Compositions and methods for modulating the activity of G protein-coupled receptor kinases GRK5 and GRK6

Abstract

G protein-coupled receptor kinases (GRK) play an important role in phosphorylating and regulating the activity of G protein-coupled receptors. Complementary DNAs (cDNAs)that encode two novel members of the G protein-coupled receptor kinase (GRK) family are provided in the present invention. These cDNAs encode GRK5 (590 amino acids) and GRK6 (576 amino acids) which represent two new members of the GRK family that have distinct tissue distribution and substrate specificity. The availability of the cDNAs enables the generation of reagents to modulate the activity of endogenous kinases. These include dominant negative mutations and antisense oligonucleotides or stably transfected antisense constructs to block expression of the kinase to generate a cell with a reduced ability to desensitize to various agents. Expression of GRK5 and GRK6 also permits identification of specific inhibitors and activators of these two kinases. Such inhibitors and activators may be used therapeutically to either directly modulate the activity of a given receptor or by augmenting the ability of a given therapeutic agent to stimulate a given receptor.

Description

[0001] This application is a divisional of U.S. patent application Ser. No. 08/464,954 filed Jun. 6, 1995 which is a continuation-in-part of U.S. patent application Ser. No. 08/076,084 which was filed Jun. 11, 1993.

BACKGROUND OF THE INVENTION

[0003] G protein-coupled receptors represent a diverse family of cell surface proteins that transduce the binding of extracellular ligands (hormones, neurotransmitters, odorants, light, etc.) into intracellular signalling events. G protein-coupled receptors modulate the activity of a wide variety of effector molecules including adenylyl cyclase, cGMP phosphodiesterase, phospholipase C, phospholipase A2, and K, Na+ and Ca++ channels. It has become apparent in recent years that G protein-coupled receptor kinases play a vital role in regulating receptor function by their unique ability to specifically phosphorylate activated forms of G protein-coupled receptors.

[0004] Two of the best characterized G protein-coupled receptors are the hormone responsive &bgr;2-adrenergic receptor (&bgr;2AR), which mediates catecholamine stimulation of adenylyl cyclase, and the visual “light receptor” rhodopsin” which mediates phototransduction in retinal rod cells. The &bgr;2AR and rhodopsin share many structural and functional similarities including a conserved protein topology (e.g., seven transmembrane domains) as well as an ability to specifically interact with G proteins upon activation. The similarities between these receptors also extend to mechanisms involved in receptor regulation. In both of these systems, a rapid loss of receptor, responsiveness occurs following activation. This rapid activation-dependent loss of responsiveness or desensitization is promoted by phosphorylation of the receptor. This phosphorylation is mediated by protein kinases that have the unique ability to recognize and phosphorylate their receptor substrates only when they are in their active conformations, i.e., when they have been stimulated and/or occupied by appropriate agonist ligands. The &bgr;-adrenergic receptor kinase (&bgr;ARK) and rhodopsin kinase have been identified as kinases involved in the agonist specific phosphorylation of the &bgr;2AR and rhodopsin, respectively. The subsequent uncoupling of the receptor and G protein is then mediated by arrest in proteins that specifically bind to the phosphorylated and activated form of the receptor. Additional lines of evidence suggest that other G protein-coupled receptors may also be regulated by similar mechanisms. These receptors include the m2 muscarinic cholinergic and &agr;2-adrenergic receptors, which inhibit adenylyl cyclase; the &agr; mating factor receptor of the yeast Saccharomyces cerevisiae and the chemotactic cAMP receptor of the slime mold Dictyostelium discoideum.

[0005] Several lines of evidence suggest that &bgr;ARK may have a broad substrate specificity and thus serve as a general G protein-coupled receptor kinase. First, direct phosphorylation studies have demonstrated that &bgr;ARK not only phosphorylates the &bgr;2AR to a high stoichiometry, but also can phosphorylate the &agr;2-adrenergic, the m2 muscarinic cholinergic and the substance P receptors. In addition, several agents appear to promote an increase in membrane-associated &bgr;ARK activity, These include &bgr;-agonists, prostaglandin E1, somatostatin, and platelet activating factor, suggesting the involvement of &bgr;ARK in the regulation of these receptors. Moreover, a growing body of evidence suggests that &bgr;ARK and rhodopsin kinase are members of a multigene family.

[0006] Structural information on the G protein-coupled receptor kinase (GRK) family was initially provided by the isolation of a cDNA encoding bovine &bgr;ARK [Benovic, J. L., DeBlasi, A., Stone, W. C., Caron, M. G., and Lefkowitz, R. J., 1989 Science, 246:235-246]. &bgr;ARK is a protein of 689 amino acids (79.6 kD) containing a central protein kinase catalytic domain flanked by large amino and carboxyl terminal domains. Additional members of the GRK family were subsequently cloned including bovine &bgr;ARK2 [Benovic, J. L., Onorato, J. J., Ariza, J. L., Stone, W. C., Lohse, M., Jenkins, N. A., Gilbert, D. J., Copeland, N. G., Caron, M. G. and Lefkowitz, R. J., 1991 J. Biol. Chem., 266:14939-14946], bovine rhodopsin kinase [Lorenz, W., Inglese, J., Palczewski, K., Onorato, J. J., Caron, M. G. and Lefkowitz, R. J., 1991 Proc. Natl. Acad. Sci. USA, 88:8715-87191, Drosophila kinases GPRK-1 and GPRK-2 [Cassill, J. A., Whitney, M., Joazeiro C. A. P., Becker, A., and Zucker, C. S., 1991 Proc. Natl. Acad. Sci. USA, 88:11067-11070], and the recently identified human IT11 [Ambrose, C., James, M., Barnes, G., Lin, C., Bates, G., Altherr, M., Duyao, M., Groot, N., Church, D., Wasmuth, J. J., Lehrach, H., Housman, D., Buckler, A., Gusella, J. F., and MacDonald, M. E., 1992 Human Mol. Genet., 1:697-703]. &bgr;ARK2 and Drosophila GPRK-1 appear to be the most similar to &bgr;ARK with amino acid identities of 84% and 64%, respectively. In contrast, rhodopsin kinase, IT11, and Drosophila GPRK-2 have significantly lower homology with &bgr;ARK (35-40% amino acid identity). Common features of these kinases include a centrally localized catalytic domain of approximately 240 amino acids which shares significant amino acid identity (46 to 95%), an N-terminal domain of 161-197 amino acids (except for GPRK-2), and a variable length C-terminal domain of 100-263 amino acids.

[0007] The tremendous diversity in the G protein-coupled receptor family suggested that there may well be other GRK members. This has, in fact, been found to be the case. In the present invention, two additional members of the GRK gene family have been identified, GRK5 and GRK6; these cDNAs have now been cloned, expressed and characterized.

SUMMARY OF THE INVENTION

[0008] G protein-coupled receptor kinases (GRK) play an important role in phosphorylating and regulating the activity of G protein-coupled receptors. Complementary DNAs (cDNAs) that encode two novel members of the G protein-coupled receptor kinase family are provided in the present invention. These Cdnas (SEQ ID NO: 1 and (SEQ ID NO: 2, respectively) encode GRK5 (590 amino acids (SEQ ID NO: 5) and GRK6 (576 amino acids) (SEQ ID NO: 6) which represent two new members the GRK family that have distinct tissue distribution and substrate specificity. These kinases have been overexpressed in Sf9 insect cells, purified and characterized. Using the purified kinases, it has been demonstrated that heparin is a potent inhibitor of GRK5 (IC50˜1 nM) and a good inhibitor of GRK6(IC50˜15 nM). These results indicate that specific inhibitors of these two kinases can be identified.

[0009] The GRK5 and GRK6 cDNAS can be expressed, purified and further characterized with regard to their substrate specificity. The availability of the cDNAs will also enable the generation of reagents to block the activity of the endogenous kinases. These include dominant negative mutations which involves mutating a critical lysine residue in the kinase to arginine. This results in a protein kinas6 which can still bind to its receptor substrate but can no longer transfer a phosphate group to the receptor. Thus, the mutant kinase can block the ability of the endogenous kinase to phosphorylate a given receptor and thereby block desensitization. An alternative strategy for blocking the activity of the endogenous kinase involves the use of antisense oligonucleotides or stably transfected antisense constructs to block expression of the kinase. This will generate a cell with a reduced ability to desensitize to various agents. Expression of GRK5 and GRK6 will also allow screening for specific inhibitors of these two kinases. It has been demonstrated that heparin is a potent inhibitor of GRK5 and a good inhibitor of GRK6. Inhibitors are identified and then utilized to specifically block the activity of GRK5 or GRK6, and thus inhibit receptor desensitization. Specific inhibitors of these kinases may be used therapeutically to either directly modulate the activity of a given receptor (by blocking endogenous desensitization of that receptor) or by augmenting the ability of a given therapeutic agent to stimulate a given receptor (again by blocking desensitization). Alternatively, a specific activator of GRK5 or GRK6 could be used to stimulate the desensitization or turn-off of an overactive (e.g., oncogenic) receptor.

DESCRIPTION OF THE DRAWINGS

[0010] FIG. 1 is a restriction map and sequencing strategy for the GRK5 clones. The restriction sites on the human heart pGRK5 clones are indicated for the enzymes NacI(N), PstI(P), XhoI(X), DraII(D), KpnI(K) and SmaI (S). Clone PGRK5 has an open reading frame of 1770 bp starting from the first in frame ATG. DNA sequencing was performed by the dideoxynucleotide chain termination technique. The extent and direction of the sequence reactions are shown by the arrows. The open reading frame is indicated as an open rectangle while untranslated sequences are indicated as a line.

[0011] FIG. 2 is a nucleotide and deduced amino acid sequence of human GRK5. The nucleotide sequence of the full length clone pGRK5 is shown. The nucleotides are numbered on the right hand side beginning at the 5′ end and ending at the 3′ end of the clone pGRK5. The predicted amino acid sequence is numbered on the right hand side just below the nucleotide sequence, beginning at the first in frame ATG (bp 221) and ending just before the first in frame stop codon (bp 1990). The predicted initiator methionine is flanked by the nucleotide sequence TCAATGG. IUPAC single letter abbreviations for amino acids are used.

[0012] FIG. 3 is a comparison of the amino acid sequence of GRK5 with related GRKs by dendogram analysis. The PILEUP program in the Wisconsin Genetics Computer Group software was used to align and compare the amino acid sequences of human GRK5 and IT11, bovine &bgr;ARK, &bgr;ARK2 and rhodopsin kinase and Drosophila GPRK-1 and GPRK-2. The overall amino acid identities of GRK5 with the other GRKs are 69%. with human IT11, 71% with Drosophila GPRK-2, 47%. with bovine rhodopsin kinase, 36%. with bovine &bgr;ARK and &bgr;ARK2, and 35%. with Drosophila GPRK-1. The PILEUP program uses a progressive pairwise alignment.

[0013] FIG. 4 is a Northern blot analysis of human mRNA from various tissues. Two &mgr;g of poly A+ RNA from human heart, brain, placenta, lung, liver, skeletal muscle, kidney and pancreas on a Multiple Tissue Northern blot was hybridized with a 1157 bp PstI ORF fragment from the clone pGRK5. The blot was washed in 0.1×SSC, 0.10% SDS at 60° C. followed by autoradiography at −80° C. for 24 hours.

[0014] FIG. 5 shows expression of GRK5 in Sf9 Insect Cells. . GRK5- and &bgr;ARK-containing recombinant baculoviruses were used to infect a monolayer of Sf9 cells. Following a 48 hour infection, the cells were harvested, lysed in 1.5 ml of buffer and then centrifuged at 40,000×g for 20 minutes. The supernatants were then used to phosphorylate urea treated ROS either in the presence or absence of room light. The reactions were stopped by the addition of 50 &mgr;l of SDS sample buffer and the samples were then electrophoresed on a 10%. SDS polyacrylamide gel. The gel was dried and autoradiographed for 30 minutes at room temperature and the rhodopsin bands were then cut and counted. The level of light-dependent rhodopsin phosphorylation in this experiment was 7 pmol/min/mg using the supernatant from wild type Sf9 cells, 1120 pmol/min/mg by Sf9 cells expressing GRK5, and 2470 pmol/min/mg by Sf9 cells expressing &bgr;ARK. B. The Sf9 cell supernatants were also used to phosphorylate urea treated ROS in the presence or absence of G protein &bgr;&ggr; subunits. The reactions were stopped, electrophoresed and autoradiographed for 120 minutes at room temperature.

[0015] FIG. 6 is a restriction map and sequencing strategy for the GRK6 clones. The restriction sites on seven of the human heart and brain CDNA clones are shown for SacI, PstI and NcoI. DNA sequencing was performed by the dideoxynucleotide chain termination method. The extent and direction of the sequence reactions are shown by the arrows. The pGRK6-hh1, -hh2 and -hb8 clones were sequenced in their entirety while the other clones were restriction mapped and partially sequenced. Sequence divergence in the GRK6-hh1 and -hh2 clones compared with the other clones is denoted by an asterisk (*) and thin line.

[0016] FIG. 7 is a nucleotide and deduced amino acid sequence of human GRK6. The composite nucleotide sequence from several of the GRK6 clones is shown. The nucleotides are numbered on the right hand side beginning with the 5′ end of clone pGRK6-hh2 and ending at the 3′ end of clone pGRK6-hb13. The amino acid sequence of GRK6 is numbered just below the nucleotide sequence and begins with the first ATG in the nucleotide sequence (bp 63) and ends just before the first in frame stop codon (bp 1790). The GXGXXGX˜16K found in all protein kinases is shown in bold type as is the sequence DLG, which is highly conserved in the GRK family. IUPAC single letter abbreviations amino acids are used.

[0017] FIG. 8 is a comparison of the amino acid sequence of human GRK6 with related GRKs by dendogram analysis. The PILEUP program in the Wisconsin Genetics Computer Group (GCG) software was used to align and compare the amino acid sequences of human GRK6, GRK5, IT11, bovine &bgr;ARK, &bgr;ARK2, rhodopsin kinase and Drosophila GPRK-1 and GPRK-2. The overall amino acid identities of GRK6 with the other kinases are 70.1% (human GRK5), 68.5% (human IT11), 66.7% (Drosophila GPRK-2), 47.1% (bovine rhodopsin kinase), 37.4% (bovine &bgr;ARK), 37.2% (bovine &bgr;ARK2), and 36.5% (Drosophila GPRK-1). PILEUP uses a progressive pairwise alignment.

[0018] FIG. 9 shows the alignment of the amino acid sequences of GRK6 with GRK5, IT11 and GPRK-2. The PILEUP program in the GCG software was used to align and compare the amino acid sequences of human GRK6, GRK5, IT11 and Drosophila GPRK-2. Only the amino acids that are different from GRK6 are shown. Identical residues are indicated as a dash (−) while gaps produced by the alignment are shown as a period (.). The amino acid sequence is numbered on the right. Overall, GRK6, GRK5, IT11 and GPRK-2 are 576, 590, 500 and 426 amino acids in length, respectively.

[0019] FIG. 10 is the expression of GRK6 in Sf9 Insect Cells. A. GRK6 and &bgr;ARK-containing baculoviruses were prepared. Sf9 cells, either uninfected (lane 1) or infected with BacPAK6 (lane 2) or GRK6-(lane 3) or &bgr;ARK-(lane 4) containing viruses, were harvested following a 48 hour infection. The cells were lysed, centrifuged and 15 &mgr;g of supernatant protein was then electrophoresed on a 10% polyacrylamide gel which was then stained with Coomassie blue. B. The supernatants (5 &mgr;g total protein) were also used to phosphorylate urea treated ROS membranes (80 pmol rhodopsin) in the presence (+) or absence (−) of light. Incubations (40 &mgr;l total volume) were at 30° C. for 10 minutes and were stopped by the addition of 50 &mgr;l of SDS sample buffer. Following polyacrylamide gel electrophoresis, the gels were dried and autoradiographed for 3 hours. From three separate experiments, the GRK6 virus infected supernatant phosphorylated bleached rhodopsin with a specific activity of 81 pmol/min/mg protein (when the BacPAK6 infected cell supernatant phosphorylation of rhodopsin was subtracted). This activity was 4.7% +/−0.4% that of the &bgr;ARK virus infected supernatant.

[0020] FIG. 11 shows Northern blot analysis of human mRNA from various tissues. A. 2 &mgr;g of poly A+ RNA from human heart, brain, placenta, lung, liver, skeletal muscle, kidney and pancreas, was used for RNA analysis. The blot was hybridized with a randomly primed 1780 bp EcoRI fragment from the clone pGRK6-hh2. The blot was washed in 0.1×SSC at 60° C. before autoradiography at −80° C. for 3 days. B. The blot was stripped and then reprobed with a ˜1556 bp EcoRI/NcoI ORF fragment from the human &bgr;ARK cDNA. The blot was again washed in 0.1×SSC at 60° C. before autoradiography at −80° C. for 3 days.

DETAILED DESCRIPTION OF THE INVENTION

[0021] To identify additional members of the GRK family, degenerate oligonucleotide primers were designed to encode several highly conserved amino acid stretches found in all of the GRKs. These oligonucleotides were then used as primers in a polymerase chain reaction (PCR) using template DNA prepared from a human heart cDNA library.

[0022] GRK5

[0023] One set of primers yielded a PCR product of the expected size (˜450 bp) as assessed by electrophoresis on a 2% agarose gel. The 450 bp DNA product was then restriction digested with EcoRI and HindIII (sites present in the two PCR primers), subcloned into EcoRI/HindIII digested pbluescript KS and sequenced using T3 and T7 primers. Of the 20 clones sequenced, 13 were identical and had 54% amino acid identity with bovine &bgr;ARK, 64% identity with bovine rhodopsin kinase and 83% identity with Drosophila GPRK-2. This PCR product also encoded the amino acid sequence Asp Leu Gly (DLG), a sequence that is highly conserved in the GRK family. The other 7 clones sequenced had no appreciable homology with the GRK family.

[0024] The GRK-related PCR product was then labeled with [32P] dCTP by random priming and used as a probe to screen the human heart cDNA library. Six different cDNA clones were isolated by this method. The relative lengths, partial restriction maps and sequencing strategy for these clones are shown in FIG. 1. Of the six clones isolated, pGRK5-1 and pGRK5-3 encode an ˜1.5 Kb ORF fragment while pGRK5-6 and pGRK5-9 encode ˜1.1 Kb of the ORF. The clone pGRK5-8 is ˜2.2 Kb in length and appears to encode the 3′ end of the ORF as well as the entire 3′ untranslated sequence. The full length clone pGRK5 is 2.55 Kb in length and contains an open reading frame of 1770 bp which is flanked by 220 bp of 5′ untranslated and 567 bp of 3′ untranslated sequence. The 5′ untranslated sequence is relatively GC rich (73%) while the 3′ untranslated sequence ends with a poly A tail. While pGRK5 does not have a good Kozak consensus; sequence for translation initiation (TCAATGG instead of ACCATGG), the comparison of GRK5 with the other known GRKs helped to determine the initiator methionine. The predicted open reading frame, beginning at the first in frame ATG and ending at the first in frame stop codon (SEQ ID NO: 1) encodes a protein of 590 amino acids (FIG. 2) comprising SEQ ID NO: 5 (GenBank accession number L15388).

[0025] The predicted molecular weight of GRK5 is 67.6 kDa, with a predicted pI of 8.75. GRK5 contains a centrally located protein kinase catalytic domain of 238 amino acids flanked by N-terminal and C-terminal regions of 193 and 159 amino acids, respectively. Comparison of the amino acid sequence of GRK5 with other known GRKs reveals a high degree of overall homology, particularly in the catalytic domain. Construction of a phylogenetic tree of all the known GRKs demonstrates that GRK5 is most closely related to human IT11 and that GRK5, IT11, GPRK-2 and rhodopsin kinase form a distinct branch of the GRK family as compared to &bgr;ARK, &bgr;ARK2 and GPRK-1 (FIG. 3). GRK5 has an overall 69% amino acid identity and 82% similarity with human IT11. The major differences between these proteins include a 33 amino acid gap in IT11 near the N-terminus, a divergent stretch in the central portion of the N-terminus (only 19% identity and 47% similarity from amino acids 97-149), and a significant. difference in the length of the C-terminus (100 and 159 amino acids for IT11 and GRK5, respectively). GRK5 also has high homology with Drosophila GPRK-2, with 71% amino acid identity and 82% similarity. However, GPRK-2 has a very short N-terminal domain (28 amino acids) compared to all other GRKs (161-197 amino acids). GRK5 also has higher homology to rhodopsin kinase (68% amino acid similarity) than to &bgr;ARK (58% similarity), both in amino acid identity and size. While the functions of the N-terminal and C-terminal domains of these kinases are largely unknown, the carboxyl terminal tail of rhodopsin kinase is farnesylated, a post-translational modification that appears to be important for its translocation to the disc membrane. By comparison, the C-terminal domain of &bgr;ARK appears to interact with G protein &bgr;&ggr; subunits. Overall, GRK5, IT11, rhodopsin kinase and GPRK-2 have significantly shorter C-terminal tails as compared to &bgr;ARK, &bgr;ARK2 and GPRK-1.

[0026] The tissue distribution of GRK5 was analyzed by Northern hybridization using a human multiple tissue Northern blot containing 2 &mgr;g of poly A+ RNA from eight different human tissues. A randomly primed 1157 bp PstI ORF fragment from PGRK5 was labeled with [32P] dCTP and used as a probe. This revealed a message of 3.0 Kb with highest levels in the human heart, placenta, lung>skeletal muscle>brain, liver, pancreas>kidney (FIG. 4). This tissue distribution is in contrast to &bgr;ARK, which has a message of 3.8 Kb, with highest levels in the brain, skeletal muscle>pancreas>heart, lung, placenta, kidney>liver. Previous studies have demonstrated that IT11 has highest message levels in testis but is also found at low levels in a wide variety of tissues. Drosophila GPRK-2 also appears to be expressed in a variety of tissues. In contrast, rhodopsin kinase is more specifically localized in the rod outer segments of the retina, with low levels also observed in the pineal gland. The rather unique tissue distribution of GRK5, as compared to the other GRKs, suggests that GRK5 likely has a unique substrate specificity and physiological role within the cell.

[0027] To initially characterize the activity of GRKS, the ORF from pGRK5 was excised by restriction digestion with NaeI and SmaI and then subcloned into the baculovirus expression vector pBacPAK1. Viral DNA and the pBacPAK-GRK5 construct were then cotransfected into Sf9 cells to obtain an isolated recombinant virus. The recombinant virus was amplified and then used to infect a monolayer of Sf9 cells. As shown in FIG. 5A, GRK5 expressed in Sf9 cells phosphorylates rhodopsin in a light dependent manner. The ability of GRK5 to phosphorylate rhodopsin was also compared to &bgr;ARK (FIG. 5A, right). Overall, while both GRK5 and &bgr;ARK phosphorylated rhodopsin well above the basal “&bgr;ARK-like” activity observed in Sf9 cells, GRK5 was less active than &bgr;ARK at phosphorylating rhodopsin. When the supernatant fractions from the GRK5 and &bgr;ARK infected cells were analyzed by polyacrylamide gel electrophoresis and Coomassie Blue staining, comparable levels of GRK5 and &bgr;ARK were observed. This suggests that the difference in activity is not due to a difference in the expression levels of the two kinases. Purified GRK5 exhibited several properties similar to many other protein kinases including a preferential requirement for Mg2+ as the divalent cation, a pH optimum of 5.5 to 7.5, and a Km for ATP of 24 &mgr;M.

[0028] The two most extensively studied GRKs, rhodopsin kinase and &bgr;ARK, appear to be regulated by different mechanisms. Farnesylation of rhodopsin kinase at its C-terminus appears to be important for the translocation of rhodopsin kinase to the disc membrane. &bgr;ARK, on the other hand, may associate with membranes via its specific interaction with G protein &bgr;&ggr; subunits. Since GRK5 does not contain a consensus sequence for isoprenylation (CAAX) at its C-terminus, the ability of G protein &bgr;&ggr; subunits to modulate the activity of GRK5 was tested. It was found that &bgr;ARK is significantly activated in the presence of G protein &bgr;&ggr; subunits (FIG. 5B, right). In contrast, the ability of GRK5 to phosphorylate rhodopsin is not affected by G protein &bgr;&ggr; subunits (FIG. 5B, left). In addition, unlike &bgr;ARK, GRK5 phosphorylation appears to be biphasic suggesting that there may be a primary site or sites that are rapidly phosphorylated by GRK5 followed by slower phosphorylation of additional sites. Thus, GRK5 appears to utilize a novel mechanism of regulation and/or cellular localization as compared to &bgr;ARK and rhodopsin kinase.

[0029] GRK5 has been found to have a complex substrate specificity that is not defined simply by the presence of absence of acidic or basic residues in the vicinity of a serine or threonine. Previous peptide phosphorylation studies revealed that &bgr;ARK preferentially phosphorylates Ser/Thr residues in an acidic environment, while rhodopsin kinase phosphorylates peptides with acidic amino acids carboxyl-terminal to the Ser/Thr. A number of peptides tested in these experiments were also found to be substrates for GRK5. However, in contrast to &bgr;ARK and rhodopsin kinase, GRK5 was found to prefer peptides that do not contain acidic amino acids in the vicinity of a serine. The kinetic parameters for the peptides tested indicate that the presence of acidic residues predominantly decreases the rate of the phosphorylation rather than the affinity of GRK5 for these peptides. In contrast, results obtained using several general protein kinase substrates have shown that the acidic proteins casein and phosvitin are weakly phosphorylated by GRK5 while the basic histones were not phosphorylated by GRK5.

[0030] The modulation of GRK5 activity by polyanions also revealed several differences as compared to &bgr;ARK and rhodopsin kinase. Although heparin and dextran sulfate have been identified as the most potent inhibitors of &bgr;ARK with IC50s of 150-2800 nm, these compounds are >150-fold more potent at inhibiting GRK5 (IC50 of ˜1 nM and ˜0.6 nM, respectively). Heparin is also a potent inhibitor of casein kinase II (IC50 of ˜20 to 60 nM); however, it is only a weak inhibitor of rhodopsin kinase (IC50 of ˜200 &mgr;M). Several other polyanions such as polyaspartic and polyglutamic acid have also been found to be much more potent inhibitors (16-80 fold) of GRK5 as compared to &bgr;ARK. This tremendous variance in the ability of polyanions to inhibit different members of the GRK family indicates that specific inhibitors of these kinases can be found. Two other compounds were also tested for their ability to inhibit GRK5. H-7, a potent inhibitor of protein kinase C and the Camp- and cGMP-dependent protein kinases, was found to be a weak inhibitor of GRK5 (IC50 of ˜170 &mgr;M). Similarly, NaCl also appears to inhibit the activity of GRK5 with an IC50 of ˜60 mM, similar to the inhibition of &bgr;ARK by NaCl.

[0031] Since polyanions were potent inhibitors of GRK5, the effects of polycations on GRK5 activity were also studied. While polycations such as spermine, spermidine and polylysine are weak inhibitors of &bgr;ARK, they are activators of rhodopsin kinase and casein kinase. Previous studies have shown that polyamines like spermine and spermidine activate casein kinase II 2-3 fold and may serve as physiological modulators of this kinase in reticulocytes. GRK5 is also significantly activated by the polycations polylysine, spermine, and spermidine. Furthermore, the extent of the activation of GRK5 by these polycations appears to be proportional to their charge, since optimal concentrations of spermidine, which contains 3 positive charges, activates GRK5 ˜1.5-fold, while spermine, which contains 4 positive charges, activates GRK5 ˜1.8-fold. Polylysine, a 14-16 amino acid peptide, is the most potent activator of GRK5 among the compounds tested, promoting a ˜2.6-fold activation at an optimal concentration of 10 &mgr;M. While polycations do not activate &bgr;ARK, at lower concentrations they are able to partially reverse the inhibition of &bgr;ARK by heparin, most likely by directly binding to heparin and thereby preventing its inhibition. A similar phenomenon was observed for GRK5, where spermine, spermidine, and polylysine effectively reversed the ability of heparin to inhibit GRK5. Accordingly, it is believed that in vivo polyamines or polycationic surfaces may act as physiological modulators of GRK5.

[0032] Further analysis of GRK5 has revealed that it undergoes rapid intramolecular autophosphorylation reaching a stoichiometry of ˜1.5 mol of phosphate/mol of kinase after ˜15 minutes. Among the GRKs identified to date, only rhodopsin kinase has been previously demonstrated to undergo extensive autophosphorylation. The characteristics of GRK5 autophosphorylation appear similar to those of receptor phosphorylation of GRK5, since both have a similar Km for ATP (21 and 24 &mgr;M, respectively) and both are inhibited by NaCl (IC50 values of ˜70 and 58 mM, respectively). However, heparin, a potent inhibitor of GRK5-mediated receptor phosphorylation, does not inhibit autophosphorylation, even at a concentration of 10 &mgr;M. Testing of additional compounds showed that crude soybean phosphatidylcholine liposomes activate the autophosphorylation. GRK5 was demonstrated to directly bind to the phospholipids, thus, indicating that the interaction between GRK5 and phosphatidylcholine liposomes may serve as a mechanism for promoting GRK5 association with membranes. This phospholipid-stimulated autophosphorylation of GRK5 is believed to provide a mechanism for both activation of kinase, as well as a potential mechanism for the in vivo targeting of GRK5 to its receptor substrates.

[0033] GRK6

[0034] To identify additional GRK cDNAs, a randomly primed, human heart cDNA library was screened with catalytic domain fragments from the bovine &bgr;ARK and &bgr;ARK2 cDNAs by low stringency hybridization. This procedure identified one clone which hybridized at low stringency but not high stringency (washed off between 0.5× and 0.2×SSC at 60° C.) A number of other clones which hybridized at high stringency (0.1×SSC at 60° C.) were presumed to be &bgr;ARK or &bgr;ARK2 and were not pursued further. The one low stringency clone, termed pGRK6-hh1, was ˜1.6 kb in length and was restriction mapped and sequenced in its entirety (FIG. 6). This clone contained an ˜840 bp stretch that had significant homology with the GRKs and, in particular, with IT11 and GRK5. However, the homology of pGRK6-hh1 with the other GRKs ended abruptly within the catalytic domain (FIG. 6) suggesting an incomplete clone. Therefore, a 0.6 kb open reading frame (ORF) DraII fragment from pGRK6-hh1 was isolated and labeled and used to rescreen the human heart cDNA library. This yielded one additional clone,.pGRK6-hh2, that was also isolated, restriction mapped and sequenced (FIG. 6). This clone was ˜3 kb in length and appeared to contain the entire ORF for GRK6. However, the sequence of the hh2 clone diverged from hh1 near the 3′ end of the ORF (marked by an * in FIG. 6). Since this sequence difference lies near an EcoRI site in the hh2 clone that is not found in hh1, the hh2 clone may have resulted from ligation of the GRK6 cDNA with another cDNA during the preparation of the human heart cDNA library. To verify the sequence of GRK6, a human brain cDNA library was screened using a 1780 bp EcoRI fragment from pGRK6-hh2. This resulted in the isolation of 13 additional high stringency clones which were restriction mapped and either partially or completely sequenced. The restriction maps and sequencing strategy for several of the GRK6 clones are depicted in FIG. 6.

[0035] Overall, these overlapping clones yielded 2848 bp of nucleotide sequence (SEQ ID NO: 2) with an open reading frame of 1728 bp flanked by 62 bp of 5′ untranslated and 1058 bp of 3′ untranslated sequence (FIG. 7). The ORF of pGRK6 encodes a protein of 576 amino acids with a predicted molecular mass of 65,968 daltons comprising SEQ ID NO: 6 (GenBank accession number L16862). The first ATG in this sequence may serve as the initiation site for translation since it has a good Kozak consensus sequence and is preceded by a GC-rich region (89% in the first 62 bp), a feature common to all of the GRKs. This predicted initiator methionine and N-terminal region of GRK6 also has significant homology with several of the other GRKs. The overall topology of GRK6 suggests an N-terminal domain of 192 amino acids, a central catalytic domain of 239 amino acids and a C-terminal domain of 145 amino acids. While the N-terminal and catalytic domains of GRK6 are similar in size to those of &bgr;ARK (197 and 239 amino acids), its C-terminus is significantly shorter than that of &bgr;ARK (253 amino acids). The catalytic domain of GRK6 contains the highly conserved GXGXXGX-16K stretch as well as all of the other amino acids that are highly conserved in protein kinases. GRK6 also contains the sequence Asp Leu Gly (residues 329-331) within the catalytic domain, a sequence unique to the GRK family (other protein kinases contain Asp Phe Gly).

[0036] The overall amino acid sequence of GRK6 was compared with those of other GRKs by construction of a phylogenetic tree (FIG. 8). This comparison suggests that there are two major branches of the GRK family tree. One branch contains bovine &bgr;ARK, &bgr;ARK2, and Drosophila GPRK-1 while the other contains bovine rhodopsin kinase, human GRKS, GRK6 and IT11 and Drosophila GPRK-2. GRK6 is most closely related to human GRK5 (70.1% amino acid identity), human IT11 (68.5% identity) and Drosophila GPRK-2 (66.7% identity). The most highly conserved region between these proteins is the protein kinase catalytic domain where GRK6 has 78.7%, 79.5% and 74.9% identity with GRK5, IT11 and GPRK-2, respectively (FIG. 9). The regions of greatest divergence lie in the amino and carboxyl terminal domains. A comparison of the amino terminal domains shows that GRK6 and GRK5 are highly conserved over the first 46 residues (87.0% identity) while over the next 122 amino acids the homology is significantly lower (46.7% identity). By comparison, GRK6 and IT11 are highly conserved over the first 17 amino acids (82.4% identity) but then IT11 has a 32 amino acid gap compared to GRK5, GRK6, &bgr;ARK, &bgr;ARK2, GPRK-1 and rhodopsin kinase. Over the next 120 residues, the amino acid identity between GRK6 and IT11 is 46.7%, comparable to that between GRK6 and GRK5. A notable feature of GPRK-2 is that it has a very short N-terminal domain of only 28 amino acids. The C-terminal domains of GRK6, GRK5, IT11 and GPRK-2 are also somewhat conserved with several stretches of very high homology. Overall, the C-terminal domains of GRK6, GRK5, IT11 and GPRK-2 are 145, 159, 100 and 1S9 amino acids in length, respectively, and share 52-61%. amino acid identity. The role of the N- and C-terminal domains of GRK6 have not been established. However, studies using monoclonal antibodies directed against different regions of rhodopsin kinase suggests that the N-terminal region of rhodopsin kinase contains a sequence involved in the recognition of photolyzed rhodopsin. Mutagenesis studies on &bgr;ARK also suggest that the N-terminal region may play a role in receptor binding. The C-terminal domain of &bgr;ARK appears to interact with G protein &bgr;&ggr; subunits while in rhodopsin kinase the C-terminus is farnesylated. GRK6 contains no consensus sequence for prenylation and it is not activated by G protein &bgr;&ggr; subunits.

[0037] The phylogenetic classification of the GRK family is further supported by functional analyses of the various GRKs. In vitro studies have demonstrated that &bgr;ARK and &bgr;ARK2 share a very similar substrate specificity both at the level of amino acid preference (they phosphorylate serine-containing peptides with N-terminal acidic residues) receptor phosphorylation (&bgr;2AR, m2 muscarinic cholinergic and substance P are good substrates in vitro), and potential mechanism of cellular activation (interaction with G protein &bgr;&ggr; subunits). &bgr;ARK, &bgr;ARK2 and Drosophila GPRK-1 also appear to be ubiquitous proteins being expressed in a variety of tissues. In contrast, the specific localization of rhodopsin kinase in rod and cone outer segments suggests that its ability to rapidly phosphorylate light-activated rhodopsin (and possibly the cone opsins) may be its major role. At present, little is known about the function of GRK6, GRK5, IT11 and Drosophila GPRK-2. However, the similarities between these four proteins suggests that they may well share common roles in the cell. As an initial approach to defining the substrate specificity of GRK6, it was expressed in Sf9 insect cells and assessed for activity using rhodopsin as a substrate.

[0038] To set up an expression system for GRKG, an ˜2 kb ORF NcoI/EcoRI fragment of pGRK6-hb8 was cloned into the baculovirus expression vector pBacPAK1. This construct was co-transfected into Sf9 insect cells with Bsu361 digested BacPAK6 virus to produce a recombinant baculovirus containing the GRK6 cDNA. Sf9 cells were then infected with either the wild type BacPAK6 virus or with the GRK6- or &bgr;ARK-containing baculovirus. The cells were harvested 48 hours post-infection, homogenized and centrifuged at 40,000×g for 20 minutes. When the supernatant fractions were assessed for protein expression by gel electrophoresis, both GRK6 and &bgr;ARK were found to be expressed at a high level with GRK6 migrating at ˜66 kDa (FIG. 10A, lane 3) and &bgr;ARK at ˜80 kDa (lane 4). The supernatant fractions were then tested for their ability to phosphorylate rhodopsin. Supernatants from both uninfected Sf9 cells and BacPAK6 infected cells phosphorylated rhodopsin to a very low extent (FIG. 10B, lanes 1-4). In contrast, the GRK6-containing cell supernatant was able to phosphorylate rhodopsin in a light dependent fashion (lanes 5 and 6). However, from three separate experiments GRK6 was only 4.7% +/−0.4% as active as &bgr;ARK (lane 8) when rhodopsin was used as the substrate.

[0039] Similar to most other protein kinases, GRK6 displays a preference for Mg2+ as the divalent metal cofactor. The optimum concentration was ˜2 mM free Mg2+ with higher concentrations being inhibitory to the kinase. This concentration is similar to the Mg2+ requirements for both &bgr;ARK and GRK5. The optimum Mn2+ concentration was 1 mM which is similar to GRK5. Unlike GRK5, however, 1 mM free Mn2+ results in GRK6 activity comparable to that observed for Mg2+ while GRK5 displays ˜23% the activity when Mn2+ is present compared to Mg2+. Similar experiments with Ca2+ or Zn2+ as the divalent cofactor show that these ions are incapable of facilitating rhodopsin phosphorylation by GRK6. Furthermore, both Ca2+ and Zn2+ inhibited GRK6 activity in the presence of Mg2+. Another difference between GRK6 and GRK5 is found in their Km for ATP. At a rhodopsin concentration of 3 &mgr;M, the Km of GRK6 for ATP is 111 &mgr;M, whereas GRK5 exhibits a Km of ˜24 &mgr;M.

[0040] The ability of purified GRK6 to phosphorylate a specific consensus amino acid sequence was analyzed using a variety of synthetic peptide substrates. The peptides, RRREEESGGG (SEQ ID NO: 7), which is a good substrate for &bgr;ARK, and RRREEESEEE (SEQ ID NO: 8), which is a good substrate for rhodopsin, were poorly phosphorylated by GRK6. These peptides are not phosphorylated by GRK5. Of all the peptides tested, RRRASAAASAA (SEQ ID NO: 9) was found to be the best substrate for GRK6. Interestingly, the same peptide containing a phosphoserine as the first serine is not phosphorylated by GRK6 indicating that GRK6 either prefers the first serine for phosphorylation or that the acidic environment created by the phosphate group prevents the kinase from phosphorylating the second serine residue. GRK6 activity was also apparent with the RRRAEASAA (SEQ ID NO: 10) peptide but not with the RRRAAAEASAAA (SEQ ID NO: 11) peptide indicating that the basic arginine residues enable GRK6 to phosphorylate the peptide if they are relatively close to the serine residue in a situation where an acidic amino acid is also present. However, kemptide (LRRASLG (SEQ ID NO: 12)), a basic peptide substrate for the cAMP-dependent protein kinase, was a poor substrate for GRK6. Overall, the pattern of peptide phosphorylation is similar to that of GRK5; however, GRK6 activity is ˜2-5 fold lower than values for GRK5. From these studies it is apparent that in contrast to &bgr;ARK and rhodopsin, GRK6, like GRK5, does not phosphorylate serine residues in an acidic environment.

[0041] To further characterize GRK6 substrate specificity, several general protein kinase substrates such as histones, casein and phosvitin were studied. As with GRK5, GRK6 was able to phosphorylate the acidic proteins casein and phosvitin while it was only weakly reactive toward the basic protein substrate histones. In experiments with the polyanions, heparin and dextran sulfate, it was demonstrated that GRK6 activity, like GRK5 activity, was strongly inhibited with IC50 values of ˜15 and 7 nM, respectively. Polyglutamic acid and polyaspartic acid were also tested on GRK6 and found to be relatively strong inhibitors of kinase activity (IC50, values of 400 and 487 nM, respectively), but were ˜16- and ˜6-fold less effective, respectively, as inhibitors compared to GRK5. The compounds H-7 and NaCl were shown to inhibit GRK6 at similar concentrations to that of GRK5.

[0042] The ability of polyanions to activate GRK6 was also examined. Spermidine enhanced GRK6 activity ˜1.4-fold over an optimal concentration range of 0.1 to 0.5 mM, while spermine elevated GRK6 phosphorylation of rhodopsin ˜1.5-fold over an optimal concentration range of 0.01 to 0.1 mM. Polylysine was found to the most potent enhancer of GRK6 activity with an ˜2.7-fold increase in phosphorylation (optimal range 1 to 10 &mgr;M). This overall pattern of polycation activation for GRK6 was similar to GRK5. In addition, low concentrations of various polycations were effective at reversing heparin inhibition of GRK6 while higher concentration activated GRK6.

[0043] In contrast to GRK5, however, GRK6 has been found to have a substantially lower ability to autophosphorylate, even in the presence of phospholipid, with a maximal stoichiometry of 0.2 to 0.3 mol of Pi/mol of GRK6. Experiments with several different sources of phospholipid showed no significant increased ability to promote GRK6 autophosphorylation compared to phosphatidylcholine. Thus, it appears that GRK6 uses a different strategy than other G protein-coupled receptor kinases such as GRK5, &bgr;ARK and rhodopsin kinase for kinase targeting to the plasma membrane and activation.

[0044] To assess the tissue distribution of GRK6, RNA blot analysis was performed with a 1780 bp ORF fragment from pGRK6-hh2. This fragment was labeled by random priming and used to probe poly A+ selected RNA from 8 human tissues. Two major RNA species of ˜3 and ˜2.4 kb were observed (was FIG. 11A). The larger message was found at comparable levels in brain, skeletal muscle and pancreas with significantly lower levels being observed in lung, placenta, heart and kidney. In contrast, the smaller message was found predominantly in skeletal muscle with lower levels in brain and still lower levels in pancreas. The significance of the two distinct messages between tissues (e.g., pancreas has predominantly the 3 kb message while skeletal muscle has predominantly the 2.4 kb message) is not known. When the RNA blot was stripped and reprobed with a human &bgr;ARK cDNA probe, a distinct ˜3.8 . kb message was seen (was FIG. 11B), similar to that previously observed for bovine &bgr;ARK. The tissue distribution of human GRK6 and &bgr;ARK appear to be very similar, with the highest levels for both in brain, skeletal muscle and pancreas and the lowest levels in liver.

[0045] GRK5 and GRK6

[0046] Thus, two new members of the G protein-coupled receptor kinase family have been identified. In vertebrates, this family presently consists of &bgr;ARK, &bgr;ARK2, rhodopsin kinase, IT11, and the newly isolated GRK5 and GRK6. Two Drosophila kinases termed GPRK-1 and GPRK-2 have also been identified. Among all the GRKs, rhodopsin kinase appears to be the most appropriately named, given its selected tissue distribution and ability to specifically phosphorylate photolyzed rhodopsin. Since rhodopsin kinase was the first identified G protein-coupled receptor kinase it might well also be termed GRK1. By comparison, &bgr;ARK appears to play a major role in the phosphorylation and desensitization of the agonist-activated &bgr;2-adrenergic receptor, although it is a ubiquitous protein whose expression does not correlate with any particular receptor. In fact, in vitro studies have demonstrated that &bgr;ARK can phosphorylate numerous receptors including the &bgr;2-adrenergic, &agr;2-adrenergic, m2 muscarinic cholinergic and substance P receptors. Thus, &bgr;ARK may well be more appropriately named GRK2, since it was the second member of this family to be identified. The third member of this family, &bgr;ARK2 could be termed GRK3, while IT11 could be appropriately represented as GRK4. In this convention the cloning of GRK5 and GRK6, the fifth and sixth members of the GRK family are described.

[0047] At present, nothing is known about the substrate specificity of the Drosophila GPRK-2 or the recently identified human GRK, IT11. However, GRK5 has been demonstrated to phosphorylate several G protein-coupled. receptors in a stimulus-dependent manner. GRK6 is also able to phosphorylate rhodopsin, &bgr;2-adrenergic and m2 muscarinic cholinergic receptors in a stimulus dependent manner, albeit to a significantly lower extent than &bgr;ARK and GRK5. However, the initial differences between &bgr;ARK, GRK5 and GRK6 in phosphorylating rhodopsin, as well as their distinct tissue distributions, is indicative of their distinct substrate specificities. The involvement of specific GRKs in stimulus dependent protein phosphorylation is believed to serve as a general mechanism for regulating many G protein-coupled receptors.

[0048] The GRK5 and GRK6 cDNAs have been expressed, purified and further characterized with regard to their substrate specificity. The availability of these cDNAs enables the generation of reagents to block the activity of endogenous kinases. These include dominant negative mutations which involves mutating a critical lysine residue in the kinase to arginine. This results in a protein kinase which can still bind to its receptor substrate but can no longer transfer a phosphate group to the receptor. Thus, the mutant kinase can block the ability of the endogenous kinase to phosphorylate a given receptor and thereby block desensitization. An alternative strategy for blocking the activity of the endogenous kinase involves the use of antisense oligonucleotides or stably transfected antisense constructs to block expression of the kinase. This will generate a cell with a reduced ability to desensitize to various agents. The ability of the reagent to inhibit the receptor kinase activity, and therefore block agonist-specific desensitization, is then determined by assaying for the ability of the kinase to phosphorylate the receptor for which it is specific. Assays for phosphorylation are well known in the art.

[0049] Expression of GRK5 and GRK6 also allows screening for specific inhibitors of these two kinases. For example, it has been demonstrated that both heparin and dextran sulfate are potent inhibitors of GRK5 and GRK6, while these polyanions are relatively weak inhibitors of &bgr;ARK. Thus, using well known screening techniques, specific inhibitors of individual G protein-coupled kinases can be identified and then utilized to specifically block the activity of a selected G protein-coupled receptor kinase, thus inhibiting receptor desensitization. Specific inhibitors of these kinases may be used therapeutically to either directly modulate the activity of a given receptor (by blocking endogenous desensitization of that receptor) or by augmenting the ability of a given therapeutic agent to stimulate a given receptor (again by blocking desensitization). Expression of GRK5 and GRK6 also allows for screening of specific activators of these two kinases. For example, the polycations, spermidine, spermine and polylysine have been identified as activators of these two kinases. In similar fashion, additional activators can be identified and then utilized to specifically stimulate the activity of GRK5 or GRK6, thus enhancing receptor desensitization. Specific activators of these kinases may be used therapeutically to modulate the activity of an overactive (e.g., oncogenic) receptor (by stimulating the endogenous desensitization of that receptor).

[0050] The invention is further illustrated by the following non-limiting examples.

EXAMPLES

[0051] Materials

[0052] A human Multiple, Tissue Northern (MTN) blot and the BacPAK baculovirus expression system were purchased from tlontech. Wildtype Spodoptera frugiperda (Sf9) cells were obtained from American Type Culture Collection. Human heart and brain cDNA libraries were obtained from Stratagene. Tissue culture reagents were purchased form Gibco and Sigma while frozen bovine retinas were from George A. Hormel and Co. Taq polymerase was purchased from Promega. Restriction endonuclease and other molecular biology reagents were from Boehringer Mannheim. [&agr;−32P] dCTP, [&agr;−35S] DATP, and [&ggr;−32P] ATP were purchased from NEN.

Example 1

Polymerase Chain Reaction (GRK5)

[0053] Polymerase chain reactions (PCR) initially contained 200 ng DNA template (human heart cDNA),100 pmol oligonucleotide primers, 50 mM KC1, 10 mM Tris-HCl, pH 8.4, 1.5 mM MgCl2 in a total reaction volume of 25 &mgr;l. The samples were heated to 99° C. for 5 minutes, cooled on ice for 3 minutes and then heated to 72° C. followed by the addition of 200 &mgr;M dNTPs and 2 units Taq polymerase. The samples were then denatured at 95° C. for one minute, annealed at 40° C. for one minute and extended at 72° C. for three minutes for 5 cycles. This was followed by 30 cycles of denaturation at 95° C. for one minute, annealing at 50° C. for one minute and extension at 72° C. for three minutes. The sense PCR primer used was 5′-ACIGGIAARATGTAYGC-3 (SEQ ID NO: 3) encoding the amino acid sequence T (G/L) KMYA while the antisense PCR primer was 5′ -YTCIGGIGCCATRWAIC-3′ (SEQ ID NO: 4) encoding G(Y/F) MAPE (I=inosine, R=A+G, Y=T+C, W=T+A) The sense primer also contained an EcoRI restriction site at the 5′ end while the antisense primer contained a HindIII restriction site at the 5′ end. This facilitated the cloning of the PCR products into EcoRI/HindIII digested pBluescript for further analysis.

Example 2

Isolation of a Full Length cDNA Clone for GRK5

[0054] In order to isolate a full length cDNA, the 450 bp PCR product was labeled with [32p] dCTP by random priming and then used to probe a human heart cDNA library in &lgr;ZAP. The cDNA library was plated at a density of 50,000 pfu/150 mm dish (106 total clones), transferred to nitrocellulose filters and hybridized with the labeled PCR probe for 48 hours at 37° C. in buffer containing 25%. formamide, 5×saline sodium citrate (SSC), 5×Denhardt's solution, 1% SDS, 0.1% sodium pyrophosphate and 100 &mgr;g/ml denatured salmon sperm DNA. The filters were initially washed in 2×SSC, 0.1% SDS at 50° C. for 1 hour followed by a high stringency wash in 0.1×SSC, 0.1%. SDS at 65° C. for 1 hour. The six clones identified by this procedure were isolated by repeated plating and screening with the labelled PCR product. The isolated clones were then rescued with a helper phage to yield the cDNAs as inserts in pbluescript SK. All six clones were restriction mapped and sequenced by the dideoxynucleotide chain termination technique using T3 and T7 primers and oligonucleotide primers synthesized to known regions of the sequence.

Example 3

Northern Blot Analysis (GRK5)

[0055] Tissue distribution of the mRNA for GRK5 was analyzed on a human Multiple Tissue Northern (MT14) blot containing 2 &mgr;g of poly A+ RNA from human heart, brain, placenta, lung, liver, skeletal muscle, kidney and pancreas. A 1157 bp PstI ORF fragment (bp 383-1540) from pGRK5 was isolated and labeled with [32P] dCTP by random priming to a specific activity of 3.5×109 cpm/&mgr;g. This probe was hybridized with the human MTN blot for 24 hours at 42° C. in a buffer containing 5×SSPE, 10×Denhardt's solution, 100 &mgr;g/ml salmon sperm DNA, 50% formamide and 2% SDS. The blot was washed with 2×SSC, 0.1% SDS at room temperature for 1 hour followed by a high stringency wash with 0.1×SSC, 0.1% SDS at 65° C. for 1 hour. The blot was then autoradiographed at −80° C. for 24 hours.

EXAMPLE 4

Expression of GRK5 Using the Baculovirus System

[0056] The GRK5 ORF was initially excised by restriction digestion of the full length clone pGRK5 using NaeI and SmaI. The resulting 1906 bp fragment, including 27 bp of 5′ untranslated and 109 bp of 3 untranslated sequence flanking the 1770 bp ORF, was isolated on a 0.8%. low melting agarose gel. The baculovirus expression vector pBacPAK1 was digested with BamHI, blunted with Klenow and dephosphorylated with calf intestinal phosphatase. The 1837 bp GRK5 ORF fragment was then ligated into the blunted BamHI site of pBacPAK1 to generate the construct pBacPAK-GRK5.

[0057] Monolayers of Sf9 cells (3×106 cells) were co-transfected with 1 &mgr;g of the pBacPAK-GRK5 construct and 0.25 &mgr;g of Bsu36I digested BacPAK6 viral DNA using the calcium phosphate precipitation technique. The cells were incubated for 4 hours at 27° C. and the medium was then replaced with complete media (TNM-FH, 10% fetal bovine serum, 2.5 &mgr;g/ml fungizone, 50 &mgr;g/ml streptomycin, 50 &mgr;g/ml penicillin). The cells were incubated for 6 days at 27° C. To obtain the isolated recombinant virus, 4×106 cells were plated on a 35 mm dish and the cells were overlaid with 1.5 ml of the diluted viral stock from the transfection. After 1.5 hours, the virus was removed and the cells were overlaid with 5 ml of 1% low melting agarose in complete media. The plates were incubated at 27° C. and isolated plaques were observed after 4-6 days. In order to confirm the presence of the GRK5 cDNA in the recombinant virus, six isolated viruses were amplified and the viral DNA was extracted and analyzed by restriction digestion and Southern blot hybridization. One of these recombinant viruses was chosen for further amplification and characterization of the expressed kinase.

[0058] For the purpose of preliminary characterization of GRK5 and to compare it with SARK, the respective recombinant baculoviruses were used to infect a monolayer of Sf9 cells in a 100 mm dish. Following a 48 hour infection, the cells were rinsed with phosphate buffered saline and were harvested by scrapping and homogenizing in 1.5 ml of ice cold buffer containing 20 mM Hepes, pH 7.2, 10 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, 20 &mgr;g/ml leupeptin, 200 &mgr;g/ml benzamidine, 250 mM NaCl, 0.02%. Triton X-100. The cells were lysed with a Brinkman tissue disrupter (30 seconds at 30,000 rpm) and were centrifuged at 40,000×g for 20 minutes. The supernatants were then assessed for the ability to phosphorylate urea treated rod outer segments.

Example 5

Preparation of G Protein &bgr;&ggr; Subunits

[0059] The GTP binding proteins Go and Gi were purified from bovine brain by successive chromatography on DEAE Sephacel, Sephacryl S200 and heptylamine-Sepharose. The G protein preparation, consisting primarily of Gi and, Go. was further purified by chromatography on a Mono Q column and then stored in 20 mM Tris-HCl, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol, 250 mM NaCl, and 0.05% Lubrol (buffer A) at −80° C. The &bgr;&ggr; subunits were isolated by chromatography of the purified G proteins on heptylamine-Sepharose in the presence of AMF (30 &mgr;M AlCl3, 6 mM MgCl2, 10 mM NaF). The AMF and cholate were removed from the &bgr;&ggr; subunit preparation by anion exchange chromatography on a Mono Q column. The purified &bgr;&ggr; subunits were stored in buffer A at −80° C.

Example 6

Receptor Phosphorylation

[0060] Rod outer segments (ROS) were prepared in accordance with methods known to those skilled in the art. Rhodopsin kinase free membranes were prepared by suspending the ROS in 50 mM Tris-HCl, pH 8.0, 5 mM EDTA, 5M urea and sonicating. The urea treated ROS were washed several times by centrifugation and resuspended in 50 mM Tris-HCl, pH 7.5 at a protein concentration of ˜1 mg/ml. Phosphorylation reactions contained 5 &mgr;g of total protein from the Sf9 cell supernatants expressing either GRK5 or &bgr;ARK, urea treated ROS (80 pmol rhodopsin), 20 mM Tris-HCl, pH 7.5, 2 mM EDTA, 6 mM MgCl2 and 0.1 mM [&ggr;-32P] ATP, in a total reaction volume of 30 &mgr;l. When the effect of G protein &bgr;&ggr; subunits was assessed, the reactions also contained 0.25 mM dithiothreitol, 0.01% Lubrol and 62 mM NaCl with or without 100 nM &bgr;&ggr; subunits. Reactions were incubated at 30° C. for 15 minutes and then stopped by the addition of 50 &mgr;l of SDS sample buffer. Samples were electrophoresed on a 10% homogeneous SDS polyacrylamide gel by the method of Laemmli which is well known to those skilled in the art. Gels were dried and autoradiographed at room temperature for 30-120 minutes. The rhodopsin bands in the gel were then cut and counted in a scintillation counter.

Example 7

Isolation and Sequencing of a &bgr;ARK-related cDNA

[0061] A randomly-primed human heart cDNA library (106 clones) in &lgr;ZAP was initially screened with catalytic domain fragments from the clones p&bgr;-ARK3A encoding bovine &bgr;ARK and p&bgr;ARK2-1B encoding bovine &bgr;ARK2. The probes were made by labeling a 717 bp SacI fragment from p&bgr;-ARK3A (bp 630-1347) and a 1210 bp HindIII fragment from p&bgr;ARK2-1B (bp 328 to 1537) by random priming using [32P] dCTP. The cDNA library was plated at a density of −50,000 plaques/150 mm plate and the DNA was then transferred to nitrocellulose filters using standard procedures. The filters were pre-hybridized for 4 hours at 37° C. in a buffer containing 25%. formamide, 5×SSC (1×=150 mM sodium chloride, 15 mM sodium citrate, pH 7.0), 5×Denhardt's solution, 1% sodium. dodecyl sulfate (SDS), 0.1% sodium pyrophosphate and 100 &mgr;g/ml denatured salmon sperm DNA and then hybridized with the labeled &bgr;ARK cDNA fragments for 48 hours at 37° C. in the same buffer. The filters were initially washed in 2×SSC, 0.1% SDS buffer at 50° C. and the stringency was then increased by increasing the washing temperature to 60° C. and successively lowering the salt concentration to 0.5×, 0.2× and 0.1×SSC. This procedure identified one clone which hybridized at low but not high stringency (washed off between 0.5× and 0.2×). A number of other clones which hybridized at high stringency (0.1×SSC at 60° C.) were presumed to be &bgr;ARK or &bgr;ARK2 and were not pursued further. The one low stringency clone, termed pGRK6-hh1, was purified by replating and screening with the labeled &bgr;ARK cDNA fragments. The isolated clone was then rescued with a helper phage to yield the insert in the vector Pbluescript. This clone was restriction mapped and sequenced using the dideoxynucleotide chain termination method. Since pGRK6-hh1 was found to be a partial clone, a ˜0.6 kb DraII fragment from this clone was isolated, labeled by random priming and used to rescreen the human heart cDNA library. This rescreening yielded one clone, pGRK6-hh2, which was also excised and sequenced. To isolate additional clones, a 1780 bp EcoRI fragment from pGRK6-hh2 was labeled and used to screen a human fetal brain cDNA library. This screening yielded a total of 13 clones which hybridized at high stringency with the pGRK6-hh2 probe. All of these clones were restriction mapped and sequenced with T3, T7 and selected oligonucleotides from the pGRK6-hh2 sequence.

Example 8

Northern Blot Analysis for GRK6

[0062] A human multiple tissue Northern (MTN) blot, containing 2 &mgr;g of poly A+ RNA from human heart, brain, placenta, lung, liver, skeletal muscle, kidney and pancreas, was used for RNA analysis. The blot was initially hybridized with a randomly primed 1780 bp EcoRI fragment from pGRK6-hh2 in 5×SSPE (1×=150 mM NaCl, 10 mM NaHPO4, 1 mM EDTA, pH 7.0), 50% formamide, 10×Denhardt's solution, 2% SDS and 100 &mgr;g/ml denatured salmon sperm DNA. The blot was washed in 0.1×SSC at 60° C. before autoradiography at ˜80° C. for 3 days. The blot was stripped by boiling in water for 10 minutes and then reprobed with a 1556 bp EcoRI/NcoI ORF fragment (bp 357-1912) the human &bgr;ARK CDNA. The blot was again washed in 0.1×SSC at 60° C. before autoradiography at ˜80° C. for 3 days.

Example 9

Expression of GRK6 Using the Baculovirus System

[0063] The clone pGRK6-hb8 (which contains bp 27 to 2064 of the sequence shown in FIG. 7) was used to express GRK6. An ORF fragment of the GRK6 cDNA was prepared by initially linearizing pGRK6-hb8 for 15 minutes with the restriction enzyme NcoI (sites at bp 61 and 748). The linearized DNA was resolved on a 0.8% low melting agarose gel, extracted and then digested x with EcoRI (sites at 5′ and 3′ ends of the cDNA. The reaction was quenched by incubation at 80° C. for 10 minutes and the DNA was then blunted with Klenow in the presence of dNTPs. The resulting 2004 bp insert (cut at bp 61 and 2064) was separated from the vector and shorter inserts by electrophoresis on a 0.8% low melting agarose gel. This insert contains 2 bp of 5′ and 274 bp of 3′ untranslated sequence surrounding the 1728 bp open reading frame of GRK6 (bp 63-1790). This insert was then ligated into the baculovirus expression vector pBacPAK1 (prepared by cutting with BamHI, blunting with Klenow, and treating with calf intestinal phosphatase).

[0064] Sf9 cells were cultured on a monolayer or in suspension (spinner flask, 70 rpm) using TNM-FH medium containing 10% fetal bovine serum and antibiotics (2.5 &mgr;g/ml fungizone, 50 &mgr;g/ml streptomycin, 50 &mgr;g/ml penicillin). To produce a recombinant baculovirus containing the GRK6 CDNA, Sf9 cells were cotransfected with 1 &mgr;g of the pBacPAK-GRK6 DNA and 0.25 &mgr;g of Bsu361 digested BacPAK6 viral DNA using the calcium phosphate precipitation technique which is well known in the art. The transfected cells were allowed to recover and produce phage particles in culture media for five days and the virus laden media was then used to infect a fresh monolayer of Sf9 cells. The infected cells were overlaid with 1% low melting agarose in complete media and after a 4 day incubation isolated viral plaques were selected. The plaques were eluted into complete media and were amplified by reinfection of a fresh monolayer of Sf9 cells. Using this procedure, it was found that greater than 90% of the viral plaques contained isolated recombinant viruses. The presence of the full length GRK6 cDNA within the genome of the recombinant baculovirus was confirmed by restriction mapping and Southern blotting.

[0065] To initially characterize the activity of GRK6 and compare it with that of &bgr;ARK, the respective recombinant baculoviruses were used to infect a monolayer of Sf9 cells (100 mm dish). Uninfected Sf9 cells and cells infected with the wild type BacPAK6 virus served as controls. The cells were harvested 48 hours post-infection by rinsing the dishes several times with phosphate buffered saline and then scraping and homogenizing the cells in 1.5 ml of ice cold buffer (20 mM Hepes, pH 7.2, 10 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, 20 &mgr;g/ml leupeptin, 200 &mgr;g/ml benzamidine, 0.02% Triton X-100, 250 mM NaCl, 1 mM dithiothreitol). The cells were lysed with a Brinkman

Claims

1. A method of screening for compounds capable of inhibiting the activity of a selected G protein-coupled receptor kinase comprising:

(a)contacting a sample containing a selected G protein-coupled receptor kinase encoded by a member of the G protein-coupled receptor kinase gene family with a selected compound under conditions such that the selected G protein-coupled receptor kinase present in the sample and the selected compound interact; and

(b)determining the ability of the selected compound to inhibit the activity of the selected G protein-coupled receptor kinase,

wherein the selected G protein-coupled receptor kinase is GRK5 or GRK6.

2. The method of claim 1, wherein the selected G protein-coupled receptor kinase is GRK5.

3. The method of claim 2, wherein said GRK5 is encoded by a nucleic acid sequence set forth in SEQ ID NO:1.

4. The method of claim 1, wherein the selected G protein-coupled receptor kinase is GRK6.

5. The method of claim 4, wherein said GRK6 is encoded by a nucleic acid sequence set forth in SEQ ID NO:2.

6. A method of inhibiting the activity of GRK5 or GRK6 comprising administering an effective amount of an inhibitor identified in claim 9 so that the activity of said GRK5 or GRK6 is inhibited.

7. A method of screening for compounds capable of inhibiting the activity of a selected G protein-coupled receptor kinase having an amino acid sequence set forth in SEQ ID NO: 1 or SEQ ID NO: 2, the method comprising:

(a) contacting a sample containing the selected G protein-coupled receptor kinase with a selected compound under conditions such that the selected G protein-coupled receptor kinase present in the sample and the selected compound interact; and

(b) determining the ability of the selected compound to inhibit the activity of the selected G protein-coupled receptor kinase.

8. A method of screening for compounds capable of enhancing the activity of a selected G protein-coupled receptor kinase comprising:

(a) contacting a sample containing a selected G protein-coupled receptor kinase encoded by a member of the G protein-coupled receptor kinase gene family with a selected compound under conditions such that the selected G protein-coupled receptor kinase present in the sample and the selected compound interact; and

(b) determining the ability of the selected compound to enhance the activity of the selected G protein-coupled receptor kinase,

wherein the selected G protein-coupled receptor kinase is GRK5 or GRK6.

9. The method of claim 8, wherein the selected G protein-coupled receptor kinase is GRK5.

10. The method of claim 9, wherein said GRK5 is encoded by a nucleic acid sequence set forth in SEQ ID NO:1.

11. The method of claim 8, wherein the selected G protein-coupled receptor kinase is GRK6.

12. The method of claim 11 wherein said GRK6 is encoded by a nucleic acid sequence set forth in SEQ ID NO:2.

13. A method of enhancing the activity of a selected G protein-coupled receptor kinase comprising administering an effective amount of a compound capable of enhancing activity of the selected G protein-coupled receptor kinase identified in claim 8 so that the activity of the selected G protein-couple receptor kinase is enhanced.

14. A method of modulating the expression of an endogenous G protein-coupled receptor kinase by a cell comprising contacting the cell capable of expressing the G protein-coupled receptor kinase selected from the group consisting of GRK5 and GRK6 with a corresponding dominant negative mutant of the G protein-coupled receptor kinase which is capable of binding to its receptor substrate but incapable of phosphorylating said receptor substrate.

15. The method of claim 14, wherein said mutant of the G protein-coupled receptor kinase carries a mutation from a lysine residue to an arginine residue.

16. A method of modulating the expression of an endogenous G protein-coupled receptor kinase by a cell comprising contacting the cell capable of expressing the G protein-coupled receptor kinase selected from the group consisting of GRK5 and GRK6 with a composition containing a selected polyanion or polycation at levels sufficient to influence the expression of the endogenous G protein-coupled receptor kinase.

17. The method of claim 16, wherein the composition comprises a polyanion selected from the group consisting of heparin, dextran sulfate, polyaspartic acid and polyglutamic acid.

18. The method of claim 16, wherein the composition comprises a polycation selected from the group consisting of polylysine, spermine and spermidine.