Mechanism of Activation of a Cyclic Adenosine 3′:5′-Monophosphate Phosphodiesterase from Bovine Heart by Calcium Ions

largely freed of the protein activator, possess low enzyme activity which is independent of Ca”+. Addition of excess protein activator stimulates the enzyme activity 6to lofold. This activation by the protein activator is shown to be completely dependent on the presence of low concentrations of Ca2+. The concentration of Ca?+ required to give 50% of the maximal activation is 2.3 NM. An equilibrium binding study has shown that “X!a binds to the protein activator. A Scatchard plot exhibits two linear regions suggesting the presence of two sets of Ca2+ binding sites on the protein with different affinities: one high affinity site and two low affinity sites per protein activator. The dissociation constants for Ca2+ bound at the high and low affinity sites are 3 and 12 pM, respectively. The results suggest that the complex of Ca2+ and the protein activator is the true activator for CAMP phosphodiesterase.

largely freed of the protein activator, possess low enzyme activity which is independent of Ca"+. Addition of excess protein activator stimulates the enzyme activity 6-to lofold.
This activation by the protein activator is shown to be completely dependent on the presence of low concentrations of Ca2+. The concentration of Ca?+ required to give 50% of the maximal activation is 2.3 NM. An equilibrium binding study has shown that "X!a binds to the protein activator.
A Scatchard plot exhibits two linear regions suggesting the presence of two sets of Ca2+ binding sites on the protein with different affinities: one high affinity site and two low affinity sites per protein activator.
The dissociation constants for Ca2+ bound at the high and low affinity sites are 3 and 12 pM, respectively.
The results suggest that the complex of Ca2+ and the protein activator is the true activator for CAMP phosphodiesterase.
The understanding of regulatory properties of ch1IP' phosphodiesterase is fundamental to the delineation of how cA;\IP effects on various physiological and biochemical functions are controlled.
Degradation of cAhIP to 5'-A1\IP by this enzyme is the only well established mechanism for the disposal of this cyclic nucleotide in mammalian cells. Sutherland and Rall (1) were the first to demonstrate this enzyme. Subsequently, Butcher and Sutherland (2) and Drummond and Perrot-Yee (3) partially purified this enzyme from bovine heart and rabbit brain, respectively.
Recent, studies have indicated that in * This work was supported by Grant MT-2381 from the Medical Research Council of Canada.
any of the mammalian tissues examined, cAhIP phosphodiesterase is present in multiple forms having different molecular weights and catalytic propcrtics (4, 5). Of particular interest is the observation of Kakiuchi and Yamazaki (6) that a rat brain cAyI1' phosphodicstcrase is inhibited by low concentrations of EGTA, thus indicating that it is a Ca2+-dependent enzyme. I\Iore recently, Miki and Yoshida (7) have shown that in the rat this Ca2f-dependent enzyme is restricted to the cerebrum.
The present study shows that a Ca2+-dcpendcnt cA1IP phosphodicsterase also exists in bovine heart. Cheung (8), Goren and Rosen (9), and Kakiuchi et al. (10) have independently shown that a specific protein activator of CAMP phosphodiesterase is present in several mammalian tissues. Recently, a procedure has been developed to purify this protein activator from bovine heart to apparent homogeneity (II). In addition, an enzyme preparation largely freed of the protein activator has been obtained from bovine heart (11). The present study shows that the activation of this CAMP phosphodiesterase is completely dependent on the addition of the protein activator and Ca2+.

MATERIALS
AKD METHODS d[uterials-Beef hearts were obtained fresh from a local slaughterhouse, Burns Food Ltd. of Winnipeg.
The fresh hearts were cut into small pieces (approsimately 1 cubic inch) and stored frozen at -20" for 1 to 4 weeks prior to use. DEAE-cellulose (medium capacity) and Chelex 100 (minus 400 mesh) were obtained from Bio-Rad.
Preparation of Activator-dejkient CAMP Phosphodiesterase-Frozen beef heart (500 g) was thawed, put through a meat mincer, and homogenized for 10 s at low speed in a Waring Blendor with 2.5 volumes of 0.1 M Tris-HCl-2 mM EDTA, pH 7.5. The homogenate was centrifuged at 10,000 x g at 4" for 20 min. The supernatant solution was titrated to pH 8.6 with 1 N NaOH, and powdered ammonium sulfate was added to give 50 y0 saturation. This was immediately centrifuged at 10,000 x g at 4" for 20 min.
The supernatant fluid was used for the preparation of the protein activator.
The pellet was dissolved in a minimal volume of a solution containing 0.02 M Tris-HCl, 0.08 M NaCl, I lllM h!Ig2+, and 1 mM imidazole, pH 7.5, and the solution was dialyzed overnight against the same buffer. The dialyzed enzyme was frozen and thawed and then centrifuged at 100,000 x g for 1 hour at 4". The supernatant solution was applied to a DEAEcellulose column (4.5 x 36 cm) which had been equilibrated with a solution containing 0.02 M Tris-HCl, 0.08 M KaCl, 1 mM ?vIg2+, and 1 m&f imidazole, pH 7.5. The column was washed with the same buffer until no more protein was eluted.
A negligible amount of cnzgme activity was eluted by this buffer. The column was then developed with the same buffer containing 0.22 M XaCl.
The eluted protein fractions which contained the activator-deficient CAMP phosphodiesterasc were pooled and dialyzed against 0.02 IM Tris-HCl containing 1 mar Mg2+ and 1 mM imidazole, pH 7.5. The dialyzed sample was stored frozen at -20" in small aliquots until use. The maximal activation of the enzyme by the protein activator varied from 6-to IO-fold depending on the preparation.
Preparation of Protein Activator of CAMP Phosphodiesterase-The protein activator was prepared from the 507, ammonium sulfate supernatant, by a procedure described previously (11). Since it was found that the purified samples sometimes contained a small amount of low molecular weight contaminant which could be removed by gel filtration on a Sephadex G-50 column, protein activator preparations used for the Ca2+ binding study were always subjected to filtration on a Sephadex G-50 column.
Assay of CAMP Phosphodiesterase-The enzyme activity was measured by the method of Butcher and Sutherland (2) with a slight modification (12). This procedure involved the conversion of 5'-,4hIP, the product of the CAMP phosphodiesterase reaction, to adenosine and inorganic phosphate by 5'-nucleotidase. The 5'Qmcleotidase reaction was carried out either concurrently with the phosphodiesterase reaction (a one-stage assay) or after the termination of the phosphodiesterase reaction by boiling (a two-stage assay). The reaction mixture, in a volume of 0.9 ml, contained, in addition to CAMP phosphodiesterase and the protein activator, 25 InM Tris, 25 mM imidazole, 3 mM magnesium acetate, 1.2 mM CAMP, and 0.2 unit of 5'nucleotidase.
In experiments  in which it was necessary to control precisely the concentration of calcium, the CAMP phosphodiesterase activity was assayed by the two-stage method.
5'-Nucleotidase retained an appreciable amount of calcium even after desalting by gel filtration with Sephadex G-25.
One unit of CAMP phosphodiesterase activity is equivalent to the amount of enzyme which, when maximally activated by both the protein activator and Ca2+, hydrolyzed 1 pmole of CAMP per min at 30" under standard conditions. Assay of Protein Activator of cAXP Phosphodiesterase-The protein activator was assayed by a previously described procedure (11) which involved the measurement of the extent of stimulation of a fixed amount of activator-deficient beef heart cAhlP phosphodiesterase and then compared with a standard curve. Since it was found during this study that Ca2f could effect the activation of cBMP phosphodiesterase by the protein activator, 0.1 rnhl Ca2+ was routinely included in the reaction mixture for activator assays. One unit of protein activator activity is defined as the amount which is required to give 50% stimulation of 0.012 unit of CAMP phosphodiesterase. Removal of Ca2+ from Reagents-Chelex 100, a resin specific for chelating divalent cations, was used for removing Ca*f from all of the stock solutions.
The resin was washed once with 1 N HCI and then with 1 N ?JaOH prior to the packing of the column. The packed columns were then washed with double distilled water.
Double distilled water, Tris-HCl (0.3 M), and imidazole (0.3 M) were separately treated for the removal of Ca2+ by passage through Chelex 100 columns (20 x 3 cm). CAMP solutions (10.8 m&f) were passed through a Chelex 100 column (6 X 1.5 cm) to remove Ca2+. Plastic columns and connections were used in the column chromatography.
The purified reagents were always stored in plastic containers and all reactions were carried out in plastic vessels. A Perkin-Elmer atomic absorption spectrophotometer, model 303, was used to monitor the concentration of calcium in these stock reagents.
The limit of detection of calcium by this instrument is 4 ppm. After Chelex 100 treatment, the calcium content of stock reagents was below this limit of detection.
Calcium was removed from the protein activator and CAMP phosphodiesterase by treatment with 0.5 mM EGTA for 20 min at 4"; these materials were then desalted by gel filtration through Sephadex G-25 columns (30 x 1.5 cm). Chelex loo-treated water and buffer were used at all steps. A4tomic absorption spectrophotometry showed that the Ca2+ in both the enzyme and the protein activator as less than 4 ppm.
Magnesium acetate and cobalt chloride were essentially free of calcium contamination. Strontium chloride was contaminated with Ca*+ to the extent of 0.00270.
Since the highest concentration of Sr2+ used was 0.22 mnl, the resultant contribution of Ca2+ was less than lo-* AI.
Binding of Ca2+ by PuriJed Protein Activator of CAMP Phosphodiesterase-The gel filtration method of IIummel and Dreyer (13), as modified by Fairclough and Fruton (14), was used to determine the binding of Ca z+ by the purified protein activator of CAMP phosphodiesterase.
A column (26 x 0.9 cm) of Sephadex G-25 was equilibrated at 24" with buffer containing 25 nlM Tris-HCl, 25 m&f imidazole, and 3 mM magnesium acetate with a known concentration of Ca2+ plus WYa2+. The column used was a plastic Pharmacia Kg/30 column. Chelex lOOtreated reagents Ivere used throughout. Desalted protein activator, 86.4 Fg in 0.4 ml, was used for each experiment.
The gel filtration was carried out at 24" at a flow rate of 3 ml per hour and 0.6-ml fractions were collected.

EGTA Inactivation
of Bovine Heart CAMP Phosphodiesterase- Kakiuchi and Yamazaki (6) have found that low concentrations of EGTA inhibit rat brain CAMP phosphodiesterase. Fig. 1 shows that this chelating agent in the presence of excess Mg2f also inactivates CAMP phosphodiesterase of crude extracts of bovine heart to a maximum of 67% inhibition.
If 30 pM Ca*+ is added to the enzyme assay mixture, much higher concentrations of EGTA are needed to inhibit the enzyme.
This result suggests that a Ca*+-dependent CAMP phosphodiesterase exists in bovine heart.
In the absence of the chelating agent, however, 30 PM Ca*+ activates the enzyme by less than 10% (Fig. I). 1. Inhibition by EGTA of CAMP phosphodiesterase in crude extract of beef heart. Dialyzed 10,000 X g supernatant solution of a homogenate of beef heart was used for this experiment. CAMP phosphodiesterase activity was measured by the one-stage assay. No Ca*+ added, 0 ; Ca2+ (30 pM) added, 0. Dialyzed activator-deficient beef heart CAMP phosphodiesterase was used. Enzyme activity was measured by the one-stage assay. No addition of the protein activator, 0; 25 units of the protein activator added, 0, Presumably, this is because a sufficient amount of endogenous Ca*+ is present in the assay mixture.
Enzyme preparations, largely freed of the protein activator by chromatography on DEAE-cellulose columns, always possess low but significant enzymic activity. Fig. 2 shows that this low basal enzymic activity is not inhibited by EGTA in the presence of excess Mg2f and is therefore independent of Ca*+. When excess protein activator is added to these activator-deficient enzyme preparations, the enzyme activity is stimulated 6-to lo-fold.
This activation by the protein activator is completely abolished by low concentrations of EGTA in the presence of excess Mg2+. This indicates that the activation of bovine heart CAMP phosphodiesterase requires the simultaneous presence of protein activator and low concentrations of Ca2+.
Ca2+ Activation of CAMP Phosphodiesterase-Although the preceding results suggest that there is a Ca2+-dependent CAMP phosphodiesterase in bovine heart, the possibility that EGTA has a direct effect on CAMP phosphodiesterase or that a metal ion other than Ca2+ is responsible for the enzyme activation cannot be excluded.
To establish unequivocally that Ca*+ activates the enzyme, removal of contaminating Ca2f in the reaction mixture and direct demonstration of the Ca2f activation are essential. Fig. 3 shows that when reagents and protein samples relatively free of the contaminating Ca*+ are used, activation of bovine heart CAMP phosphodiesterase by Ca*+ can be shown.
The dependence of the enzyme activity upon Ca2+ concentration has been examined at different levels of the protein activator.
In the absence of the protein activator, increasing the Ca2+ concentration to 0.2 mM results in little enzyme activation.
At a saturating level of the protein activator, Ca*f can bring about a lo-fold increase in the enzyme activity. At a lower level of the protein activator, the maximal Ca2+ activation of bovine heart CAMP phosphodiesterase is also lowered. In addition to the extent of activation, the Ca2+ concentration I  required to achieve 50% of the maximal activation, A50%, also depends on the amount of the protein activator in the enzyme assay. The Aso% values at 1.4 and 13 units of the protein activator are 3.6 and 2.3 PM, respectively.
The results presented above demonstrate that activation of CAMP phosphodiesterase by Ca2+ is dependent upon the presence of the protein activator, and the data of Fig. 4 show that t'he activation of the enzyme by the protein activator is dependent upon Ca2+. In the absence of Ca2+, CAMP phosphodiesterase activity is not stimulated by the protein activator.
At 100 pM Ca2+, however, CAMP phosphodiesterase is activated by increasing concentrations of the protein activator to a maximal activation of 600 "iO. Both the extent of the enzyme activation and the concentration of the protein activator required for 50% maximal activation are functions of Ca*+ concentration. At a lower concentration of Ca2+, the enzyme is activated by the protein activator to a smaller extent, and more protein activator is needed to achieve 50% maximal activation of the enzyme. Thus, results in Figs. 3 and 4 indicate that activation of CAMP phosphodiesterase is achieved only when both Ca2f and the protein activatoi are present. Furthermore, the two activators may enhance each other's efficiency in the enzyme activation.
Activation of CAMP Phosphodiesterase by Other Metal Ions-Since all enzyme assays in the present study were carried out in the presence of 3 mM ?hfg *+, the demonstration of the Ca2+ activation indicates that Mg2+ does not substitute for Ca2f in the activation of CAMP phosphodiesterase. However, Mg2+ is essential for the catalytic activity of CAMP phosphodiesterase since the enzyme is inactive in the presence of Ca2+ alone. Thus, the enzyme depends on both Mg2+ and Ca2+ for its full activity. To further study the specificity of Ca2+ activation of CAMP phosphodiesterase, the enzyme activity in the presence of 3 mM Mg*+ and one of several diralent metal ions has been examined. Table I shows that Sr*+, Co*+, and Mn2+ are the only metal ions which exhibit significant enzyme activation at a concentration of 30 PM.
The magnitudes of enzyme activation by these metals are, however, much less than that by Ca2+.
In order to determine whether the low levels of activation achieved by Sr2f and Co2f are due to low affinities of these metal 5953 The enzyme activity was measured in the presence of 13 units of Ca2+-free protein activator under conditions as described in the legend for Fig. 3. Activation by Co*, 0 ; activation by W+, A; activation by Cast, l .
ions for the binding site or low maximal enzyme activations, CAMP phosphodiesterase activity in the presence of excess protein activator has been examined as a function of the metal ion concentration. Fig. 5 shows that the enzyme is maximally activated by Sr2+ and Co*+ to 900 and 300%, respectively. Under the same condition, maximal Ca2+ activation of the enzyme is 100070. Concentrations of Sr2+ and Co2+ required to provide 507, maximal activation are 36.3 and 19.2 pM, respectively; about 10 to 20 times higher than that of Ca2+. Thus, Ca2+ appears to be the most effective metal activator for CAMP phosphodiesterase.
Interaction of Ca2+ and Protein Activator-One possible ex- planation for the mutual dependence of Ca2+ and the protein activator in the activation of CAMP phosphodiesterase is that the two activators have to combine to form a metal protein complex to activate the enzyme.
The possible formation of the Ca2+-protein activator complex has been investigated by the equilibrium binding technique on a Sephadex G-25 gel filtration column (13). Fig. 6 shows the elution profile for a typical binding experiment.
The appearance of 4%a peak and troughs in the profile is indicative of the binding of Ca2f to the protein activator.
The radioactivity peak coincides exactly with the activity peak of the protein activator.
In most experiments, double troughs have been observed in the elution profile, but the origin and significance of the double troughs are not clear. For the calculation of the amount of bound Ca2+, only the data at peak regions have been used.
In a preceding section, it has been shown that Sr2+ can replace Ca*+ in the activation of CAMP phosphodiesterase.
This metal ion, therefore, is expected to compete for Ca2+ binding sites on the protein activator if Ca2+ binding is indeed involved in the enzyme activation.
As is shown in Fig. 6, binding of Ca2f to the protein activator may be significantly reduced in the presence of 500 pM Sr2+. At a free CaZf concentration of 0.7 pM, the amount of Ca2f bound per mole of protein is changed from 0.205 in the absence of Sr*+ to 0.092 mole in the presence of Sr2+. Although the result does not show that the -two ions compete for the same binding site, it does agree with such an interpretation.
The stoichiometry of the interaction between Ca2+ and the protein activator and the dissociation constant for the complex have been determined from a Scatchard plot (Fig. 7). The Scatchard plot consists of two linear regions, thus having two different slopes. The result suggests that there are two types of Ca2+ binding sites on the protein activator having different affinities.
From the slopes, the dissociation constants of Ca2+ for the high and low affinity sites are calculated t,o be 2.9 and 11.8 PM, respectively.
Since kinetic studies indicate that Ca2f concentration required for 50% enzyme activation at a saturating \ \ amount of the protein activator is 2.3 FM, it may be suggested that only the high affinity Ca 2+ binding site is involved in the enzyme activation. The stoichiometry of the interaction between Ca*+ and the protein activator may be calculated from the intercepts on the horizontal axis of the Scatchard plot. Extrapolated lines for the high and low affinity sites intercept at 1.04 and 3.25 moles per mole of the protein activator, respectively.
This indicates that there is one high affinity Ca2+ binding site and 2 to 3 low affinity Ca*+ binding sites per molecule of the activator.

DISCUSSION
Based on experiments carried out in EGTA-Ca2+ buffer, Kakiuchi and his co-workers (6) have concluded that a Ca2+dependent CAMP phosphodiesterase is present in rat brain. In the present study, a similar enzyme is also found in bovine heart. 13rostrom et al. (15), in their study of Ca2+ activation of phosphorylase kinase, have indicated the advisability of the use of Ca2f-free reagents rather than EGTA-Ca2+ buffer to control the concentration of free Ca2+ in the study of Ca*+ effects on enzymes. This is because EGTA also chelates other metal ions, and the possibility of a direct effect of EGTA on the enzyme activity is difficult to rule out completely.
In addition, Ca2+ concentrations in enzyme assays can be more accurately determined if Ca2f-free reagents are used.
Kakiuchi and his co-workers (16) observed that the protein activator caused an increase in the maximum Ca2f activation of rat brain CAMP phosphodiesterase.
Furthermore, in the presence of the protein activator, a significantly lower concentration of Ca2f was needed to activate the rat brain CAMP phosphodiesterase (18). These observations are confirmed and extended in the present study with the bovine heart enzyme.
The activation of the enzyme requires the simultaneous presence of both Ca2+ and the protein activator. This may be explained by postulating that the enzyme is activated by the complex of the protein activator and CaZ+. Such a postulate is supported by the observation that the purified activator binds Ca2f. It is significant that the observed dissociation constant for the Ca*f bound at the high affinity site of the protein activator is very similar to the kinetic constant of Ca2+ activation of the enzyme. Furthermore, Sr2+, which can substitute for Ca*+ in the enzyme activation, is shown to reduce the Ca2f binding to the protein activator.
The fact that the functions and the metabolism of CAMP and Ca*+ are closely related has been pointed out in a review by Rasmussen et al. (17). In the heart, both Ca*+ and CAMP have been implicated in the hormone-stimulated myocardial contraction.
Several groups of investigators (18-20) have suggested that CAMP may control the free Ca2+ level in cardiac muscle by facilitating the inflow of Ca2+ into the cell. Kirchberger et al. (21) have shown that the uptake of Ca2+ by cardiac microsomes may be stimulated by CAMP in the presence of CAMP-dependent protein kinase. These observations may suggest that the inotropic effect of CAMP results from its effect on Ca2+ metabolism.
The present observation that CAMP phosphodiesterase from bovine heart is activated by Ca2+ suggests that the concentration of CAMP could in turn be regulated by Ca2+ in cardiac muscle. That this Ca*f activation of the enzyme is operative in intact hearts is supported by the observation of Yamm et al. (22) that CAMP concentrations in rat hearts can be increased or decreased upon perfusion of the hearts with a Ca2+-free or a Ca2+-rich medium, respectively.
Furthermore, the ranc.e of Ca*+ concentrations effective in the activation of d CAMP phosphodiesterase, 1 to 10 PM, also suggests that this CaZf activation may have an important regulatory role in the cardiac contraction.
It has been suggested (23) that the Ca2+ level during the myocardial contraction cycle fluctuates in the range of 0.1 to 10 PM. Although it is not clear as to how Cazf activation of c.lMP phosphodiesterase contributes to the regulation of myocardial contraction, it could conceivably be an important mechanism for the removal of the excess CAMP used for the excitation of the muscle. In addition, the Ca2+ activation of CAMP phosphodiesterase could even be involved in the control of myocardial contraction in the absence of the hormonal stimulation. Brooker (24) has recently demonstrated the fluctuation of CAMP concentration during the contraction cycle of electrically stimulated frog ventrical strips.
It is interesting to note that the activation of CAMP phosphodiesterase by Ca*+ is analogous to the Ca*+ activation of myosin ATPase in that they both depend on the binding of Ca2f to specific protein (25). Furthermore, both these Ca2+ binding proteins are highly acidic with molecular weights of about 20,000 (11,26).
Recently, we have shown that the interaction between bovine heart CAMP phosphodiesterase and the protein activator may be modulated by CAMP (11,12). That the protein activator has an absolute requirement for Ca2+ for its action suggests that Ca2+ and CAMP may interact synergetically in the activation of CAMP phosphodiesterase.
The significance of such interaction in the regulation of intracellular concentrations of CAMP and Ca2+ is, however, not clear.
Bovine heart CAMP phosphodiesterase has recently been extensively purified in two different laboratories (27,28). In both cases, two major molecular forms of the enzymes have been observed.
Hrapchak and Rasmussen (27) did not detect t,he existence of any specific intracellular protein activator. On the other hand, Goren and Rosen (28) reported that their en-zyme preparations could be activated 100% by a nondialyzable activator.
The present study shows that the protein activator can activate the activator-deficient CAMP phosphodiesterase 6-to lo-fold.
The reason for this discrepancy is not clear and is currently under investigation.