(Received for publication, July 19, 1996, and in revised form, October 30, 1996)
From The Biomedical Research Centre, The University of British Columbia, Vancouver, British Columbia V6T 1Z3, Canada and the § SmithKline Beecham Pharmaceuticals, King of Prussia, Pennsylvania 19406
The mammalian mitogen-activated protein (MAP) kinase homologue p38 has been shown to be activated by pro-inflammatory cytokines as well as physical and chemical stresses. We now show that a variety of hemopoietic growth factors, including Steel locus factor, colony stimulating factor-1, granulocyte/macrophage-colony stimulating factor, and interleukin-3, activate p38 MAP kinase and the downstream kinase MAPKAP kinase-2. Furthermore, although these growth factors activate both p38 MAP kinase and Erk MAP kinases, we demonstrate using a specific inhibitor of p38 MAP kinase, SB 203580, that p38 MAP kinase activity was required for MAP kinase-activated protein kinase-2 activation. Conversely p38 MAP kinase was shown not to be required for in vivo activation of p90rsk, known to be downstream of the Erk MAP kinases. Interleukin-4 was unique among the hemopoietic growth factors we examined in failing to induce activation of either p38 MAP kinase or MAP kinase-activated protein kinase-2. These findings demonstrate that the activation of p38 MAP kinase is involved not only in responses to stresses but also in signaling by growth factors that regulate the normal development and function of cells of the immune system.
The p38 mitogen-activated protein
(MAP)1 kinases (p38/CSBP/RK) are mammalian
homologues of the HOG-1 MAP kinase of Saccharomyces cerevisiae, necessary for their growth under hyperosmolar
conditions (1-3). p38 MAP kinase is activated by physical and chemical
stresses including UV irradiation, heat, and osmotic stress, as well as bacterial lipopolysaccharide, and the pro-inflammatory cytokines tumor
necrosis factor- and interleukin-1 (IL-1) (1, 2, 4, 5). Three
isoforms of p38 MAP kinase are generated by alternate splicing in the
human and have been termed CSBP-1, CSBP-2, and Mxi-2 (3, 6), which may
differ in their downstream targets. Recently two genes encoding other
human p38 MAP kinase family members have been identified, namely p38
(7) and ERK6 (8). Experiments with a series of dominant inhibitory and
constitutively activated mutant proteins have demonstrated that p38 MAP
kinase lies downstream of Rac and Cdc42 (9-11) and three kinases,
MKK3, MKK4, and MKK6 (12-17). These kinases phosphorylate p38 MAP
kinase on threonine and tyrosine residues in a TGY amino acid motif (4, 18), thereby increasing enzymatic activity. In vivo, p38 MAP kinase phosphorylates and activates mitogen-activated protein kinase-activated protein (MAPKAP) kinase-2, which in turn
phosphorylates the small heat shock protein (Hsp) 25/27 (2, 5). In the human, p38 MAP kinase also phosphorylates and activates MAPKAP kinase-3, which is closely related to MAPKAP kinase-2, and also at
least in vitro phosphorylates Hsp25/27 (19). As discussed below, the p38 MAP kinase family also has a role in the phosphorylation and activation of transcription factors including CHOP (20), Elk-1
(15), and ATF-2 (7).
Our interest in p38 MAP kinase was prompted by our observation that a
protein of Mr 38,000, which had a similar
electrophoretic mobility and isoelectric point to p38 MAP kinase (3),
was phosphorylated on tyrosine in primary mast cells stimulated with
either of two hemopoietic growth factors, IL-3 or Steel
locus factor (SLF) (21). These hemopoietic growth factors belong to a
family of cytokines and hormones that are characterized by a four
-helix bundle three-dimensional structure (22). The majority of
ligands of this family bind to heterodimeric receptors consisting of
subunits from the hemopoietin or cytokine receptor superfamily. The
cytoplasmic domains of these receptors characteristically lack
enzymatic activities, and signal transduction is dependent on
activation of cytoplasmic kinases, including members of the JAK and
src kinase families (22, 23). Ligands interacting with
hemopoietin receptors include IL-3 and GM-CSF. IL-3 is active on
pluripotent hemopoietic stem cells, progenitors of all erythroid and
myeloid cells, mature macrophages, eosinophils, megakaryocytes, mast
cells, or basophils (23). GM-CSF is more restricted in its activity,
targeting the progenitors and mature cells of the macrophage,
neutrophil, and eosinophil lineages (23). A minority of 4
-helix
bundle growth factors bind to conventional homodimeric protein-tyrosine
kinase receptors, resembling the platelet-derived growth factor
receptor (24). These include CSF-1, which stimulates the growth of
cells of the macrophage lineage (24, 25), and SLF, which acts on
pluripotent hemopoietic stem cells, the progenitors of a variety of
hemopoietic lineages, mature mast cells, germ cells, and cells derived
from the neural crest (25, 26).
Here we demonstrate that GM-CSF and SLF induced the tyrosine phosphorylation and activation of p38 MAP kinase in hemopoietic cells. This activation of p38 MAP kinase led to the activation of the downstream kinase MAPKAP kinase-2. Similar results were obtained with two other hemopoietic growth factors, IL-3, which like GM-CSF acts through a hemopoietin receptor, and CSF-1, which like SLF acts through a receptor with intrinsic tyrosine kinase activity. Interestingly, another cytokine of this family, IL-4, which also signals through a hemopoietin receptor (22), failed to stimulate the enzymatic activity of p38 MAP kinase or MAPKAP kinase-2. Our results demonstrate that p38 MAP kinase is activated not only by stress but also by growth factors signaling through two distinct classes of receptors.
The anti-p38 MAP kinase anti-serum was raised against full-length CSBP2 as described (3). The anti-MAPKAP kinase-2 antibody and anti-phosphotyrosine antibody 4G10 were purchased from Upstate Biotechnology Inc. (Lake Placid, NY). The anti-p38 MAP kinase antibody used for Western blotting, the anti-p90rsk antibody, and the truncated ATF-2 (1-96) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA), the anti-phospho-p38 MAP kinase-specific antibody was from New England Biolabs (Beverly, MA), recombinant murine Hsp25 was from StressGen Biotechnologies (Victoria, BC, Canada), myelin basic protein (MBP) was from Sigma, and recombinant murine CSF-1 was from R & D Systems (Windsor, ON, Canada). The monoclonal antibody 5A1, specific for CSF-1, was provided by H. Ziltener (The Biomedical Research Centre, Vancouver, BC, Canada). RPMI 1640 was purchased from Canadian Life Technologies (Burlington, ON, Canada), and fetal calf serum was from Intergen (Purchase, NY). Recombinant SLF was provided by James Wieler, and synthetic cytokines were provided by I. Clark-Lewis (BRC).
Cell Lines and Culture ConditionsPrimary mast cells were derived by culturing bone marrow cells from (C57BL/6 × DBA/2) F1 hybrid mice in RPMI 1640 supplemented with 10% fetal calf serum, 10 µM 2-mercaptoethanol, IL-3, and IL-4 for 3 weeks as described (21). The factor-dependent hemopoietic cell lines MC/9 and FD-MACII were passaged in RPMI 1640 supplemented with 10% fetal calf serum, 10 µM 2-mercaptoethanol and IL-3, or CSF-1, respectively, as described (27).
Stimulations and ImmunoprecipitationsPrior to stimulation,
factor-dependent cells were cultured overnight in one-tenth
of the concentration of growth factors in which they were normally
grown and then washed three times with phosphate-buffered saline and
incubated at 107 cells/ml in serum-free medium buffered
with 10 mM HEPES, pH 7.2, for 1 h. Cells were
stimulated with the following doses of synthetic growth factors: IL-4,
20 µg/ml; GM-CSF, 10 µg/ml; and IL-3, 20 µg/ml. All stimulations
had been shown in preliminary experiments to give maximal levels of
tyrosine phosphorylation of respective receptors or cytosolic proteins.
Cells extracts were prepared in solubilization buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 5 mM EDTA, 1% (v/v) Nonidet P-40, 1 mM sodium molybdate, 200 mM sodium orthovanadate, 1 mM sodium fluoride,
50 mM -glycerol phosphate, 10 µg/ml aprotinin, 10 µg/ml soybean trypsin inhibitor, 0.7 µg/ml pepstatin, 2 µg/ml
leupeptin, and 40 µg/ml phenylmethylsulfonyl fluoride). To monitor
the levels of tyrosine phosphorylation in the control and the treated
samples, a portion of each cell lysate was routinely resolved by
SDS-PAGE and immunoblotted with the anti-phosphotyrosine antibody 4G10.
The remaining lysate was subjected to immunoprecipitation, and the
precipitate was analyzed by immunoblotting or kinase assays.
p38 MAP kinase activity was measured
using an immune complex kinase assay with a truncated form of ATF-2 as
a substrate. The cell lysate was mixed with an anti-p38 MAP kinase
anti-serum and 20 µl (packed volume) of protein A-Sepharose beads.
After 2 h the beads were washed extensively with solubilization
buffer and once with kinase assay buffer (25 mM HEPES, pH
7.2, 25 mM magnesium chloride, 2 mM
dithiothreitol, 0.5 mM sodium vanadate, and 25 µM ATP). The kinase reaction was initiated by the
addition of 20 µl of kinase assay buffer containing 2 µg of ATF-2
and 10 µCi of [-32P]ATP and stopped after 20 min at
30 °C by the addition of SDS sample buffer.
To determine the effect of the p38 MAP kinase inhibitor SB 203580 on
the in vivo kinase activity of p38 or Erk MAP kinases, we
investigated effects on activation of the respective putative downstream enzymes MAPKAP kinase-2 and p90rsk. We incubated
107 cells/ml with or without 1 µM SB 203580 for 20 min prior to stimulation. To assay in vivo activity
of p38 MAP kinase, MAPKAP kinase-2 was precipitated using 5 µg of
anti-MAPKAP kinase-2 and 20 µl (packed volume) of protein
G-Sepharose. The kinase assay was performed as described above, except
that 5 µg of recombinant murine Hsp25 was used as substrate. To assay
in vivo activity of the Erk MAP kinase, p90rsk was
immunoprecipitated from aliquots of the same cell lysate using 5 µg
of anti-p90rsk antibody bound to 20 µl of packed protein
G-Sepharose. The kinase assay was initiated by the addition of 20 µl
of kinase assay buffer (20 mM HEPES, 5 mM
magnesium chloride, 1 mM EGTA, 5 µM
mercaptoethanol, and 2 mM sodium vanadate) containing 3 µg of MBP and 10 µCi of [-32P]ATP and stopped
after 10 min at 30 °C by the addition of SDS-sample buffer. In all
kinase assays the phosphorylated proteins were resolved by SDS-PAGE and
visualized by autoradiography.
To determine whether p38 MAP kinase
was involved in responses to IL-3 or SLF, we stimulated primary bone
marrow-derived mast cells with these factors, immunoprecipitated p38
MAP kinase, and assessed its tyrosine phosphorylation by immunoblotting
with the anti-phosphotyrosine specific antibody 4G10. Both SLF and IL-3 induced tyrosine phosphorylation of p38 MAP kinase (Fig.
1A). To test the effects of CSF-1 or GM-CSF
on the activity of p38 MAP kinase, we used murine
factor-dependent cell lines that responded to these factors
(27, 28). Stimulation of the macrophage-like FD-MACII cells with
CSF-1 resulted in rapid tyrosine phosphorylation of p38 MAP kinase
(Fig. 1B). This effect was seen following treatment with
either recombinant murine CSF-1 or a source of natural murine CSF-1,
L-cell conditioned medium. The tyrosine phosphorylation of p38 MAP
kinase was abrogated by the presence of a monoclonal antibody that
neutralizes CSF-1 activity (Fig. 1B), demonstrating that the
induced phosphorylation of p38 MAP kinase was due to CSF-1 and not, for
example, to contamination by endotoxin or to osmotic stress. Cells of
the factor-dependent hemopoietic cell line MC/9 that were
stimulated with GM-CSF likewise exhibited tyrosine phosphorylation of
p38 MAP kinase (Fig. 2, GM, 10,
).
p38 MAP Kinase Activity Is Induced by Growth Factors and Is Inhibited by SB 203580
To demonstrate that the tyrosine phosphorylation of p38 MAP kinase in cells treated with hemopoietic growth factors correlated with increased kinase activity, we immunoprecipitated p38 MAP kinase from either untreated MC/9 cells or cells treated with saturating doses of GM-CSF, SLF, or 0.2 M NaCl. We then assessed the activity of p38 MAP kinase in an in vitro kinase assay using a truncated form of ATF-2 as substrate (Fig. 2). p38 MAP kinase immunoprecipitated from cells stimulated with GM-CSF or SLF exhibited increased levels of kinase activity, which correlated with the levels of tyrosine phosphorylation of the enzyme. The activity of p38 MAP kinase was completely abolished by the inclusion of SB 203580 during the in vitro assay (Fig. 2).
Kinetics of p38 MAP Kinase Tyrosine PhosphorylationTo
examine the kinetics of SLF- and GM-CSF-induced activation of p38 MAP
kinase in MC/9 cells, we used an antibody specific for the activated
form of p38 MAP kinase. This antibody recognizes p38 MAP kinase when
phosphorylated on the tyrosine of the TGY activation motif. The
activation of p38 MAP kinase is dependent on dual phosphorylation on
both threonine and tyrosine residues as previous demonstrated (4, 18).
Treatment with growth factors resulted in a rapid increase in the
levels of tyrosine phosphorylation of p38 MAP kinase (Fig.
3). Cells treated with SLF exhibited detectable phosphorylation of p38 MAP kinase as early as 2 min and maximal phosphorylation of p38 MAP kinase between 5 and 10 min. Cells treated
with GM-CSF had maximal phosphorylation of p38 MAP kinase at around 10 min. In both instances the phosphorylation of p38 MAP kinase was
transient and returned to almost basal levels by 30 min. It is
important to note that we failed to see bimodal tyrosine
phosphorylation of p38 MAP kinase in response to either SLF or GM-CSF
at longer time courses (data not shown).
Failure of Interleukin-4 to Activate p38 MAP Kinase
It was of particular interest to investigate the effect of IL-4 because we had previously shown that unlike any other growth factor except the closely related IL-13 (29), it failed to activate Erk MAP kinases (28). IL-4 failed to induce tyrosine phosphorylation of p38 MAP kinase in primary mast cells (Fig. 1A) or in MC/9 (Fig. 2). As predicted from these results, IL-4 also failed to stimulate enzymatic activity of p38 MAP kinase (Fig. 2). Interestingly, treatment with IL-4 resulted in a small but reproducible reduction in both the levels of tyrosine phosphorylation and the enzymatic activity of p38 MAP kinase (Fig. 2) compared with untreated cells.
Activation of MAPKAP Kinase-2 by Hemopoietic Growth Factors Involves p38 MAP KinaseMAPKAP kinase-2 has been reported to be a
substrate of p38 MAP kinase (4, 5) and to be activated in cells
stimulated with GM-CSF or IL-3 (30). To determine whether the
activation of MAPKAP kinase-2 by hemopoietic growth factors was due to
activation of p38 MAP kinase, we immunoprecipitated MAPKAP kinase-2
from MC/9 cells that had been stimulated with IL-4, GM-CSF, SLF, or 0.2 M NaCl and assessed its activity in an in vitro
kinase assay using recombinant murine Hsp25 as substrate. As shown in
the previous report (30), treatment with GM-CSF resulted in activation
of MAPKAP kinase-2 (Fig. 4A). SLF also
induced strong activation of MAPKAP kinase-2 (Fig. 4A).
However, consistent with our finding that IL-4 failed to stimulate the
enzymatic activity of p38 MAP kinase, IL-4 failed to induce activation
of MAPKAP kinase-2 (Fig. 4A). Indeed, treatment with IL-4
reduced activity of MAPKAP kinase-2 below levels seen in untreated
cells (Fig. 4A), consistent with the reduction in p38 MAP
kinase activity seen in IL-4-treated cells (Fig. 2). IL-4 was active on
these cells as demonstrated by the induced tyrosine phosphorylation of
a protein known to be IRS-22 as indicated
by the arrowhead in Fig. 4B. Thus the ability of hemopoietic growth factors to activate MAPKAP kinase-2 correlated with
their ability to activate p38 MAP kinase.
MAPKAP kinase-2 has been reported to be activated by the Erk MAP kinase
family (32). GM-CSF, IL-3, and SLF induce activation of both Erk (28)
and p38 MAP kinases (Figs. 1A and 2). To investigate which
of these kinases was responsible for the activation of MAPKAP kinase-2
by these hemopoietic growth factors, we used SB 203580, a specific
inhibitor of p38 MAP kinase activity (3, 33). As shown in Fig.
5, pretreatment of cells for 20 min with 1 µM SB 203580 reduced basal activity of MAPKAP kinase-2
(Fig. 5B) and abrogated the ability of GM-CSF (Fig.
5A) or SLF (Fig. 5B) to induce activation of
MAPKAP kinase-2. In in vitro experiments, the enzymatic
activity of Erk-2 was not inhibited by concentrations of SB 203580 that
abolished the enzymatic activity of p38 MAP kinase (33). To confirm
that SB 203580 did not affect the activity of the Erk MAP kinases
in vivo, we stimulated cells with GM-CSF or SLF and
investigated the effects of the compound on the activation of
p90rsk, known to be downstream of the Erk MAP kinases (34). As
shown in Fig. 5, pretreatment of cells with SB 203580 did not affect the activation of p90rsk by GM-CSF or SLF, implying that the
Erk MAP kinase pathway was not affected by SB 203580 and that p38 MAP
kinase did not activate p90rsk.
These experiments demonstrate that a series of hemopoietic growth factors that regulate the normal development and function of cells of the immune system stimulate tyrosine phosphorylation and enzymatic activation of p38 MAP kinase. This was the case for both those hemopoietic growth factors that signal through tyrosine kinase receptors, namely SLF (Figs. 1A and 2) and CSF-1 (Fig. 1B), as well as those that signal through receptors of the hemopoietin receptor superfamily, namely IL-3 (Fig. 1A) and GM-CSF (Fig. 2). The observed correlation of tyrosine phosphorylation and activation of p38 MAP kinase (Fig. 2) is consistent with other evidence that phosphorylation of the tyrosine of the TGY activation motif of p38 MAP kinase is required for enzymatic activity (4, 18). The activation of p38 MAP kinase by GM-CSF and SLF in hemopoietic cells was rapid and transient (Fig. 3) and of the same order as that induced by hyperosmotic stress (Fig. 2). This contrasts with reports that stimulation of HeLa cells with epidermal growth factor activated p38 MAP kinase only weakly (4) and that nerve growth factor failed completely to activate p38 MAP kinase (RK) in PC-12 cells (2). This may reflect differences in the cell types used in these studies. It should be noted that we saw activation of p38 MAP kinase by hemopoietic growth factors in both cell lines and in mast cells from primary cultures.
Given that MAPKAP kinase-2 is known to be an in vivo substrate of p38 MAP kinase (2, 5), our observations that p38 MAP kinase is activated in response to hemopoietic growth factors are consistent with the earlier report that IL-3 and GM-CSF activate MAPKAP kinase 2 (30). Our results confirm this observation and extend it by demonstrating that treatment of cells with SLF also induces MAPKAP kinase-2 activity (Fig. 4). In that the specific inhibitor of p38 MAP kinase, SB 203580, completely abrogated MAPKAP kinase-2 activation by GM-CSF (Fig. 5A) and SLF (Fig. 5B), our data also extend current knowledge by indicating that the activation of MAPKAP kinase-2 by GM-CSF or SLF depended on p38 MAP kinase activity. Thus, despite the fact that treatment of cells with GM-CSF or SLF activates both Erk (28) and p38 MAP kinases (Fig. 2), SB 203580 specifically inhibited the in vivo activation of MAPKAP kinase-2 but not of p90rsk (Fig. 5). These data demonstrate that SB 203580 fails to inhibit the activity of Erk MAP kinases in vivo, consistent with in vitro data that SB 203580 fails to inhibit Erk-2 or p90rsk (33). These results are also consistent with experiments in which in vivo activation of Erk MAP kinases did not result in in vivo activation of MAPKAP kinase-2 (2).
Despite the fact that both Erk and p38 MAP kinases can activate MAPKAP kinase-2 in vitro (2, 5, 32) and both are activated in cells treated with GM-CSF or SLF, p38 MAP kinase activity was essential for activation of MAPKAP kinase-2 in vivo (Fig. 5). This evidence that MAPKAP kinase-2 is activated by p38 MAP kinase and not Erk MAP kinase suggests differences in molecular localization of these enzymes. One possibility is that inactive p38 MAP kinase has a greater affinity for MAPKAP kinase-2 than the Erk MAP kinases and that in vivo MAPKAP kinase-2 is complexed with inactive p38 MAP kinase and not Erk MAP kinases. In keeping with this notion, p38 MAP kinase has been shown to form a complex in vivo with human MAPKAP kinase-3, a close homologue of MAPKAP kinase-2 (70% amino acid identity) (19).
The failure of SB 203580 to inhibit the in vivo activation of p90rsk in the same cells in which it inhibited the in vivo activation of MAPKAP kinase-2 (Fig. 5) is consistent with the notion that p90rsk is activated only by the Erk MAP kinases and not by p38 MAP kinase. These results confirm the suggestion from unpublished in vitro experiments that p38 MAP kinase (RK) was unable to activate p90rsk (2).
It should be noted that the anti-MAPKAP kinase-2 antibody that we used was raised against a 16-amino acid peptide from mouse MAPKAP kinase-2. This peptide differs by only two closely spaced, conservative amino acid substitutions from the corresponding peptide of human MAPKAP kinase-3. If there is a murine homologue of human MAPKAP kinase-3, we would probably have precipitated it with this antibody. In the human, both MAPKAP kinase-2 and MAPKAP kinase-3 are activated by p38 MAP kinase (CSBP) (19). Moreover both enzymes act as Hsp27 kinases in vitro, although whether both phosphorylate Hsp27 in vivo is unclear.
The functional significance of activation of the p38 MAP kinase pathway by growth factors and its relationship to actions of growth factors such as promotion of cell-cycle progression or suppression of apoptosis is unclear. Two possible roles for p38 MAP kinase relate to the regulation of actin polymerization and the activation of transcription factors. Actin polymerization appears to be regulated by phosphorylation of Hsp25, which lowers the affinity of its interaction with the barbed ends of filamentous actin, thus allowing polymerization and the accumulation of filamentous actin (35). Growth factors and serum induce phosphorylation of Hsp25 on the same residues that are phosphorylated in response to stress (36, 37), consistent with the involvement of the same enzyme, most likely MAPKAP kinase-2 or MAPKAP kinase-3. Our results with hemopoietic growth factors demonstrate that inhibition of p38 MAP kinase completely blocks activation of MAPKAP kinase-2 (Fig. 5), suggesting that growth factor-stimulated phosphorylation of Hsp25 reflects activation of p38 MAP kinase. The role of actin polymerization in growth factor action is unclear, although cell growth is inhibited by cytochalasin D (38), which like unphosphorylated Hsp25 prevents elongation of filamentous actin (39). Overexpression of wild-type Hsp25 but not mutant Hsp25 that is unable to become phosphorylated renders cells relatively resistant to the growth inhibitory effect of cytochalasin D (38). This implies a requirement for Hsp25 phosphorylation and filamentous actin accumulation for cell cycle progression. Our observations that cells treated with SB 203580 underwent a rapid and reversible change in cellular morphology3 were also consistent with the notion that p38 MAP kinase regulated actin polymerization.
Transcription factors regulating cell growth may be regulated by the
p38 MAP kinase pathway through several mechanisms. In vivo
activation of p38 MAP kinase increased the transcriptional activity of
Elk-1 (15) implying a role for p38 MAP kinase in the transcriptional
regulation of proteins such as c-Fos. Interestingly p38 but not p38
MAP kinase has been shown to phosphorylate and activate ATF-2 in
vivo (7), suggesting differential regulation of transcription
factors by the p38 MAP kinase family members. Further experiments will
be required to determine whether all family members are activated by
hemopoietic growth factors. Recently p38 MAP kinase has also been shown
to regulate the phosphorylation of CREB and ATF-1 through MAPKAP kinase
2 (40). In addition, p38 MAP kinase has been shown to regulate the
production of IL-1, tumor necrosis factor-
, IL-6, and GM-CSF (3,
41), which are characterized by mRNA containing AU-rich motifs. It
is possible that the production of the transcription factors Myc, Fos,
and Jun, which are translated from mRNA containing the same AU-rich motifs characteristic of cytokine mRNA (42), may also be regulated in a p38 MAP kinase-dependent fashion.
The notion that activation of the p38 MAP kinase pathway has a critical role in cell cycle progression is supported by our unpublished observations that the specific p38 MAP kinase inhibitor SB 203580 inhibited DNA synthesis and cell growth.3 Certainly in Schizosaccharomyces pombe the p38 MAP kinase homologue Spc1 is needed for cell cycle progression under stressful conditions and overexpression of Pyp-1, a tyrosine phosphatase that inactivates Spc1 results in slowing of growth (43). It will be important to develop precise genetic tools to explore the role of p38 MAP kinase in growth factor action and cell cycle progression in mammalian cells.
A recent report implicated p38 MAP kinase in the induction of apoptosis (44). Withdrawal of nerve growth factor from PC-12 cells led to activation of p38 MAP kinase and the c-Jun N-terminal kinases, followed by apoptosis. Overexpression of a constitutively active MKK3, which at least in some cells activates p38 MAP kinase (15, 44), promoted apoptosis, whereas overexpression of a dominant negative MKK3 inhibited apoptosis (44). However these results do not discriminate between roles for p38 MAP kinase and c-Jun N-terminal kinase, because overexpression of a dominant negative MKK3 appears to block activation of both of these kinases (15). In contrast, our results show that p38 MAP kinase is activated by stimulation by growth factors (Figs. 1 and 2) not by their withdrawal. Moreover in that hemopoietic growth factors suppress apoptosis (45), our findings indicate that at least in hemopoietic cells, activation of p38 MAP kinase correlates with the suppression of apoptosis rather than its induction. Finally, treatment of cells with concentrations of SB 203580 that we have shown to inhibit p38 MAP kinase activity in vivo (Fig. 5) failed to inhibit the apoptosis induced by withdrawal of hemopoietic growth factors.3
Interleukin-4 was notable in being the only growth factor we investigated that failed to induce tyrosine phosphorylation (Figs. 1A and 2) and activation of p38 MAP kinase (Fig. 2). The inability of IL-4 to activate p38 MAP kinase accounts for its failure to activate MAPKAP kinase-2 (Fig. 4A) and correlates with its failure to activate Ras (31, 46) and Erk MAP kinase (27, 28). Further experiments will be required to determine whether the activation of p38 MAP kinase by growth factors in hemopoietic cells is Ras-dependent. Interestingly, IL-4 consistently reduced the state of tyrosine phosphorylation and enzymatic activity of p38 MAP kinase to levels below those observed in control cells (Fig. 2). Likewise treatment of cells with IL-4 reduced the in vivo activity of MAPKAP kinase-2 to levels below those in control cells (Fig. 4). However IL-4 did not markedly reduce levels of p38 MAP kinase tyrosine phosphorylation induced by maximal doses of salt,3 suggesting that the capacity of the down-regulatory signal by IL-4 was limited. The basis for the inhibition of the p38 MAP kinase pathway by IL-4 is under further investigation.
In conclusion, these experiments demonstrate that p38 MAP kinase participates in the responses to hemopoietic growth factors that interact with two structurally distinct classes of receptor. They demonstrate that p38 MAP kinase is not only involved in responses to stresses but also in the action of growth factors that regulate the development and function of hemopoietic cells.
We thank I. Clark-Lewis and J. Wieler for preparing cytokines and H. Ziltener for antibodies.