Postdoctoral Fellow, Division of Hematology/Oncology, Northwestern University Feinberg School of Medicine
Johanna Dobe Professor and Chief, Division of Hematology/Oncology, Robert H. Lurie Comprehensive Cancer Center, Northwestern University Feinberg School of Medicine
Drs. Shah and Licht indicated no relevant conflicts of interest.
Epigenetics is defined as the heritable alteration of gene expression without an accompanying change in DNA sequence. Epigenetic changes are primarily acquired through DNA methylation, which occurs at the cytosine located in a CpG dinucleotide, and post-translational histone modifications. These chromatin modifications are usually tightly regulated in development and differentiation. Generally, cancer cells exhibit global hypomethylation, which contributes to chromosomal instability, reactivation of transposable elements, and loss of imprinting. However, malignancies often demonstrate hypermethylation of many promoter-associated CpG islands and other CpG-rich regions. Hypermethylation of tumor suppressor genes (TSGs) was proposed to lead to increased cell proliferation, providing a selective advantage for cells with methylated gene promoters.
The four core histones, H2A, H2B, H3, and H4, make up the nucleosome, the main structural unit of chromatin. In recent years, numerous post-translational modifications, including methylation, acetylation, phosphorylation, and sumoylation, were identified on the “tails” of histones, a stretch of about 40 amino acids that does not directly bind to the DNA of the nucleosome. Some specific histone tail modifications, such as methylation of histone 3 lysine tail residue 4 (H3K4), are associated with activation of gene expression, while others, such as methylation of histone 3 lysine 27 (H3K27), are associated with gene repression. These marks are normally carefully controlled by the interplay of sequence-specific DNA binding transcription factors and transcriptional co-factors, many of which are histone-modifying enzymes. In this review, we will discuss recent developments in the field, and the current and future state of epigenetic therapies.
Altered Function of DNA Methyltransferases
The DNA methyltransferases (DNMTs) catalyze the conversion of cytosine to 5-methylcytosine. DNMT1 maintains DNA methylation once the cell divides, while DNMT3A and DNMT3B are de novo methyltransferases, which can add methyl groups to unmodified DNA. Mutation of the DNMT3A gene was observed in 20 percent of acute myeloid leukemia (AML) cases and correlated with a poor clinical outcome. Surprisingly, however, DNMT3A mutations do not dramatically alter 5-methylcytosine or global DNA methylation levels in AML genomes but still may have indirect effects on gene expression. Additionally, alternative, aberrantly spliced transcripts of DNMT3B were observed in primary leukemia samples and cancer cell lines, and appear to lead to the deregulation of normal methylation patterns.
TET2 Mutations in Myeloid Malignancy
In 2009, inactivating mutations of the Ten-Eleven-Translocation oncogene family member 2 (TET2) gene were identified in myelodysplastic syndrome (MDS), myeloproliferative neoplasm (MPNs), AML, and chronicmyelomonocytic leukemia (CMML). Some studies suggested that TET2 mutations in myeloid malignancy conferred a better prognosis. TET2 can convert 5-methylcytosine (5-mC) to 5-hydroxymethylcytosine (5-hmC),1 which was hypothesized to be an intermediate inthe demethylation of DNA. TET2 mutations can be found in combination with mutations of JAK2 in MPN, FLT3 in AML,and RAS in MDS, suggesting that they play a contributory role to the development of myeloid malignancy. TET2 levels increase during differentiation of myeloid cell lines,and depletion of TET2 in hematopoietic precursors leads to monocyte/macrophage differentiation, indicating that TET2is important in normal myelopoiesis.1
Epigenetic Signatures in AML
The methylation state of CpG-rich sequences within the promoters of genes can be profiled en mass through the use of microarray. This type of analysis of large cohorts of AML patients characterized 16 distinct patterns of gene methylation, some of which corresponded to specific AML subtypes, such as acute promyelocytic leukemia. However, patterns of gene methylation were found that were not associated with any known gene mutations or chromosomal translocation, yet were associated with a poorer prognosis. More recently, AML harboring isocitratedehydrogenase 1 and 2 (IDH1/2) mutations were found to display aberrant hypermethylation.2 These mutations led to the production of an abnormal metabolite in the cell, 2-hydroxyglutarate (2HG), which can inhibit the hydroxylation of 5-mC by TET2.2 It is likely, however, that IDH1/2 mutants have even more widespread epigenetic effects, since the Jumonji C class of histone is likely inhibited by 2HG.
Aberrant Histone Methylation in Hematologic Malignancy
Histone modifications, especially histone methylation, play an important role in altering gene expression in hematologic malignancies. Histone lysine methylation is generally performed by histone methyltransferases (HMT) containing the SET domain, while demethylation is carried out by LSD1 and LSD2 or demethylases containing the Jumonji C domain. MLL, the most commonly rearranged gene in AML, encodes an HMT that methylates H3K4, a mark associated with increased gene expression. MLL fusion proteins do not contain the enzymatic activity of MLL, but retain the ability to bind to MLL target genes, such as the HOX genes, which are critical for stem cell self-renewal. MLL fusion proteins work in part by recruiting transcriptional elongation complexes (which include the atypical HMT DOT1L) to HOX gene promoters, resulting in aberrant overexpression. The HMT MMSET is dramatically overexpressed in multiple myeloma (MM) patients with the t(4;14) translocation and mediates a global increase in histone 3 lysine 36 (H3K36) di-methylation, as well as a decrease of H3K27 methylation.3 This is associated with shifts in gene expression that affect cell cycle, cell death, and cell adhesion pathways.
Aberrant methylation of H3K27 is a recurrent theme in blood malignancy. Inactivating mutations in the H3K27 histone methyltransferase EZH2 were identified in patients with MDS/MPN and myelofibrosis. Gain-of-function mutations of EZH2 that increase H3K27 methylation are seen frequently in large B-cell lymphoma, suggesting that EZH2 can act as an oncogene in this context.4 Furthermore, inactivating mutations of the H3K27 demethylase UTX, which would lead to increased H3K27 methylation, are found in 10 percent of cases of multiple myeloma.
Recent data indicate a link between aberrant signaling and chromatin modification in hematologic malignancy. The JAK2 kinase, which is mutated in MPN and overexpressed in Hodgkin disease and mediastinal lymphoma, is localized to the nucleus where it directly phosphorylates histone 3 tyrosine 41 (H3Y41), a modification that prevents the methylation of histone 3 lysine 9 (H3K9), which is a mark of inactive chromatin.5 In this way, JAK2 can bind and activate specific sets of genes that stimulate malignant growth.
Epigenetic Therapy in Hematologic Malignancies
"Epigenetic therapy" refers to the use of agents intended to target chromatin processes, but whether these agents work as they are designed is not clear. The DNA hypomethylating agents 5-azacytidine and decitabine can induce significant hematologic improvement in MDS. MDS is associated with aberrant gene hypermethylation, which can be reversed by 5-azacytidine treatment, but this has not been consistently correlated with re-expression of methylated, silenced tumor suppressor genes. 5-azacytidine treatment is associated with DNA damage and may work as a low-level cytotoxic agent. Vorinostat, a histone deacetylase inhibitor (HDACi), leads to global increases in histone acetylation in many cell types but has proven to be most efficacious in cutaneous T-cell lymphoma (CTCL) for unknown reasons. The therapeutic effects of this HDACi may be related to alteration in chromatin associated with shifts in gene expression, changes in chromatin stability and DNA damage, or even non-chromatin effects, such as alterations in HSP90 protein chaperone function.
Epigenetic therapy is becoming more sophisticated and in coming years will likely work through specific, on-target actions. For example, a highly specific inhibitor of DOT1L, the enzyme that methylates H3K79, specifically killed leukemia cells harboring MLL fusion proteins as intended.6 JAK2 inhibitors being developed for the treatment of myelofibrosis may work in part through alteration of epigenetic histone modifications, reversing H3Y41 phosphorylation and increasing H3K9 methylation of JAK2-bound genes. There is much interest in the development of EZH2 inhibitors that may be useful for lymphoma and MMSET inhibitors for myeloma. IDH1/2 inhibitors that stop the production of 2HG are likely to affect the function of the TET enzymes and demethylases. As cancer genomes become sequenced and recurrent mutations characterized, it is clear that more epigenetic targets will emerge, many of which may be amenable to specific therapies. Hence, in the coming decade, epigenetic concepts are likely to change the clinical practice of hematology.
