Epigenetics in cancer
Investigation of etiology, mechanisms, and therapy
Cancer has been historically understood as a product of aberrant genes, gene products, and gene regulation that synergistically act to promote cell growth and survival. 
Oncogenes and tumor suppressor genes must become activated and silenced, respectively, often in a cancer-specific pattern, and DNA lesions are the most direct cause of activation and silencing.  A majority of DNA damage is repaired by specialized DNA repair genes, including MGMT and BRCA1, but as mutations accrue the chances of deleterious lesions increase.  Cells have sophisticated mechanisms for controlling cell growth and in dire circumstances undergo apoptosis to limit the potentially harmful effects of DNA damage and the uncontrolled cell growth characteristic of cancer. However, as life expectancy continues to increase, cancer is only becoming more common as age allows for the accumulation of deleterious mutations necessary to transform cells. 
As technology in biomedical research improved, modern cancer research has focused on discovering and understanding oncogenes and tumor suppressor genes to develop more effective and targeted cancer therapies. As our understanding of the genome increased, we discovered additional levels of gene regulation that affected gene expression as well as post-transcriptional gene products. This ‘epigenome’ is commonly defined as the heritable changes in gene expression that are not due to any alterations in the DNA sequence.  Epigenetic modifications can impact gene expression in many forms and represent promising new avenues for cancer therapeutics with increased specificity and decreased toxicity. [5, 6, 7, 8]
Epigenetic modifications and gene regulation
The genome is the backbone of evolution. It allows for heritable characteristics that can be selected for or against in a species’ history, and the genes — units of inheritance — that compose it are subject to a sophisticated system of regulation that allow for a relatively small number of genes to create indescribably complex organisms.  Above the level of direct DNA regulation, epigenetic gene regulation can affect gene expression through numerous mechanisms, including DNA methylation, acetylation, and phosphorylation, chromatin modifications, RNAi, and ncRNAs. 
DNA methylation. The most common epigenetic modification is DNA methylation, the covalent addition of a methyl group to the 5-position of cytosine, in CpG ‘islands’. These islands are in ~40–60% of mammalian gene promoters and thus represent a substantial component of gene regulation in normal and cancerous cells. [1, 2, 17] While methylation patterns — hypo- and hypermethylation — are often aberrant in cancers due to their ability to regulate oncogenes and tumor suppressor genes to favor transformation, DNA methylation is a principle factor in regulating chromatin structure, long-term gene expression, X chromosome inactivation, and intergenerational transfer of experience. [3, 9, 11]
A family of DNA methyltransferases (DNMTs) is responsible for transferring the pattern of DNA methylation onto the newly synthesized strand of DNA during replication. DNA methylation regulates gene expression by organizing chromatin structure — the physical orientation of DNA in the nucleus — into heterochromatin or euchromatin. Heterochromatin is tightly packed and transcriptionally repressed as transcriptional machinery is unable to access genes effectively. [4, 11] Methylation patterns in cancers provide evidence for the effect of epigenetic regulation of oncogenes and tumor suppressor genes in tumorigenesis and may provide a prognostic tool for targeting epigenetic therapies. [7, 8, 12] As a general model, cancer cells have a hypomethylated genome, allowing for oncogene activation, in addition to specific regions of hypermethylation, which silences tumor suppressor genes. [4, 11, 12]
An example of aberrant DNA methylation in the development of cancer is the silencing of MGMT, a DNA repair gene, through promoter hypermethylation. [3, 6, 7, 13, 14] O6-Methylguanine-DNA-methyltransferase (MGMT) encodes O6-alkylguanine-DNA-alkyltransferase (AGT), which is unique among DNA-repair proteins because it acts alone to remove DNA adducts.  MGMT promoter methylation was found in approximately half of glioblastoma cases, but perhaps unexpectedly led to increases in patient treatment success following MGMT silencing. Due to its ability to counteract alkylating agents, AGT actually impedes common chemotherapeutic drugs that rely on alkylating agents to induce apoptosis in cancerous cells.  In normal DNA repair situations, AGT is vital for repairing lesions caused by alkylating agents, which cause DNA polymerase to stall during replication and install the incorrect base (GCAT).  AGT’s repair functionality is one-use-only because AGT irreversibly binds the misplaced methyl group from a lesion and is subsequently polyubiquitinated and degraded. AGT can be mutated in numerous ways to remove its repair functionality at the protein level, but epigenetic MGMT promoter methylation remains the most common method of AGT loss. [6, 13]
DNA acetylation. Acetylation is an opposing epigenetic process to DNA methylation as it encourages euchromatin formation and more accessible histones for transcription.  Unlike epigenetic DNA methylation, acetylation patterns are controlled by the interplay of two protein families: the histone lysine acetyltransferases (KATs or HATs) and lysine deacetylases (HDACs). As expected, both of these families are often mutated in cancers due to their tumorigenic potential. HATs increase acetylation levels and thus can cause overexpression of oncogene gene products if HATs become overactive. An example is MOZ-TIF2 in leukemia, where an increase in HAT recruitment activates genes involved with stem cell like behavior, promoting the self-renewal process.  Two HATs commonly mutated in cancers are p300 and CBP. These proteins act as cofactors for many tumor suppressor gene products, including FOS, p53, Rb, and SV40 large T antigen.  Alternatively, HDACs can decrease genomic acetylation levels and cause an effect similar to hypermethylation via downregulating gene expression by blocking transcriptional machinery. Eighteen HDAC enzymes have been identified along with several HDAC inhibitors that are able to reverse aberrant acetylation patterns in cancer and restore cell cycle controllers and apoptotic pathways.  HATs and HDACs can also function on the histone and chromatin structure level as discussed later.
RNA interference. The RNAi pathway is a flexible regulatory system used by the cell to respond to endogenous and exogenous signals for post-transcriptional gene regulation for cellular processes, including proliferation, differentiation, and apoptosis. [15, 16, 17] Endogenously, non-coding RNA (ncRNA) can be used to synthesize microRNA (miRNA), and exogenously, short interfering RNA (siRNA) can be introduced naturally or by researchers; both miRNA and siRNA utilize the same RNAi pathway and result in targeted gene downregulation.  Exogenously introduced double-stranded RNA (dsRNA) is normally degraded in the cell, but with the help of Dicer, an RNase protein, dsRNA is processed into a 22-basepair long siRNA. Endogenously, the RNAi pathway is accessed via pre-miRNA processing by Drosha, which, after subsequent processing by Dicer, produces mature miRNA for use in the RNA-induced silencing complex (RISC). 
The integrated siRNA or mature miRNA of RISC allows it to target specific mRNA with which it was complementary — though siRNA was found to have more selectivity than miRNA mediated RISC activity. If the sequence match is high, the target mRNA is cleaved by RISC, but if the match is only partial then the mRNA is destabilized, but not destroyed, via readenylation.  The Argonaute protein family is the primary protein of RISC and can perform inefficient RNA interference even without the presence of other cofactors. [15, 17] As an effective form of gene regulation, it is not surprising that many components of the RNAi pathway are mutated in carcinogenesis, especially miRNAs responsible for downregulating oncogene expression.
MicroRNAs and epigenetic regulation have a cyclical relationship in carcinogenesis. Specific miRNAs are often regulated via epigenetic means to silence tumor suppressor genes or activate oncogenes; for instance, the miR-29 family is often downregulated via DNA and histone methylation in lung cancer and lymphomas. While understanding the role of these miRNAs in normal cells has proven challenging, the methylation patterns of specific miRNA promotor regions is a possible prognostic tool. [16, 17] An understudied group of miRNAs has been implicated in the loss of imprinting resulting in loss of heterozygosity in tumor suppressor genes in some cancers, including miR-296 and miR-298.  Epigenetic machinery can also be regulated by miRNAs that affect histone accessibility and DNMTs. For example, miR-449a regulates the expression of HDACs and is often downregulated in carcinogenesis to allow HDACs themselves to regulate tumor suppressor genes.  This level of “crosstalk”, as described by Poddar et al (2017), demonstrates the additional complexity of gene product regulation via genetic and epigenetic means that make cancer difficult to understand.  However, the cyclical nature of miRNA and epigenetic regulation offers new possibilities for epigenetic cancer therapies.
Histone and chromatin modifications. Epigenetic regulation affects all levels of DNA packaging, from DNA sequence to histone and chromatin structure, and is often aberrant in carcinogenesis. [12, 18] Histone methylation and acetylation patterns contribute to gene regulation by compacting DNA into an even more inaccessible form than DNA methylation alone. These modifications rely on a balance of HATs and HDACs that is often modified in cancers to activate oncogenes and silence tumor suppressor genes. For example, Tip60, a HAT, usually activates p53 through increased acetylation, but in many cancers Tip60 is downregulated, leading to silencing of p53 and the loss of the apoptotic pathway.  Histones are made of eight subunits (two dimers and a tetramer) and a charged terminal histone tail that protrudes from the histone and allows for epigenetic regulation through application of methyl, acetyl, or phosphorous group binding.  The Myc gene product has been implicated in altered histone modifications in cancers. Myc null mutants were found to have increased heterochromatin due to increased histone methylation and decreased histone acetylation, but the reintroduction of Myc rescued many of the cells. Myc regulates about 5% of all genes, therefore, its altered expression in carcinogenesis — leading to global chromatin repression — is not unexpected. 
Epigenetic regulation in response to the environment
The epigenome allows for organisms with a fixed genetic code — not accounting for mutations accrued over time — to respond to its changing environment. Additionally, epigenetic modifications are heritable, allowing for offspring to be better suited to their environment based on their parents’ experiences. [2, 5, 20] Methyltransferases and acetyltransferases allow epigenetic modifications to persist through generations and give offspring a selective advantage even though they themselves may have not experienced the stimuli that prompted the epigenetic change. A body of research on this topic has examined the effect of stress on mice and its impact on offspring though several generations. Mice exposed to stress early in their development had altered serotonin dynamics that were found in offspring even though they had not been exposed to the same stress treatment. Another example involved defenses against herbivores in many plant species; parents that were attacked by herbivores developed defensive spines to protect themselves, but even after predation they produced offspring that developed spines from germination even though they themselves had never experienced predation.  Epigenetic inheritance supplements genetic evolution, it does not replace it. Epigenetics can only regulate the existing genetic information, and relies on evolution over time for species to become more permanently adapted to their environment.
Such examples of generational epigenetic transfer of experience have been described in human populations. During World War II, famine was common in many European cities as resources were scarce and tightly regulated. Pregnant women were especially at risk and their offspring were significantly underweight because their mothers had passed on an epigenome adapted to famine conditions. Compared to siblings born after the famine conditions, children had less IGF2 gene product — an insulin-like growth factor — that likely stunted their development and health. Furthermore, a residual effect — though declining — was observed for subsequent generations even though conditions had returned to a healthy level.  In the modern world, obesity has become an epidemic in developed nations and is linked to serious metabolic diseases and carcinogenesis, especially colorectal cancers, responsible for about 10% of cancer mortality.  The proposed mechanism for obesity-induced colorectal cancer involves peroxisome proliferator-activated receptor gamma (PPARγ), a transcription factor involved with adipogenesis and the inflammatory response, that induces apoptosis, prevents proliferation and angiogenesis, and reduces local inflammation. Several miRNAs are upregulated in obesity, due to the chronic hypoxic and inflamed conditions, causing silencing of PPARγ through epigenetic means.  The link between obesity and carcinogenesis further supports the effect of lifestyle on disease prevention and presents avenues of cancer prevention similar to that of lung cancer prevention by lowering smoking rates. Additionally, the effects of epigenetic regulation can persist through several generations and thus the actions and experiences of one’s parents may predispose individuals to cancers.
The reversible nature of epigenetic modifications makes them promising prospects for cancer therapeutics. Aberrant methylation and acetylation patterns in carcinogenesis are theoretically treatable with little toxicity because they represent a return to natural patterns that should not affect non-transformed cells. [5, 6, 8, 10, 19] However, depending on the epigenetic component of carcinogenesis, the wide range of genes that are regulated epigenetically does present concerns over unintentional targets of epigenetic therapies. For example, global hypomethylation of the genome is common in cancer, and drugs are available that regulate DNMTs to combat hypomethylation. The potential issue arises when upregulation of DNMTs causes hypermethylation of tumor suppressor gene promotors or if downregulation of DNMTs — not often used in cancer therapies — causes further hypomethylation of oncogene promoters.  Promising work is being conducting using various experimental methods — primarily methylation and acetylation assays — to create cancer specific epigenetic patterns and targeted drug regiments for patients. [5, 8, 10]
Drugs targeting aberrant DNA methylation. As the most common epigenetic modification in carcinogenesis, drugs targeting the level of methylation in cancer cells were some of the earliest epigenetic therapies developed. Originally developed as cytotoxic drugs, 5-azacytidine and 5-aza-2’-deoxycytidine are nucleoside analogues effective in inhibiting DNA methylation by covalently bonding to DNA methyltransferases during replication.  By affecting DNMTs during DNA replication, aza drugs benefit from increased specificity to replicating cells, but high doses have been linked to cytotoxicity in pre-clinical models. As is often the case in cancer chemotherapy, it is likely that epigenetic treatments will be most effective when used in conjunction with other therapies. Researchers hypothesize that some cancers may be especially vulnerable to epigenetic therapy because a single epigenetic mutation can regulate many genes, thus a drug targeting that mutation is, practically speaking, targeting all of its component genes.  Various forms of azacytidine were approved by the Food and Drug Administration (FDA) in 2004 and are often used to treat myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) (National Cancer Institute Approved Drug List).
DNA methylation inhibition can also be achieved through non-nucleoside analogues, though fewer of these drugs are being pursued in clinical trials. Most non-nucleoside analogue agents are small-molecule inhibitors that can bind directly to DNMTs and cause competition or allosteric inhibition. While the possible benefits of increased specificity and not having to be incorporated into DNA are promising, pre-clinical trials of several drugs have not shown efficacy.  Other non-nucleoside analogues include antisense oligonucleotides that prevent translation of target gene products, for example, MG98 targets DNMT1, theoretically limiting DNA methylation. However, in Phase I clinical trials, MG98 showed no anti-tumor activity. This may be due to overlapping functions of the DNMT protein family, which allows for compensation following the loss of DNMT1. [4, 8]
Drugs targeting histone deacetylase. There are several classes of HDAC inhibitors, of which all induce a loss of HDAC activity and subsequent return to normal acetylation patterns in transformed cells.  A restoration of normal acetylation should reactivate tumor suppressor genes and silence oncogenes, thus allowing for normal processes regulating the cell cycle and apoptosis to resume. Therefore, HDAC inhibition should have little cytotoxicity because non-transformed cells should not be affected by the epigenetic therapy. [5, 10] However, due to the difficulty in determining the pleotropic effects of broad epigenetic therapy, like HDAC inhibition, researchers are especially cautious to limit unintentional consequences of hyperacetylation in non-cancerous cells.  A common HDAC inhibitor target is the p21 gene, a cell-cycle kinase inhibitor, that is commonly upregulated in tumor cells after HDAC inhibitor activity. p21 upregulation can compensate for a loss of p53 — a common outcome during carcinogenesis — by arresting the cell cycle in G1 or G2 phase in a similar pathway as p53. 
Combination epigenetic therapy. Simultaneous use of DNA demethylating agents and HDAC inhibitors have the potential benefit of fully restoring epigenetic patterns. The reactivation of tumor suppressor genes can have secondary effects in cancer therapy by making cancer cells more susceptible to other chemotherapies and reactivating DNA repair genes to protect the genome from accruing further mutations.  Depending on the agents used in treatment, epigenetic therapies can lower the apoptotic threshold of cancer cells.  For example, aza drugs have been used in combination with conventional chemotherapies to some clinical success in patients resistant to chemotherapy and immunotherapy. Preventative epigenetic therapy is being pursued in patients with epigenetic lesions that have not yet caused transformation to restore normal epigenetic patterns. To this end, cancer specific patterns of epigenetic modifications are increasingly important for epigenetic therapy so that epigenetic drugs can be targeted on a cancer-specific basis to reduce negative pleotropic effects. 
Obstacles to epigenetic therapy. There are potential obstacles to using epigenetic therapies effectively to treat cancer. Just as the reversible nature of epigenetic modifications allows for treatment of aberrant epigenetic patterns, epigenetic therapies are not necessarily permanent solutions. For example, many small-molecule inhibitors bind allosterically to their target protein, but if the drug is not administered at a high enough dosage or for long enough then the protein will continue to function as before treatment. Continuous treatment with drugs and/or pairing an epigenetic treatment with another form of cancer therapy could limit the potential downfall of reversibility.
Genomic changes in methylation and acetylation patterns due to epigenetic treatment can have pleiotropic effects on non-target genes. DNA methylation inhibitors and HDAC inhibitors induce cell-cycle arrest and apoptosis through upregulation of p21 and p53, which is useful in cancer treatment, but if non-target cells are affected then widespread tissue damage can occur.  Another negative effect of broad hypomethylation due to epigenetic treatment is activation of transposable elements (TE) — such as Alu sequences — that can cause genomic damage by inserting themselves into genes. Our understanding of transposable element regulation is incomplete, and while it is likely that TEs are regulated by more than just epigenetic means, these epigenetic therapies may affect TE spread. 
Concluding remarks and future directions
The potential power of epigenetic therapy for cancer is exciting. The ability to affect aberrant gene regulation without potentially harmful gene editing or broad cytotoxic treatment would represent a major step in cancer treatment and standard of life for cancer patients. As more research is conducted, it appears that epigenetic therapy is not the ‘silver bullet’ or ‘cure’ for cancer, however, it is an increasingly useful supplement for existing chemotherapies. Drug combination still represents our best effort in cancer treatment and epigenetic regulation is especially powerful in restoring many potential tumor suppressor genes. The most exciting benefit of epigenetic therapy is the ability to utilize the patients’ natural tumor suppressing mechanisms by restoring normal patterns of epigenetic modifications. [4, 8, 10]
Drug research is pursuing epigenetic treatments for many aspects of regulation, including HDACs, DNMTs, siRNAs, and HMTs, in addition to a plethora of small-molecule inhibitors.  Several drugs have already been FDA approved for treating specific cancers, and dozens more are in Phase I/II clinical trials.  Compiling cancer-specific epigenetic patterns is vital for effective cancer treatment and drug research, as well as predictions for patient response.  Not all cancers are equally susceptible to epigenetic therapies, for instance, hematopoietic cancers are more effectively treated through epigenetic therapy than solid tumors are. For example, in AML patients with TET family protein mutations, epigenetic therapy through DNA methylation inhibition was found to be clinically successful.  This is likely due to the ease of drug administration allowed by non-solid tumors, and as each tumor cell must receive the epigenetic treatment for it to be effective, this represents a major challenge for future research. 
Thanks for reading! I hope this review is helpful to those who find it. It’s an adaptation of a final article I wrote during a cancer biology seminar class at Davidson College.
 Esteller, M. Epigenetics in Cancer. N. Engl. J. Med. 358, 1148–1159 (2008).
 Lund, A. H. & Van Lohuizen, M. Epigenetics and cancer. Genes Dev. 18, 2315–2335 (2004).
 Lahtz, C. & Pfeifer, G.P. Epigenetic changes of DNA repair genes in cancer. J. Mol. Cell. Biol. 3, 51–58 (2011).
 Baylin, S.B. & Herman, J.G. DNA hypermethylation in tumorigenesis. TIG 16, 168–174 (2000).
 Egger, G., Liang, G., Aparicio, A. & Jones, P. A. Epigenetics in human disease and prospects for epigenetic therapy. Nature 429, 457–463 (2004).
 Gerson, S.L. MGMT: Its Role in Cancer Aetiology and Cancer Therapeutics. Nat. Rev. Cancer 4, 296–307 (2004).
 Hegi, M.E., Diserens, A.C., Gorlia, T., Hamou, M.F., Tribolet, N., Weller, M., Kros, J.M., Hainfellner, J.A., Mason, W., Mariani, L. et al. MGMT Gene Silencing and Benefit from Temozolomide in Gliobastoma. N. Engl. J. Med. 352, 997–1003 (2005).
 Yoo, C.B. & Jones, P.A. Epigenetic therapy of cancer: past, present and future. Nat. Rev. Drug Discov. 5, 37–50 (2006).
 Pearson, H. What is a gene? Nature 441, 398–401 (2006).
 Dawson, M.A. & Kouzarides, T. Cancer Epigenetics: From Mechanism to Therapy. Cell 150, 12–27 (2012).
 Robertson, K.D. DNA methylation, methyltransferases, and cancer. Oncogene 20, 3139–3155 (2001).
 Seligson, D.B., Horvath, S., McBrian, M.A., Mah, V., Yu, H., Tze, S., Wang, Q., Chia, D., Goodglick, L. & Kurdistani, S.K. Global Levels of Histone Modifications Predict Prognosis in Different Cancers. Am. J. Pathol. 174, 1619–1628 (2009).
 Esteller, M., Garcia-Foncillas, J., Andion, E., Goodman, S.N., Hidalgo, O.F., Vanaclocha, V., Baylin, S.B. & Herman, J.G. Inactivation of the DNA-repair gene MGMT and the clinical response of gliomas to alkylating agents. N. Engl. J. Med. 343, 1350–1354 (2000).
 Wood, R.D., Mitchell, M., Sgouros, J. & Lindahl, T. Human DNA Repair Genes. Science 291, 1284–1289 (2001).
 Hammond, S.M. Dicing and slicing: The core machinery of the RNA interference pathway. FEBS Letters 579, 5822–5829 (2005).
 Poddar, S., Kesharwani, D. & Datta, M. Interplay between the miRNome and the epigenetic machinery: Implications in health and disease. J. Cell Physiol. 232, 2938–2945 (2017).
 Sato, F., Tsuchiya, S., Meltzer, S.J. & Shimizu, K. MicroRNAs and epigenetics. FEBS J. 278, 1598–1609 (2011).
 Knoepfler, P.S., Zhang, X., Cheng, P.F., Gafken, P.R., McMahon, S.B. & Eisenman, R.N. Myc influences global chromatin structure. EMBO J. 25, 2723–2734 (2006).
 Ellis, L., Atadja, P.W. & Johnstone, R.W. Epigenetics in cancer: Targeting chromatin modifications. Mol. Cancer Ther. 8, 1409–1420 (2009).
 Harper, L.V. Epigenetic Inheritance and the Intergenerational Transfer of Experience. Psychol. Bull. 131, 340–360 (2005).
 Motawi, T.K., Shaker, O.G., Ismail, M.F. & Sayed, N.H. Peroxisome Proliferator-Activated Receptor Gamma in Obesity and Colorectal Cancer: the Role of Epigenetics. Sci. Rep. 7, 1–8 (2017).