1784
Prostaglandin Half-life Regulated by High Density Lipoprotein Is Decreased in Acute Myocardial Infarction and Unstable Angina Pectoris 12
Takeshi Aoyama, MD, Yoshiki Yui, MD, Hiroshi Morishita, MD, and Chuichi Kawai, MD To investigate the prostaglandin 12 (PGI2) half-life regulated by high density lipoprotein (HDL) in patients with coronary artery disease (CAD), we determined the stability of PGI2 and serum apolipoprotein A-I (Apo A-I) and apolipoprotein A-II (Apo A-1I) levels in four age-matched groups of patients: controls (n=17), angina pectoris (n=18), unstable angina pectoris (n=17), myocardial infarction (n = 19) (acute phase, 3.6+1.7 hours from onset; subacute phase, 75±15 hours from onset in the same patients). Serum PGI2 half-life and total serum Apo A-I levels were lower in the CAD group than in the control group. PGI2 was least stable in patients with unstable angina and the acute phase of myocardial infarction. In these patients, the molar ratio of Apo A-I to Apo A-IT and HDL-associated Apo A-I levels were decreased, and free Apo A-I levels were increased. After in vitro incubation of HDL with increasing amounts of Apo A-II, Apo A-I in HDL was displaced by Apo A-TI, with the parallel decrease in stability of PGI2. Free Apo A-I cannot stabilize PGI2. HDL-associated Apo A-I, whose amount is affected by Apo A-II, stabilized PGI2 and correlated well with stability of PGI2 in patients with CAD and control patients. Decreased PGI2 half-life may play an important role in the pathogenesis of atherosclerosis and thrombus formation in the coronary arteries, especially thrombus formation during an acute coronary event. (Circulation 1990;81:1784-1791)
rostaglandin (PGI2) is a powerful vasodilator and inhibitor of platelet aggregation and also has a cytoprotective role. It has been thought to play a role in the prevention of thrombosis and atherosclerosis.1 PGI2 is unstable in aqueous solution at physiological pH and temperature and is stabilized by binding to an unknown serum factor.23 A reduced stability of PGI2 has been observed in thrombotic disorders (thrombotic thrombocytopenic purpura4 and ischemic stroke5), and this reduction may be related to the pathogenesis of these diseases. Recently, we identified serum PGI2 stabilizing factor as apolipoprotein A-I (Apo A-I), a major apolipoprotein of high density lipoprotein (HDL).6 Our findings indicate that HDL or Apo A-I may exert its antiatherogenic and antithrombogenic effects partly through stabilization of PGI2. In this study, we found P
'2
From the Third Division, Department of Internal Medicine, Faculty of Medicine, Kyoto University, Kyoto, Japan. Supported by a research grant for cardiovascular diseases from the Ministry of Health and Welfare and grants from the Ministry of Education, Science and Culture, Tokyo, Japan. Address for correspondence: Chuichi Kawai, MD, Third Division, Department of Internal Medicine, Faculty of Medicine, Kyoto University, Kyoto 606, Tokyo, Japan. Received July 24, 1989; revision accepted January 24, 1990.
a reduced stability of PGI2 in patients with coronary artery disease (CAD), and we examined the mechanism by which such a reduction occurs.
Methods PGI2 Binding Activity [3H]-PGI2 methyl ester (New England Nuclear Corp., Boston, Massachusetts) was reduced to [3H]PGI2 sodium. [3H]-PGI2 sodium was incubated with 1.0 ml serum or HDL solution at room temperature for 3 minutes and was applied to a Sephadex G-25 See p 2013 column (1.5x25 cm) equilibrated with 50 mM TrisHCl, pH 7.4 buffer. The flow rate was 1 ml/min. Onemilliliter fractions of the gel filtrate were collected, and the radioactivity of each fraction was determined in a liquid scintillation counter (LS 1701, Beckman Instruments, Fullerton, California). Because the radioactivity recovered in all fractions was 99-100% of the radioactivity added, the recovered radioactivity count was used as the basis for determining the percentage of binding. The PGI2 was eluted in two peaks. The bound PGI2 was eluted in the first peak and the free PGI2 in the second peak. The PGI2
Downloaded from http://circ.ahajournals.org/ by guest on July 3, 2015
Aoyama et al Prostaglandin l2 Half-life Regulated by HDL
binding activity was expressed as the percentage of radioactivity in the binding peak.6 Preparation of HDL, Lipoprotein-Deficient Serum, Apo I, and Apo II HDL was isolated from human serum by preparative ultracentrifugation at densities between 1.063 and 1.21 g/ml and purified by recentrifugation.7 Lipoprotein-deficient serum (LPDS) was also obtained from human serum by ultracentrifugal flotation at densities greater than 1.25 g/ml. Apo A-I and Apo A-II were isolated from HDL by ion-exchange and gel permeation chromatographic techniques.8,9
Measurement of Cholesterol and Apo A-I and A-II Levels Cholesterol levels in serum and HDL were determined by enzymatic procedures (Wako Pure Chemical Co, Tokyo). Total serum Apo A-I and Apo A-II levels were determined by the single radial immunodiffusion method with Apo A-I and Apo A-IL plates (Daiichi Chemical Co, Ltd, Tokyo). Free Apo A-I and HDL-associated Apo A-I levels were determined by the method of Borresen and Berg.10 Measurement of Protein Protein was determined by the method of Lowry et al.11
Platelet Aggregation The biological stability of sodium PGI2 (10 ng/ml) (Ono Pharmaceutical Co, Osaka) was determined by platelet aggregation. ADP-induced platelet aggregation was measured with a Lumi-Aggregometer (model 400, Chrono-Log Corp, Havertown, Pennsylvania). Blood (9 ml) was drawn into a plastic syringe with 3.8% sodium citrate (1 ml) as the anticoagulant agent. Samples were centrifuged in plastic tubes successively at 300 and 2,000g for 15 minutes to collect platelet-rich and platelet-poor plasma, respectively, as needed to give a final platelet count of 300,000/mm3. Aggregation was induced in 0.2-ml aliquots of platelet-rich plasma by the addition of 0.02 ml ADP in saline solution to give a final concentration of 10 gM.12 Measurement of 6-Keto-PGF,a Levels 6-Keto-PGFl,, levels were determined by radioimmunoassay as previously described.13 [3H]-6-KetoPGFl, was obtained from New England Nuclear Corp, Boston, Massachusetts. Authentic 6-keto-PGF1a was
supplied by the Ono Pharmaceutical Co, Osaka. Anti6-keto-PGFla was purchased from Advanced Magnet-
ics, Inc., Cambridge, Massachusetts.
Study Population We studied male patients with chest pain of cardiac origin and an age-matched group of patients with pain due to noncardiac causes (control group). Patients with chest pain were divided into three groups: patients with stable angina pectoris, patients
1785
with unstable angina pectoris, and patients with acute myocardial infarction. The study group consisted of 71 patients: 17 control patients, 18 patients with stable angina pectoris, 17 patients with unstable angina pectoris, 19 patients with myocardial infarction (acute phase, 3.5 + 1.7 hours from the onset; subacute phase, 75 + 15 hours from the onset in the same patients). These four groups of patients were age matched. All patients underwent diagnostic coronary angiography for chest pain or suspected coronary artery disease. Coronary angiography was performed with the Judkins or Sones technique, and multiple views of the right and left coronary arteries were recorded. The angiograms were reviewed without prior knowledge of the patients' lipid, Apo A-I, and Apo A-II levels. Stable angina pectoris was defined as chest pain occurring during exercise and lasting less than 10 minutes in patients with objective evidence of coronary angiographic or electrocardiographic changes indicating ischemia during episodes of chest pain. The presence of serious stenosis (>75%) in at least one major epicardial vessel was confirmed in all patients by angiography. Unstable angina pectoris was defined as chest pain occurring at rest for a minimum of 15 minutes on at least one occasion in the absence of an increase in the MB fraction of plasma creatine kinase in patients with objective evidence of CAD; such evidence was based on coronary arteriography and electrocardiographic changes indicating ischemia during episodes of chest pain. The presence of serious stenosis (>90%) in at least one major epicardial vessel was confirmed in all patients by angiography. Postinfarction angina and vasospastic angina were excluded. Myocardial infarction was defined as chest pain occurring at rest in a patient with an associated elevation of the MB fraction of creatine kinase as well as typical electrocardiographic evidence of myocardial necrosis. Emergency coronary arteriography revealed occlusion or subtotal occlusion of a major epicardial vessel. The controls were patients with chest pain due to noncardiac causes whose coronary arteriograms were normal. Patients were excluded from the study if they had undergone any surgical procedure within the month preceding the study or were taking any drugs known to interfere with serum lipids and apolipoprotein levels. Patients receiving aspirin or other nonsteroidal anti-inflammatory agents were also excluded. Blood Sampling In patients with stable and unstable angina pectoris and control patients, venous blood samples were collected through an 8F sheath (Cordis, Co, Ltd, Miami, Florida) in a cardiac catheterization room after a 12-hour fast. In patients with acute myocardial infarction, the first blood sample was collected within 6 hours from the onset through the 8F sheath during the emergency cardiac catheterization and
Downloaded from http://circ.ahajournals.org/ by guest on July 3, 2015
1786
Circulation Vol 81, No 6, June 1990
before any medical intervention (urokinase or tissuetype plasminogen activator). The second blood sample was obtained through the 8F sheath in the coronary care unit 3-5 days after onset and a 12hour fast. Blood samples were collected before heparin administration into plastic tubes and centrifuged at 2,000g for 10 minutes and stored at 4° C. Serum was analyzed within 24 hours after sampling. In Vitro Interaction of Human HDL With Human Apo A-II and Its Relation to PGI2 Binding Activity Human serum HDL from normal fasting male donors was incubated with Apo A-II14,15: 0.8 ml of an HDL sample with Apo A-I content of 1.12 mg (concentration, 140 mg/dl) and an Apo A-II content of 0.32 mg (concentration, 40 mg/dl) was incubated with 0.2 ml of an Apo A-II solution (50 mM TrisHCI, pH 7.4) at concentrations of 0-400 mg/dl. After 1 hour of incubation at room temperature, the PGI2 binding activity and free Apo A-I level of the mixture were determined.
A1
3
:gI
B
104f*1
*
r P
U.
0 0
5, 0
PGI2 Binding Activity in Lipoprotein-Deficient Serum To investigate the ability of free Apo A-I to stabilize PGI2, we examined the PGI2 binding activity in LPDS that contained free Apo A-I.16
---
cr
0.
U.
U,i
Statistical Analysis All data are expressed as mean± SD. The data were analyzed by a multiple comparison test by NewmanKeuls method with one-way analysis of variance. For data in Figure 1, a multiple comparison test with two-way analysis of variance was used. A probability value less than 0.05 was considered significant. Results Effect of HDL on the Stabilization of PGI2 Sodium PGI2 (10 ng/ml) was incubated in 1 ml 50 mM Tris-HCI buffer at pH 7.4 and LPDS (1:9, vol/vol) solution. After 0, 10, 20, 30, 40, 60, and 90 minutes of incubation, the activity of PGI2 was assayed by platelet aggregation. After 0, 30, 60, and 90 minutes of incubation, radioimmunoassay of free PGI2 (6-ketoPGF1l) was also run in parallel with the platelet aggregation bioassay. The addition of HDL (final Apo A-I concentration, 140 mg/dl) in the LPDS (1: 9, vol/vol) significantly prolonged the biological activity of PGI2 (Figure 1A). Free PGI2 levels in LPDS were significantly higher than those in HDL-supplemented serum (Figure 1B). The biological half-life of PGI2 was prolonged in HDL-supplemented serum.
Total Cholesterol and HDL Cholesterol Levels Table 1 shows serum total cholesterol and HDL cholesterol levels in the five groups. The four groups with CAD had significantly lower serum HDL cholesterol levels than did the control group. However, serum total cholesterol levels were not significantly different among the five groups.
30
60
90 (min)
FIGURE 1. Plot of effect of high density lipoprotein (HDL) on the biological activity ofprostaglandin I2 (PGI2). Panel A: 10 ng sodium PGI2 was incubated in 1 ml of 50 mM Tris-HCl buffer at pH 7.4, lipoprotein-deficient serum (LPDS) (1:9, vol/vol) solution (e).After incubation, the activity ofPGI2 was assayed by platelet aggregation. Ten microliters of LPDS solution was used for platelet aggregation bioassay. Addition of HDL (final apolipoprotein A-I concentration, 140 mg/dl) in the LPDS (1: 9, vol/vol) significantly prolonged the biological half-life of PGI2 (o). Panel B: Free PGI2 (6-keto-PGF1J levels were also determined by radioimmunoassay in parallel with the platelet aggregation bioassay. In LPDS solution (a), PGI2 did not bind to LPDS from the start. In HDLsupplemented serum (o), free PGI2 was increased with time. Percent inhibition indicates what percent of the platelet aggregation induced by 10 MuM ADP was suppressed by PGI2containing solution. Multiple comparison test with two-way analysis of variance was used for statistical analysis. Error bar represents mean ±SD offive experiments. HDL-supplemented serum vs. LPDS, p
Prostaglandin Half-life Regulated by High Density Lipoprotein Is Decreased in Acute Myocardial Infarction and Unstable Angina Pectoris 12
Takeshi Aoyama, MD, Yoshiki Yui, MD, Hiroshi Morishita, MD, and Chuichi Kawai, MD To investigate the prostaglandin 12 (PGI2) half-life regulated by high density lipoprotein (HDL) in patients with coronary artery disease (CAD), we determined the stability of PGI2 and serum apolipoprotein A-I (Apo A-I) and apolipoprotein A-II (Apo A-1I) levels in four age-matched groups of patients: controls (n=17), angina pectoris (n=18), unstable angina pectoris (n=17), myocardial infarction (n = 19) (acute phase, 3.6+1.7 hours from onset; subacute phase, 75±15 hours from onset in the same patients). Serum PGI2 half-life and total serum Apo A-I levels were lower in the CAD group than in the control group. PGI2 was least stable in patients with unstable angina and the acute phase of myocardial infarction. In these patients, the molar ratio of Apo A-I to Apo A-IT and HDL-associated Apo A-I levels were decreased, and free Apo A-I levels were increased. After in vitro incubation of HDL with increasing amounts of Apo A-II, Apo A-I in HDL was displaced by Apo A-TI, with the parallel decrease in stability of PGI2. Free Apo A-I cannot stabilize PGI2. HDL-associated Apo A-I, whose amount is affected by Apo A-II, stabilized PGI2 and correlated well with stability of PGI2 in patients with CAD and control patients. Decreased PGI2 half-life may play an important role in the pathogenesis of atherosclerosis and thrombus formation in the coronary arteries, especially thrombus formation during an acute coronary event. (Circulation 1990;81:1784-1791)
rostaglandin (PGI2) is a powerful vasodilator and inhibitor of platelet aggregation and also has a cytoprotective role. It has been thought to play a role in the prevention of thrombosis and atherosclerosis.1 PGI2 is unstable in aqueous solution at physiological pH and temperature and is stabilized by binding to an unknown serum factor.23 A reduced stability of PGI2 has been observed in thrombotic disorders (thrombotic thrombocytopenic purpura4 and ischemic stroke5), and this reduction may be related to the pathogenesis of these diseases. Recently, we identified serum PGI2 stabilizing factor as apolipoprotein A-I (Apo A-I), a major apolipoprotein of high density lipoprotein (HDL).6 Our findings indicate that HDL or Apo A-I may exert its antiatherogenic and antithrombogenic effects partly through stabilization of PGI2. In this study, we found P
'2
From the Third Division, Department of Internal Medicine, Faculty of Medicine, Kyoto University, Kyoto, Japan. Supported by a research grant for cardiovascular diseases from the Ministry of Health and Welfare and grants from the Ministry of Education, Science and Culture, Tokyo, Japan. Address for correspondence: Chuichi Kawai, MD, Third Division, Department of Internal Medicine, Faculty of Medicine, Kyoto University, Kyoto 606, Tokyo, Japan. Received July 24, 1989; revision accepted January 24, 1990.
a reduced stability of PGI2 in patients with coronary artery disease (CAD), and we examined the mechanism by which such a reduction occurs.
Methods PGI2 Binding Activity [3H]-PGI2 methyl ester (New England Nuclear Corp., Boston, Massachusetts) was reduced to [3H]PGI2 sodium. [3H]-PGI2 sodium was incubated with 1.0 ml serum or HDL solution at room temperature for 3 minutes and was applied to a Sephadex G-25 See p 2013 column (1.5x25 cm) equilibrated with 50 mM TrisHCl, pH 7.4 buffer. The flow rate was 1 ml/min. Onemilliliter fractions of the gel filtrate were collected, and the radioactivity of each fraction was determined in a liquid scintillation counter (LS 1701, Beckman Instruments, Fullerton, California). Because the radioactivity recovered in all fractions was 99-100% of the radioactivity added, the recovered radioactivity count was used as the basis for determining the percentage of binding. The PGI2 was eluted in two peaks. The bound PGI2 was eluted in the first peak and the free PGI2 in the second peak. The PGI2
Downloaded from http://circ.ahajournals.org/ by guest on July 3, 2015
Aoyama et al Prostaglandin l2 Half-life Regulated by HDL
binding activity was expressed as the percentage of radioactivity in the binding peak.6 Preparation of HDL, Lipoprotein-Deficient Serum, Apo I, and Apo II HDL was isolated from human serum by preparative ultracentrifugation at densities between 1.063 and 1.21 g/ml and purified by recentrifugation.7 Lipoprotein-deficient serum (LPDS) was also obtained from human serum by ultracentrifugal flotation at densities greater than 1.25 g/ml. Apo A-I and Apo A-II were isolated from HDL by ion-exchange and gel permeation chromatographic techniques.8,9
Measurement of Cholesterol and Apo A-I and A-II Levels Cholesterol levels in serum and HDL were determined by enzymatic procedures (Wako Pure Chemical Co, Tokyo). Total serum Apo A-I and Apo A-II levels were determined by the single radial immunodiffusion method with Apo A-I and Apo A-IL plates (Daiichi Chemical Co, Ltd, Tokyo). Free Apo A-I and HDL-associated Apo A-I levels were determined by the method of Borresen and Berg.10 Measurement of Protein Protein was determined by the method of Lowry et al.11
Platelet Aggregation The biological stability of sodium PGI2 (10 ng/ml) (Ono Pharmaceutical Co, Osaka) was determined by platelet aggregation. ADP-induced platelet aggregation was measured with a Lumi-Aggregometer (model 400, Chrono-Log Corp, Havertown, Pennsylvania). Blood (9 ml) was drawn into a plastic syringe with 3.8% sodium citrate (1 ml) as the anticoagulant agent. Samples were centrifuged in plastic tubes successively at 300 and 2,000g for 15 minutes to collect platelet-rich and platelet-poor plasma, respectively, as needed to give a final platelet count of 300,000/mm3. Aggregation was induced in 0.2-ml aliquots of platelet-rich plasma by the addition of 0.02 ml ADP in saline solution to give a final concentration of 10 gM.12 Measurement of 6-Keto-PGF,a Levels 6-Keto-PGFl,, levels were determined by radioimmunoassay as previously described.13 [3H]-6-KetoPGFl, was obtained from New England Nuclear Corp, Boston, Massachusetts. Authentic 6-keto-PGF1a was
supplied by the Ono Pharmaceutical Co, Osaka. Anti6-keto-PGFla was purchased from Advanced Magnet-
ics, Inc., Cambridge, Massachusetts.
Study Population We studied male patients with chest pain of cardiac origin and an age-matched group of patients with pain due to noncardiac causes (control group). Patients with chest pain were divided into three groups: patients with stable angina pectoris, patients
1785
with unstable angina pectoris, and patients with acute myocardial infarction. The study group consisted of 71 patients: 17 control patients, 18 patients with stable angina pectoris, 17 patients with unstable angina pectoris, 19 patients with myocardial infarction (acute phase, 3.5 + 1.7 hours from the onset; subacute phase, 75 + 15 hours from the onset in the same patients). These four groups of patients were age matched. All patients underwent diagnostic coronary angiography for chest pain or suspected coronary artery disease. Coronary angiography was performed with the Judkins or Sones technique, and multiple views of the right and left coronary arteries were recorded. The angiograms were reviewed without prior knowledge of the patients' lipid, Apo A-I, and Apo A-II levels. Stable angina pectoris was defined as chest pain occurring during exercise and lasting less than 10 minutes in patients with objective evidence of coronary angiographic or electrocardiographic changes indicating ischemia during episodes of chest pain. The presence of serious stenosis (>75%) in at least one major epicardial vessel was confirmed in all patients by angiography. Unstable angina pectoris was defined as chest pain occurring at rest for a minimum of 15 minutes on at least one occasion in the absence of an increase in the MB fraction of plasma creatine kinase in patients with objective evidence of CAD; such evidence was based on coronary arteriography and electrocardiographic changes indicating ischemia during episodes of chest pain. The presence of serious stenosis (>90%) in at least one major epicardial vessel was confirmed in all patients by angiography. Postinfarction angina and vasospastic angina were excluded. Myocardial infarction was defined as chest pain occurring at rest in a patient with an associated elevation of the MB fraction of creatine kinase as well as typical electrocardiographic evidence of myocardial necrosis. Emergency coronary arteriography revealed occlusion or subtotal occlusion of a major epicardial vessel. The controls were patients with chest pain due to noncardiac causes whose coronary arteriograms were normal. Patients were excluded from the study if they had undergone any surgical procedure within the month preceding the study or were taking any drugs known to interfere with serum lipids and apolipoprotein levels. Patients receiving aspirin or other nonsteroidal anti-inflammatory agents were also excluded. Blood Sampling In patients with stable and unstable angina pectoris and control patients, venous blood samples were collected through an 8F sheath (Cordis, Co, Ltd, Miami, Florida) in a cardiac catheterization room after a 12-hour fast. In patients with acute myocardial infarction, the first blood sample was collected within 6 hours from the onset through the 8F sheath during the emergency cardiac catheterization and
Downloaded from http://circ.ahajournals.org/ by guest on July 3, 2015
1786
Circulation Vol 81, No 6, June 1990
before any medical intervention (urokinase or tissuetype plasminogen activator). The second blood sample was obtained through the 8F sheath in the coronary care unit 3-5 days after onset and a 12hour fast. Blood samples were collected before heparin administration into plastic tubes and centrifuged at 2,000g for 10 minutes and stored at 4° C. Serum was analyzed within 24 hours after sampling. In Vitro Interaction of Human HDL With Human Apo A-II and Its Relation to PGI2 Binding Activity Human serum HDL from normal fasting male donors was incubated with Apo A-II14,15: 0.8 ml of an HDL sample with Apo A-I content of 1.12 mg (concentration, 140 mg/dl) and an Apo A-II content of 0.32 mg (concentration, 40 mg/dl) was incubated with 0.2 ml of an Apo A-II solution (50 mM TrisHCI, pH 7.4) at concentrations of 0-400 mg/dl. After 1 hour of incubation at room temperature, the PGI2 binding activity and free Apo A-I level of the mixture were determined.
A1
3
:gI
B
104f*1
*
r P
U.
0 0
5, 0
PGI2 Binding Activity in Lipoprotein-Deficient Serum To investigate the ability of free Apo A-I to stabilize PGI2, we examined the PGI2 binding activity in LPDS that contained free Apo A-I.16
---
cr
0.
U.
U,i
Statistical Analysis All data are expressed as mean± SD. The data were analyzed by a multiple comparison test by NewmanKeuls method with one-way analysis of variance. For data in Figure 1, a multiple comparison test with two-way analysis of variance was used. A probability value less than 0.05 was considered significant. Results Effect of HDL on the Stabilization of PGI2 Sodium PGI2 (10 ng/ml) was incubated in 1 ml 50 mM Tris-HCI buffer at pH 7.4 and LPDS (1:9, vol/vol) solution. After 0, 10, 20, 30, 40, 60, and 90 minutes of incubation, the activity of PGI2 was assayed by platelet aggregation. After 0, 30, 60, and 90 minutes of incubation, radioimmunoassay of free PGI2 (6-ketoPGF1l) was also run in parallel with the platelet aggregation bioassay. The addition of HDL (final Apo A-I concentration, 140 mg/dl) in the LPDS (1: 9, vol/vol) significantly prolonged the biological activity of PGI2 (Figure 1A). Free PGI2 levels in LPDS were significantly higher than those in HDL-supplemented serum (Figure 1B). The biological half-life of PGI2 was prolonged in HDL-supplemented serum.
Total Cholesterol and HDL Cholesterol Levels Table 1 shows serum total cholesterol and HDL cholesterol levels in the five groups. The four groups with CAD had significantly lower serum HDL cholesterol levels than did the control group. However, serum total cholesterol levels were not significantly different among the five groups.
30
60
90 (min)
FIGURE 1. Plot of effect of high density lipoprotein (HDL) on the biological activity ofprostaglandin I2 (PGI2). Panel A: 10 ng sodium PGI2 was incubated in 1 ml of 50 mM Tris-HCl buffer at pH 7.4, lipoprotein-deficient serum (LPDS) (1:9, vol/vol) solution (e).After incubation, the activity ofPGI2 was assayed by platelet aggregation. Ten microliters of LPDS solution was used for platelet aggregation bioassay. Addition of HDL (final apolipoprotein A-I concentration, 140 mg/dl) in the LPDS (1: 9, vol/vol) significantly prolonged the biological half-life of PGI2 (o). Panel B: Free PGI2 (6-keto-PGF1J levels were also determined by radioimmunoassay in parallel with the platelet aggregation bioassay. In LPDS solution (a), PGI2 did not bind to LPDS from the start. In HDLsupplemented serum (o), free PGI2 was increased with time. Percent inhibition indicates what percent of the platelet aggregation induced by 10 MuM ADP was suppressed by PGI2containing solution. Multiple comparison test with two-way analysis of variance was used for statistical analysis. Error bar represents mean ±SD offive experiments. HDL-supplemented serum vs. LPDS, p
