Coexpression of VEGF and angiopoietin-1 promotes
angiogenesis and cardiomyocyte proliferation reduces
apoptosis in porcine myocardial infarction (MI) heart
Zhengxian Taoa, Bo Chena, Xiao Tana, Yingming Zhaoa, Liansheng Wanga, Tiebing Zhua, Kejiang Caoa, Zhijian Yanga,1,
Yuet Wai Kanb,1, and Hua Suc
a
Department of Cardiology, The First Afﬁliated Hospital of Nanjing Medical University, Nanjing 210029, People’s Republic of China; and bDepartment of
Medicine and cCenter for Cerebrovascular Research, Department of Anesthesia and Perioperative, University of California, San Francisco, CA 94143-0793
VEGF and angiopoietin-1 (Ang1) are two major angiogenic factors
being investigated for the treatment of myocardial infarction (MI).
Targeting VEGF and Ang1 expression in the ischemic myocardium
can increase their local therapeutic effects and reduce possible
adverse effects. Adeno-associated viral vectors (AAVs) expressing
cardiac-speciﬁc and hypoxia-inducible VEGF [AAV-myosin light
chain-2v (MLC)VEGF] and Ang1 (AAV-MLCAng1) were coinjected
(VEGF/Ang1 group) into six different sites of the porcine myocardium at the peri-infarct zone immediately after ligating the left descending coronary artery. An identical dose of AAV-Cytomegalovirus
(CMV)LacZ or saline was injected into control animals. AAV genomes
were detected in the liver in addition to the heart. RT-PCR, Western
blotting, and ELISA analyses showed that VEGF and Ang1 were predominantly expressed in the myocardium in the infarct core and
border of the infarct heart. Gated single-photon emission computed
tomography analyses showed that the VEGF/Ang1 group had better
cardiac function and myocardial perfusion at 8 wk than at 2 wk after
vector injection. Compared with the saline and LacZ controls, the
VEGF/Ang1 group expressed higher phosphorylated Akt and BclxL, less Caspase-3 and Bad, and had higher vascular density, more
proliferating cardiomyocytes, and less apoptotic cells in the infarct
and peri-infarct zones. Thus, cardiac-speciﬁc and hypoxia-induced
coexpression of VEGF and Ang1 improves the perfusion and function
of porcine MI heart through the induction of angiogenesis and cardiomyocyte proliferation, activation of prosurvival pathways, and
reduction of cell apoptosis.
coronary disease
| gene therapy
T
ransfer of angiogenic genes to the ischemic myocardium is
a promising approach under development for the treatment of
myocardial infarction (MI). Animal and clinical data indicated
that angiogenic factors such as ﬁbroblast growth factor (1–3),
VEGF (4–7), angiopoietins (8), and hepatocyte growth factor (9,
10) can induce angiogenesis in ischemic myocardium and improve
cardiac function of MI heart. VEGF and angiopoietin-1 (Ang1)
can also activate a prosurvival signaling pathway and attract stem
cell homing (11). Obstacles that prevent clinical application of
angiogenic gene therapy include short duration of action of the
factors, inadequate transduction or expression of the genes, immune reactions against the delivery vector, and unwanted angiogenesis caused by the uncontrolled expression of angiogenic factors. To overcome these obstacles, we used (i) adeno-associated
viral vector (AAV) serotype 1 to reduce the immune reaction and
increase the duration of transgene expression and (ii) cardiac
myosin light chain-2v (MLC) promoter and the hypoxia response
element (HRE) to drive VEGF and Ang1 expression in a cardiacspeciﬁc and hypoxia-inducible manner (12–15).
The notions that mammalian cardiomyocytes irreversibly
withdraw from the cell cycle soon after birth and that the mammalian heart is a terminally differentiated organ with no intrinsic
capacity to regenerate after myocardial injury have been seriously
challenged in recent years (16). Nevertheless, the innate re-
www.pnas.org/cgi/doi/10.1073/pnas.1018925108
generative ability of the adult heart is limited and not enough to
restore cardiac function adequately after a MI. Several laboratories have shown that some angiogenic factors can induce myocyte
proliferation (9, 17–21). Therefore, in addition to analyzing the
angiogenic effect of VEGF and Ang1, we have also studied their
roles in promoting cell proliferation.
In this study, we used a porcine acute MI model to study the
beneﬁts of AAV-mediated, cardiac-speciﬁc, and hypoxia-inducible
coexpression of VEGF and Ang1 and investigated the underlying
mechanisms. Similar to what we have shown in the mouse acute MI
model (14), the AAV vectors that we have developed could express
VEGF and Ang1 in a cardiac-speciﬁc and hypoxia-inducible
manner in the pig MI model. Coexpression VEGF and Ang1 in
porcine MI heart induced angiogenesis, stimulated cardiomyocyte
proliferation, and reduced apoptosis. Together, they resulted in the
improvement of cardiac functions.
Results
Cardiac-Speciﬁc and Hypoxia-Inducible VEGF and Ang1 Expression.
AAV-MLCVEGF and AAV-MLCAng1 were coinjected into six
sites of the myocardium around the ischemic area immediately
after ligation of the left anterior descending coronary artery
(LAD). Eight weeks after the injection, the distribution of AAV1MLCVEGF genome was analyzed by PCR using a pair of human
VEGF speciﬁc primers with the genomic DNA isolated from the
infarct, peri-infarct, and normal tissues of the myocardium, lung,
liver, and kidney of all animals (SI Materials and Methods). The
vector genome was detected in the DNA isolated from all three
myocardial zones (infarct, peri-infarct, and normal) and liver of
VEGF/Ang1-injected pigs (Fig. 1A). No vector was detected in
the DNA isolated from the lung or kidney. Thus, the AAVMLCVEGF and AAV-MLCAng1 were successfully delivered
into the myocardium, although some vectors leaked into liver even
when they were injected directly into myocardium.
RT-PCR (Fig. 1B) and Western blot (Fig. 1C) showed that
human VEGF and Ang1 expression were detected in all three
zones of myocardium, with a lower level of expression in the
normal zone than in the infarct and peri-infarct zones (P < 0.01).
No human VEGF and Ang1 expressions were detected in the
liver; even the vector genomes were presented (Fig. 1 B and C).
These results indicated that the gene expression mediated by
Author contributions: Y.W.K., Z.Y., and H.S. designed research; Z.T., B.C., X.T., Y.Z., L.W.,
T.Z., and K.C. performed research; Z.T., B.C., X.T., Y.Z., and Z.Y. analyzed data; and Z.T.,
Z.Y., and H.S. wrote the paper.
The authors declare no conﬂict of interest.
Freely available online through the PNAS open access option.
1
To whom correspondence may be addressed. E-mail: [email protected] or yw.
[email protected].
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1018925108/-/DCSupplemental.
PNAS Early Edition | 1 of 6
MEDICAL SCIENCES
Contributed by Yuet Wai Kan, December 20, 2010 (sent for review July 31, 2010)
Fig. 1. Vector distribution and VEGF and Ang1 expressions. (A) Distribution
of AAV genomes after intramyocardial injection. Representative pictures of
RT-PCR (B) and Western blots (C). Bar graphs are quantitative data of RT-PCR
(B) and Western blot (C), respectively. *P < 0.01 vs. normal myocardium. #P <
0.05 vs. normal myocardium. I, infarct zone; P, peri-infarct zone; N, normal
zone; K, kidney; Li, liver; L, lung; LacZ, infarct myocardium of LacZ group;
saline, infarct myocardium of saline group.
AAV-MLCVEGF and AAV-MLCAng1 was cardiac-speciﬁc and
hypoxia-inducible in the porcine acute MI model.
Improvement of Myocardial Perfusion and Cardiac Function. Gated
single-photon emission computed tomography (SPECT) was used
to compare myocardial perfusion and cardiac functions at 2 and
8 wk after treatment. The VEGF/Ang1 group had better left
ventricular ejection fraction (LVEF) at 8 wk than at 2 wk (P <
0.01) (Fig. 2B). Myocardial perfusion score of the VEGF/Ang1
group was signiﬁcantly improved at 8 wk compared with 2 wk (P <
0.01) (Fig. 2 A and C). Although the left ventricular end diastolic
volume (LVEDV) of the VEGF/Ang1 group at 2 and 8 wk was not
different (P > 0.05), the left ventricular end systolic volume
(LVESV) was signiﬁcantly reduced at 8 wk compared with 2 wk
(P < 0.05) (Fig. 2D). None of these measurements showed any
signiﬁcant changes in the saline- or LacZ-injected controls (P >
0.05) (Fig. 2 B, C, and E). Thus, coexpression of VEGF and Ang1
improves myocardial perfusion and cardiac functions of porcine
acute MI heart.
No Increase of VEGF and Ang1 Levels in the Serum. To test whether
intramyocardial injection of AAV1-MLCVEGF and AAV1MLCAng1 would increase serum levels of VEGF and Ang1, we
analyzed human VEGF and Ang1 levels in the sera of treated pigs
at 0, 1, 3, and 7 d and 2, 3, 4, and 8 wk after vector injection by
ELISA. No signiﬁcant increases in human VEGF and Ang1 levels
in VEGF/Ang1 group were detected compared with the LacZ and
saline groups (P > 0.05) (Table 1). Thus, intramyocardial injection of AAV1-MLCVEGF and AAV1-MLCAng1 did not increase the serum VEGF and Ang1 levels.
Reduction of Cardiomyocyte Apoptosis in Infarct and Peri-Infarct
Myocardium. The hearts of the VEGF/Ang1 group showed sig-
niﬁcant decrease in TUNEL-positive nuclei in the myocardium of
Fig. 2. Myocardial perfusion and cardiac function. (A) Representative
images of SPECT of saline- (Top), LacZ- (Middle), and VEGF/Ang1-treated
(Bottom) porcine acute MI hearts. Thirty-two images were acquired, 40 s
each, at every 5.6° throughout a 180° arc, beginning 45° right anterior
oblique and ending 45° left posterior oblique. There was better myocardial
perfusion in VEGF/Ang1-treated heart at 8 wk compared with 2 wk (red
arrows), but similar myocardial perfusion in saline- and LacZ-injected heart
between 2 and 8 weeks (white arrows). (B) Bar graph showing that LVEF
improved at 8 vs. 2 wk after VEGF/Ang1 injection (*P < 0.01). (C) Myocardial
perfusion score improved at 8 vs. 2 wk after VEGF/Ang1 injection (*P < 0.01).
(D) LVESV improved at 8 vs. 2 wk after VEGF/Ang1 injection (*P < 0.05). (E)
LVEDV was not changed at 8 vs. 2 wk after VEGF/Ang1 injection (P > 0.05).
infarct and peri-infarct zones compared with the LacZ and salineinjected controls (P < 0.01) (Fig. 3 A and B). The percentage of
apoptotic cardiomyocyte at the infarct and peri-infarct zones of
VEGF/Ang1 group (33 ± 5.7 and 13 ± 3.3) was signiﬁcantly lower
compared with the saline (53 ± 7.9 and 27 ± 6.7) and LacZ (56 ±
10.1 and 25 ± 6.3) groups (P < 0.01). The percentage of apoptotic
cardiomyocyte in the normal myocardial zone was similar among
the three groups (P > 0.05) (Fig. 3 C and D). Consistent with the
cytological analysis, caspase-3 expression in the infarct and periinfarct zones of VEGF/Ang1 group was the lowest among the
three groups (Fig. 3E).
To analyze the mechanism of decreased cellular apoptosis mediated by the injections, the apoptotic pathway was analyzed by
Western blots. There were more prosurvival proteins, such as
phosphorylated Akt and Bcl-xL, and less proapoptotic proteins,
such as Bad, in VEGF/Ang1-injected hearts compared with those
injected with LacZ and saline. Bcl-2 expression was similar among
all three zones of the three groups (Fig. 3E). The results suggest that
the overexpression of VEGF and Ang1 activated the prosurvival
pathways, which resulted in reduction of cardiomyocyte death.
Increase of Vascular Density in Myocardial Infarct and Peri-Infarct
Zones. To analyze the angiogenesis mediated by coexpression of
VEGF/Ang1, small blood vessels on lectin-stained sections were
counted by two investigators who were blinded to the experimental groups. The VEGF/Ang1 group had signiﬁcantly more
capillaries in the myocardial infarct (48 ± 8.3/mm2) and periinfarct (92 ± 15.5/mm2) zones than those injected with saline
Table 1. VEGF and Ang1 levels in the sera (μg/μL)
VEGF
Ang1
0d
1d
3d
7d
2 wk
3 wk
4 wk
8 wk
9 ± 2.1
22 ± 5.0
8 ± 2.8
24 ± 7.4
8 ± 2.2
23 ± 7.0
8 ± 1.8
22 ± 8.6
9 ± 3.1
23 ± 7.9
8 ± 2.5
23 ± 10.3
10 ± 4.2
26 ± 9.7
9 ± 2.6
24 ± 10.6
No signiﬁcant increase of VEGF and Ang1 levels in the sera of pigs was detected 0, 1, 3, and 7 d and 2, 3, 4, and
8 wk after the vector injection (P > 0.05).
2 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1018925108
Tao et al.
(22 ± 8.7 and 60 ± 13.9/mm2) and LacZ (23 ± 8.6 and 59 ±
16/mm2; P < 0.01) (Fig. 4 A and B). Similarly, the density of
α-smooth muscle actin (α-SMA)-positive vessels in the myocardial
infarct (15 ± 4.2/mm3) and peri-infarct (28 ± 10.1/mm2) zones of
VEGF/Ang1 group was signiﬁcantly higher than in saline (8 ± 4.4
and 18 ± 5.4/mm2) and LacZ (9 ± 4.6 and 19 ± 6.7/mm2) groups
(P < 0.01) (Fig. 4 C and D). The densities of capillaries and
α-SMA–positive vessels in the infarct and peri-infarct zones were
not signiﬁcantly different between the LacZ and saline groups
(P > 0.05) (Fig. 4 B and D).
To identify newly formed blood vessels, we used an antibody
speciﬁc to Ki67 to examine proliferating endothelial cells. We
detected signiﬁcantly more Ki67-positive endothelial cells in
VEGF/Ang1-injected hearts (12 ± 4.5/mm2) than in saline- (3 ±
2.3/mm2) and LacZ-injected (3 ± 2.3/mm2) controls (P < 0.01)
(Fig. 4 E and F). Thus, overexpression of VEGF and Ang1 induced active angiogenesis in the ischemic myocardium.
Increased Cardiomyocyte Proliferation. To determine whether
overexpression of VEGF and Ang1 increases cardiomyocyte
proliferation, we did double labeling using antibodies against
Ki67 and troponin T. As shown in Fig. 5 A and B, there were
more Ki67-positive cardiomyocytes in the VEGF/Ang1-injected
hearts (14 ± 2.2%) than that in saline- (4 ± 1.1%) and LacZ- (3 ±
0.8%) injected controls (P < 0.01).
Flow cytometry data were consistent with the immunostaining
data. There were more S phase cardiomyocytes (DipS) in the
VEGF/Ang1 group (43 ± 9.2%) compared with saline (16 ±
6.0%) and LacZ (16 ± 4.2%) groups (P < 0.01) (Fig. 5 C and D).
The expression of cell cycle proteins was also analyzed (Fig. 5E).
p27, a cell cycle inhibitor, was 2.6- and 3.5-fold lower in the infarct and peri-infarct zones of VEGF/Ang1 group compared with
the saline group, and 2.3- and 4.1-fold lower compared with the
LacZ group (Fig. 5F). The expression of cyclin D2 in the infarct
and peri-infarct zones of VEGF/Ang1 group was increased 2.4and 2.6-fold compared with the saline group and 2.2- and 2.7fold compared with the LacZ group (Fig. 5F). Similarly, cdk4
levels in the infarct and peri-infarct zones of the VEGF/Ang1
group were increased 3.6- and 1.9-fold compared with the saline
group and 4.0- and 2.0-fold compared with the LacZ group (Fig.
5F). The infarct and peri-infarct zones of VEGF/Ang1 group
Tao et al.
expressed 2.3- and 2.2-fold higher cyclin A protein than in the
saline group and 2.4- and 1.9-fold higher cyclin A protein than in
the LacZ group (Fig. 5F). The expression of cdk2 in the infarct
and peri-infarct zones of the VEGF/Ang1 group increased 3.7and 3.4-fold compared with saline group and 3.8- and 3.2-fold
Fig. 4. Vascular density. (A) Representative pictures of lectin-stained sections. (B) Bar graphs showing the quantiﬁcation of microvessels. (C) Representative pictures showed α-SMA–positive arterioles. (D) Bar graphs showing
the quantiﬁcation of α-SMA–positive vessels (50–100 μm diameter). (E) Immunoﬂuorescence of the Ki67-positive cells in the border zone of the infarct.
(F) Bar graph showing the quantiﬁcation of Ki67-positive vessels. *P < 0.01
vs. LacZ and saline groups.
PNAS Early Edition | 3 of 6
MEDICAL SCIENCES
Fig. 3. Cardiomyocyte apoptosis. (A) Representative images of TUNEL-stained sections. TUNEL-positive nuclei are green, and DAPI is blue. (B) Bar graph
showing the quantiﬁcation of the apoptotic cells. *P < 0.05 vs. LacZ and saline groups. (C) Representative images. (D) Bar graph showing the percentage of
apoptotic cardiomyocytes by ﬂow cytometry. *P < 0.05 vs. LacZ and saline groups. (E) Representative Western blots showing expressions of prosurvival and
proapoptotic genes. I, infarct zone; P, peri-infarct zone; N, normal zone.
Fig. 5. Cardiomyocyte proliferation. (A) Representative pictures of sections
stained with troponin T (green), Ki67 (red), and DAPI (blue). (B) Bar graph
showing the quantiﬁcation of Ki67-positive myocytes. (C) Flow cytometry
data. The areas denoted by the diagonal lines are cardiomyocytes in S phase
(DipS), and they show signiﬁcantly more DipS cardiomyocytes in the VEGF/
Ang1 group than in the saline and LacZ groups. (D) Bar graphs showing the
quantiﬁcation of cardiomyocytes in S phase. (E) Representative images of
Western blots showing the expression of genes related to cell cyclin. (F) Bar
graphs showing the quantiﬁcation of cyclin protein. *P < 0.01 vs. LacZ and
saline groups. N, normal zone; P, peri-infarct zone; I, infarct zone.
compared with LacZ group (Fig. 5F). Taken together, these
results suggested that VEGF/Ang1 could drive cardiomyocytes
into S phase and increase cardiomyocyte proliferation through
the Akt/cyclin D2 and cyclin A pathways.
Discussion
About 6–12% of symptomatic patients with extensive coronary
artery disease are not amenable to conventional treatment, such
as percutaneous transluminal coronary angioplasty (PTCA) or
coronary artery bypass grafting (CABG) (22). Although available
treatments for severe coronary insufﬁciency are often helpful,
restenosis of coronary vessels occurs in 30–35% of the patients
(23). Therefore, the development of new treatment strategies is
still necessary.
It has been reported that unregulated and continuous expression of VEGF can lead to angioma formation at the site of injection (24, 25). A high level of circulating VEGF in acute
myocardial infarction can induce acute cor pulmonale, resulting
in increased mortality (26). Thus, control of angiogenic factors
expression both in levels and locations is important to enhance
their local therapeutic efﬁciency and decrease their possible
adverse effects.
Many vectors can be used for in vivo gene deliveries. The AAV
vector has many advantages compared with other vectors, such as
it has the potential for mediating long-term transgene expression
and inducing less immunogenicity (27). Hoshijima et al. (28) have
used AAV2 to express a pseudophosphorylated mutant of human
phospholamban (PLN) in BIO14.6 cardiomyopathic hamsters.
The AAV2/PLN treatment enhanced myocardial sarcoplasmic
reticulum (SR) Ca2+ uptake and suppressed progressive impairment of left ventricular systolic function and contractility for 28–
30 wk, thereby protecting cardiac myocytes from cytopathic
plasmamembrane disruption (28). As to acute myocardial infarction, the early gene expression is crucial for the treatment. We
compared gene expression and therapeutic effect of AAV1-5
4 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1018925108
vectors. Our result shows that the onset of gene expression delivered by AAV serotypes 1 has minimal delay (15).
Because the porcine heart resembles the human heart in size and
coronary structure, the pig was selected as the animal model in this
study. Our results showed that, although the AAV genomes were
detected in liver after intramyocardial injection of AAVMLCVEGF and AAV-MLCAng1, no gene expression was found.
VEGF and Ang1 expressions were only indentiﬁed in the myocardial infarct, peri-infarct zone along with a low level in normal
myocardium. Also, these angiogenic factors were not increased in
the systemic circulation. These results conﬁrmed that the 250-bp
MLC promoter and nine copies of HRE in our AAV vector can
control VEGF and Ang1 gene expression speciﬁcally in the ischemic myocardium of pigs. The lower level of VEGF and Ang1
expressions detected in the normal myocardium could partially be
because of hypoxia related to anesthesia or thoracotomy surgery.
We have shown in previous publications that overexpression of
VEGF and Ang1 improves cardiac function in mouse and rat MI
heart models through their angiogenic and antiapoptotic functions
(4, 15). Here, we have showed that VEGF/Ang1 induced neovascular formation and reduced myocardial apoptosis in infarct
and peri-infarct zones of porcine acute MI heart. Hemangioma was
not detected in the VEGF/Ang1-treated heart. Myocardial perfusion score and LVEF of the VEGF/Ang1-injected pigs were
signiﬁcantly improved. The LVESV was lower at 8 wk compared
with 2 wk. No improvements were observed in the saline- and
LacZ-injected controls. In the American College of Cardiology
(ACC)/American Heart Association (AHA)/American Society for
Nuclear Cardiology (ASNC) Guidelines for the Clinical Use of
Cardiac Radionuclide Imaging, gated SPECT is an important
noninvasive method used for assessment of myocardial perfusion,
left ventricular function, and myocardial viability. Increase of
LVEF and reduction of LVESV indicate improvement of cardiac
function (29). This study showed that coexpression of VEGF and
Ang1 resulted in improvement of myocardial perfusion and cardiac
function in pigs with acute myocardial infarction.
We have previously reported that VEGF/Ang1 delivered by
adenovirus vector reduces cardiomyocyte apoptosis in vitro
through the Akt/Bcl-2 pathway (4). In this study, Akt was also
activated in VEGF- and Ang1-transduced porcine MI heart. Bad
expression was lower in the VEGF/Ang1 group than in the LacZ
and saline groups, but Bcl-xL expression was signiﬁcantly upregulated. Therefore, this study in the porcine heart conﬁrmed
our previous ﬁnding that VEGF and Ang1 provide cardiac
protection through activation of the Akt/Bcl-2 pathway.
In addition to inducing angiogenesis and reducing apoptosis, we
show in this study that coexpression of VEGF and Ang1 in MI heart
also increases cell proliferation through activation of Akt kinase,
up-regulation of cyclin D2/cdk4 and cyclin A/cdk2 expression,
driving of cardiomyocyte into S phase, and promoting proliferation
of cardiomyocytes. It has been shown that Akt kinase expands the
population of cycling myocytes in the postnatal heart and the
number of progenitor cells expressing markers of myocyte lineage in
the adult heart (30). It is also possible that VEGF/Ang1 treatment
in vivo might increase proliferation of resident stem cells in the
heart or recruit peripheral stem cells to the infarcted myocardium.
Different vector delivery routes have been tested in animal MI
models and clinical trials. It is not yet known which one is the safest
and most effective. The effect of i.v. administration of angiogenic
growth factors is minimal and is likely because of ﬁrst-pass uptake
by the lungs, resulting in considerably lower concentration in the
myocardium compared with that with intracoronary administration
(31). Intracoronary injection requires skillful technique of catheterization. Epicardial myocardial injection has been attempted
successfully with an adenoviral vector containing the VEGF121
gene that was injected directly into ischemic myocardium of pig
hearts during thoracotomy. Both myocardial perfusion and function in this model were signiﬁcantly improved (32). Thus, this apTao et al.
Materials and Methods
AAV Vector Construction and Production. AAV-MLCVEGF and AAV-CMVLacZ
constructed for previous experiments were used in this study (11–14).
AAV-MLCAng1 was constructed by replacing human VEGF gene in the
AAV-MLCVEGF vector with human Ang1 cDNA. AAV vectors were packaged in AAV serotype one capsid and prepared by the three plasmid
cotransfection system (33). More details are described in SI Materials
and Methods.
Myocardial Infarction and AAV Vector Injection. Twenty-four Yorkshire pigs
(male, 3–4 mo old, 35 ± 2.7 kg body weight) were housed in the animal care
facility of Nanjing Medical University. The animals were randomly divided
into three groups of eight: AAV1-MLCVEGF and AAV1-MLCAng1 coinjected
group (VEGF/Ang1 group), AAV1-CMVLacZ-injected group (LacZ), and Hepes
saline-injected group (saline). The detailed procedures for MI creation and
AAV vector injection were described in SI Materials and Methods.
Myocardial Perfusion and Cardiac Function. Gated SPECT was performed in the
pigs with a commercially available system (ECAM+; Siemens). Left ventricular
end diastolic volume (LVEDV), left ventricular end systolic volume (LVESV), and
left ventricular ejection fraction (LVEF) were determined by QGS software. The
pigs were habituated to the experimental environment for 2 h before the
SPECT analysis. The imaging was assessed at a dose of 0.3 m Ci/kg 99mTc sestamibi (99mTc-MIBI) at 2 and 8 wk after LAD ligation and vector injection.
Images were acquired 60–90 min after 99mTc-MIBI injection using a multihead
camera with high-resolution collimators. The camera energy window (20%)
was set on the 140 ke V photopeak of 99mTc-MIBI. Particular care was taken
to avoid movement of the pig or overlap from extracardiac activity. Thirtytwo images (64 × 64 matrix) were acquired for 40 s each with 180° rotation.
Tao et al.
The tomograms were reconstructed in the vertical and horizontal long- and
short-axis planes.
The myocardial perfusion score was calculated as follows. The left ventricular cavity was divided into 17 segments for assessment of the myocardium (34). Seven segments are assigned to the LAD distribution area. The
ﬁve-point scoring used to grade each segment was as follows: 0, absent
perfusion; 1, severe hypoperfusion; 2, moderate hypoperfusion; 3, mild
hypoperfusion; 4, normal perfusion. The normal total score of LAD distribution areas is 28 (35).
ELISA. Blood samples were collected at 0, 1, 3, and 7 d and 2, 3, 4, and 8 wk
after the LAD ligation and vector injection. The samples were centrifuged
for 10 min at 1,000 × g, and the serum was collected. The levels of human
VEGF and Ang1 in serum were quantiﬁed using an ELISA kit for human
VEGF and Ang1 (Quantikine; R&D Systems). The samples were measured
in duplicates.
PCR and RT-PCR Analysis. The TRIzol RNA isolation system (GIBCO/BRL) was used
to isolate RNA and genomic DNA from the heart, liver, lung, and kidney. We
performed PCR and RT-PCR analysis as described in SI Materials and Methods.
Immunoprecipitation and Western Blotting Analysis. The tissues were gently
homogenized in lysis buffer (50 mM Tris, pH 7.5, 10 mM sodium phosphate,
150 mM NaCl, 1% Triton X-100, 5 mM EDTA, 10 mM NaF, 5 mM iodoacetate, 1
mM benzamidine, 5 μg/mL leupeptin, 5 μg/mL aprotinin) and incubated on
ice for 30 min. The lysates were centrifuged at 10,000 × g for 20 min to
remove insoluble material, and the total protein concentration was determined by modiﬁed Bradford assay. All samples were stored at −80 °C
before electrophoresis. The detailed procedures for immunoprecipitation
and Western blotting analysis are given in SI Materials and Methods.
Histological and Immunohistochemical Analysis. Hearts were ﬁxed in 10%
formalin, embedded in parafﬁn, and sectioned. Serial sections were made
from the apex of the heart to the site of the ligation. Sections were stained
with speciﬁc antibodies against α-SMA, Ki67, and troponin T (Thermo Scientiﬁc) antibodies. Nuclei were stained with DAPI. Lectin-FITC (Santa Cruz)
was used to visualize the blood vessels. The immunohistochemical staining
was carried out according to the manufacturer’s instructions, and signals
were visualized by incubating the sections with Alexa Fluor-488– or -555–
labeled secondary antibody (Molecular Probes).
For quantiﬁcation of vessel densities in the myocardium, four equivalent
sections from each group were randomly selected, and six visual ﬁelds from
each section were observed. As a surrogate of vessel counting, vessel density
was determined by lectin optical density measurements using the QWIN100
immunohistochemical imaging analyzer system (Leica).
Cardiomyocyte Apoptosis. Cardiomyocyte apoptosis was measured by double
immunoﬂuorescence staining of TUNEL using the In Situ Cell Death Detection
Kit, Fluorescein (Roche). SI Materials and Methods contains more details.
Flow Cytometry Analysis. Cell suspensions were prepared using the Medimachine System (Becton-Dickinson). Cardiomyocyte cells were collected by
ﬂow cytometry with the use of an antibody to α-sarcomeric actinin. The cells
(1 × 106 per sample) were ﬁxed in 70% ethanol, double stained for FITC–
Annexin V binding and propidium iodide, and analyzed by ﬂow cytometry
using a FACS-SCAN apparatus (Becton-Dickinson). Cell cycle distribution was
analyzed for DNA content using propidium iodide in ﬂow cytometry.
Statistical Analysis. Statistical analysis was performed with SPSS 10.0 software.
Student t tests and one-way ANOVA testing were used to compare the
differences among groups, with statistical signiﬁcance considered if P < 0.05.
The data are presented as mean ± SD.
ACKNOWLEDGMENTS. The authors thank Drs. Li Yuan and Ying-Bin Ge for
their technical assistance, Dr. Xiao Xiao for providing the plasmid containing
AAV rep and AAV serotype 1 cap genes, Dr. Feng Jianlin and Dr. Yang
Guoping for their technical assistance, and Grant Chen for editing the
manuscript. This work was supported by a project grant of the Key Faculty of
Medical Renaissance Program of Jiangsu province, National Nature Science
Foundation of China (Grant 30971254), National Basic Research Program
(973 program; Grant 2008CB517303), National Institutes of Health Grant
HL67969 (to Y.W.K.), and American Heart Association Scientist Development
Grant 0535018N (to H.S.).
PNAS Early Edition | 5 of 6
MEDICAL SCIENCES
proach was selected in this study. One of the limitations of the
procedure is that it needs an open surgery and may not applicable
to most of the patients. Recently, we developed a set of catheterbased percutaneous endocardial intramyocardial injection system
guided by EnSite NavX for delivery of genes or cells to the myocardium. The vectors that we have tested in this paper can be delivered to patients using our newly developed method.
The limitations of this study include (i) a lack of groups that were
injected with AAV-MLCVEGF or AAV-MLCAng1 individually
because of the cost of the pig experiment. Thus, the effect of
coexpression of VEGF and Ang1 was not compared with individual expression. However, our previously published paper on
the mouse MI model showed that AAV-mediated coexpression of
Ang1 and VEGF results in a decreased vascular leakage compared
with VEGF expressed alone. That report also showed a smaller
infarct size and better cardiac function when AAV-MLCVEGF
and AAV-MLCAng1 were coinjected (11). (ii) Because Akt inhibitor was not used in this study, there is no direct evidence that
VEGF- and Ang1-induced increase of cell cyclin gene expression
and cell proliferation is dependent on Akt signaling pathways. (iii)
The functions of VEGF and Ang1 on induction of endogenous
stem cell homing, proliferation, and differentiation and endothelial
progenitor incorporation into the neovasculature were not
studied. (iv) The effects of VEGF and the other angiogenic factors
seem to be multifactorial; it is not clear how much of the therapeutic beneﬁt that we identiﬁed in this study can be attributed to
the increase of perfusion. (v) The therapy was administered immediately after coronary ligation. This study was targeted to acute
MI to reduce mortality by avoiding a second operation. Future
studies will be performed to address these issues and study the
therapeutic effects of this approach in chronic MI.
In summary, we have shown that intramyocardial coinjection of
AAV1- MLCVEGF and AAV1-MLCAng1 can result in cardiacspeciﬁc and hypoxia-inducible gene expression in porcine acute
MI heart, induce angiogenesis and cardiomyocyte proliferation,
reduce cardiomyocyte apoptosis, and improve myocardial perfusion and cardiac functions. Thus, the usefulness of AAV1MLCVEGF and AAV1-MLCAng1 in protecting patients from
cardiac injury warrants further investigation.
1. Schumacher B, Pecher P, von Specht BU, Stegmann T (1998) Induction of
neoangiogenesis in ischemic myocardium by human growth factors: First clinical
results of a new treatment of coronary heart disease. Circulation 97:645–650.
2. Giordano FJ, et al. (1996) Intracoronary gene transfer of ﬁbroblast growth factor-5
increases blood ﬂow and contractile function in an ischemic region of the heart. Nat
Med 2:534–539.
3. Harada K, et al. (1994) Basic ﬁbroblast growth factor improves myocardial function in
chronically ischemic porcine hearts. J Clin Invest 94:623–630.
4. Zhou L, et al. (2005) VEGF165 and angiopoietin-1 decreased myocardium infarct size
through phosphatidylinositol-3 kinase and Bcl-2 pathways. Gene Ther 12:196–202.
5. Pearlman JD, et al. (1995) Magnetic resonance mapping demonstrates beneﬁts of
VEGF-induced myocardial angiogenesis. Nat Med 1:1085–1089.
6. Losordo DW, et al. (1998) Gene therapy for myocardial angiogenesis: Initial clinical
results with direct myocardial injection of phVEGF165 as sole therapy for myocardial
ischemia. Circulation 98:2800–2804.
7. Takeshita S, et al. (1994) Intramuscular administration of vascular endothelial growth
factor induces dose-dependent collateral artery augmentation in a rabbit model of
chronic limb ischemia. Circulation 90:II228–II234.
8. Shyu KG, Manor O, Magner M, Yancopoulos GD, Isner JM (1998) Direct intramuscular
injection of plasmid DNA encoding angiopoietin-1 but not angiopoietin-2 augments
revascularization in the rabbit ischemic hindlimb. Circulation 98:2081–2087.
9. Yang ZJ, et al. (2006) Experimental study of bone marrow-derived mesenchymal stem
cells combined with hepatocyte growth factor transplantation via noninfarct-relative
artery in acute myocardial infarction. Gene Ther 13:1564–1568.
10. Wang W, et al. (2006) Induction of collateral artery growth and improvement of postinfarct heart function by hepatocyte growth factor gene transfer. Acta Pharmacol Sin
27:555–560.
11. Su H, et al. (2009) Additive effect of AAV-mediated angiopoietin-1 and VEGF expression on the therapy of infarcted heart. Int J Cardiol 133:191–197.
12. Su H, Lu R, Kan YW (2000) Adeno-associated viral vector-mediated vascular endothelial growth factor gene transfer induces neovascular formation in ischemic heart.
Proc Natl Acad Sci USA 97:13801–13806.
13. Su H, Arakawa-Hoyt J, Kan YW (2002) Adeno-associated viral vector-mediated
hypoxia response element-regulated gene expression in mouse ischemic heart model.
Proc Natl Acad Sci USA 99:9480–9485.
14. Su H, et al. (2004) Adeno-associated viral vector delivers cardiac-speciﬁc and hypoxiainducible VEGF expression in ischemic mouse hearts. Proc Natl Acad Sci USA 101:
16280–16285.
15. Su H, et al. (2006) AAV serotype-1 mediates early onset of gene expression in mouse
hearts and results in better therapeutic effect. Gene Ther 13:1495–1502.
16. Kajstura J, et al. (1998) Myocyte proliferation in end-stage cardiac failure in humans.
Proc Natl Acad Sci USA 95:8801–8805.
17. Engel FB, Hsieh PC, Lee RT, Keating MT (2006) FGF1/p38 MAP kinase inhibitor therapy
induces cardiomyocyte mitosis, reduces scarring, and rescues function after myocardial infarction. Proc Natl Acad Sci USA 103:15546–15551.
18. Chaudhry HW, et al. (2004) Cyclin A2 mediates cardiomyocyte mitosis in the
postmitotic myocardium. J Biol Chem 279:35858–35866.
6 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1018925108
19. Nakajima H, Nakajima HO, Tsai SC, Field LJ (2004) Expression of mutant p193 and p53
permits cardiomyocyte cell cycle reentry after myocardial infarction in transgenic
mice. Circ Res 94:1606–1614.
20. Woo YJ, et al. (2006) Therapeutic delivery of cyclin A2 induces myocardial
regeneration and enhances cardiac function in ischemic heart failure. Circulation 114
(Suppl 1):I206–I213.
21. Tamamori-Adachi M, et al. (2008) Cardiomyocyte proliferation and protection against
post-myocardial infarction heart failure by cyclin D1 and Skp2 ubiquitin ligase.
Cardiovasc Res 80:181–190.
22. Gill M, et al. (2001) Vascular trauma induces rapid but transient mobilization of
VEGFR2(+)AC133(+) endothelial precursor cells. Circ Res 88:167–174.
23. Libby P, Schwartz D, Brogi E, Tanaka H, Clinton SK (1992) A cascade model for
restenosis. A special case of atherosclerosis progression. Circulation 86(Suppl 6):
III47–III52.
24. Lee RJ, et al. (2000) VEGF gene delivery to myocardium: Deleterious effects of
unregulated expression. Circulation 102:898–901.
25. Schwarz ER, et al. (2000) Evaluation of the effects of intramyocardial injection of DNA
expressing vascular endothelial growth factor (VEGF) in a myocardial infarction
model in the rat—angiogenesis and angioma formation. J Am Coll Cardiol 35:
1323–1330.
26. Matsuno H, et al. (2002) Lack of alpha2-antiplasmin promotes pulmonary heart
failure via overrelease of VEGF after acute myocardial infarction. Blood 100:
2487–2493.
27. Kaye DM, Hoshijima M, Chien KR (2008) Reversing advanced heart failure by
targeting Ca2+ cycling. Annu Rev Med 59:13–28.
28. Hoshijima M, et al. (2002) Chronic suppression of heart-failure progression by
a pseudophosphorylated mutant of phospholamban via in vivo cardiac rAAV gene
delivery. Nat Med 8:864–871.
29. Klocke FJ, et al. (2003) ACC/AHA/ASNC guidelines for the clinical use of cardiac
radionuclide imaging—executive summary: A report of the American College of
Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/
ASNC Committee to Revise the 1995 Guidelines for the Clinical Use of Cardiac
Radionuclide Imaging). J Am Coll Cardiol 42:1318–1333.
30. Gude N, et al. (2006) Akt promotes increased cardiomyocyte cycling and expansion of
the cardiac progenitor cell population. Circ Res 99:381–388.
31. Kornowski R, Fuchs S, Leon MB, Epstein SE (2000) Delivery strategies to achieve
therapeutic myocardial angiogenesis. Circulation 101:454–458.
32. Mack CA, et al. (1998) Biologic bypass with the use of adenovirus-mediated gene
transfer of the complementary deoxyribonucleic acid for vascular endothelial growth
factor 121 improves myocardial perfusion and function in the ischemic porcine heart.
J Thorac Cardiovasc Surg 115:168–176.
33. Matsushita T, et al. (1998) Adeno-associated virus vectors can be efﬁciently produced
without helper virus. Gene Ther 5:938–945.
34. Cerqueira MD, et al. (2002) Standardized myocardial segmentation and nomenclature
for tomographic imaging of the heart: A statement for healthcare professionals from
the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American
Heart Association. Circulation 105:539–542.
35. Poornima IG, et al. (2004) Utility of myocardial perfusion imaging in patients with
low-risk treadmill scores. J Am Coll Cardiol 43:194–199.
Tao et al.