Department of Pharmacology and Cell Biophysics, University of Cincinnati College of Medicine, Cincinnati, Ohio 45267-0575 Received September 7, 1988; Revised Manuscript Received April 6, 1989 abstract: A Ca2+-calmodulin-dependent protein kinase was purified to apparent homogeneity from the cytosolic fraction of canine myocardium, with phospholamban as substrate. Purification involved sequential chromatography on DEAE-cellulose, calmodulin-agarose, DEAE-Bio-Gel A, and phosphocellulose. This procedure resulted in a 987-fold purification with a 5.4% yield. The purified enzyme migrated as a single band on native polyacrylamide gels, and it exhibited an apparent molecular weight of 550000 upongel filtration. Gel electrophoresis under denaturing conditions revealed a single protein band with Mt 55000. The purified kinase could be autophosphorylated in a Ca2+-calmodulin-dependent manner, and under optimal conditions, 6 mol of P; was incorporated per mole of 55 000-dalton subunit. The activityof the enzyme was dependent on Ca2+, calmodulin, and ATP-Mg2+. Other ions which could partially substitute for Ca2+ in the presence of Mg2+ and saturating calmodulin concentrations were Sr2+> Mn2+> Zn2+> Fe2+. The substrate specificity of the purified Ca2+-calmodulin-dependent protein kinase for cardiac proteins was determined by using phospholamban, troponin I, sarcoplasmic reticulum membranes, myofibrils, highly enriched sarcolemma, and mitochondria. The protein kinase could only phosphorylate phospholamban and troponin I either in their purified forms or in sarcoplasmic reticulum membranes and myofibrils, respectively. Exogenous proteins which could also be phosphorylated by the purified protein kinase were skeletal muscle glycogen synthase> gizzard myosin light chain> brain myelin basic protein> casein. However, phos-pholamban appeared to be phosphorylated with a higher rate as well as affinity than glycogen synthase. These findings suggest that a Ca2+-calmodulin-dependent protein kinase, composed of 10 identical subunits, may participate in regulation of sarcoplasmic reticulum function in cardiac muscle.

^ biochemical responses to regulatory stimuli are mediated by at least two intracellular signals, cAMP and calcium. Available evidence from in vitro and in vivo studies suggests that the responses to these second messengers may be interrelated. This interrelationship is particularly evident at the level of phospholamban, which is an integral proteolipid of cardiac sarcoplasmic reticulum and is the putative regulator for the Ca2+-ATPase. Phospholamban may be phosphorylated by cAMP-dependent (Kranias et al., 1980; Tada & Katz, 1982; Kranias, 1985), Ca2+-phospholipid-dependent (Movsesian et al., 1984), and Ca2+-calmodulin-dependent protein kinases (LePeuch et al., 1979; Bilezikjian et al., 1981; Kirchberger & Antonetz, 1982; Kranias, 1985; Gasser et al., 1986). These kinases phosphorylate phospholamban at distinct sites, and each phosphorylation is associated with increases in the initial rates of Ca2+ transport by cardiac sarcoplasmic reticulum. The sarcoplasmic reticulum associated Ca2+-calmodulin-dependent protein kinase has recently been par-tially purified, and it appears to consist of subunits with Mr 56000 (Molla & Demaille, 1986; Jett et al., 1987). In heart, there are at least three kinds of protein kinases, which are activated in thepresence of calmodulin. These include myosin light chain kinase (Walsh et al., 1979; Wolf & Hoffmann, 1980), phosphorylase kinase (Werth et al., 1982), and a multifunctional Ca2+-calmodulin-dependent kinase II (Kloepper & Landt, 1984; Iwasa et al., 1985 …