Received date: December 17, 2015; Accepted date: December 22, 2015; Published date: December 30, 2015
Citation: dos Reis MB, Beltrami CM, Kowaslki LP, et al. Epigenetic Alterations in well-Differentiated Thyroid Cancer. J Clin Epigenet. 2016, 1:1. DOI: 10.21767/2472-1158.100008
Copyright: © 2016 Silvia Regina Rogatto, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Genetic and epigenetic alterations have been associated with cancer development and progression. Important epigenetic modifications include posttranslational histone changes, covalent modifications of DNA bases and non-coding RNAs. Welldifferentiated thyroid carcinomas comprise papillary and follicular carcinomas, the most common subtypes. Methylation of thyroid-specific genes, including NIS and TSHR, genes involved in PI3K and MAPK pathways, as well as the tumor suppressor genes RASSF1A and PTEN and altered expression of microRNAs have been frequently reported in literature. Novel technologies using genome-wide approaches have contributed in revealing new epigenetically regulated genes involved in the molecular biology of thyroid cancer. The comprehensive analysis of genetic mutations, changes in DNA methylation, microRNA expression and posttranslational histone changes in thyroid carcinoma is critical to understand the crosstalk between these alterations and gene regulation that contributes to the development of diagnostic and prognostic tools, as well as new therapeutic strategies. The focus of this review is to present and discuss the current knowledge of molecular epigenetic data and their impact on thyroid carcinoma.
DNA methylation; MicroRNA expression; Histone PTMs; Thyroid carcinoma; Genome-wide methylation
Thyroid cancer (TC) is the most common endocrine gland malignancy, its incidence having increased over the last few decades . Well-differentiated carcinomas are derived from follicular cells and represent the majority of thyroid carcinomas , of which, papillary (PTC) and follicular (FTC) thyroid carcinomas are the most common histological subtypes representing 80-85% and 10-15% of cases, respectively .
Genetic and epigenetic alterations have been described as associated with thyroid carcinogenesis. Among the genetic alterations, mutations in effectors of the mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3-kinase (PI3KAKT) pathways have been frequently described. Rearrangements involving the RET gene (such as RET/PTC), RAS and BRAF mutations have been reported in more than 70% of PTC cases [3,4]. The most common genetic alteration is the BRAFV600E mutation (40-45% of PTC), which leads to a constitutive activation of BRAF and subsequent activation of downstream signaling transduction pathway.
In addition to genetic features, epigenetic events have been investigated in an attempt to understand other mechanisms involved in well-differentiated thyroid carcinoma. Changes in the DNA methylation pattern in genes involved in DNA damage repair, cell cycle control, tumor suppression and normal thyroid function have also been described [5-7].
In this review, the most relevant epigenetic alterations in well differentiated thyroid cancer have been described, focusing the discussion on DNA methylation and microRNA, particularly for papillary thyroid carcinoma.
Epigenetic alterations as a mechanism of gene regulation
Epigenetics is defined as the study of inherited alterations in gene regulation that do not involve changes in the DNA sequence [8,9]. The term “epigenetic” was first described by Waddington  as “the causal mechanisms between genes and their products, bring about phenotypic effects”. Cytosine DNA methylation along with chromatin modification  and regulation by noncoding RNAs  are considered effector mechanisms of epigenetic control of gene expression.
The most studied epigenetic mechanism is DNA methylation, which involves the addition of a methyl group to the 5′ position of cytosine, predominantly in the context of CpG dinucleotides . These CpGs are non-randomly distributed across the genome, including in gene promoter, gene body and intergenic region, as well as in repetitive sequences (satellite DNA and retrotransposons) [14,15]. CpG dinucleotide rich regions are known as CpG islands, which are present in approximately 60% of human genes [13,16,17].
Hypermethylation of CpG islands in promoter regions is a wellrecognized epigenetic event in tumorigenesis . Abnormal DNA methylation gains frequently occur in specific genes, which can be associated with loss of gene expression (reviewed by 19). Global loss of DNA methylation is associated with gene activation, loss of genomic imprinting and activation of DNA repetitive elements, resulting in enhanced chromosomal instability and consequently contributing to tumorigenesis [20,21].
Alterations in microRNA expression patterns also play an important role in cancer development [22,23]. MicroRNAs (miRNAs) are small non-coding RNAs that control the expression of their target genes, causing either degradation or inhibition of translation [24,25]. In tumor cells, microRNAs may function as oncogenes or tumor suppressor genes . As a result, modulation of many cellular pathways and biological process, such as those involved in embryogenesis, development timing, proliferation, apoptosis, angiogenesis, cell differentiation and other cancer-associated processes, may occur [27,28]. Similarly to classical genes, microRNAs can be subjected to epigenetic alterations including DNA methylation .
Additional disruptions of epigenetic mechanisms, such as histone modification patterns have also been reported in thyroid cancer. Histones are proteins involved in the packaging of eukaryotic DNA, comprised of chromatin. In addition to chromatin being an essential structural component, it is critical for gene regulation, DNA replication and repair, and chromosome segregation [30,31]. Post-translational modification (PTM) of histone tails includes acetylation, methylation, phosphorylation, ribosylation, sumoylation and ubiquitination . Depending on the position and type of modification that occurs in the histone, there is either activation or suppression of gene transcription. .
Gene promoter hypermethylation and inativation of tumor supressor genes are important mechanisms in thyroid cancer [34-37]. Hu et al.  reported that aberrant methylation gains in promoter regions of the TIMP3, DAPK, SLC5A8 and RARβ2 genes are associated with the BRAFV600E mutation and aggressiveness in papillary thyroid carcinoma. In PTC, similar results were reported for the RARβ2 and TIMP3 genes, as well as an inverse correlation between RASSF1A methylation and the presence of the BRAF mutation .
RASSF1A promoter methylation is one of the most common epigenetic events in human cancers, leading to gene silencing . RASSF1A promotor methylation is frequently described in follicular thyroid lesions including hyperplasia, adenoma and FTC [41,42]. Moreover, a high frequency of the RASSF1A methylation has been reported in both benign thyroid lesions (BTL; 93%) and PTC (76%) . RASSF1A hypermethylation has been suggested as an initial event associated with the development of TC [41,44].
Mutation or down-expression of PTEN, a tumor suppressor gene, is frequently reported as involved in thyroid tumorigenesis. The encoded protein acts as an antagonist of the PI3K pathway . Aberrant methylation of the PTEN promoter has been reported principally in follicular adenoma and carcinoma (83% and 84%, respectively) when compared to PTC (47%) . The PTEN hypermethylation has been associated with genetic alterations of the PI3K/AKT pathway, suggesting that gene inactivation caused by methylation may stimulate signaling pathways already activated by mutations and thus contributing to the progression of thyroid cancer .
Changes in the methylation patterns of other genes involved in the initiation or transduction of signaling in the PI3K/Akt and MAPK pathways have also been described as associated with TC (Figure 1). The tumor suppressor gene RASAL1 (RAS protein activator like-1) encodes a protein that acts as a negative modulator of the PI3K and MAPK pathways. Reduced expression of this gene mediated through CpG methylation has been described for several tumors with aberrant increased activity of the BRAF signaling pathway .
Figure 1 PI3K and MAPK protein network analysis (NAViGaTOR version 2.3). The interaction of proteins codified by PTEN, RASAL1 FGFR2 hypermethylated genes (in yellow) with several proteins of the PI3K and MAPK pathways in thyroid cancer. BRAF protein presents high level of protein interactions. Lines correspond to protein- protein interactions.
In thyroid cancer, 13 known negative modulators of the BRAF pathway (NF1, SPRY1, SPRY2, SPRED1, SPRED2, RKIP, DUSP5, DUSP6, TSC1, TSC2, LKB1, RASAL1, DAB2IP) were screened using a panel of 12 human-derived thyroid carcinoma cell lines . The only downexpressed genes identified in the tumor cell lines were DAB2IP and RASAL1. Of note, RASAL1 expression was absent in tumor cells, which was in stark contrast to normal tissue and normal thyroid cell lines. The RASAL1 gene was predominantly hypermethylated in FTC and ATC compared with BTL and PTC. Thus, based on these findings, RASAL1 alterations were suggested as affecting the regulation of the PI3K pathway over the MAPK pathway, acting as a major tumor suppressor gene in thyroid carcinoma .
Lin et al.  proposed that MIG6 acts as a tumor suppressor gene in PTC, negatively regulating the MAPK pathway. Promoter methylation was associated with MIG6 down-expression and high levels of EGFR and ERK phosphorylation. In a functional analysis, the increased expression of MIG6 inhibited proliferation and Mig-6 knockdown promoting the invasion of PTC cells . More recently, Lee et al.  investigated the DNA methylation pattern of MAPK signal-inhibiting genes. The promoter regions of DUSP4 and DUSP6 and the genes expressed in the cell lines TPC1, WRO82-1 and XTC were identified as unmethylated. In contrast, the TPC1 cell line and tumor tissues (63 of 76 cases) revealed the methylation of SERPINA5, in addition to an association with the BRAF mutation.
By using cDNA transfection in the WRO thyroid cell line, Kondo et al.  demonstrated that restoration of FGFR2-IIIb isoform expression was able to interfere in BRAF phosphorylation and decrease MAPK activation. Treatment with 5′-aza-deoxycytidine restored FGFR2-IIIb expression in WRO and TPC-1 thyroid cell lines, suggesting that methylation is a mechanism that affects FGFR2 regulation, consequently being involved in MAPK control .
The interaction between the PI3K and MAPK pathways has been well established, with epigenetic control of key factors of these pathways adding an additional layer of complexity. Overall, these data indicate that aberrant methylation of critical tumor suppressor genes may play a pivotal role in thyroid carcinogenesis.
In addition to the tumor suppressor genes, mismatch repair genes have also been described as hypermethylated in thyroid carcinoma. Guan et al.  analyzed the methylation pattern of 23 DNA repair genes in PTC samples. The authors observed methylation gains in MLH1, PCNA and OGG1 (21%, 13% and 5%, respectively). No difference was observed for MLH1 and MGMT methylation patterns in PTC, FTC and benign follicular thyroid adenomas . However, samples with mutations in BRAF, IDH1 and NRAS showed decreased MLH1 expression.
The normal function of thyroid cells requires a complex regulatory mechanism  for example, the TSHR, SLC26A4 and NIS genes involved in the uptake of iodide and thyroid hormone synthesis present aberrant methylation in their promoter regions. TSHR gene hypermethylation has been reported in 59% of PTC and 47% of FTC, while in normal and BTL samples it was unmethylated . The SLC26A4 gene encodes the pendrin protein, an anion exchanger responsible for the efflux of iodide in thyrocytes . Progressive hypermethylation of this gene has been described in a variety of lesions from adenomas to FTC, while a similar methylation pattern was seen in PTC and ATC .
Although Stephen et al.  evaluated a limited number of cases; they observed methylation of the sodium/iodide symporter (NIS) gene in PTC (11 cases), FTC (2 cases) and BTL (3 cases) when compared with normal tissue (5 cases). NIS gene promoter methylation was also reported in cold thyroid nodules, associated to mRNA down expression in 50% of hypermethylated samples .
Khan et al  described an association between TSHR promoter hypermethylation and the BRAFV600E mutation. TSHR hypermethylation was detected in 73.3% of BRAF-mutated tumors (12 PTC, 2 FTC and 1 with an unusual histology). The presence of BRAF mutation was also related to NIS down expression, mediated by DNMT1. In 30 PTC samples, Choi et al.  found significant NIS mRNA and protein down-expression in cases harboring the BRAFV600E mutation, leading the authors to suggest that high DNMT1 expression was associated with epigenetic silencing of NIS.
The findings described herein reveal that several genes with altered methylation patterns have been found in well-differentiated thyroid carcinoma. However, these DNA methylation studies are mostly restricted to candidate gene approaches. Only recently have large-scale studies addressing the gene methylation profile in thyroid cancer, most of which included PTC samples, been reported. To our knowledge, epigenetic profiles have been reported in six studies involving thyroid carcinomas (Table 1). In 2011, Hou et al.  assessed the global epigenetic alterations associated with the BRAFV600E mutation in two PTC cell lines (BCPAP and OCUT1) using shRNA knockdown. This strategy allowed the authors to describe 10 hypo- and 59 hypermethylated genes, respectively, in both cell lines following BRAF knockdown, revealing BRAFV600E as the driving force behind these genes becoming hyper or hypomethylated.
Genes or Probes
Genes or Probes
|Platform||Validation||Integrative analysis* (samples)|
|||2 PTC cell lines BRAF V600E knockdown||10 genes||59 genes||12K Human CpG-island Array chip (Microarray Center, Toronto, ON, Canada)||QMSP in the same cell lines||No|
4 cell lines
|PTCs - 13 probes
FTCs - 21
ATCs - 280
MTCs - 328
PTC and FTC cell line – 25 probes
MTC and ATC cell line – 598 probes
|PTCs – 262 probes
FTCs – 352
ATCs – 86
PTC and FTC cell line – 334 probes
MTC and ATC cell line – 198 probes
|Infinium HumanMethylation27k (Illumina)||Pyrosequencing independent samples||GEO data|
|PTCs - 0 gene||PTCs - 25 genes in 3 or more samples||Infinium HumanMethylation27k (Illumina)||Pyrosequencing independent samples||No|
|PTCs – 51 genes
FTCs – 77
FA – 9
|PTCs – 31 genes
FTCs – 416
FA – 83
|Infinium HumanMethylation27k (Illumina)||Bisulfite sequencing in test and validation set||Yes (31)|
|Classic PTC – 2582 probes
rPTCs – 2796
fvCPT – 405
|Classic PTC – 255 probes
rPTCs – 1023
fvCPT – 164
|Infinium HumanMethylation450k (Illumina)||-||No|
22 PTCs no annotated
2 groups enriched with H/K/NRAS mutations including most of
2 groups enriched for classical and tcvPTCs and BRAF mutations
|Infinium HumanMethylation450k (Illumina)||-||Yes (482)|
Table 1: Genome-wide methylation analysis in thyroid carcinomas.
Global DNA methylation analysis has been applied to identify molecular signatures in different subtypes of thyroid cancer. Rodríguez-Rodero et al. , using the 27K platform (Illumina) on a small number of samples (2 PTC, 2 FTC, 2 ATC, 2 medullary thyroid carcinomas -MTC, 2 normal samples and 4 cell lines), identified an epigenetic signature for PTC and FTC (262 and 352 hypermethylated genes and 13 and 21 hypomethylated genes, respectively). Comparing these findings with the alterations found in MTC and ATC revealed 155 and 210 hypermethylated genes and 13 and 21 hypomethylated genes exclusive to PTC and FTC, respectively. The authors identified several potential tumor suppressor genes (ADAMTS8 and HOXB4 in PTC and ZIC1 and KISS1R in FTC) and oncogenes (INSL4 and DPPA2 for MTC and TCL1B and NOTCH4 in ATC) confirmed as altered by methylation in thyroid carcinoma.
A similar platform (27K platform, Illumina) was used by Kikuchi et al.  and Mancikova et al  to describe the differential methylation profiles of TC. In the first study, 14 PTC samples were investigated according to BRAF and BRAF mutations. Mutated cases were more frequently methylated (nine of 11 samples), with six hypermethylated genes (HIST1H3J, POU4F2, SHOX2, PHKG2, TLX3 e HOXA7) being validated by pyrosequencing and presenting down-expression (except SHOX2) in at least one PTC cell line (TPC1, KTC1, and KTC3). Following treatment with 5-aza-2´deoxycytidine and/or trichostatin A, expression of these genes was restored in the cell lines . In the second study, 83 tumors (47 PTC, 18 FTC and 18 adenomas) were investigated for methylation, with the results being integrated with transcript expression data from 31 samples . The authors reported methylation profiles according to histological subtype and with the BRAF and BRAF mutation.
Ellis et al., using a more robust platform (450K platform, Illumina) on 51 cases of PTC, reported a global hypomethylation compared to normal tissue (91% of differentially methylated sites had a methylation loss) . The follicular variant showed a distinct methylation profile in comparison with the classical variant. Differences between the methylation profiles according to BRAF, BRAF and RET/PTC mutation (3.6 fold more differentially methylated compared with wild-type tumors) were also described.
Recently, a comprehensive analysis of the genome in more than 400 PTC samples from the TCGA project  revealed the involvement of several alterations associated with the development and progression of the disease. Analysis of genomic variants, transcript expression and DNA methylation revealed potential molecular drivers. Unsupervised clustering analysis with methylation data revealed four clusters of PTC classified as follicular, methylated CpG Islands, Classical 1 and Classical 2. In addition, two meta-clusters were identified according to the BRAFV600E and BRAF mutations. The BRAF mutation samples were detected in two different clusters, one with a lower number of methylation changes and the other with a large number of CpG hypermethylated in CpG islands and CpG shore regions, termed Meth-follicular and Meth-CpG Island cluster, respectively. The clusters named Classical 1 and 2 were enriched for classical and tall cell PTCs and BRAF mutations showing low levels of methylation regions outside the CpG Islands that would normally be methylated.
These results revealed changes in the methylation pattern of genes involved in DNA damage repair, and associated them with the normal function of thyrocytes. These recent approaches, as demonstrated by Ellis et al. , show that global hypomethylation is an important mechanism associated with PTC development. Moreover, these studies revealed the involvement of CpG island shores (regions with a low density of CpGs mapped at 2kb up or downstream of CpGs islands) and shelves (2 to 4kb up or downstream of CpG island shores) that have been reported as associated with gene expression regulation [69,70].
Although epigenetic alterations based on genome-wide methods have been described in well-differentiated thyroid carcinoma, a better understanding of the molecular alterations is necessary to guide the development of meaningful markers for early diagnosis, prognosis and prediction of response to therapies.
MicroRNA expression in thyroid cancer
In well-differentiated thyroid carcinomas, mainly PTC, numerous miRNAs have been described as having altered expression (Table 2). The miRNA signature based on expression levels has been used to differentiate distinct variants in PTC and FTC cases. In PTC, up-regulation of miR-21, 146b, 155, 181a, 181b, 221 and 222, among others, have been described [71-75]. In particular, the expression levels of miR-21, 146b, 181b, 221 and 222 were able to distinguish PTC from follicular adenoma and multinodular goiter . Differences in miRNA expression levels were described in the classical variant (miR-146b) and tall cell variant (miR-146b, -21 and -222) when compared with the follicular variant of PTC .
|Over expression||Down expression|
|miR-221, miR-222, miR-213, miR-220, miR-181b||let-7f-1, miR-142-as, miR-140-as, miR-199a-1, miR-151||PTC compared with normal tissue|
|miR-221, miR-222, miR-21, miR-31, miR-172, miR-34a, miR-213, miR-181b, miR-223, miR-224, miR-181a||miR-218, miR-300, miR-292-as, miR-345, miR-30c||PTC compared with MNG|
|miR-146b, miR-222, miR-221||-||PTC compared with FA miR-146b consistently overexpressed in PTC compared with non-PTC tumors (32 PTC, 24 FA, 11 HD, 2 FTC, 5NT)|
|miR-146, miR-221, miR-222, miR-21, miR-155, and miR-181a||mir-24-1, mir-9-3, mir-219-1, mir-138-1, mir-138-2, mir-345, mir-26a-1||PTC compared with normal tissue|
|let-7i-3p, miR-146-3p, miR-146-5p, miR-15a-3p, miR-181a-2-3p, miR-21, miR-221, miR-222, miR-222-5p, miR-34a, miR-34a-3p, miR-375, miR-542-5p, miR-551b, miR-99b-3p||miR-1179, miR-138, miR-144-5p, miR-199b-5p, miR-204, miR-219-5p, miR-451||fvPTC (17) compared to normal tissue|
|miR-192, miR-197, miR-328, miR-346||-||FTC compared with FA|
|miR-191||miR-191||Down-expressed in FA, FTC and fvPTC compared with NT,
Up-expressed in PTC
|-||let-7a||FA and FTC compared with NT|
|miR-146b, miR-221, miR-222, miR-155, miR-31||miR-153, miR-448, miR-325, miR-382, miR-495, miR-34b, miR-1, miR-144, miR-367, miR-130b, miR-138||PTC compared with NT
In bold,miR differentially expressed in aggressive PTC compared with non-aggressive PTC
|miR-125a-5p, miR-143, miR-23b, miR-425, miR-623, and miR-654 –3p||miR-1225–3p, miR-1238, miR-
135a, miR-150, miR-494, miR-663, miR-801, miR-923
|Comparison between tumor and normal tissues|
|miR-187, miR-221, miR-222, miR-146b, miR-155, miR-224, miR-197||-||Tumors vs. hyperplastic nodules in FNA samples|
|miR-21*, miR- 203||-||PTC BRAFV600E compared with PTC BRAF WT|
Table 2: MicroRNAs aberrantly expressed in well-differentiated thyroid carcinoma.
Alterations in microRNA expression were also associated with the mutational status of PTC. Up-regulation of miR-221 and miR-222 was demonstrated in samples with the BRAF and BRAF mutations, while overexpression of miR-146b was observed in cases carrying the BRAF mutation . The miR-221 expression observed in cases with the BRAF mutation was also associated with extra thyroidal invasion, lymph node metastasis and advanced stage of disease . Recently, using deep-sequencing, Mancikova et al.  found down-expression of miR-7 and -204 in PTC harboring the BRAFV600E mutation.
To understand the role of microRNAs in thyroid carcinogenesis, many studies have focused on functional assays to identify the targets of these molecules. In PTC cell lines, miR-221 was associated with control of cell proliferation . In addition, miR-221 and -222 potentially regulate the cell cycle targeting the CDKN1B transcript, which encodes for the p27kip1 protein . The authors showed that these miRNAs targeted the 3’UTR of CDKN1B resulting in reduced protein levels, regulating thyroid cell proliferation.
miR-181b, also up-regulated in PTC, directly targets the CYLD gene, which is associated with apoptosis . High expression levels of miR-21 were associated with cell proliferation and invasion, revealing a negative correlation with PDCD4 gene expression . Using in silico analysis, Jazdzewski et al.  showed that the 3’UTR of the THRβ gene contains binding sites for several microRNAs with aberrant expression in PTC. The authors used the CAKI-2 cell line (derived from renal carcinoma) and demonstrated a direct interaction between THRβ and miR-21 and -146a. Down-expression of this gene was detected following transfection assays with miR-21, -146a and -221.
In FTC, Weber et al.  reported up-regulation of miR-192, miR-197, miR-328 and miR-346 when compared with follicular adenoma. Overexpression of miR-197 and -346 was associated with cell proliferation. In addition, down-expression of two predicted targets of each miR (miR-197: ACVR1 and TSPAN3; miR-346: EFEMP2 and CFLAR) was found. Up-regulation of miR- 187, -224, -155, -222, and -221 were observed in conventional follicular thyroid carcinoma (cFTC), whereas in oncocytic follicular thyroid carcinoma (oFTC) up-regulation of miR-187, -221, -339, -183, -222, and -197 was found when compared with hyperplastic nodules. The authors also analyzed a panel of miRs in a small number of fine-needle aspiration biopsies (FNAB), observing high accuracy in detecting cancer samples . A higher expression of miR-96, -182, -183, -221, -222, -449a and -874 and lower expression of miR-542, -574, -455, and -199a were found in oFTC and cFTC when compared with normal tissue. miR-885 was revealed as a potential novel marker for oFTC .
Despite altered microRNA expression being well established in cancer, the molecular mechanisms underlying the regulation of these miRNAs are not completely understood. In general, miRNAs are mapped in regions involved in amplification and deletion as well as in breakpoints associated with other structural alterations, such as translocations [85,86]. Moreover, similarly to genes that encode for proteins, microRNAs are transcribed by genes that are targets for epigenetic events, which also act by modulating gene expression.
In 2011, Kunej et al.  reported that more than one hundred miRNAs were regulated by epigenetic mechanisms. The first epigenetically regulated microRNA described was miRNA-127 in bladder carcinoma cells . Following this discovery, several studies demonstrated alterations in the methylation pattern of microRNAs in gastric carcinoma , lung adenocarcinoma , gliobastoma  and colon cancer . Thyroid carcinoma is a relatively unexplored field, with only one study published by the TCGA group thus far, which revealed that microRNA genes were controlled by epigenetic mechanisms . miR-21, miR-146b, and miR-204 were aberrantly methylated in PTC and associated with altered expression of these microRNAs.
Taken together, the majority of studies have revealed miRNAs that are able to distinguish PTC and FTC from normal tissues, some of which present the potential to refine or strengthen strategies for diagnosis and management of TCs. However, the combined knowledge of the silenced miRNA genes associated with CpG island hypermethylation and those with high expression regulated by DNA hypomethylation are still an unexplored field in thyroid carcinoma.
Histone modification in thyroid cancer
Among the aberrant post-translational modifications in histones described in thyroid cancer, higher acetylation levels of lysine 9-14 (H3K9-14ac) and 18 (H3K18ac) of the H3 histone have been associated with early events in thyroid cancer progression. Puppin et al.  reported that levels of H3K18ac in non-differentiated TC were lower compared to FA, PTC and FTC, suggesting a switch off of this marker in the transition to more aggressive subtypes. In addition, high levels of H4K12ac were only found in FA.
The hypothesis of histone hypoacetylation of the NIS gene promoting its regulation and down-expression in papillary BHP 2-7 cells was reported by Kogai et al. . The authors treated the cell line with histone deacetylase (HDAC) inhibitor, trichostatin A (TSA) where no restoration of NIS expression or increased iodide accumulation by BHP cells was observed. On the other hand, Puppin et al.  demonstrated that histone acetylation controls the NIS promoter. The treatment of two papillary (BCPAP and TPC1) and two anaplastic (ARO and FRO) cell lines with TSA and Na butyrate (NaB) induced the expression of the NIS mRNA without modifying the expression of the thyroidspecific transcription factors TTF-1, TTF-2, PAX8, and HEX.
The potential of the aberrant histone PTM for treatment has been investigated. Lin et al.  evaluated a deacetylase inhibitor PXD101 as a potential treatment for thyroid cancer. Using cell lines derived from papillary, follicular, anaplastic and medullary carcinomas, the authors reported that PXD101 was efficient in inhibiting cell proliferation in all cell lines (particularly ATC cell lines), and suppressing HDACs in PTC (BHP7-13), FTC (WRO82-1) and ATC (8505C) cells depending on the dose used. The drug also induced apoptosis and double-strand DNA damage, as well as ROS accumulation (except in PTC cells), while inhibiting the BRAF/ RAF/ERK and PI3K/mTOR pathways (in ATC cell).
A phase I study with suberoylanilide hydroxamic acid (SAHA) was conducted in patients with advanced thyroid cancer. Kelly et al.  reported that HDAC inhibition promotes a partial response in patients with refractory PTC. Woyach et al.  conducted a phase II study with 16 patients with differentiated TC and 3 with medullary thyroid carcinoma receiving oral Vorinostat for 2 weeks followed by 1 week without therapy. Unfortunately, the drug was not effective for patients to achieve either partial or complete response at the dose tested.
Thus, future investigations on histone acetylation and other histone PTMs may be important for defining innovative therapeutic strategies for the treatment of thyroid cancer.
Methylation and clinical implications
Despite the significant number of epigenetic alterations described for thyroid cancer, reports of useful molecular prognostic markers for clinical practice are rare.
A strong association of DNA methylation in the TIMP3, DAPK, SLC5A8 and RARβ2 genes and classical and PTC tall-cell variants, in addition to tumor aggressiveness was demonstrated by Hu et al. . Guan et al.  reported a MLH1 methylation associated with lymph node metastasis and BRAFV600E mutation in PTC . Rap1GAP promoter hypermethylation and/or loss of heterozygosity were described in FA, PTC, FTC and ATC samples and its loss of expression was correlated with invasive and aggressive forms of thyroid cancer .
Several genes regulated by methylation have been described as associated with tumor recurrence. A recent study demonstrated that hypermethylation at specific sites of RUNX3 was capable of predicting recurrence in patients with PTC . The methylation levels at CpG sites -1397, -1406, -1415 and -1417 were able to predict prognosis with high specificity (85.1 to 95.7%), with negative predictive values of 86.1 to 88.9%, yet with low sensitivity (30.0 to 50.0%) . The recurrence was associated with unmethylated promoter regions of the TSHR gene in a study of 32 patients with PTC . Recently, Mancikova et al.  reported that methylation levels of EI24 and WT1 were associated with risk of recurrence in patients with PTC and FTC. A high percentage of recurrence-free survival was observed in patients with unmethylated genes .
Pathological grade, TNM stage and lymph node metastasis have been used to compare the methylation pattern. A significant association was described with methylation of the FHIT promoter regions in 42 PTCs and 23 FTCs . The authors also observed a difference in methylation pattern according to gender (higher in female) and histological classification (higher in FTC).
As highlighted in this review, alterations in DNA methylation, microRNA expression and aberrant histone PTMs play a role in the development of thyroid carcinoma. Most cases of DNA methylation reported in the literature were designed using analysis of strategies in the promoter region of candidate genes. There is still a very limited number of data obtained from global analysis, and even less so for miRNA methylation and histone PTMs. Therefore, a complete understanding of these alterations in TC remains to be clarified. Recent reports addressing the gene methylation profiles in appropriate sample numbers have contributed in revealing important drivers for aberrant methylation in these tumors. Two lines of evidence point out the relevance of microRNAs in TC. Firstly, recurrently altered expression in miRNAs that regulate several TC-associated transcripts and, secondly, the recent discovery of miRNA methylation. From this perspective, future studies are required to better explain the mechanisms involved in the process of thyroid carcinogenesis. Future exploratory research should focus on characterization of aberrant DNA methylation, microRNA and histone PTMs, and how this knowledge could contribute to the clinical management of well-differentiated thyroid carcinomas, in addition to identifying new targets for therapy.
This study was supported by grants from the National Institute of Science and Technology in Oncogenomics (Fundac?ao de Amparo a? Pesquisa do Estado de Sa?o Paulo – FAPESP Grant 2008/57887–9, Conselho Nacional de Desenvolvimento Cienti?fico e Tecnolo?gico Grant CNPq 573589/08–9 and fellowship CNPq (140819/2011-8).
The authors have nothing to disclose.
All Published work is licensed under a Creative Commons Attribution 4.0 International License
Copyright © 2019 All rights reserved. iMedPub LTD Last revised : October 19, 2019