Chapter 471: The Role of Epigenetics in Disease and Treatment

The term epigenetics was first coined by Waddington in 1942, as he sought to explain how changes in phenotype could occur throughout development independent of any changes to genotype. Appending the prefix epi- (Greek, meaning “over, outside of, around”) to genetics aptly describes the numerous mechanisms by which gene expression and phenotypes are influenced, independent of any changes to the underlying DNA sequence. Today, epigenetics occupies one of the most exciting topics in biology and medicine, offering profound opportunities for discovery, as well as promise for the development of new therapies for disease. Interdisciplinary by nature, the field crosses virtually all areas of science and medicine: chemistry and genetics, development and differentiation, immunology, cancer, aging, and neuroscience.

The continuous introduction of ever more powerful techniques for interrogating the epigenome has led epigenetics to become one of the most innovative fields within the biomedical sciences. Given the vast expanse of the topic and limitations of space, in this chapter we provide a broad overview of the field and highlight key areas from across the landscape of biomedicine where epigenetics plays critical roles in disease, and importantly, where epigenetics-based therapies have demonstrated success in clinical medicine.

Fundamental to epigenetic regulation is the intricate organization of each cell’s genome into chromatin (Chap. 456). The fundamental unit of the packaging into chromatin is the nucleosome, consisting of 147 base pairs of DNA wrapped around an octamer of 8 histone proteins (two copies of each of the four core histone proteins: H2A, H2B, H3, and H4). The level of compaction of this chromatin structure determines the accessibility of the DNA strand to transcription factors, the DNA repair machinery, and other DNA-binding entities. Thus, compaction has a profound influence on gene expression levels and on local DNA mutation rates. Open regions of chromatin (euchromatin) tend to be transcriptionally active, whereas compacted chromatin (heterochromatin) tends to be transcriptionally repressed.

Histones include the four core histones, which are the most abundant and most frequently found throughout the genome, and variant histones of H2A, H2B, and H3. The structure of core and variant histones include amino- and carboxyl-terminal “tails,” which are extended and unstructured, and highly conserved globular domains. The x-ray crystal structure of the nucleosome particle has illuminated the dynamic alterations of chromatin by an astounding range of regulatory mechanisms, summarized below.

The three main processes that regulate chromatin compaction, and thus access to the DNA template, include direct methylation modifications of the DNA strand itself, post-translational modifications of histones, and remodeling of nucleosomes to alter their location and composition with variant histones (Fig. 471-1). The major modification of DNA is cytosine methylation of CpG dinucleotides (5-mC), associated with gene repression and catalyzed by the DNMT1, DNMT3A, and DNMT3B enzymes. DNMT3A and 3B catalyze the addition of methyl groups on unmethylated DNA de novo at CpG dinucleotides that are typically located throughout transcribed genes and in intergenic regions, but lacking at promoters, while DNMT1 is critical for the maintenance of the methylation state after DNA replication and after transcription during the S phase of the cell cycle. To further alter and reverse methylation, the TET enzymes (TET1–3) progressively oxidize 5-methylcytosine (5-mC) to 5-hydroxymethylcytosine (5-hmC), to 5-formylcytosine (5-fC) and 5-carboxylcytosine (5-caC), which are unable to be recognized by DNMT1 but can be removed by additional enzymes. Hence, these are mechanisms to passively lose 5-mC following DNA replication, or to actively remove 5-mC, both returning to unmethylated cytosine.

Histone post-translational modifications (hPTMs) are rich sources of diverse signaling to, and marking of, the chromatin template, including at least 60 different covalent chemical modifications on the histone N- and C-terminal tails and within the globular domains. The hPTMs are added (written) and removed (erased) by enzymes, and also serve as sequence- and PTM-specific binding surfaces for effector proteins and complexes (readers) to carry out a wide range of downstream actions including transcription, replication, and DNA repair and recombination. One key point is that the staggering numbers of writers, erasers, and readers provide unlimited potential for diagnostic and therapeutic pharmacologic discovery.

Throughout this chapter, we focus on histone methylation and acetylation, the most abundant and the most well-studied hPTMs (Fig. 471-1), although several others, such as serine/threonine/tyrosine phosphorylation, lysine ubiquitination, lysine SUMOylation, and lysine ADP-ribosylation also play important roles in epigenetic regulation. For instance, histone phosphorylation targets histone H2A at Ser139 (γH2A.X), which marks DNA double-strand breaks immediately following damage and is critical for the recruitment of the DNA repair machinery. Histone mono-ubiquitination functions similarly to other hPTMs, in signaling and marking the chromatin template, in particular serving to mark the initiation region or elongation of transcribed genes for future rounds of transcription, whereas histone SUMOylation plays a role in transcriptional repression. Poly-ubiquitination serves to tag proteins for degradation by the proteasome, and dysfunction in this system may play a role in the pathogenesis of neurodegenerative diseases such as Alzheimer’s, Parkinson’s, and Huntington’s disease. ADP-ribosylation involves a class of enzymes, the poly-ADP-ribose polymerases (PARPs), which transfer ADP-ribose units from NAD+ to a variety of nuclear proteins. This PARylation alters the chromatin environment through the recruitment and modification of chromatin-associated proteins. In general, future studies of the wealth of types and functions of hPTMs will enhance our understanding of these chromatin-based mechanisms and processes and will illuminate new opportunities and targets for therapies.

In contrast, there is extensive understanding of histone lysine acetyltransferases (KATs) and methyltransferases (KMTs). KATs, previously known as HATs, were among the first identified histone modification enzymes. They attach acetyl groups on the lysine residues of histone tails and other proteins, resulting in both a novel side chain (acetyl-lysine) and an increase in negative charge (from positive charged lysine to neutral acetyl-lysine). This alteration results in loosening of chromatin structure to become more permissive to the binding of transcription factors, and it also creates a novel binding surface for the association of reader proteins. Acetylation on core histones, such as lysine 9 on histone H3 (H3K9ac) or lysine 27 (H3K27ac), is typically associated with transcriptional activation. Acetylation is very dynamic and can be rapidly removed by histone deacetylases (HDACs), of which there are multiple classes, including HDACs and sirtuins (NAD-dependent deacetylases), which return the lysine to unmodified ground state.

Methylation of histone tails by KMTs provides more nuanced regulation, in that particular methylated lysines are associated with transcriptional activation (e.g., H3K4me3, H3K36me3, H3K79me3), transcriptional repression (e.g., H3K27me3), or DNA repeat and centromeric silencing (e.g., H3K9me3). The output is strictly determined by effector protein binding, as methylation of lysine does not alter side chain electrostatic charge. Lysine methylation is a more stable chemical modification than is acetylation, and, while demethylases are identified for some of the specific methylated sites (H3K4, H3K9, H3K36, H3K27), it is provocative that H3K79 and H4K20 demethylases have not yet been discovered.

Frequently coordinating with histone modification enzymes are nucleosome remodeling enzymes, which use the energy derived from the hydrolysis of ATP to reposition and remove nucleosomes along the DNA template, and to exchange core histones and variant histones (including variants that are located at the transcriptional initiation sites [H2AZ] and over the transcribed genes [H3.3]). These complexes activate or repress transcription. The SWI/SNF family creates nucleosome-free regions for transcriptional activation, the ISWI family evenly spaces nucleosomes to repress transcription, and the INO80 family exchanges H2A with H2AZ at transcription start sites to poise transcriptional activation. Other remodeling complexes play key roles in the DNA damage response and apoptosis, among additional genomic processes.

Because multiple enzymes redundantly write, erase, and recognize many of the hPTMs, there is great complexity and the potential for fine-tuning of gene regulation. While extensive knowledge gaps remain to fully explicate mechanisms of chromatin regulation, epigenetics has become a fully established discipline within biomedical research. In the coming years, it is likely that the basic understanding of these processes will be harnessed for further betterment of human health.

Epigenetic processes are critical to organismal development and to cellular differentiation and reprogramming of cell fate (Fig. 471-1). Transcription factors establish the epigenomic landscape that enables and stabilizes cell type–specific gene expression while simultaneously ensuring stable repression of alternative cell fates. This results in chromatin profiles that display remarkable cell-type specificity in differentiated cells, particularly at the key regulatory nodes of gene enhancers. In fact, epigenome profiling of the chromatin landscape in tumors of unknown cell origin can provide a better index of the origin tissue than does sequencing the tumor itself to catalog genetic mutations.

The cell type–specific epigenetic program is first derived from the template of embryonic stem cells, where numerous genes required for differentiation exist in a “bivalent” state, marked by both the activating histone modification, H3K4me3, and the repressive modification, H3K27me3. From this state, the genes are thought to be “poised” and ready either for activation or for repression, depending on their subsequent cell fate. Critical genes directing toward a specific cell fate will be turned on, whereas genes leading toward alternative fates will be repressed. Once differentiated, an epigenetic barrier will prevent the cells from returning to the stem cell state. For example, heterochromatin in the form of H3K9me3 can serve as a barrier to cellular reprogramming when attempting to create induced pluripotent stem cells, and inhibiting the enzymes that catalyze H3K9me3, such as SUV39H1, can enhance reprogramming efficiency.

DNA methylation contributes to the specification of cell fate and to other developmental pathways. DNA methylation alterations are involved in critical processes ranging from sex chromosome dosage compensation to coordinating expression of imprinted genes. Disruption of this latter process can lead to imprinting disorders including Prader-Willi syndrome, Angelman syndrome, and Beckwith-Wiedemann syndrome.

Beyond embryonic development, epigenetics can provide the necessary coordination and balance between adult stem cell renewal and differentiation. This epigenetic control is critical, as impaired self-renewal can lead to stem cell exhaustion and premature aging, while excessive self-renewal may promote cancer. Key epigenetic regulators tend to play conserved roles across diverse tissue types. For instance, BMI1, a component of the polycomb repressive complex 1 (PRC1), is required for stem cell proliferation and self-renewal, and its ablation leads to stem cell depletion in hematopoietic, epidermal, muscle, intestinal, and mammary stem cells. Similarly, the DNA methyltransferase DNMT1 also is required for stem cell self-renewal in hematopoietic, epidermal, and mammary stem cells. HDACs 1 and 2 possess some overlapping functions and are required for normal epidermal differentiation. Likewise, a loss of these enzymes in hematopoietic stem cells can lead to failure of differentiation and severe anemia. These factors represent repressive chromatin regulation, leading to the concept that restraining specific transcription pathways related to differentiation are crucial to maintaining undifferentiated stem cell pools.

The epigenetic regulation of the tumor suppressor p16 (CDKN2A) locus during differentiation provides a prime example of this finely tuned system. For example, as mentioned above, DNMT1 is necessary for self-renewal in human epidermal stem cells. Levels of DNMT1 are high in the basal undifferentiated layer of the epidermis, decreasing progressively with epidermal stratification, leading to derepression of the tumor suppressors p16 and p15, thereby promoting cell cycle arrest and full differentiation. BMI1 displays a similar phenotype in both hematopoietic and epidermal stem cells, repressing key genes that promote differentiation, such as p16 and p19ARF. Consistently, a loss of BMI1 leads to premature differentiation and defective self-renewal. In addition to the repression provided by DNMT1 and BMI1, the p16 locus is highly decorated with the repressive H3K27me3 catalyzed by EZH2 in epidermal stem cells. Then, during epidermal differentiation, H3K27me3 is removed by the KDM6B (JMJD3) histone demethylase. Loss of this control over programmed p16 expression occurs in epithelial cancers, such as squamous cell carcinoma (SCC), where EZH2 is overexpressed and KDM6B expression is lost. In the breast, progesterone can lead to increased levels of EZH2 to promote mammary epithelial cell proliferation. However, excessive EZH2 expression occurs in breast cancers, exemplifying how epigenetics can integrate environmental signals and have a profound influence on the fine balance between stem cell maintenance and overt carcinogenesis. Indeed, loss of key chromatin regulation that promotes cell differentiation, and gain of activities that promote stemness, is a recurrent theme in cancer.

Chromatin modifying enzymes also play a major role in influencing cell type specificity. High levels of EZH2 that modify H3K27me3 promote adipogenesis while simultaneously inhibiting osteogenesis. In contrast, the H3K27me3 demethylases, KDM6A (UTX) and KDM6B (JMJD3), derepress those same genes, driving stem cells toward osteogenesis. Through interactions with tissue-specific master regulators, epigenetic modifiers also shape cell type specificity. In the epidermis, p63, the p53 family member that is a master regulator of the epidermal compartment, interacts with several chromatin regulators including HDAC1 and HDAC2, SATB1, and BRG1 to orchestrate epidermal differentiation. Similarly, the gene-activating H3K4 histone methyltransferases, MLL3 (KMT2C) and MLL4 (KMT2D), are required for adipogenesis by forming a complex with the transcriptional activator ASC2 and the transcription factor PPARγ to induce adipogenic genes.

One of the fascinating aspects of epigenetics is that it represents a mechanism for direct connection between the environment and gene expression. Numerous studies in the field of metabolism have identified a complex interplay between diet, metabolism, and the epigenome (Fig. 471-1). Seminal findings in Drosophila and mice have shown that changes in diet, particularly the paternal diet, and other environmental factors, can influence the metabolism of offspring, ultimately promoting obesity in later generations. Epidemiologic studies in humans have supported these results, as the nutritional status of grandparents has been correlated with phenotypic effects in grandchildren. In fact, diet can directly affect the levels and activity of chromatin modifiers. For instance, HDAC9 is elevated by high-fat diets, and inhibition of HDAC9 can be protective against the deleterious effects of high-fat diets in mice, including weight gain, decreased glucose tolerance, and increased insulin resistance.

These connections are driven by metabolites that constitute key cofactors for post-translational histone modifications such as acetylation and methylation. These include acetyl-CoA for histone acetylation and S-adenosylmethionine (SAM) for histone and DNA methylation. Levels of acetyl-CoA, in comparison to all measured metabolites, are indeed the best predictor of histone acetylation levels. Consistent with this, increased acetyl-CoA correlates with rising levels of total histone acetylation, including at the promoters of growth-associated genes. This increase in nuclear acetylation is associated with cell cycle progression and proliferation, and it can have clinically relevant downstream effects. For example, high levels of acetyl-CoA can delay stem cell differentiation and suppress autophagy. The oncogenes MYC and AKT can both hijack metabolic networks to enhance nutrient uptake by cancer cells, thus promoting acetyl-CoA production and resulting in both the initiation and progression of tumorigenesis.

Methylation is also altered by metabolism. Dietary factors are estimated to explain 30% of the variation in human serum methionine concentration, and this can alter histone methylation. For example, dietary methionine availability and intracellular production of SAM affects the levels of histone H3K4me3 associated with transcriptional activation. Furthermore, these fluctuations can have critical physiologic consequences: DNA methylation levels in rectal mucosa and colonic polyps are increased by higher levels of dietary folate, and a diet low in methyl donors reduces the formation of gastrointestinal cancers in mice predisposed to these tumors. Methionine metabolism and the availability of SAM regulates stem cell differentiation and contributes to carcinogenesis. For instance, cancers that display hypermethylation, including those with IDH mutations, are associated with poorly differentiated gene expression profiles. In contrast, loss of the MTAP gene, which is part of the 9p21 locus containing p16, and one of the most frequent events in human cancer, disrupts normal methionine metabolism. While this lowers methylation levels, interestingly, it can also sensitize cancer cells to inhibitors of the PRMT5 methyltransferase, therefore opening a new therapeutic opportunity. These observations illustrate how connections between epigenetics and metabolism can generate unanticipated advances in medicine.

Cancer is now understood to be a mixed genetic and epigenetic disease, as epigenetic dysregulation is pervasive in human cancers (Fig. 471-1). Beyond simple activation of oncogenes or reduced expression of tumor suppressors, epigenetic mechanisms can contribute to chemotherapy resistance and to failure of antitumor immunity. Accordingly, the development of drugs targeting epigenetic pathways is one of the most active areas of clinical and pharmaceutical development, with several compounds already approved for human use and shown to be effective in a variety of cancers. Epigenetic perturbations in cancer largely affect chromatin-regulating enzymes, which represent robust targets for development of novel small-molecule inhibitors, especially as compared with canonical oncogenic transcription factors (e.g., MYC) and tumor suppressors (e.g., p53).

Epigenetics can contribute to carcinogenesis in a variety of ways. First, on a global scale, chromatin organization is the single most influential factor in determining local mutation rate across the genome. Analysis of abundant tumor sequencing data has demonstrated that heterochromatic regions of the genome contain a higher frequency of mutations compared with more open euchromatic regions. This difference is due to the improved accessibility of the DNA repair machinery to less compact, more open regions of chromatin.

The first discovery of an epigenetic mutation was found in 1998 when the chromatin remodeler SMARCB1 was shown to drive the formation of malignant rhabdoid tumors. Extensive sequencing of human tumors from the majority of cancer types has been performed by The Cancer Genome Atlas (TCGA) consortium, and, remarkably, 25–30% of identified cancer driver mutations occur in chromatin regulatory proteins. Similar to SMARCB1, numerous other chromatin modifiers (e.g., methyltransferases MLL3 and MLL4, and acetyltransferases EP300 and CBP) and nucleosome remodeling enzymes and associated complex components (e.g., SMARCA4, SMARCA2, ARID1A) are heavily mutated and inactivated in many cancers. The majority of these mutations are loss-of-function mutations, and, indeed, enzymes like MLL4 and demethylase KDM6A possess tumor-suppressive activity. In contrast, the H3K27me3 histone methyltransferase EZH2 is an oncogene, and accordingly, it is overexpressed in many advanced-stage or metastatic solid tumors such as breast cancer, prostate cancer, and melanoma. Mechanistically, EZH2 represses the p16 tumor suppressor and other cell cycle genes required for cell cycle exit via H3K27me3 deposition. Consistent with a broad growth regulatory role, EZH2 inhibitors are therapeutically successful for a number of cancers in preclinical models and are being actively studied for B cell lymphoma, melanoma, and other solid tumors.

Recently, provocative evidence has emerged for a direct tumorigenic role of histones based on the discovery of causative mutations, such as histone H3 mutations identified in pediatric high-grade gliomas. Specifically, the majority of these mutations are in the H3 variant H3.3, where lysine 27 is replaced by methionine (K27M). Similarly, over 90% of chondroblastomas replace lysine 36 with methionine (K36M) in histone H3.3. These effects appear to be dominant negative because (1) in H3.3 these are heterozygous mutations, and (2) the mutations also occur in the canonical H3, which exists in approximately 30 orthologous genes in the human genome. Thus, a minority of H3/H3.3 mutant protein leads to global defects in the associated histone modifications (K27 or K36 methylation), possibly via irreversible inhibition of the cognate enzymes by the mutant histones. These histone mutations promote resistance to apoptosis and failure of normal differentiation in a number of pediatric cancers.

Beyond mutations, genetic translocations involving chromatin modifiers also implicate chromatin pathways as direct drivers in cancer. MLL1, the H3K4 histone methyltransferase, is a frequent translocation partner occurring in adult and pediatric acute myeloid leukemia (AML), and in approximately 80% of infant acute lymphoid leukemia (ALL) cases. MLL1 can fuse with more than 70 translocation partners, and these mutant proteins prevent normal hematopoietic differentiation. Consistent with a causative role of MLL1 in these gene fusions, drugs inhibiting the catalytic activity of MLL1 are effective in preclinical models of AML.

Given the abundance of epigenetic abnormalities in cancer combined with the inherent reversibility of epigenetic changes, extensive efforts are underway to develop epigenetic drugs. The first epigenetic therapeutic involved the use of DNA methylation inhibitors (DNMTi) to reactivate tumor suppressor genes. Interestingly, the mechanism of traditional chemotherapeutics, such as azacitidine and decitabine, is to inhibit DNMT1, thereby promoting global hypomethylation; these are currently in clinical use for myelodysplastic syndrome (MDS) and AML. In a second broad mechanism, loss of acetylation occurs in many cancers, and thus HDAC inhibitors (HDACi) are under intensive development. HDACi are effective and approved for treatment in cutaneous T cell lymphoma and multiple myeloma. Bromodomain (BRD)-containing proteins bind to lysine acetylated target proteins, including histones, and rationally designed BET inhibitors (BETi) block their binding. BETi reduce the amplified expression of oncogenes such as MYC in hematologic cancers. Current studies are focused on optimizing combinatorial therapies of these pan-DNMTi and pan-HDAC inhibitors with conventional chemotherapies and immunotherapies, particularly given the ability of epigenetic therapeutics to promote re-expression of tumor antigens and interferon (IFN)-mediated antitumor immunity.

There are several hundred chromatin enzymes and binding proteins in the human genome, and the current focus is on the identification of more specific inhibitors. Indeed, targeted inhibitors of numerous mutated chromatin regulators have been developed, with more than 30 compounds currently in various stages of development and preclinical trials. Some notable examples showing early clinical success include EZH2 inhibitors for lymphomas, sarcomas, and melanoma, IDH inhibitors for AML and gliomas carrying mutant IDH1 or IDH2 genes, LSD1 inhibitors for AML and small cell lung cancer, and DOT1L and MLL1 inhibitors for leukemias with activated MLL1.

Like many diseases of aging, human aging itself results from the complex interplay between genes and the environment. Evidence that the epigenome may be the key link between these processes derives from observations that numerous environmental stimuli and stressors—ranging from diet and exercise to hormones and circadian rhythms—contribute to both aging and epigenetic alterations (Fig. 471-1). Thus, a lifetime of exposures progressively disrupts the chromatin landscape. These age-dependent changes in chromatin organization increase the susceptibility of the genome to mutations and also reduce transcriptional fidelity. Further, provocative findings in model systems demonstrate that stress-induced epigenetic changes can be transmitted over several generations and can even affect the lifespan of later generations. Among these global epigenetic alterations, there is dysregulation of histone modifications and a general loss of histone proteins with aging across taxa. Amazingly, experimental increases in histone levels, particularly histones H3 and H4, but not H2A or H2B, can reverse these age-related changes in mammalian cells and yeast.

Thus, the sum of current evidence suggests a model of aging via a general increase in activating epigenetic modifications along with a loss of repressive modifications. Together these changes create a state of transcriptional instability and “noise” that inhibit accurate transcription. Cells from patients with Hutchinson-Gilford progeria syndrome (HGPS), the most severe form of human premature aging, display reduced levels of both H3K9me3 and H3K27me3 repressive chromatin. In another premature aging disease, Werner syndrome, DNA damage induces global loss of H3K9me3 and H3K27me3 due to the inherent absence of the Werner syndrome ATP-dependent DNA helicase, which is critical for DNA repair. Such heterochromatin loss is not limited to premature aging conditions, as aged cells derived from healthy older humans display age-dependent loss of H3K9me3 leading to aberrant expression of normally repressed transposable elements. Activation of these mobile elements correlates with neurodegenerative disorders and may also promote other aging-related phenotypes such as cancer. Human fibroblasts undergoing cellular senescence (exit from cell cycle due to replicative or other stress) undergo destabilization of compact heterochromatin adjacent to the nuclear periphery, in so-called lamin-associated-domains (LADs). At LADs, in addition to a reduction of repressive histone modifications as discussed above, there are broad new regions of the euchromatic histone modification H3K4me3.

In addition to age-associated alterations of histone modifications, direct manipulation of chromatin modifying enzymes that control these marks affects the balance between heterochromatic and euchromatic regions, and it alters the lifespan of model organisms. Inhibiting the H3K27me3 histone demethylase UTX (KDM6A) results in increased repressive H3K27me3 and extended lifespan in Caenorhabditis elegans. Consistent with this, genetic reduction of enzymes (ash-2, set-2, wdr-5) that add the activating H3K4me3 histone modification also extends lifespan in C. elegans. The consequences of these genetic manipulations nicely correspond to the observed changes in histone modifications as described above. Beyond histone modifying enzymes, dysregulation of the levels or function of chromatin remodelers can also affect lifespan in model organisms. This dysregulation occurs in humans as well, as in the nucleosome remodeling deacetylase complex (NuRD), which is reduced in HGPS fibroblasts and in aged healthy donors.

In addition to age-related changes in histone methylation, histone acetylation also contributes to aging phenotypes. Dysregulation of histone acetyltransferases (HATs) and HDACs is associated with reduced longevity across model organisms. Further, sirtuin deacetylases (class III NAD⁺-dependent HDACs) promote healthspan and lifespan across species as key mediators of pro-longevity effects of caloric restriction. Indeed, loss of Sirt6 results in premature aging in mice. As discussed previously, metabolism and acetylation are intricately linked, and the sirtuins, via NAD⁺ levels, and other HDACs may play key roles connecting the environment, gene expression, and physiologic output. For instance, exercise in humans reduces activity of HDACs 4 and 5, leading to increased H3K36ac in skeletal muscle, which likely promotes beneficial gene expression.

Epigenetic alterations with aging are not limited to histone modifications and extend to DNA methylation. Consistent with the histone patterns, DNA methylation data support the model described above—that is, general decompaction of the epigenome with aging. Specifically, levels of 5-methylcytosine (5-mC) are reduced in senescent humans cells, and global DNA hypomethylation occurs across the human genome with aging. Concurrent with this overall hypomethylated state, there are local regions of hypermethylation focused near CpGs at gene promoters, particularly at genes that maintain cellular differentiation and cell identity. This epigenetic disruption during aging thus leads to profound changes in transcription. For example, in HSCs, DNA hypermethylation blocks proper binding of transcription factors, resulting in dysregulation of normal gene expression with aging. Importantly, these patterns are not merely stochastic alterations in response to environmental stressors throughout aging. Indeed, the methylation status of a defined number of CpG sites is a highly accurate predictor of chronologic age in human tissues. This work reveals that DNA methylation status with aging outperforms previous standard biomarkers of aging, such as p16 expression levels and telomere length, and will be highly valuable in the near future to gauge effects of treatment aiming to ameliorate diseases of aging.

Brain disorders are among the greatest clinical challenges to understand and to treat. Most neurologic and psychiatric disorders result from complex dysregulation of numerous genes and pathways. In this interplay between underlying genetic predisposition and external environmental factors, aberrant epigenetic regulation is increasingly recognized as a potentially key modulator (Fig. 471-1).

More directly, however, several progressive neurodevelopmental disorders are caused by germline mutations in chromatin regulators. Mutations in methyl CpG binding protein 2 (MECP2), a protein important for binding to methylated DNA and contributing to gene repression, is the major cause of Rett syndrome. MeCP2 loss leads to overactive gene transcription in neurons and impaired presynaptic excitatory functions. Similarly, Kabuki syndrome, another progressive neurodevelopmental disorder, is caused by germline mutations in either the H3K4me1 histone methyltransferase, MLL4 (KMT2D), or the H3K27me3 demethylase, UTX (KDM6A). This disorder may derive from dysregulation of transcriptional enhancers, a major class of gene regulatory elements, as both MLL4 and UTX play a key role in activation of enhancers. Finally, the acetyltransferase CBP (CREBBP) also is important for gene enhancer function, and when mutated can lead to Rubinstein-Taybi syndrome, a cause of intellectual disability.

Beyond germline mutations, altered methylation dynamics can drive disorders of neural development and of neurodegeneration. Fragile X syndrome, characterized by learning disabilities and cognitive impairment, is caused by mutations in the FMR1 or FMR2 genes, or by hypermethylation of the transcriptional promoters regulating FMR1 or FMR2. Similarly, Prader-Willi syndrome and Angelman syndrome, neurodevelopmental conditions caused by abnormal imprinting of the paternal or maternal chromosomal region (15q11-13), respectively, are frequently caused by aberrant DNA methylation. Further, DNA hypomethylation is implicated in some neurodegenerative conditions. For instance, in Parkinson’s disease, several genes involved in pathogenesis are hypomethylated due to DNMT1 depletion, including the α-synuclein gene (SCNA). In Alzheimer’s disease (AD), DNA hypomethylation occurs at promoters of key pathogenic genes such as amyloid precursor protein (APP). Indeed, APP promoter methylation is responsive to environmental factors, including aging, a major risk factor for AD. Likewise, presenilin-1 (PSEN1) is implicated in AD and displays altered DNA methylation in response to variations in metabolic stimuli. Studies of Huntington’s disease (HD) have demonstrated DNA hypomethylation and decreased histone acetylation, in part due to altered function of the acetyl transferase CBP, leading to transcriptional dysregulation. Together these observations underscore epigenetic regulation as a crucial feature of neurodegeneration.

Additional gene regulatory proteins in the nervous system interact with and are regulated by chromatin modifiers. REST (repressor element 1–silencing transcription factor) is important in neuronal homeostasis through its ability to recruit chromatin regulatory enzymes, such as histone deacetylases and histone methyltransferases, and via its control over gene expression. REST levels increase with aging and serve a protective function in neurons against age-associated stressors and loss of cognitive function associated with AD. Similar to REST, brain-derived neurotrophic factor (BDNF), another important mediator of neural development and homeostasis, is implicated in a variety of neurologic and psychiatric disorders including depression, schizophrenia, bipolar disorder, and autism. Knockdown of BDNF in the dentate gyrus leads to depression-like behavior in mouse models, and BDNF mediates effects of antidepressant therapies. Chromatin pathways, including DNA methylation/MeCP2 and H3K27me3, play a key role in BDNF regulation as observed in brains from patients with schizophrenia.

Finally, addiction medicine is another frontier where epigenetics holds great promise to reveal connections between environmental exposure and phenotypes. Although still in its early stages in terms of mechanistic understanding, emerging evidence demonstrates disruption of epigenetic homeostasis as a consequence of addictive substances ranging from alcohol to cocaine. For example, the acetylation of regulatory elements in the FOSB gene by the histone acetyltransferase CBP is associated with behavioral effects of cocaine. Ethanol also induces histone acetylation and a decompacted chromatin structure.

Alterations in gene expression patterns are important determinants of immune-mediated disease, and in turn, epigenetics regulates immunity and inflammation (Fig. 471-1). Treatment with immune-stimulating agents such as lipopolysaccharide (LPS) and TNF-α activate expression of numerous inflammatory genes within hours, with precise gene pathways and activation kinetics determined by the cellular epigenetic state. HATs and HDACs are critical components of this response, coordinating with pro-inflammatory transcription factors such as AP-1 and NF-κB, to activate (HATs) and to repress (HDACs). For example, corticosteroids recruit HDAC2 to promoters of NF-κB-stimulated inflammatory genes to prevent activation during asthma treatment.

Type 1 IFN responses are exceptional examples of regulatory complexity governed by epigenetic control. In an unstimulated state, the H3K9 methyltransferases G9a (EHMT2) and EHMT1 suppress expression of IFN and IFN-induced genes. Upon induction of IFN-stimulated genes, STAT transcription factors recruit chromatin remodeling complexes, such as BAF (SMARCA4), and recruit HATs including p300, CBP, and GCN5 (KAT2A). In turn, chromatin remodeling and acetylation recruit chromatin binding proteins including the bromodomain protein, BRD4, which promotes transcriptional elongation and full activation.

Major regulators of adaptive immunity pathways are similarly epigenetically regulated. CD4+ and CD8+ T cells undergo extensive changes in histone modification profiles during differentiation to distinct subsets of effector T cells. For example, genes associated with effector T cell functions in CD8+ memory T cells (such as PRDM1, KLRG1, IFNG) display enrichment of H3K4me3 and low levels of H3K27me3 compared with those genes in naïve T cells. DNA methylation also plays an important regulatory role and may contribute to disease. For example, CD4+ T cells from individuals with rheumatoid arthritis (RA), systemic scleroderma, and latent autoimmune diabetes in adults display hypermethylation of the FOXP3 gene, which activates regulatory T cells that dampen immune responses. In addition, hypermethylation of the CTLA4 locus occurs in regulatory T cells from RA patients, impairing their immunosuppressive abilities.

Although mechanistic studies remain limited, there are numerous examples of epigenetic-based therapies associated with extensive effects on the immune system, underscoring the potential hope for eventual treatment of immune-related conditions. For example, the DNA methylation inhibitors azacitidine and decitabine have immunosuppressive effects possibly mediated by enhanced expression of FOXP3, which generally suppresses immune responses. HDAC inhibitors upregulate and downregulate immune genes, and they inhibit cytokine production in macrophages from patients with RA. Further, the HDAC inhibitors vorinostat and panobinostat inhibit primary B cell responses and antibody production in vitro and in vivo. Given these broad effects, it is not surprising that the HDAC inhibitor trichostatin A (TSA) has efficacy in various model systems for treatment of RA, systemic lupus erythematosus (SLE), asthma, acute kidney injury, sepsis-induced lung and cardiac damage, and acute pancreatitis. Similarly, BET inhibitors also display broad effects. They block antigen presentation and T and B cell activation, and thus have beneficial protective effects in a variety of inflammatory settings including autoimmunity, sepsis, atherosclerosis, psoriasis, periodontitis, and arthritis. Beyond these “broad-spectrum” epigenetic inhibitors, a specific inhibitor of the H3K27me3 demethylases KDM6A and KDM6B, GSK-J4, has anti-inflammatory activity, presumably by preventing loss of H3K27me3 repression over inflammatory genes.

Due to the enormity and complexity of the chromatin and epigenetics fields, and their reach into all areas of biology and medicine, it is not possible to cover such a broad scope in a single chapter. Thus, here we provide a concise snapshot highlighting key areas of development in medicine. We hope to have conveyed the tremendous excitement and promise that pervades the discipline. Indeed, given the exponential growth in uncovering the interface between the epigenome and epigenetic therapies with the environment and disease, there is little doubt that the coming years will bring important additions to this field.