The binding of skeletal muscle C-protein to F-actin, and its relation to the interaction of actin with myosin subfragment-1.

C-protein is a component of thick filaments of skeletal muscle myofibrils. It is bound to the assembly of myosin tails that forms the filament backbone. We report here that C-protein can also bind to F-actin, with a limiting stoichiometry of approximately one C-protein molecule per 3 to 5 actin subunits and a dissociation constant in the micromolar range at ionic strength 0·07. The binding is not significantly affected by ATP, calcium ions or temperature, or by the presence of tropomyosin on the actin, but it is weakened by increasing ionic strength. Myosin subfragment-1 (S-1) competes with C-protein for binding to actin. In the absence of ATP, S-1 displaces nearly all bound C-protein from actin, while in the presence of ATP, C-protein inhibits the actin activation of S-1 ATPase. Although there is no direct evidence that interaction of C-protein with actin is physiologically significant, the lenght of the C-protein molecule is sufficient so that it could make contact with the thin filaments in muscle while remaining attached to the thick filaments.     

Developmental relationship of myosin binding proteins (myomesin, connectin and C-protein) to myosin in chicken somites as studied by immunofluorescence microscopy

The developmental relationship of myosin binding proteins (myomesin, connectin and C-protein) to myosin was studied in chicken cervical somites by immunofluorescence microscopy. Muscle and non-muscle myosins initially appeared as slender rods at the same sites, and then, fused to form non-striated fibrils. As muscle myosin formed striated structures (A bands), non-muscle myosin disappeared from this structure. Myomesin (reactive with monoclonal antibodies MyB4 and MyBB78) and connectin (carboxy terminal region, reactive with monoclonal antibody T51) were seen as dots in the center of these myosin rods. These proteins then formed characteristic mature striations on non-striated fibrils of myosin. Earlier alignment of these myosin binding proteins rather than myosin indicates that the correct assembly of these proteins seems to be related to the formation of initial myosin rods as well as subsequent linear and periodic alignment of myosin molecules to form early A bands. Connectin spots reactive with 9D10 were scattered around myosin rods/myomesin dots/connectin T51 dots. These spots may represent radiating connectin filaments from these rods/dots to link myosin rods to the I-Z-I structures of myofibrils to be incorporated. Since the slow isoform of C-protein formed its characteristic bands ("doublets") prior to H zone formation within A bands by myosin, this isoform may help to precisely align myosin filaments within the A band region. The presence of the slow, then the slow and the cardiac, and finally the co-existence of the slow and the fast isoforms of C-protein may interfere with the incorporation and co-polymerization of non-adult isoforms into myofibrils.     

Immunohistochemical analysis of C-protein isoforms in cardiac and skeletal muscle of the axolotl,Ambystoma mexicanum

Of the several proteins located within sarcomeric A-bands, C-protein, a myosin binding protein (MyBP) is thought to regulate and stabilize thick filaments during assembly. This paper reports the characterization of C-protein isoforms in juvenile and adult axolotls, Ambystoma mexicanum, by means of immunofluorescent microscopy and Western blot analyses. C-protein and myosin are found specifically within the A-bands, whereas tropomyosin and -actin are detected in the I-bands of axolotl myofibrils. The MF1 antibody prepared against the fast skeletal muscle isoform of chicken C-protein specifically recognizes a cardiac isoform (Axcard1) in juvenile and adult axolotls but does not label axolotl skeletal muscle. The ALD66 antibody, which reacts with the C-protein slow isoform in chicken, localizes only in skeletal muscle of the axolotl. This slow axolotl isoform (Ax_slow) displays a heterogeneous distribution in fibers of dorsalis trunci skeletal muscle. The C₃₁₅ antibody against the chicken C-protein cardiac isoform identifies a second axolotl cardiac isoform (Ax_card2), which is present also in axolotl skeletal muscle. No C-protein was detected in smooth muscle of the juvenile and adult axolotl with these antibodies.This work was supported by NIH grants HL-32184 and HL-37702 and a grant-in-aid from the American Heart Association to L.F.L.     

Alterations in Ca2+ sensitive tension due to partial extraction of C- protein from rat skinned cardiac myocytes and rabbit skeletal muscle fibers

C-protein, a substantial component of muscle thick filaments, has been postulated to have various functions, based mainly on results from biochemical studies. In the present study, effects on Ca(2+)-activated tension due to partial removal of C-protein were investigated in skinned single myocytes from rat ventricle and rabbit psoas muscle. Isometric tension was measured at pCa values of 7.0 to 4.5: (a) in untreated myocytes, (b) in the same myocytes after partial extraction of C-protein, and (c) in some myocytes, after readdition of C-protein. The solution for extracting C-protein contained 10 mM EDTA, 31 mM Na2HPO2, 124 mM NaH2PO4, pH 5.9 (Offer et al., 1973; Hartzell and Glass, 1984). In addition, the extracting solution contained 0.2 mg/ml troponin and, for skeletal muscle, 0.2 mg/ml myosin light chain-2 in order to minimize loss of these proteins during the extraction procedure. Between 60 and 70% of endogenous C-protein was extracted from cardiac myocytes by a 1-h soak in extracting solution at 20-23 degrees C; a similar amount was extracted from psoas fibers during a 3-h soak at 25 degrees C. For both cardiac myocytes and skeletal muscle fibers, partial extraction of C-protein resulted in increased active tension at submaximal concentrations of Ca2+, but had little effect upon maximum tension. C-protein extraction also reduced the slope of the tension-pCa relationships, suggesting that the cooperativity of Ca2+ activation of tension was decreased. Readdition of C-protein to previously extracted myocytes resulted in recovery of both tension and slope to near their control values. The effects on tension did not appear to be due to disruption of cooperative activation of the thin filament, since C-protein extraction from cardiac myocytes that were 40-60% troponin-C (TnC) deficient produced effects similar to those observed in cells that were TnC replete. Measurements of the tension-pCa relationship in skeletal muscle fibers were also made at a sarcomere length of 3.5 microns which, because of the distribution of C-protein on the thick filament, should eliminate any interaction between C-protein and actin. The effects of C-protein extraction were similar at long and short sarcomere lengths. These data are consistent with a model in which C-protein modulates the range of movement of myosin, such that the probability of myosin binding to actin is increased after its extraction.     

Structural evidence for the interaction of C-protein (MyBP-C) with actin and sequence identification of a possible actin-binding domain

C-protein (MyBP-C) is a myosin-binding protein that is usually seen in two sets of seven to nine positions in the C-zones in each half of the vertebrate striated muscle A-band. Skeletal muscle C-protein is a modular structure containing ten sub-domains (C1 to C10) of which seven are immunoglobulin-type domains and three (C6, C7 and C9) are fibronectin-like domains. Cardiac muscle C-protein has an extra N-terminal domain (C0) and also some sequence insertions, one of which provides phosphorylation sites. It is conceivable that C-protein has both a structural and regulatory role within the sarcomere. The precise mode of binding of C-protein to the myosin filament has not been determined. However, detailed ultrastructural studies have suggested that C-protein, which binds to myosin, can give rise to a longer periodicity (about 435A) than the intrinsic myosin filament repeat of 429A. The reason for this has remained a puzzle for over 25 years. Here we show by modelling and computation that the presence of this longer periodicity could be explained if the myosin-binding part of C-protein binds to myosin with the expected 429A repeat, but if there are systematic interactions of the N-terminal end of C-protein with the neighbouring actin filaments in the hexagonal lattice of filaments in the A-band. We also show that if they occur these interactions would probably only arise in defined muscle states. Further analysis of the MyBP-C sequence identifies a possible actin-binding domain in the Pro-Ala-rich sequence found at the N terminus of skeletal MyBP-C and between domains C0 and C1 in the cardiac sequence.     

A new protein of the thick filaments of vertebrate skeletal myofibrils. Extractions, purification and characterization

A new protein component of skeletal myofibrils has been isolated and characterized. It is prepared from impure myosin preparations and corresponds to band C, the principal contaminant observed in sodium dodecyl sulphate polyacrylamide gel electrophoresis patterns of such preparations (Starr and Offer, 1971).The C-protein, as we term it, is deduced to be a component of the skeletal myofibril because (i) glycerinated or fresh myoflbrils contain a component with a mobility identical to C-protein on sodium dodecyl sulphate gels, (ii) this component is extracted from myofibrils by the same solvent which extracts C-protein and (iii) C-protein may be prepared from preparations of isolated myofibrils. It is presumed to be a component of the thick filaments because it binds strongly to myosin at low ionic strength; immunological evidence which confirms this view is presented elsewhere.The quantity of C-protein in the myofibril has been estimated to be 2.0% by densitometry of sodium dodecyl sulphate gels of glycerinated myofibrils using actin as an internal reference. About forty molecules of C-protein are present in a thick filament.The properties of C-protein distinguish it from the other well-characterized myoflbrillar proteins. The C-protein molecule contains a single polypeptide chain of molecular weight 140,000. The intrinsic viscosity of 13.6 ml/g suggests that the molecule is neither completely globular nor as elongated as molecules like paramyosin or tropomyosin. The α-helical content is very low and the proline content higher than the other myofibrillar proteins. The molecule associates at low ionic strength.C-protein has no ATPase activity, nor does it affect the ATPase of pure myosin. But it reduces the activity of the actin-activated myosin ATPase by about half, this inhibition being independent of the level of Ca²⁺. C-protein does not bind Ca²⁺ in the presence of Mg²⁺. Its possible location and function are discussed.     

Immunochemical analysis of porcine cardiac C-protein

C-protein has been isolated from pig heart and its immunochemical properties studied. It is extracted with myosin, and separated from the myosin on a DEAE-Sephadex column. The amount of C-protein recovered from crude myosin is approx. 3.5%. The molecular weight of C-protein is 150,000. Anti-C-protein serum reacts with crude myosin and purified C-protein but not with purified myosin in immunodiffusion plates. Cardiac C-protein does not react with anti-skeletal white muscle C-protein serum. Immunoblotting experiments show that anti-cardiac C-protein serum reacts with a Mr = 150,000 component in myofibrils or crude myosin. C-protein is located in the A-band, except the M-line region, of the myofibrils. These results indicate that C-protein is an intrinsic component of the thick filaments in pig heart myofibrils.     

Effect of C-protein on actomyosin ATPase

The effect of C-protein on the actin-activated ATPase of column-purified skeletal muscle myosin has been investigated at varied ionic strength. At ionic strengths below about 0.1, C-protein is a potent inhibitor. The inhibition is not reversed by increasing the actin concentration, showing that it is caused by C-protein bound to the myosin filaments. When the ionic strength is raised above about 0.12, on the other hand, the inhibition vanishes and C-protein becomes a mild activator of the actomyosin ATPase. Both effects appear rapidly upon addition of C-protein to pre-formed myosin filaments, so C-protein probably acts by binding to the surface of the filaments.     

Localization of C-protein isoforms in chicken skeletal muscle: ultrastructural detection using monoclonal antibodies

Monoclonal antibodies (McAbs) specific for the fast (MF-1) and slow (ALD-66) isoforms of C-protein from chicken skeletal muscle have been produced and characterized. Using these antibodies it was possible to demonstrate that skeletal muscles of varying fiber type express different isoforms of this protein and that in the posterior latissimus dorsi muscle both isoforms are co-expressed in the same myofiber (17, 18). Since we had shown that both isoforms were present in all sarcomeres, it was feasible to test whether the two isoforms co- distributed in the same 43-nm repeat within the A-band, thereby establishing a minimum number of C-proteins per repeat in the thick filaments. Here we describe the ultrastructural localization of C- protein in myofibers from three muscle types of the chicken using these same McAbs. We observed that although C-protein was present in a 43-nm repeat along the filaments in all three muscles, there were marked differences in the absolute number and position occupied by the different isoforms. Since McAbs MF-1 and ALD-66 decorated the same 43- nm repeats in the A-bands of the posterior latissimus dorsal muscle, we suggest that at least two C-proteins can co-localize at binding sites 43 nm apart along thick filaments of this muscle.     

C-protein from rabbit soleus (red) muscle.

A new form of skeletal-muscle C-protein has been isolated from rabbit soleus (red) muscle. This new form of C-protein has been purified to homogeneity by a procedure similar to that used to purify C-protein from white skeletal muscle. In soleus muscle, only this new form of C-protein could be detected, whereas in psoas (white) muscle, only the previously identified form of C-protein was detected. The content of C-protein in rabbit soleus muscle is comparable with that found in psoas muscle. Other rabbit skeletal muscles composed of a mixture of fibre types contained at least two forms of C-protein. C-Protein derived from red skeletal muscle bound to myosin isolated from either red or white tissue, with maximum binding occurring at a ratio of approximately 13 microgram of red C-protein/100 microgram of myosin. Polyacrylamide-gel electrophoresis in the presence of sodium dodecyl sulphate indicated that C-protein isolated from red skeletal muscle has a molecular weight approx. 7% greater than that of C-protein isolated from white skeletal muscle. The amino acid content of both forms of C-protein was similar but major differences in the mol % of isoleucine and threonine were found. Antiserum against C-protein from white rabbit skeletal muscle formed a single precipitin line with rabbit C-protein on double in agar. This antiserum did not form a precipitin line when diffused against red C-protein from rabbit skeletal muscle. Also, this antiserum bound specifically to the A-band region of myofibrils isolated from psoas (white) muscle, but it did not bind to myofibrils prepared from soleus (red) muscle.     

In situ reconstitution of myosin filaments within the myosin-extracted myofibril in cultured skeletal muscle cells

We studied the in situ reconstitution of myosin filaments within the myosin-extracted myofibrils in cultured chick embryo skeletal muscle cells using the electron microscope and polarization microscope. Myosin was first extracted from the myofibrils in glycerinated muscle cells with a high-salt solution containing 0.6 M KCl. When rabbit skeletal muscle myosin was added to the myosin-extracted cells in the high-salt solution, thin filaments in the ghost myofibrils were bound with myosin to form arrowhead complexes. Subsequent dilution of KCl in the myosin solution to 0.1 M resulted in the formation of thick myosin filaments within the myofibrils, increasing the birefringence of the myofibrils. When Mg-ATP was added such myosin-reassembled myofibrils were induced either to form supercontraction bands or to restore the sarcomeric arrangement of thick and thin filaments. Under the polarization microscope, vibrational movement of the myofibrils was seen transiently upon addition of Mg-ATP, often resulting in a regular arrangement of myofibrils in register. These myofibrils, with reconstituted myosin filaments, structurally and functionally resembled the native myofibrils. The findings are discussed with special reference to the myofibril formation in developing muscle cells.     

Involvement of transglutaminase in myofibril assembly of chick embryonic myoblasts in culture

Involvement of transglutaminase in myofibrillogenesis of chick embryonic myoblasts has been investigated in vitro. Both the activity and protein level of transglutaminase initially decreased to a minimal level at the time of burst of myoblast fusion but gradually increased thereafter. The localization of transglutaminase underwent a dramatic change from the whole cytoplasm in a diffuse pattern to the cross- striated sarcomeric A band, being strictly colocalized with the myosin thick filaments. For a brief period prior to the appearance of cross- striation, transglutaminase was localized in nonstriated filamental structures that coincided with the stress fiber-like structures. When 12-o-tetradecanoyl phorbol acetate was added to muscle cell cultures to induce the sequential disassembly of thin and thick filaments, transglutaminase was strictly colocalized with the myosin thick filaments even in the myosacs, of which most of the thin filaments were disrupted. Moreover, monodansylcadaverine, a competitive inhibitor of transglutaminase, reversibly inhibited the myofibril maturation. In addition, myosin heavy chain behaved as one of the potential intracellular substrates for transglutaminase. The cross-linked myosin complex constituted approximately 5% of the total Triton X-100- insoluble pool of myosin molecules in developing muscle cells, and its level was reduced to below 1% upon treatment with monodansylcadaverine. These results suggest that transglutaminase plays a crucial role in myofibrillogenesis of developing chick skeletal muscle.     

Phosphorylation of skeletal and cardiac muscle C-proteins by the catalytic subunit of cAMP-dependent protein kinase

Catecholamines are known to influence the contractility of cardiac and skeletal muscles, presumably via cAMP-dependent phosphorylation of specific proteins. We have investigated the in vitro phosphorylation of myofibrillar proteins by the catalytic subunit of cAMP-dependent protein kinase of fast- and slow-twitch skeletal muscles and cardiac muscle with a view to gaining a better understanding of the biochemical basis of catecholamine effects on striated muscles. Incubation of canine red skeletal myofibrils with the isolated catalytic subunit of cAMP-dependent protein kinase and Mg-[gamma-32P]ATP led to the rapid incorporation of [32P]phosphate into five major protein substrates of subunit molecular weights (MWs) 143,000, 60,000, 42,000, 33,000, and 11,000. The 143,000 MW substrate was identified as C-protein; the 42,000 MW substrate is probably actin; the 33,000 MW substrate was shown not to be a subunit of tropomyosin and, like the 60,000 and 11,000 MW substrates, is an unidentified myofibrillar protein. Isolated canine red skeletal muscle C-protein as phosphorylated to the extent of approximately 0.5 mol Pi/mol C-protein. Rabbit white skeletal muscle and bovine cardiac muscle C-proteins were also phosphorylated by the catalytic subunit of cAMP-dependent protein kinase, both in myofibrils and in the isolated state. Cardiac C-protein was phosphorylated to the extent of 5-6 mol Pi/mol C-protein, whereas rabbit white skeletal muscle C-protein was phosphorylated at the level of approximately 0.5 mol Pi/mol C-protein. As demonstrated earlier by others, C-protein of skeletal and cardiac muscles inhibited the actin-activated myosin Mg2+-ATPase activity at low ionic strength in a system reconstituted from the purified skeletal muscle contractile proteins (actin and myosin).(ABSTRACT TRUNCATED AT 250 WORDS)     

Connectin filaments link thick filaments and Z lines in frog skeletal muscle as revealed by immunoelectron microscopy

In an earlier study connectin, an elastic protein of striated muscle, was found to be associated with "gap filaments" originating from the thick filaments in the myofibril, but it was not clear whether it extends to Z lines or not (Maruyama, K., H. Sawada, S. Kimura, K. Ohashi, H. Higuchi, and Y. Umazume, 1984, J. Cell Biol., 99:1391-1397). In the present immunoelectron microscopic study using polyclonal antibodies against native connectin, we have concluded that the connectin structures are directly linked to Z lines from the thick (myosin) filaments in myofibrils of skinned fibers of frog skeletal muscle. There were five distinct antibody-binding stripes in each half of the A band and two stripes in the A-I junction region. Deposits of antibodies were recognized in I bands and Z lines. We suggest that connectin filaments run alongside the thick filaments, starting from a region approximately 0.15 micron from the center of the A band.     

Crystal structure of the C1 domain of cardiac myosin binding protein-C: implications for hypertrophic cardiomyopathy

C-protein is a major component of skeletal and cardiac muscle thick filaments. Mutations in the gene encoding cardiac C-protein [cardiac myosin binding protein-C (cMyBP-C)] are one of the principal causes of hypertrophic cardiomyopathy. cMyBP-C is a string of globular domains including eight immunoglobulin-like and three fibronectin-like domains termed C0-C10. It binds to myosin and titin, and probably to actin, and may have both a structural and a regulatory role in muscle function. To help to understand the pathology of the known mutations, we have solved the structure of the immunoglobulin-like C1 domain of MyBP-C by X-ray crystallography to a resolution of 1.55 Å. Mutations associated with hypertrophic cardiomyopathy are clustered at one end towards the C-terminus, close to the important C1C2 linker, where they alter the structural integrity of this region and its interactions.     

Influence of C-protein on mechano-chemical properties and structure of reconstituted actomyosin

C-protein on the mechano-chemical properties (ATPase activity and superprecipitation) of actomyosin systems has been investigated. The presence of C-protein in AM-complexes has been shown to decrease the rate of superprecipitation (SPP) and simultaneously increase the ATPase activity. Both effects of C-protein are dependent on its quantity in the system. Tropomyosin decreased considerably but does not eliminate completely the inhibitory influence of C-protein on the SPP. Electron microscopy does not reveal considerable structural differences in the initial AM-complexes depending on the presence or absence of C-protein. It is supposed that the discovered effects of C-protein on the behaviour of AM-systems are determined by the fine local structural and (or) charge changes produced by C-protein in the region of myosin cross-bridges, which in its turn results in a modification of the actin-myosin interaction. Possible participation of C-protein in the regulation of the interaction of thin and thick filaments in the muscle is discussed.     

Effects of C-protein on synthetic myosin filament structure.

In the absence of C-protein, synthetic filaments prepared from column-purified myosin exhibit the following features: individual filament diameters are uniform over a long length, but a wide distribution of diameters is apparent over the population; approximately 25% of the filaments have a frayed appearance and take up stain poorly, whereas the remaining 75% are well-stained; optical diffraction of well-stained filaments reveals a 14.3-nm subunit period and a 43-nm axial period (Koretz, 1978; Koretz, 1979). Addition of C-protein to myosin before filament formation affects all of these features in a manner related to C-protein concentration. At the physiological ratio of C-protein to myosin in the banded region of the natural thick filament, synthetic aggregates are uniform in diameter over the population and show less than 10% frays. Whereas the subunit period remains unchanged, the axial period has increased to 114.4 nm, or eight times the subunit repeat. Above and below the physiological ratio, disorder of a specific nature is apparent. Addition of C-protein after filament formation appears to coat the aggregates so that elements of backbone ultrastructure are obscured, and some evidence of axial period change is visible in diffraction patterns. A model is presented for the binding of C-protein to myosin, and its observed effects on filament structure are explained in terms of this model.     

H-protein and X-protein. Two new components of the thick filaments of vertebrate skeletal muscle

With a view to obtaining a more complete view of the composition and structure of the thick filaments of vertebrate skeletal muscle, we have isolated and characterized two new myofibrillar components, H-protein and X-protein. These were purified by hydroxyapatite column chromatography of an impure C-protein preparation itself made from impure myosin extracted from rabbit back and leg muscles. H-protein is the protein responsible for band H on sodium dodecyl sulphate/polyacrylamide gel electrophoresis of crude myosin. X-protein, although present in such preparations in significant quantities, was not detected previously since it is difficult to resolve from C-protein by sodium dodecyl sulphate/polyacrylamide gel electrophoresis. Physical-chemical parameters have been determined for the new proteins and compared with those of C-protein. The apparent chain weight of H-protein estimated by sodium dodecyl sulphate/polyacrylamide gel electrophoresis is 69,000, whereas that of X-protein (152,000) is only slightly greater than that of C-protein (140,000). The molecular weights of H- and X-proteins determined by sedimentation equilibrium centrifugation show that the molecules contain only a single polypeptide chain. The circular dichroism spectra indicate that the proteins have low alpha-helical contents. Both proteins, particularly H-protein, have a high proline content. Although X-protein is of similar chain weight to C-protein, the two show distinct differences in other properties. The sedimentation coefficient of X-protein is markedly lower than that of C-protein, suggesting X-protein is a more asymmetrical molecule. The amino acid compositions, although broadly similar, also show clear differences. Antibodies to H-protein, X-protein and C-protein have been raised in goats and shown not to cross-react.     

The myosin filament. X. Observation of nine subfilaments in transverse sections

The molecular packing of the subfilaments in muscle thick filaments has been investigated by electron microscopy. Thin (80-100 nm) transverse sections of vertebrate skeletal muscle were cut, and 129 electron microscope images of thick filaments from 15 different areas including seven to ten images in each area were analyzed by computer image processing. The transverse sections were limited to the portion of the filaments between the bare zone and the C-protein bearing region. Of the 129 images, six were discarded because they were structurally disrupted, 17 did not show evidence for the presence of subfilaments from the autocorrelation function, and four did not show evidence for three-fold rotational symmetry from the power spectrum. The remaining 102 filaments all showed evidence for three-fold rotational symmetry, consistent with other available evidence (Pepe, 1982). From the analysis of these images by rotational filtering, we have found that the vertebrate skeletal myosin filament is made up of nine subfilaments and that the image appears to have trigonal symmetry. Of these subfilaments, six are arranged with a center-to-center spacing of about 4 nm and the other three on the surface of the filament are distorted from this arrangement. Three additional densities, which together with the other nine, correspond to the pattern of 12 densities previously observed in more highly selected images (Stewart et al., 1981; Pepe and Drucker, 1972) were observed in 5% of the images. Another pattern of nine subfilaments peripherally arranged around the circumference of the filament was observed occasionally. This latter image may represent the organization of the subfilaments in the bare zone region of the filament, resulting from sampling of individual filaments displaced longitudinally relative to the other filaments in the A-band.     

3D structure of relaxed fish muscle myosin filaments by single particle analysis

To understand the structural changes involved in the force-producing myosin cross-bridge cycle in vertebrate muscle it is necessary to know the arrangement and conformation of the myosin heads at the start of the cycle (i.e. the relaxed state). Myosin filaments isolated from goldfish muscle under relaxing conditions and viewed in negative stain by electron microscopy (EM) were divided into segments and subjected to three-dimensional (3D) single particle analysis without imposing helical symmetry. This allowed the known systematic departure from helicity characteristic of vertebrate striated muscle myosin filaments to be preserved and visualised. The resulting 3D reconstruction reveals details to about 55 A resolution of the myosin head density distribution in the three non-equivalent head 'crowns' in the 429 A myosin filament repeat. The analysis maintained the well-documented axial perturbations of the myosin head crowns and revealed substantial azimuthal perturbations between crowns with relatively little radial perturbation. Azimuthal rotations between crowns were approximately 60 degrees , 60 degrees and 0 degrees , rather than the regular 40 degrees characteristic of an unperturbed helix. The new density map correlates quite well with the head conformations analysed in other EM studies and in the relaxed fish muscle myosin filament structure modelled from X-ray fibre diffraction data. The reconstruction provides information on the polarity of the myosin head array in the A-band, important in understanding the geometry of the myosin head interaction with actin during the cross-bridge cycle, and supports a number of conclusions previously inferred by other methods. The observed azimuthal head perturbations are consistent with the X-ray modelling results from intact muscle, indicating that the observed perturbations are an intrinsic property of the myosin filaments and are not induced by the proximity of actin filaments in the muscle A-band lattice. Comparison of the axial density profile derived in this study with the axial density profile of the X-ray model of the fish myosin filaments which was restricted to contributions from the myosin heads allows the identification of a non-myosin density peak associated with the azimuthally perturbed head crown which can be interpreted as a possible location for C-protein or X-protein (MyBP-C or -X). This position for C-protein is also consistent with the C-zone interference function deduced from previous analysis of the meridional X-ray pattern from frog muscle. It appears that, along with other functions, C-(X-) protein may have the role of slewing the heads of one crown so that they do not clash with the neighbouring actin filaments, but are readily available to interact with actin when the muscle is activated.