Purification and properties of prostaglandin E1/prostacyclin receptor of human blood platelets.

Activation of platelet adenylate cyclase by prostaglandin E1 or prostacyclin is initiated through the interaction of the agonists with the same receptors on membrane. Prostaglandin E1/prostacyclin receptors of human platelets were solubilized in buffer, containing 0.05% Triton X-100 and protease inhibitors. The soluble membrane protein was chromatographed on a DEAE-cellulose column and assayed by a microfiber filter by equilibrium binding technique. The active fractions eluted at 0.7 M KCl were pooled, and the receptors were purified to homogeneity by Sephadex G-200 gel filtration with an overall recovery of 30%. The isolated receptor was 2,200-fold purified over the starting platelets. As evidenced by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the receptor showed a molecular mass of 190,000 daltons and is composed of two nonidentical subunits with molecular masses of 85,000 and 95,000 daltons. The interaction of prostaglandin E1 with the purified receptor was rapid, saturable, reversible, and highly specific. Among all prostaglandins tested, only prostacyclin was capable of displacing [3H]prostaglandin E1 bound to the receptor. Scatchard analysis of [3H]prostaglandin E1 binding to the purified receptor suggested the presence of a single class of high affinity binding sites (Kd = 9.8 nM) and a second population of low affinity binding sites (Kd = 0.7 microM) in the same protein molecule. Incubation of the purified receptor with platelets stripped of the receptor by washing with low concentrations of Triton X-100 efficiently restored the ability of prostaglandin E1 and prostacyclin to activate adenylate cyclase in these cells.

Activation of platelet adenylate cyclase by prostaglandin E, or prostacyclin is initiated through the interaction of the agonists with the same receptors on membrane. Prostaglandin EJprostacyclin receptors of human platelets were solubilized in buffer, containing 0.05% Triton X-100 and protease inhibitors. The soluble membrane protein was chromatographed on a DEAE-cellulose column and assayed by a microfiber filter by equilibrium binding technique.
The active fractions eluted at 0.7 M KC1 were pooled, and the receptors were purified to homogeneity by Sephadex G-200 gel filtration with an overall recovery of 30%. The isolated receptor was 2,200-fold purified over the starting platelets. As evidenced by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the receptor showed a molecular mass of 190,000 daltons and is composed of two nonidentical subunits with molecular masses of 85,000 and 95,000 daltons. The interaction of prostaglandin El with the purified receptor was rapid, saturable, reversible, and highly specific. Among ali prostaglandins tested, only prostacyclin was capable of displacing [3H]prostaglandin E, bound to the receptor. Scatchard analysis of [3H]prostaglandin E, binding to the purified receptor suggested the presence of a single class of high affinity binding sites (& = 9.8 nM) and a second population of low affinity binding sites (& = 0.7 pM) in the same protein molecule. Incubation of the purified receptor with platelets stripped of the receptor by washing with low concentrations of Triton X-100 efficiently restored the ability of prostaglandin E, and prostacyclin to activate adenylate cyclase in these cells. ~ Aggregation of human blood platelets has been shown to be a critically important event in the process ranging from hemostasis to the extreme of thrombosis (1). The aggregation of platelets by various agonists such as ADP, l-epinephrine, or collagen is believed to be mediated in part through the intracellular synthesis of prostaglandin endoperoxide and thromboxane A2 (2). In contrast, the inhibition of platelet aggregation is mediated primarily through the increase of cyclic AMP levels by the activation of adenylate cyclase by various autacoids including PGE,,' PG12, or PGD, (3-5). The * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
ll To whom correspondence and reprint requests should be addressed.
hormone-responsive adenylate cyclase is believed to be composed of at least three subunits: an agonist-specific receptor, a guanine nucleotide binding regulatory protein, and a catalytic subunit (6). The receptors appear to be separate, mobile macromolecular protein entities residing on the outer surface of the membrane bilayer. Stimulation of adenylate cyclase is initiated through the interaction of activators including PGE, with the specific receptors on the membrane plane (7)(8)(9).
In the case of human platelets, the receptors of PGIp, PGE1, and PGD, have been identified and shown to ?x associated with the membrane structure (10,11). However, these receptors have not yet been isolated nor have their structures been characterized. We describe here the purification and characterization of adenylate cyclase-linked PGEJPGI, receptors from human platelets and the reconstitution of PGE1/PG12responsive platelet adenylate cyclase activity by the purified receptors.  (PGA,, PGA2, PGB,, PGD,, PGE,, PGE,, and  PGI,), Lubrol PX, polyoxyethylenesorbitan monostearate (Tween SO), taurocholic acid, and CHAPS were obtained from Sigma; Triton X-100 and Bio-Beads SM-2 were the products of J. T. Baker Chemical Co. and Bio-Rad Laboratories, respectively. All other chemicals used were of analytical grade.
Preparation of Platelets Typically, 10-12 g of fresh "bloody" platelets (obtained from the American Red Cross, Philadelphia) were mixed with 1.0 mM EDTA (final concentration) and centrifuged at 200 X g for 15 min at 23 "C to remove residual erythrocytes and leukocytes from the platelet suspension. The supernatant platelet-rich plasma was next centrifuged at 2000 X g for 20 min at 23 "C, and the platelet pellet thus obtained was washed three times by centrifugation after resuspending the pellet in 50 mM Tris-HC1 buffer, pH 7.4 (1 g wet weight/ml of buffer) containing 1.0 mM EDTA and 0.15 M NaCI.

Solubilization of PGE, Receptors
To evaluate the efficacy of various detergents in solubilizing the PGEI receptors, the platelet suspension (1.0 ml) was treated separately with the following compounds in the above buffer containing 1.0 mM PMSF, 5 mM DTT, 5 mM MgCl,, and 0.3 M sucrose (final concentration shown in the parentheses): CHAPS (0.2%), Lubrol PX (0.2%), taurocholic acid (0.2%), Triton X-100 (0.05%), and Tween 60 (0.2%). After the addition of the detergents, the platelet suspension was incubated at 0 "C (in ice) for 30 min with occasional mild shaking.
Care was taken to avoid frothing. After incubation, the detergenttreated platelet suspension was centrifuged at 35,000 X g for 30 min at 0 "C. The PGEI binding activities of both the supernatant and the pellet were subsequently determined.
Of all the detergents tested, Triton X-100 seemed to be the most 12685  (Table I). Except in the case of Triton X-100, the PGE, binding activities of the different extracts were determined without removing the detergents from the same, and the degree of solubilization of the receptor by individual agent may not be comparable. Thus, the data were useful only as a guideline for further purification. In the case of Triton X-100, the removal of the detergent by Bio-Beads SM-2 increased the binding of [3H]PGE1 to the solubilized protein by 10-12-fold, indicating that the agent itself is inhibitory to the ligand receptor interaction. Over 90% of the PGE, binding activity was solubilized by the detergent under experimental conditions as described when compared with the untreated washed platelets. The centrifugation of the solubilized receptor at 35,000 X g at 0 "C was found to be sufficient for the complete removal of all the intact membranes, since when the detergent-solubilized preparation was centrifuged at 100,000 X g for 1 h at 0 "C, the specific activity of the supernatant increased by only 5%. It was found that the receptors once solubilized by Triton X-100 remained in the solubilized state even in the absence of the free detergent in the solvent.
Purification of PGE, Receptor DEAE-cellulose Chromatography-All the chromatographic procedures described below were carried out at 4 "C. Typically, approximately 12 g of washed platelets suspended in 12 ml of the above buffer was treated with Triton X-100 to solubilize PGE, receptor. The supernatant containing solubilized receptors (12 ml) was reduced to 6 ml by covering the samples in a dialysis bag with powdered polyethylene glycol (M, 15,000-20,000) at 0 "C and periodically removing the wet solid. The concentrated solution was clarified by centrifugation at 8.000 X g at 0 "C and immediately applied to a DEAE-cellulose column (2.5 X 60 cm) equilibrated with 10 mM Tris-HCl buffer, pH 7.4, containing 0.05% Triton X-100, 1.0 mM PMSF, 1.0 mM DTT, 5 mM MgCl,, and 0.3 M sucrose and fractionated by stepwise elution with increasing concentrations of KC1 (0.1-0.7 M) in the same buffer except that the concentration of the buffer was increased to 50 mM. The eluates from the washing of the column with 75 ml each of 0, 0.1, 0.2, and 0.5 M KC1 in 50 mM Tris-HC1 buffer did not show the presence of any PGE, binding activity and were discarded (not shown). At 0.7 M KCI, the PGE, binding activity began to emerge from the column in a single peak (Fig. 1). Fractions (12-32) which contained the PGE, binding protein with highest specific activity were pooled and concentrated (6.0 ml). Further elution of the column with 1.0, 1.25, and 1.5 M KC1 (75 ml each) did not show the presence of any PGE, binding activity in the eluates (not shown).
Sephadex G-200 Chromatography-The concentrated fractions from the DEAE-cellulose column were applied to a Sephadex G-200 column (1.0 X 40 cm) equilibrated with the buffer system used in the DEAE-cellulose chromatography except that KC1 was omitted. The elution was carried out with the same buffer. The eluates containing PGE, binding activity emerged in a single peak in fractions 27-35 (8 ml) (Fig. 2). The active fractions were pooled and concentrated to 0.4 ml as described above. The purified receptor was kept at -70 "C in Removal of Triton X-100"Triton X-100 was removed from the protein sample using Bio-Beads SM-2 as described by Holloway (12). Optimum conditions for the removal of the detergents with minimum loss of protein were standardized by treating the samples (0.8 ml) with different amounts of moist beads for 30 min at 4 "C. The samples were then centrifuged at 35,000 X g for 30 min, and the supernatants was analyzed for protein and Triton X-100 content (13). Complete removal of the excess detergent was achieved using 1.1 ml of moist beads per 0.8 ml of sample. The removal of Triton X-100 from the receptor preparation increased the [3H]PGE, binding by 12-fold. However, the detergent has a protective action on the prostaglandin binding activity of the sample, and the samples were routinely stored with 0.05% Triton X-100.
Determination of Proteins-Proteins were determined according to Lowry et al. (14) with bovine serum albumin as the standard. The standard curve for the determination of the protein was constructed by adding 0-0.05% of Triton X-100 in the assay mixture. The presence of the above concentrations of the detergent do not interfere with the protein assay which was further verified by fluorescamine assay (15).
Polyacrylamide Gel Electrophoresis-Homogeneity of the purified receptor protein was tested by polyacrylamide gel electrophoresis under alkaline conditions (16). Molecular weight and the subunit composition of the purified receptor were determined by SDS-polyacrylamide (7.5%) gel electrophoresis with nonreduced and reduced protein using 0.1 M DTT (17). Typically, 20 pg of the purified protein were reduced by boiling the sample for 2 min in the presence of 2% SDS and 0.1 M DTT. In the case of the unreduced protein, the samples were similarly treated in the absence of the reducing agent. The gels were stained with 0.02% Coomassie Brilliant Blue in 10% acetic acid and destained in a 10% acetic acid, 10% isopropyl alcohol (1:l) mixture. The gels were also stained with AgNO, as described by Merril et al. (18). The molecular weights of the PGE, receptor and its subunits were determined by using marker proteins of known molecular weights (Bio-Rad).
In some phases of the work, it was necessary to elute the protein from the gel and assay its [,H]PGEI binding activity after the electrophoresis. In these cases, the gels were cut into slices (approximately 1.0-mm thickness) and homogenized in 0.2 ml of 50 mM Tris-HC1 buffer, pH 7.4, containing 5 mM M&Iz for 5 min at 0 "C. The homogenate was centrifuged at 8000 X g at 4 "C. The supernatant was collected, and the PGE, binding activity was determined.
pH]PGE1 Binding Assay-Unless otherwise stated, the binding of [3H]PGE, to the solubilized fractions from the platelets was performed by incubating 20-100 pg of protein for 30 min at 23 "C in 50 mM Tris-HC1 buffer containing 5 mM M&~z in a total volume of 200 pl with 0.3 p~ PGE, containing 0.1 pCi of [3H]PGE1. Since less than 1% of the PGE, bound to the preparation, the concentration of the free ligand was essentially constant throughout the incubation. Unless otherwise indicated, parallel experiments were run using 50-fold excess labeled PGE, in the above incubation mixture to determine the nonspecific binding. This value was subtracted from the total PGE, bound to calculate the specific binding. At the end of the incubation, 1.0 ml of the above buffer (0 "C) was added to each tube, and the mixture was filtered under vacuum through a Whatman glass microfiber filter (GF/C, 2.9 cm in diameter) which had been presoaked with the assay buffer. The solubilized receptor remained bound to the filter which was washed with 15 ml of the buffer at 0 "C. The filters were then dried and suspended in 10 ml of scintillation fluid (Amersham, ACS-11) and counted in a Beckman Scintillation Spectrometer (LS-8000) with 45% efficiency for 3H. The possibility that [,H]PGE, might be degraded during the period of binding was tested by incubating the radiolabeled autacoid with solubilized proteins in the same incubation mixture as described above. After incubation, the suspension was acidified with 1% HCOOH and extracted with 3 volumes of a CHC13/CH30H (95:5) mixture three times at 0 "C. The extracted materials were then analyzed by high pressure liquid chromatography using a pBondapak (particle size, 10 pm) reverse phase (C,) column (Waters Associates), and H20/CH3CN/CH3COOH (76.7:23:0.2, v/v/ v) solvent system as described earlier (19).

Analysis of Equilibrium Binding of PGE, to the Purified Receptors
The interaction of PGE, with the purified receptors was analyzed by the method of Scatchard (20). The dissociation constant (Kd) and the number of binding sites (n) were obtained from nonlinear regression analysis of equilibrium binding by a nonweighted, iterative, least squares algorithm analysis using a Radio Shack TRS 80, Model 4 microcomputer.

Assay of Adenylate Cyclase
Adenylate cyclase activity of platelet homogenate was determined by incubating 1.0 mM ATP containing 2 pCi of [c~-~'P]ATP, 2 mM MgS04, 10 mM theophylline, 1 mM creatine phosphate, 1 unit of creatine phosphokinase, 1.4 p~ PGE,, and 50 mM Tris-HC1 buffer, pH 7.4, in a total volume of 0.1-ml reaction mixture. After incubation at 37 "C for 5 min, the reactions were stopped by adding SDS (1% final concentration). The radioactive cyclic AMP was separated according to Salomon et al. (21). Unlabeled cyclic AMP (1.0 mM) was added to the reaction mixture to facilitate the recovery of the nucleotide. The heat-inactivated enzyme preparation was made by heating the enzyme preparation with the above buffer in a boiling water bath before (22). for 5 min. Platelet homogenate was prepared by the method described

RESULTS
Purification of Human Blood Platelet PGE, Receptor-The purification protocol of the human platelet PGE, receptor is summarized in Table 11. PGE, receptors were solubilized by treating fresh platelets with 0.05% Triton X-100. Over 90% of the PGEl binding activity of the platelets was recovered in the detergent extract. The detergent extract was centrifuged, and the resultant supernatant was first fractionated on a DEAE-cellulose column followed by chromatography of the active fractions on a Sephadex G-200 column as described under "Experimental Procedures." PGE, binding activity of the different fractions in each step was measured after the removal of the detergent by Bio-Beads SM-2. The procedure described above resulted in an increase of specific binding of [3H]PGEl from 2.2 pmol/mg of protein to 4,900 pmol/mg of protein which showed over 2,200-fold purification of PGEl receptor with 30% yield compared to the intact platelets.
Electrophoretic Homogeneity: Molecular Weight and Subunit Composition of PGE, Receptor Protein-Alkaline gel electrophoresis under nonreducing conditions, the PGE, binding protein from the Sephadex G-200 column showed that the isolated product was homogeneous when the gels were stained either with AgN03 (Fig. 3A) or with Coomassie Brilliant Blue (Fig. 3B). In a parallel experiment, an identical gel which was not stained with the dye to avoid denaturation was cut into slices, the binding protein from the slices was eluted, and the ['H]PGE1 binding activity of the eluates was determined. The position of the gel slices showing the [3H]PGE1 binding activities corresponded exactly to the protein band in the gel stained by the dye except that in the case of the unstained gel which was not fixed, the receptor protein apparently diffused more when compared to the dye-stained gel (Fig. 3B). Polyacrylamide gel electrophoresis of the purified protein, in the presence of SDS but in the absence of DTT, showed the molecular weight of the receptor to be 190,000 (Fig. 4). Upon reduction with DTT, the receptor protein resolved into two subunits migrating closely in the SDS-polyacrylamide gel.  (Fig. 6)   ( n ) . The best fit, as described by computer (Fig. 7), occurs with high affinity, low capacity binding sites exhibiting a dissociation constant ( K d l ) of 9.8 nM and binding capacity ( n ) of 0.082 mol/mol of the protein ( n = 3). The same analysis yields low affinity, high capacity binding sites, has a dissociation constant (Kd2) of 0.7 PM and binding capacity, n2, of 0.89 mol/mol ( n = 3) of the receptor. It is not known whether the downward curvature of the Scatchard plot was due to heterogeneity of' the binding sites or due to negative cooperativity within a single class of binding sites in the same protein molecule (23).

Effect of pH and Divalent Metal Ions on C'HJPGE, Binding to the Purified Receptors-
The optimum pH for the binding of ["HIPGE, to the receptor was found to be approximately 7.5, and the presence of Mg" was needed for the optimum binding of the ligand to the purified receptors (Fig. 8)

Reconstitution of PGE,/PGI,-responsive Platelet Adenylate
Cyclase by the Purified Receptor-To test the biological activity of the purified receptor, attempts were made to reconstitute PGE,-responsive adenylate cyclase of platelets with the purified receptors. When a platelet homogenate prepared from the cells washed with Triton X-100 to remove PGE, receptors was examined, little stimulation of adenylate cyclase activity in response to the autacoid challenge was demonstrated when compared with the control (Table 111). The purified receptors themselves did not show measurable adenylate cyclase activity. However, when the receptor was added to the detergentwashed platelets and subsequently the PGE,-responsive adenylate cyclase activity of the homogenate was assayed, it was found that the purified receptor was capable of regenerating the response of the enzyme to PGE,. Addition of 10 fig of the purified receptor, the quantity approximately equivalent to the amount of the receptor present in 2 X 10" platelets, to the detergent-treated platelet restored the activation of adenylate cyclase of the homogenate by PGE, by 57% compared to the control. PG12, like PGE,, also stimulated the activity of platelet homogenate adenylate cyclase reconstituted with PGE, receptor (Table 111).
The effect of 0.05% Triton X-100 on the removal of PGE, receptors from platelets was found to be highly specific, since the treatment of these cells with similar concentrations of the detergent under identical conditions did not reduce the stimulation of adenylate cyclase activity by either PGD, or adenosine (Table 111). Consequently, the addition of these agonists to the assay mixture containing the reconstituted adenylate cyclase produce no further stimulation of the enzyme over the control platelet homogenate (Table 111).
Stability of the Purified PGE, Receptor-The purified receptor (1.0 mg/ml), when kept at -70 "C in 50 mM Tris-HC1 buffer, pH 7.4, containing 0.3 M sucrose and 0.05% Triton X-100, was found to have a half-life of approximately 3 months. The presence of detergent apparently had a protective action on the stability of the receptor. In the absence of the detergent, the purified receptor lost 50% of its binding activity in less than a month at -70 "C.

DISCUSSION
The interaction between the ligand and the receptor has been shown to be a prerequisite for the activation of adenylate the purified receptors Platelets (2 X 10'' cells) were washed with 0.05% Triton X-100 as described under "Experimental Procedures." The detergent-treated platelet pellet was subsequently washed four times with 3 ml of 50 mM Tris-HC1, pH 7.4, by centrifugation at 2000 X g at 0 "C. After the final wash, the platelets were either homogenized or suspended in 0.5 ml of the above buffer and incubated with the purified receptor (10 pg) for 15 min at 23 "C. The platelets were then homogenized (22), and the stimulation of adenylate cyclase by various agonists was determined. Typically, 0.5 mg of homogenate protein was incubated with 1.0 mM ATP containing 2 pCi of [w3'P]ATP, 2 mM MgSO,, 10 mM theophylline, 1.0 mM creatine phosphate, 1 unit of creatine phosphokinase in 50 mM Tris-HC1 buffer, pH 7.4, in a total volume of 0.1 ml in the presence of various agonists as indicated for 5 min at 37 "C. The formation of cyclic [32P]AMP was determined according to the method of Salomon et al. (21). In the control experiments, the platelets were similarly washed with the above buffer except that it did not contain any detergent. Results shown are mean f S.E. of six  (25)(26)(27)(28)(29). Our results showed that the receptor involved in the PGE,/PGI*-stimulated adenylate cyclase in human platelet is a separate membrane protein. The receptor protein could be easily removed from the cell surface by mild detergent treatment. Light microscopic analysis of the Triton X-100 (0.05% final concentration)-treated platelets showed little visible damage, yet over 90% of the total PGEJPGI, receptor could be removed by the detergent washing. The ease with which the receptors are removed supported the concept that adenylate cyclaselinked PGE, receptors are located on the outer surface of the membrane bilayer. Although the presence of the detergent is necessary for solubilization of the receptor, once in solution, the receptor remained soluble even in the absence of the free detergent in the solvent. As opposed to /?-adrenergic receptors (30,31), the PGE1/PGI, receptors survived solubilization in the absence of the agonist, although the detergent itself appears to protect the receptors from denaturation.
The platelet adenylate cyclase in also activated by PGD, and adenosine (4, 10). However, the receptors of these agonists, unlike PGE1/PGIz receptors, cannot be removed by treating these cells with 0.05% Triton X-100. The ease of removal of PGE,/PG12 receptors from platelet membrane suggests that receptor protein is probably attached peripherally to the outer surface of the membrane bilayer. In contrast, the receptors of PGDz and adenosine might be more deeply embedded on the membrane. Although PGE1/PGI2 receptors might be only peripherally attached to the membranes, the macromolecules, nevertheless, are platelet proteins. Since platelet-free plasma does not contain any PGEI/PGIz receptor protein,' the possibility that platelets could pick up these receptors from the plasma was ruled out.