DNA methylation in thyroid cancer
in Endocrine-Related Cancer
Authors: Carles Zafon 1 , 2 , Joan Gil 3 , Beatriz Pérez-González 3 and Mireia Jordà 2 , 3
View Less
1 Diabetes and Metabolism Research Unit (VHIR) and Department of Endocrinology, University Hospital Vall d’Hebron and Autonomous University of Barcelona, Barcelona, Spain 2 Consortium for the Study of Thyroid Cancer (CECaT), Catalonia, Spain 3 Program of Predictive and Personalized Medicine of Cancer, Germans Trias i Pujol Research Institute (PMPPC-IGTP), Barcelona, Spain
Correspondence should be addressed to M Jordà: mjorda@igtp.cat
DOI: https://doi.org/10.1530/ERC-19-0093
Page(s): R415–R439
Volume/Issue: Volume 26: Issue 7
Article Type: Review Article
Online Publication Date: Jul 2019
Copyright: © 2019 Society for Endocrinology 2019
Free access
Download PDF
Check for updates
Citation Alert Citation Alerts
Get Permissions
Abstract/Excerpt
Full Text
PDF
Supplementary Materials
Abstract
In recent years, cancer genomics has provided new insights into genetic alterations and signaling pathways involved in thyroid cancer. However, the picture of the molecular landscape is not yet complete. DNA methylation, the most widely studied epigenetic mechanism, is altered in thyroid cancer. Recent technological advances have allowed the identification of novel differentially methylated regions, methylation signatures and potential biomarkers. However, despite recent progress in cataloging methylation alterations in thyroid cancer, many questions remain unanswered. The aim of this review is to comprehensively examine the current knowledge on DNA methylation in thyroid cancer and discuss its potential clinical applications. After providing a general overview of DNA methylation and its dysregulation in cancer, we carefully describe the aberrant methylation changes in thyroid cancer and relate them to methylation patterns, global hypomethylation and gene-specific alterations. We hope this review helps to accelerate the use of the diagnostic, prognostic and therapeutic potential of DNA methylation for the benefit of thyroid cancer patients.

Introduction
Thyroid cancer, the most prevalent endocrine malignancy, covers the full range of phenotypes from indolent to the worst forms of human cancer. It is categorized into differentiated thyroid cancer (DTC), poorly differentiated thyroid cancer (PDTC) and undifferentiated or anaplastic thyroid cancer (ATC), all of which are derived from thyroid follicular cells, and into medullary thyroid cancer (MTC), which is derived from parafollicular cells. Moreover, DTC has three basic subtypes: papillary thyroid cancer (PTC), follicular thyroid cancer (FTC) and Hurthle cell thyroid cancer (HCTC). Globally, DTC accounts for 95% of all thyroid carcinomas. During the last few decades, several epidemiologic studies have reported that DTC incidence has increased worldwide (Lim et al. 2017). The reasons for this are not clear; it has been attributed both to a true increase in the incidence of DTC and to the improvement and more extensive use of imaging techniques such as neck ultrasound (Kitahara et al. 2017).

Currently, total thyroidectomy, the removal of the affected neck lymph nodes of the central compartment, radioiodine (RAI) therapy for the ablation of thyroid remnants or metastases and TSH suppression with l-thyroxin are the treatment schedule for a large portion of DTC patients (Haugen et al. 2016). With these therapeutic approaches, the majority of DTC patients exhibit good prognosis with a >98% 5-year survival rate. However, a subset of tumors progress to display more aggressive behavior, and some of these undergo a progressive process of dedifferentiation that makes them less capable of producing thyroglobulin and concentrating iodine, producing a poor response to RAI. To identify patients with a progressive course of the disease, a number of prognostic factors and clinical scores have been proposed that are mainly age, the histological variant, the initial extent of the disease and the size of the primary tumor (Asa 2017). However, these prognostic factors have some limitations.

In recent years, the management of thyroid cancer in patients is shifting toward more personalized medicine to avoid the overdiagnosis and overtreatment of tumors with an indolent course and, at the same time, to identify those tumors that will progress (Dralle et al. 2015). The final goal is to deliver the most effective but least aggressive treatment. A better understanding of the molecular mechanisms underlying thyroid cancer progression may be key to tailor the management of this disease. In this regard, significant progress has been made in the last 20 years (Riesco-Eizaguirre & Santisteban 2016). The major event in PTC carcinogenesis is the constitutive activation of mitogen-activated protein kinase (MAPK), whereas the PI3K/AKT pathway is involved in the progression of FTC. Recently, the genetic landscape of some thyroid cancer histotypes has been largely deciphered (Cancer Genome Atlas Research Network 2014,  Kunstman et al. 2015), and some of these genetic alterations have been used both as diagnostic tools (in the study of thyroid nodules) and as prognostic tools. Importantly, for the first time, the last set of American Thyroid Association (ATA) guidelines recommended the use of mutations in the BRAF gene and the TERT promoter as prognostic factors in PTC (Haugen et al. 2016).

However, cancer is not only caused by genetic abnormalities but also by epigenetic alterations (reviewed in Jones & Baylin 2007). The most widely used definition for epigenetics is ‘the study of changes in gene function that are mitotically and/or meiotically heritable and that do not entail a change in DNA sequence’ (Wu & Morris 2001, Bird 2007). Epigenetics can explain how two identical genotypes can lead to different phenotypes. There are several epigenetic mechanisms: DNA methylation, posttranslational modifications of histones, chromatin remodeling or non-coding RNAs. These mechanisms have been reviewed elsewhere (Bannister & Kouzarides 2011,  Holoch & Moazed 2015, Längst & Manelyte 2015, Allis & Jenuwein 2016, Feinberg et al. 2016), and here we will focus on DNA methylation.

What is DNA methylation?
DNA methylation was the first discovered epigenetic modification (Hotchkiss 1948,  Holliday & Pugh 1975, Riggs 1975) and consists of the covalent addition of a methyl group to the 5-carbon of the cytosine, giving rise to 5-methylcytosine (5mC) (reviewed in  Portela & Esteller 2010). In humans, DNA methylation occurs almost exclusively within CpG dinucleotides, which are underrepresented (i.e., found in a lower than expected proportion based on the G/C content) and not evenly distributed throughout the genome (Bird 1980). Most human genome (approximately 60–80% of CpG sites) is methylated, except for some CpG-rich regions called CpG islands (CGIs), which are often unmethylated and encompass the promoters of approximately 60% of protein-coding genes (Ehrlich et al. 1982, Bird 1986, Lister et al. 2009).

DNA methylation is frequently described as a repressive epigenetic mark. However, DNA methylation function varies depending on the genomic context (reviewed in Jones 2012,  Baubec & Schubeler 2014) (Supplementary Fig. 1, see section on supplementary data given at the end of this article). DNA methylation in proximal and distal regulatory elements (i.e., promoters and enhancers, respectively) represses transcription by affecting the binding of transcription factors and/or recruiting enzymes that modify chromatin structure. Conversely, DNA methylation of the gene body may enhance transcriptional elongation and affect splicing. In the case of repetitive elements, which are densely methylated, DNA methylation is the major repression mechanism.

Therefore, DNA methylation is a key player in the regulation of gene expression and is implicated in many cellular processes such as imprinting (Reik et al. 1987, Swain et al. 1987), X-chromosome inactivation (Mohandas et al. 1981) and the establishment and maintenance of cell type-specific expression programs (reviewed in Suelves et al. 2016). DNA methylation is also essential for the maintenance of genome stability by modeling chromatin structure (reviewed in  Madakashira & Sadler 2017) as well as by silencing repetitive sequences to prevent chromosomal rearrangements (Gaudet et al. 2003) and the expression and expansion of transposable elements (reviewed in Belancio et al. 2010).

Furthermore, it is noteworthy that DNA methylation is an important source of promising cancer biomarkers for many reasons: DNA methylation is stable even in fixed samples over time, easily detected by well-established techniques (Fig. 1 and Supplementary Table 1), present in various bodily fluids, and cell type specific (Koch et al. 2018).

Figure 1
Download Figure
Download figure as PowerPoint slide
Figure 1
Main DNA methylation techniques according to the type of DNA methylation measured (global or sequence-specific) and the principle of DNA methylation discrimination (physicochemical properties, 5mC antibody affinity, methylation-sensitive restriction enzymes and bisulfite conversion) (more detailed information in Supplementary Table 1). In recent decades, a large number of techniques to measure DNA methylation have been developed that can be classified into two main groups: (i) those providing unique value as a global measure of DNA methylation and (ii) those that are sequence-specific and measure the levels of DNA methylation in particular regions at single CpG resolution. Moreover, global methods can be subdivided into those measuring the DNA methylation of the entire genome and those measuring the DNA methylation of a compartment of the genome used as surrogate reporter of the genome (e.g., repeat sequences such as LINE-1 and Alu elements, which comprise 20% and 10% of the human genome, respectively). Sequence-specific methods can also be subdivided into those that are genome-wide (mostly based on bead arrays or NGS) and those measuring specific regions of interest (mostly based on PCR). A more comprehensive list of available techniques can be found elsewhere (Jordà et al. 2009,  Toraño et al. 2012). Importantly, techniques to analyze DNA methylation do not differentiate 5mC from 5hmC; thus, in recent years, some approaches to specifically measure 5hmC have been developed (Skvortsova et al. 2017). AIMS, analysis of DNA methylation by amplification of intermethylated sites; AuNPs, Au nanoparticles; BS, bisulfite; COBRA, combined bisulfite restriction analysis; ELISA, enzyme-linked immunosorbent assay; HPCE, high-performance capillary electrophoresis; IHC, immunohistochemistry; LC-MS, liquid chromatography coupled with mass spectrometry; LUMA, luminometric methylation assay; MALDI-TOF-MS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry; MCA, methylated CpG island amplification; MeDIP, methylated DNA immunoprecipitation; MSP, methyl-sensitive PCR; MS-AFLP, methylation-sensitive amplification length polymorphism; NGS, next-generation sequencing; NSUMA, next-generation sequencing of unmethylated Alu; RE, restriction enzyme; RP-HPLC, reversed-phase high-performance liquid chromatography; RRBS, reduced representation bisulfite sequencing; QUAlu, quantification of unmethylated Alu; WGBS, whole genome bisulfite sequencing.

Citation: Endocrine-Related Cancer 26, 7; 10.1530/ERC-19-0093

Writers, erasers and readers of DNA methylation
DNA methylation does not function alone but is involved in a complex crosstalk with many other players to reinforce specific regulatory programs. Although how DNA methylation is interpreted in the context of genome regulation is not completely understood, there are some proteins known to modulate DNA methylation. They are classified as writers, erasers and readers of DNA methylation.

DNA methylation writers are proteins that establish and maintain DNA methylation patterns through development and differentiation. These proteins, called DNA methyltransferases (DNMTs), transfer a methyl group to cytosine residues (reviewed in Goll & Bestor 2005) (Fig. 2).

Figure 2
Download Figure
Download figure as PowerPoint slide
Figure 2
Process of DNA methylation and demethylation. DNA methyltransferases (DNMTs) catalyze the methylation of cytosine by adding a methyl group to C5 position. DNMT3A and DNMT3B are responsible for de novo methylation, while DNMT1 maintains DNA methylation patterns by copying the 5mC pattern on the newly synthesized strand after DNA replication. DNA demethylation can occur through different mechanisms: passive demethylation due to the impairment of the DNA methylation maintenance machinery that results in the dilution of DNA methylation after multiple rounds of replication, and active demethylation involving several proteins considered to be erasers of DNA methylation. Specifically, active demethylation occurs through the iterative oxidation of 5mC to 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC) mediated by ten-eleven-translocation (TET) proteins. Then, these oxidized forms can be subsequently diluted during DNA replication, or 5fC and 5caC can be excised by thymine DNA glycosylases (TDG) coupled with base excision repair (BER) (reviewed in Wu & Zhang 2017). Even though other mechanisms of active demethylation have been proposed, the TDG-BER mechanism has gained the most support (Wu & Zhang 2010, Bochtler et al. 2017).

Citation: Endocrine-Related Cancer 26, 7; 10.1530/ERC-19-0093

Despite the high stability of DNA methylation, 5mC can be demethylated by passive or active mechanisms, the latter mediated by erasers that generate DNA demethylation intermediates, such as 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC) (reviewed in Wu & Zhang 2017) (Fig. 2). Interestingly, although the levels of 5hmC in human cells are very low (in general, 14-fold lower than the levels of 5mC) and vary greatly between different tissues, 5hmC is present at a relatively stable abundance, suggesting that it is not just a DNA demethylation intermediate. 5hmC is associated with gene transcription, although this relationship is not fully understood (reviewed in Shi et al. 2017) (Supplementary Fig. 1).

Finally, DNA methylation readers are proteins that specifically bind to methylated CpGs and coordinate the crosstalk between DNA methylation, histone modifications and chromatin organization to reinforce downstream regulatory programs. The paradigm of these proteins are the methyl-CpG-binding domain (MBD) family proteins, which have the ability to recruit chromatin remodelers, histone deacetylases (HDAC) and DNMTs to methylated CpGs associated with gene repression (reviewed in Du et al. 2015) (Supplementary Fig. 1).

Aberrant DNA methylation is a hallmark of cancer
Disruption of DNA methylation is a common feature in human disease, both in noncancerous diseases and in cancer (Fernandez et al. 2012). Although the first discovered alteration of DNA methylation in cancer was an overall reduction in 5mC levels, that is, a global hypomethylation in tumoral cells compared to the methylation in normal cells (Feinberg & Vogelstein 1983,  Gama-Sosa et al. 1983), this epigenetic alteration has been ignored for decades mainly due to the technical complexity of its analysis. Conversely, research has focused on hypermethylation (i.e., the increase in the methylation of CpG sites compared to normal cells), which often but not exclusively occurs in localized sequences within regulatory elements associated with CGIs (Herman et al. 1995, Melki et al. 1999, Esteller et al. 2000, Nguyen et al. 2001). Both global hypomethylation and focal hypermethylations are constant features of the cancer genome and often coexist in tumoral cells, but although there is interplay between them, their underlying mechanisms seem to be independent.

The main consequence of the hypermethylation of promoters and enhancers is the repression of the expression of genes functionally important in the neoplastic process, whose silencing may have a tumor-promoting effect. The hypermethylation profile is tumor specific and affects all cellular pathways (reviewed in Esteller 2007). Some genes such as p16INK4A and MLH1 are frequently hypermethylated in many cancers including thyroid cancer (Schagdarsurengin et al. 2006, Guan et al. 2008) while others are tumor specific (e.g., the sodium iodide symporter gene – SLC5A5 or NIS – in thyroid cancer) (Neumann et al. 2004,  Galrão et al. 2014). In contrast to global hypomethylation, an overall increase in 5mC levels in tumors compared to its levels in normal tissues is much less common, despite the high number of local hypermethylations (Ehrlich 2002).

For several decades, it has been widely accepted that the hypomethylation of repetitive sequences is responsible for global hypomethylation (Ehrlich 2009). However, recent approaches enabling the mapping of DNA methylation at the genome scale have shown that global hypomethylation affects large genome domains including both repetitive and unique sequences (Berman et al. 2011, Hansen et al. 2011, Timp et al. 2014). In this context, hypomethylation can encompass regulatory elements and affect gene expression (reviewed in Wilson et al. 2007). Nevertheless, hypomethylation-dependent transcriptional activation is less frequent than hypermethylation-dependent transcriptional silencing. In contrast, numerous studies indicate that global hypomethylation is associated with chromosomal instability and the reactivation of transposable elements (Gaudet et al. 2003).

5hmC is also perturbed in cancer in a similar way as 5mC, i.e., there is a strong global loss of this epigenetic mark in tumors. However, its involvement in thyroid cancer is completely unknown. A recent study that analyzed 5hmC in circulating cell-free DNA and in tumoral and normal tissues from different cancer types, including thyroid cancer, found that 5hmC was mainly distributed in transcriptionally active regions (Li et al. 2017). Importantly, they identified cancer-specific 5hmC signatures. To the best of our knowledge, this is the only study of 5hmC in thyroid cancer; thus, 5hmC has opened a new field to explore in this disease.

Disruption of epigenetic pathways in cancer
The recent whole exome sequencing of thousands of tumoral samples of different cancer types has revealed that many genes controlling the epigenome are mutated, which can lead to epigenetic aberrations (reviewed in You & Jones 2012). This is the case for mutations in the writers, erasers and readers of DNA methylation. Mutations in the DNMT and TET genes have been identified in different cancers; for example, DNMT3A and TET2 are frequently mutated in hematologic malignancies. In thyroid cancer, mutations in these genes are rare (<1.5% in PTC and <3% in ATC and PDTC) (data from The Cancer Genome Atlas Research Network;  http://www.cbioportal.org/) (Cerami et al. 2012). However, the expression of some of these genes is altered (Supplementary Fig. 2) and could contribute to the dysregulation of DNA methylation in thyroid cancer, although further studies should be performed to understand the underlying relationship. On the other hand, MBD proteins are mutated in several cancers, including PTC (Du et al. 2015). Although they represent less than 5% of patients, the study of these proteins in thyroid cancer could be a promising field.

DNA methylation changes in thyroid cancer: drivers of disease progression and biomarkers
DNA methylation has been extensively studied in many cancers, such as colorectal and breast cancer, and due to technological advances, enormous progress has been made in the understanding of the epigenetic landscape of these tumors. Conversely, the role of DNA methylation in thyroid cancer has received comparatively less attention (Fig. 3). The first DNA methylation studies in thyroid cancer were based on candidate gene approaches assessing the DNA methylation levels of specific gene promoters (Supplementary Table 2). It was not until 2011 that Hou et al. performed the first array-based, genome-wide DNA methylation study using two PTC cell lines to analyze the effect of the BRAF(V600E) mutation on DNA methylation (Hou et al. 2011). To our knowledge, 11 more array-based studies using different platforms (Goldengate, 27K or 450K) to profile the methylomes of tissue samples from patients with thyroid cancer have been published since then (Table 1). All these studies showed that thyroid cancer is not an exception and exhibits DNA methylation alterations. However, different pan-cancer analyses based on data from the Cancer Genome Atlas Research Network revealed that PTC has one of the lowest frequency of DNA methylation alterations. Specifically, Yang et al. performed a differential DNA methylation analysis between normal and tumoral samples (n = 5480) for 15 cancer types showing high variability in the numbers of differentially methylated CpGs, which ranged from 3722 in PTC to 57,290 in uterine corpus endometrial carcinoma (Yang et al. 2016a). Accordingly, another DNA methylation pan-cancer study focused on promoters found that PTC exhibited one of the lowest frequencies in both hypomethylation and hypermethylation events (Saghafinia et al. 2018) (Fig. 4). In addition, these authors introduced the concept of DNA methylation instability, which was found to be very low in PTC. In contrast, ATC exhibits a high frequency of DNA methylation alterations (10-fold higher than PTC; Bisarro dos Reis et al. 2017).

Figure 3
Download Figure
Download figure as PowerPoint slide
Figure 3
Comparison of the overall number of science citation indexed publications in the field of DNA methylation in different cancer types over the last 25 years (1990 to 2017). A search was carried out on the Web of Science (Thompson Reuters) on Nov 2018 using a date-restricted search (1990–2017) and ‘DNA methylation’ AND ‘colorectal cancer’ OR ‘colon cancer’, ‘DNA methylation’ AND ‘breast cancer’, ‘DNA methylation’ AND ‘lung cancer’, ‘DNA methylation’ AND ‘prostate cancer’ or ‘DNA methylation’ AND ‘thyroid cancer’ as search terms. BRCA, breast cancer; CRC, colorectal cancer; LUAD, lung cancer; PRAD, prostate cancer PRAD; THCA, thyroid cancer.

Citation: Endocrine-Related Cancer 26, 7; 10.1530/ERC-19-0093

Figure 4
Download Figure
Download figure as PowerPoint slide
Figure 4
Epigenetic and genetic alterations by tumor type. (A) Hypermethylation and (B) hypomethylation event frequencies in 6010 human tumors across 24 cancer types. Frequencies are estimated as the percentage of probes found hypermethylated (out of 64,414) or hypomethylated (out of 3423) compared to normal tissues. (C) Somatic mutation frequencies in 3025 tumor-normal sample pairs across 27 cancer types. Tumor types are sorted by their median hypermethylation, hypomethylation and somatic mutation frequencies. Papillary thyroid tumors (red bars) are among tumors with the lowest frequency of epigenetic and genetic alterations, mostly leukemias and pediatric cancers. The x-axis gives the number of samples for each tumor type. Data from Lawrence et al. (2013) and  Saghafinia et al. (2018). ACC, adrenocortical carcinoma; AML, acute myeloid leukemia; BLCA, bladder carcinoma; BRCA.B, basal breast invasive carcinoma; BRCA.L, luminal breast invasive carcinoma; Car, carcinoid tumors; CESC, cervix squamous cell carcinoma; CLL, chronic lymphocytic leukemia; CRC, colon and rectum carcinoma; DLBCL, diffuse large B-cell lymphoma; ESCA, esophageal carcinoma; EwngSRC, Ewing sarcoma; GBM, glioblastoma multiforme; HNSC, head and neck squamous cell carcinoma; KICH, kidney chromophobe carcinoma; KIRC, kidney renal clear cell carcinoma; KIRP, kidney renal papillary cell carcinoma; LGG, low grade glioma; LIHC, liver hepatocellular carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; MD, medulloblastoma; NB, neuroblastoma; MM, multiple myeloma; OV, ovarian carcinoma; PAAD, pancreatic adenocarcinoma; PRAD, prostate adenocarcinoma; PTC, papillary thyroid cancer; RhD, Rhabdoid tumor; SKCM, skin cutaneous melanoma; STAD, stomach adenocarcinoma; UCEC, uterine corpus endometrial carcinoma; UCS, uterine carcinosarcoma.

Citation: Endocrine-Related Cancer 26, 7; 10.1530/ERC-19-0093

Table 1
Summary of studies in thyroid cancer using methods for genome-wide analysis of DNA methylation.

Study No. Method Discovery series (n) Hypera Hypob Integration expression Identified genes of interest References
1 MCA/CGI array Cell lines (2) NA NA RT-qPCR HMGB2, FGD1 Hou et al. (2011)
2 GoldenGate NT (25); TC (36) NA NA – Hansen et al. (2011)
3 27K NT (10); PTC (14) NA NA RT-qPCR HIST1H3J, POU4F2, SHOX2, PHKG2, TLX3, HOXA7 Kikuchi et al. (2013)
4 27K NT (2) GEO array data ADAMTS8, HOXB4, ZIC1, KISS1R, INSL4, DPPA2, TCL1B, NOTCH4, MAP17 Rodríguez-Rodero et al. (2013)
PTC (2) 309 14
FTC (2) 408 24
ATC (2) 114 174
cell lines (4)
5 27K NT (8) GEO array data COL4A2, DLEC1, KLK10, EI24, WT1 Mancikova et al. (2014)
FA (18) 89 9
PTC (42), fvPTC (5) 39 53
FTC (18) 460 83
6 450K NT (8) – Ellis et al. (2014)
PTC (29) 255 2582
fvPTC (15) 164 405
recurrent PTC (7) 1023 2796
7 450K NT (56); cPTC (324); fvPTC (99); tcPTC (35); Other PTC (38) NA NA RNA-Seq; miRNA-Seq Cancer Genome Atlas Research Network (2014)
8 450K NT (12) – Timp et al. (2014)
DTC (24) 24130 19857
9 450K NT (16) – CDKN1B, mir-146, PDGF, SERPINA1, TGFB1, TPO, DUSP5, ERBB3, FGF1, FGFR2, GABRB2, HMGA2 White et al. (2016)
PTC (13) 165 1061
10 450K NT (41) GEO array data ERBB3, FGF1, FGFR2, GABRB2, HMGA2, RDH5 Beltrami et al. (2017)
PTC (41) 645 5425
11 450K NT (50) – PFKFB2, LAIR2, THSD7B, OR52B2, OR2T6, FFAR2, RTN3, DCD, ADGRE2, HRH1, GPR21, MBP, YPEL4, ATP6V0C Bisarro dos Reis et al. (2017)
BTLs (17) 1531 222
PTC (60) 242 2773
FTC (8); HCTC (2) 4100 1475
PDTC (1); ATC (3) 6195 28252 Affinito et al. (2019)
12 450K NT (7) RNA-Seq
FA (10) 0 0
FTC (11) 2979 585
13 RRBS NT (3) RNA-Seq CXCL12, FBLN7, FAM3B, PROX1, COL23A1, GJB3, LAD1 Zhang et al. (2017)
PTC (3) 639 907
14 RRBS NT (39) – Diagnostic DNA methylation signature (DDMS) Yim et al. (2019)
BTLs (28) NA NA
PTC (39) NA NA
aHyper, number of hypermethylation events compared to normal tissue; bHypo, number of hypomethylation events compared to normal tissue.

27K, Infinium HumanMethylation27 BeadChip; 450K, Infinium HumanMethylation450 BeadChip; ATC, anaplastic TC; BTLs, benign thyroid lesions; cPTC, classical PTC; DTC, differentiated TC; FA, follicular adenoma; FTC, follicular TC; fvPTC, follicular variant of PTC; GEO array; Gene Expression Omnibus; MCA/CGI, methylated CpG island amplification/CpG island; NA, not available; HCTC, Hurthle cell TC; NT, normal tissue; PDTC, poorly differentiated TC; PTC, papillary TC; RRBS, reduced representation bisulfite sequencing; TC, thyroid cancer, tcPTC, tall cell PTC.

Interestingly, these epigenetic differences between PTC and ATC also can be found at the level of genetic alterations. A recent pan-cancer analysis on whole exome sequencing revealed that the mutation frequency in PTC was one of the lowest (approximately 1 change/Mb across the entire exome) among solid tumors (Lawrence et al. 2013) (Fig. 4), while the mutation frequency in ATC was at the opposite extreme and was closer to that in melanoma and lung cancer, exceeding 100 changes/Mb (Kunstman et al. 2015, Riesco-Eizaguirre & Santisteban 2016). As mutations are largely caused by errors in DNA replication (Tomasetti et al. 2017), some researchers propose that the cell division rate also participates in shaping the cancer DNA methylation landscape (Yang et al. 2016b). Thus, the different proliferation rate between PTC (low) and ATC (high) could explain, at least in part, the different frequencies of their DNA methylation alterations.

Thyroid cancer DNA methylation patterns
Since the initial studies assessing the DNA methylation of a single locus, there has been accelerating technological progress providing a plethora of DNA methylation techniques that have allowed the generation of single CpG resolution maps (Fig. 1 and Supplementary Table 1). These maps have improved our understanding of DNA methylation and have shown that DNA methylation patterns, the so-called methylomes, are tissue specific, allowing us to distinguish different normal tissues from each other (Hansen et al. 2011, Fernandez et al. 2012). Moreover, methylomes differ largely between normal and tumoral cells and between different types of tumors, which is key from a translational point of view. An example of the clinical use of this specificity is that methylomes allow the identification of the tissue of origin in carcinomas of unknown primary origin (CUPs) (Moran et al. 2016).

Association between methylomes and histology in thyroid cancer
Genome-wide studies to profile thyroid cancer methylomes, most of which used BeadArrays (Table 1), revealed histology-associated DNA methylation profiles. Specifically, PTC is characterized by a higher number of hypomethylations (most of them outside promoter regions) than hypermethylations in comparison to normal thyroid tissues (Ellis et al. 2014,  Mancikova et al. 2014, White et al. 2016, Beltrami et al. 2017, Bisarro dos Reis et al. 2017) (Supplementary Fig. 3). Only the study by Rodríguez-Rodero et al. identified more hypermethylations than hypomethylations, which could be explained by the low number of analyzed PTCs (Rodríguez-Rodero et al. 2013). In contrast to PTC, FTC exhibits more hypermethylations than hypomethylations (most of them outside promoter regions) (Rodríguez-Rodero et al. 2013, Mancikova et al. 2014, Bisarro dos Reis et al. 2017,  Affinito et al. 2019) as well as follicular adenomas (FA), although the number of DNA alterations in these benign tumors is low, thus resembling normal thyroid methylomes (Supplementary Fig. 3). There is debate within the field about whether FA and FTC are distinct molecular entities or represent a biological continuum (Arora et al. 2008, Krause et al. 2011, Yoo et al. 2016). Interestingly, Mancikova et al. showed that most of the FA-associated promoter hypermethylations that they identified were also found in FTC, suggesting a progressive gain of hypermethylations along the tumorigenic process from adenomas to carcinomas, thereby reinforcing the hypothesis that some FAs have the malignant potential to give rise to FTC (Mancikova et al. 2014). Accordingly, unsupervised clustering analysis in the study by Bisarro dos Reis et al. showed that FAs clustered with FTCs, and a recent study by Affinito et al. found that FAs displayed an intermediate DNA methylation profile between FTCs and normal thyroid tissues (Bisarro dos Reis et al. 2017, Affinito et al. 2019).

The majority of genome-wide studies are focused on PTC. Interestingly, some of them specify the PTC variants used, revealing differential DNA methylation profiles. Ellis et al. found that classical PTC (cPTC) displayed a high number of DNA alterations, most of which were hypomethylations, whereas follicular variant of PTC (fvPTC) exhibited a smaller proportion of hypomethylations (Ellis et al. 2014). In this regard, this study, as well as those by Mancikova et al. and the Cancer Genome Atlas Research Network, found that fvPTC exhibited a methylome that was not as different from that of normal thyroid tissue (Mancikova et al. 2014,  Cancer Genome Atlas Research Network 2014). Apart from cPTC and fvPTC, the Cancer Genome Atlas Research Network also analyzed tall cell PTC (tcPTC). Based on an unsupervised clustering analysis, they classified tumors into four groups: two groups enriched by fvPTC (Meth-follicular, which exhibited few methylation changes compared to normal tissue and Meth-CGI, which was characterized by the hypermethylation of numerous CGIs) and two groups enriched by cPTC and tcPTC (Meth-classical 1 and Meth-classical 2, which were characterized by hypomethylations outside of CGIs). Interestingly, a small subset of fvPTCs resembled tcPTC and cPTC. Mancikova et al., who also included FA and FTC in the study, showed that fvPTC methylomes were more similar to follicular tumors than to cPTC (Mancikova et al. 2014). These results indicate that PTCs with follicular architecture are different from PTCs with papillary architecture. In the future, it will be of great interest to investigate whether the new noninvasive follicular neoplasm with papillary-like nuclear features (NIFTP) entity shows a specific methylation profile.

Two of the array-based studies included several PDTC and ATC samples (Table 1 and Supplementary Fig. 3). While the study by Rodríguez-Rodero et al. identified fewer DNA methylation alterations in ATC than in DTC, Bisarro dos Reis et al. identified a drastically higher number of DNA methylation alterations in PDTC and ATC (six-fold higher than in FTC and ten-fold higher than in PTC) (Rodríguez-Rodero et al. 2013, Bisarro dos Reis et al. 2017). The contradictory results of these two studies are probably due to the low number of analyzed samples and the use of different arrays (27K vs 450K platforms) that cover different regions of the genome (Rodríguez-Rodero et al. 2013, Bisarro dos Reis et al. 2017). However, both studies concluded that PDTC and ATC exhibited more hypomethylation than hypermethylation events, suggesting the association of hypomethylation with dedifferentiation. The similarity between the methylomes of ATC, PDTC, extensively invasive FTC and lymphocytic thyroiditis found by Bisarro dos Reis et al. is noteworthy (Bisarro dos Reis et al. 2017); as these aggressive tumors are characterized by a high level of immune cell infiltration (Ryder et al. 2008), these results suggest that part of the DNA methylation alterations in these tumors may come from infiltrating immune cells.

BeadArrays are widely used in genome-wide DNA methylation studies due to their low cost, but they are limited to the CpGs covered by the array. However, reduced representation bisulfite sequencing (RRBS), which is based on next-generation sequencing, has a higher sensitivity, resolution and coverage than BeadArrays. There are two studies that investigated PTC methylomes using RRBS (Table 1). Zhang et al. focused on the methylation of mRNA and lncRNA promoters and confirmed previous results showing more hypomethylation than hypermethylation events in PTC (Zhang et al. 2017). However, when DNA methylation and expression data from the same samples were crossed, only 19 mRNAs were upregulated/hypomethylated, and 26 mRNAs and 3 lncRNAs were downregulated/hypermethylated. These findings, which are consistent with results from Mancikova et al. and Affinito et al. (Mancikova et al. 2014, Affinito et al. 2019), suggested that DNA methylation in promoters does not have a widespread role in controlling gene expression in PTC. On the other hand, by using RRBS, Yim et al. identified a unique DNA methylation signature of 4,575 CpGs specific to benign nodules, most of which were hypermethylated compared to adjacent normal tissues and malignant nodules, that may have important diagnostic applications (Yim et al. 2019).They also analyzed specimens with lymphocytic thyroiditis and, in accordance with the results from Bisarro dos Reis et al., suggested that the presence of immune-infiltrating cells in tumors may affect DNA methylation patterns (Bisarro dos Reis et al. 2017).

Globally, most DNA methylation alterations in thyroid cancer occur outside promoter regions and are specifically associated with histology.

Association between DNA methylation and genetic drivers in thyroid cancer
Another important finding derived from these studies is the relationship between DNA methylation profiles and mutations in BRAF and RAS genes; BRAF-mutated tumors harbor more hypomethylations (which is expected since this mutation is almost exclusively detected in PTC), while RAS-mutated tumors harbor more hypermethylations (which is expected since this mutation mostly occurs in fvPTC and FTC). These observations were confirmed by the pan-cancer study by Saghafinia et al. who found a significant association between NRAS mutation and hypermethylation events and between BRAF mutation and hypomethylation events in thyroid cancer (Saghafinia et al. 2018). Interestingly, although RAS mutations are common in many types of tumors such as lung adenocarcinoma and prostate cancer this genetic–epigenetic relationship was not detected in other cancers or even in melanomas that mostly harbor NRAS mutations such as thyroid cancer. Therefore, an association between DNA hypermethylation and a specific RAS isoform could be discarded. The BRAF(V600E) mutation is also frequent in colorectal cancer and melanoma, but there is no association between this mutation and hypomethylation events. Conversely, BRAF mutation is strongly associated with hypermethylation events in colorectal cancer (Weisenberger et al. 2006, Cancer Genome Atlas Research Network 2012,  Saghafinia et al. 2018). Specifically, in colorectal cancer, BRAF(V600E) has been associated with the so-called ‘CpG island methylator phenotype’ (CIMP) (including tumors that exhibit an exceptionally high frequency of the hypermethylation of some CGIs) (Toyota et al. 1999). From studies done in melanoma, there are some controversial results, but most studies do not find any significant association between BRAF mutation and hyper- or hypomethylations (Lauss et al. 2015, Saghafinia et al. 2018).

Altogether, these findings show a cancer type-specific relationship between BRAF or RAS mutations and aberrant DNA methylation. However, whether these events are dependent on one another requires further studies in controlled experimental systems (e.g., cell lines, mouse models). In this regard, in BRAF-mutated colorectal tumors, it has been reported that MAFG mediates hypermethylation by binding to target gene promoters and recruiting a corepressor complex that includes DNMT3B (Fang et al. 2014). In thyroid cancer, there are no studies focusing on this genetic–epigenetic relationship. Hou et al. knocked down BRAF in two PTC-derived cell lines by shRNA and found numerous hypermethylated and underexpressed genes, suggesting that these genes were hypomethylated and overexpressed in the presence of BRAF(V600E) and thus pointing out a causal relation (Hou et al. 2011). However, further investigation is required.

DNA methylation and the BRAF-like and RAS-like phenotypes
Together, these results highlight the association of DNA methylation profiles with histology and mutations in the BRAF and RAS genes, at least for DTC. However, the relationship between histology, genotype and methylome does not fit perfectly (Fig. 5). For example, fvPTCs harbor more hypermethylations than hypomethylations, except for a subset of fvPTCs that harbor more hypomethylations than hypermethylations; some PTCs that do not contain mutations in BRAF or RAS exhibit more hypermethylations than hypomethylations, while others exhibit more hypomethylations than hypermethylations. This can be resolved with the BRAF-like and RAS-like phenotypes defined by the Cancer Genome Atlas Research Network (Cancer Genome Atlas Research Network 2014). The Cancer Genome Atlas Research Network developed a scoring system based on the expression of 71 genes that classifies PTCs into two groups called BRAF-like and RAS-like tumors depending on whether their gene expression profile more closely resembles BRAF-mutated tumors or RAS-mutated tumors. The Cancer Genome Atlas Research Network shows that these two groups of tumors are different at the genetic and epigenetic levels (Fig. 5), resulting in a different expression program that activates different pathways. BRAF-like tumors are characterized by the overactivation of the MAPK/ERK pathway (preferentially via BRAF), while RAS-like tumors exhibit concurrent activation of the PI3K/AKT and MAPK/ERK pathways (the latter of which is activated at a lower level than that in BRAF-like tumors and preferentially via RAF1 (also known as C-Raf)). In the previous examples, if we take into account the BRAF-like and RAS-like phenotypes, we can see that most fvPTCs are RAS-like and harbor more hypermethylations than hypomethylations, but the subset of fvPTC that is BRAF-like (with or without BRAF mutation) exhibits more hypomethylations than hypermethylations; those tumors that do not contain mutations in BRAF or RAS exhibiting more hypomethylations than hypermethylations are BRAF-like while those exhibiting more hypermethylations than hypomethylations are RAS-like. This is in agreement with the study from Chen et al., although they do not specifically use the BRAF-like and RAS-like terms (Chen et al. 2017). These findings suggest that hypo- and hypermethylation events may be downstream of the pathways overactivated in the BRAF-like and RAS-like phenotypes, respectively. Further investigation should be conducted in this area.

Figure 5
Download Figure
Download figure as PowerPoint slide
Figure 5
Genetic and epigenetic alterations associated with DTC according to histology, BRAF and RAS mutational state and BRAF-like and RAS-like phenotypes. 1Genetic alterations were considered highly recurrent when present in >5% tumors based on Cancer Genome Atlas Research Network data. 2Genetic alterations were considered lowly recurrent when present in <5% tumors based on Cancer Genome Atlas Research Network data. 3Hyper, more hypermethylations than hypomethylations; hypo, more hypomethylations than hypermethylations. 4FTC data from Yoo et al. (2016). cPTC, classical PTC; FTC, follicular thyroid cancer; tcPTC, tall cell PTC fvPTC, follicular variant of PTC.

Citation: Endocrine-Related Cancer 26, 7; 10.1530/ERC-19-0093

Global DNA hypomethylation in thyroid cancer
A wide range of techniques to analyze global DNA methylation have been developed (Fig. 1 and Supplementary Table 1) (reviewed in Jordà & Peinado 2010, Toraño et al. 2012), but there are some key points to take into account when interpreting results that are summarized in the Supplementary material. As explained above, global DNA hypomethylation is a common epigenetic feature of cancer. Interestingly, in many cancer types, the degree of global DNA hypomethylation is strongly associated with the tumor grade and stage, which has attracted great interest for its potential clinical use. Nevertheless, little is known about global DNA hypomethylation in thyroid cancer. As shown in Table 2, as far as we know, a total of nine studies have assessed global DNA methylation levels in thyroid tumors and report conflicting results that may be partly explained by the low number of samples included in some studies and the different methods used.

Table 2
Summary of studies in thyroid cancer analyzing global DNA methylation.

Study No. Method Global methylationa Discovery series (n)b Result References
1 COBRA-LINE1 Entire genome NT (7); PTC (7) No differences Chalitchagorn et al. (2004)
2 IHC with 5mC antibody Entire genome NT (9);NG (1); PTC (3); fvPTC (1); FTC (2) Global hypomethylation in tumors de Capoa et al. (2004)
3 IHC with 5mC antibody Entire genome NT (17);NG (19); HCA (10); FA (16); PTC (17); FTC (6) Global hypomethylation in PTC and FTC Galusca et al. (2005)
4 LINE-1 pyrosequencing/LUMA Compartment NT (21); FTC (21) No differences Lee et al. (2008)
5 ELISA with 5mC antibody Entire genome NT (10); NG (24) No differences Brown et al. (2014)
6 COBRA-LINE1/IHC with 5mC antibody Compartment/entire genome NT (50); FA (15); PTC (17); FTC (18) No differences in LINE-1 methylation but global hypermethylation in tumors Keelawat et al. (2015)
7 QUAlu Compartment NT (9); cPTC (31); FTC (14) Global Alu hypomethylation in PTC and FTC Buj et al. (2016)
8 QUAlu Compartment NT (20); PTC (40); FTC (21); PDTC (7); ATC (9); M1 (24); pediatric PTC (13) Global Alu hypomethylation in distant metastatic PTC and distant metastatic FTC as well as the paired M1, and in PDTC and ATC Klein Hesselink et al. (2018)
9 ELISA with 5mC antibody Entire genome Blood: controls (6); PTC (12) No differences Ceolin et al. (2018)
aPart of the genome in which global DNA methylation has been analyzed. bDNA methylation was assessed in postsurgical tissue unless otherwise stated.

5mC, 5-methylcytosine; ATC, anaplastic thyroid cancer; COBRA, combined bisulfite restriction analysis; cPTC, classical PTC; FA, follicular adenoma; fvPTC, follicular variant of PTC; FTC, follicular thyroid cancer; IHC, immunohistochemistry; HC, Hurthle cell adenoma; LUMA, luminometric methylation assay; M1, distant metastasis; NG, nodular goiter; NT, normal tissue; PDTC, poorly differentiated thyroid cancer; PTC, papillary thyroid cancer; QUAlu, quantification of unmethylated Alu.

Five of the studies used techniques based on repetitive sequences. They revealed that LINE-1 elements of normal and tumoral samples did not show different levels of DNA methylation (Chalitchagorn et al. 2004, Lee et al. 2008, Keelawat et al. 2015), while Alu elements were slightly hypomethylated in PTC and FTC (Buj et al. 2016). A deeper study of the hypomethylation of Alu elements showed that it occurred in distant metastatic DTC, PDTC and ATC but not in low-risk DTC and pediatric PTC (Klein Hesselink et al. 2018), suggesting the involvement of global hypomethylation of Alu elements in thyroid cancer progression and dedifferentiation. This is in agreement with studies in other cancer types, such as hepatocellular carcinoma or cervical cancer, in which global hypomethylation correlates with disease progression (Lin et al. 2001,  Yegnasubramanian et al. 2008). These findings in thyroid cancer may have important prognostic applications, especially in preoperative fine-needle aspiration biopsies (FNAB), which would help in treatment planning. The differences between the results about LINE-1 and Alu elements could be explained by the fact that the studies analyzing LINE-1 elements included FA, PTC and FTC but did not include aggressive tumors. We cannot discard the use of different techniques with different sensitivity and accuracy as being responsible for the different results. On the other hand, as explained in the Supplementary material, these apparently opposite findings are biologically plausible. In this regard, the different methylation between LINE-1 and Alu elements also occurs in other cancers (Benard et al. 2013, Park et al. 2014).

Another interesting result from these studies was that while global methylation of LINE-1 elements varied between different normal tissues, especially in the normal thyroid, all normal tissues displayed similar levels of unmethylated Alu elements (Chalitchagorn et al. 2004, Buj et al. 2016). Conversely, tumors showed a broad variation of the DNA methylation of both LINE-1 and Alu elements. For example, colon and lung cancer exhibited 2- to 3-fold higher levels of unmethylated Alu elements than thyroid cancer. On the other hand, the Alu hypomethylation was similar between distant metastases and matched primary tumors, suggesting that Alu methylation remained stable during metastatic spread in thyroid cancer (Klein Hesselink et al. 2018).

Four more studies used antibodies that recognize 5mC to evaluate global hypomethylation in thyroid cancer, and three of them were based on immunohistochemistry while one used ELISA. The analyses of benign lesions (hyperplasia, FA and Hurthle adenoma) did not find differences between benign lesions and normal thyroid tissues (Galusca et al. 2005, Brown et al. 2014,  Keelawat et al. 2015), except for de Capoa et al., but they only analyzed one sample (de Capoa et al. 2004). Thus, global DNA hypomethylation in thyroid cancer does not seem to be an early event as described in other cancer types (Ehrlich 2009). The results on global DNA hypomethylation in DTC were more variable. de Capoa et al. and Galusca et al. showed global hypomethylation in malignant tumors compared to normal tissues (de Capoa et al. 2004,  Galusca et al. 2005). In contrast, Keelawat et al. did not find global hypomethylation but rather found global hypermethylation (Keelawat et al. 2015). These inconsistent results are probably due to technical issues that mainly include the use of different antibodies with different sensitivities. In this regard, the two studies showing global hypomethylation used the same antibody. Although further investigation is needed, these results suggest the diagnostic potential of global DNA methylation measured by specific antibodies. Moreover, the fact that benign lesions do not display global alterations is very promising, especially for thyroid nodules with indeterminate cytology.

Only one of the studies analyzed the relationship between global hypomethylation and mutations in BRAF and RAS, and it found an association in distant metastatic DTC but not in low-risk DTC (Klein Hesselink et al. 2018). Specifically, BRAF-mutated distant metastatic DTC showed hypomethylation of the Alu elements, but RAS-mutated tumors did not. Interestingly, distant metastatic DTC harboring no mutations in BRAF or RAS showed notable variability, which could reflect the BRAF-like and RAS-like phenotypes.

Several studies indicate that the global DNA methylation in peripheral blood leukocytes differs significantly between healthy individuals and patients with different cancer types (Li et al. 2012). There is only one study analyzing global DNA methylation in blood from PTC patients and control individuals (using a 5mC DNA ELISA), and they did not find differences (Ceolin et al. 2018).

Altogether, these findings indicate that global hypomethylation plays a role in thyroid tumorigenesis. However, further investigation is required. The analysis of global hypomethylation in different compartments of the genome using different techniques in a large cohort of thyroid samples would provide enlightening insight.

Focal DNA methylation alterations in thyroid cancer: gene-specific studies
Gene-specific DNA methylation has been broadly studied. Accordingly, many investigators in thyroid cancer have focused on the hypermethylation of specific tumor suppressor genes (TSGs) as an alternative to mutational inactivation (Supplementary Table 2). Some of the most recurrently hypermethylated TSGs in thyroid cancer are Ras association domain family 1, isoform A (RASSF1A), cyclin-dependent kinase inhibitor 2A (CDKN2A or P16INK4A) and death-associated protein kinase1 (DAPK) (Table 3). The RASSF1A gene encodes a signaling protein containing a Ras association domain and is involved in multiple apoptotic and cell cycle checkpoint pathways. The main mechanism of RASSF1A inactivation, a frequent event in many cancers, appears to be through promoter methylation rather than mutational events (Dammann et al. 2000,  Agathanggelou et al. 2005). Schagdarsurengin et al. found for the first time that RASSF1A was hypermethylated in thyroid cancer with a slightly higher frequency in more aggressive hystotypes (Schagdarsurengin et al. 2002). Many other studies confirmed RASSF1A hypermethylation in thyroid tumors, including a recent meta-analysis, although most results revealed a considerable overlap in methylation levels between benign and malignant tumors (Nakamura et al. 2005,  Hou et al. 2008, Mohammadi-asl et al. 2011, Stephen et al. 2011, Niu et al. 2017) (Table 3), suggesting that RASSF1A hypermethylation may be an early epigenetic event in thyroid carcinogenesis (Xing et al. 2004, Brown et al. 2014). Different studies have shown the potential of RASSF1A hypermethylation as a biomarker of aggressive tumors, while other authors have failed to find any relationship between RASSF1A hypermethylation and prognostic factors (Schagdarsurengin et al. 2006, Mohammadi-asl et al. 2011, Brait et al. 2012, Niu et al. 2017). Thus, further studies are required in larger series of samples using more quantitative techniques. P16INK4A is a cell cycle regulator that induces G1 phase arrest, whose functional loss is frequent in cancer. Mutations in this gene are rarely observed in primary thyroid tumors (Calabrò et al. 1996, Yane et al. 1996), whereas promoter hypermethylation, which causes gene silencing, is quite common (Table 3). Many studies, including a meta-analysis based on 17 case–control studies (804 thyroid cancer patients, 487 controls), confirmed the significantly higher frequency of P16INK4A hypermethylation in thyroid cancer than in normal samples (Boltze et al. 2003, Schagdarsurengin et al. 2006, Melck et al. 2007, Zafon et al. 2008,  Wu et al. 2015). However, its usefulness as a prognostic marker remains questionable, as summarized in Table 3. DAPK1, which belongs to the DAPK family of calcium/calmodulin-dependent kinases, participates in many cellular processes such as apoptosis, autophagy, and cell survival, and is also involved in cancer (reviewed in Farag & Roh 2019). DAPK1 promoter hypermethylation has been associated with an increased risk of developing cancer and poor prognosis in several cancer types (Dai et al. 2016, Qi & Xiong 2018, Yang et al. 2018). As shown in Table 3, DAPK1 is hypermethylated in benign and malignant thyroid tumors but the clinical value of DAPK1 hypermethylation is not well established.

Table 3
Summary of candidate approach DNA methylation studies on classical tumor suppressor genes.

Gene Study No. Method Discovery series (n) Methylation statusa Potential clinical value Reference
NT BTLs DTC PDTC/ATC
RASSF1A 1 MSP NT (4), NG (1), PTC (13), FTC (10), PDTC (1), ATC (9), CL (9) U U M M Higher hypermethylation frequency in aggressive variants Schagdarsurengin et al. (2002)
2 qMSP NT (14), BTLs (9), FTC (12), PTC (30) U hyper hyper – Early event in thyroid tumorigenesis. Hypermethylation mutually exclusive with BRAF mutation Xing et al. (2004)
3 MSP, qMSP NT (15), NG (20), FA (24), PTC (23), fvPTC (10), CL (5) M hyper hyper – Hypermethylation mutually exclusive with BRAF mutation Hoque et al. (2005)
4 MSP NT (42), FA (3), HTT (23),PTC (42), FTC (4), ATC (12), CL (3) U M M M Early event in thyroid tumorigenesis. No relationship with BRAF mutation Nakamura et al. (2005)
5 MSP NG (12), FA (10), PTC (13), FTC (10), ATC (9), CL (9) – M M M Higher methylation frequencies in older patients Schagdarsurengin et al. (2006)
6 Pyrosequencing NT (21), FTC (21) U – hyper – – Lee et al. (2008)
7 qMSP FA(42), FTC (65), ATC (36), CL (5) – M M M – Hou et al. (2008)
8 COBRA BTLs (25), PTC (25) – M hyper – Early event in thyroid tumorigenesis Mohammadi-asl et al. (2011)
9 MSP NG (20), cPTC (27), fvPTC (15), tcPTC (3) M M M – Early event in thyroid tumorigenesis Czarnecka et al. (2011)
10 MS-MLPA NT (5), NG (3), PTC (11), FTC (2) M M M – Early event in thyroid tumorigenesis Stephen et al. (2011)
11 qMSP NT (15), BTLs (44), cPTC (17), fvPTC (10), FTC (7), HCTC (2) M M hyper – Hypermethylation mutually exclusive with BRAF mutation Brait et al. (2012)
12 PCR + MS-RE NT (18), PTC (41) U – hyper – Related to aggressiveness Kunstman et al. (2013)
13 MS-RE + qPCR NT (29), BTLs (23), FA (10), FTC (10) U hyper hyper – Early event in thyroid tumorigenesis Brown et al. (2014)
14 qMSP HCTC (26), FTC (27) – – M – Hypermethylation in HCTC compared to FTC Stephen et al. (2015)
15 qMSP NT (71), FA (83), HCTC (44), FTC (46), fvPTC (42), cPTC (53) U hyper hyper – Good discrimination between NT and tumors Stephen et al. (2018)
P16INK4A (CDKN2A) 1 MSP FA (8), PTC (12), CL (4) – M M – – Elisei et al. (1998)
2 MSP NT (15), FA (18) PTC (16), FTC (18), PDTC (12), ATC (13) U M M M Higher hypermethylation frequency in aggressive variants. Related to N1 and M1 Boltze et al. (2003)
3 MSP, qMSP NT (15), NG (20), FA (24), PTC (23), fvPTC (10), CL (5) U U U – No clorrelation with clinical any parameters Hoque et al. (2005)
4 MSP NG (12), FA (10), PTC (13), FTC (10), ATC (9), CL (9) – U M M Higher hypermethylation frequency in aggressive variants Schagdarsurengin et al. (2006)
5 MSP PTC (39) – – M – Related to progression Ishida et al. (2007)
6 COBRA BTLs (25), PTC (25) – U hyper – No clorrelation with clinical any parameters Mohammadi-asl et al. (2011)
7 MSP NG (20), cPTC (27), fvPTC (15), tcPTC (3) – M M – – Czarnecka et al. (2011)
8 qMSP NT (15), BTLs (44), cPTC (17), fvPTC (10), FTC (7), HCTC (2) U U U – No clorrelation with clinical any parameters Brait et al. (2012)
9 MSP NT (21), PTC (74) U – M Related to aggressiveness Wang et al. (2013)
10 qMSP NT (71), FA (83), HCTC (44), FTC (46), fvPTC (42), cPTC (53) U U U – – Stephen et al. (2018)
DAPK1 1 MSP, qMSP NT (15), NG (20), FA (24), PTC (23), fvPTC (10), CL (5) M M M – – Hoque et al. (2005)
2 qMSP cPTC (127), fvPTC (82), tcPTC (22) – – M – Higher hypermethylation frequency in in cPTC and tcPTC. Related with multifocality Hu et al. (2006)
3 MSP NG (12), FA (10), PTC (13), FTC (10), ATC (9), CL (9) – U M M Higher hypermethylation frequency in aggressive variants Schagdarsurengin et al. (2006)
4 qMSP NT (15), BTLs (44), cPTC (17), fvPTC (10), FTC (7), HCTC (2) M M M – – Brait et al. (2012)
5 qMSP HCTC (26), FTC (27) – – U – Stephen et al. (2015)
6 qMSP NT (71), FA (83), HCTC (44), FTC (46), fvPTC (42), cPTC (53) U U U – – Stephen et al. (2018)
aMethylation status of gene promoter: M, methylated; U, unmethylated; hyper, hypermethylated compared to normal tissue.

ATC, anaplastic thyroid cancer; BTLs, benign thyroid lesions; CL, cell lines; COBRA, combined bisulfite restriction analysis; cPTC, classical PTC; FA, follicular adenoma; FTC, follicular thyroid cancer; fvPTC, follicular variant of PTC; HCTC, Hurthle cell thyroid cancer; HTT, hyalinizing trabecular tumor; MSP, methyl-sensitive PCR; MS-RE, methylation-sensitive restriction enzyme; NG, nodular goiter; NT, normal tissue; PDTC, poorly differentiated thyroid cancer; PTC, papillary thyroid cancer; qMSP, quantitative MSP.

In addition, the methylation of thyroid-specific genes, such as thyroid-stimulating hormone receptor (TSHR) or NIS, has also been extensively investigated in thyroid cancer (Table 4 and Supplementary Table 2). TSHR plays a central role in the regulation of thyroid growth and function. Somatic TSHR mutations have been found in both benign and malignant thyroid neoplasms (Davies et al. 2010), but the involvement of TSHR genetic status in thyroid carcinogenesis is not clear. A recent study showed that TSHR mutations may be associated with an increased cancer risk when present at high allelic frequency (Mon et al. 2018). In thyroid cancer, TSHR expression is also repressed by aberrantly methylation of gene promoter (Table 4). However, several studies found TSRH methylated in bening lesions (Hoque et al. 2005,  Schagdarsurengin et al. 2006, Brait et al. 2012, Kartal et al. 2015, Stephen et al. 2018), which limits the discriminatory power for diagnostic purposes. Interestingly, TSHR methylation has been found inversely associated with tumor recurrence (Smith et al. 2007). NIS is a transmembrane glycoprotein that mediates the active transport of iodide from the bloodstream into the follicular thyroid cells and is mainly regulated by the thyroid-stimulating hormone (TSH). The role of NIS is key for effective diagnosis and treatment of thyroid cancer since RAI accumulation is primarily mediated by NIS. Accordingly, the decreased NIS expression and/or impairment in NIS plasma membrane trafficking (De la Vieja & Santisteban 2018) are well demonstrated factors showing poor prognosis in thyroid cancer. However, the relationship between NIS and thyroid cancer is complex and not well understood (de Morais et al. 2018). Mutations in the NIS gene do not appear to be a major cause for reduced NIS expression/function in thyroid cancer (Russo et al. 2001). In contrast, many studies have reported the methylation of NIS promoter although results are controversial (Table 4). Interestingly, Galrao et al. identified a distal enhancer that was hypermethylated in DTC regulating NIS expression (Galrão et al. 2014). This new finding provides a basis for further investigation in the epigenetic regulation of NIS.

Table 4
Summary of candidate approach DNA methylation studies on thyroid-specific genes.

Gene Study No. Method Discovery series (n)a Methylation statusb Potential clinical value Reference
NT BTLs DTC ATC
TSHR 1 MSP, qMSP FA (8), PTC (39), FTC (15), ATC (11), CL (6) – U hyper hyper Diagnostic marker Xing et al. (2003)
2 MSP NG (12), FA (10), PTC (13), FTC (10), ATC (9), CL (9) – M M M – Schagdarsurengin et al. (2006)
3 MSP NT(2), NG (15), FA (10), PTC (30) U M hyper – Inverse correlation with recurrence Smith et al. (2007)
4 MSP FNAB: BTLs (35), FA (4), PTC (28), FTC (6) – M M – More frequent methylation in PTC Kartal et al. (2015)
5 qMSP NT (71), FA (83) cPTC (53), fvPTC (42), HCTC (44), FTC (46) M M M – – Stephen et al. (2018)
6 qMSP NT (15), NG (20), FA (24), PTC (23), fvPTC (10), CL (5) M M M – – Hoque et al. (2005)
7 qMSP NT (15), BTLs (44), cPTC (17), fvPTC (10), FTC (7), HCTC (2) M M hyper – – Brait et al. (2012)
NIS (SLC5A5) 1 MSP NT(2), NG (15), FA (10),PTC (30) U U hyper – No prognostic factor Smith et al. (2007)
2 MS-MLPA NT (5), NG (3), PTC (11), FTC (2) M M M – Early event Stephen et al. (2011)
3 MSP NT (30), BTLs (10), PTC (18), FTC (2) M M M – – Galrão et al. (2013)
4 BS-sequencing NT (30), BTLs (10), PTC (18), FTC (2) M M hyper – Correlation with expression Galrão et al. (2014)
5c MSP, BS-sequencing NT (24), PTC (24) M – hyper – Related to BRAF(V600E) Choi et al. (2014)
6 qMSP HCTC (26), FTC (27) – – M – No differences between HCTC and FTC Stephen et al. (2015)
aDNA methylation was assessed in postsurgical tissue unless otherwise stated. bMethylation status of gene promoter: hyper, hypermethylated compared to normal tissue; M, methylated; U, unmethylated. cMethylation status of NIS enhancer.

ATC, anaplastic thyroid cancer; BS-sequencing, bisulfite sequencing; BTLs, benign thyroid lesions; CL, cell lines; cPTC, classical PTC; FA, follicular adenoma; FNAB, fine-needle aspiration biopsy; FTC, follicular thyroid cancer; fvPTC, follicular variant of PTC; HCTC, Hurthle cell thyroid cancer; MSP, methyl-sensitive PCR; NG, nodular goiter; PTC, papillary thyroid cancer; qMSP, quantitative MSP.

Focal DNA methylation alterations in thyroid cancer: genome-wide studies
In addition to validating results from candidate approach studies, genome-wide DNA methylation studies have allowed the identification of novel differentially methylated sequences that may regulate the expression of genes involved in thyroid cancer tumorigenesis (Table 1). In this regard, Rodríguez-Rodero et al. showed that in ATC, the membrane-associated protein 17 (MAP17) gene was hypomethylated in its promoter region and overexpressed compared to normal tissues. They showed that overexpression of MAP17 induced tumor growth in vitro and in vivo (Rodríguez-Rodero et al. 2013). Zhang et al. identified 14 novel genes regulated by DNA methylation in PTC (Table 1) that were used to construct a core cofunction network that revealed the potential of the C-X-C motif chemokine ligand (CXCL12), a chemokine involved in the immune response, as a key player in thyroid tumorigenesis (Zhang et al. 2017). Moreover, the expression levels of these 14 genes gave the ability to discriminate between PTC patients and healthy individuals. By integrating DNA methylation and transcriptomic data, Beltrami et al. found 185 genes with a negative correlation between methylation and expression that mostly affected fibroblast growth factor (FGF) and retinoic acid (RA) signaling pathways (Beltrami et al. 2017). Other interesting hypomethylated genes that were identified in genome-wide DNA methylation analyses were high-mobility group box 2 (HMGB2), which may play a role in PTC cell proliferation, and FYVE, RhoGEF and PH domain-containing 1 (FDG1), which may be involved in cell invasion (Hou et al. 2011). Additionally, Lin et al. identified HORMA domain-containing 2 (HORMAD2) and showed that its hypermethylation and repression induced the progression of thyroid cancer, while its hypomethylation and overexpression retarded cell growth and mobility and facilitated apoptosis (Lin et al. 2018).

From a translational point of view, genome-wide DNA methylation studies are an important source of new biomarkers to develop algorithms and tools with diagnostic and prognostic value. For example, Mancikova et al. identified two putative biomarkers associated with recurrence-free survival, etoposide-induced 2.4 (EI24) and Wilms’ tumor 1(WT1) (Mancikova et al. 2014). They also found kallikrein 10 (KLK10) to be hypomethylated and overexpressed in BRAF-mutated tumors. Further analyses based on KLK10 allowed the development of an algorithm related to BRAF- and RAS-like phenotypes with prognostic implications in thyroid cancer (Buj et al. 2018). On the other hand, Bisarro dos Reis et al. developed a prognostic algorithm based on 21 CpGs able to predict recurrence in DTC with high specificity but low sensitivity (Bisarro dos Reis et al. 2017). However, the series of samples used contained a low number of recurrent cases; thus, further analyses are required to validate its potential for prognostic use. Finally, Yim et al., who profiled PTC DNA methylation by RRBS, developed a new diagnostic method, the so-called diagnostic DNA methylation signature (DDMS) approach, which is based on 373 differentially methylated regions with tissue-specific DNA methylation patterns in benign and malignant nodules (Yim et al. 2019). A notable proportion of these markers were associated with active enhancers and cancer-related genes. Importantly, the DDMS approach distinguishes benign from malignant nodules with high sensitivity and specificity and thus has the potential to provide outstanding diagnostic accuracy for thyroid nodules, which may decrease overdiagnosis and unnecessary thyroidectomies.

DNA methylation as a therapeutic target in thyroid cancer
As explained, the aberrant methylation of DNA can play a key role in tumorigenesis. In addition, DNA methylation is inherently reversible, which makes targeted therapies against it very attractive for cancer treatment. Therefore, much effort has been made to study the potential of drugs that inhibit this type of epigenetic modification to induce the re-expression of silenced genes in different malignancies. Over the past few decades, different demethylating drugs have been developed and tested in different human neoplasms. There are two different classes of demethylating agents: nucleoside DNMT inhibitors and non-nucleoside DNMT inhibitors. Treatment with these agents causes a reduction in global DNA methylation rather than demethylation in specific regions (reviewed in Mani & Herceg 2010).

The most commonly used demethylating agents are the first ones described: 5-azacytidine (azacitidine, AZA) (Sorm et al. 1964) and 5-aza-2′-deoxycytidine (decitabine, DAC), both of which are nucleoside DNMT inhibitors. Around 1970, clinical trials in Europe and the United States using AZA began focusing on the treatment of both solid and blood neoplasms (Sorm & Vesely 1968). The results showed the effectiveness of treating patients with acute myeloid leukemia (AML) resistant to conventional treatment and/or with relapse with AZA and DAC. In contrast, no significant responses were found in other types of blood cancers or in solid tumors to those drugs. At that time, the US Food and Drug Administration (FDA) did not approve AZA due to its high levels of toxicity relative to its antitumor efficacy. Nearly 40 years later, in 2004, after adjusting the dosage to reduce toxicity and increase efficiency, it was approved for clinical use to treat myelodysplastic syndromes (MDS) (Kaminskas et al. 2005). In 2006, DAC was also approved for the treatment of MDS (Kantarjlan et al. 2006). More recently, other agents have been identified such as zebularine or procaine, and their potential use in demethylating therapy is being tested (Villar-Garea et al. 2003, Marquez et al. 2005, Mani & Herceg 2010).

Curiously, demethylating agents are more effective in treating hematologic cancers than solid tumors despite the large amount of evidence showing that aberrant DNA methylation is a trait common to all tumorigenic processes (Sharma et al. 2010). Such trouble in accomplishing therapeutic effectiveness could be due to a variety of reasons, such as lower DNMT activity in solid tumors (Lin et al. 2009) or that the starting level of aberrant methylation in hematological malignancies is higher than that in solid tumors (Issa et al. 1997). Another limitation is that these agents need actively dividing cells to take action. Therefore, slow-growing tumors might require longer dosing schedules due to their short half-life or improvement in drug delivery and plasma stability (Howell et al. 2010).

It may seem that the encouraging effects seen in trials to treat hematological malignancies and preclinical data on solid tumors will never reach a clinical application. However, several studies have shown the association between the overexpression of DNMTs and chemoresistance (Wang et al. 2001, Qiu et al. 2002, Segura-Pacheco et al. 2006), and the treatment of cancer cell lines with DNMT inhibitors can revert this resistance to therapy (Qiu et al. 2005). In light of these findings, clinical trials with promising results have been conducted to test whether demethylating drugs can enhance susceptibility to other therapies when administered in combination, especially in resistant tumors (Linnekamp et al. 2017).

Thyroid cancer is not an exception to all the details explained above. Initially, in vitro studies focused on the ability of demethylating drugs to restore the expression of different genes to sensitize thyroid cancer cells to RAI treatment. Venkataraman et al. reported an increase in the mRNA and gene expression of NIS in thyroid cancer cell lines treated with AZA compared to those observed without AZA treatment. The increased expression of the NIS gene was correlated with an increase in RAI uptake in some of the cell lines (Venkataraman et al. 1999). Nevertheless, other similar studies could not show a significant increase in RAI uptake when different thyroid cell lines were treated with AZA or DAC, highlighting that the mechanism may depend on the methylation and differentiation status of the cell (Tuncel et al. 2007, Miasaki et al. 2008). Although significant cell redifferentiation is not achieved, DAC and zebularine are able to inhibit cell proliferation and migration in thyroid cancer cell lines (Miasaki et al. 2008, Kim et al. 2013).

There have only been two clinical trials focusing on the treatment of thyroid cancer patients with demethylating agents to sensitize tumors to RAI (ClinicalTrials.gov Identifier: NCT00085293 and NCT00004062). Both included patients with recurrent and/or metastatic DTC that were resistant to RAI. Unfortunately, no partial or complete responses were observed, while treatment caused serious side effects. Therefore, as with other types of malignancies, efforts have also been made toward exploring the potential of these agents in combination with other treatments. Significant redifferentiation accompanied by growth inhibition and cell apoptosis was observed when cells were treated with DAC and RA. However, no increase in RAI uptake was observed due to the cytoplasmic localization of the NIS protein (Vivaldi et al. 2009). Other studies use demethylating drugs to increase the sensitivity of thyroid cancer cells to other agents such as TNF-related apoptosis-inducing ligand (TRAIL), which induces apoptosis (Siraj et al. 2011) or to upregulate immune-related genes in cancer cells to enhance their response to cancer immunotherapies (Gunda et al. 2013, 2014). A strong synergistic effect was also seen when DAC was combined with everolimus (an mTOR inhibitor) to treat thyroid cancer cells, opening a promising scenario to overcome drug resistance (Vitale et al. 2017). However, the most explored and promising option, not just in thyroid cancer, is the combination of demethylating agents with HDAC inhibitors such as trichostatin A (TSA), sodium butyrate or valproic acid. In vitro, this combination can restore NIS transcription to levels approaching those present in RAI-responder tumors (Li et al. 2007) and even increase RAI uptake (Provenzano et al. 2007, Massimino et al. 2018). In addition, they can also inhibit cell growth and invasion (Mitmaker et al. 2011).

Finally, it is important to mention the role that demethylating agents have been playing through the years as important tools to discover and study new prognostic and diagnostic biomarkers (Murgo 2005, Zuo et al. 2010, Latini et al. 2011, Moraes et al. 2016, Wu et al. 2016,  Cao et al. 2018).

In conclusion, there is little evidence of the effectiveness of demethylating agents in thyroid cancer. Most studies have tested these drugs in a variety of cancer cell lines obtaining promising results that have not been translated into clinical practice. However, despite the unsuccessful results in clinical trials with their use as solo agents, they may be a potentially useful therapy when combined with other drugs.

DNA methylation in thyroid cancer cell lines
Established human thyroid cancer cell lines are the most widely used models to study thyroid tumorigenesis, including studies aimed at understanding the DNA methylation landscape. However, it has been shown that cell lines derived from DTC, both PTC and FTC, display mRNA expression profiles closer to dedifferentiated in vivo thyroid tumors (i.e., ATC) than to differentiated ones (van Staveren et al. 2007, Saiselet et al. 2012). This can be explained by the prior selection of initiating cells and the in vitro evolution of the cell lines. Interestingly, some of the genes commonly upregulated in ATC and thyroid cancer cell lines are related to DNA replication, which is in accordance with their high proliferation rate.

Considering that DNA methylation is involved in the regulation of gene expression, how does cellular immortalization affect DNA methylation in thyroid cancer cell lines? Although there are few studies profiling DNA methylation in thyroid cancer cell lines, they note that DNA methylation follows a similar pattern as gene expression. Rodero-Rodríguez et al. analyzed four cell lines (one derived from PTC, one from FTC, one from ATC and one from MTC), and all of them exhibited methylomes that more closely resembled undifferentiated tumors than differentiated ones (Rodríguez-Rodero et al. 2013). Typically, immortalized cell lines exhibit hypermethylation (Smiraglia et al. 2001). However, this is not the case for thyroid cancer cell lines, which, based on the study by Rodero-Rodríguez et al., show more hypomethylation than hypermethylation events (Rodríguez-Rodero et al. 2013). This result is in agreement with Klein Hesselink et al. who found that the global hypomethylation level of Alu elements in PTC- and FTC-derived cell lines was similar to that observed in ATC-derived cell lines and in vivo PDTC and ATC samples (Klein Hesselink et al. 2018).

However, some gene-specific DNA methylation studies found a good agreement between in vivo DTC tumors and cell lines. Therefore, despite the limitations of the use of cell lines, they provide a good model for controlled experiments, for example, to study the DNA methylation-mediated regulation of candidate genes identified by genome-wide studies or to investigate the effect of specific treatments, such as the ability of demethylating drugs to restore the expression of different genes to sensitize thyroid cancer cells to RAI.

Altogether, these studies indicate that thyroid cancer cell lines are an important tool for thyroid cancer research, but the differences in gene expression and DNA methylation compared to in vivo tumors should be taken into account when extrapolating results obtained from these cells.

Concluding remarks and perspectives
Numerous studies on DNA methylation in thyroid cancer have improved our understanding of thyroid carcinogenesis. However, we still do not have a complete picture of the methylation landscape, especially for histological subtypes other than PTC. The huge catalog of DNA methylation alterations, the association of DNA hypomethylation with cancer progression and dedifferentiation, the existence of different methylomes related to different clinical and molecular phenotypes and the influence of immune-infiltrating cells in tumor DNA methylation patterns are some of the recent findings that will most likely define the direction of future research in the field of DNA methylation in thyroid cancer. In addition, numerous studies confirm the importance of DNA methylation as a source of novel biomarkers in thyroid cancer. Indeed, some studies propose potential diagnostic and prognostic markers, although the combination of DNA methylation alterations with other epigenetic and/or genetic alterations may improve their clinical value. Finally, DNA methylation is also a fundamental area of interest from a therapeutic perspective. Therefore, further in vitro and in vivo functional experiments to better understand the implications and underlying mechanisms of DNA methylation alterations in thyroid cancer as well as the evaluation of candidate biomarkers through case–control studies and prospective trials are warranted.

Supplementary data
This is linked to the online version of the paper at https://doi.org/10.1530/ERC-19-0093.

Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of this review.

Funding
This work was supported by a grant from the Instituto de Salud Carlos III, co-funded by ERDF/ESF, ‘Investing in your future’ (FIS PI18/00654 to M J).

References
AffinitoOSalernoPD’AlessioAMCuomoMFlorioECarlomagnoFProiettiAGianniniRBasoloFChiariottiL 2019 Association between DNA methylation profile and malignancy in follicular-patterned thyroid neoplasms. Endocrine-Related Cancer 26 451–462. (https://doi.org/10.1530/ERC-18-0308)

Crossref
Search Google Scholar
Export Citation
AgathanggelouACooperWNLatifF 2005 Role of the Ras-association domain family 1 tumor suppressor gene in human cancers. Cancer Research 65 3497–3508. (https://doi.org/10.1158/0008-5472.CAN-04-4088)

Crossref
PubMed
Search Google Scholar
Export Citation
AllisCDJenuweinT 2016 The molecular hallmarks of epigenetic control. Nature Reviews: Genetics 17 487–500. (https://doi.org/10.1038/nrg.2016.59)

Crossref
PubMed
Search Google Scholar
Export Citation
AroraNScognamiglioTZhuBFaheyTJIII 2008 Do benign thyroid nodules have malignant potential? An evidence-based review. World Journal of Surgery 32 1237–1246. (https://doi.org/10.1007/s00268-008-9484-1)

Crossref
PubMed
Search Google Scholar
Export Citation
AsaSL 2017 The evolution of differentiated thyroid cancer. Pathology 49 229–237. (https://doi.org/10.1016/j.pathol.2017.01.001)

Crossref
PubMed
Search Google Scholar
Export Citation
BannisterAJKouzaridesT 2011 Regulation of chromatin by histone modifications. Cell Research 21 381–395. (https://doi.org/10.1038/cr.2011.22)

Crossref
PubMed
Search Google Scholar
Export Citation
BaubecTSchubelerD 2014 Genomic patterns and context specific interpretation of DNA methylation. Current Opinion in Genetics and Development 25 85–92. (https://doi.org/10.1016/j.gde.2013.11.015)

Crossref
Search Google Scholar
Export Citation
BelancioVPRoy-EngelAMDeiningerPL 2010 All y’all need to know ‘bout retroelements in cancer’. Seminars in Cancer Biology 20 200–210. (https://doi.org/10.1016/J.SEMCANCER.2010.06.001)

Crossref
Search Google Scholar
Export Citation
BeltramiCMdos ReisMBBarros-FilhoMCMarchiFAKuasneHPintoCALAmbatipudiSHercegZKowalskiLPRogattoSR 2017 Integrated data analysis reveals potential drivers and pathways disrupted by DNA methylation in papillary thyroid carcinomas. Clinical Epigenetics 9 45. (https://doi.org/10.1186/s13148-017-0346-2)

Crossref
PubMed
Search Google Scholar
Export Citation
BenardAvan de VeldeCJHLessardLPutterHTakeshimaLKuppenPJKHoonDSB 2013 Epigenetic status of LINE-1 predicts clinical outcome in early-stage rectal cancer. British Journal of Cancer 109 3073–3083. (https://doi.org/10.1038/bjc.2013.654)

Crossref
PubMed
Search Google Scholar
Export Citation
BermanBPWeisenbergerDJAmanJFHinoueTRamjanZLiuYNoushmehrHLangeCPvan DijkCMTollenaarRA 2011 Regions of focal DNA hypermethylation and long-range hypomethylation in colorectal cancer coincide with nuclear lamina-associated domains. Nature Genetics 44 40–46. (https://doi.org/10.1038/ng.969)

PubMed
Search Google Scholar
Export Citation
BirdAP 1980 DNA methylation and the frequency of CpG in animal DNA. Nucleic Acids Research 8 1499–1504. (https://doi.org/10.1093/nar/8.7.1499)

Crossref
PubMed
Search Google Scholar
Export Citation
BirdAP 1986 CpG-rich islands and the function of DNA methylation. Nature 321 209–213. (https://doi.org/10.1038/321209a0)

Crossref
PubMed
Search Google Scholar
Export Citation
BirdA 2007 Perceptions of epigenetics. Nature 447 396–398. (https://doi.org/10.1038/nature05913)

Crossref
PubMed
Search Google Scholar
Export Citation
Bisarro dos ReisMBarros-FilhoMCMarchiFABeltramiCMKuasneHPintoCALAmbatipudiSHercegZKowalskiLPRogattoSR 2017 Prognostic classifier based on genome-wide DNA methylation profiling in well-differentiated thyroid tumors. Journal of Clinical Endocrinology and Metabolism 102 4089–4099. (https://doi.org/10.1210/jc.2017-00881)

Crossref
Search Google Scholar
Export Citation
BochtlerMKolanoAXuGL 2017 DNA demethylation pathways: additional players and regulators. BioEssays 39 1–13. (https://doi.org/10.1002/bies.201600178)

PubMed
Search Google Scholar
Export Citation
BoltzeCZackSQuednowCBettgeSRoessnerASchneider-StockR 2003 Hypermethylation of the CDKN2/p16INK4A promotor in thyroid carcinogenesis. Pathology Research and Practice 199 399–404. (https://doi.org/10.1078/0344-0338-00436)

Crossref
PubMed
Search Google Scholar
Export Citation
BraitMLoyoMRosenbaumEOstrowKLMarkovaAPapagerakisSZahurakMGoodmanSMZeigerMSidranskyD 2012 Correlation between BRAF mutation and promoter methylation of TIMP3, RARβ2 and RASSF1A in thyroid cancer. Epigenetics 7 710–719. (https://doi.org/10.4161/epi.20524)

Crossref
PubMed
Search Google Scholar
Export Citation
BrownTCJuhlinCCHealyJMPrasadMLKorahRCarlingT 2014 Frequent silencing of RASSF1A via promoter methylation in follicular thyroid hyperplasia: a potential early epigenetic susceptibility event in thyroid carcinogenesis. JAMA Surgery 149 1146–1152. (https://doi.org/10.1001/jamasurg.2014.1694)

Crossref
PubMed
Search Google Scholar
Export Citation
BujRMallonaIDíez-VillanuevaABarreraVMauricioDPuig-DomingoMReverterJLMatias-GuiuXAzuaraDRamírezJL 2016 Quantification of unmethylated Alu (QUAlu): a tool to assess global hypomethylation in routine clinical samples. Oncotarget 7 10536–10546. (https://doi.org/10.18632/oncotarget.7233)

PubMed
Search Google Scholar
Export Citation
BujRMallonaIDíez-VillanuevaAZafónCMateJLRocaMPuig-DomingoMReverterJLMauricioDPeinadoMA 2018 Kallikreins stepwise scoring reveals three subtypes of papillary thyroid cancer with prognostic applications. Thyroid 28 601–612. (https://doi.org/10.1089/thy.2017.0501)

Crossref
Search Google Scholar
Export Citation
CalabròVStrazzulloMLa MantiaGFedeleMPaulinCFuscoALaniaL 1996 Status and expression of the p16INK4 gene in human thyroid tumors and thyroid-tumor cell lines. International Journal of Cancer 67 29–34. (https://doi.org/10.1002/(SICI)1097-0215(19960703)67:1<29::AID-IJC7>3.0.CO;2-1)

Crossref
PubMed
Search Google Scholar
Export Citation
Cancer Genome Atlas Research Network 2012 Comprehensive molecular characterization of human colon and rectal cancer. Nature 487 330–337. (https://doi.org/10.1038/nature11252)

PubMed
Search Google Scholar
Export Citation
Cancer Genome Atlas Research Network 2014 Integrated genomic characterization of papillary thyroid carcinoma. Cell 159 676–690. (https://doi.org/10.1016/j.cell.2014.09.050)

PubMed
Search Google Scholar
Export Citation
CaoYMGuJZhangYSWeiWJQuNWenDLiaoTShiRLZhangLJiQH 2018 Aberrant hypermethylation of the HOXD10 gene in papillary thyroid cancer with BRAFV600E mutation. Oncology Reports 39 338–348. (https://doi.org/10.3892/or.2017.6058)

PubMed
Search Google Scholar
Export Citation
CeolinLGoularteAPPFerreiraCVRomittiMMaiaAL 2018 Global DNA methylation profile in medullary thyroid cancer patients. Experimental and Molecular Pathology 105 110–114. (https://doi.org/10.1016/j.yexmp.2018.06.003)

Crossref
PubMed
Search Google Scholar
Export Citation
CeramiEGaoJDogrusozUGrossBESumerSOAksoyBAJacobsenAByrneCJHeuerMLLarssonE 2012 The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discovery 2 401–404. (https://doi.org/10.1158/2159-8290.CD-12-0095)

Crossref
PubMed
Search Google Scholar
Export Citation
ChalitchagornKShuangshotiSHourpaiNKongruttanachokNTangkijvanichPThong-ngamDVoravudNSriuranpongVMutiranguraA 2004 Distinctive pattern of LINE-1 methylation level in normal tissues and the association with carcinogenesis. Oncogene 23 8841–8846. (https://doi.org/10.1038/sj.onc.1208137)

Crossref
PubMed
Search Google Scholar
Export Citation
ChenYCGoteaVMargolinGElnitskiL 2017 Significant associations between driver gene mutations and DNA methylation alterations across many cancer types. PLoS Computational Biology 13 e1005840. (https://doi.org/10.1371/journal.pcbi.1005840)

Crossref
PubMed
Search Google Scholar
Export Citation
ChoiYWKimH-JKimYHParkSHChwaeYJLeeJSohEYKimJ-HParkTJ 2014 B-RafV600E inhibits sodium iodide symporter expression via regulation of DNA methyltransferase. Experimental and Molecular Medicine 1 46.e120. (https://doi.org/10.1038/emm.2014.68)

Search Google Scholar
Export Citation
CzarneckaKPastuszak-LewandoskaDMigdalska-SekMNawrotEBrzezinskiJDedecjusMPomorskiLBrzezianskaE 2011 Aberrant methylation as a main mechanism of TSGs silencing in PTC. Frontiers in Bioscience 3 137–157. (https://doi.org/10.2741/e228)

Search Google Scholar
Export Citation
DaiLMaCZhangZZengSLiuATangSRenQSunYXuC 2016 DAPK promoter methylation and bladder cancer risk: a systematic review and meta-analysis. PLoS ONE 11 e0167228. (https://doi.org/10.1371/journal.pone.0167228)

Crossref
Search Google Scholar
Export Citation
DammannRLiCYoonJHChinPLBatesSPfeiferGP 2000 Epigenetic inactivation of a RAS association domain family protein from the lung tumour suppressor locus 3p21.3. Nature Genetics 25 315–319. (https://doi.org/10.1038/77083)

Crossref
PubMed
Search Google Scholar
Export Citation
DaviesTFYinXLatifR 2010 The genetics of the thyroid stimulating hormone receptor: history and relevance. Thyroid 20 727–736. (https://doi.org/10.1089/thy.2010.1638)

Crossref
PubMed
Search Google Scholar
Export Citation
de CapoaAGrappelliCVolpinoPBononiMMusolinoACiardiACavallaroACangemiV 2004 Nuclear methylation levels in normal and cancerous thyroid cells. Anticancer Research 24 1495–1500.

PubMed
Search Google Scholar
Export Citation
De la ViejaASantistebanP 2018 Role of iodide metabolism in physiology and cancer. Endocrine-Related Cancer 25 R225–R245. (https://doi.org/10.1530/ERC-17-0515)

Crossref
Search Google Scholar
Export Citation
de MoraisRMSobrinhoABde Souza SilvaCMde OliveiraJRda SilvaICRde Toledo NóbregaO 2018 The role of the NIS (SLC5A5) gene in papillary thyroid cancer: a systematic review. International Journal of Endocrinology 2018 9128754. (https://doi.org/10.1155/2018/9128754)

PubMed
Search Google Scholar
Export Citation
DralleHMachensABasaJFatourechiVFranceschiSHayIDNikiforovYEPaciniFPasiekaJLShermanSI 2015 Follicular cell-derived thyroid cancer. Nature Reviews: Disease Primers 1 15077. (https://doi.org/10.1038/nrdp.2015.77)

PubMed
Search Google Scholar
Export Citation
DuQLuuPLStirzakerCClarkSJ 2015 Methyl-CpG-binding domain proteins: readers of the epigenome. Epigenomics 7 1051–1073. (https://doi.org/10.2217/epi.15.39)

Crossref
PubMed
Search Google Scholar
Export Citation
EhrlichM 2002 DNA methylation in cancer: too much, but also too little. Oncogene 21 5400–5413. (https://doi.org/10.1038/sj.onc.1205651)

Crossref
PubMed
Search Google Scholar
Export Citation
EhrlichM 2009 DNA hypomethylation in cancer cells. Epigenomics 1 239–259. (https://doi.org/10.2217/epi.09.33)

Crossref
PubMed
Search Google Scholar
Export Citation
EhrlichMGama-SosaMAHuangLHMidgettRMKuoKCMcCuneRAGehrkeC 1982 Amount and distribution of 5-methylcytosine in human DNA from different types of tissues of cells. Nucleic Acids Research 10 2709–2721. (https://doi.org/10.1093/nar/10.8.2709)

Crossref
PubMed
Search Google Scholar
Export Citation
EliseiRShioharaMKoefflerHPFaginJA 1998 Genetic and epigenetic alterations of the cyclin-dependent kinase inhibitors p15INK4b and p16INK4a in human thyroid carcinoma cell lines and primary thyroid carcinomas. Cancer 83 2185–2193. (https://doi.org/10.1002/(SICI)1097-0142(19981115)83:10<2185::AID-CNCR18>3.0.CO;2-4)

Crossref
PubMed
Search Google Scholar
Export Citation
EllisRJWangYStevensonHSBoufraqechMPatelDNilubolNDavisSEdelmanDCMerinoMJHeM 2014 Genome-wide methylation patterns in papillary thyroid cancer are distinct based on histological subtype and tumor genotype. Journal of Clinical Endocrinology and Metabolism 99 E329–E337. (https://doi.org/10.1210/jc.2013-2749)

Crossref
Search Google Scholar
Export Citation
EstellerM 2007 Epigenetic gene silencing in cancer: the DNA hypermethylome. Human Molecular Genetics 16 R50–R59. (https://doi.org/10.1093/hmg/ddm018)

Crossref
Search Google Scholar
Export Citation
EstellerMSilvaJMDominguezGBonillaFMatias-GuiuXLermaEBussagliaEPratJHarkesICRepaskyEA 2000 Promoter hypermethylation and BRCA1 inactivation in sporadic breast and ovarian tumors. Journal of the National Cancer Institute 92 564–569. (https://doi.org/10.1093/jnci/92.7.564)

Crossref
PubMed
Search Google Scholar
Export Citation
FangMOuJHutchinsonLGreenMR 2014 The BRAF oncoprotein functions through the transcriptional repressor MAFG to mediate the CpG island methylator phenotype. Molecular Cell 55 904–915. (https://doi.org/10.1016/J.MOLCEL.2014.08.010)

Crossref
PubMed
Search Google Scholar
Export Citation
FaragAKRohEJ 2019 Death-associated protein kinase (DAPK) family modulators: current and future therapeutic outcomes. Medicinal Research Reviews 39 349–385. (https://doi.org/10.1002/med.21518)

Crossref
PubMed
Search Google Scholar
Export Citation
FeinbergAPVogelsteinB 1983 Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature 301 89–92. (https://doi.org/10.1038/301089a0)

Crossref
PubMed
Search Google Scholar
Export Citation
FeinbergAPKoldobskiyMAGöndörA 2016 Epigenetic modulators, modifiers and mediators in cancer aetiology and progression. Nature Reviews: Genetics 17 284–299. (https://doi.org/10.1038/nrg.2016.13)

Crossref
PubMed
Search Google Scholar
Export Citation
FernandezAFAssenovYMartin-SuberoJIBalintBSiebertRTaniguchiHYamamotoHHidalgoMTanACGalmO 2012 A DNA methylation fingerprint of 1628 human samples. Genome Research 22 407–419. (https://doi.org/10.1101/gr.119867.110)

Crossref
PubMed
Search Google Scholar
Export Citation
GalrãoALSodréAKCamargoRYFrigugliettiCUKulcsarMALimaEUMedeiros-NetoGRubioIGS 2013 Methylation levels of sodium-iodide symporter (NIS) promoter in benign and malignant thyroid tumors with reduced NIS expression. Endocrine 43 225–229. (https://doi.org/10.1007/s12020-012-9779-8)

Crossref
PubMed
Search Google Scholar
Export Citation
GalrãoALCamargoRYFrigugliettiCUMoraesLCeruttiJMSerrano-NascimentoCSuzukiMFMedeiros-NetoGRubioIGS 2014 Hypermethylation of a new distal sodium/iodide symporter (NIS) enhancer (NDE) is associated with reduced NIS expression in thyroid tumors. Journal of Clinical Endocrinology and Metabolism 99 E944–E952. (https://doi.org/10.1210/jc.2013-1450)

Crossref
Search Google Scholar
Export Citation
GaluscaBDumollardJMLassandreSNiveleauAPradesJMEstourBPeoc’hM 2005 Global DNA methylation evaluation: potential complementary marker in differential diagnosis of thyroid neoplasia. Virchows Archiv 447 18–23. (https://doi.org/10.1007/s00428-005-1268-5)

Crossref
Search Google Scholar
Export Citation
Gama-SosaMASlagelVATrewynRWOxenhandlerRKuoKCGehrkeCWEhrlichM 1983 The 5-methylcytosine content of DNA from human tumors. Nucleic Acids Research 11 6883–6894. (https://doi.org/10.1093/nar/11.19.6883)

Crossref
PubMed
Search Google Scholar
Export Citation
GaudetFHodgsonJGEdenAJackson-GrusbyLDausmanJGrayJWLeonhardtHJaenischR 2003 Induction of tumors in mice by genomic hypomethylation. Science 300 489–492. (https://doi.org/10.1126/science.1083558)

Crossref
PubMed
Search Google Scholar
Export Citation
GollMGBestorTH 2005 Eukaryotic cytosine methyltransferases. Annual Review of Biochemistry 74 481–514. (https://doi.org/10.1146/annurev.biochem.74.010904.153721)

Crossref
PubMed
Search Google Scholar
Export Citation
GuanHJiMHouPLiuZWangCShanZTengWXingM 2008 Hypermethylation of the DNA mismatch repair gene hMLH1 and Its association with lymph node metastasis and T1799A BRAF mutation in patients with papillary thyroid cancer. Cancer 113 247–255. (https://doi.org/10.1002/cncr.23548)

Crossref
PubMed
Search Google Scholar
Export Citation
GundaVCogdillAPBernasconiMJWargoJAParangiS 2013 Potential role of 5-aza-2′-deoxycytidine induced MAGE-A4 expression in immunotherapy for anaplastic thyroid cancer. Surgery 154 1456–1462; discussion 1462. (https://doi.org/10.1016/j.surg.2013.07.009)

Crossref
PubMed
Search Google Scholar
Export Citation
GundaVFrederickDTBernasconiMJWargoJAParangiS 2014 A potential role for immunotherapy in thyroid cancer by enhancing NY-ESO-1 cancer antigen expression. Thyroid 24 1241–1250. (https://doi.org/10.1089/thy.2013.0680)

Crossref
PubMed
Search Google Scholar
Export Citation
HansenKDTimpWBravoHCSabunciyanSLangmeadBMcDonaldOGWenBWuHLiuYDiepD 2011 Increased methylation variation in epigenetic domains across cancer types. Nature Genetics 43 768–775. (https://doi.org/10.1038/ng.865)

Crossref
PubMed
Search Google Scholar
Export Citation
HaugenBRAlexanderEKBibleKCDohertyGMMandelSJNikiforovYEPaciniFRandolphGWSawkaAMSchlumbergerM 2016 2015 American Thyroid Association management guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American Thyroid Association Guidelines Task Force on thyroid nodules and differentiated thyroid cancer. Thyroid 26 1–133. (https://doi.org/10.1089/thy.2015.0020)

Crossref
PubMed
Search Google Scholar
Export Citation
HermanJGMerloAMaoLLapidusRGIssaJPDavidsonNESidranskyDBaylinSB 1995 Inactivation of the CDKN2/p16/MTS1 gene is frequently associated with aberrant DNA methylation in all common human cancers. Cancer Research 55 4525–4530.

PubMed
Search Google Scholar
Export Citation
HollidayRPughJE 1975 DNA modification mechanisms and gene activity during development. Science 187 226–232. (https://doi.org/10.1126/science.1111098)

Crossref
PubMed
Search Google Scholar
Export Citation
HolochDMoazedD 2015 RNA-mediated epigenetic regulation of gene expression. Nature Reviews: Genetics 16 71–84. (https://doi.org/10.1038/nrg3863)

Crossref
PubMed
Search Google Scholar
Export Citation
HoqueMORosenbaumEWestraWHXingMLadensonPZeigerMASidranskyDUmbrichtCB 2005 Quantitative assessment of promoter methylation profiles in thyroid neoplasms. Journal of Clinical Endocrinology and Metabolism 90 4011–4018. (https://doi.org/10.1210/jc.2005-0313)

Crossref
Search Google Scholar
Export Citation
HotchkissRD 1948 The quantitative separation of purines, pyrimidines, and nucleosides by paper chromatography. Journal of Biological Chemistry 175 315–332.

Search Google Scholar
Export Citation
HouPJiMXingM 2008 Association of PTEN gene methylation with genetic alterations in the phosphatidylinositol 3-kinase/AKT signaling pathway in thyroid tumors. Cancer 113 2440–2447. (https://doi.org/10.1002/cncr.23869)

Crossref
PubMed
Search Google Scholar
Export Citation
HouPLiuDXingM 2011 Genome-wide alterations in gene methylation by the BRAF V600E mutation in papillary thyroid cancer cells. Endocrine-Related Cancer 18 687–697. (https://doi.org/10.1530/ERC-11-0212)

Crossref
PubMed
Search Google Scholar
Export Citation
HowellPMLiuZKhongHT 2010 Demethylating agents in the treatment of cancer. Pharmaceuticals 3 2022–2044. (https://doi.org/10.3390/ph3072022)

Crossref
PubMed
Search Google Scholar
Export Citation
HuSLiuDTufanoRPCarsonKARosenbaumECohenYHoltEHKiseljak-vassiliadesKRhodenKJTolaneyS 2006 Association of aberrant methylation of tumor suppressor genes with tumor aggressiveness and BRAF mutation in papillary thyroid cancer. International Journal of Cancer 119 2322–2329. (https://doi.org/10.1002/ijc.22110)

Crossref
PubMed
Search Google Scholar
Export Citation
IshidaENakamuraMShimadaKHiguchiTTakatsuKYaneKKonishiN 2007 DNA hypermethylation status of multiple genes in papillary thyroid carcinomas. Pathobiology 74 344–352. (https://doi.org/10.1159/000110028)

Crossref
PubMed
Search Google Scholar
Export Citation
IssaJPBaylinSBHermanJG 1997 DNA methylation changes in hematologic malignancies: biologic and clinical implications. Leukemia 11 S7–S11.

Search Google Scholar
Export Citation
JonesPA 2012 Functions of DNA methylation: islands, start sites, gene bodies and beyond. Nature Reviews: Genetics 13 484–492. (https://doi.org/10.1038/nrg3230)

Crossref
PubMed
Search Google Scholar
Export Citation
JonesPABaylinSB 2007 The epigenomics of cancer. Cell 128 683–692. (https://doi.org/10.1016/j.cell.2007.01.029)

Crossref
PubMed
Search Google Scholar
Export Citation
JordàMPeinadoMA 2010 Methods for DNA methylation analysis and applications in colon cancer. Mutation Research 693 84–93. (https://doi.org/10.1016/j.mrfmmm.2010.06.010)

Crossref
PubMed
Search Google Scholar
Export Citation
JordàMRodríguezJFrigolaJPeinadoMA 2009 Analysis of DNA methylation by amplification of intermethylated sites (AIMS). Methods in Molecular Biology 507 107–116. (https://doi.org/10.1007/978-1-59745-522-0_9)

Crossref
Search Google Scholar
Export Citation
KaminskasEFarrellAAbrahamSBairdAHsiehLSLeeSLLeightonJKPatelHRahmanASridharaR 2005 Report from the FDA approval summary: azacitidine for treatment of myelodysplastic syndrome subtypes. Clinical Cancer Research 11 3604–3608. (https://doi.org/10.1158/1078-0432.CCR-04-2135)

Crossref
Search Google Scholar
Export Citation
KantarjlanHIssaJPJRosenfeldCSBennettJMAlbitarMDiPersioJKlimekVSlackJDe CastroCRavandiF 2006 Decitabine improves patient outcomes in myelodysplastic syndromes: results of a phase III randomized study. Cancer 106 1794–1803. (https://doi.org/10.1002/cncr.21792)

Crossref
PubMed
Search Google Scholar
Export Citation
KartalKOnderSKosemehmetogluKKilickapSTezelYGKaynarogluV 2015 Methylation status of TSHr in well-differentiated thyroid cancer by using cytologic material. BMC Cancer 15 824. (https://doi.org/10.1186/s12885-015-1861-1)

Crossref
PubMed
Search Google Scholar
Export Citation
KeelawatSThornerPSShuangshotiSBychkovAKitkumthornNRattanatanyongPBoonyayothinWPoumsukURuangvejvorachaiPMutiranguraA 2015 Detection of global hypermethylation in well-differentiated thyroid neoplasms by immunohistochemical (5-methylcytidine) analysis. Journal of Endocrinological Investigation 38 725–732. (https://doi.org/10.1007/s40618-015-0246-2)

Crossref
PubMed
Search Google Scholar
Export Citation
KikuchiYTsujiEYagiKMatsusakaKTsujiSKurebayashiJOgawaTAburataniHKanedaA 2013 Aberrantly methylated genes in human papillary thyroid cancer and their association with BRAF/RAS mutation. Frontiers in Genetics 4 271. (https://doi.org/10.3389/fgene.2013.00271)

PubMed
Search Google Scholar
Export Citation
KimWGZhuXKimDWZhangLKebebewEChengSY 2013 Reactivation of the silenced thyroid hormone receptor beta gene expression delays thyroid tumor progression. Endocrinology 154 25–35. (https://doi.org/10.1210/en.2012-1728)

Crossref
PubMed
Search Google Scholar
Export Citation
KitaharaCMDevesaSSSosaJA 2017 Increases in thyroid cancer incidence and mortality-reply. JAMA 318 390–391. (https://doi.org/10.1001/jama.2017.7910)

Crossref
PubMed
Search Google Scholar
Export Citation
Klein HesselinkENZafonCVillalmanzoNIglesiasCvan HemelBMKlein HesselinkMSMontero-CondeCBujRMauricioDPeinadoMA 2018 Increased global DNA hypomethylation in distant metastatic and dedifferentiated thyroid cancer. Journal of Clinical Endocrinology and Metabolism 103 397–406. (https://doi.org/10.1210/jc.2017-01613)

Crossref
Search Google Scholar
Export Citation
KochAJoostenSCFengZde RuijterTCDrahtMXMelotteVSmitsKMVeeckJHermanJGVan NesteL 2018 Analysis of DNA methylation in cancer: location revisited. Nature Reviews: Clinical Oncology 15 459–466. (https://doi.org/10.1038/s41571-018-0004-4)

PubMed
Search Google Scholar
Export Citation
KrauseKPrawittSEszlingerMIhlingCSinzASchierleKGimmODralleHSteinertFSheuSY 2011 Dissecting molecular events in thyroid neoplasia provides evidence for distinct evolution of follicular thyroid adenoma and carcinoma. American Journal of Pathology 179 3066–3074. (https://doi.org/10.1016/J.AJPATH.2011.08.033)

Crossref
Search Google Scholar
Export Citation
KunstmanJWKorahRHealyJMPrasadMCarlingT 2013 Quantitative assessment of RASSF1A methylation as a putative molecular marker in papillary thyroid carcinoma. Surgery 154 1255–1261; discussion 1261–1262. (https://doi.org/10.1016/j.surg.2013.06.025)

Crossref
PubMed
Search Google Scholar
Export Citation
KunstmanJWJuhlinCCGohGBrownTCStenmanAHealyJMRubinsteinJCChoiMKissNNelson-WilliamsC 2015 Characterization of the mutational landscape of anaplastic thyroid cancer via whole-exome sequencing. Human Molecular Genetics 24 2318–2329. (https://doi.org/10.1093/hmg/ddu749)

Crossref
PubMed
Search Google Scholar
Export Citation
LängstGManelyteL 2015 Chromatin remodelers: from function to dysfunction. Genes 6 299–324. (https://doi.org/10.3390/genes6020299)

Crossref
PubMed
Search Google Scholar
Export Citation
LatiniFRMHemerlyJPFreitasBCGOlerGRigginsGJCeruttiJM 2011 ABI3 ectopic expression reduces in vitro and in vivo cell growth properties while inducing senescence. BMC Cancer 11 11. (https://doi.org/10.1186/1471-2407-11-11)

Crossref
PubMed
Search Google Scholar
Export Citation
LaussMRingnérMKarlssonAHarbstKBuschCGeislerJLønningPEStaafJJönssonG 2015 DNA methylation subgroups in melanoma are associated with proliferative and immunological processes. BMC Medical Genomics 8 73. (https://doi.org/10.1186/s12920-015-0147-4)

Crossref
PubMed
Search Google Scholar
Export Citation
LawrenceMSStojanovPPolakPKryukovGVCibulskisKSivachenkoACarterSLStewartCMermelCHRobertsSA 2013 Mutational heterogeneity in cancer and the search for new cancer-associated genes. Nature 499 214–218. (https://doi.org/10.1038/nature12213)

Crossref
PubMed
Search Google Scholar
Export Citation
LeeJJGeliJLarssonCWallinGKarimiMZedeniusJHöögAFoukakisT 2008 Gene-specific promoter hypermethylation without global hypomethylation in follicular thyroid cancer. International Journal of Oncology 33 861–869. (https://doi.org/10.3892/ijo_00000074)

PubMed
Search Google Scholar
Export Citation
LiWVenkataramanGMAinKB 2007 Protein synthesis inhibitors, in synergy with 5-azacytidine, restore sodium/iodide symporter gene expression in human thyroid adenoma cell line, KAK-1, suggesting trans-active transcriptional repressor. Journal of Clinical Endocrinology and Metabolism 92 1080–1087. (https://doi.org/10.1210/jc.2006-2106)

Crossref
Search Google Scholar
Export Citation
LiLChoiJYLeeKMSungHParkSKOzeIPanKFYouWCChenYXFangJY 2012 DNA methylation in peripheral blood: a potential biomarker for cancer molecular epidemiology. Journal of Epidemiology 22 384–394. (https://doi.org/10.2188/jea.JE20120003)

Crossref
Search Google Scholar
Export Citation
LiWZhangXLuXYouLSongYLuoZZhangJNieJZhengWXuD 2017 5-Hydroxymethylcytosine signatures in circulating cell-free DNA as diagnostic biomarkers for human cancers. Cell Research 27 1243–1257. (https://doi.org/10.1038/cr.2017.121)

Crossref
PubMed
Search Google Scholar
Export Citation
LimHDevesaSSSosaJACheckDKitaharaCM 2017 Trends in thyroid cancer incidence and mortality in the United States, 1974–2013. JAMA 317 1338–1348. (https://doi.org/10.1001/jama.2017.2719)

Crossref
PubMed
Search Google Scholar
Export Citation
LinCHHsiehSYSheenISLeeWCChenTCShyuWCLiawYF 2001 Genome-wide hypomethylation in hepatocellular carcinogenesis. Cancer Research 61 4238–4243.

PubMed
Search Google Scholar
Export Citation
LinJGilbertJRudekMAZwiebelJAGoreSJiemjitAZhaoMBakerSDAmbinderRFHermanJG 2009 A phase I dose-finding study of 5-azacytidine in combination with sodium phenylbutyrate in patients with refractory solid tumors. Clinical Cancer Research 15 6241–6249. (https://doi.org/10.1158/1078-0432.CCR-09-0567)

Crossref
Search Google Scholar
Export Citation
LinQHouSGuanFLinC 2018 HORMAD2 methylation-mediated epigenetic regulation of gene expression in thyroid cancer. Journal of Cellular and Molecular Medicine 22 4640–4652. (https://doi.org/10.1111/jcmm.13680)

Crossref
PubMed
Search Google Scholar
Export Citation
LinnekampJFButterRSpijkerRMedemaJPvan LaarhovenHWM 2017 Clinical and biological effects of demethylating agents on solid tumours – a systematic review. Cancer Treatment Reviews 54 10–23. (https://doi.org/10.1016/j.ctrv.2017.01.004)

Crossref
PubMed
Search Google Scholar
Export Citation
ListerRPelizzolaMDowenRHHawkinsRDHonGTonti-FilippiniJNeryJRLeeLYeZNgoQM 2009 Human DNA methylomes at base resolution show widespread epigenomic differences. Nature 462 315–322. (https://doi.org/10.1038/nature08514)

Crossref
PubMed
Search Google Scholar
Export Citation
MadakashiraBPSadlerKC 2017 DNA methylation, nuclear organization, and cancer. Frontiers in Genetics 8 76. (https://doi.org/10.3389/fgene.2017.00076)

Crossref
PubMed
Search Google Scholar
Export Citation
MancikovaVBujRCastelblancoEInglada-PérezLDiezADe CubasAACurras-FreixesMMaravallFXMauricioDMatias-GuiuX 2014 DNA methylation profiling of well-differentiated thyroid cancer uncovers markers of recurrence free survival. International Journal of Cancer 135 598–610. (https://doi.org/10.1002/ijc.28703)

Crossref
PubMed
Search Google Scholar
Export Citation
ManiSHercegZ 2010 DNA demethylating agents and epigenetic therapy of cancer. Advances in Genetics 70 327–340. (https://doi.org/10.1016/B978-0-12-380866-0.60012-5)

Crossref
PubMed
Search Google Scholar
Export Citation
MarquezVEKelleyJAAgbariaRBen-KasusTChengJCYooCBJonesPA 2005 Zebularine: a unique molecule for an epigenetically based strategy in cancer chemotherapy. Annals of the New York Academy of Sciences 1058 246–254. (https://doi.org/10.1196/annals.1359.037)

Crossref
PubMed
Search Google Scholar
Export Citation
MassiminoMTirròEStellaSFrascaFVellaVSciaccaLPennisiMSVitaleSRPumaARomanoC 2018 Effect of combined epigenetic treatments and ectopic NIS expression on undifferentiated thyroid cancer cells. Anticancer Research 38 6653–6662. (https://doi.org/10.21873/anticanres.13032)

Crossref
PubMed
Search Google Scholar
Export Citation
MelckAMasoudiHGriffithOLRajputAWilkinsGBugisSJonesSJMWisemanSM 2007 Cell cycle regulators show diagnostic and prognostic utility for differentiated thyroid cancer. Annals of Surgical Oncology 14 3403–3411. (https://doi.org/10.1245/s10434-007-9572-8)

Crossref
PubMed
Search Google Scholar
Export Citation
MelkiJRVincentPCClarkSJ 1999 Concurrent DNA hypermethylation of multiple genes in acute myeloid leukemia. Cancer Research 59 3730–3740.

PubMed
Search Google Scholar
Export Citation
MiasakiFYVivaldiACiampiRAgatelLCollecchiPCapodannoAPincheraAEliseiR 2008 Retinoic acid receptor β2 re-expression and growth inhibition in thyroid carcinoma cell lines after 5-aza-2′-deoxycytidine treatment. Journal of Endocrinological Investigation 31 724–730. (https://doi.org/10.1007/BF03346422)

Crossref
PubMed
Search Google Scholar
Export Citation
MitmakerEJGriffNJGroganRHSarkarRKebebewEDuhQYClarkOHShenWT 2011 Modulation of matrix metalloproteinase activity in human thyroid cancer cell lines using demethylating agents and histone deacetylase inhibitors. Surgery 149 504–511. (https://doi.org/10.1016/j.surg.2010.10.007)

Crossref
PubMed
Search Google Scholar
Export Citation
Mohammadi-aslJLarijaniBKhorgamiZTavangarSMHaghpanahVKheirollahiMMehdipourP 2011 Qualitative and quantitative promoter hypermethylation patterns of the P16, TSHR, RASSF1A and RARβ2 genes in papillary thyroid carcinoma. Medical Oncology 28 1123–1128. (https://doi.org/10.1007/s12032-010-9587-z)

Crossref
Search Google Scholar
Export Citation
MohandasTSparkesRSShapiroLJ 1981 Reactivation of an inactive human X chromosome: evidence for X inactivation by DNA methylation. Science 211 393–396. (https://doi.org/10.1126/science.6164095)

Crossref
PubMed
Search Google Scholar
Export Citation
MonSYRiedlingerGAbbottCESeethalaROhoriNPNikiforovaMNNikiforovYEHodakSP 2018 Cancer risk and clinicopathological characteristics of thyroid nodules harboring thyroid-stimulating hormone receptor gene mutations. Diagnostic Cytopathology 46 369–377. (https://doi.org/10.1002/dc.23915)

Crossref
PubMed
Search Google Scholar
Export Citation
MoraesLGalrãoALRRubióICeruttiJM 2016 Transcriptional regulation of the potential tumor suppressor ABI3 gene in thyroid carcinomas: interplay between methylation and NKX2-1 availability. Oncotarget 7 25960–25970. (https://doi.org/10.18632/oncotarget.8416)

PubMed
Search Google Scholar
Export Citation
MoranSMartínez-CardúsASayolsSMusulénEBalañáCEstival-GonzalezAMoutinhoCHeynHDiaz-LagaresAde MouraMC 2016 Epigenetic profiling to classify cancer of unknown primary: a multicentre, retrospective analysis. Lancet: Oncology 17 1386–1395. (https://doi.org/10.1016/S1470-2045(16)30297-2)

Crossref
Search Google Scholar
Export Citation
MurgoAJ 2005 Innovative approaches to the clinical development of DNA methylation inhibitors as epigenetic remodeling drugs. Seminars in Oncology 32 458–464. (https://doi.org/10.1053/j.seminoncol.2005.07.004)

Crossref
PubMed
Search Google Scholar
Export Citation
NakamuraNCarneyJAJinLKajitaSPallaresJZhangHQianXSeboTJEricksonLALloydRV 2005 RASSF1A and NORE1A methylation and BRAFV600E mutations in thyroid tumors. Laboratory Investigation 85 1065–1075. (https://doi.org/10.1038/labinvest.3700306)

Crossref
Search Google Scholar
Export Citation
NeumannSSchuchardtKReskeAReskeAEmmrichPPaschkeR 2004 Lack of correlation for sodium iodide symporter mRNA and protein expression and analysis of sodium iodide symporter promoter methylation in benign cold thyroid nodules. Thyroid 14 99–111. (https://doi.org/10.1089/105072504322880337)

Crossref
PubMed
Search Google Scholar
Export Citation
NguyenCLiangGNguyenTTTsao-WeiDGroshenSLübbertMZhouJHBenedictWFJonesPA 2001 Susceptibility of nonpromoter CpG islands to de novo methylation in normal and neoplastic cells. Journal of the National Cancer Institute 93 1465–1472. (https://doi.org/10.1093/jnci/93.19.1465)

Crossref
PubMed
Search Google Scholar
Export Citation
NiuHYangJYangKHuangY 2017 The relationship between RASSF1A promoter methylation and thyroid carcinoma: a meta-analysis of 14 articles and a bioinformatics of 2 databases (PRISMA). Medicine 96 e8630. (https://doi.org/10.1097/MD.0000000000008630)

Crossref
Search Google Scholar
Export Citation
ParkSYSeoANJungHYGwakJMJungNChoNYKangGH 2014 Alu and LINE-1 hypomethylation is associated with HER2 enriched subtype of breast cancer. PLoS ONE 9 e100429. (https://doi.org/10.1371/journal.pone.0100429)

Crossref
PubMed
Search Google Scholar
Export Citation
PortelaAEstellerM 2010 Epigenetic modifications and human disease. Nature Biotechnology 28 1057–1068. (https://doi.org/10.1038/nbt.1685)

Crossref
PubMed
Search Google Scholar
Export Citation
ProvenzanoMJFitzgeraldMPKragerKDomannFE 2007 Increased iodine uptake in thyroid carcinoma after treatment with sodium butyrate and decitabine (5-aza-dC). Otolaryngology: Head and Neck Surgery 137 722–728. (https://doi.org/10.1016/j.otohns.2007.07.030)

Crossref
Search Google Scholar
Export Citation
QiMXiongX 2018 Promoter hypermethylation of RARβ2, DAPK, hMLH1, p14, and p15 is associated with progression of breast cancer: a PRISMA-compliant meta-analysis. Medicine 97 e13666. (https://doi.org/10.1097/MD.0000000000013666)

Crossref
PubMed
Search Google Scholar
Export Citation
QiuYYMirkinBLDwivediRS 2002 Differential expression of DNA-methyltransferases in drug resistant murine neuroblastoma cells. Cancer Detection and Prevention 26 444–453. (https://doi.org/10.1016/S0361-090X(02)00116-2)

Crossref
PubMed
Search Google Scholar
Export Citation
QiuYYMirkinBLDwivediRS 2005 Inhibition of DNA methyltransferase reverses cisplatin induced drug resistance in murine neuroblastoma cells. Cancer Detection and Prevention 29 456–463. (https://doi.org/10.1016/j.cdp.2005.05.004)

Crossref
PubMed
Search Google Scholar
Export Citation
ReikWCollickANorrisMLBartonSCSuraniMA 1987 Genomic imprinting determines methylation of parental alleles in transgenic mice. Nature 328 248–251. (https://doi.org/10.1038/328248a0)

Crossref
PubMed
Search Google Scholar
Export Citation
Riesco-EizaguirreGSantistebanP 2016 ENDOCRINE TUMOURS: Advances in the molecular pathogenesis of thyroid cancer: lessons from the cancer genome. European Journal of Endocrinology 175 R203–R217. (https://doi.org/10.1530/EJE-16-0202)

Crossref
Search Google Scholar
Export Citation
RiggsAD 1975 X inactivation, differentiation, and DNA methylation. Cytogenetic and Genome Research 14 9–25. (https://doi.org/10.1159/000130315)

Crossref
Search Google Scholar
Export Citation
Rodríguez-RoderoSFernandezAFFernandez-MoreraJLCastro-SantosPBayonGFFerreroCUrdinguioRGGonzalez-MarquezRSuarezCFernandez-VegaI 2013 DNA methylation signatures identify biologically distinct thyroid cancer subtypes. Journal of Clinical Endocrinology and Metabolism 98 2811–2821. (https://doi.org/10.1210/jc.2012-3566)

Crossref
Search Google Scholar
Export Citation
RussoDManoleDArturiFSuarezHGSchlumbergerMFilettiSDerwahlM 2001 Absence of sodium/iodide symporter gene mutations in differentiated human thyroid carcinomas. Thyroid 11 37–39. (https://doi.org/10.1089/10507250150500649)

Crossref
PubMed
Search Google Scholar
Export Citation
RyderMGhosseinRARicarte-FilhoJCMKnaufJAFaginJA 2008 Increased density of tumor-associated macrophages is associated with decreased survival in advanced thyroid cancer. Endocrine-Related Cancer 15 1069–1074. (https://doi.org/10.1677/ERC-08-0036)

Crossref
PubMed
Search Google Scholar
Export Citation
SaghafiniaSMinaMRiggiNHanahanDCirielloG 2018 Pan-cancer landscape of aberrant DNA methylation across human tumors. Cell Reports 25 1066.e8–1080.e8. (https://doi.org/10.1016/j.celrep.2018.09.082)

Search Google Scholar
Export Citation
SaiseletMFloorSTarabichiMDomGHebrantAvan StaverenWCMaenhautC 2012 Thyroid cancer cell lines: an overview. Frontiers in Endocrinology 3 133. (https://doi.org/10.3389/fendo.2012.00133)

PubMed
Search Google Scholar
Export Citation
SchagdarsurenginUGimmOHoang-VuCDralleHPfeiferGPDammannR 2002 Frequent epigenetic silencing of the CpG island promoter of RASSF1A in thyroid carcinoma. Cancer Research 62 3698–3701.

PubMed
Search Google Scholar
Export Citation
SchagdarsurenginUGimmODralleHHoang-VuCDammannR 2006 CpG island methylation of tumor-related promoters occurs preferentially in undifferentiated carcinoma. Thyroid 16 633–642. (https://doi.org/10.1089/thy.2006.16.633)

Crossref
PubMed
Search Google Scholar
Export Citation
Segura-PachecoBPerez-CardenasETaja-ChayebLChavez-BlancoARevilla-VazquezABenitez-BribiescaLDuenas-GonzálezA 2006 Global DNA hypermethylation-associated cancer chemotherapy resistance and its reversion with the demethylating agent hydralazine. Journal of Translational Medicine 4 32. (https://doi.org/10.1186/1479-5876-4-32)

Crossref
PubMed
Search Google Scholar
Export Citation
SharmaSKellyTKJonesPA 2010 Epigenetics in cancer. Carcinogenesis 31 27–36. (https://doi.org/10.1093/carcin/bgp220)

Crossref
PubMed
Search Google Scholar
Export Citation
ShiD-QAliITangJYangW-C 2017 New insights into 5hmC DNA modification: generation, distribution and function. Frontiers in Genetics 8 100. (https://doi.org/10.3389/fgene.2017.00100)

Crossref
PubMed
Search Google Scholar
Export Citation
SirajAKHussainARAl-RasheedMAhmedMBaviPAlsobhiSADSAl-NuaimAUddinSAl-KurayaK 2011 Demethylation of TMS1 gene sensitizes thyroid cancer cells to TRAIL-induced apoptosis. Journal of Clinical Endocrinology and Metabolism 96 E215–E224. (https://doi.org/10.1210/jc.2010-0790)

Crossref
Search Google Scholar
Export Citation
SkvortsovaKZotenkoELuuPLGouldCMNairSSClarkSJStirzakerC 2017 Comprehensive evaluation of genome-wide 5-hydroxymethylcytosine profiling approaches in human DNA. Epigenetics and Chromatin 10 16. (https://doi.org/10.1186/s13072-017-0123-7)

Crossref
Search Google Scholar
Export Citation
SmiragliaDJRushLJFrühwaldMCDaiZHeldWACostelloJFLangJCEngCLiBWrightFA 2001 Excessive CpG island hypermethylation in cancer cell lines versus primary human malignancies. Human Molecular Genetics 10 1413–1419. (https://doi.org/10.1093/hmg/10.13.1413)

Crossref
PubMed
Search Google Scholar
Export Citation
SmithJAFanCYZouCBodennerDKokoskaMS 2007 Methylation status of genes in papillary thyroid carcinoma. Archives of Otolaryngology: Head and Neck Surgery 133 1006–1011. (https://doi.org/10.1001/archotol.133.10.1006)

Crossref
Search Google Scholar
Export Citation
SormFVeselyJ 1968 Effect of 5-aza-2′-deoxycytidine against leukemic and hemopoietic tissues in AKR mice. Neoplasma 15 339–343.

PubMed
Search Google Scholar
Export Citation
SormFPiskalaACihakAVeselyJ 1964 5-Azacytidine, a new, highly effective cancerostatic. Experientia 20 202–203. (https://doi.org/10.1007/BF02135399)

Crossref
PubMed
Search Google Scholar
Export Citation
StephenJKChitaleDNarraVChenKMSawhneyRWorshamMJ 2011 DNA methylation in thyroid tumorigenesis. Cancers 3 1732–1743. (https://doi.org/10.3390/cancers3021732)

Crossref
PubMed
Search Google Scholar
Export Citation
StephenJKChenKMMerrittJChitaleDDivineGWorshamMJ 2015 Methylation markers for early detection and differentiation of follicular thyroid cancer subtypes. Cancer and Clinical Oncology 4 1–12. (https://doi.org/10.5539/cco.v4n2p1)

PubMed
Search Google Scholar
Export Citation
StephenJKChenKMMerrittJChitaleDDivineGWorshamMJ 2018 Methylation markers differentiate thyroid cancer from benign nodules. Journal of Endocrinological Investigation 41 163–170. (https://doi.org/10.1007/s40618-017-0702-2)

Crossref
PubMed
Search Google Scholar
Export Citation
SuelvesMCarrióENúñez-ÁlvarezYPeinadoMA 2016 DNA methylation dynamics in cellular commitment and differentiation. Briefings in Functional Genomics 15 443–453. (https://doi.org/10.1093/bfgp/elw017)

PubMed
Search Google Scholar
Export Citation
SwainJLStewartTALederP 1987 Parental legacy determines methylation and expression of an autosomal transgene: a molecular mechanism for parental imprinting. Cell 50 719–727. (https://doi.org/10.1016/0092-8674(87)90330-8)

Crossref
PubMed
Search Google Scholar
Export Citation
TimpWBravoHCMcDonaldOGGogginsMUmbrichtCZeigerMFeinbergAPIrizarryRA 2014 Large hypomethylated blocks as a universal defining epigenetic alteration in human solid tumors. Genome Medicine 6 61. (https://doi.org/10.1186/s13073-014-0061-y)

Crossref
PubMed
Search Google Scholar
Export Citation
TomasettiCLiLVogelsteinB 2017 Stem cell divisions, somatic mutations, cancer etiology, and cancer prevention. Science 355 1330–1334. (https://doi.org/10.1126/science.aaf9011)

Crossref
PubMed
Search Google Scholar
Export Citation
TorañoEGPetrusSFernandezAFFragaMF 2012 Global DNA hypomethylation in cancer: review of validated methods and clinical significance. Clinical Chemistry and Laboratory Medicine 50 1733–1742. (https://doi.org/10.1515/cclm-2011-0902)

PubMed
Search Google Scholar
Export Citation
ToyotaMAhujaNOhe-ToyotaMHermanJGBaylinSBIssaJP 1999 CpG island methylator phenotype in colorectal cancer. PNAS 96 8681–8686. (https://doi.org/10.1073/pnas.96.15.8681)

Crossref
PubMed
Search Google Scholar
Export Citation
TuncelMAydinDYamanETazebayUHGüçDDoğanALTaşbasanBUğurO 2007 The comparative effects of gene modulators on thyroid-specific genes and radioiodine uptake. Cancer Biotherapy and Radiopharmaceuticals 22 281–288. (https://doi.org/10.1089/cbr.2006.319)

Crossref
Search Google Scholar
Export Citation
van StaverenWCSolisDWDelysLDuprezLAndryGFrancBThomasGLibertFDumontJEDetoursV 2007 Human thyroid tumor cell lines derived from different tumor types present a common dedifferentiated phenotype. Cancer Research 67 8113–8120. (https://doi.org/10.1158/0008-5472.CAN-06-4026)

Crossref
Search Google Scholar
Export Citation
VenkataramanGMYatinMMarcinekRAinKB 1999 Restoration of iodide uptake in dedifferentiated thyroid carcinoma: relationship to human Na+/I- symporter gene methylation status. Journal of Clinical Endocrinology and Metabolism 84 2449–2457. (https://doi.org/10.1210/jcem.84.7.5815)

Search Google Scholar
Export Citation
Villar-GareaAFragaMFEspadaJEstellerM 2003 Procaine is a DNA-demethylating agent with growth-inhibitory effects in human cancer cells. Cancer Research 63 4984–4989. (https://doi.org/10.3389/fpsyg.2014.00533)

PubMed
Search Google Scholar
Export Citation
VitaleGDicitoreAPepeDGentiliniDGrassiESBorghiMOGelminiGCantoneMCGaudenziGMissoG 2017 Synergistic activity of everolimus and 5-aza-2′-deoxycytidine in medullary thyroid carcinoma cell lines. Molecular Oncology 11 1007–1022. (https://doi.org/10.1002/1878-0261.12070)

Crossref
PubMed
Search Google Scholar
Export Citation
VivaldiAMiasakiFYCiampiRAgateLCollecchiPCapodannoAPincheraAEliseiR 2009 Re-differentiation of thyroid carcinoma cell lines treated with 5-aza-2′-deoxycytidine and retinoic acid. Molecular and Cellular Endocrinology 307 142–148. (https://doi.org/10.1016/j.mce.2009.03.020)

Crossref
PubMed
Search Google Scholar
Export Citation
WangCMirkinBLDwivediRS 2001 DNA (cytosine) methyltransferase overexpression is associated with acquired drug resistance of murine neuroblastoma cells. International Journal of Oncology 18 323–329. (https://doi.org/10.3892/ijo.18.2.323)

PubMed
Search Google Scholar
Export Citation
WangPPeiRLuZRaoXLiuB 2013 Methylation of p16 CpG islands correlated with metastasis and aggressiveness in papillary thyroid carcinoma. Journal of the Chinese Medical Association 76 135–139. (https://doi.org/10.1016/j.jcma.2012.11.007)

Crossref
Search Google Scholar
Export Citation
WeisenbergerDJSiegmundKDCampanMYoungJLongTIFaasseMAKangGHWidschwendterMWeenerDBuchananD 2006 CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer. Nature Genetics 38 787–793. (https://doi.org/10.1038/ng1834)

Crossref
PubMed
Search Google Scholar
Export Citation
WhiteMGNagarSAschebrook-KilfoyBJasmineFKibriyaMGAhsanHAngelosPKaplanELGroganRH 2016 Epigenetic alterations and canonical pathway disruption in papillary thyroid cancer: a genome-wide methylation analysis. Annals of Surgical Oncology 23 2302–2309. (https://doi.org/10.1245/s10434-016-5185-4)

Crossref
PubMed
Search Google Scholar
Export Citation
WilsonASPowerBEMolloyPL 2007 DNA hypomethylation and human diseases. Biochimica et Biophysica Acta 1775 138–162. (https://doi.org/10.1016/j.bbcan.2006.08.007)

PubMed
Search Google Scholar
Export Citation
WuCtMorrisJR 2001 Genes, genetics, and epigenetics: a correspondence. Science 293 1103–1105. (https://doi.org/10.1126/science.293.5532.1103)

Crossref
PubMed
Search Google Scholar
Export Citation
WuSCZhangY 2010 Active DNA demethylation: many roads lead to Rome. Nature Reviews: Molecular Cell Biology 11 607–620. (https://doi.org/10.1038/nrm2950)

Crossref
PubMed
Search Google Scholar
Export Citation
WuXZhangY 2017 TET-mediated active DNA demethylation: mechanism, function and beyond. Nature Reviews: Genetics 18 517–534. (https://doi.org/10.1038/nrg.2017.33)

Crossref
PubMed
Search Google Scholar
Export Citation
WuWYangSFLiuFFZhangJH 2015 Association between p16 promoter methylation and thyroid cancer risk: a meta-analysis. Asian Pacific Journal of Cancer Prevention 16 7111–7115. (https://doi.org/10.7314/APJCP.2015.16.16.7111)

Crossref
Search Google Scholar
Export Citation
WuWZhangLLinJHuangHShiBLinXHuangZWangCQiuJWeiX 2016 Hypermethylation of the HIC1 promoter and aberrant expression of HIC1/SIRT1 contribute to the development of thyroid papillary carcinoma. Oncotarget 7 84416–84427. (https://doi.org/10.18632/oncotarget.12936)

PubMed
Search Google Scholar
Export Citation
XingMUsadelHCohenYTokumaruYGuoZWestraWBTongBCTalliniGUdelsmanRCalifanoJA 2003 Methylation of the thyroid-stimulating hormone receptor gene in epithelial thyroid tumors: a marker of malignancy and a cause of gene silencing. Cancer Research 63 2316–2321.

Search Google Scholar
Export Citation
XingMCohenYMamboETalliniGUdelsmanRLadensonPWSidranskyD 2004 Early occurrence of RASSF1A hypermethylation and its mutual exclusion with BRAF mutation in thyroid tumorigenesis. Cancer Research 64 1664–1668. (https://doi.org/10.1158/0008-5472.CAN-03-3242)

Crossref
PubMed
Search Google Scholar
Export Citation
YaneKKonishiNKitahoriYNaitoHOkaichiKOhnishiTMiyaharaHMatsunagaTHiasaY 1996 Lack of p16/CDKN2 alterations in thyroid carcinomas. Cancer Letters 101 85–92. (https://doi.org/10.1016/0304-3835(96)04117-1)

Crossref
PubMed
Search Google Scholar
Export Citation
YangXGaoLZhangS 2016a Comparative pan-cancer DNA methylation analysis reveals cancer common and specific patterns. Briefings in Bioinformatics 18 bbw063. (https://doi.org/10.1093/bib/bbw063)

Search Google Scholar
Export Citation
YangZWongAKuhDPaulDSRakyanVKLeslieRDZhengSCWidschwendterMBeckSTeschendorffAE 2016b Correlation of an epigenetic mitotic clock with cancer risk. Genome Biology 17 205. (https://doi.org/10.1186/s13059-016-1064-3)

Crossref
Search Google Scholar
Export Citation
YangXYZhangJYuXLZhengGFZhaoFJiaXJ 2018 Death-associated protein kinase promoter methylation correlates with clinicopathological and prognostic features in nonsmall cell lung cancer patients: a cohort study. Journal of Cancer Research and Therapeutics 14 S65–S71. (https://doi.org/10.4103/0973-1482.158197)

Crossref
Search Google Scholar
Export Citation
YegnasubramanianSHaffnerMCZhangYGurelBCornishTCWuZIrizarryRAMorganJHicksJDeWeeseTL 2008 DNA hypomethylation arises later in prostate cancer progression than CpG island hypermethylation and contributes to metastatic tumor heterogeneity. Cancer Research 68 8954–8967. (https://doi.org/10.1158/0008-5472.CAN-07-6088)

Crossref
PubMed
Search Google Scholar
Export Citation
YimJHChoiAHLiAXQinHChangSTongS-WTChuPKimBWSchmolzeDLewR 2019 Identification of tissue-specific DNA methylation signatures for thyroid nodule diagnostics. Clinical Cancer Research 25 544–551. (https://doi.org/10.1158/1078-0432.CCR-18-0841)

Crossref
Search Google Scholar
Export Citation
YooSKLeeSKimSJJeeHGKimBAChoHSongYSChoSWWonJKShinJY 2016 Comprehensive analysis of the transcriptional and mutational landscape of follicular and papillary thyroid cancers. PLoS Genetics 12 e1006239. (https://doi.org/10.1371/journal.pgen.1006239)

Crossref
PubMed
Search Google Scholar
Export Citation
YouJSJonesPA 2012 Cancer genetics and epigenetics: two sides of the same coin? Cancer Cell 22 9–20. (https://doi.org/10.1016/j.ccr.2012.06.008)

Crossref
PubMed
Search Google Scholar
Export Citation
ZafonCObiolsGCastellvíJRamon y CajalSBaenaJAMesaJ 2008 Expression of p21Cip1, p27Kip1, and p16INk4a cyclin-dependent kinase inhibitors in papillary thyroid carcinoma: correlation with clinicopathological factors. Endocrine Pathology 19 184–189. (https://doi.org/10.1007/s12022-008-9037-z)

Crossref
PubMed
Search Google Scholar
Export Citation
ZhangSWangYChenMSunLHanJElenaVKQiaoH 2017 CXCL12 methylation-mediated epigenetic regulation of gene expression in papillary thyroid carcinoma. Scientific Reports 7 44033. (https://doi.org/10.1038/srep44033)

Crossref
PubMed
Search Google Scholar
Export Citation
ZuoHGandhiMEdreiraMMHochbaumDNimgaonkarVLZhangPDiPaolaJEvdokimovaVAltschulerDLNikiforovYE 2010 Downregulation of Rap1GAP through epigenetic silencing and loss of heterozygosity promotes invasion and progression of thyroid tumors. Cancer Research 70 1389–1397. (https://doi.org/10.1158/0008-5472.CAN-09-2812)

Crossref
PubMed
Search Google Scholar
Export Citation