C-terminal actin-binding sites of smooth muscle caldesmon switch actin between conformational states

Caldesmon is a component of the thin filaments of smooth muscles where it is believed to play an essential role in regulating the thin filaments’ interaction with myosin and hence contractility. We studied the effects of caldesmon and two recombinant fragments CaDH1 (residues 506–793) and CaDH2 (residues 683–767) on the structure of actin–tropomyosin by making measurements of the fluorescence polarisation of probes specifically attached to actin. CaDH1, like the parent molecule caldesmon, is an inhibitor of actin–tropomyosin interaction with myosin whilst CaDH2 is an activator. The F-actin in permeabilised and myosin free rabbit skeletal muscle ‘ghost’ fibres was labelled by tetramethyl rhodamine-isothiocyanate (TRITC)–phalloidin or fluorescein-5′-isothiocyanate (FITC) at lysine 61. Fluorescence polarisation measurements were made and the parameters Φ_A, Φ_E, Θ_1/2 and N were calculated. Φ_A and Φ_E are angles between the fiber axis and the absorption and emission dipoles, respectively; Θ_1/2 is the angle between the F-actin filament axis and the fiber axis; N is the relative number of randomly oriented fluorophores. Actin–tropomyosin interaction with myosin subfragment-1 induced changes in the parameters of the polarised fluorescence that are typical of strong binding of myosin to actin and of the ‘on’ conformational state of actin. Caldesmon and CaDH1 (as well as troponin in the absence of Ca²⁺) diminished the effect of S-1, whereas CaDH2 (as well as troponin in the presence of Ca²⁺) enhanced the effect of S1. Thus the structural evidence correlates with biochemical evidence that C-terminal actin-binding sites of caldesmon can modulate the structural transition of actin monomers between ‘off’ (caldesmon and CaDH1) and ‘on’ (S-1 and CaDH2) states in a manner analogous to troponin.     

Localization of the calmodulin- and the actin-binding sites of caldesmon

Expression of the C-terminal third of chicken gizzard caldesmon in Escherichia coli, using the Nagai vector (Nagai, K., and Th?gersen, H.V. (1987) Methods Enzmol. 153, 461-481), produces a cII-caldesmon fusion protein (27 kDa) with caldesmon sequence beginning at Lys579. Degradation during purification yields five peptides with molecular masses of 24, 22, 19 (two peptides), and 15 kDa. The 24-kDa peptide begins at Phe581; the 22-kDa peptide begins at Leu597, the two 19-kDa peptides begin at Phe581 and Val629, respectively; the 15-kDa peptide also begins at Val629. We estimate that the 15-kDa and one of the 19-kDa peptides end near Leu710. Site-directed mutagenesis was used to produce truncated peptides with known C termini; one peptide (17 kDa) terminates at Asn675. Digestion of the fragments with chymotrypsin generates a second 15-kDa fragment that begins at Ser666 (15K'). All of the peptides, with the exception of 15K', bind Ca(2+)-calmodulin-Sepharose and share a common 37-amino acid peptide between Val629 and Ser666, suggesting this contains the calmodulin binding site. Comparison with published sequences (Takagi, T., Yazawa, M., Ueno, T., Suzuki, S., and Yagi, K. (1989) J. Biochem. (Tokyo) 106, 778-783 and Bartegi, A., Fattoum, A., Derancourt, J., and Kassab, R. (1990) J. Biol. Chem. 265, 15231-15238) for other calmodulin-binding fragments further restricts the binding site to 7 residues, Trp-Glu-Lys-Gly-Asn-Val-Phe, between Trp659 and Ser666. All of the fragments, except the two 15-kDa peptides, co-sediment with F-actin, indicating that there are two segments in the C-terminal third of caldesmon that can interact with F-actin: one between Leu597 and Val629, the other between Arg711 and Pro756. Although separated in the primary sequence, these domains may interact with the calmodulin-binding region in the folded structure.     

C-terminal sites of caldesmon drive ATP hydrolysis cycle by shifting actomyosin itermediates from strong to weak binding of myosin and actin

Polarized fluorimetry technique and ghost muscle fibers containing tropomyosin were used to study effects of caldesmon (CaD) and recombinant peptides CaDH1 (residues 506-793), CaDH2 (residues 683-767), CaDH12 (residues 506-708) and 658C (residues 658-793) on the orientation and mobility of fluorescent label 1.5-IAEDANS specifically bound to Cys-707 of myosin subfragment-1 (S1) in the absence of nucleotide, and in the presence of MgADP, MgAMP-PNP, MgATPgammaS or MgATP. It was shown that at modelling different intermediates of actomyosin ATPase, the orientation and mobility of dye dipoles changed discretely, suggesting a multi-step changing of the myosin head structural state in ATP hydrolysis cycle. The maximum difference in orientation and mobility of the oscillator (4 degrees and 30%, respectively) was observed between actomyosin in the presence of MgATP, and actomyosin in the presence of MgADP. Caldesmon actin-binding sites C and B' inhibit formation of actomyosin strong binding states, while site B activates it. It is suggested that actin-myosin interaction in ATP hydrolysis cycle initiates nucleotide-dependent rotation of myosin motor domain, or that of its site for dye binding as well as the change in myosin head mobility. Caldesmon drives ATP hydrolysis cycle by shifting the equilibrium between strong and weak forms of actin-myosin binding.     

The interaction of caldesmon with the COOH terminus of actin

Caldesmon interacts with the NH2-terminal region of actin. It is now shown in airfuge centrifugation experiments that modification of the penultimate cysteine residue of actin significantly weakens its binding to caldesmon both in the presence and absence of tropomyosin. Furthermore, as revealed by fluorescence measurements, caldesmon increases the exposure of the COOH-terminal region of actin to the solvent. This effect of caldesmon, like its inhibitory effect on actomyosin ATPase activity, is enhanced in the presence of tropomyosin. Proteolytic removal of the last three COOH-terminal residues of actin, containing the modified cysteine residue, restores the normal binding between caldesmon and actin. These results establish a correlation between the binding of caldesmon to actin and the conformation of the COOH-terminal region of actin and suggest an indirect rather than direct interaction between caldesmon and this part of actin.     

Interaction of caldesmon and myosin subfragment 1 with the C-terminus of actin.

The interactions of caldesmon and S1 with the C-terminus of actin were examined in co-sedimentation experiments using proteolytically truncated actin. It is shown that removal of 6 residues from the C-terminus of actin reduces the binding of caldesmon by about 50% while improving the binding of S1 to actin. We also show that S1 protects actin's C-terminus from enzymatic cleavage. Both S1 and caldesmon binding to actin are decreased in the presence of an actin C-terminal peptide. These results emphasize the importance of the C-terminus of actin in binding to S1 and caldesmon.     

Phosphorylation of caldesmon prevents its interaction with smooth muscle myosin

Caldesmon is known to bind to smooth muscle myosin. Ca2+/calmodulin-dependent phosphorylation of caldesmon completely blocks its interaction with myosin. Cleavage of caldesmon at its 2 cysteine residues by 2-nitro-5-thiocyanobenzoic acid (NTCB) occurs initially at one site to yield 108-kDa and 21.2-kDa peptides and subsequently at the second site within the 108-kDa peptide to yield 85-kDa and 23.5-kDa fragments. The 23.5-kDa peptide retains the ability to bind to myosin. The N-terminal (95 kDa) and C-terminal (42 kDa) chymotryptic peptides of caldesmon were isolated and digested with NTCB: the C-terminal actin- and calmodulin-binding peptide was not cleaved, indicating that it does not contain either of the cysteine residues, whereas the 95-kDa N-terminal peptide was cleaved at two sites to yield 56-kDa, 23.5-kDa, and 21.2-kDa fragments. The arrangement of NTCB fragments in caldesmon is, therefore: 21.2 kDa/23.5 kDa/85 kDa from N to C terminus. Digestion of phosphorylated caldesmon with NTCB suggested a single phosphorylation site in the 21.2-kDa peptide and three sites in the 23.5-kDa peptide. These results lead to the development of a model whereby caldesmon may cross-link actin to myosin and such cross-linking is blocked by phosphorylation of caldesmon. This mechanism may explain the formation of reversible "latch bridges" which permit force maintenance at low levels of myosin phosphorylation in intact smooth muscles.     

Stimulation of the interaction between actin and myosin by Physarum caldesmon-like protein and smooth muscle caldesmon.

We have purified an actin-binding protein from the plasmodia of a lower eukaryote, Physarum polycephalum, with an apparent molecular mass of 210,000 daltons on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. This protein bound to actin filaments with a stoichiometry of 1:7-8 in a Ca(2+)-calmodulin-dependent manner. Antibody raised against caldesmon from smooth muscle cross-reacted with the 210-kDa protein. In vitro motility assay revealed that the 210-kDa protein increased the sliding velocity of actin filaments on Physarum myosin. The 210-kDa protein more than doubled the actin-activated ATPase activity of Physarum myosin under comparative conditions of in vitro motility assay. Further increases in the concentration of the 210-kDa protein decreased its stimulatory effects. Ca(2+)-calmodulin prevented the stimulatory effects of the 210-kDa protein. Unexpectedly, smooth muscle caldesmon also increased the sliding velocity of actin filaments on smooth muscle myosin at lower concentrations. The well-known inhibitory effect of smooth muscle caldesmon on the actin-myosin interaction was observed with this motility assay when the concentration of the caldesmon was increased further. The stimulatory and inhibitory effects were confirmed by measurements of actin-activated ATPase activity of smooth muscle myosin. From estimations of the intracellular concentrations of the 210-kDa protein and smooth muscle caldesmon in vivo, it appears that effects of the former and the latter on actin-myosin interactions in vivo are stimulatory and inhibitory, respectively.     

Effect of caldesmon on the ATPase activity and the binding of smooth and skeletal myosin subfragments to actin

We have previously shown that inhibition of the ATPase activity of skeletal muscle myosin subfragment 1 (S1) by caldesmon is correlated with the inhibition of S1 binding in the presence of ATP or pyrophosphate (Chalovich, J., Cornelius, P., and Benson, C. (1987) J. Biol Chem. 262, 5711-5716). In contrast, Lash et al. (Lash, J., Sellers, J., and Hathaway, D. (1986) J. Biol. Chem. 261, 16155-16160) have shown that the inhibition of ATPase activity of smooth muscle heavy meromyosin (HMM) by caldesmon is correlated with an increase in the binding of HMM to actin in the presence of ATP. We now show, in agreement, that caldesmon does increase the binding of smooth muscle HMM to actin-tropomyosin while decreasing the ATPase activity. The effect of caldesmon on the binding of smooth HMM is reversed by Ca2+-calmodulin. Caldesmon strengthens the binding of smooth S1.ATP and skeletal HMM.ATP to actin-tropomyosin but to a lesser extent than smooth HMM.ATP. Furthermore, this increase in binding of smooth S1.ATP and skeletal HMM.ATP does not parallel the inhibition of ATPase activity. In contrast, in the absence of ATP, all smooth and skeletal myosin subfragments compete with caldesmon for binding to actin. Thus, the effect that caldesmon has on the binding of myosin subfragments to actin-tropomyosin depends on the source of myosin, the type of subfragment, and the nucleotide present. The inhibition of actin-activated ATP hydrolysis by caldesmon, however, is not greatly different for different smooth and skeletal myosin subfragments. Evidence is presented that caldesmon inhibits actin-activated ATP hydrolysis by attenuating the productive interaction between myosin and actin that normally accelerates ATP hydrolysis. The increased binding seen by some myosin subfragments, in the presence of ATP, may be due to binding of these subfragments to a nonproductive site on actin-caldesmon. The subfragments which show an increase in binding in the presence of ATP and caldesmon appear to bind directly to caldesmon as demonstrated by affinity chromatography.     

Mapping of actin-binding sites on the heavy chain of myosin subfragment 1

When the rigor complex of actin and myosin subfragment 1 (S1) was treated with a zero-length cross-linker, 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide, covalently linked complexes of actin and S1 heavy chain with apparent molecular weights of 165,000 and 175,000 were generated. Measurements of the molar ratio of actin to S1 heavy chain in the 165K and 175K products showed that they were 1:1 complexes of actin and S1 heavy chain. Chemical cleavages of the cross-linked products followed by peptide mappings revealed that two distinct segments of S1 heavy chain spanning the 18K-20K region and the 27K-35K region from its C terminus participated in cross-linking with actin. Cross-linking of actin to the former site generated the 165K peptide while the latter site was responsible for generating the 175K peptide.     

Calcium-dependent interaction of actin filaments with actin binding protein in the presence of calmodulin and caldesmon

Caldesmon, calmodulin-, and actin-binding protein of chicken gizzard did not affect the process of polymerization of actin induced by 0.1 M KCl. Caldesmon binds to F-actin, thus inhibiting the gelation action of actin binding protein (ABP; filamin). Low shear viscosity and flow birefringence measurements revealed that in a system of calmodulin, caldesmon, ABP, and F-actin, gelation occurs in the presence of micromolar Ca2+ concentrations, but not in the absence of Ca2+. Electron microscopic observations showed the Ca2+-dependent formation of actin bundles in this system. These results were interpreted by the flip-flop mechanism: in the presence of Ca2+, a calmodulin-caldesmon complex is released from actin filaments on which ABP exerts its gelating action. On the other hand, in the absence of Ca2+, caldesmon remains bound to actin filaments, thus preventing the action of ABP.     

Interaction between myosin and F-actin. Correlation with actin-binding sites on subfragment-1

F-Actin bindings to subfragment-1 (S-1) and S-1 after limited proteolysis by trypsin (S-1t) were studied in the absence and presence of ATP by means of ultracentrifugation. No significant difference in the affinities for F-actin was observed between S-1 and S-1t in the absence of ATP. In contrast, the affinity for F-actin in the presence of ATP was decreased about 50 times by the limited proteolysis of the S-1 heavy chain. The S-1 whose SH1 and SH2 groups were cross-linked by N,N'-p-phenylenedimaleimide bound F-actin weakly. The affinity for F-actin was similar to that of unmodified S-1 in the presence of ATP and was also decreased markedly by limited proteolysis of the cross-linked S-1. Reciprocals of the dissociation constant of acto-S-1 complex decreased markedly with increase of ionic strength in the presence of ATP, but decreased only slightly at the rigor state. All these results are consistent with our proposal that S-1 has two different actin binding sites, as reported previously (Katoh, T., Imae, S., & Morita, F. (1984) J. Biochem. 95, 447-454). The mechanism of activation of S-1 ATPase by F-actin is discussed.     

The influence of caldesmon on strong binding of myosin with actin in denervated rat skeletal muscles

The effect of caldesmon (CaD) on conformational changes in F-actin modified by fluorescent probe TRITC-phalloidin was investigated by polarized fluorimetry. Changes were induced by a subfragment-1 (S-1) of myosin in the absence or presence of CaD in ghost muscle fibers obtained from intact and denervated slow (SOL) and fast (EDL) skeletal muscles of rats. S-1 binding to actin of both SOL and EDL muscles was shown to cause changes in polarized parameters of TRITC-phalloidin typical for a strong actin-myosin binding as well as of transition ofactin subunits from "off" to "on" state. CaD inhibits this significantly. Denervation atrophy inhibits the effect of S-1 as well but does not affect the capability of CaD decreasing the formation of strong binding in actomyosin complex. It is supposed that CaD "freezes" F-actin structure in "off" state. The denervation atrophy has no effect on CaD responsibility to bind thin filaments and to switch "off" actin monomers.     

Phosphorylation of high-Mr caldesmon by protein kinase C modulates the regulatory function of this protein on the interaction between actin and myosin

High-Mr caldesmon, which is involved in smooth muscle contraction, was phosphorylated by protein kinase C. By chymotryptic digestion, actin- and calmodulin-binding assays and immunoprecipitation with the antibody to the C-terminal 35-kDa fragment, we have identified that all phosphate groups are incorporated exclusively into this fragment, which is the functional domain for binding actin and calmodulin. Phosphorylation of high-Mr caldesmon and its C-terminal 35-kDa fragment reduced their binding abilities to both F-actin and calmodulin. Further, their inhibitory effects on the actin-activated ATPase activity of gizzard myosin were also reversed in proportion to the degree of phosphorylation. These results suggest that phosphorylation of high-Mr caldesmon by protein kinase C, which is restricted within the C-terminal 35-kDa domain, results in the modulation of its activity in the smooth muscle actin--myosin interaction.     

Effect of actin-binding protein on the sedimentation properties of actin

Actin and actin-binding protein (ABP) have recently been purified from human platelet cytoskeletons (S. Rosenberg, A. Stracher, and R.C. Lucas, 1981, J. Cell Biol. 91:201-211). Here, the effect of ABP on the sedimentation of actin was studied. When ABP was added to preformed F- actin filaments, it bound until a maximum ratio of 1:9 (ABP:actin, mol:mol) was reached. however, when actin was polymerized in the presence of ABP, two and a half times more ABP was able to bind to the actin- that is, every 3.4 actin monomers were now bound by an ABP dimer. ABP was not able to induce the sedimentation of actin under nonpolymerizing conditions but was able to reduce the time and concentration of actin required for sedimentation under slow polymerizing conditions. ABP, therefore, exerts its effect of G-actin by either nucleating polymerization or by cross-linking newly formed oligomers into a more sedimentable form.     

Inhibition of the relative movement of actin and myosin by caldesmon and calponin.

Contractile activity of myosin II in smooth muscle and non-muscle cells requires phosphorylation of myosin by myosin light chain kinase. In addition, these cells have the potential for regulation at the thin filament level by caldesmon and calponin, both of which bind calmodulin. We have investigated this regulation using in vitro motility assays. Caldesmon completely inhibited the movement of actin filaments by either phosphorylated smooth muscle myosin or rabbit skeletal muscle heavy meromyosin. The amount of caldesmon required for inhibition was decreased when tropomyosin is present. Similarly, calponin binding to actin resulted in inhibition of actin filament movement by both smooth muscle myosin and skeletal muscle heavy meromyosin. Tropomyosin had no effect on the amount of calponin needed for inhibition. High concentrations of calmodulin (10 microM) in the presence of calcium completely reversed the inhibition. The nature of the inhibition by the two proteins was markedly different. Increasing caldesmon concentrations resulted in graded inhibition of the movement of actin filaments until complete inhibition of movement was obtained. Calponin inhibited actin sliding in a more "all or none" fashion. As the calponin concentration was increased the number of actin filaments moving was markedly decreased, but the velocity of movement remained near control values.     

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.     

Effect of nucleotide on interaction of the 567-578 segment of myosin heavy chain with actin

To probe the effect of nucleotide on the formation of ionic contacts between actin and the 567-578 residue loop of the heavy chain of rabbit skeletal muscle myosin subfragment 1 (S1), the complexes between F-actin and proteolytic derivatives of S1 were submitted to chemical cross-linking with 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide. We have shown that in the absence of nucleotide both 45 kDa and 5 kDa tryptic derivatives of the central 50 kDa heavy chain fragment of S1 can be cross-linked to actin, whereas in the presence of MgADP.AlF4, only the 5 kDa fragment is involved in cross-linking reaction. By the identification of the N-terminal sequence of the 5-kDa fragment, we have found that trypsin splits the 50 kDa heavy chain fragment between Lys-572 and Gly-573, the residues located within the 567-578 loop. Using S1 preparations cleaved with elastase, we could show that the residue of 567-578 loop that can be cross-linked to actin in the presence of MgADP.AlF4 is Lys-574. The observed nucleotide-dependent changes of the actin-subfragment 1 interface indicate that the 567-578 residue loop of skeletal muscle myosin participates in the communication between the nucleotide and actin binding sites.     

Effect of ethylene glycol on the interaction of different myosin subfragment 1.nucleotide complexes with actin

To elucidate the structure of the cross-bridge intermediates in the actomyosin ATPase cycle, several laboratories have added both ethylene glycol and AMP-PNP to muscle fibers. These studies suggested that ethylene glycol shifts the structure of myosin.AMP-PNP toward the weak-binding conformation, i.e., toward the structure of myosin.ATP. Since only the weak-binding conformation of myosin subfragment 1 (S-1) binds with no apparent cooperativity to the troponin-tropomyosin-actin complex (regulated actin), we used this as a probe to examine the conformation of various S-1.nucleotide complexes in ethylene glycol. Our results show that ethylene glycol markedly weakens the binding strength of S-1, S-1.ADP, and S-1.AMP-PNP to actin but has almost no effect on the binding strength of S-1.ATP. As in muscle fibers, at 40% ethylene glycol, the binding strength of S-1.AMP-PNP to actin becomes very similar to the binding strength of S-1.ATP. In the presence of troponin-tropomyosin, the binding of S-1.AMP-PNP to actin shows no apparent cooperativity in 40% ethylene glycol. Therefore, our results confirm that ethylene glycol shifts the structure of the myosin.AMP-PNP toward the weak-binding conformation. However, our results also suggest that ethylene glycol has a direct effect on the regulated actin complex. This is shown by the fact that ethylene glycol markedly increases the cooperative binding of S-1.ADP to regulated actin both in the presence and in the absence of Ca2+.(ABSTRACT TRUNCATED AT 250 WORDS)     

Identification of polyphosphate recognition sites communicating with actin sites on the skeletal myosin subfragment 1 heavy chain

Using the thrombin-cut [68-30 kilodalton (kDa)] myosin subfragment 1 (S-1) whose heavy chain has been selectively split within the central 50-kDa region, at Lys-560, with concomitant specific alterations of the ATPase and actin binding properties [Chaussepied, P., Mornet, D., Audemard, E., Derancourt, J., & Kassab, R. (1986) Biochemistry 25, 1134-1140; Chaussepied, P., Mornet, D., Barman, T., Travers, F., & Kassab, R. (1986) Biochemistry 25, 1141-1149], we have isolated and renatured the COOH-terminal 30-kDa fragment associated with the alkali light chains by the procedure recently described [Chaussepied, P., Mornet, D., Audemard, E., Kassab, R., Goodearl, J., Levine, B., & Trayer, I. P. (1986) Biochemistry 25, 4540-4547]. The 30-kDa peptide preparation was found to exhibit a crucial feature of the native S-1; namely, it interacts with F-actin in an adenosine 5'-triphosphate (ATP)-dependent manner. Studies by ultracentrifugation, turbidity measurements, and chemical cross-linking experiments showed that the acto-30-kDa peptide complex was dissociated almost completely by the gamma-phosphoryl group containing ligands ATP, 5'-adenylyl imidodiphosphate, and pyrophosphate, to a lesser extent by ADP, and not at all by AMP and inorganic phosphate. The maximal dissociating effect is operating with the thrombic 30-kDa entity, whereas the 22-kDa fragment produced by staphylococcal protease is only slightly dissociated. In contrast, the tryptic 20-kDa fragment binds irreversibly to actin.(ABSTRACT TRUNCATED AT 250 WORDS)