G Protein Subunits with Altered Prenylation ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY © 1996 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 271, No. 31, Issue of August 2, pp. 18582–18587, 1996 Printed in U.S.A.

G Protein g Subunits with Altered Prenylation Sequences Are Properly Modified When Expressed in Sf9 Cells* (Received for publication, February 21, 1996, and in revised form, May 2, 1996)

Margaret A. Lindorfer‡, Nicholas E. Sherman§, Karen A. Woodfork¶, Julia E. Fletcher, Donald F. Hunt§i, and James C. Garrison From the Departments of Pharmacology, iPathology, and §Chemistry, University of Virginia, Charlottesville, Virginia 22908

The a and g subunits of heterotrimeric G proteins1 undergo post-translational modification with lipids that are essential for their interaction with receptors and downstream effectors (1). The a subunits are known to be modified with myristate and palmitate groups. The 11 known g subunit isoforms all have CaaX motifs at their carboxyl terminus (2– 4) which direct post-translational modifications including attachment of a prenyl group to the cysteine side chain, proteolytic cleavage of the aaX tripeptide, and methylation of the resulting carboxyl terminus (5). The identity of the prenyl group attached appears to be directed by the last amino acid in the CaaX motif (6 – 8). The prenylation of the cysteine side chain not only targets the bg dimer to the membrane (8 –11), it also plays a major role in determining protein-protein interactions (12). Prenylation of * This work was supported in part by National Institutes of Health Grants R01-DK-19952 (to J. C. G.) and R01-GM-37537 (to D. F. H.). ‡ To whom correspondence should be addressed: Dept. of Pharmacology, Box 448, University of Virginia Health Science Center, Charlottesville, VA 22908. Tel.: 804-924-9977; Fax: 804-982-3878. ¶ Present address: Dept. of Biology, Washington and Jefferson College, Washington, PA 15301. 1 The abbreviations are: G proteins, guanine nucleotide-binding regulatory proteins; Sf9 cells, Spondoptera frugiperda cells (ATCC CRL 1711); CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid; Tricine, N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine; CaaX, prenylation signal sequence; g1-S74L, G protein g1 subunit with amino acid residue 74 changed from serine to leucine; g2-L71S, G protein g2 subunit with amino acid residue 71 changed from leucine to serine; CAD, collision-activated dissociation.

the g subunit is necessary for the formation of an active transducin abg complex (13, 14), for stimulation of phospholipase C-b by the b1g1 dimer (15), and for translocation of the b-adrenergic receptor kinase to the plasma membrane (16). In addition, the ability of the bg subunit to support ADP-ribosylation of Gai subunits by pertussis toxin and to regulate adenyl cyclase activity both require the presence of the prenyl modification (17). Prenylated synthetic peptides corresponding to the carboxyl terminus of the g subunit can inhibit ADP-ribosylation of a subunits, but the same peptides lacking the prenyl thioether have no effect (18). Thus, the carboxyl terminus of the g subunit is an important domain for the interaction between the a subunit and the bg dimer. We have taken advantage of the ability of the baculovirus Sf9 cell system to produce mg quantities of highly purified bg subunits to investigate further this central role of prenylation in bg function. We have altered the g1 and g2 subunit cDNAs to code for alternate prenyl modifications by changing the codon for the last amino acid in the CaaX motif. We report here the expression of these constructs in the baculovirus Sf9 cell system and purification to homogeneity of the wild type and altered g subunits complexed with the b1 subunit. We have verified using mass spectrometric techniques that wild type and altered CaaX sequences result in the expected post-translational modifications in the baculovirus Sf9 cell expression system. The accompanying report describes the effect of these prenyl modifications on the ability of the heterotrimeric G protein to interact with the A1 adenosine receptor (19). EXPERIMENTAL PROCEDURES

Construction of g Prenylation Mutants—Polymerase chain reaction amplification was used to perform site-directed mutagenesis on the cDNA clones for bovine g1 (20) and g2 (21). For g1, oligonucleotide primers were used to insert a 59 XbaI site and change the last codon from TCA to TTA. This changed the amino acid sequence of the CaaX box from CVIS to CVIL (abbreviated g1-S74L). For g2, oligonucleotide primers were used to insert a 39 EcoRI site and change the last codon from CTT to TCT. This changed the CaaX box sequence from CAIL to CAIS (abbreviated g2-L71S). Polymerase chain reaction products were ligated into the TA cloning vector (Invitrogen) and from there were cut (with XbaI for g1; with EcoRI for g2) and re-ligated into the baculovirus transfer vector pVL1393. The CaaX box mutations were confirmed in pVL1393 by sequence analysis with the dideoxy method (22) using the Sequenase version 2.0 DNA Sequencing Kit (U. S. Biochemical Corp.). Recombinant baculoviruses were obtained by calcium phosphate transfection of Sf9 cells with the recombinant pVL1393 plasmids using linear wild type baculovirus DNA (Invitrogen) as described previously (23). Recombinant viruses were isolated by four rounds of plaque purification (24). Expression and Purification of Recombinant bg Complexes—The construction of recombinant baculoviruses encoding b1 and g2 was described previously (25). Recombinant baculoviruses encoding g1 and g3 were gifts from Dr. Narasimhan Gautam, Washington University School of Medicine, and Dr. Janet Robishaw, Weiss Center, Danville, PA, respectively. Sf9 cells were coinfected with the appropriate b and g

18582

Downloaded from http://www.jbc.org/ by guest on August 10, 2016

The g subunits of heterotrimeric G proteins undergo post-translational prenylation and carboxylmethylation after formation of the bg dimer, modifications that are essential for a-bg, bg-receptor, and bg-effector interactions. We have determined the specific prenyl group present on the b1g1, b1g2, and b1g3 dimers purified from baculovirus-infected Sf9 cells by specific binding to G protein a subunits immobilized on agarose. These recombinant dimers undergo the same post-translational modifications determined for g1 and g2 isolated from mammalian tissues. Furthermore, infection of Sf9 cells with a recombinant baculovirus encoding an alteration of the g1 CaaX sequence (g1-S74L) resulted in geranylgeranylation of the resulting g1 subunit, and alteration of the g2 CaaX sequence to CAIS (g2-L71S) resulted in farnesylation. Both of these altered g subunits were able to associate stably with b1, and the resulting bg dimer bound tightly to a-agarose and eluted specifically with aluminum fluoride. These results indicate that Sf9 insect cells properly process the CaaX motif in G protein g subunits and are a useful model system to study the role of prenylation in the protein-protein interactions in which the bg subunits participate.

Prenylation of the G Protein g Subunit in Sf9 Cells

FIG. 1. Sodium dodecyl sulfate-polyacrylamide electrophoresis of purified G protein bg subunits. Purified recombinant bg subunits of defined subtypes were electrophoresed on a 16% acrylamide, 10% (v/v) glycerol, sodium dodecyl sulfate-Tricine gel and stained with silver. The mobility of the molecular mass markers is indicated on the right. The positions of the b1 and the various g subunits are indicated on the left. blots were visualized with alkaline phosphatase-conjugated mouse anti-rabbit IgG. Materials—All reagents used in the preparation of bg dimers were as described previously (23). Trifluoroacetic acid was protein sequencing grade from Applied Biosystems. Genapol C-100 was from Calbiochem; CHAPS was from Boehringer Mannheim. Acetonitrile was HPLC grade from Baxter. Yeast a-factor was the gift of Prof. Jeffrey M. Becker, University of Tennessee. All other reagents were the highest purity available. RESULTS

This study had two objectives: first, to determine the stoichiometry and fidelity of prenylation of the baculovirus Sf9 cell system for wild type G protein g subunits; and second, to determine the effect of alterations in the CaaX sequence of the g subunit on subsequent prenylation. The results presented here help define g subunit prenylation in the Sf9 cell system and provide the basis upon which to investigate the role prenylation plays in the interaction of the bg dimer with receptors and with effectors. Fig. 1 shows purified recombinant bg complexes separated on a Tricine gel as described under “Experimental Procedures” and stained with silver. These bg dimers were purified by ion exchange and a-agarose affinity chromatography, and all dimers shown are essentially free of contaminating proteins. Approximately 90% of the bg dimers extracted from the Sf9 cells with Genapol were recovered from the a-agarose affinity column. In this gel system, the differences in mobility of g1, g2, and g3 are apparent. The g1 mutant, g1-S74L, which differs from g1 only by the exchange of farnesyl for geranylgeranyl, does not exhibit a detectable difference in mobility compared with g1. Similarly, the g2 mutant, g2-L71S, which is predicted to be farnesylated, has electrophoretic mobility similar to that of g2. To determine the nature of the post-translational modifications, purified bg subunits were analyzed by electrospray ionization mass spectrometry. Under the conditions employed, only signals for the g subunit were observed. Mass spectra of g1 are displayed in Fig. 2, A and B. Fig. 2A shows the profile of ions due to multiply protonated g subunits. Fig. 2B displays the deconvoluted mass spectrum containing only (M1H)1 ions for each of the molecules in the mixture. The g1 subunit, when expressed in Sf9 cells, shows an (M1H)1 ion at m/z 8,334 Da, in agreement with the molecular mass predicted from the DNA sequence, including the post-translational modifications of: removal of the amino-terminal methionine, attachment of a farnesyl group to cysteine 71, removal of the carboxyl-terminal


baculovirus at a multiplicity of infection of 3 each and harvested 48 h after infection (25). bgs were detergent extracted and purified as described previously (23). Briefly, frozen cell pellets were thawed in 15 times their weight in ice-cold homogenization buffer (10 mM Tris, pH 8.0; 25 mM NaCl; 10 mM MgCl2; 1 mM EGTA; 1 mM dithiothreitol; 0.1 mM phenylmethylsulfonyl fluoride; 20 mg/ml benzamidine; and 2 mg/ml each of aprotinin, leupeptin, and pepstatin) and burst by nitrogen cavitation on ice. All subsequent steps were performed at 4 °C. The crude lysate was mixed with an equal volume of homogenization buffer supplemented with 0.2% (w/v) Genapol C-100 (Calbiochem) and stirred for 1 h. The Genapol extract was centrifuged at 100,000 3 g for 60 min, and the supernatant was decanted and snap frozen in 15-ml aliquots. Genapol extract (60 –90 ml) was thawed and applied to a DEAE HR40 column (2 3 10 cm) equilibrated with TED/CHAPS (50 mM Tris, 0.02 mM EDTA, 1 mM dithiothreitol, 0.6% (w/v) CHAPS, pH 8.0) and eluted with a 96-ml linear gradient from 0 to 300 mM NaCl in TED/CHAPS. The fractions containing bg were identified by dot blots using an anti-b common antibody (NEI-808, DuPont NEN) (23), pooled (30 ml), and GDP added to a final concentration of 5 mM. The pool was applied at a flow rate of 0.5 ml/min to a 3-ml recombinant Gai-agarose column (26) as described (23). The bg-loaded ai-agarose column was washed with 2 3 2 ml of wash buffer (0.6% (w/v) CHAPS, 20 mM Hepes, pH 8.0, 200 mM NaCl, 1 mM EDTA, 3 mM dithiothreitol, 5 mM GDP), and 2 3 2 ml of wash buffer plus 200 mM NaCl. The bg dimers were specifically eluted with elution buffer (wash buffer plus 50 mM MgCl2, 10 mM NaF, 30 mM AlCl3). Two ml of elution buffer was applied to the top of the column and the flow stopped for 15 min. An additional 3 3 2 ml of elution buffer was applied at a rate of 0.5 ml/min. Elution fractions were concentrated with a Centricon 30 concentrator (Amicon) to a final volume of 100 –500 ml giving a final protein concentration of 50 –250 ng/ml and stored at 270 °C. The overall recovery of bg dimer in this procedure was estimated by measuring the amount of b subunit on Western blots using the anti-b common antibody. Approximately half of the bg dimer present in the crude Sf9 cell lysate is extracted with 0.1% Genapol and applied to the DEAE column. About 90% of this population of bg subunits is recovered from the DEAE column and binds specifically to the a affinity column (i.e. can be eluted with aluminum fluoride). Mass Spectroscopic Techniques—A triple quadrupole mass spectrometer (Finnigan MAT) equipped with an electrospray ionization source (Finnigan MAT) and a solvent delivery system (Applied Biosystems, Inc.) were used. Samples of purified bg dimer in elution buffer were introduced by injection onto a 75-mm 3 15-cm fused silica capillary column packed with Poros R2/H (PerSeptive Biosystems) and eluted with a 12-min binary gradient of 0 – 80% solvent B, where solvent A is 0.1 M acetic acid in water, and solvent B is acetonitrile, at a flow rate of 0.5 ml/min. To prepare the g subunit for chemical or enzymatic digestion, the g subunit was isolated from bg dimers by reverse phase chromatography on a microbore system (Applied Biosystems model 120A). About 200 ml of bg at 100 ng/ml in elution buffer was injected directly onto a Poros R2/H column (2.1 3 30 mm, PerSeptive Biosystems) and eluted at a flow rate of 1 ml/min with a 10-min binary gradient of 0 –100% solvent B where solvent A is 0.1% trifluoroacetic acid, and solvent B is 0.1% trifluoroacetic acid in 80% acetonitrile. Absorbance was monitored at 214 nm, and fractions were collected by hand. The g subunit eluted in a sharp peak at 6 min after injection. Chemical and Enzymatic Digestion—Reverse phase elution fractions containing g1 were lyophilized and resuspended in 20 ml of 50 mM ammonium acetate, pH 8.3. Trypsin (0.2 mg/tube, sequencing grade modified trypsin, Promega) was added and the digestion allowed to proceed at 37 °C for 16 h. The reaction was stopped by the addition of 1 ml of acetic acid (Pierce). The g2 subunit was isolated by reverse phase chromatography as described above. Reverse phase elution fractions containing g2 were lyophilized and redissolved in 70% trifluoroacetic acid. Excess cyanogen bromide was added and the cleavage reaction carried out at room temperature, in the dark, under argon, for 18 h. The remaining cyanogen bromide and trifluoroacetic acid were evaporated with a stream of nitrogen, and the fragments were resuspended in 1% acetic acid. The digestion products were screened by electrospray ionization mass spectrometry, as described above, and sequenced by CAD triple quadrupole mass spectrometry (27). Electrophoretic Methods—Tricine gels were run according to the procedure of Schagger and von Jagow (28). The separating gel used contained 16.5% total acrylamide, 6% bisacrylamide, and 10% (v/v) glycerol; the spacer gel 10% total acrylamide and 3% bisacrylamide; and the stacking gel 4% total acrylamide, 3% bisacrylamide. Separating gel dimensions were 0.075 3 10 3 14 cm. Gels were run approximately 900 volt h at 5–10 °C. Gels were silver stained by the method of Morrissey (29), with a 15-min incubation with 5 mg/ml dithiothreitol. Western

18583

18584


tripeptide, and methylation of the new carboxyl terminus. Fig. 2, C and D, displays the multiply charged ion spectrum and deconvoluted spectrum, respectively, for the g1-S74L subunit. This change in the CaaX sequence from CAIS to CAIL is predicted to direct prenylation with a geranylgeranyl group. The observed m/z value for the (M1H)1 ion at 8,403 Da is consistent with exchange of the farnesyl group for geranylgeranyl and clearly demonstrates the predicted change in posttranslational processing. Thus, in the context of the b1g1 dimer, this alteration of the last amino acid in the CaaX sequence is sufficient to alter prenylation. Mass spectra displayed in Fig. 2, A and C, show two sets of signals due to multiply protonated g subunits. One corresponds to the fully processed molecule, and the second set (indicated by asterisks) corresponds to the fully processed molecule minus the prenyl group. Since there is strong evidence that the prenylation reaction precedes endoproteolysis and carboxylmethylation (5, 30), and previous fast atom bombardment mass analysis of prenylated synthetic peptides (yeast a-factor and analogs) has shown the allylic thioether bond to be labile (31), we investigated the possibility that the loss of the prenyl group was occurring in the mass spectrometer as a consequence of the analysis. We first extracted ion chromatograms corresponding to the 110 ion of the g1 subunit, with and without the farnesyl group, from the mass spectrum shown in Fig. 2A. As shown in Fig. 3A, the signals corresponding to the 110 ion of the g1 subunit, with and without farnesyl, both appear in the mass spectrum recorded at the elution time of the farnesylated parent molecule. If both molecular species were present in the original sample, they would be expected to show considerably

different chromatographic retention times on the microcapillary HPLC and thus show up in different scans of the mass spectrometer. We next introduced intact b1g1 into the mass spectrometer as before, but now the 19 ion corresponding to the fully processed g1 subunit was selected for analysis in the second quadrupole under CAD conditions and the resulting mass spectrum analyzed for fragmentation corresponding to the loss of the farnesyl group. The predominant fragment in the resulting CAD spectrum corresponded to the 18 ion for the g1 without the farnesyl group (data not shown). Thus, we conclude that the farnesylated molecule fragments with loss of the farnesyl molecule in the collision cell and also during the ionization process itself. A similar analysis was also performed on a synthetic sample of yeast a-factor. This synthetic sample had been purified by HPLC, and the presence of the farnesyl and carboxyl methyl groups had been verified (31, 32). We observed two families of multiply protonated molecules in the microcapillary HPLC, electrospray ionization mass spectrum of the synthetic yeast a-factor (data not shown). These corresponded to a-factor with and without the farnesyl group. The chromatography conditions used to introduce the a-factor peptide into the electrospray source are capable of resolving peptides with the large difference in hydrophobicity predicted for a farnesylated and nonfarnesylated peptide. Fig. 3B shows that these two species coelute from the microcapillary used to introduce the sample. Therefore, based on the observation of a similar fragmentation with the a-factor and the g1 subunit, and coelution of the intact and fragmented ions, we conclude that the loss of the farnesyl group occurs during the ionization process and is not due to an


FIG. 2. Electrospray ionization mass spectra of recombinant b1g1 and b1g1-S74L combinations. For each analysis, purified, recombinant bg subunits were injected onto a reverse phase microcapillary column and the g subunit eluted into the electrospray source of a Finnigan MAT TSQ 700 mass spectrometer using a 0 – 80% acetonitrile gradient in 0.1 M acetic acid, as described under “Experimental Procedures.” Panel A, profile of ions due to multiply protonated g1 subunits. Signals marked 16 through 112 correspond to farnesylated g1, and signals marked with an asterisk (*) correspond to g1 lacking the farnesyl group. Panel B, reconstructed mass spectrum obtained by employing a deconvolution algorithm to convert families of related ion signals in panel A to single peaks representing the singly protonated molecules in the mixture. 8,334 Da, g1, fully modified; 8,129 Da, g1, fully modified, minus the farnesyl group (g1-C15). Panel C, profile of multiply protonated g1-S74L subunits. Signals marked 17 through 113 correspond to geranylgeranylated g1-S74L, and signals marked with an asterisk (*) correspond to g1-S74L lacking the geranylgeranyl group. Panel D, reconstructed mass spectrum obtained by employing a deconvolution algorithm to convert families of related ion signals in panel C to single peaks representing the singly protonated molecules in the mixture. 8,403 Da, g1-S74L, fully modified; 8,130 Da, g1-S74L, fully modified, minus the geranylgeranyl group (g2-C20).


18585

DISCUSSION

FIG. 3. Panel A, ion chromatograms of two m/z values, corresponding to the 110 fully processed g1 subunit (833– 835 Da) and the g1 subunit without a farnesyl (812– 814 Da), plotted as a function of scan number. Time for each mass scan was 1.5 s. Microcapillary HPLC was employed to introduce the sample to the mass spectrometer as described in the legend of Fig. 2 and under “Experimental Procedures.” Panel B, ion chromatograms of two m/z values corresponding to the (M12H)12-C15 ion for the yeast a-factor minus the farnesyl group, and the (M12H)12 ion for the farnesylated and methylated yeast a-factor. Time for each mass scan was 1.5 s. Experimental conditions for recording the mass spectra were as described in the legend of Fig. 2 and under “Experimental Procedures.”

inherent heterogeneity of processing on the part of the Sf9 cell. To determine the amino acid sequence of the recombinant g1 subunit, it was prepared for tryptic digestion by direct injection of the bg complex in elution buffer onto a reverse phase column. Under the conditions of the separation, the g subunit is eluted in a well defined peak, but we and others (33–36) have been unable to detect the elution of the b subunit. Presumably the acidic conditions of the separation result in not only the dissociation of the bg dimer but also in the irreversible adsorption of the denatured b subunit to the hydrophobic stationary phase. The isolated g1 subunit was digested with trypsin, and the resulting mixture of peptides was analyzed by CAD in the triple quadrupole mass spectrometer. Amino acid sequence information deduced from the tryptic fragments is summarized in Table I. To confirm these data, CAD spectra were also recorded on methyl ester and acetylated derivatives of these peptides. All g1 tryptic peptides were isolated except the peptides corresponding to residues 30 – 40 and the carboxyl-terminal peptide. There was no evidence of any additional side chain modifications. The amino-terminal methionine has been removed from the g1 subunit, and the new amino-terminal proline is not acetylated. A similar but less extensive analysis was performed on the g2 subunit. The isolated g2 subunit was cleaved with cyanogen bromide and the amino-terminal peptide sequenced by CAD (Table I). Amino acid sequence infor-

The prenyl group on the G protein g subunit plays a central role in bg function. We and others have demonstrated that the baculovirus Sf9 cell system produces homogeneous bg complexes of defined composition (17, 25, 37, 38). Several groups have demonstrated by labeling with tritiated mevalonolactone that g subunits, when coexpressed with b subunits, are prenylated (25, 39, 40). However, these studies do not give information on the stoichiometry of prenylation. We have demonstrated here, by direct mass spectral analysis, the exact isoprenoid modification and stoichiometry for three wild type and two g subunits with altered CaaX sequences. We have also obtained direct structural information for the g1 and g2 subunits expressed in Sf9 cells. The wild type g1 subunit, when coexpressed with the b1 subunit and subjected to a-agarose affinity purification, is modified as expected with the farnesyl group. Greater than 95% of the purified material was farnesylated. Likewise, more than 90% of the wild type g2 subunit was geranylgeranylated. The g3 subunit was completely geranylgeranylated (Table III). Previous mass spectral analysis of Ras proteins expressed in the Sf9 cell system (41, 42) demonstrated that the CaaX motif in the context of the Ras sequence was appropriately prenylated. Our results with the G protein bg dimers show that the Sf9 prenyltransferases also recognize the bg dimer as a substrate, with the same prenyl specificity as displayed for the Ras sequences. The g1 subunit isolated from rod outer segments has been subjected to proteolytic digestion and the resulting peptides completely sequenced by fast atom bombardment mass spectrometry (13). The amino-terminal methionine has been removed, revealing a free amino-terminal proline, the carboxylterminal cysteine has a farnesyl thioether, and the carboxyl terminus is methylated (13). The mass spectroscopic results presented here (see Fig. 4) for g1 purified from Sf9 cells coinfected with recombinant b1 and g1 baculovirus are consistent with the post-translational modifications observed for g1 isolated from rod outer segments. bg isolated from bovine brain is a complex mixture. To date, b1, b2, g2, g3, g5, and g7 have been identified (33, 34, 43, 44). g2, g3, g5, and g7 all have a CaaX motif that is predicted to direct


mation deduced from the cyanogen bromide fragment is summarized at the bottom of Table I. The amino-terminal methionine is removed, and the newly revealed amino-terminal alanine is acetylated. Electrospray ionization mass spectra were also recorded for the g2, g3, and g2-L71S subunits. Mass measurements are summarized in Table II. Observed (M1H)1 ions for the major species are all consistent with the amino acid sequence deduced from the corresponding DNA sequence, the expected prenyl modifications (farnesyl for g2-L71S and geranylgeranyl for g3, g2, and g1-S74L), endoproteolysis, carboxylmethylation, the removal of the amino-terminal methionine for g1, g2, g1-S74L, and g2-L71S, and the amino-terminal acetylation of g2 and g2-L71S. These post-translational modifications are summarized in Fig. 4. In each g subunit studied, the appropriately processed species accounts for 87–100% of the purified material (Table III). Minor amounts of alternately prenylated species were observed and identified based on the observed m/z value for the (M1H)1 ion. The identity and percentage of these species are summarized in Table III. Relative abundances of the molecular entities were estimated from the percent total ion current carried by each species in the deconvoluted mass spectrum. The amino termini were homogeneously processed as described above, with the exception of g1-S74L, which was 27% acetylated at the amino terminus (data not shown).


18586

TABLE I Mass assignments and sequences of peptides generated by tryptic digestion of g1 and cyanogen bromide cleavage of g2 Isolated g1 subunits were subjected to tryptic digestion, and isolated g2 subunits were cleaved with cyanogen bromide. The resulting mixture of peptides was introduced into the electrospray ionization source by microcapillary HPLC as described under “Experimental Procedures.” Individual peptides in the mixture were sequenced by the technique of CAD mass spectrometry on a triple quadrupole instrument. Fragments corresponding to residues 30 – 40 and 68 –70 of g1 were not identified. Tryptic peptides from g1 Observed mass

m/z

Position

Sequence

747 845 1,516 1,345

374 422 506 448 (674)

24–29 47–54 55–67 57–67

992 1,756 810 864 1,144

332 585 (439) 406 432 382

16–23 1–15 41–46 16–22 55–64

EVTLER SGEDPLVK GIPEDKNPFKELK PEDKNPFKELK (fragment) MEVDQLKK PVINIEDLTEKDKLK DYVEER MEVDQLK GIPEDKNPFK

Da

Cyanogen bromide fragment from g2 Observed mass

m/z

Position

Sequence

724

1–20

AcASNNTASIAQARKLVEGKLM

Da

2,171

TABLE III Percentages of isoprenoid species present Percentage of isoprenoid species present in a-agarose-purified bg preparations was determined from percent total ion currents carried by (M1H)1 ions in deconvoluted mass spectra recorded on the mixture of post-translationally modified g subunits. 1Farnesyl %

Calculated mass

g1 g2 g1-S74L g2-L71S g3

Observed mass

Da

Da

g1

8,333

8,334

g2

7,752

7,752

g1-S74L

8,401

8,403

g2-L71S

7,684

7,685

g3

8,295

8,296

FIG. 4. Summary of the post-translational modifications of g subunits expressed in Sf9 cells as determined by mass spectrometric analysis. The amino acid sequences and post-translational modifications for the g1 subunit and the amino-terminal g2 fragment were determined as described under “Experimental Procedures.” The post-translational modifications for the g1-S74L, g2-L71S, g2, and g3 subunits were deduced from the mass of the (M1H)1 ion as described under “Results.” Ac, amino-terminal acetyl; CH3, carboxylmethyl; farnesyl, geranylgeranyl, thioether linkage through the indicated cysteine side chain. Numbers in the far right column are the numbers of amino acid residues in the molecule.

prenylation with a geranylgeranyl moiety. Biochemical studies on purified bg subunits isolated from bovine brain have shown that the carboxyl terminus of brain g subunits is geranylgeranylated (26, 34, 35). Wilcox et al. (33) have subjected g2 isolated from bovine brain to trypsin digestion and electrospray ioniza-

a

96 9 13 95 NDa

1Geranylgeranyl %

4 91 87 5 100

ND, none detected.

tion mass spectrometry to obtain direct structural information. Their work rigorously demonstrates the sequence and posttranslational modifications of bovine brain g2: acetylation of the amino-terminal alanine, and geranylgeranylation, endoproteolysis, and carboxylmethylation of the carboxyl terminus. Our results on recombinant g2 expressed in Sf9 cells (see Fig. 4) demonstrate the same post-translational modifications observed for g2 isolated from bovine brain. Competition studies with synthetic tetrapeptides (7), in vitro measurements with mutated proteins (45), and mutational studies in various cell types (46) have led to the conclusion that in mammalian cells, for polypeptides capable of prenylation by farnesyl transferase or geranylgeranyltransferase I, isoprenoid specificity is contained within the CaaX sequence. Our results extend this observation to G protein bg dimers expressed in Sf9 cells. The change of CaaX from CVIS to CVIL in g1 is sufficient to change the prenyl modification from farnesyl to geranylgeranyl. Likewise, mutation of the g2 CaaX sequence from CAIL to CAIS is sufficient to switch geranylgeranylation to farnesylation. Studies with tetrapeptide substrates have shown that at sufficiently high concentrations of substrate, both farnesyltransferase and geranylgeranyltransferase I are able to prenylate CaaX peptides, regardless of the identity of the last amino acid (7, 47). When the yeast a-factor CaaX sequence was altered to CVIL, both geranylgeranylation and farnesylation were observed (48). Baculovirus coding for heterologous proteins under control of the polyhedrin promoter usually result in high levels of expression, but subsequent requirement for posttranslational modification sometimes yields a low percentage of


TABLE II Molecular mass of recombinant g subunits Purified bg dimers in elution buffer were introduced to the electrospray ionization source of the mass spectrometer by microcapillary HPLC. Signals due to multiply protonated molecules were deconvoluted to generate the experimentally observed (M1H)1 shown below. The calculated m/z values for (M1H)1 ions are based on the modifications shown in Fig. 4.


Acknowledgments—We thank Dr. Narasimhan Gautam for baculovirus encoding g1, Dr. Janet Robishaw for baculovirus encoding g3, and Dr. Jeffrey M. Becker for the yeast a-factor peptide. REFERENCES 1. Wedegaertner, P. B., Wilson, P. T., and Bourne, H. R. (1995) J. Biol. Chem. 270, 503–506 2. Ray, K., Kunsch, C., Bonner, L. M., and Robishaw, J. D. (1995) J. Biol. Chem. 270, 21765–21771 3. Kalyanaraman, S., Kalyanaraman, V., and Gautam, N. (1995) Biochem. Biophys. Res. Commun. 216, 126 –132 4. Morishita, R., Nakayama, H., Isobe, T., Matsuda, T., Hashimoto, Y., Okano, T., Fukada, Y., Mizuno, K., Ohno, S., Kozawa, O., Kato, K., and Asano, T. (1995) J. Biol. Chem. 270, 29469 –29475 5. Casey, P. J. (1992) J. Lipid Res. 33, 1731–1740 6. Reiss, Y., Stradley, S. J., Gierasch, L. M., Brown, M. S., and Goldstein, J. L. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 732–736

7. Moores, S. L., Schaber, M. D., Mosser, S. D., Rands, E., O’Hara, M. B., Garsky, V. M., Marshall, M. S., Pompliano, D. L., and Gibbs, J. B. (1991) J. Biol. Chem. 266, 14603–14610 8. Cox, A. D., and Der, C. J. (1992) Curr. Opin. Cell Biol. 4, 1008 –1016 9. Simonds, W. F., Butrynski, J. E., Gautam, N., Unson, C. G., and Spiegel, A. M. (1991) J. Biol. Chem. 266, 5363–5366 10. Muntz, K. H., Sternweis, P. C., Gilman, A. G., and Mumby, S. M. (1992) Mol. Biol. Cell 3, 49 – 61 11. Casey, P. J., Solski, P. A., Der, C. J., and Buss, J. E. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 8323– 8327 12. Marshall, C. J. (1993) Science 259, 1865–1866 13. Fukada, Y., Takao, T., Ohguro, H., Yoshizawa, T., Akino, T., and Shimonishi, Y. (1990) Nature 346, 658 – 660 14. Ohguro, H., Fukada, Y., Takao, T., Shimonishi, Y., Yoshizawa, T., and Akino, T. (1991) EMBO J. 10, 3669 –3674 15. Dietrich, A., Meister, M., Brazil, D., Camps, M., and Gierschik, P. (1994) Eur. J. Biochem. 219, 171–178 16. Pitcher, J. P., Inglese, J., Higgins, J. B., Arriza, J. L., Casey, P. J., Kim, C., Benovic, J. L., Kwatra, M. M., Caron, M. G., and Lefkowitz, R. J. (1992) Science 257, 1264 –1267 17. Iniguez-Lluhi, J. A., Simon, M. I., Robishaw, J. D., and Gilman, A. G. (1992) J. Biol. Chem. 267, 23409 –23417 18. Matsuda, T., Takao, T., Shimonishi, Y., Murata, M., Asano, T., Yoshizawa, T., and Fukada, Y. (1994) J. Biol. Chem. 269, 30358 –30363 19. Yasuda, H., Lindorfer, M. A., Woodfork, K. A., Fletcher, J. E., and Garrison, J. C. (1996) J. Biol. Chem. 271, 18588 –18595 20. Kisselev, O., and Gautam, N. (1993) J. Biol. Chem. 268, 24519 –24522 21. Robishaw, J. D., Kalman, V. K., Moomaw, C. R., and Slaughter, C. A. (1989) J. Biol. Chem. 264, 15758 –15761 22. Sanger, F., Niklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 5463–5467 23. Graber, S. G., Lindorfer, M. A., and Garrison, J. C. (1996) Methods Neurosci. 29, 207–226 24. Summers, M. D., and Smith, G. E. (1987) Texas Agric. Exp. Stn. Bull. 1555, 1–56 25. Graber, S. G., Figler, R. A., Kalman-Maltese, V. K., Robishaw, J. D., and Garrison, J. C. (1992) J. Biol. Chem. 267, 13123–13128 26. Mumby, S. M., Casey, P. J., Gilman, A. G., Gutowski, S., and Sternweis, P. C. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 5873–5877 27. Hunt, D. F., Alexander, J. E., McCormack, A. L., Martino, P. A., Michel, H., Shabanowitz, J., and Sherman, N. (1991) in Techniques in Protein Chemistry II (Villafranca, J. J. ed) pp. 441– 454, Academic Press, New York 28. Schagger, H., and von Jagow, G. (1987) Anal. Biochem. 166, 368 –379 29. Morrissey, J. H. (1981) Anal. Biochem. 117, 307–310 30. Higgins, J. B., and Casey, P. J. (1994) J. Biol. Chem. 269, 9067–9073 31. Tuinman, A. A., Thomas, D. A., Cook, K. D., Xue, C., Naider, F., and Becker, J. M. (1991) Anal. Biochem. 193, 173–177 32. Xue, C., Ewenson, A., Becker, J. M., and Naider, F. (1990) Int. J. Peptide Protein Res. 36, 362–373 33. Wilcox, M. D., Schey, K. L., Busman, M., and Hildebrandt, J. D. (1995) Biochem. Biophys. Res. Commun. 212, 367–374 34. Morishita, R., Fukada, Y., Kokame, K., Yoshizawa, T., Masuda, K., Niwa, M., Kato, K., and Asano, T. (1992) Eur. J. Biochem. 210, 1061–1069 35. Yamane, H. K., Farnsworth, C. C., Xie, H., Howald, W., Fung, B. K., Clarke, S., Gelb, M. H., and Glomset, J. A. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 5868 –5872 36. Fukada, Y. (1995) Methods Enzymol. 250, 91–105 37. Boyer, J. L., Graber, S. G., Waldo, G. L., Harden, T. K., and Garrison, J. C. (1994) J. Biol. Chem. 269, 2814 –2819 38. Ueda, N., Iniguez-Lluhi, J. A., Lee, E., Smrcka, A. V., Robishaw, J. D., and Gilman, A. G. (1994) J. Biol. Chem. 269, 4388 – 4395 39. Wildman, D. E., Tamir, H., Leberer, E., Northup, J. K., and Dennis, M. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 794 –798 40. Kalman, V. K., Erdman, R. A., Maltese, W. A., and Robishaw, J. D. (1995) J. Biol. Chem. 270, 14835–14841 41. Page, M. J., Aitken, A., Cooper, D. J., Magee, A. I., and Lowe, P. N. (1990) Methods: Companion to Methods Enzymol. 1, 221–230 42. Lowe, P. N., Skinner, R. H., Cooper, D. J., Bradley, S., Sydenham, M., and Page, M. J. (1992) Biochem. Soc. Trans. 20, 484 – 487 43. Asano, T., Morishita, R., Matsuda, T., Fukada, Y., Yoshizawa, T., and Kato, K. (1993) J. Biol. Chem. 268, 20512–20519 44. Wilcox, M. D., Schey, K. L., Dingus, J., Mehta, N. D., Tatum, B. S., Halushka, M., Finch, J. W., and Hildebrandt, J. D. (1994) J. Biol. Chem. 269, 12508 –12513 45. Seabra, M. C., Reiss, Y., Casey, P. J., Brown, M. S., and Goldstein, J. L. (1991) Cell 65, 429 – 434 46. Buss, J. E., Quilliam, L. A., Kato, K., Casey, P. J., Solski, P. A., Wong, G., Clark, R., McCormick, F., Bokoch, G. M., and Der, C. J. (1990) Mol. Cell. Biol. 11, 1523–1530 47. Yokoyama, K., Goodwin, G. W., Ghomashchi, F., Glomset, J. A., and Gelb, M. H. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 5302–5306 48. Caldwell, G. A., Wang, S., Naider, F., and Becker, J. M. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 1275–1279


fully processed molecules (46). Our observation in the Sf9 cell system of a very small but detectable percentage of misprenylated molecules may be due to the relatively high concentration of heterologous bg substrate overwhelming the Sf9 prenyltransferase activity. Our results are not the same as the results of Kalman et al. (40), who observed marked heterogeneity of prenylation of both wild type g1 and a mutant g2 with an altered CaaX sequence (L71S). The differences observed may be due to the differences in Sf9 cell culture conditions and/or analytical techniques used in the two studies. All of the bg preparations characterized here were b and g coinfections, performed in suspension culture under a 50% oxygen, 50% nitrogen atmosphere, and were harvested at 48 h after infection, when cell viability was generally between 70 and 80%. We estimate, based on Western blots of the b subunit, that the initial detergent extract of the crude cell lysate recovers 50% of the b subunit present. Subsequent chromatographic steps result in approximately 90% recovery, giving an overall yield of 45%. bg dimers lacking the prenyl modification would not be isolated by the a-agarose affinity step in our purification procedure and would therefore not be observed in our mass spectral analysis. In contrast, Kalman et al. (40) used tritiated mevalonolactone to radiolabel g subunits. Crude cell lysates were subjected to gel electrophoresis, the region corresponding to g subunit excised, and the prenylthioether released by treatment with Raney nickel catalyst. Thus, differences in level of heterologous expression, Sf9 cell viability (and therefore potentially Sf9 prenyltransferase activity), substrate (bg dimer versus g alone), sample preparation (crude versus affinity-purified), and method of analysis may all have contributed to the observed differences. In conclusion, baculovirus-infected Sf9 cells accurately and specifically perform the same post-translational modifications observed in g1 and g2 subunits isolated from mammalian tissues. In the case of coinfection with recombinant baculovirus encoding b1, alteration of the g1 CaaX sequence to CVIL or the g2 CaaX sequence to CAIS is sufficient to result in geranylgeranylation and farnesylation, respectively. The resulting bg dimers, b1g1-S74L and b1g2-L71S, remain stably associated during the conditions of purification and are capable of sufficiently tight binding to a-agarose to allow specific elution with aluminum fluoride. Thus, the baculovirus Sf9 cell system provides an excellent model system for study of the role of prenylation in heterotrimeric G protein function. The accompanying paper reports on the effect of changes in prenylation on abgreceptor interaction (19).

18587

G Protein γ Subunits with Altered Prenylation Sequences Are Properly Modified When Expressed in Sf9 Cells Margaret A. Lindorfer, Nicholas E. Sherman, Karen A. Woodfork, Julia E. Fletcher, Donald F. Hunt and James C. Garrison J. Biol. Chem. 1996, 271:18582-18587. doi: 10.1074/jbc.271.31.18582

Access the most updated version of this article at http://www.jbc.org/content/271/31/18582 Alerts: • When this article is cited • When a correction for this article is posted

This article cites 47 references, 29 of which can be accessed free at http://www.jbc.org/content/271/31/18582.full.html#ref-list-1


Click here to choose from all of JBC's e-mail alerts