Hair-forming activity of human lymphocyte specific protein 1 requires cooperation between its caldesmon-like domains and the villin headpiece-like domains.

LSP1 is an F-actin binding with multiple F-actin binding domains. Overexpression of LSP1 in NAD 47/89 patient's neutrophils created hair-like projections on the patient's neutrophil cell surfaces and inhibited neutrophil cell motility and transfection of LSP1 in serial cell lines recreate the NAD 47/89 phenotype and produce branching hair-like surface projections. Although LSP1 contains hair-forming ability and LSP1 F-actin binding domains have been defined, the LSP1 domains responsible for its hair-forming activity, the relationship to the F-actin binding domains, and the required domain interactions, if any, for hair formation are not well understood. To define the hair-forming domains of LSP1, the relationship to the known F-actin binding domains, and binding domain interactions, LSP1 truncates, which include or exclude the different F-actin binding domains, were created by PCR. LSP1 mutants were created by site-directed mutagenesis to define the amino acids important for hair formation. Sf9 cells were infected with recombinant baculovirus expressing the cDNA of LSP1 truncates and mutants, and the morphology of infected Sf9 cells was documented by DIC optics. Results show that (1) the hair-forming activity of LSP1 is localized to the basic C-terminal half of the molecule, which contains all of the F-actin binding domains; (2) both the caldesmon-like domains and the villin headpiece-like domains are required for the hair-forming activity of LSP1; (3) basic amino acids in the villin headpiece regions are crucial for the hair-forming activity of LSP1 molecule. The results suggest cooperation between the caldesmon-like domains and the villin headpiece-like domains are required for the hair-forming activity of human LSP1 in cells.     

The lymphocyte-specific protein LSP1 binds to F-actin and to the cytoskeleton through its COOH-terminal basic domain

The lymphocyte-specific phosphoprotein LSP1 associates with the cytoplasmic face of the plasma membrane and with the cytoskeleton. Mouse LSP1 protein contains 330 amino acids and contains an NH2-terminal acidic domain of approximately 177 amino acids. The COOH-terminal half of the LSP1 protein is rich in basic residues. In this paper we show that LSP1 protein which is immunoprecipitated with anti-LSP1 antibodies from NP-40-soluble lysates of the mouse B-lymphoma cell line BAL17 is associated with actin. In vitro binding experiments using recombinant LSP1 (rLSP1) protein and rabbit skeletal muscle actin show that LSP1 binds along the sides of F-actin but does not bind to G-actin. rLSP1 does not alter the initial polymerization kinetics of actin. The highly conserved COOH-terminal basic domains of mouse and human LSP1 share a significant homology with the 20-kD COOH-terminal F-actin binding fragment of caldesmon. A truncated rLSP1 protein containing the entire COOH-terminal basic domain from residue 179 to 330, but not the NH2-terminal acidic domain binds to F-actin at least as well as rLSP1. When LSP1/CAT fusion proteins are expressed in a LSP1-negative T-lymphoma cell line, only fusion proteins containing the basic COOH-terminal domain associate with the NP-40-insoluble cytoskeleton. These data show that LSP1 binds F-actin through its COOH-terminal basic domain and strongly suggest that LSP1 interacts with the cytoskeleton by direct binding to F-actin. We propose that LSP1 plays a role in mediating cytoskeleton driven responses in lymphocytes such as receptor capping, cell motility, or cell-cell interactions.     

High-resolution crystal structures of villin headpiece and mutants with reduced F-actin binding activity

Villin-type headpiece domains are approximately 70 amino acid modular motifs found at the C terminus of a variety of actin cytoskeleton-associated proteins. The headpiece domain of villin, a protein found in the actin bundles of the brush border epithelium, is of interest both as a compact F-actin binding domain and as a model folded protein. We have determined the high-resolution crystal structures of chicken villin headpiece (HP67) at 1.4 A resolution as well as two mutants, R37A and W64Y, at 1.45 and 1.5 A resolution, respectively. Replacement of R37 causes a 5-fold reduction in F-actin binding affinity in sedimentation assays. Replacement of W64 results in a much more drastic reduction in F-actin binding affinity without significant changes in headpiece structure or stability. The detailed comparison of these crystal structures with each other and to our previously determined NMR structures of HP67 and the 35-residue autonomously folding subdomain in villin headpiece, HP35, provides the details of the headpiece fold and further defines the F-actin binding site of villin-type headpiece domains.     

Phosphoinositide-binding peptides derived from the sequences of gelsolin and villin.

The polyphosphoinositides phosphatidylinositol 4-monophosphate (PIP) and phosphatidylinositol 4,5-bisphosphate (PIP2) inactivate the actin filament-severing proteins villin and gelsolin and dissociate them from monomeric and polymeric actin. A potential polyphosphoinositide- (PPI) binding site of human plasma gelsolin regulating filament severing has been localized to the region between residues 150-169 and to the corresponding region in villin which occurs in the second of six homologous domains present in both proteins. Synthetic peptides based on these sequences bind tightly to both PIP and PIP2, in either micelles or bilayer vesicles, compete with gelsolin for binding to PPIs, and dissociate gelsolin-PIP2 complexes, restoring severing activity to the protein. These peptides also bind with moderate affinity to F-actin, suggesting that inactivation of the severing function of the intact proteins by PPIs results from competition between actin and PPIs for a critical binding site on gelsolin-villin. The PPI-binding peptides contain numerous basic amino acids, but their effects on PPIs are far greater than those of Arg or Lys oligomers, a highly basic peptide derived from the calmodulin-binding site of myristoylated, alanine-rich kinase C substrate protein, or the 5-kDa actin-binding protein thymosin beta-4, suggesting that specific aspects of the primary and secondary structure of these basic peptides are important for their interaction with the acidic headgroups of PPIs. In addition to elucidating the structure of PIP2-binding sites in gelsolin, the results describe a sensitive assay for phosphoinositide-binding molecules based on their ability to prevent inhibition of gelsolin function.     

Effect of the C-terminal actin-binding sites of caldesmon on the interaction of actin with myosin

TRITC-phalloidin or FITC-labeled F-actin of ghost muscle fibers was bound to tropomyosin and C-terminal recombinant fragments of caldesmon CaDH1 (residues 506-793) or CaDH2 (residues 683-767). After that the fibers were decorated with myosin subfragment 1. In the absence of caldesmon fragments, subfragment 1 interaction with F-actin caused changes in parameters of polarized fluorescence, that were typical of "strong" binding of myosin heads to F-actin and of the "switched on" conformational state of actin. CaDH1 inhibited, whereas CaDH2 activated the effect of subfragment 1. It is suggested that C-terminal part of caldesmon may modulate the transition of F-actin subunits from the "switched on" to the "switched off" state.     

Identification of a functional switch for actin severing by cytoskeletal proteins

Actin severing is vital for the organization of the actin cytoskeleton during cell motility. Severing of F-actin by the homologous proteins villin and gelsolin requires unphysiologically high calcium concentrations (20-200 microM). Here we demonstrate that high calcium releases an autoinhibited conformation in villin that is maintained by two low affinity calcium binding sites (aspartic acids 467 and 715) that interact with a cluster of basic residues in the S2 domain of villin. Mutation of either of these sites as well as tyrosine phosphorylation alters the conformation of villin resulting in a protein that can sever actin in nanomolar calcium. These results suggest that tyrosine phosphorylation rather than high calcium may be the mechanism by which villin and other related proteins sever actin in vivo.     

Domain mapping of chicken gizzard caldesmon

Limited proteolysis, affinity chromatography, and immunoblotting have been used to define the domains of chicken gizzard caldesmon, caldesmon120, that interact with calmodulin, F-actin, and a monoclonal antibody prepared using human platelet caldesmon. Treatment of caldesmon120 with chymotrypsin produces groups of fragments near 100, 80, 60, 38, and 20 kDa. Further digestion produces peptides between 40 and 50 kDa. The 100- and 80-kDa peptides cross-react with the monoclonal antibody; the smaller polypeptides do not. The kinetics of cleavage and the antibody studies indicate that the 38- and 80-kDa fragments are the two major pieces of the 120-kDa protein. The 38-kDa fragment, purified by high performance liquid chromatography, and several of its subfragments at 21 and 25 kDa sediment with F-actin, bind to calmodulin-Sepharose in the presence of Ca2+, and are displaced from F-actin by Ca2+-calmodulin. The 80-kDa fragments did not interact with F-actin or calmodulin. We have tentatively placed the 38-kDa fragment at the C-terminal using polyclonal antibodies selected against a beta-galactosidase-caldesmon120 fusion protein produced by a lambda gt11 lysogen. The 38-, 25-, and 21-kDa fragments cross-react with these antibodies; the 80- and 60-kDa fragments do not. Caldesmon77 from human platelets also cross-reacts with these selected antibodies. The results suggest that interacting calmodulin and F-actin binding sites are localized on a 38-kDa C-terminal fragment of caldesmon. The smallest subfragment of this peptide that binds to both F-actin and calmodulin-Sepharose is about 21 kDa. The monoclonal antibody epitope is tentatively localized near the N-terminal of caldesmon77 and must be within 50 kDa of the N-terminal on caldesmon120.     

In vivo analysis of functional domains from villin and gelsolin

Transfected CV1 cells were used to compare the in vivo effects of various domains of villin and gelsolin. These two homologous actin modulating proteins both contain a duplicated severin-like sequence. Villin has in addition a carboxy-terminal domain, the headpiece, which accounts for its bundling activity. The effects of the villin-deleted mutants were compared with those of native villin. Our results show that essential domains of villin required to induce the growth of microvilli and F-actin redistribution are present in the first half of the core and in the headpiece. We also show that the second half of the villin core cannot be exchanged by its homolog in gelsolin. When expressed at high levels of CV1 cells, full length gelsolin completely disrupted stress fibers without change of the cell shape. Addition of the villin headpiece to gelsolin had no effect on the phenotype induced by gelsolin alone. Expression of the first half of gelsolin induced similar modifications as capping proteins and rapid cell mortality; this deleterious effect on the cell structure was also observed when the headpiece was linked to the first half of gelsolin. In cells expressing the second half of gelsolin, a dotted F-actin staining was often seen. Moreover elongated dorsal F-actin structures were observed when the headpiece was linked to the second gelsolin domain. These studies illustrate the patent in vivo severing activity of gelsolin as well as the distinct functional properties of villin core in contrast to gelsolin.     

Amino acid sequence studies on cyanogen bromide peptides of chicken caldesmon which bind to calmodulin

Three major calmodulin-binding cyanogen bromide peptides (fragments A, B, and D) were isolated from chicken gizzard muscle caldesmon and their amino acid sequences were determined. The molecular masses of fragments A, B, and D were estimated to 16, 12, and 9 kDa, respectively, by SDS-urea polyacrylamide gel electrophoresis. Fragment A was composed of 102 amino acid residues and contained homoserine at the C terminus. The amino acid sequence from the 37th residue of fragment A corresponds to the N-terminal sequence of the 15 kDa peptide which was obtained by thrombin digestion [Mornet, D., Audemard, E., & Derancourt, J. (1988) Biochem. Biophys. Res. Commun. 154, 564-571]. Thrombin 15 kDa peptide binds to F-actin but does not bind to calmodulin. Thus the N-terminal 36 residues and the C-terminal part from the 37th residue of fragment A are supposed to bind to calmodulin and F-actin, respectively. The sequences of fragments B and D were identical, but fragment D was composed of 64 amino acid residues and ended with tryptophan, whereas fragment B was of 98 or 99 amino acid residues and ended with proline. Both fragments B and D are supposed to be the C-terminal peptides of chicken caldesmon. Fragment B had heterogeneous sequences at the C-terminal region. These results can explain the reported heterogeneity of chicken caldesmon in charge and molecular mass.     

The isolated sixth gelsolin repeat and headpiece domain of villin bundle F-actin in the presence of calcium and are linked by a 40-residue unstructured sequence

Villin is an F-actin regulating, modular protein with a gelsolin-like core and a distinct C-terminal "headpiece" domain. Localized in the microvilli of the absorptive epithelium, villin can bundle F-actin and, at higher calcium concentrations, is capable of a gelsolin-like F-actin severing. The headpiece domain can, in isolation, bind F-actin and is crucial for F-actin bundling by villin. While the three-dimensional structure of the isolated headpiece is known, its conformation in the context of attachment to the villin core remains unexplored. Furthermore, the dynamics of the linkage of the headpiece to the core has not been determined. To address these issues, we employ a 208-residue modular fragment of villin, D6-HP, which consists of the sixth gelsolin-like domain of villin (D6) and the headpiece (HP). We demonstrate that this protein fragment requires calcium for structural stability and, surprisingly, is capable of Ca2+-dependent F-actin bundling, suggesting that D6 contains a cryptic F-actin binding site. NMR resonance assignments and 15N relaxation measurements of D6-HP in 5 mM Ca2+ demonstrate that D6-HP consists of two independent structural domains (D6 and HP) connected by an unfolded 40-residue linker sequence. The headpiece domain in D6-HP retains its structure and interacts with D6 only through the linker sequence without engaging in other interactions. Chemical shift values indicate essentially the same secondary structure elements for D6 in D6-HP as in the highly homologous gelsolin domain 6. Thus, the headpiece domain of villin is structurally and functionally independent of the core domain.     

Characterization of the carboxyl-terminal 10-kDa cyanogen bromide fragment of caldesmon as an actin-calmodulin-binding region

A pair of 10-kDa peptides, designated CB-a and CB-b, was isolated by calmodulin-Sepharose chromatography from a total CNBr digest of turkey gizzard caldesmon. CB-a encompasses the COOH-terminal segment of residues 659-756, according to the sequence of adult chicken gizzard caldesmon (Bryan, J., Imai, M., Lee, R., Moore, P., Cook, R.G., and Lin, W.G. (1989) J. Biol. Chem. 264, 13873-13879), whereas CB-b comprises the same structure but was a few amino acids shorter at its COOH terminus. Both peptides cosedimented with F-actin, and their binding was increased by smooth muscle tropomyosin. The Kd values were 1.3 and 0.5 microM, in the absence and presence of tropomyosin, respectively, with a maximum binding capacity of 6.9 actins/mol of peptides. The CB-a/CB-b fragments inhibited, in a tropomyosin-sensitive and Ca2(+)-calmodulin-dependent manner, the skeletal actomyosin subfragment 1 ATPase activity to a level close but not identical to that observed for the parent caldesmon. Ca2(+)-calmodulin was selectively cross-linked to either caldesmon or the CNBr peptides with 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide producing 1:1 covalent complexes that were retained neither by phenyl-Sepharose nor by immobilized calmodulin. Moreover, the cross-linked caldesmon bound weakly to F-actin and did not inhibit the actomyosin subfragment 1 ATPase in the absence of Ca2+. The results suggest that the CB-a/CB-b peptide region contains major regulatory determinants of caldesmon.     

Actin-binding proteins are conserved from slime molds to man

DNA clones encoding the actin-binding proteins alpha-actinin and severin from Dictyostelium discoideum were isolated and sequenced. Comparisons of the deduced amino acid sequences with proteins from other species showed striking similarities at distinct regions. The F-actin cross-linking molecule alpha-actinin carries two characteristic EF-hand structures highly homologous to the Ca2+-binding loops of proteins from the calmodulin superfamily. An N-terminal region that is conserved in alpha-actinin from D. discoideum and vertebrates is also related to parts of the dystrophin sequence and might represent the F-actin binding site. Severin, gelsolin, villin, and fragmin share homologous sequences that are believed to participate in the severing activity of these proteins.     

Human and mouse LSP1 genes code for highly conserved phosphoproteins

With use of the mouse LSP1 cDNA we isolated a human homologue of the mouse LSP1 gene from a human CTL cDNA library. The predicted protein sequence of human LSP1 is compared with the predicted mouse LSP1 protein sequence and regions of homology are identified in order to predict structural features of the LSP1 protein that might be important for its function. Both the human and mouse LSP1 proteins consist of two domains, an N-terminal acidic domain and a C-terminal basic domain. The C-terminal domains of the mouse and human LSP1 proteins are highly conserved and include several conserved, putative serine/threonine phosphorylation sites. Immunoprecipitation of LSP1 protein from 32P-orthophosphate-loaded cells show that both the mouse and human LSP1 proteins are phosphoproteins. The sequences of the putative Ca2(+)-binding sites present in the N-terminal domain of the mouse LSP1 protein are not conserved in the human LSP1 protein; however, a different Ca2(+)-binding site may exist in the human protein, indicating a functional conservation rather than a strict sequence conservation of the two proteins. The expression of the human LSP1 gene follows the same pattern as the expression of the mouse LSP1 gene. Southern analysis of human genomic DNA shows multiple LSP1-related fragments of varying intensity in contrast to the simple pattern found after similar analysis of mouse genomic DNA. By using different parts of the human LSP1 cDNA as a probe, we show that most of these multiple bands contain sequences homologous to the conserved C-terminal region of the LSP1 cDNA. This suggests that there are several LSP1-related genes present in the human genome.     

Functional comparison of villin and gelsolin. Effects of Ca2+, KCl, and polyphosphoinositides

A family of homologous actin-binding proteins sever and cap actin filaments and accelerate actin filament assembly. The functions of two of these proteins, villin and gelsolin, and of their proteolytically derived actin binding domains were compared directly by measuring their effects, under various ionic conditions, on the rates and extents of polymerization of pyrene-labeled actin. In 1 mM Ca2+ and 150 mM KCl, villin and gelsolin have similar severing and polymerization-accelerating properties. Decreasing [Ca2+] to 25 microM greatly reduces severing by villin but not gelsolin. Decreasing [KCl] from 150 to 10 mM at 25 microM Ca2+ increases severing by villin, but not gelsolin, over 10-fold. The C-terminal half domains of both proteins have Ca2+-sensitive actin monomer-binding properties, but neither severs filaments nor accelerates polymerization. The N-terminal halves of villin and gelsolin contain all the filament-severing activity of the intact proteins. Severing by gelsolin's N-terminal half is Ca2+-independent, but that of villin has the same Ca2+ requirement as intact villin. The difference in Ca2+ sensitivity extends to 14-kDa N-terminal fragments which bind actin monomers and filament ends, requiring Ca2+ in the case of villin but not gelsolin. Severing of filaments by villin and its N-terminal half is shown to be inhibited by phosphatidylinositol 4,5-bisphosphate, as shown previously for gelsolin (Janmey, P.A., and Stossel, T.P. (1987) Nature 325, 362-364). The functional similarities of villin and gelsolin correlate with known structural features, and the greater functional dependence of villin on Ca2+ compared to gelsolin is traced to differences in their N-terminal domains.     

MAPKAPK2-mediated LSP1 phosphorylation and FMLP-induced neutrophil polarization

In neutrophils, the major substrate of MAPKAPK2 (MK2) is an F-actin binding protein LSP1. Studies using mutants of the two potential Serine phosphorylation sites in LSP1 C-terminal F-actin binding region indicated that the major phosphorylation site for MK2 is Ser243 in murine neutrophils (Ser252 in humans). Human phosphoLSP1 antibodies that recognize phosphoSer252 site were prepared and revealed fMLP-induced neutrophil LSP1 phosphorylation. The phosphorylation was inhibited by p38 MAPK (upstream kinase for MK2) inhibitor SB203580. The antibodies also detect LSP1 phosphorylation in murine neutrophils. Immunostaining revealed that in WT murine neutrophils phosphoLSP1 was localized in F-actin enriched lamellipodia and oriented toward the fMLP gradient while non-phosphoLSP1 failed to colocalize with F-actin. In suspension, WT neutrophils exhibited persistent F-actin polarization following fMLP stimulation, while MK2(-/-) neutrophils exhibited transient F-actin polarization. These studies suggest that MK2-regulated LSP1 phosphorylation is involved in stabilization of F-actin polarization during neutrophil chemotaxis.     

Blockage of AMV reverse transcriptase by antisense oligodeoxynucleotides.

Wild type chicken gizzard caldesmon (756 amino acids) was expressed in a T7 RNA polymerase-based bacterial expression system at a yield of 1 mg pure caldesmon per litre bacterial culture. A mutant composed of amino acids 1-578 was also constructed and expressed. The wild type and mutant caldesmon were purified and compared with native chicken gizzard caldesmon. Native and wild type expressed caldesmon were indistinguishable in assays for inhibition of actin-tropomyosin activation of myosin ATPase, reversal of inhibition by Ca²⁺-calmodulin and binding to actin, actin-tropomyosin, Ca²⁺-calmodulin, tropomyosin and myosin. The mutant missing the C-terminal 178 amino acids had no inhibitory effect and did not bind to actin or Ca²⁺-calmodulin. It bound to tropomyosin with a 5-fold reduced affinity and to myosin with a greater than 10-fold reduced affinity.     

Phosphatidylserine liposomes can be tethered by caldesmon to actin filaments.

Rotary shadowing electron microscopy revealed that attachment of caldesmon to phosphatidylserine (PS) liposomes was mainly through its C-terminal end. To determine the PS-binding sites of caldesmon, we have made use of synthetic peptides covering the two C-terminal calmodulin binding sites and a recombinant fragment corresponding to the N-terminal end of the C-terminal domain that contains an amphipathic helix. Interactions of these peptides with the PS liposomes were studied by nondenaturing gel electrophoresis and fluorescence spectroscopy. The results showed that both calmodulin-binding sites of caldesmon were able to interact with PS. The affinity (Kd) of PS for these sites was in the range of 1.8-14.3 x 10(-5) M, compared to 0.69 x 10(-5) M for the whole caldesmon molecule. Fragments located outside of calmodulin-binding sites bound PS weakly (3.85 x 10(-4) M) and thus may contain a second class of lipid-binding sites. Binding of PS induced conformational changes in regions other than the C-terminal PS-binding sites, as evidenced by the changes in the susceptibility to proteolytic cleavages. Most significantly, the presence of caldesmon greatly increased binding of PS to F-actin, suggesting that caldesmon may tether PS liposomes to actin filaments. These results raise the possibility that caldesmon-lipid interactions could play a functionally important role in the assembly of contractile filaments near the membranes.     

Modes of caldesmon binding to actin: sites of caldesmon contact and modulation of interactions by phosphorylation

Smooth muscle caldesmon binds actin and inhibits actomyosin ATPase activity. Phosphorylation of caldesmon by extracellular signal-regulated kinase (ERK) reverses this inhibitory effect and weakens actin binding. To better understand this function, we have examined the phosphorylation-dependent contact sites of caldesmon on actin by low dose electron microscopy and three-dimensional reconstruction of actin filaments decorated with a C-terminal fragment, hH32K, of human caldesmon containing the principal actin-binding domains. Helical reconstruction of negatively stained filaments demonstrated that hH32K is located on the inner portion of actin subdomain 1, traversing its upper surface toward the C-terminal segment of actin, and forms a bridge to the neighboring actin monomer of the adjacent long pitch helical strand by connecting to its subdomain 3. Such lateral binding was supported by cross-linking experiments using a mutant isoform, which was capable of cross-linking actin subunits. Upon ERK phosphorylation, however, the mutant no longer cross-linked actin to polymers. Three-dimensional reconstruction of ERK-phosphorylated hH32K indeed indicated loss of the interstrand connectivity. These results, together with fluorescence quenching data, are consistent with a phosphorylation-dependent conformational change that moves the C-terminal end segment of caldesmon near the phosphorylation site but not the upstream region around Cys(595), away from F-actin, thus neutralizing its inhibitory effect on actomyosin interactions. The binding pattern of hH32K suggests a mechanism by which unphosphorylated, but not ERK-phosphorylated, caldesmon could stabilize actin filaments and resist F-actin severing or depolymerization in both smooth muscle and nonmuscle cells.     

Characterization of mitotically phosphorylated caldesmon.

Mitosis-specific phosphorylation by cdc2 kinase causes nonmuscle caldesmon to dissociate from microfilaments (Yamashiro, S., Yamakita, Y., Ishikawa, R., and Matsumura, F. (1990) Nature 344, 675-678; Yamashiro, S., Yamakita, Y., Hosoya, H., and Matsumura, F. (1991) Nature 349, 169-172). To explore the function of mitosis-specific phosphorylation of caldesmon, in vivo- and in vitro-phosphorylated caldesmons have been characterized. We have found that both in vivo and in vitro phosphorylation of caldesmon causes similar changes in the properties, including reduction in actin, calmodulin, and myosin binding of caldesmon, and a decrease in the inhibition of actomyosin ATPase by caldesmon. Rat non-muscle caldesmon is phosphorylated in vitro up to a ratio of 7 mol/mol of protein. Actin-binding constants of both a high affinity (K a = 1.2 x 10(7) M-1) and a low affinity (K a = 1 x 10(6) M-1) site of unphosphorylated caldesmon are reduced to less than 10(5) M-1 with 5 mol of phosphate incorporation per mol of protein. Actin-bound caldesmon can be phosphorylated by cdc2 kinase, which results in the dissociation of caldesmon from F-actin. Caldesmon has a second myosin-binding site in the C terminus, in addition to the N terminus myosin-binding domain previously reported, because the bacterially expressed C terminus of caldesmon shows binding to myosin. Phosphorylation of the C-terminal fragments decreases their myosin-binding affinity as observed with intact caldesmon. These results suggest that caldesmon loses most of its in vitro functions during mitosis as a result of phosphorylation, which may be required for the reorganization of microfilaments during mitosis.