The guanine nucleotide exchange factor p63RhoGEF, a specific link between Gq/11-coupled receptor signaling and RhoA.

The monomeric GTPase RhoA, which is a key regulator of numerous cellular processes, is activated by a variety of G protein-coupled receptors, through either G12 or G(q) family proteins. Here we report that p63RhoGEF, a recently identified RhoA-specific guanine nucleotide exchange factor, enhances the Rho-dependent gene transcription induced by agonist-stimulated G(q/11)-coupled receptors (M3-cholinoceptor, histamine H1 receptor) or GTPase-deficient mutants of G alpha(q) and G alpha11. We further demonstrate that active G alpha(q) or G alpha11, but not G alpha12 or G alpha13, strongly enhances p63RhoGEF-induced RhoA activation by direct protein-protein interaction with p63RhoGEF at its C-terminal half. Moreover, the activation of p63RhoGEF by G alpha(q/11) occurs independently of and in competition to the activation of the canonical G alpha(q/11) effector phospholipase C beta. Therefore, our results elucidate a new signaling pathway by which G alpha(q/11)-coupled receptors specifically induce Rho signaling through a direct interaction of activated G alpha(q/11) subunits with p63RhoGEF.

The Rho GTPase family belongs to the Ras superfamily and comprises more than 20 distinct proteins. The best characterized members (RhoA, Rac1, and Cdc42) were first identified in the early 1990s as regulators of actin cytoskeleton rearrangements. RhoA, Rac1, and Cdc42 induce stress fiber, lamellipodia, and filopodia formation, respectively (1). Meanwhile, it became evident that Rho family proteins play a pivotal role in a variety of cellular processes, including secretion, smooth muscle contraction, migration, neurite retraction, and gene transcription (2,3). As monomeric GTPases, Rho proteins cycle between an inactive GDP-bound and an active GTP-bound state. The activation step, i.e. the exchange of GDP by GTP, is catalyzed by a group of accessory proteins: the guanine nucleotide exchange factors (GEFs). 1 About 60 different GEFs for Rho family members (RhoGEFs) are described so far. Most of them belong to the Dbl protein family, which share the typical tandem motif consisting of a Dbl homology (DH) and a pleck-strin homology (PH) domain (4,5). Besides this tandem motif, RhoGEFs often contain one or more additional signal transduction domains, such as SH2, SH3, PDZ, and additional PH domains. Therefore, they often function as molecular bridges between different signal transduction pathways (4,5).
It is well established that apart from receptor tyrosine kinases, a large variety of G protein-coupled receptors (GPCRs), particularly those coupling to the G 12/13 type of heterotrimeric G proteins, are upstream regulators of Rho proteins (6,7). A family of RhoA-specific GEFs, consisting of p115RhoGEF, PDZ-RhoGEF, and leukemia-associated RhoGEF (LARG), which mediates this activation process, has been identified (8 -10). All these proteins contain, in addition to the DH/PH tandem motif, a regulator of G protein signaling (RGS) homology domain for direct interaction with and activation by G 12 type G proteins. Recently, clear evidence has been provided that G␣ q and G␣ 11 as well as G q/11 -coupled receptors can induce potent RhoA activation (11)(12)(13). This process is apparently independent of phospholipase C ␤ (PLC␤) isozymes, the canonical G␣ q/11 effectors, and their downstream signaling, i.e. Ca 2ϩ mobilization and protein kinase C (PKC) activation (11,13). Although the G 12/13 -activated LARG has been reported to additionally mediate G␣ q -induced RhoA activation (13,14), a GEF specifically and selectively linking G␣ q/11 proteins and G q/11 -coupled receptors to RhoA activation has not been identified so far.
We and others (15,16) have recently identified a new Rho-GEF of 580 amino acids in length and an apparent molecular mass of 63 kDa. It was therefore termed p63RhoGEF. The expression of p63RhoGEF in human heart and brain tissue was confirmed by both groups of investigators using Northern blot analysis and immunohistochemistry, and it was clearly demonstrated that p63RhoGEF activates specifically RhoA but not Rac1 or Cdc42. Sequence analysis revealed that this protein does not contain other distinct functional domains besides the typical DH/PH tandem motif.
As we tried to identify regulatory mechanisms inducing a p63RhoGEF-mediated RhoA activation, we studied the influence of the stimulation of a variety of GPCRs on p63RhoGEF activity. The data presented herein will provide evidence that p63RhoGEF links specifically G q/11 -coupled receptors to RhoA by a direct interaction with GTP-liganded G␣ q/11 proteins. This RhoGEF therefore represents a so far unknown G␣ q effector molecule.
Cell Culture and Transfection-Culture of HEK-293 cells and COS-7 cells and transfection of the cells (250 ng of total DNA/well on a 48-well plate for SRF activation and up to 2 g of DNA/6-well plate for RhoA pull-down assays) were performed as described before (15). Assays were performed 48 h after transfection in serum-starved cells.
Preparation of RNA and Reverse Transcription-PCR-Total RNA from HEK-293 and COS-7 cells was prepared with RNeasy Minikit (Qiagen, Hilden, Germany). The RNA was transcribed into cDNA using oligo(dT) primers and a first strand synthesis kit (Roche Applied Science). PCR conditions for the amplification of p63RhoGEF cDNA fragments were as follows (50 l): primer 0.4 M each, dNTP 0.2 mM, 1ϫ PCR buffer, Taq polymerase plus Q-Solution (Qiagen), and 2 l of cDNA. 35 cycles (denaturation 95°C, 30 s, annealing 56°C, 30 s, elongation 72°C, 90 s) followed by a final acquisition of 5 min at 72°C were performed. Primer sequences were as follows: p63RhoGEF forward, 5Ј-GATGGTTGGATCATTCCAAACA-3Ј; p63RhoGEF reverse, 5Ј-GT-TACAGCTCATCTTCATCCA-3Ј. The sequences of these primers are conserved in rat, mouse, and man.
Assay of SRF Activation-HEK-293 cells or COS-7 cells seeded on 48-well plates were co-transfected with the indicated expression plasmids together with the pSRE.L-luciferase reporter plasmid and the pRL-TK control reporter vector. 48 h after transfection, cells were washed once with phosphate-buffered saline and lysed with passive lysis buffer (Promega). Luciferase activities were determined with the Dual-Luciferase reporter assay system (Promega) as described (15,17). The activity of the experimental reporter was normalized against the activity of the control vector.
Pull-down Assay of Activated RhoA-The cellular level of GTPloaded RhoA was determined using a GST fusion protein containing the Rho binding domain of rhotekin (GST⅐RBD) (15,18). In brief, subconfluent monolayers of HEK-293 cells were transfected with the indicated amounts of plasmid DNA or the corresponding empty vectors using Polyfect (Qiagen) and cultured for 48 h. Thereafter, the cells were lysed in a buffer containing 1% Nonidet P-40, and the particular fraction was pelleted by centrifugation. 1 mg of the GTPase-containing supernatant was then incubated for 1 h at 4°C with 40 g of GST⅐RBD (expressed in and purified from Escherichia coli) bound to glutathione-Sepharose beads. After three times washing of the beads, bound proteins were eluted with sample buffer and separated by SDS-PAGE. RhoA was then detected by immunoblotting with a specific monoclonal antibody (Santa Cruz Biotechnology).
Coimmunoprecipitation of G␣-proteins with p63RhoGEF-HEK-293 cells were seeded in 6-well plates and transfected at a confluence of 80% with 2 g of the indicated cDNA constructs. 48 h after transfection, the cells were solubilized in 600 l of immunoprecipitation buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 M Pefabloc SC, 0.1% Triton X-100). After incubation on ice for 20 min and centrifugation (26,000 ϫ g for 10 min), 2 g of anti-c-Myc antibody (clone 9E10, Sigma) or polyclonal anti-EE-antiserum (1:150 dilution, Covance) were added to the clear supernatant (300 g of protein) and incubated for 1 h at 4°C. After the addition of a 40-l 1:1 (v/v) slurry of anti-mouse-IgG-agarose conjugate (Sigma) or protein-A-Sepharose (Amersham Biosciences), the mixture was gently shaken for 4 h at 4°C. Beads were washed three times with Tris-buffered saline, 0.1% Triton X-100, and 1 M phenylmethylsulfonyl fluoride, and bound proteins were eluted with 25 l of sample buffer for 5 min at 95°C. Precipitated proteins were loaded onto a 12% polyacrylamide gel. After SDS-PAGE and transfer to nitrocellulose membranes, immunoprecipitated proteins were detected by immunoblotting with anti-c-Myc (clone 9E10, Sigma) and anti-G␣ antibodies (Gramsch Laboratories, Schwabhausen, Germany).
Phospholipase C Assay-For measurement of PLC activity, transfected COS-7 cells seeded on 12-well plates were incubated with 1 of Ci/ml myo-[ 3 H]inositol (Amersham Biosciences) 48 h prior to the assay. Thereafter, the cells were washed once with Hanks' balanced salt solution followed by the addition of fresh Hanks' balanced salt solution supplemented with 10 mM LiCl, and if indicated, 1 mM carbachol. After a 30-min incubation at 37°C, the reactions were stopped, and the formed [ 3 H]inositol phosphates were determined as described (19).
Statistics-Statistical analysis was performed by analysis of variance followed by Tukey's multiple comparison test. A p value Ͻ0.05 was considered significant. Concentration-response curves were analyzed using iterative nonlinear regression analysis (GraphPAD Prism).

Synergistic Activation of Rho-mediated Gene Transcription by p63RhoGEF and Stimulation of G q/11 -coupled Receptors-To
study whether p63RhoGEF is activated by G q/11 -coupled receptors, we coexpressed p63RhoGEF together with the M 3 R or the H 1 R in HEK-293 cells, in which the mRNA encoding p63RhoGEF could be detected by reverse transcription-PCR (Fig. 1A, inset). Both GPCRs preferentially couple to G q/11 proteins but can additionally activate other G proteins (20 -22). To monitor RhoA activation in intact cells, we measured transcription of a serum-response element (SRE)-controlled reporter gene (SRE.L-luciferase). Overexpression of p63RhoGEF caused a slight increase in luciferase expression, whereas agonist (carbachol, histamine) activation of the M 3 R or H 1 R increased transcriptional activity about 35-fold (Fig. 1A). The combined expression of p63RhoGEF and the respective GPCR led to a strong enhancement of agonist-induced luciferase expression (about 90-fold). As shown in Fig. 1B, the agonistinduced transcriptional activity of the M 3 R and the H 1 R was p63RhoGEF, a G␣ q/11 -regulated RhoA-specific RhoGEF further enhanced by the additional expression of G␣ q but not of G␣ 12 . In addition, coexpression of G␣ q but not G␣ 12 unmasked the known constitutive activity of the H 1 R (23,24). A similar effect was detected upon coexpression of the H 1 R with p63RhoGEF (Fig. 1A). Coexpression of RGS2, which acts as GTPase-activating protein (GAP) and therefore as an inhibitor of G␣ q/11 , but not G␣ 12/13 proteins (25,26), nearly eliminated the luciferase production induced by the activated M 3 R alone and largely suppressed the synergistic stimulation detected upon coexpression of p63RhoGEF (Fig. 1C). In contrast, the expression of the RGS domain of the mouse ortholog of p115RhoGEF Lsc (Lsc-RGS), which is a G␣ 12/13 -specific GAP and thereby suppresses G␣ 12/13 -mediated responses in HEK-293 cells (22), only slightly reduced (about 10 -20%) the M 3 Rinduced luciferase production. The transcriptional activity induced by the agonist-stimulated M 3 R coexpressed with p63RhoGEF was not affected by Lsc-RGS expression. The expression of both RGS proteins was verified by immunoblotting (not shown).
Synergistic Activation of Rho-mediated Gene Transcription by p63RhoGEF and G␣ q/11 but Not G␣ 12/13 Proteins-As the data obtained so far argued for an activation of p63RhoGEF by G q/11 proteins, we studied the influence of various G␣ subunits on p63RhoGEF-induced luciferase expression. Overexpression of wild-type G␣ q or G␣ 11 only marginally (2-3-fold) increased luciferase production. Upon coexpression of p63RhoGEF, the transcriptional activity was increased by 20 -40-fold (Fig. 2, A  and B). As reported before (11,17), overexpression of wild-type G␣ 12 and G␣ 13 induced strong increases in luciferase production, by 25-and 75-fold, respectively (Fig. 2C). Coexpression of p63RhoGEF with either G␣ 12 or G␣ 13 did not lead to further increases in transcriptional activity. The ineffectiveness of p63RhoGEF to enhance gene transcription by G␣ 12 and G␣ 13 was not due to a ceiling effect in the experimental setting. Expression of the GTPase-deficient mutants of G␣ q (G␣ q RC) and G␣ 11 (G␣ 11 QL) induced transcriptional activity in a range similar to that observed with G␣ 12 or G␣ 13 (compare Fig. 2, A and C). Upon coexpression of p63RhoGEF, the luciferase production induced by G␣ q RC and G␣ 11 QL was enhanced in a synergistic manner reaching levels of more than 100-fold. In contrast, the strong transcriptional activity (more than 100fold) induced by the GTPase-deficient mutants of G␣ 12 (G␣ 12 QL) and G␣ 13 (G␣ 13 QL) was not enhanced but was even slightly reduced by coexpression of p63RhoGEF (Fig. 2C). As reported before (15), the p63RhoGEF-induced luciferase production, even when stimulated by G␣ q RC or G␣ 11 QL, was largely suppressed by the RhoA-C-inactivating C3 ADP-ribosyl transferase (27) (data not shown).
Activated G␣ q/11 Proteins Largely Stimulate p63RhoGEFinduced RhoA Activation-The measurements of Rho-dependent gene transcription indicated a stimulation of p63RhoGEF by activated G␣ q/11 but not by activated G␣ 12/13 proteins. To substantiate this hypothesis, the amount of GTP-liganded RhoA was analyzed using the RhoA binding domain of rhotekin (GST⅐RBD) for pull-down of RhoA-GTP (18) in lysates of transfected HEK-293 cells (15). Under the conditions used (submaximally effective amounts of plasmid DNA encoding p63RhoGEF and G␣ subunits), expression of either p63RhoGEF, G␣ q RC, or G␣ 11 QL alone only slightly increased the amount of RhoA-GTP (Fig. 3). Combined expression of p63RhoGEF and G␣ q RC or G␣ 11 QL induced a substantial increase in RhoA-GTP. In contrast, the combined overexpression of p63RhoGEF with either G␣ 12 QL or G␣ 13 QL did not induce a synergistic activation of RhoA. In accordance with the data observed in the SRF activation assay, coexpression of p63RhoGEF slightly inhibited the RhoA activation by the permanent active G␣ 12/13 mutants.
Activated G␣ q/11 Proteins Interact Directly with the C-terminal Half of p63RhoGEF-To investigate whether p63RhoGEF interacts with activated G␣ q/11 proteins directly, we expressed various G␣-proteins and p63RhoGEF together in HEK-293 FIG. 2. p63RhoGEF enhances G␣ q/11 -but not G␣ 12/13 -induced SRF activation. A and C, luciferase production was measured in HEK-293 cells transfected with control vectors alone (Con), p63RhoGEF (p63, 100 ng), and the indicated G␣ proteins or GTPasedeficient mutants (50 ng of DNA each). B, luciferase production was measured in HEK-293 cells transfected with control vectors, p63RhoGEF (100 ng), and the indicated amounts of G␣ q plasmid. Mean Ϯ S.D.; n ϭ 6 -20. Inset in B, immunoblot detection of G␣ q/11 . FIG. 3. p63RhoGEF enhances G␣ q/11 -but not G␣ 12/13 -induced RhoA activation. The levels of RhoA-GTP, total RhoA, G␣ q/11 , G␣ 12 , G␣ 13 , and p63RhoGEF (p63) were determined in lysates of HEK-293 cells transfected with control vectors, p63RhoGEF, G␣ q RC, G␣ 11 QL, G␣ 12 QL, and G␣ 13 QL (2 g of DNA each) at 48 h after transfection as indicated. A representative experiment of several similar experiments is shown. p63RhoGEF, a G␣ q/11 -regulated RhoA-specific RhoGEF cells. Upon precipitation of the N-terminally Myc-tagged p63RhoGEF with anti-c-Myc antibodies, the immunoprecipitates were analyzed for coprecipitated G␣ proteins. As shown in Fig. 4B, the GTPase-deficient mutants G␣ q RC or G␣ 11 QL were coprecipitated by the anti-c-Myc antibody from cells transfected with the respective eukaryotic expression vectors. In contrast, G␣ q was not detected in the p63RhoGEF immunoprecipitates from cells overexpressing wild-type G␣ q . These data indicate that only activated G␣ q and G␣ 11 apparently exhibit high affinity binding to p63RhoGEF and thus can be precipitated in a complex with p63RhoGEF. To verify this hypothesis, we overexpressed an EE-tagged version of G␣ q QL (EE-G␣ q QL) together with p63RhoGEF and precipitated EEtagged proteins. As shown in Fig. 4C, p63RhoGEF was coprecipitated together with EE-G␣ q QL from lysates of cells coexpressing EE-G␣ q QL and p63RhoGEF.
To identify the part of p63RhoGEF in which this interaction takes place, we performed similar experiments with two truncated mutants of p63RhoGEF, i.e. p63-DH (amino acids 138 -379), which mainly consists of the DH domain, responsible for the guanine nucleotide exchange activity at RhoA (15,16) and p63-⌬N, consisting of the C-terminal half (amino acids 295-580) of p63RhoGEF with the PH domain but lacking the DH domain (Fig. 4A). The mutant p63-DH was precipitated to a similar extent as full-length p63RhoGEF (p63-FL) by the antic-Myc antibody but did not form a detectable complex with any of the G␣ q/11 proteins. In contrast, similar to p63-FL, p63-⌬N coprecipitated G␣ q RC and G␣ 11 QL but not wild-type G␣ q (Fig. 4, B and D). No coprecipitation with any p63RhoGEF construct was observed with either wild-type G␣ 12 and G␣ 13 (not shown) or their GTPase-deficient mutants, G␣ 12 QL and G␣ 13 QL (Fig. 4D).
The Recombinant G␣ q/11 Binding Domain of p63RhoGEF Inhibits G␣ q/11 and M 3 R-induced Rho-mediated Gene Transcription-As the N-terminally truncated mutant p63-⌬N specifically bound activated G␣ q/11 proteins, we used the expression of this mutant to study the role of endogenous RhoGEFs in RhoA activation by G␣ proteins and the M 3 R. In contrast to p63-FL, coexpression of p63-⌬N with G␣ q RC largely reduced the luciferase production induced by the GTPase-deficient G␣ q mutant (Fig. 5A). On the other hand, p63-⌬N and p63-FL both weakly inhibited the transcriptional activity induced by G␣ 13 . These data exclude a nonspecific inhibition of RhoA activation by p63-⌬N. Most important, expression of p63-⌬N, which by itself did not induce any transcriptional activity, potently suppressed (by up to 90%) the luciferase production induced by the carbachol-activated M 3 R (Fig. 5B).
p63RhoGEF Competes with PLC␤ for Activated G␣ q/11 Proteins-The direct interaction of p63RhoGEF with G␣ q/11 proteins finally prompted us to study whether p63RhoGEF and PLC␤ isozymes, which also directly interact with G␣ q/11 proteins (28), may compete with each other for activation. For this, the effect of p63RhoGEF on G␣ q -and M 3 R-induced luciferase production and PLC activation was examined in COS-7 cells, which are better suited to measure agonist-stimulated PLC activity in response to transient M 3 R expression than HEK-293 FIG. 4. Direct interaction of activated G␣ q/11 -proteins with p63RhoGEF. A, a schematic view of fulllength p63RhoGEF (p63-FL) and its truncated mutants, p63-DH and p63-⌬N. B-D, HEK-293 cells were transfected with control vectors, p63-FL, p63-DH, or p63-⌬N and wild-type G␣ q , G␣ q RC, EE-G␣ q QL, G␣ 11 QL, G␣ 12 QL, or G␣ 13 QL (2 g of DNA each) as indicated. Cells were lysed 48 h after transfection, and p63RhoGEF and its variants were immunoprecipitated (IP) with the anti-c-Myc-antibody (B and D). EE-G␣ q QL was precipitated with the anti-EE-antiserum (C). Precipitated proteins were separated by SDS-PAGE and immunoblotted with the anti-c-Myc antibody (upper panels) and G␣ subtypespecific antibodies (lower panels). The content of the G␣ subunits in the cell lysates (Lysate) is shown as a loading control. WB, Western blot. p63RhoGEF, a G␣ q/11 -regulated RhoA-specific RhoGEF cells. Similar to HEK-293 cells, COS-7 cells express p63RhoGEF endogenously (Fig. 1A, inset). Also, in these cells, coexpression of p63RhoGEF largely enhanced the G␣ q RC-induced transcriptional activity, by about 10-fold (Fig. 6A). On the other hand, p63RhoGEF, which had no effect on basal PLC activity, significantly reduced the PLC stimulation induced by G␣ q RC by about 40% (Fig. 6B). An even stronger inhibition, of about 70%, by p63RhoGEF was observed on the M 3 R-induced PLC stimulation (Fig. 6D). Vice versa, the effect of an overexpression of PLC␤2 on M 3 R-induced PLC activity and luciferase production was studied in COS-7 cells. Overexpression of PLC␤2 significantly increased the carbachol-induced inositol phosphate production from 4-to 5.5-fold (data not shown). As shown in Fig. 6C, the carbachol-induced luciferase production was reduced by about 50% in cells coexpressing PLC␤2 and p63RhoGEF. To further exclude a contribution of the PLC/PKC pathway on the regulation of p63RhoGEF, we additionally studied the influence of the PLC inhibitor U-73122 and the broad range PKC inhibitor bisindolylmaleimide IX on G␣ q QLand p63RhoGEF-induced transcriptional activity in COS-7 cells overexpressing PLC␤2. Neither U-73122 (2.5 M) nor bisindolylmaleimide IX (0.5 M) altered the transcriptional activity induced by p63RhoGEF, G␣ q QL, or their combination (data not shown).

DISCUSSION
It is meanwhile well documented that activation of heterotrimeric G proteins of the G q/11 subfamily induce Rho activation in a variety of cells and tissues (11)(12)(13)29). There is increasing evidence that this activation of Rho occurs independently of the PLC/PKC pathway. Most likely, so far unidentified GEFs are involved (11,13). In this report, we presented several lines of evidence that p63RhoGEF, a novel member of the Dbl family of GEFs, might represent the specific GEF, or at least one of the GEFs, mediating this response. Firstly, p63RhoGEF largely enhanced the transcriptional activity of the primarily G q/11 -coupled M 3 R and H 1 R in a manner sensitive to RGS2, a negative regulator of G q/11 activity (25,26), but insensitive to the G 12/13 -specific Lsc-RGS (22). Secondly, it largely enhanced the Rho-mediated gene transcription induced by the GTPase-deficient mutants of G␣ q and G␣ 11 but not of G␣ 12 and G␣ 13 . This increase in transcriptional activity was completely sensitive to the RhoA-inactivating C3 ADP-ribosyl transferase of Clostridium botulinum (27). Thirdly, coexpression of p63RhoGEF and activated mutants of G␣ q and G␣ 11 largely and synergistically increased the cellular amount of activated RhoA proteins. Finally, p63RhoGEF directly interacted with activated G␣ q and G␣ 11 proteins. This interaction could be detected by means of coimmunoprecipitations as well as by functional inhibition of G␣ q -induced gene transcription by the G␣ q/11 binding domain of p63RhoGEF. Therefore, our data indicate that p63RhoGEF is the first known GEF that is specifically and so far exclusively regulated by activated G␣ subunits of the G q family. Furthermore, the GEF-deficient mutant p63-⌬N efficiently suppressed Rho-mediated gene transcription induced by the activated M 3 R and G␣ q RC, but not G␣ 13 , indicating that p63RhoGEF is in fact specifically involved in RhoA activation by G q type G proteins and G q/11coupled GPCRs, e.g. the M 3 R and H 1 R. Thus, p63-⌬N apparently binds to activated G␣ q/11 proteins and thereby prevents binding and activation of endogenous p63RhoGEF (or another functionally related RhoGEF) required for RhoA activation by these G␣ proteins and the G q/11 -coupled M 3 R. This hypothesis is on the one hand corroborated by the finding that agonistinduced gene transcription by these receptors was enhanced by coexpression of G␣ q but not by G␣ 12 . On the other hand, the stimulatory effect of the M 3 R was strongly, but not fully, blunted by either p63-⌬N or RGS2, which inactivates and traps activated G␣ q/11 proteins (25,30). The rather small remaining part of the M 3 R action might therefore represent coupling to p63RhoGEF, a G␣ q/11 -regulated RhoA-specific RhoGEF G 12 type G proteins, which is in line with the known G protein coupling specificity of this GPCR (21,22). This interpretation is further corroborated by the small inhibitory effect (about 15%, Fig. 1C) of the G␣ 12/13 -specific GAP Lsc-RGS (22).
Screening data bases of the human genome project and the mouse genome project revealed that p63RhoGEF, in contrast to other known RhoA-specific GEFs, has no close homolog. Besides the existence of a DH/PH tandem motif, p63RhoGEF is especially not related to the members of the p115RhoGEF family, which mediate the activation of RhoA by G 12/13 proteins (8 -10,31). Accordingly, we could not detect any interaction of p63RhoGEF with or activation by G␣ 12 family members. Moreover, p63RhoGEF does not contain the RGS homology domain, which is involved in the interaction of the p115 RhoGEF family members with activated G␣ subunits. The activation of p63RhoGEF is apparently induced by a direct interaction of activated G␣ q/11 proteins with the C-terminal half of the p63RhoGEF molecule, containing the PH domain. This interaction likely results in the relief of the described autoinhibition of the GEF activity (15,32). Therefore, our data indicate that the C-terminal G␣ q/11 binding site of p63RhoGEF has to be quite distinct from the interaction site by which p115 family members interact with activated G␣ subunits. With regard to this interaction site, it is noteworthy, however, that the p115RhoGEF family member, LARG, has been shown to interact with and be activated by G␣ q (13,14). Whether this interaction is mediated by the RGS homology domain of LARG is still a matter of debate. One report indicated a binding of activated G␣ q to the RGS domain of LARG and a productive coupling of activated G␣ q to RhoA via LARG (14). Another report showed that a LARG mutant lacking the DH domain but not the RGS homology domain interfered with G␣ q -induced RhoA activation (13). However, a third publication did not detect an interaction of activated G␣ q with the RGS homology domain of LARG, which also failed to inhibit G␣ q QL-induced luciferase production (11). In contrast to the data on the interaction of G␣ q with the LARG-RGS homology domain, the interaction of G␣ 12 family members with this domain and a productive coupling to RhoA were found by both groups (11,14).
Another line of evidence for p63RhoGEF to be a so far unknown effector molecule for G q proteins resulted from the experiments analyzing the influence of the canonical G␣ q effector molecule PLC␤ (28) on the activation of p63RhoGEF and vice versa. The overexpression of p63RhoGEF inhibited G␣ q -stimulated PLC activity as well as the overexpression of PLC␤2 inhibited the G␣ q -induced gene transcription. These data argue for a direct competition of p63RhoGEF and PLC␤ isoforms for activated G␣ q/11 proteins. In line with the inhibitory effect of PLC␤2 and in accordance with previous observations, which indicated the existence of a G q/11 -activated RhoGEF not regulated by PLC and the subsequent Ca 2ϩ mobilization and PKC activation (11), the M 3 R-and G␣ q QL-induced gene transcription was not influenced by the inhibition of PLC and PKC activity, by U73122 (33) and bisindolylmaleimide IX (34), respectively.
In summary, the data presented herein define a new signal-ing pathway for GPCRs, with p63RhoGEF serving as a direct G␣ q/11 effector molecule. It directly links these GPCRs to RhoAand RhoA-dependent cellular processes, apparently in competition with the canonical PLC␤/PKC pathway.