Effect of MDM2 SNP309 and p53 codon 72 polymorphisms on lung cancer risk and survival among non-smoking Chinese women in Singapore

The TP53 tumour suppressor pathway plays a critical role in cell cycle regulation and apoptosis in many cancers, including lung carcinomas [1], and variation in the genes that regulate this pathway may exert an important influence on tumour development, and hence, cancer risk. Recent interest has focused on the murine double minute-2 protein (MDM2), a nuclear phospoprotein that inhibits p53 activity by promoting its degradation [2]. A single nucleotide polymorphism (SNP309) in the MDM2 promoter has been found to influence transcription of this gene via a greater affinity for the SP1 transcription factor, and hence, individuals with the GG genotype have higher MDM2 levels leading to attenuation of the p53 pathway [3, 4]. This has been especially so in the case of females, due to the involvement of the MDM2 SNP in the estrogen receptor signaling pathway [3].

Several epidemiologic studies have evaluated this association with varying results. No overall association between MDM2 SNP309 and lung cancer risk was observed in recent studies in a European [5], North American [6, 7] and two Asian [8, 9] populations. However, the GG genotype was observed to increase risk in a study in China [10], Norway [11] and among non-Hispanic Whites in the US [12]. In some cases, even though there was no overall association noted with the MDM2 SNP309 and cancer susceptibility, substratification of groups led to some association. For example, an increased risk was correlated with the GG genotype only in the adenocarcinomas, but not in the small cell, squamous cell or non-small cell lung cancers (NSCLCs), or when all lung cancer subtypes were grouped [9]. Other studies pointed to an interaction with smoking status - although no association was found between all lung cancers and the MDM2 SNP309, there was an elevated risk of cancer susceptibility in smokers compared to non-smokers in the same study [10]. Conversely, one study found that the risk of lung cancer was higher among individuals with the TT genotype - on stratification, this association was restricted to males and to smokers [12]. A meta-analysis of seven studies concluded that the GG genotype conferred a small but significant increased risk (summary odds ratio [OR] 1.27; 95% confidence interval [CI] 1.12-1.44) [13]. Because of the capacity of the polymorphism to enhance the inhibitory action of MDM2 on the p53 pathway, investigators have hypothesized that the GG genotype would also negatively influence prognosis in lung cancer. To date, two studies have demonstrated that the GG genotype is associated with a lower level of p53 expression [14] and with poorer survival [14, 15].

A polymorphism at codon 72 of the p53 gene resulting in an arginine-to-proline (Arg/Pro respectively) substitution has also been shown to be varyingly associated with cancer predisposition, possibly due to differences in ethnicity and in background risk due to smoking [16‐18]. A meta-analysis of 13 epidemiologic studies which examined this association reported that the summary odds ratio of lung cancer associated with the Pro/Pro genotype was 1.18 (95% CI 0.99-1.41), and was 1.02 (95% CI 0.86-1.20) for Pro-carriers [19]. The Pro/Pro genotype has also been associated with poorer prognosis and less favorable clinical outcome in lung cancer patients [14, 20, 21]. Moreover, significantly higher or a lower prevalence of p53 gene mutations have been noted in lung cancers among carriers of the Pro allele in the Polish [22] or a Norwegian population [23], respectively. These data highlight that ethnicity and other factors may have an impact on p53 functionality, with respect to the SNP and mutational status.

The influence of gender and ethnic differences on genetic determinants of lung cancer is increasingly being recognized. Smoking accounts for the overwhelming majority of cases in Western populations and among men, but lung cancer among non-smokers is now known to be a distinct clinico-pathologic entity that is as yet poorly understood [24, 25]. We therefore sought to examine if MDM2, through its role in the p53 pathway, is associated with risk and prognosis of lung cancer among Singapore Chinese women, a primarily non-smoking population. In addition, as MDM2 SNP309 status has been demonstrated to affect the overall age of onset of various cancers, with differential effects in different populations [4, 26], we have explored its effects in this unique population.

Table 1 provides a description of relevant demographic and other background characteristics of cases and controls. They were similar in respect of age, years of education, and cases were more likely to have been born outside of Singapore than controls. They were also less likely to consume fruit and vegetables, or to be exposed to environmental tobacco smoke at home. Approximately three-quarters of the cases were adenocarcinomas.

A total of 123 (97.6%) cases and 159 (98.1%) controls were successfully genotyped for both polymorphisms. Genotype frequencies were comparable with the published literature and in keeping with the lower prevalence of the T allele reported in Asian, relative to Caucasian, populations [7, 8, 10‐12]. The distribution of genotype frequencies conformed to that expected under the Hardy-Weinberg equilibrium for both MDM2 SNP309 (p = 0.62) and TP53 codon 72 (p = 0.19).

In summary, we show here that the MDM2 SNP309 TT genotype is associated with a higher risk of lung cancer among never smoking women in the Chinese population. The presence of the p53 codon 72 Pro allele did not lead to a statistically significant interaction or materially affect the magnitude of this risk. Neither of the MDM2 nor p53 polymorphisms evaluated showed any impact on survival or age of onset of lung cancer in this population.

Our study population is unique in consisting solely of non-smoking women, who could potentially be informative with regard to risk factors that operate independently of tobacco smoking. The finding that the MDM2 TT genotype increases risk was unexpected, since as described earlier, many studies that have reported associations have found increased risk with the GG genotype that are consistent with the impact of this genotype on MDM2 RNA and protein levels, and hence, on inhibition of p53 [3]. To our knowledge, only one study has examined the combined effect of MDM2 and p53 polymorphisms and lung cancer, and this was also conducted in a Chinese population [10]. The results showed a multiplicative effect, with a 4.5-fold higher risk in those with the GG, Pro/Pro genotypes relative to TT, Arg/Arg. However, our findings suggesting that TT genotype can be a susceptibility factor are not entirely surprising, given the recent observations. Recent work using large numbers of breast cancer samples in the Chinese female population also revealed that the TT genotype predisposes to accelerated onset of cancers [26], and that the GG genotype reduced risk of leukemia [28]. Moreover, one other study has noted association of increased lung cancer risk with the TT genotype in the Chinese population [12], supporting our data. Thus, it is possible that the association of the G or the T alleles with increased cancer risk or onset age may be influenced by ethnicity and by environmental factors unique to that particular population. More in depth larger scale studies are now required to further extend these findings.

How does the TT genotype lead to increased cancer risk is at present not well understood, but there is evidence to suggest that the role of MDM2 in tumorigenesis may vary in a gender-specific manner, and between smokers and never smokers. Thus, it is not inconceivable that the mechanisms by which MDM2 modulates lung cancer risk can differ between populations. The role of hormones in carcinogenesis is rapidly emerging as a complex, but important pathway. Estrogen signaling, in particular the estrogen receptor (ER), is known to play a direct role in MDM2 transcription, regulating its expression [29]. The G allele of SNP309 increases the affinity of the MDM2 promoter for Sp1, a co-transcriptional activator for the estrogen receptor, and was associated with increased MDM2 expression. Consistently, Bond et. al. showed that the effect of the MDM2 polymorphism (GG) was gender-specific and enhanced in women with active estrogen signaling pathways [30]. Therefore, one would envisage that the TT genotype is associated with reduced MDM2 levels. However, MDM2 has also been shown to negatively regulate ER expression [31]. Thus, it is possible that reduced MDM2 levels expected in TT individuals could lead to elevated ER expression, and hence may elevate cancer risk - a hypothesis that requires further investigation. Given that estrogen receptors are expressed in lung tumours, particularly adenocarcinomas, and these cells are responsive to estrogen [32, 33], it is possible that hormonal pathways provide an alternative mechanism by which MDM2 influences lung cancer risk in this population. Alternatively, other MDM2-dependent pathways could be modified in the TT subjects, leading to cancer susceptibility. In this respect, no studies have evaluated if specific signaling pathways are associated with the MDM2 SNPs. Given the emergence of an association of the TT genotype with cancer susceptibility in various studies, such analyses would provide further mechanistic insights.