(Received for publication, February 4, 1997, and in revised form, April 28, 1997)
From the Department of Pharmacology, University of Colorado Health Sciences Center, Denver, Colorado 80262 and Denver Veteran Affairs Medical Center, Denver, Colorado 80220
p38 is a member of the mitogen-activated protein (MAP) kinase superfamily activated by stress signals and implicated in cellular processes involving inflammation and apoptosis. Unlike the extracellular signal-regulated kinases (p42 and p44 MAP kinases), which are stimulated by insulin in many cell types, p38 activity is inhibited by insulin in postmitotic fetal neurons for which insulin is a potent survival factor (Heidenreich, K. A., and Kummer, J. L. (1996) J. Biol. Chem. 271, 9891-9894). These data suggested that insulin's effects on neuronal survival are mediated by inhibition of a p38-mediated apoptotic pathway. To better understand the relationship between p38 activity and cell survival, we induced apoptosis in two cell lines and examined the ability of insulin or a specific p38 inhibitor (a pyridinyl imidazole compound PD169316) to block p38 activity and cell death. In Rat-1 fibroblasts grown in the presence of serum, p38 activity was undetectable by immune complex assays, and the number of apoptotic cells was very low (<0.5%). After the removal of serum for 16 h, p38 activity was markedly elevated, and apoptosis increased by 14-15-fold. Insulin (50 ng/ml) inhibited p38 activity by ~70% and blocked apoptosis by at least 80%. PD169316 also blocked p38 enzyme activity and apoptosis by approximately 80%. Similar results were obtained in differentiated PC12 cells that were deprived of nerve growth factor (NGF) for 16 h. In the presence of NGF, p38 activity and the number of apoptotic cells was very low (~1.0%). After NGF withdrawal, p38 activity was selectively elevated and apoptosis increased to 15%. Both insulin and PD169316 markedly blocked the increase in p38 activity and apoptosis. The MAP kinase kinase inhibitor, PD98059, had no effect on apoptosis in Rat-1 fibroblasts and only partially blocked apoptosis in PC12 cells. PD98059 did not influence insulin's ability to block apoptosis, indicating that the extracellular signal-regulated kinase pathway does not mediate insulin's survival effects. These data further support the role of p38 in cellular apoptosis and support the hypothesis that insulin promotes cell survival, at least in part, by inhibiting the p38 pathway.
Mammalian MAP1 kinases are serine/threonine kinases that mediate intracellular signal transduction (for reviews, see Refs. 1-3). Members of the MAP kinase family share conserved structural domains and are activated by dual phosphorylation within a Thr-X-Tyr motif by specific upstream MAP kinase kinases. The mammalian MAP kinases can be subdivided into the extracellular signal-regulated kinases (ERKs), the Jun N-terminal kinases (JNKs), and the p38 MAP kinases. ERK1 and ERK2 are regulated by mitogens and growth factors via a Ras-dependent pathway involving sequential activation of a MAP kinase kinase kinase and a MAP kinase kinase (MEK) (4). The JNK and p38 kinases are activated by proinflammatory cytokines, hyperosmolarity, heat shock, endotoxin, and other cellular stresses (5-8). The activation cascades for JNK and p38 appear to be distinct. The MAP kinase kinase MEK4 activates JNK but not p38 in vivo (9), whereas MEK3 and MEK6 activate p38 and not JNK (10, 11).
p38 phosphorylates and activates the transcription factors, ATF-2 and
MEF2C (11, 12), indicating a role for p38 in transcriptional regulation. p38 activates MAPK-activated protein kinase 2, which, in
turn, phosphorylates the small heat shock protein 27 (8). Other
investigators have provided evidence that p38 regulates cytokine
production (13), platelet aggregation (14), and neuronal apoptosis (15,
16). p38 was identified as the target for a group of pyridinyl
imidazole compounds that block the production of interleukin-1 and
tumor necrosis factor- from monocytes stimulated with endotoxin
(17). These compounds, such as SB202190, fail to block the activity of
ERK1/ERK2, JNK, and a number of other protein kinases (17-19). A
recent paper describing the crystal structure of p38 in complex with a
pyridinyl imidazole inhibitor describes the structural basis for the
specificity of these compounds (19).
Our laboratory recently reported that insulin inhibits the activity and tyrosine phosphorylation state of p38 MAP kinase in primary neuronal cultures for which insulin is a potent survival factor (20). These data taken together with recent findings by Xia et al. (15). and Ichijo et al. (16) suggested that insulin promotes neuronal survival by inhibiting an apoptotic pathway regulated by p38 MAP kinase. Here we report that inhibition of p38 MAP kinase by insulin or by a pyridinyl imidazole inhibitor blocks apoptosis induced by trophic factor withdrawal in a non-neuronal and neuronal cell line. These data provide direct evidence for a key role of p38 MAP kinase in cellular apoptosis.
Porcine monocomponent insulin was generously supplied by Dr. Ronald Chance of Lilly. Rat-1 fibroblasts were obtained from Dr. Jerrold Olefsky (UCSD, San Diego, CA), and rat pheochromocytoma (PC12) cells were obtained from Dr. Gary Johnson (National Jewish Hospital, Denver, CO). Tissue culture media, fetal calf serum (FCS), and nerve growth factor (NGF, 2.5 S) were obtained from Life Technologies, Inc. Horse serum and newborn calf serum were obtained from Gemini Bio Products, Inc. (Calabasas, CA). All other chemicals were obtained from Sigma. Polyclonal anti-p38 antibodies (C-20) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Apotag, a TUNEL assay kit for detecting DNA fragmentation, was purchased from Oncor (Gaithersburg, MD). The MEK inhibitor, PD98059, and the p38 inhibitor, 4-[5-(4-fluorophenyl)-2-(4-nitrophenyl)-2H-imidazol-4-yl]-pyridine (patent WO95-03297), also referred to as PD169316, were generous gifts from Dr. Alan Saltiel (Parke-Davis Pharmaceuticals, Ann Arbor, MI).
Cell CultureRat embryo fibroblasts (Rat-1) were maintained at 37 °C in a humidified atmosphere (5% CO2 in air) in Dulbecco's minimal essential medium containing 10% FCS, 100 µg/ml penicillin, and 100 µg/ml streptomycin. Rat PC12 cells were plated on dishes coated with polylysine (0.1 mg/ml) in Dulbecco's minimal essential medium containing 5% FCS, 2.5% horse serum, 2.5% newborn calf serum, 100 µg/ml penicillin, and 100 µg/ml streptomycin. The PC12 cells were differentiated by the removal of FCS and the addition of 50 ng/ml NGF (21). Cells were used 7-10 days after the addition of NGF.
p38 MAP Kinase AssaySixteen hours after the removal of
serum from Rat-1 fibroblasts or NGF from differentiated PC12 cells, the
cells were incubated in the absence or presence of insulin (50 ng/ml)
for 15 min at 37 °C. After washing with 2 ml of ice-cold PBS, the
cells were solubilized in 400 µl of ice-cold immunoprecipitation
buffer containing 10 mM Tris, pH 7.4, 1% Triton X-100,
0.5% Nonidet P-40, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 0.2 mM sodium orthovanadate, and 0.2 mM phenylmethylsulfonyl fluoride. The cell lysates were
centrifuged to remove insoluble material, and 200 µg of the
supernatant protein (400 µl, total volume) were incubated with 1 µg
of anti-p38 antibodies for 1 h at 4 °C followed by incubation
with 30 µl of Protein G Plus/Protein A-agarose for an additional
hour. The immunocomplexes were pelleted and washed twice in
immunoprecipitation buffer and then once in kinase wash buffer (50 mM -glycerolphosphate, 1 mM EGTA, 20 mM MgCl2, 100 µM sodium
orthovanadate). The protein kinase assay was initiated by the addition
of 20 µl of 2× reaction buffer (50 mM
-glycerolphosphate, 1 mM EGTA, 20 mM
MgCl2, 100 µM sodium orthovanadate, 0.1 mg/ml
ATF-2 (N-terminal half), 50 µg/ml IP20, a peptide inhibitor of c-AMP
dependent protein kinase, 200 µM ATP, and 0.9 mCi/ml
[32P]ATP) to 20 µl of immune complex. The reaction was
allowed to proceed for 10 min at 30 °C and then terminated by the
addition of 2× Laemmli sample buffer and analyzed by
SDS-polyacrylamide gel electrophoresis using 12% acrylamide gels.
After electrophoresis, the gels were dried and subjected to
phosphoimaging (Bio-Rad, GS-100).
After treating the cells with or
without insulin and various inhibitors, the cells were rinsed with 2 ml
of PBS, pH 7.4, and solubilized in lysis buffer (pH 7.2) containing 50 mM -glycerolphosphate, 5 mM EGTA, 2 mM EDTA, 1 mM dithiothreitol, 2 mM
sodium orthovanadate, 10 µM leupeptin, 5 µg/ml
aprotinin, and 1 mM phenylmethylsulfonyl fluoride. Cell
extracts (3-5 mg of protein) were applied at a flow rate of 0.8 ml/min
to a Mono Q column (Pharmacia FPLC HR5/5) equilibrated with Buffer A
(50 mM
-glycerolphosphate, 5 mM EGTA, 2 mM EDTA, 1 mM dithiothreitol, 2 mM
sodium orthovanadate, pH 7.2). The column was developed at the same
rate with a linear NaCl gradient (0-800 mM) in Buffer A. Fractions (200 µl) were collected and assayed for myelin basic
protein phosphotransferase activity as described previously (20).
Apoptosis was routinely
measured by counting the number of cells with condensed or fragmented
chromatin as described previously (22). Briefly, the cells were washed
in PBS and then fixed in 1% paraformaldehyde for 2 min at room
temperature followed by 70% ethanol in glycine buffer for 10 min at
20 °C. Following fixation, the cells were rinsed 3 times in PBS
and then incubated with the 33258 Hoescht stain (8 µg/ml) for 15 min
at room temperature. The cells were rinsed 3 times in PBS and viewed
under a fluorescent microscope. The amount of apoptosis was expressed
as the percentage of the total number of attached cells that showed
condensed chromatin in eight randomly chosen fields. Typically, at
least 1000 cells per condition were counted. In initial experiments the
number of cells that showed condensed chromatin was compared with the number of cells that stained for DNA fragmentation using the Tunel method (Apotag). The values were found to be identical, although the
Apotag method had higher backgrounds.
To determine if insulin inhibits p38 MAP kinase activity in cells
other than primary neurons and to ascertain the involvement of p38 MAP
kinase in apoptosis, we examined two cell lines in this study. Rat-1
fibroblasts were used as they represent a non-neuronal cell line that
has been used extensively for characterizing insulin signaling (23).
PC12 cells were used because they represent a neuronal cell line where
p38 MAP kinase has been implicated in apoptosis (15). In Rat-1
fibroblasts grown in the presence of 10% FCS, basal p38 activity was
undetectable when assayed by an immunoprecipitation assay. After the
removal of serum in Rat-1 cultures (16 h) p38 activity was markedly
elevated as indicated by a heavily phosphorylated ATF-2 band. The
amount of radioactivity in this band was considered 100%. The addition
of insulin to these cultures inhibited p38 MAP kinase activity by about
70% when quantified by phosphoimaging (Fig.
1). Very similar results were obtained in
PC12 cells. In the presence of NGF, differentiated PC12 cells had very
little p38 activity (phosphorylated ATF-2 was either absent or very
faint). When NGF was removed from PC12 cells, p38 activity was markedly
elevated and consequently inhibited by insulin. Thus, after trophic
factor withdrawal, insulin negatively regulates the activity of p38 MAP
kinase in non-neuronal and neuronal cell lines, and the degree of
inhibition was similar to that previously observed in primary neuronal
forebrain cultures (20). The low basal activity of p38 in both
serum-fed fibroblasts and in NGF-treated PC12 cells is in contrast to
our previous studies using primary neurons where basal p38 activity was
high and accompanied by a high rate of neuronal cell death during the
initial phases of culture (20).
The effect of insulin on apoptosis was measured under the same
conditions that p38 MAP kinase activity was determined, except that
insulin was added at the time of trophic factor withdrawal (Fig.
2). In the presence of 10% FCS, there
was very little apoptosis detected in Rat-1 fibroblasts (<0.5%). When
serum was removed, apoptosis increased by approximately 15-fold.
Insulin added at the time of serum withdrawal protected cells against
apoptosis by about 84%. In PC12 cells, similar observations were made.
The level of apoptosis in differentiated PC12 cells was about 1%. After 16 h of NGF withdrawal, apoptosis increased to 15% of the cell population in agreement with previous data by Xia et
al. (15). As was seen for Rat-1 cells, insulin markedly prevented apoptosis in PC12 cells. Thus, under conditions where apoptosis was
very low, i.e. either in the presence of FCS or in the
presence of NGF, p38 MAP kinase activity was barely detectable. When
FCS or NGF was withdrawn, apoptosis increased and p38 MAP kinase
activity was elevated. Insulin blocked apoptosis induced by trophic
factor withdrawal in both Rat-1 fibroblasts and PC12 cells and
inhibited p38 activity in the same cultures.
At the low concentrations of insulin used in these studies, the effect of insulin in preventing apoptosis is likely mediated by the insulin receptors on these cell lines, although we cannot entirely rule out the involvement of IGF-I receptors (24, 25). Although early studies suggested that insulin's effects on cell survival were mediated by the IGF-I receptor, it is now generally believed that the insulin receptor itself can mediate growth effects in a number of different cell types. This is particularly evident in the nervous system, where insulin receptors lack acute metabolic functions but have potent survival and growth effects (26).
To better understand the relationship of p38 MAP kinase activity and
apoptosis, we examined the effect of a pyridinyl imidazole inhibitor of
p38 MAP kinase inhibitor on apoptosis induced by serum withdrawal in
fibroblasts or by NGF withdrawal in PC12 cells. The p38 MAP kinase
inhibitor used for these studies, PD169316, is a member of the class of
pyridinyl imidazole compounds such as SB202190 and SB203580, which are
selective inhibitors of p38 MAP kinase (17-19). To confirm the
in vivo specificity of the p38 inhibitor in our studies,
PC12 cells were treated with or without the inhibitor (10 µM) at the time of NGF withdrawal. After 16 h, the
cells were solubilized, and cell lysates were fractionated by ion
exchange chromatography. The cell fractions were then assayed for MAP
kinase activity using myelin basic protein as the substrate. In the
presence of NGF, the major peak of MAP kinase activity in PC12 cells
eluted in fractions 23-25 (Fig. 3,
open circles) and was identified as ERK1/2 by immunoblotting
with monoclonal anti-ERK1/2 antibodies (data not shown). After NGF
withdrawal there was little change in ERK activity but a large increase
in a peak eluting in fractions 33-36 (Fig. 3, closed
circles). This peak eluted at the same salt concentration that p38
MAP kinase had previously eluted (20) and was immunoreactive with p38
antibodies (data not shown). If cells were treated with the p38
inhibitor, PD169316, at the time of NGF withdrawal, there was a
complete inhibition of p38 MAP kinase activity with no significant
change in ERK activity (Fig. 3, open boxes). Under our assay
and elution conditions, JNK activity was not detected under basal
conditions or after NGF withdrawal. The data obtained from the column
chromatography illustrate two points. First, the removal of NGF from
differentiated PC12 cells results in the selective activation of p38
MAP kinase with no change in ERK activity. Second, PD169316
specifically blocks p38 MAP kinase activity without effecting ERK
activity.
Experiments were then carried out to examine the effect of the p38
inhibitor on apoptosis in PC12 cells (Fig.
4A) and Rat-1 fibroblasts
(Fig. 4B). As previously observed, withdrawal of NGF from
differentiated PC12 cells increased the number of apoptotic PC12 cells
by about 13-fold. Treatment of PC12 cells with a maximally effective
concentration of PD169316 (10 µM) at the time of serum withdrawal blocked apoptosis induced by serum withdrawal by 83%. Its
effect on blocking apoptosis was similar to the effects of insulin.
Under these conditions where maximal concentrations of insulin and
PD169316 were used, there was no additivity between the two compounds.
In contrast to the p38 MAP kinase inhibitor, the MEK inhibitor, an
upstream inhibitor of ERK, only partially blocked apoptosis by 38%.
Interestingly, the MEK inhibitor did not influence insulin's ability
to inhibit apoptosis, indicating that the ERK pathway is not necessary
for the inhibition of apoptosis by insulin. Thus, under these
conditions, blockade of p38 MAP kinase alone prevents apoptosis induced
by NGF withdrawal. In Rat-1 fibroblasts (Fig. 4B) a similar
picture was observed. Withdrawal of serum induced apoptosis and this
increase in apoptosis was attenuated 70% by the p38 inhibitor and 85%
by insulin. The MEK inhibitor did not block apoptosis, unlike the
partial inhibition seen in PC12 cells, and did not influence
insulin's ability to block apoptosis.
These experiments provide direct evidence that p38 MAP kinase plays a critical role in cellular apoptosis in both a non-neuronal transformed cell line and a more differentiated neuronal cell line. When trophic factors are removed, apoptosis increases and p38 activity is elevated in both cell lines. Selective blockade of p38 MAP kinase in both types of cells prevents apoptosis. These data complement previous findings by Xia et al. (15) that overexpression of constitutively active MEK3, an upstream regulator of p38 MAP kinase, promotes apoptosis, whereas overexpression of a dominant-negative MEK3 prevents apoptosis induced by NGF withdrawal in PC12 cells. Furthermore, at the time of preparing this paper, it was reported that ASK1, a mammalian MEK kinase that activates JNK and P38 signaling pathways, induces apoptosis in lung epithelial cell lines (16). An important distinction between the two previous reports and the present study is that the previous reports could not distinguish between JNK effects and p38 effects because overexpression of the upstream regulators blocks both pathways, whereas the inhibitor used in our studies is specific for p38. p38 has also been implicated in Fas-induced apoptosis in Jurkat cells (27). Fas activation of p38 required the action of interleukin-converting enzyme-like proteases, whereas the activation of p38 by sorbitol or etoposide did not (27), suggesting that there are multiple pathways for activating p38 by different apoptotic stimuli.
The data are less straightforward concerning the role of ERK in neuronal survival. In the studies by Xia et al. (15), overexpression of a constitutively active MEK1, the upstream regulator of ERK, prevented apoptosis in PC12 cells induced by NGF withdrawal. This led the authors to suggest that cell fate is controlled by the balance of activity in the ERK pathway, which promotes proliferation, and the stress pathways that activate programmed cell death. The current study suggests that blockade of the p38 pathway alone is sufficient to prevent apoptosis.
Regarding insulin's ability to support cell survival, one could
propose that the major way insulin supports cell survival is to inhibit
apoptosis by negatively regulating p38 MAP kinase. ERK does not appear
to play a major role in neuronal survival in response to insulin.
Previous studies have shown that ERK is not stimulated by insulin in
primary neuronal cultures (20), and the present study shows that
blockade of the ERK pathway does not influence insulin's ability to
attenuate apoptosis. The latter data are analogous to recent findings
by Creedon et al. (28) showing that the ERK activation is
not required for the actions of cAMP or NGF on neuronal survival.
However, it remains possible that in addition to inhibiting an
apoptotic pathway, insulin stimulates a survival pathway. Recent
studies by Greenberg et al. (29) indicate that insulin and
IGFs support survival of primary cerebellar neurons by activation of a
serine-threonine protein kinase Akt (also known as PKB- or RAC-
).
Akt is a widely expressed kinase that appears to be activated by a
phosphatidylinositol 3-kinase-dependent mechanism. It is
likely that the sum of the activities of multiple signaling pathways is
critical for determining cell fate and that the contribution of one
pathway versus another is cell-specific.
We thank Dr. Lynn Heasley for the recombinant glutathione S-transferase-ATF-2 and Dr. Alan Saltiel for the MEK and p38 inhibitors.