Archives of Biochemistry and Biophysics Vol. 371, No. 2, November 15, pp. 207–219, 1999 Article ID abbi.1999.1448, available online at http://www.idealibrary.com on
Biochemical Characterization of the Ras-Related GTPases Rit and Rin 1 Haipeng Shao, Keiko Kadono-Okuda, 2 Brian S. Finlin, and Douglas A. Andres 3 Department of Biochemistry, University of Kentucky College of Medicine, Lexington, Kentucky 40536-0084
Received May 10, 1999, and in revised form August 20, 1999
We report the biochemical characterization of Rit and Rin, two members of the Ras superfamily identified by expression cloning. Recombinant Rit and Rin bind GTP and exhibit intrinsic GTPase activity. Conversion of Gln to Leu at position 79 (for Rit) or 78 (for Rin) (equivalent to position 61 in Ras) resulted in a complete loss of GTPase activity. Surprisingly, significant differences were found when the guanine nucleotide dissociation constants of Rit and Rin were compared with the majority of Ras-related GTPases. Both proteins display higher k off values for GTP than GDP in the presence of 10 mM Mg 21. These GTP dissociation rates are 5- to 10-fold faster than most Ras-like GTPases. Despite these unique biochemical properties, our data support the notion that both Rit and Rin function as nucleotide-dependent molecular switches. To begin to address whether these proteins act as regulators of distinct signaling pathways, we examined their interaction with a series of known Ras-binding proteins by yeast two-hybrid analysis. Although Rit, Rin, and Ras have highly related effector domain sequences, Rit and Rin were found to interact with the known Ras binding proteins RalGDS, Rlf, and AF-6/ Canoe but not with the Raf kinases, RIN1, or the p110 subunit of phosphatidylinositol 3-kinase. These interactions were GTP and effector domain dependent and suggest that RalGDS, Rlf, and AF-6 are Rit and Rin effectors. Their biochemical properties and interaction with a subset of known Ras effector proteins sugThe nucleotide sequences for human Rit and mouse Rin have been submitted to the Genebank Data Bank with Accession Nos. AF084462 and AF084463. 1 This work was supported in part by National Institutes of Health Grant EY11231. 2 Present address: National Institute of Sericultural and Entomological Science 1-2, Ohwashi, Tsukuba, 305 Japan. 3 To whom correspondence should be addressed at Department of Biochemistry, College of Medicine, University of Kentucky, 800 Rose St., Lexington, KY 40536-0084. Fax: 606-323-1037. E-mail:
[email protected]. 0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
gest that Rit and Rin may play important roles in the regulation of signaling pathways and cellular processes distinct from those controlled by Ras. © 1999 Academic Press
Key Words: Ras; GTP-binding protein; GTPase.
The Ras superfamily of low-molecular-weight GTPbinding proteins constitutes a large family of regulatory molecules (1, 2). To date, six subfamilies of the Ras superfamily have been identified: Ras, Rho, Rab, Ran, ARF, and the Rem/Rad/Gem proteins (3). These broad subfamilies are related by primary sequence relationships but also by regulation of common cellular activities, including cell growth (Ras), cytoskeletal organization (Rho), nucleocytoplasmic transport (Ran), or vesicular transport (Rab and ARF). All guanosine triphosphate phospatases (GTPases) 4 of the Ras superfamily contain five conserved amino acid motifs (G1– G5) involved in guanine nucleotide recognition and binding as detailed by extensive mutagenesis and structural studies. The Ras-related GTPases share the ability to cycle between an active GTP-bound and inactive GDP-bound structural state. They respond to external stimuli by exchanging GTP for bound GDP, thereby triggering intracellular signaling cascades through their interaction with a variety of protein effectors (1, 4). Their activity is thus regulated by their intrinsic rates of 4 Abbreviations used: GST, glutathione S-transferase; GAP, GTPase activating protein; GDS, guanine nucleotide dissociation stimulator; GTPase, guanosine triphosphate phosphatase; PI-3 kinase, phosphatidylinositol 3-kinase; RID, Ras interaction domain; GTPgS, guanosine 59-3-O-(thio)triphosphate; DTT, dithiothreitol; HA, influenza hemagglutinin epitope; BSA, bovine serum albumin; FTase, protein farnesyltransferase; FPP, farnesyl diphosphate; AMP, ampicillin; Kan, kanamycin; IPTG, isopropyl-l-thio-b-D-galactopyranosidase; PMSF, phenylmethylsulfonyl fluoride; BSA, bovine serum albumin; EST, expressed sequence tag.
207
208
SHAO ET AL.
GTP hydrolysis and GTP/GDP exchange. These rates are normally quite low, but can be increased considerably by their interaction with regulatory proteins. The signal is terminated when bound GTP is hydrolyzed to GDP in a reaction which is stimulated by GTPase activating proteins (GAPs) (2). Additional regulatory proteins, including factors which stimulate guanine nucleotide exchange (GDSs), serve to further modulate the GTPase cycle. In most cases, membrane association of Ras-related proteins is essential to their biological activity (5). In addition, most Ras family members share conserved COOH-terminal cysteine-rich motifs needed for covalent modification by isoprenoid lipids (prenylation) (6). Prenylation is the initial step in the attachment of these proteins to the cytoplasmic leaflets of a variety of cellular organelles. However, specific membrane localization often requires additional targeting signals, provided either by a cluster of basic amino acids or the palmitoylation of internal cysteine residues (7). Mutations which eliminate prenylation as well as inhibitors of cellular prenyltransferase enzymes, block the transforming activity of Ras (8). To date, more than 60 unique cDNAs belonging to the Ras-related GTPases superfamily have been identified (3). A number of studies have implicated Rasregulated signal transduction pathways in the control of retinal development and function (9). Based on these studies and our interest in characterizing the role of small Ras-related proteins in retinal signaling, we developed an expression cloning strategy designed to isolate novel Ras-related proteins from human retina. We report here the characterization of two such proteins, Rit and Rin, which on the basis of structural criteria and biochemical properties are members of the Ras subfamily of GTPases. In addition to the absence of known lipidation signals, these proteins displayed a series of unique characteristics not found in other Ras proteins, including a novel effector (G2) domain. The biochemical properties, particularly the kinetics of guanine nucleotide dissociation and GTPase activity, of the purified recombinant proteins were determined and are consistent with the notion that the Rit/Rin proteins undergo a cycle of regulated nucleotide exchange and GTP hydrolysis. Yeast two-hybrid analysis found both proteins to interact in a GTP- and effector domain-dependent manner with the previously identified Ras effector proteins RalGDS, Rlf, and AF6, but importantly not with a series of additional Ras effectors including the Raf kinases. These studies suggest that Rit and Rin constitute a new subfamily of the Ras-related GTPases and likely are involved in regulating distinct signal transduction pathways in both general and neuronal tissues.
EXPERIMENTAL PROCEDURES General methods. Standard molecular biological techniques were used (10). cDNA clones were subcloned to plasmid pBluescript II vectors (Stratagene) and sequenced by the dideoxy chain termination method (10) using gene specific primers. Nick-translated probes were synthesized using a labeling kit (GIBCO-BRL). Human retina cDNA library. The vector pGEX-KG N/K was constructed by inserting a 1.8-kb stuffer fragment into pGEX-KG (11), generating a unique 59 NotI and 39 KpnI restriction site in the polylinker of the vector. In addition, a kanamycin resistance cassette was introduced to a unique PstI site in the plasmid. Human retina cDNA, isolated by NotI/KpnI digest from a human fetal retinal Uni-ZAP XR cDNA library (Stratagene) which had been converted to pBluescript plasmid using helper phage excision, was ligated into NotI/KpnI digested pGEX-KG N/K. Ligation reactions were introduced to ElectroMAX DH10B bacteria (GibcoBRL) by electroporation using the Bio-Rad Gene Pulser protocol. Expression cloning of Rit. Expression cloning by farnesyl radiolabeling was done as described previously (12, 13) with minor modifications. Approximately 0.5–1 3 10 4 Escherichia coli strain DH10B transformants (primary colonies of the pGEX-KG N/K human fetal retina library ligation, see above) were spread on 82-mm-diameter Magna-Lift nitrocellulose filters (Micron Separations Inc.), which were overlaid on LB plates containing 100 mg/ml ampicillin (Amp) and 50 mg/ml kanamycin (Kan) and incubated overnight at 37°C. A single replica was generated on Nytran 1 filters (Schleicher and Schuell), and fusion protein synthesis was initiated by placing the original filter onto Whatman No. 1 filter paper saturated with LB containing 100 mg/ml Amp, 50 mg/ml Kan, and 1 mM isopropyl-1thio-b-D-galactopyranosidase (IPTG) for 6 h at 37°C. Replica filters were placed on fresh LB (Amp/Kan) plates and allowed to recover for 6 h at 37°C before being placed on Whatman filters soaked in LB (Amp/Kan) containing 10% glycerol, stored at 280°C, and later thawed to recover putative positives clones. Lysis of bacterial colonies was accomplished by incubating filters sequentially in the following solutions: A (150 mM NaCl, 100 mM Tris, pH 8.0, 5 mM MgCl 2, 2 mg/ml DNase I, 50 mg/ml lysozyme) for 20 min, B (150 mM NaCl, 100 mM NaOH, 0.1% SDS) for 5 min, and C (150 mM NaCl, 100 mM Tris–HCl, pH 6.5) for 5 min. Filters were dried and stored at room temperature for up to 1 week before proceeding to prenylation assays. To begin farnesyl radiolabeling, filters were placed in blocking buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 20 mg/ml bovine serum albumin, and 0.05% Tween 20) and incubated for 1 h with gentle shaking at 25°C (10 ml/filter, four buffer changes). The filters were then rinsed, the bound proteins were subjected to [ 3H]FPP labeling using recombinant protein farnesyltransferase (FTase), and unincorporated label was washed away. These steps were carried out at 30°C with vigorous shaking (to ensure that all filters were wetted) in the following solutions: (i) TBS (50 mM Tris, pH 8.0, 150 mM NaCl), 10 ml/82 mM filter, 4 3 5-min washes; (ii) prenylation buffer (50 mM Hepes, pH 7.6, 20 mM MgCl 2, 5 mM DTT, and 5 mM ZnCl 2), 10 ml/filter, 10 min; (iii) prenylation reaction mixture (prenylation buffer containing 0.27 mM [ 3H]farnesyl pyrophosphate (30 –50 Ci/ mmol, American Radiochemical) and 15 ng/ml recombinant FTase (12), 1 ml/filter, 1 h; (iv) TBS, 10 ml/filter, 6 3 5 min; (v) 50% EtOH, 10 ml/filter, 3 3 5 min; (vi) EtOH/HCl (9/1, vol/vol), 10 ml/filter, 5 min; and (vi) 100% EtOH, 10 ml/filter, 3 3 5 min. Filters were then air-dried, dipped in Amplify (Amersham Corp.) fluorographic reagent, placed on plastic backing, dried 1 h at 50°C, and exposed with an intensifying screen to Kodak X-Omat AR film at 270°C for 7–21 days. Cells from areas of filters where positive signals were obtained were recovered and rescreened by dilution cloning as described previously (12) to isolate individual bacterial colonies expressing FTase substrates. The molecular mass of the GST fusion proteins expressed in these purified bacterial colonies was determined by immunoblot-
BIOCHEMICAL ANALYSIS OF THE Rit AND Rin GTPases ting of IPTG-induced cultures as described previously (12). The majority of the fusion proteins were in the 30- to 38-kDa size range and were not significantly larger than the 27-kDa size of the unfused GST protein. Twenty-five of the largest of these clones were selected for further characterization. DNA sequence analysis found that all of the GST fusion proteins contained authentic CAAX C-terminal motifs, which proved the utility of the expression cloning strategy. However, the majority of the recombinant proteins (64%) were found to have arisen from either inappropriate reading frame cDNA fusions from within the coding region or the 39 untranslated region of previously characterized cDNA clones. The remaining clones were found to encode previously unidentified cDNAs, as indicated by their absence from the DNA Data Banks. The analysis of these clones will be described in a later publication. Recombinant protein production. Recombinant Rit and Rin were expressed as GST fusion proteins. In a series of preliminary experiments it was determined that short C-terminal deletions were needed to allow the expression of large amounts of intact and soluble Rit and Rin recombinant proteins. The smallest deletion that allowed stable protein expression (deletion of the final C-terminal 18 amino acids) was therefore introduced to both recombinant proteins. To construct the pGEX-hRitC plasmid, PCR was performed on the EST N36448 using primers 59 GAAGATCTATGGATTCTGGAACTCGCCCA 39 and 59 GGAGATCTAGTTTTTGGGCTTAGATTTTTTC 39 to create a cDNA flanked by BglII restriction sites and resulting in the deletion of the last 18 amino acids from the carboxyl terminus of the predicted hRit protein. The PCR product was subcloned to BamHI-digested pGEX-KG and sequenced to verify the amplified product. To construct the pGEX-mRinC plasmid, PCR was performed on the EST W51063 using primers 59 CGGATCCATGGAAGCAGAAAACGAAGCC39 and 59 GGAATTCAGTCCTTCCTCTTTAATTTC 39 to introduce a 59 BamHI and 39 EcoRI restriction site and to delete the final 18 amino acids from the C-terminus of the predicted protein. The product was sequenced to verify the fidelity of the amplified cDNA and subcloned to BamHI/EcoRI-digested pGEX-KG (11) to create pGEXmRinC. pGEX-hRitC and pGEX-mRinC were transformed into DH5a bacteria and recombinant proteins expressed and purified by glutathione–agarose affinity column as described previously (14). Briefly, transformed bacteria were grown in 23 YT medium containing 50 mg/ml carbenicillin at 37°C to an OD 600 of 1.0, shifted to room temperature, and allowed to grow for an additional 4 h in the presence of 0.5 mM IPTG to induce protein production. The cell pellet from a 1-liter culture was resuspended in 20 ml of TES (50 mM Tris–Cl pH 7.5, 40 mM EDTA, 25% sucrose) and centrifuged, and the washed pellet was resuspended in 10 ml of buffer B (10 mM Tris–Cl pH 7.4, 1 mM EDTA, 10 mM MgCl 2, 1 mM DTT, 1 mM phenylmethanesulfonyl fluoride (PMSF), 1 mg/ml leupeptin, 1 mg/ml pepstatin, 1 mg/ml aprotonin, 100 mM GDP). The cells were then lysed by three passages through a French pressure cell, 25 ml of Buffer C (20 mM Hepes pH 7.6, 100 mM KCl, 10 mM MgCl 2, 100 mM GDP, 20% glycerol, 1 mM DTT, 1 mM EDTA, 1 mM PMSF, 1 mg/ml leupeptin, 1 mg/ml pepstatin, and 1 mg/ml aprotonin) was added, and the lysate was centrifuged in a JA 25.500 rotor (Beckman) at 16,000g for 30 min. The cleared supernatant was incubated with glutathione–agarose beads (Sigma) at 4°C for 20 min and the GST fusion proteins purified as described (14). The fractions that contain the eluted protein were combined and dialyzed against 20 mM Tris–HCl pH 7.5, 100 mM NaCl, 1 mM DTT, 1 mM MgCl 2, and 10% glycerol for 6 h at 4°C and stored in multiple aliquots at 270°C. Protein concentrations were determined by the Bradford assay (Bio-Rad), using bovine serum albumin as a standard. Recombinant Rit was judged to be 15–20% active and recombinant Rin 8 –12% active as measured by radiolabeled GTP and GDP binding using the rapid filtration assay (see below). Rit and Rin proteins bearing N-terminal His 6 sequences were produced using the pTrcHisB expression vector (Invitrogen). To make pTrcHisBRitC, the PCR product used to generate pGEX-hRitC
209
(as described above) was digested with BglII and ligated to the BglII site of pTrcHisB. The BamHI/EcoRI fragment of pGEX-mRinC was inserted into BglII/EcoRI-digested pTrcHisB to generate pTrcHisBRinC. The pRSET-H-Ras plasmid expressing recombinant H-Ras was kindly provided by Dr. Guy James (University of Texas Health Science Center at San Antonio), and all three His-tagged proteins were expressed and purified as described (15). The Rlf-RID domain was expressed as a GST fusion protein for binding studies. To generate the pGEX-Rlf-RID construct, PCR was performed on pMTZ-HA-Rlf (provided by Dr. Ginell Post, University of Kentucky) with primers 59 CGGGAT CCTCTGATTGCCGAATCATC 39 and 59 GGAATTCAGAACAGTGCCCGTGC 39 to introduce a 59 BamHI and 39 EcoRI site flanking the Rlf-RID domain (amino acids 648 –778). The PCR product was digested with BamHI/EcoRI and ligated to similarly digested pGEX-KG. Protein production was carried out in DH5a cells induced with 0.5 mM IPTG and allowed to continue for 20 h at room temperature with shaking. In vitro mutagenesis. Oligonucleotide site-directed mutagenesis with the Transformer Mutagenesis kit (Clontech) was used to generate the single amino acid substitution mutants Rit(Q79L) and Rin(Q78L). pGEXKGhRitC(Q79L) was generated using the synthetic mutagenic oligonucleotide primer 59 CAGCTGGACTGGCAGAGTTTACAG 39 and a selection oligo according to the manufacturer’s protocol. pGEXKGmRinC(Q78L) was produced using the Rin mutagenic primer 59 CACTGCCGGTCTGGCAGAGTT C 39. The same method was used to generate each of the point mutants described in these studies. PCR was used to generate C-terminal deleted versions of each mutant (see above). Each of the mutant cDNAs was mapped using restriction enzymes and the entire open reading frame subjected to DNA sequence analysis. After the identity of each mutant was confirmed, recombinant fusion protein was produced as described above. Guanine nucleotide binding assays. Nucleotide binding was determined by the rapid filtration technique (16). Binding buffer (50 ml total reaction volume) (20 mM Tris, pH 7.5, 100 mM NaCl, 1 mM EDTA, 1 mM DTT, and 1 mg/ml BSA) containing 1 mM GST fusion protein, 2 mM [ 35S]GTPgS, or [ 3H]GDP (each at 2 Ci/mmol, NEN), and the indicated concentration of MgCl 2 was prepared on ice and transferred to 30°C to initiate the reaction. The free magnesium concentration was calculated based on the following equation: [Mg 21] Total 5 [Mg 21] Free 3 {1 1 [EDTA] Total/([Mg 21] Free 1 K EDTA z Mg)}, and K EDTA z Mg 5 1 mM (17). After incubation at 30°C for the indicated times, the samples were diluted in 2 ml of ice-cold washing buffer (20 mM Tris pH 7.5, 100 mM NaCl, 10 mM MgCl 2, 1 mM DTT), filtered immediately through 0.45-mm nitrocellulose filters (BA85 nitrocellulose, Schleicher and Schuell), and then washed twice with 4 ml of ice-cold washing buffer. Filters were then dried and immersed in 4 ml scintillation fluid (ScintiSafe Econo 1, Fisher Scientific), and the amount of bound guanine nucleotide was measured by scintillation counting. Data points in all figures represent the mean of duplicate determinations from a representative experiment that was repeated at least twice. To determine the specificity of nucleotide binding, recombinant GST-mRin or GST-hRit protein was incubated in the reaction mixture with or without competing nucleotides. The reaction mixture contained 20 mM Tris, pH 7.5, 100 mM NaCl, 1 mM EDTA, 11 mM MgCl 2, 1 mM DTT, 1 mg/ml BSA, 1 mM GST fusion protein, 1 mM [ 35S]GTPgS (2 Ci/mmol, NEN), and 20 mM of competing unlabeled nucleotide (ATP, UTP, CTP, GTP, GDP, or GTPgS). After incubation at 30°C for 30 min, the samples were filtered on BA85 nitrocellulose membrane and counted as described above. The amount of [ 35S]GTPgS bound in the absence of competitors was set as the 100% binding value and was used to compare the binding in the presence of specific nucleotide competitors. To determine the GTPgS and GDP saturation binding curve for Rit and Rin, the GST-hRit or GST-mRin proteins were incubated with increasing concentration of [ 35S]GTPgS or [ 3H]GDP in standard
210
SHAO ET AL.
binding buffer containing 1 mM [Mg] Free as described above. The reaction mixture was incubated at 30°C for 45 min, filtered, and counted as above. Guanine nucleotide dissociation. GTPgS and GDP dissociation was measured as described previously (18). Briefly, GST-hRit or GST-mRin was incubated with 2 mM [ 35S]GTPgS or [ 3H]GDP at 30°C for 30 min to allow protein to become loaded with radiolabeled nucleotide, then cold GTPgS or GDP was added to the reaction buffer to a final concentration of 1 mM to initiate the dissociation assay. At the indicated times after the addition of cold GTPgS or GDP the samples were diluted in ice-cold wash buffer, filtered, and counted as above. As a control to determine whether the release of nucleotides was due to protein denaturation during the time course, no competing cold nucleotide was added to the reaction buffer after the protein was loaded with [ 35S]GTPgS or [ 3H]GDP, and the experiment was performed as above. GTPase assay. Steady-state GTP hydrolysis was measured because the rapid release rate of GTP from both Rit and Rin proteins, even in the presence of high Mg 21 concentrations, did not allow the isolation of the stable radiolabeled GTP complexes necessary for single turnover measurements of hydrolysis. Affinity-purified Rit and Rin fusion proteins were found to contain a contaminating phosphatase activity which resulted in a low background rate of [a- 32P]GTP hydrolysis. Unlabeled ATP (20 mM) was therefore added to each reaction to competitively inhibit this nonspecific phosphatase activity. GTP hydrolysis assays were performed in buffer containing 20 mM Tris, pH 7.5, 100 mM NaCl, 1 mM DTT, 1 mM MgCl 2, 0.2 mM active Rit/Rin GST fusion protein (as determined by filter binding assay), 10 mM [a- 32P]GTP, and either 40 mM ATP or 20 mM ATP and 10 mM GTP (19). Reactions were incubated at 30°C and 1-ml aliquots were removed at the indicated times and spotted directly onto poly(ethylene)imine– cellulose plates (EM Separations). Chromatograms were developed in 1 M LiCl and 1 M formic acid and exposed to X-OMAT AR film (Kodak) for 15 h. The migration of authentic GTP and GDP standards was visualized using ultraviolet light (254 nm). Two-hybrid analysis of Rit and Rin with known Ras effectors. The yeast strain PJ69-4A was used for all two-hybrid analyses essentially as described (20). Rit/Rin deletion mutants (lacking their final 18 amino acids) and CAAX-motif-deficient H-Ras were subcloned into the yeast vector pGAD-C1 and expressed as GAL4 DNA binding domain fusion proteins. Raf RBD, A-Raf, B-Raf, C-Raf, RalGDS-RID, Rlf-RID, and P110 of PI-3 kinase were expressed as VP16 activation domain fusions, while AF-6/Canoe and Rin were expressed as GAL4 activation domain fusions (the activation domain plasmids were the kind gift of Dr. C. Der, University of North Carolina at Chapel Hill, and Dr. A. Vojtek, University of Michigan). PJ69-4A cells were cotransformed with GAL4 DNA-binding and GAL4 or VP16 activation domain fusion plasmids and plated on selective synthetic medium lacking tryptophan and leucine. After 3 days of growth at 30°C, colonies were restreaked onto selective media lacking tryptophan, leucine, adenine, and histidine but containing 1 mM 3-AT. Plates were incubated for four days at 30°C and growth was assessed. In vitro interaction of Rlf-RID with Rit and Ha-Ras. Nucleotide loading was accomplished as follows. Either His-Rit or His-Ha-Ras in a total volume of 400 ml was loaded with either [ 35S]GTPgS or [ 3H]GDP in standard binding buffer containing either 500 mM (Rit) or 10 mM (Ha-Ras) [Mg] Free as described above. Nucleotide loading was terminated by adjusting [Mg] Free to 20 mM. A sample (10 ml) from each reaction was then subjected to nitrocellulose filtration to determine the extent of guanine nucleotide binding. For the binding studies, 1 nmol of GST or GST-Rlf RID was incubated with increasing amount of [ 35S]GTPgS- or [ 3H]GDP-bound Rit or [ 35S]GTPgSbound Ha-Ras and 30 ml of a 1/1 (v/v) slurry of glutathione–agarose beads (Sigma) in 300 ml of binding buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 20 mM MgCl 2, 20 mM imidazole, 1% NP-40, 2 mg/ml BSA)
at 4°C for 1 h. Reactions were washed five times with 1 ml of binding buffer and the bound guanine nucleotide determined by scintillation counting. For each quantity of Rit and Ha-Ras used, the specific binding was determined by subtracting counts bound to GST alone from counts bound by the Rlf-RID GST fusion protein. Interaction of in vitro translated Rlf with Rit, Rin, and Ha-Ras. Rlf was transcribed and translated in vitro in the presence of [ 35S]methionine using the Single Tube Protein System 3 (STP3) Kit (Novagen) with pCITE-4CRlf as template. To generate pCITE-4CRlf, the plasmid pMT2-HARLF (a generous gift from Dr. Ginell Post, University of Kentucky) was digested with SalI and NotI and ligated to pCITE-4C (Novagen). For binding experiments, 20 ml of a 50% glutathione agarose beads bound to 5 mg of GST, GSTHa-Ras, GSThRitC, or GSTmRinC protein were used. The proteins were each preloaded with either GTPgS or GDP as described above. The beads were then diluted into 180 ml interaction buffer (50 mM Tris–Cl, pH 7.5, 120 mM NaCl, 10 mM MgCl 2, 1% Triton X-100) containing 20 mM of the appropriate nucleotide and incubated with gentle rotation for 2 h at 4°C with 5 ml of the in vitro translated [ 35S]methionine-labeled Rlf protein. Beads were concentrated by brief centrifugation and washed five times with 1 ml ice-cold interaction buffer. The bound protein was eluted from the beads by the addition of 15 ml release buffer (PBS containing 25 mM glutathione). Samples were analyzed by SDS–polyacrylamide gel electrophoresis on 8% polyacrylamide gels; after staining with Coomassie Blue to detect the GST and GST fusion proteins, gels were treated with Amplify (Amersham Pharmacia Biotech), dried, and exposed to film.
RESULTS
Cloning of Rit and Rin An expression cloning screen was performed on a human retinal cDNA expression library as described in detail under Experimental Procedures in order to identify novel farnesylated proteins. A size-fractionated cDNA library was prepared from human retinal mRNA in a bacterial expression vector that produces glutathione S-transferase (GST) fusion proteins. The library was plated to nitrocellulose filters at ;10,000 colonies/ filter and induced to produce protein, and the filters were treated as described under Experimental Procedures. Approximately 100,000 cDNA-containing clones were screened by prenylation replica colony protein blotting with [ 3H]FPP in the presence of recombinant farnesyltransferase, resulting in the identification of 97 putative prenylated GST fusion clones. Bacterial colonies from regions of the filters where positives were detected were isolated and subsequently rescreened. Of the colonies which produced a signal in the first screening, 82% were positive in the second screen. The molecular mass of each prenylated GST fusion protein was determined by immunoblotting with antiGST antibodies. The majority of the fusion proteins were in the 30- to 38-kDa size range and were not significantly larger than the 27-kDa size of the unfused GST protein. Twenty-five of the largest of these clones were selected for further characterization. DNA sequence analysis found that all of the GST fusion proteins contained authentic CAAX C-terminal motifs,
BIOCHEMICAL ANALYSIS OF THE Rit AND Rin GTPases
which proved the utility of the expression cloning strategy. However, the majority of the recombinant proteins (64%) were found to have arisen from either inappropriate reading frame cDNA fusions from within the coding region or the 39 untranslated region of previously characterized cDNA clones. The remaining clones were found to encode previously unidentified cDNAs, as indicated by their absence from the DNA Data Banks. The analysis of these clones will be described in a later publication. The exception to this was clone FT-57, which encoded a short 21-amino-acid open reading frame ending in the sequence Cys–Ser–Phe– Thr, CSFT, a previously unrecognized CAAX motif. When we queried the NCBI expressed sequence tag (EST) database with the FT-57 sequence using the BLAST local multiple alignment algorithm (21, 22), we identified several expressed sequence tag clones (including ESTs T58089, N78919, and N75973) which appeared to contain the 39 end of a novel Ras family member. Using these initial EST sequences, it was possible through repeated databank searches to identify an EST encoding the entire putative open reading frame encoded for by FT-57. This EST (N36448) was ordered and sequenced. The resulting cDNA clone contained an open reading frame that predicted a 219amino-acid Ras-related protein; however, the predicted open reading frame failed to terminate in a CAAX motif. Sequencing of additional ESTs as well as PCR amplification of the gene from a variety of cDNA libraries confirmed the original EST cDNA sequence and showed that the FT-57 GST protein arose from an inappropriate gene fusion which placed a portion of the 39 untranslated region of the gene in frame to GST. Thus, our expression cloning strategy fortuitously identified a novel member of the Ras superfamily which lacks a C-terminal CAAX motif and is therefore not expected to be subject to posttranslational isoprenylation. During this analysis, a FT-57 homologue was identified from both mouse and human ESTs. The largest of these EST clones was ordered and sequenced. The deduced amino acid sequence of the mouse EST (W51063) was found to encode a second novel Ras family member. During the cloning, two groups Wes et al. (23) and Lee et al. (24) reported the cloning of the RIC, Rin, and Rit, Ras-related genes. The Drosophila RIC (Ras-related protein which interacted with calmodulin) gene was isolated using a screen designed to identify retinal expressed calmodulin-binding proteins (23) and this group identified the Rit (Ras-like expressed in many tissues) and Rin (for Ras-like in neurons) genes in an EST databank search for RIC-related genes. The Rin and Rit genes were independently isolated from mouse retina using a degenerate RT-PCR based cloning strategy (24). The human Rit and FT-57 proteins share complete amino acid identity (99.8% nucleotide iden-
211
tity) and therefore encode the same protein, while the murine Rin and FT-57 homologue nucleotide sequences are 96.6% identical in their coding regions (98% amino acid identity) and most likely encode the same protein or highly related proteins. We will therefore refer to these proteins as Rit and Rin. Finally, an additional homolog termed RIBA (Genebank U78166) was identified in Data Bank searches and appears to encode an alternatively spliced form of the human Rin gene. Expression and Guanine Nucleotide-Binding Characteristics of Recombinant Rit and Rin While GTP overlay assays have demonstrated that both Rin and Rit are capable of GTP binding (24), detailed characterization of their nucleotide binding properties, nucleotide exchange rates, and whether these proteins catalyze GTP hydrolysis had not been determined. To begin this analysis, recombinant Rit and Rin proteins were expressed as GST fusions in E. coli (see Experimental Procedures). The production of full-length soluble Rit and Rin proteins in bacteria proved to be quite difficult. Attempts to express the full-length proteins resulted in the production of a heterogeneous mixture of recombinant protein, apparently the result of C-terminal directed proteolysis (data not shown). However, deletion of the C-terminal 18 amino acids from both Rit and Rin allowed high-level expression of intact protein. As shown in Fig. 1, GSThRit (lane 1) and GST-mRin (lane 3) fusion proteins were purified to near homogeneity as revealed by SDS– PAGE and Coomassie blue staining. Binding of guanine nucleotides to these fusion proteins was assayed by rapid filtration on nitrocellulose filters (16). It is important to note that both Rit and Rin were purified in the presence of GDP to stabilize the proteins and likely exist as GDP-bound complexes at the start of these binding studies. The association therefore actually measures the rate of exchange of labeled nucleotide for prebound unlabeled GDP. As seen in Fig. 2, both Rit and Rin proteins exhibited Mg 21- and time-dependent [ 35S]GTPgS binding activity. As with other GTP-binding proteins (18, 25), the association of guanine nucleotides with both Rit and Rin was greatly affected by the concentration of magnesium ions. Removal of free Mg 21 in the assay mixture with EDTA completely abolished GTPgS binding (Fig. 2). Binding of [ 35S]GTPgS and [ 3H]GDP was found to be dose dependent (data not shown). The ability of various nucleotides to compete for the binding of [ 35S]GTPgS to Rit and Rin was also examined. Binding was specific for guanine nucleotides, since an excess (20-fold) of nonradioactive GDP, GTP, or GTPgS, but not ATP, CTP, or UTP, competed with [ 35S]GTPgS (Fig. 3). These studies were repeated with recombinant Rit and
212
SHAO ET AL.
class of nucleotide-binding site (Fig. 4). The half-time for [ 3H]GDP dissociation from GST-Rit was 15.1 min in the presence of 10 mM Mg 21, a value that is quite similar to that for many Ras-related proteins (3). The half-time for GDP dissociation from GST-Rin was 38.6 min under the same conditions (Table I). At a lower Mg 21 concentration (1 mM), the half-life for the RinGDP complex was reduced approximately 2-fold to 15.4 min. However, reducing the Mg 21 concentration to 1 mM led to a half-life of only 3.4 min for Rit, a reduction of approximately 4.5-fold. We also measured the dissociation of [ 35S]GTPgS from Rit and Rin. As seen in Fig. 4, both Rit and Rin release GTPgS more rapidly than GDP at 10 mM Mg 21. The half times for GTP dissociation for Rit and Rin were 6.8 and 3.2 min, respectively. These GTP dissociation rates are faster than those of most Ras-like proteins (3). Under the same assay conditions, less than 10% of prebound GTPgS was released from recombiFIG. 1. SDS–polyacrylamide gel electrophoresis of purified GSTRit and GST-Rin proteins. Affinity-purified recombinant wild-type GST-Rit and GST-Rin and mutants (1 mg) were subjected to electrophoresis on a 8% SDS–polyacrylamide gel, and the protein bands were detected with Coomassie blue as described under Experimental Procedures. Lane 1, wild-type GST-Rit; lane 2, [Leu 79]GST-Rit; lane 3, wild-type GST-Rin; lane 4, [Leu 78]GST-Rin. Each recombinant protein contained an 18-amino-acid C-terminal deletion to aid bacterial expression (see Experimental Procedures for details). The molecular mass for marker protein standards are shown on the left.
Rin proteins which bore six histidine affinity tags at their N-termini. The biochemical properties of these recombinant proteins were identical to those determined for the GST proteins (data not shown) and suggest that the Rit/Rin proteins are not adversely affected by being expressed as N-terminal fusions. Because affinity chromatography using glutathione agarose resulted in more highly purified proteins, Rit/ Rin-GST fusion proteins were used in the remaining biochemical analyses. Kinetics of [ 3H]GDP and [ 35S]GTPgS Dissociation The replacement of GDP by GTP constitutes an essential step in the activation of Ras-related proteins, since the GDP-bound form of the protein represents an inactive state and GDP release is often the rate-limiting step in their activation (1). Rate constants for the dissociation of guanine nucleotide (k off) from Rit and Rin were obtained by prebinding [ 3H]GDP or [ 35S]GTPgS to GST-Rit and GST-Rin, adding a 500-fold excess of unlabeled nucleotide and then assaying for the loss of radioactivity at 30°C in either 10 mM or 1 mM Mg 21 as described under Experimental Procedures. Under these conditions, the dissociation curves follow a single exponential, as expected for a single
FIG. 2. Association of [ 35S]GTPgS with GST-Rit and GST-Rin. Recombinant GST-Rit (top) and GST-Rin (bottom) proteins (1 mM) were incubated in a reaction volume of 50 ml containing 2 mM of [ 35S]GTPgS (2 Ci/mmol) in the presence of the indicated MgCl 2 concentration: none (filled circle), 10 mM (open square), 100 mM (filled diamond), 1 mM (open circle), or 10 mM (filled square). After incubation for the indicated period at 30°C, the amount of [ 35S]GTPgS bound to the Rit and Rin protein was determined as described under Experimental Procedures. A blank value determined in parallel reactions containing no recombinant Rit and Rin protein was subtracted from each value. Each value is the average of duplicate incubations and are representative of a typical experiment repeated three times.
BIOCHEMICAL ANALYSIS OF THE Rit AND Rin GTPases
213
in the recombinant protein preparations and unlabeled GTP was used as a control to show that any remaining nucleotide hydrolysis activity was GTP specific. The GTPase activity in our Rit and Rin preparations was not due to contaminating proteins from E. coli because GST protein purified under identical conditions did not hydrolyze GTP (data not shown). To further confirm the GTPase activity of these proteins, we used sitedirected mutagenesis to generate putative GTPase-defective Rit and Rin mutants. We replaced Gln 79 in Rit and Gln 78 in Rin with leucine (Fig. 1). The cognate Q61L mutation in Ras results in a protein with reduced GTPase activity and is often found to be mutated, alone or together with Val 12, in transforming forms of Ras (26). Initial characterization of these mutant proteins showed that both Rit(Q79L) and Rin(Q78L) bound guanine nucleotides and had rates of GTPgS dissociation
FIG. 3. Rit and Rin are specific GTP-binding proteins. GST-Rit (1 mM, top) or GST-Rin (1 mM, bottom) were incubated at 30°C with [ 35S]GTPgS (1 mM) and 10 mM MgCl 2 for 30 min in the absence (Control) or the presence of the indicated nucleotides (20 mM) and subjected to a filter-binding assay to quantitate bound radioactivity. Each value is the average of duplicate incubations and is representative of three separate experiments.
nant Ha-Ras (Fig. 4, bottom). Reduction of the Mg 21 concentration to 1 mM resulted in an approximate 2-fold reduction in the half-time of release for both Rit/Rin proteins (Table I). Thus, at 10 mM Mg 21 both Rit and Rin have a lower k off for GDP than for GTP, while at 1 mM Mg 21 only Rin displays a lower k off for GDP. [ 32P]GTP Hydrolysis Because GTP hydrolysis plays an important role in the regulation of GTP-binding proteins, we analyzed the intrinsic GTPase activity of both Rit and Rin. The rapid release of GTP from both recombinant enzymes precluded the use of single-turnover GTP hydrolysis experiments. Therefore, we measured steady-state hydrolysis by incubating the Rit and Rin proteins with [a- 32P]GTP and followed GTP hydrolysis by measuring the generation of [a- 32P]GDP. As shown in Fig. 5, nucleotide analysis by PEI thin-layer chromatography revealed that both recombinant Rit and Rin were capable of slowly hydrolyzing [a- 32P]GTP in the presence of an excess of unlabeled ATP, but not in the presence of both unlabeled ATP and GTP. Unlabeled ATP was used to eliminate a small amount of contaminating phosphatase activity found
FIG. 4. Dissociation of [ 3H]GDP and [ 35S]GTPgS from GST-Rit and GST-Rin in the presence of 10 mM or 1 mM MgCl 2. 1 mM Rit (top) and 1 mM Rin (bottom) were incubated in the presence of either 2 mM [ 3H]GDP (squares) or 2 mM [ 35S]GTPgS (circles) at 30°C for 30 min in reaction buffer containing either 1 mM (filled symbol) or 10 mM free Mg 21 (open symbol) as described under Experimental Procedures. To initiate dissociation, unlabeled GDP or GTPgS were added to 1 mM, and at the indicated times 50-ml aliquots were removed and the protein-bound radioactivity was determined in a filter-binding assay as described under Experimental Procedures. As a positive control, 1 mM Ha-Ras (bottom) was incubated in the presence of either radiolabeled GDP (open triangles) or GTPgS (filled triangles) and assayed for nucleotide release as described above. Each value is the average of duplicate incubations and is representative of four independent experiments.
214
SHAO ET AL. TABLE I
Guanine Nucleotide Dissociation Constants for Wild-Type and Mutant Rit/Rin Proteins In the presence of 1 mM Mg 21 GTPgS
In the presence of 10 mM Mg 21 GTPgS
GDP
GDP
Protein
t 1/2 (min)
k off (min 21)
t 1/2 (min)
k off (min 21)
t 1/2 (min)
k off (min 21)
t 1/2 (min)
k off (min 21)
Rit Rin RitQ79L RinQ78L
4.0 1.6 3.3 1.0
0.2 0.4 0.2 0.7
3.4 15.4 3.2 4.2
0.2 0.05 0.2 0.2
6.8 3.2
0.1 0.2
15.1 38.6
0.05 0.02
Note. The dissociation rate constants were determined as described in the legend to Fig. 7 and as described under Experimental Procedures. They represent the average of two independent experiments; the standard deviation was less than 7%.
which were only modestly higher than those of the corresponding wild-type protein (Table I). However, both Rin and Rit mutant proteins had dramatically reduced rates of GTP hydrolysis and did not hydrolyze GTP detectably during a 1-h incubation (Fig. 5, lanes
7–12, top and bottom). In addition, Rin(Q79L) had a significantly increased rate (3.5-fold) of GDP dissociation (Table I). These results strongly suggest that Gln 79/78 plays an important role in the intrinsic GTPase activity of the Rit and Rin proteins. In addition, these
FIG. 5. [a- 32P]GTP hydrolysis of wild-type and mutant Rit and Rin. Rit and Rin(Q79L) (top) or Rin and Rin (Q78L) proteins (bottom) (0.2 mM active fusion protein) were incubated with [a- 32P]GTP (10 mM, 2 Ci/mmol) and 1 mM Mg 21 in the presence of either 40 mM unlabeled ATP or 20 mM unlabeled ATP and 10 mM unlabeled GTP. Aliquots were removed from the mixture at the indicated times and analyzed by PEI thin-layer chromatography plates and autoradiography as described under Experimental Procedures. The migration of authentic GTP and GDP standards is indicated. The results are representative of five independent experiments.
215
BIOCHEMICAL ANALYSIS OF THE Rit AND Rin GTPases TABLE II
Interaction of Rit, Rin, and H-Ras with Ras Effectors in the Yeast Two-Hybrid System Cell growth a for transactivation domain fusion GAL4 DNAbinding domain fusion
RafRBD
A-Raf
B-Raf
C-Raf
RalGDSRID
RLF-RID
AF-6
RIN1
PI3K p110
Rit RitS35N RitQ79L Rin RinS34N RinQ78L H-Ras pGBD-C1
1 2 1 1 2 1 1 2
2 2 2 2 2 2 1 2
2 2 2 2 2 2 1 2
2 2 2 2 2 2 1 2
1 2 1 1 2 1 1 2
1 2 1 1 2 1 1 2
1 2 1 1 2 1 1 2
2 2 2 2 2 2 1 2
2 2 2 2 2 2 1 2
a The yeast strain PJ69-4A was cotransformed with the indicated Ras effector proteins and wild-type or mutant GTPases and assayed after 4 days for growth on selection plates. Cell growth is indicated by a 1, and lack of growth is shown as a 2.
mutants confirm that the observed GTP hydrolysis was due to the intrinsic GTPase activity of the Rit/Rin proteins rather than contaminating bacterial phosphatases. Interaction of Rit and Rin with Putative Ras Effectors Because Rit, Rin, and Ras share a great deal of amino acid identity within their effector domain sequences, and the core effector domain (Ras residues 32– 40) plays a critical role in the interaction of Ras proteins with a wide range of cellular effector proteins (27), we characterized the interaction between Rit and Rin and a series of known Ras effectors. Since Ras interaction with these proteins had been demonstrated in the yeast two-hybrid assay system, we used this method to examine interaction with our GTPases. The yeast strain PJ69-4A was cotransfected with a plasmid containing the Gal4 DNA-binding domain fused to the C-terminally truncated forms of Rit and Rin (pGBD-Rit and pGBD-Rin) and a plasmid containing the Gal4 (or VP16) transcriptional activation domain fused to cDNAs encoding a series of previously characterized Ras effector proteins (20). We initially tested the interaction between Rit and Rin and the best-characterized Ras effector, Raf, using expression plasmids encoding the VP16 acidic activation domain fused to the full-length sequences of Raf-1, B-Raf, and C-Raf or to a truncated Raf-1 fragment which contains only the minimal Ras-binding domain (Raf-1 residues 51–130; designated Raf-RBD). Surprisingly, whereas both Ha-Ras and Rit/Rin showed interaction with Raf-RBD, only Ha-Ras interacted with the full-length versions of the three Raf proteins (Table II). To extend this analysis we next examined the interaction between Rit/Rin and transcriptional activation domain fusions expressing the full-length candidate Ras effectors AF6/Canoe (28), RIN1 (29), phosphatidylino-
sitol 3-kinase (p110 catalytic subunit) (30), and the minimal C-terminal Ras interaction domain (residues 647–768 and 648 –778 respectively; RID) of RalGDS and Rlf (two GDSs for the Ras-related GTPase Ral) (31). Although Rit/Rin and Ras share a great deal of sequence identity within their effector domains, Rit and Rin were only found to interact with AF6 and the RID domains of RalGDS and Rlf, while wild-type HaRas was capable of interaction with each of these proteins (Table II). Although the known candidate Ras effectors comprise a diverse set of structurally and functionally distinct proteins, they all show preferential affinity for activated GTP-bound Ras, and require an intact effector domain. To determine if RalGDS and AF6 possess the properties of authentic effectors, these proteins were tested with various wild-type and mutated GTPases. Wild-type and constitutively activated versions of Rit and Rin (Q79L and Q78L respectively) and wildtype Ha-Ras interact with each of the putative effector proteins (Table II). As shown in Table III, effector domain mutants of the constitutively activated forms of Rit and Rin (Rit Q79LT53S, Rit Q79LE55G, and Rin Q78LT52A) were unable to interact with AF6. These results suggest that AF6 binding requires an intact effector domain. Interestingly, similar results were seen with RalGDS-RID except that the Rit effector domain mutant E55G did interact (Table III). Since the equivalent E37G mutation in Ras inhibits its ability to interact with Raf, but not RalGDS, it is predicted that Rit/Rin and Ras have similar structural requirements for binding to this common effector (32). Rit S35N and Rin S34N, which by analogy with the Ras S17N dominant negative mutant are thought to have predominantly GDP-bound, showed no detectable interaction with the effector proteins, suggesting that AF6, Rlf-RID, and RalGDS-RID
216
SHAO ET AL.
DISCUSSION
TABLE III
Interaction of RalGDS, RLF, and AF-6 with Rit and Rin in the Yeast Two-Hybrid System Cell growth a for transactivation domain fusion
GAL4 DNAbinding domain fusion
RalGDS-RID
RLF-RID
AF-6
Rit RitQ79L Rit79L53S Rit79L55G Rin RinQ78L Rin78L52A
1 1 2 1 1 1 2
1 1 2 1 1 1 2
1 1 2 2 1 1 2
In this paper we describe the identification and initial biochemical characterization of two small Ras-related GTPases, Rit and Rin. The predicted amino acid sequence of Rit and Rin are most highly homologous with members of the Ras subfamily of small GTPases, and by virtue of this similarity we include these proteins as the newest additions to this gene family. However, both proteins have biochemical properties and unique structural and biological characteristics that
a The yeast strain PJ69-4A was cotransformed with the indicated Ras effector proteins and wild-type or mutant GTPases and assayed after 4 days for growth on selection plates. Cell growth is indicated by a 1, and lack of growth is shown as a 2.
show preferential binding to the active GTP-bound forms of Rit and Rin (Table II). To confirm and extend the observations from the two-hybrid binding analyses, we generated a GST fusion protein containing the Rlf-RID for in vitro binding experiments. We measured the relative affinity of RlfRID for GTPgS-bound Rit and Ha-Ras. As expected, Rlf-RID bound the active forms of both proteins in a concentration-dependent manner. However, Rlf-RID exhibited greater binding to Ha-Ras when compared with Rit at all points tested (Fig. 6A). The failure of RalGDS-RID, Rlf-RID, and AF-6 to bind to either Rit S35N or Rin S34N in two-hybrid assays suggested that these proteins preferentially interact with the GTPbound form of Rit and Rin. To test this, Rit was loaded with either GDP or GTPgS and binding reactions performed to compare the relative affinity of Rlf-RID for the two nucleotide dependent conformations of Rit. The results in Fig. 6B clearly demonstrate that Rlf-RID binds exclusively to Rin-GTPgS. Within the range of Rit tested (50 –500 pmol) we were unable to detect any specific binding to Rit-GDP. To assess the ability of full-length Rlf to interact with Rit or Rin, Rlf was transcribed and translated in vitro in the presence of [ 35S]methionine, and its ability to bind a series of Ras-related proteins fused to GST, immobilized on glutathione-agarose beads, and loaded with either GDP or GTPgS was assessed. Figure 7 shows that this assay displayed results similar to those obtained using the isolated Rlf-RID domain in the yeast two-hybrid system. Rlf interacted with Ras, Rit and Rin in a GTPgS-dependent manner. Hence, RalGDSs, through the nucleotide-dependent binding of their C-terminal RID domains constitute potential effectors for Rit and Rin.
FIG. 6. Rlf-RID binding to Rit and Ras. Each assay contained (in a volume of 300 ml) Rlf-RID GST fusion protein (1 nmol), 30 ml of a 1/1 (v/v) suspension of glutathione–agarose, and the indicated amount of guanine nucleotide bound G-protein. After incubation for 1 h at 4°C, the samples were pelleted and washed, and the amount of bound G-protein was determined by scintillation counting as described under Experimental Procedures. (A) Binding curves comparing RlfRID binding to Rit-[ 35S]GTPgS (open circles) and Ha-Ras[ 35S]GTPgS (filled circles). (B) Binding curves comparing Rlf-RID binding to Rit-[ 35S]GTPgS (open circles) and Rit-[ 3H]GDP (filled squares). The results are representative of two experiments done in duplicate.
BIOCHEMICAL ANALYSIS OF THE Rit AND Rin GTPases
217
FIG. 7. Interaction of Rlf with activated Rit, Rin, and Ras. In vitro translated Rlf was mixed with GST-small G proteins immobilized to glutathione–agarose beads and preloaded with either GDP or GTPgS. The interacting Rlf was coeluted with GST-small G proteins by addition of glutathione. Aliquots (15 ml) of the eluates were subjected to SDS–PAGE and vacuum-dried followed by autoradiography. Lane 1, in vitro translated Rlf; lane 2, GTPgS z GST; lane 3, GDP z GST; lane 4, GTPgS z GST-Ras; lane 5, GDP z GST-Ras; lane 6, GTPgS z GST-Rit; lane 7, GDP z GST-Rit; lane 8, GTPgS z GST-Rin; lane 9, GDP z GST-Rin. The arrow denotes the position of full-length Rlf. The results are representative of two independent experiments.
distinguish them from the majority of the Ras protein family. The deduced structures of the Rit and Rin proteins (219 and 217 amino acids, respectively) are larger than the majority of the Ras superfamily but are highly conserved in the five regions of Ras that constitute the core region which has been shown to be central to GTP binding and hydrolysis (4), whereas the flanking Nand C-terminal sequences are unrelated to Ras. These extended regions account for the larger molecular mass of Rit and Rin (25.1 and 24.8 kDa, respectively) when compared to the majority of the Ras family. The G4 and G5 consensus sequences involved in binding the guanine ring as well as the G1 and G3 domains involved in phosphate binding are well conserved in all Ras family members and found in both Rit and Rin proteins. The G2 consensus domain within Rit and Rin, however, differs from those of previously characterized Ras-related proteins. The putative G2 effector domain (HDPTIEDAY) is absolutely conserved between Rit and Rin, and there is only a single amino acid substitution within the G2 domain of Drosophila RIC (HDPTIEDSY). While the Rit and Rin protein effector domain is unique among known GTPases, it is most related to the effector domains in Ras subfamily members (YDPTIEDSY) which conserve seven of the nine amino acids. Because the activity of Ras-related proteins is regulated by their ability to exchange bound guanine nucleotides and to hydrolyze GTP, we examined whether Rit and Rin are GTPases. Both Rit and Rin were expressed as GST fusion proteins in E. coli, and the purity of each protein was confirmed by SDS–PAGE analysis and Coomassie blue staining (Fig. 1). We confirmed that the purified recombinant Rit and Rin proteins specifically bind GTP and GDP (Fig. 2 and 3) and, for the first time, demonstrated that each exhibits an intrinsic GTP hydrolysis activity (Fig. 5). To further characterize this GTPase activity, a single amino acid
mutation was introduced to each protein in a region known to alter the GTPase activity of other Ras proteins. Both Rit(Q79L) and Rin(Q78L) lost measurable GTPase activity (Fig. 5). The equivalent GTPase-defective mutant in Ras is a dominant promoter of cellular transformation (33). Because Ras-like proteins are presumably active in their GTP bound form, these hydrolysis-defective mutants may provide a valuable tool for future studies designed to address the biological role of Rit and Rin proteins in vivo. A surprising finding of the current study was the rapid uncatalyzed guanine nucleotide dissociation rates of both Rit and Rin. In particular, the GTPgS dissociation rate of Rit and Rin were approximately 5–10 times faster than that of most Ras subfamily members (Table I, Fig. 4). The rates of GDP dissociation were also quite rapid (Table I). These data are unexpected because there is no obvious alteration in the primary amino acid sequence that occurs in regions thought to be involved in the regulation of guanine nucleotide binding and release between Rit and Rin and other members of the Ras family. However, other GTPases without obvious amino acid substitutions have also been reported to have relatively high rates of nucleotide exchange. Indeed, the Ras-related GTPase TC21 has been reported to display a high intrinsic rate of guanine nucleotide exchange (34). The guanine nucleotide dissociation rate for Ras p21 is quite low, and the release of GDP is rate-limiting in its guanine nucleotide controlled activation cycle (1). Considering that the intracellular concentration of Mg 21 is approximately 10 mM (1–30 mM) (35) and that [GTP] far exceeds [GDP] (36), it is conceivable that a high percentage of both Rit and Rin proteins may be maintained in a GTP-bound active state in the normal magnesium concentrations of the cell. However, by analogy with other Ras-like GTPases, we would predict that each step of the Rit and Rin GTPase cycle is likely to be
218
SHAO ET AL.
modified substantially in vivo by GAPs, GDSs, and/or GDIs (2). It is becoming increasingly clear that Ras-dependent cellular transformation is mediated by the interaction and activation of multiple effector proteins (37). Because the core effector domains of Ras-related proteins are known to play a central role in defining effector protein interactions (3), and the effector loop of Rit and Rin is most homologous to that found in Ras proteins, in the present study we used yeast two-hybrid analysis to screen known Ras binding proteins as candidate effectors for both Rit and Rin. The interaction of the proteins RalGDS, RLF, and AF6 with Rit and Rin were shown to be both GTP- and effector domain-dependent (Tables II and III, Figs. 6 and 7). These proteins are therefore the first putative Rit and Rin binding proteins to be identified and may serve as important effector molecules. Indeed, RalGEF family proteins have been shown to stimulate the GDP-GTP exchange of Ral proteins in a Ras-dependent manner and a dominant negative form of Ral blocks Ras-dependent transformation of NIH 3T3 cells (38, 39). Although the cellular functions of Ral are not fully understood, a role for Ral proteins in the regulation of phospholipase D, developmental cell shape changes in Drosophila, and liganddependent receptor-mediated endocytosis have recently been described (40 – 42). Therefore, positive regulation of the RalGDS-Ral pathway contributes to the transforming actions of Ras and may play an important role in the cellular actions of Rit and Rin. The finding of AF6 association with Rit and Rin is also intriguing. Unlike other known Ras effectors, AF6 lacks a described enzymatic function and would appear to act as a scaffold for the assembly of defined protein complexes. AF6 was first described as a fusion partner of ALL-1 in acute myeloid leukemia (43), and its Drosophila homolog is genetically linked to the Notch cascade and other signaling pathways (44). It has a PDZ domain which has been shown to bind the Eph3B receptor, as well as a N-terminal Ras-binding site (28, 45). ZO-1, a protein involved in the formation of tight junctions, also binds to the N-terminus of AF6 and competes with Ras binding (46). These studies suggest that AF6 may participate in the regulation of cell– cell contacts by Ras-mediated interaction with ZO-1. Additional studies designed to confirm the in vivo interaction of Rit and Rin with AF6 are in progress. As important as the identification of potential Rit and Rin binding proteins was the observation that the known Ras effectors c-Raf-1, A-Raf, B-Raf, RIN1, and p110 (the catalytic subunit of phosphatidylinositol 3-kinase) do not interact with active GTP-bound Rit or Rin in these assays (Table II). The ability of Rit and Rin to bind to an isolated Ras-binding sequence of Raf but not to full-length versions of the three Raf kinases is understandable in light of recent studies in which
the interaction of Ras with Raf-1 was found to be dependent upon two distinct binding regions on each protein (47). We suspect that Rit and Rin fail to interact with both of these Raf-1 domains. However, while similar results have been reported for the Ras-related GTPase TC21 (34), a recent report using in vivo MAPK activation assays suggests that TC21 is capable of both Raf binding and activation (48). These studies highlight a limitation of two-hybrid analysis. It will be important to extend our analysis using in vivo MAPK activation assays to confirm the inability of these proteins to activate Raf. Thus, while Rit and Rin appear to share some binding partners with Ras, they apparently fail to interact with the majority of known Ras effectors including the Raf kinases. Taken together, these results show that Rit and Rin possess functions that are both common to and distinct from those of Ras. Studies are underway to determine whether chronically active, GTP-bound mutant versions of Rit and Rin possess oncogenic activity. While our studies have not yet defined a cellular function for either Rit or Rin, precedent from the study of other Ras-like GTPases suggests a role for both proteins in the regulation of signal transduction cascades by controlling the assembly of protein signaling complexes at specific cellular membrane locations. Although a large number of Ras-like GTPases have been identified, they participate in regulating many aspects of cell physiology, and inclusion in this superfamily does not define a cellular role for the Rit and Rin proteins. Recently, it has been shown that Ras subfamily members are involved in growth factor-dependent signaling in many tissues including neurons. Given the close structural homology, particularly within the effector binding loop, between the Rit/Rin proteins and Ras, and their ability to interact with effector proteins that are both common to but also likely distinct from those for Ras, it is plausible that the Rit and Rin proteins may serve as important signal transducers for as yet undetermined cellular stimuli in controlling cell proliferation or differentiation. ACKNOWLEDGMENTS We thank Stephan Koh for help with the initial prenyl library screening, Dr. Ian Macara for the kind gift of the pKH3 expression vector, Nan Guo for help in producing recombinant Rit and Rin proteins, Dr. Peter Spielmann for assistance with curve fitting our nucleotide release data, Dr. Channing Der and Dr. Anne Vojtek for the kind gift of the Ras effector protein yeast two-hybrid bait vectors, Dr. Philip James for the PJ69-4A yeast and yeast two-hybrid vectors, and members of the laboratory for comments on the manuscript.
REFERENCES 1. Bourne, H. R., Sanders, D. A., and McCormick, F. (1990) Nature 348, 125–132. 2. Boguski, M. S., and McCormick, F. (1993) Nature 366, 643– 654.
BIOCHEMICAL ANALYSIS OF THE Rit AND Rin GTPases 3. Zerial, M., and Huber, L. A. (1995) in Guidebook to the Small GTPases, Guide Book Series (Zerial, M., and Huber, L. A., Eds.), Oxford Univ. Press, Oxford, UK. 4. Bourne, H. R., Sanders, D. A., and McCormick, F. (1991) Nature 349, 117–127. 5. Hancock, J. F., Magee, A. I., Childs, J. E., and Marshall, C. J. (1989) Cell 57, 1167–1177. 6. Glomset, J. A., and Farnsworth, C. C. (1994) Annu. Rev. Cell. Biol. 10, 181–205. 7. Hancock, J. F., Paterson, H., and Marshall, C. J. (1990) Cell 63, 133–139. 8. Cox, A. D., and Der, C. J. (1997) Biochim. Biophys. Acta 1333, F51–F71. 9. Wassarman, D., Therrien, M., and Rubin, G. (1995) Curr. Opin. Genet. Dev. 5, 44 –50. 10. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd Ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 11. Hakes, D. J., and Dixon, J. E. (1992) Anal. Biochem. 202, 293– 298. 12. Andres, D. A., Shao, H., Crick, D. C., and Finlin, B. S. (1997) Arch. Biochem. Biophys. 346, 113–124. 13. Biermann, B. J., Morehead, T. A., Tate, S. E., Price, J. R., Randall, S. K., and Crowell, D. N. (1994) J. Biol. Chem. 269, 25251–25254. 14. Finlin, B. S., and Andres, D. A. (1997) J. Biol. Chem. 272, 21982–21988. 15. James, G. L., Goldstein, J. L., and Brown, M. S. (1995) J. Biol. Chem. 270, 6221– 6226. 16. Northup, J. K., Smigel, M. D., and Gilman, A. G. (1982) J. Biol. Chem. 257, 11416 –11423. 17. Pan, J. Y., Sanford, J. C., and Wessling-Resnick, M. (1996) J. Biol. Chem. 271, 1322–1328. 18. Hall, A., and Self, A. J. (1986) J. Biol. Chem. 261, 10963–10965. 19. Lerosey, I., Chardin, P., de Gunzburg, J., and Tavitian, A. (1991) J. Biol. Chem. 266, 4315– 4321. 20. James, P., Halladay, J., and Craig, E. A. (1996) Genetics 144, 1425–1436. 21. Altschul, S. F., and Lipman, D. J. (1990) Proc. Natl. Acad. Sci. USA 87, 5509 –5513. 22. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990) J. Mol. Biol. 215, 403– 410. 23. Wes, P. D., Yu, M., and Montell, C. (1996) EMBO J. 15, 5839 – 5848. 24. Lee, C. H. J., Della, N. G., Chew, C. E., and Zack, D. J. (1996) J. Neurosci. 16, 6784 – 6794. 25. Gilman, A. G. (1987) Annu. Rev. Biochem. 56, 615– 649. 26. Ren, M., Drivas, G., D’Eustachio, P., and Rush, M. G. (1993) J. Cell Biol. 120, 313–323. 27. Joneson, T., and Bar-Sagi, D. (1997) J. Mol. Med. 75, 587–593. 28. Kuriyama, M., Harada, N., Kuroda, S., Yamamoto, T., Nakafuku, M., Iwamatsu, A., Yamamoto, D., Prasad, R., Croce, C.,
29. 30.
31. 32.
33. 34.
35. 36. 37. 38. 39. 40. 41.
42.
43.
44.
45.
46.
47.
48.
219
Canaani, E., and Kaibuchi, K. (1996) J. Biol. Chem. 271, 607– 610. Han, L., and Colicelli, J. (1995) Mol. Cell. Biol. 15, 1318 –1323. Rodriguez-Viciana, P., Warne, P. H., Dhand, R., Vanhaesebroeck, B., Gout, I., Fry, M. J., Waterfield, M. D., and Downward, J. (1994) Nature 370, 527–532. Feig, L. A., Urano, T., and Cantor, S. (1996) Trends Biochem. Sci. 21, 438 – 441. Rodriguez-Viciana, P., Warne, P. H., Khwaja, A., Marte, B. M., Pappin, D., Das, P., Waterfield, M. D., Ridley, A., and Downward, J. (1997) Cell 89, 457– 467. Barbacid, M. (1987) Annu. Rev. Biochem. 56, 779 – 827. Graham, S. M., Vojtek, A. B., Huff, S. Y., Cox, A. D., Clark, G. J., Cooper, J. A., and Der, C. J. (1996) Mol. Cell. Biol. 16, 6132– 6140. Zhu, J., Reynet, C., Caldwell, J. S., and Kahn, C. R. (1995) J. Biol. Chem. 270, 4805– 4812. McCormick, F. (1989) Cell 56, 5– 8. Vojtek, A. B., and Der, C. J. (1998) J. Biol. Chem. 273, 19925– 19928. Urano, A., Emkey, R., and Feig, L. A. (1996) EMBO J. 15, 810 – 816. Kishida, S., Koyama, S., Matsubara, K., Kishida, M., Matsuura, Y., and Kikuchi, A. (1997) Oncogene 15, 2899 –2907. Jiang, H., Luo, J.-Q., Urano, T., Frankel, P., Lu, Z., Foster, D. A., and Feig, L. A. (1995) Nature 378, 409 – 412. Nakashima, S., Morinaka, K., Koyama, S., Ikeda, M., Kishida, M., Okawa, K., Iwamatsu, A., Kishida, S., and Kikuchi, K. (1999) EMBO J. 18, 3629 –3642. Sawamoto, K., Winge, P., Koyama, S., Hirota, Y., Yamada, C., Miyao, S., Yoshikawa, S., Jin, M.-H., Kikuchi, A., and Hideyuki, O. (1999) J. Cell Biol. 146, 361–372. Prasad, R., Gu, Y., Alder, H., Nakamura, T., Canaani, O., Saito, H., Huebner, K., Gale, R. P., Nowell, P. C., Kuriyama, K., Mikazaki, Y., Croce, C. M., and Canaani, E. (1993) Cancer Res. 53, 5624 –5628. Miyamoto, H., Nihonmatsu, I., Kondo, S., Ueda, R., Togashi, S., Hirata, K., Ikegami, Y., and Yamamoto, D. (1995) Genes Dev. 9, 612– 625. Hock, B., Bohme, B., Karn, T., Yamamoto, T., Kaibuchi, K., Holtrich, U., Holland, S., Pawson, T., Rubsamen-Waigmann, H., and Strebhardt, K. (1998) Proc. Natl. Acad. Sci. USA 95, 9779 – 9784. Yamamoto, T., Harada, N., Kano, K., Taya, S., Canaani, E., Matsuura, Y., Mizoguchi, A., Ide, C., and Kaibuchi, K. (1997) J. Cell Biol. 139, 785–795. Brtva, T. R., Drugan, J. K., Ghosh, S., Terrell, R. S., CampbellBurk, S., Bell, R. M., and Der, C. J. (1995) J. Biol. Chem. 270, 9809 –9812. Rosairio, M., Paterson, H. F., and Marshall, C. J. (1999) EMBO J. 18, 1270 –1279.