Epigenetic Regulation of Cardiovascular Physiology: Role of DNA Methylation and Histone Modification in Heart Failure Pathogenesis and Therapy

Cardiovascular diseases collectively account for roughly 17.9 million deaths each year — approximately one-third of all recorded global mortality — a burden that places them in a category of public health urgency unmatched by any other disease family. Within this spectrum, heart failure commands particular clinical attention: it is simultaneously the shared endpoint toward which nearly all structural and functional cardiac disorders converge, and a distinct syndrome whose prevalence continues to climb as ageing populations accumulate the interacting comorbidities — systemic hypertension, type 2 diabetes, obesity, coronary atherosclerosis — that progressively erode the reserve capacity of the myocardium. Contemporary guideline-directed therapy, built around inhibition of the renin–angiotensin–aldosterone axis, beta-adrenergic receptor blockade, mineralocorticoid receptor antagonism, and sodium–glucose cotransporter-2 inhibition, has measurably improved survival and reduced hospitalisation in patients with reduced ejection fraction. Yet absolute mortality remains high, and for the roughly half of HF patients with preserved ejection fraction, no pharmacological strategy has yet convincingly altered the natural history of the disease. These limitations reflect the still-incomplete conceptualisation of HF as a disease rooted in transcriptional and metabolic reprogramming, not simply one of pressure overload and neurohormonal excess.

The science of epigenetics — the study of heritable, reversible alterations in gene expression that do not involve modification of the primary DNA sequence — has substantially repositioned our understanding of how the cardiac genome responds to the haemodynamic and biochemical stresses of disease. Epigenetic mechanisms serve as the molecular interface between environmental stimuli and gene regulatory responses, converting haemodynamic and metabolic signals into durable changes in chromatin accessibility and transcriptional output that persist across cell generations and reinforce pathological phenotypes. The three principal mechanisms — DNA methylation, histone post-translational modifications, and non-coding RNA-mediated regulatory circuits — function cooperatively to establish and maintain chromatin states that determine whether individual cardiac gene loci are accessible to transcription or sequestered in compact, transcriptionally inert chromatin.

What distinguishes epigenetic perturbations from genetic mutations, and what most strongly motivates therapeutic interest in this field, is their inherent reversibility. DNA methylation marks are installed and erased by enzymatic systems amenable to pharmacological modulation; histone acetylation is governed by the competing activities of acetyltransferases and deacetylases; and histone methylation is balanced between dedicated methyltransferases and demethylases. This reversibility transforms the therapeutic objective: rather than correcting permanent genomic lesions — a challenge that requires gene-editing technologies still far from routine clinical deployment — modulating epigenetic dysregulation in the failing heart may restore near-physiological gene expression patterns using agents already in clinical use for haematological and oncological indications.

This review examines the molecular foundations of epigenetic regulation in cardiovascular physiology, characterises the specific epigenetic disruptions that define failing myocardium at the levels of DNA methylation and histone modification, explores the functional consequences of these disruptions for cardiomyocyte biology, and evaluates the translational landscape encompassing epigenetic biomarkers and emerging therapeutic candidates. The scope extends from core principles of chromatin biology through to the most advanced clinical data on epigenetic drug candidates, with sustained attention to the mechanistic logic linking defined chromatin marks to the cellular hallmarks of HF — hypertrophy, fibrosis, apoptosis, metabolic maladaptation, and electrophysiological remodelling.

Figure 1. Overview of epigenetic mechanisms in cardiovascular physiology. The three principal regulatory layers — DNA methylation, histone post-translational modifications, and non-coding RNAs — converge on the chromatin template to control cardiac gene expression. Dysregulation of each layer contributes to the pathological transcriptional programmes underlying heart failure, cardiac hypertrophy, and myocardial fibrosis. Pharmacological targets at each node are indicated.

2. Fundamentals of Epigenetic Regulation

2.1 Chromatin Architecture and Transcriptional Accessibility

The human genome, encompassing approximately 3.2 billion base pairs, must be compacted roughly 10,000-fold to reside within a nucleus measuring about six micrometres in diameter. This feat is accomplished through hierarchical packaging around nucleosomes — the fundamental repeating units of chromatin, each consisting of 147 base pairs of DNA wound 1.65 turns around an octameric histone scaffold (two copies each of H2A, H2B, H3, and H4). The resulting nucleosomal arrays fold into higher-order chromatin fibres, topologically associating domains (TADs), and ultimately discrete chromosome territories. The accessibility of any given genomic locus to transcription factors, RNA polymerase, and the epigenetic machinery is therefore determined both locally — by the modification state of adjacent nucleosomal histone tails — and globally, by the three-dimensional spatial arrangement of chromatin domains.

Two broad chromatin states govern transcriptional competence. Euchromatin — characterised by loosely arranged, nucleosome-depleted chromatin enriched in acetylated histones and H3K4me3 — supports active transcription and is predominantly situated in the nuclear interior. Heterochromatin — a tightly compacted, transcriptionally suppressed state enriched in H3K9me3 and H3K27me3 — is generally located at the nuclear periphery and encompasses centromeric, telomeric, and gene-sparse chromosomal regions. Beyond these constitutive states, facultative heterochromatin — reversibly silenced chromatin whose repression is context-sensitive — plays an important role in developmental gene regulation, including the suppression of foetal cardiac gene programmes in the adult heart that are pathologically re-engaged during HF.

2.2 The Epigenetic Code: Writers, Readers, and Erasers

Epigenetic marks are deposited by enzymes termed writers, recognised by reader proteins bearing specialised binding domains, and removed by eraser enzymes. DNA methyltransferases (DNMTs) write methylation marks on cytosine residues; ten-eleven translocation (TET) enzymes progressively oxidise 5-methylcytosine (5-mC) through a chain of intermediates as part of the active demethylation pathway. Histone acetyltransferases (HATs) append acetyl groups to specific lysine residues on histone tails; histone deacetylases (HDACs) remove them. Lysine methyltransferases install methyl groups; lysine demethylases strip them away. Additional modifications — phosphorylation, ubiquitination, sumoylation, and ADP-ribosylation — expand the biochemical vocabulary of chromatin regulation. Reader proteins containing bromodomains (which bind acetylated lysines), chromodomains (which recognise methylated lysines), or methyl-CpG-binding domains (MBDs) translate the modification state of chromatin into functional outputs by recruiting transcriptional activators, repressors, or chromatin-remodelling complexes.

The combinatorial nature of histone modification — different marks at distinct positions acting in concert — generates regulatory logic of considerable complexity, partially decoded by chromatin immunoprecipitation sequencing (ChIP-seq) and single-cell approaches. Specific combinations of marks, rather than isolated modifications, most reliably predict transcriptional states: active enhancers characteristically carry H3K4me1 and H3K27ac; active promoters are distinguished by H3K4me3 and H3K27ac; Polycomb-repressed genes are marked by H3K27me3; and transcriptional elongation is accompanied by H3K36me3 along gene bodies. Identifying which combinations are disrupted in the failing heart, and through which enzymatic mechanism, provides the rational foundation for epigenetic therapeutic targeting.

2.3 Non-Coding RNAs as Epigenetic Regulators

Beyond DNA methylation and histone modification, the cardiac genome is regulated by a rich landscape of non-coding RNAs (ncRNAs). MicroRNAs (miRNAs), typically 20–24 nucleotides in length, suppress gene expression post-transcriptionally by directing the RISC complex to partially complementary sequences within the 3′ untranslated regions of target mRNAs, triggering their destabilisation or translational arrest. Long non-coding RNAs (lncRNAs), exceeding 200 nucleotides, engage diverse regulatory mechanisms including recruitment of PRC2 to specific genomic loci to deposit H3K27me3, scaffolding of enhancer–promoter contacts, and sequestration of miRNAs through competing endogenous RNA (ceRNA) activity. Circular RNAs (circRNAs) — covalently closed molecules generated by back-splicing — have proven to be particularly stable and functionally significant in cardiac biology, with several demonstrated to act as miRNA decoys that modulate hypertrophic and fibrotic cascades. PIWI-interacting RNAs (piRNAs), though principally characterised in the germline, are also expressed in cardiomyocytes and contribute to transposable element silencing.

2.4 Developmental Epigenetics and the Foetal Gene Programme

A defining feature of pathological cardiac hypertrophy is the re-expression of a foetal gene programme — a transcriptional shift in which genes normally silenced in the adult heart after birth, including β-myosin heavy chain (β-MHC/MYH7), atrial natriuretic peptide (ANP/NPPA), brain natriuretic peptide (BNP/NPPB), and skeletal muscle α-actin (ACTA1), are restored to active expression. This developmental regression is not the product of DNA mutation but of epigenetic reprogramming: the Polycomb-mediated H3K27me3 marks that silence these foetal genes in the healthy adult heart are erased under pathological stress, and H3K4me3 marks that enable active transcription are re-established at their promoters. The lncRNA MHRT normally shields the MYH6 locus from pathological epigenetic rewriting by occupying the helicase domain of the chromatin remodeller Brg1; its downregulation under pressure overload permits Brg1 to silence α-MHC while activating β-MHC, contributing directly to contractile dysfunction. Appreciating this epigenetic developmental framework is therefore inseparable from understanding the molecular pathology of acquired heart disease.

3. DNA Methylation in Cardiovascular Physiology

3.1 CpG Methylation: Biochemistry and Distribution

In the mammalian genome, DNA methylation occurs predominantly at cytosines immediately preceding a guanine — CpG dinucleotides — where a methyl group is appended at the fifth carbon position of cytosine to yield 5-methylcytosine (5-mC). Non-CpG methylation at CpA, CpT, and CpC contexts is also detectable in cardiac tissue and has been reported to influence cardiomyocyte gene expression, though its functional significance is less thoroughly characterised. Approximately 70–80% of all CpG dinucleotides in the human cardiac genome are methylated, with the unmethylated fraction concentrated in CpG-dense stretches — CpG islands (CGIs) — that are predominantly associated with the promoters of housekeeping and tissue-specific genes. Methylation of CGI promoters enforces stable, heritable transcriptional silencing, maintaining cell-type identity and preventing inappropriate gene expression across cell divisions.

Three principal DNMT enzymes catalyse methyl group transfer from S-adenosyl methionine (SAM) to cytosine. DNMT1 — the maintenance methyltransferase — preferentially methylates hemi-methylated DNA arising from replication, faithfully copying the parental methylation pattern to daughter strands and thereby perpetuating the methylation landscape through successive cell divisions. DNMT3A and DNMT3B — de novo methyltransferases — establish new methylation marks during development and in response to environmental cues, acting on previously unmethylated CpG substrates. DNMT3L, catalytically inert in itself, serves as an essential cofactor enhancing DNMT3A/3B activity. The stoichiometry and genomic targeting of these enzymes, shaped by interactions with sequence-specific transcription factors and histone modification reader domains, collectively determine the cardiac methylation map.

3.2 TET Enzymes and Active Demethylation

For decades DNA methylation was viewed as an essentially permanent epigenetic mark, reversed only through passive dilution during replication in the absence of maintenance methyltransferase activity. The discovery of the TET dioxygenase family — TET1, TET2, and TET3 — transformed this view. TET enzymes catalyse the sequential oxidation of 5-mC to 5-hydroxymethylcytosine (5-hmC), 5-formylcytosine (5-fC), and 5-carboxylcytosine (5-caC) via molecular oxygen and α-ketoglutarate as co-substrates in an iron(II)-dependent reaction. 5-fC and 5-caC are subsequently recognised and excised by thymine DNA glycosylase (TDG), followed by base excision repair that restores unmodified cytosine — completing the active demethylation cycle. This pathway is now recognised as integral to the dynamic adjustment of cardiac gene expression in response to physiological and pathological stimuli.

5-Hydroxymethylcytosine warrants separate discussion not merely as a demethylation intermediate, but as a potentially independent functional epigenetic mark. Unlike 5-mC, 5-hmC is not efficiently recognised by the MBD proteins and repressor complexes that execute methylation-dependent silencing. Work on human cardiac tissue has demonstrated that 5-hmC is enriched at actively transcribed cardiac genes — particularly those encoding sarcomeric proteins and metabolic enzymes — and that this enrichment is diminished in failing hearts from patients with dilated cardiomyopathy, implying that impaired TET activity contributes to transcriptional suppression of cardiac function genes in HF by reducing 5-hmC at gene bodies.

3.3 Methylation Dynamics During Cardiac Development

The cardiac methylome undergoes extensive, temporally orchestrated remodelling throughout embryonic development. Genome-wide bisulphite sequencing studies of murine and human cardiac tissue across developmental stages reveal successive waves of de novo methylation and demethylation corresponding to distinct developmental transitions — from cardiac progenitor to foetal cardiomyocyte to neonatal and ultimately to adult myocardium. The transition from foetal to adult cardiac gene expression — marked by the shift from β-MHC to α-MHC isoforms, from glycolytic to oxidative metabolism, and from proliferative to terminally differentiated phenotype — is accompanied by progressive methylation of foetal gene promoters and demethylation of adult-specific gene loci. This developmental programme is disturbed in pathological hypertrophy, where foetal gene loci are demethylated and adult-specific loci hypermethylated, recapitulating aspects of the foetal transcriptome in the physiologically stressed adult heart.

Haemodynamic stimuli directly modulate DNMT activity and downstream methylation patterns. Mechanical stretch of cardiomyocytes and cardiac fibroblasts activates angiotensin II signalling, which induces DNMT3A nuclear translocation and de novo methylation at the promoters of cardioprotective transcription factors such as Krüppel-like factor 4 (KLF4), effectively suppressing anti-hypertrophic signalling. Conversely, physiological cardiac hypertrophy induced by regular exercise training is associated with distinct methylation patterns at metabolic and angiogenic gene loci that diverge substantially from those of pathological hypertrophy — providing epigenomic evidence for the well-recognised functional distinction between adaptive and maladaptive cardiac growth.

Figure 2. DNA methylation dynamics in cardiac physiology and heart failure pathogenesis. Left panel: contrasting CpG island methylation patterns in the normal heart (balanced DNMT/TET activity sustaining cardioprotective gene expression) versus failing myocardium (DNMT3A/3B upregulation and TET downregulation producing hypermethylation at RASAL1, SERCA2a, and eNOS promoters). Right panel: the DNMT–TET balance governing methylation homeostasis and its disruption in heart failure.

4. DNA Methylation Alterations in Heart Failure

4.1 Genome-Wide Methylation Profiling in Failing Myocardium

The advent of whole-genome bisulphite sequencing (WGBS) and reduced representation bisulphite sequencing (RRBS) has made it possible to map the cardiac methylome at single-nucleotide resolution, revealing that HF is characterised by broad methylation landscape disruption rather than gene-isolated changes. In a foundational study utilising RRBS on cardiac tissue from patients with end-stage dilated cardiomyopathy compared with non-failing donor hearts, Haas and colleagues identified over 4,000 differentially methylated regions (DMRs), with approximately two-thirds showing hypermethylation and one-third hypomethylation relative to controls. The hypermethylated loci were significantly enriched for genes involved in calcium handling, mitochondrial fatty acid oxidation, and contractile protein regulation — functional categories directly relevant to the contractile and metabolic deficits that define HF. Hypomethylated loci were enriched for inflammatory cytokine genes, extracellular matrix remodelling components, and elements of the foetal gene programme, consistent with the transcriptional activation of these pathways in failing myocardium.

A consistently replicated finding across methylome studies is hypermethylation of the RASAL1 promoter in failing hearts. RASAL1 encodes a GTPase-activating protein that restrains RAS pathway signalling by accelerating Ras-GTP hydrolysis to Ras-GDP; its silencing through promoter methylation therefore sustains constitutive RAS/MAPK activation, promoting cardiomyocyte hypertrophy and myofibroblast activation. Experimental restoration of RASAL1 expression in pressure-overloaded murine hearts through DNMT inhibitor treatment attenuated hypertrophic remodelling and improved cardiac function, providing mechanistic proof-of-concept for the functional significance of this methylation event and the potential for pharmacological reversal.

4.2 SERCA2a Promoter Methylation and Calcium Dysregulation

Impaired intracellular calcium cycling is among the most consistent cellular defects observed in HF, and downregulation of SERCA2a — the principal calcium pump responsible for returning cytosolic Ca²? to the sarcoplasmic reticulum during diastole — is a mechanistically critical contributor. While SERCA2a downregulation in HF has been attributed to reduced mRNA stability, microRNA-25-mediated translational suppression, and reduced phospholamban phosphorylation, mounting evidence implicates CpG hypermethylation within the ATP2A2 promoter as an additional transcriptional silencing mechanism. DNMT3A has been shown to occupy the ATP2A2 promoter in hypertrophied cardiomyocytes, and its pharmacological inhibition partially restores SERCA2a expression in rodent HF models. This is of therapeutic significance given that gene therapy-based restoration of SERCA2a via AAV1 vectors advanced to Phase 3 clinical evaluation, and epigenetic reactivation of endogenous SERCA2a expression represents a potentially more scalable delivery approach.

4.3 eNOS Methylation and Vascular Dysfunction

Endothelial nitric oxide synthase (eNOS), encoded by NOS3, is the primary generator of vascular nitric oxide (NO) — a molecule central to endothelium-dependent vasodilation, inhibition of platelet aggregation, vascular smooth muscle relaxation, and cardioprotection against ischaemia–reperfusion injury. eNOS expression is regulated by a promoter region containing multiple CpG sites whose methylation state reflects vascular health. Hypermethylation of the NOS3 promoter has been documented in endothelial cells from patients with coronary artery disease and in the myocardial microvasculature of patients with HF with preserved ejection fraction (HFpEF), reducing eNOS transcription and impairing NO bioavailability. The resulting endothelial dysfunction diminishes coronary microvascular reserve, contributes to hypertension, and promotes the diastolic stiffness characteristic of HFpEF — mechanistically linking cardiovascular risk factor exposure, vascular endothelial epigenetic ageing, and the pathophysiology of HFpEF.

4.4 Promoter Methylation of Anti-Fibrotic Genes

Myocardial fibrosis — the substitution of cardiomyocytes by interstitial collagen produced by activated cardiac fibroblasts and myofibroblasts — is a major determinant of diastolic dysfunction, cardiac stiffness, and arrhythmia susceptibility in HF. Several genes whose products normally restrain cardiac fibroblast activation are epigenetically silenced through promoter methylation in fibrotic myocardium. RASSF1A, a tumour suppressor that limits Ras-mediated fibroblast proliferation, undergoes promoter hypermethylation in hearts from patients with hypertrophic cardiomyopathy. SMAD7, which ordinarily restrains TGF-β-driven collagen synthesis, is also silenced partly through DNMT-mediated methylation of its regulatory region in cardiac fibroblasts exposed to angiotensin II or TGF-β1. These methylation events therefore function as positive feedback mechanisms that sustain fibrosis once initiated — and represent potential targets for DNMT inhibitor-based interventions aimed at arresting progressive fibrotic remodelling.

5. Histone Modifications and Cardiac Function

5.1 Lysine Acetylation: HATs and HDACs in the Heart

Histone lysine acetylation — the transfer of an acetyl group from acetyl-CoA to the ε-amino group of specific lysine residues — neutralises the positive charge on the lysine side chain, weakening its electrostatic interaction with the negatively charged DNA phosphate backbone and facilitating chromatin decompaction. Acetylation at H3K9, H3K14, H3K18, H4K5, H4K8, H4K12, and H4K16 is strongly associated with transcriptional activation, while acetylation at H3K56 and H4K91 participates in nucleosome reassembly during DNA replication. This dynamic, rapidly reversible modification is catalysed by a family of HATs (p300/CBP, GCN5/PCAF, MOZ, Tip60, and others) and counteracted by eighteen HDAC enzymes organised into four classes: class I (HDAC1, 2, 3, 8), class IIa (HDAC4, 5, 7, 9), class IIb (HDAC6, 10), and class IV (HDAC11), with the class III sirtuins (SIRT1–7) forming a structurally distinct, NAD?-dependent deacetylase family.In the healthy adult heart, the acetylation landscape at cardiac gene loci is maintained by constitutive HAT activity — particularly p300/CBP-mediated acetylation at MEF2, GATA4, and NKX2.5 binding sites — supporting expression of contractile proteins, metabolic enzymes, and ion channel subunits. The acetyltransferase p300 occupies a particularly important position: it acetylates both H3K18/K27 and the GATA4 transcription factor directly, and its cardiac-specific deletion produces dilated cardiomyopathy in mice, confirming non-redundant roles in maintaining cardiomyocyte gene expression programmes.

5.2 HDAC Activity in Cardiac Hypertrophy and Fibrosis

Elevated HDAC activity in the hypertrophied and failing heart was among the first epigenetic abnormalities characterised in cardiac disease, and its pharmacological reversal remains among the most intensively investigated therapeutic strategies in cardiac epigenetics. Class I HDACs — particularly HDAC2 — are upregulated at both mRNA and protein levels in hypertrophied cardiomyocytes from rodent pressure overload models and from patients with end-stage HF. HDAC2 deacetylates H3K9 at the promoters of anti-hypertrophic genes including Inpp5f and Hsp70, silencing these protective pathways and permitting pro-hypertrophic signalling through PI3K/AKT to proceed without restraint. Genetic deletion of HDAC2 in mice subjected to transverse aortic constriction attenuates hypertrophic remodelling, while overexpression is sufficient to induce hypertrophy — establishing its causal role beyond mere association.Class IIa HDACs — notably HDAC4, HDAC5, and HDAC9 — regulate cardiac gene expression through a distinct mechanism that involves nucleocytoplasmic shuttling rather than direct chromatin deacetylation. Under basal conditions, class IIa HDACs reside in the nucleus in association with MEF2 transcription factors, suppressing MEF2-driven gene activation. Upon pathological stimulation, calmodulin-dependent kinase II and protein kinase D phosphorylate class IIa HDACs, triggering their cytoplasmic export and relieving MEF2 repression — activating the hypertrophic gene programme. The subsequent reactivation or dysregulated nuclear re-entry of class IIa HDACs contributes to the oscillating transcriptional dynamics that characterise the hypertrophic response.

5.3 Bromodomain Proteins and BET Inhibition

Bromodomain and Extra-Terminal (BET) domain proteins — BRD2, BRD3, BRD4, and the testis-specific BRDT — contain tandem bromodomains that recognise and bind acetylated lysine residues on histones, particularly H4K5ac, H4K8ac, and H4K12ac at active enhancers. By serving as readers of the acetylation mark, BET proteins recruit the positive transcription elongation factor b (P-TEFb) to active genes, promoting RNA polymerase II pause-release and productive transcriptional elongation. BRD4, the most extensively studied BET family member in cardiovascular biology, accumulates at super-enhancers — dense clusters of enhancers with exceptionally high transcription factor and mediator occupancy — that drive master regulatory gene networks controlling hypertrophic and fibrotic programmes in both cardiomyocytes and cardiac fibroblasts.

The demonstration that the small-molecule BET inhibitor JQ1 attenuates pressure overload-induced cardiac hypertrophy by displacing BRD4 from super-enhancers and interrupting NF-κB-dependent inflammatory transcription represented a pivotal proof-of-concept in cardiac epigenetics. Subsequent work established that BRD4 accumulates at the super-enhancer controlling Nppb gene expression in hypertrophied cardiomyocytes, and that its inhibition reduces BNP production — a finding with direct relevance to the clinical use of natriuretic peptides as HF biomarkers. BET inhibition has also been shown to reduce TGF-β-driven cardiac fibroblast activation and collagen synthesis, suggesting utility that extends across cardiomyocyte and stromal cell compartments.

5.4 Histone Methylation: H3K4 and H3K27 in the Cardiac Epigenome

Histone lysine methylation exerts context-dependent effects on transcription, with the regulatory outcome determined by both the specific lysine residue targeted and the degree of methylation (mono-, di-, or trimethylation). H3K4 trimethylation (H3K4me3) marks actively transcribed gene promoters and is catalysed principally by COMPASS complex components MLL1–4, SET1A, and SET1B. In the adult heart, H3K4me3 is enriched at sarcomeric, metabolic, and calcium-handling gene promoters whose sustained expression is required for normal systolic and diastolic function. Its redistribution in HF — with loss at structural gene loci and gain at foetal gene promoters — tracks the transcriptional reprogramming of the failing cardiomyocyte with remarkable fidelity, rendering it a potentially powerful epigenomic indicator of gene expression state.

H3K27 trimethylation (H3K27me3), deposited by the Polycomb Repressive Complex 2 (PRC2) through its catalytic subunit EZH2, is the defining mark of developmental gene silencing and plays a central role in suppressing the foetal cardiac gene programme in the adult heart. EZH2 expression is elevated in pressure-overloaded and failing hearts, and its accumulation at anti-hypertrophic gene loci — including those encoding SERCA2a regulatory proteins and metabolic transcription factors — reinforces the maladaptive transcriptional state of HF. Pharmacological EZH2 inhibition with compounds such as GSK126 and tazemetostat reduces H3K27me3 at these loci, reactivates their expression, and attenuates hypertrophic remodelling in preclinical models, establishing EZH2 as a tractable therapeutic target in cardiac pathology.

Figure 3. Histone modification landscape in cardiac physiology and heart failure. The four major post-translational modification types — acetylation, methylation, phosphorylation, and ubiquitination — are depicted at the nucleosome level with their cognate writer and eraser enzymes. Heart failure-specific perturbations are highlighted, including elevated HDAC4/5 activity, EZH2-mediated H3K27me3 accumulation at cardioprotective gene loci, and γH2AX accumulation reflecting cardiomyocyte DNA damage.

Table 1. Key Histone Modifications in Cardiac Physiology and Their Dysregulation in Heart Failure

IEG = immediate early gene; PRC1/2 = Polycomb Repressive Complex 1/2; HF = heart failure. ↑ = increased, ↓ = decreased in failing versus non-failing myocardium. Status reflects the predominant direction in human HF tissue and validated rodent models.

6. Histone Modification Dysregulation in Heart Failure

6.1 The HDAC–HAT Imbalance

A recurring theme in experimental HF models and in clinical HF tissue is the disruption of the HDAC–HAT balance in favour of excess deacetylase activity. This shift manifests as a global reduction in H3K9ac and H3K14ac at cardiac gene promoters, producing the net hypoacetylation that constrains transcriptional output at these loci. The mechanistic basis of this imbalance is multifactorial: HDAC2 is transcriptionally upregulated in response to haemodynamic and neurohormonal stress; class IIa HDACs exhibit altered nuclear–cytoplasmic distribution under conditions of elevated cytosolic calcium and CaMKII activation; and HAT activity — particularly that of p300 — may be constrained by reduced acetyl-CoA availability in the metabolically compromised failing heart, where the shift from fatty acid to glucose oxidation alters the acetyl-CoA/CoA ratio.

The HDAC2-mediated hypoacetylation of anti-hypertrophic gene promoters is compounded by the simultaneous recruitment of HDAC-containing NuRD and CoREST co-repressor complexes to specific genomic loci by hypertrophy-activated transcription factors including NRSF/REST, Hop/HHEX, and the co-repressor NCoR. These complexes act as local hubs of chromatin compaction that silence the transcriptional programmes required for adult cardiomyocyte identity — including adult isoforms of MHC, SERCA2a, and ion channel subunits. Restoring the HDAC–HAT balance through pharmacological HDAC inhibition, HAT activation, or disruption of HDAC-containing co-repressor complexes has therefore emerged as a leading strategy for reversing epigenetic reprogramming in HF.

6.2 Polycomb and Trithorax Group Protein Antagonism

The opposing activities of Polycomb Group (PcG) protein complexes — which deposit repressive H3K27me3 and H2AK119ub1 marks — and Trithorax Group (TrxG) protein complexes — which establish active H3K4me3 and H3K27ac marks — constitute the core epigenetic bivalency switch governing whether developmental gene loci are held in repressed or active chromatin states. In the adult heart, many foetal cardiac gene promoters are maintained in a bivalent configuration — simultaneously bearing H3K27me3 (repressive) and H3K4me3 (permissive) — poised for activation upon pathological stress. Under pressure overload, this balance shifts toward PcG dominance at adult cardiac gene loci and TrxG dominance at foetal gene loci, reinforcing the transcriptional reversion to a foetal programme.EZH2, the PRC2 catalytic component, co-immunoprecipitates with cardiac transcription factors GATA4 and MEF2 in hypertrophied cardiomyocytes, supporting a model of transcription factor-guided EZH2 recruitment to specific cardiac loci rather than random genome-wide H3K27me3 spreading. JARID2, a non-catalytic PRC2 subunit that modulates EZH2 targeting and activity, is elevated in failing myocardium and contributes to EZH2-dependent silencing. Conversely, KMT2D (MLL4), the principal H3K4 mono-methylase at cardiac enhancers, is downregulated in failing hearts, diminishing H3K4me1 at anti-hypertrophic enhancer elements and reducing their activity. The epigenetic landscape that emerges — elevated H3K27me3 at active adult gene loci combined with reduced H3K4me1 at their enhancers — creates a self-reinforcing chromatin state hostile to normal adult cardiac gene expression.

6.3 Histone Phosphorylation in Cardiac Stress

While histone acetylation and methylation have attracted the majority of research attention in cardiac epigenetics, histone phosphorylation contributes critical and often underappreciated functions to the cardiac stress response. H3S10 phosphorylation, catalysed by aurora B kinase and MSK1/2, is associated with chromatin condensation during mitosis but also functions as a transcriptional activator at immediate-early response genes during interphase, governing expression of c-Fos, c-Jun, and other transcription factors that initiate the hypertrophic programme in response to mechanical stretch or neurohormonal stimulation. H2AX phosphorylation at serine 139 (γH2AX), the canonical marker of DNA double-strand breaks, is markedly elevated in cardiomyocytes from failing hearts, reflecting genotoxic burden from reactive oxygen species, mitochondrial dysfunction, and telomere attrition. Accumulation of γH2AX correlates with activation of the DNA damage response and p53-dependent apoptotic signalling in cardiomyocytes, linking histone phosphorylation to cardiomyocyte loss as a driver of HF progression.

6.4 SIRT1 and the NAD?-Dependent Deacetylase Network

The class III HDACs — the seven mammalian sirtuins (SIRT1–7) — require NAD? as a co-substrate for deacetylase activity, directly coupling the energetic and redox state of the cardiomyocyte to its epigenetic landscape. SIRT1, the most extensively characterised sirtuin in cardiac biology, deacetylates both H3K9 and non-histone targets including p53, FOXO transcription factors, and PGC-1α, integrating metabolic sensing with transcriptional control of mitochondrial biogenesis, antioxidant defence, and cardiomyocyte survival. SIRT1 activity is reduced in aged myocardium and in HF, partly through declining NAD?/NADH ratios as mitochondrial oxidative phosphorylation is compromised. This reduction has downstream consequences for mitochondrial gene expression — through impaired PGC-1α deacetylation — and for cardiomyocyte apoptosis through hyperacetylation of p53, heightening its pro-apoptotic transcriptional activity. NAD? precursor supplementation with nicotinamide riboside or nicotinamide mononucleotide has shown benefit in murine HF models and is under active clinical evaluation.

7. Crosstalk Between DNA Methylation and Histone Modifications

7.1 Mechanistic Integration at the Chromatin Level

DNA methylation and histone modifications do not operate as parallel and independent regulatory systems; instead, they are deeply interconnected through multiple molecular mechanisms that collectively define the transcriptional state of individual genomic loci. This crosstalk is bidirectional: DNA methylation can direct histone modification patterns, and histone modification states can influence the recruitment and activity of DNMTs. Appreciating the regulatory logic of this crosstalk is essential for predicting the combined consequences of epigenetic perturbations and for designing rational combinatorial therapeutic strategies.The best-characterised pathway linking DNA methylation to histone modification involves MBD proteins, which recognise 5-mC and subsequently recruit HDAC-containing co-repressor complexes. MBD2, for example, is a component of the NuRD complex, whose HDAC1/2 and CHD4 chromatin-remodelling activities cooperate to compact chromatin at methylated loci. MECP2 recruits the Sin3A–HDAC1 co-repressor complex to methylated CpG loci in the heart, contributing to the silencing of neurohormonal response genes. The DNA methylation → MBD → HDAC pathway thus represents a direct biochemical mechanism through which promoter CpG methylation leads to chromatin compaction — explaining why methylated genes are not merely occluded from the transcriptional machinery but actively maintained in compact chromatin through histone-modifying activities.

7.2 DNMT Recruitment by Histone Methylation

The reverse direction of crosstalk — from histone modification to DNA methylation — is exemplified by the interaction between H3K9me2/3 and DNMT3A/3B. The ADD domain of DNMT3A binds H3K9me2/3-marked tails with high affinity while being repelled by H3K4me3. This means that H3K9me2/3-enriched heterochromatic regions actively recruit DNMT3A, ensuring that loci marked by repressive histone methylation also acquire the more stable and heritable DNA methylation mark — a mechanism for converting transient transcriptional repression into permanent, DNA-methylation-based silencing.EZH2-mediated H3K27me3 has been reported to recruit DNMT3A and DNMT3B to specific loci in cancer contexts, raising the possibility that PRC2-dependent repression in the failing heart may be reinforced by concurrent de novo DNA methylation — a dual-lock silencing arrangement more resistant to pharmacological reversal than histone modification alone. If validated in cardiac cells, this mechanism would suggest that EZH2 inhibitors could be most effective when combined with DNMT inhibitors, preventing residual DNA methylation from sustaining gene repression after H3K27me3 removal.

7.3 Bivalent Chromatin States at Cardiac Developmental Loci

Bivalent chromatin domains — characterised by simultaneous H3K4me3 and H3K27me3 at the same promoter — were originally described in embryonic stem cells as a mechanism for holding developmental genes in a poised state: repressed yet primed for rapid activation upon differentiation. In the adult heart, foetal cardiac gene loci (MYH7, NPPA, NPPB, ACTA1) adopt bivalent configurations that permit their rapid reactivation under pathological stress — a biological design that enables an adaptive transcriptional response to injury but whose persistence contributes to the maladaptive phenotype of chronic HF. The balance between EZH2 and KDM6A (UTX) at these loci determines the extent of H3K27me3 and thereby regulates the foetal gene expression programme in the stressed heart.

7.4 Enhancer–Promoter Looping and 3D Chromatin Organisation

The interplay between DNA methylation, histone modification, and non-coding RNAs ultimately operates within the three-dimensional chromatin architecture of the cardiomyocyte nucleus. Active enhancers — marked by H3K4me1, H3K27ac, and CBP/p300 — physically contact their target promoters through chromatin loops stabilised by the cohesin complex and the architectural protein CTCF. CTCF binding is directly sensitive to CpG methylation at its recognition motif: methylation of even a single CpG within the CTCF binding site prevents CTCF occupancy, severing the chromatin loop and isolating the target gene from its enhancer. In the failing heart, aberrant CTCF binding site methylation has been proposed to disrupt the enhancer–promoter contacts required for sustained expression of calcium-handling and contractile protein genes — a three-dimensional epigenetic mechanism with profound functional consequences extending beyond local chromatin modification.

8. Non-Coding RNAs in Cardiac Epigenetic Regulation

8.1 MicroRNAs in Cardiac Hypertrophy and Fibrosis

MicroRNAs constitute one of the most functionally consequential regulatory layers in cardiac biology, with individual miRNA species controlling entire transcriptional networks by simultaneously suppressing hundreds of target mRNAs with partial sequence complementarity. In cardiac hypertrophy, miR-21 — the most abundantly expressed and intensively studied miRNA in the stressed heart — exerts pro-fibrotic effects by silencing SPRY1 and PTEN in cardiac fibroblasts, sustaining fibroblast survival, proliferation, and ECM secretion. MiR-21 is secreted from cardiac fibroblasts in exosomes taken up by cardiomyocytes, representing a paracrine epigenetic signalling mechanism of considerable physiological significance.The miR-1/miR-133 family comprises the most cardiac-specific miRNAs, together constituting approximately 40% of total cardiac miRNA expression. MiR-1 targets Gata4, Hand2, and connexin 43, modulating hypertrophy and electrical coupling; miR-133 suppresses RhoA, CDC42, and NELFA, restraining the foetal gene programme and cardiomyocyte apoptosis. Both are markedly downregulated in hypertrophied and failing hearts, and their restoration in rodent models reduces hypertrophy and improves cardiac function. MiR-499, derived from an intron of the MYH7B gene and co-expressed with it during foetal cardiac development, represses calcineurin/NFAT hypertrophic signalling through targeting MAP3K3 and Id1. Its sustained expression in the adult heart constrains calcineurin-mediated hypertrophic signalling; its loss in HF removes this brake.

8.2 Long Non-Coding RNAs in Heart Failure

Long non-coding RNAs have emerged as key regulators of the cardiac epigenome, functioning through chromatin scaffolding, enhancer RNA activity, ceRNA sponge activity, and direct interaction with epigenetic enzymes. MHRT, transcribed from the MYH7 locus in an antisense orientation, is among the most functionally characterised cardiac lncRNAs. In the healthy adult heart, MHRT maintains the α-MHC-dominant phenotype by occupying the helicase domain of the chromatin remodeller Brg1, preventing Brg1 from targeting the Myh6 promoter and driving isoform switching toward foetal β-MHC. Under pressure overload, the MHRT–Brg1 interaction is disrupted, allowing Brg1 to silence Myh6 and activate Myh7 — contributing directly to contractile dysfunction. MHRT forms a feedback loop with MYH7 expression itself, ensuring proportional self-regulation of contractile isoform composition.

CHAST promotes cardiomyocyte hypertrophy by functioning as a ceRNA sponge for the anti-hypertrophic miR-93, de-repressing the autophagy regulator PLEKHM1 and impairing autophagic clearance of hypertrophic stimuli. CARL sponges miR-539, protecting cardiomyocytes from mitochondria-related apoptosis by shielding PHB2 from miRNA-mediated suppression. LIPCAR is of particular clinical relevance: measurable in plasma, its circulating levels predict mortality and cardiac remodelling in HF patients, marking it as a potentially useful non-invasive epigenetic biomarker.

8.3 Circular RNAs in Cardiac Pathophysiology

Circular RNAs — generated by back-splicing of pre-mRNA and protected from exonucleolytic degradation by their covalently closed structure — have rapidly accumulated a compelling evidence base in cardiac biology. CircHIPK3, upregulated in hypoxic cardiomyocytes, functions as a miR-29a sponge that prevents miR-29a from suppressing DNMT3A — creating an indirect mechanism through which circRNA dysregulation can alter the cardiac DNA methylation landscape. CircACTA2 sponges miR-548f-5p to promote cardiac fibroblast activation, and its abundance is elevated in hypertrophic and ischaemic cardiomyopathy. The cardiac-specific circRNA HRCR suppresses hypertrophy by sponging miR-223, preserving ARC (apoptosis repressor with caspase recruitment domain) levels and protecting cardiomyocytes against angiotensin II-induced hypertrophic signalling.

9. Epigenetic Biomarkers in Cardiovascular Disease

9.1 Circulating DNA Methylation Signatures

The presence of circulating cell-free DNA (cfDNA) in plasma and serum — released from dying cells — provides a non-invasive window into tissue-specific methylation patterns. Since cardiomyocyte-derived cfDNA carries the methylation signature characteristic of cardiac tissue at cardiomyocyte-specific gene loci, targeted bisulphite sequencing or methylation-specific PCR applied to cardiac-specific CpG markers in circulating cfDNA can identify cardiomyocyte injury with tissue specificity. In patients with acute myocardial infarction, cardiac-specific cfDNA methylation signatures are detectable within hours of coronary occlusion and correlate with infarct size assessed by cardiac MRI.

In chronic HF, genome-wide methylation profiling of peripheral blood mononuclear cells has identified methylation signatures that distinguish HF patients from controls and, within the HF population, associate with left ventricular ejection fraction, BNP levels, and exercise tolerance. A methylation-based epigenetic clock — estimating biological age from the methylation status of age-associated CpG sites — has demonstrated accelerated epigenetic ageing in HF patients that predicts adverse outcomes independently of chronological age and conventional risk factors.

9.2 Histone Modification Biomarkers

Circulating histones and modified histone fragments released from damaged or dying cells are measurable in plasma by mass spectrometry or ELISA-based approaches. Histone H3 citrullination (H3Cit), generated by PAD4 as part of neutrophil extracellular trap formation, is elevated in patients with acute decompensated HF and correlates with disease severity and in-hospital mortality. Plasma H4 acetylation at K12 and K16 has been proposed as an indicator of global chromatin acetylation state with implications for in vivo HDAC activity assessment. Technical challenges of plasma histone measurement — including contribution from multiple cell types, ex vivo mark instability, and haemolysis — remain barriers to clinical translation, but mass spectrometry-based approaches are progressively overcoming these analytical limitations.

9.3 MicroRNA Biomarkers

Circulating miRNAs — stabilised by Argonaute protein binding, lipoprotein packaging, or encapsulation within exosomes — have attracted intense interest as cardiac biomarkers since seminal reports of elevated plasma miR-1 and miR-133 in acute myocardial infarction. In HF, distinctive miRNA profiles characterise different aetiological subtypes: miR-423-5p is elevated in HF with reduced ejection fraction and correlates with NT-proBNP levels; miR-21 is raised in cardiac fibrosis and predicts adverse remodelling after myocardial infarction; the miR-29 family is downregulated in both ischaemic and non-ischaemic HF and inversely correlates with the degree of myocardial fibrosis; and miR-210 elevation reflects cardiac hypoxia and predicts worse HF outcomes. Multi-miRNA panels incorporating these markers provide diagnostic and prognostic information complementary to conventional biomarkers such as troponin and BNP.

Table 2. Summary of Epigenetic Biomarkers in Cardiovascular Disease: Potential for Diagnosis and Prognosis

HF = heart failure; MI = myocardial infarction; MACE = major adverse cardiovascular event; LVEF = left ventricular ejection fraction; BNP = brain natriuretic peptide; 6MWT = six-minute walk test; ICU = intensive care unit; HCM = hypertrophic cardiomyopathy; cfDNA = cell-free DNA; NET = neutrophil extracellular trap; DMR = differentially methylated region; EF = ejection fraction.

10. Therapeutic Implications: Targeting the Cardiac Epigenome

10.1 HDAC Inhibitors in Cardiac Disease

The most extensively investigated epigenetic therapeutic approach in HF is pharmacological HDAC inhibition. Pioneering experiments by Antos and colleagues in 2003 demonstrated that treating mice subjected to pressure overload with the pan-HDAC inhibitor trichostatin A prevented cardiac hypertrophy, fibrosis, and systolic dysfunction — establishing that HDAC activity is required for pathological remodelling, not merely a downstream consequence of it. Subsequent studies with more clinically applicable agents — vorinostat (SAHA), scriptaid, mocetinostat, and entinostat — have consistently shown attenuation of hypertrophy, reduction of fibrosis, preservation of systolic function, and mitigation of adverse remodelling in rodent and large-animal models across diverse HF aetiologies including pressure overload, ischaemia–reperfusion, and neurohormonal activation.

The cardiac benefit of HDAC inhibitor treatment extends beyond histone hyperacetylation. Treatment also acetylates non-histone targets including Hsp90 (impairing its chaperone support for hypertrophic signalling kinases), MnSOD (enhancing mitochondrial antioxidant capacity), α-tubulin (modulating cardiomyocyte mechanosensing), and PCNA (regulating DNA repair). Anti-fibrotic effects are partly mediated through acetylation and functional impairment of TGF-β pathway components in cardiac fibroblasts. Class I-selective HDACi demonstrate anti-hypertrophic and anti-fibrotic efficacy in preclinical models at doses below those associated with haematological toxicity, suggesting a potential therapeutic window for cardiac application broader than initially extrapolated from the oncology experience.Clinical translation is at an early but active stage. The HDAC inhibitor valproic acid, approved for epilepsy and bipolar disorder, has documented cardiac effects: epidemiological analyses indicate lower HF hospitalisation rates in patients receiving valproate for neurological indications. Prospective trials of class-selective HDACi in HF with reduced and preserved ejection fraction are in planning or early recruitment phases, building on the cardiac safety data accumulated in haematological oncology.

10.2 DNMT Inhibitors: Restoring Cardiac Gene Expression

The nucleoside analogue DNMT inhibitors azacitidine and decitabine achieve DNA demethylation by incorporating into DNA during replication and irreversibly trapping DNMT1, depleting active DNMT through proteasomal degradation. Their use in cardiac contexts is motivated by the prospect of reactivating methylation-silenced cardioprotective genes — RASAL1, SERCA2a, eNOS, RASSF1A — to restore anti-hypertrophic, anti-fibrotic, and vasoprotective functions. In murine pressure overload models, azacitidine treatment reduces cardiac hypertrophy and fibrosis partly through RASAL1 reactivation; in a rat infarction model, decitabine reduced infarct size and improved ventricular function in association with SERCA2a reactivation and reduced inflammatory gene expression.Non-nucleoside DNMT inhibitors — including RG108, SGI-1027, and hydralazine, which has DNMT inhibitory activity alongside its vasodilatory effects — offer alternatives that are less dependent on cell division and potentially less genotoxic than nucleoside analogues. Hydralazine, long used in HF therapy, may achieve some of its cardiac benefits through epigenetic mechanisms — a hypothesis that, if validated, would represent inadvertent epigenetic pharmacology predating the field itself. SAM-competitive DNMT inhibitors, targeting the methyl-donor substrate binding site rather than the cytosine substrate, represent a mechanistically distinct class with potential for more selective de novo methylation inhibition.

10.3 BET Bromodomain Inhibitors

BET bromodomain inhibitors displace BET proteins from acetylated chromatin by competing for binding within the acetylated lysine-recognising bromodomain pocket. Their mechanism is fundamentally different from HDACi — they do not alter the histone acetylation mark but prevent its reading by transcriptional co-activators. JQ1, the prototypic BET inhibitor, has shown notable preclinical cardiac efficacy: it attenuates pressure overload-induced hypertrophy by preventing BRD4-mediated transcriptional elongation at pro-hypertrophic super-enhancers, reduces BNP expression, suppresses NF-κB-driven inflammatory programmes in cardiomyocytes, and inhibits TGF-β-induced myofibroblast transdifferentiation. Its anti-hypertrophic effect in vivo appears disproportionately focused on a small number of super-enhancer-regulated transcription factors that control the broader hypertrophic programme — consistent with the concept that BRD4 inhibition at super-enhancers has outsized effects on master regulatory gene expression.

Clinical BET inhibitors — molibresib, OTX015, and CPI-0610 — have been evaluated principally in haematological malignancies. A Phase 1/2 trial of CPI-0610 specifically in HF with reduced ejection fraction provided early clinical data reporting dose-dependent reductions in BNP and NT-proBNP alongside improvements in left ventricular volumes — a signal warranting further evaluation. Achieving BET inhibition in the heart at doses below those causing systemic toxicity, principally thrombocytopenia and anaemia from haematopoietic progenitor cell dependence on BRD4, will determine the therapeutic window for cardiac application.

10.4 EZH2 Inhibitors

EZH2, the histone methyltransferase component of PRC2 that deposits H3K27me3, represents a mechanistically specific and pharmacologically tractable target in HF given its elevated expression in failing myocardium and its documented role in silencing anti-hypertrophic and pro-metabolic gene loci. GSK126, a highly selective EZH2 inhibitor competing with SAM at the methyltransferase active site, reduces H3K27me3 at cardiac metabolic gene promoters and restores their expression in pressure-overloaded mice; EZH2-deficient cardiomyocytes are protected against pressure overload-induced contractile dysfunction. Tazemetostat, an EZH2 inhibitor approved by the FDA for epithelioid sarcoma and follicular lymphoma, carries a well-characterised oncology safety profile and is a candidate for cardiac repurposing.

The cardiac application of EZH2 inhibitors must account for the dual role of EZH2 in cardiac biology. While pathologically elevated EZH2 activity in HF appears detrimental, EZH2 is required for normal cardiac development, and its complete loss produces congenital heart defects. A therapeutic strategy employing partial, dose-titrated EZH2 inhibition — sufficient to reduce excess H3K27me3 at specific maladaptive loci without globally disrupting PRC2 function — represents a pharmacologically demanding but mechanistically justified objective.

Figure 4. Pharmacological targeting of the cardiac epigenome. Left panel: hub-and-spoke representation of the six principal therapeutic approaches targeting the failing heart epigenome, with representative compounds for each class. Right panel: current clinical development pipeline for epigenetic agents with cardiac applications, stratified by development stage from preclinical to Phase II.