Creatine Kinase in Serum: 3. Further Study of Adenylate Kinase

Transcription

1 CLIN.CHEM.23/10, (1977) Creatine Kinase in Serum: 3. Further Study of Adenylate Kinase Inhibitors Gabor Szasz, Willie Gerhardt,2 and Wolfgang Gruber3 In search of an appropriate inhibitor to suppress the interference of adenylate kinase with the creatine kinase assay, we found that the combination diadenosine pentaphosphate (10 imol/iiter) and AMP (5 mmol/liter) Is a better inhibitor than is fluoride (25 mmol/iiter). The latter inhibits adenylate kinase uncompetitively and weakly (K1 = 2.5 mmol/liter), and must be incorporated in the starting reagent, and at 30 #{176}C it becomes fully effective only after a lag phase of 6 mm. In this concentration, fluoride inhibits adenylate kinase from erythrocytes, muscle, liver, or platelets by 94, 92, 88, and 87%, respectively, and creatine kinase by 8%. Bromide and chloride also inhibit creatine kinase. Attempts to replace AMP by a specific inhibitor of liver adenylate kinase failed. Homologs of diadenosine pentaphosphate with either fewer or more phosphoryl groups in the polyphosphate bridge inhibited even more weakly than did the pentaphosphate. Platelets can significantly contribute to adenylate kinase activity in serum. The inhibitor combination inhibited adenylate kinase from platelets by 90%. AdditIonal Keyphrases interference with enzyme activity platelets To minimize the interference of adenylate kinase (ATP:AMP phosphotransferase; EC ) with the assay of creatine kinase (ATP:creatine N-phosphotransferase; EC ) activity in serum, we recently recommended an inhibitor combination of diadenosine pentaphosphate and AMP (1). While our paper was in press, Rosano et al. (2) proposed the use of sodium fluoride to suppress the interference from adenylate kinase. Fluoride has been first applied for this purpose by Aitken et al. (3) and has been an ingredient of a commercially available kit for several years (4). We have extended the experiments with fluoride to the individual adenylate kinase isoenzymes in serum and assessed the type of inhibition. We also compared the effect of fluoride on creatine kinase with that of other halides and of AMP. We further include here our studies with the homologs of diadenosine pentaphosphate (5). The extremely potent inhibition of adenylate kinase by diadenosine pentaphosphate indicates that the stereochemistry of this compound is such that one molecule can block the binding sites of the enzyme for both substrates, whereas two molecules of AMP are necessary. This is true, however, only for the enzyme from erythrocytes and muscle; the association constant for the binding of the inhibitor to liver adenylate kinase is about 100 times higher (1). We therefore thought it logical to try whether homologs of this compound with fewer or more phosphoryl groups in the polyphosphate bridge would better fit into the space between the two binding sites and thus more potently inhibit the liver enzyme. In the course of studying the interference of endogenous adenylate kinase with measurement of creatine kinase activity in serum (1), there arose the question as to the source of this enzyme in the serum. Adenylate kinase is almost ubiquitous in the tissues (6) and so one can assume that any in the serum originates from the tissue actually damaged, e.g., muscle or liver. In nonhemolyzed sera of healthy individuals, a mixture of adenylate kinase isoenzymes is probably present, especially from the skeletal muscle, liver, and erythrocytes, but perhaps also from the platelets during clotting (7). We previously investigated the inhibition of the adenylate kinase from skeletal muscle, liver, and erythrocytes (1), but neglected that froni the platelets. In the present paper we describe the estimation of the portion of adenylate kinase in serum of healthy persons that originates from the platelets, and its inhibition. Presented in part at the Joint Annual Meeting of the German and Dutch Societies for Clinical Chemistry, Aachen, Germany, March 9-11, J. Clin. Chem. Clin. Biochem. 15, 190 (1977). Abstract. The first two papers in this series are refs. 8 and 1 here. Institute of Clinical Chemistry, University Medical School, Klinikstrasse 32b, 6300 Giessen, Germany. 2Department of Clinical Chemistry, Lasarettet, Helsingborg, Sweden. 3Research Center, Boehringer Mannheim GmbH, 8132 Tutzing, Germany. Received May 2, 1977; accepted July 14, Materials and Methods Adenylate kinase isolation. Adenylate kinase from human erythrocytes and liver was obtained as described in detail in the previous part of this study (1). We isolated adenylate kinase of platelets from human blood plasma. The plasma was received from blood donors and contained the acid-citrate-dextrose (formula A) solution used in blood banks to conserve erythrocytes. To 1888 CLINICAL CHEMISTRY, Vol. 23, No. 10, 1977

2 remove even traces of erythrocytes, we first centrifuged the plasma at 800 X g for 3 mm. Then the ph of each 10 ml of supernatant plasma ( platelets per microliter) was adjusted to ph with 1 mol/liter hydrochloric acid (about 140 il) to prevent the clotting of the platelets. A second centrifugation at 150 X g for 30 mm resulted in the sedimentation of about a third of the platelets (we counted platelets per microliter of supernatant fluid). The sediment was twice frozen at -60 #{176}C and thawed at +25 #{176}C. The platelets were dissolved by shaking in 0.5 ml of human albumin (40 g/liter), and the platelet membranes were removed by centrifugation at X g for 2 mm. The adenylate kinase activity was about 500 U/liter (at 30 #{176}C) in the clear supernate, which contained 2 million platelets per microliter based on the 20-fold concentration (10 ml of plasma reduced to 0.5 ml of albumin solution). Reagents. Diadenosine tetra-, penta-, and hexaphosphate [P,P5-di(adenosine-5 )tetra-, penta-, and hexaphosphate] were from P-L Biochemicals, Inc., Milwaukee, Wis Adenylate kinase from rabbit muscle was from Boehringer Mannheim (Bio-Dynamics/bmc, Indianapolis, Ind ). Sodium fluoride, bromide, and chloride, in the purest quality available, were from E. Merck, Darmstadt, Germany. Enzyme assay methods. Creatine kinase activity was measured with an optimized modification (8) of Oliver s method (9) incorporating, however, also 10 imol of diadenosine pentaphosphate per liter of reaction mixture (1). Adenylate kinase activity was assayed with use of the same reagent composition, except that we omitted.creatine phosphate and inhibitors of adenylate kinase (AMP, diadenosine pentaphosphate, and sodium fluoride) and started the reaction either with serum or with ADP. In studying the efficacy of the different adenylate kinase inhibitors, we included AMP and diadenosine pentaphosphate in the reagent. A further (starting) reagent contained sodium fluoride. To obtain results that could be compared with those of Rosano et al. (2) we measured the activity of both enzymes at 30 #{176}C. Results Adenylate Kinase Inhibition of adenylate kinase by fluoride. The time course of the reduction of the sample blank of the creatine kinase, omitting creatine phosphate from the creatine kinase reagent, is illustrated by original recordings (Figure 1). To human serum with very low adenylate kinase activity, hemolysate was admixed to give an adenylate kinase activity of 260 U/liter. After a 5-mm pre-incubation the recording was started after adding either buffer or sodium fluoride. When fluoride was used in the starting reagent, both AMP and diadenosine pentaphosphate were omitted. Figure 1 shows that fluoride becomes fully effective after a lag phase of 6 mm at 30#{176}C (or about 8 mm at 25#{176}C or 4 mm at 37 #{176}C). After the steady state was attained, 25 mmol of fluoride per liter of assay mixture inhibited adenylate kinase from human erythrocytes by 94%, whereas our inhibitor combination (10 imol diadenosine penta- i A Star 2 L W Fig. 1. Time course of the sample blank of the creatine kinase assay Human serum with admixed hemolysate. After a 5-mm pre-incubation the recording was started after adding buffer (A) or sodlirn fluoride (B). In case B, A1 and diadenoslne pentaphosphate were omitted from the reagent S 1 + mk Sodium ttuoride [mmol / liter] Fig. 2. Inhibition of adenylate kinase from human erythrocytes by fluoride-dixon plot C E K Sodium fluoride [mmol I liter] Fig. 3. Inhibition of adenylate kinase from human erythrocytes by fluoride, plotted according to Cornlsh-Bowden (10) phosphate and 5 mmol AMP per liter) resulted in an immediate inhibition of 98%. The sample blank was 2.5-fold greater when fluoride was used, even after an adequate lag phase. The Dixon plot (Figure 2) indicated a noncompetitive inhibition of adenylate kinase from human erythrocytes by fluoride but did not provide the value of the inhibi- 15 CLINICALCHEMISTRY,Vol. 23, No. 10,

3 Table 1. Inhibition of Adenylate Klnase from Human Erythrocytes and Liver by Diadenosine Polyphosphates a DIad.nosln.- Dladenosln.- Diadenosinet.traphosphat. p.ntaphoaphat. h.xaphosphat Source of moi/iit.r adenylate klnase Erythrocytes Liver a Extent of inhibition in percent. 4, > Liver adenylate knase tion constant. The apparent K1 value was calculated from the Cornish-Bowden plot (Figure 3) (10) to be as high as 2.5 mmol/liter, which explained the weak inhibitory efficacy of fluoride. Figure 4 compares the inhibition of adenylate kinase from human liver and erythrocytes. Fluoride also inhibited the enzyme from the erythrocytes by 94% in this experiment, but that from the liver by only 88%. The corresponding inhibition of adenylate kinase from rabbit muscle amounted to about 92% (not illustrated). Inhibition of adenylate hinase by diadenosine polyphosphates. The three diadenosine polyphosphates we compared as inhibitors of adenylate kinase are shown in Table 1. The pentaphosphate most potently inhibited adenylate kinase from both human erythrocytes and liver, followed by the hexaphosphate, and (much more poorly) by the tetraphosphate. Adenylate kinase in serum derived from platelets. We tried to distinguish adenylate kinase activity in serum and plasma of healthy blood donors. With either heparin (80 mt. units/mi) or ethylenediaminetetraacetate (1 g of the disodium salt per liter) as anticoagulants, turbidity resulted when the plasma was added to the adenylate kinase reagent, making the measurement of a constant reaction rate impossible. On adding AMP and creatine phosphate-i.e., using the creatine kinase reagent (11, 12)-the turbidity disappeared. We were not able to prevent the appearance of turbidity: neither gradually decreasing magnesium acetate from 10 to 1 mmol nor increasing ADP, as magnesium captor, from 2 to 4 mmol per liter of reaction mixture, solved the problem. Because of this, we turned to platelet suspensions obtained from human blood plasma. The inhibition of 4, > U Erythrocyte adenytate krnase Sod,um Iluonde [rnmoi ter] FIg. 4. Comparison of the inhibition of adenylate kinase from human liver and erythrocytes by fluoride 300 ib mmol or jmol / liter FIg. 5. ComparIson of the Inhibition of adenylate kinase from humanplatelets by diadenosine pentaphosphate(0) (itmol/iiter), AMP (#{149}), and fluoride (X) (mmol/iiter) adenylate kinase from platelets dissolved in human albumin is illustrated in Figure 5. At the recommended inhibitor concentrations, denoted by the arrows in the figure, sodium fluoride was the most effective inhibitor, followed by diadenosine pentaphosphate and AMP, the extent of inhibition being 87, 83, and 78%, respectively. In an additional experiment we measured the sample blank of the creatine kinase assay due to adenylate kinase from platelets (Table 2). Here also sodium fluoride showed the greatest inhibitory effect when only a single inhibitor was used, but the effect was significantly greater with our inhibitor combination of diadenosine pentaphosphate and AMP. Table 2. Effect of Inhibitors on the Sample Blank of the Creatlne Kinase Assay AMP (5 mmol/ilter) Dladenosln. and dladsnoslne Sodium without AMP p.ntapho.phat. pentaphosphate fluoride inhibitor (5 mmol/ift#{149}r) (10 umoi/iiter) (10 mol/iiter) (25 mmol/ilter) U/liter Rel. value, % Human albumin (40 9/liter) with adenylate kinase from human platelets. Means of triplicate assays at 30 #{176}C. Standard reagent (11. 12)without creatine phosphate and varying the inhibitors CLINICAL CHEMISTRY. Vol. 23, No. 10, 1977

4 Table 3. InhibItion of Creatine Kinase Activity by Different Compounds Compound No addititive AMP Sodium fluoride Sodium bromide Sodium chloride Concn Activity InhibItIon at 30 #{176}C mmoi/iiter U/liter Per cent Mean creatine kinase activity of 18 sera ( U/liter) calculated by subtracting an individualsample blank. Duplicate assays. Creatine Kinase Inhibition of crea tine kinase activity by different compounds. We compared the inhibition of creatine kinase activity by AMP (at an AMP/ADP ratio of 2.5) and by sodium halides (Table 3). The data were statistically evaluated by analysis of variance (14) (two-way classification, mixed model, random fixed). The mean values with additives we compared with the mean value of the sera with no additive by using linear contrast. Use of 5 mmol of AMP and 25 mmol of sodium fluoride per liter inhibited significantly and almost equally, by about 8% in these series. The inhibition by sodium bromide and chloride amounted to 3-4% and was at the border of significance. Effect of inhibitors on the sample blank of the creatine kinase assay. The sample blanks were measured without and with different inhibitors at 30 #{176}C (Table 4), omitting creatine phosphate from the standard creatine kinase reagent (11, 12). The nonhemolyzed sera were obtained from patients with a normal enzyme profile (controls) and with muscle damage (creatine kinase: U/liter at 25 #{176}C) or liver disease (pathological liver-enzyme profile), respectively. Without any inhibitor, even the mean sample blanks were considerably high; some of the blanks reached the upper range of the reference values for creatine kinase activity (13). The height and the marked scattering of the sample blanks stressed again the necessity to use adenylate kinase inhibitors in the creatine kinase assay. Even 5 mmol of AMP per liter (AMP/ADP ratio of 2.5) decreased sample blank rates significantly, but the highest values amounted to 20% of the upper limit of the reference range for creatine kinase activity. If 10 imol of diadenosine pentaphosphate per liter was used alone instead of AMP, there was a significant improvement, especially in sera of patients with muscle damage and liver disease. The combination of both inhibitors yielded a further decrease in the sample blanks, but 25 mmol of fluoride per liter of assay mixture was almost equally effective. The sample blanks averaged about 10% of the 50th percentiles, with the highest ones at 10% of the 97.5th percentiles of the reference range for creatine kinase activity (13). Discussion In our experiments sodium fluoride inhibited adenylate kinase noncompetitively and relative weakly. The association constant for the binding of fluoride to adenylate kinase from erythrocytes amounted to 2.5 mmol per liter, whereas the comparable apparent K1 values were about 0.3 mmol AMP and only about 30 nmol of diadenosine pentaphosphate per liter of reaction mixture (1). Fluoride, 25 mmol/liter, inhibited adenylate kinase from erythrocytes, skeletal muscle, liver, or platelets by 94, 92, 88, and 87%, respectively. Our inhibitor combination (10 mol of diadenosine pentaphosphate and 5 mmol of AMP per liter of assay mixture) inhibited the enzyme from erythrocytes and muscle by 97%, from liver by 95% (1), and from platelets by 90%. Fluoride must be incorporated in a second (starting) reagent because magnesium and fluoride ions together in one solution at the recommended concentrations yield a precipitate (2). Thus a two-component reagent system is obligatory. This excludes all procedures in which a combined reagent is used and the reaction is started with serum. An additional disadvantage of fluoride is the prolonged lag phase (6 mm at 30 #{176}C) after adding the starting reagent containing fluoride. This makes the procedure awkward and in addition the linear range of activity will be limited due to the product inhibition by NAD(P)H (8, 13). Rosano et al. (2) especially preferred fluoride as in- Patients Controls n = 56 Muscle damage n = 26 Liver disease n = 27 Table 4. Effect of Inhibitors on the Serum Blank of the Creatine Kinase Assay AMP (5 mmol/ilter) Dladenoslne and diadenosine without AMP pentaphospliat. pentaphosphate InhibItor (5 mmol/iiter) (10 moi/iiter) (10 amoiiiiler) 18.6 ± 7.6 (8-48) 29.9 ± 16.3 (15-71) 24.5 ± 14.2 (8-63) 6.7 ± 1.8 (4-11) 9.4 ± 4.3 (2-19) 8.5 ± 3.1 (4-17) Means, standarddeviations, and ranges (in parentheses) of the sample blank. Standard reagent (11. 12)without creatlne phosphate. U/liter, 5.9 :1: ± 2.9 (2-10) 6.5 ± 2.4 at 30 #{176}C 4.1 ± ± ± 1.9 SodIum fluoride (25 mmol/lfter) 4.6 ± ± ± 1.9 CLINICALCHEMISTRY,Vol. 23, No. 10,

5 hibitor of adenylate kinase because they did not observe any inhibition of creatine kinase itself, in contrast to AMP. We found that both 5 mmol of AMP (AMP/ADP ratio, 2.5) and 25 mmol of fluoride per liter of assay mixture inhibit creatine kinase by about 8%, and that 25 mmol of bromide or chloride per liter inhibit by 3 to 4%. Our attempts thus far to replace AMP by a specific inhibitor for the liver adenylate kinase have failed. The homologs of diadenosine pentaphosphate-the tetraand hexaphosphate-proved to be even weaker inhibitors for adenylate kinase from both erythrocytes and liver than is the pentaphosphate. We could not explain why fluoride decreased the serum blank of the creatine kinase assay more significantly than did AMP or diadenosine pentaphosphate singly, whether we used sera of healthy individuals or of patients with muscle damage or liver disease. This observation made us curious about the origin of the adenylate kinase in serum, especially of healthy individuals, and soon focused our interest on the platelets. The portion of adenylate kinase in serum released from the platelets during clotting can be assumed to be relatively constant, depending of course on the amount of the platelets in blood and on the clotting procedure. The logical experiment-to compare the adenylate kinase activity in the serum and plasma of the same individuals, in order to determine the activity due to the enzyme from the platelets-failed because of the turbidity resulting when plasma was added to the adenylate kinase reagent, with either heparin or ethylenediaminetetraacetate as anticoagulant. Therefore we were forced to work with isolated platelets. We estimated that the complete spillage of the enzyme from the platelets into the plasma, at platelets per microliter of blood, would yield an adenylate kinase activity of 80 U/liter at 30 #{176}C. The activity of healthy persons, however, averaged about 25 U/liter and the reference limit for adenylate kinase was U/liter at 30 #{176}C (13). Thus we can assume that only 10-20% of the adenylate kinase in the platelets is released into the serum. Platelet adenylate kinase was relatively poorly inhibited (78-83%) by a single inhibitor at the recommended concentrations, but the combination of AMP and diadenosine pentaphosphate resulted in an inhibition of 90%. Thus the apparent creatine kinase activity attributable to adenylate kinase from platelets amounted to at most 2-4 U/liter at 30 #{176}C when our standard reagent (11, 12) was used. Another unexpected result was that diadenosine pentaphosphate decreased the sample blank of the creatine kinase assay more effectively than did AMP, even in sera of patients with liver disease, although adenylate kinase isolated from the liver was less inhibited by diadenosine pentaphosphate (1). We therefore guess that only a small portion of the adenylate kinase in serum originates from the liver, except in the relatively rare cases with serious acute liver cell necrosis accompanied by marked increase in activity of the mitochondrial enzyme glutamate dehydrogenase (1). We are indebted to Miss Elfriede Kinne for her excellent technical assistance and Mr. Harald Glaucke, Miss Eva-Beate Nittel, and Mr. Lucky Hettler for the help in the preparation of the manuscript. For providing us with human blood and plasma from blood donors and for helpful discussions we thank Prof. Dr. Ch. Muller-Eckhardt. This research was supported by grants Sz 25/1 of the Deutsche Forschungsgemeinschaft, and of the Segerfalska Stiftelsen. References 1. Szasz, G., Gerhardt, W., Gruber, W., and Bernt, E., Creatine kinase in serum: 2. Interference of adenylate kinase with the assay. Clin. Chem. 22, 1806 (1976). 2. Rosano, T. G., Clayson, K. J., and Strandjord, P. E., Evaluation of adenosine 5 -monophosphate and fluoride as adenylate kinase inhibitors in the creatine kinase assay. Clin. Chem. 22, 1078 (1976). 3. Aitken, W. B., Molz, R. J., and Wermus, G. R., Automated kinetic assay of creatine phosphokinase. Clin. Chern. 18, 699 (1972) (Abstract). 4. Test methodology of creatine phosphokinase. Du Pont automatic clinical analyzer. Du Pont Co., Instrument Products Division, Wilmington, Del Lienhard, G. E., and Secemski, I. I., P,P5-Di(adenosine-5 )pentaphosphate, a potent multisubstrate inhibitor of adenylate kinase. J. Biol. Chem. 248, 1121 (1973). 6. Lehmann, F. G., Schneider, K. W., and Menge, H., Die enzymatische Diagnostik des Herzinfarktes. II Mitteilung: Die Bestimmung von organspezifischen Enzymen: Kreatinphosphokinase und Myokinase. Enzymol. Biol. Clin. 6, 36 (1966). 7. Todd, J. K., and Baron, D. N., Adenylate kinase in human tissues. In Research in Muscular Dystrophy, Proc. 3rd Symp. Pitman, London, 1965, p Szasz, G., Gruber, W., and Bernt, E., Creatine kinase in serum: 1. Determination of optimum reaction conditions. Clin. Chem. 22,650 (1976). 9. Oliver, I. T., A spectrophotometric method for the determination of creatine phosphokinase arid myokinase. Biochem. J. 61, 116 (1955). 10. Cornish-Bowden, A., A simple graphical method for determining the inhibition constants of mixed, uncompetitive and non-competitive inhibitors. Biochem. J. 137, 143 (1974). 11. Recommendations of the German Society for Clinical Chemistry. Standardization of methods for the estimation of enzyme activities in biological fluids. Standard method for the determination of creatine kinase activity. Revised draft of J. Clin. Chem. Clin. Biochem. 15, 255 (1977). 12. Recommended method for the determination of creatine kinase in blood. The Committee on Enzymes of the Scandinavian Society for Clinical Chemistry and Clinical Physiology. Scand. J. Clin. Lab. Invest. 36,711 (1976). 13. Szasz, G., Laboratory measurement of creatine kinase activity. In Proc. 2nd. mt. Symp. Clin. Enzymology, N. W. Tietz, A. Weinstock, and D. 0. Rodgerson, Eds., Amer. Assoc. Clin. Chem., Washington, D. C., 1976, p Ostle, B., Statistics in Research, 2nd ed., The Iowa State University Press, Ames, Iowa, CLINICALCHEMISTRY,Vol. 23, No. 10, 1977