mapping of the mastoparan-binding site on g proteins

THE JOURNAL OF BIOLOGICAL CHEMISTRY {c) 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

Vvl. 266, No. 19, Issue of July 5 , PP. 12655-12661,1991 Printed in U.S.A.

Mapping of the Mastoparan-binding Site on G Proteins CROSS-LINKING OF [12"I-Tyr3,Cys"]MASTOPARAN TO Go*

(Received for publication, February 5, 1991)

Tsutomu Higashijima and Elliott M. Ross From the Department of Pharmacology. Southwestern Graduate School of Biomedical Sciences, University of Texas Southwester; Medical Center, Dallas,-Texas 75235-9041

Mastoparan (MP) activates GTP-binding regulatory proteins (G proteins) by promoting GDP/GTP exchange through a mechanism similar to that of G protein-coupled receptors (Higashijima, T., Burnier, J., and Ross, E. M. (1990) J. Biol. Chem. 265, 14176- 14186). [Tyr", Cys"]MP was synthesized and shown to have regulatory activity similar to that of mastoparan when assayed in the presence of dithiothreitol (DTT). Activation by [ T ~ r ~ , c y s ~ ~ ] M P in the absence of DTT was complex in its kinetics, concentration dependence, and dependence on detergents. [lZ5I- Tyr3,Cys"]MP bound covalently to the a subunit of G proteins. Cross-linking was blocked by mastoparan or [Tyr3,Cys"]MP. Cross-linking was enhanced by the addition of /3r subunits, but no cross-linking to 8-y subunits was observed. Cross-linking was inhibited by incubation of Go with guanosine 5'-O-(thiotriphosphate) and Mg2+ and was reversed by incubation with DTT or 2-mercaptoethanol. Stoichiometry of labeling was consistent with the cross-linking of one molecule of [ 1251-Tyr3,Cy~11]MP/~ subunit, and CNBr hydrolysis of the ['z51-Tyr3,Cys"]MP-ao adduct yielded one major labeled peptide fragment of -6 kDa. Amino acid sequencing of this CNBr fragment prepared from recombinant a. showed that cross-linking occurred at Cys3. No a, sequence was obtained from the same fragment prepared from bovine brain ao, which is blocked by a myristoyl group at Gly'. Regulation of Go by MP was eliminated by tryptic proteolysis of the amino-terminal region. These observations suggest that the amino- terminal region of G protein a subunits contributes to the mastoparan-binding site, which may also be the receptor-binding site, and is involved in regulation of nucleotide exchange.

G protein'-mediated signaling depends on selective mutual recognition between G proteins and receptors, each of which are members of large families of homologous proteins. G proteins are activated by binding GTP, and activation is terminated when the bound GTP is hydrolyzed to GDP.

GM40676 and GM30355 and R. A. Welch Foundation Grant 1-0982. * This work was supported by National Institutes of Health Grants

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The abbreviations used are: G protein, GTP-binding regulatory protein; MP, mastoparan; [I-Tyr:',Cys"]MP, [monoiodo-Tyr3,Cys1'] mastoparan; DTT, dithiothreitol; NEM, N-ethylmaleimide; PTH, phenylthiohydantoin; SDS, sodium dodecyl sulfate; GTPrS, guanosine 5'-O-(thiotriphosphate); Hepes, 4-(2-hydroxyethyl-l-piperazine- ethanesulfonic acid PVDF, polyvinylidenedifluoride; HPLC, high pressure liquid chromatography.

Receptors promote the activation of G proteins by catalyzing the replacement of GDP by GTP (see Refs. 1 and 2 for review).

G proteins are heterotrimers that are bound to the inner face of the plasma membrane via the @ and y subunits, which are shared among different G proteins. The unique a subunits, which bind GTP and activate effector proteins, determine the identity of G protein trimers and are the primary determi- nants of selectivity for receptors (see Ref. 1 for review). G protein-coupled receptors are integral plasma membrane proteins. They are based on a core of seven membrane-spanning helices oriented such that the amino terminus is extracellular and the carboxyl terminus is cytoplasmic (see Ref. 3 for review). Because G protein a subunits and G protein-coupled receptors are each families of homologous proteins, it is likely that the structure of the receptor-G protein interface is conserved, with only enough variability to allow receptors to selectively activate some G proteins, but not others.

Mastoparan is a peptide toxin that promotes GDP/GTP exchange on G proteins through a mechanism that is virtually identical to that of agonist-liganded receptors. Mastoparan catalyzes GDP/GTP exchange at submicromolar Mg'+ concentrations and does not alter the rate of hydrolysis of bound GTP. Its action is blocked by pertussis toxin-catalyzed ADP- ribosylation of the G protein a subunit and is enhanced by the presence of subunits (4, 5). Mastoparan also competes with receptors for binding to G proteins (5).2 Mastoparan forms an amphiphilic a helix when it binds to phospholipid membranes and presents four positive charges to the aqueous face (6). The sites on the second and third cytoplasmic loops of receptors that are thought to interact with G proteins are also positively charged and near the membrane surface (Refs. 7-12; see Ref. 3 for review). Based on these data, it seems likely that mastoparan mimics the G protein-binding domain on receptors and binds to G protein a subunits at their receptor-binding sites ( 5 ) .

Mastoparan binds and modulates the activities of several proteins other than G proteins, most notably calmodulin (13); and at high concentrations, it exerts nonspecific effects on membranes. We are currently designing mastoparan analogs that are more specific for G proteins relative to other cellular sites and that are selective among related G proteins. How- ever, before mastoparan can be safely used as a model for receptors, it is necessary to demonstrate the existence of a discrete mastoparan-binding site on G protein a subunits. Because even the most potent mastoparan congeners act at concentrations >lo0 nM, it is unlikely that equilibrium binding measurements can readily establish the existence of such a site. This report describes the cross-linking of a mastoparan analog to a unique site on the a subunit of Go through a disulfide bridge. Several aspects of cross-linking correlate well

' H. Mukai and T. Higashijima, unpublished data.

12655

12656 Cross-linking of Mastoparan to Go

with regulation by mastoparan. It is likely that the site of cross-linking forms part of a specific mastoparan-binding site, which may be the receptor-binding site as well.

EXPERIMENTAL PROCEDURES

Peptides-Mastoparan and [Tyr,",Cys'l]MP were synthesized by solid-phase methods and were purified by C4 HPLC using a 20-60% acetonitrile gradient in 0.1% trifluoroacetic acid (5). Purity was >97% according to analytical C4 HPLC. Radioiodination of [Tyr",Cys"]MP was performed using chloramine T. The free sulfhydryl group was first blocked by incubation of 2 nmol of peptide with 50 nmol of 5,5'- dithiobis(nitrobenz0ic acid). Carrier-free Na? (1 mCi, 388 pmol) and 5 p1 of 0.4 mg/ml chloramine T were added (58-pl total volume), and the mixture was incubated a t 23 "C for 1 min. The reaction was stopped by adding 5 p l of 5 mg/ml sodium metabisulfate, and the nitrobenzoic acid group was removed by incubation with 1 pmol of DTT at 30 "C for 30 min. Unreacted l2'1- and nitrobenzoic acid-DTT complex were removed on a 1-ml column of Amberlite IRA-410 (acetate form). Peptide was eluted with 10 ml of 10% acetic acid. The eluate was lyophilized, and monoiodinated peptide was purified by C, HPLC (5-mm inner diameter, 25 cm) using a linear gradient of acetonitrile (0% for solvent A, 95% for solvent B) in 0.1% trifluoroacetic acid. The original monoiodinated, and diiodinated peptides were eluted at 45.4, 47.6, and 48.9% solvent B, respectively. Nonra- dioactive [I-Tyr:',Cys"]MP was prepared as was ["sI-Tyr",Cys'L]MP, except that the amount of iodine was increased to 2 nmol. Purified peptides were spectroscopically identified, and the concentrations of the peptides were determined according to the absorbance of tyrosine (e,,,,,, = 1400 M" cm-' a t 275 nm), monoiodotyrosine (E,,, = 2750 M" cm" at 283 nm), and diiodotyrosine ( e , , , = 2750 M" cm" at 287 nm) (14).

Assays-The regulatory activity of mastoparan analogs was routinely determined by enhancement of the steady-state GTPase activity of G proteins (15). The medium contained 50 mM NaHepes (pH 8.0), 1 mM EDTA, 1.1 mM MgSO, plus [~u-"~P]GTP and other addi- tives at the concentrations shown. G proteins were assayed either in detergent solution or after reconstitution into vesicles of phosphati- dylethanolamine, phosphatidylcholine, and phosphatidylethanola- mine as described previously (5). The molar turnover number was calculated according to the total amount of G protein, which was determined by ['''SIGTPyS binding as described previously (15, 16).

Purification of G Proteins-Bovine brain a, and P-y subunits were purified separately as reported previously (5). Recombinant a , was expressed in Escherichia coli and purified according to the protocol of Linder et al. (17). For cross-linking experiments, 0.3 mol eq of subunits from bovine brain were added to the a subunit.

The amino terminus of a0 (100 p g ) was proteolyzed after activation by 20 p~ AlCI:,, 6 mM MgCI2, 10 mM NaF (AMF) at 20 "C for 10 min in 20 mM NaHepes, 1 mM EDTA, 1 mM DTT, 0.1% Lubrol 12A9. Then, 10 pg of L-1-tosylamido-2-phenylethyl chloromethyl ketone- treated trypsin was added and incubated a t 30 "C for 1 h. The 39- kDa electrophoretic band of cyo disappeared during proteolysis, and a new 37-kDa band appeared. This band corresponds to a. that lacks its amino-terminal fragment (18-20). The 37-kDa fragment was purified by chromatography on DEAE-Sephacel in the same buffer, but which contained 0.7% Lubrol (5, 21). AMF was removed, and the Lubrol concentration was reduced to 0.1% by chromatography on columns of hydroxyapatite (3 ml) and Sephadex G-25 (8 ml) (5, 21).

Cross-linking-Cross-linking of [I-Tyr,",Cys"]MP (3 p ~ ) to Go (1 p ~ ) was performed in 50 mM NaHepes (pH 8.0), 1 mM EDTA, 1.3 mM MgCI, or 1.1 mM MgSO,, 0.1 mM GTP, 0.1% Lubrol 12A9 for 15 min at 30 "C unless mentioned otherwise. The concentrations of a and B-y subunits and other additions are described in the text. Cross- linking was quenched with 1 or 5 mM NEM. Samples were analyzed by SDS-polyacrylamide gel electrophoresis using either uniform polyacrylamide gels (22) or gradient gels at increased ionic strength (23).

CNBr Hydrolysis-After cross-linking [12sII-Tyr',Cys1']MP to Go, the adduct was incubated under argon in 70% formic acid with 100 mg/ml CNBr and 1 mg/ml tryptophan for 16 h a t room temperature in the dark. After the incubation, an equal volume of water was added, and the solution was lyophilized. Lyophilization was repeated twice to remove residual formic acid.

CNBr fragments were purified by HPLC on a 0.41 X 10-cm SynChrom C, column or a 0.41 X 25-cm Vydac C, column using a 19-57% acetonitrile gradient in 0.1% trifluoroacetic acid a t a 1 ml/ min flow rate over 60 min. Fractions (1 ml) were collected, and 10-pl aliquots were analyzed by gradient gel electrophoresis (23) after

0 5 I I

I

0 0 0 1 1 10

D l 1 (mM)

0 4 , 1

0 0 I 0 1 1 10 100

[C.Y]MP +MI

FIG. 1. St imulat ion of steady-state GTPase act ivi ty of G, b y mastoparan and [Tyr3,Cys"]MP. A , GTPase activity of Go (30 nM a,, 90 nM Py subunits) was assayed in the presence of 5 mM DTT and the concentrations shown of either mastoparan (0, V) or [Tyr',Cys"]MP (0, V). Assays contained Lubrol PX a t either 20 ppm (V, V) or 0.1% (0,O). B, effect of DTT following preincubation with mastoparan and [Tyr',Cys"]MP. Go (300 nM cy,, 900 nM Py subunits) was incubated a t 30 "C for 15 min with 30 p~ mastoparan (e, 01, with 30 p~ [Tyr',Cys"]MP (0, 0), or without peptide (A, A) in 50 mM NaHepes (pH 8), 1 mM EDTA, 1.1 p~ GTP. Go was then diluted 5-fold into GTPase assay medium that contained 2 p~ [-y-:12P]GTP with 0.1% Lubrol 12A9 (0, 0, A) or without added Lubrol (0, +, A) (5 ppm of residual Lubrol) and the concentrations of DTT (0.1-10 mM) shown on the abscissa. GTP hydrolysis was measured at 20 "C for 30 min. GTPase activity is shown as a molar turnover number calculated according to the total Go determined by GTPyS binding (4). C, effect of preincubation on the stimulatory activity of [Tyr',Cys"]MP. [Tyr',Cys"]MP (1 mM) was incubated in 50 mM NaHepes (pH 8.0) and 1 mM EDTA a t 30 "C for 30 min under ambient atmosphere. The pretreated [Tyr:',Cys"]MP was then added to GTPase assays at the final concentrations shown. Assay mixtures contained 27 nM Go, 1 p~ [y-"PIGTP, and 0.1% Lubrol (a), 10 mM DTT (V), both Lubrol and DTT (0), or neither additive (V). Assays were carried out for 10 min a t 20 "C.

Cross-linking of Mastoparan to Go 12657

removal of solvent. Based on autoradiography of the gel, two to three fractions that contained the ['Y'I-Tyr:',Cys'']MP-G. adduct were pooled.

SDS-Polyacrylamide Gel Electrophoresis-For preparative pur- poses, a 1.5-mm gel was run at 15 mA overnight. Separated peptides were transferred to PVDF paper electrophoretically at 50 V for 2 h in 10 mM Hepes (pH 8) and 20% methanol. Typically, -50% of the labeled fragment was transferred to the first PVDF paper and 20% to a second paper, and 30% remained in the gel. The PVDF paper was stained with Amido Black, washed with 50% methanol several times, and dried. After the protein band at -6 kDa was confirmed to have radioactivity by autoradiography, the hand on the PVDF paper was cut for sequencing.

Amino Acid Sequence-Samples on PVDF paper were analyzed by automated amino acid sequencing with an Applied Biosystems Model 470A gas-phase sequenator equipped with a Model 120A on-line PTH- derivative analyzer, as reported previously (24).

Miscellaneous-Protein was assayed by the method of Schaffner and Weissmann (25) with bovine serum albumin as the standard. Chemicals were obtained from sources reported previously (5).

RESULTS

Regulation of G Proteins by [Tyr~,Cys'']MP-["sI- Tyr",Cys"]MP was synthesized for use as a covalent probe of the mastoparan-binding site on G proteins. As expected from the structure-activity relationships of other mastoparan analogs (5), the replacement of Lys" by Cys and of Leu' by Tyr did not substantially reduce the ability of mastoparan to promote GDP/GTP exchange by G, (or Gi (data not shown)) when assays were performed in the presence of DTT (Fig. 1A). Stimulation of nucleotide exchange was measured according to enhancement of the steady-state GTPase activity of Go, which is a good measure of GDP/GTP exchange at low exchange rates where hydrolysis is not rate-limiting (15, 26).

(A)

e -

8 -

1 2 3 4 5 6 7 8

(B)

MW 9 10 11

FIG. 2. Cross-linking of ['*"I-Tyr',Cys"]MP to Go. A, Go (0.25 nmol of bovine brain CY", 0.75 nmol of By subunits in 50 pl) was incubated with 30 nM [""I-Tyr',Cys"]MP a t 30 "C in 50 mM NaHepes (pH 8.0), 1 mM EDTA, 1.3 mM MgCI2, 0.1% Lubrol, 60 p~ GTP for 1 min (lanes 1 and 2), 3 min (lane 3 ) , 10 min (lanes 6-8). 30 min (lane 4 ) , or 16 h (lane 5). The samples in lanes 7 and 8 also contained 1 and 100 p~ mastoparan, respectively. After the incubation, 50 pl of electrophoresis sample buffer without reducing agent but with 1 mM NEM was added, and 40-pl aliquots were electrophoresed on a 9% polyacrylamide gel. Lanes 2-8 are an autoradiograph of the gel (overnight exposure). Lane 1 shows Coomassie Brilliant Blue staining of 111e gel thai curresponds LO lune 2. B, Go (23 pmol of (yo, 8 pmol of py subunits in 12 pl) was incubated with 5 p~ ['Y51-Tyr:',Cys'']MP (radiochemical specific activity diluted 1:167) for 20 min at 30 "C as described above. The sample was analyzed by gel electrophoresis followed by silver staining (lane 10) and autoradiography (lane 11 ). Lane 9 shows a control sample of Go that was electrophoresed on the same gel and silver-stained.

The assays shown in Fig. 1A used G, that had not been reconstituted into phospholipid vesicles and were performed in medium that contained either 0.1% Lubrol 12A9 or 20 ppm of residual Lubrol. Consequently, maximum stimulation of GTPase activity was only -7-fold, less than the 20-fold maximum that is obtained with reconstituted G proteins (see Ref. 5 for comparison).

In the absence of DTT, [Tyr:',Cys"]MP evoked complex nonlinear kinetic behavior that we did not attempt to analyze. However, if Go was preincubated with [Tyr",Cys"]MP in the absence of DTT, GTPase activity measured in a subsequent assay was linear with time. Preincubation with [T~r:~,Cys"] M P had no effect on GTPase activity when the assay was performed in the absence of both added thiol and excess Lubrol (Fig. 1B). The residual 6 PM [Tyr:',Cys"]MP did not cause any stimulation, as might have been expected from the data shown in Fig. 1A. However, assay in the presence of 10 mM DTT, but without Lubrol, revealed a 4-fold stimulation by the prior exposure to [Tyr",Cys"]MP. When GTPase was assayed in the presence of 0.1% Lubrol, the effect of DTT on the activity of [Tyr'',Cys'']MP-pretreated G, was reversed. Activity was stimulated %fold by pretreatment when assayed in the absence of DTT but with Lubrol, which is more than would be predicted from the data of Fig. 1A. There was no effect of pretreatment (or of carried over [Tyr",Cys"]MP) when the assay medium contained both Lubrol and 10 mM DTT. For comparison, preincubation with 30 PM mastoparan caused slight stimulation that was consistent with the carry- over of 6 PM mastoparan into the assay volume. The GTPase activity of Go after preincubation with mastoparan or without peptide was not altered by DTT.

The interacting effects of [Tyr',Cys"]MP and DTT on Go did not result from the covalent dimerization of [Tyr:',Cys"] M P during the assay. Incubation of [Tyr:',Cys"]MP for 30 min a t 30 "C in mock assay medium did not alter its maximal effect or potency in a short (10 min) GTPase assay at 20 "C, and activity was not altered by the inclusion of DTT in the assay (Fig. IC). Fig. 1C also shows that Lubrol decreased the potency of [Tyr:',Cys"]MP with or without DTT, as we reported previously for mastoparan (5).

The effects of preincubation with [Tyr',Cys"]MP on the GTPase activity of Go can perhaps be rationalized by suggesting that [Tyr',Cys"]MP and Go form a covalent adduct that can be cleaved by DTT and that assumes an active conformation in the presence of excess Lubrol. Regardless of rationalization, the data of Fig. 1 indicate that [Tyr:',Cys''] M P stimulates nucleotide exchange by Go and that at least some of its effects are prolonged in the absence of DTT.

Cross-linking of r"I-Tyr',Cys"]MP to Go-The complex interactions between [Tyr",Cys"]MP and DTT shown in Fig. 1 suggested that [Tyr:',Cys'']MP may form a disulfide cross- link with G,. To detect covalent cross-linking, G, and [1251- Tyr',Cys"]MP were incubated under ambient atmosphere at 30 "C in GTPase assay medium without :'lP and in the absence of DTT. Samples of the reaction mixture were analyzed by SDS gel electrophoresis after the addition of 5 mM NEM to inhibit subsequent thiol-disulfide exchange. Autoradiography of the gel showed that the (Y subunit of Go was selectively labeled by ["'I-Tyr'',Cys"]MP (Fig. 2). Labeling of a0 was time-dependent (Fig. 2 A , lanes 2-6) and was largely complete by 30 mln. If NEM was nul added befure dena~ura~lun, a fa1111 band appeared that corresponded to minor labeling of the p subunit (data not shown). Preincubation of Go with 5 mM NEM blocked the cross-linking of [I-Tyr",Cys"]MP to the (Y

subunit (data not shown). Covalent labeling of a" was inhibited competitively by 100


1 2 3 4 5

(B) a a (C) e a - a - 8

6 7 8 9

FIG. 3. Cross-linking of [IzSI-Tyr3,Cys"]MP to Go. A, effects of By subunits and guanine nucleotides. CY, (25 pmol in 25 pl) was incubated with ["'I-Tyr:',Cys"]MP for 15 min under the conditions described for Fig. 2. Each sample contained different amounts of By subunits (lane I , none; lane 2, 8 pmol; lanes 3 and 5, 25 pmol; lane 4 , 50 pmol). The sample shown in lane 5 contained 60 p~ GDP instead of GTP. B, effect of GTPrS. G, (38 pmol of CY", 80 pmol of By subunits in 50 p l ) was incubated without ["iI-Tyr2,Cys"]MP for 10 min under the same conditions as described for A, except that the medium contained 1.1 mM MgCI, and either 10 p~ GTP (lane 6 ) or 10 p~ GTPyS (lane 7). ["'I-Tyr:',Cys"]MP (30 nM) was then added, and the incubation was continued for 30 min. Samples were treated as described for A. Shown is an autoradiograph of the gel. C, effect of ADP-ribosylation. Go was ADP-ribosylated with pertussis toxin as described previously (4) (lane 9). A control sample (lane 8 ) was incubated with toxin, but without NAD. Samples were then reconstituted into phosphatidylcholine vesicles. Go and ADP-ribosylated Go (40 pmol in 25 pi) were incubated with ["'I-Tyr:',Cys'']MP for 30 min under the conditions described for A. The cross-linking reactions shown in A-C were quenched with 5 mM NEM as described for Fig. 2; 40-pl samples were electrophoresed on 12% gels; and the autoradiograph was developed after 9 h of exposure at -80 "C.

-43

-29 - 18.4 - 14.3 - 63 L I 1I

"_

1 2 3 4 MW

FIG. 4. CNBr hydrolysis of [1251-Tyr3,Cy~11]MP cross- linked to ao. CY" (100 pmol) either without (lanes I and 2 ) or with (lanes 3 and 4 ) 33 pmol of By subunits was incubated with 3 ptM ['"'I- Tyr',Cys'']MP (0.25 pCi) in a total volume of 50 p1 for 15 min under the conditions described for Fig. 2 (except 1.1 mM MgCIZ). The samples shown in lanes 1 and 3 (20 pl) were quenched with sample buffer. The samples in lanes 2 and 4 (20 pl) were hydrolyzed by CNBr as described under "Experimental Procedures." All samples were analyzed on gradient gels. The autoradiograph was exposed for 1 day.

pM mastoparan, but was not significantly inhibited by 1 p~ mastoparan (Fig. M). [Tyr',Cys"]MP also inhibited labeling at similar concentrations. This concentration dependence is consistent with the relatively high ECS0 displayed by mastoparan or [Tyr:',Cys"]MP in 0.1% Lubrol (Fig. L4; see also Ref. 5). Labeling was also inhibited by inclusion of 5 mM DTT in the reaction mixture and was reversed by subsequent addition of 5 mM DTT or 1% 2-mercaptoethanol (data not shown). Although B-y subunits of G, were not labeled by [I2'I- Tyr',,Cys'']MP, /Py subunits enhanced the labeling of the a subunit by 3-5-fuld (Fig. 3A) . Enhancement of labeling by subunits saturated a t -0.3 mol of B-y subunits/mol of a subunit. These data suggest that B-y subunits act catalytically.

The efficiency of labeling was dependent on the identity of the nucleotide bound to G,. Routinely, GTP was added to labeling reactions to stabilize G, against denaturation. Similar

labeling was observed with GTP or GDP (Fig. 3A) , however, and labeling was also observed in the absence of added nucleotide (data not shown). (Purified G, retains nearly stoichi- ometric bound GDP (27).) In contrast to GTP, the nonhydro- lyzable GTP analog GTP-yS inhibited labeling by 60-80% (Fig. 3 B ) . Such inhibition is consistent with the observation that GTPyS-bound a. dissociates from B-y subunits (26, 28). Although most cross-linking experiments were performed in 0.1% Lubrol, labeling of Go with ['*'I-Tyr:',Cys'']MP appeared in a few experiments to be slightly more efficient in 50 ppm Lubrol (data not shown). G, was also efficiently labeled after reconstitution into phosphatidylcholine vesicles (Fig. 3C) or vesicles composed of mixed phospholipids (data not shown). ADP-ribosylation of a,, by pertussis toxin inhibited labeling by ['*sI-Tyr:',Cys'']MP slightly, if a t all (Fig. 3C).

Calculations based on the radiochemical specific activity of [1251-Tyr:',Cys'']MP indicated that 0.5-0.8 mol of peptide was cross-linked per mol of a, under optimum conditions. Because cross-linking decreases the electrophoretic mobility of a,, stained gels could be used to estimate independently the fraction of a, that was cross-linked (Fig. 2B). Based on silver staining, -50% of a. was cross-linked under these conditions. Only two silver-stained bands of a0 were detected after cross- linking, the upper one of which corresponded to the band on the autoradiograph. These data suggest that ['251-Tyr',Cy~''] M P cross-links to only one site on a,.

Peptide Mapping of r'"I-Tyi',Cys'']MP Labeling Site on a,-To estimate the number of ['2sI-Tyr3,Cys'']MP labeling sites on a. in more detail, cross-linked a. was digested with CNBr, and the hydrolysate was analyzed by SDS gel electrophoresis. Autoradiography of the gel showed a single major labeled peptide of -6 kDa above the smear of radioactivity from the [""I-Tyr3,Cys1']MP-NEM adduct and/or the [1251- Tyr:',Cys"]MP dimer (Fig. 4). Labeling of a, in the presence of 0.3 mol eq of By subunits increased the intensity of the 6- kDa band, in agreement with the enhanced labeling of intact a,. Two faintly labeled bands can also be seen above the major 6-kDa fragment in Fig. 4. When lower concentrations of CNBr were used, the band at 6 kDa was fainter, and these bands were darker, suggesting that the bands at higher molecular mass are due to incomplete hydrolysis of a, by CNBr. In several experiments, recovery of '"1 in the 6-kDa fragment ranged from 50 to 80% of that originally present in the intact ['2nI-Tyr3,Cys'1]MP-a, adduct. Thus, whereas smaller, less

(A: .29 - 18.4 - 14.3 - 6.3

-29 - 18.4 I14.3 - 6.3

Fr 27 4 5 M w

FIG. 5. HPLC separation of CNBr fragments of adduct of [12SI-Tyr3,Cys'1]MP and bovine brain ao. ['YiI-Tyr',Cys"]MP was cross-linked to bovine brain Go. The reaction mixture was hydrolyzed with CNBr, and the hydrolysate was analyzed by reverse-phase HPLC (SynChrom C4 column, 5-mm diameter, 10 cm long, 1 ml/min now rate). Solvents were U.l% trllluoroacetlc acld ( A ) and 95% acetonitrile plus 0.1% trifluoroacetic acid ( B ) . The elution profile was: 5 min, 20% solvent B; 40 min, linear gradient to 60% solvent €3; and 10 min, linear gradient to 100% solvent B. Ten pl of each fraction was lyophilized, resuspended in sample buffer, and electrophoresed on a polyacrylamide gradient gel. Protein was stained with silver ( A ) , and ""I was visualized by autoradiography (16 h a t -80 "C) ( R ) .


efficiently labeled fragments of a, may be hidden in the smear at the bottom of the gel, it is likely that the 6-kDa fragment is the only labeled fragment of Go and that, consequently, only one molecule of [1231-TyrR,Cy~II]MP was cross-linked per molecule of ao.

Identification of r’sI-Tyr’,Cys’lJMP Labeling Site-To identify the site of [I-Tyr’,Cys”]MP labeling on a”, the [Iz5I- Tyr”,Cys”]MP-labeled CNBr fragment was purified in larger quantity. Bovine brain a, (8 nmol) and bovine brain @y subunits (2.5 nmol) were incubated with 5 PM [’251-Tyr3,Cys”] MP; the mixture was hydrolyzed by CNBr as described above; and the hydrolysate was fractionated by reverse-phase HPLC. HPLC fractions were analyzed by SDS gel electrophoresis. Fig. 5 shows both protein staining and autoradiography of the HPLC fractions that contained peptides (fractions 27-45). The 6-kDa fragment labeled with [1251-Tyr:’,Cys’1]MP was eluted at fractions 39-41 (5446% solvent B). Because these three fractions contained several peptides of different molecular masses, they were pooled, and the peptides were separated by electrophoresis on a 1.5-mm-thick gradient gel. After elec- trotransfer to PVDF paper, the 6-kDa labeled band was cut out and subjected to automated amino acid analysis. The amino acid sequence corresponding to [I-Tyr:’,CysIL]MP was detected with reasonable yield according to protein staining or radioactivity. However, no other amino acid residues were detected. Three separate experiments gave the same result. This suggested that the amino-terminal residue of the fragment of cyo to which [“‘1-Tyr3,Cys”]MP was cross-linked may be blocked.

The amino terminus of mature bovine brain a, (Gly2) is blocked by N-myristoylation (29). Because recombinant a subunits produced in E. coli have free amino termini (17), the same procedures of [‘251-TyrD,Cys”]MP cross-linking, hydrolysis, and purification of the cross-linked fragment were performed using recombinant a. mixed with bovine brain @? subunits. Fig. 6 shows protein staining and autoradiography of the SDS-polyacrylamide gels of the relevant C4 HPLC fractions, 27-45, from one such experiment. The 6-kDa labeled fragment of recombinant a, eluted a t a lower concentration of acetonitrile (46-48% solvent B, fractions 31-33) than did the fragment of a, from bovine brain, which indicates that i t is less hydrophobic. Fractions 31-33 were pooled, fractionated on a preparative gradient gel, and transferred to PVDF paper. The 6-kDa labeled fragment was subjected to amino acid sequencing. The residues detected in the sequenator, shown in Table I, corresponded reasonably well to those predicted for an equimolar mixture of [I-Tyr’,Cys”]MP and the amino-terminal CNBr fragment of a, beginning a t Gly’. Two independent preparations and analyses gave essentially identical results. The amino-terminal CNBr fragment is expected to have a molecular mass of 5.46 kDa, which is con-

(A) - 43

.29 m 18.4

14.3 ‘ 6.3

-43

-29

-18.4 -14.3 - 0.J

Fr 27 45MW

FIG. 6. HPLC separation of CNBr fragments of adduct of [’“‘1-Tyr3,Cys”]MP and recombinant a,. [”511-Tyr3,Cys”]MP was cross-linked to recombinant cy, in the presence of bovine brain [fr subunits. The product was analyzed as described for Fig. 5.

TABLE I Amino acid sequence analysis of the ~‘sI-Tyr’.Cys’’]MP-labeled

CNRr fragment of cyo

Fractions 31-33 of Fig. 6 were pooled, fractionated on a preparative gradient gel, and transferred to PVDF paper. Based on the autoradiograph of the PVDF paper, the area that contained the labeled fragment was cut and used for sequencing. PTH-derivatives were identified according to the elution time of standard PTH-derivatives. Predicted amino acids of [I-Tyr:’,Cys”]MP and the amino-terminal CNBr fragment of cy, are also shown. In cycle 3, a residue was found that eluted at the position of PTH-I-Tyr, which coincides with that of PTH-Tm.

Cycle Detected Predicted amino acid amino acid” [I-TyS,Cys”]MP IT,

1 Gly (531, Ile (16), Ser (9) Ile G ~ Y 2 Gly (12), Asn (6) Asn CYS 3 Thr (6), Trp/I-Tyr I-Tyr Thr 4 Leu (18), Lys (24) LY s Leu 5 Ala (29), Ser (5) Ala Ser 6 Ala (23), Leu (19) Leu Ala 7 Ala (25), Glu (8) Ala Glu 8 Ala (21), Glu (14) Ala Glu 9 Leu (8). Arg (trace) Leu Arg

10 Ala (21) Ala Ala 11 Ala (23) CYS Ala 12 Leu (16), Lys (trace) LY s Leu 13 Glu (5), Ile (2) Ile Glu 14 Arg (5), Leu (4) Leu Ark? 15 Ser (2) Ser 16 Lys (2) LY s 17 Ala (3) Ala 18 Ile (3) Ile Values in parentheses represent picomoles.

TABLE I1 Effect of mastoparan and the 0-y subunits on the GTPase activity of

intact and amino-terminally proteolyzed cy. The GTPase activity of either intact CY” (0.9 nM) or the -37-kDa

tryptic fragment of a. (0.6 nM) was assayed with or without 0-y subunits (0.9 nM) in the presence or absence of 100 p~ MP. Activities, which were linear for 20 min a t 20 “C, are expressed as molar turnover numbers.

GTPase Activity

Control Proteolyzed

-Br +Br -Br +Br rnin-l

No peptide 0.21 0.11 0.28 0.27 MastoDaran 0.42 0.22 0.29 0.28

-43

29 18.4

14.3

6.3

1 2 3 4 5 6 7 8 M W

FIG. 7. Cross-linking of [1251-Tyr3,Cys11]MP to G protein a subunits. [‘2sI-Tyr:i,Cys”]MP was reacted with G protein cy subunits in 20 p1 of the medium described for Fig. 2 that also contained 0.5 pg of bovine brain subunits. Reaction was for 20 min a t 30 “C. The cy subunits were 2 pg of bovine brain a. (lane 1 ), 1 pg of recombinant

(lane 2), 1 pg of recombinant c q 2 (lane 3), 0.5 pg of recombinant ai.? (lane 4). 0.3 pg or recornblnanL a, long rurrn (lurtr 3), 0.3 pg uf recombinant CY, short form (lane 6), 1 pg of the PT mutant of cy, (32) (lane 7), and 2 pg of recombinant a. (lane 8). The reaction was stopped by the addition of 5 mM NEM, and the mixture was hydrolyzed with 100 mg/ml CNBr for 1 day at room temperature. After lyophilization, samples were analyzed by electrophoresis on gradient gels. The autoradiograph was exposed for 1 day a t -80 “C.


sistent with the electrophoretic mobility of the cross-linked fragment. It contains only 1 Cys residue, the second residue from the amino-terminal Gly.

M P Does Not Regulate Go That Has Been Proteolyzed at Amino Terminus of a Subunit-Trypsin selectively proteo- lyzes the amino termini of activated a subunits (18, 30). To determine the importance of the amino terminus in respond- ing to mastoparan, a. was activated with AMF, treated with trypsin, and then deactivated. Formation of the 37-kDa product was monitored by gel electrophoresis (data not shown). As shown in Table 11, the GTPase activity of amino-terminally proteolyzed a. was insensitive to both stimulation by mastoparan and inhibition by P-y subunits. No response to mastoparan was observed at concentrations up to 100 PM in either the presence or absence of P-y subunits. In parallel experiments, intact a. displayed both these expected re- sponses, although stimulation by mastoparan was only -2- fold because the assay was performed in Lubrol solution. The insensitivity of proteolyzed a subunits to regulation by P-y subunits has been reported previously (20,31). The unrespon- siveness of proteolyzed a. to mastoparan is consistent with the importance of the amino terminus for regulation by mastoparan.

Cross-linking of p'51-Tyr3,Cys11]MP to Other G Proteins- To test the generality of the results obtained with Go, we incubated [1zsI-Tyr3,Cys"]MP with the three forms of Gi, the long and short forms of G,, and an a. mutant that contains a pertussis toxin substrate site near its carboxyl terminus (32). Go was included as a positive control. Each G protein was selectively labeled in its a subunit (data not shown). CNBr hydrolysis of the adducts followed by electrophoresis of the hydrolysis products on a gradient gel showed that the band with covalently bound [1z51-Tyr3,Cys"]MP appeared at about the same position for each G protein: -6 kDa for a, and the ai proteins and -8 kDa for both forms of a. (Fig. 7). Some larger fragments were also observed. These were probably products of incomplete hydrolysis (see legend to Fig. 7). These data are consistent with the covalent cross-linking of ['"I- Tyr',Cys"]MP to Cys3 in each of the G protein a subunits. The Cys3 residue is conserved among all the a subunits tested, as is the Met residue (Met53 in ao) that is cleaved to produce the -6-kDa CNBr fragment (see Ref. 1 for sequences). There is a 7-amino acid insertion near the amino terminus of aR, which presumably accounts for the greater apparent size of its labeled fragment. The three ai proteins also contain a Met at position 18, which should have produced a fragment too small to be resolved from the smear of [1251-Tyr3,Cys11]MP in the electrophoresis system used here. We assume that the -6- kDa fragment of the [1251-Tyr3,Cys"]MP-ai adduct shown in Fig. 7 is the result of incomplete hydrolysis.

DISCUSSION

The activation of G proteins by mastoparan is mechanisti- cally similar to activation by receptors (4, 5 ) . The structure of membrane-bound mastoparan, an a helix with four positive charges directed toward the aqueous medium (6), is also reminiscent of cationic regions on the cytoplasmic surface of G protein-coupled receptors that are thought to interact with G proteins (33). Mastoparan may therefore be useful as a low molecular mass model of receptors for studying receptor-(= protein interactions. However, cationic amphipathic peptides exert many nonspecific effects, some of which may reflect merely disordering of cell membranes. The utility of mastoparan as a model for receptors would therefore be supported by the definition of a specific mastoparan-binding site on G proteins.

The data shown here indicate that [I-Tyr3,Cys"]MP spon- taneously cross-links to a. through a disulfide bridge between Cys" of [I-Tyr3,Cys"]MP and Cys3 of ao. Mastoparan itself, which has no Cys residue, blocked formation of the cross- linked product when present at a concentration adequate to catalyze activation of Go. G protein P-y subunits catalytically enhanced the cross-linking of [I-Tyr3,Cys"]MP to a. without themselves undergoing cross-linking. Activation of Go with GTP-yS inhibited cross-linking, which is consistent with the tendency of activated a subunits to dissociate from P-y subunits. Cross-linking was selective for a, even in preparations that were <lo% pure (data not shown), which indicates that [1251-Tyr3,Cys"]MP does not act nonspecifically as a sulfhydryl reagent. These observations all support the specificity of cross-linking.

We could identify the [1251-Tyr3,Cys"]MP labeling site on a, as Cys3 because this is the only Cys residue in the amino- terminal CNBr fragment of a,,. The identity of the fragment was determined by amino acid sequence analysis of the CNBr fragment of the adduct of [1z51-Tyr3,Cys"]MP and recombinant ao. The two PTH-derivatives released at each sequenator cycle were those predicted from the sequences of [I- Tyr3,Cys"]MP and ao, and their yields were similar, suggesting a 1:l stoichiometry. Other residues were not detected. Identification was supported by the fact that no sequence other than that of [I-Tyr3,Cys"]MP was obtained from the same-sized CNBr fragment of the adduct of ['z51-Tyr3,Cys11] MP and a, from bovine brain, whose amino terminus is blocked by myristoylation of Glyz (29). The labeled CNBr fragment from cross-linked bovine brain a, was also eluted from the C4 HPLC column at a higher concentration of acetonitrile than was the fragment from the recombinant ao, consistent with greater hydrophobicity from the myristoyl blocking group. Other silver-stained fragments of recombinant and natural a, were eluted at about the same positions.

These data suggest, but do not prove, that [1z51-Tyr3,Cys"] MP cross-linked to a single site on ao. First, calculations based on the radiochemical specific activity of [1251-Tyr3,Cys"] MP indicated that 0.5-0.8 mol of [1251-Tyr3,Cys"]MP could be covalently bound per molecule of a,. In addition, cross- linking altered the electrophoretic mobility of a, in SDS to produce a single new band of increased size that corresponded to the single band on the autoradiographs. Last, we identified only a single labeled peptide after CNBr hydrolysis of the [1251-Tyr3,Cys11]MP-ao adduct, and this fragment accounted for 50-80% of the radioactivity that was incorporated into ao. In our experience, such recovery after CNBr digestion is quite good. Other sites of labeling, if they exist, are therefore minor. Of the four other predicted cysteine-containing CNBr fragments, two are larger than the major labeled fragment and, if labeled, would have been noticed on the gradient gels. The two others, of M, 2390 and 4690, may have been missed at the bottom of the gels if they were only weakly labeled and were not separated from the mass of [1251-Tyr3,Cys"]MP by HPLC.

Mastoparan mimics the function of G protein-coupled receptors and would be expected to bind to an a subunit at its receptor-binding site. However, previous studies suggested that the amino-terminal region of G protein a subunits is primarily important for interaction with P-y subunits. Prote- olytic cleavage of the amino terminus of the a subunit inhibits interaction with P-y subunits (Refs. 20 and 31; also Table 111, and amino-terminal myristoylation of a i and a. enhances interaction with P-y subunits (34).

Receptors have been assumed to bind to a site on G proteins near the carboxyl terminus of the a subunit (1,2). This is the


site of pertussis toxin-catalyzed ADP-ribosylation of the ai proteins and a. (35, 36) and the site of the unc mutation in an (37), both of which block regulation by receptors. In addition, peptides based on the carboxyl-terminal sequence of at stabilize the meta-I1 conformation of rhodopsin (38), as does G, itself; and antibodies to carboxyl-terminal epitopes of at inhibit rhodopsin-G, interaction (39). Mastoparan action is also linked to carboxyl-terminal structures. ADP-ribosylation by pertussis toxin blocks regulation by mastoparan (4, 40); adenylylcyclase in membranes of the unc mutant of S49 lymphoma cells (37, 41) is insensitive to G,-activating peptides3; and an antibody against the carboxyl terminus of ai,* blocked regulation of Gi,2 by mastoparan (40) (confirmed for Go in this laboratory).

Although our data argue for the importance of the amino terminus of the a subunit for regulation by mastoparan, they do not exclude the possibility that mastoparan also binds to the carboxyl-terminal region of a subunits. With the present approach, [1251-Tyr3,Cys11]MP bound near the carboxyl terminus would not have been cross-linked to a, unless there were a Cys residue that was correctly oriented relative to Cys" of the bound peptide. Taken together, the data presented here and the data cited above suggest that both amino- and carboxyl-terminal regions of G protein a subunits are essential for regulation by mastoparan.

There are at least two ways to interpret the present data. First, amino and carboxyl termini of an a subunit may be juxtaposed within the three-dimensional structure of the protein. Recent chemical cross-linking studies of G, suggest that the amino and carboxyl termini may lie close together (42). If so, one molecule of mastoparan may interact directly with both the amino- and carboxyl-terminal regions, with cross- linking occurring only at one cysteine residue. Alternatively, more than one molecule of mastoparan may bind to a single a subunit at sites near the amino and carboxyl termini, but with only the amino-terminal site providing a cysteine capable of cross-linking to [1251-Tyr3,Cys11]MP. Indeed, the high Hill coefficient (n = 2-4) of the mastoparan dose-response curve (5) suggests that multiple molecules of mastoparan must bind to promote guanine nucleotide exchange. In addition, recent site-directed mutagenesis studies of G protein-coupled receptors also suggest that multiple cationic domains on the receptors' second and third cytoplasmic loops are required for specific interaction with G proteins (7 ) . The observation that ADP-ribosylation of a, by pertussis toxin inhibited activation by mastoparan (4, 40) but did not significantly inhibit the cross-linking of [12sI-Tyr3,Cys"]MP to a. is also consistent with mastoparan's interaction with both regions, although ADP-ribosylation may inhibit coupling without physically blocking the receptor-binding site.

Our data indicate that the mastoparan-binding region of G protein a subunits consists minimally of an amino-terminal sequence, but do not exclude the involvement of a carboxyl- terminal region as well. The observation that (37 subunits enhanced the cross-linking of [1251-Tyr3,Cys11]MP to a. suggests that binding sites for mastoparan and (37 subunits are closely linked, at least for function. It will now be interesting

'I T. Higashijima, unpublished data.

to use new reactive mastoparan analogs to test directly whether carboxyl-terminal portions of a subunits actually contribute to the binding site for mastoparan or receptor.

Acknowledgments-We are indebted to Dr. Clive Slaughter and Carolyn Moomaw (of this institution) for peptide sequencing, Dr. Maurine Linder (of this department) for providing us with a detailed protocol for the purification of recombinant a, and for providing recombinant ai proteins, and Dr. Michael Freismuth (of this department) for recombinant wild-type a, and the PT mutant. We thank Jimmy Woodson for expert technical assistance and Cherlyn Lacy for preparation of the manuscript.

REFERENCES 1. Gilman, A. G. (1987) Annu. Reu. Biochem. 56,615-649 2. Bourne, H., Sanders, D., and McCormick, F. (1991) Nature 349,117-127 3. Strader, C. D., Sigal, I. S., and Dixon, R. A. F. (1989) FASEB J. 3 , 1825-

4. Higashlyma, T., Uzu, S., Nakajima, T., and Ross, E. M. (1988) J. Biol.

5. Higashijima, T., Burnier, J., and Ross, E. M. (1990) J. Biol. Chem. 2 6 5 ,

6. Wakamatsu, K., Higashijima, T., Fujino, M., Nakajima, T., and Miyazawa,

7. Wong, S. K.-F., Parker, E. M., and Ross, E. M. (1990) J. Biol. Chem. 265 ,

8. Kubo, T., Bujo, H., Akiba, I., Nakai, J., Mishina, M., and Numa, S. (1988)

9. Wess, J., Brann, M. R., and Bonner, T. I. (1989) FEBS Lett. 258,133-136 10. Strader, C. D., Dixon, R. A. F., Cheung, A. H., Candelore, M. R., Blake, A.

11. O'Dowd, B. F., Hnatowich, M., Regan, J. W., Leader, W. M., Caron, M. G.,

12. Cotecchia, S., Exum, S., Caron, M. G., and Lefkowitz, R. J. (1990) Proc.

13. Malencik, D. A,, and Anderson, S. R. (1983) Biochem. Biophys. Res.

14. Edelhoch, H. (1962) J. Biol. Chem. 237,2778-2787 15. Higashijima, T., Ferguson, K. M., Smigel, M. D., and Gilman, A. G. (1987)

16. Brandt, D. R., and Ross, E. M. (1985) J. Biol. Chem. 2 6 0 , 266-272 17. Linder, M. E., Ewald, D. A,, Miller, R. J., and Gilman, A. G. (1990) J. Biol.

18. Winslow, J., Van Amsterdam, J., and Neer, E. (1986) J. Biol. Chem. 261 ,

1832 ... Chem. 263,6491-6494

14176-14186

T. (1983) FEBS Lett. 162,123-126

6219-6224

FEBS Lett. 2 4 1 , 119-125

D., and Sigal, I. S. (1987) J. Biol. Chem. 262,16439-16443

and Lefltowitz, R. J. (1988) J. Biol. Chem. 263,15985-15992

Natl. Acad. Sci. U. S. A. 87,2896-2900

Commun. 114,50-56

J. Biol. Chem. 2 6 2 , 757-761

Chem. 265,8243-8251

7.571-7579

20. 19.

21.

22. 23.

24.

25. 26.

27.

28. 29.

30.

31. 32. 33. 34.

35.

36.

37.

38.

39.

40.

41.

42.

Navon, S., and Fung, B.K.-K. (1987) J. Biol. Chem. 262,15746-15751 Neer, E., Pulsifer, L., and Wolf, L. (1988) J. Biol. Chem. 263,8996-9000 Sternweis, P. C., Northup, J. K., Smigel, M. D., and Gilman, A. G. (1981)

Laemmli, U. K. (1970) Nature 227,680-685 Rubenstein, R. C., Wong, S. K.-F., and Ross, E. M. (1987) J. Biol. Chem.

Wong, S. K.-F., Slaughter, C., Ruoho, A. E., and Ross, E. M. (1988) J. Biol.

Schaffner, W., and Weissmann, C. (1973) Anal. Biochem. 56,502-514 Higashijima, T., Ferguson, K. M., Sternweis, P. C., Smigel, M. D., and

Ferguson, K. M., Higashijima, T., Smigel, M. D., and Gilman, A. G. (1986)

Buss, J., Mumby, S., Casey, P., Gilman, A,, and Sefton, B. (1987) Proc. Sternweis, P. C. (1986) J. B~ol. Chem. 261,631-637

Hurley, J. B., Simon, M. I., Teplow, D. B., Robishaw, J. D., and Gilman,

Fung, B. K.-K., and Nash, C. R. (1983) J. Biol. Chem. 258,10503-10510 Freissmutb, M., and Gilman, A. (1989) J. Biol. Chem. 264,21907-21914 Ross, E. M. (1989) Neuron 3 , 141-152 Linder, M., Pang, I.-H., Duronio, R., Gordon, J., Sternweis, P., and Gilman,

West, R. E., Jr., Moss, J., Vaughan, M., Liu, T., and Liu, T.-Y. (1985) J.

Van Dop, C., Yamanaka, G., Steinberg, F., Sekura, R., Manclark, C. R.,

Sullivan, K. A,, Miller, R. T., Masters, S. B., Beiderman, B., Heideman,

Hamm, H. E., Deretic, D., Arendt, A,, Hargrave, P. A., Koenig, B., and

Cerione, R., Kroll, S., Rajaram, R., Unson, C., Goldsmith, P., and Spiegel,

Weingarten, R., Ransnas, L., Mueller, H., Sklar, L., and Bokoch, G. (1990)

Haga, T., Ross, E. M., Anderson, H. J., and Gilman, A. G . (1977) Proc.

Hingorani, V. N., Tobias, D. T., Henderson, J. T., and Ho, Y.-K. (1988) J.

~~ ~~

J. Biol. Chem. 2 5 6 , 11517-11526

2 6 2 , 16655-16662

Chem. 2 6 3 , 7925-7928

Gilman, A. G. (1987) J. Bid. Chem. 262,762-766

J. Biol. Chem. 2 6 1 , 7393:7399

Natl. Acad. Scz. U. S. A. 8 4 , 7493-7497

A. G. (1984) Science 226,860-862

A. (1991) J. Biol. Chem. 266,4654-4659

Biol. Chem. 2 6 0 , 14428-14430

Stryer, L., and Bourne, H. R. (1984) J. Biol. Chem. 259,23-25

W., and Bourne, H. R. (1987) Nature 330,758-760

Hofmann, K. P. (1988) Science 241,832-835

A. (1988) J. Biol. Chem. 263,9345-9352

J. Biol. Chem. 2 6 5 , 11044-11049

Natl. Acad. Sci. U. S. A. 7 4 , 2016-2020

Biol. Chem. 263,6916-6926

mapping of the mastoparan-binding site on g proteins

Documents