 |
INTRODUCTION |
Ha-Ras is a monomeric GTPase that has two types of lipid
modifications, both of which must occur in order for the protein to
bind efficiently to the plasma membrane (1-3). A C15 farnesyl isoprenoid is attached through a permanent thioether linkage to cysteine 186 at the C terminus (4-6). Farnesylation has been shown to
be a prerequisite to, but by itself not sufficiently strong for, full
Ha-Ras membrane binding (7-9). Ha-Ras requires a second lipid to
stabilize its membrane interaction. This second lipid is the fatty acid
palmitate (10), attached through thioester bonds to cysteine 181 or 184 (11, 12). Acylation of Ha-Ras is dynamic, with the palmitates having a
half-life considerably shorter than the polypeptide and undergoing
repeated cycles of removal and replacement (13, 14). Very little is
known of how this S-acyl modification occurs or of its
possible regulation, in part because purification of enzymes that might
attach palmitate has proven to be difficult (15-18). Further, no
structural motifs or signal sequences for palmitoylation have been
identified (7). Only recently has a cytosolic acyl-protein thioesterase
(APT1) been isolated that can remove palmitate from heterotrimeric G protein 
-subunits and from Ha-Ras in vitro (19). Despite
the transience of a thioacyl group, the presence of cysteines that can
be palmitoylated dramatically increases the extent of farnesylated Ha-Ras membrane binding from ~10% to >95% (7, 20).
For Ha-Ras, palmitoylation and plasma membrane targeting are clearly
necessary for biological activity, because Ha-Ras mutants that lack
both palmitates are poorly transforming in NIH 3T3 cells (7, 8, 20).
Furthermore, an Ha-Ras that has only one site for palmitoylation is
partially misdirected to internal membranes and is only weakly active
(7). Thus, permanent interference with palmitoylation decreases Ha-Ras
function. These results suggest that regulation of palmitoylation could
provide a novel approach for controlling Ha-Ras oncogenicity.
There is growing evidence (11, 21-25) that acylation of a variety of
signaling proteins, including Ha-Ras, may be important for targeting
and organizing a portion of these proteins in specialized subdomains of
membranes ("rafts," caveoli, or detergent-resistant membranes).
However, the relationships between palmitoylation, submembrane
location, and signaling by any of these proteins (26) are difficult to
study, since there are few techniques through which the acylation state
of such signaling proteins can be varied.
In a small number of palmitoylated proteins, acylation can be regulated
by agonist stimulation (27). Isoproterenol interaction with the
-adrenergic receptor results in activation of the receptor and of
the Gs subunit, and this activation correlates with
increased palmitate turnover on both the receptor and the
Gs protein (28-32). Serotonin treatment of membranes
derived from rat brain cells is reported to increase palmitate labeling
of a number of G protein -subunits (Gq, Go,
Gi, and Gs) (33), and agonist-promoted
palmitate exchange has been reported for Gq (34). For
these G proteins, deacylation appears to be an important regulated
step in changes that occur in response to ligand. In endothelial cells,
a bradykinin-triggered increase in depalmitoylation of the endothelial
nitric-oxide synthase has been observed, coincident with activation of
the protein (35, 36). Thus, palmitoylation of some proteins is, through
as yet unknown mechanisms, responsive to external signals. However, for Ha-Ras proteins, no changes in palmitoylation following the rapid and
transient activation of growth factor-dependent pathways
have been reported. The compound 2-bromo-palmitate has recently been reported to decrease palmitate labeling of the Src family kinases Fyn
and Lck but had very little effect on membrane binding of an Ha-Ras
protein (37).
Another agent that may affect protein palmitoylation is nitric oxide
(NO· or its metabolites, NOx). The nitrosothiol compound S-nitrosocysteine
(SNC)1 has been reported to
cause a decrease in palmitate incorporation into many proteins in
neuronal cells (38) Two of these proteins were identified as the
synaptic vesicle protein, SNAP-25, and the growth cone-associated
protein, GAP-43. This work suggested that S-acylation of
neuronal growth cone proteins could be manipulated by exogenous
exposure of neurons to nitric oxide or NOx-producing compounds.
Recent studies have reported that activation of Ras proteins can also
be affected by nitric oxide. In the Jurkat T cell line, and especially
in Jurkat cells depleted of glutathione by buthionine sulfoximine
treatment, more of the endogenous Ras proteins were reported to be
GTP-bound after exposure to nitric oxide (39). In vitro
studies also indicated that the cysteine at position 118 becomes
nitrosylated during treatment of recombinant Ha-Ras with nitric oxide
or SNC (40). Additionally, in rat and mouse cortical neuron cultures,
the endogenous Ras proteins were reported to bind increased amounts of
GTP after exposure to NOx donors or stimulation of nitric-oxide
synthase activity by N-methyl-D- aspartate (41).
These results indicated that Ras itself and/or proteins that regulate
its GTP state were susceptible to nitric oxide or its metabolites (42).
However, no information on the effects of nitric oxide or other
oxidants on Ha-Ras palmitoylation has been reported.
This study was designed to determine if effects of nitric oxide on
protein palmitoylation could also be observed in cells other than
neurons and specifically whether palmitoylation of the Ha-Ras protein
could be influenced by compounds that produce nitric oxide or its
metabolites. The results identify S-nitrosocysteine and
characterize parameters for its effective use as the first method for
deliberate, external manipulation of palmitoylation of Ha-Ras.
| |
MATERIALS AND METHODS |
Cell Culture--
COS-1 cells or NIH 3T3 cells expressing
oncogenic v-Ha-Ras(Arg12, Thr59),
v-Ha-Ras(Ser181) (B. Willumsen, Copenhagen), cellular
(wild-type) Ha-RasWT, or a mutated form of Ha-RasWT, Ha-Ras(C118S) (L. Quilliam, Indianapolis) were grown at 10% CO2 in
Dulbecco's modified Eagle's medium supplemented with 10% bovine calf
serum (Hyclone, Logan, UT). RAW 264.7 cells were cultured at 10%
CO2 in Dulbecco's modified Eagle's medium supplemented
with 10% heat-inactivated fetal bovine serum (Hyclone). Undifferentiated PC-12 cells (J. H. P. Skene, Duke
University) were grown on 10 µg/ml laminin-coated dishes at 5%
CO2 in RPMI medium supplemented with 5% fetal calf serum
and 10% heat-inactivated horse serum.
Preparation of NO Donors and Oxidants--
A 100 mM
S-nitrosocysteine solution was prepared less than 30 min
before each experiment by solubilizing equimolar amounts of sodium
nitrate and L-cysteine in water and adding concentrated hydrochloric acid to initiate formation and release of nitric oxide gas
(43). Control, nitric oxide-depleted SNC was prepared by allowing a
solution of SNC to stand at least 48 h at room temperature. The
acidic SNC was neutralized with NaOH before the addition to cells, and,
to maintain pH during the various treatments, supplementary buffer (50 mM HEPES, pH 7.2) was included in all media. Additional 4 mM SNC was added every 30 min unless otherwise noted.
S-Nitrosoglutathione solution was made and used similarly,
substituting reduced glutathione (Sigma) for L-cysteine.
Stock solutions (100 mM) of SIN-1, NOC-18, NOC-15,
8-bromo-cyclic GMP (all from Alexis, San Diego), or hydrogen peroxide
(Fisher) were made in water immediately before use and diluted into the
HEPES-buffered growth medium. Stock
S-nitroso-N-acetyl-DL-penicillamine (SNAP; Alexis) was made by dilution in dimethyl sulfoxide.
Radiolabeling, Immunoprecipitation, and
Immunoblotting--
Cells were labeled with 0.5-1 mCi/ml
[3H]palmitate (NEN Life Science Products) in medium
containing nonessential amino acids, 50 mM HEPES, pH 7.2, 25 mg/ml cycloheximide, and 10% calf serum (44). Cycloheximide was
used to inhibit protein synthesis and thereby to avoid labeling newly
synthesized Ha-Ras with [3H]palmitate or palmitate that
had been metabolized to [3H]amino acids. This technique
therefore detects palmitate replacement on pre-existing proteins rather
than the initial lipid modification of newly synthesized proteins.
Depalmitoylation was measured by labeling cells with
[3H]palmitate for 3 h and then incubating for
varying time periods in similar but nonradioactive medium containing
200 µM palmitic acid (Sigma). Immunoprecipitates were
formed as described (45) by incubating cell extracts with a mouse
monoclonal antibody specific for Ha-Ras (146-3E4; Quality Biotech,
Camden, NJ), mouse anti-transferrin receptor antibody
(Zymed Laboratories Inc., South San Francisco, CA), or
rabbit anti-caveolin-1 antibody (Transduction Laboratories, Lexington,
KY). Samples were resuspended in a special electrophoresis sample
buffer without -mercaptoethanol (2% SDS, 10 mM
Na2PO4, pH 7.0, 10% glycerol, 50 mM dithiothreitol, and 0.02% bromphenol blue). Immunoblots
were developed with the Ha-Ras-specific monoclonal antibody 146-3E4,
and proteins were visualized using alkaline phosphatase-linked
detection (Vector; Burlingame, CA). Amounts of palmitoylated Ha-Ras
were quantified from fluorograms by using the program ImageQuant
(Molecular Dynamics, Inc., Sunnyvale, CA) and corrected for variations
in cell numbers between samples and in protein recovery in
immunoprecipitates by dividing those values by the amounts of Ha-Ras
similarly quantified from immunoblots of the same membrane used for fluorography.
GTP/GDP Binding to Ha-Ras--
Cells were grown overnight in
medium containing 1% dialyzed calf serum, radiolabeled with
32P-inorganic phosphate (NEN Life Science Products) for
4 h, and treated for 30 min with NO-depleted or freshly prepared
SNC. Ha-Ras proteins were captured by immunoprecipitation, and bound
nucleotides were eluted, separated by thin layer chromatography, and
detected by autoradiography as described (45, 46). The autoradiograms were scanned and quantitated using ImageQuant software (Molecular Dynamics). The percentage of GTP bound to Ha-Ras was calculated from
these values using the equation, % GTP-bound Ras = GTP/(GTP + (1.5 × GDP)).
An indirect method was also used to monitor GTP-sensitive changes in
the effector domain of Ha-Ras. NIH 3T3 or PC12 cells expressing either
Ha-Ras61L or Ha-RasWT were treated for 15 min with 4 mM SNC
as described above and then lysed in 250 µl of a high salt buffer (50 mM Tris, pH 8.0, 10 mM MgCl2, 500 mM NaCl, 1.0% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS,
and protease inhibitors), sonicated for 5 s, and clarified by
centrifugation at 4 °C for 10 min at 11,000 × g.
Glutathione beads to which were bound 2 µg of GST-RBD (a chimeric
recombinant protein consisting of the glutathione
S-transferase (GST) protein fused to the Ras binding domain
(residues 1-140) of Raf-1; a gift from S. Campbell, Chapel Hill, NC)
were incubated with NIH 3T3 or PC12 lysates (500 µg of total protein)
for 30 min at 4 °C. The beads were then washed in Hepes-buffered
saline wash buffer (20 mM HEPES, 120 mM NaCl, 1% Nonidet P-40, and 10 mM MgCl2) and
resuspended in electrophoresis sample buffer, and proteins were
resolved by SDS-PAGE. GTP-bound Ha-Ras proteins captured on the GST-RBD
were detected by immunoblotting.
Preparation of Subcellular Fractions--
Subcellular fractions
were prepared by lysis of control or treated cells in hypotonic buffer
(1 mM Tris, pH 7.4, 1 mM MgCl2, 1 mM pefabloc, 1 µM leupeptin, 2 µM pepstatin, and 0.1% aprotinin (Calbiochem)) followed
by the addition of NaCl to adjust the ionic strength to 0.15 M as described previously (47). A portion (200 µl) of the
total lysate was removed, and the remainder was subjected to
centrifugation at 100,000 × g for 30 min to separate
cytosolic from membrane-containing fractions. The proteins in the
separated fractions were precipitated with 10 ml of acetone for 1 h at 4 °C, collected by centrifugation at 3,000 rpm for 30 min, and
dissolved in 100 µl of electrophoresis sample buffer. Equal volumes
of each fraction were separated by SDS-PAGE and transferred to
polyvinylidene difluoride membrane (DuPont), and v-Ha-Ras was detected
by immunoblotting.
Activation of ERK1/ERK2--
NIH 3T3 and COS-1 cells were
incubated in medium lacking serum for 2 h. PC-12 and RAW 264.7 cells were not serum-starved, since the ERK proteins failed to become
entirely quiescent even after greater than 12 h of incubation
without serum. However, S-NO-Cys stimulation of ERK
phosphorylation in PC-12 cells was also seen in serum-starved cells
(data not shown). Fresh or nitric oxide-depleted SNC was diluted to 4 mM in serum-containing medium, and cells were treated for
the times indicated. Proteins were separated by SDS-PAGE and
transferred to polyvinylidene difluoride membranes. Both phosphorylated
and unphosphorylated ERK1 and ERK2 proteins were detected by
immunoblotting with ERK1 antibody (Santa Cruz Biotechnology, Inc.,
Santa Cruz, CA). The blot membranes were then stripped as described
(48) and were reprobed using a monoclonal antibody specific for the
phosphorylated forms of ERK1 and ERK2 (New England Biolabs, catalog no.
9106S) followed by development with ECL (Pierce) using the
manufacturer's protocol.
| |
RESULTS |
S-Nitrosocysteine Treatment Leads to both Increases and Decreases
in Protein Palmitoylation--
Previous results had shown that
palmitate incorporation into SNAP-25 and GAP-43 decreased when the
neuronal cell line, PC-12, or primary cultures of rat dorsal root
ganglion cells were labeled with [3H]palmitate in the
presence of S-nitrosocysteine (49). However, preliminary
experiments indicated that SNC caused little effect on palmitate
labeling of Ha-Ras61L or Ha-RasWT expressed in PC-12 cells (data not
shown). To determine if SNC could alter protein palmitoylation in
another cell type (NIH 3T3 fibroblasts), three different classes of
palmitoylated fibroblast proteins were examined: the transferrin
receptor, which is localized in endocytotic vesicles and
clathrin-coated pits in the plasma membrane (50); caveolin, a second
transmembrane protein, localized in caveolae (51); and cellular
(wild-type) Ha-Ras (Ha-RasWT), which is plasma membrane-associated through hydrophobic lipid modifications (4, 7).
When NIH 3T3 cells stably expressing Ha-RasWT were treated with SNC, a
small decrease in palmitate incorporation into the transferrin receptor
and a more pronounced decrease in labeling of caveolin were detected
(Fig. 1A). This effect was
similar to the broad decrease in palmitate labeling caused by SNC in
neuronal cells. In contrast, [3H]palmitate incorporation
into Ha-RasWT was increased by SNC treatment. Cells treated with
NO-depleted SNC showed no changes in Ha-RasWT palmitoylation (data not
shown), indicating that neither the cystine nor other components formed
as nitric oxide is released were responsible for the observed effects
on palmitate incorporation. No change in the amount of
[3H]palmitate that entered cells or
[3H]palmitoyl CoA that was produced had been observed
previously in the SNC-treated neurons (49). The current results further confirmed that changes in protein palmitoylation did not arise from a
general effect of NOx on uptake of radioactive fatty acid or
its activation to acyl-CoA, since two proteins responded to SNC with
decreased labeling while a third, Ha-RasWT, showed increased labeling.
These results demonstrated that SNC could alter protein palmitoylation
in NIH 3T3 cells and indicated that specific proteins showed distinct
responses to this compound.
View larger version (48K):
[in this window]
[in a new window]
|
Fig. 1.
S-Nitrosocysteine and
S-nitrosoglutathione increase
[3H]palmitate incorporation into Ha-Ras.
A, NIH 3T3 cells expressing Ha-RasWT were labeled for 1 h with [3H]palmitate and simultaneously treated with
NO-depleted (C) or 4 mM freshly prepared SNC.
The endogenous transferrin receptor, caveolin, or transfected Ha-RasWT
proteins were isolated by immunoprecipitation, separated by SDS-PAGE,
and transferred to polyvinylidene difluoride membranes. Membranes were
exposed to film for 69 days (transferrin receptor) or 28 days (caveolin
and Ha-RasWT). Recovery of an equivalent amount of protein under both
conditions was monitored after film exposure by detecting each protein
on the membrane by immunoblotting with appropriate antibodies.
B, NIH 3T3 cells expressing Ha-RasWT or v-Ha-Ras were
labeled for 30 min with [3H]palmitate in medium
containing NO-depleted SNC (C), 4 mM SNC, 4 mM SNG, 1.1 mM SNAP, 4 mM NOC-15, 2 mM SIN-1, 1 mM 8-bromo-cyclic GMP
(cG), or 200 µM H2O2.
Ha-Ras was immunoprecipitated, resolved by SDS-PAGE and detected by
fluorographic exposure for 7 days (upper panel). Palmitate
incorporation was corrected for variations in protein recovery as
quantified by immunoblotting membranes after film exposure (lower
panel). Numbers below each lane
are averages of three (SNC), or two (all others) determinations of
protein-corrected [3H]palmitate incorporation expressed
relative to that of Ha-Ras from control (C) cells.
C, NIH 3T3 cells expressing v-Ha-Ras were labeled for 1 h with [3H]palmitate in medium containing NO-depleted SNC
(C), 4 mM freshly prepared SNC, or 4 mM SNG. v-Ha-Ras was immunoprecipitated, resolved by
SDS-PAGE, and detected by fluorographic exposure for 21 days. Protein
recovery was monitored by immunoblotting with antibody 3E4-146.
D, NIH 3T3 cells expressing v-Ha-Ras were treated for
30 min with either NO-depleted SNC (C) or 4 mM
freshly made SNC. Subcellular fractions were prepared; equal portions
of the total (T), soluble (S), and particulate
(P) fractions were resolved by SDS-PAGE; and v-Ha-Ras was
detected by immunoblotting. Data are representative of three
different experiments.
|
|
S-Nitrosoglutathione Also Increases Ha-Ras Palmitate
Incorporation--
To determine if other nitric oxide producers or
oxidants might also be useful agents for study of palmitate turnover on
Ha-Ras, NIH 3T3 cells expressing Ha-RasWT were labeled with
[3H]palmitate in the presence of various compounds:
S-nitrosoglutathione (SNG), a tripeptide nitrosothiol; SNAP,
a nitrosothiol of penicillamide; SIN-1, a synthetic compound that
releases both nitric oxide and superoxide, generating peroxynitrite
(ONOO
); NOC-15, a synthetic compound that produces nitric
oxide radicals; and the oxidant, hydrogen peroxide. When palmitate
incorporation was corrected for protein recovery in the
immunoprecipitates, increased palmitate incorporation was observed with
S-nitrosoglutathione (4 mM) (Fig.
1B), although the increase was more modest than the effect
observed with SNC. The protein-corrected averages from three
experiments showed that SNG stimulated Ha-RasWT palmitate labeling
approximately half as well as SNC. SNG also increased palmitate
labeling of the oncogenic v-Ha-Ras protein (Fig. 1C). Other
NOx producers (SNAP, SIN-1, and NOC-15) and the oxidant
H2O2 had only small effects on palmitate
incorporation on Ha-RasWT in NIH 3T3 cells (Fig. 1B). These
results suggested that small molecular weight nitrosothiols like SNC
and SNG were the most effective agents for increasing palmitate
turnover on Ha-Ras in NIH 3T3 cells.
SNC Alters Ha-Ras Palmitoylation through Cyclic GMP-independent
Pathways--
A well known mechanism of nitric oxide-mediated effects
is through stimulation of guanylyl cyclase activity, which in turn increases amounts of cellular cyclic GMP (52, 53). To examine whether
the SNC-triggered increase in [3H]palmitate incorporation
in Ha-Ras might be caused through a cGMP-dependent pathway,
cells expressing Ha-RasWT were labeled with [3H]palmitate
and simultaneously treated with 8-bromo-cyclic GMP. No change in
palmitate incorporation was detected in cells treated with
8-bromo-cyclic GMP for 30 min (Fig. 1B) or in two additional experiments that tested the effect of 8-bromo-cyclic GMP on palmitate incorporation over a 2-h time course (data not shown). This implied that the observed increase in palmitate incorporation caused by SNC
occurred independently of possible effects of SNC on guanylyl cyclase.
Characteristics of SNC-triggered Increases in Ha-Ras Palmitate
Incorporation--
To further characterize the effect of SNC on Ha-Ras
palmitoylation, cells expressing Ha-RasWT were labeled with
[3H]palmitate and simultaneously treated with a broad
range of concentrations of SNC in the presence of 10% calf serum.
Serum was included in the medium during SNC treatment because it
contains growth factors whose removal would affect Ras-GTP binding and
conformation and because this better replicates the in vivo
conditions in which cells in tissues will encounter nitric oxide.
However, serum proteins (e.g. serum albumin) can be
nitrosylated and delay entry or reduce the amounts of NOx
available to enter the cells (54, 55). Millimolar amounts of SNAP have
been reported to be needed to produce micromolar amounts of free NO in
intact cells (56). Therefore, SNC concentrations ranging from
micromolar to millimolar were tested.
Using micromolar amounts of SNC, palmitate incorporation into Ha-RasWT
was unaffected (neither increased nor decreased) after a 30-min
exposure (Fig. 2). Between 1 and 2 mM SNC a stimulatory effect on palmitate labeling of
Ha-RasWT became observable. At 4 mM SNC, palmitate
incorporation was an average of 3.5-fold higher than in untreated
cells. In three additional experiments, a 1-h treatment with 4 mM SNC induced increases in Ha-RasWT palmitate incorporation that ranged from 2.5- to 6.1-fold (data not shown). At a
concentration of 8 mM, Ha-RasWT palmitate labeling was
increased 3-fold at 15 min, but loss of cellular adhesion began to
occur by 30 min, and additional samples were not collected. These data suggested that SNC could be used most effectively at 4 mM
to induce increased Ha-Ras palmitate incorporation.
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 2.
Palmitate incorporation into Ha-Ras can be
increased by increasing SNC dose. NIH 3T3 cells expressing
Ha-RasWT were labeled for 30 min with [3H]palmitate in
medium containing between 0.1 and 4 mM SNC. Ha-RasWT was
isolated by immunoprecipitation, resolved by SDS-PAGE, and detected by
9-day fluorographic exposure. [3H]palmitate incorporation
into Ha-RasWT was corrected for amount of protein in immunoprecipitates
by immunoblotting membranes after film exposure. The line
shows the averages of two independent experiments.
|
|
To determine how quickly SNC affected palmitoylation and if these
changes were transient or sustained, cells expressing Ha-RasWT were
again labeled in the presence of 4 mM SNC. In control,
untreated cells, Ha-RasWT [3H]palmitate labeling was
detectable within 15 min and reached maximal
[3H]palmitate incorporation by 1-2 h. SNC treatment
produced a rapid increase in Ha-RasWT [3H]palmitate
labeling, doubling the amount of [3H]palmitate
incorporated at the earliest time point, 15 min (Fig. 3A) and reaching maximal
incorporation by 1 h. After 2 h of treatment, the amount of
radioactive [3H]palmitate still present on Ha-RasWT in
the SNC-treated cells diminished, so that control samples and
SNC-treated samples showed nearly equal amounts of total
[3H]palmitate attached. This decrease did not reflect
death of the treated cells, because cells showed no loss of viability
after a 2-h exposure if SNC-containing medium was replaced with regular growth medium (data not shown). In addition, the cellular signaling pathway involving the ERK proteins could regain activity after SNC
exposure (see below).
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 3.
SNC rapidly increases
[3H]palmitate labeling of both cellular and
v-Ha-Ras. NIH 3T3 cells stably expressing Ha-RasWT (A)
or v-Ha-Ras (B) were labeled with
[3H]palmitate for the indicated times and simultaneously
treated with NO-depleted solution of SNC (Control) or with
additions of fresh 4 mM SNC every 30 min (SNC).
Ha-Ras proteins were detected by fluorographic exposure for 6-14 days
and quantified as in Fig. 2, assigning the sample with the highest
protein-corrected amount of [3H]palmitate a value of 1. The black line and squares indicate
control samples; SNC-treated samples are represented by the
gray line and triangles.
|
|
To examine in more detail the ability of SNC to stimulate palmitate
labeling of an activated, oncogenic Ras protein, cells expressing
v-Ha-Ras(Arg12, Thr59) were labeled in the
presence of 4 mM SNC. As with Ha-RasWT, SNC noticeably
increased the amount of [3H]palmitate incorporated into
the v-Ha-Ras protein (Fig. 3B). The increase over controls
was rapid, doubling the amount of [3H]palmitate
incorporated within 15 min. Compared with Ha-RasWT, [3H]palmitate labeling into control v-Ha-Ras appeared to
reach maximal incorporation more rapidly (within 30 min). For the
SNC-treated cells, [3H]palmitate incorporation again
declined after reaching a maximum between 30 and 60 min. Additional
experiments showed SNC-induced increases in maximal v-Ha-Ras palmitate
incorporation ranging from 1.7- to 4.5-fold. Thus, the effects of SNC
on palmitoylation were not limited to cellular Ha-Ras. SNC could be
used to increase the rate of [3H]palmitate labeling of
both Ha-RasWT and oncogenic v-Ha-Ras within minutes, and this effect
could be sustained for at least 1 h.
SNC Increases Depalmitoylation of Ha-Ras--
An increase in
[3H]palmitate labeling can reflect attachment of
[3H]palmitate to previously unmodified cysteines on
mature Ha-Ras or accelerated removal of unlabeled palmitates with rapid
replacement by [3H]palmitates. The use of cycloheximide
during the labeling periods prevented synthesis of Ha-Ras with newly
created nonacylated cysteines. However, the stoichiometry of Ha-Ras
palmitoylation is unknown, so it was unclear if mature Ha-Ras with one
nonacylated cysteine might be present and be a possible source of the
available sites for increased palmitoylation. A pulse-chase technique
was therefore used to determine if turnover of previously attached
palmitates remained constant or if SNC altered the rate of Ha-Ras depalmitoylation.
Using Ha-Ras that was prelabeled with [3H]palmitate, a
gradual loss of radioactivity from control samples was observed.
However, in Ha-Ras from SNC-treated cells, less
[3H]palmitate remained at each time point. This was true
for Ha-RasWT and both the phosphorylated and unphosphorylated forms of
v-Ha-Ras (Fig. 4A).
S-Nitrosoglutathione also stimulated removal of
[3H]palmitate from Ha-RasWT (data not shown) and thus
mimicked SNC in this effect. These results indicated that SNC hastened
removal of [3H]palmitate from already acylated Ha-Ras
protein. In the previous labeling experiments, it appeared that during
short exposure times SNC accelerated removal of unlabeled palmitates,
which were rapidly replaced with [3H]palmitate, leading
to the observed increase in palmitate labeling. It was possible that
the decline in radioactive [3H]palmitate attached to
SNC-treated Ha-Ras that was seen after longer SNC exposures (see Fig.
3) was the result of continued SNC-stimulated removal of palmitates (by
then largely [3H]palmitates) that was no longer matched
by replacement with radioactive lipid. SNC thus appeared to be able to
increase Ha-Ras depalmitoylation for 1-2 h and perhaps longer.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 4.
SNC accelerates palmitate removal from
Ha-RasWT and v-Ha-Ras. A, NIH 3T3 cells expressing
Ha-RasWT or v-Ha-Ras were labeled with [3H]palmitate for
3 h and then transferred to medium containing 200 µM
nonradioactive palmitate and either NO-depleted (C) or 4 mM fresh SNC for the times indicated. Ha-Ras proteins were
detected and quantified as in Fig. 2, after fluorographic exposure for
13-56 days. Additional experiments using 2 mM SNC showed a
similar acceleration in palmitate removal (data not shown).
B, graphs show data collected from control cells during a
24-h chase with Ha-RasWT or
v-Ha-Ras(Ser181-Cys184) and lines calculated
using a double exponential decay equation. C, graphs show
data collected from chases of Ha-RasWT (n = 2) and
v-Ha-Ras (n = 3) and lines calculated using a double
exponential decay equation. The black line and
squares represent cells treated with NO-depleted SNC;
gray line and triangles represent
SNC-treated samples. Phosphorylated v-Ha-Ras also showed faster loss of
palmitate in SNC-treated samples (not shown).
|
|
To examine in more detail the kinetics of both normal and
SNC-stimulated palmitate removal, the data from multiple pulse-chase experiments were analyzed. In control, untreated cells, two phases of
loss of [3H]palmitate from Ha-RasWT could be detected
(Fig. 4B). A more rapid rate predominated during the early
stages (0-30 min) of the chase procedure. A second, slower phase of
deacylation also appeared to be occurring during this time and became
more evident at longer time points (1-24 h). It was necessary to
follow deacylation for over 12 h in order to observe a decline of
greater than 80% of the initial [3H]palmitate levels.
Because of the compound nature of the two overlapping deacylation
curves, simple visual extrapolation of palmitate loss to the time when
50% of the initial amount was removed was difficult. In addition, a
single phase exponential decay curve fit the data poorly (data not
shown; note that the graphs have linear y axes, not
logarithmic). Similar complex kinetics of palmitate removal have also
been reported by others (13). The data were therefore analyzed using a
double exponential decay model. The calculations for a pulse-chase
experiment using control, untreated cells indicated that 50% of the
[3H]palmitate initially present in Ha-RasWT would be lost
in 2.8 h if all loss occurred at the faster rate of palmitate removal.
Because Ha-Ras has two sites for palmitoylation, it was of interest to
learn if the two phases of palmitate removal occurred because of
different rates of deacylation of Cys181 and
Cys184. To test this possibility, a pulse-chase experiment
was performed with a v-Ha-Ras(C181S) protein with only
Cys184 available for palmitoylation. Loss of ~75% of
[3H]palmitate from the single Cys184 site
occurred very rapidly, in less than 30 min, but the remaining radioactivity continued to show the slow phase of decline (Fig. 4B). This result agrees well with the much faster
deacylation of Ha-Ras(Ser181) observed in COS-1 cells (13).
This experiment indicated that Cys184 was easily deacylated
when it was the only site for palmitoylation. Despite this very rapid
palmitate removal even in untreated, control cells, the
Ser181, Cys184 protein, which is farnesylated,
retained 50-70% of the membrane binding of fully lipidated v-Ha-Ras
(Ref. 7 and data not shown). Because the farnesyl group of completely
nonacylated v-Ha-Ras(Ser181, Ser184) is, by
itself, unable to support more than 10% membrane binding (7, 20), it
therefore appeared that Cys184 of
v-Ha-Ras(Ser181, Cys184) must be
repalmitoylated rapidly and continuously in order to sustain this
significant level of membrane attachment. Further advances in detection
techniques for palmitates will be needed to distinguish if
Cys181 and Cys184 might have different but
swift rates of depalmitoylation.
For the current studies, however, it was important that deacylation of
Cys184 continued to show the prolonged phase of decline in
[3H]palmitate. This suggested that the later phase of
decrease was likely to reflect repeated removal of
[3H]palmitates and replacement with lipids of lower
specific activity. Similar to the Ha-RasWT from control cells, the
Ha-RasWT and v-Ha-Ras proteins from SNC-treated cells showed both
phases of decline in [3H]palmitate (data not shown).
Therefore, further analysis on the effect of SNC on deacylation was
focused on the initial removal of palmitates during the faster, early
phase within the first 2 h.
For Ha-RasWT, SNC treatment decreased the calculated time at which 50%
of the palmitate would be removed at the faster rate from an average of
2.8 h to 0.4 h (Fig. 4C). SNC treatment thus caused an ~6-fold acceleration in this rate of depalmitoylation of
Ha-RasWT. Depalmitoylation of v-Ha-Ras was also stimulated by SNC. The
calculated time at which 50% of the palmitate would be removed at the
faster rate for v-Ha-Ras in control cells was 1.2 h. SNC
accelerated this rate of depalmitoylation by approximately 4-fold, with
a calculated loss of half of the original [3H]palmitate
in 0.3 h (Fig. 4C). The ~2-fold faster basal rate of
depalmitoylation of v-Ha-Ras (1.2 h) compared with Ha-RasWT (2.7 h) was
reproducible and was most closely related to the activation state of
Ha-Ras.2 Thus, SNC
accelerated palmitate removal from both Ha-RasWT and v-Ha-Ras by
4-6-fold. During SNC treatment, deacylation of both cellular and
oncogenic forms occurred at similar rates. From the pulse-chase and
labeling data it therefore appeared that SNC caused substantial and
sustained acceleration of the deacylation of both Ha-RasWT and v-Ha-Ras
proteins, followed by rapid readdition of palmitates to the newly
available sites. These results therefore indicated that SNC
significantly increased overall palmitate cycling on Ha-Ras.
Examination of v-Ha-Ras membrane binding from SNC-treated cells showed
that 30 min of exposure did not cause a perceptible decrease in
membrane attachment (Fig. 1D). These data provided a further
indication that SNC-accelerated deacylation was likely to be balanced
by readdition of palmitates and that membrane binding was maintained
during modest lengths of exposure.
Ha-Ras GTP Binding Is Decreased during SNC Treatment--
SNC's
ability to stimulate Ha-Ras depalmitoylation might arise either from an
effect on the proteins that regulate attachment or removal of palmitate
or through more direct actions on the Ha-Ras protein itself. The
proteins that regulate Ha-Ras palmitoylation in the cell are not yet
identified and are thus not available for study. However, exposure of
recombinant Ha-Ras to nitric oxide in vitro had been
reported to directly modify (nitrosylate) cysteine 118 in the
GTP-binding pocket of Ha-Ras (57), and this nitrosothiol modification
has been suggested to increase Ha-Ras guanine nucleotide binding. Thus,
it was possible that SNC-triggered modification of Ha-Ras, potentially
through a change in GTP binding, might produce a conformational change
that enabled more rapid deacylation of the C terminus. This model was
evaluated in two stages: testing whether SNC would cause a change in
Ha-Ras GTP status in the setting of an intact NIH 3T3 cell and testing
if an Ha-Ras(Ser118) protein (which should be insensitive
to this direct nitrosylation) would also show SNC-triggered changes in
GTP binding or palmitoylation.
A fragment of Raf kinase that binds preferentially to GTP-Ras was used
as an affinity reagent to isolate active Ha-Ras proteins from
stimulated cells. Lysates were prepared from Ha-Ras61L-expressing NIH
3T3 cells that had been treated for 15 min with NO-depleted or freshly
prepared 4 mM SNC. Glutathione-agarose beads with a bound
fusion protein of GST-RBD of Raf kinase were added to the lysates, and
the GTP-Ha-Ras proteins that bound to the RBD domain were isolated,
displayed by SDS-PAGE, and detected by immunoblotting with an
anti-Ha-Ras antibody. In contrast to the increase in GTP binding to
Ha-Ras that had been reported, SNC caused a decrease (Fig.
5A, middle
panel), as indicated by the lower amount of Ha-Ras61L captured by the RBD in the SNC-treated sample. The assay was then scaled up so that the smaller amounts of GTP-bound cellular Ha-RasWT could be detected. The results with Ha-RasWT were similar to those with
Ha-Ras61L. SNC treatment led to a decrease in the amount of Ha-RasWT
that interacted with the Raf RBD (Fig. 5A, upper
panel).
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 5.
SNC decreases Ha-Ras GTP-binding.
A, NIH 3T3 cells expressing either Ha-RasWT (top
panel) or Ha-Ras61L (middle panel) or PC12 cells
expressing Ha-Ras61L (bottom panel) were treated for 15 min
with either NO-depleted SNC (C) or 4 mM freshly
made SNC. As a positive control, one culture of Ha-RasWT-expressing
cells was treated for 5 min with 50 ng/ml PDGF. Cells were lysed, and
samples (500 µg of total protein) were incubated with a GST fusion
protein of the RBD of Raf-kinase bound to glutathione agarose. Beads
were washed, protein was resolved by SDS-PAGE, and bound Ras-GTP was
detected with anti-Ras antibody. The data are representative of three
separate experiments. B, NIH 3T3 cells expressing Ha-RasWT,
v-Ha-Ras, or Ha-Ras(C118S) were labeled with
[32P]inorganic phosphate for 4 h and then treated
with either NO-depleted SNC or 4 mM freshly made SNC for 15 min (for C118S) or 30 min. Ha-Ras proteins were immunoprecipitated, and
the bound nucleotides were released, separated by thin layer
chromatography, and detected by autoradiography. C,
radioactive GTP and GDP were quantified by scanning films of
TLC-separated nucleotides and calculating the percentage of total
nucleotides that were GTP. The bars show average and range
of two experiments. Black columns represent
control samples, and gray columns represent
SNC-treated samples.
|
|
To determine more directly if this decrease in RBD association
reflected a SNC-triggered decrease in Ha-Ras GTP binding, NIH 3T3 cells
expressing Ha-RasWT or v-Ha-Ras were labeled with
[32P]inorganic phosphate and treated for 30 min with
either NO-depleted or 4 mM fresh SNC. Ha-Ras proteins were
immunoprecipitated, the bound nucleotides were released, and GTP and
GDP were separated by thin layer chromatography (Fig. 5B).
Radioactive GTP and GDP were then quantified by scanning the films
after autoradiography (Fig. 5C). With either 0.1 or 1 mM SNC, v-Ha-Ras GTP binding remained high (~60%), as
expected for an activated v-Ha-Ras protein, and, importantly, no
increase in GTP binding was detected (data not shown). With 2 mM SNC, a decrease in v-Ha-Ras GTP binding (to ~40%) was
observed. At 4 mM SNC, the same concentration that
accelerated deacylation, an approximately 3-fold decrease in the amount
of the GTP-bound form of v-Ha-Ras was observed, from an average of 59%
to 19% of the total v-Ha-Ras protein. SNC also decreased the amount of
GTP bound to the cellular form Ha-RasWT from 9% of total Ha-RasWT to
3%. Thus, low concentrations of SNC failed to produce an increase in
v-Ha-Ras GTP-binding and higher doses of SNC, similar to those that
affected palmitate replacement, caused a decrease in GTP binding to
either the cellular or an activated form of Ha-Ras. Therefore, the
effects of SNC on Ha-Ras expressed in NIH 3T3 cells appeared to differ
from the reported effects of nitric oxide on the endogenous Ras
proteins of neuronal or T lymphoma cells.
To address if the SNC-triggered decrease in GTP binding, like the
suggested NO-triggered increase, might occur through a mechanism that
utilized nitrosylation of Cys118, the ability of SNC to
alter GTP binding to an Ha-Ras(C118S) protein was determined. The
percentage of GTP-bound C118S protein was found to decrease from an
average of 9% to 3% after SNC exposure (Fig. 5C). This
decrease was similar to the decrease caused by SNC in Ha-RasWT. It is
essential to note that these results do not mean that
Cys118 was not being nitrosylated in the previous
experiments but do indicate that neither Cys118 nor its
nitrosylation were required for SNC to cause decreased GTP binding on
Ha-RasWT. Therefore, the mechanism through which SNC decreased GTP
binding on Ha-Ras was likely to be from an effect of SNC on some other
region of Ha-Ras or from SNC effects on the guanine nucleotide exchange
factors (GEFs) or GTPase-activating proteins (GAPs) that regulate the
Ha-Ras GTP state in NIH 3T3 cells. More importantly, these results
showed that, without manipulation of cellular glutathione levels and in
serum-containing medium, exposure of intact cells to SNC could alter
GTP binding on either cellular or activated forms of Ha-Ras.
A Cysteine at Position 118 Is Not Required for SNC to Affect Ha-Ras
Palmitoylation--
Although Cys118 was not needed in
order for SNC to cause decreased GTP binding, it might still be needed
in order for SNC to affect palmitoylation. To test this possibility,
palmitoylation of the C118S mutant was examined in SNC-treated cells.
[3H]Palmitate labeling of C118S was increased 2-3-fold
when cells were exposed to SNC (Fig.
6A). This stimulation of C118S
palmitate labeling by SNC was almost identical to the increase caused
by SNC on Ha-RasWT. Additionally, the deacylation rate of C118S was also accelerated by SNC treatment (Fig. 6B). SNC therefore
continued to produce effects on palmitate turnover in a mutant Ha-RasWT protein that lacked Cys118. These data implied that the
accelerated palmitate turnover caused by SNC was not a secondary effect
of nitrosylation of Cys118. Thus, neither the ability of
SNC to decrease Ha-Ras GTP binding nor its acceleration of Ha-Ras
depalmitoylation required Cys118 to be present.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 6.
Cysteine at position 118 is not required for
SNC to affect Ha-Ras palmitoylation in NIH 3T3 cells.
A, NIH 3T3 cells expressing Ha-Ras(C118S) were labeled for
the times indicated with [3H]palmitate in medium with no
additions ( ), with nitric oxide-depleted 4 mM SNC
(C), or with fresh 4 mM SNC re-added every 30 min. Ha-Ras was detected by fluorographic exposure for 19 days and
quantified as in Fig. 2. In graphs the hatched
black line and diamonds indicate
untreated samples; control samples treated with NO-depleted SNC are
shown by the solid black line and
squares; SNC-treated samples are designated by the
gray line and triangles. B,
NIH 3T3 cells stably expressing Ha-Ras(C118S) were labeled with
[3H]palmitate for 3 h and then incubated with medium
containing 200 µM palmitate and either NO-depleted SNC
(Control) or 2 mM freshly made SNC. SNC was
added every 30 min. Ha-Ras was detected by fluorographic exposure for 7 days and quantified as in Fig. 2. The solid black
line and squares represent the control samples,
and the gray line and triangles depict
the SNC-treated samples.
|
|
SNC Inhibits Phosphorylation of ERK1 and ERK2 in NIH 3T3
Cells--
From the above experiments, SNC could alter two important
properties of Ha-Ras: its activation state and its lipidation. To examine if SNC also altered Ras signaling, the phosphorylation of the
mitogen-activated protein kinases ERK1 and ERK2, a property that
parallels activation of these kinases, was examined in SNC-treated NIH
3T3 cells. Relative changes in phosphorylation were detected using an
antibody specific for the phosphorylated form, and in addition, the
portion of total ERK1 and ERK2 that was phosphorylated was monitored
using antibody that detected both the unphosphorylated and
phosphorylated forms that had been separated by SDS-PAGE. In the first
experiments, untransfected NIH 3T3 cells, containing only endogenous
cellular Ras proteins, were incubated in the absence of serum for
2 h to deactivate the ERK1/ERK2 proteins. The serum-free medium
was then replaced with medium containing 10% bovine calf serum (the
same amount present in the palmitoylation experiments). The control
plates received medium that contained nitric oxide-depleted SNC
solution. A second set of plates received a single treatment with
medium containing freshly made SNC. A third set of plates received
medium containing SNC and had additional SNC added every 30 min over
the course of 2 h.
In the control cells, phosphorylation of both ERK1 and ERK2 was
stimulated by the serum addition. This serum-stimulated phosphorylation was maintained for the entire 2 h (Fig.
7A). However, in the
SNC-treated cells, serum-stimulated ERK phosphorylation was inhibited.
In cells that received a single addition of SNC, decreased
phosphorylation of the ERK proteins could be detected within 5 min
(data not shown). Phosphorylation of ERK1 and ERK2 was decreased by
~75% within 30 min and remained depressed for longer than 1 h.
After 2 h, the effectiveness of SNC began to wane, and small
amounts of phosphorylated ERK1 and ERK2 became detectable. In plates
that received multiple additions of SNC (as in the palmitoylation
experiments), a complete lack of ERK phosphorylation was maintained.
Variation of the amount of SNC applied showed that inhibition of ERK1
and ERK2 required the same concentrations of SNC (2-4 mM)
that were needed to increase Ha-RasWT palmitate turnover and decrease
GTP binding (data not shown). Lower concentrations of SNC (0.1 and 1 mM) neither inhibited nor stimulated ERK1 or ERK2
phosphorylation in NIH 3T3 cells. SNC also inhibited ERK
phosphorylation in growing NIH 3T3 cells that were not serum-starved,
as well as in cells in which ERK phosphorylation was restimulated with
the specific growth factor epidermal growth factor or platelet-derived
growth factor rather than serum (data not shown). Thus, SNC interfered
with both serum- and growth factor-induced activation of the ERK
pathway in NIH 3T3 cells. Inhibitory effects of a single addition of
SNC were sustained for at least 1 h and could be maintained for
longer periods by repeated additions.
View larger version (42K):
[in this window]
[in a new window]
|
Fig. 7.
SNC effects on ERK1/ERK2 are transient and
vary in different cell types. A, untransfected NIH 3T3
cells were incubated in serum-free medium for 2 h and then
incubated in fresh, serum-containing medium with either NO-depleted SNC
(C) or 4 mM freshly prepared SNC. Every 30 min,
additional fresh SNC was added to one set of plates (2×, 4×). No
additional SNC was added to the single addition (1×) samples. NIH 3T3
cells expressing Ha-RasWT were also incubated in medium lacking serum
for 2 h and then were treated for 30 min with fresh,
serum-containing medium with either no addition (), NO-depleted SNC
(C), or fresh 4 mM SNC. Cells expressing
v-Ha-Ras were treated similarly except without incubation in serum-free
medium. Cells were lysed at the times indicated, proteins were
separated by SDS-PAGE, and ERK proteins were detected by immunoblotting
with an antibody (phospho-ERK antibody) that specifically recognizes
the phosphorylated forms of ERK1 and ERK2 (upper panel) or
an ERK1 antibody that recognizes both phosphorylated and
unphosphorylated ERK1 and ERK2 proteins (total ERK antibody), The
phosphorylated forms of ERK1 and ERK2 are indicated with
arrowheads. B, NIH 3T3 cells expressing Ha-RasWT
were incubated in medium lacking serum for 1 h, and then sodium
vanadate (1 mM) was added to the indicated plates and
incubation continued for an additional 1 h. Medium was replaced
with medium containing 10% serum and, as indicated, vanadate, 4 mM NO-depleted SNC (), or 4 mM fresh SNC, and
samples were collected after 30 min. ERK phosphorylation was detected
by immunoblotting as in A. C, COS1 cells were
incubated in serum-free medium for 2 h and then treated for 30 min
with fresh, serum-containing medium with either no addition (),
NO-depleted SNC (C), or fresh 4 mM SNC. Raw
264.7 and PC-12 cells were placed directly in fresh, serum-containing
medium with the indicated additions. ERK1/ERK2 phosphorylation was
detected as in A. Phosphorylated ERK2 in the COS1 monkey
cells has a mobility equivalent to unphosphorylated ERK1, so only three
bands are detected on the blot. Similar results were observed in
multiple experiments (n > 3).
|
|
In NIH 3T3 cells stably expressing larger amounts of Ha-RasWT, SNC also
decreased serum-stimulated phosphorylation of ERK1 and ERK2 (Fig.
7A). However, in NIH 3T3 cells transformed by the oncogenic
v-Ha-Ras protein, SNC failed to decrease ERK1 or ERK2 phosphorylation
(Fig. 7A). This lack of effect of SNC on ERK phosphorylation in transformed cells indicated that SNC was not a direct chemical depressor of ERK phosphorylation. In addition, the distinct ERK responses of cells expressing Ha-RasWT and v-Ha-Ras, despite the similar SNC-triggered effects on Ha-Ras palmitoylation, suggested that
SNC affected ERK phosphorylation and Ha-Ras palmitoylation through
separate mechanisms.
The mechanism of SNC-triggered ERK inactivation was further examined to
determine if the decrease in ERK1 and ERK2 phosphorylation reflected
only decreased activation of the Ha-Ras/Raf/MEK/ERK pathway or also a
SNC-triggered increase in activity of an ERK phosphatase. ERK
phosphorylation was evaluated using NIH 3T3 cells expressing Ha-RasWT
and the phosphatase inhibitor, sodium orthovanadate, to prevent removal
of tyrosine phosphates from the ERK proteins. Successful inhibition of
phosphatase activity by vanadate was demonstrated by elevation of ERK1
and ERK2 phosphorylation in the vanadate-treated cells (Fig.
7B). The addition of serum to the vanadate-treated cells
produced the expected even greater increase in ERK phosphorylation. In
the vanadate-treated cells, SNC still diminished serum-stimulated ERK
phosphorylation but importantly was not able to fully inhibit ERK
phosphorylation, as it could in cells in which phosphatases remained
active (Fig. 7B). Thus, when the activity of phosphatases
was limited by vanadate, the inhibition of ERK phosphorylation caused
by SNC was lessened. This suggested that vanadate-sensitive
phosphatase(s) contributed to the decreased phosphorylation of ERK1 and
ERK2 in SNC-treated cells. Therefore, it appeared that ERK deactivation
by SNC could occur from a combined result of SNC-stimulated phosphatase
activity and SNC-triggered inactivation of the Ha-Ras pathway upstream of these ERK proteins. These two effects would both contribute to the
substantial inhibition of ERK phosphorylation seen in the SNC-treated
NIH 3T3 cells.
Inhibition of ERK activity by nitric oxide has also been seen in rat
cardiac fibroblasts (58). However, increased, rather than decreased,
ERK1/ERK2 phosphorylation had been reported for a number of other cell
types treated with nitric oxide (41, 59). To learn if SNC was distinct
from nitric oxide in inhibiting ERK proteins or if ERK proteins might
respond to SNC in a cell-specific manner, the effect of SNC on ERK
phosphorylation was tested in several cell lines. In COS1 monkey kidney
cells, results similar to those in NIH 3T3 fibroblasts were observed; a
30-min treatment with 4 mM SNC decreased ERK1
phosphorylation (Fig. 7C). This showed that NIH 3T3 cells
were not unique in having an inhibitory response to SNC. However, in
monocytic RAW 264.7 cells and the neuronal PC-12 cell line, SNC
treatment stimulated ERK1/ERK2 phosphorylation (Fig. 7C).
These results showed that SNC could activate ERK phosphorylation in
these monocytic or neuronal cell lines, similar to reports using other
NO donors. To examine if SNC would increase Ha-Ras GTP binding in the
PC12 cells, to parallel its stimulation of ERK phosphorylation, the
effect of SNC on interaction of the expressed Ha-Ras with the GST-RBD
protein was determined. However, rather than an increase, SNC treatment
caused a decrease in the amount of Ras(GTP) that was captured by the
GST-RBD protein (Fig. 5A, lower
panel). This decrease was similar in magnitude to the
decrease in Ha-Ras/RBD interaction that was seen in NIH 3T3 cells.
Thus, in PC12 cells activation of the ERK proteins by SNC appeared to occur without an increase in Ras GTP. Overall, the effect of SNC on the
ERK pathway appeared to vary in different cells, while a decrease in
Ha-Ras GTP binding was seen in two cell types.
| |
DISCUSSION |
Manipulation of Ha-Ras Palmitoylation with SNC--
These studies
show that SNC can alter two important properties of Ha-Ras: its
activation state and its lipidation. These effects were observed in NIH
3T3 cells expressing the normal Ha-RasWT as well as in cells
transformed by an activated v-Ha-Ras mutant.
The effects of SNC on palmitoylation were protein-specific and caused
increased palmitate labeling and turnover on Ha-Ras but decreased
incorporation of palmitate into the transferrin receptor and caveolin
in the same cells. The nitric oxide donor compound SIN-1 has also been
reported to alter protein palmitoylation (60). As with SNC, the effects
of SIN-1 on palmitoylation were protein-selective; palmitate
incorporation into the transmembrane -adrenergic receptor was
reduced by SIN-1 treatment, whereas palmitate labeling of the
Gs -subunit coupled to the receptor was unaffected (60).
It thus appears that at least certain palmitoylated proteins, including
Ha-Ras, are susceptible to nitrosylating agents but that each will
display an individualized response, presumably resulting from
regulatory mechanisms distinct from each protein. The search for the
still elusive enzymes that regulate protein palmitoylation should
benefit from the discovery of these compounds' abilities.
Our results indicate that SNC can stimulate palmitate turnover on
Ha-Ras and that the mechanism for this increase may be primarily through acceleration of palmitate removal. Additional possible effects
through increases in palmitoylation of unoccupied sites in the mature
protein have not been excluded rigorously. Although palmitate removal
is enhanced, the overall state of Ha-Ras palmitoylation does not appear
to be decreased during short term SNC exposure, since the cellular
machinery for repalmitoylation rapidly compensates and replaces
palmitates that are removed. Perhaps as a result of this reacylation,
no general release of
Ha-RasWT3 or v-Ha-Ras from
membranes of the treated cells was detected. Lu and Hofmann (13) have
suggested previously that access of Ha-Ras palmitates to deacylation
can occur without release of the protein from membranes. The
possibility of a subtle redistribution of the Ha-Ras protein, either
into or out of subdomains of the plasma membrane remains to be explored.
At least two mechanisms by which this increased rate of
depalmitoylation might occur can be envisioned: by stimulation of the
activity of an Ha-Ras S-palmitoyl thioesterase or through modifications or conformational changes that increase accessibility of
the Ha-Ras C terminus to a thioesterase. The possibility that SNC may
change palmitoyl thioesterase activity is now amenable to study, since
a candidate thioesterase that can depalmitoylate Ha-Ras and trimeric G
proteins in vitro has recently been cloned (19). Of note is
the report that caveolin does not appear to be a substrate for the APT1
thioesterase that can deacylate Ha-Ras (35). The dissimilarity in the
effects of SNC on palmitate labeling of Ha-Ras (increase) and the
transferrin receptor and caveolin (decreases) also indicates that not
all palmitoylated proteins or the putative enzymes that regulate their
acylation will respond to SNC in the same way.
Although our results do not support an involvement of a modification of
Cys118 in the effects of SNC on depalmitoylation, the three
C-terminal lipid-modified cysteines are possible candidates for direct
modification. The other cysteine residues in Ha-Ras are buried within
the core of the folded Ha-Ras protein and do not react with nitric
oxide or several other oxidants (61, 62). At least three of these four
cysteines in Ha-RasWT can be shown to be S-nitrosylated
after exposure of the protein to S-nitrosoglutathione (61).
Of the C-terminal cysteines, Cys186 is stably modified with
isoprenoid in the mature Ha-Ras protein and is available as a target
only in the unmodified precursor form. If exposure to nitric oxide,
pharmacological inhibitors of farnesyl synthesis (e.g.
compactin), or farnesyltransferase inhibitors were to cause
accumulation of this usually minor population, then
S-nitroso modification of Cys186 could assume
larger importance, since this would at least transiently prevent
membrane attachment of Ha-Ras. It is possible that Cys181
or Cys184 are nitroso-modified, but the increase in
[3H]palmitate labeling of Ha-RasWT occurs so rapidly that
such modification, if it occurs, must be brief, since these sites are
quickly available for repalmitoylation. It will be necessary to develop
direct chemical methods to analyze C-terminal cysteines of Ha-Ras from
SNC-treated intact cells to clarify if direct nitrosative or oxidative
modifications take place.
Indirect SNC-induced conformational changes of the C terminus could
also contribute to an increased susceptibility of Ha-Ras to
deacylation. APT1, a thioesterase that can deacylate Ha-Ras, has
recently been reported to deacylate the endothelial nitric-oxide synthase and to do so more efficiently when endothelial nitric-oxide synthase is activated by Ca2+/calmodulin (35). However,
little structural information is available for the Ha-Ras C terminus
(and currently none for lipidated forms of Ha-Ras), so whether the
observed decreases in Ha-Ras GTP binding triggered by SNC might also
cause conformational changes in the C-terminal domain remains to be
clarified. More detailed studies using SNC and selected
C-terminal mutants in intact cells can now examine for how long a
singly palmitoylated or nonpalmitoylated Ha-Ras may persist and begin
to examine requirements for dynamic deacylation and repalmitoylation.
Regardless of the mechanisms involved, SNC is the first compound to be
identified that can be used to manipulate Ha-Ras palmitoylation in
living cells. These results raise the question of whether SNC is
mimicking a natural regulatory event. The present studies suggest that
Ras palmitate turnover should be explored in situations where abundant
nitric oxide is produced, such as during activation of macrophages or
neutrophils (63). Palmitate turnover on Ras proteins may be among the
several important properties of the Ras-dependent signaling
pathways that are altered during nitrosative or oxidative stress.
Effectiveness of SNC and SNG--
It is not yet clear why SNC is
such a potent inducer of Ha-Ras palmitate turnover in NIH 3T3 cells.
The increased palmitate labeling of Ha-Ras contrasts with the decreased
palmitate incorporation seen with the transferrin receptor and caveolin
and argues that SNC is not a general chemical stimulant of
palmitoylation. Although 8-bromo-cyclic GMP can mimic some effects of
nitric oxide on other signaling pathways (53), we found no evidence
that it could alter Ha-Ras palmitoylation in NIH 3T3 cells. The current
tests of several other compounds that generate nitric oxide radicals, peroxynitrite, and other oxidants do not show detectable effects on
palmitoylation, suggesting that the nitrosothiol form, rather than free
nitric oxide radical, may be producing these effects. Gross and Lane
(55) have recently reviewed the growing evidence that when nitric oxide
is produced within tissues or introduced into the blood stream that it
is cell-permeable nitrosothiols that are the form through which nitric
oxide "equivalents" enter and cause biological effects within
intact cells. Our observation that the slightly larger nitrosothiol,
S-nitrosoglutathione, stimulates Ha-Ras palmitate labeling,
but more modestly than SNC, also suggests that entry of the active
compound into the cell before it reacts with extracellular targets may
be an important factor. L-S-nitrosocysteine has
been reported to enter cells through a stereo-selective amino acid
transporter (64, 65). One preliminary experiment utilizing D-S-nitrosocysteine indicated that the
D-stereoisomer had little effect on Ha-RasWT
palmitoylation,4 suggesting
that L-S-nitrosocysteine may enter cells
efficiently and thus be particularly effective in causing changes
within NIH 3T3 cells. The amount of free nitric oxide available from
the added SNC is also likely to be limited by the presence in the medium of serum proteins that can be easily nitrosylated. The demonstration that SNC can induce changes in Ha-Ras lipidation and
activation even if serum is present should allow these results to match
closely conditions that will occur in tissues during nitrosative stress
or pharmacological interventions.
Effects of SNC on Ha-Ras Signaling--
Our data indicate that
exposure of COS-1, PC12, or NIH 3T3 cells to SNC causes a decrease in
Ha-Ras GTP binding. SNC also alters ERK phosphorylation, although this
signaling response appears to differ among cell types. These results
thus provide significant support for the important concept that
oxidative (and nitrosative) stress can influence the active state and
signaling of Ha-Ras (66, 67). Further efforts will be needed to
determine how cell type and experimental conditions, particularly
cellular redox status, influence cellular responses. Initial
experiments to examine the basis of the inhibitory response of ERK1 and
ERK2 to SNC in NIH 3T3 cells suggest that this inhibition may result
from a combination of decreased Ha-RasWT GTP binding and activation of
a vanadate-sensitive ERK phosphatase. The adjustments in phosphatase
activity that occur in stably transformed cell lines may explain why
ERK proteins in cells transformed by v-Ha-Ras are insensitive to SNC,
despite a significant SNC-triggered decrease in v-Ha-Ras GTP binding. Stimulation of a protein phosphatase would be an unusual action of SNC,
since nitric oxide producers generally inhibit tyrosine-specific phosphatases instead of activating them (68). It is likely that the
decreases in cellular Ha-RasWT GTP binding and ERK phosphorylation seen
in NIH 3T3 cells during SNC treatment reinforce each other and produce
a large effect on the Ha-RasWT signaling pathway. Since nitric oxide
has been reported to regulate cell growth and survival (56, 69, 70),
this concerted effect of SNC in NIH 3T3 cells should be useful for
studies of the role of nitric oxide or oxidants (66, 71) in the
Ha-Ras/ERK pathway's control of cell proliferation.
Connection between Ha-Ras Palmitoylation and Activation--
There
is currently a small list of proteins whose rate of palmitate turnover
can be modified by pharmacological agents or agonist stimulation: the
-adrenergic receptor, the Gs protein, and several
other G -subunit types, endothelial nitric-oxide synthase, GAP-43,
and SNAP-25 (27, 33, 34, 36, 49). These results suggest that palmitates
on certain proteins respond to external stimuli, indicating that
palmitoylation of these proteins might be an event that can be regulated.
The results presented here identify SNC as a prototype compound that
will enable study of how the two dynamic properties of Ha-Ras (GTP
binding and palmitoylation) may be connected. A possible model already
exists, based on studies with heterotrimeric G proteins. For the
heterotrimeric Gs protein, GTP binding releases the
-subunit from 
and increases depalmitoylation of the
-subunit (28, 30, 31). The model proposes that while the -subunit
is briefly nonpalmitoylated, GAP-stimulated GTP hydrolysis occurs,
inactivating the protein. Gs is then repalmitoylated and
binds again to (72). Studies of palmitoylation of
Go and Gz have shown a similar
relationship between palmitate modification and activation state (73).
Palmitoylation of Gz inhibits its interaction with the
Gz GAP and other RGS proteins that stimulate hydrolysis of Gz-bound GTP (74). When Gz is
depalmitoylated, GAP-stimulated GTP hydrolysis can occur. Our
comparisons of palmitate turnover in dominant negative, normal, and
oncogenic forms of Ha-Ras indicate that deacylation is more rapid in
GTP-bound forms of Ha-Ras2 and suggest that activation and
palmitoylation of Ha-Ras proteins may also have an unexplored alliance.
A better understanding of how Ha-Ras palmitoylation is regulated could
also provide novel approaches for controlling Ha-Ras oncogenicity.
These studies with SNC will enable cellular examination of the enzymes
and requirements for attachment and removal of this crucial lipid and
provide a novel tool to help define biochemically and biologically if
there are more active roles of dynamic acylation beyond membrane binding.
§¶,
TOP
ABSTRACT
INTRODUCTION
To whom correspondence should be addressed: Dept. of
Biochemistry, Biophysics and Molecular Biology, Iowa State University, Ames, IA 50011. Tel.: 515-294-6125; Fax: 515-294-0453; E-mail: jbuss@iastate.edu.
