genetic fate mapping identifies second heart field...

Genetic Fate Mapping Identifies Second Heart FieldProgenitor Cells As a Source of Adipocytes in

Arrhythmogenic Right Ventricular CardiomyopathyRaffaella Lombardi, Jinjiang Dong, Gabriela Rodriguez, Achim Bell, Tack Ki Leung,

Robert J. Schwartz, James T. Willerson, Ramon Brugada, Ali J. Marian

Abstract—The phenotypic hallmark of arrhythmogenic right ventricular cardiomyopathy, a genetic disease of desmosomalproteins, is fibroadipocytic replacement of the right ventricle. Cellular origin of excess adipocytes, the responsiblemechanism(s) and the basis for predominant involvement of the right ventricle are unknown. We generated 3 sets oflineage tracer mice regulated by cardiac lineage promoters �-myosin heavy chain (�MyHC), Nkx2.5, or Mef2C. Weconditionally expressed the reporter enhanced yellow fluorescent protein while concomitantly deleting the desmosomalprotein desmoplakin in cardiac myocyte lineages using the Cre-LoxP technique. Lineage tracer mice showed excessfibroadiposis and increased numbers of adipocytes in the hearts. Few adipocytes in the hearts of �MyHC-regulatedlineage tracer mice, but the majority of adipocytes in the hearts of Nkx2.5- and Mef2C-regulated lineage tracer mice,expressed enhanced yellow fluorescent protein. In addition, rare cells coexpressed adipogenic transcription factors andthe second heart field markers Isl1 and Mef2C in the lineage tracer mouse hearts and in human myocardium frompatients with arrhythmogenic right ventricular cardiomyopathy. To delineate the responsible mechanism, we generatedtransgenic mice expressing desmosomal protein plakoglobin in myocyte lineages. Transgene plakoglobin translocatedto nucleus, detected by immunoblotting and immunofluorescence staining and coimmunoprecipitated with Tcf7l2, acanonical Wnt signaling transcription factor. Expression levels of canonical Wnt/Tcf7l2 targets bone morphogeneticprotein 7 and Wnt5b, which promote adipogenesis, were increased and expression level of connective tissue growthfactor, an inhibitor of adipogenesis, was decreased. We conclude adipocytes in arrhythmogenic right ventricularcardiomyopathy originate from the second heart field cardiac progenitors, which switch to an adipogenic fate becauseof suppressed canonical Wnt signaling by nuclear plakoglobin. (Circ Res. 2009;104:1076-1084.)

Key Words: adipocytes � progenitor cells � Wnt signaling � desmosomes � heart failure

Arrhythmogenic right ventricular cardiomyopathy(ARVC) is a genetic disease characterized by the unique

phenotype of fibroadipocytic replacement of cardiac myo-cytes, predominantly in the right ventricle.1–3 Clinical mani-festations of ARVC include ventricular arrhythmias, typicallyoriginating from the right ventricle; sudden cardiac death,which often is its first manifestation; and right ventricularaneurysmal dilatation and failure.1,4 The left ventricle is alsocommonly involved in the advanced stages.1,5

ARVC is typically an autosomal dominant disease. Reces-sive forms in conjunction with palmoplantar keratoderma andwoolly hair (Naxos disease) or predominant involvement ofthe left ventricle (Carvajal syndrome) are referred to as“cardiocutaneous syndromes.”6,7 Recently, mutations in 5genes for ARVC, namely, DSP, JUP, PKP2, DSC2, andDSG2, encoding desmosomal proteins desmoplakin (Dsp),

plakoglobin (PG), plakophilin 2, desmocollin 2, and desmo-glein 2, respectively, have been identified.6,8 –12 Hence,ARVC, at least in a subset, is a disease of desmosomes,intercellular junction structure responsible for cell-cell adhe-sion in epidermal cells and cardiac myocytes.

Pathogenesis of ARVC is not fully understood. Impairedmyocyte to myocyte attachment because of defective desmo-somes may explain cardiac dysfunction.13 It does not, how-ever, explain the pathogenesis of the unique phenotype offibro-fatty replacement of the myocardium and the cellularorigin of excess adipocytes. Heart is a heterogeneous organ.It contains myocytes, fibroblasts, adipocytes, smooth musclecells, endothelial cells, and pericytes, as well as circulatingcells that implant in the myocardium. We posit the cell typethat gives rise to adipocytes in ARVC must either express themutant desmosomal protein or differentiate into adipocytes

Original received January 5, 2009; resubmission received March 9, 2009; revised resubmission received March 23, 2009; accepted March 26, 2009.From the Center for Cardiovascular Genetics (R.L., J.D., G.R., A.B., J.T.W., A.J.M.), Brown Foundation Institute of Molecular Medicine, The

University of Texas Health Science Center and Texas Heart Institute, Houston; Montreal Heart Institute and Universite de Montreal (T.K.L., R.B.),Quebec, Canada; and Institute of Biotechnology (R.J.S.), Texas A&M University, Houston.

Correspondence to A. J. Marian, MD, Brown Foundation Institute of Molecular Medicine, The University of Texas Health Sciences Center, 6770Bertner St, Suite C900A, Houston, TX 77030. E-mail [email protected]

© 2009 American Heart Association, Inc.

Circulation Research is available at http://circres.ahajournals.org DOI: 10.1161/CIRCRESAHA.109.196899

1076

by guest on June 27, 2018http://circres.ahajournals.org/

Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from

http://circres.ahajournals.org/






















through a paracrine mechanism(s) emanating from cellsexpressing the mutant desmosomal protein. In the heart, theonly cell type known to express desmosomal proteins iscardiac myocyte lineage. Adult cardiac myocytes are termi-nally differentiated and, hence, not plausible candidates todedifferentiate to adipocytes.14 In contrast, cardiac progenitorcells expressing the desmosomal proteins might have thepotential to differentiate to adipocytes. To test this hypothe-sis, we performed genetic fate-mapping experiments usingthe LoxP-Cre technology regulated by 3 cardiac lineagepromoters. We extended the results of lineage tracing exper-iments to hearts of humans with autopsy-proven ARVC. Weshow adipocytes in the lineage tracer mice and in humanhearts originate from the second heart field progenitor cells,which switch to an adipogenic fate because of suppressedcanonical Wnt signaling by nuclear PG.

Materials and MethodsAn expanded Materials and Methods section is available in theonline data supplement at http://circres.ahajournals.org. The inves-tigation conforms to the Guide for the Care and Use of LaboratoryAnimals published by the U.S. National Institutes of Health and wasapproved by the Institutional Animal Care and Use Committee.Human heart samples, stripped from the identifiers, were used in thestudy. The Institutional Review Board approved the study protocol.

Genetically Modified MiceDsp-floxed, Nkx2.5-Cre, Mef2C-Cre, and �MyHC-Cre mice havebeen published.13,15–17 R26-EYFP was a generous gift from DrF. Costantini.18 We generated the PG transgenic mice by cloning aFlag-tagged full-length PG cDNA downstream to the conventional5.5-kbp �-myosin heavy chain (�MyHC) promoter.

M-Mode, 2-Dimensional, andDoppler EchocardiographyWe performed echocardiography in 7 to 12 mice per group undersodium pentobarbital-induced anesthesia.13,19

Gross and Histological Cardiac PhenotypeAn investigator without knowledge of the genotypes analyzedmorphological and histological phenotypes in 5 to 9 age-, sex-, andbody weight–matched mice per group.13,19,20 We calculated myocytecross-sectional area in �18 000 myocytes per mouse by digitalmorphometry of fresh frozen thin myocardial sections stained withTexas red–conjugated wheat germ agglutinin (Molecular Probes Inc,Eugene, Ore).19,20

ImmunoblottingWe electrophoresed 30-�g aliquots of cardiac protein extracts on 12%SDS polyacrylamide gels, transferred the proteins to membranes andprobed them with the specific antibodies. Following stripping, wereprobed the membranes with an anti–�-tubulin antibody.

CoimmunoprecipitationWe performed coimmunoprecipitation as published previously.19 Inbrief, we mixed 6-�g aliquots of anti-Tcf7l2 monoclonal antibody(Millipore–Upstate, Billerica, Mass) with 500-�g aliquots of totalprotein extracts. We precipitated the antibody–protein complex byadding 20-�L aliquots of Protein A/G PLUS-Agarose beads. Theprimary antibodies used in immunoblotting were monoclonal anti–�-catenin (Santa Cruz Biotechnology Inc, Santa Cruz, Calif), anti-FLAG (Sigma, St Louis, Mo), and anti–total (transgene�endoge-nous) PG (Invitrogen–Zymed, Carlsbad, Calif). The secondaryantibody was donkey anti-mouse IgG horseradish peroxidase–con-jugated (Santa Cruz Biotechnology Inc).

Separation of Cardiac Protein SubfractionsWe extracted nuclear, cytosolic and membrane proteins using acommercial kit (Chemicon International, Danvers, Mass).21 In brief,we homogenized the hearts in 5 volumes of a cold buffer containingHEPES (pH 7.9), MgCl2, KCl, EDTA, sucrose, glycerol, sodiumorthovanadate, and protease inhibitors. We collected the supernatantcontaining the cytosolic proteins after centrifugation at 18 000g. Weresuspended the pellets in 100 �L of a cold buffer containingHEPES, MgCl2, NaCl, EDTA, glycerol, sodium orthovanadate, andproteinase inhibitors and centrifuged and collected the supernatantcontaining the nuclear proteins. We resuspended the residual pelletsin 100 �L of a cold buffer containing HEPES, MgCl2, KCl, EDTA,sucrose, glycerol, sodium deoxycholate, NP-40, sodium orthovana-date, and protease inhibitors and centrifuged to collect the superna-tant containing the membrane proteins.

To detect expression of Flag-tagged PG (transgene), we probedthe membranes with a rabbit anti–DYKDDDDK-Tag antibody (CellSignaling Technology Inc, Danvers, Mass). To detect the endoge-nous and the Flag-tagged PG simultaneously, we probed the mem-branes with a rabbit anti-PG antibody. The secondary antibody wasdonkey anti-rabbit IgG horseradish peroxidase–conjugated (SantaCruz Biotechnology Inc).

ImmunofluorescenceWe embedded thin myocardial sections from mid-ventricle in opti-mal cutting temperature compound (Sakura Finetek Inc, Torrance,Calif) and flash-froze them in an isopentane-liquid nitrogen bath. Westained the sections with the primary antibodies in 5% donkey orgoat normal serum (Santa Cruz Biotechnology Inc). The primaryantibodies (Santa Cruz Biotechnology Inc, unless otherwise speci-fied) were goat polyclonal anti-Islet1 (Isl1), rabbit polyclonal anti-Mef2C (Aviva System Biology LLC, San Diego, Calif), rabbitpolyclonal anti–enhanced yellow fluorescent protein (EYFP), rabbitpolyclonal anti–peroxisome proliferator-activated receptor(PPAR)-�, goat polyclonal anti–C/EBP-�, rabbit polyclonal anti–C/EBP-�, rabbit polyclonal anti-PG, rabbit anti-Flag (Cell SignalingTechnology Inc, Beverly, Mass), and rat monoclonal anti–F4/80(Abcam, Cambridge, Mass), the latter as a macrophage-specificmarker. The secondary antibodies were FITC-labeled donkey anti-rabbit IgG and anti-goat IgG, FITC-labeled goat anti-rabbit IgG,Texas red–labeled goat anti-rat IgG, and Texas red–labeled donkeyanti-goat IgG. The sections were mounted in DAPI-containing HardSet mounting medium (Vector Laboratories Inc, Burlingame, Calif)and examined under fluorescence microscopy.

Additional control groups for immunofluorescence included myo-cardial sections stained with the corresponding IgG isotypes andsecondary antibodies. Likewise, to further demonstrate the specific-ity of staining of cardiac adipocytes for expression of EYFP, westained tissue sections from visceral and subcutaneous fats from thewild-type (WT) and lineage tracer mice using the same antibodiesthat were used for staining of cardiac sections.

To quantify the number of cardiac adipocytes expressing EYFP, westained frozen thin myocardial sections from WT and lineage tracermice for oil red O and the corresponding adjacent sections for expres-sion EYFP and C/EBP�. We scored the number of cells that stainedpositive for oil red O and for coexpression of C/EBP� and EYFP in 14to 44 thin sections per mouse and in 5 to 9 mice per group.

Human Myocardial SectionsThe results of genetic fate-mapping experiments showing rarecardiac cells that coexpressed adipogenic and the second heart fieldmarkers and given that fibroadiposis in humans is an evolving andprogressive phenotype, we postulated that human hearts with ARVCmight contain rare cells in transition from a myogenic to anadipogenic fate. Therefore, we costained paraffin-embedded thinmyocardial sections from 3 human patients with autopsy-provenARVC who died of sudden cardiac death (ages 28 to 38 years) forcoexpression of Isl1 or Mef2C with C/EBP-� or PPAR-�. As acontrol, we stained thin sections of a normal heart for the abovemarkers.

Lombardi et al Origin of Adipocytes in ARVC 1077


Dow

nloaded from


Statistical AnalysisStatistical calculations (STATA-Intercooled version 10.1, StataCorpLP, College Station, Tex) were as published.22 We compared thedifferences for normally distributed continuous variables among the3 groups by ANOVA, followed by pairwise comparisons by Bon-ferroni method. We analyzed variables that violated the normalityassumption and the nonparametric variables by Kruskal–Wallis test.

ResultsGeneration of Dsp-Deficient Lineage Tracer MiceUsing the LoxP-Cre technology, we generated 3 sets oflineage tracer mice regulated by cardiac lineage promotersNkx2.5, an early and pan-specific cardiogenic marker;Mef2C, a second heart field-specific marker; and �MyHC, arelatively late marker of cardiogenesis (Figure 1). Wescreened the offspring for the presence of floxed alleles andthe Cre recombinase by PCR (oligonucleotide primers areshown in Online Table I). Distribution of the genotypes of theoffspring for Dsp-deficient mice deviated from the expectedMendelian inheritance, because no liveborn homozygousDsp-deficient mouse was identified. The finding suggestsembryonic lethality, as was reported previously in conditionaldeletion of Dsp in the heart and in the germ line.13,23

Therefore, the experiments were performed in heterozygousDsp-deficient mice, which genetically represent human auto-somal dominant ARVC.5,24,25 We confirmed expression of thereporter protein EYFP in the heart by immunoblotting usingan anti-EYFP antibody (Figure 1).

Cardiac Phenotype in Control and LineageTracer MiceWe did not detect a discernible cardiac phenotype in miceexpressing EYFP or Cre recombinase alone. Cardiac histol-ogy, analyzed by hematoxylin/eosin, Masson trichrome andoil red O staining of thin myocardial sections, was normal inthe �MyHC-Cre, Nkx2.5-Cre, Mef2C-Cre, R26-EYFPF/F,and �MyHC- or Nkx2.5- or Mef2C-Cre:R26-EYFPF/F mice.Cardiac size and function, determined by M-mode, 2D, and

Doppler echocardiography, were also normal in the abovemice. Likewise, expression of EYFP in the background ofDsp deficiency did not influence cardiac phenotype becausethere were no significant differences in cardiac phenotypebetween Nkx2.5-Cre:DspW/F and Nkx2.5-Cre:DspW/F:R26-EYFPF/F or Mef2C-Cre:DspW/F and Mef2C-Cre:DspW/F:R26-EYFPF/F or �MyHC-Cre:DspW/F and �MyHC-Cre:DspW/F:R26-EYFPF/F mice. Therefore, for brevity, data on WT(control) and 3 sets of lineage tracer mice (�MyHC- orNkx2.5- or Mef2C-Cre:DspW/F:R26-EYFPF/F) is presentedalong with relevant data on Nkx2.5-Cre:DspW/F and Mef2C-Cre:DspW/F, whenever appropriate.

Ventricular weight/body weight ratio was increased in the�MyHC-Cre:DspW/F:R26-EYFPF/F and Nkx2.5-Cre:DspW/F:R26-EYFPF/F mice but not in Mef2C-Cre:DspW/F:R26-EYFPF/F mice, as compared to WT mice (Figure I in the onlinedata supplement). Likewise, ventricular weight/body weightratio was also increased in the Nkx2.5-Cre:DspW/F (withoutEYFP) mice as was reported in the �MyHC-Cre:DspW/F

mice.13 The �MyHC and Nkx2.5 regulated lineage tracermice as well as the Nkx2.5-Cre:DspW/F mice exhibited leftventricular enlargement and dysfunction (Online Figure IIand Online Table II). In contrast, left ventricular size andfunction were normal in Mef2C-Cre:DspW/F:R26-EYFPF/F

(and Mef2C-Cre:DspW/F) mice (Online Figure II and OnlineTable II). The latter findings reflect specificity of the Mef2C-Cre in deleting Dsp in the second heart field, which gives riseto the right but not the left ventricle.

Myocardial histology in the lineage tracer mice was re-markable for excess fibroadiposis (Figure 2). Adipocytes,scored from oil red O–stained sections, comprised0.16�0.02% of the total cells examined in the hearts of 3- to6-month-old WT mice (N�5 mice, 10 800�1643 total cellsper mouse). The corresponding percentages in the hearts of�MyHC-Cre–, Nkx2.5-Cre–, and Mef2C-Cre–regulated lin-eage tracer mice were 0.28�0.14% (N�7 mice,10 000�4898 total cells per mouse), 0.44�0.26% (N�9

Figure 1. Diagram illustrating generation oflineage tracer Dsp-deficient mice. Micewith floxed exon 2, the site of the initiationcodon of Dsp (A), were crossed to EYFPmice (B), in which expression of EYFP isblocked by an upstream loxP-flankedSTOP sequence. The offspring carryingboth floxed transgenes (C) were crossedto Cre deleter mice (D), the latter beingregulated by cardiac lineage promotersNkx2.5, Mef2C, or �MyHC. E, RecombinedDsp and EYFP alleles. F, Expression of an�30-kDa band corresponding to the sizeof EYFP protein in Nkx-Cre:R26-EFYPF/F

and Nkx2.5-Cre:DspW/F:R26-EYFPF/F mice.

1078 Circulation Research May 8, 2009


Dow

nloaded from


mice, 11 625�3739 total cells per mouse), and 0.63�0.32%(N�5 mice, 21 000�6363 total cells per mouse; all pair wiseprobability values versus WT mice were �0.05). As observedin human ARVC, fibroadiposis was more prominent at theepicardium in the lineage tracer mice (Figure 2). In accor-dance with the morphological data on cardiac size, cardiacmyocyte cross-sectional area was also increased in the �MyHC-Cre:DspW/F:R26-EYFPF/F and Nkx2.5-Cre:DspW/F:R26-EYFPF/F but not in the Mef2C-Cre:DspW/F:R26-EYFPF/F mice,as compared with WT mice (Figure 2).

Lineage Tracing of Excess AdipocytesTo determine the origin of excess adipocytes in the heart oflineage tracer mice, we coimmunostained myocardial sec-tions with antibodies against EYFP and C/EBP-�, an adipo-genic transcription factor. EYFP and C/EBP-� were coex-pressed in epicardial adipocytes in the hearts of Nkx-2.5– andMef2C-regulated lineage tracer mice (Figure 3 and OnlineFigure III, A). The number of adipocytes, identified byC/EBP� stained cells, that stained positive for EYFP varied.On average, 78% of adipocytes in the hearts of Mef2C-

Figure 2. Fibroadiposis in the lineagetracer mice. Myocardial transverse sec-tions stained for oil red O (ORO), hematox-ylin/eosin (H&E), Masson trichrome (MT),and wheat germ agglutinin Texas red(WGA-TR) conjugate are shown from WT,�MyHC-Cre:DspW/F:R26-EYFPF/F, Nkx2.5-Cre:DspW/F:R26-EYFPF/F, and Mef2C-Cre:DspW/F:R26-EYFPF/F mice. Patchyareas of myocyte drop out and fibroadipo-sis were present in all 3 sets of lineagetracer mice. Fibroadiposis in the Mef2C-regulated lineage tracer mice wasrestricted to the right ventricle. Cardiacmyocyte size was increased in �MyHCand Nkx2.5 but not in Mef2C-regulatedlineage tracer mice (WT: 4458�182;�MyHC: 4830�151; Nkx2.5: 4891�96;Mef2C: 4659�42; N�3 to 6; pairwiseprobability values for �MyHC, Nkx2.5, andMef2C tracer vs WT mice were 0.024,0.009, and 0.458, respectively).

Figure 3. Coexpression of EYFP and C/EBP-� in adipocytes in the hearts of lineage tracer mice. Oil red O and corresponding DAPI,C/EBP-�, and EYFP costained sections, along with merged images, are shown. C/EBP-� and EYFP are coexpressed in adipocytes inthe hearts of Nkx2.5:DspW/F:R26-EYFPF/F and Mef2C-Cre:DspW/F:R26-EYFPF/F mice.



Dow

nloaded from


Cre:DspW/F:EYFPF/F costained positive for EYFP (N�6mice, 100�57/128�73 adipocytes per mouse). A similarpercentage of adipocytes in the hearts of Nkx2.5-Cre:DspW/F:EYFPF/F mice also expressed EYFP (N�9, 40�32/51�33adipocytes per mouse). In contrast, only few adipocytes

(4/279, 1.4%) in the �MyHC-Cre:DspW/F:EYFPF/F lineagetracer mice hearts stained positive for EYFP expression.

To establish the specificity of antibodies in detecting EYFPand C/EBP�, we stained myocardial sections with the IgGisotypes of the primary antibodies and the secondary antibod-

Figure 4. Coexpression of second heart field and adipogenic transcription factors in the hearts of lineage tracer mice. A, Oil redO–stained and coimmunostained thin myocardial sections for DNA (DAPI), Islet-1 (Isl1), a second heart field marker; and PPAR-�, anadipogenic transcription factor, are shown. Rare cells showing coexpression of Isl1 and PPAR-� in the hearts of Nkx2.5:DspW/F:R26-EYFPF/F and Mef2C-Cre:DspW/F:R26-EYFPF/F lineage tracer mice were detected. B, Individual and merged images showing coexpres-sion of second heart field marker Mef2C and adipogenic marker C/EBP-� are presented along with oil red O–stained image. As shown,Mef2C and C/EBP-� are coexpressed in a subset of adipocytes in the hearts of Nkx2.5:DspW/F:R26-EYFPF/F and Mef2C:DspW/F:R26-EYFPF/F lineage tracer mice.



Dow

nloaded from


ies. The isotypes and secondary antibodies did not react withEYFP and C/EBP� (Online Figure III, B). To further sub-stantiate the specificity of the findings, we stained visceraland subcutaneous adipose tissues with the same antibodiesagainst EYFP and C/EBP�. Adipocytes from visceral andsubcutaneous adipose tissues did not express EYFP, furtherindicating cardiac specificity (Online Figure IV).

Detection of expression of EYFP in adipocytes from theMef2C-regulated lineage tracer mice indicated an origin fromthe second heart field progenitors. We surmised that thehearts in the ARVC mice might contain cells in transitionfrom a myogenic to an adipogenic fate. To test this hypothesisand further validate the finding in Mef2C-regulated lineagetracer mice, we costained myocardial sections with antibodiesagainst Isl1 or Mef2C and C/EBP-� or PPAR-�. We detectedrare cells (0 to 6 cells per heart) in the hearts of Nkx2.5- andMef2C-regulated lineage tracer mice that coexpressed Isl1and PPAR-� or Mef2C and C/EBP-� (Figure 4A and 4B). Wedid not detect expression of second heart field markers in thefew EYFP-expressing adipocytes in the hearts of �MyHC-regulated lineage tracer mice.

Because Isl1 is not an exclusive cardiac lineage marker, toexclude the possibility that macrophages were the rare cellsthat expressed the second field markers, we costained thinmyocardial sections with an antibody against F4/80, a mac-rophage-specific marker and EYFP. We included the corre-sponding oil red O–stained section and thin sections fromspleen as controls. The results (Online Figure V) showedadipocytes in the heart of lineage tracer mice expressed EYFPbut not F4/80 macrophage marker.

Detection of Expression of Isl1 and Mef2C inAdipocytes in Human Hearts With ARVCTo corroborate the results of lineage tracing studies in mice inhuman patients, we costained thin right ventricular sectionsfrom 3 human patients with autopsy-proven ARVC forcardiac progenitor markers and adipogenic transcription fac-tors. In accordance with our findings in the genetic fate-mapping studies in mice, we detected rare cells (1 to 3 cellsper heart) in the fibroadiposis area in the myocardium thatcoexpressed Isl1 or Mef2C and the adipogenic transcriptionfactor C/EBP-� (Figure 5 and Online Figure VI).

Nuclear PG Promotes Adipogenesis ThroughSuppression of the Canonical Wnt SignalingWe have implicated PG, an armadillo member of desmo-somal protein with signaling function,26–28 in the pathogene-sis of ARVC.13 Likewise, Wnt signaling is implicated indevelopment of the second heart field (right ventricle), thepredominant site of involvement in ARVC.29,30 To elucidatea mechanism for differentiation of second heart field progen-itors to adipocytes in ARVC, we expressed Flag-tagged PG incardiac myocytes through transgenesis under transcriptionalregulation of the �MyHC promoter. Expression level of theFlag-tagged transgene PG comprised approximately 25% to35% of the total PG levels in the heart (Figure 6). The PGtransgenic mice showed excess adipocytes in the heart alongwith patchy areas of fibrosis, an intact desmosome structureand normal left ventricular function (Figure 6 and OnlineFigure VII).

Immunostaining of thin myocardial sections from theFlag-tagged PG transgenic mice as well as immunoblotting of

Figure 5. Coexpression of second heart field and adipogenic markers in the hearts of human patients with autopsy-proven ARVC.Bright field and costained sections for Isl1, C/EBP-�, and DNA (A) or Mef2C, C/EBP-�, and DNA (B) in the heart of patients with ARVCwho died suddenly. Rare cells coexpressing second heart field markers Isl1 or Mef2C and C/EBP� were detected in the human heartswith ARVC.



Dow

nloaded from


cell protein subfractions showed nuclear localization of thetransgene PG in cardiac myocytes (Figure 7A and 7B). Todetermine whether nuclear PG (also known as �-catenin)interacted with protein constituents of the canonical Wntsignaling, we analyzed binding of PG with Tcf7l2 transcrip-tion factor by coimmunoprecipitation. Immunoblotting of theimmunoprecipitates using transgene-specific anti-Flag andpan-PG antibodies showed binding of the transgene PG toTcf7l2 (Figure 7C). In contrast, binding of �-catenin toTcf7l2 in the PG transgenic hearts was reduced, implyingcompetitive interactions between PG and �-catenin for bind-ing to Tcf7l2.27

To determine the biological effects of nuclear localizationand binding of PG to Tcf7l2 on canonical Wnt signaling, wedetermined expression levels of selected canonical Wnt/�-catenin signaling target genes by quantitative PCR (N�4 pergroup, probes and primers sequence are provided in OnlineTable I and Online Figure VIII). Concordant with reducedbinding of �-catenin to Tcf7l2 in the PG transgenic mice,expression of c-Myc, a known target of activation of the

canonical Wnt signaling, was suppressed (Figure 7D). Inaddition, expression levels of adipogenic factors Wnt5b andBMP7, which are normally inhibited by the canonical Wntsignaling, were increased by 3- to 4-fold (Figure 7E and7F).31,32 Recent data implicate BMP7 as a major regulator ofswitch from myogenesis to adipogenesis.32,33 In contrast,expression level of connective tissue growth factor (CTGF),which is known to inhibit adipogenesis,34 was reduced by�5-fold (Figure 7G).

DiscussionThrough a series of genetic fate-mapping experiments inmice, we show adipocytes in ARVC originate from thesecond heart field cardiac progenitor cells. We corroboratethe findings in autopsy-proven human hearts with ARVC byshowing coexpression of second heart field markers, alongwith adipogenic transcription factors in rare cells in fibroa-dipocytic areas in the heart. At a mechanistic level, we shownuclear localization of desmosomal protein PG is associatedwith suppression of the canonical Wnt/�-catenin signaling

Figure 6. Cardiac phenotype in PG trans-genic mice. A, PG transgene constructshowing 5-Flag–tagged full-length PGcDNA positioned downstream to a 5.5-kbp�MyHC promoter. B, Immunoblots show-ing expression of transgene and endoge-nous PG proteins detected using Flag-specific and pan-PG–specific antibodies innontransgenic (NTG) and transgenic mice.C, Oil red O (ORO)-, hematoxylin/eosin(H&E)-, and Masson trichrome (MT)-stained thin myocardial sections showfibroadiposis in the heart of PG transgenicmice. Immunofluorescence staining (IF)shows localization of PG to desmosomes.

Figure 7. Pathogenesis of ARVC in PG trans-genic mice. A and B show nuclear localizationof transgene PG in myocytes by immunofluo-rescence staining, detected using a Flag-specific antibody, and by immunoblotting ofsubcellular protein fractions, detected usingFlag-specific and pan-PG–specific antibodies.C illustrates binding of transgene PG to Tcf7l2,as detected by coprecipitation with an anti-Tcf7l2 antibody, which was increased in thePG transgenic mice. The lower blot showscoprecipitation of �-catenin with an anti-Tcf7l2antibody, which was reduced in the PG trans-genic mice. D through G, Quantification ofexpression levels of selected canonical Wntsignaling by real-time PCR. Expression level ofcMyc, a known target of the canonical Wntsignaling, was reduced by �50%. Relativeexpression levels of Wnt5b and BMP7, whichpromote adipogenesis and are normally inhib-ited by the canonical Wnt signaling, wereincreased 3- to 4-fold in the transgenic mice.In contrast, expression level of CTGF, a nega-tive regulator of adipogenesis, was sup-pressed by �5-fold.



Dow

nloaded from


and a transcriptional switch to adipogenesis through activa-tion of BMP7 and Wnt5b and suppression of CTGF.31–34 Thepaucity of EYFP-expressing adipocytes in the hearts of�MyHC-regulated lineage tracer mice and a significantlyhigher percentage of EYFP-expressing adipocytes in thehearts of Mef2C- or Nkx2.5-regulated lineage tracer miceindicate that the adipogenic switch occurs before commit-ment of the progenitors to a myocyte lineage. The resultsindicate that adipocytes in ARVC originate from the secondheart field progenitor cells, which switch to adipogenesisbecause of suppressed canonical Wnt/�-catenin signalingimparted by the nuclear PG (Online Figure IX).

In human patients with ARVC, fibroadiposis predomi-nantly involves the right ventricle, a feature that has also beenan enigma.1 The right ventricle primarily originates from thesecond (anterior) heart field as opposed to the left ventricle,which originates from the primary heart field.35–37 The 2 heartfields are distinguished by expression of different sets oftranscriptional factors and signaling molecules. Accordingly,Nkx2.5 is common to both heart fields; however, Tbx5 andHand1 mark the primary heart field, whereas Isl1, Mef2C,and Hand2 characterize the second heart field.29,35–37 Identi-fication of the second heart field progenitors as the cell sourceof adipocytes in ARVC offers a plausible explanation for thepredominant involvement of the right ventricle in ARVC. Ourfindings are also in accordance with the recent data empha-sizing the significance of the canonical Wnt signaling in theformation of the right ventricle from the second heartfield.29,30,38 It is also noteworthy that the canonical Wntsignaling is considered a major regulator of a switch betweenadipogenesis and myogenesis.39 Furthermore, changes inexpression levels of BMP7, Wnt5b, and CTGF, which aretargets of the canonical Wnt signaling, favored adipogene-sis.31–33 However, the characteristics of the subset of thesecond heart field progenitors and complete molecular mech-anisms that govern their differentiation to an adipogenic fatein ARVC remain unknown.

Our focus on cardiac progenitor cells as the cell source ofadipocytes is based on the premise that in ARVC caused bydesmosomal mutations the stimulus for adipocytic differen-tiation has to originate from cells that express the desmo-somal proteins. There are other potential sources that could becategorized into 2 sets. First, cells other than the myogeniclineage in the heart may express the mutant desmosomalproteins and, hence, could directly differentiate into adipo-cytes. This seems unlikely, because the only cell type in theheart known to express desmosomal proteins is the myocytelineage. Second, cells that do not express the desmosomalproteins could differentiate into adipocytes through paracrinemechanisms emanating from desmosome-defective myo-cytes. In preliminary studies, we have cocultured myocytesisolated from the hearts of Dsp-deficient mice with cardiacfibroblasts and have not detected enhanced differentiation ofthe fibroblasts to adipocytes. Nevertheless, cellularly, theheart is a heterogeneous organ, and, hence, the possibledifferentiation of cells such as pericytes, fibrocytes, or circu-lating cells that seed at the myocardium and differentiate toadipocytes cannot be completely dismissed.

In conclusions, we have traced the origin of excess adipo-cytes in desmosomal ARVC to second heart field progenitorcells, a finding that was also corroborated in human heartswith ARVC. We implicate suppression of the canonical Wntsignaling by nuclear PG as a mechanism for switchingdifferentiation of cardiac progenitor cells to adipocytes. Thefindings also explain the predominant involvement of theright ventricle in ARVC.

AcknowledgmentsThe �MyHC-Cre, Mef2C-Cre, and R26-EYFP mice were kind giftsfrom Drs Michael D. Schneider, Brian Black, and F. Costantini,respectively.

Sources of FundingSupported by National Heart, Lung, and Blood Institute grantsR01-HL68884 and R01–088498 and a Burroughs Wellcome Awardin Translational Research (1005907).

DisclosuresNone.

References1. Corrado D, Basso C, Thiene G, McKenna WJ, Davies MJ, Fontaliran F,

Nava A, Silvestri F, Blomstrom-Lundqvist C, Wlodarska EK, Fontaine G,Camerini F. Spectrum of clinicopathologic manifestations of arrhyth-mogenic right ventricular cardiomyopathy/dysplasia: a multicenter study.J Am Coll Cardiol. 1997;30:1512–1520.

2. Corrado D, Basso C, Thiene G. Arrhythmogenic right ventricular cardio-myopathy: diagnosis, prognosis, and treatment. Heart. 2000;83:588–595.

3. Awad MM, Calkins H, Judge DP. Mechanisms of disease: moleculargenetics of arrhythmogenic right ventricular dysplasia/cardiomyopathy.Nat Clin Pract Cardiovasc Med. 2008;5:258–267.

4. Thiene G, Nava A, Corrado D, Rossi L, Pennelli N. Right ventricularcardiomyopathy and sudden death in young people. N Engl J Med.1988;318:129–133.

5. Norman M, Simpson M, Mogensen J, Shaw A, Hughes S, Syrris P,Sen-Chowdhry S, Rowland E, Crosby A, McKenna WJ. Novel mutationin desmoplakin causes arrhythmogenic left ventricular cardiomyopathy.Circulation. 2005;112:636–642.

6. McKoy G, Protonotarios N, Crosby A, Tsatsopoulou A, Anastasakis A,Coonar A, Norman M, Baboonian C, Jeffery S, McKenna WJ. Identifi-cation of a deletion in plakoglobin in arrhythmogenic right ventricularcardiomyopathy with palmoplantar keratoderma and woolly hair (Naxosdisease). Lancet. 2000;355:2119–2124.

7. Kaplan SR, Gard JJ, Carvajal-Huerta L, Ruiz-Cabezas JC, Thiene G,Saffitz JE. Structural and molecular pathology of the heart in Carvajalsyndrome. Cardiovasc Pathol. 2004;13:26–32.

8. Heuser A, Plovie ER, Ellinor PT, Grossmann KS, Shin JT, Wichter T,Basson CT, Lerman BB, Sasse-Klaassen S, Thierfelder L, MacRae CA,Gerull B. Mutant desmocollin-2 causes arrhythmogenic right ventricularcardiomyopathy. Am J Hum Genet. 2006;79:1081–1088.

9. Gerull B, Heuser A, Wichter T, Paul M, Basson CT, McDermott DA,Lerman BB, Markowitz SM, Ellinor PT, MacRae CA, Peters S,Grossmann KS, Michely B, Sasse-Klaassen S, Birchmeier W, Dietz R,Breithardt G, Schulze-Bahr E, Thierfelder L. Mutations in the des-mosomal protein plakophilin-2 are common in arrhythmogenic rightventricular cardiomyopathy. Nat Genet. 2004;36:1162–1164.

10. Norgett EE, Hatsell SJ, Carvajal-Huerta L, Cabezas JC, Common J,Purkis PE, Whittock N, Leigh IM, Stevens HP, Kelsell DP. Recessivemutation in desmoplakin disrupts desmoplakin-intermediate filamentinteractions and causes dilated cardiomyopathy, woolly hair and kera-toderma. Hum Mol Genet. 2000;9:2761–2766.

11. Alcalai R, Metzger S, Rosenheck S, Meiner V, Chajek-Shaul T. Arecessive mutation in desmoplakin causes arrhythmogenic right ventric-ular dysplasia, skin disorder, and woolly hair. J Am Coll Cardiol. 2003;42:319–327.

12. Syrris P, Ward D, Evans A, Asimaki A, Gandjbakhch E, Sen-ChowdhryS, McKenna WJ. Arrhythmogenic right ventricular dysplasia/cardiomy-



Dow

nloaded from


opathy associated with mutations in the desmosomal gene desmocollin-2.Am J Hum Genet. 2006;79:978–984.

13. Garcia-Gras E, Lombardi R, Giocondo MJ, Willerson JT, Schneider MD,Khoury DS, Marian AJ. Suppression of canonical Wnt/beta-catenin sig-naling by nuclear plakoglobin recapitulates phenotype of arrhythmogenicright ventricular cardiomyopathy. J Clin Invest. 2006;116:2012–2021.

14. Anversa P, Leri A, Kajstura J. Cardiac regeneration. J Am Coll Cardiol.2006;47:1769–1776.

15. Vasioukhin V, Bowers E, Bauer C, Degenstein L, Fuchs E. Desmoplakinis essential in epidermal sheet formation. Nat Cell Biol. 2001;3:1076–1085.

16. Moses KA, DeMayo F, Braun RM, Reecy JL, Schwartz RJ. Embryonicexpression of an Nkx2–5/Cre gene using ROSA26 reporter mice.Genesis. 2001;31:176–180.

17. Agah R, Frenkel PA, French BA, Michael LH, Overbeek PA, SchneiderMD. Gene recombination in postmitotic cells. Targeted expression of Crerecombinase provokes cardiac-restricted, site-specific rearrangement inadult ventricular muscle in vivo. J Clin Invest. 1997;100:169–179.

18. Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM,Costantini F. Cre reporter strains produced by targeted insertion of EYFPand ECFP into the ROSA26 locus. BMC Dev Biol. 2001;1:4.

19. Lombardi R, Bell A, Senthil V, Sidhu J, Noseda M, Roberts R, MarianAJ. Differential interactions of thin filament proteins in two cardiactroponin T mouse models of hypertrophic and dilated cardiomyopathies.Cardiovasc Res. 2008;79:109–117.

20. Tsybouleva N, Zhang L, Chen S, Patel R, Lutucuta S, Nemoto S, DeFreitasG, Entman M, Carabello BA, Roberts R, Marian AJ. Aldosterone, throughnovel signaling proteins, is a fundamental molecular bridge between thegenetic defect and the cardiac phenotype of hypertrophic cardiomyopa-thy. Circulation. 2004;109:1284–1291.

21. Cox B, Emili A. Tissue subcellular fractionation and protein extraction foruse in mass-spectrometry-based proteomics. Nat Protocols. 2006;1:1872–1878.

22. Lombardi R, Rodriguez G, Chen SN, Ripplinger CM, Li W, Chen J,Willerson JT, Betocchi S, Wickline SA, Efimov IR, Marian AJ. Reso-lution of established cardiac hypertrophy and fibrosis and prevention ofsystolic dysfunction in a transgenic rabbit model of human cardiomyop-athy through thiol-sensitive mechanisms. Circulation. 2009;119:1398–1407.

23. Gallicano GI, Kouklis P, Bauer C, Yin M, Vasioukhin V, Degenstein L,Fuchs E. Desmoplakin is required early in development for assembly ofdesmosomes and cytoskeletal linkage. J Cell Biol. 1998;143:2009–2022.

24. Rampazzo A, Nava A, Malacrida S, Beffagna G, Bauce B, Rossi V,Zimbello R, Simionati B, Basso C, Thiene G, Towbin JA, Danieli GA.Mutation in human desmoplakin domain binding to plakoglobin causes adominant form of arrhythmogenic right ventricular cardiomyopathy. Am JHum Genet. 2002;71:1200–1206.

25. Bauce B, Basso C, Rampazzo A, Beffagna G, Daliento L, Frigo G,Malacrida S, Settimo L, Danieli G, Thiene G, Nava A. Clinical profile of fourfamilies with arrhythmogenic right ventricular cardiomyopathy caused bydominant desmoplakin mutations. Eur Heart J. 2005;26:1666–1675.

26. Salomon D, Sacco PA, Roy SG, Simcha I, Johnson KR, Wheelock MJ,Ben Ze’ev A. Regulation of beta-catenin levels and localization by over-expression of plakoglobin and inhibition of the ubiquitin-proteasomesystem. J Cell Biol. 1997;139:1325–1335.

27. Zhurinsky J, Shtutman M, Ben Ze’ev A. Differential mechanisms ofLEF/TCF family-dependent transcriptional activation by beta-catenin andplakoglobin. Mol Cell Biol. 2000;20:4238–4252.

28. Miravet S, Piedra J, Miro F, Itarte E, Garcia de Herreros A, Dunach M.The transcriptional factor Tcf-4 contains different binding sites for beta-catenin and plakoglobin. J Biol Chem. 2002;277:1884–1891.

29. Ai D, Fu X, Wang J, Lu MF, Chen L, Baldini A, Klein WH, Martin JF.Canonical Wnt signaling functions in second heart field to promote rightventricular growth. Proc Natl Acad Sci U S A. 2007;104:9319–9324.

30. Klaus A, Saga Y, Taketo MM, Tzahor E, Birchmeier W. Distinct roles ofWnt/beta-catenin and Bmp signaling during early cardiogenesis. ProcNatl Acad Sci U S A. 2007;104:18531–18536.

31. Kanazawa A, Tsukada S, Kamiyama M, Yanagimoto T, Nakajima M,Maeda S. Wnt5b partially inhibits canonical Wnt/beta-catenin signalingpathway and promotes adipogenesis in 3T3–L1 preadipocytes. BiochemBiophys Res Commun. 2005;330:505–510.

32. Tseng YH, Kokkotou E, Schulz TJ, Huang TL, Winnay JN, TaniguchiCM, Tran TT, Suzuki R, Espinoza DO, Yamamoto Y, Ahrens MJ, DudleyAT, Norris AW, Kulkarni RN, Kahn CR. New role of bone morpho-genetic protein 7 in brown adipogenesis and energy expenditure. Nature.2008;454:1000–1004.

33. Seale P, Bjork B, Yang W, Kajimura S, Chin S, Kuang S, Scime A,Devarakonda S, Conroe HM, Erdjument-Bromage H, Tempst P, RudnickiMA, Beier DR, Spiegelman BM. PRDM16 controls a brown fat/skeletalmuscle switch. Nature. 2008;454:961–967.

34. Tan JT, McLennan SV, Song WW, Lo LW, Bonner JG, Williams PF,Twigg SM. Connective tissue growth factor inhibits adipocyte differen-tiation. Am J Physiol Cell Physiol. 2008;295:C740–C751.

35. Cai CL, Liang X, Shi Y, Chu PH, Pfaff SL, Chen J, Evans S. Isl1identifies a cardiac progenitor population that proliferates prior to differ-entiation and contributes a majority of cells to the heart. Dev Cell.2003;5:877–889.

36. Kelly RG, Brown NA, Buckingham ME. The arterial pole of the mouseheart forms from Fgf10-expressing cells in pharyngeal mesoderm. DevCell. 2001;1:435–440.

37. Laugwitz KL, Moretti A, Lam J, Gruber P, Chen Y, Woodard S, Lin LZ,Cai CL, Lu MM, Reth M, Platoshyn O, Yuan JXJ, Evans S, Chien KR.Postnatal isl1� cardioblasts enter fully differentiated cardiomyocytelineages. Nature. 2005;433:647–653.

38. Cohen ED, Wang Z, Lepore JJ, Lu MM, Taketo MM, Epstein DJ,Morrisey EE. Wnt/beta-catenin signaling promotes expansion of Isl-1-positive cardiac progenitor cells through regulation of FGF signaling.J Clin Invest. 2007;117:1794–1804.

39. Ross SE, Hemati N, Longo KA, Bennett CN, Lucas PC, Erickson RL,MacDougald OA. Inhibition of adipogenesis by Wnt signaling. Science.2000;289:950–953.



Dow

nloaded from


Schwartz, James T. Willerson, Ramon Brugada and Ali J. MarianRaffaella Lombardi, Jinjiang Dong, Gabriela Rodriguez, Achim Bell, Tack Ki Leung, Robert J.

Adipocytes in Arrhythmogenic Right Ventricular CardiomyopathyGenetic Fate Mapping Identifies Second Heart Field Progenitor Cells As a Source of

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 2009 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/CIRCRESAHA.109.1968992009;104:1076-1084; originally published online April 9, 2009;Circ Res.

http://circres.ahajournals.org/content/104/9/1076World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circres.ahajournals.org/content/suppl/2009/04/09/CIRCRESAHA.109.196899.DC1Data Supplement (unedited) at:

http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:


Dow

nloaded from

http://circres.ahajournals.org/content/104/9/1076

http://circres.ahajournals.org/content/suppl/2009/04/09/CIRCRESAHA.109.196899.DC1

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circres.ahajournals.org//subscriptions/


CIRCRESAHA/2009/196899/R1

1

Supplementary Material

Genetic Fate Mapping Identifies Second Heart Field Progenitor Cells as a Source of

Adipocytes in Arrhythmogenic Right Ventricular Cardiomyopathy

Raffaella Lombardi, Jinjiang Dong, Gabriela Rodriguez, Achim Bell, Tack Ki Leung#, Robert J.

Schwartz¶, James T. Willerson, Ramon Brugada#, AJ Marian

Center for Cardiovascular Genetics, Brown Foundation Institute of Molecular Medicine, The

University of Texas Health Science Center and Texas Heart Institute, Houston, TX

#Montreal Heart Institute and Université de Montréal, Montreal, Quebec, Canada

¶ Institute of Biotechnology, Texas A&M University, Houston, TX

Short Title: Origin of Adipocytes in ARVC

Address for Correspondence and Reprints:

AJ Marian, M.D. Brown Foundation Institute of Molecular Medicine The University of Texas Health Sciences Center 6770 Bertner Street, Suite C900A Houston, TX 77030 Phone: 713 500 2350 Fax: 713 500 2320 [email protected]


2

Online Methods

Genetically modified mice: Dsp-floxed (courtesy of Dr. Elaine Fuchs), Mef2C-Cre

(courtesy of Dr. Brian Black), α -MyHC-Cre (courtesy of Dr. Michael D Schneider) and Nkx2.5-

Cre mice have been published 1-4. R26-EYFP mouse was a generous gift from Dr. F. Costantini

5. EYFP is an enhanced yellow fluorescent protein variant of the Aequorea victoria GFP, which

is cloned into the ROSA26 locus, preceded by a loxP-flanked stop sequence. Cre recombinase-

mediated excision of the loxP-flanked transcriptional "stop" sequence results in expression of

EYFP, which functions as a faithful monitor of Cre activity.

We generated the PG transgenic mice by the conventional method. The transgene was

comprised of a 5’ end Flag-tagged wild type full-length PG which was placed downstream to a

5.5 Kbp cardiac-specific α-MyHC promoter.

M mode, 2-Dimensional, and Doppler Echocardiography: We performed

echocardiography using a HP Sonos 5500 System, equipped with a 15 MHz linear transducer

using sodium pentobarbital for anesthesia 1, 6. We calculated ventricular fractional shortening,

ejection time, left ventricular dimensions and mass and Doppler indices without knowledge of

the genotype 1, 6.

Gross and histological cardiac phenotype: We determined ventricular weight/body

weight ratio in sex-matched mice at 6 months of age. We analyzed myocardial histology by H&E

and Masson Trichrome and Oil Red O staining, the latter to detect fat droplets. An investigator

without knowledge of the genotypes performed and analyzed histology in age and sex-matched

mice and in a random order 1, 6, 7. We scored the number of adipocytes on Oil Red O stained

thin myocardial sections in approximately 10,000 -12,000 cells per mouse and in 5 to 6 mice per

group.

We calculated myocyte cross sectional area (CSA) by semi-automated planimetry in

age- and sex- and body weight-matched non-transgenic and lineage tracer mice. Freshly


3

harvested thick cardiac cross-sections were placed in optimal cutting temperature (OCT)

compound (Sakura-Finetek U.S.A. Inc., Torrace, CA, cat#4583) and frozen in isopentane (2-

Methyl Butane) cooled at -155oC in a liquid nitrogen bath. Thin myocardial sections were cut,

washed in PBS for 3 times and stained with Wheat Germ Agglutinin (WGA) conjugated with

Texas Red (Molecular Probes Inc., Eugene, OR; cat # W21405). The working concentration

was 2 µg/mL and incubation was at room temperature for 2 hours. Sections were then washed 3

times in PBS and mounted in Fluorescence Mounting Medium (Biomeda Corp., Foster City, CA;

cat # 17983-20). To quantify myocyte CSA, 6 thin sections per heart in 3 mice per each group

(wild type, α-MyHC-, Nkx2.5-, and Mef2C-lineage tracer mice) were analyzed. Each thin section

was divided into 12 approximately equal sectors and 5 high magnification (x400) microscopic

fields per sector were sampled (approximately 18,000 myocytes per mouse).

Immunoblotting (IB): We detected expression levels of the target proteins by IB. In

brief, we homogenized myocardial tissues in a relaxing lysis buffer (0.5% Nonidet P-40,

120[mM) sodium chloride, 50[mM] Tris-HCl pH: 7.4, 5% glycerol) containing proteinase

inhibitors (Roche Diagnostics, GmbH, Mannheim, Germany; cat# 11-697-498-001) and

determined the protein concentration by Bradford assay (Coomassie Protein Reagent; Pierce

Biotechnology, Rockford, IL; cat # 23200). We loaded 30 µg aliquots of total protein extracts

onto 12% SDS polyacrylamide gels, subjected to electrophoresis and transferred to cellulose

membranes. The primary antibody against EYFP was polyclonal rabbit IgG (Santa Cruz

Biotechnology, Santa Cruz, CA. #sc-8334, 1:200 dilution). The secondary antibody was

anti rabbit IgG-Horseradish Peroxidase (polyclonal, donkey IgG, General Electric Health-care,

cat #NA934V, 1:2000 dilution). We detected the signals by chemiluminescence (ECL detection

reagents and Hyperfilm by Amersham Bioscience, Piscataway, NJ, cat #RPN-2106 and 28-

9068-37, respectively).


4

We stripped the membranes in Restore PLUS Western Blot Stripping Buffer (Thermo

Scientific, Hudson, New Hampshire, cat #46430,) and reprobed the membranes with an anti α-

tubulin antibody (polyclonal, goat IgG, Santa Cruz Biotechnology Inc., Santa Cruz, CA, cat# sc-

12462, 1:600 dilution). The secondary antibody was polyclonal donkey anti goat IgG-

Horseradish Peroxidase (Santa Cruz Biotechnology cat #sc-2020, 1:2000 dilution).

Co-immunoprecipitation (Co-IP): We performed Co-IP as published 6. In brief, we

homogenized 50 mg aliquots of ventricular myocardium and prepared the protein extracts as

described above for IB. To co-immunoprecipitate, we gently mixed 6 µg aliquots of anti Tcf7l2

antibody (monoclonal, clone 6H5-3, mouse IgG2a isotype, Millipore – Upstate, Billerica, MA; cat

# 05-511) to 500 µg aliquots of total protein extracts and incubated the reaction at 4°C overnight

on a rocker platform. We added 20 µl of Protein A/G PLUS-Agarose beads (Santa Cruz

Biotechnology Inc. cat# sc 2003) to the solutions and incubated overnight on a rocker platform

at 4°C. We precipitated the proteins and resuspended the final pellets in Laemmli loading buffer

(63 mM Tris-HCl, pH:6.8, 25% glycerol, 2% SDS, 0.01% brome phenol blue) and 2.5% DTT for

IB. The primary antibodies for IB were anti β-catenin (monoclonal, clone BDI-480, mouse IgG1,

Santa Cruz Biotechnology, cat # sc-59893), anti FLAG (monoclonal, clone M2, mouse IgG1

isotype, Sigma, St Louis, MO, cat # F3165,) and anti pan-specific PG (N-terminal, aa45-114,

monoclonal, clone: 11E4, mouse IgG1-kappa isotype Invitrogen – Zymed, Carlsbad, CA, cat

#13-8500).

Separation of cardiac protein subfractions: We extracted nuclear, cytosolic and

membrane proteins using a commercial kit (Chemicon Int., Danvers, MA; cat #2145) as

described 8. In brief, we homogenized the whole hearts in 5 volumes of a cold buffer containing

HEPES (pH7.9), MgCl2, KCl, EDTA, Sucrose, Glycerol, Sodium OrthoVanadate and protease

inhibitors cocktail. We centrifuged the homogenates at 18,000g for 20 min to pellet membrane

and nuclear fractions and collected the supernatant containing the cytsolic proteins. We


5

resuspended the pellet in 100 µl of an ice cold buffer containing HEPES (pH7.9), MgCl2, NaCl,

EDTA, Glycerol, Sodium OrthoVanadate and proteinase inhibitors cocktail. After gentle mixing

and centrifugation at 18,000g, we collected the supernatant containing nuclear proteins. We

resuspended the residual pellet in 100 µl of a cold buffer containing HEPES (pH7.9), MgCl2,

KCl, EDTA, Sucrose, Glycerol, Sodium deoxycholate, NP-40, Sodium OrthoVanadate and

protease inhibitors, incubated for 20 min with gentle rocking and centrifuged at 18,000 g for 20

min to collect the supernatant containing membrane proteins fraction.

We probed the membranes with a rabbit anti DYKDDDDK-Tag antibody (Cell Signaling

Technology, Inc. Danvers, MA; cat #2368, 1:500 dilution) to detect expression of the Flag-

tagged PG (transgene) and rabbit anti PG antibody (Santa Cruz Biotechnology Inc., cat #sc

7900, 1:500 dilution) to detect expression of the endogenous and Flag-tagged PG (pan-PG

antibody). The secondary antibody was donkey anti rabbit IgG HRP-conjugated (Santa Cruz

Biotechnology Inc., cat # sc 2313, 1:3000 dilution).

Immunofluorescence (IF): We removed the hearts immediately from CO2-euthanized

mice and washed them in cold PBS to remove blood. We cut 1 mm thick sections at the level of

mid ventricle and embedded the thick sections in optimal cutting temperature compound

(Sakura-Finetek U.S.A. Inc., Torrance, CA), flash-froze in an isopentane-liquid nitrogen bath

and kept at −80 °C until use. We prepared and washed thin sections (5 µm) in PBS (x3) and

blocked the sections with 5% donkey or goat normal serum in PBS for 60 min at room

temperature (Santa Cruz Biotechnology, Inc., cat # sc-2044 and sc-2043, respectively). We

incubated the sections overnight at 4 °C with the primary antibodies at 1:200 -300 dilutions in

5% blocking solution. Antibodies were from Santa Cruz Biotechnology, Inc., unless specified.

The primary antibodies were goat polyclonal anti Islet1 (sc-23590), rabbit polyclonal anti Mef2C

(Aviva System Biology, San Diego, CA, cat #ARP37342-T100), rabbit polyclonal anti EYFP (sc-

32897), rabbit polyclonal anti PPAR-γ (sc-7196), goat polyclonal anti C/EBP-α (sc-9314), rabbit


6

polyclonal anti C/EBP-α (sc-61), rabbit polyclonal anti PG (sc-7900), and rabbit polyclonal anti-

Flag (Cell Signaling Technology, Inc. cat #2368). We treated the sections with the secondary

antibodies FITC labeled donkey anti rabbit IgG (Chemicon Int., Danvers, MA, cat #AP182-F)

and Texas red labeled donkey anti goat IgG (sc-2783) at dilutions of 1:1,000 -2,000 for 1 h at

room temperature. After the final PBS-wash, we mounted the sections in DAPI-containing Hard

SetTM mounting medium (Vector Laboratories, Inc, Burlingame, CA, cat # H-1500) and

examined under fluorescence microscopy.

To determine whether EYFP positive cells originated from a monocyte/macrophage

lineage, we co-stained heart tissues from wild type and lineage tracer mice for the expression of

EYFP and the monocyte/macrophage specific marker F4/80. The primary antibodies were rabbit

polyclonal anti EYFP (sc-32897) and rat monoclonal anti F4/80, (IgG 1, clone number: CI:A3-1,

Abcam, Cambridge, MA, cat # ab6640) at 1:200 and 1:1000 dilutions, respectively. The

secondary antibodies were goat anti rabbit IgG-FITC conjugate (sc-2012) at a 1:1000 dilution

and goat anti rat IgG-Texas Red conjugate (Abcam cat # ab6843-1) at a 1:400 dilution.

To further demonstrate the specificity of staining of cardiac adipocytes for expression of

EYFP, we isolated visceral and subcutaneous fat tissues from the wild type and lineage tracer

mice and stained tissue sections, as controls, using the same antibodies as in staining of

cardiac tissues.

To further support the specificity of immunostaining for the antigen, heart tissues were

incubated with the corresponding IgG isotypes at the same concentrations as the primary

antibody. Primary IgG controls were normal goat IgG (Santa Cruz, cat# sc-2028; normal rabbit

IgG (Santa Cruz, cat# sc-2027) and normal rat IgG (Vector Laboratories, cat # I-4000). The

secondary conjugated antibodies were also used at the same concentration as in the

corresponding staining.


7

To quantify the number of adipocytes that stained positive for expression of EYFP

reporter protein, we stained frozen thin myocardial sections from wild type, α -MyHC-, Nkx2.5-,

and Mef2C-lineage tracer mice for Oil Red O and the adjacent frozen thin sections from the

same mice for the expression of EYFP and C/EBPα. We divided each thin myocardial section

into 60 high magnification (x400) microscopic fields and scored the number of cells that stained

positive for Oil Red O (adipocytes), C/EBP-α and EYFP (the reporter protein) in each field in 14

to 44 sections and in 5 to 9 mice per group. We calculated the mean and SD of adipocytes that

stained positive for EYFP per mouse.

Human myocardial sections: Based on the findings of rare cardiac cell expressing

second heart field and adipogenic markers in the lineage tracer mice and given that fibro-

adiposis in humans with ARVC evolves over time, we postulated that human hearts with ARVC

may also contains cells expressing both lineage markers in the fibro-adiposis area. Therefore,

we co-stained paraffin-embedded thin myocardial sections from 3 human patients with autopsy

proven ARVC who were victims of SCD (ages between 28 to 38 years) for expression of

markers of cardiac progenitor cells (Isl1 or Mef2C) and adipocytes (C/EBP-α). We

deparaffinized the sections through treatment with xylene and retrieved the antigens by heating

the sections in 10 mM sodium citrate buffer (pH 6.0) in a microwave. As a control, we stained

thin sections of a normal heart for the above markers. The human heart samples were paraffin-

embedded and could not be stained for Oil Red O.

Statistical analysis: Statistical calculations (STATA-Intercooled v.10.1, StataCorp LP

College Station, Texas) were as published 6, 7, 9. We tested the variables for normality

assumptions and compared the differences for normally distributed continuous variables among

the three groups by ANOVA, followed by pairwise comparisons by Bonferroni test. We analyzed

variables that violated the normality assumption and the non-parametric variables by Kurskal-

Wallis test.


8

REFERENCES

1. Garcia-Gras E, Lombardi R, Giocondo MJ, Willerson JT, Schneider MD, Khoury DS, Marian AJ. Suppression of canonical Wnt/beta-catenin signaling by nuclear plakoglobin recapitulates phenotype of arrhythmogenic right ventricular cardiomyopathy. J Clin Invest. 2006;116:2012-2021.

2. Vasioukhin V, Bowers E, Bauer C, Degenstein L, Fuchs E. Desmoplakin is essential in epidermal sheet formation. Nat Cell Biol. 2001;3:1076-1085.

3. Moses KA, DeMayo F, Braun RM, Reecy JL, Schwartz RJ. Embryonic expression of an Nkx2-5/Cre gene using ROSA26 reporter mice. Genesis. 2001;31:176-180.

4. Agah R, Frenkel PA, French BA, Michael LH, Overbeek PA, Schneider MD. Gene Recombination in Postmitotic Cells. Targeted Expression of Cre Recombinase Provokes Cardiac-restricted, Site-specific Rearrangement in Adult Ventricular Muscle In Vivo. Journal of Clinical Investigation. 1997;100:169-179.

5. Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, Costantini F. Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol. 2001;1:4.

6. Lombardi R, Bell A, Senthil V, Sidhu J, Noseda M, Roberts R, Marian AJ. Differential interactions of thin filament proteins in two cardiac troponin T mouse models of hypertrophic and dilated cardiomyopathies. Cardiovasc Res. 2008;79:109-117.

7. Tsybouleva N, Zhang L, Chen S, Patel R, Lutucuta S, Nemoto S, DeFreitas G, Entman M, Carabello BA, Roberts R, Marian AJ. Aldosterone, through novel signaling proteins, is a fundamental molecular bridge between the genetic defect and the cardiac phenotype of hypertrophic cardiomyopathy. Circulation. 2004;109:1284-1291.

8. Cox B, Emili A. Tissue subcellular fractionation and protein extraction for use in mass-spectrometry-based proteomics. Nat Protocols. 2006;1:1872-1878.

9. Lutucuta S, Tsybouleva N, Ishiyama M, Defreitas G, Wei L, Carabello B, Marian AJ. Induction and reversal of cardiac phenotype of human hypertrophic cardiomyopathy mutation cardiac troponin T-Q92 in switch on-switch off bigenic mice. J Am Coll Cardiol. 2004;44:2221-2230.


9

Online TABLE I

Sequence of Oligonucleotide Primers for PCR and qPCR Probes

A. PCR oligonucleotide primers

Nkx2.5-Cre:

Nkx2.5 Forward: GATTAGCTTAAGCGGAGCTGGGTGTCC

Nkx2.5 Reverse: GTTCTGGAACCAGATCTTGACCTGCTGGGA

Cre recombinase: GCCGCATAACCAGTGAAACAGCATTGC

Mef2c-Cre:

Mef2c AHF Cre-Forward: CCAGGCAAAGGCAAGAATAA

Mef2c AHF Cre-Reverse: ATGTTTAGCTGGCCCAAATG

Desmoplakin:

Forward: TAAGCTCCCCTCACTTCTCCAG

Reverse: TTCTCTTTGTCTGTTGCCATGT

Enhanced Yellow Fluorescent Protein (EYFP):

Rosa26R-1 AAAGTCGCTCTGAGTTGTTAT

Rosa26R-2 GCGAAGAGTTTGTCCTCAACC

Rosa26R-3 GGAGCGGGAGAAATGGATATG

B. qPCR Probe Sequence:

BMP-7 TGTGGAGACCCTGGATGGGCAGAGC

Wnt 5b GCTGGCCGCCGGGCCGTGTATAAGA

CTGF GGAGGAAAACATTAAGAAGGGCAAA

c-myc CAGCAGCGACTCTGAAGAAGAGCAA

GAPDH GAACGGATTTGGCCGTATTGGGCGC


10

Online TABLE II

Echocardiographic Phenotype

NTG Nkx2.5-Cre: DspW/F Nkx2.5-Cre:DspW/F:EYFPF/F Mef2C-Cre:DspW/F Mef2C-Cre:DspW/F:EYFPF/F

N 12 11 11 8 7

Sex (males/females) 7/5 4/7 6/5 3/5 4/3

Age (months) 7.0 ± 2.8 7.5 ± 1.8 6.5 ± 1.2 4.6 ± 2.0 3.6 ± 0.5#

Body weight (g) 29.7 ± 4.2 31.6 ± 4.6 30.8 ± 4.6 32.4 ± 7.1 30.0 ± 2.3

Heart rate (bpm) 586 ± 84 605 ± 75 566 ± 76 625 ± 88 624 ± 54

IVST (mm) 0.96 ± 0.09 0.85 ± 0.12 0.79 ± 0.09# 0.91 ± 0.07 0.95 ± 0.11

PWT (mm) 0.96 ± 0.12 0.84 ± 0.13 0.79 ± 0.09# 0.83 ± 0.05 0.86 ± 0.09

LVEDD (mm) 2.79 ± 0.30 3.93 ± 0.91ξ 3.77 ± 0.42ξ 3.01 ± 0.30 2.87 ± 0.27

LVEDD/BW (mm/g) 0.095 ± 0.009 0.125 ± 0.027# 0.124 ± 0.019# 0.097 ± 0.023 0.096 ± 0.011

LVESD (mm) 0.97± 0.20 2.34 ± 1.11ξ 2.05 ± 0.42ξ 1.07 ± 0.14 0.99 ± 0.07

LVM (mg) 98± 36 188± 92# 156 ± 40 81 ± 18 79 ± 7

LVM/BW (mg/g) 3.2 ± 0.8 5.9± 2.6# 5.1 ± 1.2* 2.6 ± 0.7 2.6 ± 0.2

FS (%) 64.9 ± 6.4 43.1 ± 13.6ξ 46.0 ± 6.7ξ 64.5 ± 3.8 65.4 ± 2.6

E (m/s) 1.2 ± 0.5 1.1 ± 0.2 1.1 ± 0.2 0.9 ± 0.1 0.9 ± 0.1

Ao Vmax (m/s) 1.1 ± 0.6 1.1 ± 0.2 1.2 ± 0.3 1.1 ± 0.2 1.1 ± 0.1


11

ET (ms) 53 ± 11 42 ± 9 48 ± 10 50 ± 7 51 ± 5

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

-Abbreviations: bpm: Beats per minutes; IVST: Interventricular septal thickness; PWT: Posterior wall thickness; LVEDD: Left ventricular

end diastolic diameter; LVESD: Left ventricular end systolic diameter; BW: Body weight; LVM: Left ventricular mass; FS: Fractional

shortening; E: early mitral inflow velocities; Ao Vmax: Maximum aortic outflow velocity; m/s: meter per second; ET: Ejection time.

*p-value ≤ 0.05, # p-value ≤ 0.01, ξ p-value ≤ 0.001 by Bonferroni pairwise comparison with non-transgenics.

P<0.001

p<0.001

p=1.000

VW/B

W (m

g/g)

Online Figure I. Box plots showing median, 25%, 75%, outliers ( ) and adjacent values of ventricular weight/body weight (VW/BW) ratio. VW/BW ratio was increased significantly in the αMyHC- (N=12) and Nkx2.5-regulated (N=24) but not in Mef-2C-regulated (N=12) Dsp-deficient (with and without EYFP), as compared to wild type (N=26) mice. The pair wise p values are corrected for multiple comparisons by Bonferroni’ method.

Mef2C-Cre:DspW/F:R26-EYFPF/F

Nkx2.5-Cre:DspW/F:R26-EYFPF/F

αMyHC-Cre:DspW/F:R26-EYFPF/F

Wild type

Wild type (non-Transgenic)

Nkx2.5-Cre:DspW/F:R26-EYFPF/F Mef2C-Cre:DspW/F:R26-EYFPF/F

α-MyHC-Cre:DspW/F:R26-EYFPF/FA B

C D

Online Figure II. Echocardiographic phenotype. Representative M-mode echocardiograms of the leftventricle from wild type, αMyHC-Cre:DspW/F:R26-EYFPF/F , Nkx2.5-Cre:DspW/F:R26- EYFPF/F , andMef2C-Cre:DspW/F:R26-EYFPF/F mice are shown. Compared with non-transgenic mice (Panel A), Leftventricle diastolic and systolic diameters were increased and systolic function was reduced in αMyHC--Cre:DspW/F:R26-EYFPF/F and Nkx2.5-Cre:DspW/F:R26-EYFPF/F mice (Panels B and C). In contrastcardiac dimensions and function was preserved in Mef2C-Cre:DspW/F:R26-EYFPF/F mice (Panel D).

Oil Red O

IgG Isotypes

DNA:Goat IgGOil Red O

Oil Red O DNA:Rabbit IgG

Online Figure III. Panel A. Oil Red O, DAPI, C/EBP-α and EYFP co-stained sections in wild type (non-transgenic) mice are shown as controls. As shown, immunostaining did not show expression of C/EBP-αand EYFP in the heart of wild type mice.Panel B. Additional controls including stained sections with IgG isotypes from the corresponding hostanimals for the primary antibodies are shown to further substantiate the specificity of the findings.

A B

DNA

Wild type

C/EBPα EYFP

Online Figure IV. Oil Red O, DAPI, C/EBP-α and EYFP co-stained sections in visceral fat tissues fromNkx-Cre:DspW/F:R26-EYFPF/F and Mef2C-Cre:DspW/F:R26-EYFPF/F lineage tracer mice are shown ascontrols. As shown, EYFP is not expressed in visceral fat tissues from the lineage tracer mice.

Visceral FatNkx-Cre:DspW/F:R26-EYFPF/F

Visceral FatMef2C-Cre:DspW/F:R26-EYFPF/F

DNA

EYFP

Oil Red ODNAOil Red O

C/EBPαEYFPC/EBPα

Oil Red O DNA

F4/80

Nkx-Cre:DspW/F:R26-EYFPF/F (Heart)

EYFP

Nkx-Cre:DspW/F:R26-EYFPF/F (Spleen)

F4/80 (+ control)

DNA

DNA:Rat IgG (isotype) DNA:Rabbit IgG (isotype)

Online Figure V. Exclusion of macrophages as a cell source of excess adipocytes in ARVC. A. Absence ofexpression of F4/80, a specific marker of macrophage in adipocytes in Nkx2.5-Cre:DspW/F:R26- EYFPF/F micehearts. The adipocytes show expression of EYFP. B. Sections of spleen from Nkx2.5-Cre:DspW/F:R26-EYFPF/F

mice showing expression of F4/80, as a positive control for antibody. C. Myocardial sections stained for Oil RedO and specific IgG isotypes, as controls for the specificity of the antibodies tested.

A

EYFP

Oil Red OC

Oil Red O

B

Bright Field

Online Figure VI. Bright field and co-stained sections for C/EBP-α andDNA, Mef2C and DNA or Isl1 and DNA in a normal heart, as controls.

Normal Human Heart

Mef2C:DNA

C/EBP‐α:DNA

Isl1: DNA

Wild type PG transgenic mice

Online Figure VII. M-Mode echocardiograms from wild type and plakoglobin (PG) transgenicmice. The time scale represent 200 msec intervals. As shown cardiac size and function were normalin the PG transgenic mice and similar to that in the wild type (non-transgenic) mice.

Online Figure VIII. QuantitativePCR amplification plotsshowing expression levels ofGAPDH as control and fourtargets of the canonical Wntsignaling in non-transgenicand wild type PG transgenicmice. Amplification plots forGAPDH were practicallysuperimposed in the controland transgenic mice. Incontrast, expression levels ofWnt5b and BMP7, both knownto be inhibited by the canonicalWnt signaling and bothinvolved in adipogenesis, wereincreased by 3 to 4 folds. Inaccord with a shift toadipogenesis, relativeexpression level of CTGF, aninhibitor of adipogenesis wasdecreased by 5-fold. Similarly,relative expression level of c-Myc, activated by thecanonical Wnt signaling, wasdown regulated byapproximately 2-fold.

BMP‐7GAPDH

BMP‐7

GAPDH Wnt 5b

Wnt5b

GAPDH cMyc

cMyc

CTGF

CTGF

P<0.001

P<0.001 P<0.001

P<0.001

GAPDH

Online Figure IX. Pathogenesis of excess adipocytes in ARVC. Mutations in desmosomal proteins by disrupting proper desmosome assembly free plakoglobin (PG) to translocate from the desmosome to the nucleus. In the nucleus PG, also known as γ-catenin because of structural and functional similarity to it, suppresses the canonical Wnt signaling through Lef/Tcf transcription factors. The net effect is removal of the inhibitory effects of the canonical Wnt signaling on expression of BMP7 and Wnt5b, known promoters of adipogenesis and suppression of expression of CTGF, known inhibitor of adipogenesis. Together they promote differentiation of cardiac progenitor cells to adipocytes.

Adipocyte

Cardiac Progenitor Cell

ARVC mutation

PG

Expression of canonical Wntsignaling target genes

PG

Tcf/Lef

β‐Catenin

Wnt

GSK3β

β‐Catenin

Frizzled

PPAR‐γC/EBP‐α

Fat

Fat

Fat

Fat Fat

Fat

Fat

Fat

Fat

Fat

Fat

Fat

Fat

Fat

Fat

cMycBMP7

Wnt5b

CTGF

genetic fate mapping identifies second heart field...

Documents