The Rho Family Member RhoE Interacts with Skp2 and Is Degraded at the Proteasome during Cell Cycle Progression

Background: RhoE is an atypical Rho protein that lacks GTPase activity.

Results: RhoE is degraded at the G₁/S cell cycle transition in a proteasome-dependent manner.

Conclusion: Cell cycle progression requires the proteasomal degradation of RhoE.

Significance: This new mechanism of controlling RhoE protein levels can regulate cellular proliferation and may be related to cancer.

Abstract

RhoE/Rnd3 is an atypical member of the Rho family of small GTPases. In addition to regulating actin cytoskeleton dynamics, RhoE is involved in the regulation of cell proliferation, survival, and metastasis. We examined RhoE expression levels during cell cycle and investigated mechanisms controlling them. We show that RhoE accumulates during G₁, in contact-inhibited cells, and when the Akt pathway is inhibited. Conversely, RhoE levels rapidly decrease at the G₁/S transition and remain low for most of the cell cycle. We also show that the half-life of RhoE is shorter than that of other Rho proteins and that its expression levels are regulated by proteasomal degradation. The expression patterns of RhoE overlap with that of the cell cycle inhibitor p27. Consistently with an involvement of RhoE in cell cycle regulation, RhoE and p27 levels decrease after overexpression of the F-box protein Skp2. We have identified a region between amino acids 231 and 240 of RhoE as the Skp2-interacting domain and Lys²³⁵ as the substrate for ubiquitylation. Based on our results, we propose a mechanism according to which proteasomal degradation of RhoE by Skp2 regulates its protein levels to control cellular proliferation.

Introduction

Mammalian Rho GTPases comprise a family of 22 members with unique biological roles. By interacting with target proteins, Rho GTPases control a variety of cellular functions such as cell adhesion, cell cycle progression, cell migration, morphogenesis, gene expression, and actin cytoskeleton dynamics (1–4).

Many Rho GTPases cycle between an inactive GDP-bound and an active GTP-bound conformation. RhoGEFs, RhoGAPs, and RhoGDIs are the main regulators of Rho proteins by switching them between their on and off states (5–7). It has recently been shown that several Rho proteins spontaneously exchange nucleotides or lack GTPase activity, suggesting that the regulation of Rho proteins does not rely exclusively on GTP exchange factors (8, 9). Indeed, recent evidence indicates that Rho GTPases are finely tuned by nucleotide exchange-independent regulatory mechanisms that include post-translational modifications and protein degradation (6, 10, 11).

Atypical Rho proteins only found in vertebrates are “GTPase-less GTPases” and, unlike typical Rho GTPases, remain always bound to GTP in a constitutively active state (12–15). This suggests that they are regulated by other mechanisms such as the regulation of gene expression and protein stability, post-translational modifications, and protein interactions. RhoE/Rnd3 is a member of the Rnd subfamily of atypical Rho proteins (16, 17) that also includes Rnd1 and Rnd2 (13). RhoE antagonizes RhoA function by binding to and inhibiting its main effector ROCK (18). RhoE also interacts with p190RhoGAP (19) to indirectly increase the GTPase activity of RhoA. As a consequence, RhoE antagonizes RhoA functions related to actin cytoskeleton organization and, among other effects, induces cell motility (17). In addition, RhoE is involved in controlling cell cycle progression and survival in some cell lines (20–23); it participates in the response to genotoxic stress (24, 25); and it also plays a role in the development and function of the central nervous system (26–28). Furthermore, recent studies have shown that RhoE may be important for invasiveness and metastatic potential in certain tumors (23, 29–31), suggesting that RhoE could be a suitable new target in cancer therapy.

Protein ubiquitylation is a multistep process that ultimately results in the attachment of ubiquitin chains to lysine residues within target proteins. The number of ubiquitin molecules and type of chains transferred to a protein determines its final cellular destination. Although monoubiquitylation modulates protein localization, polyubiquitylation leads to proteasomal degradation (33–36). It has been demonstrated that ubiquitin conjugation is relevant to regulation of the Rho family proteins RhoA, Cdc42, and Rac1 (37–40).

In view of the important effects of RhoE on cell cycle progression, proliferation, and survival, we investigated the expression of RhoE during the cell cycle. Our results show that RhoE accumulates in cells during G₁ and when exiting the cell cycle. More importantly, RhoE is rapidly degraded at the G₁/S transition in an Akt-dependent manner. This degradation is mediated by the proteasome and requires ubiquitylation of Lys²³⁵ by Skp2. Substitution of this residue for alanine stabilizes RhoE. This regulation constitutes a novel mechanism for regulating RhoE function and directly connects the mechanisms determining its protein stability to cell cycle regulation.

EXPERIMENTAL PROCEDURES

Cell Culture and Transfections

U87, HeLa, and 293 cell lines were maintained in DMEM with 10% FBS (Invitrogen). Primary mouse embryonic fibroblasts were generated as previously described (41) and maintained in DMEM with 10% FBS for early passages (P2–P5).

Transient transfections were performed using a calcium phosphate transfection kit (Sigma). Sixteen hours after transfection, cells were washed with PBS and maintained in fresh growth medium for 48 h before extracts were obtained for Western blotting, immunoprecipitation, or GST bead pulldown assays.

Expression Vectors

The expression vector encoding full-length murine RhoE cDNA was obtained from Dr. A. Ridley (King's College, London, UK). Mutants were generated using the QuikChange system (Stratagene) according to the manufacturer's instructions and cloned into the pCMV5 FLAG vector to produce FLAG epitope-tagged proteins. All mutants were confirmed by sequencing.

Cell Cycle Analysis

HeLa cells were transfected as described above with a modified pCMV5 FLAG-RhoE vector in which we included a CMV-driven GFP cassette to identify transfected cells. Cells were fixed in 80% ethanol and processed for DNA staining with propidium iodide as described previously (42, 43). GFP-positive cells were analyzed in a FC500 flow cytometer (Beckman Coulter), and the cell cycle profile was assessed using the Cytomics^TM RXP program.

For cell cycle synchronization experiments, HeLa cells were grown to subconfluency before treatments. For G₁/S synchronization, cells were treated with 2 mm thymidine (Sigma) for 16 h, released by changing the medium for 12 h, and treated with 5 μg/ml aphidicolin (Sigma) for 16 h. For G₂/M synchronization, cells were treated with 400 ng/ml nocodazole (Sigma) for 16 h. Cells were released from cell cycle blockade by washing with PBS followed by addition of growth medium. After incubation in growth medium for the indicated times, cells were collected and processed for Western blotting or stained with propidium iodide for cell cycle analysis as described above.

Lactacystin and Cycloheximide Treatments

HeLa cells were treated with lactacystin (10 μm; Sigma) or cycloheximide (70 μg/ml; Sigma) as indicated. Cells were processed for Western blotting as described below. The levels of different Rho proteins at each time point were quantified by densitometry analysis. Half-lives were calculated from the logarithmic transformation of the densitometry data plotted against time.

Adenoviral Infections

The Skp2 expressing adenovirus (44) was a kind gift from Dr. D. A. Wolf (Sanford-Burnham Medical Research Institute, La Jolla, CA). Producer 293 cells were transfected with adenovirus expressing Skp2 or HA-tagged dominant negative Akt (T308A/S473A, dnAkt) (45) for amplification. Adenovirus titration was performed with the TCID₅₀ (tissue culture infectious dose 50%) method based on the cytopathic effect in 293 cells using end point dilutions (21). To achieve different expression levels of proteins in U87 cells, we used different amounts of virus particles according to multiplicity of infection. U87 cells were incubated with adenovirus in a one-fifth volume of the culture medium. Two hours after infection, the medium was removed, and cells were maintained in fresh medium for an additional 48 h before collection and processing for Western blotting.

Western Blotting

Cell samples were processed for Western blotting as previously described (21). Membranes were blocked in 5% nonfat dried milk in PBS containing 0.1% Tween 20 (PBT) and incubated for 1 h to overnight with gentle shaking in the following primary antibodies: anti-RhoE and anti-Rac1 (Millipore); anti-HA, anti-p27, anti-RhoA, and anti-Cdc42 (Santa Cruz Biotechnology); anti-pAKT (pSer473) (Cell Signaling); anti-Skp2 (Zymed Laboratories Inc.); and anti-FLAG, anti-Actin, and anti-β-tubulin (Sigma). After three washes in 0.1% PBT, blots were incubated for 45 min with horseradish peroxidase-conjugated secondary antibodies (Thermo Scientific). Following secondary antibody incubation, blots were washed in PBT and developed using enhanced chemiluminescence (ECL Plus; GE Healthcare).

Immunoprecipitation

HeLa cells transiently expressing FLAG-RhoE were rinsed twice with PBS and scraped into 1 ml of ice-cold lysis buffer (150 mm NaCl, 0.5% Triton X-100, 5 mm DTT, 50 mm Tris, pH 7.6) containing protease inhibitors (Complete Mini; Roche Applied Science). Cellular debris was removed by sedimentation, and protein concentrations in the supernatant were determined using a protein assay kit (Bio-Rad DC protein assay). Protein extracts were precleared with protein A-Sepharose beads for 1 h at 4 °C, and 0.5-ml aliquots of the precleared lysate containing 1 mg of total protein were incubated for 1 h at 4 °C with mouse anti-FLAG or anti-histidine (as negative control) antibodies covalently linked to agarose (Sigma). Immune complexes were washed with lysis buffer and resuspended in SDS sample buffer. Proteins were resolved on 15% SDS-PAGE gels, transferred to PVDF membranes, and incubated with anti-Skp2 antibodies. Total extract (50 μg) was used as input.

GST Pulldown Assays

Recombinant GST-Skp2 protein was produced in BL21 Escherichia coli cells and bound to glutathione-Sepharose 4B beads (GE Healthcare) for 1 h at room temperature. Sepharose beads were incubated with 1 mg of extracts from 293 cells transiently transfected with different RhoE constructs for 4 h at 4 °C with gentle rotation. Samples were washed three times with PBS and analyzed by Western blotting for the presence of RhoE.

In Vitro Ubiquitylation Assay

The in vitro ubiquitylation assay was based on a described method (46). Ubiquitylation substrates were prepared using a coupled in vitro transcription and translation system (TnT quick coupled transcription/translation System; Promega). Briefly, 3 μg of pcDNA-FLAG RhoE DNA and the different mutants used were incubated in the presence of methionine (in a final volume of 20 μl) according to the manufacturer's instructions. 10 μl of the in vitro transcription/translation reactions were added to the ubiquitin ligation reaction (in a final volume of 30 μl) containing 50 mm Tris-HCl, pH 7.5, 2 mm MgCl₂, 2 mm dithiothreitol, 2 mm ATP, 10 μg of purified ubiquitin (Sigma), 0.5 μg of ubiquitin aldehyde (Biomol; Enzo Life Sciences), and 10 mm MG132 (Biomol; Enzo Life Sciences). Nontreated rabbit reticulocyte lysate (3 μl; Promega) was added to provide components of the proteasome machinery. The reactions were incubated for 2 h at 30 °C and terminated by boiling for 5 min in SDS sample buffer. Proteins were resolved in 10% SDS-PAGE. Ubiquitylated RhoE proteins were detected by Western blotting using anti-RhoE antibodies.

Immunofluorescence

HeLa cells were grown on coverslips and transfected as described above with pCMV5 constructs expressing FLAG-tagged versions of wild-type RhoE (1–244) or the K235A mutant. 48 h post-transfection, cells were fixed in 4% paraformaldehyde and permeabilized in 0.2% Triton X-100. Cells then were incubated with anti-FLAG antibodies (Sigma), revealed with secondary FITC-conjugated anti-mouse antibodies (Sigma), and mounted with Vectashield mounting medium (Vector Laboratories), containing DAPI for nuclei staining. Slides were analyzed by confocal microscopy (Nikon Eclipse C1 plus).

Statistical Analysis

Graphs represent the means ± S.E. of at least three independent experiments. Unless otherwise stated, statistical differences were evaluated by two-way analysis of the variance and Bonferroni post hoc test using GraphPad Prism software (GraphPad). p < 0.05 was considered statistically significant.

RESULTS

RhoE Accumulates in G₁

Previous results demonstrate that overexpression of RhoE induces a G₁ cell cycle arrest in various cell lines (20–23). We could reproduce this effect in HeLa cells, in which overexpression of RhoE led to a rapid decrease of cells in S phase and the accumulation of cells in G₁ 48 h after transfection (Fig. 1A). We therefore reasoned that the expression of RhoE was regulated across the cell cycle. To test this hypothesis, we analyzed RhoE protein levels in synchronized cultures of HeLa cells. We arrested the cell cycle at the G₁/S transition with a thymidine/aphidicolin treatment and also at the G₂/M transition with nocodazole. As shown in Fig. 1B, HeLa cells treated with nocodazole were arrested at the G₂/M boundary but entered G₁ after release from this block. Thymidine/aphidicolin treatment arrested cells at the G₁/S transition, and after release, they entered the S phase. The analysis of RhoE expression in treated cells revealed that RhoE levels were low during most of the cell cycle. However, RhoE accumulated during G₁, reaching maximum levels at late G₁, 6 h after release from the nocodazole block (Fig. 1C). At the G₁/S transition, RhoE levels decreased to a minimum, showed a moderate increase during the S phase, and decreased again in G₂. Our results suggest that RhoE protein levels accumulated during the G₁ phase and were rapidly degraded before cells enter the S phase.

FIGURE 1.

RhoE accumulates in cells during G₁. A, RhoE inhibits proliferation. HeLa cells were transfected with constructs co-expressing RhoE and GFP, or only GFP as control (C). 24 and 48 h after transfection, cells were incubated with propidium iodide to stain the DNA and analyzed by flow cytometry. Only GFP-positive cells (transfected) were gated and analyzed. Cell cycle distributions from three different experiments (means ± S.E.) are shown. **, p < 0.01; ***, p < 0.001. B and C, proliferating HeLa cells were blocked at the G₂/M (nocodazole) or G₁/S (thymidine/aphydicolin, T/A) transitions and released by adding fresh culture medium. After the indicated times, cell cycle progression was monitored by flow cytometry (B). RhoE and p27 expression was analyzed by Western blotting (C, left panel). The mean expression of RhoE (relative to the expression at the time of nocodazole release) during cell cycle from three different experiments is represented in the graph (C, right panel). A.U, arbitrary units. D, primary mouse embryonic fibroblasts were seeded at 1 × 10⁵ cells/well in 6-well dishes. Culture medium was replaced 24 h later (day 0) and every 48 h thereafter. Cells were collected at days 0, 4, and 8. RhoE and p27 expression was analyzed by Western blotting. Actin was used as a loading control.

One of the main mechanisms regulating cell cycle progression is the coordinated expression and degradation of cell cycle inhibitors such as p27. p27 is abundant in quiescent and G₁ phase cells and down-regulated in proliferating cells and in S and G₂ phase cells (47). When comparing the expression of RhoE and p27 in our synchronization experiments, we found that both proteins showed the same expression pattern with a peak at late G₁ and rapid decrease at the G₁/S boundary (Fig. 1C).

Because p27 also accumulates in primary cells when they are contact-inhibited in high density cultures, we analyzed the expression of RhoE under these conditions. As shown in Fig. 1D, both RhoE and p27 accumulated in primary mouse embryonic fibroblasts as they entered quiescence. Our results indicate that RhoE was expressed in G₁ and accumulated in cells that stop proliferating. More importantly, they suggest that continuing proliferation required degradation of RhoE at the G₁/S transition and during G₂.

RhoE Is Degraded at the Proteasome

To investigate whether degradation of RhoE was involved in regulation of its protein levels, we treated HeLa cells with the protein synthesis inhibitor cycloheximide before analyzing protein levels of RhoE at different time points, along with that of other Rho proteins. As shown in Fig. 2A, RhoA, Cdc42, and Rac1, which are mainly regulated by their GTPase activity, were stably expressed with half-lives longer than 6 h. Conversely, RhoE protein levels decreased rapidly after cycloheximide treatment with an estimated half-life of ∼2.4 h. Thus, protein degradation may be a mechanism to regulate atypical but not typical Rho proteins. One of the main systems regulating protein stability in cells is the ubiquitin proteasome system. To analyze the role of the ubiquitin proteasome system in RhoE regulation, we treated HeLa cells with the proteasome inhibitor lactacystin. Treatment resulted in the accumulation of RhoE, indicating that it was degraded at the proteasome (Fig. 2B). Proteasomal degradation requires a previous ubiquitylation step. To test whether RhoE was ubiquitylated, we performed an in vitro ubiquitylation assay. When in vitro synthesized RhoE protein was incubated with ubiquitin and reticulocyte lysate, ladder patterns corresponding to ubiquitylated RhoE were observed (Fig. 2C).

FIGURE 2.

RhoE is ubiquitylated and degraded at the proteasome. A, RhoE is a short half-life protein. HeLa cells were left untreated (lane C) or treated with 70 μg/ml cycloheximide (CHX) at different times. Protein levels of RhoE, RhoA, Cdc42, and Rac1 were analyzed by Western blotting (left panel). β-Tubulin was used as a loading control. The intensities of bands from three independent experiments were quantified and plotted as logarithm of the percentage of protein amount for RhoA (closed squares), Cdc42 (open circles), Rac1 (closed circles, dotted line), and RhoE (open squares). Protein levels at time 0 (untreated) were set to 100%. The gray dotted line represents the 50% value (right panel). B, RhoE is degraded at the proteasome. Expression levels of RhoE in untreated (lane C) or lactacystin-treated (10 μm, LACT) HeLa cells were analyzed by Western blotting at the indicated times (left panel). Three independent experiments were quantified, and the amount of RhoE protein (arbitrary units, A.U) at different time points was compared with time 0 (right panel). **, p < 0.01 in a one-way analysis of the variance followed by a Tukey post hoc test. C, RhoE is ubiquitylated in vitro. Full-length RhoE protein was generated by in vitro transcription/translation and incubated with or without reticulocyte extracts (Reti) in in vitro ubiquitylation reactions (see “Experimental Procedures”). RhoE and its modified forms were detected by Western blotting using anti-RhoE antibodies. The reticulocyte extract without RhoE was used as negative control (left lane). (Ub)_n-RhoE, polyubiquitylated RhoE.

RhoE Is Regulated by the Akt Pathway through the F-box Protein Skp2

Skp2 is an oncogenic protein that targets tumor suppressors for proteasomal degradation. A major target of Skp2 is the cell cycle inhibitor p27 (48). In a previous report, we showed that the F-box protein Skp2, the substrate recognition subunit of the E3 ubiquitin ligase SCF^Skp2, is regulated by PI3K/Akt (42). In a separate study, we showed that RhoE expression is reduced in U87 cells (21). U87 cells lack PTEN expression, and as a result, the PI3K/Akt pathway in this cell line is constitutively activated. To investigate a possible functional interaction between Akt, Skp2, and RhoE, we used an HA epitope-tagged dominant negative mutant of Akt (T308A/S473A) and analyzed the expression levels of Skp2 and RhoE. Expression of dominant negative Akt in U87 cells resulted in a 22% decrease of Skp2 and a 39% increase of RhoE expression levels (**, p < 0.01) (Fig. 3A).

FIGURE 3.

RhoE stability is regulated by the Akt pathway through the F-box protein Skp2. A, U87 human glioblastoma cells were infected with increasing amounts (multiplicity of infection = 10, 30, 50, and 100) of adenovirus expressing an HA-tagged dominant negative Akt mutant (dnAkt). 48 h later, the expression of Skp2 and RhoE was analyzed by Western blotting, and the intensity of the bands was quantitated. The graph on the right shows the means ± S.E. from three different experiments. Two-way analysis of the variance followed by Bonferroni test revealed statistically significant differences with uninfected control cells. **, p < 0.01. dnAkt expression was analyzed with anti-HA antibodies. The inhibitory effect of the dnAkt mutant on Akt signaling was monitored with anti-phospho Akt (pAkt-Ser473) specific antibodies. Uninfected U87 cell extracts were used as control (lane C). B, U87 cells were infected with an adenoviral construct expressing Skp2 (Ad Skp2; multiplicity of infection = 50). At the indicated time points, cells were collected, and the levels of RhoE, Skp2 and p27 were analyzed by Western blotting. β-Tubulin was used as loading control. C, RhoE and Skp2 interact in vivo. HeLa cells were transiently transfected with a pCMV-FLAG RhoE construct. Cell lysates were incubated with anti-histidine (HIS, as negative control) or anti-FLAG antibodies covalently linked to agarose. The immune complexes were resolved by SDS-PAGE. Skp2 was detected with specific antibodies by Western blotting. Total extract (50 μg) was used as input. IP, immunoprecipitation.

The inverse relationship between Skp2 and RhoE protein levels suggested that Akt may regulate RhoE expression levels in cells by modulating its Skp2-dependent proteasomal degradation. To test this possibility, we overexpressed Skp2 in U87 cells. As shown in Fig. 3B, the overexpression of Skp2 resulted in a decrease of RhoE protein levels. This effect was similar to that observed on p27 and supports the notion that Skp2 induced degradation of RhoE. Last, to analyze protein interactions between Skp2 and RhoE, we performed immunoprecipitation assays. The results (Fig. 3C) confirmed that RhoE and Skp2 interacted in cells. In summary, our results thus far indicated that cell proliferation induced the proteasomal degradation of RhoE by Skp2-mediated ubiquitylation.

RhoE Interacts with Skp2 through its C-terminal Region

We next characterized the region of RhoE that is involved in its interaction with Skp2 and necessary for ubiquitylation. We designed a series of deletion and point mutants (Fig. 4) and tested their ability to bind to Skp2 and be ubiquitylated in vitro. We first deleted the N-terminal (residues 1–16) and C-terminal (residues 201–244) regions of RhoE to generate mutants 17–244 and 1–200. As shown in Fig. 5A, both full-length RhoE (1–244) and the N-terminally truncated mutant (17–244) interacted with Skp2. A C-terminal truncated mutant (1–200) did not interact with Skp2 in GST pulldown assays. The full-length protein and N-terminal truncation mutants were ubiquitylated in vitro, whereas a mutant lacking the C-terminal region failed to interact with Skp2 and also was not ubiquitylated (Fig. 5B).

FIGURE 4.

Schematic representation of RhoE and mutants generated in this work. Deletion mutants (residues 17–244, 1–200, 1–230, and 1–240) and specific amino acid mutants (PBR2, PBR3, PBR4, and K235A) are outlined. The sequence of the PBR is indicated. Mutated positions are highlighted in bold type.

FIGURE 5.

RhoE interacts with Skp2 and is ubiquitylated in its C-terminal region. A, protein lysates from 293 cells transfected with empty vector (lane C), full-length RhoE (residues 1–244), or mutants lacking the N-terminal (residues 17–244) or the C-terminal (residues 1–200) region of RhoE were incubated with GST-Skp2 for GST pulldown assays as described under “Experimental Procedures.” RhoE was detected by Western blotting. 40 μg of total extract was used as input. B, full-length RhoE and the mutants described in A were generated by in vitro transcription/translation reactions and incubated with or without a reticulocyte extract (Reti) in in vitro ubiquitylation reactions (see “Experimental Procedures”). RhoE proteins were detected by Western blotting using anti-RhoE antibodies. (Ub)_n-RhoE, polyubiquitylated RhoE. C, constructs expressing full-length RhoE (residues 1–244) and three C-terminal deletion mutants (residues 1–200, 1–230, and 1–240) were transfected into 293 cells. Protein extracts were incubated with GST-Skp2. RhoE was detected by Western blotting as described in A. D, the constructs used in C were tested for in vitro ubiquitylation assays as in B.

To more closely identify the RhoE region necessary for interaction with and ubiquitylation by Skp2, we generated mutants 1–230 and 1–240 that lack the 14 or 4 C-terminal amino acids of RhoE, respectively. The 1–240 mutant interacted with Skp2 (Fig. 5C) and was ubiquitylated (Fig. 5D), suggesting that the last 4 residues of RhoE were not required for Skp2-dependent RhoE degradation. In contrast, the mutant lacking the 14 C-terminal amino acids (1–230) did not interact with Skp2 (Fig. 5C) and also was not ubiquitylated in vitro (Fig. 5D). Our results suggested that the minimal motif necessary for the interaction with Skp2 and ubiquitylation was located between amino acids 231–240 at the C terminus of RhoE.

Lysine 235 Determines the Stability of RhoE

The 231–240 region of RhoE contains four basic residues (one arginine and three lysines). This polybasic region (PBR)⁶ is found in several Rho proteins and believed to be important for lipid modifications to localize Rho proteins to the plasma membrane (10). We generated a PBR4 mutant by changing all four basic residues to alanine (see Fig. 4). This construct was not ubiquitylated in vitro (Fig. 6A), indicating that at least one of the four residues may be subject to ubiquitin ligation. Alternatively, these residues could be important for the interaction with Skp2. To discern between these possibilities, we performed a GST pulldown assay with the PBR4 mutant. As shown in Fig. 6B, the mutant interacted with Skp2 to a similar extent as full-length RhoE protein. This result indicated that the four basic residues in the 231–240 region of RhoE were dispensable for the interaction with Skp2 but necessary for ubiquitylation. We then constructed two more mutants in which 2 (residues 234 and 237; PBR2) or 3 (residues 234, 237, and 239; PBR3) residues were changed to alanine. Both mutants showed the same ubiquitylation pattern as full-length RhoE, indicating that alanine residues 234, 237, and 239 were not required for RhoE ubiquitylation (Fig. 6C).

FIGURE 6.

Basic residues in the PBR of RhoE are dispensable for interaction with Skp2 but necessary for its ubiquitylation. A, wild-type RhoE (1–244) and a PBR4 RhoE mutant lacking all basic residues (Arg²³⁴, Lys²³⁵, Lys²³⁷, and Lys²³⁹) in the 230–240 region (PBR) were generated by in vitro transcription/translation reactions and incubated with or without a reticulocyte extract (Reti) in in vitro ubiquitylation assays (see “Experimental Procedures”). RhoE proteins were detected by Western blotting using anti-RhoE antibodies. (Ub)_n-RhoE, polyubiquitylated RhoE. B, GST pulldown assays were performed with the PBR4 mutant and compared with wild-type (residues 1–244) RhoE. Protein lysates from 293 cells transiently transfected with RhoE constructs were incubated with GST-Skp2 as described under “Experimental Procedures.” RhoE was detected by Western blotting. 40 μg of total extract was used as input. C, wild-type RhoE (1–244) and mutants lacking two (Arg²³⁴ and Lys²³⁷; PBR2), three (Arg²³⁴, Lys²³⁷, and Lys²³⁹; PBR3), or all four (Arg²³⁴, Lys²³⁵, Lys²³⁷, and Lys²³⁹; PBR4) basic residues within the PBR of RhoE were generated in vitro and used in ubiquitylation reactions as described for A.

By eliminating the other basic residues, Lys²³⁵ remained as the only amino acid in RhoE to be modified by ubiquitylation. We next mutated Lys²³⁵ to alanine (K235A; see Fig. 4) and performed in vitro ubiquitylation assays. As shown in Fig. 7A, laddering indicative of polyubiquitylation was markedly reduced in the K235A mutant when compared with the full-length protein. Our result suggested that Lys²³⁵ was the most likely candidate amino acid to be ubiquitylated in RhoE. Last, we wanted to test whether this residue was important for degradation of RhoE in cells. We transfected the K235A RhoE construct into HeLa cells and treated them with cycloheximide. The mutant accumulated in cycloheximide-treated cells and was more stable than wild-type RhoE (Fig. 7B). It has been reported that RhoE degradation requires its membrane association (49). This may suggest that the increased stability of the K235A mutant could be due to its altered localization. However, the RhoE K235A mutant was exclusively localized to the membrane similar to wild-type RhoE (Fig. 7C). Based on this result, we concluded that Lys²³⁵ was the key residue responsible for the short half-life of RhoE and a possible target for Skp2-mediated ubiquitylation leading to its proteasomal degradation.

FIGURE 7.

Lys²³⁵ determines the stability of RhoE. A, a RhoE mutant lacking Lys²³⁵ (K235A) and wild-type protein were generated by in vitro transcription/translation reactions and incubated with or without a reticulocyte extract (Reti) in in vitro ubiquitylation assays (see “Experimental Procedures”). RhoE proteins were detected by Western blotting using anti-RhoE antibodies. (Ub)_n-RhoE, polyubiquitylated RhoE. B, HeLa cells were transfected with pcDNA-FLAG expressing full-length RhoE (residues 1–244) or K235A mutant. After 48 h, cells were treated with 200 μg/ml cycloheximide (CHX). Untreated cells were used as control (lanes C). At the indicated time points, the levels of RhoE proteins were analyzed by Western blotting (top panel). β-Tubulin was used as loading control. The intensities of bands from three independent experiments were quantitated and plotted as the logarithm of the percentage of protein amount for wild-type (residues 1–244, open circles) or mutant (K235A, closed circles) RhoE (bottom panel). Protein levels at time 0 (untreated) were set to 100%. The gray dotted line represents the 50% value. C, membrane localization of the RhoE K235A mutant. HeLa cells were transfected with constructs expressing FLAG-tagged versions of wild-type (residues 1–244) or K235A mutant RhoE. 48 h after transfection, the cells were fixed, stained, and imaged as described under “Experimental Procedures.”

DISCUSSION

RhoE belongs to the Rnd subfamily of Rho proteins. Unlike other GTPases, Rnd proteins lack GTPase activity. Therefore, atypical Rho proteins should be regulated by other means. In this study, we characterized the degradation of RhoE at the proteasome. RhoE degradation was mediated by Skp2, which ubiquitylated RhoE at Lys²³⁵. Ubiquitylation resulted in the rapid degradation of RhoE during G₁/S transition to facilitate cell cycle progression.

Ubiquitylation and proteasomal degradation are established levels of regulation of activated forms of Rho GTPases. The interaction of RhoA with Smurf 1 or Cullin-3 promotes its ubiquitylation and degradation by the proteasome (50, 51). Rac1 also can be ubiquitylated at Lys¹⁴⁷ through a JNK-dependent process to promote its degradation (52). Tumorigenic variants of Rac1 are poorly ubiquitylated, resulting in a higher resistance to proteasomal degradation. Finally, the active form of Cdc42 is able to be ubiquitylated in a CNF-1-dependent manner (37). In all these examples of Rho protein degradation, the main regulatory mechanism remains the GDP/GTP switch. Accordingly, despite the fact that they are degraded at the proteasome, these Rho family members have long half-lives. In contrast, we demonstrate that RhoE has a short half-life. This indicates that although lacking the more common GDP/GTP switch, proteasomal degradation may be an important mechanism by which the expression and activity of RhoE are regulated.

It has been shown previously that RhoE can be degraded at the proteasome, but the mechanisms remained unknown (53). More recently, it has been shown that the stability of RhoE depends on its interaction with effector proteins such as Syx or p190RhoGAP and that Lys⁴⁵ is a critical residue in these interactions (49). Our results demonstrate that proteasomal degradation of RhoE requires interaction with Skp2 through the C-terminal domain and ubiquitylation of Lys²³⁵. Accordingly, a RhoE mutant lacking the C-terminal domain is more stable than wild-type protein (49). When either residue is mutated, the half-life of RhoE increases. This apparent redundancy of RhoE regulation by alternative ubiquitylation at Lys⁴⁵ or Lys²³⁵ may suggest that these residues are modified under specific conditions or in response to different stimuli. Also, the fainter bands seen in our in vitro ubiquitylation assays on some of the RhoE mutants may suggest that the ubiquitylation pattern is altered rather than completely abrogated.

We provide experimental evidence for ubiquitylation of RhoE in vitro, but others failed to detect ubiquitylation in vivo (49). This could be due to a rapid degradation of ubiquitylated RhoE in vivo or to masking interactions with other proteins, as suggested previously (49). In fact, it has been shown that RhoE is also regulated by its interaction with 14-3-3, which sequesters RhoE from the membrane and keeps the protein inactive in the cytosol (54). This interaction may mask RhoE for ubiquitylation. As such, alternative regulatory mechanisms of RhoE function may prevent the induction and/or detection of significant levels of ubiquitylated RhoE in vivo.

Most Skp2 targets are recognized only when phosphorylated (48). RhoE can be phosphorylated by ROCK-I and PKC_α in both the N- and C-terminal regions, and it has been suggested that phosphorylation is relevant for its stability (53, 55). In particular, ROCK-I mediated phosphorylation at Ser¹¹ has been suggested as a key event resulting in the stabilization of (overexpressed) RhoE (53). In contrast, it has been shown that immediate inhibition of ROCK-I protected (endogenous) RhoE from degradation (49). It is also important to point out that RhoE phosphorylation is not fully abrogated by ROCK or PKC inhibitors (55). Finally, we identified the region involved in RhoE degradation as the C terminus not the N terminus. In sum, our findings suggest that: (i) phosphorylation may not be relevant for the mechanism described here, and (ii) other yet to be identified phosphorylation sites could be important for regulating RhoE degradation.

The C-terminal region of Rho proteins plays an important functional role (10). First, the last four amino acids (CAAX sequence) are the substrate for post-translational addition of a lipid tail targeting the proteins to the plasma membrane. Many Rho proteins contain a PBR immediately preceding the CAAX sequence. The PBR is also relevant for interaction with the modifying machinery and cell membrane. In fact, modifications at or next to the PBR such as phosphorylation will alter the intracellular localization of these proteins. In our study, we have identified the addition of ubiquitin chains as a novel modification of the PBR, leading to proteasomal degradation of RhoE. Thus, the PBR in RhoE is not only required for its correct intracellular localization but also its stability. This new function further highlights the importance of the PBR region for the function of Rho proteins. Interestingly, 14-3-3 also binds RhoE in the PBR (54). It may be that the PBR of RhoE becomes available for interaction with Skp2 and ubiquitylation only in the absence of 14-3-3 interaction, leading to the rapid proteasomal degradation of RhoE.

Degradation of RhoE plays a major role in regulating its function in cells. Indeed, our analysis of the expression of RhoE revealed that it changes during the cell cycle. In the context of previous reports showing that overexpression of RhoE inhibits the cell cycle at G₁ (20–22), our results suggest that RhoE expression is needed during the G₁ phase. It also has to be degraded at the G₁/S transition for the cell cycle to proceed into the S phase. In addition, RhoE is induced when cells are physiologically arrested by contact inhibition, as is the case for p27, a cell cycle inhibitor mediating cell cycle arrest under these conditions. Similar to RhoE, Skp2 targets p27 for proteasomal degradation during the late G₁ phase. Finally, expression of RhoE during the G₁ phase may be related to its function as a p53 target in response to DNA damage (24).

Akt, a potent inducer of cell proliferation, induces the degradation of RhoE. There is a complex relationship between Akt, Skp2, and RhoE. Akt activity induces the expression of Skp2, leading to the degradation of RhoE. In contrast, Skp2-mediated ubiquitylation is required for Akt activation (32). It has been suggested that RhoE regulates Akt activation in esophageal squamous cell carcinoma through the PTEN phosphatase (23). It is noteworthy that in our experiments Skp2 appears as a doublet, but RhoE interacts with only one of the bands, suggesting that Skp2-dependent modifications may be required for RhoE interaction (see Fig. 3). This interplay reveals a complex regulatory feedback to control cellular proliferation.

Considering all our data, we propose that RhoE is a mediator of cell cycle arrest after DNA damage and following contact inhibition. RhoE has to be degraded, most likely by Skp2, for the cell cycle to proceed into the S phase. Thus, RhoE becomes an important component in the opposing interplay of Skp2 and p27 levels that control proliferation. From a therapeutic perspective, increasing the stability of RhoE could be a novel mechanism to attenuate proliferation that otherwise may lead to the initiation and progression of cancer.