Turkey Gizzard Myosin Research Articles

Conformation-dependent proteolysis of smooth-muscle myosin.

The folded 10 S conformation of turkey gizzard myosin is more resistant to proteolysis by papain than the extended 6 S conformation. These findings confirm those of Onishi and Watanabe (Onishi, H., and Watanabe, S. (1984) J. Biochem. (Tokyo) 95, 899-902). In addition, we suggest that the effect of phosphorylation on heavy-chain digestion by papain is related to the dependence of conformation on phosphorylation and not to a direct effect of phosphorylation itself. Proteolysis by Staphylococcus aureus protease and trypsin also is slower for 10 S compared to the 6 S conformation. Heavy chain hydrolysis by alpha-chymotrypsin is not dependent on myosin conformation. Filamentous myosin and heavy meromyosin are more resistant to papain proteolysis in the dephosphorylated compared to the phosphorylated states. The different sensitivities to proteolysis probably are caused by changes in the subfragment 1-subfragment 2 region of the molecule rather than at the heavy meromyosin-light meromyosin junction. These changes are induced as part of the 6 S-10 S transition and occur in monomeric and filamentous myosin and in heavy meromyosin. These more subtle alterations in the head-neck junctions of the molecule may be more important in modifying myosin enzymatic activity than the actual interaction of the tail and neck regions of the molecule.

Inhibition of turkey gizzard myosin light chain kinase activity by BAY K 8644

BAY K 8644, a dihydropyridine with positive inotropic effects in smooth muscle, was found to inhibit calmodulin-dependent turkey gizzard myosin light chain kinase activity with an IC50 of 80 μM. In contrast, when myosin light chain kinase was rendered calmodulin-independent by limited proteolysis, drug inhibition was markedly diminished. BAY K 8644 inhibition was additive with that of the negative-inotropic dihydropyridine felodipine. These results indicate that the inotropic effects of dihydropyridines in smooth muscle are unrelated to their calmodulin-antagonistic properties.

Isolation of the native form of chicken gizzard myosin light-chain kinase.

A simple and rapid procedure for the purification of the native form of chicken gizzard myosin light-chain kinase (Mr 136000) is described which eliminates problems of proteolysis previously encountered. During this procedure, a calmodulin-binding protein of Mr 141000, which previously co-purified with the myosin light-chain kinase, is removed and shown to be a distinct protein on the basis of lack of kinase activity, different chymotryptic peptide maps, lack of cross-reactivity with a monoclonal antibody to turkey gizzard myosin light-chain kinase, and lack of phosphorylation by the purified catalytic subunit of cyclic AMP-dependent protein kinase. This Mr-141000 calmodulin-binding protein is identified as caldesmon on the basis of Ca2+-dependent interaction with calmodulin, subunit Mr, Ca2+-independent interaction with skeletal-muscle F-actin, Ca2+-dependent competition between calmodulin and F-actin for caldesmon, and tissue content.

Correlation of intrinsic fluorescence and conformation of smooth muscle myosin.

The addition of ATP to turkey gizzard myosin causes an enhancement of the intrinsic tryptophan fluorescence. The level of fluorescence enhancement is determined by the myosin conformation. The transition of myosin from the folded (10 S) state to the extended (6 S) state is accompanied by a decrease in the fluorescence level. Phosphorylation-dephosphorylation of myosin does not directly influence fluorescence and will induce changes only if the myosin conformation is altered. Under the appropriate conditions, phosphorylation of myosin favors the transition of 10 S to 6 S. The phosphorylation dependence of the associated fluorescence decrease is not linear, and it is proposed that the phosphorylation of both light chains is required for the full transition. The tryptophan residues involved respond to the binding of ATP at the hydrolytic sites. Since the fluorescence properties of gizzard myosin are influenced by the myosin conformation, it is reasonable to assume that the active sites are also modified by the shape of the myosin molecule.

Functional interactions between smooth muscle myosin light chain kinase and calmodulin.

Calmodulin (CaM) binding by turkey gizzard myosin light chain kinase (MLCK) causes subtle changes in the fluorescence emission and polarization excitation spectra of the enzyme. Fluorescence experiments using 9-anthroyl-choline (9AC), which competes with ATP in binding, demonstrate mutually stabilizing interactions between the CaM and ATP binding sites corresponding to delta G = -0.6 to -0.7 kcal/mol. Fluorescence titrations in the presence of 9AC or 5,5'-bis[8-(phenylamino)-1-naphthalenesulfonate] confirm the stoichiometry of 1 mol of CaM/MLCK. Phosphorylation of MLCK has no effect on either the protein fluorescence or the binding of ATP and 9AC. The dissociation constant for the MLCL-CaM complex is increased approximately 500-fold on phosphorylation. Values of Kd for the phosphorylated enzyme range from 0.5 to 1.1 microM in 0.2 N KCl, pH 7.3, 25 degrees C. We showed competition between MLCK and other CaM binding proteins and peptides by using both fluorescence and catalytic activity measurements. Competition for CaM occurs with ACTH, beta-endorphin, substance P, glucagon, poly(L-arginine), myelin basic protein, troponin I, and histone H2A. Phosphorylation of the last three proteins by the adenosine cyclic 3',5'-phosphate dependent protein kinase diminishes their ability to compete. Phosphorylation of MLCK by the protein kinase gives 0.95 +/- 0.04 and 2.2 +/- 0.4 mol of incorporated 32P in the presence and absence of CaM, respectively. These stoichiometries agree with those recently reported [Conti, M. A. & Adelstein, R. S. (1981) J. Biol. Chem. 256, 3178].

Subcellular localization of myosin light chain kinase in skeletal, cardiac, and smooth muscles.

Antibodies were elicited against turkey gizzard myosin light chain kinase (MLCK), purified by affinity chromatography on the enzyme bound to Sepharose, and used to localize myosin kinase--in rabbit fast skeletal, slow skeletal, cardiac, and smooth muscles--by indirect immunofluorescence. When studied on nitrocellulose replicas of NaDodSO4/polyacrylamide gel electrophoretograms, antibodies were specific for the Mr 140,000 MLCK of gizzard smooth muscle. By using the same technique, they were shown to recognize the Mr 140,000 MLCK and a Mr 75,000 polypeptide--presumably derived from the former by proteolysis--in rat arterial and stomach smooth muscle as well as in rat thyroid cells. The same antibodies reacted only with a Mr approximately equal to 75,000 protein from rat cardiac and skeletal muscle. Antibodies inhibited the activity of smooth and skeletal myosin kinases in an in vitro assay with approximately equal to 11 mole of antibody needed for 50% inhibition of 1 mole of gizzard enzyme. The antibodies stain vascular and gizzard smooth muscle cells with no apparent segregation of the enzyme in a specific part of the cell. In contrast, sarcomeric muscles exhibit a striated staining pattern, superimposable to the staining by antiactin antibodies. This shows that (i) antibodies are not species- or tissue-specific, (ii) they recognize kinases that differ in their molecular weight and ability to be phosphorylated, probably at the level of their common catalytic and calmodulin-binding domains, and (iii) sarcomeric muscle kinases are at least in part bound to the contractile apparatus and their distribution is restricted to a specific part of the sarcomere. This raises the possibility that myosin phosphorylation may be controlled not only by the Ca2+ concentration but also by actin-myosin interaction.

Reversible phosphorylation of smooth muscle myosin, heavy meromyosin, and platelet myosin.

Smooth muscle myosin was purified from turkey gizzards with the 20,000-dalton light chains in the unphosphorylated state. The actin-activated MgATPase activity was 4 nmol/min/mg at 25 degrees C. When the myosin was phosphorylated to 2 mol of Pi/mol of myosin using purified myosin light chain kinase, calmodulin, and ATP, the actin-activated MgATPase activity rose to 51 nmol/min/mg. Complete dephosphorylation of the same myosin by a purified phosphatase lowered the activity to 5 nmol/min/mg, and complete rephosphorylation of the myosin following inhibition of the phosphatase raised it again to 46 nmol/min/mg. Human platelet myosin could be substituted for turkey gizzard myosin, with similar results. A chymotryptic fragment of smooth muscle myosin which retains the phosphorylated site on the 20,000-dalton light chain of myosin was prepared. Using the same scheme for reversible phosphorylation, this smooth muscle heavy meromyosin was found to show the same positive correlation between phosphorylation of the myosin light chain and the actin-activated MgATPase activity. The results with smooth muscle heavy meromyosin show that the effect of phosphorylation on the actin-activated MgATPase activity can be separated from the effects of phosphorylation on myosin filament assembly.

Phosphorylation-dependent regulation of Limulus myosin.

Myosin from Limulus, the horseshoe crab, is shown to be regulated by a calcium-calmodulin-dependent phosphorylation of its regulatory light chains. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis of a Limulus myosin preparation reveals three light chain bands. Two of these light chains have been termed regulatory light chains based on their ability to bind to light chain-denuded scallop myofibrils (Sellers, J. R., Chantler, P. D., and Szent-Györgyi, A. G. (1980) J. Mol. Biol. 144, 223-245). Ths other light chain does not bind to these myofibrils and is thus termed the essential light chain. Both Limulus regulatory light chains can be phosphorylated with a highly purified turkey gizzard myosin light chain kinase or with a partially purified myosin light chain kinase which can be isolated from Limulus muscle by affinity chromatography on a calmodulin-Sepharose column. Phosphorylation with both of these enzymes requires calcium and calmodulin. Limulus myosin is isolated in an unphosphorylated form. The MgATPase of this unphosphorylated myosin is only slightly activated by rabbit skeletal muscle actin plus tropomyosin. The calcium-dependent phosphorylation of the myosin results in an increase in the actin-activated MgATPase rate. Once phosphorylated, the actin-activated MgATPase rate is only slightly modified by calcium. This suggests that calcium operates mainly at the level of the myosin kinase-calmodulin system.

Direct photoaffinity labeling by nucleotides of the apparent catalytic site on the heavy chains of smooth muscle and Acanthamoeba myosins.

The heavy chains of Acanthamoeba myosins. IA, IB and II, turkey gizzard myosin, and rabbit skeletal muscle myosin subfragment-1 were specifically labeled by radioactive ATP, ADP, and UTP, each of which is a substrate or product of myosin ATPase activity, when irradiated with UV light at 0 degrees C. With UTP, as much as 0.45 mol/mol of Acanthamoeba myosin IA heavy chain and 1 mol/mol of turkey gizzard myosin heavy chain was incorporated. Evidence that the ligands were associated with the catalytic site included the observations that reaction occurred only with nucleotides that are substrates or products of the ATPase activity; that the reaction was blocked by pyrophosphate which is an inhibitor of the ATPase activity; that ATP was bound as ADP; and that label was probably restricted to a single peptide following limited subtilisin proteolysis of labeled Acanthamoeba myosin IA heavy chain and extensive cleavage with CNBr and trypsin of labeled turkey gizzard myosin heavy chain.

Phosphorylation of smooth muscle myosin light chains by cAMP-dependent protein kinase.

The 20,000-dalton light chain of chicken gizzard myosin was phosphorylated in vitro by the cAMP-dependent protein kinase (ATP: protein phosphotransferase, EC 2.7.1.37) from bovine heart. The enzyme catalyzed incorporation of 1 mol of Pi/mol of light chain in a reaction that was completely dependent upon cAMP and independent of Ca2+. Two-dimensional peptide mapping of alpha-chymotryptic digests, as well as phosphoamino acid analysis of acid hydrolysates, were used to compare the site phosphorylated by cAMP-dependent protein kinase to that phosphorylated by turkey gizzard myosin light chain kinase. The results indicate that both enzymes phosphorylate the same serine residue. However, the light chains were a better in vitro substrate for myosin light chain kinase than for cAMP-dependent protein kinase. The amino acid sequence around the phosphorylated serine is characteristic of substrates of cAMP-dependent protein kinases.

Large-scale purification and characterization of calmodulin from ram testis its metal-ion-dependent conformers

Calmodulin was isolated in large quantities from ram testis by a simple procedure involving sequentially ammonium sulfate fractionation, heat treatment, anion exchange chromatography on DEAE-cellulose and gel filtration on Sephacryl S-200. Divalent cations (Mg 2+ and/or Ca 2+) were present throughout the purification which was entirely performed in the absence of chelators. The final yield was approx. 90 mg per kg testis. Ram testis calmodulin appears to be essentially identical to the brain homologous protein by the following criteria: ultraviolet absorption spectrum, amino acid composition showing a single residue of ϵ- N-trimethyl lysine, and tryptic peptide maps obtained by high performance liquid chromatography. Turkey gizzard myosin light-chain kinase, the activation of which is extremely specific for calmodulin (Walsh, M.P., Vallet, B., Cavadore, J.C. and Demaille, J.G. (1980) J. Biol. Chem. 255, 335–337), was indeed activated by ram testis calmodulin in the presence of calcium. The isolated protein migrated at different rates upon sodium dodecyl sulfate polyacrylamide gel electrophoresis, depending on the absence or presence of divalent metals which probably induce different conformations. The relative migration rates were Ca 2+ > Mn 2+ > Mg 2+ > EDTA. In the presenceof divalent metals, the observed doublet may be ascribed to the equilibrium between ion-free and ion-saturated forms, which exhibited different Stokes radii, as already suggested (Grab, D.J., Berzins, K., Cohen, R.S. and Siekevitz, P. (1979) J. Biol. Chem. 254, 8690–8696).