Adenosine 3’ : %Monophosphate-dependent Protein Kinase-catalyzed Phosphorylation Reaction and Its Relationship to Calcium Transport in Cardiac Sarcoplasmic Reticulum*

Abstract A rapid, manyfold increase in phosphorylation of cardiac microsomes consisting primarily of sarcoplasmic reticulum was seen when these membranes were incubated in the presence of a bovine cardiac adenosine 3':5'-monophosphate (cyclic AMP)-dependent protein kinase (protein kinase) and cyclic AMP. Over 85% of the 32P associated with membrane protein under similar conditions was identified as phosphoserine and phosphothreonine. A less marked increase in phosphoprotein formation was observed when cardiac microsomes were incubated in 1 µm cyclic AMP in the absence of added protein kinase. This could be attributed to the presence of an endogenous protein kinase. When cardiac microsomes were incubated with protein kinase alone, phosphorylation also was enhanced, finally reaching the level seen with cyclic AMP and added protein kinase. The increased phosphorylation induced by protein kinase alone was attributable to the presence of an adenylate cyclase in the microsomal preparation. Epinephrine could be shown to stimulate both adenylate cyclase and phosphorylation of cardiac microsomes. Protein kinases prepared from both bovine cardiac or rabbit fast skeletal muscle catalyzed the formation of microsomal phosphoprotein. The extent of phosphoprotein formation correlated closely with the increment in the stimulation of the rate of calcium uptake by cardiac microsomes when concentrations of either protein kinase were varied, and the relationship between phosphoprotein formation and stimulation of calcium transport was independent of the source of the protein kinase. These data suggest the existence of a functional relationship between cardiac microsomal phosphorylation and an increased rate of calcium transport.

A rapid, manyfold increase in phosphorylation of cardiac microsomes consisting primarily of sarcoplasmic reticulum was seen when these membranes were incubated in the presence of a bovine cardiac adenosine 3':5'-monophosphate (cyclic AMP)-dependent protein kinase (protein kinase) and cyclic AMP. Over 85 % of the 3*P associated with membrane protein under similar conditions was identified cs phosphoserine and phosphothreonine.
A less marked increase in phosphoprotein formation was observed when cardiac microsomes were incubated in 1 PM cyclic AMP in the absence of added protein kinase.
This could be attributed to the presence of an endogenous protein kinase. When cardiac microsomes were incubated with protein kinase alone, phosphorylation also was enhanced, finally reaching the level seen with cyclic AMP and added protein kinase.
The increased phosphorylation induced by protein kinase alone was attributable to the presence of an adenylate cyclase in the microsomal preparation.
Epinephrine could be shown to stimulate both adenylate cyclase and phosphorylation of cardiac microsomes. Protein kinases prepared from both bovine cardiac or rabbit fast skeletal muscle catalyzed the formation of microsomal phosphoprotein.
The extent of phosphoprotein formation correlated closely with the increment in the stimulation of the rate of calcium uptake by cardiac microsomes when concentrations of either protein kinase were varied, and the relationship between phosphoprotein formation and stimulation of calcium transport was independent of the source of the protein kinase.
These data suggest the existence of a functional Portions of this work 'were published previously (l-3). This is the first paper in the series "Effects of a Cardiac Cyclic AMPdependent Protein Kinase on the Cardiac Sarcoplasmic Reticulum." $ New York Heart Association Research Fellow. $ Philip J. and Harriet L. Goodhart Professor of Medicine (Cardiology). relationship between cardiac microsomal phosphorylation and an increased rate of calcium transport.
The early finding of Sutherland and Rall (4) that cellular effects of epinephrine are expressed through the action of cyclic AMP' has been followed by a number of studies aimed at deciphering the biochemical sequence of events triggered by catecholamines, polypeptide hormones, and other physiological "messengers." Subsequently, others have found that cyclic AMP activates the enzyme protein kinase by combining with the binding subunit of this enzyme, thereby causing the dissociation of the binding subunit-cyclic AMP complex from the catalytic subunit of the enzyme (5-g).
Protein kinases have been identified in a variety of mammalian tissues (9); several, including those from heart and skeletal muscle, have been purified and characterized (10, 11). The substrate specificity of the muscle protein kinases, like that of other protein kinases, appears to be broad as they can catalyze the phosphorylation of several protein substrates.
Information on the molecular nature of protein kinases is rapidly expanding (12) but relatively little is known about their natural protein substrates. Several protein components of smooth, skeletal, and cardiac muscle that were co-purified with soluble protein kinase from these sources were shown to be phosphorylated but were not further identified (13 were centrifuged at 3" for 10 min at 1500 X g. The supernatant was aspirated and the pellets were dissolved in 0.1 ml of 0.5 N NaOH at room temperat,ure. An additional 2 ml of the trirhloroacetic acid solution was added, the centrifugation was repeated, and the pellets were washed with trichloroacetic acid three more times.
The final pellets were dissolved in 0.

Phosphorylation of Cardiac Microsomes by Bovine Cardiac
Protein Kinase-When microsomes were incubated with MgCl, and ATP in the absence of added cyclic AMP and protein kinase, slight but significant phosphorylation was seen (Fig. 1). In the presence of both cyclic AMP and protein kinase, phosphorylation was markedly stimulated, the initial rate of phosphorylation being increased 3-to g-fold. In the presence of cyclic AMP alone, phosphorylation was increased slightly, less than a-fold. Incubation with protein kinase alone caused little stimulation during the early part of the incubation, but after 10 to 20 min phosphorylation greatly exceeded that seen in the control reaction, eventually reaching the level obtained when cyclic AMP was added along with the protein kinase at the start of the reaction.
The dependence of protein kinase-catalyzed microsomal phosphorylation upon protein kinase concentration was studied in experiments where the microsomal protein concentration was maintained at 0.5 mg per ml and the protein kinase concentration was varied from 0 to 0.3 mg per ml (Fig. 2). In the presence of 1 pM cyclic AMP, microsomal phosphorylation increased with increasing protein kinase concentration until the protein kinase The data shown in Fig. 2 have been corrected for phosphorylation attributable to the presence of added protein kinase. At the protein kinase concentration of 0.1 mg per ml, phosphorylation attributable to the protein kinase preparation was less than 8% of the total phosphorylation observed.
The phosphorylation of microsomes was slower than that of the protein kinase preparation, so that the proportional phosphorylation of microsomes increased percentagewise after prolonged incubation. Protein kinase-catalyzed microsomal phosphorylation was found to be largely independent of Ca2f concentration in the range of 0.05 to 100 pM (Fig. 3). In these studies, concentrated calcium-EGTA buffers including a high total CaClz concentration (1 mM) were used to prevent depletion of ionized calcium due to calcium uptake by the microsomes.
The cyclic AMP concentration dependence of microsomal phosphorylation is seen in Fig. 4. Maximum activation of phosphorylation by added protein kinase occurred at about 1 PM added cyclic AMP.
Half-maximal activation by cyclic AMP in the presence of protein kinase, based on four independent experiments, was seen at 0.11 f 0.02 (SE.) pM. Slight stimulation of phosphorylation by cyclic AMP seen in the absence of added protein kinase (Fig. 4) probably reflects the presence of a protein kinase in the microsomal preparation.
Its sensitivity to cyclic AMP was similar to that catalyzed by the added protein kinase.
The temperature dependence of microsomal phosphorylation was determined under conditions where phosphorylation was linear with respect to both protein kinase concentration (0.05 mg per ml) and incubation time (30 s). An Arrhenius plot (35) of the results obtained when the temperature was varied between 4 and 40" is shown in Fig. 5.
The preceding studies were performed under conditions similar to those used to demonstrate an effect of protein kinase on calcium uptake (22), in which oxalate as well as a high concentration of KC1 to decrease the solubility product of calcium oxalate (36) were included in the incubation medium.
The effects of 2.5 InM Tris-oxalate and 0.12 M KC1 on microsomal phosphorylation therefore were examined. Oxalate typically microsomes (0.5 mg per ml) were incubated for 5 min under standard conditions. Phosphorylation was measured in the presence of increasing concentrations of cyclic AMP in the absence (0) and presence (0) of 0.1 mg of bovine cardiac protein kinase per ml.
Cardiac microsomes (0.5 mg per ml) were incubated as described under "Experimental Procedure" at temperatures ranging from 4 to 40". produced less than 6% inhibition of phosphorylation. No consistent effect of high KC1 concentrations on microsomal phosphorylation could be demonstrated.
E$ect of Epinephrine on Phosphorylation and Adenylate Cyc-Zase-Because cardiac microsomes contain endogenous adenylate cyclase activity (37)(38)(39), agents that stimulate cyclic AMP production may be expected to enhance protein kinase-catalyzed microsomal phosphorylation in the absence of added cyclic AMP. Fig. 6 shows stimulation of microsomal phosphorylation after incubation in 10 pM Z-epinephrine in the presence and absence of protein kinase.
This stimulation was completely abolished by 20 pM dl-propranolol (Table I). Further evidence that microsomal phosphorylation may be stimulated by endogenously produced cyclic AMP was obtained from measurements of adenylate cyclase activity in cardiac sarcoplasmic reticulum under conditions similar to those described for the phosphorylation studies (see "Experimental Procedure"). Z-Epinephrine (10 PM) produced an approximately 2-fold increase in cyclic AMP when measured after 15 and 30 min of incubation in the presence and absence of protein kinase (Fig. 7). The stimulation of adenylate cyclase by epinephrine was completely abolished by 20 PM dl-propranolol.
Characterimtion of Acid Precipitable [32P]Phosphate-The acidprecipitable [32P]phosphate measured under standard assay conditions was characterized with respect to its chemical stability ( Table II). Following incubation in 0.5 N NaOH at 90" for 10 min and subsequent washing with trichloroacetic acid, all ["'PIphosphate was released from phosphorylated microsomes. When microsomes were incubated in the absence of cyclic AMP and protein kinase (control microsomes), 86% of the [32P]phosphate was recovered in the precipitate after treatment with 10% trichloroacetic acid for 10 min at 90". Of the total phosphorylation observed in the presence of cyclic AMP and protein kinase, which was approximately 4 times greater than for the control microsomes (Table II), 89% was recovered in the acid precipitate obtained as described above. Treatment of control microsomes with 0.8 M hydroxylamine or 0.8 M NaCl resulted in almost complete recovery of the acid-precipitable [**PIphosphate. Similarly, the recovery in acid precipitates of [azP]phosphate formed in the presence of cyclic AMP and protein kinase was only slightly reduced after treatment with 0.8 M hydroxylamine (Table II).
The distribution of [32P]phosphate in acid hydrolysates of microsomes incubated in the presence of cyclic AMP with or without added protein kinase is summarized in Table III. When the distribution of [32P]phosphate in acid precipitates of protein kinase incubated with cyclic AMP but in the absence of microsomes was examined, a significant number of counts was found to be associated with phosphoserine and a lesser number with phosphothreonine.
In the case of microsomes alone, slight but significant incorporation of [a2P]phosphate into phosphoserine and phosphothreonine also was seen. When microsomes were incubated with cyclic AMP and the protein kinase, an almost 40-fold increase in counts at the spot on the electrophoretogram corresponding to phosphoserine was observed. A lesser increase in counts corresponding to phosphoserine, approximately 4-fold, was found when microsomes were incubated with cyclic AMP alone. In both cases an increase in counts also was observed at the spot corresponding to phosphothreonine. Relationship of Protein Kinase-catalyzed Phosphorylation to Calcium Uptake-Protein kinase isolated from both bovine heart or rabbit fast skeletal muscle increased the rate of calcium uptake by dog heart sarcoplasmic reticulum (Fig. 8). To correlate protein kinase-catalyzed microsomal phosphorylation with protein kinase-stimulated calcium uptake by cardiac sarcoplasmic   (Fig. 9). The skeletal muscle protein kinase preparations were less active in causing stimulation of phosphorylation and calcium uptake than were the cardiac protein kinase preparations, but the increasing phosphorylation of cardiac microsomes that was induced by cardiac and skeletal muscle protein kinases paralleled the increase in calcium uptake rate over a wide range of protein kinase concentrations.
A positive correlation is found for both cardiac (r = 0.96, p < 0.001) and skeletal protein kinases (r = 0.94, p < 0.901) and the best fit lines, calculated by the method of least mean squares with the aid of a PDP-8e computer, were virtually superimposable.
The best fit line using combined data is shown in Fig. 10 These chemical characteristics eliminate the possibility that [a2P]phosphate is incorporated into nucleic acid, which might be present in the microsomal preparations as a contaminant, or into the acyl phosphate ATPase intermediate of the calcium transport system. These studies thus suggest that the [32P]phosphate is incorporated into serine or threonine, amino acids which are known to be phosphorylated by protein kinases (40). According to Table III because these two amino acids are difficult to separate in the system used in the present study (41). Fluoride, an inhibitor of phosphatases (42) which catalyze the dephosphorylation of phosphoproteins, was included in the incubation medium to maximize recovery because phosphatases have been found to be present in these microsomal preparations?
Half-maximal activation of protein kinase-catalyzed microsomal phosphorylation is seen when cyclic AMP is present at a concentration of 0.11 f 0.02 (SE.) PM. The same value for half-maximal activation by added cyclic AMP was obtained for protein kinase-stimulated calcium uptake (22) and Ca2+-activated ATPase (43). The similarity of these values is consistent with the view that the enhanced rate of calcium transport into the sarcoplasmic reticulum seen in the presence of protein kinase and cyclic AMP is, in fact, the result of the membrane phosphorylation that is documented in the present report.
Oscillations were regularly seen during the initial phase of microsomal phosphorylation.
For example, after about 10 min of incubation, the rate of microsomal phosphorylation in the presence of cyclic AMP and protein kinase levels off only to increase again and reach a plateau (Fig. 1). This pattern of phosphorylation may be evidence of the presence of two or more interacting enzymes (44)) one of which may be a phosphoprotein phosphatase.
The oscillations cannot be explained on the basis of a competition between the adenylate cyclase and protein kinase for the substrate ATP because ATP concentrations remain in the millimolar range, well above the apparent K, of approximately 0.1 mM ATP for microsomal phosphorylation.3 The increase in microsomal phosphorylation seen in the presence of cyclic AMP alone (Fig. l), which represents increased incorporation of [32P]phosphate into phosphoserine and phosphothreonine (Table III), appears to be catalyzed by an endogenous protein kinase. Similar findings have recently been published by Wray et al. (45) who measured protein kinase activit'y in cardiac microsomes.
The increase in phosphorylation of cardiac microsomes by protein kinase in the absence of added cyclic AMP (Fig. 1) may reflect the activity of the adenylate cyclase that has previously been found in cardiac microsomal preparations (37)(38)(39). Based on data shown in Fig. 7, the concentration of cyclic AMP produced by the microsomal adenylate cyclase is sufficiently high for stimulation of phosphorylation.
Alternatively, the observed stimulation of phosphorylation could be the result of activation of protein kinase by substrate protein (46).
Comparison of the extent of microsomal phosphorylation in the absence and presence of high concentrations of KCl, such as are found intracellularly, shows no significant differences. These results may represent additive effects of KC1 on several enzymes known to be present in the microsomal preparation. Pronounced effects of KC1 on protein kinase activity as well as on phosphatase activity of human lymphocytes have been reported (47).
The Arrhenius plot of data obtained in the study of the effect of temperature on microsomal phosphorylation (Fig. 5) shows a discontinuity at approximately 13" and is concave downwards. The energy of activation for the microsomal phosphorylation reaction is approximately 16.6 Cal per mole at temperatures ranging from 4 to 13" and 7.2 Cal per mole at temperatures ranging from 13 to 40" as calculated from the slopes of the lines drawn by the method of least mean squares. The discontinuity at approximately 13" may be due to a change in the thermodynamic properties of bovine cardiac protein kinase although definitive conclusions may not be drawn due to the heterogeneity of the membrane preparation.
The possibility that the system of enzymes described above has physiological significance would be supported if naturally occurring agents that increase cyclic AMP also cause stimulation of phosphorylation.
The present studies show that 10 PM epinephrine increases adenylate cyclase activity (Fig. 7) as well as microsomal phosphorylation (Fig. 6) and that this is a fladrenergic effect in that this stimulation is completely abolished by dl-propranolol ( Fig. 7 and Table I'). While the increase in cyclic AMP production in response to epinephrine was approximately 2-fold, the increase in the amount of phosphorylation in response to epinephrine was relatively less. Protein-protein 3 M. A. Kirchberger, unpublished observations. interactions leading to activation of protein kinase such as were mentioned above could be one factor in producing disproportionate increases in cyclic AMP production and microsomal phosphorylation.
The relatively small increase in phosphorylation in response to epinephrine in the early part of the time course shown in Fig. 6 could be attributed to a cyclic AMP concentration attained under these conditions (approximately 0.1 /.LM based on data in Fig. 7) which is not sufficient for full stimulation of protein kinase activity.
Evidence presented in this communication, as well as by other investigators, indicates that the sarcoplasmic reticulum contains both an intrinsic protein kinase (45, 48) as well as an adenylate cyclase (37)(38)(39).
However, protein kinase present in the myoplasm that bathes the sarcoplasmic reticulum and activated by cyclic AMP produced at the plasma membrane may in the intact tissue also play a role in the phosphorylation of the sarcoplasmic reticulum.
Protein kinases derived from different mammalian species, as well as from different tissues, are capable of stimulating the calcium transport system of sarcoplasmic reticulum isolated from dog heart.
Net stimulation of calcium uptake rate paralleled the net increase in protein kinase-catalyzed phosphorylation, and this relationship is independent of the source of protein kinase (Fig. 10).
The apparent stoichiometry between maximal protein kinaseinduced phosphate incorporation into cardiac microsomes, which is about 1.5 nmoles of Pi per mg of microsomal protein (e.g. Fig.  9), and that reported for the incorporation of phosphate into the acyl phosphate ATPase intermediate (49-51) is approximately 1. Assuming 1 mole of phosphate to be incorporated per mole of a regulatory phosphoprotein whose formation is catalyzed by protein kinase, each mole of this regulatory phosphoprotein would have the ability to influence a single mole of the calcium transport ATPase.
Further studies of the proposed regulatory phosphoprotein will be needed to clarify these relationships.