Transgenic Overexpression of Platelet-Derived
Growth Factor-C in the Mouse Heart Induces
Cardiac Fibrosis, Hypertrophy, and Dilated
Cardiomyopathy
Annica Ponte´n,* Xuri Li,* Peter Thore´n,†
Karin Aase,* Tobias Sjo¨blom,‡ Arne O¨ stman,‡ and
Ulf Eriksson*
From the Ludwig Institute for Cancer Research,* Stockholm
Branch, Stockholm; the Ludwig Institute for Cancer Research,‡
Uppsala Branch, Uppsala; and the Department of Physiology and
Pharmacology,† Karolinska Institutet, Stockholm, Sweden
The platelet-derived growth factors are implicated in
development of fibrotic reactions and disease in sev-
eral organs. We have overexpressed platelet-derived
growth factor-C in the heart using the
-myosin heavy
chain promoter and created a transgenic mouse that
exhibits cardiac fibrosis followed by hypertrophy
with sex-dependent phenotypes. The transgenic mice
developed several pathological changes including car-
diac fibroblast proliferation and deposition of colla-
gen, hypertrophy, vascular defects, and the presence
of Anitschkow cells in the adult myocardium. Male
mice developed a hypertrophic phenotype, whereas
female mice were more severely affected and devel-
oped dilated cardiomyopathy, leading to heart failure
and sudden death. The vascular defects initially in-
cluded dilation of microvessels and vascular leakage.
Subsequently, a marked loss of microvessels, forma-
tion of large vascular sac-like structures, and an in-
creased density of smooth muscle-coated vessels were
observed in the myocardium. In part, the observed
vascular changes may be because of an up-regulation
of vascular endothelial growth factor in cardiac fibro-
blasts of the transgenic hearts. This unique animal
model reveals that a potent mitogen for cardiac fibro-
blasts result in an expansion of the interstitium that
induce a secondary sex-dependent hypertrophic re-
sponse in the cardiomyocytes. (Am J Pathol 2003,
163:673–682)
Myocardial disease is the most common cause of death
in humans. The reasons underlying cardiac malfunction
are diverse but often relate to thromboembolic events,
arrhythmias, or cardiomyopathies. The cardiomyopathies
can be classified in several subgroups ranging from hy-
pertrophic cardiomyopathy characterized by left ventric-
ular hypertrophy to dilated cardiomyopathy with a thin left
ventricle wall and limited contractile function. Cardiac
hypertrophy is an adaptive response to compensate for a
decreased cardiac output and may occur as a result of a
variety of stimuli including myocardial infarction, hyper-
tension, endocrine disorders, and mutations in proteins
affecting contractile function. The cardiomyocytes re-
spond to the decreased cardiac output by increasing in
size by cellular hypertrophy rather than undergoing cell
division. Although the hypertrophic response initially will
increase the cardiac output it can ultimately lead to di-
lated cardiomyopathy and heart failure. As part of the
hypertrophic response, interstitial cells, ie, cardiac fibro-
blasts, become activated and the expansion of the inter-
stitium often results in a replacement fibrotic reaction,
because the hypertrophic cardiomyocytes are unable to
regenerate damaged tissue.1–5
Several types of growth factors are able to induce
cardiac hypertrophy and fibrosis, among them tumor ne-
crosis factor (TNF)-
and transforming growth factor-.6,7
Platelet-derived growth factors (PDGFs) are potent mito-
gens for many cell types of mesenchymal origin, includ-
ing cardiac fibroblasts.8,9 The PDGF family consists of
four different members; PDGF-A, PDGF-B, PDGF-C, and
PDGF-D.9,10 The PDGF isoforms bind to and activate two
receptor tyrosine kinases, PDGFR-
and PDGFR-, that
are differentially expressed in target tissues. PDGF-A, -B,
and -C chains bind to PDGFR-
with high affinity,
whereas only PDGF-B and -D chains bind PDGFR-. The
two novel PDGFs, PDGF-C and PDGF-D, differ from
PDGF-A and -B in that they are secreted as latent factors
that require proteolytic activation by removal of the N-
terminal CUB domain for activity.11,12
In cardiovascular medicine, transgenic animals with
induced myocardial diseases mimicking those found in
humans are of great interest, and may provide unique
Supported by grants from the Swedish Research Council, the Novo
Nordisk Foundation, and the Karolinska Institutet.
Accepted for publication May 5, 2003.
Present address of K. A.: Department of Oncology, Karolinska Hospital,
CCK, Karolinska Institutet, S-171 76 Stockholm, Sweden.
Address reprint requests to Dr. Ulf Eriksson, Ludwig Institute for Cancer
Research, Stockholm Branch, Box 240, S-171 77 Stockholm, Sweden.
E-mail: ueri@licr.ki.se.
American Journal of Pathology, Vol. 163, No. 2, August 2003
Copyright © American Society for Investigative Pathology
673

opportunities to explore diverse aspects of the dis-
eases, including drug development. Despite the well
known roles of the PDGFs as potent mitogens for fibro-
blasts, potentially being directly involved in fibrotic
reactions in a variety of organs,9 only a few attempts to
overexpress PDGFs in vivo using transgenic tech-
niques have been performed.13–16 In this work we have
targeted transgenic overexpression of PDGF-C to the
heart using the
-myosin heavy chain (
-MHC) pro-
moter and have created a mouse model in which a
potent fibroblast mitogen is overexpressed in the heart,
leading to a progressive fibrosis and hypertrophy. Fe-
male mice were shown to develop a lethal dilated
cardiomyopathy, whereas the males develop a more
progressive hypertrophic phenotype. As a control,
transgenic mice overexpressing the CUB domain only
were also generated. These mice appeared primarily
normal suggesting that the activated PDGF-C core
domain is responsible for the pathological changes
seen in the mice expressing full-length PDGF-C.
Materials and Methods
Generation of Transgenic Mice
Transgenic mice overexpressing full-length PDGF-C in
the heart were generated as earlier described using
CBA/C57BL/6 mice.11 Transgenic mice were maintained
by backcrossing onto C57BL/6 animals the first genera-
tion and then by brother-sister matings. The transgenic
expression and the observed phenotypes were stable for
at least five generations. All experiments were performed
using heterozygous and wild-type animals.
To generate transgenic mice overexpressing the CUB
domain of PDGF-C, mouse PDGF-C cDNA was subjected
to polymerase chain reaction (PCR) mutagenesis using
the following primers: 5
CGC GGT CGA CGC CCC AGT
CAG CCA A (forward) including a flanking SalI site and 5

CGC GAA GCT TTT ACA AGT CTT CTT CAG AAA TGA
GCT TTT GTT CTG TGA CTT GTG GCA (reverse) includ-
ing the human c-Myc epitope and a flanking HindIII site.
The PCR generated a 572-bp fragment that was digested
with SalI/HindIII and cloned into the
-MHC vector17
(kindly provided by K. Alitalo, Helsinki University, Hel-
sinki, Finland). The linearized and purified transgene
fragment was injected into male pronuclei of fertilized
mouse oocytes (Mouse Camp, Karolinska Institutet,
Stockholm, Sweden). Tail lysate prepared from the result-
ing pups was screened for the presence of the transgene
by PCR using the CUB domain-specific primer used in
the cloning 5
CGC GGT CGA CGC CCC AGT CAG CCA
A (forward) and the c-Myc/CUB-specific primer 5
ATG
AGC TTT TGT TCT GTG AC (reverse). A 542-bp fragment
was amplified in the transgenic founders. The animal
experiments in this study were approved by the local
committee of the Swedish National Board for Laboratory
Animals (CFN).
Histology and Immunohistochemistry
Wild-type and heterozygous littermates of different ages
were sacrificed and weighed. The hearts were dissected
and weighed or frozen on dry ice for protein and RNA
extraction, or fixed in 4% paraformaldehyde in phos-
phate-buffered saline overnight at  4°C and processed
for paraffin-embedding using standard protocols. A few
hearts were collected from females that died from heart
failure. Microphotographs of whole hearts were taken
with a Nikon Digital Camera finePix S1 pro. The paraffin-
embedded hearts were sectioned (6 m) and the sec-
tions stained with hematoxylin and eosin (H&E), or with
Masson’s trichrome, using standard procedures. Visual-
ization of myofibrils was performed using a phosphotung-
stic acid hematoxylin staining kit (Bio Optica, Milano,
Italy). Immunolocalization of PDGF-C was performed us-
ing TSA indirect (NEN Life Science, Boston, MA, USA)
with an affinity-purified rabbit Ig fraction to human
PDGF-C core domain protein (1 g Ig/ml).11 Endothelial-
specific staining was performed as above using rat anti-
mouse CD31/platelet-endothelial cell adhesion molecule
(PECAM) antibody (Pharmingen, San Diego, CA, USA) as
earlier described.18
-Smooth muscle actin (SMA) stain-
ing was performed using a monoclonal anti-human

-SMA antibody 1A4 (DAKO, Glostrup, Denmark). Anti-
gen retrieval was obtained by heating the tissue slides in
0.01 mol/L of citrate buffer, pH 6.0, at 94°C for 20 min-
utes. Vascular endothelial growth factor (VEGF) staining
was performed using a rabbit polyclonal anti-VEGF anti-
body A 20 (Santa Cruz, Biotechnology, Santa Cruz, CA,
USA). Antigen retrieval was obtained as above. Detection
of proliferating cells was performed using a mouse mono-
clonal anti-proliferating cell nuclear antigen antibody
(Chemicon, Temecula, CA, USA). For SMA, VEGF, and
proliferating cell nuclear antigen stainings, Elite ABC
Vectastain (Vector Laboratories, Burlingame, CA, USA)
was used.
RNase Protection Assay
Total RNA extracted from hearts was analyzed by RNase
protection assay using [32P]UTP-labeled riboprobes as
earlier described.19 A -actin probe was used as an
internal control. The results were visualized and quanti-
fied using a phosphorimager (Fuji Bas 1500).
Immunoblot Analysis
Proteins were extracted from heart tissue in lysis buffer
including protease inhibitor mix (Complete; Roche,
Mannheim, Germany) according to standard protocols.
Aliquots were subjected to sodium dodecyl sulfate-poly-
acrylamide gel electrophoresis (SDS-PAGE) (12%) under
reducing conditions. Immunoblotting was performed us-
ing the affinity-purified rabbit Ig fraction to human
PDGF-C core domain PDGF-C (see above, 1 g Ig/ml) or
a rabbit polyclonal Ig against human c-Myc for detection
of transgenic CUB domain (1/500, Santa Cruz). Bound
antibodies were visualized using ECL (Amersham Bio-
674 Ponte´n et al
AJP August 2003, Vol. 163, No. 2