CalDAG-GEFIII Activation of Ras, R-Ras, and Rap1*

We characterized a novel guanine nucleotide exchange factor (GEF) for Ras family G proteins that is highly homologous to CalDAG-GEFI, a GEF for Rap1 and R-Ras, and to RasGRP/CalDAG-GEFII, a GEF for Ras and R-Ras. This novel GEF, referred to as CalDAG-GEFIII, increased the GTP/GDP ratio of Ha-Ras, R-Ras, and Rap1 in 293T cells. CalDAG-GEFIII promoted the guanine nucleotide exchange of Ha-Ras, R-Ras, and Rap1 in vitroalso, indicating that CalDAG-GEFIII exhibited the widest substrate specificity among the known GEFs for Ras family G proteins. Expression of CalDAG-GEFIII was detected in the glial cells of the brain and the glomerular mesangial cells of the kidney by in situhybridization. CalDAG-GEFIII activated ERK/MAPK most efficiently, followed by CalDAG-GEFII and CalDAG-GEFI in 293T cells. JNK activation was most prominent in cells expressing CalDAG-GEFII, followed by CalDAG-GEFIII and CalDAG-GEFI. Expression of CalDAG-GEFIII induced neuronal differentiation of PC12 cells and anchorage-independent growth of Rat1A cells less efficiently than did CalDAG-GEFII. Thus, co-activation of Rap1 by CalDAG-GEFIII apparently attenuated Ras-MAPK-dependent neuronal differentiation and cellular transformation. Altogether, CalDAG-GEFIII activated a broad range of Ras family G proteins and exhibited a biological activity different from that of either CalDAG-GEFI or CalDAG-GEFII.

* This work was supported in part by grants from the Ministry of Education, Science, Sports and Culture, the Health Science Foundation, and the Princess Takamatsu Cancer Research Fund, Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AB020653 (for KIAA0846) and AB002349 (for KIAA0351).
Antibodies-Anti-Rap1 polyclonal antibody, anti-Pan-Ras antibody, and anti-FLAG M2 monoclonal antibody were purchased from Santa Cruz Biotechnology (Santa Cruz, CA), Calbiochem, and Sigma. Anti-HA monoclonal antibody was obtained from Roche Molecular Biochemicals.
Cell Culture and Transfection-Human 293T embryonic kidney cells and Rat-1A cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (Nissui, Tokyo) supplemented with 10% fetal calf serum. PC12 cells were maintained in DMEM plus 10% fetal calf serum and 5% horse serum. 293T cells were transfected with plasmid DNA by the calcium phosphate precipitation method.
Analysis of Guanine Nucleotide Exchange Activity in 293T Cells-Guanine nucleotide exchange activity of GEFs was analyzed in 293T cells as described previously (26). 293T cells in dishes 3.5 cm in diameter were transfected with pEBG-derived vectors encoding Ras family G proteins and with pCXN2-FLAG-derived vectors encoding GEFs. Twenty four hours after transfection, cells were labeled for 4 h with 32 P i in phosphate-free minimum Eagle's medium (Life Technologies, Inc.). Cells were lysed in TLC lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 20 mM MgCl 2 , 1 mM Na 3 VO 4 , 1% Triton X-100) and cleared by centrifugation. The supernatant was incubated with glutathione-Sepharose beads at 4°C for 1 h. The beads were then washed twice with TLC lysis buffer and once with washing buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 20 mM MgCl 2 ). Beads were resuspended in 2% SDS and boiled. The supernatant was spotted on polyethyleneimine-coated plates (Macherey & Nagel). The TLC plate was developed with 0.5 M H 3 PO 4 -KH 2 PO 4 , pH 3.4, and analyzed on a BAS1000 image analyzer (Fuji Photo Film Co. Ltd., Tokyo, Japan).
For the measurement of GEF activity, 400 nM labeled Ha-Ras, R-Ras, or Rap1A was incubated with or without 100 nM GEFs in reaction buffer (50 mM Tris-HCl, pH 7.5, 5 mM MgCl 2 , and 2 mM dithiothreitol) at 20°C. The reaction was started by the addition of GTP to 200 M, and the decrease in fluorescence was monitored in a luminescence spectrometer, JASCO FP-750 (JASCO, Japan), with excitation and emission wavelengths of 366 and 450 nm, respectively.
In Vitro ERK and JNK Assay-The activity of ERK and JNK was examined as described previously (34). 293T cells in dishes 3.5 cm in diameter were cotransfected with pEBG-ERK or pEBG-JNK and with expression vectors for GEFs. Twenty four hours post-transfection, cells were lysed in kinase lysis buffer (25 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 10 mM sodium pyrophosphate, 10 mM ␤-glycerophosphate, 10% glycerol). Glutathione-Sepharose beads were added to the cleared lysates and rotated at 4°C for 1 h. The beads were washed with KLB and kinase buffer (50 mM HEPES-NaOH, pH 7.4, 10 mM MgCl 2 ) and incubated with 15 l of kinase reaction mixture at 30°C for 10 min. Kinase reaction mixture is kinase buffer containing 100 M ATP, 5 Ci of [␥-32 P]ATP, and 0.2 mg/ml myelin basic protein (Sigma) for ERK or GST-c-Jun for JNK. The reaction was terminated by the addition of 5 l of 5ϫ Laemmli sample buffer, and proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The radioactivity of the substrates was quantitated with a BAS1000 image analyzer.
Neurite Outgrowth of PC12 Cells-PC12 cells were transfected with expression plasmids by the use of LipofectAMINE 2000 (Life Technologies, Inc.) and 48 h later observed under a microscope (34). PC12 cells that extended neurites longer than 2-fold of the diameter of the cell body were counted as differentiated cells.
Analysis of Rat-1A Cell Lines Expressing CalDAG-GEFs-Rat-1A cells were transfected with pCXN2-FLAG-derived CalDAG-GEF expression vectors by use of LipofectAMINE (Life Technologies, Inc.). Cells were selected with 1 mg/ml G418 (Life Technologies, Inc.), and well isolated colonies were maintained in medium containing 400 g/ml G418. A colony formation assay was performed as described previously (35). Briefly, 10 5 cells before and after cloning were cultured in DMEM containing 0.8% agarose with 2 or 10% fetal bovine serum at 37°C in 5% CO 2 . Two weeks later, colonies larger than 0.5 mm in diameter were counted.
Analysis of GTP-bound Ras Family G Proteins in Rat-1A Cells-Analysis of GTP-bound Ras family G proteins was performed by the method of Bos and co-workers (29). Rat-1A cells were serum-starved for 36 h, lysed in pull-down lysis buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM MgCl 2 , 1% Nonidet P-40, 0.5% deoxycholic acid, 0.1% SDS, 1 mM Na 3 VO 4 ), and clarified by centrifugation. The supernatant was incubated with GST-tagged RalGDS-RBD or GST-tagged Raf-RBD ϩ CRD prebound to glutathione-Sepharose beads at 4°C. GST-RalGDS-RBD and GST-Raf-RBD ϩ CRD were used for the detection of GTP-Rap1 and GTP-Ras, respectively. Beads were washed twice with pulldown lysis buffer, resuspended in Laemmli's sample buffer, and subjected to SDS-PAGE, followed by immunoblotting analysis with anti-Ras or anti-Rap1 antibody.

RESULTS
Identification of CalDAG-GEFIII-We carried out a data base search for new GEFs for the Ras family G proteins and identified KIAA0846. Through the entire coding region, KIAA0846 showed high amino acid sequence homology to both CalDAG-GEFI and CalDAG-GEFII (Fig. 1). Based on this high sequence homology, KIAA0846 was designated as CalDAG-GEFIII. CalDAG-GEFIII contained a CDC25 homology domain, an EF hand, and a C1 domain, as do CalDAG-GEFI and CalDAG-GEFII. The phylogenetic tree of the CDC25 domains of GEFs for Ras family G proteins proved that CalDAG-GEFIII was most closely related to CalDAG-GEFII. Whereas this CalDAG-GEF branch contains GEFs for Ras, R-Ras, and Rap1, GEFs in another branch consisting of nRap GEP/PDZ-GEF1/ RA-GEF, GFR, Epac/cAMP-GEFI, and cAMP-GEFII are specific for Rap1.
Distribution of CalDAG-GEFIII-In reverse transcriptase-PCR assay, expression of CalDAG-GEFIII has been detected in brain, heart, lung, and kidney. We confirmed the expression of CalDAG-GEFIII mRNA by Northern blotting analysis in the brain but not in the other organs. 2 We further examined the cell type-specific expression of CalDAG-GEFIII by in situ hybridization and compared it with those of CalDAG-GEFI and CalDAG-GEFII. All CalDAG-GEFs were expressed in the brain. However, we found a remarkable difference in the distribution of the cells expressing each of the CalDAG-GEFs (Fig.  2). As reported previously, CalDAG-GEFI was expressed most abundantly in the basal ganglia (16). High level expression of CalDAG-GEFII was observed in the pyramidal cells of the hippocampus and the Purkinje cells of the cerebellum (15,16). In contrast, CalDAG-GEFIII was expressed most prominently in the glial cells of the cerebral and cerebellar white matter. These cells were negative for glial fibrillary acidic protein, a marker for astrocyte, suggesting that CalDAG-GEFIII was expressed mostly in the oligodendroglia. 3 In the kidney also, expression of CalDAG-GEFs did not overlap each other. CalDAG-GEFI was expressed in the interstitial cells, CalDAG-GEFII in the epithelium of the distal convoluted tubules and collecting tubules, and CalDAG-GEFIII in the mesangial cells of the glomeruli. Thus, this observation showed that the three CalDAG-GEFs play a role in the cell type-specific regulation of Ras family G proteins by calcium and diacylglycerol.
Guanine Nucleotide Exchange Activity of CalDAG-GEFIII in Vivo-The substrate specificity of CalDAG-GEFIII was examined first in 293T cells (Fig. 3A). The expression of CalDAG-GEFIII significantly increased the GTP-bound form of Ha-Ras, R-Ras, Rap1A, and Rap2A. CalDAG-GEFIII did not activate RalA; therefore, the activity of CalDAG-GEFIII was restricted to the Ras, R-Ras, and Rap subfamilies. We used KIAA0351 as a positive control for the Ral guanine nucleotide exchange factor. As reported previously (15,16), CalDAG-GEFI activated Rap1 and, to a lesser extent, R-Ras, and CalDAG-GEFII activated both Ha-Ras and R-Ras.
To exclude the possibility that the regulatory domain modifies the substrate specificity of CalDAG-GEFs, we constructed chimeric proteins between CalDAG-GEFI and CalDAG-GEFII. CalDAG-GEFI/II consisted of the catalytic domain of CalDAG-GEFI and the regulatory domain of CalDAG-GEFII, whereas CalDAG-GEFII/I consisted of the catalytic domain of CalDAG-GEFII and the regulatory domain of CalDAG-GEFI. As shown in Fig. 3B, CalDAG-GEFI/II and CalDAG-GEFII/I showed substrate specificity identical to that of CalDAG-GEFI and CalDAG-GEFII, respectively, indicating that the regulatory domains did not alter the substrate specificity of CalDAG-GEFI or CalDAG-GEFII.
Guanine Nucleotide Exchange Activity of CalDAG-GEFIII in Vitro-We confirmed the guanine nucleotide exchange activity of CalDAG-GEFIII in vitro. For this purpose, we expressed and purified the catalytic domains of CalDAG-GEFs from Escherichia coli. Rap1A, R-Ras, and Ha-Ras were loaded with mant-GDP, and its dissociation in the presence of CalDAG-GEFs was monitored with a fluoro-spectrometer (Fig. 4). Rate constants were calculated as described previously (Table I) (33). Release of the mant-GDP from Ha-Ras was promoted by CalDAG-GEFIII and, to a lesser extent, CalDAG-GEFII, whereas the release of mant-GDP from Rap1A was promoted by CalDAG-GEFI and, less efficiently, by CalDAG-GEFIII. Dissociation of mant-GDP from R-Ras was promoted by CalDAG-GEFIII, followed by CalDAG-GEFII and CalDAG-GEFI. Thus, the  CalDAG-GEFIII promoted guanine nucleotide exchange of Ha-Ras, R-Ras, and Rap1 both in vivo and in vitro.
Activation of ERK/MAPK and JNK by CalDAG-GEFIII-To understand the role of the broad substrate specificity of CalDAG-GEFIII, we measured the activity of ERK and JNK in 293T cells expressing CalDAG-GEFs (Fig. 5). Activation of ERK/MAPK was most prominent by CalDAG-GEFIII, followed by CalDAG-GEFII. In contrast, JNK was activated most strongly by CalDAG-GEFII, followed by CalDAG-GEFIII. Activation of ERK or JNK by CalDAG-GEFI was marginal. Because we did not find any significant difference in the expression levels of CalDAG-GEFs, the difference in the activation of ERK/MAPK and JNK seemed to reflect the difference in the substrate specificity.
Neurite Outgrowth of PC12 Cells by CalDAG-GEFIII-It is known that constitutive activation of ERK/MAPK induces neuronal differentiation of PC12 cells (36). To gain further insight into the biological activity of CalDAG-GEFIII, we introduced CAAX box-containing active CalDAG-GEFs into PC12 cells (Fig. 6). Neuronal differentiation of PC12 cells, which was evaluated by the number and the length of the neurites, was most prominent when we expressed CalDAG-GEFII-CAAX, followed by CalDAG-GEFIII-CAAX. CalDAG-GEFI-CAAX did not induce neurite outgrowth. Interestingly, whereas CalDAG-GEFII-expressing cells exhibited typical neurite extension, cells expressing CalDAG-GEFIII showed spreading of the cytoplasm with wide neurites. The transfection efficiency of the expression vectors in PC12 cells exceeded 50% in all cases when the cells were immunostained with anti-FLAG antibody (data not shown). The amount of each protein was also within a comparable level.
Anchorage-independent Growth of Rat-1A Cells Expressing CalDAG-GEFIII-Next, we compared the colony-forming activity of Rat-1A cells transfected with CalDAG-GEFs. In our initial experiment, we transfected Rat-1A cells with CalDAG-GEF expression vectors, selected with G418, and examined the bulk of the drug-resistant colonies for the anchorage-independent growth in soft agar. As shown in Fig. 7A, CalDAG-GEFIII induced colony formation less efficiently than did CalDAG-GEFII and CalDAG-GEFII/I. CalDAG-GEFI and CalDAG-GEFI/II, both of which could not activate Ras, did not induce anchorage-independent growth. To avoid possible bias in the expression level of each CalDAG-GEF, we isolated several independent drug-resistant colonies and obtained cell lines with equivalent expression levels (Fig. 7C). Again, we found that the colony formation efficiency of cells expressing CalDAG-GEFIII was significantly lower than that of cells expressing CalDAG-GEFII (Fig. 7B). We observed similar levels of increase in GTP-Ras in cells expressing CalDAG-GEFII and CalDAG-GEFIII (Fig. 7D). The level of GTP-Rap1 was highest in cells expressing CalDAG-GEFI, followed by CalDAG-GEFIII-expressing cells. This result suggests that the activation of Rap1 by CalDAG-GEFIII attenuated the Ras-dependent growth of Rat-1A cells in soft agar. DISCUSSION We have characterized CalDAG-GEFIII, which shows the broadest substrate specificity among GEFs for Ras family G proteins, i.e. this is the first GEF that activates G proteins of the Ras, R-Ras, and Rap subfamilies. The phylogenetic analysis of the catalytic domain of GEFs of Ras family G proteins showed that CalDAG-GEFI, CalDAG-GEFII, and CalDAG-GEFIII made up a subfamily. All three CalDAG-GEFs promoted the guanine nucleotide exchange of R-Ras, although the levels of activation differed significantly for each CalDAG-GEF. CalDAG-GEFI and CalDAG-GEFII differed in that the former stimulated Rap1, whereas the latter stimulated Ras. Identification of CalDAG-GEFIII, which activated Ras, R-Ras, and Rap1, strongly suggests that the prototype CalDAG-GEF catalyzed the guanine nucleotide exchange of all of the Ras, R-Ras, and Rap subfamilies and that CalDAG-GEFI and CalDAG-GEFII lost the capability to catalyze Ras and Rap1, respectively, during evolution. Interestingly, there is only one CalDAG-GEF in Caenorhabditis elegans. Determination of the substrate specificity of C. elegans CalDAG-GEF will provide insights into the evolution CalDAG-GEFs.
GEFs that activate both R-Ras and Rap1 or R-Ras and Ras have already been reported (23); however, the finding that CalDAG-GEFIII promoted the guanine nucleotide exchange of these three subfamilies of G proteins questioned about the specificity in the signaling from GEFs to Ras family G proteins. The Ras, R-Ras, and Rap1 subfamilies of G proteins are expressed ubiquitously (37), suggesting that CalDAG-GEFIIIexpressing cells also express all of these three Ras subfamily G proteins. Although we showed that CalDAG-GEFIII activates Ras, R-Ras, and Rap1 in vitro and in cultured cells, this does not necessarily confirm that CalDAG-GEFIII promotes the guanine nucleotide exchange reaction of these G proteins in vivo, particularly in highly differentiated cells.
There are two possibilities to explain the redundancy in the substrate specificity. CalDAG-GEFIII may be used to activate simultaneously all of the Ras, R-Ras, and Rap1 subfamilies. Alternatively, there may be a mechanism that restricts the substrate specificity of CalDAG-GEFIII in vivo. For example, Rat-1A cells were transfected with CalDAG-GEF expression vectors and selected with 1 mg/ml G418. A, the bulk of the G418-resistant cells were trypsinized and counted. 10 5 cells were plated in 0.8% agarose-containing DMEM per 6-cm dish. Two weeks later, colonies larger than 0.5 mm in diameter were counted. Bars indicate S.D. B, after selection by G418, well isolated colonies were trypsinized and cultured in medium containing 400 g/ml G418. 10 5 cells were analyzed for anchorage-independent growth as shown in A. C, equivalent amounts of cells used in B were lysed in lysis buffer, separated by SDS-PAGE, and analyzed by immunoblotting with anti-FLAG antibody. D, equivalent amounts of cells were lysed and incubated with GST-Raf-RBD ϩ CRD. Proteins bound to the beads (GTP-Ras) and total lysate (Ras) were analyzed by immunoblotting with anti-Ras antibody (upper two columns). Similarly, GTP-Rap1 bound to GST-RalGDS and Rap1 in the total lysates was detected with anti-Rap1 antibody (lower two columns).
Rap1 is believed to be localized in the Golgi apparatus, whereas Ras is localized mostly at the plasma membrane (38). Thus, CalDAG-GEFIII may be used to activate Ras at the plasma membrane and Rap1 at the Golgi apparatus. To solve these problems, we must develop methods to stimulate GEFs without overexpression and to measure the level of GTP-binding form without disrupting the structure of the cell.
CalDAG-GEFIII, like CalDAG-GEFII, retained biological phenotypes attributable to the activation of Ras, including the activation of ERK/MAPK, the induction of neuronal differentiation of PC12 cells, and the transformation of Rat-1A cells (36). However, there were apparent differences between CalDAG-GEFII and CalDAG-GEFIII in the level of these biological effects. CalDAG-GEFIII consistently activated Ras and ERK/ MAPK more strongly than did CalDAG-GEFII in 293T cells or in vitro. Nevertheless, CalDAG-GEFIII induced neuronal differentiation of PC12 cells and transformation of Rat-1A cells less efficiently than did CalDAG-GEFII. Therefore, the attenuated phenotype of CalDAG-GEFIII appeared to result from its activity toward Rap1, which is known to antagonize Ras (7).
Rap1 is reported to induce neuronal differentiation of PC12 cells via B-Raf activation (39). However, none of Rap1, C3G, Epac, and CalDAG-GEFI induced neurite outgrowth of the PC12 cells used in our experiments, negating the positive contribution of Rap1 in CalDAG-GEFIII-induced neuronal differentiation of PC12 cells used in our laboratory. 2 Rather, our observation suggested that co-stimulation of Rap1 by CalDAG-GEFIII attenuated the effect of Ras. It has been reported that Rap1 inhibits Ras transformation by the competitive binding to c-Raf (40,41). However, more recently, it has also been reported that Rap1 may enhance cell attachment by activating integrins, suggesting that Rap1 exerts its anti-Ras effect apart from c-Raf inhibition (9,19,24,42).
In conclusion, we characterized CalDAG-GEFIII, which showed the broadest substrate specificity among the known GEFs of Ras family G proteins. Mutually exclusive expression patterns of CalDAG-GEFs in the brain suggest that each CalDAG-GEF plays an important role in the regulation of higher order brain functions.