Nov 25, 2015 - T. Evans, unpublished data and used at a 1:3,000 dilution. cytosolic ..... Dr. Tony Evans (Department of Cell Biology, Genentech Inc., South.
THEJOURNAL (c)
OF
VOl. 267, No. 33, Issue of November 25, pp . 23575-23582.1992 Printed in U.S.A.
BIOLOGICAL CHEMISTRY
1992 by The American Society for Biochemistry and Molecular Biology, Inc.
Purification and Characterization of Rac 2 A CYTOSOLICGTP-BINDINGPROTEINTHATREGULATESHUMANNEUTROPHILNADPH
OXIDASE*
(Received for publication, June 24, 1992)
Ulla G. KnausSS, Paul G. Heyworthy, B. Therese Kinsellall, John T. Curnuttev, and Gary M. BokochS** From the Departmentsof $Immunology and Cell Biology and (Molecular and Experimental Medicine, The Scripps Research Institute, LaJolla, California 92037 and the [[Departmentof Medicine and Experimental Therapeutics, Center for Cardiovascular Research, Uniuersity College Dublin, Mater Hospital, 41 Eccles St., Dublin, Ireland
Human neutroDhils and other phagocytes generate superoxide anion (0;) as a means of destroying ingested microorganisms. 0; is produced byan NADPHconsuming oxidase composed of membrane and cytosolic components. Activation of the NADPH oxidase is absolutely dependentupon GTP, indicating the requirement for a GTP-binding protein in this process. We have utilized a five-step chromatographic procedure to isolate a GTP-binding protein from human neutrophil cytosol which can stimulate NADPH oxidase activity in a cell-free assay. Oxidase enhancing activity wasshown to coisolate with thisGTP-binding component, which was purified to apparent homogeneity. The GTP-binding protein was identified as Rac 2 by immunologicalreactivity and amino acid sequencing. Thus, Rac 2 appears to be a third cytosolic component required for human neutrophil NADPH oxidase activation. Recombinant Rac 2 was shown to bind guanine nucleotides in a Mg2+-dependentfashion. GDP dissociation rates weredetermined and shown to be regulated by the free Mg2+concentration. Rac 2 was found to possess the highest rate of intrinsic GTP hydrolysis of any of the characterized members ofthe Ras superfamily. The biochemical properties of Rac 2 indicate it is likely to be subject toregulatory cofactors in vivo.
obtained using antibodies and peptides, there has been no evidence for the direct activityof these or related low molecular weight GTP-binding proteins (LMWG)’ incell-free systems that arereadily manipulatable at the biochemical level. Human neutrophils, eosinophils,monocytes, and macrophagesmakeupanimportantcomponent of the cellular immune response which protects the body against invading microorganisms. After respondingto chemoattractant stimuli, mobilizing to sites of infection, and engulfing the foreign material, these phagocytes are able to generate large quantities of superoxide anion (0;) and derivedtoxic oxygen metabolites to destroy the ingested bacteria. The formation of 0; occurs viathe actionof an NADPHoxidase in a reaction referred to as the “respiratory burst.” Activated by a stimulus from a dormant state, this multicomponent electron transport system catalyzes the NADPH-dependentreduction of molecular oxygen to form 0; (2, 3). The active,0;-generating NADPH oxidase complex is located at the plasma membrane, where a cytochrome bssx(4, 5 ) serves as the terminal electron donor. The activeoxidase has also been shown to involve two cytosolic protein components, p47-phox and p67-phox, which have been cloned (6, 7). It is possible thatother,asyet unidentified, cytosolic proteins are involved as well (8-10). There isevidence that a GTP-binding proteinregulates the NADPH oxidase (11).In a number of early studies, GTPyS was shown to potentiate by 2-4-fold the rate of 0; formation The Rassuperfamily of GTP-binding proteinsis a group of observed in cell-free oxidase systems (9, 12-14). Peveri et al. 20-28-kDa proteins whose members are thought to play im- (15)and Uhlinger et al. (16)demonstrated with guanine for GTP portant roles in cell regulation (1).Although these proteins nucleotide-depleted cytosol the absolute requirement have been implicated in a wide variety of cellular processes, in oxidase function. We have recently shown (17) that isoit has been difficult to obtain direct evidence forthe regulatory prenoid metabolism is required for activation of the respiraroles of the majority of these proteins in mammalian systems. tory burst in dimethyl sulfoxide-differentiated HL60 cells, Thus, although the Rasproto-oncogenes are clearly involved likely because of the requirement for post-translational procin normal cell growth and differentiation, as determined by essing of an oxidase-associated LMWG.Inthesestudies, mutational analysis andmicroinjection studies, and the Rab/ complementation analysis with a cell-free NADPH oxidase Arf proteins play roles in intracellular vesicular/protein traf- assay localized the modified LMWG to thecytoplasm. ficking, based upon morphological localization as well as data We have reported(18)that a GTP-binding proteinpurified * This work was supported inpart by National Institutes of Health from human neutrophilcytosol and identified as Rac 2 is able Grants HL48008 (to G. M. B.), AI24838 (to J. T. C.), and RR00833 to stimulate NADPH oxidase activity in a cell-free assay. In (to the Scripps Research Institute General Clinical Research Center), the present paper,we document the purificationof Rac 2 and as well as California Tobacco-related Disease ResearchProgram show it to bea regulatory component of the NADPH oxidase. Grant 2RT0079. The costs of publication of this articlewere defrayed T o begin to understand Rac 2 function at the biochemical in part by the payment of page charges. This article must therefore level, we have characterized its guanine nucleotide binding he hereby marked “aduertisement” in accordancewith 18 U.S.C. and hydrolysis properties. Section 1734 solely to indicate this fact. § Supported by apostdoctoral fellowship fromDeutscheForschungsgemeinschaft. ** To whom correspondence should be addressed: IMM-14, The Scripps Research Inst., 10666 N. TorreyPines Rd., La Jolla, CA 92037.
The abbreviations used are: LMWG, low molecular weight GTPbinding proteins of the Ras superfamily; GTPyS, guanosineFi’-O-(3thiotriphosphate); PIPES, 1,4-piperazinediethanesulfonicacid; GDI, GDP dissociation inhibitor; GDS, GDP/GTPdissociation stimulator.
23575
23576
Neutrophil Human EXPERIMENTALPROCEDURES
Biological Materials Purification of human neutrophils and preparationof cytosol and membranes were as described previously (18, 19). Briefly, leukapheresis was used to obtain large numbers of white blood cells from normal donors.Afterremoval of erythrocytes by hypotonic lysis, neutrophils were purified by differential centrifugation through Ficoll-Hypaque. Purified neutrophils were treated with 3 mM diisopropylfluorophosphate for 15 min at 4 "C, washed with phosphatebuffered saline, and subjected to nitrogen cavitation (450 p s i , 20 min, 4 "c)in a buffer consisting of 100 mM KC], 3 mM NaCl, 1 mM ATP, 3.5 mM MgCI,, 10 mM PIPES (pH 7.3), 1 mM phenylmethylsulfonyl fluoride, and 100 kallikrein inhibitory unitsof aprotinin/ml. Cavitated cells were collected into sufficient EGTA to give a final concentration of 1 mM, centrifuged a t low speed (1,000 X g, 10 min) t o remove unbroken cells and nuclei, and then fractionated on discontinuous 15/40/60% sucrose gradients in 25 mM HEPES (pH 8.0), 1 mM EGTA, 1 mM EDTA buffer. The cytosol overlay was collected and immediatelyfrozen a t -70 "C, whereas the 1 5 4 0 % interface containing plasma membranes was washed, repelleted, and stored in aliquots a t -70 "C in 25 mM Hepes (pH 8.0) and 20 mM sucrose. Purification of G,,,/Rac 2 from Human Neutrophils with high NADPH DEAE-Sephacel Chromatography-Cytosols oxidase activity, as assessed in the cell-free system (19), were combined to a volume of200-250 ml (4 X 10" cell equivalents)and concentrated 10-fold by Amicon filtration using a 10,000 molecular weight cutoff filtrationmembrane. The cytosol was thensupplemented to final concentrations of 1 mM 2-mercaptoethanol, 0.1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 1p~ leupeptin, 1 p~ pepstatin, 100 kallikrein inhibitory units of aprotinin/ml, and 100 p~ l-chloro-3-tosylamido-7-amino-2-heptanone. After a subsequent 10-fold dilution in 25 mM Tris-HC1 (pH 7.5),1 mM EDTA, 0.1 mM dithiothreitol, 5 mMMgC12, 0.5 mM phenylmethylsulfonyl fluoride, 1 mM 2-mercaptoethanol, 1.5 mM ATP (TEDMPM buffer/ATP 1.5), the cytosol was applied a t a flow rate of 25 ml/h to a column (2.5 X 25 cm) of DEAE-Sephacel (Pharmacia LKB Biotechnology Inc.) equilibratedwith the samebuffer. The columnwas washed with TEDMPM buffer/ATP 1.5 until the monitored absorbance at 280 n m reached the base line and then eluted with a linear gradient of NaCl (280 ml, 0-180 mM NaC1) in TEDMPM/ATP 0.1. After completion of the gradient, the columnwas eluted further with150 ml of 1 M NaCl in TEDMPM buffer. Fractions of 3.5 ml were collected. Gel Filtration on Sephacryl S-200 HR-The IjEAE-Sephacel fractions containing ["S]GTPrS binding activity and the oxidase enhancing activity (peak 11) were pooled, concentrated to 600 pl, and applied to a Sephacryl S-200 HR column (1.5 X 120 cm, Pharmacia) equilibrated with 100 mM NaCl in TEDMPM buffer/ATP 0.15. The column was eluted a t a flow rate of 13 ml/h, and fractions of 2 ml were collected. The column had also beenpreevaluated withmolecular weight standards under the same elution conditions. Mono Q Ion Exchange Chromatography-The [35S]GTPyS-binding and oxidase enhancing fractions from the Sephacryl S-200 column were pooled, concentrated to 2ml, adjusted to 10 mM NaCI, and injected onto a Mono Q HR 5/5 column connected to a fast protein liquid chromatography system (Pharmacia). The column was then washed with the equilibration buffer (25 mM Tris-HC1 (pH 7.5), 1 mM EDTA, 0.1 mM dithiothreitol, 5 mM MgCl,, 1 mM 2-mercaptoethanol) and eluted with a shallow linear NaCl gradient (25 ml, 0120 mM NaCl), followed by a steeper linear NaCl gradient (10 ml, 120-250 mM NaCI) and a 1M NaCl wash (5 ml). The chromatography was performed at a flow rate of 0.5 ml/min, and the fractionsize was 1 ml. Heptylamine-Sephal ose Chromatography-The peak Mono Q fractions with ["'SIGTPyS-binding and oxidase stimulatory activitywere pooled, adjusted to 0.2% cholate, and injected onto a heptylamineSepharose column (1.5 X 20 cm, fast protein liquid chromatography system), equilibrated with 25 mM Tris-HC1 (pH 7.5), 1 mM EDTA, 0.1 mM dithiothreitol, 5 mM MgCl,, 1 mM 2-mercaptoethanol (TEDMM buffer),100 mM NaC1, and 0.2% cholate. The column was then washed with 50 ml of equilibration buffer a t a flow rate of 0.2 ml/min. The elution was performed withtwo successivelinear cholate gradients in TEDMM buffer (10 ml, 0.2-1.0% cholate; 25 ml, 1.01.6% cholate) followed by two cholate step gradients (15 ml of 1.6% cholate and 30 ml of 2.0% cholate in TEDMM buffer). At the same time a negative NaCl gradient in TEDMM buffer (250-0 mM NaCU
Rac 2 Purification was performed. Fractions of 2 ml were collected and assayed for [3sS] G T P r S binding activity. The active fractions were combined, concentrated, extensively dialyzed in TEDMM buffer, and tested in the cell-free oxidase assay. Phenyl-Superose Chromatography-The dialyzed heptylamineSepharose pool was diluted with an equal amountof TEDMM buffer containing 1.5 M ammonium sulfate and 5% ethylene glycol. After injection onto a phenyl-Superose HR 5/5 fast protein liquid chromatography column (Pharmacia) equilibrated with the same buffer, the column was washed with 8 ml of equilibration buffer until the absorbance base line was obtained. The elution was performed with a simultaneous reverse lineargradient of ammoniumsulfatein TEDMM (12 ml, 1.5 M-0 ammonium sulfate) and a linear ethylene glycol gradient (12 ml, 5-50% ethylene glycol). This was immediately followed by a 3-ml wash with TEDMM buffer/50% ethylene glycol and 5 ml of TEDMM/6O% ethylene glycol. The elution was performed a t a flow rate of 0.1 ml/min, and fractionsof 0.5 ml were collected. The results shown aretypical of more than 10 purifications of GoJ Rac 2 conducted over the course of 2 years. Preparation and Purificationof Recombinant Rac 2 The full-length cDNA coding for Rac 2 was subcloned into the NdeI-BamHI restriction sites on the expression vector pET3a under the control of the isopropyl 1-thio-/3-D-galactopyranoside-inducible gene 10 promoter, as reported for Rac 1 in (20). The recombinant plasmid was named aspWO1. For protein expression, a single recombinant colony was inoculated into LB medium (10 g of NaC1, 10 g of tryptone, 5 g of yeast extract), containing 100 pg/ml ampicillin and 15 pg/ml chloramphenicol and grown overnight a t 37 "C. The overnight culture was diluted 50-fold with the same medium, incubated for 2.5 h a t 37 "C, then isopropyl 1-thio-P-D-galactopyranosidewas added to a final concentration of 1 mM to induce protein expression. The cells were grown for an additional 2 h at 37 "C, then harvested by centrifugation at 5,000 X g, 15 min, 4 "C and suspended in25 mM Tris-HC1 (pH 8.0), 1 mM EDTA, 5 mM dithiothreitol, 5 mM MgCl,, 25 mM NaC1, 1 mM phenylmethylsulfonyl fluoride. After two freeze/ thaw cycles the suspension was subjected to nitrogen cavitation (450 p s i . , 20 min, 4 "C) and centrifuged a t 15,000 X g, 20 min, 4 "C. The clear supernatant obtained (60 ml)was applied to a DEAE-Sephacel column (2.5 X 20) equilibrated with 25 mM Tris-HC1 (pH 8.0), 1 mM EDTA, 1 mM dithiothreitol, 5 mM MgCl,, 0.5 mM phenylmethylsulfonyl fluoride. After a 150-ml wash with this buffer, a linear NaCl gradient (400 ml, 0-500 mM NaCl) was applied, followed by a 150-ml 1 M NaCl wash, both in equilibrationbuffer. The Rac 2 protein eluted in the flow-through and the wash of the column, as determined by ["S]GTPrS bindingactivity. SDS-polyacrylamidegel electrophoresis revealeda proteinwith amolecular mass of 22,000 Daandan estimated purity of 8 5 9 0 % by silver staining. Further purification was obtained by phenyl-Superose fast proteinliquid chromatography (see above), asnecessary. Cell-free NADPH Oxidase Activation Assay Chromatography fractionswere assayed for their ability to enhance NADPH oxidase activity ina fully soluble cell-freesystem containing a threshold concentration of neutrophil cytosol that provided a small amount of each of the cytosolic oxidase components (8). Reaction mixtures contained 0.1 mM cytochrome c, 6.5 mM MgCl,, 87 mM KC], 2.6 mM NaC1, 8.7 mM PIPES (pH 7.3), 10 p M GTPyS, 0.16 mM NADPH, 40 p~ SDS, 1.25 X lo6 cell equivalents of deoxycholatesolubilized membranes, and 6.3 X lo5 cell equivalents of cytosol. TO account for 0;-independent cytochrome c reduction, parallel reaction mixtures also contained 9 pg of superoxide dismutase. Cytosol and membranes were prepared as described (19,211. Reactions were done in 96-well microplates in a total volume of 150 pl and initiatedby the addition of 40 p~ SDS. Maximal rates of 0; generation were measured a t 25 "C by following thechangeinabsorbance caused by reduction of cytochrome c a t 550 nm using akinetic microplate reader (Molecular Devices Corp., Menlo Park, CA) and Softmax software (Molecular Devices, release 2.01) (8). Analysis of Guanine Nucleotide Binding Binding of [%]GTPrS to GTP-binding proteins during all chromatography steps was determined with the rapid filtration technique as described (22). For thestandardassay, 10 p1 ofsamplewas incubated for 5 min a t 30 "C in 90p1 of reaction mixture containing 50 mM HEPES (pH 8.0), 1 mM dithiothreitol, 2 mM EDTA, 0.1% Lubrol Px, and1p~ ["S]GTPyS (1-2 X lo4cpm/pmol). The reaction
Neutrophil Human was terminated by the addition of 2 ml of ice-cold stop mixture (25 mM Tris-HC1 (pH 8.0), 100 mM NaC1, 30 mMMgC12, 2 mM dithiowas quantitated threitol, 1mg/ml bovine serum albumin), and binding by vacuum filtration on BA-85 nitrocellulose filters and liquid scintillation counting. Time courses of GTPyS binding at different magnesium concentrations were performed with 40 pmol of Rac 2 protein (in 25 mM Tris-HC1 (pH7.5), 1 mM EDTA, 0.1 mM dithiothreitol, 5 mM MgC12) in 500 pl of HEPES (pH 8.0), 2 mM EDTA, 1 mM dithiothreitol, 100 pg/ml bovine serum albumin, 1p M [35S]GTPrS(10,000-15,000 cpm/ pmol), and varying amountsof MgC1, (0, 0.8,1.8, and 2 mM) up until 30 min a t room temperature. The free magnesium ion concentrations under the conditions described were calculated according to Higashijima et al. (23). The ability of Rac 2 to bind GTP+ at varyingMg2' concentrations was subsequently determined as follows. A basic buffer of 50 mM HEPES (pH 7.5), 1 mM dithiothreitol, 100 pg/ml bovine serum albuminwas supplemented with2 or 10 mM EDTA andvarious amounts of MgC1, (100 pM to 5 mM). This resulted in mixtures with 11 different free Mg2' concentrations, as indicated in the legend to Fig. 8. 1 p~ ["S]GTPyS (2 X lo4 cpm/pmol) and 10 pmol of Rac 2 protein in 25 mM Tris-HCI (pH 7.5), 1 mM EDTA, 0.1 mM dithiothreitol,and 5 mMMgC1, were added to give a finalincubation volume of 100 pl. The sample was incubated for 5min a t room temperature, then aliquots of 25 p1 were removed for binding analysis. at a final Dissociation Rate of GDP from Rac 2 Protein-Rac 2 concentration of 100 nM was incubated with 25 mM Tris-HC1 (pH 7.5), 4.735 mM EDTA, 1 mM dithiothreitol, 4.687 mMMgC1, (free Mg'f = 36 nM), 100 pg/ml bovine serum albumin, and 10 p M [3H] GDP (3,000-4,000 cpm/pmol) a t 30 "C for 4 min. The reaction was stopped by the additionof sufficient MgCl, to raise the concentration of free M$+ to 50 p~ and the incubation placed on ice until several minutespriorto use. Therate of dissociation of [3H]GDP was assessed by adding 700 pl o f the prelabeledRac 2 to a test tube of sufficient EDTA containing 200 pM unlabeled GTP in the presence and MgCl, toadjustthe free Mg2' concentrationstothe values indicated in the text. The incubation was carried out a t room temperature, and aliquots of 50 pl were taken at the indicated times for filtration analysis of the amount of ['HIGDP bound to protein. The results shown represent one of two separate but similar experiments. Analogous experiments were performed using [w3'P]GTP to assess the dissociation rate for GTP. Assay of GTP Hydrolysis by Rac 2-The rate of intrinsic GTP hydrolysis by Rac 2 was determined by first preparing the [y-"PI GTP-bound protein atlow Mg2' and then initiating G T P hydrolysis by the addition of high M$+. The [y-"PIGTP-Rac 2 complex was formed by incubating the protein(250 nM) with 25 mM Tris-HC1 (pH 7.5), 4.735 mM EDTA, 1 mM dithiothreitol, 4.687 mM MgCI, (free M$+ = 36 nM), 5 mM dimyristoylphosphatidylcholine,and 20 p~ [y"'PIGTP (10,000-15,000 cpm/pmol) for 4 min a t 30 "C. Hydrolysis was initiated at 30 "C or after a brief equilibration to room temperature by the addition of 0.05 volume of a mixture containing 400 mM MgCl, and 4 mM GTP, and aliquotswere taken to assess[y-"'PIGTP remaining protein-bound at the indicated times. Under these conditions the dissociation of the [y-:j2P]GTP was negligible, allowing us t o assess the rate of GTP hydrolysis directly. The results shown are from a single experiment representative of three similar experiments.
Rac 2 Purification
23577
cytosolic fraction of phagocytic cells (12, 13, 27). Our laboratoryhad localized a compactin-sensitivecomponent absolutely required for the respiratory burstresponse in dimethyl sulfoxide-differentiated HL60 cells to the cytosolic fraction (17). To identify this putative GTP-binding protein,we fractionated human neutrophil cytosol in a five-step chromatographic procedure, assessing both the binding of GTPyS and thestimulation of NADPH oxidase activityin acell-free assay containing only a basal level of cytosol, as described under "Experimental Procedures." DEAE-Sephacel chromatography of combined cytosols resolved the NADPH oxidase enhancing activity into threemajor peaks: one located in the flow-through of the column and two others in the salt gradient with peptide (Fig. 1). Immunoblotting of the column fractions antibodies directed against p47-phox and p67-phox revealed the p47-phox component in the unbound fractions (peak I), as well as at low levels in the fractions 75-80. Peak 111 was found to contain high levels of p67-phox (18). These previously identified oxidase components were not detected in the second peak of NADPH oxidase enhancing activity. Rapid filtration assays of GTPyS binding at low magnesium concentrations revealed the presence of GTPyS binding activity in the DEAE column flow-through and in two overlapping peaks in the salt gradient. The firstof these peaks, eluted a t 110 mM NaC1, correlated with the oxidase enhancing peak 11. ADP-ribosylation experiments with the botulinum toxin C3 ADP-ribosyltransferase showed that the substrate for this toxin, presumably Rho (28), was clearly resolved from the oxidase enhancing activity in peak 11. Protein immunoblots a common consensus region of G T P with an antibody against binding in Ras-related proteins (142-24305), as well as specific RaplA and CDC42Hs antibodies, did not recognize the GTP-binding protein in the peak I1 fractions (18). Furthermore, the GTP-binding protein in this peak was unable to bind [ L ~ - ~ * P ] Gspecifically TP after SDS-polyacrylamide gel electrophoresis/transfer to nitrocellulose, a characteristic of some, but notall, low molecular weightGTP-binding proteins (29). DEAEpeak I1 was then subjected to gel filtrationon Sephacel S-200(Fig.2). Theoxidase-stimulatoryactivity coeluted with GTPyS binding activityin a single peak with a relative molecular weight of 160,000. This suggests that Rac 2 may exist as a complex with other cytosolic protein(s) a t this stage. Further purification of this activity was achieved by ion exchange chromatography on Mono Q (Fig. 3). Again a single peak of oxidase enhancing activity was obtained in 100 mM NaCl which correlated with thesaltgradientat 40
I
Miscellaneous
7 3.0
[:"SS]GTPyS (1,200 Ci/mmol), ['HIGDP (36.3 Ci/mmol), [,-"'PI G T P (3,000 Ci/mmol), and [y3'P]GTP (6,000 Ci/mmol) were from Du Pont-New England Nuclear Research Products. SDS-polyacrylamide gel electrophoresis was performed according to Laemmli (24) using 13%polyacrylamide gels and molecular weight markers by BioRad. Silver staining (25) and immunoblotting (26) were as described. Peptide antibodies to Rac 2 and Rac 1 were used as previously (18). Rho GDPdissociation inhibitor (GDI) was detected with an affinitypurified peptide antibody generated to amino acids 17-28 of Rho GDI.' RESULTS
Purification of an NADPHOxidase-stimulatory GTP-binding Protein-A number of earlier studies had suggested that a putativeGTP-bindingcomponent which could enhance NADPH oxidase activity in cell-free systems resided in the
' T. Evans,
unpublished data andused a t a 1:3,000 dilution.
0
20
40
60
140 80 120 100
160
""
Fraction Number
FIG. 1. DEAE-Sephacel chromatography of human neutrophil cytosol. 0; generation (U)and GTPyS binding (A)activities were assayed as described under"Experimental Procedures." The absorbance of the eluted protein a t 280 nm was monitored (- - -). Peaks I, 11, and 111 represent collected pools with oxidase enhancing activity.
23578
Neutrophil
Rac 2 Purification
Human
acrylamide gels revealed a 22-kDa protein, purified to near homogeneity (Fig. 5) after this chromatographic step. This GTP-binding protein was able to augment the basal rate of NADPH oxidase activity 2.5-fold a t a final concentration of 73 150 nM (18). In the particular preparationdepicted in Fig. 5, 400 lane 5, twobands were visible a t approximately 22 kDa, m whereas ina number of other preparations a single band was 300 visible. Both bands of the purified preparation shown were Ic7 sequenced after trypsin cleavage, and both were identified as 200 In the low molecularweight GTP-bindingprotein Rac 2, as 0, 100 indicated by amino acid identity intwo peptides derived from regions (amino acids 18-49 and 175-183) where Rac 1 and n 0 10 20 30 40 50 60 Rac 2 differ and in two peptides identical in both proteins Fraction Number (18, 30). It is possible that the lower bandrepresents a FIG.2. Sephacryl S-200 H R chromatography of the concen- proteolytic breakdown product of the upperRac 2band, since trated peak I1 (fractions 106-114 of DEAE-Sephacel). 0; the abundance of the lower band was perceived to increase generation (M) and GTPyS binding (A)activities were assayed as during the purification process. Alternatively, the lower band described under “Experimental Procedures.” The absorbance of the could be a differentially processed form of the protein. eluted protein a t 280 nm was monitored (- - -). Fractions 33-39 were Immunoblotswith affinity-purified antibodies directed pooled and concentrated (Go%). againstcarboxyl-terminalpeptides specific forRac2 and Rac 1 (18) were performed with neutrophil membranes, cy40 1000 I I I I 110 tosol, and the purified Go, preparation. Only the Rac 2 antibody recognized Go,, as shown in Fig. 6 (data of Rac 1 immunoblot notshown). It is difficult to calculate the actual yield of Rac 2 from the starting material because of the presence of multiple GTPbinding proteins in cytosol and an increase in the apparent recovery of Rac 2, reflected in anincrease in GTPySbinding, observed during later chromatography steps. Wereasoned that the lattercould be explained by the presence of a regulatory protein such aasGDI in the Rac 2-containing fractions. We therefore probed the Rac 2 pools from each column with a n affinity-purifiedpeptideantibody directed againstRho 0 10 20 30 40 GDI. This antibodyrecognized a single protein with the same Fraction Number FIG.3. Mono Q H R 5/5 chromatography of Gox.The Sepha- molecular mass as Rho GDI (26 kDa, Fig. 7) in neutrophil cry1 S-200 pool was adjusted to 10 m M NaCl, injected onto a Mono Q cytosol and in theRac 2 peaks from the DEAE-Sephacel and column, and elutedwith two successive salt gradients. 0;generation subsequent gel filtration chromatographies. The Mono Q peak (M), GTPyS binding (A),and absorbance (- - -) a t 280 nm were fractions showed only a faintly cross-reactive band, which monitored. See “Experimental Procedures”for further explanation. was resolved from Rac 2 by the heptylamine-Sepharose chromatography. The purified GTP-binding proteinGo, (or Rac 2) used in the cell-free assays to determine oxidase enhancing activity contained no traceof protein reacting with Rho GDI antibody. We routinely obtain about3-5 pg of essentially pure Rac 2 at theheptylamine-Sepharose stagefrom -700 mg of starting cytosolic protein. This purified Rac 2 will bind [”S]GTP.yS 12.0
700 L
Y
kDa 9766
0
10
20
30
40
50
-
-
60
Fraction Number FIG.4. Heptylamine-Sepharosechromatographyof Gox.The pooled and concentrated fractions 20-24 obtained from the Mono Q column were applied to a heptylamine-Sepharose column and eluted as described under “Experimental Procedures.” GTPyS binding (A) was analyzed, andabsorbance (- - -) was monitored a t 280 nm. Fractions 46-48 were pooled (see bar above peak),concentrated, dialyzed in TEDMM buffer (pH 7.5), and assayed for 0;generation. See “ExperimentalProcedures” for further details.
GTPyS binding. After hydrophobic interaction chromatography on heptylamine-Sepharose, the GTPyS binding activity was elutedat 1.2%cholate (Fig. 4). The pooled, concentrated, and dialyzed fractions of this peak exhibited NADPH oxidase enhancing activity in the cell-free assay. Silver-stained poly-
P I
14-
m1
2
0
3
4
5
6
7
FIG.5. SDS-polyacrylamide gel electrophoresis of the peak fractions for 0;generation and GTPyS binding from a complete purification sequence. Proteins were visualized by silver staining.Thesamples are: lane 1 , molecular massmarkers (low molecular mass, Bio-Rad); lane 2, peak 11, DEAE-Sephacel; lane 3, Sephacryl S-200 H R lane 4, Mono Q HR 5/5; lane 5, heptylamineSepharose; lane 6,protein complex of MRP-8 and MRP-14(4.8 pg of protein); lane 7, phenyl-Superose HR 5/5.
H u m a n Neutrophil Rac 2 Purification
Rac2-
1
2
3
4
FIG. 6. Immunoblot of purified Go, with anti-Rac 2 antibody. The samples are: lane I, Rac 2-expressing 293 cells to serve as a positive control; lane 2, human neutrophil cytosol (95 pg of protein); lane 3, human neutrophil membranes (120 pg of protein); lane 4, purified Gox,represented by the peak fraction of the heptylamineSepharose chromatography step (0.9 pg ofGTP-bindingprotein). Immunoblots were performedas described under “Experimental Procedures.”
Rho GDI-
1 2 3 4 5 6 7 FIG.7. Immunoblot of the peak fractions for oxidase enhancing activity).( and GTP-yS binding (A)from a complete purification sequence with anti-Rho GDI antibody. The samples are: lane 1, Rho GDI-expressing h’. coli cells to serve as apositive control; lane 2, human neutrophil membranes (105 pg of protein); lane 3, total human neutrophil cytosol (75 pg of protein); lane 4 , peak 11, DEAE-Sephacel(2.5 p g of GTP-binding protein); lane5, Sephacryl S-200 HR peak (1.6 pg of GTP-binding protein); lane 6, Mono Q HR 5/5 peak (1.4 pg of GTP-binding protein); lane 7, heptylamineSepharose peak (1.5 pg of GTP-binding protein). Immunoblotting was performed as described under “Experimental Procedures.”
with a stoichiometry of 0.7-0.9 mol/mol protein, indicating a highlyactive preparation which contains a singleGTP-yS binding site. After we had carried out a number of G,./Rac 2 purifications, we observed the presence of two bands that seemed to coelute with the GTP-binding protein during several purification steps. They migrated in 13% polyacrylamide gel electrophoresis with a relative molecular weight of 14,000 and 8,000 (Fig. 5, lane 6). As shown in Fig. 5, both bands are still associated with the GTP-binding protein after the second ion exchange step but areonly faintly visible after the heptylamine chromatography. Silver-stained gels of every fraction of the Mono Q column revealed that the majorityof these two bands eluted in fractions17-21, directly in frontof the GTP binding peak (fractions 20-24, data not shown). I t was clear i n every purification that theoxidase enhancing activitycorrelated exactly with the GTP-binding protein and not with the 8- and 14-kDa proteins, nor was the activity associated with Rac 2 dependent upon the presence of either of these two proteins in the preparation. Although we were able to resolve these two contaminants from Rac 2 completely by a
23579
second hydrophobic interaction chromatography on phenylSuperose (see Fig. 5, lane 7), this purification step reduced the proteinyield significantly and was therefore notroutinely performed. Sequence analysis of the trypsin-digested 14-kDa protein and immunoblots with a polyclonal antibody that detected both proteins enabled us identify to the 14-kDa protein as the calcium-binding protein MRP-14 and the 8-kDa protein as MRP-8(datanotshown).These two proteins have been previously purified and cloned (31-33) and their expression shown to be associated with myeloid cell differentiation (32, 34) and inflammation (31). Biochemical Characterization of Rac 2-The limited quantities of Rac 2 available by purification from human neutrophil cytosol prompted us touse a bacterial expression system to obtain larger quantities of this protein for the purpose of characterizing its biochemical activity and properties. The bulk of the expressed Rac 2 protein in Escherichia coli was localized in inclusion bodies, which were found, after solubilization, subsequent purification and renaturation, to exhibit little biochemical activity (re: GTPyS binding). The soluble Rac 2 protein, isolated by a two-step purification procedure, was found to be active and at least 90% pure as judged by silver staining of SDS-polyacrylamide gels. Examination of the GTP-yS binding properties of the recombinant protein gave results comparable to those obtained with the native protein. We therefore used the bacterially expressed protein throughout thefollowing biochemical analysis. Kinetics of GTP-yS Binding-The kinetics of GTP-yS binding to Rac 2 a t various magnesium concentrations were examined (Fig. 8 A ) . In the presence oflow concentrations of free M e (26 and 730 nM), the binding to Rac 2 was very rapid, reaching the maximal value between 4 and 5 min of incubation at room temperature. Higher levels of free M e ) in a slower rate of binding, but a t this (10 p ~ resulted concentration maximal GTP-yS bindingcould still be reached within 30 .min.Binding of GTP-yS a t free Mg2‘ values above this concentration, as shown for 69 p~ in Fig. 8A, occurred so slowly that the saturation pointwas not reached after 60 min. We evaluated in more detail the dependence on M e for optimal binding at short times.An optimal range from 0.1-1 p~ free M e was observed for binding (Fig. 8B). Decreasing as well as increasing amounts of free M e resulted in a reduced GTP-yS binding ability of Rac 2 under the experimental conditions (5-min incubation at room temperature). Studies with other low molecular weight GTP-binding proteins have demonstrated that the binding kinetics of the isolated proteins are determinedby the presence of preassociated GDP (35-37). The dissociation rate of bound GDP is very dependent upon the concentrationof free Mg2‘ (35-37). This was confirmed with Rac 2 in Fig. 9, where the presence of Mg” modulated the release of GDP from the protein. At concentrations of 10 pM (not shown) to2 mM free M e , 30% or less of prebound GDP was released after 15 min, whereas the dissociation rate was very rapid a t concentrations below 1p ~with , essentially all of the prebound[’HIGDP dissociated within 5 min. The rateof GDP dissociation was estimated to be 0.95-1.5/min at 400 nMMg2‘ and 0.024-0.032/min at 2 mM M e . Dissociation rates of [a-”P]GTP were considerably slower than for [‘HIGDP under the same conditions (not shown), likely reflecting a higher affinity of Rac 2 for GTP than GDP. GTP Hydrolysis by Rac 2-Fig. 10 showsthe timecourse of the intrinsic GTPhydrolytic activity of the recombinantRac 2 protein a t two different temperatures. The ratefor [-y-”’PI
Human Neutrophil Rac 2 Purification
23580
120 110 100
90
80 70 60 50 40
30 20 10
0 0
5
10
15
20
25
30
35
Time (Min)
16
lo’
15’
Time (min)
FIG. 10. GTPase activity of the Rac2 protein. Intrinsic GTP hydrolysis was determined by evaluating the amount of residual [y32P]GTP remaining bound to the Rac 2 protein at different time points and temperatures, as described under “Experimental Procedures.” The experiment shown is representative of three similar experiments. RT, room temperature.
Free Mg2* (M)
FIG. 8. Panel A , time course of GTPyS binding at varying MgZ+ concentrations. Rac 2 protein was incubated at room temperature until 30 min in buffers with varying amounts of Mg2+ as described under “Experimental Procedures.” Free magnesium concentrations (in PM) were calculated as: A, 0.026; m, 0.73; A, 10.8; and 0,68.6. The data shown are of a single representative experiment out of three similar experiments. Panel B, effect of M e on the rate of GTPyS binding to Rac 2. Rac 2 protein was incubated a t room temperature for 5 min in buffers with varying Mg2’ concentrations. Free Mg2+ values were calculated as described under “ExperimentalProcedures.” The GTPySbinding values represent duplicate determinations from a single experiment, which is representative of three such experiments.
GTP hydrolysis was calculated from the initial portionof the time course to be 0.095 mol/min/mol of GTP bound at room temperature. This rateincreased to 0.198 mol/min/mol when the reaction was carried out at 30 “C. The intrinsic rate of GTP hydrolysis by the Rac 2 protein is approximately 5-10fold higher than the rate determined for the GTP-binding protein CDC42Hs (0.01-0.015 mol/min/mol) (38) and much higher when compared with the hydrolysis rates of other low molecular weight GTP-binding proteins such as Ras (0.005 mol/min/mol) (39), Rho (0.005 mol/min/mol) (28), Ral(O.007 mol/min/mol) (40), or RaplA (0.01 mol/min/mol) (36), even when measured at higher temperatures. DISCUSSION
In this paper we report thepurification to nearhomogeneity of a cytosolic GTP-binding protein that has stimulatoryactivityfor the human neutrophil NADPH oxidase. We observed the copurification of GTPyS bindingactivity and oxidase enhancing activity over DEAE-Sephacel, Sephacryl S-200 HR, andMono Q chromatographies. Oxidase enhancing activity resulting from the p47-phox and p67-phox components were completely resolved at the initialDEAE chromatography step. The purified GTP-binding protein obtained from the heptylamine-or phenyl-Sepharose chromatographies exhibited the ability to stimulate oxidase activity by 2.5-fold a t a final concentration of 150 nM in the cell-free assay. The identity ofGo, as Rac 2 was established by amino acid sequence analysis of proteolytic fragments obtained from the purified protein (18), as well as by cross-reactivity with Racspecific antiserum (Fig. 6). We have confirmed the requirement for Rac 2 innormalNADPH oxidase function by demonstrating oxidase inhibition by a Rac2-specific antibody (18). It has been reported recently byAb0 et al. (41) that a component of the 01 factor isolated from guinea pig macrophages was Rac 1and thatRac 1 exhibited oxidase enhancing activity (%fold increase) in a similar cell-free assay. Rac 1 Time (min) was found to copurify with the GDI for Rho (42) in their FIG. 9. Kinetics of the GDP dissociation from Rac 2. The purification protocol. We were able to isolate Rac 2 which release of bound [“HI GDP from Rac 2 was calculated by measuring was not complexed with Rho GDI or other molecules. Accordthe amounts of Rac 2-bound [3H]GDP after incubation at:A, 2 mM; 0 , l p M ; and 0,0.4 ,UM as described,under “Experimental Procedures.” ing to theimmunoblot in Fig. 7, Rac 2appears tobe associated The result shown is from a single experiment representative of two with Rho GDI or asimilar molecule in human neutrophil cytosol, but this complex can be dissociated in the later steps total experiments.
Human Neutrophil Rac 2 Purification of our isolation procedure. The absence of detectable Rac 2 in neutrophil membrane (Fig. 6, lane 3 ) suggests that neutrophil Rac 2, which is likely to be isoprenylated (171, is maintained as a cytoplasmic form by complexation with the GDI. Our results demonstrate that Rac 2 possesses oxidase stimulatory activity when isolated free of any associated GDI. It should be noted, however, that GDI is present in the cytosol added to thecell-free assay as determinedby immunoblotting and GDI activity assay.3 One would, simplistically, expect the presence of Rho GDI with Rac 1 in the a1 complex to inhibit the activity of Rac 1by preventing exchange of GTP for GDP. Such an inhibitory effect of Rho GDI upon NADPH oxidase activity has been recently reported (43). Further analysis of the function(s) of Rac 1, Rac 2, and Rho/Rac GDI in thecellfree NADPH oxidase system will benecessary to address such questions. Interestingly, two proteins which we identified by partial amino acid sequence analysis and immunoblotting as MRP-8 and MRP-14 (31-33) copurified with Rac 2 through three chromatographic steps. These two proteins have been cloned (31, 32) and shown to be members of the S-100 family of calcium-binding proteins. Both proteins have been isolated from the cytosol of human monocytes and granulocytes (33), and the levels of MRP-8 andMRP-14 expression in myeloid/ monocytic cell precursors increase upon differentiation (31, 33). We were able to resolve the bulk of MRP-8 and-14 from Rac 2 after the second ion exchange chromatography and detected no oxidase enhancing activity associated with the fractions containing these two proteins. Rac 2 itself was active when resolved from these two proteins. Clark et al. (44) recently reported that MRP-14 is able to enhance p47-phoxand p67-phox-dependent oxidase activity in a cell-free system. Although there is no requirement for calcium in the cell-free NADPH oxidase assay, it is possible that the association of MRP-8 and MRP-14 with Rac 2 is not fortuitous. Guanine Nucleotide Binding Properties of Rac 2-We have initiated a biochemical analysis of the guanine nucleotide to binding properties of Rac 2 to relatetheseproperties regulation of the NADPH oxidase by Rac 2. It is apparent from the data of Fig. 8 that the binding of GTPyS to Rac 2 is regulated by the free Mg2+ concentration in the reaction. This situation is similar to thatwhich we previously observed with purified human neutrophil RaplA (36) and which has also been reported for purified Ras (35) and related proteins (40). The slow binding kinetics observed at high Mg2+ concentrations reflect the slow release of GDP which is bound to the purified LMWG. Hall and Self (35) showed that low Mg2+ increased the exchange rate of bound guanine nucleotide on the H-Ras protein. We observe a similar situation with Rac 2: there is a marked enhancement of the rate of GDP dissociation at low Mg2+concentrations (Fig. 9). These data suggest that theexchange of GTP for GDP on Rac 2 must be catalyzed by a GDP/GTP dissociation stimulator (GDS) in the intact cell. The results shown in Fig. 10 demonstrate that Rac 2 has a very rapid intrinsic rate of GTP hydrolysis. At room temperature, Rac 2 hydrolyzes GTP at a rate of -0.095 mol/min/ mol. This is 40-50-fold higher than H-Ras (39) and 6-10-fold greater than the closely related CDC42Hs protein (38). This rapid rate of GTP hydrolysis begins to approach the level of activity seen with GTPase-activating protein-stimulated HRas (45). The consensus amino acids for GTP binding found in H-Ras and other LMWG are well conserved overall in Rac 2 (30). An exception is a threonine for asparagine substitution at position 116 (the NKXD consensus sequence). :I
T.-H. Chuang and G. M. Bokoch, unpublished observations.
23581
CDC42Hs, which also has this Thr + Asn substitution, has a GTP hydrolysis rate nearly 10-fold greater than H-Ras(38). This region is not thoughtto contribute significantly to GTP hydrolysis activity in Ras, however. Conformational changes in Ras after GTP binding involving loop 4 (amino acids 5777) and loop 2 (amino acids 21-48) regions are the probable rate-limiting steps in the GTP hydrolysis reaction (46, 47). The conformation of Rac 2 in these regions may besuch that GTP hydrolysis can occur at an inherently greater rate. Mutational analysis willbe required to identify the region responsible for the intrinsically higher GTPase activity of Rac 2. A high rate of GTP hydrolysis by Rac 2 may confer unique regulatory properties to theRac 2-regulated NADPH oxidase system. If such a high rate of hydrolysis is manifest in uiuo, conversion of GTP to GDPmight not be a rate-limiting step in the activation cycle of Rac 2. GTPase activity for Rac 1 has been shown to reside in the breakpoint cluster region protein associated with chronic myelogenous leukemia, as well as in the brain protein N-chimaerin (48). It is not known whether these Rac 1-active GTPase-activating proteins are also able to catalyze GTP hydrolysis by Rac 2. Under cellfree oxidase assay conditions (i.e. in the presence of cytosol +- arachidonic acid or SDS) we observed no stimulation of Rac 2 GTP hydrolysis. We have, however, detected a Rac 2 GTPase enhancing activity in the plasma membrane of human neutrophils, which was unaffected by the presence of arachidonic acid or SDS. The role of this protein in determining the activity of Rac 2 in the NADPH oxidase system remains to be evaluated. Rac 2 Is a Regulatory Componentof the Phagocyte NADPH Oxidase-We have demonstrated that Rac 2 copurifies with an NADPH oxidase enhancing activity present in human neutrophil cytosol. We have also determined that recombi.~ lines of evidence nant Rac 2 exhibits similar a ~ t i v i t ySeveral indicate that Rac 2 is necessary for complete NADPH oxidase function. We have shown (18)that a specific Rac 2 antibody can inhibit by up to 70% the generation of 0;in a fully active cell-free system. Drugs able to inhibit the isoprenylation of low molecular weight GTP-binding proteins can completely prevent the respiratory burstresponse of differentiated HL60 cells to either chemoattractantsor phorbol esters (17). Preliminary data indicate that post-translationally processed Rac 2 can reconstitute the oxidative response of such inhibited cells.5 It thus appears that Rac 2 can be considered a third required cytosolic cofactor of the NADPH oxidase system of human neutrophils. Rac 1 has been found to enhance activity of the guinea pig macrophage NADPH oxidase (41). Studies are under way to compare the relative activities of Rac 1 and Rac 2 in the human neutrophil system. Because the cell-free assay we have utilized in the current study contains abasal level of neutrophil cytosol, it is possible that proteins present in the added cytosol contribute to the oxidase enhancing activity of Rac 2. Additionally, Rac 2 regulatory proteins, such asGTPase-activating proteins, GDIs, and GDSs, wouldbe expected to exert significant control over the activity of the system. It will be important to direct future studies to identifying and understanding the interaction of these other components with Rac 2 and the NADPH oxidase. Acknowledgments-We acknowledge Benjamin P. Bohl for technical assistance, Dr. Xue Min Xu for assistance in Rac 2 expression,
P. G. Heyworth, U. G. Knaus, X. M. Xu, G. M. Bokoch, and J. T. Curnutte, unpublished observations. U. G. Knaus and G. M. Bokoch, unpublished observations.
’
Neutrophil Human
23582
and Elizabeth Cromer for excellent secretarial assistance. We thank Dr. Tony Evans (Department of Cell Biology, Genentech Inc., South San Francisco, CA) for kindly providing Rac 2, Rac 1, and Rho GDI antibodies. We thank Dr. Eliezer Huberman (Argonne National Laboratory, Argonne, IL) for kindly providing protein standards and antibody against MRP-8 and MRP-14. REFERENCES 1. Hall, A. (1990) Science 249,635-640 2. Clark, R. A. (1990) J. Infect. Dis. 161,1140-1147 3. Babior. B. M. (1978) N. End. J. Med. 298. . ~ . 659-668 4. Senal. A. W. (i987) Nature"326.88-92 5. PGkos, C. A , Allen, R. A., Cochrane, C. G., and Jesaitis, A. J. (1987) J. Clin. Invest. 80,732-742 6. Volpp, B. D., Nauseef, W. M., Donelson, J. E., Moser, D. R., and Clark, R. A. (1989) Proc. Natl. Aead. Sei. U.S.A. 86.7195-7199 7. Leto,'T. L., Lomax, K. J., Volpp, B. D.,~Nunoi,H.; Sechler, J. M. G., Nauseef, W. M., Clark, R. A., Gallin, J. I., and Malech, H. I.,, (1990) Science 248, 727-730 8. C u ~ u t t eJ. , T., Scott, P. J., and Mayo, L. A. (1989) Proc. h'atl. Acad. Sci, U. A. 86.825-829 9. Bolscher, B. 6. J. M., Denis, S. W., Verhoeven, A. J., and Roos, D. (1990) J. Biol. Chem. 2 6 5 , 15782-15787 10. Nunoi, H., Rotrosen, D., Gallin, J. I., and Malech, H.L. (1988) Science 242.1298-1301 11. Quilfiam, L.A.. and Bokoch. G. M. f19921 Si~naET r u ~ d ~ t i o innInflammatory C e l k I (Gimbrone; M. A,, 'and Cocgrane, C. G., eds) Voi. 3, pp, 25-56, Academic Press, San Diego, CA 12. Gabig, T. G., English, D., Akard, L. P., and Schell, M. J. (1987) J. Biol. Chem. 262. 1685-1690 13. Ligeti, E., Doussiere, J., and Vignais, P. V. (1988) Biochem. J. 27,193-200 14. Seifert, R., Schultz, G., and Rosenthat, W. (1986) FEBS Lett. 205, 161~
~~~~
- I
s.
1fi5
15. Pei&i, P., Heyworth, P. G., and Curnutte, J. T. (1992) Proc. Nutl. Acad. Sci. U. S . A. 89,2494-2498 16. Uhlineer, D. J., Burnham. D. N.. and Lambeth. J. D. 11991) J. Biol. Chem. 266,20990-20997 17. Bokoch, G. M., and Prossnitz, V. (1992) J. Clin. Inuesf. 89, 402-408 18. Knaus, U. G., Heyworth, P. G., Evans, T., Curnutte, d . T., and Bokoch, G. M. (1991) Science 254. 1512-1515 19. Curnutte, J. T., Kuver, R., and Scott, P. J. (1987) J. Biol. Chem. 262, 5563-5569 20. Kinseila, B. T.,and Maltese, W. A. (1992) J. Bioi. Chem. 267,3940-3945
Rae 2 Purification 21. Curnutte, J. T., Kuver, R., and Babior, B. M. (1987) J. Bid. Chem. 262, 6450-6452 22. Northup, J. K., Smigel, M. D., and Gilman, A. G. (1982) J. Biol. Chem.
."".
257. 11416-11423 "".
23. Hi ashijima, T., Ferguson, K. M., Sternweis, P. C., Smigel, M.D., and &man, A. G . (1987) J. Bcol. Chem. 262,762-766 24. Laemmli, U. K. (1970) Nature 227,680-685 25. Wray, W., Boulikas, T., Wray, P., and Hancock, R. (1981) Anal. Biochem. 1 1 8 , 197-203 26. Bokoch, G. M., Bickford, K., and Bohl, B. P. (1988) J. Cell Biol. 106, 1927-1936 27. Sha'ag, D., and Pick, E. (1990) Biochim. Bio h s Acta 1037,405-412 28. Quilliam, L. A,, Lacal, *J.-C., and Bokoch, (1989) FEBS Lett. 247,
&
-..
311-93G - "
29. Bokoch, G. M., and Parkos, C. A. (1988) FEBS Lett. 227,66-70 30. Didsbury, J., Weber, R. F., Bokoch, G. M., Evans, T., and Snyderman, R. (1989) J. Biol. Chem. 264, 16378-16382 31. Odink, K., Cerletti, N., Brueggen, J., Clerc, R. G., Tarcsay, L., Zwadlo, G., Gerhards, G., Schlegel, R., and Sorg, C. (1987) Nature 330,80-82 32. Lagasse, E., and Cierc, R. G. (1988) MOL. Cell. Biol. 8,2402-2410 33. Murao, S., Collart, F. R., and Huberman, E. (1989) J. Biol. Chem. 264, 8356-8360 34. Murao, S., Collart, F., and Huberman, E.(1990) Cell Growth Differentiation 1,447-454 35. Hall, A. and Self, A. J. (1986)J. Bid. Chem 261,10963-10965 36. Bokoch,'G. M., and V l l i a m , L: A. (1990) Biochem. J. 267,407-411 37. Goody, R. S., Frech, ., and Wlttlnghofer, A. (1991) Trends Blochem. Sei. 16,327-328 38. Hart M. J. Shinjo, K. Hall A., Evans, T., and Cerione, R. A. (1991) J. Si&. Cheh.266,20&0-26848 39. Gibbs, J. B., Schaber, M. D., Allard, W. J., Sigal, I. S., and Scolnick, E. M. (1984) Proc. Natl. Acad. SCL.U.S.A. 81,5704-5708 40. Frech, M., Schlichting, I., Wittinghofer, A,, and Chardin, P. (1990) J. Biol. Chem. 265,6353-6359 41. Abo, A,, Pick, E., Hall, A., Tot.ty, N., Teahan, C. G., and Segal, A. W. (1991 I. Nnturp. 353. ~~. ._, - - - fifi%670 , --42. Fukumoto, Y., Kaibuchi, K., Hori, Y., Fujioka, H., Araki, S., Ueda, T., Kikuchi, A., and Takai, Y. (1990) Oncogene 5, 1321-1328 43. Mizuno, T., Kaibuchi K. Ando, S., Musha, T., Hiraoka, K Takaishi, K. Asada, M., Nunoi, H., Matsuda, I., and Takai, Y. (1992)'k Biol. Chem!
_""
2 ~ 7 I""" I W I L I ~ ~ R
44. Clark R. A. Iyer, S., Pearson, D. W., Volpp, B. D., and Nauseef, W. M. (1962) C&. Ees. Abstr. 4 0 , 351 (abstr.) 45. Trahe M and McCormick, F. (1988) Science 238,542-545 46. Grandjk. A. and Owen D. (1991) Biochem. J. 279,609-631 47. Neal, S. F., Ecileston, J. FI, and Webb, M. R. (1990) Proc. Nutl. Acad. Sci. U. S.A. 87,3562-3565 48. Diekmann, D., Brill, S., Garret, M. D., Totty, N., Hsuan, J., Monthes, C., Hall, C., Lim. L., and Hall, A. (1991) Nature 351,400-402
a.
