Epigenetic defects in cancer


Mammalian Developmental Epigenetics Group, Genetics and Developmental Biology Unit
Institut Curie, INSERM

Epigenetic defects in cancer

by R. Chaligné and E. Heard, France

Epigenetics concerns heritable changes in gene expression that are not linked to changes in the DNA sequence. Potential epigenetic regulators range from chromatin-associated proteins to DNA methylation and noncoding RNAs. Groundbreaking research over the last half century has revealed the importance of epigenetic mechanisms in development and disease. The investigation of normal processes such as X-chromosome inactivation and genomic imprinting has enabled a deeper understanding of the molecular basis of epigenetic mechanisms, and this knowledge can now be used to explore disease. Indeed, disruption of epigenetic control is frequent in cancer and the potential for reversal of epigenetic changes, contrary to genetic changes, means that epigenetic-based therapies are increasingly being considered in the treatment of cancer. Here, we provide an overview of some of the links between epigenetics and cancer.

Medicographia. 2014;36:293-299 (see French abstract on page 299)

Epigenetics was first coined by Conrad Waddington and, in one of its more recent definitions, concerns heritable changes in gene expression or gene function that are not due to changes in DNA sequence. Our understanding of epigenetic mechanisms included in this definition, has radically increased in the last few years with the arrival of breakthrough technologies. Although all cells within an organism contain the same genomic DNA, numerous epigenetic regulators and transcription factors organize the genome into accessible and closed chromatin, to ensure a correct and specific transcriptional program in a given cell type. Chromatin is a macromolecular complex of DNA and proteins, which not only provides a scaffold for the packaging of our genome into the cell nucleus, but also influences gene expression and genome functions, such as DNA replication and DNA repair. It can also act as a scaffold for epigenetic information that is heritable through cell divisions, both at the level of the DNA (eg, DNA methylation), or at the level of histones and other proteins associated with them. The nucleosome is defined as the basic functional unit of this macromolecular complex. Basically, 147 base pairs of DNA are wrapped around 8 histone proteins, composed of two of histone H2A, H2B, H3, and H4. We generally divide chromatin into two different states, heterochromatin that is highly condensed, late replicating, and principally enriched in repeats and/or inactive genes, and euchromatin that is more open and enriched in active genes (Figure 1). Heterochromatin can itself be divided into constitutive and facultative forms, the former being present in all cells types, for example in centromeric and repetitive chromosomal regions; while the latter is best exemplified by the inactive X chromosome (Barr body) that is established during female development and ensures the dosage compensation for Xlinked genes.

Figure 1
Figure 1. Chromatin organization in normal and cancer cells.

(A) The two main types of chromatin present in normal cells: euchromatin, enriched for “active” chromatin
marks, and heterochromatin enriched for “repressive” chromatin modifications. In normal cells,
constitutive heterochromatin lies mainly at the nuclear periphery and around the nucleolus; facultative
heterochromatin (Barr body) is at the nuclear or nucleolar periphery. (B) Perturbed chromatin states
frequently observed in cancer cells. For example, the Barr body is disrupted or lost, and new, aberrant
heterochromatic compartments are sometimes observed. Chromatin modifiers are also frequently disrupted
in cancer, either through aberrant expression or mutations, and this can impact heterochromatin
maintenance. The reorganization of heterochromatin and broad epigenetic instability has the potential
to impact gene expression profiles and increase the risk of cancerogenesis. Heterochromatic regions
of the nucleus are shown in dark brown; euchromatin is shown in light brown.
Abbreviations: LAD, lamin-associated domain; LOCK, large organized chromatin K-modification; RNAP,
RNA polymerase.

The components of the different types of heterochromatin and euchromatin are subject to covalent modifications that are thought to contribute to their specificities. Today, at least four different DNA modifications, numerous histone variants, and several tens of different classes of histone modifications have been identified (Figure 2).1-4 Such variants and modifications can influence chromatin organization and DNA accessibility to transcription factors, by changing noncovalent interactions within and between nucleosomes. They can also act as docking sites for “reader” proteins that bind to them, either alone or in combination, and such “reader” proteins can recruit further chromatin modifiers and remodeling chromatin complexes. The ensemble of DNA-based processes, including transcription, DNA replication, and DNA repair, can be influenced by such chromatin states. Mutations or abnormal expression level of chromatin regulators can have a major impact on the regulation of epigenetic mechanisms, leading in some case to abnormal development or to disease. Indeed, there is an increasing realization that epigenetic changes may be tightly linked to cancer. On the one hand, epigenetic modifications such as global changes in DNA methylation and chromatin structure clearly accompany tumorigenesis. This correlates with aberrant gene expression as well as genetic instability, although whether such epigenetic changes are cause or consequence is not always clear. On the other hand, epigenetic changes may underlie tumor initiation, through the appearance of epimutations that, for example, result in aberrant repression of tumor suppressor genes or expression of oncogenes. The degree to which epigenetic changes are involved in cancer and can be used to assess the tumor state are active areas of research.

Figure 2
Figure 2 (opposite page).
Epigenetic regulators affecting chromatin.

The tables on the right represent examples of mutated
genes found in a cancer context. (A) The 5-carbon of
of cytosine nucleotides are methylated (5mC) by a family
of DNA methyl transferases (DNMTs). The 5-methylcytosine
hydroxylase (TET family) of DNA hydroxylases
metabolizes 5mC into several oxidative intermediates,
including 5-hydroxymethylcytosine (5hmC), 5-formylcytosine
(5fC), and 5-carboxylcytosine (5caC). These
intermediates may participate in the process of active
DNA demethylation. (B) Schematic view of action of the
histone acetyltransferases (HATs), deacetylases (HDACs),
and “reader” proteins that bind histone-acetyl group.
(C) Schematic view of action of histone methyltransferases
(HMTs), demethylases (HDMs), and “reader”
proteins that bind histone-methyl group. (D) Schematic
view of the action of the kinases, phosphatases, and
reader proteins that bind the histone-phosphate group.
(E) Multisubunit chromatin remodeling complexes, such
as SWI-SNT, bind chromatin and perturb histone-DNA
contacts. This complex modifies nucleosome positioning
and structure by sliding and evicting nucleosomes to
make the DNA more accessible to transcription factors
and other chromatin regulators.
Abbreviations: AML, acute myeloid leukemia; ALL,
acute lymphoid leukemia; B-NHL, B-cell non–Hodgkin’s
lymphoma; CML, chronic myeloid leukemia;
DLBCL, diffuse large B-cell lymphoma; FL, follicular
lymphoma; HNSCC, head and neck squamous cell
carcinoma; MDS, myelodysplastic syndromes;
MPD, myeloproliferative diseases; NHL, non–Hodgkin’s
lymphoma; OCC, ovarian clear cell carcinoma; T-PLL,
T cell prolymphocytic leukemia; TCC, transitional cell
carcinoma of the urinary bladder.

Epigenetics and epigenomics of cancer

The recent decrease in next generation DNA sequencing costs, coupled with the improvement in techniques, such as chromatin immunoprecipitation, have provided an unprecedented view of the genomes and epigenomes of normal and cancer cells.5 We can now readily analyze DNA modifications,6 histone variants and posttranslational modifications,7 transcription factor binding sites,8 and chromosome conformation,9 and have a more comprehensive view of nucleosome positioning.10 Transcriptome analyses, using RNAseq, have revealed that a large portion of our genome is transcribed, and several non–protein-coding RNAs have been identified. Given work on processes such as X inactivation and imprinting, where long noncoding RNA (lncRNA) play key roles,11 the discovery of numerous other lncRNAs suggests that some of them may have an important function in epigenetic regulation.12 The first suggestion that epigenetic instability might be involved in cancer dates from the 1970s and 1980s.13 Subsequent analyses of gene expression and DNA methylation patterns revived interest in epigenetic misregulation in cancers, as they revealed both broad and more punctual alterations in epigenomic landscapes (see reference 14 for review). Nevertheless, these studies were mainly correlative up until recently. Insights into a direct role of epigenetic processes in cancer came from recent whole genome sequencing studies of numerous cancer types, notably through The Cancer Genome Atlas (TCGA) consortium, where sets of somatic mutations in epigenetic regulators were identified. A high frequency of somatic mutations in genes coding for chromatin-associated proteins that are known to regulate DNA methylation patterns, histone posttranslational modifications, and chromatin remodeling have been found. Such changes might be “driver” mutations in the process of carcinogenesis. For example, KDM6A, an X-linked histone demethylase, is mutated in up to 12 distinct cancers (Figure 2).17 Another example is the histone methyltransferase, KMT2B which has been found to be mutated in almost 90% of non– Hodgkin lymphoma patients.18 In glioma, the histone variant H3.3 has been found mutated with single codon changes within its N-terminal tail, which is the target for posttranslational modifications.19 Another important recent discovery is that patients with leukemia often have mutations in genes such as TET2, IDH1, IDH2, and DNMT3A, which are all involved in regulating DNA methylation patterns. This has provided insight into why patients with leukemia show a significant response to DNA methylation inhibitors (see below), and represents a promising avenue for future patient stratification strategies. Polycomb group proteins, such as the EZH2 histone methyltransferase, together with the histone H3K27 methylation modification it lays down, have been frequently correlated with cancer, although their exact roles are still not clear. Indeed, EZH2 is thought to promote or inhibit tumorigenesis in a context-dependent manner. This may also be the case for several epigenetic modifiers that may have rather different influences on gene expression and cell proliferation depending on their exact partners, protein complex stoichiometry, as well as their target chromatin landscape.

In addition to the identification of epigenetic modifiers as potential drivers in cancer, the genome-wide mapping of chromatin modifications, thanks to highly specific antibodies, has provided further insights into the nature and extent of epigenetic abnormalities in cancer. For example, comparisons of DNA methylation profiles, and the binding regions of chromatin regulators and histone modifications in human cancers has revealed a link between hypermethylated gene promoters and genes with a particular “bivalent” status for histone modifications in cancer cells.20,21 Indeed, bivalent genes show an enrichment of both the H3K4me3 mark (usually associated with euchromatin) and the H3K27me3 mark (associated with facultative heterochromatin; Figure 1).22 This bivalent chromatin state had been previously associated with genes in embryonic stem cells that are poised to show lineage specific expression patterns during differentiation and opened up the possibility that in cancer, the bivalent state represents a similar poised and possibly dedifferentiated state. Furthermore, by comparing normal and tumor tissue from the same individual, epigenomic studies have discovered intriguing profiles showing altered DNA methylation and H3K9 methylation profiles, in large regions spanning several hundreds of kilobases, termed large organized chromatin K-modifications (LOCKs), often corresponding to lamin-associated domains (LADs).23-25 During normal development or cellular stress processes, such as following injury, changes in LOCKs/LADs are thought to reflect or even contribute to cellular plasticity. It has been proposed that such changes in cancer cells may reflect or contribute to their increased plasticity. Furthermore, large domains of H3K27me3, H3K9me2/3, and DNA methylation, identified in bladder and other cancers,23,26,27 have been associated with coordinated repression of the genes within them, and this has been correlated with particular types of tumor prognosis in some cases.28,29 Why and how these particular genomic regions are more vulnerable to epigenomic perturbations is still a mystery, but could be due to aberrant targeting of epigenetic complexes to regions that are clustered in the nucleus, such as topologically-associated domains, or to perturbation of long-range regulatory sequences (see reference 30 for review).

Epigenetics and nuclear disorganization in cancer

Global perturbations in heterochromatin and unusual nuclear architecture are hallmarks of cancer that have been tradition ally used by pathologists in tumor classification (Figure 1). An example of this concerns the disappearance or disruption of the heterochromatic Barr body, or inactive X chromosome, that has long been associated with the most aggressive breast tumors by clinicians.31,32 The extent to which the disappearance of the Barr body is linked to more general nuclear disorganization, chromatin disruption as opposed to physical loss of the inactive X chromosome, remains to be found. Although some studies reported that in certain types of tumor (eg, basal-like molecular subtype breast tumors), the Barr body loses association with the non–coding X-inactive specific transcript (XIST; responsible for triggering its inactivation) and becomes euchromatic,33 others have reported that the Barr body remains inactive and XIST RNA coating can still be found, even in basal-like tumors.34-36 Nevertheless, the epigenetic instability and degree of gene reactivation from the inactive X chromosome and other types of facultative heterochromatin in cancer have not been systematically investigated to date.

Epigenetic therapy in cancer

Given the evidence that epigenetic changes are a frequent feature in cancer, and that in some cases the misregulation of cancer-related genes may be at the epigenetic rather than at the genetic level (epimutation vs DNA sequence mutation), this provides great promise for cancer treatment. Indeed, epimutations have the potential to be reversed via chemical agents, known as “epidrugs”, unlike genetic mutations, which are essentially irreversible. Such epidrugs are at various stages of development and some are already being used as therapeutic agents in cancer treatment. The first epidrugs that were approved by the US food and Drug Administration (FDA) for cancer therapy target DNA methylation: azacytidine (5-azactydine) and decitabine (5-aza-2’-deoxycytidine), both being nucleoside analogues and irreversible inhibitors of the DNA methyltransferase enzymes DNMT1 and DNMT3. These drugs are being used as first-line treatments for patients with myelodysplastic syndrome (see reference 37 for review). Subsequently, chemical inhibitors targeting histone deacetylases, HDACs (eg, SAHA and romidepsin or FK228), have been FDA approved for treatment of refractory cutaneous T-cell lymphomas (see reference 38 for review). Such drugs are clearly successful in the clinic, although it is still not entirely clear what their relevant target genes are. For example, HDACs show poor enzyme specificity, their mechanism of action is still not fully understood,39 and so far there is no striking gene expression signature or profile that can predict whether a patient will benefit from the use of HDAC inhibitors. The situation is very similar for DNMT inhibitors, which have been shown to lead to global hypomethylation, although we still do not know their precise mechanism of action in a clinical context. For both types of epidrug, the lack of reliable molecular biomarkers for predicting either clinical activity or resistance is a serious drawback, limiting clinicians’ ability to achieve the vision of “personalized medicine.” Furthermore, so far these drugs have only been used with success in specific hematological cancers and their use for treatment of solid tumors has had limited success. Several pharmaceutical and biotech company research groups have also developed highly potent, selective, small molecule inhibitors of the H3K27me3 histone methyltransferase, EZH2, although there are potential drawbacks in inhibiting this enzyme given its context-specific action (see above). Thus, a major challenge for the use of such inhibitors in cancer will be to gain a better understanding of their mechanisms of action, and the biology of their target proteins.

Nevertheless, the investment so far in developing epidrugs has clearly paid off, as exemplified by the recent preclinical success of small molecule inhibitors of bromodomain-containing protein 4 (BRD4), an acetyl-lysine chromatin-binding protein.40-42 Recent studies revealed that several such bromodomain and extraterminal (BET) proteins are involved in cancer, and that they can be targeted by small molecule antagonists that directly bind to them and prevent the interaction of the bromodomain “reader” module to acetylated histones, thus preventing assembly of active gene transcriptional complexes at genes. Insights into the mechanism underlying the efficiency of inhibiting the BRD4 protein came from studies on the fusion of the bromodomain protein BRD4 with a nuclear protein in testis (NUT). This fusion-protein leads to the development of aggressive NUT midline carcinoma.43 Aberrant regulation of BRD4 has also been reported in other cancers, such as colon and breast tumors. The use of cell-based, high throughput screening of chemical libraries has led to the development of compounds that can selectively inhibit each of the four known BET proteins. The development of selective inhibitors that target such epigenetic “reader” proteins thus represents a very promising horizon for cancer therapy.

Concluding remarks

Epigenetic mechanisms are essential for normal development and are often perturbed in the context of cancer. The advent of new technologies has enabled hypotheses concerning the molecular origins of cancer to be confirmed, but has also opened up a new era of research. The belief that cancer is driven by genetic abnormalities remains true, however it is clear now that epigenetic pathways play an important function in cancer development. The hallmarks of cancer, such as self-renewal, differentiation blockade, escape from cell death, and invasiveness, may all be influenced by epigenetic processes in tumor cells. The development of genomic techniques, single-cell profiling, and highly specific tools for exploring epigenomic changes, as well as the use of specific inhibitors of epigenetic modifiers, opens up new horizons for an understanding of the molecular mechanisms underlying carcinogenesis and exciting perspectives for cancer treatment.■

Acknowledgments: This work is supported by grants from l’Association pour la Recherche sur le Cancer (ARC; post-doc fellowship to R. Chaligné), FRM (Equipe FRM to E. Heard), Equipe labellisée “La Ligue Contre Le Cancer” (Equipe Labéllisé to E. Heard), and LABEX DEEP (E. Heard).

1. Baylin SB, Jones PA. A decade of exploring the cancer epigenome—biological and translational implications. Nature Rev Cancer. 2011;11:726-734.
2. Kouzarides T. Chromatin modifications and their function. Cell. 2007;128:693- 705.
3. Tan M, Luo H, Lee S, et al. Identification of 67 histone marks and histone lysine crotonylation as a new type of histone modification. Cell. 2011;146:1016-1028.
4. Wu H, Zhang Y. Mechanisms and functions of Tet protein-mediated 5-methylcytosine oxidation. Genes Dev. 2011;25:2436-2452.
5. Park PJ. ChIP-seq: advantages and challenges of a maturing technology. Nature Rev Genet. 2009;10:669-680.
6. Laird PW. Principles and challenges of genomewide DNA methylation analysis. Nature Rev Genet. 2010;11:191-203.
7. Rando OJ, Chang HY. Genome-wide views of chromatin structure. Annu Rev Biochem. 2009;78:245-271.
8. Farnham PJ. Insights from genomic profiling of transcription factors. Nature Rev Genet. 2009;10:605-616.
9. de Wit E, de Laat W. A decade of 3C technologies: insights into nuclear organization. Genes Dev. 2012;26:11-24.
10. Segal E, Widom J. What controls nucleosome positions? Trends Genet. 2009; 25:335-343.
11. Fatica A, Bozzoni I. Long non-coding RNAs: new players in cell differentiation and development. Nature Rev Genet. 2014;15:7-21.
12. Amaral PP, Dinger ME, Mercer TR, Mattick JS. The eukaryotic genome as an RNA machine. Science. 2008;319:1787-1789.
13. Holliday R. The inheritance of epigenetic defects. Science. 1987;238:163-170.
14. Feinberg AP, Tycko B. The history of cancer epigenetics. Nature Rev Cancer. 2004;4:143-153.
15. Forbes SA, Bindal N, Bamford S, et al. COSMIC: mining complete cancer genomes in the Catalogue of Somatic Mutations in Cancer. Nucleic Acids Res. 2011;39:D945-950.
16. Stratton MR, Campbell PJ, Futreal PA. The cancer genome. Nature. 2009;458: 719-724.
17. van Haaften G, Dalgliesh GL, Davies H, et al. Somatic mutations of the histone H3K27 demethylase gene UTX in human cancer. Nat Genet. 2009;41: 521-523.
18. Morin RD, Mendez-Lago M, Mungall AJ, et al. Frequent mutation of histonemodifying genes in non-Hodgkin lymphoma. Nature. 2011;476:298-303.
19. Schwartzentruber J, Korshunov A, Liu XY, et al. Driver mutations in histone H3.3 and chromatin remodelling genes in paediatric glioblastoma. Nature. 2012; 482:226-231.
20. Easwaran H, Johnstone SE, Van Neste L, et al. A DNA hypermethylation module for the stem/progenitor cell signature of cancer. Genome Res. 2012;22: 837-849.
21. Ohm JE, McGarvey KM, Yu X, et al. A stem cell-like chromatin pattern may predispose tumor suppressor genes to DNA hypermethylation and heritable silencing. Nat Genet. 2007;39:237-242.
22. Bernstein BE, Mikkelsen TS, Xie X, et al. A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell. 2006;125:315-326.
23. Hansen KD, Timp W, Bravo HC, et al. Increased methylation variation in epigenetic domains across cancer types. Nat Genet. 2011;43:768-775.
24. Berman BP, Weisenberger DJ, Aman JF, et al. Regions of focal DNA hypermethylation and long-range hypomethylation in colorectal cancer coincide with nuclear lamina-associated domains. Nat Genet. 2012;44:40-46.
25. Hinoue T, Weisenberger DJ, Lange CP, et al. Genome-scale analysis of aberrant DNA methylation in colorectal cancer. Genome Res. 2012;22:271 282.
26. Frigola J, Song J, Stirzaker C, Hinshelwood RA, Peinado MA, Clark SJ. Epigenetic remodeling in colorectal cancer results in coordinate gene suppression across an entire chromosome band. Nat Genet. 2006;38:540-549.
27. Stransky N, Vallot C, Reyal F, et al. Regional copy number-independent deregulation of transcription in cancer. Nat Genet. 2006;38:1386-1396.
28. Vallot C, Stransky N, Bernard-Pierrot I, et al. A novel epigenetic phenotype associated with the most aggressive pathway of bladder tumor progression. J Natl Cancer Inst. 2011;103:47-60.
29. Wolff EM, Chihara Y, Pan F, et al. Unique DNA methylation patterns distinguish noninvasive and invasive urothelial cancers and establish an epigenetic field defect in premalignant tissue. Cancer Res. 2010;70:8169-8178.
30. Nora EP, Dekker J, Heard E. Segmental folding of chromosomes: a basis for structural and regulatory chromosomal neighborhoods? Bioessays. 2013;35: 818-828.
31. Barr ML, Moore KL. Chromosomes, sex chromatin, and cancer. Proc Can Cancer Conf. 1957;2:3-16.
32. Perry M. Evaluation of breast tumour sex chromatin (Barr body) as an index of survival and response to pituitary ablation. Br J Surg. 1972;59:731-734.
33. Ganesan S, Silver DP, Greenberg RA, et al. BRCA1 supports XIST RNA concentration on the inactive X chromosome. Cell. 2002;111:393-405.
34. Sirchia SM, Ramoscelli L, Grati FR, et al. Loss of the inactive X chromosome and replication of the active X in BRCA1-defective and wild-type breast cancer cells. Cancer Res. 2005;65:2139-2146.
35. Sirchia SM, Tabano S, Monti L, et al. Misbehaviour of XIST RNA in breast cancer cells. PloS One. 2009;4:e5559.
36. Vincent-Salomon A, Ganem-Elbaz C, Manie E, et al. X inactive-specific transcript RNA coating and genetic instability of the X chromosome in BRCA1 breast tumors. Cancer Res. 2007;67:5134-5140.
37. Griffiths EA, Gore SD. Epigenetic therapies in MDS and AML. dv Exp Med Biol. 2013;754:253-283.
38. Rangwala S, Zhang C, Duvic M. HDAC inhibitors for the treatment of cutaneous T-cell lymphomas. Future Med Chem. 2012;4:471-486.
39. Minucci S, Pelicci PG. Histone deacetylase inhibitors and the promise of epigenetic (and more) treatments for cancer. Nat Rev Cancer. 2006;6:38-51.
40. Dawson MA, Prinjha RK, Dittmann A, et al. Inhibition of BET recruitment to chromatin as an effective treatment for MLL-fusion leukaemia. Nature. 2011;478: 529-533.
41. Delmore JE, Issa GC, Lemieux ME, et al. BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell. 2011;146:904-917.
42. Zuber J, Shi J, Wang E, et al. RNAi screen identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature. 2011;478:524-528.
43. Filippakopoulos P, Qi J, Picaud S, et al. Selective inhibition of BET bromodomains. Nature. 2010;468:1067-1073.

Keywords: Barr body; cancer; epigenetic; epigenomic; epimutation