Epigenetic Alterations in Parathyroid Cancers

Parathyroid cancers (PCas) are rare malignancies representing approximately 0.005% of all cancers. PCas are a rare cause of primary hyperparathyroidism, which is the third most common endocrine disease, mainly related to parathyroid benign tumors. About 90% of PCas are hormonally active hypersecreting parathormone (PTH); consequently patients present with complications of severe hypercalcemia. Pre-operative diagnosis is often difficult due to clinical features shared with benign parathyroid lesions. Surgery provides the current best chance of cure, though persistent or recurrent disease occurs in about 50% of patients with PCas. Somatic inactivating mutations of CDC73/HRPT2 gene, encoding parafibromin, are the most frequent genetic anomalies occurring in PCas. Recently, the aberrant DNA methylation signature and microRNA expression profile have been identified in PCas, providing evidence that parathyroid malignancies are distinct entities from parathyroid benign lesions, showing an epigenetic signature resembling some embryonic aspects. The present paper reviews data about epigenetic alterations in PCas, up to now limited to DNA methylation, chromatin regulators and microRNA profile.


Introduction
Tumors of the parathyroid glands frequently occur, showing a prevalence of 0.1%-0.4% in the general population, increasing up to 4% in postmenopausal women [1]. Parathyroid tumors are often associated with parathormone (PTH) hypersecretion determining primary hyperparathyroidism (PHPT), which represents the third most common endocrine disease following diabetes and thyreopathies. In PHPT, PTH hypersecretion due to tumor parathyroid cell proliferation induces hypercalcemia by increasing calcium mobilization from bone and calcium reabsorption from kidney. Most tumors of parathyroid glands are benign lesions, involving one or multiple parathyroid glands. Parathyroid benign tumor-related morbidity comes from PTH inappropriate and uncontrolled release.
Similarly to other human cancers, parathyroid carcinomas (PCas) display local vascular and tissue invasiveness, as well as metastatic localization. At variance with the most common human cancers, PCas are rare, accounting for approximately 0.005% of all cancers. Moreover, precancerous lesions, potentially evolving in cancer, have not been identified in parathyroid glands. These peculiarities suggest that parathyroid cells are highly resistant to cancer, though parathyroid cells are stimulated by several conditions, whose persistence induces cell proliferation: (1) pregnancy increases the rate of PTH release in the first trimester to face the embryonic needs; (2) chronic calcium and vitamin D deficiencies induce benign proliferative cell response; (3) congenital inactivating mutations of the calcium sensing receptor (CASR) gene are associated with parathyroid glands hyperplasia. However, none of these conditions are associated with PCa development.

DNA Methylation in Parathyroid Cancers
National and international mapping projects, such as those conducted by the U.S. National Institutes of Health (NIH) Roadmap Epigenomics Mapping Consortium [17], the International Human Epigenome Consortium [18], the Cancer Genome Atlas Network [19], European initiative to establish epigenomic maps of blood cells (BLUEPRINT) and the International Cancer Genome Consortium, have defined the genome-wide distribution of epigenetic markers in many fetal and adult normal and cancerous tissues. Unfortunately, no data about parathyroid normal or cancer tissues have been produced from the derived databases. Few studies have investigated epigenetic signature in PCas; however, available data provide interesting insights into parathyroid tumorigenesis. Here, we report these data attempting to define the landscape of epigenetic alterations in PCas.

Global DNA Methylation Pattern in Parathyroid Cancers
Gene silencing by DNA promoter cytosine phosphate guanine (CpG) islands methylation is the main and most well-studied epigenetic mechanism in humans, and it is intimately involved in cancer development [20]. DNA methylation is catalyzed by DNA methyltransferases (DNMTs), while the erasure of DNA methylation is achieved by ten-eleven translocation (TET) methylcytosine dioxygenases. TET enzymes convert 5-methylcytosine (5mC) to 5-hydroxymethylcytosine and subsequently to formyl or carboxyl cytosine. Then, modified cytosine is excised by thymine-DNA glycosylase (TDG) and repaired with an unmodified cytosine. The TET2 genes are frequently mutated in human cancers [21]. Spontaneous hydrolytic deamination of 5mC to thymine induces a high percentage of point mutations in germ or somatic cells, contributing to the genetic changes.
Comparing the DNA methylome profiles of seven metastatic PCas, whose CDC73 status was not investigated, and three normal parathyroid glands, 175 differentially-methylated genes were identified, while comparing the methylome profiles of PCas with those of 14 parathyroid benign adenomas, 263 genes with distinct methylation levels were detected [22]. Considering the top 100 differentially-methylated CpG islands, PCas showed the hypermethylation of all examined CpG islands with a pattern clearly distinguishable from that detected in normal parathyroid glands, which displayed low levels of CpG methylation. However, global hypermethylation were not detected in three PCas harboring CDC73 mutations analyzed using long interspersed nucleotide element-1 (LINE-1), a surrogate marker of genome-wide methylation changes [23]. Indeed, LINE-1 has been considered more sensitive in detecting decreases in DNA methylation rather than increases [24]. The authors admitted that the discrepancy could be explained by the different position of the analyzed CpG dinucleotides, the tumor heterogeneity, as well as the cut-offs and algorithms used for the detection of differential methylation used in both methods.
In line with the global hypermethylation detected by Starker and collaborators, a recent study reports a consistent reduction of the global 5-hydroxymethylcytosine (5hmC) levels in the analyzed PCas compared to parathyroid normal glands [25]. All 17 PCas stained negatively for 5hmC, as well as for TET1.

Regional DNA Methylation
Quantification of the methylation density by the pyrosequencing technique at the promoter of several candidate genes in CDC73-mutated PCas [23] showed that APC, SFRP1 and RASSF1A were hypermethylated, while CTNNB1/β-catenin was not affected. These findings confirmed previously reported hypermethylation of the APC gene promoter associated with reduced APC mRNA and protein expression levels and with increased levels of active unphosphorylated β-catenin in five PCas [22,26]. Other genes regulating the Wnt/β-catenin pathway have been found hypermethylated in PCas: SFRP1 [22,23], a potent antagonist of the Wnt signaling pathway, SFRP4 and SFRP2 [22]. It is worth noting that SFRP1 silencing by methylation can constitutively activate the Wnt signaling pathway. Furthermore, rat sarcoma (RAS) association family domain 1 (RASSF1A) is an important tumor suppressor gene involved in the regulation of cell growth and proliferation, as well as DNA repair and hypermethylation were frequently reported in many human cancers [27]. Hypermethylated genes in PCas included also [22]: (1) tumor suppressor genes involved in the cell cycle control, such as CDKN2B/p15INK4b, CDKN2A/p16INK4; (2) genes involved in apoptosis, such as PYCARD, SOCS3; (3) phosphatases, such as DUSP8 and PTPN20; (4) transcription factors, including HOXC11, WT1, GATA4 and HIC1 [28]. In particular, HIC1 encodes a transcriptional repressor shown to participate in complex regulatory loops resulting in increased p53 activation and inhibition of E2 transcription factor 1 (E2F1) through direct and indirect interactions with sirtuin 1 (SIRT1). In PCas, hypermethylation of HIC1 is associated with reduced gene expression, due to repressive histone H3K27me2/3 modifications induced by the polycomb repressor complex 2 (PRC2) member Enhancer of Zeste Homolog 2 (EZH2) [28].
No MEN1 or CDC73 promoter hypermethylation could be detected in PCas [23,29,30], despite a previous study reporting methylation of the CDC73 gene promoter in two out of 11 PCa samples [31]. These findings suggest that the mutational status of these two genes is unlikely to direct the tumors toward a different methylation profile. Similarly, no alteration of the CpGs methylation has been detected in the promoter region of the calcium sensing receptor (CASR) and vitamin D receptor (VDR) genes, two key molecules of parathyroid cells, conferring the sensitivity to extracellular calcium and precociously downregulated in parathyroid tumors [32,33].
A minority of aberrant methylated genes in PCas compared with normal parathyroid glands displayed hypomethylation [22]. In particular, hypomethylation of the promoter region of the microRNAs cluster on chromosome 19 (C19MC) has been detected in more than a half of PCas, though it did not correlate with C19MC microRNA expression levels [34].
The available data, summarized in Table 1, highlighted a specific "methylome" in PCas: (1) PCas are characterized by hypermethylation rather than reduced levels of methylation, consistent with the loss of tumor suppressor genes as a hallmark of parathyroid tumorigenesis; (2) All of the hypermethylated genes in PCas are hypermethylated also in benign parathyroid tumors; indeed, in PCas, the hypermethylation levels are more consistent; (3) PCas show hypermethylation of the promoter region of genes commonly hypermethylated in human cancers, namely CDKN2B/p15, CDKN2A/p16, SFRPs, RASSF1, HIC1 and APC [35]; (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway (Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway (Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway (Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].

Chromatin Regulators in Parathyroid Carcinomas
The genome sequencing efforts of thousands of uncultured tumors have revealed the frequent existence of mutations in writers, readers and erasers, thus establishing a causative role for an altered epigenome in cancer. Here, aberrant expression of chromatin regulatory molecules identified in PCas are reported.

Histones Modifications in Parathyroid Cancers
Replication-dependent histones are tightly regulated throughout the cell cycle. Interestingly, histone expression is mainly regulated by the processing of the 3′ end of histone transcripts.
In the absence of parafibromin in CDC73-mutated PCas, replication-dependent histone transcripts are not cleaved and contain a poly(A) tail. Loss of parafibromin in HCT116 cells and HeLa cells upregulates the replication-dependent histone family; of these, histones H1, H2A, H2B and H3 mRNAs are increased. The abnormally polyadenylated histone transcripts display an increased mRNA stability. Therefore, parafibromin emerges as a regulator of a posttranscriptional pathway critical to cell-cycle progression [39]. In line with these observations, histone H1.2 (histone cluster 1 H1 family member c; HIST1H1C (6p22.2)) has been reported up-regulated and highly overexpressed in seven sporadic PCas, together with the upregulation of histone H2A (HIST1H2AC Figure 1. The Wnt/β-catenin pathway is potentially deregulated in PCas. Schematic representation of the molecules involved in the Wnt/β-catenin signaling in the inactive (left) and active (right) state: molecules, whose expression may be affected by genetic and epigenetic modifications in PCas, are indicated.

Chromatin Regulators in Parathyroid Carcinomas
The genome sequencing efforts of thousands of uncultured tumors have revealed the frequent existence of mutations in writers, readers and erasers, thus establishing a causative role for an altered epigenome in cancer. Here, aberrant expression of chromatin regulatory molecules identified in PCas are reported.

Histones Modifications in Parathyroid Cancers
Replication-dependent histones are tightly regulated throughout the cell cycle. Interestingly, histone expression is mainly regulated by the processing of the 3 end of histone transcripts.
Moreover, in PCas there is loss of monoubiquitinated H2B at lysine 120 (K120)(H2Bub1). Parafibromin is required for the maintenance of H2B-K120 monoubiquitination [41] (Figure 2), while the level of H2B is consistently high in all parathyroid tumors independent of CDC73 expression. H2Bub1 is involved in RNA elongation, while losses of H2Bub1, as well as of nuclear CDC73 expression do not affect H3K4me3.

Aberrant Expression of Methyltransferases in Parathyroid Cancers
Unmethylated CpG islands are key factors in controlling H3K4me3 levels through recruitment of H3K4 methyltransferases. CpG islands likewise play an important role in establishing and maintaining H3K27me3 at bivalent domains. H3K27me3, a pivotal marker in the establishment of repressive chromatin in both early development and adult organisms, is activated by the polycomb group (PcG) PCR2. The histone 3 lysine 27 methyltransferase Enhancer of Zeste Homolog 2 (EZH2) is a member of PCR2. EZH2 mRNA and protein are overexpressed in PCas due to gene amplification [42], while EZH2 mutations have not been detected in 23 sporadic PCas [43]. EZH2 is involved in H3K27 methylation, and EZH2-mediated epigenetic control of RNA polymerase II (Pol II) transcribed coding gene transcription has been well established [44,45] (Figure 2). Moreover, the histone methyltransferase SUV39H1 is potentially deregulated in PCas as its recruitment and induction of H3K9 methylation are dependent on parafibromin [46] (Figure 2). Unfortunately, SUV39H1 expression and function have not been investigated so far in PCas.
Considering these limited data, PCas may be supposed to have elevated global H3K27me3 levels. Indeed, increased EZH2 activity redistributes the H3K27me3 marker across the genome with a complex effect on transcription, including a loss of H3K27me3 that is associated with increased transcription at many loci [47].

MicroRNAs Deregulated in Parathyroid Cancers
MicroRNAs (miRNAs), single-stranded non coding RNAs of 19-25 nucleotides in length, are drivers or suppressors of the hallmarks of malignant cells [48]. In general, cancer cells show a large alteration of miRNA expression compared to their normal counterpart. All tumors show a specific miRNA signature, referred to as "miRNome". Each miRNA is predicted to repress the expression of thousand mRNAs, but, in turn, each mRNA can be targeted by several hundreds of different miRNAs. Currently, there are about 1800 annotated human miRNA precursor genes that are processed into more than 2500 mature miRNA sequences (http://www.mirbase.org). miRNA signature has been investigated in PCas compared with normal parathyroid glands [34,49,50] and with parathyroid benign tumors [34,50]. These studies identified a global miRNA downregulation in PCas compared to normal parathyroid glands, reflecting a deregulated pattern common to human cancers. Nonetheless, the different technical strategies employed produced different sets of significantly deregulated miRNAs. Among the downregulated miRNAs, miR-296-5p [49], miR-139-3p [49], miR-126-5p [50], miR-26b [50] and miR-30b [50] were the most significantly varied in PCas. Few miRNAs were upregulated in PCas; the most significantly varied were miR-222 [49,50], miR-503 [49] and miR-517c [34,49] (Table 2).
PCa deregulated miRNAs could be detected with a similar pattern of expression in their distant metastasis: miR-296-5p was downregulated in lung metastasis of a PCa [34], and C19MC miRNAs and miR-372 were upregulated in samples derived from five PCa metastases [34].
A number of molecular mechanisms may deregulate miRNAs expression in PCas. Loss of parafibromin, which interacts with RNA polymerase II (Pol II) [51], may potentially alter miRNAs expression, being a key molecule in miRNA transcription ( Figure 2). Unfortunately, the relationship between loss of parafibromin in PCas and miRNAs expression have not been investigated so far. The transcriptional repression of miRNAs by hypermethylation of their corresponding promoter loci is a common feature of all human tumors [52]. Interestingly, PCas are characterized by the deregulation of microRNAs belonging to methylated genomic regions, namely the C19MC, miR-371-373 and GNAS loci.
C19MC cluster: PCas showed miR-519d, miR-518e and miR-517c median expression levels comparable with those in human placentas, the only human adult tissue physiologically expressing C19MC miRNAs. Two thirds of PCas showed gains in DNA copies either in the miRNA cluster regions or at a distant position along chromosome arm 19q, suggesting a genetic mechanism responsible for the aberrant expression of C19MC miRNAs at least in a subset of PCas [34]. Additionally, the C19MC promoter was hypomethylated in about half of PCa samples. However, the epigenetic status of the C19MC promoter was not correlated with miRNAs expression levels [34]. C19MC is the largest human miRNA gene cluster. It maps on chromosome 19q13.42, spanning about 100 kb and consisting of 46 genes encoding a total of 56 mature miRNAs [53]. miRNAs belonging to the C19MC cluster are expressed in human embryonic stem cells and rapidly downregulated during the differentiation process [54]. Expression of C19MC microRNAs increases significantly in trophoblasts from the first to third trimester of gestation [55], while they are silenced in the majority of adult normal cells by hypermethylation [56]. C19MC miRNAs are aberrantly expressed in infantile hemangioma [57], melanotic neuroectodermal tumor of infancy [58], embryonal tumors with multilayered rosettes [59], testicular germ cell tumors (in particular, in non-seminoma aggressive tumors) [60], glioblastoma [61], hepatic mesenchymal hamartomas [62], suggesting that C19MC reactivation characterizes tumors with embryonic features [63].
miR-372: miR-372, which is highly expressed in almost all PCas, belongs to the miR-371-373 cluster on chromosome 19q13.42, which is close to the C19MC cluster. Similarly to C19MC miRNAs, the miR-371-373 cluster is an "embryonic" cluster. miR-372 is deregulated in many human cancers [64]. miR-296: miR-296-5p is located at the imprinted locus GNAS (guanine nucleotide binding protein (G protein), α stimulating activity polypeptide 1) on chromosome 20q13. miR-296-5p is expressed from the paternally-derived allele, arising from the long, noncoding antisense transcript, GNAS antisense RNA 1 (GNAS-AS1) [65]. It is a tumor suppressor in breast and lung cancers [66,67]. In PCas, miR-296-5p is downregulated, showing the best predictive value in distinguishing cancers from normal parathyroid glands [49]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway (Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway (Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway (Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].  (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway (Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38]. (4) The promoter regions of the tumor suppressor genes known to be involved in parathyroid tumorigenesis, namely CDC73, MEN1, CASR and VDR, are not affected by increased methylation; therefore, methylation is not the major molecular mechanism inducing their loss in parathyroid tumor cells; (5) Hypermethylation affects the promoter of genes encoding molecules of the Wnt/β-catenin pathway ( Figure 1). Wnt/β-catenin is potentially deregulated in PCas by loss of parafibromin, which is a member of the polymerase II complex interacting with β-catenin [36]. Therefore, Wnt/β-catenin deregulation has been suggested as a "hub" of parathyroid tumorigenesis [37] ( Figure 1). Indeed, accumulation of β-catenin is controversial in PCas, with studies reporting constitutive accumulation of active unphosphorylated β-catenin [26] and others failing in the detection of total β-catenin at the nuclear level [38].