Enhancement by Mg2+ of domain specificity in Ca2+-dependent interactions of calmodulin with target sequences

Mg2+ binds to calmodulin without inducing the changes in secondary structure that are characteristic of Ca2+ binding, or the exposure of hydrophobic surfaces that are involved in typical Ca2+-dependent target interactions. The binding of Mg2+ does, however, produce significant spectroscopic changes in residues located in the Ca2+-binding loops, and the Mg-calmodulin complex is significantly different from apo-calmodulin in loop conformation. Direct measurement of Mg2+ binding constants, and the effects of Mg2+ on Ca2+ binding to calmodulin, are consistent with specific binding of Mg2+, in competition with Ca2+. Mg2+ increases the thermodynamic stability of calmodulin, and we conclude that under resting, nonstimulated conditions, cellular Mg2+ has a direct role in conferring stability on both domains of apo-calmodulin. Apo-calmodulin binds typical target sequences from skeletal muscle myosin light chain kinase and neuromodulin with Kd approximately 70-90 nM (at low ionic strength). These affinities are virtually unchanged by 5 mM Mg2+, in marked contrast to the strong enhancement of peptide affinity induced by Ca2+. Under conditions of stimulation and increased [Ca2+], Mg2+ has a role in directing the mode of initial target binding preferentially to the C-domain of calmodulin, due to the opposite relative affinities for binding of Ca2+ and Mg2+ to the two domains. Mg2+ thus amplifies the intrinsic differences of the domains, in a target specific manner. It also contributes to setting the Ca2+ threshold for enzyme activation and increases the importance of a partially Ca2+-saturated calmodulin-target complex that can act as a regulatory kinetic and equilibrium intermediate in Ca2+-dependent target interactions.     

Effects of trifluoperazine on calcium binding by calmodulin. Heat capacity and entropy changes

The possible structural changes of the calmodulin-trifluoperazine (TFP) complex caused by Ca2+ binding have been analyzed by microcalorimetric titrations. Titrations of calmodulin with Ca2+ in the presence of 8-fold molar excess TFP have been made both in the absence and presence of Mg2+, at pH 7.0, and at 5, 15, and 25 degrees C. At high concentrations of TFP calmodulin forms a complex with TFP even in the absence of Ca2+. The reaction of the calmodulin-TFP complex with Ca2+ is exothermic, both in the presence and absence of Mg2+. In the presence of Mg2+ the reaction is driven almost entirely by a favorable enthalpy change. The magnitudes of the hydrophobic and internal vibrational contributions to the heat capacity and entropy changes of this complex on Ca2+ binding have been estimated by the empirical method of Sturtevant (Sturtevant, J. M. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 2236-2240). In the presence of Mg2+, the vibrational as well as hydrophobic entropy is slightly increased in a parallel manner by Ca2+ binding to each of the binding sites. In contrast, when Mg2+ is absent, the hydrophobic entropy gradually increases on Ca2+ binding, but the vibrational entropy decreases. These changes of entropy indicate the assembling of non-polar groups on the surface of the complex and suggest that the overall structure is loosened in the presence of Mg2+, but tightened in the absence of Mg2+.     

Changes in the structure of calmodulin induced by a peptide based on the calmodulin-binding domain of myosin light chain kinase

Small-angle X-ray and neutron scattering data were used to study the solution structure of calmodulin complexed with a synthetic peptide corresponding to residues 577-603 of rabbit skeletal muscle myosin light chain kinase. The X-ray data indicate that, in the presence of Ca2+, the calmodulin-peptide complex has a structure that is considerably more compact than uncomplexed calmodulin. The radius of gyration, Rg, for the complex is approximately 20% smaller than that of uncomplexed Ca2+.calmodulin (16 vs 21 A), and the maximum dimension, dmax, for the complex is also about 20% smaller (49 vs 67 A). The peptide-induced conformational rearrangement of calmodulin is [Ca2+] dependent. The length distribution function for the complex is more symmetric than that for uncomplexed Ca2+.calmodulin, indicating that more of the mass is distributed toward the center of mass for the complex, compared with the dumbell-shaped Ca2+.calmodulin. The solvent contrast dependence of Rg for neutron scattering indicates that the peptide is located more toward the center of the complex, while the calmodulin is located more peripherally, and that the centers of mass of the calmodulin and the peptide are not coincident. The scattering data support the hypothesis that the interconnecting helix region observed in the crystal structure for calmodulin is quite flexible in solution, allowing the two lobes of calmodulin to form close contacts on binding the peptide. This flexibility of the central helix may play a critical role in activating target enzymes such as myosin light chain kinase.     

Fluorescence investigations of calmodulin hydrophobic sites

Calmodulin activation of target enzymes depends on the interaction between calmodulin hydrophobic regions and some enzyme areas. The Ca2+ induced exposure of calmodulin hydrophobic sites was studied by means of 2-p-toluidinylnaphthalene-6-sulfonate, a fluorescent probe. Scatchard and Job plots showed that the calmodulin-Ca42+ complex bound two molecules of this hydrophobic probe, with KD congruent to 1.4 X 10(-4) M. These sites are not totally exposed until calmodulin has bound four Ca2+ per molecule, so the conformational change is not over before the four specific Ca2+ - binding sites are saturated with Ca2+.     

Structural characterization of the interactions between calmodulin and skeletal muscle myosin light chain kinase: effect of peptide (576-594)G binding on the Ca2+-binding domains

Calcium-containing calmodulin (CaM) and its complex with a peptide corresponding to the calmodulin-binding domain of skeletal muscle myosin light chain kinase [skMLCK(576-594)G] have been studied by one- and two-dimensional 1H NMR techniques. Resonances arising from the antiparallel beta-sheet structures associated with the calcium-binding domains of CaM and their counterparts in the CaM-skMLCK(576-594)G complex have been assigned. The assignments were initiated by application of the main chain directed assignment strategy. It is found that, despite significant changes in chemical shifts of resonances arising from amino acid residues in this region upon binding of the peptide, the beta-sheets have virtually the same structure in the complex as in CaM. Hydrogen exchange rates of amide NH within the beta-sheet structures are significantly slowed upon binding of peptide. These data, in conjunction with the observed nuclear Overhauser effect (NOE) patterns and relative intensities and the downfield shifts of associated amide and alpha resonances upon binding of peptide, show that the peptide stabilizes the Ca2+-bound state of calmodulin. The observed pattern of NOEs within the beta-sheets and their structural similarity correspond closely to those predicted by the crystal structure. These findings imply that the apparent inconsistency of the crystal structure with recently reported low-angle X-ray scattering profiles of CaM may lie within the putative central helix bridging the globular domains.     

Binding of both Ca2+ and mastoparan to calmodulin induces a large change in the tertiary structure

The technique of small-angle X-ray scattering has been employed to examine the solution conformation of calmodulin and its complexes with Ca2+ alone, and with both Ca2+ and mastoparan. The radius of gyration decreased by 3.1 +/- 0.3 A upon binding of both 4 mol Ca2+/mol of protein and 1 mol mastoparan/mol of protein to form the ternary complex. A smaller increase was found for the separate binding of 4 mol Ca2+/mol of protein in the absence of mastoparan (0.6 +/- 0.3 A). The analyses of pair distance distribution function showed that the maximal pair distance in calmodulin complex with both Ca2+ and mastoparan decreased by 20-30% in comparison with calmodulin or its complex with Ca2+, and a shoulder near 40 A, which characterizes the dumbbell-shaped molecule of calmodulin, disappeared. These results indicate that the two globular domains of the calmodulin complex with Ca2+ and mastoparan come close together by 8.0-9.5 A on average, if the size and the overall shape of the globular domains are the same in Ca2+-calmodulin-mastoparan complex as in calmodulin or Ca2+-calmodulin complex.     

Glycation of calmodulin: chemistry and structural and functional consequences

In the presence of Ca2+ and glucose, calmodulin incorporates 2.5 mol of glucose/mol of protein. In the absence of Ca2+, only 1.5 mol of glucose is incorporated per mole of calmodulin. Glycation of calmodulin is associated with variable reductions in its capacity to activate three Ca2+/calmodulin-dependent brain target enzyme systems, including adenylyl cyclase, phosphodiesterase, and protein kinase. In addition, glycated calmodulin exhibits a 54% reduction in its Ca2+ binding capacity. Isolated CNBr cleavage fragments of glycated calmodulin suggest that glycation follows a nonspecific pattern in that each of seven available lysines is susceptible to modification. A limit observed on the extent of glycation appears related to the accompanying increase in negative charge on the protein. Glycation results in minimal structural rearrangements in calmodulin, and the Ca2+-induced increase in alpha-helix content and radius of gyration is the same for glycated and unmodified calmodulin. Since glycated calmodulin's Ca2+ binding capacity is reduced, this implies that the Ca2+-induced conformational changes in calmodulin do not require all four Ca2+ binding sites to be occupied. Examination of the lysine positions in calmodulin suggests that Ca2+ binding to domains II and IV is sufficient to induce these changes. The functional consequences of calmodulin glycation therefore cannot be attributed to inhibition of these conformational changes. An alternative explanation is that the inhibition arises from interference at the target enzyme binding site by bound glucose. While glycation shows minimal structural effects, a large pH dependence is observed for the alpha-helix content of unmodified calmodulin.(ABSTRACT TRUNCATED AT 250 WORDS)     

Complex of calmodulin with a ryanodine receptor target reveals a novel, flexible binding mode

Calmodulin regulates ryanodine receptor-mediated Ca(2+) release through a conserved binding site. The crystal structure of Ca(2+)-calmodulin bound to this conserved site reveals that calmodulin recognizes two hydrophobic anchor residues at a novel "1-17" spacing that brings the calmodulin lobes close together but prevents them from contacting one another. NMR residual dipolar couplings demonstrate that the detailed structure of each lobe is preserved in solution but also show that the lobes experience domain motions within the complex. FRET measurements confirm the close approach of the lobes in binding the 1-17 target and show that calmodulin binds with one lobe to a peptide lacking the second anchor. We suggest that calmodulin regulates the Ca(2+) channel by switching between the contiguous binding mode seen in our crystal structure and a state where one lobe of calmodulin contacts the conserved binding site while the other interacts with a noncontiguous site on the channel.     

Calcium binding domains of calmodulin. Sequence of fill as determined with terbium luminescence

Terbium, a trivalent lanthanide, effectively substituted for Ca2+ in calmodulin as judged by several criteria: intrinsic fluorescence spectra, altered mobilities on polyacrylamide gel electrophoresis, formation of a stable complex with troponin I or calcineurin, and stimulation of phosphodiesterase. Calmodulin harbors four Ca2+ binding domains; domains I and II contain no tyrosine, whereas domains III and IV each have one tyrosine. The binding of Tb3+ to calmodulin was followed by the increase of Tb3+ fluorescence at 545 nm upon binding to calmodulin. This fluorescence was elicited either by exciting Tb3+ directly at 222 nm or by exciting the calmodulin tyrosine at 280 nm with resulting energy transfer from tyrosine to Tb3+. Fluorescence generated by direct excitation measures binding of Tb3+ to any of the Ca2+ binding domains, whereas energy transfer through indirect excitation is effective only when Tb3+ is within 5 A of tyrosine, indicating that Tb3+ necessarily occupies a Ca2+ binding domain that contains tyrosine. A judicious use of the direct and indirect excitation could reveal the sequence of fill of the binding domains. Our results suggest these domains are filled in the following sequence: 1) domain I or II; 2) domains III and IV; and 3) domain II or I that has not been filled initially.     

Construction and molecular dynamics simulation of calmodulin in the extended and in a bent conformation.

Analysis of sequence similarity and comparison of the three-dimensional (3D) structures of troponin C and calmodulin have revealed a sequence in the central helix of calmodulin with a high probability for bending. The three amino acids known to form a bend in the N-terminal portion of troponin C are also found in the central helix of calmodulin. The modelling of a bent calmodulin structure, using the dihedral angles of the three residues in the bend of troponin C as a 3D template, results in a conformation of calmodulin where the N- and C-terminal domains are able to form contacts. Dynamics simulations starting from the X-ray structure of calmodulin and from the modelled bent calmodulin were carried out to compare flexibility and correlated movements of Ca2+ in the binding loops. Both conformations of calmodulin remained stable during the period of simulation. In the simulation of calmodulin in the extended form, the motions of the Ca2+ atoms in the two domains (Ca2+1 and Ca2+2 in one domain, and Ca2+3 and Ca2+4 in the other) are correlated. In the simulation of the bent form, an additional correlation between the Ca atoms in the two different domains is observed. The results are compatible with the occurrence of a bent conformation of calmodulin in the presence of targets, and with increased Ca2+ affinity and cooperativity of the Ca(2+)-binding loops in the calmodulin-peptide complexes.     

Communication between two globular domains of calmodulin in the presence of mastoparan or caldesmon fragment. Ca2+ binding and 1H NMR

Ca2+ binding to calmodulin was measured in the presence of mastoparan or caldesmon fragment. Mastoparan and caldesmon fragment were used as model compounds of enzymes and cytoskeleton proteins, respectively, working as the target of calmodulin. Although the Ca2+ bindings of the two globular domains of calmodulin occur independently in the absence of the target peptide (or proteins), mastoparan and caldesmon fragment increased the affinity of Ca2+ and, at the same time, produced the positive cooperative Ca2+ bindings between the two domains. The result of Ca2+ binding was compared with 1H NMR spectra of calmodulin in the presence of equimolar concentration of mastoparan. It is known that a conformation change of the C-terminal half-region (C-domain) occurs by the Ca2+ binding to C-domain. A further change in conformation of C-domain was demonstrated by the Ca2+ binding to the N-terminal half-region (N-domain) in the presence of mastoparan. It indicates that the two domains of calmodulin get into communication with each other in the associated state with the target, and we concluded that the Ca2+ binding to the N-domain is responsive to the development of calmodulin function.     

The crystal structure of the sorcin calcium binding domain provides a model of Ca2+-dependent processes in the full-length protein

Sorcin is a 21.6 kDa calcium binding protein, expressed in a number of mammalian tissues that belongs to the small, recently identified penta-EF-hand (PEF) family. Like all members of this family, sorcin undergoes a Ca2+-dependent translocation from cytosol to membranes where it binds to target proteins. For sorcin, the targets differ in different tissues, indicating that it takes part in a number of Ca2+-regulated processes. The sorcin monomer is organized in two domains like in all PEF proteins: a flexible, hydrophobic, glycine-rich N-terminal region and a calcium binding C-terminal domain. In vitro, the PEF proteins are dimeric in their Ca2+-free form, but have a marked tendency to precipitate when bound to calcium. Stabilization of the dimeric structure is achieved by pairing of the uneven EF-hand, EF5. Sorcin can also form tetramers at acid pH.The sorcin calcium binding domain (SCBD, residues 33-198) expressed in Escherichia coli was crystallized in the Ca2+-free form. The structure was solved by molecular replacement and was refined to 2.2 A with a crystallographic R-factor of 22.4 %. Interestingly, the asymmetric unit contains two dimers.The structure of the SCBD leads to a model that explains the solution properties and describes the Ca2+-induced conformational changes. Phosphorylation studies show that the N-terminal domain hinders phosphorylation of SCBD, i.e. the rate of phosphorylation increased twofold in the absence of the N-terminal region. In addition, previous fluorescence studies indicated that hydrophobic residues are exposed to solvent upon Ca2+ binding to full-length sorcin. The model accounts for these data by proposing that Ca2+ binding weakens the interactions between the two domains and leads to their reorientation, which exposes hydrophobic regions facilitating the Ca2+-dependent binding to target proteins at or near membranes.     

The synthesis and reaction of a specific affinity label for the hydrophobic drug-binding domains of calmodulin

An affinity-labeling reagent for the two hydrophobic drug-binding domains of calmodulin has been prepared and its reaction with calmodulin characterized. The reagent, 10-(3-propionyloxysuccinimide)-2-(trifluoromethyl)phenothiazine, was shown to be very specific labeling reagent for these domains. Its specificity was demonstrated by the following observations. 1) Previous reports have shown that Ca2+ is required for phenothiazine binding to calmodulin, and here we show that the affinity-labeling reagent reacts with and inactivates calmodulin in the presence of Ca2+, but not in its absence. 2) Inclusion of trifluoperazine, fluphenazine, W-7, or 10-(3-aminopropyl)-2-(trifluoromethyl)phenothiazine in the reaction mixture protected calmodulin from inactivation by the reagent. 3) Inactivation by the reagent yielded calmodulin that was no longer retained on a phenothiazine-Sepharose column under conditions in which unreacted calmodulin was retained. 4) The measured stoichiometry of the reaction in the presence of excess reagent was 2.1 mol of reagent per mol of calmodulin which agrees well with previous reports of two high-affinity phenothiazine-binding sites on calmodulin. 5) The stoichiometry of the reaction was further confirmed by tryptic peptide maps which show two phenothiazine-labeled peptides unique to the fully reacted protein. 6) The spectral properties of the reagent, while attached to calmodulin, change in the presence of Ca2+ in a manner consistent with the known effects of Ca2+ binding by calmodulin on these hydrophobic domains. The specificity of the reagent makes it useful for further characterization of these hydrophobic binding domains on calmodulin.     

Molecular dynamics simulations of calcium-free calmodulin in solution

A 4-ns molecular dynamics simulation of calcium-free calmodulin in solution has been performed, using Ewald summation to treat electrostatic interactions. Our simulation results were mostly consistent with solution experimental studies, including NMR, fluorescence and x-ray scattering. The secondary structures within the N- and C-terminal domains were conserved in the simulation, with trajectory structures similar to the NMR-derived model structure 1CFD. However, the relative orientations of the domains, for which there are no NMR restraints, differed in details between the simulation and the 1CFD model. The most interesting information provided by the simulations is that the dynamics of calcium-free calmodulin in solution is dominated by slow rigid body reorientations of the domains. The interdomain distance fluctuated between 29 and 39 A, and interdomain orientation angle, defined as the pseudo-dihedral formed by the four calcium binding sites, varied between -2 degrees and 108 degrees. Similarly, the domain linker region also exhibited significant fluctuations, with its length varying in the 34-45 A range and its bend angle in the 10-100 degrees range. The simulations are in accord with fluorescence results suggesting that calcium-free calmodulin is more compact and more flexible than the calcium activated form. Surprisingly, quite similar solvent accessibilities of the hydrophobic patches were seen in the calcium-free trajectory described in this work and previously generated calcium-loaded calmodulin simulations. Thus, our simulations suggest a reexamination of the standard model of the structural change of calmodulin upon calcium binding, involving exposure of the hydrophobic patches to solvent.     

Molecular Dynamics Simulations of Calcium-Free Calmodulin in Solution

Abstract

A 4-ns molecular dynamics simulation of calcium-free calmodulin in solution has been performed, using Ewald summation to treat electrostatic interactions. Our simulation results were mostly consistent with solution experimental studies, including NMR, fluorescence and x-ray scattering. The secondary structures within the N- and C-terminal domains were conserved in the simulation, with trajectory structures similar to the NMR-derived model structure 1CFD. However, the relative orientations of the domains, for which there are no NMR restraints, differed in details between the simulation and the 1CFD model. The most interesting information provided by the simulations is that the dynamics of calcium-free calmod- ulin in solution is dominated by slow rigid body reorientations of the domains. The interdomain distance fluctuated between 29 and 39 Å, and interdomain orientation angle, defined as the pseudo-dihedral formed by the four calcium binding sites, varied between ?2° and 108°. Similarly, the domain linker region also exhibited significant fluctuations, with its length varying in the 34–45 Å range and its bend angle in the 10–100° range. The simulations are in accord with fluorescence results suggesting that calcium-free calmodulin is more compact and more flexible than the calcium activated form. Surprisingly, quite similar solvent accessibilities of the hydrophobic patches were seen in the calcium-free trajectory described in this work and previously generated calcium-loaded calmodulin simulations. Thus, our simulations suggest a reexamination of the standard model of the structural change of calmodulin upon calcium binding, involving exposure of the hydrophobic patches to solvent.     

Two trifluoperazine-binding sites on calmodulin predicted from comparative molecular modeling with troponin-C

Among the known regulatory proteins that are conformationally sensitive to the binding of calcium ions, calmodulin and troponin-C have the greatest primary sequence homology. This observation has led to the conclusion that the most accurate predicted molecular model of calmodulin would be based on the X-ray crystallographic coordinates of the highly refined structure of turkey skeletal troponin-C. This paper describes the structure of calmodulin built from such a premise. The resulting molecular model was subjected to conjugate gradient energy minimization to remove unacceptable intramolecular non-bonded contacts. In the analysis of the resulting structure, many features of calmodulin, including the detailed conformation of the Ca2+-binding loops, the amino- and carboxy-terminal hydrophobic patches of the Ca2+-bound form, and the several clusters of acidic residues can be reconciled with much of the previously published solution data. Calmodulin is missing the N-terminal helix characteristic of troponin-C. The deletion of three residues from the central helical linker (denoted D/E in troponin-C) shortens the molecule and changes the orientation of the two domains of calmodulin by 60 degrees relative to those in troponin-C. The molecular model has been used to derive two possible binding sites for the antipsychotic drug trifluoperazine, a potent competitive inhibitor of calmodulin activity.     

The complex structure of calmodulin bound to a calcineurin peptide

The activity of the protein phosphatase calcineurin (CN) is regulated by an autoinhibition mechanism wherein several domains from its catalytic A subunit, including the calmodulin binding domain (CaMBD), block access to its active site. Upon binding of Ca2+ and calmodulin (Ca2+/CaM) to CaMBD, the autoinhibitory domains dissociate from the catalytic groove, thus activating the enzyme. To date, the structure of the CN/CaM/Ca2+ complex has not been determined in its entirety. Previously, we determined the structure of a fusion protein consisting of CaM and a 25-residue peptide taken from the CaMBD, joined by a 5-glycine linker. This structure revealed a novel CaM binding motif. However, the presence of the extraneous glycine linker cast doubt on the authenticity of this structure as an accurate representation of CN/CaM binding in vivo. Thus, here, we have determined the crystal structure of CaM complexed with the 25-residue CaMBD peptide without the glycine linker at a resolution of 2.1 A. The structure is essentially identical to the fusion construction which displays CaM bound to the CaMBD peptide as a dimer with an open, elongated conformation. The N-lobe from one molecule and C-lobe from another encompass and bind the CaMBD peptide. Thus, it validates the existence of this novel CaM binding motif. Our experiments suggest that the dimeric CaM/CaMBD complex exists in solution, which is unambiguously validated using a carefully-designed CaM-sepharose pull-down experiment. We discuss structural features that produce this novel binding motif, including the role of the CaMBD peptide residues Arg-408, Val-409, and Phe-410, which work to provide rigidity to the otherwise flexible central CaM helix joining the N- and C-lobes, ultimately keeping these lobes apart and forcing "head-to-tail" dimerization to attain the requisite N- and C-lobe pairing for CaMBD binding.     

Static and kinetic studies of calmodulin and melittin complex.

Ca2+ binding to calmodulin triggers conformational change of the protein which induces exposure of hydrophobic surfaces. Melittin has been believed to bind to Ca(2+)-bound calmodulin through the exposed hydrophobic surfaces. However, tryptophan fluorescence measurements and gel chromatography experiments with the melittin-calmodulin system revealed that melittin bound to calmodulin at zero salt concentration even in the absence of Ca2+; addition of salt removed melittin from Ca(2+)-free calmodulin. This means not only the hydrophobic interaction but also the electrostatic interaction contributes to the melittin-calmodulin binding. The fluorescence stopped-flow studies of the dissociation reaction of melittin-calmodulin complex revealed that Ca2+ removal from the complex induced a conformational change of calmodulin, resulting in reduction of the hydrophobic interaction between melittin and calmodulin, but the electrostatic interaction kept melittin still bound to calmodulin for a subsecond lag period, after which melittin dissociated from calmodulin. The fluorescence stopped-flow experiments on the dissociation reaction of complex of melittin and tryptic fragment(s) of calmodulin revealed that the lag period of the melittin dissociation reaction was attributable to the interaction between the C-terminal half of calmodulin and the C-terminal region of melittin.     

Solution X-ray scattering reveals a novel structure of calmodulin complexed with a binding domain peptide from the HIV-1 matrix protein p17

The solution structures of complexes between calcium-saturated calmodulin (Ca (2+)/CaM) and a CaM-binding domain of the HIV-1 matrix protein p17 have been determined by small-angle X-ray scattering with use of synchrotron radiation as an intense and stable X-ray source. We used three synthetic peptides of residues 11-28, 26-47, and 11-47 of p17 to demonstrate the diversity of CaM-binding conformation. Ca (2+)/CaM complexed with residues 11-28 of p17 adopts a dumbbell-like structure at a molar ratio of 1:2, suggesting that the two peptides bind each lobe of CaM, respectively. Ca (2+)/CaM complexed with residues 26-47 of p17 at a molar ratio of 1:1 adopts a globular structure similar to the NMR structure of Ca (2+)/CaM bound to M13, which adopted a compact globular structure. In contrast to these complexes, Ca (2+)/CaM binds directly with both CaM-binding sites of residues 11-47 of p17 at a molar ratio of 1:1, which induces a novel structure different from known structures previously reported between Ca (2+)/CaM and peptide. A tertiary structural model of the novel structure was constructed using the biopolymer module of Insight II 2000 on the basis of the scattering data. The two domains of CaM remain essentially unchanged upon complexation. The hinge motions, however, occur in a highly flexible linker of CaM, in which the electrostatic residues 74Arg, 78Asp, and 82Glu interact with N-terminal electrostatic residues of the peptide (residues 12Glu, 15Arg, and 18Lys). The acidic residues in the N-terminal domain of CaM interact with basic residues in a central part of the peptide, thereby enabling the central part to change the conformations, while an acidic residue in the C-terminal domain interacts with two basic residues in the two helical sites of the peptide. The overall structure of the complex adopts an extended structure with the radius of gyration of 20.5 A and the interdomain distance of 34.2 A. Thus, the complex is principally stabilized by electrostatic interactions. The hydrophobic patches of Ca (2+)/CaM are not responsible for the binding with the hydrophobic residues in the peptide, suggesting that CaM plays a role to sequester the myristic acid moiety of p17.