The sites of phosphorylation of rabbit cardiac troponin I by adenosine 3'':5''-cyclic monophosphate-dependent protein kinase. Effect of interaction with troponin C.

1. Troponin I prepared from rabbit hearts contains 1.0-1.5 mol of P/mol when isolated by affinity chromatography. Most of the covalently bound phosphate is located in residues 1-48 of the molecule. 2. 3':5'-Cyclic AMP-dependent protein kinase catalyses phosphorylation at serine-20 and serine-146. Serine-20 is more rapidly phosphorylated than serine-146. 3. In troponin I prepared from frozen hearts by affinity chromatography about 0.3-0.5 mol of P/mol is associated with serine-20 and 0.8-1.0 mol of P/mol with other site(s) in residues 1-48 of the molecule. 4. Phosphorylation at serine-20 and servine-146 is not significantly inhibited by troponin C. 5. The mechansim of the interaction of troponin C with cardiac troponin I is discussed in the light of these results.     

Phosphorylation of rabbit cardiac-muscle troponin I by phosphorylase kinase. The effect of adrenaline.

1. Incubation of rabbit cardiac-muscle troponin I with phosphorylase b kinase leads to the incorporation of .07-1.2 mol of Pi/mol. 2. The major site of phosphorylation is a serine residue at position 72. 3. Lesser amounts of phosphate are incorporated into threonine-138, threonine-162 and serine 20. 4. Serine-20 is the only site that contains a significant amount of phosphate before incubation with phosphorylase b kinase. 5. Unlike the situation with serine-20, the extent of phosphorylation of serine-72 and threonine-138 in the perfused rabbit heart does not change when the heart is exposed to adrenaline (4 microM).     

Structure-function relationships in cardiac troponin T

Regions of rabbit and bovine cardiac troponin T that are involved in binding tropomyosin, troponin C and troponin I have been identified. Two sites of contact for tropomyosin have been located, situated between residues 92-178 and 180-284 of troponin T. A cardiac-specific binding site for troponin I has been identified between residues 1-68 of cardiac troponin T, within a region of the protein that has previously been shown to be encoded by a series of exons that are expressed in a tissue-specific and developmentally regulated manner. The binding site for troponin C is located between residues 180-284 of cardiac troponin T. When isolated from fresh bovine hearts, cardiac troponin T contained 0.21 +/- 0.11 mol phosphate per mol; incubation with phosphorylase kinase increased the phosphate content to approx. 1 mol phosphate per mol. One site of phosphorylation was identified as serine-1; a second site of phosphorylation was located within peptide CB3 (residues 93-178) and has been tentatively identified as serine-176. Addition of troponin C to cardiac troponin T does not inhibit the phosphorylation of this latter protein that is catalysed by phosphorylase b kinase.     

The phosphorylation of troponin I from cardiac muscle.

1. Troponin I isolated from fresh cardiac muscle by affinity chromatography contains about 1.9 mol of covalently bound phosphate/mol. Similar preparations of white-skeletal-muscle troponin I contain about 0.5 mol of phosphate/mol. 2. A 3':5'-cyclic AMP-dependent protein kinase and a protein phosphatase are associated with troponin isolated from cardiac muscle. 3. Bovine cardiac 3':5'-cyclic AMP-dependent protein kinase catalyses the phosphorylation of cardiac troponin I 30 times faster than white-skeletal-muscle troponin I. 4. Troponin I is the only component of cardiac troponin phosphorylated at a significant rate by the endogenous or a bovine cardiac 3':5'-cyclic AMP-dependent protein kinase. 5. Phosphorylase kinase catalyses the phosphorylation of cardiac troponin I at similar or slightly faster rates than white-skeletal-muscle troponin I. 6. Troponin C inhibits the phosphorylation of cardiac and skeletal troponin I catalysed by phosphorylase kinase and the phosphorylation of white skeletal troponin I catalysed by 3':5'-cyclic AMP-dependent protein kinase; the phosphorylation of cardiac troponin I catalysed by the latter enzyme is not inhibited.     

Studies on the phosphorylation of the inhibitory subunit of troponin during modification of contraction in perfused rat heart.

1. Rat hearts were perfused with 32Pi, and contractile force was increased by positive inotropic agents (agents that increase contractility). The inhibitory subunit of troponin (troponin I) was then isolated by affinity chromatography in 8M-urea, and its 32P content measured. Incorporation of phosphate into the subunit was calculated on the basis of the [gamma-32P]ATP specific radioactivity in the hearts. 2. When hearts were perfused with 30 nM-DL-isoprenaline (N-isopropylnoradrenaline), there was an increase in contractile force over 30s which was paralleled by an increase in troponin I phosphorylation. When hearts were perfused for 25s with increasing concentrations of isoprenaline from 1 NM to 0.6 muM, there was again a parallel increase in contractile force and troponin I phosphorylation. The maximum phosphorylation observed was 1.5 mol of phosphate/mol of troponin I, which was reached after 25s with 0.1 muM-isoprenaline. 3. Hearts were stimulated with a 15s pulse perfusion of 30nM-DL-isoprenaline. There was an increase in contractile force which was followed by a return to the control value within 50s. Troponin I phosphorylation increased to a plateau value which was reached within 30s, and remained constant for 60s after the isoprenaline pulse. Phosphorylase a and 3':5'-cyclic AMP concentration showed changes similar to that of the contractile force. There was no change in 3':5'-cyclic GMP concentration. 4. When hearts stimulated with a 15S pulse of isoprenaline were subsequently perfused with 0.6 muM-acetylcholine, the changes in contractile force, phosphorylase a and 3':5'-cyclic AMP were very similar to those seen with the 15s pulse of isoprenaline alone. Troponin I phosphorylation increased to a maximum 30s after the end of the isoprenaline pulse, but then rapidly decreased during the subsequent 30s. This decrease was preceded by a 60% increase in the concentration of 3':5'-cyclic GMP. 5. Hearts were perfused with 0.2 muM-glucagon for periods up to 60s. Contractile force showed little change for the first 30s, but then increased rapidly. This was paralleled by changes in 3':5'-cyclic AMP concentration. Troponin I phosphorylation increased slowly, but the increase in contractile force had reached a maximum before significant phosphorylation had occurred. 6. It is concluded that under certain conditions, e.g. immediately after beta-adrenergic stimulation, there is a good correlation between contractile force and troponin I phosphorylation. However, under other conditions, e.g. when contractile force is decreasing after removal of beta-adrenergic stimulation or in the presence of glucagon, contractile force and troponin I phosphorylation are not well correlated. These results suggest that mechanisms for modifying cardiac contractility, other than troponin I phosphorylation, must be present in rat heart.     

Sites phosphorylated in bovine cardiac troponin T and I. Characterization by 31P-NMR spectroscopy and phosphorylation by protein kinases

Bovine cardiac troponin isolated in a highly phosphorylated form shows four 31P-NMR signals [Beier, N., Jaquet, K., Schnackerz, K. & Heilmeyer, L.M.G. Jr (1988) Eur. J. Biochem. 176, 327-334]. Troponin I, which contains phosphate covalently linked to serine-23 and/or -24 [Swiderek, K., Jaquet, K., Meyer, H. E. & Heilmeyer, L. M. G. Jr (1988) Eur. J. Biochem. 176, 335-342], shows three resonances. Mg2(+)-saturation of holotroponin shifts these troponin I resonances to higher fields. Direct binding of Mg2+ to the phosphate groups can be excluded. Both these serine residues of troponin I, 23 and 24, are substrates for cAMP- and cGMP-dependent protein kinases as well as for protein kinase C. Isolated bovine cardiac troponin T contains 1.5 mol phosphoserine/mol protein, indicating that minimally two serine residues are phosphorylated. One phosphoserine residue is located at the N-terminus. An additional phosphoserine is located in the C-terminal cyanogen bromide fragment, CN4, which contains covalently bound phosphate. Protein kinase C phosphorylates serine-194, thus demonstrating exposure of this residue on the surface of holotoponin.     

Phosphorylation of troponin and the effects of interactions between the components of the complex

1. The troponin complex from skeletal muscle contains approximately 1 mol of phosphate/80000g of complex, covalently bound to the troponin T component. 2. On prolonged incubation of the troponin complex or troponin T with phosphorylase kinase the phosphate content of troponin T was increased to approx. 3mol/mol. 3. On prolonged incubation of troponin I with phosphorylase kinase up to 1.6mol of phosphate/mol were incorporated. 4. Phosphorylation of troponin I was greatly inhibited by troponin C owing to the strong interaction between these proteins. Thus in the troponin complex troponin T was the main substrate for phosphorylase kinase. The phosphorylation of isolated troponin T was also inhibited by troponin C. 5. Troponin I was phosphorylated when the troponin complex was incubated with a bovine cardiac 3′:5′-cyclic AMP-dependent protein kinase. Troponin T either in its isolated form or in the troponin complex was not phosphorylated by bovine protein kinase to any significant extent under the conditions used. 6. If the troponin complex was dephosphorylated to 0.2mol/mol, or phosphorylated up to 2.5mol/mol there was no significant effect on the ability of normal concentrations to confer Ca²⁺ sensitivity on the adenosine triphosphatase of densensitized actomyosin.     

Phosphorylation of cardiac troponin by cyclic adenosine 3':5'-monophosphate-dependent protein kinase.

The purpose of this investigation was to characterize the phosphorylation of bovine cardiac troponin by cyclic AMP-dependent protein kinase. The purified troponin-tropomyosin complex from beef heart contained 0.78 +/- 0.15 mol of phosphate per mol of protein. Analysis of the isolated protein components indicated that the endogenous phosphate was predominately in the inhibitory subunit (TN-I) and the tropomyosin-binding subunit (TN-T) of troponin. When cardiac troponin or the troponin-tropomyosin complex was incubated with cyclic AMP-dependent protein kinase and [gamma-32P]ATP, the rate of phosphorylation was stimulated by cyclic AMP and inhibited by the heat-stable protein inhibitor of cyclic AMP-dependent protein kinase. The 32P was incorporated specifically into the TN-I subunit with a maximal incorporation of 1 mol of phosphate per mol of protein. The maximal amount of phosphate incorporated did not vary significantly between troponin preparations that contained low or high amounts of endogenous phosphate. The Vmax of the initial rates of phosphorylation with troponin or troponin-tropomyosin as substrates was 3.5-fold greater than the value obtained with unfractionated histones. The rate or extent of phosphorylation was not altered by actin in the presence or absence of Ca2+. The maximal rate of phosphorylation occurred between pH 8.5 and 9.0. At pH 6.0 and 7.0 the maximal rates of phosphorylation were 13 and 45% of that observed at pH 8.5, respectively. These results indicate that cyclic AMP formation in cardiac muscle may be associated with the rapid and specific phosphorylation of the TN-I subunit of troponin. The presence of endogenous phosphate in TN-T and TN-I suggests that kinases other than cyclic AMP-dependent protein kinase may also phosphorylate troponin in vivo.     

The effect of adrenaline on the phosphorylation of the P light chain of myosin and troponin I in the perfused rabbit heart.

1. Two-dimensional electrophoresis has been used to study the extent of phosphorylation of the P light chain of myosin and troponin I in the rabbit beating heart. 2. A procedure has been developed that eliminates endogenous protein phosphatase activity during homogenization and sample preparation for electrophoresis. 3. Evidence has been obtained for two unphosphorylated forms of the P light chain in myosin from the ventricle of the rabbit, guinea pig and cow. 4. In vivo and in the rabbit perfused beating heart about 25% of the P light-chain fraction is in the phosphorylated form. 5. Intervention with adrenaline produced a slight increase in the extent of phosphorylation that reached a maximum after the peak in inotropic response. A similar increase was obtained with ischaemia in the absence of adrenaline. 6. The changes in phosphorylation of the major forms of troponin I identified by electrophoresis occurred after the peak of response to adrenaline and were compatible with previous results.     

Cardiac troponin I, isolated from bovine heart, contains two adjacent phosphoserines. A first example of phosphoserine determination by derivatization to S-ethylcysteine

Bovine cardiac troponin containing approximately 3 mol P/mol protein could be separated into its subunits without loss of phosphate. Troponin I and troponin T each contain about 1.5 mol P/mol protein. In troponin I two phosphorylated serine residues could be localized in the N-terminal region by conversion of phosphoserine to S-ethylcysteine. They are located in adjacent positions in the following sequence: -Arg-Arg-Ser(P)-Ser(P)-Ala-Asn-Tyr-Tyr-Arg-Ala-Tyr-Ala-Thr-Glu-Pro- His-Ala-Lys. This sequence shows that the first phosphoserine residue in bovine cardiac troponin I occupies a homologous position to phosphoserine-20 of rabbit cardiac troponin I.     

Phosphorylation of troponin in the heart and skeletal muscle by Ca2+-phospholipid-dependent protein kinase

The phosphorylation of the whole troponin complex and of the cardiac and skeletal troponin components by Ca2+-phospholipid-dependent protein kinase was studied. The activity of enzyme isolated from rat brain by ion-exchange chromatography on DEAE-Sephadex and by affinity chromatography on phosphatidylserine immobilized on polyacrylamide gel was shown to be completely dependent on Ca2+ and phospholipids and was equal to 0.4-0.6 mumol of phosphate/min.mg protein with histone H1 as substrate. The resulting preparation of Ca2+-phospholipid-dependent protein kinase was able to phosphorylate the isolated troponin I; the amount of phosphate transferred per mol of cardiac and skeletal troponin I was equal to 1.1 and 0.4, respectively. The maximal degree of phosphorylation of isolated troponin T by Ca2+-phospholipid-dependent protein kinase was 0.6 mol of phosphate per mol of troponin T both for skeletal and cardiac proteins. The rate and degree of phosphorylation were independent of the initial level of troponin T phosphorylation. Ca2+-phospholipid-dependent protein kinase did not phosphorylate the first serine residue of troponin T, i.e., the site which was phosphorylated in the highest degree after isolation of troponin T from skeletal muscles. The data obtained and the fact that the rate and degree of phosphorylation of troponins I and T within the whole troponin complex are 10-20 times less than those for isolated components provide little evidence for the participation of protein kinase C in troponin phosphorylation in vivo.     

Phosphorylation of myosin light chains in perfused rat heart. Effect of adrenaline and increased cytoplasmic calcium ions.

1. A method was developed for the isolation of essentially pure myosin light chains from perfused rat heart. The phosphorylation of the P-light chains was estimated by hydrolysis and measurement of phosphate released, by electrophoresis in 8 M-urea and by 32P incorporation in perfusion with [32P]Pi. 2. In control perfusions there was 0.5-0.6 mol of phosphate/mol of P-light chain. This was not changed by perfusion with 5 microM-adrenaline for 10-40s. Perfusion for 1 min with medium containing 7.5 mM-CaCl2, or for 30s with medium containing 118 mM-KCl, also did not change the phosphorylation of P-light chains. 3. It is concluded that phosphorylation of P-light chains is not important in mediating the action of inotropic agents in the heart.     

Phosphorylation of an actin.tropomyosin.troponin complex from human skeletal muscle.

A human skeletal actin.tropomyosin.troponin complex was phosphorylated in the presence of [gamma-32 P]ATP, Mg2+, adenosine 3':5'-monophosphate (cyclic AMP) and cyclic AMP-dependent protein kinase (protein kinase). Phosphorylation was not observed when the actin complex was incubated in the absence of protein kinase or 1 microM cyclic AMP. In the presence of 10(-7) M Ca2+ and protein kinase 0.1 mole of [32P]phosphate per 196 000 g of protein was incorporated. This was two-fold higher than the [32P]phosphate content of a rabbit skeletal actin complex but two-fold lower than that of a bovine cardiac actin complex. At high Ca2+, 5.10(-5) M, little change in the phosphorylation of a human skeletal actin complex occurred. Phosphoserine and phosphothreonine were identified in the [32P]phosphorylated actin complex. Polyacrylamide gel electrophoresis in sodium dodecyl sulfate showed that 60% of the label was associated with the tropomyosin binding component of troponin. The inhibitory component of troponin contained 16% of the bound [32P]phosphate. Increasing the Ca2+ concentration did not significantly decrease the [32P]phosphate content of the phosphorylated proteins in the actin complex. No change in the distribution of phosphoserine or phosphothreonine was observed. Half maximal calcium activation of the ATPase activity of reconstitute human skeletal actomyosin made with the [32P] phosphorylated human skeletal actin complex was the same as a reconstituted actomyosin made with an actin complex incubated in the absence of protein kinase at low or high Ca2+.     

Phosphorylation of phospholipids in isolated guinea pig hearts stimulated with isoprenaline.

Phosphorylation of phospholipids was studied in Langendorff perfused guinea pig hearts subjected to beta-adrenergic stimulation. Hearts were perfused with Krebs-Henseleit buffer containing [32P]Pi and freeze-clamped in a control condition or at the peak of the inotropic response to isoprenaline. 32P incorporation into total phospholipids, individual phospholipids and polyphosphoinositides was analysed in whole tissue homogenates and membranes, enriched in sarcoplasmic reticulum, prepared from the same hearts. Isoprenaline stimulation of the hearts did not result in any significant changes in the levels of phosphate incorporation in the total phospholipid present in cardiac homogenates (11.6 +/- 0.4 nmol of 32P/g for control hearts and 12.4 +/- 0.5 nmol of 32P/g for isoprenaline-treated hearts; n = 6), although there was a significant increase in the degree of phospholipid phosphorylation in sarcoplasmic reticulum (3.5 +/- 0.3 nmol of 32P/mg for control hearts and 6.7 +/- 0.2 nmol of 32P/mg for isoprenaline-treated hearts; n = 6). Analysis of 32P incorporation into individual phospholipids and polyphosphoinositides revealed that isoprenaline stimulation of the hearts was associated with a 2-3-fold increase in the degree of phosphorylation of phosphatidylinositol monophosphate and bisphosphate as well as phosphatidic acid in both cardiac homogenates and sarcoplasmic reticulum membranes. In addition, there was increased phosphate incorporation into phosphatidylinositol in sarcoplasmic reticulum membranes. Thus, perfusion of guinea pig hearts with isoprenaline is associated with increased formation of polyphosphoinositides and these phospholipids may be involved, at least in part, in mediating the effects of beta-adrenergic agents in the mammalian heart.     

Additional PKA phosphorylation sites in human cardiac troponin I.

We used mass spectrometry to monitor cAMP-dependent protein kinase catalysed phosphorylation of human cardiac troponin I in vitro. Phosphorylation of isolated troponin I by cAMP-dependent protein kinase resulted in the covalent incorporation of phosphate on at least five different sites on troponin I, and a S22/23A troponin I mutant incorporated phosphates on at least three sites. In addition to the established phosphorylation sites (S22 and S23) we found that S38 and S165 were the other two main sites of phosphorylation. These 'overphosphorylation' sites were not phosphorylated sufficiently slower than S22 and S23 that we could isolate pure S22/23 bisphosphorylated troponin I. Overphosphorylation of troponin I reduced its affinity for troponin C, as measured by isothermal titration microcalorimetry. Phosphorylation of S22/23A also decreased its affinity for troponin C indicating that phosphorylation of S38 and/or S165 impedes binding of troponin I to troponin C. Formation of a troponin I/troponin C complex prior to cAMP-dependent protein kinase treatment did not prevent overphosphorylation. When whole troponin was phosphorylated by cAMP-dependent protein kinase, however, [(32)P]phosphate was incorporated only into troponin I and only at S22 and S23. Mass spectrometry confirmed that overphosphorylation is abolished in the ternary complex. Troponin I bisphosphorylated exclusively at S22 and S23 troponin I showed reduced affinity for troponin C but the effect was diminished with respect to overphosphorylated troponin I. These results show that care should be exercised when interpreting data obtained with troponin I phosphorylated in vitro.     

Regulation of cardiac glycogen synthase by phosphorylation: catalysis of inactivation by cAMP-dependent and cAMP-independent protein kinases and comparison with the phosphorylation of the skeletal muscle enzyme

Glycogen synthase has been purified from bovine heart to near homogeneity by a procedure including zonal sucrose gradient ultracentrifugation. The purified enzyme had a subunit molecular weight of 88,000 ± 2000, an $\text{I}\text{D}$ ratio of between 0.8 and 1.0, and contained less than 0.1 mol of covalently bound phosphate per mole of subunit. The rates, extent, and sites of phosphorylation of the cardiac enzyme were compared with those of skeletal muscle glycogen synthase as catalyzed by both the cardiac cAMP-dependent and a cardiac cAMP-independent protein kinases. The cardiac glycogen synthase was phosphorylated up to 1 mol of phosphate/mol of subunit by the cAMP-dependent protein kinase, to at least 2 mol of phosphate/mol of subunit by the cAMP-independent protein kinase, and to at least 3 mol of phosphate/mol of subunit with the two protein kinases together. There was a linear correlation between the extent of phosphorylation and conversion of cardiac synthase I to the glucose 6-phosphate-dependent form. This correlation was independent of which kinase(s) catalyzed the phosphorylation. Maximum inactivation occurred at an incorporation of 2 mol of phosphate per subunit. Under equivalent conditions, the rates of phosphorylation of cardiac and skeletal muscle glycogen synthase by the cAMP-dependent protein kinase were identical. In contrast, the cardiac enzyme was phosphorylated at a faster rate by the homologous cardiac cAMP-independent protein kinase than was the skeletal muscle synthase by the latter cardiac protein kinase. Analysis of the sites of phosphorylation of the cardiac and skeletal muscle glycogen synthases by CNBr cleavage and trypsin hydrolysis indicated minor differences in the derived phosphopeptides.     

Phosphorylation of cardiac troponin by guanosine 3':5'-monophosphate-dependent protein kinase.

Homogeneous cGMP-dependent protein kinase catalyzes the rapid incorporation of phosphate, specifically into the inhibitory subunit of purified cardiac troponin with a maximal incorporation of 1 mol of phosphate/mol of troponin. When troponin was incubated in the presence of both cGMP- and cAMP-dependent protein kinases, a maximal incorporation of 1 mol of phosphate/mol of troponin was observed which suggested phosphorylation of the same site by the two kinases. Both cyclic nucleotide-dependent kinases had similar Km values for troponin, but the Vmax value for the phosphorylation reaction catalyzed by cAMP-dependent protein kinase was 12-fold greater than the value obtained for cGMP-dependent protein kinase.     

Stretch activates myosin light chain kinase in arterial smooth muscle

Stretching of porcine carotid arterial muscle increased the phosphorylation of the 20 kDa myosin light chain from 0.23 to 0.68 mol [32P]phosphate/mol light chain, whereas stretching of phorbol dibutyrate treated muscle increased the phosphorylation from 0.30 to 0.91 mol/mol. Two-dimensional gel electrophoresis followed by two-dimensional tryptic phosphopeptide mapping was used to identify the enzyme involved in the stretch-induced phosphorylation. Quantitation of the [32P]phosphate content of the peptides revealed considerable light chain phosphorylation by protein kinase C only in the phorbol dibutyrate treated arterial muscle, whereas most of the light chain phosphorylation was attributable to myosin light chain kinase. Upon stretch of either the untreated or treated muscle, the total increment in [32P]phosphate incorporation into the light chain could be accounted for by peptides characteristic for myosin light chain kinase catalyzed phosphorylation, demonstrating that the stretch-induced phosphorylation is caused by this enzyme exclusively.     

Exchange of 20-kDa myosin light chain-bound phosphate during sustained contraction of arterial smooth muscle

K(+)-contracted porcine carotid arterial muscles containing phosphorylated 20-kDa myosin light chains (LC) were exposed to carrier-free [32P]orthophosphate in K(+)-stimulating solution during sustained contraction. The covalently bound LC phosphate was completely replaced by [32P]phosphate, indicating that myosin light chain phosphatase and kinase have ready access to the bound phosphate during the sustained contraction. On average, 0.38 mol [32P]phosphate was incorporated per mole LC during the sustained K+ contraction. This value was about half of the maximal value for [32P]phosphate incorporation into LC, 0.74 mol/mol, in muscles contracted with K+ for 1 min. Assuming that sustained contraction involves the maximal number of cross-bridges attached to actin, the data suggest that half of the attached cross-bridges contain phosphorylated LC.     

Studies on (Na+ + K+)-activated ATPase. XXXVIII. A 100 000 molecular weight protein as the low-energy phosphorylated intermediate of the enzyme.

Phosphorylation of NaI-treated bovine brain cortex microsomes by inorganic phosphate in the presence of Mg2+ and ouabain has been studied at 0 degrees C (pH 7.4) and 20 degrees C (pH 7.0). Nearly maximal (90%) and half-maximal phosphorylation are achieved at 20 degrees C within 2 min with 50--155 and 5.6--17 muM 32Pi, respectively, and at 0 degrees C within 75 s with 300--600 and 33--66 muM 32Pi, respectively. Maximal phosphorylation yields 146 pmol 32P - mg-1 protein. Without ouabain (20 degrees C, pH 7.0) less than 25% of the incorporation observed in the presence of ouabain is reached. Preincubation of the native microsomes with Mg2+ and K+, in order to decompose possibly present high-energy phosphoryl-bonds prior to ouabain treatment, does not affect the maximal phosphate incorporation. This indicates that the inorganic phosphate incorporation is not due to an exchange with high-energy phosphoryl-bonds, which might have been preserved in the microsomal preparations. Phosphorylation of the native microsomes by ATP in the presence of Mg2+ and Na+ reaches 90 and 50% maximal levels within 15--30 s at 0 degrees C and pH 7.4 at concentrations of [gamma-32P]ATP of 5--32 and 0.5--3.5 muM, respectively. The maximal phosphorylation level is 149 pmol 32P-mg-1 protein, equal to that of ouabain-treated microsomes phosphorylated by inorganic phosphate. Both inorganic phosphate and ATP phosphorylate on site per active enzyme subunit of 135 000 molecular weight. From the equilibrium constants for the phosphorylation of ouabain-treated microsomes by inorganic phosphate at 0 degrees C and 20 degrees C standard free-energy changes of --5.4 and --6.8 kcal/mol, respectively, are calculated. These values yield a standard enthalpy change of 14 kcal/mol and an entropy change of 70 cal/mol - degree K. This characterizes the reaction as a process driven by an entropy change. The intermediate formed by phosphorylation with Pi has maximal stability at acidic pH, as is the case for the intermediate formed with ATP. Solubilization in sodium dodecyl sulfate stabilizes the phosphoryl-bond in the pH range of 4--7. The non-solubilized preparation has optimal stability at pH 2--4, the level of which is equal to that of detergent-solubilized intermediate. Sodium dodecyl sulfate gel electrophoresis of the microsomes at pH 3, following incorporation of 32Pi yields 11 protein bands, only one of which (mol. wt 100 000--106 000) carries the radioactive label. This protein has the same molecular weight as the protein, which is phosphorylated by ATP in the presence of Mg2+ and Na+.