Background: Tumor protein 53 plays important role in tumor suppress by binding to the regulator of its target genes. Single nucleotide polymorphism located in p53 binding regions is likely to affect the expression of p53 target genes and may contribute to the susceptibility of humans to common diseases. Role of genetic variations in esophageal squamous cell carcinoma (ESCC) has been well explored. However, the role of p53 binding region variations in esophageal cancer is poorly understood.
Methods: We investigated the association of six p53 binding region polymorphisms with susceptibility of 400 ESCC cases and 400 cancer-free controls in a southwest China population using SNapShot method. Differences in frequencies of the genotypes of the SNPs variant between the cases and controls were evaluated by using the chi-square test.
Results: We found that C allele of rs1009316 in Bax and rs762624 in CDKN1A can decrease the risk of ESCC. In the multiple genetic model, rs2395655 in CDKN1A is related with risk of ESCC, G allele increase susceptibility to ESCC (OR: 1.364; 95% CI: 1.104-1.685). We carried out a stratification analysis between allele and risk of ESCC according to clinical stage. There is no relationship between these SNPs and clinical stage.
CONCLUSION: P53 binding region polymorphisms may modulate risk of ESCC in southwest China population.
Introduction
Esophageal cancer is the eighth most common malignancies worldwide and the incidence rate is increasing significantly especially in the developing countries [1, 2]. Patients with ESCC have no obvious clinical features and are always diagnosed at moderate or later stage. Overall survival of this disease is less than 10%, and the 5-year survival rate is 20%-40% after surgery. Therefore, more efforts are needed to look for new markers for early diagnosis of this disease. Esophageal cancer is a complex disease involving genetic and environmental factors, while genetic factors play important roles in ESCC. P53 is the most important tumor suppressor gene, which can prevent cancer by maintaining genomic stability, inducing apoptosis, inhibiting angiogenesis and other ways through regulation of many downstream target genes. P53 as a transcriptional factor can bind to the promoter or intron of its target genes functioning as a cancer suppressor. P53 can promote cell apoptosis through up-regulating Bax and down-regulating Bcl-2. P53 plays important roles in cycle arrest and genome repair through regulating target genes CDKN1A and GADD45A expression.
Single nucleotide polymorphism (SNP) is the most common variation across the human genome, which is the ideal genetic marker in association studies of complex diseases. Candidate gene study and genome-wide association study (GWAS) are the main strategies applied in SNP-based association studies, with their own strengths and weaknesses. Candidate gene study must assume the causative gene, and screening is confined. However, GWAS requires a lot of manpower and resources, and positive SNPs are quite difficult to explain their biological functions. Therefore, how to choose disease-related SNPs for association analysis has become the key issue of complex disease studies.
In this study, we speculate that SNPs lying in the binding regions of p53 may affect its regulation of target genes, and then cause the carriers with different alleles with varying degrees of susceptibility for the tumor. Hence, these SNPs in p53 binding regions are expected to become the novel risk markers for the early diagnosis of ESCC. Our study began with a genome-wide screening of SNPs in p53 binding regions using data mining. Then, we chose the SNPs existing in Chinese Han population and tumor-related genes as target SNPs. Case-control study was applied to evaluate the relevance between these SNPs and susceptibility to ESCC based on the population of Chongqing. At last, we intended to explore the function of positive SNPs and provide evidence for the early diagnosis of ESCC.
Methods
Study sample
All person were genetically unrelated Han Chinese and were living in Chongqing City of southwest China for at least 20 years. The esophageal cancer patients were histopathologically diagnosed and conformed at the Southwest Hospital, the first affiliated hospital of Third Military Medical University. The exclusion criteria included previous cancer, metastasized cancer, and previous radiotherapy or chemotherapy patients. The controls were healthy individuals who participated in a physical examination in Chongqing who had no history of cancer. Each participant was scheduled for an interview after written informed consent, and the interviewers collected the information about the demographic data and environmental exposure history using a questionnaire. After interview, about 5 mL venous blood was collected from each participant.
SNP selection
We began with a genome-wide screening of SNPs in p53 binding regions in UCSC (http://genome.ucsc.edu/) and dbSNP (http://www.ncbi.nlm.nih.gov/projects/SNP/) databases, 12301 SNPs lies in P53 binding sites, 2688 of which exist in Chinese population according to HapMap database and were regarded as SNP pool. According to the Gene Ontology database (http://www.geneontology.org/) and previous studies have confirmed that Bax, CDKN1A, GADD45A genes are regulated by P53 and played a critical role in carcinogenesis. Supplementary figure 1 shows the location of genes, p53 binding regions and SNPs. Then, we focused on SNPs in the three loci and chose SNPs with Minor Allele Frequency (MAF) ≥ 5% in Chinese population based on the HapMap CHB database. The information of these SNPs is in supplementary table1.
Genotyping
We used the SNaPshot assay to genotype the SNPs. The detailed method may be found in our previous studies. PCR and SNaPshot primers are shown in supplementary table 2. The genotyping results are validated by direct sequencing.
Statistical Analysis
The χ2 test for Hardy-Weinberg equilibrium was applied to the SNP among controls (http://ihg.gsf.de/cgi-bin/hw/hwa1.pl). Differences between cases and controls were evaluated by the Student's t test for continuous variables and x2 test for categorical variables. The associations between SNPs and esophageal cancer risk were estimated by computing the odds ratios and 95% confidence intervals (CI) using co-dominant, dominant, recessive, multiple, and over-dominant genetic models. All the statistical analyses were done with Statistical Package for the Social Sciences software (SPSS 16.0). P <0.05 means significantly difference.
Results
Selected characteristics of participants genotyped in this association study are summarized in table 1. In all there were 400 esophageal squamous cell carcinoma cases and 400 cancer-free controls in this study. Patients involved in our study are from Department of Cardiothoracic Surgery, Southwest Hospital, while health controls are from Physical Examination Center in the same hospital. Cases and controls are Han population with long-term living in Chongqing, matching in age (p=0.47), gender (p=0.22) and without kinship. The pathological stages of esophageal cancer cases were 7 (1.75%) stage I (T1N0M0), 335 (83.75%) stage IIa (T2-3N0M0), 37 (9.25%) stage IIb (T1-2N1M0), 13 (3.25%) stage III (T3N1M0 or T4NanyM0) and 8 (2.0%) stage IV (TanyNanyM1).
Table 2 shows the genotype and allele frequency distribution of these SNPs in cases and controls and the results of Hardy-Weinberg equilibrium (HWE) test in controls. HWE has been used to better understand genetic characteristics of populations. We have used Hardy-Weinberg tests for genotyping quality control. Three SNPs (rs3783468, rs681673, rs532446) are deviation from HWE test, while the other three SNPs (rs1009316, rs762624, rs2395655) passed HWE, with p values of 0.340, 0.258, and 0.297, respectively. The SNPs showing departure from Hardy-Weinberg equilibrium are removed in further analysis.
Table 3 summarizes the association of these polymorphisms with ESCC risk. Co-dominant, dominant, recessive, multiple, and over-dominant genetic models were used to evaluate the relationship between SNPs and susceptibility to ESCC. We found that CC genotype and C allele of rs1009316 and rs762624 can decrease the risk of ESCC in all the models except dominant genetic model. While, rs2395655 is related with risk of ESCC in co-dominant, dominant, recessive, and multiple genetic model. In the multiple genetic model, p value is the least (0.004) and compared with A allele, G allele increase susceptibility to ESCC (Odds Ratio: 1.364; 95% Confidence Interval: 1.104-1.685）
To examine whether the genetic variant was modified by epidemiological factors, we performed stratified analyses based on gender and age by logistic regression analysis and we found no significant interactions between this genotype and gender, age, and differentiation. To evaluate the gene-clinical stage interactions, we carried out a stratification analysis between allele and risk of ESCC according to clinical stage. We found no significant interactions between SNPs and ESCC clinical stages (Table 4).
Discussion
Esophageal cancer is a complex disease which results from gene-gene, gene-environment interactions. Genetic factor plays important roles in esophageal cancer genesis and it may be very important for the early diagnosis of this disease. SNP is the most common variation across the human genome, which is the ideal genetic marker in association studies of complex diseases. Recently, though there have been many ESCC association studies, including the GWAS. In our previous study, a SNP in the pre-miR-196a was associated with susceptibility of esophageal squamous cell carcinoma risk in this Chinese Han population and a SNP in RAP1A 3'UTR could affect miR-196a regulation contributing to ESCC risk.
P53 is the most extensively studied gene involved in human cancers including breast cancer, colorectal carcinoma, ovarian carcinoma, head and neck SCC42. P53 can bind to specific DNA sequence, activate its downstream genes expression, and control cell-cycle arrest, DNA repair, and apoptosis8. When cells were stimulated, p53 could promote cell apoptosis through up-regulating Bax expression and down-regulating Bcl-2 expression [43-45]. Furthermore, p53-mediated induction of CDKN1A (which encodes WAF1, also known as p21) might also inhibit apoptosis 31 and may function as Cell-cycle arrest and DNA repair through GADD45A, and these functions of p53 might actually protect cells from death during the early stages of tumor progression46. So, the SNP located in the p53 binding site of its target gene may affect the regulation of p53 and contributing to cancer susceptibility. From a relatively large case-control study provide evidence that SNPs in the promoter regions of BAX and Bcl-2 may affect their regulation by p53 and their association with risk of squamous cell carcinoma of the head and neck (SCCHN)47.
Hardy-Weinberg equilibrium (HWE) has been used for better understanding genetic characteristics of populations. HWE testing is commonly used for quality control of large-scale genotyping and is one of the few ways to identify systematic genotyping errors in unrelated individuals (Gomes et al. 1999; Hosking et al. 2004). In our HWE testing, we found that the p values of rs1009316ã€rs762624ã€rs2395655 are 0.240ã€0.258ã€0.297 respectively but the three SNPs rs3783468ã€rs681673ã€rs532446 are not fit the HWE. Some reasons for this: first, the sample size is small. Second, the three SNPs which are departure from HWE are located in the gene GADD45A which may be under natural selection in this population.
The optimal method for considering different genetic models in association studies is not clear when the inheritance pattern of the causal allele is unknown. One apparent solution to this problem is to test several genetic models. Until now, there are only few reports on the association between Bax (rs1009316) polymorphisms and risk of cancer. Rs1009316 CC genotype decreases the risk of ESCC in co-dominant, recessive and multiple genetic models. Rs1009316 located in the first intron of Bax and linked with the SNP rs4546878. Some studies have found that rs4645878 was associated with squamous cell carcinoma of the head and neck in Chinese population47 and associated with Chronic Lymphocytic Leukemia in Spain population48. And this SNP was associated with lower mean Bax expression in those CLL patients [52]. Liu zhigang has revealed that SNP rs4546878 of Bax gene may not be one susceptible genotype for esophageal cancer49 this may be different from our results. This may be caused by different population and different pathological types.
SNPs rs762624 and rs2395655 located in the promoter of CDKN1A gene are associated with the risk of ESCC. We speculated that these two SNPs might affect p53 binding to the promoter of CDKN1A and contribute to the susceptibility of ESCC. Ma Hongxia suggested possible association of polymorphisms rs2395655 of CDKN1A with the prognosis of NSCLC in a Chinese population50. Kim conclude that the minor allele A at rs762624 of CDKN1A is associated with increased susceptibility to SLE and lupus nephritis, and decreased cellular levels of p21 and luciferase reporter gene assay found that allele A displayed ~15% lower activity than that with the non-risk allele C 51. In our study, allele A at rs762624 of CDKN1A is associated with increased susceptibility to ESCC. The underline mechanism is allele A enhanced the regulation of p53 and increased the expression of CDKN1A. So, the individual carrying allele C may have low level of CDKN1A and high risk of ESCC.
In conclusion, we found three SNPs rs1009316ã€rs762624ã€rs2395655 located in the p53 binding regions were associated with the risk of ESCC and provide an experimental evidence for the early diagnosis of ESCC. However, there are some limitations in this study: Sample size and sample range is limited, and need to expand the sample size to further validate this result; the biological mechanism which the positive SNPs affect susceptibility to esophageal squamous cell is unclear. More experiments are needed to clarity the biological functions of the positive SNPs sites, and provide theoretical basis of the mechanism for the early diagnosis of esophageal squamous cell carcinoma.
