling of melanin. This biologically plausible mechanism may also explain our observation that the wild type NQO1 C allele and NQO1 CC genotype is significantly associated with melanoma development.

This study used a candidate gene approach to analyse genetic factors involved in disease susceptibility. A major potential problem in this approach is the possibility of finding associations by chance. We have addressed this by applying the conservative Bonferroni correction for multiple comparisons and by only considering an association significant if it maintained a P < 0.05 following correction. Although the data presented in this study indicate a role for NQO1 in individual susceptibility to CMM further work is required with larger sample numbers to confirm this association. In addition, work to investigate levels of ROS in NQO1 wild type and mutant melanoma cells is planned.

In conclusion, this study has identified a genetic factor that may influence an individual’s susceptibility to CMM. The data presented also provide further insight into the possible biological role of NQO1 in CMM tumour progression. Antioxidants have been shown to inhibit both growth in both melanoma cell lines and tumours in animal models (e.g. Calagirone et al, 2000, Tanaka et al, 2000, Liu et al, 2001, Malafa et al, 2002 Kogure et al, 2003, reviewed by Sander et al, 2004). Therefore, NQO1 wild type individuals and those with elevated NAD(P)H oxidase may represent prime candidates for the future testing of novel therapeutic strategies involving the use of antioxidants for interruption of oxidant signalling pathways involved in melanoma growth.