The relationship between P16INK4A and TP53 promoter methylation and the risk and prognosis in patients with oesophageal cancer in Thailand

DNA methylation can regulate the expression of tumour suppressor genes P16 and TP53, environmental factors, which are both important factors related to an increased risk and prognosis of oesophageal cancer (EC). However, the association between these two genes methylation status, as well as the effects of gene-environment interactions, EC risk remains unclear. A Hospital-based case–control study data were collected from 105 new EC cases and 108 controls. Promoter methylation status was investigated for P16 and TP53 genes using methylation-specific polymerase (MSP) chain reaction methods with SYBR green. Logistic and Cox regression models were used to analyse the association of P16 and TP53 promotor methylation status with EC risk and prognosis, respectively. Our results suggest P16, TP53 methylation significantly increased the risk of EC (OR = 5.24, 95% CI: 2.57–10.66, P < 0.001; OR = 3.38, 95% CI: 1.17–6.67, P < 0.001, respectively). In addition, P16 and TP53 promoter methylation status and the combined effects between environmental factors and its methylations in tissue were correlated with the EC risk and prognosis of EC patients. As a new biomarker, the methylation of P16 and TP53 can serve as a potential predictive biomarker of EC.


Materials and methods
Study design and population. This study was a hospital-based case-control study with 105 EC cases and 108 cancer-free controls. All subjects were recruited during the same period. All cases and controls were recruited from Srinagarind Hospital, Khon Kaen Province, Northeast Thailand, from 2007 to 2017. A total of 105 newly diagnosed EC patients were included in this study. All cases were histologically confirmed and diagnosed according to the International Classification of Diseases for Oncology (ICD-O 3rd), and the histological diagnosis was reviewed in each case confirmed by pathologists. A medical report was obtained from the pathology department. The control subjects selected were healthy individuals confirmed by physical examination and clinical and biochemical analysis during the same period of case recruitment. Control individuals with a history of malignant tumours or oesophageal malignancy were excluded. Control subjects were randomly sampled from patients undergoing routine endoscopy for investigation of presumed non-malignant conditions, such as gastroesophageal reflux disease. Areas of oesophagus biopsied were macroscopically normal in appearance. All subjects gave written informed consent for their participation in this study. This project was approved by the Human Research and Ethics Committee of Khon Kaen University (Reference No. HE621269).
Data collection. Data on subjects were obtained with an interviewer-based structured questionnaire, and data were collected from the recruited patients by a face-to-face interview with a trained interviewer using a standardised questionnaire. The questionnaire was considered for content validity by specialists in the field of EC and developed by researchers. All subjects completed a face-to-face investigation questionnaire to obtain demographic characteristic variables and information about social habits and lifestyle. The clinicopathology data were obtained from the electronic medical record system. All cases were followed up until death or the end of the www.nature.com/scientificreports/ study (31 October, 2019). Overall survival was considered the primary outcome. Thus, the classical endpoint in this study is survival time of EC, and the status of each patient was checked from medical records and by linkage with the death registry of the Thai national statistics database.
Tissue samples. Tissue  Laboratory method. DNA extraction and quality control. DNA extraction was from FFPE tissue. The FFPE oesophagus tissue was cut into ten μM sections in a 1.5 microcentrifuge tube. Eight sections from each sample were used for DNA extraction. Prior to sectioning, the microtome and accessories were cleaned with 70% ethanol alcohol. The DNA of the FFPE tissue samples were extracted using a commercially available system, DNeasy blood and tissue kits (Qiagen, Hilden, Germany), according to the manufacturer's instructions.
The yields and quality of the DNA isolated by the kit were measured by use of a Nano Drop ND-1000 Spectrophotometer (Thermo Scientific), and DNA quality was assessed. The housekeeping human beta actin (β-actin) gene served as an endogenous control to guarantee the DNA quality and was detected using a specific primer as previously described 40 . The integrity of the extracted DNA samples was confirmed by amplifying a housekeeping human beta actin (β-actin) gene using SYBR green real-time polymerase chain reaction (PCR). All DNA samples were stored at − 20 ๐ C for further analyses.
Bisulphite modification. Sodium bisulphite modification is based on the selective deamination of unmethylated cytosines to uracil, whereas methylated cytosines remain unchanged. The chemical reaction exhibits specific changes in the DNA sequences that reflect the methylation status of individual cytosine residues. Tissue DNA containing 1 μg was treated with bisulphite using an EZ DNA methylation Gold Kit (Zymo Research Corp., Irvine, CA, USA) according to the manufacturer's instructions. Briefly, a final volume of 20 µl bisulfitedmodified DNA was obtained and used immediately as a template for real-time methylation-specific PCR (MSP) analysis or stored at − 80 °C no longer than 4 weeks.
Real-time methylation-specific PCR. The methylation status of P16 and TP53 was performed using real-time MSP chain reaction with two primer sets, methylated primer and unmethylated primer. The specific primers of CpG islands related to promoters of these genes were designed 8 and were checked using Methyl Primer Express software version 1.0 (Applied Biosystems) and Bisearch (http:// bisea rch. enzim. hu/?m= genom psear ch). The methylated primer is specific to fully methylated sequences, while the unmethylated primer is specific to fully unmethylated sequences. PCR amplification and melting curve analysis were performed on an Applied Biosystems 7500 flats system. The PCR reaction mixture was performed in a final volume of 20 μl consisting of 8 ng of bisulphite-modified DNA1XSYBR qPCR Mix (Bio-Rad, Hercules, CA, USA), 2.5 mM MgCl2, 500 µM of each dNTP and 0.3 μM of each forward and reverse primer. The optimal conditions of MSP are shown in Table  Supplementary (S1). The amplification steps were as follows: initial denaturation at 94 °C for 5 min, 40 cycles of denaturation at 94 °C for 45 s, annealing at optimal temperature for 40 s and extension at 72 °C for 40 s, and a final extension at 72 °C for 5 min. The melting step was obtained from 60 to 95 °C in duplicate to confirm repeatability, and they were averaged following analysis. Human peripheral blood leukocyte DNA was treated with SssI methyl transferase (New England Biolabs, Inc., Beverly, MA, USA) as fully methylated DNA, while the leukocyte DNA was non-treated as fully unmethylated DNA. Each sample was done in duplicate. The melting temperature (Tm) of fully unmethylated DNA controls with the specific primer sets was defined as the cut-off point for methylation status. MSP assay of P16 and TP53 methylation in tissue DNA samples was determined by using fully methylated controls and fully unmethylated DNA controls with individual specific primer set is shown in Figs. S1 and S2. The data shows the amplification curve and melting curve, which represent the cutoff point of Tm for determining methylation status. Thus, these figures present the standard melting curves and melting peaks, for the definition P16 and TP53 methylation status.
Statistical analysis. The demographic characteristics of the subjects were summarised using descriptive statistics. Statistical analysis was performed to examine the differences in characteristics between cases and control group using the chi-squared test for categorical variables, and continuous variables were analysed using t-tests or rank-sum tests. Logistic regression analysis was applied to explore the relation of DNA methylation and environmental factors with EC. Odds ratios (ORs) and 95% confidence intervals (95% CIs) were applied to assess the strength of association between P16 and TP53 methylation status along with environmental factors and EC risk derived from unconditional logistic regression. Univariate and multivariate logistic regression analyses were approved to evaluate the relationships between gene methylation, environmental factors and EC risk, and the relationships between gene methylation and environmental factors. The interactions of gene methylation and environmental factors on EC risk were estimated on a multiplicative scale with a product-term coefficient using multivariate logistic regression. The combined effects of gene methylation and environmental factors on EC risk were calculated by crossover analysis. Multivariable logistic regression was used to compute adjusted odds ratios (ORadj.) and 95% confidence intervals (95% CIs) for the association between DNA methylation, www.nature.com/scientificreports/ environmental factors and EC, while controlling for the effects of confounding variables. The backward stepwise elimination method was used as the model fitting strategy. A likelihood ratio test was performed to assess the goodness-of-fit of the final model. The survival probabilities were determined by the Kaplan-Meier method. The statistic used to compare survival between groups was performed by using the log-rank test. Univariate and multivariate Cox proportional hazard regression models were used to estimate the association between explanatory gene methylation and clinical characteristics and prognosis of EC patients, and the results were presented in the form of crude and adjusted hazard ratios (HRs) and their 95% CIs. HR was assessed to investigate the magnitude and direction of the effect. All statistical analysis was performed using Stata® software (version 13.0), with the test statistics two-sided, and a p-value less than 0.05 was considered statistically significant.

Ethics declarations. This present study was approved by the Khon Kaen University Ethics Committee for
Human Research, based on the Declaration of Helsinki and the ICH Good Clinical Practice Guidelines; reference number HE621269.

Results
Demographic characteristics of subjects. The study included 105 cases and 108 controls. The characteristics of all subjects are provided in Table 1. The gender and age distribution differed between the cases and controls (men, 60 and 47%, respectively; women, 40 and 53%, respectively). The mean age (± SD) of the cases and controls was 60.4 years (± 8.8) and 60.0 years (± 9.2), respectively. The demographics of the subjects' lifestyle habits were evaluated. Drinking status, smoking and betel chewing were more prevalent among the cases than the controls. The distribution frequency of a family history of cancer between cases and controls statistically differed (P < 0.05). The results showed no difference between cases and controls in terms of body mass index (BMI) and marital status (P = 0.681 and P = 0.945, respectively).

Association between environmental factors and EC risk.
The correlations of environmental exposure with EC risk are presented in Table S2. The primary outcomes of the multivariable analysis are depicted in this table, and after backward conditional selection analysis, our results showed that alcohol drinking, smoking and family history of cancer significantly increase the risk of EC (P < 0.05). BMI, marital status and betel chewing were not significantly associated with EC.

The effect of interactions between P16 and TP53 methylation and their interactions with environmental factors on the risk of EC.
The results found that the combined effects between P16 methylation and alcohol drinking and smoking on EC risk existed (OR c = 2.43, 95% CI: 1.41-4.38, P = 0.002; OR c = 2.75, 95% CI: 1.73-5.10, P = 0.001 respectively), whereas no interaction between P16 methylation and environmental factors on the EC risk was showed, as seen in Table S5. As for the TP53 gene, its methylation and some environmental factors have combined effects on the risk of EC (P < 0.05) (Table S6). Additionally, the results in Table S7 illustrate that P16 methylation did interact with TP53 methylation on EC risk.

Characteristics of EC patients.
Of the 105 subjects who were recruited as EC cases in this study, all EC patients were included in this 10-year follow-up study. The association between demographic clinicopathological and EC prognosis was analysed, as presented in Table S8. Although the correlation between each demographic characteristic and prognosis of EC patients was not statistically significant, age, gender, BMI, family history of cancer, marital status and betel chewing were still used as the adjustment factors in analysing the relationship between clinical characteristics and EC prognosis, and these factors were common confounder factors in this study. Multivariate analysis based on Cox proportional hazard regression revealed that TNM stage, histology grading (poor differentiation) and lymph node metastasis were statistically significantly associated with EC prognosis (P < 0.05) ( Association between P16 and TP53 methylation status and EC prognosis. In this study, the potential impact of P16 and TP53 methylation on EC prognosis was investigated. Our data are shown in Table 4. When compared with non-methylation, both P16 and TP53 methylation were strongly associated with EC prognosis (HR = 2.82, 95% CI: 1.13-5.06; HR = 2.95, 95% CI: 1.34-5.47, respectively). In Fig. 1, our results show the association between P16 and TP53 methylation status and EC patients' prognosis by the Kaplan-Meier method. Briefly, this study found that DNA methylation of P16 and TP53 was associated with patients' overall survival. In addition, EC patients who had P16 and TP53 methylation showed significantly shorter overall survival than those who had non-methylation status (P = 0.027; P = 0.007, respectively).

Discussion
To the key point, this study is the first report on the molecular characteristics of EC from FFPE tissue patients, and this is the first study showing the possible interaction and association of epigenetics with environmental factors and the risk and prognosis in patients with EC in Thailand. EC is the result of both effects of environmental factors and epigenetic susceptibility, and behaviour is one of the most important influencing factors. Moreover, our results found that the P16 and TP53 methylation status was significantly associated with poor prognosis in EC patients. In this case-control study, our objective was to investigate the association and interaction of various environmental, behaviour factors and carcinogen metabolising aberrant DNA methylation, how they   In this study, we observed the possible association and interaction of epigenetic and environmental factors with the risk of EC. The results indicated that methylation of P16 and TP53 was significantly associated with the risk of EC. In addition, alcohol consumption and smoking status may have tumour-promoting effects of aberrant DNA methylation of P16 and TP53 genes. This interesting incidence was also consistent with in previous research 10,22 . As mentioned, the consumption of alcohol and tobacco contribute to ESCC carcinogenesis. A previous study found that promoter region hypermethylation was associated with tobacco consumption by analysing a group of tumour suppressor genes in ESCC patients when compared to the control group 42 . Similar results from Ye and Xu 43 found that Benzo (α) pyrene diol epoxide, as a carcinogen present in tobacco smoke, alcohol drinking and environmental pollution, has been shown to induce aberrant DNA methylation (such as P16, TP53 and KRAS genes). Although a previous study also confirmed tobacco smoking as a predominant risk factor for ESCC, the highest risk was associated with tobacco chewing in the concerned population. Tobacco is chewed in various forms either alone or with slaked lime or betel quid, and the spit is often swallowed. Like tobacco smoke, smokeless forms of tobacco are also known to contain several carcinogenic compounds, the most potent of which are the tobacco-specific N-nitrosamines like N'-nitrosonornicotine (NNN) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) 44 . Many epidemiological studies have consistently shown that alcohol drinking and tobacco use have synergistic effects on carcinogenesis, when combined use explained more than 70% of ESCC 45,46 . Indeed, Fan et al. 47 demonstrated that various factors from exposure to the same carcinogen (i.e., nitrogen, cigarette smoke, alcohol drinking) have been identified as risk factors for EC in a high-risk population, and these modifiable factors should be part of any primary prevention strategy for this EC, which has a very poor prognosis. Although the relationship between EC risk and DNA methylation has been previously reported, evidence indicating the association of environmental factors with DNA methylation and EC risk and prognosis in EC patients remains limited. The association between methylation status of P16, TP53 and EC prognosis was also explored in this study. These results reported on the effect of DNA methylation and clinicopathological characteristics on the survival of EC patients among the Thai population. As expected, we found that tumour characteristics, such as stage of cancer and metastasis, were associated with survival rates. In addition, some of the interesting factors, such as P16 and TP53 methylation, were related to the mortality of EC cancer patients. This finding suggested that DNA methylation is an epigenetic event commonly found in EC. The DNA methylation of these genes, at least in part, may potentially lead to gene silencing, which may have an important impact on cancer cell survival and progression. This finding is also consistent with previous studies 11 . Our results showed that the stage of cancer was the greatest factor affecting the survival of EC, which is similar to several previous studies that have reported this, especially in the advanced stage of the disease 48,49 . Moreover, in our finding, stage IV of EC related to the declining of survival when compared to other stages, as mentioned above; this is consistent with the reported by Fujiwara et al. 50 . A previous study concluded that differences in pathogenesis and tumour biology can be induced given the variance of survival time in EC patients. In addition, stage IV also contains not only those patients with metastasis but also those with advanced primary tumour and region lymph node status without metastasis. Thus, EC patients with stage IV (poor differentiation) have a significantly poor prognosis. Our research showed high methylation levels of P16 and TP53 genes, and these genes were associated with EC. Briefly, this finding suggested that DNA methylation is a common form of epigenetic modification, which has a crucial role in human malignancies, such as EC 51 , lung cancer 17 and gastric cancer 52 . Abnormal methylation in the promoter region of www.nature.com/scientificreports/ tumour suppressor genes is one of the most common mechanisms of modification, and the results in target gene transcriptional silencing, so that DNA methylation has become a credible potential biomarker for early detection and diagnosis of cancer, and it may have a great impact on cancer cell survival and prognosis 18,19 . The present study showed that promoter methylation of two genes, namely P16 and TP53, were not only tumour suppressor specific, but also correlated with a patient's overall survival time. An important aspect of this study is the potential methylation status of P16 and TP53, which can serve as predictive biomarkers for EC. P16 methylation is a favourable predictive marker, and TP53 methylation is clear to use for predictions; when used in combination, it may strengthen their potential marker. In particular, promoter methylation of P16 and TP53 are involved in the progression of EC, and clinical pathology, such as stage of disease and metastasis, is related to poor survival outcomes and shorter survival. Therefore, further large-scale prospective studies are needed to contribute to a deeper and more extensive understanding of epigenetic and environmental factors and their interaction in the different stages of carcinogenesis of EC. There were several limitations in the present study. First, recall bias might be inevitable for gathering information on environmental factors, although we attempted to reduce this bias. In particular, the collection data about the frequency and duration of use of alcohol and smoking are not detailed, and these data could affect the methylation of candidate genes. Second, we cannot determine the possible mechanisms that affect methylation differences in these genes between the cases and control group. This study had important strengths, a case-control study was conducted to apply molecular epidemiology pathology methods to the association and interaction of environmental factors and epigenetic biomarkers with the survival of EC.

Conclusion
In conclusion, our findings provide insight into aberrant DNA methylation of P16 and TP53 as an epigenetic event of EC, and these results indicated that P16 and TP53 promoter methylation status and the combined effects between environmental factors and their methylations in tissue were correlated with the EC risk and prognosis of EC patients. The methylation of P16 and TP53 can serve as a potential predictive biomarker of EC.

Data availability
The datasets during and/or analyzed during the current study available from the corresponding author on reasonable request.