RGS16 Attenuates Gαq-dependent p38 Mitogen-activated Protein Kinase Activation by Platelet-activating Factor*

The large gene family encoding the regulators of G protein signaling (RGS) proteins has been implicated in the fine tuning of a variety of cellular events in response to G protein-coupled receptor activation. Several studies have shown that the RGS proteins can attenuate G protein-activated extracellular signal-regulated kinase (ERK) group of mitogen-activated protein kinases. We demonstrate herein that the production of inositol trisphosphate and the activation of the p38 group of mitogen-activated protein kinases by the G protein-coupled platelet-activating factor (PAF) receptor was attenuated by RGS16 in both CHO cells transiently and stably expressing RGS16. The inhibition was not observed with RGS2, RGS5, and a functionally defective form of RGS16, RGS16R169S/F170C. The PAF-induced p38 and ERK pathways appeared to be preferentially regulated by RGS16 and RGS1, respectively. Overexpression of a constitutively active form of Gα11 (Gα11Q209L) prevented the RGS16-mediated attenuation of p38 activity, suggesting that Gαq/11 is involved in PAF activation of p38. The Gαq/11 involvement is further supported by the observation that p38 activation by PAF was pertussis toxin-insensitive. These results demonstrate for the first time that apart from ERK, p38 activation by a G protein-coupled receptor can be attenuated by an RGS protein and provide further evidence for the specificity of RGS function in G protein signaling pathways.

RGS proteins serve as GTPase-activating proteins of a variety of G protein ␣-subunits, terminating the signaling process by G protein-coupled receptors (18 -21). To date, about 20 mammalian RGS proteins have been identified, all of which are defined by a highly conserved domain of 120 amino acid residues in length. Underscoring the significance of their sequence similarity within the RGS domain, RGS proteins seem to have functional promiscuity as assayed by both their G protein binding and functional resemblance to that of the Sst2 protein in yeast. It is noteworthy that the RGS proteins vary in size, ranging from 21 to 150 kDa, and contain divergent sequences flanking the conserved RGS domain. The divergent sequences among their flanking regions may be the specificity determinants for RGS function. Functional specificity is best demonstrated by the finding that the RGS domain-containing protein p115 RhoGEF specifically binds to G␣ 12 and G␣ 13 and regulates Rho, mediating cell morphology, adhesion, and cell proliferation (22,23). In addition, recent studies have shown that RGS4 and G␣-interacting protein block G␣ i -mediated inhibition of adenylyl cyclase (24), whereas RGS1, RGS2, RGS3T, and RGS4 attenuate G␣ i -or G␣ q -regulated activation of the ERK group of MAPK (25)(26)(27). Furthermore, RGS3, RGS4, and G␣-interacting protein suppress G␣ q -mediated synthesis of inositol trisphosphate (24, 26 -28).
We have previously identified and characterized RGS16 (also known as RGS-r; Ref. 29); and we have shown that RGS16 binds G␣ i2 , G␣ i3 , and G␣ o subunits in the transition state (30) and that RGS16 has GTPase-activating activity on these G proteins (31). In this report, we show that RGS16 inhibits platelet-activating factor (PAF)-stimulated p38 MAPK activation. The RGS inhibition of p38 can be abolished by the mutant G␣ 11 Q209L, which indicates an involvement of G␣ q/11 in the PAF signaling. Moreover, we show that RGS members have differential attenuating effects on the G protein-mediated activation of ERK and p38. Our findings show that p38 activation by a G protein-coupled receptor can be attenuated by an RGS protein and provide further evidence that individual RGS members act as distinct regulators for different G protein signaling pathways.
Transient Transfection of Cells-Chinese hamster ovary (CHO) cells were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum, 100 IU penicillin, 100 g/ml streptomycin, and 2 mM glutamine. Cells plated in 60-mm dishes were transfected with various plasmids using DOSPER according to the manufacturer's instructions (Boehringer Mannheim). In each transfection mixture, the total amount of transfected DNA was adjusted to 5 g with the empty vector pCMV5 where necessary. Two g each of pCMV5-RGS constructs was transfected together with 1 g of pCDNA3-PAFR, and 2 g of pCDNA3-FLAG-p38 MAPK or pCDNA3-FLAG-ERK2. For cotransfection with G protein plasmids, 1.5 g of each indicated G protein subunit was cotransfected with 1.5 g of pCMV5-RGS, 1.5 g of pCDNA3-FLAG-p38, and 0.5 g of pCDNA3-PAFR. The transfection medium was replaced with fresh growth medium after 24 h, and cells were harvested 40 h after transfection.
Establishment of RGS16-inducible Expression in CHO Cells-The ecdysone-inducible expression system, based on the Drosophila molting induction system and modified for mammalian cells, uses the steroid hormone ecdysone analog muristerone A to activate expression of the gene of interest via a heterodimeric nuclear receptor (Invitrogen). EcR-CHO Chinese hamster ovary cells containing the ecdysone receptor (Invitrogen) were transfected with pIND-RGS16 or pIND-RGS16 R169S/F170C constructs using DOSPER (Boehringer Mannheim), and 24 h after transfection, the cells from each dish were diluted into a 150-mm dish and selected in medium containing 800 g/ml of G418 (Life Technologies, Inc.). Clones resistant to G418 were isolated after 2-3 weeks and expanded to test for RGS16 expression in response to muristerone A induction. Muristerone A was added to a final concentration of 1 M; after 24 h, cell lysates obtained before and after induction were analyzed by SDS-polyacrylamide gel electrophoresis and immunoblotted with the RGS16 antibody (see below). Multiple were obtained, and at least two clones with little leakiness for each construct were expanded and tested for hormonal responses. These expression cell lines yielded similar experimental results in both basal and stimulated IP 3 production as well as MAPK activities to their respective lines expressing the wild type RGS16 (designated CHO-R16) Measurement of Inositol 1,4,5-Trisphosphate Production-The levels of IP 3 in CHO cells were measured by a competitive radioreceptor assay using the Biotrak TM D-myo-inositol 1,4,5-trisphosphate assay system (Amersham Pharmacia Biotech). Briefly, cells were separately treated with the indicated ligands for 10 min; IP 3 was extracted with 15% (v/v) trichloroacetic acid and neutralized with NaHCO 3 . The samples and working IP 3 standards were incubated with the binding protein in the presence of [ 3 H]IP 3 , and the amount of radioactivity bound was measured by liquid scintillation counting. The amount of IP 3 in the samples was determined by interpolation from the standard curve.
Immunoprecipitation of MAPK and Kinase Assays-Cells were serum-starved for 2 h and stimulated with the agonists indicated. After a wash with PBS, the cells were lysed in ice-cold lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton, 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerolphosphate, 1 mM sodium orthovanadate, 1 g/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride). Kinases were immunoprecipitated using either the mouse monoclonal anti-FLAG M2 (Eastman Kodak Co.), rabbit polyclonal anti-p38, or anti-ERK2 antibodies bound to Protein A/G Plus-agarose beads (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Kinase assays were performed on the washed immunoprecipitates in a 50-l reaction mixture comprising the kinase buffer (25 mM Tris-HCl, pH 7.4, 5 mM ␤-glycerolphosphate, 2 mM dithiothreitol, 0.1 mM sodium orthovanadate, 10 mM MgCl 2 ), 5 Ci of [␥-32 P]ATP (Amersham Pharmacia Biotech), kinase substrate (1 g of affinity-purified GST-ATF2 (residues 1-109; Ref. 33) for p38, 5 g of myelin basic protein (MBP) (Sigma) for ERK), and unlabeled ATP (50 M for p38; 100 M for ERK). The reactions were carried out at 30°C for 30 min (p38) or 20 min (ERK) and terminated by adding 50 l of 2ϫ SDS-polyacrylamide gel electrophoresis sample buffer. The boiled samples were separated by SDSpolyacrylamide gel electrophoresis, and the radioactivity incorporated into the substrate proteins was measured by an imaging analyzer (Molecular Dynamics model 425E) and detected by autoradiography. The amount of the kinase in each immunoprecipitate was quantified by immunoblotting.
Production of RGS16 Antibody and Western Blot Analysis-Bacterially expressed glutathione S-transferase fusion RGS16 proteins, generated as described previously (30), were used to raise antibodies in rabbits (Bioprocessing Technology Center, National University of Singapore). The specific immunoglobulins were purified from serum samples by affinity binding as before (35). For Western blotting, protein samples were separated on 10% SDS-polyacrylamide gels and transferred onto polyvinylidene difluoride membranes (Immobilon-P; Milli-

FIG. 2. PAF-stimulated p38 activation is attenuated by RGS16.
A, suppression of PAF-stimulated p38 MAPK activity by RGS16 but not RGS16 R169S/F170C . Cells before (Ϫ) and after (ϩ) induction by muristerone A were separately exposed for 90 s to 100 nM PAF or for 15 min to 0.4 M sorbitol. Endogenous p38 was immunoprecipitated and assayed for its kinase activity using GST-ATF2 as the substrate. The amount of the kinase in each immunoprecipitate was quantified by immunoblotting. B, PAF-stimulated ERK activity is less affected by RGS16. Cells were treated as above or with 160 nM PMA for 20 min. Endogenous ERK2 was immunoprecipitated and assayed for its kinase activity using MBP as the substrate. Data are expressed as -fold kinase activation compared with kinase activity produced in unstimulated control cells. The values represent the means Ϯ S.E. from three separate experiments. pore Corp.). The membranes were incubated for 2 h with rabbit polyclonal anti-RGS16 antibody (1:500), anti-FLAG, anti-p38, or anti-ERK antibodies (1:1000, Santa Cruz Biotechnology), and bound antibodies were visualized by enhanced chemiluminescence (Amersham Pharmacia Biotech) using horseradish peroxidase-conjugated antibodies. 3 Production by RGS16 -To study the biological function of RGS16 in G protein-mediated signaling pathway, we employed the ecdysone-inducible expression system to establish stable cell lines expressing RGS16 (CHO-R16) and its mutant RGS16 R169S/F170C (CHO-M18). As shown in Fig. 1, A and B (insets), induction of these cells with 1 M muristerone A significantly increased the expression of the RGS16 proteins.

Inhibition of PAF-stimulated IP
It has been shown that activation of the G protein-coupled PAF receptor can induce IP 3 production (27,36,37). We observed a 6-fold increase in IP 3 production when CHO cells were stimulated with PAF (Fig. 1A), confirming the presence of endogenous PAF receptors in CHO cells. 2 However, this PAFinduced increase in IP 3 was almost entirely suppressed by the induced expression of RGS16. On the other hand, induced expression of the mutant RGS16 R169S/F170C , which can no longer bind any G␣ subunit in vitro and has lost the ability to inhibit pheromone signaling in yeast (30), did not affect PAFinduced IP 3 production (Fig. 1B). Treatment of CHO-R16 cells with other agonists of G protein-coupled receptors, lysophosphatidic acid, N-formyl-methionyl-leucyl-phenylalanine, and interleukin-8, all marginally increased IP 3 levels; these increases were unaffected by induced RGS16 expression (Fig.  1A). These results were repeated with at least two other clones.
PAF-stimulated p38 MAPK Activation Is Attenuated by RGS16 -The G protein-coupled PAF receptor is known to activate other downstream targets besides phospholipase C; these include the p38 group of MAPK (15,38). Since RGS16 expression attenuated the PAF-induced IP 3 production, we asked if the PAF-stimulated p38 activity could be inhibited by RGS16 by measuring the activity of endogenous p38 using GST-ATF2 as substrate in stable cell lines before and after induction with muristerone A. For both RGS16-and mutant RGS16 R169S/F170C -expressing cells, experiments were conducted on at least two other clones for each construct, and the results obtained were similar to CHO-R16 and CHO-M18 cells, respectively.
In the CHO-R16 cells, stimulation with PAF increased the p38 activity 4.5-fold ( Fig. 2A, top). Pretreatment of the same cells with muristerone A to induce RGS16 expression abolished the PAF-stimulated p38 activation. By contrast, in CHO-M18 cells, induced expression of the mutant RGS16 R169S/F170C , which is defective in G protein binding, did not inhibit the p38 activation by PAF ( Fig. 2A, bottom). RGS16 and its mutant had negligible effect on p38 activation by sorbitol.
Since PAF has been demonstrated to activate ERK, another member of the MAPK family (39), we examined the endogenous ERK activity of CHO-R16 cells using MBP as the substrate. As expected, in cells without RGS16 induction, PAF treatment markedly increased ERK activity (Fig. 2B). RGS16 expression diminished the PAF-stimulated ERK activity slightly but did not impair ERK activation by the phorbol ester PMA. Therefore, compared with the extent of p38 suppression, the inhibitory effect of RGS16 on ERK was much less. We did not detect any activation of c-Jun NH 2 -terminal kinase/stress-activated protein kinase by PAF, as measured by similar immunokinase assays (data not shown), in agreement with previous results by Nick et al. (38).
Differential Regulation of PAF-activated p38 and ERK Pathways by RGS1 and RGS16 -To determine whether the inhibition of PAF-stimulated p38 activity was specific to RGS16, we separately transfected RGS1, RGS2, RGS5, or RGS16, along with PAFR and FLAG-tagged p38, into CHO cells. Assayed in the Saccharomyces cerevisiae strain YDM400 (YPH499 sst2-⌬2 strain, provided by H. G. Dohlman and J. Thorner), RGS1, RGS2, and RGS5 exhibited similar activity to that of RGS16 in the attenuation of pheromone signaling. 3  4B), it was only partially inhibited by RGS1 (Fig. 4B) and was not inhibited by RGS2 or RGS5 (Fig. 3).
It has been shown that RGS1 markedly impaired ERK activation by PAF (25). We asked whether ERK activation by PAF was affected by RGS1 and RGS16 in a similar manner as p38 activation. In cells cotransfected with FLAG-tagged ERK, pronounced activation of ERK was observed upon treatment with the phorbol ester PMA, which was not affected by RGS1 or RGS16 (Fig. 4A). The inability of RGS1 and RGS16 to inhibit PMA-induced activation of ERK is consistent with the proposed role for RGS proteins as direct regulators of G proteins. As previously reported (25), RGS1 dramatically suppressed ERK activation by PAF (Fig. 4A). However, inhibition of PAF-stimulated ERK activity by RGS16 was to a much lower extent (Fig.  4A), suggesting a differential regulation by RGS proteins on PAF-activated MAPK pathways.
PAF-induced p38 Activation Is Pertussis Toxin-insensitive-ERK activation by the heptahelical PAF receptor has been shown to be mediated by both pertussis toxin (PTX)-sensitive and -insensitive G proteins (39). In the case of p38 activation by PAF, it is unclear which G proteins are involved. We compared the effects of PTX on the PAF-induced signals in transiently transfected CHO cells. Preincubation with 100 ng/ml PTX for 24 h partially abolished PAF-stimulated ERK activity (Fig. 5). However, p38 activation by PAF was insensitive to PTX treatment, indicating that p38 activation by PAF is independent of G␣ i . G␣ 11 Q209L, but Not G␣ i2 Q205L, Can Overcome the Suppression Effect of RGS16 on PAF-stimulated p38 Activity-Mutations in the catalytic domain of G␣ subunits that inhibit their intrinsic GTPase activity are known to render these proteins constitutively active (40). To identify the G␣ subunit(s) involved in the RGS16-mediated inhibition, we tested the effect of PAF on p38 MAPK in CHO cells transiently transfected with GTPase-deficient mutants of G␣ 11 (G␣ 11 Q209L) or G␣ i2 (G␣ i2 Q205L).
In the absence of PAF stimulation, the basal levels of p38 activities in cells transfected with G␣ i2 , G␣ i2 Q205L, or G␣ 11 were similar (Fig. 6). Stimulation by PAF in the G␣ i2 -, G␣ i2 Q205L-, or G␣ 11 -transfected cells increased p38 activity by about 5-fold each. RGS16 expression significantly diminished p38 activity to levels close to basal, whereas RGS1 showed less effect, consistent with results in Fig. 4B.
In contrast, in the absence of PAF treatment, cells overexpressing the GTPase-deficient mutant G␣ 11 Q209L showed a high basal level of p38 activity, which was not affected by the expression of RGS1 or RGS16 (Fig. 6). Stimulation by PAF in these G␣ 11 Q209L-transfected cells did not further increase the activity of p38, suggesting that PAF stimulation of p38 MAPK occurs via G␣ q/11 and that RGS16 interacts with G␣ q/11 to FIG. 5. Effects of pertussis toxin on PAF-stimulated p38 and ERK activation. CHO cells were transiently transfected with 2 g of RGS16 or vector plus 2 g of FLAG-p38 or FLAG-ERK2 and 1 g of PAFR. Transfected cells were treated for 20 h with (ϩ) or without (Ϫ) 100 ng/ml PTX before stimulation for 90 s with 100 nM PAF. Immunokinase assays were performed as described in the legend to Fig. 4. Data are expressed as -fold kinase activation compared with kinase activity produced in unstimulated, vector-transfected cells. The values represent the means Ϯ S.E. from three separate experiments.
FIG. 6. G␣ 11 Q209L, but not G␣ i2 Q205L, can overcome the suppression effect of RGS16 on PAF-induced p38 activation. CHO cells were transiently transfected with 1.5 g of FLAG-p38, 1.5 g of vector, RGS1 or RGS16, plus 1.5 g of G␣ i2 , G␣ i2 Q205L, G␣ q/11 , or G␣ q/11 Q209L, and 0.5 g of PAFR. Transfected cells were unstimulated (Ϫ) or stimulated (ϩ) for 90 s with 100 nM PAF. Following immunoprecipitation of FLAG-p38, the kinase activities were assayed using GST-ATF2, and the amount of the kinase in each immunoprecipitate was quantified by immunoblotting. Data are expressed as -fold p38 activation compared with p38 activity produced in unstimulated, vectortransfected cells. The values represent the means Ϯ S.E. from three separate experiments. attenuate p38 activation induced by PAF.
RGS16 Does Not Affect G␤ 1 ␥ 2 -mediated p38 Activation-It has been reported that overexpression of G␤ 1 ␥ 2 can stimulate p38 MAPK activity (16). We asked whether accelerated G␣ inactivation by RGS16 could also lead to suppression of the G␤␥-stimulated p38 activity. We included G␤ 1 and G␥ 2 in the cotransfections and assayed for p38 activity in the presence or absence of overexpressed RGS16. Cells cotransfected with G␤ 1 ␥ 2 showed approximately 3-fold higher p38 activity than the control (Fig. 7). Overexpression of either RGS1 or RGS16 did not affect the G␤ 1 ␥ 2 -stimulated p38 activity. Furthermore, PAF treatment in the G␤ 1 ␥ 2 -overexpressing cells resulted in an additive increase in p38 activity. Consistent with observations above (Figs. 4B and 6), this additive increase was only partially suppressed by RGS1 expression but was suppressed to uninduced levels by the expression of RGS16. In other words, the PAF induction component was entirely suppressed by RGS16. These results also suggest that the p38 activation pathway mediated by G␤ 1 ␥ 2 is different from that activated by PAF. DISCUSSION Our results demonstrate for the first time that activation of p38 MAPK by the G protein-coupled PAF receptor can be attenuated by an RGS family member, RGS16. Such an inhibitory effect was not observed with RGS2, RGS5, or the mutant RGS16 R169S/F170C which is defective in G protein binding. RGS1 and RGS16 showed preferential regulation for ERK and p38 MAPK, respectively, in response to PAF stimulation. The RGS16 attenuation of p38 can be inhibited by the GTPasedeficient mutant G␣ 11 Q209L, but not G␣ i2 Q205L, indicating that G␣ q/11 mediates signaling between the G protein-coupled PAF receptor and p38 MAPK.
We and others have shown that RGS proteins are rather promiscuous with respect to various G protein ␣-subunits in manifesting their GTPase-activating activity. RGS4, for instance, binds to and serves as a GTPase-activating protein for, G␣ i1 , G␣ i2 , G␣ i3 , G␣ o , G␣ t , G␣ z , and G␣ q (41)(42)(43)(44). Similarly, RGS16 binds to G␣ i2 , G␣ i3 , G␣ o , G␣ t (29,30), and G␣ q . 3 The promiscuity of RGS proteins is also suggested by the fact that many if not all of the RGS proteins can complement the func-tion of Sst2 in yeast pheromone signaling when assayed using a ⌬sst2 strain 3 (25,30,45,46). However, investigations of RGS function in mammalian cells indicate that individual RGS family members seem to be selective in the regulation of specific G protein-linked signaling pathways. RGS1, RGS2, RGS3, and RGS4 differ in their ability to impair interleukin-8 receptor signaling of ERK, with RGS4 showing the greatest inhibition followed by RGS3, RGS1, and RGS2 (25). G protein-gated inward rectifier potassium channels evoked by agonist activation of muscarinic m2 receptors are dramatically accelerated by coexpression of RGS1, RGS3, or RGS4, but not RGS2 (47). In a separate study, RGS3, but not RGS1, RGS2, or RGS4, can suppress the IP 3 responses induced by the gonadotropin-releasing hormone (28). The present findings that PAF-stimulated p38 MAPK activity is substantially attenuated by RGS16 and partially attenuated by RGS1 (but not by RGS2, RGS5, or RGS16 R169S/F170C ) and that RGS1 is more effective than RGS16 in blocking PAF-stimulated ERK activity provide further evidence of selectivity. Moreover, it should be noted that assays in yeast may not completely reflect what actually occurs in mammalian cells, since yeasts do not contain the same repertoire of components. Indeed, we have identified a mammalian membrane protein that specifically binds to the NH 2 -terminal portion of a small subset of RGS proteins. 4 It is conceivable that this membrane protein may somehow determine RGS functional specificity in G protein interaction or cross-talk with other signaling pathways. In addition, it is likely that other factors such as tissue and cell type distribution, temporal expression, and post-translational modification act in concert to dictate the specificity of RGS function.
Several lines of evidence suggest that p38 activation by G protein-coupled receptor can be mediated by G␣ q/11 and G␤␥ (16,48,49). G␣ q can directly stimulate the nonreceptor Bruton's tyrosine kinase Btk, which is required for the activation of p38, as demonstrated in cells deficient for Btk (48). The G␣ q/ 11-coupled receptor agonist phenylephrine activates p38 MAPK in perfused rat heart (49). Moreover, p38 MAPK activation by 4 C. Chen and S.-C. Lin, unpublished results. FIG. 7. RGS16 does not affect G␤ 1 ␥ 2mediated p38 activation. CHO cells were transiently transfected with 1.5 g of FLAG-p38, 1.5 g of RGS1 or RGS16, plus vector alone or with 0.75 g each of G␤ 1 and G␥ 2 and 0.5 g of PAFR. Transfected cells were unstimulated (Ϫ) or stimulated (ϩ) for 90 s with 100 nM PAF. Immunokinase assays were performed as described in the legend to Fig. 6. Data are expressed as -fold p38 activation compared with p38 activity produced in unstimulated, vector-transfected cells. The values represent the means Ϯ S.E. from three separate experiments. m1 muscarinic acetylcholine receptor involves both G␣ q/11 and G␤␥, while m2 muscarinic acetylcholine and ␤-adrenergic receptors act through G␤␥ in human embryonic kidney 293 cells (16). It is intriguing that RGS4 can inhibit G␤␥-activated inwardly rectifying potassium channels (47), whereas RGS16 appears not to play a role in the modulation of G␤ 1 ␥ 2 -stimulated p38 MAPK activation. The difference may be explained by the possibility that different RGS proteins may be linked to different receptors, although it cannot be ruled out that RGS16 may interfere with signaling pathways mediated by other G␤␥ dimers.
The functional significance of RGS16 inhibition of PAF-stimulated p38 MAPK activity is as yet unclear. PAF exhibits a wide variety of physiological and pathophysiological effects in various cells and tissues, such as proto-oncogene expression in neuronal cells, respiratory and cardiovascular functions, and inflammatory and immune responses (50). It is conceivable that p38 mediates many aspects of the PAF signaling. p38 MAPK is thought to play an important role in the regulation of cellular responses during infection (51)(52)(53). Perhaps RGS16 is recruited to fine tune the immune response through its effects on the expression of proinflammatory molecules. Indeed, we have observed changes in RGS16 expression when lymphoid cell lines are challenged with PAF. 5 As RGS1 and RGS16 display differential regulation of PAF-activated MAPK pathways, it is possible that both RGS1 and RGS16 are necessary for fine tuning the exquisitely orchestrated events elicited by PAF. Taken together, our data show that RGS proteins display differential regulation of G protein-mediated p38 and ERK pathways, pointing to distinct modulatory activities of different RGS proteins in G protein-regulated signal transduction.