Role of GSTM3 Polymorphism in the Risk of Developing Esophageal Cancer

Nutritional deficiency coupled with alcohol and tobacco abuse (both smoked and smokeless forms), environmental carcinogens, and occupational exposure have been shown to be associated with the development of esophageal cancer (1, 2). Glutathione S-transferases (GST) constitute a family of xenobiotic detoxifying phase II enzymes that catalyze the conjugation of glutathione to a variety of electrophilic compounds, including polyaromatic hydrocarbons, which are widely present in the human environment and are known to be carcinogenic. At least five mammalian GST gene families have been identified to be polymorphic, and mutations or deletions of these genes contribute to the predisposition for several diseases, including cancer. The gene cluster of GSTM1-GSTM5 has been reported to be localized on chromosome 1p13 and spans a length of nearly 100 kb. Polymorphic genes of GST family, like GSTM3 and GSTP1, have been shown to modulate the cancer risk. In the GSTM3 gene, the GSTM3*A wild-type allele and the GSTM3*B variant allele have been described (3). GSTM3 plays a role in the metabolism of harmful agents, like polyaromatic hydrocarbons benzo(a)pyrene, and has overlapping substrate specificity with GSTM1 (4). The GSTM3 polymorphism could, therefore, confer different efficiencies in the metabolism of carcinogens. In a recent study, an increase in larynx cancer risk associated with GSTM3 AA genotype was suggested (5). In the GSTM3 gene, the GSTM3*A wild-type allele and GSTM3*B variant allele have deletion of 3 bp in intron 6, resulting in the generation of a recognition sequence for the YY1 (Ying Yang) transcription factor. As a result of this mutation, the expression of GSTM3 can be influenced. The functional consequence of this is unclear, but both negative and positive regulatory effects have been suggested (6-8). However, possible influence of GSTM3 has not been studied in esophageal cancer. Altered detoxification due to GSTM3 polymorphism alone or in combination with GSTM1 null genotypes might influence cancer susceptibility. Therefore, the present study was designed to analyze the association of GSTM3 genotypes and interaction with GSTM1 null genotypes and environmental factors with risk of developing esophageal cancer.

This study included a total of 149 untreated, histologically confirmed, esophageal cancer patients with squamous or adenocarcinomas, referred to the Departments of Gastroenterology and Radiotherapy of a tertiary referral hospital in Northern India along with 200 nonmalignant unrelated controls, during a 2.5-year period (2003 to mid-2006). The ethnic background of the subjects was similar (i.e., from Northern India). Patient data were collected through an interview where demographic features, clinical details, and environmental exposure were recorded using a standard clinical proforma. Written informed consent was obtained from each patient. Environmental carcinogens included smoking of cigarette, bidi (Indian cigarette made of tobacco wrapped in tendu leaf), or hukka (Indian pipe). Smokeless tobacco use consisted of gutka (sweet flavourful chewing tobacco) or zarda (aromatic chewing tobacco). Smoking was measured in pack-years (1 pack-year = 20 sticks per day for 1 year). Smokeless tobacco use was measured as chewing-year (1 chewing-year = taking chewable tobacco quid once in a day for 1 year). Alcohol was either consumed or not. “Occupational exposure” consisted of being exposed to coal, smoke, or petroleum products at work and included exposure of housewives to smoke emanating from domestic cooking fuels. The study was accorded approval by the ethical committee of the Institute.

Genomic DNA was prepared from frozen blood samples collected in EDTA. DNA was isolated as described elsewhere (9, 10). Genotype frequencies at the GSTM3 and GSTM1 loci were determined by PCR. GSTM1 null polymorphism genotyping was done using method described elsewhere (9). GSTM3 reactions were carried out in a total volume of 25 μL containing 50 ng DNA template, 20 pmol GSTM3 primers (10, 11), 10 mmol/L deoxynucleotide triphosphates, 10× buffer containing Tris-HCl (pH 8.6), 50 mmol/L KCl, 15 mmol/L MgCl₂, and 1.5 units Taq polymerase (Bangalore Genei, Bangalore, India). The reaction conditions for GSTM3 were 94°C for 3 min followed by 35 cycles of denaturation at 94°C for 1 min, annealing at 62°C for 1 min, extension at 72°C for 1 min, and final extension for 10 min. Intron 6 GSTM3 PCR product was separated on 20% polyacrylamide gel (Fig. 1). The band sizes corresponding to 79 bp (GSTM3*A allele) or 76 bp (GSTM3*B allele) were recorded.

The demographic characteristics of patients and controls were described as percentages or summary measures. Statistical significance of frequency differences between patients and control groups was evaluated using the χ² test or t test as appropriate. Deviation from the Hardy-Weinberg equilibrium in controls was assessed using the χ² test, and P was considered significant at <0.05 level. Unconditional logistic regression analysis was used to fit statistical models to predict the association of GSTM3 genotypes with susceptibility to esophageal cancer, histopathology, and site of tumor, lymph node presence, and environmental factors, like tobacco habits, alcohol consumption, and occupational exposure. A case-only analysis was used for estimating gene-environment interaction, and interaction terms were included in the logistic regression model. The likelihood ratio test was used to test the goodness of fit, and the Wald's test was used to assess the coefficient of significance. Association was expressed as odds ratios (OR) as risk estimates with 95% confidence intervals (95% CI). ORs were adjusted for confounding factors, such as sex, age, smoking, and alcohol usage. Statistical analysis was done using SPSS Software version 10.0 (SPSS, Chicago, IL).

The characteristics of the study subjects are listed in Table 1. Mean age ± SD (years) and gender distribution were similar in patients (56.5 ± 12.4 years and 76% males) and controls (55.6 ± 9.4 years and 71% males). The majority of patients harbored a squamous cell carcinoma (94.1%), located in the middle third of the esophagus (52.7%), with nodal metastasis in 67.2%. Prevalence of smoking and smokeless tobacco users was around 80% (Table 1). The proportion of drinkers and occupationally exposed patients is also summarized in Table 1.

The GSTM3 genotype distributions in the control population were in Hardy-Weinberg's genetic equilibrium model in relation to observed and expected genotypic frequencies (P = 0.40). The genotype frequencies are presented in Table 2. The GSTM3 AB + BB genotype was associated with a 2-fold risk of esophageal cancer, due to a higher frequency of the GSTM3 AB genotype and a lower frequency of GSTM3 AA genotype, in cancer patients compared with controls (Table 2). The frequency of the GSTM3 B allele was also higher and imposed significantly risk for esophageal cancer (OR, 2.2; 95% CI, 1.4-3.3; P = 0.00). Effect of the GSTM3 AB + BB genotype was slightly modified by GSTM1 null genotypes (OR, 2.3; 95% CI, 0.97-4.2; P = 0.01).

After gender stratification according to GSTM3 genotypes, risk for esophageal cancer increased in males (Table 2). Analysis of clinical characteristics in patients with squamous cell carcinoma histology revealed higher GSTM3 AB + BB genotype frequency (21.2%), which increased further in adenocarcinoma (44.4%), compared with controls (Table 3). Association of lymph node metastasis with GSTM3 AA + BB genotypes showed risk in patients with absence of spread to lymph nodes (Table 3). GSTM3 AB + BB genotype frequency and risk for esophageal cancer was higher in patients with middle and lower third location (OR, 2.2; 95% CI, 1.1-4.4; P = 0.01; OR, 2.6; 95% CI, 1.2-5.6; P = 0.01). Interaction of GSTM3 AA + AB genotypes with tobacco habits (OR, 1.0; 95% CI, 0.38-2.8; P = 0.91), smokers (OR, 1.3; 95% CI, 0.59-2.9; P = 0.48), smokeless tobacco users (OR, 0.92; 95% CI, 0.48-1.7; P = 0.80), alcohol consumption (OR, 0.55; 95% CI, 0.21-1.4; P = 0.09), and occupational exposure (OR, 1.7; 95% CI, 1.0-7.3; P = 0.21) did not modulate the risk further.

Esophageal cancer is known to be associated with environmental carcinogens. GSTM3 is a phase II enzyme that plays a role in conjugation and detoxification of environmental and occupational carcinogens. Lack of glutathione conjugation could be an important risk factor for development of esophageal cancer. A 3-bp deletion in GSTM3 intron 6 results in two alleles: GSTM3*A and GSTM3*B. The deletion generates a site for the binding of transcription factor YY1. The frequency of GSTM3 B mutant allele has been rated as 15% to 24% (5) in Caucasians and 68% in African Americans (11). In the present study, frequency of GSTM3 B allele was 11.5% in controls, which was similar to other studies from India (10, 12). Heterozygous carriers of the mutated allele GSTM3*B were more frequent and at risk in the group of esophageal cancer patients. This is similar to the other studies on larynx (5), colorectal (13), bladder (6), and breast cancer (14). In contrast, GSTM3 AA genotype has been a risk factor in oral cancer (12). According to our results, the GSTM3*B allele could be considered a risk allele as it has a YY1 site. There are many promoters regulated by YY1 either by initiation, activation, or repression. The repression of cytokines or stress response genes like IFN-γ by YY1 transcription factor may also enhance cancer susceptibility (15). The GSTM3*B allele has been reported to lead to variable expression in cytosol (16), which may lead to alteration in detoxification efficiency of carcinogens predisposing an individual for cancer susceptibility.

After subgrouping patients and controls based on gender, the GSTM3 AB + BB genotypes were at higher risk for esophageal cancer in males compared with females. The differences may be due to gender-associated expression patterns of the GST family of enzymes or the influence of sex hormones for which GST regulation is established in rodent models (17, 18). Lifestyle and dietary factors may also influence the difference in susceptibility to disease. Dietary influence of potentially protective vegetable diets are known to have a more pronounced effect on GST activity in females compared with males, which may explain lower risk in females (19).

The association of GSTM3 AB + BB genotypes with clinical characteristics, such as histology, showed higher risk in patients with adenocarcinoma compared with those with squamous cell carcinoma, suggesting contribution of altered expression of GSTM3 genotypes in susceptibility for histology specificity in esophageal cancer. The association with the absence of lymph node metastasis and GSTM3 AB + BB genotypes may not reflect a true association as the sample size in the subgroup of patients with no nodal metastasis was small, with a real possibility of a type I error. However, the variable risk for tumor development due to GSTM3 AB + BB genotype with middle and lower third location of tumor seems to support a previous theory of site-specific genetic alterations (20).

A case-only analysis was used to explore gene-environment interaction. Case-only approaches are believed to be better than case-control studies as the latter may suffer from common biases arising from the selection of controls (21, 22). Tobacco carcinogen exposure and alcohol use have strong association with esophageal cancer. The present study did not reveal significant modulation of risk with respect to GSTM3 genotypes in tobacco users (i.e., smokers or smokeless tobacco users) or in alcohol users. No significant associations were observed between GSTM3 genotypes and occupational exposure, thus indicating no potential interaction of GSTM3 variants and environmental carcinogens. This finding is similar to other studies on larynx (5) and oral cancers (23, 24). One common flaw may have been small sample sizes in these studies with low power to detect real differences.

Nakajima et al. have shown influence of GSTM1 on GSTM3 genotype expression levels (16). GSTM1 null homozygotes expressed less GSTM3 protein than individuals with other GSTM1 genotypes. These findings indicate that the effect of GSTM1 on susceptibility may be influenced by expression of GSTM3. In the present study, only marginal increase in risk due to GSTM1 null genotype was observed. We had earlier shown GSTM1 null genotype to be a risk factor for developing adenocarcinoma (9).

In summary, this study suggests that GSTM3 AB + BB genotypes apparently play a significant role in mediating susceptibility for developing esophageal cancer. The risk is slightly modulated due to presence of GSTM1 null genotype. Furthermore, GSTM3 polymorphism also seems to modulate the risk for adenocarcinoma histology subtype and in middle and lower third esophageal anatomic regions.

We thank the assistance of Dr. Sandeep Bhattacharya (King George Medical University).