INTRODUCTION

Phosphorylase b kinase (PbK)¹ is a regulatory enzyme of glycogenolysis that integrates metabolic, hormonal, and neural signals (for review, see Refs. 1 and 2). In skeletal muscle, its dependence on Ca²⁺ ions for activity couples contraction with energy production (3). The enzyme has four copies each of four different subunits ()₄. The  subunit, with a mass of 44.7 kDa (4), is catalytic; and the , , and  subunits, with masses of 138.4, 125.2 (5, 6), and 16.7 (7) kDa, respectively, are regulatory. The  subunit is an endogenous molecule of tightly bound calmodulin (CaM) (8) that is undoubtedly responsible for the Ca²⁺ dependence of the enzyme activity, given that complexes containing the  subunit (, , and PbK) are stimulated by Ca²⁺ (9), whereas the free  subunit is not (10). Two distinct, high affinity binding domains for CaM/Ca²⁺ have been identified near the COOH terminus of the  subunit (11); the  subunit has been shown to interact with  in the holoenzyme (12); and, CaM/Ca²⁺ stimulates the activity of free isolated  subunit (10). Thus, the primary site of interaction for the  subunit within the holoenzyme is presumed to be on the catalytic  subunit. PbK can be activated through a variety of mechanisms, including phosphorylation (13), proteolysis (14), and allosterically by ADP (15), but none of the variously activated forms of the enzyme loses the capacity to be stimulated by Ca²⁺ ions. Although there have been a large number of studies on the relationship of Ca²⁺ to activity, little is known about the effect of Ca²⁺ ions on the structure of the holoenzyme, either activated or nonactivated. A recent communication has suggested that Ca²⁺ increases the accessibility of specific loci within the COOH-terminal region of the  subunit of nonactivated enzyme (16).

In addition to the stimulatory effect of Ca²⁺ mediated by the endogenous CaM ( subunit), which is essentially bound irreversibly to PbK, Ca²⁺ is also required for the reversible binding of exogenous CaM to a different site on the holoenzyme (17-19). This exogenous CaM is termed , and it binds in a stoichiometry of one molecule/each protomer (12, 20). This binding of /Ca²⁺ further stimulates activity past that obtained with Ca²⁺ alone, especially for nonactivated PbK (19, 21). Based on cross-linking and peptide binding studies (12, 20, 22), both the  and  subunits apparently contribute to the binding site for /Ca²⁺, although it is again the  subunit that is ultimately stimulated. Even though the binding sites for  and are distinct and on different subunits, skeletal muscle troponin C can substitute for both  in activating isolated  subunit (23) and in activating the holoenzyme (21, 24, 25). The structural effects induced by the binding of /Ca²⁺ to PbK are only slightly more fully characterized than the effects of Ca²⁺ alone. This laboratory has found that /Ca²⁺ increases the binding to nonactivated PbK (26) of a monoclonal antibody specific for an epitope (27) that occurs at the base of the peptide binding lobe of the  subunit (28) and also increases the incorporation of putrescine into that subunit by transglutaminase (29). As in the case of the structural influence of Ca²⁺ alone cited above (16), both of these effects were interpreted as manifestations of increased accessibility of particular regions of the  subunit induced by the binding of /Ca²⁺ (26, 29). Inasmuch as the transglutaminase used in that study required Ca²⁺, the effect of /Ca²⁺ on the structure of  that it detected was necessarily greater than that caused by Ca²⁺ alone. The incorporation of putrescine into the  and  subunits was also influenced by /Ca²⁺, with modification of  decreased and  increased (29). Although the effect on  could be due to direct steric inhibition caused by the specific binding of /Ca²⁺, the stimulatory effect on modification of  indicates a conformational change in that subunit induced by /Ca²⁺.

In this study, we have used mono- and bifunctional modifying agents as conformational probes to compare the effects of Ca²⁺ alone versus /Ca²⁺ on the structure of PbK. Using the monofunctional reagents, we have further addressed the issue of relative changes in the conformation of the  subunit induced by the two activators. The bifunctional reagents have allowed a screening for relative changes in the interactions of all subunits (as detected by cross-linking) initiated by the binding of Ca²⁺ to the  subunit or of /Ca²⁺ to the / subunits. These conformational probes have also been used to evaluate whether activation of the enzyme through other mechanisms alters the structural changes induced by Ca²⁺ and /Ca²⁺. The results obtained indicate that, although Ca²⁺ has structural effects characteristic of other activators, it also has distinct effects, and these are observed with both nonactivated and activated enzyme; /Ca²⁺ appears, for the most part, to amplify the structural changes brought about by Ca²⁺ alone. A preliminary account of this work has been published (30).

EXPERIMENTAL PROCEDURES

Nonactivated and autophosphorylated PbK used in this study were described in the accompanying report (31). All experiments described herein were repeated a minimum of three times using three different PbK preparations. Phosphorylase b and bovine serum albumin were obtained as described (31), as were the four mAbs and their detection conjugates. Bovine brain CaM was isolated as described previously (32). Biotinylated CaM (at Lys-94) was prepared by treatment with N-hydroxysuccinimidyl biotin at pH 6.0 and purified over DEAE-Spherogel by the procedure of Mann and Vanaman (33). The concentrations of PbK, phosphorylase b, bovine brain CaM, and BtCaM were determined as described (31, 33).

Reductive methylation of PbK was carried out for 30 min at 30 °C essentially as described (34). Final concentrations in the reaction were: 1.73 µM PbK protomers, 50 mM Hepes, pH 6.8 or 8.2, 1.0 mM EDTA, 2.5 mM NaCNBH₃, and 2.92 mM [³H]CH₂O (1.28 Ci/mol, American Radiochemicals, St. Louis). The enzyme was also modified under identical conditions, but in the presence of 1.25 mM CaCl₂ ± 1.73 µM CaM.

Carboxymethylation of PbK with [³H]iodoacetic acid was carried out for 20 min at 30 °C. Final concentrations in the reaction were: 1.73 µM PbK protomers, 50 mM Hepes, pH 6.8 or 8.2, 1.0 mM EDTA, and 2.5 mM [³H]iodoacetic acid (28.6 Ci/mol, American Radiochemicals). Concentrations of Ca²⁺ and CaM identical to those used in the reductive methylation were also used, where indicated, in the alkylation reaction.

Methylation and carboxymethylation of the kinase subunits were quenched by dilution of an aliquot of the assay mixture into an equivalent volume of SDS buffer (0.125 M Tris, pH 6.8, 20% glycerol, 5% -mercaptoethanol, 4% SDS), followed by brief mixing. After heating at 80 °C for 10 min, the samples were run on SDS-polyacrylamide gradient (4-20%) gels (35) and stained with Coomassie Blue. All gels were destained in 40% methanol, 10% acetic acid (2 h) and 7% acetic acid, 4% methanol (15 h). The integrated optical density of the protein bands was determined on a BioImage whole band analyzer. Each band was then excised, solubilized, and decolorized by heating in 250 µl of 30% H₂O₂ for 2 h at 80 °C. The samples and blanks, which contained equivalent amounts of polyacrylamide and H₂O₂, were diluted with 7 ml of Ecoscint scintillation mixture (ICN), and the ³H content was determined.

Standard conditions for cross-linking were developed by optimizing time and cross-linker concentration to allow formation of sufficient amounts of intermediate sized complexes of interest for convenient quantification, while the amounts formed were still capable of increasing or decreasing in a linear manner in response to effectors. With o-phenylenedimaleimide (o-PDM, 17.3 µM), m-maleimidobenzoyl-N-hydroxysuccinimide ester (MBS, 8.6 µM), and 1,1-(methylenedi-4,1-phenylene)bismaleimide (mdPDM, 17.3 µM), the cross-linking was carried out at 30 °C for 2 min with the indicated final concentrations of cross-linkers. In the case of the photocross-linker N-5-azido-2-nitrobenzoyloxysuccinimide (ANB·NOS, 86.0 µM), it was first incubated with PbK for 30 min in the dark, and cross-linking was then initiated by irradiation with UV light (360 nm) for 1 min at 4 °C. Besides cross-linkers, the final concentrations of the other components in the reactions were: 1.73 µM PbK protomer, 50 mM Hepes, pH 8.2, and 1.0 mM EDTA. These same conditions were used to test the effects of Ca²⁺ (1.25 mM CaCl₂, i.e. 250 µM in excess of chelator), Ca²⁺/CaM (1.25 mM/1.73 µM), and Ca²⁺/BtCaM (1.25 mM/1.73 µM), except for the experiment shown in Fig. 2, where CaM was used at the indicated concentrations. Cross-linking was quenched with SDS buffer, and the subunits were resolved by SDS-PAGE as described above. As was observed previously with PDM (31), cross-linking of PbK by MBS and ANB·NOS was intramolecular, as judged by coelution of the indicated conjugates with native enzyme on Sepharose 6B (data not shown); mdPDM was not evaluated in this aspect.

Effect of /Ca²⁺ on the cross-linking of PbK by reagents of variable spans and chemistries.

Subunit composition of cross-linked species was analyzed by Western blotting using subunit-specific mAbs as described previously (29, 31). The extent of cross-linking for individual bands was determined using transmissive and reflective densitometry to measure bands stained with Coomassie Blue in SDS-PAGE and alkaline phosphatase in Western blots, respectively. The electroblotting conditions used to achieve quantitative transfer of the various conjugates onto nitrocellulose resulted in a relatively low recovery of monomeric CaM, presumably because of its low molecular weight. For the detection of BtCaM, streptavidin-alkaline phosphatase (Southern Biotechnology) was exposed to blots and assayed with an alkaline phosphatase kit from Bio-Rad, following the supplier's suggested protocol. Apparent molecular masses of cross-linked species were determined from comparison with the migration of commercial protein standards (29-250 kDa) on 4-20% linear gradient PAGE (31, 34). Also, the migration of known dimer, prepared by cross-linking PbK with transglutaminase (29), was used in identifying dimers formed by the chemical cross-linkers and as the maximum molecular mass standard (no attempt was made to estimate the masses of oligomers with apparent molecular masses greater than that of the dimer).

The assays at pH 6.8 for the phosphorylase conversion activity of PbK, with and without cross-linking, were performed exactly as described in the previous report (31).

RESULTS

Ca²⁺ and CaM/Ca²⁺ Alter the Conformation of the Catalytic  Subunit

To screen for perturbation of the catalytic  subunit induced by the binding of Ca²⁺ to endogenous CaM () or by the binding of exogenous CaM/Ca²⁺ () to the ()₄ holoenzyme, PbK was incubated with radioactive, general chemical modifiers as conformational probes either alone (control), with Ca²⁺, or with /Ca²⁺ (equimolar to protomers), and the incorporation of label into the  subunit was followed. Carboxymethylation by iodoacetate ([³H]ICH₂CO₂), which is selective for thiols, and reductive methylation by formaldehyde ([³H]CH₂O), which is selective for amines, were used for the modifications, which were performed at both pH 6.8, where the nonactivated enyzme has little activity, and pH 8.2, where it is nearly fully active. At either pH, there was a linear carboxymethylation of the  subunit for 20 min (data not shown), which was enhanced by Ca²⁺ and /Ca²⁺, respectively, by 2.1 × and 2.8 × at pH 6.8 and by 1.8 × and 2.4 × at pH 8.2 (Fig. 1A). Similarly, under conditions where reductive methylation of the  subunit increased linearly with time, Ca²⁺ enhanced its modification by 1.5 × at pH 6.8 and by 2.0 × at pH 8.2; however, for this particular conformational probe, /Ca²⁺ had little effect over that of Ca²⁺ alone at either pH (Fig. 1B). These data suggest that regardless of the activity state of the enzyme as defined by pH, Ca²⁺ increases the accessibility of at least one thiol group and multiple amine groups on the  subunit (Fig. 1), which is consistent with the fact that catalytic activity is Ca²⁺-dependent at both pH values.

Stimulation by Ca²⁺ and /Ca²⁺ of the chemical modification of the  subunit.

Influence of /Ca²⁺ on the Cross-linking of PbK by Diverse Cross-linkers

Chemical cross-linkers with different chemistries and spans were evaluated for their ability to detect changes in subunit interactions (in addition to those described above) that were induced by Ca²⁺ and /Ca²⁺. Because the effects of /Ca²⁺ seemed to be overlaid on those of Ca²⁺ alone (Fig. 1), our initial screening was with /Ca²⁺, where the excess Ca²⁺ would simultaneously saturate the  subunit. Cross-linking of control enzyme by o-PDM, MBS, ANB·NOS, and mdPDM (Fig. 2) resulted in the formation of low and intermediate molecular mass cross-linked species of known subunit composition, of dimers, and of high molecular mass oligomers containing all four subunits, but in indeterminate amounts (denoted as a and b in Figs. 2, 3 and 6). As before (29, 31), the subunit composition and stoichiometry of cross-linked conjugates were determined by their masses and cross-reactivities against subunit-specific mAbs. The /Ca²⁺ promoted significant changes in the rates of subunit cross-linking of PbK by all of the cross-linkers. With o-PDM and mdPDM, the cross-linking of , , and  was increased (i.e. their rates of disappearance increased), with this effect more pronounced at higher concentrations of and with the longer cross-linker (Fig. 2, B and E). In contrast, with ANB·NOS or MBS, /Ca²⁺ protected the  subunit from cross-linking, with higher concentrations of being more effective (Fig. 2, C and D).

Effect of Ca²⁺ and /Ca²⁺ on the cross-linking of PbK by o-PDM.

Effect of Ca²⁺ and /Ca²⁺ on the cross-linking of autophosphorylated phosphorylase kinase.

In addition to altering the rates of cross-linking, /Ca²⁺ also changed the patterns of subunit cross-linking, with new cross-linked species formed that contained CaM: with o-PDM and CaM with MBS (Fig. 2, B and C, and Figs. 3 and 4). In all such complexes, the CaM could, of course, represent either  or ; in many cases, as is described in a later section, we were able to distinguish between the two alternatives by utilizing derivatized CaM (BtCaM) as . As a result, whenever possible throughout this report, conjugates are specifically denoted as containing either  or (e.g. above); if  could be present with or if the two are alternatively present in a conjugate (depending on cross-linking conditions), then the conjugate is simply denoted as containing CaM (e.g. CaM above). In figures, for the labeling of a given band that might contain  under one condition (one gel lane) but not another (a different gel lane), the "CaM" nomenclature is also used. The complication regarding the presence in conjugates of  versus is relevant, of course, only for cross-linking carried out in the presence of /Ca²⁺, not Ca²⁺ alone, where only the  subunit is present. Because cross-linking by o-PDM and MBS demonstrated the most significant changes in response to /Ca²⁺, we focused primarily on these cross-linkers as probes for targeting interactions in the holoenzyme mediated by  and .

Effect of Ca²⁺ and /Ca²⁺ on the cross-linking of PbK by MBS.

CaM

To establish reference data against which to determine the effects of Ca²⁺ and /Ca²⁺ on subunit interactions, we characterized as fully as possible the cross-linking of nonactivated, control PbK by o-PDM and MBS. For o-PDM (4.8 Å cross-linking span), the conditions for cross-linking were as described previously (31), and correspondingly, the cross-linked species formed by a 10-fold molar excess of the cross-linker over protomers were as before, namely predominant doublets of dimers and trimers, plus small amounts of an trimer and a doublet with the mass of an dimer, but which cross-reacted only with anti- mAb (Fig. 3, lane 2). As was discussed previously (31, 36), the presence of doublets most likely results from intramolecular cross-linking within the large  and  subunits. In addition to the cross-linked complexes, the degradation product of  (_frag) that commonly occurs in small amounts in purified preparations of the enzyme (14) also showed the ability to cross-react with more than one mAb; its predominant interaction was, as expected, with the anti- mAb, but it also showed highly variable cross-reactivity with the anti-CaM mAb (Fig. 3, lane 2). Because of this cross-reactivity, any species that, based on mass and cross-reactivity with the different mAbs, could have possibly contained the _frag was eliminated from further analysis. The variability in the cross-reactivity may be related to epitope presentation in the blotting process itself because in all cases there was no cross-reactivity by the anti-CaM mAb with the intact  subunit, only with the _frag.

When nonactivated, control enzyme was cross-linked with MBS (9.9 Å cross-linking span), the majority of the cross-linked complexes contained the  subunit (Fig. 4, lane 2). The predominant species formed were a trimer (5.0% error) and an dimer (2.3% error) and in smaller amounts, a dimer (mass_theor = 250 kDa; 3.0% error) and an only partially classified conjugate termed X₂₂₅ that contained  and  and migrated with a mass of 225 kDa, slightly slower than . The mass and cross-reactivity of this last complex do not correspond to any straightforward combination of subunits. Several other complexes containing the  subunit that were present in only small amounts had masses that most closely corresponded to a dimer (mass_theor = 142 kDa) but did not cross-react with the anti-CaM mAb. One of these, designated *1 (Fig. 4, lane 2; 9.8% error), migrated slightly below the  subunit; another, *2 (Fig. 4, lane 2, designated CaM*2 in the figure; 5.6% error), migrated just above the  subunit. The faster migrating *1 may represent intramolecular cross-linking of the  subunit, in that such cross-linking has been shown previously to cause more rapid migration of this subunit (31, 36). Even though these putative complexes did not cross-react with the anti-CaM mAb and even though a relatively large number of theoretical permutations of subunit cross-linking are possible, there are no other combinations of subunits that are consistent with the observed masses and cross-reactivities of these complexes. Apparently the epitope recognized by the anti-CaM mAb, which is at the COOH terminus of CaM (37), is masked in the blots of these particular complexes containing the  subunit, the smallest of the four PbK subunits.

Inclusion of Ca²⁺ caused two significant changes in the cross-linking of PbK by o-PDM (Fig. 3, lane 3): a large increase in the formation of trimer (2.2% error), which was formed in only trace amounts in the absence of Ca²⁺; and formation of a new conjugate, tentatively identified as an tetramer (mass_theor = 217 kDa; 3.2% error). Although this latter complex cross-reacted with only the anti- and anti- mAbs, it comigrated exactly with the better defined complex formed in the presence of /Ca²⁺. These structural effects of Ca²⁺, as well as those of /Ca²⁺, are summarized in Table I. Stimulation by Ca²⁺ of the formation of the conjugate, a marker for conformations induced by other activators of PbK (31), was also observed when using m- or p-PDM as the cross-linker (data not shown); however, the effect was more clearly observed with the o-isomer, as it forms the least amount of this trimer in Ca²⁺-free controls (31). Other changes induced by Ca²⁺ in the cross-linking by o-PDM were increased formation of conjugates heavier than the dimer, plus a modest increase in formation of the dimer (Fig. 3, lane 3). The results with o-PDM indicate that the binding of Ca²⁺ by the  subunit of PbK promotes a conformational change in the holoenzyme that, as a minimum, perturbs interactions among the , , and  subunits. To ask whether Ca²⁺ also perturbs interactions involving the  subunit, cross-linking was performed with MBS, which preferentially forms conjugates containing this subunit.

Table I. Structural perturbations of nonactivated and autophosphorylated PbK by Ca2+ and /Ca2+ The relative structural changes induced by Ca2+ and /Ca2+ in a given conjugate or in the  subunit are compared with controls for the two forms of the holoenzyme. The comparison control condition for the effects of Ca2+ is the absence of any added effector, whereas the reference control condition for the effects of /Ca2+ is the presence of Ca2+ alone. The following symbols denote the relative changes observed under the test condition compared with its control condition: , an increase in amount, with more arrows denoting a greater increase; New, formation of conjugate not observed under control condition; NX, no change in amount; , different regions of the  subunit cross-linked; , conjugate not observed; , a decrease in amount, with more arrows denoting a further decrease; , exogenous CaM a component of the conjugate; and ND, not determined. All conclusions regarding the relative amounts of each conjugate were based on optical densitometry, as described under "Experimental Procedures." Probe and protein targets Nonactivated PbK Autophosphorylated PbK Ca2+  /Ca2+ Ca2+  /Ca2+ o-PDM        NX      New   New              MBS   *1 NX   NX     CaM*2           CaM*3   New   New   X225            NX   NX          New   IC3H2COOH    modified         C3H2O    modified   NX ND ND

Including Ca²⁺ in the cross-linking with MBS caused significant changes mostly in the anti-CaM blot of the cross-linked products (Fig. 4, lane 3), although two of the observed changes are not readily interpretable. One of these two is a new cross-reactive band that migrates in the area of the _frag; but, as was discussed previously, species that could possibly contain this fragment were eliminated from further analysis because of its anomalous cross-reactivity. The second concerns the previously described X₂₂₅ conjugate containing at least the  and  subunits, which shows enhanced anti-CaM cross-reactivity when the cross-linking includes Ca²⁺. The interpretable changes in the cross-linked products containing  involve the  and  subunits. The heaviest of these is an trimer (mass_theor = 172 kDa; 2.9% error), whose formation is slightly increased by Ca²⁺; for the related dimer (mass_theor = 155 kDa; 5.8% error), Ca²⁺ caused an increase in cross-reactivity. The remaining change was in *2, which now cross-reacted with the anti-CaM mAb, unlike the situation when cross-linking was carried out in the absence of Ca²⁺. These last results suggest that Ca²⁺ promotes a conformational change that results in the cross-linking of different regions of  to the  subunit, no longer masking the epitope on  for the anti-CaM mAb.

Influence of /Ca²⁺ on Cross-linking

Compared with the results observed with Ca²⁺ alone, including /Ca²⁺ in the cross-linking of PbK by MBS caused several changes (Fig. 4, lane 4): new bands were formed, and there were changes in the cross-reactivities of other bands, especially in the anti-CaM blot. There was an increase in the cross-reactivities of the bands corresponding to CaM, (CaM)₂, *1, and CaM*2, as well as of X₂₂₅, and a decrease in the cross-reactivity of the band. New bands formed in the presence of /Ca²⁺ included several that migrated between the  subunit and the _frag, which as discussed previously, were not considered further, plus a new CaM dimer, *3 (1.4% error). The three differently migrating CaM complexes presumably result from different amounts or regions of inter- and intramolecular cross-linking, particularly of the  subunit in the latter case.

To determine if the CaM in the CaM-containing complexes represented  or (i.e. endogenous or exogenous CaM), we used as the source for a tagged (monobiotinylated) CaM derivative, BtCaM (33), which activates the PbK holoenzyme in parallel with bovine brain CaM (data not shown). Even though the exchange rate of the endogenous  subunit for exogenous CaM has been shown to be barely detectable, especially in the presence of Ca²⁺ (12), we nevertheless incubated the BtCaM/Ca²⁺ with equimolar PbK for only 2 min prior to cross-linking to eliminate further any possibility of exchange. Cross-linking in the presence of BtCaM/Ca²⁺ resulted in the same cross-linking pattern and cross-reactivities against the anti-CaM mAb as were observed with nonderivatized CaM as (Fig. 4, lane 5). Avidin-alkaline phosphatase cross-reacted with CaM*3, suggesting that this band may be entirely , especially given that it is not observed in the absence of exogenous CaM. Some bands that were initially observed to form in the absence of exogenous CaM, such as *2, cross-reacted with the avidin probe when formed in the presence of BtCaM/Ca²⁺; this indicates either that /Ca²⁺ flips the cross-linking of  from  to or that these bands contain both and . In contrast, the *1 band, which is formed both in the presence and absence of BtCaM/Ca²⁺, did not appear to cross-react with avidin, suggesting that it is composed entirely of . It is noteworthy that although /Ca²⁺ causes only a small increase over Ca²⁺ alone in the amount of *1 observed with anti-, it causes a dramatic increase in the amount of *1 that cross-reacts with anti-CaM. These results suggest that in the *1 dimer different regions of  are cross-linked in the presence of /Ca²⁺ than in the presence of Ca²⁺ alone, i.e. that exogenous CaM affects the endogenous CaM of PbK.

In contrast to its effects on the cross-linking by MBS, /Ca²⁺ (compared with Ca²⁺ alone) caused relatively small changes in the cross-linking by o-PDM (Fig. 3, lane 4). The two most significant changes were a large increase in the amount of formed and in the number of cross-reactive bands observed in the anti-CaM blot. In the latter case, however, the mass and cross-reactivity of the bands other than and did not allow their unambiguous identification. As with MBS, use of BtCaM as did not cause additional changes; however, the and formed in its presence did not cross-react with the avidin probe (data not shown), suggesting that these species did not contain , only . Consequently, increases the formation of by o-PDM and apparently alters the cross-linked regions in both and , allowing greater cross-reactivity with the anti-CaM mAb.

Activation of PbK by Cross-linking in the Presence of Ca²⁺ and /Ca²⁺

In the previous report (31) we demonstrated that when PbK is cross-linked by PDM in the presence of the allosteric activators ADP and GDP it remains activated even after dilution of the allosteric effectors to ineffective concentrations, i.e. the active conformers were trapped by cross-linking. In addition, formation of increased along with activation. Because both Ca²⁺ and /Ca²⁺ are activators that also caused increased formation of by o-PDM (Fig. 3), we asked whether cross-linking in the presence of these effectors could similarly lock the enzyme in an active conformation. When enzyme was cross-linked at pH 8.2 in the presence of Ca²⁺ or /Ca²⁺ and then assayed at pH 6.8, there was a 4.2- and 5.9-fold increase, respectively, in its activity, following dilution of the effectors (Fig. 5, closed bars). In contrast, the non-cross-linked control enzyme did not show significant activation when assayed with identical carryover concentrations of Ca²⁺ and /Ca²⁺ (Fig. 5, open bars). Therefore, the irreversible activation in response to cross-linking results from the direct action of o-PDM on the Ca²⁺ and /Ca²⁺ complexes of PbK. Cross-linking with MBS, which does not form complexes, did not activate PbK or its Ca²⁺ or /Ca²⁺ complexes (data not shown). These results, now obtained with different activators, corroborate the previous finding that formation of the trimer by PDM is a marker for active conformers of PbK (31).

Influence of Ca²⁺ and /Ca²⁺ on the activity of enzyme cross-linked by o-PDM.

Influence of Ca²⁺ and /Ca²⁺ on the Conformation of Autophosphorylated Enzyme

The maximal activation of PbK is achieved through its autophosphorylation (38), but even this activity remains Ca²⁺-dependent, which suggests that Ca²⁺-induced effects on conformation, which were observed with nonphosphorylated, nonactivated PbK, may also occur with autophosphorylated enzyme. We first tested the abilities of Ca²⁺ and /Ca²⁺ to stimulate the carboxymethylation of the  subunit of autophosphorylated enzyme at pH 6.8. Similar to the results obtained with the nonactivated enzyme (Fig. 1A), Ca²⁺ and /Ca²⁺ enhanced the carboxymethylation of the  subunit by 1.6 × and 1.7 ×, respectively; however, autophosphorylation itself caused no increase in carboxymethylation above that observed with the control nonactivated enzyme (data not shown). Using cross-linking by o-PDM as the conformational probe, Ca²⁺ and /Ca²⁺ had the same effects on the cross-linking of both nonactivated and autophosphorylated PbK, namely increased formation of and and induction of (Fig. 6A). Furthermore, as with nonactivated PbK, when the latter two complexes were formed in the presence of BtCaM as , they did not cross-react with avidin (Fig. 6A), suggesting that in phosphorylated and in nonactivated PbK, has similar effects on . With MBS as the cross-linker, the behavior of autophosphorylated and nonactivated PbK was again very similar, but with some minor differences. The cross-linking patterns in the absence of effectors were the same, and Ca²⁺ and /Ca²⁺ had the same influence on the cross-linking of both forms of the enzyme; but, with the phosphorylated enzyme, there was diminished cross-reactivity in the anti-CaM blot for the *2, , and complexes formed in the presence of Ca²⁺ (Fig. 6B). In summary (see Table I), Ca²⁺ significantly alters the structure of the  subunit of autophosphorylated PbK, as indicated by the apparent accessibility difference detected by carboxymethylation, and perturbs the cross-linking of the enzyme. These effects are very similar to those observed with nonactivated PbK and suggest, therefore, that the effects of Ca²⁺ on PbK are essentially independent of its state of activation. The results with /Ca²⁺ indicate that autophosphorylation does not block the binding of exogenous CaM nor its being cross-linked to the regulatory subunits nor its ability to promote differences in the cross-linking of the  subunit (Table I); however, the influence of /Ca²⁺ on the chemical modification of the  subunit was less with the autophosphorylated enzyme.

DISCUSSION

The influence of Ca²⁺ ions on the actions of the monofunctional and bifunctional chemical probes used in this study to monitor PbK conformation indicates extensive communication, either direct or indirect, between the  subunit (endogenous CaM) and the remaining subunits of the holoenzyme, regardless of its state of activation. Because the activity of the holoenzyme expressed by the  subunit is Ca²⁺-dependent and because direct interactions between the  and subunits have been documented (10, 12), we first evaluated the ability of Ca²⁺ to alter the conformation of the catalytic  subunit, as monitored by chemical modification. Both carboxymethylation and reductive methylation of  were increased by Ca²⁺. Because Ca²⁺ also influenced the formation of several cross-linked complexes containing  (Table I), we think that the increased modification of  likely represents increased accessibility of this subunit, rather than simply increased reactivity of particular side chains. An increased accessibility of  in response to Ca²⁺ is also consistent with previous reports that Ca²⁺ enhances the binding to the holoenzyme of antipeptide antibodies against carboxyl-terminal regions of the  subunit (16) and increases the affinity of PbK for its macromolecular substrate, phosphorylase b (39). With the bifunctional probes, Ca²⁺ influenced the formation of cross-linked species containing the inhibitory  and  subunits (Table I). With respect to the subunit, Ca²⁺ promoted increased formation of and the new conjugate (both by o-PDM) and a second new conjugate, (by MBS). This influence by Ca²⁺ on the interactions among the , , and  subunits suggests an activation mechanism for the holoenzyme that includes a linkage in which the binding of Ca²⁺ to  perturbs constraining quaternary interactions imposed by  upon ; support for such a linkage comes from the work of Chan and Graves (9), who reported that the activity of the complex showed a less stringent requirement for Ca²⁺ than did the complex. It should be noted, however, that Ca²⁺ also affects the  subunit, as indicated by the results with MBS (Table I) or by the use of partial proteolysis as a probe of conformation (40).

Although Ca²⁺ displays the property common to other activators of PbK of promoting formation by PDM of the trimer, it also has effects that are distinct from those of the other activators. For instance, neither ADP nor autophosphorylation caused formation of the tetramer by o-PDM (31) or stimulated carboxymethylation of the  subunit (41) or enhanced the affinity of PbK for phosphorylase b (39). Not only were the structural effects of Ca²⁺ observed in this study distinct, they were nearly the same for both nonactivated and autophosphorylated PbK: similar stimulation of carboxymethylation of the  subunit and similar perturbation of cross-linking by either o-PDM or MBS (Table I). That the activator Ca²⁺ has unique structural effects that occur with both nonactivated and activated PbK is totally consistent with the fact that all other activators of the enzyme stimulate its Ca²⁺-dependent activity but do not eliminate the requirement for Ca²⁺. One might envision a hierarchy of tiered conformational transitions leading to the activation of PbK, with the most fundamental being that induced by the binding of Ca²⁺ to the  subunit. Conformational transitions triggered by additional activating events targeted to the remaining subunits (phosphorylation and proteolysis to  and  (1); binding of to / (12, 20, 22); binding of ADP, probably to  (42, 43); and binding of excess Mg²⁺, at l