Post-translational modifications of Ras and Ral are important for the action of Ral GDP dissociation stimulator.

Ral GDP dissociation stimulator (RalGDS) is a GDP/GTP exchange protein of Ral and a new effector protein of Ras. Therefore, there may be a new signaling pathway from Ras to Ral. In this paper, we examined the roles of the post-translational modifications of Ras and Ral on this new signal transduction pathway. The post-translationally modified form of Ras bound to RalGDS more effectively than the unmodified form. The modification of Ras was required to regulate the distribution of RalGDS between the cytosol and membrane fractions in COS cells. The post-translational modification of Ral enhanced the activities of RalGDS to stimulate the dissociation of GDP from and the binding of GTP to Ral. Furthermore, the modified form of Ral bound to Ral-binding protein 1 (RalBP1), a putative effector protein of Ral, more effectively than the unmodified form. Taken together with the observations that Ras and Ral are localized to the membranes, these results suggest that the post-translational modifications of Ras and Ral play a role for transmitting the signal effectively on the membranes in the signal transduction pathway of Ras/RalGDS/Ral/RalBP1.

and this second cysteine is palmitoylated. Ki-Ras does not have the second cysteine, but it has a polybasic region on the Nterminal side of the farnesylated cysteine instead of the second cysteine. Among these post-translational modifications, farnesylation of Ras is most important for its transformation and membrane binding activities. Furthermore, it has been shown that the post-translational modification of Ki-Ras is required for the activation of yeast adenylate cyclase and mitogen-activated protein kinase (26,27) and that those of Ha-Ras and Ki-Ras enhance the actions of smgGDS, mCDC25, and hSOS, which are GDP/GTP exchange proteins of Ras (28 -30). We also found that the post-translational modification of Ha-Ras is essential for the Raf activation (31). However, the role of the post-translational modification of Ras on the RalGDS action is not known. Ral has a CAAL motif at the C-terminal (L is leucine) and has been found to be geranylgeranylated (32). The observation that Ral is localized to the membranes suggests that Ral binds to the membranes through its post-translational modification (1,4,5,8). However, other roles of the posttranslational modification of Ral are not clear.
As described above, the signal from Ras to Ral could regulate various cell functions. In this paper we examined the roles of the post-translational modifications of Ras and Ral on this new signaling pathway. We show that the post-translational modifications of Ras and Ral enhance their binding activities to the effector proteins, RalGDS and RalBP1, that the modification of Ras is necessary for the localization of RalGDS on the membranes in intact cells, and that the modification of Ral is important for the RalGDS action but not for the RalGAP action. Furthermore, we show the actions of RalGDS and RalGAP for various Ral mutants.

EXPERIMENTAL PROCEDURES
Materials and Chemicals-The RalGDSb and RalB cDNAs were provided by Dr. R. Weinberg (Whitehead Institute for Biomedical Research) (10). The c-Ha-Ras and RalA G23V cDNAs were provided by Drs. J. Downward (Imperial Cancer Research Institute, London, United Kingdom) and P. Chardin (Institut de Pharmacologie Moleculaire et Cellulaire, Valbonne, France), respectively (33). Mammalian expression vectors, pBJ-1 and pCGN, and the mouse anti-influenza virus HA1 monoclonal antibody 12CA5 were provided by Dr. Q. Hu (University of California San Francisco) (34). pCGN was designed to express a 16amino acid epitope from influenza virus HA fused to protein. pV-IKS were provided by Dr. D. Mirda (University of California San Francisco) (31). pV-IKS was designed to express GST and HA fused to proteins in Sf9 cells. All procedures of passage, infection, and transfection of Sf9 cells and the isolation of baculoviruses were carried out as described (35). Ras G12V (in which valine was substituted for glycine at position 12) and Ras G12V⌬ (in which the C-terminal four amino acids of Ras G12V were deleted) were synthesized as described (15,31). RalB S28N , RalB T46A , and RalB G86E were synthesized by PCR (36). The RalBP1 cDNA was isolated by reverse transcriptase PCR using rat brain mRNA as a template. pMALc-2 was purchased from New England Biolabs (Bevery, MA). COS cells were obtained from the American Type Culture Collection. The anti-Ras antibodies, Y13-238 for immunoprecipitation assay and F235 for immunoblot analysis, were from Oncogene Science Inc. Plasmid Constructions-pGEX/RalB, pCGN/RalGDS, and pBJ/ Ras G12V were constructed as described (15). To construct pMAL/ RalGDS-(764 -864), the 0.3-kb fragment encoding RalGDS-(764 -864) with BamHI sites at 5Ј-and 3Ј-ends was synthesized by PCR. The fragment was digested with BamHI and inserted into the BamHI cut pMALc-2 to generate pMAL/RalGDS-(764 -864). To construct pMAL/ RalBP1-(364 -647), the full length of RalBP1 cDNA with BamHI sites synthesized by reverse transcriptase PCR was digested with BamHI and inserted into the BamHI cut pUC19. The RalBP1-(364 -647) fragment was digested with PmaCI and HindIII and then inserted into the SmaI and HindIII cut pBSKS to generate pBSKS/RalBP1-(364 -647). pBSKS/RalBP1-(364 -647) was digested with BamHI and inserted into the BamHI cut pMALc-2 to generate pMAL/RalBP1-(364 -647). To construct pGEX-2T encoding RalGDS, pGAD/RalGDS (11) was digested with BamHI and EcoRI and then blunted with Klenow. The 2.7-kb fragment encoding RalGDS was inserted into pGEX-2T which was digested with EcoRI and then blunted with Klenow. To construct pV-IKS/Ras, the 0.6-kb fragment encoding Ras with EcoRI and PstI sites was synthesized by PCR. This fragment was digested with EcoRI and PstI and inserted into the EcoRI and PstI cut pV-IKS to generate pV-IKS/Ras. To construct pV-IKS/RalB, the 0.6-kb fragment encoding RalB with XbaI and PstI sites was synthesized by PCR. This fragment was digested with XbaI and PstI and inserted into the XbaI and PstI cut pV-IKS to generate pV-IKS/RalB. To construct pBJ/Ras G12V⌬ , pVL1393/ Ras G12V⌬ (31) was digested with BamHI and PstI and then blunted with T4 DNA polymerase. This fragment was inserted into pBJ-1 which was digested with EcoRI and then blunted with Klenow to generate pBJ/ Ras G12V⌬ . To construct pGEX-2T encoding RalA G23V , ptac/RalA G23V was digested with EcoRI and HindIII and blunted with Klenow. The 0.6-kb fragment encoding RalA G23V was inserted into pGEX-2T which was digested with EcoRI and blunted with Klenow to generate pGEX/ RalA G23V . To construct pGEX-2T encoding RalB S28N , the 0.6-kb fragment encoding RalB S28N with XbaI and BamHI sites was synthesized by PCR. This fragment was digested with XbaI and BamHI, blunted with mung bean nuclease, and then inserted into the SmaI cut pGEX-2T to generate pGEX/RalB S28N . To construct pGEX-2T encoding RalB T46A and RalB G86E , the 0.6-kb fragments containing these Ral mutants with BamHI sites were synthesized by PCR. These fragments were digested with BamHI and inserted into the BamHI cut pGEX-2T to generate pGEX/RalB T46A and pGEX/RalB G86E .
Expression and Purification of GST Fused to Proteins and MBP Fused to Proteins from Sf9 Cells and Escherichia coli-To purify GST fused to Ha-Ras (GST-Ras) and RalB (GST-Ral) from Sf9 cells, monolayers of Sf9 cells (2 ϫ 10 7 cells) were infected with high-titer recombinant baculoviruses (1 ϫ 10 8 plaque-forming units/ml) at a multiplicity of infection of 5/cell. After 72-h post-infection, the cells were washed with cold phosphate-buffered saline and suspended in 1 ml of Buffer A (20 mM Tris/HCl (pH 7.5), 5 mM MgCl 2 , 1 mM EGTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride, 20 g/ml aprotinin, and 10 g/ml leupeptin). The suspension was sonicated and centrifuged at 700 ϫ g for 5 min to remove unbroken cells and nuclei. The homogenate was centrifuged at 100,000 ϫ g for 30 min. The supernatant was used as the cytosol fraction. The pellet was resuspended in 1 ml of Buffer A containing 1% CHAPS, rocked for 1 h, and then centrifuged at 100,000 ϫ g for 30 min. The supernatant was used as the membrane fraction. The post-translationally modified and unmodified forms of GST-Ras or GST-Ral were purified from the membrane and cytosol fractions, respectively, using glutathione-Sepharose 4B columns in accordance with the manufacturer's instructions. To purify MBP fused to RalGDS-(764 -864) (MBP-RalGDS-(764 -864)) and MBP fused to RalBP1-(364 -647) (MBP RalBP1-(364 -647)) from E. coli, transformed E. coli were initially grown in Luria-Bertani's broth at 37°C to an absorbance of 0.8 (optical density at 600 nm) and subsequently transferred to 25°C. Then isopropyl-1-␤-D-thiogalactopyranoside was added at a final concentration of 100 M, and further incubation was carried out for 10 h at 25°C. To purify GST fused to RalGDS (GST-RalGDS) from E. coli, transformed E. coli were initially grown at 37°C in superbroth containing 1% glucose to an absorbance of 1.8 (optical density at 600 nm) and subsequently transferred to 25°C. Then isopropyl-1-␤-D-thiogalactopyranoside was added at a final concentration of 40 M and further incubation was carried out for 4 h at 25°C. GST fused to NF1 (GST-NF1) and RalB (GST-Ral) were produced in E. coli as described (11,15). The GST and MBP fused to proteins were purified from E. coli by affinity chromatographies in accordance with the manufacturer's instructions.
Triton X-114 Separation-GST-Ras and GST-Ral (6 pmol each) purified from Sf9 cells in 1 ml of reaction mixture (20 mM Tris/HCl (pH 7.5), 1% Triton X-114, 5 mM MgCl 2 , 1 mM EDTA, 0.038% CHAPS, and 1 mM DTT) were warmed for 2 min at 37°C (37). The turbid solution was centrifuged at 10,000 ϫ g for 2 min at room temperature. The upper phase was the aqueous phase (detergent-depleted), and the lower phase was the detergent-enriched phase. After the separation, the volumes of the samples were equalized. Aliquots of each sample were subjected to SDS-PAGE and probed with the anti-HA antibody.
Partial Purification of RalGAP-Bovine brain was obtained from the heads of freshly slaughtered cattle and frozen at Ϫ80°C until use. Cerebral tissue (approximately 140 g, wet weight) was homogenized in a Potter-Elvehjem Teflon-glass homogenizer with 280 ml of Buffer B (25 mM Tris/HCl (pH 7.5), 1 mM MgCl 2 , 1 mM EGTA, 1 mM DTT, and 1 mM phenylmethylsulfonyl fluoride). The homogenate was centrifuged at 20,000 ϫ g for 30 min, and the supernatant was centrifuged again at 100,000 ϫ g for 1 h. The supernatant (200 ml, 1500 mg of protein) was applied to DE52 column (4.3 ϫ 11 cm) equilibrated with Buffer B. After the column was washed with 800 ml of Buffer B and 480 ml of Buffer B containing 100 mM NaCl, the elution was performed with Buffer B containing 200, 300, and 500 mM NaCl in a stepwise manner. All the RalGAP activity appeared in the fractions eluted by 200 mM NaCl. Solid (NH 4 ) 2 SO 4 was added to this eluate (24 ml, 144 mg of protein) at a final concentration of 40% saturation. The sample was centrifuged at 20,000 ϫ g for 20 min. Most RalGAP activity was recovered in the precipitate. The precipitate was dissolved in 2.4 ml of Buffer B. After dialysis overnight against a large volume of Buffer B, the sample was centrifuged at 100,000 ϫ g for 1 h. The supernatant (2 ml, 40 mg of protein) was applied to a Sephacryl S-300 column (1.5 ϫ 98 cm) equilibrated with Buffer B. The elution was performed with Buffer B and fractions of 2 ml each were collected. RalGAP activity appeared in Fractions 34 -43. The partial purified RalGAP did not contain the GAP activity for Ras or Rho, but contained the weak GAP activity for Rap1.
Localization of RalGDS Regulated by Ras in COS Cells-COS cells (60 -70% confluent in a 10-cm dish) were transfected with pCGN/Ral-GDS and pBJ/Ras G12V or pBJ/Ras G12V⌬ by the DEAE-dextran method (38). Sixty hours after transfection, the cells were washed with cold phosphate-buffered saline and suspended in 1 ml of Buffer A. This suspension was sonicated and centrifuged at 700 ϫ g for 5 min to remove unbroken cells and nuclei. The homogenate was centrifuged at 100,000 ϫ g for 30 min. The supernatant was used as the cytosol fraction. The pellet was resuspended in 1 ml of Buffer A containing 1% Nonidet P-40, rocked for 1 h, and then centrifuged at 100,000 ϫ g for 30 min. The supernatant was used as the membrane fraction. Aliquots (20 g of protein) of the cytosol and membrane fractions were subjected to SDS-PAGE and probed with the anti-HA or Ras (F235) antibody. For immunoprecipitation assay, aliquots (200 g of protein) of the cytosol and membrane fractions were immunoprecipitated with the anti-Ras antibody (Y13-238), subjected to SDS-PAGE, and probed with the anti-HA or Ras (F235) antibody as described (11,31).
RasGAP Assay-The modified or unmodified forms of GST-Ras (3 pmol) was preincubated for 10 min at 30°C in 9 l of the preincubation mixture (50 mM Tris/HCl (pH 7.5), 2 M [␥-32 P]GTP (8,000 -12,000 cpm/pmol), 5 mM MgCl 2 , 10 mM EDTA, 0.3% CHAPS, 1 mM DMPC, 1 mM DTT, and 1 mg/ml BSA). After the incubation, 1 l of 340 mM MgCl 2 was added. To this preincubation mixture, 30 l of reaction mixture (50 mM Tris/HCl (pH 7.5), 1.3 mM GTP, 0.3 mM MgCl 2 , and 27 mg/ml BSA) containing GST-NF1 (0.8 pmol) was added, and the second incubation was performed for 15 min at 24°C in the presence of the indicated concentrations of MBP-RalGDS-(764 -864). Assays were quantified by rapid filtration on nitrocellulose filters. The GAP activity was calculated from the decrease of the radioactivity of [␥-32 P]GTP compared with the reaction performed in the absence of GST-NF1, and the GAP inhibition activity was expressed as percent decrease of the GAP activity.
Other Assays-The K d for GDP or GTP␥S of, dissociation rate of GDP from, and steady-state and actual catalytic rates of GTP hydrolysis of Ral mutants were determined as described (39 -41).

Purification of the Post-translationally Modified and Unmodified Forms of Ras and
Ral-To examine the roles of the post-translational modifications of Ras and Ral on the signaling pathway from Ras to Ral, we produced GST-Ras and GST-Ral in Sf9 cells and purified them from both the cytosol and membrane fractions of Sf9 cells. These proteins were tagged with an HA epitope. By the Triton X-114 phase-partition method, the proteins purified from the cytosol fraction of Sf9 cells were recovered in the aqueous phase, while the proteins purified from the membrane fraction were recovered in the detergent-enriched phase (Fig. 1). These results indicate that GST-Ras and GST-Ral purified from the cytosol fraction of Sf9 cells are post-translationally unmodified and those from the membrane fraction are post-translationally modified.
Effect of the Post-translational Modification of Ras on Its Binding to RalGDS-We originally found that RGL was an effector protein of Ras and that the C-terminal region of RGL (RGL-(632-734)) bound to the post-translationally unmodified GTP-bound form of Ras (11,42). The C-terminal region of RalGDS (RalGDS-(764 -864)) shared high homology with RGL-(632-734) and was found to bind to the post-translationally unmodified GTP-bound form of Ras (12). However, the effect of the post-translational modification of Ras for its binding to RalGDS was not yet studied. We compared the binding activity of the post-translationally modified forms of Ras to MBP-RalGDS-(764 -864) with that of the unmodified form to MBP-RalGDS- (764 -864). The post-translationally modified form of Ras bound to RalGDS-(764 -864) two to three times more than the unmodified form ( Fig. 2A). When 25 pmol of the modified and unmodified forms of Ras were used in this assay, 1.5 pmol of the modified form and 0.8 pmol of the unmodified form bound to 20 pmol of MBP-RalGDS-(764 -864). As consistent with the previous observations (11,13), RalGDS-(764 -864) inhibited the GAP activity of NF1 for the post-translationally unmodified form of Ras (Fig. 2B). Furthermore, RalGDS-(764 -864) inhibited the GAP activity of NF1 for the post-translationally modified form more effectively (Fig. 2B). The IC 50 values of the GAP inhibitory activity of RalGDS-(764 -864) for the modified and unmodified forms are 170 and 350 nM, respectively. These results suggest that the post-translational modification of Ras may render it effectively bind to RalGDS.
Localization of RalGDS Regulated by Ras in COS Cells-Next, we examined the effect of the post-translational modification of Ras on its interaction with RalGDS in intact cells. When Ras G12V was expressed in COS cells, the unmodified and modified forms of Ras appeared in the cytosol and membrane fractions, respectively (Fig. 3A, lanes 3 and 4). The modified form migrated faster than the unmodified form on SDS-PAGE. On the while, Ras G12V⌬ appeared in the cytosol fraction but not in the membrane fraction (Fig. 3A, lanes 5 and 6). When Ral-GDS alone was transfected in COS cells, most of RalGDS, which was tagged with an HA epitope at the N terminus, was detected in the cytosol fraction of COS cells (Fig. 3A, lanes 1  and 2). When RalGDS was transfected with Ras G12V , RalGDS was detected in both the cytosol and membrane fractions, while when RalGDS was transfected with Ras G12V⌬ , most of RalGDS was still recovered in the cytosol fraction (Fig. 3A, lanes 3-6). We also examined whether RalGDS interacted with Ras on the membranes. When the lysates expressing RalGDS alone were immunoprecipitated with the anti-Ras antibody, RalGDS was not detected in the precipitates (data not shown). As expected, RalGDS made a complex with Ras in both the cytosol and membrane fractions of COS cells expressing RalGDS and Ras G12V (Fig. 3B, lanes 1 and 2), while RalGDS made a complex with Ras in the cytosol fraction but not in the membrane fraction of COS cells expressing RalGDS and Ras G12V⌬ (Fig.  3B, lanes 3 and 4). As consistent with the previous observations (11, 15), nonimmune immunoglobulin did not immunoprecipi-tate a RalGDS-Ras complex (data not shown). These results suggest that the modified form of Ras localized to the membranes may recruit RalGDS from the cytosol to the membranes, resulting in placing RalGDS in the vicinity of Ral, which is also present on the membranes through its post-translational modification.
Effect of Mutations of Ral on the Actions of RalGDS and RalGAP-To analyze the role of the signaling pathway from Ras to Ral, Ral mutants are useful tools. However, the characterization of Ral mutants has not yet been systematically done. We constructed wild type Ral and four Ral mutants (Ral G23V , Ral S28N , Ral T46A , and Ral G86E ) in bacterial expression vector and purified them as GST fused to proteins. These Ral mutants correspond to Ras mutants, Ras G12V , Ras S17N , Ras T35A , and Ras G75E . The characterization of these Ral mutants were summarized in Table I. The K d values of wild type Ral for GDP and GTP␥S were similar and about 20 nM. Ral G23V and Ral T46A also showed the similar K d values for both GDP and GTP␥S. The affinity of Ral S28N for GDP was 10-fold higher than for GTP␥S, while Ral G86E exhibited the higher affinity for GTP␥S than GDP. The GDP dissociation constants (K Ϫ1 ) of Ral mutants were variable. The K Ϫ1 values of Ral, Ral G23V , Ral S28N , Ral T46A , and Ral G86E were 0.0043, 0.0011, 0.13, 0.0067, and 0.0017, respectively. RalGDS stimulated the dissociation of GDP from Ral 10-fold and showed the similar effect on Ral T46A . RalGDS stimulated the dissociation of GDP from Ral G23V 4-fold, but affected that from Ral S28N or Ral G86E very little. The steady-state rates (K ss ) of GTPase activity of Ral, Ral G23V , Ral T46A , and Ral G86E were 0.021, 0.0084, 0.015, and 0.025, respectively. RalGAP stimulated the actual catalytic rates (K cat ) of GTPase activity of Ral and Ral G86E 3-4-fold, but not those of Ral G23V or Ral T46A . These results indicate that Ser 28 and Gly 86 of Ral are important for the action of RalGDS and that Gly 23 and Thr 46 are important for the action of RalGAP.
Effect of the Post-translational Modification of Ral on the Actions of RalGDS and RalGAP-It has been reported that RasGDSs such as smgGDS, mCDC25, and hSOS act more effectively on the post-translational modified form of Ras than the unmodified form (28 -30). We examined the effect of posttranslational modification of Ral on the RalGDS activity. The modified and unmodified forms of Ral bound to GDP and

FIG. 3. Localization of RalGDS regulated by Ras in COS cells.
A, translocation of RalGDS by Ras G12V but not by Ras G12V⌬ . COS cells expressing RalGDS alone (lanes 1 and 2), both RalGDS and Ras G12V (lanes 3 and 4), or both RalGDS and Ras G12V⌬ (lanes 5 and 6) were disrupted and separated into the cytosol (lanes 1, 3, and 5) and membrane (lanes 2, 4, and 6) fractions. Aliquots of each sample were subjected to SDS-PAGE and probed with the anti-HA and Ras (F235) antibodies. B, interaction of Ras with RalGDS on the membranes. The proteins of the cytosol (lanes 1 and 3) and membrane (lanes 2 and 4) fractions of COS cells expressing both RalGDS and Ras G12V (lanes 1 and  2) or both RalGDS and Ras G12V⌬ (lanes 3 and 4)  GTP␥S with the similar efficiency (data not shown). GDP was dissociated from both the modified and unmodified forms of Ral with the similar efficiency in a time-dependent manner (Fig.  4A). The action of RalGDS to stimulate the dissociation of GDP from Ral was more effective on the modified form of Ral than on the unmodified form (Fig. 4A). RalGDS stimulated the binding of GTP␥S to the modified form of Ral more effectively than to the unmodified form (Fig. 4B). When the binding of GTP␥S to Ral was assayed in the presence of RalGDS (50 nM) using various amounts of Ral, the apparent K m values of RalGDS for the modified and unmodified forms of Ral were estimated to be 380 nM and 3.2 M, respectively. The V max values of RalGDS for the modified and unmodified forms were 1.5 and 1.6 nmol/min/ nmol, respectively. These results taken together with the results of Table I suggest that the post-translational modification of Ral is not essential for the action of RalGDS but that the modification is important for increasing the affinity of RalGDS for Ral. GTP was hydrolyzed in both the modified and unmodified forms of Ral with the similar efficiency in a time-dependent manner (Fig. 5). In contrast to the case of RalGDS, the post-translational modification of Ral did not affect the Ral-GAP activity (Fig. 5).
Effect of the Post-translational Modification of Ral on the Binding to RalBP1-RalBP1 has been identified as an effector protein of Ral by yeast two-hybrid system (16 -18). The Ralbinding domain of RalBP1 has been found to be localized in its C-terminal half. We purified Ral-binding domain of RalBP1 as MBP-RalBP1-(364 -647). As consistent with the previous observations (16 -18), the post-translationally unmodified GTPbound form of Ral interacted with RalBP1-(364 -647) (Fig. 6A). The post-translationally modified form of Ral bound to RalBP1-(364 -647) two to three times more than the unmodified form (Fig. 6A). When 25 pmol of the modified and unmodified forms of Ral were used in this assay, 2.5 pmol of the modified form and 1.2 pmol of the unmodified form bound to 20 pmol of MBP-RalBP1-(364 -647). RalBP1 has been shown not to bind to Ral D49N , which is a mutant in the putative effector loop (16). As shown in Table I, RalGAP did not act on Ral T46A , which is also a mutant in the putative effector loop. These results indicate that RalGAP and RalBP1 share a common binding domain on Ral. We examined the effect of RalBP1 on the RalGAP activity for the post-translationally modified and unmodified forms of Ral. RalBP1-(364 -647) slightly inhibited the endogenous GTPase activities of both the modified and unmodified forms of Ral with the similar efficiency (data not shown). RalBP1-(364 -647) inhibited the RalGAP activity in a dose-dependent manner (Fig. 6B). This inhibitory action of RalBP1-(364 -647) was stronger for the modified form of Ral than for the unmodified form. The IC 50 values of the GAP inhibitory action of RalBP1-(364 -647) for the modified and unmodified forms are 200 and 450 nM, respectively. These results suggest that RalBP1 competes with RalGAP for binding to the putative effector loop of Ral and that the post-translational modification of Ral may render it effectively bind to RalBP1.

DISCUSSION
In this paper we have shown that the post-translational modification of Ras enhances its binding activity to RalGDS and is required for the localization of RalGDS on the membranes in intact cells and that the post-translational modification of Ral is important for the RalGDS action and enhances its binding activity to RalBP1.
Small G proteins undergo a series of post-translational modification, which result in their localization to the membrane fractions such as plasma membrane, endoplasmic reticulum, and Golgi apparatus, and which are essential for their biological actions (20,21,23). It has been demonstrated that the post-translational modifications of small G proteins are important for the actions of their GDP/GTP exchange proteins. smg-GDS promotes the GDP/GTP exchange of the modified forms of Ras, Rap, RhoA, Rac, but not that of the unmodified form (28). mCDC25 and hSOS are active on the modified form of Ras S]GTP␥S in 100 l of reaction mixture (50 mM Tris/HCl (pH 7.5), 5 mM MgCl 2 , 1 mM DTT, and 1 mg/ml BSA) (40). To determine the RalGDS activity for Ral mutants, [ 3 H]GDP-bound form of Ral mutants (100 nM) were incubated with or without 200 nM RalGDS for various periods of time at 30°C. GDP dissociation constant (K Ϫ1 ) was determined as described (41). The steady-state rate (K ss ) of GTPase activity was determined by incubating Ral mutants (100 nM) with 1 M [␥-32 P]GTP for various periods of time at 30°C and expressed as turnover number (40). The actual catalytic rate (K cat ) of GTPase activity was determined in the presence or absence of RalGAP (7 g of protein) as described (39). GTPase activity of Ral S28N was not determined because most of [␥- 32   more effectively than on the unmodified form (29,30). RabGDI and RhoGDI inhibit the GDP/GTP exchange of the modified forms of the members of Rab subfamily and Rho subfamily, respectively, but not that of the unmodified forms (43,44). Consistent with these observations, we have demonstrated that RalGDS stimulates the GDP/GTP exchange of the modified form of Ral more effectively than that of the unmodified form. Since the K m value of RalGDS for the modified form of Ral is smaller than that for the unmodified form, the posttranslational modification of Ral could increase the affinity of RalGDS for Ral. In contrast to RalGDS, RalGAP activity is not affected by the post-translational modification of Ral. These results are also consistent with the previous observations that the post-translational modifications of small G proteins do not affect the activities of RasGAP, RapGAP, and RhoGAP (44 -46).
It has been known that the post-translational modifications of small G proteins are critical for their functions (21,23). Deletion or mutation of the C-terminal region of Ras abolishes the transforming activity (21,23,47). The modified form of Ras stimulates adenylate cyclase more effectively than the unmodified form in vitro (26). The modified form of Ras activates B-Raf in the presence of 14-3-3 protein, but the unmodified form does not in vitro (48). We have also found that the posttranslational modification of Ras is essential for the Raf activation in intact cells (31). Rab family members require the post-translational modification to regulate the vesicle transport (49 -51). We have demonstrated that the modified forms of Ras and Ral bind to their effector proteins, RalGDS and RalBP1, two to three times more than the unmodified forms.
We have also shown that RalGDS and RalBP1 inhibit the activities of NF1 and RalGAP, respectively. The inhibitory actions of RalGDS and RalBP1 are more effective on the modified form than on the unmodified form. These results suggest that the post-translational modifications of Ras and Ral may increase their affinities for the effector proteins.
We have not yet demonstrated whether the bindings of Ras to RalGDS and Ral to RalBP1 regulate the activities of RalGDS and RalBP1. The binding of Ras to Raf is not sufficient for the activation of Raf (31). The role of the post-translational modification of Ras to activate Raf could be to bring Raf to the membranes where some activator may be present (52,53). Therefore, the post-translational modifications of Ras and Ral might be important for the subcellular localization of RalGDS and RalBP1, resulting in transmitting the signals to downstream molecules. Indeed, our results have demonstrated that the post-translational modification of Ras is necessary for the localization of RalGDS on the membranes. Since Ral is localized to the membranes through the post-translational modification, it is possible that RalGDS on the membranes acts on Ral. RalBP1 has been found to show the GAP activities for CDC42 and Rac (16 -18). CDC42 and Rac regulate the cytoskelton and the stress-activated protein kinase activity (54 -57). Since CDC42 and Rac are also localized to the membranes, Ral might regulate the subcellular localization of RalBP1 to act on CDC42 and Rac on the membranes. It has been reported that RalGDS-mediated GDP/GTP exchange of Ral is activated by Ras in COS cells and that a RalGDS mutant in which Rasinteracting domain is deleted fails to respond to Ras (58). These results indicate that Ras-induced RalGDS activation is dependent upon the two protein binding and that Ras mediates the redistribution of RalGDS to Ral. This is analogous to the mechanism to receptor activation of SOS, a GDP/GTP exchange protein of Ras (59,60). SOS is in a complex with Grb2 in the cytosol in the absence of growth factors. After a growth factor induces autophosphorylation of its receptor, the Grb2-SOS complex translocates from the cytosol and associates with the receptor on the membranes thereby placing it in the vicinity of Ras. Therefore, our results suggest that the post-translational modifications of Ras and Ral are also important for regulating the localization of their effector proteins in the cells, resulting in bringing them into contact with their substrates on the membranes. However, since our experiments have been done by overexpression of RalGDS and/or Ras in COS cells, we cannot exclude the possibility that endogenous RalGDS is localized on the membranes, that overexpressed RalGDS is in the cytosol simply because RalGDS is already saturated on the membranes, and that addition of Ras G12V takes this excess RalGDS to the membranes. Therefore, it is necessary to determine subcellular localization of endogenous RalGDS.
We have also characterized several Ral mutants. Ral G23V , Ral S28N , Ral T46A , and Ral G86E are analogous to Ras G12V , Ras S17N , Ras T35A , and Ras G75E , respectively. Ral S28N exhibits a higher affinity for GDP than GTP and markedly increases the intrinsic GDP dissociation activity thereby suppressing the effect of RalGDS. Ral G86E shows a higher affinity for GTP than GDP and is not sensitive to RalGDS. These properties of Ral S28N and Ral G86E are similar to those of Ras S17N and Ras G75E . It is known that Ser 17 of Ras binds to Mg 2ϩ and that Ras S17N acts as a dominant negative form (61). Therefore, it is possible that Ser 28 of Ral binds to Mg 2ϩ and that Ral S28N acts as a dominant negative form. The substitution of Gly 75 with Glu of Ras abolishes the growth inhibitory activity of Ras S17N , indicating that Gly 75 is important for the interaction with hSOS or mCDC25 (62). By analogy with Ras, it is likely that Ser 28 of Ral is required for RalGDS stimulation and that Gly 86 of Ral is important for RalGDS binding. However, RalGDS acts on Ral T46A as well as on wild type of Ral. This result is different from the case of mCDC25 and Ras, since mCDC25 does not stimulate the GDP/GTP exchange of Ras T35A (63). Our results clearly show that RalGAP acts on Ral G86E , but not on Ral G23V or Ral T46A . Therefore, the residues of Ral interacting with RalGAP might be different from those with RalGDS. These mutants will be useful for analyzing the functions of Ral.
It is conceivable that RalGDS mediates the signal from Ras to Ral (11)(12)(13)58). It has been reported that a dominant negative mutant of Rac reverses the v-Ras-dependent transformation (64,65). Although the functions of Ral are not clear, it is intriguing to speculate that there is a new signaling pathway consisting of Ras-RalGDS-Ral-RalBP1-CDC42/Rac. Furthermore, it has been recently reported that Ral interacts with PLD (19). PLD has been known to be localized to the membranes. Taken together, the signal from Ras to Ral could be transmitted on the membranes and regulate various cell functions and the post-translational modifications of Ras and Ral may be important for transmitting the signal effectively. Further studies are necessary to understand the whole picture of this new signaling pathway of Ras through RalGDS.