Cardiovascular Research (2012) 93, 181–189
doi:10.1093/cvr/cvr280
Mechanical stretch induces the apoptosis regulator
PUMA in vascular smooth muscle cells
Wen-Pin Cheng 1, Bao-Wei Wang 1, Shih-Chung Chen 2,3, Hang Chang 4,
and Kou-Gi Shyu 1,5*
Division of Cardiology, Shin Kong Wu Ho-Su Memorial Hospital, 95 Wen-Chang Road, Taipei 111, Taiwan; 2Division of Cardiology, New Taipei City Hospital, New Taipei City, Taiwan;
Graduate Institute of Medical Sciences, College of Medicine, Taipei Medical University, Taipei, Taiwan; 4Graduate Institute of Injury Prevention and Control, College of Medicine,
Taipei Medical University, Taipei, Taiwan; and 5Graduate Institute of Clinical Medicine, College of Medicine, Taipei Medical University, Taipei, Taiwan
3
Received 11 April 2011; revised 7 October 2011; accepted 18 October 2011; online publish-ahead-of-print 21 October 2011
Time for primary review: 16 days
Aims
The expression of PUMA (p53-up-regulated modulator of apoptosis), an apoptosis-regulating gene, increases during
endoplasmic reticulum stress. The mechanisms by which cyclic stretch influences the regulation of PUMA in vascular
smooth muscle cells (VSMCs) during apoptosis remain unclear. We hypothesized that cyclic stretch enhances PUMA
expression in VSMCs undergoing apoptosis.
.....................................................................................................................................................................................
Methods
Human VSMCs grown on a Flexcell I flexible membrane base were stretched via vacuum to 20% of elongation at a
frequency of 1 Hz. An in vivo model of volume overload with aorta-caval shunt and pressure overload with aortic
and results
banding in adult rats was used to study PUMA expression. Cyclic stretch markedly enhanced PUMA protein and
gene expression after stretch. Addition of c-jun N-terminal kinase (JNK) inhibitor SP600125 and interferon-g
(IFN-g) antibody 30 min before stretch inhibited PUMA expression. Gel shift assay demonstrated that stretch
increased the DNA binding activity of interferon regulatory factor-1 (IRF-1). SP600125, JNK small interfering
RNA, and IFN-g antibody attenuated the DNA binding activity induced by stretch. PUMA-Mut plasmid, SP600125,
and IRF-1 antibody attenuated the promoter activity. Stretch increased secretion of IFN-g from VSMCs, and conditioned media from stretched VSMCs increased PUMA protein expression. The in vivo model of aorta-caval shunt and
aortic banding also showed increased PUMA protein expression in the aorta.
.....................................................................................................................................................................................
Conclusion
Cyclic mechanical stretch increases PUMA expression in cultured human VSMCs. The PUMA expression induced by
stretch is mediated by IFN-g, JNK, and IRF-1 pathways. These findings suggest that PUMA is an important mediator in
VSMC apoptosis induced by stretch.
----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords
Apoptosis † Smooth muscle cell † p53-Up-regulated modulator of apoptosis † Cyclic stretch
1. Introduction
Although prevention strategies for atherosclerosis have significantly
progressed, atherosclerosis is still a major and increasing health
concern in developed countries. Thus, the development of novel
therapeutic agents for patients with atherosclerosis remains a major
research priority.1 Atherosclerosis is characterized by lipid accumulation within vessel walls, inflammation, and apoptosis of both macrophages and vascular smooth muscle cells (VSMCs).2 VSMC
apoptosis plays an important role in both normal and pathological
remodelling of vessel walls.3 VSMC apoptosis has been increasingly
implicated in both the development and the outcome of atherosclerotic disease.4 Accelerated VSMC apoptosis may induce thinning of
fibrous cap (atheroma caused by vessel inflammation, resulting in
lumen narrowing) and is very important in plaque stability.5 VSMC
apoptosis is rare in normal blood vessels; however, apoptosis is
increased in unstable human atherosclerotic plaques. Apoptosis of
VSMCs is sufficient to induce plaque vulnerability in atherosclerosis.6
Therefore, reducing VSMC apoptosis is a promising strategy for attenuating plaque instability and atherosclerosis.7
Endoplasmic reticulum (ER)-initiated apoptosis is implicated in the
pathophysiology of plaque rupture and atherosclerosis in animal and
human studies.8 One component of the ER stress-mediated apoptosis
pathway is the p53-up-regulated modulator of apoptosis (PUMA), also
known as Bcl-2 binding component 3.9 PUMA is a Bcl-2 homology 3
(BH3)-only Bcl-2 family member and is similar to other BH3-only
* Corresponding author. Tel: +886 2 2833 2211; fax: +886 2 2836 5775, Email: [email protected]
Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2011. For permissions please email: [email protected].
Downloaded from http://cardiovascres.oxfordjournals.org/ at Academia SinicaLife Science Library on March 29, 2012
1
182
2. Methods
2.1 Culture of VSMCs
Human coronary artery VSMCs were purchased from PromoCell (Heidelberg, Germany). Cells were cultured in smooth muscle cell growth medium
2 containing 20% fetal calf serum, 0.1 mmol/L non-essential amino acids,
1 mmol/L sodium pyruvate, 4 mmol/L L-glutamine, 100 U/mL penicillin,
and 100 mg/mL streptomycin at 378C in an atmosphere of 5%CO2 – 95%
air in a humidified incubator. When confluent, VSMC monolayers were passaged every 6 – 7 days after trypsinization and were used for experimentation from the third to sixth passages. These third to sixth passage cells
were then cultured in Flexcell I flexible membrane dishes coated with
Collagen I (FLEX Iw CULTURE PLATES COLLAGEN I, Flexcell International, Hillborough, NC, USA) in smooth muscle cell growth medium 2 containing 0.5% fetal calf serum, and the cells were incubated for an additional
2 days to render them quiescent prior to initiating each experiment.
2.2 In vitro cyclic stretch of cultured VSMCs
The Flexcell FX-2000 strain unit (Flexcell International, Hillborough, NC,
USA) consists of a vacuum unit linked to a valve controlled by a computer
program. VSMCs cultured on the flexible membrane base were subjected
to cyclic stretch produced by this computer-controlled application of sinusoidal negative pressure with a peak level of 15 kPa at a frequency
of 1 Hz for various periods of time.
for reverse transcription-PCR (RT-PCR) is further described in detail in
the Supplementary material online, Methods.
2.5 Real-time PCR
The primers used were as follows: PUMA, 5′ -d(CCCTGGAGGGT
CCTGTACAA)-3′ (forward) and 5′ -d(CTCTGTGGCCCCTGGGTA
A)-3′ (reverse); and glyderaldehyde-3-phosphate dehydrogenase
(GAPDH), 5′ -d(CATCACCATCTTCCAGGAGC) (forward) and 5′ -d(G
GATGATGTTCTGGGCTGCC)-3′ (reverse). Details of the procedures
are further described in the Supplementary material online, Methods.
2.6 Electrophoretic mobility shift assay (EMSA)
Nuclear protein concentrations from cultured VSMCs were determined
by Bio-Rad (Hercules, CA, USA) protein assay. Consensus and control oligonucleotides (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were labelled by polynucleotides kinase incorporation of [g32-p]ATP. After the
IRF-1 was radiolabelled, the nuclear extracts (4 mg of protein in 2 mL of
nuclear extract) were mixed with 20 pmol of the appropriate
[g32-p]ATP-labelled consensus or mutant oligonucleotide in a total
volume 20 mL for 30 min at room temperature. The samples were then
resolved on a 4% polyacrylamide gel. Gels were dried and imaged by autoradiography. Controls were performed in each case with mutant or cold
oligonucleotides to compete with labelled sequence.
2.7 RNA interference
VSMCs were transfected with 800 ng small interfering RNA (siRNA) of
PUMA or c-jun N-terminal kinase (JNK; Dharmacon, Lafayette, CO,
USA). PUMA or JNK siRNA are target-specific 19 or 21 nucleotide
siRNAs according to a computer program provided by Dharmacon. The
PUMA and JNK targeted base sequences were as follows: sense,
5′ -AAGAAGAGCAACAUCGACA and 5′ -CGUGGAUUUAUGGUCU
GUGdTdT-3′ , respectively; and antisense, 5′ -P.ACAAGAAGAGC
AACAUCGA and 5′ -CACAGACCAUAAAUCCACCdTdT-3, respectively. Green fluorescent protein (GFP) siRNA was used as a negative control,
with the following base sequences: sense, 5′ -P.GGCUACGU
CCAGGAGCGCACC-3′ and antisense, 5′ -P.UGCGCUCCUGGACG
UAGCCUU-3′ (Dharmacon). After overnight incubation, cells were
stretched and subjected to analysis by western blot, EMSA, immunohistochemistry and detection of apoptosis.
2.8 Promoter activity assay
A human PUMA promoter construct was generated as previously
described (human PUMA: 22183 to 21490, containing an IRF-1
binding site17 and 2318 to 215, containing a p53 binding site19). The
plasmids were transfected into VSMCs using a low-pressure-accelerated
gene gun (Bioware Technologies, Taipei, Taiwan), essentially following
the protocol from the manufacturer. Test plasmid (2 mg) and control
plasmid (pGL4-Renilla luciferase; 0.02 mg) were co-transfected with a
gene gun in each well, and then replaced with normal cultured medium.
Following 6 h of cyclic stretch, cell extracts were prepared for the DualLuciferase Reporter Assay System (Promega, Madison, WI, USA) and
measured for dual luciferase activity by luminometer (Turner Designs,
Sunnyvale, CA, USA).
2.3 Western blot analysis
A western blot was performed as previously described.18 Antibodies used
for the western blot are described in detail in the Supplementary material
online, Methods.
2.4 Reverse transcription-PCR
Total RNA was isolated from VSMCs using the single-step acid guanidinium thiocyanate– phenol – chloroform extraction method. The method
2.9 Measurement of IFN-g concentrations
Conditioned medium from VSMCs subjected to cyclic stretch and
medium from control (unstretched) cells were collected for IFN-g measurement. IFN-g levels were measured using a quantitative sandwich
enzyme immunoassay technique (R&D Systems, Minneapolis, MN, USA).
The lower limit of detection of IFN-g was 8.0 pg/mL. Both the
intra-observer and the interobserver coefficients of variance were ,10%.
Downloaded from http://cardiovascres.oxfordjournals.org/ at Academia SinicaLife Science Library on March 29, 2012
proteins, which serve as proximal signalling molecules that transduce
death signals to the mitochondria. Previously, Yu and Zhang have
demonstrated that PUMA is a critical mediator of p53-dependent
and –independent apoptosis in response to various stimuli in
several tissues and cells.10 It has been implicated in the pathophysiological mechanisms of several diseases, including cancer, acquired
immunodeficiency syndrome and ischaemic brain disease.11
Mechanical force overload is capable of inducing inflammatory mediators and is known to cause ventricular hypertrophy.12 Furthermore,
using a cyclic strain system on cultured cells subjects them to repetitive
stretching and relaxation at rates comparable to dynamic stretch
overload in vivo. This system has been applied widely in studying the
molecular mechanisms of gene expression and signal transduction in
many cell types.13 Moreover, it has been demonstrated that cyclic
stretch could induce apoptosis of cardiovascular cells.14 PUMA plays
an important role in apoptosis of cardiomyocytes following ER stress
and ischaemia–reperfusion.15 Additionally, PUMA is involved in apoptosis of VSMCs.16 However, few studies have reported on whether
cyclic stretch affects the expression of PUMA in VSMCs undergoing
apoptosis. Interferon regulatory factor-1 (IRF-1) also plays an important
role in apoptosis. In response to interferon-g (IFN-g), PUMA is transactivated by IRF-1.17 Thus, the aim of the present study was to test the
hypothesis that cyclic stretch enhances PUMA expression in VSMCs
undergoing apoptosis. We also investigated the role of IRF-1 in mediating the stretch-induced expression of PUMA.
W-P. Cheng et al.
Regulation of PUMA by mechanical stretch
183
2.10 Cytotoxicity studies
Cell viability after application of cyclic stretch was monitored using a
trypan blue staining procedure and the 3- (4,5-cimethylthiazol-2-yl)-2,5diphenyl tetrazolium bromide (MTT) assay to detect for stretch-induced
cell injury. Cytotoxicity studies were performed as previously described.18
2.11 Flow cytometric analysis and caspase 3
activities for apoptotic quantification
2.12 Terminal deoxynucleotidyl
transferase-mediated dUTP nick-end
labelling assay
DNA fragmentation was determined via terminal deoxynucleotidyl
transferase-mediated dUTP nick-end labelling (TUNEL) using the
ApopTag peroxidase in situ apoptosis detection kit (Chemicon International, Temecula, CA, USA). The methodology for the TUNEL assay
is further described in detail in the Supplementary material online,
Methods.
2.13 The aorta-caval shunt and aortic banding
rat model
Aorta-caval (AV) shunt and aortic banding was performed on rats to
induce volume and pressure overload, respectively. On the day of
surgery, Wistar rats weighing 280 – 330 g were anaesthetized with 2% isoflurane, and the vena cava and aorta were exposed via abdominal mid-line
incision after confirming a fully anaesthetized state (e.g. no response to
toe pinching). Both procedures are further described in the Supplementary material online, Methods. The rats were killed with an overdose of
isoflurane. All study protocols were approved by our Institutional Committee of Animal Care and Use (protocol number #0990816001) and
were carried out in accordance with the Guide for the Care and Use of
Laboratory Animals (NIH publication no. 86-23, revised 1996).
2.14 Statistical analysis
All results are expressed as means + SEM. Statistical significance was evaluated using analysis of variance (ANOVA; GraphPad Software Inc.,
San Diego, CA, USA). Dunnett’s test was used to compare multiple
groups with a single control group. The Turkey – Kramer comparison
was used for pairwise comparisons between multiple groups following
the ANOVA. A value of P , 0.05 was considered as significant.
3. Results
3.1 Cyclic stretch enhances PUMA protein
and mRNA expression in VSMCs
PUMA protein levels began to increase as early as 14 h after stretch to
20% of elongation, reached a maximal increase that was 3.5-fold
higher than control values by 18 h, and remained elevated for up to
24 h. When VSMCs were stretched to 10% of elongation, PUMA
protein levels were similar to those of control cells without stretch
(Figure 1A and B). Real-time PCR demonstrated that PUMA mRNA
Figure 1 Effects of cyclic stretch on PUMA mRNA and protein
expression in VSMCs. (A) Representative western blots for PUMA
in VSMCs subjected to cyclic stretch by 20 or 10% for various
periods of time. (B) Quantitative analysis of PUMA protein levels.
The values from stretched VSMCs have been normalized to
matched a-tubulin measurements and then expressed as a ratio of
normalized values to protein in the control group (n ¼ 5 per
group). (C) Quantitative analysis of PUMA mRNA levels. The
values from stretched VSMCs have been normalized to matched
GAPDH measurements and then expressed as a ratio of normalized
values to mRNA in the control group (n ¼ 5 per group). *P , 0.01,
**P , 0.05 vs. control.
increased significantly after 18 h of stretch at 20% of elongation
(Figure 1C). These results indicate that cyclic stretch induces PUMA
expression in VSMCs.
3.2 Stretch-induced PUMA protein
expression in VSMCs is mediated by JNK
VSMCs were stretched by 20% for 18 h in the presence and absence
of various inhibitors or siRNA to determine the signalling pathway
mediating the stretch-induced increases in PUMA expression in
VSMCs. As shown in Figure 2, stretch-induced increases of PUMA proteins were significantly attenuated after the addition of SP600125 (a
potent, cell-permeable, selective, and reversible inhibitor of JNK)
30 min before stretching of VSMCs. PUMA proteins induced by
stretch were not affected by the addition of PD98059 (a specific
and potent inhibitor of extracellular signal-regulated kinase) and
SB203580 (a highly specific, cell-permeable inhibitor of p38 kinase),
but were partly blocked by the addition of N-acetylcysteine (a free
radical scavenger). JNK siRNA was transfected into VSMCs prior to
Downloaded from http://cardiovascres.oxfordjournals.org/ at Academia SinicaLife Science Library on March 29, 2012
Apoptotic cells were quantified as the percentage of cells with hypodiploid DNA (sub-G1). VSMCs were fixed with 70% ethanol and treated
with RNase. Then nuclei were stained with propidium iodide (Molecular
Probes, Eugene, OR, USA) and fluorescein isothiocyanate– annexin
V. DNA content was measured using by a FACSCalibur flow cytometer
and Cell Quest software (Becton Dickinson, Franklin Lakes, NJ, USA).
Ten thousand cells were counted in all assays. Measurement of caspase
3 activity is described in further detail in the Supplementary material
online, Methods.
184
W-P. Cheng et al.
activity induced by cyclic stretch. Exogenous administration of
IFN-g to VSMCs without stretch increased IRF-1 –DNA binding activity (Figure 3A; see Supplementary material online, Figure S2A). These
results demonstrate that stretch enhances IRF-1 binding activity in
VSMCs. Addition of JNK siRNA and SP600125 significantly decreased
IFN-g-induced IRF-1 binding activity without stretch (Figure 3B; see
Supplementary material online, Figure S2B). Addition of PD98059,
SB203580, angiotensin II, or tumour necrosis factor-a antibody
(5 mg/mL; purchased from R&D Systems) 30 min prior to stretch
had no effect on the IRF-1 binding activity induced by cyclic stretch
(Figure 3B; see Supplementary material online, Figure S2B).
Figure 2 Effects of MAP kinase inhibitors on PUMA protein
expression induced by cyclic stretch in VSMCs. (A) Representative
western blots for PUMA protein levels in VSMCs subjected to
cyclic stretch in the absence or presence of MAP kinase inhibitors,
N-acetylcysteine (NAC), siRNA and vehicle (dimethyl sulfoxide,
DMSO, 0.1%). CM, conditioned medium. (B) Quantitative analysis
of PUMA protein levels. The values from stretched VSMCs have
been normalized to matched a-tubulin measurements and then
expressed as a ratio of normalized values to protein in the control
group (n ¼ 4 per group). *P , 0.01 vs. control.
cyclic stretch to determine whether the JNK mitogen-activated
protein (MAP) kinase pathway mediates the expression of PUMA.
JNK siRNA completely blocked stretch-induced increases in PUMA
expression. Furthermore, conditioned medium alone also had a
similar effect on PUMA expression levels to cyclic stretch, whereas
dimethyl sulfoxide alone, as a vehicle control, and control siRNA
did not affect PUMA expression levels following cyclic stretch. Additionally, both JNK expression and phosphorylation increased following cyclic stretch, and both SP600125 and IFN-g antibody inhibited
JNK phosphorylation induced by cyclic stretch (see Supplementary
material online, Figure S1A and B). Addition of JNK siRNA significantly
decreased JNK phosphorylation induced by stretch. These findings
suggest that the JNK pathway, but not p42/p44 or p38 MAP kinase
pathways, mediates the induction of PUMA proteins by stretch
in VSMCs.
3.3 Cyclic stretch increases IRF-1 binding
activity
Cyclic stretch of VSMCs for 2–6 h significantly increased the DNA–
protein binding activity of IRF-1 (Figure 3A; see Supplementary material online, Figure S2A). An excess of unlabelled IRF-1 oligonucleotide
competed with the probe for binding IRF-1 protein, whereas an oligonucleotide containing a 2 bp substitution in the IRF-1 binding site did
not compete for binding. Addition of SP600125, JNK siRNA, and
IFN-g antibody (5 mg/mL; purchased from R&D Systems) 30 min
prior to stretch significantly inhibited the DNA–protein binding
The promoter region of human PUMA was cloned, and a luciferase
reporter plasmid (pGL3-Luc) was constructed to determine
whether the stretch-induced expression of PUMA is regulated at
the transcriptional level. The PUMA promoter construct contains
cAMP response element B (CREB), nuclear factor of activated T
cells (NFAT), nuclear factor-kB (NF-kB), and IRF-1 binding sites
(Figure 3C). As shown in Figure 3D, a transient transfection of this reporter gene into VSMCs revealed that stretch for 6 h significantly
induces PUMA promoter activation. This finding indicates that
PUMA expression is induced at the transcriptional level during
cyclic stretching of VSMCs. Furthermore, when the IRF-1 binding
sites were mutated, the increases in promoter activity induced by
stretch were abolished. Moreover, addition of the IFN-g antibody
and SP600125 resulted in an inhibition of transcription. Addition of
pifithrin (p53 inhibitor) prior to stretch did not inhibit the expression
of PUMA induced by cyclic stretch (see Supplementary material
online, Figure S2C). When the IRF-1 binding sites were mutated (wildtype, CTGCAAGTCCTGACTTGTCC; and mutant, CTGCAAGT
TAGCACTTGTCC), the increase in promoter activity induced by
stretch was similar to that of wild-type (see Supplementary material
online, Figure S2D and E). This result indicates that increases in the
expression of PUMA in VSMCs are mediated by IRF-1, and are
independent of p53.
3.5 Cyclic stretch stimulates the secretion
of IFN-g from VSMCs
As shown in Figure 4A, cyclic stretch significantly increased the IFN-g
secretion from VSMCs at 1 h, and it remained elevated for 6 h.
This result indicates that stretch stimulates the secretion of IFN-g
from VSMCs.
3.6 Exogenous IFN-g increases PUMA
protein expression
To investigate the direct effect of IFN-g on PUMA expression in
VSMCs, different concentrations of IFN-g were administrated into
the culture medium for 18 h. As shown in Figure 4B and C, the
effects of IFN-g on PUMA protein expression were dose dependent.
Addition of the IFN-g monoclonal antibody 30 min prior to stretch
significantly blocked the expression of PUMA induced by cyclic
stretch. Furthermore, addition of SP600125 and JNK siRNA attenuated IFN-g-induced expression of PUMA. These findings suggest
that IFN-g enhances PUMA expression by cyclic stretch and that
JNK is required for IFN-g-dependent PUMA expression.
Downloaded from http://cardiovascres.oxfordjournals.org/ at Academia SinicaLife Science Library on March 29, 2012
3.4 Cyclic stretch increases PUMA
promoter activity through IRF-1
Regulation of PUMA by mechanical stretch
185
3.7 Cyclic stretch-induced apoptosis is
mediated by PUMA in VSMCs
As shown in Supplementary material online, Figure S3A and B, cyclic
stretch not only increased the death rate but also decreased the viability of VSMCs as determined via a cell counter and the MTT assay.
These results suggest that cyclic stretch induces VSMC cell death. As
shown in Figure 5A and Supplementary material online, Figure S3C,
apoptosis was assessed by propidium iodide –annexin V double staining and FACS analysis. The percentage of cells stained with annexin V
was elevated following stretch for 18 h and the addition of IFN-g.
Addition of PUMA and JNK siRNA before stretch significantly
decreased the percentage of annexin V-positive cells induced by
stretch. Caspase 3 activity was induced by cyclic stretch in VSMCs
(see Supplementary material online, Figure S3D). Addition of IFN-g
(200 pg/mL) to VSMCs alone without stretch had a similar effect on
caspase 3 activity to that of cyclic stretch. Conversely, addition of
PUMA and JNK siRNA abolished the induction of caspase 3 activity
induced by stretch. A significant increase in TUNEL-positive nuclei
was present after stretch for 18 h and the addition of IFN-g
(Figure 5B; see Supplementary material online, Figure S3E). These
increases in TUNEL-positive nuclei of VSMCs induced by stretch
were significantly reversed by PUMA and JNK siRNA. PUMA siRNA
significantly reduced the expression of PUMA protein (see Supplementary material online, Figure S4). These findings demonstrate that
PUMA mediates stretch-induced apoptosis of VSMCs.
3.8 In vivo AV shunt and aortic banding
increases the expression of aortic PUMA
protein
AV shunt and aortic banding was performed to explore whether
PUMA expression is increased during volume and pressure overload,
respectively, in vivo. As shown in Figure 6, the expression of PUMA
protein in rat aorta increased significantly by 5 days after induction
of the AV shunt and 3 weeks after aortic banding. PUMA levels
Downloaded from http://cardiovascres.oxfordjournals.org/ at Academia SinicaLife Science Library on March 29, 2012
Figure 3 Effects of cyclic stretch on PUMA binding and promoter activity in VSMCs. (A and B) Representative EMSA showing protein binding to
IRF-1 oligonucleotide in nuclear extracts of VSMCs following cyclic stretch for various periods of time and in the absence or presence of JNK, extracellular signal-regulated kinase, p38 inhibitors, or siRNA, and IFN-g, angiotensin II (Ang II), and tumour necrosis factor-a (TNF-a) antibody. Arrows
indicate the mobility of the complex. Similar results were found in another two independent experiments. Cold oligo (Cold) means unlabelled IRF-1
oligonucleotide. (C) Constructs of PUMA promoter. (D) Quantitative analysis of PUMA promoter activity. VSMCs were transiently transfected with
pPUMA-Luc by a gene gun. The luciferase activity in cell lysates was measured and was normalized with Renilla activity (n ¼ 3 per group). *P , 0.01 vs.
control.
186
were 3.8- and 3.6-fold greater, respectively, compared with the sham.
Additionally, the AV shunt and aortic banding induced the apoptosis
of VSMCs (see Supplementary material online, Figure S5). No neointima formation was found. PUMA expression was found in apoptotic
cells. The increase in PUMA expression within the vessel wall after AV
shunt is localized in the media area (see Supplementary material
online, Figure S6). PUMA expression was found in the a-actin-positive
cells, but not in the CD31-positive cells. PUMA expression was found
in the media but not in the intima area.
4. Discussion
In the present study, we demonstrated the following findings: (i) cyclic
stretch up-regulates PUMA expression in human VSMCs; (ii) cyclic
stretch induces IFN-g expression in VSMCs; (iii) IFN-g acts as an
autocrine factor to mediate the increase in PUMA expression
induced by cyclic stretch; (iv) JNK MAP kinase and the IRF-1 transcription factor are involved in the signalling pathway of PUMA induction;
(v) cyclic stretch induces apoptosis of VSMCs via PUMA; and (vi) in
vivo acute haemodynamic overload increases the expression of
aortic PUMA. Furthermore, PUMA was up-regulated in both a timeand load-dependent manner by cyclic stretch. Cyclic stretch of
VSMCs increased the expression of PUMA at the protein and
mRNA levels.
Our findings demonstrate that stretch induces IFN-g, which, in
turn, mediates cyclic stretch-induced expression of PUMA in
VSMCs. Moreover, exogenous addition of IFN-g to non-stretched
VSMCs was sufficient to induce PUMA protein expression and
VSMC apoptosis, similar to that observed in stretched VSMCs.
These results indicate that IFN-g acts as an autocrine mediator in response to cyclic stretch in VSMCs. Our results also demonstrated that
JNK is required for IFN-g-dependent PUMA expression. Mechanical
stretch may have a direct effect on IFN-g release. The effect of
stretch on PUMA expression could be entirely attributed to IFN-g.
Rosner et al. demonstrated that IFN-g primes VSMCs for Fas-induced
apoptosis, and this may be mediated by phosphatidyl inositol 3-kinase
(PI3K), Akt, and Janus-Associated Kinase-2/signal transducer and activator of transcription 1 (Jak-2/Stat1).20 Furthermore, IFN-g-induced
VSMC apoptosis was found to be accomplished by the overexpression of regulated upon activation normal T-cell expressed, and presumably secreted (RANTES).21 These observations are consistent
with our data regarding IFN-g-induced VSMC apoptosis.
We also demonstrated that cyclic stretch up-regulates PUMA
protein expression and that IRF-1 –DNA binding activity required
the phosphorylation of JNK, because IRF-1 binding activity was abolished by both the JNK inhibitor and JNK siRNA. SP600125 inhibited
stretch-induced PUMA expression. N-Acetylcysteine, a free radical
scavenger, had a partial inhibitory effect, whereas SB203580 did not
have any inhibitory effect. It has been previously demonstrated that
the induction of PUMA via palmitate is JNK1 dependent in primary
murine hepatocytes.22 However, p38 MAP kinase also appears to
be an important intracellular signalling pathway that regulates
PUMA. Gomez-Lazaro et al. noted that treatment with a p38 MAPK
inhibitor (SKF86002) potently inhibited PUMA expression induced
by 6-Hydroxydopamine (6-OHDA).23 Furthermore, Zou et al.
reported that insulin-like growth factor-1 effectively protects PC-12
neuronal cells against PUMA-mediated apoptosis enhanced by ER
stress through the p38 MAP kinase pathway.24 Stretch increased
JNK phosphorylation, whereas SP600125 and JNK siRNA significantly
inhibited the phospho-JNK protein induced by stretch. Our data indicate that the JNK MAP kinase pathway, and not the p38 MAP kinase
pathway, is the major pathway involved in the induction of PUMA by
stretch and mediates the increase in binding activity of IRF-1.
In the present study, we also demonstrated that cyclic stretchinduced PUMA expression acts via the IRF-1 pathway. IRF-1 can be effectively induced in most cell types after exposure to cytokine-like
TNF-a, interleukin-1 and IFN-g. IRF-1 binds to specific sequences in
the promoters of genes and controls the transcription of these genes
involved in mediating the immunomodulatory, antiviral and antiproliferative effects. Growing evidence suggests that IRF-1 plays an important role in apoptosis. Gao et al. have demonstrated that in response
to IFN-g, PUMA is transactivated by IRF-1.17 The involvement of
PUMA in stress-induced apoptosis was mostly regulated by the
tumour suppressor, p53.10 Wang et al. noted that the binding of
nuclear p53 to the specific sites within the PUMA promoter is essential
for its ability to induce DNA damage and apoptosis.25 It has been previously demonstrated that PUMA is up-regulated in response to reactive
oxygen species or anoxia, and that p53 is responsible for this
Downloaded from http://cardiovascres.oxfordjournals.org/ at Academia SinicaLife Science Library on March 29, 2012
Figure 4 Effects of human IFN-g on PUMA expression in VSMCs.
(A) Cyclic stretch increased release of human IFN-g from VSMCs
subjected to cyclic stretch by 20% for various periods of time
(n ¼ 4 per group). *P , 0.01 vs. control. (B) Exogenous administration of human IFN-g increases PUMA protein expression. Representative western blots for PUMA in VSMCs after exogenous
administration of IFN-g. (C) Quantitative analysis of PUMA protein
levels. The values from treated VSMCs have been normalized to
matched a-tubulin measurements and then expressed as a ratio of
normalized values to the control cells (n ¼ 4 per group).
*P , 0.01 vs. control. AV shunt and aortic banding were performed
in different animals.
W-P. Cheng et al.
Regulation of PUMA by mechanical stretch
187
Downloaded from http://cardiovascres.oxfordjournals.org/ at Academia SinicaLife Science Library on March 29, 2012
Figure 5 Effect of PUMA on stretch-induced apoptosis in VSMCs. (A) VSMCs were subjected to cyclic stretch for 18 h, with the addition of 200 pg/
mL IFN-g or siRNA before stretch. Quantification of the apoptotic fractions was performed using FACScan. Cells that stain negative for both
annexin V and propidium iodide are alive. Cells that stain positive for annexin V and negative for propidium iodid are undergoing apoptosis. Cells
that stain positive for both annexin V and propidium iodid are in the end stage of apoptosis, called second apoptosis (n ¼ 3). *P , 0.01 vs.
control. **P , 0.01 vs. stretch 18 h. (B) Representative microscopy images of VSMCs after cyclic stretch for 18 h, addition of 200 pg/mL IFN-g, or
siRNA before being stretched then stained with a TUNEL kit. Similar results were observed in another two independent experiments. Arrows indicate
TUNEL-positive cells.
188
up-regulation.26 However, PUMA is also activated by other stimuli and
transcription factors that initiate p53-independent apoptotic responses
to non-genotoxic stimuli, including growth factor/cytokine deprivation,
ischaemia–reperfusion and ER stress.10 Harford et al. demonstrated
that the muscle regulatory transcription factor MyoD-mediated regulation of the apoptotic response to various stimuli correlates with the
level of PUMA induction.27 Wang et al. reported that PUMA is a
direct target of nuclear factor-kB and that it mediates TNF-a-induced
apoptosis.28 PUMA is also activated by oxidative stress in response to
ischaemia–reperfusion to promote p53-independent apoptosis in the
small intestine via the mitochondrial pathway.29 Unlike its well-known
function in p53-dependent apoptosis, the role of PUMA in
p53-independent apoptosis remains to be fully elucidated. Our findings
suggest that an addition of pifithrin (p53 inhibitor) prior to stretch did
not inhibit the expression of PUMA induced by cyclic stretch. Furthermore, we cloned the promoter region of human PUMA and constructed a luciferase reporter plasmid (pGL3-Luc) containing a p53
binding site. When the IRF-1 binding sites were mutated, there was an
increase in the promoter activity induced by stretch that was similar
to that of wild-type. This observation indicates that cyclic stretchinduced expression of PUMA in VSMCs is mediated by IRF-1 and is
p53 independent.
It has been argued that calpain counteracts the mechanically
induced excessive VSMC apoptosis through its p53 degrading
properties.30 We previously demonstrated that cyclic mechanical
stretch-induced VSMC apoptosis is mediated by GADD153 (hallmarks of unfolded protein response, which is up-regulated by the
ER stress).18 PUMA siRNA did not inhibit stretch-mediated
GADD153 expression (see Supplementary material online, Figure
S3B), indicating that the expressions of PUMA and GADD153
during stretch are not related. Morrow et al. noted that cyclic strain
enhances smooth muscle cell apoptosis, at least in part, via the regulation of the Notch receptor.14 Moreover, Su et al. demonstrated that
stretch-induced VSMC apoptosis is regulated via Bcl-2-associated
death factor.31 In the present study, our findings suggest that the ER
stress-related pro-apoptotic protein, PUMA, may be involved in
VSMC apoptosis induced by mechanical stretch. Several studies
have previously indicated that PUMA appears to function as a direct
activator of BAX and BAK (members of Bcl-2 family), which initiate
mitochondrial apoptosis32 and directly antagonize the well-known
anti-apoptotic Bcl-2 family members.33 However, the correlation
between PUMA and apoptosis is still poorly understood. We found
that cyclic stretch enhances caspase 3 activity and that PUMA
siRNA abolishes the induction of caspase 3 by stretch. This finding
indicates that caspase 3 is involved in the PUMA-induced apoptosis
of VSMCs during cyclic stretch.
The increase in PUMA protein expression levels after acute haemodynamic overload may contribute to the regulation of vascular repair,
remodelling and atherosclerosis, all of which involve VSMC proliferation and apoptosis. Toth et al. noted that myocardial PUMA levels
are increased in the ischaemia–reperfusion model and that cardiac
function is significantly better in Puma2/2 mice than in their wildtype littermates;11 however, the role of PUMA in this animal model
is not well understood, and therefore, future research is warranted
to resolve this issue. Although VSMCs play a crucial role in atherosclerosis, it is not known whether prevention of VSMC apoptosis
alone can reduce atherosclerosis. Further studies are needed to
clarify whether therapeutic targeting of PUMA will also prevent apoptosis of inflammatory cells in the vessel wall with potential detrimental
outcome.
In summary, the present study demonstrates, for the first time, that
cyclic stretch enhances the expression of PUMA in cultured human
VSMCs. The stretch-induced increase in PUMA expression is
mediated by IFN-g, JNK MAP kinase, and IRF-1 pathways. Additional
research focusing on these aspects would be of great interest and
value in understanding the role of VSMC apoptosis and its influence
in atherosclerosis.
Supplementary material
Supplementary material is available at Cardiovascular Research online.
Conflict of interest: none declared.
Funding
This study was sponsored in part from Shin Kong Wu Ho-Su Memorial
Hospital, Taipei, Taiwan.
References
1. Koh KK, Han SH, Oh PC, Shin EK, Quon MJ. Combination therapy for treatment or
prevention of atherosclerosis: focus on the lipid-RAAS interaction. Atherosclerosis
2010;209:307 –313.
Downloaded from http://cardiovascres.oxfordjournals.org/ at Academia SinicaLife Science Library on March 29, 2012
Figure 6 Effect of in vivo model of aorta-caval shunt (AV shunt)
and aortic banding on aortic PUMA protein levels. (A) Representative western blots for PUMA in rat aorta after short-term induction
of AV shunt. (B) Representative western blots for PUMA in rat aorta
after aortic banding. (C) Quantitative analysis of PUMA protein
levels. The values have been normalized to a-tubulin measurements
and then expressed as a ratio of normalized values to PUMA protein
in the sham group (n ¼ 5 per group). *P , 0.01, **P , 0.05 vs.
control.
W-P. Cheng et al.
Regulation of PUMA by mechanical stretch
18. Cheng WP, Hung HF, Wang BW, Shyu KG. The molecular regulation of GADD153
in apoptosis of cultured vascular smooth muscle cells by cyclic mechanical stretch.
Cardiovasc Res 2008;77:551 –559.
19. Luo X, He Q, Huang Y, Sheikh MS. Transcriptional up-regulation of PUMA modulates
endoplasmic reticulum calcium pool depletion-induced apoptosis via Bax activation.
Cell Death Differ 2005;12:1310 – 1318.
20. Rosner D, Stoneman V, Littlewood T, McCarthy N, Figg N, Wang Y et al. Interferon-g
induces Fas trafficking and sensitization to apoptosis in vascular smooth muscle cells
via a PI3K- and Akt-dependent mechanism. Am J Pathol 2006;168:2054 –2063.
21. Peng YD, Cheng LX, Zeng QT, Shi Y, Bao XX, Mao XB et al. The role of RANTES as a
crucial downstream cytokine in calcineurin-dependent VSMC apoptosis stimulated by
INFg and CD40L. Cell Biol Int 2010;34:447–453.
22. Cazanave SC, Mott JL, Elmi NA, Bronk SF, Werneburg NW, Akazawa Y et al.
JNK1-dependent PUMA expression contributes to hepatocyte lipoapoptosis. J Biol
Chem 2009;284:26591 –26602.
23. Gomez-Lazaro M, Galindo MF, Concannon CG, Segura MF, Fernandez-Gomez FJ,
Llecha N et al. 6-Hydroxydopamine activates the mitochondrial apoptosis pathway
through p38 MAPK-mediated, p53-independent activation of Bax and PUMA.
J Neurochem 2008;104:1599 –1612.
24. Zou CG, Cao XZ, Zhao YS, Gao SY, Li SD, Liu XY et al. The molecular mechanism of
endoplasmic reticulum stress-induced apoptosis in PC-12 neuronal cells: the protective effect of insulin-like growth factor I. Endocrinology 2009;150:277 – 285.
25. Wang P, Yu J, Zhang L. The nuclear function of p53 is required for PUMAmediated apoptosis induced by DNA damage. Proc Natl Acad Sci USA 2007;104:
4054– 4059.
26. Li YZ, Lu DY, Tan WQ, Wang JX, Li PF. p53 initiates apoptosis by transcriptionally
targeting the antiapoptotic protein ARC. Mol Cell Biol 2008;28:564–574.
27. Harford TJ, Shaltouki A, Weyman CM. Increased expression of the pro-apoptotic
Bcl2 family member PUMA and apoptosis by the muscle regulatory transcription
factor MyoD in response to a variety of stimuli. Apoptosis 2010;15:71–82.
28. Wang P, Qiu W, Dudgeon C, Liu H, Huang C, Zambetti GP et al. PUMA is directly
activated by NF-kB and contributes to TNF-a-induced apoptosis. Cell Death Differ
2009;16:1192 –1202.
29. Wu B, Qiu W, Wang P, Yu H, Cheng T, Zambetti GP et al. p53 independent induction
of PUMA mediates intestinal apoptosis in response to ischaemia –reperfusion. Gut
2007;56:645 –654.
30. Sedding DG, Homann M, Seay U, Tillmanns H, Preissner KT, Braun-Dullaeus RC.
Calpain counteracts mechanosensitive apoptosis of vascular smooth muscle cells
in vitro and in vivo. FASEB J 2008;22:579–589.
31. Su BY, Shontz KM, Flavahan NA, Nowicki PT. The effect of phenotype on mechanical
stretch-induced vascular smooth muscle cell apoptosis. J Vasc Res 2006;43:229–237.
32. Kim H, Tu HC, Ren D, Takeuchi O, Jeffers JR, Zambetti GP et al. Stepwise activation
of BAX and BAK by tBID, BIM, and PUMA initiates mitochondrial apoptosis. Mol Cell
2009;36:487 –499.
33. Letai A. Puma strikes Bax. J Cell Biol 2009;185:189 –191.
Downloaded from http://cardiovascres.oxfordjournals.org/ at Academia SinicaLife Science Library on March 29, 2012
2. Clarke MC, Talib S, Figg NL, Bennett MR. Vascular smooth muscle cell apoptosis
induces interleukin-1-directed inflammation: effects of hyperlipidemia-mediated inhibition of phagocytosis. Circ Res 2010;106:363 –372.
3. Hotamisligil GS. Endoplasmic reticulum stress and atherosclerosis. Nat Med 2010;16:
396 –399.
4. Kim HJ, Kim MY, Jin H, Kim HJ, Kang SS, Kim HJ et al. Peroxisome proliferatoractivated receptor d regulates extracellular matrix and apoptosis of vascular
smooth muscle cells through the activation of transforming growth factor-b1/
Smad3. Circ Res 2009;105:16 –24.
5. Clarke MC, Littlewood TD, Figg N, Maguire JJ, Davenport AP, Goddard M et al.
Chronic apoptosis of vascular smooth muscle cells accelerates atherosclerosis and
promotes calcification and medial degeneration. Circ Res 2008;102:1529 –1538.
6. Sigala F, Savvari P, Liontos M, Sigalas P, Pateras IS, Papalampros A et al. Increased expression of FGF is associated with carotid atherosclerotic plaques instability engaging
the NF-kB pathway. J Cell Mol Med 2010 DOI: 10.1111/j.1582-4934.2010.01082.x
7. Lyon CA, Johnson JL, Williams H, Sala-Newby GB, George SJ. Soluble N-cadherin
overexpression reduces features of atherosclerotic plaque instability. Arterioscler
Thromb Vasc Biol 2009;29:195 –201.
8. Minamino T, Komuro I, Kitakaze M. Endoplasmic reticulum stress as a therapeutic
target in cardiovascular disease. Circ Res 2010;107:1071 –1082.
9. Ren D, Tu HC, Kim H, Wang GX, Bean GR, Takeuchi O et al. BID, BIM, and PUMA
are essential for activation of the BAX- and BAK-dependent cell death program.
Science 2010;330:1390 –1393.
10. Yu J, Zhang L. PUMA, a potent killer with or without p53. Oncogene 2008;27:
S71 –S83.
11. Toth A, Jeffers JR, Nickson P, Min JY, Morgan JP, Zambetti GP et al. Targeted deletion
of Puma attenuates cardiomyocyte death and improves cardiac function during
ischemia-reperfusion. Am J Physiol Heart Circ Physiol 2006;291:H52 –H60.
12. Honsho S, Nishikawa S, Amano K, Zen K, Adachi Y, Kishita E et al. Pressure-mediated
hypertrophy and mechanical stretch induces IL-1 release and subsequent IGF-1 generation to maintain compensative hypertrophy by affecting Akt and JNK pathways. Circ
Res 2009;105:1149 –1158.
13. Thodeti CK, Matthews B, Ravi A, Mammoto A, Ghosh K, Bracha AL et al. TRPV4
channels mediate cyclic strain-induced endothelial cell reorientation through
integrin-to-integrin signaling. Circ Res 2009;104:1123 –1130.
14. Morrow D, Sweeney C, Birney YA, Cummins PM, Walls D, Redmond EM et al. Cyclic
strain inhibits Notch receptor signaling in vascular smooth muscle cells in vitro. Circ
Res 2005;96:567 –575.
15. Nickson P, Toth A, Erhardt P. PUMA is critical for neonatal cardiomyocyte apoptosis
induced by endoplasmic reticulum stress. Cardiovasc Res 2007;73:48 –56.
16. Solé M, Hernandez-Guillamon M, Boada M, Unzeta M. p53 phosphorylation is
involved in vascular cell death induced by the catalytic activity of membrane-bound
SSAO/VAP-1. Biochim Biophys Acta 2008;178:1085 –1094.
17. Gao J, Senthil M, Ren B, Yan J, Xing Q, Yu J et al. IRF-1 transcriptionally upregulates
PUMA, which mediates the mitochondrial apoptotic pathway in IRF-1-induced apoptosis in cancer cells. Cell Death Differ 2010;17:699 – 709.
189