Download Print - Circulation

Molecular Cardiology
␣-E-Catenin Inactivation Disrupts the Cardiomyocyte
Adherens Junction, Resulting in Cardiomyopathy and
Susceptibility to Wall Rupture
Farah Sheikh, PhD; Yinghong Chen, MD, PhD; Xingqun Liang, MD, PhD; Alain Hirschy, PhD;
Antine E. Stenbit, MD, PhD; Yusu Gu, MS; Nancy D. Dalton, RDCS; Toshitaka Yajima, MD, PhD;
Yingchun Lu, PhD; Kirk U. Knowlton, MD; Kirk L. Peterson, MD;
Jean-Claude Perriard, PhD; Ju Chen, PhD
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Background—␣-E-catenin is a cell adhesion protein, located within the adherens junction, thought to be essential in
directly linking the cadherin-based adhesion complex to the actin cytoskeleton. Although ␣-E-catenin is expressed in
the adherens junction of the cardiomyocyte intercalated disc, and perturbations in its expression are observed in models
of dilated cardiomyopathy, its role in the myocardium remains unknown.
Methods and Results—To determine the effects of ␣-E-catenin on cardiomyocyte ultrastructure and disease, we generated
cardiac-specific ␣-E-catenin conditional knockout mice (␣-E-cat cKO). ␣-E-cat cKO mice displayed progressive dilated
cardiomyopathy and unique defects in the right ventricle. The effects on cardiac morphology/function in ␣-E-cat cKO
mice were preceded by ultrastructural defects in the intercalated disc and complete loss of vinculin at the intercalated
disc. ␣-E-cat cKO mice also revealed a striking susceptibility of the ventricular free wall to rupture after myocardial
infarction.
Conclusions—These results demonstrate a clear functional role for ␣-E-catenin in the cadherin/catenin/vinculin complex
in the myocardium in vivo. Ablation of ␣-E-catenin within this complex leads to defects in cardiomyocyte structural
integrity that result in unique forms of cardiomyopathy and predisposed susceptibility to death after myocardial stress.
These studies further highlight the importance of studying the role of ␣-E-catenin in human cardiac injury and
cardiomyopathy in the future. (Circulation. 2006;114:&NA;-.)
Key Words: genes 䡲 myocytes 䡲 structure 䡲 cardiomyopathy 䡲 death, sudden
I
ntercalated discs are highly organized structural components of cardiac muscle that are thought to maintain
structural integrity and synchronize contraction of cardiac
tissue. Although disorganization of the cardiac intercalated
disc has been considered a “hallmark” of cardiac disease,1,2 to
date, there is limited information on whether alterations in its
structure and/or components play a causal role.
cardiomyopathies and other fatal defects.1,2,4 – 8 Components
of this complex include (1) cadherins, which are transmembrane proteins responsible for Ca2⫹-dependent homophilic
cell-cell adhesion9; (2) catenins (ie, ␣-, ␤-, and ␥-catenin),
which are cytoplasmic cadherin-binding partners that regulate
cadherin adhesive activity9 and are implicated in signaling10;
and (3) other catenin-related proteins, including vinculin and
␣-actinin.3,11–13 The role of cadherins in the postnatal heart
was recently established in studies that demonstrated that
cardiac-specific loss of N-cadherin resulted in loss of the
cardiac intercalated disc, dilated cardiomyopathy, and arrhythmogenic defects in the postnatal heart.14,15 Because all
junctional components were lost/reduced in N-cadherin–null
hearts,14 however, the mechanism and contribution of the
catenins remain to be determined in the myocardium and
disease.
Cadherin-mediated cell adhesion is regulated by distinct
proteins found within the extracellular and cytoplasmic do-
Clinical Perspective p 䡵䡵䡵
The intercalated disc is a complex entity that consists of an
array of proteins, which are categorized into 3 major junctional complexes: adherens junctions, desmosomes, and gap
junctions. Adherens junctions are anchoring junctions between cells that link the cell membrane to the cytoplasmic
actin cytoskeleton to provide strong cell adhesion.3 Considerable attention has focused on the adherens junction proteins
because human and mouse genetic studies have revealed that
mutations/deficiencies in their components are linked to
Received April 17, 2006; revision received June 13, 2006; accepted June 30, 2006.
From the Department of Medicine (F.S., Y.C., X.L., A.E.S., Y.G., N.D.D., T.Y., Y.L., K.U.K., K.L.P., J.C.), University of California-San Diego, La
Jolla, Calif, and Institute of Cell Biology (A.H., J.-C.P.), ETH, Swiss Federal Institute of Technology, Zurich, Switzerland.
The online-only Data Supplement can be found with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.106.634469/DC1.
Correspondence to Ju Chen, University of California-San Diego, 9500 Gilman Dr, La Jolla, CA 92093-0613. E-mail [email protected]
© 2006 American Heart Association, Inc.
Circulation is available at http://www.circulationaha.org
DOI: 10.1161/CIRCULATIONAHA.106.634469
1
2
Circulation
September 5, 2006
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
mains of cadherin.3 ␣-Catenins are key cytoplasmic molecules that are thought to indispensably link the cytoplasmic
domain of cadherin to the actin cytoskeleton.16 ␣-Catenins are
thought to provide this link by binding to ␤- or ␥-catenin
through their N-terminal domain and to actin directly through
their C-terminus or indirectly by binding to actin-binding
proteins, such as vinculin and ␣-actinin.11 On the basis of
recent structural studies, new models have suggested a
dynamic as opposed to static role for ␣-catenin and vinculin
in the cadherin/catenin/actin complex.11–13 Three ␣-catenin
subtypes have been described, which include ␣-E, -N and -T
catenins.17–19 ␣-E-catenin is ubiquitously expressed in tissues, mainly in epithelial cells, but also at high levels in the
heart.1,20 On the other hand, ␣-N-catenin expression is restricted to neural tissues,18 whereas ␣-T-catenin expression is
restricted to testis, muscle, and brain.20 ␣-E-catenin is the
most widely studied, and although it is highly expressed in
the heart, its functional importance is most notable in tumor
cells that have lost functional ␣-E-catenin protein, with
concomitant loss of cell aggregation and thus gain of invasive
capacity.21,22 In skin cells, studies have revealed dual functions for ␣-E-catenin, not only in adherens junction formation
but also in signaling.10
Although ␣-E-catenin has been studied primarily in the
context of cancer, there is evidence to suggest that it may play
a role in cardiac disease. Specifically, ␣-E-catenin is highly
expressed in the adherens junction of the cardiac intercalated
disc, and perturbations in its expression are associated with
dilated cardiomyopathy, a disease associated with and shown
to influence the cytoskeleton.1,2,20,23,24 Because ␣-E-catenin is
dynamically linked to the cytoskeleton in other cell types, we
hypothesized that ␣-E-catenin would play an essential role in
cardiac adherens junction ultrastructure and function in the
myocardium in vivo. ␣-Catenin is required for early vertebrate embryonic development8 and is ubiquitously expressed;
thus, we utilized a conditional approach to explore its
functional role in the myocardium. Specifically, we generated
cardiomyocyte-specific ␣-E-catenin conditional knockout
mice (␣-E-cat cKO) using the well-established ventricular
form of myosin light chain 2 (MLC2v)–Cre knock-in mice.25
We demonstrate that cardiomyocyte-specific inactivation
of ␣-E-catenin resulted in a unique form of cardiomyopathy,
which included intercalated disc ultrastructural defects and
complete loss of vinculin from the intercalated disc. We also
demonstrate that ␣-E-cat cKO mice are predisposed to
ventricular free wall rupture. These findings demonstrate a
clear functional role for ␣-E-catenin within the cadherin/
catenin/vinculin complex in the myocardium in vivo.
Methods
Experimental Animals
Floxed ␣-E-catenin mice (kind gift from Dr Elaine Fuchs, Rockefeller University, New York, NY) have been characterized previously10 and were used to generate cardiac-specific ␣-E-cat cKO
mice, as illustrated in Figure 1A. All animal procedures were in full
compliance with the guidelines approved by the University of
California–San Diego Animal Care and Use Committee.
Adult Mouse Cardiomyocyte Isolation
Please refer to the online-only Data Supplement.
RNA Blot Analysis
Total RNA was extracted from cardiomyocytes with Trizol (Invitrogen Corp, Carlsbad, Calif). Total RNA (20 ␮g) was electrophoresed,
blotted, and hybridized with full-length ␣-E-catenin cDNA, as
described previously.26
Protein Analysis
Total protein was prepared from cardiomyocytes, ventricles, and
atria as described previously.1 Membrane fractions were prepared by
layering ventricular extracts onto a linear 5% to 20% sucrose density
gradient and subjecting them to ultracentrifugation, fractionation,
and concentration as described previously.27 Immunodetection of
␣-catenin (Sigma-Aldrich, Inc, St Louis, Mo), ␣-T catenin (kind gift
from Dr Frans van Roy, Ghent University, Ghent, Belgium),
␤-catenin (Affiniti Research Products Ltd, Plymouth Meeting, Pa),
vinculin (Sigma-Aldrich), dystrophin,27 pan-cadherin (SigmaAldrich), desmoplakin (Serotec, Raleigh, NC), plakoglobin (Transduction Laboratories, BD Biosciences, San Jose, Calif), connexin 43
(Chemicon, Temecula, Calif), and all actin (Sigma-Aldrich), was
performed as described previously.1,28
Immunofluorescence Microscopy
Adult heart cryosections and cardiomyocytes were fixed with 4%
paraformaldehyde and stained with primary antibodies against
␣-catenin (1:500; Sigma-Aldrich), sarcomeric ␣-actinin (1:2000;
Sigma-Aldrich), ␤-catenin (1:2000; Affiniti Research Products Ltd),
and vinculin (1:2000; Sigma-Aldrich). Cells were then stained with
secondary antibodies (1:250; Molecular Probes, Invitrogen) and
Hoechst nuclear stain, followed by visualization of signals by
deconvolution imaging (Deltavision, Applied Precision, Inc, Issaquah, Wash).
Histology
Transverse sections were obtained at 8-␮m intervals and stained with
hematoxylin and eosin as described previously.28 Masson trichrome
(Sigma-Aldrich) and terminal deoxynucleotidyl transferasemediated biotinylated UTP nick end labeling (Roche Applied Science, Indianapolis, Ind) staining assays were performed on sections
according to the manufacturer’s instructions. DNA fragmentation
was verified under ultraviolet illumination with the DAPI nuclear
counterstain.
Echocardiography
Adult mice were anesthetized with 1% isoflurane and subjected to
echocardiography as described previously.29
Hemodynamic Analysis
Mice were anesthetized with ketamine/xylazine and ventilated, and a
1.4F Millar catheter-tip micromanometer catheter was inserted
through the right jugular vein into the right ventricle of the mouse
heart to obtain and record right ventricular pressures, as similarly
described for the left ventricle (LV).30
Electron Microscopy
Cardiac ventricles were processed for electron microscopy as described previously.25,28
Myocardial Infarction Model
A previously established myocardial ischemia/reperfusion protocol31
was modified to induce permanent left anterior descending branch
ligation in the mouse heart. Quantitative measurement of area at risk,
infarct size, and ratio of the infarct area/area at risk was done as
described previously.31
Statistical Analysis
Data presented in the text and figures are expressed as mean⫾SEM,
unless otherwise stated. Significance has been evaluated by the
2-tailed Student t test.
Sheikh et al
Role of ␣-E-Catenin in Cardiomyopathy/Rupture
3
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Figure 1. Generation of ␣-E-cat cKO mice. A, Schematic representation of mating strategy to generate ␣-E-cat cKO mice. B, RNA blot
analysis of ␣-E-catenin expression in WT and ␣-E-cat cKO purified ventricular cardiomyocytes. 28S RNA was used as a loading control. C, Protein blot analysis of ␣-E-catenin expression in WT and ␣-E-cat cKO atria and purified right ventricular and LV cardiomyocytes (CM). Sarcomeric ␣-actinin expression was used as a loading control. D, Immunofluorescence staining of ␣-E-catenin in WT
(closed arrow) and ␣-E-cat cKO (open arrow) cardiomyocytes. Adult cardiomyocytes were isolated from ␣-E-cat cKO mice and double
labeled with antibodies against ␣-catenin (red) and ␣-sarcomeric actinin (green), as well as being counterstained with Hoechst nuclear
stain (blue). Areas in cultures that contain both ␣-E-catenin positively and negatively stained ventricular cardiomyocytes were specifically selected. Bars represent 20 ␮m.
The authors had full access to the data and take full responsibility
for its integrity. All authors have read and agree to the manuscript as
written.
Results
Generation of ␣-E-cat cKO Mice
To specifically ablate ␣-E-catenin in ventricular cardiomyocytes, floxed ␣-E-catenin mice were bred with MLC2v-Cre
knock-in mice, which have been shown to mediate DNA
recombination of genes, specifically in ventricular cardiomyocytes.25 A breeding strategy was used to generate wildtype (WT) and cardiac-restricted ␣-E-cat cKO mice, as
illustrated in Figure 1A. ␣-E-cat cKO mice were born at
mendelian ratios and survived until adulthood.
To assess the efficiency of Cre, ␣-E-catenin expression
was assessed in purified cardiomyocytes by RNA and protein
analysis at 16 weeks. A significant 80⫾3.5% (n⫽3) reduction
in ␣-E-catenin RNA was observed in ␣-E-cat cKO purified
ventricular cardiomyocytes compared with controls (Figure
1B). A 90⫾7.0% (n⫽3) and 90⫾5.0% (n⫽3) decrease in
␣-E-catenin protein expression was observed in purified left
and right ventricular cardiomyocytes, respectively, but not in
atria of ␣-E-cat cKO mouse hearts compared with controls
(Figure 1C). Immunofluorescence staining demonstrated that
positively stained myocytes in ␣-E-cat-cKO mice, representative of WT cardiomyocytes, specifically expressed ␣-Ecatenin at the distal polar ends of the cardiomyocytes, which
indicates localization to the intercalated disc (Figure 1D).
Negatively stained cardiomyocytes within ␣-E-cat cKO mice
demonstrated absence of ␣-E-catenin staining at the intercalated disc (Figure 1D).
Progressive Ventricular Cardiomyopathy and
Cardiac Dysfunction in ␣-E-cat cKO Mice
At 32 weeks, ␣-E-cat cKO mice revealed modest dilation of
the LV and thinning of the right ventricle anterior wall
(Figure 2A). At 60 weeks, ␣-E-cat cKO mouse hearts showed
significant dilation of the LV and thinning of the right
ventricle anterior wall compared with controls (Figure 2A).
These results were reflected by the significant increase in
LV/body weight ratios and the decrease in right ventricle/
4
Circulation
September 5, 2006
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Figure 2. Regional right ventricle thinning
and LV dilation in ␣-E-cat cKO mice. A,
Paraffin sections from ␣-E-cat cKO and
control mice at 32 and 60 weeks were
stained for nuclei and cytoplasm with
hematoxylin and eosin, respectively.
Black arrow denotes the progressive
right ventricle thinning in ␣-E-cat cKO
mouse hearts. Black bar indicates
2.5 mm. B, Quantitative assessment of
transferase-mediated biotinylated UTP
nick end labeling–positive nuclei in ␣-Ecat cKO (n⫽3) and control mouse hearts
(n⫽3). *P⬍0.05. C, Masson trichrome
stain of ␣-E-cat cKO and control mouse
heart at 60 weeks. Bar represents 50 ␮m.
body weight ratios in ␣-E-cat cKO mice at 60 weeks
(Figure 3A) compared with controls. We also observed a
significant increase in cardiomyocyte apoptosis in ventricles of ␣-E-cat cKO mouse hearts at 60 weeks (Figure 2B).
Interestingly, we did not observe significant fibrosis in the
LV of ␣-E-cat cKO mouse hearts at 60 weeks; however,
fibrosis could be detected in the thinned regions of the
right ventricle (Figure 2C). No significant differences in
cardiomyocyte proliferation, as assessed by phosphorylated histone 3–positive cardiomyocyte nuclei, was detected
between ␣-E-cat cKO and control mouse hearts at 60
weeks (data not shown).
Echocardiography further revealed that ␣-E-cat cKO mice
had significant abnormalities in cardiac dimensions and
function, suggestive of dilated cardiomyopathy, at 60 weeks.
Figure 3B is a representative echocardiographic M-mode
image demonstrating the significant increase in LV internal
dimensions in ␣-E-cat cKO mice compared with control.
There were no significant differences in body weight (WT,
n⫽8, 35.3⫾0.9 g; ␣-E-cat cKO, n⫽10, 34.8⫾1.7 g) or heart
rate (WT, n⫽8, 468⫾14 bpm; ␣-E-cat cKO, n⫽10, 444⫾18
bpm) between ␣-E-cat cKO and WT mice. However, ␣-E-cat
cKO mice had significantly increased LV chamber size, as
determined by LV end-diastolic dimension and LV end-systolic dimension (Figure 3A). In addition, there was significant
wall thinning, as determined by the decrease in systolic
dimensions in interventricular septal wall thickness and LV
posterior wall thickness, in ␣-E-cat cKO mouse hearts compared with controls (Figure 3C). Most striking was the
presence of reduced LV function, as evidenced by reduced
percent fractional shortening in ␣-E-cat cKO mouse hearts
(Figure 3C). Despite functional changes in the LV, there was
no evidence of pulmonary edema, as assessed by lung
weights in ␣-E-cat cKO mice (WT, n⫽5, 0.214⫾0.02 g;
␣-E-cat cKO, n⫽6, 0.214⫾0.0.2 g). Hemodynamic analysis
of the right ventricle also revealed no significant differences
in right ventricular function between ␣-E-cat cKO mice and
controls at 60 weeks of age (see online Data Supplement). No
obvious cardiac phenotype was observed in heterozygous
floxed ␣-E-cat cKO mice.
Ultrastructural Abnormalities of the Cardiac
Intercalated Disc in ␣-E-cat cKO Mice
Electron microscopy was performed on ␣-E-cat cKO mouse
hearts to examine the integrity of the intercalated disc at the
ultrastructural level. LV and right ventricular regions were
obtained from ␣-E-cat cKO and WT mouse hearts at 32
weeks and assessed for ultrastructural changes in longitudinally oriented myocytes. Severe disorganization in intercalated disc structure was observed in the ␣-E-cat cKO mouse
heart compared with controls (Figure 4). Cardiac muscle from
␣-E-cat cKO mice revealed strikingly large, highly convoluted intercalated disc membranes with disorganized/lessaligned mitochondria. In contrast, cardiac muscle from WT
mice showed well-aligned/attached arrays of myofibrils at
Z-lines and intercalated discs, with visibly aligned
mitochondria.
Sheikh et al
Role of ␣-E-Catenin in Cardiomyopathy/Rupture
5
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Figure 3. Echocardiographic assessment of cardiac size and function in ␣-E-cat cKO and WT mice at 60 weeks. A, LV and right ventricular weight to body weight ratios in ␣-E-cat cKO (n⫽10) and WT (n⫽10) mice. B, Transthoracic M-mode echocardiography tracings
revealing dilated (arrows) end-systolic (LVIDs) and end-diastolic (LVIDd) LV internal dimensions in a representative ␣-E-cat cKO mouse
heart compared with a control mouse heart. HR indicates heart rate. C, Echocardiographic measurements, which were obtained from
transthoracic M-mode tracings of WT (n⫽12; average age 59.5 weeks) and ␣-E-cat cKO (n⫽12; average age 60.8 weeks) mice. IVSs
indicates interventricular septal wall thickness at end systole (mm); LVPWs, LV posterior wall thickness at end systole (mm); LVIDd and
LVIDs, LV internal dimension at end diastole and systole (mm); and LV %FS, LV percent fractional shortening. *P⬍0.05.
Alterations Within the Adherens Junction Protein
Complex in ␣-E-cat cKO Mouse Hearts
To determine the influence of ␣-E-catenin ablation on the
expression of other cardiac intercalated disc proteins, we
assessed the expression levels of other adherens junction
(pan-cadherin, ␣-T-catenin, ␤-catenin, vinculin, and plakoglobin), desmosome (desmoplakin and plakoglobin), and gap
junction (connexin 43) proteins in ␣-E-cat cKO and WT
mouse ventricles at 20 weeks of age. ␣-E-catenin protein
levels in ␣-E-cat cKO whole ventricles were 35.3⫾8.9%
(n⫽3; P⬍0.01; mean⫾SD) of control levels (Figure 5),
which was comparable to the decrease observed in purified
ventricular cardiomyocytes (Figure 1C), because ␣-E-catenin
is also expressed in noncardiomyocytes. Overall, no significant differences in ␣-T-catenin, plakoglobin, desmoplakin,
and connexin 43 levels were observed in ␣-E-cat cKO
ventricles (Figure 5). However, a significant increase in
␤-catenin levels (163.6⫾35.9% of WT levels; n⫽3; P⬍0.05)
and decrease in pan-cadherin levels (85.7⫾4.4% of WT
levels; n⫽3; P⬍0.05) were detected in ␣-E-cat cKO ventricles (Figure 5). The most striking difference was the significant reduction in vinculin levels (40⫾5.9% of WT levels; n⫽3;
P⬍0.01) observed in ␣-E-cat cKO ventricles (Figure 5).
Loss of ␣-E-Catenin/Vinculin Complex at Cardiac
Intercalated Disc of ␣-E-Cat cKO Mice
Figure 4. Ultrastructural analysis of ␣-E-cat cKO and WT mouse
hearts. Representative electron micrographs from the ventricular
myocardium (right ventricle) from WT and ␣-E-cat cKO mice at
32 weeks of age. Bar represents 0.73 ␮m.
Because vinculin is expressed in various compartments
within cardiomyocytes,32,33 we determined the subcellular
localization of loss of vinculin within ␣-E-cat cKO cardiomyocytes. Cardiomyocytes that stained positively for ␣-Ecatenin in WT and ␣-E-cat cKO mice revealed specific
6
Circulation
September 5, 2006
␣-E-catenin, ␤-catenin, and vinculin cofractionate in fractions 15 to 16 of the gradient (Figure 7), which suggests that
they are physically associated. However, in ␣-E-cat cKO
mouse ventricles, both ␣-E-catenin and vinculin were lost in
fractions 15 to 16 (Figure 7), which suggests physical
dissociation of the ␣-E-catenin/vinculin complex. Because
vinculin was still expressed in fractions that were shown not
to cofractionate with ␣-E-catenin,7–14 this presumably reflected vinculin expression at other subcellular locations
(Figure 7). In addition, ␤-catenin was still expressed in
fractions 15 to 16, where vinculin and ␣-E-catenin were lost
(Figure 7), which demonstrates that the loss of ␣-E-catenin
did not result in loss of expression of other adherens junction
proteins in these fractions. The presence of the plasma
membrane protein dystrophin in fractions 5 to 8 in both WT
and ␣-E-cat cKO cardiac membrane fractions further reinforced the efficiency in compartmentalization of different
proteins in the sucrose gradient and the specificity of the loss.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
␣-E-cat cKO Mice Are Susceptible to Ventricular
Wall Rupture
Figure 5. Alterations in adherens junction protein expression in
␣-E-cat cKO mouse whole ventricles. Top, Protein blot analyses
of pan-cadherins, ␣-E-catenin, ␣-T-catenin, ␤-catenin, vinculin,
desmoplakin, plakoglobin, and connexin 43 levels in control
(n⫽3) and ␣-E-cat cKO mouse (n⫽3) ventricular extracts at 20
weeks of age. All actin levels were assessed in mouse ventricles
to normalize for protein loading. Bottom, Densitometry analyses
of various intercalated disc proteins in ␣-E-cat cKO vs control
mice revealed significant changes in pan-cadherin, ␣-E-catenin,
␤-catenin, and vinculin. To avoid loading errors, the relative
expression level of each protein was first normalized to corresponding actin levels, and then the ␣-E-cat cKO results were
combined and expressed as a percentage of combined WT
results. 100% represented no change between ␣-E-cat cKO and
WT expression levels. **P⬍0.01, *P⬍0.05.
expression of vinculin at the intercalated disc and costameres
(green arrows in Figure 6). However, cardiomyocytes lacking
␣-E-catenin expression at the intercalated disc had loss of
vinculin expression at the intercalated disc (Figure 6). The
presence of vinculin at the costameres in ␣-E-cat cKO
myocytes further suggested that the membrane integrity of
the ␣-E-cat cKO myocyte was not compromised. Unlike
vinculin, we demonstrated that ␤-catenin was overexpressed
at the intercalated disc in ␣-E-cat cKO cardiomyocytes
(Figure 6), which reinforced the increase in expression
observed by protein blotting (Figure 5). The same localization
of ␤-catenin was observed in adult cardiomyocytes isolated
from WT mice (Figure 6).
To assess the physical integrity of the ␣-E-catenin/vinculin
complex in ␣-E-cat cKO mice, membrane fractions from
mouse ventricles were separated by sucrose density gradient
ultracentrifugation to analyze for the presence of ␣-E-catenin,
␤-catenin, vinculin, and dystrophin. In WT mouse ventricles,
Because plakoglobin (␥-catenin) conventional null mouse
hearts exhibited signs of rupture during embryonic development,4 we hypothesized that ␣-E-cat cKO mouse hearts
would also have increased vulnerability to cardiac rupture
after stress. Cardiac ruptures occur in 1% to 6% of patients
with acute myocardial infarction (MI),34 and therefore, we
assessed the effects of MI in ␣-E-cat cKO (n⫽14) and WT
(n⫽14) female mice at 20 weeks.31 No significant differences
in early mortality (within 24 hours of surgery) were observed
between the 2 groups after MI. In addition, all sham animals
of both genotypes survived the surgery and postoperative
period up to 7 days after MI. However, 82%, 54%, and 0% of
␣-E-cat cKO mice and 100% of WT mice were found to
survive at 5, 6, and 7 days after MI, respectively (Figure 8A).
Macroscopic assessment of ␣-E-cat cKO mice on autopsy
revealed the presence blood in the chest cavity in all mice
assessed (Figure 8B), suggestive of death as a result of
rupture. In most cases, the site of rupture in ␣-E-cat cKO
mouse hearts could be identified macroscopically by the
visible subepicardial hemorrhage that was localized at border
zones of the infarct area and right ventricle of the heart
(Figure 8C). ␣-E-cat cKO mouse hearts revealed the presence
of intramural hematomas (accumulation of multiple hemorrhagic foci; Figure 8D), which is a classic feature of rupture.35
Because the timing of rupture could not be predicted in
␣-E-cat cKO mice, we analyzed infarct size/area at risk in
␣-E-cat cKO and WT mice at 4 days after MI, a time point at
which all mice survive. No significant differences in infarct
size/area at risk could be detected between ␣-E-cat cKO
(40⫾4.5%, n⫽6) and WT (36⫾2.9%, n⫽6) mice.
Discussion
Recent studies have implicated a role for ␣-catenins in the
postnatal heart.1,2,20 However, because ␣-catenin plays a vital
role in early embryonic development,8 this has precluded its
ability to be studied in the postnatal heart with conventional
strategies. We utilized a conditional approach to determine
the role of ␣-E-catenin in cardiomyocytes in vivo. Using
Sheikh et al
Role of ␣-E-Catenin in Cardiomyopathy/Rupture
7
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Figure 6. Loss of vinculin expression at
the intercalated disc of ␣-E-cat cKO cardiomyocytes. Adult cardiomyocytes were
double labeled with antibodies against
␣-catenin (red) and vinculin (green) or
␤-catenin (green) and counterstained
with Hoechst nuclear stain (blue) to demonstrate specific loss of vinculin (red
arrow) and increased ␤-catenin (red
arrow) in ␣-E-cat cKO cardiomyocytes.
WT cardiomyocytes demonstrate colocalization of vinculin/␣-E-catenin (white
arrow) and ␤-catenin/␣-E-catenin (white
arrow). Green arrows denote costameric
vinculin staining. Areas in cultures that
contain both ␣-E-catenin positively and
negatively stained ventricular cardiomyocytes were specifically selected in ␣-Ecat cKO. Bars represent 15 or 20 ␮m.
MLC2v-Cre mice, we demonstrated the successful ablation of
␣-E-catenin expression in cardiomyocytes in ␣-E-cat cKO
mice.
Our studies initially suggested that ␣-E-catenin is dispensable for early cardiac development, because ␣-E-cat cKO
mice survive until adulthood up to 20 weeks with no
significant effects on cardiac ultrastructure and function.
However, because maximal ␣-E-catenin protein reduction
was not achieved until 16 weeks, we could not rule out the
possibility that the lack of a cardiac phenotype at earlier
stages of development could be due to incomplete Cremediated recombination during development.
Structural alterations in intercalated discs have long been
thought to be “hallmarks” of cardiac disease.1,2 We now
provide evidence from ␣-E-cat cKO adult mice demonstrating that ␣-E-catenin is a critical regulator of cardiac intercalated disc structure. Unlike cardiac-specific N-cadherin
knockout mice and keratinocyte-specific ␣-E-catenin knockout mice, in which intercalated discs or adherens junctions do
not form,10,14 ␣-E-cat cKO mouse ventricles display enlarged
and highly convoluted intercalated disc membranes. These
structural defects recapitulate the intercalated disc defects
observed in mouse and human models of dilated cardiomyopathy.1,7 It has been previously suggested that the increase in
membrane convolution between neighboring cardiomyocytes
can lead to impaired alignment between myocytes and thus
decrease the flexibility/stability/integrity of the contractile
tissue, resulting in decreased cardiac output that leads to
dilated cardiomyopathy.2 We have provided morphological
and functional evidence that demonstrates that ␣-E-cat cKO
mouse hearts do indeed progress to dilated cardiomyopathy.
Because the ultrastructural defects precede the morphological/functional defects in ␣-E-cat cKO mouse hearts, these
results provide clear evidence that defects in ␣-E-catenin, and
thus intercalated disc alterations, do indeed cause
cardiomyopathy.
The most surprising morphological defect within ␣-E-cat
cKO mice was the appearance of anterior right ventricle wall
8
Circulation
September 5, 2006
Figure 7. ␣-E-catenin deficiency causes
loss of vinculin from the adherens junction
enriched fractions. The integrity of the ␣-Ecatenin/vinculin complex within the cardiac
membrane fraction was studied with
sucrose density gradient ultracentrifugation. A, In WT mouse hearts, ␣-E-catenin,
␤-catenin, and vinculin cofractionate in
fractions 15 to 16 of the gradient. Vinculin,
on the other hand, could be detected in a
broad range of fractions.7–16 B, In ␣-E-cat
cKO mouse hearts, loss of ␣-E-catenin in
fractions 15 to 16 resulted in loss of vinculin in fractions 15 to 16 yet preservation of
␤-catenin in fractions 15 to 16 of the gradient. Vinculin could still be detected in fractions 6 to 14. Nonfractionated WT mouse
cardiac ventricular extracts (H) were used
as a positive control to detect endogenous
␣-E-catenin, ␤-catenin, vinculin, and
dystrophin.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
thinning. Despite morphological and ultrastructural defects in
␣-E-cat cKO mouse right ventricle, no significant defect in
right ventricular function could be detected. It is possible that
hemodynamic analyses may lack the sensitivity to detect
regional differences in right ventricular function. It is also
possible that the severity of the right ventricular defects may
not be sufficient to cause changes in overall function, which
is supported by the fact that pulmonary congestion, which is
a symptom of right heart failure, was not evident in ␣-E-cat
cKO mice. Right ventricular cardiomyopathies, such as arrhythmogenic right ventricular dysplasia, have been ascribed
to defects/mutations in proteins present within the desmosome complex, including desmoplakin and plakoglobin.6,36
However, we could not detect differences in desmoplakin or
plakoglobin expression in ␣-E-cat cKO mouse ventricles
compared with controls. We also did not detect differences in
␣-E-catenin expression between the right and left adult
cardiomyocytes in WT mouse hearts, thereby ruling out the
possibility that the endogenous pattern of ␣-E-catenin expression could account for the right ventricular defects. Therefore, it remains to be determined whether other right ventricle
factors or pathways exist that can influence ␣-E-catenin
dynamics.
The most striking molecular defect observed in ␣-E-cat
cKO mice was the complete loss/dissociation of vinculin at
the cardiac intercalated disc. Although structural studies have
proposed shared roles for ␣-catenin and vinculin in regulating
actin dynamics,11–13 to date, there is no information on the
shared roles of ␣-catenin and vinculin in the cadherin-catenin
complex in vivo. Human genetic studies have implied a role
for metavinculin at the intercalated disc and in disease.7,37
Interestingly, the intercalated disc defects found in human
patients with metavinculin mutations7 have striking similarities to the intercalated disc defects found within ␣-E-cat cKO
mouse ventricles. On the basis of our studies, we have
provided evidence that vinculin has shared functional roles
Figure 8. ␣-E-cat cKO mice show susceptibility to wall rupture after MI. A, Kaplan-Meier survival curve analysis of WT and ␣-E-cat
cKO mice up to 7 days after MI. Note the significant death in ␣-E-cat cKO mice at 5, 6, and 7 days post-MI. B, Macroscopic appearance of a representative ␣-E-cat cKO mouse heart within the mouse at 7 days post-MI, after autopsy. Note the blood in the chest cavity, dilated right ventricle, and most importantly, visible region of rupture (white arrows). C, Microscopic analysis of ruptured region in a
representative ␣-E-cat cKO mouse heart. Note the visible subepicardial hemorrhage within the ventricle (black arrow). Infarct area is
shown by *. D, Cardiac sections from ␣-E-cat cKO at 7 days after MI were stained for nuclei and cytoplasm with hematoxylin and
eosin, respectively. Note the large hematoma present within the ventricular free wall of a representative ␣-E-cat cKO mouse heart
(black arrow).
Sheikh et al
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
with ␣-E-catenin at the intercalated disc and more specifically, that ␣-catenin is required for vinculin localization to the
cardiac intercalated disc. Because defects in intercalated disc
ultrastructure appear after the molecular defects, we believe
that the loss of ␣-E-catenin and concomitant loss of vinculin
are critical for the stability of the cardiac intercalated disc in
vivo.
The significance of the loss of ␣-E-catenin/vinculin is
more clearly illustrated in MI studies. We showed that
␣-E-cat cKO adult mice have a striking susceptibility to free
wall rupture. Ruptures were not only found at the border
zones of the infarct but also within regions of the right
ventricle. It is well established that intercalated disc remodeling is observed in the heart after injury.5,38 The results of the
present study suggest that the loss of ␣-E-catenin/vinculin
may also be critical in intercalated disc remodeling at the
border zone of infarcts. The role of the right ventricle has
largely been ignored in MI studies; however, recent studies
have revealed that diminished right ventricular function has
been observed as a consequence of the ischemic LV.39 As a
result, we believe that the stress incurred in the ␣-E-cat cKO
right ventricle by the MI was sufficient to cause free wall
rupture. Because similar infarct sizes were observed in
␣-E-cat cKO and WT mice, this further reinforced that the
molecular disruptions within the adherens junction complex
are key factors regulating the structural susceptibility of the
ventricle. Heterozygous vinculin mice also exhibit cardiac
intercalated disc defects and predisposed susceptibility to
death after stress.33 Because we demonstrate specific loss of
vinculin at the intercalated disc of ␣-E-cat cKO myocytes, the
present study highlights the critical role of the cadherincatenin-vinculin complex in cardiac intercalated disc structure/integrity after stress in vivo.
In conclusion, the present study further highlights the
importance of studying the role of ␣-E-catenin in human
cardiac injury and cardiomyopathy. There is already some
evidence that suggests that ␣-E-catenin may play an important role in human heart rupture after MI.40 Thus, identification of ␣-E-catenin mutations within the human population
could be of critical importance in predicting populations
prone to inherited forms of cardiac disease and/or cardiac
rupture.
Acknowledgments
We thank Drs Elaine Fuchs (The Rockefeller University, New York,
NY) and Frans van Roy (Ghent University, Ghent, Belgium) for
providing us with the ␣-E-catenin floxed mice and ␣-T-catenin
antibodies, respectively. We thank Drs Sylvia Evans and MarieLouise Bang (UCSD, La Jolla, Calif) for critically reading the
manuscript.
Sources of Funding
This work was supported by National Institutes of Health grants (Dr
Ju Chen and Dr Knowlton). Dr Sheikh is a recipient of an American
Heart Association postdoctoral fellowship.
Disclosures
None
Role of ␣-E-Catenin in Cardiomyopathy/Rupture
9
References
1. Ehler E, Horowits R, Zuppinger C, Price RL, Perriard E, Leu M, Caroni
P, Sussman M, Eppenberger HM, Perriard JC. Alterations at the intercalated disk associated with the absence of muscle LIM protein. J Cell
Biol. 2001;153:763–772.
2. Perriard JC, Hirschy A, Ehler E. Dilated cardiomyopathy: a disease of the
intercalated disc? Trends Cardiovasc Med. 2003;13:30 –38.
3. Jamora C, Fuchs E. Intercellular adhesion, signalling and the
cytoskeleton. Nat Cell Biol. 2002;4:E101–E108.
4. Bierkamp C, McLaughlin KJ, Schwarz H, Huber O, Kemler R.
Embryonic heart and skin defects in mice lacking plakoglobin. Dev Biol.
1996;180:780 –785.
5. Ferreira-Cornwell MC, Luo Y, Narula N, Lenox JM, Lieberman M,
Radice GL. Remodeling the intercalated disc leads to cardiomyopathy in
mice misexpressing cadherins in the heart. J Cell Sci. 2002;115:
1623–1634.
6. McKoy G, Protonotarios N, Crosby A, Tsatsopoulou A, Anastasakis A,
Coonar A, Norman M, Baboonian C, Jeffery S, McKenna WJ. Identification of a deletion in plakoglobin in arrhythmogenic right ventricular
cardiomyopathy with palmoplantar keratoderma and woolly hair (Naxos
disease). Lancet. 2000;355:2119 –2124.
7. Olson TM, Illenberger S, Kishimoto NY, Huttelmaier S, Keating MT,
Jockusch BM. Metavinculin mutations alter actin interaction in dilated
cardiomyopathy. Circulation. 2002;105:431– 437.
8. Torres M, Stoykova A, Huber O, Chowdhury K, Bonaldo P, Mansouri A,
Butz S, Kemler R, Gruss P. An alpha-E-catenin gene trap mutation
defines its function in preimplantation development. Proc Natl Acad Sci
U S A. 1997;94:901–906.
9. Aberle H, Schwartz H, Kemler R. Cadherin-catenin complex: protein
interactions and their implications for cadherin function. J Cell Biochem.
1996;61:514 –523.
10. Vasioukhin V, Bauer C, Degenstein L, Fuchs E. Hyperproliferation and
defects in epithelial polarity upon conditional ablation of alpha-catenin in
skin. Cell. 2001;104:605– 617.
11. Drees F, Pokutta S, Yamada S, Nelson WJ, Weis WI. Alpha-catenin is a
molecular switch that binds E-cadherin-beta-catenin and regulates actinfilament assembly. Cell. 2005;123:903–915.
12. Janssen ME, Kim E, Liu H, Fujimoto LM, Bobkov A, Volkmann N,
Hanein D. Three-dimensional structure of vinculin bound to actin filaments. Mol Cell. 2006;21:271–281.
13. Yamada S, Pokutta S, Drees F, Weis WI, Nelson WJ. Deconstructing the
cadherin-catenin-actin complex. Cell. 2005;123:889 –901.
14. Kostetskii I, Li J, Xiong Y, Zhou R, Ferrari VA, Patel VV, Molkentin JD,
Radice GL. Induced deletion of the N-cadherin gene in the heart leads to
dissolution of the intercalated disc structure. Circ Res. 2005;96:346 –354.
15. Li J, Patel VV, Kostetskii I, Xiong Y, Chu AF, Jacobson JT, Yu C,
Morley GE, Molkentin JD, Radice GL. Cardiac-specific loss of
N-cadherin leads to alteration in connexins with conduction slowing and
arrhythmogenesis. Circ Res. 2005;97:474 – 481.
16. Rimm DL, Koslov ER, Kebriaei P, Cianci CD, Morrow JS. Alpha 1(E)catenin is an actin-binding and -bundling protein mediating the
attachment of F-actin to the membrane adhesion complex. Proc Natl Acad
Sci U S A. 1995;92:8813– 8817.
17. Herrenknecht K, Ozawa M, Eckerskorn C, Lottspeich F, Lenter M,
Kemler R. The uvomorulin-anchorage protein alpha catenin is a vinculin
homologue. Proc Natl Acad Sci U S A. 1991;88:9156 –9160.
18. Hirano S, Kimoto N, Shimoyama Y, Hirohashi S, Takeichi M. Identification of a neural alpha-catenin as a key regulator of cadherin function
and multicellular organization. Cell. 1992;70:293–301.
19. Nagafuchi A, Takeichi M, Tsukita S. The 102 kd cadherin-associated
protein: similarity to vinculin and posttranscriptional regulation of
expression. Cell. 1991;65:849 – 857.
20. Janssens B, Goossens S, Staes K, Gilbert B, van Hengel J, Colpaert C,
Bruyneel E, Mareel M, van Roy F. alphaT-catenin: a novel tissue-specific
beta-catenin-binding protein mediating strong cell– cell adhesion. J Cell
Sci. 2001;114:3177–3188.
21. Breen E, Clarke A, Steele G Jr, Mercurio AM. Poorly differentiated colon
carcinoma cell lines deficient in alpha-catenin expression express high
levels of surface E-cadherin but lack Ca(2⫹)-dependent cell– cell
adhesion. Cell Adhes Commun. 1993;1:239 –250.
22. Morton RA, Ewing CM, Nagafuchi A, Tsukita S, Isaacs WB. Reduction
of E-cadherin levels and deletion of the alpha-catenin gene in human
prostate cancer cells. Cancer Res. 1993;53:3585–3590.
10
Circulation
September 5, 2006
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
23. Bowles NE, Bowles KR, Towbin JA. The “final common pathway”
hypothesis and inherited cardiovascular disease: the role of cytoskeletal
proteins in dilated cardiomyopathy. Herz. 2000;25:168 –175.
24. Chen J, Chien KR. Complexity in simplicity: monogenic disorders and
complex cardiomyopathies. J Clin Invest. 1999;103:1483–1485.
25. Chen J, Kubalak SW, Chien KR. Ventricular muscle-restricted targeting
of the RXRalpha gene reveals a non-cell-autonomous requirement in
cardiac chamber morphogenesis. Development. 1998;125:1943–1949.
26. Chu PH, Ruiz-Lozano P, Zhou Q, Cai C, Chen J. Expression patterns of
FHL/SLIM family members suggest important functional roles in skeletal
muscle and cardiovascular system. Mech Dev. 2000;95:259 –265.
27. Lee GH, Badorff C, Knowlton KU. Dissociation of sarcoglycans and the
dystrophin carboxyl terminus from the sarcolemma in enteroviral cardiomyopathy. Circ Res. 2000;87:489 – 495.
28. Zhou Q, Chu PH, Huang C, Cheng CF, Martone ME, Knoll G, Shelton
GD, Evans S, Chen J. Ablation of Cypher, a PDZ-LIM domain Z-line
protein, causes a severe form of congenital myopathy. J Cell Biol. 2001;
155:605– 612.
29. Tanaka N, Dalton N, Mao L, Rockman HA, Peterson KL, Gottshall KR,
Hunter JJ, Chien KR, Ross J Jr. Transthoracic echocardiography in
models of cardiac disease in the mouse. Circulation. 1996;94:1109 –1117.
30. Arber S, Hunter JJ, Ross J Jr, Hongo M, Sansig G, Borg J, Perriard JC,
Chien KR, Caroni P. MLP-deficient mice exhibit a disruption of cardiac
cytoarchitectural organization, dilated cardiomyopathy, and heart failure.
Cell. 1997;88:393– 403.
31. Chen Y, Davis-Gorman G, Watson RR, McDonagh PF. Chronic ethanol
consumption modulates myocardial ischaemia-reperfusion injury in
murine AIDS. Alcohol Alcohol. 2003;38:18 –24.
32. Danowski BA, Imanaka-Yoshida K, Sanger JM, Sanger JW. Costameres
are sites of force transmission to the substratum in adult rat cardiomyocytes. J Cell Biol. 1992;118:1411–1420.
33. Zemljic-Harpf AE, Ponrartana S, Avalos RT, Jordan MC, Roos KP,
Dalton ND, Phan VQ, Adamson ED, Ross RS. Heterozygous inactivation
of the vinculin gene predisposes to stress-induced cardiomyopathy.
Am J Pathol. 2004;165:1033–1044.
34. Becker RC, Gore JM, Lambrew C, French W, Rogers WJ. A composite
view of cardiac rupture in the United States National Registry of Myocardial Infarction. J Am Coll Cardiol. 1996;27:1321–1326.
35. Gao XM, Xu Q, Kiriazis H, Dart AM, Du XJ. Mouse model of postinfarct ventricular rupture: time course, strain- and gender-dependency,
tensile strength, and histopathology. Cardiovasc Res. 2005;65:469 – 477.
36. Alcalai R, Metzger S, Rosenheck S, Rosenheck S, Meiner V,
Chajek-Shaul T. A recessive mutation in desmoplakin causes arrhythmogenic right ventricular dysplasia, skin disorder, and woolly hair. J Am
Coll Cardiol. 2003;42:319 –327.
37. Maeda M, Holder E, Lowes B, Valent S, Bies RD. Dilated cardiomyopathy associated with deficiency of the cytoskeletal protein metavinculin.
Circulation. 1997;95:17–20.
38. Matsushita T, Oyamada M, Kurata H, Masuda S, Takahashi A, Emmoto
T, Shiraishi I, Wada Y, Oka T, Takamatsu T Formation of cell junctions
between grafted and host cardiomyocytes at the border zone of rat
myocardial infarction. Circulation. 1999;100(suppl II):II-262–II-268.
39. Jerzewski A, Steendijk P, Pattynama PM, Leeuwenburgh BP, de Roos A,
Baan J. Right ventricular systolic function and ventricular interaction
during acute embolisation of the left anterior descending coronary artery
in sheep. Cardiovasc Res. 1999;43:86 –95.
40. van den Borne SWM, Lijnen P, Daemen MJAP, Smits JFM, Blankesteijn
WM. Alterations in alpha catenin in the adherens junctions of cardiomyocytes in infarct rupture in humans. Presented at: Keystone Symposia:
Molecular Biology of Cardiac Disease/Cardiac Development and Congenital Heart Disease; 2004; Keystone, Colo. Abstract 318.
CLINICAL PERSPECTIVE
End-to-end connections between myocardial cells, known as intercalated discs, are thought to maintain structural integrity
and synchronize contraction of cardiac tissue. Although abnormalities in intercalated disc structure are considered
“hallmarks” of cardiomyopathy, to date there is limited information on whether alterations in its structure or components
play a causal role. Thus, an improved understanding of the molecular interactions that are important for formation of the
intercalated disc may allow for the identification of novel targets for therapy for certain forms of heart disease and
identification of hereditary causes of the disease. The intercalated disc is thought to connect adjacent myocytes through
transmembrane adhesion molecules such as cadherin, which then interact with the contractile apparatus via intracellular
linking molecules such as the catenins and vinculin. Recent studies have implied a role for the connector molecule
␣-E-catenin in cardiomyopathy; its role in the myocardium remains unclear, however. In the present study, we provide
genetic evidence that implies a role for ␣-E-catenin in dilated cardiomyopathy and susceptibility to cardiac rupture after
myocardial infarction. These studies further demonstrate that the proper formation of intercalated discs requires
␣-E-catenin inside the cell to act as a link to vinculin and the contractile apparatus. These studies also suggest the clinical
importance of studying the role of ␣-E-catenin in human cardiac injury and cardiomyopathy in the future. Thus, strategies
to identify ␣-E-catenin and vinculin mutations within the human population could be of critical importance in predicting
populations prone to inherited forms of cardiomyopathy or cardiac rupture.
α-E-Catenin Inactivation Disrupts the Cardiomyocyte Adherens Junction, Resulting in
Cardiomyopathy and Susceptibility to Wall Rupture
Farah Sheikh, Yinghong Chen, Xingqun Liang, Alain Hirschy, Antine E. Stenbit, Yusu Gu,
Nancy D. Dalton, Toshitaka Yajima, Yingchun Lu, Kirk U. Knowlton, Kirk L. Peterson,
Jean-Claude Perriard and Ju Chen
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Circulation. published online August 21, 2006;
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2006 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circ.ahajournals.org/content/early/2006/08/21/CIRCULATIONAHA.106.634469.citation
Data Supplement (unedited) at:
http://circ.ahajournals.org/content/suppl/2006/08/21/CIRCULATIONAHA.106.634469.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located,
click Request Permissions in the middle column of the Web page under Services. Further information about
this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/