(Received for publication, January 23, 1997, and in revised form, April 15, 1997)
From the Eicosanoid Biochemistry Section, Laboratory of Molecular
Carcinogenesis, NIEHS, National Institutes of Health, Research Triangle
Park, North Carolina 27709 and Division of Basic Medical
Sciences, Mercer University School of Medicine,
Macon, Georgia 31207
In Syrian hamster embryo (SHE) fibroblasts,
epidermal growth factor receptor (EGFR) tyrosine kinase activity
regulates the metabolism of endogenous linoleic acid to
(13S)-hydroperoxyoctadecadienoic acid
(13S)-HPODE). (13S)-HPODE stimulates
EGF-dependent mitogenesis in a SHE cell phenotype, which
expresses tumor suppressor genes (supB+), but was not
effective in a variant that does not express these suppressor genes
(supB). In the present study, we have investigated the
potential effects of this lipid metabolite on the EGFR signaling
pathways in these two SHE cell lines. Treatment of quiescent SHE cells
with EGF produced a rapid, transient increase in the tyrosine
phosphorylation of EGFR. Dependence on EGF concentration for EGFR
tyrosine phosphorylation was similar in both SHE cell lines, but a more
prolonged phosphorylation was detected in the supB
variant. Incubation of supB+ cells with
(13S)-HPODE and EGF increased EGFR autophosphorylation and
tyrosine phosphorylation on several signaling proteins with Src
homology-2 domains including GTPase-activating protein. The lipid
metabolite did not significantly alter EGF-dependent
tyrosine phosphorylation in the supB
variant. Tyrosine
phosphorylation of mitogen-activated protein (MAP) kinase was also
measured. The addition of (13S)-HPODE increased the extent
and duration of MAP kinase tyrosine phosphorylation in
supB+ cells but not in the supB
variant. MAP
kinase activity in supB+ cells, as measured in
immunoprecipitates from cells after the addition of EGF, was increased
by the presence of (13S)-HPODE. The addition of
(13S)-HPODE did not directly alter EGFR kinase activity or
the internalization of the EGFR. However, the addition of
(13S)-HPODE to supB+ cells extended the
tyrosine phosphorylation of the EGFR in response to EGF. The
dephosphorylation of the EGFR was measured directly, and a slower rate
was observed in the supB
compared with the
supB+ cells. Incubation of the supB+ cells with
(13S)-HPODE attenuated the dephosphorylation of the EGFR.
Thus, (13S)-HPODE stimulates EGF-dependent
mitogenesis and up-regulation of EGF-dependent tyrosine
phosphorylation by inhibiting the dephosphorylation of the EGFR. This
study shows that a metabolite of an essential dietary fatty acid,
linoleic acid, can modulate tyrosine phosphorylation and activity of
key signal transduction proteins in a growth factor mitogenic
pathway.
Several lines of evidence suggest that metabolism of the cis-polyunsaturated fatty acids, arachidonic acid and linoleic acid, by prostaglandin H synthase and lipoxygenases generate metabolites that modulate the EGF1 mitogenic signal in fibroblasts. In Balb/c 3T3 fibroblasts, mitogenic stimulation by EGF induced the formation of prostaglandin E2 (1) and the expression of c-myc (2). Inhibition of prostaglandin H synthase partially blocked both mitogenesis and c-myc expression, which was restored and enhanced by the addition of exogenous prostaglandins. In these studies lipoxygenase inhibitors were very effective inhibitors of mitogenesis and also inhibited the metabolism of linoleic acid by these cells (1, 3). We did not detect lipoxygenase metabolites of arachidonic acid, but did observe that linoleic acid was readily metabolized to (13S)-hydroxyoctadecadienoic acid (HODE). Linoleic acid metabolites were potent stimulators of EGF-dependent mitogenesis in these cells (3).
A more extensive investigation was conducted with Syrian hamster embryo fibroblasts (SHE) cells (4). These cells metabolized arachidonic acid to low amounts of prostaglandin E2 with little or no lipoxygenase metabolites observed. The addition of prostaglandins inhibited EGF-dependent mitogenesis and the expression of the proto-oncogenes, c-myc (5), c-jun, and Jun-B (6). Indomethacin did not inhibit EGF-dependent mitogenesis, while lipoxygenase inhibitors were particularly effective for inhibiting mitogenesis. EGF is a potent mitogen in these cells but does not enhance arachidonic acid metabolism nor the expression of prostaglandin H synthase-2 (7). EGF stimulates the metabolism of exogenous as well as endogenous linoleic acid to (13S)-HPODE/(13S)-HODE catalyzed by an apparent 15-lipoxygenase. The activity of this lipoxygenase is regulated by the EGF receptor tyrosine kinase (7), thus linking the activation of lipoxygenase metabolism of linoleic acid to the tyrosine kinase activity of the epidermal growth factor receptor (EGFR). Other growth factors, including fibroblast growth factor (8) and insulin and platelet-derived growth factor (2), do not activate linoleic acid metabolism, suggesting that the stimulation of linoleic acid metabolism is unique to the EGF signaling pathway.
The SHE cells are used as a model for neoplastic progression since two
phenotypes are available, which, when fused with tumor cells, either
suppress (supB+) or do not suppress (supB)
tumorigenic phenotypes in cell-cell hybrids (9-11). The addition of
(13S)-HODE or its hydroperoxide precursor,
(13S)-HPODE, greatly potentiated the mitogenic response of
supB+ SHE cells to EGF (4). At concentrations of
10
9 to 10
6 M,
(13S)-HPODE and (13S)-HODE increased EGF
dependent mitogenesis 3-4-fold in the supB+ cells with
(13S)-HPODE being 10 times more potent than
(13S)-HODE. The linoleic acid metabolites did not alter the
EGF mitogenic response in the supB
variant. In the
absence of EGF, these lipid metabolites were not mitogenic. Further
structural characterization of the lipid-dependent enhancement of mitogenesis indicated (13S)-HPODE and
(13S)-HODE have exclusive activity (11). For example, the
(13R)-HODE enantiomer and the analogous 15-lipoxygenase
metabolites of arachidonic acid, (15S)-HPETE/15-HETE, were
not active (11). This finding suggests a specific interaction of
(13S)-HPODE with either a receptor or with a intracellular
signal transduction protein of the EGF signaling pathway.
These results indicate that the formation of
(13S)-HPODE/(13S)-HODE is a component of the EGF
signaling pathway and this lipid metabolite potentiates
EGF-dependent mitogenesis in the supB+ cell
line. Polypeptide growth factors such as EGF regulate cellular growth
and metabolism by binding to specific cell surface receptors that have
tyrosine kinase activity (12, 13). Occupation of the EGFR by EGF and
other ligands causes dimerization of the receptor and stimulates the
receptor's intrinsic tyrosine kinase activity (14). The intrinsic
tyrosine kinase activity of the EGFR is the critical biochemical signal
involved in cellular mitogenic responses to EGF (17-19). The kinase
activity autophosphorylates the EGFR, phosphorylates key tyrosine
residues present in several signaling proteins, and initiates a series
of tyrosine, serine and threonine protein phosphorylations which link
the cell surface with the nucleus (15). One of the final steps in the
EGF cascade appears to be the activation of the mitogen-activated
protein (MAP) kinase family, which in turn regulates other biochemical events including activation of transcription factors (16). One possible
explanation for the enhancement of EGF-dependent
mitogenesis by (13S)-HPODE/(13S)-HODE is that
this lipid metabolite alters phosphorylation events in the EGFR
pathway. Furthermore, our findings suggest that important differences
may exist in the EGF signaling pathway between these two SHE cell lines
and suggest that during the neoplastic progression a mutation or
deletion occurred that alters the signaling transduction process. In
this report we have tested these hypotheses by examining the EGF
signaling pathway in the two SHE cell phenotypes. In the
supB+ cell line, (13S)-HPODE specifically
enhanced the phosphorylation of EGFR and other signaling proteins of
the pathway leading ultimately to enhanced MAP kinase activity. The
(13S)-HPODE exerts this response by inhibiting the
dephosphorylation of the EGF receptor. In contrast, the linoleate
metabolite did not modulate receptor dephosphorylation in
supB cells and thus did not alter
EGF-dependent tyrosine phosphorylation in the
supB
cell line.
Bovine serum albumin (BSA), IBR modified
Dulbecco's modified Eagle's medium (IBR), phosphate-buffered saline
(PBS), trypsin, and gentamicin were from Life Technologies, Inc. Fetal
calf serum was purchased from HyClone Laboratories (Logan, UT).
75-cm2 flasks were obtained from Costar (Cambridge, MA).
100-mm dishes were from Falcon (Plymouth, United Kingdom).
(13S)-HPODE, (13S)-HODE, (15S)-HETE,
(15S)-HPETE, and (12S)-HETE were from Cayman
Chemical Co. (Ann Arbor, MI). (13R)-HODE was obtained from
Oxford (Oxford, MI). EGF was obtained from Collaborative Research
Associates (Bedford, MA). Tris, glycerol, ecolume and glycine were
purchased from ICN (Aurora, IL). Sodium chloride and methanol were from
J.T. Baker. The magnesium chloride and trichloroacetic acid were from
Mallinckrodt. The acrylamide and bisacrylamide were purchased from
Amresco (Solon, OH). The nitrocellulose was obtained from Schleicher & Schuell. The anti-GAP antibody was from Transduction Laboratories
(Lexington, KY). The Ig conjugates, Hyperfilm, ECL reagents,
[-33P] adenosine triphosphate (2000 Ci/mmol) and
125I-EGF were purchased from Amersham. The anti-MAP
antibodies (ERK-1, ERK-2) were from Santa Cruz Biotechnology (Santa
Cruz, CA), while anti-GAP antibodies were purchased from Transduction
Laboratories (Lexington, KY). The P81 chromatography paper was from
Whatman. The ethanol was from Pharmaco (Brookfield, CT). The protein
A-Sepharose beads were from Pharmacia Biotech (Uppsala, Sweden). The
methyl 2,5-dihydroxycinnamate was from Biomol (Plymouth Meeting, PA). The anti-phosphotyrosine antibody and all other reagents were obtained
from Sigma, while a EGFR antibody raised in rabbits against the
carboxyl region of rat EGFR was a gift from Dr. G. Clark, NIEHS.
These experiments were performed with Syrian
hamster embryo fibroblast cell lines, 10WsupB+ 8 (supB+) and 10WsupB 1 (supB
),
as described previously (4). Cells were maintained at 37 °C in a
humidified 5% CO2, 95% air atmosphere. The cells were
cultured in IBR containing 10% fetal calf serum and gentamicin (10 µg/ml). Trypsin (0.05%) was used to subculture the cells. Both
variants were grown under identical conditions and passaged at 70-80%
confluence. Passage 10-17 was used in these experiments. Passage
number and degree of confluence (70-80%) were identical in
experiments comparing supB+ to supB
cells.
After reaching 70-80% confluence on 100-mm dishes, cells were serum-deprived for 20 h to synchronize cell cycles in Go. All experiments were done in serum-free medium at 37 °C unless otherwise stated. Cells were treated with (13S)-HPODE or other lipids at 10 nM to 1 µM for 1 h prior to the addition of EGF. After stimulation by EGF, cells were harvested at various time points. For the room temperature experiments, the cells were allowed to equilibrate for 30 min before the addition of the linoleic acid metabolite.
Western Blot AnalysisTreated cells were washed twice with
ice-cold PBS and lysed in boiling sample buffer: 5 mM
sodium phosphate, pH 6.8, 2% sodium dodecyl sulfate, 0.1 M
dithiothreitol, 10% glycerol, 5% -mercaptoethanol, 0.2%
bromphenol blue, 1 mM sodium orthovanadate, 1 mM sodium fluoride. The lysates from an equal number of
cells were then sheared five times through a 25-gauge needle, boiled
for 5 min, and immediately loaded onto a gel or the protein estimated
by the Bradford reagent and the same amount of protein loaded onto the
gel. Samples were run on SDS-PAGE (6%, 8%, or 10%) and
electrophoretically transferred to a nitrocellulose membrane in 25 mM Tris, 192 mM glycine, 20% methanol with 200 µM sodium orthovanadate using Hoefer electrophoresis equipment. Membranes were blocked in Tris-buffered saline containing 0.2% Tween 20 (TBST) with 5% BSA. The blots were then incubated for
1 h with an anti-phosphotyrosine antibody (1:2000), anti-EGFR antibody (1:2000), or an anti-GAP antibody (1:250) with 1% BSA at room
temperature. The blots were washed five times in TBST with 0.1% BSA,
then incubated with peroxidase-conjugated anti-mouse Ig (1:10000) or
anti-rabbit Ig (1:10000 or 1:2000) in 1% BSA. The blots were again
washed five times in TBST with 0.1% BSA and visualized using the
Amersham ECL system.
When cells reached 70-80%
confluence, they were starved in Dulbecco's modified Eagle's medium
without antibiotics containing 0.05% fetal bovine serum overnight.
Three dishes of each experimental group were set aside to count the
cell number. The medium was aspirated, and the cells were rinsed with
10 ml of PBS containing 0.1% BSA and 2 mM sodium vanadate
and lysed with 0.8 ml of ice-cold lysis buffer: 1% Triton X-100, 50 mM Tris, pH 8.5, 150 mM NaCl, 5 mM
EDTA, 5 mM PMSF, 50 mM NaF, 2.0 mM
sodium vanadate, 20 µg/ml leupeptin, 20 µg/ml aprotinin. After
scraping of the cells, the lysate was collected, 50 µl of 5 M NaCl added, sonicated on ice for 7 s twice, and
centrifuged in a microcentrifuge for 10 min at 4 °C. The supernatant
was transferred to a new tube and an aliquot removed to measure the
protein concentration by BCA protein assay reagent. 2.0 mg of cellular
protein from 3 × 106 cells was precleared with 5 µl
of normal rabbit serum and 100 µl of protein A-Sepharose 4B beads.
The mixture was tumbled at 4 °C for 30 min and centrifuged at
10,000 × g for 10 min, and the supernatant removed. To
the supernatant was added 7 µl of EGFR antibody, tumbled overnight at
4 °C, and 100 µl of protein A-Sepharose 4B beads added and tumbled
for an additional 2 h. After centrifugation at 10,000 × g for 5 min, the pellet was washed sequentially with 1.0 ml
of IP2, IP3, and IP4 buffers.
(IP2: 0.5% Triton X-100, 50 mM Tris, pH 8.5, 500 mM NaCl, 1 mM EDTA, 1 mM sodium
vanadate, 20 µg/ml leupeptin, 20 µg/ml aprotinin, 40 µg/ml PMSF;
IP3: 0.5% Triton X-100, 50 mM Tris pH 8.5, 150 mM NaCl, 1 mM EDTA, 1 mM sodium
vanadate, 20 µg/ml leupeptin, 20 µg/ml aprotinin, 40 µg/ml PMSF;
IP4: 0.1% Triton X-100, 10 mM Tris, pH 8.5).
To the final pellet was added equal amounts of 10 mM Tris,
pH 8.5, and 2 × protein sample buffer, boiled for 8 min and
centrifuged at 10,000 × g for 5 min. The supernatant
was frozen at 70 °C for later use on SDS-PAGE (8%).
The EGFR in cell membrane preparations
was used to estimate the EGFR kinase activity using a modification of
the method described by Griswold-Prenner et al. (20).
Briefly, membranes prepared from SHE cells were incubated in 0.25 mM Hepes buffer, pH 7.4, 50 nM EGF, 11 mM MgCl2, 1 mg/ml angiotensin II, 20 µM ATP, and 10 µCi of [-33P]ATP. The
incubation mixture was incubated 37 °C for the times indicated. The
reaction was stopped by the addition of 5% trichloroacetic acid,
vortexed, and maintained on ice. The samples were centrifuged for 5 min
at 15,000 × g and the supernatant spotted on P-81
phosphocellulose chromatography paper. The paper was washed with 0.2%
phosphoric acid six times to remove unincorporated ATP and the filters
counted. The incorporation of 33P into angiotensin II was
used as a measure of EGFR kinase activity.
Treated cells were washed twice
with ice-cold PBS and lysed in ice-cold lysis buffer: 50 mM
HEPES, pH 7.4, 50 mM sodium pyrophosphate, 50 mM sodium fluoride, 50 mM sodium chloride, 5 mM EDTA, 5 mM EGTA, 2 mM sodium
orthovanadate, 0.5 mM PMSF, 10 µg/ml leupeptin, and
0.01% Triton X-100. An aliquot (20 µl) of each MAPK antibody and 40 µl of 50% protein A-Sepharose beads were added to each of the cell
lysates to immunoprecipitate both p42 and p44 forms. The incubation
mixtures from an equal number of cells were allowed to tumble overnight
at 4 °C. Immunoprecipitates were washed three times in lysis buffer
and then resuspended in 50 µl of the reaction buffer: 50 mM -glycerophosphate, pH 7.3, 1.5 mM EGTA,
0.1 mM sodium orthovanadate, 1 mM
dithiothreitol, 10 µM calmidazolium, 10 mM
MgCl2, 10 µg/ml leupeptin, 10 µg/ml aprotinin, 2 µg/ml pepstatin A, 1 mM benzamidine, 0.3 mM
adenosine triphosphate, and 0.5 mg/ml myelin basic protein. 10 µCi of
[
-33P]ATP was added to the reaction. The reaction was
allowed to proceed for 30 min at 37 °C and was quenched by adding 50 µl of ice-cold trichloroacetic acid. The mixture was then spotted on
P-81 phosphocellulose chromatography paper and allowed to dry. The
filters were washed five times in 0.5% phosphoric acid, followed by a
final wash in ethanol (21). The filters were allowed to dry and the
radioactivity counted on a Packard 2000CA scintillation counter.
EGFR down-regulation was measured
using a modification of the procedure described by Carpenter (22). 80%
confluent 60-mm dishes were rinsed twice with IBR and treated with or
without 1 µM (13S)-HPODE for 1 h at
37 °C. Cells were then treated with cold EGF (10 ng/ml) plus or
minus 1 µM (13S)-HPODE for 0, 15, 30, 60, and
120 min at 37 °C. At the end of incubation period, the plates were
washed twice with PBS, and incubated 2 min in 0.2 M acetic
acid, 0.5 M NaCl (ice-cold) followed by a second incubation
for 20 s to strip cold EGF. Cells were then rinsed twice with PBS
and incubated with 125I-EGF (1 nM) in IBR plus
0.1% BSA on ice for 2 h. Cells were then washed three times with
PBS and scraped, and the bound 125I-EGF was quantified by
-counting.
70-80% confluent
cells in 100-mm dishes were starved for 18 h at 37 °C in IBR
containing 0.1% fetal bovine serum with no antibiotics. The media were
replaced with Hepes-buffered IBR (0.027% BSA) either containing
(13S)-HPODE (the final concentration of ethanol was less
than 0.02%) or vehicle. The cells were incubated for 1 h. EGF (10 ng/ml) was added and the cells incubated for 1 min at room temperature.
The EGF receptor kinase activity was blocked by addition of 50 µM methyl 2,5-dihydroxycinnamate, a protein-tyrosine kinase inhibitor, in Me2SO and the cells harvested at the
times indicated. To harvest, the cells were moved onto ice, the medium removed and replaced with ice-cold PBS containing BSA (0.1%) and sodium orthovanadate (2 mM). The medium was removed and the
cells lysed immediately with boiling sample buffer and boiled for 5 min. The lysate was dispersed with a 25-gauge needle and centrifuged, and the supernatant frozen at 70 °C for Western analysis with the
anti-phosphotyrosine monoclonal antibody. Gels were loaded with lysate
from 4 × 105 cells.
The
addition of EGF to serum-starved SHE cells stimulated a rapid
autophosphorylation of the EGFR (170 kDa) as measured by Western
analysis using an antibody for phosphorylated tyrosine residues
(anti-Tyr(P)). The tyrosine phosphorylation peaked between 0.5 and 1.5 min after the addition of EGF and then rapidly declined (data not
shown). An EGF concentration dependence was then determined in the
supB+ and supB variants measured at the peak
of EGFR autophosphorylation. The dose-response relationship between EGF
concentration and EGFR autophosphorylation was very similar for both
variants (data not shown). These findings are in agreement with
previous data that report the number of EGFR, and their affinity for
EGF are approximately the same for both SHE cell variants (4). We
observed a number of other proteins, which were also
tyrosine-phosphorylated in response to EGF. One of the most intense
bands migrated on SDS-PAGE at 120 kDa. The tyrosine phosphorylation of
the 120-kDa band also showed a concentration-dependent
effect of EGF similar to that observed for EGFR autophosphorylation in
both variants.
We attempted to estimate the levels of the EGFR in lysates prepared
from both the supB+ and supB cells by Western
analysis, but the low level of the receptor and the poor
cross-reactivity with anti-EGFR antibodies prevent adequate estimation.
Lysates were then prepared from the same number of cells, the protein
estimated, and EGFR immunoprecipitated. The immunocomplexes were then
analyzed by Western analysis using anti-EGFR antibody for blotting. As
illustrated in Fig. 1A, the level of the EGFR
based on protein or cell number was the same in lysates prepared from
both SHE cell variants.
13(S)-HPODE and EGFR Tyrosine Phosphorylation
Previously, we
reported that the addition of the linoleic acid metabolite,
(13S)-HPODE/(13S)-HODE, enhanced
EGF-dependent mitogenesis in the supB+ SHE
cells but not in the supB variant (9). Other structurally
related lipid metabolites were ineffective (11). (13S)-HPODE
was approximately 10 times more potent than (13S)-HODE in
enhancing EGF-dependent mitogenesis in supB+
cells. Both the hydroperoxide and the alcohol were used in experiments to examine possible modulation of the EGF signaling pathway and the
concentrations adjusted to account for the difference in potency.
Since EGF binding to its receptor initiates tyrosine phosphorylation of
signaling proteins in the pathway, we measured
EGF-dependent tyrosine phosphorylation by Western analysis
using an antibody that detects phosphorylated tyrosine residues. The
addition of EGF to either the supB+ or supB
cells increased tyrosine phosphorylation of the EGFR and a band at 120 kDa (Fig. 1B) tentatively characterized as GAP. GAP
identification was based on the apparent molecular weight and
immunoreactivity with specific GAP antibody. After immunoprecipitation
of cell lysates with the anti-GAP antibodies, the 120-kDa
tyrosine-phosphorylated band was not detected in subsequent analysis of
the supernatant (data not shown). The addition of (13S)-HODE
significantly increased EGF-dependent tyrosine
phosphorylation in the supB+ cells. In the
supB
variant, the addition of (13S)-HODE
caused significantly less increase in tyrosine phosphorylation than
observed for the supB+ cells with EGF and
(13S)-HODE treatment. This finding suggests possible
up-regulation of the EGF signaling pathway in supB+ cells
by (13S)-HODE.
Initiation of the EGF signal
transduction pathway leads to the downstream activation of MAP kinases,
which are likely to mediate many of the nuclear activation events
associated with EGF-dependent mitogenesis (16). MAP kinase
is activated by threonine and tyrosine phosphorylation catalyzed by the
dual functional MAP kinase kinase (MAPKK or MEK) (23-25). If
(13S)-HODE/(13S)-HPODE up-regulates the EGF
signaling pathway, then (13S)-HPODE when added to the cells
should eventually result in an increase in MAP kinase tyrosine phosphorylation and activity in the supB+ and not in the
supB variant. We thus measured tyrosine phosphorylation
of p42 MAPK and p44 MAPK stimulated by EGF in the presence and absence
of (13S)-HPODE, 10 nM to 1 µM. The
results for treatment with 1 µM (13S)-HPODE
are shown in Fig. 2. In the absence of EGF or on
addition of (13S)-HPODE alone, little or no tyrosine
phosphorylation of either p42 or p44 MAPK was detected in the quiescent
SHE cells. Treatment with 10% fetal calf serum or EGF (10 ng/ml)
stimulated tyrosine phosphorylation of both the p42 and p44 proteins. A
biphasic response to EGF was observed, with initial enhancement in the tyrosine phosphorylation observed at early times (2-30 min) followed by a down-regulation of the signal. A second phase of increased tyrosine phosphorylation of MAP kinase occurs at 90-100 min. The addition of (13S)-HPODE enhanced the magnitude of MAP kinase
tyrosine phosphorylation at all time points examined in the
supB+ cell line (Fig. 2). Furthermore, tyrosine
phosphorylation of MAP kinase was prolonged in the presence of
(13S)-HPODE. The addition of (13S)-HPODE did not
alter the tyrosine phosphorylation of either p44 or p42 in the
supB
variant (data not shown). Thus,
(13S)-HPODE appears to up-regulate the entire EGF signaling
pathway exclusively in supB+ cells.
To determine if the observed changes in the phosphotyrosine state of
MAP kinase correlates with changes in MAP kinase activity, the activity
was measured in immunoprecipitates from supB+ SHE cells
using anti-MAP kinase antibodies, which react with both the p42 and p44
proteins. Quiescent cells were treated with EGF for 2 or 90 min ± (13S)-HPODE (pretreatment for 1 h prior to EGF
addition). The MAP kinase activity of the immunoprecipitates was
determined by measuring the incorporation of [-33P]ATP
into myelin basic protein as described under "Experimental Procedures." As shown in Table I, the addition of EGF
stimulated the MAP kinase activity as expected. At the 2-min time
point, the combination of EGF and (13S)-HPODE increased MAP
kinase activity in the immunocomplex 1.5-fold compared with the
corresponding cells incubated with EGF alone. After 90 min of
treatment, a greater than 2-fold increase in MAP kinase activity was
observed in the immunocomplex obtained from cells treated with the
lipid metabolite and EGF as compared with cells treated with EGF alone.
Thus, the addition of the linoleic acid metabolite to supB+
cells increased and prolonged EGF-dependent MAP kinase
tyrosine phosphorylation and MAP kinase activity.
|
One possible explanation for the
up-regulation of the EGF signaling pathway in the supB+
cells by (13S)-HPODE is the selective increase in the EGFR
kinase activity in the supB+ cell line. The EGFR kinase
activity in SHE cell lysates was estimated by measuring the
incorporation of [-33P]ATP into angiotensin II, which
is a good substrate for the EGFR kinase (20). As shown in Fig.
3, little or no difference was observed in the EGFR
kinase activity of the two SHE cell lines. Furthermore, incubation
in vitro with (13S)-HPODE did not directly alter
EGFR kinase activity. Thus (13S)-HPODE does not up-regulate the EGF pathway in the supB+ cells by direct activation of
the receptor kinase activity.
EGFR Down-regulation
The linoleic acid metabolite could also
act by altering the rate of receptor internalization. To test this
hypothesis cells were incubated with EGF in the presence and absence of
(13S)-HPODE for different times at 37 °C and EGFR
internalization measured by a method published previously (22). The
rate and extent of receptor down-regulation was very similar in the
supB+ and supB cells (data not shown).
Pretreatment of the supB+ cells with either vehicle (Fig.
4A) or (13S)-HPODE (Fig.
4B) for 1 h prior to the incubation with EGF also did
not change the rate and extent of receptor down-regulation. Similar
results were obtained with supB
cells.
EGFR Autophosphorylation
The phosphorylation and
dephosphorylation of the EGFR was examined in both cell lines in the
presence and absence of the lipid metabolite (Fig. 5).
The EGFR was rapidly tyrosine-phosphorylated in response to EGF with
maximum phosphorylation observed at 1-2 min after the addition of EGF
and then rapidly declined. The duration of EGF-dependent
tyrosine phosphorylation appears to be greater in the
supB than in the supB+ variant.
The addition of (13S)-HPODE 1 h prior to the addition
of EGF to quiescent SHE cells augmented EGFR tyrosine phosphorylation in supB+ but not in supB cells. In the
supB+ cells, (13S)-HPODE clearly stimulated an
increase in autophosphorylation of EGFR at the 1-10-min time points
(Fig. 5). In contrast, (13S)-HPODE did not significantly
alter the phosphotyrosine state of the EGFR in supB
cells
(Fig. 5). Shown in Fig. 5 are the results of a typical experiment in
which the cells were treated with 1 µM lipid, which produced maximum enhancement of tyrosine phosphorylation. A
concentration-dependent effect was observed from 10 nM to 1 µM (data not shown). This effect of
(13S)-HPODE appears to be on both the magnitude and duration
of the tyrosine phosphorylation signal in the supB+ cells.
Treatment of cells with (13S)-HPODE alone did not alter the
tyrosine phosphorylation pattern in either SHE supB cell (data not
shown). Other corresponding arachidonic acid metabolites (15-HETE and
12-HETE) did not increase tyrosine phosphorylation in EGF stimulated
supB+ cells. (13R)-HODE, which was inactive for
enhancement of EGF-dependent mitogenesis, also did not
increase EGF-dependent tyrosine phosphorylation (data not
shown). In addition to modulating EGFR tyrosine phosphorylation, the
addition of (13S)-HODE or (13S)-HPODE to the
supB+ cells alters the magnitude and the duration of
tyrosine phosphorylation of the protein band at 120 kDa (immunoreactive
with anti-GAP antibodies). This observation fits the hypothesis that
the linoleic acid metabolite up-regulates the EGF pathway in
supB+ cells starting with the EGFR and transmitted
downstream to GAP, MAP kinase, and other signal transduction
proteins.
The biological effect of the linoleic acid metabolite on EGFR
autophosphorylation can be demonstrated even more dramatically in
experiments conducted at room temperature rather than 37 °C. Shifting to lower temperature slows down the rate of EGFR
dephosphorylation and ligand-induced receptor internalization and
degradation (14). The room temperature experiments allow one to study
the tyrosine phosphorylation signal at much slower kinetics. As shown
in Fig. 6 (panel A versus panel C),
(13S)-HPODE enhanced EGF-stimulated tyrosine phosphorylation
of both EGFR and GAP as compared with cells treated with EGF alone. The
action of (13S)-HPODE increased both the magnitude and
duration of protein-tyrosine phosphorylation in this analysis. Western
immunoblots using antibodies specific for GAP indicated that the level
of this protein did not change during the course of the experiment as
shown in panel B of Fig. 6. Thus, the
(13S)-HPODE-stimulated changes noted with the
anti-phosphotyrosine immunoblots appear to be due to an effect on the
tyrosine phosphorylation state of GAP and not due to alterations in
protein levels. These results suggest that (13S)-HPODE
attenuates the dephosphorylation of the EGFR, which subsequently
results in enhancement of EGF-dependent tyrosine
phosphorylation of key signal transduction proteins in the
supB+ but not in the supB variant.
EGFR Dephosphorylation
The rate of EGFR dephosphorylation was
estimated using a modification of the assay described by Bohmer
et al. (26). EGFR dephosphorylation occurred very rapidly
(Fig. 7) in the supB+ cell line.
Dephosphorylation of the EGFR appeared to reach completion after 2 min.
In contrast, the dephosphorylation of the EGFR was significantly slower
in the supB variant. Dephosphorylation was nearing
completion at 6-8 min after EGF binding. The difference in the rate of
EGFR dephosphorylation observed in this experiment is in agreement with
the other data on EGF-dependent tyrosine phosphorylation.
The effect of (13S)-HPODE on dephosphorylation of the EGFR
of the supB+ and supB
cells was examined
following preincubation of the cells for 1 h with the lipid
metabolite prior to the addition of EGF. As shown in Fig.
8, (13S)-HPODE significantly altered the rate
of receptor dephosphorylation in the supB+ cells. No
significant response was observed in the supB
cell (data
not shown). In the presence of (13S)-HPODE, receptor dephosphorylation for the supB+ cells was not complete
until 7 min after EGF addition, while in the absence of
(13S)-HPODE, dephosphorylation was complete by 3 min. The
effect of (13S)-HPODE on EGFR dephosphorylation in the
supB+ and supB
cells at several
concentrations of (13S)-HPODE was examined by measuring the
EGFR tyrosine phosphorylation 2 min after the addition of EGF. With
increasing concentration of (13S)-HPODE, a greater intensity
of tyrosine phosphorylation of the receptor was observed (Fig.
9) in the supB+ cells, indicating inhibition
of receptor dephosphorylation was dependent on the concentration of
(13S)-HPODE. Little or no effect of (13S)-HPODE
on EGFR dephosphorylation was observed in the supB
cells.
Thus, (13S)-HPODE appears to up-regulate the EGF pathway in
supB+ cells by inhibiting dephosphorylation of the EGFR,
resulting in an increased activation of the EGFR tyrosine kinase and
downstream kinase signaling.
The binding of EGF to its receptor activates the receptor tyrosine
kinase activity, which in turn tyrosine phosphorylates other key
signaling proteins and begins a series of phosphorylation events that
culminates in activation of transcription factors leading to cell
division. For Syrian hamster embryo fibroblasts, the addition of EGF to
serum-starved quiescent cells initiates cell proliferation (4) and
stimulates the metabolism of linoleic acid to (13S)-HPODE.
In recent years considerable progress was made in understanding this
transduction pathway, but the nature and the importance of linoleic
acid metabolism in the signaling pathway is not recognized nor fully
understood. The formation of (13S)-HPODE and
(13S)-HODE by an apparent 15-lipoxygenase appears to be
regulated by the EGFR tyrosine kinase activity (7), suggesting that
formation of (13S)-HPODE/(13S)-HODE is a
component of the EGF signaling pathway in these cells. The addition of
lipoxygenase inhibitors that block the 15-lipoxygenase without effects
on the EGFR kinase activity attenuates the EGF response as measured by a mitogenesis assay (4). Furthermore, exogenous (13S)-HPODE and (13S)-HODE enhanced the EGF-dependent
mitogenesis 4-5-fold in the SHE supB+ phenotype, but did
not alter EGF-dependent mitogenesis in the supB variant (4). The linoleic acid metabolites were not
mitogens alone, but appeared to specifically enhance the EGF mitogenic signal. Other metabolites of arachidonic acid and linoleic acid formed
by either prostaglandin H synthase or lipoxygenases were either
inactive or inhibited mitogenesis, suggesting that
(13S)-HPODE/(13S)-HODE have exclusive ability to
up-regulate the EGF signaling pathway in these cells (11).
The EGF signaling pathway in both SHE cell lines was examined in an
effort to determine mechanisms by which the linoleic acid metabolites
enhance EGF-dependent proliferation and to explore potential differences in the signaling pathway in the two cell types.
No differences were observed in either the number of EGF receptors, the
kinase activity of the receptor, or the internalization and degradation
of the receptor in the two SHE cell lines. (13S)-HPODE in
the presence of EGF did not directly alter either the kinase activity
or internalization of EGFR but a more intense tyrosine phosphorylation
of the EGFR and GAP was observed in the supB+ cells,
indicating that (13S)-HPODE up-regulated the EGF signaling pathway at an early event in the cascade. This up-regulation in supB+ cells resulted in an increase in MAP kinase tyrosine
phosphorylation and MAP kinase activity. In the presence of
(13S)-HPODE, the supB+ cells appear to resemble
the supB variant response to EGF. The up-regulation of
the EGF signaling pathway is consistent with previously reported
modulation of EGF-dependent mitogenesis in SHE fibroblasts
by the linoleic acid metabolite (4). As observed for
EGF-dependent mitogenesis in supB+ cells, the
ability to modulate EGF-dependent tyrosine phosphorylation is exclusive for (13S)-HODE and (13S)-HPODE.
Other arachidonic acid and linoleic acid metabolites were inactive.
This suggests a specific but uncharacterized biological target for the
metabolite.
An important difference in the EGF signaling pathway of
supB+ and supB cells is the observation of a
longer duration of EGF-dependent tyrosine phosphorylation
in the supB
variant. The extended duration of tyrosine
phosphorylation was noted for both EGFR and GAP. EGFR
dephosphorylation, measured directly in intact cells, was slower in the
supB
variant compared with the supB+
phenotype, an observation that provides a rational explanation for the
more prolonged tyrosine phosphorylation of signaling proteins of the
EGFR cascade. Furthermore, incubation of the supB+ cells
with (13S)-HPODE resulted in a
concentration-dependent attenuation of EGFR
dephosphorylation, while no effect on EGFR dephosphorylation was
observed in the supB
cells. Thus, it appears that an
important difference between supB+ and supB
cells is the rate of EGFR dephosphorylation, which also appears to be
the site for (13S)-HPODE modulation of the EGF signaling pathway. The mechanism for the attenuation of EGFR dephosphorylation by
(13S)-HPODE is not known and is made more complex by a poor general understanding of EGFR dephosphorylation. Negative regulation of
growth factor receptor by receptor-directed protein-tyrosine phosphatases has been suggested. Differences in functional activity of
tyrosine phosphatases specific for the EGF pathway could provide an
explanation for the phenotypic difference between the two SHE cell
lines and may be a target for the enhancement of EGFR signaling by
(13S)-HPODE. Our current hypothesis is that
(13S)-HPODE modulates the interaction of a tyrosine
phosphatase with the EGFR. On-going studies are designed to test this
hypothesis.
The lack of a response to (13S)-HPODE in the
supB variant and the difference in the
EGF-dependent tyrosine phosphorylation of the two SHE cell
lines suggest that during neoplastic progression of the SHE cells, gene
deletions or mutations of EGF signaling proteins have occurred with the
loss of the tumor suppressor phenotype (14-16). One of these proteins
appears to be the site or the "receptor" responsible for the
modulation of the EGF signaling pathway by (13S)-HODE/(13S)-HPODE. Further studies of the
EGF signaling pathway in the two SHE cell lines should shed light on
the molecular events associated with the loss of the tumor suppressor
phenotype and should provide the necessary information to understand
how the linoleic acid metabolites modulate the EGF signaling
pathway.
Ligand-induced activation of the EGFR tyrosine kinase appears to be the essential biochemical event for further EGF mitogenic signal transduction (17-19). In the work reported here, we demonstrate that the linoleic acid metabolite (13S)-HPODE modulates the tyrosine phosphorylation of the EGFR signaling pathway. Downstream tyrosine phosphorylation of MAP kinase is enhanced. Phosphorylation and subsequent activation of MAP kinase allows for transduction of the mitogenic signal from the cell membrane to the nucleus (5). The cytosolic form of phospholipase A2 (cPLA2) is a substrate for MAP kinase, with phosphorylation of cPLA2 resulting in increased enzymatic activity (27, 28). Thus, (13S)-HPODE enhancement of MAP kinase activity could potentially result in phospholipase activation generating additional lipid mediators involved in a positive feedback loop for mitogenic signaling. In earlier work we found that another cis-polyunsaturated fatty acid, arachidonic acid, can augment MAP kinase activity (21). In these previous studies with vascular smooth muscle cells, arachidonic acid at micromolar levels was found to directly increase MAP kinase tyrosine phosphorylation and activity. This effect was mediated in part by the 15-lipoxygenase metabolite, 15-hydroxy-eicosatetraenoic acid (15-HETE). In these cells the arachidonate compounds stimulated MAP kinase in the absence of growth factors. The implication from this work is a role for arachidonic acid in smooth muscle cell proliferation in both normal and pathological conditions such as atherosclerosis. Our current finding of modulation of EGF-dependent MAP kinase activity by a linoleic acid metabolite in SHE fibroblasts further demonstrates the importance of specific lipid compounds in regulating mitogenesis in multiple cell types and in various physiological processes. (13S)-HPODE appears to be a highly specific modulator of the tyrosine phosphorylation state of the EGF pathway from EGFR autophosphorylation to MAP kinase activity by altering the balance between kinase and phosphatase activities (29-31) associated with this signaling pathway.
The importance of arachidonic acid and linoleic acid metabolites in regulating cell proliferation is supported by both animal and human epidemiology studies. For example, nonsteroidal anti-inflammatory drugs, such as aspirin and sulindac, reduce the incidence and mortality of colon cancer and will induce regression of rectal polyps in patients with familial adenomatous polyposis (32-34). Aspirin and other nonsteroidal anti-inflammatory drugs inhibit the activity of prostaglandin H synthase, which can convert arachidonic acid to prostaglandins but can also metabolize linoleic acid to (9R)- and (13S)-HODE (35-37). A risk factor associated with the etiology of breast cancer arises from animal and human epidemiology studies linking high dietary fat with increased incidence of breast cancer. Linoleic acid is the major polyunsaturated fatty acid consumed in the human diet and has been demonstrated to stimulate cell proliferation and metastasis in human breast carcinoma cells (38, 39). Our investigation indicates that linoleic acid metabolism may play a central role in transduction of the EGF mitogenic signal from the cell surface to the nucleus and may provide a useful model for understanding neoplastic progression at a molecular level.