Stimulus-dependent phosphorylation of G-protein-coupled receptors by casein kinase 1alpha.

We have previously demonstrated that the phospholipase C-coupled m3-muscarinic receptor is phosphorylated in an agonist-sensitive manner by a protein kinase of approximately 40 kDa purified from porcine cerebellum (Tobin, A. B., Keys, B., and Nahorski, S. R. (1996) J. Biol Chem. 271, 3907-3916). This kinase, called muscarinic receptor kinase (MRK), is distinct from second messenger-regulated protein kinases and from beta-adrenergic receptor kinase and other members of the G-protein-coupled receptor kinase family. In the present study we propose that MRK is casein kinase 1alpha (CK1alpha) based on the following evidence: 1) the amino acid sequence from two proteolytic peptide fragments derived from purified MRK corresponded exactly to sequences within CK1alpha. 2) Casein kinase activity co-eluted with MRK activity from the final two chromatography steps in the purification of porcine brain MRK. 3) Recombinant CK1alpha expressed in Sf9 cells is able to phosphorylate both casein and the bacterial fusion protein, Ex-m3, that contains a portion of the third intracellular loop of the m3-muscarinic receptor downstream of glutathione S-transferase. 4) Partially purified CK1alpha increased the level of muscarinic receptor phosphorylation in an agonist-sensitive manner when reconstituted with membranes from Chinese hamster ovary-m3 cells expressing the human recombinant m3-muscarinic receptor. 5) Partially-purified CK1alpha phosphorylated rhodopsin, contained in urea-treated bovine rod outer segment membranes, and the extent of phosphorylation was increased in the presence of light. These data demonstrate that the kinase previously called MRK is CK1alpha, and that CK1alpha offers an alternative protein kinase pathway from that of the G-protein-coupled receptor kinase family for the stimulus-dependent phosphorylation of the m3-muscarinic receptor, rhodopsin, and possibly other G-protein-coupled receptors.

Intensive research over the last decade have revealed that many GPCR 1 subtypes are phosphorylated in response to ago-nist stimulation (1). These receptors include those coupled to either the adenylate cyclase or phospholipase C (PLC) pathways and suggests that receptor phosphorylation is a common regulatory mechanism employed by all but a few GPCR's (1). For the majority of these receptors the cellular protein kinases involved in agonist-mediated receptor phosphorylation have yet to be determined. However, this is not the case for the extensively studied ␤-adrenergic receptor where agonist-dependent phosphorylation and receptor desensitization is mediated by the receptor-specific kinase, ␤-adrenergic receptor kinase (␤-ARK) (2,3).
Studies using purified or partially purified receptor preparations reconstituted in phospholipid vesicles with purified ␤-ARK, have demonstrated that ␤-ARK is also able to phosphorylate both cyclase-coupled (e.g. m2-muscarinic (4)) and PLCcoupled (e.g. substance P receptor (5)) receptors in an agonistdependent manner. Furthermore, the use of dominant negative mutants of ␤-ARK to inhibit endogenous ␤-ARK activity (6) has suggested that ␤-ARK is the endogenous kinase responsible for the phosphorylation of recombinant PLC-coupled ␣ 1B -adrenergic receptors expressed in COS-7 cells and rat-1 fibroblasts (7), angiotensin II receptors in HEK 293 cells (8), and the cyclasecoupled ␦-opioid receptors in HEK 293 cells (9). These studies have indicated that ␤-ARK may have a broad receptor substrate specificity that extends beyond ␤-adrenergic receptors.
However, recent evidence suggests that receptor phosphorylation mediated by ␤-ARK and the GRKs may be inhibited by the PLC signaling pathway. For example, increased intracellular calcium concentrations and depletion of the phospholipid, phosphoinositide 4,5,-bisphosphate (PIP 2 ), may contribute to a reduction in the activity of GRKs (see "Discussion"). This raises the possibility that agonist-dependent phosphorylation of PLCcoupled receptors may be mediated by an alternative protein kinase pathway from that of ␤-ARK and the other GRKs.
Our early studies on the PLC-coupled m3-muscarinic receptor indicated that the rapid serine phosphorylation observed * This research is supported in part by Wellcome Trust Grants 047600/96 and 16895/96. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
following agonist stimulation (15) was mediated by a receptor kinase that was distinct from ␤-ARK (15,16). We have recently purified a 40-kDa protein kinase from porcine cerebellum that is able to phosphorylate a glutathione S-transferase bacterial fusion protein containing a portion of the third intracellular loop of the m3-muscarinic receptor (17). Furthermore, this 40-kDa protein kinase was able to enhance the agonist-dependent phosphorylation of the muscarinic receptor present in membranes obtained from CHO-m3 cells transfected with the human m3-muscarinic receptor cDNA (17). The molecular weight, chromatographic properties, and protein kinase inhibitor studies demonstrated that the 40-kDa protein kinase was distinct from the second messenger-regulated protein kinases (e.g. protein kinase C) and from ␤-ARK and other members of the GRK family (17). These findings indicated that the 40-kDa protein kinase, called muscarinic receptor kinase (MRK), represents a previously unidentified receptor-specific kinase that offers an additional/alternative protein kinase pathway for the phosphorylation of the m3-muscarinic receptor and possibly other GPCRs.
In the present paper we present evidence that MRK is a member of the casein kinase 1 family, namely casein kinase 1␣ (CK1␣), and reveal the ability of the recombinant kinase to phosphorylate m3-muscarinic receptors and rhodopsin in an agonist/stimulus-dependent manner.

EXPERIMENTAL PROCEDURES
Cell Culture-CHO (Chinese hamster ovary) cell cultures stably transfected with human m3-muscarinic receptor cDNA (CHO-m3 cells, a kind gift from Dr. N. J. Buckley, Dept. Pharmacology, University College, London, UK) contained ϳ2100 fmol of receptor/mg of protein.
Purification of Muscarinic Receptor Kinase (MRK) from Porcine Brain-The procedure used for purification of MRK from porcine cerebellum has been previously described (17).
Preparation of the Bacterial Fusion Protein Ex-m3-Preparation of the bacterial fusion protein, Ex-m3, where amino acids Ser 345 -Leu 463 of the human m3-muscarinic receptor third intracellular loop are fused with glutathione S-transferase has previously been described (15).
Amino Acid Sequencing-The excised Coomassie-stained protein band corresponding to MRK was digested with trypsin. Peptides were extracted for 2 h in a sonicating water bath. After concentration, the peptides were resolved using a Relasil C18 column with a guard precolumn packed with AX-300 on a Michrom HPLC system. Peptides were sequenced at the low picomole level using a modified ABI 477a sequencer employing fast cycle chemistry (18).
Preparation of Recombinant Baculovirus-The coding sequence for bovine CK1␣ (a kind gift from Dr. Melanie Cobb, Department of Pharmacology, University of Texas Southwestern Medical Center, Dallas, TX) was amplified using the polymerase chain reaction (PCR) using primers that inserted a Kozak (GCC ACC) sequence upstream of an epitope, FLAG-TAG (MDYKDDDDK), at the N terminus of CK1␣ (5Јprimer, CCCAAGCTTGCCACCATGGACTACAAGGACGACGATGAC-AAGATGGCGAGCAGCAGCGGC; 3Ј-primer, CCCGGATCCTTAGAA-ACCTGTGGGGGTTTGGGC). The resulting polymerase chain reaction product was cloned into the XbaI-BamHI sites of pVL1392 (Pharmingen). Recombinant baculovirus was generated using the BaculoGold system (Pharmingen).
The coding sequence for the m3-muscarinic receptor was cloned into the BamHI-EcoRI sites in pVL1393. Recombinant baculovirus was then generated using the BaculoGold system. This baculovirus was used in control infections. (Note that controls were also run using non-infected Sf9 cells with similar results to cells infected with control virus. In this paper experiments with control infected cells are reported.) Purification of CK1␣ from Infected Sf9 Cells-Sf9 cells (1 liter at 1 ϫ 10 6 cells/ml) were infected with baculovirus containing the CK1␣ or control baculovirus (1-4 plaque-forming units/cell). After 4 days the cells were harvested and resuspended in 45 ml of TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 7.4) containing protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 10 g/ml soybean trypsin inhibitor, 1 g/ml leupeptin, 1 g/ml pepstatin A, 100 g/ml benzamine, 100 g/ml iodoacetamide). After 10 min on ice the cells were homogenized. The homogenate was centrifuged at 20,000 ϫ g for 15 min and a high speed supernatant fraction (S200 fraction) obtained by further centrifugation at 200,000 ϫ g for 35 min. The sample was applied to a 1-ml Resource S (Pharmacia) column which was eluted using a linear gradient of 0 -1.0 M NaCl over 20 bed volumes (flow rate ϭ 2 ml/min). 1-ml fractions were collected. The kinase activity eluted as a single peak at ϳ0.3 M NaCl. These fractions were combined and passed through a 1-ml heparin-Sepharose column equilibrated with 0.32 M NaCl. The column was eluted using a linear gradient of 0.32-1.75 M NaCl over 15 ml. The kinase activity eluted as a single peak at ϳ0.94 M NaCl. The peak activity was dialyzed against 0.3 M NaCl in TE buffer and stored at 4°C. The kinase in this preparation was ϳ4 ng/l (0.1 pmol/l), and represents ϳ3.5% of the total protein in the fraction.
Control purification protocol involved identical purification steps from Sf9 cells infected with the m3-muscarinic receptor baculovirus. The total protein in fractions obtained from the heparin purification of the control and CK1␣-infected cells were very similar, 100 g/ml total protein in the control, and 112 g/ml from the CK1␣-infected cells.
Assay for Muscarinic Receptor Kinase and Casein Kinase Activity-Assay for the muscarinic receptor kinase involved using the muscarinic receptor fusion protein Ex-m3 as a substrate for the kinase (17).
Where ␣-casein was used, the reaction was stopped by the addition of 100% trichloroacetic acid (11 l). The precipitated proteins were pelleted by centrifugation in a Microfuge for 10 min at 13,000 ϫ g. The protein pellet was washed with acetone (Ϫ20°C) and resuspended in 20 l of 2 ϫ SDS-PAGE sample buffer.
The proteins from the Ex-m3 or casein assay were resolved by 12% SDS-PAGE. To ensure the equal recovery of the protein substrates and to confirm their relative positions, gels were stained with Coomassie Blue. Gels were then dried and autoradiographs obtained and/or bands corresponding to the peptide substrates were excised and counted.
Phosphorylation of m3-muscarinic Receptors in Membrane Preparations from CHO-m3 Cells-Membrane phosphorylations were carried out as described previously (17). Briefly, crude CHO-m3 cell membranes were prepared and resuspended in kinase buffer at 1 mg of protein/ml. 50 l of membranes (ϳ0.1 pmol of receptor) were used in a phosphorylation reaction mixture that contained final concentrations of 20 mM Tris-HCl, pH 7.4, 10 mM MgCl 2 , 1 mM EGTA, 100 M [␥-32 P]ATP (1-4 cpm/fmol ATP), Ϯ 1 mM carbachol and Ϯ20 M atropine. To this reaction mixture 10 l of partially purified CK1␣ (0.5-1 pmol) or control extract was added. Total volume was 100 l. Reactions were started by the addition of ATP and continued at 32°C for 10 min. Reactions were stopped by centrifugation at 13,000 ϫ g for 30 s. The supernatant was removed by aspiration and membranes solubilized with 1 ml of solubilization buffer (10 mM Tris-HCl, pH 7.4, 10 mM EDTA, 500 mM NaCl, 1% Nonidet P-40, 0.1% SDS, 0.5% deoxycholate) for 30 min on ice. m3-Muscarinic receptors were then immunoprecipitated with a specific antiserum (332) as described previously (15).
Phosphorylation of Rod Outer Segment Membranes-Urea-treated bovine rod outer segment membranes (30 pmol of rhodopsin/reaction; a kind gift from Dr. Martin Lohse, Institute of Pharmacology and Toxicology, University of Wurzburg, Versbacher Strasse 9, D-97078 Wurzburg, Germany) were added to kinase buffer containing 50 M ATP (1 cpm/fmol). To this partially purified CK1␣ (0.3 pmol), purified bovine ␤-ARK (0.3 pmol; a gift from Dr. Martin Lohse), or control extract were added. Total volume was 30 l. The above reagents were combined at 4°C (where necessary under a safe light). Reactions were started by placing the tubes in a water bath at 32°C either under a safe light or in room fluorescent light for a given time period. Reactions were stopped by addition of 10 l of 2 ϫ SDS-PAGE sample buffer. Proteins were then resolved on a 12% gel. Gels were stained with Coomassie Blue, dried, and autoradiographs obtained. Bands corresponding to rhodopsin were excised and counted.
Miscellaneous Procedures-Determination of relative intensities of phosphorylated bands was carried out using a Bio-Rad GS 670 densitometer. Western blotting of whole cell extracts from Sf9 cells used a CK1␣-mediated Phosphorylation of G-protein-coupled Receptors commercially available antiserum (FLAG-M2, Kodak) that recognizes the FLAG epitope cloned onto the N terminus of CK1␣.

RESULTS
Amino Acid Sequence Analysis-Proteolytic fragments of MRK purified from porcine brain were isolated by high performance liquid chromatography. The amino acid sequences derived from two peptides were determine to be: 1, WYGQEK; and 2, IEYVHTK. These sequences matched exactly to sequences found exclusively within CK1␣ (peptide 1, Trp 78 -Lys 83 ; and peptide 2, Ile 124 -Lys 130 ). Furthermore, the predicted molecular mass of CK1␣ is 37.5 kDa, which corresponds closely with the ϳ40-kDa mass suggested for MRK (17).
Co-purification of MRK Activity and Casein Kinase Activity-To determine if the kinase we previously defined as MRK was CK1␣, samples from fractions obtained from the purification of MRK were assayed for both MRK activity, which is defined by the ability of fractions to phosphorylate the bacterial fusion protein Ex-m3 (glutathione S-transferase:m3-third intracellular loop; see "Experimental Procedures"), and casein kinase activity. Fractions from the final two column purification steps, namely the Resource S and heparin-Sepharose fractionation (see Ref. 17), were analyzed. The casein kinase activity was found to elute from the Resource S (data not shown) and heparin-Sepharose columns in an identical manner to MRK activity (Fig. 1). The peak of MRK and casein kinase activity co-eluted from the Resource S and heparin-Sepharose columns at ϳ0.37 and ϳ0.87 M NaCl, respectively.
Expression of Recombinant CK1␣ in Sf9 Cells-Attempts to express recombinant bovine CK1␣ in mammalian cells (COS-7, HEK 294, and CHO cells) have not proven successful. The reason for this is unclear and is currently under investigation. To obtain a source of recombinant CK1␣, we turned to the insect cell baculovirus expression system.
Recombinant CK1␣ was expressed in infected Sf9 cells as determined by Western blotting of whole cells using an antiserum against the FLAG-TAG epitope engineered onto the N terminus of CK1␣ ( Fig. 2A). The activity of the recombinant kinase was confirmed by an increase in casein kinase activity in cytosolic extracts from CK1␣-infected Sf9 cells compared with the control cell infection (Fig. 2B). Correlating with an increase in casein kinase activity was an increase in the ability of the cell extract obtained from the CK1␣-infected cells to phosphorylate the m3-muscarinic receptor fusion protein, Ex-m3 (Fig. 2B).
To confirm that the recombinant CK1␣ contained in infected FIG. 1. Co-purification of MRK activity and casein kinase activity. Fractions from the heparin-Sepharose chromatography column step in the purification of porcine brain MRK were tested for MRK activity (i.e. the ability to phosphorylate the m3-muscarinic receptor fusion protein, Ex-m3) and casein kinase activity. The top panel shows an autoradiograph of fractions tested for phosphorylation of the fusion protein Ex-m3. The bottom panel shows the same fractions used to phosphorylate casein (the data are expressed as counts/10 min incorporated into casein).

FIG. 2. Identification of recombinant CK1␣ expression and kinase activity from infected Sf9 cells.
A, Western blot of whole cell extract (30 g of protein/lane) from cells infected with control baculovirus or CK1␣ baculovirus. The antiserum used was a commercially available monoclonal antiserum against a FLAG epitope engineered on the N terminus of CK1␣. B, high speed supernatant extracts from Sf9 cells infected with control virus or CK1␣ virus were tested for the ability to phosphorylate the m3-muscarinic receptor fusion protein Ex-m3 (3.5 g) and casein (15 g). The relative positions of casein and Ex-m3 as determined by Coomassie Blue staining are indicated. C, high speed supernatant extracts from Sf9 cells infected with control virus or CK1␣ virus were fractionated on a Resource S column and fractions tested for the ability to phosphorylate Ex-m3 and casein.

CK1␣-mediated Phosphorylation of G-protein-coupled Receptors
cell extracts was able to phosphorylate both casein and Ex-m3, the cell extracts were fractionated on a Resource S column. The peak of casein kinase activity contained in CK1␣-infected cells co-eluted with the peak of MRK activity (Fig. 2C). The peak of recombinant kinase activity eluted at ϳ0.3 M NaCl. Note, the presence of endogenous casein kinase activity is evident in the control fractionation (Fig. 2C). MRK activity and casein kinase activity also co-eluted during heparin-Sepharose chromatography (data not shown).

Phosphorylation of m3-Muscarinic Receptors and Rhodopsin by Recombinant CK1␣-Previous
studies have demonstrated that membranes from CHO-m3 cells contained an endogenous protein kinase able to phosphorylate the m3muscarinic receptor in an agonist-dependent manner (17). In these earlier studies, addition of cerebellum-derived MRK to the CHO-m3 membrane preparation resulted in an enhancement of agonist-sensitive m3-muscarinic receptor phosphorylation (17). A parallel experiment was conducted here using partially purified extracts from infected Sf9 cells as the source of exogenous kinase. Fig. 3 shows that in the presence of control extract, purified on the Resource S column, the m3-muscarinic receptor (ϳ0.1 pmol of receptor/reaction) contained in CHO-m3 membranes undergoes agonist-sensitive phosphorylation which can be inhibited by the muscarinic antagonist atropine. Addition of Resource S purified extract from Sf9 cells infected with CK1␣ baculovirus (0.5-1 pmol of CK1␣/reaction) resulted in an ϳ2.3-fold increase in agonistsensitive m3-muscarinic receptor phosphorylation with no significant change in the basal phosphorylation of the receptor (Fig. 3). Furthermore, the increase in agonist-sensitive muscarinic receptor phosphorylation, mediated by CK1␣, was inhibited by the antagonist atropine (Fig. 3).
To further test the ability of CK1␣ to phosphorylate GPCR's, rhodopsin, contained in urea-treated rod outer segments, was used as substrate for CK1␣. Heparin-Sepharose-purified CK1␣ (ϳ300 fmol) was incubated with rod outer segment membranes (containing rhodopsin at 30 pmol/reaction) either under a safe light or under room fluorescent lights. Rhodopsin phosphorylation was observed only in the extract obtained from cells infected with CK1␣ baculovirus and not from control extracts (Fig. 4A). The CK1␣ extract did phosphorylate rhodopsin under dark conditions and this phosphorylation was enhanced in the presence of light (118 fmol of phosphate incorporated/5-min reaction) (Fig. 4A). In comparison, ␤-ARK (ϳ300 fmol/reaction) did not phosphorylate rhodopsin in the absence of light but in the presence of light rhodopsin was phosphorylated to a similar extent as that seen for light-mediated CK1␣ phosphorylation (122 fmol of phosphate incorporated/5-min reaction) (Fig. 4A). The time course for CK1␣-mediated phosphorylation of rhodopsin was found to be similar to the time course for ␤-ARKmediated phosphorylation (Fig. 4B).

DISCUSSION
The present study has revealed that the 40-kDa protein kinase purified previously from porcine cerebellum, called MRK due to its ability to phosphorylate the m3-muscarinic receptor in an agonist-dependent manner (17), can be identified as CK1␣. These data, together with the ability of recombinant CK1␣ to phosphorylate the m3-muscarinic receptor and rhodopsin in a stimulus-dependent manner, indicate that CK1␣ offers an alternative protein kinase pathway, from that of ␤-ARK and the other GRKs, for the agonist-sensitive phosphorylation and the potential regulation of GPCR's.
Due to the presence of endogenous receptor kinase activity in membranes from CHO-m3 cells, the enhancement of muscarinic receptor phosphorylation observed in the presence of CK1␣ may be due to an activation of the endogenous kinase rather than a direct phosphorylation of the receptor. This, however, appears unlikely since CK1␣ is also able to phosphorylate rhodopsin in a stimulus-dependent manner, in ureatreated rod outer segment membranes where there is no endogenous kinase activity. These data also suggest that like ␤-ARK, CK1␣ acts predominantly on the activated form of the receptor. This conclusion is supported by dose-response analysis of m3-muscarinic receptor phosphorylation in intact CHO-m3 cells which have demonstrated a close correlation between receptor occupancy and receptor phosphorylation (19). Furthermore, the ability of CK1␣ to mediate light-dependent phosphorylation of rhodopsin suggests that the substrate specificity of CK1␣ is broader than just the m3-muscarinic receptor. Experiments are presently in progress to test the receptorsubstrate specificity of CK1␣ against a range of cyclase-and PLC-coupled receptors.
Casein kinase activity was one of the first protein kinase activities identified in mammalian cells and is attributed to two enzymes called casein kinase I and casein kinase II (20,21). The true "biological" casein kinase responsible for phosphorylating casein in mammary glands is a transmembrane protein kinase of the Golgi apparatus that bears no relationship to casein kinase I or II. In this regard casein kinase I and II are misnomers since the biological substrates for these kinases is not casein. The casein kinase I gene family consists of at least six members; CK1␣, CK1␤ (22), CK1␥ 1-3 (23), and CK1␦ (24). A number of in vitro substrates for this kinase family have been identified including: glycogen synthase, SV40 large T antigen (20), p53 tumor suppressor protein (25), and DARPP-32 (26). However, it is not clear which are the biologically relevant substrates. The present study is the first to demonstrate that CK1␣ is able to phosphorylate GPCR's in a stimulus-dependent manner.
The broad receptor substrate specificity of ␤-ARK (5, 27) and other members of the GRK family (11)(12)(13)(14), as determined in reconstituted systems, has implicated this protein kinase family in the phosphorylation of PLC-coupled receptors. Furthermore, expression of the dominant negative mutant of ␤-ARK inhibits agonist-sensitive phosphorylation of PLC-coupled ␣ 1Badrenergic receptors (7) and type 1A angiotensin II receptors (8). However, recent studies have indicated that the intracellular environment following activation of PLC-coupled receptors would not favor the membrane translocation of ␤-ARK and other GRK's. It is now clear that ␤-ARK activity in reconstituted systems is completely dependent on the presence of the phospholipid, PIP 2 (28). Furthermore, translocation of ␤-ARK to the plasma membrane is dependent on the synergistic action of G-protein ␤␥-subunits and PIP 2 at a site within the pleckstrin homology domain at the C terminus of ␤-ARK (28,29). PIP 2 is also involved in the anchoring of GRK-4, GRK-5, and GRK-6 to the plasma membrane by interacting with sites at the C terminus (30). Therefore, receptors that regulate the level of PIP 2 could have a profound influence on the activation and translocation of ␤-ARK and other members of the GRK family. Stimulation of the m3-muscarinic receptor in CHO-m3 cells 3 and in the human neuroblastoma cell line SH-SY5Y cells results in a 80% fall in the levels of PIP 2 within the first 10 s of agonist stimulation and this is sustained for 10 min in the presence of agonist (31,32). Similar rapid falls in the level of PIP 2 have been reported to occur following stimulation of m1muscarinic (33,34), vasopressin (35), and thrombin receptors (36,37). PLC-coupled receptor activation may, therefore, discourage membrane translocation/association of the GRK's by depleting PIP 2 .
In addition, recent studies indicate that calmodulin, in a calcium-dependent manner, can inhibit the association of ␤-ARK and GRK-5 with the plasma membrane (38,39). ␤-ARK translocation is inhibited by the ability of calcium/calmodulin to compete for the binding of ␤␥-subunits, and GRK-5 association with the membrane is disrupted by a direct interaction with the kinase (38). Therefore, PLC-coupled receptors, by virtue of their ability to mobilize intracellular calcium stores and mediate calcium entry across the plasma membrane, could again discourage GRK activity.
Overall, it could be anticipated that the intracellular environment following signal transduction via PLC-coupled receptors would not be conducive for GRK translocation and receptor phosphorylation. We would suggest, on the basis of the present study, that CK1␣ offers an attractive alternative route for the phosphorylation of PLC-coupled receptors. In this regard it is interesting to note that casein kinase I isolated from erythrocytes is inhibited by PIP 2 (40). Therefore, in contrast to the GRK's, it is possible that PLC-coupled receptors may increase the activity of CK1␣ by the hydrolysis of membrane PIP 2 . Other regulatory features may also be important such as autophosphorylation which is reported to occur on both tyrosine and serine residues on CK1␣ (41) classifying this kinase as a dual kinase. We have identified autophosphorylation of MRK purified from porcine brain (17) and are presently in the process of determining if autophosphorylation plays any regulatory role in CK1␣-mediated receptor phosphorylation. Studies are also underway to determine the mechanism of membrane association and the receptor substrate specificity of CK1␣.