Special Issue: The Actin-Myosin Interaction in Muscle: Background and Overview

Abstract

Muscular contraction is a fundamental phenomenon in all animals; without it life as we know it would be impossible. The basic mechanism in muscle, including heart muscle, involves the interaction of the protein filaments myosin and actin. Motility in all cells is also partly based on similar interactions of actin filaments with non-muscle myosins. Early studies of muscle contraction have informed later studies of these cellular actin-myosin systems. In muscles, projections on the myosin filaments, the so-called myosin heads or cross-bridges, interact with the nearby actin filaments and, in a mechanism powered by ATP-hydrolysis, they move the actin filaments past them in a kind of cyclic rowing action to produce the macroscopic muscular movements of which we are all aware. In this special issue the papers and reviews address different aspects of the actin-myosin interaction in muscle as studied by a plethora of complementary techniques. The present overview provides a brief and elementary introduction to muscle structure and function and the techniques used to study it. It goes on to give more detailed descriptions of what is known about muscle components and the cross-bridge cycle using structural biology techniques, particularly protein crystallography, electron microscopy and X-ray diffraction. It then has a quick look at muscle mechanics and it summarises what can be learnt about how muscle works based on the other studies covered in the different papers in the special issue. A picture emerges of the main molecular steps involved in the force-producing process; steps that are also likely to be seen in non-muscle myosin interactions with cellular actin filaments. Finally, the remarkable advances made in studying the effects of mutations in the contractile assembly in causing specific muscle diseases, particularly those in heart muscle, are outlined and discussed.

Keywords: M-band; Z-band; actin filament compliance; actin filaments; cross-bridge compliance; dilated cardiomyopathy; fluorescence methods; hypertrophic cardiomyopathy; myosin cross-bridge cycle; myosin filament compliance; myosin filaments; rigor muscle; sarcomere compliance; spin probe methods; strong-binding states; the sarcomere; time-resolved X-ray diffraction; weak-binding state.

Figure 1

Summary of the main muscle types of (a) vertebrates and (b) invertebrates [3]. Muscle types which are commonly used in muscle research are highlighted in red. Vertebrate skeletal muscles occur in two distinct structural types, simple lattice and superlattice, but they are also distinguished physiologically in terms of different fibre types (e.g., fast, slow, intermediate). Insect muscles are classified in terms of the number of actin filaments that there are for each half myosin filament. Insect flight muscles often have 3 actins per half myosin filament, whereas leg muscles, for example, have more actins.

Figure 2

(a) Electron micrograph of a vertebrate striated muscle sarcomere (here about 2.3 µm long) which runs between two Z-lines (Z) to which actin filaments (A) are attached. The A-band contains myosin (thick) filaments (M), cross-linked at the M-band, and which the actin filaments partly overlap. (b) Simplified diagram of a myosin II molecule; a long rod on the end of which are two myosin heads which bind and hydrolyse ATP and bind actin. Myosin molecules (b) aggregate to form myosin filaments (c); the rods are in the backbone and the heads are almost helically arranged on the filament surface. The head ends of the rods, myosin S2 (subfragment-2) can lift from the filament surface to enable myosin head attachment to actin. Actin (thin-) filaments (d) comprise a twisted helical array of globular (G-)actin monomers along which run two strands of tropomyosin on which is also the troponin complex. For details see text. (Adapted from [7]).

Figure 3

Structure of the myosin head from X-ray crystallography [12]. The top part is the motor domain, part of the myosin heavy chain, which can bind and hydrolyse ATP and can also bind to actin, which accelerates the ATPase. The lower part is the lever arm. It contains a central α–helix, also, part of the heavy chain, around which are wrapped two light chains, the essential light chain (ELC, yellow) and the regulatory light chain (RLC, maroon). Between the motor and the lever arm is the converter domain, which can be thought of as a kind of gear box. In the muscle contractile cycle the motor domain binds to actin, products of hydrolysis (ADP and Pi) are released and the lever arm tends to rotate around the converter domain; it will swing to a new position if it is free to do so, or provide a constant force if it is restrained and the lever arm cannot move. Adapted from [7].

Figure 4

3D reconstruction of part of the bridge region of myosin filaments from human cardiac muscle [20]. The image shows a length of about 450 Å, containing three crowns of head pairs. The possible location of strands of titin and the accessory protein C-protein (MyBP-C) are shown in yellow and mauve respectively. The myosin head densities have been fitted with interacting head motif structures ([21]; discussed later). This conformation is supposed to be the position of the heads in what has been called the super-relaxed state when the ATP turnover rate is very low (see Section 2.4). Adapted from [20].

Figure 5

Structure of the G-actin monomer [37], but here from human cardiac muscle (3DAW). (a) Ribbon diagram and (b) surface representation of G-actin showing the four sub-domains and the central ATP-binding pocket. In actin filaments (e.g., Figure 2d or Figure 9a) sub-domains 3 and 4 lie on the inside where they interact with sub-domains 3 and 4 of neighbouring actin monomers, and sub-domains 1 and 2 lie on the outside of the filament. Myosin heads bind preferentially to sub-domain 1. Generated using the Swiss-Prot database [48].

Figure 6

Wall-eyed stereo image of the interacting head motif structure seen on most relaxed myosin filaments (see Figure 4). Here M represents the motor domains and L represents the lever arms. The two lever arms come together at the first part of the coiled-coil myosin rod, known as subfragment-2 or S2. The motor domain (M2) of head 2 interacts with the M1 motor domain through its actin-binding site. Head M2 is therefore referred to as the docked head. The other head (M1), where the actin binding site is exposed, is termed the free head.

Figure 7

(a) Structure of the ATP molecule showing the adenine group, the ribose ring, and the three terminal phosphate groups. Also indicated is the position of the chelated magnesium ion on the last two phosphates. This is the normal Mg-ATP form of ATP found in muscle. (b) Structural representation of the ATP hydrolysis reaction in which ADP and inorganic phosphate (Pi) are the products. Reproduced from [3].

Figure 8

A simplified ATP-driven contractile cycle according to the scheme of Lymn and Taylor [55]. For details see text. Adapted from [3]. Red, myosin head; brown, myosin filament backbone, blue/green actin filament. (a) initial attached state, (b) end state (rigor-like) after product (ADP and Pi) release, (c) detached state induced by ATP binding, (d) myosin head after ATP hydrolysis with products ADP and Pi bound.

Figure 9

3D reconstructions of thin filaments with and without Ca²⁺ bound [39]. (a–c) Data from Ca²⁺-treated filaments, (d–f) data from Ca²⁺-free filaments. (a,d) Wire mesh representation of the single particle based reconstructions of the thin filament in the two states. An atomic F-actin model [38] is docked into the reconstruction and each subunit is colour coded. The barbed end (Z-line end) of the actin filament is at the bottom of the figure. (b,e) Difference density maps calculated by subtracting the docked F-actin model (grey) from the single particle reconstructions. This leaves the density attributable to the regulatory proteins troponin and tropomyosin (both orange). (c,f) Difference density maps calculated by subtracting docked F-actin (grey) and tropomyosin (orange) models from the single particle reconstructions leaving density attributable only to troponin (blue). Reproduced from [39] with permission.

Figure 10

Actin filament structure and regulation: (a) An untwisted version of an F-actin filament, (b) coiled-coil strands of tropomyosin. (c) Troponin structures with and without Ca²⁺ bound. (d–f) The whole thin filament (untwisted) to illustrate the assembly of the three components in (a–c) and their changing configurations in different states: (d) Off state, no Ca²⁺. (e) Closed state, Ca²⁺ bound to troponin alters troponin which shifts tropomyosin strands to where some heads can attach. (f) With more attached heads the tropomyosin is pushed over further to the open state and yet more heads can then attach to actin. (d–f) adapted from Paul et al. [39], with permission.

Figure 11

(a) Different time scales of the twitch responses in (1) a mammalian gastrocnemius muscle, (2) a mammalian soleus muscle, and (3,4) a frog sartorius muscle at 10 °C (3) and 7 °C (4). (The maximum tensions here have been normalised to the same scale). (b) Variation of tension response in a twitch muscle with different stimuli: (A) response to a single stimulus (pulse), (B) response to two closely–spaced stimuli, (C) unfused tetanus produced by repetitive stimulation at moderate frequences, and (D) fused tetanus induced by repetitive stimuli arriving faster than the fusion frequency. T_max is the maximum tension, sometimes called P_o. The horizontal axis is time (seconds). Adapted from [3] after [75].

Figure 12

Summary of the results of Huxley and Simmons and their collaborators [76,77,78]. Representation of the experiment showing the tension transient in a fully active intact frog muscle fibre (b) after a rapid shortening step (a) of about 8 nm. (b) shows the point where the T₁ and T₂ tensions were recorded. (c) the T₁ and T₂ plots from experiments as in (a,b), but for different shortening steps (filament displacement) and shown at two different sarcomere lengths—solid lines full overlap, dashed lines 3.1 µm (0.39% of full overlap). Figure adapted from [79] after [76,77,78].

Figure 13

Methods in structural biology: (a) Outline diagram showing the steps in an optical microscope. The incoming light interacts with the object and is scattered (diffracted) into spreading beams which enter the objective lens. At the plane marked DP (diffraction plane) the diffraction pattern can be seen. The lens focusses the diffracted beams in such a way that the path length is the same for each of them. These beams recombine to give an enlarged image. Preserving the path length is important since each of the scattered beams of light has a sinusoidally varying amplitude and a relative phase (i.e., are the amplitudes in different scattered beams changing in the same way at the same time—or, if not, what is the lag between them?). If the beams are recombined with the correct amplitude and phase then a faithful image is produced. (b) In an electron microscope the same principle applies, but now there is a beam of electrons being scattered and the lenses are electromagnetic. Electrons have wave properties and, once again, they need to be recombined with the correct amplitude and phase to give a good image. (c) With X-rays the problem is that X-rays cannot easily be focused; it is difficult to make good X-ray lenses. However, the first part of the procedure can be carried out as with light and electrons. X-ray beams are diffracted and, at the diffraction plane, a detector records the diffraction pattern. The spots on the diffraction pattern have intensities that are the square of the amplitudes of the beams, but, by recording the pattern, the phase information is lost (the phase problem). So, the second half of the process, the recombination of beams to give an image, is done in a computer using various tricks to find the phase information. This is the basis of protein crystallography.

Figure 14

Probes: (a) Energy levels in fluorescence probes, (b) A: EPR: Spin up and spin down (M_s = ± 1/2) energy levels in a magnetic field B_0, which become more widely separated the stronger the field_, and a microwave source on the right that can flip the spins. Inset (B), the microwave absorption spectrum (top) from the spin flip in A and its first derivative (below). For details see text.

Figure 15

Apparatus used for rapid reaction kinetics: (a) stopped flow apparatus, (b) quenched flow apparatus. In (a) the reactants are held in syringes A and B and are forced into the mixing chamber C within a few msec. The mixing chamber can be monitored optically to follow the progress of the reaction. In (b) the reactants A and B, as before, are mixed in a small chamber (mixer 1). The flow is continuous, and the reaction proceeds in the tube between mixer 1 and mixer 2, The reaction is quenched in the mixer 2 chamber by the addition of acid from syringe C. (c) Example of a quenched flow result showing phosphate accumulation when rabbit myosin was mixed with Mg-ATP [55]. The initial ATP concentration was 32 M and the magnitude of the Pi burst was 1.2 moles Pi/ mole myosin. (Adapted from [3] after White and Thorson [112]).

Figure 16

Simultaneous measurement of length (a), force (b) and phosphate release (c) in a single skinned muscle fibre, illustrating the acceleration of Pi release rate during shortening. A permeabilized muscle fibre was mounted between two hooks, one attached to a length-adjusting motor and the other to a force transducer. The Figure shows two consecutive measurements on a single fibre. The fibre was initially at rest length in a rigor solution (no ATP). At zero time a contraction was initiated by the release into a muscle fibre of around 1.5 mM ATP by laser photolysis of caged ATP. (a) The length of the fibre as controlled by the motor. At either 0.2 or 0.4 s, the fibre was allowed to shorten by 8% of its length. (b) Force measurements: after approximately 0.2 s, the fibre, which initially was prevented from shortening, reached its maximum level of force development. (c) The amount of phosphate bound to the Pi-sensor incubated with the muscle fibre. The trace shows clearly that the rate of Pi liberation was increased during shortening steps (adapted from Figure 5 of He et al. [116]).

Figure 17

In vitro motility assay and optical tweezers set-up for single-molecule mechanics. (a) Flow cell of two cover-slips sandwiched on top of each other via spacers; (b) magnified view of rectangular area in (a) indicating the principle for the gliding in vitro motility assay with surface-adsorbed myosin heads that propel an actin filament. The curved red arrow indicates a possible filament sliding path; (c) schematic diagram of the three-bead optical tweezers assay where an actin filament is suspended between two beads held in optical traps. The filament is then lowered down to allow the actin filament to interact with single myosin motor fragments adsorbed to a third bead or another type of pedestal as indicated here. Reproduced from Special Issue Review [125] with permission.

Figure 18

Diagram of the environmental chamber in the electron microscope. The chamber is sealed top and bottom by carbon films through which the electrons can penetrate. ATP is applied to the specimen iontophoretically by passing a current through the ATP electrode, containing 100 mM ATP. Reproduced from Sugi et al. [126] with permission.

Figure 19

Various stages of the actin-myosin ATPase cycle (cf. Figure 8) as expanded by Houdusse and Sweeney [127]. For details see text. Reproduced from [127] with permission.

Figure 20

Results from use of the electron microscope environmental chamber to study myosin head conformations under different conditions. (a–c) Histograms showing the amplitude distribution of ATP-induced myosin head movement in the recovery stroke in the absence of actin at the distal (a), and the proximal (b) regions of the myosin head motor domain, and at the two regulatory light chains located at the proximal lever arm region of the myosin head (c). (d,e) Diagrams illustrating the ATP-induced changes in the myosin head configuration in the absence (d) and in the presence (e) of actin filaments. These are based on histograms such as (a–c) and similar histograms for the movement in the presence of actin. In (e) the heads start in rigor and then ATP is added. The heads are thought to come off actin, hydrolyse ATP, rebind to actin and then go through a power stroke. Sites 1 and 2 in (d,e) are the distal and proximal labelling positions discussed in (a,b), and sites 3 and 3′ are the light chain labelling positions discussed in (c). Adapted from ref [126] with permission.

Figure 21

Schematic diagram showing the different states that have been identified from electron microscopy and low-angle X-ray diffraction studies: (a) the fully relaxed (super-relaxed) state; (b) the normal (activated) relaxed state in intact muscle; (c) the attached part of the weak-binding state where heads are in a rapid detached-attached equilibrium; (d) the strong states during which the lever arm is supposed to swing axially towards the rigor state (e); (f) the rapid detachment caused by binding of ATP to the rigor-like heads in (e); (g) the resetting state where ATP is hydrolysed and the lever arm returns towards its configuration in resting muscle. All transitions are reversible, but some are more likely in the forward direction around the cycle. For discussion see text.

Figure 22

Part of the IHM structure in Figure 4 [21] as detailed by Marston [144]. Strands of titin are shown in dark blue, three domains of MyBP-C in orange, the myosin subfragment-2 (S2) in pink, and the myosin heads in pale blue (docked head) and green (free head). The myosin mesa on the docked head is circled in orange and the free head converter domain interacting with the docked head is indicated by a red ellipse. Adapted from Figure 2c of Marston [144] using structural data from [20].

Figure 23

Examples of measurements of (a) tension and (b) ATPase as a function of the concentration of calcium [Ca²⁺]. Both follow a sigmoid curve centred between 10⁻⁷ and 10⁻⁶ M Ca²⁺. Also shown is the EC₅₀ value where the parameter in question (e.g., tension or ATPase or other) has reached 50% of its peak value. Shifts of the sigmoid curve to the left or right indicate increased or decreased sensitivity respectively. (Adapted from Squire [3] after [150,151]).