SEARCH

SEARCH BY CITATION

Keywords:

  • bead-based assays;
  • protein–protein interaction;
  • Rho GTPases;
  • Rho GDI;
  • p21-activated kinase;
  • post-translational modification

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. Acknowledgements
  9. REFERENCES

Bead-based interaction assays are excellently suited to study protein–protein interactions, as they require only minimal amounts of sample material. Miniaturized protein–protein interaction assays were designed to analyze Rho GTPase activation based on its interaction with Rho GDI or p21-activated kinase (PAK).

Rho GDI plays a key role in the regulation of a variety of cellular functions through its interaction with Rho GTPases. Rho GDI is frequently overexpressed in many human cancers. Therefore, there is a growing and as yet unfulfilled demand for screening assays to identify biologically active compounds that may inhibit the Rho GTPase–Rho GDI interaction. Bead-based interaction assays provide an interesting alternative that facilitate such assays to be performed faster with only small amounts of material compared to routinely used co-immunoprecipitation followed by Western Blot analysis.

Bead-based protein interaction assays for overexpressed HA-tagged Rho GTPases were established to study the GTPγS-dependent interaction of five different Rho GTPases with the regulatory protein Rho GDIα and the downstream effector PAK1. In addition, it was demonstrated that the ability of Rho GTPases to interact with Rho GDI in this experimental system was markedly, but differentially sensitive to post-translational modification of their carboxyl terminus. Importantly, this modification also notably affected the ability of Rac1 and Rac2, but not of Cdc42, to interact with PAK1. Copyright © 2010 John Wiley & Sons, Ltd.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. Acknowledgements
  9. REFERENCES

Rho GTPases are members of the Ras superfamily of small GTPases (20–30 kDa) that act as binary molecular switches in a variety of signaling pathways. They are involved in the reorganization of the actin cytoskeleton, in the control of cell growth, and in the regulation of gene transcription, vesicle trafficking, and apoptosis. Small GTPases cycle between an inactive GDP-bound and an active GTP-bound state. The GTP-bound molecule is able to interact with its downstream effectors (Van Aelst and D'Souza-Schorey, 1997; Forget et al., 2002; Wennerberg and Der, 2004; Caruso et al., 2005; Ellenbroek and Collard, 2007) and the GDP-bound form interacts with the regulator Rho GDI. The C-terminal post-translational modifications of Rho GTPases, including the attachment of an isoprenoid moiety, are important for subcellular localization, membrane association, and effector regulation (Wennerberg and Der, 2004). The lipid modification occurs at the cysteine residue of a C-terminal CAAX motif (A = aliphatic, X = any amino acid), which is the recognition sequence for farnesyltransferase and geranylgeranyltransferase I. Following isoprenylation, the AAX part of the motif is proteolytically removed and the new C-terminus is carboxymethylated (Wennerberg et al., 2005).

The family of Rho proteins interacts with Rho GDI proteins (guanine nucleoside diphosphate dissociation inhibitors), which serve as key regulators of Rho proteins (Hoffman et al., 2000; Scheffzek et al., 2000; DerMardirossian and Bokoch, 2005; Dransart et al., 2005). Rho GDIs inhibit both GDP dissociation and GTP hydrolysis by binding to the N-terminal regulatory arm within the switch I and II regions of the GTPases. In addition, the prenyl moiety of the Rho GTPase inserts into a hydrophobic pocket within an immunoglobulin-like domain of the Rho GDIs, which leads to membrane release and thus the inactivation of the GTPase (Hori et al., 1991; Hancock and Hall, 1993; Keep et al., 1997; Hoffman et al., 2000). The complex between Rho GTPase and Rho GDI is stabilized by a magnesium ion (Mg2+) in the guanine nucleotide binding pocket (Scheffzek et al., 2000; Shimizu et al., 2000).

The p21-activated kinase 1 (PAK1) is a downstream effector of the Rho GTPases Rac and Cdc42 that is involved in cytoskeletal remodeling and cell motility. PAK1 serine/threonine kinase activity and subcellular distribution are tightly regulated by Rac and Cdc42 and by the interaction with several adaptor proteins (Knaus and Bokoch, 1998; Stofega et al., 2004; Gururaj et al., 2005). GTPase binding disrupts an autoinhibitory interaction between the regulatory and catalytic domain of PAK1, resulting in PAK1 activation and phosphorylation of target proteins (Bagrodia and Cerione, 1999; Bishop and Hall, 2000).

Inactivation of Rho GTPases occurs after infection with pathogenic Yersinia strains such as Yersinia enterocolitica, Yersinia pseudotuberculosis, and Yersinia pestis. These Gram-negative bacteria inject effector proteins into their eukaryotic target cells (Cornelis, 2002a, b). One of these effectors, YopT, a cysteine protease, cleaves the isoprenyl moiety of the Rho GTPases directly at the C-terminal cysteine. This leads to the release of the active GTPases from the membrane, and the resultant inactivation disrupts the actin cytoskeleton of the target cells and leads to cell rounding (Shao et al., 2002; Aepfelbacher et al., 2003; Shao et al., 2003; Fueller et al., 2006; Fueller and Schmidt, 2008; Shao, 2008).

The different protein interactions of Rho GTPases with activating and inactivating binding partners can be analyzed using the bead-based xMAP technology (Luminex Corp., Austin, Texas). Appropriate capture molecules were immobilized on color-coded microspheres and incubated with the interaction partners Rho GDI or p21-activated kinase–CRIB domain (PAK-CD). Bound interaction partners were visualized using a fluorescence-based reporter system.

The bead-based xMAP technology used in this study allows the analysis of the activity of Rho GTPases with minimal amount of sample material in a time- and cost-saving manner. Adaption of the assays to 96-well filter plates allows an easy bead handling and removal of unbound sample. Commercially available Rho GTPase activation assay kits utilize the PAK Binding Domain immobilized on a resin to selectively isolate and pull-down the active form of GTPase from dozens to hundreds of µg's of purified samples or cell lysates. Read-out of the precipitated GTPase is usually performed by time-consuming SDS-Page and Western Blot analysis. Compared to this, using a 96-well plate format for the bead-based setup allows a better sample throughput, which might be interesting for the screening of biologically active compounds in a parallel manner.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. Acknowledgements
  9. REFERENCES

Materials

GTPγS was purchased from Jena Bioscience GmbH (Jena, Germany), GDP from Sigma-Aldrich (Hamburg, Germany). R-phycoerythrin-conjugated goat anti-GST antibody was purchased from ProZyme Inc. (San Leandro, CA, USA) and the rat anti-HA antibody from Roche Diagnostics GmbH (Mannheim, Germany).

Expression and purification of recombinant proteins

The HA-tagged Rho GTPases were expressed in Sf9 cells with recombinant baculoviruses (Cerione et al., 1995). Cells were pelleted at 250 × g, resuspended in wash buffer (100 mM NaCl, 20 mM Tris pH 8.0, 3.75 mM MgCl2, 1 mM EDTA, 1 mM DTT, 3 µM GDP, 0.1 mM PMSF, 1 µg/mL aprotinin, and 1 µg/mL leupeptin) and lysed by vortexing and drawing the lysate 10x up and down with a syringe through an 18 gauge needle. Cell debris was removed by centrifugation at 300 × g for 10 min. The supernatant was centrifuged for another 15 min at 12 000 × g and indicated as cytosol in the following part. The membrane pellet was resuspended in cholate-extraction buffer (100 mM NaCl, 20 mM Tris pH 8.0, 3.75 mM MgCl2, 1 mM EDTA, 1 mM DTT, 3 µM GDP, 0.1 mM PMSF, 1 µg/mL aprotinin, and 1 µg/mL leupeptin, 1% (w/v) Na-cholate) and incubated for 90 min. Membrane debris was removed by centrifugation at 12 000 × g for 15 min. The supernatant is indicated as membrane extract. All incubation steps were performed on ice and centrifugation at 4°C.

GST–Rho GDIα and GST–PAK/CRIB (where CRIB means Cdc42/Rac interactive binding) were expressed in E. coli strain BL21 (DE3). A crude extract was used for the experiments.

The GST-YopT expression plasmid was provided by the group of Jürgen Heesemann (Max von Pettenkofer-Institut für Medizinische Mikrobiologie, LMU Munich). YopT was expressed in E. coli strain BL21 (DE3). The strain containing the corresponding plasmid was grown in LB media with 100 µg/mL ampicillin to a density of 0.6 (A600). Protein expression was induced with 0.4 mM isopropyl-β-D-thiogalactopyranoside (IPTG) for 6 h at 28°C. Cells were harvested and lysed by incubating them for 30 min in PBS (phosphate-buffered saline) with 1% (v/v) Triton X-100 and 0.4 mg/mL lysozyme and subsequent sonication. The GST fusion protein was purified batchwise by affinity precipitation using glutathione sepharose. The GST-tag was cleaved by incubation with thrombin in PBS, 0.1% (v/v) Triton X-100, and 2.5 mM CaCl2 for 4 h at room temperature. Expression, purification, and thrombin cleavage were monitored using SDS-PAGE (NuPage®, Invitrogen, Carlsbad, CA, USA).

Coupling of antibodies to carboxylated Luminex microspheres

Anti-HA antibodies were coupled to xMAPTM-carboxylated Luminex microspheres using a carbodiimide method. Two hundred µL of bead suspension (corresponds to 2.5 × 106 beads) were washed with activation buffer (sodium phosphate, pH 6.2) and incubated for 20 min with 50 mg/mL EDC (N-3-(dimethylaminopropyl)-N'-ethylcarbodiimide) and 50 mg/mL sulfo-NHS (sulfo-N-hydroxysuccinimide) on a shaker in the dark. The beads were incubated with 200 µg/mL anti-HA antibody in 2-N-morpholino-ethansulfonic acid (MES), pH 5.0, for 2 h in the dark. After washing with PBS/0.05% (v/v) Tween 20, pH 7.4, the beads were collected in PBS with 1% (w/v) BSA and 0.05% (w/v) sodium azide and stored at 4°C in the dark.

Luminex assays

The Luminex assays were performed in 96-well microtiter plates (Millipore, Billerica, MA, USA) with a final reaction volume of 80 µL. All incubations were performed at 22°C and 650 rpm on a shaker (Thermomixer comfort, Eppendorf, Hamburg, Germany) unless otherwise specified.

Rho GTPase interaction assays

Beads with immobilized anti-HA antibody were incubated for 20 min in a 96-half well plate with 50 µg/mL of either membrane detergent extract or cytosolic protein of baculovirus-infected insect cells expressing HA-tagged Rho GTPases. Intrinsically bound GDP was replaced by increasing concentrations of GTPγS in exchange buffer (100 mM NaCl, 20 mM Tris, pH 8.0, 2 mM EDTA, 1 mM MgCl2, 1% (w/v) BSA, 30 µM GDP, 50 µg/mL aprotinin, 1 µg/mL leupeptin, 0.1 mM PMSF) in the case of the GTPase–Rho GDI interaction assay and in Rac Ripa buffer (150 mM NaCl, 50 mM Tris, pH 7.5, 2 mM EDTA, 1 mM MgCl2, 1% (w/v) BSA, 3 µM GDP, 0.2% (v/v) Triton X-100, 0.2% (v/v) Na-cholate, 50 µg/mL aprotinin, 1 µg/mL leupeptin, 0.1 mM PMSF) for the GTPase–PAK interaction assay. The guanine nucleotide exchange was stopped by addition of 5 mM MgCl2, followed by a further 30 min incubation. Beads and bound protein were transferred to a pre-wet filter plate (2000 beads per well) and unbound protein was removed by washing three times with 100 µL wash buffer (100 mM NaCl, 20 mM Tris, pH 8.0, 3.75 mM MgCl2, 1 mM EDTA, and 1% (w/v) BSA). Complex formation was initiated by adding 4 µg/mL GST–Rho GDIα in complex buffer (100 mM NaCl, 20 mM Tris, pH 8.0, 3.75 mM MgCl2, 1 mM EDTA, 50 µg/mL aprotinin, 1 µg/mL leupeptin, 0.1 mM PMSF, and 1% (w/v) BSA) or 4 µg/mL GST–PAK-CD in lysis buffer (150 mM NaCl, 50 mM Tris, pH 7.5, 5 mM MgCl2, 0.2 (v/v) % Triton X-100, 1% (w/v) BSA, 50 µg/mL aprotinin, 1 µg/mL leupeptin, and 0.1 mM PMSF), respectively. The incubation was performed for 1.5 h at 22°C. After washing twice with 100 µL wash buffer, 30 µL of R-phycoerythrin-conjugated goat anti-GST antibody (2.5 µg/mL) was added. After 1.5 h of incubation, unbound detection antibody was removed by washing twice with 100 µL wash buffer. The beads were resuspended in 100 µL wash buffer and the samples were subsequently analyzed using a Luminex 100 system (Luminex Corp., Austin, TX, USA). The parameters were set to 60 s detection time, 80 µL sample volume, and 100 events per bead sort.

YopT cleavage assay

The YopT cleavage assay was carried out as described in the Rho GTPase–Rho GDI interaction assay, except that 25 µg/mL of the Rho GTPases were incubated with 5 µL of purified YopT for 30 min at 37°C before the mixture was added to the beads.

Statistical analysis

The data was fit by non-linear least squares curve fitting to a three parameter variant of the four parameter logistic function established for the analysis of families of sigmoidal curves (DeLean et al., 1978) using GraphPad Prism, version 4.03 (GraphPad Software, San Diego California USA). The global curve fitting procedure contained in GraphPad Prism was used to determine whether the best-fit values of selected parameters differed between data sets. The simpler model was selected unless the extra sum of squares F-test had a p value less than 0.05.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. Acknowledgements
  9. REFERENCES

A bead-based protein–protein interaction assay was developed to investigate the interaction between five different Rho GTPases, RhoA, RhoC, Rac1, Rac2, and Cdc42 (isoform 2, NP 426359) and the regulator protein Rho GDIα or the Rho GTPase binding domain of the downstream effector PAK-CD, respectively.

Post-translational modification of the Rho GTPase carboxyl terminus differentially affects their interaction with Rho GDIα

The isoprenoid moiety of Rho GTPases is crucial for the interaction with Rho GDI. The geranylgeranyl moiety inserts into a hydrophobic pocket within an immunoglobulin-like domain of the Rho GDI, leading to membrane release of the GTPase (; Hancock and Hall, 1993; Keep et al., 1997; Hoffman et al., 2000; Anderson et al., 2004).

The influence of C-terminal post-translational modification of the five Rho GTPases RhoA, RhoC, Rac1, Rac2, and Cdc42 on the Rho GTPase–Rho GDI interaction was determined using recombinant expressed proteins. The modified forms of the GTPases are mainly present in the membrane fraction of the cell lysates, whereas the non-modified GTPases are mainly found in the cytosolic fraction. HA-tagged GTPases present in membrane extracts or in the cytosolic fraction were immobilized on beads carrying a rat anti-HA antibody. The immobilized GTPases were then incubated with increasing concentrations of GTPγS. Complex formation was initiated by adding GST–Rho GDIα and bound GST–Rho GDIα was visualized using an R-phycoerythrin-conjugated goat anti-GST antibody (Figure 1, a).

thumbnail image

Figure 1. Schematic representation of the Rho GTPase interaction assays. In a singleplex approach, five HA-tagged Rho GTPases from cytosols or membrane extracts of Sf9 cells were captured onto rat-α-HA-coated beads and competitive exchange of GTPγS for GDP was performed with increasing concentrations of GTPγS. GST–RhoGDIα (a) or GST–PAK-CD (c) was added and the formed heterodimeric complex was detected with goat-α-GST-PE. YopT treatment leads to the release of the post-translational modification (b) and influences the interaction with GST–RhoGDIα (1) or GST–PAK-CD. The readout was done using a Luminex 100 system.

Download figure to PowerPoint

Consistent with the known reduction of the affinities of GTP-bound Rho GTPases to Rho GDIs when compared to GDP-bound Rho GTPases (Sasaki et al., 1993; Robbe et al., 2003) all modified forms of the Rho GTPase present in the membrane extracts revealed an interaction with Rho GDIα that was markedly sensitive to preincubation of the Rho GTPases with GTPγS. Thus, GTPγS—concentration-dependent—caused a signal loss of 80–90% for all Rho GTPases. There were appreciable differences in the IC50 values of GTPγS, which followed the order RhoC (0.36 µM) < RhoA (0.48 µM) < Rac1 (0.78 µM) < Rac2 (1.1 µM) < Cdc42 (4.0 µM). Note that these concentrations are influenced by the concentrations of free and Rho GTPase-bound GDP present in the incubation medium. As expected, all five non-modified Rho GTPases showed a weaker interaction with Rho GDIα than their modified counterparts. Most interestingly, however, there were marked differences between the five GTPases to do so. Specifically, while non-modified Cdc42 and Rac2 showed reductions of Rho GDIα interaction by approximately 94 and 85%, respectively, these reductions were only 72, 63, and 47, for RhoA, Rac1, and RhoC. As a result, the stimulatory influence of C-terminal Rho GTPase modification on the Rho–Rho GDIα interaction ranges from > 10-fold (Cdc42) to < 2-fold (RhoC). This indicates that the ability of Rho GTPases to interact with Rho GDIα in this experimental system was differentially sensitive to post-translational modification of their carboxyl terminus. Note that Rho GDIα interaction was sensitive to GTPγS in all cases. There were no statistically significant differences in the IC50 values of GTPγS between modified and non-modified Rho GTPases (Figure 2). The corresponding p values were: Cdc42, 0.25; RhoA, 0.16; RhoC, 0.91; Rac1, 0.05; and Rac2, 0.37.

thumbnail image

Figure 2. Rho GTPase–Rho GDI interaction assay. Panels A–E (A) Cdc42, (B) RhoA, (C) RhoC, (D) Rac1, and (E) Rac2) show the GTPγS-dependent interaction of Rho GTPases with Rho GDIα. Post-translationally modified GTPases (full circles) interact with Rho GDIα to a considerably higher extent than non-modified GTPases (open circles). The differences between the maximal fluorescence intensities obtained for the non-modified versus modified Rho GTPases in the absence of GTPγS were statistically significant at p < 0.0001 in all five cases. Each experiment was performed twice with similar results (intra-assay CV 8–15%, inter-assay CV 17–26%). Means of duplicate determinations ± SEM are presented.

Download figure to PowerPoint

In additional experiments (results not shown), we found that the abundance of the Rho GTPases was similar in the five pairs of detergent extract and cytosolic protein preparations and that the different abilities of non-modified Rho GTPases to interact with Rho GDIα did not correlate with the slight differences observed for their abundance in these preparations. Furthermore, increasing the quantities of cytosolic Rho GTPases during the preincubation did not increase the median fluorescence intensity (data not shown).

Treatment of Rho GTPases with YopT differentially affects the Rho GTPase-Rho GDIα interaction

Rho GTPases post-translationally modified at their C-terminus have previously been shown to be differentially sensitive [RhoA > Rac1 > Cdc42 (isoform 1)] to proteolytic cleavage upstream of the isoprenylated, terminal cysteine residue by the cysteine protease YopT (Fueller and Schmidt, 2008). Therefore, the five Rho GTPases examined in this study were treated with recombinant YopT and used to perform the Rho GTPase–Rho GDIα interaction assay (Figure 1, b). Figure 3 shows that YopT treatment led to about 60% loss of median fluorescence intensity in the case of RhoA, followed by 50% for RhoC, 40% for Rac1, 30% for Rac2, and only about 15% for Cdc42 (isoform 2) compared to non-treated GTPases. These results show the differential affects of YopT on Rho GTPases which is in accordance with results reported by Fueller and Schmidt, 2008. In addition, the loss of the isoprenyl moiety of a distinct Rho GTPase leads to the decrease in relative binding affinity to Rho GDI and supports the findings described in the previous section.

thumbnail image

Figure 3. YopT cleavage assay. The experiment was performed as described in figure 1 except that the Rho GTPases were incubated with YopT prior to Rho GDI binding. Signal decrease of YopT-treated compared to non-treated GTPases is about 60% for RhoA, 50% for RhoC, 40% for Rac1, 30% for Rac2, and only about 15% for Cdc42. The graphs are the mean result ± SEM of four independent experiments done in duplicate. The percentage of control response determined in the presence of 30 µM GDP and absence of YopT is depicted.

Download figure to PowerPoint

PAK-CD interaction with Rho GTPases depends on the post-translational modification state of the GTPase

The C-terminal modification(s) of Rho GTPases enhance(s) their interaction with Rho GDIα (Figures 2 and 3). Therefore, it was of great interest to see whether the modifications also play a role in the interaction of the Rho GTPases with their binding site present on their downstream effector PAK1. To this end, HA-tagged Rho GTPases from either membrane extracts or cytosols were captured onto anti-HA-coated beads and activated by incubation with increasing concentrations of GTPγS. Subsequently, GST–PAK-CD was added to the beads and the formed heterodimeric complexes were visualized using an R-phycoerythrin-labeled detection antibody directed against the GST-tag (Figure 1, c).

Figure 4 shows that the interaction of Rho GTPases with their downstream target site PAK-CD can in fact be detected and quantified using the experimental approach described here. Specifically, incubation of beads harboring Cdc42 with GTPγS led to a concentration-dependent, very robust, i.e., more than 20-fold signal increase. Half-maximal effects were observed at approximately 40 nM GTPγS. Again, free and Rho GTPase-bound GDP may have affected these concentrations. Identical GTPγS-concentration-response curves were obtained for Cdc42 from detergent extract and cytosolic preparations. Very interestingly and quite unexpectedly, however, the responses observed upon GTPγS-mediated activation of both Rac1 and Rac2 from these two preparations were clearly distinct. In both cases, the GTPγS-activated Rac GTPase from the cytosolic preparation caused an about two-fold higher signal increase than its detergent extract counterpart. Furthermore, the stimulatory effects of both cytosolic Rac GTPases was observed at lower GTPγS concentrations (Rac1, 25 nM; Rac2, 120 nM) than the effects of detergent extracted, membranous Rac (Rac1, 50 nM; Rac2, 770 nM). The Rho family member GTPases did not interact with PAK-CD at all (data not shown) (Knaus and Bokoch, 1998; Bagrodia and Cerione, 1999; Bishop and Hall, 2000).

thumbnail image

Figure 4. Rho GTPase-PAK-CD interaction assay. The post-translationally modified (full circles) as well as the non-modified (open circles) forms of Cdc42 (A), Rac1 (B), and Rac2 (C) interact with PAK-CD in a GTPγS-dependent manner. Post-translationally modified Rac (full circles) interact with PAK-CD to a considerably higher extent than non-modified Rac (open circles). No difference was detected for Cdc42 (isoform2). The figures show means of duplicate determinations ± SEM of one representative experiment (intra-assay CV 5–8%, inter-assay CV 13–26%).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. Acknowledgements
  9. REFERENCES

In this study, a bead-based, miniaturized assay was developed to investigate the activities of Rho GTPases by assessing the protein–protein interactions of the five Rho GTPases RhoA, RhoC, Rac1, Rac2, and Cdc42 with their regulator and effector proteins Rho GDIα and the CRIB domain of PAK1. The influence of the C-terminal post-translational modification of the Rho GTPases on these interactions was investigated. The isoprenoid moiety of Rho GTPases is crucial for their interaction with Rho GDIs. The latter proteins remove the Rho GTPases from the membrane by inserting the geranylgeranyl moiety into a hydrophobic pocket within an immunoglobulin-like domain (Hori et al., 1991; Hancock and Hall, 1993; Keep et al., 1997; Hoffman et al., 2000).

For the interaction assay, the five different HA-tagged Rho GTPases were captured onto anti-HA beads and incubated with increasing concentrations of GTPγS. The GTPase-Rho GDIα complex was formed by adding GST–Rho GDIα and binding was visualized using a PE-labeled anti-GST antibody and a Luminex 100 system.

The expected GTPγS-dependent interaction of the five different Rho GTPases with Rho GDIα occurred (Sasaki et al., 1993). In the assay setup used, a 20 min incubation of GTPγS was sufficient to replace the intrinsically bound GDP nucleotides. This result is consistent with previous findings (Kikuchi et al., 1992; Leonard et al., 1992).

A considerably lower maximum signal was observed in the absence of GTPγS with the cytosolic fractions, containing non-modified prenylated Rho GTPases, compared to detergent extracts of the membrane fractions, containing modified Rho GTPases. Depending on the Rho GTPase, the signal loss was 50–90%. This result highlights the importance of the post-translational modification of Rho GTPases on the interaction with Rho GDI and is, therefore, consistent with previously published data (Hori et al., 1991). However, our findings also support previous findings suggesting that C-terminal post-translational modification of Rho GTPases is not generally required for a high affinity Rho GTPase–Rho GDI interaction (Faure and Dagher, 2001). The results presented here add RhoC and Rac1 to the Rho GTPases that, together with RhoA (Faure and Dagher, 2001) appear to interact with Rho GDIs to a readily appreciable extent even in the absence of such a modification. The remaining signal for Rac1 and RhoC might be also due to higher amounts of modified Rho GTPases in the cytosolic fraction.

It is known that Cdc42 interacts with the CRIB domain of PAK1 via the switch I and switch II regions and some N-terminal helix and β-strand structures (Morreale et al., 2000). The finding that PAK-CD apparently discriminates between modified and unmodified Rac1 and Rac2, but not of Cdc42, is unprecedented and intriguing. However, neither the carboxyl-terminal-most 7 residues of Cdc42 (isoform1) nor the polybasic region immediately upstream of this sequence were required for the interaction of the activated GTPase with PAK-CD (Li et al., 1999). In contrast, the polybasic regions of Rac1 and Rac2 have previously been shown to be critical determinants of PAK1 binding and activation (Knaus et al., 1998). Specifically, Rac1 bound to and stimulated the protein kinase activity of PAK1 approximately 2- and 4–5-fold, respectively, better than Rac2. Mutation of these six basic residues in Rac1 to neutral amino acids dramatically decreased the ability of Rac1 to bind PAK1 and almost completely abolished its ability to stimulate PAK activity. Moreover, replacing the highly charged polybasic domain of Rac1 with the less charged domain of Rac2 (and vice versa) completely reversed the PAK binding/activation properties of the two Rac isomers. While the former finding is consistent with the different IC50 values of GTPγS to results shown in Figure 4(B) and (C), the latter findings suggest that the polybasic domain differences account for the disparate abilities of Rac1 and Rac2 to bind to and activate PAK1. Isoprenylation of H-Ras has previously been shown to perturb several residues within the three-dimensional structure of the protein, including the extreme C-terminus of the mature protein immediately upstream of the isoprenylated cysteine residue (Thapar et al., 2004). Likewise, geranylgeranylation of purified Rab3A led to a time-dependent loss in the ability of the protein to bind [35S]GTPγS (Musha et al., 1992), and isoprenylation of the large hepatitis delta antigen altered the conformation of an epitope present at least 15 amino acids upstream of the isoprenylated cysteine residue (Hwang and Lai, 1993; Hwang and Lai, 1994). It thus appears possible that post-translational modification of the Rac1 and Rac2 carboxyl termini affects the structures of the polybasic regions immediately adjacent to the very C-terminus and that these changes modify the abilities of the two proteins to interact with PAK-CD. The high variability of PAK autophosphorylation used to monitor PAK activity and, hence, high variability of the results may have precluded detection of these differences in previous work (Kreck et al., 1996).

Since there are no suitable antibodies against Rho GTPases commercially available we used the HA-fusion proteins to capture the different Rho GTPases. Using tag-specific capture antibodies, rather than Rho GTPase specific antibodies, seems to be advantageous, since Rho GTPases are small proteins containing numerous functional regions responsible for the protein conformation and single amino acids are needed for the interaction with PAK-CD and Rho GDI (Morreale et al., 2000; Grizot et al., 2001).

The bead-based protein interaction assays enable the analysis of activity of HA-tagged overexpressed Rho GTPases with minimal amount of sample in a time- and cost-saving manner. The assay is already adapted to the 96-well microtiter plate format, which easily increases sample throughput capabilities compared with co-immunoprecipitation followed by Western Blot analysis, which are also offered as commercially available kits for the measurement of the activity of small GTPases. Usually the respective binding domain of the downstream effector for each small GTPase is expressed as a GST-fusion protein as it was used in the bead-based experiments described above. An alternative for the measurement of Rho GTPase activity is described in Caruso et al., 2005. Caruso et al. used GSH-coated 96-well FlashPlates (Perkin Elmer) to capture GST–GTPases. However, the assay was designed to measure the [γ-35S]GTPγS hydrolysis activity. The described bead-based assay setup allows working with non-radioactive substances.

Such interaction assays can be implemented into screening processes for the identification of biologically active compounds interfering with Rho GTPase-mediated signaling, both at the level of Rho GTPase activation and Rho GTPase effector interaction.

Multiplexing these protein–protein interaction assays will further reduce the amount of sample required and, furthermore, allow determination of activation and effector interaction of several Rho GTPases as well as testing them in their non-modified and C-terminally modified form in a single sample. This needs further development efforts.

Applying the assay described here in the latter format for drug screening purposes appears a most intriguing and appealing possibility, since it would allow, at least in certain cases, to discriminate between compounds addressing distinct sites on the different Rho GTPases, their regulatory proteins, and their effectors in one experimental setup.

CONFLICT OF INTEREST

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. Acknowledgements
  9. REFERENCES

Dr Thomas O. Joos is a member of the scientific advisory board of Luminex Corp., Austin, TX, USA and of the scientific advisory board of Rules Based Medicine Inc., Austin, TX, USA.

The authors declare that all relevant relationships have been disclosed.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. Acknowledgements
  9. REFERENCES

This work was supported by the Deutsche Forschungsgemeinschaft (German Research Foundation, DFG)/JO 687/2-1 and GI138/5-1.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. Acknowledgements
  9. REFERENCES