(Received for publication, August 23, 1996, and in revised form, December 24, 1996)
From the Department of Biochemistry, Hiroshima University School of Medicine, 1-2-3 Kasumi, Minami-ku, Hiroshima 734, Japan and the § Department of Virology II, National Institute of Health, 1-23-1 Toyama, Shinjyuku-ku, Tokyo 162, Japan
Ral GDP dissociation stimulator-like (RGL) has been identified to be a possible effector protein of Ras. RGL shares 50% amino acid identity with Ral GDP dissociation stimulator and contains the CDC25-like domain in the central region and the Ras-interacting domain in the C-terminal region. Since the modes of activation and action of RGL have not yet been clarified, in this paper we have analyzed the functions of RGL. In COS cells, RGL interacted with RasG12V/E37G (a Ras mutant in which Gly-12 and Glu-37 were changed to Val and Gly, respectively) which failed to bind to Raf, but not with RasG12V/T35S which bound to Raf. Raf did not inhibit the binding of RGL to RasG12V/E37G under the condition that Raf inhibited that of RGL to RasG12V. Expression of either RGL or Raf into NIH3T3 cells slightly activated c-fos promoter, while coexpression of both proteins greatly stimulated the c-fos promoter activity. RGL stimulated the GDP/GTP exchange of Ral and this action was enhanced by the post-translational modification of Ral. However, RGL was not active on Ras, Rac, CDC42, Rap, or Rho. Furthermore, this action of RGL to stimulate the GDP/GTP exchange of Ral was dependent on Ras in COS cells. These results suggest that RGL constitutes another Ras-signaling pathway which is distinct from the Raf pathway and indicate that the RGL pathway regulates the c-fos promoter activity and the GDP/GTP exchange of Ral.
Ras, a member of the small G protein1
superfamily, is an essential component in the transduction of
extracellular signals that induce proliferation and differentiation (1,
2). Evidence shows that Ras exerts its functions through multiple
effector proteins. These include Raf (3, 4), RalGDS (5-7), RGL (5), phosphatidylinositol 3-OH-kinase (8, 9), Rin1 (10), stress-activated protein kinase (11), MAP kinase kinase kinase (12), protein kinase C
(13), and AF-6 (14). Among these effector proteins, Raf has been
extensively studied. Ras binds to Raf and promotes the activation of
Raf, which in turn activates MAP kinase kinase, which then triggers the
activation of MAP kinase. Activated MAP kinase in turn translocates to
the nucleus, where it phosphorylates and regulates the activities of
nuclear transcription factors which cause changes in gene expression
that controls cell proliferation. Recent biochemical and biological
evidence has implicated a second Ras-mediated signaling pathway which
is distinct from the Raf/MAP kinase pathway and involves Rho family
members, Rho, Rac, and CDC42 (15-19). Furthermore, the experiments
using the effector loop mutants of Ras have suggested that Ras
interacts with diverse effector proteins in the same cells (19-22).
In vitro and yeast two-hybrid studies have shown that
RasG12V/T35S (a Ras mutant in which Gly-12 and Thr-35 are
changed to Val and Ser, respectively) binds to Raf, RalGDS, RGL, and
AF6, that RasG12V/E37G interacts with RalGDS, RGL, yeast
byr2, adenylyl cyclase, and AF6, and that RasG12V/Y40C
associates with AF6. Furthermore, RasG12V/T35S and
RasG12V/Y40C activate Raf and Rac, respectively, and these
Ras effector loop mutants cooperate to promote DNA synthesis in
mammalian cells. Consistent with these results, coexpression of Raf and
Rac induces transformation synergistically (15). Therefore, the Ras
effector pathways could cooperate to exert the functions of Ras.
We have identified RGL as a Ras-interacting protein (5). RGL shares 50% amino acid identity with RalGDS (5, 23). RGL contains the Ras-interacting domain in the C-terminal region, named RID. RID binds to an active form of Rap, which is a member of the small G protein superfamily and possesses the same amino acid sequence as the effector loop of Ras (24). Expression of RID into Ras- but not Raf-transformed NIH3T3 cells reverses the malignant phenotypes such as growth in low serum condition and anchorage independence (25), and injection of RID into Xenopus oocytes inhibits Ras-dependent maturation and MAP kinase activation (26). Thus, RID has been well characterized and known to interact with the effector loop of an active form of Ras. Therefore, RGL could mediate some of Ras functions. The amino acid sequence of RGL shows that it contains the CDC25-like domain. Several GDP/GTP exchange proteins including CDC25, SOS, SCD25, Ste6, BUD5, LTE1, C3G, and RalGDS share the CDC25-like domain (23, 27, 28). This domain of CDC25 and SOS stimulates the GDP/GTP exchange of Ras and induces transformation in NIH3T3 cells (29-31). C3G and RalGDS stimulate the GDP/GTP exchange of Rap and Ral, respectively (23, 32). However, the functions of RGL are not known since the studies using the full-length of RGL have not yet been done.
In this paper we have examined the roles of RGL in the Ras-signaling pathway. We show that RGL constitutes the Ras-mediated signaling pathway which is distinct from the Raf pathway and that it cooperates with Raf to activate c-fos promoter. Furthermore, we demonstrate that RGL stimulates the GDP/GTP exchange of Ral, that the post-translational modification of Ral is important for this action of RGL, and that Ras regulates the RGL activity. These results indicate that RGL mediates the Ras signals to stimulate the gene expression and activate Ral.
pBJ-1, pCGN, pfos-luc (a
reporter vector in which the expression of luciferase is controlled by
c-fos promoter) (33), the anti-HA and Myc antibodies, and
wild type NIH3T3 cells were provided by Drs. Q. Hu, A. Klippel, and H. Cen (Chiron Corp., Emeryville, CA). pmt/RafCAAX and pmt/RafSAAX,
pEXV/RacG12V, the anti-Ras antibody (Y13-259), pEF-Bos,
pcDNAIAmp/RasG12V and
pcDNAIAmp/RasG12V/T35S, and pGEX-2T/Rac and
pEF-Bos/CDC42 were generous gifts from Drs. C. Marshall (Institute of
Cancer Research, London), A. Hall (University College London, London),
J. Downward (Imperial Cancer Research Fund, London), S. Nagata (Osaka
University, Suita, Japan) (34), T. Kataoka (Kobe University, Kobe,
Japan) (21), and K. Kaibuchi (Nara Institute of Science and Technology,
Ikoma, Japan), respectively. Baculoviruses producing
RasG12V, RasS17N, and GST-fused to small G
proteins were generated as described (35, 36). Sf9 cells, pBlueBacHisB,
and Baculogold linearized baculovirus DNA were purchased from
Pharmingen (San Diego, CA). All procedures of passage, infection, and
transfection of Sf9 cells and the isolation of recombinant
baculoviruses were carried out as described (37). The anti-Ras
antibodies (Y13-238 for immunoprecipitation assay, F235 for immunoblot
analysis) were from Oncogene Science, Inc. (Manhasset, NY) and Santa
Cruz Biotechnology, Inc. (Santa Cruz, CA). [3H]GDP and
[35S]GTPS were from Amersham (Buckinghamshire, United
Kingdom).
RGL tagged with the HA epitope at the
N terminus was synthesized by PCR to have BamHI and
NcoI sites. This fragment was digested with BamHI
and NcoI, and the NcoI site was blunted with
Klenow fragment. The fragment was inserted into pBlueBacHisB
which was digested with HindIII, blunted with
Klenow fragment, and digested with BamHI to
generate pBlueBacHisB/RGL. pGEX-2T/Rac and pEF-Bos/CDC42 were digested
with BamHI and the fragments were inserted into the
BamHI cut pUC19 to generate pUC19/Rac and pUC19/CDC42. To construct pVIKS/Rac and pVIKS/CDC42, pUC19/Rac and pUC19/CDC42 were
digested with XbaI and EcoRI. The fragments were
inserted into the XbaI and EcoRI cut pVIKS. To
construct pBJ-Myc, pBJ-1 was digested with XbaI, blunted
with Klenow fragment, and self-ligated to generate
pBJ-1XbaI. The fragment encoding the Myc
epitope with EcoRI and BamHI sites was
synthesized by PCR, digested with EcoRI and
BamHI, and inserted into the EcoRI and
BamHI cut pBJ-1
XbaI to generate
pBJ-Myc. To construct pGEX-2T/RGL and pBJ-Myc/RGL, the fragment
encoding RGL with BamHI sites was synthesized by PCR. The
fragment was digested with BamHI and inserted into the BamHI cut pGEX-2T and BamHI cut pBJ-Myc. To
construct pGEX-2T/RGL-(1-505) (amino acids 1-505) and
pGEX-2T/RGL-(88-505), the fragments encoding RGL-(1-505) and
RGL-(88-505) with BamHI and EcoRI sites were
synthesized by PCR. These fragments were digested with BamHI
and EcoRI and inserted into the BamHI and
EcoRI cut pGEX-2T. pGEX-2T/RGL-(1-505) was digested with
BamHI and EcoRI and inserted into the
BamHI and EcoRI cut pBSKS to generate
pBSKS/RGL-(1-505). pBSKS/RGL-(1-505) was digested with
EcoRI, blunted with Klenow fragment, and digested with XbaI. The fragment was inserted into pBJ-Myc, which was
digested with BamHI, blunted with Klenow
fragment, and digested with XbaI to generate
pBJ-Myc/RGL-(1-505). To construct pBJ-myc/RGL-(88-505), the fragment
encoding RGL-(88-505) with XbaI and EcoRI sites
was synthesized by PCR and digested with EcoRI, blunted with
Klenow fragment, and digested with XbaI. This
fragment was inserted into pBJ-Myc, which was digested with
BamHI, blunted with Klenow fragment, and digested
with XbaI. To construct pCGN/RGL, pBJ-Myc/RGL was digested
with BamHI, and the fragment was inserted into the
BamHI cut pBSKS to generate pBSKS/RGL. pBSKS/RGL was
digested with XhoI, blunted with Klenow fragment,
and digested with XbaI. The fragment was inserted into the
XbaI and SmaI cut pCGN. To construct
pGEX-2T/RhoA, the fragment encoding RhoA with BamHI sites
was synthesized by PCR. The fragment was digested with BamHI
and inserted into the BamHI cut pGEX-2T. To construct
pBJ/ RasG12V and pBJ/RasG12V/T35S,
pcDNAIAmp/RasG12V and
pcDNAIAmp/RasG12V/T35S were digested with
XbaI and BamHI then blunted with
Klenow fragment. The fragments were inserted into the
EcoRV cut pBJ-1. To synthesize RasG12V/E37G and
RasG12V/Y40C, pcDNAIAmp/RasG12V was
digested with XbaI and BamHI, then blunted with
Klenow fragment. The fragment was inserted into the
SmaI cut pKF-19, and site-directed mutagenesis was performed
using the Mutan-Express Km system (Takara Shuzo Co., Ltd., Ohtsu,
Japan). To construct pEF-Bos/RasG12V/E37G and
pEF-Bos/RasG12V/Y40C, pKF-19/RasG12V/E37G and
pKF-19/RasG12V/Y40C were digested with KpnI and
XbaI and blunted with T4 DNA polymerase. The fragments were
inserted into pEF-Bos, which was digested with XbaI and
blunted with Klenow fragment. To construct
pGEX-2T/RasG12V and pGEX-2T/RasG12V/T35S,
pcDNAIAmp/RasG12V and
pcDNAIAmp/RasG12V/T35S were digested with
XbaI, blunted with Klenow fragment, and digested with BamHI. The fragments were inserted into pGEX-2T, which
was digested with EcoRI and blunted with Klenow
fragment, then digested with BamHI. To construct
pGEX-2T/RasG12V/E37G and pGEX-2T/RasG12V/Y40C,
pKF-19/RasG12V/E37G and pKF-19/RasG12V/Y40C
were digested with KpnI and XbaI and blunted with
T4 DNA polymerase, and the fragments were inserted into the
SmaI cut pGEX-2T. To construct pCGN/RalB, the fragment
encoding RalB with XbaI and BamHI sites was
synthesized by PCR and digested with XbaI and BamHI, then inserted into the XbaI and
BamHI cut pCGN. pGEX-2T/Rap1, pGEX-2T/Raf-(1-149), and
pMAL-c2/RGL-(602-768) were constructed as described
previously (24, 38).
The post-translationally modified and
unmodified forms of GST-fused to Ral (GST-Ral), GST-Ras, GST-Rac, and
GST-CDC42 were purified from the membrane and cytosol fractions of Sf9
cells as described (36). GST-RGL, GST-RGL-(1-505), GST-RGL-(88-505), GST-Raf-(1-149), and MBP-fused to RGL-(602-768) (MBP-RGL-(602-768)) were produced in Escherichia coli. Briefly, the transformed
E. coli were initially grown at 37 °C to an absorbance of
0.5 (optical density at 600 nm) and subsequently transferred to
20 °C, then isopropyl-1--D-galactopyranoside was
added at a final concentration of 0.1 or 0.3 mM and further
incubation was carried out for 16 h at 20 °C. The expressed
proteins were purified by affinity chromatographies in accordance with
the manufacturer's instructions. GST-Rap1 and GST-RhoA were produced
in E. coli and purified as described (24).
Monolayers
of Sf9 cells (2 × 107 cells on a 15-cm diameter
plate) were infected with recombinant baculoviruses expressing RGL and
Ras (35, 39). After 72 h, the cells were lysed as described (5,
35) and the proteins of the lysates (240 µg of protein) were
immunoprecipitated with the anti-Ras antibody. The immunoprecipitates were subjected to SDS-polyacrylamide gel electrophoresis (40), transferred to nitrocellulose filters, and probed with the anti-HA or
Ras (F235) antibody. COS cells (60-70% confluent on a 60-mm diameter
plate) were transfected with pBJ and pEF-Bos derived constructs
described above using the DEAE-dextran method (41). After 60 h,
the cells were lysed as described (38) and proteins of the lysates (160 µg of protein) were immunoprecipitated with the anti-Ras antibody.
Y13-238 was used in the immunoprecipitation experiments except for
Fig. 1B (lane 7), where Y13-259 was used. The
precipitates were probed with the anti-Myc or Ras (F235) antibody.
Interaction Assay of RGL and Ras Effector Loop Mutants in Vitro
To make immobilized RGL-(602-768) on amylose resin, 130 µg of MBP-RGL-(602-768) was incubated with 200 µl of amylose resin in 500 µl of reaction mixture (20 mM Tris-HCl, pH 7.5, 1 mM DTT, and 200 mM NaCl) for 2 h at
4 °C. The resin was precipitated by centrifugation and washed with
10 mM Tris-HCl (pH 7.5) three times. To make the
[35S]GTPS-bound form of Ras effector loop mutants,
each of Ras mutants (25 pmol) was incubated with 10 µM
[35S]GTP
S (20,000 cpm/pmol) for 15 min at 30 °C in
40 µl of reaction mixture (50 mM Tris-HCl, pH 7.5, 10 mM EDTA, 5 mM MgCl2, and 1 mM DTT). After the incubation, 1 M
MgCl2 was added at a final concentration of 20 mM. The [35S]GTP
S-bound form of Ras
mutants were incubated for 30 min at 4 °C with immobilized
RGL-(602-768) (50 pmol) in 140 µl of reaction mixture (50 mM Tris-HCl, pH 7.5, 10 mM EDTA, 20 mM MgCl2, 1 mM DTT, and 10 µM GTP
S) in the presence or absence of
GST-Raf-(1-149). Immobilized RGL-(602-768) were precipitated by
centrifugation, the precipitated resin was washed, and the remaining
radioactivities were counted.
NIH3T3 cells (40-50% confluent on a 35-mm diameter plate) were transfected with the plasmid DNA (2 µg) containing pfos-luc (1 µg) and the indicated plasmids using LipofectAMINE according to the manufacturer's instructions (Life Technologies, Inc., Gaithersburg, MD). After 60 h, the cells were lysed and luciferase activity was measured using the PicaGene luciferase assay system (Toyo B-NET. Co., Ltd., Tokyo, Japan) according to the manufacturer's instructions. Luciferase activity was expressed as the fold increase compared with the level observed with empty vector. To examine the transfection efficiency, pGL2 (a reporter vector in which the expression of luciferase is controlled by SV40 promoter) was cotransfected instead of pfos-luc as a control in a separate set of experiments.
GDP/GTP Exchange Assay of RGL in VitroTo measure the
activity of RGL to stimulate the binding of GTPS to small G
proteins, the post-translationally modified or unmodified form of
GST-fused to small G proteins (100 nM) were incubated with
2 µM [35S]GTP
S (1,000-1,500 cpm/pmol)
for 7 or 15 min at 30 °C in 50 µl of reaction mixture (50 mM Tris-HCl, pH 7.5, 10 mM MgCl2, 5 mM EDTA, 0.12% CHAPS, 1 mM
DL-
-dimyristoylphosphatidylcholine, 1 mM
DTT, and 1 mg/ml bovine serum albumin) containing the indicated concentrations of GST-RGL and its deletion mutants. To measure the
activity of RGL to stimulate the dissociation of GDP from Ral, the
post-translationally modified or unmodified form of GST-Ral (5 pmol)
was preincubated with 2 µM [3H]GDP
(1,500-3,000 dpm/pmol) for 15 min at 30 °C in 20 µl of reaction mixture (50 mM Tris-HCl, pH 7.5, 5 mM
MgCl2, 10 mM EDTA, 0.3% CHAPS, 1 mM DL-
-dimyristoylphosphatidylcholine, 1 mM DTT, and 1 mg/ml bovine serum albumin). After the
incubation, 1 µl of 400 mM MgCl2 was added.
To this preincubation mixture, 29 µl of reaction mixture (50 mM Tris-HCl, pH 7.5, 170 µM GTP, and 1 mg/ml
bovine serum albumin) containing GST-RGL (15 pmol) was added and the mixture was further incubated for the indicated periods of time at
30 °C. Assays were quantified by rapid filtration on nitrocellulose filters (23, 36, 38).
COS cells (60-70% confluent on a 35-mm diameter plate) were transfected with pCGN, pBJ, and pEF-Bos derived constructs described above using the DEAE-dextran method (41). After 48 h, the cells were metabolically labeled with 32Pi (0.2 mCi/ml) in phosphate-free RPMI 1640 for 12 h. Then the cells were lysed as described (38) and Ral was immunoprecipitated with the anti-HA antibody. After extensive washing of the immunoprecipitates, bound guanine nucleotides were eluted, separated by ascending thin-layer chromatography (42, 43), and analyzed with Fuji BAS 2000 image analyzer. The relative molar ratio of GTP and GDP was corrected for the number of phosphates per mol of guanosine, and the amount of the GTP bound form of Ral was expressed as percentage of the total amount of Ral.
Other AssaysProtein concentrations were determined with bovine serum albumin as a standard (44).
RGL contains the CDC25-like domain which consists of 6 blocks in its central region and RID in the C-terminal region (5, 23) (Fig. 1A). Since experiments using the full-length of RGL have not yet been done, we coexpressed RGL with Ras in Sf9 cells to examine whether the full-length of RGL interacts with Ras in intact cells. RGL was tagged with the HA epitope at the N terminus. The expression level of transfected RGL in Sf9 cells expressing RGL alone was similar to that in the cells coexpressing RGL with RasG12V, an active form, or RasS17N, a dominant negative form, as assessed by immunoblotting (Fig. 1B, lanes 1-3). When the lysates coexpressing RGL with RasG12V were immunoprecipitated with the anti-Ras antibody, both RGL and Ras were detected in the Ras immune complex (Fig. 1B, lane 5). When the lysates expressing RGL alone were immunoprecipitated with the anti-Ras antibody, RGL was not detected (Fig. 1B, lane 4). When the lysates coexpressing RGL with RasS17N were immunoprecipitated with the anti-Ras antibody, RGL was not coprecipitated with RasS17N (Fig. 1B, lane 6). Neither RGL nor RasG12V was immunoprecipitated with non-immune immunoglobulin from the lysates expressing both proteins (data not shown). We used Y13-238 as the anti-Ras antibody to immunoprecipitate Ras for these experiments. Another anti-Ras antibody, Y13-259 which is known to be the neutralizing antibody (45), was tested for its ability to immunoprecipitate the Ras-RGL complex. In contrast to Y13-238, Y13-259 could not immunoprecipitate the Ras-RGL complex from the cell lysates coexpressing RGL with RasG12V under the same condition (Fig. 1B, lane 7), although Y13-259 and Y13-238 immunoprecipitated the similar amounts of Ras (Fig. 1B, lanes 5 and 7). These results indicate that the full-length of RGL interacts with an active form of Ras through its effector loop in intact cells.
RID of RGL is known to interact with the GTP-bound form of Ras through its effector loop (5, 25, 26). To examine whether RID is the only domain which is required for the association with Ras, two deletion mutants, RGL-(1-505) and RGL-(88-505), were generated (Fig. 1A). RGL-(88-505) encodes only the CDC25-like domain. RasG12V was coexpressed with RGL, RGL-(1-505), or RGL-(88-505) in COS cells (Fig. 1C, lanes 1-3). RGL and its deletion mutants were tagged with the Myc epitope at the N terminus. Although RGL was coprecipitated with RasG12V, either RGL-(1-505) or RGL-(88-505) was not (Fig. 1C, lanes 4-6). Taken together with the previous observations (5, 25, 26), RID is sufficient and necessary for the association of RGL with Ras.
Interaction of RGL with Ras Effector Loop MutantsYeast
two-hybrid experiments have shown that RGL binds to
RasG12V/T35S and RasG12V/E37G but not to
RasG12V/Y40C (19). To examine whether RGL interacts with
these Ras effector loop mutants in mammalian cells, RGL was coexpressed
with the Ras effector loop mutants in COS cells (Fig.
2A). RGL was coprecipitated with
RasG12V and RasG12V/E37G but not with either
RasG12V/T35S or RasG12V/Y40C (Fig.
2B, lanes 2-5). In COS cells without ectopically
expressed Ras, RGL was not precipitated with the anti-Ras antibody
(Fig. 2B, lane 1). These results were not
partially consistent with the observations obtained from the yeast
two-hybrid experiments. The specific binding of RGL to
RasG12V/E37G was confirmed by in vitro
experiments (Fig. 3). RGL-(602-768), the
Ras-interacting domain of RGL, was purified as MBP-fused to protein
from E. coli. Raf-(1-149), RasG12V,
RasG12V/T35S, RasG12V/E37G, and
RasG12V/Y40C were purified as GST-fused to proteins from
E. coli. These Ras mutants showed the same GTPS binding
activity and the Kd of GTP
S for these Ras mutants
were almost similar (data not shown). RGL-(602-768) bound to the
GTP
S-bound form of RasG12V and RasG12V/E37G,
although the binding activity of RGL-(602-768) to
RasG12V/E37G was weaker than to RasG12V (Fig.
3A). However, RGL-(602-768) interacted with neither the GTP
S-bound form of RasG12V/T35S nor
RasG12V/Y40C (Fig. 3A). Raf-(1-149) which
contains the Ras-binding domain inhibited the binding of RGL-(602-768)
to RasG12V in a dose-dependent manner, while
Raf-(1-149) was unable to displace RGL-(602-768) for binding to
RasG12V/E37G (Fig. 3B). These results clearly
indicate that the binding manner of RGL to Ras is different from that
of Raf to Ras and suggest that RGL constitutes another Ras-signaling
pathway which is distinct from the Raf pathway.
Synergistic Activation of c-fos Promoter by RGL and Raf
It
has been reported that the Raf/MAP kinase and Rac pathways synergize to
induce focus formation in Rat1 cells and DNA synthesis in REF52 cells
(15, 22). Since the Raf and Rac pathways are important for gene
expression (46, 47), we examined whether the RGL pathway cooperates
with other Ras effector pathways to activate c-fos promoter
using c-fos luciferase. RafCAAX is an active Raf kinase
targeted to the plasma membranes by virtue of the addition of a
C-terminal membrane localization signal from Ki-Ras (48, 49). RafSAAX,
in which the cysteine residue of CAAX box is changed to serine to
prevent farnesylation, is a control protein (48). Transfection of
RasG12V stimulated the c-fos luciferase
expression. Transfection of either RGL or RafCAAX alone slightly
stimulated the c-fos luciferase expression, while
cotransfection of RafCAAX and RGL greatly stimulated the expression
(Fig. 4). RafSAAX did not stimulate the c-fos
luciferase expression, and the expression level induced by
cotransfection of RafSAAX and RGL was the same as that induced by
transfection of RGL alone. Cotransfection of RGL and
RacG12V, an active form of Rac, synergized to stimulate the
c-fos luciferase expression, but the extent was weaker than
that of RGL and RafCAAX (Fig. 4). These results indicate that the RGL
pathway activates c-fos promoter synergistically with the
Raf pathway.
RGL Activity to Regulate the GDP/GTP Exchange of Small G Proteins
Since RGL contains the domain highly homologous with
CDC25, we examined whether RGL regulates the GDP/GTP exchange of small G proteins. Furthermore, we also investigated the effect of
post-translational modification of small G proteins on the RGL activity
to regulate the GDP/GTP exchange since it has been known that the
GDP/GTP exchange regulatory proteins such as CDC25, SOS, smgGDS,
RabGDI, and RhoGDI act more effectively on the post-translationally
modified form of small G proteins than the unmodified form (30, 31, 39,
50). The post-translationally modified and unmodified forms of Ras,
Ral, Rac, and CDC42 were purified from the membrane and cytosol
fractions of Sf9 cells, respectively, as GST fused to proteins. RGL did
not stimulate the binding of GTPS to the post-translationally
modified or unmodified form of Ras (Fig. 5). RGL was not
active on the post-translationally modified or unmodified form of Rac
or CDC42, either (Fig. 5). However, RGL stimulated the binding of
GTP
S to both the post-translationally modified and unmodified forms
of Ral (Fig. 5). RGL dissociated GDP from the modified form of Ral more
effectively than the unmodified form (Fig.
6A). RGL stimulated the binding of GTP
S to
Ral in a dose-dependent manner and the action was more
effective on the modified form than the unmodified form (Fig.
6B). When the binding of GTP
S to Ral was assayed in the
presence of RGL (40 nM) using various amounts of Ral, the
apparent Km value of RGL for the modified form was
estimated to be 380 nM. That of RGL for the unmodified form
was calculated to be more than 4 µM (data not shown). As
shown in Fig. 6B, a large amount of RGL stimulated the
GDP/GTP exchange of the post-translationally unmodified form of Ral.
Under the same conditions, RGL was not active on the
post-translationally unmodified form of either Rap or Rho (data not
shown). These results indicate that RGL regulates the GDP/GTP exchange
of Ral but not that of Ras, Rap, Rac, CDC42, or Rho. This
characteristic of RGL is similar to that of RalGDS (23). To determine
the region which is responsible for the GDP/GTP exchange activity, we
examined the actions of RGL-(1-505) and RGL-(88-505) on Ral.
RGL-(1-505) showed almost the same activity as RGL on Ral and the
action of RGL-(1-505) to stimulate the binding of GTP
S to Ral was
enhanced by the post-translational modification of Ral (Fig.
7). However, RGL-(88-505) did not stimulate the binding
of GTP
S to either the modified or unmodified form of Ral (Fig. 7).
These results indicate that RID of RGL is not necessary for the
activity to stimulate the GDP/GTP exchange of Ral but that the
CDC25-like domain alone is not sufficient for the activity.
Effect of Ras on the RGL Activity to Stimulate the GDP/GTP Exchange of Ral
Finally we examined whether the association of Ras with
RGL regulates the RGL activity in intact cells. Ral, which was tagged with the HA epitope at the N terminus, was coexpressed with RGL and Ras
in COS cells. The expression level of Ral was not changed by the
coexpression of RGL, RasG12V, or RasG12V/Y40C
(Fig. 8A, lanes 1-4). The cells
were labeled with 32Pi and Ral was
immunoprecipitated with the anti-HA antibody. The proportion of GTP
bound to Ral was quantified by the thin-layer chromatography method
(42, 43) (Fig. 8B). When Ral was expressed alone, the
percentage of the GTP-bound form of Ral was about 4%. Coexpression of
RGL increased the proportion of GTP bound to Ral from 4 to 12%.
Furthermore, RasG12V enhanced the RGL-dependent
increase of the proportion of GTP bound to Ral to 18%, but not did
RasG12V/Y40C which failed to bind to RGL. To confirm that
the association of Ras with RGL regulates the RGL activity,
RGL-(1-505) which lacks the Ras-interacting domain was studied. Ral
and RGL-(1-505) were expressed with or without RasG12V
(Fig. 8A, lanes 5 and 6). RGL-(1-505) increased
the percentage of the GTP-bound form of Ral, but its activity did not
raise by coexpression with RasG12V. These results indicate
that the association of Ras with RGL enhanced the GDP/GTP exchange
activity of RGL in intact cells.
Previously we isolated cDNA of RGL and found that the C-terminal region of RGL interacts with Ras (5, 24-26). However, the functions of RGL have not yet been clear. Using the full-length of RGL, we have clarified that RGL binds to a Ras effector loop mutant which activates neither the Raf nor Rac pathway, that it cooperates with Raf to activate c-fos promoter, and that it stimulates the GDP/GTP exchange of Ral in a Ras-dependent manner. These results are the first demonstration showing that RGL constitutes another Ras-signaling pathway which is distinct from the Raf pathway and that the signal from RGL leads to the gene expression and Ral activation. These findings provide new insight into the functional significance of Ras and RGL interaction and support the importance of RGL as a real effector protein of Ras.
Evidence has been accumulated that Ras transmits the signal to at least two pathways, the Raf and Rac pathways (15-19, 22). It has been also demonstrated that RasG12V/T35S and RasG12V/Y40C activate Raf and Rac, respectively, and that these Ras effector loop mutants cooperate to promote DNA synthesis in REF52 cells (22). Our results have clearly shown that RGL specifically binds to RasG12V/E37G, but not to either RasG12V/T35S or RasG12V/Y40C. It has been reported that RGL interacts with both RasG12V/T35S and RasG12V/E37G in yeast two-hybrid experiments (19). Although we do not know the reasons for the discrepancy between our and other's results, it might be due to the difference of the assay methods. Furthermore, we have demonstrated that RGL activates c-fos promoter synergistically with Raf. These results suggest that the RGL pathway is a Ras-signaling pathway which is different from the Raf and Rac pathways. The Raf and Rac pathways have been shown to increase c-fos promoter activity through ternary complex factor and serum response factor, respectively (47). Furthermore, it has been demonstrated that the signal through serum response factor acts synergistically with the signal through ternary complex factor which is activated by MAP kinase (46, 47). Therefore, RGL could activate transcription factors other than ternary complex factor. It has been reported that RalGDS induces focus formation in NIH3T3 cells synergistically with Raf (51). Therefore, RalGDS family members could act downstream of Ras and regulate cell functions cooperatively with the Raf pathway.
We have also found that RGL stimulates the GDP/GTP exchange of Ral as well as RalGDS. This activity of RGL is specific for Ral as far as we have examined. Although RGL is homologous with RalGDS, about 100 amino acids in the region between blocks 1 and 2 of the CDC25-like domain are lacking (5). Therefore, these 100 amino acids are not essential for the RGL activity to stimulate the GDP/GTP exchange of Ral. CDC25 and SOS share the CDC25-like domain containing blocks 1 to 6 (23). The region consisting of blocks 1 to 6 of CDC25 and SOS is sufficient for stimulating the GDP/GTP exchange of Ras (30, 31). However, we have found that RGL-(1-505) but not RGL-(88-505) has the activity to stimulate the GDP/GTP exchange of Ral although RGL-(88-505) covers the CDC25-like domain. These results indicate that the CDC25-like domain of RGL is not sufficient for its activity to stimulate the GDP/GTP exchange of Ral and that the region of RGL-(1-87) may be necessary to keep RGL in a structure to regulate the GDP/GTP exchange of Ral. The mode of regulation in the GDP/GTP exchange of Ral by RGL could be different from that of Ras by CDC25 and SOS. We have shown that RID of RGL is necessary and sufficient for the association with Ras but that it is not directly required for the action to stimulate the GDP/GTP exchange of Ral in vitro. We have also demonstrated that RasG12V but not RasG12V/Y40C stimulates RGL-mediated GDP/GTP exchange of Ral in COS cells and that the activity of RGL-(1-505) is not enhanced by RasG12V. These results indicate that Ras-induced RGL activation is dependent upon the two proteins binding and that Ras mediates the redistribution of RGL to Ral which is present on the membranes. This characteristic of RGL is similar to that of RalGDS (43). Indeed RalGDS translocates from the cytosol to the membranes to bind to Ras which is present on the membranes (36). Therefore, it is possible that RID of RGL is not important for the catalytic activity itself but essential for the determination of subcellular localization. However, we cannot exclude the possibility of the existence of a factor on the membranes which activates RGL.
Small G proteins undergo a series of post-translational modification which regulates their localization to the membrane fractions such as plasma membrane, endoplasmic reticulum, and Golgi apparatus, and which is essential for their biological actions (2, 50, 52). It has been demonstrated that the post-translational modifications of small G proteins are important for the actions of their GDP/GTP exchange proteins (30, 31, 39, 50). Consistent with previous observations, RGL stimulates the GDP/GTP exchange of the modified form of Ral more effectively than that of the unmodified form. Since the Km value of RGL for the modified form of Ral is smaller than that for the unmodified form, the post-translational modification of Ral could be important for not only its binding to the membranes but also its affinity for RGL. We have demonstrated that RalGDS is more active on the post-translationally modified form of Ral than the unmodified form (36). Therefore, the post-translational modification of Ral is important for the action of RalGDS family (RalGDS and RGL) to stimulate the GDP/GTP exchange of Ral.
Although it is clear that Ral is activated by RalGDS and RGL, the functions of Ral have long remained elusive. It has been shown that Ral is required for Src- and Ras-dependent activation of phospholipase D (53). Furthermore, it has been reported that Ral is involved in Ras-dependent transformation in NIH3T3 cells and that this small G protein regulates the initiation of border cell migration in Drosophila oogenesis (43, 54). Thus, evidence has been accumulated that Ral is an important small G protein in the signal transduction system. Rlf has been recently identified as a novel Ras- and Rap1-associating protein (55). Rlf contains the CDC25-like domain and RID and it shares 34 and 24% amino acid identity with RalGDS and RGL, respectively. Although the functions of Rlf are not determined, it could have a GDP/GTP exchange activity of Ral from the structural similarity with RalGDS and RGL. The reasons of the existence of three GDP/GTP exchange proteins for Ral are not known at present. The difference of the modes of action and activation of RalGDS family remains to be clarified. Further studies are necessary to understand the whole picture of the Ras-signaling pathway through RalGDS family.
We are grateful to Drs. Q. Hu, A. Klippel, H. Cen, C. Marshall, A. Hall, J. Downward, S. Nagata, T. Kataoka, and K. Kaibuchi for the plasmids and thank the Research Center for Molecular Medicine, Hiroshima University School of Medicine, for the use of their facilities.