Phosphorylation of Phosducin and Phosducin-like Protein by G Protein-coupled Receptor Kinase 2*

G protein-coupled receptor kinase 2 (GRK2) is able to phosphorylate a variety of agonist-occupied G protein-coupled receptors (GPCR) and plays an important role in GPCR modulation. However, recent studies suggest additional cellular functions for GRK2. Phosducin and phosducin-like protein (PhLP) are cytosolic proteins that bind Gβγ subunits and act as regulators of G-protein signaling. In this report, we identify phosducin and PhLP as novel GRK2 substrates. The phosphorylation of purified phosducin and PhLP by recombinant GRK2 proceeds rapidly and stoichiometrically (0.82 ± 0.1 and 0.83 ± 0.09 mol of Pi/mol of protein, respectively). The phosphorylation reactions exhibit apparent K m values in the range of 40–100 nm, strongly suggesting that both proteins could be endogenous targets for GRK2 activity. Our data show that the site of phosducin phosphorylation by GRK2 is different and independent from that previously reported for the cAMP-dependent protein kinase. Analysis of GRK2 phosphorylation of a variety of deletion mutants of phosducin and PhLP indicates that the critical region for GRK2 phosphorylation is localized in the C-terminal domain of both phosducin and PhLP (between residues 204 and 245 and 195 and 218, respectively). This region is important for the interaction of these proteins with Gβγ subunits. Phosphorylation of phosducin by GRK2 markedly reduces its Gβγ binding ability, suggesting that GRK2 may modulate the activity of the phosducin protein family by disrupting this interaction. The identification of phosducin and PhLP as new substrates for GRK2 further expands the cellular roles of this kinase and suggests new mechanisms for modulating GPCR signal transduction.

From the ‡Departamento de Biología Molecular and Centro de Biología Molecular Severo Ochoa, Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, E-28049 Madrid, Spain and the §Institut fü r Pharmakologie und Toxikologie der Universitä t Wü rzburg, Versbacher Strasse 9, 97078 Wü rzburg, Germany G protein-coupled receptor kinase 2 (GRK2) is able to phosphorylate a variety of agonist-occupied G proteincoupled receptors (GPCR) and plays an important role in GPCR modulation. However, recent studies suggest additional cellular functions for GRK2. Phosducin and phosducin-like protein (PhLP) are cytosolic proteins that bind G␤␥ subunits and act as regulators of G-protein signaling. In this report, we identify phosducin and PhLP as novel GRK2 substrates. The phosphorylation of purified phosducin and PhLP by recombinant GRK2 proceeds rapidly and stoichiometrically (0.82 ؎ 0.1 and 0.83 ؎ 0.09 mol of P i /mol of protein, respectively). The phosphorylation reactions exhibit apparent K m values in the range of 40 -100 nM, strongly suggesting that both proteins could be endogenous targets for GRK2 activity. Our data show that the site of phosducin phosphorylation by GRK2 is different and independent from that previously reported for the cAMP-dependent protein kinase. Analysis of GRK2 phosphorylation of a variety of deletion mutants of phosducin and PhLP indicates that the critical region for GRK2 phosphorylation is localized in the C-terminal domain of both phosducin and PhLP (between residues 204 and 245 and 195 and 218, respectively). This region is important for the interaction of these proteins with G␤␥ subunits. Phosphorylation of phosducin by GRK2 markedly reduces its G␤␥ binding ability, suggesting that GRK2 may modulate the activity of the phosducin protein family by disrupting this interaction. The identification of phosducin and PhLP as new substrates for GRK2 further expands the cellular roles of this kinase and suggests new mechanisms for modulating GPCR signal transduction.
Agonist occupancy of G protein-coupled receptors (GPCR) 1 triggers interaction with different types of cellular proteins. Interaction with heterotrimeric G proteins promotes GTP/GDP exchange and dissociation into G␣ and G␤␥ subunits, both of which modulate different membrane-bound effector proteins. This signal transduction process is regulated at different levels. Agonist-activated GPCRs are phosphorylated by a family of specific G protein-coupled receptor kinases (GRKs). This is followed by binding of regulatory proteins termed arrestins to the phosphorylated receptor leading to uncoupling from G proteins, a process known as desensitization (reviewed in Refs. [1][2][3]. GRK2 is a ubiquitous member of the GRK family, which is able to phosphorylate a variety of GPCR (4,5). Recent data indicate that GRK2 and ␤-arrestins play additional roles in GPCR regulation and signaling. These proteins also initiate receptor internalization by facilitating the interaction of the receptor complex with the endocytic machinery. This is followed by receptor dephosphorylation and recycling (reviewed in Refs. 6, 7). Moreover, recent data suggest that GRK2 and ␤-arrestin may directly participate in GPCR-mediated activation of the mitogen-activated protein kinase (MAPK) cascade by allowing the recruitment of signaling proteins such as Srctyrosine kinase (8 -10), further stressing their relevance in GPCR signaling.
GRK2 cellular levels, activity, and subcellular localization appear to be tightly controlled through different mechanisms (1,5,11), consistent with the idea that this kinase plays different important roles in cellular signaling. Agonist-induced translocation of GRK2 to the plasma membrane involves its interaction with receptor domains and binding of the C-terminal domain of GRK2 to both free G␤␥ subunits and phosphatidylinositol-4,5-bisphosphate (12,13). This binding process results in an enhanced activity of GRK2 toward its substrates (14,15). The functionality of GRK2 is also modulated by protein kinase C or Src-mediated phosphorylation (16,17,18) or by means of its interaction with caveolin (19) or calmodulin (20). In addition, an as yet unidentified anchoring protein present in microsomal membranes also modulates GRK2 localization and activity (21,22). The occurrence of different cellular pools of GRK2 suggests the existence of additional substrates and cellular functions for this kinase.
Other regulatory mechanisms of GPCR signaling involve direct modulation of G-protein activities. Interaction of regulators of G-protein signaling family members with activated G␣ subunits increases its GTPase activity and modulates termination of the signal (23). On the other hand, phosducin and phosducin-like proteins (PhLP) are able to preferentially bind to G␤␥ subunits, thus interfering with G␤␥-mediated signaling pathways and with re-association with G␣⅐GDP (24 -30). Phosducin is a ϳ33-kDa soluble phosphoprotein that was first identified in the retina (24) and later shown to be ubiquitously expressed (25,31). The homologous PhLP also shows a broad pattern of expression (28,29). Both proteins have been reported to compete for G␤␥ binding with other targets such as adenylyl cyclases or phospholipase C␤2 resulting in modulation of GPCR signaling (25,29,30,(32)(33)(34). In the case of phosducin, its ability to interact with G␤␥ dimers is attenuated upon phosphorylation by the cAMP-dependent protein kinase (PKA), thus providing a mechanism for regulating its function (35,36), whereas regulation of PhLP remains unclear because PKA does not phosphorylate PhLP (37).
Despite this data, very little is known about how these regulatory mechanisms at the receptor and G-protein levels are integrated during cellular responses to agonists. Because both GRK2 and phosducins are able to interact with G␤␥ subunits, it is tempting to suggest that they may influence one another. In fact, phosducins have been shown to inhibit G␤␥-mediated GRK2 activation and translocation (29,38). In this report, we show that phosducin and PhLP are high-affinity substrates for GRK2 and that such phosphorylation results in inhibition of the G␤␥ binding ability. Together with recent data that identify tubulin as a substrate for GRK2 (39,40), our results corroborate the existence of non-GPCR substrates for this kinase and suggest a key role for GRK2 in the modulation of GPCR signaling at different levels.
Phosphorylation Assays-GRK2 phosphorylation of phosducin-His 6 , GST⅐PhLP, and the truncated constructs was performed in 50 l of 20 mM Tris-HCl, pH 7.5, 5 mM MgCl 2 , 1 mM EDTA, 1.4 mM EGTA, 25-60 M ATP (2-3 cpm/fmol) (phosphorylation buffer), and the desired concentrations of GRK2, phosducin, or PhLP as indicated in the figure legends. In some experiments, 1 mM dithiothreitol, 10% glycerol, 0.8 M heparin (Sigma) or 100 units of the peptide PKA inhibitor PKI (Sigma) were included in the assay buffer. After incubation for the desired time at 30°C, the reaction was stopped by adding 25 l of SDS-sample buffer and heating for 5 min at 95°C. Samples were resolved by 10 -12.5% SDS-PAGE except for the C-terminal truncated constructs of phosducin that were resolved by using a 16.5% Tricine-SDS gel. The phosphorylated proteins were detected by autoradiography. Cerenkov counting of the excised bands was used to quantitate the phosphorylation and determine stoichiometry.
To compare the capacity of GRK2, GRK3, and GRK5 to phosphoryl- Ϫ , 1 mM GDP) with 20 nM GRK2 for 15 min at 37°C as described previously (22). The phosphorylation assay was initiated by adding 2 M rhodopsin and phosphorylation buffer as detailed under "Experimental Procedures" followed by incubation for 20 min at 30°C in the presence or absence of 600 nM purified phosducin. Phosphorylated proteins were resolved by 10% SDS-PAGE and analyzed by autoradiography. B and C, GRK2 phosphorylates phosducin. Purified phosducin-His 6 (100 nM) was incubated in phosphorylation buffer as detailed under "Experimental Procedures" with 25 nM purified GRK2 for 60 min at 30°C in the presence or absence of 50 units of PKI or of the GRK2 inhibitor heparin (C, 0.8 M) as indicated. After stopping the reaction, proteins were resolved by SDS-PAGE and analyzed by autoradiography. The first lane in B shows GRK2 autophosphorylation. D, phosducin is phosphorylated by GRK2 and PKA at different and independent sites. Phosducin-His 6 (100 nM) was preincubated in 25 l of phosphorylation buffer with unlabeled ATP either with 25 nM GRK2 for 60 min (left) or with 5 units of PKA and 5 M cAMP for 10 min (right). Control phosducin aliquots were preincubated under the same conditions without any kinase present. The pre-phosphorylated phosducin was used as a substrate for a subsequent phosphorylation assay with the other kinase. Reactions were initiated by adding 25 l of phosphorylation buffer including [␥-32 P]ATP to a final specific activity of 4 -6 cpm/fmol, and the indicated additions of PKA or GRK2 to the final concentrations that are detailed above. Proteins were resolved and analyzed as in previous panels. E, phosphorylation of phosducin and the S73A phosducin mutant by PKA and GRK2. His 6 -tagged wild-type phosducin and a S73A mutant lacking the PKA phosphorylation site (100 nM) were incubated as described in other panels with PKA (5-10 units) or purified recombinant GRK2 (100 nM) as indicated. Proteins were resolved and analyzed as detailed above. A representative autoradiogram of three experiments with similar results is shown. F, additive phosphorylation of phosducin by PKA and GRK2. His 6 -tagged phosducin (100 nM) was incubated as detailed above in the presence of the indicated kinases, and phosphorylation was analyzed by SDS-PAGE and autoradiography. ate phosducin, each GRK was dialyzed in salt-free phosphorylation buffer for 12 h just before usage to exclude the potential interference of different buffer conditions on phosducin phosphorylation. Aliquots of the GRK dialysates were used in a rhodopsin phosphorylation assay as described previously (38) to assess its activity, and equally effective amounts were then used in the phosducin phosphorylation assays.
For PKA-mediated phosphorylation assays, 25-100 nM of purified phosducin-His 6 was incubated for 20 min at 30°C in the same phosphorylation buffer used with GRK2 in the presence of 5-10 units of PKA (Sigma) and 5 M cAMP (Sigma) in a final volume of 50 l in the presence or absence of 80 units of PKI. Phosphorylation assays were stopped, resolved, and analyzed as detailed above. In some experiments, phosducin was subjected to phosphorylation by either PKA or GRK2 in the absence of radiolabeled ATP prior to its incubation with the other kinase in the presence of [␥-32 P]ATP, to investigate whether these were independent processes (see legend to Fig. 1).
Two-dimensional Gel Electrophoresis-To analyze the number of Ser/Thr residues in phosducin and PhLP that were phosphorylated by GRK2, 25 nM phosducin was incubated with 25 nM GRK2 for 30 min at 30°C in phosphorylation buffer. The resultant phosphoprotein was resolved by two-dimensional gel electrophoresis as described previously (44). The isoelectric focusing gel was made using a wide range of ampholytes (Bio-Rad) by mixing 5-7:3-10 ampholytes in a ratio of 40:10 in 8M urea, 4% acrylamide/bisacrylamide gel. The second dimension was resolved by 12% SDS-PAGE, and the phosphoproteins were detected by autoradiography.
Effect of Phosducin Phosphorylation on G␤␥-Phosducin Interaction-After phosphorylation of purified His 6 -tagged phosducin (200 nM) by recombinant GRK2 (2 M) in 200 l of 20 mM Tris, pH 7.5, 6 mM MgCl 2 , 50 M ATP, 1 mM EDTA, 1 mM dithiothreitol, and 10% glycerol for 30 min at 30°C, most of the GRK2 was removed by incubation with 10 l of SP-Sepharose for 10 min on ice followed by centrifugation. The resultant supernatant showed an Ӎ100-fold reduction in GRK2 as determined by immunoblot analysis (data not shown). Non-phosphorylated phosducin controls were treated exactly in the same way, except that ATP was omitted during the incubation. G␤␥ binding was performed with 25 l of the supernatant containing phosducin, phosphophosducin, or background controls in 100 l of 20 mM Hepes, pH 7.9, 100 mM NaCl, 50 mM NaF, 30 M heparin, 10 mM 2-mercaptoethanol, 0.1% Lubrol, 1 mg/ml dialysed E. coli lysate, and different concentrations of purified G␤␥ subunits for 10 min at 4°C. 10 l of Ni 2ϩ -NTAagarose (Qiagen) were then added for an additional 10 min at 4°C to bind phosducin-His 6 and the associated G␤␥ subunits as described (45). The resin was then pelleted by centrifugation and washed twice with 500 l of 20 mM Hepes, pH 7.9, 100 mM NaCl, 0.1 mM EDTA, 20 mM imidazole, 0.1% Lubrol, and 10 mM 2-mercaptoethanol. The final pellets were then resuspended in SDS-sample buffer followed by 12% SDS-PAGE after heating at 95°C for 5 min. G␤ was identified by Western blotting with specific antibodies (Signal Transduction Laboratories). Peroxidase-coupled secondary antibodies and enhanced chemiluminescence reagents (ECL, Amersham Pharmacia Biotech) were used to develop the blots. Scanning with densitometric analysis was used to quantify the bands.

RESULTS
Identification of Phosducin as a GRK2 Substrate-We have previously described that GRK2 binds to intracellular membranes under basal conditions by means of an as yet unidentified anchoring protein and that this interaction inhibits its phosphorylation activity (21,22). Several lines of evidence suggest that membrane-bound GRK2 activity could be enhanced by stimulation of endogenous heterotrimeric G proteins with aluminum fluoride (22). Because phosducin has been described as an inhibitor of G␤␥-stimulated GRK2 activity toward agonist-activated receptors by competing for binding with free G␤␥ subunits (31,34), purified bovine phosducin was included in our assays to test the participation of endogenous G␤␥ subunits in the observed stimulation of GRK2 activity. As expected, phosducin (600 nM) did inhibit light-dependent phosphorylation of rhodopsin (2 M) by GRK2 (Fig. 1A). However, it was noted that a band of ϳ33 kDa was heavily phosphorylated under these conditions (Fig. 1A, lane 2). This band co-migrated with purified phosducin and was recognized by antibodies raised against purified recombinant phosducin (Ref. 31, data not shown). To test the possibility that GRK2 was the kinase directly responsible for the observed phosducin phosphorylation, we performed additional experiments using recombinantpurified GRK2 and hexahistidine-tagged phosducin. As shown in Fig. 1B, phosducin was markedly phosphorylated in the presence of GRK2. Additional experiments were carried out to confirm that this phosphorylation was a result of GRK2 activity and could not be attributed to contaminating PKA, which might be present at very low concentrations in the purified GRK2 preparation. The addition of high concentrations of the PKA inhibitor PKI during the phosphorylation assay had no effect on the phosphorylation of phosducin by recombinant GRK2 (Fig. 1B). Phosphorylation of phosducin by purified PKA was completely inhibited by PKI in similar experimental conditions (data not shown). Moreover, phosducin phosphorylation by GRK2 was completely abolished in the presence of heparin (Fig. 1C), a potent inhibitor of GRK2 (46). Overall, these results indicated that phosducin was an "in vitro" substrate for GRK2.
Because both GRK2 and PKA were able to phosphorylate phosducin, we next investigated whether the prior phosphorylation by one of these kinases had any effect on the subsequent phosphorylation by the other kinase. To answer this question, 100 nM phosducin was incubated under phosphorylating conditions in the absence of [␥-32 P]ATP with recombinant GRK2 (Fig. 1D, left) or purified PKA (Fig. 1D, right), for periods of time (60 and 10 min, respectively) known to allow the completion of these reactions (data not shown). Control phosducin aliquots were incubated in the same buffer conditions in the absence of any kinase. Next, the pre-phosphorylated phosducin was tested as a substrate for the other kinase by including purified PKA or GRK2 in the reaction together with [␥-32 P]ATP. As shown in Fig. 1D, prior phosducin phosphorylation by GRK2 had no effect on PKA activity toward phosducin (left), nor did phosphorylation by PKA significantly alter the subsequent phosphorylation by GRK2 (right). The fact that the observed phosphorylation is not due to remaining activity of the first kinase was confirmed by the absence of radioactive labeling observed upon addition of [␥-32 P]ATP when no additional kinase was present. To further demonstrate that GRK2 and PKA phosphorylate phosducin at different sites, we tested the ability of GRK2 to phosphorylate a phosducin mutant (S73A) that lacks the phosphorylation site for PKA (35). Fig. 1E shows that wild-type phosducin and the S73A mutant were phosphorylated to a similar extent by GRK2 whereas PKA phosphorylation was completely abolished in the mutant. Moreover, simultaneous incubation with PKA and GRK2 led to an additive incorporation of phosphate (Fig. 1F). Overall, these data indicate that phosducin phosphorylation by both kinases is not mutually exclusive and takes place at different residues, suggesting that PKA and GRK2 may be independent modulators.
Characterization of Phosducin Phosphorylation by GRK2- Fig. 2A shows that phosducin phosphorylation by GRK2 proceeds rapidly (t1 ⁄2 ϭ 15 min), stoichiometrically (0.82 Ϯ 0.1 mol of P i /mol of phosducin, when both phosducin and GRK2 are present at 25 nM during the phosphorylation assay), and is dependent on phosducin concentration (Fig. 2B). Double-reciprocal plot analysis (Fig. 2C) reveals that GRK2 displays an apparent K m for phosducin of 46 Ϯ 7 nM, with a V max of 0.72 Ϯ 0.04 nmol P i /min/mg of protein. The observed affinity is well below the physiological range (ϳ1 M) of phosducin concentrations (31) and similar or higher than that reported for other GRK2 substrates (39,40,47). Finally, in agreement with the stoichiometric data, two-dimensional gel electrophoresis analysis of GRK2-phosphorylated phosducin indicates that most of Purified GRK2, GRK3, and GRK5 preparations were dialyzed and pre-tested for rhodopsin phosphorylation activity to normalize GRK activity as detailed under "Experimental Procedures." Kinase aliquots were then incubated with phosducin-His 6 (100 nM) in 50 l of phosphorylation buffer as described in Fig. 1, in the presence or absence of the GRK inhibitor heparin as indicated. Proteins were resolved by SDS-PAGE and analyzed by autoradiography. An autoradiogram representative of three experiments is shown. this protein incorporates a single phosphate in the presence of GRK2 (Fig. 2D). Overall, these data indicate that phosducin is a high-affinity GRK2 substrate.
PhLP Is a Substrate for GRK2, and Other GRKs Are Able to Phosphorylate Phosducin-PhLP has 65% amino acid homology to phosducin (28) and displays similar functions as modulator of G␤␥ signaling and as inhibitor of G␤␥-mediated stimulation of GRK2 activity toward GPCR (29,30). Unlike phosducin, however, regulation of PhLP activity by phosphorylation is unclear (37). Therefore, we tested the ability of GRK2 to phosphorylate purified GST⅐PhLP. As observed with phosducin, GRK2 clearly caused PhLP phosphorylation, and the process was completely abolished in the presence of heparin (data not shown). Time-course analysis (Fig. 3A) revealed that GRK2 phosphorylates PhLP very rapidly (t1 ⁄2 ϭ 5 min), reaching a stoichiometry of 0.83 Ϯ 0.09 mol of P i /mol of GST⅐PhLP. Double-reciprocal plot analysis (Fig. 3B) revealed kinetic parameters (K m of 103 Ϯ 4 nM; V max of 1.12 Ϯ 0.15 nmol/min/mg protein) in the range observed for phosducin. GRK2 also phosphorylates PhLP preferentially in one site, as indicated by bidimensional-gel electrophoresis analysis (Fig. 3C).
We then explored whether phosducin was a specific substrate for GRK2 or whether it could also be phosphorylated by other members of the GRK family. To address this issue, aliquots of GRK2, GRK3, and GRK5 that were equally effective in phosphorylating rhodopsin (data not shown) were tested for their activity toward purified phosducin. Fig. 4 shows that all tested GRKs were able to phosphorylate phosducin although with slightly different activities.
Localization of Phosphorylation Sites and Functional Conse-quences of Phosducin Phosphorylation by GRK2-We next attempted to localize the region of phosducin and PhLP that contains the site(s) of phosphorylation by GRK2. For this purpose, we tested the ability of GRK2 to phosphorylate a variety of purified phosducin and PhLP constructs in which different domains of these proteins had been deleted. The N-terminal truncated phosducin construct phosducin-(64 -245) was efficiently phosphorylated by GRK2 (Fig. 5A, panel I). The phosducin-(138 -245) deletion construct was clearly phosphorylated by GRK2 (Fig. 5A, panel I), although it displayed a ϳ2.5-fold increase in the phosphorylation K m with respect to full-length phosducin (data not shown). In contrast, phosducin-(1-204) was poorly phosphorylated by GRK2 even in the presence of a higher GRK2 concentration (Fig. 5A, panel II). Overall, our results indicate that the C-terminal region of phosducin (residues 205-245) contains the main determinants required for GRK2 phosphorylation (the phosphorylation site itself and/or docking regions critical for the action of the kinase). Similar experiments with a variety of GST⅐PhLP constructs (Fig. 5B) also showed that the region essential for phosphorylation by GRK2 resides in the C-terminal region of the protein. FIG. 6. Binding of GRK2-phosphorylated phosducin to G␤␥ subunits. A, purified phosducin-His 6 was incubated with recombinant GRK2 in the absence (-) or presence (ϩ) of ATP as detailed under "Experimental Procedures." GRK2 was removed by SP-Sepharose and centrifugation, and phosducin was then incubated with the indicated concentrations of purified brain G␤␥ subunits as described under "Experimental Procedures." The phosducin-His 6 complexes were separated with Ni 2ϩ -NTA-agarose columns, and binding of G␤␥ subunits was assessed by immunoblot analysis of G␤. On the right is a control for the nonspecific binding of G␤␥ subunits to the beads in the absence of phosducin (with or without GRK2). An immunoblot representative of three independent experiments is shown. B, binding of purified G␤␥ subunits to phosducin (open bars) and GRK2-phosphorylated phosducin (filled bars) was performed as in A and quantified by densitometric scanning of the G␤-specific bands in the immunoblots. Data are mean Ϯ S.D. of three independent experiments. *, p Ͻ 0.05 when compared with control.
The C-terminal region of both phosducin and PhLP have been shown to play a critical role in their interaction with G␤␥ subunits (26,27,30,42). The fact that these regions are also critical for GRK2 phosphorylation suggests that the activity of this kinase may modulate the ability of phosducin to bind G␤␥ dimers. To test this hypothesis, we investigated the effect of preincubating phosducin with GRK2 under phosphorylating or non-phosphorylating conditions on the subsequent interaction of histidine-tagged phosducin with purified brain G␤␥ subunits. After removal of GKR2 with SP-Sepharose, the ability of unphosphorylated (no ATP) or GRK2-phosphorylated (ϩATP) phosducin to interact with two different concentrations of G␤␥ subunits was assessed (Fig. 6A). A marked reduction in G␤␥ binding to phosphorylated phosducin (ϳ50%, Fig. 6B) was noted at 50 nM G␤␥, whereas such a decrease was not apparent at high G␤␥ concentrations (1000 nM), thus suggesting that GRK2 phosphorylation decreases the affinity of the G␤␥-phosducin interaction. Whereas G␤␥ subunits did not bind to Ni 2ϩ -NTA-agarose on its own, GRK2 and its G␤␥ binding ability were still detectable in the background control (Fig. 6A). However, the signals were much weaker than those observed in the presence of phosducin and were visible only at high concentrations of G␤␥ (1000 nM).

DISCUSSION
In this report, we have identified phosducin and PhLP as new substrates for GRK2. In line with recent reports that show that tubulin is phosphorylated by this kinase (39,40), our data further indicate the existence of soluble substrates for GRK2. The emerging evidence that the activity of this kinase is not restricted to agonist-occupied GPCR suggests that GRK2 may also act as an effector of GPCR signaling and have additional cellular roles.
Several lines of evidence support the hypothesis that GRK2mediated phosphorylation of phosducin proteins may have physiological relevance. The phosphorylation reactions proceed rapidly and stoichiometrically. The apparent K m (40 nM for phosducin and 103 nM for PhLP) displays far higher affinity (10,000-fold) than that reported for peptide substrates of GRK2 (with K m in the range of 1. . Taken together, our results indicate that phosducin and PhLP are high-affinity substrates for GRK2. Though the interaction of GRK2 with agonist-occupied receptors and G␤␥ subunits promotes an allosteric activation of the kinase (14,15), phosducin phosphorylation by GRK2 is not altered in the presence of bleached rhodopsin (data not shown), suggesting that it is not the activated receptor itself that promotes phosducin phosphorylation. Future experiments will be directed to investigate the mechanisms of this phosphorylation process in living cells and its potential modulation by GPCR-mediated signaling pathways.
Our results indicate that the critical determinants for GRK2 phosphorylation are located within the C-terminal domain of both phosducin and PhLP (between residues 205 and 245 and 196 and 218, respectively), thus suggesting a similar regulatory mechanism in which these regions contain the phosphorylation sites and/or required docking sites for the action of the kinase. These regions contain some serine and threonine resi-dues in the vicinity of negative-charged amino acids, consistent with previous reports showing that GRK2 preferentially phosphorylates substrates containing acidic residues N-terminal to the phosphorylation site (48). These C-terminal domains are critically involved in the interaction of phosducin and PhLP with its main physiological target, the G␤␥ subunits of heterotrimeric G proteins. The crystal structure of phosducin complexed with the ␤␥ subunits of transducin (26) suggests that phosducin binds with its C-terminal domain to the side of the ␤-propeller, whereas its less well defined N terminus is stretched out on the face of the propeller, covering sites where G␤␥ interacts with G␣. In fact, a small peptide fragment within the C terminus of phosducin (amino acids 215-232) appears to be sufficient for high-affinity G␤␥ binding and is able to disrupt G␤␥-mediated functions (27). Moreover, point mutations in this region markedly reduce functional activity of phosducin (27). Based on sequence homology, a similar model of PhLP-G␤␥ interaction has been predicted (26,30), and several types of functional studies also indicate that the C-terminal part of PhLP plays a key role in G␤␥ binding (30,42).
Through interaction with free G␤␥ subunits, phosducin and PhLP have been shown to interfere with a variety of GPCR signaling pathways triggered by different G proteins (25,29,31,32,42). The reported modulation of phosducin function by PKA phosphorylation provides a clear feedback regulatory mechanism in the case of receptor signaling leading to changes in cellular cAMP concentrations. GRK2-mediated phosphorylation may represent an additional mechanism for the modulation of phosducin interaction with G␤␥ subunits released upon receptor-mediated G i or G o activation. GRK2-mediated regulation of phosducins further differs from that by PKA in three important ways. (a) The phosphorylation of phosducin by PKA and GRK2 is independent and takes place at different sites. (b) GRK2 has more subtle effects on phosducin function than PKA. (c) Although the effect of PKA is restricted to phosducin, GRK2 also phosphorylates PhLP. These data suggest independent control of phosducin functions by the two different types of kinases.