Introduction

Multiple lines of evidence indicate that the development of the cancers is the results of both cumulative effect of environmental risk factors and genetic susceptibility of the individual [1]. The environment–gene interaction in carcinogenesis is also well reflected by metabolic enzymes involved in the inactivation and/or detoxification of environmental carcinogens. Most of the carcinogens are metabolically inactivated by detoxification enzymes. Therefore, inherited variations in genes encoding the carcinogen-metabolizing enzymes may alter enzymatic activity and subsequently the carcinogens activation and/or deactivation [2]. Individual susceptibility to cancer is likely to be affected by the genotypes of biotransformation enzymes which represent significant ethnic differences in the frequency of alleles [3].

Cytochrome P450s (CYP450 or CYPs) are phase I biotransformation enzymes and these enzymes play a key role in the metabolism of drugs and environmental chemicals including food contaminants. Several CYP enzymes metabolically activate precarcinogens to genotoxic intermediates [4]. It has been hypothesized that polymorphism in the CYP genes could affect the individual’s capacity in converting precarcinogens into carcinogens and this be one of the important contributing factor to individual’s susceptibility to cancer development [5]. CYP1A1 participates in the metabolism of a vast number of xenobiotics, among which hydroxylation at a vacant position of an aromatic ring is considered to be the hallmark of the initiation of carcinogenesis; it has been reported that the formation of highly reactive intermediates can cause oncogenic mutations in experimental animals and humans [6]. It was revealed experimentally that the polymorphism of Msp I restriction site owing to a T–C variation in the 3′ non-coding region of CYP1A1 allele is associated with increased catalytic activity and increased DNA adducts in cord blood and placenta of newborns [7]. CYP2D6 is involved in the metabolism of various therapeutic drugs and is a possible susceptibility factor for certain environmental reagent-induced diseases [8]. Ethnicity is an important contributing factor in the occurrence of CYP2D6 variability. In African-Americans, the frequency of CYP2D6 with poor metabolize activity is greater than that in the white populations. The occurrence of CYP2D6 with ultra rapid metabolizers appears to be greater amongst Middle Eastern and North African populations [9, 10]. Among the variants, the detrimental mutation of the CYP2D6Ch gene is C188-T, causing a Pro34-Ser amino acid substitution in the highly conserved region. This substitution results in lower CYP2D6 stability, and thus, decreased enzymatic activity [11]. Johansson [12] demonstrated that the lower capacity to metabolize the carcinogens in Chinese was well correlated with a higher frequency of the C188-T mutation in CYP2D6 gene.

Glutathione S-transferases (GSTs) are phase II biotransformation enzymes, which catalyze the conjugation of reduced glutathione via a sulfhydryl group to electrophilic centers on a wide variety of substrates [13]. This activity is responsible for detoxifying both xenobiotics and endogenously generated compounds such as peroxidized lipids. Therefore, GSTs play a pivotal role in the elimination of toxic chemicals [14]. The GSTM1 and GSTT1 enzymes have been shown to alleviate cellular damage from cytotoxic substances in vitro [15]. Individuals with total or partial deletion of GSTM1 and/or GSTT1 showed reduced ability in detoxifying electrophilic carcinogens [16]. The GSTM1 or GSTT1 null, which results from the entire GSTM1 or GSTT1 gene deletion mutation, results in decreased capability in detoxifying some of the environmental carcinogens. It has been observed that GSTM1 null may affect individual susceptibility to cancer [17]. GSTT1 null has also been proved to be a risk factor for individual susceptibility to some cancers [7, 18].

Alcohol consumption, tobacco smoking, as well as some dietary factors believed to be related to the development of gastric carcinoma. It is conceivable that polymorphisms of the metabolic enzymes responsible in metabolizing dietary or smoking carcinogens are associated with gastric carcinoma [19]. The reported mechanisms of carcinogenic role of alcohol are production of acetaldehyde, which is a weak mutagen and carcinogen; induction of CYP2E1 and associated oxidative stress and conversion of procarcinogens to carcinogens; depletion of S-adenosylmethionine and consequently, induction of global DNA hypomethylation; induction of increased production of inhibitory guanine nucleotide regulatory proteins and components of extracellular signal-regulated kinase-mitogen-activated protein kinase signaling etc. [20]. Nicotine (major component of tobacco smoking) may promote carcinogenesis through activation of extracellular signal-regulated kinase/cyclooxygenase-2/vascular endothelial growth factor signaling pathway [20].

Therefore, numerous researches about the relationship between the polymorphisms of metabolic enzymes and cancer susceptibility have been conducted. However, the findings are controversial due to different reasons including the populations selected and their ethnicities. We were particularly interested in the risk factors of gastric carcinoma in the Han ethnic population in Changsha City of Hunan Province of China. A molecular epidemiological investigation aimed to figure out the possible relationship between the genetic polymorphisms of the genes, CYP1A1, CYP2D6, GSTM1, and GSTT1, and the susceptibility of gastric cancer was conducted. The genotype analyses of the two-phase metabolic enzymes would provide a rational strategy for the risk estimation and prevention of gastric carcinoma in the Han ethnic population of Central South China.

Materials and methods

Grouping of participants

This case-controlled study included 123 patients with gastric carcinoma and 129 healthy controls. The healthy controls in this study were defined as people who are cancer- and hematological disease-free. Both the patients and controls were selected from the Han ethnic group born and lived in Changsha City of Hunan Province, P. R. China. The 123 patients were all pathologically confirmed at Xiangya Hospital, of them, 51 were women and 72 were men with an average age of 55.2 years old. Among 123 patients, 35 were with high/high–middle differentiation (low grade tumor), 22 with middle differentiation (intermediate grade), 17 with middle–low differentiation (high grade), and 49 were with low differentiation (highest grade) tumor type. The National Cancer Institute Fact sheet was followed for differentiation typing [http://www.cancer.gov/images/documents/21b8921a-b275-46b4-8dde-957b7218f2ab/Fs5_9.pdf]. The 129 controls were recruited from citizens taking part in a health screening survey carried out at Xiangya School of Medicine, of them, 49 were women and 80 were men with an average age of 53.7 years old.

Genomic DNA isolation from peripheral blood cells

Heparinized peripheral blood samples were collected and the genomic DNAs were purified using standard sodium dodecyl sulfate-proteinase K treatment and phenol–chloroform extraction [21]. The purified DNAs were dissolved in tris–ethylenediaminetetraacetic acid buffer (TE buffer, pH 7.4) and the concentration of DNA was estimated by absorbance at 260 nm and adjusted to 100 ng/ml. The quality of DNA was assessed by agarose gel electrophoresis.

Amplification of CYP1A1, CYP2D6, GSTM1, and GSTT1 genes

The allele sequences of CYP1A1, CYP2D6, GSTM1, and GSTT1 polymorphic variants were obtained from Genbank. PCR primers (Table 1) for various alleles of each variant mentioned above were designed using software Primer Premier 5. PCR was carried out in a total volume of 50 ml containing 100 ng of DNA template, 10 pmol/L of each forward (F) and reverse (R) primer, 200 mmol/L of each deoxyribonucleotide triphosphate and 2 units of Taq polymerase (Roche Basel, Switzerland). The reaction conditions for CYP1A1 and GSTT1 variant allele amplification were as follows. The mixture was incubated for 5 min at 94°C to denature the template DNA; and then the PCR reactions were carried out for 30 cycles by denaturing at 94°C for 30 s, annealing for 30 s at 58°C, and extension for 30 s at 72°C, and a final extension period of 5 min at 72°C was performed after the PCR cycles finished. For analysis of CYP2D6 and GSTM1 genes, annealing temperature was set at 62°C and cycle number was 32. Five microliters of PCR products were loaded on a 1.5% agarose gel with 0.5 mg/ml of ethidium bromide. The DNA bands were visualized and photographed under UV light after electrophoresis.

Table 1 PCR primers for amplification of CYP1A1, CYP2D6, GSTM1, and GSTT1 genes
Full size table

Mutation analysis of CYP1A1 and CYP2D6 genes

To detect the mutation in CYP1A1 and CYP2D6 alleles, the related PCR product was digested with Msp I or Kpn I and the restriction fragment length polymorphism (RFLP) was conducted. Ten microliters of PCR product and 10 units of restriction enzyme were used for each digestion reaction. All the reactions were carried out overnight at 37°C. All digested products were separated on 2% or 3% agarose gel with 0.5 mg/ml of ethidium bromide. For PCR-RFLP analysis, the corresponding sequenced PCR products were used as controls.

Mutation analysis of GSTM1 and GSTT1 genes

The absence of GSTM1- or GSTT1-specific fragments indicated the corresponding null genotype, which was confirmed by using β-actin-specific primers to amplify the specific fragment as internal control. To further verify the gene deletion, a second PCR was performed using another pair of PCR primers in extension for the same gene (Table 1).

Statistics analysis

Genotype frequencies were reported as percentages. The difference in frequencies between groups was compared by using χ 2 test. A P value of 0.05 was considered statistically significant. The correlation between genetic polymorphisms of the studied metabolic enzymes and risk of gastric carcinoma was assessed by means of the odds ratio (OR) with 95% confidence intervals (95% CI) calculated by unconditional multivariate logistic regression. All computations were calculated by using the statistical package of SPSS (SPSS Inc, Chicago, Illinois, USA) version 13.

Results

The demographic characteristics of the gastric carcinoma cases and healthy controls are shown in Table 2.

Table 2 Sex and age distribution in gastric carcinoma cases and controls
Full size table

The PCR products and the Msp I (C▼CGG) restriction map showed three genotypes of CYP1A1 alleles which were wild type (W) without Msp I restriction site, heterozygous mutant (W/M) with Msp I restriction site in one allele and homozygous mutant (M/M) with Msp I restriction sites in both alleles, respectively. Furthermore, PCR-RFLP analysis by digesting the PCR products of CYP2D6 alleles with Kpn I (GGTAC▼C) revealed that there were also three genotypes in this gene, wild type (W) with Kpn I restriction sites in both alleles, heterozygous mutant (W/M) with Kpn I restriction site in one of the alleles and homozygous mutant (M/M) without Kpn I restriction site (Fig. 1). GSTM1 and GSTT1 genotypes were analyzed by PCR amplification as described in materials and methods. The absence of GSTM1 or GSTT1-specific fragments indicated the deletion mutation (Fig. 2). The β-actin gene was used as an internal control to confirm the presence of amplifiable DNA template in the reaction mixture.

Fig. 1
figure 1

Agarose gel electrophoresis of PCR products from a CYP1A1 alleles and the Msp I restriction fragments [1 wild type, 2 heterozygous mutant, 3 homozygous mutant, 4 PCR products without Msp I digestion]. b CYP2D6 alleles and the Kpn I restriction fragments [1 wild type, 2 heterozygous mutant, 3 homozygous mutant, 4 PCR products without Kpn I digestion]

Full size image
Fig. 2
figure 2

Agarose gel electrophoresis of PCR products from a GSTM1 alleles and β-actin gene [M 100 bp DNA ladder marker, 1 GSTM1-F1/R1, 2 GSTM1-F1/R1 (deletion), 3 GSTM1-F2/R2/β-actin, 4 GSTM1-F2/R2 (deletion)/β-actin]. b GSTT1 alleles and β-actin gene [M 100 bp DNA ladder marker, 1 GSTT1-F1/R1, 2 GSTT1-F1/R1 (deletion), 3 GSTT1-F2/R2/β-actin, 4 GSTT1-F2/R2 (deletion)/β-actin]

Full size image

There was no significant difference among the frequencies of CYP1A1 and/or CYP2D6 gene’s wild type (W), heterozygous mutant (W/M) and homozygous mutant (M/M) between the gastric carcinoma group and healthy control group. The difference in the frequencies of GSTM1 null (GSTM1−/−) was significant (P = 0.001), and there were significant differences in the frequencies of GSTM1 null (GSTM1−/−) in four kinds of clinical adenocarcinoma pathology types: high/high–middle differentiated tumor (77.1%, P = 0.030), middle differentiated tumor (81.8%, P = 0.034), middle–low differentiated tumor (70.6%, P = 0.019) and low differentiated tumor (73.5%, P = 0.039), separately. The difference in GSTT1 null (GSTT1−/−) was also significant (P = 0.039), but there were no significant differences (P > 0.05) in the frequencies of GSTT1 null in the mentioned four kinds of clinical adenocarcinoma pathology types: high/high–middle differentiated tumor (60.0%, P = 0.328), middle differentiated tumor (68.2%, P = 0.149), middle–low differentiated tumor (64.7%, P = 0.333), and low differentiated tumor (61.2%, P = 0.191), respectively.

Excluding the influence from other factors, GSTM1 null showed an increased risk of middle–low differentiated gastric carcinoma (OR = 2.532, 95% C.I.: 1.738 − 6.692) and low differentiated gastric tumor (OR = 1.960, 95% C.I.: 1.911 − 4.220). However, GSTT1 null showed no influence to development of the four pathological types of gastric carcinoma mentioned above and the influence of homozygous or heterozygous mutation of CYP1A1 Msp I site and/or CYP2D6 Kpn I site was not significant on any type of differentiated gastric tumors (Table 3).

Table 3 Frequency of CYP1A1, CYP2D6, GSTM1, and GSTT1 genotypes and influence of the genotypes to the risk of gastric carcinomas
Full size table

Discussion

Biotransformation of toxic xenobiotics to their nontoxic forms is considered as the most important step in the processing of the potentially toxic chemical substances toward their elimination [7]. Xenobiotics are metabolized typically in a two-phase reaction: activation and deactivation. In phase I, carcinogens are activated mainly by the CYP450 enzyme family and in phase II, usually under the catalysis of conjugation enzymes, like GSTs, the activated carcinogen is processed into more hydrophilic metabolites which are much easier for excretion [22]. The conjugation enzymes such as GSTs usually play important roles in the latter phase. Genetic polymorphisms of metabolic enzymes resulted from either partial or complete deficiencies significantly affect their catalytic activity and subsequently the levels of activated carcinogens. On the other hand, these activated carcinogens could affect the expression and function of the genes in DNA-repairing, oncogenes and/or tumor suppressor genes. Altogether, these effects could lead to cell damage and tumor genesis by affecting cellular growth, differentiation and apoptosis [23]. Thus, inter-individual variability in xenobiotic metabolism has been associated with different susceptibility to toxicity in response to a given environmental factor. Individuals with certain genotypes that evoke increased metabolic activation of carcinogens and decreased detoxification are more susceptible to environmental carcinogens and have higher risk of cancer development.

Genetic predisposition of CYP enzymes is known to be associated with gastro-esophageal cancer [24], however the association is controversial. There were even reports showed that there was no significant association between the polymorphic expression of CYP1A1, CYP2D6, and other CYP enzymes and gastric carcinoma risk [4, 25]. But significant association between CYP1A1 polymorphism and gastric cancer risk was found in European population [26]. Polymorphism in CYP1A1 was also not found to be associated with gastric carcinoma patients in USA [27] and Japan [28]. Possible explanations to the controversial results include sample size and detecting method, ethnicity, and random scatter around the null hypothesis. In our study, we did not find any significant relationship among the genetic polymorphisms in CYP1A1 and/or CYP2D6 between gastric carcinoma patients and healthy controls from the central southern region of China. It may be that the polymorphisms of these two genes are not susceptible to gastric carcinoma risk because of some specific genetic background or environmental influences. Further investigations are required to clarify the exact cause.

GST enzymes are known to metabolize tobacco-related carcinogens. In this study, we report a clear and significant association between GSTM1 null (deletion) genotype and all types of differentiated gastric carcinoma. Genetic polymorphisms of GSTM1 and/or GSTT1, especially their null genotypes have been reported consistently to be involved in different cancers. It may be that the polymorphisms of these two genes are susceptible to gastric carcinoma risk because of some specific genetic background or environmental influences. Further investigations are required to clarify the exact cause. The involvement of GSTM1 and/or GSTT1 null genotypes in esophageal squamous cell carcinoma and gastric cardiac carcinoma among Caucasian population has been reported previously [29]. But there was no significant difference in the GSTM1, GSTT1, and GSTP1 polymorphisms between the gastric carcinoma cases and healthy controls from Korean population [30]. Also, Wideroff [27] found no evidence that the risk of adenocarcinomas of the esophagus or gastric cardia or noncardia stomach varies with GSTP1, GSTM1, or GSTT1 genotype among USA population. Like CYP polymorphisms, possible explanations for controversial findings include sample size and detecting method, ethnicity, random scatter around the null hypothesis and different lifestyles.

In this case control study, we have reported a clear pattern of involvement of phase I and phase II metabolic enzymes in gastric carcinoma in Han ethnic population in Hunan, China. However, there are some limitations in our study. As a pilot project, the sample size was limited. Although most of the patients and controls are mainly selected from Changsha city in the province, the socioeconomic levels of this city is relatively high in Hunan Province. Therefore, the socioeconomic effects may not be very well reflected in our study. Investigation in the genetic polymorphism of metabolic enzymes and gastric carcinoma should be further extended with a bigger and more reprehensive population in future studies.