J. Lowy

The structure of F-actin and of actin filaments isolated from muscle

The filaments of the contractile apparatus have been isolated from a wide variety of smooth and striated muscles and examined in negatively-stained preparations in the electron microscope. In all cases there are thin filaments which are indistinguishable from the filaments in F-actin preparations. The filament consists of two helically-wound strands composed of subunits which appear to be alike and approximately spherical. The arrangement of the subunits corresponds to that of the scattering centres in one of the possible structures for actin deduced by Selby & Bear (1956) from the moderate-angle X-ray diffraction pattern of intact dried muscle. The number of globular subunits per turn of the helix (if integral) is 13 (cf. either 13 or 15 in the models proposed from diffraction data). The spacing of the subunits along each strand is 56·5 A (cf. 55 A in the models proposed from diffraction data). The cross-over points of the two twisted strands are spaced at intervals of 349 A along the filament (cf. either 351 A or 406 A in the models proposed from diffraction data). The overall diameter of the filament is about 80 A. It is shown in the case of rabbit skeletal muscle that this result is consistent with the quantity of actin in that muscle. There is good evidence that each of the globular subunits seen in the electron microscope represents one actin monomer. The structure of actin alone does not account for (i) the approximately 400 A axial periodicity observed in the I-substance of fibrils in the electron microscope, or (ii) the reflection at about 400 A observed in the axial diffraction pattern of intact muscle. It is suggested that both (i) and (ii) could be due to the combination of actin with other material, possibly tropomyosin B.

Low-angle X-ray diffraction studies of living striated muscle during contraction

Abstract The spacings and intensities of the small-angle equatorial X-ray reflexions from living toad sartorius muscle were studied in the resting and contracted state. On contraction, there is a small and nearly constant decrease of 6 to 12 A in the lattice spacing over the whole range of muscle lengths studied and the 1,1 reflexion becomes slightly more intense relative to the 1,0 reflexion. In both resting and contracting muscle, the centre-to-centre distance between actin and myosin filaments depends on sarcomere length, being about 80 A larger in the short muscles (sarcomere length 2.1 μ) than in the long ones (sarcomere length 3.6 μ). In another set of experiments with the toad sartorius muscle, it was found that on contraction no change occurred in the spacings of the meridional X-ray reflexions from either the actin or the myosin filaments. We have also observed a decrease in intensity of the myosin layer lines (given by the projections on the myosin filaments) as first reported by Huxley, Brown & Holmes (1965). By indicating that there is no change in length in the actin or myosin filaments during contraction, these results provide general support for the sliding filament hypothesis.

ELECTRON MICROSCOPE STUDIES OF BACTERIAL FLAGELLA.

Flagella of Pseudomonas fluorescens, Pseudomonas rhodos, Proteus vulgaris, Salmonella typhimurium and Bacillus subtilis were prepared for electron microscopy by several negative-contrast methods. Flagella attached to the cell and short detached pieces all gave the same results. Two types of structure (A and B) and a possible intermediate form were found; sometimes ( Ps. rhodos ) the structure changes along a single flagellum. The A structure shows helically connected globules aligned in longitudinal rows; sometimes there are also indications of longitudinal connexions. The B structure shows neither globules nor helices, but has thick longitudinal lines, the number of which in a given species is the same as the number of rows of globules in the A form. The globules in type A flagella alternate in adjacent longitudinal rows and are connected helically so that the number per turn of each helix equals the number of longitudinal rows. In the models proposed, this number is 10 in Ps. fluorescens , but 8 in all the other species examined. The spacing of the globules along each row is about 50 A. This could account for the reflexions at 25·6 A, etc., on the meridian of diffraction patterns ( Astbury, Beighton & Weibull, 1955 ). The significance of the finding ( Astbury, Beighton & Weibull, 1955 ) that the reflexions from Proteus flagella indicate a 410 A periodicity (as in the act in-containing filaments of muscle) is discussed in the light of our results which indicate that the pitch of the globule helix is about 200 A in Proteus , and about 250 A in Ps. fluorescens . The question whether flagella, like muscles, might be two-component contractile systems is considered. It is interesting that helices of globular units (actin molecules), comparable in size to the globules in flagella, are present in all kinds of muscle. It remains to be shown if the structural changes found in the globule system in flagella have anything to do with contraction.

THE STRUCTURE OF ACTIN FILAMENTS AND THE ORIGIN OF THE AXIAL PERIODICITY IN THE I-SUBSTANCE OF VERTEBRATE STRIATED MUSCLE.

Certain advances due mainly to H. E. Huxley (see Huxley 1961, 1963) have made it possible to use the electron microscope to study the detailed structure of the filaments in the contractile apparatus. The results of our work on actin filaments have already been published (Hanson & Lowy 1962, 1963). We shall now examine some of the consequences of these findings, including certain unsolved problems which they raise. Actin in the polymerized form (F-actin) has been prepared from rabbit skeletal muscle by the usual methods and examined in negatively stained preparations in the electron microscope (Hanson & Lowy 1963). It has been found that solutions of F-actin are, in fact, suspensions of filaments. These consist of globular subunits arranged in a characteristic helical manner (figure 15).

Structure and Function in Smooth Tonic Muscles of Lamellibranch Molluscs

Muscles like the smooth adductors of lamellibranch molluscs, and the anterior byssus retractor of Mytilus (ABRM) possess two remarkable features which have been studied intensively during recent years. They can maintain a steady level of tension (tonic contraction) for longer than any other muscle type (see Marceau 1909), and their contractile apparatus contains, in addition to the usual actomyosin, large quantities of a protein (tropomyosin A—TMA) apparently not found in vertebrate striated muscle (Bailey 1956; Kominz, Saad & Laki 1957). As in vertebrate striated muscle, the contractile apparatus of these smooth molluscan muscles contains two kinds of filaments, thin ones and thicker ones (Hanson & Lowy 1959; Philpott, Kahlbrock & Szent-Györgyi 1960). In both muscle types the thin filaments are of similar diameter and contain actin; but the diameter of the thick filaments is about ten times greater in the molluscan muscles, and it is these filaments which contain TMA (see Lowy & Hanson 1962). When large quantities of TMA were found in these molluscan muscles, their outstanding capacity for tonic contraction was not unnaturally connected with the presence of that protein (Bailey 1957). A hypothesis was proposed according to which there exist in such muscles two independent systems acting in parallel: the usual actomyosin system which produces tension, and a unique system (TMA filaments) which, by becoming rigid, can maintain this tension with negligible energy expenditure, after the actomyosin system has relaxed (Rüegg 1958; Johnson, Kahn & Szent-Györgyi 1959). The evidence in support of this ‘independent catch hypothesis’ is based on studies with preparations of TMA, on experiments with glycerol-extracted muscles, and on certain experiments with the living ABRM treated with thiourea. Results obtained from structural studies and from extensive investigations on living tonic molluscan muscles have suggested an alternative hypothesis, the ‘linkage hypothesis’. This holds that, as in vertebrate striated muscle, tension in tonic molluscan muscles is both generated and maintained by the same system, which is one that involves interaction between adjacent actin- and myosin-containing filaments in a sliding filament system (Lowy & Millman 1959b, 1963).

Low-angle X-ray reflections from living molluscan muscles

A method is described for obtaining low-angle X-ray reflections from living muscles, checking the state of the muscle throughout the exposure. In the present work certain molluscan muscles were used in a relaxed condition. X-ray diffraction patterns of the muscles when alive showed vestiges of an equatorial streak, together with a single clear equatorial reflection at ∼ 120 A. When the muscle died the clear reflection disappeared, and the specimens then gave only the equatorial streak. Several meridional reflections were observed from all the muscles used, regardless of whether they were alive and relaxed or dead. In the interpretation of the results it is suggested that both the equatorial and the meridional reflections come from diffracting systems situated in the paramyosin filaments, and that the equatorial reflection comes from the planes of ribbon-shaped sub-filaments stacked face-to-face within the paramyosin filaments and separated by 120 A. The finding that the meridional reflections are present in relaxed muscles is discussed in relation to the problem of the prolonged maintenance of tension in the molluscan muscles studied.

Proteins of the Contractile Mechanism of Mammalian Smooth Muscle and their Possible Location in the Cell [and Discussion]

Dorothy M. Needham speaking. Since the pioneer work of Csapo and his colleagues, beginning about fifteen years ago, it has been realized that from uterine smooth muscle can be extracted a protein closely resembling skeletal-muscle actomyosin in its viscous behaviour, sedimentation rate and electrophoretic mobility. (See, for example, Csapo 1948, 1949, 1950, 1959; Csapo, Erdos, Naeslund & Snellman 1950; Naeslund & Snellman 1951). Later work, in which the properties of purified preparations of myosin, actin and actomyosin have been studied, bears out these earlier conclusions. Thus, for example, we have shown (Needham & Williams 1963b) that skeletal-muscle myosin will react normally with uterus actin to give the highly viscous actomyosin; and similarly uterus myosin with skeletal-muscle actin. In both types of experiment the results indicated that the two proteins associated together in about the same proportions as when both are derived from skeletal muscle. Uterus actomyosin may be fragmented by carefully controlled trypsin treatment giving light and heavy meromyosins which, so far as they have been studied, show similar properties to the meromyosins from skeletal-muscle actomyosin (Needham & Williams 1959; Cohen, Lowey & Kucera 1961). Smooth muscle, however, does contain very strikingly less actomyosin than striated muscle, only 6 to 10 mg/g wet wt as compared with about 70 mg/g wet wt in skeletal muscle (Needham & Williams 1963 a).