A review of progress in a field as broad as cell motility will be a selection of topics that reflects the interest of the reviewer. I will restrict the subject even further by concentrating on a single problem, the progress in our understanding of the physicalchemical basis of motility. One measure of progress in cell biology is the extent to which the description of a phenomenon is replaced by a chemical mechanism. An understanding of the mechanisms of individual processes studied in isolation still comes far short of a science of cell biology, which must deal with the integration of processes into a description of the cell as a functional unit. Considerable progress has been made in the understanding of individual processes but their integration in the cell remains a problem to be studied over the next 20 years.

Studies on a wide variety of cells and tissues have led to the important concept that the vast majority of the examples of motile behaviour can be accounted for by a small number of basic mechanisms for the conversion of chemical energy into mechanical work. Although there are a few exceptions, such as spasmonemes (Amos, 1975) and myonemes (Huang & Pitelka, 1973), motile behaviour can generally be ascribed to the action of actin and myosin or microtubules plus an enzyme such as dynein or kinesin. In both cases the source of energy is the hydrolysis of ATP. The fundamental questions are how is chemical energy converted into force or movement and whether the underlying mechanisms are similar in the different systems.

Striated muscle provides the best model for the understanding of an actomyosin-based motile system and the concepts developed in the studies of muscle have been the basis for an understanding of motility in non-muscle cells. The major advantage of striated muscle is the high degree of order, which provides the possibility of determining the structural change that leads to force development, and of correlating the structural change with biochemical steps in the enzyme mechanism. However, much of the structure is a specialization for the particular problem for which the muscle was designed and we wish to extract from the mechanism the properties of the minimum motile system that can serve as the basis for cell functions.

Myosins from a variety of organisms ranging from amoeba to mammals have similar properties and it is reasonable to infer that the energy conversion mechanism is essentially the same in all organisms. A possible exception is the presence in some amoebae of two types of myosin, the normal two-headed myosin and a second myosin having a single head (Pollard & Korn, 1973).

Progress in the study of the mechanism of muscle contraction has been the subject of numerous reviews (Goody & Holmes, 1982; Webb & Trentham, 1983). The structural elements of the sarcomere are an array of actin filaments of the same polarity in each half sarcomere and a bipolar myosin thick filament. The mechanism can be discussed in terms of a single myosin cross-bridge and a single actin filament, since the bridges appear to act independently as tension generators (Huxley & Simmonds, 1971). A mechanochemical scheme was put forward in the early seventies based on structural, mechanical and biochemical evidence (Huxley, 1969; Huxley & Simmonds, 1971; Lymn & Taylor, 1971). It was proposed that the acto-myosin complex (AM) could exist in two states in which the cross-bridge head is bound to actin at an angle of 45° or 90°. Rotation of the head from the 45° to the 90° (rigor) orientation produces a relative sliding of the filaments or stretches a spring in the myosin molecule if the filaments are held at fixed length. The three important properties of the enzyme reaction are: (1) nucleotide binding to AM alters the conformation, leading to a very rapid dissociation of the protein complex, AM + ATP ⇀ A + M-ATP; (2) the nucleotide is hydrolysed fairly rapidly by myosin but the products dissociate slowly from the enzyme site, M-ATP’— M ADP Pi ⇀ M + ADP ⇀ Pi; (3) actin activates myosin ATPase by increasing the rate of product dissociation, AM-ADPP, —AM + ADP + Pi. It was proposed that reaction (1) corresponds to the dissociation of the cross-bridge at the end of the forcegenerating step, that the hydrolysis step alters the structure of the free bridge in reaction (2) so that it binds at a 90° orientation, and that release of products in reaction (3) is coupled to rotation to the 45° state to complete the cycle. This simple model was important in the development of the subject since it appeared to provide a simple explanation of the structural changes in the contraction cycle and to answer the question of how ATP hydrolysis could be coupled to movement.

During the last dozen years extensive studies have been devoted to testing the assumptions and predictions of this model. It is evident that the actual mechanism is more complex. Studies by X-ray diffraction of muscle and by spectroscopy using fluorescence and spin-labels (Huxley et al. 1983 : Yanigita, 1981 ; Cooke, Crowder & Thomas, 1982; Burghardt, Ando & Borejdo, 1983) have failed to provide clear evidence for a 90° orientation of the cross-bridge. The results of these studies are not in agreement and in only one case (Burghardt et al. 1983) has any difference in angle been obtained for active versus rigor muscle. The large rotation was based on the mechanical evidence that the cross-bridge range was of the order of 10 nm and the head was about 10 nm long; consequently, a 45° rotation is required. A 90° orientation of the cross-bridge was also inferred from the diffraction pattern of relaxed muscle (Huxley & Brown, 1967), but this interpretation has not been confirmed. Recent evidence from crystals of myosin (Winkelmann, Mekeel & Rayment, 1985) indicates that the head has a length of 16 nm. The minimum range of the elastic element is 4nm; larger ranges are model-dependent conclusions (Ford, Huxley & Simmonds, 1977). Consequently, the actual rotation may be only 15° to 20° (Huxley & Kress, 1985) and it has so far escaped detection. The head is also considerably bent in shape and may not rotate as a rigid body attached to a spring. The head may bend in the region distal to the sites to which fluorescence or spinlabels have been attached since these labels generally fail to detect any rotation of the head. In this case the spring is part of the globular head. A satisfactory answer has not been obtained to the question of whether the head rotates or bends and by how much.

The biochemical model has also become more complex with the finding of additional intermediate states in the mechanism. More important is the evidence that the equilibrium constant of the hydrolysis step is small, and that hydrolysis in solution can occur without dissociation of actomyosin at a comparable or somewhat slower rate than for myosin (Stein, Chock & Eisenberg, 1984; Rosenfeld & Taylor, 1984). The weakly bound M-ATP and M-ADP-P; states attach and detach rapidly on the time scale of the cycle, hence they are essentially in equilibrium with the corresponding bound states. The same situation is probably the case in muscle. Thus, the reaction cannot be described by a simple cycle. In its simplest form the reaction is :
formula
where, T, D and P refer to ATP, ADP and inorganic phosphate, respectively.

In the absence of structural evidence it is not clear how the biochemical states are to be assigned to cross-bridge orientations. A better model of the cycle is to regard the weakly attached states, M-T and M-D-P, as a pool of bridges that are in rapid equilibrium with the corresponding attached states. Since these states can change actin sites rapidly they exert negligible force and whether they have the same or different orientation has little effect on the properties of the system. The AM-D-P state is still equivalent to the 90° state of the original model and the strongly bound states AM-D and AM correspond to the 450 orientation (Eisenberg & Greene, 1980). The behaviour is illustrated in Fig. 1. The cross-bridge enters the pool by the rapid detachment by ATP. The rate of transit through the pool is determined by the rate of the hydrolysis step and the attached bridge leaves the pool by the transition from AM-D-P to AM-D with a change in orientation or deformation. The model retains most of the features of the original scheme but we have no information on the relative orientation of the weakly bound states. Consequently, we can no longer assign a change in the orientation of binding to a structural effect of the hydrolysis step.

A problem with contraction models is that the significance of two heads has not been explained. It is generally assumed that the heads act independently. A singleheaded myosin prepared by proteolytic digestion can form threads with actin that exert the same tension per head as normal myosin (Cooke & Franks, 1978). The single-headed myosin of Acanthamoeba produces bead movement in the model system described below (Albanesi et al. 1985). Although the evidence suggests that a single-headed myosin is sufficient to produce movement, the efficieney of such systems is unknown. The special requirements of muscle for high energy-conversion efficiency and the development of large forces may necessitate a more subtle mechanism involving interaction of the two heads. At present we do not understand why myosin and dynein have at least two heads.

THE MICROTUBULE-DYNEIN SYSTEM OF CILIA

Cilia and flagella are discussed by Brokaw in this volume but we are concerned with the relationship between the mechanisms of force generation in the microtubule—dynein and actomyosin systems. This problem has been the subject of a recent review (Johnson, 1985) and only the main conclusions need be repeated here.

Dynein, isolated from various organisms, is a two-or three-headed molecule. It consists of two (or three) heavy chains with molecular weights in the 300 000 Mr range, two (or three) intermediate chains of 70000 to 100000 Mr and three or four light chains in the 15 000 to 20000 Mr range. The globular heads are connected by flexible strands to a root-like base that appears to bind to the A tubule. The heads interact with the B tubule in the enzyme cycle. Thus, dynein has a superficial resemblance to myosin and the heads may undergo a cycle of attachment, movement and detachment.

The three properties of actomyosin ATPase discussed in the last section are also properties of the microtubule-dynein system. The heads are rapidly detached from the microtubule by the binding of ATP, the free dynein has a relatively rapid hydrolysis step followed by a slower release of products, although the rate is much faster than with myosin, and microtubules can activate dynein ATPase in solution. As discussed in detail by Johnson (1985), the similarities in kinetic properties suggest that the nucleotide-binding and hydrolysis steps have similar functions in driving the microtubule-dynein and actomyosin systems.

What is the minimum structure necessary to produce movement? In smooth muscle the thick and thin filaments are not arranged in a regular lattice, yet the muscle can exert a tension comparable to striated muscle. In muscle the thick filament is a bipolar structure, which normally interacts with two actin filament bundles of opposite polarity. The heads of myosin are able to rotate about the base of the globular region, including an axial rotation, since the two curved heads of heavy meromyosin are observed in a parallel orientation in the actin-heavy meromyosin complex (Craig et al. 1980). Since the molecule is a dimer, a relative axial rotation of the two heads must occur. Although there may be some constraint on this rotation, which reduces the probability of interaction of heads with actin filaments of the wrong polarity at short sarcomere lengths, it is primarily the polarity of the actin filament that determines the direction of sliding. An array of myosin molecules that are not organized into a bipolar filament may be expected to act additively in generating a force in a given direction determined by the polarity of actin filaments. The polarity of actin filaments relative to a Z line, dense body or membrane is defined by the direction specified by decoration of the filaments with myosin, which gives an arrowhead structure. The arrow points away from the Z line. The opposite or barbed end is referred to as the plus end because it has a higher rate of elongation in the polymerization reaction. Thus, myosin molecules or thick filaments move toward the plus end of actin filaments.

Movement has been demonstrated in reconstituted actomyosin mixtures. Individual thick filaments slide along polarized bundles of actin filaments (Higashi-Fujime, 1982). A rotary motor was constructed by polymerizing actin filaments on the trailing surface of four paddles arranged at right angles. The addition of myosin fragments and ATP caused a rotation of the paddles (Yano, Yamomoto & Shimizu, 1982). A more interesting model, suitable for quantitative studies, was constructed by covalent attachment of myosin to small beads. The coated beads moved along polarized actin filament bundles obtained from Nitella (Sheetz & Spudich, 1983; Sheetz, Chasan & Spudich, 1984). The velocity of movement corresponded closely to the velocity of unloaded shortening of the sarcomeres of the muscles used as the source of the myosin. Although there is some uncertainty as to movement produced by heavy meromyosin from muscle, the single-headed myosin will function in the bead system (Albanesi et al. 1985), and endogenous vesicles of Acanthamoeba, which have the myosin bound to them, are also transported by actin filaments (Adams & Pollard, 1985). The important conclusion from these studies is that the minimum motile system consists of a polarized bundle of actin filaments and independent myosin molecules attached to a suitable substrate.

Various authors have been intrigued by the possibility that a microtubule-dyneinlike system might be acting in intracellular movement. Although the cilium uses a doublet microtubule that is not found in cytoplasm, the enzyme of cilia will bind to cytoplasmic microtubules (Johnson, 1985). Dynein-like molecules have been isolated from cytoplasm (Pratt, 1980; Hisanga & Sakai, 1983) but convincing evidence for the participation of these components in a motile system has not been obtained. Recently, the movement of vesicles along single microtubules has been observed using extracts of axoplasm (Allen et al. 1985; Vale et al. 1985a ; Lasek & Brady, 1985) and a protein that is necessary for this movement has been purified (Vale et al. 19856 ; Brady, 1985). The protein, named kinesin, binds to microtubules and to vesicles or to carboxylated latex beads. It is dissociated from microtubules by ATP. Although the enzymic activity is very low the properties of the protein suggest that it may be the component of the system responsible for motility. It appears to be distinct from ciliary dynein, since the MT is about 600000 and it lacks the heavy polypeptide chains characteristic of dynein. Although it is too soon to evaluate this important discovery it appears that a motile system can consist of a single microtubule and an enzyme, presumably kinesin, attached to a suitable substrate. So, cells contain two motile systems capable of producing the movement of organelles and the order that is required for unidirectional movement is provided by the polarity of the microtubule or actin filament.

The wide variety of the motile behaviour of cells is divided into three classes for the purpose of discussion. Hopefully the classes correspond to distinguishable differences in mechanism. (1) Organelle movements, including the movement of vesicles in axons, possibly secretory processes in cells, melanophores and some aspects of chromosome movements; (2) changes in cell shape, spreading and cleavage; (3) streaming and amoeboid locomotion.

The classification is not based on whether the movement is caused by an actomyosin or a microtubule system, because the mechanism of a minimum motile system is likely to be the same for both cases.

Organelle movements

In some respects this category may be the simplest to understand. The movements of vesicles and chromosomes at velocities of a few /tm per second require a very low rate of energy production to overcome viscous drag (Taylor, 1964). The summation of the action of a small number of myosin molecules is sufficient to meet the energy requirement. The experiments with coated beads show that binding of myosin or other enzymes to the beads does not require a specific structural arrangement of the myosin. The direction of movement is determined by the polarity of the actin filament or microtubule. The properties of the actomyosin system are well suited to this type of movement. A small number of myosin molecules positioned to interact with actin, probably between 10 and 100, are sufficient to maintain the attachment by summation of weak interactions of M-ATP and M-ADP P states. The rapid detachment of these weakly bound states does not impede the movement. Hydrolysis of ATP at an appreciable rate occurs only for the myosins that interact with actin to complete the hydrolysis cycle. In muscle the activation factor is 500—1000; consequently, the system is reasonably efficient if only a small fraction of the myosins on the surface of the bead are able to interact with actin. Cytoplasmic myosin is activated by calcium binding to a calmodulin-dependent kinase, which phosphorylates a myosin light chain; consequently, the system is controlled by calcium.

This system accounts for organelle movement in a plant cell such as Nitella but it has not been shown to be responsible for vesicle movements in animal cells. The axon provides the most useful system for the study of vesicle movement in animal cells (Schliwa, 1984) and in this case the movement depends on a microtubule system. Fast transport occurs at a rate of 2–3 μms-1, which is comparable to the velocity of sliding of filaments in a muscle. Beads or foreign synaptic vesicles injected into an axon are transported at comparable rates, indicating the presence of a soluble factor (Adams & Bray, 1983; Schroer, Brady & Kelly, 1985). The motion of beads and axoplasmic vesicles has been demonstrated in a reconstituted system consisting of synthetic microtubules and the protein kinesin prepared from the soluble fraction of extracts of brain axoplasm (Vale et al. 1985b). Velocities of movement of particles along single microtubules are comparable to the rate of fast transport in axons.

Movements in the reconstituted system are unidirectional along a particular microtubule but in axons different classes of vesicles are transported in both directions. In axons the microtubules appear to be unipolar (Burton & Paige, 1981), with the plus end pointing down the axon as expected from the polarity of microtubule growth from the centrosome. This poses a problem, since the direction of movement is expected to be determined by the polarity of the microtubule. In crude extracts of axoplasm bidirectional movement of vesicles was observed along a single microtubule (Allen et al. 1985; Vale et al. 1985a). It is difficult to imagine a mechanism of bidirectional movement produced by a single enzyme, which makes it necessary to suppose that two different enzymes are involved.

The movement of chromosomes in mitosis has obvious similarities to the movement of vesicles in axons by a microtubule system. Saltatory movements of particles along astral rays and along the surface of the mitotic spindle are commonly observed in cells (Taylor, 1964). This phenomenon may be explained by the same mechanism as axon transport if the particle dissociates from the microtubule at the end of the saltation.

The mechanism of chromosome movement has long been debated in spite of the absence of evidence. Recent observations, aided by improvements in fluorescence microscopy, have begun to provide important evidence on chromosome movements (Mitchison & Kirschner, 1985a,b). The kinetochore region of the metaphase chromosome is able to capture a microtubule and the plus end of the microtubule points toward the chromosome. In the presence of ATP the chromosome moves toward the plus end as the microtubule elongates by polymerization. The nature of the enzyme is unknown but a kinesin-like factor is an obvious candidate. Although the authors of this exciting work registered some disappointment that the chromosome moves away from the pole, the results suggest an explanation for much of the movement that occurs in mitosis, the movement of chromosomes to the metaphase plate and the oscillation of chromosomes on the plate during metaphase. A second enzyme could be involved in movement to the pole, as in the axon system.

The striking progress in the study of organelle movement promises to provide a biochemical explanation for some of the outstanding problems of cell biology within the next few years.

Cell shape

Immunofluorescence microscopy has revealed the complex arrangement of actin filaments, microtubules and intermediate filaments in the cytoplasm of the cell. A prominent feature of cells, particularly when spread on a surface, is the network of stress fibres (Byers, White & Fujiwara, 1984). The fibres consist of bundles of actin filaments, which in some cases show periodic staining with anti-α-actinin and antitropomyosin. Staining with anti-myosin showed a periodicity of 0·5 μm (Herman & Pollard, 1981). Thus, some stress fibres have a sarcomere-like structure. Active shortening of stress fibres has not been commonly observed in vivo, although contraction can be induced by ATP in isolated fibres (Isenberg et al. 1976).

Stress fibres often terminate at or near focal contacts of the cell with the underlying substrate, which serve to anchor the cell (Izzard & Lochner, 1976). The inner plaque of the adherens junction contains α-actinin, which binds to actin, and vinculin, which is implicated in attachment to membranes (Geiger, Zafrira, Kreis & Schlessinger, 1984). Although there is some uncertainty as to the arrangement of these proteins in the plaque, the structure appears to be responsible for attachment of the stress fibre to the junction. Thus, the stress fibres appear to provide a set of connections among the points of attachment of the cell to the surface. Although there does not appear to be direct evidence that the fibres maintain a state of tension, this would act to stiffen the connections (Byers et al. 1984).

The stress fibre, among the various arrangements of these proteins in cells, comes closest to reproducing the muscle-like organization of actomyosin and it is surprising that it appears to have a structural rather than a contractile function in the cell. However, a structure resembling a stress fibre may function indirectly in cell motility. Moving fibroblasts often taper down to a tail-like projection at the trailing end of the cell. The tail appears to be anchored to the substrate and is pulled off the surface by what appears to be an active contraction of a stress fibre. Retraction requires ATP and activation of the myosin by phosphorylation by myosin light-chain kinase (Yenna & Goldman, 1978; Yenna, Askoy, Hartshorn & Goldman, 1978). Thus, breaking of focal contacts by contraction of a stress fibre may be necessary for movement in cells making strong contacts with the substrate but this contraction does not generate the force for movement.

Cell cleavage can be considered to be a sarcomere-type contraction mechanism. Myosin and a ring of actin filaments are found in the cleavage furrow (Fujiwara & Pollard, 1978). Bundles of actin filaments project at intervals from the membrane and appear to interact with myosin thick filaments (Sanger & Sanger, 1980). This arrangement corresponds to the bipolar sarcomere structure and would act like a draw-string.

Thus, actin and myosin can be arranged in a sarcomere-like structure in the cell but the function of this is in maintaining and changing cell shape, rather than in locomotion. In addition to stress fibres, actin filaments occur in small bundles and in a more or less random orientation, suggesting a gel structure. This type of organization is found adjacent to the plasma membrane and is particularly prominent in the leading lamella of migrating fibroblasts. It is this type of organization, rather than the stress fibre, that is associated with cell movement.

Amoeboid locomotion and streaming

Movement by protoplasmic streaming is exhibited by amoeboid cells and slime moulds and is probably the most difficult to understand of the three classes. The motive force is undoubtedly generated by actomyosin, but the lack of any well-defined structure suggests only a negative conclusion, that a highly ordered structure is not necessary. Small actin filament bundles are observed in the cortex and associated with the membrane but actin is rather uniformly distributed throughout the ectoplasmic gel and streaming sol (Taylor, Wang & Heiple, 1980a). We are left with the conclusion that force is generated by the contraction of a relatively uniform three-dimensional gel of actomyosin. A further difficulty is that the gel must dissolve and be transported in the stream, and again re-form a gel in order to continue the motion in one direction. This would require a fine balance of the factors controlling both contraction and the sol-gel transformation. This double requirement is the basis of the solation—contraction hypothesis, which stresses that contraction and solation are linked reactions (Taylor, Hellewell, Virgin & Heiple, 1979).

The more difficult problem is not the contraction but the control of the state of actin and myosin in the cell. Progress in understanding streaming has come from the extensive studies on the state of actin and myosin in cells.

Myosins of smooth muscle and non-muscle cells, including the macrophage, a typical amoeboid cell, have similar properties. Activation by actin is controlled by phosphorylation of a myosin light chain. Phosphorylation also favours the formation of thick filaments (Suzuki, Onishi, Takahashi & Watanabe, 1978; Trybus, Huiatt & Lowey, 1982; Craig, Smith & Kendrick-Jones, 1983). The tail of myosin folds around the head in the unphosphorylated state and inhibits filament formation. Thus, myosin is soluble at the ionic strength present in a cell and individual myosin molecules are very weakly bound to actin at this ionic strength in the presence of ATP. Thick filament formation can occur in regions of the cell that have a sufficiently high calcium concentration to activate the kinase. In a region of lower calcium concentration, filaments would break down by being dephosphorylated by a phosphatase.

A striking development over the last few years is the isolation of a large number of proteins that interact with actin and control its state of aggregation in the cell. In the earlier studies of muscle actin it was recognized that o’-actinin was bound at or near the barbed end of actin filaments and that β-actinin affected filament length (Maruyama, 1971), but the significance of these proteins was not appreciated. A variety of proteins have been isolated from non-muscle cells that bind to the barbed end, the pointed end and preferentially to actin monomers and thus determine the degree of polymerization of cellular actin. Other proteins induce the formation of bundles of filaments, the formation of three-dimensional gels and breakage of filaments (lists are given by Shliwa, 1981 ; Craig & Pollard, 1982). The polymerization of pure actin as studied in the test tube may not be relevant to what happens in the cell, but recent studies of actin polymerization call attention to an interesting property. The rates of addition of monomers differ by a factor of about 10 for the two ends (Pollard & Mooseker, 1981) and the coupling of ATP hydrolysis to polymerization leads to treadmilling in the steady state. What may be more important is that the rate of dissociation of subunits is larger for an end unit containing bound ADP than bound ATP ; consequently, at a steady state some filaments can depolymerize while others protected by an actin-ATP cap can continue to grow (Pantaloni, Carlier & Korn, 1985). This effect (dynamic instability) is more dramatic for microtubules (Mitchison & Kirschner, 1984) but may also be important for actin filaments.

In order to deal with the bewildering list of proteins that interact with actin it is assumed that proteins with similar properties, which keep turning up in a variety of organisms, are probably important in controlling the state of actin, even though they may have different names. Proteins with similar functions may differ in details such as molecular weight but they will be considered to be essentially the same protein. The types of proteins found in different organisms are: (1) a barbed-end capping protein, which nucleates polymerization and requires calcium to bind to actin. It reduces the length of actin filaments by preventing reannealing of broken filaments or actively breaks filaments and immediately seals the new barbed end. (2) A large flexible dimeric protein, which cross-links actin filaments to yield a threedimensional gel (in which every actin filament is connected to at least one other actin filament). It does not require calcium. (3) A smaller protein, which forms bundles or weak gels with actin. A pointed-end capping protein, requiring calcium, has been described in macrophages (Southwick, Tatsumi & Stossel, 1982). Although it may be important, pointed-end capping proteins have not been described in other cells except for muscle β-actinin. In the macrophage (1) and (2) are gelsolin and actin-binding proteins, respectively. In the sea urchin they are the 45000Mr capping protein (Wang & Spudich, 1984) and the 200000Mr protein (Bryan & Kane, 1978). The corresponding proteins have been described in smooth muscle, Acanthamoeba, Physarum andDictyostelium (Schliwa, 1981; Craig & Pollard, 1982).

The capping and cross-linking proteins can function to produce a calciumdependent sol-gel transformation (Stossel, 1983). In a region of low calcium concentration the cross-linking protein forms a three-dimensional gel with actin. At a higher calcium concentration the capping protein binds to barbed ends and the filament length is reduced by dissociation of actin from the pointed end and by preventing end-to-end association. It may also cut filaments. As the filament length is reduced the three-dimensional gel breaks down into small aggregates of capped filaments connected by the cross-linking protein. Myosin molecules form small thick filaments and interact with actin to bring about contraction of the aggregate. The loose gel of actomyosin, and possibly a bundling protein, probably breaks up into smaller aggregates as contraction proceeds and enters the streaming region. In the amoeba, Chaos carolinensis, the calcium concentration is relatively high in the tail region, where the force of contraction is probably developed, and the concentration decreases toward the head (Taylor, Blinks & Reynolds, 19806). As the actomyosin aggregates are moved forward by the stream, the thick filaments dissociate and the actin re-forms a three-dimensional gel as the capping protein is released. This description, as given by Stossel, Taylor and their collaborators, is certainly incomplete but it is the beginning of a molecular description of this complex phenomenon.

Approximately 20 years have passed since the first isolation of tubulin, dynein and non-muscle actomyosin. During this period actomyosin, microtubule—dynein or microtubules plus other ATPases (kinesin) have been implicated in almost all cellular motile phenomena. A possible exception is that polymerization of actin in a preferred direction could be responsible for extending the leading edge of a fibroblast. We now tend to look for an explanation of a motile process in terms of the cross-bridge cycle or some variation on this general mechanism. It will be of considerable interest to determine whether the new system for organelle movement fits this concept.

In spite of great progress in identifying the basis of motility, fundamental problems remain unsolved. The nature of the conformation change of actomyosin in the cross-bridge cycle is still unknown and the solution of this problem may require the determination of the three-dimensional structure of the actin-myosin-nucleotide complex. We are still at the stage of classifying proteins that bind to actin and microtubules. The control of the dynamic state of actin and the microtubules in the cell is still poorly understood. Recent studies give hope that it will be possible to understand the mechanism of mitosis, but the biochemical study of this problem has only just begun.

This work was supported by Program Project grant HL 20592 of the Heart, Lung and Blood Institute of the National Institutes of Health and by the Muscular Dystrophy Society of America.

Adams
,
R. J.
&
Bray
,
D.
(
1983
).
Rapid transport of foreign particles microinjected in to crab axons
.
Nature, Lond
.
303
,
718
720
.
Adams
,
R. J.
&
Pollard
,
T. D.
(
1985
).
Organelle movements on actin bundles
.
J. CellBiol
.
101
,
389a
.
Albanesi
,
J. P.
,
Fujisaki
,
H.
,
Hammer
,
J. A.
,
Kqrn
,
E. D.
,
Jones
,
R.
&
Sheetz
,
M. P.
(
1985
).
Monomeric acanthamoeba myosin I supports movement in vitro
.
J. biol. Chem
.
260
,
8649
8652
.
Allen
,
R. D.
,
Weiss
,
D. E.
,
Hayden
,
J. H.
,
Brown
,
D. T.
,
Fijisaki
,
H.
&
Simpson
,
M.
(
1985
).
Gliding movement of and bidirectional organelle transport along single native microtubules
.
J. CellBiol
.
100
,
1736
1752
.
Amos
,
W. B.
(
1975
).
Contraction and calcium binding in the vorticelled ciliates
.
In Molecules and Cell Movement
(ed.
S.
Inoue
&
P. E.
Stephens
), pp.
411
436
.
New York
:
Raven Press
.
Brady
,
S. T.
(
1985
).
Novel brain ATPase with properties expected for a fast axonal transport motor
.
Nature, Lond
.
317
,
73
75
.
Bryan
,
J.
&
Kane
,
R. E.
(
1978
).
Separation and interaction of the major components of sea urchin actin gel
.
J. molec. Biol
.
175
,
207
244
.
Burghardt
,
T. P.
,
Ando
,
T.
&
Borejdo
,
J.
(
1983
).
Evidence for cross bridge order in contraction of glycerinated skeletal muscle
.
Proc. natn. Acad. Sci. U.SA
.
80
,
7515
7519
.
Burton
,
P. R.
&
Paige
,
J. L.
(
1981
).
Polarity of axoplasmic microtubules in the olfactory nerve of the frog
.
Proc. natn. Acad. Sci. U.SA
.
78
,
3269
3273
.
Byers
,
H. R.
,
White
,
G. E.
&
Fujiwara
,
K.
(
1984
).
Organization and function of stress fibers in cells
.
In Cell and Muscle Motility
(ed.
J. W.
Shay
), vol.
5
, pp.
83
125
.
New York
:
Plenum Press
.
Cooke
,
R.
,
Crowder
,
M. S.
&
Thomas
,
D. D.
(
1982
).
Orientation of spin-labels attached to cross-bridges in contracting muscle fibers
.
Nature, Lond
.
300
,
776
778
.
Cooke
,
R.
&
Franks
,
K.
(
1978
).
Generation of force by single-headed myosin
.
J. molec. Biol
.
120
,
361
.
Craig
,
S. W.
&
Pollard
,
T. D.
(
1982
).
Actin binding proteins
.
Trends Biochem. Sci
.
7
,
88
92
.
Craig
,
R.
,
Smith
,
R.
&
Kendrick-Jones
,
J.
(
1983
).
Light-chain phosphorylation controls the conformation of vertebrate non-muscle and smooth muscle myosin molecules
.
Nature, Lond
.
302
,
436
439
.
Craig
,
R.
,
Szent-Gyôrgyi
,
A. G.
,
Bease
,
L.
,
Flicker
,
P.
,
Vibert
,
P.
&
Cohen
,
C.
(
1980
).
Electron microscopy of thin filaments decorated with a Ca-regulated myosin
.
J. molec. Biol
.
140
,
35
55
.
Eisenberg
,
E.
&
Greene
,
L. E.
(
1980
).
Relation of muscle biochemistry to muscle physiology
.
A. Rev. Physiol
.
42
,
293
309
.
Ford
,
L. E.
,
Huxley
,
A. F.
&
Simmonds
,
R. M.
(
1977
).
Tension responses to sudden length change in stimulated frog muscle fibres near slack length
.
J. Physiol
.
269
,
441
515
.
Fujiwara
,
K.
&
Pollard
,
T. D.
(
1978
).
Fluorescent antibody localization of myosin in the cytoplasm, cleavage furrow and mitotic spindle of human cells
.
J. Cell Biol
.
71
,
848
875
.
Geiger
,
B.
,
Zafrira
,
A.
,
Kreis
,
T. E.
&
Schlessinger
,
J.
(
1984
).
Dynamics of cytoskeletal organization in areas of cell contact
.
Cell and Muscle Motility
(ed.
J. W.
Shay
), vol.
5
, pp.
195
224
.
New York
:
Plenum Press
.
Goody
,
R. S.
&
Holmes
,
K. C.
(
1983
).
Cross-bridges and the mechanism of muscle contraction
.
Biochim biophys. Acta
726
,
13
39
.
Herman
,
I. M.
&
Pollard
,
T. D.
(
1981
).
Electron microscopic localization of cytoplasmic myosin
.
J. Cell Biol
.
88
,
346
351
.
Higashi-Fujime
,
S.
(
1982
).
Active movement of bundles of actin and myosin filaments from muscle
.
Cold Spring Harbor Symp. quant. Biol
.
46
,
69
75
.
Hlsanga
,
S.
&
Sakai
,
H.
(
1983
).
Cytoplasmic dynein of the sea urchin egg
.
J. Biochem
.
93
,
87
98
.
Huang
,
B.
&
Pitelka
,
D.
(
1973
).
Contractile process in the ciliate, Stentor coerulens. Role of microtubules and filaments
.
J. Cell Biol
.
57
,
704
722
.
Huxley
,
A. F.
&
Simmonds
,
R. M.
(
1971
).
Proposed mechanisms of force generation in striated muscle
.
Nature, Lond
.
233
,
533
538
.
Huxley
,
H. E.
(
1969
).
Mechanism of muscle contraction
.
Science
164
,
1356
1366
.
Huxley
,
H. E.
&
Brown
,
W.
(
1967
).
Low angle X-ray diagram of vertebrate striated muscle
.
J. molec. Biol
.
30
,
383
434
.
Huxley
,
H. E.
&
Kress
,
M.
(
1985
).
Cross-bridge behavior during muscle contraction
.
J. Muscle Res. Cell Motil
.
6
,
153
164
.
Huxley
,
H. E.
,
Simmonds
,
R. M.
,
Faruqi
,
A. R.
,
Kress
,
M.
,
Bordes
,
J.
&
Koch
,
M. H. J.
(
1983
).
Changes in X-ray reflections from contracting muscle during rapid mechanical transients
.
J. molec. Biol
.
169
,
469
506
.
Isenberg
,
G.
,
Rathke
,
P. C.
,
Hulsmann
,
N.
,
Franke
,
W. W.
&
Wohlfarth-Bottermann
,
K. E.
(
1976
).
Cytoplasmic actomyosin fibrils in tissue culture cells
.
Cell Tiss. Res
.
166
,
427
443
.
Izzard
,
C. S.
&
Lochner
,
L. R.
(
1976
).
Cell to substrate contacts in living fibroblasts
.
J. Cell Sci
.
21
,
129
159
.
Johnson
,
K. A.
(
1985
).
Pathway of the microtubule-dynein ATPase and the structure of dynein
.
A. Rev. biophys. biophys. Chem
.
14
,
161
188
.
Lasek
,
R. J.
&
Brady
,
S. T.
(
1985
).
Attachment of transported vesicles to microtubules in axoplasm is facilitated by AMPPNP
.
Nature, Lond
.
316
,
645
646
.
Lymn
,
R. W.
&
Taylor
,
E. W.
(
1971
).
Mechanism of adenosine triphosphate hydrolysis by actomyosin
.
Biochemistry
10
,
4617
4624
.
Maruyama
,
K.
(
1971
).
A study of β-actinin
.
J. Biochem. Tokyo
69
,
369
375
.
Mitchison
,
T.
&
Kirschner
,
M.
(
1984
).
Microtubule dynamics and cellular morphogenesis
.
In Molecular Biology of the Cytoskeleton
(ed.
G. G.
Borisy
,
D. W.
Cleveland
&
D. B.
Murphy
), pp.
27
44
.
New York
:
Cold Spring Harbor Laboratory Press
.
Mitchison
,
T. J.
&
Kirschner
,
M. W.
(
1985a
).
Properties of kinetochore in vitro. I
.
J. Cell Biol
.
101
,
755
765
.
Mitchison
,
T. J.
&
Kirschner
,
M. W.
(
1985b
).
Properties of kinetochore in vitro. II. T
.
Cell Biol
.
101
,
766
777
.
Pantaloni
,
D.
,
Carlier
,
M.
&
Korn
,
E. D.
(
1985
).
Interaction between ATP-actin and ADP-actin
.
J. biol. Chem
.
260
,
6572
6578
.
Pollard
,
T. D.
&
Korn
,
T. D.
(
1973
).
Acanthamoeba myosin
.
J. biol. Chem
.
248
,
4682
4690
.
Pollard
,
T. D.
&
Mooseker
,
M. S.
(
1981
).
Direct measurement of actin polymerization rate constants by electron microscopy
.
J. Cell Biol
.
88
,
654
659
.
Pratt
,
M. M.
(
1980
).
The identification of a dynein ATPase in unfertilized sea urchin eggs
.
Devi Biol
.
74
,
364
378
.
Rosenfeld
,
S. S.
&
Taylor
,
E. W.
(
1984
).
ATPase mechanism of skeletal and smooth muscle actosubfragment-1
.
J. biol. Chem
.
259
,
11908
11919
.
Sanger
,
J. M.
&
Sanger
,
J. W.
(
1980
).
Banding and polarity of actin filaments in interphase
.
J. Cell Biol
.
86
,
568
575
.
Schliwa
,
M.
(
1981
).
Proteins associated with cytoplasmic actin
.
Cell
25
,
587
590
.
Schliwa
,
M.
(
1984
).
Mechanisms of intracellular organelle transport
.
In Cell and Muscle Motility
(ed.
J. W.
Shay
), vol.
5
, pp.
1
66
.
New York
:
Plenum Press
.
Schroer
,
T. A.
,
Brady
,
S. T.
&
Kelly
,
R. B.
(
1985
).
Fast transport of foreign synaptic vesicles in squid axoplasm
.
J. Cell Biol
.
101
,
568
572
.
Sheetz
,
M. P.
,
Chasan
,
R.
&
Spudich
,
T. A.
(
1984
).
ATP-dependent movement of myosin in vitw. Characterization by a quantitative assay
.
J. Cell Biol
.
99
,
1867
1871
.
Sheetz
,
M. P.
&
Spudich
,
J. A.
(
1983
).
Movement of myosin-coated fluorescent beads on actin cables in vitro
.
Nature, Lond
.
303
,
31
35
.
Southwick
,
F. S.
,
Tatsumi
,
N.
&
Stossel
,
T. P.
(
1982
).
Acumentin, an actin modulating protein of pulmonary macrophages
.
Biochemistry
21
,
6321
6326
.
Stein
,
L. A.
,
Chock
,
P. B.
&
Eisenberg
,
E.
(
1984
).
Rate-limiting step in the actomyosin ATPase cycle
.
Biochemistry
23
,
1555
1563
.
Stossel
,
T. P.
(
1983
).
Spatial organization of cortical cytoplasm in macrophages
.
In Modern Cell Biology
(ed.
J. R.
Mcintosh
), vol.
2
, pp.
203
224
.
New York
:
Alan R. Liss
.
Suzuki
,
H.
,
Onishi
,
H.
,
Takahashi
,
K.
&
Watanabe
,
S.
(
1978
).
Structure and function of chicken gizzard myosin
.
J. Biochem. Tokyo
84
,
1529
1542
.
Taylor
,
D. L.
,
Blinks
,
J. R.
&
Reynolds
,
G.
(
1980b
).
Aequorin luminescence during amoeboid movement, endocytosis and capping
.
J. Cell Biol
.
86
,
599
607
.
Taylor
,
D. L.
,
Hellewell
,
S. B.
,
Virgin
,
H. W.
&
Heiple
,
J.
(
1979
).
The solation-contraction coupling hypothesis of cell movements in cell motility: molecules and organization
(ed.
S.
Matano
,
H.
Ishikawa
&
M.
Sato
), pp.
363
377
.
Tokyo
:
University of Tokyo Press
.
Taylor
,
D. L.
,
Wang
,
Y.
&
Heiple
,
J. M.
(
1980a
).
Distribution of fluorescently-labelled actin in living amoebas
.
J. Cell Biol
.
86
,
590
598
.
Taylor
,
E. W.
(
1964
).
Brownian and saltatory movements of cytoplasmic granules and the movement of anaphase chromosomes
.
Proc. 4th Int. Conf. Rheol
. pp.
175
191
.
New York
:
J. Wiley
.
Trybus
,
K. M.
,
Huiatt
,
T. W.
&
Lowey
,
S.
(
1982
).
A bent monomeric conformation of myosin from smooth muscle
.
Proc. natn. Acad. Sci. U.SA
.
79
,
6151
6155
.
Vale
,
R. D.
,
Reese
,
T. S.
&
Sheetz
,
M. P.
(
1985b
).
Identification of a novel free-generating protein, kinesin, involved in microtubule-based motility
.
Cell
42
,
39
50
.
Vale
,
R. D.
,
Schnapp
,
B. J.
,
Reese
,
T. S.
&
Sheetz
,
M. P.
(
1985a
).
Movement of organelles along filaments dissociated from axoplasm of squid giant axon
.
Cell
40
,
449
454
.
Wang
,
L.
&
Spudich
,
J. A.
(
1984
).
A 45,000mol. wt. protein from unfertilized sea urchin eggs severs actin filaments in a calcium-dependent manner
.
J. Cell Biol
.
99
,
844
851
.
Webb
,
M. F.
&
Trentham
,
D. R.
(
1983
).
Chemical mechanism of myosin catalyzed ATP hydrolysis
.
In Handbook of Physiology: Skeletal Muscle
(ed.
L. D.
Peachey
,
H.
Adrian
&
S. R.
Geiger
), pp.
237
255
.
Baltimore
:
Waverley Press
.
Winkelmann
,
D. A.
,
Mekeel
,
H.
&
Rayment
,
I.
(
1985
).
Packing analysis of crytalline myosin sub-fragment-1
.
J. molec. Biol
.
181
,
487
501
.
Yanigita
,
T.
(
1981
).
Angles of nucleotides bound to cross bridges in glycerinated muscle fibers
.
J. molec. Biol
.
146
,
539
560
.
Yano
,
M.
,
Yamamoto
,
Y.
&
Shimizu
,
H.
(
1982
).
An actomyosin motor
.
Nature, Lond
.
299
,
557
559
.
Yenna
,
M.
,
Aksoy
,
M.
,
Hartshorn
,
D. J.
&
Goldman
,
R. D.
(
1978
).
BHK-21 myosin. Isolation, biochemical characterization and intracellular localization
.
J. Cell Sci
.
31
,
411
429
.
Yenna
,
M.
&
Goldman
,
R. D.
(
1978
).
Ca sensitive regulation of actin and myosin interaction in cultured BHK-21
.
J. Cell Biol. (Abstr
.)
79
,
274a
.