INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The Ras family of G proteins consists of Ras (H-, N-, and K-), Rap1, Rap2, R-Ras, TC21, Ral, Rheb, and M-Ras (R-Ras3) (5). Compared to Ras, which has been extensively studied as a pivotal protein in cell growth and differentiation (4, 8), much less is known about the other Ras-family G proteins. Rap1, which shares its effector domain with Ras, antagonizes Ras in many aspects. Overexpression of Rap1 suppresses Ras-induced transformation of NIH 3T3 cells (30), Ras-induced c-fos activation (47), and Ras-dependent inhibition of muscarinic K⁺ channels (53). At least some of these effects appear to be due to the suppression of Ras-induced activation of Raf serine/threonine kinase and mitogenic ERK/mitogen-activated protein kinase (ERK/MAPK) (10, 19). In concordance with these findings, constitutive activation of Rap1 inhibits interleukin-2 (IL-2) gene production and causes T-cell anergy (7), and inhibition of Rap1 by insulin or lysophosphatidic acid stimulates Ras (40). However, Rap1 may activate the MAPK cascade in different milieus. Rap1 activates ERK/MAPK via the activation of B-Raf in neuronal cells (52, 54) and induces DNA synthesis and oncogenic transformation of Swiss 3T3 cells (1, 55).

Rap1 circulates between GTP-bound active and GDP-bound inactive states. The activation is induced by guanine nucleotide exchange factors (GEFs); these include C3G, CalDAG-GEFI, and Epac (or cyclic AMP [cAMP]-GEF), which are activated by tyrosine kinases, Ca and diacylglycerol, and cAMP, respectively (12, 27, 28, 50). Thus, many signals converge at Rap1 via different GEFs. There are four GTPase-activating proteins (GAPs) of Rap1: rap1GAP, SPA-1, GAP1^IP48P, and tuberin (6). Little is known about the regulation of these GAPs, except for that of an isoform of rap1GAP, rap1GAPII, which has recently been shown to bind to and transduce signal from the  subunit of heterotrimeric G_i protein (37).

Knowledge about Rap2, the amino acid sequence of which shares 60% identity with Rap1, is limited. Rap2 is reported to localize mainly in the endoplasmic reticulum (ER), whereas Rap1 localizes at the Golgi apparatus (2, 3). Unlike Rap1, Rap2 cannot reverse Ras-induced transformation of NIH 3T3 cells (23), and no biological phenotype has been linked to Rap2 in the literature. The regulation of Rap2 also remains unknown. rap1GAP stimulates Rap2 GTPase activity in vitro, albeit significantly more weakly than Rap1 (22). An attempt to purify a GAP specific to Rap2 culminated in the isolation of rap1GAP (22). Very recently, de Rooij et al. reported that a newly isolated GEF for Rap1, PDZ-GEF1, also activates Rap2 and that GTP-bound Rap2 makes up more than 50% of the Rap2 in A14 and COS1 cells (11).

Rap2 shares most of the effector proteins with Ras and Rap1, except for a recently identified RPIP8 (21, 38), suggesting that there is cross-talk among these three proteins. Here, we show that GEFs and GAPs are shared between Rap1 and Rap2 and that, unlike the other Ras family proteins, the GTP-bound active form makes up at least 50% of Rap2 in adherent cells due to a low sensitivity to GAPs.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Plasmids. The cDNA fragment of Rap2A was amplified from a human spleen cDNA library by PCR with primers 5'-CCCTCGAGATGCGCGAGTACAAAGTGGTG-3' and 5'-TTGCGGCCGCCTATTGTATGTTACATGCAGAACA-3'. A constitutively active mutant of Rap2 was designed by analogy to Ras: Gly 12 was substituted for by Val in Rap2V12 by PCR-mediated mutagenesis. Wild-type Rap1A (Krev1) and a constitutively active mutant, Rap1V12, were obtained from M. Noda (Kyoto University, Kyoto, Japan) (30). pCEV-c-Ha-ras and pCEV-c-Ha-ras V12 were obtained from K. Kaibuchi (NAIST, Nara, Japan). The coding sequences of human H-Ras, Rap1A, and Rap2A were subcloned into pEBG (49); pCXN2-Flag (39); pCAGGS-EGFP, pCAGGS-ECFP, and pGBT9 (Clontech); and pGEX4T-3 (Amersham-Pharmacia Biotech). pCAGGS-EGFP and pCAGGS-ECFP are derivatives of pCAGGS (39) and encode enhanced green fluorescent protein (EGFP) and enhanced cyan fluorescent protein (ECFP) (Clontech), respectively. Expression plasmids of the wild type and a constitutively active mutant of c-Raf1, pSR-Raf and pSR-SKR, respectively, were obtained from S. Hattori (NCNP, Tokyo, Japan). The DNA fragment covering the Ras-binding domain (RBD) and cysteine-rich domain (CRD) of c-Raf1 was amplified by PCR with primers 5'-CTCGAGCCTTCTAGACAAGCAACACT-3' and 3'-GCGGCCGCGACTCCACTATCACCAATAGT-5' and subcloned into pGEX4T-3 and pGAD424 (Clontech) to generate pGEX-Raf-RBD+CRD and pGAD424-Raf-RBD+CRD, respectively. pGEX-RalGDS-RBD and pmt2-sm-ha Epac were obtained from J. L. Bos (Utrecht University) (12, 14). The entire coding region of Epac was amplified by PCR with primers 5'-GTCGACATGGTGTTGAGAAGGATGCACCGG-3' and 3'-GCGGCCGCTCATGGCTCCAGCTCTCGGGAG-5'. The entire coding region of mouse CalDAG-GEFI cDNA was amplified by PCR from a mouse spleen cDNA library with primers 5'-GGTCGACATGGCGAGCACTCTGGACCTGGA-3' and 5'-AGTCACAGCGTCTTATAATTGGATG-3'. cDNAs of KIAA0277 (GFR) and KIAA0313 (PDZ-GEF1) were provided by N. Nomura (Kazusa Institute, Kisarazu, Japan). cDNAs of Epac, mouse CalDAG-GEFI, GFR, and PDZ-GEF1 were subcloned into pCXN2-Flag and pGEX-4T3. pCXN2-rap1GAPII has been described previously (37). The entire coding region of rap1GAPII was subcloned into pAcSG2-His (Pharmingen) to generate pAcSG2-His-rap1GAPII. pSR-SPA-1 was obtained from N. Minato (Kyoto University) (31). pYFP-ER and pYFP-Golgi were purchased from Clontech.

Cell culture and transfection. The cell lines used in this study were 293T (obtained from B. J. Mayer, Harvard Medical School), NIH 3T3 (JCRB 0615), HT1080 (ATCC CCL121), 3Y1 (JCRB 0734), SR-3Y1 (26), HR-3Y1 (JCRB 0743), Crk-3Y1 (34), NY72-3Y1, MDCK (ATCC CCL34), and Jurkat (ATCC TIB152). NY72-3Y1 cells express a temperature-sensitive mutant of v-Src, v-Src^tsNY72-4 (36). Adherent cells were cultured in Dulbecco's modified Eagle medium (MEM) (Nissui, Tokyo) supplemented with 10% fetal calf serum (FCS). Jurkat cells were cultured in RPMI 1640 (Nissui) with 10% FCS. Expression plasmids were introduced into 293T cells by the calcium-phosphate precipitation method, into NIH 3T3 cells with Superfect (Qiagen), and into MDCK cells with FuGENE6 (Roche Molecular Biochemicals).

Cell lines expressing Rap2 and Rap1. NY72-3Y1 cells were transfected with pCXN2-Flag-Rap2WT, pCXN2-Flag-Rap2V12, pCXN2-Flag-Rap1WT, pCXN2-Flag-Rap1V12, and pCAGGS and with a hygromycin resistance gene, followed by selection in the medium containing 200 µg of hygromycin B per ml at 40°C. After 10 days, cells were transferred to an incubator maintained at 33°C for the induction of the active v-Src and cultured further for 24 h. We observed the morphology of the colonies under the microscope and scored transformed and nontransformed colonies as described previously (30).

Antibodies. Anti-Rap2 monoclonal antibody, anti-Flag M2 monoclonal antibody, and anti-Rap1 polyclonal antibody were purchased from Transduction Laboratories, Sigma, and Santa Cruz, respectively. Anti-pan-Ras monoclonal antibody was supplied by S. Hirohashi (24). Anti-rap1GAP and anti-C3G antibodies were developed in our laboratory (37, 50). Horseradish peroxidase-conjugated antiphosphotyrosine antibody, RC20, was purchased from Transduction Laboratories.

Preparation of GST-tagged proteins. Recombinant proteins fused to glutathione S-transferase (GST) were expressed in Escherichia coli from pGEX-derived vectors and purified as described previously (35).

Purification of rap1GAPII. Recombinant baculovirus carrying rap1GAPII cDNA was produced by the cotransfection of pAcSG2-His-rap1GAPII and baculo-Gold DNA (Pharmingen) according to the manufacturer's protocol. The recombinant baculovirus was inoculated into Sf9 cells cultured in TC100 (Gibco BRL) containing 10% fetal bovine serum. Forty-eight hours after infection, cells were suspended in lysis buffer, disrupted by freeze-thawing, cleared by centrifugation, and loaded onto a HiTrap chelating column (Amersham-Pharmacia). After washing with 20 mM imidazole (pH 8.0) containing 100 mM NaCl, His-tagged rap1GAPII was eluted in 200 mM imidazole (pH 8.0) containing 100 mM NaCl and dialyzed against 20 mM Tris (pH 7.5), 150 mM NaCl, 5 mM MgCl₂, and 1 mM dithiothreitol (DTT).

Analysis of guanine nucleotides bound to Rap1 and Rap2. Guanine nucleotides bound to Rap1 and Rap2 were analyzed essentially as described previously (16). Briefly, 293T cells were transfected with pEBG-Rap1 or Rap2 with or without other expression plasmids. Twenty-four hours after transfection, cells were labeled for 2 to 4 h with ³²P_i in phosphate-free MEM (GibcoBRL). GST-tagged Rap1 and Rap2 were collected on glutathione-Sepharose beads. In another experiment, 293T cells were transfected with pCXN2-Flag-Rap1 or Rap2. After ³²P_i labeling, Flag-tagged Rap1 and Rap2 were immunoprecipitated with anti-Flag M2 monoclonal antibody and protein G-Sepharose. Guanine nucleotides bound to Rap1 and Rap2 were separated by thin-layer chromatography (TLC) and quantitated with a BAS-1000 image analyzer (Fuji-Film).

Analysis of the guanine nucleotide exchange and GTP hydrolysis in vivo. 293T cells (2 × 10⁵) transfected with pEBG-Rap1 or pEBG-Rap2 were labeled with 2 MBq of ³²P_i for 30 min in phosphate-free MEM and chased with complete medium. At each time point, cells were lysed and guanine nucleotides bound to G proteins were quantitated as described above. We also measured the radiolabeling efficiency of the guanine nucleotides as follows. The labeled 293T cells were freeze thawed in 800 µl of hypotonic buffer (10 mM Tris-HCl [pH 8.0], 1.5 mM MgCl₂, 10 mM KCl, 0.1 mM DTT) and clarified by centrifugation at 10,000 × g for 15 min. Two hundred microliters of the resulting supernatant was added to equal amounts of 2× exchange buffer (40 mM Tris [pH 8], 30 mM EDTA, 2 mM DTT). After the addition of 1 µg of recombinant GST-H-Ras and incubation at 37°C for 10 min, the loading reaction was terminated by the addition of MgCl₂ at 20 mM. GST-H-Ras loaded with guanine nucleotides was collected with glutathione-Sepharose, and the ³²P-labeled guanine nucleotides were separated by TLC and quantitated. We used recombinant GST-H-Ras because it showed the highest in vitro loading efficiency among various G proteins tested by us.

Detection of GTP-bound Ras family G proteins. Detection of GTP-bound Ras-family G proteins was performed by the Bos method with slight modifications (14). Briefly, cells were lysed in lysis buffer (50 mM Tris [pH 7.5], 150 mM NaCl, 5 mM MgCl₂, 1% NP-40, 0.5% deoxycholic acid, 0.1% sodium dodecyl sulfate [SDS], 1 mM Na₃VO₄), clarified by centrifugation, and incubated with either GST-RalGDS-RBD or GST-Raf-RBD+CRD prebound to glutathione-Sepharose beads for 1 h at 4°C. Beads were washed twice with lysis buffer and resuspended in SDS-sample buffer. Proteins bound to the beads were separated by SDS-polyacrylamide gel electrophoresis (PAGE) and analyzed by immunoblotting with either anti-Rap1 or anti-Rap2 antibodies. Bound antibodies were detected by the ECL enhanced chemiluminescence system (Amersham Pharmacia) and analyzed with an LAS-1000 image analyzer (Fuji-Film). In another experiment, 293T cells were transfected with pCXN2-Flag-derived vectors encoding Rap2, Rap1, and H-Ras or with pCXN2-Flag-rap1GAPII. After 24 h, cells were lysed in lysis buffer and processed as described previously, except that anti-Flag M2 antibody was used for detection of the recombinant G proteins.

In vitro binding of Rap2 to Raf and RalGDS. GST-Rap2 was cleaved with thrombin. After the removal of GST with glutathione-Sepharose, 2.5 µg of Rap2 was loaded with either 0.5 mM GTP or 0.5 mM GDP in exchange buffer (50 mM Tris-HCl [pH 7.5], 25 mM EDTA, 0.5 mg of bovine serum albumin [BSA] per ml, 1 mM DTT) at 37°C for 20 min. The reaction was terminated by the addition of MgCl₂ at 50 mM. The nucleotide-bound Rap2 was incubated with 20 µg of GST, GST-Raf-RBD+CRD, or GST-RalGDS-RBD, prebound to glutathione-Sepharose in binding buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 5 mM MgCl2, 25 µM ZnCl₂, 1 mM DTT, 0.2% BSA, 1% Triton X-100) for 1 h at 4°C. After extensive washing with binding buffer, proteins applied or bound to the beads were analyzed by immunoblotting with anti-Rap2 antibody.

Guanine nucleotide exchange reaction in vitro. A fluorescent analogue of GDP, 2', 3'-bis(O)-(N-methylanthranolol)-GDP (mGDP), was purchased from Dojin Kagaku (Kumamoto, Japan). GST-Rap2 and GST-Rap1 were loaded with mGDP as described previously (32). The mGDP loading efficiency for Rap1A was between 80 and 90%, and for Rap2A, it was between 40 and 50%. For the measurement of GEF activity, 400 nM labeled Rap2A or Rap1A was incubated with or without 100 nM GEFs in reaction buffer (50 mM Tris-HCl [pH 7.5], 5 mM MgCl₂, 2 mM DTT) at 20°C. For activation of Epac, an analogue of cAMP, Sp-adenosine 3',5'-cyclic monophothioate triethylamine salt (Sp-cAMPS, Research Biochemical International), was included at 100 µM. The reaction was started by addition of GTP at 200 µM. The decrease in fluorescence was monitored in a JASCO FP-750 fluorescence spectrometer, with excitation and emission wavelengths of 366 and 450 nm, respectively.

In vitro GAP assay. Cells were washed in TBS (20 mM Tris [pH 7.5], 150 mM NaCl, 1 mM Na₃VO₄) and freeze-thawed in TBS-Mg (20 mM Tris [pH 7.5], 150 mM NaCl, 5 mM MgCl₂, 2 mM DTT, 1 mM Na₃VO₄). After centrifugation at 10,000 × g for 15 min, supernatants were used as a crude cytosolic fraction. GAP activity was measured in vitro as described previously (18). GST-Rap1A or GST-Rap2A (5 µM each) was loaded with [-³²P]GTP in loading buffer (20 mM Tris [pH 8], 1.5 µM GTP, 10 mM 2-mercaptoethanol, 5 mM MgCl₂, 20 mM EDTA, 10% glycerol, 0.5 mg of BSA per ml) at 30°C for 15 min, followed by the addition of MgCl₂ to 20 mM. Purified rap1GAPII or crude cell lysates were added to 250 nM GST-Rap1A or GST-Rap2A in reaction buffer (20 mM Tris, 5 mM MgCl₂, 0.5 mg of BSA per ml) at 30°C. The reaction was terminated by addition of ice-cold washing buffer (20 mM Tris-HCl [pH 8], 5 mM MgCl₂, 100 mM NaCl). Samples were adsorbed to nitrocellulose filters (S&S). The filters were washed three times with washing buffer, dried, and analyzed with a BAS-1000 image analyzer.

Fluorescence microscopy. HT1080 cells grown on fibronectin-coated coverslips were fixed with 90% ethanol. Alternatively, cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) and permeabilized with 0.2% Triton X-100 in PBS for 5 min. Cells were preincubated in PBS containing 1% BSA for 30 min and incubated with anti-Rap2 monoclonal antibody or anti-Rap2 preadsorbed to GST-Rap2, followed by incubation with Alexa 488 goat anti-mouse antibody (Molecular Probes, Leiden, The Netherlands). 293T and MDCK cells were transfected with pCAGGS-EGFP-Rap2, pCAGGS-EGFP-Rap1, or pCAGGS-EGFP-H-Ras and observed with an LSM-510 confocal microscope (Carl Zeiss). In other experiments, MDCK cells were cotransfected with pCAGGS-ECFP-Rap2 and pYFP-ER or pYFP-Golgi and observed with the confocal microscope.

Immunoelectron microscopy. Immunoelectron microscopy was performed essentially as described previously (46). HT1080 cells were harvested with a cell scraper, fixed in 0.1 M phosphate buffer (pH 7.4) containing 3% paraformaldehyde and 0.05% glutaraldehyde for 1 h at room temperature, dehydrated in ethanol, and embedded in Lowicryl K4M (Polysciences, Inc., Eppelheim, Germany) at 40°C for 24 h. After UV polymerization, blocks were trimmed and mounted in the microtome. Sections were collected on uncoated nickel grids, preincubated in PBS containing 1% BSA for 10 min at room temperature, and incubated in the same buffer containing anti-Rap2 monoclonal antibody overnight at 4°C. After being washed in PBS, the sections were further incubated with goat anti-mouse antibody conjugated with 10-nm-diameter colloidal gold (British Biocell, Cardiff, Wales) followed by counterstaining with 1% osmium tetroxide and with 2% uranyl acetate and Millonig's lead acetate. Sections were examined under a JEOL JEM-1200-EX transmission electron microscope.

Yeast two-hybrid analysis. Yeast strain y-190 was double transformed with pGBT9-derived vectors and pGAD424-derived vectors. Transformants were grown on plates containing y-NB-Ura-Leu-Trp. Grown colonies were transferred to nitrocellulose filters and examined for -galactosidase activity.

Reporter assay. Activation of the Elk1 transcription factor was assayed by use of a PathDetect kit (Stratagene). Briefly, 2 × 10⁵ 293T cells were transfected with 1 µg of pFR-Luc reporter plasmid, 0.1 µg of pFA-Elk1 encoding the activator domain of the Elk1 transcription factor, 0.5 µg of pCMV-gal, and 0.5 to 1 µg of test plasmids. After 24 h, cells were lysed in lysis buffer, and luciferase activity was measured with the Promega Luciferase Assay System. The activity of -galactosidase was measured for normalization of the transfection efficiency.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

High GTP/GDP ratio on Rap2. As an initial characterization of Rap2, the GTP/GDP ratio on Rap2 was determined. We labeled NIH 3T3 cells and 293T cells expressing either Flag- or GST-tagged Rap2 protein with ³²P_i and quantitated by TLC the guanine nucleotides bound to Rap2 (Fig. 1A). In contrast to Rap1, 5 to 19% of which existed as GTP-bound form in repeated experiments, between 50 and 70% of the labeled Rap2 protein was in the GTP-bound form. Both the Flag and GST tags yielded similar results. However, it was still possible that the high GTP/GDP ratio on Rap2 might result from tagging of the proteins at the N terminus. Thus, we examined the effect of the overexpression of GEFs for Rap1, which, as we shall describe later, also promote the guanine nucleotide exchange of Rap2. The endogenous GTP-bound Rap2 was recovered from the cell lysates by the use of GST-RalGDS and quantitated by the Bos method (14). As shown in Fig. 1B, the increase in GTP-bound Rap2 did not exceed 30% of the control level after the expression of C3G, CalDAG-GEFI, or Epac/cAMP-GEFI. In contrast, expression of these GEFs increased GTP-bound Rap1 from four- to sevenfold. The transfection efficiency exceeded at least 50% when it was monitored with the GFP expression vector. Thus, this result supports the idea that the basal level of endogenous GTP-Rap2 was already high before the overexpression of GEFs.

View larger version (52K): [in this window] [in a new window]   FIG. 1.   Analysis of guanine nucleotides bound to Rap2. (A) 293T cells or NIH 3T3 cells were transfected with expression vectors encoding the proteins listed at the top and labeled with 32Pi. Guanine nucleotides bound to the expressed Rap1 or Rap2 were separated by TLC. The radioactivity of GTP and GDP was quantitated, and the percentage of GTP [GTP/(GDP + GTP)] is shown at the bottom. (B) 293T cells were transfected with Flag-tagged expression vectors as indicated at the top. In one sample, cells were incubated with 100 µM Sp-cAMPs for 10 min before harvest (indicated as cAMP). Endogenous GTP-Rap2 and GTP-Rap1 were detected by the Bos method as described in the text (top). The intensity of the bands was quantitated by an image analyzer, and the fold increase is shown. The lower panels show immunoblotting (IB) of total cell lysates with anti-Rap2, anti-Rap1, or anti-Flag antibody.

GEFs for Rap2. To understand the mechanism by which Rap2 retains a high GTP/GDP ratio, we examined whether GEFs for Rap1 also stimulated guanine nucleotide exchange of Rap2. 293T cells expressing GEFs were labeled with ³²P_i, and the guanine nucleotides bound to Rap1 or Rap2 were quantitated by TLC (Fig. 2A). The amount of GTP-bound Rap2 was increased by the coexpression of GEFs for Rap1. Epac/cAMP-GEFI in the presence of cAMP analog and PDZ-GEF1 showed the highest activity, followed by GFR, CalDAG-GEFI, and C3G. mSos, Ras-GRF, and Ras-GRP did not increase the amount of GTP-bound Rap2 (Y. Ohba and M. Matsuda, unpublished data).

View larger version (51K): [in this window] [in a new window]   FIG. 2.   GEF-dependent guanine nucleotide exchange reaction of Rap2. (A) 293T cells expressing GST-Rap2 or GST-Rap1 with GEFs indicated at the top were labeled with 32Pi, and guanine nucleotides bound to GST-Rap2 or GST-Rap1 were analyzed as described. In four samples, cells were incubated with 100 µM Sp-cAMPS for 10 min before harvest (indicated as cAMP). (B) Rap1 or Rap2 bound to mGDP was incubated at 20°C with or without GEFs. Epac was stimulated with 100 µM Sp-cAMPS (indicated as cAMP). The decrease in fluorescence emission at 450 nm was monitored as a function of time.

Guanine nucleotide exchange rate in the cells. Because the GTP/GDP ratio on Ras family G proteins is determined primarily by GEFs and GAPs, the high GTP/GDP ratio on Rap2 may occur because of either a high guanine nucleotide exchange rate or low GTPase activity in the cells. We first analyzed the turnover rate of guanine nucleotides on Rap1 and Rap2 in the cells. For this purpose, cells were labeled with ³²P_i for 30 min and chased up to 3 h. The labeling efficiency of the cytosolic guanine nucleotides and the radioactivity of the ³²P_i-labeled guanine nucleotides on Rap2 and Rap1 were quantitated at each time point and plotted (Fig. 3A). Because cytosolic ³²P-labeled guanine nucleotides decreased faster than ³²P-labeled guanine nucleotides bound to Rap1 or Rap2, the guanine nucleotide exchange reaction, but not the synthesis of the guanine nucleotides, was the limiting step in the loading of the radiolabeled guanine nucleotides to Rap1 or Rap2. In this condition, the levels of radiolabeled guanine nucleotides on Rap1 and Rap2 decreased with a similar time course. Thus, the velocity of the guanine nucleotide exchange reaction does not account for the high GTP/GDP ratio on Rap2. When we plotted the percentage of GTP on Rap1 or Rap2, we found that GTP bound to Rap2 decreased remarkably slower than GTP bound to Rap1 (Fig. 3B). This result demonstrates that low GTPase activity of Rap2 is the principal cause of the high GTP/GDP ratio in the cells.

View larger version (22K): [in this window] [in a new window]   FIG. 3.   Guanine nucleotide exchange and GTP hydrolysis of Rap1 and Rap2 in vivo. 293T cells were labeled with 32Pi for 30 min and chased with phosphate-containing medium. Cells were harvested at the indicated time points, and the cytosolic fraction was used to load recombinant H-Ras protein with 32P-labeled guanine nucleotides in vitro. The cytosolic guanine nucleotides bound to H-Ras in vitro were separated by TLC and quantitated. In a parallel experiment, 293T cells expressing GST-Rap1 or GST-Rap2 were lysed, and the labeled guanine nucleotides bound to GST-Rap1 or GST-Rap2 were separated by TLC and quantitated. (A) The sum of radioactivity of GTP and GDP at each time point was plotted at a ratio to the radioactivity at 30 min. (B) The percentage of GTP on GST-Rap2 or GST-Rap1 at each time point was plotted. Bars indicate standard deviations.

Stimulation of Rap2 GTPase by GAPs. We next examined whether GAPs for Rap1 stimulated Rap2 GTPase in the cells (Fig. 4A). Both rap1GAPII and SPA-1 stimulated Rap2 GTPase, albeit less efficiently than they did Rap1. The sensitivity of Rap2 and Rap1 to rap1GAPII was further compared in vitro (Fig. 4B). rap1GAPII stimulated GTPase activity of both Rap1 and Rap2 in a dose- and time-dependent manner; however, Rap1 was significantly more sensitive than Rap2. We next used the cytosolic fractions of Jurkat, HT1080, and 293T cells as a source of GAP, because the GAP activity of Rap1 and Rap2 exists mostly in the cytosolic fraction (22). As shown in Fig. 4C, we could not detect GAP activity toward Rap2 in the cytosolic fraction of the cells, whereas, under the same conditions, Rap1 GTPase was stimulated. These results support the proposal that the low GTPase activity of Rap2 causes a high GTP/GDP ratio on Rap2.

View larger version (36K): [in this window] [in a new window]   FIG. 4.   Activation of Rap2 GTPase by GAPs. (A) 293T cells expressing either GST-Rap2 or GST-Rap1 and GEFs and GAPs indicated at the top were labeled with 32Pi, and guanine nucleotides bound to GST-Rap1 or GST-Rap2 were separated by TLC. (B) Rap1 and Rap2 were loaded with [-32P]GTP and incubated with the indicated amounts of rap1GAPII for 10 min (left panel). Similarly, the 32P-labeled Rap1 and Rap2 were incubated with (solid symbols) or without (open symbols) 1 µg of rap1GAPII for the indicated periods (right panel). The radioactivity remaining on Rap1 or Rap2 was quantitated and plotted. (C) [-32P]GTP-loaded Rap1 and Rap2 were incubated with buffer alone, purified rap1GAPII, or cell lysates of Jurkat cells, HT1080 cells, or 293T cells for 20 min. The radioactivity remaining on Rap1 or Rap2 was quantitated and plotted. Bars indicate standard errors.

Subcellular localization of Rap2. For understanding the function of Rap2, its subcellular localization was determined by immunomicroscopy, GFP tagging, and immunoelectron microscopy. The specificity of the anti-Rap2 antibody was examined by immunoblotting (Fig. 5A). The Rap2 protein was detected as a major protein at about 22 kDa both in HT1080 and in 293T cells, although multiple faint bands were also detected in the higher-molecular-mass range. Notably, only the 22-kDa band had disappeared after preadsorption of the antibody to the recombinant Rap2. In addition, the proteins in the higher-molecular-mass range were also detected when we omitted the primary antibody. This result indicates that the anti-Rap2 antibody used in this study was specific to Rap2 and that the proteins detected in the higher-molecular-mass region were due to nonspecific binding by the secondary antibody.

View larger version (37K): [in this window] [in a new window]   FIG. 5.   Subcellular localization of Rap2. (A) Total cell lysates of HT1080 cells and 293T cells were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes. Filters were incubated with anti-Rap2 antibody, anti-Rap2 antibody preincubated with GST-Rap2, or buffer alone. Filters were further incubated with a peroxidase-conjugated antimouse antibody, which was detected by the ECL enhanced chemiluminescence system. (B) HT1080 cells grown on fibronectin-coated coverslips were fixed with paraformaldehyde and permeabilized by Triton X-100 (PFA) or fixed with ethanol (EtOH). The cells were incubated with anti-Rap2 antibody or anti-Rap2 antibody preadsorbed by GST-Rap2. Bound antibodies are detected by the use of antimouse Alexa 488-conjugated antibody. (C) 293T and MDCK cells grown on poly-L-lysine-coated glass dishes were transfected with expression vectors of EGFP-tagged Rap2, Rap1, and H-Ras for 24 h and observed under a confocal microscope. (D) MDCK cells cultured on poly-L-lysine-coated glass dishes were transfected with expression vectors encoding ECFP-Rap2 and EYFP-ER or EYFP-Golgi for 24 h and observed with a confocal microscope. (E) HT1080 cells were collected and embedded in Lowicryl. Ultrathin sections were prepared and incubated with anti-Rap2 antibody or buffer alone (), followed by incubation with antimouse antibody conjugated with immunogold. Cells were arbitrarily divided into plasma membrane (PM), ER, mitochondria (Mi), cytoplasm (Cy), and nucleus (Nu). Immunogold particles in these regions were counted and plotted.

Binding of Rap2 to Raf and RalGDS. Many of the Ras effector proteins, including Raf and RalGDS, also bind to Rap1 (4). We examined whether Rap2 also binds to Raf and RalGDS by using a yeast two-hybrid assay and a pull-down assay. In the yeast two-hybrid assay, both the wild-type and the constitutively active Rap2 associated with the RBDs of Raf and RalGDS (Fig. 6A), as did the wild-type and constitutively active Ras or Rap1. In the pull-down assay, the active forms of Ras and Rap1 preferentially bound to the RBDs of Raf and RalGDS. In contrast, both the wild-type and the constitutively active form of Rap2 bound to the RBDs of Raf and RalGDS with similar efficiency (Fig. 6B). To confirm the GTP dependency in the binding of Rap2 to Raf and RalGDS, we performed an in vitro binding assay. As shown in Fig. 6C, GTP-Rap2 bound to Raf and RalGDS. GDP-Rap2 did not bind to RalGDS; however, it bound to Raf-RBD+CRD, although less efficiently than did GTP-Rap2. This weak binding may be ascribable to the CRD, because the CRD of Raf binds to Ras and Rap1 in a GTP-independent manner (19, 20). We further confirmed the GTP-dependent binding in 293T cells expressing rap1GAPII. As shown in Fig. 6D, the binding of Rap2 to Raf and RalGDS was significantly reduced by the expression of rap1GAPII, indicating that Rap2 bound to the effector in a GTP-dependent manner.

View larger version (37K): [in this window] [in a new window]   FIG. 6.   Binding of Rap2 to Raf and RalGDS. (A) Yeast strain y-190 was cotransformed with the pGAD424-derived plasmids denoted at the top and with the pGBT9-derived bait plasmids listed to the left. His activity of the transformants was assayed on histidine-deficient plates, followed by a -galactosidase assay. (B) 293T cells expressing the Flag-tagged G proteins denoted at the top were lysed in lysis buffer and incubated with GST alone, GST-Raf-RBD+CRD, or GST-RalGDS-RBD. Proteins bound to these GST fusion proteins and the total cell lysates were separated by SDS-PAGE and probed with anti-Flag monoclonal antibody. (C) GST-Raf-RBD+CRD, GST-RalGDS-RBD, or GST alone was incubated with buffer alone () or recombinant Rap2 protein loaded with GDP (D) or GTP (T). Proteins bound to the beads were separated by SDS-PAGE and analyzed by immunoblotting (IB) with anti-Rap2 antibody. (D) 293T cells expressing Flag-Rap2WT alone or Flag-Rap2WT and rap1GAPII were harvested and incubated with GST-Raf-RBD+CRD or GST-RalGDS-RBD. Proteins bound to the GST fusion proteins and total cell lysates were analyzed by immunoblotting with anti-Flag antibody.

Rap2 inhibition of Ras-dependent transcriptional activity. To examine whether Rap2 interferes with Ras signaling, as does Rap1, we monitored the Ras-dependent transcriptional activation of Elk1, which is a direct downstream target of ERK/MAPK. As shown in Fig. 7, both the wild-type and constitutively active Rap2 inhibited the Ras-dependent transcriptional activation of Elk1, as did the constitutively active Rap1. In contrast, Elk1 activation by the constitutively active Raf was not abolished by the expression of either form of Rap2. This result shows that overexpression of Rap2 inhibits Ras-dependent Elk1 activation, probably by inhibiting Raf activation, as does Rap1 (10, 19).

View larger version (33K): [in this window] [in a new window]   FIG. 7.   Inhibition of Ras-dependent transcription by Rap2. 293T cells were transfected with pFR-luc, pFA-Elk1, and pCXN2-Flag-RasV12 (H-RasV12; solid columns) or pSR-SKR (Raf-SKR; shaded columns) in combination with wild-type (WT) and active Rap2 or Rap1 expression vectors. After 24 h, luciferase activity was measured. In this assay system, expression of luciferase is driven by the Elk1 transcription factor. Mean values obtained from three experiments are shown with standard errors.

Determination of the numbers of Rap2, Rap1, and Ras molecules in a single cell. If Rap2 antagonizes Ras signaling by competitive binding to Ras effectors, the numbers of these Ras family molecules in a cell must be critical in the physiologic milieu. Thus, we determined the quantity of Rap2, Rap1, and Ras by using recombinant Rap2, Rap1, and H-Ras as standards (Fig. 8). Thirty micrograms of proteins from 7.4 × 10⁴ 293T cells and 20 µg of proteins from 9.2 × 10⁴ HT1080 cells were applied to SDS-polyacrylamide gel and analyzed by immunoblotting. The anti-Rap2 antibody used here recognizes both Rap2A and Rap2B. Similarly, the anti-Rap1 antibody recognizes both Rap1A and Rap1B, and anti-Ras antibody reacts with H-, K-, and N-Ras proteins. The calculated numbers of Rap2, Rap1, and Ras molecules in a 293T cell were 4.6 × 10⁶, 1.5 × 10⁶, and 5.7 × 10⁵, respectively; those in an HT1080 cell were 7.8 × 10⁶, 1.1 × 10⁶, and 5.6 × 10⁵, respectively. Thus, Rap2 exceeds Ras and Rap1 in number in 293T and HT1080 cells.

View larger version (39K): [in this window] [in a new window]   FIG. 8.   Quantitative immunoblotting (IB) of Ras, Rap1, and Rap2. Thirty micrograms of proteins from 7.4 × 104 293T cells and 20 µg of proteins from 9.2 × 104 HT1080 cells and recombinant Rap2, Rap1, or H-Ras, the amounts of which are indicated at the bottom of each column, were separated by SDS-PAGE and analyzed by immunoblotting with the antibodies indicated to the right. With recombinant proteins used as a standard, the amounts of Rap2, Rap1, and H-Ras in the cell lysates were determined and are shown at the bottom of each lane of cell lysates.

Decreased level of GTP-Rap2 in transformed fibroblasts. The high basal level of GTP-Rap2 evoked the question of whether the level of GTP on Rap2 varies under physiologic conditions. Because Rap2 antagonized Ras-dependent transcription, we speculated that the level of GTP-Rap2 might be low in transformed cells. Compared to the level in parental rat 3Y1 fibroblasts, the level of GTP-Rap2 was significantly lower in three cell lines transformed by v-Src (SR-3Y1), v-Ras (HR-3Y1), or v-Crk (Crk-3Y1) (Fig. 9A). To confirm that the level of GTP-Rap2 decreases by cellular transformation, we used 3Y1 cells expressing a temperature-sensitive mutant of v-Src, v-Src^tsNY72-4 (Fig. 9B). The amount of GTP-Rap2 in NY72-3Y1 cells at a permissive temperature, 33°C, was similar to that in SR-3Y1 cells expressing wild-type v-Src. However, as expected, the amount of GTP-Rap2 increased when NY72-3Y1 cells were incubated at a nonpermissive temperature, 40°C.

View larger version (42K): [in this window] [in a new window]   FIG. 9.   Level of GTP-Rap2 in transformed cells. (A) 3Y1 cells and 3Y1-derived transformed cells denoted at the top were lysed, and 500 µg each of the lysates was incubated with GST-RalGDS prebound to glutathione-Sepharose beads. Proteins bound to the beads and 10 µg of total cell lysates were analyzed by SDS-PAGE and immunoblotting with anti-Rap2 antibody. (B) NY72-3Y1 cells and SR-3Y1 cells expressing a temperature-sensitive mutant and wild-type v-Src, respectively, were maintained at 33°C. After a temperature shift to 40°C, cells were lysed at the times indicated and analyzed as in panel A. (C) Soluble cytosolic fraction was prepared from NY72-3Y1 cell cultures at either 40 or 33°C. [-32P]GTP-loaded Rap2 and Rap1 were incubated with buffer alone, purified rap1GAPII, or 30 µg of the soluble cytosolic fractions for 20 min. Radioactivity retained by Rap2 and Rap1 was quantitated and plotted. Bars indicated standard errors from three samples. (D) Cell lysates used in panel C were separated by SDS-PAGE and blotted with anti-rap1GAP (top). The levels of GTP-Rap2 and GTP-Rap1 in these cell lysates were analyzed as in panel A.

Inhibition of Src transformation by Rap2. Finally, we examined whether GTPase-deficient Rap2 could inhibit morphological transformation by v-Src. For this purpose, we introduced wild-type and GTPase-deficient Rap1 or Rap2 into NY72-3Y1 cells and examined the temperature-dependent transformation of the cells. We did not find a remarkable difference in the morphology of the transfected cells at a nonpermissive temperature (Fig. 10A). However, at a permissive temperature, expression of Rap1 or Rap2 significantly inhibited the morphological transformation of NY72-3Y1 cells by v-Src. We counted the number of transformed or nontransformed colonies at a permissive temperature and found that wild-type and GTPase-deficient Rap2 inhibited transformation at similar levels (Fig. 10B). Thus, the transformation can be blocked by high GTP-Rap2 levels.

View larger version (71K): [in this window] [in a new window]   FIG. 10.   Inhibition of morphological transformation by Rap2. NY72-3Y1 cells were transfected with pCAGGS or pCAGGS-derived Flag-tag expression vector for wild-type Rap2 (Rap2WT), Rap2V12, Rap1WT, or Rap1V12 with an expression vector of the hygromycin resistance gene. After selection with hygromycin B at 40°C for 10 days, cells were cultured at 33°C overnight. (A) Morphology of the representative colonies at 40 and 33°C. (B) Hygromycin-resistant colonies consisting of transformed spindle cells or nontransformed flat cells were scored under the microscope. Mean values obtained from two independent experiments are shown with standard deviations. (C) Equal amounts of cell lysates were analyzed by immunoblotting (IB) with anti-Flag antibody.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The basal level of the GTP-bound form of the Ras family G proteins is less than 20% in many cell types (14-16, 42, 48). This low basal level of the GTP-bound form enables the Ras family G proteins to transduce signals only when the proper stimuli activate GEFs (29). We found that the basal GTP/GDP ratio of Rap2 is unexpectedly high and that this results from the low GTPase activity of Rap2. The example closest to Rap2 may be RhoE, which does not have detectable GTPase activity and remains mostly in the GTP-bound form (13, 17). It has been proposed that RhoE acts to inhibit signaling downstream of RhoA, altering some RhoA-regulated responses (17). Rap2 may play a similar inhibitory role in the Ras-signaling cascade for the following reason. First, previous reports have demonstrated that the effector molecules are mostly shared between Rap2 and Ras (21, 38), and we showed that the wild-type Rap2 binds to Raf as efficiently as the GTPase-deficient Rap2 mutant in the cells. Second, expression of Rap2 antagonized Ras-dependent transcriptional activity, as did expression of Rap1. Third, at least half of Rap2 localized at the plasma membrane, where most of Ras localized. Fourth, the level of GTP-Rap2 decreased in a manner dependent on cellular transformation, in which Ras plays a pivotal role. Finally, overexpression of Rap2 inhibited morphological transformation by v-Src.

Most GEFs for Rap1 that have been reported thus far, Epac/cAMP-GEF, CalDAG-GEFI, PDZ-GEF1, and GFR, have been shown to promote guanine nucleotide exchange of Rap2 in cells. Similar to GEFs, SPA-1 and rap1GAP activate Rap2 GTPase, although less efficiently than Rap1 GTPase. Moreover, most of the effectors are also shared between Rap1 and Rap2 (21, 38). This co-usage of GEFs, GAPs, and effectors between Rap1 and Rap2 implies that Rap1 and Rap2 function in the same signaling cascade. The conspicuous difference, however, is the low sensitivity of Rap2 to GAPs, which results in the long half-life of the GTP-bound form and high GTP/GDP ratio in the cells. These characteristics render the signaling from Rap2 to effectors weak (less than twofold compared to the basal level) and prolonged, whereas Rap1 transduces strong (several-fold) and transient signals. Thus, in the Rap signaling cascade, Rap2 may determine the basal level, and Rap1 transduces signals rapidly and transiently in response to external stimuli.

It has been proposed that, unlike Ras, Rap1 and Rap2 are localized mainly in the intracellular membrane compartments, such as the ER and Golgi apparatus (2, 3). These previous studies utilized a sucrose density gradient and indirect immunofluorescence for the characterization. A low resolution of the sucrose density gradient cannot exclude the localization of Rap1 and Rap2 in the plasma membrane. In addition, as we showed in this study, fixatives significantly affect the staining pattern of Rap2 in indirect immunofluorescence. We used indirect immunofluorescence, GFP tagging, and immunoelectron microscopy to determine the subcellular localization of Rap2 and found that Rap2 is both at the plasma membrane and in the intracellular compartments. This observation provides a basis for the Rap2 inhibition of Ras-induced Raf activation at the plasma membrane.

A recent report demonstrated that sequences unique to each Ras family protein preceding the CAAX box affect the efficiency of translocation from the ER or Golgi apparatus to the plasma membrane (9), providing a reason why the distribution of each Ras family G protein differs in the cells. However, it is not well established whether this difference in the C-terminal amino acid sequence plays a critical role in the signaling from Ras, Rap1, and Rap2. A study using chimeras between H-Ras and Rap1 demonstrated that the effector domain, but not the carboxyl-terminal region including the CAAX box, determines the antioncogenic activity of Rap1 (56). Moreover, recent reports that B-Raf is activated by both Ras and Rap1 suggest that the difference in the subcellular localizations of Ras and Rap1 may not be critical in the activation of downstream effectors, such as B-Raf (41, 52, 54). In contrast to these reports, Matsubara et al. have shown that the localization of Rap1 overexpressed at the perinuclear region causes its inability to activate Ral via RalGDS, because Ral is localized mostly in the plasma membrane (33). It is noteworthy that, in native cells, the site of GEF activation is regulated. Although we demonstrated that Rap2 is localized both on the plasma membrane and within the cells, the site of Rap2 activation by GEFs remains unknown. A method to detect the GTP-bound Rap2 in the cells is awaited to solve this question.

In several adherent cells tested, we could not detect any remarkable increase or decrease in GTP-Rap2 upon various types of stimulation that increase GTP-Rap1, such as serum, lysophosphatidic acid, and 12-O-tetradecanoylphorbol 13-acetate (TPA). Reedquist and Bos reported that GTP-Rap2 increased rapidly when human T cells were stimulated with TPA (44). Preliminary data in our laboratory with Jurkat cells have suggested that the amount of GTP-Rap2 in Jurkat cells is significantly lower than that in fibroblasts (Y. Ohba and M. Matsuda, unpublished data). These results may be interpreted as showing that the GTP/GDP ratio on Rap2 is low in T cells. Recently, it has been shown that Rap1 activates integrin (25, 45). Attachment of fibroblasts to fibronectin-coated dishes induces a rapid and transient increase in GTP-Rap1 (43). Thus, Rap1 may be required in the initial step of cell adhesion, which is then maintained by GTP-Rap2. This hypothesis is supported by our observation that the v-Src-induced transformation was inhibited by the overexpression of Rap2, because a hallmark of morphologic transformation of cells is loose attachment to the substrate.

In conclusion, Rap2 and Rap1 share GEFs, GAPs, and effectors in common, indicating that both are components of the same signaling cascade. The difference in the sensitivity to GAPs suggests that, in this Rap signaling cascade, Rap1 and Rap2 function as the fast and slow molecular switches, respectively.