Problems, Mechanisms and Potential Solutions
Epigenetic therapy presents several clinical and translational challenges. In particular, while DNA methylation inhibitors and HDIs are considered targeted agents, this approach is generally nonspecific because of the numerous downstream effects on gene expression. Also, despite documented epigenetic effects, there is no consensus as to whether or not the responses are mediated strictly via epigenetic mechanisms. Thus, translational research will be critical to clarify the precise mechanisms of the in vivo action of the drugs—and mechanisms of resistance. Furthermore, as new data emerge regarding the response markers and mechanisms of action of these drugs, combination therapy involving epigenetic agents together with conventional or targeted agents is increasingly seen as an attractive opportunity. Therefore, preclinical combinatorial resistance or synthetic lethality screens are likely to be instrumental in selecting the most potent drugs and combinations.
While several studies have documented the epigenetic effects of decitabine and azacitidine in vitro and in vivo,[31-33] other studies have elucidated other mechanisms. As an example, it was recently shown that DAC upregulates p21 and induces cell cycle arrest in a DNA methyltransferase (DNMT)-independent fashion through DNA damage ATM/P53 axis, implicating the role of these pathways in responses. However, the same group has demonstrated that neither DNA damage nor methylation changes were predictive of response to demethylating therapy. Further, decitabine effects were studied by whole-genome expression assays in melanoma cells; p21 induction was observed and WNT pathway activation was found to be contributing to resistance. Interestingly, inhibition of the proteasome increased the growth arrest of DAC-treated cells, implying that proteasome inhibitors might act in synergy with epigenetic modifiers. An interesting study looking at cytotoxic CD8+ T-cell responses to MAGE antigens has demonstrated that epigenetic therapy with a DNMT inhibitor induced such responses in vivo and was correlating with clinical responses in 8 out of 11 patients; thus, it is possible that unmasking MAGE antigens with epigenetic drugs caused immune clearance of leukemic cells. Epigenetic drugs are known to induce differentiation, and a study recently described how both cytarabine(Drug information on cytarabine) and DAC were able to induce differentiation of embryonic cancer stem cells via caspase-mediated degradation of EZH2 as well as NANOG and OCT-4 stem cell factors, suggesting yet another mechanism of action of nucleoside analogues and epigenetic drugs operating at the stem cell level.
To resolve these issues regarding the mechanism(s) of action of epigenetic drugs, the identification of predictors or markers of clinical responses is of paramount importance. However, we are at an early stage in understanding the process of identifying predictors/response markers. A recent study of elderly patients with AML who were treated with decitabine showed that higher pretreatment levels of microRNA-29b (miR-29b), previously shown to target DNMTs, correlated with clinical responses; thus, if these results are validated, miR-29b levels could be used for stratification of patients. Several HDI studies were recently published that proposed explanations of responses seen in CTCLs. Using a genome-wide loss of function screen with a short hairpin RNA (shRNA) library, HR23B, a component of the proteasome, was identified as a determinant of HDI sensitivity. HR23B is expressed at high levels in CTCL, and there was a correlation between HR23B expression and clinical response to HDIs. Further, a mechanistic study of panobinostat showed that this agent induces CXCR4 degradation via the proteasomal pathway and has synergy with CXCR4 inhibition, supporting the rationale for testing such a combination in vivo. The same research group reported another combination study of the EZH2 inhibitor 3-deazaneplanocin A (DZNep) with panobinostat. This study demonstrated induction of more apoptosis in AML cells with the combination therapy than with either agent alone; the combination also prolonged survival in a mouse model. Another study of neuroblastoma cells exhibiting silencing of caspase-8 demonstrated that the combination of various HDIs with interferon-gamma was effective in overcoming the resistance of such cells to tumor necrosis factor–related apoptosis-inducing ligand (TRAIL), thus prompting further investigations. Finally, a study addressing fundamental mechanisms of drug resistance identified a subpopulation of cancer cells that upregulated insulin-like growth factor 1(IGF-1) signaling and acquired an altered chromatin state via RBP2/KDM5A/Jarid1A demethylase upregulation and concomitant decreases in H3K4 methylation and H3K14 acetylation. Interestingly, drug-tolerant cells could be ablated by IGF-1 or HDIs. The role of RBP2 in mediating clinical sensitivity to epigenetic drugs needs testing; this approach may yield a novel therapeutic opportunity to use epigenetic therapy to reverse drug resistance in patients in whom resistance to chemotherapy is developing.
Novel Epigenetic Drugs
There is great enthusiasm in both academia and industry for developing new epigenetic compounds. Some of the recent preclinical data regarding several of these potential drugs are summarized below.
S110 is a 5-aza-2'-deoxycytidine–containing dinucleotide that releases decitabine intracellularly and has shown improved stability in vitro, as well as demonstrating hypomethylating and anti-tumor effects in a cellular model. Clinical trials with S110 are planned for the near future. CP-4200, an elaidic acid derivative of azacitidine was recently developed. This compound has demonstrated strong epigenetic effects in cell lines; it causes depletion of DNA methyltransferase, genome-wide DNA demethylation, and robust reactivation of epigenetically silenced tumor suppressor genes. Importantly, the cellular uptake of CP-4200 is substantially less dependent on the nucleoside transporters, and the compound showed significantly higher antitumoral activity than azacitidine in an orthotopic mouse tumor model of acute lymphocytic leukemia. Thus, these early data suggest that elaidic acid modification may improve the therapeutic efficacy of azacitidine. The first DNMT3B-selective inhibitor, nanaomycin A, has been described; this compound exhibited genomic demethylation and antiproliferative effects in different cell lines.
Several new HDIs have recently shown promise in preclinical testing. Resminostat is a potent inhibitor of HDACs 1, 3 and 6, and in low micromolar concentrations it abrogated cell growth and strongly induced apoptosis in multiple myeloma cell lines and primary cells. It has shown synergistic effects in vitro in combinations with melphalan(Drug information on melphalan) and proteasome inhibitors. AR-42 is a novel HDI that recently showed efficacy against B cells in vivo and in vitro. It demonstrated inhibition of primary leukemia and lymphoma cell lines, and it significantly reduced leukocyte counts and/or prolonged survival in three separate mouse models of B-cell malignancy, without evidence of toxicity. Novel HDIs have also been shown to be active against solid tumors, in particular lung cancer. For instance, the novel pan-HDAC inhibitor, OSU-HDAC-44, exhibits three to four times more effectiveness than vorinostat in suppressing cell viability in various NSCLC cell lines. It inhibited cytokines, induced mitochondria-mediated apoptosis, and inhibited xenograft tumor growth; based on these data, it is a promising chemotherapeutic drug for NSCLC.
Histone methyltransferase inhibitors
EZH2 is a histone methyltransferase that is overexpressed in multiple cancer types and that has an activating mutation in lymphomas that makes it an attractive therapeutic target.[51,52] DZNep has been shown to deplete components of the PRC2 complex, including EZH2, and to induce apoptosis in cancer cells. Inhibition of EZH2 in glioblastoma stem cells by DZNep resulted in impairment of cell renewal and of the tumor-initiating capacity of these cells, in part via c-myc downregulation. DZNep has also been shown to be especially effective in killing BRCA-deficient breast cancer cells. Another histone methyltransferase, G9a, which targets a different residue, has previously been shown to be important for cancer cell survival. Two inhibitors of G9a, the diazepin-quinazolin-amine derivative BIX-01294 and the 2,4-diamino-6,7-dimethoxyquinazoline derivative named UNC0224, were recently described but have not yet been tested in vivo.[57-59] Coactivator-associated arginine methyltransferase 1 (CARM1) is an arginine methyltransferase that has been implicated in prostate cancer development. Ellagic acid has recently been shown to be a specific CARM1 inhibitor that reduced H3R17 levels. It is expected that several histone methyltransferase inhibitors will enter clinical trials in the next few years.
Histone demethylase inhibitors
To date, two classes of histone lysine(Drug information on lysine) demethylases have been identified. One class includes the lysine-specific demethylases LSD1 and LSD2, and a second comprises the recently discovered Jumonji domain–containing protein (JMJD) histone demethylases. LSD1 has been shown to be overexpressed and/or critical for cell proliferation in prostate cancer,[62-64] breast cancer, and neuroblastoma. LSD1 inhibitors were recently developed and have demonstrated reactivation of silenced genes as well as inhibition of colon cancer growth.[67-69] Similarly, JMJD demethylases have been implicated in cell growth in prostate, breast, and esophageal cancers. JMJD inhibitors have also recently been described; these have demonstrated inhibition of prostate and colon cancer cell growth. Further characterization of these compounds is needed in anticipation of potential clinical trials.
Histone acetyltransferase (HAT) inhibitors
HATs are involved in many critical cellular proliferation pathways, and several chromosomal translocations observed in leukemias directly disrupt these enzymes. Translocation t(8;16)(p11;p13), a cytogenetic hallmark for the M4/M5 subtype of AML, fuses the CREB-binding protein (CBP) on chromosome 16 with a chromosome 8 HAT, MOZ. MOZ-CBP might induce leukemia by antagonizing the function of the AML1 complex, since MOZ-driven acetylation plays a role in controlling a desirable balance between proliferation and differentiation during hematopoiesis. Further, t(11;16)(q23;p13.3), a common translocation in acute leukemia and secondary myelodysplasia, results in a fusion of the MLL gene with the CBP gene. The chimeric fusion product retains the HAT domain of CBP and may lead to leukemia by promoting histone acetylation of genomic regions targeted by MLL, which in turn leads to transcriptional deregulation via aberrant chromatin organization. These findings make HATs an attractive target for cancer therapy, and various inhibitors have been described, although few have been tested against cancer cells.[76,77] A recent study described a novel HAT inhibitor, a derivative of spermidinyl–coenzyme A, which was shown to be effective as a cancer-specific chemo- and radiosensitizer. Another study described the successful blocking of hyperacetylation of histone H3 in oral squamous cell cancers by a curcumin derivative, resulting in inhibited cancer cell growth. Another HAT inhibitor, C646, was shown to inhibit lung and melanoma cancer cells without affecting a normal fibroblast cell line.
The use of epigenetic drugs allows restoration of the expression of genes that are deregulated in cancer. These agents help patients with MDS live longer with fewer side effects than is possible with standard chemotherapy. Several early trials have shown promising results in patients with solid tumors, especially NSCLC, where vorinostat appears to enhance the efficacy of chemotherapy. HDIs are also now being tested as radiosensitizing agents. Components of the proteasome have been identified as likely sensitivity determinants for HDIs, whereas immune responses and degradation of stem cell factors have been reported to be likely sensitivity determinants for DNA methylation inhibitors. Finally, several novel epigenetic drugs targeting histone-modifying enzymes are emerging and have entered preclinical testing; hopes are high that these agents will be seen in clinical trials in the near future.
Financial Disclosure: Dr. Issa receives research support from Eisai, Merck, and Celgene; and honoraria from Celgene, Novartis, and Johnson & Johnson. He serves as a consultant for GlaxoSmithKline and Syndax. Dr. Boumber has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.