 |
INTRODUCTION |
Angiotensin II (AngII)1
has pleiotropic acute and chronic effects on many cell types and plays
an important role in the pathophysiology of cardiovascular diseases,
including hypertension, atherosclerosis, and heart failure (1). A
significant body of evidence supports a role for the intracellular
production of reactive oxygen species (ROS) in the signal transduction
of AngII-dependent cellular responses via activation of
redox-sensitive signaling cascades (2-4). Recent studies (5) suggest
that a significant source of intracellular ROS in cardiovascular cells
is a phagocyte-type NADPH oxidase. AngII activates NADPH oxidases in
vascular smooth muscle (5, 6), fibroblasts (7), endothelial cells (EC)
(8, 9), and cardiomyocytes (10). Furthermore, NADPH oxidase activation and increased ROS production are implicated in AngII-stimulated vascular smooth muscle hypertrophy (3, 5) and in
AngII-dependent hypertension and the associated endothelial
dysfunction (11-13). The mechanisms through which AngII activates
NADPH oxidase are therefore of interest.
To date, NADPH oxidase has been best characterized in neutrophils where
it is integral to nonspecific host defense. The neutrophil oxidase
comprises a membrane-bound cytochrome b558
composed of a
p22phox-gp91phox
heterodimer and several cytosolic subunits
(p47phox, p40phox,
p67phox, and Rac) (14). The enzyme is normally
dormant but upon neutrophil stimulation, the cytosolic subunits
translocate to the membrane and associate with cytochrome
b558, resulting in rapid activation of the
oxidase. Phosphorylation of the regulatory subunit
p47phox plays a key role in this process. The
kinetics of NADPH oxidase activation parallel the kinetics of
p47phox phosphorylation (15). Phosphorylation of
p47phox is thought to induce conformational
changes that allow subsequent binding of
phospho-p47phox to cytochrome
b558 and the other cytosolic subunits
(14-16).
NADPH oxidase in non-phagocytic cells such as EC and vascular smooth
muscle exhibits significant differences from the neutrophil enzyme. In
particular, in contrast to the neutrophil enzyme, the oxidase in
non-phagocytic cells is constitutively active at a low level even in
unstimulated cells, yet it can be further stimulated acutely by
agonists such as AngII and cytokines (5). Recent studies (17, 18) have
indicated the presence of a number of isoforms of
gp91phox, termed Noxs (for NADPH oxidase), and
it has been suggested that the substitution of
gp91phox (also known as Nox2) by Nox1 or Nox4
may account for the different behavior of non-phagocytic enzymes. For
example, it has been shown in rat aortic smooth muscle cells that the
predominant Nox isoforms are Nox4 and Nox1 with very low to
undetectable levels of gp91phox expressed (18).
In EC, however, all the classic NADPH oxidase subunits, including
gp91phox, are expressed (19, 20). Recently, we
have reported that, in contrast to neutrophils, in unstimulated
quiescent cultured EC a substantial proportion of the NADPH oxidase is
present as already fully preassembled complexes in a predominantly
perinuclear location associated with the intracellular cytoskeleton
(20). Thus, the "cytosolic" subunits
p47phox, p40phox,
p67phox, and Rac1 could be co-immunoprecipitated
down with p22phox and
gp91phox in unstimulated EC. The presence of
these preassembled complexes is likely to account for the constitutive
NADPH oxidase activity of unstimulated EC. However, if the oxidase is
already preassembled in unstimulated EC, what is the precise nature of
the mechanism(s) underlying the increase in oxidase activity following
addition of agonists such as AngII? Recently, we reported that the
stimulation of EC NADPH oxidase activity by TNF
or phorbol ester
(PMA) required the presence of p47phox, although
the mechanisms involved were not defined in that study (21). Likewise,
in vascular smooth muscle cells, the p47phox
subunit has been shown to be essential for ROS production in response
to PMA, AngII, thrombin, and platelet-derived growth factor (22, 23),
and compelling evidence has been provided for its involvement in
atherosclerosis (23). Potential mechanisms that could underlie the
increase in NADPH oxidase activity induced by AngII may include
(a) an increase in the number of fully assembled oxidase
complexes, (b) the translocation of
p47phox to partially assembled complexes, and/or
(c) the phosphorylation of p47phox
(or other subunits) that are already part of the assembled oxidase complex in EC.
In the present study, we have therefore investigated the mechanisms of
AngII-induced EC NADPH oxidase activation, focusing on the role of
p47phox, its phosphorylation, and its possible
translocation and association with cytochrome
b558. Our results provide an insight into the fundamental mechanisms of AngII-induced EC NADPH oxidase activation and
ROS production.
 |
MATERIALS AND METHODS |
Reagents--
5-(and
6)-chloromethyl-2',7'-dichlorohydrofluorescein diacetate (DCF) was
purchased from Molecular Probes. Goat polyclonal antibodies to
p22phox and p47phox and
the corresponding blocking peptides were from Santa Cruz Biotechnology.
Affinity-purified rabbit polyclonal antibodies to
p47phox and p22phox were
a kind gift from Dr. F. Wientjes (University College London, UK). The
anti-phosphoserine-specific monoclonal antibody was from Sigma. All
other reagents were from Sigma except where specified.
Cell Culture--
p47phox null mice
(p47phox
/
) on a 129 sv background were
generated as described (24) and were kindly provided by Dr. Jurgen Roes (University College London, UK). All studies conformed with the Guidance on the Operation of Animals (Scientific Procedures) Act, 1986 (Her Majesty's Stationery Office, London, UK). Coronary microvascular EC (CMEC) were isolated from the hearts of 8-10-week-old
p47phox
/
and matched wild-type mice exactly
as described previously (25). Cultured CMEC were used at passage 2. Human microvascular EC (HMEC-1) (26) were obtained from the Centers for
Disease Control and Prevention (Atlanta, GA). HL-60 cells were from the
American Type Culture Collection. They were differentiated to
neutrophils by incubation with 1.3% Me2SO for seven days
and stimulated with PMA (100 ng/ml) for 30 min to activate NADPH
oxidase (21).
Detection of ROS Production--
A number of complementary
methods and assays were used for the detection of ROS (either
O
or H2O2). For visualization
of ROS production, EC were cultured onto chamber slides, washed in
Hanks' buffer, and then exposed to AngII (100 nmol/liter) or buffer
alone for 10 min. Cells were incubated with 10 µmol/liter DCF in
Hanks' buffer for 30 min at room temperature, and DCF fluorescence
(which detects mainly H2O2) was then acquired as described (21). Fluorescence intensity was quantified
microscopically from at least three random fields (1024 × 1022 pixels; 269.7 × 269.2 µm) per slide (50 cells assessed per
slide), three slides per experimental condition.
In some experiments, ROS generation in DCF-loaded cells was also
measured by flow cytometry on a FACScalibur instrument (BD PharMingen). For these studies, EC just reaching confluence
were detached by trypsin and stimulated with AngII (100 nM). Cellular fluorescence was quantitated by the geometric
means of data distributions (GMean).
Specific NADPH-dependent O
production was
assessed by lucigenin (5 µmol/liter) chemiluminescence as described
previously (19, 21). We measured this both in living cells
and in total EC homogenate. For the latter, cells stimulated with AngII
(100 nmol/liter) or buffer in serum-free medium were scraped into
ice-cold Hank's balanced salt solution supplemented with
MgCl2 0.8 mmol/liter and CaCl2 1.8 mmol/liter. Cells were disrupted by rapid freezing in liquid nitrogen followed by
sonication. O
production was measured in the presence of
NADPH (100 µmol/liter) for 20 min and was expressed as mean arbitrary
light units per min (19, 21). In experiments with an AT1 or
AT2 receptor antagonist (losartan and PD123319,
respectively), these were added 30 min prior to AngII.
Immunoprecipitation and Immunoblotting--
Immunoprecipitation
was performed as described previously using protein samples (250 µg)
diluted in Tris-HCl 0.05 mol/liter buffer (pH 7.4) with protease
inhibitors (20). p22phox or
p47phox was immunoprecipitated down with
affinity-purified goat polyclonal antibodies coupled to protein
G-agarose beads. Immunoblotting of immunoprecipitates was undertaken
with affinity-purified rabbit polyclonal antibodies to
p47phox and p22phox or an
anti-phosphoserine-specific monoclonal antibody. Protein isolated from
differentiated, PMA-stimulated HL-60 cells was used as a positive control.
Subcellular Protein Fractionation--
Differential
centrifugation was undertaken exactly as described previously (20). The
following fractions were separated: (a) nuclei-enriched (N
fraction) pelleted at 1,475 × g; (b)
primary mitochondria and other large organelles (C fraction) pelleted at 10,800 × g; (c) submitochrondrial
particles, smaller organelles, and some microsomes (D fraction)
pelleted at 29,000 × g; (d) microsomes, microperoxisomes, and membrane fractions (E fraction) pelleted at
100,000 × g; and (e) soluble cytoplasmic
protein (S fraction). The "purity" of fractions was confirmed by
marker enzyme activities as described previously (20).
Sense and Antisense p47phox cDNAs Transfection
and Confocal Microscopy--
Full-length neutrophil sense and
antisense p47phox cDNAs were generated and
CMEC were transfected as described previously (21). After 72 h of
transfection, cells were examined by confocal microscopy for the
expression of p47phox protein using a BioRad
1024 microscope. Optical sections were taken at 0.5-µm intervals, and
images were captured and stored digitally for analysis.
Statistics--
Data are presented as mean ± S.D. of at
least three different experiments for each condition. Six mice per
group were used for each CMEC isolation, and at least three independent
isolates were studied. Comparisons were made by one-way analysis of
variance or unpaired Student's t test as appropriate.
p < 0.05 was considered statistically significant.
 |
RESULTS |
AngII-induced ROS Production Is Inhibited in EC Lacking
p47phox--
To investigate whether
p47phox is necessary for AngII-stimulated
increases in NADPH oxidase activity, wild-type and
p47phox
/
CMEC were studied in parallel (Fig.
1A). Using flow cytometry in
DCF-loaded cell suspensions, basal ROS production was detected in
unstimulated EC of both types. Unstimulated
p47phox
/
CMEC had significantly higher DCF
fluorescence than wild-type cells (GMean 207.6 ± 11.26 versus 162.9 ± 13.44; p < 0.01).
Wild-type EC incubated with AngII for 30 min showed a significant
increase in fluorescence signal (GMean 241.1 ± 12.40 versus 162.9 ± 13.44; p < 0.01). In
marked contrast, AngII treatment induced no increase in fluorescence in
p47phox
/
cells; on the contrary, DCF
fluorescence was reduced after AngII treatment (GMean 168.8 ± 13.45 versus 207.6 ± 11.26; p < 0.01).

View larger version (26K):
[in this window]
[in a new window]
|
Fig. 1.
ROS production by wild-type and
p47phox / CMEC detected by DCF
fluorescence. A, ROS generation in EC suspensions
detected by flow cytometry. The experiment was repeated five times and
quantitative analyses performed on geometric means of the data
distributions. B, representative microscopic images of DCF
fluorescence in adherent wild-type and
p47phox / EC with and without AngII.
C, mean data for DCF fluorescence intensity in adherent EC,
quantified microscopically (n > 300 cells from three
independent experiments). *, p < 0.05 compared with
respective wild-type or p47phox /
control.
|
|
Similar results were obtained in experiments on adherent CMEC cultured
onto chamber slides and examined by fluorescence microscopy (Fig. 1,
B and C). A small amount of DCF fluorescence was
detected in a mainly perinuclear distribution in unstimulated EC from
both groups, with a higher level in p47phox
/
cells. AngII significantly increased fluorescence (by ~50%) in wild-type EC. However, no increase was observed after AngII in p47phox
/
cells. The addition of the
cell-permeable O
scavenger, Tiron, abolished DCF fluorescence
in both groups, indicating that the primary source of
H2O2 detected by DCF was likely to be
O
.
As an alternative approach, NADPH-dependent
O
production with and without AngII was examined by lucigenin
chemiluminescence both in living cells and cell homogenates (Fig.
2). Unstimulated
p47phox
/
cells had slightly but
significantly higher NADPH-dependent O
production
than wild-type cells. AngII significantly increased NADPH-dependent O
production in intact wild-type
cells (by 38.5 ± 9.8%), whereas this response was completely
blocked in p47phox
/
cells.
NADPH-dependent O
production was fully inhibited
by DPI (50 µmol/liter) in both groups (Fig. 2) but was
unaltered by rotenone, oxypurinol, or L-NAME (data not shown).
Similar results were obtained with cell homogenate (Fig. 2,
middle panel). Again basal O
production was higher
in p47phox
/
cells compared with wild-type.
AngII caused an ~2-fold increase in basal O
production in
wild-type cells, whereas no increase was observed in
p47phox
/
cells. AngII-induced
NADPH-dependent O
production by wild-type cell
homogenate was inhibited by DPI, superoxide dismutase, or Tiron, but
was unaffected by rotenone, oxypurinol, or L-NAME (Fig. 2,
middle and lower panels), consistent with NADPH
oxidase as the source. Collectively, these results suggest that
the p47phox subunit is essential for the
AngII-induced increase in NADPH oxidase activity in EC.

View larger version (17K):
[in this window]
[in a new window]
|
Fig. 2.
NADPH-dependent O
production by intact wild-type and p47phox /
CMEC or cell homogenate measured by lucigenin chemiluminescence.
The top panel shows results for intact cells, the
middle and lower panels for cell homogenate. MLU,
mean light units. *, p < 0.05 compared with the
respective wild-type or p47phox / control. In
the bottom panel, the effect of a number of inhibitors or
antagonists on the maximal AngII-induced NADPH-dependent
O production by wild-type EC homogenate is shown. *,
p < 0.05 compared with oxidase activity in the absence
of inhibitors.
|
|
Effects of p47phoxSense or Antisense
cDNA Transfection on the Oxidase Response to AngII--
To confirm
that the loss of ROS response to AngII stimulation was directly
attributable to the absence of p47phox, we
undertook experiments to acutely deplete or restore
p47phox levels in EC by transfection of
antisense or sense p47phox cDNA,
respectively. Fig. 3A shows a
representative example of the effect of antisense
p47phox cDNA transfection on
p47phox expression in wild-type CMEC, assessed
by confocal microscopy. There was an ~60.4 ± 7.3% reduction in
p47phox fluorescence (mean data from 300 cells
from three independent transfections). In wild-type EC transfected with
vector control, AngII significantly increased
NADPH-dependent O
production; this effect was
totally lost in antisense cDNA transfected cells (Fig.
3B, top panel).

View larger version (38K):
[in this window]
[in a new window]
|
Fig. 3.
Effect of sense and antisense
p47phox cDNA transfection of EC.
A, representative confocal micrographs showing wild-type or
knockout cells stained for p47phox after
transfection with empty vector or cDNA as labeled. B,
AngII-induced changes in NADPH-dependent O
production (lucigenin chemiluminescence) in wild-type and
p47phox / cells. *, p < 0.05 comparing AngII versus respective control.
|
|
On the other hand, transfection of p47phox
/
cells with sense p47phox cDNA resulted in
significant expression of p47phox (Fig.
3A, lower panel). In parallel experiments,
NADPH-dependent O
production in response to AngII was restored after sense p47phox cDNA
transfection (Fig. 3B, bottom panel). These
experiments clearly indicate an essential role of
p47phox in AngII-induced EC NADPH oxidase
activation and ROS production.
AngII Induces p47phox Phosphorylation and
Increases Binding to p22phox--
To examine the
mechanism through which p47phox is involved in
AngII-induced NADPH oxidase activation, we studied
p47phox phosphorylation and the binding of
p47phox to p22phox. These
studies were undertaken in HMEC-1, which were available in
greater numbers than the primary murine EC isolates. Adherent cells
were stimulated with AngII for 1-60 min,
p47phox was immunoprecipitated from cell
homogenate, and serine phosphorylation of
p47phox was detected with a
phosphoserine-specific antibody. In unstimulated cells, no significant
serine phosphorylation of p47phox was detected
(Fig. 4A, upper
panel). AngII induced rapid p47phox
phosphorylation within 1 min, peaking at ~15 min and thereafter declining gradually. The same membrane was reprobed with an
anti-p22phox polyclonal antibody to assess the
relationship between p47phox phosphorylation and
p47phox-p22phox complex
formation (Fig. 4A, lower panel).
p22phox protein was readily detectable even in
the p47phox immunoprecipitate of unstimulated
EC, indicating complex formation in these cells in the absence of
significant p47phox phosphorylation. However,
the amount of p22phox co-immunoprecipitated with
p47phox rapidly increased in AngII-stimulated
EC, peaking at ~15 min.

View larger version (28K):
[in this window]
[in a new window]
|
Fig. 4.
AngII-induced phosphorylation of
p47phox and
p22phox-p47phox complex
formation in HMEC-1. A, upper panel,
p47phox was immunoprecipitated (IP)
and then immunoblotted (IB) for serine phosphorylation; the
lower band, which is also present in the IgG control lane, is a
nonspecific band. Lower panel, the same membrane was
re-probed for p22phox co-immunoprecipitated down
with p47phox. B, upper and
middle panels, mean data for changes in
p47phox phosphorylation and complex formation
after AngII treatment. Immunoblots were scanned densitometrically, and
results were expressed as arbitrary absorbance units. Lower
panel, NADPH-dependent O production by cell
homogenates was measured by chemiluminescence and expressed as mean
light units (MLU) per mg protein. *, p < 0.05 compared with unstimulated cells (i.e. time 0).
|
|
Fig. 4B shows the mean data from three different experiments
and also shows data for NADPH oxidase activity measured in parallel in
EC homogenates. It is evident that AngII-induced increases in
p47phox phosphorylation and
p22phox-p47phox complex
formation were paralleled by significant increases in oxidase activity.
Effect of AT Receptor Antagonists on p47phox
Phosphorylation and Binding to
p22phox--
Because both AT1 and
AT2 receptors are expressed on EC and could mediate
different biological actions (8, 9, 27), we examined their role in
AngII-induced p47phox phosphorylation. HMEC-1
were preincubated for 30 min with either the AT1 antagonist
losartan (1 µmol/liter) or the AT2 antagonist PD123319
(0.1 µmol/liter) before treatment with AngII.
p47phox was immunoprecipitated and examined for
serine phosphorylation (Fig.
5A, upper panel).
Both losartan or PD123319 inhibited AngII-induced p47phox phosphorylation, although the
AT1 antagonist was more effective (Fig. 5B,
upper panel).

View larger version (21K):
[in this window]
[in a new window]
|
Fig. 5.
Effect of AT1 and AT2
antagonists on p47phox phosphorylation and
p22phox-p47phox complex
formation. A, upper panel,
p47phox was immunoprecipitated (IP)
and immunoblotted (IB) for serine phosphorylation.
Lower panel, p22phox was
immunoprecipitated and detected for the co-existence of
p47phox. B, upper and
middle panels, mean data for changes in
p47phox phosphorylation and complex formation.
Immunoblots were scanned densitometrically and expressed as arbitrary
absorbance units. Lower panel, NADPH-dependent
O production by cell homogenates measured by
chemiluminescence and expressed as mean light units (MLU)
per min per mg protein. *, p < 0.05 compared with
unstimulated cells (control).
|
|
To examine complex formation, we immunoprecipitated
p22phox and then probed for the co-existence of
p47phox (Fig. 5A, lower
panel). After 30 min of AngII treatment, the amount of
p47phox co-immunoprecipitated with
p22phox had increased significantly
(p < 0.05; Fig. 5B, top panel).
Interestingly, p47phox was detected as a doublet
(Fig. 5A, lower panel); the additional band may
represent phospho-p47phox because it was not
detected in unstimulated EC and was decreased in cells preincubated
with losartan or PD123319. Both losartan and PD123319 significantly
inhibited AngII-induced
p47phox-p22phox complex
formation, with losartan again being more effective (Fig. 5B, middle panel). This was accompanied by
inhibition of AngII-induced NADPH oxidase activity, which occurred to a
greater extent with losartan than with PD123319 (Fig. 5B,
lower panel).
Subcellular Redistribution of p47phox after AngII
Treatment--
Finally, we examined the effect of AngII treatment on
the subcellular distribution of p47phox in
HMEC-1. In unstimulated cells, after fractionation by differential centrifugation, the majority of p47phox (44 ± 3%) was detected in the 100,000 × g (microsomal
and membrane) E fraction with a significant proportion (30 ± 4%)
also found in the 1457 × g (nuclei-enriched) N
fraction (Fig. 6A, upper
panel). After AngII stimulation, 59 ± 4% of
p47phox was detected in the N fraction and was
clearly detected as a doublet band (Fig. 6A, lower
panel).

View larger version (20K):
[in this window]
[in a new window]
|
Fig. 6.
Distribution of
p47phox studied by subcellular
fractionation. A, p47phox
expression in each fraction before and after AngII stimulation,
assessed by immunoblotting. B, NADPH-dependent
O production by each fraction assessed by chemiluminescence.
*, p < 0.05 compared with results without AngII.
|
|
In line with these results, NADPH-dependent O
production was significantly increased in the N fraction in
AngII-stimulated EC. A significant increase was also found in the E
fraction (Fig. 6B).
To assess whether a change in p47phox
subcellular distribution could also be detected by a complementary
method, we undertook immunocytochemical staining and confocal
microscopy in HMEC-1 with and without AngII stimulation (Fig.
7). In unstimulated cells, p47phox had a mainly perinuclear and rather
punctate distribution. After AngII stimulation, the perinuclear
p47phox labeling appeared more organized in a
reticular pattern, and there was also clear labeling of the cell
surface membranes (Fig. 7, lower panels, white
arrows), consistent with a redistribution of
p47phox.

View larger version (79K):
[in this window]
[in a new window]
|
Fig. 7.
Confocal micrographs showing
immunocytochemical staining of HMEC-1 for
p47phox with and without AngII
stimulation. The right panels show higher
magnification images. The white arrows indicate areas of
cell surface membrane labeling in AngII-stimulated cells.
|
|
 |
DISCUSSION |
AngII is well known to increase EC ROS production, and recent
studies have implicated the stimulation of NADPH oxidase as an
important component of this response (4, 5, 8, 9). In neutrophils, the
activation of NADPH oxidase involves the translocation and binding of
cytosolic subunits (p47phox,
p67phox, p40phox, and
rac) to cytochrome b558; the phosphorylation of
p47phox is considered to be a key step in this
process (14). In EC, however, we have recently shown that a significant
proportion of the oxidase is present as fully preassembled
intracellular complexes, even in unstimulated ("resting") cells
(20). The main new findings of the present study are that
(a) AngII-stimulated increases in EC NADPH oxidase activity
and ROS production are absolutely dependent on the
p47phox subunit; (b) the
phosphorylation of p47phox and its binding to
the cytochrome b558 to form new complexes is
required for the AngII response; and (c) in unstimulated
cells, the p47phox that is bound to cytochrome
b558 appears to be in an unphosphorylated or
minimally phosphorylated state.
The requirement for p47phox in the AngII
response was clearly demonstrated in experiments on EC isolated from
p47phox
/
and wild-type mice. Furthermore,
in vitro transfection of p47phox
cDNA into p47phox
/
EC successfully
restored the ROS response to AngII, whereas in vitro
depletion of p47phox from wild-type EC by
antisense cDNA transfection resulted in loss of the ROS response.
In the present study, we used lucigenin chemiluminescence or DCF
fluorescence for ROS detection either in adherent cells, cell
suspensions, and/or cell homogenates. Lucigenin chemiluminescence
detects both intracellular and extracellular O
production,
whereas DCF fluorescence detects mainly intracellular
H2O2 (28). Evidence that the primary species
leading to the H2O2 detected by DCF
fluorescence in this study was O
was provided by the finding
that the cell-permeable O
scavenger Tiron substantially
inhibited the DCF signal (Fig. 1C). The data from these
different assays (i.e. Figs. 1-3) were broadly comparable
in indicating a requirement for p47phox in the
AngII response; the small quantitative differences between assays are
likely to reflect the differing experimental conditions (i.e. adherent versus trypsinized cells and
intact cells versus cell homogenate).
The oxidase subunit p47phox contains several
serine residues clustered within the carboxyl terminus that form
potential protein kinase C phosphorylation sites (29). Native
p47phox is a highly basic protein that undergoes
stepwise charge shifts with successive phosphorylation events (29, 30).
In neutrophils, p47phox phosphorylation is
thought to be a prerequisite for its interaction with
p22phox and the formation of a functional
oxidase complex (15, 16). In EC, however, a significant proportion of
the NADPH oxidase is present as fully preassembled intracellular
complexes even in unstimulated ("resting") cells (20).
Nevertheless, the oxidase can be further stimulated by agonists such as
AngII. In the present study, we confirmed that
p47phox could be co-immunoprecipitated with
p22phox in the absence of AngII stimulation. An
interesting and new finding was that the p47phox
associated with p22phox in unstimulated EC
appeared to be largely unphosphorylated (or minimally phosphorylated).
Although it is unclear how unphosphorylated p47phox may bind to
p22phox in EC, it has previously been reported
that unphosphorylated p47phox can bind to
p22phox in in vitro assays in the
presence of arachidonic acid but that this binding only supports low
O
production (31). The binding of unphosphorylated rather
than phosphorylated p47phox to cytochrome
b558 may therefore account for the fact that
oxidase activity in unstimulated EC is relatively low despite the
presence of fully assembled complexes.
The mechanism of further increases in oxidase activity in a setting
where there are already existing preassembled complexes could either
involve the phosphorylation of already bound
p47phox or the formation of additional (new)
complexes through the association of p47phox
with other oxidase components. In neutrophils, initial
p47phox phosphorylation is thought to occur in
the cytosol; additional p47phox phosphorylation
occurs upon interaction with membrane components, which then leads to
immediate and irreversible binding of p47phox to
cytochrome b558 (15). In the present study, we
found that AngII very rapidly induced p47phox
phosphorylation (<1 min), which was accompanied by an increase in
p47phox-p22phox complex
formation and O
production. These data are consistent with a
mechanism where p47phox phosphorylation promotes
stable binding of p47phox to
p22phox (and thus oxidase activation), perhaps
by inducing appropriate conformational changes in
p47phox (16, 31), i.e. that
p47phox phosphorylation promotes the formation
of new oxidase complexes. Consistent with this, we found that AngII
stimulation was associated with an ~20% translocation of
p47phox from other fractions to the N
(nuclei-enriched) fraction, where we have previously shown that the
majority of oxidase components are found in EC (20). Interestingly,
AngII also induced a small but significant translocation of
p47phox to the membrane-enriched E fraction,
consistent with some of the oxidase being present also on the plasma
membrane. Indeed, immunocytochemical staining and confocal microscopy
demonstrated p47phox labeling of the plasma
membrane in AngII-treated cells as well as a suggestion of a more
organized reticular labeling in the perinuclear region. The temporal
relationship between p47phox phosphorylation,
binding of phospho-p47phox to
p22phox, and the increase in NADPH oxidase
activity suggests that, as in neutrophils (14-16), the serine
phosphorylation of p47phox is the key step for
oxidase activation by AngII.
AngII exerts its biological effects through the activation of two
distinct classes of receptors, AT1 and AT2.
Both AT1 and AT2 receptors are known to be
expressed in endothelial cells (9, 27, 32). Most of the stimulatory
effects of AngII on cardiovascular cells are mediated by
AT1 receptors, whereas there is less data on the
AT2 receptor subtype, which in some settings can be
antagonistic. The results of the present study suggested that both
AT1 and AT2 receptors were involved in
AngII-induced p47phox phosphorylation and NADPH
oxidase activation because the latter could be inhibited either by
losartan or PD123319, which are known to be specific inhibitors of the
AT1 and AT2 receptor, respectively. However,
the effects of losartan appeared to be more potent than those of
PD123319, in accordance with a previous study in EC (9).
An interesting finding in the present study was that unstimulated
p47phox
/
cells maintained a significant
level of NADPH-dependent O
production and, in
fact, had slightly but significantly higher ROS production
than wild-type cells, as we have reported previously (21). The
underlying basis for this "background" ROS production remains to be
established. One interpretation of these results would be that the
p47phox subunit is not essential for "basal"
NADPH oxidase activity in EC. Consistent with this possibility, it has
been reported that in cell-free systems NADPH oxidase can be activated
in the absence of p47phox if high concentrations
of p67phox and Rac1 are present (33, 34).
However, an alternative interpretation would be that the background ROS
production emanates from a completely different source. The findings of
the present and previous (21) studies that the background
NADPH-dependent O
production is abolished by DPI
but is unaffected by rotenone, oxypurinol, or L-NAME indicates that any
such source would have to be an NADPH-dependent
flavoprotein enzyme. Because DPI is not a specific inhibitor just of
NADPH oxidase, definitive analysis of the underlying source of this
background ROS production in EC may require studies in which multiple
NADPH oxidase subunits are knocked out.
In summary, this study demonstrates that AngII-induced ROS production
by EC is critically dependent on the NADPH oxidase subunit p47phox. Serine phosphorylation of
p47phox, followed by
p47phox translocation and the stable binding of
phospho-p47phox to
p22phox are the key steps that initiate
AngII-induced NADPH oxidase activation. This process appears to
involve the formation of new oxidase complexes. ROS generated
intracellularly in the perinuclear region may serve as signaling
molecules for the activation of redox-sensitive enzymes and
transcription factors (1, 5), whereas extracellular ROS may contribute
to endothelial dysfunction (11-13).