genetics and genomics of single-gene cardiovascular diseases · genetics and genomics of...

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 AM 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 VO L . 6 8 , N O . 2 5 , 2 0 1 6

ª 2 0 1 6 B Y T H E AM 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 F O UN 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 6 . 0 9 . 9 6 8

FOCUS SEMINAR: GENETICS

STATE-OF-THE-ART REVIEW

Genetics and Genomics ofSingle-Gene Cardiovascular DiseasesCommon Hereditary Cardiomyopathies as Prototypesof Single-Gene Disorders

Ali J. Marian, MD,a Eva van Rooij, PHD,b,c Robert Roberts, MDd

ABSTRACT

Fro

Sci

UtdU

Th

Ma

This is the first of 2 review papers on genetics and genomics appearing as part of the series on “omics.” Genomics pertains

to all components of an organism’s genes, whereas genetics involves analysis of a specific gene or genes in the context of

heredity. The paper provides introductory comments, describes the basis of human genetic diversity, and addresses the

phenotypic consequences of genetic variants. Rare variants with large effect sizes are responsible for single-gene

disorders, whereas complex polygenic diseases are typically due to multiple genetic variants, each exerting a modest

effect size. To illustrate the clinical implications of genetic variants with large effect sizes, 3 common forms of

hereditary cardiomyopathies are discussed as prototypic examples of single-gene disorders, including their genetics,

clinical manifestations, pathogenesis, and treatment. The genetic basis of complex traits is discussed in a separate paper.

(J Am Coll Cardiol 2016;68:2831–49) © 2016 by the American College of Cardiology Foundation.

T he human nuclear genome is composed of3.2 billion base pairs and 20,576 protein-coding genes, which are arranged in 22 pairs

of somatic and a pair of sex (X and Y) chromosomes(NCBI Homo sapiens Annotation Release 107). Chro-mosome 1 is the largest chromosome, containingapproximately w249 million base pairs and about4,000 genes (Assembly hg38, UCSC Genome Browser,University of California, Santa Cruz, California). Chro-mosome 21 is the smallest, with about 48 million basepairs and 250 protein-coding genes. Over 90% of thegenome is transcribed, predominantly into noncod-ing ribonucleic acids (ncRNAs), and only w1% of thegenome is translated into protein.

The nuclear genome also contains about 18,000genes that are transcribed into ncRNAs. The ncRNAsare transcribed from active chromatin, poly-adenylated, and capped, but typically are not

m the aCenter for Cardiovascular Genetics, Brown Foundation Institute of

ence Center, and Texas Heart Institute, Houston, Texas; bHubrecht Insti

recht, the Netherlands; cDepartment of Cardiology, University Medical

niversity of Arizona College of Medicine, Phoenix, Arizona. Dr. van Ro

erapeutics Inc. All other authors have reported that they have no relationsh

nuscript received August 2, 2016; revised manuscript received Septembe

translated into proteins. The ncRNAs are commonlyclassified into small ncRNAs, which are usually <200nucleotides in length, and long noncoding ribonucleicacids (lncRNAs), which are >200 nucleotides long.Microribonucleic acids (miRNAs) are the best-studiedncRNAs. They are initially transcribed as longertranscripts, and then are cleaved into mature22-nucleotide-long miRNAs. miRNAs repress geneexpression by binding to a recognition sequence,typically within the 30 untranslated regions of targetmessenger ribonucleic acids (mRNAs). lncRNAsregulate gene expression through a broad range ofeffects, including forming complexes with proteinsand as sponges for other transcripts.

The human genome also contains pseudogenes,which no longer code for proteins. In addition,repetitive deoxyribonucleic acid (DNA) elementsoccupy more than 50% of the human nuclear genome.

Molecular Medicine, The University of Texas Health

tute, KNAW and University Medical Center Utrecht,

Center Utrecht, Utrecht, the Netherlands; and the

oij has served on the advisory board for miRagen

ips relevant to the contents of this paper to disclose.

r 14, 2016, accepted September 19, 2016.

https://s3.amazonaws.com/ADFJACC/JACC6825/JACC6825_fustersummary_08




http://crossmark.crossref.org/dialog/?doi=10.1016/j.jacc.2016.09.968&domain=pdf

http://dx.doi.org/10.1016/j.jacc.2016.09.968

ABBR EV I A T I ON S

AND ACRONYMS

AC = arrhythmogenic

cardiomyopathy

ARVC = arrhythmogenic right

ventricular cardiomyopathy

DCM = dilated cardiomyopathy

GV = genetic variant

HCM = hypertrophic

cardiomyopathy

lncRNA = long noncoding

ribonucleic acid

LOF = loss-of-function

miRNA = microribonucleic acid

SNV = single-nucleotide

variant

Marian et al. J A C C V O L . 6 8 , N O . 2 5 , 2 0 1 6

Genetics of Single-Gene Disorders D E C E M B E R 2 7 , 2 0 1 6 : 2 8 3 1 – 4 9

2832

Functions of the repetitive elements andnoncoding regions of the nuclear genome arelargely unknown.

DNA is wrapped around a set of octomericprotein complexes comprising 4 highlyconserved core (H2A, H2B, H3, and H4) his-tones, and 2 linker (H1 and H5) histones(Central Illustration). This compacted DNAand protein complex is referred to as chro-matin. Histones pack the DNA into units ofapproximately 150 base pairs, referred to asnucleosomes. Tightly packed nucleosomesare referred to as heterochromatin, and areinaccessible to the transcription machinery(inactive transcription). In contrast, looselypacked nucleosomes, referred to as euchro-

matin, are accessible to the transcription machinery,and hence are actively transcribed.

Histones undergo extensive modifications,including acetylation and methylation, which regu-late the chromatin open and closed states, and henceaccess of the transcription machinery to DNA. Forexample, trimethylation of lysine residue 27 on his-tone H3 (H3K27me3) is a chromatin marker for sup-pression of gene expression. In contrast, acetylationof the same residue (H3K27ac) marks the chromatinfor active transcription. Given the large number ofresidues that could undergo post-translational mod-ifications and various forms of modification, histonesare considered major regulators of gene expression.Post-translational histone changes and chemicalmodifications to DNA (not nucleotide changes), suchas CpG methylation, are collectively considered theepigenome.

STRUCTURE OF A GENE

About 1% of the genome, containing approximately30 million base pairs, codes for proteins. Eachprotein-coding gene has a 50 transcriptional regula-tory region; protein-coding segments referred to asexons; intervening regions between exons, calledintrons; and a 30 untranslated region or regulatoryregion (Central Illustration). Proteins that bind to en-hancers, silencers, and promoters proximal to thetranscription initiation site regulate transcription.The primary transcripts of genes are spliced toexclude introns and produce mRNAs. Splicing of eachprimary transcript is not uniform, and often multiplesplice variants are generated, of which 1 is the pre-dominant isoform. Each unit of 3 bases, referred to asa codon, encodes a specific amino acid. There are 61amino acid codons and 3 stop codons in the nucleargenome. Thus, each amino acid has multiple codons.

MITOCHONDRIAL GENOME

Each cell also contains mitochondrial deoxy-ribonucleic acid (mtDNA), a circular DNA composed of16,700 nucleotides. The mitochondrial genome con-tains 37 genes, which code for 13 proteins, 2 ribosomalRNAs, and 22 transfer RNAs. Each cell contains a largenumber of mitochondria, and each mitochondriontypically contains several copies of mtDNA. The co-dons are largely identical between the nuclear andmitochondrial genomes, except that there are 60 co-dons for amino acids and 4 for stop codons in themtDNA. In addition, transcription of mtDNA iscontinuous, as opposed to the discontinuous tran-scription of the nuclear genome.

GENERATION OF GENETIC VARIANTS

The replication machinery introduces rare errorsduring each round of DNA replication, which areempirically calculated to occur at a rate of w1.1 � 10�8

per base pair per generation (1–3). Given the size ofthe human genome, each DNA replication (meiosis)introduces about 40 to 60 new genetic variants (GVs).Accordingly, each newborn adds 40 to 60 new GVs tothe human genetic pool as de novo variants (i.e., ab-sent in the parents). The explosive growth of thehuman population during the last thousand years orso has introduced a massive number of GVs into thepopulation genetic pool, rendering humans exceed-ingly diverse at the genetic level (Table 1). The mu-tation rate of the mtDNA is several orders ofmagnitude higher, likely because of higher oxidativestress, and compromised function of the DNA repli-cation and repair system.

Each nuclear genome contains approximately 4million GVs, of which w3.5 million are single-nucleotide variants (SNVs), also referred to assingle-nucleotide polymorphisms, and several thou-sand small insertions/deletions, referred to as indels(4–7). In addition, the nuclear genome contains largeinsertions, deletions, and rearrangements, which arereferred to as structural variants. Structural variantsthat increase or decrease the number of chromosomesegments or genes are referred to as copy numbervariants. The vast majority of variants in eachgenome are rare. In addition, rare variants are typi-cally population-specific and hence, vary significantlyamong people with different ethnic backgrounds.

Approximately 12,000 SNVs change the amino acidsequence and are referred to as nonsynonymousSNVs. Computational programs that incorporatepopulation frequencies of the variants and evolu-tionary conservation of the involved codons, among

CENTRAL ILLUSTRATION Basic Structure of the Genome

Marian, A.J. et al. J Am Coll Cardiol. 2016;68(25):2831–49.

The nuclear genome is organized into chromosomes and wrapped in histone proteins, condensed into chromatin, and arranged in segments referred to as nucleosomes.

Histone modifications affect the chromatin state and gene expression. Each gene contains a 50 regulatory or untranslated region (UTR) where the transcription factors

bind and initiate transcription of the primary transcript. Each gene also contains exons and the intervening regions referred to as introns. Splicing machinery binds to

exon-intron boundaries and splices out introns from the coding transcript, referred to as messenger ribonucleic acid (mRNA). The first coding exon contains the signal

for initiation of translation, namely ATG, and the last coding exon contains a translation termination codon, such as TAG (UAG in the mRNA). The 30 regulatory region

or 30 UTR contains 1 or more polyadenylation signals, which signal for adding a poly-A tail to the mRNA. It is also the binding site for microribonucleic acids (miRNAs).

miRNAs and long noncoding ribonucleic acids (lncRNAs) are transcribed from introns or intergenic regions. SNP ¼ single-nucleotide polymorphism.

J A C C V O L . 6 8 , N O . 2 5 , 2 0 1 6 Marian et al.D E C E M B E R 2 7 , 2 0 1 6 : 2 8 3 1 – 4 9 Genetics of Single-Gene Disorders

2833

others, predict the presence of several thousands offunctional nonsynonymous SNVs in each genome. Asubgroup of GVs that result in gain or loss of a stopcodon, aberrant splicing, or frameshift are consideredloss-of-function (LOF) variants. They are expected toimpart biological effects, albeit not all cause clinicallydiscernible diseases. Each genome contains a couple

of hundred heterozygous and a few homozygous LOFvariants.

EFFECT SIZE OF DNA GENETIC VARIANTS

The effect of a GV on the phenotype (effect size) fol-lows a gradient ranging from none or indiscernible to

TABLE 1 Genetic Variations in the Human Genome

DNA sequence variants 4 � 106

SNVs 3.5 � 106

Nonsynonymous SNVs 12,000

LOF variants 500

Stop codon variants 100

Functional variants Several thousands

Pathogenic variants Few thousands

SVs/CNVs 104–105

CNV ¼ copy number variants; DNA ¼ deoxyribonucleic acid; LOF ¼ loss offunction; SNV ¼ single nucleotide variants; SV ¼ structural variants.

FIGURE 1 Effect Si

RareSmall

Large

Effec

t siz

e of

the

varia

nt

SingDis

The population frequ

more likely to impart

small or clinically ne

single-gene diseases

oligogenic diseases,

complex phenotypes



2834

large and clinically consequential (Figure 1). The vastmajority of the 4 million GVs do not seem to exertbiologically or clinically discernible effects. A smallfraction of the GVs, typically rare in the populationgenome, exerts clinically detectable effects. Theyaccount for the rare, single-gene diseases. GVs on theextreme end of large phenotypic effects are respon-sible for the single-gene disorders that exhibit Men-delian patterns of inheritance. Complex andpolygenic disorders are primarily due to a large

ze of GVs and Clinical Phenotype

Population frequency of the genetic variants Common

le Geneorders

OligogenicTraits

Complex Traits

ency of GVs is depicted against their effect sizes. Rare variants are

larger effect sizes than common variants, which typically impart

gligible effect sizes. Rare variants with very large effect sizes cause

, multiple uncommon variants with moderate effect sizes cause

and a very large number of common variants are responsible for

.

number of common GVs, each exerting small andoften indiscernible effects, but collectively, andthrough interactions with nongenetic factors, theyinfluence the risk of the complex phenotype.

CATEGORIZATION OF GVs FOR CLINICAL USE. Toenhance the clinical utility of GVs, various algorithms,typically on the basis of the predicted effects of the GVson protein structure and function, are used. It isimportant to note that none of the existing algorithmscan accurately predict the pathogenicity of GVs, and itis best to use multiple in silico programs. It is alsoimportant to recognize that the biological and clinicalsignificance of the GVs are context-dependent, couldvary among different genetic backgrounds, and areinfluenced by other GVs, as well as by environmentalfactors. This complexity poses significant challengesto clinical application of GVs.

As a general guide, LOF variants, which are definedas GVs resulting in gain or loss of a stop codon,aberrant splicing of transcripts, or frameshifts in theprotein sequence, are considered to be more patho-genic. This category of GVs has the highest risk ofexerting biologically and clinically discernible phe-notypes. In general, a rare variant, typically definedas a variant with a frequency of <1% in the popula-tion, is more likely to be pathogenic than a commonvariant. Missense variants could also result in gain- orloss-of-function, and are expected to be less patho-genic than stop codon, splice, or frameshift variants.Synonymous variants, as well as variants in introns,are considered generally to be nonpathogenic. Thereare, however, a number of exceptions. A notable oneis a synonymous variant in the LMNA gene that re-sults in aberrant splicing of lamin A/C and is respon-sible for Hutchinson-Gilford progeria syndrome (8).Finally, variants in intergenic regions are generallyconsidered clinically inconsequential, albeit thoseaffecting ncRNAs might impart biological functions. Asubset of functional variants that result in survivaland reproduction disadvantages is considereddeleterious.

GVs are classified functionally as pathogenic, likelypathogenic, of uncertain significance, likely benign,and benign. In the context of single-gene disordersand from a clinical point of view, GVs can be catego-rized as disease-causing, probably disease-causing,disease-associated, functional variants not associ-ated with disease, and variants with unknown func-tion or clinical significance (Table 2). Humanmolecular genetic studies offer the most robust evi-dence of disease causality of the GVs (9–11). Func-tional and mechanistic studies in model organismsprovide supportive evidence (9–11).

TABLE 2 Clinical Classification of GVs

Category Definition

Disease-causing variants LOF and pathogenic variants conclusively shown to cause thedisease through cosegregation and linkage analysis inlarge families.

Probably (likely) disease-causingvariants

Evidence for the causal role is on the basis of statisticalenrichment in small families and trios with the disease.Robust linkage is hindered by the small size of families orsporadic nature of the disease. To reduce the possibilityof random cosegregation in small families, findingsrequire testing for replication in independentpopulations.

Disease-associated variants Case–control studies show an association between GVs andthe phenotype. Replication in an independent studypopulation is necessary. Disease-associated variantsmight be in linkage disequilibrium with the actualpathogenic variants, requiring additional studies toidentify the actual pathogenic variants.

Functional variants notassociated with a disease

This class of functional variants might affect expressionlevels, structure, and function of the respective proteins,but they have not been associated with a disease.

Variants with no knownbiological or clinicalsignificance

This category comprises the vast majority of the 4 millionGVs in each genome.

GV ¼ genetic variant; LOF ¼ loss of function.


2835

GENETIC VARIANTS IN NONCODING RNA. Most ofthe GVs associated with complex diseases throughgenome-wide association studies map to noncodingregions. Variants mapped by a genome-wide associ-ation study might interfere with expression of aneighboring, causative protein-coding gene. Suchvariants might reside in ncRNAs and affect expressionby disrupting pairing of the ncRNAs with the targetRNAs. Recent data suggest that functional elementsin lncRNAs have a much lower variation frequency,almost comparable to protein-coding exons, indi-cating their potential biological significance (12).

GENETIC AND GENOMIC BASIS OF SELECTED

SINGLE-GENE DISORDERS. Single-gene disordersconstitute a group of genetic diseases caused by raremutations (<1% of the population). Accordingly, asingle mutation is sufficient to cause the phenotype,typically with an age-dependent penetrance. Thus,this group of genetic disorders represents theextreme of the effect size of the GVs, as opposed tothe complex diseases, which are often due to a verylarge number of common variants with very smalleffect sizes (Figure 1). Given the large effect sizes ofthe causal mutations, single-gene disorders typicallyexhibit Mendelian patterns of inheritance, namelydominant, recessive, or X-linked. The phenotype insingle-gene disorders is also influenced by a largenumber of noncausal GVs, which are referred to asmodifier variants, as well as by nongenetic environ-mental factors. Hereditary cardiomyopathies repre-sent prototypic forms of single-gene disorders.

GENETIC BASIS OF

HEREDITARY CARDIOMYOPATHIES

Hereditary cardiomyopathies comprise a group ofsingle-gene disorders wherein the primary defect is incardiac myocytes. The common forms of cardiomy-opathies, classified according to their phenotypicexpression, include hypertrophic cardiomyopathy(HCM), dilated cardiomyopathy (DCM), arrhythmo-genic right ventricular cardiomyopathy (ARVC), andrestrictive cardiomyopathy (Figure 2). The cardinalmanifestation in a subset of primary DCM is ventric-ular arrhythmias occurring early, disproportionately,and even prior to cardiac dysfunction. This subset isreferred to as arrhythmogenic DCM. ArrhythmogenicDCM together with classic ARVC comprise the broadercategory of arrhythmogenic cardiomyopathy (AC).The genetic basis of 3 common forms of hereditarycardiomyopathies is discussed.

GENETIC BASIS OF HCM. HCM is defined by thepresence of cardiac and typically left ventricularhypertrophy (LVH), occurring in the absence of a

known secondary cause, such as hypertension oraortic stenosis, in conjunction with normal globalcardiac systolic function (13). A left ventricular wallthickness $13 mm in adults and a z-score >2 in chil-dren are used to define LVH. A wall thickness $15 mmoffers a higher specificity, but reduces sensitivity ofdetection. As defined in the previous text, HCM is arelatively common disease with an estimated preva-lence of approximately 1 in 500 in young adults (14).Given the age-dependent phenotypic expression ofLVH, the population frequency of HCM might behigher in an older population.

LVH is commonly concentric, involving the inter-ventricular septum and left ventricular posterior andlateral walls. In about one-third of the patients, hy-pertrophy predominantly involves the interventric-ular septum, leading to asymmetric septalhypertrophy. Less commonly, hypertrophy also in-volves the right ventricle. Phenotypic expression ofcardiac hypertrophy is age-dependent, and acceler-ates during puberty and adolescence. Cardiac hyper-trophy typically manifests by the third and fourthdecades of life. De novo cardiac hypertrophy rarelydevelops in individuals older than 60 years. Thecardiac apex is the predominant site of involvementin a minority of cases, resulting in apical HCM.Involvement of cardiac apex, whether isolated or anextension of septal and lateral wall hypertrophy,leads to deep T-wave inversion in the precordialleads, which is characteristic of apical HCM. Myocar-dial tissue Doppler velocities are typically reduced inpatients with HCM and also in mutation carriers,

FIGURE 2 Common Forms of Primary Cardiomyopathies

Primary Cardiomyopathies

HCM DCM ARVC RCM Unclassified

Classic DCM ArrhythmogenicDCM

AC

The diagram illustrates common forms of primary cardiomyopathies, including

hypertrophic cardiomyopathy (HCM), dilated cardiomyopathy (DCM), arrhythmogenic

right ventricular cardiomyopathy (ARVC), and restricted cardiomyopathy (RCM). A subset

of DCM predominantly manifests with ventricular arrhythmias and is referred to as

arrhythmogenic DCM. The latter category, along with classic ARVC, encompasses

arrhythmogenic cardiomyopathy (AC).



2836

sometimes referred to as genotype positive-phenotype negative individuals (15–17). In the famil-ial setting, tissue Doppler imaging might help inidentification of the mutation carrier (16,17).

Cardiac myocyte disarray is the pathological char-acteristic of HCM and typically occurs in conjunctionwith myocyte hypertrophy and interstitial fibrosis(18). The extent of myocyte disarray in HCM varies,but typically involves more than 20% of the myocar-dium and often is more prominent in the septum(19,20). Interstitial fibrosis, typically detected bydelayed gadolinium enhancement in the myocar-dium, is relatively common in HCM, and is associatedwith the increased risk of sudden cardiac death (SCD),mortality, and morbidity (21,22).

Individuals with HCM are typically asymptomaticor minimally symptomatic. Exercise intolerance andreduced cardiopulmonary capacity are often present.The cardinal symptoms are palpitations, pre-syncope,and syncope due to cardiac arrhythmias. Syncope isoften a harbinger of SCD, and requires detailed eval-uation and management. The second set of symptomsrelates to diastolic dysfunction and heart failure withpreserved ejection fraction. These symptoms includedyspnea, orthopnea, and peripheral edema, which areoften responsive to diuretic agents, but could be re-fractory to medical therapy, as in other forms of heartfailure with preserved ejection fraction. Patients withHCM often experience chest pain that does not havethe typical features of chest pain of coronary origin.

Chest pain and related symptoms are commonlymanaged with medical therapy, including the use ofbeta-blockers and nondihydropyridine calcium-channel blockers.

HCM may infrequently evolve into dilated cardio-myopathy (DCM), leading to global systolic dysfunc-tion and heart failure symptoms. Left ventricularoutflow obstruction, which is present in approxi-mately 25% of patients, is a risk factor for heart failure(23). Atrial fibrillation is relatively common, and isassociated with heart failure and adverse clinicaloutcomes (24).

HCM is among the most common causes of SCD inthe young, particularly in competitive athletes(25,26). Table 3 lists major predictors of the risk ofSCD in HCM. Syncope is commonly due to ventriculararrhythmias, and less commonly due to orthostatichypotension. It is a major risk factor for SCD (27–30).Sustained or repetitive nonsustained ventriculartachycardia is also a major risk factor for SCD,requiring implantation of a defibrillator. Moreover,severe cardiac hypertrophy, commonly defined as awall thickness $30 mm, and severe interstitialfibrosis, involving multiple segments in cardiacmagnetic resonance, are associated with an increasedrisk of SCD. Despite the concern about the risk of SCD,HCM is a relatively benign disease, with an estimatedannual mortality of about 1% in the adult population(31–33). The focus in clinical practice is to identifythose who are at an increased risk of SCD and inter-vene accordingly. This is a challenging task, as noneof the clinical or genetic predictors reliably predictsthe risk of SCD. Thus, a comprehensive approachtailored to the characteristics of each individual andhis/her family is required.Causa l genes for HCM. HCM, a prototypic single-gene disorder, is commonly familial, exhibiting anautosomal dominant mode of inheritance. It is spo-radic in about one-third of cases. The genetic basis ofHCM has been partially elucidated, and causal mu-tations in over a dozen genes encoding sarcomereproteins have been identified (Table 4). MYH7 andMYBPC3 are the 2 most common causal genes, andtogether are responsible for approximately one-halfof HCM, particularly familial HCM (13,34–37). Theyencode myosin heavy chain 7 or b-myosin heavychain and myosin binding protein C, 2 major compo-nents of the sarcomeres, respectively. Insertion/deletion and frameshift mutations are more commonin the MYBPC3 than in the MYH7 gene (34,36–38).TNNT2, TNNI3, and TPM1, encoding the thin filamentproteins cardiac troponin T, cardiac troponin I,and a-tropomyosin, respectively, are relatively un-common causes of HCM, accounting for <10% of

TABLE 3 Clinical Risk Factors for SCD in Patients With HCM

History of sudden cardiac arrest (aborted SCD)

History of (arrhythmic) syncope

Sustained VT and repetitive NSVT

Family history of SCD (>1 victim of SCD)

Severe cardiac hypertrophy

Left ventricular outflow tract obstruction

Interstitial fibrosis and myocyte disarray

HCM ¼ hypertrophic cardiomyopathy; NSVT ¼ nonsustained ventriculartachycardia; SCD ¼ sudden cardiac death; VT ¼ ventricular tachycardia.


2837

cases/families (37). Mutations in genes coding forseveral other sarcomere and sarcomere-associatedproteins also have been identified as likely causes ofHCM, including titin (TTN), cardiac a-actin (ACTC),telethonin (TCAP), myosin light chain 2 (MYL2),myosin light chain 3 (MYL3), myozenin 2 (MYOZ2),and ubiquitin E3 ligase tripartite motif protein 63(TRIM63) (13,39–42). Overall, the causal genes forapproximately 60% of HCM have been identified.Thus, the remainder, which are typically responsiblefor sporadic HCM or HCM occurring in small families,have yet to be identified (43).

Mutations in genes other than those coding forsarcomere proteins also could cause primary cardiachypertrophy that clinically resembles HCM caused bymutations in sarcomere proteins. The pathophysi-ology of LVH in such conditions typically differs fromthat of HCM caused by mutations in genes encodingsarcomere proteins; hence, these conditions areconsidered to phenocopy HCM. There are severalnotable examples, including Fabry (Anderson-Fabry)disease, and glycogen storage diseases. A partial listof HCM phenocopy conditions is presented in Table 5.

Fabry disease is an X-linked lysosomal storagedisease with pleiotropic phenotypic manifestations,including HCM, angiokeratoma, corneal deposits,renal insufficiency, proteinuria, peripheral neuropa-thy, coronary artery disease, and cerebrovascularevents (44–46). Often the phenotype is restricted tothe heart, rending it indistinguishable from HCMcaused by sarcomere protein mutations (47,48).Findings such as severe hypertrophy, particularly in amale patient; high QRS voltage; conduction defects;and the presence of D-wave mimicking pre-excitationon a 12-lead electrocardiogram suggest screening forFabry disease.

Fabry disease is caused by LOF mutations in GLAencoding a-galactosidase. The enzyme is responsiblefor hydrolysis of a-D-galactose residues in glyco-sphingolipids (49). Mutations result in the accumu-lation of globotriaosylceramide (GB3), the maina-galactosidase substrate, in lysosomes in multiple

organs. Because it is an X-linked disease, it predom-inantly affects male patients, but female mutationcarriers might show a mild phenotype. Despite theavailability of genetic testing and a-galactosidaseA activity assays in whole blood or leukocytes,the diagnosis could be challenging (50,51).Enzyme replacement therapy has shown somesuccess in slowing progression and attenuating thephenotype (52–55).

Mutations in the PRKAG2 gene, which encodes thenoncatalytic g2 regulatory subunit of adenosinemonophosphate–activated protein kinase, also causeHCM phenocopy (56–60). Cardiac hypertrophy resultspredominantly from storage of glycogen in myocytesand, to a lesser degree, from the cardiac response toaltered myocyte function (61). The cardiac hypertro-phy phenotype is often associated with conductiondefects and a pattern mimicking pre-excitation(56,59,60).Genotype-phenotype assoc iat ion . Patients withHCM exhibit considerable variability in the severity ofcardiac hypertrophy and risk of SCD (62,63). Pheno-typic variability (expressivity), variable penetrance,and pleiotropy (multiple phenotypes associated witha single gene or GV) compound genotype-phenotypecorrelation and impede the predictive utility of ge-netic testing in HCM. Phenotypic variability is in partbecause of the multiplicity of phenotypic de-terminants, involving genetic and nongenetic factors.The causal mutation imparts the largest effect size onthe phenotype. However, a large number of geneticand nongenetic determinants also contribute toexpression of the phenotype in each individual,diluting the overall effect of a single determinant.Several candidate genes and loci are implicated inaffecting the severity of cardiac hypertrophy and riskof SCD (64–70). By and large, however, GVs andepigenetic factors that influence the phenotypicvariability of HCM, as well as other forms of heredi-tary cardiomyopathies, are largely unknown.Pathogenes is of HCM. Several mechanisms areimplicated in the pathogenesis of HCM, includingaltered actomyosin cross-bridging, ATPase activity,and the calcium sensitivity of the myofilaments(Figure 3). Most HCM-causing mutations are missensemutations, and the mutant proteins incorporate intothe myofibrils, although the efficiency of transcrip-tion, translation, and incorporation into myofibrilsmight be reduced. Premature truncation mutationsoften lead to unstable transcripts and proteins,resulting in their degradation and hence hap-loinsufficiency (34,36,37,71). Mutations in thick fila-ment proteins typically affect ATPase activity,whereas those in the thin filament proteins alter the

TABLE 4 Genetic Basis for HCM

Protein Gene Locus Frequency

Established causal genes

b-myosin heavy chain MYH7 14q1 w25%

Myosin binding protein C MYBPC3 11q1 w25%

Cardiac troponin T TNNT2 1q3 <5%

Cardiac troponin I TNNI3 19p13.2 <5%

a-tropomyosin TPM1 15q1 <5%

Likely causal genes

Cardiac alpha-actin ACTC1 15q11 <5%

Myozenin 2 MYOZ2 4q25-26 Rare

Tripartite motif containing 63 TRIM63 1p34-33 Rare

Myosin light chain 3 MYL3 3p Rare

Myosin light chain 2 MYL2 12q Rare

Titin TTN 2q13-33 Rare

Telethonin TCAP 17q12 Rare

Myosin light chain kinase 2 MYLK2 20q13.3 Rare

a-myosin heavy chain alpha MYH6 14q12 Rare

Troponin C TNNC1 3p21 Rare

Caveolin 3 CAV3 3p25 Rare

Phospholamban PLN 6p22.1 Rare

Lamin A/C LMNA 21.2-q21.3 Rare

Calsequestrin CASQ2 1p13.1 Rare

Junctophilin 2 JPH2 20q13.12 Rare

HCM ¼ hypertrophic cardiomyopathy.

TABLE 5

Gene

PRKAG2

PTPN11

SOS1

GLA

GAA

LAMP2

FRDA

DMPK

mtDNA

AMP ¼ ade



2838

calcium sensitivity of the myofilaments (71), Thesedefects instigate a cascade of molecular changes suchas altered transcriptomics, including expression ofncRNAs, activation of signaling pathways, andexpression of trophic and mitotic factors, whichcollectively lead to cardiac hypertrophy and fibrosis(69,72–76). Regardless of the primary defect, cardiachypertrophy and fibrosis are considered secondary,and therefore are potentially preventable andreversible (77).

HCM Phenocopy Conditions

Protein Phenotype

AMP-activated protein kinase,g2 subunit

Glycogen storage disease

Protein-tyrosine phosphatase,nonreceptor type II

Noonan and Leopold syndromes

Son of sevenless-1 Noonan syndrome

a-galactosidase A Fabry disease

Glucosidase, a, acid Glycogen storage disease, type II

Lysosome-associated membraneprotein 2

Danon disease

Frataxin Friedreich ataxia 1

Dystrophia myotonica proteinkinase

Myotonic dystrophy

Mitochondrial DNA Compound metabolic phenotype

nosine monophosphate; HCM ¼ hypertrophic cardiomyopathy.

Potential new therapeutic interventions. Currentpharmacological treatment of patients with HCM isprimarily on the basis of relieving symptoms, and isnot directed toward reversing, attenuating, and pre-venting the phenotype. A number of new therapeuticapproaches are designed to target the underpinningmechanisms involved in the pathogenesis of HCM.Among the notable clinical trials is the LIBERTY-HCM(Effect of Eleclazine [GS-6615] on Exercise Capacity inSubjects With Symptomatic Hypertrophic Cardiomy-opathy) study of eleclazine, which targets the latesodium channel in cardiac myocytes. The study isdesigned to test the safety and tolerability of elec-lazine, as well as its effects on exercise capacity andquality of life. The HALT-HCM (HypertrophicRegression With N-Acetylcysteine in HCM) study is apilot double-blinded randomized clinical trial,designed to target oxidative stress in HCM. The studytests whether treatment with N-acetylcysteine im-parts a beneficial effect on established cardiachypertrophy and fibrosis. In addition, several pilotstudies have tested the potential utility of angio-tensin II receptor blockers, beta-hydroxy-beta-methylglutaryl-coenzyme A reductase inhibitors,and antioxidants in prevention, attenuation, andreversal of cardiac phenotypes in humans and inanimal models of HCM (78–83). The data in humanpatients are considered preliminary and have shown,at best, a modest (if any) effect (84–90). In addition,diltiazem, an L-type calcium-channel blocker, hasbeen used in pre-clinical mutation carriers to test theeffects on echocardiographic indexes of cardiac sizeand function (91). Finally, gene therapy approachesalso have been applied in experimental models, withsome success (92).

GENETIC BASIS OF PRIMARY DCM. Primary DCM isdefined by the presence of left ventricular dilationand dysfunction in the absence of any discerniblesecondary cause, such as coronary artery disease. Theleft ventricular ejection fraction is typically <0.45and the end-diastolic diameter is >2.7 cm/m2. Pri-mary DCM is an uncommon disease with a prevalenceof approximately 1 per 2,500 individuals, and anincidence of 5 to 8 cases per 100,000 persons (93–95).Primary DCM is familial in about one-third to one-halfof cases, and sporadic in the remainder (96–98). Themode of inheritance is typically autosomal dominant,but X-linked inheritance, such as in Duchenne andBecker muscular dystrophies, and autosomal reces-sive forms also have been reported (50,99,100).

Clinical manifestations of DCM include heart fail-ure and, less often, cardiac arrhythmias. Cardiac ar-rhythmias typically occur late in the course of the

FIGURE 3 Schematic Illustration of Pathogenesis of HCM

Mutation inSarcomere proteins

Functional defects:MechanicalBiochemicalSignalingMetabolicOthers

Trophic and Mitoticpathways

Myocytes stress&

Dysfunction

HCM

Mutations in genes encoding sarcomere proteins impart a diverse array of initial defects, such as altered myofibrillar mechanical or

biochemical functions, leading to expression and activation of trophic and mitotic pathways, including cardiac hypertrophy and interstitial

fibrosis. HCM ¼ hypertrophic cardiomyopathy.


2839

disease, in contrast to AC, wherein cardiac arrhyth-mias are the cardinal manifestations. A notablefeature of DCM is the absence of symptoms in theearly stages of the disease (99). In familial cases,routine clinical screening leads to the diagnosis ofDCM in approximately 10% of asymptomatic familymembers (97,98).Causa l genes for DCM. DCM is genetically a veryheterogeneous disease. TTN, encoding the giantprotein titin, is the most common causal gene(Table 6). TTN mutations are responsible for about25% of familial DCM (101,102). An accurate estimate ofthe prevalence of TTN mutations in DCM is somewhatchallenging because of its large size (>30,000 aminoacids in length) and the abundance of GVs in TTN.Thus, GVs that affect the length of TTN proteins,including nonsense, frameshift, and splice site vari-ants, are considered pathogenic. TTN truncatingvariants are also present in about 1.5% of the generalpopulation. In addition, a subset of missense muta-tions might also be pathogenic. GVs in the A band,N2BA, and N2B transcripts are more likely to bepathogenic (102). Overall, the clinical significance ofGVs in TTN seems to relate to the location of thevariants and exon usage (whether and how commonlythe involved exons are included in the transcriptduring splicing) (102).

Mutations in genes encoding sarcomere proteins,such as MYH7 and TNNT2, are also established causesof DCM (Table 6). In this respect, HCM and DCMpartially share genetic etiology. The findings alsopoint to the phenotypic plasticity (pleiotropic effects)of GVs in sarcomere genes, which hampers a reliableprediction of the phenotype. The contrasting pheno-types of HCM and DCM likely result from the topog-raphy of the mutations, divergence of the initialdefects, differential interactomes, and the geneticbackgrounds in which the mutations operate(103,104).

ACTC, which encodes cardiac a-actin, was the firstcausal gene identified for autosomal dominant DCM(105). Likewise, mutations in cytoskeletal proteinsdelta sarcoglycan, beta-sarcoglycan, and dystrophinwere subsequently identified in families with DCM(94). Mutations in genes encoding cysteine- andglycine-rich protein 4 (CSRP3) and LIM domain-binding protein 3 (LDB3) are also known to causeDCM. More recently, the gene coding for the RNA-binding motif protein 20 (RBM20), which regulatessplicing of multiple cardiac transcripts, has emergedas an important causal gene for DCM (106,107).

LMNA encoding the nuclear membrane proteinlamin A/C is among the most intriguing causalgenes for DCM. LMNA mutations exhibit extreme

TABLE 6 Causal Genes for Primary DCM

Gene Protein Comments/Phenotypic Plasticity

TTN Titin Giant sarcomere protein, responsible for w25%of primary DCM, also causes HCM

MYH7 Myosin heavy chain 7 (b) HCM

TNNT2 Cardiac troponin T HCM

TNNI3 Cardiac troponin I HCM

TNNC1 Cardiac troponin C HCM

TPM1 a-tropomyosin HCM

ACTC Cardiac a-actin HCM

TNNI3K Troponin I interacting kinase Conduction defect, atrial fibrillation

LMNA Lamin A/C Nuclear envelope protein responsible for over 1dozen phenotypes

EMD Emerin Emery-Dreifuss syndrome

RBM20 RNA-binding motif protein 20 Targets splicing of several cardiac genes

SGCA a-sarcoglycan Involves skeletal muscle

SGCB b-sarcoglycan

SGCD d-sarcoglycan

DMD Dystrophin Duchenne muscular dystrophy

CSRP3 Cysteine and glycine rich protein 3

ANKRD1 Ankyrin repeat domain 1

DES Desmin Desminopathy

CRYAB aB-crystallin Protein aggregation myopathy

ACTN2 Alpha-actinin 2

TCAP Telethonin (T-cap)

LDB3 LIM domain binding 3 (Z-bandalternatively spliced PDZ motif)

VCL Vinculin

BAG3 BCL2-associated athanogene 3 Cochaperone

SCN5A Sodium voltage-gated channel Also causes Brugada syndrome and conductiondefects

ABCC9 SUR2 subunit of potassium channels

PLN Phospholamban Inhibits SERCA2

KCNQ1 Potassium channel

DCM ¼ dilated cardiomyopathy; HCM ¼ hypertrophic cardiomyopathy; RNA ¼ ribonucleic acid; SERCA ¼ sarco/endoplasmic reticulum calcium ATPase.



2840

phenotypic diversity, encompassing over a dozendistinct phenotypes, including Hutchinson-Gilfordprogeria syndrome, muscular dystrophy, and lip-odystrophy, among others, which are collectivelyreferred to as laminopathies (108). DCM caused bymutations in LMNA is typically associated with con-duction defects, such as atrial bradyarrhythmias andleft bundle branch block (108).

Another notable feature of the genetics of DCMpertains to overlapping conditions with cardiac ar-rhythmias, resulting from mutations in genes encod-ing ion channels. Notable examples are mutations inSCN5A, which encodes sodium voltage-gated channelalpha subunit 5, and ABCC9, which encodes the reg-ulatory SUR2A subunit of the cardiac KATP channels(109,110). Mutations in SCN5A are known to causelong QT syndrome type 3, Brugada syndrome, andfamilial heart block (111). Similarly, mutations inABCC9 have been associated with DCM and atrialfibrillation (112).

DCM could be a phenotypic consequence of proteinaggregation in the myocardium, referred to as pro-teotoxic cardiomyopathy (113,114). Mutations in thegenes encoding desmin and alpha/B-crystallin causeproteotoxic DCM (114). Phenotypes caused by DESmutations are also referred to as desminopathies, andoften involve skeletal muscle as well.

Mutations in gene encoding alpha-sarcoglycan(adhalin) cause autosomal recessive DCM, whichtypically occurs in conjunction with limb-girdlemuscular dystrophy (115). Cardiac involvement,manifesting as DCM, is a common feature of severalforms of limb-girdle muscular dystrophy (115).Genotype–phenotype association. Extreme geneticheterogeneity, variable penetrance, and pleotropiceffects of the GVs prohibit robust genotype-phenotype correlation studies. However, a notablefinding is the presence of concomitant phenotypes ina significant number of patients with DCM, such asconduction defects or skeletal myopathy, whichoffers clues to its genetic etiology. The presence ofleft bundle branch block hints at LMNA, and skeletalmyopathy hints at DMD (dystrophin) (108,116). Like-wise, DCM observed in conjunction with the expan-sion of triplet repeats, such as in myotonic dystrophy,is typically associated with skeletal myopathy,conduction defects, and cardiac arrhythmia (117).Nevertheless, given the considerable phenotypicvariability of the causal genes, accurate clinical pre-diction is unreliable. Moreover, availability of genetictesting on a routine basis has abrogated the need foretiological speculation on the basis of the clinicalmanifestations of DCM.Pathogenes is of DCM. Considering the geneticheterogeneity of DCM, affecting proteins with adiverse array of functions, no unifying mechanism isresponsible for its pathogenesis. The causal muta-tions impart a diverse set of primary defects,including changes in cytoskeletal integrity, mechan-ical force generation and transmission, and myofila-ment sensitivity to calcium. Mutations in sarcomereproteins, particularly those in the thin filaments,reduce calcium sensitivity of the myofilaments, thosein actin and cytoskeletal proteins affect force trans-mission, and those in thick filaments affect ATPaseactivity (94,118). Mutations in RBM20 affect splicingof transcripts of cardiac genes, and those in DES andCRYAB lead to accumulation of protein aggregatesand proteotoxicity in the heart (107,113). The molec-ular pathogenesis of DCM caused by mutations invarious genes, including the most common gene,TTN, remains largely unknown.Novel therapeut ic approaches . Current pharma-cological and nonpharmacological treatment of DCM

FIGURE 4 Fibro-Adipocytes Replacing Cardiac

Myocytes in ARVC

The photomicrograph illustrates thin myocardial sections

stained with Masson trichrome, showing excess adipocytes and

fibrosis (blue). The pathognomonic feature of classic ARVC is

excess fibro-adipocytes replacing cardiac myocytes predomi-

nantly in the right ventricle. ARVC ¼ arrhythmogenic right

ventricular cardiomyopathy.


2841

is similar to that of heart failure with reducedleft ventricular ejection fraction. The list includesb-blockers and inhibitors of the renin-angiotensin-aldosterone system. Gene therapy approaches arenot quite successful in suppressing expression of themutant allele. Recently, the clustered regularlyinterspaced short palindromic repeats (CRISPR)-Cas9system has been used to delete the mutant exonthrough an exon-skipping mechanism, which leads tostable dystrophin expression levels and improvementin skeletal muscle and cardiac function (119–122).Genet i c bas i s of AC . AC is a new term coined todefine a set of primary cardiomyopathies whose car-dinal manifestation is ventricular arrhythmiasoccurring in the presence of, but disproportionate to,ventricular dysfunction. AC is distinct from conven-tionally defined DCM, whose primary manifestation iscardiac dysfunction and heart failure, whereas ven-tricular arrhythmias are often late manifestationsoccurring in the presence of severe left ventriculardysfunction. Thus, AC, in a sense, encompasses asubgroup of DCM and classic ARVC, formerly knownas arrhythmogenic right ventricular dysplasia(Figure 2). AC is a global ventricular disease involvingboth the right and left ventricles, whereas ARVCpredominantly and classically (but not exclusively)involves the right ventricle. In advanced stages ofARVC, the left ventricle is commonly involved (123).The characteristic pathological feature of ARVC isexcess fibro-adipocytes in the ventricle replacingcardiac myocytes (Figure 4) (124–127).

The clinical phenotype of AC includes ventriculararrhythmias, SCD, and heart failure; the latter occursin advanced cases. The left ventricle is commonlyinvolved, and its involvement in the ARVC subsetportends a poor prognosis (127). Initial presentationof patients with AC is ventricular arrhythmias and,less commonly, cardiac arrest, particularly duringexercise (126,128,129). In the young, AC is an impor-tant cause of SCD, being responsible for 3% to 5% ofall SCDs (25). AC accounts for up to 25% of SCD casesin young athletes in certain parts of Italy (130–132). Asubset of AC occurs in conjunction with skin abnor-malities, namely keratosis, wooly hair, and some-times baldness. This subset is referred to ascardiocutaneous syndrome (133,134).

Electrocardiographic features include the charac-teristic (and yet uncommon) epsilon-wave, depolari-zation and repolarization abnormalities in the rightprecordial leads, and ventricular arrhythmias origi-nating from the right ventricle (135,136). Echocardio-graphic findings in classic ARVC are notable for rightventricular dilation and dysfunction, as well asaneurysm formation, but these findings are

uncommon and are typically observed in theadvanced stages. Likewise, cardiac magnetic reso-nance is a valuable diagnostic test in a limited num-ber of patients and might show evidence of fibro-fattyinvolvement of the right ventricle, as well as struc-tural and functional abnormalities.

A pathological hallmark of AC is fibro-adipocytesreplacing cardiac myocytes, predominantly in theright ventricle, and in both ventricles in advancedcases (Figure 5). This phenotype is typically associ-ated with atrophic myocytes, myocardial wall thin-ning, dilation, and aneurysm formation (137).However, none of the clinical or histological findingsare sufficiently sensitive or specific for an accuratediagnosis of AC. Typically, the presence of 2 major, or1 major and 2 minor, or 4 minor criteria are requiredfor the diagnosis of AC (138).Causal genes for AC. The discovery of a 2-base-pairdeletion mutation in the JUP gene, which encodes thejunction protein plakoglobin, provided the first clueto the genetic causes of AC (133). The discovery wasmade in a family with an autosomal recessive diseasecharacterized by woolly hair, keratosis, and ARVC,who resided in the island of Naxos off the coast ofmainland Greece. The discovery had a watershed ef-fect, as it led to identification of causal mutations inPKP2, DSP, DSG2, and DSC2, coding for the desmo-some or intercalated disc proteins plakophilin 2,desmoplakin, desmoglein 2, and desmoscollin 2,respectively (139–144). The PKP2 gene is the most

FIGURE 5 Pathogenesis of AC

The diagram illustrates the mechanisms involved in the pathogenesis of cardiac dysfunction, arrhythmias, and excess fibro-adipocytes in AC. Mutations impair the mechanical

integrity of the intercalateddiscs and lead to cardiac dysfunction.Mechanical dysfunction also activatesmechanosensitive signal transductionpathways, leading to suppressionof

gene expression through the YAP-TEAD and CTNNB1 (b-catenin)-TCF7L2 pathways. A transcriptional switch in a subset of fibro-adipocyte progenitor cells results in their

differentiation to fibro-adipocytes, and also contributes to cardiac dysfunction. AC¼ arrhythmogenic cardiomyopathy; MAPK¼ mitogen-activated protein kinase.



2842

TABLE 7 Causal Genes for Arrhythmogenic Cardiomyopathy

Locus Gene Protein Prevalence

Intercalated disc proteins

12p11 PKP2 Plakophilin 2 w20%

18q12 DSG2 Desmoglein 2 w10%

18q12 DSC2 Desmocollin 2 w10%

6p24 DSP1 Desmoplakin 10%

17q21 JUP Plakoglobin Uncommon

Others

3p23 TMEM43 Transmembrane protein 43 Rare

1q21.2 LMNA Lamin A/C Rare

14q24 TGFB3? Transforming growth factor b3 Rare

6p22.1 PLN? Phospholamban Rare

11p15.5 KCNQ1? Iks channels Rare


2843

common causal gene for ARVC, accounting for up toone-quarter of the ARVC cases (140). Overall, genescoding for intercalated disc proteins account forabout one-half of the ARVC cases, and the remainingcausal genes are unknown (Table 7).

The TMEM43 gene, encoding transmembrane pro-tein 43, has been identified as a causal gene for ARVCthrough linkage analysis and positional cloning, andsubsequently through screening in additional fam-ilies (145,146). Biological functions of TMEM43 andthe mechanisms responsible for AC have remainedlargely unexplored.

Mutations in LMNA, PLN, and TTN are also impli-cated as causes of AC (147–149). Moreover, noncodingvariants in the TGFB3 gene have been associated withARVC (150). Likewise, a missense mutation in KCNQ1,encoding a potassium channel, has been reported in apatient with AC (151). Mutations in RYR2, encodingryanodine receptor 2, cause the phenotype of cate-cholaminergic (stress-induced) polymorphic ventric-ular tachycardia, which is a phenocopy of AC.

Mutations in JUP and DSP are known to causecardiocutaneous subforms of ARVC (133,134,152,153).A recessive 2-base pair deletion in JUP, which leads topremature truncation of JUP protein, is a cause ofNaxos disease (133,134). A left ventricular dominantform of cardiocutaneous syndrome, referred to asCarvajal syndrome, is caused by mutations in the DSPgene (124,134,153).Pathogenes i s of AC . The pathogenesis of AC hasbeen enigmatic. Identification of causal mutations inthe intercalated disc proteins provided the first clue.Intercalated discs are conventionally recognized asmechanical structures responsible for cell–cellattachment, predominantly in epithelial cells andmyocytes. Consequently, part of the underlyingmechanism in the pathogenesis of AC is impairedmechanical integrity of cardiac myocytes. In accordwith these functions, a number of protein constitu-ents of the intercalated discs, particularly the junc-tion protein JUP, are dislocated, redistributed, ordegraded in AC (154–156). Dislocation of JUP fromthe junction might serve as a molecular marker forAC, albeit with limited utility (155,157). Molecularremodeling and impaired assembly of the interca-lated discs not only affects global cardiac function,but also instigates mechanosensitive signalingpathways, including the Hippo pathway, which isactivated in AC (154) (Figure 5). Activation of theupstream molecules of the Hippo pathway leads to acascade of phosphorylation of a number of inter-mediary molecules and the downstream effectormolecule YAP, and to cytoplasmic sequestrationof YAP and its binding partner b-catenin of the

canonical Wnt signaling pathway. The net effect issuppression of gene expression through theYAP-TEAD and b-catenin–TCF7L2 transcriptionalcomplexes (154,156). These molecular events couldexplain the atrophic nature of cardiac myocytes inAC, as the Hippo pathway is a major regulator of celland organ size, a process that is regulated throughcell–cell contact (158).

Perturbation of the Hippo and the canonical Wntsignaling pathways in AC results in differentiationof a subset of cardiac cells to fibro-adipocytes,explaining the mechanistic basis of this enigmaticphenotype (154,156,159–161). In addition, membersof the intercalated disc proteins, such as JUP andPKP2, also possess functions beyond serving asstructural proteins. JUP is known to partially trans-locate from the cell membrane to the nucleus andregulate gene expression, in part through competi-tion with b-catenin for binding to the TCF7L2 tran-scription factor (156,159). Similarly, PKP2 is alsoknown to translocate into the nucleus and interactwith RNA polymerase III, influencing gene expression(162). Collectively, the data indicate a major role forintercalated discs as signaling hubs regulating variousmechanotransduction-activated pathways.

Recently, a subset of cardiac mesenchymal pro-genitor cells, referred to as cardiac fibro-adipocyteprogenitors (FAPs), has been identified as a cellsource of excess adipocytes in AC (161). Accordingly,FAPs are bipotential cells with a capacity to differ-entiate to adipocytes or fibroblasts, depending on theexpression of desmosome proteins and the externalstimulus. The majority of FAPs express fibroblastmarkers, and they have a natural tendency to differ-entiate into fibroblasts in culture. However, a subsetof FAPs expresses desmosome proteins and differ-entiates to adipocytes (159,161,163). The responsiblemechanisms involve activation of the Hippo and



2844

suppression of the canonical Wnt signaling pathways(161). In addition, a minority of the excess adipocytesin AC might originate from the second heart field andkitþ cardiac progenitor cells (159,160). Moreover, in-hibition of Rho-kinase during cardiac development isalso implicated in the pathogenesis of excess adipo-cytes in AC (164).

Finally, miR-184, which is predominantlyexpressed in embryonic myocytes and cardiacmesenchymal progenitor cells, is down-regulated inAC models (165). Down-regulation of miR-184 isimplicated in cellular proliferation and differentia-tion of a subset of cardiac progenitor cells to adipo-cytes (165).Novel therap ies . There is no effective pharmaco-logical therapy for AC. Treatment is typically targetedto the clinical manifestations of the disease, such asablation of cardiac arrhythmias. Recently, pharma-cological inhibition of GSK3B has been shown to exertbeneficial effects in experimental models of AC(166,167).

PERSPECTIVE ON SINGLE-GENE DISEASES

Breakthroughs during the last 3 decades have usheredin the next phase of genetic discoveries, and theirapplication to diagnosis and management of patientswith single-gene disorders. The field is transitioningfrom initial genetic discoveries to the next phase ofdiscovery, applications, and genetic-based in-terventions. A few challenges are briefly discussed.

THE “MISSING” CAUSAL GENES. Despite enormousprogress in elucidation of the molecular genetic basisof single-gene disorders during the last 3 decades,and despite recent technological advances, such aswhole-exome sequencing, the causal genes inapproximately one-third to one-half of single-genedisorders have remained elusive. There are a num-ber of obstacles in identification of the “missing”causal genes for single-gene disorders, including he-reditary cardiomyopathies. The difficulty is, in part,inherent to the complexity of the GVs in the humangenome and the sporadic nature of a significant pro-portion of single-gene diseases or the small size of thefamilies. These inherent complexities render unam-biguous ascertainment of causality in a single indi-vidual or in small families with single-gene disordersalmost impossible. Consequently, to fully use thegenetic information and advance the yield of geneticscreening, novel approaches are needed to identifythe causal variant among myriad potentially patho-genic variants.

CLINICAL APPLICATIONS OF GENETIC DISCOVERIES.

Genetic testing has emerged at the forefront of the

clinical management of patients with single-genedisorders. Genetic testing is commercially availablefor a large number of single-gene disorders,including cardiomyopathies. The common approachis whole-exome sequencing, followed by identifica-tion of pathogenic GVs in genes that have beenpreviously implicated in cardiomyopathies, whichcurrently encompasses about 100 genes. Genetictesting should be pursued in all patients withhereditary cardiomyopathies and, whenever appli-cable, extended to all related family members.This is the key point in genetic testing, as unam-biguous ascertainment of causality in a single caseis almost impossible. Extending the studies tophenotypically affected and unaffected familymembers empowers detection of the causal variants.The approach also often leads to identificationof mutation carriers, referred to as genotypepositive–phenotype negative individuals. Earlyidentification of mutation carriers could lead to closemonitoring through exquisite phenotyping, andearly interventions, which might prove to be bene-ficial in prevention and attenuation of the pheno-type, and reducing the risk of SCD.

Despite the utility of cascade genetic screeningof family members, the “missing” causal genes havesubstantially hampered its widespread clinical utility.Moreover, as indicated, ascertaining causality insmall families and probands is quite challenging, ifnot impossible. In addition, the extreme phenotypicvariability of the GVs has limited a straightforwardgenotype-phenotype correlation. Considering thesechallenges, the most important application of thegenetic discoveries might be in elucidation ofnovel therapeutic approaches to treatment of single-gene disorders. Although large-scale sequencingand population-based genotype-phenotype correla-tion studies would offer valuable informationat the population level, the next phase of discoverieswould mandate focusing on the individual. Compre-hensive cataloging of GVs, genomic determinants,and nongenetic factors would be required, along withdetailed phenotypic characterization, to garnermeaningful information in a single individual.

ROLE OF NONCODING RNA. The role of lncRNAs inthe pathogenesis of hereditary cardiomyopathieshas been difficult to decipher for various reasons,including poor annotation and the lack of cross-species conservation. Additionally, the interactionsof lncRNAs with other biomolecules are not welldefined. Moreover, effects of the GVs on structureand function of lncRNAs, as well as their interactionswith other molecules, are difficult to predict.


2845

Despite the scant available data, the relevance oflncRNAs in biology and disease is indubitable. Largeralterations, such as chromosomal rearrangements(translocations, amplifications, or deletions), thataffect the expression of lncRNAs would be expectedto influence the pathogenesis of hereditarycardiomyopathies.

CORRECTION OF THE UNDERLYING GENETIC

DEFECTS. The ultimate application of the geneticdiscoveries in single-gene disorders is to correct theunderlying causal mutation. The convergence of 2major sets of discoveries, namely elucidation of themolecular genetic basis of single-gene cardiovasculardiseases and development of gene-editing tech-niques, such as the CRISPR-Cas9 system, has usheredin the possibility of effective correction of the un-derlying causal mutations (168). Cas9 is a bacterialnuclease that is guided by an RNA molecule to thedesired sites on the genome, where it induces adouble-strand break (DSB). This property enables se-lection of the specific guide RNA to target Cas9 to thedesired specific nucleotide, albeit restrained by therequirement of a specific sequence referred to as aprotospacer adjacent motif sequence, and the loca-tion of the mutation site. In conjunction with theseadvances, the advent of adeno-associated virus 9vectors has enabled efficient gene transfer into car-diac myocytes (169,170). Thus, a combination ofadeno-associated virus 9 constructs and the CRISPR-Cas9 system has the potential for specific targetingof selected mutations in single-gene disorders. Aprototypic clinical application of these powerful toolsis the recent data showing attenuation of cardiac and

skeletal muscle phenotypes in mouse models ofmuscular dystrophy using the RNA-guided CRISPR-Cas9–based gene editing (120,122,171).

Precise correction of the underlying genetic defectin cardiomyopathies, however, would requireenhancing the homology-directed repair followingCRISPR-Cas9–mediated DSBs (172). However,homology-directed repair does not occur in termi-nally differentiated, nondividing, mature cardiacmyocytes, which renders specific mutation correctionuntenable in the foreseeable future. Thus, one has torely on the alternative approach of DNA repair,namely nonhomologous end-joining (NHEJ), which isthe canonical homology-independent pathway forrepair of DSBs. NHEJ is highly effective in genesilencing (knockdown), but is ineffective for geneediting with specific nucleotide precision. Therefore,the NHEJ repair mechanism is potentially desirablefor specific silencing of a limited number of mutationsin single-gene disorders. These approaches are in theearly stages of development, and the pace of discov-eries is astounding. In addition, new nucleases arebeing discovered and applied for targeted editing ofdesired sequences, such as Cpf1, a single RNA-guidedendonuclease with potential for precise targeting ofthe genome in nondividing cells (173).

REPRINT REQUESTS AND CORRESPONDENCE: Dr.Ali J. Marian, Center for Cardiovascular Genetics,Brown Foundation Institute of Molecular Medicine,The University of Texas Health Science Center, 6770Bertner Street, DAC900, Houston, Texas 77030.E-mail: [email protected].

RE F E RENCE S

1. 1000 Genomes Project Consortium. A map ofhuman genome variation from population-scalesequencing [Erratum in Nature 2011;473:544].Nature 2010;467:1061–73.

2. Kong A, Frigge ML, Masson G, et al. Rate of denovo mutations and the importance of father’s ageto disease risk. Nature 2012;488:471–5.

3. Roach JC, Glusman G, Smit AF, et al. Analysis ofgenetic inheritance in a family quartet by whole-genome sequencing. Science 2010;328:636–9.

4. Kidd JM, Cooper GM, Donahue WF, et al. Map-ping and sequencing of structural variation fromeight human genomes. Nature 2008;453:56–64.

5. Wang J, Wang W, Li R, et al. The diploid genomesequence of an Asian individual. Nature 2008;456:60–5.

6. Wheeler DA, Srinivasan M, Egholm M, et al. Thecomplete genome of an individual by massivelyparallel DNA sequencing. Nature 2008;452:872–6.

7. Francioli LC, Polak PP, Koren A, et al. Genome-wide patterns and properties of de novo mutationsin humans. Nat Genet 2015;47:822–6.

8. De Sandre-Giovannoli A, Bernard R, Cau P, et al.Lamin A truncation in Hutchinson-Gilford progeria.Science 2003;300:2055.

9. Marian AJ, Belmont J. Strategic approaches tounraveling genetic causes of cardiovascular dis-eases. Circ Res 2011;108:1252–69.

10. Marian AJ. Genetic causality in complex traits:the case of uric acid. J Am Coll Cardiol 2016;67:417–9.

11. Marian AJ. Causality in genetics: the gradientof genetic effects and back to Koch’s postulates ofcausality [Erratum in Circ Res 2014;114:e29]. CircRes 2014;114:e18–21.

12. Bhartiya D, Jalali S, Ghosh S, et al. Distinctpatterns of genetic variations in potential func-tional elements in long noncoding RNAs. HumMutat 2014;35:192–201.

13. Marian AJ. Hypertrophic cardiomyopathy: fromgenetics to treatment. Eur J Clin Invest 2010;40:360–9.

14. Maron BJ, Gardin JM, Flack JM, et al. Preva-lence of hypertrophic cardiomyopathy in a generalpopulation of young adults. Echocardiographicanalysis of 4111 subjects in the CARDIA Study.Circulation 1995;92:785–9.

15. Nagueh SF, Kopelen HA, Lim DS, et al. TissueDoppler imaging consistently detects myocardialcontraction and relaxation abnormalities, irre-spective of cardiac hypertrophy, in a transgenicrabbit model of human hypertrophic cardiomyop-athy. Circulation 2000;102:1346–50.

16. Nagueh SF, Bachinski LL, Meyer D, et al. TissueDoppler imaging consistently detects myocardialabnormalities in patients with hypertrophic car-diomyopathy and provides a novel means for anearly diagnosis before and independently ofhypertrophy. Circulation 2001;104:128–30.

mailto:[email protected]

http://refhub.elsevier.com/S0735-1097(16)36688-8/sref1






























































2846

17. Nagueh SF, McFalls J, Meyer D, et al. TissueDoppler imaging predicts the development ofhypertrophic cardiomyopathy in subjects withsubclinical disease. Circulation 2003;108:395–8.

18. Varnava AM, Elliott PM, Mahon N, et al.Relation between myocyte disarray and outcomein hypertrophic cardiomyopathy. Am J Cardiol2001;88:275–9.

19. Maron BJ, Sato N, Roberts WC, et al. Quanti-tative analysis of cardiac muscle cell disorganiza-tion in the ventricular septum. Comparison offetuses and infants with and without congenitalheart disease and patients with hypertrophic car-diomyopathy. Circulation 1979;60:685–96.

20. Maron BJ, Roberts WC. Quantitative analysisof cardiac muscle cell disorganization in the ven-tricular septum of patients with hypertrophiccardiomyopathy. Circulation 1979;59:689–706.

21. Chan RH, Maron BJ, Olivotto I, et al. Prognosticvalue of quantitative contrast-enhanced cardio-vascular magnetic resonance for the evaluation ofsudden death risk in patients with hypertrophiccardiomyopathy. Circulation 2014;130:484–95.

22. Ismail TF, Jabbour A, Gulati A, et al. Role oflate gadolinium enhancement cardiovascularmagnetic resonance in the risk stratification ofhypertrophic cardiomyopathy. Heart 2014;100:1851–8.

23. Maron MS, Olivotto I, Betocchi S, et al. Effectof left ventricular outflow tract obstruction onclinical outcome in hypertrophic cardiomyopathy.N Engl J Med 2003;348:295–303.

24. Guttmann OP, Rahman MS, O’Mahony C, et al.Atrial fibrillation and thromboembolism in patientswith hypertrophic cardiomyopathy: systematicreview. Heart 2014;100:465–72.

25. Maron BJ, Shirani J, Poliac LC, et al. Suddendeath in young competitive athletes. Clinical, de-mographic, and pathological profiles. JAMA 1996;276:199–204.

26. Bagnall RD, Weintraub RG, Ingles J, et al.A prospective study of sudden cardiac deathamong children and young adults. N Engl J Med2016;374:2441–52.

27. O’Mahony C, Jichi F, Pavlou M, et al., for theHypertrophic Cardiomyopathy Outcomes In-vestigators. A novel clinical risk prediction modelfor sudden cardiac death in hypertrophic cardio-myopathy (HCM risk-SCD). Eur Heart J 2014;35:2010–20.

28. Christiaans I, van Engelen K, van Langen IM,et al. Risk stratification for sudden cardiac death inhypertrophic cardiomyopathy: systematic reviewof clinical risk markers. Europace 2010;12:313–21.

29. Nienaber CA, Hiller S, Spielmann RP, et al.Syncope in hypertrophic cardiomyopathy: multi-variate analysis of prognostic determinants. J AmColl Cardiol 1990;15:948–55.

30. Elliott PM, Poloniecki J, Dickie S, et al. Suddendeath in hypertrophic cardiomyopathy: identifica-tion of high risk patients. J Am Coll Cardiol 2000;36:2212–8.

31. Maron BJ, Rowin EJ, Casey SA, et al.Hypertrophic cardiomyopathy in adulthood asso-ciated with low cardiovascular mortality with

contemporary management strategies. J Am CollCardiol 2015;65:1915–28.

32. Maron BJ, Ommen SR, Semsarian C, et al.Hypertrophic cardiomyopathy: present and future,with translation into contemporary cardiovascularmedicine [Erratum in J Am Coll Cardiol 2014;64:1188]. J Am Coll Cardiol 2014;64:83–99.

33. Klarich KW, Attenhofer Jost CH, Binder J, et al.Risk of death in long-term follow-up of patientswith apical hypertrophic cardiomyopathy [Erratumin Am J Cardiol 2013;112:1271]. Am J Cardiol 2013;111:1784–91.

34. Millat G, Bouvagnet P, Chevalier P, et al.Prevalence and spectrum of mutations in a cohortof 192 unrelated patients with hypertrophic car-diomyopathy. Eur J Med Genet 2010;53:261–7.

35. Kaski JP, Syrris P, Esteban MT, et al. Preva-lence of sarcomere protein gene mutations inpreadolescent children with hypertrophic cardio-myopathy. Circ Cardiovasc Genet 2009;2:436–41.

36. Erdmann J, Daehmlow S, Wischke S, et al.Mutation spectrum in a large cohort of unrelatedconsecutive patients with hypertrophic cardiomy-opathy. Clin Genet 2003;64:339–49.

37. Richard P, Charron P, Carrier L, et al. Hyper-trophic cardiomyopathy: distribution of diseasegenes, spectrum of mutations, and implicationsfor a molecular diagnosis strategy [Erratum inCirculation 2004;109:3258]. Circulation 2003;107:2227–32.

38. Marian AJ, Yu QT, Mares A Jr., et al. Detectionof a new mutation in the beta-myosin heavy chaingene in an individual with hypertrophic cardio-myopathy. J Clin Invest 1992;90:2156–65.

39. Kimura A. Molecular genetics and pathogen-esis of cardiomyopathy. J Hum Genet 2016;61:41–50.

40. Ho CY, Charron P, Richard P, et al. Geneticadvances in sarcomeric cardiomyopathies: state ofthe art. Cardiovasc Res 2015;105:397–408.

41. Osio A, Tan L, Chen SN, et al. Myozenin 2 is anovel gene for human hypertrophic cardiomyop-athy. Circ Res 2007;100:766–8.

42. Chen SN, Czernuszewicz G, Tan Y, et al.Human molecular genetic and functional studiesidentify TRIM63, encoding Muscle RING FingerProtein 1, as a novel gene for human hypertrophiccardiomyopathy. Circ Res 2012;111:907–19.

43. Marian AJ. The case of “missing causal genes”and the practice of medicine: a Sherlock Holmesapproach of deductive reasoning. Circ Res 2016;119:21–4.

44. Kozor R, Grieve SM, Tchan MC, et al. Cardiacinvolvement in genotype-positive Fabry diseasepatients assessed by cardiovascular MR. Heart2016;102:298–302.

45. Linhart A, Kampmann C, Zamorano JL, et al.Cardiac manifestations of Anderson-Fabry disease:results from the international Fabry outcomesurvey. Eur Heart J 2007;28:1228–35.

46. Monserrat L, Gimeno-Blanes JR, Marin F, et al.Prevalence of Fabry disease in a cohort of 508unrelated patients with hypertrophic cardiomy-opathy. J Am Coll Cardiol 2007;50:2399–403.

47. Thomas AS, Hughes DA. Fabry disease. PediatrEndocrinol Rev 2014;12 Suppl 1:88–101.

48. Elstein DA, Altarescu G, Beck M, editors. FabryDisease. New York, NY: Springer, 2010.

49. Romeo G, Migeon BR. Genetic inactivation ofthe a-galactosidase locus in carriers of Fabry’sdisease. Science 1970;170:180–1.

50. Kamdar F, Garry DJ. Dystrophin-deficientcardiomyopathy. J Am Coll Cardiol 2016;67:2533–46.

51. Marian AJ. Challenges in the diagnosis ofAnderson-Fabry disease: a deceptively simple andyet complicated genetic disease. J Am Coll Cardiol2016;68:1051–3.

52. Schiffmann R, Pastores GM, Lien YH, et al.Agalsidase alfa in pediatric patients with Fabrydisease: a 6.5-year open-label follow-up study.Orphanet J Rare Dis 2014;9:169.

53. Hughes DA, Elliott PM, Shah J, et al. Effects ofenzyme replacement therapy on the cardiomyop-athy of Anderson-Fabry disease: a randomised,double-blind, placebo-controlled clinical trial ofagalsidase alfa. Heart 2008;94:153–8.

54. Weidemann F, Breunig F, Beer M, et al.Improvement of cardiac function during enzymereplacement therapy in patients with Fabry dis-ease: a prospective strain rate imaging study.Circulation 2003;108:1299–301.

55. Schiffmann R, Kopp JB, Austin HA III, et al.Enzyme replacement therapy in Fabry disease: arandomized controlled trial. JAMA 2001;285:2743–9.

56. Gollob MH, Green MS, Tang AS, et al. Identi-fication of a gene responsible for familial Wolff-Parkinson-White syndrome [Erratum in N Engl JMed 2001;345:552 and N Engl J Med 2002;346:300]. N Engl J Med 2001;344:1823–31.

57. Murphy RT, Mogensen J, McGarry K, et al.Adenosine monophosphate-activated protein ki-nase disease mimicks hypertrophic cardiomyopa-thy and Wolff-Parkinson-White syndrome: naturalhistory. J Am Coll Cardiol 2005;45:922–30.

58. Arad M, Maron BJ, Gorham JM, et al. Glycogenstorage diseases presenting as hypertrophic car-diomyopathy. N Engl J Med 2005;352:362–72.

59. Arad M, Benson DW, Perez-Atayde AR, et al.Constitutively active AMP kinase mutations causeglycogen storage disease mimicking hypertrophiccardiomyopathy. J Clin Invest 2002;109:357–62.

60. Blair E, Redwood C, Ashrafian H, et al. Muta-tions in the g2 subunit of AMP-activated proteinkinase cause familial hypertrophic cardiomyopa-thy: evidence for the central role of energycompromise in disease pathogenesis. Hum MolGenet 2001;10:1215–20.

61. Kim M, Hunter RW, Garcia-Menendez L, et al.Mutation in the g2-subunit of AMP-activatedprotein kinase stimulates cardiomyocyte prolifer-ation and hypertrophy independent of glycogenstorage. Circ Res 2014;114:966–75.

62. Marian AJ. On genetic and phenotypic vari-ability of hypertrophic cardiomyopathy: natureversus nurture. J Am Coll Cardiol 2001;38:331–4.

63. Lopes LR, Rahman MS, Elliott PM.A systematic review and meta-analysis of




































































































































































































2847

genotype-phenotype associations in patients withhypertrophic cardiomyopathy caused by sarco-meric protein mutations. Heart 2013;99:1800–11.

64. Daw EW, Lu Y, Marian AJ, et al. Identifyingmodifier loci in existing genome scan data. AnnHum Genet 2008;72:670–5.

65. Daw EW, Chen SN, Czernuszewicz G, et al.Genome-wide mapping of modifier chromosomalloci for human hypertrophic cardiomyopathy. HumMol Genet 2007;16:2463–71.

66. Brugada R, Kelsey W, Lechin M, et al. Role ofcandidate modifier genes on the phenotypicexpression of hypertrophy in patients with hy-pertrophic cardiomyopathy. J Investig Med 1997;45:542–51.

67. Marian AJ, Yu QT, Workman R, et al.Angiotensin-converting enzyme polymorphism inhypertrophic cardiomyopathy and sudden cardiacdeath. Lancet 1993;342:1085–6.

68. Lechin M, Quiñones MA, Omran A, et al.Angiotensin-I converting enzyme genotypes andleft ventricular hypertrophy in patients with hy-pertrophic cardiomyopathy. Circulation 1995;92:1808–12.

69. Lim DS, Roberts R, Marian AJ. Expressionprofiling of cardiac genes in human hypertrophiccardiomyopathy: insight into the pathogenesis ofphenotypes. J Am Coll Cardiol 2001;38:1175–80.

70. Christodoulou DC, Wakimoto H, Onoue K,et al. 50RNA-Seq identifies Fhl1 as a geneticmodifier in cardiomyopathy. J Clin Invest 2014;124:1364–70.

71. Marston S, Copeland O, Gehmlich K, et al. Howdo MYBPC3 mutations cause hypertrophic cardio-myopathy? J Muscle Res Cell Motil 2012;33:75–80.

72. Hwang JJ, Allen PD, Tseng GC, et al. Micro-array gene expression profiles in dilated and hy-pertrophic cardiomyopathic end-stage heartfailure. Physiol Genomics 2002;10:31–44.

73. 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.

74. Fang L, Ellims AH, Moore XL, et al. CirculatingmicroRNAs as biomarkers for diffuse myocardialfibrosis in patients with hypertrophic cardiomy-opathy. J Transl Med 2015;13:314.

75. Kuster DW, Mulders J, Ten Cate FJ, et al.MicroRNA transcriptome profiling in cardiac tissueof hypertrophic cardiomyopathy patients withMYBPC3 mutations. J Mol Cell Cardiol 2013;65:59–66.

76. Derda AA, Thum S, Lorenzen JM, et al. Blood-based microRNA signatures differentiate variousforms of cardiac hypertrophy. Int J Cardiol 2015;196:115–22.

77. Marian AJ. Pathogenesis of diverse clinical andpathological phenotypes in hypertrophic cardio-myopathy. Lancet 2000;355:58–60.

78. Senthil V, Chen SN, Tsybouleva N, et al. Pre-vention of cardiac hypertrophy by atorvastatin in atransgenic rabbit model of human hypertrophiccardiomyopathy. Circ Res 2005;97:285–92.

79. Patel R, Nagueh SF, Tsybouleva N, et al.Simvastatin induces regression of cardiac hyper-trophy and fibrosis and improves cardiac functionin a transgenic rabbit model of human hypertro-phic cardiomyopathy. Circulation 2001;104:317–24.

80. Lombardi R, Rodriguez G, Chen SN, et al.Resolution of established cardiac hypertrophy andfibrosis and prevention of systolic dysfunction in atransgenic rabbit model of human cardiomyopathythrough thiol-sensitive mechanisms. Circulation2009;119:1398–407.

81. Marian AJ. Experimental therapies in hyper-trophic cardiomyopathy. J Cardiovasc Transl2009;2:483–92.

82. Lim DS, Lutucuta S, Bachireddy P, et al.Angiotensin II blockade reverses myocardialfibrosis in a transgenic mouse model of humanhypertrophic cardiomyopathy. Circulation 2001;103:789–91.

83. Marian AJ, Senthil V, Chen SN, et al. Anti-fibrotic effects of antioxidant N-acetylcysteine in amouse model of human hypertrophic cardiomy-opathy mutation. J Am Coll Cardiol 2006;47:827–34.

84. Nagueh SF, Lombardi R, Tan Y, et al. Ator-vastatin and cardiac hypertrophy and function inhypertrophic cardiomyopathy: a pilot study. Eur JClin Invest 2010;40:976–83.

85. Bauersachs J, Störk S, Kung M, et al. HMG CoAreductase inhibition and left ventricular mass inhypertrophic cardiomyopathy: a randomizedplacebo-controlled pilot study. Eur J Clin Invest2007;37:852–9.

86. Yamazaki T, Suzuki J, Shimamoto R, et al.A new therapeutic strategy for hypertrophic non-obstructive cardiomyopathy in humans. A ran-domized and prospective study with anangiotensin II receptor blocker. Int Heart J 2007;48:715–24.

87. Shimada YJ, Passeri JJ, Baggish AL, et al. Ef-fects of losartan on left ventricular hypertrophyand fibrosis in patients with nonobstructive hy-pertrophic cardiomyopathy. J Am Coll Cardiol HF2013;1:480–7.

88. Penicka M, Gregor P, Kerekes R, et al., for theCandesartan use in Hypertrophic And Non-obstructive Cardiomyopathy Estate (CHANCE)Study Investigators. The effects of candesartan onleft ventricular hypertrophy and function in non-obstructive hypertrophic cardiomyopathy: a pilot,randomized study. J Mol Diagn 2009;11:35–41.

89. MacDonald KA, Kittleson MD, Larson RF, et al.The effect of ramipril on left ventricular mass,myocardial fibrosis, diastolic function, and plasmaneurohormones in Maine Coon cats with familialhypertrophic cardiomyopathy without heart fail-ure. J Vet Intern Med 2006;20:1093–105.

90. Axelsson A, Iversen K, Vejlstrup N, et al. Ef-ficacy and safety of the angiotensin II receptorblocker losartan for hypertrophic cardiomyopathy:the INHERIT randomised, double-blind, placebo-controlled trial. Lancet Diabetes Endocrinol 2015;3:123–31.

91. Ho CY, Lakdawala NK, Cirino AL, et al. Diltia-zem treatment for pre-clinical hypertrophic car-diomyopathy sarcomere mutation carriers: a pilot

randomized trial to modify disease expression.J Am Coll Cardiol HF 2015;3:180–8.

92. Jiang J, Wakimoto H, Seidman JG, et al. Allele-specific silencing of mutant Myh6 transcripts inmice suppresses hypertrophic cardiomyopathy.Science 2013;342:111–4.

93. Codd MB, Sugrue DD, Gersh BJ, et al. Epide-miology of idiopathic dilated and hypertrophiccardiomyopathy. A population-based study inOlmsted County, Minnesota, 1975–1984. Circula-tion 1989;80:564–72.

94. McNally EM, Golbus JR, Puckelwartz MJ.Genetic mutations and mechanisms in dilatedcardiomyopathy. J Clin Invest 2013;123:19–26.

95. Manolio TA, Baughman KL, Rodeheffer R,et al. Prevalence and etiology of idiopathic dilatedcardiomyopathy (summary of a National Heart,Lung, and Blood Institute workshop. Am J Cardiol1992;69:1458–66.

96. Petretta M, Pirozzi F, Sasso L, et al. Reviewand metaanalysis of the frequency of familialdilated cardiomyopathy. Am J Cardiol 2011;108:1171–6.

97. Michels VV, Moll PP, Miller FA, et al. Thefrequency of familial dilated cardiomyopathy in aseries of patients with idiopathic dilated cardio-myopathy. N Engl J Med 1992;326:77–82.

98. Goerss JB, Michels VV, Burnett J, et al. Fre-quency of familial dilated cardiomyopathy. EurHeart J 1995;16 Suppl O:2–4.

99. Mestroni L, Rocco C, Gregori D, et al., for theHeart Muscle Disease Study Group. Familialdilated cardiomyopathy: evidence for genetic andphenotypic heterogeneity. J Am Coll Cardiol 1999;34:181–90.

100. Mangin L, Charron P, Tesson F, et al. Familialdilated cardiomyopathy: clinical features in Frenchfamilies. Eur J Heart Fail 1999;1:353–61.

101. Herman DS, Lam L, Taylor MR, et al. Trun-cations of titin causing dilated cardiomyopathy.N Engl J Med 2012;366:619–28.

102. Roberts AM, Ware JS, Herman DS, et al. In-tegrated allelic, transcriptional, and phenomicdissection of the cardiac effects of titin trunca-tions in health and disease. Sci Transl Med 2015;7:270ra6.

103. Lombardi R, Bell A, Senthil V, et al. Differ-ential interactions of thin filament proteins in twocardiac troponin T mouse models of hypertrophicand dilated cardiomyopathies. Cardiovasc Res2008;79:109–17.

104. Davis J, Davis LC, Correll RN, et al. A tension-based model distinguishes hypertrophic versusdilated cardiomyopathy. Cell 2016;165:1147–59.

105. Olson TM, Michels VV, Thibodeau SN, et al.Actin mutations in dilated cardiomyopathy, aheritable form of heart failure. Science 1998;280:750–2.

106. Guo W, Schafer S, Greaser ML, et al. RBM20,a gene for hereditary cardiomyopathy, regulatestitin splicing. Nat Med 2012;18:766–73.

107. Maatz H, Jens M, Liss M, et al. RNA-bindingprotein RBM20 represses splicing to orchestratecardiac pre-mRNA processing. J Clin Invest 2014;124:3419–30.









































































































































































































2848

108. Cattin ME, Muchir A, Bonne G. ’State-of-the-heart’ of cardiac laminopathies. Curr Opin Cardiol2013;28:297–304.

109. Bienengraeber M, Olson TM, Selivanov VA,et al. ABCC9 mutations identified in human dilatedcardiomyopathy disrupt catalytic KATP channelgating. Nat Genet 2004;36:382–7.

110. McNair WP, Sinagra G, Taylor MR, et al., forthe Familial Cardiomyopathy Registry ResearchGroup. SCN5A mutations associate witharrhythmic dilated cardiomyopathy and commonlylocalize to the voltage-sensing mechanism. J AmColl Cardiol 2011;57:2160–8.

111. Veerman CC, Wilde AA, Lodder EM. The car-diac sodium channel gene SCN5A and its geneproduct NaV1.5: role in physiology and patho-physiology. Gene 2015;573:177–87.

112. Olson TM, Alekseev AE, Moreau C, et al. KATPchannel mutation confers risk for vein of Marshalladrenergic atrial fibrillation. Nat Clin PractCardiovasc Med 2007;4:110–6.

113. McLendon PM, Robbins J. Desmin-relatedcardiomyopathy: an unfolding story. Am J PhysiolHeart Circ Physiol 2011;301:H1220–8.

114. Dalakas MC, Park KY, Semino-Mora C, et al.Desmin myopathy, a skeletal myopathy with car-diomyopathy caused by mutations in the desmingene. N Engl J Med 2000;342:770–80.

115. Thompson R, Straub V. Limb-girdle musculardystrophies—international collaborations fortranslational research. Nat Rev Neurol 2016;12:294–309.

116. Muntoni F, Wilson L, Marrosu G, et al.A mutation in the dystrophin gene selectivelyaffecting dystrophin expression in the heart. J ClinInvest 1995;96:693–9.

117. Hermans MC, Pinto YM, Merkies IS, et al.Hereditary muscular dystrophies and the heart.Neuromuscul Disord 2010;20:479–92.

118. Marston SB. How do mutations in contractileproteins cause the primary familial cardiomyopa-thies? J Cardiovasc Transl Res 2011;4:245–55.

119. Long C, Amoasii L, Mireault AA, et al. Post-natal genome editing partially restores dystrophinexpression in a mouse model of muscular dystro-phy. Science 2016;351:400–3.

120. Long C, McAnally JR, Shelton JM, et al. Pre-vention of muscular dystrophy in mice by CRISPR/Cas9-mediated editing of germline DNA. Science2014;345:1184–8.

121. Nelson CE, Hakim CH, Ousterout DG, et al.In vivo genome editing improves muscle functionin a mouse model of Duchenne muscular dystro-phy. Science 2016;351:403–7.

122. Tabebordbar M, Zhu K, Cheng JK, et al.In vivo gene editing in dystrophic mouse muscleand muscle stem cells. Science 2016;351:407–11.

123. Corrado D, Basso C, Thiene G, et al. Spectrumof clinicopathologic manifestations of arrhythmo-genic right ventricular cardiomyopathy/dysplasia:a multicenter study. J Am Coll Cardiol 1997;30:1512–20.

124. Dunn KE, Ashley EA. Arrhythmogenic rightventricular cardiomyopathy: toward a modern

clinical and genomic understanding. Circ Car-diovasc Genet 2015;8:421–4.

125. Lombardi R, Marian AJ. Arrhythmogenic rightventricular cardiomyopathy is a disease of cardiacstem cells. Curr Opin Cardiol 2010;25:222–8.

126. Thiene G. The research venture in arrhyth-mogenic right ventricular cardiomyopathy: aparadigm of translational medicine. Eur Heart J2015;36:837–46.

127. Mast TP, Teske AJ, vd Heijden JF, et al.Left ventricular involvement in arrhythmogenicright ventricular dysplasia/cardiomyopathyassessed by echocardiography predicts adverseclinical outcome. J Am Soc Echocardiogr 2015;28:1103–13.e9.

128. Groeneweg JA, Bhonsale A, James CA, et al.Clinical presentation, long-term follow-up, andoutcomes of 1001 arrhythmogenic right ventricu-lar dysplasia/cardiomyopathy patients and familymembers. Circ Cardiovasc Genet 2015;8:437–46.

129. Corrado D, Basso C, Schiavon M, et al.Screening for hypertrophic cardiomyopathy inyoung athletes. N Engl J Med 1998;339:364–9.

130. Tabib A, Loire R, Chalabreysse L, et al. Cir-cumstances of death and gross and microscopicobservations in a series of 200 cases of suddendeath associated with arrhythmogenic rightventricular cardiomyopathy and/or dysplasia.Circulation 2003;108:3000–5.

131. Pilmer CM, Kirsh JA, Hildebrandt D, et al.Sudden cardiac death in children and adolescentsbetween 1 and 19 years of age. Heart Rhythm2014;11:239–45.

132. Semsarian C, Ingles J, Wilde AA. Sudden car-diac death in the young: the molecular autopsyand a practical approach to surviving relatives. EurHeart J 2015;36:1290–6.

133. McKoy G, Protonotarios N, Crosby A, et al.Identification of a deletion in plakoglobin inarrhythmogenic right ventricular cardiomyopathywith palmoplantar keratoderma and woolly hair(Naxos disease). Lancet 2000;355:2119–24.

134. Protonotarios N, Tsatsopoulou A. Naxos dis-ease and Carvajal syndrome: cardiocutaneous dis-orders that highlight the pathogenesis andbroaden the spectrum of arrhythmogenic rightventricular cardiomyopathy. Cardiovasc Pathol2004;13:185–94.

135. Marcus FI. Epsilon waves aid in the prognosisand risk stratification of patients with ARVC/D.J Cardiovasc Electrophysiol 2015;26:1211–2.

136. Platonov PG, Calkins H, Hauer RN, et al. Highinterobserver variability in the assessment ofepsilon waves: Implications for diagnosis ofarrhythmogenic right ventricular cardiomyopathy/dysplasia. Heart Rhythm 2016;13:208–16.

137. Burke AP, Farb A, Tashko G, et al. Arrhyth-mogenic right ventricular cardiomyopathy andfatty replacement of the right ventricularmyocardium: are they different diseases? Circula-tion 1998;97:1571–80.

138. Marcus FI, McKenna WJ, Sherrill D, et al.Diagnosis of arrhythmogenic right ventricularcardiomyopathy/dysplasia: proposed modificationof the Task Force Criteria. Eur Heart J 2010;31:806–14.

139. Rampazzo A, Nava A, Malacrida S, et al.Mutation in human desmoplakin domain bindingto plakoglobin causes a dominant form ofarrhythmogenic right ventricular cardiomyopathy.Am J Hum Genet 2002;71:1200–6.

140. Gerull B, Heuser A, Wichter T, et al. Muta-tions in the desmosomal protein plakophilin-2 arecommon in arrhythmogenic right ventricular car-diomyopathy. Nat Genet 2004;36:1162–4.

141. van Tintelen JP, Entius MM, Bhuiyan ZA, et al.Plakophilin-2 mutations are the major determinantof familial arrhythmogenic right ventriculardysplasia/cardiomyopathy. Circulation 2006;113:1650–8.

142. Heuser A, Plovie ER, Ellinor PT, et al. Mutantdesmocollin-2 causes arrhythmogenic right ven-tricular cardiomyopathy. Am J Hum Genet 2006;79:1081–8.

143. Syrris P, Ward D, Evans A, et al. Arrhythmo-genic right ventricular dysplasia/cardiomyopathyassociated with mutations in the desmosomalgene desmocollin-2. Am J Hum Genet 2006;79:978–84.

144. Pilichou K, Nava A, Basso C, et al. Mutationsin desmoglein-2 gene are associated witharrhythmogenic right ventricular cardiomyopathy.Circulation 2006;113:1171–9.

145. Haywood AFM, Merner ND, Hodgkinson KA,et al. Recurrent missense mutations in TMEM43(ARVD5) due to founder effects cause arrhyth-mogenic cardiomyopathies in the UK and Canada.Eur Heart J 2013;34:1002–11.

146. Merner ND, Hodgkinson KA, Haywood AFM,et al. Arrhythmogenic right ventricular cardiomy-opathy type 5 is a fully penetrant, lethalarrhythmic disorder caused by a missense muta-tion in the TMEM43 gene. Am J Hum Genet 2008;82:809–21.

147. Quarta G, Syrris P, Ashworth M, et al. Muta-tions in the Lamin A/C gene mimic arrhythmogenicright ventricular cardiomyopathy. Eur Heart J2012;33:1128–36.

148. van der Zwaag PA, van Rijsingen IAW,Asimaki A, et al. Phospholamban R14del mutationin patients diagnosed with dilated cardiomyopathyor arrhythmogenic right ventricular cardiomyopa-thy: evidence supporting the concept of arrhyth-mogenic cardiomyopathy. Eur J Heart Fail 2012;14:1199–207.

149. Taylor M, Graw S, Sinagra G, et al. Geneticvariation in titin in arrhythmogenic right ventric-ular cardiomyopathy-overlap syndromes. Circula-tion 2011;124:876–85.

150. Beffagna G, Occhi G, Nava A, et al. Regula-tory mutations in transforming growth factor-b3gene cause arrhythmogenic right ventricular car-diomyopathy type 1. Cardiovasc Res 2005;65:366–73.

151. Xiong Q, Cao Q, Zhou Q, et al. Arrhythmo-genic cardiomyopathy in a patient with a rare loss-of-function KCNQ1 mutation. J Am Heart Assoc2015;4:e001526.

152. Norgett EE, Hatsell SJ, Carvajal-Huerta L,et al. Recessive mutation in desmoplakin disruptsdesmoplakin-intermediate filament interactionsand causes dilated cardiomyopathy, woolly hair






































































































































































































2849

and keratoderma. Human Mol Genet 2000;9:2761–6.

153. Alcalai R, Metzger S, Rosenheck S, et al.A recessive mutation in desmoplakin causesarrhythmogenic right ventricular dysplasia, skindisorder, and woolly hair. J Am Coll Cardiol 2003;42:319–27.

154. Chen SN, Gurha P, Lombardi R, et al. TheHippo pathway is activated and is a causal mech-anism for adipogenesis in arrhythmogenic cardio-myopathy. Circ Res 2014;114:454–68.

155. Asimaki A, Tandri H, Huang H, et al. A newdiagnostic test for arrhythmogenic right ventric-ular cardiomyopathy. N Engl J Med 2009;360:1075–84.

156. Garcia-Gras E, Lombardi R, Giocondo MJ,et al. Suppression of canonical Wnt/b-cateninsignaling by nuclear plakoglobin recapitulatesphenotype of arrhythmogenic right ventricularcardiomyopathy. J Clin Invest 2006;116:2012–21.

157. Marian AJ. On the diagnostic utility of junctionplakoglobin in arrhythmogenic right ventricular car-diomyopathy. Cardiovasc Pathol 2013;22:309–11.

158. Heallen T, Zhang M, Wang J, et al. Hippopathway inhibits Wnt signaling to restrain car-diomyocyte proliferation and heart size. Science2011;332:45–61.

159. Lombardi R, da Graca Cabreira-Hansen M,Bell A, et al. Nuclear plakoglobin is essential fordifferentiation of cardiac progenitor cells to adi-pocytes in arrhythmogenic right ventricular car-diomyopathy. Circ Res 2011;109:1342–53.

160. Lombardi R, Dong J, Rodriguez G, et al.Genetic fate mapping identifies second heart fieldprogenitor cells as a source of adipocytes inarrhythmogenic right ventricular cardiomyopathy.Circ Res 2009;104:1076–84.

161. Lombardi R, Chen SN, Ruggiero A, et al. Car-diac fibro-adipocyte progenitors express desmo-some proteins and preferentially differentiate toadipocytes upon deletion of the desmoplakingene. Circ Res 2016;119:41–54.

162. Mertens C, Hofmann I, Wang Z, et al. Nuclearparticles containing RNA polymerase III complexesassociated with the junctional plaque proteinplakophilin 2. Proc Natl Acad Sci U S A 2001;98:7795–800.

163. Sommariva E, Brambilla S, Carbucicchio C,et al. Cardiac mesenchymal stromal cells are asource of adipocytes in arrhythmogenic cardio-myopathy. Eur Heart J 2016;37:1835–46.

164. Ellawindy A, Satoh K, Sunamura S, et al. Rho-kinase inhibition during early cardiac developmentcauses arrhythmogenic right ventricular cardio-myopathy in mice. Arterioscler Thromb Vasc Biol2015;35:2172–84.

165. Gurha P, Chen X, Lombardi R, et al. Knock-down of plakophilin 2 downregulates miR-184through CpG hypermethylation and suppressionof the E2F1 pathway and leads to enhanced adi-pogenesis in vitro. Circ Res 2016;119:731–50.

166. Asimaki A, Kapoor S, Plovie E, et al. Identifi-cation of a new modulator of the intercalated discin a zebrafish model of arrhythmogenic cardio-myopathy. Sci Transl Med 2014;6:240ra74.

167. Chelko SP, Asimaki A, Andersen P, et al.Central role for GSK3b in the pathogenesis ofarrhythmogenic cardiomyopathy. JCI Insight 2016;1:e85923.

168. Sternberg SH, Doudna JA. Expanding thebiologist’s toolkit with CRISPR-Cas9. Mol Cell2015;58:568–74.

169. Yang L, Jiang J, Drouin LM, et al.A myocardium tropic adeno-associated virus (AAV)evolved by DNA shuffling and in vivo selection.Proc Natl Acad Sci U S A 2009;106:3946–51.

170. Ruozi G, Bortolotti F, Falcione A, et al.AAV-mediated in vivo functional selection oftissue-protective factors against ischaemia. NatCommun 2015;6:7388.

171. Ousterout DG, Kabadi AM, Thakore PI, et al.Multiplex CRISPR/Cas9-based genome editing forcorrection of dystrophin mutations that causeDuchenne muscular dystrophy. Nat Commun 2015;6:6244.

172. Pardo B, Gómez-González B, Aguilera A.DNA repair in mammalian cells: DNA double-strand break repair: how to fix a brokenrelationship. Cell Mol Life Sci 2009;66:1039–56.

173. Zetsche B, Gootenberg JS, Abudayyeh OO,et al. Cpf1 is a single RNA-guided endonuclease ofa class 2 CRISPR-Cas system. Cell 2015;163:759–71.

KEY WORDS cardiomyopathy, mutation,noncoding RNA































































































genetics and genomics of single-gene cardiovascular diseases · genetics and genomics of...

Documents