Caldesmon structure and function: sequence analysis of a 35 kilodalton actin- and calmodulin-binding fragment from the C-terminus of the turkey gizzard protein

Inhibition of cross-bridge binding to actin by caldesmon fragments in skinned skeletal muscle fibers

Several regions within the 35-kDa COOH-terminal portion of caldesmon have been implicated in the ability of caldesmon to inhibit actin-activated myosin ATPase activity. To further define the functional regions of caldesmon, we have studied the effects of three chymotryptic fragments, one fragment produced by CNBr digestion and two fragments produced by digestion with submaxillaris arginase C protease, on the relaxed stiffness and active force of rabbit psoas fibers. Each of the regions of caldesmon studied had either direct or indirect effects on single-fiber mechanics. The 35-kDa and 20-kDa fragments of caldesmon, like intact caldesmon, were effective inhibitors of fiber stiffness, a measure of cross-bridge attachment. The 7.3-kDa and 10-kDa fragments, which constitute the NH2 and COOH halves of the 20-kDa fragment, inhibited both relaxed fiber stiffness and active force production, but with a reduced efficacy compared to the 20-kDa fragment. These results suggest that several regions within the 35-kDa COOH-terminal region of caldesmon are required for optimum function of caldesmon and that function includes inhibition of weak cross-bridge attachment and force production.

Strong interaction between caldesmon and calponin

Caldesmon was labeled at either Cys-153 in the NH2-terminal domain or Cys-580 in the COOH-terminal domain with a 6-acryloyl-2-dimethylaminonaphthalene (acrylodan) fluorescence probe. The addition of smooth muscle calponin to Cys-580-labeled caldesmon resulted in an 18% drop in fluorescence intensity, which titrated with a stoichiometry of 0.9 and a binding constant of 9.5 x 10(7) M-1. For Cys-153-labeled caldesmon, there was no change in fluorescence upon adding calponin. These findings indicate strong binding between calponin and the COOH-domain of caldesmon. The association was sensitive to ionic strength, suggesting that ionic interactions between calponin, a basic protein, and caldesmon, an acidic protein, contribute to the stabilization of the protein complex. That non-muscle acidic calponin interacts with caldesmon with a much reduced association constant of 3.5 x 10(6) M-1 supports such a model. The binding between acidic calponin and caldesmon is strengthened to 1.8 x 10(7) M-1 in the presence of Ca2+, which might bind to acidic residues of the calponin and partially neutralize its negative charge. The strong, specific binding between calponin and caldesmon suggests that this interaction occurs within smooth muscle cells and possibly plays a role in the regulation of contraction.

In vitro motility analysis of smooth muscle caldesmon control of actin-tropomyosin filament movement

We have used the in vitro motility assay to investigate the effect of caldesmon on the movement of actin-tropomyosin filaments over thiophosphorylated smooth muscle myosin and skeletal muscle heavy meromyosin. Using either motor, incorporation of up to 8 nM caldesmon inhibited filament movement by decreasing the proportion of filaments motile from > 85% to < 30%. There was a minimal effect on filament attachment and a modest decrease in motile filament velocity in this concentration range. The reduction in the proportion of filaments motile could be completely reversed by incorporation of an excess of calmodulin at pCa 4.5. The expressed C-terminal fragment, 606C, which retains caldesmon's inhibitory capacity but does not bind to myosin, decreased the proportion of filaments motile but had no effect on velocity. We conclude that the velocity reduction by whole caldesmon is due to actin-myosin cross-linking. A significant decrease in filament attachment was observed when caldesmon was added to an excess over actin (> 10 nM). In the absence of tropomyosin, addition of an excess of caldesmon caused a similar decrease in the filament density, but there was no effect on the proportion of filaments that were motile. Our results demonstrate that caldesmon can switch actin-tropomyosin from motile to non-motile states without controlling velocity of movement or weak binding affinity and show the inhibitory action of caldesmon in the motility assay to be functionally indistinguishable from that reported for troponin.

Identification of functioning regulatory sites and a new myosin binding site in the C-terminal 288 amino acids of caldesmon expressed from a human clone

A partial clone of caldesmon, coding for the C-terminal 288 amino acids, was isolated from a human fetal liver cDNA library and sequenced. Expression of the clone in Escherichia coli produced a peptide called H1 (M(r) 32,549), which inhibited tropomyosin-enhanced actomyosin Mg(2+)-ATPase activity by 90% with half maximal inhibition at 0.03-0.04 mol H1 per mol actin. The inhibition could be reversed by Ca(2+)-calmodulin. H1 bound actin, Ca(2+)-calmodulin and tropomyosin and smooth muscle myosin with high affinities. This latter finding shows the presence of a second myosin-binding site in caldesmon. This was confirmed in thrombic digests of native sheep aorta and chicken gizzard caldesmon.

A mosaic multiple-binding model for the binding of caldesmon and myosin subfragment-1 to actin

Binding of caldesmon to actin causes a decrease in the quantity of bound myosin and results in a reduction in the rate of actin-activated adenosine triphosphate hydrolysis. It is generally assumed that the binding of caldesmon and myosin to actin is a pure competitive interaction. However, recent binding studies of enzyme digested caldesmon subfragments directed at mapping the actin binding site of caldesmon have shown that a small 8-kD fragment around the COOH-terminal can compete directly with the myosin subfragment 1 (S-1) binding to actin; at least one other fragment that binds to actin does not inhibit the actin-activated adenosine triphosphate activity of myosin. That is, only a part of the caldesmon sequence may be responsible for directly blocking the binding of S-1 to actin. This prompts us to question the actual mode of binding of intact caldesmon and myosin S-1 to actin: whether the entire intact caldesmon molecule is competing with S-1 binding (pure competitive model) or just a small part of it (mosaic multiple-binding model). To answer this question, we measured the amount of myosin S-1 and caldesmon bound per actin monomer as a function of the total concentration of S-1 added to the system at constant concentrations of actin and caldesmon. A formalism for calculating the titration data based on the pure competitive model and a mosaic multiple-binding model was then developed. When compared with theoretical calculations, it is found that the binding of caldesmon and S-1 to actin cannot be pure competitive if no cooperativity exists between S-1 and caldesmon. In contrast, the mosaic multiple-binding model can fit the binding data rather well regardless of the existence of cooperativity between S-1 and caldesmon.

Structural relationships of actin-binding proteins

The sequences of a large number of actin-binding proteins have been compared. These findings, together with the results of protein-chemical analysis, peptide synthesis and site-directed and deletion mutagenesis, have led to the assignment of actin-binding sites. Within these segments, small actin-binding motifs have been delineated. Most actin-binding proteins interact with actin subdomain-1 but our analyses reveal neither primary nor secondary structure homology among these proteins, suggesting that actin binding does not follow simple structural principles.

The molecular anatomy of caldesmon

Images

Identification of the site phosphorylated by casein kinase II in smooth muscle caldesmon

Phosphorylation of avian gizzard caldesmon by casein kinase II was investigated. The enzyme incorporates about 1 mol of phosphate per mol of caldesmon. All sites of phosphorylation are located in short chymotryptic peptides with Mr 25-27 kDa or in the short N-terminal peptide formed after cleavage of chicken gizzard caldesmon at Cys153. The primary structure of the tryptic peptide containing the main site of duck gizzard caldesmon phosphorylation is S-E-V-N-A-Q-N-X-V-A-E-D-E-T-K, where X is an unidentified residue, presumed to be phosphoserine. Thus, Ser73 is the main site phosphorylated by casein kinase II in avian gizzard caldesmon.

Chemical modification of the sole histidine residue of smooth muscle caldesmon

Caldesmon was stoichiometrically N-carbethoxylated specifically at the only histidine residue (His-610) with diethylpyrocarbonate. Carbethoxylation of a 1:1 molar complex of caldesmon and calmodulin in the presence of Ca2+ resulted in the stoichiometric N-carbethoxylation of His-610 of caldesmon and His-107 of calmodulin. Carbethoxy-caldesmon, like the unmodified protein, bound to immobilized calmodulin (in the presence of Ca2+) and to immobilized tropomyosin (at low ionic strength). The affinity of F-actin for carbethoxy-caldesmon (Kd = 1.29 x 10(-6)M) was similar to that for unmodified caldesmon (Kd = 0.88 x 10(-6)M), and the modified protein was as effective as control caldesmon in the inhibition of the actin-activated MgATPase of skeletal muscle myosin. We conclude that the predicted basic amphiphilic alpha-helical sequence (Arg-593-His-610) does not represent the calmodulin-binding site of caldesmon. Furthermore, His-610 does not play a major role in the interaction of caldesmon with F-actin or tropomyosin.

Conformational change of turkey-gizzard caldesmon induced by specific chemical modification with carbodiimide

Water soluble 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide was used to internally cross-link carboxyl and lysyl groups of caldesmon. The modification did not involve the two cysteines of the molecule which were previously labelled with N-iodoacetyl-N'-(5-sulfo-1-naphthyl)ethylenediamine. The modified caldesmon exhibited a smaller Stokes radius (4.0 nm instead of 6.3 nm) and its electrophoretic mobility corresponded to an apparent molecular mass of approximately 82 kDa, appreciably lower than that of the native molecule (120 kDa), but more similar to the reported true molecular mass of 86,974 Da of chicken-gizzard caldesmon (Bryan, J., Imai, M., Lee, R., Moore, P., Cook. R. G. & Lin, W. (1989) J. Biol. Chem. 264, 13,873-13,879). Comparative circular dichroism analysis indicated a decrease of the alpha-helix content from 43% to 36% resulting from the chemical modification. The 1H-NMR spectra of the native and modified caldesmon showed that the covalent cross-linking affected mainly the central and N-terminal parts of the molecule. The C-terminal part, rich in aromatic amino acids, was unmodified by the carbodiimide treatment. This was also corroborated by the continued ability of the modified caldesmon to bind to actin and calmodulin, and by the property of the 90-kDa proteolytic N-terminal fragment to give an internally cross-linked species of 60 kDa. Using electron microscopy, the modified protein was shown to have a more compact shape and a reduced capacity to induce tight and long F-actin bundles. These conformational changes were obtained when the carbodiimide reaction was conducted at pH 6.0 and were not observed at pH 8.0. This suggests that local variation of the pH might affect the conformation of caldesmon which changes from an elongated to more compact shape, stabilized by electrostatic interactions. It is proposed that the flexibility of caldesmon might be involved in the regulatory function of this protein in the smooth muscle and might favour tightly packed F-actin bundles or weaker interactions between actin filaments.

Epitope mapping of monoclonal antibodies against caldesmon and their effects on the binding of caldesmon to Ca++/calmodulin and to actin or actin-tropomyosin filaments

The effects of monoclonal anti-caldesmon antibodies, C2, C9, C18, C21, and C23, on the binding of caldesmon to F-actin/F-actin-tropomyosin filaments and to Ca++/calmodulin were examined in an in vitro reconstitution system. In addition, the antibody epitopes were mapped by Western blot analysis of NTCB (2-nitro-5-thiocyanobenzoic acid) and CNBr (cyanogen bromide) fragments of caldesmon. Both C9 and C18 recognize an amino terminal fragment composed of amino acid residues 19 to 153. The C23 epitope lies within a fragment ranging from residues 230 to 386. Included in this region is a 13-residue repeat sequence. Interestingly this repetitive sequence shares sequence similarity with a sequence found in nuclear lamin A, a protein which is also recognized by C23 antibody. Therefore, it is likely that the C23 epitope corresponds to this 13-residue repeat sequence. A carboxyl-terminal 10K fragment contains the epitopes for antibodies C2 and C21. Among these antibodies, only C21 drastically inhibits the binding of caldesmon to F-actin/F-actin-tropomyosin filaments and to Ca++/calmodulin. When the molar ratio of monoclonal antibody C21 to caldesmon reached 1.0, a maximal inhibition (90%) on the binding of caldesmon to F-actin filaments was observed. However, it required double amounts of C21 antibody to exhibit a maximal inhibition of 70% on the binding of caldesmon to F-actin-tropomyosin filaments. These results suggest that the presence of tropomyosin in F-actin enhances caldesmon's binding. Furthermore, C21 antibody also effectively inhibits the caldesmon binding to Ca++/calmodulin. The kinetics of C21 inhibition on caldesmon's binding to Ca++/calmodulin is very similar to the inhibition obtained by preincubation of caldesmon with free Ca++/calmodulin. This result suggests that there is only one Ca++/calmodulin binding domain on caldesmon and this domain appears to be very close to the C21 epitope. Apparently, the Ca++/calmodulin-binding domain and the actin-binding domain are very close to each other and may interfere with each other. In an accompanying paper, we have further demonstrated that microinjection of C21 antibody into living chicken embryo fibroblasts inhibit intracellular granule movement, suggesting an in vivo interference with the functional domains [Hegmann et al., 1991: Cell Motil. Cytoskeleton 20:109-120].

Some properties of duck gizzard caldesmon

The domain structure of duck gizzard caldesmon was investigated. A single thiol group is located in the vicinity of the C-terminus of the protein. A simple method for the purification of a short (21 kDa) C-terminal peptide formed after chemical cleavage of caldesmon at cysteine residues was evolved. The C-terminal peptide of caldesmon interacts with calmodulin with an affinity one order of magnitude higher than that of native caldesmon. The Ca2+/phospholipid-dependent protein kinase (protein kinase C) transfers about 2 mol of phosphate per mol of caldesmon. All sites phosphorylated by protein kinase C are located in the short (21 kDa) C-terminal peptide of caldesmon. Phosphorylation does not affect the interaction of caldesmon with calmodulin.

Interaction of smooth muscle caldesmon with phospholipids

Taking into account the perimembrane localization of caldesmon [(1986) Nature 319, 68] and its ability to participate in the regulation of receptor clusterization [(1989) J. Biol. Chem. 264, 496], we studied the interaction of duck gizzard caldesmon with soybean phospholipids (azolectin). By using four independent methods, i.e. light scattering, gel-electrophoresis, gel-filtration and ultracentrifugation, we showed a Ca-independent complex formation between caldesmon and azolectin. Interacting with caldesmon, calmodulin is shown to dissociate the caldesmon-azolectin complex. It is supposed that the caldesmon-phospholipid interaction may affect caldesmon phosphorylation by Ca-phospholipid-dependent protein kinase. This effect may be important for various cell motility processes.

Structural study of gizzard caldesmon and its interaction with actin. Binding involves residues of actin also recognised by myosin subfragment 1

The interaction between actin and caldesmon that is associated with the inhibition of actomyosin ATPase activity in smooth muscle has been studied using 1H-NMR spectroscopy. Binding studies using the intact molecules were complemented by the use of thrombic cleavage fragments of both turkey and chicken gizzard caldesmon as well as defined peptides of actin, in order to investigate the conformational properties of caldesmon and to localise regions of the primary structures that participate in protein-protein contacts. The binding of caldesmon is shown to involve distinct segments on the N-terminal region (residues 1-44) of actin, as previously observed for the inhibitory component of the thin filament of striated muscle, troponin I [Levine et al. (1988) Eur. J. Biochem. 153, 389-397]. The comparable structural properties of these tissue-specific inhibitors of actomyosin ATPase and the similarities in their mode of interaction at the N-terminal region of actin suggest common aspects to the structural mechanism for thin-filament regulation in smooth and striated muscle. Unlike the inhibitory interaction of troponin I, however, the binding of caldesmon to the N-terminal region of actin directly involves groups within residues 20-41 of actin that are also recognised by myosin subfragment 1. The complementary segment of caldesmon has been localised to a 15-kDa thrombic fragment (residues 483-578) derived from the N-terminal portion of a 35-kDa proteolytic cleavage product from the C-terminal of caldesmon whose interaction with actin is modulated by calmodulin. The results are discussed in relation to the calcium-mediated mechanism for thin-filament regulation in smooth and striated muscle.

Complexes between 15 kDa caldesmon fragment and actin investigated by immuno-electron microscopy

The regulatory system of smooth muscle thin filaments is thought to involve a major calcium-calmodulin and actin binding protein: caldesmon. A dissective approach was used to isolate a 35 kDa C-terminal fragment of the molecule and to produce antibodies reacting against both the intact and the 15 kDa N-terminal end of this parental fragment. While this purified 15 kDa caldesmon fragment demonstrates a weak actin association, we observed that it cross-links actin filaments into loose bundles. These structures were labelled with a selective antibody and showed regular periodic striation with repeats of approximately 40 nm. This work brings additional information to previous reports using an actin and calmodulin binding 25 kDa C-terminal fragment of the caldesmon molecule [(1989) J. Biol. Chem. 264, 2869-2875]. We demonstrate that a purified fragment corresponding to a sequence smaller than 96 amino acids, which contains no cystein residue, is able to interact with actin at a single site which is not the calmodulin modulated.

Caldesmon, calmodulin and tropomyosin interactions

Binary complex interactions between caldesmon and tropomyosin, and calmodulin and tropomyosin, and ternary complex interaction involving the three proteins were studied using viscosity, electron microscopy, fluorescence and affinity chromatography techniques. In 10 mM NaCl, caldesmon decreased the viscosity of chicken gizzard tropomyosin by 7-8 fold with a concomitant increase in turbidity (A330nm). Electron micrographs showed spindle-shaped particles in the tropomyosin-caldesmon samples. These results suggest side-by-side aggregation of tropomyosin polymers induced by caldesmon. Binding studies in 10 mM NaCl between caldesmon and chicken gizzard tropomyosin labelled with the fluorescent probe N-(1-anilinonaphthyl-4)maleimide (ANM) gave association constants from 5.3.10(6) to 7.9.10(6) M-1 and stoichiometry from 1.0 to 1.4 tropomyosin per caldesmon. Similar binding was observed for rabbit cardiac tropomyosin and caldesmon. Removal of 18 and 11 residues from the COOH ends of the gizzard and cardiac tropomyosin by carboxypeptidase A, respectively, had no significant effect on their binding to caldesmon. In the presence of Ca2+, chicken gizzard tropomyosin bound to a calmodulin-Sepharose-4B column and was eluted with a salt concentration of 140 mM. This interaction was weakened in the absence of Ca2+, and the bound tropomyosin was eluted by 65 mM KCl. ANM-labelled tropomyosin bound calmodulin in the presence of Ca2+ with a binding constant of 3.5.10(6) M-1 and a binding stoichiometry of 1 to 1.4 tropomyosin per calmodulin. In 10 mM NaCl, calmodulin reduced the specific viscosity of chicken gizzard tropomyosin in the presence of Ca2+ by 5 fold, while a 1.5-fold reduction in viscosity was observed in the absence of Ca2+. In either case, no significant increase in turbidity was observed suggesting that calmodulin reduced head-to-tail polymerization of tropomyosin. The interaction of caldesmon with the calmodulin-ANM-tropomyosin complex in the presence and absence of Ca2+ was also examined. The result is consistent with a model that in the absence of Ca2+, calmodulin binds weakly to either caldesmon or tropomyosin and has little effect on the tropomyosin-caldesmon interaction; whereas, Ca2(+)-calmodulin interacts with caldesmon and reduces its affinity to tropomyosin.

Competitive binding of the troponin T-specific pool of caldesmon antibodies and tropomyosin to skeletal troponin T and smooth muscle caldesmon

The fraction of polyclonal caldesmon antibodies cross-reacting with rabbit skeletal troponin T are shown to compete with smooth muscle tropomyosin for caldesmon and troponin T, as revealed by ELISA method. The epitope recognized by these antibodies was also found in Mr 77 kDa non-muscle caldesmon. These results provide functional confirmation for the suggestion that the regions of amino acid sequence homology in caldesmon isoforms and troponin T belong to the tropomyosin binding sites.