Nedd4 regulates ubiquitination and stability of the guanine-nucleotide exchange factor CNrasGEF.

Cyclic nucleotide ras GEF (CNrasGEF) is a guanine-nucleotide exchange factor previously isolated in a screen for Nedd4-WW domain interacting proteins (Pham, N., Cheglakov, I., Koch, C. A., de Hoog, C. L., Moran, M. F., and Rotin, D. (2000) Curr. Biol. 10, 555-558). It activates Ras in a cAMP-dependent manner and Rap-1 independent of cAMP. Here we show that CNrasGEF is a likely substrate of the ubiquitin protein ligase Nedd4. CNrasGEF possesses two PY motifs at its C terminus that are responsible for binding to Nedd4 in vitro. Moreover, Nedd4 and CNrasGEF co-immunoprecipitate from 293T cells expressing ectopic CNrasGEF and endogenous Nedd4, and this co-immunoprecipitation is abrogated in PY motif-mutated CNrasGEF (CNrasGEFDelta2PY). CNrasGEF is ubiquitinated in cells, and this ubiquitination is augmented upon overexpression of wt-Nedd4 but is inhibited in cells overexpressing a catalytically inactive Nedd4 (Nedd4(CS)) or in cells expressing CNrasGEFDelta2PY, which cannot bind Nedd4. Moreover, pulse-chase experiments have demonstrated that the half-life of CNrasGEF is reduced 5-fold (from approximately 10 to approximately 2 h) in cells co-expressing Nedd4 with CNrasGEF but not with CNrasGEFDelta2PY (t(0.5) approximately 14 h). CNrasGEF is also stabilized in cells co-expressing Nedd4(CS) or following treatment with lactacystin, indicating proteasomal degradation of this protein. Deletion/mutation of the CDC25 domain to abrogate Ras (or Rap-1) binding leads to impaired ubiquitination of CNrasGEF, suggesting that such binding is critical for ubiquitination. Treatment of cells with the cAMP analogue 8-bromo-cAMP does not affect the ability of CNrasGEF to bind Nedd4 nor its level of ubiquitination, suggesting that Ras binding per se and not its activation is the critical step in triggering ubiquitination of CNrasGEF. These results suggest that CNrasGEF is a substrate for Nedd4, which regulates its ubiquitination and stability in cells.

Nedd4 is an E3 ubiquitin protein ligase comprised of a C2 domain, three or four tandem WW domains, and a ubiquitin ligase Hect domain (1). Our earlier work has demonstrated that Nedd4 binds to the epithelial Na ϩ channel (ENaC) 1 and regulates channel stability at the cell surface (2,3). The association between these two proteins is mediated by binding of the Nedd4-WW domains to conserved PY motifs (PPPxY) found in ENaC (2,4). However, our own work 2 and that of others (e.g. Ref. 5) has suggested that Nedd4 likely has other cellular targets aside from ENaC. In our search for such putative targets, we had performed an expression screen of 16-day-old mouse embryo library with the second Nedd4-WW domain as a probe. This screen resulted in the isolation of a C-terminal fragment of the cyclic nucleotide Ras GEF (CNrasGEF (6); also known as KIAA0313, nRap-GEF1, PDZ-GEF1, and RA-GEF (7)(8)(9)).
CNrasGEF belongs to a family of guanine-nucleotide exchange factors of the small GTPase Ras superfamily (10,11). It has several domains and motifs, including a cyclic nucleotide (cAMP/cGMP)-binding domain (cNMP-BD), Ras exchange motif, PDZ domain, Ras association domain, a CDC25 domain similar to that of Sos1/2, GRF1/2, GRP, as well as Rap GEFs such as Epac, two PY motifs (PPGY and PPDY, see Fig. 1A) and a C-terminal SAV sequence conforming to a PDZ binding motif (6). Ras proteins play a key role in controlling many cellular processes, including proliferation, differentiation, and apoptosis (10 -12). Rap proteins were originally identified by their ability to antagonize Ras signaling but have been shown recently to play overlapping roles with Ras and/or have their own signaling pathways (13). Our recent work has demonstrated that CNrasGEF can activate Ras in vitro and is able to bind cAMP via its cNMP-BD and to activate Ras in cells following treatment with the cAMP analogue 8-Br-cAMP or following elevation of intracellular cAMP with forskolin plus IBMX. This activation is not dependent on protein kinase A (6). CNrasGEF also activates Rap-1, but this activation is independent of cAMP (6 -9). The PDZ domain of CNrasGEF is likely involved in targeting or retaining the protein at the plasma membrane (6).
Because we identified CNrasGEF as a Nedd4-interacting protein, the present work was aimed at investigating the possibility that CNrasGEF is a substrate for Nedd4. Here we report that CNrasGEF binds via its PY motifs to Nedd4, that it is ubiquitinated in a Nedd4-dependent manner, that the stability of CNrasGEF in cells is regulated by Nedd4, and that ubiquitination is reduced in mutant CNrasGEF which can not bind Ras. This work, therefore, suggests that CNrasGEF is a novel substrate for Nedd4 and that Ras binding to CNrasGEF is likely necessary for ubiquitination of this exchange factor.

EXPERIMENTAL PROCEDURES
Construction of Plasmids-Expression cloning of mouse CNrasGEF was previously detailed (6). The corresponding full-length cDNA of human CNrasGEF (KIAA0313) was kindly provided by T. Nagase (Kazusa Institute, Japan) and was used in all subsequent studies. The FLAG-tagged CNrasGEF and GST-CNrasGEF-Cterm (amino acids 1348 -1499) were constructed as previously described (6). The FLAGtagged CNrasGEF was used as a template to generate FLAG-CNrasGEF(⌬PY) with a Y 3 A mutation at the first PY motif (PPGY 3 PPGA) using the QuikChange kit (Stratagene). FLAG-CNrasGEF(⌬PY) was then used as template to construct FLAG-CNrasGEF(⌬2PY) with a Y 3 A mutation in the second PY motif (PPDY 3 PPDA) (Fig. 1A). CNrasGEF lacking its CDC25 domain (⌬CDC25) was constructed as previously described (6). CNrasGEF bearing a mutation in the CDC25 domain, which inhibits Ras binding (R898D), was constructed using the QuikChange kit.
Cell Culture and Transfections-HEK 293T cells were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 100 units of penicillin plus 100 g of streptomycin per ml. Cells were transfected using the calcium phosphate precipitation method as described, with a total of 20 g of DNA per 10-cm dish (equal amounts of DNA per construct). Carrier DNA was added to make up to 20 g of DNA when single transfections were used.
Co-immunoprecipitation Assays-HEK 293T cells were transfected as above. Where indicated, transfected cells were serum-starved overnight and treated with 0.5 mM 8-Br-cAMP (Sigma) for 15 min. Cells were then lysed in lysis buffer as above. Protein concentrations of FIG. 1. Binding of Nedd4 to the C terminus of CNrasGEF. A, alignment of the C-terminal region of CNrasGEF (amino acids 1348 -1499). The partial coding sequence of murine CNrasGEF (mCNrasGEF) isolated from the screen was aligned with its human orthologue (hCNrasGEF) using ClustalW (1.7). The PY motifs are shown in boldface. B, the C terminus of CNrasGEF containing two PY motifs interacts with full-length Nedd4 in a pull-down assay. Equal amount (ϳ50 g) of GST fusion protein containing the C terminus of CNrasGEF (GST-CNrasGEF-Cterm) or GST alone was incubated with 1 mg of HEK 293T cell lysates. Bound proteins were resolved on SDS-PAGE and immunoblotted with anti-Nedd4 antibodies. Lysates (ϳ20 g) of HEK 293T cells were also immunoblotted with either anti-Nedd4 antibodies to detect endogenous Nedd4, or with non-immune serum (control). The bottom panel in B depicts 5 g of the GST or GST-CNrasGEF-Cterm proteins used in the pull-down experiment, stained with Coomassie Blue. different lysates were determined using the Bradford Bio-Rad protein assay (Sigma). Equal amounts of different lysates (ϳ2 mg) were used in immunoprecipitations. FLAG-tagged CNrasGEF was immunoprecipitated using anti-FLAG-M2 agarose (Sigma) for 1.5 h. Immunocomplexes were washed with lysis buffer (2ϫ) and HNTG (3ϫ) and immunoblotted with anti-Nedd4 antibodies (as above) or with anti-Ras antibodies at 1:25,000 dilution (Quality Biotech).
Ubiquitination Assays-HEK 293T cells were transfected as above, and where indicated, were also co-transfected with HA-tagged ubiquitin (HA-Ub) (14). In some experiments, these cells were serum-starved overnight and treated with 0.5 mM 8-Br-cAMP (Sigma) for 15 min. Where indicated, transfected cells were treated (or not) with 10 M lactacystin (Sigma) for 30 min. Cells were then lysed in lysis buffer supplemented with 50 M LLnL (N-acetyl-Leu-Leu-norleucinal, Sigma). Equal amounts of different lysates (ϳ2 mg) were used for the immunoprecipitations. FLAG-tagged CNrasGEF was immunoprecipitated using anti-FLAG-M2 agarose for 1.5 h. Immunocomplexes were washed with lysis buffer (2ϫ) and HNTG (3ϫ) and immunoblotted with anti-HA antibodies to detect HA-ubiquitin conjugation, followed by anti-mouse horseradish peroxidase secondary antibodies and ECL detection. To ensure ubiquitination of CNrasGEF and not of associated proteins, in certain experiments the cell lysates were treated with 1% SDS and boiled for 5 min. These boiled lysates were then diluted 11 times with lysis buffer to dilute the SDS prior to their use in the immunoprecipitation step described above.
Pulse-chase Experiments-HEK 293T cells that had been transfected with the same plasmids were pooled and re-seeded in 10-cm dishes to normalize the variation in transfection efficiency. Cells were washed (3ϫ) with Met/Cys-deficient medium (Life Technologies, Inc.) within 30 min and then incubated in that medium containing 0.1 mCi/ml [ 35 S]Met/Cys (Promix, Amersham Pharmacia Biotech) for 2 h. Cells were washed (3ϫ) with chasing medium (Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 100 units of penicillin plus 100 g of streptomycin per ml, and 10 mM unlabeled Met and Cys) and then chased for the indicated times. They were then washed (2ϫ) with phosphate-buffered saline and lysed in lysis buffer. Protein concentrations of the different lysates were determined using the Bradford Bio-Rad protein assay, and equal amounts of total protein (ϳ1 mg) were incubated with anti-FLAG-M2-agarose to immunoprecipitate CNras-GEF. Precipitated proteins were washed as described above and resolved on SDS-PAGE. Gels were then dried and exposed to an x-ray film. The amount of labeled CNrasGEF was determined by cutting the corresponding areas on the gel and counting radioactivity by liquid scintillation counting. Where indicated, 10 M lactacystin was added (or not) during the 2-h pulse followed by 0-or 3-h chase in the absence or presence of the inhibitor. A 3-h chase period was chosen because prolonged incubation with lactacystin was toxic to the cells.
Ras Binding Assays-GST fusion protein of H-Ras was generated as previously described (6) and immobilized on glutathione-agarose beads (ϳ1 mg of protein/ml of resin). It was first rendered free of nucleotide by incubation for 20 min at 23°C in nucleotide exchange buffer (NEB: 20 mM Tris, pH 7.5, 50 mM NaCl, 5% glycerol, 1 mM dithiothreitol, 0.1% Triton X-100) supplemented with 10 mM EDTA and then incubated in NEB containing 10 mM MgCl 2 plus 200 M GDP. Nucleotide-bound H-Ras was resuspended in 500 l of NEB buffer containing 10 mM MgCl 2 plus 200 M GDP and then combined with 2.5 mg of cell lysates, which were prepared as follows: HEK 293T cells were transfected as indicated above and lysed in Nonidet P-40-lysis buffer (20 mM Tris, pH 7.5, 50 mM NaCl, 1% Nonidet P-40, 50 mM NaF, 10 mM sodium pyrophosphate, 1 mM NaVO 4 , 10 g/ml leupeptin, 10 g/ml aprotinin, 250 M PMSF). Following incubation for 2 h at 4°C, beads were washed (3ϫ) with NEB containing 10 mM MgCl 2 plus 200 M GDP. Bound CNrasGEF were detected with anti-FLAG antibodies on a Western blot.

Interaction of CNrasGEF with Nedd4 in Vitro and in
Cells-In search of proteins that bind to Nedd4, we performed an expression screen of a 16-day mouse embryo library using the second WW domain of rat Nedd4 as a probe. The screen identified a murine cDNA fragment encoding a 149-amino acid polypeptide that has 75% sequence identity (95% sequence similarity) to the C-terminal region of the human GenBank TM entry KIAA0313. We renamed KIAA0313 to CNrasGEF (cyclic nucleotide ras guanine-nucleotide exchange factor) and showed that CNrasGEF is a novel exchange factor that activates Ras in response to elevation of intracellular cAMP or cGMP (6) and Rap-1 independent of cAMP elevation (6 -9).
CNrasGEF contains two conserved PY motifs at its C terminus (Fig. 1A), which we suspected were responsible for binding to the WW domain of Nedd4 in our screen. To confirm the binding, we generated a GST fusion protein containing the C-terminal region of human CNrasGEF (GST-CNrasGEF-Cterm), which corresponds to the murine partial coding sequence isolated in the screen (Fig. 1A). We tested the ability of GST-CNrasGEF-Cterm to bind to endogenous Nedd4 expressed in HEK 293T cells. As shown in Fig. 1B, GST-CNras-GEF-Cterm, but not GST alone, was able to precipitate Nedd4 from 293T cells.
We next tested if CNrasGEF can interact with Nedd4 in cells using co-immunoprecipitation assays. Fig. 2A shows that FLAG-tagged CNrasGEF, ectopically expressed in HEK 293T cells, was able to co-immunoprecipitate endogenous Nedd4 from these cells. However, this co-immunoprecipitation was abolished in cells expressing a mutant CNrasGEF, in which the tyrosines in the PY motifs were mutated to alanines (CNrasGEF⌬2PY) (Fig. 2B), suggesting that the interaction with Nedd4 is mediated via the PY motifs of CNrasGEF. Collectively, these and our previous results (6) show that Nedd4 can bind CNrasGEF in vitro and in cells, and that this interaction is mediated by binding of one or more Nedd4 WW domains to the PY motifs of CNrasGEF.
Nedd4  Nedd4 Regulates Stability of CNrasGEF immunoprecipitated from transfected cells using anti-FLAG antibodies and subsequently immunoblotted with anti-HA antibodies to detect conjugation of HA-Ub. As shown in Fig. 3A  (top panel), a high molecular weight smear representing ubiquitinated CNrasGEF (CNrasGEF-Ub) is clearly apparent in cells expressing CNrasGEF and HA-Ub (lane 3). The smear below the 172-kDa marker in lane 3 probably represents degradation products of ubiquitinated CNrasGEF, because the protein tends to become partially degraded in these cells (Fig.  3A, bottom panel), particularly upon interaction with active Nedd4. Indeed, boiling of the samples in SDS prior to precipitation of CNrasGEF (to ensure these bands do not represent associated proteins of CNrasGEF) revealed the same pattern of ubiquitination as that of untreated samples (Fig. 3C), suggesting that the lower molecular weight ubiquitinated bands are indeed fragments of CNrasGEF that are covalently attached to ubiquitin. To test whether Nedd4 is the E3 involved or responsible for the ubiquitination of CNrasGEF, we overexpressed T7-tagged Nedd4, either wild-type (Nedd4(wt)) or catalytically inactive (Nedd4(CS)), bearing a Cys to Ser mutation at the Hect domain, with CNrasGEF and HA-Ub and analyzed the extent of CNrasGEF ubiquitination in these cells. As seen in To further confirm the involvement of Nedd4 in CNrasGEF ubiquitination, we repeated the ubiquitination experiments, only using CNrasGEF in which one (⌬PY) or both (⌬2PY) tyrosines in the PY motifs were mutated to alanines; the latter double-mutant is unable to bind Nedd4 (Fig. 2). As evident from Fig. 3B (top panel), mutation of the first PY motif (lane 4) did not significantly block ubiquitination of CNrasGEF, but mutating both PY motifs led to an almost complete blockade of CNrasGEF ubiquitination (lane 5), similar to the effect of Nedd4(CS) (Fig. 3A). This suggests that either the second or both PY motifs is/are required for Nedd4-mediated ubiquitination. Taken together, these results demonstrate that ubiquitination of CNrasGEF requires binding to Nedd4, further confirming our conclusion that Nedd4 is the E3 responsible for the ubiquitination of CNrasGEF in cells.
The Stability of CNrasGEF Is Regulated by Nedd4 -Because our results above demonstrate that Nedd4 is responsible for CNrasGEF ubiquitination, and because ubiquitination of proteins usually results in their subsequent degradation, we proceeded to test whether Nedd4 regulates the stability of CNras-GEF using pulse-chase experiments. To this end, transfected CNrasGEF was pulsed with [ 35 S]Met/Cys for 2 h and chased for 0 -18 h (Fig. 4, A and B). Our results show that CNrasGEF expressed in HEK 293T cells has a half-life of ϳ10 h. This half-life was drastically reduced to ϳ2 h upon co-expression of wt-Nedd4 together with CNrasGEF. In contrast, co-expression of Nedd4 with CNrasGEF with mutated PY motifs (CNrasGEF⌬2PY) led to stabilization of the mutant protein, with a half-life of ϳ14 h (Fig. 4B). In accord, CNrasGEF stabilization was also observed in cells co-expressing Nedd4(CS) relative to those co-expressing wt-Nedd4 (Fig. 4A, bottom).
Thus, Nedd4 appears to regulate protein stability of CNrasGEF.
CNrasGEF Is Sensitive to Proteasome Degradation-Several Nedd4 substrates appear to be transmembrane proteins that are targeted for endocytosis and lysosomal degradation follow-ing ubiquitination (e.g. ENaC, numerous yeast amino acid permeases (17)), whereas other, cytoplasmic proteins, appear to be targeted to the proteasome (e.g. Ref. 18). We thus tested whether ubiquitination of CNrasGEF, which is not a transmembrane protein, is sensitive to proteasome inhibitors. Fig. 5 shows that treatment of cells with lactacystin for 30 min enhanced the accumulation of ubiquitinated CNrasGEF (Fig. 5A) and stabilized the protein due to inhibition of the proteasome (Fig. 5B), suggesting that ubiquitinated CNrasGEF is targeted for proteasomal degradation.
Reduced Ubiquitination in CNrasGEF Mutants That Can Not Bind Ras-Wild-type CNrasGEF is able to co-immunoprecipitate Ras and Nedd4 (Fig. 6C), suggesting that the three proteins may form a complex in cells. Interestingly, CNrasGEF lacking its CDC25 (⌬CDC25) domain is unable to bind Ras (Fig.  6B). To test whether Ras binding is involved in CNrasGEF ubiquitination, we analyzed ubiquitination of the ⌬CDC25 mutant of CNrasGEF in cells. As seen in Fig. 6A, ubiquitination of the ⌬CDC25 mutant was severely impaired. Because the CDC25 domain possesses 26 lysine residues, which could serve as potential ubiquitin acceptor sites (hence their loss could theoretically lead to reduced ubiquitination irrespective of Ras binding), we generated a point mutation (R898D) in the CDC25 domain that cannot bind Ras (Fig. 6B). Similar to the ⌬CDC25 mutant, the R898D mutant showed impaired ubiquitination (Fig. 6A). These results suggest that Ras binding is critical for the ubiquitination of CNrasGEF. To test whether Ras binding to CNrasGEF affects the ability of Nedd4 to associate with this exchange factor, we tested the ability of Nedd4 to co-immunoprecipitate with the R898D mutant, which is incapable of binding Ras. Fig. 6D shows that Nedd4 can still bind the R898D mutant, suggesting that Ras binding to the CDC25 domain is not affecting Nedd4 binding to the PY motifs of CNrasGEF. Thus, Ras binding is not required for Nedd4 binding to CNras-GEF, but rather, for its subsequent ubiquitination.
Nedd4-CNrasGEF Association and CNrasGEF Ubiquitination Are Not Dependent on CNrasGEF-mediated Activation of Ras by cAMP-We have previously shown that CNrasGEF activates Ras in response to cAMP in HEK 293T cells (6). We therefore tested whether CNrasGEF activation can influence binding to and regulation by Nedd4. As depicted in Fig. 7A, CNrasGEF potently activates Ras in response to treatment with the membrane permeant cAMP analogue 8-Br-cAMP or treatment with forskolin and IBMX, which elevate intracellular cAMP levels (6). However, 8-Br-cAMP treatment had no effect on the amount of Nedd4 co-immunoprecipitating with CNrasGEF (Fig. 7B) or on CNrasGEF ubiquitination (Fig. 7C). Similar results were also obtained following treatment with Forskolin plus IBMX (not shown). Therefore, the association of CNrasGEF with Nedd4 as well as the Nedd4-mediated CNras-GEF ubiquitination are not dependent on CNrasGEF-mediated activation of Ras in response to cAMP. DISCUSSION We had identified CNrasGEF as an interacting protein of Nedd4 in an expression library screen (6) and have now characterized the interactions between these two proteins. In this report, we show that (i) Nedd4 binds in vitro and in cells to CNrasGEF; this association is mediated by one or more WW domains of Nedd4 (6) and the PY motifs of CNrasGEF; (ii) CNrasGEF is ubiquitinated in cells in a Nedd4-dependent manner, because it requires both active Hect domain of Nedd4 and intact PY motifs of CNrasGEF; (iii) Nedd4 regulates the stability of CNrasGEF in cells; (iv) removal of the Ras (or Rap-1) binding region on CNrasGEF leads to impaired ubiquitination of CNrasGEF; and (v) Binding of Nedd4 to CNras-GEF and the regulation of its ubiquitination are not dependent on cAMP-mediated activation of Ras by CNrasGEF. Only a few studies have investigated regulation of Ras exchange factors by the ubiquitin system. In an earlier report, the yeast Cdc25p was shown to possess a destruction box (DB) similar to that found in mitotic cyclins, which controls Cdc25p stability in vivo. The effect of this DB on Cdc25p ubiquitination was not investigated, however. Moreover, Cdc25p degradation appeared constitutive and not cell cycle-dependent (19). Another report has demonstrated in vitro ubiquitination and shortening of half-life of mSos2, which contains two sequences resembling the DB but not mSos1 (where the equivalent sequences are somewhat different) (20). More recently, GRF2 was shown to become ubiquitinated and destabilized in cells upon elevation of cellular calcium, effects dependent on a DB sequence present between the Ras exchange motif and CDC25 domains of this protein. Moreover, Ras binding to the CDC25 domain of GRF2 seems to be critical for its subsequent ubiquitination (21). In all three cases, a destruction box was implicated in the ubiquitination process, but the E3 involved was not identified.
The work presented here demonstrates ubiquitination-mediated regulation of the stability of the Ras/Rap-1 exchange factor CNrasGEF in cells, which is mediated by a different recognition signal: PY motifs that are recognized by WW domains of the ubiquitin ligase Nedd4. This study thus identifies the E3 involved in the regulation of stability of CNrasGEF.
Our recent work has demonstrated that CNrasGEF activates Ras in cells in response to cAMP (6) and Rap-1 independent of cAMP (6 -9). The present results showing impaired ubiquitination of CNrasGEF by mutants that are unable to bind Ras/ Rap-1 (⌬CDC25, R898D) suggest that Ras or Rap-1 binding is important for the ensuing ubiquitination of CNrasGEF by Nedd4. This is in agreement with the recent demonstration of requirement for Ras binding to GRF2 for its ubiquitination (21) and implicates Ras/Rap-1 binding as an important signal for destruction of these exchange factors. Interestingly, our work shows that stimulation of Ras activation by cAMP is not required for Nedd4 binding to, or Nedd4-mediated ubiquitination of, CNrasGEF. These results suggest that Ras (or Rap-1) binding per se, not the ensuing activation of this protein by CNras-GEF, provides the trigger for ubiquitin-mediated destruction. How the binding of Ras to CNrasGEF leads to recognition by Nedd4 is currently unknown. Clearly, Ras binding to the CDC25 domain is not affecting the ability of Nedd4 to bind the PY motifs of CNrasGEF. Thus, a possible explanation, albeit speculative, may be that Ras binding mediates conformational changes of CNrasGEF that may expose key lysine residues, thus allowing the Hect domain of Nedd4 to attach ubiquitins onto them. A similar scenario may exist for GRF2, whereby Ras binding may allow ubiquitination of exposed lysine residues by an as yet unknown E3 ligase (21).
Nedd4 and its yeast orthologue Rsp5p have been shown to FIG. 6. Reduced ubiquitination of CNrasGEF(⌬CDC25) or CNrasGEF(R898D), which cannot bind Ras. A, 293T cells were transfected with Wt or mutant CNrasGEF that either lack its CDC25 domain (⌬CDC25) or bear a point mutation in the CDC25 domain, which renders it incapable of binding Ras (R898D) (see B). As a control, Wt CNrasGEF was also transfected with catalytically inactive Nedd4 (Nedd4(CS)). Following cell lysis, CNrasGEF was immunoprecipitated with anti-FLAG antibodies, and the immune complex was immunoblotted with anti-HA antibodies to detect ubiquitinated CNrasGEF (CNrasGEF-Ub). The lower panel depicts expression of CNrasGEF or mutated CNrasGEF in lysates corresponding to those in the upper panel. B, 293T cells expressing either Wt CNrasGEF, ⌬CDC25, or R898D mutants were lysed, and the lysate was incubated with GST-Ras to co-precipitate CNrasGEF or its mutant (upper panel). The lower panel represents expression of CNrasGEF or its mutants in the corresponding lysates. C, co-immunoprecipitation of Ras and Nedd4 with CNrasGEF. 293T cells were transfected with FLAG-CNrasGEF and T7-Ras. CNrasGEF was then immunoprecipitated with anti-FLAG antibodies (bottom panel) and presence of co-immunoprecipitated Ras determined with anti-Ras antibodies (upper panel). The blot was then stripped and reprobed with anti-Nedd4 antibodies to test for co-immunoprecipitation of endogenous Nedd4 in the same complex (middle panel). The right-hand panel shows expression of Ras and endogenous Nedd4 in the lysates used for the experiment. Identical expression of Ras (in the right panel) was also seen using anti-Ras antibodies. D, Nedd4 is still able to bind mutant CNrasGEF that cannot bind Ras. FLAG-tagged Wt, ⌬CDC25, or R898D mutants of CNrasGEF, expressed in 293T cells, were immunoprecipitated and co-immunoprecipitated endogenous Nedd4 analyzed with anti-Nedd4 antibodies (upper panel). The lower panel represents expression of CNrasGEF or its mutants in the corresponding lysates. regulate ubiquitination and cell surface stability of several transmembrane proteins (17). Nedd4 is best characterized by its interactions with the epithelial Naϩ channel (ENaC) (2). ENaC is composed of three subunits (␣␤␥), each containing one PY motif (22) that can bind to the Nedd4-WW domains (2,4). Mutations or deletion of the PY motif of ␤ or ␥ ENaC cause Liddle syndrome (reviewed in Ref. 23), a genetic form of hypertension caused by increased retention and activity of the channel at the cell surface (22,24). Nedd4 was demonstrated to be a suppressor of ENaC, which regulates channels numbers at the plasma membrane (3,25), and indeed cell surface stability of ENaC is regulated by ubiquitination (26). This regulation by Nedd4 is impaired in Liddle syndrome due to impaired binding of Nedd4 to ENaC (3) and impaired endocytosis of ENaC (27). Thus, ENaC is a well-documented substrate for Nedd4. Interestingly, recent studies have reported regulation of ENaC by K-Ras-2A, which increases channel activity while decreasing its numbers at the cell surface (28). Because cAMP is a known regulator of ENaC in native tissues (29), it is possible, although speculative, that the association of Nedd4 with CNrasGEF may be somehow involved in connecting ENaC to the Ras pathway.
Although numerous studies have demonstrated regulation of endocytosis and cell surface stability of transmembrane proteins by Nedd4 and Nedd4 family members (e.g. ENaC, the chlorine channel ClC-5, the voltage-gated sodium channel H1, LMP2A, and several yeast amino acid permeases and receptors (30,31;reviewed in Refs. 17,32)), several studies have shown that Nedd4 and its family members regulate stability of intracellular proteins as well. For example, Smurf1 and -2 regulate ubiquitination and stability of Smads proteins involved in transforming growth factor-␤ signaling (33,34), and Rsp5p binds to and regulates stability of RNA polymerase II (18). Thus, although CNrasGEF is localized to the plasma membrane in a PDZ-domain-dependent manner (6), it is nevertheless an intracellular protein. Accordingly, our results here suggest that ubiquitination of CNrasGEF targets it for proteasomal degradation. Thus, Nedd4 can ubiquitinate and target its substrates for either endocytosis/lysosomal degradation or for proteasome degradation.
In summary, we have shown here that Nedd4 regulates FIG. 7. Nedd4 binding to and ubiquitination of CNrasGEF are independent of cAMP-mediated activation of Ras by CNrasGEF. A, activation of CNrasGEF by 8-Br-cAMP. HEK 293T cells were transfected (or not) with FLAGtagged CNrasGEF, serum-starved overnight, and treated (or not) with 8-Br-cAMP. Cells were lysed, and lysates were incubated with immobilized Ras binding domain (RBD) of Raf1 (GST-Raf1-RBD). Co-precipitated active Ras was then detected with anti-Ras antibodies (upper panel). The lower two panels depict the amounts of total endogenous Ras and of the transfected CNrasGEF. B, Nedd4-CNrasGEF co-immunoprecipitation is independent of 8-Br-cAMP treatment. HEK 293T cells were transfected (or not) with FLAG-tagged CNrasGEF, serum-starved overnight, and treated (or not) with 8-Br-cAMP. Cells were lysed, and CNrasGEF was immunoprecipitated with anti-FLAG antibodies.
Co-immunoprecipitated Nedd4 was detected with anti-Nedd4 antibodies. The last lane to the right depicts endogenous expression of Nedd4 in (untransfected) 293T cell lysate. C, CNras-GEF ubiquitination is not dependent of 8-Br-cAMP. HEK 293T cells were transfected with the indicated plasmids, serum-starved overnight and treated (or not) with 8-Br-cAMP. Cells were then lysed, and ubiquitinated CNrasGEF was detected by immunoprecipitation with anti-FLAG antibodies followed by immunoblotting with anti-HA antibodies (upper panel). The lower panel depicts expressed CNrasGEF in lysates from the transfected cells.