TERT promoter methylation is associated with high expression of TERT and poor prognosis in papillary thyroid cancer

The telomerase reverse transcriptase (TERT) is overexpressed and associated with poor prognosis in papillary thyroid cancer (PTC), the most common subtype of thyroid cancer. The overexpression of TERT in PTC was partially attributed to transcriptional activation by two hotspot mutations in the core promoter region of this gene. As one of the major epigenetic mechanisms of gene expression regulation, DNA methylation has been proved to regulate several tumor-related genes in PTC. However, the association of TERT promoter DNA methylation with TERT expression and PTC progression is still unclear. By treating PTC cell lines with demethylating agent decitabine, we found that the TERT promoter methylation and the genes’ expression were remarkably decreased. Consistently, PTC patients with TERT hypermethylation had significantly higher TERT expression than patients with TERT hypomethylation. Moreover, TERT hypermethylated patients showed significant higher rates of poor clinical outcomes than patients with TERT hypomethylation. Results from the cox regression analysis showed that the hazard ratios (HRs) of TERT hypermethylation for overall survival, disease-specific survival, disease-free interval (DFI) and progression-free interval (PFI) were 4.81 (95% CI, 1.61-14.41), 8.28 (95% CI, 2.14-32.13), 3.56 (95% CI, 1.24-10.17) and 3.32 (95% CI, 1.64-6.71), respectively. The HRs for DFI and PFI remained significant after adjustment for clinical risk factors. These data suggest that promoter DNA methylation upregulates TERT expression and associates with poor clinical outcomes of PTC, thus holds the potential to be a valuable prognostic marker for PTC risk stratification.


Introduction
Thyroid cancer is the most frequently observed malignancy of the endocrine system (1).Based on histology, thyroid cancer can be divided into four main subtypes, including papillary thyroid cancer (PTC), follicular thyroid cancer (FTC), anaplastic thyroid cancer (ATC), and medullary thyroid cancer (MTC).Among them, PTC has a more favorable prognosis than other subtypes of thyroid cancer (2).However, as the most common subtype, PTC had a sharp increasing incidence in the past three decades and currently accounts for about 90% of all thyroid cases (3).
Overwhelming studies have demonstrated that a number of genetic alterations, including hotspot mutations and fusions in some key cancer driver genes like BRAF, RAS, RET, EIF1AX and telomerase reverse transcriptase (TERT), orchestrated the initiation and progression of thyroid cancer by upregulating the expression of certain oncogenes and downregulating the expression of tumor suppressor genes (4)(5)(6).BRAF V600E mutation and RET fusions are well studied in PTC and it is now well established that they are main drivers of the aberrant MAPK pathway (7,8).Two hotspot mutations, occurred collectively in around 10% of PTC, in the core promoter region of TERT activates the expression of TERT at the transcription level and thus promotes the tumorigenesis and development of thyroid cancer (4,(9)(10)(11).Clinically, TERT promoter mutations had been proved to be strongly associated with aggressive characteristics and poor prognosis of PTC (12)(13)(14)(15).
In addition to genetic alterations, epigenetic alterations also involved in the pathogenesis of PTC.As one of the most widely studied epigenetic mechanism, DNA methylation usually observed in the promoter region of tumor suppressor genes (TSGs) and inhibits their expression in human cancers (16).The expression of several classic TSGs, including RASSF1A, CDKN2A and DAPK, and two thyroid-specific genes, TSHR and NIS, had been identified to be silenced by methylation (17).Interestingly, increasing studies reported that DNA hypermethylation was frequently observed in the promoter region of TERT in human cancers and the promoter methylation was emerged as an epigenetic mechanism of TERT activation (18,19).Clinically, TERT hypermethylation was correlated with tumor progression and unfavorable prognosis in several types of cancer (20)(21)(22)(23).In this study, we assessed the association of TERT promoter methylation with its expression and the clinical outcomes of PTC by analyzing 571 PTC samples in The Cancer Genome Atlas (TCGA) thyroid cancer database, and explored the effect of promoter DNA methylation on TERT expression by demethylating assays in thyroid cancer cell lines.

Clinical data
The clinical data of patients with PTC used in this study was originated from the TCGA thyroid cancer (THCA) dataset.The following clinical factors and prognostic values were collected: age at diagnosis, gender, extrathyroidal extension, pathologic T/N/M, residual tumor, pathologic stage, overall survival (OS), disease-specific survival (DSS), disease-free interval (DFI) and progression-free interval (PFI).

TERT methylation and expression analysis
The b-value, ranging from 0 to 1, was used to show the methylation level of DNA methylation.The b-value of the CpG probe cg11625005 in the promoter of TERT and the Illumina HiSeq data of TERT expression for each thyroid sample were downloaded from TCGA dataset through the UCSC Xena platform.If the b-value of cg11625005 of one sample was higher than the mean b-value + 2*SD of the normal samples, this sample was defined as a case with TERT hypermethylation.

Cell line and reagents
Papillary thyroid cancer cell lines BCPAP and TPC1 were purchased from the National Collection of Authenticated Cell Cultures (Shanghai, China) and cultured in RPMI-1640 medium (#10-040-CVRC, Corning) with 10% fetal bovine serum (#04-001-1A, Biological Industries) at 37 °C.DNA demethylating drug decitabine was purchased from Selleck Chemicals (#S1200) and used for treating the two cell lines at 5 mM for 5 days.DMSO was used as the control reagent.

Cell viability assay
Two thousand and five hundred BCPAP cells or one thousand TPC1 cells were seeded in triplicate in each well of a 96-well cell culture plate and incubated with DMSO or decitabine at concentrations ranging from 0.1 to 10 uM for 5 days.The cell culture medium and drugs were refreshed daily.At the end of the drug-treated course, the cell viability was determined by the Cell Counting Kit-8 (#K1018, ApexBio) according to the manufacturer's instructions.The absorbance was measured at 450 nm using a spectrophotometric microplate reader.

Statistical analysis
All the in vitro experiments were performed three times with each done in triplicate.Comparisons of categorical variables were performed using the chi-square test.For continuous variables, the significance of differences between two groups was assessed by the Student's t-test.Kaplan-Meier curves with log-rank tests and Cox regression analysis were used to compare survival data by methylation status.All the statistical analyses were performed by Stata software (version 12.0) and GraphPad Prism (version 8.0).All P values were 2-sided and P<0.05 was considered as statistically significant.

Increased TERT expression in PTC patients with TERT hypermethylation
We first compared the TERT methylation level in 571 thyroid cancer samples with that in 56 normal thyroid samples and found that there was no significant difference between the two groups (P=0.312, Figure 1A).According to our criteria, the cut-off b-value for TERT hypermethylation was set at 0.494, and 33 PTC samples were defined as TERT hypermethylated cases.Then we analyzed the potential effect of TERT methylation on TERT expression by comparing the TERT mRNA levels in TERT hypomethylated and hypermethylated thyroid cancer samples.As shown in Figure 1B, the patients with TERT hypermethylation had a significant high level of TERT expression (1.72 ± 0.32 vs 0.40 ± 0.04, P<0.001).

Association of TERT methylation with clinical characteristics and outcomes of PTC
We next analyzed whether TERT methylation was associated with any of the following clinical characteristics of PTC: patient age, gender, extrathyroidal extension, pathologic TNM, residual tumor, and stage.The results showed that there was no significant Promoter DNA methylation and TERT expression.By performing Kaplan-Meier and cox regression analyses, we further assessed the impact of TERT methylation on the prognosis of PTC.The overall survival (OS) curves had a modest decline in patients with TERT hypomethylation, and declined more sharper in patients with TERT hypermethylation (Figure 2A).As shown in Figures 2B-D, similar patterns were obtained when analyzing disease-specific survival (DSS), disease-free interval (DFI) and progression-free interval (PFI).Results from the cox regression analysis showed that the HRs of TERT hypermethylation for OS and DSS were 4.81 (95% CI, 1.61-14.41)and 8.28 (95% CI, 2.14-32.13),respectively.These significances were lost after adjustment for multiple risk factors of PTC.The HRs of TERT hypermethylation for DFI and PFI were 3.56 (95% CI, 1.24-10.17)and 3.32 (95% CI, 1.64-6.71),respectively, and they remained significant after multiple variables adjustment (Table 3).Then we stratified all the PTC samples into two groups according to the tumor stage, and found that TERT hypermethylation was associated poor clinical outcomes in advanced PTC patients (stage III/IV, Tables 2, 4 and Figure 3).

Demethylating agent treatment decreased TERT expression and cell viability in TERT hypermethylated PTC cells
To determine whether TERT promoter methylation regulates the gene's expression in PTC, we selected BCPAP, a PTC cell line with TERT hypermethylation (24), as a model and treated it with the classic demethylation agent decitabine and then checked the changes of TERT methylation and mRNA expression.As show in Figures 4A, B, after treatment with the demethylating agent, the methylation levels of TERT promoter regions -633 to -540 and -392 to -321 were decreased from 73% to 51% and from 60% to 32%, respectively.Importantly, the mRNA expression of TERT was significantly decreased after decitabine treatment (Figure 4C).Next, we tested whether DNA demethylation affects the growth of thyroid cancer cells and found that decitabine decreased the cell viability of BCPAP cells in a dose-dependent manner (Figure 4D).Similarly, significant TERT mRNA downregulation and cell growth inhibition were observed when another PTC cell line TPC1 was treated by Kaplan-Meier analysis of the association of TERT methylation with clinical outcomes of all PTC.Results of the analysis for overall survival (A), diseasespecific survival (B), disease-free interval (C) and progression-free interval (D).The P values labeled in each panel were calculated by log-rank tests.4C, D).These data suggest that promoter DNA methylation positively regulates TERT expression in PTC cells and demethylation treatment suppresses the growth of PTC cells.

Discussion
TERT is regarded as a key oncogene in the tumorigenesis and development of human cancer.It is the catalytic subunit of telomerase which is essential for maintaining the length of telomere in cancer cell immortalization (25).And increasingly recent studies identified that TERT is also involved in cancer progression in a telomere-independent manner.For example, knockdown of TERT induced quick inhibition of cell proliferation, migration and invasion, while overexpression of TERT variants that lack telomerase activities promotes cell proliferation (26)(27)(28).In addition, TERT regulates NF-kB -dependent gene expression by binding to the promoter region of these genes, suppresses TGF-b -dependent growth suppression, and function as a central regulator of various hallmark features of cancer, including resistance to apoptosis, angiogenesis, metabolism reprogramming, genome instability (29,30).
TERT is overexpressed in thyroid cancer and associated with poor prognosis of PTC (31).Previous studies had shown that TERT promoter mutation is one of the major contributors to the gene's overexpression in thyroid cancer (32)(33)(34).Molecular studies further revealed that the two mutations in TERT promoter generated consensus binding sites for ETS transcription factors and thereby promoted TERT activation (4,(35)(36)(37).A large number of genetic association studies have demonstrated that TERT promoter mutation is associated with multiple characteristics of aggressive tumor behaviors and patients with TERT promoter mutation had a higher risk of recurrence and mortality (12)(13)(14)(15)38).
In this study we found that there is no significant difference of TERT promoter methylation level between normal thyroid tissues and PTC tissues, suggesting that promoter DNA methylation had little contribution to the overall increase of TERT overexpression in PTC.It should be noticed that the average level of TERT methylation is low in thyroid tissues, therefore we next defined hypermethylation in the tumor samples and obtained 33 of 531 (5.8%) TERT hypermethylated samples.Compared with patients carrying hypomethylated TERT, the TERT hypermethylated patients had significantly higher TERT expression.These data suggest that hypermethylation in the TERT promoter is likely to positively regulate TERT expression.
To confirm this, we assessed the effect of DNA methylation on TERT expression in a cell line model.After treating the PTC cells with demethylating agent decitabine, we found that the methylation level of TERT promoter was remarkably decreased and the expression of TERT mRNA was correspondingly decreased significantly.To the best of our knowledge, this is the first experimental evidence to show that TERT promoter DNA    methylation positively regulates TERT expression in PTC.Our finding is consistent with previous report that the DNA methylation in the TERT hypermethylated oncological region upregulates the promoter's activity (18).There are at least three possible mechanisms underlying the methylation mediated TERT activation.First, the transcriptional activator GSC showed the most remarkable binding in PTC cells than other cell lines, suggesting GSC acts as a putative regulator for methylated TERT in PTC (24).Second, allele-specific regulation of TERT was proposed to explain its activation in human cancer, and an interplay between TERT promoter mutation, methylation, and histone modification were involved in this process (24,(39)(40)(41)(42). Third, a 1.6 kb antisense long non-coding RNA, named hTERT antisense promoter-associated (hTAPAS), in the TERT promoter region was proved to regulate TERT expression through DNA methylation in urothelial cancer (43).As the epigenetic regulation on gene expression is usually occurred in a tissue type dependent manner, whether the methylation of hTAPAS involves TERT regulation in PTC needs to be further pursued.
The associations between TERT methylation and clinical tumor behaviors and outcomes of thyroid cancer had been investigated by few preliminary studies with limited number of thyroid patients (44,45).In this study we comprehensively analyzed the correlation between TERT methylation and clinical outcomes of PTC by enrolling 571 samples from the TCGA thyroid cancer database, and found that patients with TERT hypermethylation had unfavorable OS, DSS, DFI and PFI.Importantly, the association between TERT hypermethylation and DFI and PFI is still significant after adjusting multiple clinical parameters.These data strongly indicated that DNA hypermethylation in TERT promoter is associated with poor prognosis of PTC, and it could be serving as a prognostic marker for risk stratification of PTC.Similarly, the prognostic role of TERT promoter methylation had been reported in several other types of cancer, including childhood brain tumors (20), pancreatic cancer (21), bladder cancer (22), and breast cancer (23).
In conclusion, our data showed that DNA hypermethylation in the promoter region of TERT upregulated TERT expression in PTC, and TERT promoter hypermethylated patients with PTC had a poor clinical outcome.Our findings in this study indicate a prognostic value of TERT methylation in PTC and needs to be validated in further studies.
(A) Comparison of TERT promoter methylation levels between 56 thyroid normal tissues and 571 cancer tissues.(B) Comparison of TERT mRNA expressions in thyroid cancer samples with TERT hypermethylation (n=33) or hypomethylation (n=530).Data were shown as mean ± standard error.ns, not significant; ***P<0.001.

4
FIGURE 4 Impact of demethylating treatment on TERT expression and cell viability in PTC cell lines.DNA methylation levels of TERT promoter region regions -633 to -540 (A) and -392 to -321 (B) in BCPAP cells treated with DMSO or 5 uM of decitabine for 5 days.The methylated clones were shown with red color, and the unmethylated clones were shown with white color.(C) TERT mRNA levels in PTC cell lines treated with or without decitabine.The relative mRNA expression of TERT was determined by qPCR.(D) The absorbances of cells treated with DMSO or decitabine at various doses as indicated.DAC, decitabine.***P<0.001.

3
FIGURE 3Kaplan-Meier analysis of the association of TERT methylation with clinical outcomes of advanced PTC (stage III/IV).Results of the analysis for overall survival (A), disease-specific survival (B), disease-free interval (C) and progression-free interval (D).The P values labeled in each panel were calculated by log-rank tests.

TABLE 1
The association of TERT methylation with clinical characteristics of PTC.

TABLE 2
The association of TERT methylation with clinical outcomes of PTC.

TABLE 4
Hazard ratios of TERT methylation for clinical outcomes of advanced PTC.Adjusted for age at diagnosis, sex, extrathyroidal extension, pathologic T and N. OS, overall survival; DSS, disease specific survival; DFI, disease free interval; PFI, progression free interval. *

TABLE 3
Hazard ratios of TERT methylation for clinical outcomes of all PTC.
*Adjusted for age at diagnosis, sex, extrathyroidal extension, pathologic T and N. OS, overall survival; DSS, disease specific survival; DFI, disease free interval; PFI, progression free interval.