Effect of CAMP and ATP on the Reassociation of Phosphorylated and Nonphosphorylated Subunits of the CAMP-dependent Protein Kinase from Bovine Cardiac Muscle*

We studied the extent of reassociation of phosphorylated and nonphosphorylated CAMP-binding protein (Rz) with the catalytic subunits (C) in the presence of CAMP and ATP using chromatography on w-aminohexyl agarose to resolve the various forms of the enzyme, and a purified cyclic nucleotide phosphodiesterase to bring about changes in the concentration of CAMP. The following observations were made. (a) The extent to which R, (phosphorylated and nonphosphorylated) and C reassociate is much smaller in the presence than in the absence of CAMP. (6) Both forms of Rz reassociated completely with catalytic subunits when the CAMP had been hydrolyzed by cyclic nucleotide phosphodiesterase.

From the Departments of Molecular Pharmacology and Medicine, Albert Einstein College of Medicine, Bronx, New York 10461 Self-phosphorylation of the bovine cardiac muscle adeno. sine 3':5'-monophosphate (CAMP&dependent protein kinase results in retardation of the rate of reassociation of its isolated subunits in the absence of CAMP (Rangel-Aldao, R., and Rosen, 0. M. (1976) J. Biol. Chem. 251, 3375-3380).
We have now studied the extent of reassociation of phosphorylated and nonphosphorylated CAMP-binding protein (Rz) with the catalytic subunits (C) in the presence of CAMP and ATP using chromatography on w-aminohexyl agarose to resolve the various forms of the enzyme, and a purified cyclic nucleotide phosphodiesterase to bring about changes in the concentration of CAMP. The following observations were made. (a) The extent to which R, (phosphorylated and nonphosphorylated) and C reassociate is much smaller in the presence than in the absence of CAMP. (6) Both forms of Rz reassociated completely with catalytic subunits when the added CAMP had been hydrolyzed by cyclic nucleotide phosphodiesterase.
(c) The unphosphorylated CAMP-binding protein achieved 50% reassociation at a concentration of CAMP lo-fold higher than its phosphorylated counterpart. (d) Reassociation of both the phosphorylated and unphosphorylated protein kinase involved the formation of a new species of enzyme prior to the formation of the holoenzyme tetramer (R&). This putative intermediate was assigned an RBC structure; it possessed both regulatory and catalytic subunits, a ratio of R to C activities twice that observed in the tetramer (R&j, and a molecular weight consistent with this subunit composition (137,000). (e) Millimolar concentrations of ATP limited the extent of reassociation of both phosphorylated and unphosphorylated R, with C. This effect was not dependent upon divalent cations and did not involve a phosphotransferase reaction. Similar inhibitions were observed with ADP, AMP, and adenosine. At the same concentration, these latter compounds also inhibited the phosphotransferase activity of * protein kinase, suggesting that interaction with the catalytic subunit of protein kinase may be involved in their ability to inhibit reassociation.
It is concluded that in the presence of physiological concentrations of ATP and CAMP, the extent to which protein kinase is reassociated to the inactive holoenzyme may be significantly affected by the state of phosphorylation of R,. for 60 min at 4". The recovery of catalytic activity was 75 to 85%. The concentrated fractions from each column were then subjected to electrophoresis as described under "Materials and Methods," The gels were stained and scanned at 550 nm using a Gilford spectrophotometer with scanning and recording devices.
Reassociation was quantitated as described in the legend to Fig. 1. In the scans, reading from right to left or top to bottom of the gel, the first peak is the CAMP-binding protein dimer, &, the second peak is the putative intermediate, R,C, and the third peak is the holoenzyme, R,C,. The minor peak following R,C, is a contaminant in the protein kinase preparation. A, reassociation of nonphosphorylated R.z and C; B, reassociation of phosphorylated R, and C.  tration of CAMP appears to be an important factor in the reassociation of both phosphorylated and nonphosphorylated subunits of protein kinase (Figs. 1 and 5). Second, in the presence of CAMP, R2CZ accumulates more rapidly than phosphorylated holoenzyme. The concentration of CAMP must be lowered lo-fold (Fig. 1, time period 10  The difference in reassociation of the two forms of R, was confirmed by analysis of the products of reassociation (Fig. 2) in polyacrylamide gels. These experiments suggest that the reassociation of protein kinase subunits involves the formation of a trimer (R&1 prior to formation of the holoenzyme tetramer. Designation of the R,C structure was based upon the findings that it possessed both regulatory and catalytic subunits, a ratio of R to C activities twice that observed in the holoenzyme (R,C,) and the predicted molecular weight (137,000).
The inhibitory effect of ATP on reassociation of F& with C is worth noting. In the presence of 1 mM ATP and physiological concentrations of CAMP (2 to 20 PM), R,, unlike phosphorylated &, exhibited substantial reassociation and regeneration of the inactive holoenzyme (8, 31). The inhibitory effect of ATP on the reassociation of both forms of protein kinase accentuated the differences between the interactions of R, and phosphorylated I& with C. The effect of ATP was unrelated to its ability to serve as a phosphate donor for selfphosphorylation of protein kinase since it occurred in the absence of Mg*+ under conditions which do not allow the phosphotransferase reaction to take place. With the exception of GTP, the effect was specific for adenine-containing molecules. The binding of CAMP, at equilibrium, to either form of K, was not affected by ATP. On the other hand, there was a correlation between ability to retard the onset of reassociation and ability to inhibit the catalytic activity of protein kinase. Since the most potent inhibitors, ADP and adenosine, were competitive inhibitors of ATP in the phosphotransferase reaction, it is possible that they can bind to an ATP-specific site on C even in the absence of divalent cation. Haddox et al. (32) and Hofmann et al. (12) demonstrated high affinity ATP:Mg2+ binding to the Type I protein kinase of skeletal muscle. However, Hofmann et al. (12) were unable to demonstrate these kinds of sites in the enzyme used here, Type II protein kinase from bovine cardiac muscle. Although the mechanism by which ATP limits the extent of reassociation is unknown, it is likely that at physiological concentrations of ATP (71, in uivo, the holoenzyme is phosphorylated. If the kinase is, in fact, localized in the cell, it may be only partially dissociated by physiological elevations of CAMP. The ability of CAMP to keep the enzyme activated could be potentiated by ATP. In order to inactivate protein kinase, i.e. to permit reassociation to occur in the presence of ATP, the concentration of CAMP would have to fall dramatically so that C could reassociate with phosphorylated &. Alternatively, dephosphorylation of & would enable reassociation to take place in the absence of a change in the concentration of CAMP. The discovery of a cardiac muscle phosphoprotein phosphatase that acts on the isolated phosphorylated R, rather than the phosphorylated holoenzyme (9, 33) makes this latter mechanism plausible.