Myosin filaments in vertebrate smooth muscle

Myosin and Actin Filaments in Muscle: Structures and Interactions

In the last decade, improvements in electron microscopy and image processing have permitted significantly higher resolutions to be achieved (sometimes <1 nm) when studying isolated actin and myosin filaments. In the case of actin filaments the changing structure when troponin binds calcium ions can be followed using electron microscopy and single particle analysis to reveal what happens on each of the seven non-equivalent pseudo-repeats of the tropomyosin α-helical coiled-coil. In the case of the known family of myosin filaments not only are the myosin head arrangements under relaxing conditions being defined, but the latest analysis, also using single particle methods, is starting to reveal the way that the α-helical coiled-coil myosin rods are packed to give the filament backbones.

X-ray diffraction study on mammalian visceral smooth muscles in resting and activated states

Structural changes of guinea pig taenia coli and rat anococcygeus muscle during contraction were studied by X-ray diffraction. The diffraction pattern of the taenia coli showed the 14.4-nm myosin reflection, the 5.9-nm actin layer-line and a diffuse equatorial peak at 1/11.4 nm-1. On application of carbachol, the muscle contracted and the intensity of the 14.4-nm reflection showed a concentration-dependent decrease: the maximum decrease was 24% at 2 x 10(-5) M. Such an intensity decrease was not observed in K-contracture (154 mM). The intensity of the 5.9-nm actin layer-line did not change appreciably on activation. The equatorial peak became broader during contraction. The 14.4-nm myosin reflection of the anococcygeus muscle was weak. Its intensity increased by 106% during contraction induced by 2 x 10(-5) M phenylephrine and by 75% during K-contracture. These results suggest that the number of myosin filaments may increase during contraction of rat anococcygeus muscle but not guinea pig taenia coli.

X-ray studies of order-disorder transitions in the myosin heads of skinned rabbit psoas muscles

Using x-rays from a laboratory source and an area detector, myosin layer lines and the diffuse scattering between them in the moderate angle region have been recorded. At full overlap, incubation of rigor muscles with S-1 greatly reduces the diffuse scattering. Also, three of the four actin-based layer lines lying close to the meridian (Huxley, H. E., and W. Brown, 1967. J. Mol. Biol. 30:384-434; Haselgrove, J. C. 1975. J. Mol. Biol. 92:113-143) increase, suggesting fuller labeling of the actin filaments. These results are consistent with the idea (Poulsen, F. R., and J. Lowy, 1983. Nature [Lond.]. 303:146-152) that some of the diffuse scattering in rigor muscles is due to a random mixture of actin monomers with and without attached myosin heads (substitution disorder). In relaxed muscles, regardless of overlap, lowering the temperature from 24 to 4 degrees C practically abolishes the myosin layer lines (a result first obtained by Wray, J.S. 1987. J. Muscle Res. Cell Motil. 8:62 (a). Abstr.), whilst the diffuse scattering between these layer lines increases appreciably. Similar changes occur in the passage from rest to peak tetanic tension in live frog muscle (Lowy, J., and F.R. Poulsen. 1990. Biophys. J. 57:977-985). Cooling the psoas demonstrates that the intensity relation between the layer lines and the diffuse scattering is of an inverse nature, and that the transition occurs over a narrow temperature range (12-14 degrees C) with a sigmoidal function. From these results it would appear that the helical arrangement of the myosin heads is very temperature sensitive, and that the disordering effect does not depend on the presence of actin. Measurements along the meridian reveal that the intensity of the diffuse scattering increases relatively little and does so in a nearly linear manner: evidently the axial order of the myosin heads is much less temperature sensitive. The combined data support the view (Poulsen, F. R., and J. Lowy. 1983. Nature [Lond.]. 303:146-152) that in relaxed muscles a significant part of the diffuse scattering originates from disordered myosin heads. The observation that the extent of the diffuse scattering is greater in the equatorial than in the meridional direction suggests that the disordered myosin heads have an orientation which is on average more parallel to the filament axis.

Studies of the diffuse x-ray scattering from contracting frog skeletal muscles

Using x-rays from synchrotron radiation, we studied diffuse scattering, sometimes together with the myosin layer lines. With an area detector, sartorius muscles and a time resolution of 150 ms, earlier results from semitendinosus muscles contracting isometrically at 6 degrees C (Lowy, J., and F. R. Poulsen. 1987. J. Mol. Biol. 194:595-600) were confirmed and extended. Evidence from intensity changes both in the diffuse scattering and in the myosin layer lines showed that the majority of the heads become disordered at peak tetanic tension. With a linear detector and a time resolution of 5 ms, it was found that during tension rise the intensity increase of the diffuse scattering (which amounted maximally to 12% recorded near the meridian) runs approximately 20 ms ahead of the mechanical change, comparing half-completion times. This suggests that an appreciable number of heads change orientation before peak tension is reached. In quick release experiments the diffuse scattering intensity showed very little change. Recorded near the meridian during rapid shortening, however, it decreased progressively with a half-time of approximately 40 ms. This change amounted to approximately 35% of that observed during the initial tension rise. We interpret this to indicate that during rapid shortening a certain number of heads assume an orientation characteristic of the relaxed state. Viewed in the context of the behavior of the first myosin layer line and the (1, 1) equatorial reflection in similar experiments (Huxley, H. E., M. Kress, A. R. Faruqi, and R. M. Simmons. 1988. Molecular Mechanism of Muscle Contraction), the present results provide further support for the view that the diffuse scattering is mostly due to disordered myosin heads; whilst ordered heads produce the myosin layer lines (Poulsen, F. R., and J. Lowy.1983. Nature lLond.l. 303:146-152).

Diffuse X-ray scatter from myosin heads in oriented synthetic filaments

X-ray results are presented concerning the structural state of myosin heads of synthetic filaments in threads. These were made from purified rabbit skeletal muscle myosin and studied by x-ray diffraction and electron microscopy by Cooke et al. (Cooke, P. H., E. M. Bartels, G. F. Elliott, and R. A. Hughes, 1987, Biophys. J., 51:947-957). X-ray patterns show a meridional peak at a spacing of 14.4 nm. We concentrate here on the only other feature of the axial pattern: this is a central region of diffuse scatter, which we find to be similar to that obtained from myosin heads in solution (Mendelson, R. A., K. M. Kretzschmar, 1980, Biochemistry, 19:4103-4108). This means that the myosin heads have very large random displacements in all directions from their average positions, and that they are practically randomly oriented. The myosin heads do not contribute to the 14.4-nm peak, which must come entirely from the backbone. Comparison with x-ray data from the unstriated Taenia coli muscle of the guinea pig indicates that in this muscle at least 75% of the diffuse scatter comes from disordered myosin heads. The results confirm that the diffuse scatter in x-ray patterns from specimens that contain myosin filaments can yield information about the structural behavior of the myosin heads.

The repeat distance of myosin in the thick filaments of various muscles

The meridional spacing of the X-ray diffraction peak from the repeat of myosin along the thick filament of four muscles has been remeasured on the same apparatus. The frog sartorius gave a shorter repeat distance (143.7 A) than the three invertebrate muscles, which ranged from 144.9 to 145.4 A. These results confirm earlier measurements. Provided that the myosin molecules are staggered relative to one another by a constant 98 residues, it may be inferred that in vertebrate thick filaments part or all of the tail lies at a considerable angle to the filament axis, whereas in the invertebrates the angle is smaller.

Mechanical properties of smooth muscle cells in the walls of arterial resistance vessels

1. Methods have been developed for measuring the dynamic mechanical response of arterial resistance vessels (i.d. 83--235 micrometer) with a time resolution of about 4 msec. 2. Observations of the microscope image of the smooth muscle cells in the walls of these vessels indicate that there is little intercellular compliance in this preparation, and that the mechanical properties of the activated preparation are a reflexion of the mechanical properties of the individual smooth muscle cells. 3. Under isometric conditions the force developed per unit cell area was about 350 mN/mm2. Under isotonic conditions the cells had a maximum velocity for shortening at 37 degrees C of about 0.17 lengths/sec. 4. Quick releases of activated vessels indicate that the instantaneous elastic characteristic of smooth muscle cells is approximately exponential. 5. The wall tension response to small (0.3%) square wave changes in circumference was proportional to the logarithm of the time following the start of each circumference change. 6. Active wall tension, deltaT, was varied by varying the Ca2+ concentration of the activating solution. Under these conditions the active dynamic stiffness, k, was proportional to deltaT, and was not temperature dependent. The active half response time, tau (the time, taken to recover half the tension change caused by a small change in circumference) was also proportional to deltaT, but here the constant of proportionality had a Q10 of about 1.8. 7. It is concluded that the quick release response and the square wave response are in part a function of the mechanical properties of the crossbridges between the contractile filaments. Calculations show that both these responses can be explained if it is assumed that there is a relatively compliant passive component in series with the crossbridges.

Dynamic stiffness of rabbit mesotubarium smooth muscle: effect of isometric length

The dynamic stiffness of mesotubarium smooth muscle from nonpregnant adult rabbits was measured continuously during isometric contraction by applying small (0.5 percent of the muscle length) sinusoidal length perturbations and measuring the amplitude and phase of the resulting tension perturbations. Stiffness during contraction was directly proportional to muscle tension; during relaxation stiffness at all tensions was significantly increased as compared to the values encountered during the rise of tension. Peak isometric tension and dynamic stiffness (determined at a common tension level) both decreased at shorter muscle lengths; the relative falloff in stiffness was significantly less than the tension decrease. Varying levels of muscle activation (obtained by changing stimulus strength and by applying quick releases to active muscle) had little effect on the measured elastic modulus. Comparisons of these results with published data on single-cell contractile properties imply a cellular locus for a portion of the measured stiffness.

[Intraepithelial filament bundles of the adult human kidney. Studies by light and electron microscopy and by diffraction analysis (author's transl)]

In studies of semi-thin sections of adult human kidney it is found that the structures formerly known as "Basalreifen" can be definitely related to basal cytoplasmic portions of the renal epithelium. Under the electron microscope they appear as bundles of filaments, usually having a width of 50-85 A. They are particularly well developed in the parietal layer of Bowman's capsule and are also present in the proximal and distal tubule where they are more abundant than in other parts of the nephron. The filament bundles proceed toward hemidesmosome-like structures at the basal border of the epithelia and sometimes contain dense bodies in the parietal capsular epithelium of the glomerulus; these bundles are probably responsible for local intra- and/or supravital protrusions of the cytoplasm. Findings in some diseased kidneys are also presented. Diffraction analysis of intrarenal vascular muscle cells and intraepithelial filament bundles gave comparable photographic records of the image structure spectrograms; it supports the assumption derived from the electron microscopic results that the intraepithelial filament bundles represent a contractile complex. However, the conclusiveness of image structure analysis in comparing the two biological objects is limited, as outlined in the discussion. With regard to renal function the significance of the intraepithelial filament bundles is still unclear. Periglomerular interstitial cells, which normally contain only narrow tracts of filaments in the periphery, had many of the characteristics of smooth muscle cells in a case of pyelonephritic renal cirrhosis. They are compared with myofibroblasts of a different origin.

Thick filaments in vertebrate smooth muscle

Smooth muscle cells of the mouse vas deferens fixed with 5% glutaraldehyde contained three types of filaments, namely, thin (50-80) A) filaments, intermediate (100 A) filaments and thick (120-180 A) filments. However, in 2 out of 16 experiments, under identical conditions, the cells did not contain thick filaments. With OSO4 fixation, thin filaments were not prominent, the most obvious being thick (120-250 A) and intermediate (100 A) filaments. After soaking in a modified Ringer solution under no applied tension for one hour, thick filaments (120-180 A) appeared prominently in smooth muscle cells of the mouse vas deferens and thin filaments were in ordered bundles. By 4 hours, thick filaments had increased in size and density, with thin filaments distributed randomly around them. After 8 hours in Ringer, thin filaments were diffuse and difficult to discern, while thick filaments were large (up to 300 A) and electron-dense. Intermediate (100 A) filaments were present in association with dark bodies. Physiological experiments indicated that the intracellular components responsible for the development of a mechanical response were still functional at this time. The presence of "thick filaments" is also reported in degenerating smooth muscle cells of the guinea-pig vas deferens in tissue culture.

Potassium accumulation in smooth muscle and associated ultrastructural changes

1. The water content, extracellular ((60)CoEDTA) space, ionic composition and ultrastructure of several mammalian smooth muscles were studied after incubation in solutions of varying ionic compositions and osmolarities.2. Substitution of KCl for NaCl resulted in an increase in cell water, K and Cl, accompanied by little change in total wet weight. This was due to a reduction in the extracellular space.3. Changes in extracellular osmolarity produced a wider range of cell volumes in high KCl solutions than in Krebs. The addition of 29-58 mM sucrose to high KCl prevented the swelling.4. Electron microscopy of smooth muscle swollen in high KCl solution revealed light (less electron opaque than normal) fibres of increased diameter, reduction in extracellular space, and nuclear swelling. The normal thick filament lattice was destroyed in swollen, osmium-fixed smooth muscles.5. The ultrastructural changes ascribed to swelling were absent in smooth muscles, (a) depolarized in high K(2)SO(4) solutions, (b) in high KCl solutions with 29-58 mM sucrose, and (c) returned to normal Krebs solution for recovery from swelling.6. Smooth muscles incubated in high KCl (swollen) and high K(2)SO(4) (unswollen) exhibited similar contractile responses, suggesting the filament lattice was intact until fixation, and that the contractile mechanism can operate over a relatively wide range of actin to myosin separations.7. Shrinkage of smooth muscles in high KCl solutions made hypertonic with the addition of 10% sucrose was accompanied by an aggregation of the thick filaments.8. The cell water of fixed taenia coli was reduced (a) by incubation in hypertonic solution followed by fixation in normal glutaraldehyde, or (b) by fixation of normal tissue in hypertonic glutaraldehyde. Osmotic responses during aldehyde fixation may be a source of artifact in the visualization of the normal filament lattice.

Thick filaments in vascular smooth muscle

Two sets of myofilaments were demonstrated after incubation of strips of rabbit portal-anterior mesenteric vein under moderate stretch in a physiological salt solution. Thick filaments had a mean diameter of 18 nm and reached a maximum length of 1.4 microm with a mean length of 0.61 microm. In transverse sections, 2.5-5 nm particles were resolved as subunits of the thick filaments. Thin filaments had an average diameter of 8.4 nm and generally conformed to the structure believed to represent actin filaments in smooth and striated muscles. In the areas of maximum concentration there were 160-328 thick filaments/microm(2) and the lowest ratio of thin to thick filaments was 12:1. Thick filaments were present in approximately equal numbers in vascular smooth muscle relaxed by theophylline, in Ca(++)-free solution, or contracted by norepinephrine. The same preparatory procedures used with vascular smooth muscle also enabled us to visualize thick filaments in guinea pig and rabbit taenia coli and vas deferens.

Myosin-like aggregates in trypsin-treated smooth muscle cells

Segments of the lower small intestine of the toad Bufo marinus were excised and soaked for approximately 2 hr in Ringer's solution (pH 7.4 or 7.8) containing crystalline trypsin and then fixed for electron microscopy at approximately the same pH. Thin sections of the tunica muscularis of these specimens show smooth muscle cells ranging in appearance from severely damaged at one extreme to apparently unaffected at the other. Among these are cells at intermediate stages, including some which exhibit large and conspicuous populations of thick filaments closely resembling artificially prepared aggregates of smooth muscle myosin. The thick filaments have the form of tactoids approximately 250-300 A in diameter in their middle regions and are approximately 0.5-1.0 micro in length. In some preparations they also display an axial periodicity approximating 143 A. They are usually randomly oriented and segregated from the thin filaments, which tend to form closely packed, virtually crystalline bundles at the periphery of these cells. "Dense bodies" are absent from cells showing these changes. The simplest interpretation of these data is that smooth muscle myosin normally exists among the actin filaments in a relatively disaggregated state and that trypsin induces aggregation by altering the conformation of the myosin molecule. Alternatively, trypsin may act indirectly through an effect on some other smooth muscle protein which normally forms a stable complex with relatively disaggregated myosin.

The organization of contractile filaments in a mammalian smooth muscle

Ordered arrays of thin filaments (65 A diameter) along with other apparently random arrangements of thin and thick filaments (100-200 A diameter) are observed in contracted guinea pig taenia coli rapidly fixed in glutaraldehyde. The thin-filament arrays vary from a few to more than 100 filaments in each array. The arrays are scattered among isolated thin and thick filaments. Some arrays are regular such as hexagonal; other arrays tend to be circular. However, few examples of rosettes with regular arrangements of thin filaments surrounding thick filaments are seen. Optical transforms of electron micrographs of thin-filament arrays give a nearest-neighbor spacing of the thin filaments in agreement with the "actin" filament spacing from x-ray diffraction experiments. Many thick filaments are closely associated with thin-filament arrays. Some thick filaments are hollow circles, although triangular shapes are also found. Thin-filament arrays and thick filaments extend into the cell for distances of at least a micron. Partially relaxed taenia coli shows thin-filament arrays but few thick filaments. The suggestion that thick filaments aggregate prior to contraction and disaggregate during relaxation is promoted by these observations. The results suggest that a sliding filament mechanism operates in smooth muscle as well as in striated muscle.