HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 29, 27629-27637, July 20, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/29/27629    most recent M101727200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hall, B. E. Articles by Bar-Sagi, D. Search for Related Content PubMed PubMed Citation Articles by Hall, B. E. Articles by Bar-Sagi, D. Social Bookmarking                      What's this?

Structure-based Mutagenesis Reveals Distinct Functions for Ras Switch 1 and Switch 2 in Sos-catalyzed Guanine Nucleotide Exchange^*

Brian E. Hall, Shao Song Yang^§, P. Ann Boriack-Sjodin^¶, John Kuriyan^¶, and Dafna Bar-Sagi^**

From the ^** Department of Molecular Genetics and Microbiology and the Graduate Programs in  Molecular Pharmacology and ^§ Molecular and Cellular Biology, State University of New York at Stony Brook, New York 11794-5222 and the ^¶ Laboratories of Molecular Biophysics and the Howard Hughes Medical Institute, Rockefeller University, New York, New York 10021

Received for publication, February 23, 2001, and in revised form, April 23, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Ras GTPases function as binary switches in signaling pathways controlling cell growth and differentiation. The guanine nucleotide exchange factor Sos mediates the activation of Ras in response to extracellular signals. We have previously solved the crystal structure of nucleotide-free Ras in complex with the catalytic domain of Sos (Boriack-Sjodin, P. A., Margarit, S. M., Bar-Sagi, D., and Kuriyan, J. (1998) Nature 394, 337-343). The structure demonstrates that Sos induces conformational changes in two loop regions of Ras known as switch 1 and switch 2. In this study, we have employed site-directed mutagenesis to investigate the functional significance of the conformational changes for the catalytic function of Sos. Switch 2 of Ras is held in a very tight embrace by Sos, with almost every external side chain coordinated by Sos. Mutagenesis of contact residues at the switch 2-Sos interface shows that only a small set of side chains affect binding, with the most important contact being mediated by tyrosine 64, which is buried in a hydrophobic pocket of Sos in the Ras·Sos complex. Substitutions of Ras and Sos side chains that are inserted into the Mg²⁺- and nucleotide phosphate-binding site of switch 2 (Ras Ala⁵⁹ and Sos Leu⁹³⁸ and Glu⁹⁴²) have no effect on the catalytic function of Sos. These results indicate that the interaction of Sos with switch 2 is necessary for tight binding, but is not the critical driving force for GDP displacement. The structural distortion of switch 1 induced by Sos is mediated by a small number of specific contacts between highly conserved residues on both Ras and Sos. Mutations of a subset of these residues (Ras Tyr³² and Tyr⁴⁰) result in an increase in the intrinsic rate of nucleotide dissociation from Ras and impair the binding of Ras to Sos. Based on this analysis, we propose that the interactions of Sos with the switch 1 and switch 2 regions of Ras have distinct functional consequences: the interaction with switch 2 mediates the anchoring of Ras to Sos, whereas the interaction with switch 1 leads to disruption of the nucleotide-binding site and GDP dissociation.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Ras proteins function as binary switches in signaling pathways controlling cell proliferation and differentiation by cycling between inactive GDP- and active GTP-bound states. The conversion of Ras·GDP to Ras·GTP following receptor activation is catalyzed by guanine nucleotide exchange factors (GEFs)¹; and in the context of receptor tyrosine kinases, this process is dependent on the Sos (Son of sevenless) GEF (1). Mammalian cells contain two sos genes, sos1 and sos2, encoding highly related M_r 150,000 proteins (2). The Ras-specific guanine nucleotide exchange activity of Sos proteins is mediated by a central region of ~450 amino acids that is structurally and functionally related to the catalytic domain of the Saccharomyces cerevisiae Ras exchanger Cdc25 and has been designated the Cdc25 homology domain (2-4). This domain is flanked at the N terminus by 600 amino acids containing a Dbl homology domain and a pleckstrin homology domain and at the C terminus by 300 amino acids containing proline-rich SH3 domain-binding sites. Although the Cdc25 homology domain of Sos is necessary and sufficient to catalyze guanine nucleotide exchange on Ras, both the Sos N and C termini are required for full biological activity. The C-terminal domain mediates the ligand-dependent recruitment of Sos to activated receptors via the adaptor molecule Grb2 (5-9), and the Dbl homology and pleckstrin homology domains contribute to the stable association of Sos with activated receptors (10-16). In addition, the Sos N and C termini have been implicated in the regulation of guanine nucleotide exchange activity through intramolecular interactions (12, 15, 17, 18).

Most biochemical analysis of Ras GEFs carried out to date have been performed using the Cdc25 homology domains of the yeast GEFs Cdc25 and Sdc25 and the mammalian GEF GRF/CDC25^Mm. Kinetic studies have indicated that the GEF-stimulated guanine nucleotide exchange reaction involves the transient formation of a ternary Ras·nucleotide·GEF complex, followed by the formation of a stable binary Ras·GEF complex. Nucleotide then binds to this binary complex, causing the release of GEF from the resulting ternary complex (19). Mutagenesis studies have led to the identification of three regions in Ras that are important for its activation by GEFs: the switch 1 region (amino acids 25-40), the switch 2 region (amino acids 57-75), and a short region spanning amino acids 100-110 (20-32). Although specific residues within these regions have been shown to be critical for either GEF binding or GEF responsiveness, the molecular mechanisms underlying these requirements have not been established. Likewise, the sites on GEFs that are required for interactions with Ras remain undefined.

We have previously determined the crystal structure of human Ha-Ras (Ras) complexed with the catalytic region of Sos, hereafter referred to as Sos (33). The structure reveals that the interface between Ras and Sos is extensive, encompassing >30 contacting side chains. The switch 1 region of Ras is displaced by a helical hairpin structure of Sos (-H and -I), resulting in the opening of the nucleotide-binding site. The switch 2 region of Ras is held tightly by Sos through a cluster of hydrophobic interactions surrounded by polar and charged interactions. The structural changes induced by these interactions result in the insertion of side chains of Ras and Sos into the nucleotide phosphate- and Mg²⁺-binding sites. In this study, we have utilized the information derived from the crystal structure to test specific predictions concerning the functional interactions between Ras and Sos. By analyzing the biochemical properties of Ras and Sos proteins harboring single amino acid substitutions, we have identified residues that play a role in the activation of Ras by Sos.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Plasmids-- All Ras constructs contained full-length Ha-Ras (amino acids 1-188), and all Sos constructs contained the Sos catalytic domain (amino acids 564-1049). Point mutations in both Ras and Sos were produced via a polymerase chain reaction-based strategy. Primers encoding each individual mutation were mixed with the template gene in the presence of Pfu polymerase (Stratagene) and exposed to 30 rounds of polymerase chain reaction in a Robocycler Gradient 96 (Stratagene). Mutations were confirmed by dideoxynucleotide sequencing using a T7 Sequenase kit (Amersham Pharmacia Biotech). Ras constructs were subcloned into the bacterial expression vector pGEX-2T (Amersham Pharmacia Biotech) at cloning sites BamHI and EcoRI, which allow the expression of genes fused to an N-terminal glutathione S-transferase (GST) tag. Sos catalytic domain constructs were cloned into the bacterial expression vector pET-28 (Novagen) at cloning sites BamHI and XhoI, which allow the expression of genes fused to a polyhistidine tag and a T7 epitope tag. The bacterial expression plasmids were transformed into the BL21 strain of Escherichia coli.

Protein Purification-- GST-Ras fusion proteins and polyhistidine-tagged Sos proteins were expressed in E. coli BL21 by induction with 500 µM isopropyl-1-thio--D-galactopyranoside at a cell density absorbance of A₆₀₀ = 0.5. Pellets were resuspended in buffer containing 20 mM Tris (pH 7.6), 200 mM NaCl, 2 µM phenylmethylsulfonyl fluoride, 1% aprotinin, 10 µg/ml leupeptin, 10 mM benzamidine, 10 µg/ml soybean trypsin inhibitor, and 10 µg/ml pepstatin and sonicated using a Branson Cell Disrupter 200. Expressed proteins were purified by affinity chromatography. Lysates containing polyhistidine-tagged Sos proteins were incubated with charged nickel resin (Invitrogen), and lysates containing GST-Ras fusion proteins were incubated with glutathione-Sepharose 4B resin (Amersham Pharmacia Biotech) at 4 °C for 30 min. The resins were washed five times in the resuspension buffer. Ras proteins were eluted with buffer containing 10 mM glutathione in 20 mM Tris (pH 7.6) and 200 mM NaCl, and Sos proteins were eluted with buffer containing 200 mM imidazole in 20 mM Tris (pH 7.6) and 200 mM NaCl. Eluates were dialyzed against buffer containing 20 mM Tris (pH 7.6) and 200 mM NaCl.

Ras-Sos Binding Assay-- GST-Ras fusion protein (100 nM) and the indicated amount of T7-tagged Sos protein were added to 1 ml of binding buffer (20 mM Tris (pH 7.6), 50 mM NaCl, 1 mM dithiothreitol, 5 mM EDTA, and 1% Triton X-100) and incubated at 4 °C for 30 min. Following incubation, 60 µl of 1:1 slurry of glutathione-Sepharose 4B beads resuspended in binding buffer were added to each sample. Samples were incubated for an additional 20 min at 4 °C. Beads were subsequently pelleted, washed five times with binding buffer, and resuspended in SDS-polyacrylamide gel electrophoresis sample buffer. Proteins were separated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose. Western blots were probed with anti-T7 antibody (diluted 1:10,000; Novagen) to detect Sos. Changes in binding affinity were assessed via comparing the amount of Sos precipitated by mutant and wild-type proteins.

Ras Nucleotide Dissociation Assay-- GST-Ras fusion proteins were loaded with labeled guanine nucleotide by incubating 40 pmol of Ras in 100 µl of buffer containing 20 mM Tris (pH 7.6), 50 mM NaCl, 1 mg/ml bovine serum albumin, 1 mM dithiothreitol, and 1 mM EDTA with 200 pmol of [³H]GDP (25-50 Ci/mmol; PerkinElmer Life Sciences) or 200 pmol of [-³²P]GTP (25 Ci/mmol; ICN) for 15 min at 30 °C on a Thermoshaker. Following incubation, MgCl₂ was added, and the reaction mixture was incubated on ice for 10 min. The dissociation reaction was initiated by adding 100 µl of ice-cold buffer containing 20 mM Tris (pH 7.6), 50 mM NaCl, 1 mg/ml bovine serum albumin, 1 mM dithiothreitol, 5 mM MgCl₂, and 0.1 mM GTPS with or without Sos protein. Reactions were incubated at 30 °C, with 20-µl aliquots being removed at indicated time points, and terminated by adding 1 ml of ice-cold stop solution (20 mM Tris (pH 7.6), 50 mM NaCl, and 15 mM MgCl₂). Stopped reactions were filtered through 0.45-µm nitrocellulose filter discs (Millipore HAWPO2500), which were washed twice with 15 ml of ice-cold stop solution. Radioactivity retained was measured in a Packard Instrument 1900TR liquid scintillation analyzer using 10 ml of Scintiverse scintillation fluid/filter. Data points were plotted and curve-fitted using the SigmaPlot Version 5.0 program (SPSS Inc.).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Role of Switch 2-mediated Hydrophobic Contacts in Ras-Sos Binding-- Fig. 1A illustrates the general architecture of the Ras·Sos complex. As can been seen, the interaction between the two proteins is predominantly mediated by the switch 1 and switch 2 regions of Ras (33). The crystal structure of the Ras·Sos complex shows that the switch 2 region of Ras is coordinated extensively by Sos through hydrophobic, polar, and charged interactions. We have employed site-directed mutagenesis to examine the role of these interactions in Ras-Sos binding. In the Ras·Sos complex, the side chain of Tyr⁶⁴ of Ras is inserted into a hydrophobic pocket created by Ile⁸²⁵, Leu⁸⁷², and Phe⁹²⁹ of Sos (Fig. 1B). To determine the contribution of these hydrophobic interactions to Ras-Sos binding, we first mutated Tyr⁶⁴ of Ras to Ala (Y64A). GST fusion proteins of wild-type Ras and Y64A Ras were immobilized on glutathione beads and incubated with increasing concentrations of purified T7-tagged Sos. The amount of bound Sos was then assessed by immunoblotting with anti-T7 antibodies. The Y64A mutation resulted in a decrease of at least 50-fold in the apparent binding affinities of Ras for Sos (Fig. 1C), but had no effect on the intrinsic rate of nucleotide dissociation of Ras alone (data not shown). Next, we mutated Phe⁹²⁹ of Sos to Ala (F929A) and performed a similar binding assay using wild-type Ras. Similar to the Y64A mutation, the F929A Sos mutant displayed a reduction of >50-fold in binding affinity for Ras (Fig. 1C). These observations indicate that Tyr⁶⁴ mediates hydrophobic contacts that are essential for the formation of a stable Ras·Sos complex. In contrast, polar and charged interactions appear to contribute much less to the binding affinity of Ras for Sos, as indicated by the observation that single alanine substitutions of Sos residues Glu¹⁰⁰², Thr⁹³⁵, and Arg⁸²⁶ exerted a relatively small effect on Ras binding and activation (Fig. 2, A-C).

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Role of switch 2-mediated hydrophobic interactions in Ras-Sos binding. A, ribbon diagram of Ha-Ras (Ras) with the exchange factor region of human Sos1 (Sos). Ras (residues 1-166) is shown in orange, with switch regions (residues 25-40 and 57-75) highlighted in yellow. The Sos catalytic domain (residues 752-1044) is shown in green, and the Sos N-terminal domain (residues 568-741) is shown in light green. Helices that have been mutated in this study are indicated. Images were created from Protein Data Bank code 1BKD using the Midas ribbonjr command. B, close-up view of contact residues that form the hydrophobic anchor. The backbone traces of Ras switch 2 (residues 57-70) and Sos (-B, -D, and -H; residues 822-827, 870-875, and 927-933) are shown in yellow and green, respectively. Side chain groups of Ras and Sos residues are colored red and white, respectively. C, effects of mutations of hydrophobic side chains on Ras-Sos binding. GST fusion proteins of wild-type Ras and mutant Ras (100 nM) were incubated for 20 min at 4 °C in the presence of increasing concentrations of T7-tagged wild-type or mutant Sos protein. Ras·Sos complexes were isolated by incubation with glutathione-Sepharose beads. Bound material was eluted, and the amount of associated Sos was determined by blotting with anti-T7 antibodies. Data shown are from a single experiment and are representative of three independent experiments.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Role of switch 2-mediated charged and polar interactions in Sos binding and guanine nucleotide exchange activity. A, ribbon representation of the interface between Ras switch 2 and Sos. The backbone traces of Ras switch 2 (residues 57-70) and Sos (-B, -K, and -H; residues 805-832, 935-944, and 998-1006) are shown in yellow and green, respectively. Side chain groups of Sos residues that have been mutated and the contacting Ras residues are colored white and red, respectively. B, effects of alanine mutations of Sos residues Glu1002, Arg826, and Thr935 on Ras binding. GST-wild-type Ras fusion protein (100 nM) was incubated with increasing concentrations of the indicated T7-tagged wild-type or mutant Sos proteins. Ras·Sos complexes were isolated by incubation with glutathione-Sepharose beads. Bound material was eluted, and the amount of associated Sos was determined by blotting with anti-T7 antibodies. Data shown are from a single experiment and are representative of three independent experiments. C, effects of alanine mutations of Sos residues Glu1002, Arg826, and Thr935 on guanine nucleotide exchange activity. Ras protein (40 pmol) was loaded with [3H]GDP and incubated with 40 pmol of the indicated mutant Sos proteins and excess unlabeled GTPS for 10 min at 30 °C. Following incubation, the radioactivity remaining bound to Ras was determined by a nitrocellulose filter assay. Results are expressed as percentages of the [3H]GDP dissociation stimulated by wild-type Sos. Each value is the mean ± S.D. of three independent experiments.

Role of Mg²⁺ and Phosphate Displacement in the Guanine Nucleotide Exchange Activity of Sos-- The backbone conformation of the switch 2 region of Ras is significantly altered in the presence of Sos. One of the consequences of this structural remodeling is that the orientation of Ala⁵⁹ of Ras is changed such that its methyl side chain now occupies the Mg²⁺-binding site. To investigate whether the occlusion of the Mg²⁺-binding site resulting from this conformational change is of consequence for the guanine nucleotide exchange reaction, we replaced Ala⁵⁹ of Ras with Gly (A59G). This substitution was chosen because glycine is the only residue lacking a methyl side chain and therefore would be expected not to occlude the Mg²⁺-binding site. The A59G Ras protein appeared equivalent to wild-type Ras in its ability to bind to Sos (Fig. 3A). We next tested whether the A59G mutation influences Sos-mediated guanine nucleotide exchange. For this purpose, purified wild-type and A59G Ras proteins were first complexed with ³H-labeled GDP or ³²P-labeled GTP and then diluted in buffer containing Sos and excess unlabeled GTP. Aliquots were taken at different intervals, and the amount of protein-bound radioactive nucleotide was measured by nitrocellulose filtration and scintillation counting. As shown in Fig. 3B, the rate of Sos-catalyzed GDP dissociation was not altered by the A59G mutation. In contrast, the A59G mutant displayed a >50% inhibition of Sos-catalyzed GTP dissociation. This inhibitory effect was partially abolished by lowering the free Mg²⁺ concentration from 10 mM to 1 µM (Fig. 3C), suggesting that the displacement of Mg²⁺ by Ala⁵⁹ is a critical step in Sos-catalyzed GTP (but not GDP) dissociation.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Role of Mg2+ and phosphate displacement in Sos-catalyzed guanine nucleotide dissociation. A, effects of the A59G Ras mutation on Sos binding. GST fusion proteins of wild-type Ras and A59G Ras (100 nM) were incubated with increasing amounts of T7-tagged Sos protein. Ras·Sos complexes were isolated by incubation with glutathione-Sepharose beads. Bound material was eluted, and the amount of associated Sos was determined by blotting with anti-T7 antibodies. Data shown are from a single experiment and are representative of three independent experiments. B, effects of the A59G Ras mutation on Sos-mediated GDP and GTP dissociation. Wild-type (circles) and A59G (inverted triangles) Ras proteins (40 pmol) were loaded with [3H]GDP or [32P]GTP as indicated and incubated with (open symbols) or without (closed symbols) Sos protein (40 pmol) and excess unlabeled GTPS at 30 °C. Aliquots were removed at the indicated time points, and the radioactivity remaining bound to Ras was determined by a nitrocellulose filter assay. Data shown are from a single experiment and are representative of three independent experiments. Results are presented as percentages of the radioactivity bound to Ras at time 0. The solid lines represent a single exponential fit to the data. C, Sos-catalyzed GTP dissociation from A59G Ras at low Mg2+ concentration. A59G Ras (40 pmol) was loaded with [32P]GTP and incubated with (open symbols) or without (closed symbols) Sos protein (40 pmol) and excess unlabeled GTPS at 30 °C. Incubations were carried out in the presence of 10 mM (circles) or 1 µM (inverted triangles) free Mg2+. Aliquots were removed at the indicated time points, and the radioactivity remaining bound to A59G Ras was determined by a nitrocellulose filter assay. Data shown are from a single experiment and are representative of three independent experiments. Results are presented as percentages of the radioactivity bound to Ras at time 0. The solid lines represent a single exponential fit to the data. D, effects of alanine mutations of Sos residues Leu938 and Glu942 on Ras binding. GST fusion proteins of wild-type Ras and A59G Ras (100 nM) were incubated with increasing concentrations of the indicated T7-tagged mutant Sos proteins. Ras·Sos complexes were isolated by incubation with glutathione-Sepharose beads. Bound material was eluted, and the amount of associated Sos was determined by blotting with anti-T7 antibodies. Data shown are from a single experiment and are representative of three independent experiments. E, effects of alanine mutations of Sos residues Leu938 and Glu942 on stimulation of GDP and GTP dissociation. Ras protein (40 pmol) was loaded with [3H]GDP (dark-gray bars) or [32P]GTP (light-gray bars) and incubated with 40 pmol of the indicated mutant Sos proteins and excess unlabeled GTPS for 10 min at 30 °C. Following incubation, the radioactivity remaining bound to Ras was determined by a nitrocellulose filter assay. Results are expressed as percentages of the [3H]GDP or [32P]GTP dissociation stimulated by wild-type Sos. Each value is the mean ± S.D. of three independent experiments.

Helix -H of Sos presents two residues, Leu⁹³⁸ and Glu⁹⁴², that interfere with Mg²⁺ and phosphate binding, respectively. Leu⁹³⁸ is inserted into the Mg²⁺-binding site, and Glu⁹⁴² forms a hydrogen bond with Ser¹⁷ of Ras, thereby displacing the -phosphate of GDP. To determine the contribution of these interactions to the guanine nucleotide exchange reaction, Leu⁹³⁸ and Glu⁹⁴² were both replaced by alanines. As shown in Fig. 3 (D and E), these mutations interfered only slightly with the exchange and binding activities of Sos for either wild-type Ras or the A59G mutant. These results suggest that the destabilization of Mg²⁺ and phosphate coordination resulting from the molecular contacts between Sos and the switch 2 region of Ras is not critical for Sos-catalyzed GDP dissociation.

Role of Switch 1 in Guanine Nucleotide Exchange-- The most pronounced structural change in Ras following the binding of Sos is the displacement of switch 1 from the nucleotide-binding site. This change is mediated by the insertion of a helical hairpin formed by helices H and I of Sos into the switch 1 region. Tyr³² of Ras forms hydrophobic contacts with Lys⁹³⁹ of Sos, and Tyr⁴⁰ of Ras makes a stacking interaction with His⁹¹¹ of Sos (Fig. 4A). To investigate the role of the switch 1 region in the interaction with Sos, Tyr³² and Tyr⁴⁰ of Ras were mutated to Ser (Y32S) and Ala (Y40A), respectively. As shown in Fig. 4B, both the Y32S and Y40A mutations reduced the binding of Sos to Ras, suggesting that these residues are important for Ras-Sos recognition. Consistent with this interpretation, alanine substitutions of Sos Lys⁹³⁹ and His⁹¹¹ also resulted in a decrease in Ras-Sos binding (Fig. 4B). In addition, the Y32S and Y40A Ras mutations accelerated the rate of intrinsic GDP/GTP exchange by factors of 2 and 4, respectively, indicating that the contacts mediated by these residues contribute to nucleotide stabilization (Fig. 4C). The Y40A mutation had no significant effect on Sos-catalyzed guanine nucleotide exchange (Fig. 4D), whereas the disruption of the contact between Tyr³² of Ras and Lys⁹³⁹ of Sos reduced the sensitivity of Ras to the exchange activity of Sos (Fig. 4, E and F). This effect persisted even when the concentration of Sos in the reaction was increased 10-fold (Fig. 4, E and F), indicating that the defect in Sos-mediated guanine nucleotide exchange displayed by these mutants cannot be attributed solely to a reduction in binding affinity. In an attempt to examine whether switch 1 and switch 2 mutations have an additive effect on Sos-catalyzed exchange, we generated a G59A/Y40A Ras double mutant. This mutant displayed a severe defect in Sos binding and thus could not be utilized for further analysis.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Role of switch 1-mediated interactions in Sos binding and guanine nucleotide exchange activity. A, ribbon representation of the interface between Ras switch 1 and Sos. The backbone traces of Ras switch 1 (residues 26-20) and Sos (-G, -H, and -I; residues 909-979) are shown in yellow and green, respectively. Side chain groups of Ras residues that have been mutated and the contacting Sos residues are colored red and white, respectively. B, effects of Y32S and Y40A Ras mutations and K939A and H911A Sos mutations on Ras-Sos binding. GST fusion proteins of wild-type Ras and mutant Ras (100 nM) were incubated with increasing concentrations of T7-tagged wild-type or mutant Sos proteins. Ras·Sos complexes were isolated by incubation with glutathione-Sepharose beads. Bound material was eluted, and the amount of associated Sos was determined by blotting with anti-T7 antibodies. Data shown are from a single experiment and are representative of three independent experiments. C, effects of Y32S and Y40A mutations on the intrinsic rate of guanine nucleotide dissociation. Wild-type (circles), Y32S (triangles), and Y40A (squares) Ras proteins (40 pmol) were loaded with [3H]GDP and incubated with excess unlabeled GTPS for the indicated intervals at 30 °C. Following incubation, the radioactivity remaining bound to Ras was determined by a nitrocellulose filter assay. Data shown are from a single experiment and are representative of three independent experiments. Results are presented as percentages of the radioactivity bound to Ras at time 0. The solid lines represent a single exponential fit to the data. D, effects of the Y40A Ras mutation on Sos-mediated GDP dissociation. Wild-type (circles) and Y40A (inverted triangles) Ras proteins (40 pmol) were loaded with [3H]GDP and incubated with (open symbols) or without (closed symbols) Sos protein and excess unlabeled GTPS at 30 °C. E, effects of the Y32S Ras mutation on Sos-mediated GDP dissociation. Wild-type (circles) and Y32S (inverted triangles and squares) Ras proteins (40 pmol) were loaded with [3H]GDP and incubated with 40 pmol (open circles and open inverted triangles, respectively) or 400 pmol (Ras Y32S only; open squares) of Sos protein or without Sos protein (closed symbols) and excess unlabeled GTPS at 30 °C. F, effects of the K939A Sos mutation on Sos-mediated GDP dissociation. Wild-type Ras protein (40 pmol) was loaded with [3H]GDP and incubated with 40 pmol of wild-type or mutant Sos protein (open circles and open inverted triangles, respectively) or 400 pmol of mutant Sos protein (open squares) or without Sos protein (closed circles) and excess unlabeled GTPS at 30 °C. For all experiments described for C-F, aliquots were removed at the indicated time points, and the radioactivity remaining bound to Ras was determined by a nitrocellulose filter assay. Data shown are from a single experiment and are representative of three independent experiments. Results are presented as percentages of the radioactivity bound to Ras at time 0. The solid lines represent a single exponential fit to the data.

Importance of the N-terminal Domain for the Catalytic Activity of Sos-- The crystal structure of the Ras·Sos complex identifies two structural domains within the exchange factor region of Sos: an N-terminal domain (residues 568-741) that does not have direct contacts with Ras and a C-terminal domain containing all the residues that interact with Ras. The N-terminal domain consists of six -helices, two of which, helices -1 and -2, form a hydrophobic groove containing Leu⁶⁰⁹ and Phe⁶²³, into which two conserved hydrophobic residues from the helical hairpin, Ile⁹⁵⁶ and Phe⁹⁵⁸, are inserted (Fig. 5A). These hydrophobic interactions would be predicted to be important for the stability and correct orientation of the helical hairpin structure. To investigate the significance of the interaction between the N-terminal domain and the helical hairpin structure for the catalytic activity of Sos, Ile⁹⁵⁶ of the helical hairpin and Phe⁶²³ of the N-terminal domain were individually mutated to Glu (I956E and F623E, respectively). As shown in Fig. 5 (B and C), the I956E Sos mutant was slightly defective in Ras binding, but displayed a pronounced decrease in exchange activity. The F623E Sos mutant showed no apparent defect in Ras binding (Fig. 5B), but it did display a decrease in exchange activity (Fig. 5D). These results indicate that the N-terminal domain plays a role in maintaining Sos in a catalytically active conformation.

View larger version (34K): [in this window] [in a new window]   Fig. 5.   Role of the N-terminal domain of Sos in Ras binding and guanine nucleotide exchange activity. A, ribbon representation of the interface between the helical hairpin and the N-terminal domain of Sos. The backbone traces of the helical hairpin (-H and -I; residues 921-979) and the N-terminal domain (residues 603-741) are shown in green and light green, respectively. Switch 1 of Ras (residues 26-40) is shown in yellow. Side chain groups of -1 and -2 residues that form the N-terminal domain hydrophobic groove and -I residues from the helical hairpin inserted into the groove are colored white. B, effects of the I956E and F623E Sos mutations on Ras binding. GST-wild-type Ras fusion proteins (100 nM) were incubated with increasing concentrations of the indicated T7-tagged wild-type and mutant Sos proteins. Ras·Sos complexes were isolated by incubation with glutathione-Sepharose beads. Bound material was eluted, and the amount of associated Sos was determined by blotting with anti-T7 antibodies. Data shown are from a single experiment and are representative of three independent experiments. C, effects of the I956E mutation on Sos-catalyzed GDP dissociation. Wild-type Ras protein (40 pmol) was loaded with [3H]GDP and incubated with 40 pmol of wild-type or mutant Sos protein (open circles and open inverted triangles, respectively) or 400 pmol of mutant Sos protein (open squares) or without Sos protein (closed circles) and excess unlabeled GTPS at 30 °C. D, effects of the F623E mutation on Sos-catalyzed GDP dissociation. Wild-type Ras protein (40 pmol) was loaded with [3H]GDP and incubated with 40 pmol of wild-type or mutant Sos protein (open circles and open inverted triangles, respectively) or 400 pmol of mutant Sos protein (open squares) or without Sos protein (closed circles) and excess unlabeled GTPS at 30 °C. For the experiments described for C and D, aliquots were removed at the indicated time points, and the radioactivity remaining bound to Ras was determined by a nitrocellulose filter assay. Data shown are from a single experiment and are representative of three independent experiments. Results are presented as percentages of the radioactivity bound to Ras at time 0. The solid lines represent a single exponential fit to the data.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In this study, we have investigated the interaction of Ras with its GEF Sos using structure-guided mutagenesis. In agreement with earlier predictions, the three-dimensional structure of the Ras·Sos complex identifies the switch 2 region of Ras as a potential binding site for Sos (33). The binding interface between switch 2 and Sos consists of both hydrophobic and polar intermolecular contacts. Individual replacement of contact residues with alanine showed that the binding affinity is mediated principally by Tyr⁶⁴ of Ras, which is inserted into a hydrophobic pocket of Sos formed by Phe⁹²⁹, Leu⁸⁷², and Ile⁸²⁵. Phe⁹²⁹ is highly conserved among GEFs for members of the Ras family; and in a recent study, it has been shown that Phe⁶⁴ of Rap1A is important for its interaction with the C3G exchange factor (34). In addition, the structure of Rac1 in complex with Tiam1 shows that Tyr⁶⁴ of Rac is buried in a hydrophobic pocket formed by Tiam1 (35). This suggests that the binding reaction between Ras-related proteins and their GEFs might be governed by similar affinity determinants. The hydrophobic contact region between Ras and Sos is surrounded by polar and charged residues (Glu¹⁰⁰², Arg⁸²⁶, and Thr⁹³⁵) whose mutation to alanine had little effect on binding affinity or guanine nucleotide exchange activity. Thus, rather then contributing directly to binding energy, these polar contacts might be important for the specificity and/or reversibility of the binding interaction between Ras and Sos. A mutation of the residue analogous to Arg⁸²⁶ of Sos in the yeast Ras-specific GEF Cdc25 has been reported to interfere with its binding to Ras (28), and a mutation of the residue analogous to Sos Thr⁹³⁵ in the mammalian Ras-specific GEF CDC25^Mm has been reported to abolish catalytic activity (36). The discrepancy between these findings and our observations might be due to differences in protein constructs and experimental conditions used to assay the biochemical properties of the GEF mutants.

Several structural features of the Ras·Sos complex suggest that Sos might promote the dissociation of guanine nucleotide by perturbing the Mg²⁺ and phosphate groups. Specifically, Sos inserts Leu⁹³⁸ and Glu⁹⁴² into the Mg²⁺- and phosphate-binding sites, respectively. In addition, the position adopted by the methyl side chain of Ala⁵⁹ of Ras in the Ras·Sos complex occludes Mg²⁺ binding. Surprisingly, we have found that E942A, L938A, and A59G mutations had no significant effect on the ability of Sos to stimulate GDP dissociation, suggesting that disruption of Mg²⁺ coordination by these residues is not critical for Sos-mediated exchange reaction. The fact that Glu⁹⁴² is not highly conserved among different Ras GEFs is consistent with this interpretation. That the mechanism of GEF-mediated nucleotide release from Ras is independent, at least in part, of the removal of Mg²⁺ is indicated by the finding that CDC25^Mm induces a significant increase in the GDP dissociation rate even in the presence of saturating concentrations of EDTA (19). The reason for the selective impairment of Sos-mediated GTP dissociation displayed by the A59G Ras mutant is not known. The observation that this defect was partly corrected by lowering the Mg²⁺ concentration suggests that Mg²⁺ coordination and/or GTP-binding properties of this mutant might be altered.

The interaction of Ras with Sos induces the displacement of switch 1 via injection of the helical hairpin structure, moving it away from the nucleotide-binding site. The observed increase in the intrinsic rate of nucleotide dissociation caused by mutations in switch 1 Tyr residues that form side chain-specific interactions with Sos (Y32S and Y40A) indicates the importance of these interactions in stabilizing the bound nucleotide. Moreover, the reduction in Sos-mediated guanine nucleotide exchange activity displayed by the Y32S Ras and K939A Sos mutants highlights the critical role of switch 1 conformational change in the exchange reaction. It has been noted before that substitution of residue 40 of Ras to Ile or Ser leads to an increase in intrinsic exchange rates (30). In addition, we have found that mutation of residue 40 of Ras to Cys similarly causes an increase in the intrinsic rate of nucleotide dissociation (data not shown). This is of particular interest since the Y40C Ras mutant has been extensively used as a partial loss-of-function mutant in studying the biological roles of Ras effector pathways (37). It remains to be established to what extent the signaling activities of this mutant are influenced by its altered properties with respect to guanine nucleotide exchange.

The minimal region of Ras GEFs required for catalytic activity consists of ~250 amino acids containing three regions that are highly conserved among known Ras GEFs (3). However, it has been shown that, in the case of the yeast Ras-specific GEF Cdc25, a larger fragment of ~450 amino acids is necessary for full exchange activity in vivo (38). The N terminus of this fragment (defined previously as the N-terminal domain) (33) contains an additional conserved region of ~50 amino acids found exclusively in Ras-specific GEFs. In the crystal structure of the Ras·Sos complex, two conserved residues within this region, Leu⁶⁰⁹ and Phe⁶²³, form a hydrophobic groove that accommodates the side chains of two conserved hydrophobic residues from the helical hairpin of Sos, Ile⁹⁵⁶ and Phe⁹⁵⁸. The disruption of these hydrophobic interactions was shown to cause a reduction in Sos exchange activity presumably due to the destabilization of the helical hairpin. Because the stability and correct placement of the helical hairpin are principally important for Sos-induced structural changes in switch 1, these results provide additional support for the idea that the movement of switch 1 is a critical aspect of the exchange reaction. Recently, Shimizu et al. (39) described the crystal structure of the Mg²⁺-free form of GDP-bound RhoA. It is difficult to assess the direct relevance of the information derived from this structure to the Sos-mediated nucleotide exchange mechanism because of the differences that exist between the nucleotide-binding properties of Ras and Rho GTPases. For example, the nucleotide-binding affinities of Ras are in the nanomolar range, and Mg²⁺ is required for high affinity binding. In contrast, the nucleotide-binding affinities of the Rho GTPases are in the submicromolar range, and Mg²⁺ is not essential for nucleotide binding (40). In addition, the Mg²⁺-chelating residues of RhoA·GDP are different from those of Ras·GDP, and the switch 1 region of Rho GTPases is more flexible compared with that of Ras (41-44). Nevertheless, it is of interest to note that the conformational changes in switch 1 and switch 2 of RhoA·GDP induced by the absence of Mg²⁺ bear a striking resemblance to the conformation adopted by switch 1 and switch 2 of nucleotide-free Ras when complexed to Sos (33, 39). This similarity raises the possibility that the major function of Mg²⁺ displacement from Ras might be the induction of structural changes that result in a conformation that is compatible with Sos binding. Our mutagenesis studies suggest that the disruption of Mg²⁺ coordination is not dependent on Sos. Since Mg²⁺ is in fast equilibrium with the solvent, it is possible that the initial step in the exchange reaction involves the binding of Sos to Mg²⁺-free, but nucleotide-bound Ras molecules. Based on kinetic analysis of CDC25^Mm-mediated nucleotide dissociation, it has been proposed that the binding interaction of the GEF with Ras consists of two distinct steps: a fast association reaction involving the formation of a ternary complex of a tightly bound nucleotide, Ras, and CDC25^Mm, followed by a conformational transition to a ternary complex in which the nucleotide is loosely bound (19). The formation of a complex between Sos and Mg²⁺-free Ras, if it occurs, might be equivalent to the first step identified in the kinetic studies. Although the structural features of this complex would not permit Mg²⁺ binding, they would not be sufficient to induce a decrease in nucleotide affinity, and a second step involving Sos-mediated conformational changes that destabilize the nucleotide would be required to promote nucleotide dissociation. A detailed kinetic analysis of the Sos exchange mechanism would be required to dissect these reaction steps.

In addition to Ras·Sos, there are two other structures of a GEF bound to a small GTPase that have been described recently: the Sec7 domain bound to nucleotide-free Arf1 and Tiam1 bound to nucleotide-free Rac1 (35, 45). The overall structural features of the Ras·Sos, Arf1·Sec7, and Rac1·Tiam1 complexes share similarities in that, in all complexes, the GEF interacts extensively with switch 2 and induces the displacement of switch 1. However, the postulated exchange mechanisms for the Sec7 domain and Tiam1 appear to be different from that observed for Sos. The conformational change in switch 1 of Arf1 allows the insertion of a glutamate residue (Glu⁹⁷) from the Sec7 domain into the Mg²⁺-binding site. This glutamate residue plays a critical role in the exchange reaction, as indicated by the finding that mutants of the Sec7 domain containing substitutions at Glu⁹⁷ display a >1000-fold reduction in exchange factor activity (46, 47). In contrast, mutation of a glutamate residue of Sos (Glu⁹⁴²) that appears to be structurally equivalent to Glu⁹⁷ of Sec7 had no effect on the ability of Sos to catalyze guanine nucleotide exchange activity. The structural changes in switch 2 of Rac induced by the binding of Tiam1 lead to the repositioning of Ala⁵⁹ toward the Mg²⁺-binding site. This rearrangement is supported by Lys¹¹⁹⁵ of Tiam1, and substitution of this lysine residue with alanine has been shown to impede the guanine nucleotide exchange activity of Tiam1 (35). In the Ras·Sos complex, Ala⁵⁹ of Ras undergoes a nearly identical conformational change that is supported by Thr⁹³⁵, but substitution of either Ala⁵⁹ with Gly or Thr⁹³⁵ with Ala had no effect on nucleotide exchange. Thus, although the structural transitions that accompany the binding of different GEFs to their cognate GTPases might share common features, the molecular details of the reaction mechanisms are likely to be specific for each GEF.

	ACKNOWLEDGEMENTS

We thank Amy B. Hall, Nicholas Nassar, S. Mariana Margarit, Steve Soisson, and Bjorn Schumacher for advice and help.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant CA28146 and by the Carol Baldwin Foundation.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Present address: Biogen, Inc., 14 Cambridge Center, Cambridge, MA 02142.

To whom correspondence should be addressed: Dept. of Molecular Genetics and Microbiology, State University of New York at Stony Brook, Life Science Bldg., Stony Brook, NY 11794-5222. Tel.: 631-632-9737; Fax: 631-632-8891; E-mail: barsagi@pharm.sunysb.edu.

Published, JBC Papers in Press, May 1, 2001, DOI 10.1074/jbc.M101727200

	ABBREVIATIONS

The abbreviations used are: GEFs, guanine nucleotide exchange factors; GST, glutathione S-transferase; GTPS, guanosine 5'-O-(3-thiotriphosphate).

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				N. Tuncbag, G. Kar, O. Keskin, A. Gursoy, and R. Nussinov A survey of available tools and web servers for analysis of protein-protein interactions and interfaces Brief Bioinform, May 1, 2009; 10(3): 217 - 232. [Abstract] [Full Text] [PDF]

				W. N. van Egmond, A. Kortholt, K. Plak, L. Bosgraaf, S. Bosgraaf, I. Keizer-Gunnink, and P. J. M. van Haastert Intramolecular Activation Mechanism of the Dictyostelium LRRK2 Homolog Roco Protein GbpC J. Biol. Chem., October 31, 2008; 283(44): 30412 - 30420. [Abstract] [Full Text] [PDF]

				M. Zenker, D. Horn, D. Wieczorek, J. Allanson, S. Pauli, I. van der Burgt, H.-G. Doerr, H. Gaspar, M. Hofbeck, G. Gillessen-Kaesbach, et al. SOS1 is the second most common Noonan gene but plays no major role in cardio-facio-cutaneous syndrome J. Med. Genet., October 1, 2007; 44(10): 651 - 656. [Abstract] [Full Text] [PDF]

				K. Ura, K. Obama, S. Satoh, Y. Sakai, Y. Nakamura, and Y. Furukawa Enhanced RASGEF1A Expression Is Involved in the Growth and Migration of Intrahepatic Cholangiocarcinoma. Clin. Cancer Res., November 15, 2006; 12(22): 6611 - 6616. [Abstract] [Full Text] [PDF]

				T. S. Freedman, H. Sondermann, G. D. Friedland, T. Kortemme, D. Bar-Sagi, S. Marqusee, and J. Kuriyan A Ras-induced conformational switch in the Ras activator Son of sevenless PNAS, November 7, 2006; 103(45): 16692 - 16697. [Abstract] [Full Text] [PDF]

				B. Ford, K. Skowronek, S. Boykevisch, D. Bar-Sagi, and N. Nassar Structure of the G60A Mutant of Ras: IMPLICATIONS FOR THE DOMINANT NEGATIVE EFFECT J. Biol. Chem., July 8, 2005; 280(27): 25697 - 25705. [Abstract] [Full Text] [PDF]

				A. P. Tabancay Jr., C.-L. Gau, I. M. P. Machado, E. J. Uhlmann, D. H. Gutmann, L. Guo, and F. Tamanoi Identification of Dominant Negative Mutants of Rheb GTPase and Their Use to Implicate the Involvement of Human Rheb in the Activation of p70S6K J. Biol. Chem., October 10, 2003; 278(41): 39921 - 39930. [Abstract] [Full Text] [PDF]

				V. A. Schmidt, L. Scudder, C. E. Devoe, A. Bernards, L. D. Cupit, and W. F. Bahou IQGAP2 functions as a GTP-dependent effector protein in thrombin-induced platelet cytoskeletal reorganization Blood, April 15, 2003; 101(8): 3021 - 3028. [Abstract] [Full Text] [PDF]

				K. L. Rossman, D. K. Worthylake, J. T. Snyder, L. Cheng, I. P. Whitehead, and J. Sondek Functional Analysis of Cdc42 Residues Required for Guanine Nucleotide Exchange J. Biol. Chem., December 20, 2002; 277(52): 50893 - 50898. [Abstract] [Full Text] [PDF]

				R. Jorge, N. Zarich, J. L. Oliva, M. Azanedo, N. Martinez, X. de la Cruz, and J. M. Rojas hSos1 Contains a New Amino-terminal Regulatory Motif with Specific Binding Affinity for Its Pleckstrin Homology Domain J. Biol. Chem., November 8, 2002; 277(46): 44171 - 44179. [Abstract] [Full Text] [PDF]

				B. E. Hall, D. Bar-Sagi, and N. Nassar The structural basis for the transition from Ras-GTP to Ras-GDP PNAS, September 17, 2002; 99(19): 12138 - 12142. [Abstract] [Full Text] [PDF]

				X. Tian and L. A. Feig Basis for Signaling Specificity Difference between Sos and Ras-GRF Guanine Nucleotide Exchange Factors J. Biol. Chem., December 7, 2001; 276(50): 47248 - 47256. [Abstract] [Full Text] [PDF]

				Y. Gao, J. Xing, M. Streuli, T. L. Leto, and Y. Zheng Trp56 of Rac1 Specifies Interaction with a Subset of Guanine Nucleotide Exchange Factors J. Biol. Chem., December 7, 2001; 276(50): 47530 - 47541. [Abstract] [Full Text] [PDF]

				L. Pnueli, T. Gutfinger, D. Hareven, O. Ben-Naim, N. Ron, N. Adir, and E. Lifschitz Tomato SP-Interacting Proteins Define a Conserved Signaling System That Regulates Shoot Architecture and Flowering PLANT CELL, December 1, 2001; 13(12): 2687 - 2702. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 276/29/27629    most recent M101727200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hall, B. E. Articles by Bar-Sagi, D. Search for Related Content PubMed PubMed Citation Articles by Hall, B. E. Articles by Bar-Sagi, D. Social Bookmarking                      What's this?