Organization of the Actomyosin Apparatus

Interdigitation of thick, bipolar, myosin filaments and thin actin filaments in the sarcomeres of living muscle cells is so precise (Fig. 39-3) that it yields an X-ray diffraction pattern (Fig. 39-13) revealing the spacing of the filaments and the helical repeats of their subunits to a resolution of about 3 nm. Z disks at both ends of the sarcomere anchor the barbed ends of the actin filaments, so their pointed ends are near the center of the sarcomere. Myosin heads project from the surface of thick filaments, whereas their tails are anchored in the filament backbone. Thick and thin filaments overlap, with the myosin heads only a few nanometers away from adjacent actin filaments. The alignment and interdigitation of the filaments facilitate the sliding interactions required to produce contraction.

Figure 39-3 electron micrographs and drawings of sarcomeres. A, Longitudinal thin section showing the array of thin filaments anchored to Z disks and overlapping bipolar thick filaments cross-linked in the middle at the M line. B, Longitudinal freeze-fractured, etched, and shadowed sarcomere showing myosin cross-bridges attached to thin filaments near the bare zone in the center (right) of a sarcomere. C–D, Cross-sections of insect flight muscle and vertebrate skeletal muscle showing the double hexagonal arrays of thick and thin filaments. E, Drawings indicating the polarity of the thick and thin filaments.

(A and C, Courtesy of H. E. Huxley, Brandeis University, Waltham, Massachusetts. B and D, Courtesy of J. Heuser, Washington University, St. Louis, Missouri.)

Figure 39-13 cross-bridge dynamics revealed by x-ray diffraction patterns of whole muscle. A, Electron micrograph showing the orientation of the muscle in the X-ray beam. B–C, Fiber diffraction patterns from relaxed and contracting skeletal muscles with interpretive drawings of cross-bridges in each state. Reflections from myosin heads arranged on the thick filament helix are strong in relaxed muscle. Reflections from the actin helix are stronger than the thick filament helix in contraction. The myosin and actin reflections are each labeled in only one of four equivalent quadrants. During contraction, a few myosin heads attach transiently to actin, increasing the strength of the actin helix reflections, but most are disordered.

(Micrograph and X-ray patterns courtesy of H. E. Huxley, Brandeis University, Waltham, Massachusetts.)

An important, simplifying architectural feature is that sarcomeres are symmetrical about their middles (Fig. 39-3). Consequently, the polarity of myosin relative to the actin filaments is the same in both halves of the sarcomere, allowing the same force-generating mechanism to work at both ends of the bipolar myosin filaments. Sarcomeres are organized end to end into long, cylindrical assemblies called myofibrils (Fig. 39-2) that retain their contractility even after isolation from muscle.

Thin Filaments

Thin filaments consist of actin and the tightly bound regulatory proteins troponin and tropomyosin (Fig. 39-4). When the concentration of Ca²⁺ is low in cytoplasm, troponin and tropomyosin inhibit the actin-activated ATPase of myosin. Tropomyosin, a 40-nm long coiled-coil of two α-helical polypeptides (see Fig. 3-10), binds laterally to seven contiguous actin subunits as well as head to tail to neighboring tropomyosins, forming a continuous strand along the whole thin filament. Troponin (TN) consists of three different subunits called TNC, TNI, and TNT (Table 39-1). TNT anchors troponin to tropomyosin. Like calmodulin (see Fig. 3-12 and Chapter 26), TNC is a dumbbell-shaped protein with four EF-hand motifs to bind divalent cations. In resting muscle the C-terminal globular domain of TNC binds two Mg²⁺ ions and an N-terminal α-helix of TNI, while the two low-affinity sites in the N-terminal globular domain of TNC are empty. Ca²⁺ binding to the low-affinity sites during muscle activation exposes a new binding site for TNI. The resulting conformational change in TNI allows tropomyosin to expose myosin-binding sites on the actin filament.

Figure 39-4 thin filament structure. A, Three-dimensional reconstruction from electron micrographs of a thin filament from vertebrate skeletal muscle showing actin and the position of tropomyosin in relaxed muscle. B, Drawing of a model of a thin filament from active muscle. Each tropomyosin is associated with seven actin subunits. The structure and binding sites of troponin and tropomodulin have been inferred from biochemical experiments. C, Ribbon diagrams of the atomic structures of troponin C, free and bound to a troponin I peptide. Two divalent cation-binding EF-hands are found at each end, separated by a long α-helix. In cells, two high-affinity sites at the C-terminal end are permanently occupied with Mg²⁺. Two low-affinity sites at the other end are unoccupied in relaxed muscle but bind Ca²⁺ when muscle is activated.

(A, Courtesy of W. Lehman, Boston University, Massachusetts. C, PDB files: 1AX2 and 1TROP.)

A protein meshwork in the Z disk anchors the barbed end of each thin filament (Fig. 39-5). Some cross-links between actin filaments consist of α-actinin, a short rod with actin-binding sites on each end (see Fig. 33-16). At least a half dozen structural proteins stabilize the Z disk through interactions with α-actinin, actin, and titin in the Z disk.

Figure 39-5 z disk structure. A–B, Electron micrographs of thin sections perpendicular to and in the plane of the Z disk. C, Three-dimensional reconstruction, based on electron micrographs of the Z disk, showing the network of protein cross-links that anchor the barbed ends of the yellow actin filaments.

(Courtesy of J. Deatherage, National Institutes of Health, Bethesda, Maryland; modified from Cheng NQ, Deatherage JF: Three dimensional reconstruction of the Z disk of sectioned bee flight muscle. J Cell Biol 108:1761–1774, 1989, by copyright permission of The Rockefeller University Press.)

Proteins cap both ends of thin filaments. Cap-Z, the muscle isoform of capping protein (see Fig. 33-14), binds the barbed ends of thin filaments with high affinity, limiting actin subunit addition or loss. Tropomodulin associates with both tropomyosin and actin to cap and stabilize the pointed end of thin filaments (Fig. 39-4B).

Tropomyosin and a gigantic filamentous protein, nebulin, stabilize thin filaments laterally. Nebulin consists of about 185 imperfect repeats of 35 amino acids that interact with each actin subunit, tropomyosin, and troponin along the length of thin filaments. Interactions with tropomodulin and Z disk proteins anchor nebulin at the two ends of the thin filament. Acting as a ruler and a cap, nebulin and tropomodulin help to set the length of thin filaments.

Thick Filaments

The self-assembly of myosin II (see Fig. 5-7) establishes the bipolar architecture of striated muscle thick filaments (Fig. 39-6). Some features of thick filaments are invariant, such as a backbone consisting of myosin tails, a surface array of myosin heads, the 14.3-nm stagger between rows of heads, and a central bare zone formed by antiparallel packing of tails. Filaments may vary in length, diameter, and organization of the helical array of heads in various species. Invertebrate thick filaments have a core of paramyosin, a second coiled-coil protein, which is not found in vertebrates.

Figure 39-6 structure of bipolar thick filaments. A, Electron micrograph of a thick filament isolated directly from skeletal muscle and prepared by negative staining. A myosin molecule is shown at the same magnification at the lower left. The myosin tails form the backbone of the thick filament and allow the two myosin heads to swing out from the side (see the enlarged inset on the right). B, Reconstruction from electron micrographs of part of a rabbit skeletal muscle thick filament. The surface bumps are myosin heads. C, Cross section of vertebrate skeletal muscle showing the double hexagonal arrays of thick and thin filaments. D, Electron micrograph of a highly stretched sarcomere with the M line in the middle. E, Drawing of protein links between thick filaments in the M line.

(A, Courtesy of John Trinick, University of Bristol, England. Reference: Knight P, Trinick J: Structure of the myosin projections on native thick filaments from vertebrate skeletal muscle. J Mol Biol 177:461–482, 1984. B, Courtesy of M. Stewart, MRC Laboratory of Molecular Biology, Cambridge, England. C, Courtesy of J. Heuser, Washington University, St. Louis, Missouri. D, Courtesy of H. E. Huxley, Brandeis University, Waltham, Massachusetts.)

Several accessory proteins stabilize thick filaments (Table 39-1). Thick filaments in most striated muscles are girdled at intervals by semicircular bands of a protein that is, coincidentally, called C-protein. C-protein consists of fibronectin III and immunoglobulin domains. The “M line” in the center of the sarcomere is a three-dimensional array of protein cross-links that maintains the precise registration of thick filaments. At least three structural proteins and the enzyme MM-creatine phosphokinase (which transfers phosphate from creatine-phosphate to ADP) are located in the M line.

Titin Filaments

A third array of protein filaments lies parallel to the thin and thick filaments, connecting the Z disk to the thick filaments and the M line (Fig. 39-7). Hard to preserve for electron microscopy, these diaphanous filaments were neglected for years. Each filament is a single polypeptide named titin (after mythological giants), so named because of its remarkable size: more than 30,000 amino acids folded into a linear array of 300 immunoglobulin and fibronectin II domains measuring more than 1.2 μm long. Titin is thought to be the largest protein encoded by the human genome.

Figure 39-7 titin filaments. Upper panel, Electron micrographs of single, isolated titin molecules prepared by heavy metal shadowing. Titin molecules are long enough to extend from the Z disk to the M line. Middle panel, Drawing of a sarcomere, to the same scale as the electron micrograph, with the thick filaments removed from the bottom half to illustrate how titin molecules anchor thick filaments to the Z disk and extend to the M line. Lower panel, Drawing illustrating a model for the elasticity of titin. Modest stretches within the physiological range reversibly extend the chain of Ig domains in the I-band and the PEVK domain. Extreme extension can unfold immunoglobulin domains.

(Modified from Reif M, Gautel M, Oesterhelt F, et al: Reversible unfolding of individual titin immunoglobulin domains by AFM. Science 276:1090–1092, 1997. Reference: Leake MC, Wilson D, Gautel M, Simmons RM: The elasticity of single titin molecules using a two-bead optical tweezers assay. Biophys J 87:1112–1135, 2004.)

Titin molecules are elastic, and this accounts for the passive resistance to stretching of relaxed muscle. Their connections to the Z disk and thick filaments provide elastic continuity from one sarcomere to the next and keep the thick filaments centered in the sarcomere during contraction. If titin molecules are broken experimentally, thick filaments slide out of register toward one Z disk during contraction. Two features provide the elasticity during short (∼0.3 μm per titin), physiological stretches: The irregular chain of immunoglobulin domains in the I band straightens out, and a segment of the polypeptide rich in proline, glutamic acid, valine, and lysine (the PEVK domain) stretches. Stretching decreases entropy and provides the energy for elastic recoil. (See Fig. 29-12 for another example of an entropic spring in biology.) Extreme stretching unfolds Ig domains one by one.

Intermediate Filaments

Desmin intermediate filaments (see Chapter 35) help to align the sarcomeres laterally (Fig. 39-8) by linking each Z disk to its neighbors and to specialized attachment sites on the plasma membrane. Myofibrils near the cell surface are attached to the plasma membrane at specializations called costameres. In addition to desmin, costameres contain several cytoskeletal proteins (vinculin, talin, spectrin, and ankyrin) found in focal contacts and adherens junctions of nonmuscle cells (see Figs. 30-11 and 31-7). Desmin mutations in humans cause disorganization of myofibrils, resulting in generalized muscle failure.

Figure 39-8 desmin intermediate filaments in skeletal muscle. Desmin filaments connect Z disks laterally to each other and to the plasma membrane at specializations called costameres.

(Redrawn from Lazarides E: Intermediate filaments as mechanical integrators of cellular space. Nature 283:249–256, 1980.)

Structural Proteins of the Plasma Membrane: Defects in Muscular Dystrophies

In addition to providing a permeability barrier, the plasma membrane of the muscle cell must maintain its integrity while being subjected to years of forceful contractions. Accordingly, the membrane is stabilized by a transmembrane complex of proteins, the dystroglycan-sarcoglycan complex, that links the internal membrane skeleton to the basal lamina outside (Fig. 39-9 and Table 39-2). Occasional breaches of the membrane are inevitable, so muscle cells also depend on a repair process that reseals holes. If membrane damage exceeds the repair capacity, muscle cells degenerate locally (segmental necrosis) or globally. Cell death beyond the ability of muscle stem cells to repair the tissue results in muscular dystrophy. The age of onset and clinical features of inherited muscular dystrophies depend on the molecular defect. Patients with severe defects develop progressive muscle weakness as children. Ultimately, failure of respiratory muscles is fatal.

Figure 39-9 dystrophin and associated proteins stabilize the plasma membrane of skeletal muscle. A–B, Fluorescent antibody staining of cross sections of human skeletal muscle showing the localization of dystrophin at the plasma membrane of a normal individual (B) and its absence in an individual with Duchenne’s muscular dystrophy (A). C, Model of the transmembrane complex of proteins that links dystrophin and actin filaments in cytoplasm to laminin in the basal lamina outside the cell.

(A–B, Courtesy of L. Kunkel, Harvard Medical School, Boston, Massachusetts. C, Based on a drawing by K. Amann and J. Ervasti, University of Wisconsin, Madison.)

Mutations in more than 40 human genes have been linked to muscular dystrophies (Table 39-2), so it is possible that the malfunction or lack of any molecule in the system that maintains the integrity of the plasma membrane will cause disease. Disease-causing mutations occur in genes for the proteins of the dystroglycan-sarcoglycan complex, extracellular matrix proteins, Golgi apparatus enzymes that process these structural proteins, and the membrane repair machinery. Some of these mutations also affect the nervous system. Mutations are usually autosomal-recessive and are not uncommon. About one in several thousand humans develops some form of muscular dystrophy, because they inherit mutations in both copies of one of the sensitive genes. The mechanism of disease in muscular dystrophies is similar to that in hereditary spherocytosis, in which deficiencies of the membrane skeleton make red blood cells susceptible to mechanical damage (see Fig. 6-10).

The proteins that stabilize muscle membranes escaped detection until the late 1980s, when x-linked mutations in the dystrophin gene were discovered to cause Duchenne’s muscular dystrophy, the most common human form of the disease. Dystrophin is an enormous member of the α-actinin superfamily of actin-binding proteins (see Fig. 33-16). The dystroglycan-sarcoglycan complex was found when it copurified with dystrophin after solubilizing the membrane with detergents. Loss of any protein of the dystrophin-dystroglycan-sarcoglycan complex typically leads to secondary loss of the other proteins from muscles.

Studies of mutations in muscular dystrophy patients led to the identification of most of the other genes in this system. a2 laminin is the extracellular ligand for dystroglycan in the basal lamina. Glycosylation of the transmembrane complex by Golgi apparatus glycosyltransferases is required for their mechanical functions. Proteins in cytoplasmic vesicles are used to repair damaged plasma membranes. The mechanical activity of muscle cells might make them more sensitive than other cells to deficiencies in proteins that support the nuclear envelope (lamin A/C and emerin).

Dystroglycans and a dystrophin homolog, utrophin, participate in clustering acetylcholine receptors at the neuromuscular junction, the chemical synapse between motor neurons and skeletal muscle (see Fig. 11-8). When, during development, a motor neuron arrives at the surface of its target muscle cell, the neuron secretes an adhesive protein called agrin, which is incorporated into the basal lamina, immediately adjacent to the nerve terminal. Dystroglycan binds agrin and positions associated acetylcholine receptors at the site where they receive acetylcholine secreted by the nerve in response to an action potential.

Interaction of Plasma Membrane Invaginations with the Smooth Endoplasmic Reticulum

The plasma membrane of skeletal muscle cells, like the plasma membrane of nerve cells, is excitable (see Fig. 11-6) and invaginates deeply to form so-called T tubules that run across the entire cell (Fig. 39-10). Depending on the species and type of striated muscle (skeletal versus cardiac), T tubules may be located either at the level of the Z disks or at the thick fila-ment ends. Inside the muscle cell, T tubules interact extensively with the smooth endoplasmic reticulum (SER) that surrounds each myofibril. Historically, this SER has been called sarcoplasmic reticulum. Terminal cisternae of SER are closely associated with the passing T tubules by foot processes that can be visualized by electron microscopy. Together, T tubules and SER constitute a signal-transducing apparatus that converts depolarizations of the plasma membrane into a spike of cytoplasmic Ca²⁺ that triggers contraction (Fig. 39-15).

Figure 39-10 plasma membrane specializations of striated muscles. A–B, Electron micrographs of thin sections of fish skeletal muscle showing invaginations called T tubules, which cross the whole muscle cell and associate closely with smooth endoplasmic reticulum (ER). The complex of a T tubule with smooth ER on both sides is called a triad. Foot processes, consisting of voltage-sensitive calcium channels in a T tubule paired with calcium release channels in the ER, connect the T tubule to the smooth ER. (Fig. 39-15 provides molecular details.) C–D, Drawings of the three-dimensional arrangement of T tubules and smooth ER relative to the sarcomeres in skeletal and cardiac muscle.

(A–B, Courtesy of C. Franzini-Armstrong and K. Porter, University of Pennsylvania, Philadelphia.)

Figure 39-15 mechanism of calcium release in skeletal and cardiac muscles. Both muscles use voltage-sensitive calcium channels in the T tubule membrane and calcium release channels in the smooth endoplasmic reticulum (SER). A, Direct coupling in skeletal muscle. An action potential in the T tubule (DV) activates the voltage sensor (turning from gray to blue). This direct contact opens the calcium release channel (turning from gray to pink). Cytoplasmic Ca²⁺ levels rise only briefly because calcium-ATPase pumps Ca²⁺ back into the lumen of the SER. B, Calcium-induced Ca²⁺ release in cardiac muscle. An action potential opens the voltage-sensitive Ca²⁺ channel in the T tubule, releasing Ca²⁺ into the cytoplasm. This Ca²⁺ opens the calcium release channel in the SER.

The Sliding Filament Mechanism

The key to understanding muscle contraction was the discovery that thick and thin filaments maintain constant lengths and slide past each other as sarcomeres (and the muscle) shorten (Fig. 39-11). About the same time, it was appreciated that cross-bridges (now recognized to be myosin heads) can connect actin and myosin filaments and that tension produced during contrac-tion is proportional to the overlap of actin and myosin filaments (Fig. 39-12). Supported by biochemical and ultrastructural evidence for actin-myosin interaction, these pioneering observations led to the theory that cross-bridges between the thick and thin filaments produce force for contraction. Fifty years of research on cross-bridges have yielded a detailed picture of the chemistry and molecular mechanics underlying the force-producing reactions. A review of the steps of the actomyosin-ATPase cycle (see Fig. 36-5) is helpful in understanding the contraction mechanism. Three different physiological states reveal information about cross-bridge mechanisms.

Figure 39-11 sliding filaments. Electron micrographs and interpretive drawings of longitudinal sections of a sarcomere from a relaxed muscle (A) and a contracted skeletal muscle (B). The lengths of the thin and thick filaments are constant as the sarcomere shortens, demonstrating that the filaments slide past each other during contraction. C, Cross-bridges between thick and thin filaments from a muscle in rigor.

(Micrographs courtesy of H. E. Huxley, Brandeis University, Waltham, Massachusetts.)

Figure 39-12 physiological properties of skeletal muscle. A, Dependence of maximum tension on the length of the sarcomeres. B, Interpretive drawings. Each relates to a point on graph A. C, Relationship of force and velocity during muscle contraction.

(A, Reference: Gordon AM, Huxley AF, Julian F: The variation in isometric tension with sarcomere length in vertebrate muscle fibres. J Physiol 171:28P–30P, 1964. C, From Ruch TC, Patton HD (eds): Physiology and Biophysics, 19th ed. Philadelphia, WB Saunders, 1965.)

Most myosin heads in contracting muscle have bound ATP or ADP-P_i and oscillate rapidly among the four “weakly bound” states illustrated in Figure 36-5. During some of the transient interactions of myosin-ADP-P_i with actin, phosphate dissociates from myosin, and the light-chain domain rapidly reorients (see Figs. 36-4 and 36-5). This stretches elastic elements in the myosin heads and both thick and thin filaments. Energy in these elastic elements can be used over a period of milliseconds to displace the actin filament relative to the cross-bridge and contract the muscle. When ADP dissociates from the actin-myosin-ADP intermediate, ATP rapidly binds to the actin-myosin complex, dissociating the cross-bridge and starting a new ATPase cycle.

Relationship of Cross-Bridge Behavior to the Mechanical Properties of Muscle

Under normal conditions, each sarcomere shortens less than 1 μm. However, the whole muscle shortens macroscopically because it has thousands of sarcomeres in series. For example, a human biceps muscle 20 cm long has about 80,000 sarcomeres in series from end to end. When each contracts 0.25 μm, the muscle shortens 2 cm. Because the system maintains a constant volume, each sarcomere and the whole muscle increase in diameter as they shorten. Although the individual filaments slide past each other relatively slowly (about 5 μm s⁻¹), muscles contract rapidly because the motion of each sarcomere in the series is added together. In our example, without resistance, the biceps contracts 2 cm in 100 to 200 ms.

The behavior of cross-bridges explains why the velocity of muscle contractions of an active muscle depends on the external load (Fig. 39-12). Contraction velocity is maximal when opposed by no load. Without a load, the molecular motion stored in elastic elements of each cross-bridge is largely converted into movement of actin filaments relative to myosin filaments. Under these conditions, the filaments in muscle slide past each other at a rate of about 5 μm s⁻¹, the same speed that is observed for free actin filaments moving over myosin heads in vitro (see Fig. 36-6). For this rapid sliding to occur, myosin heads that do not produce force must not impede movement. If bound tightly to actin, they would interfere mechanically with rapid sliding. This is avoided by the rapid equilibrium of the myosin intermediates between being bound to actin and being free. Myosin heads with bound ATP or ADP-P_i do not produce force when they bind transiently to a given actin subunit. Nor do these brief encounters retard sliding driven by force-producing cross-bridges.

Muscle produces maximum force when the contraction rate is zero (Fig. 39-12). The conformational change in the myosin head stretches elastic elements in the cross-bridge, but the force cannot overcome the resistance from the load on the muscle. Consequently, the filaments do not slide, and energy stored in each stretched elastic element is lost as heat when the cross-bridge dissociates at the end of the ATPase cycle. The maximum force depends on the numbers of sarcomeres in parallel, that is, the cross-sectional area of the muscle. Thus, muscles respond to strengthening exercises by growing in diameter.

Control of Skeletal Muscle by Motor Neurons

Neural stimuli that activate skeletal muscles arise in two ways (Fig. 39-14). In organisms with well-developed central nervous systems, most neural signals that activate skeletal muscles result from conscious decisions, providing voluntary control over skeletal muscles. Other signals result from reflex responses to stimula-tion of sensory nerves. Specialized muscle cells innervated with both motor and sensory nerves function as stretch receptors, relaying information about length and tension back to the spinal cord, where reflexes coordinate the motor neuron output. Neural inputs from both sources converge on motor neurons located in the brain stem and spinal cord of vertebrates. Axons of these motor neurons branch in the muscle to contact one or more muscle cells. A motor neuron together with its target muscle cells forms a motor unit. In the most precisely controlled muscles, such as the extraocular muscles, some motor neurons innervate single muscle cells.

The contractile activity of a muscle is graded in terms of the speed and force of the contraction, so individual muscles can produce both delicate and powerful movements. Nerve stimulation determines the contractile force in two ways: (1) The number of active motor units determines how many muscle cells produce force, and (2) the rate of stimulation adjusts the force produced by active cells. Every time a muscle cell is stimulated, all of the sarcomeres are activated, but the force that they produce increases as the rate of stimulation increases, up to a maximum of about 200 stimuli per second. The shortening velocity of an active muscle depends simply on force produced and the resistance (Fig. 39-12C). If a large force or high velocity of contraction is required, many motor units are called into action and stimulated repetitively. To sustain contraction, motor nerves fire repeatedly. By varying the number of active cells in a muscle and the rate of stimulation, the nervous system sets the force required for a particular movement.

Synaptic Transmission at Neuromuscular Junctions

The terminal branch of each motor neuron axon forms a synapse called the motor end plate or neuromuscular junction on the muscle surface (see Fig. 11-8). These nerve endings are filled with synaptic vesicles containing the neurotransmitter acetylcholine. Arrival of an action potential at the nerve terminal stimulates fusion of synaptic vesicles with the nerve plasma membrane, releasing acetylcholine into the cleft between nerve and muscle. In less than a millisecond, acetylcholine diffuses across the extracellular space and binds to acetylcholine receptors concentrated in the adjacent muscle plasma membrane. Acetylcholine binding opens the receptor cation channel, initiating a new action potential that spreads over the muscle cell plasma membrane and down into the T tubules.

Coupling Action Potentials to Contraction

An action potential in a T tubule triggers the release of Ca²⁺ from SER into the cytoplasm (Fig. 39-15). Ca²⁺ binding to troponin allows myosin to interact with the thin filament, initiating contraction. This signal transduction process is called excitation-contraction coupling. Ca²⁺ release in skeletal muscle is the best-characterized example of a general regulatory mechanism used by many cells (see Chapter 26).

Three transmembrane proteins located in the T tubule and the terminal cisternae of the SER cooperate to generate the transient Ca²⁺ signal (Fig. 39-16). The operation of this system is described after its components are introduced:

Figure 39-16 Ca²⁺ triggers contraction of skeletal muscle. In these experiments, the Ca²⁺-sensitive protein aequorin was injected into live muscle cells to provide a signal for the cytoplasmic Ca²⁺ concentration. A, Single stimulus. Cytoplasmic Ca²⁺ concentration increases transiently, followed by a short contraction. This brief contraction persists after cytoplasmic Ca²⁺ decreases to the resting level. B, Multiple stimuli. Each stimulus releases a new pulse of Ca²⁺, prolonging the contraction in so-called tetanus.

(Reference: Ridgway EB, Ashley CC: Calcium transients in single muscle fibers. Biochem Biophys Res Comm 29:229–234, 1967.)

An action potential in a T tubule results in a transient rise in cytoplasmic Ca²⁺, from 0.1 μM to about 2 μM (Fig. 39-16), in the following way. The action potential causes a short-lived conformational change in the DHP receptors that is transmitted directly to associated ryanodine receptor Ca²⁺ release channels. Many Ca²⁺ channels open transiently, allowing Ca²⁺ to diffuse down the steep concentration gradient from the SER lumen to cytoplasm. Physical connections between ryanodine receptors may spread their activation laterally, ensuring synchronous activation of a patch of channels. The structural changes in these channels that release Ca²⁺ are not yet understood.

After a single action potential, the rise in the cytoplasmic Ca²⁺ level lasts but a few milliseconds for three reasons. First, Ca²⁺ release channels close quickly. Second, cytoplasmic Ca²⁺ binds to troponin C and other proteins. Third, Ca²⁺ pumps efficiently transport cytoplasmic Ca²⁺ back into the lumen of the smooth endoplasmic reticulum, even before the muscle develops maximum force. Ca²⁺ pumps are continuously active, keeping the cytoplasmic Ca²⁺ concentration low. Re-peated action potentials are required to prolong the rise in cytoplasmic Ca²⁺ (Fig. 39-16B).

Transduction of the Calcium Spike into Contraction

Troponin-tropomyosin on thin filaments cooperates with myosin to turn on contraction in response to a Ca²⁺ spike. At rest, two Ca²⁺-binding sites of troponin C are largely unoccupied (owing to their low affinity for Ca²⁺ and the low Ca²⁺ concentration), and the troponin-tropomyosin complex partially blocks the binding site for myosin heads on actin (Fig. 39-17). This prevents most of the weak-binding myosin intermediates with ATP or ADP-P_i in the active site from binding the thin filament. When released into cytoplasm, Ca²⁺ binds troponin C, causing a conformational change that creates a binding site for a helical region of TNI. This interaction attracts the C-terminus of TNI away from actin and tropomyosin, allowing a small shift in the position of tropomyosin on the thin filament. This shift increases the probability that myosin-ADP-P_i heads will bind to the thin filament, dissociating their bound P_i and producing force. The initial binding of the first force-producing heads shifts the long tropomyosin molecule a bit farther away from the myosin-binding sites, allowing adjacent actins to interact freely with other myosin heads. The end-to-end association of tropomyosins facilitates this cooperative switch by exposing the myosin-binding sites on more distant actin subunits in the thin filament.

Figure 39-17 thin filament activation mechanism. Reconstructions from electron micrographs showing a short segment of thin filament (A) and a cross section of a thin filament (B). Ca²⁺ binding to troponin C partially activates the filament by moving tropomyosin away from its lateral position in relaxed muscle, where it overlaps the myosin-binding site on actin (red). Myosin binding to the partially activated filament moves tropomyosin further out of the way into the active position.

(Redrawn from data of W. Lehman, Boston University, Massachusetts.)

Thus, the combined influence of Ca²⁺ and myosin makes the thin filament receptive to myosin binding. Activation is cooperative because both Ca²⁺-binding sites on troponin C must be occupied, because the effects of Ca²⁺ binding and myosin binding are transmitted to neighboring tropomyosins through their end-to-end attachments, and because every myosin that binds accentuates the response. This cooperativity makes the on-off switch respond very sharply to a relatively small, 10- to 20-fold change in the cytoplasmic Ca²⁺ concentration. The efficiency of this switch is underscored by the fact that the energy consumption of a muscle cell increases more than 1000-fold when it is activated. Activation of slow skeletal muscle (Table 39-3) and cardiac muscle is less cooperative, as their troponin C has only one Ca²⁺-binding site.

Note the delay between the Ca²⁺ spike and the onset of tension (Fig. 39-16). The Ca²⁺-sensitive switch is sharp but slow owing to the slow response of thin filaments to Ca²⁺ binding. Note also that muscle continues to produce force well after the cytoplasmic Ca²⁺ concentration returns to resting levels. Ca²⁺ binds troponin C rapidly (milliseconds) but dissociates slowly (tens of milliseconds). Thus, the Ca²⁺ spike saturates troponin C, and the muscle remains active even after free Ca²⁺ has returned to the endoplasmic reticulum lumen. Force declines slowly as Ca²⁺ dissociates from troponin C and returns to the smooth endoplasmic reticulum without raising the cytoplasmic Ca²⁺ concentration.

A single action potential produces a short contractile “twitch” (Fig. 39-16). Maximum contractile force is produced by a series of closely spaced action potentials, leading to a sustained rise in cytoplasmic Ca²⁺ and prolonged activation of actomyosin. The extended contraction is called tetanus.

Regulation by Myosin Light Chains

The participation of skeletal muscle myosin light chains in the regulation of contraction varies among species. The skeletal muscles of mollusks are one extreme; myosin light chains bind Ca²⁺ and provide the main on/off switch for contraction. When the Ca²⁺ concentration is low in resting muscle, no Ca²⁺ binds to light chains, and the actin-myosin ATPase is off. Ca²⁺ that is released during activation binds to the light chains, turning on the ATPase and contraction. At the other extreme, the light chains of vertebrate skeletal muscle myosin do not bind Ca²⁺ and do not participate in activation. However, phosphorylation of vertebrate skeletal muscle light chains modulates contractile activity by increasing force production at suboptimal Ca²⁺ concentrations. Horseshoe crab skeletal muscle uses a dual system: Ca²⁺ binding to troponin-tropomyosin on thin filaments and Ca²⁺-regulated phosphorylation of myosin light chains both stimulate contraction.

The Contractile Apparatus

Smooth muscle cells are specialized for slow, powerful, efficient contractions under the control of a variety of involuntary mechanisms. Smooth muscle cells are generally confined to internal organs, such as blood vessels (where they regulate blood pressure), the gastrointestinal tract (where they move food through the intestines), and the respiratory system (where their excessive contraction contributes to asthma and other allergic reactions). The cytoplasm of spindle-shaped smooth muscle cells (Fig. 39-1) appears homogeneous by light microscopy because the contractile proteins are not organized in regular arrays like sarcomeres of skeletal and cardiac muscle. A basal lamina and variable amounts of collagen and elastic fibers surround each cell.

In terms of organization and biochemistry, smooth muscle cells (Fig. 39-20) resemble nonmuscle cells more than they do skeletal or cardiac muscle. For example, the gene for smooth muscle myosin arose relatively recently from a cytoplasmic myosin II gene (see Fig. 36-7). These myosins also share the same regulatory light chain. Long myosin thick filaments are interspersed among the thin filaments but not in a regular way like striated muscles. Thin filaments are composed of actin and tropomyosin, along with two regulatory proteins, caldesmon and calponin, rather than troponin. Thin filaments are arranged obliquely in the cell, some with their barbed ends attached to dense plaques on the plasma membrane, others to dense bodies in the cytoplasm. Like Z disks in striated muscles, dense bodies anchor desmin intermediate filaments, forming a continuous, inextensible, internal “tendon” running from end to end into the cell, preventing excess stretching (see Fig. 35-8).

Figure 39-20 contractile apparatus of smooth muscle. A, Electron micrograph of a thin cross section. B–C, Organization of the contractile units, which stretch across the cell between plasma membrane attachment plaques. Contractile units consist of myosin filaments connecting thin filaments attached to a dense body or plasma membrane plaque. D, High-power electron micrograph showing a dense body and cross sections of three types of filaments. E, Electron micrograph of a longitudinal section of an extracted vascular smooth muscle cell illustrating associations of actin filaments and intermediate filaments (IF) with dense bodies, and myosin filaments interacting with actin filaments.

(A, D, and E, Courtesy of A. V. Somlyo and A. P. Somlyo, University of Virginia, Charlottesville. References: Somlyo AP, Devine CE, Somlyo AV, Rice RV: Filament organization in vertebrate smooth muscle. Philos Trans R Soc Lond [Biol] 265:223–229, 1973; Bond M, Somlyo AV: Dense bodies and actin polarity in vertebrate smooth muscle. J Cell Biol 95:403–413, 1982.)

Smooth muscle cells contract like a concertina (Fig. 39-21) because tension generated by myosin and actin is applied to discrete spots on the plasma membrane. This compression can be seen in light micrographs as irregular cells with “corkscrew” nuclei. Given that smooth muscle cells have less myosin than striated muscle cells do, it is remarkable that they develop the same force. This is explained by two factors. First, the force-generating unit, the myosin filament, is larger in smooth muscle than in skeletal muscle. Deploying a given amount of myosin in large, thick filaments in a long sarcomere produces more force than does the same myosin in smaller filaments arranged in a series of short sarcomeres. Second, individual smooth muscle myosin molecules produce a larger force than skeletal muscle myosin, at least in vitro assays.

Figure 39-21 activation of smooth muscle contraction. A, The spindle-shaped smooth muscle cell becomes pleated as it contracts, owing to the attachment of the actin filaments at intervals along the plasma membrane. The graph shows the time course of activation, consisting of release of Ca²⁺ into the cytoplasm, phosphorylation of myosin regulatory light chains, and then the slow development of force. Myosin light-chain phosphorylation (LC-P) is required to initiate, but not to prolong, the contraction of smooth muscle. B, Biochemical pathways controlling phosphorylation of myosin regulatory light chains. Receptor stimulation leads to production of IP₃ by phospholipase C and release of Ca²⁺ into cytoplasm. Ca²⁺ binds calmodulin (CM), which activates myosin light-chain kinase (MLCK) by binding the kinase’s autoinhibitory peptide and displacing it from the active site. Active MLCK phosphorylates activating sites on the regulatory light chain. Light-chain phosphatase reverses phosphorylation of myosin. Activation of the small GTPase Rho with GTP stimulates Rho-kinase, which phosphorylates and inactivates light-chain phosphatase. This makes the system more sensitive to any level of Ca²⁺, as light-chain phosphorylation is prolonged. P-MLC-, phosphorylated myosin light chain.

(A, Redrawn from the work of K. Kamm and J. Stull, University of Texas Southwestern Medical School, Dallas.)

Regulation of Smooth Muscle Contraction

Stimuli that trigger smooth muscle contraction vary widely, but they all seem to act through seven-helix receptors coupled to trimeric G-proteins. Hormones stimulate contraction of the uterus, whereas motor nerves stimulate intrinsic eye muscles that close the pupil. Gap junctions couple some groups of smooth muscle cells, so their activation is propagated and coordinated within the tissue.

Depending on the muscle, Ca²⁺ for contraction enters the cytoplasm through either voltage-dependent calcium channels in the plasma membrane or IP₃ (inositol, 1,4,5-triphosphate)–induced Ca²⁺ release from smooth endoplasmic reticulum (Fig. 39-21). Drugs that block plasma membrane calcium channels can distinguish these two pathways experimentally. In intestines, parasympathetic nerves release acetylcholine to stimulate seven-helix muscarinic receptors (see Figs. 11-7 and 11-12). Associated trimeric G-proteins activate cation channels that depolarize the plasma membrane and allow Ca²⁺ to enter through voltage-sensitive calcium channels. Gap junctions couple gut smooth muscle cells, allowing excitation to spread from cell to cell. Calcium channel blockers strongly inhibit activation of gut smooth muscle. At the other end of the spec-trum, vascular smooth muscle depends on IP₃ to release Ca²⁺ from intracellular stores rather than from outside the cell.

Following stimulation, intracellular Ca²⁺ increases rapidly but transiently, declining to a value above resting level as the receptors desensitize (see Fig. 24-3). Ca²⁺ pumps in both smooth endoplasmic reticulum and plasma membrane clear the cytoplasm of Ca²⁺ so that Ca²⁺ levels decrease to resting levels and the muscle relaxes when the activating stimulus is removed. Relaxing agents, acting through cyclic guanosine monophosphate (cGMP) or cAMP (see Fig. 26-1), promote clearance of cytoplasmic Ca²⁺. Epinephrine relaxes smooth muscles of the respiratory system by another method. Stimulation of β-adrenergic receptors activates potassium channels that hyperpolarize the plasma membrane and reduce Ca²⁺ entry. This approach is widely used to treat asthma.

After a considerable delay (>200 ms) following the Ca²⁺ spike, contractile force develops slowly. The delay is attributable to the time required for a se-quence of three biochemical reactions: Ca²⁺ binding to calmodulin, calcium-calmodulin activation of myosin light-chain kinase (see Fig. 25-4), and phosphorylation of myosin regulatory light chains, turning on the myosin-actin ATPase cycle (Fig. 39-21). Unphosphorylated myosin-II from smooth muscle and vertebrate nonmuscle cells is inactive.

Phosphorylation of myosin light chains is required to initiate but not maintain contraction, so slowly cycling, unphosphorylated myosins maintain peak force with little expenditure of energy. Regulation of unphosphorylated cross-bridges is not well understood, but they appear to be activated cooperatively by a small population of phosphorylated myosin heads. Caldesmon, a calcium-calmodulin-binding protein associated with tropomyosin on actin filaments, may contribute to activation and/or allow myosin heads to cycle very slowly even in the presence of ATP.

The sensitivity of light-chain phosphorylation to Ca²⁺ depends on a parallel signaling pathway that partially inhibits myosin phosphatase, thus increasing the number of phosphorylated myosin cross-bridges and force at any given Ca²⁺ concentration (Fig. 39-21). Receptors coupled to trimeric G-proteins activate the small GTPase RhoA, which stimulates a protein kinase that inhibits myosin light-chain phosphatase. Malfunction of this Ca²⁺-sensitizing mechanism may contribute to some forms of high blood pressure, since in hypertensive animals, drugs that inhibit Rho-activated kinase relax smooth muscle and lower blood pressure.