Listen to this manuscript’s

audio summary by

JACC Editor-in-Chief

Dr. Valentin Fuster.

J O U R N A L O F T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y V O L . 7 0 , N O . 5 , 2 0 1 7

ª 2 0 1 7 B Y T H E A M E R I C A N CO L L E G E O F C A R D I O L O G Y F O U N DA T I O N

P U B L I S H E D B Y E L S E V I E R

I S S N 0 7 3 5 - 1 0 9 7 / $ 3 6 . 0 0

h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j a c c . 2 0 1 7 . 0 5 . 0 6 7

THE PRESENT AND FUTURE

STATE-OF-THE-ART REVIEW

Translational Perspective on Epigeneticsin Cardiovascular Disease

Pim van der Harst, MD, PHD,a,b Leon J. de Windt, PHD,c John C. Chambers, MD PHDd,e

ABSTRACT

Fro

the

of

Ep

dle

1R

Na

Du

en

is

rel

Ma

A plethora of environmental and behavioral factors interact, resulting in changes in gene expression and providing a basis for

the development and progression of cardiovascular diseases. Heterogeneity in gene expression responses among cells and

individuals involves epigenetic mechanisms. Advancing technology allowing genome-scale interrogation of epigenetic marks

provides a rapidly expanding view of the complexity and diversity of the epigenome. In this review, the authors discuss the

expanding landscape of epigenetic modifications and highlight their importance for future understanding of disease.

The epigenome provides a mechanistic link between environmental exposures and gene expression profiles ultimately

leading to disease. The authors discuss the current evidence for transgenerational epigenetic inheritance and summarize

the data linking epigenetics to cardiovascular disease. Furthermore, the potential targets provided by the epigenome for

the development of future diagnostics, preventive strategies, and therapy for cardiovascular disease are reviewed.

Finally, the authors provide some suggestions for future directions. (J Am Coll Cardiol 2017;70:590–606)

© 2017 by the American College of Cardiology Foundation.

A plethora of environmental and behavioralfactors are involved in the developmentand progression of cardiovascular disease

(CVD). Also, a large variety of genetic variationshave been associated with CVD. For some geneticvariations, the causal pathway is clear: for example,cardiomyopathies caused by disruptive mutationsin genes involved in the sarcomere. For manyother cardiovascular traits and diseases, especiallythose involving multiple cells or organ systems,the role of genetics is far more complex andour current understanding is limited. Examplesinclude hypertension, dyslipidemia, atherosclerosis,

m the aDepartments of Cardiology and Genetics, University of Groninge

Netherlands; bDurrer Center for Cardiovascular Research, Netherlands He

Cardiology, CARIM School for Cardiovascular Diseases, Maastricht Unive

idemiology and Biostatistics, Imperial College London, London, United K

sex, United Kingdom. This work was partially supported by the Na

01HL104125. The content is solely the responsibility of the authors and do

tional Institutes of Health. Dr. de Windt has received support from the

tch Heart Foundation, Dutch Federation of University Medical Centers, Z

ces; has been supported by grant 311549 from the European Research Coun

a cofounder and stockholder of Mirabilis Therapeutics BV. Both other au

evant to the contents of this paper to disclose.

nuscript received March 28, 2017; revised manuscript received May 30, 2

myocardial infarction, atrial fibrillation, and heartfailure.

All cells carry essentially the same genetic infor-mation. Although there are a few exceptions, thegenetic information itself does not change during thelife of an organism. However, different cell typeshave highly heterogeneous gene expression profiles,resulting in the large variety of cells, tissues, andorgans with different functions throughout thehuman body. Epigenetic mechanisms control thesedifferences in gene expression. Techniques forgenome-scale analysis of epigenetic marks are nowavailable. The complexity and diversity of the

n, University Medical Center Groningen, Groningen,

art Institute, Utrecht, the Netherlands; cDepartment

rsity, Maastricht, the Netherlands; dDepartment of

ingdom; and the eEaling Hospital NHS Trust, Mid-

tional Institutes of Health under award number

es not necessarily represent the official views of the

Netherlands CardioVascular Research Initiative: the

onMW and the Royal Netherlands Academy of Sci-

cil and by VICI award 918-156-47 from the NWO; and

thors have reported that they have no relationships

017, accepted May 31, 2017.

AB BR E V I A T I O N S

AND ACRONYM S

5mC = 5-methylcytosine

CVD = cardiovascular disease

DNMT = deoxyribonucleic acid

methyltransferase

EWAS = epigenome-wide

association study

HAT = histone

acetyltransferase

HDAC = histone deacetylase

ncRNA = noncoding RNA

J A C C V O L . 7 0 , N O . 5 , 2 0 1 7 van der Harst et al.A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6 Epigenetics in Cardiovascular Disease

591

epigenome is increasingly appreciated. Disease-related epigenetic research was pioneered in thecancer field, but more recently, the cardiovascularfield has been quickly catching up.

Three concepts of epigenetics might be of interestto our understanding of the development and pro-gression of CVD. First, epigenetics might provide amechanistic link between environmental exposuresand gene expression profiles. For example, exposureof inflammatory cells to stimuli, such as infectiousagents and lipids, influences their epigenomes andprovides a link to development of atherosclerosis(1,2). Second, the paradigm that the genetic sequencealone defines the heritability of CVD is being chal-lenged (3,4). The full spectrum and variability ofheritability might also include heritable informationencoded in epigenomic variations. For example,accumulating evidence suggests that influences ofparental environmental exposures might be trans-mitted via epigenetic mechanisms and eventuallyaffect the offspring’s risk of developing diseases.Third, the epigenome provides some novel classes oftherapeutic targets, as it can be modified by micro-nutrients, drugs, and other factors (5–7).

In this review, we provide a translational overviewof epigenetic biology and the relevance of epigeneticsto cardiovascular physiology and disease, fromfundamental concepts to the clinical, populationhealth, and pharmacotherapy perspectives.

PRIMER ON BASIC EPIGENETIC CONCEPTS

Epigenetics is the collective name for the genomicmechanisms that influence gene expression, but donot involve variation in the deoxyribonucleic acid(DNA) sequence itself. Epigenetic modifications orevents are of major importance for several key bio-logical processes, including differentiation of cells,imprinting, and inactivation of the X chromosome.Epigenetic modifications are, in general, plastic andresponsive to external stimuli. Examples of stimulithat have been identified as affecting the epigenomeinclude pre-natal malnutrition, ultraviolet radiation,and cigarette smoke. The induced changes vary fromvery transient to long-lasting. Multiple epigeneticfeatures can be propagated from 1 generation of cellsto the next. Some epigenetic features are directlylinked to the DNA molecule itself (e.g., DNA methyl-ation), others relate to the dynamic remodeling ofchromatin or modifications of its associated proteins(e.g., post-translational histone modification), andothers involve ribonucleic acid (RNA) molecules (e.g.,gene silencing by noncoding ribonucleic acids[ncRNAs] and RNA methylation) (Figure 1). Here, we

will briefly expand on 3 key epigeneticmechanisms: 1) DNA methylation; 2) post-translational histone modification; and 3)ncRNA-based mechanisms.

DNA METHYLATION

Methylation of nucleotides is widespread,and is common to both DNA and RNA (8). Thebest-studied epigenetic mechanism ismethylation of nuclear DNA (Figure 1). Of the4 DNA nucleotides (A, C, G, T), methylation ofcytosine (C) is known as 5-methylcytosine

(5mC), and has been most thoroughly investigatedin higher eukaryotes, including humans. 5mC occurspredominantly at C followed by guanine (G) at so-called CpG sites. In human cells, 60% to 80% of thew28 million CpG sites are typically methylated.Methylation is usually linked to silencing of genes, asit can decrease the accessibility of chromatin andinhibit the function of DNA-binding proteins ortranscription factors (TFs) that are required for geneexpression. For example, DNA methylation competeswith NRF1 (a TF) binding to its DNA binding-sitemotif, suggesting that these methylation-sensitiveTFs depend on the absence of low levels ofmethylation (hypomethylation) (9) to induce geneexpression. Also, tumor necrosis factor-a can in-crease DNA methylation in the SERCA2a promotorregion, which results in lower levels of SERCA2atranscripts (10). However, the role of DNA methyl-ation is more complex, with a variable directionalrelationship to gene expression that can be context-dependent (11). More recent data suggest that thesilencing function might be limited and highly spe-cific for certain targets. In the case of the CCCTC-binding factor (zinc finger protein), demethylationdid not result in extensive occupancy of the TF,suggesting that methylation is not always a keydeterminant in gene silencing or expression (12).

CpG methylation in CpG-dense gene promotor re-gions (CpG islands at 50 transcriptional start sites) hasbeen well-studied, but only makes up <10% of thetotal CpG sites and usually has a low methylationstatus (<10%). Regions up to several kb distant arereferred to as CpG island shores, have a far morevariable methylation status, and are also believed tohave regulatory relevance for gene expression andrepression of retrotransposons (e.g., retrovirus-likeDNA sequences).

5mC PROGRAMMING BY DNA METHYLTRANSFERASES.

Key in DNA programming of 5mC are the epigeneticmodifying enzymes that deposit (methylases) orremove (demethylases) these marks. Three highly-

FIGURE 1 Epigenetic Modifications and Their Location

Epigenetics is the collective name for the genetic effects that result in gene expression but do not involve variation in the DNA sequence itself. These include the

chemical modifications to DNA itself (DNA methylation), the histones (around which DNA is wound) or non-coding RNA. A total of 8 histone units form chromatin,

around which 146 base pairs of DNA is wound to form the nucleosome. (A) DNA methylation occurs predominantly at the CpG islands. (Lower box) Epigenetic marks are

an important determinant of the differentiation and cell fate during development, and certain modifications can increase accessibility to DNA. (Upper box, B) Histone

complexes can have modifications at multiple positions on the tail, jointly making up the histone code. (C) Histone-remodeling complex slide histones in directions,

making the DNA accessible or not. DNA ¼ deoxyribonucleic acid; RNA ¼ ribonucleic acid.

van der Harst et al. J A C C V O L . 7 0 , N O . 5 , 2 0 1 7

Epigenetics in Cardiovascular Disease A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6

592

conserved deoxyribonucleic acid methyltransferases(DNMTs) catalyze the addition of methyl groups to Cto generate 5mC. DNMT1 is essential to maintenanceof established methylation patterns; it recognizeshemi-methylated DNA during replication and addsmethyl groups to the nonmethylated daughter strand.

This process allows the 5mC marks to be retainedduring cell division and cell differentiation. The other2 DNMTs (DNMT3A and DNTM3B) are involved in denovo methylation (13).DNA METHYLATION BEYOND 5mC. In addition to thefrequently-measured 5mC, intermediate forms have

J A C C V O L . 7 0 , N O . 5 , 2 0 1 7 van der Harst et al.A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6 Epigenetics in Cardiovascular Disease

593

also been discovered (e.g., 5-hydroxymethylcytosine,5-formylcytosine, and 5-carboxylcytosine). These in-termediate forms are increasingly receiving moreattention (14). Technological improvements andlowering of detection limits have resulted in thedemonstration that methylation of the adenine(6mA) not only occurs in bacteria, but also is presentat very low abundance in other organisms, includingChlamydomonas (15), Drosophila (16), C. elegans (17),and even in humans (at 0.00009%) (18). 6mA appearsto prefer TAGG sites and, in contrast to 5mC, 6mAmethylation occurs widely across genomes andappears to be more depleted in exonic regions. Inhigher eukaryotes, transcription start sites showed astrong decrease in dA6m, in contrast to findings inDrosophila (16) and Chlamydomonas (15), suggestingthat this specific epigenetic modification might evenhave distinct functions across eukaryotes. The func-tional role of 6mA, especially in humans, and possiblerelationship to disease remains to be established.Finally, methylation is not unique to nuclear DNA,but is also present in mitochondrial DNA and RNA.Bidirectional cross-talk between nuclear and mito-chondrial DNA via methylation may be essentialfor some critical aspects of cell function (19), andRNA methylation occurs in both coding and non-coding RNAs, influencing post-transcriptional controlof RNA function (8).

POST-TRANSLATIONAL

HISTONE MODIFICATIONS

Four different histone proteins (H2A, H2B, H3, andH4), structurally organized as octamers, form thenucleosome around which the DNA is wound(Figure 1). The intimate interaction of histones withDNA indeed implicates them in an important regula-tory role of DNA-dependent processes. Histonemodifications emerge as fundamental players in theregulation of nucleosome structure (also called thechromatin state), and thereby, the accessibility ofthe DNA to key proteins involved in gene transcrip-tion. Numerous histone modifications at manyamino-acid residue positions of the different histonetails are involved in generating a complex regulatorycode (Figure 1). Known functional histone modifica-tions include methylation, acetylation, glycosylation,carbonylation, ubiquitination, phosphorylation, andseveral others. At least 15 different type of modifica-tions have been identified to date for at least 130different sites on the histone tails. A dedicatednomenclature has been developed to unambiguouslyspecify the specific modification by listing: 1) thespecific histone protein; 2) the modified amino-acid

residue; and 3) the type of modification (20). Exam-ples are H3K27ac (acetylation of lysine 27 of histoneH3), H3K4me1 (mono-methylation of lysine 4 of his-tone H3), or H3K9me3 (tri-methylation of lysine 9 ofhistone H3). Some of these modifications are short-lived, with half-lives ranging from a few minutes toa couple of hours, whereas others are long-lastingand are maintained during cell division and differ-entiation. These long-lasting modifications are alsobelieved to be involved in “epigenetic memory.”

Histone modifications govern the interactions ofDNA and result in the activation or repression of genetranscription. Examples of frequently studied acti-vation marks include H3K27ac, H3K4me3, H3K4me1,and H3K36me3. Repressive marks include H3K27me3and H3K9me3. Obviously, the true complexity ofthis regulatory “histone language” is that regulationdoes not originate from a single modification, butfrom the large number or combinatorial possibilitiesof modifications of the different histones, evenwithin a single octamer. When multiple modifica-tions, both activating (H3K4me3) and repressive(H3K27me3), co-occur, this is called “poised” (biva-lent), and was first identified in the promoters ofdevelopmental genes in embryonic stem cells. Tech-nological advances now allow the colocalization ofthese modifications on a single nucleosome to beidentified using high-throughput techniques. Exten-sive bivalency of hypomethylated CpGs have alsobeen discovered to coincide with inactive promotersof development regulators in nonembryonic stemcells, and indicate a more common function in geneexpression regulation (21,22). The identity of left andright ventricular tissue and their gene expressionprofiles can also be characterized by histone marks.For example, expression of the atrial natriureticpeptide (ANP) and B-type natriuretic peptide genes inthe left ventricle has been associated with higherlevels of histone acetylation and methylation,whereas SERCA2a expression does not differ be-tween the left and right ventricle, and aMHCexpression in the left ventricle relative to the rightventricle correlated with H3K4 methylation (23).

The modified tails have also been suggested tofunction as docking sites and signaling platforms forregulatory and remodeling proteins, thereby influ-encing chromatin organization.

HISTONE PROGRAMMING. A plethora of enzymes canchange histones by adding (“writers”) or removing(“erasers”) modifications. Increasing data supports arole for these writer/eraser enzymes in cardiac path-ophysiology (24). Important “writers” are histoneacetyltransferases (HATs), and important “eraser”

van der Harst et al. J A C C V O L . 7 0 , N O . 5 , 2 0 1 7

Epigenetics in Cardiovascular Disease A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6

594

enzymes for the heart are class II histone deacetylases(HDACs), which can have repressive effects on myo-cyte enhancer factor 2 (activity) and thereby affectthe cardiac hypertrophy response to stress signals(25). In addition to “eraser” enzymes, an additionalmechanism has more recently been identified thatsimply removes the histone modifications by cleaving(or “clipping”) the tail, including its modificationsignal (26). The proteins whose functions are affecteddue to these histone modifications are also referredto as “reader” proteins (14).

The complexity of the regulation is not only drivenby the many histone modification variants and com-binations, but is further increased by epigeneticmechanisms influencing each other, resulting in acomplex interplay (27). DNA methylation can beassociated with specific histone modifications, butthey can also be mutually exclusive. The epigeneticeffects affect the accessibility of the DNA via thestructural position of the nucleosome. The position ofone nucleosome affects neighboring nucleosomes,which jointly result in “open” or “closed” chromatinstructures. The many chromatin remodeling mecha-nisms are also referred to jointly as the “EpigeneticCode REplication Machinery” (28).

RNA-BASED MECHANISMS

Based on data from the Encyclopedia of DNA Elements(ENCODE), only 1.2% of human DNA is estimated toencode protein-coding exons, but the vast majority ofthe genome is transcribed at some point in at least 1cell type (29,30). This majority of ncRNA (Figure 2)can be subdivided into constitutively-expressedtranscripts involved in structural “housekeeping”processes (e.g., transfer RNAs, ribosomal RNAs, smallnuclear RNAs, and small nucleolar RNAs) on the onehand, and ncRNAs involved in regulation of geneexpression that are specifically expressed as aresponse to or during cell differentiation, on the other(31). The rapidly-expanding class of regulatoryncRNAs, including microribonucleic acids (miRNAs),small interfering RNAs, antisense RNAs, Piwi-interacting RNAs, and long noncoding RNAs(lncRNAs), can influence the regulation of the chro-matin state and influence expression of (coding) RNA(Figure 2) (31,32). Regulatory ncRNAs are generallysubdivided according to their length into short(<200 nt; e.g., miRNAs, Piwi-interacting RNAs), orlong ncRNA (>200 nt; e.g., lncRNAs, circular RNAs[circRNAs]). A substantial amount of data supportsthe notion that ncRNAs influence gene expression viaa variety of mechanisms at multiple control levels(transcription, as well as translation). miRNAs, with a

size of w20 nt, are the best-studied group of ncRNAsto date in the cardiovascular system, and their mainmechanism of action involves the repression oftranslation of target messenger ribonucleic acids(mRNAs) and concomitant suppression of theircognate proteins (Figure 2). On occasion, miRNAscan also influence other epigenetic phenomena byinhibiting relevant proteins/enzymes, remodelingchromatin, by changing availability of the requiredsubstrates, or even by targeting the promoter (onDNA) itself and thereby acting as transcriptionalactivators/repressors (33). Compared with miRNAs,the mechanistic characterization of lncRNAs is stillincomplete, in part because they are evolutionarilypoorly conserved at the nucleotide sequence leveland due to the wide variety of gene regulatorymechanisms described so far. Many lncRNAs caninteract with chromatin-modifying enzymes or tran-scription factors and play a role in chromatin regu-lation (31). Some lncRNAs may show a backsplicingphenomenon that can give rise to circular RNAs, anRNA species that is stable, more tissue-specific, andevolutionarily conserved (34,35). A subset of lncRNAsmay code for micropeptides, illustrating the need torevise the definition and classification of lncRNAspecies (36,37). Finally, the role of RNA methylationprovides another, largely unexplored, and potentiallyrelevant layer of complexity to the function and sig-nificance of ncRNA. Verification of the presenceand/or absence of methylation in the different RNAspecies, the enzymes involved, and the consequencesfor regulation of gene expression should furtherincrease our understanding of their biologicalrole (8,38).

THE EVIDENCE FOR TRANSGENERATIONAL

EPIGENETIC INHERITANCE

The relationship between early growth and future riskof diseases, including diabetes, metabolic syndrome,and CVD, has been established in epidemiologicalstudies. The placenta is involved in early growth andhas been linked to birth weight and also to multiplediseases and disorders that develop later in life,including cardiovascular disease (3). This mechanismof intrauterine exposure is often confused with“epigenetic inheritance.”

To consider epigenetic inheritance as a potentialmechanism of transgenerational influence, it isimportant to realize that the egg that develops into afetus was likely to have arisen in the ovary of itsmotherwhen shewas developing in the grandmother’swomb (3). This provides a theoretical mechanism bywhich the environment, micronutrients, and other

FIGURE 2 Noncoding RNA

(A) Schematic overview of the transcriptome and the classification of RNA as coding or noncoding, with the different species of noncoding RNA. (B) Different

mechanisms of action of noncoding RNAs in epigenetic regulation. AAAAA ¼ poly-A tail; circRNA ¼ circular ribonucleic acid; HP1 ¼ heterochromatin protein 1;

lncRNA ¼ long noncoding ribonucleic acid; Me ¼ methyl; miRNA ¼ microribonucleic acid; ORF ¼ open reading frame; PCR2 ¼ polycomb repressive complex 2;

piRNA ¼ Piwi-interacting ribonucleic acid; PIWI ¼ p-element induced wimpy testis (protein that can bind with piRNA); rasiRNA ¼ repeat-associated small interfering

ribonucleic acid; RISC ¼ ribonucleic acid–induced silencing complex; RNA ¼ ribonucleic acid; RNA pol ¼ ribonucleic acid polymerase; scaRNA ¼ small Cajal body-

specific ribonucleic acid; siRNA ¼ small interfering ribonucleic acid; snoRNA ¼ small nucleolar ribonucleic acid; snRNA ¼ small nuclear ribonucleic acid; SUZ ¼ SUZ12

polycomb repressive complex 2 subunit; tiRNA ¼ transcription initiation ribonucleic acid.

J A C C V O L . 7 0 , N O . 5 , 2 0 1 7 van der Harst et al.A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6 Epigenetics in Cardiovascular Disease

595

van der Harst et al. J A C C V O L . 7 0 , N O . 5 , 2 0 1 7

Epigenetics in Cardiovascular Disease A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6

596

factors can influence the third future generationvia intrauterine exposures. This mechanism ofintrauterine exposure should not be confused with“true” transgenerational epigenetic inheritance. Truetransgenerational epigenetic inheritance involvesmaintenance of an epigenetic mark for at least 4generations in a gestating female (or 3 via the paternalgerm line), and thus requires incomplete erasureof the epigenetic signature during developmentalreprogramming.

The difficulty with the evidence to date is thatonly a very few studies, including experimentalstudies in animals, have focused on inheritance ofepigenetic marks for this duration or beyond (4). Inhumans, there is no direct evidence of transgenera-tional epigenetic inheritance, except for the parent-of-origin specificity of genomic imprinting (4). Inaddition, a considerable amount of variability inDNA methylation appears to be explained sufficientlyby the underlying genotypic variation, which mightexceed the influence of imprinting (39). The influenceof genetic variance on methylation can be very sub-stantial, with up to 15% of sites estimated to showgenetic influence, nearly all in cis, and a majority(w75%) of the cis-methylation quantitative trait lociseems to contain only a simple variation of the 2 CpGnucleotides (40). Interindividual variability ofDNA methylation might therefore be largely driven bycis-regulatory variations in the CpGs (41). Finally,data from monozygotic, dichorionic twins (eachhaving their own placenta) displayed greater within-pair expression discordance than monozygoticmonochorionic twins (sharing a placenta), suggestingthe strong influence of the intrauterine environment(42). Taken together, the available data suggestthat the commonly perceived inheritance of DNAmethylation may be driven more by genetics andintrauterine exposures. Proof-of-principle data sup-porting widespread transgenerational epigeneticinheritance is not available (4).

Nevertheless, the idea that early life, includingearly (in utero) exposure to environmental factors,such as nicotine, hormones, and nutrients, influencesdevelopment, with long-lasting effects on organ andtissue function (a phenomenon called developmentalprogramming [43,44]), also has major implications forour understanding of CVD development and pro-gression. In mammals, DNA demethylation in theparental gametes occurs by both active demethyla-tion by translocation (TET) proteins and passive lossof methylation during subsequent cell divisions (4).DNA methylation marks are re-established during theformation of primordial germ cells. However, a sec-ond general demethylation occurs after fertilization

in the developing embryo (45). Some regions(imprinted genes and retrotransposable elements)are exempted from this phase of demethylationand remethylation, in both mice and humans, whichallows them to maintain their parent-of-originmethylation state (46). Whether this resistance tothis second demethylation and remethylationpathway provides an evolutionary mechanismfor transfer of epigenetic marks remains to be estab-lished, but in theory, it is possible that intergenera-tional inheritance of transcriptionally-activetransposable elements plays a role in reducing therisk of germline mutations (4). A frequently-citedexample is the influence of maternal dietary genis-tein on DNA methylation of agouti, resulting indifferences in fur color and obesity in offspring(47). These developmental programming effects ofobesity or diabetes via epigenetics might predisposeoffspring to develop metabolic diseases in their laterlife by transmitting adverse environmental exposuresof parents to the next generation (48).

TRANSLATIONAL PERSPECTIVE

Many studies support the concept that changes inepigenetic factors influence biological processesrelated to CVD development, including diabetes, hy-pertension, obesity, atherosclerosis, atrial fibrillation,and heart failure. Characterization of these epigeneticdifferences may enable identification of the genesand biological mechanisms involved in cardiovascu-lar physiology and disease, and may further ourunderstanding thereof.

Key in progressing our knowledge on how epige-netic changes might affect CVD is a deep under-standing about what is normal, or what is thereference. Indeed, large efforts are currently ongoingto establish a reference for the human epigenome.ENCODE (49), the Epigenome Roadmap (27),BLUEPRINT (50), and the overarching InternationalHuman Epigenome Consortium are dedicated toproviding reference maps for key cellular states.These efforts also focus on harmonizing measure-ment techniques, bioinformatics standards, datamodels, and analytical tools for organizing, inte-grating, and displaying the epigenome data generated(51,52). With availability and easy use of analysistools, it is becoming increasingly feasible for manyresearchers to study the epigenome in differentdisease states (53). Reproducibility among labora-tories is also improving, with DNA methylationprofiling now benchmarked across many laboratories,and with assays potentially sufficiently robust tobecome applicable in clinical settings (54).

J A C C V O L . 7 0 , N O . 5 , 2 0 1 7 van der Harst et al.A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6 Epigenetics in Cardiovascular Disease

597

DNA METHYLATION

DNA hypomethylation and site-specific hyper-methylation are very common in human malig-nancies, but DNA methylation differences, bothhypomethylation and hypermethylation, have alsobeen linked to many other phenotypes, includingcardiovascular phenotypes.

Genome-wide DNA methylation profiling ofatherosclerotic versus normal human aorta revealedalterations in global DNA hypermethylation in bothCpG and non-CpG contexts. The identified locationshave been mapped to genes with known roles inthe vascular wall or atherosclerosis (55,56). Globallyhypomethylated chromosomal DNA is widely presentin atherosclerotic plaques (57). Prominent gene clus-ters linked to hypomethylation were recently identi-fied at chromosomes 9p21 and 14q32, and were linkedto several clustered, up-regulated miRNAs (57).However, a significant limitation with several of thesestudies is the tissues used. These are usually frag-mented specimens, consisting of a complex mixtureof cells. Cellular composition differed betweendiseased and healthy tissues, making it difficult tounderstand whether differences in observed methyl-ation profiles are simply driven by changes in cellularcomposition.

Several studies examined the relationship of DNAmethylation in heart failure through comparison ofmyocardial tissue from patients with cardiomyopa-thies and from normal hearts. Although these studiesare all characterized by remarkably small samplesizes, they report intriguing findings. For example, inan early study evaluating 8 cases with heart failurecompared with 6 control subjects, altered myocardialexpression of 3 angiogenesis-related genes (AMOTL2,PECAM1, and ARHGAP24) was associated withdifferential methylation in their gene body and 50

regions (58). In a further study by the same group,methylated DNA immunoprecipitation sequencing ofventricular myocardial tissue from 4 patients withheart failure and 4 control subjects revealed exten-sive differences in DNA methylation in promoter andintragenic CpG islands, in gene bodies, and also inH3K36me3-enriched genomic regions between thecase and control subjects. Reduced DNA methylationin gene promotor regions was globally associatedwith up-regulated gene expression; as proof of prin-ciple, experimental manipulation of methylation inthe mouse HL1 cell line up-regulated expression ofDux, the murine homologue of DUX4, the geneshowing the strongest differential methylation be-tween heart failure case and control subjects (59).

Disturbances of both DNA methylation and geneexpression in heart failure have been confirmed inseparate studies, supporting the view of epigenomiccontrol of expression of key genes in the develop-ment of, or myocardial response to, heart failure. Thecross-sectional and observational nature of the cur-rent human studies precludes interpretation ofwhether these changes are the cause or consequenceof heart failure and, if the consequence, whether theyare adaptive or maladaptive.

To gain further insights, experiments have beenperformed in several animal models. In mice sus-ceptible to isoproterenol-induced cardiac dysfunc-tion, differences in basal methylation pattern beforeenvironmental stress were predictive of disease pro-gression (60). Ablation of DNMT3A and DNMT3B(required for de novo methylation) in geneticallyengineered mice did not affect the phenotypicresponse to left ventricular pressure overloadinduced by transverse aortic constriction, but didinfluence transcriptional responses. This impliesthat de novo DNA methylation of cardiomyocytesmay not be essential in the response to chroniccardiac pressure overload (61).

Advancing our knowledge on specific environ-mental exposures may help us to understand thepotential role of targeted epigenetic interventions.Experiments in animals show that nicotinic stimula-tion of acetylcholine receptors inhibits myocardialdifferentiation by down-regulation of Tbx5 and Gata4in both differentiating embryonic bodies and inthe offspring’s heart, and this is accompanied bypromoter hypermethylation (44). Physical exerciseinfluences the DNA methylation profile of adiposetissue in apparently healthy individuals, including ata number of genes implicated in the development oftype 2 diabetes (62). Mitochondrial DNA methylationis inversely associated with metal-rich air exposure,and modifies the relationship between exposure toair pollution and heart rate variability outcomes (63).Toll-like receptor 2 (TLR2) methylation may alsoinfluence susceptibility of cardiac autonomic activityto short-term fine particle exposure (64). Theseintriguing findings support the view that exposuresknown to influence CVD may operate through DNAmethylation, and provide the justification for furtherwork to unravel the complex relationships involved.

HISTONE MODIFICATIONS

Global deletion of the histone deacetylases HDAC1 orHDAC3 results in embryonic death. When HDAC2 issilenced globally, it results in cardiac morphological

van der Harst et al. J A C C V O L . 7 0 , N O . 5 , 2 0 1 7

Epigenetics in Cardiovascular Disease A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6

598

abnormalities and death shortly after birth. However,when silencing of HDAC1 or HDAC2 is restricted tothe cardiomyocytes, it appears to have no directeffect on heart development (24). HDAC3 loss incardiomyocytes does not directly influence cardiacdevelopment, but eventually results in hypertrophy,as does HDAC2 overexpression (24). It is clear thatmultiple pathways are controlled by the activities ofdifferent isoforms of HDAC and that multiple re-dundancies exist. Histone deacetylases have beenlinked to atherosclerosis. HDAC2 has been suggestedto have a protective function, and is linked toimproved endothelial function (65). HDAC2 can bedown-regulated by oxidized low-density lipoprotein(LDL), resulting in increased oxidative stress due toendothelial nitric oxide synthase (eNOS) uncoupling.In areas prone to atherosclerosis, increased expres-sion of HDAC3 is observed and relates to flow dis-turbances. Knockdown of HDAC3 leads to reducedendothelial cell survival, more atherosclerosis, andeven rupture of vessels, implicating a role for HDAC3in maintenance of endothelial integrity (66).

HATs have also been linked to influence on thecardiovascular system. For instance, P300/CBP-associated factor is involved in arteriogenesis, thedevelopment of pre-existing collateral arterioles intolarger arteries (67). In a systematic evaluation of denovo mutations in 362 cases of severe congenitalheart disease, an excess of protein-altering de novomutations in genes expressed in the developing heartwere identified (68). Interestingly, a marked excess ofde novo mutations in genes involved in the produc-tion, removal, or reading of H3K4 methylation orubiquitination of H2BK120 (which is required forH3K4 methylation) and SMAD2 (which regulatesH3K27 methylation) was identified (68). These ob-servations suggest a central role for histone modifi-cations in the etiology of congenital heart disease(69).

Not only are the histone deacetylase enzymeslinked to cardiovascular function, but so are thehistone modifications themselves. The archetypalendothelial gene eNOS, which has a central role inendothelial biology, is governed by a unique epige-netic signature. The eNOS proximal promotor isenriched with H3K9Ac, H4K12Ac, H3K4me2,H3K4me3, and the H2A. Dynamic changes in thissignature are associated with the activation andrepression of eNOS in endothelial cells in response toenvironmental stimuli, especially hypoxia (70). Micedeficient for the H3K36 methyltransferase exhibitvascular remodeling defects and are embryonicallylethal, suggesting a role for H3K36me3 in vascularbiology (70,71).

Several studies have also mapped differences inhistones in human cardiomyopathies comparedwith control hearts. Comparison of 10 normal with 10dilated cardiomyopathy hearts identified differencesin 4 histone modifications (dimethyl-K4, dimethyl-K79, acetyl-K14, and acetyl-K27) (72). Some datasuggest that H3 modifications may be specific fordilated cardiomyopathies. For example, H3K9 deme-thylation and heterochromatin protein 1 dissociationin the promotor regions of the ANP/ B-type natriureticpeptide genes was associated with nuclear export ofHDAC4 and reactivation ANP gene expression inresponse to hemodynamic load and end-stage humanheart failure (73). The H3K36me3 (activator) mark wasdifferent in coding regions of the genome of humanend-stage cardiomyopathies (59).

Calcium/calmodulin-dependent protein kinase II,involved in cardiomyocyte calcium ion homeostasisand reuptake, has also been implicated in epigeneticchanges, linked to the phosphorylation of histone H3at serine-10 during pressure overload hypertrophy,and might provide a mechanistic link to reactivationof the fetal cardiac gene program (74). Induction ofthe fetal contractile protein gene program has beenlinked to the nuclear export of class II HDACs, whichallow activation of fetal cardiac genes via the tran-scription factor MEF2 (75). Acetylation/deacetylationevents seems to play an important role in the regu-lation of cardiac growth and programmed geneexpression in response to acute or chronic stressors(76). Also, histone trimethyl demethylase is up-regulated in patients with hypertrophic cardiomyop-athies, and experimental data from transgenic micesuggest that overexpression exacerbates the hyper-trophic response (77).

NONCODING RNA

Data from deep-sequencing studies in the humanheart have identified >1,000 miRNAs and >3,500lncRNAs that are expressed (78). Among ncRNAs,miRNAs represent the most broadly-studied species,and are reported to be involved in cardiogenesis,vascular and blood development, and a plethora ofCVDs and conditions (79). Among those miRNAsstudied in genetic and pharmacological deletionmodels and most intensively considered for thera-peutic development approaches are miR-133,miR-132, and miR-199b, which regulate various pro-hypertrophic intracellular signaling cascades andautophagy in heart muscle cells; the muscle-specificmiRNAs miR-208a, miR-208b, and miR-499, whichare embedded in myosin genes and are responsiblefor cardiac muscle contractions; miR-21, which

J A C C V O L . 7 0 , N O . 5 , 2 0 1 7 van der Harst et al.A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6 Epigenetics in Cardiovascular Disease

599

regulates extracellular signal-related kinase–mitogen-activated protein kinase signaling in fibro-blasts and controls cardiac fibrosis; and miR-92,which is involved in angiogenesis and functionalrecovery in ischemic heart conditions (44,80).

A further characterization of distinct lncRNAs andtheir relation to CVD development is only appearingto emerge. Examples include the Braveheart lncRNA,required for cardiovascular lineage commitment andessential for the central regulatory gene-networksinvolved in governing cardiovascular cell fate.Braveheart lncRNA also functions upstream of theMesP1 (mesoderm posterior) master gene involved incardiovascular lineage commitment. The lncRNAFendrr is required for lateral mesoderm lineage dif-ferentiation. Fendrr modifies chromatin gene signa-tures by binding to the trithorax group proteins/mixed lineage leukemia complex and polycombrepressive complex 2. An lncRNA transcript (Mhrt)originating from the Myh7 locus antagonizes Brg1function. Brg1 expression is triggered by stress, and itremodels chromatin to induce an aberrant geneexpression profile, resulting in a cardiac myopathyphenotype (80).

The myocardium transcriptome is regulateddynamically in heart failure and responds to unload-ing of the heart by left ventricular assist devicetreatment (78,81). Interestingly, lncRNA (78) andmiRNA (81) expression profiles are reported to moresensitively respond to treatment compared withmRNA. ncRNA normalization of expression appearsto be insufficient to result in complete normalizationof the mRNA signature. Conceptually, this might beof interest when considering therapeutic approachesusing ncRNAs.

The investigation of ncRNAs as circulating bio-markers in body fluids represents a new diagnosticpotential, where these molecules can be secreted inexosomes, microvesicles, or apoptotic bodies, asso-ciated with high-density lipoprotein and other lipo-protein complexes, or through passive leakage.Indeed, secreted ncRNAs appear stable, and can beeasily measured in blood, serum, cerebrospinal fluid,or urine. As such, miRNAs have been demonstrated toact as intercellular communicators, where miR-21* issecreted in exosomes derived from fibroblasts andaccumulates in cardiomyocytes under hypertrophicconditions (82), exosomal miR-146a derived fromendothelial cells accumulates in cardiomyocytes inperipartum cardiomyopathy (83), and exosomal miR-143/145 secreted by endothelial cells influencesmooth muscle cell function in atherosclerosis (84).Other ncRNAs distinguish patients with heart failurewith reduced or preserved ejection fraction (85) or

different forms of hypertrophic cardiomyopathies(86), or predict survival in patients with heart failure(87). Ongoing studies validate incidental findingsin multicenter meta-studies using multiplexedbiomarker combinations and focus on the develop-ment of clinical-grade molecular diagnostics detec-tion platforms.

EPIDEMIOLOGICAL PERSPECTIVE

Hundreds, or even thousands, of CVD- and trait-associated DNA (genetic) variants have been identi-fied in the last decade by genome-wide associationstudies (GWAS). Most of the identified genetic vari-ants are in the noncoding regions of the genome,often at a remarkable distance from known protein-coding genes. The major challenge is, therefore, todetermine the likely causal gene or the causal mech-anisms. Many cardiovascular GWAS variants havebeen associated with histone marks or chromatinstates (88,89). Interactions with distant marks havebeen reported to partly overlap (share) at least 1 oftheir interactions in 60%, but almost all (>99%) alsohave lineage- or cell-type–specific interactions (90).This suggests that higher-order genome structuresundergo coordinated remodeling during lineagespecification, dynamically reshaping transcriptionaldecisions (90). GWAS focused on electrically-activemyocardial mass, for example, have identified ge-netic loci that are strongly enriched for certain his-tone marks (H3K27ac, H3K4me3, H3K4me, andH3K36me3) and are associated with chromatin statesinvolved in active enhancers, promoters, and tran-scription in the human heart; in contrast, no enrich-ment was observed for transcriptionally-repressivehistone marks (H3K27me3 and H3K9me3) or states(89). Enrichment of activating histone marks was alsoidentified during differentiation of mouse embryonicstem cells into cardiomyocytes (89). However, thegenomic maps of epigenetic modifications that areused in GWAS are usually derived from cell culturesor cells isolated from a few individuals, an importantlimitation to their utility for understanding tissue-and cell-specific processes (52).

Population-based epidemiological studies areincreasingly exploring the role of epigenetic modifi-cation in the development of CVD, and it potentialrole as a mediator of the actions of DNA sequencevariation and environmental exposures. Driven byfeasibility, the most-studied epigenetic modificationis DNA methylation, occurring predominantly at thesymmetrical dinucleotide CpG. Studies of specificgenetic loci show that genetic variants influencingblood pressure are enriched for association with

van der Harst et al. J A C C V O L . 7 0 , N O . 5 , 2 0 1 7

Epigenetics in Cardiovascular Disease A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6

600

changes in DNA methylation at multiple nearby CpGsites, suggesting that DNA methylation may lie on theregulatory pathway linking sequence variation toblood pressure (91). Separate studies suggest thathypomethylation of long interspersed nucleotideelement-1 (a repetitive sequence found across thegenome) is linked to prevalent and even incidentheart disease and stroke (92).

Epigenomic research is now rapidly progressingfrom candidates epigenetic studies in a few individ-ual cells toward the investigation of epigenomicvariation in the population (93). Epigenome-wideassociation studies (EWAS) provide a systematicapproach, comparable to GWAS, focused on uncov-ering epigenetic variants associated with diseasesor phenotypes of interest by taking advantage ofmicroarrays capable of measuring DNA methylation atup to 850,000 sites (93). These EWAS have identifiedhundreds of CpGs, including HIF3A (hypoxia-inducible factor 3A), associated with body mass in-dex (94–96), ABCG1 (ATP-binding cassette subfamilyG member 1, PHOSPHO1 (phosphoethanolamine/phosphocholine phosphatase), SOCS3 (suppressorof cytokine signaling-3), SREBF1 (sterol regulatoryelement-binding protein 1), and TXNIP (thioredoxin-interacting protein) with risk of future type 2 diabetes(97). CPT1A (carnitine palmitoyltransferase 1a),ABCG1, TXNIP, and SREBF1 were associated with lipidlevels (98–102), as well as other traits, including cor-onary heart disease (98,103) and atrial fibrillation(104). Some of the CpGs were also confirmed in rele-vant tissues and linked to gene expression levels. Forexample, the CpGs of SREBF1, ABCG1, SREBF1, andHIF3A that were identified in blood were differentfrom those in adipose tissue and linked to geneexpression levels (94,96,102). Although these genescan be linked to biology, the reasons underlying thesedifferences are not known. A recent meta-analysis ofEWAS on lipids in 3,296 individuals, followed by astepwise Mendelian randomization analysis, sug-gested that differential methylation was induced bylipids and not by DNA methylation levels on lipids (1).The same strategy also recently resulted in the sug-gestion that DNA methylation associated with bodymass index is predominantly the consequence, ratherthan the cause of adiposity (96). Thus, lipid levels oradiposity might cause epigenetic changes in immunecells, which might provide a novel link to athero-sclerotic and other inflammatory diseases. In this re-gard, it is of interest to note that CpG changesassociated with body mass index (e.g., in the ABCG1locus) do affect the future risk of type 2 diabetes innormoglycemic individuals, which might help to

identify individuals at risk and indicate a role early inthe etiology (96,97).

Multiple challenges remain regarding the design,conduct, and interpretation of EWAS. Key challengesinclude accounting for (or removing) variation incellular composition, the potential confounding ef-fect of genetic background, and resolving whetherassociations identified represent causal, consequen-tial, or bystander effects. Although there are nowwell-validated approaches for taking the cell-typemixture into account in population-based studies onblood, proposed solutions for other tissues areunverified, and may remove both true and false sig-nals (93,105). Quantitative assessment of the contri-bution of epigenetic factors to transcription showsthat cis-acting DNA sequence variation explains themajority of transcriptional variances for the majorityof genes, with relatively few independent epigeneticinfluences for a small subset of biologically-relevantgenes (106). These findings suggest that cis-geneticvariation should also be considered in EWAS.

One key question is whether blood samples, whichare usually used in epidemiological studies, areindeed relevant for all targeted tissues, consideringthe epigenetic heterogeneity at many of the identifiedloci, including the heart. EWAS studies are beginningto explore this issue through robust analysis of tissue-specific, cell-specific, and cross-tissue variations inmethylation (96). Integration of the EWAS with GWASand Mendelian randomization analysis is also rapidlyemerging as a tool for further understanding howgenetics and epigenetics are linked (93,96). Longitu-dinal studies of changes in DNA methylation and theirrelationship to phenotypic pattern may also help tobetter understand casual relationships (96).

THERAPEUTIC PERSPECTIVE

The redundancy of epigenetic regulation suggeststhat a single “magic bullet” to epigenetically treat acardiovascular condition is utopian (76). However,the concept of targeting redundant systems isfamiliar to the cardiologist. Treating patients withangiotensin-converting enzyme inhibitors, andangiotensin II, beta-adrenergic, and aldosteroneblockers is also based on multiple “hubs” of myocar-dial remodeling.

DNA METHYLATION

DNA methylation depends on dietary methyldonors, including folic acid and vitamins, for whichinterventions could have an effect in certain in-dividuals (93). Other (dietary) compounds, such as

J A C C V O L . 7 0 , N O . 5 , 2 0 1 7 van der Harst et al.A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6 Epigenetics in Cardiovascular Disease

601

polyphenols and catechins, likewise influence DNAmethylation. Consuming cacao (a polyphenol-richsource) reduces global DNA methylation in circu-lating leukocytes of patients with cardiovascularrisk factors (Central Illustration) (107). Epigeneticeffects are induced by some common drugs, includinghydralazine, which interferes with DNA methyl-transferase and inhibits DNA methylation, or procai-namide, which inhibits DNA methyltransferase I.Their shared side effect of a lupus-like autoimmunedisease is indeed thought to be due to hypo-methylation. Although these drugs are consideredin cancer therapy, the epigenetic potential for CVDremains to be determined.

Cellular data and data derived from experiments inspontaneously hypertensive rats suggest that theDNA methylation inhibitor 5-azacytidine can reducethe detrimental effects of tumor necrosis factor-aon SECRA2a expression (10) and might block expres-sion of hypertrophic cardiomyopathy genes (108).5-azacytidine (Central Illustration) might improvecardiac hypertrophy, reduce cardiac fibrosis, andpreserve diastolic dysfunction (7). Similarly, thesetting of hypertrophy induced by norepinephrineis associated with hypermethylation by an increaseof DNMT activity of the left ventricle of rats (109).5-Aza-20-deoxycytidine (a DNMT inhibitor) causes areversal of the changes in the cardiac proteome,decreased hypertrophy, improved cardiac contrac-tility, and abrogated susceptibility to ischemic injuryin rats (109). Although beyond the scope of this re-view, epigenetic interventions might also be relevantto the induction of pluripotency and reprogrammingof induced pluripotent cells (110). In the future,epigenetic modifications of cardiomyocytes might bea target of precision medicine for CVD, possibly viadown-regulation of DNMTs.

HISTONES

Inhibitors of HDACs are potential compounds rele-vant to targeting cardiovascular conditions (CentralIllustration). Even nonspecific HDAC inhibitors thatinhibit deacetylation of both histones and nonhistoneproteins, affecting sarcomeric protein acetylation,showed a promising effect in treatment of cardiachypertrophy in animal models (111).

Considering its role in activating the fetal geneprogram in heart failure, pharmacological modulationof histone acetylation might also be an interestingtherapeutic strategy (76). Statins have been reportedto inhibit HDAC activity (112), and several HDACsmall-molecule inhibitors are used in the treatmentof several malignancies and some inflammatory

disorders. HDAC inhibitors used for heart diseasemight be of interest, as they exhibit antiapoptotic,antiautophagic, anti-inflammatory, and antifibroticcharacteristics, as suggested by animal experimentalstudies (76). For example, in cellular experiments,HDAC4 inhibitors appear to block hypertrophy andfetal gene activation in cardiomyocytes (113), andhave been reported to improve cardiac functionand suppress cardiac remodeling the mouse heart(5). Also, in the setting of ischemia/reperfusion injuryand myocardial infarction, HDAC inhibitors reducemyocardial infarct size and attenuate ventricularremodeling in animal studies (114,115) and induce anincreased angiogenic response (116). The pan-HDACinhibitor vorinostat is currently approved for treat-ment of cutaneous T-cell lymphoma, has beendemonstrated to reduce ischemia/reperfusion injuryin mice and rabbits, and might be closest to a proof-of-principle study in humans (24). However, otherHDAC inhibitors (e.g., specific to HDAC1, HDAC2, orHDAC3) also have been reported to exert possibledetrimental effects on, for example, vascular andendothelial function (65,66). In addition to inhibitorsof deacetylation, inhibitors of acetylation (HAT in-hibitors) might protect against ischemia/reperfusioninjury (117,118) (Central Illustration). Therefore, it ap-pears that a carefully-selected and likely cell-specifictarget of histone modification is important. To date,human trials in CVD are lacking, but more selectiveinhibitors that are aimed at avoiding systemic sideeffects are under development (119,120).

NONCODING RNA

Influencing miRNA expression could represent atherapeutic strategy for multiple cardiovascular con-ditions. For therapeutic purposes, 2 approaches existto alter ncRNA levels in disease settings. Dependingon whether their activity is harmful or beneficial,their function can either be enhanced or mimicked bydelivery strategies, or inhibited by targeting the RNAtranscript with antisense oligonucleotides.

For example, antisense oligonucleotide adminis-tration targeting endogenous miR-92a resulted ingreater blood vessel growth and the recovery ofinjured tissue after myocardial infarction (121).Antisense-induced silencing of miR-208a duringheart failure in rats caused by hypertension pre-vented cardiac remodeling and improved cardiacfunction and survival (122). Restoration of miR-1 geneexpression regressed cardiac hypertrophy in rats andled to a marked reduction of myocardial fibrosis (123).Silencing of Chast (cardiac hypertrophy-associatedtranscript) by a chimeric antisense oligonucleotide

CENTRAL ILLUSTRATION Potential Therapeutic Epigenetic Targets in Cardiovascular Disease

van der Harst, P. et al. J Am Coll Cardiol. 2017;70(5):590–606.

Examples of targeting epigenetic mechanisms in CVD. Possible targets include modifying DNA methylation, changing the acetylation or deacetylation of histones, and

miRNA or lncRNA modifications. CVD ¼ cardiovascular disease; DNA ¼ deoxyribonucleic acid; DNMT ¼ deoxyribonucleic acid methyltransferase; HAT ¼ histone

acetyltransferase; HDAC ¼ histone deacetylase; lncRNA ¼ long noncoding ribonucleic acid; mRNA ¼ messenger ribonucleic acid; miRNA ¼ microribonucleic acid;

RNA ¼ ribonucleic acid.

van der Harst et al. J A C C V O L . 7 0 , N O . 5 , 2 0 1 7

Epigenetics in Cardiovascular Disease A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6

602

(gapmer) prevented and regressed pressure overload–induced adverse cardiac remodeling in the mousewithout signs of side effects (124). Recently, phase 1data was published (6), and phase 2 clinical data waspresented on inclisiran (a chemically synthesizedsmall interfering RNA targeting PCSK9 mRNA). Inthe phase 2 trial, 497 patients were treated with aninvestigational N-acetylgalactosamine–conjugatedRNA interference therapeutic being developed forthe treatment of hypercholesterolemia that targetsproprotein convertase subtilisin/kexin type 9(PCSK9), a genetically-validated protein regulatorof LDL receptor metabolism. LDL cholesterol levels

were reduced by >50% with injections that appearto be sufficient when administered only once every 3to 6 months.

FUTURE DIRECTIONS

Many cardiovascular risk factors and traits havebeen linked to epigenetic changes. However, muchof the available data on epigenetics, especiallypopulation-based, is from studies of DNA methyl-ation, typically using blood as a convenient source ofDNA. Future research needs to focus on more rele-vant cell types and other epigenetic modifications.

J A C C V O L . 7 0 , N O . 5 , 2 0 1 7 van der Harst et al.A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6 Epigenetics in Cardiovascular Disease

603

Technological advancements and improvements indetection limits will also enable exploration oflow-abundancy and other modifications of DNAin humans (18).

A thorough understanding of the dynamic cross-talk between epigenetic factors, the genome, andthe environment is lacking. Creating an integratedand comprehensive network of the genome, epi-genome, expressome, metabolome, proteome, andphenome is on its way, and will be essential in ourunderstanding of health and disease. The rapidly-expanding data, cataloguing of high-quality refer-ence sets, and tools to interpret and analyze thedata will greatly facilitate future meaningful in-terpretations and the generation of novel hypotheseson how epigenetic changes occur and are mechanis-tically related to altered gene expression, and even-tually, the phenotype. Advances in technology anddevelopment of novel strategies for (epi)genomeediting (e.g., using CRISPR-based methods) will allowthese novel hypotheses to be efficiently tested. Theultimate goal of increasing our epigenetic knowledgewill be to translate this to developing rational in-terventions and then put these to the test in trials.There is a need to perform translational studies totest the effects of epigenetic modifications, both asprognosticators and as therapies. Recent data linkingmethylation changes to future onset of type 2 dia-betes will allow the design, in the near future, ofan interventional trial targeting individuals at highrisk (96). Examples of successful epigenetic therapiesare available in the cancer field. In addition to themore established RNA interference therapies, otherepigenetic strategies are to be developed in the

cardiovascular arena. The pan-HDAC inhibitor vor-inostat is already approved to treat cutaneous T-celllymphomas and might be able to reduce ischemia/reperfusion injury in both mice and rabbits. Vorino-stat might be the closest to a proof-of-principle studyin humans (24). However, specific therapies targetingspecific epigenetic processes have not been designedfor CVDs.

CONCLUSIONS

We are still far from truly understanding the epi-genomic “language” and how all elements and mod-ifications “communicate” with each other. Importantuncertainties remaining include the existence andsignificance of transgenerational, nongenetic orepigenetic, inheritance in the expression of pheno-types in humans. The view of the epigenome israpidly changing, with major new discoveries re-ported each year. Epigenomics is becoming a bigdata science, with a potentially enormous effect onour understanding of CVD. Although causalityremains to be established and conflicting evidenceexists, novel opportunities may be provided forboth diagnostic and therapeutic avenues. The effectof targeting epigenetic mechanisms appears prom-ising, but these therapies for CVD still await a longroad ahead.

ADDRESS FOR CORRESPONDENCE: Dr. Pim van derHarst, Department of Cardiology and Department ofGenetics, University of Groningen, University MedicalCenter Groningen, Hanzeplein 1, 9700RB Groningen,the Netherlands. E-mail: p.van.der.harst@umcg.nl.

RE F E RENCE S

1. Dekkers KF, van Iterson M, Slieker RC, et al.Blood lipids influence DNA methylation in circu-lating cells. Genome Biol 2016;17:138.

2. Saeed S, Quintin J, Kerstens HH, et al. Epige-netic programming of monocyte-to-macrophagedifferentiation and trained innate immunity. Sci-ence 2014;345:1251086.

3. Thornburg KL. The programming of cardiovas-cular disease. J Dev Orig Health Dis 2015;6:366–76.

4. van Otterdijk SD, Michels KB. Transgenerationalepigenetic inheritance in mammals: how good isthe evidence? FASEB J 2016;30:2457–65.

5. Chen Y, Du J, Zhao YT, et al. Histone deacety-lase (HDAC) inhibition improves myocardial func-tion and prevents cardiac remodeling in diabeticmice. Cardiovasc Diabetol 2015;14:99.

6. Fitzgerald K, White S, Borodovsky A, et al.A highly durable RNAi therapeutic inhibitor ofPCSK9. N Engl J Med 2017;376:41–51.

7. Watson CJ, Horgan S, Neary R, et al. Epigenetictherapy for the treatment of hypertension-induced cardiac hypertrophy and fibrosis.J Cardiovasc Pharmacol Ther 2016;21:127–37.

8. Squires JE, Patel HR, Nousch M, et al. Wide-spread occurrence of 5-methylcytosine in humancoding and non-coding RNA. Nucleic Acids Res2012;40:5023–33.

9. Domcke S, Bardet AF, Adrian Ginno P, Hartl D,Burger L, Schübeler D. Competition between DNAmethylation and transcription factors determinesbinding of NRF1. Nature 2015;528:575–9.

10. Kao YH, Chen YC, Cheng CC, Lee TI, Chen YJ,Chen SA. Tumor necrosis factor-a decreasessarcoplasmic reticulum Ca2þ-ATPase expressionsvia the promoter methylation in cardiomyocytes.Crit Care Med 2010;38:217–22.

11. Spruijt CG, Vermeulen M. DNA methylation: olddog, new tricks? Nat Struct Mol Biol 2014;21:949–54.

12. Maurano MT, Wang H, John S, et al. Role ofDNA methylation in modulating transcriptionfactor occupancy. Cell Rep 2015;12:1184–95.

13. Wu H, Zhang Y. Reversing DNA methylation:mechanisms, genomics, and biological functions.Cell 2014;156:45–68.

14. Juan D, Perner J, Carrillo de Santa Pau E, et al.Epigenomic co-localization and co-evolutionreveal a key role for 5hmC as a communicationhub in the chromatin network of ESCs. Cell Rep2016;14:1246–57.

15. Fu Y, Luo GZ, Chen K, et al. N6-methyldeox-yadenosine marks active transcription start sites inChlamydomonas. Cell 2015;161:879–92.

16. Zhang G, Huang H, Liu D, et al. N6-methyladenine DNA modification in Drosophila.Cell 2015;161:893–906.

17. Greer EL, Blanco MA, Gu L, et al. DNAmethylation on N6-adenine in C. elegans. Cell2015;161:868–78.

van der Harst et al. J A C C V O L . 7 0 , N O . 5 , 2 0 1 7

Epigenetics in Cardiovascular Disease A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6

604

18. Koziol MJ, Bradshaw CR, Allen GE, Costa AS,Frezza C, Gurdon JB. Identification of methylateddeoxyadenosines in vertebrates reveals diversityin DNA modifications. Nat Struct Mol Biol 2016;23:24–30.

19. Castegna A, Iacobazzi V, Infantino V. Themitochondrial side of epigenetics. Physiol Geno-mics 2015;47:299–307.

20. Turner BM. Reading signals on the nucleo-some with a new nomenclature for modifiedhistones. Nat Struct Mol Biol 2005;12:110–2.

21. Kinkley S, Helmuth J, Polansky JK, et al.reChIP-seq reveals widespread bivalency ofH3K4me3 and H3K27me3 in CD4þ memory T cells.Nat Comm 2016;7:12514.

22. Shema E, Jones D, Shoresh N, Donohue L,Ram O, Bernstein BE. Single-molecule decoding ofcombinatorially modified nucleosomes. Science2016;352:717–21.

23. Mathiyalagan P, Chang L, Du XJ, El-Osta A.Cardiac ventricular chambers are epigeneticallydistinguishable. Cell Cycle 2010;9:612–7.

24. Gillette TG, Hill JA. Readers, writers, anderasers: chromatin as the whiteboard of heartdisease. Circ Res 2015;116:1245–53.

25. Zhang CL, McKinsey TA, Chang S, Antos CL,Hill JA, Olson EN. Class II histone deacetylases actas signal-responsive repressors of cardiac hyper-trophy. Cell 2002;110:479–88.

26. Santos-Rosa H, Kirmizis A, Nelson C, et al.Histone H3 tail clipping regulates gene expression.Nat Struct Mol Biol 2009;16:17–22.

27. Kundaje A, Meuleman W, Ernst J, et al., for theRoadmap Epigenomics Consortium. Integrativeanalysis of 111 reference human epigenomes.Nature 2015;518:317–30.

28. de Dieuleveult M, Yen K, Hmitou I, et al.Genome-wide nucleosome specificity and functionof chromatin remodellers in ES cells. Nature 2016;530:113–6.

29. Djebali S, Davis CA, Merkel A, et al. Landscapeof transcription in human cells. Nature 2012;489:101–8.

30. Palazzo AF, Lee ES. Non-coding RNA: what isfunctional and what is junk? Front Genet 2015;6:2.

31. Kaikkonen MU, Lam MT, Glass CK. Non-codingRNAs as regulators of gene expression and epi-genetics. Cardiovasc Res 2011;90:430–40.

32. Jin H, Sun Y, Wang S, Cheng X. Matrine acti-vates PTEN to induce growth inhibition andapoptosis in V600EBRAF harboring melanomacells. Int J Mol Sci 2013;14:16040–57.

33. Krol J, Loedige I, Filipowicz W. The widespreadregulation of microRNA biogenesis, function anddecay. Nat Rev Genet 2010;11:597–610.

34. Jeck WR, Sorrentino JA, Wang K, et al.Circular RNAs are abundant, conserved, andassociated with ALU repeats. RNA 2013;19:141–57.

35. Khan MA, Reckman YJ, Aufiero S, et al. RBM20regulates circular RNA production from the titingene. Circ Res 2016;119:996–1003.

36. Rohrig H, Schmidt J, Miklashevichs E, Schell J,John M. Soybean ENOD40 encodes two peptides

that bind to sucrose synthase. Proc Natl Acad SciU S A 2002;99:1915–20.

37. Anderson DM, Anderson KM, Chang CL, et al.A micropeptide encoded by a putative long non-coding RNA regulates muscle performance. Cell2015;160:595–606.

38. Liu RJ, Long T, Li J, Li H, Wang ED. Structuralbasis for substrate binding and catalytic mecha-nism of a human RNA:m5C methyltransferaseNSun6. Nucleic Acids Res 2017 May 22 [E-pubahead of print].

39. Bonder MJ, Luijk R, Zhernakova DV, et al.Disease variants alter transcription factor levelsand methylation of their binding sites. Nat Genet2017;49:131–8.

40. McClay JL, Shabalin AA, Dozmorov MG, et al.High density methylation QTL analysis in humanblood via next-generation sequencing of themethylated genomic DNA fraction. Genome Biol2015;16:291.

41. Taudt A, Colomé-Tatché M, Johannes F.Genetic sources of population epigenomicvariation. Nat Rev Genetics 2016;17:319–32.

42. Gordon L, Joo JH, Andronikos R, et al.Expression discordance of monozygotic twins atbirth: effect of intrauterine environment and apossible mechanism for fetal programming. Epi-genetics 2011;6:579–92.

43. Fernandez-Twinn DS, Constância M,Ozanne SE. Intergenerational epigenetic inheri-tance in models of developmental programming ofadult disease. Semin Cell Dev Biol 2015;43:85–95.

44. Jiang XY, Feng YL, Ye LT, et al. Inhibition ofGata4 and Tbx5 by nicotine-mediated DNAmethylation in myocardial differentiation. StemCell Reports 2017;8:290–304.

45. Smith ZD, Chan MM, Mikkelsen TS, et al.A unique regulatory phase of DNA methylation inthe early mammalian embryo. Nature 2012;484:339–44.

46. Guo H, Zhu P, Yan L, et al. The DNA methyl-ation landscape of human early embryos. Nature2014;511:606–10.

47. Dolinoy DC, Weidman JR, Waterland RA,Jirtle RL. Maternal genistein alters coat color andprotects Avy mouse offspring from obesity bymodifying the fetal epigenome. Environ HealthPerspect 2006;114:567–72.

48. El Hajj N, Schneider E, Lehnen H, Haaf T.Epigenetics and life-long consequences of anadverse nutritional and diabetic intrauterine envi-ronment. Reproduction 2014;148:R111–20.

49. ENCODE Project Consortium. An integratedencyclopedia of DNA elements in the humangenome. Nature 2012;489:57–74.

50. Adams D, Altucci L, Antonarakis SE, et al.BLUEPRINT to decode the epigenetic signaturewritten in blood. Nat Biotechnol 2012;30:224–6.

51. Stunnenberg HG, Hirst M. The InternationalHuman Epigenome Consortium: a blueprint forscientific collaboration and discovery. Cell 2016;167:1145–9.

52. Bujold D, Morais DA, Gauthier C, et al. TheInternational Human Epigenome Consortium dataportal. Cell Syst 2016;3:496–9.e2.

53. Fernández JM, de la Torre V, Richardson D,et al. The BLUEPRINT data analysis portal. CellSyst 2016;3:491–5.e5.

54. BLUEPRINT Consortium. Quantitative com-parison of DNA methylation assays for biomarkerdevelopment and clinical applications. Nat Bio-technol 2016;34:726–37.

55. Zaina S, Heyn H, Carmona FJ, et al. DNAmethylation map of human atherosclerosis. CircCardiovasc Genet 2014;7:692–700.

56. Valencia-Morales Mdel P, Zaina S, Heyn H,et al. The DNA methylation drift of the athero-sclerotic aorta increases with lesion progression.BMC Med Genomics 2015;8:7.

57. Aavik E, Lumivuori H, Leppänen O, et al. GlobalDNA methylation analysis of human atheroscle-rotic plaques reveals extensive genomic hypo-methylation and reactivation at imprinted locus14q32 involving induction of a miRNA cluster. EurHeart J 2015;36:993–1000.

58. Movassagh M, Choy MK, Goddard M,Bennett MR, Down TA, Foo RS. Differential DNAmethylation correlates with differential expressionof angiogenic factors in human heart failure. PloSOne 2010;5:e8564.

59. Movassagh M, Choy MK, Knowles DA, et al.Distinct epigenomic features in end-stage failinghuman hearts. Circulation 2011;124:2411–22.

60. Chen H, Orozco LD, Wang J, et al. DNAmethylation indicates susceptibility toisoproterenol-induced cardiac pathology and isassociated with chromatin states. Circ Res 2016;118:786–97.

61. Nuhrenberg TG, Hammann N, Schnick T, et al.Cardiac myocyte de novo DNA methyltransferases3a/3b are dispensable for cardiac function andremodeling after chronic pressure overload inmice. PloS One 2015;10:e0131019.

62. Rönn T, Volkov P, Davegårdh C, et al. A sixmonths exercise intervention influences thegenome-wide DNA methylation pattern in humanadipose tissue. PLoS Genet 2013;9:e1003572.

63. Byun HM, Colicino E, Trevisi L, Fan T,Christiani DC, Baccarelli AA. Effects of air pollutionand blood mitochondrial DNA methylation onmarkers of heart rate variability. J Am Heart Assoc2016;5:e003218.

64. Zhong J, Colicino E, Lin X, et al. Cardiacautonomic dysfunction: particulate air pollutioneffects are modulated by epigenetic immunoreg-ulation of Toll-like receptor 2 and dietary flavo-noid intake. J Am Heart Assoc 2015;4:e001423.

65. Pandey D, Sikka G, Bergman Y, et al. Tran-scriptional regulation of endothelial arginase 2 byhistone deacetylase 2. Arterioscler Thromb VascBiol 2014;34:1556–66.

66. Zampetaki A, Zeng L, Margariti A, et al.Histone deacetylase 3 is critical in endothelialsurvival and atherosclerosis development inresponse to disturbed flow. Circulation 2010;121:132–42.

67. Bastiaansen AJ, Ewing MM, de Boer HC, et al.Lysine acetyltransferase PCAF is a key regulator ofarteriogenesis. Arterioscler Thromb Vasc Biol2013;33:1902–10.

J A C C V O L . 7 0 , N O . 5 , 2 0 1 7 van der Harst et al.A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6 Epigenetics in Cardiovascular Disease

605

68. Zaidi S, Choi M, Wakimoto H, et al. De novomutations in histone-modifying genes in congen-ital heart disease. Nature 2013;498:220–3.

69. Blue GM, Kirk EP, Giannoulatou E, et al.Advances in the genetics of congenital heartdisease: a clinician’s guide. J Am Coll Cardiol 2017;69:859–70.

70. Yan MS, Marsden PA. Epigenetics in thevascular endothelium: looking from a differentperspective in the epigenomics era. ArteriosclerThromb Vasc Biol 2015;35:2297–306.

71. Hu M, Sun XJ, Zhang YL, et al. Histone H3lysine 36 methyltransferase Hypb/Setd2 isrequired for embryonic vascular remodeling. ProcNatl Acad Sci U S A 2010;107:2956–61.

72. Koczor CA, Lee EK, Torres RA, et al. Detectionof differentially methylated gene promoters infailing and nonfailing human left ventriclemyocardium using computation analysis. PhysiolGenomics 2013;45:597–605.

73. Hohl M, Wagner M, Reil JC, et al. HDAC4controls histone methylation in response toelevated cardiac load. J Clin Invest 2013;123:1359–70.

74. Awad S, Al-Haffar KM, Marashly Q, et al.Control of histone H3 phosphorylation by CaMKIIdin response to haemodynamic cardiac stress.J Pathol 2015;235:606–18.

75. McKinsey TA, Zhang CL, Lu J, Olson EN.Signal-dependent nuclear export of a histonedeacetylase regulates muscle differentiation.Nature 2000;408:106–11.

76. DiSalvo TG. Epigenetic regulation in heartfailure: part II DNA and chromatin. Cardiol Rev2015;23:269–81.

77. Zhang QJ, Chen HZ, Wang L, Liu DP, Hill JA,Liu ZP. The histone trimethyllysine demethylaseJMJD2A promotes cardiac hypertrophy inresponse to hypertrophic stimuli in mice. J ClinInvest 2011;121:2447–56.

78. Yang KC, Yamada KA, Patel AY, et al. DeepRNA sequencing reveals dynamic regulation ofmyocardial noncoding RNAs in failing human heartand remodeling with mechanical circulatory sup-port. Circulation 2014;129:1009–21.

79. Small EM, Olson EN. Pervasive roles ofmicroRNAs in cardiovascular biology. Nature 2011;469:336–42.

80. Beermann J, Piccoli MT, Viereck J, Thum T.Non-coding RNAs in development and disease:background, mechanisms, and therapeutic ap-proaches. Physiol Rev 2016;96:1297–325.

81. Matkovich SJ, Van Booven DJ, Youker KA,et al. Reciprocal regulation of myocardial micro-RNAs and messenger RNA in human cardiomyop-athy and reversal of the microRNA signature bybiomechanical support. Circulation 2009;119:1263–71.

82. Bang C, Batkai S, Dangwal S, et al. Cardiacfibroblast-derived microRNA passenger strand-enriched exosomes mediate cardiomyocyte hy-pertrophy. J Clin Invest 2014;124:2136–46.

83. Halkein J, Tabruyn SP, Ricke-Hoch M, et al.MicroRNA-146a is a therapeutic target andbiomarker for peripartum cardiomyopathy. J ClinInvest 2013;123:2143–54.

84. Hergenreider E, Heydt S, Tréguer K, et al.Atheroprotective communication between endo-thelial cells and smooth muscle cells throughmiRNAs. Nat Cell Biol 2012;14:249–56.

85. Watson CJ, Gupta SK, O’Connell E, et al.MicroRNA signatures differentiate preservedfrom reduced ejection fraction heart failure. Eur JHeart Fail 2015;17:405–15.

86. Roncarati R, Viviani Anselmi C, Losi MA, et al.Circulating miR-29a, among other up-regulatedmicroRNAs, is the only biomarker for both hyper-trophy and fibrosis in patients with hypertrophiccardiomyopathy. J Am Coll Cardiol 2014;63:920–7.

87. Kumarswamy R, Bauters C, Volkmann I, et al.Circulating long noncoding RNA, LIPCAR, predictssurvival in patients with heart failure. Circ Res2014;114:1569–75.

88. Maurano MT, Humbert R, Rynes E, et al. Sys-tematic localization of common disease-associated variation in regulatory DNA. Science2012;337:1190–5.

89. van der Harst P, van Setten J, Verweij N, et al.52 Genetic loci influencing myocardial mass. J AmColl Cardiol 2016;68:1435–48.

90. Javierre BM, Burren OS, Wilder SP, et al.Lineage-specific genome architecture links en-hancers and non-coding disease variants to targetgene promoters. Cell 2016;167:1369–84.e19.

91. Kato N, Loh M, Takeuchi F, et al. Trans-ancestry genome-wide association study identifies12 genetic loci influencing blood pressure andimplicates a role for DNA methylation. Nat Genet2015;47:1282–93.

92. Baccarelli A, Wright R, Bollati V, et al. Ischemicheart disease and stroke in relation to blood DNAmethylation. Epidemiology 2010;21:819–28.

93. Zhong J, Agha G, Baccarelli AA. The role ofDNA methylation in cardiovascular risk and dis-ease: methodological aspects, study design, anddata analysis for epidemiological studies. Circ Res2016;118:119–31.

94. Dick KJ, Nelson CP, Tsaprouni L, et al. DNAmethylation and body-mass index: a genome-wideanalysis. Lancet 2014;383:1990–8.

95. Demerath EW, Guan W, Grove ML, et al.Epigenome-wide association study (EWAS) of BMI,BMI change and waist circumference in AfricanAmerican adults identifies multiple replicated loci.Human Mol Genet 2015;24:4464–79.

96. Wahl S, Drong A, Lehne B, et al. Epigenome-wide association study of body mass index, andthe adverse outcomes of adiposity. Nature 2017;541:81–6.

97. Chambers JC, Loh M, Lehne B, et al.Epigenome-wide association of DNA methylationmarkers in peripheral blood from Indian Asians andEuropeans with incident type 2 diabetes: a nestedcase-control study. Lancet Diabetes Endocrinol2015;3:526–34.

98. Hedman Å, Mendelson MM, Marioni RE, et al.Epigenetic patterns in blood associated with lipidtraits predict incident coronary heart diseaseevents and are enriched for results from genome-wide association studies. Circ Cardiovasc Genet2017;10:e001487.

99. Frazier-Wood AC, Aslibekyan S, Absher DM,et al. Methylation at CPT1A locus is associatedwith lipoprotein subfraction profiles. J Lipid Res2014;55:1324–30.

100. Gagnon F, Aïssi D, Carrié A, Morange PE,Trégouët DA. Robust validation of methylationlevels association at CPT1A locus with lipid plasmalevels. J Lipid Res 2014;55:1189–91.

101. Irvin MR, Zhi D, Joehanes R, et al.Epigenome-wide association study of fastingblood lipids in the Genetics of Lipid-loweringDrugs and Diet Network study. Circulation 2014;130:565–72.

102. Pfeiffer L, Wahl S, Pilling LC, et al. DNAmethylation of lipid-related genes affects bloodlipid levels. Circ Cardiovasc Genet 2015;8:334–42.

103. Lu TP, Chuang NC, Cheng CY, et al. Genome-wide methylation profiles in coronary arteryectasia. Clin Sci (Lond) 2017;131:583–94.

104. Lin H, Yin X, Xie Z, et al. Methylome-wideassociation study of atrial fibrillation in Framing-ham Heart Study. Sci Rep 2017;7:40377.

105. Lehne B, Drong AW, Loh M, et al. A coherentapproach for analysis of the Illumina Human-Methylation450 BeadChip improves data qualityand performance in epigenome-wide associationstudies. Genome Biol 2015;16:37.

106. Chen L, Ge B, Casale FP, et al. Genetic driversof epigenetic and transcriptional variation in hu-man immune cells. Cell 2016;167:1398–414.e24.

107. Crescenti A, Solà R, Valls RM, et al. Cocoaconsumption alters the global DNA methylation ofperipheral leukocytes in humans with cardiovas-cular disease risk factors: a randomized controlledtrial. PloS One 2013;8:e65744.

108. Fang X, Robinson J, Wang-Hu J, et al. cAMPinduces hypertrophy and alters DNA methylationin HL-1 cardiomyocytes. Am J Physiol Cell Physiol2015;309:C425–36.

109. Xiao D, Dasgupta C, Chen M, et al. Inhibitionof DNA methylation reverses norepinephrine-induced cardiac hypertrophy in rats. CardiovascRes 2014;101:373–82.

110. Tonge PD, Corso AJ, Monetti C, et al. Diver-gent reprogramming routes lead to alternativestem-cell states. Nature 2014;516:192–7.

111. Gupta MP, Samant SA, Smith SH, Shroff SG.HDAC4 and PCAF bind to cardiac sarcomeres andplay a role in regulating myofilament contractileactivity. J Biol Chem 2008;283:10135–46.

112. Lin YC, Lin JH, Chou CW, Chang YF, Yeh SH,Chen CC. Statins increase p21 through inhibition ofhistone deacetylase activity and release ofpromoter-associated HDAC1/2. Cancer Res 2008;68:2375–83.

113. Antos CL, McKinsey TA, Dreitz M, et al. Dose-dependent blockade to cardiomyocyte hypertro-phy by histone deacetylase inhibitors. J Biol Chem2003;278:28930–7.

114. Lee TM, Lin MS, Chang NC. Inhibition of his-tone deacetylase on ventricular remodeling ininfarcted rats. Am J Physiol Heart Circ Physiol2007;293:H968–77.

115. Xie M, Kong Y, Tan W, et al. Histone deace-tylase inhibition blunts ischemia/reperfusion injury

van der Harst et al. J A C C V O L . 7 0 , N O . 5 , 2 0 1 7

Epigenetics in Cardiovascular Disease A U G U S T 1 , 2 0 1 7 : 5 9 0 – 6 0 6

606

by inducing cardiomyocyte autophagy. Circulation2014;129:1139–51.

116. Zhang L, Qin X, Zhao Y, et al. Inhibition ofhistone deacetylases preserves myocardial per-formance and prevents cardiac remodelingthrough stimulation of endogenous angiomyo-genesis. J Pharmacol Exp Ther 2012;341:285–93.

117. Wongcharoen W, Phrommintikul A. The pro-tective role of curcumin in cardiovascular diseases.Int J Cardiol 2009;133:145–51.

118. Liu H, Wang C, Qiao Z, Xu Y. Protective effectof curcumin against myocardium injury in ischemiareperfusion rats. Pharm Biol 2017;55:1144–8.

119. McKinsey TA. Isoform-selective HDAC in-hibitors: closing in on translational medicine forthe heart. J Mol Cell Cardiol 2011;51:491–6.

120. Narita K, Matsuhara K, Itoh J, et al. Syn-thesis and biological evaluation of novelFK228 analogues as potential isoform selectiveHDAC inhibitors. Eur J Med Chem 2016;121:592–609.

121. Bonauer A, Carmona G, Iwasaki M, et al.MicroRNA-92a controls angiogenesis and func-tional recovery of ischemic tissues in mice. Science2009;324:1710–3.

122. Montgomery RL, Hullinger TG, Semus HM,et al. Therapeutic inhibition of miR-208a improves

cardiac function and survival during heart failure.Circulation 2011;124:1537–47.

123. Karakikes I, Chaanine AH, Kang S, et al.Therapeutic cardiac-targeted delivery of miR-1reverses pressure overload-induced cardiac hy-pertrophy and attenuates pathological remodel-ing. J Am Heart Assoc 2013;2:e000078.

124. Viereck J, Kumarswamy R, Foinquinos A,et al. Long noncoding RNA Chast promotes cardiacremodeling. Sci Transl Med 2016;8:326ra22.

KEY WORDS EWAS, HAT, HDAC, histones,methylation, RNA