Rho GDP dissociation inhibitor (RhoGDI) plays an essential role in control of a variety of cellular functions through interactions with Rho family GTPases, including Rac1, Cdc42, and RhoA. RhoGDI is frequently overexpressed in human tumors and chemoresistant cancer cell lines, raising the possibility that RhoGDI might play a role in the development of drug resistance in cancer cells. We found that overexpression of RhoGDI increased resistance of cancer cells (MDA-MB-231 human breast cancer cells and JLP-119 lymphoma cells) to the induction of apoptosis by two chemotherapeutic agents: etoposide and doxorubicin. Conversely, silencing of RhoGDI expression by DNA vector–mediated RNA interference (small interfering RNA) sensitized MDA-MB-231 cells to drug-induced apoptosis. Resistance to apoptosis was restored by reintroduction of RhoGDI protein expression. The mechanism for the antiapoptotic activity of RhoGDI may derive from its ability to inhibit caspase-mediated cleavage of Rac1 GTPase, which is required for maximal apoptosis to occur in response to cytotoxic drugs. Taken together, the data show that RhoGDI is an antiapoptotic molecule that mediates cellular resistance to these chemotherapy agents.
The Rho GDP dissociation inhibitor (RhoGDI) is a cellular regulatory protein that acts primarily by controlling the cellular distribution and activity of Rho GTPases ( 1– 4). It is a member of a family of GDIs that include RhoGDI (also known as RhoGDIα or RhoGDI-1), D4-GDI, and RhoGDI-3. RhoGDI is ubiquitously expressed in cells and tissues and binds to most of the Rho GTPases thus far examined, including Rac1, RhoA, and Cdc42 ( 5). It has been postulated that RhoGDI can negatively regulate Rho proteins in three ways: (a) by shielding the membrane-anchoring domain of the GTPases, thereby restricting them to a cytosolic (nonactive) localization; (b) by blocking their interaction with guanine nucleotide exchange factors (GEF), thereby inhibiting GTPase activation; and (c) by blocking binding to downstream target molecules (e.g., various kinases regulated by small GTPases). Thus, overexpression of RhoGDI in various cell lines induces disruption of the actin cytoskeleton and loss of substratum adherence, and microinjection of RhoGDI into fibroblasts inhibits cell motility ( 6– 8). In addition, transgenic mice lacking RhoGDI have impaired development of the kidneys and reproductive organs ( 9). Cardiac-specific overexpression of RhoGDI disrupts cardiac morphogenesis and inhibits cardiomyocyte proliferation, leading to embryonic lethality ( 10). Dictyostelium species lacking GDI-1, a RhoGDI homologue, are multinucleate and display a pinocytosis defect ( 11).
RhoGDI is overexpressed in multiple types of human cancers. Proteomic results revealed the high degree of overexpression of RhoGDI in invasive ovarian cancers when compared with low malignant tumors or normal tissues (∼4-fold; ref. 12). The elevation of RhoGDI is also detected in human breast cancer cell lines and breast tumor tissues ( 13– 15). However, little is known about whether or how elevated RhoGDI promotes the cancer phenotype. Recent studies showed that RhoGDI is highly expressed in stable chemoresistant cancer cell lines, such as melanoma ( 16, 17) and ovarian cells ( 18), raising the possibility that RhoGDI might play a role in the development of drug resistance of cancer cells.
The goal of these studies was to determine whether RhoGDI modulates chemotherapy-induced apoptosis in cancer cells. Our previous research showed that caspase-mediated cleavage of Rac1 is required to achieve maximal apoptosis in response to cytotoxic drugs ( 19) and that native Rac1 inhibits apoptosis by regulating Bad kinase ( 20). Thus, we postulated that RhoGDI, which binds to and controls Rac1 function, might influence the apoptotic process as well.
We report here that RhoGDI is an antiapoptotic molecule that promotes the resistance of cancer cells to drug-induced toxicity. Overexpression of RhoGDI in human breast cancer and lymphoma cells strongly inhibited apoptosis induced by two different chemotherapeutic agents [etoposide (VP-16) and doxorubicin]. Using DNA vector–mediated RNA interference (RNAi), we established a stable MDA-MB-231 breast cancer cell line with persistently suppressed RhoGDI expression. RhoGDI knockdown cells were more susceptible to drug-induced apoptosis, and restoration of RhoGDI expression to the levels found in wild-type MDA-MB-231 cells restored the drug-resistant phenotype. In addition, we provide several lines of evidence that RhoGDI inhibits caspase-mediated cleavage of Rac1 GTPase. These results reveal a new activity for RhoGDI and a novel mechanism for the positive regulation of Rac1 function: shielding of the GTPase from apoptotic cleavage, thereby maintaining Rac1 in the intact, functional state. Expression of the caspase-3-resistant Rac1 mutant (Rac1D11E) inhibited VP-16-induced apoptosis in RhoGDI-depleted cells, suggesting that RhoGDI-mediated inhibition of apoptosis may occur in part through protection of Rac1 from caspase cleavage.
Materials and Methods
Expression vectors and reagents. The pcDNA3.1(+)-RhoGDI plasmid encoding wild-type human RhoGDI (RhoGDIα) protein was provided by Guthrie Research Institute (Sayre, PA). For expression of glutathione S-transferase (GST)–fused RhoGDI, the insert was excised and subcloned into pGEX-KG vector. The pEGFP-C3/RhoGDI vector expressing green fluorescent protein (GFP)–tagged RhoGDI was kindly provided by Dr. Mark R. Philips (New York University School of Medicine, New York, NY). The pGEX-KG-PAK1 vector, encoding amino acids 51 to 135 of human PAK1 fused to GST (GST-PBD), was described previously ( 19). For expression in mammalian cells, wild-type Rac1 was subcloned into a pcDNA4/HisMax-TOPO (Invitrogen, Carlsbad, CA) vector to be expressed as NH2-terminal His6-tagged proteins.
The small interfering RNA (siRNA) expression plasmids were custom constructed from GeneScript Corp. (Scotch Plains, NJ) using a mammalian expression vector, pRNA-U6.1/hygromycin, which contains a DNA template for the synthesis of siRNA under the control of the U6 promoter. The 19-mer candidate target sequences were selected from the open reading frame of human Rho-GDI, corresponding to nucleotides 390 to 408, 401 to 441, and 413 to 431 (relative to the start codon) of RhoGDI. Each hairpin siRNA sequence contains a 5′ BamHI cloning site followed by 19-mer target sequences, a 9-mer spacer, another 19-mer reverse complementary target sequence, the transcription terminator (TTTTT), and the HindIII (3′) cloning site. The full-length of the sequence is 70-nucleotide oligonucleotides, which were synthesized in the forward and reverse directions and annealed to form dsDNA. This dsDNA was cloned into pRNA-U6.1 vector to form pRNA-U6/siRhoGDI. Figure 2A summarizes the siRNA constructs used in this study. For the knock-in studies, the RhoGDI gene was mutated (403AAAGGCGTCAAGATTGAC420 to 403AAGGGAGTAAAAATCGAT420) to prevent destruction of exogenous mRNA by the corresponding siRNA, leaving the amino acid sequence unchanged. Mutations were introduced using the Quick Change Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA) according to the manufacturer's protocol. The resulting RhoGDI mutant was excised and subcloned into pEGFP-C3/kanamycin vector at BamHI and EcoRI sites, fusing the enhanced GFP (EGFP) to the NH2 terminus of RhoGDI construct. All constructs were verified by DNA sequencing.
Antibodies used were the following: polyclonal anti-Rac1 (C-11) and anti-RhoGDI from Santa Cruz Biotechnology (Santa Cruz, CA); anti-caspase-3 from Calbiochem (San Diego, CA); monoclonal anti-Rac1 and anti-RhoGDI from BD Transduction Laboratories (Lexington, KY); and monoclonal anti–poly(ADP-ribose) polymerase (PARP) from Oncogene Research Products (Boston, MA). Recombinant active caspase-3 enzyme was purchased from BD Biosciences (Lexington, KY). VP-16 and staurosporine were from Sigma (St. Louis, MO).
Cell culture and transfection. MDA-MB-231 human breast cancer cells were grown in DMEM/F-12 (1:1 mix) (Mediatech, Inc., Herndon, VA) supplemented with 5% fetal bovine serum (FBS), 4 mmol/L l-glutamine, 50 μmol/L β-mercaptoethanol, and 1 mmol/L sodium pyruvate at 37°C and 5% CO2. For transient transfections, cells (∼90% confluent) were grown onto 25 cm2 flasks and transfected with 8.0 μg pEGFP-C3/RhoGDI by LipofectAMINE 2000 (Invitrogen). After 48 hours, the cells were harvested, washed twice in PBS with 30% FBS, and filtered through a 40 μm nylon mesh. EGFP-positive cells (∼20%) were isolated by fluorescence-activated cell sorting (FACS) and maintained in complete medium supplemented with 1.5 mg/mL kanamycin. For stable transfection of siRNA expression vectors, MDA-MB-231 cells (50-60% confluent) were grown in six-well plates, and either 1.6 μg of pRNA-U6.1/siRhoGDI or a control plasmid pRNA-U6.1/siLuciferase was introduced using LipofectAMINE 2000 according to the manufacturer's instructions. After 72 hours, 0.45 mg/mL hygromycin (Clontech, Mountain View, CA) was added to the cultures to select for hygromycin-resistant clones. Two weeks later, independent colonies were picked using cloning cylinders (Labcor Products, Inc., Frederick, MD), subcultured, and tested for expression of RhoGDI by immunoblot analysis with antibodies against human RhoGDI as described below. The selected stable clones with decreased levels of RhoGDI (designated as MDA-MB-231−/−RhoGDI) were maintained in complete culture medium containing hygromycin (0.45 mg/mL). Transfection of Burkitt's lymphoma JLP-119 cells with pcDNA3.1(+)-RhoGDI was carried out by electroporation as described previously ( 19). Stable clones were selected with G418 and maintained in RPMI 1640 containing 10% FCS, 2 mmol/L l-glutamine, 50 μmol/L β-mercaptoethanol, and 1.5 mg/mL G418.
Cytotoxicity assays. Cells were induced to undergo apoptosis using different agents. Stock solutions of VP-16 were prepared in DMSO and stored at −20°C. Doxorubicin was prepared in water and stored at 4°C. The drugs were diluted in PBS and added to the culture medium to achieve the desired concentrations and were not washed out. Where indicated, the caspase inhibitor Z-VAD (ICN Biochemicals, Aurora, OH) was added 1 hour before addition of VP-16. The apoptotic cells were assessed by FACS analysis using FITC-Annexin V and propidium iodide (PI) or by nuclear morphology with Hoechst 33342 and PI and fluorescence microscopy as described previously ( 19, 20). Cell viability was quantified using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay. Briefly, cells were seeded in a 96-well plate (4 × 104/mL) overnight and treated with VP-16 or doxorubicin for different times. MTT (Sigma) solution (20 μL of 5 mg/mL solution) was added to each well and incubated for 4 hours at 37°C. The supernatant was removed, and the MTT-formazan crystals formed by metabolically viable cells were dissolved in 200 μL Me2SO. Absorbance at 570 nm was monitored using a microplate reader.
Protein expression and purification. The post-translationally modified form of Rac1 was purified from the membrane fraction of COS-1 cells transfected with pcDNA4/His-Rac1 encoding His6-Rac1. Heterodimer complexes composed of RhoGDI and Rac1 (1:1) were produced by coinfecting Spodoptera frugiperda (Sf9) cells with two baculoviruses encoding wild-type Rac1 and COOH-terminal His-tagged RhoGDI, and the soluble complex was purified by sequential chromatography on Superdex 75 and Mono Q columns ( 21). GST-RhoGDI and GST-PBD were expressed in Escherichia coli DH5α cells and purified by glutathione-coupled agarose chromatography as described ( 22). The quality of the proteins was judged by SDS-PAGE. Protein concentrations were determined from Coomassie blue–stained gels and by using the BCA protein assay (Pierce, Rockford, IL).
Rac1 cleavage and functional analysis. The cleavage of Rac1 GTPase was determined as described previously ( 19). For in vitro Rac1 cleavage, recombinant His-Rac1 or Rac1-RhoGDI complex (1:1) at a concentration of 100 ng/μL was incubated with purified active caspase-3 (5 ng/μL) at 37°C in caspase reaction buffer [50 mmol/L HEPES (pH 7.5), 100 mmol/L NaCl, 1 mmol/L EDTA, 10% sucrose (w/v), 0.1% CHAPS, 10 mmol/L DTT]. Where indicated, purified GST-RhoGDI proteins were added and preincubated with His-Rac1 before addition of caspase-3. At the indicated times, the reaction was terminated by the addition of the SDS sample buffer. To assess the level of endogenous Rac1, cells were lysed by nitrogen cavitation using a Parr bomb (400 p.s.i. for 5 minutes to 5 × 108 MDA-MB-231 cells) in lysis buffer [25 mmol/L HEPES (pH 7.5), 1 mmol/L EDTA, 1% NP40, 10% glycerol, and protease inhibitors (1:50 dilution of the Protease Inhibitor Cocktail Set III, Calbiochem)]. After clearance of the insoluble fractions, the supernatant was supplemented with 100 μmol/L GTPγS and incubated at room temperature for 30 minutes followed by addition of MgCl2 to a final concentration of 10 mmol/L. The intact and functional Rac1 was precipitated by GST-PBD immobilized on agarose beads. The resulting pellets were subjected to Western blot analysis with anti-Rac1 antibody. The cleavage of Cdc42 was examined in a similar manner.
Immunoblot analysis. Equal amounts of cell lysates (20 μg/lane) were resolved by electrophoresis using a 4% to 12% NuPAGE Bis-Tris gel (Invitrogen) and transferred to nitrocellulose membranes (Millipore, Bedford, MA) for immunoblot analysis. When necessary, the membrane was stripped by Restore Western Blot Stripping Buffer (Pierce) and reprobed with appropriate antibodies. Immunocomplexes were visualized by chemiluminescence using enhanced chemiluminescence (Santa Cruz Biotechnology) or SuperSignal reagent (Pierce).
Elevation of Rho GDP dissociation inhibitor expression confers resistance to apoptosis. To examine whether expression of RhoGDI would alter cellular sensitivity to apoptosis, we initially did a transient transfection of MDA-MB-231 human breast cancer cells with a GFP-RhoGDI expression plasmid, pEGFP-C3/RhoGDI, or the empty vector as a control. Following a 48-hour transfection, GFP-positive cells were isolated by FACS and maintained in a medium containing 1.5 mg/mL kanamycin. Western blot analysis revealed the expression of GFP-RhoGDI in the transfected cells, whereas the endogenous RhoGDI expression was essentially unaffected ( Fig. 1A ). Under normal growth conditions, cells overexpressing GFP-RhoGDI appeared morphologically similar to the control cells (data not shown), and growth rates, as determined by MTT assay, were indistinguishable ( Fig. 1B). These RhoGDI-overexpressing cells were then tested for their sensitivity to chemotherapeutic agents. As shown in Fig. 1C and D, GFP-RhoGDI cells exhibited increased resistance to VP-16-induced and doxorubicin-induced cell death compared with control cells. Overexpression of RhoGDI also partially impaired the apoptotic response of MDA-MB-231 cells to taxol, although less so (data not shown).
We further tested the role of RhoGDI overexpression on drug-induced apoptosis in JLP-119 human lymphoma cells. To this end, we established a stable JLP-119 cell line in which the level of RhoGDI was elevated by ∼2- to 3-fold compared with the vector control cells ( Fig. 1E). Cells were induced to undergo apoptosis with VP-16 or staurosporine and the percentage of apoptotic cells was quantified by examining nuclear morphology following staining with Hoechst 33342 and PI. Consistent with our previous observations, both agents induced extensive apoptosis in JLP-119 control cells ( Fig. 1F). In sharp contrast, the RhoGDI-overexpressing cell line was protected against apoptosis induced by both VP-16 and staurosporine ( Fig. 1F). At the 4- and 6-hour time points, the transfected cells showed ∼40% less apoptosis than control cells. These results revealed that overexpression of RhoGDI renders cancer cells more resistant to apoptosis induction by different stimuli.
Loss of Rho GDP dissociation inhibitor sensitizes MDA-MB-231 cells to drug-induced apoptosis. To elucidate the antiapoptotic role of RhoGDI, we employed a siRNA strategy to target and knockdown RhoGDI transcripts in MDA-MB-231 cells. To this end, we employed a DNA template–based vector, pRNA-U6.1 (GeneScript), directing the synthesis of a 19-nucleotide double-stranded siRNA. Because the vectors integrate into the genomes of the target cells, continuous expression of RhoGDI siRNAs containing hairpin loops is achieved in the stable transfectants ( Fig. 2A ). Three target sequences were tested, corresponding to nucleotides 390 to 408 (siRhoGDI-I), 401 to 419 (siRhoGDI-II), and 413 to 431 (siRhoGDI-III) of RhoGDI mRNA, respectively ( Fig. 2A). Plasmid that expresses siRNA against firefly luciferase, siLuciferase, was used as a control. After transfection, hygromycin-resistant cells were expanded and screened for the expression of RhoGDI by Western blot analysis (see details in Materials and Methods). Whereas plasmid pRNA-U6.1/siLuciferase had no effect on RhoGDI expression, pRNA-U6.1/siRhoGDI-II greatly diminished its expression ( Fig. 2B). In ∼20% of hygromycin-resistant clones, RhoGDI expression was reduced by at least 90% as judged by densitometry of the protein bands. Importantly, the expression of RhoGDI was persistently suppressed after the initial transfection, indicating that the knockdown phenotype is stable over time ( Fig. 2C). The RhoGDI-depleted cells (designated MDA-MB-231−/−RhoGDI) were tested for their growth properties using the MTT assay. The data showed that the proliferation rate of these cells is slightly slower than that of wild-type cells ( Fig. 2D) or cells expressing siLuciferase (data not shown). Neither siRhoGDI-I nor siRhoGDI-III had any effect on RhoGDI expression, supporting the notion that specific nucleotide sequences of siRNA are required to achieve effective gene silencing ( 23). Of note, none of these three siRNAs had an effect on RhoGDI expression when transfected into JLP-119 or BL-41 lymphoma cells, suggesting that the DNA vector–based RNAi approach may be cell type dependent.
Next, we examined the effect of RhoGDI depletion on apoptosis. As measured by flow cytometry ( Fig. 2E), cells lacking RhoGDI exhibited 2- to 3-fold more apoptosis than control cells and exhibited increased activation of caspase-3 and cleavage of PARP in response to VP-16 treatment ( Fig. 2F). Depletion of RhoGDI in MDA-MB-231 cells did not alter the baseline level of apoptosis ( Fig. 2E). Reproducible results were obtained with three independently isolated cell lines.
Restoration of Rho GDP dissociation inhibitor expression restores drug resistance. To confirm that the observed RNAi effect is gene specific, we determined that drug resistance is restored by restoring RhoGDI levels. This was accomplished by introducing an exogenous RhoGDI expression vector in the siRhoGDI cells. MDA-MB-231−/−RhoGDI cells were transfected with a rescue GFP-RhoGDI plasmid, pEGFP-C3/RhoGDI-re, in which the sequence of the RhoGDI gene (401-419) complementary to the siRNA oligonucleotide was mutated to prevent destruction of the exogenous mRNA by the siRNA ( Fig. 3A ). After a 48-hour transfection, GFP-positive cells were isolated by FACS and analyzed for RhoGDI expression by immunoblotting ( Fig. 3B). The level of GFP-RhoGDI-re was comparable with that of endogenous RhoGDI in wild-type cells ( Fig. 3B). Interestingly, the ectopic expression of GFP-RhoGDI-re did not interfere with siRhoGDI-mediated gene silencing of the native RhoGDI ( Fig. 3B). GFP-RhoGDI-re was localized properly to the cytoplasm as determined by fluorescence microscopy ( Fig. 3C). The cells expressing GFP-RhoGDI-re were resistant to the induction of cytotoxicity by VP-16 and doxorubicin as shown by the MTT assay ( Fig. 3D). The results show that ectopic expression of GFP-RhoGDI-re compensates for the loss of the endogenous protein, thus demonstrating not only the specificity of RNAi inhibition but also the activity of RhoGDI in inhibiting apoptosis.
Loss of Rho GDP dissociation inhibitor increases apoptosis in a caspase-dependent manner. To understand the mechanism underlying the enhanced apoptosis that results from RhoGDI knockdown, we examined the effect of caspase inhibitors on drug-induced apoptosis in MDA-MB-231 cells and in cells lacking RhoGDI. A general caspase inhibitor (Z-VAD) was added to cells 1 hour before VP-16 treatment. Caspase activity was almost completely blocked by Z-VAD at 10 μmol/L, as was induction of apoptosis in both MDA-MB-231 cells and cells lacking RhoGDI ( Fig. 4 ). These results indicate that the caspase activity is essential for the increased apoptosis as a result of RhoGDI depletion.
Rho GDP dissociation inhibitor blocks cleavage of Rac1 GTPase by caspase-3 in vitro and during apoptosis. Rac1 GTPase plays a critical role in cell survival in many cell types, including breast cancer cells. We have shown previously that Rac1 is a substrate for caspase-3-like caspases and that caspase-mediated Rac1 cleavage promotes drug-induced apoptosis ( 19). Knowing that RhoGDI forms a complex with Rac1 in the cytosol, we asked whether RhoGDI inhibits apoptosis, at least in part, by modulating the cleavage and inactivation of Rac1. To facilitate these studies, we purified recombinant geranylgeranylated Rac1 (Rac1-GG) from COS-1 cells, knowing that this prenyl chain is required for complex formation ( 21). As was found previously with bacterially expressed (nonprenylated) Rac1 ( 19), exposure of Rac1-GG to purified active caspase-3 resulted in the formation of two COOH-terminal fragments of ∼20 and 16 kDa ( Fig. 5A ). RhoGDI protein was completely resistant to caspase-3 ( Fig. 5A, bottom), although it contains a caspase-3 consensus sequence. We then tested the effect of RhoGDI on the Rac1 cleavage by adding RhoGDI at different concentrations. As shown in Fig. 5B, RhoGDI inhibited the cleavage of Rac1 in a dose-dependent manner. At 10 μmol/L, >90% of fragment formation was inhibited. To confirm this effect, we tested caspase-3 activity on mammalian Rac1-RhoGDI complex (1:1) purified from Sf9 cells ( 24). Rac1 was completely resistant to caspase-3 cleavage within this complex ( Fig. 5C).
To assess the cleavage of Rac1 in cells induced to undergo apoptosis with VP-16, we employed a Rac1 functional (effector binding) assay that takes advantage of the fact that Rac1 binds to its downstream effector molecules only if it is in the active state (i.e., full-length and GTP-bound; refs. 19, 22). As shown previously, cleavage of Rac1 results in loss of binding to PAK1 kinase. At different time points following VP-16 treatment, whole-cell extracts from MDA-MB-231 cells or cells lacking RhoGDI were preincubated with GTPγS, a nonhydrolyzable GTP analogue that, when bound to Rac1, converts it to the stable (active) state. As apoptosis proceeded, the amount of Rac1-GTPγS available to bind to the kinase gradually decreased in both cell lines; however, in the cells lacking RhoGDI, the rate of decrease was significantly greater ( Fig. 6, top ). Importantly, the cleavage of Rac1 occurred in parallel with the progression of apoptosis (see Fig. 2). Taken together, the results suggest that RhoGDI, which forms a tight complex with Rac1 GTPase in the cytoplasm, may play a role in shielding Rac1 from caspase-3-like proteases, thus keeping the GTPase in an intact and functional state.
Cdc42, but not RhoA, is a substrate for caspases and is cleaved in cells undergoing apoptosis in response to FasL ( 25). We examined the cleavage of these two GTPases in MDA-MB-231 cells or cells lacking RhoGDI. Cdc42 was followed by a GST-PBD pull-down assay as done for Rac1. As shown in Fig. 6 (bottom), the amount of GTPγS-Cdc42 bound to PBD gradually decreased during apoptosis in control cells, and the decrease was greatly accelerated in RhoGDI knockdown cells. These data suggest that RhoGDI also inhibits the apoptotic cleavage of Cdc42. RhoA protein was followed directly in whole-cell lysates by Western blotting using anti-RhoA antibodies. The results show that RhoA remained intact throughout the course of apoptosis regardless of RhoGDI status (data not shown).
Expression of the caspase-3-resistant Rac1 mutant restores resistance of Rho GDP dissociation inhibitor knockdown cells. We have shown previously that expression of the caspase-3-resistant Rac1 mutant (Rac1D11E) suppresses VP-16-induced apoptosis in human lymphoma cells ( 19). To understand the mechanism of RhoGDI-mediated antiapoptosis, we examined whether expression of Rac1D11E inhibits drug-induced apoptosis in the RhoGDI knockdown cells. To this end, MDA-MB-231 cells lacking RhoGDI were transfected with an empty pcDNA3.1 vector or vector encoding hemagglutinin (HA)–tagged wild-type Rac1 (Rac1wt) or the caspase-3-resistant mutant Rac1D11E followed by VP-16 treatment. As shown in Fig. 7 , the expression of Rac1D11E significantly inhibited VP-16-induced apoptosis in RhoGDI-depleted cells compared with wild-type Rac1 (P < 0.01).
RhoGDI is frequently up-regulated in different human tumors, including breast cancer ( 12– 15). Particularly high expression levels are observed in stable, chemoresistant cancer cell lines, suggesting that RhoGDI might play a role in the development of drug resistance ( 16– 18). Here, we have provided evidence for a novel function for RhoGDI, showing that it can protect cancer cells against apoptosis induced by different chemotherapy agents. Overexpression of cellular RhoGDI caused resistance to VP-16-induced or doxorubicin-induced apoptosis in breast cancer and lymphoma cells, and molecular depletion of cellular RhoGDI protein had the opposite effect. All cell lines with forced RhoGDI depletion were more sensitive to drug-induced apoptosis. Importantly, when RhoGDI protein levels were restored, resistance to VP-16-induced apoptosis was also restored, thus demonstrating that the observed effect on apoptosis was a result of RhoGDI-specific activity. Molecular studies into the mechanism of RhoGDI activity showed that RhoGDI is a potent inhibitor of Rac1 cleavage by caspase-3 both in vitro and during apoptosis. VP-16 and doxorubicin are both topoisomerase II inhibitors that induce apoptosis in many types of cancer cells. The results presented here show that the level of RhoGDI in cancer cells regulates the apoptotic response of the cells to these two chemotherapy agents as well as to staurosporine and suggest that RhoGDI overexpression gives cancer cells a survival mechanism against these antitumor chemotherapies. Consistent with this hypothesis, treatment of myeloid leukemia cells with agents that increase cell death in response to chemotherapy drugs (all-trans retinoic acid and the cyclin-dependent kinase inhibitor bohemine treatment) was associated with down-regulation of RhoGDI levels ( 26, 27).
Alteration of RhoGDI levels, either overexpression or depletion, had little effect on the monolayer growth rate of MDA-MB-231 breast cancer cells. Moreover, stable MDA-MB-231 cells lacking RhoGDI were morphologically indistinguishable from control cells. This finding contrasts the observations made with nontransformed cells, where RhoGDI levels must be under tight control in order for the cells to behave normally ( 7– 10). In cardiomyocytes, for example, increasing RhoGDI expression to four times the normal levels caused severe defective cardiac morphogenesis.
The mechanism of action of RhoGDI as an antiapoptotic molecule remains to be fully established. One potential mechanism, shown here, is that RhoGDI inhibits drug-induced apoptosis by protecting Rac1 GTPase from apoptotic cleavage. It is well accepted that chemotherapeutic drugs induce cell death by triggering activation of the caspase cascade and the subsequent cleavage of various cellular substrates ( 28). In many cell types, Rac1 GTPase plays a key role in controlling cell survival ( 20, 29, 30). Our previous work showed that caspase-mediated cleavage of Rac1 is a critical event in drug-induced apoptosis ( 19). Cleavage of native Rac1 occurs at VVGD11/G, a site that is fully covered when Rac1 is bound to RhoGDI ( 21). In this study, we show that Rac1 is completely resistant to caspase cleavage when in a 1:1 complex with RhoGDI. Thus, it is likely that RhoGDI inhibits drug-induced apoptosis, at least in part, by blocking the apoptotic cleavage of Rac GTPases. In support of this notion, expression of the caspase-3-resistant Rac1D11E mutant increased the resistance of RhoGDI-depleted cells to VP-16-induced cytotoxicity ( Fig. 7). These results, however, do not exclude the possibility of other mechanisms by which RhoGDI may exert its antiapoptotic function. Nonetheless, these findings add a novel role for RhoGDI in the regulation of Rac GTPases during apoptosis: shielding Rac1 against proteolytic degradation. It should be noted that RhoGDI also inhibits the apoptotic cleavage of Cdc42. Whether this contributes to the RhoGDI-mediated inhibition of apoptosis remains to be investigated.
RhoGDI is often thought of as a negative regulator Rho GTPases, acting through its ability to bind to and hold the GTPases in inactive, cytosolic forms that are unable to effectively interact with GEFs and/or downstream target molecules ( 1). However, several lines of evidence, including the present findings, show that RhoGDI can also positively regulate Rho GTPases. For example, Rac1 regulation of NADPH oxidase activity in neutrophils may require a complex with RhoGDI ( 24, 31, 32). Similarly, RasGRF-induced mitogen-activated protein kinase activation and Cdc42-mediated cellular transformation ( 33) may require formation of a complex of the respective GTPases with RhoGDI ( 34). It also seems that RhoGDI can serve as an escort to shuttle Rho GTPases to membrane-associated signaling complexes, which is crucial for coupling the GTPases to their downstream effector proteins ( 35). Now, we show that RhoGDI positively regulates Rac1 by protecting it from apoptosis-associated inactivation. Interestingly, the overall production of reactive oxygen species (ROS) was significantly reduced in the MDA-MB-231 cells after depletion of RhoGDI, 3 suggesting a possible role for RhoGDI in Rac1-mediated ROS production in these breast cancer cells. It will be interesting to determine if the linkage between RhoGDI and Rac1 in controlling ROS production plays a role in the RhoGDI-mediated antiapoptotic effects. Our preliminary results also support a role for RhoGDI in regulating the subcellular localization of Rho family GTPases, as both Rac1 and Cdc42 were overwhelmingly translocated from the cytosol to the plasma membrane in MDA-MB-231 cells lacking RhoGDI (data not shown). Further experiments will determine whether this change in Rho GTPase localization contributes to the effect of RhoGDI on drug-induced apoptosis.
In summary, we have shown for the first time that RhoGDI is an antiapoptotic molecule that can mediate tumor resistance to apoptosis. High levels of RhoGDI have been found in a variety of tumors, with higher levels being present in malignant rather than benign tumors. Thus, RhoGDI may become a new target for anticancer treatment.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵3 B. Zhang et al., unpublished results.
- Received January 19, 2005.
- Revision received March 31, 2005.
- Accepted May 4, 2005.
- ©2005 American Association for Cancer Research.