Am J Physiol Heart Circ Physiol 293: H1646–H1653, 2007.
First published June 1, 2007; doi:10.1152/ajpheart.01385.2006.
Involvement of cell surface ATP synthase in flow-induced ATP release
by vascular endothelial cells
Kimiko Yamamoto,1,2 Nobutaka Shimizu,1 Syotaro Obi,1 Shinichiro Kumagaya,1 Yutaka Taketani,3
Akira Kamiya,1 and Joji Ando1
1
Department of Biomedical Engineering, Graduate School of Medicine, University of Tokyo, Tokyo; 2Precursory Research for
Embryonic Science and Technology, Japan Science and Technology Agency, Saitama; and 3Department of Clinical Nutrition,
Institution of Health Biosciences, University of Tokushima Graduate School, Tokushima, Japan
Submitted 20 December 2006; accepted in final form 29 May 2007
(ECs) that line vessels are
constantly exposed to shear stress, a mechanical force generated by flowing blood. ECs recognize shear stress and transmit
the signal into the interior of the cell, where it triggers cell
responses that involve changes in a variety of cell functions
(10). For example, in response to shear stress, ECs increase
production of vasodilators, such as nitric oxide (NO) and
prostacyclin, and cell surface expression of an anti-thrombotic
molecule, thrombomodulin. Many of the alterations of EC
functions by shear stress are accompanied by changes in
expression of related genes (2). These EC responses to shear
stress are thought to play important roles in blood flow-
dependent phenomena, such as angiogenesis, vascular remodeling, and atherogenesis, as well as in the homeostasis of
circulatory functions. At present, however, shear stress signal
transduction is not fully understood.
Ca2⫹ signaling plays an important role in shear stress signal
transduction. Our previous studies demonstrated that a shear
stress-dependent Ca2⫹ influx occurs in human pulmonary arterial ECs (HPAECs) when exposed to flow, and that the
ATP-gated P2X4 ion channel is the major contributor to
flow-induced Ca2⫹ influx (29, 30). Mice lacking P2X4 channels do not show normal EC responses to flow, such as Ca2⫹
influx and subsequent production of NO, and as a result they
exhibit impaired flow-dependent control of vascular tone and
remodeling (31). We have also shown that flow-induced activation of P2X4 requires ATP, which is supplied in the form of
endogenous ATP released by ECs (32). The molecular mechanism of the shear stress-dependent ATP release by ECs,
however, remains unclear.
F1Fo ATP synthase (referred to as ATP synthase below) was
thought to be located exclusively in the mitochondria, but it has
recently been identified on the cell surface of many types of
cells, including human tumor cells, hepatocytes, keratinocytes,
adipocytes, and ECs (7, 9). However, we are only beginning to
understand the functions of cell surface ATP synthase. ATP
generated by cell surface ATP synthase may be transported
into the cells and provide an additional energy source, or it may
act through the P2X/P2Y purinoceptors, thereby activating
Ca2⫹-dependent signaling cascades that modulate cell functions. Since ATP synthase has a proton-transporting function
that regulates intracellular pH, cell surface ATP synthase may
play important roles in cell homeostasis and pathological
phenomena. Cell surface ATP synthase has been found to be
enzymatically active in the synthesis of ATP in human umbilical vein ECs (HUVECs) and to be involved in sustaining cell
proliferation (3, 19, 20), and inhibition of cell surface ATP
synthase activity with the ATP synthase inhibitor angiostatin
reduced proliferation by HUVECs (19, 20). Our previous study
(32) showed that angiostatin suppresses flow-induced ATP
release by HPAECs, and based on that finding, we hypothesized that cell surface ATP synthase plays a role in the process
of ATP release. To test our hypothesis, we investigated
whether HPAECs possess cell surface ATP synthase that is
active in ATP synthesis and, if so, the role of the cell surface
ATP synthase in flow-induced ATP release.
Address for reprint requests and other correspondence: J. Ando, Dept. of
Biomedical Engineering, Graduate School of Medicine, Univ. of Tokyo, 7-3-1
Hongo, Bunkyo-ku, Tokyo 113-0033, Japan (e-mail: [email protected]).
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
THE VASCULAR ENDOTHELIAL CELLS
H1646
0363-6135/07 $8.00 Copyright © 2007 the American Physiological Society
http://www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.246 on June 18, 2017
Yamamoto K, Shimizu N, Obi S, Kumagaya S, Taketani Y,
Kamiya A, Ando J. Involvement of cell surface ATP synthase in
flow-induced ATP release by vascular endothelial cells. Am J Physiol
Heart Circ Physiol 293: H1646–H1653, 2007. First published June 1,
2007; doi:10.1152/ajpheart.01385.2006.—Endothelial cells (ECs) release ATP in response to shear stress, a mechanical force generated by
blood flow, and the ATP released modulates EC functions through
activation of purinoceptors. The molecular mechanism of the shear
stress-induced ATP release, however, has not been fully elucidated. In
this study, we have demonstrated that cell surface ATP synthase is
involved in shear stress-induced ATP release. Immunofluorescence
staining of human pulmonary arterial ECs (HPAECs) showed that cell
surface ATP synthase is distributed in lipid rafts and co-localized with
caveolin-1, a marker protein of caveolae. Immunoprecipitation indicated that cell surface ATP synthase and caveolin-1 are physically
associated. Measurement of the extracellular metabolism of [3H]ADP
confirmed that cell surface ATP synthase is active in ATP generation.
When exposed to shear stress, HPAECs released ATP in a dosedependent manner, and the ATP release was markedly suppressed by
the membrane-impermeable ATP synthase inhibitors angiostatin and
piceatannol and by an anti-ATP synthase antibody. Depletion of
plasma membrane cholesterol with methyl-␤-cyclodextrin (M␤CD)
disrupted lipid rafts and abolished co-localization of ATP synthase
with caveolin-1, which resulted in a marked reduction in shear
stress-induced ATP release. Pretreatment of the cells with cholesterol
prevented these effects of M␤CD. Downregulation of caveolin-1
expression by transfection of caveolin-1 siRNA also markedly suppressed ATP-releasing responses to shear stress. Neither M␤CD,
M␤CD plus cholesterol, nor caveolin-1 siRNA had any effect on the
amount of cell surface ATP synthase. These results suggest that the
localization and targeting of ATP synthase to caveolae/lipid rafts is
critical for shear stress-induced ATP release by HPAECs.
CELL SURFACE ATP SYNTHASE MEDIATES ATP RELEASE
H1647
MATERIALS AND METHODS
AJP-Heart Circ Physiol • VOL
Fig. 1. Surface localization of ATP synthase, caveolin-1, and lipid rafts on human
pulmonary arterial endothelial cells (HPAECs). A: immunofluorescence photomicrograph. Nonpermeabilized HPAECs were fixed and stained with antibodies against
ATP synthase ␤-subunit, caveolin-1, or a lipid raft marker, cholera toxin subunit B.
Co-localization of ATP synthase with caveolin-1 and cholera toxin is visible in the
merged images as yellow areas. The confocal microscope images of single cells at a
higher magnification also show co-localization of ATP synthase with caveolin-1.
Representative images are shown; n ⫽ 22. B: Western blot (WB) analysis. Five
fractions were used in the immunoblot assay: postnuclear supernatant (PNS), cytosol
(Cyt), plasma membrane (PM), caveolae-rich membrane (CM), and noncaveola
membrane (NCM). A 25-␮g protein sample from each fraction was separated by
SDS-PAGE and immunoblotted with antibody against ATP synthase ␤-subunit or
caveolin-1. The CM fraction, but not the NCM fraction, contained both ATP synthase
and caveolin-1, indicating co-localization of ATP synthase and caveolin-1. The results
of 3 independent experiments were similar. C: co-immunoprecipitation of ATP
synthase with caveolin-1. The CM fraction was incubated with [immunoprecipitation
(IP) fraction] or without (control) polyclonal anti-caveolin-1 antibody (caveolin-1 pAb)
overnight at 4°C. The immunocomplexes were assessed by gel electrophoresis and
Western blot (WB) analysis against monoclonal anti-ATP synthase ␤-subunit and
anti-caveolin-1 antibodies. Experiments were repeated 3 times, with similar results.
293 • SEPTEMBER 2007 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.246 on June 18, 2017
Cell culture. HPAECs were obtained from Clonetics and grown on
a 1% gelatin-coated tissue culture flask in M199 supplemented with
15% FBS, 2 mmol/l L-glutamine (Gibco), 50 ␮g/ml heparin, and 30
␮g/ml EC growth factor (Becton Dickinson). The cells used in the
present experiments were in the 7th and 10th passages.
Flow loading experiments. A parallel plate-type apparatus was used
to apply laminar shear stress to the cells, as described previously (30).
Briefly, one side of the flow chamber consisted of a 1% gelatin-coated
glass plate on which the cultured HPAECs rested, and the other side
consisted of a polycarbonate plate. Their flat surfaces were held 200
␮m apart with a Teflon gasket. The chamber was provided with an
entrance and an exit for the fluid, and the entrance was connected to
an upper reservoir with a silicone tube. The exit was open to a lower
reservoir. The flow was driven by gravitational force. The fluid
(Hanks’ balanced salt solution) passed from the upper reservoir
through the flow chamber into the lower reservoir. The flow rate was
monitored with an ultrasonic transit time flow meter (HT107, Transonic Systems) placed at the entrance, and it was controlled by
changing the difference in height between the upper reservoir and the
exit of the flow chamber. The intensity of the shear stress (␶,
dynes/cm2) acting on the EC layer was calculated by using the
formula ␶ ⫽ 6␮Q/a2b, where ␮ is the viscosity of the perfusate
(poise), Q is the flow volume (ml/s), and a and b are the crosssectional dimensions of the flow path (cm).
Quantification of extracellular ATP. The ATP released by the
HPAECs was measured by means of a luciferin-luciferase assay. The
cells were exposed to laminar flow in a parallel plate flow chamber at
37°C. The intensity of shear stress was increased every minute in a
stepwise manner from 0 to 3, 8, and 15 dynes/cm2 by increasing the
volumetric flow rate from 0 to 2.2, 5.8, and 10.9 ml/min, respectively.
Since the perfusion was performed in an open circuit, there was no
ATP accumulation in the system. The effluent that accumulated in the
lower reservoir was collected every minute, and 100 ␮l of the fluid
were applied to the ATP assay system (Toyo, Tokyo, Japan). The ATP
assay mixture (luciferase, D-luciferin, and bovine serum albumin) was
injected into each sample, and its relative light intensity was recorded
for 10 s in a Lumat LB 9501 luminometer (Berthold) at room
temperature. A calibration curve for ATP concentrations was obtained
for each experiment by using the same batch of luciferin-luciferase
reagents. Since the perfusion had a dilution effect on ATP, the ATP
concentrations were calculated accordingly.
Measurement of the extracellular metabolism of 3H-labeled adenine nucleotide. HPAECs were seeded onto gelatin-coated 24-well
tissue culture plates at a density of 5 ⫻ 104 cells per well. Twenty-four
hours later, confluent HPAECs were rinsed twice with RPMI 1640
medium and then incubated at 37°C with gentle orbital rotation in a
starting volume of 0.25 ml of RPMI 1640 containing 50 ␮M [3H]ADP
as initial substrates. Aliquots of the mixture were periodically applied
to an Alugram TLC sheet (⬃5 ⫻ 104 dpm per spot), and adenine
nucleotides and adenosine were separated with an appropriate solvent
system as described previously (34). Radioactive areas that co-migrated with respective nucleotide/nucleoside standards were scraped
into scintillation vials, extracted from silica with 0.1 N HCl, and
quantified by scintillation counting with a Wallac-1409 ␤-spectrometer.
Immunohistochemistry. HPAECs on the coverslip were fixed with
4% paraformaldehyde (Sigma) and maintained in 1% normal bovine
serum albumin (Sigma) to block nonspecific protein-binding sites.
The cells were incubated with antibodies (5 ␮g/ml) against the
␤-subunit of ATP synthase (Molecular Probes), caveolin-1 (Transduction Laboratories), or Alexa Fluor 594-conjugated cholera toxin
subunit B (Molecular Probes). After a washing, they were incubated
with Alexa Fluor 488 goat anti-mouse IgG and Alexa Fluor 594 goat
anti-rabbit IgG (Molecular Probes) at a dilution of 1:500. Stained cells
were photographed through a confocal fluorescence microscope
H1648
CELL SURFACE ATP SYNTHASE MEDIATES ATP RELEASE
expression in HPAECs as described previously (13). A DNA fragment
flanked by the BamH I and HindIII sites containing the sense target
sequence corresponding to bases 167–186 from the open reading
frame of the human caveolin-1 (GenBank accession no. NM001753)
(5⬘-CTAAACACCTCAACGATGA-3⬘), the hairpin loop sequence
(5⬘-CTGTGAAGCCACAGATGGG-3⬘), and the antisense target sequence (5⬘-TCATCGTTGAGGTGTTTAG-3⬘) were synthesized and
inserted between the human U6 promoter and the terminator sequences of pBAsi-hU6 (Takara) to generate a stem-loop type of
siRNA in transfected cells. The randomized sequence 5⬘-TAACATGAACCACGA CTAC-3⬘ was used to construct the vector for a
negative control. HPAECs were treated with the construct using
Lipofectamine 2000 (Invitrogen) as the transfection reagent, and
experiments were conducted 48 h after transfection.
Statistical analysis. All results are expressed as means ⫾ SD.
Statistical significance was evaluated by an ANOVA and a Bonferonni adjustment applied to the results of a t-test performed with SPSS
software (SPSS). P values ⬍0.01 were regarded as statistically significant.
RESULTS
Cell surface ATP synthase is distributed in lipid rafts and
co-localized with caveolin-1. HPAECs were immunostained
with anti-ATP synthase ␤-subunit antibody to determine
whether they possess cell surface ATP synthase. It should be
noted that the cells were not permeabilized for immunofluorescence. Photomicrographs showed that the ATP synthase
␤-subunit was distributed on the surface of the HPAECs and
was concentrated at localized regions at the cell edges (Fig.
1A). Similar findings were obtained in the cells immunostained
with anti-ATP synthase ␣-subunit antibody (data not shown).
Since the punctate staining pattern of ATP synthase resembles
that of caveolae/lipid rafts, HPAECs were immunostained with
an antibody against cholera toxin subunit B, a gangliosidebinding protein used as lipid raft landmarker, or with an
Fig. 2. Extracellular ATP generation by cell surface ATP
synthase. A: TLC assay. HPAECs were incubated with
[3H]ADP, and, after incubation for the periods indicated,
aliquots of medium were spotted on Alugram TLC sheets
and analyzed by TLC. Autoradiography showed [3H]ATP
formation on the cell surface (control). [3H]ATP formation
was markedly suppressed by angiostatin (1 ␮M), a membrane-impermeable ATP synthase inhibitor, indicating
that cell surface ATP synthase is active in ATP synthesis.
The first lane shows the radiochemical purity of [3H]ADP
after a 10-min incubation in the absence of HPAECs.
B: quantitative densitometry analysis. Graph shows relative amounts of [3H]ATP in the absence (open bars) and
presence (solid bars) of angiostatin. All values are
means ⫾ SD of 4 separate experiments. *P ⬍ 0.01 vs.
control. C: radioactivity inside or outside the cells. Aliquots of the medium or cell lysate were collected during
incubation of the cells with [3H]ADP, and their radioactivity (dpm) was quantified with a scintillation counter. No
3
H radioactivity was detected in the cell lysate, indicating
that [3H]ATP was produced on the cell surface of
HPAECs.
AJP-Heart Circ Physiol • VOL
293 • SEPTEMBER 2007 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.246 on June 18, 2017
(Leica), and all images were imported into Adobe Photoshop as
JPEGs for contrast manipulation and figure assembly.
Purification of caveola fractions. Caveola membranes were prepared by the detergent-free method as described previously (24).
Briefly, cells were homogenized with the Teflon homogenizer
(Wheaton) and centrifuged at 1,000 g for 10 min. The postnuclear
supernatant fraction (PNS) was collected, and the plasma membrane
fraction was isolated by 30% Percoll-gradient fractionation (84,000 g
for 30 min) and sonicated six times for 5 s each. Caveola membranes
were separated by the OptiPrep density gradients, a continuous 23%
to 10% gradient followed by centrifugation (52,000 g for 90 min)
against a discontinuous gradient to concentrate the lightest material.
Proteins from each fraction were analyzed by Western blot.
Immunoprecipitation and Western blot analysis. Immunoprecipitation of caveolin-associated proteins was performed as previously
described (27). Briefly, equal amounts of protein in the caveolaeenriched fraction were diluted with RIPA buffer and precleared by
incubation for 2 h with protein G-Sepharose beads (Amersham).
Protein G-Sepharose beads were incubated with polyclonal anticaveolin-1 antibody (BD Transduction Laboratories) for 2 h, and,
after the beads were washed with RIPA buffer, precleared supernatants were incubated with the beads overnight at 4°C with gentle
mixing. The immune complexes were collected by centrifugation at
2,000 g for 5 min, washed, and eluted in SDS-PAGE sample buffer.
Western blot analysis was performed as previously described (32).
Briefly, cells were dissolved in RIPA buffer and centrifuged at 26,000
g for 30 min. A 30-␮g sample of total cell lysate was resolved in an
SDS-PAGE gel and then transferred to an Immobilon membrane
(Millipore) and incubated for 1 h. The membrane was probed with the
anti-ATP synthase ␤-subunit antibody (4 ␮g/ml) or anti-caveolin-1
antibody (3 ␮g/ml) and then incubated with anti-rabbit IgG horseradish peroxidase-conjugated antibody. The same membrane was reprobed with monoclonal anti-␤-actin antibody (Abcam). The densitometry values of the ATP synthase blots and caveolin-1 blots were
standardized to those of the ␤-actin blots.
Small interfering RNA preparation and transfection. Targeted
small interfering RNA (siRNA) was used to knock down caveolin-1
CELL SURFACE ATP SYNTHASE MEDIATES ATP RELEASE
AJP-Heart Circ Physiol • VOL
Fig. 3. Extracellular ATP release in response to flow. The intensity of shear
stress was increased every minute in a stepwise manner, and the effluent that
accumulated every minute was applied to ATP measurement. ATP concentrations in the perfusate (Hanks’ balanced salt solution) were measured by
luminometry. When HPAECs were exposed to shear stress, ATP release
increased in a dose-dependent manner (control). Angiostatin (1 ␮M), piceatannol (20 ␮M), or anti-ATP synthase ␤-subunit antibody (250 ng/ml) significantly suppressed the shear stress-dependent ATP release. All values are
means ⫾ SD of 5 different experiments. *P ⬍ 0.01 vs. control. Inset:
intracellular ATP concentration. Angiostatin, piceatannol, or anti-ATP synthase ␤-subunit antibody did not change the intracellular ATP concentration.
Exposure to shear stress (15 dynes/cm2) for 3 min had no effect on the
intracellular ATP concentration. Static, cells cultured under static conditions.
Results are shown as means ⫾ SD of 5 separate samples.
outline-like accumulation of caveolin-1 was visible some distance away from the cell edge. These M␤CD effects were
prevented by cholesterol pretreatment. In HPAECs incubated
with cholesterol for 24 h before M␤CD treatment, lipid rafts
appeared, and both ATP synthase and caveolin-1 were distributed in the same localized regions at the cell edge. Western
blotting revealed that neither M␤CD nor M␤CD plus cholesterol changed the amount of ATP synthase or caveolin-1 in
either the cells as a whole or their plasma membrane (Fig. 4B).
M␤CD markedly suppressed the flow-induced ATP release by
HPAECs, and it was almost completely prevented by cholesterol pretreatment (Fig. 4C). Co-immunoprecipitation showed
that M␤CD abolished the physical association between ATP
synthase and caveolin-1, whereas shear stress had no effect
(Fig. 4D). These results suggest that the localization of ATP
synthase in caveolae/lipid rafts and its association with caveolin-1 are required for shear stress-dependent ATP release by
HPAECs.
Downregulation of caveolin-1 expression inhibits flow-induced ATP release. To further examine the role of caveolin-1
in the cell surface ATP synthase-mediated ATP release, we
used siRNA to specifically knock down expression of caveolin-1. Transfection of HPAECs with caveolin-1 siRNA resulted
in a significant and specific reduction in caveolin-1 protein
expression, as judged by both immunofluorescent staining and
Western blotting (Fig. 5, A and B). The cells transfected with
293 • SEPTEMBER 2007 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.246 on June 18, 2017
antibody against caveolin-1, a key constituent protein of caveolae. The photomicrographs showed that ATP synthase was
co-localized with caveolin-1 and cholera toxin, suggesting that
cell surface ATP synthase is localized in caveolae/lipid rafts.
To confirm the localization of ATP synthase in caveolae/
lipid rafts, detergent-resistant caveolae-rich membranes were
isolated and analyzed by immunoblotting with anti-ATP synthase ␤-subunit antibody and anti-caveolin-1 antibody. The
caveolae-rich membrane fraction contained both ATP synthase
␤-subunit and caveolin-1, whereas the noncaveola fraction
contained neither (Fig. 1B). This means that ATP synthase is
localized in the caveolae.
To determine whether ATP synthase in HPAECs is physically associated with caveolin-1, we performed co-immunoprecipitation experiments. The caveolae-rich fraction was incubated with a polyclonal anti-caveolin-1 antibody to precipitate caveolin-1 and its associated proteins. The immune
complexes were collected with protein G beads and analyzed
by immunoblotting against monoclonal anti-ATP synthase and
anti-caveolin-1 antibodies. These results indicated that the
ATP synthase in the caveolae-rich fraction was co-precipitated
with caveolin-1, suggesting that cell surface ATP synthase and
caveolin-1 are physically associated (Fig. 1C).
Cell surface ATP synthase is active in extracellular ATP
generation. To explore the function of cell surface ATP synthase, HPAECs were incubated with [3H]ADP and analyzed
for subsequent ATP production by TLC. ATP generation was
detected within 1 min after incubation, peaked at 5 min, and
then declined (Fig. 2, A and B). Angiostatin, a membraneimpermeable ATP synthase inhibitor, significantly inhibited
the production of [3H]ATP. No labeled product was detected
within the cell pellets, confirming that the ATP had been
synthesized on the cell surface (Fig. 2C). These findings
indicate that the cell surface ATP synthase of HPAECs functions to synthesize ATP.
Cell surface ATP synthase is involved in flow-induced ATP
release. HPAECs were exposed to flow, and the concentration
of ATP in the perfusate was measured by a bioluminescence
assay. The ATP concentration increased in response to flow
and became increasingly greater with the intensity of the shear
stress (Fig. 3). ATP synthase inhibitors, such as angiostatin and
piceatannol (36), or anti-ATP synthase ␤-subunit antibody
markedly suppressed the shear stress-dependent ATP release
by the HPAECs, but none of these inhibitors had any effect on
the intracellular ATP concentration, suggesting that none of
these inhibitors crossed the cell membrane and inhibited mitochondrial ATP synthase. These findings indicate that cell
surface ATP synthase plays a critical role in the shear stressdependent ATP release by HPAECs.
Depletion of plasma membrane cholesterol inhibits flowinduced ATP release. Next, we investigated whether the localization of ATP synthase in caveolae/lipid rafts is critical for
flow-induced ATP release to occur. After treating HPAECs
with methyl-␤-cyclodextrin (M␤CD), which disrupts lipid rafts
by depleting plasma membrane cholesterol, we examined the
cells for changes in the distribution of ATP synthase, cholera
toxin, and caveolin-1 by confocal immunofluorescence microscopy. Treatment with M␤CD abolished the number and size of
the lipid rafts and resulted in redistribution of ATP synthase
and caveolin-1 (Fig. 4A). ATP synthase and caveolin-1 were
distributed throughout the cells in a diffuse pattern, and an
H1649
H1650
CELL SURFACE ATP SYNTHASE MEDIATES ATP RELEASE
caveolin-1 siRNA showed a clear loss of staining intensity
compared with control cells (subjected to transfection conditions alone). Western blotting showed a marked reduction of
caveolin-1, but not of ATP synthase, in both the PNS and
plasma membrane fractions, and the level of caveolin-1 protein
expression declined to nearly 15% of the control at 48 h
posttransfection (Fig. 5C). Downregulation of caveolin-1
markedly suppressed the shear stress-dependent ATP release
AJP-Heart Circ Physiol • VOL
by HPAECs (Fig. 5D). These findings suggest that the association of ATP synthase with caveolin-1 is critical for flowinduced ATP release by HPAECs.
DISCUSSION
The results of this study demonstrated that HPAECs express
cell surface ATP synthase that is active in extracellular ATP
293 • SEPTEMBER 2007 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.246 on June 18, 2017
Fig. 4. Effect of membrane cholesterol depletion on the localization of ATP synthase, lipid rafts, and caveolin-1 and on flow-induced ATP release.
A: immunofluorescence photomicrograph. HPAECs were treated with 10 mM methyl-␤-cyclodextrin (M␤CD) or pretreated with 1.3 mM cholesterol and then
with 10 mM M␤CD. Cells were then washed in PBS, fixed, and immunostained with antibodies against ATP synthase ␤-subunit, cholera toxin subunit B, or
caveolin-1. M␤CD disrupted the lipid rafts of the HPAECs and abolished the co-localization of ATP synthase with caveolin-1. Cholesterol pretreatment prevented
these effects of M␤CD. Representative images are shown; n ⫽ 16. B: Western blot analysis. Neither M␤CD nor M␤CD plus cholesterol changed the amount
of ATP synthase or caveolin-1 proteins in the PNS fraction or in the PM fraction. The cytoskeletal protein ␤-actin serves as a protein loading control.
C: flow-induced ATP release. The intensity of shear stress was increased every minute in a stepwise manner, and the effluent that accumulated every minute was
applied to ATP measurement. M␤CD markedly suppressed the shear stress-dependent ATP release, but the suppression was prevented by cholesterol
pretreatment. Results are shown as means ⫾ SD of 6 separate samples. *P ⬍ 0.01 vs. control. D: co-immunoprecipitation of ATP synthase and caveolin-1.
HPAECs were cultured under static conditions (control), treated with 10 mM M␤CD, or exposed to shear stress (15 dynes/cm2) for 3 min. Three fractions were
used in the immunoblot assay: PNS, PM, and CM (for abbreviations, see legend to Fig. 1). M␤CD abolished the association between ATP synthase and
caveolin-1, but shear stress had no effect. Experiments were repeated 3 times, with similar results.
CELL SURFACE ATP SYNTHASE MEDIATES ATP RELEASE
H1651
synthesis. Inhibition of cell surface ATP synthase with angiostatin, piceatannol, or anti-ATP synthase antibody markedly
reduced flow-induced ATP release, and since none of these
inhibitors crosses the plasma membrane, they have no effect on
mitochondrial ATP synthase activity. Thus our results suggest
that cell surface ATP synthase is involved in flow-induced
ATP release by HPAECs. The cell surface ATP synthase was
found to be localized in the caveolae/lipid rafts of HPAECs,
consistent with the findings obtained in HUVECs, hepatocytes,
and adipocytes (4, 17, 21, 26). Immunoprecipitation results
indicated that the cell surface ATP synthase and caveolin-1 are
physically associated. Depletion of plasma membrane cholesterol with M␤CD disrupted lipid rafts and abolished the association of ATP synthase with caveolin-1, which led to marked
suppression of flow-induced ATP release. Downregulation of
expression of caveolin-1 by siRNA also markedly inhibited
flow-induced ATP release. Taken together, these findings show
that the localization of ATP synthase in caveolae/lipid rafts and
its association with caveolin-1 are critical for shear stressinduced ATP release by HPAECs.
A variety of types of cells, including epithelial cells, fibroblasts, and ECs, release ATP in response to mechanical stresses
AJP-Heart Circ Physiol • VOL
(5, 12, 14, 28). However, the molecular mechanisms of mechanical stress-induced ATP release have not been fully elucidated (15). One possibility is that cells release ATP through
a process called exocytosis in which intracellular vesicles
containing ATP fuse with plasma membranes and release ATP
into the extracellular space. Bodin and Burnstock (6) demonstrated that shear stress-induced ATP release by HUVECs is
vesicular, because ATP release was significantly reduced by
monensin, an inhibitor of vesicle formation at the level of the
Golgi apparatus, and by N-ethylmaleimide, an inhibitor of
vesicle fusion to the plasma membrane. Another possible
mechanism is ATP-binding cassette (ABC) protein-mediated
ATP export across the plasma membrane. ABC proteins, such
as the cystic fibrosis transmembrane conductance regulator (a
cAMP-activated ATP-dependent Cl⫺ channel), P-glycoprotein
(a channel transporting both Cl⫺ and ATP out of the cell), and
the ATP-sensitive K⫹ channel, have been identified in ECs (8,
16, 25). Although the present study did not address these two
mechanisms, vesicular exocytosis or ABC transporters may be
involved in flow-induced ATP release by HPAECs. The results
of the present study clearly demonstrate a novel mechanism for
flow-induced ATP release mediated by cell surface ATP syn-
293 • SEPTEMBER 2007 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.246 on June 18, 2017
Fig. 5. Effect of caveolin-1 knockdown on the localization of ATP synthase and caveolin-1 and on flow-induced ATP release. A: immunofluorescence
photomicrograph. Nonpermeabilized HPAECs transfected with caveolin-1 siRNA were fixed and stained with antibodies against ATP synthase ␤-subunit or
caveolin-1. Caveolin-1 siRNA markedly decreased cell surface caveolin-1. Transfection of disordered siRNA with a scrambled nucleotide sequence had no effect
on caveolin-1 expression (data not shown). B: Western blot analysis. Control cells were treated with the transfection reagent alone without siRNA (control).
Caveolin-1 siRNA transfection significantly decreased caveolin-1 protein in the PNS and PM fractions. The cytoskeletal protein ␤-actin serves as a protein
loading control. C: quantitative densitometry analysis. Results are shown as means ⫾ SD of 6 separate samples. *P ⬍ 0.01 vs. control. D: flow-induced ATP
release. The intensity of shear stress was increased every minute in a stepwise manner, and the effluent that accumulated every minute was applied to ATP
measurement. Caveolin-1 siRNA transfection markedly suppressed the shear stress-dependent ATP release. Experiments were performed 48 h after transfection.
Results are shown as means ⫾ SD of 6 separate samples. *P ⬍ 0.01 vs. control.
H1652
CELL SURFACE ATP SYNTHASE MEDIATES ATP RELEASE
ACKNOWLEDGMENTS
We thank Yuko Sawada for technical assistance.
AJP-Heart Circ Physiol • VOL
GRANTS
This study was supported in part by Grants-in-Aid for Scientific Research
on Priority Areas and from the Ministry of Education, Culture, Sports, Science
and Technology and by a research grant for cardiovascular diseases from the
Japanese Ministry of Health, Labor and Welfare.
REFERENCES
1. Anderson RG. Plasmalemmal caveolae and GPI-anchored membrane
proteins. Curr Opin Cell Biol 5: 647– 652, 1993.
2. Ando J, Korenaga R, Kamiya A. Flow-induced endothelial gene regulation. In: Mechanical Forces and the Endothelium, edited by Lelkes PI.
London: Harwood Academic, 1999, p. 111–126.
3. Arakaki N, Nagao T, Niki R, Toyofuku A, Tanaka H, Kuramoto Y,
Emoto Y, Shibata H, Magota K, Higuti T. Possible role of cell surface
H⫹-ATP synthase in the extracellular ATP synthesis and proliferation of
human umbilical vein endothelial cells. Mol Cancer Res 1: 931–939, 2003.
4. Bae TJ, Kim MS, Kim JW, Kim BW, Choo HJ, Lee JW, Kim KB, Lee
CS, Kim JH, Chang SY, Kang CY, Lee SW, Ko YG. Lipid raft
proteome reveals ATP synthase complex in the cell surface. Proteomics 4:
3536 –3548, 2004.
5. Bodin P, Bailey D, Burnstock G. Increased flow-induced ATP release
from isolated vascular endothelial cells but not smooth muscle cells. Br J
Pharmacol 103: 1203–1205, 1991.
6. Bodin P, Burnstock G. Evidence that release of adenosine triphosphate
from endothelial cells during increased shear stress is vesicular. J Cardiovasc Pharmacol 38: 900 –908, 2001.
7. Chi SL, Pizzo SV. Cell surface F1Fo ATP synthase: a new paradigm? Ann
Med 38: 429 – 438, 2006.
8. Cordon-Cardo C, O’Brien JP, Casals D, Rittman-Grauer L, Biedler
JL, Melamed MR, Bertino JR. Multidrug-resistance gene (P-glycoprotein) is expressed by endothelial cells at blood-brain barrier sites. Proc
Natl Acad Sci USA 86: 695– 698, 1989.
9. Das B, Mondragon MO, Sadeghian M, Hatcher VB, Norin AJ. A novel
ligand in lymphocyte-mediated cytotoxicity: expression of the beta subunit of H⫹ transporting ATP synthase on the surface of tumor cell lines.
J Exp Med 180: 273–281, 1994.
10. Davies PF. Flow-mediated endothelial mechanotransduction. Physiol Rev
75: 519 –560, 1995.
11. Ferraro JT, Daneshmand M, Bizios R, Rizzo V. Depletion of plasma
membrane cholesterol dampens hydrostatic pressure and shear stressinduced mechanotransduction pathways in osteoblast cultures. Am J
Physiol Cell Physiol 286: C831–C839, 2004.
12. Furuya K, Sokabe M, Furuya S. Characteristics of subepithelial fibroblasts as a mechano-sensor in the intestine: cell-shape-dependent ATP
release and P2Y1 signaling. J Cell Sci 118: 3289 –3304, 2005.
13. Gonzalez E, Nagiel A, Lin AJ, Golan DE, Michel T. Small interfering
RNA-mediated down-regulation of caveolin-1 differentially modulates
signaling pathways in endothelial cells. J Biol Chem 279: 40659 – 40669,
2004.
14. Grierson JP, Meldolesi J. Shear stress-induced [Ca2⫹]i transients and
oscillations in mouse fibroblasts are mediated by endogenously released
ATP. J Biol Chem 270: 4451– 4456, 1995.
15. Hamill OP, Martinac B. Molecular basis of mechanotransduction in
living cells. Physiol Rev 81: 685–740, 2001.
16. Katnik C, Adams DJ. An ATP-sensitive potassium conductance in rabbit
arterial endothelial cells. J Physiol 485: 595– 606, 1995.
17. Kim BW, Choo HJ, Lee JW, Kim JH, Ko YG. Extracellular ATP is
generated by ATP synthase complex in adipocyte lipid rafts. Exp Mol Med
36: 476 – 485, 2004.
18. Lungu AO, Jin ZG, Yamawaki H, Tanimoto T, Wong C, Berk BC.
Cyclosporin A inhibits flow-mediated activation of endothelial nitricoxide synthase by altering cholesterol content in caveolae. J Biol Chem
279: 48794 – 48800, 2004.
19. Moser TL, Kenan DJ, Ashley TA, Roy JA, Goodman MD, Misra UK,
Cheek DJ, Pizzo SV. Endothelial cell surface F1–F0 ATP synthase is
active in ATP synthesis and is inhibited by angiostatin. Proc Natl Acad Sci
USA 98: 6656 – 6661, 2001.
20. Moser TL, Stack MS, Asplin I, Enghild JJ, Hojrup P, Everitt L,
Hubchak S, Schnaper HW, Pizzo SV. Angiostatin binds ATP synthase
on the surface of human endothelial cells. Proc Natl Acad Sci USA 96:
2811–2816, 1999.
293 • SEPTEMBER 2007 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.246 on June 18, 2017
thase. It remains unknown, however, whether cell surface ATP
synthase interacts with ATP-releasing mechanisms via vesicular transport or ABC proteins.
Vascular ECs express a variety of ecto-enzymes that contribute to the metabolism and/or interconversion of extracellular adenine nucleotides. Thus the ATP concentration of the
perfusate is thought to represent the net change that occurs as
a result of the release of ATP, conversion of nucleotides to
ATP, and degradation of ATP, not just the release of ATP. Our
flow experiments were done in an open circulation system, and
the perfusate was not recirculated. Since flow quickly removes
ATP from the cell surface, it seems difficult to evaluate the
degree to which ecto-enzymes degrade endogenously released
ATP. Yegutkin et al. (33) demonstrated that stimulation of
HUVECs by shear stress induced concomitant release of endogenous ATP and soluble ecto-enzymes that degrade both
ATP (ATPase) and AMP (5⬘-nucleotidase). Thus soluble ectoenzymes that were released into the effluent may have affected
the ATP concentration in the effluent. Although the present
study did not extend to the kinetics of ATP generation and
degradation, quantitative understanding of the kinetics would
provide better insight into the role of cell surface ATP synthase
in shear stress-induced ATP release by ECs.
Lipid rafts are specialized membrane microdomains that are
very rich in glycosphingolipids and cholesterol, unlike other
membrane regions, and caveolae are a subset of lipid rafts
characterized by the presence of the protein caveolin-1 and
plasma membrane invaginations (1). Caveolae/lipid rafts are
endowed with signaling functions because various receptors
and mediators for signal transduction are clustered in them
(23). There is accumulating evidence in support of the concept
that caveolae/lipid rafts have the general function of cell
surface mechanotransduction sites within the plasma membrane. Caveolae/lipid rafts must be intact for shear stress signal
transduction to occur (35). Disruption of caveolae/lipid rafts by
M␤CD blocks the activation of extracellular signal-regulated
kinase (ERK) induced by shear stress in bovine aortic ECs
(BAECs) (22). Cyclosporine A decreases the cholesterol content of the plasma membrane and also suppresses shear stressmediated activation of endothelial NO synthase (eNOS) in
BAECs (18). Depletion of plasma membrane cholesterol in
human fetal osteoblasts by M␤CD significantly dampens hydrostatic pressure- and shear stress-induced mechanotransduction, including protein tyrosine phosphorylation, activation of
ERK, and enhanced expression of c-fos (11). In the present
study, we found that M␤CD markedly inhibited shear stressdependent ATP release by HPAECs. The inhibitory effect was
prevented by the pretreatment of the cells with cholesterol.
These findings suggest that intact caveolae/lipid rafts are required for ATP synthase-mediated ATP release in response to
shear stress. Thus the cell surface ATP synthase localized at
caveolae/lipid rafts may be involved in shear stress signal
transduction through activation of P2X/P2Y purinoceptors
rather than directly contributing to ATP release by producing
ATP on the cell surface. Further study is needed to clarify the
true role of cell surface ATP synthase in shear stress-induced
ATP release by ECs.
CELL SURFACE ATP SYNTHASE MEDIATES ATP RELEASE
AJP-Heart Circ Physiol • VOL
30. Yamamoto K, Korenaga R, Kamiya A, Qi Z, Sokabe M, Ando J.
P2X(4) receptors mediate ATP-induced calcium influx in human vascular
endothelial cells. Am J Physiol Heart Circ Physiol 279: H285–H292,
2000.
31. Yamamoto K, Sokabe T, Matsumoto T, Yoshimura K, Shibata M,
Ohura N, Fukuda T, Sato T, Sekine K, Kato S, Isshiki M, Fujita T,
Kobayashi M, Kawamura K, Masuda H, Kamiya A, Ando J. Impaired
flow-dependent control of vascular tone and remodeling in P2X4-deficient
mice. Nat Med 12: 133–137, 2006.
32. Yamamoto K, Sokabe T, Ohura N, Nakatsuka H, Kamiya A, Ando J.
Endogenously released ATP mediates shear stress-induced Ca2⫹ influx
into pulmonary artery endothelial cells. Am J Physiol Heart Circ Physiol
285: H793–H803, 2003.
33. Yegutkin G, Bodin P, Burnstock G. Effect of shear stress on the release
of soluble ecto-enzymes ATPase and 5⬘-nucleotidase along with endogenous ATP from vascular endothelial cells. Br J Pharmacol 129: 921–926,
2000.
34. Yegutkin GG, Henttinen T, Jalkanen S. Extracellular ATP formation on
vascular endothelial cells is mediated by ecto-nucleotide kinase activities
via phosphotransfer reactions. FASEB J 15: 251–260, 2001.
35. Yu J, Bergaya S, Murata T, Alp IF, Bauer MP, Lin MI, Drab M,
Kurzchalia TV, Stan RV, Sessa WC. Direct evidence for the role of
caveolin-1 and caveolae in mechanotransduction and remodeling of blood
vessels. J Clin Invest 116: 1284 –1291, 2006.
36. Zheng J, Ramirez VD. Piceatannol, a stilbene phytochemical, inhibits
mitochondrial F0F1-ATPase activity by targeting the F1 complex. Biochem Biophys Res Commun 261: 499 –503, 1999.
293 • SEPTEMBER 2007 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.246 on June 18, 2017
21. Moser TL, Stack MS, Wahl ML, Pizzo SV. The mechanism of action of
angiostatin: can you teach an old dog new tricks? Thromb Haemost 87:
394 – 401, 2002.
22. Park H, Go YM, St John PL, Maland MC, Lisanti MP, Abrahamson
DR, Jo H. Plasma membrane cholesterol is a key molecule in shear
stress-dependent activation of extracellular signal-regulated kinase. J Biol
Chem 273: 32304 –32311, 1998.
23. Shaul PW, Anderson RG. Role of plasmalemmal caveolae in signal
transduction. Am J Physiol Lung Cell Mol Physiol 275: L843–L851, 1998.
24. Smart EJ, Ying YS, Mineo C, Anderson RG. A detergent-free method
for purifying caveolae membrane from tissue culture cells. Proc Natl Acad
Sci USA 92: 10104 –10108, 1995.
25. Tousson A, Van Tine BA, Naren AP, Shaw GM, Schwiebert LM.
Characterization of CFTR expression and chloride channel activity in
human endothelia. Am J Physiol Cell Physiol 275: C1555–C1564, 1998.
26. Wang T, Chen Z, Wang X, Shyy JY, Zhu Y. Cholesterol loading increases
the translocation of ATP synthase beta chain into membrane caveolae in
vascular endothelial cells. Biochim Biophys Acta 1761: 1182–1190, 2006.
27. Wang XL, Ye D, Peterson TE, Cao S, Shah VH, Katusic ZS, Sieck GC,
Lee HC. Caveolae targeting and regulation of large conductance Ca2⫹activated K⫹ channels in vascular endothelial cells. J Biol Chem 280:
11656 –11664, 2005.
28. Watt WC, Lazarowski ER, Boucher RC. Cystic fibrosis transmembrane
regulator-independent release of ATP. Its implications for the regulation of
P2Y2 receptors in airway epithelia. J Biol Chem 273: 14053–14058, 1998.
29. Yamamoto K, Korenaga R, Kamiya A, Ando J. Fluid shear stress
activates Ca2⫹ influx into human endothelial cells via P2X4 purinoceptors.
Circ Res 87: 385–391, 2000.
H1653