THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1990 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 265, No. 35, Isaue of December 15, pp. 21922-21928, 1990 Printed in U.S.A.

Purification, Characterization, and Western Blot Analysis of Human GTPase-activating Protein from Native and Recombinant Sources*

(Received for publication, June 11, 1990)

Robert Halenbeck&!, Walter J. Crosierll, Robin Clarkll, Frank McCormick7, and Kirston KothsS From the Departments of SProtein Chemistry and llMolecular Biology, Cetus Corporation, Emeryuille, California 94608

Human ras GTPase-activating protein (GAP) is a cytoplasmic factor that stimulates the GTPase activity of normal N-ras p2 1 while having no stimulatory effect on the GTPase activity of oncogenic variants of N-ras ~21. We have purified two forms of native rm GAP from human placental tissue. In addition to the M, = 120,000 type I GAP reported previously (l), an equiv- alent amount of an M, = 95,000 molecule with GAP activity was recovered and shown to have the N-ter- minal sequence expected for type II GAP. The two GAP forms in placental extracts were resolved by molecular sieve chromatography and appeared to have a mono- meric native structure.

Human recombinant type I GAP was produced intra- cellularly in Sp9 insect cells using a baculovirus expres- sion vector, and IO-mg quantities were purified to homogeneity in three steps. Comparison of the purified native and recombinant GAP molecules revealed that all three displayed similar biological specific activities in an in vitro GAP assay. A polyclonal antibody to purified recombinant GAP was prepared and shown to neutralize the activity of both native and recombinant GAPS. The antibody was also highly specific for the detection of native GAP by Western blot. Type I and II GAP species were detected in approximately equal amounts in cytoplasmic extracts of human placenta, but only type I GAP was observed when other human tissues were examined.

The ras family of proto-oncogenes encodes small guanine nucleotide-binding proteins that are thought to be involved in the regulation of cell growth (3). Point mutations in ras proteins have been found in a large number of tumors, impli- cating ras proteins in the cellular events leading to oncogen- esis in humans (4-6). Studies of the effects of these point mutations on ras function have led to the discovery of a protein that up-regulates the intrinsic GTPase activity of wild-type ras p21 while having no effect on the GTPase activity of mutated forms of ras p21 (7). This factor has been named GTPase-activating protein (GAP).’

Through the use of deletion and point mutation analyses,

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

§ To whom correspondence should be addressed: Dept. of Protein Chemistry, Cetus Coru.. 1400 53rd St.. Emervville, CA 94608.

’ The abbreviations used are: GAP, GTPase-activating protein; rGAP, recombinant GAP; PMSF, phenylmethylsulfonyl fluoride; DTT, dithiothreitol; HPLC, high performance liquid chromatogra- phy; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electro- phoresis; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid.

the effector binding site in ras p21 has been mapped. This region has been shown to be essential for m-s-GAP interac- tions (8-10). Oncogenic forms of ras ~21, whose GTPase activity is not up-regulated by GAP, have been shown to bind to GAP with an equal or greater affinity than wild-type ras p21 (11). These results and others (2, 24) implicate GAP as a downstream effector of rus ~21 action, up- or down-regulating cell growth, depending on whether GTP or GDP is bound to ras p21 in the ras. GAP complex (10).

The cDNAs coding for bovine (11) and human (1) GAPS have been isolated using oligonucleotide probes designed from peptide sequences of the corresponding GAP proteins. The bovine and human type I GAP cDNAs are 96% homologous at the amino acid level. A second, shorter form of human GAP cDNA (type II) has also been isolated from a human placental cDNA library (1). This GAP cDNA apparently is derived from a second mRNA species generated by an alter- native splicing event. The type II GAP cDNA is predicted to encode a 100,400-Da protein missing the first 180 amino acids of the type I GAP sequence and containing new amino acids at its N terminus. Although detection of native type II GAP has not been reported, analysis of various human tissues by Northern blotting has identified the mRNA for this GAP form, primarily in the placenta (1).

N-terminal deletion mutants of bovine GAP have recently been expressed as soluble lucZ fusion proteins in Escherichiu coli (12). A shortened form of GAP containing the C-terminal 343 amino acids was shown to have essentially the same specific bioactivity as full-length bovine GAP when measured in an in vitro GAP assay. These results indicate that a distinct catalytic domain exists at the C terminus of GAP.

In this report, we describe the purification of both type I and II GAPS from human placenta and compare these mole- cules to full-length human recombinant type I GAP purified from insect cells. In addition, we demonstrate the utility of a neutralizing polyclonal antibody to rGAP for the detection of native GAP in human tissue extracts by Western blotting.

MATERIALS AND METHODS

Assays for GAP Actiuity-The biological activities of native and recombinant GAP preparations were measured using either an im- munoprecipitation assay or a phosphate-release assay. The immu- noprecipitation assay (7) was based on the precipitation of N-ru.s- bound s-32P-labeled guanine nucleotide (70 nM) and measurement of GTP-to-GDP conversion by autoradiography or scintillation counting of TLC-fractionated material. This assay was used primarily to identify column fractions enriched for GAP activity during protein fractionation procedures. In certain column fractions in which non- specific GTPase activity was absent, the assay was run without an immunoprecipitation step (referred to as a non-IP assay).

The phosphate-release assay, based on release of [3ZP]-phosphate from N-ra.s-bound y”2P-labeled guanine nucleotide, was performed as a modification of the organic extraction method (13). Samples con- taining GAP activity were serially diluted in 20 mM Hepes (pH 7.6) containing 1 mM MgCl*, 2 mM DTT, 0.1% Nonidet P-40, 10 wg/ml

21922

by guest on September 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Native and Recombinant GTPase-activating Proteins

leupeptin, 200 PM PMSF, and 1 fig/ml pepstatin. 25 ~1 of each GAP dilution was equilibrated in a 25 “C water bath. Highly purified GAP preparations (1 mg/ml) were diluted 50,000-fold to measure specific

Production of rGAP Using Baculovirus Expression System-Se- quence coding for human type I GAP was recombined into the Autographa cabfornica baculovirus via the transfer vector pLP59-3. This was constructed by insertion of a synthetic sequence encoding the first 39 residues of GAP (ATG to SacII) at the NcoI site of pLP31, a derivative of the pAcC3 transfer vector (15, 16), followed by inser- tion of the S&I-BamHI fragment of pGAP-101. This construction places the GAP coding sequence directly adjacent to the polyhedrin untranslated leader sequence and thus eliminates 118 base pairs of 5’.untranslated GAP cDNA.

activities. Prior to initiating the GAP assay, purified recombinant N-ras (31)

was eauilibrated with ~Y-~*P~GTP for 10 min at 15 “C. For each reaction, 0.4 ~1 of [y-32P]GTP(Du Pont-New England Nuclear; 6000 Ci/mmol), 0.1 ~1 of N-rus (250 pg/ml), and 4.5 ~1 of 20 mM Tris (pH 8.0) containing 1 mM EDTA, 1 mM DTT, 1 mM ATP, and 1 mg/ml bovine serum albumin were used. The assay was initiated by adding 5 $1 of N-ro.s-bound [Y-~*P]GTP to the GAP samples. After 5 min at 25 “C, 200 ~1 of 5 mM silicotungstate and 1 mM HPSO, was added to stop the reaction. 300 pl of isobutyl alcohol/toluene (1:l) was then added, followed by 40 pl of 5% ammonium molybdate in 2 M H2S04. The samples were vortexed on high for 10 s and then allowed to sit for 5 min during phase separation. I50 ~1 of the organic phase (upper) was removed and quantitated by measuring Cerenkov radiation. The total amount of N-ro.s-boundlr-32P]GTP was calculated by measuring the amount of [32P)phosph&e released in tubes having a 200-fold excess of uuritied GAP (250-fold dilution of 1 ma/ml stock). A counts/ minute to picomole conversion factor was deteimined for each assay and used to calculate picomoles of GTP by comparing the manufac- turer’s specification of [-r-32P-]GTP concentration to the actual counts/minute measured.

This assay was used to measure the biological specific activities of native and recombinant GAP preparations. One unit of GAP activity was defined as 1 pmol of N-ras-bound GTP converted to GDP per min at 25 “C at an N-ro.s-bound GTP concentration of 18 nM.

Purification of Native GAP from Human Pkxenta-A fresh human placenta was obtained from a full-term delivery by Cesarian section and shown to be negative for human immunodeficiency virus and hepatitis A and B. The placenta was trimmed free of connective tissue, soaked in phosphate-buffered saline containing 10 mM EDTA and 1 pg/ml leupeptin (pH 7.4, 4 “C), and then stored in pieces at -70 “C. Approximately 300 g of frozen placenta was thawed in 1.5 liters of 10 mM sodium phosphate (pH 7.4) containing 150 mM sodium chloride, 10 mM EDTA, 2 pg/ml leupeptin, and 200 pM PMSF (lysis buffer) and then homogenized in a Waring blender. A cytoplasmic extract was prepared by centrifuging the homogenate at 10,000 x g for 15 min. Ammonium sulfate was added to 50% saturation, and the solution was held for 1 h at 4 “C. Precipitated protein was recovered by centrifugation at 10,000 x g for 10 min and resuspended in 400 ml of lysis buffer. The crude preparation was dialyzed for 4 h into 20 mM sodium phosphate (pH 6.5) containing 5 mM EDTA, 0.1 mM DTT, 1 rg/ml leupeptin, and 200 pM PMSF and chromatographed on an S-Sepharose column (5 X 10 cm; Pharmacia Fast Flow). Proteins were eluted with a 1.5-liter O-O.6 M sodium chloride gradient. GAP activity eluted as two peaks as detected using the immunopre- cipitation assay. Both peaks of GAP activity were pooled separately (peaks I and II eluted at 170-220 and 320-400 mM sodium chloride, respectively) and dialyzed into 30 mM Tris (pH 8.5) containing 1 mM EDTA, 0.1 mM DTT, 200 pM PMSF, 1 pg/ml leupeptin, and 10% glycerol. The pool of peak I GAP was divided into two portions, and each was subjected to ion-exchange chromatography using a TSK- DEAE-5-PW column (21.5 x 150 mm: Bio-Rad) eluted with a 45- min O-O.6 M sodium chloride gradient (3 ml/min). The pool of peak II GAP was purified by DEAE-HPLC in a single column run. DEAE column fractions were analyzed for GAP using the nonIP assay and Western blot analysis. The peak I and II GAP DEAE pools were dialyzed into 20 mM sodium phosphate (pH 6.5) containing 0.5 mM EDTA, 0.1 mM DTT, 100 UM PMSF, 1 &ml leuaentin, and 10% glycerol. Each GAP pool was further purified by-cation-exchange HPLC on a TSK-SP-5-PW column (7.5 x 75 mm: Bio-Rad) run in the pH 6.5 buffer described above. GAP was eluted’w~ith a 45Lmin O- 0.6 M sodium chloride gradient (1 ml/min). Tris (1 M, pH 8.0) was immediately added to each fraction to raise the pH to 7.5. To obtain a homogeneous type II GAP preparation for sequencing, type II GAP was purified by SDS-PAGE and transferred onto a polyvinylidene difluoride membrane as described below.

Characterization of Native GAP-The biological activity of purified type I GAP was stable for at least 2 weeks at 4 ‘C. The type II GAP preparation, however, lost activity upon storage. The protein concen- tration of the purified GAP preparations was determined by the method of Lowry et al. (14). Analytical size-exclusion HPLC was performed using a Superose 12 column (1 X 30 cm; Pharmacia LKB Biotechnology Inc.) run in phosphate-buffered saline (pH 7.4) con- taining 0.1 mM DTT, 0.5 pg/ml leupeptin, and 0.1% polyethylene glyCol4000 (1 ml/min).

pAcCl2 was derived from pVL941 (17) by deleting the unique &RI site at 7.15-kilobases and introducing-a polylinker sequence (GGTACCCGGGCTGCAGAATTCTAGATCTCGAGCTCCATGG- TGGATCC) at the unique BamHI insertion site.

To generate recombinant virus, 2 pg of transfer vector was cotrans- fected with 1 hg of wild-type viral DNA into Sr9 cells as described (18). Recombinant virus (occlusion-negative) was isolated from the transfection supernatant by plaque purification (19).

To produce recombinant GAP, Sr9 cells were infected with 5-10 plaque-forming units of recombinant virus/cell. Suspension cultures were grown in a stirred bio-reactor sparged with air containing 6 liters of protein-free medium (20). The cells were infected at l-l.5 X lo6 cells/ml and were harvested 48 h post-infection.

Purification of rGAP from Insect Cells-&S insect cells were pre- pared from an Il-liter culture (described above) and stored as an 85% slurry at -70 “C. The paste was thawed in 3 volumes (330 ml) of 20 mM sodium phosphate (pH 6.5) containing 5 mM EDTA, 0.1 mM DTT, 20 pM PMSF, 2 pg/ml leupeptin, and 1 *g/ml pepstatin A. Complete cell lysis was achieved by nitrogen cavitation (30 min at 250 p.s.i., 4 “C). A cytoplasmic extract was prepared by centrifuging the homogenate at 10,000 X g for 10 min at 4 “C and dialyzed for 4 h in two changes of the phosphate buffer described above. The dialyzed homogenate was recentrifuged (10,000 X g for 10 min), and the supernatant was loaded onto an S-Sepharose column (5 x 15 cm; Pharmacia Fast Flow) at 8 ml/min. The column was washed and eluted with a l-liter O-O.6 M sodium chloride gradient. Fractions enriched in M, = 120,000 GAP were identified by SDS-PAGE analysis and pooled. rGAP eluted at -200 mM NaCl. The pH of the GAP preparation was adjusted to 8.0 using 1 M Tris (pH 8.5), and the protein was concentrated 7-fold by ultrafiltration using an Amicon YM-10 membrane (final protein concentration of 3-4 mg/ml).

The pool of concentrated GAP protein (-50 ml) was divided in half, and each portion was loaded onto a Sephacryl 200 column (5 X 90 cm; Pharmacia LKB Biotechnology Inc.) equilibrated in 50 mM Tris (pH 8.5) containing 100 mM sodium chloride, 1 mM EDTA, 0.1 mM DTT, 200 /IM PMSF, 1 &g/ml leupeptin, and 1 rg/ml pepstatin A. Both columns were eluted overnight by gravity flow at 1.5 ml/min. Fractions containing GAP protein were identified by SDS-PAGE analysis on PhastGel 8-25 (Pharmacia LKB Biotechnology Inc.), pooled, and dialyzed into 30 mM Tris (pH 8.5) containing 1 mM EDTA, 0.1 mM DTT, 200 pM PMSF, 1 pg/ml leupeptin, and 1 pg/ml pepstatin A.

The dialyzed GAP protein was passed through a 0.45-p filter and pumped at 3 ml/min onto a TSK-DEAEd-PW column (21.5 x 150 mm; Bio-Rad) equilibrated in the same buffer. The column was washed and then eluted with a 60-min O-O.6 M sodium chloride gradient. Recombinant GAP eluted as two major peaks early in the column profile, as confirmed by SDS-PAGE analysis on PhastGel9 25 (Pharmacia LKB Biotechnology Inc.). Both DEAE pools were individually pooled and dialyzed into 10 mM Tris buffer (pH 7.4) containing 10% glycerol, 1 mM MgCl,, 1 mM EGTA, and 0.1 mM DTT. Recombinant GAP was filter-sterilized and stored at 4 ‘C.

Preparation of Neutralizing Polyclonal Antibody-Using the puri- fied rGAP preparation described above, a rabbit anti-rGAP antibody was prepared by Berkeley Antibody Co., Inc. (Richmond, CA). 500 pg of purified rGAP was mixed with complete Freund’s adjuvant and injected into the axial lvmph nodes of each of two New Zealand White rabbits. 21 days later, the rabbits were boosted by intramuscular injection of 250 pg of GAP in incomplete Freund’s adjuvant. 10 days later, the rabbits were bled from the ear vein, 20-30 ml of serum was prepared from each. The immunization protocol was continued for 13 months, boosting on day 21 and bleeding on day 31. Sera from bleeds 4-13 of rabbit 2 were pooled and subjected to a 50% ammonium sulfate precipitation. After centrifugation (10,000 X g for 10 min), the antibody was resuspended in the same volume, dialyzed into phos- phate-buffered saline, filter-sterilized, and stored at -70 “C.

Neutralization experiments were performed by serially diluting the anti-rGAP antibody in phosphate-buffered saline with a constant

by guest on September 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

21924 Native and Recombinant GTPase-activating Proteins

amount of native or recombinant GAP (15 units/ml), incubating overnight at 4 “C, and then assaying the remaining activity in the phosphate-release assay. One neutralizing unit is defined as the amount of antibody required to neutralize 50% of the activity of 2 units of GAP.

SDS-PAGE Purification and N-terminal Analysis-SDS-PAGE was performed using a modification of the method of Laemmli (21). Preparative isolation of native type II GAP was performed on a 6% SDS-polyacrylamide gel pre-electrophoresed with 0.1 M Tris (pH 8.9) containing 0.1% SDS after loading 15 ~1 of normal sample buffer containing 5% P-mercaptoacetic acid (final pH 6.8). The gel was run for 90 min at 40 mA. 10 Kg of DEAE-HPLC-purified type II GAP, prepared in normal sample buffer, was electrophoresed in normal running buffer containing 0.1 mM thioglycolic acid. The gel was transferred onto a polyvinylidene difluoride membrane (Applied Bi- osystems, Inc.), prewetted in methanol and then water, using the standard dry blot technique at 150 mA for 40 min for each 58.8 cm2 of membrane. The GAP protein bands were visualized by Coomassie Blue staining and subjected to automated sequence analysis using an Applied Biosystems 470A Gas-Phase Sequencer (22).

Western Blot Analysis-Western blots were prepared by electro- blotting SDS-polyacrylamide gels onto nitrocellulose membranes (Bio-Rad) and probing with the anti-rGAP antibody (1:lOOO dilution) for 4-15 h, followed with “?-protein A (Du Pont-New England Nuclear; NEX-146L; 5 &i/pg) at 0.35 &i/ml for 1 h, essentially as described by Burnette (23). The Western blots were blocked, washed, and probed at room temperature in 10 mM sodium phosphate (pH 7.4) containing 150 mM NaCl, 0.1% bovine serum albumin, 0.1% ovalbumin, 0.1% Tween 20, and 0.02% sodium azide.

Human tissues that were examined included l&week fetal brain, liver, lung, and spleen (obtained from Advanced Bioscience Re- sources, Inc., Oakland, CA). Adult tissues that were examined in- cluded male spleen, lung, and liver and female brain and uterus (obtained from the International Institute for the Advancement of Medicine, Exton, PA). Tissue homogenates were prepared in 10 mM sodium phosphate (pH 7.4) containing 10 mM EDTA, 2 Kg/ml leu- peptin, and 200 pM PMSF using a glass tissue homogenizer. The homogenates were clarified by centrifugation at 10,000 x g for 10 min and measured for protein content by the method of Lowry et al. (14). 20 rg of protein from each cytoplasmic extract was subjected to Western blot analysis.

RESULTS

Purification of Native GAP Species from Human Plucenta- Native GAP was purified from the cytoplasmic fraction of homogenized human placenta using a four-step procedure involving ammonium sulfate fractionation and a series of ion- exchange chromatography steps. The first cation-exchange column in the purification resolved GAP into two peaks of activity, as shown in Fig. 1.

Western blot analysis of the column fractions was per- formed using a polyclonal antibody generated against recom- binant GAP (described below). Fig. 1 (lower) shows two major immunoreactive bands with apparent M, values of -120,000 and -95,000 that eluted from the column in regions corre- sponding to the two GAP activity peaks. The upper band, which had a molecular weight consistent with type I GAP, eluted first from the column and trailed into the second GAP peak. The lower band, which had a molecular weight consist- ent with type II GAP, also eluted broadly, but peaked late in the gradient. The inability to completely resolve the GAP peaks from one another on this column may reflect structural or other heterogeneity in the two GAP forms. Western blot analysis also revealed several other bands with smaller molec- ular weights than either of the major GAP forms. These may represent proteolytic fragments of peak II GAP since they appear to coelute with the second peak of GAP activity.

The recent identification (30) of a nucleotide exchange- promoting activity in human placental extracts which acts on ras p21 raises the possibility that the GAP activity profile we observed may have reflected the presence of this exchange activity in our column fractions. This seems unlikely since

0 20 40 60 60 100 120 140

FRACTION NUMBER

0.6

i

o.4 c > F

Y 0.2

4

,.. 47- 6

30-- 25-

FIG. 1. Cation-exchange chromatography of crude GAP preparation. A placental cytoplasmic extract was fractionated by ammonium sulfate precipitation and chromatographed on an S-Seph- arose column. Upper, absorbance at 280 nm and GAP activity (as- sayed by the immunoprecipitation method) showing the conversion of ras-bound GTP to GDP plotted as the ratio of GDP to GDP + GTP; lower Western blot analysis of the column factions using the anti-rGAP antibody (described below) and “?-protein A. The two major GAP bands are indicated. Fractions 72-82 were pooled for peak I GAP, and fractions 108-120 for peak II GAP.

1 234 5 6 7 kDa

200- : c,

f 1

30-- ) ,_ a- - -

FIG. 2. SDS-PAGE analysis of native GAP purification. Samples were electrophoresed on an 8% SDS-polyacrylamide gel and stained with Coomassie Blue. Lane I, placental cytoplasmic extract; lane 2, resuspended 50% ammonium sulfate pellet; lane 3, S-Sepha- rose peak I; lane 4, DEAE-HPLC peak I; lane 5, sulfopropyl HPLC peak I: lane 6, S-Sepharose peak II; lane 7, DEAE-HPLC peak II. Scanning densitometric analysis of lanes 5 and 7 indicated the purity of type I and II GAPS to be 90 and 27%, respectively. The material at the dye front was included in the determination of purity.

the total amount of radioactive nucleotide bound by ras ~21 in the assay samples remained constant across the column profile.

The fractions enriched in the two GAP forms were pooled and purified separately through DEAE-HPLC and sulfopro- pyl HPLC as described under “Materials and Methods.” In these two subsequent purification steps, peak I GAP resolved as a relatively homogeneous species, whereas peak II GAP appeared to trail on both chromatography steps.

SDS-PAGE analysis of the purification of both GAP forms is shown in Fig. 2. Purified peak I GAP migrated as a single band at an apparent M, of 120,000 and had a purity of 90% (lane 5), as determined by scanning densitometry. Peak I

by guest on September 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Native and Recombinant GTPase-activating Proteins 21925

GAP migrated at a similar molecular weight under nonreduc- ing conditions (data not shown). The DEAE-purified peak II GAP preparation consisted of two bands on reducing SDS- PAGE at apparent M, of 95,000 and 80,000 respectively (lane 7). To purify peak II GAP to homogeneity for N-terminal sequence analysis, preparative SDS-PAGE was performed. After electrophoresis, the proteins were transferred onto a polyvinylidene difluoride membrane, visualized by Coomassie Blue staining, and subjected to N-terminal sequence analysis (described below).

The biological specific activity of the purified GAP species is reported in Table I. Type I GAP was purified -2,000-fold with an overall yield of 5%. The biological specific activity was 33,000 units/mg, as determined in a Pi-release assay using N-ras-bound GTP as the GAP substrate. Type II GAP was purified -700-fold with an overall yield of 11%. Since type II GAP was recovered from the DEAE-HPLC step at a purity of only 27% (determined by scanning densitometry of a Coo- massie Blue stained SDS-polyacrylamide gel), the specific activity of this GAP form was estimated to be -39,000 units/ mg, essentially equivalent to that of type I GAP.

N-terminal Sequencing of Native Type II GAP-The results of N-terminal sequence analysis of the purified M, = 95,000 GAP species are shown in Fig. 3. A single sequence was detected corresponding to the first 20 amino acids predicted to be at the N terminus of type II GAP (1). These results establish that type II GAP translation initiates near the 3’- end of the 65-base-pair insert, beginning with Met-Lys-Gly, followed by the type I GAP sequence from amino acid 181 onward, as predicted by Trahey et al. (1). The N terminus of type II GAP differs from that of type I GAP, which was found to be blocked (1). To confirm this result, type I GAP was sequenced following the preparative SDS-PAGE method used to purify type II GAP. No sequence was detected under loading conditions similar to those utilized for type II GAP.

Analysis of Native Types I and II GAPS by Molecular Sieve Chromatography-A freshly prepared placental cytoplasmic extract was subjected to molecular sieve chromatography to determine the native molecular weight of the active GAP species. The results are shown in Fig. 4. A broad peak of GAP activity was detected eluting from the column, as shown in Fig. 4 (upper). Analysis of the column fractions by Western blot using the anti-rGAP antibody (Fig. 4, lower) detected GAP species of two distinct molecular weights, corresponding to type I and II GAPS. The observed native molecular weights are consistent with each type of GAP having primarily a monomeric rather than a multimeric structure.

Expression of rGAP in Insect Cells-The expression of recombinant type I GAP was accomplished in Sf9 insect cells by infection with the recombinant baculovirus AcGAPll. This virus was produced as described (18,19) using a transfer vector in which the coding sequence of pGAP-101 (1) was fused to the A. californica NP virus polyhedrin promoter. In suspension culture, AcGAPll induced expression levels as high as 10 mg/liter. This level of expression is 2-3-fold higher than that obtained with our initial vector, AcGAP5, in which the entire GAP cDNA (including 118 base pairs of 5’-untrans- lated sequence) was inserted into the EcoRI site of pAcC12, a polylinker derivative of pVL941.

Purification of rGAP Produced in Insect Cells-Initial at- tempts to purify rGAP resulted in proteolyzed preparations containing four GAP species detected by SDS-PAGE: intact rGAP, at M, = 120,000; two partially proteolyzed GAP species at M, = 105,000 and 100,000 respectively; and a “limit-digest” form of M, = 95,000 (see Fig. 5, lane 7). Upon further storage at 4 “C, the entire preparation was converted into the M, = 95,000 GAP species (data not shown). The biological activity of rGAP remained constant despite the proteolytic degrada- tion. Four measures were subsequently employed to minimize proteolytic degradation: 1) addition of the protease inhibitor leupeptin; 2) use of NP cavitation to lyse the cells (rather than freeze/thaw/sonication) and to minimize breakage of lyso- somes; 3) minimization of exposure to pH values below 8, at which lysosomal proteases are most active; and 4) inclusion of a molecular sizing chromatography step, which was shown on an analytical scale to remove most of the protease activity.

By using the precautions described above, cell paste from 11 liters of insect culture yielded 22 mg of unproteolyzed rGAP with an overall recovery of 20% (see Table II). SDS- PAGE analysis of the purification is shown in Fig. 5. The final rGAP preparation was >90% pure and had a biological specific activity of -20,000 units/mg. This is within experi- mental error of the value measured for native type I GAP protein. rGAP remained fully active for >l year of storage at 4 “C.

In the final purification step (preparative DEAE-HPLC), rGAP was resolved into two protein peaks, both of which had essentially equivalent biological specific activities. SDS- PAGE analysis revealed that each peak contained a single band of rGAP protein at the molecular weight corresponding to unproteolyzed M, = 120,000 rGAP (Fig. 5, lanes 5 and 6). When the two DEAE peaks were separately dialyzed and rechromatographed using analytical DEAE-HPLC, they no longer appeared to be different from one another since they

TABLE I Purification of native type I and II GAPS from human placenta

Volume Total Activityb Total Specific protein” activity activity Recovered Purification

ml mg units/ml units unitslmg % -fold Placental cytoplasmic extract 1,500 11,250 120 180,000 15 100 1 Ammonium sulfate 440 2,860 560 245,000 85 136 5

Type I GAP S-Sepharose 127 219 830 105,400 380 58 24 DEAE-HPLC 8.7 2.8 5,530 48,100 17,200 27 1,080 SP’-HPLC 2.0 0.28 4,660 9,300 33,200 5 2,080

Type II GAP S-Sepharose 150 84 780 117,000 1,390 65 90 DEAE-HPLC 4.9 2.0 4,200 20,600 10,500 11 660

(L Protein concentration was determined by the method of Lowry et al. (14). ‘Units of GAP activity were determined using the Pi-release assay with N-ras-bound [r-“‘P]GTP as the

substrate. ’ SP, sulfopropyl.

by guest on September 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

21926 Native and Recombinant GTPase-activating Proteins

FIG. 3. N-terminal sequences predicted and determined for native GAP. The amino acid sequences of type I and II GAPS were predicted from the nucleotide sequences of their corresponding cDNA clones (1). No N-terminal sequence was detected for type I GAP by Edman degradation, presumably due to blockage at the N terminus. The first 20 N-terminal amino acids of GAP II were obtained by direct sequencing (out to residue 197, using the amino acid numbering for type I GAP).

1.2

1 .o

0.6 E

g 0.6

: a

0.4

0.2

0.0

15 25 35 45

FRACTION NUMBER

0.0

0.5 .

0.4 i

0.3 E 5 F

0.2 Y

% 0.1 CD

kDa

AP I AP II

FIG. 4. Separation of type I and II GAPS by molecular sieve chromatography. A placental cytoplasmic extract was fractionated by molecular sieve chromatography on a Superose 12 column. Upper absorbance at 280 nm and GAP activity (assayed by the immunopre- cipitation method) showing the conversion of r--bound GTP to GDP plotted as the ratio of GDP to GDP + GTP; her, Western blot analysis of the column fractions using the anti-rGAP antibody and ““I-protein A. Molecular weight standards were: thyroglobulin (void volume), bovine globulin (lSO,OOO), chicken ovalbumin (43,000), bo- vine myoglobin (17,000), and cyanocobalamin (1,300).

eluted with identical retention times. Since the two GAP species had identical reducing and nonreducing SDS-PAGE molecular weights (data not shown), the same biological spe- cific activities, and identical analytical DEAE-HPLC profiles, it seems possible that the heterogeneity observed on a pre- parative scale resulted from conformational rearrangements of rGAP during column chromatography or to weak interac- tions with other uncharacterized species.

Preparation of Neutralizing Polyclonnl Antibody to rGAP- A rabbit anti-human polyclonal antibody was generated against purified rGAP. To determine the neutralizing titer of the anti-rGAP antibody, serial dilutions of the antibody were incubated overnight with a constant amount of native and recombinant GAPS and then tested for GAP activity in the Pi-release assay. The antibody displayed an equivalent neu-

1234567

FIG. 5. SDS-PAGE analysis of recombinant GAP purifica- tion. Samples were electrophoresed on a 6% SDS-polyacrylamide gel and stained with Coomassie Blue. Lane 1, insect cell cytoplasmic extract; lane 2, S-Sepharose unbound fraction; lane 3, S-Sepharose; lane 4, Sephacryl 200 peak; lane 5, DEAE-HPLC peak A; lane 6, DEAE-HPLC peak B; lane 7, partially proteolyzed preparation of purified type I rGAP.

tralizing titer against all three forms of GAP tested, whereas the preimmune serum had no effect. The titer was 3300 neutralizing units/ml of serum. One neutralizing unit is de- fined as the amount of antibody required to neutralize 50% of the activity of 2 units of GAP. The antibody was shown to be specific for GAP since it had no effect on the activity of rap ~21 GAP, a human protein with functional similarity to ras GAP,2 or on the intrinsic GTPase activity of ras p21 (24).

Western Blot Analysis Using Anti-rGAP Antibody-The anti-rGAP antibody was tested for its ability to detect rGAP on Western blots (Fig. 6, lanes 1-4). As little as 2 ng of rGAP could be detected on 24-h film exposures using this antibody. The level of native GAP detected by Western blot analysis of human tissue extracts was -lo-fold higher than this limit of detection.

Various human tissues were analyzed by Western blot to characterize the relative distribution of the different GAP species. Purified type I and II GAPS were run as standards for comparison (Fig. 6, lanes 5 and 6). Analysis of a fresh placental cytoplasmic extract (lane 7) revealed two GAP species having molecular weights equivalent to those of the purified GAP standards. The two GAP bands were present in approximately equal amounts. Although placental tissue con- tains cells of fetal origin, analysis of cytoplasmic extracts of other human fetal tissues (18-week) revealed only type I GAP to be present at this stage of development in brain (lane 8), lung (lane 9), liver (lane IO), and spleen (data not shown). Similar results were observed upon analysis of cytoplasmic extracts of the same tissues from adults (data not shown) and from the human breast carcinoma cell line SKBR-3 (lane 11). Comparison of the relative amount of GAP detected in each lane indicated that placenta had significantly more GAP per microgram of cytoplasmic protein than the other tissues ex- amined. However, since we studied only the cytoplasmic frac- tion of the various human tissues, the possibility exists that differing levels of membrane-bound forms of GAP may also exist in these tissues.

DISCUSSION

We have purified two different forms of native GAP that are present in a placental cytoplasmic extract in approxi- mately equal amounts. The larger molecular weight form has been previously shown to be type I GAP (1) and represents the predominant GAP species detected in the mammalian tissues that have been examined. A smaller molecular weight form of GAP was also purified and was shown by N-terminal

* P. G. Polakis, unpublished observation.

by guest on September 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Native and Recombinant GTPase-activating Proteins

TABLE II Purification of recombinant type Z GAP from insect cells

21927

Volume Total protein” Activityb Total Specific

activity activity Recovered Purification

ml w units/ml units units/mg % -fold Insect cell cytoplasmic extract 525 1,250 4,000 21 x lo5 1,700 100 1 S-Sepharose peak 118 170 16,000 19 x lo5 11,000 90 6 Sephacryl200 peak 59 44 16,000 9.4 x lo5 21,000 45 12 DEAE-HPLC

Peak I 7.0 13.6 39,000 2.7 x lo5 20,000 13 12 Peak II 5.5 8.8 32,000 1.8 x lo5 20,000 9 12 Combined 12.5 22.4 35,500 4.5 x lo5 20,000 22

’ Protein concentration was determined by the method of Lowry et al. (14). * Units of GAP activity for DEAE-purified rGAP were determined using the P,-release assay with N-ras-bound

[y-32P]GTP as the substrate. Units of GAP activity for all other samples were determined using the immunopre- cipitation method where purified rGAP was used as a standard with a defined specific activity of 20,000 units/mg. The method used is only accurate within a factor of 2.

12 345 6 7891011

kOs

hydrophobic domain that may be involved in the association of GAP with the cell membrane. Such an association may be important in ras-GAP signal transduction (10, 12). The pro- posed hydrophobic domain is not present on type II GAP, which lacks the first 180 amino acids of the type I sequence. If type I and II GAP proteins are both expressed within the same cell type, the relative amounts of the two species may regulate signal transduction among GAP, membrane-bound ras ~21, and other cellular proteins. In addition, the 3 new amino acids present at the amino terminus of type II GAP could also affect the affinity of GAP for important regulatory or effector proteins. It is probably significant that the region immediately following these 3 N-terminal amino acids repre- sents the first of two conserved SH2 domains (1).

47-

FIG. 6. Western blot analysis using anti-rGAP antibody. Samples were electrophoresed on an 8% SDS-polyacrylamide gel, transferred to nitrocellulose by eiectroblotting, and probed with the anti-rGAP antibody detected by lz51-protein A. Lanes 1-4, 50, 10, 2, and 0.4 ng of rGAP, respectively; lane 5, purified native type I GAP, lane 6, DEAE-purified type II native GAP, lanes 7-l 1,20 rg of protein from cytoplasmic extracts of placenta, fetal brain, fetal lung, fetal liver, and human breast carcinoma cell line SKBR-3, respectively. The autoradiograph is a 4-h film exposure with a Lightning-Plus screen at -70 “C.

sequence analysis to correspond to type II GAP. The shorter type II GAP was not simply a truncated form of type I GAP because its N-terminal sequence is not present anywhere in the type I GAP sequence. Both GAP species had similar biological specific activities when assayed on N-ras p21 in vitro.

The mRNA coding for type II GAP was first identified in human placenta (1) and is thought to be derived by a differ- ential splicing event. We have shown that the protein corre- sponding to this GAP message exists at a high level in the placenta and that, based on the N-terminal sequence data, initiation of translation does indeed occur at the ATG in the 65-base pair mRNA insert, as originally proposed (1). The apparent absence of this form of GAP in other human adult or fetal tissues that apparently contain a significant level of type I GAP suggests that this splicing event may be specific to certain tissue types in placenta and may represent a mech- anism for controlling production of type II GAP. In addition, the apparent restriction of type II GAP expression to placenta may also reflect a specific function of type II GAP in control- ling certain cellular events.

An M, = 125,000 protein having GAP activity has been purified from bovine brain (25) and was subsequently shown to be homologous to the type I form of human GAP (11). Interestingly, no evidence was reported for type II GAP in bovine brain. These results are supported by our finding that only type I GAP is detectable in human adult or fetal brain.

The cDNAs coding for bovine and human GAPS have been isolated, and through amino acid and homology analyses, various GAP domains have been proposed (1, 11). The N- terminal 160 amino acids of type I GAP represent a potential

Molecular sizing analysis resolved native type I and II GAP species and showed them to have molecular weights consistent with a monomeric native structure. However, since we studied only the cytoplasmic fraction of the placenta, the existence of membrane-associated forms of GAP in complexes with other proteins was not addressed by this study.

rap GAP, a human protein that stimulates the GTPase activity of the ras homolog, rap ~21, has been purified in two forms: 1) as a soluble protein (26) and 2) from detergent- solubilized membranes.3 Although the homology between rap GAP and ras GAP is not known, the finding that a neutral- izing anti-ras GAP antibody does not affect rap GAP activity or detect rap GAP on Western blots2 suggests that the ho- mology between these two proteins may be small.

Recombinant type I GAP was produced intracellularly in insect cells to provide sufficient quantities of GAP for bio- chemical studies and to facilitate the generation of a poly- clonal antibody to rGAP. To obtain an unproteolyzed GAP preparation, the purification method had to be optimized for the removal and inactivation of contaminating proteases. Purified rGAP that was obtained had physical properties similar to those of native type I GAP and also displayed a specific activity in vitro that was essentially equivalent to that of native GAP on N-m-s ~21.

The specific activity that we obtained for purified type I GAP was apparently lo-fold higher than that reported for bovine GAP (25). This may be a reflection of the different affinities of bovine and human GAPS for human rus p21 proteins, differences in the individual GAP preparations, or other differences in the assays themselves. Nonetheless, the properties of human recombinant GAP are similar, if not identical, to those of native human GAP, validating the use

3 Polakis, P. G., Rubinfeld, B., Evans, T., and McCormick, F. (1991) Proc. Natl. Ad. Sci. U. S. A., in press.

by guest on September 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

21928 Native and Recombinant GTPase-activating Proteins

of this recombinant protein for investigation into the action of GAP in cellular events.

10. McCormick, F., Adari, H., Trahey, M., Halenbeck, R., Koths, K., Martin, G. A., Crosier, W. J., Watt, K., Rubinfeld, B. & Wong, G. (1988) Cold Spring Harbor Symp. Quant. Biol. 53,849-854

Vogel. U. S.. Dixon. R. A. F.. Schaber. M. D.. Diehl. R. E.. The generation of a highly sensitive antibody against hu- man rGAP has allowed the identification of both type I and II GAPS in crude lysates. In addition, the antibody has also been used to immunoprecipitate GAP in association with other cellular proteins that may be important in RX-GAP signal transduction (27). Similar anti-rGAP antibodies have been used to demonstrate that GAP associates with the plate- let-derived growth factor receptor in platelet-derived growth factor-treated cells (28) and with M, = 190,000 and 62,000 proteins in transformed and epidermal growth factor-stimu- lated cells (29). Thus, the specificity and high titer of this antibody make it a useful tool for understanding the complex mechanisms involved in ras-GAP actions in the cell.

Acknowledgements-We thank Ken Watt for performing N-ter- minal sequence analysis and for advice concerning preparative SDS- PAGE, Meg Trahey for help in purifying native GAP, Paul Polakis for providing rap GAP and measuring its activity, David Lowe for producing rGAP-containing insect cell paste, Lootsee Panganiban and Jim Devlin for constructing pLP.59, and Gideon Bollag for providing additional recombinant GAP. We also thank Derrek Pavey and Eric Ladner for preparation of the figures.

REFERENCES

1. Trahey, M., Wong, G., Halenbeck, R., Rubinfeld, B., Martin, G. A.. Ladner. M.. Lone. C. M., Crosier. W. J., Watt, K., Koths, K.‘& McCdrmick, F.(1988) Science 242, 1697-1700

2. Hall, A. (1990) Cell 61,921-923 3. Paterson, H., Reeves, B., Brown, R., Hall, A., Furth, M., Bos, J.,

Jones, P. & Marshall, C. (1987) Cell 51,803-812 4. Varmus, H. E. (1984) Annu. Reu. Genet. l&553-612 5. Barbacid. M. (1987) Annu. Rev. Biochem. 56,779-827 6. Bos, J. L: (1988) M&at. Res. 195.225-271 7. Trahev. M. & McCormick. F. (1987) Science 238.542-545 8. Adari,*& Lowry, D. R., Wiliums&, B. M., De;, C. J. & Mc-

Cormick (1988) Science 240,518-521 9. Cal&, C., Hancock, J. F., Marshall, C. J., & Hall, A. (1988)

Nature 332, 548-551

11.

12.

13. Shatter; E. (1984) Anal. Biochem. 138,416-420 14. Lowrv. 0. H.. Rosebroueh. N. J.. Farr. A. L. & Randall. R. J.

15.

16.

(19%i) J. B&l. Chem. l-9$,2651275 ’ Devlin, J. J., Devlin, P. E., Clark, R., O’Rourke, E. C., Levenson,

C. & Mark, D. F. (1989) Bio/Technology 7, 286-292 Luckow, V. A. & Summers, M. D. (1988) Bio/Technology 6, 47-

55 17. 18.

Luckow, V. A. & Summers, M. D. (1989) Virology 170, 31-39 Summers, M. D. & Smith, G. E. (1987) Z’ex. Agric. Exp. St. Bull.

1555 19. Smith, G. E., Summers, M. D. & Fraser, M. J. (1983) Mol. Cell.

Biol. 3,2156-2165 20. Maiorella, B., Inlow, D., Shauger, A. & Harano, D. (1988) Bio/

Technology 6.1406-1410 21. 22.

23. 24.

Laemmli, U. K. (1970) Nature 227,680-685 Hunkapiller, M. W., Hewick, R. M., Dreyer, W. J. & Hood, L. E.

(1983) Methods Enzymol. 9 1,399-413 Burnette, W. N. (1981) Anal. Biochem. 112, 195-203 Yatani, A., Okabe, K., Polakis, P., Halenbeck, R., McCormick, F.

& Brown, A. M. (1990) Cell 61,769-776 Gibbs, J. B., Schaber, M. D., Allard, W. J., Sigal, I. S. & Scolnick,

E. M. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 5026-5030 Kikuchi, A., Sasaki, T., Araki, S., Hata, Y. & Takai, Y. (1989) J.

Biol. Chem. 264.9133-9136

25.

26.

27.

28.

29.

30.

31.

Gaishall, h. S., Scolnick, El M., Sigal, I. S. & Gibds, J. B: (1988) Nature 335, 90-93

Marshall, M. S., Hill, W. S., Ng, A. S., Vogel, U. S., Schaber, M. D., Scolnick, E. M., Dixon, R. A. F., Sigal, I. S. & Gibbs, J. B. (1989) EMBOJ. 8. 1105-1110

Kaplan, D. R., Morrison, D. K., Wong, G., McCormick, F. & Williams. L. T. (19901 Cell 61. 125-133

Kazlauskas; A., illis, c., Paws& T. & Cooper, J. A. (1990) Science 247, 1578-1581

Ellis, C., Moran, M., McCormick, F. & Pawson, T. (1990) Nature 343,337-381

Downward, J., Riehl, R., Wu, L. & Weinberg, R. A. (1990) Proc. Natl. Acad. Sci. Cr. S. A. 87, 5998-6002

Trahey, M., Milley, R. J., Cole, G. E., Innis, M., Paterson, H., Marshall, C. J., Hall, A. & McCormick, F. (1987) Mol. Cell. Biol. 7,541-544

by guest on September 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from

R Halenbeck, W J Crosier, R Clark, F McCormick and K KothsGTPase-activating protein from native and recombinant sources.Purification, characterization, and western blot analysis of human

1990, 265:21922-21928.J. Biol. Chem. 

  http://www.jbc.org/content/265/35/21922Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

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

  http://www.jbc.org/content/265/35/21922.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on September 20, 2020

http://ww

w.jbc.org/

Dow

nloaded from