| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Institute for Physiological Chemistry and Pathobiochemistry, Johannes Gutenberg-University, D-55099 Mainz, Germany;
* Institute for Biochemistry II, University of Frankfurt Hospital, Frankfurt, Germany;
Institute for Anatomy and Cell Biology, Justus-Liebig-University, D-35385 Giessen, Germany; and
§ Aventis, Disease Group Cardiovascular Agents, D-65926 Frankfurt, Germany
1Correspondence: Institute for Biochemistry II, University of Frankfurt Hospital, Theodor-Stern-Kai 7, D-60590 Frankfurt, Germany. E-mail: wme{at}biochem2.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: eNOS • yeast two-hybrid system • protein–protein interaction • subcellular trafficking
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, is synthesized from
L-arginine by a family of nitric oxide synthases (NOS, EC 1.14.13.39).
Three NOS isoforms are known to date: neuronal NOS (nNOS) expressed
mainly in neuronal tissues and skeletal muscle; inducible NOS (iNOS)
originally cloned from macrophages and later discovered in many other
cell types; and endothelial NOS (eNOS) present in vascular endothelial
cells, cardiac myocytes, and blood platelets (2)
. The
activity of all three isoforms is critically dependent on calmodulin,
which binds to nNOS and eNOS at elevated intracellular calcium
concentrations ([Ca2+]i)
while it is permanently associated with iNOS even at basal calcium
levels. Thus, activity of nNOS and eNOS is modulated by changes in
[Ca2+]i leading to
transient NO production, whereas iNOS continuously releases NO
independent of [Ca2+]i
fluctuations (3)
. iNOS activity is primarily regulated via
induced gene expression, although constitutive expression of iNOS is
found in some cell types (4
, 5)
. nNOS and eNOS may be
under (post)transcriptional regulation (6)
, but
posttranslational control mechanisms clearly dominate in the regulation
of their activities.
Among the various NOS, the endothelial isoform eNOS bears some unique
features. The enzyme is cotranslationally modified at its amino
terminus by myristoylation and is further acylated by palmitoylation of
two residues next to the myristoylation site. These modifications
direct eNOS to endothelial cell caveolae, i.e., small invaginations of
the plasma membrane abundant in the transmembrane protein caveolin
(7
, 8)
. Attachment of eNOS to plasma membrane caveolae
seems to be necessary for the efficient release of nitric oxide by
external stimuli (9)
. After activation, eNOS shuttles
between caveolae and other subcellular compartments such as the
noncaveolar plasma membrane portions, Golgi apparatus, and/or
perinuclear structures, depending on cell type and mode of activation
(2)
. Transient phosphorylation and depalmitoylation
further contributes to activity modulation and subcellular targeting of
eNOS. For example, stimulation of endothelial cells by shear stress or
exposure to vascular endothelial growth factor (VEGF) results in
protein kinase B (Akt)-dependent phosphorylation at Ser-1177 of eNOS
and thus activation of the enzyme in a calcium-independent manner
(10
, 11)
.
Multiple protein–protein interactions are involved in the fine-tuning
of eNOS activity. For example, direct interaction of the oxygenase or
reductase domains of eNOS with the scaffolding domain of caveolin-1 in
endothelial cells or caveolin-3 in cardiac myocytes inhibits the enzyme
(12
13
14)
. Caveolin-bound eNOS remains inactive unless
Ca2+-calmodulin displaces caveolin from an
overlapping binding site, thereby activating the enzyme
(15)
. Recently it has been shown that eNOS activity is
inhibited through its association with the carboxyl-terminal domain of
G-protein-coupled receptors (16)
and that stimulation of
endothelial cells by VEGF, histamine, or fluid shear stress promotes
the binding of heat shock protein 90 (Hsp90) to eNOS, resulting in
allosterical activation of the enzyme (17)
. Thus, several
protein factors have been identified that directly control the
catalytic efficiency of eNOS. By contrast, the factor(s) governing the
subcellular translocation of this key enzyme is still obscure. In a
yeast two-hybrid search for cellular proteins associated with
endothelial NO synthase, we have identified a novel eNOS interacting
protein, which we name NOSIP, that modulates NO production of eNOS by a
hitherto unknown mechanism, i.e., interference with the subcellular
trafficking of the enzyme.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-tubulin from Sigma (St. Louis,
Mo.). NG-Nitro-L-arginine (L-NNA) and calcium
ionophore A23187 were from Alexis (Gruenberg, Germany), Pansorbin cells
from Calbiochem (Nottingham, U.K.), GSH-Sepharose from Pharmacia,
enzymes from New England Biolabs (Beverly, Mass.), radiochemicals from
ICN Pharmaceuticals (Irvine, Calif.), and the enhanced
chemiluminescence kit from Amersham (Little Chalfont, U.K.). Other
reagents were obtained from Sigma unless otherwise indicated.
Caveolin-1 peptides
NH2-DGIWKASFTTFTVTKYWFYR-CONH2
(residues 82–101) derived from the scaffolding domain and
NH2-DDVVKIDFEDVIAEPEGTHSF-CONH2
(residues 61–81) were synthesized on a 9050 Pep-Synthesizer
(Milligen) using Fmoc/HOBt chemistry and a Fmoc-amide polyethylene
glycerol polystyrene resin. After cleavage from the resin and
purification by reverse-phase high performance liquid chromatography,
the peptides were characterized by mass spectroscopy.
Yeast two-hybrid screening and bacterial expression of GST fusion
proteins
The yeast two-hybrid system (18)
using
Saccharomyces cerevisiae strain EGY48, which contains
reporter plasmid pSH18–34, was used for the selection procedure. To
construct the LexA-fused bait protein, we amplified human eNOS
oxygenase domain cDNA (corresponding to residues 1–486 of the protein
sequence) by polymerase chain reaction (PCR) using forward primer
5'-CACGAATTCATGGGCAACTTGAAGAGCGTGGCCC-3' with a flanking
EcoRI restriction site and reverse primer
5'-GTGCGCTCGAGCTACTTGGCGGCACTCCCCTTCCAG-3' with a flanking
XhoI restriction site. The construct was inserted into the
EcoRI and XhoI restriction sites of pEG202
(‘pEG-eNOS1–486’). EGY48 containing pSH18–34
and pEG-eNOS1–486 was transformed with the human
placenta Matchmaker cDNA library and 
2 x
107 transformants were plated on selection medium
(leu2). Colonies were screened by the ß-galactosidase
reporter gene assay (lacZ) using X-gal
(5-bromo-4-chloro-3-indolyl-ß-D-galactoside) plates and library
plasmids of positive yeast clones were rescued by transformation of
Escherichia coli KC8 strain. To test the specificity of
protein–protein interaction, EGY48/pSH18–34 was retransformed with
the rescued library plasmids. The yeast reporter strain contained as
the bait pEG-eNOS1–486, empty pEG202, or
pEG-cdNHE (controls) encoding the carboxyl-terminal domain of human
sodium proton exchanger NHE-1. Transformants were assayed for
leu2 and lacZ activity. DNA sequences of selected
library plasmids were determined by the dideoxy chain termination
method (MWG Biotech, Ebersberg, Germany).
To define the region(s) of the eNOS oxygenase domain binding to
the interacting proteins, the following segments of eNOS were PCR
amplified: residues 98–486 (protein sequence) using forward primer
5'-CACCGAATTCCGCTGCCTGGGCTCCCTGGTATTTCC-3' and reverse primer
5'-GTGCGCTCGAGCTACTTGGCGGCACTCCCCTTCCAG-3'; residues 1–366,
forward primer 5'-CACGAATTCATGGGCAACTTGAAGAGCGTGGCCC-3',
reverse primer 5'-CGGTGCTCGAGCTACAGGTTCCTCGTGCCGATCTCAG-3'; and
residues 98–366, forward primer
5'-CACCGAATTCCGCTGCCTGGGCTCCCTGGTATTTCC-3', reverse primer
5'-CGGTGCTCGAGCTACAGGTTCCTCGTGCCGATCTCAG-3'. The subfragments were
cloned into the pEG202 vector via EcoRI and XhoI
restriction sites, thus creating pEG-eNOS98–486,
pEG-eNOS1–366, and
pEG-eNOS98–366. The plasmids were used to
transform the yeast reporter strain EGY48/pSH18–34 and tested for the
interaction with the identified eNOS-associating proteins as above. For
positive control, the eNOS subfragments were cloned into the pJG4–5
vector using the same restriction sites
(pJG-eNOS1–486,
pJG-eNOS98–486,
pJG-eNOS1–366,
pJG-eNOS98–366) and tested for eNOS
homodimerization. Reporter gene expression was monitored by a liquid
culture assay using o-nitrophenyl-ß-D-galactopyranoside as
the substrate. Assays were done in triplicate throughout and the
specific ß-galactosidase activity was calculated (19)
.
For bacterial expression of identified eNOS interaction partners as GST fusion proteins, the inserts of the isolated library plasmids were excised by EcoRI and XhoI and ligated into pGEX2T previously modified for in-frame expression of EcoRI/XhoI fragments. E. coli BL21 strain was transformed with recombinant pGEX2T plasmids, and GST fusion proteins from exponentially growing bacteria were purified by GSH-Sepharose affinity chromatography according to the manufacturer’s instructions (Pharmacia).
Cell culture, stable transfection, and viral infection
Dihydrofolate reductase-deficient CHO (Chinese hamster ovary
cells) were cultured in Dulbecco’s modified Eagle’s medium (Life
Technologies, Inc., Grand Island, N.Y.) containing 10% (v/v) fetal
calf serum, 0.5% (w/v) penicillin/streptomycin supplemented with 100
µM hypoxanthine and 16 µM thymidine (HT) in a humidified 5%
CO2 atmosphere at 37°C. The cells were
transfected with the human eNOS-cDNA construct by the Lipofectamine
method (Life Technologies, Inc.). Transfectants were isolated by
single-cell cloning in the same medium without HT. eNOS expression was
analyzed by immunoblotting, and the clone with the highest expression
rate (‘CHO-eNOS’) was chosen for further experiments. For generation
of recombinant SFV particles, the 5'-hemagglutinin (HA) -tagged
NOSIP-cDNA fragment was ligated into the pSFV2 vector using appropriate
restriction sites. In vitro transcription and in
vitro packaging of recombinant viruses were done as described
(20)
. Briefly, SFV-NOSIP mRNA was mixed with SFV-helper 2
mRNA and electroporated into confluent baby hamster kidney cells. After
24 h, recombinant virus particles present in the supernatant were
activated with -chymotrypsin (20)
. CHO-eNOS cells were
grown on 6-well plates, infected with 100 µl virus solution per well,
and processed 5–15 h later. Human umbilical vein endothelial cells
(HUVECs) were isolated from human umbilical cords as described
previously (21)
.
Immunoprecipitation and immunoblotting
For determination of in vitro interactions, cultured
CHO-eNOS cells were washed 3x with ice-cold phosphate-buffered saline
(PBS), lysed for 1 h on ice in 20 mM Tris-HCl, pH 7.4, 150 mM
NaCl, 2 mM EDTA, 1% Triton X-100 (‘lysis buffer’) supplemented with
Complete protease inhibitor mixture (Boehringer Mannheim, Mannheim,
Germany), and centrifuged at 14,000 g for 10 min at 4°C.
GST-NOSIP or GST-cdNHE (5 µg/ml each) was added to the supernatant
and the mixtures were incubated for 1 h at 4°C under rotation
with 10 µl of anti-eNOSp or anti-NOSIP; for control, the
corresponding preimmune sera were used. Antibodies were bound to
Pansorbin cells and precipitated at 8000 g for 2 min at
4°C. The precipitates were washed 4x with lysis buffer and dissolved
in sample buffer (63 mM Tris-HCl, pH 6.8 containing 2.5% sodium
dodecyl sulfate (SDS), 5% glycerol, 5% ß-mercaptoethanol, 0.005%
bromphenol blue). Samples were run on SDS-polyacrylamide gel
electrophoresis (PAGE) (10% (w/v) total acrylamide) and subsequently
electrotransferred to nitrocellulose membranes. Immunoblotting was done
with anti-eNOSm or anti-GST at 1:1000, followed by a peroxidase-labeled
secondary antibody at 1:5000 and the chemiluminescence detection assay.
To analyze in vivo interactions, CHO-eNOS cells were
infected with SFV-NOSIP or SFV-mGFP expressing green fluorescent
protein (GFP) with a mitochondrial targeting sequence; noninfected
(‘native’) CHO-eNOS were used for control. After 15 h, the cells
were washed and lysed as above. For competition studies, samples were
supplemented with caveolin-1 peptides at 50 µM each. After
immunoprecipitation with 10 µl of anti-eNOSp, anti-NOSIP, or
respective preimmune sera (control), the immunoprecipitates were
resolved by SDS-PAGE. For immunoprinting, anti-eNOSm and anti-HAm were
used at 1:1000. To study the interaction of NOSIP and eNOS, freshly
isolated HUVEC were cultured in T75-flasks, washed with PBS, and
solubilized as above except that 20 mM
3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS)
replaced Triton X-100 in the lysis buffer. The cell extract was divided
into two equal fractions; one was used for immunoprecipitation with 20
µl anti-NOSIP (control: preimmune serum) and Western blotting using
anti-eNOSm. The other fraction was used for direct immunoprecipitation
of eNOS with anti-eNOSp, followed by Western blotting with anti-eNOSm
at 1:1000. To judge the completeness of eNOS precipitation, the
corresponding supernatant was subjected to another round of
immunoprecipitation with anti-eNOSp.
Measurement of eNOS activity and NO production
eNOS activity was quantified by following the conversion of
[14C]-L-arginine into
[14C]-L-citrulline (22)
. Briefly,
CHO-eNOS cells were washed 3x with ice-cold PBS, scraped into 25 mM
Tris-HCl, pH 7.4, 1 mM EDTA, 1 mM EGTA, 20 mM CHAPS including Complete
protease inhibitor mixture, and kept on ice for 45 min. After
centrifugation at 14,000 g for 10 min at 4°C, the
supernatant was removed. Cell extracts containing 10–30 µg of total
protein were mixed with 25 mM Tris-HCl, pH 7.4 including 16 mM
L-valine, 1 mM CaCl2, 1 mM NADPH, 10 µM
L-arginine, 3 µM tetrahydrobiopterine, 1 µM FAD, 1 µM FMN, and
0.2 µCi [14C]-L-arginine in the presence or
absence of 200 µM L-NNA. Endogenous calmodulin accounts for 1–2% of
the total protein, and supplementation with calmodulin did not further
increase the activity of eNOS (data not shown). After incubation for 30
min at 37°C, the reaction was quenched with 50 mM HEPES, pH 5.5, 2 mM
EDTA, 2 mM EGTA, then [14C]-L-arginine was
removed using Dowex AG 50WX-8 resin and the remaining
[14C]-L-citrulline was quantified by liquid
scintillation counting. To monitor the effects of NOSIP in intact
cells, CHO-eNOS were infected with SFV-mGFP or SFV-NOSIP for 8 h
(unless otherwise indicated), the cells were stimulated with 3 µM
A23187 in NOS assay buffer (10 mM HEPES, pH 7.4, 150 mM NaCl, 5 mM KCl,
1 mM MgSO4, 1 mM CaCl2, 1
mM glucose, 5 mM L-valine, and 100 µM L-arginine) for 3 to 60 min,
and the [14C]-L-arginine conversion assay was
done as above. NO release from native or infected CHO-eNOS was measured
by the NO chemiluminescence assay (11)
. Cells grown on
6-well plates were equilibrated with NOS assay buffer for 30 min at
37°C, followed by stimulation with 3 µM A23187 in 1 ml fresh NOS
assay buffer for 60 min at 37°C in the presence or absence of 100
µM L-NNA. Supernatant (100 µl) containing the oxidation product
NO2- was collected, treated
with 1% (w/v) sodium iodide in glacial acetic acid, and NO generation
was quantified by a chemiluminescence detector (Sievers) using ozone.
Immunofluorescence studies
CHO-eNOS cells grown on coverslips in 6-well plates were
infected with SFV-NOSIP. After 5 h, cells were washed with PBS and
fixed with ice-cold methanol for 10 min. After rinsing with PBS, cells
were blocked in PBS containing 10% heat-inactivated normal swine serum
(v/v), 0.5% Tween (v/v), and 0.1% bovine serum albumin (w/v) for
1 h to prevent unspecific binding of antibodies. Then cells were
incubated with anti-eNOSm (1:200) and anti-NOSIP (1:800) overnight.
After a washing step with PBS, cells were incubated for 1 h with
fluorescein isothiocyanate (FITC) -conjugated anti-mouse
immunoglobulins and Cy3-conjugated anti-rabbit immunoglobulins or
FITC-conjugated anti-rabbit immunoglobulins and biotinylated anti-mouse
immunoglobulins, followed by Texas Red conjugated to streptavidin.
After washing with PBS, coverslips were mounted on slides and evaluated
with an Olympus BX 60 epifluorescence microscope. NOSIP and eNOS
immunolabeling was documented at randomly selected areas with a 60x
lens using a CCD camera and stored for later examination. CHO-eNOS
cells transfected for the same period of time with SFV-mGFP were
examined as control.
Immunohistochemistry
Three adult Wistar rats were killed by chloroform inhalation.
The hearts were dissected and snap frozen, and cryostat sections (10
µm) were fixed in ice-cold acetone for 10 min and dried for 1 h.
Sections were blocked with heat-inactivated normal swine serum for
1 h, washed, and incubated with anti-NOSIP at 1:4000 overnight.
For negative control the primary antibody was omitted. Washed sections
were incubated with horseradish peroxidase-conjugated anti-rabbit
immunoglobulin (from goat). Peroxidase activity was visualized by
incubation with 0.02% diaminobenzidine, 0.08% nickel ammonium
sulfate, 0.02% H2O2 in
Tris-HCl, pH 7.6. Slides were evaluated and photographed using an
Olympus BX 60 microscope.
Subcellular fractionation and detergent extraction
Low density caveolin-enriched membrane fractions were isolated
(23)
. Briefly, noninfected and SFV-mGFP- or
SFV-NOSIP-infected CHO-eNOS cells grown on 100 mm dishes were washed
twice with PBS and directly scraped into 2 ml of 500 mM sodium
carbonate, pH 11. Cell suspensions were Dounce homogenized (10
strokes), sonicated (three 20-s bursts), adjusted to 42.5% sucrose by
the addition of 2 ml of 85% sucrose in 25 mM MES, pH 6.5, 150 mM NaCl
(MBS), and placed at the bottom of an ultracentrifuge tube. A
discontinuous sucrose gradient (5 ml of 30% sucrose and 3 ml of 5%
sucrose in MBS containing 250 mM sodium carbonate) was layered on top
of the cell suspension and centrifuged at 35,000 rpm for 18 h
using a SW40ti rotor (Beckman Instruments, Fullerton, Calif.).
Fractions of 1 ml each were collected and numbered from top to bottom,
1–12. Samples (20 µl each) were transferred to nitrocellulose
membranes by a slot blot apparatus (Schleicher & Schuell, Keene, N.H.)
and probed by anti-caveolin-1 (1:5000), anti-ß-COP (1:500),
anti--tubulin (1:1000), anti-eNOSm (1:1000), or anti-NOSIP (1:5000),
followed by the respective peroxidase-labeled secondary antibody
(1:5000) and the chemiluminescence detection kit.
To monitor the differential detergent solubility of eNOS in native or infected CHO-eNOS, the cells were equilibrated in NOS assay buffer for 30 min and stimulated with 3 µM A23187 in the same buffer. At 3 to 60 min after stimulation, cells were washed with ice-cold PBS and lysed in 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 2 mM EDTA containing 1% Triton X-100 for 1 h on ice. After centrifugation at 100,000 g for 30 min at 4°C, the supernatants (‘Triton-soluble fraction’) were mixed with 2x sample buffer and the pellets (‘Triton-insoluble fraction’) were resuspended in 1x sample buffer. Samples were subjected to SDS-PAGE and immunoblotting using anti-eNOSm at 1:1000 each, as detailed above.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 x 107 cotransformants identified seven
positive clones that demonstrated specific interaction in a rescreen
with the authentic pEG-eNOS1–486 bait but not
with two unrelated baits, pEG202 and pEG-cdNHE, used as negative
controls. Sequence analysis of one cloned cDNA revealed an open reading
frame coding for 301 amino acid residues with a calculated molecular
mass of 34 kDa; we named the corresponding protein NOSIP (eNOS
interacting protein; Fig. 1
). Database analysis revealed that the sequence of NOSIP is completely
identical to CGI25, a protein of unknown function recently identified
by comparative gene cloning with a homologue from Caenorhabditis
elegans (‘CEHOM’) as the probe. At the amino acid level, NOSIP
shows 46% identity with CEHOM and 29% identity with a plant protein
from Arabidopsis thaliana (‘ATHOM’), suggesting that a
family of NOSIP-like protein may exist in various species. The homology
between the various proteins is most prominent in their amino- and
carboxyl-terminal portions, whereas the center portion is highly
divergent (Fig. 1)
. The cellular function(s) of these proteins is yet
unknown.
|
Specificity of the interaction between eNOS and NOSIP
To demonstrate that NOSIP binds specifically to the eNOS oxygenase
domain, we performed coimmunoprecipitation experiments in
vitro and in vivo. Extracts of CHO cells stably
transfected with the human eNOS gene (CHO-eNOS) were incubated with
GST-NOSIP fusion protein (5 µg/ml, 80 nM) or GST-cdNHE for control.
Immunoprecipitation with anti-eNOSp and anti-NOSIP, followed by Western
blotting using anti-GST and anti-eNOSm, respectively, demonstrated the
in vitro coimmunoprecipitation of eNOS with GST-NOSIP and
vice versa. (Fig. 2A
). To show specific interaction of eNOS and NOSIP in
vivo, we infected CHO-eNOS with recombinant Semliki Forest virus
encoding GFP (SFV-mGFP) or HA-tagged NOSIP (SFV-NOSIP) and solubilized
the cells 15 h after infection. Immunoprecipitations were done
with anti-eNOSp and anti-NOSIP, and coimmunoprecipitated proteins were
probed by anti-HAm and anti-eNOSm, respectively. For control, the
corresponding preimmune sera were used. Western blotting of
immunoprecipitated proteins demonstrated that eNOS and NOSIP
specifically interact in a mammalian cell line (Fig. 2B
).
Stimulation of the cells with A23187 (3 µM for 15 min) had no effect
on the amount of coimmunoprecipitated proteins, indicating that the
interaction of eNOS and NOSIP is not under the control of
calcium-dependent eNOS activation.
|
Mapping of the eNOS binding region for NOSIP
To map the region in the oxygenase domain of eNOS that binds
NOSIP, we used the yeast two-hybrid system. Yeast reporter strain
EGY48/pSH18–34 was cotransformed with amino- and carboxyl-terminal
deletion constructs of the eNOS oxygenase domain, together with the
NOSIP-containing vector, and tested for the interaction by blue/white
screening and ß-galactosidase activity (Table 1
). Correct folding of the constructs was proved by homodimerization,
which was almost as efficient as for the full-length eNOS oxygenase
domain, eNOS1–486. Truncation of the 97
amino-terminal amino acid residues of eNOS oxygenase domain
(eNOS98–486) did not affect NOSIP binding. By
contrast, deletion of the carboxyl-terminal 120 residues
(eNOS1–366) and of both the amino- and
carboxyl-terminal regions (eNOS98–366) resulted
in a significant loss of binding affinity, indicating that the
carboxyl-terminal portion of the eNOS oxygenase domain is involved in
NOSIP binding. Because the binding region of the scaffolding domain of
caveolin-1 has been mapped to the same segment of the eNOS oxygenase
domain (24
, 25)
, we wondered whether caveolin-1 interferes
with NOSIP binding to eNOS. Solubilisates from CHO-eNOS cells
overexpressing NOSIP were used for coimmunoprecipitation of eNOS and
NOSIP in the presence of synthetic peptides of the caveolin-1. Addition
of 50 µM of caveolin-1 peptide (cav82–101)
covering the caveolin-1 scaffolding domain (amino acid residues
82–101) almost completely inhibited coimmunoprecipitation of NOSIP and
eNOS by anti-eNOSp, whereas a peptide comprising amino acid residues
61–81 of caveolin-1 (cav61–81) not
involved in eNOS binding or an unrelated peptide of similar size failed
to interfere with eNOS-NOSIP coimmunoprecipitation (Fig. 2C
). Hence, the eNOS oxygenase domain exposes a binding
segment for NOSIP that partially overlaps with its caveolin-1
attachment site.
|
Modulation of eNOS activity by NOSIP
To investigate whether NOSIP affects eNOS activity, we performed
[14C]-L-arginine conversion assays using
eNOS-containing extracts from CHO-eNOS cells. NOS inhibitor L-NNA (200
µM) almost completely blocked eNOS activity whereas GST-NOSIP (5
µM) failed to affect eNOS activity in this assay system (data not
shown). Because cell extracts contain detergent that might interfere
with NOSIP-eNOS complex formation in vitro, we studied the
effects of NOSIP overexpression in intact cells. CHO-eNOS cells
infected with SFV-NOSIP for 8 h, and thus overexpressing NOSIP,
demonstrated a 53% decrease in NO production after stimulation with
A23187 (3 µM for 60 min) compared to CHO-eNOS cells infected with
SFV-mGFP or control (set 100%; Fig. 3A
). Western blotting demonstrated that the amount of eNOS in
infected cells remained unchanged upon viral infection, indicating that
the lowered eNOS activity in NOSIP-overexpressing cells is not due
simply to a loss of protein (Fig. 3B
). We also asked whether
the inhibitory effect of NOSIP on eNOS activity in vivo
depends on external stimuli that might induce posttranslational
modifications of eNOS. No significant differences in eNOS activity were
seen between the cell lysates of noninfected and NOSIP-overexpressing
cells that had been exposed to A23187 for 60 min (Fig. 3C
).
Thus, our results indicate that the inhibitory capacity of NOSIP is
critically dependent on the integrity of cells and that this function
is not maintained in cell extracts as would be expected for
NOSIP-mediated posttranslational modifications of eNOS, which usually
persist after solubilization.
|
Effect of NOSIP on the subcellular redistribution of eNOS
The strict requirement of intact cells for eNOS inhibition by
NOSIP and the well-established dependence of eNOS activity on
translocation between various cell compartments prompted us to examine
the effect of NOSIP on the cellular trafficking of eNOS. Initially we
studied noninfected and SFV-mGFP- or SFV-NOSIP-infected CHO-eNOS cells,
stimulated them with 3 µM A23187 for various periods of time (0 to 60
min), and solubilized them by the nonionic detergent Triton X-100. The
presence of eNOS in the Triton-insoluble fraction was monitored by
Western blotting (Fig. 4A
; note that 30% of total cellular eNOS is associated
with this fraction in unstimulated cells). Upon stimulation, a major
fraction (43%) of eNOS present in the Triton-insoluble fraction of
native or SFV-mGFP-infected CHO-eNOS cells (set 100%) left the
detergent-insoluble fraction within 10 to 15 min of stimulation and
slowly returned after 30 to 60 min (Fig. 4B
). In contrast,
coexpression of NOSIP completely abrogated the transient loss of eNOS
from the Triton-insoluble fraction (Fig. 4A
, B
).
We tentatively conclude that NOSIP is involved in the subcellular
redistribution of eNOS, most probably between the detergent-insoluble,
caveolae-rich plasma membrane and detergent-soluble intracellular
compartments.
|
Effect of NOSIP expression on eNOS localization in CHO-eNOS cells
To determine the effect of NOSIP on subcellular localization of
eNOS, we infected CHO-eNOS cells with SFV-NOSIP. After 5 h cells
were fixed and immunolabeled for eNOS (Fig. 5A
) and NOSIP (Fig. 5B
). The infection efficiency
was reduced to allow for internal control, i.e., noninfected vs.
infected cells. For quantification, 20 randomly selected pictures of
eNOS immunolabeling and corresponding NOSIP immunolabeling were taken
in each experiment and evaluated for membrane-bound eNOS immunolabeling
and NOSIP transfection. Three different experiments performed on
different days were analyzed. Prominent staining of perinuclear
structures presumably representing the Golgi apparatus was seen in all
cells; in addition 77% of the noninfected CHO-eNOS cells displayed
eNOS immunofluorescence in the plasma membrane (Fig. 5A
). In
contrast, only 26% of NOSIP overexpressing cells showed eNOS
immunoreactivity in the plasma membrane; rather, eNOS appeared to be
restricted to intracellular compartments. Hence, overexpression of
NOSIP affects the subcellular trafficking of eNOS. In cells infected
with SFV-mGFP for the same period of time, the plasma membrane staining
was similar to that of noninfected cells (data not shown), thus ruling
out a direct virus effect on the subcellular localization of eNOS.
|
Subcellular fractionation of eNOS-expressing cells
We corroborated the notion that NOSIP affects the differential
distribution of eNOS in CHO cells by performing subcellular
fractionation studies. Sucrose floating gradients of sonicated CHO-eNOS
cells infected for 8 h with SFV-mGFP or SFV-NOSIP were collected
and probed for eNOS and NOSIP by Western blotting; for control,
noninfected CHO-eNOS cells were applied (Fig. 6
). Antibody probes against marker proteins caveolin-1 (associated with
caveolin-rich membranes), ß-COP (Golgi apparatus), and -tubulin
(cytoskeleton) were also applied. The distribution patterns of marker
proteins were similar or even identical for noninfected vs.
NOSIP-expressing cells: caveolin-1 was prominent in fractions 4 to 9,
ß-COP was present in fractions 8 to 12, and -tubulin was found in
fractions 10 to 12 (Fig. 6)
. By contrast, we observed a striking effect
of NOSIP expression on eNOS distribution: in noninfected or
SFV-mGFP-infected CHO-eNOS cells, eNOS cosedimented mainly with
caveolin-rich membrane fractions (4 to 8), whereas overexpression of
NOSIP caused a drastic shift of eNOS to fractions 8 to 12, representing
intracellular compartments such as Golgi and cytoskeleton. In infected
cells, NOSIP mapped to fractions 8 to 12 (Fig. 6
, outer right panel)
and thus cosedimented with eNOS, a finding that is compatible with the
hypothesized interaction of the two proteins. Thus, NOSIP
overexpression leads to a shift of eNOS from the plasmalemmal caveolae
to intracellular compartments, with important consequences for eNOS
activity (cf. Fig. 3B
).
|
Expression of NOSIP in endothelial cells
We applied Northern blot analysis, to study the relative
expression of NOSIP mRNA in human tissues. Significant amounts of a 1.2
kb NOSIP transcript were present in heart, brain, and lung (Fig. 7A
). A minor expression was also found in skeletal muscle,
whereas NOSIP expression in kidney, liver, pancreas, and spleen was
below detection level. The presence of a dominant 1.2 kb band suggests
that a single transcript is transcribed from the NOSIP gene although
alternative splice products of similar size cannot be ruled out.
In situ hybridization demonstrated the presence of NOSIP in
the endothelium of spleen vessels (data not shown); this finding may
indicate that NOSIP is expressed at varying levels in different
tissues. To address the cellular localization of NOSIP in
situ, we applied immunohistochemistry to a rich source of NOSIP
mRNA, i.e., the heart (see above). NOSIP immunoreactivity was abundant
in endothelial cells of the cardiac microvasculature in rat heart (Fig. 7B
). To further analyze the expression pattern of NOSIP in
endothelial cells, we performed Western blotting of total HUVEC
extracts and found a major band of 34 kDa, suggesting that a single
form of NOSIP is expressed in endothelial cells (Fig. 7C
).
Also, reverse transcription-PCR amplified a NOSIP cDNA fragment of
expected size from HUVEC total RNA (data not shown).
|
We demonstrated specific interaction of eNOS with NOSIP in endothelial
cells by coimmunoprecipitation experiments. NOSIP was
immunoprecipitated from crude extracts of HUVEC using rabbit
anti-NOSIP, and the immunoprecipitates were examined by Western
blotting with mouse anti-eNOSm. The presence of a 130 kDa band
demonstrated the specific interaction of eNOS and NOSIP in endothelial
cell extracts, whereas the control with the corresponding preimmune
serum was negative (Fig. 7D
, lanes 3 and 4). To estimate the
relative amount of total cellular eNOS interacting with NOSIP, we
directly immunoprecipitated eNOS from HUVEC extracts using rabbit
anti-eNOSp. Under the conditions of the experiment, eNOS precipitation
was almost complete since application of anti-eNOSp to the resultant
supernatant failed to produce a significant immunoprecipitate (Fig. 7D
, lanes 1 and 2). From the relative amounts of protein
obtained by direct or indirect (via NOSIP) immunoprecipitation, we
estimated that only a small fraction (2–5%) of total eNOS is
associated with NOSIP in HUVEC under basal condition.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, caveolin-1 and
-3 (12)
, and heat shock protein Hsp90 (17)
,
which directly stimulate or inhibit eNOS activity. The present study
sheds light on yet another layer of eNOS activity modulation through
regulated intracellular trafficking. Our results indicate that NOSIP, a
novel protein identified by a yeast two-hybrid screen, specifically
binds to the carboxyl-terminal region of the human eNOS oxygenase
domain. Overexpression of NOSIP in eNOS-expressing cells has two major
effects: inhibition of NO synthase activity and redistribution of eNOS
between cellular compartments.
In resting cells eNOS is located to variable extent at the plasma
membrane, where it is predominantly associated with caveolae, at the
Golgi apparatus and at perinuclear compartments (2)
. After
stimulation of bovine aortic endothelial cells by bradykinin,
estradiol, or A23187, eNOS translocates from the plasma membrane to
intracellular sites close to the nucleus and returns to the plasma
membrane upon prolonged stimulation (27
, 28)
. Subcellular
trafficking of eNOS is thought to be regulated by transient changes in
the phosphorylation and palmitoylation state of the enzyme
(29
30
31
32)
. Indeed, we have observed a similar distribution
pattern of eNOS in CHO-eNOS cells (this study), suggesting that native
endothelial cells and transfected CHO cells do not grossly differ in
their subcellular eNOS distribution. This distinct pattern is lost in
CHO-eNOS cells overexpressing NOSIP, i.e., these cells almost
completely lack plasma membrane-bound eNOS immunoreactivity (cf. Fig. 5
). In NOSIP-overexpressing cells, eNOS appears to be shifted to
intracellular sites that colocalize with Golgi and/or cytoskeletal
marker proteins. Also, the transient changes in detergent solubility of
eNOS after enzyme activation are fully arrested by overexpression of
NOSIP. Thus, our caveolin-1 displacement, immunofluorescence, and
subcellular fractionation studies converge at the conclusion that eNOS
is ‘trapped’ in the Triton-insoluble fraction, which likely
represents the structural components of the cytoskeleton and
not the caveolae-rich plasma membrane fraction (cf. Fig. 6
).
Indeed, our preliminary localization studies indicate that NOSIP is
often associated with filamentous structures in the cytoplasm of
transfected CHO-eNOS cells or HUVECs (P. König, unpublished
observations); the precise subcellular location of NOSIP in native
cells is being investigated.
Against this background it is tempting to speculate that NOSIP is a
modulator of eNOS enzyme activity by uncoupling eNOS from plasma
membrane caveolae, a hypothesis that is underlined by the finding that
the eNOS binding sites for caveolin and NOSIP overlap. Since
agonist-promoted eNOS activity is highest in the caveolar
microenvironment (9)
, uncoupling of eNOS from caveolae
should entail reduced NO production. This may explain the (partial)
inhibitory effect of NOSIP overexpression on NO release in intact
cells, which is not simply caused by the interaction of NOSIP and eNOS
and thus is not observed in activity assays in vitro.
Furthermore, NOSIP might be involved in the subcellular trafficking of
eNOS by promoting eNOS translocation from the plasma membrane to
intracellular compartments or by inhibiting the reverse transport. In
either case, overexpression of NOSIP reduces the fraction of plasma
membrane-associated eNOS, as we have observed for CHO-eNOS cells. One
may envisage that translocation of eNOS from the plasma membrane to
intracellular sites could subserve the local NO production, albeit at a
lower rate, at compartments such as the nucleus or mitochondria
(33
, 34)
. The distinct possibility remains that NOSIP
further modulates eNOS activity by direct inhibition though our
in vitro reconstitution studies do not support such a
notion.
At present we do not know whether NOSIP is a protein that has been specifically designed for eNOS regulation or whether it fulfills more general role(s) in the cell. The existence of NOSIP homologues in C. elegans and A. thaliana may favor the latter hypothesis; however, the function(s) of these proteins is completely unknown. On the other hand, NOSIP mRNA expression is most pronounced in heart, brain, and lung tissues, and NOSIP protein is present in the rat cardiac microvasculature, thus recapitulating the patterns of eNOS expression in endothelial cells of heart, brain, and lung. Expression of NOSIP in brain, and to a lesser extent in skeletal muscle, may imply a role of NOSIP also in the regulation of neuronal NOS. In fact, we have found a weak though significant interaction between the nNOS oxygenase domain and NOSIP in the yeast two-hybrid system (J. Dedio, unpublished observations).
In conclusion, we have identified a novel protein, NOSIP, which interacts with the carboxyl-terminal portion of the eNOS oxygenase domain. We suggest that NOSIP modulates activity of eNOS by uncoupling it from plasma membrane caveolae and by promoting the translocation of the enzyme from caveolae to intracellular compartments.
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 9, 2000.
Revision received June 16, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. E. Mortensen, L. N. Conley, J. Hedegaard, T. Kalstad, P. Sorensen, C. Bendixen, and A. Revhaug Regenerative response in the pig liver remnant varies with the degree of resection and rise in portal pressure Am J Physiol Gastrointest Liver Physiol, March 1, 2008; 294(3): G819 - G830. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Dudzinski and T. Michel Life history of eNOS: Partners and pathways Cardiovasc Res, July 15, 2007; 75(2): 247 - 260. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Ji, G. Ferracci, A. Warley, M. Ward, K.-Y. Leung, S. Samsuddin, C. Leveque, L. Queen, V. Reebye, P. Pal, et al. beta-Actin regulates platelet nitric oxide synthase 3 activity through interaction with heat shock protein 90 PNAS, May 22, 2007; 104(21): 8839 - 8844. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.-D. Lajoix, S. Badiou, S. Peraldi-Roux, T. Chardes, S. Dietz, C. Aknin, F. Tribillac, P. Petit, and R. Gross Protein Inhibitor of Neuronal Nitric Oxide Synthase (PIN) Is a New Regulator of Glucose-Induced Insulin Secretion Diabetes, December 1, 2006; 55(12): 3279 - 3288. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Chen, M. Medhora, J. R. Falck, K. A. Pritchard Jr, and E. R. Jacobs Mechanisms of activation of eNOS by 20-HETE and VEGF in bovine pulmonary artery endothelial cells Am J Physiol Lung Cell Mol Physiol, September 1, 2006; 291(3): L378 - L385. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Schilling, N. Opitz, A. Wiesenthal, S. Oess, R. Tikkanen, W. Muller-Esterl, and A. Icking Translocation of Endothelial Nitric-Oxide Synthase Involves a Ternary Complex with Caveolin-1 and NOSTRIN Mol. Biol. Cell, September 1, 2006; 17(9): 3870 - 3880. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Nakano, C.-J. Kwak, K. Fujii, K. Ikemura, A. Satake, M. Ohkita, M. Takaoka, Y. Ono, M. Nakai, N. Tomimori, et al. Sesamin Metabolites Induce an Endothelial Nitric Oxide-Dependent Vasorelaxation through Their Antioxidative Property-Independent Mechanisms: Possible Involvement of the Metabolites in the Antihypertensive Effect of Sesamin J. Pharmacol. Exp. Ther., July 1, 2006; 318(1): 328 - 335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Mukherjee, M. Tessema, and A. Wandinger-Ness Vesicular Trafficking of Tyrosine Kinase Receptors and Associated Proteins in the Regulation of Signaling and Vascular Function Circ. Res., March 31, 2006; 98(6): 743 - 756. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Yang and X.-F. Ming Recent advances in understanding endothelial dysfunction in atherosclerosis. Clin. Med. Res., March 1, 2006; 4(1): 53 - 65. [Abstract] [Full Text] [PDF]
|
|
