- Email: [email protected]
Regulator of G-protein signalling 3 redirects prototypical Gi-coupled receptors from Rac1 to RhoA activation
Cellular Signalling 19 (2007) 1229 – 1237 www.elsevier.com/locate/cellsig
Regulator of G-protein signalling 3 redirects prototypical Gi-coupled receptors from Rac1 to RhoA activation Andreas Vogt a , Susanne Lutz a , Ulrich Rümenapp b , Li Han b , Karl H. Jakobs b , Martina Schmidt b,1 , Thomas Wieland a,⁎ a
Department of Experimental and Clinical Pharmacology and Toxicology, University of Heidelberg, Mannheim, Germany b Department of Pharmacology, University of Duisburg-Essen, Essen, Germany Received 21 December 2006; received in revised form 8 January 2007; accepted 8 January 2007 Available online 17 January 2007
Abstract The small GTPases, Rac1 and RhoA, are pivotal regulators of several essential, but distinct cellular processes. Numerous G-protein-coupled receptors signal to these GTPases, but with different specificities. Specifically, Gi-coupled receptors (GiPCRs) are generally believed to activate Rac1, but not RhoA, a process involving Gβγ-dimers and phosphatidylinositol 3-kinase (PI3K). Here we show that, depending on the expression level of the 519 amino acid isoform of regulator of G-protein signalling 3 (RGS3L), prototypical GiPCRs, like M2 muscarinic, A1 adenosine, and α2-adrenergic receptors, activate either Rac1 or RhoA in human embryonic kidney cells and neonatal rat cardiomyocyte-derived H10 cells. The switch from Rac1 to RhoA activation in H10 cells was controlled by fibroblast growth factor-2 (FGF-2), lowering the expression of RGS3L. Activation of both, Rac1 and RhoA, seen at low and high expression levels of RGS3L, respectively, was sensitive to pertussis toxin and the PI3K inhibitor LY294002 and mediated by Gβγ-dimers. We conclude that RGS3L functions as a molecular switch, redirecting GiPCRs via Gβγ-dimers and PI3K from Rac1 to RhoA activation. Considering the essential roles of Rac1 and RhoA in many signalling pathways, this additional function of RGS3L indicates a specific role of this protein in cellular signalling networks. © 2007 Elsevier Inc. All rights reserved. Keywords: Gi proteins; Gβγ-dimers; RhoA; Rac1; FGF-2; RGS proteins; RGS3
1. Introduction Members of the Rho GTPases family, like the best characterised family members RhoA, Rac1, and Cdc42, which has been first identified as regulators of actin cytoskeleton alterations, are now considered to be pivotal regulators of many signalling networks, participating in secretion, smooth muscle contraction, cell migration, membrane trafficking, gene transcription and cell cycle progression (for recent review see [1– 3]). G-protein-coupled receptors (GPCRs) in numerous cell types have been shown to activate Rho GTPases. It is generally ⁎ Corresponding author. Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Medizinische Fakultät Mannheim, Universität Heidelberg, Maybachstrasse 14, D-68169 Mannheim, Germany. Tel.: +49 621 330030; fax: +49 621 3300333. E-mail address: [email protected] (T. Wieland). 1 Present address: Department of Molecular Pharmacology, University of Groningen, The Netherlands. 0898-6568/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.cellsig.2007.01.003
assumed that the activation of RhoA by GPCRs is mediated by Gα-subunits of the G12 and/or Gq subfamilies. By contrast, Gicoupled receptors (GiPCRs) are believed to activate Rac1, but not RhoA, via Gβγ-dimers and involvement of phosphoinositide 3-kinase (PI3K) [3,4]. Therefore, the distinct cellular responses induced by pertussis toxin (PTX)-sensitive GiPCR signalling cascades and PTX-insensitive Gq/12 PCRs regulated pathways are often attributed to Rac1 and RhoA activation, respectively, and their different spectrum of downstream effectors [3]. RGS3, a member of the regulator of G-protein signalling (RGS) family, shares a high degree of homology within its RGS domain with the R4 subfamily and thus acts as GAP for the Gα-subunits of Gi and Gq family members [5–8]. Increases in cellular levels of RGS proteins apparently contribute to alterations in GPCR signalling [5,6] as for example seen in heart failure [9,10]. In contrast to other members of the R4 subfamily, RGS3 exists in several splice variants of which at least two exhibit an extended amino-terminal domain in
1230
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
Fig. 1. The GAP deficient RGS3L mutant N460A acts as a Gβγ-scavenger (A) Single turnover GTPase activity of recombinant Gαi1 measured in the absence (Basal) and presence of recombinant RGS3L or RGS3L N460A (N460A). (B) Luciferase production was measured in HEK-TsA201 cells transfected with the permanent active but GAP-sensitive Gαq mutant (GαqRC) (1 μg of DNA) alone and together with wild-type RGS3L, RGS3L N460A, RGS2 (2 μg of DNA each) or C3 transferase (C3T, 1 μg of DNA) as indicated. Mean ± SEM; n = 8–20. Gαq and RGS protein expression was monitored by immunoblotting (inset). (C) RGS3L N460A was immunoprecipitated from cell lysates with an anti-c-myc antibody and precipitated RGS3L N460A (upper panel) and Gβ (middle panel) were visualised by immunoblotting. Total content of Gβ in the cell lysates is shown as loading control (lower panel). (D) HEK-TsA201 cells were transfected with empty vector (Basal) and Gβ1 plus Gγ2 (10 μg of DNA each) in the absence or presence of RGS3L N460A (25 μg of DNA) or βARK-CT (100 μg of DNA). (A) After 48 h, basal [3H] inositol phosphate formation was measured. Data are means ± SEM (n = 4; ⁎⁎⁎, p b 0.001 vs. Gβ1γ2). Means ± SD of assay triplicates are given.
addition to the RGS domain [11,12]. It has been shown that the 519 amino acid isoform of RGS3 (RGS3L) binds Gβγ-dimers [13]. It thereby acts as a Gβγ-scavenger and inhibits the Gβγinduced activation of phospholipase Cβ (PLCβ), Akt and mitogen-activated protein kinase independently of its GAP activity. Herein we report that the expression level of RGS3L controls whether GiPCRS activate Rac1 or RhoA in human embryonic kidney (HEK) and neonatal rat cardiac myocyte derived H10 cells via a Gβγ- and PI3K-mediated pathway.
2. Materials and methods 2.1. Plasmid construction The cDNA encoding the 519 aa isoform of human RGS3 (a kind gift by Dr. J. H. Kehrl, Bethesda, MD, USA) was subcloned in the pCMV-Tag3b vector (Stratagene). Site-directed mutagenesis of RGS3L was performed using Quikchange (Stratagene) with the mutagenic primers RGS3N460Afw: GCATGCAAGGAGGTAGCGCTGGACTCCTACACG, RGS3N460Abw: CGTGTAGGAGTCCAGCGCTACCTCCTTGCATGC according to the manufacturer's protocol. The sequence of the mutant was verified by automated sequencing. The eukaryotic expression vectors encoding C3 transferase, M2 muscarinic acetylcholine receptor (M2 mAChR), α2A-adrenoceptor (α2AAR) and β-ARKCT were kind gifts of Dr. A. Hall, London, UK, Dr. C. van Koppen, Essen,
Germany, Dr. L. Hein, Freiburg, Germany and Dr. M. Lohse, Würzburg, Germany, respectively.
2.2. Cell culture and transfection Culture of human embryonic kidney (HEK-TsA201) cells and transfection of the cells (3 μg of total DNA per well on a 12-well plate for SRF activation, and up to 250 μg of DNA per 145-mm culture dish for GTPase pull-down assays) were performed as described before [14]. Assays were performed 24 h or 48 h after transfection in serum-starved cells. H10 cells were cultured as reported before [15], and then treated for 24 h in serum-free medium without and with fibroblast growth factor-2 (FGF-2) (50 ng/ml) or PTX (100 ng/ml) as indicated in the figure legends. For transfection of H10 cells with siRNAs, the cells were cultured on 35-mm dishes or 6-well plates and transfected with 5 μl Lipofectamine 2000 in 500 μl OptiMEM and 50 pmol of each siRNA according to the manufacturer's instructions (Invitrogen). If vector DNA was co-transfected with siRNA, 800 ng of the individual vector DNA was used in a total amount of 2.5 μg. The specific siRNAs (Ambion) were si-RGS3 (1): 5′GGAUAUGAAGAAUAAGCUGtt-3′, si-RGS3 (2): 5′-GGAAGGAAUCCUUUUCAGGtt-3′. To evaluate transfection efficiency, fluorescence labelled siRNA (Alexa Fluor 488, Qiagen) was used. Scrambled siRNA served as control. The expression of surface M2 mAChR was determined by [3H] Nmethylscopolamine ([3H] NMS) binding as described before [16].
2.3. Assay of SRF activation HEK-TsA201 cells seeded on 12-well plates were transfected with different expression plasmids together with the pSRE.L-luciferase reporter plasmid
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
1231
activity of the experimental reporter was normalised against the activity of the control vector.
2.4. Pull-down assays of activated Rho GTPases The cellular levels of GTP-loaded Rac1, RhoA and Cdc42 were determined as described [17,18]. In brief, the Rac/Cdc42-binding domain of Pak1 (GSTPBD) and the Rho-binding domain of rhotekin (GST-RBD) were expressed in and purified from E. coli. Subconfluent monolayers of HEK-TsA201 cells were transfected with the indicated amounts of plasmid DNA or the corresponding empty vectors and cultured for 48 h. Thereafter, the cells were rinsed with Hank's balanced salt solution and stimulated with the indicated agonists for the indicated periods of time at 37 °C. The cells were lysed in a buffer containing 1% (by vol.) Nonidet P-40, and the particular fraction was pelleted by centrifugation. The GTPase-containing supernatant was then incubated for 2 h at 4 °C with GST-PBD or GST-RBD bound to glutathione–sepharose beads. After twice washing of the beads, bound proteins were eluted with sample buffer and separated by SDS-PAGE. Rac1, RhoA and Cdc42 were then detected by immunoblotting with specific antibodies. Pull-down of GTP-loaded Rac1 and RhoA from lysates of H10 cells, pretreated with and without FGF-2 or PTX or transfected with siRNAs and then stimulated for 5 min at 37 °C with the indicated receptor agonists, was performed as in HEK-TsA201 cells.
2.5. 2D-gel electrophoresis Fig. 2. RGS3L N460A inhibits Gβ1γ2-induced Rac1 but enhances Gβ1γ2induced RhoA activation in HEK-TsA201 cells. The levels of Rac1-GTP, total Rac1, RhoA-GTP and total RhoA were determined in HEK-TsA201 cells transfected with Gβ1 plus Gγ2 (10 μg of DNA each), RGS3L N460A (N460 A, 25 μg of DNA), βARK-CT (100 μg of DNA) or the indicated combinations. Recombinant RGS3L N460A and βARK-CT expression was monitored by immunoblot analysis. (0.5 μg of DNA) and the pRL-TK control reporter vector (0.1 μg of DNA) as described [17]. At 24 h after transfection, luciferase activities were determined in cell extracts with the Dual Luciferase Reporter Assay System (Promega). The
Cell lysates (about 50 μg protein) were mixed with 70 μl rehydration buffer and applied to 7 cm IPG gel strips (GE Healthcare) containing a linear 3–10 pH gradient. Isoelectric focusing was carried out using an Ettan IPGphor unit (GE Healthcare). The subsequent SDS-PAGE was performed on 8% polyacrylamide gels.
2.6. Western blotting, immunoprecipitation Western blotting was performed according to standard procedures [17]. Specific primary antibodies against RGS3 (H300), Gβ (T20) Gαq (C19), βARK-CT (GRK2, C15), and c-myc (Santa Cruz Biotechnology) were used according to the manufacturer's recommendations. The RGS2 antibody was a
Fig. 3. RGS3L N460A redirects GiPCRs in HEK-TsA201 cells from PI3K-dependent Rac1 to RhoA activation. The levels of Rac1-GTP, total Rac1 (A) RhoA-GTP and total RhoA (B) were determined in HEK-TsA201 cells transfected with the M2 mAChR or α2AAR (12.5 μg of DNA) without and with βARK-CT or RGS3L N460A (N460A) and stimulated for 5 min without (−) and with (+) 1 mM carbachol or 10 μM adrenaline (Adr). Where indicated, cells were treated for 16 h with 100 ng/ml PTX or 15 min with 50 μM LY294002 before agonist stimulation. Recombinant RGS3L N460A expression was monitored by immunoblot analysis.
1232
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
2.10. Data presentation and statistical analyses All effector pull-down assays shown are representative experiments which have been repeated at least three times. Statistical analysis was performed by one-way ANOVA with Bonferroni's posttest. p b 0.05 was considered statistically significant.
3. Results 3.1. RGS3L N460A inhibits Gβγ-PI3K-dependent activation of Rac1 in HEK cells RGS3L inhibits Gβ1γ2-induced activation of co-expressed PLCβ2 in COS-7 cells, independent of its GAP activity on Gα proteins [13]. To dissect the Gβγ-scavenger property of RGS3L from its GAP function, the catalytically important Asn 460 in the RGS domain was mutated to Ala [20]. In contrast to wildtype RGS3L, the RGS3L N460A mutant neither accelerated the GTPase activity of purified recombinant Gαi1 (Fig. 1A) nor inhibited serum response factor (SRF)/RhoA-mediated transcriptional activity [21] induced by expression of the constitutively active but GAP sensitive Gαq mutant GαqRC (Fig. 1B). As reported before [13], RGS3L forms a complex
Fig. 4. Influence of dominant-negative Rac1 and RhoA on M2 mAChR-induced Rac1 and RhoA activation. HEK-TsA201 cells were transfected with expression plasmids encoding M2 mAChR (12.5 μg DNA), RGS3L N460A (N460A, 25 μg DNA), Rac1N17 (A) and RhoAN19 (B) (100 μg of DNA) alone or in combinations. 48 h after transfection, the cells were stimulated for 5 min without (−) or with (+) 1 mM carbachol. Levels of Rac1-GTP and RhoA-GTP were visualised by effector pull-down assays.
kind gift of Dr. J. H. Kehrl, Bethesda, MD. Detection was carried out with suitable secondary antibodies and Lumi Light Plus (Roche Applied Science). Co-immunoprecipitation was performed as described [19].
2.7. F-actin staining Subconfluent H10 cells were cultured on Culture Slides (Becton Dickinson), transfected with siRNA or treated with or without FGF-2 (50 ng/ml). Seventytwo hours posttransfection or after 24 h FGF-2 treatment, cells were stimulated with and without 1 mM carbachol for 1 h. Staining with tetramethylrhodamine isothiocyanate (TRITC)-conjugated phalloidin was performed as described [17].
2.8. Single turnover GTPase assay The single turnover GTPase assay of recombinant Gαi1 was performed exactly as described before [20] in the presence and absence of 1 μg purified RGS3L or RGS3L N460A.
2.9. Measurement of PLC activity Formation of [3H]inositol phosphates in HEK-TsA201 cells labelled with myo-[3H]inositol was measured 48 h posttransfection for 30 min at 37 °C in the presence of 10 mM LiCl as described before [15].
Fig. 5. Effects of RGS3L isoforms and Gβ1γ2-dimers on SRF-mediated gene transcription. (A) HEK-TsA201 cells were transfected with the pSRE.Lluciferase reporter encoding firefly luciferase (0.5 μg of DNA) and the pRLTK control reporter encoding Renilla luciferase (0.1 μg of DNA) alone (Control) or together with RGS3L, RGS3L N460A (N460A), RGS3S (2 μg of DNA each) or C3 ADP-ribosyl transferase (C3T, 1 μg of DNA) as indicated. (B) Cells were transfected with RGS3L, RGS3L N460A (0.5 μg of DNA), Gβ1 plus Gγ2 (0.3 μg of DNA each), alone and in the indicated combinations without or with βARK-CT (2 μg of DNA). Firefly luciferase activities were normalised against the level of expressed Renilla luciferase. Data are means ± SEM (n = 4– 14, ⁎⁎⁎p b 0.001 vs. Basal, #p b 0.01 vs. Gβ1γ2). Expression of recombinant proteins was monitored by immunoblot analysis.
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
1233
suppressed the Gβ1γ2-induced Rac1 activation. Similar data were obtained on Rac1 activation induced by endogenous Gβγdimers derived from receptor-activated Gi-proteins. Agonist (1 mM carbachol, 10 μM adrenaline) activation of the transfected Gi-coupled M2 muscarinic acetylcholine receptor (M2 mAChR) or α2A-adrenoceptors (α2AAR) led to marked GTP loading of Rac1. The GiPCR-induced response was fully suppressed by treatment with either PTX or the PI3K inhibitor LY294002 (50 μM). Similarly, co-expression of either β-ARKCT or RGS3L N460A (Fig. 3A), completely abolished agonistinduced Rac1 activation. As judged by [3H] NMS binding, cotransfection with either RGS3L N460A or βARK-CT did not alter the M2 mAChR expression. Specific binding was 288 ± 45, 305 ± 33, and 318 ± 45 fmol/mg of protein for M2 mAChR, M2 mAChR plus RGS3L N460A and M2 mAChR plus βARK-CT expressing cells, respectively. 3.2. RGS3L N460A induces Gβγ-PI3K-dependent activation of RhoA in HEK cells It is well known that different Rho GTPases can influence each other's activities. In particular, activation of Rac1 can lead to activation of RhoA, as demonstrated in Swiss 3T3 fibroblasts [23,24], while in other cell types, e.g. NIH 3T3 and HeLa cells, activated Rac1 down-regulates RhoA activity [25,26]. We therefore studied whether RGS3L, which blunted Rac1 activation, may affect RhoA activities. Indeed, expression of Gβ1γ2 in HEK-TsA201 cells also increased the amount of
Fig. 6. Comparison of RGS3L and RGS3L N460A-dependent RhoA activation. HEK-TsA201 cells transfected with the M2 mAChR together with RGS3L (A) or RGS3L N460A (N460A, B) were treated without (−) and with (+) carbachol for the indicated periods of time. Representative immunoblots for RhoA-GTP and total RhoA are shown in the upper panels, while in the lower panels mean ± SEM (n = 4–6) are presented, with the amount of RhoA-GTP found in unstimulated control cells set to 1.
with Gβγ-dimers and thereby inhibits Gβγ-induced PLC activation. In accordance, RGS3L N460A co-immunoprecipitated co-expressed Gβ1γ2 in HEK-TsA201 cells (Fig. 1C). Expression of Gβ1γ2 strongly increased activation of endogenous PLC in HEK-TsA201 cells, by about 8-fold (Fig. 1D). Co-expression of RGS3L N460A suppressed this cellular response, similarly as co-expression of the well-known Gβγ-scavenger, βARK-CT, the Gβγ-binding domain of the β-adrenergic receptor kinase [22], by about 50%. We therefore studied whether RGS3L N460A also inhibits the Gβγ-induced activation of Rac1. As illustrated in Fig. 2, overexpression of Gβ1γ2 in HEK-TsA201 cells induced the activation of endogenous Rac1 as measured by precipitation of Rac1-GTP by the Rac/Cdc42-binding domain of Pak1 and visualisation of bound Rac1 by immunoblotting [18]. Coexpression of either βARK-CT or RGS3L N460A completely
Fig. 7. Depletion of RGS3L expression in H10 cells. (A) Serum-starved and H10 cells were treated without (control) and with FGF-2 for 24 h as indicated. (B) Serum-starved H10 cells were transfected with scrambled siRNA (control) or RGS3L-siRNA (siRGS3L). The expression of RGS3L in lysates of control and FGF-2 treated and RGS3L-siRNA-transfected cells was analysed by 2D gel electrophoresis and immunoblot. Gβ content is shown as loading control. (C) Transfection efficacy was monitored by light (LM) and fluorescence microscopy (FM) of Alexa-Fluor-siRNA-transfected H10 cells.
1234
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
RhoA-GTP, measured by precipitation with a GST-fusion protein of the Rho-binding domain of rhotekin [18] (Fig. 2). Most interestingly however, expression of RGS3L N460A but not of βARK-CT induced a similar RhoA activation than Gβ1γ2 overexpression. When RGS3L N460A was co-expressed with Gβ1γ2, the level of activated RhoA was further increased. Co-expression of the Gβγ-scavenger βARK-CT fully prevented RhoA activation induced by the combination of Gβ1γ2 and RGS3L N460A or the individual agents, Gβ1γ2 and RGS3L N460A (Fig. 2). Thus, the data suggest that RGS3L N460A induce RhoA activation in a Gβγ-dependent manner. It was therefore of major interest to test whether RGS3L may switch Gβγ-liberating GiPCR signalling to induce RhoA activation. As reported before by others [21], agonist stimulation of the M2 mAChR (Fig. 3B) did not induce RhoA activation, neither in control nor in PTX-treated HEK-TsA201 cells. However, in cells expressing RGS3L N460A the agonistactivated M2 mAChR caused a strong, further increase in RhoA activity. PTX treatment abolished the receptor response. Coexpression of βARK-CT fully abrogated the synergistic RhoA
Fig. 9. siRNA-induced RGS3L depletion in H10 cells induces a switch in GiPCR-induced RhoGTPase activation. Serum-starved H10 cells were transfected with scrambled siRNA (−) or RGS3L-siRNA (+). The level of Rac1-GTP and RhoA-GTP was determined after 5 min stimulation without or with 1 mM carbachol or 10 μM CPA. Total Rac1 and RhoA content of cell lysates (30 μg of protein) is shown as loading controls.
activation induced by RGS3L N460A and the M2 mAChR. Similar to M2 mAChR, stimulation of α2AAR induced RhoA activation in the presence of RGS3L N460A in a LY294002 sensitive manner. Activation of RhoA induced by RGS3L N460A and GiPCRs was not due to inhibition of Rac1. Expression of the dominantnegative Rac1 mutant, Rac1N17, suppressed the M2 mAChRmediated activation of endogenous Rac1, similarly as expression of RGS3L N460A (Fig. 4A). However, Rac1N17 had no effects on activation of RhoA induced by either RGS3L N460A alone or when combined with the agonist-activated M2 mAChR. Vice versa, the expression of a dominant-negative RhoA mutant, RhoAN19, slightly increased the basal Rac1 activity in HEK-TsA201 cells. Thus, the M2 mAChR-induced Rac1 activation in the absence of RGS3L N460A appeared to be diminished (Fig. 4B). Importantly, in the presence of RGS3L N460A, expression of RhoAN19 completely suppressed the basal as well as the carbachol-stimulated activation of RhoA. Additionally, we studied whether another member of the Rho GTPases, i.e. Cdc42, might be involved in the observed RGS3L N460A-induced shift from Rac1 to RhoA activation. However, neither overexpression of Gβ1γ2 nor expression of RGS3L N460A alone caused an increase in the amount of Cdc42-GTP in HEK-TsA201 cells. Furthermore, although the agonistactivated M2 mAChR induced an increase in the level of Cdc42GTP, this was not altered by co-expression of RGS3L N460A (data not shown). These data therefore indicate that RGS3L N460A overexpression is able to redirect the GiPCR/Gβγ/PI3K signalling pathway from Rac1 to RhoA activation. 3.3. Comparison of RGS3L N460A-and wild-type RGS3Lmediated RhoA activation in HEK cells
Fig. 8. Regulation of Rac1 and RhoA activation by FGF-2 and carbachol in H10 cells. Serum-starved H10 cells were treated with PTX, LY294002 and FGF-2 as indicated. Thereafter, cells were stimulated for 5 min without (−) or with (+) carbachol. Levels of Rac1-GTP (A) and RhoA-GTP (B) were analysed by effector pull-down assays. Total Rac1 and RhoA content of cell lysates (30 μg of protein) is shown as loading controls.
As the data so far presented demonstrate a GiPCR- and Gβγdependent RhoA activation only in the presence of the GAP deficient N460A mutant of RGS3L, we studied whether and to what extent wild-type RGS3L is also able to contribute to RhoA activation. For this, we first examined the effects of RGS3L and
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
RGS3L N460A and RGS3S, an N-terminally truncated variant of RGS3 [11], lacking half of the Gβγ binding domain [13], on SRF-mediated gene transcription. Expression of RGS3L, but not of RGS3S, significantly increased luciferase expression by about 2.5-fold (Fig. 5A). Expression of RGS3L N460A increased transcriptional activity even stronger up to 6.8-fold. The RGS3L- and RGS3L N460A-induced luciferase production was fully suppressed by co-expression of Clostridium botulinum C3 transferase, which specifically inactivates RhoA–C, but not Rac1 or Cdc42 [27]. As shown in Fig. 5B, the transcriptional activity induced by RGS3L or RGS3L N460A was completely suppressed by co-expression of βARK-CT and mimicked by expression of Gβ1γ2. Co-expression of RGS3L or RGS3L N460 with Gβ1γ2 caused a synergistic increase in transcriptional activity. For example, while RGS3L and Gβ1γ2 alone increased the luciferase production by about 2.5- and 4.3-fold, respectively, the transcriptional activity was enhanced up to 14-fold and 19-fold by co-expression of RGS3L plus Gβ1γ2 and RGS3L N460A plus Gβ1γ2, respectively. Activation of RhoA by the M2 mAChR was also observed with wild-type RGS3L, although with different kinetics as with RGS3L N460A (Fig. 6A). Maximal RhoA activity induced by the M2 mAChR in cells overexpressing wild-type RGS3L was observed at 2 min stimulation with carbachol, while at 10 min the receptor response was lost. In contrast, in the presence of RGS3L N460A the carbachol-induced RhoA activation was sustained, for up to 30 min. Thus, wild-type RGS3L is similarly able to mediate GiPCR-induced RhoA activation via its direct interaction with Gβγ-dimers [13]. The difference in signal extent and kinetics between RGS3L and RGS3L N460A apparently reflects the Giinactivating GAP activity of RGS3L. Therefore, it was not seen when free Gβ1γ2 dimers are overexpressed (see Fig. 5B).
1235
the endogenously expressed M2 mAChR and adenosine A1 receptor with carbachol and N6-cyclopentyladenosine (CPA, 10 μM, not shown in Fig. 8), respectively, strongly increased RhoA-GTP levels in H10 cells, without inducing any measurable Rac1 activation (Figs. 8 and 9). The activation of RhoA by the two receptor agonists was fully sensitive to PTX treatment of the cells or the inhibition of PI3K by LY294002 (Fig. 8A). The capabilities of both carbachol and CPA to induce RhoA activation were lost in the FGF2-treated H10 cells. Instead, both agonists induced a robust activation of Rac1, which was fully blunted by PTX treatment and PI3K inhibition by LY294002 (Fig. 8B). The influence of the cellular depletion of RGS3L by siRNA on carbachol- or CPA-induced Rac1 and RhoA activation is shown in Fig. 9A and B, respectively. Similar to the reduction of RGS3L expression by FGF-2, the siRNA-induced RGS3L depletion caused a switch from a carbachol- or CPA-induced RhoA to Rac1 activation. Like in HEK-TsA201 cells the expression of M2 mAChR was similar in siRNA transfected cells. Specific [3H] NMS binding to H10 cells transfected with scrambled siRNA and RGS3L-specific
3.4. RGS3L down-regulation in H10 cells switches coupling of GiPCRs from RhoA to Rac1 The data presented so far indicate that a high expression level of RGS3L is required for G iPCR-induced RhoA activation. We thus screened mammalian cells for expression levels of endogenous RGS3L. To identify this specific isoform of RGS3, we chose the 2D-gel-electrophoresis technique with subsequent immunoblotting. As shown in Fig. 6, rat RGS3L migrates at the predicted isoelectric point (Pi) and molecular mass of 4.79 and 61 kDa, respectively. The identical spot was detected by a second antibody raised against a different epitope in RGS3L (data not shown). We detected a rather high endogenous expression of RGS3L in serum-starved H10 cells (Fig. 7), an immortalized cell line derived from neonatal rat cardiomyocytes [15]. Most interestingly, like in human aortic smooth muscle cells [28], a profound down-regulation of RGS3L was observed when the cells were treated for 24 h with FGF-2 (50 ng/ml, Fig. 7A). A similar decrease was observed 72 h after transfection of H10 cells with RGS3L-specific siRNAs, whereas transfection with scrambled control siRNA led the endogenous expression level unaffected (Fig. 7B). We therefore studied which of the two Rho GTPases, Rac1 or RhoA, is activated by GiPCRs in these cells. Stimulation of
Fig. 10. Regulation of stress fibre formation by FGF-2 and carbachol in H10 cells. Serum-starved H10 cells were treated with PTX (A) and FGF-2 (B) for 24 h as indicated. Thereafter, cells were stimulated for 1 h with carbachol or solvent. Filamentous actin was stained with TRITC-phalloidin.
1236
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
siRNAs was 245 ± 35 and 275 ± 35 fmol/mg of protein, respectively. To test whether M2 mAChR activation causes RhoA typical cellular responses in living H10 cells we analysed the actin cytoskeleton. Carbachol-induced stress fibre formation in control H10 cells, which indicates RhoA activation in a PTXsensitive, thus Gi-mediated manner (Fig. 10A). Depletion of RGS3L in H10 cells by FGF-2 treatment prevented carbacholinduced stress fibre formation. FGF-2 by itself had no effect on the actin cytoskeleton (Fig. 10B). Taken together, the data on GiPCR-induced RhoA and Rac1 activation in H10 cells demonstrate that the level of endogenously expressed RGS3L directs whether GiPCRs activate Rac1 or RhoA in a PI3Kdependent manner. 4. Discussion The Rho GTPases RhoA and Rac1 play prominent roles in cellular responses to GPCRs in a variety of tissues [1–3]. It is generally believed that the activation of RhoA by these receptors is mediated by the PTX-insensitive α-subunits of G12 and/or Gq classes of G proteins, and the pathways leading to RhoA activation, specifically by the Gα12-type G proteins, are rather well defined. Activation of RhoA by Gα12/13 is mediated by the p115RhoGEF family of guanine nucleotide exchange factors (GEFs) known to contain a RGS-like domain by which they interact with the heterotrimeric G protein α-subunit and a DH domain which specifically catalyses activation of RhoA, but not of Rac1 [4]. Leukemia-associated RhoGEF can additionally be activated by Gαq/11 [29]. p63RhoGEF, a recently identified Rho-specific GEF, interacts exclusively with Gαq/11 [17,19]. In contrast, Gi proteins, generally believed not to mediate RhoA activation, have been found in several cell types to mediate activation of Rac1 by GPCRs. This GiPCR-induced Rac1 activation is mediated by Giβγ-dimers, often involves activation of PI3K and apparently requires GEFs like Tiam1, PRex and others [4]. We show here that three prototypical GiPCRs, the M2 mAChR, the α2AAR and the A1R, are able to activate both RhoGTPases, Rac1 and RhoA, in a PTX-sensitive manner. In HEK-TsA201 cells, activation of the M2 mAChR or α2AAR induced a strong PTX- and LY294002-sensitive activation of Rac1 without concomitant RhoA activation. In contrast, in H10 cells activation of M2 mAChR or A1R induced a pronounced PTX-sensitive activation of RhoA, without concomitant activation of Rac1. We obtained several lines of evidence that the preferential coupling of GiPCRs to either Rac1 or RhoA in HEK-TsA201 and H10 cells, respectively, is due to the differential expression level of RGS3L protein. 1) The expression of RGS3L in H10 cells was much higher compared to HEK-TsA201 cells. 2) Depletion of RGS3L in H10 cells by either FGF-2 or siRNA treatment redirected the GiPCR-induced activation of RhoGTPases from RhoA to Rac1. 3) We could reconstitute the GiPCR-induced RhoA activation by overexpression of RGS3L or even more efficient RGS3L N460A and thereby suppressed the GiPCR-induced Rac1 activation. In both cell types, the GiPCR-induced activation of Rac1 and RhoA seen at low and high expression level of RGS3L,
respectively, used apparently a similar signalling pathway. It was blunted by PTX and the PI3K inhibitor, LY294002. Furthermore, as studied in the HEK-TsA201 cell model, we were able to show that the RGS3L-mediated activation of RhoA involves, like the GiPCR-induced Rac1 activation, free Gβγdimers. Accordingly, a short splice variant of RGS3, i.e. RGS3S (aa 351–519) [11] which is deficient in one of the two Gβγ interaction domains identified in RGS3 (aa 313–390 and aa 391–458) [13], did not induce RhoA activation. As the RGS3Ldependent RhoA activation was accompanied by a decrease in Rac1 activation, a possible explanation would be that the inhibition of Rac1 activity is responsible for the increase in RhoA activation, a mechanism which has been proposed before [25,26]. Our data in H10 cells and HEK-TsA201 cells, however, exclude that possibility for the signalling pathway studied herein. 1) Inhibition of GiPCR-induced Rac1 activation by LY294002 did not induce concomitant RhoA activation. 2) Similarly, expression of βARK-CT suppressed activation of both RhoA and Rac1. 3) Expression of dominant-negative Rac1N17 blunted the M2 mAChR-induced Rac1 activation, but did not alter the M2 mAChR-induced RhoA activation in the presence of RGS3L N460A, which was however inhibited by dominant-negative RhoAN19. Our findings, therefore, indicate that RGS3L, independent of its catalytic GAP activity on Gα proteins, acts as an intracellular molecular switch for GiPCR signalling to the RhoGTPases Rac1 and RhoA, deciding whether these receptors, here shown for three prototypical GiPCRs, activate either Rac1 or RhoA. Although transfection studies in several cultured cell lines led to the conclusion that Gi-coupled receptors do no activate RhoA, a PTX- and LY294002-sensitive activation by formyl peptide receptors in human monocytes has been reported [30]. We demonstrate herein that two prototypical GiPCRs (M2 mAChRs, A1Rs) in H10 cells activate RhoA at a higher expression level of endogenous RGS3L and thereby induce the formation of Rho-dependent actin stress fibres. Interestingly, it has been reported that activation of the M2 mAChR leads via Gi-proteins and Rho activation to formation of actin stress fibres in cultured human airway smooth muscle cells [31,32]. Based on our data, it is feasible to assume that these cells as well as the monocytes tested in [30] express RGS3L at a rather high level. Throughout the literature there are several reports indicating an increase in RGS3 expression due to different causes. Amphetamine treatment of rats induced an about 3-fold up-regulation of RGS3 mRNA content in the striatum [33]. In human hearts with end stage heart failure an increase of RGS3 mRNA content and protein expression by 65% compared to non-failing transplants was detected [9]. Most important, our data in H10 cells as well as data in human aortic smooth muscle cells [28], demonstrate that the expression level of RGS3L and thus the prevalence by which the two RhoGTPases are activated by GiPCRs is under control of a physiological stimulus, i.e. FGF-2, which for example exerts protective effects against myocyte death and arrhythmias in myocardial infarction [34,35]. FGF-2 synthesis and secretion itself is controlled by α1-adrenergic receptors in cardiomyocytes, known inducers of hypertrophy [36]. The cellular
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
responses induced by RhoA or Rac1 activation in cardiac myocytes differ. Whereas Rac1 activation is associated with the generation of reactive oxygen species and its deleterious effects [37,38], RhoA apparently contributes to regulation of L-type Ca2+ and potassium channels [39,40]. Thus, persistent RhoA activation can cause arrhythmia [41]. On a cellular level, RhoA induces myofilament re-organisation and induces hypertrophic gene transcription by a plethora of transcriptional regulators, like SRF and GATA-4 [42,43]. Therefore, a switch from Rac1 to RhoA activation or vice versa, depending on the expression level of RGS3L, might result in drastically altered cellular responses to the GiPCR activation in the heart. As H10 cells are derived from cardiomyocytes, it is therefore tempting to speculate that the herein reported switch from Rac1 to RhoA activation by GiPCRs in response to increased RGS3L expression might contribute to the cardiac effects of FGF-2. Further studies are in progress to address this interesting question. In summary, our data demonstrate that the subset of RhoGTPases which are activated by GiPCRs is under control of the expression of RGS3L. For the very first time, we show that this isoform of RGS3 is not only a GAP for Gα-subunits, but by binding to Giβγ apparently dictates the type of RhoGTPase, Rac1 or RhoA, activated via a PI3K-dependent mechanism. Acknowledgements We thank K. Baden for the expert technical assistance. The gifts of various plasmids and antisera by A. Hall, J. H. Kehrl, M. Lohse, U. Mende, L. Hein, M. M. Chou, C. J. van Koppen, J. Mao, M. A. Schwartz and D. Wu are greatly appreciated. This work was supported by grants from the Deutsche Forschungsgemeinschaft, the Fonds der Chemischen Industrie, the Interne Forschungsförderung Essen and the Medizinische Fakultät Mannheim. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
S. Etienne-Manneville, A. Hall, Nature 420 (2002) 629. K. Burridge, K. Wennerberg, Cell 116 (2004) 167. J.H. Brown, D.P. Del Re, M.A. Sussman, Circ. Res. 98 (2006) 730. K. Rossman, C. Der, J. Sondek, Nat. Rev., Mol. Cell Biol. 6 (2005) 167. T. Wieland, C. Mittmann, Pharmacol. Ther. 97 (2003) 95. E.L. Riddle, R.A. Schwartzman, M. Bond, P.A. Insel, Circ. Res. 96 (2005) 401. M. Abramow-Newerly, A.A. Roy, C. Nunn, P. Chidiac, Cell. Signal. 18 (2006) 579. A. Scheschonka, C. Dessauer, S. Sinnarajah, P. Chidiac, C. Shi, J.H. Kehrl, Mol. Pharmacol. 58 (2000) 719. V.J. Owen, P.B. Burton, A.J. Mullen, E.J. Birks, P. Barton, M.H. Yacoub, Eur. Heart J. 22 (2001) 1015. C. Mittmann, C.H. Chung, G. Höppner, C. Michalek, M. Nose, C. Schüler, A. Schuh, T. Eschenhagen, J. Weil, B. Pieske, S. Hirt, T. Wieland, Cardiovasc. Res. 55 (2002) 778.
1237
[11] C. Mittmann, C. Schüler, C.H. Chung, G. Höppner, M. Nose, J.H. Kehrl, T. Wieland, Naunyn-Schmiedeberg's Arch. Pharmacol. 363 (2001) 456. [12] J.H. Kehrl, D. Srikumar, K. Harrison, G. Wilson, C.S. Shi, Genomics 79 (2002) 860. [13] C.S. Shi, S.B. Lee, S. Sinnarajah, C.W. Dessauer, S.G. Rhee, J.H. Kehrl, J. Biol. Chem. 276 (2001) 24293. [14] U. Rümenapp, M. Asmus, H. Schablowski, M. Wozniki, L. Han, K.H. Jakobs, M. Fahimi-Vahid, C. Michalek, T. Wieland, M. Schmidt, J. Biol. Chem. 276 (2001) 2474. [15] L. Jahn, J. Sadoshima, A. Green, C. Parker, K.G. Morgan, S. Izumo, J. Cell Sci. 109 (1996) 397. [16] O. Vögler, B. Nolte, M. Voss, M. Schmidt, K.H. Jakobs, C. van Koppen, J. Biol. Chem. 274 (1999) 12333. [17] S. Lutz, A. Freichel-Blomquist, U. Rümenapp, M. Schmidt, K.H. Jakobs, T. Wieland, Naunyn-Schmiedeberg's Arch. Pharmacol. 369 (2004) 540. [18] V. Benard, G. Bokoch, Methods Enzymol. 345 (2002) 349. [19] S. Lutz, A. Freichel-Blomquist, Y. Yang, U. Rümenapp, K.H. Jakobs, M. Schmidt, T. Wieland, J. Biol. Chem. 280 (2005) 11134. [20] T. Wieland, N. Bahtijari, X.B. Zhou, C. Kleuss, M. Simon, J. Biol. Chem. 275 (2000) 28500. [21] J. Mao, H. Yuan, W. Xie, M. Simon, D. Wu, J. Biol. Chem. 273 (1998) 27118. [22] W.J. Koch, B.E. Hawes, J. Inglese, L.M. Luttrell, R.J. Lefkowitz, J. Biol. Chem. 269 (1994) 6193. [23] C. Nobes, A. Hall, Biochem. Soc. Trans. 23 (1995) 456. [24] A. Ridley, H. Paterson, C. Johnston, D. Diekmann, A. Hall, Cell 70 (1992) 401. [25] E.E. Sander, J.P. ten Klooster, S. van Delft, R. van der Kammen, J.G. Collard, J. Cell Biol. 147 (1999) 1009. [26] A. Nimnual, L. Taylor, D. Bar-Sagi, Nat. Cell Biol. 5 (2003) 236. [27] M. Vogelsgesang, A. Pautsch, K. Aktories, Naunyn-Schmiedeberg's Arch. Pharmacol. (2006), doi:10.1007/s00210-006-0113-y Online First. [28] H. Cho, K. Harrison, O. Schwartz, J.H. Kehrl, Biochem. J. 371 (2003) 973. [29] M. Booden, D. Siderovski, C. Der, Mol. Cell. Biol. 22 (2002) 4053. [30] S. Huang, L.Y. Chen, B.L. Zuraw, R.D. Ye, Z.K. Pan, J. Biol. Chem. 276 (2001) 40977. [31] H. Togashi, C. Emala, I. Hall, C. Hirshman, Am. J. Physiol. 274 (1998) L803. [32] C. Hirshman, C. Emala, Am. J. Physiol. 27 (1999) L653. [33] S.A. Burchett, M.J. Bannon, J.G. Granneman, J. Neurochem. 72 (1999) 1529. [34] S.L. House, C. Bolte, M. Zhou, T. Doetschman, R. Klevitsky, G. Newman, J. Jel Schultz, Circulation 108 (2003) 3140. [35] S. Nishida, H. Nagamine, Y. Tanaka, G. Watanabe, Circ. J. 67 (2003). [36] Y. Jin, F. Sheikh, K.A. Detillieux, P.A. Cattini, Mol. Pharmacol. 57 (2000) 984. [37] K.S. Kim, K. Takeda, R. Sethi, J.B. Pracyk, K. Tanaka, Y.F. Zhou, Z.X. Yu, V.J. Ferrans, J.T. Bruder, I. Kovesdi, K. Irani, P. Goldschmidt-Clermont, T. Finkel, J. Clin. Invest. 101 (1998) 1821. [38] M. Ozaki, S. Haga, H.Q. Zhang, K. Irani, S. Suzuki, Cell Death Differ. 10 (2003) 508. [39] A. Yatani, K. Irie, T. Otani, M. Abdellatif, L. Wei, Am. J. Physiol. 288 (2005) H650. [40] T.G. Cachero, A.D. Morielli, E.G. Peralta, Cell 93 (1998) 1077. [41] V.P. Sah, S. Minamisawa, S.P. Tam, T.H. Wu, G.W. Dorn, J. Ross Jr., K.R. Chien, J.H. Brown, J. Clin. Invest. 103 (1999) 1627. [42] L. Wei, W. Zhou, J.D. Croissant, F.E. Johansen, R. Prywes, A. Balasubramanyam, R.J. Schwartz, J. Biol. Chem. 273 (1998) 30287. [43] F. Charron, G. Tsimiklis, M. Arcand, L. Robitaille, Q. Liang, J.D. Molkentin, S. Meloche, M. Nemer, Genes Dev. 15 (2001) 2702.
Regulator of G-protein signalling 3 redirects prototypical Gi-coupled receptors from Rac1 to RhoA activation Andreas Vogt a , Susanne Lutz a , Ulrich Rümenapp b , Li Han b , Karl H. Jakobs b , Martina Schmidt b,1 , Thomas Wieland a,⁎ a
Department of Experimental and Clinical Pharmacology and Toxicology, University of Heidelberg, Mannheim, Germany b Department of Pharmacology, University of Duisburg-Essen, Essen, Germany Received 21 December 2006; received in revised form 8 January 2007; accepted 8 January 2007 Available online 17 January 2007
Abstract The small GTPases, Rac1 and RhoA, are pivotal regulators of several essential, but distinct cellular processes. Numerous G-protein-coupled receptors signal to these GTPases, but with different specificities. Specifically, Gi-coupled receptors (GiPCRs) are generally believed to activate Rac1, but not RhoA, a process involving Gβγ-dimers and phosphatidylinositol 3-kinase (PI3K). Here we show that, depending on the expression level of the 519 amino acid isoform of regulator of G-protein signalling 3 (RGS3L), prototypical GiPCRs, like M2 muscarinic, A1 adenosine, and α2-adrenergic receptors, activate either Rac1 or RhoA in human embryonic kidney cells and neonatal rat cardiomyocyte-derived H10 cells. The switch from Rac1 to RhoA activation in H10 cells was controlled by fibroblast growth factor-2 (FGF-2), lowering the expression of RGS3L. Activation of both, Rac1 and RhoA, seen at low and high expression levels of RGS3L, respectively, was sensitive to pertussis toxin and the PI3K inhibitor LY294002 and mediated by Gβγ-dimers. We conclude that RGS3L functions as a molecular switch, redirecting GiPCRs via Gβγ-dimers and PI3K from Rac1 to RhoA activation. Considering the essential roles of Rac1 and RhoA in many signalling pathways, this additional function of RGS3L indicates a specific role of this protein in cellular signalling networks. © 2007 Elsevier Inc. All rights reserved. Keywords: Gi proteins; Gβγ-dimers; RhoA; Rac1; FGF-2; RGS proteins; RGS3
1. Introduction Members of the Rho GTPases family, like the best characterised family members RhoA, Rac1, and Cdc42, which has been first identified as regulators of actin cytoskeleton alterations, are now considered to be pivotal regulators of many signalling networks, participating in secretion, smooth muscle contraction, cell migration, membrane trafficking, gene transcription and cell cycle progression (for recent review see [1– 3]). G-protein-coupled receptors (GPCRs) in numerous cell types have been shown to activate Rho GTPases. It is generally ⁎ Corresponding author. Institut für Experimentelle und Klinische Pharmakologie und Toxikologie, Medizinische Fakultät Mannheim, Universität Heidelberg, Maybachstrasse 14, D-68169 Mannheim, Germany. Tel.: +49 621 330030; fax: +49 621 3300333. E-mail address: [email protected] (T. Wieland). 1 Present address: Department of Molecular Pharmacology, University of Groningen, The Netherlands. 0898-6568/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.cellsig.2007.01.003
assumed that the activation of RhoA by GPCRs is mediated by Gα-subunits of the G12 and/or Gq subfamilies. By contrast, Gicoupled receptors (GiPCRs) are believed to activate Rac1, but not RhoA, via Gβγ-dimers and involvement of phosphoinositide 3-kinase (PI3K) [3,4]. Therefore, the distinct cellular responses induced by pertussis toxin (PTX)-sensitive GiPCR signalling cascades and PTX-insensitive Gq/12 PCRs regulated pathways are often attributed to Rac1 and RhoA activation, respectively, and their different spectrum of downstream effectors [3]. RGS3, a member of the regulator of G-protein signalling (RGS) family, shares a high degree of homology within its RGS domain with the R4 subfamily and thus acts as GAP for the Gα-subunits of Gi and Gq family members [5–8]. Increases in cellular levels of RGS proteins apparently contribute to alterations in GPCR signalling [5,6] as for example seen in heart failure [9,10]. In contrast to other members of the R4 subfamily, RGS3 exists in several splice variants of which at least two exhibit an extended amino-terminal domain in
1230
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
Fig. 1. The GAP deficient RGS3L mutant N460A acts as a Gβγ-scavenger (A) Single turnover GTPase activity of recombinant Gαi1 measured in the absence (Basal) and presence of recombinant RGS3L or RGS3L N460A (N460A). (B) Luciferase production was measured in HEK-TsA201 cells transfected with the permanent active but GAP-sensitive Gαq mutant (GαqRC) (1 μg of DNA) alone and together with wild-type RGS3L, RGS3L N460A, RGS2 (2 μg of DNA each) or C3 transferase (C3T, 1 μg of DNA) as indicated. Mean ± SEM; n = 8–20. Gαq and RGS protein expression was monitored by immunoblotting (inset). (C) RGS3L N460A was immunoprecipitated from cell lysates with an anti-c-myc antibody and precipitated RGS3L N460A (upper panel) and Gβ (middle panel) were visualised by immunoblotting. Total content of Gβ in the cell lysates is shown as loading control (lower panel). (D) HEK-TsA201 cells were transfected with empty vector (Basal) and Gβ1 plus Gγ2 (10 μg of DNA each) in the absence or presence of RGS3L N460A (25 μg of DNA) or βARK-CT (100 μg of DNA). (A) After 48 h, basal [3H] inositol phosphate formation was measured. Data are means ± SEM (n = 4; ⁎⁎⁎, p b 0.001 vs. Gβ1γ2). Means ± SD of assay triplicates are given.
addition to the RGS domain [11,12]. It has been shown that the 519 amino acid isoform of RGS3 (RGS3L) binds Gβγ-dimers [13]. It thereby acts as a Gβγ-scavenger and inhibits the Gβγinduced activation of phospholipase Cβ (PLCβ), Akt and mitogen-activated protein kinase independently of its GAP activity. Herein we report that the expression level of RGS3L controls whether GiPCRS activate Rac1 or RhoA in human embryonic kidney (HEK) and neonatal rat cardiac myocyte derived H10 cells via a Gβγ- and PI3K-mediated pathway.
2. Materials and methods 2.1. Plasmid construction The cDNA encoding the 519 aa isoform of human RGS3 (a kind gift by Dr. J. H. Kehrl, Bethesda, MD, USA) was subcloned in the pCMV-Tag3b vector (Stratagene). Site-directed mutagenesis of RGS3L was performed using Quikchange (Stratagene) with the mutagenic primers RGS3N460Afw: GCATGCAAGGAGGTAGCGCTGGACTCCTACACG, RGS3N460Abw: CGTGTAGGAGTCCAGCGCTACCTCCTTGCATGC according to the manufacturer's protocol. The sequence of the mutant was verified by automated sequencing. The eukaryotic expression vectors encoding C3 transferase, M2 muscarinic acetylcholine receptor (M2 mAChR), α2A-adrenoceptor (α2AAR) and β-ARKCT were kind gifts of Dr. A. Hall, London, UK, Dr. C. van Koppen, Essen,
Germany, Dr. L. Hein, Freiburg, Germany and Dr. M. Lohse, Würzburg, Germany, respectively.
2.2. Cell culture and transfection Culture of human embryonic kidney (HEK-TsA201) cells and transfection of the cells (3 μg of total DNA per well on a 12-well plate for SRF activation, and up to 250 μg of DNA per 145-mm culture dish for GTPase pull-down assays) were performed as described before [14]. Assays were performed 24 h or 48 h after transfection in serum-starved cells. H10 cells were cultured as reported before [15], and then treated for 24 h in serum-free medium without and with fibroblast growth factor-2 (FGF-2) (50 ng/ml) or PTX (100 ng/ml) as indicated in the figure legends. For transfection of H10 cells with siRNAs, the cells were cultured on 35-mm dishes or 6-well plates and transfected with 5 μl Lipofectamine 2000 in 500 μl OptiMEM and 50 pmol of each siRNA according to the manufacturer's instructions (Invitrogen). If vector DNA was co-transfected with siRNA, 800 ng of the individual vector DNA was used in a total amount of 2.5 μg. The specific siRNAs (Ambion) were si-RGS3 (1): 5′GGAUAUGAAGAAUAAGCUGtt-3′, si-RGS3 (2): 5′-GGAAGGAAUCCUUUUCAGGtt-3′. To evaluate transfection efficiency, fluorescence labelled siRNA (Alexa Fluor 488, Qiagen) was used. Scrambled siRNA served as control. The expression of surface M2 mAChR was determined by [3H] Nmethylscopolamine ([3H] NMS) binding as described before [16].
2.3. Assay of SRF activation HEK-TsA201 cells seeded on 12-well plates were transfected with different expression plasmids together with the pSRE.L-luciferase reporter plasmid
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
1231
activity of the experimental reporter was normalised against the activity of the control vector.
2.4. Pull-down assays of activated Rho GTPases The cellular levels of GTP-loaded Rac1, RhoA and Cdc42 were determined as described [17,18]. In brief, the Rac/Cdc42-binding domain of Pak1 (GSTPBD) and the Rho-binding domain of rhotekin (GST-RBD) were expressed in and purified from E. coli. Subconfluent monolayers of HEK-TsA201 cells were transfected with the indicated amounts of plasmid DNA or the corresponding empty vectors and cultured for 48 h. Thereafter, the cells were rinsed with Hank's balanced salt solution and stimulated with the indicated agonists for the indicated periods of time at 37 °C. The cells were lysed in a buffer containing 1% (by vol.) Nonidet P-40, and the particular fraction was pelleted by centrifugation. The GTPase-containing supernatant was then incubated for 2 h at 4 °C with GST-PBD or GST-RBD bound to glutathione–sepharose beads. After twice washing of the beads, bound proteins were eluted with sample buffer and separated by SDS-PAGE. Rac1, RhoA and Cdc42 were then detected by immunoblotting with specific antibodies. Pull-down of GTP-loaded Rac1 and RhoA from lysates of H10 cells, pretreated with and without FGF-2 or PTX or transfected with siRNAs and then stimulated for 5 min at 37 °C with the indicated receptor agonists, was performed as in HEK-TsA201 cells.
2.5. 2D-gel electrophoresis Fig. 2. RGS3L N460A inhibits Gβ1γ2-induced Rac1 but enhances Gβ1γ2induced RhoA activation in HEK-TsA201 cells. The levels of Rac1-GTP, total Rac1, RhoA-GTP and total RhoA were determined in HEK-TsA201 cells transfected with Gβ1 plus Gγ2 (10 μg of DNA each), RGS3L N460A (N460 A, 25 μg of DNA), βARK-CT (100 μg of DNA) or the indicated combinations. Recombinant RGS3L N460A and βARK-CT expression was monitored by immunoblot analysis. (0.5 μg of DNA) and the pRL-TK control reporter vector (0.1 μg of DNA) as described [17]. At 24 h after transfection, luciferase activities were determined in cell extracts with the Dual Luciferase Reporter Assay System (Promega). The
Cell lysates (about 50 μg protein) were mixed with 70 μl rehydration buffer and applied to 7 cm IPG gel strips (GE Healthcare) containing a linear 3–10 pH gradient. Isoelectric focusing was carried out using an Ettan IPGphor unit (GE Healthcare). The subsequent SDS-PAGE was performed on 8% polyacrylamide gels.
2.6. Western blotting, immunoprecipitation Western blotting was performed according to standard procedures [17]. Specific primary antibodies against RGS3 (H300), Gβ (T20) Gαq (C19), βARK-CT (GRK2, C15), and c-myc (Santa Cruz Biotechnology) were used according to the manufacturer's recommendations. The RGS2 antibody was a
Fig. 3. RGS3L N460A redirects GiPCRs in HEK-TsA201 cells from PI3K-dependent Rac1 to RhoA activation. The levels of Rac1-GTP, total Rac1 (A) RhoA-GTP and total RhoA (B) were determined in HEK-TsA201 cells transfected with the M2 mAChR or α2AAR (12.5 μg of DNA) without and with βARK-CT or RGS3L N460A (N460A) and stimulated for 5 min without (−) and with (+) 1 mM carbachol or 10 μM adrenaline (Adr). Where indicated, cells were treated for 16 h with 100 ng/ml PTX or 15 min with 50 μM LY294002 before agonist stimulation. Recombinant RGS3L N460A expression was monitored by immunoblot analysis.
1232
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
2.10. Data presentation and statistical analyses All effector pull-down assays shown are representative experiments which have been repeated at least three times. Statistical analysis was performed by one-way ANOVA with Bonferroni's posttest. p b 0.05 was considered statistically significant.
3. Results 3.1. RGS3L N460A inhibits Gβγ-PI3K-dependent activation of Rac1 in HEK cells RGS3L inhibits Gβ1γ2-induced activation of co-expressed PLCβ2 in COS-7 cells, independent of its GAP activity on Gα proteins [13]. To dissect the Gβγ-scavenger property of RGS3L from its GAP function, the catalytically important Asn 460 in the RGS domain was mutated to Ala [20]. In contrast to wildtype RGS3L, the RGS3L N460A mutant neither accelerated the GTPase activity of purified recombinant Gαi1 (Fig. 1A) nor inhibited serum response factor (SRF)/RhoA-mediated transcriptional activity [21] induced by expression of the constitutively active but GAP sensitive Gαq mutant GαqRC (Fig. 1B). As reported before [13], RGS3L forms a complex
Fig. 4. Influence of dominant-negative Rac1 and RhoA on M2 mAChR-induced Rac1 and RhoA activation. HEK-TsA201 cells were transfected with expression plasmids encoding M2 mAChR (12.5 μg DNA), RGS3L N460A (N460A, 25 μg DNA), Rac1N17 (A) and RhoAN19 (B) (100 μg of DNA) alone or in combinations. 48 h after transfection, the cells were stimulated for 5 min without (−) or with (+) 1 mM carbachol. Levels of Rac1-GTP and RhoA-GTP were visualised by effector pull-down assays.
kind gift of Dr. J. H. Kehrl, Bethesda, MD. Detection was carried out with suitable secondary antibodies and Lumi Light Plus (Roche Applied Science). Co-immunoprecipitation was performed as described [19].
2.7. F-actin staining Subconfluent H10 cells were cultured on Culture Slides (Becton Dickinson), transfected with siRNA or treated with or without FGF-2 (50 ng/ml). Seventytwo hours posttransfection or after 24 h FGF-2 treatment, cells were stimulated with and without 1 mM carbachol for 1 h. Staining with tetramethylrhodamine isothiocyanate (TRITC)-conjugated phalloidin was performed as described [17].
2.8. Single turnover GTPase assay The single turnover GTPase assay of recombinant Gαi1 was performed exactly as described before [20] in the presence and absence of 1 μg purified RGS3L or RGS3L N460A.
2.9. Measurement of PLC activity Formation of [3H]inositol phosphates in HEK-TsA201 cells labelled with myo-[3H]inositol was measured 48 h posttransfection for 30 min at 37 °C in the presence of 10 mM LiCl as described before [15].
Fig. 5. Effects of RGS3L isoforms and Gβ1γ2-dimers on SRF-mediated gene transcription. (A) HEK-TsA201 cells were transfected with the pSRE.Lluciferase reporter encoding firefly luciferase (0.5 μg of DNA) and the pRLTK control reporter encoding Renilla luciferase (0.1 μg of DNA) alone (Control) or together with RGS3L, RGS3L N460A (N460A), RGS3S (2 μg of DNA each) or C3 ADP-ribosyl transferase (C3T, 1 μg of DNA) as indicated. (B) Cells were transfected with RGS3L, RGS3L N460A (0.5 μg of DNA), Gβ1 plus Gγ2 (0.3 μg of DNA each), alone and in the indicated combinations without or with βARK-CT (2 μg of DNA). Firefly luciferase activities were normalised against the level of expressed Renilla luciferase. Data are means ± SEM (n = 4– 14, ⁎⁎⁎p b 0.001 vs. Basal, #p b 0.01 vs. Gβ1γ2). Expression of recombinant proteins was monitored by immunoblot analysis.
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
1233
suppressed the Gβ1γ2-induced Rac1 activation. Similar data were obtained on Rac1 activation induced by endogenous Gβγdimers derived from receptor-activated Gi-proteins. Agonist (1 mM carbachol, 10 μM adrenaline) activation of the transfected Gi-coupled M2 muscarinic acetylcholine receptor (M2 mAChR) or α2A-adrenoceptors (α2AAR) led to marked GTP loading of Rac1. The GiPCR-induced response was fully suppressed by treatment with either PTX or the PI3K inhibitor LY294002 (50 μM). Similarly, co-expression of either β-ARKCT or RGS3L N460A (Fig. 3A), completely abolished agonistinduced Rac1 activation. As judged by [3H] NMS binding, cotransfection with either RGS3L N460A or βARK-CT did not alter the M2 mAChR expression. Specific binding was 288 ± 45, 305 ± 33, and 318 ± 45 fmol/mg of protein for M2 mAChR, M2 mAChR plus RGS3L N460A and M2 mAChR plus βARK-CT expressing cells, respectively. 3.2. RGS3L N460A induces Gβγ-PI3K-dependent activation of RhoA in HEK cells It is well known that different Rho GTPases can influence each other's activities. In particular, activation of Rac1 can lead to activation of RhoA, as demonstrated in Swiss 3T3 fibroblasts [23,24], while in other cell types, e.g. NIH 3T3 and HeLa cells, activated Rac1 down-regulates RhoA activity [25,26]. We therefore studied whether RGS3L, which blunted Rac1 activation, may affect RhoA activities. Indeed, expression of Gβ1γ2 in HEK-TsA201 cells also increased the amount of
Fig. 6. Comparison of RGS3L and RGS3L N460A-dependent RhoA activation. HEK-TsA201 cells transfected with the M2 mAChR together with RGS3L (A) or RGS3L N460A (N460A, B) were treated without (−) and with (+) carbachol for the indicated periods of time. Representative immunoblots for RhoA-GTP and total RhoA are shown in the upper panels, while in the lower panels mean ± SEM (n = 4–6) are presented, with the amount of RhoA-GTP found in unstimulated control cells set to 1.
with Gβγ-dimers and thereby inhibits Gβγ-induced PLC activation. In accordance, RGS3L N460A co-immunoprecipitated co-expressed Gβ1γ2 in HEK-TsA201 cells (Fig. 1C). Expression of Gβ1γ2 strongly increased activation of endogenous PLC in HEK-TsA201 cells, by about 8-fold (Fig. 1D). Co-expression of RGS3L N460A suppressed this cellular response, similarly as co-expression of the well-known Gβγ-scavenger, βARK-CT, the Gβγ-binding domain of the β-adrenergic receptor kinase [22], by about 50%. We therefore studied whether RGS3L N460A also inhibits the Gβγ-induced activation of Rac1. As illustrated in Fig. 2, overexpression of Gβ1γ2 in HEK-TsA201 cells induced the activation of endogenous Rac1 as measured by precipitation of Rac1-GTP by the Rac/Cdc42-binding domain of Pak1 and visualisation of bound Rac1 by immunoblotting [18]. Coexpression of either βARK-CT or RGS3L N460A completely
Fig. 7. Depletion of RGS3L expression in H10 cells. (A) Serum-starved and H10 cells were treated without (control) and with FGF-2 for 24 h as indicated. (B) Serum-starved H10 cells were transfected with scrambled siRNA (control) or RGS3L-siRNA (siRGS3L). The expression of RGS3L in lysates of control and FGF-2 treated and RGS3L-siRNA-transfected cells was analysed by 2D gel electrophoresis and immunoblot. Gβ content is shown as loading control. (C) Transfection efficacy was monitored by light (LM) and fluorescence microscopy (FM) of Alexa-Fluor-siRNA-transfected H10 cells.
1234
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
RhoA-GTP, measured by precipitation with a GST-fusion protein of the Rho-binding domain of rhotekin [18] (Fig. 2). Most interestingly however, expression of RGS3L N460A but not of βARK-CT induced a similar RhoA activation than Gβ1γ2 overexpression. When RGS3L N460A was co-expressed with Gβ1γ2, the level of activated RhoA was further increased. Co-expression of the Gβγ-scavenger βARK-CT fully prevented RhoA activation induced by the combination of Gβ1γ2 and RGS3L N460A or the individual agents, Gβ1γ2 and RGS3L N460A (Fig. 2). Thus, the data suggest that RGS3L N460A induce RhoA activation in a Gβγ-dependent manner. It was therefore of major interest to test whether RGS3L may switch Gβγ-liberating GiPCR signalling to induce RhoA activation. As reported before by others [21], agonist stimulation of the M2 mAChR (Fig. 3B) did not induce RhoA activation, neither in control nor in PTX-treated HEK-TsA201 cells. However, in cells expressing RGS3L N460A the agonistactivated M2 mAChR caused a strong, further increase in RhoA activity. PTX treatment abolished the receptor response. Coexpression of βARK-CT fully abrogated the synergistic RhoA
Fig. 9. siRNA-induced RGS3L depletion in H10 cells induces a switch in GiPCR-induced RhoGTPase activation. Serum-starved H10 cells were transfected with scrambled siRNA (−) or RGS3L-siRNA (+). The level of Rac1-GTP and RhoA-GTP was determined after 5 min stimulation without or with 1 mM carbachol or 10 μM CPA. Total Rac1 and RhoA content of cell lysates (30 μg of protein) is shown as loading controls.
activation induced by RGS3L N460A and the M2 mAChR. Similar to M2 mAChR, stimulation of α2AAR induced RhoA activation in the presence of RGS3L N460A in a LY294002 sensitive manner. Activation of RhoA induced by RGS3L N460A and GiPCRs was not due to inhibition of Rac1. Expression of the dominantnegative Rac1 mutant, Rac1N17, suppressed the M2 mAChRmediated activation of endogenous Rac1, similarly as expression of RGS3L N460A (Fig. 4A). However, Rac1N17 had no effects on activation of RhoA induced by either RGS3L N460A alone or when combined with the agonist-activated M2 mAChR. Vice versa, the expression of a dominant-negative RhoA mutant, RhoAN19, slightly increased the basal Rac1 activity in HEK-TsA201 cells. Thus, the M2 mAChR-induced Rac1 activation in the absence of RGS3L N460A appeared to be diminished (Fig. 4B). Importantly, in the presence of RGS3L N460A, expression of RhoAN19 completely suppressed the basal as well as the carbachol-stimulated activation of RhoA. Additionally, we studied whether another member of the Rho GTPases, i.e. Cdc42, might be involved in the observed RGS3L N460A-induced shift from Rac1 to RhoA activation. However, neither overexpression of Gβ1γ2 nor expression of RGS3L N460A alone caused an increase in the amount of Cdc42-GTP in HEK-TsA201 cells. Furthermore, although the agonistactivated M2 mAChR induced an increase in the level of Cdc42GTP, this was not altered by co-expression of RGS3L N460A (data not shown). These data therefore indicate that RGS3L N460A overexpression is able to redirect the GiPCR/Gβγ/PI3K signalling pathway from Rac1 to RhoA activation. 3.3. Comparison of RGS3L N460A-and wild-type RGS3Lmediated RhoA activation in HEK cells
Fig. 8. Regulation of Rac1 and RhoA activation by FGF-2 and carbachol in H10 cells. Serum-starved H10 cells were treated with PTX, LY294002 and FGF-2 as indicated. Thereafter, cells were stimulated for 5 min without (−) or with (+) carbachol. Levels of Rac1-GTP (A) and RhoA-GTP (B) were analysed by effector pull-down assays. Total Rac1 and RhoA content of cell lysates (30 μg of protein) is shown as loading controls.
As the data so far presented demonstrate a GiPCR- and Gβγdependent RhoA activation only in the presence of the GAP deficient N460A mutant of RGS3L, we studied whether and to what extent wild-type RGS3L is also able to contribute to RhoA activation. For this, we first examined the effects of RGS3L and
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
RGS3L N460A and RGS3S, an N-terminally truncated variant of RGS3 [11], lacking half of the Gβγ binding domain [13], on SRF-mediated gene transcription. Expression of RGS3L, but not of RGS3S, significantly increased luciferase expression by about 2.5-fold (Fig. 5A). Expression of RGS3L N460A increased transcriptional activity even stronger up to 6.8-fold. The RGS3L- and RGS3L N460A-induced luciferase production was fully suppressed by co-expression of Clostridium botulinum C3 transferase, which specifically inactivates RhoA–C, but not Rac1 or Cdc42 [27]. As shown in Fig. 5B, the transcriptional activity induced by RGS3L or RGS3L N460A was completely suppressed by co-expression of βARK-CT and mimicked by expression of Gβ1γ2. Co-expression of RGS3L or RGS3L N460 with Gβ1γ2 caused a synergistic increase in transcriptional activity. For example, while RGS3L and Gβ1γ2 alone increased the luciferase production by about 2.5- and 4.3-fold, respectively, the transcriptional activity was enhanced up to 14-fold and 19-fold by co-expression of RGS3L plus Gβ1γ2 and RGS3L N460A plus Gβ1γ2, respectively. Activation of RhoA by the M2 mAChR was also observed with wild-type RGS3L, although with different kinetics as with RGS3L N460A (Fig. 6A). Maximal RhoA activity induced by the M2 mAChR in cells overexpressing wild-type RGS3L was observed at 2 min stimulation with carbachol, while at 10 min the receptor response was lost. In contrast, in the presence of RGS3L N460A the carbachol-induced RhoA activation was sustained, for up to 30 min. Thus, wild-type RGS3L is similarly able to mediate GiPCR-induced RhoA activation via its direct interaction with Gβγ-dimers [13]. The difference in signal extent and kinetics between RGS3L and RGS3L N460A apparently reflects the Giinactivating GAP activity of RGS3L. Therefore, it was not seen when free Gβ1γ2 dimers are overexpressed (see Fig. 5B).
1235
the endogenously expressed M2 mAChR and adenosine A1 receptor with carbachol and N6-cyclopentyladenosine (CPA, 10 μM, not shown in Fig. 8), respectively, strongly increased RhoA-GTP levels in H10 cells, without inducing any measurable Rac1 activation (Figs. 8 and 9). The activation of RhoA by the two receptor agonists was fully sensitive to PTX treatment of the cells or the inhibition of PI3K by LY294002 (Fig. 8A). The capabilities of both carbachol and CPA to induce RhoA activation were lost in the FGF2-treated H10 cells. Instead, both agonists induced a robust activation of Rac1, which was fully blunted by PTX treatment and PI3K inhibition by LY294002 (Fig. 8B). The influence of the cellular depletion of RGS3L by siRNA on carbachol- or CPA-induced Rac1 and RhoA activation is shown in Fig. 9A and B, respectively. Similar to the reduction of RGS3L expression by FGF-2, the siRNA-induced RGS3L depletion caused a switch from a carbachol- or CPA-induced RhoA to Rac1 activation. Like in HEK-TsA201 cells the expression of M2 mAChR was similar in siRNA transfected cells. Specific [3H] NMS binding to H10 cells transfected with scrambled siRNA and RGS3L-specific
3.4. RGS3L down-regulation in H10 cells switches coupling of GiPCRs from RhoA to Rac1 The data presented so far indicate that a high expression level of RGS3L is required for G iPCR-induced RhoA activation. We thus screened mammalian cells for expression levels of endogenous RGS3L. To identify this specific isoform of RGS3, we chose the 2D-gel-electrophoresis technique with subsequent immunoblotting. As shown in Fig. 6, rat RGS3L migrates at the predicted isoelectric point (Pi) and molecular mass of 4.79 and 61 kDa, respectively. The identical spot was detected by a second antibody raised against a different epitope in RGS3L (data not shown). We detected a rather high endogenous expression of RGS3L in serum-starved H10 cells (Fig. 7), an immortalized cell line derived from neonatal rat cardiomyocytes [15]. Most interestingly, like in human aortic smooth muscle cells [28], a profound down-regulation of RGS3L was observed when the cells were treated for 24 h with FGF-2 (50 ng/ml, Fig. 7A). A similar decrease was observed 72 h after transfection of H10 cells with RGS3L-specific siRNAs, whereas transfection with scrambled control siRNA led the endogenous expression level unaffected (Fig. 7B). We therefore studied which of the two Rho GTPases, Rac1 or RhoA, is activated by GiPCRs in these cells. Stimulation of
Fig. 10. Regulation of stress fibre formation by FGF-2 and carbachol in H10 cells. Serum-starved H10 cells were treated with PTX (A) and FGF-2 (B) for 24 h as indicated. Thereafter, cells were stimulated for 1 h with carbachol or solvent. Filamentous actin was stained with TRITC-phalloidin.
1236
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
siRNAs was 245 ± 35 and 275 ± 35 fmol/mg of protein, respectively. To test whether M2 mAChR activation causes RhoA typical cellular responses in living H10 cells we analysed the actin cytoskeleton. Carbachol-induced stress fibre formation in control H10 cells, which indicates RhoA activation in a PTXsensitive, thus Gi-mediated manner (Fig. 10A). Depletion of RGS3L in H10 cells by FGF-2 treatment prevented carbacholinduced stress fibre formation. FGF-2 by itself had no effect on the actin cytoskeleton (Fig. 10B). Taken together, the data on GiPCR-induced RhoA and Rac1 activation in H10 cells demonstrate that the level of endogenously expressed RGS3L directs whether GiPCRs activate Rac1 or RhoA in a PI3Kdependent manner. 4. Discussion The Rho GTPases RhoA and Rac1 play prominent roles in cellular responses to GPCRs in a variety of tissues [1–3]. It is generally believed that the activation of RhoA by these receptors is mediated by the PTX-insensitive α-subunits of G12 and/or Gq classes of G proteins, and the pathways leading to RhoA activation, specifically by the Gα12-type G proteins, are rather well defined. Activation of RhoA by Gα12/13 is mediated by the p115RhoGEF family of guanine nucleotide exchange factors (GEFs) known to contain a RGS-like domain by which they interact with the heterotrimeric G protein α-subunit and a DH domain which specifically catalyses activation of RhoA, but not of Rac1 [4]. Leukemia-associated RhoGEF can additionally be activated by Gαq/11 [29]. p63RhoGEF, a recently identified Rho-specific GEF, interacts exclusively with Gαq/11 [17,19]. In contrast, Gi proteins, generally believed not to mediate RhoA activation, have been found in several cell types to mediate activation of Rac1 by GPCRs. This GiPCR-induced Rac1 activation is mediated by Giβγ-dimers, often involves activation of PI3K and apparently requires GEFs like Tiam1, PRex and others [4]. We show here that three prototypical GiPCRs, the M2 mAChR, the α2AAR and the A1R, are able to activate both RhoGTPases, Rac1 and RhoA, in a PTX-sensitive manner. In HEK-TsA201 cells, activation of the M2 mAChR or α2AAR induced a strong PTX- and LY294002-sensitive activation of Rac1 without concomitant RhoA activation. In contrast, in H10 cells activation of M2 mAChR or A1R induced a pronounced PTX-sensitive activation of RhoA, without concomitant activation of Rac1. We obtained several lines of evidence that the preferential coupling of GiPCRs to either Rac1 or RhoA in HEK-TsA201 and H10 cells, respectively, is due to the differential expression level of RGS3L protein. 1) The expression of RGS3L in H10 cells was much higher compared to HEK-TsA201 cells. 2) Depletion of RGS3L in H10 cells by either FGF-2 or siRNA treatment redirected the GiPCR-induced activation of RhoGTPases from RhoA to Rac1. 3) We could reconstitute the GiPCR-induced RhoA activation by overexpression of RGS3L or even more efficient RGS3L N460A and thereby suppressed the GiPCR-induced Rac1 activation. In both cell types, the GiPCR-induced activation of Rac1 and RhoA seen at low and high expression level of RGS3L,
respectively, used apparently a similar signalling pathway. It was blunted by PTX and the PI3K inhibitor, LY294002. Furthermore, as studied in the HEK-TsA201 cell model, we were able to show that the RGS3L-mediated activation of RhoA involves, like the GiPCR-induced Rac1 activation, free Gβγdimers. Accordingly, a short splice variant of RGS3, i.e. RGS3S (aa 351–519) [11] which is deficient in one of the two Gβγ interaction domains identified in RGS3 (aa 313–390 and aa 391–458) [13], did not induce RhoA activation. As the RGS3Ldependent RhoA activation was accompanied by a decrease in Rac1 activation, a possible explanation would be that the inhibition of Rac1 activity is responsible for the increase in RhoA activation, a mechanism which has been proposed before [25,26]. Our data in H10 cells and HEK-TsA201 cells, however, exclude that possibility for the signalling pathway studied herein. 1) Inhibition of GiPCR-induced Rac1 activation by LY294002 did not induce concomitant RhoA activation. 2) Similarly, expression of βARK-CT suppressed activation of both RhoA and Rac1. 3) Expression of dominant-negative Rac1N17 blunted the M2 mAChR-induced Rac1 activation, but did not alter the M2 mAChR-induced RhoA activation in the presence of RGS3L N460A, which was however inhibited by dominant-negative RhoAN19. Our findings, therefore, indicate that RGS3L, independent of its catalytic GAP activity on Gα proteins, acts as an intracellular molecular switch for GiPCR signalling to the RhoGTPases Rac1 and RhoA, deciding whether these receptors, here shown for three prototypical GiPCRs, activate either Rac1 or RhoA. Although transfection studies in several cultured cell lines led to the conclusion that Gi-coupled receptors do no activate RhoA, a PTX- and LY294002-sensitive activation by formyl peptide receptors in human monocytes has been reported [30]. We demonstrate herein that two prototypical GiPCRs (M2 mAChRs, A1Rs) in H10 cells activate RhoA at a higher expression level of endogenous RGS3L and thereby induce the formation of Rho-dependent actin stress fibres. Interestingly, it has been reported that activation of the M2 mAChR leads via Gi-proteins and Rho activation to formation of actin stress fibres in cultured human airway smooth muscle cells [31,32]. Based on our data, it is feasible to assume that these cells as well as the monocytes tested in [30] express RGS3L at a rather high level. Throughout the literature there are several reports indicating an increase in RGS3 expression due to different causes. Amphetamine treatment of rats induced an about 3-fold up-regulation of RGS3 mRNA content in the striatum [33]. In human hearts with end stage heart failure an increase of RGS3 mRNA content and protein expression by 65% compared to non-failing transplants was detected [9]. Most important, our data in H10 cells as well as data in human aortic smooth muscle cells [28], demonstrate that the expression level of RGS3L and thus the prevalence by which the two RhoGTPases are activated by GiPCRs is under control of a physiological stimulus, i.e. FGF-2, which for example exerts protective effects against myocyte death and arrhythmias in myocardial infarction [34,35]. FGF-2 synthesis and secretion itself is controlled by α1-adrenergic receptors in cardiomyocytes, known inducers of hypertrophy [36]. The cellular
A. Vogt et al. / Cellular Signalling 19 (2007) 1229–1237
responses induced by RhoA or Rac1 activation in cardiac myocytes differ. Whereas Rac1 activation is associated with the generation of reactive oxygen species and its deleterious effects [37,38], RhoA apparently contributes to regulation of L-type Ca2+ and potassium channels [39,40]. Thus, persistent RhoA activation can cause arrhythmia [41]. On a cellular level, RhoA induces myofilament re-organisation and induces hypertrophic gene transcription by a plethora of transcriptional regulators, like SRF and GATA-4 [42,43]. Therefore, a switch from Rac1 to RhoA activation or vice versa, depending on the expression level of RGS3L, might result in drastically altered cellular responses to the GiPCR activation in the heart. As H10 cells are derived from cardiomyocytes, it is therefore tempting to speculate that the herein reported switch from Rac1 to RhoA activation by GiPCRs in response to increased RGS3L expression might contribute to the cardiac effects of FGF-2. Further studies are in progress to address this interesting question. In summary, our data demonstrate that the subset of RhoGTPases which are activated by GiPCRs is under control of the expression of RGS3L. For the very first time, we show that this isoform of RGS3 is not only a GAP for Gα-subunits, but by binding to Giβγ apparently dictates the type of RhoGTPase, Rac1 or RhoA, activated via a PI3K-dependent mechanism. Acknowledgements We thank K. Baden for the expert technical assistance. The gifts of various plasmids and antisera by A. Hall, J. H. Kehrl, M. Lohse, U. Mende, L. Hein, M. M. Chou, C. J. van Koppen, J. Mao, M. A. Schwartz and D. Wu are greatly appreciated. This work was supported by grants from the Deutsche Forschungsgemeinschaft, the Fonds der Chemischen Industrie, the Interne Forschungsförderung Essen and the Medizinische Fakultät Mannheim. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
S. Etienne-Manneville, A. Hall, Nature 420 (2002) 629. K. Burridge, K. Wennerberg, Cell 116 (2004) 167. J.H. Brown, D.P. Del Re, M.A. Sussman, Circ. Res. 98 (2006) 730. K. Rossman, C. Der, J. Sondek, Nat. Rev., Mol. Cell Biol. 6 (2005) 167. T. Wieland, C. Mittmann, Pharmacol. Ther. 97 (2003) 95. E.L. Riddle, R.A. Schwartzman, M. Bond, P.A. Insel, Circ. Res. 96 (2005) 401. M. Abramow-Newerly, A.A. Roy, C. Nunn, P. Chidiac, Cell. Signal. 18 (2006) 579. A. Scheschonka, C. Dessauer, S. Sinnarajah, P. Chidiac, C. Shi, J.H. Kehrl, Mol. Pharmacol. 58 (2000) 719. V.J. Owen, P.B. Burton, A.J. Mullen, E.J. Birks, P. Barton, M.H. Yacoub, Eur. Heart J. 22 (2001) 1015. C. Mittmann, C.H. Chung, G. Höppner, C. Michalek, M. Nose, C. Schüler, A. Schuh, T. Eschenhagen, J. Weil, B. Pieske, S. Hirt, T. Wieland, Cardiovasc. Res. 55 (2002) 778.
1237
[11] C. Mittmann, C. Schüler, C.H. Chung, G. Höppner, M. Nose, J.H. Kehrl, T. Wieland, Naunyn-Schmiedeberg's Arch. Pharmacol. 363 (2001) 456. [12] J.H. Kehrl, D. Srikumar, K. Harrison, G. Wilson, C.S. Shi, Genomics 79 (2002) 860. [13] C.S. Shi, S.B. Lee, S. Sinnarajah, C.W. Dessauer, S.G. Rhee, J.H. Kehrl, J. Biol. Chem. 276 (2001) 24293. [14] U. Rümenapp, M. Asmus, H. Schablowski, M. Wozniki, L. Han, K.H. Jakobs, M. Fahimi-Vahid, C. Michalek, T. Wieland, M. Schmidt, J. Biol. Chem. 276 (2001) 2474. [15] L. Jahn, J. Sadoshima, A. Green, C. Parker, K.G. Morgan, S. Izumo, J. Cell Sci. 109 (1996) 397. [16] O. Vögler, B. Nolte, M. Voss, M. Schmidt, K.H. Jakobs, C. van Koppen, J. Biol. Chem. 274 (1999) 12333. [17] S. Lutz, A. Freichel-Blomquist, U. Rümenapp, M. Schmidt, K.H. Jakobs, T. Wieland, Naunyn-Schmiedeberg's Arch. Pharmacol. 369 (2004) 540. [18] V. Benard, G. Bokoch, Methods Enzymol. 345 (2002) 349. [19] S. Lutz, A. Freichel-Blomquist, Y. Yang, U. Rümenapp, K.H. Jakobs, M. Schmidt, T. Wieland, J. Biol. Chem. 280 (2005) 11134. [20] T. Wieland, N. Bahtijari, X.B. Zhou, C. Kleuss, M. Simon, J. Biol. Chem. 275 (2000) 28500. [21] J. Mao, H. Yuan, W. Xie, M. Simon, D. Wu, J. Biol. Chem. 273 (1998) 27118. [22] W.J. Koch, B.E. Hawes, J. Inglese, L.M. Luttrell, R.J. Lefkowitz, J. Biol. Chem. 269 (1994) 6193. [23] C. Nobes, A. Hall, Biochem. Soc. Trans. 23 (1995) 456. [24] A. Ridley, H. Paterson, C. Johnston, D. Diekmann, A. Hall, Cell 70 (1992) 401. [25] E.E. Sander, J.P. ten Klooster, S. van Delft, R. van der Kammen, J.G. Collard, J. Cell Biol. 147 (1999) 1009. [26] A. Nimnual, L. Taylor, D. Bar-Sagi, Nat. Cell Biol. 5 (2003) 236. [27] M. Vogelsgesang, A. Pautsch, K. Aktories, Naunyn-Schmiedeberg's Arch. Pharmacol. (2006), doi:10.1007/s00210-006-0113-y Online First. [28] H. Cho, K. Harrison, O. Schwartz, J.H. Kehrl, Biochem. J. 371 (2003) 973. [29] M. Booden, D. Siderovski, C. Der, Mol. Cell. Biol. 22 (2002) 4053. [30] S. Huang, L.Y. Chen, B.L. Zuraw, R.D. Ye, Z.K. Pan, J. Biol. Chem. 276 (2001) 40977. [31] H. Togashi, C. Emala, I. Hall, C. Hirshman, Am. J. Physiol. 274 (1998) L803. [32] C. Hirshman, C. Emala, Am. J. Physiol. 27 (1999) L653. [33] S.A. Burchett, M.J. Bannon, J.G. Granneman, J. Neurochem. 72 (1999) 1529. [34] S.L. House, C. Bolte, M. Zhou, T. Doetschman, R. Klevitsky, G. Newman, J. Jel Schultz, Circulation 108 (2003) 3140. [35] S. Nishida, H. Nagamine, Y. Tanaka, G. Watanabe, Circ. J. 67 (2003). [36] Y. Jin, F. Sheikh, K.A. Detillieux, P.A. Cattini, Mol. Pharmacol. 57 (2000) 984. [37] K.S. Kim, K. Takeda, R. Sethi, J.B. Pracyk, K. Tanaka, Y.F. Zhou, Z.X. Yu, V.J. Ferrans, J.T. Bruder, I. Kovesdi, K. Irani, P. Goldschmidt-Clermont, T. Finkel, J. Clin. Invest. 101 (1998) 1821. [38] M. Ozaki, S. Haga, H.Q. Zhang, K. Irani, S. Suzuki, Cell Death Differ. 10 (2003) 508. [39] A. Yatani, K. Irie, T. Otani, M. Abdellatif, L. Wei, Am. J. Physiol. 288 (2005) H650. [40] T.G. Cachero, A.D. Morielli, E.G. Peralta, Cell 93 (1998) 1077. [41] V.P. Sah, S. Minamisawa, S.P. Tam, T.H. Wu, G.W. Dorn, J. Ross Jr., K.R. Chien, J.H. Brown, J. Clin. Invest. 103 (1999) 1627. [42] L. Wei, W. Zhou, J.D. Croissant, F.E. Johansen, R. Prywes, A. Balasubramanyam, R.J. Schwartz, J. Biol. Chem. 273 (1998) 30287. [43] F. Charron, G. Tsimiklis, M. Arcand, L. Robitaille, Q. Liang, J.D. Molkentin, S. Meloche, M. Nemer, Genes Dev. 15 (2001) 2702.