Current Epigenetic Perspective on Diabetes - Who Regulates the Regulators?

Samuel T. Keating and Assam El-Osta

This article was originally published in Cellular & Molecular Medicine: Open Access on December 18 2015 under a Creative Commons Attribution 4.0 International License

Abstract

Intensified interest in the development of pharmacological compounds that manipulate gene control highlights the importance of understanding the key players and molecular events driving gene function. Epigenetic research too frequently focuses on a single chromatinized modification, often at a limited number of loci, and without context for other determinants of transcription. This perception is problematic because it implies an oversimplification of the vastly complex and multidimensional network of gene control. It overlooks the interactions of chromatin modifying enzymes and transcription factors, and seldom addresses the molecular events and signaling cues that influence the executive enzymatic machinery that regulate the epigenome. Here we discuss the connectivity and complexity of epigenetic regulation in the context of chromatin modifications and transcription factors. Using the example of the Set7 (also Set9 or Set7/9) methyltransferase, we describe recent observations that expand the understanding of chromatin biology.

Dynamic Chromatin

Site-selective DNA-binding proteins interact with genomic regulatory regions as a central mechanism to direct cell-specific transcriptional programs. Though initially considered a benign scaffold to package DNA, the nucleoprotein structures of chromatin are now recognized as integral to the accessibility of transcription factors to gene regulatory elements [1]. A relaxed euchromatin structure where the DNA template is loosely associated with histone proteins allows the transcriptional machinery of the cell to access and read the genetic code. This contrasts with transcriptionally repressive conformations where DNA is tightly associated with histones and buried within the chromatin. By influencing the chromatin-penetrating potential of transcription factors, the structural re-organization of chromatin underpins the selective activation and silencing of gene programs that give rise to an astounding array of cell types from a single source of genetic information. Just as importantly, chromatin dynamics underlie a cell’s ability to manage environmental variations with rapid changes in gene activity [2].

Active and silent regions of the genome are distinguished by small chemical modifications to histones and the DNA itself that markedly shape the chromatin architectures by recruiting or precluding other factors that remodel the histone-DNA complex. Because the chemical tags can regulate changes in gene expression independently of the underlying DNA sequence, they are called epigenetic (literally in addition to genetic). These molecular signatures have an enormous capacity to control the functional state of chromatin and transcriptional activity of genes encoded therein for both dividing and terminally differentiated cells. Chromatin modifications can be influenced by various environmental factors, hence epigenetic grammar contextualizes the language of the DNA code [3].

The importance of this environment-epigenome axis is emerging for many disease states [4-8]. This is exemplified by recent large clinical studies of diabetes. Indeed the pathogenesis of type 2 diabetes has a particularly strong association to environmental factors that are proposed to alter the chromatin landscape. Likewise, diabetes-associated perturbations in metabolism and hemodynamics are known to influence the development of vascular complications by epigenetic modifications. Moreover, some of these chromatinized changes are implicated in the phenomenon of metabolic memory in type 1 and type 2 diabetes, where antecedent periods of hyperglycemia drive persistent vascular complications many years after blood glucose control is achieved [9]. To this end, epigenetic profiling of circulating blood monocytes has revealed a persistent histone acetylation signature at genes implicated in diabetes complications that was closely associated with glycemic history in patients with type 1 diabetes [10].

While numerous enzymes and specific modifications have been described, defining the cell’s ability to sense the variety of diabetic signaling cues at the chromatin level remains an important challenge. Precisely how is information communicated to the orchestra of factors controlling the chromatin landscape? How does the epigenetic machinery interact with transcription factors to control gene expression? Who regulates the genome regulators?  Regulating chromatin accessibility and gene expression. Cytosine bases covalently modified by addition of methyl groups are recognized by methyl-CpG binding proteins that associate with chromatin remodeling factors to establish transcriptionally incompetent chromatin architectures. The nucleosome is comprised of approximately 147 bp of DNA encircling an octamer of 2 copies of each of the 4 core histone proteins H2A, H2B, H3, and H4. Post-translational modification (PTM) of nucleosomal histones further regulates transition between active, structurally open euchromatin and silent, condensed heterochromatin. Shown here is a summary of acetyl (purple) and methyl (red) modifications of specific amino acid residues on N-terminal tails of the heavily modified H3 and H4 histones. Click image to enlarge.   Oxidative stress alters the vascular chromatin landscape  The functional relationship between chromatin architecture and changes in gene expression conferred by chronic and prior hyperglycemia has proven to be an important avenue of investigation for explaining persistent vascular complications of diabetes. We previously described the critical role of H3 histones lysine 4 mono-methylation (H3K4m1) in the high glucose-mediated transcriptional activation of human endothelial NFκB-p65 (encoded by the RELA gene), a key pro-inflammatory transcription factor that regulates the expression of genes implicated in inflammation associated with vascular complications of diabetes [11,12]. Moreover this specific chromatin signature, written by the Set7 lysine methyltransferase, persisted for up to 6 days in normal glucose conditions, suggesting it could confer future cell memories. The clinical relevance of these seminal in vitro findings was recently validated in the peripheral blood mononuclear cells of a cohort of patients with type 2 diabetes [13]. Identification of the methyl writer in the chromatinization of glucose signaling cues raised a new question. How are changes in ambient glucose transmitted to Set7? Indeed Set7 is mobilized to the nucleus with increasing glucose concentration [14]. Mitochondrial overproduction of superoxide has long been known to initiate many hyperglycemia-induced mechanisms related to the pathogenesis of diabetic complications [15]. Accordingly, the up-regulation of RELA induced by transient hyperglycemia was abolished by overexpression of either uncoupling protein-1 (UCP-1) or manganese superoxide dismutase (MnSOD), both of which prevent hyperglycemia-induced superoxide accumulation [11]. Further, Paneni and colleagues identified epigenetic changes driving up-regulation of the mitochondrial adapter protein and critical mediator of oxidative stress p66Shc in vascular endothelial cells cultured in high glucose conditions [16]. The resulting superoxide production activates PKCβII, which in turn maintains elevated p66Shc levels, ultimately stimulating and sustaining epigenetic changes by enzymes such as Set7 [17]. While this mechanism has the capacity to explain the sustained legacy of hyperglycemia in diabetes, it raises further challenges in identifying how high glucose signals to the epigenetic machinery regulating p66Shc expression. Chromatin modifiers interact with transcription factors Large datasets reveal striking overlaps between transcription factor binding sites and chromatin modifications [18]. Coregulatory interactions between transcription factors and chromatin-modifying enzymes may at least partly account for this co-localization [19]. Set7 was shown to be co-recruited with TAF10 to activating gene promoters [20]. In fact, along with its role in methylating histones, Set7 interacts with numerous transcription factors across various cell types, often promoting methylation reactions on regulatory lysine residues at the surface of the transcription factor [21]. Our recent characterization of human vascular endothelial cells depleted of Set7 revealed widespread changes in gene expression across numerous pathways associated with vascular function that were only partly explained by changes in H3K4m1 at promoters and distal enhancer regions [22]. By intersecting the transcriptome profile with publicly available datasets, we identified strong associations between deregulated genes and six transcription factors previously described as Set7 methylation substrates: NFκB, STAT3, IRF1, p53, ERα, and TAF7 [21]. In addition, many deregulated genes were associated with numerous transcription factors not previously connected with Set7 function. By applying a consensus formula derived from Set7 substrates previously used to accurately predict several biochemically validated in vivo non-histone substrates [23], we predicted that Set7 post-translationally regulates transcription factors associated with vascular endothelial expression through the presence of Set7 amino acid methylation motifs. Amino hydrophobicity analysis indicated most predicted sites to be accessible to post-translational modification. Further, in vitro peptide methylation assays suggest that Set7 can indeed modify a predicted site on the STAT1 transcription factor, demonstrating the predictive value of our method to identify novel candidate substrates to analyze in vivo. Further characterization of putative substrates identified in these studies has the capacity to identify not only functional modulation of transcription factors, but also substrate-driven co-recruitment of the enzyme to specific promoters to potentiate H3K4m1 enrichment. Like the regulatory function of acetylation, lysine methylation has emerged as an important post-translational modification for modulation of transcription factors [24]. Our novel method of mapping transcriptional changes to transcription factors for the identification of putative substrates with strong associations to functional changes is applicable to substrate prediction for other broad-substrate histone modifiers [22]. Both the histone and non-histone-modifying activities of epigenetic enzymes are important considerations for future strategies of pharmacological targeting in the clinic. Chromatin modifiers regulate each other A growing body of evidence points to post-translational modifications as regulators of chromatin modifying enzymes themselves [2]. The prevalence of these modifications suggests a highly ordered and dynamic network of components capable of writing, reading, and erasing modifications at both the chromatin template as well as each other. Counted among the expanding catalogue of experimentally validated methylation substrates of Set7 are SUV39h1 and DNMT1 - enzymes that methylate histones (at a distinct site to Set7) and DNA respectively. Methylation at lysines 105 and 123 by Set7 impairs the repressive histone modifying capacity of SUV39h1 [25], whereas the stability of DNMT1 is regulated by Set7-mediated lysine methylation [26]. Similarly Set7 also methylates multiple lysines on the p300/CBP-Associated Factor (PCAF) histone acetyltransferase [27]. A single epigenetic enzyme therefore has the ability to control many chromatin modifications. Mapping the inter-enzyme modification network of epigenetic regulators has an enormous capacity to increase our understanding of gene regulation and further raises important considerations for therapeutic strategies aimed at editing the epigenome. Conclusion Immense interest surrounds efforts to modulate gene expression, to restore the activity of silenced genes or attenuate unscheduled gene expression. However, the complexity of gene regulation is vast and researchers are only beginning to gain an appreciation for the biochemical determinants and genome-wide interconnectivity of epigenetic enzymes and transcription factors. Using the example of Set7, we have described recent observations that expand the understanding of chromatin biology beyond the immediate histone methyl-writing event. Indeed this is just a scratch on the surface and further studies are required to completely understand this important enzyme. Similar questions of biochemistry and interactivity remain for many other classes of chromatin modifiers considered useful in the clinic, including methylases, demethylases, acetylases and deacetylases, as well as protein components responsible for reading the chromatin mark such as bromodomains. While chromatin modifications can be informative of gene regulation at specific loci, the challenge of understanding their cell-specific function remains unmet. Myeloid-specific genetic deletion of histone deacetylase 3 is associated with stable atherosclerotic plaques [28], whereas deletion of the same enzyme in endothelial cells enhances atherosclerosis in mice [29]. More recently emerged concepts could offer further insight into epigenomic regulation. For example several intermediates of cellular metabolism are critical substrates for chromatin modifying enzymes. Fluctuating levels of these metabolites could therefore signal for continual adjustment and contextualization of gene expression (recently reviewed [2]). In addition long noncoding RNA molecules could play a role in the localization of chromatin signatures. By simultaneously recruiting two different histone modifiers to the chromatin – one a writer and the other an eraser of histone methylation – the HOTAIR long noncoding RNA facilitates the coordinated addition of a repressive modification and removal of an activating one to silence specific genes [30]. Technological and scientific advances have rapidly expanded the field of epigenetics to the point where chromatin modifiers are seriously considered as therapeutic targets for numerous diseases. The challenge for the next decade is to develop a comprehensive understanding of the biochemical and molecular events controlling the genome’s regulators. References Voss TC, Hager GL (2014) Dynamic regulation of transcriptional states by chromatin and transcription factors. Nat Rev Genet 15:69-81 Keating ST, El-Osta A (2015) Epigenetics and metabolism. Circ Res 116:715-736 Turner BM (2007) Defining an epigenetic code. Nature Cell Biology 9:2-6 Brown TA et al., (2015) Alterations in DNA methylation corresponding with lung inflammation and as a biomarker for disease development after MWCNT exposure. Nanotoxicology 16:1-9 Silver MJ et al., (2015) Independent genomewide screens identify the tumor suppressor VTRNA2-1 as a human epiallele responsive to periconceptional environment. Genome Biol 16:118 Ma L et al., (2015) Histone Methylation in Nickel-Smelting Industrial Workers. PLoS One 10:e0140339 Saeed S et al., (2014) Epigenetic programming of monocyte-to-macrophage differentiation and trained innate immunity. Science 345:1251086 Nilsson E et al., (2015) Epigenetic alterations in human liver from subjects with type 2 diabetes in parallel with reduced folate levels. J Clin Endocrinol Metab 100:E1491-501 Keating ST, El-Osta A (2013) Glycemic memories and the epigenetic component of diabetic nephropathy. Curr Diab Rep 13:574-581 Miao F et al., (2014) Evaluating the role of epigenetic histone modifications in the metabolic memory of type 1 diabetes. Diabetes 63:1748-1762 El-Osta A et al., (2008) Transient high glucose causes persistent epigenetic changes and altered gene expression during subsequent normoglycemia. J Exp Med 205:2409-2417 Brasacchio D et al., (2009) Hyperglycemia induces a dynamic cooperativity of histone methylase and demethylase enzymes associated with gene-activating epigenetic marks that coexist on the lysine tail. Diabetes 58:1229-1236 Paneni F et al., (2015) Adverse epigenetic signatures by histone methyltransferase Set7 contribute to vascular dysfunction in patients with type 2 diabetes mellitus. Circ Cardiovasc Genet 8:150-158 Okabe J et al., (2012) Distinguishing hyperglycemic changes by Set7 in vascular endothelial cells. Circ Res 110:1067-1076 Brownlee M (2005) The pathobiology of diabetic complications: a unifying mechanism. Diabetes 54:1615-1625 Paneni F et al., (2012) Gene silencing of the mitochondrial adaptor p66(Shc) suppresses vascular hyperglycemic memory in diabetes. Circ Res 111:278-289 Paneni F et al., (2013) SIRT1, p66(Shc), and Set7/9 in vascular hyperglycemic memory: bringing all the strands together. Diabetes 62:1800-1807 The ENCODE Project Consortium (2012) An integrated encyclopedia of DNA elements in the human genome. Nature 489:57-74 Benveniste D et al., (2014) Transcription factor binding predicts histone modifications in human cell lines. Proc Natl Acad Sci USA 111:13367-13372 Kouskouti A et al., (2004) Gene-specific modulation of TAF10 function by SET9-mediated methylation. Mol Cell 14:175-182 Keating ST, El-Osta A (2013) Transcriptional regulation by the Set7 lysine methyltransferase. Epigenetics 8:361-372 Keating ST et al., (2014) Deep sequencing reveals novel Set7 networks. Cell Mol Life Sci 71:4471-4486 Dhayalan A et al., (2011) Specificity analysis-based identification of new methylation targets of the SET7/9 protein lysine methyltransferase. Chem Biol 18:111-120 Carr SM et al., (2015) Post-translational control of transcription factors: methylation ranks highly. FEBS J 282: 4450-65 Wang D et al., (2013) Methylation of SUV39H1 by SET7/9 results in heterochromatin relaxation and genome instability. Proc Natl Acad Sci USA 110:5516-5521 Esteve PO et al., (2009) Regulation of DNMT1 stability through SET7-mediated lysine methylation in mammalian cells. Proc Natl Acad Sci USA 106:5076-5081 Masatsugu T, Yamamoto K (2009) Multiple lysine methylation of PCAF by Set9 methyltransferase. Biochem Biophys Res Commun 381:22-26 Swirski FK et al., (2007) Ly-6Chi monocytes dominate hypercholesterolemia-associated monocytosis and give rise to macrophages in atheromata. J Clin Invest 117:195-205 Zampetaki A et al., (2010) Histone deacetylase 3 is critical in endothelial survival and atherosclerosis development in response to disturbed flow. Circulation 121:132-142 Tsai MC et al., (2010) Long noncoding RNA as modular scaffold of histone modification complexes. Science 329:689-693.

The chromatin code

A dynamic landscape

The chromatin polymer is a dynamic assembly of DNA and various nuclear proteins that spatially regulates eukaryotic gene transcription by structural adaptation and genome compartmentalisation. 

Gene expression is primarily regulated by chromatin accessibility. The repeating unit of chromatin is the nucleosome, comprised of approximately 147 base pairs of DNA encircling an octamer of two copies of each of the four core histone proteins: H2A, H2B, H3, and H4. A relaxed chromatin structure where DNA is loosely associated with nucleosomes (euchromatin) is permissive to the transcriptional machinery, allowing the synthesis of RNA. By contrast, genes are silenced when DNA is tightly wrapped around histones and buried within the chromatin (heterochromatin). It is by selective activation and silencing of different genes that more than 200 distinct cell types of the human body are derived from a single source of genetic information. Furthermore, chromatin dynamics provides the adaptive agency for cells to manage environmental variations with rapid changes in gene expression.

Small chemical modifications that decorate the chromatin landscape distinguish active and silent genomic regions. Modifications frequently occur at specific amino acid residues on histones as well as the DNA itself to collectively comprise the epigenome. 

By several distinct mechanisms, the chemical signatures of chromatin architectures are fundamental to gene expression. Individual modifications can recruit functional complexes responsible for the structural reorganisation of chromatin. Other modifications directly affect the stability of the histone-DNA complex to promote an open conformation. These gene-regulatory effects in combination with substantial functional interplay between histone modifications has lead to the proposition of a chromatin code that greatly extends the information potential of the genetic code by shaping transcriptional competency. 

The epigenome is diverse

Methylation of DNA predominantly at cytosine nucleotides adjacent to guanine residues (CpG) at gene promoters is predominantly associated with gene suppression and corresponds most closely to the etymological interpretation of epigenetics. Regulatory proteins mechanistically interpret this prevalent epigenetic modification by recruiting chromatin-remodeling complexes to establish transcriptionally repressive chromatin. Recent discovery of the Ten Eleven Translocation family of dioxyegnases that oxidise methyl-cytosine to generate hydroxymethyl-cytosine have inspired strong interest in transitions between methylated and unmethylated DNA in mammalian cells.

Equally important are the wealth of chemical groups dynamically written to and erased from the protruding N-terminal tails of histones by specialised enzymes and complexes. Here the methyl modification also occupies a prominent role in gene regulation when assigned to lysine (K) and arginine (R) residues. Moreover the effect on transcription is dependent on the site of modification. For example, methylation at lysines 9 and 36 of H3 histones associates with gene repression and activation respectively. Furthermore, diverse degrees of methylation (lysine residues can be mono-, di-, or tri-methylated, and arginine residues mono-methylated and asymmetrically or symmetrically di-methylated) are differentially distributed across chromatin and ascribed distinct functional roles in gene regulation. A prominent example is the range of methyl modifications occurring on lysine 4 of H3 histones. The tri-methylated form of this modification punctuates the start of actively transcribed genes. On the other hand, di-methylation is dispersed throughout gene bodies, and mono-methylation is frequently observed at both proximal and distal regulatory regions of active genes. Contradicting the activities of histone methylases are enzymes responsible for the removal of methyl groups, such as JmjC-domain-containing proteins and LSD1.   

In contrast, site-specific histone lysine acetylation is ubiquitously associated with open chromatin and transcriptional activation. Regulated by the opposing activities of histone acetylases (HATs) and deacetylases (HDACs), acetyl groups are thought to disrupt the charge states of histone tails, altering their contact with the nucleosome as well as reducing their affinity for DNA. Additionally, acetylation of lysines 9 and 27 of H3 histones directly competes with the transcriptionally repressive methylation of these residues.

Other important histone modifications include phosphorylation, ubiquitination, SUMOylation, ADP-ribosylation, and the recently emerged O-linked addition of N-acetylglucosamine (O-GlcNAcylation).

Gene function is primarily regulated by chromatin accessibility (click to enlarge). Shown here is a summary of acetyl (purple) and methyl (red) modifications of specific arginine (R) and lysine (K) residues on N-terminal tails of the heavily modified H3 and H4 histones, as well as DNA methylation at active and silent regions of chromatin. Click image to enlarge

Who regulates the regulators?

The enzymes responsible for the transfer and removal of histone modifications demonstrate strong specificity toward particular amino acid positions within histone tails. Research has characterised a substantial catalogue of enzymes that catalyse more than 100 histone tail marks. Indeed, several biological factors are known to influence the expression of epigenetic modifiers. However a number of questions remain unanswered regarding the precise mechanism underlying their enzymatic function. Specifically, how are chromatin-modifying reactions regulated?

A current topic in the field is the effects of metabolism on the epigenome. Several components of the epigenetic machinery require intermediates of cellular metabolism for enzymatic function. Dietary factors such as folate and B vitamins influence cellular concentrations of S-adenosyl methionine, the universal methyl-donor to methylation reactions. Similarly, acetyl-CoA generated from glucose and fatty acid metabolism is the essential acetyl group donor to lysine acetylation reactions. Furthermore, some DNA and histone demethylase enzymes utilise specific intermediates of the citric acid cycle.

Secondly, post-translational modifications are not restricted to the chromatin. In fact, the same enzymes that write and erase epigenetic marks modify a diverse array of non-histone proteins. These include not only transcription factors, which can independently influence gene expression, but also the activities of other histone modifiers. For example, acetylation enhances the enzymatic functions of the P/CAF, p300, and MYST acetylases, as well as lowering the methylase activity of SUV39H1.  On the other hand, SUV39H1 is activated, and the DNA methylase DNMT1 destabilized by lysine methylation. Characterisation of this network of modifications controlling enzyme function and consequent epigenetic chromatin modulation holds immense potential to further our understanding of gene regulation.

Finally, an important role for non-coding RNA molecules in the establishment of cell-type and gene-specific chromatin modification patterns has recently emerged. Advances in nucleic acid sequencing technologies have revealed that while approximately 90% of the human genome is transcribed, only 1-2% of RNA transcripts encode proteins. Non-coding RNA vary in length and function, with short transcripts such as mircoRNAs playing numerous regulatory roles in gene expression primarily at the mRNA level. Importantly, long non-coding RNAs (lncRNAs) interact with epigenetic enzymes to direct their chromatin binding. For example, the HOTAIR lncRNA simultaneously recruits the Polycomb Repressive Complex 2 and the LSD1 demethylase for coordinated methylation of lysine 27 and demethylation of lysine 4 on H3 histones. Future research is anticipated to uncover many scaffolding and tethering roles for lncRNAs in the specific localisation of chromatin changes.

Finally, an important role for non-coding RNA molecules in the establishment of cell-type and gene-specific chromatin modification patterns has recently emerged. Advances in nucleic acid sequencing technologies have revealed that while approximately 90% of the human genome is transcribed, only 1-2% of RNA transcripts encode proteins. Non-coding RNA vary in length and function, with short transcripts such as mircoRNAs playing numerous regulatory roles in gene expression primarily at the mRNA level. Importantly, long non-coding RNAs (lncRNAs) interact with epigenetic enzymes to direct their chromatin binding. For example, the HOTAIR lncRNA simultaneously recruits the Polycomb Repressive Complex 2 and the LSD1 demethylase for coordinated methylation of lysine 27 and demethylation of lysine 4 on H3 histones. Future research is anticipated to uncover many scaffolding and tethering roles for lncRNAs in the specific localisation of chromatin changes.

Mapping the chromatin landscape

Whether these observations constitute a true code is often questioned.  The existence of a strict chromatin code implies that definite patterns of post-translational histone modifications instruct a rigid functional outcome.  Unlike the causal nature of the genetic code, it is more likely that combinatory patterns of histone modifications create a biased chromatin landscape that generally favours a particular transcriptional outcome. Nonetheless, genetic knock-out animals and similar approaches in cells demonstrate the necessity of epigenetic regulation.

Chromatin modifications are increasingly studied in development and disease and considerable interest surrounds the pharmacological targeting of epigenetic enzymes for therapy. Particularly HDAC inhibitors are rigorously investigated for their clinical use in the treatment of human disease such as cancer and heart disease. Clearly, the thorough characterisation of the epigenome holds immense promise for the clinic. 

Increased availability and use of massive parallel sequencing has allowed the epigenetic analysis of various cell types and biological contexts. Accelerated generation of epigenomic datasets has driven the accumulation of large repositories of data.This shift to systems-level perspective signifies a fundamental change in the way cell biology is investigated, rapidly propelling our modern knowledge of medicine and biology. The generation of cell-specific epigenomic maps and transcriptome profiles is fundamental to a truly comprehensive understanding of gene regulation.

References

1. Jenuwein T. & Allis CD. (2001) Translating the histone code. Science 293(5532): 1074-80

2. Delatte B. et al., (2014) Playing TETris with DNA modifications. EMBO 33(11): 1198-211

3. Bernstein BE. et al., (2005) Genomic maps and comparative analysis of histone modifications in human and mouse. Cell 120(2): 169-81

4. Heintzman ND. et al., (2007) Distinct and predictive chromatin signatures of transcriptional promoters and enhancers in the human genome. Nat Genet 39(3): 311-8

5. Klose RJ. et al., (2006) JmjC-domain-containing proteins and histone demethylation. Nat Rev Genet 7(9): 715-27

6. Verdone L. et al., (2005) Role of histone acetylation in the control of gene expression. Biochem Cell Biol 83(3): 344-53

7. Hanover JA. (2010) Epigenetics gets sweeter: O-GlcNAc joins the "histone code". Chem Biol 17(12): 1272-4

8. Donohoe DR & Bultman SJ. (2012) Metaboloepigenetics: interrelationships between energy metabolism and epigenetic control of gene expression. J Cell Physiol 227(9): 3169-77

9. Santos-Rosa H. et al., (2003) Mechanisms of P/CAF auto-acetylation. Nucleic Acids Res 31(15): 4285-92

10. Thompson PR. et al., (2004) Regulation of the p300 HAT domain via a novel activation loop. Nat Struct Mol Biol 11(4): 308-15

11. Yuan H. et al., (2012) Myst protein acetyltransferase activity requires active site lysine autoacetylation. EMBO 31(1): 58-7012. Vaquero A. et al., (2007) Sirt1 regulates the histone methyl-transferase suv39h1 during heterochromatin formation. Nature 450(7168): 440-4

13. Wang D. et al., (2013) Methylation of SUV39H1 by SET7/9 results in heterochromatin relaxation and genome instability. PNAS 110(14): 5516-21

14. Esteve PO, et al., (2009) Regulation of DNMT1 stability through SET7-mediated lysine methylation in mammalian cells. PNAS 106(13): 5076-81

15. Mercer TR, et al., (2009) Long non-coding RNAs: insights into functions. Nat Rev Genet 10(3): 155-9

16. Clark MB, et al., (2011) The reality of pervasive transcription. PLoS Biol 9(7)

17. Tsai MC, et al., (2010) Long noncoding RNA as modular scaffold of histone modification complexes. Science 329(5992): 689-93

18. Mathiyalagan P, et al., (2014) Interplay of chromatin modifications and non-coding RNAs in the heart. Epigenetics 9(1): 101-12