Purification and Characterization of a GTP-binding ... - Semantic Scholar

Vol. 263, No. 20, Issue of July 15, pp. 9926-9932,1988 Printed in U.S.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc

Purification and Characterization of a GTP-binding Protein with a Molecular Weightof 20,000 in Bovine Brain Membranes IDENTIFICATION AS THE rho GENE PRODUCT* (Received for publication, December 15, 1987)

Katsuhiko Yamamoto, Jun KondoS, Tadashi HishidaS, Yutaka TeranishiS, andYoshimi TakaiQ From the Department of Biochemistry, Kobe University School of Medicine, Kobe 650 and the $Research Center, Mitsubishi Chemical Industries, Yokohama 227, Japan

We have determinedthat there are at least sixGTP- and inhibitory (Gi) G proteinsof adenylate cyclase and transbinding proteins (G proteins) with M, values between ducin, which links rhodopsin to cyclic GMP phosphodiester20,000 and 25,000 in the crude membrane fraction of ase (for reviews, see Refs. 1-4): Two additional G proteins bovine brain and have purified one of them with a M, designated as Go (5, 6) and G40 (7, 8) have recently been of about 24,000 (24K G) to nearhomogeneity (Kikuchi, identified and shown to belong to this family, although their A., Yamashita, T., Kawata, M., Yamamoto, K., Ikeda, definitive functions have not yet been clarified. Each G proK., Tanimoto, T., and Takai, Y. (1988)J. Biol. Chem. tein has a specific a, a common p, and a common y subunit 263, 2897-2904). In this study,we have purified another G protein with a M, of about 20,000 (20K G) to (1-6,8). The M , values of the respective a subunits areslightly near homogeneity and have characterized it. 20K G different and arebetween 39,000 and 52,000 (1,2,4-8). There bound maximally about 1.0 mol of [3sS]guanosine5’- are two forms of the P subunit with M , values of about 36,000 (3-0-thio)triphosphate (GTPyS)/molof protein, with a and 35,000 (1,2,4-6, 8,9). TheM , of the y subunit is about K d value of about 50 nM. [“SIGTPrS binding to 20K 10,000 (1, 2,4-6, 8). All the a subunits have both GTPG was inhibited by GTP and GDP, but not by other binding and GTPase activities (1-6, 8) and amino acid senucleotides such as ATP, UTP, and CTP; it was also quences corresponding to the GTP-binding and GTPase doinhibited by pretreatmentwith N-ethylmaleimide. mains described for elongation factor Tu (for reviews,see 20K G hydrolyzed GTP to liberatePi, with a turnover Refs. 1, 3, and 10). The GTP- and GDP-bound forms of the number of about 0.01 min-l, and was not copurified CY subunits have been shown to be active and inactive, respecwith the fir subunits of the regulatory G proteins of tively, for their specific effectors (1-4). adenylate cyclase. 20K G was not recognized by the In addition to this family of G proteins, there is another antibody against the ADP-ribosylation factor for the family of G proteins with M , values of about 20,000 in mamstimulatory regulatory G protein of adenylate cyclase. malian tissues. This family includes the G proteinsof the ras, Peptide map analysis showed that 20K G was not a rho, and ral gene products and other novel G proteins. Three proteolytic product of 24K G. The partial amino acid highly homologous ras genes (Ki-, Ha-, andN-ras) have been sequence of 20K G was almost identical with that identified (11-13). The amino acid sequences deduced from deduced from the rho gene. The amino acid composition the respective genes indicate that theproteins (ras p21) have of 20K G was similar to that of the rho gene product. M , values of about 21,000 and sequences corresponding to the These results suggest that 20K G is the rho gene product and that this G protein is present in bovine brain GTP-binding and GTPase domains which are homologous with those found in the G proteins described above. In fact, membranes. Ki- and Ha-ras p21s produced in bacteria have been shown to possess GTP-binding and GTPase activities (14-16). Although the protein molecules of the rho and ral genes have There is a family of G proteins’ with a subunit structureof not been identified in mammalian tissues and their properties aPy which serves as transducers for membrane receptors in have not been studied, these two genes are homologous with mammalian tissues. This family includes the stimulatory (G,) the ras genes (17-19). The amino acid sequences deduced from these genes indicate that theAplysia rho and simian ral * This work was supported in part by research grants from the gene products have M , values of about 21,500 and 23,500, Scientific Research Fund of the Ministry of Education, Science, and respectively, with sequences corresponding to the GTP-bindCulture, Japan, the Investigation Committee on Abnormalities in Hormone Receptor Mechanisms, the Ministry of Health and Welfare, ing and GTPase domains (17-19). Recently, a novel G protein Japan, the Yamanouchi Foundation for Research on Metabolic Dis- with a M , of about 21,000 designated as ARF was purified to orders, and the Research Program on Cell Calcium Signals in the near homogeneity from rabbit liver and bovine brain memCardiovascular System. The costs of publication of this article were branes (20, 21). Although the physiological function of ARF defrayed in part by the payment of page charges. This article must remains to be clarified, this G protein has been shown to serve therefore be hereby marked “advertisement” in accordance with 18 as a cofactor for the cholera toxin-dependent ADP-ribosylaU.S.C. Section 1734 solely to indicate this fact. tion of G, (20, 21). A G protein with a M , of 25,000 has also § To whom reprint requests should be addressed. The abbreviations used are: G proteins, GTP-binding proteins; been partially purified from human placentaand bovine brain G. and Gi, stimulatory and inhibitory regulatory GTP-binding pro- membranes (22, 23). This G protein is designated as G, (22, teins of adenylate cyclase; ARF, ADP-ribosylation factor; GTPrS, 23). ARF and G, have been shown to be different from ras guanine 5’43-0-thi0)triphosphate;SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; NEM, N-ethylmaleimide; p21 (21, 22). In this paper, the Gproteins with the apy CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-l-propanesulfo-subunit structure and those with M , values of about 20,000 are designated as large and small M , G proteins,respectively. nate.

9926

Characterization Purification and In a preceding report from our laboratory (24),we separated at least six G proteins with M , values between 20,000 and 25,000 from the crude membrane fraction of bovine brain and purified one of them with a M, of about 24,000 (24K G) to near homogeneity. The characterization of24K Ghas revealed that this G protein is distinct from ras p21, ARF, and G,. Moreover, we have recently cloned and sequenced the cDNA of 24K G.' Homology analysis of the complete amino acid sequence deduced from the cDNA of 24K G indicates that this Gprotein is a novel one distinct from other G proteins described previously. Extending these observations, we have purified another G protein with a M, of 20,000 to near homogeneity from the crude membrane fraction of bovine brain and have characterized it. This G protein is designated here as 20K G. The properties of 20K G suggest that this G protein is the rho gene product. This paper described for the first time the purification procedures and properties of the G protein which is likely to be the rho gene product. EXPERIMENTAL PROCEDURES

Materials and Chemicals-The crude membrane fraction was prepared from bovine brain as describedpreviously (24). 24K Gwas purified to near homogeneity from the crude membrane fraction of bovine brain as described (24). A Bakerbond W P octyl column (4.6 X 250 mm) was purchased from J. T. Baker Chemical Co. Other materials and chemicals were obtained from the same sources as described previously (24). p'S]GTPyS-binding and GTPmeA S S ~ ~ S - [ ~ ~ S ] G binding T P ~ S to G proteins was determined by use of the nitrocellulose filter method under the conditions specified previously (24). GTPase activity of 20K G was estimated by the liberationof 32Pifrom [Y-~'P]GTP under the conditions specified (24). Peptide Map Analysis-20K G or 24K G (15 pg of each protein) was desaltedon a Sephadex '2-25 (PD-10)columnwith distilled water. Each sample was digested with Achromobacter protease I a t a molar ratio of 1:300 (Achromobacter protease 1:each sample) for 6 h a t 37 "C in 200 pl of 50 mM Tris/HCl a t p H 9.0. The reaction was terminated by the addition of formic acid a t a final concentration of 1%. The Achromobacter protease I-digested peptides (200 pl) were applied to a Bakerbond W P octyl column preequilibrated with 0.1% trifluoroacetic acid. The elution was performed with a 60-ml linear gradient of acetonitrile (0-60%) in a 0.1% trifluoroacetic acid a t a flow rate of 1 ml/min. Peptides were detected by measuring absorbance at 215 nm. Analysis of Amino Acid Sequence-Two major peptides of 20K G separated by a Bakerbond W P octyl column were sequenced using an Applied Biosystems Model 470A automated Gas-Phase Sequencer. Phenylthiohydantoins were identified by high performance liquid chromatography according to the methodof Tsunasawa et al. (25). Analysis of Amino Acid Composition-Amino acid analysis of 20K G was carried out with an amino acid analyzer (Mitsubishi Chemical Industries, Model MCI AAO1). A purified preparation of 20K G (5.4 pg of protein) washydrolyzedfor 24 and 72 h in30 p1 of 4 M methanesulfonic acid containing 0.2% 3-(2-aminoethyl)indole a t 110 "C in vacuo. Molar ratios of serine and threonine were corrected for destruction by extrapolation of hydrolysis for 0 h. Molar ratiosof the sample hydrolyzed for 72 h were used for those of valine, isoleucine, and leucine. Determinations-The radioactivity of 35S-and 32P-labeled samples was determined using a liquid scintillation system (Beckman Instruments, Model LS 3801). Protein was determined by the method of Lowry et al. (26) with bovine serum albumin as a standard. Where specified, protein was alternatively determinedby densitometric tracing of protein bands stained by Coomassie Brilliant Blue on a SDSpolyacrylamide gel by the method of Weber et al. (27) with bovine serum albumin as a standard.

of 20K G

9927

Fraction Number

FIG. 1. Mono Q HR5/5 column chromatography. [36S] GTPyS-bindingactivity of 5 p1 of eachfraction was assayed. M, [35S]GTPySbound; - - -, absorbance a t 280 nm; -, NaCl concentration.

brain. The extracted G proteinswere subjected to gel filtration on an Ultrogel AcA-44 column, by which G proteins with M, values between 20,000 and 25,000 were separated from large M, Gproteins,such as G., Gi, and Go. After small M , G proteins were further purified by phenyl-Sepharose CL-4B column chromatography, the active fractions were subjected to hydroxylapatite column chromatography, which yielded two peaks of G proteins.The active fractions of the first peak were collected, concentrated, and then subjected to Mono Q HR5/5 column chromatography. When each fraction was assayed for GTPyS-binding activity, four peaks appeared, as shown in Fig. l.3All of these purification procedures were the same as those described previously (24), except that a YM-5 filter was used instead of a PM-10 filterfor the concentration of the G proteins with an Amicon ultrafiltration apparatus because ARF partly passed through a PM-10 filter. The active fractions of the third peak (fractions 46-50) were pooled. The pooled fractions (2.5 ml, 66 pg of protein) were diluted with 22.5ml of20mM Tris/HCl at pH 8.0 containing 1 mM EDTA, 1 mM dithiothreitol, 5 mM MgCL, and 0.6% CHAPS to decrease the NaCl concentration and were then subjected to rechromatography on a Mono Q HR5/ 5 column preequilibratedwith the same buffer. After the column was washed with 5 ml of the same buffer, elution was performed with a 30-ml linear gradient of NaCl (50 mM to 0.25 M ) in the same buffer at a flow rate of0.5 ml/min. Fractions of 0.5 ml each were collected. When each fraction was assayed for GTPyS-binding activity, two peaks appeared, as shown in Fig. 2 A . Fig. 2B shows the protein staining of each fraction on SDS-polyacrylamide gel. The second peak was apparently homogeneous. The active fractions of the second peak (fractions 30-34) were pooled and characterized. Since this G proteinshowed a M, of about 20,000 as described below, it is tentatively designated as 20KG. The proteins corresponding to the (3 subunits (Mr = 36,000 and 35,000) of large M , G proteins were not observed in the 20K G fractions. The first peak of this column chromatography was not purified further here. A summary of typical purification of20K G is shown in

RESULTS

Purification of 20K G-G proteins were extracted with sodium cholate from the crude membrane fraction of bovine Y. Matsui, A. Kikuchi, J. Kondo, T. Hishida, Y. Teranishi, and Y. Takai, manuscript submitted toJ . Biol. Chem.

We showed previously that five GTPyS-binding peaks were separated by Mono Q HR5/5columnchromatography(24).Inthis experiment, the second and third peaks of this column chromatographyshowninour previous paper(24) were notseparatedand appeared in one broad peak (second peak). poor This resolution might be due to the decrease in the quality of our Mono Q HR5/5 column.

5

20K G

Characterization Purification andof

9928

a single band was observed, as shown in Fig. 3A. The M,of 20K G was estimated to be about 20,000, as shown in Fig. 3B. In this experiment, 20K G was boiled in the presence of 2mercaptoethanol and subjected to SDS-PAGE in this reducing agent. When the same experiment was done under nonreducing conditions, the same result was obtained (data not shown). p5S]GTPyS-binding Actiuity"20K G bound [35S]GTPyS in a time-dependent manner. Fig. 4 shows the comparison of the time courses of 20K G and 24K G. [35S]GTPySbound to 20K G more slowly than to24K G. 20K G bound [35S]GTPyS in a dose-dependent manner. Scatchard plot analysis showed that both 20K G and 24K G bound [35S]GTPySmaximally a t 1.0 f 0.1 mol (mean f S.E.)/mol of protein, with K d values of 50 f 5 and 46 f 5 nM (mean f S.E.), respectively. The K d value of20K G was 2-4-fold higher than those of G.,Gi, B transducin, Go, and ras p21 (1,2,4, 14, 15,28). 36Specificity for Nucleotide~-[~~S]GTPyS binding to 20K G 2 was progressively inhibited by increasing concentrations of x nonradioactive GTPyS,GTP,or GDP. Among thethree t200.nucleotides, GTPyS was 2-3-fold more active than GTP and GDP.Both GTP and GDP showed no difference in this 18 20 22 211 26 28 73 32 34 36 38 40 capacity. Other nucleotides such as ATP, CTP, and UTP Fraction Elcsber were not effective. FIG.2. Re-Mono Q HR5/5 column chromatography. A, [35S] Inhibition of p5S]GTPyS Binding to 20K G by Pretreatment GTPyS-binding activity. [35S]GTP$3-bindingactivity of 20 pl of each fraction was assayed. The symbols are the same as those used with NEM-It has been shown that [3sS]GTPyS binding to in Fig. 1. B , protein staining of each fraction on SDS-polyacrylamide Ha-rus p21 is inhibited by pretreatment with NEM (29). We gel. The indicated fractions (40pl each) were subjected to SDS-PAGE have previously confirmed this observation with Ki-ras p21 (12% polyacrylamide). The proteinswere visualized by silver staining. (24). [35S]GTPyS binding to 20K G was also inhibited by The protein markers used were glyceraldehyde-3-phosphate dehydro- pretreatment with NEM, and this inhibition was blocked by genase (M, = 36,000) and trypsin inhibitor (M. = 20,100). the simultaneous presence of dithiothreitol. In contrast to ras p21 and 20K G, [35S]GTPyS binding to 24K G was not Table I. When 20K G was kept frozen a t -20 "C,GTP-binding inhibited by pretreatment with NEM, as described previously activity was not lost, a t least for 2 weeks. (24). The amount of 20K G in the cholate extract from bovine GTPase Activity-20K G showed GTPase activity as estibrain membraneswas estimated to be less than thoseof other mated by the liberation of 32Pifrom [y-32P]GTP.Fig. 5 shows well-known G proteins. About 21 pg of 20K G was purified the time courses of 20K G and 24K G for the GTPaseactivity. from 858 mg of the cholate extract from the crude membrane There were lag periods for the GTPase activities of both G fraction. It is difficult to calculate the exact yield of 20K G proteins, but the lag period for 20K G was longer than that by our purification procedures since at least six small M , G for 24K G. This difference might be due to the fact that [35S] proteins were separated by column chromatography. How- GTPyS binding to 20K G required a longer time than that to ever, on the assumption that these six small M , G proteins 24K G (see Fig. 4). The turnovernumber of20K G was are purified with the same yields, the yield of the purification calculated to be about 0.01 f 0.002min" (mean f S.E.), of 20K G is calculated to be about 11%. Therefore, it could whereas that of 24K G was about 0.04 f 0.006 min" (mean be estimated that about 0.02% of the total proteins in the k S.E.), as described (24). The value of 20K G is far less than cholate extract of the crude membrane fraction was 20K G. those of G., Gi, tranducin, and Go, but is comparable to that The amount of 20K G is one-fifth those of 24K G, G., and of c-rus p21 (1,2,4,6, 16). ARF and much less (at least 40-fold less) than those ofGi Immunological Cross-reactiuity of 20K G with Antibody and Go. against ARF-We have previously shown that ARF is eluted Physical Properties-When the purified preparation of 20K on the first peak of Mono Q HR5/5 column chromatography G was subjected to SDS-PAGE (8-16% polyacrylamide) and (24). This result indicates that 20K G is separable from ARF the protein was stained with silver after electrophoresis, only by this column chromatography. In order to obtainadditional

--

TABLE I Purification of 20K G Purification step

Total volume

ml

Total protein mR

Total GTP+ binding amount nmol

Specific activity nmollmg

Yield %

Cholate extract 0.56858 484 100 Ultrogel AcA-44 (second peak) 200 46 90 2.5 223 CL-4B Phenyl-Sepharose 120 32 142 4.4 29 11 peak) Hydroxylapatite (first 240 6.8 51 7.5 peak) (third Mono Q HR5/5 2.5 0.08" 2.9 36 0.6 0.21 40 (second 1.0Deak) 0.025" 2.5 Re-Mono Q HR515 a Values were determined by densitometric tracing of the gel after SDS-PAGE asdescribed under "Experimental Procedures."

Purification and Characterization of 20K G

9929

A

1

2

, 4

Molecular Weight ( x

I

a

FIG. 3. Analysis of 20K G on SDS-PAGE. A, protein staining, 20K G (200 ng of protein) was subjected to SDS-PAGE (8-16% polyacrylamide). The protein was visualized with silver staining. B, estimation of Mr. The proteinmarkers usedwere bovine serum albumin (M,= 66,000) (point I ) , glyceraldehyde-3-phosphate dehy= drogenase (point2), trypsin inhibitor (point3 ) ,and myoglobin (M, 17,000) (point4 ) . The arrow indicates the position of 20K G.

r"

Incubatlon Tlme (h) FIG. 4. Time courses of ['?3]GTPrS binding to 20K G and 24K G . [3sSS]GTPyS-bindingactivity of 20K G or 24K G (50 ng of each protein) was assayed for various periods of time. The radioactivity of ["SIGTPyS maximally bound to 20K G or 24K G was about 4000 cpm. 0,20K G; 0,24K G. The results shown are representative of three independent experiments.

evidence that 20K G is distinct from ARF, cross-reactivity of 20K G with the polyclonal antibody against ARF was examined. Fig. 6A shows protein staining of 20K G and ARF on SDS-polyacrylamide gel. ARF migrated slightly faster than 20K G.4 Fig. 6B shows the immunoblot analysis of both proteins using the antibody against ARF. This antibody did not recognize 20K G under the conditions where it reacted markedly with ARF. Peptide MapAnalysis of 20K G and 24K G-In order to rule out that20K G is a proteolytic product of 24K G, we compared 'Although the M,of ARF was estimated to be about 21,000 by Kahn and Gilman (20,21), it was estimated to be about 19,500 under our conditions. The exact reason for this difference is not known, but may be due to the different molecular markers used in their and our laboratories.

FIG. 5. GTPase activities of 20K G and 24K G . GTPase activity of 20K G or 24K G (100ng of each protein) was assayed with 1 p~ [-p3'P]GTP for various periods of time. 0, 20K G; 0, 24K G. The results shown are representative of three independent experiments.

1

2

1

2

FIG.6. Immunological cross-reactivity of 20K G with antibody against ARF. A, protein staining. 20K G or ARF (200 ng of each protein) was subjected to SDS-PAGE (8-16% polyacrylamide). The proteins were visualized with silver staining. Lane 1, 20K G; lane 2, ARF. B, immunoblot analysis of 20K G and ARF. 20K G or ARF (200 ng of each protein) was subjected to SDS-PAGE (8-16% polyacrylamide). After the protein on the gel was transferred to a nitrocellulose sheet, the sheet was immunoblotted with the polyclonal antibody against ARF. Lane I , 20K G; lane 2, ARF.

the peptide maps of both proteins. Pure samples of 20K G and 24K G were digested completely by Achromobacter protease I. This protease has been shown to hydrolyze specifically lysyl bonds (30). The digested samples were separately subjected to Bakerbond WP octyl column chromatography under the same conditions. The peptides were monitored spectrophotometrically at 215 nm. The peptide peaks of 20K G did not coincide with those of 24K G at all, as shown in Fig. 7. Partial Amino Acid Sequence and Amino Acid Composition of 2OK G-The major peaks, peaks 1 and 2 shown in Fig. 7, were sequenced by use of an automatic Sequencer. The analysis of sequence homology revealed that the amino acid se-

Purification and Characterization of 20K G

9930

TABLEI1 Comparison of amino acid compositions of 2OK G and the Aplysia rho gene product Amino acid

20K Go

Aplysio rho gene productb

no. residueslrnolecule

21.1

Asx Thr Ser Glx Pro G~Y Ala CYS Val Met Ile Leu TYr Phe LYS His Arg

21.9 15.8 19.0

Time (mln)

FIG. 7. Peptide map analysisof digested peptides generated from 20K G and 24K G . 20K G or 24K G (10 pg of each protein) was digested with Achromobacter protease I. The digested protein was subjected to Bakerbond WP octyl column chromatography. Peptide peaks were monitored spectrophotometrically at 215 nm. The peaks indicated by arrows were not peptides, but CHAPS included in the purified preparations of 20K G and 24K G. A, 20K G; B, 24K G.

21 9 8 21 10.1 11 12.2 10 13 ND' 6 19 O.Od 5 8.3 10 15.6 15 6.5 6 6.3 6 13.0 19 1 2.7 11.8 10 Trp 0.3d 2 The number of residues/molecule was calculated according to a M, of 20,000 for 20K G. * Values were calculated from the data of Madaule and Axel (17). e ND, not determined. Methionine and tryptophan were destroyed by oxidation under our conditions.

9.2 11.0

human rho gene products share 35% amino acid homology Ha-ras p21 (17), and therho and ras gene products have with MAAIRKKLVIVGDGACGKTCLLIVFSKDQFPEVYVPTVFENYVADIEMG Aplysia EFPEVYVPTVFENYVADIEMG common properties. 1) Both have similarM , values of about Human 20K G 20,000; 2) both have amino acid sequences corresponding t o the GTP-binding and GTPase domains; 3) both have car70 80 90 100 60 Aplysia e KQVELALWDTAGQEDYDRLRPLSYPDTDVILMCFSIDSPDSLENIPEKWT boxyl-terminal sequences required for attachment to the inner Human rho KQVELALWDTAGQEDYDRLRPLSYPDTDVILMCFSMSPDSLENIPEKWV a leaflet of the cytoplasmic membranes; and 4) both have 20K G QVELALWDTAGQEDYDRLRYLSY (Peak 21 wv cysteine residue in the vicinity of the GTP-binding domain 120 110 130 140 150 (17). In spiteof these similaritiesbetween both proteins, they Aplysia PEVRHFCPNVPIILVGNKKDLRNDESTKRELMKMKQEPVRPEDGRAMAEK Human rho PEVKHFCPNVPIILVANKKDLRSDEHVRTELARMKQEPVRTDDGWVR have variable regions. Particularly, maximaldivergence is 20K G PEVK (Peak 1 ) observed near the carboxyl-terminal aminoacid (17). In this paper,we have purified 20K G to near homogeneity 160 170 180 190 Aplysia INAYSYLECSAKTKEGVRDVFETATRAALOVKKKKKGGCW and characterized it. 20K G is a GTP-binding protein witha Human IQAYDYLECSAKTKEGVREVFETATRAALQKRYGSQNGCINCCKVL Kd value for GTPyS of about 50 nM and has GTPase activity 20K G FIG. 8. Comparison of amino acid sequence of 2 0 K G with with a turnover number of about 0.01 min". 20K G is found those deduced from Aplysia and humanrho cDNAs. The amino in the crude membrane fraction of bovine brain and is exacid sequences of peaks 1and 2 are shown in Fig. 7A were determined. tracted withsodium cholate from the membranes. The partial The sequence data of the Aplysia and human rho cDNAs were taken amino acid sequence of 20K G is almost identical t o those of from the data of Madaule and Axel (17). the Aplysia and human rho gene products. Moreover, the amino acid composition of 20K G is similar to that of the quences of peaks 1 and 2 were almost identical to those of Aplysia rho gene product, with the exceptionof a few amino residues 99-104 and 52-74, respectively, of the Aplysia and acids. The contentof lysine in 20K G was considerably lower human rho gene products deducedfrom their cDNAs, as than that in the Aplysia rho gene product. This difference shown in Fig. 8. The amino acid composition of 20K G was may be due to the species difference since the Aplysia rho also similar to thatof the Aplysia rho gene product, with the gene product has a high lysine content, but the human rho exception of a few amino acids, as shown in Table 11. The gene product does not, in the variable regions near the caramino acid composition of 20K G wasnot compared with thatboxyl-terminal amino acid of each protein (17). These propof the human rho gene product since the total amino acid erties of 20K G strongly suggest that this G protein is the sequence of the human rho gene product was not determined bovine brain rho gene product, although the possibility cannot (17). be neglected completely at present that the 20K G gene is highly homologous with, but distinct from, therho gene until DISCUSSION the totalnucleotide sequence of the 20K G gene is determined. a cysteine residue in the vicinity The rho gene (the ras-homologous gene) was first isolated It has been shown (29) that from a cDNA library from the abdominal ganglia of Aplysia of the GTP-binding domain of Ha-ras p21 is sensitive t o a by Madaule and Axel (17). This gene was also detected in thiol-specific reagent, NEM, and that pretreatment of this yeast, Drosophila, rat, and man (17). The rho gene is con- protein with NEM diminishes its GTP-binding activity. Simserved, and amino acid sequences of the Aplysia and human ilarly, the GTPyS-binding activity of 20K G is inhibited by gene products are 85% homologous (17). The Aplysia and pretreatment with NEM. This result is consistent with the 10

20

30

40

50

Purification Characterization and fact that the Aplysia and human rho gene products have a cysteine residue in the vicinity of the GTP-binding domain (17). Bovine brain 20K G shows a M , of about 20,000. This M, is slightly different from that of the Aplysia rho gene product (Mr = 21,500). The slight difference of the M , values between bovine 20K G and theAplysia rho gene product may be due to the difference of amino acid sequences. During the preparation of this manuscript, Anderson and Lacal (31) purified the Aplysia rho gene product overexpressed in Escherichia coli and characterized it. The Aplysia rho protein shows GTP-binding and GTPaseactivities, with a K d value for GTP of 303 nM and a turnover number of 0.106 min", respectively. These values are quite different from those of 20K G purified from the crude membrane fraction of bovine brain in this paper. The exact reason for these differences is not clear, but may be due to the different origins or assay conditions or to differences between the proteins expressed bacterially and those purified from a mammalian tissue. 20K G shows properties similar to those of ras p21 as predicted from the structure of the rho gene product (17). 1) Both have similar M , values of about 20,000; 2) both have GTP-bindingandGTPase activities; 3)the GTP-binding activities of both are sensitive to pretreatment with NEM; and 4) both are localized in the membranes. In spite of the similarities between 20K G and ras p21, the partial amino acid sequence of 20K G is different from that of ras p21. Moreover, c-ras p21 is mostly eluted in the second peak and is partly eluted in the first peak, whereas 20K G is eluted in the first peak by hydroxylapatite column chromatography, as described previously (24). c-ras p21slightly contaminates the third peak (20K G) of the first Mono Q HR5/5 column chromatography, but this contaminating c-ras p21 is separated from 20K G by re-Mono Q HR5/5 column chromatography (data not shown). Moreover, several monoclonal antibodies against ras p21 are notcross-reactive with 20K G (data notshown).Theseresults indicate that 20K G is clearly distinct from c-ras p21. In addition to the rho gene, three ras-related genes have been identified. One is designated as the ral gene and has been identified in simian B-lymphocytes (19). The ral gene codes for a protein with a M, of about 23,500 and shares 50% amino acid homology with the Ki-, Ha-, and N-rasgene p21s (19). The protein molecule of the ral gene has been neither identified nor produced in bacteria, so its properties have not been studied. However, its amino acid sequence indicates that the ral gene product is also a G protein (19).Two other genes are designated as YPTl and SEC4 and have recently been found in yeast (32-34). These yeast genes encode G proteins with similar M , values of about 23,500 (32-34).The yptlgene, which codes for a G protein with 71%homology with the yeast YPTl gene product, has more recently been identified in mouse F9 cells (35); but the presence of the SEC4 gene in mammalian tissues has not been determined. The partially determined amino acid sequence of 20K G is different from that deduced from the cDNA of ral, YPT1, yptl, or SEC4. Moreover, two additional small M , G proteins have been purified and well characterized: one is ARF purified from the crude membranes of rabbit liver and bovine brain by Kahn and Gilman (20, 2l), and the other is 24K G purified from the crude membranes of bovine brain in our laboratory (24). 20K G is distinguishable from ARF by the following properties. 1) 20K G and ARF are separable by Mono Q HR5/5 column chromatography and SDS-PAGE; 2) 20K G is not recognized by the polyclonal antibody against ARF; and 3) 20K G shows GTPase activity, whereas ARF has been described to lack this activity (21). 20K G is also distinguishable

of 20K G

993 1

from 24K G by the following properties. 1)20K G and 24K G are separable by Mono Q HR5/5 column chromatography and SDS-PAGE (data not shown); 2) GTPyS-binding activity of 20K G is sensitive to pretreatment with NEM, whereas that of 24K G is resistant to this pretreatment; 3) the time course of GTPyS binding to 20K G is slower than that of24K G; and 4) the turnover number of20K G for GTPase is onefourth that of24KG. It is also evident from peptide map analysis that 20K G is not a proteolytic product of 24K G. Since 20K G is most likely to be the rho gene product and 20K G is distinct from ARF and 24K G, it could be concluded that ARF and 24K G are not rho gene products. In fact, we have recently cloned and sequenced the cDNA of24KG.' Homology analysis of the complete amino acid sequence deduced from the cDNA of 24K G indicates that this G protein is different from the rho gene product. There is another small M , G proteindesignated as G, which has been partially purified first from human placenta and subsequently from bovine brain (22, 23). The M , ofG, is about 25,000 (23). The properties ofG, have not been well studied, but this G protein has been shown to be copurified with the Py subunits similar to those of large M , G proteins. It has been suggested, on the basis of this observation, that G, is associated with the By subunits. Although the exact relationship between 20K G and G, is not known at present, it is likely that 20K G is different from G, since 20K G is not copurified with the Py subunits. The physiological functions of large M , G proteins have been investigated extensively, and this family of G proteins has been clarified to serve as transducers in transmembrane signaling (1-4). In contrast, the physiological functions of small M, G proteins including 20K G have not been clarified. The yeast RAS2 gene product has been shown to be involved in the regulation of adenylate cyclase (36). The mammalian ras and yeast RASl gene products have been suggested to be involved in the regulation of the phospholipase C-mediated hydrolysis of phosphoinositides (37-39). Therefore, it is conceivable that small M , G proteins are also involved in transmembrane signaling, but presumably in a manner different from that of large M , G proteins. Besides the functions of small M , G proteins in transmembrane signaling, the YPTl gene product has been shown to be involved in the regulation of microtubular arrangement (33), and theSEC4 gene product in the secretory processes in yeast (34). By analogy, some mammalian small M , Gproteins may also be related to cytoskeletal arrangement andthe regulation of secretory processes. Investigation of the functions of small M , G proteinsis essential to understand the new regulatory mechanisms of various cell functions. Acknowledgments-We thank Dr. R. A. Kahn (National Institutes of Health) for providing bovine brain ARF and its specific polyclonal antibody and Dr. T. Masaki (Ibaragi University, Ibaragi, Japan) for supplying pure Achromobacter protease I. We are grateful to Junko Yamaguchi for skillful secretarial assistance. REFERENCES

1. Ui, M. (1984) Trends Pharmacol. Sci. 5, 277-279 2. Bourne, H. R. (1986) Nature 321, 814-816 3. Gilman, A. G. (1987) Annu. Reu. Biochern. 56,615-649

4. Stryer, L.(1986) Annu. Rev. Neurosci. 9,87-119 5. Sternweis, P. C., and Robishaw, J. D. (1984) J. Biol. Chem. 2 5 9 , 13806-13813 6. Katada, T., Oinuma, M., and Ui, M. (1986) J. Biol. Chern. 2 6 1 , 8182-8191 7. Gierschik, P., Falloon, J., Milligan, G., Pines, M., Gallin, J. I., and Spiegel, A. (1986) J. Biol. Chem. 261,8058-8062 8. Oinuma, M., Katada, T., and Ui, M. (1987) J. Biol. Chem. 262, 8347-8353

9932

Purification and Characterization of 20K G

9. Manning, D. R., and Gilman, A. G. (1983) J. Biol. Chem. 2 5 8 , 7059-7063 10. Jurnak, F. (1985) Science 2 3 0 , 32-36 11. Ellis, R.W., DeFeo, D., Shih, T. Y., Gonda, M. A., Young, H. A., Tsuchida, N., Lowy, D. R., and Scolnick, E. M. (1981) Nature 292,506-511 12. Hall, A., Marshall, C. J., Spurr, N. K., and Weiss, R. A. (1983) Nature 303, 396-400 13. Shimizu, K., Goldfarb, M., Suard, Y., Perucho, M., Li, Y., Kamata, T., Feramisco, J., Stavnezer, E., Fogh, J., andWigler, M. H. (1983) Proc. Natl. Acad. Sci. U. S. A . 80, 2112-2116 14. Scolnick, E. M., Papageorge, A. G., and Shih, T. Y. (1979) Proc. Natl. Acad. Sci. U. S. A . 7 6 , 5355-5359 15. Shih, T. Y., Papageorge, A. G., Stokes, P. E., Weeks, M. O., and Scolnick, E. M. (1980) Nature 287, 686-691 16. Gibbs, J. B., Sigal, I. S., Poe, M., and Scolnick, E. M. (1984) Proc. Natl. Acad. Sci. U. S. A . 81, 5704-5708 17. Madaule, P., and Axel, R. (1985) Cell 41,31-40 18. Madaule, P., Axel,R., and Myers, A. M. (1987) Proc.Natl.Acad. Sci. U. S. A . 84, 779-783 19. Chardin, P., and Tavitian, A. (1986) EMBO J. 5 , 2203-2208 20. Kahn, R. A., and Gilman, A. G. (1984) J. Biol. Chem. 259,62286234 21. Kahn, R. A., and Gilman, A. G. (1986) J. Biol. Chem. 261,79067911 22. Evans, T., Brown, M. L., Fraser, E. D., and Northup, J. K. (1986) J. Biol. Chem. 261,7052-7059 23. Waldo, G.L., Evans, T., Fraser, E. D., Northup, J. K., Martin, M. W., and Harden, T. K. (1987) Biochem. J. 246,431-439 24. Kikuchi, A., Yamashita, T., Kawata, M., Yamamoto, K., Ikeda,

25. 26. 27. 28. 29. 30.

31.

K., Tanimoto, T., and Takai, Y. (1988) J. Biol. Chem. 2 6 3 , 2897-2904 Tsunasawa, S., Kondo, J., and Sakiyama, F. (1985) J. Biochem. (Tokyo)97, 701-704 Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275 Weber, K., Pringle, J. R., and Osborn, M. (1972) Methods Enzymol. 26,3-27 Northup, J. K., Smigel, M. D., and Gilman, A.G. (1982) J. B i d . Chem. 257,11416-11423 Hattori, S., Yamashita, T., Copeland, T. D., Oroszlan, S., and Shih, T. Y. (1986) J. Biol. Chem. 2 6 1 , 14582-14586 Masaki, T., Fujihashi, T., Nakamura, K., and Soejima, M. (1981) Biochim. Biophys. Acta660,51-55 Anderson, P. S., and Lacal, J. C. (1987) Mol. Cell. Biol. 7, 3620" "

3WL8

32. Gallwitz, D., Donath, C., and Sander, C. (1983) Nature 306,704707 33' Schmitt9 H' D', Wagner' p'p pfaff*E'p and D. (1986) Cell 47,401-412 34. Salminen, A., and Novick, P. J. (1987) Cell 49,527-538 35. Haubmck, H., ~ i ~c., ~wagner, l ~ p,, , and Gallwitz, D. (1987) EMBO J. 6,4049-4053 36. Broek, D., Samiy, N., Fasano, O., Fujiyama, A., Tamanoi, F., Northup, J., and Wigler, M. (1985) Cell 4 1 , 763-769 37. Fleischman, L. F., Chahwala, S. B., and Cantley, L. (1986) Science 231,407-410 38. Wakelam, M. J. O., Davies, S. A., Houslay, M.D., McKay, I., Marshall, C. J., and Hall, A. (1986) Nature 323, 173-176 39. Kaibuchi, K., Miyajima, A., Arai, K., and Matsumoto, K. (1986) Proc.Natl.Acad.Sci. U.S. A . 83,8172-8176