Download Westermann D, Mersmann J, Melchior A, Petrik C, Freudenberger T

Biglycan Is Required for Adaptive Remodeling After Myocardial Infarction
D. Westermann, J. Mersmann, A. Melchior, T. Freudenberger, C. Petrik, L. Schaefer,
R. Lüllmann-Rauch, O. Lettau, C. Jacoby, J. Schrader, S.-M. Brand-Herrmann, M.F.
Young, H.P. Schultheiss, B. Levkau, H.A. Baba, T. Unger, K. Zacharowski, C.
Tschöpe and J.W. Fischer
Circulation 2008;117;1269-1276; originally published online Feb 25, 2008;
DOI: 10.1161/CIRCULATIONAHA.107.714147
Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX
72514
Copyright © 2008 American Heart Association. 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/cgi/content/full/117/10/1269
Subscriptions: Information about subscribing to Circulation is online at
http://circ.ahajournals.org/subscriptions/
Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters
Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax:
410-528-8550. E-mail:
[email protected]
Reprints: Information about reprints can be found online at
http://www.lww.com/reprints
Downloaded from circ.ahajournals.org at SWETS SUBSCR SVC/25338706 on September 15, 2008
Heart Failure
Biglycan Is Required for Adaptive Remodeling After
Myocardial Infarction
D. Westermann, MD*; J. Mersmann, MD*; A. Melchior, MS*; T. Freudenberger, MS*; C. Petrik;
L. Schaefer, PhD; R. Lüllmann-Rauch, MD; O. Lettau, MS; C. Jacoby, PhD; J. Schrader, MD;
S.-M. Brand-Herrmann, MD, PhD; M.F. Young, PhD; H.P. Schultheiss, MD; B. Levkau, MD;
H.A. Baba, MD; T. Unger, MD; K. Zacharowski, MD; C. Tschöpe, MD; J.W. Fischer, PhD
Background—After myocardial infarction (MI), extensive remodeling of extracellular matrix contributes to scar formation
and preservation of hemodynamic function. On the other hand, adverse and excessive extracellular matrix remodeling
leads to fibrosis and impaired function. The present study investigates the role of the small leucine-rich proteoglycan
biglycan during cardiac extracellular matrix remodeling and cardiac hemodynamics after MI.
Methods and Results—Experimental MI was induced in wild-type (WT) and bgn⫺/0 mice by permanent ligation of the left
anterior descending coronary artery. Biglycan expression was strongly increased at 3, 7, and 14 days after MI in WT
mice. bgn⫺/0 mice showed increased mortality rates after MI as a result of frequent left ventricular (LV) ruptures.
Furthermore, tensile strength of the LV derived from bgn⫺/0 mice 21 days after MI was reduced as measured ex vivo.
Collagen matrix organization was severely impaired in bgn⫺/0 mice, as shown by birefringence analysis of Sirius red
staining and electron microscopy of collagen fibrils. At 21 days after MI, LV hemodynamic parameters were assessed
by pressure-volume measurements in vivo to obtain LV end-diastolic pressure, end-diastolic volume, and end-systolic
volume. bgn⫺/0 mice were characterized by aggravated LV dilation evidenced by increased LV end-diastolic volume (bgn⫺/0,
111⫾4.2 ␮L versus WT, 96⫾4.4 ␮L; P⬍0.05) and LV end-diastolic pressure (bgn⫺/0, 24⫾2.7 versus WT, 18⫾1.8 mm Hg;
P⬍0.05) and severely impaired LV function (EF, bgn⫺/0, 12⫾2% versus WT, 21⫾4%; P⬍0.05) 21 days after MI.
Conclusion—Biglycan is required for stable collagen matrix formation of infarct scars and for preservation of cardiac
hemodynamic function. (Circulation. 2008;117:1269-1276.)
Key Words: collagen 䡲 myocardial infarction 䡲 extracellular matrix 䡲 fibrosis 䡲 proteoglycans
E
xtensive remodeling of the extracellular matrix (ECM)
occurs during all phases of infarct healing, including
initial degradation of ECM by matrix metalloproteinases
(MMPs), the formation of provisional ECM by de novo synthesis, and finally the deposition of collagen-forming scar tissue.1
During scar maturation, inflammatory cells and fibroblasts
disappear as a result of apoptosis, and collagen becomes increasingly crosslinked to stabilize the scar and to provide tensile
strength until a hypocellular, collagen-rich scar is formed.2 The
composition of the ECM during the individual phases of infarct
healing not only determines the mechanical stability but also
modulates the cellular phenotype.3 In the long term, adverse left
ventricular (LV) remodeling, including fibrosis and ventricular
dilatation, often causes progressive heart failure associated with
considerable morbidity and mortality after acute myocardial
infarction (MI).
Clinical Perspective p 1276
The role of MMPs during postinfarct remodeling has been
studied extensively. After MI, an imbalance between MMPs
Received May 8, 2007; accepted December 24, 2007.
From the Abteilung für Kardiologie und Pneumologie, Charite-Universitätsklinikum Berlin, Campus Benjamin Franklin, Berlin, Germany (D.W., O.L.,
H.P.S., C.T.); Molecular Cardioprotection and Inflammation Group, Klinik für Anästhesiologie (J.M.) and Molekulare Pharmakologie, Institut für
Pharmakologie und Klinische Pharmakologie (A.M., T.F., J.W.F.), Universitätsklinikum Düsseldorf, Düsseldorf, Germany; Center for Cardiovascular
Research, Institut für Pharmakologie und Toxikologie, Campus Charite-Mitte, Charite-Universitatsmedizin Berlin, Berlin, Germany (C.P., T.U.);
Nephropharmakologie, Allgemeine Pharmakologie und Toxikologie, Klinikum der Johann Wolfgang Goethe Universität Frankfurt, Frankfurt, Germany
(L.S.); Institut für Anatomie, Christian-Albrechts-Universität Kiel, Kiel, Germany (R.L.-R.); Institut für Herzkreislauf-Physiologie, Heinrich-HeineUniversität Düsseldorf, Düsseldorf, Germany (C.J., J.S.); Leibniz-Institute for Arteriosclerosis Research, Department of Molecular Genetics of
Cardiovascular Disease, University of Muenster, Muenster, Germany (S.-M.B.-H.); Craniofacial and Skeletal Diseases Branch, National Institute of
Dental and Craniofacial Research, National Institutes of Health, Bethesda, Md (M.F.Y.); Institut für Pathophysiologie (B.L.) and Institut für Pathologie
(H.A.B.), Universitätsklinikum Essen, Essen, Germany; and Molecular Cardioprotection and Inflammation Group, Department of Anaesthesia, University
Hospital Bristol, Bristol Royal Infirmary, Bristol, UK (K.Z.).
*The first 4 authors contributed equally to this work.
The online Data Supplement, which contains an expanded Methods section and figures, can be found with this article at http://circ.ahajournals.org/
cgi/content/full/CIRCULATIONAHA.107.714147/DC1.
Correspondence to Jens W. Fischer, Molecular Pharmacology, Institut für Pharmakologie and Klinische Pharmakologie, Heinrich-Heine-Universität
Düsseldorf, Moorenstrasse 5, D-40225 Düsseldorf, Germany. E-mail [email protected]
© 2008 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org
DOI: 10.1161/CIRCULATIONAHA.107.714147
1269
Downloaded from circ.ahajournals.org at SWETS
SUBSCR SVC/25338706 on September 15, 2008
1270
Circulation
March 11, 2008
and tissue inhibitors of MMPs (TIMPs) is thought to be a key
factor in promoting adverse remodeling of the ECM and LV
dilatation. However, collagen synthesis and fibrillogenesis
also underlie a complex regulation and might in turn be
equally important during infarct healing. Self-assembly of
procollagen ␣ chains forms triple-helical collagen molecules
that associate into collagen fibrils. This process is regulated
by accessory ECM molecules such as the family of small
leucine-rich proteoglycans, including biglycan and decorin.4
By controlling lateral assembly of triple helixes into fibrils
and the arrangement of collagen fibers, biglycan and decorin
are key regulators of collagen fibrillogenesis and fibril
diameter.4,5 Biglycan and decorin are capable of specific
binding to collagen types I, II, III, and VI6,7 via the core
proteins near the d band of collagen fibrils.8,9 The role of
small leucine-rich proteoglycans in LV remodeling is only
partially understood. It has been reported previously that
biglycan and decorin are induced after MI in rats and that
biglycan associates with cardiac fibroblasts in the infarcted
myocardium.10,11 bgn⫺/0 mice have reduced skeletal growth
and develop an osteoporosis-like phenotype resulting from a
defect in bone formation.12 Furthermore, the diameter of
collagen fibrils was reduced in bgn⫺/0 mice.13 Other functions
of biglycan include regulation of transforming growth factor
␤-1 activity,14 cell proliferation, apoptosis,15 and activation of
inflammatory responses mediated through activation of Tolllike receptor-2 and -4.16 Taken together, biglycan plays
important roles in collagen matrix assembly, control of
growth factor activity, and cellular phenotype, including
inflammatory cells. However, it is still unknown whether
biglycan plays a role in cardiac physiology or during infarct
healing. The aim of the present study was to investigate the
role of biglycan during remodeling and functional adaptation
after experimental MI. For this purpose, bgn⫺/0 mice were
subjected to permanent occlusion of the left anterior descending coronary artery, followed by analysis of heart morphology, ECM, and hemodynamic function.
Hemodynamic Measurements
Mice were anesthetized by injection of thiopental (125 mg/kg IP),
intubated, and ventilated with a respirator (type 7025, Ugo Basile,
Comerio, Italy). A 1.4F microconductance pressure catheter (ARIA
SPR-719, Millar Instruments Inc, Houston, Tex) was positioned in
the LV via the right carotid artery for continuous registration of LV
pressure-volume loops in closed-chest animals.18 Systolic function
and myocardial contractility were quantified by LV end-systolic
pressure, peak rate of rise in LV pressure (dP/dtmax), ejection
fraction, cardiac output, end-systolic volume, and stroke volume.
Diastolic performance was measured by LV end-diastolic pressure,
peak dP/dtmin, end-diastolic volume, and time constant of isovolumic pressure relaxation.
Analysis of Human MIs
Tissue sections derived from patients who died of ischemic MI
presenting either as an acute, primary event (⬍5 days, n⫽4)) or
reoccurring infarction on the basis of old, healed MIs (n⫽4) were
analyzed for biglycan by immunostaining (LF 51) and for collagen
by Sirius red staining. The samples were collected for diagnostic
purposes during autopsy and were examined by a senior pathologist
(H.B.). Informed consent was obtained from patients or relatives.
The present study was performed according to the Declaration of
Helsinki, and the study protocol was approved by the Ethics
Committee of the University Hospital of Essen.
Statistical Analysis
For multiple comparisons, 2-way ANOVA, followed by Bonferroni
posttest, was performed. Values of P⬍0.05 were considered significant.
Data are expressed as mean⫾SEM. Survival curves were computed by the
Kaplan-Meier method and were compared by use of the log-rank test.
The authors had full access to and take responsibility for the
integrity of the data. All authors have read and agree to the
manuscript as written.
Results
Increased Biglycan Expression After MI
Real-time reverse-transcription polymerase chain reaction
(RT-PCR) revealed strongly increased bgn mRNA expression
in the LV after experimental MI, peaking 7 days after MI
(Figure 1A). These results were confirmed by immunostaining for biglycan, showing that biglycan accumulation occurs
in the infarcted ischemic area and border zone at 7 days
(Figure 1C). Furthermore, at 7 days after MI, fibroblasts and
inflammatory cells were detected in areas enriched with
Methods
A
A detailed description of the methods and materials is available in
the online Data Supplement.
Animals
bgn⫺/0 mice (bred at the University of Düsseldorf) with a targeted
deletion of the bgn gene12 and male wild-type (WT; C57BL/6)
littermates were compared in this study. All procedures were carried
out in accordance with the Association for Assessment and Accreditation of Laboratory Animal Care guidelines and the Guide for the
Care and Use of Laboratory Animals (Department of Health and
Human Services, National Institutes of Health, publication No.
86-23) and were approved by the ethics and research boards of the
University of Düsseldorf. Animals (12 to 16 weeks old) were
randomized as indicated.
Myocardial Infarction
At the age of 12 weeks, WT and bgn⫺/0 mice were anesthetized by
injection of pentobarbital (100 mg/kg IP). Mice were subjected to
permanent left anterior descending coronary artery occlusion, followed by recovery for 7 or 21 days. All procedures were carried out
as recently described in detail.17
B
C
D
Figure 1. Time course of biglycan expression after MI. A, Realtime RT-PCR revealed significantly increased bgn mRNA expression, peaking 7 days after MI. B, Immunohistochemistry of biglycan
in LV of sham-operated WT mice (inset, negative control) and WT
mice after left anterior descending coronary artery occlusion at 7
(C) and 14 (D) days. n⫽6. Magnification ⫻100. *P⬍0.05.
Downloaded from circ.ahajournals.org at SWETS SUBSCR SVC/25338706 on September 15, 2008
Westermann et al
biglycan (online Data Supplement), in line with previous
results in rats.10,11 At 14 days after MI, bgn mRNA levels
were still 5-fold higher than in sham animals (Figure 1A), and
biglycan immunostaining was still increased in the infarct and
peri-infarct zones (Figure 1D).
Role of Biglycan During Post-MI Remodeling
1271
A
Cardiac Phenotype of Adult bgnⴚ/0 Mice
bgn⫺/0 mice were used as a model to study the effect of
biglycan on cardiac function. First, the cardiac phenotype was
analyzed in healthy, noninfarcted animals. At 6 months of
age, morphological and hemodynamic analyses of the heart
did not reveal any differences between WT and bgn⫺/0 mice.
In 15-month-old mice, the body weight (WT, 48.5⫾4.5 g
versus bgn⫺/0, 41.3⫾5.2 g [mean⫾SEM]; P⬍0.05) and ratio
of heart to body weight (Data Supplement) were slightly
decreased. At 15 months, hemodynamic measurements in
these mice revealed no differences in LV end-diastolic
pressure, end-diastolic volume, end-systolic pressure, endsystolic volume (not shown), and heart rate (Data Supplement). However, ejection fraction showed a trend toward
lower values, and end-diastolic pressure-volume ratios were
decreased in bgn⫺/0 mice (end-diastolic pressure-volume ratio: WT, 0.32⫾0.11 mm Hg/␮L; bgn⫺/0, 0.19⫾0.07 mm Hg/
␮L [mean⫾SEM]; n⫽6; P⬍0.05; Data Supplement). In
addition, 15-month-old bgn⫺/0 mice showed a trend toward
decreased collagen type I mRNA expression but increased
collagen type III and decorin expression as determined by
real-time RT-PCR (data not shown). Morphological analysis
and collagen staining showed no obvious difference between
WT and bgn⫺/0 mice at 15 months (Data Supplement),
suggesting that both the morphology and cardiac ECM
showed no major difference in noninfarcted bgn⫺/0 mice.
Decreased Survival After MI in bgnⴚ/0 Mice
Kaplan-Meier analysis revealed higher mortality in bgn⫺/0
mice after experimental MI. This increase in mortality was
obvious after 7 days and continued until 3 weeks at the end of
the study. At 3 weeks, the survival rates were 60% and 20%
in WT and bgn⫺/0 mice, respectively (Figure 2A). To investigate the cause of death, an additional group of animals
(n⫽8) underwent experimental MI and morphological analysis 7 days after MI. During this 7-day period, the animals
were closely monitored, and mice that died underwent autopsy immediately. Analysis of this group revealed that the
frequency of LV ruptures was ⬇33% in WT and 66% in
bgn⫺/0 mice, explaining the lower survival (Figure 2B). Figure
3 shows that the area at risk and infarct size were not different
between the 2 genotypes as determined by Evans blue and
p-nitro blue tetrazolium 7 days after MI. These data exclude
the possibility that increased area at risk or infarct size is
responsible for the higher frequency of rupture.
B
C
Figure 2. Survival after experimental MI is decreased in bgn⫺/0
mice. A, Mortality after permanent left anterior descending coronary artery occlusion in WT and bgn⫺/0 mice up to 21 days after
MI. Mortality was significantly increased in bgn⫺/0 mice (P⬍0.01,
log-rank-test; WT, n⫽20, 8 nonsurvivors; bgn⫺/0, n⫽33, 27 nonsurvivors). B, bgn⫺/0 mice died more often as a result of ventricular
rupture than their WT littermates within 7 days after MI. C, Heart of
a bgn⫺/0 mouse. Arrows/thread indicate the LV ruptured area.
birefringence analysis, Sirius red staining revealed increased
density of collagen in WT mice as indicated by the bright red
(Figure 4C and 4D; Sirius red birefringence expressed as area
fraction: bgn⫺/0, 3.9⫾1% versus WT, 10.8⫾1.7%; n⫽4;
P⬍0.05). Furthermore, Movat’s pentachrome staining revealed a higher proportion of proteoglycans in scars of bgn⫺/0
mice at 7 days after MI (Figure 4E and 4F), which was
corroborated by Alcian blue staining and subsequent quanti-
A
B
C
D
Matrix Remodeling 7 Days After MI
To investigate the mechanisms that caused LV ruptures, ECM
remodeling was characterized in both groups. Collagen content of infarct scars was similar between genotypes as
determined by Sirius red staining (bright field, Sirius red
expressed as area fraction: WT, 21.9⫾6.7% versus bgn⫺/0,
37.9⫾7%; n⫽4; P⫽1.6). In addition, immunohistochemistry
of collagen type 1 revealed no quantitative differences between groups (data not shown). In polarized light using the
Figure 3. Infarct size 7 days after MI. Area at risk (A; %LV) and
infarct size (B; % LV) were not different between the genotypes and
controls (WT, n⫽8; bgn⫺/0, n⫽8). Representative Evans blue/p-nitro
blue tetrazolium–stained sections of WT (C) or bgn⫺/0 (D) mice.
Downloaded from circ.ahajournals.org at SWETS SUBSCR SVC/25338706 on September 15, 2008
1272
Circulation
March 11, 2008
A
WT
B
bgn-/0
C
WT
D
bgn-/0
E
WT
F
bgn-/0
G
WT
H
bgn-/0
in the infarcted area compared with the noninfarcted area in both
genotypes (Figure 5). The expression of MMP-2 and -13 was
significantly higher in the infarcted area of bgn⫺/0 mice compared with WT. However, MMP-9 was decreased and TIMP-1
and -4 were significantly upregulated in bgn⫺/0 mice after
experimental MI. The differences detected by immunostaining
were largely paralleled by mRNA expression (Data Supplement). In situ zymography indicating gelatinolytic activity
mainly through MMP-2 and -920 revealed no significant difference between genotypes (Figure 5P). Furthermore, the mRNA
expression of procollagen-prolyl-4-hydroxylase, procollagenlysyl-hydroxylase 2, procollagen-lysyl-hydroxylase 1, and tissue
transglutaminase, key enzymes in the regulation of collagen and
ECM crosslinking, was not differentially regulated between the
genotypes after MI (Data Supplement).
Hemodynamic Parameters
Figure 4. ECM remodeling 7 days after MI. A, B, Sirius red
staining, bright field. C, D, Sirius red viewed under polarized
light revealing less densely packed collagen in the infarct scars
of bgn⫺/0 mice. E, F, Movat’s pentachrome staining revealed a
proteoglycan-rich ECM in bgn⫺/0 mice vs WT. G, H, Decorin immunostaining (LF-113) was markedly increased in bgn⫺/0 mice.
Representative stainings of WT (n⫽8) and bgn⫺/0 (n⫽8). Magnification ⫻100.
fication by image analysis (Alcian blue expressed as area
fraction: WT, 73.3⫾4.4% versus bgn⫺/0, 87.5⫾3%; n⫽4;
P⬍0.05). Decorin was upregulated after MI in WT mice as
described previously.19 Notably, post-MI immunostaining of
decorin was much stronger in bgn⫺/0 mice (Figure 4G and
4H). Taken together, the infarct scars of bgn⫺/0 mice were
characterized by loosely packed collagen fibers and a higher
amount of proteoglycans, including decorin.
The expression patterns of MMP-2, -8, -9, and -13 and
TIMP-1, -2, and -4 were analyzed by immunostaining (Figure
5) and by real-time RT-PCR (Data Supplement). MMP-2, -8,
-9, and -13 and TIMP-1, -2, and -4 were strongly upregulated
No difference in cardiac hemodynamic parameters was detected between 16-week-old sham-operated WT and bgn⫺/0
animals (Table), confirming the analysis of WT versus bgn⫺/0
mice under normal conditions as mentioned above. In all
animals, as predicted, MI caused a decrease in ejection
fraction (%) and an increase in LV end-diastolic pressure
(mm Hg), end-diastolic volume (␮L), and end-systolic volume (␮L) (Figure 6). Additional hemodynamic parameters
are presented in the Table, indicating severe impairment of
hemodynamic functions of WT and bgn⫺/0 mice 3 weeks after
MI such as a decrease in LV end-systolic pressure (mm Hg),
dP/dtmax (mm Hg/s), and dP/dtmin (mm Hg/s) and increased
␶ (ms). The most important finding is that bgn⫺/0 mice
demonstrate significantly increased post-MI LV dilatation
with increased LV end-diastolic volume and end-systolic
volume and decreased ejection fraction compared with WT
controls 21 days after MI (Figure 6A through 6D). Furthermore, bgn⫺/0 mice demonstrated reduced stroke volume and
cardiac output despite significantly increased heart rate (bpm)
(Table), suggesting compensatory tachycardia. As an additional sign of cardiac dilatation and failure, LV end-diastolic
Figure 5. Immunohistochemistry of MMPs
and TIMPs. Immunostaining and quantification (area fraction [AF]) in the infarct scar 7
days after MI of MMP-2 (A, B, and I), MMP-8
(C, D, and J), MMP-9 (E, F, and K), and
MMP-13 (G, H, and L). For TIMP-1 (M),
TIMP-2 (N), and TIMP-4 (O), only the results
of the image analysis (AF) are given. In the
scar (MI), MMP-2, MMP-13, TIMP-1, and
TIMP-4 were increased in bgn⫺/0 mice,
whereas MMP-9 was decreased. P, Gelatinolytic activity as determined by in situ
zymography (WT, n⫽5; bgn⫺/0, n⫽4).
*P⬍0.05, ***P⬍0.001 vs WT-MI; #P⬍0.05 vs
respective noninfarcted regions.
Downloaded from circ.ahajournals.org at SWETS SUBSCR SVC/25338706 on September 15, 2008
Westermann et al
Table.
Role of Biglycan During Post-MI Remodeling
1273
Hemodynamic Measurements at 21 Days After MI
Heart
Rate, bpm
Sham
WT (n⫽7)
bgn⫺/0 (n⫽8)
End-Systolic
Pressure, mm Hg
dP/dtmax,
mm Hg/s
dP/dtmin,
mm Hg/s
␶, ms
Stroke
Volume, ␮L
Cardiac Output,
␮L/min
Lung Wet/Dry
Weight
442⫾12
105⫾2
7554⫾818
⫺5231⫾485
13.8⫾0.7
36.6⫾3.4
13160⫾1560
4.48⫾0.1
428⫾8
105⫾5
6890⫾345
⫺5763⫾429
14.0⫾0.7
35.4⫾3.2
13640⫾1728
4.56⫾0.1
4045⫾414†
3240⫾406†
⫺3446⫾314†
⫺2757⫾346†
17.4⫾0.8†
17.0⫾1.4
19.6⫾1.1†
14.1⫾1.2†
8450⫾441†
6723⫾716†
MI
WT (n⫽10)
bgn⫺/0 (n⫽6)
433⫾11
489⫾25*†
73⫾4†
64⫾2†
5.5⫾0.2†
6.4⫾0.2*†
*P⬍0.05, WT vs bgn⫺/0; †P⬍0.05, sham vs MI.
pressure was significantly increased compared with WT
controls. This increase in LV end-diastolic pressure is further
verified by increased ratios of lung wet weight to lung dry
weight after MI compared with WT as an additional indicator
for LV heart failure (Table). These data clearly indicate that
bgn⫺/0 mice develop impaired hemodynamic function in
response to ischemic injury. In addition, a subset of animals
was investigated with MRI. The longitudinal MRI images
(Figure 6E and 6F) showed stronger LV dilation and are
consistent with the hemodynamic data. Calculation of LV
end-diastolic wall stress revealed increased wall stress in
bgn⫺/0 mice after MI compared with WT (14.84⫾1.2⫻103
dynes · mm⫺2 [P⬍0.05 compared with WT MI] versus
9.09⫾1.8⫻103 dynes · mm⫺2).
Matrix Remodeling 21 Days After MI
Increased frequency of cardiac ruptures and decreased hemodynamic function of bgn⫺/0 mice after MI and less tightly
packed collagen at 7 days (Figure 4) point toward perturbed
ECM remodeling in the infarct scar of bgn⫺/0 mice. Therefore,
ECM remodeling also was characterized at 21 days after MI.
MI induced a marked upregulation of collagen type 1 and 3
mRNA as determined by real-time RT-PCR, which was
significantly stronger in the bgn⫺/0 mice (Figure 7A). Decorin
mRNA showed a trend toward upregulation in WT versus
bgn⫺/0 mice 21 days after MI (Figure 7B). The overall
collagen content was similar as evidenced by Sirius red
staining (Figure 7C and 7D; total Sirius red expressed as area
fraction: WT, 41⫾3.9% versus bgn⫺/0, 43.8⫾11.7%; n⫽4;
P⬍0.05) and collagen type 1 immunostaining (not shown). A
striking difference appeared with respect to collagen organization (Figure 7C and 7D). Under polarized light, Sirius red
staining revealed that the collagen present in the scar of bgn⫺/0
mice was characterized by a lower percentage of tightly
packed collagen (Figure 7E and 7F; Sirius red birefringence
expressed as area fraction: WT, 11.1⫾0.3% versus bgn⫺/0,
5.9⫾1.4%; n⫽4; P⬍0.05). Immunohistochemistry showed
still increased decorin accumulation (not shown); however,
the difference was not as pronounced as detected at 7 days
(Figure 4). Analysis of the collagen matrix by electron
microscopy revealed tight and parallel packing of collagen
fibrils in the scars of WT mice. In contrast, collagen fibril
organization was perturbed in bgn⫺/0 mice (Figure 7H). The
tensile force to rupture LV rings ex vivo was determined to
investigate whether the differences in matrix composition and
organization correlate with changes of tensile strength. Under
control conditions without MI, no difference was evident
between LV rings from WT and bgn⫺/0 mice that ruptured at
84 and 90 mN, respectively (Figure 7I). However, 21 days
after MI, the force to rupture was dramatically decreased in
LV rings derived from bgn⫺/0 mice (69 mN) compared with
WT mice (146 mN).
Biglycan Accumulation in Human MI
Figure 6. Hemodynamic parameters 21 days after MI. A through
D, Hemodynamic measurements in WT vs bgn⫺/0 under control
conditions (sham) and 21 days after MI. After MI, bgn⫺/0 mice
were characterized by decreased ejection fraction (EF; A) and
increased LV end-diastolic pressure (Ped; B), LV end-diastolic
volume (Ved; C), and LV end-systolic volume (Ves; D) (sham
WT, n⫽7; sham bgn⫺/0, n⫽8; MI WT, n⫽10; MI bgn⫺/0; n⫽6).
Additional hemodynamic parameters are given in the Table.
E and F, MRI representative longitudinal sections showing
enhanced LV dilatation in bgn⫺/0 mice (F; n⫽7) vs WT (E; n⫽8)
at 21 days after MI. Arrows point toward an area of marked
ventricular wall thinning. *P⬍0.05, ***P⬍0.001, #P⬍0.05 vs
respective noninfarcted controls.
Tissue sections derived from patients who died of ischemic
MI presenting as either an acute, primary event or reoccurring
acute infarction on the basis of old healed MI were examined.
Figure 8A through 8D represents an area from an acute MI
characterized by acute myocyte necrosis with eosinophilic
cytoplasm and without visible cardiomyocyte nuclei. Hardly
any immunostaining for biglycan nor Sirius red staining of
collagen was detected under these conditions. In contrast,
Figure 8E through 8H depicts a healed infarction with fibrotic
regions and some remaining vital cardiomyocytes. In these
healed infarcts, strong accumulation of both biglycan and
densely packed collagen as shown by birefringence analysis
Downloaded from circ.ahajournals.org at SWETS SUBSCR SVC/25338706 on September 15, 2008
1274
Circulation
March 11, 2008
Figure 7. ECM remodeling 21 days after MI. Quantitative realtime RT-PCR of collagen (coll) type 1 (A) and decorin (dcn) (B)
mRNA expressed as fold of sham-operated WT 21 days after
MI. C and D Sirius red staining and morphology of the infarct
scar in WT mice was similar to that in bgn⫺/0 mice. Magnification ⫻40. E and F, Sirius red staining viewed under polarized
light. Compared with WT (E), less densely packed collagen was
detected in the infarct scar of bgn⫺/0 mice (F). A through F,
Sham WT, n⫽7; sham bgn⫺/0, n⫽8; MI WT, n⫽10; MI bgn⫺/0,
n⫽6. Magnification ⫻40. G and H, Ultrastructure of collagen
fibrils derived from the infarct scar as visualized by electron
microscopy. Infarcts of WT mice showed tightly packed collagen fibrils aligned mostly in parallel (G). In contrast, organization
of collagen fibrils in the infarct scar of bgn⫺/0 mice (H) was perturbed. Magnification ⫻12 000. Representative images of 4 for
each genotype 21 days after MI. I, Quantitative analysis of collagen fibril arrangement. Collagen confluence indicates the percentage of tight, parallel arranged collagen fibrils viewed by
electron microscopy. J, Tension to rupture (mN) measured ex
vivo on 1-mm rings of the LV of noninfarcted mice (control: WT,
n⫽5; bgn⫺/0, n⫽8) and 21 days after MI (WT, n⫽5; bgn⫺/0, n⫽7).
***P⬍0.001, #P⬍0.05 vs respective noninfarcted controls.
of Sirius red staining occurs (Figure 8F and 8H). Notably,
biglycan colocalizes with tightly packed collagen. Thus,
biglycan accumulates during healing and remodeling and
associates with collagen after MI in human hearts.
Discussion
The key results of this study are increased mortality, high
frequency of cardiac ruptures, and aggravated cardiac failure
in bgn⫺/0 mice after experimental MI, which were likely the
result of perturbed collagen remodeling in the infarct scar.
Seven days after experimental MI, thinning of the infarct area
and formation of a collagen matrix occur in mice. After 14 days, a
A
B
C
D
E
F
G
H
Figure 8. Immunohistochemistry of biglycan in human MI. A
through D, Consecutive sections of acute, human MI characterized by myocyte necrosis. A, Hematoxylin and eosin; B, biglycan (LF51); C, Sirius red; D, Sirius red polarized. Arrows indicate
the same coronary artery for orientation. E through H, Consecutive sections of healed MI with collagenous scar tissue showing
strong accumulation of biglycan (F) in the fibrotic interstitium in
colocalization with densely packed collagen (G and H). Magnification ⫻40. The presented cases are representative of each of
4 cases of acute and healed infarct scars. *Areas of colocalization between biglycan and densely packed collagen as evidenced by polarized light (H).
mature collagen-rich scar is established, and at this time point,
collagen levels peak.2,21 Collagen type 1 is the major collagen of the
myocardial ECM. Much attention has been focused on regulators of
collagen expression and the breakdown of collagen matrixes during
postinfarct remodeling. However, little is known about the role of
small ECM molecules that govern collagen fibril formation. Among
these small ECM molecules are the small leucine-rich proteoglycans such as biglycan and decorin. Small leucine-rich proteoglycans
can be regarded as accessory ECM molecules at the interface of
matrix assembly and signaling, which might have potential for
therapeutic interventions. The aim of the present study was to
investigate whether biglycan affects cardiac phenotype in healthy
animals and, importantly, after experimental MI.
No differences were detected with respect to the cardiac
phenotype at 6 months in WT versus bgn⫺/0 mice. At 15 months,
the relaxation parameter ␶ and the end-diastolic pressure-volume
ratio were decreased in bgn⫺/0 mice. Although other parameters22
Downloaded from circ.ahajournals.org at SWETS SUBSCR SVC/25338706 on September 15, 2008
Westermann et al
may influence the end-diastolic pressure-volume ratio, this
finding suggests that the intrinsic myocardial stiffness is reduced
in aged bgn⫺/0 mice. Taken together, these findings suggest that
during physiological conditions bgn deficiency has only minor
effects on LV function, which starts to induce hemodynamic
changes only during aging.
After experimental MI, biglycan expression was strongly
upregulated in WT mice, as shown previously in rats.10,11
Interestingly, the upregulation was transient, peaking at 7
days, at the same time that collagen content also is highest.2
Using immunohistochemistry, we found biglycan in the
infarct scar and border zone of the infarct at 7 and 14 days
after MI, in line with a role of biglycan in collagen remodeling. During the whole experimental period, bgn⫺/0 mice
showed higher mortality after MI, which was due in large part
to LV ruptures. In addition, spontaneous death without
ruptures occurred more often in bgn⫺/0 mice. This might be a
consequence of aggravated cardiac failure and LV dilatation
as indicated by increased LV end-diastolic volume, end-diastolic pressure, and ratios of wet lung weight to dry lung
weight in bgn⫺/0 mice after MI.
bgn⫺/0 mice and WT mice showed no differences in the
area at risk and infarct size, which excludes the possibility
that decreased survival of bgn⫺/0 mice after MI is a consequence
of increased ischemia or necrosis. Birefringence analysis of
Sirius red–stained sections revealed that the packing of collagen
fibrils was disturbed in bgn⫺/0 mice 7 and 21 days after MI. This
was confirmed by analysis of the ultrastructure of collagen in
the infarct scar, showing that the array of collagen was
heavily perturbed in bgn⫺/0 mice. The quantity of collagen
accumulating in the scar was not decreased as analyzed by
immunohistochemistry of collagen type 1 and Sirius red
staining. In contrast, mRNA expression of collagen type 1
and collagen type 3 was increased even in bgn⫺/0 mice, which
might represent a compensatory mechanism.
Increased MMP activity is known to contribute to ventricular dilatation after MI, and expression of MMPs is highly
increased in the infarcted area.23,24 The upregulation of
MMP-2 and MMP-13 was opposed by upregulation of
TIMP-1 and TIMP-4 and downregulation of MMP-9. MMP-9
is a key player in adverse ventricular remodeling after MI
because MMP-9 – dependent collagen degradation is associated with adverse LV myocardial remodeling in animal
models and patients.25,26 An additional mechanism of how the
absence of biglycan could increase collagen cleavage is
conceivable. It was recently demonstrated in vitro that the
noncovalent binding of decorin and biglycan to fibrillar
collagen type 1 inhibits the cleavage of collagen type 1 by
collagenase-1 (MMP-1) and collagenase-3 (MMP-13).27 The
interaction of biglycan core protein with collagen fibrils
renders collagen type 1 fibrils less accessible to MMPmediated cleavage. Of note, biglycan inhibited collagen
breakdown without changing the activation or expression of
MMPs, which would obscure this mechanism even after
analysis of the MMP/TIMP system as we performed in the
present study. Taken together, it is conceivable that changes in
the MMP/TIMP levels and/or increased susceptibility to breakdown in the absence of biglycan contribute to the cardiac
phenotype of bgn⫺/0 mice described in the present study.
Role of Biglycan During Post-MI Remodeling
1275
Taken together, in bgn⫺/0 mice, disturbed collagen matrix
assembly might result in decreased mechanical strength of
myocardial scar tissue after MI. Indeed, ex vivo measurements
revealed severely reduced tensile strength of LV rings derived
from bgn⫺/0 mice compared with WT mice. The reduced tensile
strength of LV and the increase in LV end-diastolic wall stress
might explain the frequent ruptures in bgn⫺/0 mice.
Interestingly, the dcn⫺/⫺ mice also have a strong cardiac
phenotype after experimental MI.19 dcn⫺/⫺ mice developed
impaired LV function after experimental MI that was attributed
to abnormal collagen fibril formation. However, ventricular
ruptures did not occur in dcn⫺/⫺ mice.19 Therefore, from the
present study, it appears that the cardiac phenotype is even
stronger in bgn⫺/0 mice compared with dcn⫺/⫺ mice. Furthermore, upregulation of decorin as observed in bgn⫺/0 mice 7 days
after MI was not sufficient to compensate for the loss of biglycan
as it does with respect to the osteoporosis phenotype.28
Analysis of scars of MI from humans revealed a strong
accumulation of biglycan in close association with collagen
in fibrotic areas of healed scars, whereas biglycan was not
detected after acute human MI. It is therefore possible that the
present findings are of clinical relevance and that biglycan
might participate in the evolution of a collagenous ECM
during healing of human MI. However, the present data do
not allow a conclusion to be drawn as to whether biglycan
deposition is characteristic for postinfarct remodeling or is
associated with fibrotic remodeling and hypertrophy in general. LV wall rupture as observed in bgn⫺/0 mice also is a
complication of acute MI in patients that is responsible for
⬇20% of infarct-related deaths.29,30 Therefore, it might be of
interest to investigate the patterns of biglycan expression in
human MI and the relation to cardiac rupture in future studies.
From the present study, it is obvious that the complete absence
of biglycan is deleterious and prevents the establishment of a
stable infarct scar and hemodynamic adaptation, suggesting that
biglycan is required for proper infarct healing. However, future
studies should address how much biglycan is needed for proper
scar formation and whether too much biglycan might also have
adverse effects such as enhanced fibrosis.
Sources of Funding
This study was in part supported by Deutsche Forschungsgemeinschaft SFB612, SFB TR19, GRK1089, and GRK 754; the Interdisciplinary Center for Clinical Research, Münster; and the Excellence
Cluster Cardio-Pulmonary System.
Disclosures
None.
References
1. Welch MP, Odland GF, Clark RA. Temporal relationships of F-actin
bundle formation, collagen and fibronectin matrix assembly, and
fibronectin receptor expression to wound contraction. J Cell Biol. 1990;
110:133–145.
2. Dobaczewski M, Bujak M, Zymek P, Ren G, Entman ML, Frangogiannis
NG. Extracellular matrix remodeling in canine and mouse myocardial
infarcts. Cell Tissue Res. 2006;324:475– 488.
3. Gray AJ, Bishop JE, Reeves JT, Laurent GJ. A alpha and B beta chains
of fibrinogen stimulate proliferation of human fibroblasts. J Cell Sci.
1993;104(pt 2):409 – 413.
4. Danielson KG, Baribault H, Holmes DF, Graham H, Kadler KE, Iozzo
RV. Targeted disruption of decorin leads to abnormal collagen fibril
morphology and skin fragility. J Cell Biol. 1997;136:729 –743.
Downloaded from circ.ahajournals.org at SWETS SUBSCR SVC/25338706 on September 15, 2008
1276
Circulation
March 11, 2008
5. Corsi A, Xu T, Chen XD, Boyde A, Liang J, Mankani M, Sommer B, Iozzo RV,
Eichstetter I, Robey PG, Bianco P, Young MF. Phenotypic effects of biglycan
deficiency are linked to collagen fibril abnormalities, are synergized by decorin
deficiency, and mimic Ehlers-Danlos-like changes in bone and other connective
tissues. J Bone Miner Res. 2002;17:1180–1189.
6. Douglas T, Heinemann S, Bierbaum S, Scharnweber D, Worch H. Fibrillogenesis of collagen types I, II, and III with small leucine-rich proteoglycans decorin and biglycan. Biomacromolecules. 2006;7:2388 –2393.
7. Wiberg C, Hedbom E, Khairullina A, Lamande SR, Oldberg A, Timpl R,
Morgelin M, Heinegard D. Biglycan and decorin bind close to the n-terminal
region of the collagen VI triple helix. J Biol Chem. 2001;276:18947–18952.
8. Vogel KG, Paulsson M, Heinegard D. Specific inhibition of type I and
type II collagen fibrillogenesis by the small proteoglycan of tendon.
Biochem J. 1984;223:587–597.
9. Schonherr E, Witsch-Prehm P, Harrach B, Robenek H, Rauterberg J,
Kresse H. Interaction of biglycan with type I collagen. J Biol Chem.
1995;270:2776 –2783.
10. Doi M, Kusachi S, Murakami T, Ninomiya Y, Murakami M, Nakahama M,
Takeda K, Komatsubara I, Naito I, Tsuji T. Time-dependent changes of
decorin in the infarct zone after experimentally induced myocardial infarction
in rats: comparison with biglycan. Pathol Res Pract. 2000;196:23–33.
11. Yamamoto K, Kusachi S, Ninomiya Y, Murakami M, Doi M, Takeda K,
Shinji T, Higashi T, Koide N, Tsuji T. Increase in the expression of
biglycan mRNA expression co-localized closely with that of type I
collagen mRNA in the infarct zone after experimentally-induced myocardial infarction in rats. J Mol Cell Cardiol. 1998;30:1749 –1756.
12. Xu T, Bianco P, Fisher LW, Longenecker G, Smith E, Goldstein S,
Bonadio J, Boskey A, Heegaard AM, Sommer B, Satomura K,
Dominguez P, Zhao C, Kulkarni AB, Robey PG, Young MF. Targeted
disruption of the biglycan gene leads to an osteoporosis-like phenotype in
mice. Nat Genet. 1998;20:78 – 82.
13. Ameye L, Aria D, Jepsen K, Oldberg A, Xu T, Young MF. Abnormal
collagen fibrils in tendons of biglycan/fibromodulin-deficient mice lead
to gait impairment, ectopic ossification, and osteoarthritis. FASEB J.
2002;16:673– 680.
14. Hildebrand A, Romaris M, Rasmussen LM, Heinegard D, Twardzik DR,
Border WA, Ruoslahti E. Interaction of the small interstitial proteoglycans biglycan, decorin and fibromodulin with transforming growth
factor beta. Biochem J. 1994;302:527–534.
15. Schaefer L, Beck KF, Raslik I, Walpen S, Mihalik D, Micegova M,
Macakova K, Schonherr E, Seidler DG, Varga G, Schaefer RM, Kresse
H, Pfeilschifter J. Biglycan, a nitric oxide-regulated gene, affects
adhesion, growth, and survival of mesangial cells. J Biol Chem. 2003;
278:26227–26237.
16. Schaefer L, Babelova A, Kiss E, Hausser HJ, Baliova M, Krzyzankova
M, Marsche G, Young MF, Mihalik D, Gotte M, Malle E, Schaefer RM,
Grone HJ. The matrix component biglycan is proinflammatory and
signals through Toll-like receptors 4 and 2 in macrophages. J Clin Invest.
2005;115:2223–2233.
17. Petzelbauer P, Zacharowski PA, Miyazaki Y, Friedl P, Wickenhauser G,
Castellino FJ, Groger M, Wolff K, Zacharowski K. The fibrin-derived
peptide Bbeta15-42 protects the myocardium against ischemiareperfusion injury. Nat Med. 2005;11:298 –304.
18. Westermann D, Knollmann BC, Steendijk P, Rutschow S, Riad A,
Pauschinger M, Potter JD, Schultheiss HP, Tschope C. Diltiazem
treatment prevents diastolic heart failure in mice with familial hypertrophic cardiomyopathy. Eur J Heart Fail. 2006;8:115–121.
19. Weis SM, Zimmerman SD, Shah M, Covell JW, Omens JH, Ross J Jr, Dalton N,
Jones Y, Reed CC, Iozzo RV, McCulloch AD. A role for decorin in the
remodeling of myocardial infarction. Matrix Biol. 2005;24:313–324.
20. Frederiks WM, Mook OR. Metabolic mapping of proteinase activity with
emphasis on in situ zymography of gelatinases: review and protocols.
J Histochem Cytochem. 2004;52:711–722.
21. Cleutjens JP, Verluyten MJ, Smiths JF, Daemen MJ. Collagen
remodeling after myocardial infarction in the rat heart. Am J Pathol.
1995;147:325–338.
22. Burkhoff D, Mirsky I, Suga H. Assessment of systolic and diastolic
ventricular properties via pressure-volume analysis: a guide for clinical,
translational, and basic researchers. Am J Physiol Heart Circ Physiol.
2005;289:H501–H512.
23. Mukherjee R, Brinsa TA, Dowdy KB, Scott AA, Baskin JM, Deschamps
AM, Lowry AS, Escobar GP, Lucas DG, Yarbrough WM, Zile MR,
Spinale FG. Myocardial infarct expansion and matrix metalloproteinase
inhibition. Circulation. 2003;107:618 – 625.
24. Peterson JT, Li H, Dillon L, Bryant JW. Evolution of matrix metalloprotease and tissue inhibitor expression during heart failure progression in
the infarcted rat. Cardiovasc Res. 2000;46:307–315.
25. Ducharme A, Frantz S, Aikawa M, Rabkin E, Lindsey M, Rohde LE,
Schoen FJ, Kelly RA, Werb Z, Libby P, Lee RT. Targeted deletion of
matrix metalloproteinase-9 attenuates left ventricular enlargement and
collagen accumulation after experimental myocardial infarction. J Clin
Invest. 2000;106:55– 62.
26. Webb CS, Bonnema DD, Ahmed SH, Leonardi AH, McClure CD, Clark
LL, Stroud RE, Corn WC, Finklea L, Zile MR, Spinale FG. Specific
temporal profile of matrix metalloproteinase release occurs in patients
after myocardial infarction: relation to left ventricular remodeling.
Circulation. 2006;114:1020 –1027.
27. Geng Y, McQuillan D, Roughley PJ. SLRP interaction can protect collagen
fibrils from cleavage by collagenases. Matrix Biol. 2006;25:484 – 491.
28. Young MF, Bi Y, Ameye L, Chen XD. Biglycan knockout mice: new
models for musculoskeletal diseases. Glycoconj J. 2002;19:257–262.
29. Moreno R, Lopez de Sa E, Lopez-Sendon JL, Garcia E, Soriano J,
Abeytua M, Elizaga J, Botas J, Rubio R, Moreno M, Garcia-Fernandez
MA, Delcan JL. Frequency of left ventricular free-wall rupture in patients
with acute myocardial infarction treated with primary angioplasty.
Am J Cardiol. 2000;85:757–760.
30. Hutchins KD, Skurnick J, Lavenhar M, Natarajan GA. Cardiac rupture in
acute myocardial infarction: a reassessment. Am J Forensic Med Pathol.
2002;23:78 – 82.
CLINICAL PERSPECTIVE
Known mechanisms leading to heart failure after myocardial infarction include subsequent cardiac dilatation and infarct
thinning, also called infarct expansion, which may lead to the development of aneurysm and left ventricle rupture.
Therefore, adaptive changes in cardiac remodeling are essential so that the ventricle forms a stable scar rather than
rupturing after myocardial infarction. Collagenous networks elaborated by cardiac fibroblasts provide the mechanical
support for the scar tissue. It seems important to consider that the stability and 3-dimensional arrangement of cardiac
collagen matrixes are controlled by accessory molecules such as biglycan. Biglycan is a proteoglycan that binds to the d
bands of collagen type 1 fibrils and controls fibrillogenesis and proteolysis of collagen by matrix metalloproteinases. The
present study reveals that after myocardial infarction, the scar tissue of biglycan-deficient mice was characterized by
perturbed collagen fibril arrangement and that ventricular ruptures occur more frequently compared with wild-type mice.
Furthermore, the surviving biglycan-deficient mice showed impaired hemodynamic function with subsequent cardiac dilatation,
suggesting that in the absence of biglycan, adverse ventricular remodeling is aggravated. Therefore, biglycan appears to play a
critical role in the formation of stable infarct scars and the preservation of hemodynamic function after myocardial infarction.
Because biglycan accumulates in the scar tissue of human ischemic infarctions, it is possible that the present findings extend to
human postinfarct remodeling. In the future, it will be of interest to determine whether differences in cardiac biglycan expression
may be modulated by pharmacological treatment and whether this may change clinical outcome.
Downloaded from circ.ahajournals.org at SWETS SUBSCR SVC/25338706 on September 15, 2008