gtp-blot analysis of small gtp-binding proteins : the c-terminus is involved in renaturation of...

Eur. J. Biochem. 225, 99-105 (1994) 0 FEBS 1994

GTP-blot analysis of small GTP-binding proteins The C-terminus is involved in renaturation of blotted proteins

Franz-Josef KLINZ’.’ ‘ Institut fur Pharmakologie, Freie Universitat Berlin, Germany

Lehrstuhl fiir Molekulare Neurobiochemie, Ruhr-Universitat Bochum, Germany

(Received June 2, 1994) - EJB 94 079213

Recombinant c-Ha-ras, ralA and rap2, but not raplA or raplB proteins retained their ability to bind [a-32P]GTP after SDSPAGE and transfer to nitrocellulose. Recombinant c-Ha-ras missing the C-terminal 23 amino acid residues failed to bind [a-32P]GTP after the blot, and the ability of recombinant ralA missing the C-terminal 28 amino acid residues to bind [a-32P]GTP was decreased manyfold. The presence of nonionic detergents of the polyoxyethylene type such as Tween 20, Triton X-100, Nonidet P40 or Lubrol PX in the incubation buffer was necessary to induce renaturation of blotted recombinant c-Ha-ras protein, whereas other types of detergents were ineffective. We pro- pose that detergents of the polyoxyethylene type induce the refolding of some types of blotted small GTP-binding proteins and that the C-terminus is involved in the refolding process.

Membranes from NIH3T3 fibroblasts overexpressing c-Ha-ras protein showed much weaker binding of [a-32P]GTP as expected from the level of ras immunoreactivity. Treatment of fibroblasts with lovastatin, an inhibitor of hydroxymethylglutaryl-coenzyme A reductase, caused the accumulation of the unfarnesylated form of c-Ha-ras in the cytosol. Examination of [a-32P]GTP-binding and immunoreactivity for cytosolic and membrane-bound c-Ha-ras revealed that binding of [CX-~~P]GTP to unprocessed c-Ha-ras was increased about threefold compared to the same amount of processed c-Ha-ras.

Our results demonstrate that detection and quantification of small GTP-binding proteins in eukaryotic cells by GTP-blot analysis is hampered by the fact that these proteins differ strongly in their ability to renature after blotting to nitrocellulose.

Guanine nucleotide binding regulatory proteins such as ras proteins and a-subunits of heterotrimeric GTP-binding proteins exert their function by cycling between an inac- tive, GDP-bound, and an active, GTP-bound, conformation [l-31.

Members of the mammalian ras protooncogene group, c-Ha-ras, c-Ki-ras and N-ras each encode small guanine nucleotide binding proteins of 21 kDa which are associated with the inner side of the plasma membrane. Mammalian ras genes acquire transformation-inducing properties by single point mutations within their coding sequences [ l , 31. Cloning and sequencing of cDNAs has revealed that a large and di- verse family of ras-related small GTP-binding proteins exist with calculated molecular masses between 21 -25 kDa. Groups belonging to this family include the rap, ral, rho, rac and rab proteins [4,5]. Members of the family of small GTP- binding proteins show an extraordinary conservation of sequences dedicated to guanine nucleotide binding.

To crystallize the human c-Ha-ras protooncogene product, John et al. [6] expressed the protein with a truncated C- terminus of 23 amino acids. The resulting c-Ha-ras (amino acid residues 1-166) was crystallized as a complex with the slowly hydrolyzing GTP analogue guanosine 5’-[P,y- imido]triphosphate (GppNHp) and analyzed by X-ray crys-

Correspondence to F.-J. JSlinz, Lehrstuhl fur Molekulare Neurobiochemie, Ruhr-Universitat Bochum, Universitatsstr. 150, D-44780 Bochum, Germany

tallography [7]. The truncated c-Ha-ras has biochemical properties very similar to those of the full-length protein, but thermal stability and binding to nitrocellulose filters are different [ 81.

Soon after discovery of the ras proteins it was shown by McGrawth et al. [9] that ras proteins in lysates from ras- producing Escherichia coli cells retain their ability to bind [a-32P]GTP after SDSPAGE and transfer to nitrocellulose. Later it was shown that, in contrast to blotted a-subunits of heterotrimeric GTP-binding proteins, at least some of the blotted low-molecular-mass GTP-binding proteins present in mammalian cells are able to bind [(x-~~PIGTP [lo]. Identified recombinant proteins with this property include ralA [ l l ] , rabl-6 [12] and rap2B [13].

Ras proteins are localized at the inner side of the plasma membrane in eukaryotic cells, whereas most other small GTP-binding proteins seem to be associated with intracellu- lar membranes. The C-terminus of ras proteins is essential for attaching the proteins to the plasma membrane, in large part due to modifications of the last four amino acids, the so- called CAAX motif. These modifications include attachment of a farnesyl group to the cysteine of the CAAX motif, removal of the final three amino acids, and methylation of the newly exposed a-carboxyl group [14]. Inhibition of hydroxymethylglutaryl-coenzyme A reductase by lovastatin blocks farnesylation and subsequently cleavage and rnethyl- ation at the C-terminus of c-Ha-ras [15]. In the case of c-Ha-

100

ras, avidity of binding to the cell membrane is further increased by attachment of a palmitoyl group to Cysl81 and Cys184 [15].

Results described in this paper denionstrate that renaturation and binding of [a-"PIGTP to blotted c-Ha-ras protein is influenced by the C-terminal structure and by the type of detergent used for renaturation.

MATERIALS AND METHODS Materials

Materials were obtained from the following sources : [a-32P]GTP (800 Ci/mmol) from New England Nuclear; nitrocellulose membranes (BA83, 0.2 pm) from Schleicher & Schiill ; Hyperfilm-MP from Amersham Buchler; sodium dodecyl sulfate (SDS) and sodium deoxycholate from Serva; urea from Fluka; Tween 20 and materials for SDS/PAGE from BioRad; n-octyl glucoside and nucleotides from Boeh- ringer (Mannheim) ; molecular mass marker proteins, bovine serum albumin and other detergents used from Sigma. Lovas- tatin was generously supplied by Dr A. Alberts (Merck, Sharp and Dohme, Rahway, NJ, USA). All other chemicals were purchased from Merck.

Purified small GTP-binding proteins used for this study were more than 95% pure and bind 1 mol nucleotide/mol of protein (A. Wittinghofer, personal communication).

Expression of small GTP-binding proteins in E. coli Recombinant raplA [16], truncated raplA (amino acid

residues 1-168), rap2 [17], truncated rap2 (amino acid residues 1 - 167) and ralA [18] were produced in E. coli using the ptaq 32 vector. Colonies were grown overnight at 37°C in 50 ml of Luria-Bertani medium containing 50 pg/ml of both ampicillin and kanamycin. A 10-ml sample of this culture was then used to inocculate 1000 ml medium. The culture was allowed to grow at 37°C to an A,,, of about 0.5. Isopropyl thio-P-D-galactoside was added to a final concentration of 0.5 mM and the culture was incubated for 3 h at 37 "C. Bacteria were harvested by centrifugation, and the cell sediment was lysed at room temperature in sample buffer according to Laemmli [19]. The lysate was cleared by centrifugation, and an aliquot of the supernatant was applied to SDSPAGE, [a-"PIGTP blot analysis and subsequently immunoblot analysis.

Overexpression of ras proteins in NIH3T3 fibroblasts NIH3T3 fibroblasts transfected with the human c-Ha-ras

protooncogene or the Ha-ras (Va112) oncogene under control of a glucocorticoid-inducible mouse mammary tumor virus long terminal repeat were cultured, as described by Jaggi et al. [20]. If indicated, dexamethasone was added for 24 h at a concentration of 1 pM, either alone or in combination with lovastatin at a concentration of 20 pM. Cells were harvested and membranes (25000Xg sediment) were prepared as described by Vachon et al. [21].

Cytosol (200000Xg sediment) was isolated by spinning the 25000Xg supernatant for 1 h at 200000Xg and 4°C. Pro- tein concentration was estimated by the method of Peterson [22] using bovine serum albumin as a standard.

SDSPAGE If not Otherwise indicated, recombinant small GTP-bind-

ing proteins and membranes together with molecular mass

marker proteins were denatured for 5 min at 95 "C in sample buffer according to Laemmli [19]. Proteins were then sepa- rated on SDS/polyacrylamide gels prepared by the method of Laemmli [19]. The slab gels (1 mm thick) consisted of a 5% stacking gel (10mm long) and a 12.5% separating gel (50 mm long).

Western blot Subsequently to SDSPAGE, gels were incubated for

30 min in protein transfer buffer, pH 8.3, containing 25 mM Tris/HC1 192 mM glycine, 20% (by vol.) methanol and 0.05% (mass/vol.) SDS [lo, 231. Proteins were electrotransferred onto nitrocellulose membranes for 2 h at 100 V in transfer buffer. Proteins on blots were stained with Poinceau S to ascertain that comparable amounts of protein were transferred to nitrocellulose.

[a-32P]GTP blot analysis If not otherwise stated, nitrocellulose blots were washed

for 10 min at room temperature in binding buffer, pH 7.5, containing 50mM Tris/HCl, 20 pM MgC1, and 0.3% (by vol.) Tween 20. Blots were then incubated for 60 min at room temperature in binding buffer, containing 1 pCi/ml [a-32P]GTP (800 Ci/mmol). After washing five times for 10 min in binding buffer, blots were air-dried, and bound [a-3'P]GTP was detected by autoradiography at 4 "C. Expo- sure times were between 3-19 h, with the exception of autoradiographs for Figs 5 and 6, where blots were exposed for 2 days to visualize minor bands. After exposure, GTP- blots were stored at room temperature. For some experiments, labelled bands were cut from the nitrocellulose and radioactivity was measured by liquid scintillation ; values for blank strips of the same size were subtracted for the back- ground. For some experiments, intensity of bands on autoradiographs was evaluated using a video densitometer (Biotek- Fischer, Reiskirchen, Germany). All experiments were re- peated two or three times with similar results.

Immunoblot analysis Subsequently to [a-"PIGTP blot analysis, nitrocellulose

blots were preincubated in buffer, pH 8.0, containing 10 mM Tris/HCI, 150 mM NaCI, 0.05% (by vol.) Tween 20 and 1% (mass/vol.) bovine serum albumin and then reacted with the appropriate dilution of mouse monoclonal antibody followed by anti-(mouse IgG) - alkaline-phosphatase conjugate as specified by the manufacturer's instructions (Promega). Two mouse monoclonal antibodies were used for this study: Tu- mark-ras 11 (New England Nuclear) [24] and 142-24E05 (Microbiological Associates, Bethesda, MD, USA) [25]. To demonstrate that comparable amounts of recombinant c-Ha- ras protein were present on nitrocellulose after GTP-blot analysis, blots were reacted with ras 11 monoclonal antibody. For some experiments, intensity of stained bands on immu- noblots was evaluated using a video densitometer.

RESULTS

Recombinant c-Ha-ras protein [26] was denatured, subjected to SDSPAGE and blotted to nitrocellulose. We first tested the effect of different types of detergents on binding of [a-'*P]GTP to blotted recombinant c-Ha-ras. Binding buffer

1 2 3 4 5 1 2 3 4 5 B C

Fig. 1. Binding of [a-3ZP]GTP to blotted recombinant c-Ha-ras protein: effect of (A) different types of detergents, (B) concentration of Tween 20 and (C) concentration of MgCI,. 1 pg purified recombinant c-Ha-ras proteiflane was subjected to SDSPAGE and transferred to nitrocellulose. (A) Strips were incubated in buffer containing 50mM Tris/HCl pH7.5, 20 pM MgCl,, 1 pCi/ml [a-”PIGTP and 0.1% (by vol.) Tween 20 (lane I), 0.1% (by vol.) Triton X-100 (lane 2), 0.1% (by vol.) Nonidet P40 (lane 3), 0.1% (by vol.) Lubrol PX (lane 4), 0.1% (masdvol.) n-octyl glucoside (lane 5), 0.1% (masshol.) sodium deoxycholate (lane 61, 0.1% (masshol.) SDS (lane 7) or 2 M urea (lane 8). (B) Strips were incubated in buffer pH 7.5 containing 50 mM Tris/HCl, 20 pM MgC1, and 1 pCi/ml [LZ-~~P]GTP in the absence (lane 1) or presence of Tween 20 at concentrations (by vol.) of 0.03% (lane 2), 0.1% (lane 3), 0.3% (lane 4) and 1% (lane 5). (C) Strips were incubated in buffer pH 7.5 containing 50 mM Tris/HCl, 0.3% (by vol.) Tween 20 and 1 pCi/ml [~x-~’P]GTP in the absence (lane 1) or presence of MgCl, at concentrations of 0.2 pM (lane 2), 2 pM (lane 3), 20 pM (lane 4) and 200 pM MgC1, (lane 5). Autoradiographs are shown. Values on the right indicate the position of molecular mass marker proteins.

contained 50 mM Tris/HCl pH 7.5, 20 pM MgC1, and 0.1% detergent. In the presence of Tween 20, Triton X-100, Noni- det P40 or Lubrol PX, binding of [a-”PIGTP to blotted c-Ha-ras was easily detectable (Fig. 1A). All these detergents belong to the polyoxyethylene type. In contrast, binding of [a-”’P]GTP to blotted recombinant c-Ha-ras was not observed in the presence of the non-ionic detergent n-octyl glucoside and the anionic detergents sodium deoxycholate or sodium dodecyl sulfate (SDS) (see Fig. 1A). Other detergents tested that did not allow the binding of [a-32P]GTP to blotted recombinant c-Ha-ras were the non-ionic detergent digitonin and the zwitterionic detergent Chaps (data not shown). Also incubation in 2 M urea did not result in binding of [a-32P]GTP to blotted recombinant c-Ha-ras (see Fig. 1A).

We next tested the effect of different concentrations of Tween 20 and MgCl, on binding of [a-32P]GTP to blotted recombinant c-Ha-ras protein. In the presence of 0.03% (by vol.) Tween 20 little binding of [a-32P]GTP to blotted c-Ha- ras was detectable (Fig. 1 B). Increasing the concentration of Tween 20 to 1 % resulted in increased binding of [a-32P]GTP to blotted c-Ha-ras. In standard GTP-blot experiments we used Tween 20 at a concentration of 0.3% [lo].

It was demonstrated that the presence of Mg2+ is necessary for binding of [a-3ZP]GTP to blotted small GTP-binding proteins [9, lo]. In the case of blotted c-Ha-ras, weak binding of [a-32P]GTP was observed at a concentration of 0.2 pM MgC1, (Fig. 1C). Under the concentrations tested, 20 pM MgC1, was most effective to support binding of [a-”P]GTP to blotted c-Ha-ras ; therefore this concentration was chosen for standard GTP-blot experiments. Binding of [a-”P]GTP to blotted recombinant c-Ha-ras could also be performed in the presence of 200 pM MnCl, (data not shown).

1 2 3 A

1 2 3 B

101

4

Fig. 2. Binding of [a-”P]GTP to blotted recombinant c-Ha-ras protein: effect of other nucleotides (A) and denaturation temperature (B). 1 pg purified recombinant c-Ha-ras proteinllane was subjected to SDSPAGE and transferred to nitrocellulose. (A) Strips were incubated in buffer pH 7.5 containing 50 mM Tris/HCl, 20 pM MgCl,, 0.3% (by vol.) Tween 20 and 1 pCi/ml [a-”PIGTP and either no other nucleotide (lane l), 10 pM GTP (lane 2) , 10 pM GDP (lane 3) or 10 pM ATP (lane 4). (B) Before being applied to SDSPAGE, purified recombinant c-Ha-ras protein was denatured for 5 min in sample buffer according to Laemmli [19] at 25°C (lane I), 60°C (lane 2) and 95°C (lane 3). Autoradiographs are shown.

To test the specificity of binding of [a-3ZP]GTP to blotted recombinant c-Ha-ras, we added an excess of unlabelled nucleotide to the binding reaction. Addition of 10 pM GTP or 10 pM GDP to the standard incubation abolished binding of [a-32P]GTP completely, whereas 10 pM ATP was less effective (Fig. 2A). Similar results were obtained with platelet membranes, where binding of [a-32P]GTP to small GTP- binding proteins was also partially blocked in the presence of 10 pM ATP [lo].

We have also analysed whether the denaturation temperature has an influence on binding of [a-32P]GTP to blotted recombinant c-Ha-ras. Incubation of recombinant c-Ha-ras in sample buffer according to Laemmli 1191 for 5 min at 25”C, 60°C or 95°C did not result in different binding of [a-3ZP]GTP after blotting to nitrocellulose (Fig. 2 B).

To examine the relationship between concentration of blotted recombinant c-Ha-ras and binding of [a-32P]GTP, we electrotransferred increasing amounts of c-Hams to nitrocellulose. As can be seen in Fig. 3, binding of [a-32P]GTP to blotted c-Ha-ras is concentration-dependent, but not linear over the concentration range tested. We observed the binding of about 10000 cpm [a-32P]GTP (800 Ci/mmol) to 1 pg blotted recombinant c-Ha-ras, the amount used for most experiments. About 2400 cpm [a-32P]GTP binds to 0.1 pg recombinant c-Ha-ras, leading to the conclusion that 0.03% of blotted c-Ha-ras is able to renature and bind [a-32P]GTP.

The experiments described provide no evidence on whether nonionic detergents of the polyoxyethylene type induce renaturation of the denatured recombinant c-Ha-ras protein after blotting to nitrocellulose, thereby leading to an increased binding of [a-32P]GTP. Alternatively, detergents like Tween 20 may facilitate binding of [a-3zP]GTP to blotted recombinant c-Ha-ras in a more direct way. To discriminate between these two alternatives, we performed an experiment in which blotted c-Ha-ras was preincubated without [a-32P]GTP in the presence or absence of Tween 20 followed by incubation with [a-32P]GTP in the presence or absence of Tween 20. As expected, omission of Tween 20 from both preincubation and incubation reaction resulted in no binding of [a-32P]GTP to blotted c-Ha-ras (Fig. 4). Preincubation of

102

15 r 1

c-Ha-ras Cpg)

Fig. 3. Concentration-dependent binding of [a-3ZP]GTP to blotted recombinant c-Ha-ras protein. Recombinant c-Ha-ras protein in amounts of 0.05 pg (lane l), 0.1 pg (lane 2), 0.2 pg (lane 3), 0.5 pg (lane 4), 1 pg (lane 5 ) , 1.5 pg (lane 6) or 2 pg (lane 7) was subjected to SDSPAGE and electrotransferred to nitrocellulose. The blot was incubated with [a-32P]GTP under standard conditions. The autoradiograph is from a representative GTP-blot, which was processed for quantification by liquid scintillation counting. The plot shows the concentration-dependent increase in binding of [a-32P]GTP to recombinant c-Ha-ras; values ? SEM (n = 4).

Preincubation

Tween20 - - + + - - + + Minutes 10 60 10 60 10 60 10 60

Incubation

Tween20 - - - - + + + + Minutes 60 60 60 60 60 60 60 60

Fig. 4. Effect of incubation conditions on renaturation and/or binding of [a-3ZP]GTP to blotted recombinant c-Ha-ras protein. 1 pg purified recombinant c-Ha-ras proteidane was subjected to SDSPAGE and transferred to nitrocellulose. Strips were first incubated for the given time in buffer pH 7.5 containing 50 mM Trid HCl and 20 pM MgCl, in the absence (-) or presence (+) of 0.3% (by vol.) Tween 20. Strips were then incubated for 60 min in buffer pH 7.5 containing 50 mM Tris/HCl, 20 pM MgC1, and 1 pCi/ml [a-32P]GTP in the absence (-) or presence (+) of 0.3% (by vol.) Tween 20. An autoradiograph is shown.

blotted c-Ha-ras in the presence of Tween 20 followed by incubation in the absence of Tween 20 resulted in stronger binding of [a-3’P]GTP than preincubation in the absence of Tween 20 followed by incubation in the presence of Tween 20 (see Fig. 4). This result supports the idea that Tween 20 enables renaturation of blotted recombinant c-Ha-ras on nitrocellulose. Preincubation and incubation of blotted c-Ha-ras in the presence of Tween 20 resulted in only slightly increased binding of [a-3ZP]GTP compared to the presence of Tween 20 in the preincubation alone (see Fig. 4).

We next tested binding of [a-32P]GTP to several blotted recombinant ras-related proteins in lysates from E. coli or

1 2 3 4 5 6 7 a 9 1 0 1 1 kDa - 66 - 45 - 36 - 29 - 24

- 20 - 14

Fig.5. [a-32P]GTP-blot analysis of lysates from E. coli or yeast cells producing recombinant small GTP-binding proteins. The following samples were subjected to SDSPAGE and transferred to nitrocellulose: 1 pg purified recombinant c-Ha-ras protein (lane 1) ; a lysate from uninduced E. coli (lane 2); a lysate from E. coli producing raplA (lane 3); an extract from yeast S. cerevisiae (lane 4); an extract from yeast S. cerevisiae expressing raplB (lane 5 ) ; a lysate from uninduced E. coli (lane 6) ; a lysate from E. coli producing rap2 (lane 7); a lysate from uninduced E. coli (lane 8); a lysate from E. coli producing rap2 (amino acid residues 1-167) (lane 9); a lysate from uninduced E. coli (lane 10); a lysate from E. coli producing ralA (lane 11). The blot was incubated with [a-32P]GTP under standard conditions. An autoradiograph is shown. Values on the right indicate the position of molecular mass marker proteins.

yeast Saccharomyces cerevisiae. Using an inducible system for expression of ras-related proteins in E. coli, we analysed lysates from uninduced and induced bacteria. As shown in Fig. 5, no binding of [a-3’P]GTP to lysates from E. coli overexpressing raplA was observed. In contrast, weak binding of [a-32P]GTP to a 22-kDa band in lysates from E. coli overexpressing rap2 became visible (see Fig. 5) . Overexpression of truncated rap2 (amino acid residues 1 - 167), missing the C- terminal 16 amino acids, resulted in loss of binding of [a-”P]GTP. In lysates from E. coli overexpressing ralA, strong binding of [a-3ZP]GTP to a 26-kDa band was observed (see Fig. 5). Even without induction, basal expression of ralA was high. In lysates from yeast S. cerevisiae overexpressing raplB no binding of [cI-~*P]GTP to a band in the 22-kDa range was observed (see Fig. 5). The 25-kDa band, visible in both lysates from control cells and raplB-expressing cells, is caused by endogenous small GTP-binding proteins known to be present in the yeast S. cerevisiae.

We also tested binding of [a-”P]GTP to blotted purified recombinant small GTP-binding proteins containing or missing specific C-terminal sequences. Both blotted c-Ha-ras and Ha-ras (Va112) bound [c~-~’P]GTP, whereas truncated c-Ha- ras (amino acid residues 1-166) and Ha-ras (Va112, amino acid residues 1 - 166) were no longer able to bind [a-”P]GTP after blotting to nitrocellulose (Fig. 6A). Also blotted recombinant ralA showed strong binding of [a-”P]GTP, whereas binding to truncated ralA (amino acid residues 1 - 178) was hardly visible. Purified recombinant rapl A and truncated raplA (amino acid residues 1-167) were unable to bind [a-”P]GTP after blotting to nitrocellulose (see Fig. 6A), con- firming our result obtained with lysates from E. coli overexpressing raplA protein.

We found that monoclonal antibody 142-24E05, which was raised against a synthetic peptide consisting of amino acid residues 96-118 of ras proteins 1251, also crossreacted with ralA protein, as was shown recently for raplA [27]. Incubation of the GTP-blot with this antibody revealed that c-Ha-ras, ralA and rapl A full-length and truncated proteins were still present in comparable amounts on nitrocellulose

103

A

B

1 2 3 4 5 6 7 8 kDa - 45 - 36 - 29 - 24

- 20

- 14 - 45 - 36 - 29 - 24

- 20

- 14

Fig. 6. [a-3ZP]GTP-blot and immunoblot analysis of recombinant small GTP-binding proteins. 1 pg each of different purified recombinant small GTP-binding proteins was subjected to SDSPAGE and transferred to nitrocellulose : c-Ha-ras (lane 1) ; truncated c-Ha-ras (amino acid residues 1-166) (lane 2); Ha-ras (Vall2) (lane 3); truncated Ha-ras (Val12, amino acid residues 1-166) (lane 4); ralA (lane 5 ) ; truncated ralA (amino acid residues 1-178) (lane 6); raplA (lane 7); truncated raplA (amino acid residues 1-168) (lane 8). (A) The blot was incubated with [a-32P]GTP under standard conditions. An autoradiograph is shown. (B) The GTP-blot was subsequently reacted with 142- 24E05 mAb and alkaline-phosphatase- conjugated anti-mouse IgG. A photograph of the stained blot is shown. Values on the right indicate the position of molecular mass marker proteins.

after incubation with [a-32P]GTP (Fig. 6B). In conclusion, only defined small GTP-binding proteins bind [cz-’~P]GTP after transfer to nitrocellulose, but they loose this ability if C-terminal sequences are removed.

Even in membranes from a tissue with a high concentration of ras proteins, proteins in the 22-kDa range can hardly be detected that bind [a-32P]GTP after transfer to nitrocellulose [28]. Since the GTP-blot technique is sensitive enough to detect small amounts of recombinant c-Ha-ras (see Figs 3 and 8A), we asked whether post-translational modifications may be responsible for the decreased ability of membrane- associated ras proteins to bind [a-32P]GTP after blotting to nitrocellulose. We therefore used membranes from NIH3T3 cells expressing either c-Ha-ras or Ha-ras (Va112) under control of a glucocorticoid-inducible promoter. Membranes from uninduced and induced NIH3T3 fibroblasts revealed a GTP- binding band in the 24-kDa range representing endogenous small GTP-binding proteins. In membranes from cells that were induced to overexpress ras protein, an additional band of 22 kDa bound [a-”P]GTP weakly (Fig. 7A). The blot was subsequently incubated with a monoclonal antibody directed against ras proteins (Fig. 7B). Surprisingly, membranes from NIH3T3 fibroblasts overexpressing c-Ha-ras protein showed much weaker binding of [a-”P]GTP than was expected from the level of ras immunoreactivity.

We therefore treated NIH3T3 fibroblasts overexpressing c-Ha-ras with lovastatin to examine whether inhibition of farnesylation, and subsequent processing reactions of c-Ha- ras in intact cells, is followed by increased binding of [a-32P]GTP after blotting to nitrocellulose. Membrane and cytosolic fractions were prepared, subjected to SDSPAGE and electrotransferred to nitrocellulose. Binding of [a-”P]- GTP to blotted proteins revealed that lovastatin treatment of NIH3T3 fibroblasts overexpressing c-Ha-ras caused the dis- appearance of a 22-kDa band from the membrane fraction,

A

B

1 2 3 4 5 6 kDa - 45 - 36 - 29 - 24

- 20 - 14

- 45 - 36 - 29 - 24

- 20

- 14

Fig. 7. [a-32P]GTP blot and immunoblot analysis of ras proteins in membranes from NIH3T3 fibroblasts overexpressing c-Ha- ras. The following samples were subjected to SDSPAGE and transferred to nitrocellulose: 1 pg purified recombinant c-Ha-ras (lane 1) or truncated c-Ha-ras (amino acid residues 1-166, lane 2); 50 p g membrane proteins from uninduced NIH3T3 fibroblasts (lane 3) ; 50 pg membrane proteins from NIH3T3 fibroblasts overexpressing c-Ha-ras (lane 4); 50 pg membrane proteins from uninduced NIH3T3 fibroblasts (lane 5 ) ; 50 pg membrane proteins from NIH3T3 fibroblasts overexpressing Ha-ras (Val1 2) (lane 6). (A) The blot was incubated with [a-32P]GTP under standard conditions. An autoradiograph is shown. (B) The GTP-blot was subsequently reacted with ras 11 mAb and alkaline-phosphatase-conjugated anti- mouse IgG. A photograph of the stained blot is shown. Values on the right indicate the position of molecular mass marker proteins.

1 2 3 4 5 6 7 % kDa - 36 - 29 - 24

A - 20 - 14

B

- 36 - 29 - 24

- 20 - 14

Fig. 8. [a-3ZP]GTP blot and immunoblot analysis of ras proteins in membranes and cytosol from lovastatin-treated fibroblasts overexpressing c-Ha-ras. The following samples from NIH3T3 fibroblasts overexpressing c-Ha-ras protein were subjected to SDS/ PAGE and transferred to nitrocellulose: 20 pg membrane proteins from control cells (lane 1); 20 pg membrane proteins from lovastatin-treated cells (lane 2); 20 pg cytosolic proteins from control cells (lane 3); 20 pg cytosolic proteins from lovastatin-treated cells (lane 4); 20 ng (lane 5), 50 ng (lane 6), 100 ng (lane 7) and 200 ng (lane 8) of purified recombinant c-Ha-ras. (A) The blot was incubated with [a-32P]GTP under standard conditions. An autoradiograph is shown. (B) The GTP-blot was subsequently reacted with ras 11 mAb and alkaline-phosphatase-conjugated anti-mouse IgG. A photograph of the stained blot is shown. Values on the right indicate the position of molecular mass marker proteins.

whereas a GTP-binding protein with a slightly lower electrophoretic mobility appeared in the cytosolic fraction (Fig. 8 A). Increasing amounts of recombinant c-Ha-ras were coelectrophoresed and blotted from the same gel. As can be

104

seen in Fig. 8, electrophoretic mobility of recombinant c-Ha- ras is slightly increased compared to the cytosolic form of c- Ha-ras, suggesting that the attachment of two palmitoyl groups to cytosolic c-Ha-ras leads to this difference in electrophoretic mobility.

The GTP-blot was subsequently incubated with a monoclonal antibody against ras proteins to estimate the amount of overexpressed c-Ha-ras protein in membrane and cytosolic fractions (Fig. 8 B). Quantitative evaluation by a video densitometer revealed that the unprocessed cytosolic form of c-Ha-ras binds about threefold more [a-”P]GTP than the processed membrane-bound form of c-Ha-ras but that the binding of [a-’2P]GTP is comparable between the cytosolic form of c-Ha-ras and recombinant c-Ha-ras (referred to the same amount of ras immunoreactivity).

The conclusion obtained from these experiments supports our finding that the C-terminal region of c-Ha-ras is involved in renaturation of the protein after blotting to nitrocellulose.

DISCUSSION

Binding of [a-32P]GTP to blotted ras and ras-related proteins is now widely used for analysis of small GTP-binding proteins and was reported to occur in the presence of MgC1, [9, 101 and nonionic detergents like Nonidet P40 [9] or Tween 20 [lo]. Ligand blotting of a,-macroglobulin to its receptor was also performed after renaturation of the blotted receptor in the presence of Nonidet P40 [29].

We have tested the effect of different types of nonionic and ionic detergents on binding of [a-’”PIGTP to blotted recombinant c-Ha-ras and found that binding is promoted only by nonionic detergents of the polyoxyethylene type (Tween 20, Nonidet P40, Triton X-100, Lubrol PX), but not by nonionic detergents of other types like n-octyl glucoside and digitonin or by ionic detergents. Our experiments suggest that detergents of the polyoxyethylene type stimulate renaturation of blotted recombinant c-Ha-ras on nitrocellulose and thereby promote the binding of [a-3ZP]GTP. The reason why only detergents with a polyoxyethylene structure induce refolding of blotted c-Ha-ras is unclear at the moment. As was outlined by O’Farrell [30], in the presence of Nonidet P40, SDS comes off the proteins and forms mixed micelles with Nonidet P40. Possibly, only detergents of the polyoxyethylene type are able to displace efficiently the protein-bound SDS and thereby allow renaturation of blotted proteins.

The renaturation in the presence of Tween 20 was seen only with some types of blotted recombinant small GTP- binding proteins. Whereas Tween 20 induced renaturation of c-Ha-ras, ralA and rap2, no effect was observed for raplA and raplB.

Binding of [a-”P]GTP to blotted recombinant c-Ha-ras is concentration-dependent, but not linear over the concentration range tested. An explanation for this result may be that the binding capacity of nitrocellulose for recombinant c-Ha- ras becomes limiting. Even under our optimized conditions, stoichiometry of binding of [a-32P]GTP to blotted recombinant c-Ha-ras is relatively low. Despite the fact that only 0.03% of blotted c-Ha-ras is able to renature and bind [a-32P]GTP, recombinant c-Ha-ras in the nanogram range can be easily detected by the GTP-blot technique.

We found that C-terminal sequences of ras and ras-related proteins are involved in the renaturation process induced by Tween 20, because deletion of the C-terminal 23 amino acid residues in c-Ha-ras and the C-terminal 16 amino acid resi-

dues in rap2 prevented renaturation completely, whereas deletion of the 28 amino acid residues from the extreme C- terminus of ralA decreased renaturation and thereby binding of [a-32P]GTP manyfold. The behaviour of truncated c-Ha- ras (amino acid residues 1 - 166) blotted to nitrocellulose was strongly different from that in solution where biochemical properties like binding of guanine nucleotides were essen- tially unaltered compared to the full-length protein [8]. How- ever, it should be noted that thermal stability and binding to nitrocellulose filters were different [ 81. These results demonstrate that, in solution, C-terminal sequences of c-Ha-ras are not important for binding of GTP but influence other parame- ters like thermal stability. Our data show that C-terminal sequences of different small GTP-binding proteins are important for renaturation of these protein after blotting to nitrocellulose.

Small GTP-binding proteins differ mostly from each other in C-terminal amino acid sequences. The fact that both blotted recombinant raplA and raplB do not renature and bind [a-32P]GTP in the presence of Tween 20 may be ex- plained by unfavourable C-terminal sequences that are highly similar between these two proteins. As was reported by Ohmstede et al. [13], recombinant rap2B protein binds [a-”P]GTP after blotting to nitrocellulose. This fits well with our data showing that blotted recombinant rap2 is also able to bind [a-32P]GTP, because both proteins have highly similar C-terminal sequences. In this context, it is interesting to note that recombinant rabl-6 all bind [a-”P]GTP after the blot [12], suggesting that the ability for renaturation depends on C-terminal amino acid sequences characteristic for subgroups of small GTP-binding proteins.

The C-terminus of ras proteins is post-translationally modified before insertion into the plasma membrane of eukaryotic cells. In the case of c-Ha-ras these modifications include attachment of a farnesyl group to Cysl86, removal of the final three amino acids, methylation of the newly exposed carboxyl group and attachment of a palmitoyl group to Cysl81 and Cys184 [15]. Since ras proteins were not easily detectable by GTP-blot analysis in membranes from brain tissue with high ras content [28], we considered whether post-translational modification(s) may be critical for renaturation and binding of [a-32P]GTP to ras proteins blotted from membranes of eukaryotic cells. To compare directly binding of [a-32P]GTP to the unfarnesylated and farnesylated form of c-Ha-ras from mammalian cells, we treated NIH3T3 fibroblasts overexpressing c-Ha-ras with lovastatin, an inhibitor of isoprenylation. As was expected, lovastatin treatment caused the accumulation of unprocessed c-Ha-ras in the cytosol. We found that, referred to the same amount of blotted c-Ha-ras immunoreactivity, binding of [a-32P]GTP to unprocessed cytosolic c-Ha-ras was comparable to recombinant c-Ha-ras but was increased about threefold compared to processed membrane-bound c-Ha-ras.

These data suggest that the C-terminus of ras and proba- bly ras-related proteins determines renaturation efficiency of blotted proteins by detergents of the polyoxyethylene type. It is unclear at the moment whether the presence of modifying groups linked to Cys186 of processed c-Ha-ras may result in decreased renaturation efficiency of blotted ras proteins by detergents of the polyoxyethylene type. Alternatively, the presence of three additional amino acids present in unprocessed eukaryotic as well as bacterially expressed c-Ha-ras may increase renaturation efficiency.

In conclusion, our results show that detergents of the polyoxyethylene type induce renaturation of several blotted

105

small GTP-binding proteins and that the structure of the C- terminus is important for this process. The fact that renaturation does not occur with all blotted small GTP-binding proteins and, furthermore, is different between eukaryotic and bacterially expressed proteins restricts the application of the GTP-blot technique for analysis and quantification of ras and ras-related proteins in eukaryotic cells.

I thank Prof. G. Schultz (Berlin, Germany) and Prof. R. Heu- mann (Bochum, Germany) for their support. I am grateful to Prof. A. Wittinghofer (Dortmund, Germany) for purified recombinant c-Ha-ras, truncated c-Ha-ras (amino acid residues 1 - 166), Ha-ras (Va112), truncated Ha-ras (Va112, amino acid residues 1 -166), raplA, truncated raplA (amino acid residues 1-168), ralA and truncated ralA (amino acid residues 1-178) proteins, E. coli strains producing raplA and ralA, for stimulating discussions and for critically reading the manuscript. I thank Dr V. Pizon (Paris, France) for an extract from yeast S. cerevisiae expressing raplB and for E. coEi strains producing rap2 and truncated rap2 (amino acid residues 1 - 167). NIH3T3 fibroblasts overexpressing ras proteins were kindly provided by Prof. B. Groner (Freiburg, Germany). The technical assistance of R. Bonnet (Berlin, Germany) and D. Peter (Bochum, Germany) is gratefully acknowledged. I thank Dr H.-H. Kiltz (Bochum, Germany) for critically reading the manuscript and K. Grabert (Bochum, Germany) for photographic work. This work was supported by grants from the Deutsche Forschungsgemeinschaft to Prof. G. Schultz (Berlin, Germany) and to Prof. R. Heumann (Bochum, Germany).

REFERENCES 1. Santos, E. & Nebreda, A. R. (1989) Structural and functional

properties of ras proteins, FASEB J. 3, 2151-2163. 2. Bourne, H. R., Sanders, D. A. & McCormick, F. (1991) The

GTPase superfamily : conserved structure and molecular mechanism, Nature 349, 117-126.

3. Grand, R. J. A. & Owen, D. (1991) The biochemistry of ras p21, Biochem. J. 279, 609-631.

4. Hall, A. (1990) The cellular functions of small GTP-binding proteins, Science 249, 635-640.

5. Downward, J. (1990) The ras superfamily of small GTP-binding proteins, Trends Biochem. Sci. 15, 469-472.

6. John, J., Schlichting, I., Schiltz, E., Rosch, P. & Wittinghofer, A. (1989) C-terminal truncation of p21H preserves crucial ki- netic and structural properties, J. Biol. Chem. 264, 13086- 13 092.

7. Pai, E. F., Kabsch, W., Krengel, U., Holmes, K. C., John, J. & Wittinghofer, A. (1 989) Structure of the guanine-nucleotide- binding domain of the Ha-ras oncogene product p21 in the triphosphate conformation, Nature 341, 209-214.

8. Wittinghofer, A. & Pai, E. F. (1991) The structure of ras protein: a model for a universal molecular switch, Trends Biochem. Sci. 16, 382-387.

9. McGrath, J. P., Capon, D. J., Goeddel, D. V. & Levinson, A. D. (1984) Comparative biochemical properties of normal and activated human ras p21 protein, Nature 310, 644-649.

10. Bhullar, R. P. & Haslam, R. J. (1987) Detection of 23-27 kDa GTP-binding proteins in platelets and other cells, Biochem. J. 245, 617-620.

11. Bhullar, R. P., Chardin, P. & Haslam, R. J. (1990) Identification of multiple ral gene products in human platelets that account for some but not all of the platelet G,-proteins, FEBS Lett. 260,48-52.

12. Zahraoui, A, Touchot, N., Chardin, P. & Tavitian, A. (1989) The human rab genes encode a family of GTP-binding proteins

related to yeast YPTI and SEC4 products involved in secre- tion, J. Biol. Chem. 264, 12394-12401.

13. Ohmstede, C.-A,, Farrell, F. X., Reep, B. R., Clemetson, K. J. & Lapetina, E. G. (1990) Rap2B: a ras related GTP-binding protein from platelets, Proc. Natl Acad. Sci. USA 87, 6527- 6531.

14. Clarke, S. (1992) Protein isoprenylation and methylation at carboxyl-terminal cysteine residues, Annu. Rev. Biochem. 61, 355 -386.

15. Hancock, J. F., Magee, A. I., Childs, J. E. & Marshall, C. J. (1989) All ras proteins are polyisoprenylated but only some are palmitoylated, Cell 57, 1167-1177.

16. Frech, M., John, J., Pizon, V., Chardin, P., Tavitian, A., Clark, R., McCormick, F. & Wittinghofer, A. (1990) Inhibition of GTPase activating protein stimulation of ras-p21 GTPase by the krev-I gene product, Science 249, 169- 171.

17. Lerosey, I., Chardin, P., de Gunzburg, J. & Tavitian, A. (1991) The product of the rap2 gene, member of the ras superfamily: biochemical characterization and site-directed mutagenesis, J. Biol. Chem. 266, 4315-4321.

18. Frech, M., Schlichting, I., Wittinghofer, A. & Chardin, P. (1990) Guanine nucleotide binding properties of the mammalian ralA protein produced in Escherichia coli, J. Biol. Chem. 265, 6353-6359.

19. Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227, 680-685.

20. Jaggi, R., Salmons, B., Muellener, D. & Groner, B. (1986) The v-mos and H-ras oncogene expression represses glucocorticoid hormone-dependent transcription from the mouse mammary tumor virus LTR, EMBO J. 5,2609-2616.

21. Vachon, L., Costa, T. & Herz, A. (1987) GTPase and adenylate cyclase desensitize at different rates in NG108-15 cells, Mol. Pharmacol. 31, 159-168.

22. Peterson, G. L. (1983) Determination of total protein, Methods Enzymol. 91, 95-119.

23. Towbin, H., Staehelin, T. & Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications, Proc. Nut1 Acad. Sci. USA 76,4350-4354.

24. Carney, W. P., DeLellis, R. A., Wolfe, H., Cooper, G. & Rabin, H. (1987) A biochemical analysis of carcinoma cells using a panel of monoclonal antibodies to normal and activated ras proteins, Proc. 78th Ann. Meet. Am. Assoc. Cancer Res. 28, 397.

25. Chesa, P. G., Rettig, W. J., Melamed, M. R., Old, L. J. & Niman, H. L. (1987) Expression of ~21"' in normal and malignant human tissues: lack of association with proliferation and ma- lignancy, Proc. Natl Acad. Sci. USA 84, 3234-3238.

26. Tucker, J., Sczakiel, G., Feuerstein, J., John, J., Goody, R. S. & Wittinghofer, A. (1986) Expression of p21 proteins in Escher- ichia coli and stereochemistry of the nucleotide-binding site, EMBO J. 5, 1351-1358.

27. Quinn, M. T., Parkos, C. A,, Walker, L., Orkin, S. H., Dinauer, M. C. & Jesaitis, A. J. (1989) Association of a ras-related protein with cytochrome b of human neutrophils, Nature 342, 198 -200.

28. Klinz, F.-J. (1 990) Development-dependent expression of low molecular mass GTP-binding proteins in embryonic chicken brain, Biochem. Biophys. Res. Commun. 172, 633-637.

29. Frey, J. & Afting, E.-G. (1983) Isolation and renaturation of a,-macroglobulin receptor from diploid human fibroblasts, Biochem. J. 214, 629-631.

30. O'Farrell, P. H. (1975) High resolution two-dimensional electro- phoresis of proteins, J. Biol. Chem. 250, 4007-4021.

gtp-blot analysis of small gtp-binding proteins : the c-terminus is involved in renaturation of...

Documents