Investigation of the Mechanisms of G Protein: Effector Coupling by the Human and Mouse Prostacyclin Receptors

We recently identified a novel mechanism explaining how the mouse (m) prostacyclin receptor (IP) couples to Gαs, Gαi, and Gαq(Lawler, O. A., Miggin, S. M., and Kinsella, B. T. (2001) J. Biol. Chem. 276, 33596–33607) whereby mIP coupling to Gαi and Gαq is dependent on its initial coupling to Gαs and subsequent phosphorylation by cAMP-dependent protein kinase A (PKA) on Ser357. In the current study, the generality of that mechanism was investigated by examining the G protein coupling specificity of the human (h) IP. The hIP efficiently coupled to Gαs/adenylyl cyclase and to Gαq/phospholipase C activation but failed to couple to Gαi. Coupling of the hIP to Gαq, or indeed to Gαs or Gαi, was unaffected by the PKA or protein kinase C (PKC) inhibitors H-89 and GF 109203X, respectively. Thus, mIP and hIP exhibit essential differences in their coupling to Gαi and in their dependence on PKA in regulating their coupling to Gαq. Analysis of their primary sequences revealed that the critical PKA phosphorylation site within the mIP, at Ser357, is replaced by a PKC site within the hIP, at Ser328. Conversion of the PKC site of the hIP to a PKA site generated hIPQL325,326RP that efficiently coupled to Gαs and to Gαi and Gαq; coupling of hIPQL325,326RP to Gαi but not to Gαs or Gαq was inhibited by H-89. Abolition of the PKC site of the hIP generated hIPS328A that efficiently coupled to Gαs and Gαq but failed to couple to Gαi. Finally, conversion of the PKA site at Ser357 within the mIP to a PKC site generated mIPRP354,355QL that efficiently coupled to Gαs but not to Gαi or Gαq. Collectively, our data highlight critical differences in signaling by the mIP and hIP that are regulated by their differential phosphorylation by PKA and PKC together with contextual sequence differences surrounding those sites.

The prostanoid prostacyclin (prostaglandin I 2 ) is the major product of arachidonic acid metabolism in the vascular endothelium (1). It plays a key role in the local control of vascular hemostasis, acting as a potent inhibitor of platelet aggregation and as a vasodilator (2), and exhibits pro-inflammatory and anti-proliferative properties in vitro (3,4). The actions of prostacyclin generally counteract those of thromboxane A 2 , and the relative levels of these two prostanoids regulate platelet/endo-thelium/vascular smooth muscle interactions (5). Prostacyclin is abundantly produced during cardiac ischemia/reperfusion, conferring a cytoprotective effect (6), and, from studies carried out in prostacyclin receptor-deficient mice, is known to exert a protective effect on cardiomyocytes independent of its effects on platelets and neutrophils (7). Perturbations in prostacyclin and/or prostacyclin receptor signaling have been implicated in the pathogenesis of conditions such as ischemic heart disease, atherosclerosis, renal failure, and systemic and pregnancyinduced hypertension (5, 8 -10).
Prostacyclin signals through activation of its cell surface G protein-coupled receptor (GPCR), 1 termed the prostacyclin receptor or IP (11). The IP is subject to post-translational modifications such as N-glycosylation and phosphorylation that play central roles in regulating receptor function (12,13). The IP appears to be somewhat unique among GPCRs in that it is isoprenylated (14). Although isoprenylation is not required for ligand binding or membrane localization, it is required for efficient IP:G protein coupling and may regulate agonistinduced IP internalization (14 -16).
A number of independent studies have demonstrated that, although the IP primarily couples to activation of adenylyl cyclase, mediating prostacyclin inhibition of platelet aggregation and vascular tone (5,17), it may also regulate a number of other effector systems, perhaps in a tissue-and/or speciesspecific manner. In the rat medullary thick ascending limb, IP couples to G i , but not to G s , suggesting inhibition rather than activation of adenylyl cyclase (18). Iloprost, a stable prostacyclin analogue, can activate calcium-activated potassium channels (K Ca channels) in rat small arteries (19) and can stimulate the opening of ATP-sensitive K ϩ channels resulting in hyperpolarization and relaxation of canine carotid artery (20). The cloned mouse (m) and human (h) IPs couple to both G S and to G q , leading to phospholipase C (PLC) activation and to mobilization of intracellular calcium (12)(13)(14)(15)(16)21). More recently, we have demonstrated that, in addition to its coupling to G s , the mIP also couples to both G i , leading to inhibition of adenylyl cyclase, and to G q , leading to PLC activation, through a novel G protein switching mechanism (22). In this mechanism, mIP coupling to both G i and G q are dependent upon its initial coupling to G s /adenylyl cyclase activation and consequent cAMPdependent protein kinase A (PKA) phosphorylation of Ser 357 within the carboxyl-terminal cytoplasmic (C) tail of the mIP, thereby switching mIP coupling from G S to G i and to G q signaling (22).
Thus, it is evident that the IP is capable of coupling to multiple G protein/effectors in a species-and/or tissue-specific manner. However, the molecular basis of this species-specific coupling has not been investigated in detail. Thus, in the present study, in view of the central roles of prostacyclin within the human vasculature, we sought to define the G protein coupling specificity of the hIP. Moreover, given the central role of PKAmediated phosphorylation of the mIP in determining its G protein specificity and mechanisms of signaling, we sought to investigate whether the hIP undergoes a similar G protein switching mechanism accounting for its patterns of G protein coupling and intracellular signaling. Our data highlight critical differences in the signaling of the mIP and hIP that are regulated by their differential phosphorylation by the second messenger-regulated kinases PKA and PKC together with surrounding contextual sequence differences within those kinase recognition sites, thereby accounting for essential species-dependent differences in signaling by the mIP and hIP.

Methods
Site-directed Mutagenesis of the hIP and mIP-All site-directed mutagenesis procedures were performed using the Stratagene QuikChange site-directed mutagenesis kit. Conversion of Ser 328 of the hIP to Ala 328 , herein designated hIP S328A , was performed using pHM-hIP (15) as template and mutator oligonucleotides: 5Ј-CTT TCC CAG CTC GCC GCC GGG AGG AGG GAC C-3Ј (sense primer) and 5Ј-G GTC CCT CCT CCC GGC GGC GAG CTG GGA AAG-3Ј (antisense primer; the sequence complimentary to mutator Ser (TCC) to Ala (GCC) codon is underlined). Conversion of Gln 325 and Leu 326 of the hIP to Arg 325 and Pro 326 , herein designated hIP QL325,326RP , was performed using pHM-hIP as template and oligonucleotides: 5Ј-CAG ACA CCC CTT TCC CGG CCC GCC TCC GGG AGG AG-3Ј (sense primer) and 5Ј-CT CCT CCC GGA GGC GGG CCG GGA AAG GGG TGT CTG-3Ј (antisense primer; the sequence complimentary to mutators Gln and Leu (CAG CTC) to Arg and Pro (CGG CCC) codons are underlined). Conversion of Arg 354 and Pro 355 of the mIP to Gln 354 and Leu 355 , herein designated mIP RP354,355QL , was performed using pHM-mIP (14) as template and oligonucleotides: 5Ј-CAG GCG CCC CTT TCC CAA CTT GCA TCG GGG AGA AG-3Ј (sense primer) and 5Ј-CT TCT CCC CGA TCG AAG TTG GGA AAG GGG CGC CTG-3Ј (antisense primer; the sequence complimentary to mutators Arg and Pro (AGA CCT) to Gln and Leu (CAA CTT) codon are underlined). All resulting plasmids pHM-hIP S328A , pHM-hIP QL325,326RP , and pHM-mIP RP354,355QL and mutations were verified by automated double-stranded DNA sequencing and encode hemagglutinin (HA) epitope-tagged forms of hIP S328A , hIP QL325,326RP , and mIP RP354,355QL , respectively.
HEK 293 cells were transfected with 10 g of pADVA and 25 g of pCMV-or pHM-based vectors using the calcium phosphate/DNA coprecipitation procedure (24). For transient transfections, cells were harvested 48 h after transfection. HEK.mIP and HEK.hIP cell lines overexpressing HA epitope-tagged forms of the wild type mIP and hIP, respectively, have been previously described (22,15). To create the HEK.mIP RP354,355QL , HEK.hIP QL325,326RP , and the HEK.hIP S328A stable cell lines, HEK 293 cells were transfected with 10 g of ScaI-linearized pADVA plus 25 g of PvuI-linearized pHM:mIP RP354,355QL , pHM: hIP QL325,326RP , or pHM:hIP S328A , respectively. Forty eight hours post transfection, G418 (0.8 mg/ml) selection was applied, and after ϳ21 days, G418-resistant colonies were selected and individual pure clonal stable cell lines/isolates were examined for IP expression by radioligand binding.
Preparation of Platelets-Blood was drawn via venipuncture from normal human volunteers, who had not taken any medication for at least 10 days, into syringes containing indomethacin (10 M) and 3.8% sodium citrate (9:1 v/v) (final concentration, 0.38% sodium citrate). The blood was centrifuged for 10 min at 160 ϫ g; the platelet-rich plasma was removed and recentrifuged for 10 min at 160 ϫ g to remove contaminating red blood cells. Washed platelets were prepared, following the addition of 5 mM EDTA as anti-coagulant, by centrifuging the platelet-rich plasma at 900 ϫ g for 15 min and resuspending the platelets in a modified Tyrode's albumin buffer (26) containing 10 M indomethacin.
Measurement of cAMP-cAMP assays were carried out as previously described (14). Briefly, cells were harvested by scraping and washed three times in ice-cold PBS. Cells (ϳ1-2 ϫ 10 6 cells) or washed human platelets (1.85 ϫ 10 6 platelets/l) were resuspended in 200 l of HEPES-buffered saline (HBS; 140 mM NaCl, 4.7 mM KCl, 2.2 mM CaCl 2 , 1.2 mM KH 2 PO 4 , 11 mM glucose, 15 mM HEPES-NaOH, pH 7.4) containing 1 mM 3-isobutyl-1-methylxanthine and were preincubated at 37°C for 10 min. Thereafter, ligands (50 l) were added and cells were stimulated at 37°C for 10 min in the presence of the ligand (1 M cicaprost, 10 M forskolin, or 1 M cicaprost plus 10 M forskolin). For concentration response studies, cells were stimulated with 10 Ϫ12 to 10 Ϫ6 M cicaprost. As controls, cells were incubated in the presence of 50 l of HBS in the absence of ligand. To investigate the effect of pertussis toxin (PTx), cells were preincubated with PTx (50 ng/ml) for 16 h prior to stimulation with cicaprost plus forskolin. To investigate the effect of protein kinase inhibitors on cAMP generation, cells were preincubated in the presence of GF 109203X (50 nM), H-89 (10 M), or vehicle (HBS) at 37°C for 10 min prior to stimulation with cicaprost plus forskolin. In separate experiments, to examine the effect of co-transfection of G␣ s on cAMP generation, HEK 293, HEK.hIP, HEK.hIP QL325,326RP , HEK.hIP S328A , and HEK.mIP RP354,355QL cells were transiently co-transfected with pCMV-G␣ s (25 g/10-cm dish) plus pADVA (10 g/10-cm dish). After 48 h, cells were harvested and stimulated with 1 M cicaprost or vehicle (HBS).
In each case, cAMP reactions were terminated by heat inactivation at 100°C for 5 min and the level of cAMP produced was quantified using the cAMP binding protein assay (27). Levels of cAMP produced by ligand-treated cells over basal stimulation, determined in the presence of HBS, were expressed as pmol cAMP/mg cell protein Ϯ standard error of the mean (S.E.). Results are expressed as -fold stimulation relative to basal (-fold increase Ϯ S.E.). Data were analyzed using the unpaired Student's t test. p values of less than or equal to 0.05 were considered to indicate a statistically significant difference.
Measurement of IP 3 Levels-Intracellular IP 3 levels were measured as previously described (28). Briefly, cells were transiently co-transfected with pCMV-G␣ q (25 g/10-cm dish) plus pADVA (10 g/10-cm dish). After 48 h, cells were harvested, washed twice in ice-cold PBS and were then resuspended at ϳ5 ϫ 10 6 cells/ml in HBS containing 10 mM LiCl. Cells (200 l) were then preincubated at 37°C for 10 min. To investigate the effect of the protein kinase inhibitors, H-89 (10 M) or GF 109203X (50 nM) were added and the cells were incubated for 2 min at 37°C, 5% CO 2 prior to stimulation with cicaprost. Thereafter, cells were stimulated for 2 min at 37°C in the presence of cicaprost (1 M) or, for concentration response studies, were stimulated with cicaprost 10 Ϫ6 to 10 Ϫ12 M. To determine basal IP 3 levels, cells were incubated in the presence of an equivalent volume (50 l) of the vehicle HBS. The IP 3 levels produced were determined using the IP 3 binding protein assay (28). Levels of IP 3 produced by ligand-stimulated cells over basal stimulation, in the presence of HBS, were expressed in picomoles of IP 3 /mg of cell protein Ϯ standard error (pmol/mg Ϯ S.E.), and results are presented as -fold stimulation over basal (-fold increase Ϯ S.E.). The data presented are representative of four independent experiments, each performed in duplicate.
Measurement of Intracellular [Ca 2ϩ ] Mobilization-Measurements of [Ca 2ϩ ] i mobilization in Fura2/AM-preloaded cells was carried out essentially as previously described (24). Where appropriate, the PKA Measurement of Agonist-mediated IP Phosphorylation in Whole Cells-Agonist-mediated IP phosphorylations in whole HEK.hIP, HEK.hIP QL325,326RP , HEK.hIP S328A , HEK.mIP, and HEK. mIP RP354,355QL cells were carried out essentially as previously described (22). Briefly, cells were washed once in phosphate-free Dulbecco's modified Eagle's medium containing 10% dialyzed FBS and were metabolically labeled for 1 h in the same media (1.5 ml per 10-cm dish) containing 100 Ci/ml [ 32 P]orthophosphate (8000 -9000 Ci/mmol) at 37°C, 5% CO 2 . Where appropriate, H-89 (10 M), GF 109203X (50 nM), or the vehicle HEPES-buffered saline (HBS) was added for the duration of the labeling period. Thereafter, 1 M cicaprost or vehicle HBS was added for 10 min at 37°C, 5% CO 2 . Metabolic labeling of cells was terminated by transferring the dishes to ice and aspiration of the medium. Thereafter, cells were quickly washed once in ice-cold PBS (2 ml/dish) and were lysed with 0.6 ml of radioimmune precipitation buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40 (v/v), 0.5% sodium deoxycholate (w/v), 0.1% SDS (w/v) containing 10 mM sodium fluoride, 25 mM sodium pyrophosphate, 10 mM ATP, 1 g/ml leupeptin, 10 g/ml soybean trypsin inhibitor, 1 mM benzamidine hydrochloride, 0.5 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate). Following 15-min incubation on ice, cells were harvested and disrupted by sequentially passing through hypodermic needles of decreasing bore size (20-, 21-, 23-, and 26-gauge), and soluble cell lysates were harvested by centrifugation for 15 min at 13,000 ϫ g at room temperature. HA epitope-tagged IP receptors were immunoprecipitated using the anti-HA antibody (101R, 1:300 dilution) at room temperature for 2 h followed by the addition of 10 l of protein G-Sepharose 4B (Sigma) and further incubation at room temperature for 1 h. Immune complexes were collected by centrifugation at 13,000 ϫ g at room temperature for 5 min and were washed three times in 0.3 ml of radioimmune precipitation buffer and were finally resuspended in 1ϫ solubilization buffer , 50 mM Tris-HCl, pH 6.8, 60 l). Samples were boiled for 5 min and then loaded onto 10% polyacrylamide gels, analyzed by SDS-PAGE, and thereafter electroblotted onto PVDF mem-branes. Electroblots were then exposed to Eastman Kodak Co. X-Omat XAR film to detect 32 P-labeled proteins. Thereafter, blots were subject to PhosphorImage analysis, and the intensities of phosphorylation relative to basal phosphorylation were determined and were expressed in arbitrary units of intensity relative to basal levels. In parallel experiments, cells were incubated under identical conditions in the absence of [ 32 P]orthophosphate; hIP, hIP QL325,326RP , hIP S328A , mIP, and mIP RP354,355QL receptors were immunoprecipitated, and immunoblots were screened using the anti-HA 3F10 horseradish peroxidase conjugate (1:500) to check for quantitative recovery of each receptor type. Immunoreactive proteins were visualized using the chemiluminescence detection system (22).
Data Analyses-Statistical analysis was carried out using the unpaired Student's t test using the GraphPad Prism version 2.0 program (GraphPad Software Inc., San Diego, CA). p values of less than or equal to 0.05 were considered to indicate a statistically significant difference. Amino acid sequence alignments were carried out using the ClustalW software (29), where sequences were aligned to show maximum homology.

Human Prostacyclin Receptor:G protein Coupling to Adenylyl
Cyclase and to Phospholipase C-Previous studies have demonstrated that, although the mIP couples to G␣ s , mediating increases in cAMP generation, it may also couple to G␣ i , leading to inhibition of adenylyl cyclase and cAMP generation and to G␣ q -mediated PLC activation through a novel PKA-dependent mechanism (22). In the present study, we sought to investigate the G protein coupling specificity of the hIP and the mechanism of its regulation thereof.
To this end, the effect of the selective agonist cicaprost on IP-mediated cAMP generation in HEK.hIP cells (16), which stably overexpress the human (h)IP (Table I), was investigated and compared with that which occurred in HEK.mIP cells (14) stably overexpressing the mouse (m) IP (Table I). Stimulation of HEK.mIP cells (Fig. 1A, p Ͻ 0.001) and HEK.hIP cells (Fig.  1B, p Ͻ 0.001) with cicaprost each resulted in significant increases in cAMP generation consistent with coupling of both the mIP and the hIP to G␣ s . Although stimulation of both cell types with forskolin resulted in significant, receptor-independent increases in cAMP generation (Fig. 1, A and B, p Ͻ 0.001), cicaprost significantly inhibited forskolin-induced cAMP generation in HEK.mIP cells, consistent with mIP coupling to G␣ i (Fig. 1A, p Ͻ 0.001). In contrast, in HEK.hIP cells, cicaprost failed to inhibit (Fig. 1B, p Ն 0.05) but, rather, augmented (Fig.  1B, p Ͻ 0.05) forskolin-induced cAMP generation indicative of hIP coupling to G␣ s , but not to G␣ i . Moreover, cicaprost showed efficient concentration-dependent increases in cAMP generation in HEK.hIP cells, in the absence or presence of forskolin, throughout the range of cicaprost concentrations used (10 Ϫ6 to 10 Ϫ12 M) but failed to reduce forskolin-induced cAMP generation (data not shown). Furthermore, stimulation of hIP endogenously expressed in human erythroleukemic 92.1.7 (HEL) cells and in human platelets exhibited efficient cicaprostinduced rises in cAMP generation in both the presence and absence of forskolin, consistent with coupling of the hIP to G␣ s , but not to G␣ i or to inhibition of adenylyl cyclase activity (Fig.  1, C and D, respectively). Whereas mIP coupling to G␣ i was inhibited by the PKA inhibitor H-89 (Fig. 1A) (22), the ability of the hIP expressed in HEL cells or in human platelets (data not shown) or in HEK.hIP cells to couple to G␣ s or to G␣ i was unaffected by H-89 (p ϭ 0.94; Fig. 1E) or by the PKC inhibitor GF 109203X (p ϭ 0.64, data not shown). These data indicate that, although the mIP couples to both G␣ s and G␣ i through a PKA-dependent mechanism, the hIP couples to G␣ s but not to G␣ i .
To investigate the ability of the hIP to couple to G␣ q -mediated PLC activation, cicaprost-induced IP 3 generation and concomitant increases in [Ca 2ϩ ] i mobilization were investigated and were compared with that of the mIP. Stimulations of HEK.hIP, HEL, and HEK.mIP cells each resulted in efficient cicaprost-induced rises in [Ca 2ϩ ] i mobilization (Fig. 2, A-C, respectively), and increases in IP 3 generation ( affected by H-89 or by GF 109203X. Thus, although the hIP and mIP both couple to Gq-mediated PLC activation, they display essential differences in their dependence on PKA and, hence, in their mechanism of Gq coupling whereby hIP independently couples to G␣ q , whereas mIP coupling is dependent on its PKA phosphorylation and consequent switching from G␣ s to G␣ q . To fully exclude the possibility that cicaprost-induced [Ca 2ϩ ] i mobilization might be mediated through G i -derived G␤␥ subunits, the effect of PTx and the G␤␥ sequestrant peptide, ␤ARK1 459 -689 (23) (22). Overexpression of the carboxyl-terminal residues of ␤ARK1 was confirmed by Western blot analysis using anti-GRK 2 (Santa Cruz Biotechnology, As #C-15, data not shown).
Role of Second Messenger Kinases in hIP:G protein Coupling-We have previously established that mIP coupling to both G␣ i and G␣ q is dependent on its initial coupling to G␣ s and subsequent cAMP-dependent PKA phosphorylation, where Ser 357 within the C-tail region of mIP was identified as the target residue for PKA phosphorylation (22). Optimal alignment of the primary sequences of the C-tail regions of the mIP (21) and the hIP (30,31) and their computational analyses for putative phosphorylation sites (32) revealed that the previously identified consensus target site for PKA phosphorylation occurring at Ser 357 within the mIP, with the sequence RPAS 357 GRR, is replaced by a consensus target site for PKC phosphorylation within the hIP, with the sequence QLAS 328 GRR (Fig. 4). Thus, in view of the differential G protein coupling specificities of the mIP and the hIP to G␣ i and the differential dependence on PKA in regulating their PLC activation, we sought to determine whether alteration of the consensus PKC recognition sequence of the hIP surrounding Ser 328 may impact on its G protein coupling specificity and intracellular signaling. Thus, site-directed mutagenesis of hIP was performed to generate hIP QL325,326RP , a variant of hIP whereby the critical residues Gln 325 -Leu 326 , within the consensus PKC phosphorylation site (QLAS 328 GRR), were mutated to Arg 325 -Pro 326 , thereby generating a putative PKA consensus site (RPAS 328 GRR), identical to that of the mIP in this region.
Initial characterization of HEK.hIP QL325,326RP cells, recombinant HEK 293 cells stably overexpressing hIP QL325,326RP by saturation radioligand binding confirmed high level receptor expression (Table I). Consistent with its ability to couple to G␣ s , the hIP QL325,326RP exhibited concentration-dependent increases in cicaprost-induced cAMP generation (Fig. 5A, p Ͼ 0.05), and transient co-transfection of HEK.hIP QL325,326RP cells with pCMV-G␣ s resulted in a significant augmentation (1.70fold, p Ͻ 0.05) in cAMP generation. In contrast to that of the hIP, cicaprost significantly inhibited forskolin-induced cAMP generation in HEK.hIP QL325,326RP cells (Fig. 5B), indicative of hIP QL325,326RP coupling to G␣ i . Furthermore, pretreatment of HEK.hIP QL325,326RP cells with PTx abolished hIP QL325,326RP coupling to G␣ i (Fig. 5B). Although preincubation of HEK.hIP QL325,326RP cells with H-89 had no effect on cicaprostinduced cAMP generation, it blocked hIP QL325,326RP coupling to G␣ i and to inhibition of adenylyl cyclase (Fig. 5C). The PKC inhibitor GF 109203X had no effect on hIP QL325,326RP coupling to G␣ s or to G␣ i (data not shown).
Thereafter, the ability of hIP QL325,326RP to couple to G␣ q and to PLC activation was investigated. Stimulation of HEK.hIP QL325,326RP cells with cicaprost led to significant increases in IP 3 generation (Fig. 5D, p Ͻ 0.05) and [Ca 2ϩ ] i mobilization (Fig. 5E) at levels that were not significantly different from those of the hIP. Moreover, neither hIP QL325,326RP -mediated IP 3 generation nor [Ca 2ϩ ] i mobilization were affected by preincubation with H-89 (Fig. 5, D and E, p Ͼ 0.61) or GF 109203X (Fig. 5D and data not shown, p Ͼ  0.75). Thus, coupling of hIP QL325,326RP to G␣ q and PLC activation is independent of PKA and PKC. Taken together, these data suggest that by conversion of a putative PKC phosphorylation site within the hIP to a putative PKA site within the hIP QL325,326RP facilitates hIP coupling to PTx-sensitive G␣ i but does not influence its ability to couple to G␣ s or to G␣ q .
To further explore the essential role of PKA in regulating hIP, or more specifically hIP QL325,326RP , coupling to G␣ i , sitedirected mutagenesis was used to generate hIP S328A , a variant of hIP whereby the critical phospho-targeted residue Ser 328 was converted to Ala 328 , thereby destroying the putative phosphorylation site. Initial characterization of HEK.hIP S328A cells, recombinant HEK 293 cells stably overexpressing hIP S328A , by saturation radioligand binding confirmed high level receptor expression (Table I). Consistent with its ability to couple to G␣ s , hIP S328A exhibited efficient, concentration-dependent increases in cAMP in response to cicaprost stimulation (Fig. 6A), which were not significantly different from that obtained for the hIP (Fig. 1C). In addition, transient co-transfection of HEK.hIP S328A cells with pCMV-G␣ s significantly augmented cicaprost-induced cAMP generation by 2.07-fold (p Ͻ 0.005). In contrast to that of hIP QL325,326RP , cicaprost did not significantly inhibit, but rather augmented, forskolin-induced cAMP generation in HEK.hIP S328A cells (Fig. 6B), indicative of hIP S328A coupling to G␣ s but not G␣ i . Moreover, neither PTx nor H-89 affected cAMP generation by HEK.hIP S328A cells (Fig. 6, B and C, p Ͼ 0.28).
Next, the ability of the hIP S328A to couple to G␣ q and to PLC activation was investigated. Stimulation of HEK.hIP S328A cells with cicaprost led to significant increases in IP 3 generation (Fig. 6D, p Ͻ 0.001) and [Ca 2ϩ ] i mobilization (Fig. 6E) that were not significantly different from those of the wild type hIP (Figs. 2 and 3). Moreover, neither IP 3 generation nor [Ca 2ϩ ] i mobilization were affected by preincubation with H-89 (Fig. 6, D and E, p Ͼ 0.05) or GF 109203X (Fig. 6D and data not shown, p Ͼ 0.85). Thus, taken together, these data suggest that the hIP S328A can couple to G␣ s and to G␣ q through a PKA-and PKC-independent mechanism but that hIP S328A may not couple to G␣ i .
Whole cell phosphorylations, through metabolic labeling studies, established that the hIP underwent cicaprost-induced phosphorylation as evidenced by the detection of a broad radiolabeled band between the 46-and 66-kDa molecular size markers in the immunoprecipitates from HEK.hIP cells (Fig. 7A, lane 2) but not from vehicle-treated HEK.hIP cells or from non-transfected HEK 283 cells (Fig. 7A, lanes 1 and 5, respectively). Consistent with previous reports (12,13), cicaprostinduced hIP phosphorylation was unaffected by H-89 but was inhibited by the PKC inhibitor GF 109203X (Fig. 7A, lanes 3  and 4, respectively). Although the hIP QL325,326RP underwent cicaprost-induced phosphorylation, it was unaffected by GF 109203X but was inhibited by the PKA inhibitor H-89 (Fig. 7B). FIG. 4. Alignment of human and mouse prostacyclin receptor amino acid sequences. The deduced amino acid sequences of the carboxyl-terminal cytoplasmic (C) tail regions of the hIP (residues 299 -386) and mIP (residues 328 -417), aligned to show maximum homology using the ClustalW software (29), are shown. The consensus PKC phosphorylation site within the hIP is underlined, and the putative phosphotarget residue Ser 328 is highlighted in boldface (32). The consensus PKA phosphorylation site within the hIP is underlined, and the putative phospho-target residue Ser 357 is highlighted in boldface (32). Throughout the alignment, gaps, indicated by the dash symbol (-), were inserted to optimize the alignment; identical amino acids are indicated by an asterisk; conservative substitutions are represented by a colon, and semiconservative substitutions are indicated by a period. Sequences for the hIP and mIP are based on published sequences (21,30,31).
On the other hand, the hIP S328A did not undergo significant cicaprost-induced phosphorylation (Fig. 7C). The identities of the broad radiolabeled band between 46 and 66 kDa were confirmed to be those of the immunoprecipitated HA-tagged hIP and its variant hIP QL325,326RP in parallel immunoprecipitations whereby efficient, quantitative recovery of each receptor type was detected by Western blot analysis using the anti-HA 3F10 peroxidase antibody (Fig. 7D, lanes 1 and 2,  respectively). Moreover, quantitative recovery of hIP S328A from HEK.hIP S328A cells confirmed that the absence of significant phosphorylation of hIP S328A was not attributed to reduced receptor expression or due to reduced immunoprecipitation (Fig.  7D, lane 3).
Role of Second Messenger Kinases in mIP:G protein Coupling-As previously stated, coupling of the mIP to both G␣ i and to G␣ q is dependent on its initial coupling to G␣ s and consequent PKA phosphorylation of mIP at Ser 357 thereby switching mIP coupling from G␣ s coupling to G␣ i and G␣ q coupling (22). In the present study, we sought to fully explore the role of PKA in mediating mIP switching by investigating whether alteration of the consensus PKA target site surrounding Ser 357 to a PKC target site would alter the G protein coupling specificity of the mIP. Thus, mIP RP354,355QL a variant of mIP was generated whereby Arg 354 -Pro 355 , within the PKA consensus sequence RPAS 357 GRR, were converted to Gln 354 -Leu 355 , to generate a putative PKC consensus sequence QLAS 357 GRR (32).
Next, the ability of the mIP RP354,355QL to couple to G␣ q and PLC activation was investigated and was compared with the mIP. Unlike that of the mIP, stimulation of HEK. mIP RP354,355QL cells with cicaprost did not result in significant increases in cicaprost-induced IP 3 generation (Fig. 8E, p Ͼ 0.32) or in [Ca 2ϩ ] i mobilization (Fig. 8F, p Ͼ 0.50). Moreover, neither H-89 nor GF 109203X affected mIP RP354,355QL -mediated IP 3 generation (Fig. 8E) or [Ca 2ϩ ] i mobilization (data not shown) but inhibited mIP-mediated Gq coupling and PLC activation (22). Whole cell phosphorylation studies established that, unlike the mIP (22), the mIP RP354,355QL did not undergo PKA-or PKC-induced phosphorylation in response to cicaprost stimulation; despite this, high levels of expression of mIP RP354,355QL were confirmed by radioligand binding studies (Table I) and by the efficient, quantitative recovery of the HA-tagged mIP RP354,355QL in the immunoprecipitates from HEK.mIP RP354,355QL cells but not from control HEK 293 cells (Fig. 8G).
Thus, taken together, these data highlight the essential role of PKA phosphorylation of Ser 357 of the mIP in independently mediating its coupling to both G␣ i and to G␣ q , whereby conversion of critical determinants of the PKA recognition site to that of a defined PKC phosphorylation site inhibits cicaprostinduced mIP phosphorylation and switching from G␣ s and, thereby, inhibits its coupling to G␣ i and to G␣ q . DISCUSSION Prostacyclin plays a central role in the local control of vascular hemostasis acting as an endothelium-derived inhibitor of platelet aggregation and as a vasodilator (5). Mice deficient in the IP show increased susceptibility to thrombotic stimuli, exhibit diminished pain perception and inflammatory responses (3,33), and develop more pronounced hypertension and vascular remodeling following hypoxic exposure relative wild type mice (34). Although IPs are thought to primarily couple to adenylyl cyclase (5,21), in certain species/cell types they are reported to couple to diverse effectors, including PLC activation (12-14, 16, 21, 22) and/or inhibition of adenylyl cyclase (18). Several GPCRs have been established to activate dual or, indeed, multiple G protein:effector systems (35)(36)(37)(38)(39), and a number of independent mechanisms have been proposed (40,41). We have recently established that the mIP couples to both activation and to inhibition of adenylyl cyclase, via G␣ s and G␣ i , respectively, and to activation of PLC, via G␣ q (22). Through detailed mechanistic studies, we established that coupling to both G␣ i and G␣ q was dependent on initial mIP coupling to G␣ s and adenylyl cyclase activation and subsequent cAMP-dependent PKA phosphorylation of the mIP at Ser 357 within its carboxyl-terminal (C)-tail region, thereby terminating mIP coupling to, and association with, G␣ s and switching mIP coupling to both G␣ i and to G␣ q (22). In the current study, we sought to test the generality of this model by examining the G protein coupling specificity and mechanisms of signaling of the hIP, comparing it to that of the mIP.
Stimulation of HEK.mIP, HEK.hIP, and HEL cells and human platelets with cicaprost led to concentration-dependent increases in cAMP generation, consistent with mIP and hIP coupling to G␣ s . Whereas stimulation of the mIP with cicaprost inhibited forskolin (Fsk)-induced cAMP generation, consistent with mIP coupling to G␣ i , stimulation of the hIP augmented Fsk-induced cAMP generation indicative of the hIP coupling to G␣ s , but not to G␣ i . Moreover, although PTx and H-89 abolished mIP coupling to G␣ i , cAMP generation by the hIP was not affected by these agents.
Stimulation of HEK.mIP and HEK.hIP cells with cicaprost also led to increases in IP 3 generation and to mobilization of [Ca 2ϩ ] i consistent with both mIP (14,21,22) and hIP (12,15,16) coupling to PLC activation. Although stimulation of HEL cells with cicaprost also yielded significant increases in IP 3 generation and [Ca 2ϩ ] i mobilization, hIPs expressed in human platelets did not couple to PLC activation (Ref. 42 and data not shown). These data are consistent with previous findings demonstrating that, although hIPs expressed in the more immature megakaryoblastic cells, such as HEL and MEG-01 cells, retain the ability to couple to PLC activation, those expressed in the more mature platelet have lost that function, largely owing to changes in the profile of G protein expression during the progression of megakaryocytopoiesis (43). Although cicaprost-induced IP 3 generation and [Ca 2ϩ ] i mobilization by the mIP were inhibited by H-89 (22), neither IP 3 generation nor [Ca 2ϩ ] i mobilization by the hIP were affected by H-89. Hence, there are fundamental differences in signaling by the mIP and hIP, in that the mIP coupling to G␣ i and G␣ q is dependent on its initial coupling to G␣ s and subsequent PKA-mediated phosphorylation and switching (Fig. 9A (22)); the hIP, on the other hand, independently couples to G␣ s and G␣ q but does not undergo PKA phosphorylation and switching and does not couple to G␣ i (Fig. 9B).
While the mIP and rIP exhibit extensive amino acid sequence identity (94% identity), the hIP and mIP only share 73% overall identity, exhibiting greater divergence within their Ctail regions (66% identity). Alignment of the C-tail regions of the mIP (21) and the hIP (30 -32) revealed that the previously identified PKA site at Ser 357 within the mIP (RPAS 357 GRR), is replaced by a consensus target site for PKC within the hIP (QLAS 328 GRR) whereby the Arg-Pro (RP) versus Gln-Leu (QL) represent the only divergent residues and therefore act as the determinants of PKA versus PKC phosphorylation of the targeted Ser. Although the sequence RPAS 357 GRR within the mIP may actually be predicted to act as both a PKA or PKC site (32,21), we have confirmed that this site acts as a PKA, but not a PKC, site, and as stated, cicaprost-induced PKA phosphorylation of Ser 357 mediates mIP switching from G␣ s to G␣ i and G␣ q (22). Moreover, Ser 328 of the hIP has been confirmed to be a direct target for cicaprost-induced PKC, but not PKA, phosphorylation (12,13). Thus, in view of the differential G protein coupling specificities of the mIP and the hIP to G␣ i and their differential dependence on PKA in regulating their PLC activation, we investigated whether alteration of the determinant residues within the consensus PKC recognition sequence of the hIP surrounding Ser 328 might alter its G protein coupling specificity and intracellular signaling. Thus, the hIP QL325,326RP was generated whereby the consensus PKC phosphorylation site was mutated to a putative PKA site, identical to that of the mIP in this region.
The hIP QL325,326RP exhibited both G␣ s and G␣ i coupling, with G␣ i coupling inhibited by PTx and H-89. This receptor also exhibited G␣ q coupling that was unaffected by H-89 or by GF 109203X or PTx (data not shown), consistent with the independent coupling of hIP QL325,326RP to G␣ q and PLC activation. Although the hIP underwent cicaprost-induced phosphorylation, consistent with previous reports (12,13), this phosphorylation was unaffected by H-89 but was inhibited by GF 109203X. Although the hIP QL325,326RP also underwent cicaprost-induced phosphorylation, it was unaffected by GF 109203X but was inhibited by H-89. Hence, conversion of the PKC site within the hIP to that of a PKA site generated hIP QL325,326RP , which can switch from G␣ s to G␣ i through a PKA-dependent mechanism (Fig. 9C), and implies that the contextual sequence surrounding Ser 328 in the hIP QL325,326RP is essential for G␣ i coupling following its phosphorylation by PKA.
To further explore the essential requirement of the latter PKA phosphorylation at Ser 328 in mediating hIP QL325,326RP switching and coupling to G␣ i , hIP S328A was generated. Although the hIP S328A coupled to G␣ s , it failed to exhibit coupling to G␣ i . The hIP S328A also mediated increases in IP 3 generation and [Ca 2ϩ ] i mobilization consistent with its independent coupling to G␣ q /PLC activation. Whole cell phosphorylations established that, unlike the hIP or the hIP QL325,326RP , the hIP S328A did not undergo significant cicaprost-induced phosphorylation. Hence, the hIP S328A independently couples to both G␣ s and G␣ q , similar to the hIP, but does not couple to G␣ i .
As previously stated, coupling of the mIP to both G␣ i and to G␣ q is dependent on its initial coupling to G␣ s and consequent PKA phosphorylation of mIP at Ser 357 thereby switching mIP from G␣ s coupling to G␣ i and G␣ q coupling (22). Herein, we also investigated whether alteration of the critical divergent residues within the consensus PKA target site surrounding Ser 357 , which converts it to a PKC target site identical to that found within the hIP, would alter the G protein coupling specificity of the mIP. Although the mIP RP354,355QL coupled to G␣ s , it failed to exhibit coupling to G␣ i . Additionally, the mIP RP354,355QL failed to couple to G␣ q , and whole cell phosphorylations established that mIP RP354,355QL did not undergo measurable cicaprost-induced phosphorylation. Hence, conversion of the PKA site to a PKC site within the mIP yielded mIP RP354,355QL that can independently couple to G␣ s but that does not undergo agonist-induced phosphorylation on Ser 357 and, hence, cannot switch coupling from G␣ s to G␣ i or to G␣ q (Fig. 9D).
Taken together, these data with the hIP and the mIP and their mutants have indicated the critical requirement for the dipeptide RP, as opposed to a QL, in the Ϫ3 and Ϫ2 positions in addition to a phosphorylated Ser (representing ϩ1) in determining mIP coupling to G␣ i and G␣ q and in determining hIP coupling to G␣ i , but not to G␣ q . Further experiments are re-quired to dissect the importance of the contextual nature of the individual residues within the dipeptide RP sequence and/or any knock-on or consequent structural changes in determining that G␣ i /G␣ q interaction and coupling. A similar type of PKAdependent mechanism is involved in regulating the human ␤ 2 adrenergic receptor switching from G␣ s to G␣ i (40). In this case, a 14-amino acid peptide containing a PKA recognition sequence RRSS within the third intracellular loop of the ␤ 2 adrenergic receptor could stimulate GTP␥S binding by G␣ s but not by G␣ i ; on the other hand, the PKA-phosphorylated peptide showed weak G␣ s activation and strong G␣ i activation (44). Similar to our findings with the mIP sequence (22), although the RRSS could be predicted to act as both a PKA and PKC recognition site, it was preferentially phosphorylated by PKA and not by PKC (44).
In essence, these studies highlight critical differences in mIP and hIP:G protein coupling and intracellular signaling. Although the mIP can couple to G␣ s , and to G␣ i and G␣ q , coupling to the latter G protein:effector systems is dependent on PKAmediated phosphorylation and switching (Fig. 9A). The hIP, on the other hand, can independently couple to G␣ s and to G␣ q , FIG. 9. Model of mIP and hIP coupling to G␣ s , G␣ i , and G␣ q . A, ligand-activated mIP stimulates: (i) G␣ s -mediated activation of AC, leading to increases in cAMP generation and, in turn, activation of PKA. Activated PKA phosphorylates the mIP at Ser 357 switching mIP coupling from G␣ s to (ii) pertussis toxin (PTx)-sensitive G␣ i and inhibition of AC, and to (iii) G␣ q and activation of phosphatidyl inositol-specific PLC (PI-PLC) leading to increases in IP 3 generation and mobilization of [Ca 2ϩ ] i . B, ligand-activated hIP independently couples to (i) G␣ s -mediated activation of AC and to (iii) G␣ q -mediated activation of PLC. The hIP does not couple to (ii) G␣ i . C, ligand-activated hIP QL325,326RP (hIP QL ) independently couples to (iii) G␣ q -mediated activation of PLC and to (i) G␣ s -mediated activation of AC, leading to increases in cAMP generation and, in turn, activation of PKA. Activated PKA phosphorylates the hIP QL325,326RP at Ser 328 switching its coupling from G␣ s to (ii) G␣ i and inhibition of AC. D, ligand-activated mIP RP354,355QL (mIP RP ) stimulates (i) G␣ s -mediated activation of AC, leading to increases in cAMP generation and, in turn, activation of PKA. Unlike the mIP, mIP RP354,355QL does not undergo PKA phosphorylation at Ser 357 and, hence, cannot switch coupling from G␣ s to (ii) G␣ i or to (iii) G␣ q . The steps inhibited by H-89 are indicated in A-D.
but cannot couple to G␣ i (Fig. 9B). mIP coupling to G␣ i , versus that of hIP, is primarily regulated by a targeted cicaprostinduced PKA consensus sequence at Ser 357 of the mIP, which is replaced by a PKC consensus sequence in the hIP, at Ser 328 . Conversion of the PKC recognition site within the hIP to a PKA site within hIP QL325,326RP facilitates its coupling to G␣ i (Fig.  9C), therefore emphasizing the importance of the contextual nature of the Arg 325 -Pro 326 residues in addition to phosphorylation of Ser 328 in regulating G␣ i coupling. mIP coupling to G␣ q is also regulated by a targeted cicaprost-induced PKA phosphorylation site at Ser 357 and conversion of that PKA recognition sequence from a PKA to a PKC site (in mIP RP354,355QL ) abolished mIP coupling to G␣ i and G␣ q (Fig. 9D). The hIP independently couples to G␣ q and is not regulated by PKA or PKC phosphorylation of hIP at Ser 328 as neither (a) conversion of the PKC site of the hIP to a PKA recognition sequence within hIP QL325,326RP (Fig. 9C) nor (b) abolition of the PKC site through site-directed mutagenesis within hIP S328A impaired the ability of the hIP to couple to G␣ q /PLC activation. Thus, it appears that other sequence differences within the hIP, as opposed to the mIP, act as determinants in facilitating its independent coupling to G␣ q . Given the extent of sequence variation that exists between the hIP and the mIP, particularly within their C-tail sequences and third intracellular loop regions, it is likely that these domains may also contain the sequence determinants of G␣ q interactions with the hIP.
These critical differences in mechanisms of intracellular signaling by the hIP and the mIP may confer important speciesdependent differences to the cellular responses to prostacyclin under both physiologic/pathophysiologic settings, and an appreciation of these mechanisms may offer a basis for the differential targeting of prostacyclin-signaling in certain disease settings. Finally, data presented herein greatly expands our existing knowledge of the general mechanisms of GPCR signaling, particularly in relation to their ability to couple to multiple G protein:effector systems.