(Received for publication, October 1, 1996, and in revised form, November 15, 1996)
From the Faculty of Bioscience and Biotechnology, Tokyo Institute of Technology, Yokohama 226, Japan
Activation of the c-Jun N-terminal kinase
(JNK)/stress-activated protein kinase pathway in response to
stimulation of the interleukin (IL)-3 or granulocyte-macrophage
colony-stimulating factor (GM-CSF) receptor was examined in mouse
hematopoietic BaF3-derived cell lines (BaF3-N6 and -V2 cells).
Significant increase in the activity of JNK1 was observed within 30 min
following IL-3 or GM-CSF stimulation at physiological concentrations.
Dominant-negative Ras(S17N), which is conditionally expressed in the
presence of isopropyl-1-thio--D-galactoside in BaF3-N6
cells, prevented the IL-3 stimulation of JNK1, whereas
anisomycin-induced JNK1 activation was unaffected. Furthermore, a
deletion mutant of the common
subunit for IL-3 and GM-CSF receptors
that consists of only the membrane-proximal region, including box 1 and
box 2 motifs, was incapable of facilitating JNK1 activity as well as
Ras activation. These results provide evidence that Ras is required for
IL-3-stimulated JNK1 activation. We also examined if constitutively
active Ras(G12V) alone could stimulate JNK1 activity by using the
inducible expression system.
Isopropyl-1-thio-
-D-galactoside induction of Ras(G12V) in the BaF3-V2 cell line caused no significant increase in JNK1 activity, which could be activated by IL-3 or anisomycin. On the contrary, the extracellular signal-regulated kinase/mitogen-activated protein kinase pathway was fully activated following Ras(G12V) induction. Together with these results, it seems likely that the Ras
protein is indispensable for the IL-3 stimulation of JNK1 although Ras
activation by itself is insufficient for JNK1 activation.
Ras family GTP-binding proteins play a pivotal role in regulating proliferation of fibroblast cells, particularly in tyrosine kinase receptor-mediated signaling pathways (1, 2). A serine/threonine kinase cascade comprised of Raf family kinases, mitogen-activated protein kinase (MAPK)1/extracellular signal-regulated kinase (ERK) kinase (MEK), and ERK/MAPKs, functions downstream of Ras, and this cascade is essential for Ras-mediated growth promotion in fibroblast cells (3, 4). Activation of the ERK/MAPK pathway ultimately results in transcriptional activation of a serum response element-containing promoter, for example, the c-fos promoter, through phosphorylation of Elk-1 (5).
Stimulation with various cytokines, including interleukin (IL)-2, IL-3, IL-5, and granulocyte-macrophage colony-stimulating factor (GM-CSF), triggers Ras activation as evidenced by accumulation of an active GTP-bound form, implying that Ras acts as a signal transducer also in hematopoietic cells (6, 7). In the IL-3 signaling system, Raf and ERK/MAPK activation as well as induction of the c-fos promoter are demonstrated to be Ras-dependent (8). In addition to the Ras pathway, several distinct signaling pathways, for instance, the Janus kinase (JAK)/signal transducers and activators of transcription-mediated pathway, are stimulated by the IL-3 receptor (9).
The c-Jun N-terminal kinase (JNK)/stress-activated protein kinase
(SAPK) pathway is stimulated by various kinds of stresses, such as
ultraviolet (UV) irradiation, the high-temperature shock, and
hyperosmolarity. Moreover, protein synthesis inhibitors
(e.g. anisomycin and cyclohexamide) and inflammatory
cytokines (e.g. IL-1 and tumor necrosis factor ) activate
JNK/SAPK (5). More recently, JNK/SAPK activation by stimulation of G
protein-coupled receptors (10, 11), GTPase-deficient mutants of certain
types of G
(12, 13), as well as overproduced G
subunits (14) have been reported.
JNK/SAPK is phosphorylated and subsequently activated by a specific kinase, JNKK1/SEK1/MKK4, which is also activated following phosphorylation by other protein kinases, such as MEKK1 (5), Tpl-2 (15), and MLK-3 (16). Although the precise regulatory mechanism of this kinase cascade is still obscure, involvement of Cdc42 and Rac GTP-binding proteins in regulation of the JNK/SAPK pathway has recently been demonstrated using a transfection assay (17-19). Furthermore, a significant role of a mammalian homologue of the Saccharomyces cerevisiae Ste20 protein, p21-activated kinase, which is directly activated by a GTP-bound form of Cdc42 or Rac (20), has been suggested (18, 21-23). Another Ste20 homologue, germinal center kinase, is also a candidate for an activator of the JNK/SAPK pathway (24).
In this paper, we show activation of JNK1 following IL-3 treatment of
IL-3-dependent hematopoietic cell lines. Conditionally expressed dominant-negative Ras abrogated the activation. Moreover, a
deletion mutant of the IL-3 receptor c subunit that is
unable to stimulate the Ras pathway failed to cause JNK1 activation. On
the other hand, a constitutively active mutant of Ras was unable to
induce JNK1 activation by itself despite its full activity to stimulate
ERK2. Thus, we conclude that Ras is required, but not sufficient, for
JNK1 activation in IL-3-dependent hematopoietic cells.
A plasmid for expression of glutathione S-transferase (GST)-c-Jun(1-223) in Escherichia coli was kindly provided by Michael Karin (University of California, San Diego, La Jolla, CA). GST-c-Jun(1-223) was purified using a glutathione column. Purified mouse IL-3 and human GM-CSF are generous gifts of Robert Kastelein and Satish Menon (DNAX Research Institute of Molecular and Cellular Biology, CA). Antibodies against ERK2 (sc-154), JAK2 (sc-278), and JNK1 (sc-474) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). An anti-phosphotyrosine antibody, 4G10 (05-321), was obtained from Upstate Biotechnology, Inc. (Lake Placid, NY).
Cell CultureBaF3-N6 and BaF3-V2 cell lines (8) were
cultured in RPMI 1640 supplemented with fetal calf serum (10%, v/v),
mouse IL-3 (approximately 1 nM), G418 (1 mg/ml), and
hygromycin (1 mg/ml). BaF3/c and
BaF3/
544 cells, which were kindly provided by Atsushi Miyajima (The University of Tokyo, Tokyo, Japan), were cultured in RPMI
1640 supplemented with fetal calf serum (10%, v/v), mouse IL-3
(approximately 1 nM), and hygromycin (1 mg/ml). For
starvation, cells were incubated in RPMI 1640 supplemented with bovine
serum albumin (1 mg/ml) for 3 h.
Cells (1-2 × 106 per point) were lysed in kinase IP buffer (20 mM Tris-HCl (pH 7.5), 0.5% (v/v) Nonidet P-40, 250 mM NaCl, 3 mM EDTA, 3 mM EGTA, 3 mM -glycerophosphate, 0.1 mM
Na3VO4, 10 µg/ml leupeptin, 1 mM
phenylmethylsulfonyl fluoride), and the supernatant of centrifugation
(15,000 × g) for 10 min at 4 °C was mixed with
protein A-Sepharose CL-4B (Phamacia Biotech Inc.) and an anti-ERK2
(sc-154, 1 µg) or anti-JNK1 (sc-474, 0.25 µg) antibody. The mixture
was incubated for 1.5 h at 4 °C with gentle mixing, and the
precipitate was washed twice with kinase IP buffer, twice with kinase
wash buffer (20 mM Hepes-NaOH (pH 7.6), 0.05% (v/v) Triton
X-100, 50 mM NaCl, 0.1 mM EDTA), and once with
kinase reaction buffer (25 mM Hepes-NaOH (pH 7.6), 20 mM MgCl2, 20 mM
-glycerophosphate, 0.1 mM
Na3VO4, 20 mM
p-nitrophenyl phosphate, 2 mM dithiothreitol).
The precipitated proteins were subjected to the kinase assay within
kinase reaction buffer (30 µl) containing 0.25 mg/ml myelin basic
protein (for ERK/MAPK) or 0.05 mg/ml GST-c-Jun(1-223) (for JNK/SAPK),
and 20 µM [
-32P]ATP (307 TBq/mol) for 20 min at 30 °C. The proteins were separated by SDS-polyacrylamide gel
electrophoresis (13.5% (w/v) (for ERK/MAPK) or 10% (w/v) (for
JNK/SAPK) polyacrylamide) and the radioactivity incorporated into each
substrate was quantitated by an image analyzer (BAS2000, Fuji Film,
Japan).
Cells
(107 per point) were dissolved in IP buffer (50 mM Hepes-NaOH (pH 7.3), 150 mM NaCl, 10% (v/v)
glycerol, 1% (v/v) Nonidet P-40, 2 mM MgCl2, 1 mM EDTA, 100 mM sodium fluoride, 10 mM NaPPi, 20 mM
-glycerophosphate, 1 mM Na3VO4,
1 mM phenylmethylsulfonyl fluoride, 10 µg/ml aprotinin, 1 µg/ml leupeptin), and the supernatant of centrifugation (15,000 × g) for 10 min at 4 °C was used as a cell lysate.
Protein A-Sepharose CL-4B (Phamacia Biotech Inc.) and an anti-JAK2
antibody (sc-278, 2 µg) were mixed gently with the lysate for
1.5 h at 4 °C, and the precipitate was washed twice with IP
buffer and twice with a wash buffer (50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 20 mM MgCl2).
Then, the precipitated proteins were separated by SDS-polyacrylamide
gel electrophoresis and transferred onto a nitrocellulose membrane. The
membrane was stained with an anti-phosphotyrosine antibody (05-321, 2 µg/ml) and enhanced chemiluminescence detection reagents (DuPont
NEN).
As a first step to examine
possible involvement of JNK/SAPK in IL-3-stimulated signal transduction
pathways, we measured the activity of endogenous JNK1 in
IL-3-stimulated as well as unstimulated BaF3-N6 cells. The BaF3-N6 cell
line was isolated from mouse IL-3-dependent hematopoietic
BaF3 cells as a stable transfectant of an inducible expression system
for dominant-negative Ras(S17N) (8). Serum and IL-3-starved BaF3-N6
cells were stimulated with various concentrations of mouse IL-3 for 40 min. Endogenous JNK1 was collected by immunoprecipitation using a
specific antibody, and the activity was measured by an in
vitro kinase assay using GST-c-Jun(1-223) as a substrate. As shown in Fig. 1, A and B,
dose-dependent increase of JNK1 activity was observed,
which paralleled the induction of cell growth as measured by a
colorimetric assay using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (8). Incubation of the cells without mouse IL-3 for
40 min exhibited no effect on JNK1 activity, indicating that the
increased kinase activity was not due to nonspecific stress during the
incubation (data not shown). Time course of JNK1 activation following
IL-3 treatment is demonstrated in Fig. 1, C and
D. An increase in JNK1 activity was measurable at 20 min
after stimulation. The activity reached a maximum at 40 min, and
remained at plateau for at least another 20 min. Increase of JNK1
activity was rather slower than the case of ERK2 (Fig. 1, E
and F). The activity of immunoprecipitated ERK2 was measured by an in vitro kinase assay using myelin basic protein as a
substrate, where the activation was already detectable after 5 min
stimulation. The observation is well correlated with time course of
hyperphosphorylation described previously (8). Moreover, we observed
similar delayed activation of JNK1 compared with ERK2 in another
BaF3-derived stable clone.2
IL-3 Receptor-mediated JNK/SAPK Activation Is Ras-dependent
Since the Ras protein seems to play an
important role for regulation of JNK/SAPK activity in a variety of
signal transduction pathways, including those initiated by stimulation
with epidermal growth factor, nerve growth factor, and G
protein-coupled receptor ligands (14, 17, 18, 25), we next assessed the
role of Ras in IL-3-stimulated JNK1 activation. As described previously (8), dominant-negative Ras(S17N) was induced in BaF3-N6 cells by
incubation with isopropyl-1-thio--D-galactoside (IPTG)
for 16 h. The IL-3 stimulation of JNK1 activity was found to be
diminished considerably after induction of Ras(S17N), whereas JNK1
activation in response to a translational inhibitor, anisomycin, which
has been shown to be Ras-independent in COS7 cells (14), remained unaffected (Fig. 2A). Likewise, ERK2
stimulation by IL-3 was inhibited by Ras(S17N) (Fig. 2B),
which is consistent with our previous results that Ras(S17N) inhibited
c-Raf-1 activation as well as ERK2 hyperphosphorylation. These results
suggest that both JNK1 and ERK2 pathways are regulated by Ras. Further
examination of the assumption that the JNK1 pathway is dependent on Ras
was performed by using a mutant receptor for GM-CSF. It has been shown
that human GM-CSF is capable of stimulating proliferation of a
BaF3-derived transfectant of
and
c subunits of the
human GM-CSF receptor (26).
544 is a mutant of the
c subunit, which is composed of intact extracellular and
transmembrane domains and a C-terminal truncated cytoplasmic domain (a
membrane-proximal region including box 1 and box 2 motifs) (26). The
544 receptor is incapable of activating the Ras pathway
as determined by a lack of Ras·GTP formation and impaired increase in
ERK2 activity, whereas human GM-CSF can stimulate the Ras pathway to an
extent similar to mouse IL-3 in BaF3/
c cells (Ref. 27
and data not shown). In contrast, the
544 receptor is
considered intact in terms of stimulation of the JAK/signal transducers
and activators of transcription pathway because human GM-CSF-induced
tyrosine phosphorylation of JAK2 was detected also in
BaF3/
544 cells (Fig. 3A,
see also Refs. 28 and 29). Using these transfectants, we evaluated the
role of the Ras pathway in JNK1 activation. As demonstrated in Fig.
3B, the
c receptor was fully active in
promoting the JNK1 pathway upon human GM-CSF stimulation in contrast to
the
544 receptor, which failed to activate JNK1. Mouse
IL-3 could induce JNK1 activation similarly in both cell lines. Taken
together with the above results, it is likely that Ras is indispensable for stimulation of the JNK1 pathway.
Constitutively Active Ras(G12V) Is Insufficient to Induce JNK1 Activation
BaF3-V2 cells grow in an IL-3-dependent
manner, and produce constitutively active Ras(G12V) following IPTG
induction (8). The induced Ras(G12V) can stimulate c-Raf-1 kinase
activity, ERK2 hyperphosphorylation, as well as transcriptional
activation of the c-fos promoter (Ref. 8 and data not
shown). To examine the effect of Ras(G12V) expression on the induction
of JNK1 activation, we first measured JNK1 activation in response to
mouse IL-3 and anisomycin. Both IL-3 and anisomycin induced JNK1
activation to a similar extent as in the case of BaF3-N6 cells (Figs.
2A and 4A). Time course of the
activation also was similar to BaF3-N6 cells as illustrated in Figs.
1D and 4B. Next, the activity of ERK2 and JNK1
was quantitated in parallel, and the results are described in Fig.
4C. No JNK1 activation was observed in response to IPTG
induction of Ras(G12V), while ERK2 activity was markedly augmented in
accordance with Ras(G12V) accumulation. The results suggest that
Ras(G12V) by itself is insufficient for activation of JNK1, where an
additional signal may be required.
It has been demonstrated that a variety of cytokines, including
IL-2, IL-3, IL-5, GM-CSF, and erythropoietin can trigger Ras activation
in hematopoietic cell lines (6, 7). Although the results suggest a
significant role of Ras in cytokine-mediated intracellular signal
transduction, it still remains to be solved whether Ras is implicated
in growth-promoting or differentiation-inducing signaling pathways. In
IL-3 signal transduction, analyses using a series of deletion mutants
of the receptor c subunit have shown that the Ras
pathway is responsible for prevention of apoptotic cell death (30).
Moreover, conditional expression of a dominant-negative Ras mutant
exhibited no inhibitory effect on IL-3-dependent
proliferation of BaF3 cells (8). Likewise, the Ras pathway does not
seem to be necessary for IL-2 induction of cell proliferation (31-33). On the other hand, in another hematopoietic cell line 32Dcl3, Ras was
required for cell growth in response to IL-3, but not for granulocyte
colony-stimulating factor-induced differentiation (34). As for
granulocyte colony-stimulating factor receptor-mediated signaling, the
Ras pathway was found to be associated with the proliferative response
in myeloid leukemia cell lines, such as NFS-60 (35). Taken together,
the role of Ras in cells of the hematopoietic lineage is still
controversial, and thus, further investigation of Ras-regulated
signaling pathways is required for a final conclusion.
In addition to the Ras protein, members of the Cdc42/Rac/Rho GTP-binding protein family have recently been shown to be important for cellular responses, particularly for the regulation of cytoskeletal organization, in various kinds of mammalian cells (36-39). They are also implicated in T-cell polarization (40), cell cycle progression (19), cell motility (41), and morphogenesis of neuronal cells (42). Furthermore, their dominant-negative-type mutants prevent Ras-induced transformation of fibroblast cells, suggesting a significant role at a point downstream of Ras (43, 44). However, the role of Cdc42, Rac, and Rho proteins in cytokine-mediated signal transduction remains to be defined.
In this report, we provide evidence that the JNK1 pathway is activated upon IL-3 stimulation, where Ras may be implicated as a regulator. Furthermore, we have obtained data showing that dominant-negative mSos1 that is capable of interfering with IL-3-induced Ras activation could block JNK1 activation upon IL-3 treatment.2 Collectively, it is conceivable that a Ras-dependent JNK1 pathway exists within BaF3 cells, which may play an important role in IL-3-induced cellular responses.
Although ERK/MAPK and JNK/SAPK pathways converge at the point of transcriptional activation by the AP-1 complex (5), the precise physiological role of each pathway is still obscure. Some growth factor signals preferentially stimulate the ERK/MAPK pathway, while others stimulate only the JNK/SAPK pathway. Growth factors like epidermal growth factor as well as IL-3, as illustrated in this paper, are able to activate both. Additionally, in the case of T-cell activation signaling, for example, two apparently independent signals from the T-cell receptor and the CD28 costimulatory receptor are necessary to augment JNK/SAPK activity, whereas ERK/MAPK is fully activated in response to T-cell receptor stimulation alone (45). In PC12 pheochromocytoma cells, ERK/MAPK is responsible for nerve growth factor-induced differentiation and proliferation, whereas the JNK/SAPK pathway is modulated by a signal of nerve growth factor withdrawal, which causes apoptotic cell death (46). Taking into consideration these findings, it is possible to speculate that JNK/SAPK may be responsible for cellular responses distinct from those regulated by ERK/MAPK in cytokine receptor-mediated signal transduction.
We have shown that IL-3-induced JNK1 activation occurs in a Ras-dependent manner, which is also the case in other types of cells, including PC12, HeLa, NIH 3T3, and COS7 cells (14, 17, 18, 25, 47). However, at present, the mechanism of JNK/SAPK activation through Ras has not been fully understood. It may be possible that MEKK1, which binds directly to a GTP-bound form of Ras (48), is activated by Ras, thereby stimulating a kinase cascade comprising of JNKK1/SEK1/MKK4 and JNK/SAPK (5). Ras and MEKK1-dependent JNK/SAPK activation was, in fact, observed in HeLa cells (25). Instead, Rac may be involved between Ras and the JNK/SAPK cascade because Rac is thought to act as a downstream element of Ras (43, 44), and regulate the JNK/SAPK pathway (17-19). These possibilities are currently under examination.
As shown in Fig. 4, Ras(G12V) is incapable of promoting JNK1 activity by itself in contrast to ERK2, which is strongly activated by Ras(G12V) alone in BaF3-V2 cells (see also Ref. 8). In HeLa and NIH 3T3 cell lines, an activated form of Ras is able to induce JNK/SAPK activation significantly, but less than an activated Cdc42 or Rac protein (18), whereas, in COS7 cells, v-Ras induces only a slight increase of JNK/SAPK activity (17). Another paper has reported that an activated Ras induced only a partial increase in JNK/SAPK activity, which could be further enhanced by irradiation of UV light (47). Furthermore, as described above, stimulation of the T-cell receptor, which is sufficient to induce Ras activation in T-lymphocytes, fails to activate JNK/SAPK, suggesting the activated Ras protein is unable to stimulate JNK/SAPK by itself. Taken together with these observations, a second signal, besides a signal from Ras, may be required for full activation of JNK/SAPK, especially in hematopoietic and lymphoid cells. The elucidation of the biochemical mechanism by which Ras regulates the JNK/SAPK pathway will contribute to our understanding of the role of this pathway in growth and differentiation signaling in hematopoietic cells.
We are grateful to Atsushi Miyajima for BaF3 cell lines expressing the human GM-CSF receptor and Michael Karin for the GST-c-Jun plasmid.