Regulation of GTP-binding Protein αq(Gαq) Signaling by the Ezrin-Radixin-Moesin-binding Phosphoprotein-50 (EBP50)*

Although ezrin-radixin-moesin-binding phosphoprotein 50 (EBP50) is a PDZ domain-containing protein known to bind to various channels, receptors, cytoskeletal elements, and cytoplasmic proteins, there is still very little evidence for a role of EBP50 in the regulation of receptor signal transduction. In this report, we show that EBP50 inhibits the phospholipase C (PLC)-β-mediated inositol phosphate production of a Gαq-coupled receptor as well as PLC-β activation by the constitutively active Gαq-R183C mutant. Coimmunoprecipitation experiments revealed that EBP50 interacts with Gαq and to a greater extent with Gαq-R183C. Agonist stimulation of the thromboxane A2 receptor (TP receptor) resulted in an increased interaction between EBP50 and Gαq, suggesting that EBP50 preferentially interacts with activated Gαq. We also demonstrate that EBP50 inhibits Gαq signaling by preventing the interaction between Gαq and the TP receptor and between activated Gαq and PLC-β1. Investigation of the EBP50 regions involved in Gαq binding indicated that its two PDZ domains are responsible for this interaction. This study constitutes the first demonstration of an interaction between a G protein α subunit and another protein through a PDZ domain, with broad implications in the regulation of diverse physiological systems.

EBP50 1 (also known as NHERF1), a 55-kDa phosphoprotein, was first identified as a cofactor essential for protein kinase A-mediated inhibition of Na ϩ /H ϩ exchanger isoform 3 (NHE3) (1). EBP50 contains two PDZ domains (PDZ1 and PDZ2) implicated in multiple protein-protein interactions, and an ERM domain, which binds to the actin-associated ERM proteins (ezrin, radixin, moesin, and merlin) (2,3). EBP50 was also found to interact with a small number of transmembrane proteins such as the cystic fibrosis transmembrane conductance regulator (CFTR) (4), the P2Y1 purinergic receptor (5), the platelet-derived growth factor receptor (6), the ␤ 2 -adrenergic receptor (5), the B1 subunit of the H ϩ -ATPase (7), and the type IIa sodium phosphate cotransporter (8). EBP50 also associates with the phospholipases C (PLC)-␤1/␤2, and with the TRP4 and TRP5 calcium channels to form a PLC-␤1/2-TRP4/5-EBP50 protein complex (9). The physiological role of this interaction on the regulation of PLC-␤1/2 remains undefined. The EBP50 protein can also bind through its PDZ domains to various intracellular proteins, including GRK6A (10), EPI64 (11), and Yes-associated protein 65 (12). A close relative of EBP50 has been identified and is known as E3KARP (13), SIP-1 (14), and NHERF2 (15). EBP50 and NHERF2 share 52% amino acid identity and a conserved domain architecture (13). It has been shown recently that the PDZ domains of EBP50 can homooligomerize and also hetero-oligomerize with the PDZ domains of NHERF2 (16). Despite the growing evidence suggesting the role of EBP50 as a scaffolding protein involved in the formation of signaling complexes, there is still little evidence for a role of EBP50 in the regulation of transmembrane receptor signaling pathways.
Our laboratory has been interested in the regulation of the thromboxane A 2 (TXA 2 ) receptors (TP receptors). TXA 2 has a variety of pharmacological effects that modulate the physiological responses of several cells and tissues (17). Binding of TXA 2 to its receptor induces platelet aggregation, constriction of vascular and bronchiolar smooth muscle cells, as well as mitogenesis and hypertrophy of vascular smooth muscle cells (18). The TP receptor is part of the G protein-coupled receptor (GPCR) superfamily. Two TP receptor isoforms were identified that are generated by the alternative splicing of a single gene, TP␣ (343 amino acids) and TP␤ (407 amino acids), which share the first 328 amino acids (19,20). Different experiments demonstrated that agonist-induced production of the second messenger inositol phosphates by the TP receptors results from the activation of the G q/11 family of the G␣ subunits (17). G␣ q -mediated production of inositol phosphates involves the stimulation of PLC-␤ isoforms in the following order of potency: PLC-␤1 Ն PLC-␤3 Ͼ PLC-␤2 (21). Furthermore, PLC-␤ isoenzymes accelerate the intrinsic GTP hydrolysis by G␣ subunits acting as a GTPase-activating protein (GAP) (22). More recently, it was shown that PLC-␤s form dimers in vivo and suggested that dimerization could mediate their interaction with GTP-bound G␣ q (23).
Although the signaling pathways of the TP receptors are being described in increasing detail, their regulation remains to be characterized. Knowing that the TP receptor signaling cascade involves the activation of G␣ q and the stimulation of the downstream effector PLC-␤, we investigated the effect of EBP50 on the signaling of the agonist-stimulated TP receptor through the G␣ q pathway. Surprisingly, our experiments revealed that EBP50 regulates the G␣ q signaling pathway by preferentially binding through its PDZ domains to the activated G␣ q and by preventing its interaction with PLC-␤. Furthermore, we show that EBP50 also interferes in the TP receptor-G␣ q coupling. This novel interaction between PDZ domains and G␣ q has broad implications in intracellular signaling regulation. Growing evidence demonstrates the role of EBP50 in a wide variety of physiological events. The regulation of these events could possibly be modulated by the G␣ q -EBP50 interaction.

EXPERIMENTAL PROCEDURES
Reagents-The Myc, G␣ q/11 , and PLC␤1-specific polyclonal antibodies were from Santa Cruz Biotechnology. Myc-specific monoclonal antibody was a gift from Dr. J. Stankova (Université de Sherbrooke). Hemagglutinin (HA)-specific monoclonal antibody was from Babco, and EBP50-specific monoclonal antibody was from BD Biosciences. ECL reagents were purchased from Amersham Biosciences. Protein A-agarose and FuGENE 6 TM were purchased from Roche Molecular Biochemicals.
Cell Culture and Transfection-HEK293 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum at 37°C in a humidified atmosphere containing 5% CO 2 . Transient transfections of HEK293 cells grown to 75-90% confluence were performed using FuGENE 6 TM according to the manufacturer's instructions. Empty pcDNA3 vector was added to keep the total DNA amount added per plate constant.
Immunoprecipitations-6-Well plates of HEK293 cells were transfected with pcDNA3-EBP50-HA, pcDNA3-G␣ q , pcDNA3-G␣ q -R183C, pcDNA3-Myc-TP␣, pcDNA3-Myc-TP␤, pcDNA3-PDZ1-HA, pcDNA3-PDZ2-HA, pcDNA3-PDZ1-PDZ2-HA, pcDNA3-PDZ2-ERM-HA, and pcDNA3-ERM-HA, in the different combinations indicated under the "Results." Transfected cells were maintained as described above for 48 h. The cells transfected with TP␣ or TP␤ were incubated for 0 -30 min at 37°C in the presence or the absence of 100 nM U46619 prior to harvesting. The cells were then rinsed with ice-cold phosphate-buffered saline and harvested in 800 l of lysis buffer (150 mM NaCl, 50 mM Tris, pH 8, 1% Nonidet P-40,0.5% deoxycholate, 0.1% SDS, 10 mM Na 4 P 2 O 7 , 5 mM EDTA) supplemented with protease inhibitors (9 nM pepstatin, 9 nM antipain, 10 nM leupeptin, and 10 nM chymostatin) (Sigma). After the cells were incubated in lysis buffer for 60 min at 4°C, the lysates were clarified by centrifugation for 20 min at 14,000 rpm at 4°C. One to four g of specific monoclonal or polyclonal antibodies were added to the supernatant. After 60 min of incubation at 4°C, 50 l of 50% protein A-agarose pre-equilibrated in lysis buffer was added, followed by an overnight incubation at 4°C. Samples were then centrifuged for 1 min in a microcentrifuge and washed three times with 1 ml of lysis buffer. Immunoprecipitated proteins were eluted by addition of 50 l of SDS sample buffer followed by a 30-min incubation at room temperature. Initial lysates and immunoprecipitated proteins were analyzed by SDS-PAGE and immunoblotting using specific antibodies.
Inositol Phosphates Measurement in Cells-Inositol phosphates measurements were performed as described previously (24). 2 ϫ 10 5 HEK293 cells were grown overnight in 12-well plates. The cells were then cotransfected as described above with the indicated constructs. The cells were labeled the following day for 18 -24 h with 4 Ci/ml of myo-[ 3 H]inositol in Dulbecco's modified Eagle's medium (high glucose, without inositol). The cells were washed once in phosphate-buffered saline and incubated in pre-warmed Dulbecco's modified Eagle's medium (high glucose, without inositol) supplemented with 0.5% bovine serum albumin, 20 mM Hepes, pH 7.5, and 20 mM LiCl for 10 min. Cells were then stimulated for 5-30 min with 500 nM U46619 in the case of the TP receptor-transfected cells. After stimulation, the medium was removed, and the reactions were terminated by addition of 0.8 ml of 0.4 M chilled perchloric acid. Cells were than collected in Eppendorf tubes, and 0.5 volume of 0.72 N KOH, 0.6 M KOHCO 3 was added. Tubes were mixed and centrifuged for 5 min at 14,000 rpm in a microcentrifuge. Inositol phosphates were separated on Bio-Rad AG 1-X8 columns. Total labeled inositol phosphates were then measured by liquid scintillation counting.
Construction of Epitope-tagged EBP50 Domains-The pcDNA-EBP50-HA construct was a kind gift of Dr. Mark Von Zastrow (University of California, San Francisco). The different HA-tagged domains of EBP50 (PDZ1, PDZ2, ERM, PDZ1-PDZ2, and PDZ2-ERM) were obtained by the use of polymerase chain reaction using the Expand High Fidelity PCR System (Roche Molecular Biochemicals), subcloned in the pcDNA3 vector, and the integrity of the coding sequence was confirmed by dideoxy sequencing.
Recombinant EBP50 Production-cDNA encoding for EBP50 was subcloned in the pRSET A vector (Invitrogen), and the construct was used to produce a His 6 -tagged recombinant EBP50 in the Escherichia coli K-12 PR745 by following the manufacturer's instruction. The recombinant EBP50 was purified using the QIAexpressionist TM kit (Qiagen Inc, Canada) as indicated by the manufacturer. The purified recombinant EBP50 was analyzed by SDS-PAGE and was immunoblotted by using an EBP50-specific monoclonal antibody as described above. The purified recombinant EBP50 was later used in the G␣ q interaction and GTPase assays.
Binding Assay-The His-EBP50-G␣ q -R183C binding assay was performed using recombinant EBP50 produced as described above. His-EBP50 (50 g) was incubated for 30 min at room temperature with 100 l of Ni-NTA resin (Qiagen, Canada) in binding buffer (20 mM Tris-HCl, pH 8, 2 mM MgSO 4 , 6 mM ␤-mercaptoethanol, 100 mM NaCl, 0.05% Nonidet P-40, 5% glycerol). Thereafter, the resin was washed three times with the binding buffer. The Ni-NTA-bound EBP50 was then incubated for 1 h at 4°C with cellular extracts from HEK293 cells transfected with either pcDNA3-G␣ q -R183C or with pcDNA3 in binding buffer. The binding reactions were then washed three times with binding buffer. SDS sample buffer was added to the binding reactions, and the tubes were boiled for 5 min. The binding reactions were analyzed by SDS-PAGE and immunoblotting using the indicated specific antibodies.
GTPase Assay-G␣ GTPase activity was assessed in solution using a single turnover assay essentially as described previously (25). After incubation of G␣ q -R183C in the presence of [␥-32 P]GTP, the GTP-loaded G␣ q -R183C was filtered in G-25 Sephadex. Thereafter, G␣([␥-32 P]GTP) was incubated in the absence or the presence of RGS4 (100 nM), EBP50 (100 nM), EBP50 (250 nM), or control buffer. Reactions were quenched with 9 volumes of 5% (w/v) Norit A Charcoal in 50 mM NaH 2 PO 4 . The charcoal was pelleted, and the 32 P i -containing supernatant was counted.

EBP50 Inhibits TP Receptor-mediated PLC-␤ Activation-To
investigate the effect of EBP50 on TP(␣/␤) signaling, we measured total inositol phosphate production in HEK293 cells cotransfected with either pcDNA3-HA-TP␣ or pcDNA3-HA-TP␤ with an empty pcDNA3 vector (100% control) or pcDNA3-EBP50 following a U46619 (a stable TP receptor agonist) stimulation. GRK2 has been shown to inhibit the inositol phosphate response of both TP receptor isoforms (26) and was thus included in these experiments as a positive control. Fig. 1A shows that EBP50 expression results in a significant loss in the agonist-induced inositol phosphate production by both TP␣ and TP␤, reaching only 41 and 45% of their maximal response, respectively. Receptor expression was determined to be 1.2 Ϯ 0.2 pmol of receptor/mg of protein for each construct combination shown in Fig. 1A.
Inhibition of G␣ q -mediated Signaling by EBP50 -Because the TP receptors activate PLC-␤ through the G␣ q protein, we looked at whether EBP50 could directly inhibit G␣ q signaling. G␣ q -R183C, a GTPase-deficient G␣ q mutant that leads to the constitutive activation of PLC-␤ (27), was cotransfected with either an empty pcDNA3 vector (100% control), EBP50, GRK2, GRK5, or GRK6 in HEK293 cells. Although it was shown that GRK2 inhibited G␣ q -R183C mediated activation of PLC-␤, GRK5, and GRK6 on the other hand, showed no such effect (26). GRK2, GRK5, and GRK6 were thus used as experimental controls. Surprisingly, EBP50 inhibits significantly the inositol phosphate production induced by G␣ q -R183C-mediated activation of PLC-␤. Indeed, we observed a loss of more than 60% of the inositol phosphate production when G␣ q -R183C was cotransfected with EBP50 as compared with its transfection with an empty vector (Fig. 1B). These results suggest that EBP50 mediates the inhibition of the production of inositol phosphates following TP␣ and TP␤ stimulation by preventing the TP(␣/␤)activated G␣ q from activating PLC-␤.
EBP50 Preferentially Interacts with Activated G␣ q -EBP50 is a PDZ-domain containing protein known for its ability to bind to a wide variety of other proteins. We were thus interested in determining if EBP50 could interact with G␣ q . Consequently, coimmunoprecipitation experiments were performed in HEK293 cells cotransfected with pcDNA3-EBP50-HA and either G␣ q , G␣ q -R183C, or empty pcDNA3. Cell lysates were then incubated with a G␣ q -specific polyclonal antibody. Blotting of the immunoprecipitation reactions with an HA-specific monoclonal antibody revealed that EBP50-HA could be coimmunoprecipitated with G␣ q and G␣ q -R183C. However, the amount of EBP50-HA that coimmunoprecipitated with G␣ q -R183C was greater than with G␣ q (Fig. 2A, upper panel). The presence of a faint EBP50-HA band in absence of overexpressed G␣ q indicates that endogenous G␣ q proteins are sufficient for coimmunoprecipitation of EBP50. Immunoblotting results of the cell lysates with the HA-specific antibody showed that EBP50-HA expression levels were similar in the cells trans-fected with pcDNA3-EBP50-HA (data not shown). The G␣ q -EBP50 interaction was further verified in cell lysates immunoprecipitated with an EBP50-specific antibody, and the resulting material was blotted with the G␣ q -specific antibody to visualize the coprecipitated G␣ q (Fig. 2B, upper panel). As it can be seen, G␣ q -R183C and G␣ q could be immunoprecipitated with either endogenous or transfected EBP50. As expected from Fig. 2A, G␣ q -R183C coimmunoprecipitated with EBP50 in a larger amount than the wild type G␣ q . The equivalent expression of EBP50 in the cell lysates was confirmed by immunoblotting using an EBP50-specific monoclonal antibody (data not shown). The EBP50-G␣ q -R183C interaction was further confirmed by running HEK293 cell extracts expressing G␣ q -R183C over His-tagged EBP50-purified protein bound to a Ni 2ϩ column (Fig. 3). The upper panel of Fig. 3 shows that G␣q-R183C binds to EBP50, whereas it does not bind to the free Ni-NTA resin. Furthermore, no binding was observed when cellular extracts from HEK293 transfected with pcDNA3 were used. The lower panel of Fig. 3 shows immunoblotting of the binding reactions with anti-EBP50 monoclonal antibody.
TP␤ Promotes the G␣ q -EBP50 Interaction-The results de-FIG. 1. Inhibition of G␣ q signaling by EBP50. A, HEK293 cells expressing either the TP␣ or TP␤ receptors along with pcDNA3 vector (control), GRK2 (positive control), or EBP50 constructs were metabolically labeled with myo-[ 3 H]inositol and then stimulated with 500 nM U46619 for 5 min. Total [ 3 H]inositol phosphates were isolated as described under "Experimental Procedures," measured by liquid scintillation counting, and expressed as a percent of control. B, HEK293 cells expressing G␣ q -R183C along with vector (control), GRK2 (positive control), GRK5, and GRK6 as negative controls and EBP50 constructs were metabolically labeled with myo-[ 3 H]inositol. Total [ 3 H]inositol phosphate production was measured as described under "Experimental Procedures" and plotted as above. All values are mean Ϯ S.E. from 3 to 10 separate experiments. Data were analyzed by using the Student's t test (*, p Ͻ 0.05 compared with cells transfected with receptor and pcDNA3). scribed in Fig. 2 showed that EBP50 coimmunoprecipitated preferentially with G␣ q -R183C (a constitutively active form of G␣ q ) than it did with G␣ q (the assumed "inactive" form). Thus, our data strongly suggest that the interaction of EBP50 with the activated form of G␣ q could be implicated in the mechanism leading to the inhibition of the agonist-induced inositol phosphate production by TP␤. To verify this, immunoprecipitation experiments were performed using HEK293 cells transfected with pcDNA3-Myc-TP␤ and pcDNA3-EBP50-HA. The cells were incubated with the TP receptor agonist U46619 (500 nM) for different incubation times from 0 to 30 min. Cell lysates were then incubated with the G␣ q -specific antibody, and the immunoprecipitation reactions were assessed by immunoblotting with the HA-specific monoclonal antibody. As shown in Fig. 4, only a small amount of EBP50 could be coimmunoprecipitated with G␣ q in the absence of TP␤ agonist stimulation, consistent with results seen in Fig. 2A. However, a greater amount of EBP50 could be coimmunoprecipitated with G␣ q after a 5-min TP␤ stimulation, suggesting that the agonist treatment resulted in the activation of endogenous G␣ q leading to an increased interaction with EBP50 (Fig. 4, upper panel). Basal levels of EBP50 coimmunoprecipitated with G␣ q when TP␤ was stimulated for longer periods, suggesting an inactivation of GTP-bound G␣ q and a loss of G␣ q -EBP50 interaction.
Our data indicate that activation of G␣ q by a GPCR results in a transiently increased G␣ q -EBP50 interaction. Similarly, activation of G␣ s by the TP and ␤ 2 -adrenergic receptors showed that this G␣ protein subunit could also bind transiently to EBP50 when activated, whereas other G␣ proteins tested so far (G␣ i/o/t/z , G␣ 12/13 , and G␣ 16 ) failed to interact with EBP50 in our system, 2 showing a specificity in the EBP50 interaction with G␣ subunits. HEK293 cells coexpressing EBP50-HA and either G␣ q -R183C or G␣ q were harvested and lysed as described under "Experimental Procedures." A, immunoprecipitation (IP) experiments from cell extracts were performed by incubating with a G␣ q -specific polyclonal antibody followed by incubation with protein A-agarose. Immunoprecipitated proteins were eluted from protein A-agarose with SDS sample buffer, and elutions and cell extracts were subjected to SDS-PAGE. Immunoblotting (IB) was performed with an HA-specific monoclonal antibody (upper panel). The amount of immunoprecipitated G␣ q in each samples was verified by immunoblotting with the G␣ q -specific polyclonal antibody (lower panel). B, immunoprecipitation experiments were carried out as above using an EBP50-specific monoclonal antibody and immunoblotting with the G␣ q -specific polyclonal antibody (upper panel). The amount of G␣ q present in each cell extract was evaluated by immunoblotting with the G␣ q -specific polyclonal antibody (lower panel). Representative Western blots are shown. cells were cotransfected with different combinations of pcDNA3-Myc-TP␤, pcDNA3-Myc-TP␣, pcDNA3-G␣ q , pcDNA3-EBP50-HA, and pcDNA3 as described in Fig. 5. The cells were then stimulated or not with 500 nM U46619 for 5 min, and the cell lysates were incubated with a Myc-specific monoclonal antibody. Blotting of the immunoprecipitation reactions with the G␣ q -specific polyclonal antibody revealed that, in the absence of EBP50 overexpression, G␣ q coimmunoprecipitated with TP␣ and TP␤ demonstrating physical coupling of the receptors to G␣ q (Fig. 5, upper panel). After agonist stimulation of TP␤, the amount of coimmunoprecipitated G␣ q decreased due to the activation and the subsequent uncoupling of G␣ q . This uncoupling was observed in a lesser extent for the TP␣ receptor. These observations argue for differential TP␣/␤ coupling to G␣ q . This difference is not surprising because TP␣ and TP␤ are also subject to differential regulation of other events such as phosphorylation (17) or constitutive and agonist-induced internalization (24,28). However, in the HEK293 cells transfected with either TP␤ or TP␣ with G␣ q and EBP50, EBP50 seemed to inhibit the coupling of both receptors to G␣ q , because we observed a significant decrease in the amount of G␣ q that could be coimmunoprecipitated with both receptors.
Once more, the negative effect of EBP50 on TP␤ coupling to G␣ q was more significant than on the TP␣ coupling to G␣ q . Thus, it appears that EBP50 interferes in the receptor-G␣ q interaction.
EBP50 Prevents Activated G␣ q from Binding to PLC-␤1-It is well known that the GTP-bound (active) form of G␣ q enhances the phospholipase activity of PLC-␤1 by directly binding to its C2 carboxyl-terminal domain (29). In order to investigate whether the EBP50-(GTP-G␣ q ) interaction results in a sequestration of the activated G␣ q , thus preventing it from activating the downstream effectors such as PLC-␤1, we performed immunoprecipitation experiments of endogenous PLC-␤1 in HEK293 cells cotransfected with pcDNA3-G␣ q -R183C or pcDNA3-G␣ q and either pcDNA3-EBP50-HA or an empty pcDNA3. The cell lysates were then incubated with a PLC-␤1-specific polyclonal antibody. Blotting of the immunoprecipitation reactions with a G␣ q -specific antibody revealed that a large amount of G␣ q -R183C could be coimmunoprecipitated with PLC-␤1 in the absence of transfected EBP50 (Fig.  6A). As we expected, these data indicate that G␣ q -R183C interacts with PLC-␤1 resulting in the inositol phosphate production shown in Fig. 1B. The interaction between G␣ q -R183C and PLC-␤1 was almost completely inhibited when EBP50-HA was present. This result demonstrates that EBP50 interferes with the G␣ q -dependent activation of PLC-␤1 by preventing the GTP-bound G␣ q -PLC-␤1 interaction. It can also be observed that only G␣ q -R183C, and not the inactive form of G␣ q , could bind to PLC-␤1 (Fig. 6A). Fig. 6D demonstrates that EBP50 also binds to PLC-␤1, as was recently shown by Tang et al. (9).
EBP50 Binds to G␣ q -R183C through Its PDZ Domains-We were next interested in studying the EBP50 domains involved in the interaction with activated G␣ q . Different deletion constructs of EBP50 were thus generated, as schematically represented in Fig. 7A. HEK293 cells were transfected with pcDNA3-G␣ q -R183C and either of the indicated EBP50 mutant constructs. Cell lysates were then incubated with the G␣ qspecific antibody, and the immunoprecipitation reactions were analyzed by blotting with an HA-specific monoclonal antibody.  4. Coimmunoprecipitation of EBP50 with G␣ q following TP␤ stimulation. HEK293 cells coexpressing MycTP␤ and EBP50-HA were incubated with 500 nM U46619 for the indicated times. Immunoprecipitation experiments were carried out as described in Fig. 2A using a G␣ q -specific polyclonal antibody to precipitate the endogenous G␣ q , and immunoblotting (IB) was performed with an HA-specific monoclonal antibody (upper panel). The amount of immunoprecipitated (IP) G␣ q in each sample was verified by immunoblotting with the G␣ qspecific polyclonal antibody (lower panel).

FIG. 7. Identification of the EBP50 domains interacting with G␣ q -R183C.
A, schematic representation of the different EBP50 constructs that were generated for the identification of the domains responsible for the G␣ q -R183C interaction. B, HEK293 cells coexpressing different HA-tagged EBP50 domains and G␣ q -R183C were harvested and lysed as described under "Experimental Procedures." The cell lysates were incubated with a G␣ q -specific polyclonal antibody followed by incubation with protein A-agarose. The immunoprecipitation (IP) reactions were analyzed by immunoblotting (IB) using an HA-specific monoclonal antibody (upper panel). Expression of each of the EBP50s was confirmed in the cell lysates by immunoblotting with the HA-specific monoclonal antibody (middle panel). Equivalent immunoprecipitation of G␣ q -R183C in the reactions was verified by Western blotting with a G␣ q -specific polyclonal antibody (lower panel).

FIG. 6. EBP50 inhibits the interaction of G␣ q -R183C with PLC-␤1.
HEK293 cells expressing G␣ q -R183C alone or coexpressing G␣ q -R183C and EBP50-HA were harvested and lysed as described under "Experimental Procedures." The cell lysates were incubated with a PLC-␤1-specific polyclonal antibody followed by incubation with protein A-agarose. The immunoprecipitation (IP) reactions were analyzed by Western blotting with a G␣ q -specific polyclonal antibody. The results showed that EBP50 prevented the binding of G␣ q -R183C to PLC-␤1 (A). B and C represent the immunoblotting (IB) of the cell extracts with G␣ q -and PLC-␤1-specific polyclonal antibody showing equal amounts of G␣ q -R183C and PLC-␤1, respectively, present in the cell extracts. D, coimmunoprecipitation of EBP50-HA with PLC-␤1 as detected by Western blotting of the same immunoprecipitation reactions with the HA-specific monoclonal antibody.
The results obtained have shown that the ERM domain does not coimmunoprecipitate with G␣ q -R183C (Fig. 7B). On the other hand, PDZ1 and PDZ2 coimmunoprecipitated with G␣ q -R183C, indicating that these domains are implicated in the EBP50-G␣ q -R183C interaction. However, the PDZ1-PDZ2 construct seems to coimmunoprecipitate in a greater amount than did each individual PDZ domain (PDZ1 and PDZ2) or the PDZ2-ERM mutant and to a similar level as the full-length EBP50 (Fig. 7B, upper panel). Thus, it appears that the PDZ1-PDZ2 domains are responsible for the full interaction with activated G␣ q .
PDZ1 and PDZ2 Interfere with the Physical Coupling of the TP Receptor to G␣ q and with the Activation of PLC-␤1 by G␣ q -R183C-EBP50 is thought to be an important link for several membrane proteins to the actin cytoskeleton through its ERM domain, which is involved in the binding to the cytoskeleton-binding ERM proteins. The results described above have shown that the PDZ1 and PDZ2 domains are involved in the EBP50-(G␣ q -R183C) interaction, whereas the ERM domain is not. However, we wondered if the ERM domain of EBP50 is important for its inhibition of the G␣ q binding to PLC␤1 and the TP receptor. HEK293 cells were transfected with pcDNA3-G␣ q -R183C and either pcDNA3-PDZ1-HA, pcDNA3-PDZ2-HA, or pcDNA3-ERM-HA. The cell lysates were then incubated with a PLC-␤1-specific polyclonal antibody. The immunoprecipitation reactions were analyzed by blotting with the G␣ qspecific antibody. Surprisingly, the PDZ1 and PDZ2 domains inhibited the binding of G␣ q -R183C to PLC-␤1, whereas the ERM domain did not (Fig. 8A).
The effect of each EBP50 domain on the interaction between TP␤ and G␣ q was then investigated. HEK293 cells were trans-fected with pcDNA3-Myc-TP␤, pcDNA3-G␣ q , and either pcDNA3-EBP50-HA, pcDNA3-PDZ1-HA, pcDNA3-PDZ2-HA, or pcDNA3-ERM-HA. The cell lysates were then incubated with a Myc-specific monoclonal antibody, and the immunoprecipitation reactions were analyzed by immunoblotting with the G␣ q -specific antibody. The results obtained showed that EBP50, PDZ1, and PDZ2 could inhibit the physical coupling of TP␤ to G␣ q , whereas the ERM showed no effect (Fig. 9A). Taken together, these results have shown that the PDZ1 and the PDZ2 domains are sufficient to regulate the coupling of TP␤ to G␣ q as well as the binding of the activated G␣ q to PLC-␤1.
EBP50 Has No GAP Activity for G␣ q -Finally, we wanted to assess if EBP50 showed any GAP activity for G␣ q . GDP dissociation, and therefore GTP loading, in the absence of activated GPCR is particularly inefficient compared with k cat for GTP hydrolysis for G␣ q precluding use of the single turnover assay (30). However, the k cat for GTP hydrolysis of G␣ q -R183C is significantly reduced allowing GTP loading to occur more efficiently (31). Furthermore, it was recently shown that the GTPase activity of G␣ q -R183C can be promoted by RGS4 in a single turnover assay (31). Thus, we utilized this assay to monitor the GTPase activity of G␣ q -R183C in the absence or presence of purified RGS4 or EBP50. Although 100 nM RGS4 promoted rapid GTP hydrolysis releasing up to 7-8 fmol of P i , FIG. 8. Inhibition of the interaction between G␣ q -R183C and PLC-␤1 by the individual EBP50 domains. HEK293 cells were cotransfected with G␣ q -R183C and either pcDNA3, PDZ1-HA, PDZ2-HA, or ERM-HA. The cells were then harvested and lysed as described under "Experimental Procedures." The cell lysates were then incubated with a PLC-␤1-specific polyclonal antibody followed by incubation with protein A-agarose. A, Western blotting of the immunoreaction samples with a G␣ q -specific polyclonal antibody. B and C, equal amounts of PLC-␤1 and G␣ q -R183C could be detected in the cell extracts using the PLC-␤1-specific and the G␣ q -specific polyclonal antibodies, respectively. D, the expression of EBP50-HA and the different HA-tagged EBP50 domains was confirmed by Western blotting of the cell extracts with an HA-specific monoclonal antibody. IP, immunoprecipitation; IB, immunoblot.
FIG. 9. TP␤ receptor-G␣ q interaction blockade by the PDZ1 and PDZ2 domains of EBP50. HEK293 cells expressing Myc-TP␤ and G␣ q and either EBP50-HA, PDZ1-HA, PDZ2-HA, or ERM-HA domains of EBP50 were harvested and lysed as described under "Experimental Procedures." The cell lysates were then incubated with a Mycspecific monoclonal antibody followed by incubation with protein A-agarose. A, Western blotting of the immunoprecipitation (IP) reactions with a G␣ q -specific polyclonal antibody showed that EBP50 and its first and second PDZ domains, but not the ERM domain, interfere with G␣ q binding to the TP␤ receptor. B, equivalent expression of G␣ q in cell extracts was verified for each reaction by blotting with the G␣ q -specific polyclonal antibody. C, expression of each of the HA-tagged EBP50 constructs in the cell lysates was confirmed by immunoblotting (IB) with the HA-specific monoclonal antibody. D, quantities of immunoprecipitated Myc-TP␤ receptors were the same for all reactions as revealed by Western blotting with the Myc-specific monoclonal antibody. 100 and 250 nM EBP50 failed to enhance GTP hydrolysis over 15 min of incubation (Fig. 10). Thus, it appears that EBP50 has no ability to function as a GAP for G␣ q . DISCUSSION This study demonstrates a direct regulation of the G␣ q signaling pathway by EBP50. We have shown that EBP50 inhibited the inositol phosphate production induced by TP␣ and TP␤ agonist stimulation. We showed that EBP50 decreased PLC-␤dependent production of inositol phosphates induced by G␣ q -R183C, a constitutively active mutant of G␣ q . We demonstrated that EBP50 interacts in a greater extent with G␣ q -R183C than with G␣ q . Moreover, stimulation of the TP␤ receptor promoted the G␣ q -EBP50 interaction, strongly indicating that EBP50 interacts better with the activated G␣ q . It was observed that, beyond 5 min of agonist stimulation of TP␤, the amount of EBP50 that coimmunoprecipitated with G␣ q decreased to levels similar to what was observed in the absence of agonist. It is known that the activation of G␣ q by GPCRs is a rapid process involving the release of GDP by G␣ q and its exchange for a GTP. Thereafter, rapid hydrolysis of the GTP by the intrinsic GTPase activity of G␣ q results in the inactivation of the G␣ q subunit and its return to the inactive GDP-bound state. In addition, rapid desensitization of GPCRs occurs after agonist stimulation resulting in their inability to activate G protein subunits. Taken together, this could explain the loss of EBP50-G␣ q interaction after prolonged stimulation of TP␤. These results strongly suggest that the regulation of the TP receptors signaling by EBP50 is caused by the binding of the activated G␣ q to EBP50. However, because activation of the G␣ subunit results in its dissociation from the ␤␥ subunits, the strong binding of the active form of G␣ q to EBP50 could be due to its dissociation from the ␤␥ subunits and not to its GTPbound active state per se. Arguing against the last statement is the fact that the overexpressed inactive G␣ q did not interact as well as G␣ q -R183C with EBP50, unless the endogenous ␤␥ subunits can account for the difference. Furthermore, the activated G␣ q -EBP50 interaction after receptor stimulation is transient, suggesting that the G␣ q comes off the EBP50 protein upon GTP hydrolysis, which could indicate that EBP50 indeed binds preferentially to the active form of G␣ q . More experiments need to be performed to verify if the ␤␥ subunits compete with EBP50 for the binding of G␣ q .
Our results showed that EBP50 interferes with the coupling of the TP receptors to G␣ q which could be explained by the binding of EBP50 to both the active and the inactive forms of G␣ q leading to the possible sequestration of G␣ q . It can be argued that EBP50 could also bind directly to the receptor, as reported for other GPCRs (5), and further interfere in the G␣ q coupling this way. The EBP50-TP␤ receptor interaction is currently studied in our laboratory.
GTP-bound G␣ q preferentially activates PLC-␤1 as described in the Introduction. Singer et al. (23) have shown that the GTP bound-G␣ q directly binds and activates PLC-␤. Because on the one hand EBP50 inhibited the inositol phosphate production and on the other hand had a greater affinity for the active form of G␣ q , the effect of EBP50 on the G␣ q -PLC-␤1 interaction was investigated. We observed an almost complete inhibition in the interaction of G␣ q -R183C with PLC-␤1 in the presence of EBP50. Taken together, the results described above indicate that EBP50 regulates the TP receptor signaling through G␣ q at different levels. First, we have shown that EBP50 interferes with the coupling of the TP receptors to G␣ q by binding and preventing the GDP-bound G␣ q (inactive form) from binding to the TP receptor. Second, we have shown that EBP50 also interferes with the coupling of the activated G␣ q to the downstream effector PLC-␤1.
GAPs regulate the heterotrimeric G protein signaling by increasing the rate of GTP hydrolysis by G␣ subunits up to 2000-fold. Both RGS proteins (regulators of G protein signaling) and G protein effectors (such as PLC-␤ and p115RhoGEF, which lack an RGS domain) belong to the family of GAP proteins. RGS proteins specifically bind the GTP-bound G␣ subunits and share an RGS domain involved in the regulation of the intrinsic GTPase activity of G␣. Because our results have shown that EBP50 binds preferentially the GTP-bound G␣ q , and that this interaction is transient after activation of a GPCR, there was the possibility that EBP50 could modulate the intrinsic GTP hydrolysis activity of G␣ q . Analysis of the EBP50 amino acid sequence and its alignment with the RGS proteins failed to show any significant homology, indicating that EBP50 does not contain an RGS domain. Our results showed that EBP50 does not modulate the intrinsic GTPase activity of G␣ q . Regulation of G␣ q signaling by EBP50 must then rely principally on the mechanisms discussed above. We cannot, however, totally rule out the GAP activity of EBP50. Indeed, it was shown that the GRK2 GAP activity is apparent only in the presence of an activated GPCR (26), which suggests the possibility that receptors could have a critical role in potentiating GAP activity in cells. Moreover, it has been demonstrated that the ability of RGS2 to inhibit G␣ q -mediated signals in cells is highly dependent on the nature of the receptors that are being stimulated (32). The authors of this study suggested that regulatory selectivity may be conferred by specific receptor-RGS complexes. Thus, further investigation of the role of receptors in modulating EBP50-G␣ q interactions seem warranted.
EBP50 binds to NHE3 and is involved in its regulation by cAMP, as well as in its down-regulation (33). EBP50 may also be required to localize the H ϩ -ATPase in both apical and basolateral membranes in renal intercalated cells (33). Furthermore, EBP50 serves as a membrane retention signal for CFTR (34) and facilitates the dimerization of CFTR that promotes the full expression of chloride channel activity (35). Moreover, recent experiments (6) have shown that EBP50 is involved in the dimerization of the platelet-derived growth factor receptor. Among its numerous functions, EBP50 also regulates the sorting of the ␤ 2 -adrenergic receptor to either recycling endosomes or lysosomal degradatory pathways (36). Since its isolation, EBP50 has been associated with a broad array of biological systems, all of which depend on the interaction of EBP50 through its PDZ1 and PDZ2 domains with the different target proteins. Our study identifies G␣ q as a novel interacting partner for the PDZ domains of EBP50. This finding is important in terms of G protein signaling regulation. The study of possible interactions between other PDZ domain-containing proteins with G␣ subunits constitutes a new and exciting field of research. It is also easy to envision the great impact of the G␣ q -EBP50 interaction in signaling "cross-talk" because an EBP50 molecule bound to G␣ q could possibly be unable to engage in the interaction with another binding partner. Indeed, one can imagine that multiple biological systems associated with EBP50 could be modulated by the activated G␣ q -EBP50 interaction.