Effects of the binding of myosin light chain kinase on the reactivities of calmodulin lysines

Calmodulin: The switch button of calcium signaling

Calmodulin (CaM), a calcium sensor, decodes the critical calcium-dependent signals and converts them into the driving force to control various important cellular functions, such as ion transport. This small protein has a short central linker to connect two globular lobes and each unit is composed of a pair of homologous domains (HD) which are responsible for calcium binding. The conformation of each HD is sensitive to the levels of the intracellular Ca²⁺ concentrations while the flexible structure of the central domain enables its interactions with hundreds of cellular proteins. Apart from calcium binding, posttranslational modifications (PTMs) also contribute to the modulations of CaM functions by affecting its protein-protein interaction networks and hence drawing out the various downstream signaling cascades. In this mini-review, we first aim to elucidate the structural features of CaM and then overview the recent studies on the engagements of calcium binding and PTMs in Ca²⁺/CaM-mediated conformational alterations and signaling events. The mechanistic understanding of CaM working models is expected to be a key to decipher the precise role of CaM in cardiac physiology and disease pathology.

Protein engineering and NMR studies of calmodulin

The calcium regulatory protein calmodulin (CaM) plays a role as an on-off switch in the activation of many enzymes and proteins. CaM has a dumbbell shaped structure with two folded domains, which are connected by a flexible linker in solution. The calmodulin-binding domains of the target proteins are contained in 20 residue long amino acid sequences, that share no obvious amino acid sequence homology. In this contribution, we discuss the features of CaM, which allow it to be rather promiscuous, and bind effectively to all these distinct domains. In particular, we describe the role of the methionine-rich hydrophobic surfaces of the protein in providing a malleable and sticky surface for binding many hydrophobic peptides. The enzyme activation properties of various Met --> Leu mutants of CaM are discussed. In addition, the role of the flexible linker region that connects the two domains is also analyzed. Finally, we describe various NMR and spectroscopic experiments that aid in determining the CaM-bound structures of synthetic peptides containing various CaM-binding domains. All structures analyzed to date are alpha-helical when bound to CaM, and they interact with CaM only through amino acid sidechains. This form of protein-protein interaction is rather unique, and may contribute to CaM's capacity to bind effectively to such a wide range of distinct partners.

Carbon-13 NMR studies of the lysine side chains of calmodulin and its proteolytic fragments

The pH-titration and dynamic behaviour of the seven lysine side chains in bovine calmodulin were studied by carbon-13 NMR. The amino groups of the calcium saturated protein and its proteolytic fragments TR1C (1-75) and TR2C (78-148) were dimethylated with carbon-13 labeled formaldehyde; this modification did not alter the protein's structure or its ability to activate the enzyme cyclic nucleotide phosphodiesterase. Tentative assignments for 5 out of the 7 dimethyl lysine resonances could be obtained by comparing spectra of the fully and partially modified protein, with those of the proteolytic fragments. The pKa values measured for calcium saturated calmodulin ranged between 9.5 (Lys 75) and 10.2 (Lys 13); two residues (Lys 94 and Lys 13) showed a biphasic titration curve suggesting their possible involvement in ion-pairs. The dynamic behavior of the lysine side chains was deduced from spin lattice relaxation measurements. All side chains were flexible and this was not influenced by the removal of calcium, or the addition of the calmodulin antagonist trifluoperazine. The latter data suggest that the lysine side chains are not directly involved in calmodulin's target binding sites.

Activation of enzymes by calmodulins containing intramolecular cross-links

We have reacted calmodulins containing cysteines substituted at positions 3 and 146 or 5 and 146 with bismaleimidohexane (BMH) to generate intramolecularly cross-linked proteins termed BMHCM or BMHCM1, respectively. Reactions were also performed with N-ethylmaleimide (NEM) in place of BMH to generate corresponding S-ethylsuccinimidylated proteins termed NEMCM or NEMCM1. The abilities of these proteins to activate plant NAD kinase, erythrocyte Ca(2+)-ATPase and bovine brain calcineurin activities were assessed. The BMH- or NEM-reacted proteins activate calcineurin activity as does control calmodulin. Kact values for Ca(2+)-ATPase activation by BMHCM and BMHCM1 are increased 10-fold relative to the control value, with no corresponding change in Vmax values. Activation of this enzyme by NEMCM or NEMCM1 is not different from the control. In NAD kinase activation experiments BMHCM and BMHCM1 are associated with a 10 to 20-fold increase in Kact values and a 60-75% reduction in Vmax values relative to the control. NEMCM1 is not associated with any apparent changes in NAD kinase activation, however, NEMCM is associated with a 10-fold increase in the Kact value. NEM-reacted calmodulin containing a cysteine only at position 3 is not associated with an increased Kact value, implying that this change is due to interactions between S-(ethylsuccinimido)cysteines at positions 3 and 146. In conclusion, cross-linking and associated distortions in the structure of calmodulin appear to have little or no effect on activation of calcineurin enzyme activity. However, bending in the central helix and/or steric restrictions associated with cross-linking increase significantly the Kact value for Ca(2+)-ATPase and NAD kinase activation, and dramatically reduce maximal activation of NAD kinase activity.

A mass spectrometry method for mapping the interface topography of interacting proteins, illustrated by the melittin-calmodulin system

The shielding of lysine groups from acetylation by acetic anhydride has been used to identify the regions of calmodulin in contact with melittin in the 1:1 complex. The estimation of the degree of acetylation was done by examining cyanogen bromide and cyanogen bromide/trypsin digests by mass spectrometry. Evidence was obtained that lysines-21, -75, and -148 are protected to some extent, with the implication that both the N- and C-terminal lobes and the connecting strand are involved in the interaction.

The secondary structure of turkey gizzard myosin light chain kinase and the nature of its interaction with calmodulin

The enzymatic activities of native myosin light chain kinases are subject to modification by interaction with Ca2(+)-calmodulin (CaM). The interaction between myosin light chain kinase isolated from turkey gizzard (tgMLCK) and calmodulin isolated from bovine testes (CaMbt) and wheat germ (CaMwg) has been examined by means of the intrinsic tryptophan fluorescence of tgMLCK and the fluorescence of extrinsic fluorescent labels located at Cys-27 and Tyr-139 of CaMwg and Tyr-99 of CaMbt. Static and dynamic fluorescence measurements provide evidence for the involvement of the former two sites in the zone of contact with lesser involvement of the site marked by the probe at Tyr-99. Complex formation protected the primary cleavage site in CaMbt (Lys-77) from proteolysis by trypsin. These results are consistent with involvement of the N- and C-terminal lobes of CaM in stabilization of the complex with tgMLCK, but cannot rule out participation of the connecting strand in the interaction. CD measurements extending to 175 nm, obtained using synchroton radiation, indicate the following secondary structure content for tgMLCK: 17 +/- 2% alpha-helix, 22 +/- 3% antiparallel beta-sheet, 3 +/- 1% parallel beta-sheet, 24 +/- 2% beta-turns, and 34 +/- 2% random coil. Similar measurements of the CD spectra of CaMbt and of the 1:1::CaMbt:tgMLCK complex presently indicate that neither protein undergoes major secondary structure rearrangement during their interaction, although subtle changes in the CD spectrum of tgMLCK appear to be correlated with the interaction with CaM.

The phosphorylation of calmodulin and calmodulin fragments by kinase fractions from bovine brain

The phosphorylation of intact calmodulin and of fragments obtained by trypsin digestion was studied, using a protein kinase partially purified from bovine brain. Brain extracts were made in the presence of the detergent CHAPS (3-[3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate). The protein kinase catalyzed the incorporation of nearly 1 mol of 32P from [gamma-32P]ATP into calmodulin fragment 1-106. Incorporation was exclusively into serine 101. With fragment 78-148, the extent of phosphorylation was somewhat less and 32P appeared mainly in threonine residues. Fragment 1-90 was also a fairly good substrate, but the phosphorylation of intact calmodulin never exceeded 0.01 mol per mol. Little or no phosphorylation was seen with parvalbumin, the brain Ca2+-binding protein (CBP-18) and intestinal calcium-binding protein. The protein kinase had no requirement for cAMP or phospholipids. High levels of Mg2+ (60-70 mM) stimulated phosphorylation of the fragments 20-fold. Millimolar concentrations of Ca2+ were inhibitory. It is suggested that the calmodulin fragments were in a conformation more favorable for phosphorylation than intact soluble calmodulin.

Characterization of the interaction of ophiobolin A and calmodulin

1. The fungal toxin ophiobolin A reacts with the epsilon-amino group of lysine to give a conjugated enamine produce with lambda max at 272 nm and a molar extinction of 19,200 per M/cm. 2. Bovine brain calmodulin reacts with ophiobolin A to give a lambda max at 272 nm. 3. One mol of calmodulin reacts with two moles of ophiobolin A. Reaction of 1 mol of ophiobolin A inactivates 1 mol of calmodulin. 4. Ophiobolin A-treated calmodulin is resistant to tryptic cleavage at lysine 77. 5. Ophiobolin A also inhibits Dictyostelium calmodulin which has glutamine instead of lysine at residue 77.

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.

Isolation of a calcium-sensitive, 35,000-dalton microfilament- and liposome-binding protein from ascites tumor cell microvilli: identification as monomeric calpactin

Microvilli isolated from the MAT-C1 ascites subline of the 13762 rat mammary adenocarcinoma contain a major calcium-sensitive microfilament-binding protein, AMV-p35 (ascites microvillar p35). Association of AMV-p35 with microfilament cores during Triton X-100 extraction of the microvilli is half-maximal at 0.1-0.2 mM calcium. The protein, which comprises 6% of the total microvillar protein, can be isolated from microfilament cores prepared in the presence of calcium by extraction with EGTA and purification by ion-exchange chromatography. Alternatively, the protein can be isolated from Triton extracts of microvilli prepared in the absence of calcium by precipitation with calcium, solubilization of the precipitate with EGTA, and chromatography on an ion-exchange column. AMV-p35 binds to phosphatidylserine liposomes and F-actin with half-maximal calcium concentrations of about 10 microM and 0.2 mM, respectively. Treatment of AMV-p35 with chymotrypsin yields a 33,000-dalton fragment, behavior similar to the tyrosine kinase substrates calpactins I and II and lipocortins I and II. Immunoblot analyses using antibodies directed against calpactin I, lipocortin I, and lipocortin II showed strong reactivity of AMV-p35 with anti-calpactin I and anti-lipocortin II, but little reactivity toward anti-lipocortin I. The close relationship between AMV-p35 and calpactin I was verified by amino acid sequence analyses of peptides isolated from cyanogen bromide digests of AMV-p35. By gel filtration and velocity sedimentation analyses purified AMV-p35 is a 35,000-dalton monomer. Moreover, AMV-p35 extracted directly from microvilli in Triton/EGTA also behaves as a 35,000-dalton menomer. These findings indicate that AMV-p35 is closely related to the pp60src kinase substrate calpactin I (p36). However, AMV-p35 occurs in the microvilli as a monomer rather than as the heterotetrameric calpactin found in several other cell types.

Sequence homology in the metalloproteins; purple acid phosphatase from beef spleen and uteroferrin from porcine uterus

The primary structures of purple acid phosphatase and uteroferrin, two iron-binding glycoproteins isolated from beef spleen and porcine uterine fluids, respectively, have been examined by a combination of tandem mass spectrometry and classical Edman sequencing methods. Reported here are amino acid sequence data covering more than 90% of the primary structures for these two proteins. The sequence data reveal an unexpectedly high degree of homology, greater than 90%, for these two proteins.

Affinity selection of chemically modified proteins: role of lysyl residues in the binding of calmodulin to calcineurin

In affinity selection, calcineurin selects from a population of randomly modified calmodulins those species with which it prefers to interact. The method shows that acetylation of lysines affects calmodulin so as to interfere with its ability to interact with calcineurin. Monoacetylation of any lysine of calmodulin reduces its affinity for calcineurin by 5-10-fold. Multiple acetylations amplify the loss of affinity; none of the modifications are imcompatible with activity. The lack of selectivity of calcineurin against any particular modified lysine indicates that the loss of affinity reflects changes induced by the removal of the charged groups and suggests an important role for electrostatic interactions in the cooperative structural transitions which calmodulin undergoes upon binding its target proteins or calcium. In the presence of calcineurin, a large and specific decrease in the rate of acetylation of Lys-75 and -148 of calmodulin is observed. The reactivity of the same residues is greatly increased in the presence of calcium alone [Giedroc, D. P., Sinha, S. K., Brew, K., & Puett, D. (1985) J. Biol. Chem. 260, 13406-13413]. Lys-75, located in the central helix, and the C-terminal Lys-148 [Babu, Y. S., Sacks, J. S., Greenhouse, T. J., Bugg, C. E., Means, A. R., & Cook, W. J. (1985) Nature (London) 315, 37-40] may act as sensors of the calmodulin allosteric transitions. Their reactivity changes in opposite directions in response to calcium-induced or calcineurin-induced structural changes. The reactivity of other residues such as Lys-21, decreased in the presence of calcineurin but not calcium, is also affected by a conformational change which is induced specifically by calcineurin.

Association of calmodulin and smooth muscle myosin light chain kinase: application of a label selection technique with trace acetylated calmodulin

A method is described for rapidly surveying the effects of modifying individual amino acid residues of a protein on its ability to interact specifically with another macromolecule. The procedure has been used to examine the individual roles of the seven lysyl residues of calmodulin in its ability to bind to smooth muscle myosin light chain kinase; previous studies by Jackson et al. (J. Biol. Chem. 261:1226-12232, 1986) have suggested that certain lysines may be located close to the interaction site. Trace [3H]-acetylated calmodulin, consisting predominantly of molecules acetylated at single sites together with unmodified protein, was incubated in excess (five- to 20-fold) with smooth muscle MLC kinase to allow the modified and unmodified molecules to compete for binding to the enzyme. Subsequently, the calmodulin-enzyme complex was separated from unbound calmodulin, and the level of acetylation of each of the seven lysines of the bound fraction of calmodulin was determined and compared to that of each corresponding group of the starting preparation. Significant changes were found at only two of the lysines, 21 and 75, where the extent of acetylation in the bound fraction was three- and fivefold lower, respectively, than that in the original preparation. These results were reproducible in three separate selection experiments employing both chicken and turkey gizzard MLC kinase. It is concluded that acetylation of calmodulin at either lysine 21 or 75 markedly reduces its affinity for MLC kinase, but acetylation at any of the other lysines (13, 30, 77, 94, or 148) has only minor effects.(ABSTRACT TRUNCATED AT 250 WORDS)

Calcium effects on calmodulin lysine reactivities

The differential reactivities of individual lysines on porcine testicular calmodulin were determined by trace labeling with high specific activity [3H]acetic anhydride as a function of the molar ratio of Ca2+ to calmodulin. In progressing from the Ca2+-depleted form of the protein to a Ca2+:calmodulin molar ratio of 5:1, six of the seven lysyl residues exhibited a modest 1.5- to 3.0-fold increase in reactivity. Lys 75, in contrast, was enhanced in reactivity greater than 20-fold. When the change in reactivity of each lysine was normalized as a percentage of the maximum change, most of the residues were found to fall into two distinct classes. One class, comprising lysines 94 and 148 from the two carboxy terminal Ca2+-binding domains 3 and 4, respectively, exhibited about 90% of their reactivity change when the Ca2+:calmodulin molar ratio was 2:1, and these residues were perturbed very little upon further addition of Ca2+. The other class, encompassing lysines 13, 21, and 30 from the amino terminal domain 1 and Lys 75 from the extended helix connecting the two globular lobes of calmodulin, underwent most of their overall reactivity change (55-70%) between 2 and 5 equivalents of Ca2+ per mol of calmodulin. Lys 77 was distinct in its pattern of change, undergoing approximately equal changes with each Ca2+ increment. These results are consistent with a model where Ca2+ first binds to the two carboxy terminal sites of calmodulin with no apparent preference, concomitant with minor alterations in the microenvironments of lysines in the unoccupied amino terminal domains. The third and fourth Ca2+ ions then bind to these latter two domains, again with no evidence of preference, with little change in the lysine reactivities at the carboxy terminus of the molecule. The environments of groups in the central helix appear to undergo changes in a manner that reflects their proximity to the amino and carboxy terminal domains. In the course of this work, it was found that Lys 94 in apocalmodulin is specifically perturbed by the addition of EGTA, suggesting that the chelating agent may interact with calmodulin at or near the third Ca2+-binding domain.