Impaired Function of Inhibitory G Proteins During Acute Myocardial Ischemia of Canine Hearts and Its Reversal During Reperfusion and a Second Period of Ischemia
Possible Implications for the Protective Mechanism of Ischemic Preconditioning
Abstract A brief antecedent period of myocardial ischemia and reperfusion can delay cellular injury during a subsequent ischemic condition. Recent observations suggest that this protective mechanism depends on the continued activation of adenosine A1 receptors and Gi proteins. During acute myocardial ischemia, sufficient amounts of adenosine for maximal activation of adenosine A1 receptors are released, independent of a preconditioning ischemia. Hence, the protective mechanism of ischemic preconditioning may not exclusively be explained by activation of adenosine A1 receptors. As a working hypothesis, an increased responsiveness of Gi proteins toward receptor-mediated activation, leading to an increased response of Gi-regulated effectors, was tested in this study. In 47 anesthetized dogs, ischemia was induced by proximal ligation of the left anterior descending coronary artery. Animals underwent either a single period of 5 minutes of ischemia (n=9), a single period of 15 minutes of ischemia (n=10), 5 minutes of ischemia followed by 15 minutes of reperfusion (n=8), 15 minutes of ischemia followed by 60 minutes of reperfusion (n=5), or 5 minutes of ischemia followed by 15 minutes of reperfusion and a second period of 5 minutes of ischemia (n=15). Sarcolemmal membranes were prepared from the central ischemic area and from the posterior left ventricular wall, which served as the control. During ischemia, carbachol-stimulated GTPase decreased by 38% (control, 33.5±17.7; ischemia, 24.2±15 pmol · min−1 · mg protein−1; n=9; P<.001). The decrease in carbachol-stimulated GTPase activity was associated with a 45% decrease in carbachol-mediated inhibition of adenylyl cyclase (control, 28.9±2.4% maximal inhibition; ischemia, 15.1±2.6% maximal inhibition; n=5; P<.001). Prolongation of the ischemic period to 15 minutes did not lead to a further reduction of the Gi-mediated signal transduction. The binding properties of muscarinic receptors were not affected by ischemia. Furthermore, as demonstrated by carbachol-stimulated binding of [γ-35S]GTP to sarcolemmal membranes, high- and low-affinity binding sites for the muscarinic antagonist carbachol, the EC50 for carbachol-stimulated GTPase activity and the substrate dependency of the high-affinity GTPase, the interaction between muscarinic receptors and inhibitory G proteins, and GTP binding to G proteins were not altered (n=14). Immunoblotting with αi- and αi2-specific antibodies did not indicate a loss of Gi proteins during ischemia that could explain the reduced GTPase activity. During 15 minutes of reperfusion, carbachol-stimulated GTPase activity increased to 147% of the control value (control, 33.7±20.6; reperfusion, 49.1±22.5 pmol · min−1 · mg protein−1; n=7; P=.012). Maximal inhibition of adenylyl cyclase by carbachol increased similarly (control, 21±6.8% maximal inhibition; reperfusion, 26.4±7.6% maximal inhibition; n=8; P=.016). After 15 minutes of ischemia and 60 minutes of reperfusion, carbachol-stimulated GTPase activity remained increased. When the 5-minute ischemia and 15-minute reperfusion periods were followed by a second period of 5-minute ischemia, carbachol-stimulated GTPase activity and inhibition of adenylyl cyclase remained elevated (GTPase: control, 38.4±16.7; second ischemia, 49.2±20.1 pmol · min−1 · mg protein−1; n=13; P=.009; adenylyl cyclase: control, 24.2±6.8% maximal inhibition; second ischemia, 28.6±8% maximal inhibition; n=15; P=.003). In conclusion, the responsiveness of Gi proteins toward receptor activation decreased rapidly during the first 5 minutes of ischemia. During a following 15-minute period of reperfusion, this decreased responsiveness was reversed completely, exceeding control activities. The increased responsiveness of this signaling pathway was maintained during a subsequent second ischemic period. This suggests that the underlying mechanism of ischemic preconditioning is the increased responsiveness of Gi proteins after a brief period of ischemia and reperfusion.
- myocardial ischemia
- ischemic preconditioning
- G proteins
- muscarinic receptors
- adenylyl cyclase
Single or repetitive brief periods of ischemia and reperfusion delay cellular injury,1 slow energy metabolism,2 and reduce the incidence of ventricular tachyarrhythmias3 during a subsequent more sustained period of myocardial ischemia. The infarct size–limiting effect of ischemic preconditioning is lost after pretreatment with adenosine receptor antagonists.4 Pretreatment with pertussis toxin also prevents ischemic preconditioning, indicating that the activation of Gi proteins is crucial.5 In addition, pretreatment with A1-selective agonists (but not A2-selective agonists6 ) or with the muscarinic M2 receptor agonist carbachol5 before ischemia causes a delay in cell necrosis similar to ischemic preconditioning. This finding is also consistent with a role for Gi proteins in the mechanism of ischemic preconditioning, since in the myocardium both receptors are coupled to Gi proteins.
However, the underlying mechanisms of myocardial protection in the preconditioned myocardium during ischemia are still unknown. It has been shown very recently that adenosine receptor stimulation not only induces preconditioning during the initial brief period of ischemia but also mediates the protective effect during a subsequent sustained ischemic period.7 However, during every episode of ischemia increased amounts of adenosine are released,8 which are sufficient for maximal activation of A1 receptors.9 Hence, ischemic preconditioning should result in an improved effectiveness of adenosine, and an increased responsiveness of Gi-mediated signaling pathways may be suggested as the underlying mechanism. Accordingly, the influence of ischemia, reperfusion, and subsequent ischemia on the Gi-mediated signal transduction was investigated in the present study.
Materials and Methods
Animal Protocol and Experimental Protocol
Male beagles (age, 1 to 2 years; body weight, 10 to 16 kg) were premedicated by injection of 0.12 mg/kg IM methylparaben (Combelen, Bayer Leverkusen) and anesthetized by application of 15 mg/kg IV sodium pentobarbital (Nembutal, Ceva Düsseldorf). After oropharyngeal intubation the dogs were ventilated with a respirator (Engström Respirator System 300, LKB Medical) with a mixture of 5 L N2 per minute and 2 L O2 per minute. A left thoracotomy was performed in the fourth left intercostal space. An ECG (leads I through III, aVR through aVF) was recorded with subcutaneous needles serving as limb electrodes. The proximal part of the left anterior descending coronary artery (LAD) was prepared and ligated for 5 or 15 minutes. Regional myocardial ischemia was verified by ST-segment deviations in the ECG, the occurrence of the typical epicardial cyanotic area,10 and the use of a research ultrasonic transit-time flowmeter (T106, Transonic Systems Inc). For flow detection, a perivascular flow probe was used that was mounted around the LAD distal to the ligating thread. The average flow under control conditions was 12 to 15 mL/min; no flow could be detected during ligation of the LAD. During early reperfusion flow rapidly increased up to 60 to 100 mL/min and then gradually decreased to baseline levels during a period of 10 to 15 minutes. After the indicated time intervals the hearts were excised and immediately placed in ice-cold saline (0.75 mol/L NaCl, 20 mmol/L HEPES, pH 7.8). Subsequently, the central ischemic area supplied by the LAD (1.5 to 2.5 g) and an equal-sized part from the posterior wall of the left ventricle were separated. The same procedure was performed in sham-operated animals, in which the LAD was not ligated. From each of these two myocardial regions highly purified sarcolemmal vesicles were prepared according to the method of Jones et al11 with slight modifications.12 This membrane preparation has been extensively characterized under control conditions.12 Protein content of the sarcolemmal membranes was determined with the use of the BIO-RAD Bradford dye-binding assay. The animal protocol has been approved by the official animal care committee.
To rule out effects that were not related to ischemia (eg, anesthesia), each animal served as its own control, and ischemic and nonischemic tissues were prepared under exactly the same conditions.
Binding of [3H]N-Methylscopolamine
Equilibrium binding of muscarinic receptor antagonists to purified sarcolemmal membranes was determined by incubation of [3H]N-methylscopolamine with 3 μg sarcolemmal protein suspended in 500 μL of a reaction mixture containing 5 mmol/L MgCl2, 20 mmol/L Tris HCl (pH 7.5), and 5 mmol/L NaH2PO4 (120 minutes at 30°C). The amount of sarcolemmal protein was within the range of a linear increase of [3H]N-methylscopolamine binding, regardless of whether sarcolemmal membranes were from ischemic or nonischemic myocardial tissue. Binding reactions were stopped by rapid filtration through Whatmann GF/B filters with the use of a Brandel cell harvester. Filters were washed three times with 3 mL of ice-cold 50 mmol/L Tris HCl (pH 7.5) and placed in 3 mL scintillation liquid (Aquasure High Performance LSC Cocktail, Du Pont). Radioactivity was determined by liquid scintillation spectroscopy (Multi-User LSC LB 5004 6F Betaszint). Specific binding was defined as the difference in [3H]N-methylscopolamine bound in the absence and presence of 10 μmol/L atropine sulfate.
Binding of Carbachol
Binding of carbachol was indirectly measured by competition experiments. Sarcolemmal membranes were preincubated for 30 minutes (pH 7.5, 30°C) in a reaction medium containing 50 mmol/L Na2HPO4/NaH2PO4, 5 mmol/L MgCl2, the ionophore alamethicin (wt alamethicin/wt protein=1.0), either with or without 10 μmol/L (β,γ-imido)-guanosine-5′-triphosphate [Gpp(NH)p]. The reaction was started by simultaneous addition of [3H]N-methylscopolamine (final concentration, 500 pmol/L) and increasing amounts of carbachol (final concentrations, 5 nmol/L to 50 μmol/L). The other experimental steps were performed according to the description of [3H]N-methylscopolamine binding.
Binding of [γ-35S]GTP
To determine binding of [γ-35S]GTP, 0.8 to 1.0 μg of sarcolemmal proteins were incubated in a reaction buffer (100 μL final volume) containing 1 μmol/L GDP, 50 mmol/L Tris HCl, 5 mmol/L MgCl2, 1 mmol/L EDTA, 150 mmol/L NaCl, and [γ-35S]GTP/[γ-S]GTP at increasing concentrations (pH 7.5, 30°C). Unspecific binding was determined in the presence of 10 μmol/L unlabeled [γ-S]GTP. The binding reaction was started by addition of sarcolemmal membranes and stopped after 60 minutes by rapid filtration through Whatmann GF/B filters with the use of a Brandel cell harvester. The incubation time of 60 minutes was enough to reach binding equilibrium under ischemic and nonischemic conditions. The dependence of [γ-35S]GTP binding on protein concentration was linear up to 1 μg protein per 100 μL. The filters were washed four times with 3 mL of an ice-cold buffer containing 50 mmol/L Tris HCl (pH 7.5) and 5 mmol/L MgCl2 and subsequently placed in 3 mL scintillation liquid (Aquasure High Performance LSC Cocktail; Du Pont). Radioactivity was determined by liquid scintillation spectroscopy as described above.
GTPase activity was determined as described by Fleming and Watanabe.13 GTP hydrolysis was directly assessed by measuring the release of [32Pi] from [γ-32P]GTP. The transfer of [γ-32P] from GTP to ADP was suppressed by an ATP-regenerating system. The enzyme reaction was started by the addition of sarcolemmal membranes (3 μg protein per 100 μL assay volume) to the reaction medium containing 1 mmol/L EDTA, 40 mmol/L Tris HCl (pH 7.5), 0.5 mmol/L ATP, 0.5 mmol/L (β,γ-imido)-adenosine-5′-triphosphate [App(NH)p], 1 mmol/L dithiothreitol, 5 mmol/L MgCl2, 5 mmol/L creatine phosphate, 4 U/L creatine kinase, 1 mg/mL bovine serum albumin, 0.2 μmol/L GTP, and 36 to 40 nCi [γ-32P]GTP (37°C). App(NH)p was used for the inhibition of nonspecific nucleoside triphosphatases. To suppress basal β-adrenoreceptor–Gs interaction, 10 μmol/L alprenolol was added to the assay. The reaction was stopped after 10 minutes by the addition of 900 μL ice-cold activated charcoal suspension (Norit A, 20 mmol/L phosphoric acid, 5% wt/vol). This mixture subsequently was centrifuged at 2500g (10 minutes, 4°C), and 500 μL of the supernatant was added to 4 mL Tris HCl (pH 7.5). The amount of hydrolyzed [32P]phosphate in the solution was determined by liquid scintillation spectroscopy as described above. Specific GTPase activity was calculated as the difference between total activity and background activity, obtained in the absence of sarcolemmal membranes in the assay. “Basal” GTPase activity is defined as activity in the absence of a receptor agonist (carbachol).
The values of GTPase activity as well as the enzyme characteristics measured in the present study under control conditions were similar to those previously reported by Fleming and Watanabe.13 Basal (in the absence of carbachol) and total (in the presence of 1 mmol/L carbachol) GTPase activity were linearly dependent on protein concentration (0.8 to 6.0 μg per assay). The amount of hydrolyzed GTP linearly increased within the time range from 1 to 30 minutes in the presence and absence of carbachol under control conditions as well as after 15 minutes of ischemia. The dependence of basal and total GTPase activity on GTP concentration also exhibited saturation kinetics and was measured under ischemic and nonischemic conditions (see “Results”). The effect of carbachol (1 mmol/L) on high-affinity GTPase activity was blocked completely by the addition of 100 μmol/L atropine sulfate.
Treatment of Sarcolemmal Vesicles With Alamethicin
The sidedness of the present preparation of sarcolemmal vesicles has been extensively studied by Colvin et al12 under control conditions. To exclude unspecific effects resulting from different amounts of latent enzyme activities due to different sidedness of vesicles gained from ischemic and nonischemic tissues, most experiments were performed in the presence of the antibiotic ionophore alamethicin. Alamethicin was preincubated with sarcolemmal membranes (20 minutes, 25°C; wt alamethicin/wt protein=1.0) before starting the biochemical reaction.
Adenylyl cyclase activity was determined by measuring the conversion of [α-32P]ATP to [32P]cAMP according to Jakobs et al.14 The assay volume was 100 μL, containing 0.1 mmol/L ATP with 2×105 cpm of [α-32P]ATP, 5 mmol/L MgCl2, 0.1 mmol/L cAMP, 10 μmol/L GTP, 1 mmol/L EDTA, 0.5 mmol/L dithiothreitol, 5 mmol/L creatine phosphate, 2 U creatine kinase, 0.05 mg bovine serum albumin, and 75 mmol/L Tris HCl (pH 7.6). The reaction was started by the addition of sarcolemmal membranes (2.0 to 3.0 μg protein) and continued for 10 minutes at 37°C. Specific adenylyl cyclase activity was calculated as the difference between total activity and background activity, obtained in the absence of sarcolemmal membranes in the assay. To unmask latent adenylyl cyclase activity, membranes (2.0 to 3.0 μg protein) were preincubated with alamethicin for 20 minutes at 25°C as described above. A sample that contained [3H]cAMP instead of labeled ATP was run in parallel to determine the yield of cAMP in the assays. Under these conditions adenylyl cyclase activity was linear with time (1 to 20 minutes) and protein content (0.1 to 3.0 μg). Carbachol (1 mmol/L)–mediated effects on adenylyl cyclase activity were blocked completely by the addition of 100 μmol/L atropine sulfate.
Experiments with sham-operated animals showed that there was no inherent difference of specific adenylyl cyclase activity between the left ventricular anterior and posterior walls. Furthermore, adenylyl cyclase activity was similar in the control area of the left ventricular posterior wall regardless of whether the LAD was ligated or not. The use of highly purified sealed sarcolemmal vesicles and alamethicin, which was added to each assay to unmask latent enzyme activities, was associated with an increased variability of basal adenylyl cyclase activity (defined as activity in the presence of 10 μmol/L GTP and absence of a receptor ligand) among different assays (see “Results”). However, the SD within each assay (triplicate determinations) did not exceed 5% of the mean value. Hence, since each animal served as its own control and membrane preparation and all assays were run in parallel for every animal, the effects of ischemia and reperfusion were highly reproducible, with only slight variations among all animals studied.
Quantification of Giα, Giα2, and Gsα
Sarcolemmal proteins were separated by disc sodium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis. Thirty micrograms of sarcolemmal proteins was suspended in 200 μL buffer containing 2% SDS, 40 mmol/L Tris HCl (pH 6.8), 10% glycerol, 0.006% bromphenol blue (solubilized in ethanol), and 256 mmol/L β-mercaptoethanol before application to the gel. For reference, molecular weight standards (30 to 200 kD) and transducin were used. Electrophoresis was performed for 6 hours with a current of 30 mA with a running buffer consisting of 0.1% SDS, 200 mmol/L glycine, and 25 mmol/L Tris HCl. The separated proteins were transferred to nitrocellulose filters BA 83 (Schleicher & Schüll) with a transfer buffer consisting of 27.5 mmol/L Tris HCl, 190 mmol/L glycine, and 20% methanol. Unspecific binding sites of nitrocellulose were blocked by 2.5% bovine serum albumin in 0.9% NaCl and 50 mmol/L Tris HCl (pH 7.4). The nitrocellulose filters then were incubated for 3 hours at room temperature with specific peptide antibodies against Gsα and Giα (NEN–Du Pont). Both primary antisera were probed with the nitrocellulose blots simultaneously. The antibody against Giα is specific for the subtypes 1 and 2, cross-reacts with transducin, and does not recognize Giα3 or Go. The amount of Gsα and Giα on the nitrocellulose filters was visualized by autoradiography with protein A-3-(4-hydroxy,5-[125I]iodo-phenyl)propionamid as radioactive marker. Unspecific binding of labeled protein A was eliminated by washing the filters three times with a solution consisting of 150 mmol/L NaCl, 20 mmol/L Tris HCl (pH 7.4). The nitrocellulose filters then were exposed at −80°C for 3 days with the use of Kodak X-Omat films. The autoradiograms were analyzed by computer-assisted laser densitometry (LKB-2400-Gel-Scan XL, Fa Pharmacia). Quantitative analysis revealed that Western blot analysis was linear in a range of 10 to 100 μg sarcolemmal proteins applied (data not shown). Because myocardial Giα is predominantly represented by the subunit Giα2,15 this procedure was repeated with the same immunoblots after washing off previous peptide antibodies specific for Giα1,2 (and Gsα) with glycine (0.1 mol/L, pH 2.5, 1 hour). Completion of the washing procedure was verified by autoradiograms after incubating the nitrocellulose filters with [125I]protein A. Thereafter, the washed nitrocellulose blots were exposed to the specific antibodies against subtype Giα2. The following labeling of the membranes with Giα2 and [125I]protein A was as described above.
Samples from the control and ischemic areas of the same heart were always run in parallel in the same experiment, using the same primary antiserum.
A total of 54 animals were studied. Seven animals were sham-treated, and 47 animals were used for the interventions. The following protocols were used: 5 minutes of ischemia (n=9); 15 minutes of ischemia (n=10); 15 minutes of ischemia/60 minutes of reperfusion (n=5); 5 minutes of ischemia/15 minutes of reperfusion (n=8); and 5 minutes of ischemia/15 minutes of reperfusion/5 minutes of ischemia (n=15).
Because of the small sample size within the ischemic myocardium and the low yield of sarcolemmal protein from the purification, not all experiments could be performed with each animal. Carbachol-stimulated GTPase and/or adenylyl cyclase inhibition was determined in all animals. Binding of [3H]N-methylscopolamine was determined in 4 dogs subjected to 15 minutes of ischemia. Quantification of Gi proteins was assessed in 6 animals subjected to 15 minutes of ischemia (n=3 for pertussis toxin–catalyzed ADP ribosylation and immunoblotting with Gi1,2- and Gi2-specific antibodies, respectively).
The experiments to assess the interaction between receptor and G protein were carried out in 14 animals, as follows: (1) high- and low-affinity binding sites for carbachol: n=4, 5 minutes of ischemia; (2) carbachol-stimulated binding of [γ-35S]GTP: n=4, 15 minutes of ischemia; (3) EC50 for carbachol-stimulated GTPase: n=4, 15 minutes of ischemia; and (4) substrate dependency of the high-affinity GTPase with and without carbachol: n=3, 15 minutes of ischemia.
All values are given as mean±SD. Student’s paired t test was commonly used for statistical analysis. Statistical analysis of the dose-response effects (Figs 4⇓, 6⇓, and 7⇓) was performed by means of ANOVA followed by the Student-Newman-Keuls test. Each dog used in this study served as its own control in that the results obtained from the left ventricular posterior wall (control area) were compared with those obtained from the myocardial area supplied by the LAD. For each series of experiments a total of at least four measurements with different membrane preparations from at least three animals were performed; n always indicates the number of individual animals in each set of experiments. Furthermore, all experiments were carried out by means of triplicate determinations. The SD within each triplicate determination in the assay did not exceed 5% of the mean value.
All chemicals were reagent grade. [3H]N-methylscopolamine, guanosine-5′-[γ-32P]triphosphate, [γ-35S]guanosine-5′-O-(3-thiotriphosphate), and antibodies against Giα1,2 and Gsα were obtained from NEN–Du Pont. Unlabeled nucleotides were obtained from Boehringer. If not indicated otherwise in the text, the other chemicals were purchased from Sigma Chemie GmbH. Antibodies against Giα2 were obtained from Gramsch Laboratories.
Binding of [3H]N-Methylscopolamine
In sham-treated animals (n=4), binding characteristics for the muscarinic antagonist [3H]N-methylscopolamine were similar in sarcolemmal membranes purified from the anterior and posterior left ventricular wall (Table 1⇓).
After 15-minute ligation of the LAD, neither binding capacity nor binding affinity of muscarinic receptors was changed (Table 1⇑, Fig 1⇓; n=4). Unspecific binding was <10% of total binding and did not change under ischemic conditions.
Binding of Carbachol
Under control conditions and after 5 minutes of ischemia, computerized curve-fitting analysis detected two binding states for carbachol (control: Kdhigh=36.7 nmol/L, 62% of total binding capacity, Kdlow=3.3 μmol/L; ischemia: Kdhigh=36.8 nmol/L, 68% of total binding capacity, Kdlow=6.1 μmol/L; Fig 2⇓). Under both conditions the presence of Gpp(NH)p resulted in a conversion of high-affinity sites to low-affinity sites. According to curve-fitting analysis, in the presence of Gpp(NH)p the two-site model was not superior to the one–binding site model (control: Kd=1.2 μmol/L; ischemia: Kd=0.9 μmol/L; Fig 2⇓; n=4). Similar results were obtained in sham-treated animals (n=4) in membranes obtained from the anterior and posterior left ventricular wall.
Binding of [γ-35S]GTP
Binding of [γ-35S]GTP to sarcolemmal membranes was studied up to a concentration of 30 nmol/L. Carbachol stimulated the binding of [γ-35S]GTP to sarcolemmal membranes only at low concentrations of [γ-35S]GTP. The maximal carbachol-induced increase in [γ-35S]GTP binding was observed at 2 nmol/L of [γ-35S]GTP. At higher concentrations (15 to 30 nmol/L) carbachol-stimulated binding was no longer observed, indicating that carbachol only increased binding affinity without affecting binding capacity for [γ-35S]GTP. In addition, at concentrations >4 to 8 nmol/L, binding of [γ-35S]GTP reincreased because of the coexistence of distinct binding sites in the sarcolemmal membranes. After 15 minutes of myocardial ischemia, neither basal nor carbachol (1 mmol/L)–stimulated [γ-35S]GTP binding was significantly changed (Table 2⇓). These data were obtained by Scatchard analysis of the high-affinity binding site. Unspecific binding of [γ-35S]GTP was <10% of total binding and similar under both conditions. Similar results were obtained with sham-treated animals (n=4).
Quantification of G Proteins
Western blot analysis that used specific peptide antibodies resulted in single bands with a molecular weight of 38 kD for transducin, 41 kD for Giα, and 42 kD for Gsα. Quantitative analysis of the amount of Giα1,2 and Gsα in sarcolemmal membranes did not reveal significant differences between the ischemic and the nonischemic areas of the left ventricle (Fig 3⇓). Similarly, comparing the nonischemic myocardium of the left and the right ventricles with the ischemic myocardial area of the anterior wall of the left ventricle, subgroup analysis that used specific antibodies against Giα2 did not reveal significant differences in the amount of Giα2 (Table 3⇓).
In accordance with these results, pertussis toxin–catalyzed ADP ribosylation of sarcolemmal proteins remained unchanged after 15 minutes of ischemia.16
Basal GTPase activity, defined as the enzymatic activity in the absence of carbachol, was 27.4±3.5 pmol · min−1 · mg protein−1 in controls (posterior wall of the left ventricle; n=4). Similar values were obtained from the anterior wall of the left ventricle in sham-operated animals (28.3±3.1 pmol · min−1 · mg protein−1; n=4).
High-Affinity GTPase Activity During a Single Period of Ischemia
Fifteen-minute ligation of the LAD resulted in a 23% reduction of basal GTPase activity in sarcolemmal membranes obtained from the anterior wall (21.1±2.2 pmol · min−1 · mg protein−1; n=4; P<.05 versus control; Fig 4⇓).
GTPase activity could be stimulated in a dose-dependent manner by the muscarinic receptor agonist carbachol in both nonischemic and ischemic myocardium (Fig 4⇑). The concentration of carbachol needed for half maximal stimulation was similar under both conditions (control, 8.7±6.3 μmol/L; 15 minutes of ischemia, 5.1±2.0 μmol/L; P=NS). Total GTPase activity in the presence of 1 mmol/L carbachol was 38.4±3.5 pmol · min−1 · mg protein−1 at control conditions and 27.6±2.0 pmol · min−1 · mg protein−1 after 15 minutes of ischemia (n=4; P<.05 ischemia versus control; Fig 4⇑).
The amount of muscarinic receptor–coupled GTPase activity (total GTPase activity in the presence of carbachol minus basal GTPase activity) was significantly reduced by 45% after ischemia (control, 11.0±1.6 pmol · min−1 · mg protein−1; ischemia, 6.5±1.4 pmol · min−1 · mg protein−1; P<.05; carbachol concentration, 1 mmol/L; Fig 4⇑).
Treatment of sarcolemmal membranes with alamethicin resulted in a twofold to fivefold increase of basal and total GTPase activity after stimulation with carbachol (Fig 4⇑, inset). After 5 (n=9) and 15 (n=6) minutes of ischemia, the amount of muscarinic receptor–coupled GTPase activity (total GTPase activity in the presence of carbachol minus basal GTPase activity) again was significantly decreased by 39% to 46% compared with the control area (Figs 5⇓, 6⇓, and 8⇓). Between 5 and 15 minutes there was no significant progression of ischemia-induced GTPase impairment (Fig 5⇓).
The dependence of high-affinity GTPase activity on GTP concentration, measured in the presence of alamethicin, exhibited saturation kinetics (Km=192±11 nmol/L; Vmax=138±16 pmol · min−1 · mg protein−1; n=3). Carbachol increased maximal velocity but did not change Km; ie, the GTP concentration needed for half maximal velocity remained unchanged (Km=200±14 nmol/L, P=NS; Vmax=218±20 pmol · min−1 · mg protein−1, P<.05 [plus carbachol versus without carbachol]; n=3). The ischemia-induced inhibition of GTPase activity exhibited the characteristics of a noncompetitive inhibition; ie, Vmax was reduced, whereas Km remained unchanged (Km=185±15 nmol/L; Vmax without carbachol=102±13 pmol · min−1 · mg protein−1, P<.05 [ischemia versus control]; Vmax plus carbachol=154±19 pmol · min−1 · mg protein−1, P<.05 [ischemia versus control]; n=3; Fig 6⇑).
Carbachol-Stimulated GTPase Activity After a Single Period of Ischemia Followed by Reperfusion
When the LAD was ligated for 15 minutes and subsequently reperfused for 1 hour, the ischemia-induced reduction of GTPase activity was not observed. The level of carbachol-stimulated GTPase activity in the area of intervention exceeded the GTPase activity of the control area by 20% (control, 18.0±2.4; reperfusion, 22.4±3.2 pmol · min−1 · mg protein−1, no alamethicin added; n=5; P<.05; Fig 7⇓).
When the 5-minute ischemic period was followed by 15 minutes of reperfusion, carbachol-stimulated GTPase activity in the reperfused area again exceeded the activity in the control area (33.7±20.6 versus 49.1±22.5 pmol · min−1 · mg protein−1; n=7; P=.012; Fig 8⇓).
Carbachol-Stimulated GTPase Activity During a Second Period of Ischemia After a Short Period of Ischemia and Reperfusion
After 5 minutes of ischemia followed by 15 minutes of reperfusion, a second 5-minute period of ischemia was induced. During this second ischemic period carbachol-stimulated GTPase activity remained elevated (38.4± 16.7 versus 49.2±20.1 pmol · min−1 · mg protein−1; n=13; P=.009; Fig 8⇑).
Isoproterenol-Induced Adenylyl Cyclase Stimulation
Stimulation of GTP-activated adenylyl cyclase by a submaximally active concentration of the β-receptor agonist isoproterenol was not affected by the first or second ischemic period or during reperfusion. After 5 minutes of ischemia, basal adenylyl cyclase activity in the presence of 10 μmol/L GTP increased to 228% by the addition of 100 μmol/L isoproterenol (control, from 382±147 to 869±290 pmol · min−1 · mg protein−1; ischemia, from 258±112 to 609±225 pmol · min−1 · mg protein−1). In the reperfusion group, adenylyl cyclase activity was increased to 223% (control, from 422±272 to 807±485 pmol · min−1 · mg protein−1; reperfusion, from 328±290 to 751±759 pmol · min−1 · mg protein−1), and in the preconditioned group it was increased to 247% by the addition of isoproterenol (control, from 541±254 to 1349±807 pmol · min−1 · mg protein−1; preconditioning, from 388±203 to 1024±730 pmol · min−1 · mg protein−1).
Carbachol-Induced Adenylyl Cyclase Inhibition
Maximal inhibition of isoproterenol (100 μmol/L)–stimulated adenylyl cyclase activity by carbachol was determined. The decrease in carbachol-stimulated GTPase activity observed during 5 minutes of ischemia was accompanied by a similar decrease of the carbachol-induced adenylyl cyclase inhibition. Carbachol reduced isoproterenol-stimulated adenylyl cyclase activity by 28.9±2.4% from 869±290 to 617±192 pmol · min−1 · mg protein−1 in the control group and by 15.1±2.6% from 609±225 to 517±172 pmol · min−1 · mg protein−1 during 5 minutes of ischemia (Fig 9⇓; n=5; P<.001).
The increased carbachol-stimulated GTPase activity during the reperfusion period was accompanied by a similar increase in maximal carbachol-induced adenylyl cyclase inhibition. Carbachol reduced isoproterenol-stimulated adenylyl cyclase activity by 21±6.8% from 807±485 to 636±374 pmol · min−1 · mg protein−1 in the control group and by 26.4±7.6% from 751±759 to 553±579 pmol · min−1 · mg protein−1 in the reperfusion group (n=8; P=.016; Fig 9⇑). This increased inhibition of adenylyl cyclase activity was also maintained during a second period of ischemia. Carbachol reduced isoproterenol-stimulated adenylyl cyclase activity by 24.2±6.8% from 1349±807 to 1022±672 pmol · min−1 · mg protein−1 in the control group and by 28.6±8% from 1024±730 to 737±537 pmol · min−1 · mg protein−1 in the preconditioned group (n=15; P=.003; Fig 9⇑).
Time to Onset of Ischemia-Induced ST-Segment Deviation
The onset of significant ST-segment deviation (>0.1 mV) in the surface ECG was determined during the initial preconditioning ischemia and compared with the second ischemic period. During the initial ischemic period, the onset of significant ST-segment deviations was observed after 44±20 seconds. During the subsequent second 5-minute period of ischemia, the onset of ST-segment deviations was delayed, occurring after 93±37 seconds (n=5; P=.036).
M2 Receptor–Stimulated GTPase Activity and Adenylyl Cyclase Inhibition
To study the responsiveness of Gi proteins toward receptor activation, the carbachol-stimulated GTPase activity and carbachol-induced inhibition of adenylyl cyclase activity were evaluated. In purified sarcolemmal membranes a high-affinity GTPase (Km=0.2 μmol/L) can be detected. The high-affinity GTPase corresponds to the GTPase activity displayed by the α-subunits of heterotrimeric G proteins. Only the high-affinity GTPase responds to stimulation by receptor agonists. By evaluating the carbachol-stimulated GTPase activity (GTPase activity in the presence of carbachol minus basal GTPase activity), only the GTPase of G proteins coupling to muscarinic receptors is assessed. Muscarinic receptor–stimulated high-affinity GTPase in cardiac sarcolemma has been demonstrated to be susceptible to pertussis toxin and furthermore to be closely related to inhibition of adenylyl cyclase.13 Hence, carbachol-stimulated GTPase activity is predominantly expressed by Gi proteins.
Within the cycle of G protein activation and inactivation, the GTPase activity of the α-subunit is the inactivation mechanism (“turn-off” reaction). However, measurement of the carbachol-stimulated GTPase reaction is a reflection of hormone-stimulated GDP release with resulting stimulation of GTP binding and hydrolysis. Hence, it is a direct measure of muscarinic stimulation of the G protein. Accordingly, the changes in GTPase activity due to ischemia or reperfusion were accompanied by concordant changes in adenylyl cyclase inhibition, indicating that the altered activation of the G protein cycle led to alterations of the effector regulation.
During ischemia, activation of Gi proteins is expected to be predominantly induced by activation of adenosine A1 receptors, due to the increased myocardial release of adenosine. With regard to the mechanisms underlying ischemic preconditioning, however, the activation of Gi proteins rather than activation of adenosine A1 receptors seems crucial in that ischemic preconditioning is lost after pretreatment with pertussis toxin.5 Pertussis toxin does not interact with the receptor but catalyzes ADP ribosylation of Gi and Go proteins, thereby uncoupling them from receptor-mediated activation. In addition, cardioprotection during ischemia can be induced by pretreatment with carbachol as well as with A1 agonists.
Because of the very low density of adenosine receptors in sarcolemmal membranes from canine ventricles (≈25 fmol/mg protein in the membrane preparations used in this study; data not shown) and the low yield of sarcolemmal protein from the small ischemic area (1.5 to 2.5 g of tissue), it was not possible to assess the characteristics of the adenosine A1 receptor–coupled signal transduction in this set of experiments. In contrast, M2 receptor density in these purified sarcolemmal membranes amounts to 3 pmol/mg protein (Table 1⇑), and M2-mediated signal transduction is easily assessable. Therefore, in this study the responsiveness of Gi-mediated signal transduction toward receptor activation was studied with the use of the M2 agonist carbachol. The detection of carbachol-stimulated high-affinity GTPase activity afforded the use of a highly purified membrane preparation. Despite the advantages of using highly purified membrane preparations, it is recognized that the purification steps are associated with a loss of membrane proteins, including fractions of the signal transduction components studied here.
Impaired Gi-Mediated Signal Transduction During Ischemia
During early myocardial ischemia, the activation of adrenergic receptors is increased because of an accumulation of catecholamines in the interstitial space.17 In addition, the sensitivity of adrenergic signal transduction may be likewise augmented due to a paradoxical increase in α- and β-adrenergic receptor density18 19 and a receptor-independent sensitization of adenylyl cyclase.20 Activation of adrenergic receptors leads to an increased energy demand with extension of myocardial necrosis21 and may induce malignant ventricular tachyarrhythmias.22 23
Under physiological conditions, an effective endogenous control system exists to antagonize excessive adrenergic signal transduction. This endogenous antagonism of adrenergic signal transduction is mediated by Gi proteins, which can be activated by various receptors such as adenosine A1 and muscarinic M2 receptors. Although a sufficient amount of adenosine for maximal activation of A1 receptors is already released during the first minutes of myocardial ischemia, the endogenous antagonism of adrenergic signal transduction remains insufficient, and exogenous antagonism with β-adrenergic receptor blocking agents remains an effective therapeutic approach.24 This may indicate an impaired signal transduction of Gi proteins during early myocardial ischemia. This concept is supported by the finding that the inhibition of canine cardiac adenylyl cyclase is significantly reduced during the first 5 minutes of myocardial ischemia.16
The activation of Gi proteins leads to a regulation of various effectors, eg, the inhibition of adenylyl cyclase. In addition, Gi proteins mediate activation of atrial potassium channels,25 activation of ventricular ATP-sensitive potassium channels,26 direct inhibition of L-type calcium channels,27 and inhibition of the isoproterenol-stimulated sodium current.28 ATP-sensitive potassium channels may be of major importance because ischemic preconditioning is lost after pretreatment with glibenclamide,29 an inhibitor of these channels. However, in rabbits blockade of the ATP-dependent potassium channels did not prevent ischemic preconditioning.30 Potentially, the increased Gi-mediated regulation of any of the aforementioned effectors could be “protective” during ischemia by decreasing excessive calcium loading and cellular energy demand and by increasing electric stability.
An impairment of the Gi-mediated signal transduction during ischemia, as shown in this study, would hamper the potential benefit of a cholinergic system activation, which has been shown to prevent ischemia-induced malignant tachyarrhythmias.31 32 33 34 35 36 Accordingly, as shown in an earlier study, carbachol elevated the ventricular fibrillation threshold in nonischemic but not in ischemic areas of canine ventricular myocardium.37 Furthermore, it recently has been shown that carbachol-mediated inhibition of postsynaptic norepinephrine release is attenuated in the ischemic myocardium.38
The ischemia-induced impairment of muscarinic receptor signal transduction could be attributed neither to a reduction of muscarinic receptor density or affinity nor to an impaired interaction between muscarinic receptors and inhibitory G proteins, since muscarinic receptor–stimulated GTP binding, agonist binding to muscarinic receptors, and the EC50 for stimulation of GTPase by carbachol all remained unaltered during ischemia. It is known that unoccupied receptors can stimulate basal high-affinity GTPase activities. This activity of unoccupied receptors can be reversed with specific receptor antagonists. A decrease of this “empty receptor” activation of Gi proteins as the cause of decreased basal high-affinity GTPase activity during ischemia was excluded because atropine decreased high-affinity GTPase activity similarly in ischemic and nonischemic membranes. A specific, ischemia-induced loss of the α-subunit population of inhibitory G proteins was also excluded. Although no direct assessment of the levels of βγ-subunits was performed, a selective loss of these subunits despite unchanged levels of the α-subunit within 5 minutes of ischemia is very unlikely and should have been detected by autoradiography that used pertussis toxin for the labeling of Gi proteins with [32P]ADP.16 This is because pertussis toxin catalyzes only the ADP ribosylation of the complete heterotrimeric G protein. In addition, carbachol-stimulated binding of [γ-S]GTP would be altered by a selective loss of βγ-subunits because the receptor can interact only with the heterotrimeric G protein.39 A selective loss of βγ-subunits, relevant to Gi-mediated signal transduction, would reduce the availability of heterotrimeric G proteins, leading to a decrease in receptor-stimulated binding of [γ-S]GTP.
From the above findings it may therefore be concluded that reduced carbachol-stimulated high-affinity GTPase activity is the result of a reversible change of Gi protein function that involves reaction steps after GTP binding. It has been shown that after binding of GTP, G proteins must undergo at least one other reaction step, which is associated with subunit dissociation, before they can interact with their responsive effectors. This reaction step is rate limiting for the activation of the G protein. The noncompetitive characteristics of ischemia-induced inhibition of GTPase (Fig 6⇑) indicate that a part of the available enzyme pool may be reversibly converted to a functional state that is able to bind GTP but that cannot be activated to mediate inhibition of adenylyl cyclase and subsequently catalyze GTP hydrolysis. The characteristics of such a hypothetical functional state of Gi proteins would resemble those recently observed with the mutant cell line H21a.40 Reduction of muscarinic receptor–coupled GTPase activity is a rapid event; it could be observed after 5 minutes of ischemia with no further progression if ischemia was prolonged up to 15 minutes.
In conclusion, the present results suggest an ischemia-induced functional alteration of inhibitory G proteins leading to an impairment of muscarinic receptor–mediated signal transduction.
Increased Gi-Mediated Signal Transduction During Reperfusion and Second Ischemia
Several previous studies did not differentiate between the mechanism initiating ischemic preconditioning (during the preconditioning brief period of ischemia) and the protective mechanism during the subsequent sustained ischemic period. More recently it has been shown that a continued activation of A1 receptors during sustained ischemia is required for the protective mechanism of ischemic preconditioning.7 Maximal activation of adenosine A1 receptors is obtained at concentrations of ≈100 nmol/L adenosine.9 During myocardial ischemia (even in the absence of preconditioning), interstitial adenosine concentration is in the micromolar range.8 Hence, it can be assumed that the effect of preconditioning may be due to an increased responsiveness of the adenosine A1 receptor pathway. Although several findings suggest that the release of adenosine during the preceding brief period of ischemia is essential for the initiation of ischemic preconditioning in most species (except in rats41 ), the present study emphasizes the importance of the reperfusion period for the protective mechanism of preconditioning. The responsiveness of Gi-mediated signal transduction increases only on reperfusion, whereas it decreases during the initial period of ischemia.
Regarding ischemic preconditioning, as stated above, a continued activation of adenosine A1 receptors is required for the protective mechanism. In the present study an increased responsiveness of Gi proteins toward receptor activation during reperfusion and during a second ischemic period was accompanied by an increased carbachol-mediated inhibition of adenylyl cyclase. The relevance of adenylyl cyclase inhibition for the protective mechanism during ischemic preconditioning is not yet established. Indirect evidence may be derived from findings that administration of β-adrenergic receptor blocking agents reduce infarct size42 and mortality from myocardial infarction,24 whereas activation of β-adrenergic receptors and an intracellular increase of cAMP during ischemia are associated with an increase in life-threatening ventricular arrhythmias22 23 and an extension of myocardial necrosis.21
Although it is not definitely known which of the Gi-regulated effectors are involved in the cardioprotective mechanism of ischemic preconditioning, the present study suggests that Gi proteins are sensitized during reperfusion after a single 5-minute period of ischemia and maintain an increased responsiveness toward receptor activation during a following period of ischemia. This sensitization leads to an improved adenylyl cyclase inhibition and may involve other effectors as well.
This study was supported by the Deutsche Forschungsgemeinschaft (SFB 320/B1 and Ni/327/2-1). We wish to thank Ulrike Oehl and Elke Braunwell for excellent technical assistance.
Presented in part at the 63rd to 66th Scientific Sessions of the American Heart Association (1990-1993).
- Received February 14, 1994.
- Accepted January 23, 1995.
- © 1995 American Heart Association, Inc.
Murry CE, Jennings RB, Reimer KA. Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation. 1982;66:1150-1153.
Murry CE, Richard VJ, Reimer KA, Jennings RB. Ischemic preconditioning slows energy metabolism and delays ultrastructural damage during a sustained ischemic episode. Circ Res. 1990;66:913-931.
Shiki K, Hearse DJ. Preconditioning of ischemic myocardium: reperfusion induced arrhythmias. Am J Physiol. 1987;253:H1470-H1476.
Liu GS, Thornton J, Van Winkle DM, Stanley AWH, Olsson RA, Downey JM. Protection against infarction afforded by preconditioning is mediated by A1 adenosine receptors in rabbit heart. Circulation. 1991;84:350-356.
Thornton JD, Liu GS, Olsson RA, Downey JM. Intravenous pretreatment with A1-selective adenosine analogues protects the heart against infarction. Circulation. 1992;85:659-667.
Toombs CF, McGee DS, Johnston WE, Vinten-Johanson J. Myocardial protective effects of adenosine: infarct size reduction with pretreatment and continued receptor stimulation during ischemia. Circulation. 1992;86:986-994.
Olsson RA. Changes in content of purine nucleoside in canine myocardium during coronary occlusion. Circ Res. 1970;26:301-306.
Londos C, Cooper DMF, Schlegel W, Rodbell M. Adenosine analogs inhibit adipocyte adenylate cyclase by a GTP-dependent process: basis for actions of adenosine and methylxanthines on cyclic AMP production and lipolysis. Proc Natl Acad Sci U S A. 1978;75:5362-5366.
Maroko PR, Kjekshus JK, Sobel BE, Watanabe T, Covell JW, Ross J Jr, Braunwald E. Factors influencing infarct size following experimental coronary artery occlusions. Circulation. 1971;43:67-82.
Jones LR, Maddock SW, Besch HR Jr. Unmasking effect of alamethicin on the (Na+,K+)-ATPase, beta-adrenergic receptor-coupled adenylate cyclase, and cAMP-dependent protein kinase activities of cardiac sarcolemmal vesicles. J Biol Chem. 1980;255:9971-9980.
Fleming JW, Watanabe AM. Muscarinic cholinergic-receptor stimulation of specific GTP hydrolysis related to adenylate cyclase activity in canine cardiac sarcolemma. Circ Res. 1988;64:340-350.
Jakobs KH, Saur W, Schultz G. Reduction of adenylate cyclase activity in lysates of human platelets by the alpha-adrenergic component of epinephrine. J Cyclic Nucl Res. 1976;2:381-392.
Holmer RH, Stevens S, Homcy CJ. Tissue- and species-specific expression of inhibitory guanine nucleotide-binding proteins: cloning of a full-length complementary DNA from canine heart. Circ Res. 1989;65:1136-1140.
Schömig A, Dart AM, Dietz R, Mayer E, Kübler W. Release of endogenous catecholamines in the ischemic myocardium of the rat. Circ Res. 1984;55:689-701.
Corr PB, Shayman JA, Kramer JB, Kipnis RJ. Increased alpha-adrenergic receptors in ischemic cat myocardium. J Clin Invest. 1981;67:1232-1236.
Mukherjee A, Wong TM, Buja M, Lefkowitz RJ. Beta adrenergic and muscarinic cholinergic receptors in canine myocardium: effects of ischemia. J Clin Invest. 1979;64:1423-1428.
Strasser RH, Braun-Dullaeus R, Walendzik H, Marquetant R. α1-receptor–independent activation of protein kinase C in acute myocardial ischemia: mechanisms for sensitization of the adenylyl cyclase system. Circ Res. 1992;70:1304-1312.
Schulz R, Guth BD, Pieper K, Martin C, Heusch G. Recruitment of an inotropic reserve in moderately ischemic myocardium at the expense of metabolic recovery: a model of short-term hibernation. Circ Res. 1992;70:1282-1295.
Corr PB, Witkowski FX, Sobel BE. Mechanisms contributing to malignant dysrhythmias induced by ischemia in the cat. J Clin Invest. 1987;61:109-119.
Hjalmarson Å, Elmfeldt D, Herlitz J, Holmberg S, Málek I, Nyberg G, Rydén L, Swedberg K, Vedin A, Waagstein F, Waldenström A, Waldenström J, Wedel H, Wilhelmsen L, Wilhelmsson C. Effect on mortality of metoprolol in acute myocardial infarction: a double blind randomised trial. Lancet. 1981;2:823-827.
Codina J, Yatani A, Grenet D, Brown AM, Birnbaumer L. The alpha subunit of the GTP binding protein Gk opens atrial potassium channels. Science. 1987;236:442-445.
Kirsch GE, Codina J, Birnbaumer L, Brown AM. Coupling of ATP-sensitive K+ channels to A1 receptors by G proteins in rat ventricular myocytes. Am J Physiol. 1990;259:H820-H826.
Keung CE, Karliner JS. Complex regulation of calcium current in cardiac cells. J Clin Invest. 1990;85:950-954.
Matsuda JJ, Lee HC, Shibata EF. Acetylcholine reversal of isoproterenol-stimulated sodium currents in rabbit ventricular myocytes. Circ Res. 1993;72:517-525.
Gross GJ, Auchampach JA. Blockade of ATP-sensitive potassium channels prevents myocardial preconditioning in dogs. Circ Res. 1992;70:223-233.
Thornton JD, Thornton CS, Sterling DL, Downey JM. Blockade of ATP-sensitive potassium channels increases infarct size but does not prevent preconditioning in rabbit hearts. Circ Res. 1993;72:44-49.
Myers RW, Pearlman AS, Hyman RM, Goldstein RA, Kent KM, Goldstein RE, Epstein SE. Beneficial effects of vagal stimulation and bradycardia during experimental acute myocardial ischemia. Circulation. 1974;49:943-947.
Corr PB, Gillis RA. Role of the vagus nerve in cardiovascular changes induced by coronary occlusion. Circulation. 1974;49:86-97.
Zuanetti G, De Ferrari GM, Priori SG, Schwartz PJ. Protective effect of vagal stimulation on reperfusion arrhythmias in cats. Circ Res. 1987;61:429-435.
Collins MN, Billman GE. Autonomic response to coronary occlusion in animals susceptible to ventricular fibrillation. Am J Physiol. 1989;257:H1886-H1894.
Billman GE. Effect of carbachol and cyclic GMP on susceptibility to ventricular fibrillation. FASEB J. 1990;4:1668-1673.
Du XJ, Dart AM, Riemersma RA, Oliver MF. Failure of the cholinergic modulation of norepinephrine release during acute myocardial ischemia in the heart. Circ Res. 1990;66:950-956.
Bourne HR, Masters SB, Miller RT, Sullivan KA, Heideman W. Mutations probe structure and function of G-protein α chains. Cold Spring Harb Symp Quant Biol. 1988;53:221-228.
Li Y, Kloner RA. Ischemic preconditioning is not mediated by adenosine receptors in the rat heart. Circulation. 1992;86(suppl I):I-24. Abstract.