Scientific reports 一作

https://www.wendangku.net/doc/db7873696.html,/scientificreports

Polymorphisms in the 5′ upstream regulatory region of p21WAF1/CIP1 and susceptibility to oesophageal squamous cell carcinoma

Wenjun Y ang 1,2, Y ong Li 1,3, Tao Ning 1, Hong Cai 1, Zhiqiang Chen 4, Y ing Dong 2 & Y ang Ke 1

This study aims to scan the 5′-upstream regulatory region of the p21 gene to identify all putative functional single nucleotide polymorphisms (SNPs) and to evaluate the contribution of p21 variants to oesophageal squamous cell carcinoma (ESCC) in the Chinese Han population. Common SNPs were identified, and both locus-based and haplotype-based association tests were used to evaluate the potential risk of these p21 gene polymorphisms for ESCC. Immunohistochemistry assay was further performed to detect the P21 protein expression in ESCC specimens. Twenty three SNPs were identified and seven Tagging SNPs were chosen to represent all 23 SNPs. Univariate analysis indicated that the rs3829963 C and the rs2395655 G alleles increased susceptibility to ESCC (OR = 1.606 and OR = 1.572, respectively). The rs3829963 C and rs2395655 G alleles, combined with cigarette smoking, could further increase the risk for ESCC (OR = 2.657 and OR = 2.828, respectively). Additionally, the rs2395655 G allele appeared to elevate the positive rate of P21 expression in ESCC tissues, as compared to the A allele. This report demonstrates for the first time that rs3829963 and rs2395655, in the promoter of the p21 gene are potentially functional, modulating susceptibility to ESCC among the high-risk cigarette-smoking Chinese population.

Oesophageal cancer is the sixth most common cause of cancer-related deaths worldwide 1 and the fourth among both men and women in China 2. The incidence of oesophageal cancer has been found to show marked geo-graphical variation. Anyang County in Henan Province of China, located in the southern portion of the Taihang Mountains, exhibits the highest incidence and mortality rates of oesophageal squamous cell carcinoma (ESCC) globally 3. Environmental and genetic factors have been implicated in oesophageal carcinogenesis 4,5. In high-risk areas, ESCC tends to cluster strongly in families 6,7, suggesting that genetic susceptibility contributes to the high rates of ESCC in these areas 8–10, in combination with exposure to environmental risk factors, including smoking, alcohol drinking, and nutritional deficiencies, among others. Despite extensive research, the aetiology of ESCC remains inconclusive. Some of the most promising candidate genes appear to be involved in the regulation of cellular activity. The p21 WAF1/CIP1 gene (OMIM: 116899; hereafter, p21), a main downstream effecter of p53, is one of these promising candidate genes that have been shown to play vital roles in carcinogenesis. Because more than 50% of human cancers exhibit p53 mutations 11, it is conceivable that the existence of natural variants or mutations in the p21 gene could also be linked to the development of specific cancers. The natural genetic variants of the p21 gene have therefore emerged as a resource for studies aimed at understanding differences in cancer risk between individuals. By scanning the coding region of the p21 gene, our group previously demonstrated that only three SNPs existed in the coding region and 3′ untranslated region (3′ UTR) in a Chinese population 12. Many previous reports investigating the effects of p21 polymorphisms on cancer susceptibility have also mainly concentrated on these three SNPs 13,14. It has been reported that P21 expression can be controlled at the transcriptional level

1

Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education), Department of Genetics, Peking University School of Oncology, Beijing Cancer Hospital and Institute, Beijing 100142, P. R. China. 2Key Laboratory of Fertility Preservation and Maintenance (Ministry of Education), Cancer Institute of the General Hospital, Ningxia Medical University, Y inchuan, Ningxia, 750004, P. R. China. 3Department of Laboratory Animal, Peking University School of Oncology, Beijing Cancer Hospital and Institute, Beijing 100142, P. R. China. 4Radiology Department of General Hospital, Ningxia Medical University, Y inchuan, Ningxia, 750004, P . R. China. Correspondence and requests for materials should be addressed to Y .K. (email: keyang@https://www.wendangku.net/doc/db7873696.html,)

R eceived: 26 June 2015A ccepted: 17 February 2016P ublished: 02 March 2016

OPEN

by both p53-dependent and p53-independent mechanisms. That is, apart from p53, a variety of other factors are known to activate or repress P21 expression by binding to specific cis -acting elements located within the p21 pro-moter region 15, including Sp1/Sp3, Smads, Ap2, and E2F. Therefore, sequence variations, such as single nucleotide polymorphisms (SNPs) in the p21 promoter region, likely alter P21 expression and lead to abnormal biological responses, including cancer.

In the present study, we first scanned the 5′ -upstream regulatory region of the p21 gene to identify all putative functional polymorphisms within 96 Chinese Han people from Henan Anyang County. The pattern of link-age disequilibrium (LD) across the 5′ -upstream regulatory region was characterised, and tagging SNPs (tSNPs) were selected. Then, these tSNPs were further genotyped in the overall study population. Both locus-based and haplotype-based association tests were employed to evaluate any potential association of the p21 gene with ESCC.

Results

SNP identification. A total of 23 polymorphisms were detected in the 5′ -upstream regulatory region of the

p21 gene. In the 5′ flanking region, 13 polymorphisms were identified upstream of the transcriptional start point (rs4135234–rs4135239). No polymorphism was discovered in the first exon. Ten polymorphisms were identified in intron 1 (rs3176320–rs3176337). Position information and mutation types for all SNPs are presented in table 1, along with the minor allele frequencies. No significant deviation from Hardy-Weinberg equilibrium was observed for any polymorphism.

Linkage disequilibrium, haplotype structure and selection of tSNPs. Figure 1 displays the pairwise

LD coefficients (D ′ ) of the 23 SNPs. Overall, the variants in this 5 Kb region exhibited strong LD. Because pair-wise LD tests within a gene are not independent, the P values of the D ′ values were adjusted based on the number of SNP pairs in the p21 gene using an FDR correction. Notably, most of the polymorphisms in the 5′-upstream regulatory region (rs4135234–rs4135237–rs3829966–rs3829968–rs762623–rs4135239–rs3176330–rs3176331, rs3829963–rs3829965, rs730506–rs4151702–rs3176326, and rs3176320–rs3176322–rs3176323–rs4135240–rs3176334) were in perfect LD with one another.

Table 2 shows the inferred haplotypes (frequency > 1%) of the 23 SNPs in the 96 subjects. Four truly common haplotypes each presented a frequency of over 10%, constituting 70% of the total. The frequencies of the remain-ing three haplotypes observed were approximately 5%.

After screening, a total of seven tSNPs (rs4135234, rs3829963, rs3829964, rs762624, rs2395655, rs730506, and rs3176320) were selected that showed a D ′> 0.95 using Haploview software, which were the same as those that exhibited an R 2

H threshold of 0.9 using GOLD software (Table 2).

Table 1. Summary data on 23 polymorphisms discovered in 96 Chinese subjects. a rs No. was shown if

present in the dbSNP database. b 5′ flanking means the upstream from the first transcribed nucleotide. c The base immediately preceding the start of transcription numbered as ‘?1’.d With major allele given first and minor allele given second. e Using 96 Chinese Han subjects.

Genotypes and risk of ESCC. Seven tSNPs were genotyped in all subjects. The characteristics of all subjects

and the genotype distributions of the tSNPs of the p21 gene are shown in Tables 3 and 4. No significant differences were found in age or gender (both p > 0.05) between the cases and controls. However, the distribution of cigarette smoking and alcohol consumption between the cases and controls was significantly different (both p <0.05).Single SNP analyses indicated that both rs3829963 and rs2395655 were significantly associated with ESCC. Compared with rs3829963 AA, the rs3829963 C allele was associated with higher susceptibility to ESCC, even after adjustment for multiple conventional ESCC risk factors (OR = 1.606, 95% CI =1.148–2.247, p <0.05).

Similarly, the rs2395655 G allele was associated with higher susceptibility to ESCC compared with rs2395655 AA

Figure 1. p21 gene structure and linkage disequilibrium (LD) map for the SNPs in the promoter region of p21. genotyped in the current study. (A ) Gene map and polymorphisms in the p21 gene on chromosome 6p21.2. Coding exons are indicated with black blocks and 5′ - and 3′ -UTRs with white blocks. The first base of the transcription start site is denoted as +1. (B ) LD blocks between SNPs in the promoter region of p21 in 96 healthy Han Chinese individuals. Each diamond represents the correlation (D ′ ) between each pair of SNPs. A standard colour scheme is used to display the LD pattern, where dark grey indicates very strong LD; white indicates no LD; and bright grey and shades of grey represent intermediate LD, with an increasing intensity of

grey indicating an increasing degree of LD.

Table 2. Haplotype structure of the p21 gene in Chinese population. *Tagging SNPs. a Haplotype designation. b Frequency of the common haplotypes estimated by E-M algorithm. c

Cumulative frequency.

Table 3. Comparison of characteristics between cases and controls. Means ± S D values for continuous variables. *Two-sided χ 2 test.

homozygotes, even after adjustment for multiple conventional ESCC risk factors (OR = 1.572, 95% CI =1.129–2.190, p < 0.05). No association was found with the other tSNPs (Table 4).

After eliminating rare haplotypes (with an estimated frequency of < 0.03), 7 remaining haplotypes required evaluation. The haplotype-based test was performed using the haplo.stats program. For haplo.score analy-sis, p =2×10?5 for the global distribution of all haplotypes, with 9 degrees of freedom (df) between cases and controls, which was identical to the empirical P value (p = 0.01) based on 1,000 simulation repetitions. Haplotype-based association analyses revealed that the frequency of Hap 1 GACCAGA was lower in cases than in controls (11.3% vs. 14.3%, p = 0.036) after adjusting for covariates (Supplemental Table 2).

To identify possible risk haplotypes and evaluate the effect of each haplotype, we ran the haplo.glam model, in which Hap 1 GACCAGA, the protective haplotype, was chosen as the baseline. The results of model fitting are shown in supplemental Table 3. Compared with this haplotype, no other haplotypes were found to significantly increase the risk of ESCC (p >0.05).

We further statistically investigated the association analyses of the rs3829963 or rs2395655 genotypes and other potential risk factors, including smoking and alcohol consumption, on the risk of ESCC. Using non-smokers carrying rs3829963 AA as a reference, smokers harbouring the AA genotype (OR = 1.845, 95% CI =0.953–3.576) and non-smokers with the CA or CC genotype (OR = 1.460, 95% CI = 1.002–2.174) were found to display an increased risk for ESCC. Furthermore, smokers with the CA or CC genotype presented a significantly increased risk for ESCC (OR = 2.657, 95% CI = 1.702–4.149). Similarly, using non-smokers carrying rs2395655 AA as a ref-erence, smokers with the AA genotype (OR = 2.068, 95% CI = 1.109–4.248) and non-smokers with the GA or GG genotype (OR = 1.581, 95% CI =

1.059–

2.360) were observed to show an increased risk for ESCC. Furthermore,

Table 4. Adjusted OR for the association of tSNPs with ESCC. *Adjusted for age, gender, smoking and alcohol drinking status.

the crossover analyses showed that smokers with the GA or GG genotype exhibited a significantly increased risk for ESCC (OR = 2.828, 95% CI = 1.818–4.389) (shown in Table 5). However, no significant result was observed between the tSNPs and alcohol consumption via logistic analysis (P > 0.05, data not shown).

Association of the rs3829963 or rs2395655 genotypes with the P21 expression of ESCC patients. To further analyze the effects of the rs3829963 or rs2395655 genotypes on P21 expression in human

tissues, we examined the expression of P21 protein in 93 cases of ESCC tissues. The representative examples of P21 staining negative, positive I and above were shown in Fig. 2. The brownish signals represent the positive stain-ing and were found mainly in the nucleus of tumor cells and the para-basal layer of normal esophageal mucosa. As shown in Table 6, 34 (36.6%) out of 93 ESCC specimens were scored P21 staining positive I and above. The positive rate of P21 protein was much higher in ESCC tissues carrying the rs2365955 G allele than tissues with the A allele (31.2% vs. 5.4%; p = 0.001), indicating significant association between the rs2365955 genotypes and P21 protein expression. However, no association was shown between the rs3829963 genotypes and P21 expression (p >0.05).

Discussion

The p21 gene, located on chromosome 6p21.2, is also known as W AF1, CIP1, SDI-1, MDA6, and Cap20 because its cDNA was cloned independently by several groups using a number of different screening strategies 16. The gene consists of three exons and two introns and encodes a 21 kDa protein. The translation region lies mainly in exon217. As the first cyclin-dependent kinase inhibitor (CDKI) to be identified to date, p21 can bind cyclin/CDK complexes and negatively modulates cell cycle progression. It is also part of quaternary complexes that contain cyclin, CDK, and PCNA, a processivity factor for DNA polymerase δ(Pol δ ). The formation of these complexes is essential for cell cycle progression. Through a direct interaction with PCNA, p21 prevents DNA synthesis and regulates DNA methylation 18. Varieties of transcriptional factors activate or repress P21 expression by binding to specific cis -acting elements located within the p21 promoter region 16. Naturally occurring SNPs in the p21 promoter likely have a great impact on its transcriptional activity by altering its ability to bind to certain nuclear proteins. In the present study, we applied tSNP selection and haplotype-based analysis methods to identify func-tional SNPs in the 5′ -upstream regulatory region of the p21

gene in a Chinese population and investigated their

Table 5. Risk of ESCC associated with p21 rs3829963 and rs2395655 by smoking status. ΔP <0.01. a Adjusted

for age, gender, and alcohol drinking status.

Figure 2. Immunohistochemistry staining of the P21 protein in ESCC tissues. (A ) No–anti p21, negative control with primary antibody replaced by PBS. (B ) Anti–p21, the brown signals represent positive staining for P21 protein and the staining was scored on a scale as indicated in the Materials and Methods. Positive cases were defined as those with over 10% of examined cells stained. Magnification, ×4

00.

Table 6. Association of rs3829963 and rs2365955 with P21 expression of ESCC patients. a Two-sided χ 2 test.

association with the risk of developing ESCC. This was the first systematic scanning study of polymorphisms in the 5′-upstream regulatory region of the p21 gene in a Chinese population.

Deletion analysis of the p21 promoter revealed a distal (positions -2,300 to -2,100 bp) and a proximal (posi-tions ?124 to ?61 bp) region that act synergistically to achieve high levels of constitutive expression19. Two p53 binding regions have also been previously confirmed at approximate positions of ?1,400 bp and ?2,300 bp20. Therefore, we screened the 5′flanking region up to 2.5 Kb upstream and the region extending 2.6 Kb down-stream from the transcriptional start site. We identified 23 polymorphisms within a distance of just over 5 Kb from the p21 gene. We also found that complete LD or even perfect LD existed among most of these variants, indicating that all of polymorphisms might exist in one large haplotype block within the 5′-upstream regulatory region of the p21 gene. Using tSNPs software, we chose seven tSNPs (rs4135234, rs3829963, rs3829964, rs762624, rs2395655, rs730506, and rs3176320) to represent all other SNPs and subsequently detected them within the over-all study population. Analyses were conducted based on both single SNPs and haplotypes. The single SNP univar-iate analysis indicated that the rs3829963 C allele and the rs2395655 G allele conferred risk for ESCC, even after adjustment for multiple conventional ESCC risk factors, and they may therefore likely correspond to susceptible forms of ESCC. We also found that the rs3829963 C allele and the rs2395655 G allele, combined with cigarette smoking, further increased the risk for ESCC. Haplotypes are the particular combinations of alleles observed in a population. In this report, haplotype-based association analyses revealed that the frequency of the GACCAGA haplotype was lower in cases than in controls after adjusting for covariates, indicating that the GACCAGA hap-lotype may play a protective role in ESCC.

In the present study, we confirmed 23 SNPs in the 5′-upstream regulatory region of the p21 gene in a Chinese population. In general, mutations in biologically important genes may readily lead individuals to a high risk for diseases under balanced selection. Therefore, embryonic cells carrying these mutations may die off in an early stage, and only those mutations with mild effects on biological functions will be maintained and transmitted to daughter cells, some of which may inevitably result in susceptibility to diseases under certain living conditions. By scanning most of the p21 gene in a Chinese population, our group identified 3 polymorphisms within 3 Kb of the coding region and 23 polymorphisms within 5 Kb of the regulatory region, which suggested that the coding region of p21 was highly conserved, while still presenting numerous variants in the promoter region. Although all 23 polymorphisms have been recorded in the NCBI SNP database, we report the first association with ESCC in a Chinese population to the best of our knowledge. In this study, we also found that cigarette smoking was a risk factor for ESCC. When smoking was combined with the rs3829963 C allele or the rs2395655 G allele, the risk for ESCC increased further. We used 93 cases of paraffin-embedded ESCC specimens and performed immu-nohistochemistry assay to detect the P21 protein expression. In these ESCC specimens, positive rate of the P21 protein expression in ESCC tissues with the rs2365955 G allele was raised obviously in comparison with the A allele, suggesting that the rs2365955 genotypes were associated with the P21 protein expression in ESCC tissues. All in all, this was the first systematic polymorphism scanning study in p21 gene in Chinese ESCC population, which provided us some valuable information about two potential functional SNPs. However, there are some limitations in our work. First, all SNPs have only been tested in ESCC population in limited samples. Therefore, replication in a larger sample of ESCC or other cancers is still required. Second, since all our subjects, includ-ing patients and controls were only selected from either the Henan Anyang Cancer Hospital or communities of Henan Province, China, selection bias cannot be ignored. Last, further functional work is also needed, which will provide additional information about the functional polymorphisms in cancers.

Materials and Methods

The workflow chart of the whole study was showed in Supplemental Figure 1.

Subjects for the case-control Study. A total of 490 patients with ESCC and 600 cancer-free controls were enrolled in this study. All of the subjects were unrelated ethnic Han Chinese. Patients with histopatholog-ically confirmed ESCC were recruited from January 2004 to December 2007 at the Anyang Cancer Hospital of Henan Province, China. Demographic data, including the age at diagnosis, gender, cigarette consumption and alcohol use of the subjects, were obtained from medical records. Subjects with previous cancer and those who had received chemotherapy or radiotherapy were excluded.

Controls matching the cases for gender and age (±5 years) were randomly selected in the same regions as the patients from a community-based cancer screening program for the early detection of oesophageal cancer. The recruited subjects were asymptomatic residents with ages from 40 to 75 years with no contraindications regard-ing oesophageal cytology or endoscopic examinations. Eligible subjects who agreed to participate and signed an informed consent form were assigned a study identification number and referred for examinations. Each sub-ject was interviewed to obtain demographic and risk factor information and underwent an oesophageal balloon cytology examination and an endoscopy examination. Exfoliated oesophageal cells were collected in the balloon cytology examinations. Cells on the surface of the balloon were smeared onto slides for cytological diagnosis.

Blood samples from all subjects were collected for DNA preparation and p21 gene geno-typing. All experimental protocols in this study were approved by the Peking University Ethical Review Committee. All the methods were carried out in accordance with the approved guidelines All participants pro-vided written informed consent according to local regulations.

SNP identification. Searches for SNPs were performed in DNAs from 96 blood samples, which were ran-domly selected from 600 cancer-free subjects. These samples included 192 chromosomes, providing at least a 95% confidence level for the detection of alleles with frequencies >1%. A total of 5,171 bp were screened, including the

5′flanking region up to 2,642 bp upstream of the transcriptional start site and extending 2,529 bp downstream of the transcriptional start site.

The 5′flanking region starting 2,642 bp upstream of the transcriptional start site, the first exon, and the exon 1/intron 1 boundary were screened via direct DNA sequencing. Five sets of PCR primers were designed for SNP screening with Oligo 6.0, and five fragments were amplified from each subject (Supplemental Table 1). PCR was performed in a 25 μl reaction mixture containing 50 ng of DNA, 0.1 μM each primer, 0.2 mM dNTPs, 2.0 mM MgCl2, 0.5 units of Taq DNA polymerase (Promega, Madison, WI), and 1 ×reaction buffer. The PCR profile consisted of an initial melting step of 2 min at 95 °C, followed by 35 cycles of 45 s at 94 °C, 40 s at 59–62.5 °C (variable Tm values according to different fragments), and 45 s at 72 °C, with a final elongation step of 10 min at 72 °C. SNPs were identified by directly sequencing the obtained PCR products with ABI PRISM Dye Terminator Sequencing kits (Applied Biosystems, Foster City, CA) and loading the samples onto an ABI 3700 sequencer. SNP candidates were then confirmed and analysed by two independent observers. We further confirmed these SNP positions and genotypes by re-amplifying and re-sequencing the SNP sites from the opposite strands.

We compared the SNPs which were detected in the direct DNA sequencing with those in the dbSNP-NCBI in the same region (2642 bp upstream, the 1st exon, boundary of exon 1 and intron 1). All detected SNPs were the same as SNPs in dbSNP. Therefore, to screen GC-rich intron1 region for polymorphisms, SNPs (with frequency >1%) were selected from the dbSNP-NCBI (https://www.wendangku.net/doc/db7873696.html,/entrez/query.fcgi?db=s np) public database and further confirmed using Beckman GenomeLabTM SNPstream UHT (Ultra-High Throughput) Genotyping System (Beckman/Coulter, Fullerton, Calif.) by Shanghai Bio Star Genechip Inc.. Primers and probes were supplied directly by Beckman Coulter. Sequences are available on request. The target genomic sequences containing the SNPs of interest were amplified in 12-plex PCR reactions. After enzymatic clean-up, PCR products were subjected to an extension reaction using 5′-tagged extension primers and fluorescent dye-labeled termi-nators. In a thermal cycled extension step, the primers hybridized to the specific amplicons 1 base adjacent to the SNP site and were extended by one base at the 3′end with a fluorescently labeled nucleotide. These reaction products were transferred to an array plate, where each of the 12 extension products in the multiplex reaction was sorted by hybridization of the unique 5′-tag sequence to its complementary probe immobilized in a mini-array within each well. The reader imaged the microarray plates. The 2-color detection allowed SNP detection by com-paring signals from the 2-fluorescent dyes. The image signals were then transferred to genotyping software that translated the images of the arrays into genotype calls. The error rate on the basis of blind replicates was 0.1% to 2.0% for the SNPs examined in this study.

Linkage disequilibrium, haplotype structure and selection of tSNPs. All patterns of pairwise LD among the SNPs and the correlation coefficient D′for each pair of polymorphic sites were measured using both Haploview software (version 4.1, https://www.wendangku.net/doc/db7873696.html,/mpg/haploview) and GOLD software (http://www. https://www.wendangku.net/doc/db7873696.html,/csg/abecasis/GOLD/). Haplotype frequencies were estimated using the ‘partition-ligation’ E-M algorithm, as implemented in the tSNPs program of GOLD software. The tSNPs software was used to identify an optimal set of tSNPs that optimise the predictability of common haplotypes utilising the R2H statistic. We ran the program with the following parameters: common haplotypes were defined as the minimal set of haplotypes that covers 70% of existing haplotypes, and sets of tSNPs resolving the common haplotypes were selected at an R2H threshold of 0.9 21,22.

Genotyping. Genomic DNA was obtained from peripheral blood lymphocytes from all controls and patients via standard phenol-chloroform extraction. Seven tSNPs were selected and genotyped from all subjects. All gen-otyping was performed via direct DNA sequencingor single-base-pair extension on the Beckman GenomeLabTM SNPstream UHT Genotyping (Beckman/Coulter, Fullerton, Calif) platform. To assess genotype concordance, rs4135234 and rs3176320 were double-genotyped using both methods. All results were 100% concordant.

ESCC tissue samples. To investigate the association of P21 protein expression and genotypes of rs 3829963 or rs 2365955, 93 patients, who were retrieved from the surgical patients of 490 ESCC patients mentioned above from the Anyang Cancer Hospital of Henan Province, China. The enrolled patients involved 72 men and 21 women and aged 39 to 75 years. In all of these cases, the primary treatment was surgical and no patients had suffered from severe postoperative complications. All the specimens had been routinely formalin-fixed, paraffin-embedded, serially sectioned at 4 μm in thickness.

Immunohistochemical analysis. P21 protein expression in ESCC tissues was examined using S-P immu-nohistochemical method. Briefly, after routine deparaffinization and hydration, slides were treated with 1% hydrogen dioxide and then heated in EDTA (pH 8.0) for antigen retrieval. Following 10% goat serum blocking, tissue sections were then incubated with Mouse Anti-Human P21WAF1 protein, (Biogenex, San Raman, CA) at 4 °C overnight. After rinsing, sections were subsequently incubated with goat anti-mouse biotin-conjugated IgG for 15 min and then with streptavidin-peroxidase conjugate for 15 min. The signal was developed with diaminoben-zidine, and slides were counterstained with 5% hematoxylin. The brown signals located in the nuclear represent positive staining for P21 protein. Sections without monoclonal antibody treatment were used as negative controls. The staining was scored on a scale from 0 to III as follows: 0, less than 10% cells were stained; I, 10–24% cells were stained; II, 25–49% cells were stained; and III, 50–100% cells were stained. Scores I–III were classified as positive, while score 0 was negative.

Association analyses. The main purpose of our analyses was to test the association of tSNPs or haplotype variation in the p21 gene with ESCC. Analyses were conducted separately for each tSNP, followed by haplotype analyses.

Descriptive statistical analyses and unconditional logistic regression were performed with Statistical Analysis System software (version 8.01; SAS Institute, Cary, NC). Hardy-Weinberg equilibrium was tested by comparing the expected and observed genotype frequencies using χ2 statistics. Logistic regression analyses by dominant model were used to estimate the association between genotypes and ESCC risk. Odds ratios (ORs) were adjusted for gender, age, cigarette smoking, and alcohol drinking. Crossover analyses were used to conduct association analyses by cigarette smoking or alcohol drinking status. A P value of <0.05 was used as the criterion for statisti-cal significance, and all statistical tests were two-sided. To test the association of statistically inferred haplotypes with ESCC, we used the haplo.score approach outlined by Schaid et al.23. This method assigns a probability to each haplotype pair in each individual and then directly models the phenotype of the individual as a function of each inferred haplotype pair, weighted by their estimated probability to account for haplotype ambiguity. This program can adjust for covariates and compute simulation P values both globally and for each haplotype. The number of simulations for empirical P values was set to 1,000. Because haplo.score does not provide the mag-nitude of the effect of each haplotype, haplo.glm was employed to calculate adjusted ORs and 95% confidence intervals (CIs) for each haplotype24. Both haplo.score and haplo.glm were implemented in haplo.stats software, developed using the R language25.

References

1. Ferlay, J. et al. GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon,

France: International Agency for Research on Cancer; 2013. Available from: http://globocan.iarc.fr, accessed on 07/05/2014.

2. Chen, W. et al. Annual report on status of cancer in China, 2010. Chin J Cancer Res. 26, 48–58 (2014).

3. Wang, L. D. et al. Cytological screening and 15 years′follow-up (1986-2001) for early esophageal squamous cell carcinoma and

precancerous lesions in a high-risk population in Anyang County, Henan Province, Northern China. Cancer Detect Prev. 29, 317–322 (2005).

4. Hongo, M., Nagasaki, Y. & Shoji, T. Epidemiology of esophageal cancer: Orient to Occident. Effects of chronology, geography and

ethnicity. J Gastroenterol Hepatol.24, 729–735 (2009).

5. Cheung, W. Y. & Liu, G. Genetic variations in esophageal cancer risk and prognosis. Gastroenterol. Clin North Am. 38, 75–91 (2009).

6. Guohong, Z. et al. Genetic heterogeneity of oesophageal cancer in high-incidence areas of southern and northern china. PLoS ONE.

5. e9668 (2010).

7. Zhang, W. et al. Segregation analysis of esophageal cancer in a moderately high-incidence area of northern China. Am J Hum Genet.

67, 110–119 (2000).

8. Morita, M. et al. Alcohol drinking, cigarette smoking, and the development of squamous cell carcinoma of the esophagus:

epidemiology, clinical findings, and prevention. Int J Clin Oncol. 15, 126–134 (2010).

9. Yang, C. X. et al. Risk factors for esophageal cancer: a case-control study in South-western China. Asian Pac J Cancer Prev. 6, 48–53

(2005).

10. Siassi, F. & Ghadirian, P. Riboflavin deficiency and esophageal cancer: a case control-household study in the Caspian Littoral of Iran.

Cancer Detect Prev. 29, 464–469 (2005).

11. Bellini, M. F. et al. Alterations of the TP53 gene in gastric and esophageal carcinogenesis. J Biomed Biotechnol. 2012, 891–961 (2012).

12. Xi, Y. G. et al. p53 polymorphism and p21WAF1/CIP1 haplotype in the intestinal gastric cancer and the precancerous lesions.

Carcinogenesis. 25, 2201–2206 (2004).

13. Pietsch, E. C., Humbey, O. & Murphy, M. E. Polymorphisms in the p53 pathway. Oncogene. 25, 1602–1611 (2006).

14. Yang, W. et al. p21 Waf1/Cip1 polymorphisms and risk of esophageal cancer. Ann Surg Oncol. 17, 1453–1458 (2010).

15. Gartel, A. L. & Radhakrishnan, S. K. Lost in transcription: p21 repression, mechanisms, and consequences. Cancer Res. 65,

3980–3985 (2005).

16. Gartel, A. L. & Tyner, A. L. Transcriptional regulation of the p21((WAF1/CIP1)) gene. Exp Cell Res. 246, 280–289 (1999).

17. el-Deiry, W. S. et al. W AF1, a potential mediator of p53 tumor suppression. Cell. 75, 817–25 (1993).

18. Strzalka, W. & Ziemienowicz, A. Proliferating cell nuclear antigen (PCNA): a key factor in DNA replication and cell cycle regulation.

Ann Bot. 107, 1127–1140, (2011).

19. Moustakas, A. & Kardassis, D. Regulation of the human p21/WAF1/Cip1 promoter in hepatic cells by functional interactions between

Sp1 and Smad family members. Proc Natl Acad Sci. 95, 6733–6738 (1998).

20. Saram?ki, A. et al. Regulation of the human p21(waf1/cip1) gene promoter via multiple binding sites for p53 and the vitamin D3

receptor. Nucleic Acids Res. 34, 543–554 (2006).

21. Stram, D. O. et al. Modeling and EM estimation of haplotype-specific relative risks from genotype data for a case-control study of

unrelated individuals. Hum Hered.55, 179–190 (2003).

22. Stram, D. O. et al. Choosing haplotype-tagging SNPS based on unphased genotype data using a preliminary sample of unrelated

subjects with an example from the Multiethnic Cohort Study. Hum. Hered. 55, 27–36 (2003).

23. Schaid, D. J. et al. Score tests for association between traits and haplotypes when linkage phase is ambiguous. Am J Hum Genet. 70,

425–434 (2002).

24. Lake, S. L. et al. Estimation and tests of haplotype-environment interaction when linkage phase is ambiguous. Hum Hered. 55, 56–65

(2003).

25. Boston, R. C. & Sumner, A. E. STATA: a statistical analysis system for examining biomedical data. Adv Exp Med Biol. 537, 353–369

(2003).

Acknowledgements

We thank Wiley for the linguistic assistance during the preparation of this manuscript. We thank all patients, volunteers, and their families who participated in this study. This work was supported by ‘973’ Project of National Ministry of Science and Technology Grant (2011CB504300 and 2012CB910800); National Natural Science Foundation of China (30872937, 30930102, 30871366, 81460434 and 81160249).

Author Contributions

Conceived and designed the experiments: Y.K. Performed the experiments: W.Y. and Y.L. Analyzed the data: W.Y., Y.L., H.C., Z.C. and Y.D. Contributed reagents/materials/analysis tools: T.N. Wrote the paper: W.Y Prepared all figures: Y.L. and Y.D. All authors read and approved the final manuscript.

Additional Information

Supplementary information accompanies this paper at https://www.wendangku.net/doc/db7873696.html,/srep

Competing financial interests: The authors declare no competing financial interests.

How to cite this article: Yang, W. et al. Polymorphisms in the 5′ Upstream Regulatory Region of p21WAF1/CIP1 and Susceptibility to Oesophageal Squamous Cell Carcinoma. Sci. Rep.6, 22564; doi: 10.1038/srep22564 (2016).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images

or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit https://www.wendangku.net/doc/db7873696.html,/licenses/by/4.0/