Multi-susceptibility genes associated with the risk of the development stages of esophageal squamous cell cancer in Feicheng County

Background The purpose of this study was to evaluate the association of multi-genotype polymorphisms with the stepwise progression of esophageal squamous cell cancer (ESCC) and the possibility of predicting those at higher risk. Methods A total of 1,004 subjects were recruited from Feicheng County, China, between Jan. 2004 and Dec. 2007 and examined by endoscopy for esophageal lesions. These subjects included 270 patients with basal cell hyperplasia (BCH), 262 patients with esophageal squamous cell dysplasia (ESCD), 226 patients with ESCC, and 246 controls with Lugol-voiding area but diagnosed as having normal esophageal squamous epithelial cells by histopathology. The genotypes for CYP2E1 G1259C, hOGG1 C326G, MTHFR C677T, MPO G463A, and ALDH2 allele genes were identified in blood samples collected from all participants. Results The alleles ALDH2 and MTHFR C677T were critical for determining individual susceptibility to esophageal cancer. Compared to the ALDH 1*1 genotype, the ALDH 2*2 genotype was significantly associated with increased risks of BCH, ESCD, and ESCC. However, the TT genotype of MTHFR C677T only increased the risk of ESCC. Further analysis revealed that the combination of the high-risk genotypes 2*2/1*2 of ALDH 2 and TT/TC of MTHFR C677T increased the risk of BCH by 4.0 fold, of ESCD by 3.7 fold, and ESSC by 8.72 fold. The generalized odds ratio (ORG) of the two combined genotypes was 1.83 (95%CI: 1.55-2.16), indicating a strong genetic association with the risk of carcinogenic progression in the esophagus. Conclusions The study demonstrated that the genotypes ALDH2*2 and MTHFR 677TT conferred elevated risk for developing esophageal carcinoma and that the two susceptibility genotypes combined to synergistically increase the risk.


Background
Esophageal cancer is the fourth most common cause of cancer-related death in China. Esophageal cell carcinoma (ESCC) is by far the most common subtype of esophageal cancer, followed distantly by adenocarcinoma, which accounts for less then 3% of all esophageal cancers in high incidence areas of China [1].
The pathogenesis of ESCC is thought to include a stepwise progression from basal cell hyperplasia (BCH) to low-grade dysplasia (LGD), high-grade dysplasia (HGD), carcinoma in situ, and finally invasive carcinoma. Japanese and Chinese pathologists prefer to classify LGD as mild-dysplasia or moderate-dysplasia, and HGD as severe-dysplasia. They defined esophageal dysplasia as a precancerous lesion of ESCC [2][3][4][5]. Epidemiological studies indicated that esophageal dysplasia was associated with a significantly increased risk of developing invasive ESCC [6][7][8][9][10].
Endoscopic screening with the Lugol dye method combined with pathologic evaluation has proven useful in screening for early esophageal cancer and in ascertaining the different stages of esophageal carcinogenesis in high-incidence areas [11]. This test is relatively expensive, however, and patients may experience pain and discomfort, so only a small percentage of the population may be willing to participate in this testing program. With the development of molecular techniques, however, it is now possible to employ useful biomarkers to identify subjects at high-risk who should receive this pathological screening test.
In the present study, we developed a program for evaluation of esophageal lesions using endoscopic screening with the Lugol dye method. The screening ran from Jan. 2005 to Dec. 2007 in Feicheng County, China. In addition to endoscopic evaluation for various types of esophageal lesions, the multi-genotype polymorphisms of CYP2E1 G1259C, hOGG1 C326G, MTHFR C677T, MPO G463A, and ALDH2 genes were identified for each individual subject. Therefore, this program provided a valuable opportunity to first examine the potential association between a specific combination of genotypes and the carcinogenic progression during ESCC, and then to evaluate the possibility of using the genotype combination as a biomarker to predict ESCC risk.

Study subjects
The human subject protocol for this study was approved by the Ethics Committee of the Medical Faculty of the Shandong Academy of Medical Sciences. Written informed consent was obtained from all participants.
For subject recruitment, a questionnaire interview was first conducted to collect information, such as sociodemographic characteristics, alcohol intake, tobacco use, and family history of esophageal carcinoma. Then, a routine physical examination, electrocardiogram, and abdomen ultrasound were provided to all candidates. Those who had liver diseases, cardiovascular diseases, lung diseases, or head-and-neck diseases were excluded for further consideration. All candidate subjects were given an endoscopic staining examination with 1.2% iodine solution for evaluation of esophageal lesions. Furthermore, for persons with a non-staining area of the mucosa, random 4-quadrant biopsy specimens were obtained at 2-cm intervals.
Specimens were processed by the standard procedure [5] and histopathological diagnoses were made by two independent pathologists.
A total of 10125 persons took part in the endoscopic staining examination. Of these patients, 1364 had a nonstaining area of the mucosa and were diagnosed by histopathologic analysis of biopsy tissue. These patients included 280 with basal cell hyperplasia (BCH), 577 with esophageal squamous cell dysplasia (ESCD), 94 with esophageal squamous cell cancer (ESCC) at early stage, and 267 controls with normal esophageal squamous epithelial cells. We randomly selected 50% of the ESCD for further study using the program SPSS. Another 146 patients with ESCC were referred from the Hospital of Feicheng County. In total, 1004 subjects with pathological data were available for the analysis, including 270 patients with BCH, 262 patients with ESCD, 226 patients with ESCC, and 246 controls with Lugol-voiding area but diagnosed as normal by histopathology of esophageal squamous epithelial cells. For genotype assays, 5-8 ml of blood were collected from each participant in a sterile cryogenic vial and stored at -70 ºC until processed.
The sample size was calculated by the Power 3.0 software (http://dceg.cancer.gov/bb/tools/power). We recalculated the sample size based on the following parameters: design: case-control study; case: control = 1:1; probability of esophageal cancer = 0.001 for the aged 40-69; observed minor allele frequency 0.1-0.25; a moderate genetic risk effect (OR = 1.2-1.3); P-value = 0.05; 1-β = 0.8; additive effect model OR = 3.0 of two genes exposure, the sample size = 265. According to the actual sample sizes in the study, the posterior powers (1-β), were 0.735, 0.796, 0.808 and 0.769 for the above four groups of ESCC, ESLD, BCH and controls, respectively.
PCR-RFLP Polymorphism for MTHFR C677T, MPO G463A and CYP2E1 G1259C The PCR reaction was carried out in gradient PCR instrument (Eppendorf, Germany). The reaction mixture contained 25 ng DNA, 10 mM dNTP 0.5 μl, 10 × PCR Buffer 2.5 μl, 10uM of each primer (Table 1)  PCR conditions were 95°C for minutes, followed by 35 cycles of 95°C for 60 seconds optimal annealing temperature ×°C (Table 1) for 60 seconds and 72°C for 60 seconds followed by a final extension step of 72°C for 7 minutes.
The PCR product of the MPO G463A gene digested by an AciI restriction enzyme yielding three possible genotypes which were defined by three distinct banding patterns: A/A 289 and 61bp fragments, A/G 289, 169, 120, and 6lbp fragments, and G/G 169, 120, and 6lbp fragments [20].

PCR for ALDH2
The two pairs of primers shown in Table 1 were used in the PRC assay. F1 and R1 were used to amplify the ALDH2*1 allele (296bp), and F2 and R2 to amplify the ALDH2*2 allele (203bp). Two 25 μL reaction tubes were needed for each specimen to amplify ALDH2*1 (G) and ALDH2*2 (A) respectively, each containing 100 ng DNA, 0.12 mmol/L dNTPs, 12.5 pmol F1 (or R1) primer, 12.5 pmol F2 (or R2) primer, 0.5 U Taq polymerase, and 2.5 μL 10 × PCR buffer (containing 15 mmol/L MgCl2). The reaction tubes were heated to 95°C for 5 min followed by 30 cycles of 95°C for 60 s, 60°C for 60 s, 72°C for 60 s, and 72°C for 45 s, and then followed by a final extension of 5 min at 72°C. 10 μL PCR products were used in agarose gel electrophoresis and the electrophoresis result was photographed.
Two lanes were used for each specimen. If one showed 296 bp band and the other showed no band, the corresponding genotype was ALDH2*1/2*1 (G/G); if one showed 296 bp band and the other showed 203 bp band, the corresponding genotype was ALDH2*1/2*2 (G/A); and if one showed 203 bp band and the other showed no band, the corresponding genotype was ALDH2*2/2*2 (A/A) [23].

PCR-SSCP analysis of hOGG1 C326G
The PCR product was denatured with formamide at 95°C for 15 min, quenched on ice, and loaded to polyacrylamide gels under several conditions. Visualization was performed with a silver stain kit (Wako, Osaka, Japan) as described [22]. The PCR product of hOGG1 C326G gene is digested by Fnu4HI on the polyacrylamide gels showed C/C genotype is a band at 200bp, G/G genotype is a band at 100bp, and C/G genotype is two bands at 200bp and 100bp.

Quality Control
The genotypes of the DNA samples were identified without knowledge of the case or control status; 5% were randomly selected as a sample set of cases and controls that were genotyped by different investigators, and the reproducibility was 100%. Each PCR procedure was performed with a blank control (without DNA template) and positive and negative controls. Cycle sequencing PCR product was performed to confirm the accuracy of this method of single-nucleotide polymorphisms (SNP) identification. When any of these controls failed, the PCR was repeated for the batch of samples.

Statistical Analysis
Pearson's Chi-Square and Kruskal-Wallis H tests were used to examine differences in sociodemographic characteristics, alcohol intake, tobacco use, and family history of esophageal cancer among the four diagnostic groups (Control, BCH, ESCD, and ESCC). Smoking index represents the number of cigarettes per day multiplied by the years of smoking. Alcohol drinking index equals the amount of alcohol consumed per month multiplied by drinking years. Allele frequencies were calculated using allele counting tests for Hardy-Weinberg equilibrium and were analyzed by the Chi-square test. Odds ratios (ORs) and 95% confidence interval (95%CI) were calculated in the multinomial Logistic model after adjusting for the variables of age, smoking index, and drinking index.
An additional analysis based on the generalized odds ratio (OR G ) was also performed. The OR G utilizes the complete genotype distribution and provides an estimate of the magnitude of the association between disease status and genotype [24]. The OR G and 95%CI were  calculated using the software "ORGGASMA" (downloaded from http://biomath.med.uth.gr). All other statistical analysis were performed using SPSS (version 15.0), and P <0.05 (two-sided) was accepted as statistically significant.

Characteristics of demographic and selected variables
Demographic characteristics and selected variables are shown in Table 2. All seven variables, including gender, age, school years, income per year, smoking index, alcohol drinking status, and family history of esophageal cancer were significantly different among the four groups. Each variable also has a significant linear by linear association with the carcinogenic stages in transition from normal esophageal mucosa to carcinoma.
Associations of CYP2E1 G1259C, MPO G463A, MTHFR C677T, hOGG1 C326G, and ALDH2 genotypes with BCH, ESCD, and ESCC The frequency distribution of CYP2E1 G1259C, MPO G463A, MTHFR C677T, hOGG1 C326G, and ALDH2 genotypes are shown in Table 3. The Hardy-Weinberg test for the control group showed that the distributions of the five genotypes were in equilibrium.
As shown in Table 4 after adjusting for the potential confounders gender, age, school years, income per year, smoking index, alcohol drinking status, and family history of esophageal cancer, we found that polymorphism of the ALDH2 genotype was associated with BCH, ESCD, and ESCC. Compared to the ALDH 1*1 genotype, the ALHD 2*2 genotype was associated with significantly increased risks of BCH, ESCD, and ESCC (with the adjusted OR = 4.15, 95% CI, 2.33-7.40 for BCH, OR = 4.54, 95% CI 2.32-8.89 for ESCD, and OR = 3.38, 95% CI 1.64-6.95 for ESCC). Furthermore, the TT genotype of MTHFR C677T increased the relative risk in the ESCC group, while the GG genotype of hOGG1 C326G increased the risk in the ESCD group.
Based on the values of OR G , only the MTHFR C677T genotype (OR G = 1.16; 95%CI: 1.00-1.35) and ALDH 2 (OR G = 1.52; 95%CI: 1.30-1.77) genotype showed significant genetic association with the risk of carcinogenic progression of the esophagus.

Combing two-susceptibility genotypes analysis
Subjects with either homozygous or heterozygous variant alleles (2*2 or 1*2) of ALDH 2 had increased risk of developing BCH, ESCD, and ESCC compared to those who had wild type ALDH 2 (Table 4). Furthermore, TT and CT genotypes of MTHFR C677T were found to enhanced susceptibility to ESCC compared to the CC genotype.
The frequencies of the various combinations of the susceptible genotypes of ALDH 2 and MTHFR C677T genes were calculated and analyzed for their associated risks of diseases (Tables 5 and 6). The OR values for the associations of the combined susceptibility genotypes with esophageal lesions was significantly higher than for the individual genotypes (Table 6). For example, after adjusting for the aforementioned seven potential confounders, the combinations of ALDH 2 2*2/1*2 and MTHFR TT/TC genotypes were associated with significantly increased risks for BCH, ESCD, and ESCC compared to patients with the 1*1 ALDH 2 and CC MTHFR C677T genotype. The ORs (95%CI) were 4.03(2.14-7.57) for BCH, 3.70(1.74-7.87) for ESCD, and 8.72(3.24-23.48) for ESCC. Furthermore, the OR G of the two combined genotypes was 1.83(95%CI: 1.55-2.16), indicating a significant genetic association with the risk of carcinogenic progression in the esophagus.

Discussion
Feicheng County has a high incidence of esophageal cancer compared to the rest of China. Worldwide mortality rates have decreased from 75.82 per 100,000 in 1970-1974 to 57.22 per 100,000 in 2000-2004 [25]. In the present study, we demonstrate that specific allelic combination significantly increased the risk for esophageal cancer (by as much as 8-fold). While reproducibility of studies linking genotype to disease risk is often problematic, there are several strengths of this study. First, the subjects in the study were diagnosed by biopsy, so misclassification bias was very low. Our DNA collection method avoided biases that may arise from single-center or multi-center collection. Furthermore, several steps were taken to ensure high quality and repeatability of  results. These included initial DNA sequencing of SNP regions to prove the reliability, blinding of the operator to the case-control status of samples to reduce observer bias, and simultaneous analysis of case and control samples to avoid differential misclassification. Moreover, the allele frequencies reported among normal controls in this study were similar to those reported in previous studies of Chinese subjects. In sum, these controls indicated that our findings have high validity and reliability.
In the present study, we found that the ALDH2 genotype was associated with BCH, ESCD, and ESCC, the main stages of carcinogenic transition in the esophagus. Acetaldehyde is formed by the oxidation of ethanol by alcohol dehydrogenase (ADH), and is eliminated by aldehyde dehydrogenase (ALDH). The ALDH2 gene carries two alleles, ALDH2*1 and ALDH2*2, with different kinetic properties and distinct distributions among ethnicities [23,26]. The ALDH2*2 allele is found at a frequency of only 50% in Orientals, while the ALDH2*1 allele is more predominant in Caucasians [26]. The ALDH2*2 allele codes for an inactive ALDH2, and is closely associated with alcohol related cancers in the upper aerodigestive tract [27][28][29]. The accumulation of acetaldehyde plays a protective role against excessive alcohol consumption as it causes unpleasant reactions, including "Oriental facial flushing" and other symptoms due to alcohol sensitivity, such as headache, nausea, vomiting, tachycardia, hypotension, and sleepiness [30]. Essentially, a person harboring the ALDH2*2 allele may not become a heavy drinker. Indeed, genetic epidemiologic studies have indicated that the ALDH2*2 allele inhibits the development of alcoholism. However, many studies demonstrated that patients harboring ALDH2*2 allele who are heavy drinkers were at increased risk of ESCC. It is unknown why patients harboring the ALDH2*2 allele became heavy drinkers despite the unpleasant reaction to acetaldehyde [28][29][30][31].
The TT genotype of the MTHFR gene had a significantly positive association with ESCC (OR = 1.85, 95% CI 1.02-3.34) but not with BCH or ESCD. There was also a significant association between esophageal cancer and the MTHFR TT genotype with which the patient was also a heavy smoker. Associational studies linking polymorphisms of the MTHFR C677T genotype with ESCC risk have yielded inconsistent result. A meta-analysis of studies examined the association of the MTHFR C667T polymorphism with risk of esophageal cancer [34].
The association between esophageal cancer and MTHFR enzyme activity is most likely related to the metabolism of folic acid, as risk of esophageal cancer depends on the status of folic acid intake. When folic acid intake is sufficient, individuals with the MTHFR CT or TT genotypes may actually have a decreased risk of esophageal lesions because the lower MTHFR activity of the 677TT allele may lead to an elevation in 5, 10methylenetetrahydrofolate, facilitating DNA synthesis. In contrast, both impaired DNA methylation and DNA synthesis/repair may become the primary mechanisms of carcinogenesis in the presence of low folic acid intake [16,[35][36][37]. However, the TT genotype was not related to BCH or ESCD, suggesting a weaker or absent linkage.
We present evidence that two susceptibility genes, ALDH2*2 and MTHFR 677T, contribute to the process of esophageal carcinogenesis. The combination of the two high-risk genotypes 2*2/1*2 of ALDH 2 and TT/TC of MTHFR C677T resulted in a 4-fold higher risk of developing BCH, a 3.7-fold increased risk of ESCD, and a 8.72 times higher ESSC risk. The ALDH 2 and MTHFR C677T genes showed a significant association with ESCC in our population. In contrast to ALDH 2 and MTHFR, polymorphisms of CYP2E1 G1259C, MPO G463A, and hOGG1 C326G genes were not associated with BCH, ESCD, or ESCC risk in this study. These negative results may be attributable to the fact that the study population came from the same community where residents share a similar life style and diet. This homogeneity may cause an overmatch, such that the association of these two metabolic enzyme genes (CYP2E1, MPO) and one repair gene (hOGG1) with lesions of the esophagus cannot be demonstrated or is too low to estimate. In addition, because these alleles were associated with smaller ORs (<2.0) for risk of the diseases, this effects would not be detected due to an allele null frequency less than 0.10. The sample size, therefore, may not have been large enough to detect an association.

Conclusion
The ALDH2*2 and MTHFR 677T alleles were associated with higher susceptibility to esophageal cancer. Compared with the ALDH 1*1 genotype, the ALHD 2*2 genotype was associated with significantly increased risks for BCH, ESCD, and ESCC, while the TT genotype of MTHFR C677T increased the risk of ESCC. The generalized odds ratio analysis confirmed these findings. Further analysis revealed that the combination of these high-risk genotypes (2*2/1*2 of ALDH 2 and TT/TC of MTHFR C677T) significantly increased susceptibility for BCH, ESCD, and ESSC (by 4.0, 3.7 and 8.72 fold, respectively). The OR G of the two genotypes combined was 1.83(95%CI: 1.55-2.16), indicating a significant genetic association between this combined genotype and cancer of the esophagus.