Population-prevalent desmosomal mutations predisposing toarrhythmogenic right ventricular cardiomyopathyAnnukka M. Lahtinen, MSc,*† Eero Lehtonen, MD, PhD,‡§ Annukka Marjamaa, MD, PhD,*†
Maija Kaartinen, MD, PhD,� Tiina Heliö, MD, PhD,� Kimmo Porthan, MD,� Lasse Oikarinen, MD, PhD,�
Lauri Toivonen, MD, PhD,� Heikki Swan, MD, PhD,� Antti Jula, MD, PhD,¶
Leena Peltonen, MD, PhD,#**††§§ Aarno Palotie, MD, PhD,#**††‡‡ Veikko Salomaa, MD, PhD,¶
Kimmo Kontula, MD, PhD*†
From the *Research Program for Molecular Medicine, Biomedicum Helsinki, University of Helsinki, Helsinki, Finland;†Department of Medicine, University of Helsinki, Helsinki, Finland; ‡Department of Pathology, University of Helsinki, Helsinki,Finland; §Laboratory Animal Centre, University of Helsinki, Helsinki, Finland; �Department of Cardiology, University of
elsinki, Helsinki, Finland; ¶National Institute for Health and Welfare, Helsinki, Finland; #Wellcome Trust Sanger Institute,Wellcome Trust Genome Campus, Cambridge, UK; **Institute for Molecular Medicine Finland, University of Helsinki,Helsinki, Finland; ††The Broad Institute of MIT and Harvard, Boston, Massachusetts; ‡‡Department of Medical Genetics,
niversity of Helsinki and Helsinki University Hospital, Helsinki, Finland; §§Deceased.
t3o02
Ctat
Km
Ad
BACKGROUND Arrhythmogenic right ventricular cardiomyopathy(ARVC) is a progressive myocardial disorder caused by mutations ofdesmosomal cell adhesion proteins. The prevalence of these vari-ants in the general population is unknown.
OBJECTIVE This study examined the spectrum and populationprevalence of desmosomal mutations predisposing to ARVC inFinland.
METHODS We screened 29 Finnish ARVC probands for mutationsin the DSP, DSG2, and DSC2 genes. All Finnish-type ARVC-associ-ated mutations, including those 3 previously identified in PKP2 inthe same patient group, were analyzed in the population-basedHealth 2000 cohort of 6,334 individuals and tested for associationwith electrocardiographic variables.
RESULTS We detected 2 novel mutations: DSG2 3059_3062delAGAGand DSP T1373A. DSG2 3059_3062delAGAG was present in a family
ith 5 mutation carriers. The endomyocardial samples of the DSG2eletion carrier showed reduced immunoreactive signal for des-oglein-2, plakophilin-2, plakoglobin, and desmoplakin. DSP1373A was found in 1 proband with typical right ventricularisease and exercise-related ventricular tachycardia. In the pop-
ified in the 29 ARVC patients (PKP2 Q62K, Q59L, N613K, DSG2059_3062delAGAG, and DSP T1373A) was 31 of 6,334 individuals,r 0.5%. The apparent founder mutation PKP2 Q59L is present in.3% of Finns and was previously shown to have an approximately0% disease penetrance.
ONCLUSION One of 200 Finns carries a desmosomal mutationhat may predispose to ARVC and its clinical sequelae. ARVC-ssociated mutations may thus be more prevalent in the popula-ion than expected based on the published ARVC prevalence data.
EYWORDS Arrhythmia; Arrhythmogenic right ventricular cardio-yopathy; Cell adhesion; Desmosome; Genetics
BBREVIATIONS ARVC � arrhythmogenic right ventricular car-iomyopathy; DNA � deoxyribonucleic acid; DSC2 � desmocol-
lation sample, the collective prevalence of all 5 mutations iden- All rights reserved.
This study was supported by research grants from the Academy of Finlandto Dr. Kontula and Dr. Salomaa (grant number 129494); the Sigrid JuséliusFoundation to Dr. Kontula; the Finnish Foundation for Cardiovascular Re-search to Drs. Kontula, Heliö, and Porthan; the Finnish Cultural Foundation toDr. Lahtinen; the Emil Aaltonen Foundation to Dr. Marjamaa; the FinnishMedical Foundation to Dr. Heliö; the Aarne Koskelo Foundation to Dr.Porthan; and the special governmental subsidy for health sciences research toDr. Heliö. Address reprint requests and correspondence: Dr. KimmoKontula, Department of Medicine, University of Helsinki, FIN-00290Helsinki, Finland. E-mail address: [email protected] (Received Feb-
IntroductionArrhythmogenic right ventricular cardiomyopathy (ARVC)may cause severe ventricular arrhythmias and is a prev-alent cause of sudden cardiac death among young ath-letes.1,2 In this disorder, the right ventricular myocardiums progressively replaced by adipose and fibrous tissue.1,3
Recently, biventricular and predominantly left ventricu-lar forms have also been recognized.4 The current diag-
ostic criteria consist of categories for ventricular struc-ure, tissue replacement, electrocardiographic (ECG)
bnormalities of repolarization and depolarization, ar-
1215Lahtinen et al Population-Prevalent ARVC Mutations
rhythmias, and family history.5 The estimated prevalencef ARVC is 1:1,000 to 1:5,000.6,7
In approximately half of all cases, ARVC is familial.3,8,9
Inheritance is usually dominant with markedly reduced pen-etrance, although compound heterozygous and digenicheterozygous mutation carriers may have a higher risk ofdeveloping ARVC.10–13 Rare recessive syndromic formsnvolving hair, skin, and cardiac abnormalities have alsoeen recognized.14 The most common genetic causes of
ARVC are mutations in desmosomal proteins: plakophilin-2(encoded by PKP2),15 desmoplakin (DSP),16 plakoglobin(JUP),17 desmoglein-2 (DSG2),18,19 and desmocollin-2DSC2).20 These mutations lead to cell adhesion defects by
disrupting the proper organization of desmosomal junctionsor may interfere with Wnt signaling pathways through re-distribution of plakoglobin from junctional to cytoplasmic/nuclear pools.21 Nondesmosomal genes associated withARVC pathogenesis have also been identified, includingthose encoding cardiac ryanodine receptor,22 transformingrowth factor �3,23 transmembrane protein 43,24 and
desmin.25
We have previously reported PKP2 mutations to accountfor approximately 10% of Finnish ARVC patients.10 Ourim was to supplement the molecular genetic studies of theame 29 ARVC probands by search for DSP, DSG2, and
DSC2 mutations. The cohort was not screened for mutationsin JUP because although previously implicated in the patho-genesis of ARVC, they seem to be extremely rare.12,13 Theprevalence and phenotypic associations of those desmo-somal gene mutations and variants that could potentially berelated to disease development and expression were inves-tigated in a large population sample (n � 6,334).
MethodsPatients and controlsDesmosomal screening was performed in 29 consecutiveARVC patients described previously10 and diagnosed ac-ording to the International Task Force26 between 1998 and004 at the Department of Cardiology, University of Hel-inki, Finland.27 Control deoxyribonucleic acid (DNA)amples of �250 blood donors (kindly provided by Dr. Tomrusius, The Finnish Red Cross Blood Service, Finland)
epresented the background population.
General population sampleThe National Public Health Institute collected a nationallyrepresentative 2-stage cluster sample (N � 8,028) from theFinnish population (age �30) (http://www.terveys2000.fi/).This Health 2000 survey included DNA samples (n �6,334), ECGs (n � 6,299), clinical examinations (n �6,771), and extensive health questionnaires. The mutationcarriers were followed up until the end of 2008 (meanfollow-up 7.6 � 1.4 years) through the National Hospital
ischarge Register, the National Causes-of-Death Register,nd the National Drug Reimbursement Register. Self-re-orted arrhythmia was defined as knowledge of prior man-
festation or demonstration of arrhythmias at past physical w
xamination. Clinical assessment of arrhythmia and heartailure was based on an interview of the participant, clinicalxamination, ECG, previous medical records, and informa-ion on current medication. This study complied with theeclaration of Helsinki, and all participants gave their writ-
en informed consent. The study was approved by the Ethicsommittees of the Hospital District of Helsinki and Uusi-aa and of the National Public Health Institute.Digital standard 12-lead ECGs were recorded with Mar-
uette MAC 5000, and heart rate with QT Guard softwareGE Marquette Medical Systems, Milwaukee, WI). QT in-erval and QRS duration were assessed from a medianRS-T complex with custom-made software based on areviously validated algorithm,28 and the ECG measure-ents were reviewed by K.P. in a blinded fashion. Theean QT interval (from QRS onset to T-wave end) of all 12
eads was used in the current study. The PR interval wasalculated from a median complex with Magellan softwareGE Marquette Medical Systems).
All 6,334 subjects with DNA samples available werencluded in molecular genetic analyses and calculation ofutation prevalences. However, for the analyses of ECG
ariables, subjects with pacemaker (n � 12) or completeight bundle branch block (RBBB, n � 75) or left bundleranch block (LBBB, n � 68) were excluded. Subjects withotentially QRS-altering (n � 65), QT-prolonging (n �44),29 and heart rate– or PR-altering medication (n �
1,028) were excluded from the analyses of the respectiveECG measurements.
Molecular genetic analysesIn the DNA samples of the 29 ARVC probands, all pro-tein-coding exons and exon-intron junctions of DSP I(NM_004415), DSG2 (NM_001943), DSC2a (NM_024422),nd DSC2b (NM_004949) were amplified by polymerasehain reaction and sequenced with BigDye Terminator v3.1nd ABI 3730 DNA Analyzer (Applied Biosystems, Fosterity, CA). Large deletions were searched for using the Mul-
iplex Ligation-dependent Probe Amplification (MLPA) kit168 (MRC-Holland, Amsterdam, The Netherlands). ARVC-ssociated mutations were screened for in 250 blood donorontrol samples, using restriction enzyme assays.
All 5 Finnish ARVC mutations and 2 PKP2 polymor-hisms were genotyped in all available DNA samples of theopulation material (n � 6,334) using MALDI-TOF masspectrometry (Mass Array Compact Analyzer, Sequenomnc., San Diego, CA). Hardy-Weinberg equilibrium wasonfirmed by �2 test or Fisher’s exact test.
Histopathologic and electron microscopic analysesEndomyocardial biopsy and aneurysm resection samples ofan ARVC patient were compared with endomyocardialsamples from 2 non-ARVC patients (with hypertrophic car-diomyopathy and catecholaminergic polymorphic ventric-ular tachycardia [VT], respectively). Samples were pro-cessed as described previously.10 Immunohistochemistry
moplakin 1/2 (Progen, Heidelberg, Germany), mouse N-cadherin (Sigma-Aldrich, St. Louis, MO), and rabbit con-nexin-43 antibodies (Sigma-Aldrich) was performed usingthe Vectastain Elite ABC kit (Vector Laboratories, Burlin-game, CA). The immunoreaction was quantified using Al-exa Fluor 594 detection (Invitrogen, Carlsbad, CA), ZeissAxiophot 2 microscope, and a Qimaging Retiga-4000Rcamera. The integrated signal density was measured in theareas of immunopositive signal using the ImageJ program.The mean signal intensity per total cell area in 3 to 4 fieldsfrom ARVC myocardium was compared with an equivalentnumber of fields from control myocardium. For electronmicroscopy, endomyocardial tissue was retrieved from par-affin blocks and examined in a Jeol JEM 1400 electronmicroscope. The analysis included 44 high-power fieldsfrom ARVC samples and 24 fields from control myocar-dium.
Statistical analysesNormality of distributions for continuous variables wasevaluated visually. In linear regression, normally distributedPR interval, QRS duration, and QT interval were adjustedfor age, sex, and heart rate. Heart rate was adjusted for ageand sex. PKP2 polymorphisms L366P and I531S weretested by additive model (number of minor alleles coded as0, 1, or 2). Discrete variables (occurrence of arrhythmia,T-wave inversion in leads V2-V3, and heart failure) weretested by �2 test or Fisher’s exact test. The prevalencestimates were calculated from the weighted study popula-ion as described previously.29 Statistical analyses were
performed with SPSS 15.0/16.0 (SPSS Inc., Chicago, IL).Two-tailed P �.05 was considered statistically significant.
ResultsNovel desmosomal gene mutations in ARVCfamiliesScreening of 29 ARVC probands for mutations in DSP,DSG2, and DSC2 revealed 1 ARVC patient (3%) withDSG2 3059_3062delAGAG and another (3%) with DSP
1373A. These 2 mutations are novel and were not presentn 250 control samples. No mutations were detected inSC2. The occurrence of large chromosomal rearrange-
Figure 1 Family D with DSG2 3059_3062delAGAG (A) and electropheroutation carriers (�) fulfill either the Task Force criteria (solid symbols
�) are unaffected (open symbols). Squares symbolize males, circles femre shown (B). N/A � no genetic or clinical information available.
ents in PKP2 was excluded by MLPA (see Online Sup-
plementary Figure 1). Because 3 of the 29 probands werepreviously found to carry PKP2 mutations,10 a total of 5patients (17%) carried desmosomal gene mutations in theFinnish ARVC sample. Several polymorphisms occurred inthe desmosomal genes (see Online Supplementary Table 1).
The DSG2 3059_3062delAGAG mutation is predicted toresult in frameshift (E1020AfsX18) abolishing 99 carboxy-terminal amino acids of desmoglein-2. The proband II:1 infamily D (Figure 1, Table 1) was a heterozygous carrier of3059_3062delAGAG, as well as the PKP2 polymorphismsE58D, L366P, and I531S. Starting at age 24, he experiencedrepeated VT paroxysms. Examinations revealed VT withLBBB morphology, T-wave inversion in leads V1-V3, latepotentials in signal-averaged ECG, right ventricular dilata-tion, and an aneurysm in the right ventricular outflow tract.Four additional family members carried DSG2 3059_3062delAGAG. Two of them (II:5 and II:6) fulfilled theproposed modified criteria for first-degree family members,8
and 1 (II:3) had minor repolarization abnormalities. Allnoncarriers in family D were unaffected.
Proband E was heterozygous for the novel DSP T1373A4117A�G) missense substitution. The mutated amino acid1373 is conserved across different species (Figure 2). Atge 29, proband E, a male athlete, experienced exercise-elated VT. Clinical examination revealed epsilon waves,BBB-type VT, and enlarged right ventricle, with no signsf left ventricular involvement. No family members werevailable.
Histopathologic and electron microscopic analysesThe endomyocardial samples from the proband with DSG23059_3062delAGAG showed moderate fatty infiltration andmild fibrosis. In immunohistochemical staining (Figure 3),these ARVC samples showed reduced immunoreactive sig-nal for desmoglein-2 (99% reduction of signal density),plakophilin-2 (94% reduction), plakoglobin (91% reduc-tion), and desmoplakin (70% reduction) compared with thecontrol samples. No significant differences were noticed instainings for the gap junctional protein connexin-43.
For transmission electron microscopy, endomyocardialtissue was retrieved from paraffin blocks. Compared withthe non-ARVC samples (Figure 4A), the intercalated disk
a wild type and a heterozygote (B). Three of 5 DSG2 3059_3062delAGAGmodified criteria for family members (shaded symbols). All noncarriersnd crossed symbols deceased individuals. Partial amino acid sequences
gram of) or the
ales, a
regions tended to appear more vacuolated in the ARVC
TsbQ
m
ccfEQipadrdohbt
tscLirt
ED
1217Lahtinen et al Population-Prevalent ARVC Mutations
samples (Figure 4B). No consistent differences in the num-ber or general organization of the junctions, such as lengthof desmosomes or distance between the 2 plaques, wereobserved. However, the junctions and the associated fila-ments sometimes appeared disorganized in the ARVC sam-ples.
Population prevalence and phenotypicassociations of the desmosomal mutationsThe prevalence and related ECG phenotypes of FinnishARVC-related desmosomal gene alterations (PKP2 Q62K,Q59L, and N613K,10 DSG2 3059_3062delAGAG, and DSP
1373A) were investigated in a large Finnish populationample (n � 6,334). Functional evidence for causality haseen reported in epithelial cell lines for PKP2 Q59L and62K30 and in histopathologic studies for PKP2 N613K,
Q62K,10,31 and DSG2 3059_3062delAGAG (the present
Table 1 ARVC diagnostic criteria according to McKenna et al26
ARVC � arrhythmogenic right ventricular cardiomyopathy; ND � not*Age at clinical examination (years). †T-wave inversion in leads V1-V2 and§Ventricular extrasystoles (�500/24 h).
Figure 2 Location of DSP T1373A and conservation of the mutatedamino acid. T1373A is located in the desmoplakin I rod domain involvedin homodimerization. The region is evolutionary conserved, and the mu-tated amino acid is invariant between different species. Sequences arealigned according to UCSC Genome Browser Vertebrate Multiz Alignment
& Conservation. Frog denotes Xenopus tropicalis.
study). We also investigated the PKP2 polymorphismsL366P and I531S10 as possible disease-modifying factors.
The mean age (54 � 16) and gender distribution (52%ale) of desmosomal mutation carriers (n � 31) were
similar to those of the noncarriers (mean age 53 � 15, 45%male). PKP2 Q59L was the most prevalent mutation with aarrier frequency of 0.3% (Table 2). Clustering of Q59Larriers in the southeastern part of Finland suggests aounder effect (see Online Supplementary Figure 2). InCG, inverted T waves in V2-V3 occurred in 2 of the 1959L carriers and a widened QRS complex (�110 ms) in 4,
n 1 due to RBBB. Heart failure was by clinical assessmentresent in 2 (11%) Q59L carriers (vs. 2.2% of noncarriers)nd 1 additional carrier developed left ventricular failureuring follow-up. Five (26%) carriers had self-reported ar-hythmia (vs. 15% of noncarriers). One of them had aiagnosis of VT and died suddenly out of hospital at the agef 46 years. Medicolegal autopsy suggested obstructiveypertrophic cardiomyopathy as the cause of death, butecause a written document was not available for review,here remains the possibility of a misdiagnosis.
PKP2 Q62K has a prevalence of approximately 0.1% inhe Finnish population (Table 2). One of 6 carriers hadelf-reported arrhythmia, and ECG recording of anotherarrier showed frequent premature ventricular complexes ofBBB morphology. Age-, sex-, and heart rate–adjusted QT
nterval was 14 ms longer in Q62K carriers than in noncar-iers (P � .16, Table 3). PKP2 N613K was not present inhe population sample.
Only 1 DSG2 3059_3062delAGAG carrier was detectedin the population sample (prevalence estimate 0.02%). TheECG of this subject, aged 55, showed LBBB with border-line prolonged PR interval (200 ms).
The prevalence of DSP T1373A was estimated to be0.08% in the population (Table 2). Two of 5 T1373Acarriers had self-reported arrhythmia, and 1 had a clinicaldiagnosis of paroxysmal tachycardia and was treated with
ned (no biopsy).Not applicable due to complete or incomplete right bundle branch block.
in the
polarinorma
ajor M
����������
determiaVR. ‡
propranolol medication. Age-, sex-, and heart rate–adjusted
a(w(
igP
do
i
3tpttt
atIthe
t3lptqc
1218 Heart Rhythm, Vol 8, No 8, August 2011
PR interval was 33 ms longer in T1373A carriers comparedwith noncarriers (P �.005, Table 3).
The minor allele frequencies of the PKP2 polymor-phisms L366P (19.5%) and I531S (2.2%) did not differfrom those in our ARVC sample (17% and 2%, respec-tively).10 In linear regression, L366P was associated withPR-interval duration (P � .036), whereas no significantssociation was detected with QT-interval durationP �.074, Table 3). Genotype-associated differencesere also detected in the occurrence of arrhythmias
Figure 3 Immunostaining of desmoglein-2 (A–B), plakophilin-2 (C–D),plakoglobin (E–F), desmoplakin (G–H), and N-cadherin (I–J) at the in-ercalated disks of endomyocardial samples. The ARVC patient (DSG2059_3062delAGAG) shows reduced immunoreactive signal for desmog-ein-2 (B), plakophilin-2 (D), plakoglobin (F), and desmoplakin (H) com-ared with a non-ARVC subject (A, C, E, G). Immunoreactive signal forhe nondesmosomal adhesion protein N-cadherin, used to control for tissueuality, was strong in the ARVC samples (J) and indistinguishable fromontrols (I). Bar � 50 �m.
6.2% in the genotype L366L, 4.6% in L366P, and 4.3% I
n P366P, P �.028) and inverted T waves (0.6% in theenotype L366L, 0.2% in L366P, and 0.0% in P366P,� .040).
DiscussionOur study is the first to directly address the prevalence ofARVC-predisposing mutations at the population level. Thepathophysiologic role of the investigated mutations wassupported by their biochemical nature, previously publishedfunctional studies, or associated histopathological altera-tions. Our representative sample of the general population(n � 6,334) showed a remarkably higher prevalence ofesmosomal mutation carriers (1:200) than expected basedn the estimated ARVC prevalence (1:1,000 to 1:5,000).6,7
The occurrence of founder mutations may be particularlystriking in the genetically isolated Finnish population,which also shows a high prevalence (1:250) of foundermutations underlying another arrhythmic disorder, the longQT syndrome.29 Our data also support the findings of re-duced penetrance observed in clinical ARVC materials.10,32
However, our study importantly demonstrates that a greatnumber of individuals may be at risk of developing ARVC-related myocardial alterations and severe arrhythmia.
DSG2 and DSP mutationsDSG2 3059_3062delAGAG cosegregated with cardiac ab-normalities in family D, as all noncarriers were healthy and60% of mutation carriers fulfilled the Task Force or modi-fied criteria for ARVC. The only mutation carrier in thepopulation sample had a conduction defect. This carboxy-terminal deletion leads to truncation of the desmoglein-specific cytoplasmic region including the end of the re-peated unit domain, with a putative protein kinase Cphosphorylation site, and the desmoglein-specific terminaldomain.33 The corresponding region of desmoglein-1 isnvolved in interaction with other desmosomal proteins.34
DSG2 3059_3062delAGAG may thus induce disintegrationof the desmosomal structure, as supported by the reducedimmunoreactive signal for desmoglein-2, plakophilin-2,plakoglobin and desmoplakin, and looser and often disor-ganized intercalated disk areas in the proband’s biopsysamples. Because the immunoreactive signal for desmog-lein-2 was reduced by over 50%, it is possible that DSG2059_3062delAGAG acts in a dominant negative way in-erfering with the degradation or localization of the proteinroduct of the wild-type allele. Connexin-43 immunoreac-ive signal showed no reduction, but this does not excludehe possibility that gap junctional remodeling could occur inhe myocardium of the DSG2 deletion carrier.
Only 1 ARVC patient (3%) carried a potentially disease-ssociated desmoplakin variant, DSP T1373A. This substi-ution has a prevalence of 0.08% in the Finnish population.t changes a polar threonine residue to nonpolar alanine inhe central coiled-coil rod domain involved in desmoplakinomodimerization.35 The population sample showed somevidence for a role of DSP T1373A in risk of tachycardias.
n addition, T1373A was associated with PR-interval pro-
pppafr
de
ct
.cwg
pi
P
1219Lahtinen et al Population-Prevalent ARVC Mutations
longation, an abnormality in the atrioventricular conduction.However, the exact functional significance of this variantremains unclear. It might require another trigger, such asphysical exertion in the case of our proband, for diseaseprogression.
PKP2 mutations and polymorphismsApproximately 0.3% of Finns carry the proposed foundermutation PKP2 Q59L. It has been shown to disrupt theinteraction between plakophilin-2 and desmoplakin invitro.30 We previously reported PKP2 Q59L in 2 ARVCfamilies with a penetrance of 20%.10 Data on its possiblehenotypic effects did not reach statistical significance,robably due to low penetrance. However, Q59L mightredispose the carriers to heart failure, cardiac arrhythmias,nd even sudden cardiac death constituting a potential riskor the 0.3% of the population carrying the mutation. Itemains to be seen whether similar population-prevalent
Figure 4 Electron micrographs showing intercalated disk regions inatient (B, DSG2 3059_3062delAGAG) show relatively well preservedn the ARVC sample (asterisks in B). Note the disorganized structure (
material. Bar � 500 nm.
Table 2 Prevalence of desmosomal mutations in the population
Gene Amino acid Nucleotide Heter
KP2 Q59L 176A�T 19Q62K 184C�A 6N613K 1839C�G 0
DSP T1373A 4117A�G 5DSG2 E1020AfsX18 3059_3062delAGAG 1Total of all mutation carriers 31
CI � confidence interval.
esmosomal gene mutations associated with ARVC occurlsewhere in the world.
PKP2 Q62K has been reported as an unclassified variant inompound heterozygous patients.9,10,13 We previously de-ected this variant together with PKP2 N613K.10 In epithelial
cells, plakophilin-2 with Q62K was degraded more rapidlythan wild-type protein and was unable to recruit desmoplakinto desmosomal sites.30 In our population sample, Q62K asso-ciated with a suggestive 14 ms QT-interval prolongation (P �16). Further evidence is needed to decide whether this variantan cause ARVC by itself. We did not find a single individualith PKP2 N613K in our sample of 6,334 individuals, sug-esting that it is a very rare ARVC-causing mutation.
In addition, 2 common PKP2 polymorphisms (L366Pand I531S)10 were studied in the population sample. Inter-estingly, the P366 minor allele appeared as a putative pro-tective variant in that it was associated less frequently (P �
yocardial samples. Both the non-ARVC subject (A) and the ARVCl junction arrays. The intercalated disk region appears more vacuolatedin B), apparently consisting of junctional plaque and attached filament
.028) with arrhythmias as assessed by clinical examinations.The mechanism of this association remains unknown, al-though L366P variation may profoundly affect the structureof plakophilin-2, because proline is known to interrupt thehelix formation in protein structures. It will be important toobtain additional data on the possible phenotypic associa-tions of these 2 PKP2 polymorphisms in other populations.
Study limitationsThe major limitation of the present study is the limitedphenotypic information of the population sample. AlthoughECGs were available for all mutation carriers, the studysubjects did not undergo extensive cardiologic examina-tions. Because the population sample comprised only adults�30 years, the exact prevalence in children (possibly evenhigher than in surviving adults) cannot be estimated. Inaddition, we cannot exclude the occurrence of disease-caus-ing intronic or regulatory mutations in our ARVC probands.Because the population sample was only screened for theknown Finnish desmosomal mutations and variants, occur-rence of other ARVC-causing mutations cannot be ex-cluded.
ConclusionIn summary, the present study describes novel desmosomalmutations in ARVC patients and for the first time reports theprevalence of mutations predisposing to ARVC in a randompopulation sample. We show that in Finland, the prevalenceof ARVC-related desmosomal mutations is considerablyhigher (approximately 1:200) than the estimated ARVCprevalence (1:1,000 to 1:5,000).6,7 Remarkably, the PKP2
59L mutation, with substantial evidence favoring itsathogenic role, is present in 1 of 330 Finns. These desmo-
Table 3 Electrocardiographic variables (mean � SE) in each ge
Mutation/variant and genotype Heart rate (beats/min)
QTc � QT interval corrected for heart rate according to the Bazett for*With left bundle branch block. †P � .005 in linear regression. ‡P � .03
omal mutations may predispose to ARVC and ensuing
ymptoms with reduced penetrance and variable expressiv-ty. We believe that our data further encourage screening foresmosomal gene mutations in patients with symptoms re-embling ARVC and families with cases of sudden cardiaceath.
AcknowledgementsThe authors thank Susanna Saarinen, Hanna Nieminen, SiniWeckström, Pirkko Alha, Sirkka Rinne, Harri Rissanen,Ulla Kiiski, and Svetlana Zueva for expert technical assis-tance, and Päivi Lahermo and Mari Kaunisto for help inSequenom genotyping. Heikki Väänänen and Juhani Dabekare acknowledged for the ECG measurement software.
Supplementary dataSupplementary data associated with this article can befound, in the online version, at doi:10.1016/j.hrthm.2011.03.015.
References1. Thiene G, Nava A, Corrado D, Rossi L, Pennelli N. Right ventricular cardio-
myopathy and sudden death in young people. N Engl J Med 1988;318:129–133.2. Tabib A, Loire R, Chalabreysse L, et al. Circumstances of death and gross and
microscopic observations in a series of 200 cases of sudden death associatedwith arrhythmogenic right ventricular cardiomyopathy and/or dysplasia. Circu-lation 2003;108:3000–3005.
3. Nava A, Thiene G, Canciani B, et al. Familial occurrence of right ventriculardysplasia: a study involving nine families. J Am Coll Cardiol 1988;12:1222–1228.
4. Sen-Chowdhry S, Syrris P, Prasad SK, et al. Left-dominant arrhythmogeniccardiomyopathy: an under-recognized clinical entity. J Am Coll Cardiol 2008;52:2175–2187.
5. Marcus FI, McKenna WJ, Sherrill D, et al. Diagnosis of arrhythmogenic rightventricular cardiomyopathy/dysplasia: proposed modification of the task forcecriteria. Circulation 2010;121:1533–1541.
6. Rampazzo A, Nava A, Danieli GA, et al. The gene for arrhythmogenic right
1221Lahtinen et al Population-Prevalent ARVC Mutations
7. Peters S, Trümmel M, Meyners W. Prevalence of right ventricular dysplasia-cardiomyopathy in a non-referral hospital. Int J Cardiol 2004;97:499–501.
8. Hamid MS, Norman M, Quraishi A, et al. Prospective evaluation of relatives forfamilial arrhythmogenic right ventricular cardiomyopathy/dysplasia reveals aneed to broaden diagnostic criteria. J Am Coll Cardiol 2002;40:1445–1450.
9. van Tintelen JP, Entius MM, Bhuiyan ZA, et al. Plakophilin-2 mutations are themajor determinant of familial arrhythmogenic right ventricular dysplasia/cardio-myopathy. Circulation 2006;113:1650–1658.
10. Lahtinen AM, Lehtonen A, Kaartinen M, et al. Plakophilin-2 missense mutationsin arrhythmogenic right ventricular cardiomyopathy. Int J Cardiol 2008;126:92–100.
11. Bhuiyan ZA, Jongbloed JD, van der Smagt J, et al. Desmoglein-2 and desmo-collin-2 mutations in Dutch arrhythmogenic right ventricular dysplasia/cardio-myopathy patients: results from a multicenter study. Circ Cardiovasc Genet2009;2:418–427.
12. den Haan AD, Tan BY, Zikusoka MN, et al. Comprehensive desmosomemutation analysis in North Americans with arrhythmogenic right ventriculardysplasia/cardiomyopathy. Circ Cardiovasc Genet 2009;2:428–435.
13. Xu T, Yang Z, Vatta M, et al. Compound and digenic heterozygosity contributesto arrhythmogenic right ventricular cardiomyopathy. J Am Coll Cardiol 2010;55:587–597.
14. Sen-Chowdhry S, Syrris P, McKenna WJ. Genetics of right ventricular cardio-myopathy. J Cardiovasc Electrophysiol 2005;16:927–935.
15. Gerull B, Heuser A, Wichter T, et al. Mutations in the desmosomal proteinplakophilin-2 are common in arrhythmogenic right ventricular cardiomyopathy.Nat Genet 2004;36:1162–1164.
16. Rampazzo A, Nava A, Malacrida S, et al. Mutation in human desmoplakindomain binding to plakoglobin causes a dominant form of arrhythmogenic rightventricular cardiomyopathy. Am J Hum Genet 2002;71:1200–1206.
17. Asimaki A, Syrris P, Wichter T, Matthias P, Saffitz JE, McKenna WJ. A noveldominant mutation in plakoglobin causes arrhythmogenic right ventricular car-diomyopathy. Am J Hum Genet 2007;81:964–973.
18. Awad MM, Dalal D, Cho E, et al. DSG2 mutations contribute to arrhythmogenicright ventricular dysplasia/cardiomyopathy. Am J Hum Genet 2006;79:136–142.
19. Pilichou K, Nava A, Basso C, et al. Mutations in desmoglein-2 gene areassociated with arrhythmogenic right ventricular cardiomyopathy. Circulation2006;113:1171–1179.
20. Syrris P, Ward D, Evans A, et al. Arrhythmogenic right ventricular dysplasia/cardiomyopathy associated with mutations in the desmosomal gene desmocol-lin-2. Am J Hum Genet 2006;79:978–984.
21. Garcia-Gras E, Lombardi R, Giocondo MJ, et al. Suppression of canonicalWnt/beta-catenin signaling by nuclear plakoglobin recapitulates phenotype ofarrhythmogenic right ventricular cardiomyopathy. J Clin Invest 2006;116:2012–
2021.
22. Koop A, Goldmann P, Chen SR, Thieleczek R, Varsányi M. ARVC-related
mutations in divergent region 3 alter functional properties of the cardiacryanodine receptor. Biophys J 2008;94:4668–4677.
3. Beffagna G, Occhi G, Nava A, et al. Regulatory mutations in transforminggrowth factor-beta3 gene cause arrhythmogenic right ventricular cardiomyopa-thy type 1. Cardiovasc Res 2005;65:366–373.
4. Merner ND, Hodgkinson KA, Haywood AF, et al. Arrhythmogenic right ven-tricular cardiomyopathy type 5 is a fully penetrant, lethal arrhythmic disordercaused by a missense mutation in the TMEM43 gene. Am J Hum Genet2008;82:809–821.
5. Otten E, Asimaki A, Maass A, et al. Desmin mutations as a cause of rightventricular heart failure affect the intercalated disks. Heart Rhythm 2010;7:1058–1064.
6. McKenna WJ, Thiene G, Nava A, et al. Diagnosis of arrhythmogenic rightventricular dysplasia/cardiomyopathy. Task Force of the Working Group Myo-cardial and Pericardial Disease of the European Society of Cardiology and of theScientific Council on Cardiomyopathies of the International Society and Feder-ation of Cardiology. Br Heart J 1994;71:215–218.
7. Kaartinen M, Heliö T, Lehtonen A, et al. Characterization of familial andsporadic arrhythmogenic right ventricular cardiomyopathy in Finland. Ann Med2007;39:312–318.
8. Oikarinen L, Paavola M, Montonen J, et al. Magnetocardiographic QT intervaldispersion in postmyocardial infarction patients with sustained ventricular tachy-cardia: validation of automated QT measurements. Pacing Clin Electrophysiol1998;21:1934–1942.
9. Marjamaa A, Salomaa V, Newton-Cheh C, et al. High prevalence of four longQT syndrome founder mutations in the Finnish population. Ann Med 2009;41:234–240.
0. Hall C, Li S, Li H, Creason V, Wahl JK 3rd. Arrhythmogenic right ventricularcardiomyopathy plakophilin-2 mutations disrupt desmosome assembly and sta-bility. Cell Commun Adhes 2009;16:15–27.
1. Fidler LM, Wilson GJ, Liu F, et al. Abnormal connexin-43 in arrhythmogenicright ventricular cardiomyopathy caused by plakophilin-2 mutations. J Cell MolMed 2009;13:4219–4228.
2. Dalal D, James C, Devanagondi R, et al. Penetrance of mutations in plakophi-lin-2 among families with arrhythmogenic right ventricular dysplasia/cardiomy-opathy. J Am Coll Cardiol 2006;48:1416–1424.
3. Schäfer S, Koch PJ, Franke WW. Identification of the ubiquitous human des-moglein, Dsg2, and the expression catalogue of the desmoglein subfamily ofdesmosomal cadherins. Exp Cell Res 1994;211:391–399.
4. Kami K, Chidgey M, Dafforn T, Overduin M. The desmoglein-specific cyto-plasmic region is intrinsically disordered in solution and interacts with multipledesmosomal protein partners. J Mol Biol 2009;386:531–543.
5. Green KJ, Parry DA, Steinert PM, et al. Structure of the human desmoplakins.Implications for function in the desmosomal plaque. J Biol Chem 1990;265:
