Role of C-terminal Serines in Desensitization and Phosphorylation of the Mouse Thromboxane Receptor*

To investigate the role of C-terminal hydroxyamino acids in desensitization of the receptor for thromboxane A2 (TxA2), we created a mutant TxA2 receptor (TP receptor) in which serines at positions 321, 322, and 328 were replaced with either alanine or glycine. Mutant and wild type receptors were expressed in a mesangial cell line, and clones expressing similar numbers of receptors were studied. Affinity and specificity of TxA2 binding to the mutant receptor were identical to wild type receptors. In contrast, TxA2-induced inositol trisphosphate generation by the mutant receptor was enhanced compared with the wild type. Prior treatment with the TxA2agonist U46619 reduced subsequent U46619-induced increases in inositol trisphosphate generation by both receptors; however, the extent of desensitization was significantly reduced in the receptor mutant. Protein kinase C (PKC) inhibitors attenuated TxA2-induced desensitization of wild type receptors, but had little effect on TxA2-induced desensitization of mutant receptors. Pretreatment with the phorbol ester phorbol 12,13-dybutyrate (PDBu) (100 nm) decreased subsequent responsiveness of wild type but not mutant TP receptors. U46619-induced desensitization of wild type receptors was associated with enhanced phosphorylation of receptor proteins. This agonist-specific phosphorylation of the TP receptor was largely prevented by inhibitors of PKC. Treatment with 100 nm PDBu increased phosphorylation of both wild type and mutant TP receptors, but the extent of phosphorylation of the receptor mutant was reduced compared with the wild type. Increasing the concentration of PDBu from 100 nm to 1 μmPDBu reduced responsiveness of both mutant and wild type receptors without enhancing phosphorylation of either of the receptor proteins. These data suggest that 1) phosphorylation of C-terminal serines contributes to agonist-specific desensitization of the TP receptor, 2) PKC-induced desensitization of TP receptors is caused, in part, by phosphorylation of C-terminal serines, and 3) desensitization of TP receptors by PKC is complex and involves mechanisms that may not require direct phosphorylation of receptor proteins.

To investigate the role of C-terminal hydroxyamino acids in desensitization of the receptor for thromboxane A 2 (TxA 2 ), we created a mutant TxA 2 receptor (TP receptor) in which serines at positions 321, 322, and 328 were replaced with either alanine or glycine. Mutant and wild type receptors were expressed in a mesangial cell line, and clones expressing similar numbers of receptors were studied. Affinity and specificity of TxA 2 binding to the mutant receptor were identical to wild type receptors. In contrast, TxA 2 -induced inositol trisphosphate generation by the mutant receptor was enhanced compared with the wild type. Prior treatment with the TxA 2 agonist U46619 reduced subsequent U46619-induced increases in inositol trisphosphate generation by both receptors; however, the extent of desensitization was significantly reduced in the receptor mutant. Protein kinase C (PKC) inhibitors attenuated TxA 2 -induced desensitization of wild type receptors, but had little effect on TxA 2 -induced desensitization of mutant receptors. Pretreatment with the phorbol ester phorbol 12,13-dybutyrate (PDBu) (100 nM) decreased subsequent responsiveness of wild type but not mutant TP receptors. U46619-induced desensitization of wild type receptors was associated with enhanced phosphorylation of receptor proteins. This agonist-specific phosphorylation of the TP receptor was largely prevented by inhibitors of PKC. Treatment with 100 nM PDBu increased phosphorylation of both wild type and mutant TP receptors, but the extent of phosphorylation of the receptor mutant was reduced compared with the wild type. Increasing the concentration of PDBu from 100 nM to 1 M PDBu reduced responsiveness of both mutant and wild type receptors without enhancing phosphorylation of either of the receptor proteins. These data suggest that 1) phosphorylation of C-terminal serines contributes to agonist-specific desensitization of the TP receptor, 2) PKCinduced desensitization of TP receptors is caused, in part, by phosphorylation of C-terminal serines, and 3) desensitization of TP receptors by PKC is complex and involves mechanisms that may not require direct phosphorylation of receptor proteins.
Thromboxane A 2 (TxA 2 ) 1 is a labile lipid mediator with po-tent platelet aggregating and vasoconstrictor actions that have been implicated in thrombotic and vasospastic disorders affecting the heart, lungs, kidneys, and peripheral vascular system (1)(2)(3). Its effects are mediated by activating specific cell surface receptors and are subject to regulatory controls (4 -6). Prior exposure to TxA 2 results in decreased receptor responsiveness (4,6). This loss of receptor responsiveness or desensitization has been extensively studied in the cAMP-coupled ␤ 2 -adrenergic receptor and the phosphodiesterase-coupled receptor rhodopsin. In these receptor systems, desensitization is largely caused by direct phosphorylation of receptors at serine and threonine residues, often in the C terminus (reviewed in Dohlman et al. (7)). Receptor phosphorylation is mediated by both receptor specific kinases and general kinase systems such as protein kinase A and C (8). These kinase systems provide a mechanism for regulating receptor activity through negative feedback loops (homologous desensitization) as well as for modulating receptor responsiveness through cross-talk between different receptor systems (heterologous desensitization).
The receptor for TxA 2 belongs to the large superfamily of heptahelical G protein-coupled receptors (9 -11). In most cell systems, TP receptors couple to phospholipase C through pertussis toxin insensitive G proteins (12,13). More recently, two isoforms of the human TP receptor have been described (14) which, in addition to coupling to phospholipase C, oppositely regulate adenylyl cyclase activity (15). Previous studies suggest that desensitization of TP receptors is mediated, in part, through activation of PKC (4,6). PKC-induced desensitization may involve direct phosphorylation of C-terminal domains of the TP receptor because 1) truncation of the C terminus attenuates desensitization mediated by direct activation of PKC by phorbol esters (16), 2) PKC can phosphorylate C-terminal sequences of the TP receptor in vitro (5), and 3) phorbol esterinduced activation of PKC causes phosphorylation of the human TP receptor isoforms in the intact cell (17). While the C terminus of the TP receptor contains several potential PKC phosphorylation sites, it is not known if phosphorylation of these domains contributes to TP receptor desensitization. We therefore constructed a mutant TP receptor (mutant 1) in which serines with favorable phosphorylation motifs for PKC (positions 321, 322, and 328) were replaced with either alanine or glycine. Both mutant and wild type receptors were epitope-tagged to permit rapid isolation of receptor proteins and assessment of their phosphorylation state. As an additional control, we compared desensitization of mutant 1 with another mutant receptor (mutant 2) in which a C-terminal serine lacking a consensus motif for phosphorylation by PKC (position 339) was replaced with glycine. Using these constructs, we investigated the role of C-terminal serines in regulating TP receptor responsiveness by direct phosphorylation of receptor proteins.

MATERIALS AND METHODS
Isolation and Mutagenesis of a Genomic Clone Encoding the Mouse TP Receptor-A genomic clone encoding the mouse TP receptor was isolated as described previously (16). We subcloned a XhoI/ApaI fragment of the genomic clone containing the complete amino acid-encoding regions into the mammalian expression vector pcDNA 3 (Invitrogen, San Diego, CA). Mutagenesis of the wild type receptor was performed using the polymerase chain reaction (PCR) (18). To change serines at positions 321, 322, and 328 to either alanines or glycine, we replaced nucleotides at positions 961, 964, and 982 with guanines using the primer pairs encompassed nucleotides 952-997 (CCGCGGTTCGGTG-CACAGCTCCAGGCTGTGGCCTTGCGCCGGCCT) and 1030 -1009 (TCCTTCAGGGTCCAGTGAGCAT) of the mouse TP receptor cDNA (11). To change serine at position 339 to glycine, we replaced the nucleotide at position 1015 with guanine using the primer pairs encompassed nucleotides 790 -809 (GCAGACTTTGTTGCAGACAC) and 1030 -1001 (CCTTCAGGGTCCACCGAGCATGGCCTGGGC) of the mouse TP receptor cDNA (11). PCR products were cloned into the TA cloning vector (Invitrogen) using the manufacturer's directions. To create mutant receptors lacking C-terminal serines, we took advantage of an unique SacII restriction site located at nucleotide 955 (11) in the C terminus of the TP receptor. Appropriately sized SacII/ApaI fragments were isolated by polyacrylamide gel electrophoresis (19) from the TA cloning vector containing our PCR products. These SacII/ApaI fragments were subcloned into the mammalian expression vector pcDNA 3 (Invitrogen) containing our genomic clone. Mutant constructs were sequenced using the dideoxy method (19) to confirm the desired mutations.
Epitope-tagged TP receptors were created using PCR to insert the 12CA5 epitope (20) at both the N and C termini of the TP receptor as follows. For the N-terminal epitope, we used primer pairs encompassed nucleotides Ϫ12-18 (GGCTTAGGAGCCATGTACCCATAC-GACGTC-CCAGACTACGCTTGGCCCAATGGCACC) and 728 -709 (TGAACCAT-CATCACC-ACCTC) of the TP receptor cDNA (11) where the primer encompassing nucleotides Ϫ12-18 contained the 12CA5 epitope sequence. For the C-terminal epitope, we used primer pairs encompassed nucleotides 790 -809 (GCAGACTTTGTTGCAGACAC) and 1035-1009 (GACTGTCCTTCAGGCGTAGTCTGGGACGTCGTATGGGTAGGGTC-CACTGAGCAT) of the TP receptor cDNA (11), where the primer encompassing nucleotides 1035-1009 contained the 12CA5 epitope sequence. Templates for the PCR reactions were the vectors containing either the wild type or mutant TP receptor lacking three C-terminal serines. PCR products were cloned into the TA cloning vector (Invitrogen) using the manufacturer's directions. To create epitope-tagged receptors, we took advantage of an unique SFiI restriction site located at nucleotide 373 of the TP receptor as well as an unique SacII restriction site located at nucleotide 955 in the C terminus of the TP receptor. First, appropriately sized EcoRI/SFiI fragments were isolated by polyacrylamide gel electrophoresis (19) from the TA cloning vector containing PCR products with the N-terminal epitope. These EcoRI/SFiI fragments were subcloned into the mammalian expression vector pcDNA 3 (Invitrogen) containing the wild type TP receptor. Following insertion of the N-terminal epitope, appropriately sized SacII/ApaI fragments were isolated by polyacrylamide gel electrophoresis (19) from the TA cloning vector containing PCR products derived from either the wild type or mutant receptor lacking three C-terminal serines. These SacII/ApaI fragments were subcloned into the mammalian expression vector pcDNA 3 (Invitrogen) containing the TP receptor with the N-terminal epitope. Mutant constructs were sequenced using the dideoxy method (19) to confirm the desired mutations.
Culture and Transfection of a Mouse Mesangial Cell Line-Mouse mesangial cells derived from SV40 transgenic mice (21) were obtained from American Type Culture Collection (Rockville, MD). Cells were grown and subcultured as described previously (16). To create cell lines stably expressing wild type or mutant TP receptors, our pcDNA 3 expression vector containing either the wild type or mutant constructs was transfected into SV40-transformed mouse mesangial cells by the calcium-phosphate method (19). To isolate permanent transfectants, G418-resistant cells were selected in complete medium containing 500 g/ml G418. Following G418 selection, individual clones were screened for TxA 2 binding as described below.
Culture and Transfection of Human Embryonic Kidney (HEK) 293 Cells-HEK293 cells were obtained from American Type Culture Collection (Rockville, MD). Cells were grown in 75% Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum, penicillin (100 units/ml), and streptomycin (100 g/ml) (all from Life Technologies, Inc.) at 37°C in a humidified atmosphere of 95% air and 5% CO 2 . HEK293 cells were subcultured every week after becoming confluent using 0.25% trypsin with 1 mM EDTA (Life Technologies, Inc.). Cell viability was assessed by standard dye exclusion techniques (0.1% trypan blue) and was always greater than 95%.
For transfection, HEK293 cells were plated in 100-mm dishes and grown to approximately 80% confluency. Cells were then transfected using LipofectAMINE (Life Technologies, Inc.) according to the manufacturer's directions. For each transfection we used 8 g of DNA and 24 l of LipofectAMINE based on preliminary experiments that suggested these amounts optimized the level of receptor expression and transfection efficiency (Ϸ40%). Briefly, either DNA or LipofectAMINE was mixed in 0.8 ml of DMEM. The DNA and LipofectAMINE solutions were then combined, mixed, and incubated at room temperature. After 30 min, 6.4 ml of DMEM was added to the LipofectAMINE and DNA mixture. This solution was mixed and added to the 100-mm dishes containing washed HEK293 cells. Following an overnight incubation, the LipofectAMINE-DNA solution was replaced with DMEM supplemented with 10% heat-inactivated fetal calf serum, penicillin (100 units/ml), and streptomycin (100 g/ml). HEK293 cells were studied 48 h following transfection.
Ligand Binding Assays-Whole cell ligand binding assays were performed as described previously (16) using the stable radiolabeled thromboxane receptor antagonist 3 H-labeled SQ29548 (22) (NEN Life Science Products) and the following unlabeled compounds: SQ29548 (Squibb Institute, Princeton, NJ), the thromboxane agonists U46619 (23) (Cayman Chemicals, Ann Arbor MI), or [ 127 I]BOP (24) (Cayman Chemicals), the inactive thromboxane metabolite thromboxane B 2 (Advanced Magnetics Inc., Cambridge, MA), or prostaglandin E 2 (Advanced Magnetics). Equilibrium binding data were analyzed by the method of Scatchard (25) to give estimates of the maximal number of specific binding sites (B max ) and apparent equilibrium K d by fitting the data to a nonlinear model using the ENZFITTER computer program (Elsevier-Biosoft, Cambridge, UK). For the competitive binding assays, data were analyzed by the method of Cheng and Prusoff (26) to calculate the dissociation constant for each inhibitor (K i ). Protein concentration used in the binding assays was determined using the method of Bradford (27).

Measurement of [ 3 H]Inositol
Phosphate Generation-Inositol phosphates were measured as described previously using anion exchange chromatography (16). For the desensitization experiments, cells were pretreated for 10 min with the indicated concentrations of agonists, inhibitors, or their vehicles in 2-ml Krebs-Ringer buffer (KRB) containing 118 mM NaCl, 4.6 mM KCl, 24.9 mM NaHCO 3 , 1 mM KH 2 PO 4 , 11.1 mM glucose, 1.1 mM MgSO 4 , 1.0 mM CaCl 2 , 5 mM HEPES, and 0.1% bovine albumin, pH 7.4, at 37°C. After desensitization, cells were washed three times with KRB and then incubated in 2 ml of KRB for 4 min before adding 2 M lithium chloride to a final concentration of 20 mM. This 4-min time period allowed inositol phosphate levels to return to baseline following treatment with U46619 (data not shown). One minute after adding the lithium chloride solution, cells were stimulated with the indicated concentrations of U46619 or its vehicle for 2 min. The reaction was stopped, and samples were processed as described previously (16).
Immunoprecipitation of Epitope-tagged TxA 2 Receptors-Forty-eight hours following transfection, cells were washed with KRB without KH 2 PO 4 and then incubated in KRB without KH 2 PO 4 but containing 0.1% bovine serum albumin (fraction V) and 0.1-0.2 mCi of 32 P. After 90 min, cells were stimulated with agonist for 10 min at the indicated concentrations. The reaction was stopped by washing the cells with ice cold Dulbecco's phosphate-buffered saline and then scrapping the cells into 1 ml of ice-cold lysis buffer containing 150 mM NaCl, 50 mM Tris, pH 8.0, 1.0% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 2 mM EDTA, 10 mM sodium fluoride, 10 mM sodium pyrophosphate, 100 nM calyculin, 1 g/ml leupeptin, 1 g/ml pepstatin, 1 g/ml aprotinin, and 100 g/ml phenylmethanesulfonyl fluoride. The lysate was transferred to a 1.5-ml microcentrifuge tube and then rocked for 30 min at 4°C. Insoluble material was removed by centrifugation at 10,000 ϫ g for 4 min. One milliliter of supernatant was transferred to a 1.5-ml microcentrifuge tube, and 10 g of 12CA5 monoclonal antibody (Boehringer Mannheim) were added. After rocking at 4°C for 1 h, 70 l of 20% (v/v) protein A-Sepharose was added, and the samples rocked at 4°C for 1 h. The protein A-Sepharose was washed twice in lysis buffer supplemented with 0.1% ammonium sulfate and once in lysis buffer without ammonium sulfate. SDS-sample buffer (100 l) was added to the pellet and boiled for 10 min. Proteins were separated on 12% polyacrylamide gel with 0.1% SDS as described by Laemmli (28). After drying the gels, phosphorylated proteins were detected by autoradiography.
Statistical Analysis-Data are presented as the mean Ϯ S.E. of the mean. For comparisons between two groups, statistical significance was assessed using an unpaired t test. For comparisons between more than two groups, statistical analysis was performed by analysis of variance followed by Bonferroni's procedure for multiple pairwise comparisons (29).

The Absence of C-terminal Serines of the TP Receptor Does
Not Affect TxA 2 Binding-We used PCR to create 2 mutant TP receptors as described under "Materials and Methods." As shown in Fig. 1, serines at positions 321, 322, and 328 were changed to either alanine or glycine in mutant 1. For mutant 2, the serine at position 339 was changed to glycine (Fig. 1). Plasmids containing either the wild type or the mutant constructs were transfected into a mouse mesangial cell line (21). Following G418 selection, we selected clones expressing either wild type receptors, mutant 1, or mutant 2 with similar levels of TxA 2 binding as shown in Table I. Competitive displacement assays found that the wild type and mutant receptor displaced TxA 2 ligands with similar K i values (Table II), suggesting that the absence of the C-terminal serines does not affect binding of TxA 2 to its receptor. TxA 2 binding was less than 20 fmol/mg of protein in nontransfected cells or cells transfected with vector alone (data not shown).

PKC-induced Desensitization of TP Receptors Is Attenuated by the Absence of C-terminal Serines-
To investigate the role of C-terminal serines in desensitization of the TP receptor, mesangial cells expressing wild type or mutant TP receptors were incubated with vehicle or 1 M U46619 prior to washing and rechallenge with 1 M concentrations of the TxA 2 agonist. As shown in Table III, IP 3 generation in vehicle-treated cells tended to be increased in mutant 1 compared with wild type receptors or mutant 2. Prior incubation with U46619 reduced subsequent TxA 2 -induced increases in IP 3 generation by both wild type receptors and mutant 2. Pretreatment with U46619 also caused some reduction in IP 3 generation by mutant 1, but this decrease in responsiveness did not reach statistical significance. In order to normalize for differences in the baseline response, data for cells pretreated with U46619 were expressed as the percent response in vehicle treated cells (Table III). When the data are normalized in this fashion, there was significantly less desensitization in cells expressing mutant 1 compared with wild type receptors. These data suggest that desensitization of IP 3 responses is attenuated by the absence of C-terminal serines at positions 321, 322, and 328 of the mouse TP receptor. The similar pattern of desensitization in wild type receptors and mutant 2 also suggests that homologous desensitization of TP receptors is not affected by the absence of the serine at position 339.
Previous studies from this laboratory (16) found that the absence of the C-terminal 22 amino acids of the mouse TP receptor attenuates both agonist-specific and PKC-induced desensitization. Because this truncation mutant and mutant 1 both lack serines at positions 321, 322, and 328, we determined the role of PKC in the attenuated homologous desensitization observed with mutant 1. For these experiments, mesangial cells expressing wild type or mutant TP receptors were incubated with vehicle or 1 M U46619 in the presence or absence of the kinase inhibitor staurosporine (200 nM) (30) prior to washing and rechallenge with 1 M concentrations of the TxA 2 agonist. As seen in Fig. 2, staurosporine attenuated homologous desensitization in clones expressing the wild type TP receptor. In contrast, staurosporine had little effect on subsequent responsiveness of mutant 1 following exposure to a desensitizing stimulus. These data suggest that a staurosporinesensitive kinase contributes to homologous desensitization of wild type, but not mutant, TP receptors.
To further investigate the role of C-terminal domains in PKC-induced desensitization of the TP receptor, clones ex-FIG. 1. Membrane topology of the mouse TP receptor. The proposed membrane topology of the mouse TP receptor as well as the C-terminal amino acids of the mouse, rat, and human TP receptor (9 -11). Putative intracellular domains are rich in hydroxyl amino acids (shown by black circles), which are potential phosphorylation sites for protein kinases. Serines at positions 321, 322, and 328 were changed to either alanine or glycine in mutant 1 (shown by asterisks). The serine at position 339 was changed to glycine in mutant 2 (shown by the arrow head). These hydroxyamino acid residues are highly conserved in the mouse, rat, and human TP receptor. were loaded with 32 P, prior to stimulation with 1 M U46619 and immunoprecipitation of receptor proteins as described under "Materials and Methods." For our initial studies, we transfected mouse mesangial cells with our epitope-tagged receptor; however, despite preclearing of the lysate and stringent wash conditions, we continued to have high backgrounds which made interpretation of the autoradiograms difficult. We therefore changed to HEK293 cells for our immunoprecipitation studies since these cells have been used successfully by other investigators to study direct phosphorylation of eicosanoid receptors (17,31). Using HEK293 cells, we successfully visualized phosphorylated receptor proteins as shown in Fig. 4. Stimulation of transfected HEK293 cells with 1 M U46619 caused phosphorylation of a broad band at Ϸ44 kDa. The band was not seen in cells transfected with the wild type receptor (panel A) or when the immunoprecipitation was performed in the presence of the hemagglutinin peptide, which is the epitope recognized by the 12CA5 antibody (panel B), suggesting that the Ϸ44 kDa band represents the TP receptor.
To determine whether HEK293 cells were a suitable model system for studying TP receptor regulation, we transiently transfected cells with the wild type receptor or mutant 1. Two days following transfection, cells were incubated with either vehicle, 1 M U46619 or 100 nM PDBu, prior to washing and rechallenge with 1 M concentrations of the TxA 2 agonist. As shown in Table IV and Table V, IP 3 generation in vehicletreated cells was increased in mutant 1 compared with wild type receptors, similar to the pattern of TP responsiveness in our mesangial cell clones. Prior treatment with either U46619 or PDBu reduced subsequent TxA 2 -induced increases in IP 3 generation in HEK293 cells expressing wild type receptors or mutant 1. However, when data were normalized for differences in the baseline response, there was significantly less desensitization in cells expressing mutant 1 compared with wild type receptors (Tables IV and V). These data suggest that the pattern of TP receptor responsiveness and desensitization is similar in HEK293 cells transiently transfected with TP receptors and in our mouse mesangial cell clones stably expressing TP receptors.  We next determined the role of PKC in direct phosphorylation of TP receptor proteins. For these studies, cells were incubated with 1 M U46619 in the presence or absence of the specific PKC inhibitor GF109203X (1 M) (32). As seen in Fig.  5, GF109203X inhibited, but did not prevent, phosphorylation of TP receptor proteins. Inhibition of TP receptor phosphorylation by GF109203X suggests that phosphorylation of the mouse TP receptor is mediated to a large extent by PKC. In data not shown, similar results were obtained with the general kinase inhibitor staurosporine (200 nM) (30).
To study the role of C-terminal serines in agonist-induced phosphorylation of the TP receptor, we created an epitopetagged mutant receptor lacking serines at positions 321, 322, and 328 of the mouse TP receptor (ET mutant 1). Cells express-ing epitope-tagged wild type receptors or ET mutant 1 were loaded with 32 P, prior to stimulation with 1 M U46619 and immunoprecipitation of receptor proteins as described under "Materials and Methods." Basal phosphorylation of wild type and mutant receptors was similar, and these data were combined for the densitometric analysis. As seen in Fig. 6, U46619induced phosphorylation of the TP receptor was reduced in cells transfected with ET mutant 1 compared with cells transfected with wild type receptors. The apparent difference in agonist-induced phosphorylation of ET mutant 1 was not due to decreased expression of the receptor mutant, because in experiments performed in parallel, TxA 2 binding was similar in cells transfected with either wild type receptors or ET mutant 1 (325 Ϯ 14 fmol/mg of protein (WT) versus 314 Ϯ 13 fmol/mg of protein (ET mutant 1); p ϭ NS, n ϭ 3 experiments). Thus, the absence of serines 321, 322, and 328 attenuates both agonistinduced phosphorylation and homologous desensitization (Table III) of TP receptors, suggesting that phosphorylation of one or more of these serine residues contributes to agonist-specific desensitization of the mouse TP receptor. Although U46619-

FIG. 4. U46619-induced phosphorylation of the TP receptor.
HEK293 cells were transfected with either wild type TP receptors (WT) or TP receptors that had been tagged with the 12CA5 epitope (ET). Two days following transfection, 32 P-loaded cells were stimulated with 1 M U46619 for 10 min followed by immunoprecipitation of receptor proteins as described under "Materials and Methods." Stimulation with 1 M U46619 enhanced phosphorylation of a broad band at Ϸ44 kDa. The band was not seen in cells transfected with the wild type receptor (panel A) or when the immunoprecipitation was performed in the presence of the HA peptide which is the epitope recognized by the 12CA5 antibody (panel B), suggesting that the Ϸ44-kDa band represents the TP receptor.   5. Effect of GF109203X on U461619-induced phosphorylation of TP receptors. HEK293 cells were transfected with wild type TP receptors that had been tagged with the 12CA5 epitope. Two days following transfection, 32 P-loaded cells were incubated with the PKC inhibitor GF109203X (1 M) or its vehicle for 10 min prior to stimulation with 1 M U46619 for an additional 10 min. Epitope-tagged TP receptors were then isolated by immunoprecipitation as described under "Materials and Methods." GF109203X inhibited phosphorylation of TP receptor proteins. induced phosphorylation of ET mutant 1 was reduced compared with the wild type, there was still a significant increase in phosphorylation compared with its basal phosphorylation state (see Fig. 6). This U46619-induced phosphorylation of ET mutant 1 suggests that the TP receptor can be phosphorylated at residues other than serine 321, 322, or 328.
To study the role of PKC in direct phosphorylation of Cterminal serines, cells expressing epitope-tagged wild type receptors or ET mutant 1 were loaded with 32 P, prior to stimulation with 100 nM PDBu and immunoprecipitation of receptor proteins as described under "Materials and Methods." As in the U46619-induced phosphorylation experiments, basal phosphorylation of wild type and mutant receptors was similar, and these data were combined for the densitometric analysis. As seen in Fig. 7A, phosphorylation of TP receptor proteins was decreased in cells transfected with ET mutant 1 compared with cells transfected with the wild type receptor. Thus, the deletion of serines 321, 322, and 328 of the mouse TP receptor attenuates direct phosphorylation of the receptor protein following activation of PKC.
Prior incubation with 1 M PDBu caused a similar decrease in subsequent TP receptor responsiveness (Fig. 3). This finding indicates that regulation of TP receptor activity by PKC is complex and also involves mechanisms which do not require C-terminal serines at positions 321, 322, and 328. Possible regulatory mechanisms activated by higher levels of PKC stimulation include: 1) phosphorylation of additional domains of the TP receptor or 2) phosphorylation of other substrates such as downstream components of the signaling pathway. We therefore determined if TP receptor phosphorylation is further enhanced by 1 M PDBu. As in the previous experiments, basal phosphorylation of wild type and mutant receptors was similar and the data were combined for the densitometric analysis. As shown in Fig. 7B, phosphorylation of ET mutant 1 by 1 M PDBu was decreased compared with PDBu-induced phospho-rylation of wild type receptors. The level of phosphorylation of either the wild type receptor or ET mutant 1 was not significantly different at the 100 nM and 1 M concentrations of PDBu. These data suggest that increasing PDBu from 100 nM to 1 M does not result in significant increases in TP receptor phosphorylation.

DISCUSSION
TxA 2 plays a key role in diseases affecting the heart, lungs, kidneys, and peripheral vascular system (1)(2)(3). Its effects are regulated by both PKC-dependent and PKC-independent mechanisms (4,6). Previous studies have suggested that Cterminal domains are required for regulation of the TP receptor by PKC (17), perhaps by direct phosphorylation of the receptor protein (5,17). In the present study, we investigated the role of the C-terminal serines in the rapid regulation of TP receptor responsiveness. We found that TP receptors lacking serines at positions 321, 322, and 328 in the C terminus are resistant to both U46619-and PKC-induced desensitization. Attenuated desensitization of the receptor mutant is associated with decreased phosphorylation of receptor proteins compared with wild type receptors. Taken together, these data suggest that both agonist-specific and PKC-induced desensitization of the TP receptor are mediated, at least in part, by phosphorylation of C-terminal hydroxyamino acids.
The absence of the C-terminal serines at positions 321, 322, and 328 inhibited, but did not prevent, U46619-induced phosphorylation of the TP receptor. These findings indicate that the TP receptor is phosphorylated in an agonist-dependent fashion at residues other than serines 321, 322, and 328. It is likely that this agonist-dependent phosphorylation of mutant 1 is mediated by either receptor specific kinases or general kinase systems such PKC. With regard to the latter hypothesis, the specific PKC inhibitor GF109203X markedly attenuated, but did not prevent, U46619-induced phosphorylation of the TP receptor. Similar results were found using the more general kinase inhibitor staurosporine. Thus, the TP receptor is phosphorylated in an agonist-dependent manner by a staurosporine-and GF109203X-resistant kinase. These finding are consistent with the notion that receptor specific kinases may contribute to desensitization of the mouse TP receptor. This family of kinases has been shown to play a major key role in regulating receptor responsiveness in other receptor systems (7,8). Although the consensus motif for phosphorylation by receptor-specific kinases is not fully elucidated, serine or threonine residues flanked by acid amino acids (D/ES/T or D/EXS/T) appear to be preferred substrates for G protein-coupled receptor kinases (33). Motifs of this configuration are found in the third intracellular loop of the mouse, rat, and human TP receptor (9 -11) and might play a role in regulating TP receptor responsiveness. Indeed, preliminary studies from this laboratory suggest that phosphorylation of the TP receptor by G protein-coupled receptor kinases may contribute to TP receptor regulation (34).
The absence of serines 321, 322, and 328 prevented desensitization of TP receptors by 100 nM PDBu. In contrast, 1 M PDBu caused desensitization of wild type and mutant receptors to a similar extent. One possibility to explain this observation is that 1 M PDBu enhances phosphorylation of domains of the TP receptor outside of the C-terminal 22 amino acids. Indeed, hydroxyamino acids in the proximal C terminus as well as the second and third intracellular loops have favorable motifs for phosphorylation by PKC (11,35). Moreover, the absence of serines 321, 322, and 328 inhibits, but does not prevent, phosphorylation of the TP receptor by 100 nM PDBu. We therefore determined if treatment with 1 M PDBu further enhances TP receptor phosphorylation compared with 100 nM PDBu. In these experiments, increasing PDBu from 100 nM to 1 M did not significantly enhance phosphorylation of either wild type or mutant TP receptors. While we cannot exclude the possibility that small increases in receptor phosphorylation were not detected by our assay system, it seems more likely that 1 M PDBu caused phosphorylation of other substrates leading to attenuation of TP receptor signaling. In this regard, PKC might: 1) phosphorylate and inactivate other downstream components of the signaling cascade or 2) phosphorylate and activate other regulatory mechanisms. Evidence for each of these possibilities has been demonstrated in other receptor systems. For example, activation of PKC by phorbol esters has been shown to phosphorylate and inactivate phospholipase-␤ (36). Moreover, G protein-coupled receptor kinase 2 can be phosphorylated by PKC in vitro resulting in 2-fold activation of the enzyme toward its substrate (37). Thus, PKC might regulate multiple components of the signaling cascade.
U46619-stimulated IP 3 generation was higher in mutant 1 compared with either wild type receptors or mutant 2. A possible explanation for this enhanced PI hydrolysis is that general kinase systems may inhibit TP receptor responsiveness under basal conditions. The absence of C-terminal serines might attenuate this inhibition by these kinase systems. In support of this hypothesis, we previously found that PKC inhibitors tended to enhance TP receptor responsiveness in cells that express an endogenous receptor for TxA 2 (6). Alternatively, the absence of regulatory serines may result in more sustained PI hydrolysis in mutant 1 compared with wild type receptors. In this regard, agonist-induced elevations in IP 3 levels are prolonged in other receptor systems following removal of C-terminal regulatory domains (38). Further studies will be needed to determine if either of these possibilities contributes to enhanced PI hydrolysis by TP receptors lacking serines 321, 322, and 328.
Although the PKC inhibitor GF109203X did not completely prevent U46619-induced phosphorylation of the mouse TP receptor, agonist-induced phosphorylation was markedly reduced in the presence of the PKC inhibitor. In contrast to our findings, Habib et al. (17) found that phosphorylation of human TP receptors was only modestly affected by GF109203X in HEK293 cells expressing TP receptor isoforms. Possibilities to explain these differing results include: 1) species specific differences in TP receptor regulation or 2) methodological differences between the studies. With regard to the latter, Habib et al. (17) immunoprecipitated human TP receptors using isoform-specific anti-peptide antibodies from cells stably expressing high levels of TP receptors. In contrast, we studied phosphorylation of 12CA5-tagged TP receptors in a transient transfection system expressing more modest levels of TP receptor proteins. While it is not known if these methodologic differences can account for the differing results, both these studies were performed in artificial expression systems that might affect normal regulatory mechanisms. It will therefore be important to determine whether results using overexpression systems are generalizable to cells that endogenously express receptors for TxA 2 .
Mutating serine at position 339 did not affect signaling or desensitization of the mouse TP receptor. While we did not directly study the effects of this mutation on receptor phosphorylation, serine 339 does not have a favorable motif for phosphorylation by PKC, protein kinase A, protein kinase G, or calcium-calmodulin-dependent protein kinases (34). In addition, this amino acid residue is not flanked by acid amino acids, which tend to be favored sites for phosphorylation by G proteincoupled receptor kinases (33). It therefore seems unlikely that serine 339 is an important target for phosphorylation by protein kinases.
In summary, the present studies indicate that the absence of serines at positions 321, 322, and 328 inhibits both U46619and PKC-induced desensitization of the TP receptor. Agonistspecific desensitization is associated with enhanced phosphorylation of receptor proteins. Both U46619-induced desensitization and phosphorylation are attenuated by either PKC inhibitors or the absence of serines 321, 322, and 328. PKCinduced desensitization caused by 100 nM PDBu is also associated with increased phosphorylation of receptor proteins. This PKC-mediated phosphorylation was inhibited by deleting serines, 321, 322, and 328. Finally, increasing the concentration of PDBu from 100 nM to 1 M PDBu reduced responsiveness of wild type receptors and mutant receptors lacking serines 321, 322, and 328 without enhancing phosphorylation of either of the receptor proteins. These data suggest that 1) phosphorylation of C-terminal serines contributes to both agonist-specific and PKC-dependent desensitization of the TP receptor, and 2) PKC-induced desensitization of the TP receptor is complex and involves both phosphorylation of C-terminal serines as well as mechanisms that may not require direct phosphorylation of receptor proteins.