Abstract
Background/Aim: Anaplastic thyroid cancer (ATC) is one of the most aggressive human malignancies, remaining generally incurable. Histone deacetylase (HDAC) seems to play a role in regulating transcription of genes involved in ATC, making HDAC inhibitors (HDACI) promising anticancer drugs for ATC. The purpose of this review was to evaluate the role of HDACIs in ATC treatment and describe the latest trends of current research on this field. Materials and Methods: This literature review was performed using the MEDLINE database. The keywords/phrases were; thyroid cancer, anaplastic, HDAC, histone, deacetylase*, HDACI. Results: Compounds, such as SuberoylAnilide Hydroxamic Acid, valproic acid, sodium butyrate, butyrate, phenylbutyrate, trichostatin A, AB1-13, panobinostat or LBH589, belinostat, MS-275, depsipeptide, CUDC101, CUDC907, N-Hydroxy-7-(2-naphthylthio)-Hepanomide (HNHA), and PXD101 have shown promising antitumor effects against ATC. Conclusion: HDACIs represent a promising therapy for ATC management, both as monotherapy and in combination with other anticancer drugs.
Anaplastic thyroid carcinoma (ATC), is one of the most aggressive human malignancies. It is an undifferentiated thyroid cancer responsible for most thyroid cancer deaths, with an overall survival rate as low as 13%, despite accounting for only about 1% of the thyroid cancer incidence (1-4). Furthermore, early tumor dissemination results in 20-50% of patients having distant metastases and in 90% having adjacent tissue invasion on presentation (5). Thus, most often patients present with mass effects that cause compressive symptoms, without hormone-related symptoms (6).
ATC is refractory to current conventional treatments with radioiodine ablation and chemotherapy, while radiation therapy with thyroidectomy is being considered only for airway decompression (6). The median survival from diagnosis ranges from 3 to 7 months and the one- and five-year survival rates are 20-35% and 5-14%, respectively (7-11). Therefore, due to its oncological behavior, there is an urgent need for novel, more innovative immunological and gene therapies (12).
Histone deacetylase inhibitors (HDACIs) are a promising class of drugs in thyroid cancer. Histone deacetylases (HDACs) are enzymes that remove the acetyl groups, a process called deacetylation, from histone lysine residues. The process of deacetylation creates more positive charges on the histones and thus increases the interaction of the positively charged N termini of histones with the negatively charged phosphate groups of DNA. These interactions transform DNA into a more condensed form, making it less accessible to the cell's transcriptional machinery (13, 14). Deacetylation-mediated inhibition of tumor suppressor genes leads to the progression of tumors (15).
There are many types of HDACs, mainly divided into four classes: (a) class I includes HDACs 1, 2, 3 and 8; (b) class II includes HDACs 4, 5, 6, 7, 9 and 10; (c) class III HDACs are the sirtuins (SIRT1-7); and (d) class IV contains HDAC 11 (16, 17). Different HDAC subtypes are considered to be associated with different cancer characteristics and behaviors like HDAC-1, -4, -6 being closely related to tumor size; HDAC-4 is also related to capsular invasion, and HDAC-2 is considered to play a role in lymphatic and vascular invasion of malignant neoplasms (18).
HDACIs have been extensively studied as potential drugs for treating cancer with vorinostat (SuberoylAnilide Hydroxamic Acid) being the first to receive approval for clinical use in cutaneous T-cell lymphoma (19). One of the first well-studied HDACIs for treating thyroid malignancies is suberoylanilide hydroxamic acid (SAHA), which has been already approved by the FDA for the treatment of several neoplastic diseases (20, 21) and has antitumor activities against thyroid cancer (22).
HDACIs can induce tumor growth arrest, differentiation, and apoptosis (23). Moreover, they can sensitize tumor cells to radiation, increase radioiodine uptake and intratumoral radioiodine accumulation, organification, and chemotherapy (24). However, the most crucial role of these drugs is the deacetylation of the NH2 terminal of lysine residues leading to the change of DNA configuration regulating gene expression (25).
In vitro and in vivo studies have shown their effectiveness against human thyroid cancer cell lines. Even though there is only one report of a randomized multicenter phase II/III clinical trial testing the effect of HDACIs in patients with ATC (26), more than 50 articles describing the effects of HDACIs in cell lines have shown promising results. These drugs are being tested either alone or in combination with other cytotoxic or targeted cancer treatments on human thyroid cell lines in vitro. The cell lines that were mostly used are ARO cells, ARO81-1, FRO, DR090-1, CAL-62, 8305C, GSA1 and 2, BHT-101, SW1736, and C643. Each cell line has unique features, expresses different genes and has a different response to each medication. Furthermore, some studies presented results from cell lines which derived from human cancer biopsies and expressed a unique set of different genes. The main characteristics of the different cell lines are summarized in Table I.
Some data suggest that single-agent HDACIs may not be a viable treatment when used alone for ATC (27). As such, a number of studies suggest the usage of two or more HDACIs or the combination of HDACIs with tyrosine kinase inhibitors (TKI) in order to maximize the therapeutic effect (28). This review discusses all the main compounds used for the treatment of ATC, their mechanism of action and their effect on tumor progression. These include: SuberoylAnilide Hydroxamic Acid, valproic acid, sodium butyrate, butyrate, phenylbutyrate, trichostatin A, AB1-13, panobinostat or LBH589, belinostat, MS-275, depsipeptide, CUDC101, CUDC907, N-Hydroxy-7-(2-naphthylthio)-Hepanomide (HNHA), and PXD101. The major mechanisms of action of these agents are summarized in Table II.
SuberoylAnilide Hydroxamic Acid
SuberoylAnilide Hydroxamic Acid (SAHA) or Vorinostat is a hydroxamic acid and anilide derivative that acts as an HDACI of class I and II HDACs. It has antineoplastic activity on many cancer cells including anaplastic thyroid cancer cells. Treatment of these cancer cells with 5 μM of SAHA for 30 min induced histone acetylation, leading to increased expression of Na/I symporter (NIS), thyroid stimulating hormone receptor (TSHR), thyroid peroxidase (TPO) and thyroglobulin (Tg) (29, 30). Their effect on SW1736, Hth7 and C643 thyroid cells was increased when they were combined with RDEA119, a MEK inhibitor, temsirolimus, an mTOR inhibitor, and perifosine, an AKT inhibitor. With the addition of TSH, there was an up-regulation of NIS, TSHR, TPO, Tg, PAX8, FOXE1, and TTF1. In addition, the combination of SAHA, perifosine and TSH increased the sensitivity of iodide uptake (29). Furthermore, SAHA down-regulated HuR via decreasing NF-kB on SW1736 cells (31).
On the other hand, when ATC 8505C cells were treated with 10 nM SAHA for 48 h, the concentration of NIS remained stable, but when used on C643, there was a strong re-expression of NIS. Interestingly, when the cell line's basal NIS concentration was low, like in C643 cell line, SAHA caused re-expression of NIS leading to increased protein levels. However, when the NIS concentration was not low from the beginning, the concentration remained stable after the treatment. The same effect was observed when the cell lines were treated with Trichostatin A and Panobinostat (32).
Its anti-proliferative effects and the up-regulation of TSHR mRNA were potentiated by the PARP inhibitor PJ34 (33). However, when used in combination with 1,25-dihydroxy vitamin D3, this additive effect was only observed in Hth74 and C643 cell lines and ranged from 25% to 35% inhibition of growth (34). In FRO cells, SAHA induced apoptosis via inhibiting degradation and thus increasing TRAIL protein levels (same mRNA levels), especially when combined with the MS-275 HDACI (35).
Also, SAHA caused cell death by suppressing PI3K/Akt/mTOR, inactivating survivin, an anti-apoptotic protein, and activating of DNA damage-related proteins via decreasing acetylation of H3, H4. When used in combination with SNX5422, a heat shock protein 90 inhibitor, these effects were potentiated (36).
Valproic Acid
Valproic acid (VPA) is a branched fatty acid that is commonly used for its anticonvulsant activity in seizures, epilepsy. However, a new indication for VPA as an HDACI was discovered, specifically for the treatment of malignancies including thyroid neoplasms. In vitro studies have shown that VPA increases expression of NIS in ARO cells (37) and E-cadherin expression on BHT101, and 8350C cells (38). On the other hand, studies on 8505C cells have shown that VPA and Sodium Butyrate, another HDACI, increase the sensitivity of ATC cells to radiation via impairing DNA damage repair processes, apart from increasing NIS concentration (39). Moreover, it increases p21 mRNA and induces G1 cell-cycle arrest only when used in combination with imatinib in ARO cells. When used alone, neither VPA nor imatinib induced a similar effect (40). The combined action of VPA and doxorubicin on ARO and Cal-62 cells caused increased cytotoxicity, halting apoptosis and increasing accessibility of DNA. Their combination also resulted in G2 arrest and reduced production of free radicals, suggesting fewer complications due to doxorubicin toxicity (41).
However, in contrast to in vitro human cell studies, in a phase II/III clinical trial of the combination of VPA with paclitaxel (TAX) in patients with anaplastic thyroid cancer along, there was no benefit in the final outcome nor in TAX's pharmacokinetics (26).
Trichostatin A
Trichostatin A (TSA), is a commonly used antifungal drug that can also inhibit HDACs and, therefore can be used in the treatment of different malignancies. Furthermore, TSA sensitized SW579 cells to radiation with iodide but to a lesser extent than silencing HDACs with shRNA (short-heparin RNA) (42). When combined with SNX5422, they synergistically suppressed PI3K/Akt/mTOR and survivin, and activated DNA damage-related proteins (36). When used along with sodium butyrate, another HDACI, they increased NIS expression in FRO, ARO, and SW1736 cell lines and expression of PAX8 only in ATC ARO cells (43). Furthermore, the combination of TSA (300 nM) with sodium butyrate (3 mM) suppressed HDAC-1 and 2 and increased mRNA alternative splicing of prohibitin (PHB), a ubiquitous protein, in FRO thyroid cells. The PHB transcript is normally spliced into two isoforms, the 3’UTR and PHB SHORT. However, the treatment combination of TSA and sodium butyrate induced a specific increase in the 3’UTR isoform through alternative splicing. 3’UTR's action, though, encodes for another isoform that contains an untranslated sequence that eventually induces a proliferation blockage (44). TSA induced tumor suppression and this effect was independent of p53. Also, it increased apoptosis by caspases activation and a decrease in cyclin A, B levels and also increased the levels of cyclin-dependent kinases inhibitors p21, p27. These effects result in cell cycle arrest at the G1 phase in DR090-1 cells along with a pRb hypophosphorylation and cell cycle arrest at the G2 phase in ARO81-1 cells (45).
Moreover, TSA increased the levels of Rap1GAP and RAP2, regulators of HDAC expression (46), and modulated E-cadherin expression on BHT101, CAL62 and 8305C cells (38).
N-Hydroxy-7-(2-naphthylthio)-Hepanomide
N-hydroxy-7-(2-naphthylthio)-Hepanomide (HNHA) is a potent HDACI and is being tested in many different cancer types including breast cancer, renal cell carcinoma, medulloblastoma, and anaplastic thyroid cancer. Specifically, on anaplastic thyroid cancer cells. It inhibited progression to G0/G1 phase leading to cell-cycle arrest, reduction of Bcl-2 and Ki-67 and apoptosis. Moreover, along with lenvatinib, an FGFR inhibitor, and sorafenib, a tyrosine kinases inhibitor, HNHA was more toxic and decreased proliferation of GSA1, GSA2 cells. The combination with lenvatinib resulted in inhibition of the FGFR pathway, a decrease in PKC, MEK, p-ERK1/2, MET and a dramatical increas in E-cadherin more than HNHA alone or in combination with sorafenib (47). HNHA induced an endoplasmic reticulum stress-dependent cell cycle arrest, an increase in p53 and p21, and a decrease in cyclin D1, CDK4 and CDK6 levels in 8505X, SNU80 and GSA1 anaplastic thyroid cells. When combined with sorafenib, cell viability was further decreased and cells arrested at G0/G1 phase, Apaf-1, NF-kB, p65, and Bcl-2 levels increased, and apoptosis was induced through the cleavage of caspase-3 (48). HNHA induced acetylation of H3 and compared to TSA and SAHA, it led to a more significant reduction in cell proliferation along with increased cytotoxicity. Finally, HNHA induced a higher increase in the levels of cytoplasmic calcium than the other drugs and led to a stronger induction of apoptosis (49).
Butyrate
Butyrate is a short-chain fatty acid that competitively binds to the zinc sites of class I and II HDACs inducing cell cycle arrest in G1 or G2/M phase. Specifically, in anaplastic thyroid cancer decreased the levels of cyclin A and B, and by increasing cyclin-dependent kinase inhibitors p21 and p27 induced cell cycle arrest at G1 phase in both DR090-1 and ARO81-1 cells. Furthermore, butyrate induced caspase-dependent apoptosis independently of p53 (45).
Sodium butyrate (NaB), the most commonly used, increases Rap1GAP protein levels which are commonly low in cancers. It also decreased the expression of RAP2 via inhibiting regulators like RapGEFs and RapGAPs in Hth7, Hth74, Hth83, Hth104, Hth112 and SW1736 cells (46). In combination with TSA, they increased NIS expression in FRO, ARO, and SW1736 cells and expression of PAX8 only in ARO cells (43).
The combination of phenylbutyrate (Phb), another butyrate analogue, with COX inhibitors and cisplatin derivatives resulted in higher levels of cytotoxicity compared to both cisplatin and oxaliplatin. They were 51 times more selective for thyroid cells than cisplatin. The use of VPA instead of Phb led to the same results with the exception that the VPA effects were not exerted through HDAC inhibition (50).
CUDC101
CUDC-101 is a novel inhibitor of class I HDACs. The combination of CUDC-101 and carfilzomib resulted in higher inhibition of cell growth and proteasome in 8505C, SW1736, C643 anaplastic thyroid cells compared to treatment with each one separately. Specifically, in 8505C and C-643 the two medications synergistically induced cell cycle arrest at G2/M. What is more, they dramatically increased expression of p21 and PARP, an apoptotic marker cleaved by caspase 3 (51).
CUDC907
CUDC907 is both a tyrosine kinase inhibitor and an HDACI. It increases the acetylation of H3 and decreases HAC2 protein levels. In 8505C thyroid anaplastic cancer cells, it decreased cell migration and invasion by decreasing the expression of TWIST-1 protein levels, which normally repress E-cadherin's expression. Moreover, it induced cell cycle arrest at G2/M via increasing p21 expression and reducing expression of cyclin B1, AURKA, B, and PLK1. Furthermore, CUDC907 increased the apoptotic rate by inducing caspase 3/7 activity and increasing p27 and survivin protein levels by increasing caspases 3 and 9 activity. In addition, it decreased the PI3K/Akt pathway and phospho-ERK protein levels leading to reduced activation of the RAF/RAS/MEK/ER pathway (52).
Depsipeptides
Depsipeptides (FK228) are bicyclic peptides mainly produced by Chromobacterium violaceum during fermentation (53). They have been mainly tested on ARO anaplastic thyroid cancer cells and there is evidence that they induced re-expression of NIS, TPO, Tg mRNAs and increased radioiodide uptake. The re-sensitization was not caused by TSH because its mRNA levels were low but probably by increasing NIS protein levels. Depsipeptides increased TTF-1 levels but not the levels of Pax-8 and showed cytotoxicity only at higher doses, at around 3-10 ng/ml (54).
AB1-13
AB1-13 are HDACIs that have conserved hydroxamic acid metal chelating groups that bind to the zinc moiety on HDACs and a conserved anisole hydrophobic cap. AB2, 3, 10 induce apoptosis and cell cycle arrest via increasing the levels of cyclin-dependent kinase inhibitor p21, p27 and decreasing the levels of cyclin D1. Moreover, they increased cleaved PARP and cleaved caspase 3 in 8505C cells. In Hth and 8505C cells, the AB2, 3, 10 increased mRNA levels of NIS. In Hth cells there was an increase in TSHR after the use of all members of the AB family but in 8505C cells, only AB5 and 10 had this effect. AB2, 3, 10 led to increased mRNA levels of PAX8 and TTF2 in Hth cells. AB3 was more selective for HDAC1, 2, 3 just like AB2, but was less potent than the latter (55).
Panobinostat
Panobinostat (LBH589), is a non-selective HDACI hydroxamic acid that has shown promising results in many experiments with anaplastic thyroid cancer cells. It led to increased apoptosis and cell cycle arrest not through Bcl2, Bcl-xl or Bax and Bak. It increased the number of cells at the G1 phase of the cell cycle. Furthermore, panobinostat increased p21 expression and reduced the levels of cyclin D1 in Cal-62 and 8505C cells (56). Importantly, treatment of BRAFV600+ ATC cells of an 83-year-old patient at stage IVC pT4aN1aM1 along with panobinostat and sorafenib, resulted in reduction of cell viability by nearly 100%. It was shown that after 24 h of treatment with 10 nM panobinostat the SLC5A5C (NIS) protein concentration was increased leading to up-regulation of radioiodine retention (57). The NIS concentration after 24 h of treatment with 10 nM panobinostat increased but decreased if the treatment persisted for 48 h even though the RAI-U remained high (57).
Slightly different results were produced when panobinostat was used to treat ATC cells of a 42-year-old with a BRAFV600-tumor at IVB stage and pT4bN0M0 who survived 282 weeks after diagnosis. Sorafenib and panobinostat acted synergistically to increase NIS protein levels and RAI-U, and this increase was inversely correlated with the levels of hsa-let7g-5p mRNA, a tumor suppressor miRNA in thyroid cancer (57).
Belinostat
Belinostat (PXD101) acetylates H3 and induces the expression of p21 in both SW1736 and Cal62 anaplastic thyroid cancer cells. Furthermore, it decreased RAS-RAF, PI3K-mTOR, pERK, and Akt pathways. Treatment with 100 nM PXD101 resulted in 11-37% apoptosis compared to the treatment with 50nM panobinostat that led to 43-68% apoptosis (58).
Furthermore, in combination with SNX5422, it led to higher levels of cytotoxicity, further suppression of the PI3K/Akt/mTOR pathway, inactivation of survivin and activation of DNA damage-related proteins by reducing the acetylation of H3 and H4 (36). When combined with NVP-AUY922, it reduced the levels of phospho-ERK1, 2 and total and phosphor-Akt without affecting the hsp90 or hsp70 protein levels. However, this combination did not lead to the reduction of H3, H4 acetylation as seen with SNX5422, but to decreased levels of ATM and ATR proteins (59).
Belinostat showed promising results when used along with gemigliptin, a dipeptidyl peptidase-IV inhibitor in SW1736, 8505C anaplastic thyroid cancer cells. The combination synergistically augmented cell death and increased the number of apoptotic cells, the levels of Bcl2 protein, and cleaved poly ADP-ribose polymerase while it reduced the levels of xIAP and survivin proteins. Moreover, the levels of phospho-Akt and phospho-AMPK were elevated and cell migration was reduced (60).
Triptolide
Triptolide, a diterpenoid epoxide produced by the Tripterygium wilfordii, has shown antitumor activities in a broad range of solid tumors like pancreatic, breast and thyroid cancers. Treatment of anaplastic thyroid cancer with the combination of triptolide and BIIB021, a heat shock protein inhibitor, resulted in increased cytotoxicity, decreased cell viability, decreased levels of survivin protein and activation of DNA damage response. Moreover, they increased the levels of cleaved caspase 3, xIAP, cIAP and acetyl H4 (61).
Conclusion
One of the main reasons for the chemotherapy resistance observed in many cancers, including ATC, is the epigenetic modification of DNA leading to changes in gene expression. Histone acetylation/deacetylation is a significant mechanism that induces DNA methylation and chromatin remodeling, modifying gene transcription. The delicate balance between histone transacetylases and deacetylases is often disturbed in cancer, altering expression of tumor protooncogenes and tumor suppressor genes (62). The acetylation of histones is maintained by histone acetyltransferase and HDAC enzymes. However, HDAC enzymes are often over-functioning in cancer cells, indirectly leading to the acetylation of tumor protooncogenes and increased proliferation of cells (63-66). Therefore, HDACIs are drugs that specifically target cancer cells that have higher levels of acetylation, avoiding the complications of other not so targeted therapies like chemotherapy (67). Moreover, HDACIs have many more functions including invigoration of DNA repair mechanisms, induction of apoptosis and cell cycle arrest (68).
Even though there are many HDACIs under clinical trials, none is available for the treatment of ATC. The results of many in vitro experiments in cell lines are promising, with many HDACIs showing signs of cancer regression (69-71). These medications exert better results when used along with other anticancer drugs, but in some specific cases, monotherapy could also be considered.
Moreover, HDACIs may be even more valuable in ATC treatment, due to the failure of standard treatment used so far and its disappointing survival rate. The studies presented in this review justify further research in the field of HDACIs, in order to learn more about their functions and how they interact with other medications in order to be better utilized for the treatment of many cancers including ATC.
Footnotes
↵* These Authors contributed equally to this study.
Authors' Contributions
Drafting of the manuscript: Spartalis E. and Athanasiadis D.I. Literature search and analysis: Athanasiadis D.I., Chrysikos D. Data extraction: Spartalis M. and Schizas D. Boutzios G performed quality assessment. Table drafting: Garmpis N. and Damaskos C. Manuscript editing: Paschou S.A. and Ioannidis A. Critical revision of the manuscript for important intellectual content: Tsourouflis G. and Dimitroulis D. Supervising professor: Nikiteas N.I.
This article is freely accessible online.
Conflicts of Interest
None of the Authors have any conflicts of interest to disclose regarding this study.
- Received January 25, 2019.
- Revision received February 15, 2019.
- Accepted February 18, 2019.
- Copyright© 2019, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved