Fibrous Proteins: Muscle and Molecular Motors (Advances in Protein Chemistry)

CONTENTS

PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

Comparative Motile Mechanisms in Cells John M. Squire and David A. D. Parry

I. II. III. IV.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cellular Locations and Roles of Motor Proteins . . . . . . . . . . . . Common Features in Many Motor Proteins . . . . . . . . . . . . . . . . How Do ATPases Produce Movement? . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 2 7 12 14

Molecular Architecture in Muscle Contractile Assemblies John M. Squire, Hind A. Al-Khayat, Carlo Knupp, and Pradeep K. Luther

I. II. III. IV.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Actin Filament Structure and the Z-Band . . . . . . . . . . . . . . . . . Myosin Filament Structure and the M-Band . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

17 19 34 51 79

Titin and Its Associated Proteins: The Third Myofilament System of The Sarcomere Henk L. Granzier and Siegfried Labeit

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Functional Genomics of Titin. . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Molecular Mechanism of Titin Elasticity and Its Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. The Titin Filament System in the Sarcomere and Its Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v

89 91 98 110

vi

CONTENTS

Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

114 114

Regulation of Muscle Contraction by Tropomyosin and Troponin: How Structure Illuminates Function Jerry H. Brown and Carolyn Cohen

I. II. III. IV. V. VI.

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Periodic Features. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aperiodic Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Turning on Troponin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Turning on the Thin Filament . . . . . . . . . . . . . . . . . . . . . . . . . . . Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

121 124 130 139 142 149 151

The Molecular Mechanism of Muscle Contraction Michael A. Geeves and Kenneth C. Holmes

I. II. III. IV. V.

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structure of the Crossbridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Crossbridge Polymorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biochemistry and Kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

161 165 166 178 188 189

X-Ray Diffraction Studies of Muscle and the Crossbridge Cycle John M. Squire and Carlo Knupp

I. II. III. IV. V.

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modeling of Rigor Muscle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Time-Resolved Events in Contracting Muscles . . . . . . . . . . . . . . X-Ray Interference Measurements and Their Implications . . . Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

195 222 230 234 246 247

CONTENTS

vii

Microtubules and Maps Linda A. Amos and Daniel Schlieper

I. II. III. IV. V. VI. VII.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microtubule Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dynamic Instability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structural MAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microtubule Destabilizers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proteins That Control Microtubule Location . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

258 258 268 272 279 282 290 291

The Structure of Microtubule Motor Proteins A. Marx, J. MU¨ller, and E. Mandelkow

I. II. III. IV. V. VI. VII.

Kinesin Classes, Domain Structure, and Nomenclature. . . . . . Kinesin-1 as Prototypical Motor . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison of Kinesin Structures . . . . . . . . . . . . . . . . . . . . . . . . Conformational Switching in Kinesin . . . . . . . . . . . . . . . . . . . . . Structures of Kinesin-Related Domains. . . . . . . . . . . . . . . . . . . . Dynein Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

299 301 310 328 333 334 338 340

Rotary Molecular Motors Stephan Wilkens

I. Introduction: The F-, V-, and A-ATPases and Their Function in the Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Overall Structural Features of the F-, V-, and A-ATPases . . . . . III. Mechanistic Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

346 352 362 372 372

viii

CONTENTS

Cytoskeleton Dynamics Powers Nematode Sperm Motility Murray Stewart and Thomas M. Roberts

I. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Comparison Between MSP and Actin . . . . . . . . . . . . . . . . . . . . . III. Locomotion in Nematode Sperm is Coupled to Assembly and Disassembly of the Cytoskeleton . . . . . . . . . . . . . IV. Reconstitution of Lamellipodial Protrusion In Vitro . . . . . . . . . V. Retraction is Also Required for Crawling . . . . . . . . . . . . . . . . . . VI. Reconstitution of Retraction In Vitro . . . . . . . . . . . . . . . . . . . . . . VII. A Push-Pull Model for Nematode Sperm Amoeboid Motility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

384 385 387 389 391 393 395 397 397

Structure and Mechanism of DNA Polymerases Paul J. Rothwell and Gabriel Waksman

I. II. III. IV. V. VI. VII.

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biological Diversity of DNA Polymerases. . . . . . . . . . . . . . . . . . . The Nucleotide Incorporation Pathway. . . . . . . . . . . . . . . . . . . . The E State: Basic Architecture of DNA Polymerases . . . . . . . . Primer/Template DNA Binding and Recognition. . . . . . . . . . . Formation of the E:p/t:dNTP Complex . . . . . . . . . . . . . . . . . . . Conformational Transition to a Catalytically Active Ternary Complex: The E’:p/t:dNTP Complex. . . . . . . . . . . . . . VIII. Phosphoryl Transfer Reaction, Product Release, and Translocation of the Primer/Template DNA . . . . . . . . . . . . . . . IX. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

402 403 406 409 414 417

AUTHOR INDEX. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

441 467

SUBJECT INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

419 428 430 431

COMPARATIVE MOTILE MECHANISMS IN CELLS By JOHN M. SQUIRE* AND DAVID A. D. PARRY{ *Biological Structure and Function Section, Biomedical Sciences Division, Imperial College London, London SW7 2AZ, United Kingdom { Institute of Fundamental Sciences, Massey University, Private Bag 11–222, Palmerston North, New Zealand

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Cellular Locations and Roles of Motor Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Common Features in Many Motor Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. How Do ATPases Produce Movement? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I.

1 2 7 12 14

Introduction

Cell maintenance, cell division, and wholesale movement of the cell body involve a variety of molecular motors. Some movement is along actin filaments and involves the action of one of the seventeen types of myosin (Geeves and Holmes, 2005; Squire and Knupp, 2005; Squire et al., 2005). The most highly organized arrangement of actin and myosin filaments occurs in muscle, which is a sophisticated and elaborately organized mechanical system also involving specialized elastic filaments known as titin (Granzier and Labeit, 2005). Different muscles also require careful regulation of activity as appropriate for their differing functions (Brown and Cohen, 2005). Other movement is along microtubules (Amos and Schlieper, 2005) and involves the motor proteins kinesin or dynein (Marx et al., 2005). Actin filaments or microtubules in each case act as filamentous tracks along which their appropriate motors step in processes controlled by adenosine triphosphate (ATP) hydrolysis. A special case of the microtubule system is found in eukaryotic flagella where dynein side‐arms walk along constrained arrays of microtubules to produce controlled and systematically varying bending and wave motions along the flagellum (Marx et al., 2005). Flagella in prokaryotes involve a very different structure; they do not contain microtubules, but are composed of the protein flagellin, with a complex rotary motor where the flagellum emanates from the cell body (Suzuki et al., 2004). Many of the myosin and kinesin motors are double‐headed. Some of these motors, like myosin in muscle, are nonprocessive, whereas others, like some double‐headed non‐muscle myosins and the kinesins, are processive. In other words, since many of these non‐muscle motors may work ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71001-3

1

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

2

SQUIRE AND PARRY

in isolation, double‐headed motor systems can stay attached to their track by at least one of their two heads at all times so that they and their cargo can progress along the track. Whether the motors are processive depends on careful control of the ATPase cycle (see discussions in Geeves and Holmes, 2005; and Marx et al., 2005). However, some myosins and some members of the kinesin family appear to be monomers, but they may still function as processive motors; perhaps they work together in groups in some other way. Another ATP‐driven motor—but this time a rotary, not a linear system—is exemplified by the F1F0ATPase in mitochondria and related structures (Wilkens, 2005). Here the motor can function in the forward or reverse direction, in one case using the energy from ATP hydrolysis to establish a proton gradient across the mitochondrial membrane, and in the other case using such a gradient in reverse to power the synthesis of ATP from adenosine diphosphate (ADP). Other motile systems involve aggregation of globular subunits into filaments. Actin and tubulin polymerization are well‐documented examples of this, but the action of the major sperm protein in nematodes (Stewart and Roberts, 2005) has similar characteristics despite forming filaments, which, unlike actin and microtubules, do not have a polarity. Finally, some of the proteins that help to process the necessary activities of nucleic acids in cells, the polymerases, also show motile features and are powered by ATP (Rothwell and Waksman, 2005). This article introduces these systems, provides a background for the detailed descriptions in later articles, and highlights some of the common molecular themes that recur throughout the volume.

II.

Cellular Locations and Roles of Motor Proteins

For obvious reasons, since its effects are appreciable on a macroscopic scale, the first motile system to be studied was muscle, with the first documented theory of muscle contraction postulated by Erasistratus (ca. 304–250 bc). This theory lasted for about 2000 years, until it was tested experimentally in the sixteenth century by Vesalius and was found to be wanting. It is now known, thanks largely to the efforts about 50 years ago of two unrelated Huxleys, Hugh Huxley (Huxley and Hanson, 1954) and Andrew Huxley (Huxley and Niedergerke, 1954), that muscles of all types contain the proteins myosin and actin and it is the relative sliding of these filaments that produces muscle shortening (see Squire et al., 2005). Actin forms long filaments and myosin motors ‘‘walk’’ along these filaments to produce force that creates this relative sliding. As discussed previously, the

FIBROUS PROTEINS: MOTILE SYSTEMS

3

notion of tracks along which molecular motors can walk is one that is well used in nature. Active motility occurs in every cell (Fig. 1), not just in muscle, and non‐muscle myosins (cytoplasmic myosins) carry cargoes along actin filaments (Baker and Titus, 1998). A number of other motors walk along microtubules (Amos and Schlieper, 2005). As described in Marx et al. (2005), these include the kinesin family, many of which transport cell organelles along microtubules in a processive manner (Hirokawa et al., 1998), and cytoplasmic dynein, which carries out a related role (Gibbons, 1995; Karki and Holzbaur, 1999). Different members of the myosin and kinesin families can even transport organelles in opposite directions along the same chosen tracks. Some cargoes appear to be transported along microtubules to their ends, and some of these tracks are very long, as in the axons of nerve cells (Fig. 1C). There they are transferred to actin tracks close to the cell membrane; clearly their action is coordinated. In some cases it appears that particular members of the myosin family can

Fig. 1. (A) Illustration of a generalized animal cell including many of the organelles found in muscle. (B) The structures of cilia and flagella in eukaryotes. (C) The structures of nerve cells with their very long axons down which cellular material needs to be transported. From Squire (1986).

4

SQUIRE AND PARRY

interact with one of the kinesin family to enable this change of track (Delacruz et al., 2003). In addition to the highly organized arrangement of actin and myosin filaments in muscle (discussed in detail in Squire et al., 2005), there are also highly organized arrays of microtubules in the cilia and flagella of eukaryotes that either pass fluid across the cell surface or act as the propellers for spermatazoa (Fig. 1B). Ordered arrays of actin also occur in microvilli on the surfaces of cells from the brush border of intestinal epithelia. Here, the highly cross‐linked actin arrays appear to be tensioned by interaction with myosin at the bases of the microvilli in the terminal web (Fig. 2B). The actin filament and microtubule tracks are both composed of globular proteins. Actin monomers (
43 kDa) aggregate in a helical fashion to produce actin filaments that are about
100 A˚ in diameter and very long (often in the 1‐ to 10‐mm range). Microtubules (Fig. 2A) are hollow tubes, about twice the diameter of actin filaments, composed of two very similar types of globular subunits, a‐ and b‐tubulin, each
50 kDa, which alternate to form the long, axially oriented, protofilaments that form the walls of the microtubules. There are often 13 protofilaments in naturally occurring microtubules. In cilia and eukaryotic flagella, microtubules are arrayed in a so‐called 9 þ 2 structure with the outer nine components being double microtubules, one complete and the other opened out and linked onto the back of the first one (Fig. 2C). Adjacent microtubule pairs interact through a single‐headed form of dynein (Marx et al., 2005), which shears adjacent outer microtubules axially and thus enables the flagellum to create a wavelike motion that drives the cell through the surrounding fluid. Bacterial flagella are quite different and are composed largely by the protein flagellin (Suzuki et al., 2004). Here there is a complex rotary motor at the base of the flagellum that drives the motion and shape of the hairlike flagellum in a variety of ways to control the movement of its parent cell. Other rotary motors occur in quite a different guise. Some are in the mitochondrial membrane, for example, where ATP hydrolysis in the F1F0ATPase rotary motor is related to proton exchange across the mitochondrial membrane. In fact, the F1F0ATPase is the first such motor for which much of the structure has been largely solved (Wilkens, 2005). The process of cell division is a nice example of many of these motors working together. Figure 3 illustrates this in a simple schematic. Here, from the parent cell (A), duplication of the centriole occurs (B), and the centrioles migrate to opposite ends of the cell (C). Chromosome duplication and condensation follow (D), the nuclear envelope breaks down, and the two sets of chromosomes are drawn to the centrioles at opposite ends of the cell (E). Cell cleavage then takes place (F and G), and two

FIBROUS PROTEINS: MOTILE SYSTEMS

5

Fig. 2. (A) The structure of microtubules. They usually consist of 13 rows of protofilaments of the proteins a‐tubulin and b‐tubulin, forming a closed cylindrical tube. (Left) View down the icrotubule axis shows its hollow center, part of which may be occupied by microtubule‐associated proteins (MAPS), which can also occur on the outer surface of the microtubule. (Right) Axial view shows the helical arrangement of the tubulin subunits. (B) The structure of microvilli, with their internal network of aligned, highly cross‐connected actin filaments. The terminal web (TW) contains myosin filaments as well as actin, and provides tension to stabilize the projecting microvilli. (C) The 9 þ 2 arrangement of microtubules with their dynein side‐arms in the axonemes of eukaryotic cilia and flagella. From Squire (1986).

identical daughter cells are formed. This process, therefore, consists of a number of distinct steps, all of which involve motors. In more detail, after replication of DNA, a process enabled and checked, among other things, by polymerases (Steitz, 1999; Rothwell and Waksman, 2005), the

6

SQUIRE AND PARRY

Fig. 3. Schematic illustrations of distinct steps in cell division show the central role of contractile motor action in the process of mitosis. (A) to (C) replication (prophase); (D) formation of the mitotic spindle (metaphase); (E) and (F) chromosome migration (anaphase) and building of the nuclear envelopes, and (G) formation of the contractile ring containing actin and myosin, forming the cleavage furrow and eventually two separate daughter cells. CE, centriole pair; A, aster of microtubules; N, nucleus; M, microtubules; C, chromosomes; K, kinetochore; NR, remnant of nuclear envelope; NE, nucelar envelope reforming; CR, contractile ring; CM, cell membrane]. From Squire (1986).

chromosomes are linked to the kinetocore from which kinetocore microtubules emanate. The centrioles at each end of the cell (pole) are surrounded by polar (astral) microtubules. Other interpolar microtubules may extend between one centriole and the other, although many do not reach the centrioles themselves. This whole assembly is the mitotic spindle, which has condensed chromosomes at its equator. The two sets of chromosomes are then drawn toward opposite centrioles, a process involving kinesins, while the centrioles themselves move apart, a process involving cytoplasmic dyneins. At the same time, new nuclear envelopes are assembled. Finally, cleavage of the cell into two daughters (cytokinesis) occurs by means of the contractile ring, which consists of actin and myosin filaments. At first the actin‐myosin ring forms a cleavage furrow, which then reduces in diameter to pinch off the two daughter cells. This process may be activated by a calcium transient that may turn on myosin light chain kinase, which itself can then activate the myosin II in the contractile ring.

FIBROUS PROTEINS: MOTILE SYSTEMS

III.

7

Common Features in Many Motor Proteins

One of the characteristic features of many motors that walk along tracks is that they are so‐called P‐loop ATPases (Goodson et al., 1994; Sack et al., 1999; Saraste et al., 1990; Sellers, 1999; Smith and Rayment, 1996). They are very close to the G‐proteins and have a similar arrangement of so‐called P‐loop, switch 1, and switch 2 regions. Myosin and kinesin are quite closely related; they are both members of the myosin–kinesin superfamily in the TRAFAC class of GTPases, with the third ATPase motor, dynein, a more distant relative in the AAAþ class of P‐loop NTPases (Leipe et al., 2002). For completeness, the structure of the high‐energy phosphate ATP is illustrated in Fig. 4. ATP hydrolysis consists of the cleavage of the terminal (g) phosphate group in ATP in the presence of water to give ADP and inorganic phosphate (usually referred to as Pi). ATP in muscle is usually associated with a chelated magnesium ion, which links to the last two phosphate groups. It is ATP hydrolysis to ADP that provides the energy for the action of these molecular motors. The P‐loop structure in crystals of Dictyostelium myosin II is shown in Fig. 5 (Smith and Rayment, 1996). This structure provides an environment where MgATP is coordinated with a large number of the surrounding amino acid side groups. For example, the Mg2þ ion, in addition to interacting with the two terminal phosphates of ATP, is coordinated directly to

Fig. 4. (A) The structure of ATP and (B) the process of ATP hydrolysis produce ADP and inorganic phosphate (Pi).

8

SQUIRE AND PARRY

Fig. 5. Illustration of the P‐loop (red), switch 1 (black), and switch 1I (gray) regions in myosin (sequence and numbering as in Dictyostelium myosin II). The locations of ATP (green) and the chelated Mg2þ ion (yellow) within this grouping are also indicated. For discussion see text.

Ser 237 and Thr 186, both through the side chain OH groups. It is also linked through water molecules to Asp 454 and Asp 455 on one side and Asn 235 (side chain O and backbone O) on the other. The terminal (g)‐phosphate group is coordinated with Ser 237 through the backbone O and NH groups, with Ser 236 through the side chain OH, with Asn 233 through the side‐chain NH2, and with Gly 457 through the backbone NH group. It is also linked through a water molecule to the backbone O of Gly 457 and the side chain O of Glu 459, which also links across to the NH2 of Arg 238, which in turn links to the OH of Ser 236. The first and second phosphate groups (a and b) of ATP are linked to the sequence Gly‐Lys‐Thr‐Glu (residues 184–187). Clearly, in the crystal, the entire region of binding of the phosphate groups, in the so‐called phosphate tube, represents a very tight‐knit set of interactions. It is considered very

FIBROUS PROTEINS: MOTILE SYSTEMS

9

unlikely that the triple‐phosphate region can thread its way into this phosphate tube, implying that the tube must be open before ATP can bind, and that binding causes closure of the tube. For example, Pate et al. (1993), using analogies of ATP with large moieties at the g‐phosphate position, showed that the phosphate tube must initially be rather open, since the large moieties appeared to be able to compete successfully with MgATP for the same binding site. Their binding then causes the movement (closure) of switch 2 toward the P‐loop and switch 1 (see Geeves and Holmes, 2005). In contrast to the tight constraints in the phosphate tube in the crystal structure, the links to the base and ribose groups are fewer. Tyr 135 (not shown in Fig. 5) forms a hydrogen bond with the adenine ring, and Asn 127 forms a hydrogen bond to the ribose ring. Both of these residues are totally conserved in the myosin II family. In addition, a few other interactions occur via water molecules. It is thought that the limited interactions with this part of ATP can explain why myosin can accommodate and hydrolyze many different nucleotides and also nonnucleotide triphosphates (Gulick et al., 1997; Pate et al., 1993; Ruppel and Spudich, 1996). The characteristic (consensus) sequence of P‐loops (the Walker A motif; Walker et al., 1982) is Gly‐x‐x‐x‐x‐Gly‐Lys‐Thr/Ser (the region in red in Fig. 5); this sequence is often used in bioinformatic searches to identify proteins related to this family. Each myosin and kinesin has a single P‐loop. For example, Dictyostelium myosin II has the sequence as in Fig. 5: (179) G‐E‐S‐G‐A‐G‐K‐T (186). On the other hand, dynein, with a heavy chain that partly forms a ringlike core complex of six AAAþ domains, has P‐loop motifs in the first four of these domains (e.g., G‐P‐A‐G‐P‐G‐K‐T). There may be a complex series of interactions between these various sites to generate movement, but the P‐loop in the third domain has been shown to be essential for dynein motor function (Silvanovich et al., 2003). The adjacent regions of the chain, in addition to the P‐loop (G1 in the G‐proteins), are the so‐called switch 1 and switch 2 regions (Fig. 5). Switch 1 is N‐x‐x‐S‐S‐R in the motor proteins (233 to 238 in Fig. 5 for Dictyostelium myosin II), and is D‐xn‐T (G–2) in the G‐proteins. Switch 2 is D‐x‐x‐G‐x‐E in the motor proteins (454 to 459 in Fig. 5), and D‐x‐x‐G (G-3) in the G‐proteins (Walker et al., 1982). In G‐proteins these two regions change conformation as a result of hydrolysis of guanosine triphosphate (GTP) to guanosine diphosphate (GDP). In myosin, switch 2 changes as a result of ATP hydrolysis, whereas switch 1 moves rather little. Note that in the kinesin crystal structure, unlike myosin, switch 1 is pulled away from the triphosphate region, whereas switch 2, the P‐loop, and the phosphates are close together. (See detailed discussions by Geeves and Holmes, 2005, on myosin; and Marx et al., 2005, on kinesin about how the action of ATP

10

SQUIRE AND PARRY

Fig. 6. (A) Schematic illustration of the mitochondrial F1F0ATPase (ATP synthase), which uses a proton gradient across the mitochondrial membrane in domain ‘‘a’’ to drive the rotation of the g‐subunit rotor inside a ring of six alternating a‐ and b‐subunits. These in turn synthesize ATP from ADP (see text). (B) The basal region of a bacterial

FIBROUS PROTEINS: MOTILE SYSTEMS

11

hydrolysis causes structural changes in this region, changes that are propagated through the different molecular motors to produce force and movement.) Note, finally, that the mitochondrial F1F0ATPase (ATP synthase) is a rotary motor, not a linear one, but it is associated with ATP synthesis from ADP. The main body of the motor (the F1 part) is a ring of six subunits, three a‐subunits and three b‐subunits, which alternate in position around the ring (Fig. 6A). All subunits have nucleotide binding sites, but it appears that the b‐subunits are mainly involved in ATP synthesis. Like myosins and kinesins, these binding sites also have a Walker A (P‐loop) motif structure (Walker et al., 1982) of sequence G‐G‐A‐G‐V‐G‐K‐T, and a Walker B region (R‐X6–8‐H4‐D), where X can be any residue and H here is any hydrophobic residue. In this structure (see Wilkens, 2005), there is a central spindle, the g‐subunit, which rotates in the ab ring such that at any instant the three b‐subunits are going through different phases of their ATP‐generating cycles. The g‐subunit spindle is linked to a trans‐membrane region (C10 in the mitochondrial ATP synthase); these are part of the F0 structure. The C10 acts as a rotor, which turns under the influence of the proton gradient across the mitochondrial membrane through the neighboring ‘‘A’’ subunit, which comprises several trans‐membrane helixes. Thus a proton gradient can itself produce a rotary force. The use of such a proton gradient to drive another system is evident in bacterial flagella (Fig. 6B). Here the rod of the flagellum passes out through a complex assembly of rings in the cytoplasmic membrane, the cell wall, and the outer membrane. Outside the cell the rod joins to a hook region, and then to the main part of the flagellar filament, which is composed of flagellin. The cytoplasmic end of the assembly, in addition to having a part that is involved in exporting flagellin through a central channel (which goes all the way to the tip of the flagellum), also is surrounded by a sheath of so‐called MotA or MotB torque‐generating units (Zhou et al., 1998). These are driven by a proton gradient (or a Naþ gradient in other systems) to induce the rotor protein FliG in the C‐ring or MS‐ring to rotate. How such proton gradients are actually capable of producing such torques remains to be seen (Zhou et al., 1998).

flagellum showing the rotor (the MS‐ring and c‐ring), which is driven to rotate by the proton gradient (or Naþ gradient in some systems) across the MotA or MotB units. Flagellin, from which the main part of the flagellar filament is built, is exported through the basal body and into the central channel, which extends to the tip of the flagellum.

12

SQUIRE AND PARRY

IV.

How Do ATPases Produce Movement?

Figure 7 summarizes the general structures of (A) myosin IIs, (B) kinesins, and (C) cytoplasmic dyneins that are found in association with dynactin. The myosin has a rod‐shaped region, which is an a‐helical coiled‐coil, on the end of which are two usually identical globular domains. Each globular domain has a motor region that binds ATP and interacts with actin and is linked back to the rod by a so‐called neck region or lever arm. It is thought to be movements of the lever arm on the actin‐attached motor domain that move the myosin rod past the actin filament. Some myosins, instead of having a long rod as in myosin II, have a coiled‐coil linker region and then a cargo‐binding domain, not unlike the arrangement shown in Fig. 7B for kinesin. Here two kinesin motor domains, which are each smaller than that of myosin, but where the ATP‐binding P‐loop regions are very similar to that in myosin, are joined through rather flexible linkers to a coiled‐coil stalk and a cargo‐binding domain. Kinesins may ‘‘walk’’ processively along microtubules in a type of hand‐over‐hand process. A third kind of translational motor, shown in Fig. 7C, is cytoplasmic dynein, which is a very complicated structure and relatively less well understood than myosin or kinesin. However, the six‐domain rings of dynein mentioned earlier, of which there are two copies in cytoplasmic dyneins (Fig. 7C), are linked to leverlike stalks that interact with and pull on the microtubule. The cargo that dynein must transport is first linked to dynactin, which binds both to dynein and to the microtubule. It is thought that the presence of dynactin also enables the dynein to be processive. The other kind of dynein, that associated with eukaryotic cilia and flagella, has only a single six‐domain ring, also with a projecting, leverlike, microtubule‐ binding stalk. At the other end is a stem that interacts with the cargo. Very informative electron microscopy and image processing has shown the possible relative movement of the stalk on the six‐domain ring (head) in this kind of dynein (Burgess et al., 2003). Examining these structures and the fact that they are all powered by ATP, the question remains as to how force is actually produced. Geeves and Holmes (2005) argue that myosin acts by the specific coupling between different myosin head states and different positions of the lever arm on the motor domain, so that, once attached to actin, the myosin acts as an ATP‐ driven motor where the energy released by ATP hydrolysis is directly coupled to the performance of mechanical work. However, Marx et al. (2005) argue that in some cases the kinesins appear to act as thermal ratchets. In this case, the attachment of a second head, once the first head has bound, is an event controlled by thermal motion, but, presumably for steric reasons, the head is more likely to bind to the microtubule in the

FIBROUS PROTEINS: MOTILE SYSTEMS

13

Fig. 7. General structures of a number of different motors: (A) myosin II interacting with actin, (B) kinesin carrying a cargo and interacting with a microtubule, and (C) cytoplasmic dynein, with its associated cargo‐laden dynactin complex, interacting with a microtubule.

14

SQUIRE AND PARRY

forward direction than in the reverse direction. There must then be some coupling of the attached state of the second head with the P‐loop ATP‐binding site of the first head to let it know that it can now let go. Once again this head will then find that it is easier to reattach in the forward position rather than staying where it is. In this case, the energy for the movement would be purely thermal, and the ATP hydrolysis cycle, rather than directly producing mechanical work, would simply provide a mechanism that tells the kinesin heads whether to bind to the microtubule or to detach. Do some cytoplasmic myosins also operate in this way? Are there different ways of using ATP in different motors depending on the particular system? Or are we missing something somewhere and, for example, are all ATPase motors actually thermal ratchets where the ATPase cycle is simply a controlling mechanism? Evidence presented in the following articles sheds light on these questions.

References Amos, L. A., and Schlieper, D. (2005). Microtubules and MAPs. Adv. Protein Chem. 71, 257–298. Baker, J. P., and Titus, M. A. (1998). Myosins: Matching functions with motors. Curr. Opin. Cell Biol. 10, 80–86. Brown, J. H., and Cohen, C. (2005). Regulation of muscle contraction by tropomyosin and troponin: How structure illuminates function. Adv. Protein Chem. 71, 121–159. Burgess, S. A., Walker, M. L., Sakakibara, H., Knight, P. J., and Oiwa, K. (2003). Dynein structure and power stroke. Nature 421, 715–718. Delacruz, J., Brown, J. R., and Langford, G. M. (2003). Interactions between recombinant conventional squid kinesin and native myosin‐V. Biol. Bull. 205, 188–190. Geeves, M. A., and Holmes, K. C. (2005). Molecular mechanism of muscle contraction. Adv. Protein Chem. 71, 161–193. Gibbons, I. R. (1995). Dynein family of motor proteins: Present status and future questions. Cell Motil. Cytoskel. 32, 136–144. Goodson, H. V., Kang, S. J., and Endow, S. A. (1994). Molecular phylogeny of the kinesin family of microtubule motor proteins. J. Cell Sci. 107, 1875–1884. Granzier, H. K., and Labeit, S. (2005). Titin and its associated proteins: The third myofilament system of the sarcomere. Adv. Prot. Chem. 71, 89–119. Gulick, A. M., Bauer, C. B., Thoden, J. B., and Rayment, I. (1997). X‐ray structures of the MgADP, MgADPgammaS, and MgAMPPNP complexes of the Dictyostelium discoideam myosin motor domain. Biochemistry 36, 11619–11628. Hirokawa, N., Noda, Y., and Okada, Y. (1998). Kinesin and dynein superfamily proteins in organelle transport and cell division. Curr. Opin. Cell Biol. 10, 67–73. Huxley, A. F., and Niedergerke, R. (1954). Structural changes in muscle during contraction: Interference microscopy of living muscle fibres. Nature 173, 971–973. Huxley, H. E., and Hanson, J. (1954). Changes in the cross‐striations of muscle during contraction and stretch and their structural interpretation. Nature 173, 973–976. Karki, S., and Holzbaur, E. L. (1999). Cytoplasmic dynein and dynactin in cell division and intracellular transport. Curr. Opin. Cell Biol. 11, 45–53.

FIBROUS PROTEINS: MOTILE SYSTEMS

15

Leipe, D. D., Wolf, Y. I., Koonin, E. V., and Aravind, L. (2002). Classification and evolution of P‐loop GTPases and related ATPases. J. Mol. Biol. 317, 41–72. Marx, A., Muller, J., and Mandelkow, E. (2005). The structure of microtubule motor proteins. Adv. Protein Chem. 71, 299–304. Pate, E., Franks‐Skiba, K., White, H., and Cooke, R. (1993). The use of differing nucleotides to investigate crossbridge kinetics. J. Biol. Chem. 268, 10046–10053. Rothwell, P. J., and Waksman, G. (2005). Structure and mechanism of DNA polymerises. Adv. Protein Chem. 71, 401–440. Ruppel, K. M., and Spudich, J. A. (1996). Structure‐function analysis of the motor domain of myosin. Annu. Rev. Cell Dev. Biol. 12, 543–573. Sack, S., Kull, F. J., and Mandelkow, E. (1999). Motor proteins of the kinesin family: Structures, variations and nucleotide binding sites. Eur. J. Biochem. 262, 1–11. Saraste, M., Sibbald, P. R., and Wittinghofer, A. (1990). The P‐loop: A common motif in ATP‐ and GTP‐binding proteins. Trends Biochem. Sci. 15, 430–434. Sellers, J. R. (1999). ‘‘Myosins.’’ 2nd ed. Oxford University Press, Oxford. Silvanovich, A., Li, M‐G., Serr, M., Mische, S., and Hays, T. S. (2003). The third P‐loop domain in cytoplasmic dynein heavy chain is essential for dynein motor function and ATP‐sensitive microtubule binding. Mol. Biol. Cell 14, 1355–1365. Smith, C. A., and Rayment, I. (1996). Active site comparisons highlight structural similarities between myosin and other P‐loop proteins. Biophys. J. 70, 1590–1602. Squire, J. M. (1986). ‘‘Muscle: Design, Diversity and Disease.’’ Menlo Park, CA: Benjamin‐Cummings. Squire, J. M., Al‐Khayat, H. A., Knupp, C., and Luther, P. K. (2005). Molecular architecture in muscle contractile assemblies. Adv. Protein Chem. 71, 17–87. Squire, J. M., and Knupp, C. (2005). X‐ray diffraction studies of muscle and the cross‐ bridge cycle. Adv. Protein Chem. 71, 195–255. Steitz, T. A. (1999). DNA polymerases: Structural diversity and common mechanisms. J. Biol. Chem. 274, 17395–17398. Stewart, M., and Roberts, T. M. (2005). Cytoskeleton dynamics powers nematode sperm motility. Adv. Protein Chem. 71, 383–399. Suzuki, H., Yonekura, K., and Namba, K. (2004). Structure of the rotor of the bacterial flagellar motor revealed by electron cryomicroscopy and single‐particle image analysis. J. Mol. Biol. 337, 105–113. Walker, J. E., Saraste, M., Runswick, M. J., and Gay, N. (1982). Distantly related sequences in the alpha‐ and beta‐subunits of ATP synthase, myosin, kinases and other ATP‐ requiring enzymes and a common nucleotide binding fold. EMBO J. 1, 945–951. Zhou, J., Lloyd, S. A., and Blair, D. A. (1998). Electrostatic interactions between rotor and stator in the bacterial flagellar motor. Proc. Natl. Acad. Sci. USA 95, 6436–6441.

PREFACE Fibrous proteins represent a substantial subset of the human proteome. They include the filamentous structures found in animal hair that act as a protective and thermoregulatory outer material. They are responsible for specifying much of an animal’s skeleton, and connective tissues such as tendon, skin, bone, cornea, and cartilage all play an important role in this regard. Fibrous proteins are frequently crucial in locomotion and are epitomised by the muscle proteins myosin and tropomyosin and by elastic structures like titin. Yet again the fibrous proteins include filamentous assemblies, such as actin filaments and microtubules, where these provide supporting structures and tracks for the action of a variety of molecular motors. In some cases, when things go wrong, they are involved in the generation of disease; the study of amyloids, prions, and b-proteins is not only fascinating, but has significant medical importance. It is nearly 20 years since this field was fully reviewed, and there have been very significant advances in that time. The present book, therefore, represents one of a set of three volumes in the Elsevier ‘Advances in Protein Chemistry’ series covering the entire fibrous protein field. These are entitled: Fibrous Proteins: Coiled-Coils, Collagen and Elastomers Fibrous Proteins: Muscle and Molecular Motors Fibrous Proteins: Amyloids, Prions and b-Proteins The present volume covers ‘Muscle and Molecular Motors’. The first few chapters describe the ultrastructures of striated muscles and of various muscle filaments (myosin, actin, titin), they discuss the regulation of muscle contractile activity, and they explore the mechanism of force production and movement. The book then sets out to survey other kinds of motor systems; microtubules and their interactions with both microtubule associated proteins (MAPs) and the motor proteins kinesin and dynein, the major sperm protein in nematodes, the rotary ATPases driven by or driving proton gradients, and the action of motor enzymes, polymerases, on nucleic acids. The aim throughout is to explore different molecular mechanisms of motor action in order to identify common themes. The complete set of three books will provide a compendium of up-tothe-minute information on the entire fibrous protein field. Each chapter, which is clearly written, fully illustrated, and with a comprehensive citation list, is by an acknowledged authority in the field. It is our hope that, ix

x

PREFACE

together, these books will enable valuable comparisons to be made, they will allow general principles to be elucidated and they will help to take the fibrous protein field forward in good shape into the 21st Century era of post-genomics, molecular medicine, and nanoscience. John Squire Imperial College London United Kingdom David Parry Massey University New Zealand

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES By JOHN M. SQUIRE,* HIND A. AL‐KHAYAT,* CARLO KNUPP,* AND PRADEEP K. LUTHER{ *Biological Structure and Function Section, Biomedical Sciences Division, Imperial College London, London SW7 2AZ, United Kindom; Biophysics Group, Department of Optometry and Vision Sciences, Redwood Building, Cardiff University, Cardiff CF10 3NB, Wales

{

I. II.

III.

IV.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Components and Organization of the Sarcomere . . . . . . . . . . . . . . . . . . . . . B. Actin Filaments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Vertebrate A‐Band Lattices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. The Sliding Filament Model and the Crossbridge Cycle . . . . . . . . . . . . . . . Actin Filament Structure and the Z‐Band . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. The Actin Monomer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. F‐Actin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. The Thin Filament and Troponin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Filament Organization in the Contractile Units of Different Muscle Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. The Z‐Band. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Filament Organization in the Vertebrate I‐Band . . . . . . . . . . . . . . . . . . . . . . Myosin Filament Structure and the M‐Band . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. The X‐Ray Diffraction Approach to Myosin Filament Structure . . . . . . . . B. Myosin Head Organization in Relaxed Vertebrate Myosin Filaments . . . C. Further A‐Band Analysis: C‐Protein, Titin, and the Vertebrate M‐Band .... D. Invertebrate Myosin Filaments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Crossbridge Arrangements on Different Myosin Filaments: Variations on a Theme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Conclusion: Implications about the Crossbridge Mechanism. . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I.

17 19 23 29 30 31 34 34 36 38 40 42 49 51 51 56 61 71 74 77 79

Introduction

The pioneering days of molecular biology in the 1950s and 1960s saw the determination of a variety of unknown molecular structures using an array of then novel techniques. The double helix of DNA was solved using high‐angle X‐ray fiber diffraction data (Watson and Crick, 1953; Wilkins et al., 1953). The technique of protein crystallography came of age when structures of key globular proteins were determined for the first time (Kendrew, 1963; Muirhead and Perutz, 1963). At the same time, electron microscopy and low‐angle X‐ray fiber diffraction helped to define the ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71002-5

17

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

18

SQUIRE ET AL.

Fig. 1. (A) Schematic illustration of the hierarchy of muscle. Skeletal muscle is composed of fibers about 20 to 100 mm in diameter and very long. Fibers in the light microscope appear cross‐striated and the muscles they come from are known as striated

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

19

structure of the muscle sarcomere in terms of separate actin and myosin filaments that slide past each other when a muscle shortens and that interact through cross‐connections (Huxley, 1957, 1969). Since then the muscle story has progressed in leaps and bounds so that in some ways it is now one of the best understood tissues. But this progress has been a story of using all of the techniques mentioned above in a correlated way; each method has illuminated the application of the other techniques and without any single one of them, our knowledge of muscle would be much the poorer. This article summarizes current knowledge about the major muscle components, the actin and myosin filaments. We also briefly discuss a third set of filaments, the titin filaments, which have remarkable properties and play a central role in integrating sarcomere structure (see Granzier and Labeit, 2005) We further describe how these filaments are organized in the muscle repeating unit, the sarcomere, through the cross‐linking M‐band and Z‐band structures. Finally, we discuss the filament arrangements in invertebrate muscles. This articles serves as an introduction to the detailed articles on titin (Granzier and Labeit, 2005), on muscle regulation (Brown and Cohen, 2005), and two articles on the contractile mechanism (Geeves and Holmes, 2005; Squire and Knupp, 2005).

II.

Hierarchy

Anyone eating a steak or a slice carved from roast beef knows that meat is fibrous in texture. These fibers,
20 to 100 mm in diameter and very long, are the multinucleate muscle cells of which skeletal muscles are composed (Fig. 1A, B). Such fibers in the light microscope appear cross‐ striated and the muscles from which they are derived are known as striated muscles. The term striated also covers the muscles in animal hearts (Fig. 1C), but here the cells (the myocytes) are much shorter, they contain a single nucleus, and they are linked end to end by special structures known as muscles. Striations are from repeating units, the sarcomeres, with A‐band and I‐band regions. Each sarcomere extends between successive Z‐bands and is about 2.2 to 2.3 mm long in a resting muscle (Bloom and Fawcett, 1975). (B) Groups of muscle fibers (F), showing they are multinucleated (N ), and composed of the myofibrils (MF). Myofibrils may be about 2 to 5 mm in diameter. (C) Representation of the muscle cell arrangement in animal hearts, showing similar striations to those seen in skeletal muscles, but here the cells (the myocytes) are much shorter, they contain a single nucleus (N ), and they are linked end‐to‐end by special structures known as intercalated disks (D), which provide mechanical and electrical continuity between cells. (D) A typical smooth muscle that can be found surrounding the blood vessels and various hollow organs apart from the heart. These visceral muscles do not have cross‐striations. M, mitochondria, Z, Z‐band, N, Nucleus.

20

SQUIRE ET AL.

intercalated disks, which provide mechanical and electrical continuity between cells. Other muscles in animals surround the blood vessels and various hollow organs apart from the heart. In vertebrates these visceral muscles are smooth muscles; they do not have cross‐striations (Fig. 1D). A closer look at striated muscle fibers shows that they themselves are assemblies of fine, hairlike structures known as myofibrils (Fig. 1A, B). Myofibrils may be about 2 to 5 mm in diameter, with cell organelles such as mitochondria and membranous systems called T‐tubules and the sarcoplasmic reticulum (SR) sandwiched between them (Fig. 2B). Vertebrate skeletal and cardiac muscles have a striated appearance because the myofibrils themselves are cross‐striated; they have a repeating unit along them known as the muscle sarcomere (Fig. 1A). In vertebrate muscles this repeat is approximately 2.2 to 2.3 mm long in a resting muscle but varies in length as the muscle stretches or shortens. The sarcomere, taken to extend between successive Z‐bands (Z‐discs) along the myofibril (Fig. 1A), is really the ‘‘business part’’ of the muscle where force is generated. Large muscles are assemblies of millions of sarcomeres all working together. Understanding the basic contractile mechanism in muscle therefore requires an understand how the sarcomere itself works. Sarcomere structure and function are the main topics of this article and those by Granzier and Labeit, Brown and Cohen, Geeves and Holmes, and Squire and Knupp in this volume. Before going into such detail, a description of some of the properties of muscle as a whole is warranted. Muscle innervation mechanisms and the kinds of contractile response that are produced are the topics of the rest of this section. Muscle contraction in the so‐called voluntary muscles, of which most skeletal muscles are examples, is initiated when a nerve action potential arrives at the nerve‐muscle (neuromuscular) junction (Fig. 2A). This in turn causes a depolarization of the muscle outer membrane, the sarcolemma, which causes the release of calcium ions into the interior of the muscle. However, muscle fibers are relatively large in diameter and it would take far too long for calcium to diffuse to the muscle interior to activate the centrally located myofibrils. For this reason, there are invaginations of the sarcolemma into the interior of the fiber. These invaginations are the T‐tubules (Fig. 2B). Thus, when the sarcolemma is depolarized, the depolarization is also propagated down the T‐tubules, which in turn interact with the terminal cisternae of the SR (Fig. 2B) to trigger the release of calcium ions locally into the adjacent myofibrils. The induced calcium release is mediated through the ryanodine receptor in the SR and subsequent sequestration of calcium when the muscle relaxes is accomplished by calcium pumps (Franzini‐Armstrong, 1999).

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

21

Fig. 2. (A) A set of fibers in a motor unit, activated by a single nerve axon also known as a motor nerve or motor neuron. (B) A group of muscle myofibrils showing the T‐tubular network, which is continuous with the fiber membrane (sarcolemma). When the sarcolemma is depolarized, this depolarization is also propagated down the T‐tubules, which in turn interact with the terminal cisternae (tc) of the sarcoplasmic reticulum to trigger the release of calcium ions locally into the adjacent myofibrils (Peachey, 1965). (C) Normalized tension records (twitch responses, see D) showing different contractile characteristics seen in different muscles and different fibers in the same muscle. (D) The tension response (pulse A) if a muscle is stimulated electrically by a very short pulse. In this twitch response the tension rises rapidly to a low value and then decays rapidly back to zero. If a second pulse arrives (pulse B) before the first twitch has finished, then there is a build‐up of tension so that the peak level in the second pulse is higher than the first. With repetitive pulses, the tension achieved gradually increases to a plateau level (D), which may be the maximum tension that the muscle can produce. This kind of maximal, sustained contraction is known as a tetanus and such a fused plateau response is known as a tetanic contraction. If the frequency is not quite high enough (C), the top tension level may still be bumpy and is known as an unfused tetanus. Only if the pulse train has sufficiently high frequency will successive peaks ‘‘fuse’’ to give a steady tension plateau (D).

22

SQUIRE ET AL.

A nerve axon activating a particular muscle fiber is known as a motor nerve or motor neuron (Fig. 2A). This axon can in fact be just one branch of a large nerve that also interacts with many other fibers in the muscle. The set of fibers activated by a single nerve is known as a motor unit. Fibers in a motor unit need not be adjacent to each other, but may be distributed through a muscle. If a muscle is stimulated electrically by a very short pulse, then the response is known as a twitch (Fig. 2D). The tension rises rapidly to a certain value and then decays rapidly back to zero. However, if a second pulse arrives before the first twitch has finished, then there is a buildup of tension so that the peak level in the second pulse is higher than the first (wave B in Fig. 2D). With repetitive pulses the tension achieved can gradually increase to a plateau level, which may be the maximum tension that the muscle can produce. However, the top tension level may still be bumpy (trace C in Fig. 2D). Only if the pulse train has sufficiently high frequency will successive peaks ‘‘fuse’’ to give a steady tension plateau (wave D). This kind of maximal, sustained contraction is known as a tetanus. The bumpy version at a lower frequency (C) is an unfused tetanus. A fused plateau response is known as a tetanic contraction. Note that in muscle jargon a contraction can occur without the muscle changing length. This is then an isometric contraction. Muscles can also be studied while shortening under constant load—an isotonic contraction. In everyday use the load and length of our muscles vary continuously, so the isometric and isotonic contractions often used by muscle researchers are useful ‘‘steady‐state’’ simplifications of what normally would be a constantly varying process. Finally, note in this quick overview of muscle properties that different muscles and different fibers in the same muscle may have different contractile characteristics (Fig. 2C). There are so‐called fiber types that can confer on individual muscles the particular responses that are needed for their differing functions (McComas, 1996). Some muscles are needed for posture, so they need to use energy slowly, they fatigue slowly, and they change length rather little. Other muscles are used for rapid spurts of activity and it may not matter that they tire quickly. The energy supply comes from the muscle mitochondria or from metabolism of glycogen granules. The speed of a muscle also depends on the particular isoforms of the main muscle proteins that they express. Different fibers may contain different isoforms of the same proteins (e.g., myosin or titin), which confer different contractile properties to their sarcomeres. In general, postural muscles are slower in their contractile response (e.g., with slow myosins), they have more mitochondria (they have a greater supply of the muscle oxygen‐carrier myoglobin which gives them a reddish colour), and they fatigue relatively slowly. Muscles required for rapid action have faster

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

23

isoforms of myosin and other sarcomeric proteins, fewer mitochondria, more glycogen, and they fatigue relatively rapidly. They are much less red in colour than slow fibers. Between these two extremes are wide gradations of fiber properties, and many muscles also contain a mixture of fiber types so that their responses are fine‐tuned to their functional role.

A. Components and Organization of the Sarcomere 1. Introduction The muscle sarcomere contains the principal contractile proteins myosin and actin (Fig. 3A to C), which on their own can produce force and movement, together with a number of cytoskeletal and regulatory proteins. The latter include titin, C‐protein (MyBP‐C), tropomyosin, troponin, a‐actinin, myomesin, M‐protein, and so on. Some of these help to organize the myosin and actin filaments in the sarcomere, some to define the filament lengths and structure, some to regulate activity, and some to modulate the actin–myosin interaction when the muscle is active.

2. Myosin Filaments Myosin filaments are composed of myosin molecules (Fig. 3B, C). It is now known that there are many varieties of myosin (currently about 17 types; http://www.mrc‐lmb.cam.ac.uk/myosin/trees/trees.html), which are used in a diverse panoply of cellular roles (see Squire and Parry, 2005; and Geeves and Holmes, 2005). Muscle myosin is myosin type II. Myosin II has a long two‐chain, parallel, coiled‐coil a‐helical rod about 1500 A˚ long and about 20 A˚ in diameter with a globular region or head on the N‐terminal end of each chain (Fig. 3C). One chain of the rod part of myosin, together with the rest of the same chain that forms the bulk of one of the heads, is known as the heavy chain (Fig. 4A). Two light chains, the so‐called essential light chain (ELC) and the regulatory light chain (RLC), are associated with each myosin head. The globular domain is an ATPase; it catalyses the hydrolysis of adenosine triphosphate (ATP) to adenosine diphosphate (ADP) and inorganic phosphate (Pi). This region of myosin is also the part that binds to actin and this binding enhances the ATPase activity of the head; the ATPase of myosin is said to be activated by actin. Studies of the myosin head, the globular part of the heavy chain together with the ELC and RLC, were dramatically transformed when the isolated myosin head from chicken skeletal muscle myosin was crystallized and its structure solved using protein crystallography by Rayment et al. (1993b; Fig. 4A). This showed that the head consists of a globular

24

SQUIRE ET AL.

Fig. 3. (A) Actin filament composed of actin molecules (A), two tropomyosin stands (TM ), and troponin molecule complexes (TN ). (B) Myosin filament composed of myosin molecules shown in (C) with the rod of the myosin molecules forming the backbone of the filament and the myosin heads are arranged on the surface of the filament backbone. (D) The bipolar packing of the myosin molecules showing the anti‐ parallel arrangement giving rise to a heads‐free bare zone region at the centre of the filament. This is also illustrated in (E). (F) Sarcomere structure extending between two successive Z‐bands (M, myosin, A, actin). (G–J) Cross‐sectional views through different parts of the sarcomere, showing (G) the square lattice of actin filaments in the I‐band; (H) the hexagonal lattice between overlapping arrays of actin and myosin filaments in the A‐band; (I, J) the hexagonal lattice of myosin filaments in the M‐band (I) and bare‐ zone ( J ) regions, with the extra M‐protein density linking the myosin filaments at the M‐region in the center of the sarcomere (I).

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

25

region, all part of the myosin heavy chain, containing the ATP‐binding site and the actin‐binding face, from which extends the next part of the heavy chain in the form of a very long a‐helical region, linking to the junction between the head and the myosin rod. Wrapped around this long a‐helix are the ELC and RLC, which serve both to stiffen the otherwise unsupported main chain a‐helix and in some muscles to regulate contractile activity. Because of its ATPase and actin‐binding properties, the globular part of the myosin head is sometimes called the catalytic domain, sometimes the motor domain. The thinner light‐chain binding region that joins the motor domain to the myosin rod is sometimes referred to as the neck region, or, for reasons that will become apparent later, the lever arm. In striated muscles, myosin molecules aggregate to form bipolar myosin filaments (see Fig. 3D). In the middle of each myosin filament the myosin rods pack in an anti‐parallel fashion, whereas outside this central region the rods pack parallel to each other. This gives the bipolar myosin filaments a head‐free region in the middle (the bare zone) on each side of which the heads are arrayed with opposite polarity on the filament surfaces in the bridge regions (see Fig. 3D, E). In different kinds of striated muscles, especially between vertebrate muscles and the striated muscles found in invertebrates, the size and symmetry of the bipolar myosin filaments can vary. However, in all vertebrate striated muscles so far studied, the myosin filaments all appear to be very similar; they are about 1.6 mm long, the bare zone length is just slightly longer (
1600 A˚ ) than the length of the myosin rod, and the myosin filament backbone is
140 A˚ in diameter with the heads projecting out from this. It was shown in the early studies of H. E. Huxley by electron microscopy and low‐angle X‐ray diffraction that there is a periodicity along the vertebrate myosin filaments that seemed to be associated with a helical array of myosin projections, now known to be the myosin heads (Huxley, 1963; Huxley and Brown, 1967). The repeat of the helical array was found to be 429 A˚ , with a sub‐periodicity of 143 A˚ . This was deduced from the spacings associated with layer‐lines in the X‐ray diffraction pattern from frog muscle (for details of muscle diffraction, see Squire and Knupp, 2005). The layer‐lines (ML1, ML2, ML3, etc.) appear as a set of parallel lines of intensity running at right angles to the muscle long axis in recorded diffraction patterns (Fig. 4B). The myosin heads were thought to give rise to layer lines with spacings that were orders of a repeat of 429 A˚ , with the third order (ML3) at 143 A˚ having particularly strong intensity on the central ‘‘vertical’’ axis of the diffraction pattern known as the meridian. This meridional reflection on the third layer‐line has been labeled the M3 reflection and is the subject of further discussion in Geeves and Holmes (2005) and Squire and Knupp (2005).

26

SQUIRE ET AL.

Fig. 4. (A) Structure of the myosin head, S1, determined by X‐ray crystallography for chicken skeletal muscle without nucleotide bound. It consists of a heavy chain forming a globular part where the actin‐binding site is located, and also a long a‐helical chain

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

27

Huxley and Brown (1967) originally suggested that the surface array of heads was described by two coaxial helices, each of pitch 2  429 A˚ , with a subunit axial translation along the helix of 143 A˚ and six head pairs in one pitch. Thinking of the helix as a spiral staircase, the subunit axial translation is the vertical rise on each step and the pitch is the height up the staircases where one is immediately over the lowest step. For further definitions of these terms and an introduction to helical diffraction theory, see the documentation for the HELIX or MusLABEL programs at www.ccp13. ac.uk (Knupp & Squire, 2004; Squire & Knupp, 2004). For more details of helical diffraction theory, see Squire (2000) or Chandrasekaran and Stubbs (2001). After this enormous step forward in myosin filament analysis, it was soon realized by Squire (1971, 1972), by comparison of myosin head arrays from different muscle types, that this postulated symmetry was likely to be wrong. In due course the evidence pointed to a three‐start (or three‐ stranded) helix of head pairs (Squire, 1972; see Fig. 3B, E), each helix having a pitch of 3  429 A˚ , the same subunit axial translation of 143 A˚ as Huxley and Brown had suggested, but with nine head pairs in one pitch. Because the number of strands in the helix and the number of head pairs in one helix repeat are both divisible by 3, the axial repeat of the new structure is still 429 A˚ as required by the X‐ray diffraction data. So, vertebrate muscle myosin filaments are three‐stranded and electron micrographs show that they have threefold rotational symmetry in any cross‐ section (Luther et al., 1981). The term crown is used to describe the ring of myosin heads within each 143 A˚ subunit axial translation. We show later that myosin filaments in invertebrate muscles have different rotational symmetries from threefold and in some cases different axial repeats, although all show the common ‘‘crown’’ subunit axial translation of
143 to 145 A˚ , which appears to be a signature of molecular packing between myosin II molecules. In the myosin‐containing part of the sarcomere, the A‐band, the myosin filaments are cross‐linked at the M‐band by various additional proteins (Fig. 3F, I)). Details of these are given later in Section II.C. Along the myosin filaments there is also part of the titin molecule (Fig. 5B, D). Titin is the largest known protein with a molecular weight of 3 MDa. It is anchored with its C‐terminus at the M‐band and its N‐terminus at the Z‐band. Titin shown in red surrounded by two light chains, the essential light chain (ELC) and regulatory light chain (RLC). (B) X‐ray diffraction pattern recorded at Spring-8 by Dr. J. J. Harford from relaxed fish muscle showing the myosin layer lines (ML) at orders of 429 A˚ repeat. The third‐order myosin layer line has a reflection on the meridian referred to as M3 at a spacing of 143 A˚ . Layer lines are sampled by the hexagonal lattice arrangement of actin and myosin filaments giving rise to (vertical) row‐lines (arrows).

28

SQUIRE ET AL.

Fig. 5. (A) Electron micrograph picture showing a whole sarcomere from fish muscle in relaxing conditions (Z to Z distance about 2.3 mm). (B) Schematic diagram showing the sarcomere with titin molecules (green and blue) with the N‐terminus of each titin molecule located at the Z‐band and the C‐terminus at the M‐band. (C) Arrangement of the C‐protein within the A‐band showing seven stripes on each

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

29

stretches across half a sarcomere or
1.1 to 1.2 mm at the muscle rest length. In the A‐band it is closely associated with the myosin filaments. It then extends from the A‐band, through the I‐band part of the sarcomere, to the Z‐band (Fig. 5D). In the I‐band region titin displays remarkable elastic properties; it confers on the sarcomere much needed structural integrity and helps to maintain the central location of the A‐band in the sarcomere. Particular regions of the I‐band titin of interest are the PEVK region and the N2 region (Fig. 5D; see Granzier and Labeit, 2005). At the ends of the myosin filaments, at the A‐band–I‐band junction, titin forms the so‐called end filaments (Fig. 5D). Also in the A‐band, in the middle of each of the bridge regions, the myosin filaments carry the extra protein originally known as C‐protein (now often called myosin‐binding protein C: MyBP‐C; Offer et al., 1973), which occurs in two sets of 7 to 9 stripes in the C‐zones in each half of the A‐band (Fig. 5C; Bennett et al., 1986; Sjostrom and Squire, 1977). The structure, interactions, and role of C‐protein are discussed in Section IV.C.

B.

Actin Filaments

Actin filaments were first visualized by Hanson and Lowy (1963) in one of the pioneering applications of the negative staining technique for electron microscopy. The filaments appeared as two strings of beads that gradually twisted around each other, the spacing between crossovers being
370 A˚ (Fig. 3A). Each bead was in fact one actin monomer of molecular weight
42 kDa. Since then actin filament structure has been studied in some detail as described later. At an early stage it was shown that the actin filaments in striated muscles carry the additional proteins tropomyosin and troponin, which are involved in regulating sarcomere activity (also shown in Fig. 3A). The tropomyosin molecule is rather like a short version of the rod part of the myosin molecule; it is a two‐chain, parallel, coiled‐coil a‐helical rod
400 A˚ long. It was also found that there is one tropomyosin molecule to one troponin complex and seven actin monomers (Ebashi et al., 1969). Since the axial repeat of the actin monomers along the filament was known from X‐ray diffraction to be
55 A˚ , it was suggested that the tropomyosin rods might lie along each of the twisting strings of actin beads. One troponin complex would then be associated with each tropomyosin molecule, giving an axial repeat of 7  55 A˚ or 385 A˚ half of the myosin filament. (D) Magnified version of the area between the Z‐band and I‐band showing the PEVK and N2 domains of titin and the end filament at the tip of the myosin filament.

30

SQUIRE ET AL.

for the whole tropomyosin/troponin assembly. Further details of this structure are given in Section III and in Brown and Cohen (2005).

C. Vertebrate A‐Band Lattices In a cross‐section through the A‐bands of vertebrate striated muscles the myosin filaments lie on a hexagonal lattice, and the actin filaments are at the so‐called trigonal points, midway between three mutually adjacent myosin filaments (see Fig. 3H–J). However, an early suggestion by Huxley and Brown (1967) was that there is actually a more complicated arrangement than this, a so‐called superlattice. This means that adjacent myosin filaments do not all have the same orientations. Since the symmetry that they perceived for the myosin filaments was two‐stranded, the superlattice they suggested was not quite right. A careful study of muscle cross‐sections in electron micrographs of frog sartorius muscle by Luther and Squire (1980) showed that in fact the threefold symmetric myosin filaments occur in one of two possible orientations 60 apart (Fig. 6B), forming a statistical superlattice. It was found to be impossible to generate the superlattice in a regular way, but, by applying two packing rules, Luther and Squire found that a statistical superlattice of the required size was automatically produced, even though it contained a large amount of disorder. About the same time it was also found that some vertebrate striated muscles, particularly those of bony fish, do not have a superlattice at all;

Fig. 6. (A) The simple lattice myosin filament arrangement that occurs in all the teleost (bony fish) muscles so far studied, in some muscles of cartilaginous fish and also in some primitive fish such as sturgeons and bowfin. (B) The superlattice arrangement present in the muscles of all the higher vertebrates, namely mammals (including humans), amphibians, birds, reptiles, and some muscles of cartilaginous fish.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

31

in these muscles all the myosin filaments have exactly the same rotations, forming what is known as the simple lattice (Fig. 6A). Because this simple lattice is so regular, it makes structural studies of bony fish muscle relatively advantageous; it is much easier to analyze structural data in a rigorous way (e.g., Harford and Squire, 1986; Hudson et al., 1997; Luther and Crowther, 1984). Luther et al. (1996) conducted a systematic study of the occurrence of the simple lattice and superlattice across the vertebrate kingdom. Superlattices are present in the muscles of all the higher vertebrates, namely, in mammals (including humans), in amphibians, in birds, in reptiles, and in some muscles of cartilaginous fish. Simple lattices occur in all the teleost (bony fish) muscles so far studied, in some muscles of cartilaginous fish, and also in some primitive fish such as sturgeons and bowfin.

D.

The Sliding Filament Model and the Crossbridge Cycle

At the time of writing it is exactly 50 years since the original major breakthrough in understanding sarcomere function, namely, the postulation of the sliding filament model whereby during muscle shortening myosin and actin filaments slide past each other without much change in length (Huxley and Hanson, 1954; Huxley and Niedergerke, 1954). Previously it had been assumed that actomyosin was some form of continuous filamentous structure extending throughout the sarcomere; a structure that would stretch or shorten in a similar way to the stretching of hair (keratin) under steam and the transition within the keratin chains from a short a‐helical structure to a longer b‐sheet structure. The idea of sliding filaments was revolutionary. The sliding filament model moved the central dilemma in muscle from ‘‘how does actomyosin shorten?’’ to ‘‘what makes the actin and myosin filaments slide past each other?’’ At an early stage it was realized that the myosin projections must be heavily involved in this process. Several studies then illuminated what must be happening. Reedy et al. (1965) showed that insect flight muscle (from Lethocerus) fixed for electron microscopy in two different states showed crossbridges between the myosin and actin filaments in the overlap part of the A‐band, but the angle of the crossbridges was different in the two states. Muscles fixed in the relaxed state showed crossbridges more or less at 90 to the myosin and actin filament long axes, whereas muscles fixed in the rigor state induced by the removal of ATP showed crossbridges at an angle closer to 45 . It was known that in the absence of ATP the myosin heads became firmly attached to actin. Only on the addition of ATP to such a system would the heads come off actin. ATP

32

SQUIRE ET AL.

Fig. 7. (A) The crossbridge cycle showing the way the myosin head attaches and detaches from actin filaments. In the rigor state (top right) heads are rigidly attached to actin in a specific conformation at a 45 angle forming the AM rigor complex. When ATP is added, the myosin head is released from actin (1) and hydrolysis of ATP into its products, ADP and Pi, occurs, with both products still attached to the head (2). The hydrolysis of ATP is assumed to be accompanied by a conformational change of the heads from 45 to 90 . It is at the M.ADP.Pi state that can rebind to actin (step 3) with the heads still at a 90 angle and forming AM.ADP.Pi. The transition from AM.ADP.Pi to AM.ADP to AM, possibly with some isomerization steps within each state, is associated with force production and movement. The swinging of the elongated attached head

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

33

loss gives rise to the rigid state that occurs on death when the ATP supply stops, the state known as rigor mortis. The nucleotide‐free state of myosin heads on actin in the crossbridge cycle is therefore known as the rigor state. This and other evidence was put into a biochemical version of the crossbridge cycle by Lymn and Taylor (1971; shown in Fig. 7A), with a structural interpretation summarized by Huxley (1969) involving the 90 to 45 tilt of the actin‐attached myosin heads as seen by Reedy et al. (1965). Here the elongated myosin head (M; at that stage the detailed head shape in Fig. 4A was not known) is shown bound to actin (A) in the rigor, nucleotide‐free, state (AM). Subsequent binding of ATP brings the attached head rapidly off actin (the M.ATP state). The ATP on the detached head then hydrolyses to ADP and Pi (M.ADP.Pi), but the products ADP and Pi remain on the myosin. If the muscle is not switched off, a myosin head as M.ADP.Pi can reattach to actin to give the AM.ADP.Pi state. In Fig. 7A the head is shown attaching to actin at a 90 angle. It is the transition from AM.ADP.Pi to AM. ADP to AM, possibly with some isomerization steps within each state as well, that is associated with force production and movement. Clearly a swinging of the elongated attached head from 90 to 45 causes relative sliding of the myosin and actin filaments. An early test of the sliding filament model was the very careful measurement by Gordon et al. (1966) of the active tension produced by the muscle at different sarcomere lengths (Fig. 7B–D). If the myosin heads or crossbridges act as independent force generators, then, as the sarcomere length from 90 to 45 will cause relative sliding of the myosin and actin filaments. (C) The active tension produced by the muscle at different sarcomere lengths (B, C, D from Gordon et al., 1966) is shown in (B, D). If the myosin heads or crossbridges act as independent force generators, then, as the sarcomere length (S) is increased and the overlap of the actin and myosin filaments reduces, the tension produced by the muscle should gradually reduce in proportion to the overlap. A linear reduction in tension was observed as the sarcomere length changed from about 2.2 mm to about 3.6 mm labeled as (1). Since the actin filaments are about 1 mm long and separated by an estimated Z‐band thickness of 0.05 mm, and since the myosin filament length is about 1.6 mm, it would be expected that there would be zero overlap and hence zero tension when S  3.65 mm (¼ 1.6 mm þ 1.0 mm þ 1.0 mm þ 0.05 mm). As the sarcomere length is reduced the overlap will gradually increase until the two bridge regions of the myosin filaments are fully overlapped by actin. This will occur at a sarcomere length of about 2.25 mm (2  1.0 mm for the actin filaments plus 0.05 mm for the Z‐band, plus the size of the bare zone of be about 0.2 mm, labeled as (2). Reduction of S below this value would not increase the overlap any further so there will be an active tension plateau as observed between 2 and 3. After this there are complications to the simple analysis; first the actin filaments meet the M‐band, then there is overlap of anti‐parallel actin filaments, then the actin filaments start overlapping myosin bridge regions with the wrong polarity in the other half of the A‐band, and finally the myosin filaments bump up against the Z‐bands, so the observed tension gradually reduces below S ¼ 2.0 mm.

34

SQUIRE ET AL.

(S) is increased and the overlap of the actin and myosin filaments reduces, the tension produced by the muscle should gradually reduce in proportion to the overlap. The tension in a muscle that is not stimulated, known as the resting tension, also changes with sarcomere length, so in the analysis by Gordon et al. the resting tension was subtracted from the total tension for a given sarcomere length to give the active tension. Their result is shown in Fig. 7C. They observed a beautifully linear reduction in tension as the sarcomere length changed from
2.2 mm to
3.6 mm. Since the actin filaments are
1 mm long and separated by an estimated Z‐band thickness of 0.05 mm, and since the myosin filament length is
1.6 mm, it would be expected that there would be zero overlap and hence zero tension when S  3.65 mm (¼ 1.6 mm þ 1.0 mm þ 1.0 mm þ 0.05 mm). As the sarcomere length is reduced, the overlap gradually increases until the two bridge regions of the myosin filaments are fully overlapped by actin. This occurs at a sarcomere length of
2.25 mm (2  1.0 mm for the actin filaments plus 0.05 mm for the Z‐band, plus the size of the bare zone taken by Gordon et al. to be
0.2 mm). Reduction of S below this value would not increase the actin overlap with myosin heads any further, so there will be an active tension plateau as observed (Fig. 7C). After this point, there are complications to the simple analysis. First, the actin filaments meet the M‐band; then there is overlap of anti‐parallel actin filaments; the actin filaments then start overlapping myosin bridge regions with the wrong polarity in the other half of the A‐band; and finally the myosin filaments bump up against the Z‐bands, so the observed tension gradually reduces below S ¼ 2.0 mm. (Note that these exact sarcomere length positions depend on the precise actin filament length, which can vary slightly between different vertebrate muscle types, and on the thickness of the Z‐bands and bare zones, which are discussed in later sections. The A‐band length always appears to be the same in vertebrate striated muscles.) In a later study, Huxley and Simmons (1971a,b) used tension records from single muscle fibers undergoing very rapid step changes in length to elucidate details of the crossbridge mechanism. The crossbridge properties that they observed also scaled exactly with sarcomere length. There could be little doubt that the myosin heads were acting as independent force generators.

III.

Actin Filament Structure and the Z‐Band A. The Actin Monomer

As in the case of the myosin head, knowledge of actin filament structure, or thin filament structure as it is termed when tropomyosin and troponin are present, also progressed rapidly when the structure of the globular actin (G‐actin) monomer was determined by protein crystallography in

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

35

the work of Kabsch et al. (1990), who studied co‐crystals of actin bound to DNase 1. Since 1990 a number of other G‐actin crystal structures have been published (Chik et al., 1996; Graceffa and Dominguez, 2003; Otterbein et al., 2001), but the structures are essentially the same. The actin monomer (Fig. 8) is a four‐domain structure with two large domains known as 1 and 3 and two smaller domains 2 and 4. The N‐ and

Fig. 8. Illustration of the actin monomer structure solved by X‐ray crystallography (Kabsch et al., 1990) showing the four structural subdomains of the actin monomer, labeled, in the front (A) and back (B) view with N, C‐termini labeled 1–4. Also labeled is the loop linking residues 262 and 309 between subdomains 3 and 4.

36

SQUIRE ET AL.

C‐termini of the protein chain are in subdomain 1. The chain then goes up into subdomain 2, back down to subdomain 1, across to subdomain 3, up to subdomain 4, back down through subdomain 3, and across to subdomain 1 again. In some ways the two halves of the monomer, subdomains 1 and 2 in one half and subdomains 3 and 4 in the other half, are rather similar in structure. The largest subdomains (1 and 3) contain a central five‐stranded b‐sheet structure with short a‐helices on each face. The smaller subdomains (2 and 4) are primarily b‐sheet with a small amount of a‐helix in subdomain 2 (the smallest subdomain) and slightly more a‐helix in 4. Centrally placed between the two halves of the monomer there is an ATP‐binding site. Actin is itself an ATPase, and the state of hydrolysis of ATP is an important factor in determining actin filament growth in cell motility and other general cell functions (see review in Sheterline et al., 1998).

B. F‐Actin G‐actin monomers in muscle aggregate to form filamentous actin (or F‐actin) filaments as shown in Fig. 9B. This shows in molecular detail the twisting strands of globular beads seen originally by Hanson and Lowy (1963; cf. Fig. 3A). Holmes et al. (1990) modeled the G‐actin structure into filaments with the correct helical symmetry for F‐actin and tested such models against high‐angle fiber diffraction data from oriented gels of F‐actin to
8 A˚ resolution. The actin monomer is probably oriented with subdomains 3 and 4 on the inside of the filament where they would interact with subdomains 3 and 4 from other monomers. Subdomains 1 and 2 were placed toward the outside of the filament. It was later shown that the primary binding site for the myosin head is, in fact, on subdomain 1, so the exterior location of this subdomain makes good sense. This model for F‐actin, which must still be considered as a model since it has not been demonstrated at high resolution, has been tested in several studies (e.g., Al‐Khayat et al., 1995; Lorenz et al., 1993; Oda et al., 1998), and it is generally agreed that the Holmes et al. model is more or less correct. Subsequent studies have refined the structure slightly, but its essential features remain (Holmes et al., 2003). The structure of F‐actin can be described in terms of a helix of actin monomers with approximately 13 monomers in 6 left‐handed turns of the genetic helix (a 13/6 helix). The helix pitch is 59.6 A˚ , and the subunit axial translation is 27.5 A˚ . The fact that the pitch is only slightly different from twice the subunit axial translation gives rise to the very slow twist of the filament as observed by Hanson and Lowy (1963). The twist of this so‐called long‐period helix is right‐handed. The crossover repeat for a 13/6 actin helix is 13  27.5 A˚ ¼ 357.5 A˚ (the pitch of each long‐period

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

37

Fig. 9. (A) Actin monomer with each of its four structural subdomains shown in different colors: subdomain 1(red); subdomain 2 (green); subdomain 3 (blue); and subdomain 4 (yellow). (B) The helical arrangement of the actin monomers along the actin filament, F‐actin, according to Holmes et al. (1990). (C) Tropomyosin molecules consist of a two‐chain a‐helical coiled‐coil. (D) Crystal structure of Tn‐C (Herzberg and James, 1985, 1988). (E) Schematic of the whole troponin complex with a globular region composed of Tn‐C and Tn‐I and a rod region of Tn‐T. (Diagram modified from Squire and Morris, 1998.)

strand is twice this value). Note that actin filaments can have different twists from this depending on their source and on which other proteins are bound to them. In insect flight muscle the actin filament symmetry is that of a 28/13 helix. The subunit axial translation is still 27.5 A˚ , but the crossover repeat is now 385 A˚ and the long‐period pitch is 2  385 A˚ . Actin filaments with various non‐muscle actin‐binding proteins on them can have crossover repeats as little as 320 A˚ (e.g., McGough et al., 1997). Note also that even in a particular actin filament type the crossover position may be rather variable. Actin filaments have been observed to have what is

38

SQUIRE ET AL.

known as random variable twist, meaning that they are azimuthally rather flexible and that the quoted crossover spacing is an average value (Egelman et al., 1982). One aspect of the Holmes et al. structure, which was not in the G‐actin structure but was modeled by hand since it seemed to make sense, is the loop between residues F262 and I309 situated at the interface between subdomains 3 and 4 (see Fig. 8). The F‐actin structure as originally modelled seemed to have rather few links across from one strand to the other, which would stabilize the structure. Holmes et al. suggested that this loop might remodel itself and insert between subdomain 4 of the monomer one below along the genetic helix (i.e., in the other long‐period strand) and subdomain 3 of the monomer one above in the genetic helix. These proposed cross‐linking loops can be seen across the gap between the two long‐period strands in the F‐actin model of Fig. 9B.

C. The Thin Filament and Troponin The components of the full thin filament, including the regulatory proteins tropomyosin and troponin, are shown in Fig. 9. Tropomyosin (Fig. 9C) consists of two chains of molecular weight 32.8 kDa (Sodek et al., 1972; Stone and Smillie, 1978) folded into a parallel coiled‐coil a‐helical structure. There is a pseudo‐repeat in the sequence along the tropomyosin chains, yielding reasonably equivalent sequences where it interacts with the seven actin monomers in the 385 A˚ repeat (Parry and Squire, 1973; Stewart and McLachlan, 1975). The sequence is such that there are 14 bands of acidic residues that occur in two alternating types called a and b (McLachlan and Stewart, 1976). Since the twisting tropomyosin coiled‐ coil is labeling seven actin monomers that themselves lie on helical tracks, it turns out that to an outside observer the tropomyosin strands only complete six half turns in the 385 A˚ repeat. This gives the tropomyosin coiled‐coil a pitch of
385/3 ¼ 128 A˚ , which is similar to those of other fibrous proteins such as paramyosin, a‐keratin, and honeybee silk (see discussion in Squire, 1981; Geeves and Holmes, 2005). The 385 A˚ thin filament repeat also contains a single copy of the troponin complex along each strand (Fig. 9E). This comprises troponin‐I (TnI) of molecular weight 23 kDa, which on its own can inhibit contraction, troponin‐C (TnC) of molecular weight 18 kDa, which reversibly binds to Ca2þ ions in the physiology range of concentrations and was the first EF‐hand calcium‐binding protein to be solved by protein crystallography (Herzberg and James, 1985, 1988; Fig. 9D), and troponin‐T (TnT) of molecular weight 30.5 kDa, which binds to tropomyosin and to TnC and TnI. The whole troponin complex is known to be elongated,
200 A˚ long, with a

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

39

globular region at one end that contains TnC, TnI, and part of TnT (known as T2), with the rest of TnT (T1) forming a long tail. The structure of the globular part of troponin has been the subject of many studies, including nuclear magnetic resonance (NMR; Lindhout and Sykes, 2003) and solution scattering (King et al., 2005; Stone et al., 1998) and has led in recent times to partial crystal structures (Takeda et al., 2003). The full troponin structure and the nature of its interactions with tropomyosin and actin are gradually emerging. Brown and Cohen (2005; this volume) present a discussion of the latest findings on troponin and tropomyosin structures. Note that it was shown by Huxley (1972), Haselgrove (1972), and Parry and Squire (1973) that the effect of Ca2þ binding to troponin when calcium is released from the SR following arrival of a nerve‐stimulated action potential along the T‐tubules could be explained by a change in the position of tropomyosin on the long‐period actin strands, thus uncovering or at least modifying the binding site on subdomain 1 of actin where the myosin heads attach to go through their force‐generating cycle on actin. This model, known as the steric blocking model, with the implication that tropomyosin regulates by virtue of its changing position on actin, has gained general acceptance, although there are undoubtedly subtleties, foreseen in Parry and Squire (1973), such as the simultaneous propagation of information through the actin monomers, which have yet to be fully described. Vertebrate striated muscle sarcomeres contain two remarkable molecular rulers. One of these is titin, mentioned briefly earlier (see Fig. 5B and D), which runs from the Z‐band to the M‐band and provides the sarcomere with mechanical continuity. Titin is discussed further in Section III.F. The other is nebulin, which is an I‐band protein anchored in the Z‐band. Nebulin is largely a‐helical, has a repeating sequence along it with a 55 A˚ axial repeat (assuming it is 35 residues in an a‐helix), which may fit in with the axial repeat along the long‐period strands in actin filaments; it seems to run the whole length of skeletal muscle actin filaments, and it may define the thin filament length. The stoichiometry of nebulin to actin is not yet known, but there appears to be at least one and probably two chains to each actin filament (Zhang et al., 1998). Because of the two strands of the actin filament, the number of nebulins might also be expected to be two (or a multiple). However, as discussed later, neither the Z‐line nor the actin filament itself has exact twofold symmetry, so the number of nebulins need not necessarily be even. At the actin filament tip (the pointed end away from the Z‐band) there is a capping protein called tropomodulin (Krieger et al., 2002; Littlefield and Fowler, 1998) that also lacks twofold symmetry. Finally, there is at least one more molecular component of muscle thin filaments that should be mentioned. This is phalloidin, which does not

40

SQUIRE ET AL.

bind to G‐actin, but binds to and stabilizes F‐actin, where it also inhibits release of Pi from actin‐hydrolyzed ATP and it resists actin depolymerization (Dancker & Hess, 1990; Dancker et al., 1975; Estes et al., 1981). Phalloidin location on F‐actin was modeled by Lorenz et al. (1993) to lie at the contact region of three adjacent actin monomers between the two long‐period actin strands and thus to form a direct stabilizing link across the filament. Fiber diffraction studies of oriented F‐actin gels under various conditions and with different ligands have been performed by Oda et al. (1998, 2005) and further work in press (Oda et al., 2005).

D. Filament Organization in the Contractile Units of Different Muscle Types The A‐band lattices in different kinds of striated muscles have distinct arrangements. As shown in Fig. 3 and reproduced in simpler form in Fig. 10A and B, vertebrate striated muscle A‐bands have actin filaments at the trigonal points of the hexagonal myosin filament array. As discussed prevously, this array also occurs in two types, the simple lattice and superlattice. The ratio of actin filaments to myosin filaments in each unit cell is 2:1. In both cases the center‐to‐center distance between adjacent myosin filaments is
470 A˚ , but this varies as a function of overlap, becoming smaller as the sarcomere lengthens, giving an almost constant volume to the sarcomere (April et al., 1971). Insect fibrillar flight muscles (IFM) have a highly developed stretch‐ activation mechanism and contract in an oscillatory fashion at a frequency determined by the resonance of the wing thorax assembly that is higher than the nervous stimulus (Pringle, 1957). For this reason they are sometimes referred to as asynchronous muscles. In IFM sarcomeres the filament arrangement is different from vertebrates. As shown later the myosin filaments have fourfold rotational symmetry, they are longer than vertebrate myosin filaments, about 2 mm, and although the myosin crown spacing (the subunit axial translation) is almost the same (145 A˚ ), the filament repeat is 1160 A˚ rather than 429 A˚ (Al‐Khayat et al., 2003; Morris et al., 1990; Reedy, 1968). In the IFM A‐bands there is also a hexagonal array of myosin filaments, but the actin filaments are midway between two adjacent myosin filaments giving a thin‐to‐thick filament ratio of 3:1 (Fig. 10C). The center‐ to‐center myosin filament spacing is larger than in vertebrate muscles at
520 A˚ . Different kinds of insect asynchronous flight muscles have slightly different arrays of myosin filaments within the hexagonal lattice (e.g., between Drosophila [fly] and Lethocerus [water bug]), but the myosin filaments all seem to have fourfold rotational symmetry. In the water bug, the myosin filaments appear to have their crossbridge arrays in orientational

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

41

Fig. 10. (A–F) A‐band filament lattices in different striated muscles showing the threefold myosin filaments in vertebrates, the fourfold myosin filaments in some invertebrates, and the sevenfold myosin filaments in scallop striated adductor muscle. The ratio of actin to myosin filaments within the different lattices is also shown. Two smooth muscle types are also illustrated; the face polar or side‐polar myosin filaments in vertebrate smooth muscle (G) and the large paramyosin‐containing filaments in molluskan smooth muscles (H).

42

SQUIRE ET AL.

register at least locally (Freundlich and Squire, 1983; Schmitz et al., 1994), which, like bony fish muscle, gives the sarcomere a simple A‐band lattice with a high degree of three‐dimensional (3D) order. Note that in these insect muscles the Z‐band is also hexagonal, unlike vertebrate Z‐bands, which are tetragonal (see Fig. 3G). Like IFM, insect leg muscle myosin filaments also appear to have fourfold symmetry, but they often have many more actin filaments around them than given by the 3:1 ratio in the flight muscles. The filament ratio can be as high as 5:1 or 6:1 (Fig. 10D and E), and obviously the lattice spacing is larger than in IFM. This is also true of other arthropod leg and trunk muscles and some synchronous flight muscles. Many of these muscles also have myosin filaments with fourfold rotational symmetry, but the axial repeat of the crossbridge array, still with a 143 to 145 A˚ intercrown spacing, can vary. For example, four‐stranded myosin filaments occur in tarantula leg muscle (axial repeat 435 A˚ ; Crowther et al., 1985; Padron et al., 1998) and in Limulus (horseshoe crab) telson muscle myosin filaments (axial repeat 435 A˚ ; Stewart et al., 1981, 1985). The scallop striated adductor muscle has even larger myosin filaments with sevenfold rotational symmetry (axial repeat 1440 A˚ ; Craig et al., 1991; Vibert, 1992; Fig. 10F). All of these myosin filaments in invertebrate muscles contain the protein paramyosin, which is rather like the rod part of myosin without the heads (Cohen et al., 1987). In fact, in some invertebrate muscles, particularly in mollusks, paramyosin is the major protein and forms a large central core on the surface of which is a layer of myosin molecules (Squire, 1971; Fig. 10H). In a given muscle the size of these paramyosin filaments can be variable, but they can reach more than 1000 A˚ in diameter. The actin filament arrays in this case are essentially rings around the thick filament circumference. The major factor here is that the interactin center‐to‐ center spacing is often
120–140 A˚ , presumably a spacing that more or less matches the lateral spacing of myosin crossbridges on the thick filament surface with which they interact (see Section II.D,E). The actin filaments are also placed
150–200 A˚ from the thick filament surface, presumably to position them within easy reach of projecting myosin heads. A similar spacing occurs between actin filaments in vertebrate smooth muscles (Fig. 10G).

E. The Z‐Band The vertebrate striated muscle Z‐band is a cross‐linking structure that links actin filaments of opposite polarity in successive sarcomeres along a myofibril. One of the curious things about it is that, unlike the A‐band,

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

43

which is based on a hexagonal array of filaments, the Z‐band lattice is approximately square in cross‐section (see Fig. 3G). This means that through the I‐band the actin filaments must gradually change from a hexagonal arrangement in the A‐band to a square arrangement in the Z‐band. The strands of titin must also change from myosin‐linked and hexagonally arranged in the A‐band to anchoring in some systematic way into the tetragonal Z‐band. Before seeing how this might work, the main features and components of the Z‐band itself are described. The main cross‐linking protein between anti‐parallel actin filaments is a‐actinin. a‐Actinin is one of the spectrin families of proteins (Broderick and Winder, 2002). Each chain contains an actin‐binding globular end domain (two calponin homology domains) that continues into a rod‐shaped region where the chain is folded into a string of four 3‐chain a‐helical coiled‐coil domains, each one known as a spectrin‐repeat (Fig. 11A). The end of the chain then forms EF hands in a globular region which in some a‐actinins is Ca2þ binding. Two such chains interact in an anti‐parallel fashion to give the full a‐actinin molecule. In other words, in muscle, the a‐actinin molecule is an anti‐parallel dimer with actin‐binding domains at each end. The structure of the a‐actinin rod has been defined by protein crystallography (Djinovic‐Carugo et al., 1999; Ylanne et al., 2001). The actin‐binding region in a‐actinin has also been determined (Franzot et al., 2005), and is very similar to equivalent domains in dystrophin and utrophin (Keep et al., 1999). In the Z‐band, the a‐actinin molecules form cross‐links between anti‐parallel actin filaments. In muscle cross‐sections (Fig. 11B, C) the square Z‐band structure appears in two forms, the basketweave and the small square lattice. These appearances seem to depend on the shape of the cross‐links between adjacent anti‐parallel actin filaments (called up [U] and down [D] in Figs. 11–14). Some authors believe that the appearance depends on the physiological state of the muscle (e.g., Goldstein et al., 1988), although this is not certain. A possible mechanism relating the two may be that the small square lattice is simply a basketweave in which the a‐actinin cross‐links have become bent (Yamaguchi et al., 1985). Different muscles and fiber types have Z‐bands of different thicknesses (Fig. 12A–H). This seems to depend on the exact number of levels of a‐ actinin cross‐links that are present. In longitudinal section the Z‐band has a zigzag appearance, and different Z‐band types have different numbers of zigzags. However, detailed analysis of Z‐band structure (Fig. 12I) shows that the zigzags usually correspond to the overlapping of two levels of a‐actinin links (Luther et al., 2003). Other proteins in the vertebrate muscle Z‐band are parts of nebulin and titin. When the sequence of the Z‐band part of titin was analyzed, it was

44

SQUIRE ET AL.

Fig. 11. The structure of a‐actinin and the two vertebrate Z‐band lattices. (A) The ubiquitous protein a‐actinin is an anti‐parallel homodimer. Each
100 KDa monomer comprises four central spectrin repeats (S1 to S4) an EF‐hand domain and two calponin homology domains (CH) at the N‐terminus. The EF‐hand domains bind calcium in non‐muscle cells. One a‐actinin molecule binds two actin filaments via the calponin homology domains. a‐Actinin binds titin via EF‐hand domains. (B, C) The Z‐band is the site where actin filaments from adjacent sarcomeres overlap in a tetragonal lattice and are crosslinked by a‐actinin molecules. The polarity and origin of the actin filaments is indicated by U (up) and D (down). The appearance of the Z‐band in cross‐section is typically basketweave‐like (B) or small square‐like (C). The appearance is reported to transform between the two appearances depending on the state of the muscle.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

45

found that there are repeating sequences of about 45 residues that have become known as Z‐repeats. These can occur in different numbers in titins from different muscles and fibers due to differential splicing (Gautel et al., 1996). There does seem to be a good correlation between the number of titin Z‐repeats, the thickness of the Z‐band, and the number of a‐actinin levels (Luther et al., 2002). It was shown by Atkinson et al. (2000) that the axial extent of a Z‐repeat is likely to be
120 to 150 A˚ . However, by careful measurement of Z‐band spacing in electron micrographs, Luther and Squire (2002) were able to show that the axial spacing between zigzag levels is
190 A˚ , which is too large to be spanned by a single Z‐repeat. It was concluded (Fig. 13) that there must be two Z‐repeats between zigzag levels, that the titin Z‐repeat region must span only half of the Z‐band, and that two anti‐parallel titins coming into the Z‐band from adjacent sarcomeres must be needed to define the whole Z‐band array. The nebulin strands that run along the actin filaments are anchored in the Z‐band and appear to start at the Z‐band edges (Millevoi et al., 1998). Other proteins in the Z‐band are telethonin (Mues et al., 1998) and PDZ‐LIM proteins, several of which bind to a‐actinin (Zhou et al., 1999). The presence of these proteins in the Z‐band suggests that they not only help to stabilize Z‐band structure, but that the Z‐band itself has more than just the passive role of transmitting tension from one sarcomere to the next along the myofibril. Also in the Z‐band is an actin filament ‘‘capping protein,’’ otherwise known as CapZ or b‐actinin in skeletal muscle (Papa et al., 1999) and as Cap32/34 in Dictyostelium (Haus et al., 1991). The crystal structure of CapZ has been determined (Yamashita et al., 2003). In a model of binding of CapZ to actin, a single molecule of CapZ binds to the barbed end of actin where it interacts with two or three actin monomers (Wear and Cooper, 2004). The most detailed Z‐band structure to date has come from the extended Z‐crystals found in muscles of patients with nemaline myopathy (Morris et al., 1990). This showed the unit cell of the structure to have the symmetry 432121 in which the actin filaments themselves have 43 screw symmetry (i.e., they are on a left‐handed 4/1 helix). In an earlier section the helical symmetry of typical actin filaments was described as a 13/6 helix of repeat
357.5 A˚ . However, this is only one of a family of closely related symmetries. The actin filaments in insect flight muscle have 28/13 symmetry (28 actin subunits in 13 turns) and a repeat of 770 A˚ . The normal 13/6 helix could also be called a 26/12 helix (with a repeat of 715 A˚ ), which shows the similarity between the two structures; one is a slightly unwound version of the other. One of the features of the 28/13 helix is that 28 is divisible by 4, which means that, starting from a given actin monomer, there must be three other monomers with exactly 90 azimuthal

46

SQUIRE ET AL.

Fig. 12. Z‐band modular structures. The axial width of the Z‐band varies with fiber type: fast muscles typically have narrow Z‐bands of width 4070 nm, and slow and cardiac muscles have wide Z‐bands of width
100 nm. (A–D) Electron micrographs of

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

47

separations around the actin filament from the first monomer and separated axially by 770 A˚ /4 ¼ 192.5 A˚ . In fact, these four actin monomers on their own are related by 43 screw symmetry (they are on a left‐handed 4/1 helix). This, of course, is just what is needed to generate the Z‐crystal structure. The four actin subunits identified above will be the binding sites of four a‐actinin cross‐links pointing out at exactly 90 to the four neighboring anti‐parallel a-actin filaments. Since the actin molecules are themselves dimers, the Z‐crystal structure can also have twofold rotation axes perpendicular to the actin filament axis. Vertebrate muscle Z‐bands are clearly not extensive in the way that Z‐crystals are, but they do have interesting plane group symmetries. For a long time it was thought that the vertebrate Z‐band must have axial separations between the a actinin layers, which were directly related to the actin repeat measured from X‐ray diffraction patterns to be
360 A˚ . However, the careful measurements by Luther and Squire (2002) for comparison of the zigzag spacing with the size of the titin Z‐repeats showed that, as in the Z‐crystals, the axial separation of the layers of cross‐links in a variety of Z‐bands is
190 A˚ ; much closer to one‐quarter of the 28/13 actin filament repeat than to one‐half of the 360 A˚ repeat found elsewhere in the vertebrate sarcomere. The implication is that the perfect fourfold helical symmetry intrinsic to a filament with 28/13 symmetry fits in much better with a–actinin crosslinking in the Z‐band than does the 13/6 helix. It could even be that the a‐actinin and other Z‐band proteins pull the actin filament locally into the 28/13 geometry, whereas the rest of the actin filament, which does not make the same interactions, is closer to being that of a 13/6 helix. In summary, the actin and a‐actinin parts of the various vertebrate Z‐bands appear to be organized as subsets of the Z‐crystal structure. In each subset different numbers of layers from the Z‐crystal are present, the number presumably being determined by the titin Z‐repeats and possibly the Z‐band part of nebulin. Why the control of Z‐band thickness goes wrong when Z‐crystals are produced is not yet clear, but nemaline longitudinal sections of the Z‐band in (A) fish body white muscle (Luther, 1991); (B) fish fin muscle (Luther, 2000); (D) frog sartorius muscle (Luther et al., 2003); and (D) bovine neck (slow) muscle (Luther et al., 2002). The left and right panels show the primary orthogonal lattice views obtained by tilting the sections by 90 in the electron microscope. (E–H) The corresponding schematic views of the electron micrographs in A–D. Luther et al. (2003) proposed that these modular patterns are determined by the number of a‐actinin layers within the width of the Z‐band. (I) Stereo view of a model of a 6‐a‐actinin layer Z‐band as found in slow muscle (Luther et al., 2003). Pairs of a‐actinin layers occur close together in longitudinal projection and give rise to three zigzag layers observed in electron micrographs, as shown by the model image superimposed on a corresponding electron micrograph (J).

48

SQUIRE ET AL.

Fig. 13. The titin Z‐repeats and a Z‐band assembly model. The N‐termini of titin filaments from adjoining sarcomeres overlap in the Z‐band. This part of titin comprises a modular region of so‐called Z‐repeats, each about 45 residues long, the number of which is related to fiber type: 24 repeats occur in fast muscles, 57 occur in slow and cardiac muscles. This correlates with the Z‐band appearance since fast and slow fibers have narrow and wide Z‐bands, respectively, as shown in Fig. 12. The measured axial spacing between a‐actinin bridges is about 19.2 nm (Luther and Squire, 2002), a distance that is too long for a single Z‐repeat (A) to stretch to. Perhaps the bridge separation is related to two levels of Z‐repeats (C).

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

49

myopathy can be produced by mutations in actin, in troponin, in tropomyosin, and in nebulin (Wallgren‐Petterson and Laing, 2003). Since troponin is only found outside the body of the Z‐band, it could well be that it is the molecular organization at the Z‐band edges, where labeling of the actin filament by tropomyosin/troponin stops and nebulin insertion into the Z‐band edge occurs, that is important in controlling Z‐band thickening.

F.

Filament Organization in the Vertebrate I‐Band

As mentioned earlier, vertebrate muscle sarcomeres have a strange geometric transition through the I‐band from the hexagonal A‐band lattice to the tetragonal Z‐band structure. It was Pringle (1968) who demonstrated that this kind of transition could be achieved by equal displacements of the actin filaments if the Z‐band lattice is not exactly square but has an included angle of 83 . (Clearly an equivalent systematic change could occur if the Z‐band is in fact perfectly square, but the A‐band is not exactly hexagonal). The vertebrate A‐band has two actin filaments to one‐half myosin filament in a typical A‐band unit cell (see Fig. 10A and B). Since it has now been shown that there are probably six titin chains per half myosin filament (Liversage et al., 2001), the implication is that there are three titin strands per actin filament in the Z‐band. This number seems rather strange in view of the pseudo–two‐strandedness of the actin filaments and the fact that in the Z‐band they seem to make a‐actinin cross‐ links to four anti‐parallel actin filaments. However, the possible distribution of titin through the I‐band has been rationalized by Knupp et al. (2002) as illustrated in Fig. 14. Here the six titin strands from a single half myosin filament, which aggregate to form the end‐filaments at the A‐band edge, divide into two diametrically opposite pairs and two individual strands. The pairs pass to the Z‐band and interact with actin filaments of one polarity, whereas the single titin strands interact with actin of the opposite polarity in the Z‐band. This seems to be a plausible scheme and leads to an even distribution of titin interactions with every actin filament; all have two parallel titin strands and one anti‐parallel strand interacting with them in the Z‐band. It also would explain some of the lateral forces that are known to be present in the sarcomere. However, it is, of course, an oversimplification in that titin is known to make interactions with actin filaments in certain parts of the I‐band, possibly at the PEVK and N2 regions (Fig. 5D; and see Granzier and Labeit, 2005), and this will modify the simple scheme in Fig. 14.

50

SQUIRE ET AL.

Fig. 14. Stereo pairs of the transverse structure (A) and the axial structure (B) of a 3D model relating successive half sarcomeres in vertebrate‐striated muscles. In both images, the wide blue and brown cylinders represent actin filaments, the gray cross‐links

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

IV.

51

Myosin Filament Structure and the M‐Band

A. The X‐Ray Diffraction Approach to Myosin Filament Structure As described in an earlier section, the myosin filaments in vertebrate striated muscles are bipolar assemblies of myosin molecules with threefold rotational symmetry. In the case of the filaments in bony fish muscle, they give the kind of X‐ray diffraction pattern shown in Fig. 4B. There is an axial repeat along the filaments of 429 A˚ and a subunit axial translation (the axial separation of myosin head crowns) of 143 A˚ . The shape of the myosin head was determined first by Rayment et al. (1993b; Fig 4A), but since then numerous other myosin head crystal structures have been determined (see Geeves and Holmes, 2005). A major feature of these is that the catalytic or motor domains are usually rather similar, and the neck regions or lever arms are also rather similar, but there may be a variety of angles between the two (Fig. 15). It is evident that there is a hinge within the myosin head, actually around residue 780 (Fig. 16A). It is this hinge that led to the idea of the neck region acting as a lever arm responding to ATP‐induced conformational changes in the catalytic domain (Rayment et al., 1993a). One of the tasks of structural biologists studying muscle contraction is to determine the organization and shapes of the myosin head in muscle under different physiological conditions. The technique of low‐angle X‐ray diffraction has unique advantages in this process, particularly since it can be applied to living muscle, which can be stimulated to produce active force or can be studied under a variety of different steady‐state conditions. The main problem with X‐ray fiber diffraction, as detailed in Squire and schematically represent a‐actinin bridges between anti‐parallel actin filaments, the narrow green and purple cylinders represent titin strands in one half sarcomere, and the narrow dark blue and red cylinders represent titin strands in the other half sarcomere. The narrow green and red cylinders represent single titin strands, whereas the narrow purple and dark blue cylinders represent paired titin strands. The myosin filaments are shown as brown cylinders. At the bottom of the structure (more obvious in B), there are seven myosin filaments, one central and six surrounding this, and at the top of the structure, for clarity, there is just a single myosin filament. Since the actin/a‐ actinin Z‐band assembly appears to be a strong structure and is an early development in myogenesis, one can alternatively look on the 3D model as showing that there are balanced lateral forces keeping the myosin filament tips in the A‐band in the correct lateral positions relative to the actin filaments emanating from the Z‐band. At the same time the titin strands are held relatively clear of the actin filaments throughout most of the I‐band, thus allowing them to behave freely as parallel elastic elements. The opposite tilting of the titin and actin filament displacements in a given I‐band and the generally balanced 3D distribution of forces on the Z‐band actin filaments make this a very attractive model. (Modified from Knupp et al., 2002.)

52

SQUIRE ET AL.

Fig. 15. Stereo views of the different myosin head, S1, structures showing their variable conformations in different crystal structures. (A) The heads with their motor domains superimposed and oriented as if interacting with a vertical actin filaments in the rigor conformation, Z‐band bottom and M‐band top. (B) The same structures in a view down the actin filament long axis, looking from the M‐band towards the Z‐band. Blue is the Dominguez et al. (1998) structure of S1 in chicken smooth muscle with ADP.AlF4 bound, orange is the insect flight muscle S1 in the ADP.Pi state (Al‐Khayat et al., 2003), yellow is scallop S1 crystal structure in the ADP.VO4 state (Houdusse et al., 1999), and green is the chicken skeletal muscle with no nucleotide bound (Rayment et al., 1993a).

Knupp (2005; this volume), is one of interpretation. Unlike X‐ray crystallography, low‐angle fiber diffraction does not lead directly to images of the diffracting object or to the generation of electron density maps. Setting up models with whatever prior knowledge is available is required, and then the unknowns must be set up as parameters and searched for the best values of these parameters to provide a model that satisfactorily accounts for the observed diffraction patterns.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

53

Fig. 16. (A) Myosin head S1 showing the two domains referred to as the motor domain and the lever arm with a hinge at around residue 780 between the two domains that allows the relative movement between the two domains. (B) The arrangement of the myosin heads on the surface of the myosin filament backbone lying approximately on three co‐ axial helices of subunit translation 143 A˚ and repeat 429 A˚ . If the origins of each pair of myosin heads in each myosin molecule are labeled as black circles on a cylindrical piece of paper; when this is flattened out (C) a radial net is obtained (D) showing the three helices of myosin head origins in different colors, red, blue, and yellow.

54

SQUIRE ET AL.

Fig. 17. (A) Principle of the searching routine showing various parameters describing the arrangement of the myosin heads on the myosin filament backbone and also changes in shape of the head by rotating the motor domain relative to the lever arm. (B) Illustration of the simulated annealing process showing the variation of R‐factor in a hypothetical parameter space. Points B and F are local minima; G and C are local maxima. The best parameter value is at point D, which is the so‐called global minimum,

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

55

A typical low‐angle diffraction pattern from relaxed bony fish muscle is shown in Fig. 4B. Much of the intensity that is seen comes from the organization of the myosin heads on the myosin filaments in the resting state (probably mainly M.ADP.Pi). We know that the myosin heads lie approximately on three co‐axial helices of subunit translation 143 A˚ and repeat 429 A˚ . This is most easily represented by the radial net shown in Fig. 16B–D. The radial net in D is like an opened‐out surface view of the filament in B. Here the helical tracks become straight lines, and the black blobs represent the origins on the myosin filament surface of the pairs of myosin heads in each myosin molecule. From early studies it is known that the three crowns within the 429 A˚ repeat are not exactly the same and that there is a perturbation. Figure 17 illustrates the basis of the problem faced in modeling the observed X‐ray pattern as in Fig. 4B. We know the myosin head shape in that it appears to consist of two major domains that at low resolution appear to act as rigid bodies. We know the approximate arrangement of myosin head origins on the filament surface, taken initially to be perfectly helically organized, but we know that within a 429 A˚ repeat, successive crowns are not the same. It is therefore necessary to set up a model in which not only the relative orientations of adjacent heads can vary, but also their shape. The head organization must also be permitted to be different on the three crown levels within a 429 A˚ repeat. Figure 17A illustrates some of the parameters that are involved, such as the head tilt and the changes of shape produced by the catalytic domain slew, tilt, and rotation. In analysis of the bony fish muscle myosin filament structure, Hudson et al. (1997) set up a computer model of this structure with about 22 variable parameters; they stripped diffraction patterns such as that in Fig. 4B to give 56 independent X‐ray intensities and then used a simulated annealing search to define the ‘‘best’’ structure. For each set of parameters, they used the computer to generate model diffraction patterns to compare with the observed diffraction patterns. The term ‘‘best’’ above was defined as that giving the lowest R‐factor, where the R‐factor is an objective measure of agreement between the observed and calculated diffraction patterns; it is a kind of ‘‘goodness of fit’’ factor in which the smaller the value the better the fit. One possible way to conduct such a search is to systematically change every single parameter in appropriately small steps and to calculate the diffraction pattern in each case. However, with as many as 20 or so as opposed to local minima at B or F. The simulated annealing approach can be likened to ‘‘heating up’’ the system so that the parameter values can bounce around more freely in parameter space. It permits the search to jump out of local minima (B, F) and to find the global minimum (D).

56

SQUIRE ET AL.

parameters and with many steps required for each parameter, the computing time involved soon reaches many tens of years. Another approach is needed. Practical approaches include refinement methods such as the Powell method and the downhill simplex method (Brent, 1973; Nelder and Mead, 1965). However, these methods have a problem illustrated in Fig. 17B. These refinement methods, which change the parameter values down local gradients in R‐factor space, will take one from say point G or point A down to B at the bottom of the well that they are next to, but the refinement then gets stuck. The best parameter value is actually somewhere else at point D, which is the so‐called global minimum, as opposed to local minima at B or F. Another approach, known as simulated annealing (Metropolis et al., 1958), which can be likened to ‘‘heating up’’ the system so that the parameter values can bounce around more freely in parameter space, permits the search to jump out of local minima and to find the global minimum. Subsequent application of one of the other local refinement methods (Powell or simplex above) then ensures that the point D is reached at the bottom of the global well.

B. Myosin Head Organization in Relaxed Vertebrate Myosin Filaments The results of modeling bony fish muscle myosin filaments using the simulated annealing approach (Hudson et al., 1997) are shown in Fig. 18. The top half of the observed X‐ray diffraction pattern, after data reduction and correction using CCP13 software (www.ccp13.ac.uk), is shown as Fig. 18A. The best model found after repeated simulated annealing searches (Fig. 18C) gave the computed diffraction pattern in Fig. 18B. In this structure the pairs of heads on one origin on the filament surface are relatively close to each other, but the three crowns are clearly different; the perturbation is evident. In fact, on two levels, the two heads in each pair lie one above the other, whereas on the third level the heads are side by side. On the face of it, this appears to be a very sensible structure, but there are two major problems in such analysis. The first is that vertebrate myosin filaments are known to carry extra proteins like C‐protein (MyBP‐C) and titin and the X‐ray model as described did not include these proteins. The second is that the diffraction pattern from the structure in Fig. 18C would be exactly the same if the structure was turned over top to bottom. In other words, there is no information in the X‐ray pattern about the polarity of the structure in Fig. 18C relative to the actin filaments with which the myosin interacts. Apart from testing the correctness of the X‐ray model, it is in these areas that electron microscopy comes into its own. Figure 19A shows a field of negatively stained isolated myosin filaments from bony fish muscle, together with actin filaments in the background.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

57

Fig. 18. (A) Observed top half of the low‐angle X‐ray diffraction pattern from relaxed fish (see Fig. 4B) showing the three myosin layer lines (ML) at orders of 429 A˚ repeat and the third‐order meridional reflection, M3. (B) Calculated diffraction pattern from the best model shown in (C). The equator (Eq) and meridian (M) are labeled. (C) The model for the arrangement of myosin heads on the myosin filament that gave the best fit to the observed pattern shown in (A). (Based on Hudson et al., 1997.)

58

SQUIRE ET AL.

Fig. 19. (A) Electron micrograph image of isolated myosin filaments and actin filaments from goldfish body muscle negatively stained and under relaxing conditions (scale bar ¼ 0.5 mm). (B) An isolated myosin filament from (A) showing two selected segments boxed in white and black rectangles and spaced at a multiple of the 429 A˚ repeat distance (arrows). The M‐band region is shown at the bottom. (C) A series of class‐averages from the particles selected and (D) reprojections of the final 3D reconstruction projected at the angles assigned to the corresponding classes shown in (C). (From Al‐Khayat et al., 2005a.)

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

59

One of the image‐processing techniques that has been applied to such myosin filaments is helical reconstruction (Eakins et al., 2002; Moore et al., 1970), where the computed Fourier transform (diffraction pattern) from a selected filament is processed to give a 3D reconstruction. As is clear from our need to use two eyes to get stereo vision, in order to generate a 3D reconstruction in electron microscopy it is necessary to obtain images of the same object viewed in different directions. Fortunately, with a helical structure (e.g., an actin filament or a perfectly helical myosin filament), a single filament image shows many subunits at different angles around the filament axis, so in this case a single image is sufficient to generate a 3D reconstruction. Such reconstructions of invertebrate myosin filaments are discussed later. The problem with vertebrate myosin filaments is simply that they are not exactly helical; there is the perturbation discussed above. For this reason, another approach is needed. In fact, the new technique of single particle analysis (e.g., van Heel et al., 2000) has been found to be very effective. Here the assumption is that there are many electron micrograph images available of exactly the same structure randomly oriented on the grid and therefore viewed in different directions. These images are then selected and a computer program is used to sort them into ‘‘classes’’ of views that look the same. The members of each ‘‘class’’ are then averaged. The problem that remains is to find the relative viewing directions of the different class–averages. Computer programs such as IMAGIC (van Heel et al., 1996), SPIDER (Frank et al., 1996), or EMAN (Ludtke et al., 1999) make use of the common‐line projection theorem to estimate these angles. Once these have been determined, the particle structure can be put back together using a back‐projection procedure (Radermacher, 1992). This technique has been very successful for the analysis of globular particles, where remarkably high resolutions have been achieved:
10 A˚ in some cases and even higher with viruses that have a great deal of built‐in symmetry (van Heel et al., 2000). However, the method has a special problem when applied to filaments or other elongated particles (Paul et al., 2004). In these cases, the filaments normally lie down on the grid with their long axes parallel to the plane of the grid so the only rotations available are those around the filament long axis. Normally, tilts around more than a single axis are necessary for the computer algorithm that determines the particle viewing angles to work properly. However, with special precautions, evaluated by Paul et al. (2004) and Patwardhan et al. (2004), the approach can be applied successfully to elongated structures such as actin and myosin filaments. Figure 19B shows an isolated myosin filament from a field such as that in A and the rectangular boxes show ‘‘particles’’ that have been cut out from the filament images at 429 A˚ intervals to form a data set for single particle

60

SQUIRE ET AL.

Fig. 20. (A) Cross‐sectional views of the three crowns of the X‐ray model shown in Fig. 18C. (B) Cross‐sectional views of the three crowns of the EM 3D map shown in E. (C) Cross‐sectional views of the densities between the crowns of the EM 3D map shown in E. (D) Stereo view of the X‐ray model of Fig. 18C reconstructed to 50 A˚ resolution. (E) Stereo view of the 3D reconstruction of the myosin filament structure obtained by single particle analysis using the classes shown in Fig. 19C (from Al‐Khayat et al., 2005a). The

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

61

analysis (Al‐Khayat, 2004, 2005a). Class averages from such particles are shown in C. These were then subjected to angle‐assignment and a 3D reconstruction generated as in Fig. 20E. This 3D density map can also be reprojected in the same directions as the original classes as a kind of test of how well the procedure is working. Figure 19D shows the reprojected classes for comparison with C; the correlation is good. Figure 20 also compares the results from this single particle analysis with the results from the X‐ray analysis of Hudson et al. (1997). Parts A and B show the projected densities in the three different crowns within a 429 A˚ repeat, A from the X‐ray model in Fig. 18B and C from the electron microscopic reconstruction in Fig. 20E. In both cases there is a striking feature. If the myosin filament was properly helical, there would be rotations of exactly 40 between the crowns and, apart from this rotation, the structures of each crown would be the same. However, in Fig. 20A and B it is clear that level 1 in both cases has a triangular profile pointing to the right. Level 2, on the other hand, points to the left and level 3 points to the right again. To a first approximation, levels 1 and 3 point in the same direction. Rather than having three equal angular steps of 40 , there are two steps of 60 and one of 0 . This is the azimuthal aspect of the crossbridge perturbation. The other feature, most obvious from the X‐ray structure (Fig. 20D) and the electron microscopic reconstruction in axial view (Fig. 20E), is that the crowns are not evenly spaced axially. This is the axial part of the head perturbation. The only other possible perturbation would be radial, but there seems to be very little in the way of a radial perturbation; all the crowns seem to have heads at about the same radius. Note finally that there is a difference between the X‐ray model, where only myosin is included, and the electron microscopic reconstruction, where everything is included. In part E the arrows highlight density features not seen in D that appear to lie between the levels of projecting heads. These positions are also shown in cross‐section in C. Presumably these features are associated with proteins such as titin and C‐protein that label the myosin filaments. The locations and properties of these proteins are briefly described in the next section.

C. Further A‐Band Analysis: C‐Protein, Titin, and the Vertebrate M‐Band Titin is a very long string of immunoglobulin‐like (Ig) and fibronectin‐ like (Fn3) domains with some intervening unique sequences such as the three crown levels are labeled and extra density that could be attributed to non‐myosin density (e.g., titin or C‐protein) are shown by arrows. The M‐band is at the bottom in (D and E). In A to E, levels 1, 2, and 3 refer to the crown level.

62

SQUIRE ET AL.

Fig. 21. (A) The A‐band part of titin consists of a repeating pattern of 11 Ig and Fn3 domains grouped in 11 copies from C‐terminus to N‐terminus as (Fn‐Fn‐Fn‐Ig‐Fn‐Fn‐ Fn‐Ig‐Fn‐Fn‐Ig) starting at the edge of the bare zone. It then continues as six copies of the repeat (Fn‐Fn‐Fn‐Ig‐Fn‐Fn‐Ig) plus a few extra domains out to the myosin filament tip. (B) C‐protein (MyBP‐C) in cardiac and skeletal muscle; common features are 10 domains, which are Fn3‐like, and Ig‐like, numbered C1 to C10. The C‐terminal domains C8, C9, C10 are associated with myosin and titin binding. Between C1 and C2 is a sequence that binds to myosin S2. Cardiac C‐protein has an additional Ig domain (C0)

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

63

PEVK region (see Fig. 5D) in the I‐band part of titin and a titin kinase region close to the M‐band (see Granzier and Labeit, 2005). In the A‐band (Fig. 21A), titin occurs in close association with the myosin filaments and through much of its A‐band part it has a repeating pattern of 11 Ig and Fn3 domains grouped in 11 copies from C‐terminus to N‐terminus as (‐Fn‐Fn‐ Fn‐Ig‐Fn‐Fn‐Fn‐Ig‐Fn‐Fn‐Ig‐) starting at the edge of the bare zone. It then continues as six copies of the repeat (Fn‐Fn‐Fn‐Ig‐Fn‐Fn‐Ig‐) to the myosin filament tip. There are six titin molecules per myosin filament in half of the sarcomere (Liversage et al., 2001). These six molecules converge to form the so‐called end‐filaments at the A‐band edge (Trinick, 1981), after which the titin strands cross the I‐band through to the Z‐band, as described earlier (see Fig. 14). The I‐band part of titin is essentially composed of Ig domains apart from the major insertion rich in the amino acids P, E, V and K (proline, glutamic acid, valine, and lysine). As described in Granzier and Labeit (2005), this part of titin has remarkable elastic properties. Because all the subdomains are of about the same size (40 A˚ ), the 11‐subdomain repeat in the A‐band part of titin is
440 A˚ long. In fact, this is thought to coincide with the 429 A˚ repeat of the myosin head array, giving a titin subdomain repeat of 429/11 ¼ 39 A˚ . Such a repeat, the eleventh order of 429 A˚ , has in fact been observed in freeze‐fractures of muscle A‐bands (Cantino et al., 2002) and in single particle analysis of myosin filaments (A‐Khayat et al., 2005a), and the eleventh‐order meridional peak in muscle X‐ray diffraction patterns is also relatively strong (Oshima et al., 2003). Note, however, that some aggregates of parts of the myosin rod also give a strong eleventh‐order peak in their computed Fourier transforms (diffraction patterns) even in the absence of titin (Bennett, 1981). C‐protein (MyBP‐C) has a substructure that is in many ways similar to parts of titin. Some of its domains are Fn3‐like and Ig‐like domains, but there are also unique sequences. The cardiac and skeletal forms of C‐protein are also different (Fig. 21B). The skeletal form has 10 domains in which the C‐terminal domains C8, C9, and C10 are associated with myosin and titin binding. C10 is the main myosin binding site, but C8 and C9 are needed to give C‐protein its proper location on the myosin filament. Between C1 and C2 is a sequence that can bind to myosin S2

at its N‐terminus. Phosphorylation sites are indicated by a rectangle in the S2 binding site and in C5. A pro‐Ala–rich domain is present at the N‐terminus in both isoforms; it links C0 and C1 in the cardiac molecule. Also shown are the related structures of Hprotein, myomesin, and M-protein. (C) Interactions can occur between domains 5 and 8 and between domains 7 and 10 of different C‐protein molecules (from Moolman‐ Smook et al., 2002). These interactions may suggest that three C‐protein molecules form a collar around the myosin filament.

64

SQUIRE ET AL.

(Kunst et al., 2000). At the N‐terminal end of skeletal C‐protein is a sequence rich in proline and alanine, and therefore known as the Pro‐ Ala domain. Cardiac C‐protein has an additional Ig domain (C0) at its N‐ terminus and has some sequence insertions both in the S2‐binding region, which can also be phosphorylated, and in C5. It has been found that interactions can occur between domains 5 and 8 and between domains 7 and 10 (Fig. 21C) of different C‐protein molecules (Moolman‐Smook et al., 2002), and these interactions were used to suggest that three C‐protein molecules form a collar around the myosin filament. On the basis of some puzzling X‐ray diffraction and electron microscopy evidence, Squire et al. (2003) have now suggested that while the C‐terminal end of C‐protein binds to myosin, the N‐terminal ends may bind to actin filaments, at least in resting muscle. The existence of an actin interaction would explain the observation of a C‐protein peak at
442 A˚ rather than at the 429 A˚ repeat of the myosin filament (Haselgrove, 1975; Squire et al., 1982). Electron microscopy has also suggested the presence of two repeats in the C‐zones of vertebrate muscles (Sjostrom and Squire, 1977). Figure 22 shows how C‐protein molecules based on the myosin filament repeat, with three molecules at each 429 A˚ –spaced level, would extend out to actin and bind actin‐binding sites which would give the actin end a rather different repeat from 429 A˚ . Since there is also good biochemical and physiological evidence for actin binding (Kulikovskaya et al., 2003; Moos et al., 1978) and now further X‐ray diffraction evidence (Squire et al., 2004), it does seem fairly certain that C‐protein can bind to actin as well as myosin. Note that this region of C‐protein also binds to myosin (Kunst et al., 2000). For reasons outlined in Squire and Knupp (2005), it also seems possible that the C‐terminal ends of C‐protein may run axially along the titin strands with which they interact, rather than forming a collar as suggested by Moolman‐Smook et al. (2002). By comparison of the sequence of the Pro‐Ala domain of C‐protein and the sequences of similar regions in one of the light chains of myosin, shown by Trayer and colleagues to bind to actin (Timson et al., 1998), it was suggested by Squire et al. (2003) that this region of C‐protein might be part of the actin binding site. This idea remains to be tested. In the central region of vertebrate myosin filaments is the bare zone (Fig. 3D). It is bare in the sense that there are no myosin heads there, but it is not actually bare. In muscle there is a strong cross‐linking structure, the M‐band, in the middle of this region and titin strands continue from the bridge region into the M‐band. Because it is not bare in the sarcomere, but contains the M‐band proteins and other structures, this region is sometimes called the M‐region (Sjostrom and Squire, 1977). In electron micrographs of muscle cross‐sections, the M‐band appears as dense myosin

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

65

Fig. 22. (A) Radial net of an array of six actin filaments (light blue) surrounding a myosin filament as seen from the axis of the myosin filament backbone looking radially outward. Open circles and lines represent the radial net of the MyBP‐C protein origin positions on the thick filament surfaces. This consists of seven levels of C‐protein molecules axially separated by 429 A˚ and with 3 C‐proteins 120 apart at each level. The orange circles show which actin monomers might be labeled by the N‐termini of the C‐protein molecules whose origins are represented by the open black circles. (B) A possible model for the C‐protein geometry within the sarcomeric A‐band of resting vertebrate striated muscle. A thick filament backbone (brown) is surrounded by six actin filaments (dark gray). C‐protein molecules with domains C7 to C10 running axially along the myosin filament backbone interact with actin filaments through the C‐protein N‐terminal Pro‐Ala–rich domains (green ovals). A few myosin heads are drawn as ‘‘transparent’’ ghosts to indicate their proximity to the C‐protein array. (After Squire et al., 2003.)

filament profiles arranged on a hexagonal lattice and cross‐linked to their six myosin filament neighbors by bridging structures (see Fig. 3I), which sometimes show an increased density halfway along them. In axial views, the M‐band can vary between muscles and fiber types, but there appears to

66

SQUIRE ET AL.

be a symmetrical pattern of up to five lines of M‐bridges (Fig. 23A, B and D, and Figs. 24 and 25). These lines are separated axially by about 220 A˚ . Because of the presence of some weaker lines between them, these main M‐band lines were termed M6, M4, M1 (centrally located), M40 , and M60 (Sjostrom and Squire, 1977). Of the weaker lines in different M‐bands, those that were particularly consistent and strong were the lines at M3, which were thought by Luther and Squire (1978) to possibly represent weaker bridging structures in the M‐band. The M‐region transverse lines observed by Sjostrom and Squire (1977) extended out to M9 and M90 at the edges of the M‐region, with the next observed densities thought to be the first level (P1) of myosin heads in the proximal zone of the bridge region (Fig. 23B). Different fiber and muscle types have M‐bands with different relative densities of the strong M‐band lines (Carlsson et al., 1990; Sjostrom and Squire, 1977). In particular, relatively fast fibers have rather weak lines at M6 and M60 , giving a three‐line M‐band (Fig. 23D), whereas very slow fibers have relatively weak density at M1, giving a four‐line M‐band. Other muscles have line densities between these extremes, sometimes with all lines relatively strong, giving a five‐line M‐band. In all cases the M4, M40 lines are strong, suggesting that the structures giving rise to M4 are the primary elements defining M‐band strength. One intriguing aspect of the M‐band is that it presumably is the structure that defines the relative rotations of adjacent myosin filaments in the A‐band hexagonal lattice. In other words, it determines whether the A‐band has a superlattice or a simple lattice. In fact, the generation of the two lattice types can be rationalized in terms of the molecular interactions that might occur at the M4 level (Pask et al., 1994; based on Luther and Squire, 1980). At this position the myosin filament cross‐sectional profiles appear triangular (Fig. 23E and F). Since they effectively make six cross‐linking interactions with their six myosin filament neighbors, the binding sites for the cross‐links must either come from a triangle tip or a triangle side. If the tip interactions are called ‘‘A’’ or ‘‘a’’ and the side interactions are called ‘‘B’’ or ‘‘b,’’ then the lattice type is, in fact, defined by the interactions between these sites on adjacent filaments. If an ‘‘A’’ always interacts with a ‘‘B’’ (Fig. 23E), then all the triangles must have the same orientation and a simple lattice is automatically generated. On the other hand, if an ‘‘a’’ site prefers to bind to another ‘‘a’’ site, or a ‘‘b’’ to a ‘‘b,’’ then such interactions cannot be systematically generated across the myofibril, but optimization of the number of such interactions automatically generates the kind of statistical superlattice defined by Luther and Squire (1980; Fig. 6B). Figure 24A shows three cross‐sectional slices through the fish simple lattice M‐band reconstruction of Luther and Crowther (1984). Here it

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

67

Fig. 23. (A) 3D Reconstruction of fish muscle M‐band/bare zone obtained from single particle analysis (Al‐Khayat et al., unpublished data). (B) 1D Density profile of (A) showing the main stripes, M60 , M40 , M1, M4, and also M6, M8 (titin kinase?), M9, and P1 (the first myosin crown). (C) The anti‐parallel packing of myosin rods in the bare zone, together with the main M‐bridge levels. M1 and M6 are of variable density giving rise to different M‐band types (D). M‐band interactions at the M4 level can explain the generation of simple and superlattice A‐bands (E and F). (After Pask et al., 1994.)

68

SQUIRE ET AL.

Fig. 24. M‐band structure from electron microscopy of both simple lattice and superlattice muscles. (A) 3D Reconstruction of fish muscle M‐band. Three distinct layers were observed in the reconstruction, at each of the M‐bridge levels: M40 , M1, and M4 (M and B label the myosin filaments and M‐bridges, respectively.) The observed 32‐point group symmetry has been imposed on the 3D map. (B) Part of the M‐band as modeled by Luther and Squire (1978). M1 and M4 bridges are seen connecting adjacent myosin filaments. Halfway along the M‐bridges and running parallel to the myosin filaments are the M‐filaments. M3 marks a further level of secondary Y‐shaped bridges. (C) A slice

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

69

can be seen that the M1 profile is rather sixfold, whereas the M4 and M40 profiles are rather triangular with the bridge density on the points of the triangles rather stronger than on the triangle sides. The whole M‐band and bipolar myosin filament has the symmetry of the 32‐point group (Luther et al., 1981). However, M4 and M40 do show a slight thickening of the bridges halfway across, which has been interpreted in terms of the presence of axially aligned density, the so‐called M‐filaments, running through the M‐band halfway between adjacent myosin filaments. Fig. 24B summarizes the general features of one half of the M‐band, excluding M6 (Luther and Squire, 1978). The myosin filaments, M‐filaments, M1 bridges, and M4 bridges form a 3D network, with densities at M3 possibly acting as secondary links between the M‐filaments. The existence of axially aligned density between the myosin filaments is supported by a new reconstruction of the M‐band in frog muscle (Fig. 24C). An oblique view of a 3D reconstruction of the superlattice M‐band in frog sartorius muscle is shown in Fig. 24D, where both lateral and axial densities can be seen (Luther et al., 2005). Apart from the M‐band part of titin, the bridges in the M‐band appear to be composed of at least two proteins, myomesin and M‐protein. These two rather similar proteins have a domain structure shown in Fig. 21B, which is clearly assembled from Ig and FN3 domains as in titin and C‐protein. There are also unique sequences at the N‐terminal ends. Antibody‐labeling studies have suggested the molecular arrangement illustrated in Fig. 25F (Obermann et al., 1996, 1997). Here titin forms part of the axially aligned density at the M‐filament position, and associated with this is part of myomesin. The N‐terminal part of myomesin then forms the M4 bridges. M‐protein, on the other hand, appears to be mainly confined to the M1 position. Other proteins in the M‐band include the muscle form of creatine kinase (MM‐CK; Hornemann et al., 2003). Since the M‐region part of titin also contains a kinase domain (Mayans et al., 1998) it is clear that, like the Z‐band, the M‐band does not just have a structural role but also has important metabolic properties. That the M‐band structure is clearly related to function is demonstrated by the appearance of the M‐band in the heart muscles of different animals with different heart rates (Pask et al., 1994). Figure 25A shows micrographs and their density profiles from a variety of cardiac muscles, going from through a 3D reconstruction of frog sartorius muscle. The M‐band features are noisier than in the fish case because of the statistical superlattice structure (see text). Myosin filaments (M) are observed with M‐bridges (B) and longitudinal running M‐filaments. (D) Stereo view of the 3D reconstruction of the frog muscle (superlattice) M‐band, showing M‐bridges and M‐filaments (Luther et al., 2005).

70

SQUIRE ET AL.

Fig. 25. Variety of cardiac M‐band structures and their molecular origin. (A–E) Electron micrographs of M‐region segments from (A) cow, (B) guinea pig, (C) rabbit, (D) plaice, and (E) carp. The profile plots on the right show the main stripes, M1, M4, and M6. M1 is a strong stripe in A through C but missing from E through E the fish hearts. (F) Molecular arrangement of the major M‐band proteins, myomesin (left) and

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

71

one with a clearly prominent M1 line (A) to one with M1 almost absent (E). Presumably the M‐band in E has little M‐protein, A has much M‐protein, and presumably M4, which remains strong in all M‐bands, is the site of myomesin which is always present. MM‐CK probably also occurs in variable amounts on and around M1 to M4, but its location is relatively ill defined.

D. Invertebrate Myosin Filaments Myosin filaments in muscles other than vertebrates appear to be relatively simple, in that they all appear to have their myosin heads arranged in regular helices, apart from at the edges of the bare zones where perturbations in the regular molecular packing are to be expected. Because of their helicity, analysis of electron micrographs of such filaments by helical reconstruction has been relatively successful. 3D reconstructions of the filaments in insect flight muscle (Morris et al., 1991), tarantula leg muscle (Crowther et al., 1985), limulus telson muscle (Stewart et al., 1981, 1985), and scallop striated muscles (Craig et al., 1991; Vibert, 1992) have been published. In the case of insect flight muscle, the myosin filament structure has also been determined from X‐ray diffraction modeling (Al‐Khayat et al., 2003). The muscle is so well ordered that, as in the case of fish muscle, the X‐ray diffraction patterns are beautifully sampled (Fig. 26A). The simulated annealing approach of Hudson et al. (1997) was adopted for this study, except that, since the structure was assumed to be perfectly helical, many fewer unknown parameters were needed than with vertebrate myosin filaments. Figure 26B and C compare the observed and calculated diffraction patterns for the insect case, and Fig. 26D shows the resulting model of the insect (Lethocerus) four‐stranded myosin filaments. In this case, the subunit axial translation (i.e., between crowns) is much like that in vertebrates at just below 145 A˚ . However, the repeat of the structure is after eight crowns at 1160 A˚ (see Fig. 27C), a number that fits in nicely with the 386.7 A˚ crossover repeat of the 28/13 actin filaments in this muscle (3  386.7 ¼ 1160) and helps to explain why this muscle is so well ordered in three dimensions. Unlike the vertebrate structure, here the heads are very closely confined axially to 145 A˚ –spaced ‘‘shelves’’ of density. However, half of the heads are projecting from the filament surface and are apparently supported by the other heads, which are wrapped around the filament circumference. A circumferential head from one myosin M‐protein (right), elucidated using immunolabeling (Obermann et al., 1997). Fast muscles have both proteins, but slow muscles have only myomesin. This correlates with the observed M‐band stripes where M1 is missing in slow muscles.

72

SQUIRE ET AL.

Fig. 26. (A) Observed X‐ray diffraction pattern recorded at the APS BioCAT Argonne Synchrotron Radiation Source for relaxed insect flight muscle (Lethocerus) showing a beautifully ‘‘crystalline’’ sampled pattern with the 145 A˚ meridional reflection labeled. (B) The bottom right quadrant of the observed pattern (OBS) shown in (A) with the myosin layer lines at orders of 1160 A˚ repeat. (C) The calculated diffraction pattern (CALC) for the best model that gave the best fit to the observed pattern (B). (D) 3D Model for the best X‐ray model for the myosin filament in relaxed insect flight muscle shown with its M‐band at the bottom and with the filament backbone shown simply as a cylinder (Al‐Khayat et al., 2003). (E and F) 3D Reconstruction of myosin filaments in relaxed scallop muscle obtained from single particle analysis (Al‐Khayat et al., 2005b) and viewed from different directions, M‐band bottom in E and F and viewed from the M‐band toward the Z‐band in G.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

73

molecule supports a projecting head from a neighboring molecule, thus providing the potential for cooperative interactions, possibly involved in regulation (Al‐Khayat et al., 2003; Squire et al., 2005). Other four‐stranded myosin filaments occur in tarantula, in the leg muscles of other insects, and, for example, in the Limulus telson muscle (Fig. 27B). In these cases, the intercrown spacing is still
145 A˚ , but, as in vertebrate filaments, the axial repeats of the filaments are after three

Fig. 27. Radial net of various crossbridge lattices from different species along with their corresponding computed Fourier transform. The myosin filament is three‐ stranded in (A) vertebrate muscle, four‐stranded in invertebrates (B and C), and seven‐stranded in scallop muscle (D). This figure shows the similarity of the surface lattices of the myosin head origins on the myosin filaments in different muscles although the myosin heads have different slew, tilt, and rotations. Images were created using the program HELIX (Knupp and Squire, 2004).

74

SQUIRE ET AL.

crowns at 435 A˚ . Reconstructions of these filaments show strong continuity of myosin head density along the long‐period strands of the helical surface net, once again showing interactions between the heads of different myosin molecules, but this time between crowns rather than within a crown as in insect flight muscle (e.g., Padron et al, 1998; Stewart et al, 1985). The myosin filaments in scallop striated adductor muscle are much larger in diameter than vertebrate myosin filaments and have seven co‐axial strands of myosin heads (Fig. 27D). As in the other cases of helical filaments, the scallop structure has also been determined by 3D helical reconstruction (Vibert, 1992). The reconstruction had a resolution of
70 A˚ , as for the other filaments reconstructed by this method. The single particle analysis technique described earlier for analysis of the nonhelical vertebrate myosin filaments can also be applied to purely helical structures and, even using the same micrographs, it has the potential to yield much higher resolution. This technique is being applied to electron micrographs of insect flight muscle thick filaments (Al‐Khayat et al., 2004) and also to scallop striated muscle filaments (Al‐Khayat et al., 2005b), in both cases achieving
50 A˚ resolution. The result in the latter case is shown in Fig. 26E–G. Here it can be seen that, although the head density appears to be aligned along the long‐period helical tracks (part E), it does appear to be such that both myosin heads would be accommodated within the main density peaks (F and G).

E. Crossbridge Arrangements on Different Myosin Filaments: Variations on a Theme One of the early suggestions about myosin filament structure made by Squire (1971, 1972) in his ‘‘General Model of Myosin Filament Structure’’ was that, since all muscle myosin filaments are made up from muscle myosin molecules that are rather similar in size, shape, and structure, perhaps myosin filaments, although different in different muscles, would have closely related packing arrangements. If this is considered in terms of the molecular packing of the myosin rods, then the implication is that, whatever the tilts, slews, and rotations of the heads, the underlying surface lattices of the myosin head origins on the different myosin filaments would be rather similar. As evidence has accumulated about different myosin filaments, this general theme appears to have been confirmed. The different filament lattices shown in Fig. 28 have different axial repeats, they have different numbers of strands of myosin heads, but if looked at locally, the relationships between adjacent heads are very similar. In each case, they are axially separated by
143–145 A˚ axially and the lateral separation is in the

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

75

Fig. 28. Classification of crossbridge configurations in myosin filaments in different muscles. In each case, the axial separation is
143–145 A˚ and the lateral separation is
120–150 A˚ . There are three main classes: (A) Class I, where the interaction is between heads of the same molecule as in vertebrate striated muscles; (B) Class II, where interaction occurs between heads of adjacent myosin molecules in the same crown, as seen in insect (Lethocerus) flight muscles; and (C) Class III, where the interaction appears to be between heads in different crowns, as seen in tarantula and Limulus.

region of 120 to 150 A˚ , a dimension that correlates well with the lateral separation of actin filaments in different muscles (see Fig. 10). The lateral separations would not be expected to be identical, because, if the flat layer of closely packed myosin rods, as envisaged by Squire (1973), was bent by varying amounts to produce filaments with different rotational symmetries, then the outer part of the layer, on which the heads are located, would open out by different amounts depending on the filament symmetry. Flat filaments in vertebrate smooth muscle (see Fig. 10G) or very large diameter filaments in molluskan smooth muscles (see Fig. 10H) would have relatively undistorted myosin layers and might have closer head separations than in other filaments where the layers are bent around to give small diameter filaments (see the filament backbone in Fig. 29A). It is in the muscles with flat or large myosin filaments that the actin filaments are relatively long and in excess over the myosin‐containing filaments and form pools of actin organized into actin lattices with an interactin filament spacing of
120 A˚ (Lowy et al., 1970; Small and Squire, 1972). In striated muscles the actin filament separations are much larger than this and the myosin heads radiate out toward them from tightly curved layers of rods. It is clear that

76

SQUIRE ET AL.

Fig. 29. (A) View down the filament long axis of one full myosin 429 A˚ repeat of the X‐ray model of the fish muscle myosin filament shown in Fig. 18C along with the backbone model from Squire (1986) and Chew and Squire (1995). (B) View down the fiber axis of one full myosin 429 A˚ repeat in the fish muscle A‐band unit cell, in which two actin filaments have been included. The head perturbations create a local environment of actin‐binding sites (blue) on the myosin heads (yellow) around the actin filaments (green).

myosin filaments as a family do indeed have a great deal in common, as envisaged by Squire (1971) and reviewed in Squire (1986). In the discussions above, it was noted that in some cases the interaction between heads in resting muscle myosin filaments is between heads of the same myosin molecule (vertebrate), whereas in other cases (insect flight), it is between adjacent molecules in the same crown. In a third type

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

77

(tarantula and Limulus), the interaction appears to be between heads in different crowns. We have just seen that the underlying surface lattices of myosin head origins are rather similar (see Fig. 28). It is convenient in discussing the different head organizations on these relatively uniform underlying lattice geometries to call them Class I, Class II, and Class III, respectively, as in Fig. 28. The different classes may have a distinct role in defining the regulation or other functions of the heads in these muscles, and their classification should help to define and understand their different functional roles in different muscle types. Note that according to early published models (Vibert, 1992), scallop muscle myosin filaments would have been classified as Class III, but the reconstruction in Fig. 26E–G may suggest that these are actually Class I filaments.

F.

Conclusion: Implications about the Crossbridge Mechanism

Several of the articles that follow this one, particularly those by Geeves and Holmes (2005) and Squire and Knupp (2005), are concerned with the

Fig. 30. Comparison of the X‐ray–modeled myosin head arrays in relaxed fish muscle (A) and relaxed insect flight muscle (B), with the motor (catalytic) domain of outer myosin heads in each model circled to show the close similarity of their configurations in the two different species. The M‐band is at the bottom in both models.

78

SQUIRE ET AL.

myosin crossbridge mechanism in active muscle. The present article has set the scene in which this interesting molecular mechanism takes place. Many of the studies of the contractile mechanism have involved studies of vertebrate striated and insect flight muscles, so this article summarizes what is known from these muscles. Figure 29A shows that the filament backbone has the myosin rods in a tightly packed and physically plausible arrangement that agrees with the high‐angle X‐ray diffraction patterns from muscle (Squire, 1973, 1986) better than any other published model (Chew and Squire, 1995). On the surface of this rod array, the myosin head arrangement is shown as determined by Hudson et al. (1997) for the A‐band unit in resting fish muscle. The single particle analysis work of Al‐Khayat et al. (2004, 2005a) has confirmed many of the features of this model (see Fig. 20). When the structure in Fig. 29A is put into the fish muscle A‐band unit cell (Fig. 29B; note that the analysis of Hudson et al. [1997] defined the absolute orientation of the filament within the A‐band lattice), it can be seen that the actin‐binding sites on the myosi n heads are already close to

Fig. 31. Implications about the contractile mechanism in insect flight muscle. Blue is insect flight muscle S1 shape in pre‐powerstroke state (Al‐Khayat et al., 2003), and green is chicken skeletal muscle S1 in the rigor state with no nucleotide bound (Rayment et al., 1993a). The actin filament (right) is shown with the Z‐band at the bottom and M‐band at the top. A transition from the pre‐powerstroke/resting S1 shape to the rigor/end of post‐powerstroke shape would involve an axial swing of the lever arm by 100 A˚ , resulting in the sliding of the actin filaments past the myosin filaments and toward the M‐band.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

79

actin even in resting muscle. If this structure is now looked at in axial view (Fig. 30A), it can be seen that the catalytic domains of the outer heads have almost the same orientation as the equivalent domains in the insect flight muscle myosin filament structure (Fig. 30B; shown earlier in Fig. 26D). In fact, the head shapes in these two structures are different, but both assemblies contrive to get the catalytic domains of the outer heads in similar orientations relative to actin. The myosin head shape in the insect muscle is very close to the pre‐powerstroke shape thought to be the shape in which the heads attach initially to actin (Dominguez et al. 1998; Houdusse et al., 2000). According to Al‐Khayat et al. (2003), the catalytic domain in resting insect flight muscle is almost in the correct orientation for easy attachment to actin to occur. A small rotation would put the head in the situation shown in Fig. 31. Binding to actin with this head shape, loss of ADP and Pi and conversion to the rigor head shape would be associated with a 100 A˚ swing of the outer end of the lever arm and a working stroke would have been completed (Fig. 31). The articles by Geeves and Holmes and Squire and Knupp (this volume) illuminate these basic ideas about the contractile mechanism.

Acknowledgments Much of the work reported here has been supported by grants from the UK Medical Research Council, the Biology and Biotechnology Research Council, the Wellcome Trust, the Leverhulme Trust, and the British Heart Foundation. We are indebted to these organizations for their support. We also thank our many colleagues who have contributed to much of the work reported here, in particular Dr. Michael Sjostrom, Dr. Edward Morris, Dr. Liam Hudson, Dr. Richard Denny, the late Dr. Helen Pask, Dr. Jeffrey Harford, Dr. Michael Reedy, and Dr. Tom Irving. None of the synchrotron X‐ray work could have been carried out without the help of the beamline staff on lines 2.1. and 16.1 at the Daresbury Laboratory, UK; on line ID–02 at the ESRF in Grenoble, France; and on the BioCAT beamline at the Advanced Photon Source at the Argonne National Laboratory, Argonne, Illinois, USA.

References Al‐Khayat, H. A., Yagi, N., and Squire, J. M. (1995). Structural changes in actin‐tropomyosin during muscle regulation: Computer modelling of low‐angle X‐ray diffraction data. J. Mol. Biol. 252, 611–632. Al‐Khayat, H. A., Hudson, L., Reedy, M. K., Irving, T. C., and Squire, J. M. (2003). Myosin head configuration in relaxed insect flight muscle: X‐ray modeled resting cross‐ bridges in a pre‐powerstroke state are poised for actin binding. Biophys. J. 85, 1063–1079. Al‐Khayat, H. A., Morris, E. P., and Squire, J. M. (2004). Single particle analysis: A new approach to solving the 3D structure of nyosin filaments. J. Mus. Res. Cell Motil. 25, 635–644.

80

SQUIRE ET AL.

Al‐Khayat, H. A., Morris, E. P., Powell, A. S., Kensler, R. W., and Squire, J. M. (2005a). 3D structure of vertebrate (fish) muscle myosin filaments by single particle analysis. J. Mol. Biol. (in revision). Al‐Khayat, H. A., Morris, E. P., and Squire, J. M. (2005b). 3D structure of relaxed scallop striated muscle myosin filaments by electron microscopy and single particle analysis. (in prep.) April, E. W., Brandt, P. W., and Elliott, G. F. (1971). The myofilament lattice: Studies on isolated fibers. I. The constancy of the unit‐cell volume with variation in sarcomere length in a lattice in which the thin‐to‐thick myofilament ratio is 6:1. J. Cell Biol. 51, 72–82. Atkinson, R. A., Joseph, C., Dal Piaz, F., Birolo, L., Stier, G., Pucci, P., and Pastore, A. (2000). Binding of alpha‐actinin to titin: Implications for Z‐disk assembly. Biochemistry 39, 5255–5264. Bennett, P. M. (1981). The structure of spindle‐shaped paracrystals of light meromyosin. J. Mol. Biol. 146, 201–221. Bennett, P., Craig, R., Starr, R., and Offer, G. (1986). The ultrastructural location of C‐protein, X‐protein and H‐protein in rabbit muscle. J. Muscle Res. Cell Motil. 7, 550–567. Brent, R. P. (1973). ‘‘Algorithms for Minimisation Without Derivatives.’’ Prentice‐Hall, Englewood Cliffs, NJ. Broderick, M. J., and Winder, S. J. (2002). Towards a complete atomic structure of spectrin family proteins. J. Struct. Biol. 137, 184–193. Brown, J. H., and Cohen, C. (2005). Regulation of muscle contraction by tropomyosin and trophonin: How structure illuminates function. Adv. Protein Chem. 71, 121–159. Cantino, M. E., Chew, M. W., Luther, P. K., Morris, E., and Squire, J. M. (2002). Structure and nucleotide‐dependent changes of thick filaments in relaxed and rigor plaice fin muscle. J. Struct. Biol. 137, 164–175. Carlsson, E., Grove, B. K., Wallimann, T., Eppenberger, H. M., and Thornell, L. E. (1990). Myofibrillar M‐band proteins in rat skeletal muscles during development. Histochemistry 95, 27–35. Chandrasekaran, R., and Stubbs, G. (2001). Fibre diffraction Kluwer, Dordrecht. Chew, M. W., and Squire, J. M. (1995). Packing of alpha‐helical coiled‐coil myosin rods in vertebrate muscle thick filaments. J. Struct. Biol. 115, 233–249. Chik, J. K., Lindberg, U., and Schutt, C. E. (1996). The structure of an open state of beta‐ actin at 2.65 A˚ resolution. J. Mol. Biol. 263, 607–623. Cohen, C., Lanar, D. E., and Parry, D. A. D. (1987). Amino acid sequence and structural repeats in schistosome paramyosin match those of myosin. Biosci. Rep. 7, 11–16. Craig, R., Padron, R., and Alamo, L. (1991). Direct determination of myosin filament symmetry in scallop striated adductor muscle by rapid freezing and freeze substitution. J. Mol. Biol. 220, 125–132. Crowther, R. A., Padron, R., and Craig, R. (1985). Arrangement of the heads of myosin in relaxed thick filaments from tarantula muscle. J. Mol. Biol. 184, 429–439. Dancker, P., Low, I., Hasselbach, W., and Wieland, T. (1975). Interaction of actin with phalloidin: Polymerization and stabilization of F‐actin. Biochim. Biophys. Acta 400, 407–414. Dancker, P., and Hess, L. (1990). Phalloidin reduces the release of inorganic phosphate during actin polymerization. Biochim. Biophys. Acta 1035, 197–200. Djinovic‐Carugo, K., Young, P., Gautel, M., and Saraste, M. (1999). Structure of the alpha‐actinin rod: Molecular basis for cross‐linking of actin filaments. Cell 98, 537–546.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

81

Dominguez, R., Freyzon, Y., Trybus, K. M., and Cohen, C. (1998). Crystal structure of a vertebrate smooth muscle myosin motor domain and its complex with the essential light chain: Visualization of the pre‐power stroke state. Cell 94, 559–571. Eakins, F., Al‐Khayat, H. A., Kensler, R. W., Morris, E. P., and Squire, J. M. (2002). 3D Structure of fish muscle myosin filaments. J. Struct. Biol. 137, 154–163. Ebashi, S., Endo, M., and Otsuki, I. (1969). Control of muscle contraction. Q. Rev. Biophys. 2, 351–384. Egelman, E. H., Francis, N., and De Rosier, D. J. (1982). F‐actin is a helix with a random variable twist. Nature 298, 131–135. Estes, J. E., Selden, L. A., and Gershman, L. C. (1981). Mechanism of action of phalloidin on the polymerization of muscle actin. Biochemistry 20, 708–712. Frank, J., Radermacher, M., Penczek, P., Zhu, J., Li, Y., Ladjadj, M., and Leith, A. (1996). SPIDER and WEB: Processing and visualization of images in 3D electron microscopy and related fields. J. Struct. Biol. 116, 190–199. Franzini‐Armstrong, C. (1999). The sarcoplasmic reticulum and the control of muscle contraction. FASEB J. 13(Suppl. 2), S266–S270. Franzot, G., Sjoblom, B., Gautel, M., and Djinovic Carugo, K. (2005). The crystal structure of the actin binding domain from alpha‐actinin in its closed conformation: Structural insight into phospholipid regulation of alpha‐actinin. J. Mol. Biol. 348, 151–165. Freundlich, A., and Squire, J. M. (1983). Three‐dimensional structure of the insect (Lethocerus) flight muscle M‐band. J. Mol. Biol. 169, 439–453. Gautel, M., Goulding, D., Bullard, B., Weber, K., and Furst, D. O. (1996). The central Z‐disk region of titin is assembled from a novel repeat in variable copy numbers. J. Cell. Sci. 109, 2747–2754. Goldstein, M. A., Michael, L. H., Schroeter, J. P., and Sass, R. L. (1988). Structural states in the Z band of skeletal muscle correlate with states of active and passive tension. J. Gen. Physiol. 92, 113–119. Gordon, A. M., Huxley, A. F., and Julian, F. J. (1966). The variation in isometric tension with sarcomere length in vertebrate muscle fibres. J. Physiol. 184, 170–192. Graceffa, P., and Dominguez, R. (2003). Crystal structure of monomeric actin in the ATP state. Structural basis of nucleotide‐dependent actin dynamics. J. Biol. Chem. 278, 34172–34180. Granzier, H. L., and Labeit, S. (2005). Titlin and its associated proteins: The third myofilament system of the sarcomere. Adv. Protein Chem. 71, 89–119. Hanson, J., and Lowy, J. (1963). The structure of F‐actin and of actin filaments isolated from muscle. J. Mol. Biol. 6, 48–60. Harford, J. J., and Squire, J. M. (1986). The ‘‘crystalline’’ myosin crossbridge array in relaxed bony fish muscles. Biophys. J. 50, 145–155. Haselgrove, J. C. (1972). X‐ray evidence for a conformational change in the actin‐ containing filaments of vertebrate striated muscle. Cold Spring Harbor Symp. Quant. Biol. 37, 341–352. Haselgrove, J. C. (1975). X‐ray evidence for conformational changes in the myosin filaments of vertebrate striated muscle. J. Mol. Biol. 92, 113–143. Haus, U., Hartmann, H., Trommler, P., Noegel, A. A., and Schleicher, M. (1991). F‐actin capping by cap32/34 requires heterodimeric conformation and can be inhibited with PIP2. Biochem. Biophys. Res. Commun. 181, 833–839. Herzberg, O., and James, M. N. (1985). Structure of the calcium regulatory muscle protein troponin‐C at 2.8 A˚ resolution. Nature 313, 653–659.

82

SQUIRE ET AL.

Herzberg, O., and James, M. N. (1988). Refined crystal structure of troponin C from turkey skeletal muscle at 2.0 A˚ resolution. J. Mol. Biol. 203, 761–779. Holmes, K. C., Popp, D., Gebhard, W., and Kabsch, W. (1990). Atomic model of the actin filament. Nature 347, 44–49. Holmes, K. C., Angert, I., Kull, F. J., Kahn, W., and Schroder, R. R. (2003). Electron cryo‐ microscopy shows how strong binding of myosin to actin releases nucleotide. Nature 425, 423–427. Hornemann, T., Kempa, S., Himmel, M., Hayess, K., Furst, D. O., and Wallimann, T. (2003). Muscle‐type creatine kinase interacts with central domains of the M‐band proteins myomesin and M‐protein. J. Mol. Biol. 332, 877–887. Houdusse, A., Kalabokis, V. N., Himmel, D., Szent‐Gyorgyi, A. G., and Cohen, C. (1999). Atomic structure of scallop myosin subfragment S1 complexed with MgADP: A novel conformation of the myosin head. Cell 97, 459–470. Houdusse, A., Szent‐Gyorgyi, A. G., and Cohen, C. (2000). Three conformational states of scallop myosin S1. Proc. Natl. Acad. Sci. USA 97, 11238–11243. Hudson, L., Harford, J. J., Denny, R. C., and Squire, J. M. (1997). Myosin head configuration in relaxed fish muscle: Resting state myosin heads must swing axially by up to 150 A or turn upside down to reach rigor. J. Mol. Biol. 273, 440–455. Huxley, A. F., and Niedergerke, R. (1954). Structural changes in muscle during contraction: Interference microscopy of living muscle fibres. Nature 173, 971–973. Huxley, A. F., and Simmons, R. M. (1971a). Mechanical properties of the cross‐bridges of frog striated muscle. J. Physiol. 218, 59P–60P. Huxley, A. F., and Simmons, R. M. (1971b). Proposed mechanism of force generation in striated muscle. Nature 233, 533–538. Huxley, H. E. (1957). The double array of filaments in cross‐striated muscle. J. Biophys. Biochem. Cytol. 3, 631–648. Huxley, H. E. (1963). Electron microscope studies on the structure of natural and synthetic protein filaments from striated muscle. J. Mol. Biol. 77, 281–308. Huxley, H. E. (1969). The mechanism of muscular contraction. Science 164, 1356–1365. Huxley, H. E. (1972). Structural changes in the actin‐ and myosin‐containing filaments during contraction. Cold Spring Harbor Symp. Quant. Biol. 37, 361–376. Huxley, H. E., and Brown, W. (1967). The low‐angle x‐ray diagram of vertebrate striated muscle and its behaviour during contraction and rigor. J. Mol. Biol. 30, 383–434. Huxley, H. E., and Hanson, J. (1954). Changes in the cross‐striations of muscle during contraction and stretch and their structural interpretation. Nature 173, 973–976. Kabsch, W., Mannherz, H. G., Suck, D., Pai, E. F., and Holmes, K. C. (1990). Atomic structure of the actin:DNase I complex. Nature 347, 37–44. Keep, N. H., Norwood, F. L., Moores, C. A., Winder, S. J., and Kendrick‐Jones, J. (1999). The 2.0 A˚ structure of the second calponin homology domain from the actin‐ binding region of the dystrophin homologue utrophin. J. Mol. Biol. 285, 1257–1264. Kendrew, J. C. (1963). Myoglobin and the structure of proteins. Science 139, 1259–1266. King, W. A., Stone, D. B., Timmins, P. A., Narayanan, T., von Brasch, A. A. M., Mendelson, R. A., and Curmi, P. M. G. (2005). Solution structure of the chicken skeletal muscle troponin complex via small‐angle neutron and X‐ray scattering. J. Mol. Biol. 345, 797–815. Knupp, C., and Squire, J. M. (2004). HELIX: A helical diffraction simulation program. J. Appl. Cryst. 37, 832–835. Knupp, C., Luther, P., and Squire, J. (2002). Titin organisation and the 3D architecture of the vertebrate‐striated muscle I‐band. J. Mol. Biol. 322, 731–739.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

83

Krieger, I., Kostyukova, A., Yamashita, A., Nitanai, Y., and Maeda, Y. (2002). Crystal structure of the C‐terminal half of tropomodulin and structural basis of actin filament pointed‐end capping. Biophys. J. 83, 2716–2725. Kulikovskaya, I., McClellan, G., Flavigny, J., Carrier, L., and Winegrad, S. (2003). Effect of MyBP‐C binding to actin on contractility in heart muscle. J. Gen. Physiol. 122, 761–774. Kunst, G., Kress, K. R., Gruen, M., Uttenweiler, D., Gautel, M., and Fink, R. H. (2000). Myosin binding protein C, a phosphorylation‐dependent force regulator in muscle that controls the attachment of myosin heads by its interaction with myosin S2. Circ. Res. 86, 51–58. Lindhout, D. A., and Sykes, B. D. (2003). Structure and dynamics of the C‐domain of human cardiac troponin C in complex with the inhibitory region of human cardiac troponin I. J. Biol. Chem. 278, 27024–27034. Littlefield, R., and Fowler, V. M. (1998). Defining actin filament length in striated muscle: Rulers and caps or dynamic stability? Annu. Rev. Cell. Dev. Biol. 14, 487–525. Liversage, A. D., Holmes, D., Knight, P. J., Tskhovrebova, L., and Trinick, T. (2001). Titin and the sarcomere symmetry paradox. J. Mol. Biol. 305, 401–409. Lorenz, M., Popp, D., and Holmes, K. C. (1993). Refinement of the F‐actin model against X‐ray fiber diffraction data by the use of a directed mutation algorithm. J. Mol. Biol. 234, 826–836. Lowy, J., Poulsen, F. R., and Vibert, P. J. (1970). Myosin filaments in vertebrate smooth muscle. Nature 225, 1053–1054. Ludtke, S. J., Baldwin, P. R., and Chiu, W. (1999). EMAN: Semiautomated software for high‐resolution single‐particle reconstructions. J. Struct. Biol. 128, 82–97. Luther, P. K. (1991). Three‐dimensional reconstruction of a simple Z‐band in fish muscle. J. Cell. Biol. 113, 1043–1055. Luther, P. K. (2000). Three‐dimensional structure of a vertebrate muscle Z‐band: Implications for yitin and a‐actinin binding. J. Struct. Biol. 129, 1–16. Luther, P. K., and Crowther, R. A. (1984). Three‐dimensional reconstruction from tilted sections of fish muscle M‐band. Nature 307, 566–568. Luther, P., and Squire, J. M. (1978). Three‐dimensional structure of the vertebrate muscle M‐region. J. Mol. Biol. 125, 313–324. Luther, P. K., and Squire, J. M. (1980). Three‐dimensional structure of the vertebrate muscle A‐band. II. The myosin filament superlattice. J. Mol. Biol. 141, 409–439. Luther, P. K., and Squire, J. M. (2002). Muscle Z‐band ultrastructure: Titin Z‐repeats and Z‐band periodicities do not match. J. Mol. Biol. 319, 1157–1164. Luther, P. K., Munro, P. M., and Squire, J. M. (1981). Three‐dimensional structure of the vertebrate muscle A‐band. III. M‐ region structure and myosin filament symmetry. J. Mol. Biol. 151, 703–730. Luther, P. K., Squire, J. M., and Forey, P. L. (1996). Evolution of myosin filament arrangements in vertebrate skeletal muscle. J. Morphol. 229, 325–335. Luther, P. K., Barry, J. S., and Squire, J. M. (2002). The three‐dimensional structure of a vertebrate wide (slow muscle) Z‐band: Lessons on Z‐band assembly. J. Mol. Biol. 315, 9–20. Luther, P. K., Padron, R., Ritter, S., Craig, R., and Squire, J. M. (2003). Heterogeneity of Z‐band structure within a single muscle sarcomere: Implications for sarcomere assembly. J. Mol. Biol. 332, 161–169.

84

SQUIRE ET AL.

Luther, P. K., Hillenbrand, G., and Squire, J. M. (2005). Three dimensional structure of the superlattice M-band in vertebrate muscle. J. Mol. Biol. (in revision). Lymn, R. W., and Taylor, E. W. (1971). Mechanism of adenosine triphosphate hydrolysis by actomyosin. Biochemistry 10, 4617–4624. Mayans, O., van der Ven, P. F., Wilm, M., Mues, A., Young, P., Furst, D. O., Wilmanns, M., and Gautel, N. (1998). Structural basis for activation of the titin kinase domain during myofibrillogenesis. Nature 395, 863–869. McComas, A. J. (1996). Skeletal Muscle: Form and Function. Human Kinetics Europe Ltd. McGough, A., Pope, B., Chiu, W., and Weeds, W. (1997). Cofilin changes the twist of F‐actin: Implications for actin filament dynamics and cellular function. J. Cell. Biol. 138, 771–781. McLachlan, A. D., and Stewart, M. (1976). The 14‐fold periodicity in alpha‐tropomyosin and the interaction with actin. J. Mol. Biol. 103, 271–298. Metropolis, N., Rosenbluth, A. W., Rosenbluth, M. N., Teller, A. H., and Teller, E. (1958). Equations of state calculations by fast computing machines. J. Chem. Phys. 21, 1087–1092. Millevoi, S., Trombitas, K., Kolmerer, b., Kostin, S., Schaper, J., Pelin, K., Granzier, H., and Labeit, S. (1998). Characterization of nebulette and nebulin and emerging concepts of their roles for vertebrate Z‐discs. J. Mol. Biol. 282, 111–123. Moolman‐Smook, J., Flashman, E., De Lange, W., Li, Z., Corfield, V., Redwood, C., and Watkins, H. (2002). Identification of novel interactions between domains of myosin binding protein‐C that are modulated by hypertrophic cardiomyopathy missense mutations. Circ. Res. 91, 704–711. Moore, P. B., Huxley, H. E., and De Rosier, D. J. (1970). Three‐dimensional reconstruction of F‐actin, thin filaments and decorated thin filaments. J. Mol. Biol. 50, 279–295. Moos, C., Mason, C. M., Besterman, J. M., Feng, I. N., and Dubin, J. H. (1978). The binding of skeletal muscle C‐protein to F‐actin, and its relation to the interaction of actin with myosin subfragment–1. J. Mol. Biol. 124, 571–586. Morris, E. P., Nneji, G., and Squire, J. M. (1990). The three‐dimensional structure of the nemaline rod Z‐band. J. Cell. Biol. 111, 2961–2978. Morris, E. P., Squire, J. M., and Fuller, G. W. (1991). The 4‐stranded helical arrangement of myosin heads on insect (Lethocerus) flight muscle thick filaments. J. Struct. Biol. 107, 221–226. Mues, A., van der Ven, P. F., Young, P., Furst, D. O., and Gautel, M. (1998). Two immunoglobulin‐like domains of the Z‐disc portion of titin interact in a conformation‐dependent way with telethonin. FEBS Lett. 428, 111–114. Muirhead, H., and Perutz, M. F. (1963). Structure of haemoglobin. A three‐dimensional Fourier synthesis of reduced human haemoglobin at 5–5 A˚ resolution. Nature 199, 633–638. Nelder, J. A., and Mead, R. (1965). The simplex method for function minimisation. Comput. J. 7, 308–313. Obermann, W. M., Gautel, M., Steiner, F., van der Ven, P. F., Weber, K., and Furst, D. O. (1996). The structure of the sarcomeric M band: Localization of defined domains of myomesin, M‐protein, and the 250‐kD carboxy‐terminal region of titin by immunoelectron microscopy. J. Cell. Biol. 134, 1441–1453. Obermann, W. M., Gautel, M., Weber, K., and Furst, D. O. (1997). Molecular structure of the sarcomeric M band: Mapping of titin and myosin binding domains in myomesin and the identification of a potential regulatory phosphorylation site in myomesin. EMBO J. 16, 211–220.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

85

Oda, T., Makino, K., Yamashita, I., Namba, K., and Maeda, Y. (1998). Effect of the length and effective diameter of F‐actin on the filament orientation in liquid crystalline sols measured by X‐ray fiber diffraction. Biophys. J. 75, 2672–2681. Oda, T., Namba, K., and Maeda, Y. (2005). Position and orientation of phalloidin in Factin determined by X-ray fiber diffraction analysis. Biophys. J. 88, 2727–2736. Offer, G., Moos, C., and Starr, R. (1973). A new protein of the thick filaments of vertebrate skeletal myofibrils. Extractions, purification and characterization. J. Mol. Biol. 74, 653–676. Oshima, K., Takezawa, Y., Sugimoto, Y., Kiyotoshi, M., and Wakabayashi, K. (2003). Modeling analysis of myosin‐based meridional X‐ray reflections from frog skeletal muscles in relaxed and contracting states. Adv. Exp. Med. Biol. 538, 243–249. Otterbein, L. R., Graceffa, P., and Dominguez, R. (2001). The crystal structure of uncomplexed actin in the ADP state. Science 293, 708–711. Padron, R., Alamo, L., Murgich, J., and Craig, R. (1998). Towards an atomic model of the thick filaments of muscle. J. Mol. Biol. 275, 35–41. Papa, I., Astier, C., Kwiatek, O., Raynaud, F., Bonnal, C., Lebart, M. C., Roustan, C., and Benyamin, Y. (1999). Alpha actinin‐CapZ, an anchoring complex for thin filaments in Z‐line. J. Muscle Res. Cell Motil. 20, 187–197. Parry, D. A. D., and Squire, J. M. (1973). Structural role of tropomyosin in muscle regulation: Analysis of the X‐ray diffraction patterns from relaxed and contracting muscles. J. Mol. Biol. 75, 33–55. Pask, H. T., Jones, K. L., Luther, P. K., and Squire, J. M. (1994). M‐band structure, M‐bridge interactions and contraction speed in vertebrate cardiac muscles. J. Muscle Res. Cell Motil. 15, 633–645. Patwardhan, A., Paul, D., Al‐Khayat, H. A., and Morris, E. P. (2004). A measure for the angle between projections based on the extent of correlation between corresponding central sections. J. Mol. Biol. 344, 707–724. Paul, D., Patwardhan, A., Squire, J. M., and Morris, E. P. (2004). Single particle analysis of filamentous and highly elongated macromolecular assemblies. J. Struct. Biol. 148, 236–250. Pringle, J. W. S. (1957). Insect Flight. Cambridge University Press, Cambridge, UK. Pringle, J. W. S. (1968). Mechano‐chemical transformation in striated muscle. In ‘‘Aspects of Cell Motility,’’ Vol. 22, pp. 67–86. Cambridge University Press, Cambridge, UK. Radermacher, M. (1992). ‘‘Weighted Back‐Projection Methods.’’ Plenum, New York. Rayment, I., Holden, H. M., Whittaker, M., Yohn, C. B., Lorenz, M., Holmes, K. C., and Milligan, R. A. (1993a). Structure of the actin‐myosin complex and its implications for muscle contraction. Science 261, 58–65. Rayment, I., Rypniewski, W. R., Schmidt‐Base, K., Smith, R., Tomchick, D. R., Benning, M. M., Winkelmann, D. A., Wesenberg, G., and Holden, H. M. (1993b). Three‐ dimensional structure of myosin subfragment–1: A molecular motor. Science 261, 50–58. Reedy, M. K. (1968). Ultrastructure of insect flight muscle. I. Screw sense and structural grouping in the rigor cross‐bridge lattice. J. Mol. Biol. 31, 155–176. Reedy, M. K., Holmes, K. C., and Tregear, R. T. (1965). Induced changes in orientation of the cross‐bridges of glycerinated insect flight muscle. Nature 207, 1276–1280. Schmitz, H., Lucaveche, C., Reedy, M. K., and Taylor, K. A. (1994). Oblique section 3‐D reconstruction of relaxed insect flight muscle reveals the cross‐bridge lattice in helical registration. Biophys. J. 67, 1620–1633.

86

SQUIRE ET AL.

Sheterline, P., Clayton, J., and Sparrrow, J. C. (1998). ‘‘Actin.’’ Oxford University Press, Oxford. Sjostrom, M., and Squire, J. M. (1977). Fine structure of the A‐band in cryo‐sections. The structure of the A‐band of human skeletal muscle fibres from ultra‐thin cryo‐sections negatively stained. J. Mol. Biol. 109, 49–68. Small, J. V., and Squire, J. M. (1972). Structural basis of contraction in vertebrate smooth muscle. J. Mol. Biol. 67, 117–149. Sodek, J., Hodges, R. S., Smillie, L. B., and Jurasek, L. (1972). Amino acid sequence of rabbit skeletal tropomyosin and its coiled-coil structures. Proc. Nat. Acad. Sci. USA 69, 3800–3804. Squire, J. (1981). ‘‘The Structural Basis of Muscular Contraction’’ Plenum Press, New York. Squire, J. M. (1971). General model for the structure of all myosin‐containing filaments. Nature 233, 457–462. Squire, J. M. (1972). General model of myosin filament structure. II. Myosin filaments and cross‐bridge interactions in vertebrate striated and insect flight muscles. J. Mol. Biol. 72, 125–138. Squire, J. M. (1973). General model of myosin filament structure. 3. Molecular packing arrangements in myosin filaments. J. Mol. Biol. 77, 291–323. Squire, J. M. (1986). Muscle myosin filaments: Internal structure and crossbridge organisation. Comments Molec. Cel. Biophy. 3, 155–177. Squire, J. M. (2000). Fibre and muscle diffraction. In ‘‘Structure and Dynamics of Biomolecules’’ (E. Fanchon, E. Geissler, L.‐L. Hodeau, J.‐R. Regnard, and P. Timmins, Eds.). Oxford University Press, Oxford. Squire, J. M., Harford, J. J., Edman, A. C., and Sjostrom, M. (1982). Fine structure of the A‐band in cryo‐sections. III. Crossbridge distribution and the axial structure of the human C‐zone. J. Mol. Biol. 155, 467–494. Squire, J. M., and Knupp, C. (2004). MusLABEL: A program to model striated muscle A‐band lattices, to explore crossbridge interaction geometries and to simulate muscle diffraction patterns. J. Mus. Res. Cell Motil. 25, 423–438 Squire, J. M., and Knupp, C. (2005). X‐ray diffraction studies of the muscle and the crossbridge cycle. Adv. Protein Chem. 71, 195–255. Squire, J. M., Knupp, C., Roessle, M., AL‐Khayat, H. A., Irving, T. C., Eakins, F., Mok, N. S., Harford, J. J., and Reedy, M. K. (2005). X‐ray diffraction studies of striated muscles. In ‘‘Mysteries about the Sliding Filament Mechanism: Fifty Years After its Proposal’’ , (H. Sugi, Ed.) Adv. Exp. Med. Biol. Kluwer, In press. Squire, J. M., Luther, P. K., and Knupp, C. (2003). Structural evidence for the interaction of C‐protein (MyBP‐C) with actin and sequence identification of a possible actin‐ binding domain. J. Mol. Biol. 331, 713–724. Squire, J. M., and Morris, E. P. (1998). A new look at thin filament regulation in vertebrate skeletal muscle. FASEB J. 12, 761–771. Squire, J. M., and Parry, D. A. D. (2005). Comparative motile mechanisms in cells. Adv. Protein Chem. 71, 1–15. Squire, J. M., Roessle, M., and Knupp, C. (2004). New X‐ray diffraction observations on vertebrate muscle: Organisation of C‐protein (MyBP‐C) and troponin and evidence for unknown structures in the vertebrate A‐band. J. Mol. Biol. 343, 1345–1363. Stewart, M., and McLachlan, A. D. (1975). Fourteen actin‐binding sites on tropomyosin? Nature 257, 331–333. Stewart, M., Kensler, R. W., and Levine, R. J. (1981). Structure of Limulus telson muscle thick filaments. J. Mol. Biol. 153, 781–790.

MOLECULAR ARCHITECTURE IN MUSCLE CONTRACTILE ASSEMBLIES

87

Stewart, M., Kensler, R. W., and Levine, R. J. (1985). Three‐dimensional reconstruction of thick filaments from Limulus and scorpion muscle. J. Cell. Biol. 101, 402–411. Stone, D., and Smillie, L. B. (1978). The amino acid sequence of rabbit skeletal a‐tropomyosin: The N‐terminal half and complete sequence. J. Biol. Chem. 253, 1137–1148. Stone, D. B., Timmins, P. A., Schneider, D. K., Krylova, I., Ramos, C. H. I., Reinach, F. C., and Mendelson, R. A. (1998). The effect of regulatory Ca2þ on the in situ strucutres of troponin C and troponin I: A neutron scattering study. J. Mol. Biol. 281, 689–704. Takeda, S., Yamashita, A., Maeda, K., and Maeda, Y. (2003). Structure of the core domain of human cardiac troponin in the Ca(2þ)‐saturated form. Nature 424, 35–41. Timson, D. J., Trayer, H. R., and Trayer, I. P. (1998). The N‐terminus of A1‐type myosin essential light chains binds actin and modulates myosin motor function. Eur. J. Biochem. 255, 654–662. Trinick, J. A. (1981). End‐filaments: A new structural element of vertebrate skeletal muscle thick filaments. J. Mol. Biol. 151, 309–314. van Heel, M., Gowen, B., Matadeen, R., Orlova, E. V., Finn, R., Pape, T., Cohen, D., Stark, H., Schmidt, R., Schatz, M., and Patwardhan, A. (2000). Single‐particle electron cryo‐microscopy: Towards atomic resolution. Q. Rev. Biophys. 33, 307–369. van Heel, M., Harauz, G., Orlova, E. V., Schmidt, R., and Schatz, M. (1996). A new generation of the IMAGIC image processing system. J. Struct. Biol. 116, 17–24. Vibert, P. (1992). Helical reconstruction of frozen‐hydrated scallop myosin filaments. J. Mol. Biol. 223, 661–671. Wallgren‐Petterson, C., and Laing, N. G. (2003). 109th ENMC International Workshop: 5th Workshop on nemaline myopathy, 11–13 October 2002, Naarden, The Netherlands. Neuromuscular Disorders 13, 501–507. Watson, J. D., and Crick, F. H. C. (1953). Molecular structure of nucleic acids; a structure for deoxyribose nucleic acid. Nature 171, 737–738. Wear, M. A., and Cooper, J. A. (2004). Capping protein: New insights into mechanism and regulation. Trends Biochem. Sci. 29, 418–428. Wilkins, M. H. F., Stokes, A. R., and Wilson, R. H. (1953). Molecular structure of deoxypentose nucleic acids. Nature 171, 738–740. Yamaguchi, M., Izumimoto, M., Robson, R. M., and Stromer, M. H. (1985). Fine structure of wide and narrow vertebrate muscle Z‐lines. A proposed model and computer simulation of Z‐line architecture. J. Mol. Biol. 184, 621–643. Yamashita, A., Maeda, K., and Maeda, Y. (2003). Crystal structure of CapZ: Structural basis for actin filament barbed end capping. EMBO J. 22, 1529–1538. Ylanne, J., Scheffzek, K., Young, P., and Saraste, M. (2001). Crystal structure of the alpha‐ actinin rod reveals an extensive torsional twist. Structure (Camb). 9, 597–604. Zhang, J. Q., Weisberg, A., and Horowits, R. (1998). Expression and purification of large nebulin fragments and their interaction with actin. Biophys. J. 74, 349–359. Zhou, Q., Ruiz‐Lozano, P., Martone, M. E., and Chen, J. (1999). Cypher, a striated muscle‐restricted PDZ and LIM domain‐containing protein, binds to alpha‐ actinin–2 and protein kinase C. J. Biol. Chem. 274, 19807–19813.

TITIN AND ITS ASSOCIATED PROTEINS: THE THIRD MYOFILAMENT SYSTEM OF THE SARCOMERE By HENK L. GRANZIER* AND SIEGFRIED LABEIT{ *Department of Veterinary and Comparative Anatomy, Pharmacology and Physiology, Washington State University, Pullman, Washington; { Institut fu¨r Ana¨sthesiologie und Operative Intensivmedizin, Universita¨tsklinikum Mannheim, Mannheim 68167, Germany

I.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Early History of Titin Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Functional Genomics of Titin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Structure and Features of the Human Titin Gene . . . . . . . . . . . . . . . . . . . . . B. Differential Splicing of Titin: Exon Shuffling Creates Functionally Diverse Titin Isoforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. A Titin Exon Microarray as a Novel Tool to Analyze Differentially Expressed Titin Isoforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Molecular Mechanism of Titin Elasticity and Its Regulation . . . . . . . . . . . . . . . . A. Titin Subsegments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. The Titin Filament System in the Sarcomere and Its Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Titin‐Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I.

89 89 91 91 94 97 98 98 110 110 114 114

Introduction

This articles focuses on current knowledge of titin, the third myofilament of the sarcomere. The early history that led to the discovery of titin is discussed first. Then we turn to functional genomics of titin, including an extensive discussion of differential splicing in the I‐band region of the molecule. This region is elastic and we discuss the molecular mechanism of this elasticity and how it can be modulated by differential splicing and posttranslational modifications. The article ends with a discussion of titin‐ binding proteins, including the possible roles of titin‐based protein complexes in cell signaling.

A.

Early History of Titin Research

The two‐filament model of the sarcomere was proposed half a century ago (Huxley and Hanson, 1954; Huxley and Niedergerke, 1954) and has been proved to be highly successful in explaining many features of contracting muscle. However, it was realized early that the model is unable to ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71003-7

89

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

90

GRANZIER AND LABEIT

account for features of passive muscle, including long‐range muscle elasticity (reviewed in Wang, 1985). This elasticity is independent of interaction between thin and thick filaments and its existence is demonstrated well by stretched relaxed single myofibrils that rapidly shorten on release. Electron microscopy during the 1950s and 1960s provided compelling evidence for the existence of a third type of endosarcomeric filament with possible elastic properties (Huxley and Peachey, 1961; Locker and Leet, 1976; Trombitas and Tigyi‐Sebes, 1974). The clearest evidence was provided in sarcomeres stretched beyond overlap of thin and thick filaments that revealed slender filaments emanating from the ends of the thick filaments. We know now that these so‐called gap filaments are the visible portions of a more extensive elastic filament system. Studies directed at understanding the basis of muscle elasticity conducted by Maruyama and colleagues revealed that extensively washing myofibrils with a range of extraction solutions gives rise to an extraction‐resistant residue that resembles an elastic and insoluble protein gel (Maruyama, 1976). The largest protein purified from this residue was named connectin (Maruyama et al., 1981). Studies conducted at approximately the same time by Wang and colleagues, with the aim of searching for filamin‐like proteins in vertebrate striated muscle, failed to find filamin but instead uncovered a megadalton‐ sized protein that was named titin (Wang et al., 1979). It is now clear that connectin and titin are the same elastic protein. Important advances during the first decade of titin research include establishing the layout of the titin filament in the sarcomere and the discovery that the filament is elastic as revealed by the passive force that is developed by titin in response to sarcomere stretch. The overall layout was established with immunoelectron microscopy of muscles that were stretched to various lengths and labeled with monoclonal antibodies raised against titin (Furst et al., 1988; Itoh et al., 1988; Trombitas and Pollack, 1993; Wang et al., 1991b; Whiting et al., 1989). This revealed that a single titin polypetide spans the full distance from the Z‐disk to the M‐line region of the sarcomere and, importantly, that the I‐band region of the molecule extends as the sarcomere is stretched, but that the A‐band region is inextensible (Fig. 1). Sequencing studies conducted during the early 1990s revealed that titin’s A‐band portion is composed of regular patterns of immunoglobulin (Ig)‐like and fibronectin type 3 (Fn3) repeats (Labeit and Kolmerer, 1995; Labeit et al., 1990). Titin’s extensible region is mainly composed of tandem Ig segments (tandemly arranged Ig‐like domains), and the so‐called PEVK segment (rich in proline [P], glutamate [E], valine [V], and lysine [K] amino acid residues (see below). It has been speculated that the inextensible A‐band region of titin may perform structural roles such as thick‐filament length regulation (Trinick, 1994; Wang,

TITIN AND ITS ASSOCIATED PROTEINS

91

1996). Support for the proposal that titin is elastic and develops force in response to sarcomere stretch was obtained by Podolsky and colleagues who used low doses of ionizing radiation to degrade titin and showed that this greatly reduced the ability of relaxed skeletal muscle fibers to generate passive force (Horowits et al., 1986). This effect was accompanied by axial misalignment of thick filaments. Thus, since the introduction of the two‐ filament model of the sarcomere, an additional elastic filament has been discovered that consists of the giant elastic protein titin (or connectin). Early research showed that single polypeptides of titin span from Z‐disk to M‐line, and that titin performs roles both in passive tension generation and in maintaining the central position of the thick filament in the sarcomere, ensuring a balance of forces in the two half sarcomeres during contractile cycles. Next we discuss the properties of titin in greater detail and its additional functions discovered more recently.

II.

Functional Genomics of Titin

A. Structure and Features of the Human Titin Gene 1. Organization at the Protein Level Titin is an extreme example of a modular protein. Of the 4200 kDa titin polypeptide mass encoded by the titin gene, ~90% is organized into modular repeats, spread throughout the molecule. For example, within titin’s Z‐disk region, seven copies of 45‐residue Z‐repeats are present, each of which share ~50% sequence homology (Gautel et al., 1996). The I‐band region of titin is composed of both tandemly arranged Ig domains and 28–30‐residue PEVK repeats (also called PPAK repeats, Greaser et al., 2002). Finally, titin’s A‐band portion is composed of regular patterns of Ig and Fn3 repeats. Only a small fraction of titin’s sequence is unique with no clear homology to other protein sequences in the databases. Titin’s modular organization is found back in its exon organization at the genomic level. For example, titin’s exons 8 to 14 correspond to ~135 bp exons that code for ~45‐residue Z‐repeats (Zr#1 to Zr#7). Similarly, most of the tandemly arranged Ig domains and the 28–30‐residue PEVK repeats are organized into single exons. This organization provides extensive opportunity for creating titin isoforms through differential splicing. Isoform diversity of titin had been suggested previously both by gel electrophoresis and cDNA transcript analysis (for review, see Granzier and Labeit, 2004). Knowledge of titin structure now reveals a series of exon‐shuffling pathways as the structural basis for titin isoform diversity, and consequently during the past

92 GRANZIER AND LABEIT

TITIN AND ITS ASSOCIATED PROTEINS

93

3 years, the genome sequencing projects have had an important impact on functional titin studies.

2. Organization at the Genomic Level Titin is encoded by a single gene that is highly conserved within vertebrates. In human and mouse it has been shown that the gene is located on the long arm of chromosome 2, region 2q31 (Labeit et al., 1990; Muller‐ Seitz et al., 1993; Rossi et al., 1994). Already, during the initial draft of the human genome, it was noted that the titin locus corresponds to a region with 174 exons, including the largest known exon (17 kb) of the human genome (Lander et al., 2001). The completed human titin gene sequence, together with its alignment with sequenced cDNAs (such as from adult skeletal and cardiac muscles; (Labeit and Kolmerer, 1995) and fetal muscle cDNAs (Ma and Wang, 2003), identified >100 additional PEVK exons. These PEVK exons are 70–110 bp in length, have atypical codon usage, and were therefore missed by exon search programs used in the early draft assembly of the human genome. The current version of the titin gene (accession AJ277892.2) lists 363 exons within a 280 kb genomic segment. Exons 45, 46, and 48 of the titin gene sequence had been missed during previous cDNA sequencing projects because they are one to two orders of magnitude less abundant than other titin exons. These exons, referred to as novel exons (Novex) 1‐III are located within titin’s extensible I‐band region and are not included in conventional titin isoforms. (Note that the novex exons are referred to by Roman numerals, but that for the protein [see below] Arabic numerals are used.) Consistent with this, we identified novex‐III transcripts with poly‐A‐tails (Bang et al., 2001). Thus, two exons within the transcriptional unit of titin (exon 48 (novex‐III) and exon 363) drive polyadenylation. Novex‐III polyadenylation leads to a ~650 kDa titin isoform (known as novex-3 titin), while the conventional titins are obtained by exon-48 exclusion and exon-363 polyadenylation. Only the exon-363 polyadenylated titins correspond to the abundant, half‐ sarcomere spanning, muscle protein. In contrast, the novex-3 titin isoform

Fig. 1. Layout of titin in the sarcomere. Center panel: electron micrograph of sarcomere of stretched soleus muscle fiber, labeled with anti‐titin antibodies that demarcate the tandem Ig and PEVK spring elements of titin’s extensible I‐band region. Superimposed are two schematic titin molecules (one for each half sarcomere). Top and bottom panels: domain structure of I‐band and A‐band sequence of titin, respectively (from Labeit and Kolmerer, 1995). Bottom left: length of tandem Ig segment (proximal þ distal segment) and PEVK segment in human soleus fibers, as function of sarcomere length (based on Trombitas et al., 1998b).

94

GRANZIER AND LABEIT

is much less abundant and less well characterized. Immunoelectron microscopy studies suggest that the novex-3 titin isoform projects about 100 nm into the I‐band and is extensible. The novex‐III exon contains binding motifs for obscurin, another giant muscle protein with multiple signaling motifs (Bang et al., 2001; Young et al., 2001) and it has been speculated that the novex-3 titin isoform serves as a biomechanically active signaling molecule (Bang et al., 2001). Support for this is derived from the elevated expression of novex-3 titin in cardiomyopathies (Wu et al., 2002).

B. Differential Splicing of Titin: Exon Shuffling Creates Functionally Diverse Titin Isoforms Before the cloning of titin, low‐percentage gradient gels suggested the presence of titin isoforms in different skeletal muscle tissues (Hu et al., 1986; Wang et al., 1991a). With the availability of titin cDNA sequences from a variety of species, it became possible to perform qualitative surveys for titin isoform expression and, for example, to compare skeletal and cardiac titin isoforms. This work suggested that several regions exist within the titin gene where extensive differential splicing occurs. The comparison of the gene sequence with the transcript variants revealed several distinct exon shuffling pathways, which are discussed below.

1.

Exon Shuffling Pathways

a. Exons 8–14, Z‐repeats. Exons 8–14 code for seven copies of ~45‐ residue repeats. Immunoelectron microscopy has indicated that these repeats are located centrally within the Z‐band lattice, and biochemical studies indicated that Z‐repeats interact with a‐actinin (Gregorio et al., 1998; Young and Gautel, 2000). These ~45‐residue repeats have been called Z‐repeats (Gautel et al., 1996). Titin’s Z‐repeats are differentially spliced: of the tissues studied so far, only mammalian heart tissues express all seven Z‐repeats and skeletal muscle tissues skip one or three Z‐repeats (Sorimachi et al., 1997). An attractive hypothesis is that the differential splicing of titin’s Z‐repeats regulates the different Z‐band widths found in vertebrate striated muscles (Gautel et al., 1996). However, the finding that the span of Z‐repeats does not correspond to the Z‐band periodicity (Luther and Squire, 2002) is inconsistent with a one‐to‐one correspondence of the Z‐repeats and a‐actinin cross‐links in the Z‐band, although two Z‐repeats per cross‐link might be possible. Alternatively, the differential expression of titin’s Z‐repeats may modulate the biomechanical properties of the Z‐disk lattice.

TITIN AND ITS ASSOCIATED PROTEINS

95

b. Exon 48, Novex–3 Pathway. As described previously, exon 48 corresponds to the novex‐III exon that serves as an alternative C‐terminus. Therefore, exclusion of exon 48 results in a full‐length >3000 kDa half‐ sarcomere spanning titin, and inclusion in an unconventional ~650 kDa titin (Bang et al., 2001). The physiological function of this binary switch in titin is unknown, but might relate to modulating titin’s interaction with obscurin. c. Exons 49–224, Titin’s Central I‐band Region. This segment contains three distinct spring elements: (1) exon 49 codes for the cardiac‐specific so‐called N2B spring element, (2) exons 50–101 code for tandemly repeated Ig domains (exons 102–108 for the N2A element), and (3) exons 109–224 encode the PEVK region of titin (Bang et al., 2001). Most PEVK exons code for conserved ~28‐residue PEVK repeats and 10 are more complex and encode E‐rich motifs (Bang et al., 2001). Within the segment that contains exons 49–224, a series of highly complex splice decisions are taken that control the inclusion of these three different spring elements and their copy numbers. The multiple splice pathways in titin’s central I‐band region give rise to isoforms with distinct spring compositions. All isoforms identified so far contain a constitutively expressed region that consists of 15 Ig domains near the Z‐disk (that are part of the so‐called proximal tandem Ig segment), 22 Ig domains located close to the A‐band (that make up the distal tandem Ig segment), and a ~180‐residue–containing PEVK segment (Freiburg et al., 2000). In addition, all skeletal muscle isoforms contain the N2A element, a variable number of additional Ig domains located in the proximal tandem Ig segment, and a variable number of additional PEVK residues. As for cardiac titins, two classes of isoform exist: the small 2970 kDa cardiac isoform known as N2B titin (so named because it contains the N2B element) and the large so‐called N2BA titin isoform (the name reflects the presence of both N2B and N2A elements). The N2BA titins contain additional PEVK residues and a variable number of additional Ig domains. As a result, N2BA titins are larger than N2B titin and vary in size from ~3200 kDa to ~3400 kDa. Due to their longer extensible I‐band region, N2BA titins are less stiff (a given degree of sarcomere stretch results in a lower fractional extension of titin, also see below) than N2B titin. Large mammals co‐express N2B and N2BA titins, with co‐expression occurring at the level of the half‐sarcomere (Trombitas et al., 2001). The co‐expression ratio is not fixed but can vary; a prominent example of this occurs during fetal and neonatal heart development (also see below). How the precise developmental and tissue‐specific regulation of splicing is achieved remains to be determined.

96

GRANZIER AND LABEIT

Fig. 2. (A) Analysis of human adult skeletal (HAS) and cardiac muscle (HAC) transcripts by an oligonucleotide array representing all 363 exons of the human titin gene. Left: Comparison of results obtained with HAS (top) and HAC (bottom) transcripts reveals large blocks of exons that are only positive in skeletal muscle. Right: Examples of constitutively expressed exons (exon 5 and 7), cardiac specific exons (11 and 49), and skeletal muscle specific exons (156 and 210). 5MM: a 50mer from exon 5 including five base‐pair mismatches as a control for hybridization

TITIN AND ITS ASSOCIATED PROTEINS

97

d. Exon 362, M‐line Titin. Cardiac titin isoforms include exon 362, whereas skeletal muscles skip exon 362 to different extents; the ratio of exon–362 expression differs in rabbit from roughly about 90% inclusion in soleus muscle to about 10% inclusion in extensor digitorum longus muscle (Kolmerer et al., 1996). Exon 362 binds calpain 3 (also called p94) (Sorimachi et al., 1995). Therefore, skipping or inclusion of exon 362 might be linked to regulation of calpain 3–dependent signaling pathways. C. A Titin Exon Microarray as a Novel Tool to Analyze Differentially Expressed Titin Isoforms The multiple splice pathways described above can generate highly complex titin splice isoforms. Currently, it is unknown whether the distinct exon shuffling pathways within the Z‐disk, I‐band, and M‐line regions of titin are coordinated. Answering this issue will require powerful methods to type titin isoforms. Analysis by reverse transcriptase polymerase chain reaction (RT‐PCR) with specific primer pairs is laborious and might miss the global picture. Recently, a high‐resolution agarose gel system has become available for separation of titin isoforms, and this has more firmly established that the different titin splice variants are indeed translated into distinct titin polypeptides (Warren et al., 2003b). However, for titin‐sized proteins, these gels can only resolve isoforms that differ >~100 kDa. We recently developed an exon microarray with which the expression levels of all 363 titin exons can be determined simultaneously (Fig. 2A). When using all three methods in conjunction (i.e., high‐ resolution protein gels, RT‐PCR, and the titin exon array), detailed insights can be obtained into the differential expression of cardiac titin isoforms in health and disease. We used this approach and analyzed expression of cardiac titin isoforms during fetal and postnatal heart development. We discovered fetal N2BA titin isoforms, characterized by additional spring elements both in the tandem Ig and PEVK region of the molecule (Fig. 2B). The fetal N2BA isoform predominates in fetal myocardium and gradually disappears during postnatal development with a time course that varies (from days to weeks) in different species (Lahmers et al., stringency. (B) Differential expression of titin exons in human fetal cardiac versus human adult cardiac muscle. Left: examples of differentially expressed exons. The two rows per sample show duplicate spots and reveal reproducibility. Exon 50 (top) is an example of a constitutively expressed exon. Left, bottom: domain organization of I‐band region of largest N2BA cardiac isoform known, with Ig domains in red and PEVK in yellow. Many of the exons that are upregulated in fetal myocardium have not been previously described in cardiac transcripts, as indicated by insertion arrows. (Figure based on Lahmers et al., 2004.)

98

GRANZIER AND LABEIT

2004). As expected, passive stiffness is low if expression of fetal titin isoforms dominates (discussed further below).

III.

Molecular Mechanism of Titin Elasticity and Its Regulation

Ultrastructural studies with antibodies specific to titin’s I‐band region have demonstrated that titin’s central I‐band region behaves extensibly on myofibrillar stretch. Analysis of titin’s primary structure revealed specific motif families within its I‐band region that are now known to act as molecular springs: (1) tandem Ig segments, (2) the PEVK segment, and (3) the N2B‐Us segment. The first two elements are found in both skeletal muscle and cardiac muscle titins, but the N2B‐Us is cardiac specific. In slack sarcomeres, titin’s extensible region is in a shortened state and on sarcomere stretch; extension of the tandem Ig segments initially dominates, followed by dominating PEVK extension (see Fig. 1, bottom left). The unique sequence of the cardiac‐specific N2B element is a major source of extensibility toward the upper limit of the physiological sarcomere length range of cardiac muscle. Thus, titin’s extensible region consists of subsegments with distinct extensible behaviors.

A. Titin Subsegments 1.

Titin as Wormlike Chain

In order to understand the molecular basis of titin’s elasticity, the mechanical properties of titin have been explored in single‐molecule manipulation experiments, using laser tweezers and atomic force microscopy (AFM). An important finding of early work was that titin molecules behave as wormlike‐chain (WLC) entropic springs (Kellermayer et al., 1997; Rief et al., 1997; Tskhovrebova et al., 1997). The WLC model describes the molecule as a deformable continuum of persistence length P (measure of bending rigidity, or minimal distance along chain for orientations to be uncorrelated) and contour length C. The end‐to‐end length of the molecule is z and the fractional extension is z/C. The force (F) developed in response to stretch is a function of the fractional extension and persistence length: FP z 1 ¼ þ kB T C 4ð1 z=C Þ2

1 4

ð1Þ

TITIN AND ITS ASSOCIATED PROTEINS

99

(Bustamante et al., 1994). According to this equation, force increases in a nonlinear fashion with fractional extension, and the force response is inversely proportional to the persistence length, P . Note that if the flexibility of a chain is high, P is short, and a relatively high force is needed to extend the chain.

2. Tandem Ig Segments Early work on single titin molecules also focused on unfolding of titin’s globular domains. Titin’s Ig domains are ~4 nm in length and consist of the b‐sandwich fold characteristic of the immunoglobulin superfamily, more specifically of the intermediate I‐set of the Ig‐superfamily (for a recent review, see Tskhovrebova and Trinick, 2004). The single‐molecule experiments revealed that application of mechanical force by stretching the molecule initiates unfolding by breaking the inter‐and intra‐sheet bonds (Kellermayer et al., 1997; Rief et al., 1997; Tskhovrebova et al., 1997) and that, importantly, once unfolded, the domains remain unfolded until a low force is reached during release (Kellermayer et al., 1997). Subsequent studies (Carrion‐Vazquez et al., 1999a,b, 2000; Watanabe et al., 2002a) focused on characterizing unfolding in multidomain constructs of Ig domains, using engineered protein fragments and force‐ measuring AFM (Fig. 3A). Multidomain constructs when stretched generate force‐length curves that consist of a wave of peaks that reflect sequential unfolding of domains (see Fig. 3B). Furthermore, it was shown that the unfolding force of Ig domains varies with stretch speed, and ranges from approximately ~150–300 pN for stretch rates from 1–1000 nm/s (Watanabe et al., 2002a). These results led to the realization that Ig domains unfold with a probability that increases with increasing force and passing time. The unfolding forces measured on single molecules in vitro are high compared to physiological forces encountered in muscle (estimated at ~0–10 pN/titin molecule (Watanabe et al., 2002b). Unfolding occurance in vivo was experimentally tested by stretching and holding relaxed soleus muscle fibers for extended durations (up to 64 hours) and by performing immunoelectron microscopy (IEM), mechanics, and Monte Carlo simulations that take into account unfolding/refolding kinetics of Ig domains (Trombitas et al., 2003). Tandem Ig segments were found to greatly extend when slack sarcomeres are stretched to intermediate sarcomere lengths (in soleus muscle from ~2.0–2.6 mm) and to attain a nearly constant length in sarcomeres stretched to longer lengths (here PEVK extension dominates, see below). Although the findings do not exclude that a few domains unfold (and contribute to, for example, passive force hysteresis),

100

GRANZIER AND LABEIT

Fig. 3. (A) Schematic of force‐measuring atomic force microscopy. A molecule is attached between the tip of the cantilever and a gold‐coated microscope slide. Extension of the molecule by retraction of the piezoelectric positioner generates force in the molecule that deflects the cantilever. Cantilever deflection is measured from the force generated in the molecule that can be calculated by using the experimentally determined cantilever stiffness. (B) Examples of stretch (light gray) release (black) curves of an 8‐Ig domain containing fragment from the distal tandem Ig segment. The stretch curve displays sawtooth‐like force peaks, indicating that repetitive structural transitions occurred during stretch with each sudden force drop representing the unfolding of an Ig domain. (At the imposed stretch rate of 0.5 mm/sec the mean unfolding force is ~230 pN.) (C) and (D) Examples of force‐extension curves (stretch in light gray and release in black) of PEVK (C) and N2B‐US (D). Note that stretch and release curves largely overlap and that hysteresis is small. Wormlike‐chain fitted to release curve shown in black, with persistence length (P) indicated on figure. (Based on Watanabe et al., 2002a,b.)

evidence indicates that under physiological loading conditions, unfolding is unlikely to be a major source of elasticity (Trombitas et al., 2003). Extension of the tandem Ig segments in short to intermediate long sarcomeres results from unbending of sequences that serially link Ig domains. Recent laser‐tweezer experiments used antibody pairs to demarcate tandem Ig segments within single titin molecules (Leake et al., 2004)

TITIN AND ITS ASSOCIATED PROTEINS

101

and, using a force range and stretch speed that is unlikely to result in unfolding, measured the force‐extension curve of native tandem Ig segments. Results indicate that the persistence length (P) of tandem Ig segments is ~3.5 nm. This is less than the values estimated earlier from mechanical measurements on whole titin molecules 4.6 nm (Tskhovrebova et al., 1997), and much less than values based on image analysis of single molecules (~13.5 nm; Tskhovrebova and Trinick, 2001); 18.9 nm (Kellermayer et al., 2003), or on light scattering of a suspension of full‐length molecules ~15 nm (Higuchi et al., 1993). The reason for the discrepancy of values is unclear but may reflect differences in errors of the various methods that have been used. Considering that the two first mentioned values (3.5 and 4.6 nm) are similar to the length of Ig domains (Holden et al., 1992), they imply that two successive domains have orientations that are nearly independent of each other. Considering the short linker sequence that separates domains (Witt et al., 1998), such a high degree of freedom seems unlikely and the short P values might thus be underestimated. This notion is supported by nuclear magnetic resonance studies of multi‐Ig domain constructs that indicate that interdomain connections are relatively rigid, allowing limited twisting and bending between successive domains, but not orientations that are uncorrelated (Improta et al., 1998). Taking the mean of all P values cited above results in a value of ~10 nm, which suggests that domains that are separated by at least two other domains have independent orientations. Because this seems reasonable, we will assume henceforth a P of tandem Ig segments of ~10 nm. Using this value to calculate tandem Ig extension (according to Eq. [1] above) shows that the tandem Ig segments easily extend under low force. For example, a force of 1 pN (in cardiac muscle, this at the lower end of the physiological force range) extends the segments to ~50% of their contour length.

3. PEVK Extension A major source of extensibility in titin in intermediate to long sarcomere lengths is the PEVK segment (see Fig. 1, left bottom). In early immunoelectron microscopy the extension of the PEVK segment was measured in human soleus muscle fibers stretched to different sarcomere lengths (Trombitas et al., 1998a,b). In highly stretched sarcomeres, the PEVK region was found to extend to a length of ~750 nm, a value that is close to the 826 nm contour length of the PEVK segment, assuming that the PEVK is fully unfolded (maximal residue spacing in an unfolded polypeptide [0.38 nm] times the number of residues in the soleus PEVK [2174]). Furthermore, the extensible behavior of the PEVK segment measured at a wide range of sarcomere lengths can be simulated well by assuming WLC

102

GRANZIER AND LABEIT

behavior without unfolding/refolding transitions (Trombitas et al., 1998b). These observations led to the notion that the PEVK segment behaves as an unfolded polypeptide (random coil) at all sarcomere lengths, consistent with the preponderance of proline residues and charge clusters along the PEVK sequence that are likely to prevent the formation of stable structures (Labeit and Kolmerer, 1995). More recent work has shown that the PEVK segment is not a simple random coil but also contains polyproline type II (PPII) helices (Ma et al., 2001). In the soleus PEVK segment, sequence‐ based estimates suggest that ~10% of the residues are part of PPII motifs (Watanabe et al., 2002b), although locally this percentage might be much higher (~ 50% of residues in a 28‐residue fetal titin fragment used by Ma et al. (Ma et al., 2001). Single‐molecule experiments have been carried out with AFM on constructs that represent the PEVK segment of cardiac titin (N2B isoform), by Li et al. (Li et al., 2001) using an engineered polyprotein (I91‐PEVK)3 and by Watanabe et al. (Watanabe et al., 2002b) using the I27‐PEVK‐I84 fragment. These experiments showed that stretching the PEVK segment from a short length to close to its unfolded contour length (~70 nm for the N2B PEVK) gives rise to a force‐extension curve that does not reveal abrupt force fluctuations (unlike tandem Ig segments), but that is relatively smooth and that can fit well to a single WLC equation (Fig. 3C shows an example). The featureless appearance of the force‐extension curve indicates the absence of major structural transitions and suggests that the mechanical behavior of the PEVK segment is largely dominated by unfolded polypeptide. The persistence length values of the WLC fits obtained in a large number of experiments had a relatively wide range that varied from ~0.3–2.1 nm, and the persistence length histogram was multimodal with peaks at ~1.4 nm, ~0.8 nm, and ~0.45 nm (Watanabe et al., 2002b). We explained this multimodal distribution by assuming that the persistence length peak at ~1.4 nm reflects the persistence length of the single molecule and that the peaks at ~0.8 and ~0.45 reflect the behavior of doublets and triplets, respectively. A broad P distribution was also obtained by Li et al. (Li et al., 2001). Li et al. used the unfolding characteristics of the Ig domains that flanked their PEVK fragment as a ‘‘single molecule finger print,’’ making it possible to determine which results were derived from single molecules. The observed broad P distribution was explained by assuming that the structure of the polyproline helix conformation of the PEVK segment can be trapped in either a type I or type II conformation and that differences in persistence length of the two types underlie the wide persistence length range of the cardiac PEVK region. The mean P that was obtained was 0.91 nm (Li et al., 2002). A subsequent AFM study on

TITIN AND ITS ASSOCIATED PROTEINS

103

a skeletal muscle PEVK construct with flanking Ig domains (in order to obtain during stretch a ‘‘single molecule finger print’’) revealed a P of ~1.4 nm (Labeit et al., 2003). Finally, using laser‐tweezers, Leake and colleagues studied the native PEVK segment of skeletal muscle titin as well as a skeletal muscle PEVK construct (Leake et al., 2004). Values obtained were ~1.4 nm and ~2.2 nm for native titin and the construct, respectively (values obtained from their Fig. 8B, at an ionic strength [IS] similar to that of the above studies: 200 mM). Thus, mechanically, the PEVK region behaves as a WLC with a mean P value of all studies conducted so far (IS 200 mM) of ~1.5 nm. This relatively low mean P value for the PEVK segment indicates that high forces are required to stretch it, and, thus, at low passive force levels (short sarcomeres) PEVK extension is limited (at a force of 1 pN, the extension is 20% of the PEVK contour length, several fold less than obtained for native tandem Ig segments; see above).

4. N2B‐Unique Sequence (N2B‐Us) The cardiac‐specific N2B element consists of three Ig domains and a centrally located ~570 residue unique sequence (N2B‐Us). Immunoelectron microscopy has shown that the N2B‐Us is the third spring element in cardiac titins and that it is the major source of extensibility at the upper range of physiological sarcomere lengths in the heart (Helmes et al., 1999; Linke et al., 1999; Trombitas et al., 2000). We mechanically characterized an Ig‐(N2B‐Us)–Ig fragment, using AFM (Watanabe et al., 2002b). During stretch‐release cycles N2B‐Us displayed little hysteresis and generated force‐extension curves that follow WLC behavior (an example is shown in Fig. 3D). Persistence length histograms revealed a main peak at ~0.65 nm and a second peak at half of this value (Watanabe et al., 2002b). Assuming that the peak at the longest length reflects single molecules and that the peak at half this length reflects doublets results in a P for N2B‐Us of ~0.65 nm. In subsequent work Li et al. (Li et al., 2002) studied a polyprotein composed of a single N2B module flanked on either side by three tandem Ig domains that were used to create a ‘‘mechanical fingerprint’’ during stretch. The force‐extension curve derived from N2B‐Us was smooth and featureless and had mechanical properties of a random coil. The persistence length averaged 0.66 nm, which is in close agreement with the single‐molecule value (0.65 nm) deduced by Watanabe et al. (Watanabe et al., 2002b). Thus, of all three spring elements found in titin, N2B‐Us has the lowest P, and, for example, extends less than 10% of its contour length under a 1 pN force. Although good progress has been made in understanding the molecular basis of titin’s elasticity, many details remain to be established, such as the

104

GRANZIER AND LABEIT

origin of a small degree of hysteresis discernible in some results (Fig. 3C and D and results of Leake et al., 2004), the possible mechanical role of PPII helices in the PEVK segment, and the presence and role of other possible structures in either the N2B‐Us or PEVK (see also below). Because the AFM has a resolution of ~3–5 pN, AFM‐based studies do not exclude the presence of structures at low force with structural transitions that take place hidden in the force noise. Finally, the PEVK constructs that have been studied so far cover ~30% of the full PEVK sequence and could have missed important features. Thus follow‐up work is required using ultra–high‐resolution techniques. The data currently available indicate that titin’s extensible region in the sarcomere may be modeled as serially linked WLCs with different contour and persistence lengths. The tandem Ig segments extend by straightening of sequences that link folded Ig domains, and these segments have a P of ~10 nm. Their contour length (C) in the various isoforms may be calculated from the combined number of Ig domains contained in proximal, distal, and middle tandem Ig segments (N2B, N2BA, fetal cardiac titin, psoas, and soleus: 40, 64, 72, 73, and 93 domains, respectively) and an average domain spacing of 5 nm (Trombitas et al., 1998b). As for the PEVK region, its elasticity is likely to be derived mainly from straightening of random coil sequences and the average persistence length (at physiological ionic strength) is ~1.5 nm. The contour length of the PEVK region in different isoforms may be obtained from the number of amino acid residues (approximate numbers as estimated from sequencing studied for N2B, N2BA, fetal cardiac titin, psoas, and soleus: 188, 800, 1125, 1200, and 2174, respectively) multiplied by the maximal residue spacing in an unfolded polypeptide (0.38 nm). Finally, the available data for the N2B‐Us segment suggest that this extensible element also behaves as a random coil. It has a P of ~0.65 nm and a contour length that can be calculated from its number of amino acids (572 in human) and the maximal residue spacing (0.38 nm) as 217 nm.

5.

Regulating Elasticity

a. Differential Splicing. Differential splicing of the titin message results in isoforms with length variants of tandem Ig and PEVK segments and that either include (cardiac) or exclude (skeletal muscle) the N2B‐Us. The effect of differential splicing on the force–sarcomere length relation can be revealed by calculating the force–sarcomere length relations of known isoforms (for details, see Watanabe et al., 2002b). Results shown in Fig. 4 reveal that expressing length variants of titin’s spring elements is an effective means of varying passive force. It is also interesting to note that the

TITIN AND ITS ASSOCIATED PROTEINS

105

Fig. 4. Predicted force extension curves of a single titin molecule in sarcomeres expressing cardiac titin isoforms or skeletal muscle isoforms.

most compliant cardiac isoform (fetal cardiac titin) has a force–Sarcomere length relation that is similar to one of the stiffest skeletal muscle isoforms (psoas), but that this is achieved in a different way: in skeletal muscle by serially linking only tandem Ig and PEVK segments and in cardiac muscle by also splicing in of the N2B‐Us segment. Thus, if the only function of titin’s extensible region were to be passive force generation, fetal cardiac muscle could suffice with two spring elements only (tandem Ig and PEVK), with lengths similar to those found in psoas muscle. The fact that instead the N2B‐element is spliced in suggests that this element not only performs mechanical functions but may also be required for different reasons (see the signaling section below). In summary, passive properties of striated muscle can be effectively modulated by altering the expression of titin isoforms with length variants of the PEVK and tandem Ig segments and by including or excluding the N2B element. Whether changes occur in differential splicing during heart development or disease has been studied during the past several years. To understand how splicing responds to chronic mechanical challenge of the heart, a canine tachycardia‐induced model of Dilated candiomyopathy has been

106

GRANZIER AND LABEIT

used in which rapid pacing results in chamber dilation and elevated chamber stiffness (Bell et al., 2000). Two weeks of pacing causes an exaggerated transmural titin isoform ratio gradient (Bell et al., 2000), and 4 weeks of pacing results in elevated N2B (stiff) titin expression and down‐ regulation of N2BA (compliant) titin, accompanied by increased titin‐ based passive stiffness (Wu et al., 2002). A recent study of the spontaneously hypertensive rat model (SHR) has shown a reduced expression of N2BA titin in response to pressure overload, consistent with elevated passive stiffness of SHR cardiac myocytes (Warren et al., 2003a). The adjustment in cardiac isoform expression is not restricted to animal models but also occurs in human patients with coronary artery disease (CAD). Interestingly, in CAD, an increase of the more compliant N2BA cardiac titin isoform occurs (Neagoe et al., 2002). The role of titin in patients with heart failure was recently investigated in patients with end‐ stage heart failure caused by nonischemic dilated cardiomyopathy (Nagueh et al., 2004). Results revealed small N2B (stiff) and large N2BA (compliant) cardiac titin isoforms with a mean N2BA:N2B expression ratio that was significantly increased in patients with heart failure. Mechanical measurements on left ventricular (LV) muscle strips dissected from these hearts revealed that passive muscle stiffness was significantly reduced in patients with high N2BA:N2B expression ratios (Fig. 5A). Clinical correlations support the relevance of these changes for LV function. LV function was assessed by invasive hemodynamics and Doppler echocardiography. Positive correlations were found between the N2BA:N2B titin isoform ratio and various parameters of chamber compliance. Thus, in end‐stage failing hearts, the more compliant N2BA isoform comprised a greater percentage of titin, and changes in titin isoform expression in heart failure patients with DCM significantly affects diastolic filling by lowering myocardial stiffness (Nagueh et al., 2004). The N2BA:N2B expression ratio positively correlated with increased exercise tolerance, suggesting that the titin isoform shift might be compensatory in nature and beneficial for the patients. The reduction in passive stiffness seen in heart failure patients is the reverse of the process that occurs during normal fetal and neonatal heart development (Lahmers et al., 2004; Opitz et al., 2004; Warren et al., 2004) where titin expression switches from a predominance of compliant fetal N2BA titin to the more stiff N2B isoform, giving rise to increased passive stiffness (Fig. 5B). The mechanisms that underlie changes in titin splicing during physiological and pathological adaptation remain obscure to date, and, considering the high clinical relevance, warrant future investigations.

TITIN AND ITS ASSOCIATED PROTEINS

107

Fig. 5. Schematic illustration of adjustments in titin expression that decrease titin‐ based passive myocardial stiffness during DCM heart failure (A) and increase titin‐based passive myocardial stiffness during normal neonatal development (B). Titin gels are shown in the inset. They reveal that normal adult myocardium (control in A) expresses more N2B titin than N2BA titin and that N2BA titin in upregulated in myocardium of heart failure (HF) patients. In fetal myocardium (inset of B), N2BA titin also dominates. (Figure based on Nagueh et al., 2004 and Lahmers et al., 2004.)

108

GRANZIER AND LABEIT

b. Calcium. Passive properties of striated muscle can also be adjusted over the short term by modulating the mechanical properties of titin’s extensible region. Earlier work showed that it is likely that the PEVK segment binds calcium with high affinity (Tatsumi et al., 2001), and recent single‐molecule mechanical experiments indicate that calcium lowers the bending rigidity (P) of the PEVK segment (Labeit et al., 2003). Importantly, the effect requires the presence of E‐rich motifs. E‐rich motifs are conserved domains found in the PEVK sequence, and ~30%–50% of their primary structure is composed of glutamic acid. The mechanism by which they respond to calcium is unknown. Perhaps in the absence of calcium the E‐rich motif and flanking sequences form an ordered structure that on calcium binding becomes disordered, giving rise to a more flexible chain with a lower PL. Alternatively, the E‐rich exon may be disordered in the absence of calcium and the negatively charged glutamates may give rise to a relatively stiff chain as a result of electrostatic repulsion between charges. Calcium binding to glutamates may then reduce the ‘‘electrostatic stiffening effect’’ and increase the flexibility of the chain (i.e., reduce the PL). Several lines of evidence support the notion that skeletal muscle titin increases its stiffness on Ca2þ binding. Increase in passive stiffness occurs in activated intact frog muscle fibers, well ahead of active tension development (Bagni et al., 1994, 2002), whereas stretch during tetani of intact cat muscle results in passive force enhancement (Herzog et al., 2003). Evidence was also obtained in skinned mouse soleus muscle fibers (which express nine E‐rich motifs) from which thin filaments had been extracted and that when stretched showed a calcium‐dependent passive force response (Labeit et al., 2003). As for cardiac muscle, a recent study found no calcium effect in N2B‐expressing myocardium, whereas in N2BA‐expressing myocardium passive stiffness was significantly increased by calcium (Fujita et al., 2004). Considering that the PEVK segment of N2B cardiac titin is composed solely of PEVK repeats (Greaser et al., 2002) and that the PEVK of N2BA titin contains several E‐rich motifs (Greaser et al., 2002; Labeit and Kolmerer, 1995), the findings support the conclusion that calcium‐sensitive passive stiffness requires the presence of E‐rich motifs. Previous work on N2B titin‐expressing myocardium has revealed that the N2B PEVK region and F‐actin interact (Kulke et al., 2001; Yamasaki et al., 2001), enhancing passive stiffness. Evidence suggests that this interaction can be inhibited by S100, provided that calcium is present (Yamasaki et al., 2002). Whether the PEVK segment of N2BA titin also interacts with actin has not been studied. Considering the different composition of N2B

TITIN AND ITS ASSOCIATED PROTEINS

109

and N2BA PEVKs (with E‐rich motifs in only the latter), differences in actin‐binding properties are to be expected. Indeed, a recombinant PEVK fragment from skeletal muscle titin consisting of PEVK repeats and an E‐rich motif does not bind actin under physiological conditions (Yamasaki et al., 2001) and the ability to interact with actin may be most pronounced in sarcomeres that express high levels of N2B titin. Thus, by splicing in certain PEVK exons and excluding others, unique molecules can be constructed, some of which interact with F‐actin (N2B PEVK) and some of which have mechanical properties that are calcium sensitive (N2BA titin and skeletal muscle titin isoforms).

c. Phosphorylation. The N2B unique sequence is a substrate for protein kinase A (PKA) and PKA‐based titin phosphorylation reduces passive tension of cardiac myocytes (Yamasaki et al., 2002). Passive tension reduction may be explained by assuming that phosphorylation increases the length of the N2B element, possibly due to destabilizing native structures (for details see Yamasaki et al., 2002). In turn, at a given sarcomere length, this would reduce the fractional extension (z/L) of titin’s extensible region and, hence, lower passive force (see Eq. [1]). A study of the effect of phosphorylation on passive tension of myocardium that expresses the two cardiac isoforms at different ratios showed that the passive tension reduction was largest in muscles that express high levels of N2B titin (Fukuda et al., 2004). This indicates that the effect of phosphorylation on passive tension generated by N2BA titin is less than that of N2B titin. This is consistent with the notion that the passive tension reduction is due to a reduction in the fractional extension (end‐to‐end length divided by the contour length) of titin’s extensible region. Because the N2BA isoform has a much longer PEVK segment and contains an additional tandem Ig segment, its contour length is longer than that of N2B titin, diminishing the effect of the phosphorylation‐induced contour length gain of the N2B unique sequence on the fractional extension of titin’s ‘extensible region. In summary, titin’s elasticity can be adjusted by altering differential splicing, a process that occurs during normal heart development and during heart disease. In addition, more rapid adjustment mechanisms exist. A cardiac‐specific mechanism is based on phosphorylation and has the largest effect on the force generated by the N2B cardiac isoform. Calcium can regulate titin–actin interactions in the N2B‐expressing myocardium. Furthermore, in N2BA cardiac titin and in skeletal muscle isoforms, calcium increases passive stiffness, an effect that requires the presence of E‐rich motifs in the PEVK segment. Clearly, the passive properties of muscle are highly adaptable.

110

GRANZIER AND LABEIT

IV.

The Titin Filament System in the Sarcomere and Its Ligands

Despite rapid progress in understanding titin at the molecular level, many questions remain about titin’s behavior in the sarcomere, including how exactly titin is associated with the thick and thin filaments. Monitoring different titin epitopes by immunoelectron microscopy has suggested that in the A‐band, titin is a component of the thick filament. Whether titin is inside the thick filament backbone, or on its surface remains to be established, although the accessibility of titin for antibodies favors the latter possibility. It also remains to be established whether titin follows the helical crossbridge path or runs straight along the thick filament. As for stoichiometry, Trinick and colleagues performed mass measurements of end‐filaments using scanning transmission electron microscopy and provided evidence for six titin molecules per half‐thick filaments (Liversage et al., 2001), a value consistent with earlier estimates based on quantitative gel electrophoresis (Cazorla et al., 2000). Because the unit cell structure of the half sarcomere contains two thin filaments per thick filament, the number of titin molecules per thin filament is likely to be 3. Considering the twofold symmetry of the thin filament this value is unexpected (2 of 4 would be expected). Thus the number of titin molecules is insufficient to satisfy symmetries of both thick (threefold) and thin filaments (twofold). A solution to this paradox has been suggested that is based on interaction between titin molecules and thin filaments that enter from the adjacent sarcomere (Knupp et al., 2002; see Liversage et al., 2001; Squire et al., 2005). Expression of novex-3 titin that integrates in the Z‐disk but that does not extend to the A‐band region could be an alternative mechanism for matching the number of titin molecules to the threefold symmetry of the thick filament and the twofold symmetry of the thin filament (Bang et al., 2001). Further research is required to fully understand how titin is integrated in the sarcomere.

A. Titin‐Ligands To better understand how titin integrates into the sarcomere, systematic surveys for interaction between titin and other myofibrillar proteins were initially carried out using dotblot assays with purified muscle proteins and titin fragments. This showed that titin interacts with myosin and MyBP‐C (Houmeida et al., 1995; Labeit et al., 1992). With the introduction of the yeast two‐hybrid technology, an unbiased approach for the survey of protein and protein interactions has become available. The yeast two‐hybrid method is based on a genetic selection for clones interacting with a given

TITIN AND ITS ASSOCIATED PROTEINS

111

‘‘bait’’ cDNA expressed in yeast. This method can identify low abundance and previously unknown proteins as potential novel factors interacting with the protein of interest. In the past several years the yeast two‐hybrid approach has identified about a dozen novel titin‐ligands, and binding partners of theses ligands, with possible roles in, for example, protein turnover (P94/Calpain-3; Muscle specific ring finger protein), regulating potassium channel function (TCap/telethonin and MinK), positioning the sarcoplasmic reticulum (sANK-1 and obscurin), compartmentalization of metabolic enzymes (DRAL/FLH-2), stretch‐sensing and controlling gene expression (muscle LIM protein [MLP], cardiac ankyrin repeat protein [CARP], and ankyrin‐repeat domain 2 [Ankrd2]).

Fig. 6. Layout of titin in cardiac sarcomere with titin‐binding ligands located in the Z‐disk, I‐band, and M‐line. The N‐terminal region of titin interacts with T‐cap and MLP, forming part of a putative MLP‐dependent stretch‐sensing complex, as well as s‐ANK1, which may have a structural role. Titin’s N2B element (found in both N2B and N2BA isoforms) interacts with DRAL/FLH–2, a protein involved in recruiting multiple metabolic enzymes to the I‐band. Titin’s N2A element (only present in N2BA and N2A titins) interacts with CARP, a potential regulator of gene expression during stretch‐induced myocyte hypertrophy, and the calpain protease p94(calpain–3). CARP also interacts with the muscle‐specific nuclear and sarcomeric protein, myopalladin. The M‐line region of titin contains a kinase domain of unknown function, and a MURF‐interacting region. Additionally titin’s M‐line region interacts with myomesin, a structural protein, and P94. (Note that the figure is not to scale.)

112

GRANZIER AND LABEIT

Table I Functional Features and Recently Identified Mutations in Human Titin Exons Exon 1 2–4

6 8–14

13 18 45, 46 48

49

51–224

104/105

107–109 111–224 129

Features Untranslated. Code for the Ig domains Z1þZ2, interacting with T‐cap (telethonin; Gregorio et al., 1998). T‐cap in turn interacts in vitro with the stretch‐dependent IsK ion channel (Furukawa et al., 2001), MLP (Knoll et al., 2002), and myostatin (Nicholas et al., 2002). The titin Z1‐Z2/T‐cap complex may function as a versatile adaptor complex that couple titin’s N‐terminus to myofibrillar signaling. Localization: edge of the Z‐line. Unique sequence that codes for potential serine/proline residues that can be phosphorylated (Gautel et al., 1996). Code for seven copies of 45‐residue Z‐repeats (Gautel et al., 1996). Z‐repeats interact in vitro with alpha‐actinin (Sorimachi et al., 1997). Ultrastructurally, the titin Z‐repeats are inside the Z‐line lattice (Gregorio et al., 1998). Z‐repeat copy #6. Heart‐specific exon (Sorimachi et al., 1997). Codes for Ig repeat Z3. A point mutation in Z3 causes dilated cardiomyopathy (Gerull et al., 2002). Novex 1 and II exons. Excluded in full‐length titin isoforms (Bang et al., 2001). Novex‐III exon includes the Ig domains I–18 to I–23. In vitro, this exon interacts with the giant muscle protein obscurin (Bang et al., 2001). Novex‐III acts as an internal polyadenylation site in titin. Novex–3 titin isoforms are expressed at low levels in muscle, since this exon is skipped in half‐sarcomere spanning titins (Bang et al., 2001). Cardiac‐specific N2B exon. Codes for the Ig domains I–24, I–25, and I–26. The ~570‐residue unique sequence insertion between I–25 and I–26 acts as a molecular spring element by extending its length upon stretch (Helmes et al., 1999; Linke et al., 1999). Exon 49 may also interact with ab‐crystalline (Bullard et al., 2004). Segment contains tandem Ig repeats and PEVK repeats, both acting as molecular spring elements. This segment is extensively differentially spliced. In the shortest cardiac‐specific N2B titin isoform (linked to inclusion of exon 49), exon 50 is spliced directly to exon 225. In more compliant cardiac and skeletal titin isoforms, exon 50 is spliced to exon 51 (linked to additional inclusion of exons in the segment 52–224). Code for a domain that interacts with the muscle ankyrin repeat proteins CARP, ankrd2, and Diabetes related ankynin repeat protein (Miller et al., 2003). Code for Ig I82/82, acting as binding site for the muscle specific protease calpain–3 (also called p94 or Calpain 3; Sorimachi et al., 1995). Code for a proline, glutamate, valine and lysine‐rich spring region in titin (Labeit and Kolmerer, 1995). Codes for a glutamate‐rich insertion within the PEVK region. This E‐rich motif mechanically alters the titin PEVK spring region in the presence of calcium (Labeit et al., 2003).

TITIN AND ITS ASSOCIATED PROTEINS

113

Table I (continued ) Exon 256–258

326

357

358

362

363

Features Code for a highly immunogenic region of titin, termed ‘‘main immunogenic region (MIR)’’ of titin (corresponds to I–109/I–110). Patients with myasthenia gravis and thymus tumors develop autoantibodies directed to this titin epitope (Gautel et al., 1993). Ultrastructurally, the ‘‘MIR’’ epitope localizes close to the end of the A‐band. A frame shift in this exon causes hereditary dilated cardiomyopathy (Gerull et al., 2002). This exon is the largest exon of the human genome. Codes for the FN3 repeat A167 and the Ig repeat A168. These two domains interact with the RING finger protein MURF–1 (Centner et al., 2001). Codes for a serine/threonine kinase domain (‘‘the titin kinase domain’’ (Labeit et al., 1992). For its structure and the identification of a potential substrate (T‐cap/telethonin), see Mayans et al., 1998. Ultrastructurally, the titin kinase domain is located at the periphery of the M‐line lattice (Obermann et al., 1997). Codes for a unique sequence domain. This sequence (also called Mex5 for M‐line exon 5) is differentially spliced and skipped in some skeletal muscles (Kolmerer et al., 1996). Mex5 provides a binding site for calpain–3 (Sorimachi et al., 1995). Codes for titin’s C‐terminal Ig repeat M10. Ultrastructurally, this domain localizes at the edge of the M‐line lattice (Obermann et al., 1997). A mutation in the M10 domain causes the skeletal muscular dystrophy Tibial muscular dystrophy, a disease relatively common in Finland (Hackman et al., 2002).

The binding sites of titin ligands are not randomly distributed along the titin filament, but instead are arranged in three ‘‘hot spots’’: the Z‐disk, the central I‐band region, and the M‐line region of the molecule (Fig. 6; see also Table I). These ‘‘hot spots’’ may represent complexes of proteins that cooperate to achieve certain functions. Furthermore, there appears to be crosstalk between the ‘‘hot spots.’’ For example, titin‐cap (T‐cap) is present at the periphery of the Z‐disk (Gregorio et al., 1998) but also occurs transiently in the M‐line region (Mayans et al., 1998). Similarly, calpain–3/ p94–binding sites have been detected in the M‐line region of titin and in the I‐band hot spot. In response to stretch, expression of ankrd2 (Kemp et al., 2000) and CARP (Miller et al., 2003) are upregulated and target to the sarcomere. Although the factors that are responsible for recruiting and distributing titin ligands along the molecule remain to be elucidated, an attractive hypothesis is that stretching of titin is involved in initiating these

114

GRANZIER AND LABEIT

processes and that restructuring/redistribution of titin‐based protein complexes may in turn control myofibrillar signaling.

Abbreviations T‐cap s‐ANK1 MLP DRAL/FLH–2

CARP Ankrd2 p94 PEVK N2B‐Us

titin‐cap (binds to titin’s N‐terminus) small ankyrin 1 muscle LIM protein member of the LIM protein family known as DRAL/ FHL–2, functions as an adaptor protein that binds metabolic enzymes cardiac ankyrin repeat protein ankyrin‐repeat domain 2 skeletal muscle‐specific calcium‐dependent cysteine protease, also known as calpain 3 proline (P)‐, glutamate (E)‐, valine (V)‐, and lysine (K)‐rich domain in titin unique sequence that is part of titin’s exon 49 (N2B element)

Acknowledgments This work was supported by a grants from NIH HL61497/62881 (H.G.), and Deutsche Forschungsgemeinschaft La 668/7-2 (S.L.).

References Bagni, M. A., Cecchi, G., and Colombini, B. (2002). A non‐cross‐bridge stiffness in activated frog muscle fibers. Biophys. J. 82, 3118–3127. Bagni, M. A., Cecchi, G., and Colomo, F. (1994). Development of stiffness precedes cross‐bridge attachment during the early tension rise in single frog muscle fibers. J. Physiol. 481(Pt 2), 273–278. Bang, M. L., Centner, T., Fornoff, F., et al. (2001). The complete gene sequence of titin, expression of an unusual approximately 700‐kDa titin isoform, and its interaction with obscurin identify a novel Z‐line to I‐band linking system. Circ. Res. 89, 1065–1072. Bell, S. P., Nyland, L., and Tischler, M. D. (2000). Alterations in the determinants of diastolic suction during pacing tachycardia. Circ. Res. 87, 235–240. Bullard, B., Ferguson, C., and Minajeva, A. (2004). Association of the chaperone alphaB‐ crystallin with titin in heart muscle. J. Biol. Chem. 279, 7917–7924. Bustamante, C., Marko, J. F., and Siggia, E. D. (1994). Entropic elasticity of lambda‐ phage DNA. Science 265, 1599–1600. Carrion‐Vazquez, M., Marszalek, P. E., and Oberhauser, A. F. (1999a). Atomic force microscopy captures length phenotypes in single proteins. Proc. Natl. Acad. Sci. USA 96, 11288–11292.

TITIN AND ITS ASSOCIATED PROTEINS

115

Carrion‐Vazquez, M., Oberhauser, A. F., and Fisher, T. E. (2000). Mechanical design of proteins studied by single‐molecule force spectroscopy and protein engineering. Prog. Biophys. Mol. Biol. 74, 63–91. Carrion‐Vazquez, M., Oberhauser, A. F., and Fowler, S. B. (1999b). Mechanical and chemical unfolding of a single protein: A comparison. Proc. Natl. Acad. Sci. USA 96, 3694–3699. Cazorla, O., Freiburg, A., and Helmes, M. (2000). Differential expression of cardiac titin isoforms and modulation of cellular stiffness. Circ. Res. 86, 59–67. Centner, T., Yano, J., and Kimura, E. (2001). Identification of muscle specific ring finger proteins as potential regulators of the titin kinase domain. J. Mol. Biol. 306, 717–726. Freiburg, A., Trombitas, K., and Hell, W. (2000). Series of exon‐skipping events in the elastic spring region of titin as the structural basis for myofibrillar elastic diversity. Circ. Res. 86, 1114–1121. Fujita, H., Labeit, D., and Gerull, B. (2004). Titin‐isoform dependent adjustment of myocardial tension by calcium. Am. J. Physiol. (Heart) 287(6), H2528–H2534. Fukuda, N., Wu, Y., and Granzier, H. (2004). Phosphorylation of titin modulates passive stiffness of cardiac muscle in a titin isoform‐dependent manner. J. Gen. Physiol. 125(3), 257–271. Furst, D. O., Osborn, M., and Nave, R. (1988). The organization of titin filaments in the half‐sarcomere revealed by monoclonal antibodies in immunoelectron microscopy: A map of ten nonrepetitive epitopes starting at the Z line extends close to the M line. J. Cell. Biol. 106, 1563–1572. Furukawa, T., Ono, Y., and Tsuchiya, H. (2001). Specific interaction of the potassium channel beta‐subunit minK with the sarcomeric protein T‐cap suggests a T‐tubule‐ myofibril linking system. J. Mol. Biol. 313, 775–784. Gautel, M., Goulding, D., and Bullard, B. (1996). The central Z‐disk region of titin is assembled from a novel repeat in variable copy numbers. J. Cell. Sci. 109(Pt 11), 2747–2754. Gautel, M., Lakey, A., and Barlow, D. P. (1993). Titin antibodies in myasthenia gravis: Identification of a major immunogenic region of titin. Neurology 43, 1581–1585. Gerull, B., Gramlich, M., and Atherton, J. (2002). Mutations of TTN, encoding the giant muscle filament titin, cause familial dilated cardiomyopathy. Nat. Genet. 30, 201–204. Granzier, H. L., and Labeit, S. (2004). The giant protein titin: A major player in myocardial mechanics, signaling, and disease. Circ. Res. 94, 284–295. Greaser, M. L., Berri, M., and Warren, C. M. (2002). Species variations in cDNA sequence and exon splicing patterns in the extensible I‐band region of cardiac titin: Relation to passive tension. J. Muscle. Res. Cell Motil. 23, 473–482. Gregorio, C. C., Trombitas, K., and Centner, T. (1998). The NH2 terminus of titin spans the Z‐disc: Its interaction with a novel 19‐kD ligand (T‐cap) is required for sarcomeric integrity. J. Cell. Biol. 143, 1013–1027. Hackman, P., Vihola, A., and Haravuori, H. (2002). Tibial muscular dystrophy is a titinopathy caused by mutations in TTN, the gene encoding the giant skeletal‐ muscle protein titin. Am. J. Hum. Genet. 71, 492–500. Helmes, M., Trombitas, K., and Centner, T. (1999). Mechanically driven contour‐length adjustment in rat cardiac titin’s unique N2B sequence: Titin is an adjustable spring. Circ. Res. 84, 1339–1352. Herzog, W., Schachar, R., and Leonard, T. R. (2003). Characterization of the passive component of force enhancement following active stretching of skeletal muscle. J. Exp. Biol. 206, 3635–3643.

116

GRANZIER AND LABEIT

Higuchi, H., Nakauchi, Y., and Maruyama, K. (1993). Characterization of beta‐ connectin (titin 2) from striated muscle by dynamic light scattering. Biophys. J. 65, 1906–1915. Holden, H. M., Ito, M., and Hartshorne, D. J. (1992). X‐ray structure determination of telokin, the C‐terminal domain of myosin light chain kinase, at 2.8 A resolution. J. Mol. Biol. 227, 840–851. Horowits, R., Kempner, E. S., and Bisher, M. E. (1986). A physiological role for titin and nebulin in skeletal muscle. Nature 323, 160–164. Houmeida, A., Holt, J., and Tskhovrebova, L. (1995). Studies of the interaction between titin and myosin. J. Cell. Biol. 131, 1471–1481. Hu, D. H., Kimura, S., and Maruyama, K. (1986). Sodium dodecyl sulfate gel electrophoresis studies of connectin‐like high molecular weight proteins of various types of vertebrate and invertebrate muscles. J. Biochem. (Tokyo) 99, 1485–1492. Huxley, A. F., and Niedergerke, R. (1954). Structural changes in muscle during contraction; interference microscopy of living muscle fibers. Nature 173, 971–973. Huxley, A. F., and Peachey, L. D. (1961). The maximum length for contraction in vertebrate straiated muscle. J. Physiol. 156, 150–165. Huxley, H., and Hanson, J. (1954). Changes in the cross‐striations of muscle during contraction and stretch and their structural interpretation. Nature 173, 973–976. Improta, S., Krueger, J. K., and Gautel, M. (1998). The assembly of immunoglobulin‐like modules in titin: Implications for muscle elasticity. J. Mol. Biol. 284, 761–777. Itoh, Y., Suzuki, T., and Kimura, S. (1988). Extensible and less‐extensible domains of connectin filaments in stretched vertebrate skeletal muscle sarcomeres as detected by immunofluorescence and immunoelectron microscopy using monoclonal antibodies. J. Biochem. (Tokyo) 104, 504–508. Kellermayer, M. S., Bustamante, C., and Granzier, H. L. (2003). Mechanics and structure of titin oligomers explored with atomic force microscopy. Biochim. Biophys. Acta 1604, 105–114. Kellermayer, M. S., Smith, S. B., and Granzier, H. L. (1997). Folding‐unfolding transitions in single titin molecules characterized with laser tweezers. Science 276, 1112–1116. Kemp, et al. (2000). 100(23), 13716–13721 Knoll, R., Hoshijima, M., and Hoffman, H. M. (2002). The cardiac mechanical stretch sensor machinery involves a Z disc complex that is defective in a subset of human dilated cardiomyopathy. Cell 111, 943–955. Knupp, C., Luther, P. K., and Squire, J. M. (2002). Titin organisation and the 3D architecture of the vertebrate‐striated muscle I‐band. J. Mol. Biol. 322, 731–739. Kolmerer, B., Olivieri, N., and Witt, C. C. (1996). Genomic organization of M line titin and its tissue‐specific expression in two distinct isoforms. J. Mol. Biol. 256, 556–563. Kulke, M., Fujita‐Becker, S., and Rostkova, E. (2001). Interaction between PEVK‐titin and actin filaments: Origin of a viscous force component in cardiac myofibrils. Circ. Res. 89, 874–881. Labeit, S., Barlow, D. P., and Gautel, M. (1990). A regular pattern of two types of 100‐residue motif in the sequence of titin. Nature 345, 273–276. Labeit, S., Gautel, M., and Lakey, A. (1992). Towards a molecular understanding of titin. EMBO J. 11, 1711–1716. Labeit, S., and Kolmerer, B. (1995). Titins: Giant proteins in charge of muscle ultrastructure and elasticity. Science 270, 293–296.

TITIN AND ITS ASSOCIATED PROTEINS

117

Labeit, D., Watanabe, K., and Witt, C. (2003). Calcium dependent molecular spring elements in the giant protein titin. Proc. Natl. Acad. Sci. USA Oct 30. Lahmers, S., Wu, Y., and Call, D. R. (2004). Developmental control of titin isoform expression and passive stiffness in fetal and neonatal myocardium. Circ. Res. 94, 505–513. Lander, E. S., Linton, L. M., and Birren, B. (2001). Initial sequencing and analysis of the human genome. Nature 409, 860–921. Leake, M. C., Wilson, D., and Gautel, M. (2004). The elasticity of single titin molecules using a two‐bead optical tweezers assay. Biophys. J. 87, 1112–1135. Li, H., Linke, W. A., and Oberhauser, A. F. (2002). Reverse engineering of the giant muscle protein titin. Nature 418, 998–1002. Li, H., Oberhauser, A. F., and Redick, S. D. (2001). Multiple conformations of PEVK proteins detected by single‐molecule techniques. Proc. Natl. Acad. Sci. USA 98, 10682–10686. Linke, W. A., Rudy, D. E., and Centner, T. (1999). I‐band titin in cardiac muscle is a three‐element molecular spring and is critical for maintaining thin filament structure. J. Cell. Biol. 146, 631–644. Liversage, A. D., Holmes, D., and Knight, P. J. (2001). Titin and the sarcomere symmetry paradox. J. Mol. Biol. 305, 401–409. Locker, R. H., and Leet, N. G. (1976). Histology of highly‐stretched beef muscle. IV. Evidence for movement of gap filaments through the Z‐line, using the N2‐line and M‐line as markers. J. Ultrastruct. Res. 56, 31–38. Luther, P. K., and Squire, J. M. (2002). Muscle Z‐band ultrastructure: Titin Z‐repeats and Z‐band periodicities do not match. J. Mol. Biol. 319, 1157–1164. Ma, K., Kan, L., and Wang, K. (2001). Polyproline II helix is a key structural motif of the elastic PEVK segment of titin. Biochemistry 40, 3427–3438. Ma, K., and Wang, K. (2003). Malleable conformation of the elastic PEVK segment of titin: Non‐cooperative interconversion of polyproline II helix, beta‐turn and unordered structures. Biochem. J. 374, 687–695. Maruyama, K. (1976). Connectin, an elastic protein from myofibrils. J. Biochem. (Tokyo) 80, 405–407. Maruyama, K., Kimura, S., and Ohashi, K. (1981). Connectin, an elastic protein of muscle. Identification of ‘‘titin’’ with connectin. J. Biochem. (Tokyo) 89, 701–709. Mayans, O., van der Ven, P. F., and Wilm, M. (1998). Structural basis for activation of the titin kinase domain during myofibrillogenesis. Nature 395, 863–869. Miller, M. K., Bang, M. L., and Witt, C. (2003). The muscle ankyrin repeat proteins: CARP, ankrd2/Arpp and DARP as a family of titin filament based stress response molecules. J. Mol. Biol. 333, 951–964. Muller‐Seitz, M., Kaupmann, K., and Labeit, S. (1993). Chromosomal localization of the mouse titin gene and its relation to ‘‘muscular dystrophy with myositis’’ and nebulin genes on chromosome 2. Genomics 18, 559–561. Nagueh, S. F., Shah, G., and Wu, Y. (2004). Altered titin expression, myocardial stiffness, and left ventricular function in patients with dilated cardiomyopathy. Circulation 110, 115–162. Neagoe, C., Kulke, M., and del Monte, F. (2002). Titin isoform switch in ischemic human heart disease. Circulation 106, 1333–1341. Nicholas, G., Thomas, M., and Langley, B. (2002). Titin‐cap associates with, and regulates secretion of, myostatin. J. Cell. Physiol. 193, 120–131.

118

GRANZIER AND LABEIT

Obermann, W. M., Gautel, M., and Weber, K. (1997). Molecular structure of the sarcomeric M band: Mapping of titin and myosin binding domains in myomesin and the identification of a potential regulatory phosphorylation site in myomesin. EMBO J. 16, 211–220. Opitz, C. A., Leake, M. C., and Makarenko, I. (2004). Developmentally regulated switching of titin size alters myofibrillar stiffness in the perinatal heart. Circ. Res. 94, 967–975. Rief, M., Gautel, M., and Oesterhelt, F. (1997). Reversible unfolding of individual titin immunoglobulin domains by AFM. Science 276, 1109–1112. Rossi, E., Faiella, A., Zeviani, M., et al. (1994). Order of six loci at 2q24‐q31 and orientation of the HOXD locus. Genomics 24, 34–40. Sorimachi, H., Freiburg, A., and Kolmerer, B. (1997). Tissue‐specific expression and alpha‐actinin binding properties of the Z‐disc titin: Implications for the nature of vertebrate Z‐discs. J. Mol. Biol. 270, 688–695. Sorimachi, H., Kinbara, K., and Kimura, S. (1995). Muscle‐specific calpain, p94, responsible for limb girdle muscular dystrophy type 2A, associates with connectin through IS2, a p94‐specific sequence. J. Biol. Chem. 270, 31158–31162. Squire, J. M., Al‐Khayat, H. A., Knupp, C., and Luther, P. K. (2005). Moleuclar architecture in muscle contractile assemblies. Adv. Protein Chem. 71, 17–87. Tatsumi, R., Maeda, K., and Hattori, A. (2001). Calcium binding to an elastic portion of connectin/titin filaments. J. Muscle Res. Cell Motil. 22, 149–162. Trinick, J. (1994). Titin and nebulin: Protein rulers in muscle? Trends Biochem. Sci. 19, 405–409. Trombitas, K., Greaser, M., and French, G. (1998a). PEVK extension of human soleus muscle titin revealed by immunolabeling with the anti‐titin antibody 9D10. J. Struct. Biol. 122, 188–196. Trombitas, K., Greaser, M., and Labeit, S. (1998b). Titin extensibility in situ: Entropic elasticity of permanently folded and permanently unfolded molecular segments. J. Cell. Biol. 140, 853–859. Trombitas, K., and Pollack, G. H. (1993). Elastic properties of the titin filament in the Z‐line region of vertebrate striated muscle. J. Muscle Res. Cell Motil. 14, 416–422. Trombitas, K., Redkar, A., and Centner, T. (2000). Extensibility of isoforms of cardiac titin: Variation in contour length of molecular subsegments provides a basis for cellular passive stiffness diversity. Biophys. J. 79, 3226–3234. Trombitas, K., and Tigyi‐Sebes, A. (1974). Direct evidence for connecting C filaments in flight muscle of honey bee. Acta Biochim. Biophys. Acad. Sci. Hung. 9, 243–253. Trombitas, K., Wu, Y., and Labeit, D. (2001). Cardiac titin isoforms are coexpressed in the half‐sarcomere and extend independently. Am. J. Physiol. Heart Circ. Physiol. 281, H1793–H1799. Trombitas, K., Wu, Y., and McNabb, M. (2003). Molecular basis of passive stress relaxation in human soleus fibers: Assessment of role of immunoglobulin domain unfolding. Biophys. J. 85, 3142–3153. Tskhovrebova, L., and Trinick, J. (2001). Flexibility and extensibility in the titin molecule: Analysis of electron microscope data. J. Mol. Biol. 310, 755–771. Tskhovrebova, L., and Trinick, J. (2004). Properties of titin immunoglobulin and fibronectin–3 domains. J. Biol. Chem. 279(45), 46351–46354. Tskhovrebova, L., Trinick, J., and Sleep, J. A. (1997). Elasticity and unfolding of single molecules of the giant muscle protein titin. Nature 387, 308–312. Wang, K. (1985). Sarcomere‐associated cytoskeletal lattices in striated muscle. Review and hypothesis. Cell Muscle Motil. 6, 315–369.

TITIN AND ITS ASSOCIATED PROTEINS

119

Wang, K. (1996). Titin/connectin and nebulin: Giant protein rulers of muscle structure and function. Adv. Biophys. 33, 123–134. Wang, K., McCarter, R., and Wright, J. (1991a). Regulation of skeletal muscle stiffness and elasticity by titin isoforms: A test of the segmental extension model of resting tension. Proc. Natl. Acad. Sci. USA 88, 7101–7105. Wang, K., McClure, J., and Tu, A. (1979). Titin: Major myofibrillar components of striated muscle. Proc. Natl. Acad. Sci. USA 76, 3698–3702. Wang, S. M., Sun, M. C., and Jeng, C. J. (1991b). Location of the C‐terminus of titin at the Z‐line region in the sarcomere. Biochem. Biophys. Res. Commun. 176, 189–193. Warren, C. M., Jordan, M. C., and Roos, K. P. (2003a). Titin isoform expression in normal and hypertensive myocardium. Cardiovasc. Res. 59, 86–94. Warren, C. M., Krzesinski, P. R., and Campbell, K. S. (2004). Titin isoform changes in rat myocardium during development. Mech. Dev. 121, 1301–1312. Warren, C. M., Krzesinski, P. R., and Greaser, M. L. (2003b). Vertical agarose gel electrophoresis and electroblotting of high‐molecular‐weight proteins. Electrophoresis 24, 1695–1702. Watanabe, K., Muhle‐Goll, C., and Kellermayer, M. S. (2002a). Different molecular mechanics displayed by titin’s constitutively and differentially expressed tandem Ig segments. J. Struct. Biol. 137, 248–258. Watanabe, K., Nair, P., and Labeit, D. (2002b). Molecular mechanics of cardiac titin’s PEVK and N2B spring elements. J. Biol. Chem. 277, 11549–11558. Whiting, A., Wardale, J., and Trinick, J. (1989). Does titin regulate the length of muscle thick filaments? J. Mol. Biol. 205, 263–268. Witt, C. C., Olivieri, N., and Centner, T. (1998). A survey of the primary structure and the interspecies conservation of I‐band titin’s elastic elements in vertebrates. J. Struct. Biol. 122, 206–215. Wu, Y., Bell, S. P., and Trombitas, K. (2002). Changes in titin isoform expression in pacing‐induced cardiac failure give rise to increased passive muscle stiffness. Circulation 106, 1384–1389. Yamasaki, R., Berri, M., and Wu, Y. (2001). Titin‐actin interaction in mouse myocardium: Passive tension modulation and its regulation by calcium/S100A1. Biophys. J. 81, 2297–2313. Yamasaki, R., Wu, Y., and McNabb, M. (2002). Protein kinase: A phosphorylates titin’s cardiac‐specific N2B domain and reduces passive tension in rat cardiac myocytes. Circ. Res. 90, 1181–1188. Young, P., Ehler, E., and Gautel, M. (2001). Obscurin, a giant sarcomeric Rho guanine nucleotide exchange factor protein involved in sarcomere assembly. J. Cell. Biol. 154, 123–136. Young, P., and Gautel, M. (2000). The interaction of titin and alpha‐actinin is controlled by a phospholipid‐regulated intramolecular pseudoligand mechanism. EMBO J. 19, 6331–6340.

REGULATION OF MUSCLE CONTRACTION BY TROPOMYOSIN AND TROPONIN: HOW STRUCTURE ILLUMINATES FUNCTION By JERRY H. BROWN AND CAROLYN COHEN Rosenstiel Basic Medical Sciences Research Center, Brandeis University, Waltham, Massachusetts 02454

I. II.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Periodic Features. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Heptad Repeat and Coiled‐Coil Structure (Overview) . . . . . . . . . . . . . . . . . B. Core Residues and Tropomyosin Bending: Generalized Flexibility or Specific Design? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Tropomyosin Surface Residues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Aperiodic Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. The Head‐to‐Tail Overlap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. The Troponin Complex and the TnT–Tropomyosin Interaction . . . . . . . IV. Turning on Troponin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Turning on the Thin Filament. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Binding Sites in Relation to Steric Blocking. . . . . . . . . . . . . . . . . . . . . . . . . . . B. Atomic Structures and Molecular Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Why Does Tropomyosin Move to a Third Position in the Fully‐Activated State? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Additional Recent Results on Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I.

121 124 124 127 130 130 131 134 139 142 144 145 147 148 149 151

Introduction

The thin filaments of vertebrate striated muscle are periodic structures composed of three proteins with strikingly different designs that function together for the regulation of contraction. Actin provides the framework of the thin filament and is probably the oldest of the three in terms of evolution. Actin is a highly conserved, yet versatile, member of a large superfamily (including hexokinase and HSP70, as well as Arp2 and 3), which binds nucleotide. The protein is widespread and binds to many other proteins. Each G‐actin subunit (375 residues) is ~55 A˚ in length and width, and self‐assembles, upon the addition of salt, into a ‘‘double helix’’ to form the F‐actin filament (Pollard, 1999). (The structure and function of actin filaments are reviewed by Squire et al. in this volume and are also described in the ‘‘Additional Recent Results on Regulation’’ section of this article.) ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71004-9

121

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

122

BROWN AND COHEN

Fig. 1. Periodic features of tropomyosin are displayed at various levels. (A) Tropomyosin amino acid sequences (including the chicken striated‐muscle isoform shown [P02559]) display a short‐range seven‐residue long (heptad) motif where the a and d positions are generally apolar and located in the core of the a‐helical coiled‐coil. There is also a long range periodicity where features in roughly 40‐residue long segments (each one divided into an a [shaded] and subsequent b zone [unshaded]) are repeated seven times in the tropomyosin molecule (McLachlan and Stewart, 1976). The repeats of certain surface (b and f position) acidic residues (in black squares) appear most regular

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

123

Tropomyosin is a rather more specialized and deceptively ‘‘simple’’ molecule, designed, as it were, to interact with and match the actin helix. This semi‐periodic, two‐chain protein has been seen as the paradigm of the a‐helical coiled‐coil family of fibrous proteins (Crick, 1953). The molecule in muscle has 284 pairs of residues (Stone et al., 1975) and is ~400 A˚ in length. Tropomyosin assembles into filaments by a short ‘‘head‐ to‐tail’’ (N‐ to C‐) overlap (McLachlan and Stewart, 1975). When bound to the actin helix, one tropomyosin spans seven consecutive (identical) actin monomers, so that each half‐turn of the tropomyosin coiled‐coil makes equivalent (or quasi‐equivalent) interactions with each actin monomer (Fig. 1; McLachlan and Stewart, 1976; Parry, 1975; Phillips et al., 1986). The actin–tropomyosin interaction involves periodic features in tropomyosin’s sequence and structure. Aperiodic features of tropomyosin are related both to the head‐tail overlap and to other functions that vary in different cells. The best understood is its interaction with the irregularly shaped (~265 A˚ long) troponin complex in vertebrate striated muscles. Here, one tropomyosin molecule binds one troponin complex, so that the tropomyosin filament positions troponin regularly along the thin filament producing a periodicity of 385 A˚ . Troponin’s role in the thin filament of vertebrate striated muscles is primarily that of regulation. The three subunits of this complex form what has been described as a Ca2þ‐sensitive ‘‘latch’’ that fixes tropomyosin’s position on the actin helix in the ‘‘off’’ state of contraction (Lehman et al., 2001). One subunit of the complex, troponin T (TnT), maintains an invariant linkage to tropomyosin, and another, troponin I (TnI), a variable linkage to actin. The third subunit, troponin (TnC) is the Ca2þsensor of the complex and indeed of the myofibril itself. The latch is opened or closed depending on the level of Ca2þ. Correspondingly, a series of conformational changes takes place in the entire complex and in the thin

and may be involved in binding consecutive monomers of F‐actin (Phillips et al., 1986). The semi‐regularly repeated clusters of core alanines (in triangles) are implicated in producing the bends in the tropomyosin filament necessary for its winding around the actin filament (Brown et al., 2001). Muscle tropomyosins are generated from nine exons. Each one is depicted between black lines and is roughly 21 or 42 codons in length (Ruiz‐ Opazo and Nadal‐Ginard, 1987). The regularity and lengths of the alanine clusters, as well as the general level of correspondence between the exons and protein period repeats, are less apparent in the C‐terminal third of the molecule; much of this region of tropomyosin also displays diminished thermal stability and binds troponin T (see text). (B) Model of tropomyosin and its periodic features (blue helices contain core alanines, black spheres correspond to the surface acidic residues of the alpha zones) on a simplified depiction of F‐actin. (Drawn by G. Johnson (fivth.com)). Figures adapted from Brown et al., 2001.

124

BROWN AND COHEN

filament. The proteins of the thin filament—of widely disparate design— thus appear to regulate myosin/actin interactions by means of a steric‐ blocking mechanism (Huxley, 1972; Parry and Squire, 1973), whose detailed workings are beginning to be understood at the molecular level. This review begins with a brief description of tropomyosin’s periodic and aperiodic structural features related to their function. This analysis is greatly aided by very recent nuclear magnetic resonance (NMR) and high‐resolution crystal structures, as well as a number of biochemical experiments that were not available at the time of previous reviews (e.g., Perry, 2001). We also include a description of the structure of troponin, relying on major crystallographic advances of the past 2 years (Takeda et al., 2003; Vinogradova et al., 2005). These studies—together with new information on F‐actin—are then related to the molecules’ respective roles in regulating the contraction of vertebrate striated muscle.

II. A.

Periodic Features

Heptad Repeat and Coiled‐Coil Structure (Overview)

Tropomyosin displays two types of periodicity. In addition to a long‐range ~40‐residue repeat (discussed below), tropomyosin is the paradigm of the a‐helical coiled‐coil class of proteins: it shows the corresponding short‐ range, seven‐residue (so‐called heptad) repeat of the form (a‐b‐c‐d‐e‐f‐g) in its sequence (Fig. 1) and structure (Figs. 2, 3). The various designs of a‐ helical coiled‐coils in proteins are described in detail by Lupas in a sister volume in this series (Volume 70), but the current article provides a brief overview. In these proteins, the a and d residues are generally apolar, found for the most part in the interior of the structure, and are in large part responsible for determining many aspects of the coiled‐coil’s conformation. These residues form a left‐handed stripe along the surface of right‐ handed a‐helices. The winding of the helices around one another allows these core residues to interlock in a ‘‘knobs‐into‐holes’’ fashion (Fig. 3) (Crick, 1953), which confers stability on highly charged polypeptide chains in an aqueous environment (for review, see Cohen and Parry, 1990). The b, c, and f positions, which are on the surface (see Fig. 2), are often polar or charged and confer solubility to the molecule. These residues are also in position to make direct contacts with ligands. It was, in fact, the relative ratio of apolar to polar residues in tropomyosin, nearly 2 to 5, that originally supported Crick’s basic packing model for the coiled‐coil (Crick, 1953). Tropomyosin was also the first a‐helical coiled‐coil to be sequenced (Stone et al., 1975), and most isoforms (yeast excepted) have been shown to have

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

125

Fig. 2. Heptad repeat unit of the a‐helical coiled‐coil. The schematic projection of a short segment of a two‐stranded a‐helical coiled‐coil viewed down its axis shows the locations of the different heptad positions relative to one another and relative to the core interface between the helices. (Adapted from Figure 1a of Stewart, 2001.)

an unbroken series of forty continuous heptads (see Fig. 1), now known to be unusual in fibrous proteins. Analysis of this sequence also suggested that additional stabilization of the coiled‐coil was provided by salt links between e and g residues on neighboring chains, which flank the core, favoring a parallel in‐register arrangement for two‐chain molecules (McLachlan and Stewart, 1976; Parry, 1975). This basic design of a parallel dimeric a‐helical coiled‐coil was confirmed by the crystal structure of the GCN4 ‘‘leucine zipper’’—a relatively short a‐helical coiled‐coil (O’Shea et al., 1991). This atomic resolution structure revealed a straight coiled‐coil in which the two a‐helices are in axial register (within approximately 0.3 A˚ ) and relatively close packed (~10 A˚ diameter). The side chains are arranged in alternating ‘‘layers’’ with distinctive orientations in the core a and d positions. The core side chains, primarily medium‐sized branched aliphatics, have the right size and shape as ‘‘knobs’’ to fit snugly into ‘‘holes’’ formed by four side chains of the neighboring helix (see Fig. 3 for similar packing observed in segments of tropomyosin). The simplicity of the linearly periodic coiled‐coil (in

126

BROWN AND COHEN

Fig. 3. How to bend a coiled‐coil. Local bending of the tropomyosin molecule, necessary for its winding around actin, may be described as having components that can occur about the (A) narrow edge or (B) broad face of the a‐helical dimer (McLachlan and Stewart, 1976; Stewart, 2001). The latter type of bending has been observed in a crystal structure of an N‐terminal fragment of tropomyosin. This bend occurs away from a ‘‘wedge’’ (triangular shape) formed at the junction of (C) two types of coiled‐coil segments, one with in‐register (red) and the other with ~1.2 A˚ axially staggered (blue) a‐helices (Brown et al., 2001). This difference is promoted by the (D) different ‘‘holes’’ into which core leucines and alanines preferentially fit (see also Gernert et al., 1995 and Walther et al., 1996).

contrast to three‐dimensional globular folds) has also permitted the visualization of conformational effects of different sequences or residues, especially those a and d residues in the core, on most of the structural parameters of coiled‐coils, including pitch, parallel versus antiparallel nature, and multimeric state (Harbury et al., 1993; Lupas, 1996; Woolfson and Alber, 1995), as well as the detailed axial register between the a‐helices (Gernert et al., 1995; Brown et al., 2001; Day and Alber, 2000, and see below). Recent calorimetric studies have shown the influence of specific core a and d residues (e.g., Kwok and Hodges, 2004; Lu and Hodges, 2004; Singh

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

127

and Hitchcock‐DeGregori, 2004; Tripet et al., 2000), as well as the flanking (interhelical salt‐bridge–forming) e and g residues (e.g., Burkhard et al., 2002; Meier et al., 2002) on the stability of the coiled‐coil. So‐called trigger sequences, about two‐heptad long motifs that promote a‐helical coiled‐coil formation, have also been identified in many coiled‐coil proteins, including tropomyosin; the key structural feature here is a particular positional relationship in the primary sequence between salt bridges on the flanking e and g positions and hydrophobic residues in the core (Kammerer et al., 1998; Steinmetz et al., 1998; Wu et al., 2000). The proper positioning of such a salt bridge along the sequence apparently provides key stabilization by shielding apolar core residues from solvent; their exposure to solvent in a segment of scallop myosin subfragment 2 lacking interhelical salt bridges helps explain this coiled‐coil’s instability and function (Li et al., 2003). The a‐helical coiled‐coil conformation illustrates special features of a‐helical packing but also exemplifies general principles of protein design. Isolated a‐helices, unless they have special sequences, are marginally stable in aqueous solution. Although a‐helical regions in globular proteins may not have a systematic pattern of bonding as in a coiled‐coil, nevertheless they display both side‐chain and backbone interactions that stabilize this conformation (see ‘‘Core Residues and Tropomyosin’’ section). ‘‘Single‐chain’’ regions in proteins that lack such stabilization are key points of liability in the structure and often function as hinges or have other dynamic roles (Cohen, 1966; Cohen and Holmes, 1963; see ‘‘Structure of Troponin’’ section). The seven distinct amino acid positions and associated interactions that are produced from the a‐helical coiled‐coil provide the basic structural unit of tropomyosin. These elements superimpose on the longer, roughly 40‐residue, functional unit of tropomyosin (see Fig. 1), and patterns of residues found both in the core of the coiled‐coil as well as on its surface are repeated seven times in a full‐length tropomyosin molecule and play a role in the periodic binding of tropomyosin to actin.

B. Core Residues and Tropomyosin Bending: Generalized Flexibility or Specific Design? Repeating surface residues of tropomyosin appear to interact directly with recognition sites on consecutive monomers of F‐actin. We first discuss, however, the repeat in the core, which, although less regular than that of certain surface residues (see below), has a clearly understood effect on the conformational and dynamic properties of the coiled‐coil.

128

BROWN AND COHEN

In contrast to other a‐fibrous proteins, alanine is the most common residue in the core d position of vertebrate tropomyosins (Conway and Parry, 1990). Inspection of these sequences shows that these small residues (triangles in Fig. 1) are clustered into seven groups (of one to three d‐position alanines each) that are separated from one another by segments rich in d‐position leucines and other medium or large‐sized side chains (Brown et al., 2001; McLachlan and Stewart, 1976). This sequence repeat was connected early on with a presence of periodic flexible regions in tropomyosin (McLachlan and Stewart, 1976), and mutational studies indicate a decreased stability for coiled‐coils containing alanines rather than canonical leucines or valines in the core (Singh and Hitchcock‐DeGregori, 2003, 2004; Tripet et al., 2000). Flexibility appears to be necessary for a number of functions of tropomyosin, including its capacity to wind around actin (see Phillips and Chacko, 1996; Phillips et al., 1986) and to occupy different positions on its surface (see below). Stabilization of tropomyosin with extra core leucines does indeed reduce actin affinity (Singh and Hitchcock‐DeGregori, 2003, 2004). It had long been assumed that limited interhelical contacts resulting from the small alanine side chains would lead to a generalized local instability and hence flexibility of the coiled‐coil. Correspondingly, it has been suggested that tropomyosin, when bound to actin, bends either about the two‐stranded coiled‐coil’s narrow edge or broad face (see Fig. 3A and B and McLachlan and Stewart, 1976). Analyses of the recent crystal structures of a‐tropomyosin’s N‐terminal 80 residues (Brown et al., 2001) and of other coiled‐coils, together with mutational experiments (see above and Kwok and Hodges, 2003; Lu and Hodges, 2004) suggest, however, a more specific design—in which the arrangement, not simply the number, of core alanines affects the coiled‐coil stability, and in which the alanine ‘‘clusters’’ actually promote, rather than just tolerate, a specific bending of the coiled‐coil about its broad face. The crystal structure of the N‐terminal 80 residues of tropomyosin (Brown et al., 2001) contains its first alanine cluster and displays two specific consequences of this motif for the main‐chain geometry of the coiled‐coil. One is that the coiled‐coil in this segment becomes locally narrow, to ~8.0 A˚ diameter, as would be expected from alanine’s small size. This feature is directly related to the stability of the coiled‐coil: in a 10 A˚ wide dimeric coiled‐coil, a pair of core alanines from the opposite helices would generally leave unfilled spaces in the interior; these spaces become smaller as the main chains of the helices approach each other and the core becomes more close‐packed. Recent studies of model coiled‐coils with identical amino acid compositions, but different arrangements,

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

129

show that clustering of same‐sized core residues, similar to those in the N‐terminal fragment of tropomyosin just described, produces a more stable coiled‐coil than one in which consecutive core positions alternate between leucines and alanines (Lu and Hodges, 2004; Kwok and Hodges, 2003). Although the atomic structures of these designed peptides are not available, inspection of the cortexillin crystal structure (Burkhard et al., 2000, PDB i.d. 1d7m) shows that its coiled‐coil narrows only to 9.0 A˚ diameter at the location of a single d‐position alanine (residue 335); the a‐helices apparently are propped partially apart here by the immediately adjacent larger (leucine and lysine) core residues found on either side. In the case of tropomyosin, the clustering of similarly sized core side chains over a number of helical turns, which appears to permit relatively extensive close packing for the small as well as large side chains, may reduce instability caused by its unusually high content of core alanines. In fact, the temperature factors of the alanine cluster present in the crystal structure of the N‐terminal fragment of tropomyosin (Brown et al., 2001) show no significant differences from those in the neighboring (leucine‐rich) segments. The second consequence of core alanines on the main‐chain geometry of the N‐terminal region of tropomyosin is that they cause a local, symmetry‐breaking, axial stagger of about 1.2 A˚ between the helices (Brown et al., 2001) (see Fig. 3C). The stagger occurs because an alanine (unlike a leucine) is generally too small to make contacts with all four side chains forming the ‘‘hole’’ of the partner helix; the two main chains are axially staggered so that the alanine is positioned instead into the center of a three‐residue hole (see Fig. 3D). This difference in bonding preferences of alanine and leucine side chains occurs in parallel and heteromeric antiparallel coiled‐coils (Brown et al., 2001; Gernert et al., 1995) and occurs in globular proteins as well (Walther et al., 1996). Here again the coiled‐coil illustrates a general principle in protein design. The functional importance of axial staggering in tropomyosin arises from the joining of this segment with an adjacent in‐register segment rich with canonical core d‐position leucines. This arrangement may be pictured as creating a wedgelike ‘‘insertion,’’ of almost one residue, in one helix relative to the other; this insertion stabilizes a specific local bend in the coiled‐coil axis (on average about 6 ) away from the wedge about the broad face of the coiled‐coil (see Fig. 3B). Such a design at the junctions of core d‐position alanine and leucine clusters has a simple chemical/mechanical basis and is found in diverse proteins (Brown et al., 2001) and Brown, in preparation). Local helical staggering and resulting ‘‘wedge bending’’ (previously termed ‘‘alanine bending’’; Brown et al., 2001) therefore appear to be intrinsic

130

BROWN AND COHEN

properties of core alanine‐containing coiled‐coils,* suggesting that these features will also occur at other alanine clusters of tropomyosin that have not yet been visualized at atomic resolution. In this manner, core alanines can by specific design promote the winding conformation necessary for tropomyosin’s interaction with F‐actin. Some recent findings, however, have prompted an alternative interpretation of the alanine clusters. Here mutant tropomyosins were expressed in which most of an alanine cluster was replaced with a similarly destabilizing ‘‘amide cluster’’ (glutamines and asparagines) (Singh and Hitchcock‐ DeGregori, 2003, 2004). Since these mutants appeared to have the same stability and actin affinity as native tropomyosins, it was concluded that the effect of the alanine clusters was simply general coiled‐coil flexibility produced by destabilizing residues in its core. Note, however, that the APC coiled‐coil, which also has an amide‐rich core, displays significant (0.8 A˚ ) axial staggering between its a‐helices (Day and Alber, 2000). Helical staggering may thus also occur in the amide clusters of the mutant tropomyosins cited above, and resulting wedge bending about the coiled‐coil’s broad face could also account for their actin‐binding properties. Taken together, the results suggest that specific design features in the core, in addition to coiled‐coil destabilization, are important for tropomyosin function.

C. Tropomyosin Surface Residues Analyses of the tropomyosin sequence have shown long‐range periodicities of certain surface acidic and apolar residues that are likely to be recognition sites for actin. These features are discussed below in relation to their role in regulation (see ‘‘Turning on the Thin Filament’’ section).

III.

Aperiodic Features

In addition to the semi‐regular sites on the surface and core of tropomyosin, related to the binding to actin, there are other specialized binding regions on the molecule. One such feature is described below, and another is discussed in the troponin section. * Note that two aspects of this design have been misunderstood in the recent literature: it is the axial staggering and not the local narrowing that directly leads to this wedge‐bending design. Also, the bends in the crystal structure do not occur within the (alanine‐associated) axially staggered segments, but rather at their ends.

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

A.

131

The Head‐to‐Tail Overlap

A distinctive feature of tropomyosin is its strong tendency to self‐ assemble into filaments by means of a specific head‐to‐tail overlap. The length of this overlap between the N‐ and C‐termini was originally deduced to be about 8–11 residues so that the periodicity of consecutive tropomyosin molecules remains in phase as the filament winds around the actin helix (McLachlan and Stewart, 1975). This overlap has so far been observed only as extra density in low‐resolution (7 A˚ ) maps of native tropomyosin filaments (Whitby and Phillips, 2000). But high‐resolution crystal structures (Brown et al., 2001; Li et al., 2002) and NMR results (Greenfield et al., 1998, 2003) have been determined for individual striated‐muscle tropomyosin N‐ and C‐terminal fragments (Fig. 4). These fragments display structural features that allow some inferences to be made about the conformation of this critical linkage. In native tropomyosin from muscle, the main chain at the N‐terminus is acetylated, and this region is fully helical, according to the NMR results of a 14‐residue N‐terminal fragment stabilized by a short leucine zipper (LZ) (Greenfield et al., 1998) (see Fig. 4). The crystal structure of an unacetylated, bacterially expressed, N‐terminal fragment of tropomyosin, however, reveals an extended conformation for residues 1 and 2 (Brown et al., 2001) (see Fig. 4 and legend). Since the head‐tail overlap is strengthened by acetylation (Cho and Hitchcock‐DeGregori, 1991), these structures indicate that the overlap requires a stable coiled‐coil in this region. By contrast, the crystal and NMR structures of C‐terminal fragments (Li et al., 2002; Greenfield et al., 2003) show that the last 15–21 residues adopt non–coiled‐coil (see Fig. 4), hence more flexible, structures (see Fig. 4D legend). Replacing d‐position glutamine 263 with leucine indeed greatly increases the stability of a C‐terminal peptide and somewhat reduces its ability to form a head‐tail overlap complex with an N‐terminal peptide (Greenfield et al., 2002). Taken together, the structures of the N‐ and C‐terminal fragments of striated‐muscle tropomyosin show significant differences in the paths of their chains, and the terminal regions appear to require different levels of a‐helical coiled‐coil stability for formation of the head‐tail overlap (Paulucci et al., 2004). In addition to the different designs of the two termini, differences in charge may be significant to the formation of the head‐tail overlap. Polymerization of tropomyosin is ionic‐strength dependent (Kay and Bailey, 1960; Sousa and Farah, 2002), and a model has been suggested that involves interaction between positive charges in the N‐terminal region and negative charges in the C‐terminal region (McLachlan and Stewart, 1975). In support of this model, increasing the acidity of the C‐terminus,

132

BROWN AND COHEN

Fig. 4. The structures of N‐ and C‐terminal fragments of striated‐muscle alpha‐ tropomyosin show different designs. (A) The N‐terminus of native (N‐acetylated) tropomyosin is a fully a‐helical coiled‐coil (based on the NMR results of a synthesized 14‐residue long fragment [Greenfield et al., 1998]). (B) When not acetylated, the N‐terminal two residues adopt a nonhelical (extended) conformation (based on the crystal structure of a bacterially expressed 80‐residue long N‐terminal fragment [Brown et al., 2001]; only 14 pairs of residues shown). Unblocked positively charged N termini at a positions have been predicted to locally destabilize the coiled‐coil (Monteiro et al., 1994). (C) and (D) A relatively long region (15–21 residues) at striated‐muscle tropomyosin’s C‐terminus adopts a non–coiled‐coil conformation, with either (C) the a‐helices lying parallel to one another (in the NMR structure of a 34‐residue long C‐terminal fragment [Greenfield et al., 2003]) or (D) the a‐helices splaying out from one another and being stabilized by symmetry contacts (in the crystal structure of a 31‐residue long C‐terminal fragment [Li et al., 2002]). The two C‐terminal peptide structures are viewed in panels C and D from different vantage points. The non–coiled‐ coil and poorly ordered nature of these C‐terminal residues are likely to be intrinsic features of the isolated (striated‐muscle) tropomyosin molecule. The C‐termini of native tropomyosin, unlike the N‐termini, are not blocked, and repulsion between these d‐position (see Fig. 1A) negatively charged main‐chain carboxylates could destabilize a coiled‐coil structure. Also, the large or polar core side chains (e.g., Gln 263 and Tyr 267) appear to destabilize the striated‐muscle tropomyosin coiled‐coil just prior to the expected overlap region (Li et al., 2002; Greenfield et al., 2003). Note that the peptides studied here have been stabilized either by an attached leucine zipper (Greenfield et al., 1998; Li et al., 2002), disulfide bond (Brown et al., 2001), or mutation (Greenfield et al., 2003).

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

133

either by phosphorylation of serine 283 (Heeley et al., 1989), which is the molecule’s only phosphorylation site (Mak et al., 1978), or by mutation of this residue to glutamate (Sano et al., 2000), strengthens the head‐tail overlap. In this connection, electrostatics is also thought to play a key role in an unusually strong artificial head‐to‐‘‘tail’’ overlap produced in a striated‐muscle tropomyosin by deleting the native C‐terminal 24 residues: the C‐terminal region of this truncated tropomyosin has a greater density of negatively charged residues than the C‐terminal region of native tropomyosin (Paulucci et al., 2004). In addition to ionic interactions, apolar contacts between the termini have also been used to model the head‐tail overlap (McLachlan and Stewart, 1975; Palm et al., 2001). Some of the results cited above (Greenfield et al., 2002; Li et al., 2002) indicate that, in striated‐muscle tropomyosin, the near C‐terminal region (residues 263–275) also appears to play an important role in the formation of the overlap, and this region may also be important in smooth‐muscle tropomyosin isoforms. Here, the head‐tail overlap is stronger in the a/b heterodimer than in either homodimer ( Jancso and Graceffa, 1991), see also (Sanders and Smillie, 1984), and these two otherwise highly similar smooth‐muscle tropomyosin gene products differ at positions 268, 269, 270, and 273. As described in a later section, in striated‐muscle tropomyosin this region is also important for binding TnT. Several tentative models of the overlap have been proposed. The structure may involve a parallel four‐helix bundle—perhaps where the N‐termini form a tightly packed coiled‐coil to which more flexible C‐terminal a‐helices bind (see also Palm et al., 2001). Note that an early model (McLachlan and Stewart, 1975) in which both termini were assumed to be simple two‐stranded a‐helical coiled‐coils, is inconsistent with the available atomic structures of the individual fragments. Whatever its structure may be, the head‐tail overlap promotes formation of extended tropomyosin filaments, whose binding to actin, albeit weak, is greater than that of the monomer (Wegner, 1979). Indeed, removal of parts or all of the native N‐ or C‐terminal overlap region results in lower polymerization and actin affinity (Heald and Hitchcock‐DeGregori, 1988; Maytum et al., 2000, 2001). The increased F‐actin affinity of polymerized over monomeric tropomyosins is but one example of a common principle in macromolecular structure: the binding strength of a molecular species is increased due to the linking together of individual, otherwise weak, binding sites. The tropomyosin filament is relatively weakly bound, but it appears to stabilize (stiffen) the F‐actin helix by a factor of 50% (Kojima et al., 1994). Additional aspects of the functional role of tropomyosin, such as cooperativity, are discussed in the ‘‘Turning On the Thin Filament’’ section.

134

BROWN AND COHEN

B. The Troponin Complex and the TnT–Tropomyosin Interaction 1.

Structure of Troponin

Contraction in skeletal and cardiac muscles is triggered by the binding of Ca2þ to the troponin C (TnC) subunit (Fig. 5) of the troponin complex. The other two subunits are troponin I (TnI), which binds to F‐actin and inhibits actomyosin ATPase (for review, see Perry, 1999), and troponin T (TnT), which binds to tropomyosin, linking it to the rest of the troponin

Fig. 5. Ribbon diagram of TnC (from Herzberg and James, 1985). The structural C‐terminal lobe (consisting of alpha helices E, F, G, and H in domains III and IV) has higher affinity for Ca2þ ions than does the regulatory N‐terminal lobe (consisting of alpha helices A, B, C, and D in domains I and II). The D/E linker between the two lobes is often flexible, but when ordered it adopts an a‐helical conformation with three turns fully exposed to solvent (see text).

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

135

complex (for review, see Perry, 1998, and below). The highly asymmetric ~265 A˚ long structure of troponin consists of a globular ‘‘head’’ containing TnC, TnI, and the C‐terminal part of TnT (TnT2) and a long ‘‘tail’’ containing primarily the N‐terminal part of TnT (TnT1) that extends away from the head region (Fig. 6A) (Flicker et al., 1982). The troponin complex and its functional role were discovered by Ebashi and co‐workers more than three decades ago (Ebashi and Endo, 1968; Ebashi et al., 1969), but only recently has the structure of most of the head region in both cardiac and skeletal muscle been revealed by X‐ray crystallography (Takeda et al., 2003; Vinogradova et al., 2005). Some key structural features, found in both skeletal and cardiac troponin, were certainly expected from previous structural and biochemical studies. The TnC subunit is a well‐characterized member of the calmodulin (CaM) superfamily whose atomic structure (see Fig. 5) has been determined in both muscle types (Herzberg and James, 1988; Houdusse et al., 1997; Li et al., 2000; Satyshur et al., 1988; Sia et al., 1997). The molecule consists of two lobes connected by a linker: a regulatory N‐terminal lobe that specifically binds Ca2þ ions and a structural C‐terminal lobe that binds Ca2þ and Mg2þ ions (Potter and Gergely, 1975). Each lobe consists of a pair of helix-loop-helix (EF‐hand) domains, but the regulatory N‐terminal lobe of cardiac TnC contains only one functional domain. The binding of Ca2þ ions has been shown to open a hydrophobic cleft in its regulatory N‐terminal lobe (Gagne et al., 1995; Houdusse et al., 1997; Strynadka et al., 1997). In cardiac TnC, Ca2þ alone is not sufficient to stabilize the open conformation, so that a target peptide or a drug molecule is necessary (Li et al., 1999, 2000). The orientation of this lobe with respect to the structural C‐terminal lobe varies, with the central linker that connects them adopting a partially disordered or fully a‐helical conformation. This linker was expected to be flexible from biochemical studies of TnC (Grabarek et al., 1981; Slupsky and Sykes, 1995) as well as work on calmodulin (Walsh et al., 1977; Zhang et al., 1995). When not disordered, the three‐turn fully solvent‐exposed a‐helical linker had been considered unusual since a‐helices in proteins are generally stabilized by interactions with other chains, as exemplified by the a‐helical coiled‐coil. The other two a‐helices connected to the linker come from the EF hands of the N‐ and C‐terminal lobes, respectively, that are stabilized by multiple interactions within the lobe. Compared to TnC, far less was known about the structures of TnI and TnT. Like TnC, they were both predicted to be highly a‐helical. Moreover, long stretches of heptad repeats were recognized in TnI and TnT so that a specific ‘‘IT’’ a‐helical coiled‐coil was suggested for this part of the structure (Parry, 1981; Pearlstone and Smillie, 1981, 1985;

136 BROWN AND COHEN

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

137

Stefancsik et al., 1998). X‐ray crystallographic and NMR structures of binary complexes between TnC and certain fragments of TnI have also revealed how additional a‐helical segments of TnI might be stabilized (Li et al., 1999; Vassylyev et al., 1998). These complexes in the Ca2þ‐saturated state revealed the locations of certain segments of TnI (see below for definitions), most importantly, that of the ‘‘switch’’ region in the open hydrophobic cleft of the TnC regulatory N‐lobe (Li et al., 1999). As described below, the movements of these segments of TnI are critical in the Ca2þ‐activation of the thin filament. The actual picture of the troponin head, however, could not have been fully anticipated before the recent crystal structures of the Ca2þ‐ activated cardiac (not shown) and skeletal troponin ‘‘head’ region (Fig. 6C and D), both of which contain all three subunits (Takeda et al., 2003)* (Vinogradova et al., 2005). To be sure, one can easily recognize the previously observed or predicted features such as the IT coiled‐coil in these structures, as well as the contacts made between the TnC lobes and * Note that Takeda et al. refer to the globular head region of troponin as the ‘‘core’’ domain, which they further subdivide into a ‘‘regulatory head’’ (consisting of a short helix of TnI bound to the N‐lobe of TnC) and an ‘‘IT’’ arm (consisting of long segments of TnI and TnT2 and the C‐lobe of TnC).

Fig. 6. The structure of troponin and possible linkages to the thin filament. (A) Schematic depiction of troponin bound to tropomyosin based on early electron microscopic studies (adapted from Flicker et al., 1982). The globular head region consists of TnC, TnI, and the C‐terminal part of TnT (TnT2) and is bound to the central region of tropomyosin near residue 190. The tail of troponin consists of the N‐terminal region of TnT (‘‘TnT1’’ or ‘‘T1’’), which interacts with the C‐terminal region of tropomyosin. (The ‘‘continuous’’ interaction depicted between TnT and tropomyosin is oversimplified; see text). The box represents the region illustrated in subsequent panels. (B) Schematic depiction of the head region of skeletal muscle troponin (TnC, red; TnT2, orange; TnI, blue) in the absence of Ca2þ, based on the 7 A˚ resolution crystal structure (Vinogradova and Fletterick, personal communication, 2004), with its TnI inhibitory segment (black dashes) modeled as interacting with actin (wide gray bar). TnT1 was not included in this structure but in this view would extend to the left. (C) 3 A˚ resolution crystal structure of the head region in the presence of Ca2þ. Here, the TnI switch segment is bound in the cleft (now open) of the TnC regulatory domain and the inhibitory segment interacts with the TnC linker (now helically ordered). (D) Schematic of this structure modeled next to tropomyosin‐actin suggesting that the release of the TnI inhibitory segment from actin permits tropomyosin (green bar) to move azimuthally along the actin surface. A similar overall structure of the head region in the Ca2þ state and model of conformational changes that may occur during the switching have been published for the cardiac isoform (Takeda et al., 2003), but certain aspects of the two structures differ (see text). (Figures in panels B, C, D are courtesy of M. Vinogradova and R. Fletterick.)

138

BROWN AND COHEN

the N‐terminal and switch segments of TnI. But now for the first time the relative orientations of these various regions, as well as the connections between them and peptide segments that extend away from them, can be visualized. For example, in TnT2, two a‐helices form a bridge‐like V‐shape that is anchored by the coiled‐coil interaction with TnI; as described below, this conformation in TnT2 constrains the positions of adjacent segments of TnT that bind tropomyosin. Moreover, in the skeletal head structure, the inhibitory segment of TnI is observed to interact with TnC, including its fully helical interlobe linker, providing additional stabilization to the central region of the a‐helix. In the cardiac head structure, both the TnI inhibitory region and TnC interlobe linker are disordered. (Note, however, that recent neutron‐scattering experiments suggest that this linker is a‐helical when attached to the thin filament [Matsumoto et al., 2004]). A key advance has also been made in visualizing the skeletal troponin head in the Ca2þ‐free state with the determination of an interpretable 7 A˚ resolution structure (Vinogradova et al., 2005), as described below (see Fig. 6B).

2.

Tropomyosin‐TnT Interactions

The overall arrangement of troponin bound on tropomyosin was visualized by electron microscopy over 20 years ago (Fig. 6A, Flicker et al., 1982). Troponin, including its long TnT tail, appears to span over half of tropomyosin’s length, with the head region of troponin binding at tropomyosin’s central region around residues 150–190 and the tail of troponin extending toward tropomyosin’s C‐terminus. Moreover, a low‐resolution (17 A˚ ) X‐ray crystallographic study showed that TnT1 extends about 10–30 residues past the head‐tail overlap (White et al., 1987). Additional biochemical experiments have revealed further aspects of this interaction. There appear to be two well‐separated sites on TnT that interact with distinct regions on tropomyosin (Mak and Smillie, 1981). TnT1, and in particular its CB2 fragment (residues 71–151), interact in a Ca2þ‐insensitive, conserved manner with the C‐terminal region of tropomyosin; here, residues (258–275) just prior to the head‐tail overlap appear to be most critical (Hammell and Hitchcock‐DeGregori, 1996; Mak and Smillie, 1981). The linkage between TnT2 in the head region, in particular its C‐terminal residues 243–259, and the middle of tropomyosin near cys–190 is variable, however, being strong only in the absence of Ca2þ (Chong and Hodges, 1982; Pearlstone and Smillie, 1982; Tanokura et al., 1983). Based on the early electron microscope images (Flicker et al., 1982), initial depictions (see Fig. 6A) of the tropomyosin‐troponin interaction had TnT binding continuously along the tropomyosin coiled‐coil. The separation of TnT’s binding sites on tropomyosin is now accounted for,

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

139

in part, by the non‐linear V‐shaped conformation of TnT2 observed in the troponin head crystal structures (see above and Fig. 6B–D; see also results of Oliveira et al., 2000). Recent results on tropomyosin suggest that the initial depiction of the Ca2þ‐insensitive interaction between tropomyosin and TnT1 as a triple‐ stranded coiled‐coil (see e.g., Nagano et al., 1982) needs to be modified in some regions. It appears that, although tropomyosin is an a‐helical coiled‐coil throughout most of its length, a departure from regular coiled‐ coil structure and resulting flexibility at its C‐terminus (see Fig. 4) are critical features for its binding to TnT1. The C‐terminal 31–34 residues of striated‐muscle a‐tropomyosin are the only part of the tropomyosin‐TnT interface whose structure has been determined at atomic resolution, and the a‐helices in its last 15–21 residues either splay apart or lie parallel to one another in its X‐ray crystal (Li et al., 2002) and NMR (Greenfield et al., 2003) structures (see Fig. 4). Such irregularity in the structure appears to be an intrinsic feature of this region in striated‐muscle tropomyosin, which contains many bulky and polar core residues unfavorable for coiled‐coil formation (Li et al., 2002). (Stabilization of the striated‐muscle coiled‐ coil by mutation of core d‐position residue 263 from glutamine to leucine indeed results in decreased interaction with a fragment of TnT1 (Greenfield et al., 2002). The unusual structure near the C‐terminus of vertebrate striated‐muscle tropomyosin contrasts with the corresponding region of tropomyosin in vertebrate smooth muscle, where troponin is lacking. In smooth‐muscle tropomyosin this region has relatively small or medium‐sized core residues and has been predicted to have a relatively regular a‐helical coiled‐coil structure (Li et al., 2002). The fact that the C‐terminal 27 residues are the products of one of only two exons that are alternatively expressed between a vertebrate striated‐muscle and smooth‐ muscle tropomyosin (see Lees‐Miller and Helfman, 1991, and Fig. 1) provides further evidence that the irregular C‐terminal structure in striated‐muscle tropomyosin is specifically designed for interaction with TnT1. In summary, the special features of tropomyosin sequence, which has both periodic features for the binding of F‐actin and aperiodic features for a number of functions including the binding of troponin, are necessary for its special linkages to the striated‐muscle thin filament.

IV. Turning on Troponin The regular distribution of troponin molecules along F‐actin determined by tropomyosin sets the stage for the sequence of events that begins with the binding of Ca2þ to TnC and ends with activation of the thin

140

BROWN AND COHEN

filament. To describe this process, as far as we know it, we begin with the ‘‘off’’ structure of the troponin head. The most detailed structure of the troponin head in the absence of Ca2þ is at 7 A˚ resolution (Vinogradova et al., 2005), available only in the skeletal muscle isoform and at a considerably lower resolution than the Ca2þ‐ activated troponin head structures (see Fig. 6). Nevertheless, much of the map appears to be interpretable and informative. In this structure, and in contrast to the skeletal Ca2þ‐activated structure, the TnC interlobe linker is disordered. Moreover, the inhibitory region (skeletal residues 104–115) and switch region (residues 116–131) appear to be displaced from the positions they occupy in the Ca2þ‐activated structure, and some residues of the inhibitory region may form a helix located between the N‐terminal domain of TnC and the IT coiled‐coil. A variety of experiments on the thin filament in the off state have indicated that the inhibitory segment, as well as C‐terminal residues (141–181) of TnI, both bind actin (Farah et al., 1994; Talbot and Hodges, 1981; Takeda et al., 1997). In this state, as well, the C‐ terminus of TnT2 appears to bind to tropomyosin, near its middle region (Chong and Hodges, 1982; Tanokura et al., 1983; see above). It thus appears that in the off state, key linkages within and between the subunits of the troponin head are weak; at the same time, the troponin head makes strong linkages to tropomyosin‐actin. In going from the off state to the Ca2þ‐activated state, the reverse appears to be true. The N‐lobe of TnC is the first part of the complex to sense the presence of Ca2þ. A reorientation of pairs of a‐helices in the N‐lobe results in the opening of its hydrophobic cleft (Herzberg et al., 1986; Houdusse et al., 1997; Strynadka et al., 1997), into which the switch segment of TnI now binds (Li et al., 1999; Takeda et al., 2003; Vinogradova et al., 2005). Biochemical experiments have shown that in the presence of Ca2þ, the inhibitory region as well as the C‐terminus of TnI move away from actin and toward TnC (Li et al., 2001; Luo et al., 2000). In the skeletal Ca2þ‐activated structure, the inhibitory region of TnI and the TnC interlobe linker interact with one another and are now seen to be ordered in extended and helical conformations, respectively (Vinogradova et al., 2005). These results support a model in which the binding of Ca2þ to TnC, and its subsequent binding of the TnI switch segment, ‘‘drags’’ the rest of the C‐terminal region of TnI away from actin (Takeda et al., 2003; Vinogradova et al., 2005). This general mechanism appears to apply to both skeletal and cardiac troponins. Although the TnI inhibitory segment and the TnC interlobe linker do not appear to be as ordered in the Ca2þ‐activated cardiac troponin head crystal structure as they are

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

141

in the skeletal isoform, the cardiac TnC linker may nevertheless be ordered in the native thin filament according to the neutron‐scattering experiments mentioned above. Yet, some differences in the details of the crystal structures between the two isoforms may be related to specialized sequences, such as residues that inactivate one of the Ca2þ binding sites in the cardiac TnC N‐lobe, and the cardiac‐specific N‐terminal extension of TnI (Vinogradova et al., 2005). It remains to be determined which differences between the isoforms persist in the actin‐associated state. Although precise structural information is lacking on changes in the TnT2‐tropomyosin interaction during switching, the IT coiled‐coil (which includes skeletal TnT residues 200–245 and TnI residues 58–102) provides a good clue. Both the inhibitory region of TnI and the C‐terminus of TnT, which bind actin in the absence of Ca2þ, are immediately adjacent in sequence to the IT coiled‐coil. It is therefore possible that the position of the C‐terminus of TnT may be affected by the motions of the inhibitory region of TnI described above. This speculation suggests a structural coupling between the weakening of the linkages between TnI and actin and the similar weakening between TnT2 and tropomyosin. These current studies provide a plausible picture of how the binding of Ca2þ to TnC leads to the release of tropomyosin from its ‘‘off’’ position on the actin filament. At this point, a review of the current knowledge of how the troponin complex acts as a Ca2þ‐dependent switch may be useful. Previous references have been made to conserved and variable linkages in the complex. This concept is helpful in understanding how a molecular machine or switch functions (Cohen et al., 1972). In a variety of molecules or molecular assemblies, an essential pattern of connections provides the framework that organizes the moving parts. Conserved linkages in troponin are found within the ‘‘IT arm,’’ which consists of the TnC C‐lobe bound on one side by the long N‐terminal a‐helix of TnI and on the other side by the IT coiled‐coil (Takeda et al., 2003; Vinogradova et al., 2005); these regions together form an internally rigid subdomain of the complex. Variable linkages that depend on Ca2þ concentration include those between TnC N‐lobe and TnI switch segment, and the interlobe TnC linker. Two different states of the complex produced by the variable linkages are observed at critical concentrations of Ca2þ ions, but the conserved linkages are the same in both states. Similarly, the linkage between TnT1 and tropomyosin appears to be another invariant connection. Just as tropomyosin is the link between troponin and actin, the TnT1 portion of TnT is the link between the regulatory subunits of the complex and the thin filament.

142

BROWN AND COHEN

V.

Turning on the Thin Filament

A series of striking observations, beginning in the early 1960s, led to the formulation of the so‐called steric‐blocking model, which has been the major paradigm for the mechanism of thin filament activation (reviewed by Cohen and Vibert, 1987). In 1963, Hanson and Lowy (1963) directly visualized the F‐actin helix in the electron microscope, using the then‐new technique of negative staining. At about the same time, using this same technique, Hugh Huxley (1963) showed the polarity of the actin monomer by decorating thin filaments with myosin heads in the rigor state. This eloquent image of the filaments had a unique ‘‘arrowhead’’ appearance (Fig. 7A). The development of three‐dimensional reconstruction by DeRosier and Klug led to a model of the decorated filament revealing the shape and orientation of the myosin heads at a low resolution (Fig. 7B; Moore et al., 1970). An additional finding of large changes in intensity of the layer lines associated with actin by X‐ray diffraction of muscle fibers (Huxley, 1972) was interpreted as showing that tropomyosin moves when contraction is switched on. All this information was then accounted for by Huxley (1972) and by Parry and Squire (1973) in terms of a so‐called steric‐blocking model (Fig. 7C). Steric blocking is a simple idea, which by its very nature, specifies certain spatial relations between proteins (Fig. 7C). To quote Cohen and Vibert (1987): . . . at rest, tropomyosin is positioned by troponin at the edge of the actin groove and blocks the site(s) on actin at which myosin crossbridges attach. At the critical calcium level, troponin changes its conformation and tropomyosin moves deeper into the groove of the actin helix and exposes these binding sites, which then interact with myosin. The muscle is switched ‘‘on’’: ATP splitting can take place and force is developed. When the calcium level is reduced, the troponin complex undergoes its shape change, and tropomyosin rolls out to its ‘‘blocking’’ position. The muscle is switched ‘‘off’’ and is at rest. This model accounts also for the cooperativity discovered by Bremel and Weber (Bremel et al., 1972), whereby the binding of calcium ions to troponin switches on about seven actin monomers. Tropomyosin has an effective length of about 40 nm in muscle, and so spans seven subunits. One troponin could then effect the blocking or unblocking of seven actin monomers because the tropomyosn molecule is linked to this number of subunits.

At the time that this simple two‐state mechanism was conceived—and, indeed, in most attempts to model unknown structures—the proteins were pictured as smooth and symmetrical. In keeping with this image, the motions associated with both regulation and contraction were pictured by many as being simple and machine‐like (but see an alternative view in Squire, 1975).

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

143

Fig. 7. Early views of the decorated thin filament and a model for steric blocking. (A) Electron micrograph of actin filaments ‘‘decorated’’ with S1 heads from rabbit myosin. Each crossover repeat of the actin filament looks like an arrowhead because each actin monomer carries an angled elongated myosin head. (Figure from Craig et al., 1980.) (B) A model for the decorated filament derived from three‐dimensional image processing. The myosin heads are tilted by about 45 degrees to the filament axis. (Figure from Moore et al., 1970.) (C) Early two‐state model of steric blocking by tropomyosin of the attachment of myosin heads to actin. The open circles show the position of tropomyosin in resting muscle; the solid circles show its new position after activation. (Figure from Huxley, 1972.) (All figures reproduced from Cohen and Vibert, 1987).

144

BROWN AND COHEN

A. Binding Sites in Relation to Steric Blocking Shortly after the early steric‐blocking mechanism was proposed, the amino acid sequence of tropomyosin was determined (Hodges et al., 1972; Sodek et al., 1972), revealing an unbroken short‐range, seven‐residue repeat of the a‐helical coiled‐coil along the entire length of the molecule. Analysis of this sequence (McLachlan and Stewart, 1976; Parry, 1975; Stewart and McLachlan, 1975), assuming a one‐dimensional structure, indicated that there were fourteen quasi‐equivalent groups or ‘‘zones’’ of charged and nonpolar residues along the molecule, each about 29 A˚ long. (An equivalent period was observed experimentally by Parry and Squire [1973] in the mg paracrystals of tropomyosin discovered by Cohen and Longley [1996].) As it winds around actin, the regular tropomyosin supercoil would have a pitch of 137 A˚ and would make seven half‐twists relative to actin in a repeat length of 410 A˚ (see Fig. 1B). As indicated previously, the head‐to‐tail overlap in the molecule was estimated to be nine residues in length. On this picture, the fourteen quasi‐equivalent groups of residues were divided into two sets of alternative recognition sites, or so‐called ‘‘zones,’’ which were 90 degrees apart azimuthally on the tropomyosin molecule. The more regular set (a sites) was thought to correspond to recognition sites in the off state of the muscle, while the less regular set (b sites) was considered to correspond to those for the on state. In agreement with Parry (1975), McLachlan and Stewart (1976) pictured the switching between the two states, as described in the steric‐blocking mechanism, as due to a simple quarter‐turn rolling of a regular, symmetrical tropomyosin molecule. Another view of this mechanism was advanced by Phillips and colleagues (Phillips et al., 1986) based on the 15 A˚ resolution crystal structure of the highly hydrated ‘‘Bailey’’ form of the tropomyosin crystals (~95% solvent content). Here it was shown that with an average pitch of ~140 A˚ in the supercoiled filaments in the crystal (the pitch varied from ~135–145 A˚ along the length of the molecule), each half‐turn of the tropomyosin coiled‐coil would make equivalent interactions with each actin monomer of the actin helix. Moreover, the molecule was shown to have both regular and irregular features, and the molecule’s motions were shown to depend on its interaction with actin, troponin, and bound myosin heads. In contrast to the linear sequence analysis cited above, Phillips (appendix to Phillips et al., 1986) extended this study to take into account the relative azimuthal positions of the various side chains and showed that the sevenfold periodicity of the a sites is far more regular than the periodicity of the b sites. These results, together with biochemical evidence (Weber and Murray, 1973), led to a three‐state, rather than a two‐state, model of regulation. In the off state, it was pictured that it is primarily the troponin

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

145

complex, in the absence of Ca2þ, which holds tropomyosin in a position that blocks myosin from interacting strongly with actin. (The b sites would have only a minor, if any, role here.) With increasing Ca2þ, troponin changes its conformation, releasing tropomyosin, which now moves to a new position on the thin filament because of the a sites, which now bind (albeit weakly) to complementary receptor sites on actin.* In this ‘‘Ca2þ state,’’ weak, noncooperative binding of myosin may occur, and the flexible tropomyosin filaments make marked excursions from the actin helix, so that the a sites are not saturated. In the third ‘‘fully‐activated’’ state, called ‘‘potentiated’’ by Weber and Murray (1973), tropomyosin is now more regularly attached to actin at the a sites, and myosin heads bind strongly and cooperatively to actin, so that the muscle is fully switched on. ‘‘Contraction is therefore seen as a statistical mechanism, requiring at least three distinct average conformations for the tropomyosin on the actin helix’’ (Phillips et al., 1986).

B.

Atomic Structures and Molecular Models

By this time, a number of controversies plagued various aspects of the steric‐blocking mechanism. An early challenge concerned the interpretation of reconstructions of decorated thin filaments. Here, density assignments were interpreted differently by different groups, but these difficulties led to improvements in image‐processing techniques, together with improvements in resolution. Major advances in our understanding of the atomic structure of G‐actin (Kabsch et al., 1990), as well as myosin (Rayment et al., 1993), have been critical as well. Actin was visualized as a four‐subdomain monomer (Fig. 8), which has since been shown to exist in both an ‘‘open‐cleft’’ (Chik et al., 1996) and ‘‘closed‐cleft’’ (Graceffa and Dominguez, 2003; Otterbein et al., 2001) conformation—as well as intermediates—depending on nucleotide content and/or the binding of various proteins. Following the first atomic model for F‐actin by Holmes and collaborators (Holmes et al., 1990), a number of other, slightly different models have now been proposed (Holmes et al., 2003; Lorenz et al., 1993; Tirion et al., 1995). In all of these models, the crystal structure of G‐actin was refined against the same X‐ray fiber diffraction intensities (Popp and Holmes, 1992). The results appear to indicate that the monomer is in the closed‐cleft conformation when in F‐actin. (Note that the current model of Egelman and colleagues [Orlova et al., 2001], derived from electron * Note that this role for the a sites was formerly ascribed by McLachlan and Stewart (1996) to the b sites.

146

BROWN AND COHEN

Fig. 8. (A) Schematic of the three states of thin filament regulation. Surface view (~25 A˚ resolution) of a three‐dimensional reconstruction of negatively stained F‐actin (adapted from Figure 4a of Tobacman et al., 2002) superimposed on models of tropomyosin (Tm). The axial positions of tropomyosin (relative to actin) are depicted arbitrarily, but the relative azimuthal positions are derived from thin filament reconstructions (Craig and Lehman, 2001). In the Ca2þ‐free ‘‘off’’ state, tropomyosin (red) contacts actin’s outer domain (which consists of subdomains 1 and 2). In the presence of Ca2þ, tropomyosin (yellow) moves azimuthally away from most of actin’s outer domain. Tropomyosin (green) is positioned over actin’s inner domain (consisting of subdomains 3 and 4) as a result of the strong binding of myosin (not shown). Note that although troponin is present in the thin filaments used by Craig and Lehman (2001), it was not resolved by helical reconstruction. We expect the structure of F‐actin to be somewhat different in each of the three states (see text). (B) and (C) Ribbon representations of the four‐subdomain G‐actin (B) from a profilin‐actin crystal structure in an ‘‘open‐cleft’’ conformation (Chik et al., 1996) (C) and from uncomplexed actin in a ‘‘closed‐cleft’’ conformation (Otterbein et al., 2001), the latter probably being most similar to the actin shown in panel (A).

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

147

microscopy of F‐actin using negative staining, indicates that the monomer is in an open‐cleft conformation, similar to that found in the crystal structure of the actin‐profilin complex with bound adenosine triphosphate [ATP] [Chik et al., 1996].) These results have been essential to the work of Lehman, Craig and Vibert, who, beginning in 1994 (Craig and Lehman, 2002; Lehman et al., 1994, 1995), directly visualized the effects of Ca2þ and the binding of myosin in rigor on the position of tropomyosin on the thin filament. Using the simple technique of negative staining, they have confirmed and extended the basic features of the steric‐blocking mechanism. One major finding of this work is that in vertebrate striated muscle there are three, rather than two, distinct azimuthal positions for tropomyosin on the thin filament (Fig. 8A). The first corresponds to the off state, originally envisaged in the steric‐blocking mechanism. Here, tropomyosin is located at the center of myosin’s binding site on actin so that it blocks strong myosin attachment (at low Ca2þ, troponin present). At higher Ca2þ concentrations, tropomyosin moves to a new position, which only partially blocks strong, specific myosin binding. The remarkable result they have found is a third azimuthal position for tropomyosin, which fully uncovers the myosin binding site and allows strong myosin binding to actin—switching the muscle to the fully on state (Fig. 8A). (Note that 20 years earlier, Weber and Murray [1973] had suggested that the shift in tropomyosin’s position in this fully‐activated state might be too small to be resolved by X‐ray diagrams of muscle.) To be sure, these results, which depend on the methods used in the three‐dimensional reconstructions, need further refinement by the use of improved models for F‐actin and single particle averaging techniques, but the results, even at this stage, lend strong credence to the steric‐blocking model (see also Squire and Morris, 1998).

C. Why Does Tropomyosin Move to a Third Position in the Fully‐Activated State? These structural findings can now be correlated with the results of Phillips and colleagues (Phillips et al., 1986) and relatively recent biochemical and kinetic studies of McKillop, Geeves, and colleagues (Head et al., 1995; McKillop and Geeves, 1991, 1993), who envisaged three states of regulation in the presence of troponin and myosin (so‐called blocked, closed, and open). In the third state, a number of actin subunits are cooperatively activated by strong, specific myosin binding. In fact, recent studies (Lehrer et al., 1997) indicate that the cooperative unit size appears to depend on the strength of the tropomyosin head‐to‐tail overlap in various isoforms.

148

BROWN AND COHEN

A related aspect of the F‐actin structure that has been recognized in the past decade is that F‐actin appears to undergo cooperative allosteric transitions in response to the binding of tropomyosin and troponin and the strong binding of myosin. For example, myosin and tropomyosin, or tropomyosin/troponin, greatly enhance each other’s binding to actin (Tobacman and Butters, 2000; Tobacman et al., 2002). A number of authors (e.g., Ansari et al., 2003; Perry, 2003) have attempted to establish that regulation can, in fact, be accounted for by these conformational changes in F‐actin (see also Reedy et al., 1994). On this view, the movements of tropomyosin are not essential for regulation, but rather incidental to the actual mechanism. Alternatively, the tropomyosin position might act indirectly by modulating the actin structure to promote myosin binding and cycling as anticipated by Parry and Squire (1973). Another way of looking at these results, however, is that these conformational changes may, in fact, be the key to understanding the third position of tropomyosin on the thin filament (Tobacman et al., 2002). We assume, from the sequence and structural studies on tropomyosin (reviewed previously), that there probably is just one regular set of recognition sites for actin—the so‐called a sites (Phillips et al., 1986). And it is likely that these sites bind weakly to actin in the presence of calcium (the second position of tropomyosin). With the strong, specific binding of myosin, the position of these complementary sites on actin may change cooperatively along the thin filament, promoting the azimuthal movement of the tropomyosin filament to a new position, which allows cooperative strong binding of myosin heads and the switching on of the muscle.

D. Additional Recent Results on Regulation Another area of current research that has relevance for regulation concerns the nature of the conformational changes in actin produced by various effectors. Egelman has been a pioneer in the area of actin dynamics and has produced convincing evidence of domain motions in F‐actin by electron microscopy (for example, see Egelman and Orlova, 1995; Orlova et al., 2001). Correspondingly, a number of studies in solution indicate multiple states of F‐actin, especially those involved in intermolecular contacts in the filament (for example, see Kim and Reisler, 2000). A recent article by Dominguez (2004), based largely on crystal structures of actin complexes with actin‐binding proteins, emphasizes the significance of the ‘‘adaptable’’ hydrophobic cleft formed between actin subdomains 1 and 3 (Fig. 8B and C). This region and, in particular, a hydrophobic pocket in subdomain 1, are suggested to be key targets for G‐ and F‐actin–binding proteins, including actin itself.

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

149

Although we do not yet have a crystal structure of F‐actin—nor, indeed, of its ‘‘nucleus,’’ the closely related actin trimer—a recent study of F‐actin at ~8 A˚ resolution by Oda, Maeda, and colleagues (Maeda, 2004, personal communication) appears to be a major step in arriving at an atomic model. Here, greatly improved X‐ray fiber diagrams of oriented gels of F‐actin (Oda et al., 1998) have been used to provide amplitude information, while phase information was obtained by cryoelectron microscopy of F‐actin using single‐particle averaging. These various improvements, while in overall agreement with the earlier low‐resolution Holmes model (Holmes et al., 1990), reveal that the actin monomer is indeed in the closed‐cleft conformation, that the DNAse I loop is not helical, and that the packing between two actin strands (between subdomain 4 and subdomain 1 of the diagonally related upper subunit) is considerably closer than in previous models. At present, this so‐called Oda model is probably the most reliable picture of F‐actin currently available. Attempts are currently being made to determine experimentally the nature of the movement of tropomyosin on actin during regulation. For example, a simple azimuthal ‘‘rolling’’ (as opposed to sliding) of tropomyosin on the surface of actin (as in McLachlan and Stewart, 1976; Parry, 1976) has been invoked to explain recent fluorescence quenching (Holthauzen et al., 2004) and Forster resonance energy transfer experiments (Bacchiocchi and Lehrer, 2002; Bacchiocchi et al., 2004) on actin complexed with modified tropomyosins. The interpretations of all these experiments, however, are based on a very limited number of mutated residues and on a simple, smooth, and conformationally static (surface) representation of tropomyosin and actin. Moreover, any simple rolling model of an extended stretch of tropomyosin would require additional conformational changes within tropomyosin (such as semiflexibility provided by core alanines [Brown et al., 2001]) so that consecutive periods of tropomyosin would remain adjacent to actin. At this stage, the exact residues of tropomyosin involved in the recognition of actin are not known, and the conclusions drawn from these experiments can only be considered preliminary.

VI.

Perspective

Our thesis in this review is that an understanding of structure leads to an understanding of function. But from this account of advances in the areas of tropomyosin, troponin, and muscle regulation, it should be evident that this path is not always straightforward. Of course, general ideas about function often spring from relatively low‐resolution information. This was certainly true for the sliding filament theory of muscle

150

BROWN AND COHEN

contraction, as well as for the steric‐blocking model of regulation. Detailed knowledge of the protein filaments was not necessary for either proposal. In the first model, the idea emerged from the discovery of two interdigitating arrays of filaments that maintain constant length during shortening; in the second, from the inferred changes in the position of tropomyosin on the thin filament during contraction. Higher‐resolution data about more detailed aspects of structure, however, can provide a fuller picture. Thus, only now is information about the detailed conformational changes in crossbridges as they interact with actin being extracted from high‐resolution results of S1 crystals. The physical basis for general theories can emerge, then, from knowledge at a finer level (see Cohen, 1983). It is also often true that very exact information can lead to broad truths. The current resolution of each of the three ‘‘structures’’ we have described roughly corresponds in an inverse manner to the complexity of the structure itself. In the case of the ‘‘simple molecule’’ of tropomyosin, discovered by Kenneth Bailey in 1946 (Bailey, 1946), it has been shown, only as late as 2001, that alanine ‘‘clusters’’ in the two chains of the coiled‐coil produce an axial stagger (of ~1.2 A˚ ) between chains promoting periodic bending of the molecule around actin. In the case of troponin, it took some 40 years after Ebashi discovered this complex (Ebashi and Ebashi, 1964) to visualize some of the interactions among its three subunits (Takeda et al., 2003). And only now do we see how key joints and segments might undergo order–disorder transition in the two states of regulation (Vinogradova et al., 2005). As described previously, the steric‐blocking mechanism involves a series of structures at a higher level of organization than the proteins comprising its machinery and includes actin and myosin as well. Here, electron microscopy has been useful in clarifying the three positions of tropomyosin on actin and the relationship of the structures to the biochemical results. In all these cases, seeing an image—of whatever kind and at whatever resolution—often poses an enigma, and the solution of that enigma allows ‘‘seeing’’ to become ‘‘knowing.’’ One of the implicit themes in the previous discussion was the need for an ‘‘informed eye’’ in looking at images. In this article, we have tried to show how concepts, such as ‘‘knobs into holes’’ packing, requirements of stabilization for a‐helices, and ‘‘specific recognition,’’ all of which are fundamental to protein structure, are keys to interpreting many images. It is also clear that in future a new level of detail will be obtained for the proteins and protein complexes we have described. These results, together with other information about the

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

151

properties of these molecules and assemblies, will lead to improved models for thin filament regulation.

Acknowledgments We are grateful to many colleagues for comments on various sections, including L. Tobacman, Z. Grabarek, W. Lehman, E. Egelman, M. Geeves, S. Lehrer, and D. Parry. This work has been supported by grants from the National Institutes of Health and the Muscular Dystrophy Association (C. C.).

References Ansari, S., El‐Mezgueldi, M., and Marston, S. (2003). Cooperative inhibition of actin filaments in the absence of tropomyosin. J. Muscle Res. Cell. Motil. 24, 513–520. Bacchiocchi, C., Graceffa, P., and Lehrer, S. S. (2004). Myosin‐induced movement of alphaalpha, alphabeta, and betabeta smooth muscle tropomyosin on actin observed by multisite FRET. Biophys. J. 86, 2295–2307. Bacchiocchi, C., and Lehrer, S. S. (2002). Ca(2þ)‐induced movement of tropomyosin in skeletal muscle thin filaments observed by multi‐site FRET. Biophys. J. 82, 1524–1536. Bailey, K. (1946). Tropomyosin: A new asymmetric protein component of muscle. Nature 157, 368–369. Bremel, R. D., Murray, J. M., and Weber, A. (1972). Cooperative behavior in actin. Cold Spring Harbor Symp. Quant. Biol. 37, 267–275. Brown, J. H., Kim, K.‐H., Jun, G., Greenfield, N. J., Dominguez, R., Volkmann, N., Hitchcock‐DeGregori, S. E., and Cohen, C. (2001). Deciphering the design of the tropomyosin molecule. Proc. Natl Acad. Sci. USA 98, 8496–8501. Burkhard, P., Ivaninskii, S., and Lustig, A. (2002). Improving coiled‐coil stability by optimizing ionic interactions. J. Mol. Biol. 318, 901–910. Burkhard, P., Kammerer, R. A., Steinmetz, M. O., Bourenkov, G. P., and Aebi, U. (2000). The coiled‐coil trigger site of the rod domain of cortexillin I unveils a distinct network of interhelical and intrahelical salt bridges. Structure 8, 223–230. Chik, J. K., Lindberg, U., and Schutt, C. E. (1996). The structure of an open state of beta‐actin at 2.65 A resolution. J. Mol. Biol. 263, 607–623. Cho, Y. J., and Hitchcock‐DeGregori, S. E. (1991). Relationship between alternatively spliced exons and functional domains in tropomyosin. Proc. Natl Acad. Sci. USA 88, 10153–10157. Chong, P. C., and Hodges, R. S. (1982). Photochemical cross‐linking between rabbit skeletal troponin and alpha‐tropomyosin. Attachment of the photoaffinity probe N‐(4‐azidobenzoyl‐[2–3H]glycyl)‐S‐(2‐thiopyridyl)‐cysteine to cysteine 190 of alpha‐tropomyosin. J. Biol. Chem. 257, 9152–9160. Cohen, C. (1966). Design and function of fibrous proteins. In ‘‘Principles of Biomolecular Organizations’’ (G. E. W. Wolstenholme and M. O’Connor, Eds.), pp. 101–129. J. & A. Churchill Ltd., London. Cohen, C., and Lonaley, W. (1996). Tropomyosin paracrystals formed by divalent crystals. Science 152, 794–796.

152

BROWN AND COHEN

Cohen, C. (1983). Structural and dynamic aspects of thin filament regulation. In ‘‘Biological structures and coupled flows’’ (A. Oplatka and M. Balaban, Eds.), pp. 181–205. Academic Press, New York. Cohen, C., Caspar, D. L., Parry, D. A., and Lucas, R. M. (1972). Tropomyosin crystal dynamics. Cold Spring Harb. Symp. Quant. Biol. 36, 205–216. Cohen, C., and Holmes, K. C. (1963). X‐ray diffraction evidence for alpha‐helical coiled‐ coils in native muscle. J. Mol. Biol. 6, 423–432. Cohen, C., and Parry, D. A. D. (1990). Alpha‐helical coiled‐coils and bundles: How to design an alpha‐helical protein. Proteins: Structure, Function and Genetics 7, 1–15. Cohen, C., and Vibert, P. J. (1987). Actin filaments: Images and models. In ‘‘Fibrous Protein Structure’’ (J. M. Squire and P. J. Vibert, Eds.). Academic Press, New York. Conway, J. F., and Parry, D. A. D. (1990). Structural features in the heptad substructure and longer range repeats of two‐stranded alpha‐fibrous proteins. Int. J. Biol. Macromol. 12, 328–334. Craig, R., and Lehman, W. (2001). Crossbridge and tropomyosin positions observed in native, interacting thick and thin filaments. J. Mol. Biol. 311, 1027–1036. Craig, R., and Lehman, W. (2002). The ultrastructural basis of actin filament regulation. Results Probl. Cell. Differ. 36, 149–169. Craig, R., Szent‐Gyorgyi, A. G., Beese, L., Flicker, P., Vibert, P., and Cohen, C. (1980). Electron microscopy of thin filaments decorated with a Ca2þ‐regulated myosin. J. Mol. Biol. 140, 35–55. Crick, F. H. C. (1953). The packing of alpha‐helices: Simple coiled‐coils. Acta Crystallogr. 6, 689–697. Day, C. L., and Alber, T. (2000). Crystal structure of the amino‐terminal coiled‐coil domain of the APC tumor suppressor. J. Mol. Biol. 301, 147–156. Dominguez, R. (2004). Actin‐binding proteins—a unifying hypothesis. Trends Biochem. Sci. 29, 572–578. Ebashi, S., and Ebashi, F. (1964). A new protein component participating in the superprecipitation of myosin B. J. Biochem. (Tokyo) 55, 604–613. Ebashi, S., and Endo, M. (1968). Calcium ion and muscle contraction. Prog. Biophys. Mol. Biol. 18, 123–183. Ebashi, S., Endo, M., and Otsuki, I. (1969). Control of muscle contraction. Q. Rev. Biophys. 2, 351–384. Egelman, E. H., and Orlova, A. (1995). New insights into actin filament dynamics. Curr. Opin. Struct. Biol. 5, 172–180. Farah, C. S., Miyamoto, C. A., Ramos, C. H., da Silva, A. C., Quaggio, R. B., Fujimori, K., Smillie, L. B., and Reinach, F. C. (1994). Structural and regulatory functions of the NH2‐ and COOH‐terminal regions of skeletal muscle troponin I. J. Biol. Chem. 269, 5230–5240. Flicker, P. F., Phillips, G. N. J., and Cohen, C. (1982). Troponin and its interactions with tropomyosin. An electron microscope study. J. Mol. Biol. 162, 495–501. Gagne, S. M., Tsuda, S., Li, M. X., Smillie, L. B., and Sykes, B. D. (1995). Structures of the troponin C regulatory domains in the apo and calcium‐saturated states. Nat. Struct. Biol. 2, 784–789. Gernert, K. M., Surles, M. C., Labean, T. H., Richardson, J. S., and Richardson, D. C. (1995). The Alacoil: A very tight, antiparallel coiled‐coil of helices. Protein Sci. 4, 2252–2260.

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

153

Grabarek, Z., Drabikowski, W., Vinokurov, L., and Lu, R. C. (1981). Digestion of troponin C with trypsin in the presence and absence of Ca2þ. Identification of cleavage points. Biochim. Biophys. Acta 671, 227–233. Graceffa, P., and Dominguez, R. (2003). Crystal structure of monomeric actin in the ATP state. Structural basis of nucleotide‐dependent actin dynamics. J. Biol. Chem. 278, 34172–34180. Greenfield, N. J., Montelione, G. T., Farid, R. S., and Hitchcock‐DeGregori, S. E. (1998). The structure of the N‐terminus of striated muscle a‐tropomyosin in a chimeric peptide: Nuclear magnetic resonance structure and circular dichroism studies. Biochemistry 37, 7834–7843. Greenfield, N. J., Palm, T., and Hitchcock‐De Gregori, S. E. (2002). Structure and interactions of the carboxyl terminus of striated muscle alpha‐tropomyosin: It is important to be flexible. Biophys. J. 83, 2754–2766. Greenfield, N. J., Swapna, G. V., Huang, Y., Palm, T., Graboski, S., Montelione, G. T., and Hitchcock‐DeGregori, S. E. (2003). The structure of the carboxyl terminus of striated alpha‐tropomyosin in solution reveals an unusual parallel arrangement of interacting alpha‐helices. Biochemistry 42, 614–619. Hammell, R. L., and Hitchcock‐DeGregori, S. E. (1996). Mapping the functional domains within the carboxyl terminus of alpha‐tropomyosin encoded by the alternatively spliced ninth exon. J. Biol. Chem. 271, 4236–4242. Hanson, J., and Lowy, J. (1963). The structure of F‐actin and of actin filaments isolated from muscle. J. Mol. Biol. 6, 46–60. Harbury, P. B., Zhang, T., Kim, P. S., and Alber, T. (1993). A switch between two‐, three‐, and four‐stranded coiled‐coils in GCN4 leucine zipper mutants. Science 262, 1401–1407. Head, J. G., Ritchie, M. D., and Geeves, M. A. (1995). Characterization of the equilibrium between blocked and closed states of muscle thin filaments. Eur. J. Biochem. 227, 694–699. Heald, R. W., and Hitchcock‐DeGregori, S. E. (1988). The structure of the amino terminus of tropomyosin is critical for binding to actin in the absence and presence of troponin. J. Biol. Chem. 263, 5254–5259. Heeley, D. H., Watson, M. H., Mak, A. S., Dubord, P., and Smillie, L. B. (1989). Effect of phosphorylation on the interaction and functional properties of rabbit striated muscle alpha‐tropomyosin. J. Biol. Chem. 264, 2424–2430. Herzberg, O., and James, M. N. (1985). Structure of the calcium regulatory muscle protein troponin‐C at 2.8 A˚ resolution. Nature 313, 653–659. Herzberg, O., and James, M. N. (1988). Refined crystal structure of troponin C from turkey skeletal muscle at 2.0 A˚ resolution. J. Mol. Biol. 203, 761–779. Herzberg, O., Moult, J., and James, M. N. (1986). A model for the Ca2þ‐induced conformational transition of troponin C. A trigger for muscle contraction. J. Biol. Chem. 261, 2638–2644. Hodges, R. S., Sodek, J., Smillie, L. B., and Jurasek, L. (1972). Tropomyosin: Amino acid sequence and coiled‐coil structure. Cold Spring Harb. Symp. Quant. Biol. 37, 299–310. Holmes, K. C., Angert, I., Kull, F. J., Jahn, W., and Schroder, R. R. (2003). Electron cryo‐ microscopy shows how strong binding of myosin to actin releases nucleotide. Nature 425, 423–427. Holmes, K. C., Popp, D., Gebhard, W., and Kabsch, W. (1990). Atomic model of the actin filament. Nature 347, 44–49.

154

BROWN AND COHEN

Holthauzen, L. M., Correa, F., and Farah, C. S. (2004). Ca2þ‐induced rolling of tropomyosin in muscle thin filaments: The alpha‐ and beta‐band hypothesis revisited. J. Biol. Chem. 279, 15204–15213. Houdusse, A., Love, M. L., Dominguez, R., Grabarek, Z., and Cohen, C. (1997). Structures of four Ca2þ‐bound troponin C at 2.0 A resolution: Further insights into the Ca2þ‐switch in the calmodulin superfamily. Structure 5, 1695–1711. Huxley, H. E. (1963). Electron microscope studies on the structure of natural and synthetic protein filaments from striated muscle. J. Mol. Biol. 77, 281–308. Huxley, H. E. (1972). Structural changes in actin‐ and myosin‐containing filaments during contraction. Cold Spring Harb. Symp. Quant. Biol. 37, 361–376. Jancso, A., and Graceffa, P. (1991). Smooth muscle tropomyosin coiled‐coil dimers. Subunit composition, assembly, and end‐to‐end interaction. J. Biol. Chem. 266, 5891–5897. Kabsch, W., Mannherz, H. G., Suck, D., Pai, E. F., and Holmes, K. C. (1990). Atomic structure of the actin:DNase I complex. Nature 347, 37–44. Kammerer, R. A., Schulthess, T., Landwehr, R., Lustig, A., Engel, J., Aebi, U., and Steinmetz, M. O. (1998). An autonomous folding unit mediates the assembly of two‐stranded coiled‐coils. Proc. Natl Acad. Sci. USA 95, 13419–13424. Kay, C. M., and Bailey, K. (1960). Light scattering in solutions of native and guanidinated rabbit tropomyosin. Biochim. Biophys. Acta 40, 149–156. Kim, E., and Reisler, E. (2000). Intermolecular dynamics and function in actin filaments. Biophys. Chem. 86, 191–201. Kojima, H., Ishijima, A., and Yanagida, T. (1994). Direct measurement of stiffness of single actin filaments with and without tropomyosin by in vitro nanomanipulation. Proc. Natl Acad. Sci. USA 91, 12962–12966. Kwok, S. C., and Hodges, R. S. (2003). Clustering of large hydrophobes in the hydrophobic core of two‐stranded alpha‐helical coiled‐coils controls protein folding and stability. J. Biol. Chem. 278, 35248–35254. Kwok, S. C., and Hodges, R. S. (2004). Stabilizing and destabilizing clusters in the hydrophobic core of long two‐stranded alpha‐helical coiled‐coils. J. Biol. Chem. 279, 21576–21588. Lees‐Miller, J. P., and Helfman, D. M. (1991). The molecular basis for tropomyosin isoform diversity. Bioessays 13, 429–437. Lehman, W., Craig, R., and Vibert, P. (1994). Ca(2þ)‐induced tropomyosin movement in Limulus thin filaments revealed by three‐dimensional reconstruction. Nature 368, 65–67. Lehman, W., Rosol, M., Tobacman, L. S., and Craig, R. (2001). Troponin organization on relaxed and activated thin filaments revealed by electron microscopy and three‐ dimensional reconstruction. J. Mol. Biol. 307, 739–744. Lehman, W., Vibert, P., Uman, P., and Craig, R. (1995). Steric‐blocking by tropomyosin visualized in relaxed vertebrate muscle thin filaments. J. Mol. Biol. 251, 191–196. Lehrer, S. S., Golitsina, N. L., and Geeves, M. A. (1997). Actin‐tropomyosin activation of myosin subfragment 1 ATPase and thin filament cooperativity. The role of tropomyosin flexibility and end‐to‐end interactions. Biochemistry 36, 13449–13454. Li, M. X., Spyracopoulos, L., and Sykes, B. D. (1999). Binding of cardiac troponin‐ I147–163 induces a structural opening in human cardiac troponin‐C. Biochemistry 38, 8289–8298. Li, Y., Brown, J. H., Reshetnikova, L., Blazsek, A., Farkas, L., Nyitray, L., and Cohen, C. (2003). Visualization of an unstable coiled‐coil from the scallop myosin rod. Nature 424, 341–345.

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

155

Li, Y., Love, M. L., Putkey, J. A., and Cohen, C. (2000). Bepridil opens the regulatory N‐terminal lobe of cardiac troponin C. Proc. Natl Acad. Sci. USA 97, 5140–5145. Li, Y., Mui, S., Brown, J. H., Strand, J., Reshetnikova, L., Tobacman, L. S., and Cohen, C. (2002). The crystal structure of the C‐terminal fragment of striated‐muscle alpha‐ tropomyosin reveals a key troponin T recognition site. Proc. Natl Acad. Sci. USA 99, 7378–7383. Li, Z., Gergely, J., and Tao, T. (2001). Proximity relationships between residue 117 of rabbit skeletal troponin‐I and residues in troponin‐C and actin. Biophys. J. 81, 321–333. Lorenz, M., Popp, D., and Holmes, K. C. (1993). Refinement of the F‐actin model against X‐ray fiber diffraction data by the use of a directed mutation algorithm. J. Mol. Biol. 234, 826–836. Lu, S. M., and Hodges, R. S. (2004). Defining the minimum size of a hydrophobic cluster in two‐stranded alpha‐helical coiled‐coils: Effects on protein stability. Prot. Sci. 13, 714–726. Luo, Y., Wu, J. L., Li, B., Langsetmo, K., Gergely, J., and Tao, T. (2000). Photocrosslinking of benzophenone‐labeled single cysteine troponin I mutants to other thin filament proteins. J. Mol. Biol. 296, 899–910. Lupas, A. (1996). Coiled‐coils: New structures and new functions. Trends Biochem. Sci. 21, 375–382. Mak, A., Smillie, L. B., and Barany, M. (1978). Specific phosphorylation at serine–283 of alpha tropomyosin from frog skeletal and rabbit skeletal and cardiac muscle. Proc. Natl. Acad. Sci. USA 75, 3588–3592. Mak, A. S., and Smillie, L. B. (1981). Structural interpretation of the two‐site binding of troponin on the muscle thin filament. J. Mol. Biol. 149, 541–550. Matsumoto, F., Makino, K., Maeda, K., Patzelt, H., Maeda, Y., and Fujiwara, S. (2004). Conformational changes of troponin C within the thin filaments detected by neutron scattering. J. Mol. Biol. 342, 1209–1221. Maytum, R., Geeves, M. A., and Konrad, M. (2000). Actomyosin regulatory properties of yeast tropomyosin are dependent upon N‐terminal modification. Biochemistry 39, 11913–11920. Maytum, R., Konrad, M., Lehrer, S. S., and Geeves, M. A. (2001). Regulatory properties of tropomyosin effects of length, isoform, and N‐terminal sequence. Biochemistry 40, 7334–7341. McKillop, D. F., and Geeves, M. A. (1991). Regulation of the acto.myosin subfragment 1 interaction by troponin/tropomyosin. Evidence for control of a specific isomerization between two acto.myosin subfragment 1 states. Biochem. J. 279(Pt 3), 711–718. McKillop, D. F. A., and Geeves, M. A. (1993). Regulation of the interaction between actin and myosin subfragment 1: Evidence for three states of the thin filament. Biophys. J. 65, 693–701. McLachlan, A. D., and Stewart, M. (1975). Tropomyosin coiled‐coil interactions: Evidence for an unstaggered structure. J. Mol. Biol. 98, 293–304. McLachlan, A. D., and Stewart, M. (1976). The 14‐fold periodicity in alpha‐tropomyosin and the interaction with actin. J. Mol. Biol. 103, 271–298. Meier, M., Lustig, A., Aebi, U., and Burkhard, P. (2002). Removing an interhelical salt bridge abolishes coiled‐coil formation in a de novo designed peptide. J. Struct. Biol. 137, 65–72. Monteiro, P. B., Lataro, R. C., Ferro, J. A., and Reinach, F. D. C. (1994). Functional alpha‐tropomyosin produced in Escherichia coli: A dipeptide extension can substitute the amino‐terminal acetyl group. J. Biol. Chem. 269, 10461–10466.

156

BROWN AND COHEN

Moore, P. B., Huxley, H. E., and De Rosier, D. J. (1970). Three‐dimensional reconstruction of F‐actin, thin filaments and decorated thin filaments. J. Mol. Biol. 50, 279–295. Nagano, K., Miyamoto, S., Matsumura, M., and Ohtsuki, I. (1982). Prediction of a triple‐ stranded coiled‐coil region in tropomyosin‐troponin T complex. J. Theor. Biol. 94, 743–782. O’Shea, E. K., Klemm, J. D., Kim, P. S., and Alber, T. (1991). X‐ray structure of the GCN4 leucine zipper, a two‐stranded, parallel coiled‐coil. Science 254, 539–544. Oda, T., Makino, K., Yamashita, I., Namba, K., and Maeda, Y. (1998). Effect of the length and effective diameter of F‐actin on the filament orientation in liquid crystalline sols measured by x‐ray fiber diffraction. Biophys. J. 75, 2672–2681. Oliveira, D. M., Nakaie, C. R., Sousa, A. D., Farah, C. S., and Reinach, F. C. (2000). Mapping the domain of troponin T responsible for the activation of actomyosin ATPase activity. Identification of residues involved in binding to actin. J. Biol. Chem. 275, 27513–27519. Orlova, A., Galkin, V. E., Van Loock, M. S., Kim, E., Shvetsov, A., Reisler, E., and Egelman, E. H. (2001). Probing the structure of F‐actin: Cross‐links constrain atomic models and modify actin dynamics. J. Mol. Biol. 312, 95–106. Otterbein, L. R., Graceffa, P., and Dominguez, R. (2001). The crystal structure of uncomplexed actin in the ADP state. Science 293, 708–711. Palm, T., Graboski, S., Hitchcock‐DeGregori, S. E., and Greenfield, N. J. (2001). Disease‐ causing mutations in cardiac troponin T: Identification of a critical tropomyosin‐ binding region. Biophys. J. 81, 2827–2837. Parry, D. A. (1975). Analysis of the primary sequence of alpha‐tropomyosin from rabbit skeletal muscle. J. Mol. Biol. 98, 519–535. Parry, D. A. (1981). Analysis of the amino acid sequence of a tropomyosin‐binding fragment from troponin‐T. J. Mol. Biol. 146, 259–263. Parry, D. A. D. (1976). Movement of tropomyosin during regulation of vertebrate skeletal muscle: A simple physical model. Biochem. Biophys. Res. Comm. 68, 323–328. Parry, D. A. D., and Squire, J. M. (1973). Structural role of tropomyosin in muscle regulation analysis of the x‐ray diffraction patterns from relaxed and contracting muscles. J. Mol. Biol. 75, 33–55. Paulucci, A. A., Katsuyama, A. M., Sousa, A. D., and Farah, C. S. (2004). A specific C‐terminal deletion in tropomyosin results in a stronger head‐to‐tail interaction and increased polymerization. Eur. J. Biochem. 271, 589–600. Pearlstone, J. R., and Smillie, L. B. (1981). Identification of a second binding region on rabbit skeletal troponin‐T for alpha‐tropomyosin. FEBS Lett. 128, 119–122. Pearlstone, J. R., and Smillie, L. B. (1982). Binding of troponin‐T fragments to several types of tropomyosin. Sensitivity to Ca2þ in the presence of troponin‐C. J. Biol. Chem. 257, 10587–10592. Pearlstone, J. R., and Smillie, L. B. (1985). The interaction of rabbit skeletal muscle troponin‐T fragments with troponin‐I. Can. J. Biochem. Cell. Biol. 63, 212–218. Perry, S. V. (1998). Troponin T: Genetics, properties and function. J. Muscle Res. Cell. Motil. 19, 575–602. Perry, S. V. (1999). Troponin I: Inhibitor or facilitator. Mol. Cell. Biochem. 190, 9–32. Perry, S. V. (2001). Vertebrate tropomyosin: Distribution, properties and function. J. Muscle Res. Cell. Motil. 22, 5–49. Perry, S. V. (2003). What is the role of tropomyosin in the regulation of muscle contraction? J. Muscle Res. Cell. Motil. 24, 593–596.

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

157

Phillips, G. N., and Chacko, S. (1996). Mechanical properties of tropomyosin and implications for muscle regulation. Biopolymers 38, 89–95. Phillips, G. N., Fillers, J. P., and Cohen, C. (1986). Tropomyosin crystal structure and muscle regulation. J. Mol. Biol. 192, 111–131. Pollard, T. D. (1999). Actin and associated proteins: Introduction. In ‘‘Guidebook to the Cytoskeletal and Motor Proteins’’ (T. Kreis and R. D. Vale, Eds.), pp. 3–110. Oxford University Press, Oxford. Popp, D., and Holmes, K. C. (1992). X‐ray diffraction studies on oriented gels of vertebrate smooth muscle thin filaments. J. Mol. Biol. 224, 65–76. Potter, J. D., and Gergely, J. (1975). The calcium and magnesium binding sites on troponin and their role in the regulation of myofibrillar adenosine triphosphatase. J. Biol. Chem. 250, 4628–4633. Rayment, I., Rypniewski, W. R., Schmidt‐Ba¨ se, K., Smith, R., Tomchick, D. R., Benning, M. M., Winkelmann, D. A., Wesenberg, G., and Holden, H. M. (1993). Three‐ dimensional structure of myosin subfragment–1: A molecular motor. Science 261, 50–58. Reedy, M. K., Reedy, M. C., and Schachat, F. (1994). Does resolution lead to reconciliation? Curr. Biol. 4, 624–626. Ruiz‐Opazo, N., and Nadal‐Ginard, B. (1987). Alpha‐tropomyosin gene organization. Alternative splicing of duplicated isotype‐specific exons accounts for the production of smooth and striated muscle isoforms. J. Biol. Chem. 262, 4755–4765. Sanders, C., and Smillie, L. B. (1984). Chicken gizzard tropomyosin: Head‐to‐tail assembly and interaction with F‐actin and troponin. Can. J. Biochem. Cell. Biol. 62, 443–448. Sano, K.‐I., Maeda, K., Oda, T., and Maeda, Y. (2000). The effect of single residue substitutions of serine-283 on the strength of head‐to‐tail interaction and actin binding properties of rabbit skeletal muscle a‐tropomyosin. J. Biochem. 127, 1095–1102. Satyshur, K. A., Rao, S. T., Pyzalska, D., Drendel, W., Greaser, M., and Sundaralingam, M. (1988). Refined structure of chicken skeletal muscle troponin C in the two‐calcium state at 2‐A resolution. J. Biol. Chem. 263, 1628–1647. Sia, S. K., Li, M. X., Spyracopoulos, L., Gagne, S. M., Liu, W., Putkey, J. A., and Sykes, B. D. (1997). Structure of cardiac muscle troponin C unexpectedly reveals a closed regulatory domain. J. Biol. Chem. 272, 18216–18221. Singh, A., and Hitchcock‐DeGregori, S. E. (2003). Local destabilization of the tropomyosin coiled‐coil gives the molecular flexibility required for actin binding. Biochemistry 42, 14114–14121. Singh, A., and Hitchcock‐DeGregori, S. E. (2004). What makes the tropomyosin coiled‐ coil an actin binding protein [annual meetings abstracts]. Biophys. J., 86, 361a. Slupsky, C. M., and Sykes, B. D. (1995). NMR solution structure of calcium‐saturated skeletal muscle troponin C. Biochemistry 34, 15953–15964. Sodek, J., Hodges, R. S., Smillie, L. B., and Jurasek, L. (1972). Amino‐acid sequence of rabbit skeletal tropomyosin and its coiled‐coil structure. Proc. Natl Acad. Sci. USA 69, 3800–3804. Sousa, A. D., and Farah, C. S. (2002). Quantitative analysis of tropomyosin linear polymerization equilibrium as a function of ionic strength. J. Biol. Chem. 277, 2081–2088. Squire, J. M. (1975). Muscle filament structure and muscle contraction. Ann. Rev. Biophys. Bioeng. 4, 137–163.

158

BROWN AND COHEN

Squire, J. M., and Morris, E. P. (1998). A new look at thin filament regulation in vertebrate skeletal muscle. FASEB J. 12, 761–771. Stefancsik, R., Jha, P. K., and Sarkar, S. (1998). Identification and mutagenesis of a highly conserved domain in troponin T responsible for troponin I binding: Potential role for coiled‐coil interaction. Proc. Natl Acad. Sci. USA 95, 957–962. Steinmetz, M. O., Stock, A., Schulthess, T., Landwehr, R., Lustig, A., Faix, J., Gerisch, G., Aebi, U., and Kammerer, R. A. (1998). A distinct 14 residue site triggers coiled‐coil formation in cortexillin I. EMBO J. 17, 1883–1891. Stewart, M. (2001). Structural basis for bending tropomyosin around actin in muscle thin filaments. Proc. Natl Acad. Sci. USA 98, 8165–8166. Stewart, M., and McLachlan, A. D. (1975). Fourteen actin‐binding sites on tropomyosin? Nature 257, 331–333. Stone, D., Sodek, J., Johnson, P., Smillie, L.B. (1975). In: Proc. 9th FEBS Meeting 31, 125–136. Budapest. Strynadka, N. C., Cherney, M., Sielecki, A. R., Li, M. X., Smillie, L. B., and James, M. N. (1997). Structural details of a calcium‐induced molecular switch: X‐ray crystallographic analysis of the calcium‐saturated N‐terminal domain of troponin C at 1.75 A resolution. J. Mol. Biol. 273, 238–255. Takeda, S., Kobayashi, T., Taniguchi, H., Hayashi, H., and Maeda, Y. (1997). Structural and functional domains of the troponin complex revealed by limited digestion. Eur. J. Biochem. 246, 611–617. Takeda, S., Yamashita, A., Maeda, K., and Maeda, Y. (2003). Structure of the core domain of human cardiac troponin in the Ca(2þ)‐saturated form. Nature 424, 35–41. Talbot, J. A., and Hodges, R. S. (1981). Synthetic studies on the inhibitory region of rabbit skeletal troponin I. Relationship of amino acid sequence to biological activity. J. Biol. Chem. 256, 2798–2802. Tanokura, M., Tawada, Y., Ono, A., and Ohtsuki, I. (1983). Chymotryptic subfragments of troponin T from rabbit skeletal muscle. Interaction with tropomyosin, troponin I and troponin C. J. Biochem. (Tokyo) 93, 331–337. Tirion, M. M., ben‐Avraham, D., Lorenz, M., and Holmes, K. C. (1995). Normal modes as refinement parameters for the F‐actin model. Biophys. J. 68, 5–12. Tobacman, L. S., and Butters, C. A. (2000). A new model of cooperative myosin‐thin filament binding. J. Biol. Chem. 275, 27587–27593. Tobacman, L. S., Nihli, M., Butters, C., Heller, M., Hatch, V., Craig, R., Lehman, W., and Homsher, E. (2002). The troponin tail domain promotes a conformational state of the thin filament that suppresses myosin activity. J. Biol. Chem. 277, 27636–27642. Tripet, B., Wagschal, K., Lavigne, P., Mant, C. T., and Hodges, R. S. (2000). Effects of side‐chain characteristics on stability and oligomerization state of a de novo‐ designed model coiled‐coil: 20 amino acid substitutions in position ‘‘d’’. J. Mol. Biol. 300, 377–402. Vassylyev, D. G., Takeda, S., Wakatsuki, S., Maeda, K., and Maeda, Y. (1998). Crystal structure of troponin C in complex with troponin I fragment at 2.3‐A resolution. Proc. Natl. Acad. Sci. USA 95, 4847–4852. Vinogradova, M. V., Stone, D. B., Malania, G. G., Karatzaferi, C., Cooke, R., Mendelson, R. A., and Fletterick, R. J. (2005). Ca2t‐regulated structural changes in troponin. Proc. Natl. Acad. Sci. USA 102, 5038–5043.

STRUCTURE AND FUNCTION OF TROPOMYOSIN AND TROPONIN

159

Walsh, M., Stevens, F. C., Kuznicki, J., and Drabikowski, W. (1977). Characterization of tryptic fragments obtained from bovine brain protein modulator of cyclic nucleotide phosphodiesterase. J. Biol. Chem. 252, 7440–7443. Walther, D., Eisenhaber, F., and Argos, P. (1996). Principles of helix‐helix packing in proteins: The helical lattice superposition model. J. Mol. Biol. 255, 536–553. Weber, A., and Murray, J. M. (1973). Molecular control mechanisms in muscle contraction. Physiol. Rev. 53, 612–673. Wegner, A. (1979). Equilibrium of the actin‐tropomyosin interaction. J. Mol. Biol. 131, 839–853. Whitby, F. G., and Phillips, G. N. (2000). Crystal structure of tropomyosin at 7 A˚ ngstroms resolution. Proteins: Structure, Function and Genetics 38, 49–59. White, S. P., Cohen, C., and Phillips, G. N., Jr. (1987). Structure of co‐crystals of tropomyosin and troponin. Nature 325, 826–828. Woolfson, D. N., and Alber, T. (1995). Predicting oligomerization states of coiled‐coils. Prot. Sci. 4, 1596–1607. Wu, K. C., Bryan, J. T., Morasso, M. I., Jang, S. I., Lee, J. H., Yang, J. M., Marekov, L. N., Parry, D. A., and Steinert, P. M. (2000). Coiled‐coil trigger motifs in the 1B and 2B rod domain segments are required for the stability of keratin intermediate filaments. Mol. Biol. Cell 11, 3539–3558. Zhang, M., Tanaka, T., and Ikura, M. (1995). Calcium‐induced conformational transition revealed by the solution structure of apo calmodulin. Nat. Struct. Biol. 2, 758–767.

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION By MICHAEL A. GEEVES* AND KENNETH C. HOLMES{ *Department of Biosciences, University of Kent, Canterbury, Kent CT2 7NJ, United Kingdom; {Max Planck Institute for Medical Research, Heidelberg 69120, Germany

I.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. General Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. The Lymn–Taylor Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. The Swinging Lever Arm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Expression Helps Crystallization and Kinetic Studies . . . . . . . . . . . . . . . . . II. Structure of the Crossbridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Naming of Parts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Crossbridge Polymorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Three Primary Conformations Have Been Characterized . . . . . . . . . . . . . B. Post‐Rigor and Pre‐Powerstroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Structural Origin of the Powerstroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Rigor State: Actin Binding Closes the 50K Cleft . . . . . . . . . . . . . . . . . . . . . . E. Myosin V Provides a Model for the Rigor State. . . . . . . . . . . . . . . . . . . . . . . F. Twisting the b‐Sheet Takes the Kink Out of the Relay Helix. . . . . . . . . . G. A Model of the Strongly Bound Pre-Powerstroke . . . . . . . . . . . . . . . . . . . . . H. Classifying the Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Biochemistry and Kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Structural Correlations of the Lymn–Taylor Cycle . . . . . . . . . . . . . . . . . . . . B. From Rigor to Post‐Rigor: ATP Binding and Actin Binding to the Negatively Linked Crossbridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. From Post‐Rigor to Pre‐Powerstroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. From Pre‐Powerstroke to Powerstroke: Pi Release . . . . . . . . . . . . . . . . . . . . . . . . E. Unloaded Powerstroke: Acto‐S1 in Solution . . . . . . . . . . . . . . . . . . . . . . . . . . F. Isometric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. ADP Release and Back to Rigor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. The Relay Helix Kink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Actin Binding Appears to Drive the Powerstroke via a b‐Sheet Twist . . C. Phosphate Release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I.

Introduction

A.

General Overview

161 161 162 163 164 165 165 166 166 169 171 171 172 173 175 177 178 178 179 181 182 185 186 187 188 188 188 189 189

Myosin from muscle (myosin II) consists of two long polypeptide chains (heavy chains) combined with four light chains. The long C‐terminal portions of the heavy chains dimerize to form an a‐helical coiled‐coil. ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71005-0

161

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

162

GEEVES AND HOLMES

The N‐terminal portions of the heavy chains together with the light chains form two globular ‘‘heads’’ or ‘‘crossbridges.’’ In cross‐striated muscle the tails of the molecules pack together to form the thick filaments (see Squire et al., 2005), while the crossbridges, which are ATPases, point away from the thick filaments and cyclically interact with the actin filaments, moving them along by a kind of rowing action. The fuel for this process is provided by the hydrolysis of adenosine triphosphate (ATP). Each interaction produces ~10 nm of sliding movement or, if the muscle is constrained, a few piconewtons of force. Many other non‐muscle myosins have been discovered. All myosins appear to function by the same mechanism. During the past 15 years the structures of the myosin crossbridge and actin filaments have been elucidated. Furthermore, the myosin structure has been solved in a number of different conformations. It appears that actin is the more passive partner—its primary function is to provide binding sites for crossbridges. In skeletal muscles the actin filament contains associated proteins, troponin and tropomyosin, that regulate muscle contraction: the actin filament is ‘‘turned on’’ when calcium binds to troponin (see Brown and Cohen, 2005). However, for studies of the underlying basis of the force production, the focus is on myosin. The myosin crossbridge may be cleaved from the myosin molecule as a soluble molecular fragment (myosin subfragment 1 [S1], ~120 kD; Margossian and Lowey 1973a,b), which contains three polypeptide chains, one heavy and two light. Myosin S1 is a fully competent, actin‐activated, ATPase that transports actin in in vitro motility assays. Moreover, in the absence of nucleotide, it forms a tight (rigor) bond to actin filaments. Biochemical and structural investigations have therefore concentrated on myosin S1 and its interactions with actin as a minimal model for muscle. Much is known about the steps in the biochemical reaction of ATP breakdown by myosin and how these relate to the production of force by the crossbridge. However, since it is no longer attached to the myosin thick filament, myosin S1 cannot be an adequate model for a strained crossbridge. Thus data from muscle fibers (e.g., the dependence of phosphate affinity on strain) must also be considered. In this review we attempt to summarize the currently known structural data on myosin and produce a synthesis of this with the biochemical data. We start with an analysis of the polymorphism of the myosin crossbridge and relate this to the crossbridge cycle proposed by Lymn and Taylor (1971).

B.

The Lymn–Taylor Cycle

Myosin is a product‐inhibited ATPase with an active site and enzyme mechanism similar to that of the G‐proteins, including the P‐loop

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

163

nucleotide‐binding site and the flanking switch 1 (SW1) and switch 2 (SW2) motifs (see Squire and Parry, 2005). The substrate is MgATP. The ATPase is strongly stimulated by binding to actin, which is a nucleotide‐exchange factor for myosin. In the absence of nucleotide, the myosin crossbridge binds tightly to the actin filament to form the ‘‘strong’’ or ‘‘rigor’’ complex. The binding of ATP to the ATPase site on the myosin crossbridge rapidly dissociates the actomyosin complex. Myosin then hydrolyzes ATP and forms a stable myosin‐products complex (ADP.Pi). This reaction primes the crossbridge, which then reattaches to a neighboring actin site. Binding to actin causes a crossbridge to change its shape so as to move the actin approximately 10 nm, the ‘‘powerstroke’’ or ‘‘working stroke.’’ Binding to actin also enables the release of products. First, the phosphate is released from the crossbridge and then the ADP. At the end of the powerstroke, ADP is released, which allows a new ATP molecule to bind to the myosin. A rapid release of the crossbridge from the actin filament follows and the cycle starts again. This crossbridge cycle was first proposed by Lymn and Taylor (1971; Fig. 1A). In the absence of ATP (rigor mortis), the crossbridge binds tightly to actin in the end of powerstroke conformation. Because there is an abundance of ATP present in a living muscle cell, this state is only transitorily present in the active Lymn–Taylor cycle.

C. The Swinging Lever Arm X‐ray crystallographic studies show the myosin II crossbridge to have a long a‐helical tail at its C‐terminus that looks like a lever arm. This lever arm binds the two calmodulin‐like ‘‘light chains,’’ which apparently stabilize the a‐helix and make it stiff. The lever arm has been found by X‐ray crystallography in two conformations, 60
apart, that appear to be the two ends of the powerstroke. This and many other experiments have led to the consensus that most of the movement occurring during the powerstroke arises from a rotation of the lever arm, which, when the crossbridge attaches to actin, lies distal to the actin filament (the swinging lever arm hypothesis; Fig. 1B). In addition, recombinant DNA technology may be used to change the crossbridge structure in defined ways. Most dramatically, using recombinant DNA technology to alter the length of the lever arm produces a sliding velocity proportional to the length of the lever arm (Uyeda et al., 1996; Warshaw et al., 2000). Warshaw et al. also measured the myosin step size from single molecules in an optical trap. This method also shows that the step size is directly proportional to the length of the lever arm; see also (Purcell et al., 2002). However, different myosins may have different step‐sizes for the same length of lever arm (Ko¨ hler et al., 2003).

164

GEEVES AND HOLMES

Fig. 1. (A) The crossbridge cycle (Lymn and Taylor, 1971), with annotation. The binding of ATP to the actin–myosin complex (1) leads to rapid dissociation of the crossbridge from actin, but without hydrolysis of ATP (2). The crossbridge then undergoes a conformational change (recovery stroke) that puts the lever arm in the pre‐ powerstroke or ‘‘UP’’ conformation (3). This form is the ATPase. Subsequent rebinding to actin (4) leads to product release and the moving ‘‘DOWN’’ of the lever arm (powerstroke). Species 1, 2, and 3 have been named on the basis of three crystal structures described in the following text. No crystal structure analogous to species 4 has been defined to date and its exact form is poorly defined. This is discussed in detail in ‘‘From Pre‐Powerstroke to Rigor.’’ (B) The conformational change in the crossbridge consists of a rotation of the distal lever arm (swinging lever arm) about an axis that passes through the SH2 residue (after Cooke, 1986). Most of the mass of the crossbridge takes does not take part in this motion. The axis of rotation is about 30 A˚ from the actin surface.

These experiments allow the radial position of the fulcrum to be estimated. It lies at a radius that is fully consistent with the proposals arising from crystallographic studies and lends no support to earlier proposals that the crossbridge rolled on the surface of actin.

D. Expression Helps Crystallization and Kinetic Studies There is a great deal of literature on the kinetics of myosin S1 (crossbridge) ATPase. Originally biochemical studies were restricted to rabbit skeletal myosin. However, recent studies have often been conducted on

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

165

Dictyostelium myosin II, since this myosin can be expressed in plasmids and its sequence manipulated (Manstein et al., 1989). Similarly, a number of myosin fragments have been successfully expressed in insect cells (expressed using baculovirus; Onishi et al., 1995; Pato et al., 1996; Sweeney et al., 1998). This has permitted kinetic studies to be extended to a range of different myosins. Sometimes the kinetic properties lead to unusual myosin conformations being dominant, a situation that can be exploited to explore new structural states (Sweeney and Houdusse, 2004). Changes in the intrinsic protein fluorescence play a central role in signaling conformational states of myosin and linking these to catalytic events. Expression may be used to remove or insert tryptophan markers at strategic points (Kovacs et al., 2002). The structures of a number of different conformers of the crossbridge have been characterized by X‐ray crystallography (see, for example, Bauer et al., 2000; Coureux et al., 2003; Dominguez et al., 1998; Houdusse et al., 1999; Rayment et al., 1993a; Smith and Rayment, 1996). The ability to express myosin constructs in plasmids has been of great importance in the crystallography of myosin. In particular, it allows crystallographers to work with truncated forms of the crossbridge, mostly without the lever arm. Whereas the crystallization of the full‐length crossbridge was a heroic task, truncated myosins often appear to crystallize rather easily. Structural data, however, can only yield the stable end states of a dynamic process. Moreover, so far there are no X‐ray crystallographic data on the actin–myosin complex, meaning that electron microscopy data must be used to study the actin–myosin interaction. These data are of limited resolution and can only be interpreted by combining them with high‐resolution structural data derived from myosin crossbridges not bound to actin. Therefore, we depend on kinetic and physiological measurements to correlate the crystallographic data with the crossbridge cycle.

II.

Structure of the Crossbridge A. Naming of Parts

The proteolytic fragment S1 comprises the first 843 residues of the heavy chain together with the two light chains. It is the morphological crossbridge and contains all of the enzymatic activity of myosin and is the smallest fragment of myosin with motor activity similar to the parent myosin. Further limited proteolysis breaks the S1 into three fragments named after their apparent molecular weights: 25K (N‐terminal), 50K (middle), and 20K (C‐terminus). These fragments were originally thought

166

GEEVES AND HOLMES

to represent subdomains of the S1. The structure of S1 (see below) shows them rather to mark the positions of flexible loops (loop 1 and loop 2) in the S1. The first crystallographic structure solved was of subfragment 1 (S1) from chicken skeletal myosin crystallized at high salt and without bound nucleotide (Rayment et al., 1993a). This structure is now recognized to be the post‐rigor state (see Fig. 1A). The myosin crossbridges have an elongated head, with a seven‐stranded b‐sheet and associated a‐helices forming a deep cleft at one end. The cleft separates two subdomains of the molecule, referred to as the upper 50K and lower 50K domains, both of which are involved in actin‐binding. Rayment et al. (1993b) suggested that actin binding might involve cleft closure, which does seem to be the case. The ATP binding site, which lies close to the apex of the cleft, consists of a ‘‘P‐loop’’ motif flanked by SW1 and SW2 segments, similar to those found in the G‐proteins. The SW2 segment, which is structurally part of the lower 50K domain, connects with the ‘‘relay helix.’’ The C‐terminal region of the crossbridge, sometimes called the ‘‘neck’’ but now mostly referred to as the ‘‘lever arm,’’ provides the connection to the thick filament (in myosin II) or to the cargo in non‐muscle myosins. This region forms an extended a‐helix containing repeated ‘‘IQ motifs,’’ each of which binds a calmodulin‐like light chain (or calmodulin in other myosins). The number of IQ motifs varies with myosin type, as does the length of the lever arm. In the case of myosin II, there are two IQ motifs; in the case of myosin V there are six. The N‐terminal or proximal end (referring to the complex with the actin helix; see Fig. 5) of the lever arm is anchored in the converter domain that is firmly attached to the distal end of the relay helix.

III. A.

Crossbridge Polymorphism

Three Primary Conformations Have Been Characterized

To date, three primary conformations of the myosin crossbridge that can be associated with states in the Lymn–Taylor cycle have been identified. These have been named the post‐rigor structure (Fig. 2 and state 2 in Fig. 1), the pre‐powerstroke structure (corresponding to the myosin products complex, M.D.Pi, state 3 in Fig. 1), and the rigor‐like (or rigor structure if it is associated with actin) state (shown as state 1 in Fig. 1). A comparison of these structures leads to the identification the following important conformationally flexible elements:  

The positions of the converter domain The kink in the relay helix

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

167

Fig. 2. The post‐rigor structure of the myosin motor domain (Rayment et al., 1993a). The structure of the myosin crossbridge is shown as a ribbon diagram in the orientation it would take on binding to actin viewed from the pointed (‐) end of the actin filament (see Fig. 5A). The N‐terminus is shown in green and the nucleotide binding P‐loop and adjoining helix are shown in yellow. The upper 50K domain is red, the lower 50K domain is gray. Note the cleft separating the upper and lower 50K domains. The lower 50K domain appears to be the primary actin‐binding site. The N‐terminal boundary of the upper 50K domain comprises the disordered loop 1 (between the points marked A and B). The upper and lower 50K domains are also connected by a disordered loop (loop 2 between C and D). The C‐terminal long helix (dark blue) carries two calmodulin‐like light chains and joins onto the thick filament. The relay helix and converter domain are shown. In this conformation of the crossbridge (post‐rigor state), and the lever arm is in the post‐powerstroke position or ‘‘DOWN,’’ as in the rigor state. The colors correspond with subdomain boundaries. For clarity, the proximal end of the relay helix is shown light blue, although it is actually part of the lower 50K domain. The distal end (beyond the kink) is firmly attached to the converter domain. In the post‐rigor structure the relay helix is straight (no kink).

   

The positions of the switch segments SW1 and SW2 The status of the actin‐binding cleft: open, half open, or shut The position of the P‐loop The degree of twist of the central b‐sheet.

These attributes are not independent (Table I). SW1 opening and the actin cleft fully closing are strongly linked. The position of the P‐loop and the b‐sheet twist appear to be strongly linked. Actin cleft half closing and SW2 ‘‘closed’’ broadly describe the same phenomenon, although the

168

Table I Definition of Main Conformationsa SW1

Outer Cleft

Inner Cleft

SW2

1 Rigor A.M 2 Post‐rigor M#ATP 3 Pre‐power stroke M*ATP/M*ADP.Pi 4 Strongly attached top M*ADP.Pi

Open

Shut

C1

C1

Maximum twist

Up

Straight

Down

Closed

Open

Open

Open

No‐twist

Down

Straight

Down

Closed

Open

C2

C2

No‐twist

Down

Kinked

Up

Closed

Closed

C2

C2

No‐twist

Down

Kinked

Up

a

b‐sheet

P‐loop

Relay Helix

Converter‐ Lever

Labeled 1–3 The values of the eight movable elements that have been identified in the text for the conformations defined in Fig. 1. C1 and C2 are two forms of ‘‘closed’’ structure found in rigor and in the pre‐powerstroke. The assumed properties of conformation 4 (shown in italics) are extrapolated from the adjoining states in the Lymn–Taylor cycle.

GEEVES AND HOLMES

State

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

169

details are different. The first two elements—position of the converter, kink in the relay helix—appear to be strongly linked (and moreover, are strongly linked to the intrinsic protein fluorescence, W510 in chicken skeletal myosin or W501 in Dictyostelium myosin II). We first describe these states and then present biochemical and kinetic arguments for assigning them to the positions in the Lymn–Taylor cycle shown in Fig. 1.

B. Post‐Rigor and Pre‐Powerstroke The majority of crystal structures of the myosin crossbridge solved so far fall into two broad classes depending on the position of the converter domain, which can rotate through 60
with respect to the body of the crossbridge. Since the distal lever arm is firmly attached to the converter domain, this leads to a 60
rotation of the lever arm. In the case of myosin II, this amounts to a change in position of the end of the lever arm of about 10 nm, which is often referred to as the powerstroke (sometimes as the working stroke). The two classes that appear to represent the states of the crossbridge at the two ends of the powerstroke are typified by the structural states now known as the pre‐powerstroke state and post‐rigor state. The post‐ rigor state (see above) was actually the first structure of the myosin crossbridge to be solved and, since there was no nucleotide in the active site, it was thought to be close to the true rigor state. However, it is now recognized as a weak actin‐binding state and is therefore not a suitable model for the rigor state. The post‐rigor state is probably the form myosin takes after the rebinding of ATP at the end of the powerstroke. The recently solved structure of a nucleotide‐free myosin V construct appears to be very close to the true rigor state (Holmes et al., 2004; Sweeney and Houdusse, 2004) and is a good model for the end of the powerstroke (thus this class of structures is referred to as rigor-like; see below). This state shares one important property with the post‐rigor state: the converter domain is in a position that appears to be at the end of the powerstroke. In the post‐rigor state, the SW2 segment is ~5 A˚ away from the g‐ phosphate of the ATP. This is referred to as the SW2 open structure (see Geeves and Holmes 1999; also Chapter 1). In the pre‐powerstroke state, the SW2 segment is in (H‐bond) contact with the g‐phosphate (Pi). This is referred to as the SW2 closed structure. Since SW2 is part of the lower 50K domain, in going from the post‐rigor to the pre‐powerstroke structures the whole of the lower 50K domain actually rotates about the W‐helix through about 5
inwards toward the b‐sheet. However, since the relay helix is anchored in the lower 50K domain, one effect of this movement is to force the relay helix against the b‐sheet. The relay helix responds by forming a kink that rotates the converter domain through 60
into the

170

GEEVES AND HOLMES

pre‐powerstroke conformation (Smith and Rayment, 1996). In the pre‐ powerstroke state, the b‐sheet is held in the same untwisted conformation as in the post‐rigor state by the nucleotide bound in the active site, which constrains the P‐loop and stops the twisting of the b‐sheet. The phenomena are illustrated in Fig. 3.

Fig. 3. Details of the seven‐stranded b‐sheet and associated structures (A and B) in the post‐rigor conformation and (C and D) in the pre‐powerstroke conformation. The orientation of A and C is at right angles to that shown in Fig. 2. When attached to actin, it corresponds to that shown in Fig. 5B. The colors are as in Fig. 2. The views shown in B and D are at right angles to A and C looking out radially from the axis of the actin helix. Note the kink in the relay helix shown in C and D that leads to a 60
rotation of the converter domain. This in turn rotates the lever arm 60
. The P‐loop (which constitutes the ATP‐binding site) and the adjoining a‐helix are shown in yellow. The flanking switch sequences (1 and 2) are also shown. The strands of the b‐sheet are numbered from the N‐terminal (distal) end of the sheet. The lower part of strand 5 (light blue) constitutes switch 2. In the post‐rigor state, switch 2 lies out of the plane of the b‐sheet (open) and in the pre‐powerstroke state switch 2 is in the plane of the b‐sheet (closed).

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

171

It is important to note that the binding of ATP places the gPi into the pocket formed by SW1 and the P‐loop, which promotes the moving in of SW2. The effects of this movement are twofold: the relay helix is kinked and thereby ‘‘re‐primed,’’ and the enzymatically active site is formed. In the pre‐powerstroke state, the nucleotide is completely enclosed by the positioning of SW1 and SW2. This conformation appears to be the enzymatically active form of myosin. Moreover, this is the preferred stable structure of myosin in the presence of nucleotide.

C. Structural Origin of the Powerstroke The position of the converter domain depends on whether the relay helix has a kink near its middle point. The kink in the relay helix occurs in the pre‐powerstroke state. The kink leads to a rotation of the converter domain through 60
. Removing the kink causes the lever arm to rotate back by 60
, which is the elementary structural event in the powerstroke. Conversely, creating the kink is the priming action necessary to reach the start of the powerstroke.

D. Rigor State: Actin Binding Closes the 50K Cleft Incubating actin filaments with isolated crossbridges without ATP produces the rigor complex known as decorated actin. Decorated actin provides a model system for studying the strong interaction between actin and myosin. Unfortunately, there are no crystallographic data on the actin–myosin complex. Cryoenergy‐filter electron microscopy has recently yielded a 14 A˚ resolution map of rabbit skeletal actin decorated with chicken skeletal S1. This can be interpreted by fitting known crystal structures. These studies show that the cleft in the actin‐binding site is closed on strong binding to actin (Holmes et al., 2003, 2004). On strong binding to actin the upper 50K domain moves to close the actin‐binding cleft and to open the nucleotide‐binding pocket. The closure appears to entail the movement of the upper 50K domain as a rigid body. The effects of cleft closure include the movement of SW1 away from the phosphate groups, which opens the nucleotide binding pocket. Conversely, the binding of ATP restores SW 1 to its initial position, thereby opening the actin‐binding cleft. The movement of the upper 50K domain therefore provides a geometrical explanation of the strong negative linkage between actin affinity and ATP affinity.

172

GEEVES AND HOLMES

E. Myosin V Provides a Model for the Rigor State Myosin V is an unconventional myosin involved in transporting vesicles along actin cables in the cell. Like myosin II, it is a two‐headed molecule, but it differs in having a long lever arm (6 IQ motifs) and a binding site for its cargo at the C‐terminus. Myosin V appears to proceed along the actin filament by a ‘‘hand‐over‐hand’’ mechanism (Purcell et al., 2002) so that at any one time at least one head is attached to the actin filament (it is a processive motor). Myosin V is kinetically tuned to allow movement along actin filaments: nucleotide‐free myosin V appears to be constitutively in the strong binding form (which is not the case for myosin II). Moreover, the X‐ray crystallographic structure of the nucleotide‐free myosin V crossbridge (a truncated construct) shows the cleft between the upper and lower 50K domains to be shut (Coureux et al., 2003). Figure 4 shows the changes that occur to the actin‐binding cleft in proceeding from the myosin V (rigor‐like) structure through the pre‐powerstroke state to the post‐rigor structure. When compared with post‐rigor or pre‐powerstroke states the structural effects of cleft closure appear to include the movement of SW1, which opens the nucleotide‐binding pocket, together with a twist of the central b‐sheet, which is associated with a large movement of the P‐loop that considerably modifies the nucleotide binding site. Partial closure of the actin‐binding cleft and a very similar twisting of the b‐sheet were also seen in the nucleotide‐free structure of Dictyostelium myosin II reported by Reubold et al. (2003). The myosin V atomic model can be fitted without deformation into the electron microscope three‐dimensional (3D) reconstruction of decorated actin (Holmes et al., 2004). For this and other

Fig. 4. The actin‐binding cleft between the upper (red) and lower (gray) 50K domains (orientation as in Fig. 5A). In A (rigor‐like), the cleft is shut. In B (pre‐powerstroke), the outer end of the cleft (that forms the actin‐binding site) is fully open, but the apex or inner end of the cleft (next to the nucleotide‐binding pocket; ATP is shown in B) is closed. This closure is brought about by the switch 2 element (SW2) being in the ‘‘closed’’ conformation. In C (post‐rigor), both the outer end and the inner end are open. SW2 is ‘‘open.’’ In A and B the dispositions of SW2 are similar, but not identical. We refer to them as closed 1 (C1) and closed 2 (C2), respectively.

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

173

reasons (Sweeney and Houdusse, 2004), myosin V appears to be structurally the strong binding form and may therefore be used as a model of myosin in the rigor complex (it is therefore referred to as rigor‐like). This allows a detailed description of the actin–myosin interface and also suggests a mechanism (by twisting the central b‐sheet) whereby actin binding may straighten the relay helix and bring about the powerstroke. Using the fit of myosin V into the cryoelectron microscope density as the basis of the rigor complex, by adding back the missing lever arm one can generate the model of the end‐of‐powerstroke actin–myosin complex shown in Fig. 5 (Holmes et al., 2004).

F. Twisting the b‐Sheet Takes the Kink Out of the Relay Helix The rigor‐like state (typified by nucleotide‐free myosin V) has a straight relay helix and the lever arm is in the end‐of‐powerstroke configuration, although the lower 50K domain and SW2 have not swung out as is typical for the post‐rigor state. In fact, the orientation of the lower 50K domain with respect to the proximal part of the central b‐sheet is similar to that found in the pre‐powerstroke state and for this reason we refer to it as closed. Coureux et al. (2004) point out that, although the geometry of the SW2 closed structure as found in the pre‐powerstroke state and the disposition of the lower 50K domain in the rigor‐like state are similar, the details are different. To make this distinction we refer to the ‘‘closed’’ form found in the rigor‐like state as C2 and the form found in the pre‐powerstroke state as C1. However, in the post‐powerstroke state, SW2 is open, which had been taken to mean that the opening of SW2 is strongly linked to the execution of the powerstroke (and loss of Pi). This appears to be incorrect. We believe it is most probable that SW2 remains closed during strong binding to actin (in the geometric sense of its position with respect to the proximal end of the b‐sheet). Its relationship to the gPi is, however, dramatically altered by the P‐loop swinging out from the position found in the pre‐powerstroke or post‐rigor structures; it is carried by the twist of the distal end of the b‐sheet into an open position (i.e., it is the P‐loop that ‘‘opens’’ rather than SW2). Myosin therefore appears to possess two mechanisms for relieving the kink in the relay helix: either by moving out the lower 50K domain (SW2 closed to SW2 open) or by twisting the b‐sheet, which gives the distal end of the relay helix more space. The first mechanism (in the reverse order, SW2 open to SW2 closed) occurs in the absence of actin after ATP binds in the nucleotide pocket. This is the ‘‘recovery stroke’’ leading to the priming of the powerstroke. However, when the myosin is bound to actin, it appears that only the second mechanism is relevant (indeed, it would be rather

174

GEEVES AND HOLMES

Fig. 5. An atomic model of the rigor or strong‐binding complex between the myosin crossbridge and actin. (A) Looking down the actin helix from the pointed end. (B) At right angles to A. The actin monomers are colored gray and gray‐blue. The myosin crossbridge colors are the same as in Fig. 2. Note that both the upper (red) and lower (gray) 50K domains make contact with actin. This model is based on the electron micrographs of decorated actin and the crystal structures of actin and myosin V. (See Holmes et al., 2004 for details.)

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

175

Fig. 6. The b‐sheet can twist. A comparison of the pre‐powerstroke state (A) with the rigor‐like state (B). For this comparison, strands 6 and 7 of the b‐sheet (shown edge on) have been aligned (this corresponds closely with an alignment of the lower 50K domains). The SW2 elements (light blue, below strand 5) in the two states are in very similar positions, as is the proximal part of the relay helix (light blue). In A, the outer part of the relay helix (dark blue) is kinked. In B, the relay helix is straight. A comparison of A and B shows that the distal end of the b‐sheet (strands 1–4) twists through 7–8
. Note that this leads to a displacement of the P‐loop.

difficult for SW2 to open when the lower 50K is bound to actin without disrupting the actin‐binding site). This second mechanism was only recently discovered for the nucleotide‐free forms of myosin V and for Dictyostelium myosin II (Coureux et al., 2003; Reubold et al., 2003). The mechanism is illustrated in Fig. 6. A further effect of the twist of the central b‐sheet is to displace the P‐loop into an open environment where nucleotide exchange would be rapid. Conversely, the rebinding of ATP would force the P‐loop back into the position found in the post‐rigor state and would untwist the b‐sheet. The implication is that actin binding drives the primed myosin crossbridge toward this structure—presumably as a result of closing the cleft and the subsequent loss of Pi.

G. A Model of the Strongly Bound Pre-Powerstroke The strongly bound pre‐powerstroke state or top‐of‐powerstroke state is the transitory state labeled 4 in Fig. 1. It is experimentally difficult to characterize this either kinetically or structurally. At present, the structure can only be guessed at by an extrapolation of the properties of the adjoining structures. It seems very likely that the actin‐binding cleft closes on strong binding in the top‐of‐powerstroke state. Comparison of the structures of the pre‐powerstroke and post‐rigor states with the nucleotide‐free

176

GEEVES AND HOLMES

Fig. 7. The strongly bound top‐of‐powerstroke state. Shown is the truncated myosin crossbridge without the lever arm. The orientation is as in Fig. 5A. (A) Pre‐powerstroke state with the upper 50K domain shown in yellow. (B) The rigor‐like state with the upper 50K domain of the pre‐powerstroke state (yellow) superimposed on the upper 50K domain of the rigor‐like state (red). (C) The model produced by taking the superimposed orientation of the upper 50K domain and combining it with the original pre‐powerstroke coordinates. This generates a pre‐powerstroke state with a shut actin‐ binding cleft that serves as a model of the ephemeral strongly bound top‐of‐powerstroke state.

Fig. 8. Switch 1 (SW1) opens on strong binding to actin. The orientation as in Fig. 5B, with the lower 50K domain and the lever arm omitted. (A) The weakly attached pre‐powerstroke crossbridge. (B) Model of the strongly attached top‐of‐powerstroke structure (for details see text). Note that the SW1 element has been pulled away from the environment of the phosphates by strong binding to actin.

myosin V shows that the upper 50K domain moves very much as a solid body (Fig. 7B). Thus a model may be generated by taking a pre‐powerstroke state structure (Fig. 7A) and superimposing the upper and lower 50K domains on those of the rigor‐like state. The result of this operation is shown in Figs. 7C and 8B. The most characteristic property of this model is the open SW1—the nucleotide‐binding pocket has been transformed into a nucleotide‐binding groove. This would certainly facilitate nucleotide exchange, although the structure gives little indication about the order of the release of products.

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

177

Fig. 9. The powerstroke. (A) A model of the strongly bound pre‐powerstroke state. (B) The rigor state (as in Fig. 5).

The rest of the nucleotide binding site, P‐loop and SW2, are intact. However, the subsequent twisting of the b‐sheet during the powerstroke moves the P‐loop and considerably modifies the nucleotide‐binding site. The modeled structure may also be used to generate the attached top‐ of‐powerstroke state. This is shown in Fig. 9A compared with the rigor conformation (Fig. 9B). The same geometry for the attachment to actin has been used, as was found by electron microscopy for the binding of myosin V to actin. The lever arm from chicken skeletal myosin has been used to complete the model.

H.

Classifying the Structures

Coureux et al. (2004) discuss the differences among rigor‐like, post‐ rigor, and pre‐powerstroke states in detail. Their conclusions are based on the crystallographic structures of myosin V with and without bound nucleotide. We use a somewhat simplified scheme to describe the differences and similarities among the three conformers of myosin that we have identified as being represented in the Lymn–Taylor cycle. These are summarized in Table I. Two conditions appear to be simultaneously necessary for attaining the pre‐powerstroke state: SW2 must be closed and the b‐sheet not twisted. Relaxing either of these conditions allows the relay helix to straighten (the converter domain is in the end of powerstroke conformation). Thus, both the rigor‐like and the post‐rigor structures have a straight relay helix. This ability of the myosin crossbridge to respond to two (or maybe more) input parameters has prompted Coureux et al. (2004) to refer to the b‐sheet as the transducer. According to their notation, it is distorted in rigor and

178

GEEVES AND HOLMES

relieved in post‐rigor, along with the other structural elements that allow this distortion. The transducer is the ‘‘central processor’’ of the motor. The relay/converter conformation is the readout of the result of these inputs. In going from the post‐rigor to pre‐power states, the only signal is SW2 going from open to closed, which produces the relay kink and the converter up. To go from the pre‐power to rigor states requires the b‐twist, which is triggered by actin binding and cleft closure.

IV.

Biochemistry and Kinetics

A. Structural Correlations of the Lymn–Taylor Cycle The apparent positions in the Lymn–Taylor cycle of the structural states that we refer to as the pre‐powerstroke, rigor‐like, and post‐powerstroke conformations are shown in Fig. 10. Thus it appears that three of the four

Fig. 10. The structural correlates of the Lymn–Taylor cycle. Numbering of states as in Fig. 1A. See Section IV for a detailed description.

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

179

important states in the crossbridge cycle are now known in atomic detail. A few crystal structures cannot be allocated to states in the Lymn–Taylor cycle. Our present estimate is that these structures do not represent conformations of the crossbridge found in the crossbridge cycle; their importance lies rather in demonstrating the rich repertoire of possible crossbridge conformations. A table of all known crystal structures and their properties can be found in Coureux et al. (2004; supplementary material). The fourth actin‐bound, top‐of‐powerstroke state is ephemeral. In the original Lymn–Taylor model it is not clear if this fourth state is strongly or weakly bound to actin, but Pi release, ADP release, and force generation all occur during the transition 4 to 1. Thus, in the following discussion, we attempt to break this transition down into a series of elementary events and to explore if it is possible to order the biochemical and mechanical events and to correlate them with structural changes. We first review the biochemical evidence for the events from state 1 to 3 (rigor to pre‐powerstroke) before considering the events associated with 3 to 1 (pre‐powerstroke to rigor).

B. From Rigor to Post‐Rigor: ATP Binding and Actin Binding to the Negatively Linked Crossbridge Starting from the rigor‐like conformation, ATP binds into the nucleotide pocket, which results in a dissociation of actin and myosin and formation of M.ATP, the post‐rigor conformation. Since the original work of Lymn and Taylor, it has been known that actin dissociation precedes ATP hydrolysis (in solution at high ATP concentrations). The recent data using the mutants of Dictyostelium myosin II also show that the relay helix/ converter ‘‘repriming’’ event (monitored by the tryptophan fluorescence on the relay loop) also occurs after actin dissociation (Malnasi‐Csizmadia et al., 2001; Zeng et al., 2004). Dissociation itself can be monitored via either light scattering (sensitive to the size of the actomyosin complex) or the fluorescence of a pyrene label attached to Cys 374 of actin (Criddle et al., 1985). Thus, the relay helix and the converter/lever‐arm movements associated with the closing of SW2 are not directly linked to the marked reduction of affinity for actin that accompanies ATP binding. It seems likely, therefore, that ATP binding to the nucleotide pocket causes SW1 to close, and this results in the opening of the 50K cleft and disruption of the actin/myosin interface. The binding of ATP will also result in the repositioning of the P‐loop into the ‘‘down’’ position. Is opening of the cleft and loss of the interaction between the upper 50K

180

GEEVES AND HOLMES

domain and actin sufficient to account for the major loss of actin affinity? This question has not been addressed directly, but the upper 50K actin‐ binding site does contain two regions of specific interest that interact with actin. The upper 50K domain carries a conserved arginine that is associated with a cardiomyopathy when mutated to a glutamine (R430Q) in human cardiac myosin (Geisterfer‐Lowrance et al., 1990). This region also contains the so‐called TEDS site, which in myosin II is a conserved negatively charged Asp or Glu residue (Bement and Mooseker, 1995; Brzeska and Korn, 1996). In several non‐muscle myosins a serine or threonine is found in this position and it must be phosphorylated to switch on the actomyosin ATPase activity and motility. Thus, there is a requirement for a negatively charged residue at this site and mutations to eliminate this negative charge result in loss of motility and reduced actin‐activated ATPase activity (De La Cruz et al., 2001; Fujita‐Becker et al., 2004). Where it has been measured, both the lack of phosphorylation at the TEDS site and the cardiomyopathy mutation result in reduced actin affinity in rigor conditions and a lower Km for actin. Thus, disruption of the upper 50K domain contact site with actin may be sufficient to induce dissociation of actin from myosin, but additional effects of ATP on the lower 50K — actin interaction cannot be ruled out. Studies of the pyrene label on actin indicate that the ATP‐induced dissociation of actin occurs via a three‐step process (steps 10 , 20 , and c, Scheme 1; Millar and Geeves, 1983). Initial fast equilibrium binding of ATP (controlled by K10 and probably diffusion limited) is followed by an isomerization that alters the environment of the pyrene label on actin. This isomerization is fast in myosin II and much slower for some non‐muscle myosins (~500 s–1 for Dictyostelium myosin II; Kurzawa et al., 1997; >1000 s–1 for rabbit skeletal myosin II, Geeves and Jeffries, 1988; 100 s–1 for myosin 1b at 20
C; Coluccio and Geeves, 1999) and is accompanied by a very rapid

Scheme 1.

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

181

dissociation of actin at >1000 s–1, since no lag between the light scattering and pyrene‐actin fluorescence signals is observed. Probes in the nucleotide binding pocket (methylanthranilolyl derivatives of ATP [mant‐ATP], Woodward et al., 1991; single tryptophan, Zeng et al., 2004) suggest that tight ATP binding and the pyrene label are perturbed at the same time. This is compatible with ATP‐binding inducing closure of SW1, to give tightly bound ATP, which simultaneously opens the 50k cleft, thereby weakening actin affinity and perturbing the pyrene signal on actin. Some local changes in the nucleotide binding pocket may occur after actin dissociates. The repriming of the converter/relay helix associated with the perturbation of W501 occurs after actin has dissociated (Zeng et al., 2004). If we assume that the M#ATP conformation (post‐rigor) is the same whether formed by ATP binding to M or to A.M, then we can define the equilibrium constants in scheme 1 and therefore the free energy associated with the disruption of the upper 50K binding domain. The affinity of actin for myosin in rabbit muscle myosin II (Ka in Scheme 1), is of the order of 108 M–1 (Kurzawa and Geeves, 1996) and this falls to 104–103 M–1 when ATP is tightly bound (Kc; Millar and Geeves, 1983). The binding of ATP to form M.T and A.M.T is a relatively weak, rapid equilibrium, reaction with an equilibrium constant of the order of 103 M–1 in both cases (K1 and K10 ¼ 103 M–1; Bagshaw and Trentham, 1974; Millar and Geeves, 1983). Therefore, detailed balance requires that Kb ¼ Ka ¼ 108 M–1. ATP binding to myosin is known to be very tight and almost irreversible (~10–11 M; Trentham et al., 1976) and therefore K2 is approximately 108. Detailed balance then requires that K20 ¼ 103–104 and the free energy change on this actomyosin isomerisation step is 3–4  2.303RT kJ mol–1 (where ‘R’ is the gas constant and ‘T’ is the temperature). The logarithm of the ratio of K2/K20 (K2/K20 ¼ 104–105) defines the energy used to break the A.M contact and open the 50K cleft.

C.

From Post‐Rigor to Pre‐Powerstroke

The events associated with SW2 closure followed by ATP hydrolysis have been described above. The idea originally explored in 1999 (Geeves and Holmes, 1999), that SW2 must close onto the gPi of ATP before ATP hydrolysis takes place, has been elegantly demonstrated for both Dictyostelium myosin II and rabbit fast muscle myosin II (Malnasi‐Csizmadia et al., 2001; Urbanke and Wray, 2001). The kinetic and spectroscopic studies are all compatible with the post‐rigor and pre‐powerstroke crystal structures. The details of the events are worth setting out for what they reveal about the nature of the conformational changes within the myosin motor

182

GEEVES AND HOLMES

domain. For the Dictyostelium myosin, SW2 closure is a rapid (~1000 s–1) event with an equilibrium constant close to 1 (
Go ~ 0 kJ mol–1) and is associated with a change in the environment of the W501 on the end of the relay helix (Malnasi‐Csizmadia et al., 2001). Thus, SW2 is rapidly and reversibly exploring the nucleotide pocket. If it finds the gPi of ATP present, SW2‐closed is stabilized (by the H‐bond to the gPi and the salt bridge between SW1 and SW2). Cleavage of the b‐g Pi bond of ATP follows within 30 ms for Dictyostelium myosin II or 6 ms for rabbit skeletal muscle myosin II. ATP hydrolysis stabilizes the closed conformation of SW2, but not to a great extent (Khydrolysis ~ 10). However, the cleaved Pi is tightly bound with an average lifetime before escape (if there is no interaction with actin) of 20 s (kþ4 ¼ 0.05 s–1; Trentham et al. 1976). Apparently a relatively rare structural fluctuation is required to allow Pi to dissociate. Once released, Pi does not readily rebind to the myosin‐ADP complex (KPi ¼ 0.1 M). In the absence of actin it appears that SW2 closure occurs at the same time as the relay‐helix/converter/lever‐arm repriming so that the entire conformational change occurs on the sub‐milliscecond time scale and can be monitored through changes in fluorescence from W501 on the relay helix. Thus, all of the ‘‘repriming’’ events occur rapidly and reversibly and result in a cleaved ATP with a tightly bound Pi. Thus, myosin has no strong preference between the post‐rigor or pre‐powerstroke conformations and the two conformations are constantly being sampled. This may mean that the conformation of each of the mobile elements of the structure cannot be precisely defined for either the post‐rigor or the pre powerstroke state. The crystal structures are well defined because the crystal packing is unlikely to allow the converter/relay‐helix transition to take place and therefore the associated changes in SW2, the b‐sheet, and the P‐loop may also be constrained. The use of fluorescence lifetime studies and spin probes shows that locations around the nucleotide pocket are highly dynamic in solution (Malnasi‐Csizmadia et al., 2001; Ostap et al., 1995). However, the fact that the free energy differences between the conformations are very small means that the introduction of probes is likely to affect the equilibrium between conformers.

D.

From Pre‐Powerstroke to Powerstroke: Pi Release

This is the least well-defined of the events in the crossbridge cycle, but the one where the most interesting events take place—the generation of force and movement. The overall scenario is generally agreed, but there remains a wide range of opinions on the details. The essential events are initiated by binding to actin. The binding of actin to myosin

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

183

Scheme 2.

in solution has been well described and consists of three major events with the affinity of myosin for actin increasing in each step (Scheme 2) (Geeves and Conibear, 1995). The initial diffusion‐limited complex formation (step I) has a marked dependence on ionic strength and is probably driven by long‐range ionic interactions. This is followed by an isomerization that has little ionic strength dependence but can be inhibited by small concentrations of organic solvent and therefore could involve the formation of stereo‐specific hydrophobic interactions. The third step is particularly sensitive to the presence of nucleotide in the myosin pocket and is accompanied by a large volume change in the complex indicative of the displacement of a large volume of bound water from the actomyosin complex. These three events are compatible with step I being driven through the positively charged loop 2 (connecting the U50 and L50K domains), interacting with the negative charge of the N‐terminus of actin. The second step is associated with a predominantly hydrophobic interaction between actin and the lower 50K domain, and the third step with cleft closure (sensitive to the presence of nucleotide) and formation of additional interactions between actin and the upper 50K domain. These events are likely to have equivalent processes in the pre‐powerstroke to rigor transition, but in this case they are coupled with product release and the powerstroke. Starting from the pre‐powerstroke state myosin complex with ADP and Pi tightly bound (summarized structure in Table I), the M.D.Pi is in rapid equilibrium with an actin‐bound state on the microsecond–millisecond time scale. This is very dependent on ionic strength (Furch et al., 2000; White and Taylor, 1976) and is therefore probably a non‐stereo‐specific weak binding state and is controlled by the ionic interactions between loop 2 and the N‐terminus of actin. Other ionic interactions may also be involved. This loose association between actin and myosin probably does not alter the overall conformation of myosin. The next stage involves the formation of a stereo‐specific interaction via the lower 50K domain and is dominated by hydrophobic interactions as described above for step II. Formation of the initial stereo‐specific interaction is quickly followed by closure of the cleft and formation of the complete actomyosin contact surface (strongly bound rigor‐like actin–myosin interface), which is equivalent to step III above. This has been modeled above in Figs. 7 through 9.

184

GEEVES AND HOLMES

The cleft is probably dynamic on the microsecond time scale and in the absence of actin (or in the weakly actin‐bound conformation with a stereo‐ specific interaction with the lower 50K domain), the cleft may be rapidly opening and closing with an equilibrium strongly favoring the open conformation by 100:1 (influenced by the presence of gPi). Formation of the second stereo‐specific interaction with the upper 50K domain stabilizes the closed conformation. The extent to which the closed cleft is stabilized may depend on what happens in the rest of the myosin molecule (in particular, the fate of the gPi). This rapid opening and closure of the cleft and formation of the second actin interaction site may be analogous to the opening and closure of SW2 in the M.ATP/M.ADP.Pi post‐rigor and pre‐powerstroke complexes. The cleft is constantly exploring the closed conformation (similar to the way SW2 explores its open and closed forms). In the case of the 50K cleft, the presence of gPi is incompatible with the cleft being in a stable, closed structure. Equally, the closed cleft destabilizes the gPi‐binding site (and ADP binding) via changes in SW1 and the P loop. If Pi dissociates, then the closed cleft is stable. While the cleft remains open, the bound gPi is stable. Consider the following potential sequence of events (Scheme 3): State 1. There is a weak actin–myosin attachment (through the lower 50K domain), which is in rapid equilibrium with the detached crossbridge. It is also in rapid equilibrium with State 2, in which the cleft closes (an equilibrium constant of 0.1–1.0). State 2. The 50K cleft is closed, the b sheet distorted. This is the strained complex. The strain can be dissipated by a step back to state 1 or by P‐loop movement and loss of Pi via Steps II and III. States 3 and 4. In States 3 and 4 the cleft closure and b‐sheet distortion promote P‐loop, SW1 and SW2 changes, but if the lever‐arm cannot complete its movement because of an attached load, then a strained state remains. In this case, all States 2 through 4 are strain‐holding complexes and the occupancy of these states is a function of Pi concentration. At very low Pi (<100 mM) State 4 predominates; at Pi > 20 mM, States 1 and 2 predominate. If we assume that Step III is a reversal of a simple ligand‐(Pi) binding event with no associated conformational change, then this step is not strain dependent. But transitions 1‐ to -2, 2‐ to -3 and the step allowing ADP

Scheme 3.

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

185

release are all strain dependent and we cannot define the equilibrium constant for each step without defining the strain dependence of each step (for a recent discussion, see Smith and Sleep, 2004). Under isometric load, the apparent equilibrium constants between States 1 and 2 and between States 2 and 3 may be close to a value of unity. Studies of Pi‐induced tension transients suggest that, in an isometric muscle fiber, the equilibrium constant between States 2 and 3 has a value of ~0.1, but this may reflect a subset of crossbridges bearing the highest load. For the outline in Scheme 3, two situations can be considered: the isometric case where a large load attached to the myosin lever‐arm prevents the lever‐arm from completing its swing, and the unloaded situation as seen in solution for a single motor domain.

E. Unloaded Powerstroke: Acto‐S1 in Solution At high actin concentrations in solution, actin binding to M.ADP.Pi is rapid and complete. Actin binding would induce the 50K cleft closure, leading to the opening of SW1 and movement of the P‐loop, allowing Pi escape. The loss of Pi is associated with relieving the strain in the upper 50K b‐sheet and the swing of the converter and lever‐arm. All of the events following Pi release are rapid compared to Pi release (or the conformational change allowing Pi release), and it is not currently possible to define the exact sequence of events. It is only possible to speculate about which events are required to allow Pi to escape and which ones may follow. Information from probes such as the pyrene on actin (cys 374), the mant group on the ribose of ADP, and W501 seem to change at the same time as Pi release (~100 s–1; White et al., 1997). If the pyrene‐actin and mant‐ADP signal changes are linked to the 50k cleft closing and SW1 opening, then this could be interpreted as cleft closure limiting the rate of Pi release. However, as discussed in Scheme 3 above, an alternative possibility is that the cleft conformation is unstable (small Keq) and, therefore, incomplete until Pi escapes. Thus, stable cleft closure and Pi escape occur at the same time, but Pi release is induced by cleft closure. Similarly, because Pi escape and the change in the W501 environment occur at the same time, it is not possible to define which is the cause and which the result. The structural question remains: What does cleft closure does to allow Pi escape? A movement of the P‐loop and SW1 would be sufficient to destabilize Pi binding. However, an obvious exit route for Pi from the bottom of the nucleotide pocket is not apparent from current structures. Pi release is, however, relatively irreversible for the unloaded crossbridge. It therefore remains possible that an exit route for Pi from the A.M.ADP.Pi state may only be available transiently. Once the lever‐arm and converter complete

186

GEEVES AND HOLMES

their 60
swing, the route for Pi to rebind may be inaccessible. This is different from the isometric case.

F.

Isometric

In the totally isometric case, very few results are available from probes around the nucleotide‐binding site, but data available from mechanical experiments that allow some correlation between the release of Pi and the development of force and movement. For a crossbridge under load, the release of Pi is a reversible step and therefore the strained A.M.D state must allow reversible binding of Pi (Dantzig et al., 1992; Fortune et al., 1991; Kawai and Halvorson, 1991; Ranatunga et al., 2002; Tesi et al., 2000). This reversible escape of Pi becomes almost irreversible once the load as dissipated. Pi rebinding is associated with a loss of force at ~100 s–1 (Dantzig et al., 1992; Fortune et al., 1991; Kawai and Halvorson, 1991; Ranatunga et al., 2002; Tesi et al., 2000). Thus, Pi rebinding to a load‐bearing A.M.D state has been interpreted as a reversal of the powerstroke, resulting in detachment of the crossbridge. Isotope exchange studies suggest that the rebinding of Pi to the crossbridge allows the reformation of myosin‐bound ATP so that the energy associated with ATP hydrolysis is conserved (Sleep and Hutton, 1980). Thus the isometric case starts with the A‐M.D.Pi state and the lower 50K cleft bound to actin. Closure of the cleft is fast and results in a distortion of the b‐sheet of the upper 50K domain. This tends to destabilize the P‐loop, leading to Pi release, but, if the movement of the lever‐arm and converter is not complete, then the whole process is not completed and at mM Pi concentration Pi can rebind, leading to the loss of force and detachment. When the crossbridge is held isometric, all steps in Scheme 3 are in a quasi‐stable dynamic equilibrium. If State 2 has a relatively low occupancy, the question of whether force is developed before or after Pi release may not be very meaningful. The interpretation will depend on how Pi release and force are measured and whether State 2 is sampled in the measurements. The loss of force through Pi rebinding is due to formation of State 1 as much as to formation of State 2. There is a general problem in defining a structural change into a strained state. Such a structural change may only involve changes in the position of a small number of residues by a few tenths of an angstrom and will therefore remain undetected by current structural analysis. Alternatively, the strained state may represent a quasi‐equilibrium between two or more conformations, none of which is stable under the conditions discussed here. In the absence of load, the transition between the states can

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

187

be defined in terms of rate and equilibrium constants, but for a protein conformational change the free energy landscape between two conformations can be very complex. In the absence of load, the lowest free energy mean path may be uniquely defined, but the application of a distorting force with a particular direction may change that landscape significantly. The lowest free energy path under zero load may then no longer be a reliable guide to the path under a load.

G.

ADP Release and Back to Rigor

Completion of the cycle back to the rigor conformation requires the release of ADP. The ADP release step has come into focus in recent years for a combination of reasons. In the past few years, it has become clear that the rate of ADP release is important for defining how fast a muscle can contract (Pereira et al., 2001; Siemankowski et al., 1985; Weiss et al., 2001), for defining the ability of a myosin to sustain a load with low ATP turnover (Cremo and Geeves, 1998), and in the coordination of the two heads of a processive myosin such as myosin V (Coureux et al., 2003; Mehta et al., 1999). These properties are linked by the concept that the efficient coupling of the motor activity of a myosin to the rate of ATP breakdown (and the Fenn effect; Fenn, 1923) requires that the myosin remains bound to actin until the myosin completes its working stroke and force is dissipated. This can be achieved by having strain‐sensitive ADP release, strain‐sensitive ATP binding, or both. The observation that ADP release from certain types of myosin is associated with an additional movement of the myosin tail (in the same direction as the powerstroke) has given a low‐resolution structural interpretation of such a strain‐dependent ADP‐ release mechanism ( Jontes et al., 1995; Veigel et al., 1999). The biochemical evidence has been reviewed recently and is not be repeated here (Nyitrai and Geeves, 2004). To date, crystal structures have given few indications of the structural details of an acto–myosin complex with a tightly bound ADP. Such a complex may not exist in the absence of actin. Moreover, in the presence of actin, it would only be observed under mechanical load. Analogies with the G‐proteins have led to suggestions that disruption of Mg coordination may be the signal for ADP release (Fujita‐Becker et al., 2004; Rosenfeld et al., 2004), but the triggering structural change remains to be identified. There is evidence that ADP binding is sensitive to the nature of loop 1 near the entrance to the nucleotide binding pocket (Kurzawa‐Goertz et al., 1998; Murphy and Spudich, 1998; Sweeney et al., 1998), to the conformation of the converter and lever‐arm (Veigel et al., 1999, 2003; Whittaker et al., 1995), and to the structure of the SH1‐SH2 helices (Batra and Manstein, 1999). It is therefore possible that the

188

GEEVES AND HOLMES

nucleotide‐binding pocket integrates a series of structural/mechanical signals before allowing release of ADP.

V. Summary A.

The Relay Helix Kink

The position of the converter domain depends on whether the relay helix has a kink at its middle point. The kink in the relay helix occurs (exclusively) in the pre‐powerstroke state. The kink leads to a rotation of the converter domain through 60
. Removing the kink causes the lever arm to rotate back by 60
, which is the elementary structural event in the powerstroke. Conversely, creating the kink is the priming action to come to the start of the powerstroke. The mechanism by which actin binding brings about the straightening of the kinked relay helix becomes a central issue in understanding muscle contraction. In going from the pre‐power to post‐rigor structure (the recovery stroke), SW2 moves out and the relay helix straightens. In going from the pre‐ powerstroke to rigor‐like states (power stroke), SW2 stays roughly where it is and the distal part of the b‐sheet rotates counterclockwise, thereby relieving pressure on the relay helix and allowing it to straighten.

B. Actin Binding Appears to Drive the Powerstroke via a b‐Sheet Twist It appears that during strong binding the straightening of the relay helix (removal of the kink) comes about through the twisting of the b‐sheet rather than a substantial movement of SW2. If the cleft closure at the actin‐ binding‐site causes the twisting of the b‐sheet, this suggests a mechanism for how actin binding drives the powerstroke. It seems likely that strong binding to actin causes the b‐sheet to twist, which in turn brings about the rotation of the converter domain by allowing the relay helix to straighten. From physiological experiments it is known that force production happens on strong binding to actin before phosphate or ADP are released. The structural implication of this experiment is that the b‐sheet is put under strain by strong binding. In solution this would lead to a rapid transition to the rigor state (limited by the rates of release of products). However, in an isometric muscle the crossbridge is tethered and the b‐sheet cannot twist, so that no substantial movement of the P‐loop takes place. Thus we begin to have a geometrical explanation for the dependence of the rate of phosphate release on the tension on the crossbridge.

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

C.

189

Phosphate Release

The structural changes associated with Pi release remain the key to understanding the powerstroke of the crossbridge cycle. The first event is a structural change induced by actin binding to myosin that promotes Pi release. This may be force‐producing, but while Pi release and rebinding are reversible, the process is reversible and will not lead to the production of net external work. The second is the structural change associated with completion of the working stroke that changes Pi release from being reversible to almost irreversible. This ‘‘commits’’ the crossbridge to turning chemical energy into mechanical work. We have tried to show how the current structural, biochemical, and mechanical evidence can be used to provide an outline of how actin binding, Pi release, and the powerstroke may be coupled. However, concrete evidence remains in short supply. The structural states will be hard to define if they are coupled to equilibrium constants with values close to unity and fast rate constants (100–1000 s–1). Only in the presence of a large load, which brings the system close to an isometric condition, will the structural states have significant occupancy during a steady state. The use of stalled processive motors such as myosins V or VI may be one way of trapping the relevant structures.

References Bagshaw, C. R., and Trentham, D. R. (1974). The characterization of myosin‐product complexes and of product‐release steps during the magnesium ion‐dependent adenosine triphosphatase reaction. Biochem. J. 141, 331–349. Batra, R., and Manstein, D. J. (1999). Functional characterisation of Dictyostelium myosin II with conserved tryptophanyl residue 501 mutated to tyrosine. Biol. Chem. 380, 1017–1023. Bauer, C. B., Holden, H. M., Thoden, J. B., Smith, R, and Rayment, I. (2000). X‐ray structures of the apo and MgATP‐bound states of Dictyostelium discoideum myosin motor domain. J. Biol. Chem. 275, 38494–38499. Bement, W. M., and Mooseker, M. S. (1995). TEDS rule: A molecular rationale for differential regulation of myosins by phosphorylation of the heavy chain head. Cell Motil. Cytoskeleton 31, 87–92. Brown, J. H., and Cohen, C. (2005). Regulation of muscle contraction by tropomyosin and troponin: How structure illuminates function. Adv. Protein Chem. 71, 121–159. Brzeska, H., and Korn, E. D. (1996). Regulation of class I and class II myosins by heavy chain phosphorylation. J. Biol. Chem. 271, 16983–16986. Coluccio, L. M., and Geeves, M. A. (1999). Transient kinetic analysis of the 130‐kDa myosin I (MYR-1 gene product) from rat liver. A myosin I designed for maintenance of tension? J. Biol. Chem. 274, 21575–21580. Cooke, R. (1986). The mechanism of muscle contraction. CRC Crit. Rev. Biochem. 21, 53–118.

190

GEEVES AND HOLMES

Coureux, P. D., Sweeney, H. L., and Houdusse, A. (2004). Three myosin V structures delineate essential features of chemo‐mechanical transduction. EMBO J. 23, 4527–4537. Coureux, P. D., Wells, A. L., Menetrey, J., Yengo, C. M., Morris, C. A., Sweeney, H. L., and Houdusse, A. (2003). A structural state of the myosin V motor without bound nucleotide. Nature 425, 419–423. Cremo, C. R., and Geeves, M. A. (1998). Interaction of actin and ADP with the head domain of smooth muscle myosin: Implications for strain‐dependent ADP release in smooth muscle. Biochemistry 37, 1969–1978. Criddle, A. H., Geeves, M. A., and Jeffries, T. (1985). The use of actin labeled with N‐(1‐ pyrenyl)iodoacetamide to study the interaction of actin with myosin subfragments and troponin/tropomyosin. Biochem. J. 232, 343–349. Dantzig, J. A., Goldman, Y. E., Millar, N. C., Lacktis, J., and Homsher, E. (1992). Reversal of the crossbridge force‐generating transition by photogeneration of phosphate in rabbit psoas muscle fibres. J. Physiol. 451, 247–278. De La Cruz, E. M., Ostap, E. M., and Sweeney, E. M. (2001). Kinetic mechanism and regulation of myosin VI. J. Biol. Chem. 276, 32373–32381. Dominguez, R., Freyzon, Y., Trybus, K. M., and Cohen, C. (1998). Crystal structure of a vertebrate smooth muscle myosin motor domain and its complex with the essential light chain: Visualization of the pre‐powerstroke state. Cell 94, 559–571. Fenn, W. O. (1923). A quantitative comparison between the energy liberated and the work performed by the isolated sartorius muscle of the frog. J. Physiol. 58, 175–203. Fortune, N. S., Geeves, M. A., and Ranatunga, K. W. (1991). Tension responses to rapid pressure release in glycerinated rabbit muscle fibers. Proc. Natl. Acad. Sci. USA 88, 7323–7327. Fujita‐Becker, S., Durrwang, U., Erent, M., Clark, R. J., Geeves, M. A., and Manstein, D. J. (2005). Changes in Mg2þ‐ion concentration and heavy chain phosphorylation regulate the motor activity of a class‐I myosin. J. Biol. Chem. 280, 6064–6071. Furch, M., Remmel, B., Geeves, M. A., and Manstein, D. J. (2000). Stabilization of the actomyosin complex by negative charges on myosin. Biochemistry 39, 11602–11608. Geeves, M. A., and Conibear, P. B. (1995). The role of three‐state docking of myosin S1 with actin in force generation. Biophys. J. 68, 194S–199S; discussion 199S–201S. Geeves, M. A., and Holmes, K. C. (1999). Structural mechanism of muscle contraction. Annu. Rev. Biochem. 68, 687–728. Geeves, M. A., and Jeffries, T. E. (1988). The effect of nucleotide upon a specific isomerization of actomyosin subfragment 1. Biochem. J. 256, 41–46. Geisterfer‐Lowrance, A. A., Kass, S., Tanigawa, G., Vosberg, H. P., McKenna, W., Seidman, C. E., and Seidman, J. G. (1990). A molecular basis for familial hypertrophic cardiomyopathy: A beta cardiac myosin heavy chain gene missense mutation. Cell 62, 999–1006. Holmes, K. C., Angert, I., Kull, F. J., Jahn, W., and Schroder, R. R. (2003). Electron cryo‐ microscopy shows how strong binding of myosin to actin releases nucleotide. Nature 425, 423–427. Holmes, K. C., Schro¨ der, R. R., Sweeney, H. L., and Houdusse, A. (2004). The structure of the rigor complex and its implications for the powerstroke. Phil. Trans. R. Soc. B 359, 1819–1828.

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

191

Houdusse, A., Kalabokis, V. N., Himmel, D., Szent‐Gyorgyi, A. G., and Cohen, C. (1999). Atomic structure of scallop myosin subfragment S1 complexed with MgADP: A novel conformation of the myosin head. Cell 97, 459–470. Jontes, J. D., Wilson‐Kubalek, E. M., and Milligan, V. N. (1995). A 32 degree tail swing in brush border myosin I on ADP release. Nature 378, 751–753. Kawai, M., and Halvorson, H. R. (1991). Two step mechanism of phosphate release and the mechanism of force generation in chemically skinned fibers of rabbit psoas muscle. Biophys. J. 59, 329–342. Ko¨ hler, D., Ruff, C., Meyho¨ fer, E., and Ba¨ hler, M. (2003). Different degrees of lever arm rotation control myosin step size. J. Cell Biol. 161, 237–241. Kovacs, M., Malnasi‐Csizmadia, A., Woolley, R. J., and Bagshaw, R. J. (2002). Analysis of nucleotide binding to Dictyostelium myosin II motor domains containing a single tryptophan near the active site. J. Biol. Chem. 277, 28459–28467. Kurzawa, S. E., and Geeves, M. A. (1996). A novel stopped‐flow method for measuring the affinity of actin for myosin head fragments using microgram quantities of protein. J. Muscle Res. Cell Motil. 17, 669–676. Kurzawa, S. E., Manstein, D. J., and Geeves, M. A. (1997). Dictyostelium discoideum myosin II: Characterization of functional myosin motor fragments. Biochemistry 36, 317–323. Kurzawa‐Goertz, S. E., Perreault‐Micale, C. L., Trybus, K. M., Szent‐Gyorgyi, A. G., and Geeves, M. A. (1998). Loop I can modulate ADP affinity, ATPase activity, and motility of different scallop myosins. Transient kinetic analysis of S1 isoforms. Biochemistry 37, 7517–7525. Lymn, R. W., and Taylor, E. W. (1971). Mechanism of adenosine triphosphate hydrolysis by actomyosin. Biochemistry 10, 4617–4624. Malnasi‐Csizmadia, A., Kovacs, M., Woolley, R. J., Botchway, S. W., and Bagshaw, C. R. (2001). The dynamics of the relay loop tryptophan residue in the Dictyostelium myosin motor domain and the origin of spectroscopic signals. J. Biol. Chem. 276, 19483–19490. Malnasi‐Csizmadia, A., Pearson, D. S., Kovacs, M., Woolley, R. J., Geeves, M. A., and Bagshaw, C. R. (2001). Kinetic resolution of a conformational transition and the ATP hydrolysis step using relaxation methods with a Dictyostelium myosin ii mutant containing a single tryptophan residue. Biochemistry 40, 12727–12737. Manstein, D. J., Ruppel, K. M., and Spudich, J. A. (1989). Expression and characterization of a functional myosin head fragment in Dictyostelium discoideum. Science 246, 656–658. Margossian, S. S., and Lowey, S. (1973a). Substructure of the myosin molecule. III. Preparation of single‐headed derivatives of myosin. J. Mol. Biol. 74, 301–311. Margossian, S. S., and Lowey, S. (1973b). Substructure of the myosin molecule. IV. Interactions of myosin and its subfragments with adenosine triphosphate and actin. J. Mol. Biol. 74, 313–330. Mehta, A. D., Rock, R. S., Rief, M., Spudich, J. A., Mooseker, M. S., and Cheney, R. E. (1999). Myosin‐V is a processive actin‐based motor. Nature 400, 590–593. Millar, N. C., and Geeves, M. A. (1983). The limiting rate of the ATP‐mediated dissociation of actin from rabbit skeletal muscle myosin subfragment 1. FEBS Lett. 160, 141–148. Murphy, C. T., and Spudich, J. A. (1998). Dictyostelium myosin 25–50K loop substitutions specifically affect ADP release rates. Biochemistry 37, 6738–6744. Nyitrai, M., and Geeves, M. (2004). ADP and strain sensitivity in myosin motors. Phil. Trans Roy. Soc. B 359, 1867–1877.

192

GEEVES AND HOLMES

Onishi, H., Maeda, K., Maeda, Y., Inoue, A., and Fujiwara, K. (1995). Functional chicken gizzard heavy meromyosin expression in and purification from baculovirus‐infected insect cells. Proc. Natl. Acad. Sci. USA 92, 704–708. [Published erratum appears in Proc. Natl. Acad. Sci. USA 1995: 92, 3076]. Ostap, E. M., Barnett, V. A., and Thomas, D. D. (1995). Resolution of three structural states of spin‐labeled myosin in contracting muscle. Biophys. J. 69, 177–188. Pato, M. D., Sellers, J. R, Preston, Y. A., Harvey, E. V., and Adelstein, R. S. (1996). Baculovirus expression of chicken nonmuscle heavy meromyosin II‐B. Characterization of alternatively spliced isoforms. J. Biol. Chem. 271, 2689–2695. Pereira, J. S., Pavlov, D., Nili, M., Greaser, M., Homsher, E., and Moss, R.‐L. (2001). Kinetic differences in cardiac myosins with identical loop 1 sequences. J. Biol. Chem. 276, 4409–4415. Purcell, T. J., Morris, C., Spudich, J. A., and Sweeney, H. L. (2002). Role of the lever arm in the processive stepping of myosin V. Proc. Natl. Acad. Sci. USA 99, 14159–14164. Ranatunga, K. W., Coupland, M. E., and Mutungi, G. (2002). An asymmetry in the phosphate dependence of tension transients induced by length perturbation in mammalian (rabbit psoas) muscle fibres. J. Physiol. 542, 899–910. Rayment, I., Holden, H. M., Whittaker, M., Yohn, C. B., Lorenz, M., Holmes, K. C., and Milligan, R. A. (1993b). Structure of the actin‐myosin complex and its implications for muscle contraction. Science 261, 58–65. Rayment, I., Rypniewski, W. R., Schmidt‐Base, K., Smith, R., Tomchick, D. R., Benning, M. M., Winkelmann, D. A., Wesenberg, G., and Holden, H. M. (1993a). Three‐ dimensional structure of myosin subfragment-1: A molecular motor. Science 261, 50–58. Reubold, T., Eschenburg, F. S., Becker, A., Kull, F. J., and Manstein, D. J. (2003). A structural model for actin‐induced nucleotide release in myosin. Nat. Struct. Biol. 10, 826–830. Rosenfeld, S. S., Houdusse, A., and Sweeney, H. L. (2005). Magnesium regulates ADP dissociation from myosin V. J. Biol. Chem. 280, 6072–6079. Siemankowski, R. F., Wiseman, M. O., and White, H. W. (1985). ADP dissociation from actomyosin subfragment 1 is sufficiently slow to limit the unloaded shortening velocity in vertebrate muscle. Proc. Natl. Acad. Sci. USA 82, 658–662. Sleep, J. A., and Hutton, R. L. (1980). Exchange between inorganic phosphate and adenosine 50 ‐triphosphate in the medium by actomyosin subfragment 1. Biochemistry 19, 1276–1283. Smith, C. A., and Rayment, I. (1996). X‐ray structure of the magnesium(II). ADP. vanadate complex of the Dictyostelium discoideum myosin motor domain to 1.9 A resolution. Biochemistry 35, 5404–5417. Smith, D. A., and Sleep, J. (2004). Mechanokinetics of rapid tension recovery in muscle: The myosin working stroke is followed by a slower release of phosphate. Biophys. J. 87, 442–456. Squire, J. M., Al‐Khayat, H. A., Knupp, C., and Luther, P. K. (2005). Molecular architecture in muscle contractile assemblies. Adv. Protein Chem. 71, 17–87. Squire, J. M., and Parry, D. A. D. (2005). Comparative motile mechanisms in cells. Adv. Protein Chem. 71, 1–15. Sweeney, H. L., and Houdusse, A. (2004). The motor mechanism of myosin V: Insights for muscle contraction. Phil. Trans. R. Soc. B 359, 1829–1841. Sweeney, H. L., Rosenfeld, S. S., Brown, F., Faust, L., Smith, J., Xing, J., Stein, L. A., and Sellers., J. R. (1998). Kinetic tuning of myosin via a flexible loop adjacent to the nucleotide binding pocket. J. Biol. Chem. 273, 6262–6270.

THE MOLECULAR MECHANISM OF MUSCLE CONTRACTION

193

Tesi, C., Colomo, F., Nencini, S., Piroddi, N., and Poggesi, C. (2000). The effect of inorganic phosphate on force generation in single myofibrils from rabbit skeletal muscle. Biophys. J. 78, 3081–3092. Trentham, D. R., Eccleston, J. F., and Bagshaw, C. R. (1976). Kinetic analysis of ATPase mechanisms. Q. Rev. Biophys. 9, 217–281. Urbanke, C., and Wray, J. (2001). A fluorescence temperature‐jump study of conformational transitions in myosin subfragment 1. Biochem. J. 358, 165–173. Uyeda, T. Q., Abramson, P. D., and Spudich, J. A. (1996). The neck region of the myosin motor domain acts as a lever arm to generate movement. Proc. Natl. Acad. Sci. USA 93, 4459–4464. Veigel, C., Coluccio, L. M., Jontes, J. D., Sparrow, J. C., Milligan, R. A., and Molloy, J. E. (1999). The motor protein myosin‐I produces its working stroke in two steps. Nature 398, 530–533. Veigel, C., Molloy, J. E., Schmitz, S., and Kendrick‐Jones, J. (2003). Load‐dependent kinetics of force production by smooth muscle myosin measured with optical tweezers. Nat. Cell Biol. 5, 980–986. Warshaw, D. M., Guilford, W. H., Freyzon, Y., Krementsova, E., Palmiter, K. A., Tyska, M. J., Baker, J. E., and Trybus, K. M. (2000). The light chain binding domain of expressed smooth muscle heavy meromyosin acts as a mechanical lever. J. Biol. Chem. 275, 37167–37172. Weiss, S., Rossi, R., Pellegrino, M. A., Bottinelli, R., and Geeves, M. A. (2001). Differing ADP release rates from myosin heavy chain isoforms define the shortening velocity of skeletal muscle fibers. J. Biol. Chem. 276, 45902–45908. White, H. D., Belknap, B., and Webb, M. R. (1997). Kinetics of nucleoside triphosphate cleavage and phosphate release steps by associated rabbit skeletal actomyosin, measured using a novel fluorescent probe for phosphate. Biochemistry 36, 11828–11836. White, H. D., and Taylor, E. W. (1976). Energetics and mechanism of actomyosin adenosine triphosphatase. Biochemistry 15, 5818–5826. Whittaker, M., Wilson‐Kubalek, E. M., Smith, J. E., Faust, L., Milligan, R. A., and Sweeney, H. L. (1995). A 35‐A movement of smooth muscle myosin on ADP release. Nature 378, 748–751. Woodward, S. K., Eccleston, J. F., and Geeves, M. A. (1991). Kinetics of the interaction of 20 (30 )‐O‐(N‐methylanthraniloyl)‐ATP with myosin subfragment 1 and actomyosin subfragment 1: Characterization of two acto‐S1‐ADP complexes. Biochemistry 30, 422–430. Zeng, W., Conibear, P. B, Dickens, J., Cowie, R., Wakelin, S., Malnasi‐Csizmadia, A., and Bagshaw, C. (2004). Dynamics of actomyosin interac mmtions in relation to the crossbridge cycle. Phil. Trans. R. Soc. B 359, 1843–1855.

X‐RAY DIFFRACTION STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE By JOHN M. SQUIRE* AND CARLO KNUPP{ *Biological Structure & Function Section, Biomedical Sciences Division, Imperial College Faculty of Medicine, London SW7 2AZ London, United Kingdom; { Structural Biophysics Group, School of Optometry and Visi00on Sciences, Cardiff University, Cardiff University, Cardiff CF10 3NX, United Kingdom

I.

II.

III. IV.

V.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Muscle Structure and Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Factors Affecting Diffraction Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Diffraction from Actin Filaments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Diffraction from the Myosin Head Array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. The Z‐Band Contribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Diffraction from the Myosin Filament Backbone . . . . . . . . . . . . . . . . . . . . . . Modeling of Rigor Muscle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Labeling of Actin in Rigor Muscle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. The Weak‐Binding State. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Summary: Take‐Home Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Time‐Resolved Events in Contracting Muscles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Changes in the Equatorial Reflections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X‐Ray Interference Measurements and Their Implications . . . . . . . . . . . . . . . . . A. Introduction to the Idea of Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Interference Observations and Their Possible Interpretations. . . . . . . . . . C. Interference Changes During Muscle Transients . . . . . . . . . . . . . . . . . . . . . . D. Further Details of the Interference Experiments . . . . . . . . . . . . . . . . . . . . . . E. Temperature Jump Experiments and Isotonic Contractions . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I.

195 196 197 206 215 217 221 222 222 225 228 229 230 230 234 234 237 239 240 245 246 247

Introduction

The basic assembly of striated muscles has been described in detail in Squire et al. (2005), Granzier and Labeit (2005), and Brown and Cohen (2005) and ideas about the crossbridge cycle were described in Geeves and Holmes (2005). Many of the ideas about the crossbridge mechanism have come from X‐ray diffraction studies of muscle, but to many people such studies are difficult to understand and, even to some practitioners of the technique, it is easy to be misled by certain types of results. The technique itself is immensely powerful and may be one of the very few approaches that can actually probe molecular structural changes that occur in intact muscles while they are undergoing their usual function, namely, tension ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71006-2

195

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

196

SQUIRE AND KNUPP

production and shortening. This article explores some of the issues involved in X‐ray diffraction studies of muscle and attempts to separate what has actually been shown unambiguously from what may be merely either conjecture or just one plausible interpretation among a panoply of very different ones.

A. Muscle Structure and Diffraction The discussion here starts with the structure of vertebrate striated muscles, the basic components of which were described in Squire et al. (2005; see also Squire, 1981, 1992, 1997). The vertebrate sarcomere consists of bipolar myosin filaments with threefold rotational symmetry and myosin heads organized in the bridge regions on three co‐axial roughly helical strands with nine head pairs per turn of the helix, giving a true repeat of 429 A˚ and an axial separation between crowns of heads averaging
143 A˚ (see Squire et al., 2005, Fig. 16). The myosin filaments are organized in hexagonal arrays across the myofibril, with those in bony fish muscle having identical orientations around their long axes in a simple lattice arrangement and those in higher vertebrates having a statistical mixture of two orientations 60 apart giving a superlattice (see Squire et al., 2005, Fig. 6). A‐band lattice structure is defined and stabilized by bridging structures in the M‐band (Squire et al., 2005, Figs. 23 and 24). Interdigitating with these myosin filaments are actin filaments (Squire et al., 2005, Fig. 9) with a helical symmetry that in vertebrate striated muscles is close to, but not exactly, that of a 13/6 helix of actin subunits, which would have an axial repeat of 357.5 A˚ . The actin filaments carry strands of tropomyosin molecules and troponin complexes (Squire et al., 2005, Fig. 2.9, and see Brown and Cohen, 2005). The actin filaments are arranged at the trigonal points of the hexagonal lattice of myosin filaments in the A‐band, but through the I‐band there is a transition to a square array in the Z‐band. The giant protein titin (Granzier and Labeit, 2005) interacts systematically with myosin filaments in the A‐band and then spans the I‐band and interacts with the Z‐band through its Z‐repeats and other adjacent sequences. In thinking about X‐ray diffraction from this assembly, a number of the sarcomere components contribute to the observed patterns in ways that have been the subject of detailed analysis. In the A‐band, these include the myosin filament backbone, where the coiled‐coil a‐helical myosin rods pack together, the myosin head arrays in the bridge regions of the myosin filaments, the non‐myosin A‐band proteins titin and C‐protein (MyBP‐C), and the A‐band parts of the actin filaments. Very little has been seen in X‐ray patterns so far that appears to be related to the M‐band, probably

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

197

because the M‐band carries relatively little mass. In the I‐band, the contributors are the remainder of the actin filaments, the tetragonal Z‐band, and possibly the I‐band parts of titin. In order to explain and understand the various contributions that these different structures make, the following section assesses the diffraction effects of each of the major components. The background ideas about muscle structure and the crossbridge cycle, together with some historical perspectives, are discussed in this volume in Squire et al. (2005) and Geeves and Holmes (2005) and also, for example, in Huxley (1969, 2004), Holmes (1997), Geeves and Holmes (1999), and the special Royal Society issue on ‘‘Myosin, Muscle and Motility’’ (Phil. Trans. Roy. Soc. B. volume 359, pp 1811–1964).

B.

Factors Affecting Diffraction Patterns

1. General Ideas About Diffraction Although this volume is not primarily intended as a techniques book, it will probably be helpful here to summarize in a qualitative way for those who are non‐experts in the basic aspects of diffraction theory that are used later in this article. Those who already have a thorough understanding of the technique should proceed directly to Section I.C. Figure 1 summarizes the basic concept underlying all diffraction methods. Whether the radiation being used is X‐rays, neutrons, electrons, or visible light, it can be described in terms of a wave of oscillating amplitude (y) with a maximum amplitude a, and with a wavelength l. These waves can be imagined as propagating across space in the x direction as in Fig. 1A. If two such waves with the same wavelength arrive at the same point with their peaks and troughs in step (they are said to be in phase), then the amplitudes add and a wave of larger amplitude results (Fig. 1B). This is called constructive interference. However, if the peaks of one coincide with the troughs of the other (they are exactly out of phase), then destructive interference occurs (Fig. 1C). Often such peaks will be neither exactly in phase nor exactly out of phase and they sum to give an intermediate or partial amplitude as in Fig. 1D. If a beam of light with a wavelength of, say
5000 A˚ , falls onto a card with two small holes a distance d apart (Fig. 2), then each hole will scatter the light in all directions. In a particular direction at the angle f the ways that the light scattered from the two holes adds up depends on how in phase or out of phase the two beams are. This can be determined by the size of d sinf, which is the extra distance the beam from X must travel relative to the beam from Y. Obviously, if the two waves add up after being

198

SQUIRE AND KNUPP

Fig. 1. Summary of the ideas of interference. (A) Wave profile with wavelength l and amplitude a. (B) Constructive interference from waves moving in the x direction where the amplitude (a) of the two waves varies in step (or in phase). (C) destructive interference by waves out of step by half a wavelength (l/2), and (D) partial reinforcement by waves not exactly in phase or out of phase.

shifted by a whole wavelength or any integer multiple of the wavelength (nl), then constructive interference will occur (Fig.1B). So if d sinf is the same as nl, then an intense peak will be seen on a screen placed on the right hand side of Fig. 2A. In general, if d sinf 6¼ nl, then partial or destructive interference occurs. Fig. 2B–E show various situations where the path difference d sinf is (B) 0 (n ¼ 0), (C) l/2 (n ¼ 0.5), (D) l (n ¼ 1), and (E) 2l (n ¼ 2). In summary, the condition d sinf ¼ nl produces a series of intensity peaks on the screen for varying values of n. This is described as the diffraction pattern from the array of holes. The condition d sinf ¼ nl, sometimes known as the grating equation, has some interesting implications. For a given wavelength of radiation, if the separation of the holes d is larger, then the value of f needed for constructive interference is smaller and vice versa (i.e., sinf / 1/d). This is often described in terms of the reciprocal nature of diffraction. Also noteworthy is that if the wavelength increases, then the whole diffraction pattern gets bigger as well. Most importantly, if f can be measured and l is known, then the distance d can be calculated.

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

199

Fig. 2. (A) Scattering from two small holes at points X and Y a distance d apart in an opaque card, and (B) to (E) the effects of the angle of scatter (f) on whether the two waves are in phase or out of phase. Path differences are (B) zero (in phase), (C) l/2 (out of phase), (D) l (in phase), (E) 2l (in phase).

2. Diffraction from Crystals The next step to think about is what happens if the radiation is, say, an X‐ray beam (with a wavelength of 1–2 A˚ ) and this falls onto an array of atoms in a crystal rather than holes. This situation is illustrated in Fig. 3A. What happens when the X‐rays arrive at the atoms is that the electrons in the atoms are caused to oscillate by the alternating electric field in the X‐ray beam and such oscillating charged particles themselves radiate at the same wavelength as the incident radiation, but in all directions. Considering the top plane of atoms in Fig. 3A, including the atom at O, then each of these atoms will be stimulated by the incoming X‐ray beam and will radiate in all directions at the same wavelength l. It is easy to show that the radiation incident at a particular angle y is scattered most strongly in a

200

SQUIRE AND KNUPP

Fig. 3. (A) Scattering of an X‐ray beam from planes of atoms in a crystal. Scattering from the atoms (atom planes) at O and B has a path difference of 2d siny, giving rise to Bragg’s law (nl ¼ 2d siny). The angle of incidence is y; the diffraction angle between the incident and diffracted beam is 2y. (B) Geometry of a typical fiber diffraction experiment. The radiation used (e.g., X‐rays or neutrons) comes in from the left and passes through the fiber. Molecules in the fiber scatter the radiation onto a film or detector at a distance D from the fiber. The pattern of spots on the detector can be related to the organization of the molecules in the fiber. Spots at a position S from the center of the diffraction pattern are diffracted through an angle given by Tan 2y ¼ S/D. This, combined with Bragg’s law can yield the value of d corresponding to the peak at S. Typical fiber patterns have a meridian, parallel to the fiber axis through the undiffracted beam direction (center), an equator at right angles to this through the center, and a series of layer lines (horizontal) parallel with the equator. These layer lines may have continuous intensity along them, or if the fiber is well ordered they may be sampled on vertical row‐lines to give diffraction spots along the layer lines.

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

201

direction that is also at an angle y to the plane of atoms. In most other directions there is destructive interference. It is almost as if the X‐rays have been reflected by the plane of atoms. What happens now if there is a second plane a distance d below the first plane and reflecting like the first? Fig. 3A shows the geometry involved. In this case, the two beams going off to the right at angle y are such that the beam from B has had to travel the extra distance AB þ BC relative to the beam from O to reach a screen or film on the right. Obviously if AB þ BC is a whole number of wavelengths, l, then a constructive interference would be expected to occur. From the figure it is clear that AB and BC are both d sin y, so we now have the rule for diffraction (constructive interference) from a crystal: 2d siny ¼ nl. This is Bragg’s law. Note that as well as the reciprocal relationship that we had before, and the change of size of the diffraction pattern with l, there is an additional condition in this case. The condition is that if the crystal planes of spacing d are not at the angle y as in Bragg’s law above, then those particular planes will not diffract. A single crystal needs to be turned relative to the incident direction of a monochromatic (single wavelength) X‐ray beam in order to get diffraction from particular planes of atoms. However, when diffraction is seen, measurement of the angle of diffraction (the angle 2y between the incident and diffracted beams) then permits calculation of the value of d for those planes, assuming that the wavelength is known (Fig. 3B).

3. Description of Lattice Planes A 3D crystal has its atoms arranged such that many different planes can be drawn through them. It is convenient to be able to describe these planes in a systematic way and Fig. 4 shows how this is done. It illustrates a 2D example, but the same principle applies to the third dimension. The crystal lattice can be defined in terms of vectors a and b, which have a defined length and angle between them (it is c in the third dimension). The box defined by a and b (and c for 3D) is known as the unit cell. The dashed lines in Fig. 4A show one set of lines that can be drawn through the 2D lattice (they would be planes in 3D). It can be seen that these lines chop a into 1 piece and b into 1 piece, so these are called the 11 lines. The lines in B, however, chop a into 2 pieces, but still chop b into 1 piece, so these are the 21 lines. If the lines are parallel to an axis as in C, then they do not chop that axis into any pieces so, in C, the lines chopping a into 1 piece and which are parallel to b are the 10 lines. This is a simple rule. The numbers that are generated are known as the Miller indices of the plane. Note that if the structure in Fig. 6.4 was a 3D crystal viewed down the c axis, the lines would be planes. In these cases, the third Miller index would be zero (i.e., the planes would be the 110 planes in A, the 210 planes in B, and the 100

202

SQUIRE AND KNUPP

Fig. 4. Demonstration of the definitions if Miller indices describing different planes through a lattice. For details, see text.

planes in C). The Miller indices are also represented in general by the italic letters h, k, and l.

4.

Intensity of Diffraction Peaks: The Convolution Theorem

One of the obvious features of diffraction patterns is that the diffracted peaks do not all have the same intensity. We can see partly why this is from the illustration in Fig. 5. Figure 5A represents a 2D lattice (as in Fig. 4), and D shows the sort of diffraction pattern that would be observed if the scattering object in Fig. 2A was a mask of holes arrayed as in Fig. 5A and the diffraction pattern was viewed on the screen. The positions of the spots in Fig. 5D are totally defined by the arrangement of the objects in A, that is, by a, b, and the angle between them. However, real crystals have some

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

203

Fig. 5. Illustration of the convolution theorem applied to a crystal structure and its diffraction pattern. (A) is a lattice and (B) is the motif or repeating unit on the lattice. The full crystal (C) is a convolution of (A) and (B). The diffraction pattern (F) of the crystal (C) is the product of the diffraction patterns (Fourier transforms) (D) and (E) from (A) and (A), respectively. For details, see text. (Based on Squire, 1981.)

interesting objects (e.g., Fig. 5B) within each unit cell (Fig. 5C). It is convenient to think about the full crystal (C) as what is termed a convolution. Literally this means ‘‘folding together.’’ If the object in B is picked up and placed on every point on the lattice in A, then the structure in C is generated. This is called the convolution ( * ) of B with A. In short, C ¼ (A) * (B). It can be shown mathematically, and is illustrated in Fig. 5D–F, that the diffraction pattern (G(C)) from C is then the product of the diffraction patterns G(A) and G(B) from A and B, respectively. In short: If (c) ¼ (a) * (c), then G(c) ¼ G(a)  G(c); or, in words, as in Fig. 5, ‘‘IF lattice * object ¼ crystal, THEN lattice transform  object transform ¼ crystal transform.’’ The term transform here is the mathematical equivalent of the diffraction pattern. This rule is known as the convolution theorem. The way that the product of the two diffraction patterns D and E is produced is such that they are placed on top of each other with their centers together and then multiplied point‐for‐point. This means that where there is zero in either pattern there is zero also in F. The result is

204

SQUIRE AND KNUPP

that F consists of peaks in the positions defined by D, but the intensities of these peaks are determined by the intensities in E at those particular positions. The ‘‘take‐home message’’ from this is that the positions of observed diffraction peaks tell us about the lattice and unit cell in the crystal (A), whereas the intensities of the peaks tell us about the structure of the object (B) on the lattice. This object is variously described in the literature as the motif, or the unit cell contents, or the asymmetric unit. Following from this, if the intensities of the peaks in (F) can be measured for a real crystal, then it is possible to work out the structure of the object in (B). This is what is done in the technique of protein crystallography (e.g., Blow, 2002).

5.

Effects of the Extent of a Lattice

Before embarking on the diffraction patterns produced by the objects in the muscle sarcomere, it is necessary to illustrate one other feature of diffraction patterns. Figure 6 shows various objects as in Fig. 2A and their computed diffraction patterns. The objects are like the mask of holes a distance d apart as in Fig. 2A, but this time there are different numbers of holes. In Fig. 6A there are 3 holes, in C there are 7 holes, and in E there are 10 holes. Since it is assumed that the wavelength of the light being scattered in these examples is the same, and all the objects have the same interhole spacing d, the grating equation d sinf ¼ nl from Fig. 2A still applies to each of them. This means that the diffraction patterns on the right of Fig. 6 all have peaks in exactly the same place. However, the width of the peaks changes. If there are only three objects as in Fig. 6A, then there are many directions each side of the main peaks in the pattern where the partial interference still leaves a significant amount of intensity. The peaks in Fig. 6B are therefore quite broad. On the other hand, as the number of objects increases, the partials become much weaker and the diffraction peaks become progressively narrower. If the length of the whole array is W, then the width of the observed peaks can be expected to be related to 1/W (the reciprocal nature of diffraction). 1/W10 is therefore very small compared with 1/W3, with 1/W7 lying between the two. In summary, the width of an observed diffraction peak can be related to the length of the array giving rise to the peak. Short arrays give broad peaks, large arrays give sharp peaks. This armory affords consideration of the diffraction from the components of the muscle sarcomere. Note first that muscles are not single crystals of the kind illustrated in Fig. 3A. The sarcomeres themselves can have varying degrees of order; some, like insect flight muscle and bony fish muscle, are almost ‘‘crystalline’’ within an A‐band or sarcomere. But, whatever the muscle, both different myofibrils within a fiber, and different

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

205

Fig. 6. Illustration of the effects of array length on the thickness (width) of observed diffraction peaks. The arrays on the left all have the same spacing (d) between points. The peaks in the diffraction patterns on the right therefore have peaks centered in the same places, despite the varying array lengths. What changes is the thickness of the lines; short arrays (A) give thick diffraction peaks (B); long arrays (E) give relatively narrow diffraction peaks (F); and intermediate length arrays (E) have intermediate effects (D). Calculated and displayed using MusLABEL (Squire and Knupp, 2004).

fibers within a whole muscle, can have random rotations about the fiber long axis, thus presenting different Bragg planes to the incoming X‐ray beam. This means that the muscle diffraction pattern is actually a so‐called fiber pattern that possesses rotational symmetry around the fiber axis direction (the meridian in Fig. 3B). The effect of this is that Bragg’s law

206

SQUIRE AND KNUPP

is satisfied simultaneously for different diffraction planes in myofibrils with different orientations around the fiber axis, and, ideally, all possible diffraction peaks out to a certain resolution can be seen in a diffraction pattern from a stationary muscle or fiber. In other words, it is not usually necessary to adjust the orientation of a muscle or fiber to see all of the low‐angle X‐ray diffraction pattern.

C. Diffraction from Actin Filaments Actin filaments are typical of many elongated biological particles in that they have their subunits arranged in a helical manner. Figure 7 illustrates some general features of helical structures. First, the helix is thought of first as a continuous uniform wire. The wire will turn around the axis by some angle and, at the same time, will rise along this axis by an amount determined by the angle. The amount traveled along the helix axis in one complete turn of the helix around its axis is the helix pitch (P). Helical molecules or filaments are made up from subunits, such as an actin monomer or a myosin molecule, and the axial separation between monomers (h) is called the subunit axial translation or unit rise. It may be that there is an exact number of subunits in one pitch, in which case the helix is said to be integral. B‐DNA is an example of this with exactly 10 nucleotide pairs (each being one unit) in one pitch. It is described as a 101 or 10/1 helix. However, other structures are not integral and it may be necessary to traverse several helix pitches before a subunit is reached that is at an exactly equivalent azimuthal position (i.e., at the same angle measured around the filament long axis) to the starting subunit. In this case a repeat C is defined. The structure in Fig. 7A has exactly five subunits in two turns of the helix. It is a 52 or 5/2 helix where C ¼ 2P ¼ 5h. It is not appropriate to go into the full mathematics of helical diffraction theory, which is done well elsewhere (Chandrasekaran and Stubbs, 2001; Harford and Squire, 1997; Holmes and Blow, 1965; Squire, 2000). However, some of the important features can be noted. The diffraction pattern has the form illustrated in Fig. 7B. One can imagine a fiber of helical molecules or filaments being in the path of an X‐ray beam (Fig. 7C) and diffracting the X‐rays onto a screen or detector a distance D from the fiber (Fig. 3B). The axis in the diffraction pattern that is parallel to the fiber axis and goes through the undiffracted beam position (i.e., the center of the pattern) is called the meridian. The axis at right angles to this through the center is called the equator. The helical diffraction pattern as exemplified by Fig. 7B consists of a series of spots along the meridian at positions related to m/h where h is the subunit axial translation as above and m is an integer number, which can be positive, negative, or zero. The spots at m ¼ 0 and m

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

207

Fig. 7. Geometry of a helical structure (A) and the form of its diffraction pattern (B). In (A), the pitch (P) of the helix is like the wavelength of a sine wave. The radius (r) of the helix is like the amplitude of the sinewave. The subunit axial translation (h) is the rise along the helix axis from one monomer to the next. If there is not a whole number of monomers in one turn of the helix (said to be a non‐integral helix), then there may be a longer repeat (C). In the case illustrated C ¼ 2P. Dimensions in the helix in (A) have their counterparts in the diffraction pattern illustrated in (B), but dimensions in (B) are reciprocal to those in (A). Meridional reflections occur at positions m/h from the equator, where m is an integer. Each of these positions is the center of a so‐called helix cross consisting of layer lines, which are n/P up or down from the meridional peaks, where n is another integer. All of the resulting layers of intensity can be related to orders of l/C, where C is the repeat of the helix and l is the layer line number.

¼ 1 are shown in the figure. For a perfectly helical structure, these and peaks at other values of m are the only intensities that occur along the meridian. Off the meridian is a series of so‐called layer lines (horizontal), which occur at axial positions (i.e., measured in the direction parallel to the meridian), which are orders of the repeat C: they are at l/C, where l, known as the layer line number, is an integer that is positive, negative, or zero. Particularly strong layer lines occur when l/C is equal to 1/P (l ¼ 2), 2/P (l ¼ 2), and so forth. They also occur at the same distances above

208

SQUIRE AND KNUPP

and below the meridional reflections at different values of m. In other words there are strong peaks at: l ¼ m/h þ n/P, where n is another integer that can be positive, negative, or zero (in mathematical treatments n is actually the order of a Bessel function describing the intensity along the layer line). Measurement of the axial positions of the meridional reflections (related to 1/h) and of the layer lines (related to 1/C) can determine the number of subunits h in one repeat C. An explanation of the origins of the helical diffraction pattern is given in Fig. 8 and its legend. In the case of 13/6 actin filaments, the subunit axial translation h is 27.5 A˚ and the repeat C is after 13 subunits at 357.5 A˚ , which is also after 6 turns of the helix. The strongest layer lines are where l ¼ 6 or 7. This is because the pitch P of the 13/6 actin helix is C/6 at
59 A˚ . The sixth layer line is therefore 1/P up from the origin, where m ¼ 0, and the seventh layer line at 51 A˚ is 1/P down (i.e., 6 layer lines down) from the m ¼ þ1 position, which occurs on layer line 13. These structural features of an actin filament are illustrated in Fig. 9 (see, for example, Parry and Squire, 1973; Holmes et al., 1990; Lorenz et al., 1993). This figure also shows a convenient representation of helical structures in terms of a radial projection or helical net. This is obtained (A) by imagining a piece of paper wrapped round the filament and marking wherever there is a subunit on the helix. This paper is then unwrapped (B) and the helical strands of the filament become straight lines. The left‐ and right‐hand edges of the radial projection represent the same line along the filament. Marked on the radial net in Fig. 9C are the 27.5 A˚ subunit axial translation between actin monomers and the pitch of the genetic helix (i.e., the helix that runs through all the monomers) of 59 A˚ . The fact that the azimuthal rotation between successive subunits along the helix is not far from 180 (it is at 360  6/13 ¼ 166.15 ) shows why the actin filament appears as two slowly twisting strands of monomers, like two strings of beads twisting around each other (see Fig. 3A and Fig. 9 in Squire et al. of this volume). In terms of spacing, the first actin layer line (A1 in Fig. 10B) from a 13/6 helix occurs at 1/357.5 A˚ –1, the second (A2) at 2/357.5 ¼ 1/178.75 A˚ 1, the sixth (A6) at 1/59.58 A˚ –1, the seventh at 1/51.07 A˚ –1, and the thirteenth (A13) at 1/27.5 A˚ –1. Not all actin filaments have 13/6 helical symmetry. For example, in insect flight muscle, as exemplified by Lethocerus, the actin filaments form a 28/13 helix. This also occurs in the vertebrate striated muscle Z‐band (Squire et al., this volume. Section III.E; Luther and Squire, 2002). The differences between the diffraction patterns from helices with 13/6 and 28/13 symmetry are illustrated in Fig. 10. The diffraction patterns were generated by the program HELIX (Knupp and Squire, 2004), but the program MusLabel can also be used (Squire and Knupp, 2004). Another

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

209

Fig. 8. Generation of the form of the helical diffraction pattern. (A) shows that a continuous helical wire can be considered as a convolution of one turn of the helix and a set of points (actually three‐dimensional delta‐functions) aligned along the helix axis and separated axially by the pitch P. (B) shows that a discontinuous helix (i.e., a helical array of subunits) can be thought of as a product of the continuous helix in (A) and a set of horizontal density planes spaced h apart, where h is the subunit axial translation as in Fig. 7. This discontinuous set of points can then be convoluted with an atom (or a more complicated motif ) to give a helical polymer. (C)–(F) represent helical objects and their computed diffraction patterns. (C) is half a turn of a helical wire. Its transform is a cross of intensity (high intensity is shown as white). (D) A full turn gives a similar cross with some substructure. A continuous helical wire has the transform of a complete helical turn, multiplied by the transform of the array of points in the middle of (A), namely, a set of planes of intensity a distance n/P apart (see Fig. 7). This means that in the transform in (E) the helix cross in (D) is only seen on the intensity planes, which are n/P apart. (F) shows the effect of making the helix in (E) discontinuous. The broken helix cross in (E) is now convoluted with the transform of the set of planes in (B), which are h apart. This transform is a set of points along the meridian of the diffraction pattern and separated by m/h. The resulting transform in (F) is therefore a series of helix crosses as in (E) but placed with their centers at the positions m/h from the pattern center. (Transforms calculated using MusLabel or HELIX.)

210

SQUIRE AND KNUPP

Fig. 9. Representation of a 13/6 actin filament together with its illustration by means of a radial net. In (A) an imaginary piece of paper is wrapped round the filament and on it are marked all the positions of the actin monomers. The paper is then unwrapped as in (B) and the helical tracks in (A) become straight lines. The final result in (C) is the radial projection or radial net. The 59 A˚ pitch length (P) and 27.5 A˚ subunit axial translation (h) are indicated in (C).

way of thinking about a 13/6 helix is that it has 26 subunits in 12 turns of the helix. A helix with 28 subunits in 13 turns is therefore very closely related—just obtained by slight untwisting of a 26/12 (i.e., 13/6) helix. For this reason, the two diffraction patterns in Fig. 10 are remarkably similar and distinguishing between them experimentally is not easy. Note that there is also a further important feature of actin filaments in muscle, namely, that they are rather flexible. Not only can their long axes be curved, but they can also twist azimuthally in a random way so that even along a helix that is a 13/6 filament on average there can be local twisting and untwisting, which makes the so‐called coherence length along the filaments rather short (Egelman et al., 1982; it is known as the random variable twist). The coherence length can be thought of as being a length

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

211

Fig. 10. Comparison of 13/6 (A) and 28/13 (C) actin filaments and their diffraction patterns, (B) and (D), respectively. The pattern in (B) shows characteristic features such as a strong first layer line (A1; P ¼ 357.5 A˚ ); a significant second layer line (A2; 179 A˚ ); strong layer lines, which are the sixth (59.58 A˚ ) and seventh (51.07 A˚ ) orders of the repeat C (357.5 A˚ ); and the first meridional reflection at h ¼ 27.5 A˚ on layer line 13. The pattern in (D) has similar features to (B), but the layer line numbering is different. The first few layer lines are the A2 at 385 A˚ , A4 at 192.5 A˚ , A13 at 59.2 A˚ , A15 at 51.33 A˚ , and A28 at 27.5 A˚ .

212

SQUIRE AND KNUPP

over which the positions of, say, the crossovers of different parts of the two long‐pitched strands in actin are well enough related for the diffraction from these two parts of the filament to interfere in phase. It should be remembered that diffracting arrays that are relatively short (see Fig. 6) give diffraction peaks that are relatively broad. For this reason, the layer lines from the short coherence length actin filaments in muscle diffraction patterns are often quite broad axially. This makes it even harder to measure their axial position exactly and thus to define the helical symmetry precisely. In fact, vertebrate striated muscle diffraction patterns appear to be such that the crossover spacing is actually slightly longer than the 357.5 A˚ expected for an exact 13/6 helix; it is more likely 360–370 A˚ (Huxley and Brown, 1967; Harford and Squire, 1986). Note that in terms of Bragg’s law, and assuming a wavelength l for the X‐rays of say 1.5 A˚ , which is quite common for both laboratory‐based and synchrotron X‐ray sources, the value of the angle of diffraction 2y for an actin filament repeat of 357.5 A˚ is very small. It is 2y ¼ 2  sin–1 (1.5/ 2  357.5) ¼ 0.24 . Since the myosin filament repeat in vertebrate striated muscles is of the same order of magnitude as the actin repeat (it is 429 A˚ ), diffraction from muscle filaments that shows their helical structure is often called low‐angle diffraction (LAD) or small‐angle diffraction (SAD). Such diffraction patterns need to be recorded with the specimen to detector distance (see D in Fig. 3B) long enough (1 to 10 m depending on the X‐ray beam diameter) for the pattern to expand to a useful size that matches the pixel resolution in the detector being used. As discussed later, useful high‐ angle diffraction information can also be obtained from muscle (Squire, 1986). This arises from the myosin filament backbone structure and also from the detailed internal structure of actin filaments (Holmes et al., 1990). As seen earlier (in Fig. 5), it is the lattice shape and size in a crystal that define the location of the diffraction spots and it is the repeating unit on the lattice that defines the relative intensities of these spots. So it is with actin filaments. It is the symmetry of the array of actin subunits that determines where the layer lines will be (as in Fig. 10), but it is the shape and orientation of the actin subunits on each point along the helix that define the relative intensities of these actin layer lines. The actin monomer (as discussed in Squire et al. in this volume; see Section II.A.3 and Fig. 9) has four subdomains in which subdomains 3 and 4 lie close to the helix axis and subdomains 1 and 2 are on the outside of the helix. Subdomain 1 is relatively large and is where myosin heads bind, whereas subdomain 2 is relatively small and its precise role is not clear. However, it can be shown (e.g., Harford and Squire, 1997) that even quite small movements of subdomain 2 can have their effect on the intensities of the low‐angle actin layer lines. In terms of resolution, which is often all‐important in structural

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

213

studies, the actin layer lines very definitely occur at low resolution. However, the crystal structure of the actin monomer is known, so we have a great deal of additional very high‐resolution information. When the low‐angle layer lines from the actin filament are modeled in terms of the relative positions of the actin subdomains, the intensities of the low‐angle actin layer lines are seen to be sensitive to sub‐domain movements of only a few angstroms. The sensitivity of the technique is therefore very high. Before leaving the discussion of actin filaments it is necessary to discuss the effects of tropomyosin and troponin. Tropomyosin strands run along the long‐period helices of actin where one tropomyosin molecule interacts pseudo‐equivalently with seven actin monomers. Successive tropomyosin molecules link end to end along the filament to form more or less continuous strands of density. In the case of a 13/6 actin helix, one of these strands has a pitch of 2  357.5 ¼ 715 A˚ , but because there are two essentially continuous strands of tropomyosin, the tropomyosin repeat appears to be halved to 357.5 A˚ . The tropomyosin strands therefore contribute strongly to the low‐angle actin layer lines A1 at 1/357.5 A˚ –1, A2 at 2/357.5 A˚ –1, A3 at 3/357.5 A˚ –1, and so on. The tropomyosin strands do not contribute strongly elsewhere because the density appears continuous along the strands, so there are no helix crosses apart from the one at m ¼ 0. However, the tropomyosin diffraction pattern and the actin diffraction pattern interfere with each other and this is how it was realized that the tropomyosin strands might move across the actin monomers as a result of muscle activation (Haselgrove, 1972; Huxley, 1972; Parry and Squire, 1973). The A2 reflection was weak in patterns from resting muscle but became stronger in patterns from active muscle when the A3 layer line became slightly weaker. This was consistent with a movement of the tropomyosin from a position well out of the ‘‘groove’’ between the two strands of actin to one where it was closer into the groove (see analysis in Parry and Squire, 1973). This was because the actin and tropomyosin patterns interfered with each other in different ways when the tropomyosin was in the two positions. This kind of model now enjoys a great deal of support (Al‐Khayat et al., 1995; Brown and Cohen, 2005; Craig and Lehman, 2002; Squire and Morris, 1998; Vibert et al., 1997). The diffraction pattern from troponin is very different. Here, there is one troponin complex for each tropomyosin molecule, but the end‐to‐end repeat along the tropomyosin strands is about 385 A˚ It is longer than the actin filament crossover repeat of just over 357 A˚ in vertebrate muscles (Fig. 11A and B). Because much of the troponin complex is globular, unlike tropomyosin, it shows very marked discontinuous density every 385 A˚ along each strand of the actin filament, with the troponins in opposite strands axially shifted by the actin monomer subunit translation h of

214

SQUIRE AND KNUPP

Fig. 11. (A) Radial net for an actin filament (blue) with tropomyosin strands (green) and troponin added at 385 A˚ intervals along each strand (red) and a possible line of nebulin (gray) shown here along one strand only of the long‐period actin helices. In fact, there may be nebulins on both strands. (B) As in (A) but shown as a 3D model with the same color code. (C) Left‐hand half of a low‐angle X‐ray diffraction pattern from bony fish muscle, and (D) calculation of the diffraction pattern of a structure similar to that in (B), but with a slightly different pitch to show that some features in (C) could be due to troponin. (Based on Squire et al., 2004.)

27.5 A˚ . In vertebrate striated muscles, the troponin repeat is slightly longer than the actin crossover repeat, so the troponin subunits appear to lie on two very slowly twisting helices around the actin filament. The effect on the diffraction pattern is shown in Fig. 11C and D. The troponin complexes can be thought of as lying on helices of pitch 2  357.5 ¼ 715 A˚ , but with a subunit axial translation of 385 A˚ (Fig. 11A and B). Thus, part of the diffraction pattern consists of a series of meridional reflections at orders m ¼ 0, 1, 2, 3, etc. of 385 A˚ . These occur at spacings 1/385 A˚ –1, 2/ 385 [¼ 1/192.5 A˚ –1], 3 /385 [¼ 1/128.3 A˚ –1], etc. In addition, there are layer lines up and down from these positions by an amount n/715 A˚ –1. Since 385 is not far from half of 715, the layer lines (n ¼ 1, 3, etc.) appear about halfway between the meridional reflections (Fig. 11C and D). However, the layer lines for n ¼ 2, 4, etc. appear very close to the equator and to the meridional reflections for different values of m, so the whole pattern appears as meridional peaks with very closely spaced layer lines above and below them forming a very shallow cross, together with off‐ meridional layer lines about halfway between the meridional reflections.

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

215

Note that in insect flight muscle, since the actin filaments have 28/13 symmetry with a repeat of 385 A˚ (which is the same as the troponin repeat), in this muscle the troponins lie in regular positions along each side of the actin filament, almost like a ladder; they fit precisely into the regular A‐band lattice geometry (see Fig. 10C in Squire et al., this volume) where each actin filament lies exactly between two myosin filaments. This gives rise to interesting interference effects between the troponins and any myosin heads that bind to actin (Tregear et al., 1998).

D.

Diffraction from the Myosin Head Array

Myosin filament structure has been described by Squire et al. (2005). In vertebrate striated muscles the myosin filaments can be described approximately as three‐stranded 9/1 helices. The helix pitch is 1287 A˚ , but, because there are three strands and nine subunits in each strand, the structure repeats after C ¼ 1287/3 ¼ 429 A˚ . Figure 12 shows the expected form of the low‐angle diffraction pattern from such filaments. The modeling of this structure by X‐ray diffraction was described by Squire et al. in terms of the three crowns of heads within each 429 A˚ repeat. The crown repeat of 143 A˚ gives rise to an m ¼ þ1 meridional reflection, which has been labeled as the M3 reflection in many muscle studies (as in Fig. 12). The myosin head array also gives rise to layer lines at orders of the repeat of 429 A˚ . The first myosin layer line (ML1) is at 1/429 A˚ –1, the second (ML2) at 2/429 ¼ 1/ 214.5 A˚ –1, and so on. The M3 reflection occurs on the third layer line at 3/429 ¼ 1/143 A˚ –1. It was shown in Fig. 27 of Squire et al. (this volume) that the myosin filaments in different muscle types, particularly in invertebrate muscles, have their heads arranged on different surface lattices. There can be different numbers of helical strands and also different axial repeats. However, in all of these other cases the head arrays appear to be perfectly helical. The vertebrate striated muscle myosin filaments are different in that their heads do not lie on perfect helical tracks; there is a perturbation (described in Squire et al., this volume, see Fig. 20) that makes the three crowns within a 429 A˚ repeat nonequivalent. This shows up very obviously in the X‐ray diffraction patterns from vertebrate striated muscles. If the structure was helical, meridional reflections would be expected to occur only at positions where m ¼ 0, 1, 2, etc., that is, at multiples of 3/429 A˚ –1. In fact, meridional reflections are observed on all of the first few orders of the 429 A˚ repeat, in particular with a strong peak on the second myosin layer line at 214.5 A˚ (M2), and with others at M4, M5, M7, and so on. These are all indications that the helix is not quite perfect.

216

SQUIRE AND KNUPP

Fig. 12. (A) The left half of a low‐angle diffraction pattern from bony fish muscle (similar to that in Fig. 11C) but showing the layer lines, which are thought to come from the array of myosin heads on the myosin filaments. This array has a repeat C of 429 A˚ and a subunit axial translation (intercrown spacing) of 143 A˚ . The observed layer lines are therefore ML1 at 429 A˚ , ML2 at 214.5 A˚ , and so on with meridional reflections at M3 (h ¼ 143 A˚ ). Other ‘‘forbidden’’ meridional reflections occur at M2, M4, M5, and so on. (B) Simulation of the layer line pattern in (A) produced by MusLabel for the myosin filament structure illustrated in Fig. 16 in Squire et al. (this volume) and in 3D in Fig. 14 here.

The reason for this perturbation could be that the myosin filaments in vertebrate striated muscles carry titin, which has a repeating pattern along it which itself probably has a 429 A˚ spacing, and C‐protein (MyBP‐C), which in the C‐zone region of the A‐band appears to label every third crown level. Squire et al. (2003d) showed that the outer, N‐terminal, end of C‐protein could also bind to actin, in addition to the C‐terminal part binding to myosin and to titin, and this may be why there are particular features of the X‐ray pattern from striated muscles with an apparent

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

217

spacing that is just longer than the 429 A˚ repeat of the myosin filaments (Fig. 13). In particular, there is a meridional reflection at about 440 A˚ , which has been shown to come from C‐protein (Rome et al., 1973a), and there is evidence from electron microscopy that appears to show C‐protein with a slightly longer spacing than myosin (Sjostrom and Squire, 1977; Squire et al., 1982). The diffraction pattern from C‐protein is discussed further when the effects of interference between the two halves of the A‐band are considered. To summarize the results so far, Fig. 14 shows a stereo view of part of the A‐band in bony fish muscle with the myosin heads organized on the myosin filament surface as in the X‐ray analysis of Hudson et al. (1997), with the actin filaments as described in Fig. 6.10, and with the C‐protein strands running with the C6–C10 C‐terminal region aligned parallel to the myosin filament axis and the N‐terminal C5–C0 regions crossing over to binding sites on adjacent actin filaments. Note that this arrangement of the C‐terminal region is different from that shown in Fig. 13A, B. Based on evidence for binding between domains 7 and 10 and 5 and 8, Moolman‐ Smook et al. (2002) have suggested that this part of C‐protein may form a collar around the myosin filament backbone. The alternative arrangement in Fig. 14 is based on X‐ray diffraction evidence (Squire et al., 2003d). As yet it is not known which model is correct.

E. The Z‐Band Contribution As detailed by Squire et al. (this volume; see Section II.C.6), one of the curious features of the vertebrate muscle sarcomere is that, although the A‐band lattice is hexagonal, the Z‐band lattice is square. The transition between the two structures in the I‐band has been discussed by Squire et al. (see their Fig. 14, this volume). There is relatively very little material in the Z‐band itself to give very strong diffraction unless the Z‐band is very thick (e.g., in the midshipman fish swim bladder; Lewis et al., 2003). However, the actin filaments in the muscle I‐band appear to be relatively straight for 0.1 mm on each side of the Z‐band and these regions might well diffract strongly. In muscle diffraction patterns the equator is the part of the pattern that shows what the sarcomere looks like in a view down the filament axis. The hexagonal lattice in the A‐band gives rise to equatorial reflections, the first few of which index as the 100, 110, 200, 210, and 300 reflections from a hexagonal lattice of side a ¼ b of
420 to 450 A˚ , depending on the muscle type and the sarcomere length. Such a pattern is shown in Fig. 15. Note that often for brevity the third index is omitted, so these reflections are often referred to as the 10, 11, 20, 21, and 30 reflections.

218

SQUIRE AND KNUPP

Fig. 13. Illustrations of the possible arrangement of C‐protein (MyBP‐C) on the myosin filament backbone in projection down the axis (A) and in axial view (B). Of particular importance here is the possibility that the N‐terminal half of C‐protein extends out and binds to actin in relaxed muscle. (C) Simulation of the possible interactions of C‐protein with binding sites on actin generated using the program MusLABEL (Squire and Knupp, 2004). (D) Left: left half of the low‐angle X‐ray diffraction pattern from bony fish muscle (as in Fig. 11C), showing (right) the possible positions where the C‐protein array in (D) might contribute. (From Squire et al., 2003d.)

Also apparent in Fig. 15A is a peak (shown with arrows) that is not part of the A‐band series, but in fact comes from the Z‐band/I‐band part of the sarcomere. This peak at a spacing of
290 A˚ is between the 10 and 11

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

219

Fig. 14. Model of the A‐band filament lattice in bony fish muscle, based on Hudson et al. (1997) and Squire et al. (2003d), showing the central myosin filament with its projecting myosin heads, together with C‐protein in orange and actin filaments colored as in Fig. 11A and B.

A‐band peaks, which to avoid confusion can be referred to as AB10 and AB11. The 290 A˚ peak is the 10 reflection from the tetragonal I/Z lattice and so can be called the Z10 reflection. As realized by Harford et al. (1994), this spacing puts the Z‐band 11 reflection in exactly the same spot as the A‐band 20 reflection; the Z11 and AB20 peaks overlap. This can be a nuisance if one is trying to use the equatorial reflections to determine the mass distribution in the A‐band by carrying out what is known as Fourier synthesis (e.g., Harford et al., 1994; Squire, 1981; Yu and Brenner, 1989). This is paffiffi means of using the observed intensities (actually the amplitudes / I), and their estimated relative phases, to compute the

220

SQUIRE AND KNUPP

Fig. 15. Intensity profiles along the equator of the bony fish muscle low angle X‐ray diffraction pattern from muscles at rest (A), fully active (B), and in rigor (C). The indexing in (A) is based on the hexagonal A‐band lattice, and the arrows indicate peaks that come from the Z‐band. (C) to (F) are computed electron density maps based on the amplitudes of the A‐band peaks in (A) to (A), respectively. The simple lattice unit cell is outlined in (D). (From Harford and Squire, 1997.)

contents of the unit cell. Harford et al. (1994) devised a method to unscramble the AB20 and Z11 peaks and were able to compute separate density maps for the A‐band (Fig. 15D–F) and Z‐band (not shown).

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

221

It was mentioned by Squire et al. (2005) that the myosin filament lattice spacing changes when the sarcomere length changes to give an almost constant volume for the sarcomere. When the sarcomere length is increased, the A‐band equatorial reflections move to smaller d‐spacings (further from the meridian). An interesting observation by Yu et al. (1977) was that the Z‐band spacing also changes in the same direction and by proportionately the same amount, so that the AB10/Z10 ratio remains constant. There presumably are lateral forces on the Z‐band that make this happen, probably due to the organization of titin in the I‐band (see Squire et al., 2005; Granzier and Labeit, 2005).

F.

Diffraction from the Myosin Filament Backbone

Before leaving this discussion of the basic contributors to the muscle diffraction pattern, there are two other parts of the sarcomere that make a significant contribution at high diffraction angles. One is the internal structure of the actin monomers in the actin filaments, which can give rise to a rich and quite sharp meridional pattern out to d‐spacings of a few angstroms, and also some diffuse off‐meridional high‐angle intensities. This rich, high‐angle pattern can be recorded from oriented gels of actin filaments (Holmes et al., 1990), but can also be seen in high‐angle patterns from muscle (see Squire, 1981). The other contributor is the backbone of the myosin filaments where the myosin rods are packed. These rods are two‐chain coiled‐coil a‐helical structures, each of diameter
20 A˚ , which pack together quite tightly to give a sturdy filament backbone. Evidence from paramyosin filaments suggests that the rods might prefer to pack with the wide parts of the coiled‐coil of one molecule side onto the narrow part of its neighbor, thus giving rise to a pseudo‐body centered lattice. This may be (pseudo‐) tetragonal in molluskan muscles with large diameter thick filaments in which paramyosin is abundant (see Squire et al., this volume, Figs. 10 and 27), but must be more close‐packed in the much smaller diameter, tighter myosin surface layer, filaments of vertebrate or insect flight muscles (see Squire, 1973). Low‐angle X‐ray diffraction data has revealed the distribution of myosin heads on the myosin filament surfaces of different muscles, and the myosin rods must be organized in the backbone so as to produce this arrangement of heads on the surface. However, the precise mode of rod organization is not fully determined. Various models have been proposed that have been tested in depth by Chew and Squire (1995) against the quite rich high‐angle X‐ray diffraction patterns from bony fish and

222

SQUIRE AND KNUPP

other muscles (Squire, 1986). Here there are two main features; one is the 5.1 A˚ meridional reflection from the slightly distorted a‐helix pitch in the coiled‐coil structure and the other is a set of rather diffuse equatorial peaks at
20 A˚ and 10 A˚ due to the lateral packing of the myosin rods, together with diffuse near‐equatorial peaks due to the axial repeat of the coiled‐coils. Chew and Squire found that the model proposed by Squire (1973) for the packing of these rods was in fact the most consistent with the observations of all the models tested, but direct proof is still needed. This model is also consistent with the molecular overlaps predicted from the distribution of charges along the myosin rods. In addition to this high‐angle X‐ray diffraction, the myosin rods together with titin and C‐protein undoubtedly contribute to the low‐angle meridional diffraction pattern. Part of this is due to the peaks already discussed as coming from C‐protein (
440 A˚ ; see later discussion). However, they also contribute to some of the otherwise forbidden meridional reflections, such as M2, M4, M5, etc., and to the ‘‘proper’’ meridionals at M3, M6, M9, and so on. Since the M3 reflection and some of its orders are used to probe the movement of myosin heads in active muscle, knowledge of the contribution from the myosin filament backbone is important. Note that the eleventh‐order meridional peak (M11) at 429/11 ¼ 39 A˚ is relatively strong, possibly because of the myosin packing itself (see Squire et al., 2005) and possibly because of the eleven domains in the 429 A˚ repeat along the titin (Labeit and Kolmerer, 1995; Cantino et al., 2002).

II.

Modeling of Rigor Muscle A.

Introduction

With the structure of resting muscle established as far as we know it, it is appropriate to think about how the myosin heads can interact with actin. Before thinking about contracting muscle and force production, this section briefly describes current knowledge of the well‐documented ‘‘static’’ state after the relaxed state, namely, the nucleotide‐free rigor state. This can be induced in intact muscle by allowing the muscle to die (as would occur in rigor mortis), but a more controlled and useful procedure is to skin the muscle in some way, perhaps by glycerination or by detergent‐ skinning, and then to bathe the muscle in an ATP‐free solution so that all of the heads end up in the AM (rigor) state. A controlled reduction of the ATP level and the use of such things as BDM (2,3‐butanedione monoxime) or NEM (N‐ethylmaleimide) can help to produce a relatively well‐ordered

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

223

rigor structure (e.g., Bershitsky et al., 1996; Yagi, 1992). The myosin head interaction with actin in this state in skinned muscles or fibers has been studied by a variety of methods including X‐ray diffraction, electron microscopy, and analysis of the orientation of fluorescent probes. However, the most detailed modeling of the rigor head conformation on actin has come from image processing of isolated actin filaments labeled with S1 moieties (e.g., Rayment et al., 1993a,b). A reconstruction of this type is shown as Fig. 16C. The most recent analysis of S1‐labeled actin using cryoelectron microscopy and energy‐filtering (Holmes et al., 2003; 2004) produced a density map at
14–17 A˚ resolution and showed both the tilted configuration of the rigor head, as in Fig. 16C, and the suggestion that the cleft in the motor domain should be closed in this strong actin‐binding configuration (see discussion and references in Geeves and Holmes, 2005). This rigor conformation is more or less consistent with previous maps from electron microscopy. So, if we know how isolated myosin heads like to interact with actin in the rigor conformation, the question then is, what happens in rigor muscle? The most direct indication of mass transfer from myosin to actin in rigor muscle comes from analysis of the equatorial part of the low‐angle X‐ray diffraction pattern (e.g., Harford and Squire, 1992; Harford et al., 1994; Haselgrove and Huxley, 1973; Huxley, 1968; Millman and Irving, 1988; Yu et al., 1977, and many others). Fig. 15A–C show the intensity profiles along the equator of bony fish muscle in three different states: relaxed, active, and rigor. The AB10 and AB11 reflections from the A‐band hexagonal lattice are seen with the AB10 relatively strong and the AB11 relatively weak in patterns from relaxed muscle (A), but the intensities reversed in patterns from rigor muscle (C). Alongside these profiles are computed electron density maps based on these equatorial patterns and showing the changing mass distribution in the A‐band unit cell in the different states (D–F). These maps depend on an assumption about the phases of the reflections as discussed in detail in, for example, Harford et al. (1994) and Harford and Squire (1997). In the maps in Fig. 15, there is relatively little mass at the actin filament position in relaxed muscle (Fig. 15D), there is a great deal of mass at actin in rigor (Fig. 15F), and there is something in between in active muscle (Fig. 15E). The equatorial reflections can therefore be taken as indicators of head attachment to actin. However, it should be noted that it is also true that for a given number of myosin heads attached, the actual shape and configuration of the heads on actin will also change these equatorial intensities (e.g., Lymn, 1978); in principle, the equatorial intensities can also be used to monitor head shape as well as attachment number.

224

SQUIRE AND KNUPP

Fig. 16. (A) Low‐angle X‐ray diffraction pattern from permeabilized bony fish muscle with exogenous myosin heads (S1) added under rigor conditions. The actin layer lines are enhanced, but no new layer lines appear compared with patterns from relaxed muscle. The arrowed peak is the M3 reflections at 143 A˚ . (B) A similar pattern to (A) but from permeabilized fish muscle in rigor, but without added exogenous myosin heads. A new set of layer lines appears. (C) Helical reconstruction of labeled actin

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

B.

225

Labeling of Actin in Rigor Muscle

Figure 16A shows a low‐angle X‐ray diffraction pattern from a fish muscle into which an excess of exogenous myosin S1 had been soaked under rigor conditions (from Squire et al., 2005b). Such a procedure should randomly label any available actin sites, at the limit producing actin filaments through the I‐band and overlap region of the A‐band that are totally S1‐labeled. The effect on the diffraction pattern is as predicted. The actin layer lines (Fig. 16A) all become stronger and their peak intensity moves in toward the meridian, consistent with the heads producing an actin‐like helix of relatively large radius. No new layer lines are seen, since there is no new lattice geometry present. The 3D reconstruction in Fig. 16C is actually a helical reconstruction based on the observed X‐ray amplitudes from Fig. 16A combined with phases obtained from Dr. R. Milligan. The diffraction pattern from the S1‐labeled structure should be compared with Fig, 16B, which is from a skinned fish muscle without any exogenous heads, but in the rigor state. Of course, the original actin layer lines are enhanced once again, but, in addition, the M3 reflection is much stronger. It also extends along the layer line away from the meridian, and there are new layer lines that are not present in patterns from either relaxed muscle or S1‐labeled muscle. The reason that these new features are present is that in the overlap region of the vertebrate A‐band there are not enough heads to fully decorate actin filaments. The positions of attachment of the heads that do find a binding site are heavily dependent on the relative positions of the origin of the myosin head on the myosin filament and the position of the nearest available binding site on actin (Squire, 1972). Since the rigor attachment is one that is described as ‘‘stereo‐specific’’ (i.e., its geometry is fully determined in three dimensions; Fig. 16C, D), it is easier for a given head to attach to some actin monomers, which present the binding site in an appropriate orientation for myosin attachment, than it is to attach to other actin monomers, which may be relatively hard to reach. This constraint was fully described and illustrated by Squire (1972) in his analysis of the results of Reedy (1968) on insect flight muscle. It gave rise to what are termed actin target areas within which it is relatively easy for heads to bind. Subsequent independent analysis has supported the idea of target areas (e.g., Squire and Harford, 1988; Yagi, 1996; Koubassova and Tsaturyan, 2002; Steffen et al., 2001).

filaments based on the X‐ray amplitudes from (A) and phases courtesy of Dr. R. Milligan. This shows the general form of the rigor head shape on actin as in the green head shape in (D). Also shown in (D) is the lever arm in the pre‐powerstroke position (blue) discussed in detail in Geeves and Holmes (2005).

226

SQUIRE AND KNUPP

The new layer lines observed in X‐ray patterns from rigor vertebrate muscle can be explained approximately as orders of a repeat of about 5  429 A˚ ¼ 3  715 A˚ ¼ 2145 A˚ , where 429 A˚ is the myosin repeat, 715 A˚ is the actin filament pitch, and 2145 A˚ is the beat period between the two. The first myosin layer line (ML1 in Fig. 11) would then be the fifth order of this; the first actin layer line (A1; Fig. 9) would be the sixth order, and the M3 or ML3 layer line would be the fifteenth order. In the observed rigor pattern, intensity is seen on orders 5, 6, 9, 10, 11, 12, and 15 of the roughly 2145 A˚ repeat. The tenth‐ and eleventh‐order peaks are relatively close to the meridian (shown with arrows in Fig. 16D). One way to analyze this process is to use the program MusLabel, mentioned earlier, and to set up the appropriate geometry for the vertebrate muscle A‐band overlap region. The search ranges for the axial and azimuthal reach of a myosin head and the size of the actin target area can then be chosen and the program used to show what happens to the computed diffraction pattern. We know from previous studies (e.g., Lovell et al., 1981) that in rigor vertebrate striated muscles there is labeling of actin by virtually 100% of the available heads, so the MusLabel program can be used to determine which geometrical constraints on head movement are needed to yield 100% labeling. In fact, Squire et al. (2005b) assumed 98% labeling and obtained results such as those in Fig. 17, which shows a set of computed diffraction patterns from just the actin‐binding sites that are labeled by at least 98 % of the myosin heads under different search conditions. The first point is that there are many different ways to get virtually 100% labeling of actin by the heads. However, not all of these ways give strong layer lines where they are observed in the pattern from rigor muscle. In other words, the observed pattern can start to tell us which factors are dominant when a head is searching for a binding site on actin. Three distinct regions are outlined in the set of computed diffraction patterns shown in Fig. 17. Even though these patterns are computed from putative sites on actin filaments, region A is such that it gives rise to strong layer lines at positions 5, 10, and 15 (i.e., 429 A˚ , 215 A˚ , and 143 A˚ ), which are the same as those from resting myosin filaments (see Fig. 12). In other words, myosin‐like layer lines can be produced by myosin heads on actin. The unifying feature of these patterns in region A is that the angular size of the target area (twice the inset number in each frame) is generally very large. In summary, if the actin target area concept and the idea of stereo‐specificity are relaxed, the myosin heads just reach out and bind to the nearest actin monomer regardless of its azimuthal orientation. Heads attached to actin non–stereo‐specifically will give rise to a strong M3 reflection and to other myosin‐like layer lines, as well as making a contribution to the first actin layer line (A1).

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

227

Fig. 17. Sets of computed diffraction pattern simulations for different patterns of labeling of myosin heads on actin, defined by the head angular search range 
y, the head axial search range 
Z, and the actin target area angular size (twice the large number on each pattern), in each case with at least 98% of the available myosin heads bound to actin. For details of parameters and regions A, B, and C, see text. (Based on Squire et al., 2005b.)

Area B in Fig. 17 shows that the M3 layer line is very weak or absent, the other myosin‐like layer lines and the new low‐radius layer lines on orders 9–12 of 2145 A˚ are practically nonexistent, but the actin‐like layer lines (i.e., 6, 12 [high radius], 18, etc; 357.5, 179, 119 A˚ ) are very strong and at ever increasing radius. The particular point about this region is that the target areas on actin are very small and the heads need to search a large distance axially to reach a binding site. In doing this, they effectively wash

228

SQUIRE AND KNUPP

out the 143 A˚ ‐ness of their crown spacing and the myosin‐like features of the pattern disappear. The remaining region C is somewhere between these two extremes. All patterns in C have a strong first actin layer line (order 6 of 2145), they have a strong fifteenth order at 143 A˚ , they have intensity at low radius on layer lines 9 and 12, and in some cases they have a weak remnant of the fifth layer line (429 A˚ ) at relatively low radius, as observed. The extended nature of the observed 143 A˚ layer line in Fig. 6.15 is reproduced in region C, where the axial excursion (
Z) of the heads is small, the azimuthal excursion (
y) is quite large, and the actin targets for the better models are of the order of 2  75 ¼ 150 . (The lowest box in region C of Fig. 17 has the fifteenth layer line much stronger than the sixth, which is not observed). In one half‐turn of a long‐period actin helix containing 6.5 actin monomers, the actin monomer azimuth changes by 180 , so this analysis suggests that a given head can bind within a region that is 4 or 5 actin monomers long on one strand only (obviously the nearer) of the filament. Note finally that all of this analysis so far is independent of the shape of the myosin head. The analysis is like defining the lattice on which the heads are arranged (see Fig. 5) thereby defining which layer lines will be dominant. The shape of the head will then be a kind of motif convoluted with this lattice of binding sites. Electron tomographic analysis of rigor insect flight muscle by Chen et al. (2002) and Liu et al. (2004) has suggested that the strains on different myosin heads in rigor muscle will make the neck parts of the heads tilt back in different directions toward their own particular origins on the myosin filament, so not all attached heads will necessarily have the unstrained conformation in Fig. 15C and D. If this is generally true, this distribution of shapes will have its effect on the relative intensities of the rigor layer lines, but it will not generate ‘‘new’’ layer lines in addition to those on the 2145 A˚ lattice. A more detailed examination of head shapes in rigor will require full modeling of the rigor X‐ray diffraction pattern from either bony fish muscle or insect flight muscle where the lattice sampling is good. This has yet to be done.

C.

The Weak‐Binding State

The analysis of the crossbridge cycle in Geeves and Holmes (2005) emphasized the probable existence of a weak‐binding (non–stereo‐specific, pre‐ powerstroke) myosin head state before transition to a strong‐binding (stereo‐specific; post‐powerstroke) state. It would make sense if myosin heads can first dock loosely to actin, then become stereo‐specifically bound, and then go through their force‐producing cycle on actin. A highly populated weak‐binding state was, in fact, observed and studied in skinned

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

229

rabbit psoas muscle fibers by immersing the fibers in low ionic strength relaxing solution (Brenner et al., 1984; Yu and Brenner, 1989). The existence of this state was deduced from the following observations. First, the muscle was not ‘stiff’ to slow stretches, but became stiffer as the speed of stretch was increased. This implied that a substantial number of heads were reversibly and transiently bound to actin even in relaxing solution, so that in slow stretches a subset of actin‐attached heads could quickly come off actin and others could reattach somewhere else so that little stiffness would be recorded. On the contrary, in very fast stretches, many of the attached subset of heads would be attached essentially for the whole time and the muscle would appear stiff. Even faster stretches would trap even more heads and the muscle would appear stiffer. Second, equatorial diffraction patterns from these low–ionic strength fibers showed relative intensities for the AB10 and AB11 equatorial reflections, which were midway between relaxed and rigor; they were in fact not unlike patterns from active muscle (see Fig. 15B) with the AB10 and AB11 intensities about the same as each other, even though no tension was being produced. Third, even though these heads were clearly attaching to actin, hence the rapid stiffness, they still gave rise to myosin‐like layer lines in the low‐angle X‐ray patterns (e.g., Xu et al., 2002). We would interpret this (Squire and Harford, 1988) as indicating the presence of non‐stereo‐specifically‐bound actin‐attached heads which, as in region A of Fig. 17 (Squire et al., 2005b), would give myosin‐like layer lines.

D. Summary: Take‐Home Messages Work by Xu et al. (1999; and references to others there) has tended to confirm that good helical ordering of the myosin heads on rabbit psoas muscle thick filaments requires the heads to be in the M.ADP.Pi state at temperatures above
20  C. This probably occurs in others muscles, too, but fish and frog muscles, for example, normally operate at much lower temperatures (5 –15  C) than rabbit (37  C), which precludes the possibility of taking them over 17  C down from their normal operating temperature to test this. However, assuming that M.ADP.Pi is the predominant state in ordered resting muscles, then the next step in the Lymn– Taylor (1971) scheme is AM.ADP.Pi, a state discussed in some detail in Geeves and Holmes (2005). The AM.ADP.Pi state is taken to be initially a non‐stereo‐specific, transiently attached, state largely controlled by electrostatic interactions (Wong et al., 1999). The strong states such as AM.ADP and AM follow from this. A number of attempts to trap heads in the elusive AM.ADP.Pi state in intact muscle, sometimes using ATP analogues (e.g., AMP.PNP), as well as

230

SQUIRE AND KNUPP

other studies with heads largely in defined states (e.g., Xu et al., 2002; Yagi et al., 1998; and others) have tended to confirm this general picture of varying populations of non‐stereo‐specific states and stereo‐specific states in the crossbridge cycle. Results from ‘‘steady‐state’’ muscles have led to the following general conclusions: 1. Stereo‐specifically attached heads as in rigor give rise to a mixed population of myosin‐like and actin‐like layer lines, including a 143 A˚ (M3) reflection. They show good evidence for actin target areas. 2. Non‐stereo‐specifically attached (such as weak‐binding) heads, even though on actin, give a myosin‐like layer line pattern including a 143 A˚ (M3) reflection. This also includes a contribution to the first myosin layer line (ML1) and the first actin layer line (A1; Ferenczi et al., 2005; Harford and Squire, 1992; Xu et al., 1999). 3. Both stereo‐specifically attached and non‐stereo‐specifically attached heads can change the equatorial intensities. For a given attachment number, the observed intensities depend on the nature of the binding and shape of the myosin head (Lymn, 1978; Squire and Harford, 1988). The AB11/AB10 intensity ratio increases with increasing attachment number, and, for a given attachment number, it increases with increased relative populations of rigor‐like or stereo‐specifically attached heads. 4. Following on from (3) above, non‐stereo‐specifically attached heads give an AB11/AB10 ratio of about 1, as in the low ionic strength relaxed state of Brenner et al. (1984). Transition to the strongly attached state reduces the AB10 intensity, which therefore changes in step with tension, but does not greatly alter the AB11 intensity. 5. In active or rigor muscle, heads that do not overlap actin filaments become disordered (Cantino et al., 2002; Padron and Craig, 1989). They therefore contribute little to the observed low‐angle X‐ray diffraction patterns. This population increases with increasing sarcomere length (reduced filament overlap).

III.

Time‐Resolved Events in Contracting Muscles A.

Changes in the Equatorial Reflections

Following the pioneering, laboratory‐based, time‐resolved X‐ray diffraction studies of muscle by H. E. Huxley and separately by G. F. Elliott (Elliott et al., 1965; 1967; Huxley, 1972; Huxley and Brown, 1967; Huxley et al., 1965), a number of groups, including that of Huxley, have followed

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

231

up these experiments on frog and several other muscle types, in recent years using much stronger and brighter synchrotron beam‐lines and very high‐speed electronic detectors (e.g., RAPID; Lewis et al., 1996, 1997). These include further studies by Huxley et al. (1980, 1981, 1983, 2003) and by many others, including Bagni et al. (2001); Bordas et al. (1993); Cecchi et al. (1991, 2003); Martin‐Fernandez et al. (1994); Piazzesi et al. (1999); Juanhix et al. (2001); and Kraft et al. (1999; 2002). Further studies of very fast transients in intact frog single fibers have been carried out by (Dobbie et al., 1998; Irving et al., 1992, 1995, 2000; Linari et al., 2000; Lombardi et al., 1995; Piazzesi et al., 2002; Reconditi et al., 2003, 2004) and studies of permeabilized rabbit fibers with applied temperature jumps and other interventions (Bershitsky et al., 1996; 1997; Ferenczi et al., 2005; Tsaturyan et al., 1999a,b). We have made use of the muscles of bony fish, because of their high degree of order and well‐sampled diffraction patterns (Harford and Squire, 1986; 1990; 1992; 1997; Harford et al., 1991; Mok et al., 2005; Squire, 1998; Squire et al., 1994; 2003a,b,c; 2005a) and have carried out related time‐resolved experiments with tetanically contracting intact muscles. The reasons for preferring bony fish muscles is the very high degree of simple lattice order in the bony fish muscle A‐band (discussed in Squire et al., 2005; Luther and Squire, 1980; Luther et al., 1981; Harford and Squire, 1986), which makes the diffraction patterns highly sampled and therefore much more amenable to proper interpretation (e.g., Hudson et al., 1997) than superlattice muscles. Figure 18 shows a difference fish muscle diffraction pattern, active minus relaxed, between the tetanus plateau and the resting phase before stimulus, and Fig. 19 shows intensity and tension time courses on a millisecond time scale in the early part of fish muscle tetanic contractions (from Mok et al., 2005). Note that these are all shown as normalized changes, with rest as 0% and the plateau set as 100% for intensities that increase on activation and vice versa for decreasing intensities; they are not absolute changes. Both figures illustrate the kinds of changes that are observed in most of the quoted X‐ray studies. A number of characteristic features are observed in the difference map in Fig. 18. Generally, as a result of activation, the myosin layer lines become weaker (black arrows), whereas much of the actin pattern becomes stronger (white arrows), particularly the second actin layer line (A2), which is associated with the tropomyosin shift discussed in Squire et al. (2005) and Brown and Cohen (2005). However, the outer part of the first actin layer line drops a little (black double arrow). These features are put on a quantitative basis in Fig. 19. As expected from the intensity changes seen in Fig. 15(C–E), the effect on the equator is that during the rising phase of tension the AB10 intensity decreases and the AB11 intensity increases as more heads become attached to actin. However,

232

SQUIRE AND KNUPP

Fig. 18. Difference intensity map between diffraction patterns from a fully active and a relaxed fish muscle (Mok et al., 2005). Generally, dark colors show intensity drops and green, yellow, and red show intensity increases. Generally, the myosin layer lines have dropped in intensity (black arrows), and many of the actin layer lines have increased in intensity (white arrows), especially layer line A2. However, the outer part of A1 has dropped in intensity (double‐headed black arrow). There are also clearly some shifts in axial spacing of the peaks; these are especially visible along the meridian.

an interesting observation is that the time courses of the intensity changes of these two reflections are not the same. The AB10 intensity changes almost in step with tension (even though one increases while the other reduces), whereas the change in the AB11 reflection is well ahead of tension (about 20  1 ms before tension (T) at level of 50% change; Fig. 19A). The time to 50% tension and AB10 change was 37  1 ms after activation. Another feature that increases rapidly and well before tension is the intensity of the second actin layer line (25  1 ms before T; A2 in Fig. 19A), interpreted as showing that tropomyosin moves as an early event

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

233

Fig. 19. Tension, intensity, and spacing time courses from the rising phase of tetanic contractions in bony fish muscle (from Mok et al., 2005). All changes have been normalized to be 0% in resting muscle and 100% at the tetanus plateau for all changes that are increases, and vice versa for all changes that are decreases. (A) Shows the changes of the tension (T), the A1 and A2 actin layer lines, and the A11 (11) equatorial reflection. (B) Shows the M3 spacing and intensity relative to tension (T). (C) Shows the changes of the myosin layer line ML3 and of ML1 at the A11 and A20 positions. For details, see text.

in muscle activation (Kress et al., 1986). Other changes shown in Fig. 19C are the drop of the myosin layer lines, ML1 and ML3. The meridional part of ML3, the much studied M3 reflection, is actually a set of closely spaced peaks produced by interference effects, as discussed in the next section.

234

SQUIRE AND KNUPP

However, if these components cannot be resolved, it appears that the M3 reflection changes spacing as a peak at
143.4 A˚ gradually reduces and a peak centered at
145.7 A˚ develops. Fig. 19B shows that the intensity drop of M3 is very rapid (26  1 ms before T), but that the apparent spacing change is later (9  2 ms before T), showing that the increase of the 145.7 A˚ peak is a later event than the decrease of the 143.4 peak. All of these observations are consistent with the following scenario. On activation, Ca2þ ions are released into the sarcomere where they bind to troponin C, causing a tropomyosin shift that shows up as an early change in the A2 actin layer line (Fig. 19A). As soon as the thin filament is switched on, a small population of heads, which are presumably attaching transiently to actin even in resting muscle, can now move into stronger actin interactions, other heads are pulled over to replace them, and the myosin layer lines gradually become weaker. This loss of helical order in the myosin head array is presumably quite cooperative. Heads labeling the actin filament will generally make the actin layer lines stronger (as in the rigor diffraction pattern in Fig. 16B), but it should be noted that an effect of the tropomyosin shift that increases the A2 intensity is also to decrease the A1 intensity, and evidence for this is also seen in Fig. 18 (Mok et al., 2005; Parry and Squire, 1973; Squire et al., 2005a ; Yagi, 2003). Our evidence that the AB11 intensity change precedes the AB10 change in time‐resolved studies of bony fish muscle (Harford and Squire, 1992) suggests that the heads must attach to actin in an initial conformation, which gives a relatively strong AB11 reflection, but changes the AB10 peak rather little, and that they must then move over to another conformation on actin, which changes (reduces) the AB10 reflection. Since the weak‐binding bridges observed by Brenner et al. (1984) changed the AB11 intensity rather more than the AB10, we interpreted the first attached state as being like a weak‐ binding, non‐stereo‐specific, state, whereas the second, later, attached state must be more like rigor.

IV.

X‐Ray Interference Measurements and Their Implications A.

Introduction to the Idea of Interference

Striated muscles are so beautifully ordered in the axial direction that interference effects can be seen in the observed low‐angle diffraction patterns. The first clear evidence for this came from a heroic study by the late John Haselgrove (1975) using a very long camera on a laboratory X‐ray generator. Figure 20 shows a trace of the meridional part of the low‐angle X‐ray diffraction pattern from frog muscle recorded by Haselgrove. What is

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

235

Fig. 20. Meridional intensity profile from frog sartorius muscle recorded on a long laboratory‐based X‐ray camera by Haselgrove (1975). This shows the effects of interference on peaks arising from C‐protein, troponin, and myosin. For details, see text.

apparent is that the reflection referred to so far as the M3 meridional reflection is in fact a set of closely spaced peaks, with that at 143.4 A˚ dominant. The M2, M4, and M5 peaks are also multiples. In addition, there is a strong pair of peaks at
442 A˚ and 418 A˚ and another pair at
396 A˚ and 382 A˚ . Rome et al. (1973a,b) showed by antibody‐labeling that these latter two pairs were from C‐protein and troponin, respectively. The origin of these closely spaced peaks was very quickly shown to be the interference effects observed within the sarcomere because, in the case of C‐ protein, the diffraction patterns from the two C‐zones in a single A‐band would interfere, and, in the case of troponin, the diffraction patterns from the two troponin arrays across the Z‐band would interfere. Also, in the case of the M3 multiple, the diffraction from the myosin heads in the two bridge regions of a single A‐band would interfere (see summary in Squire, 1981; pages 576– 582). In the case of the C‐zone interference, illustrated in Fig. 21A–C), the diffraction intensity profile from a single C‐zone (A) would have a prominent peak at
430 A˚ , but the two C‐zones in one A‐band would be centered a distance L apart (Fig. 21C). The two C‐zones could then be considered as

236

SQUIRE AND KNUPP

Fig. 21. The origin of the interference effects seen in Fig. 20. (A–C) show the analysis of the C‐protein array by Squire et al. (1982). Each C‐zone has an array of seven C‐protein peaks (C) in each half of the A‐band. One such array will give to a diffraction peak at around 434 A˚ (A). The two C‐zones will be spaced L apart and the simple effect of this is to modulate the pattern in (A) by a set of fringes (Cos‐squared fringes) at orders of 1/L. The effect of this is to split the main peak in (A) into subpeaks at
441 and 416 A˚ . Putting the underlying peak at 434 A˚ rather than myosin repeat of 429 A˚ has the effect of making the 441 A˚ peak stronger than the 416 A˚ peak as observed (see Fig. 20). (D) and (E) A similar scenario to (A–C) but for the interference effects across the Z‐band for the troponin arrays on the thin filaments, where the interference distance in (D) is called R and the troponin peak in (E) is split into components at 396 and 382 A˚ . (F) and (G) Changes in the interference distance L that might occur with tilting myosin heads or tilting lever arms. The original interference distance L1 reduces to L2 if the heads or lever arms tilt. (From Squire, 1981.)

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

237

a diffracting object as in Fig. 2, with the separation of the holes in the mask equal to L rather than the distance in Fig. 2. In Fig. 21C, the C‐protein array can be considered to a first approximation as a convolution of one C‐zone with a set of two points a distance L apart. To get the full diffraction pattern it is then necessary to multiply the pattern from one C‐zone, as in Fig. 21A, by the diffraction pattern of the two points. This consists of a series of peaks (Cos‐squared fringes) at positions related to 1/L. Since L is very large, these peaks will be very close together. The result (Fig. 21B) is that the pattern in (A) from one C‐zone is crossed by a set of closely spaced interference fringes of separation related to 1/L. In their analysis of the C‐zone, Squire et al. (1982) found that they needed to put the underlying spacing of the C‐protein stripes in each C‐zone to
432 A˚ , rather than at the myosin repeat of 429 A˚ , in order to make the fringe at 441 A˚ much stronger than that at 416 A˚ . The reason for this remained a puzzle for an extended time—one raised initially by Sjostrom and Squire (1977) from their analysis of C‐zones in electron micrographs of cryo‐sectioned muscle. It was not until it was realized that, although anchored on myosin, part of C‐protein might also bind to actin in resting muscle that the longer repeat could be explained (Squire et al., 2003b; see Figs. 13 and 14). The interference effects of the troponin peaks and the intensities from the myosin heads were explained in the same general way as the C‐protein peaks, but with different values for the interference spacing L. The idea for the I‐band is shown in Fig. 21D and E, where it can be seen that the interference spacing, here called R, causes the sampling of the underlying 385 A˚ troponin peak at two positions (
396 and 382 A˚ ). In the case of the myosin heads, Squire (1981) raised an interesting point that has now been followed up by systematic studies as discussed in the next sections. Fig. 21G represents the myosin head arrays in the two bridge regions of a single A‐band. This would have an interference spacing called L2. However, if the myosin heads were to tilt toward the M‐band, as for example might occur in pulling an actin filament toward the M‐band, then the interference spacing would reduce by some amount down to L1. (Note that Squire [1981] originally discussed this in terms of stretching rigor muscle, but the principle is exactly the same.) It has now been found that such changes in interference spacing provide an exquisitely sensitive probe of myosin head motion, as discussed in the next section.

B.

Interference Observations and Their Possible Interpretations

Much of the recent work on interference measurements has been specifically aimed at probing crossbridge activity; for this reason the main targets for study have been the M3 and M6 meridional peaks. Looking

238

SQUIRE AND KNUPP

first at the M3 peak, we have seen that the dominant peak from resting frog muscle is at 143.4 A˚ . This peak will be referred to as M3rel. In patterns from active muscle the 143.4 A˚ peak gets weaker and a new set of peaks of mean spacing 145.7 A˚ gets stronger (Linari et al., 2000; Mok et al., 2005; Oshima et al., 2003; Piazzesi et al., 1999). Evidence was presented in Fig. 19B for the relatively late apparent change of spacing of the unresolved M3 peak. This has been taken to mean that the 143.4 A˚ component drops relatively quickly as the resting myosin head helix is lost, and the new set of peaks with a mean spacing of 145.7 A˚ characteristic of active muscle takes some time to increase. Linari et al. (2000) interpreted these observations in terms of a cooperative spacing change along the whole myosin filament. In this study the active peak of mean spacing 145.7 A˚ was observed to be split because of the interference between the two bridge regions into peaks at 144.6 A˚ and 146.7 A˚ . The final result in fully active muscle is a pair of peaks of about the same intensity. We refer to these two peaks as M3i (146.7A˚ ) and M3o (144.6 A˚ ). To summarize, in resting muscle M3rel  M3i and M3o and at the plateau of an isometric tetanus M3rel < M3o M3i. The measured intensity ratio of M3o to M3i in a tetanus was 0.76 to 0.8 (Huxley et al., 2003; Reconditi et al., 2003). Note that the M6 meridional reflection was also split into two peaks centered at 71.7 A˚ in patterns from resting muscle and 72.9 A˚ in patterns from active muscle. Again there were two components, and the active intensity ratio M6o (72.61 A˚ ) to M6i (73.12 A˚ ) was 0.55 (Reconditi et al., 2003). The elegant and informative studies of Irving, Lombardi, and their collaborators (e.g., Linari et al., 2000; Piazzesi et al., 1999; Reconditi et al., 2003) have suggested a possible set of crossbridge events that might be involved in contraction. First of all they showed that the intensities of the M3o/M3i pair both gradually reduced as the sarcomere length increased (Linari et al., 2000); there was an almost linear reduction which scaled with the amount of filament overlap which directly mimicked the change of tension with overlap. This implied that the intensities were coming from myosin heads attached in some way to actin and with a constant fraction producing force. It also implied that the myosin heads that were not overlapped by actin were too disordered to contribute strongly to the observed M3 peaks. Second, they showed that as the sarcomere length increased, the spacings of the two M3 components gradually changed, consistent with the length L of the interference function gradually increasing. The starting value of L at short sarcomere lengths was
8669 A˚ with the two M3 peaks being the fifth‐ninth (M3i) and sixtieth (M3o) orders. Since it was envisaged that the increase in sarcomere length would result in crowns moving out from the overlap region in symmetrical pairs on each side of the

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

239

A‐band, the interference function L would gradually increase to LN ¼ L þ 145.7 N for N crown pairs lost from overlap. In other studies, Dobbie et al. (1998) and Bagni et al. (2001) reported the effects of applying sinusoidal length changes to single frog muscle fibers. They found that, with total length changes of 50 A˚ per half‐sarcomere and a frequency of
3100 Hz (320‐ms period), the M3 intensity changed in direct anti‐phase to the force. An increase in length produced a lower M3 intensity and vice versa. However, when similar length changes (40 A˚ per half sarcomere, 320‐ms period) were applied to rigor fibers, the M3 intensity changes were in‐phase with the length change. Further studies of this kind were performed by Cecchi et al. (2003), who applied sinusoidal length changes of various amplitudes and frequencies to intact fiber bundles of frog muscle. They found that the shape of the intensity change was not necessarily sinusoidal, but with the application of large‐amplitude oscillations the intensity showed double peaks at the top of each cycle but unsplit peaks at the bottom. This effect was less obvious at high frequencies where the imposed sarcomere length change was less. Recent studies on other muscle types include beautifully conducted studies of the flight muscles of Drosophila actually flying in the synchrotron X‐ray beam (Dickinson et al., 2005). This showed dramatic oscillatory changes in the 145 A˚ and 72 A˚ meridional reflections and in some of the layer lines such as the 387 and 193 A˚ actin layer lines. The 145 A˚ and 72 A˚ reflections increased in intensity during the upbeat of the wing (muscle stretch) and reduced on the downbeat (muscle shortening). The 193 A˚ layer line behaved in a similar way except for being double‐peaked at the top of the beat, whereas the 387 A˚ reflection changed in anti‐phase to this, without any splitting. All of these observations were interpreted largely in terms of coordinated myosin lever‐arm movements, as detailed below.

C. Interference Changes During Muscle Transients Several groups have studied the effects on the muscle low‐angle diffraction pattern of applying various mechanical perturbations to steady‐state structures, either isometric contractions or rigor muscle at various strain levels. Huxley et al. (1981, 1983) used whole frog muscles and followed the effects of step changes of length of various amplitudes applied at the plateau of an otherwise isometric tetanus. They studied the effects on the M3 intensity as a whole. More recently, with the two components of the active M3 resolved, Huxley et al. (2003) and Reconditi et al. (2003) have studied the separate behavior of these components. Huxley et al. (2003) found that the intensity ratio of the M3o to M3i varied from an initial value

240

SQUIRE AND KNUPP

of
0.8 (cf. 0.76 from Reconditi et al., 2003), down to a value of about 0.3 for a length step of about 60 to 70 A˚ . A close study showed that the total M3 intensity increased slightly (10%) for shortening steps up to 15 A˚ , and then decreased systematically for larger steps. These various observations, including the effects of applied oscillations in the case of vertebrate muscle have been interpreted in terms of the swinging lever arm hypothesis. In isometric muscle the lever arm is supposed to be such that the projected myosin head density is just slightly to one side of the position that would give maximum M3 intensity. Oscillations of the muscle here would tilt the lever arm by different amounts and would increase the M3 intensity for releases and decrease it for stretches. On the other hand, with a highly tilted lever arm in rigor muscle (Reconditi et al., 2003), an increased M3 intensity will be produced by a stretch and a decreased intensity by a release. Cecchi et al. (2003) then argued that, with higher‐amplitude stretches, the release side of the oscillation at the isometric tetanus plateau would take the lever arm over the maximum intensity, and the intensity would drop slightly at the maximum displacement position before returning back through the maximum intensity position and down again. The intensity behavior during stretches and releases would then be different. The final aspect of this story is what happens to the relative intensities of the two components of the M3 reflection. Linari et al. (2003) argued that during a quick release the lever arms would tilt inwards toward the M‐band from their original isometric positions, thus at first increasing the M3 intensity as observed by themselves, and by Huxley and others, but this would then change the mean interference spacing to a slightly shorter value, thus moving the positions in the underlying M3 peak that are being sampled by the interference peaks. If the value of L reduces (Fig. 21F compared with G), then the M3o to M3i will change; M3i will move toward the middle of the underlying peak and M3o will move toward the edge so that the ratio M3o to M3i will drop. In fact, the observed starting ratio of
0.8 drops to
0.3 for a length step of about
60 to 70 A˚ (Huxley et al., 2003). In summary, although there are still some anomalies, all of these observations can be interpreted in terms of the swinging lever arm hypothesis.

D. Further Details of the Interference Experiments Note that the response of the M6 reflection is very different to that of the M3. It has been concluded, quite reasonably, by Huxley et al. (2003) and by Linari et al. (2000) that most of the M6 intensity must come from densities other than the myosin heads, perhaps the myosin filament

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

241

backbone. The M6 intensity appears to change rather little with changing overlap or during transients, so it clearly is from a different source from the M3. Huxley and others had the problem that, to get satisfactory modeling of the M3 reflection, the M6 intensity from the heads came out to be much too high and they had to introduce an artificial disorder factor (dispersion) to knock down the M6 peak. In fact, this is not necessary. The problem encountered by these authors was that they were using a simplistic model of the head array on actin—one in which the heads are assumed to be evenly spaced at 145.7 A˚ intervals along the thin filament. In fact, the lack of intensity of the M6 peak from active heads is a simple result of the fact that none of the myosin heads attached to actin are actually separated by 145.7 A˚ . There are no actin monomers that are this distance apart. The possible separations of actin monomers in a single target area on one strand are 55, 110, and 165 A˚ . It is the fact that the heads are coming from a 145.7 A˚ repeat on myosin, which means that they label actins in such a way that on average the repeat on actin is
145.7 A˚ , even though no two actin‐attached heads are actually separated by this distance. Because of this, there is a kind of true built‐in dispersion of the myosin head positions on actin away from a simple 145.7 A˚ repeat, and, if the M6 is calculated from heads on properly defined actin‐binding sites (for example, using MusLABEL; Squire and Knupp, 2004), then it is found that the M6 peak contribution from the heads is automatically very weak. In summary, it seems very probable that the M6 reflection has very little contribution from the myosin heads and that what is seen must be coming mostly from other densities associated with the myosin filament, such as the backbone. But it is not necessary to use an artificial dispersion term to account for this. Following from this, it is not correct to calculate interference effects as though there are two identical sets of head arrays in the two halves of the A‐band, apart from the lever arms tilting in opposite directions on each side of the M‐band. As shown in Fig. 22, if the two arrays (W) are identical, then they can be related by a simple displacement L (Fig. 22A) and the interference function is a simple Cos‐squared function (Fig. 22C). In other words, for every peak in the array on the right at a distance of, say, x from the middle of the M‐line, there is a peak in the left hand array at x – L. This is analogous to structure illustrated in Fig. 21C. However, the labeling pattern of heads on actin, as discussed in the section on rigor muscle, does not have a simple, uniform axial periodicity of 145.7 A˚ ; because of the mismatch of the myosin and actin arrays there must be a much more complex pattern of labeling. Such labeling patterns can be explored using MusLABEL for different numbers of heads bound. The important point, however, is that under any particular conditions, and assuming identical overlap in the two halves of the same sarcomere, the pattern of labeling in

242

SQUIRE AND KNUPP

Fig. 22. The effect of the form of the scattering object (W) on the interference function. In (A) the two arrays labeled W are related by simple translational symmetry. The array on the right is obtained by shifting the array on the left by L toward the right. This means that the full transform of the array in (A), shown as (B), is a product of the transform from one W array and a set of equidistant Cos‐squared fringes as in (C). The fringe spacing is related to 1/L. In (D) the diffracting object is one array W on the right and a similar array ‐W on the left, except that they are mirror images; they are not related by a simple translation. Interference effects still occur (E), but the interference function is far from simple (F); it consists of unevenly‐spaced peaks whose positions are not easily predicted. (From Knupp and Squire, 2005.)

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

243

one half of the A‐band will have a definite polarity. But this polarity will reverse in the other half of the same A‐band. The densities of the two arrays in projection onto the muscle long axis, giving the density distributions that give rise to the meridional intensities like M3, will not be related by a simple axial translation, but will be mirror images of each other. This can have profound effects on the interference function. This is shown in Fig. 22D–F for a very simple array, but where the array on the left (W) has the opposite polarity to the array on the right (‐W); they are not related by a simple translation L. In fact, this time, for every peak (head) on the right of the structure at þx from the middle of the M‐line, there is an equivalent peak at –x on the left (not at x – L as above). Figure 22E shows the resulting, complex, full, interference pattern that results, and Fig. 22F shows the interference function alone. The peaks in the latter are clearly very unevenly distributed compared to the Cos‐squared function in Fig. 22C. This might be an extreme example, and these effects will certainly be less marked in the muscle case where the two interfering arrays are quite long, but since the muscle meridional interference is so sensitive and is being interpreted so fully, it is well to model it correctly. This can now be done in a new version of MusLABEL, which includes interference effects (MusLABEL2; Knupp and Squire, 2005). Note finally that, for reasons given earlier, in the sarcomere length series of Linari et al. (2000), as the sarcomere length increases, the axial extent of the crossbridge arrays on actin will gradually reduce (Fig. 23). For this reason the M3 peak will weaken in line with sarcomere length, as observed by Linari et al., but the underlying 145.7 A˚ –spaced peak, which is being sampled by the interference function, will gradually become broader. Its size is inversely related to the array width (see Fig. 5). Figure 23C–E show what would be expected; each is a profile of the M3 region calculated using MusLABEL2 for different sarcomere lengths (S). Although a double peak is seen for rest length muscle (S ¼ 2.2 mm; Fig. 23C), two strong satellites appear on each side of this pair at 2.8 mm (Fig. 23D), and a wide array of 4 to 6 quite strong peaks appears at S ¼ 3.2 mm (Fig. 23E). In the published profiles of Linari et al. (2000), such an increase in the number of satellite peaks was not obvious. It may be that there was, in fact, a gradually broadening dispersion in the sarcomere lengths in their preparations as the mean sarcomere length increased; each sarcomere length would then give a slightly different interference effect. In this case what would be observed experimentally would be a mixture of slightly different patterns all centered on the same double peaks, and the outer peaks might then tend to smear out. This needs to be tested, but, in any case, the profiles in Fig. 23 show a predicted difference from observation that needs to be accounted for.

244

SQUIRE AND KNUPP

Fig. 23. Simulation of the effects of changing sarcomere length on the M3 reflection. As the sarcomere length increases from Sa in (A) to Sb in (B), the axial extent (W) of the overlapped region of A‐band reduces from Wa to Wb. The effect of this on the M3 region (C) to (E) is that the peak being sampled by the interference function gradually broadens (see Fig. 6). An essentially double peak in (C) at S ¼ 2.2 mm has two strong satellites for S ¼ 2.8 mm (D) and becomes a set of up to six quite strong peaks at S ¼ 3.2 mm. All simulations were carried out using MusLABEL2 (Knupp and Squire, 2005).

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

245

E. Temperature Jump Experiments and Isotonic Contractions So far all the results that have been discussed (except those from Drosophila) have come from isometrically contracting muscles with or without rapid applied length steps. Interesting alternative probes of crossbridge behavior, apart from imposing length steps, are either to rapidly change the temperature in a muscle fiber and to monitor the effect (Bershitsky and Tsaturyan, 2002; Ferenczi et al., 2005; Griffiths et al., 2002; Tsaturyan et al., 1999a) or to allow the muscle to shorten under a fixed load (Reconditi et al., 2004). The temperature jump experiments (e.g., from 5 to 30 C in rabbit psoas fibers) produce a gradual (almost exponential) increase in tension over a 30‐ to 40‐ms period, without at the same time greatly changing the fiber stiffness. At the same time there is an intensity increase of the actin first layer line (A1) and of the M3 meridional reflection. However, the latter reduces slightly in the first few milliseconds before increasing substantially with a similar time course to tension. These results were interpreted in terms of myosin heads already bound to actin in an initial weak‐binding, non‐stereo‐specific state changing to a force‐producing stereo‐specific state after the temperature jump. It was argued that the non‐stereo‐specifically attached heads would contribute relatively little to the actin first layer line and that the observed increase reflected a changeover to the stereo‐specific state. It was then argued that the change in M3 would be due to lever‐arm swinging as in previous studies. In the other kind of experiment, Reconditi et al. (2004) used an applied tension step to produce isotonic shortening in an initially isometrically contracting fiber. Since the tension step is very fast, everything that happens after it is at constant load, which has the great advantage that there will be very much reduced alterations in compliance in the system. The myosin and actin filaments are known to stretch slightly under load (see Huxley et al., 1994; Wakabayashi et al., 1994) and this compliance needs to be considered in interpreting data from fibers producing different tensions. The isotonic contractions avoid this. Reconditi et al. (2004) found different behavior for different levels of load after the tension step. For low loads (0.25 T0) the fibers obviously shortened quite rapidly, whereas the speed of shortening gradually reduced for higher loads. They also found that the behavior of the two components of the M3 peak (M3o to M3i) was different in each case. In each case, the ratio (RM3 ¼ M3o/M3i) dropped down and then increased again over different elapsed times, but the whole curve for the high load (0.75 T0) dropped and increased again within a length change of
70 A˚ , whereas for 0.5 T0 this extended to
100 A˚ , and for 0.25 T0 it extended to
140 A˚ or more. It was argued that this reflected

246

SQUIRE AND KNUPP

different sizes of the myosin head working stroke under different load conditions.

V. Summary In conclusion, we note that many people tend to think about the crossbridge cycle in terms of local events between a single myosin head and an actin monomer. This is the philosophy in Geeves and Holmes (2005) where crystal structures of myosin heads in various states, together with electron microscopy of actin filaments labeled with myosin heads, are used to analyze the crossbridge cycle. This is, indeed, a powerful approach. However, to fully understand how muscle works, especially in the aspects of how strain and other factors influence the crossbridge cycle in situ, techniques such as time‐resolved X‐ray diffraction are essential. Our argument is that in this case it is not sufficient to treat the sarcomere in a simplistic way; this will lead to misleading conclusions. The A‐band can, in fact, be treated much more thoroughly and realistically, and in a way that can be trusted, if all our current knowledge of muscle ultrastructure is brought to bear on the problem. This applies to the labeling pattern of heads on actin in rigor muscle, it applies to similar patterns that are likely to occur in active muscle, it applies to analysis of the interference effects seen on the meridian, and it applies to the changing with time, temperature, and tension of the observed myosin and actin layer lines. We finish where we started by emphasizing the enormous power of the X‐ray diffraction technique applied to muscle. But we also caution those using the technique not just to follow the current fashions, which will not help the field to move forward, but to consider any alternative possible explanations for their observations in detail and to test these alternatives against each other so that the conclusions that are reached and published are much more likely to stand the test of time. We also still have a great deal to learn about structures in the sarcomere. The recent X‐ray diffraction work by Squire et al. (2004) on line ID–02 at the ESRF in Grenoble, France, using a 10m camera and the highly ordered bony fish muscle has revealed a whole set of diffraction peaks not previously observed in X‐ray patterns from any vertebrate muscle. It is evident, from the good sampling in the observed fish muscle patterns, that the new features come from the A‐band. Some of these features can be explained by structures that are already known, such as by troponin on the A‐band parts of the thin filaments and by C‐protein, but other peaks still await a sensible interpretation. It is essential to use the most highly ordered muscles (bony fish muscle or insect flight muscle; see Squire et al., 2005) to fully

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

247

solve the structures of the sarcomeres in different muscle types and states (e.g., Al‐Khayat et al., 2003; Hudson et al., 1997; Squire et al., 1998) and thus to reach sensible conclusions about the myosin crossbridge cycle that will stand the test of time. Clearly there is a still much to learn about the contractile mechanism, force production, the strain dependence of the crossbridge cycle and, indeed, how force and movement are actually generated. As shown in Squire et al., Granzier and Labeit, Brown and Cohen, and Geeves and Holmes (all in this volume), and in this article, a great deal is already known about muscle structure and function. However, there is also little doubt that the sarcomere still has some major surprises in store for us.

Acknowledgments Much of the work on fish muscle and the development of the programs HELIX, MusLABEL, and MusLABEL2 reported here has been supported by grants from the UK Medical Research Council, the Biotechnology and Biological Sciences Research Council, the Wellcome Trust, the Leverhulme Trust, and the British Heart Foundation. We acknowledge our indebtedness to these organizations for their support. We also thank our many colleagues who have contributed to much of the work reported here, in particular Dr. Michael Sjostrom, Dr. Edward Morris, Dr. Liam Hudson, Dr. Richard Denny, the late Dr. Helen Pask, Dr. Jeffrey Harford, Dr. Michael Reedy, Dr. Tom Irving, Ngai‐Shing Mok, Ganeshalingam Rajkumar, and Dr. Andrew He. None of the synchrotron X‐ray work could have been carried out without the help of the beam‐line staff on lines 2.1. and 16.1 at the Daresbury Laboratory, UK; on line ID–02 at the ESRF in Grenoble, France; and on the BioCAT beam‐line at the Advanced Photon Source at the Argonne National Laboratory, USA. We greatly appreciate their many and varied contributions. In particular, we acknowledge the technical excellence of the development team and the foresight of the funding agencies shown by the production of the RAPID X‐ray detector at the CLRC Daresbury Laboratory (Lewis et al., 1996; 1997) ,without which millisecond and sub‐millisecond time‐resolved two‐dimensional X‐ray diffraction studies of muscle would not be possible.

References Al‐Khayat, H. A., Yagi, N., and Squire, J. M. (1995). Structural changes in actin‐tropomyosin during muscle regulation: Computer modelling of low‐angle X‐ray diffraction data. J. Mol. Biol. 252, 611–632. Al‐Khayat, H. A., Hudson, L., Reedy, M. K., Irving, T. C., and Squire, J. M. (2003). Myosin head configuration in relaxed insect flight muscle: X‐ray modelled resting crossbridges in a pre‐powerstroke state are poised for actin binding. Biophys. J. 85, 1063–1079. Bagni, M. A., Colombini, B., Amenitsch, H., Bernstorff, S., Ashley, C. C., Rapp, G., and Griffiths, P. J. (2001). Frequency‐dependent distortion of meridional intensity changes during sinusoidal length oscillations of activated skeletal muscle. Biophys. J. 80, 2809–2822.

248

SQUIRE AND KNUPP

Bershitsky, S. Y., and Tsaturyan, A. K. (2002). The elementary force generation process probed by temperature and length perturbations in muscle fibers from the rabbit. J. Physiol. 540, 97–988. Bershitsky, S. Y., Tsaturyan, A. K., Bershitskaya, O. N., Mashanov, G. I., Brown, P., Burns, R., and Ferenczi, M. A. (1997). Muscle force is generated by myosin heads stereo‐ specifically attached to actin. Nature 388, 186–190. Bershitsky, S., Tsaturyan, A., Bershitskaya, O., Mashanov, G., Brown, P., Webb, M., and Ferenczi, M. A. (1996). Mechanical and structural properties underlying contraction of skeletal muscle fibers after partial 1‐ethyl‐3‐[3‐dimethylamino)propyl]carbodiimide cross‐linking. Biophys. J. 71, 1462–1474. Bordas, J., Diakun, G. P., Diaz, F. G., Harries, J. E., Lewis, R. A., Lowy, J., Mant, G. R., Martin‐ Fernandez, M. L., and Towns‐Andrews, E. (1993). Two‐dimensional time‐ resolved X‐ray diffraction studies of live isometrically contracting frog sartorius muscle. J. Muscle Res. Cell Motil. 14, 311–324. Blow, D. M. (2002). Outline of Crystallography for Biologists. Oxford University Press, Oxford, UK. Brenner, B., Yu, L. C., and Podolsky, R. J. (1984). X‐ray diffraction evidence for crossbridge formation in relaxed muscle fibers at various ionic strengths. Biophys. J. 46, 299–306. Brown, J. H., and Cohen, C. (2005). Regulation of muscle contraction by tropomyosin and troponin: How structure illuminates function. Adv. Protein Chem. 71, 121–159. Cantino, M. E., Chew, M. W., Luther, P. K., Morris, E., and Squire, J. M.. (2002). Structure and nucleotide‐dependent changes of thick filaments in relaxed and rigor plaice fin muscle. J. Struct. Biol. 137, 164–175. Cecchi, G., Bagni, M. A., Colombini, B., Ashley, C. C., Amenitsch, H., Bernstorff, S., and Griffiths, P. J. (2003). Use of sinusoidal length oscillations to detect myosin conformation by time‐resolved X‐ray diffraction. Adv. Exp. Med. Biol. 538, 267–277; discussion 277. Cecchi, G., Griffiths, P. J., Bagni, M. A., Ashley, C. C., and Maeda, Y. (1991). Time‐ resolved changes in equatorial X‐ray diffraction and stiffness during rise of tetanic tension in intact length‐clamped single muscle fibers. Biophys. J. 59, 1273–1283. Chandrasekaran, R., and Stubbs, G. (2001). Fiber diffraction. In ‘‘International Tables Crystallography,’’ Volume F: Crystallography of Biological Macromolecules, pp. 444–450. Kluwer Academic Publishers, Dordrecht, The Netherlands. Chen, L. F., Winkler, H., Reedy, M. K., Reedy, M. C., and Taylor, K. A. (2002). Molecular modeling of averaged rigor cross‐bridges from tomograms of insect flight muscle. J. Struct. Biol. 138, 92–104. Chew, M. W., and Squire, J. M. (1995). Packing of alpha‐helical coiled‐coil myosin rods in vertebrate muscle thick filaments. J. Struct. Biol. 115, 233–249. Craig, R., and Lehman, W. (2002). The ultrastructural basis of actin filament regulation. Results Probl. Cell Differ. 36, 149–169. Dickinson, M., Farman, G., Frye, M., Bekyarova, T., Gore, D., Maughan, D., and Irving, T. (2005). Molecular dynamics of cyclically contracting insect flight muscle in vivo. Nature 433, 330–334. Dobbie, I., Linari, M., Piazzesi, G., Reconditi, M., Koubassova, N., Ferenczi, M. A., Lombardi, V., and Irving, M. (1998). Elastic bending and active tilting of myosin heads during muscle contraction. Nature 396, 383–387. Egelman, E. H., Francis, N., and De Rosier, D. J. (1982). F‐actin is a helix with a random variable twist. Nature 298, 131–135.

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

249

Elliott, G. F., Lowy, J., and Millman, B. M. (1965). X‐ray diffraction from living striated muscle during contraction. Nature 206, 1357–1358. Elliott, G. F., Lowy, J., and Millman, B. M. (1967). Low‐angle X‐ray diffraction from living striated muscle during contraction. J. Mol. Biol. 25, 31–45. Ferenczi, M. A., Bershitsky, S. Y., Koubassova, N., Siththanandan, V., Helsby, W. I., Panine, P., Roessle, M., Narayanan, T., and Tsaturyan, A. K. (2005). The ‘‘roll and lock’’ mechanism of force generation in muscle. Structure 13, 131–141. Geeves, M. A., and Holmes, K. C. (1999). Structural mechanism of muscle contraction. Annu. Rev. Biochem. 68, 687–728. Geeves, M. A., and Holmes, K. C. (2005). Molecular mechanism of muscle contraction. Adv. Protein Chem. 71, 161–193. Granzier, H. L., and Labeit, S. (2005). Titin and its associated proteins: the third myofilament system of the sarcomere. Adv. Protein Chem. 71, 89–119. Griffiths, P. J., Bagni, M. A., Colombini, B., Amenitsch, H., Bernstorff, S., Ashley, C. C., and Cecchi, G. (2002). Changes in myosin S1 orientation and force induced by a temperature increase. Proc. Natl. Acad. Sci. USA 99, 5384–5389. Gu, J., Xu, S., and Yu, L. C. (2002). A model of crossbridge attachment to actin in the A*M*ATP state based on X‐ray diffraction from permeabilized rabbit psoas muscle. Biophys. J. 82, 2123–2133. Harford, J. J., and Squire, J. M. (1986). The ‘‘crystalline’’ myosin cross‐bridge array in relaxed bony fish muscles. Biophys. J. 50, 145–155. Harford, J. J., and Squire, J. M. (1990). Static and time-resolved X‐ray diffraction studies of contracting fish muscle. In ‘‘Molecular Mechanisms in Muscular Contraction’’ (J. M. Squire, Ed.), pp. 287–320. Macmillan Press, Basingstoke & London, UK. Harford, J. J., and Squire, J. M. (1992). Evidence for structurally different attached states of myosin cross‐bridges on actin during contraction of fish muscle. Biophys. J. 63, 387–396. Harford, J. J., and Squire, J. M. (1997). Time‐resolved studies of muscle using synchrotron radiation. Rep. Prog. Phys. 60, 1723–1787. Harford, J. J., Chew, M. W., Squire, J. M., and Towns‐Andrews, E. (1991). Crossbridge states in isometrically contracting fish muscle: Evidence for swinging of myosin heads on actin. Adv. Biophys. 27, 45–61. Harford, J. J., Luther, P. K., and Squire, J. M. (1994). Equatorial A‐band and I‐band X‐ray diffraction from relaxed and active fish muscle: Further details of myosin crossbridge ehaviour. J. Mol. Biol. 239, 500–512. Haselgrove, J. C. (1972). X‐ray evidence for a conformational change in the actin‐ containing filaments of vertebrate striated muscle. Cold Spring Harbor Symp. Quant. Biol. 37, 341–352. Haselgrove, J. C. (1975). X‐ray evidence for conformational changes in the myosin filaments of vertebrate striated muscle. J. Mol. Biol. 92, 113–143. Haselgrove, J. C., and Huxley, H. E. (1973). X‐ray evidence for radial crossbridge movement and for the sliding filament model in actively contracting skeletal muscle. J. Mol. Biol. 77, 549–68. Holmes, K. C. (1997). The swinging lever‐arm hypothesis of muscle contraction. Curr. Biol. 7, R112–R118. Holmes, K. C., and Blow, D. M. (1965). The use of X‐ray diffraction in the study of protein and nucleic acid structure. Methods Biochem. Anal. 13, 113–239. Holmes, K. C., Angert, I., Kull, F. J., Jahn, W., and Schroder, R. R. (2003). Electron cryo‐ microscopy shows how strong binding of myosin to actin releases nucleotide. Nature 425, 423–427.

250

SQUIRE AND KNUPP

Holmes, K. C., Popp, D., Gebhard, W., and Kabsch, W. (1990). Atomic model of the actin filament. Nature 347, 44–49. Holmes, K. C., Schro¨ der, R. R., Sweeney, H. L., and Houdusse, A. (2004). The structure of the rigor complex and its implications for the power stroke. Phil. Trans. R. Soc. B 359, 1819–1828. Hudson, L., Harford, J. J., Denny, R. J., and Squire, J. M. (1997). Myosin head configurations in relaxed fish muscle: Resting state myosin heads swing axially by 150A˚ or turn upside down to reach rigor. J. Mol. Biol. 273, 440–455. Huxley, H. E. (1968). Structural difference between resting and rigor muscle; evidence from intensity changes in the low‐angle equatorial X‐ray diagram. J. Mol. Biol. 14, 507–520. Huxley, H. E. (1969). The mechanism of muscular contraction. Science 164, 1356–1365. Huxley, H. E. (1972). Structural changes in the actin‐ and myosin‐containing filaments during contraction. Cold Spring Harbor Symp. Quant. Biol. 37, 361–376. Huxley, H. E. (2004). Recent X‐ray diffraction studies of muscle contraction and their implications. Phil. Trans Roy. Soc. Lond. B 359, 1879–1882. Huxley, H. E., and Brown, W. (1967). The low‐angle X‐ray diagram of vertebrate striated muscle and its behaviour during contraction and rigor. J. Mol. Biol. 30, 383–434. Huxley, H. E., Brown, W., and Holmes, K. C. (1965). Constancy of spacings in frog sartorius muscle during contraction. Nature 206, 1358. Huxley, H. E., Faruqi, A. R., Bordas, J., Koch, M., and Milch, J. R. (1980). The use of synchrotron radiation in time‐resolved X‐ray diffraction studies of myosin layer line reflections during muscle contraction. Nature 284, 140–143. Huxley, H. E., Reconditi, M., Stewart, A., and Irving, T. (2003). X‐ray interference evidence concerning the range of cross‐bridge movement, and backbone contributions to the meridional pattern. Adv. Exp. Med. Biol. 538, 233–241. Huxley, H. E., Simmons, R. M., Faruqi, A. R., Kress, M., Bordas, J., and Koch, M. H. (1981). Millisecond time‐resolved changes in X‐ray reflections from contracting muscle during rapid mechanical transients, recorded using synchrotron radiation. Proc. Natl. Acad. Sci. USA 78, 2297–2301. Huxley, H. E., Simmons, R. M., Faruqi, A. R., Kress, M., Bordas, J., and Koch, M. H. (1983). Changes in the X‐ray reflections from contracting muscle during rapid mechanical transients and their structural implications. J. Mol. Biol. 169, 469–506. Huxley, H. E., Stewart, A., Sosa, H., and Irving, T. (1994). X‐ray diffraction measurements of the extensibility of actin and myosin filaments in contracting muscle. Biophys. J. 67, 2411–2421. Irving, M., Lombardi, V., Piazzesi, G., and Ferenczi, M. A. (1992). Myosin head movements are synchronous with the elementary force‐generating process in muscle. Nature 357, 156–158. Irving, M., Piazzesi, G., Lucii, L., Sun, Y.‐B., Harford, J. J., Dobbie, I. M., Ferenczi, M. A., Reconditi, M., and Lombardi, V. (2000). Conformation of the myosin motor during force generation in skeletal muscle. Nat. Struct. Biol. 7, 482–485. Irving, M., St. Claire Allen, T., Sabido‐David, C., Craik, J. S., Brandmeier, B., Kendrick‐ Jones, J., Corrie., J., Trentham, D. R., and Goldman, Y. (1995). Tilting of the light chain region of myosin during step length changes and active force generation in skeletal muscle. Nature 375, 688–691. Juanhuix, J., Bordas, J., Campmany, J., Svensson, A., Bassford, M. L., and Narayan, T. (2001). Axial disposition of myosin heads in isometrically‐contracting muscles. Biophys. J. 80, 1429–1441.

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

251

Knupp, C., and Squire, J. M. (2004). HELIX: A helical diffraction simulation program. J. Appl. Cryst. 37, 832–835. Knupp, C., and Squire, J. M. (2005). MusLabel2: A new version of MusLabel including interference effects. (in prep.) Koubassova, N. A., and Tsaturyan, A. K. (2002). Direct modelling of X‐ray diffraction pattern from skeletal muscle in rigor. Biophys. J. 83, 1082–1097. Kraft, T., Xu, S., Brenner, B., and Yu, L. C. (1999). The effect of thin filament activation on the attachment of weak binding cross‐ bridges: A two‐dimensional X‐ray diffraction study on single muscle fibers. Biophys. J. 76, 1494–1513. Kraft, T., Mattei, T., Radocaj, A., Piep, B., Nocula, C., Furch, M., and Brenner, B. (2002). Structural features of crossbridges in isometrically contracting skeletal muscle. Biophys. J. 82, 2536–2547. Kress, M., Huxley, H. E., Faruqi, A. R., and Hendrix, J. (1986). Structural changes during activation of frog muscle studied by time‐resolved X‐ray diffraction. J. Mol. Biol. 188, 325–342. Labeit, S., and Kolmerer, B. (1995). Titins: Giant proteins in charge of muscle structure and elasticity. Science 270, 293–296. Lewis, M. K., Nahirney, P. C., Chen, V., Adhikari, B. B., Wright, J., Reedy, M. K., Bass, A. H., and Wang, K. (2003). Concentric intermediate filament lattice links to specialized Z‐band junctional complexes in sonic muscle fibers of the type I male midshipman fish. J. Struct. Biol. 143, 56–71. Lewis, R. A., Hall, C., Parker, B., Jones, A., Helsby, W., Sheldon, J., Clifford, P., Hillen, M., and Fore, N. (1996). The ‘‘RAPID’’ high rate area X‐ray detector system. Fiber Diffraction Review 5, 30–34. Lewis, R. A., Helsby, W. I., Jones, A. O., Hall, C. J., Parker, B., Sheldon, J., Clifford, P., Hillen, M., Sumner, I., Fore, N. S., Jones, R. W. M., and Roberts, K. (1997). The ‘‘RAPID’’ high rate large area X‐ray detector system. Nucl. Inst. Methods Phys. Res. A 392, 32–41. Linari, M., Piazzesi, G., Dobbie, I., Koubassova, N., Reconditi, M., Narayanan, T., Diat, O., Irving, M., and Lombardi, V. (2000). Interference fine structure and sarcomere length dependence of the axial X‐ray pattern from active single muscle fibers. Proc. Natl. Acad. Sci. USA 97, 7226–7231. Liu, J., Reedy, M. C., Goldman, Y. E., Franzini‐Armstrong, C., Sasaki, H., Tregear, R. T., Lucaveche, C., Winkler, H., Baumann, B. A., Squire, J. M., Irving, T. C., Reedy, M. K., and Taylor, K. A. (2004). Electron tomography of fast frozen, stretched rigor fibers reveals elastic distortions in the myosin crossbridges. J. Struct. Biol. 147, 268–282. Lombardi, V., Piazzesi, G., Ferenczi, M. A., Thirlwell, H., Dobbie, I., and Irving, M. (1995). Elastic distortion of myosin heads and repriming of the working stroke in muscle. Nature 374, 553–555. Lorenz, M., Popp, D., and Holmes, K. C. (1993). Refinement of the F‐actin model against X‐ray fiber diffraction data by the use of a directed mutation algorithm. J. Mol. Biol. 234, 826–836. Lovell, S. J., Knight, P. J., and Harrington, W. F. (1981). Fraction of myosin heads bound to thin filaments in rigor fibrils from insect flight and vertebrate muscles. Nature 293, 664–666. Luther, P. K., and Squire, J. M. (1980). Three‐dimensional structure of the vertebrate muscle A‐band. II. The myosin filament superlattice. J. Mol. Biol. 141, 409–439. Luther, P. K., and Squire, J. M. (2002). Muscle Z‐band ultrastructure: Titin Z‐repeats and Z‐band periodicities do not match. J. Mol. Biol. 319, 1157–1164.

252

SQUIRE AND KNUPP

Luther, P. K., Munro, P. M., and Squire, J. M. (1981). Three‐dimensional structure of the vertebrate muscle A‐band. III. M‐region structure and myosin filament symmetry. J. Mol. Biol. 151, 703–730. Lymn, R. W. (1978). Myosin subfragment‐1 attachment to actin. Expected effect on equatorial reflections. Biophys. J. 21, 93–98. Lymn, R. W., and Taylor, E. W. (1971). Mechanism of adenosine triphosphate hydrolysis by actomyosin. Biochemistry 10, 4617–4624. Martin‐Fernandez, M. L., Bordas, J., Diakun, G., Harries, J., Lowy, J., Mant, G. R., Svennson, A., and Towns‐Andrews, E. (1994). Time‐resolved X‐ray diffraction studies of myosin head movements in live frog sartorius muscle during isometric and isotonic contractions. J. Mus. Res. Cell Motil. 15, 319–348. Millman, B. M., and Irving, T. C. (1988). Filament lattice of frog striated muscle. Radial forces, lattice stability, and filament compression in the A‐band of relaxed and rigor muscle. Biophys. J. 54, 437–47. Mok, N. S., Knupp, C., and Squire, J. M. (2005). Ultrastructural changes during the development of isometric contraction in plaice fin muscle. Moolman‐Smook, J., Flashman, E., De Lange, W., Li, Z., Corfield, V., Redwood, C., and Watkins, H. (2002). Identification of novel interactions between domains of myosin binding protein‐C that are modulated by hypertrophic cardiomyopathy missense mutations. Circ. Res. 91, 704–711. Oshima, K., Takezawa, Y., Sugimoto, Y., Kiyotoshi, M., and Wakabayashi, K. (2003). Modelling analysis of myosin‐based meridional X‐ray reflections from frog skeletal muscles in relaxed and contracting states. Adv. Exp. Med. Biol. 538, 243–249. Padron, R., and Craig, R. (1989). Disorder induced in non‐overlap myosin crossbridges by loss of adenosine triphosphate. Biophys. J. 56, 927–933. Parry, D. A. D., and Squire, J. M. (1973). Structural role of tropomyosin in muscle regulation: Analysis of the X‐ray diffraction patterns from relaxed and contracting muscles. J. Mol. Biol. 75, 33–55. Piazzesi, G., Reconditi, M., Dobbie, I., Linari, M., Boesecke, P., Diat, O., Irving, M., and Lombardi, V. (1999). Changes in conformation of myosin heads during the development of isometric contraction and rapid shortening in single frog muscle fibers. J. Physiol. 514, 305–312. Piazzesi, P., Reconditi, M., Linari, M., Lucii, L., Sun, Y.‐B., Narayanan, T., Boesecke, P., Lombardi, V., and Irving, M. (2002). Mechanism of force generation by myosin heads in skeletal muscle. Nature 415, 659–662. Rayment, I., Holden, H. M., Whittaker, M., Yohn, C. B., Lorenz, M., Holmes, K. C., and Milligan, R. A. (1993b). Structure of the actin‐myosin complex and its implications for muscle contraction. Science 261, 58–65. Rayment, I., Rypniewski, W. R., Schmidt‐Base, K., Smith, R., Tomchick, D. R., Benning, M. M., Winkelmann, D. A., Wesenberg, G., and Holden, H. M. (1993a). Three‐ dimensional structure of myosin subfragment‐1: A molecular motor. Science 261, 50–58. Reconditi, M., Koubassova, N., Linari, M., Dobbie, I., Narayan, T. Diat,O., Piazzesi, G., Lombardi, V., and Irving, M. (2003). The conformation of myosin head domains in rigor muscle determined by X‐ray interference. Biophys. J. 85, 1098–1110. Reconditi, M., Linari, M., Lucii, L., Stewart, A., Sun, Y.‐B., Boesecke, P., Narayanan, T., Fischetti, R. F., Irving, T., Piazzesi, G., Irving, M., and Lombardi, V. (2004). The myosin motor in muscle generates a smaller and slower working stroke at higher load. Nature 428, 578–581.

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

253

Reedy, M. K. (1968). Ultrastructure of insect flight muscle. I. Screw sense and structural grouping in the rigor crossbridge lattice. J. Mol. Biol. 31, 155–176. Rome, E., Hirabayashi, T., and Perry, S. V. (1973b). X‐ray diffraction of muscle labelled with antibody to troponin‐C. Nature New Biol. 244, 154–155. Rome, E., Offer, G., and Pepe, F. A. (1973a). X‐ray diffraction of muscle labelled with antibody to C‐ protein. Nature New Biol. 244, 152–154. Sjostrom, M., and Squire, J. M. (1977). Fine structure of the A‐band in cryo‐sections. The structure of the A‐band of human skeletal muscle fibers from ultra‐thin cryo‐ sections negatively stained. J. Mol. Biol. 109, 49–68. Squire, J. M. (1972). General model of myosin filament structure. II. Myosin filaments and cross‐ bridge interactions in vertebrate striated and insect flight muscles. J. Mol. Biol. 72, 125–38. Squire, J. M. (1973). General model of myosin filament structure. III. Molecular packing arrangements in myosin filaments. J. Mol. Biol. 77, 291–323. Squire, J. M. (1981). ‘‘The Structural Basis of Muscular Contraction.’’ Plenum Press, New York. Squire, J. M. (1986). Muscle myosin filaments: Internal structure and crossbridge organization. Comments Molec. Cel. Biophysics 3, 155–177. Squire, J. M. (1992). The structures of striated and smooth muscles related to their function. In ‘‘Muscle Contraction and Cell Motility: Molecular and Cellular Aspects’’ (H. Sugi, Ed.), pp. 87–131. Springer‐Verlag, Berlin and Heidelberg. Squire, J. M. (1997). Architecture and function in the muscle sarcomere. Curr. Op. Struct. Biol 7, 247–257. Squire, J. M. (1998). Time‐resolved X‐ray Diffraction. In ‘‘Current Methods in Muscle Physiology: Advantages, Problems and Limitations’’ IUPS Commission on Muscle Physiology (H. Sugi, Ed.), pp. 241–285. Oxford University Press. Squire, J. M. (2000). Fiber and muscle diffraction. In ‘‘Structure and dynamics of biomolecules’’ (E. Fanchon, E. Geissler, L.‐L. Hodeau, J.‐R. Regnard, and P. Timmins, Eds.). Oxford University Press, Oxford. Squire, J. M., Al‐Khayat, H. A., Knupp, C., and Luther, P. K. (2005). Molecular architecture in muscle contractile assemblies. Adv. Protein Chem. 71, 17–87. Squire, J. M., and Harford, J. J. (1988). Actin filament organisation and myosin head labelling patterns in vertebrate skeletal muscles in the rigor and weak-binding states. J. Mus. Res. Cell Motil. 9, 344–358. Squire, J. M., and Knupp, C. (2004). MusLABEL: A program to model striated muscle A‐band lattices, to explore cross‐bridge interaction geometries and to simulate muscle diffraction patterns. J. Muscle Res. Cell Motil. 25, 423–438. Squire, J. M., and Morris, E. P. (1998). A new look at thin filament regulation in vertebrate skeletal muscle. FASEB J. 12, 761–771. Squire, J. M., Al‐Khayat, H. A., Harford, J. J., Hudson, L., Irving, T. C., Knupp, C., Mok, N. S., and Reedy, M. K. (2003a). Myosin filament structure and myosin cross‐bridge dynamics in fish and insect muscles. In ‘‘Molecular and Cellular Aspects of Muscle Contraction’’ (H. Sugi, Ed.). Kluwer: Adv. Exp. Med. Biol. 538, pp. 251–266. Squire, J. M., Al‐Khayat, H. A., Harford, J. J., Hudson, L., Irving, T., Knupp, C., and Reedy, M. K. (2003b). Modelling muscle motor conformations using low‐angle X‐ray diffraction. IEE Proc. BioNanotechnol. 150, 103–110. Squire, J. M., Cantino, M., Chew, M., Denny, R., Harford, J. J., Hudson, L., and Luther, P. K. (1998). Myosin rod packing schemes in vertebrate muscle thick filaments. J. Struct. Biol. 122, 128–138.

254

SQUIRE AND KNUPP

Squire, J. M., Harford, J. J., Edman, A. C., and Sjostrom, M. (1982). Fine structure of the A‐band in cryo‐sections. III. Crossbridge distribution and the axial structure of the human C‐zone. J. Mol. Biol. 155, 467–494. Squire, J. M., Harford, J. J., and Al‐Khayat, H. A. (1994). Synchrotron radiation in time‐ resolved X‐ray diffraction studies of molecular movements in muscle. In ‘‘Synchrotron Radiation in the Biosciences.’’ Oxford University Press, Oxford, UK. Squire, J. M., Harford, J. J., Al‐Khayat, H. A., and Knupp, C. (2005b). Actin target area labelling in rigor vertebrate muscle. (In prep.) Squire, J. M., Knupp, C., Al‐Khayat, H. A., and Harford, J. J. (2003c). Millisecond time‐ resolved low‐angle X‐ray diffraction: A powerful, high‐sensitivity technique for modelling real‐time movements in biological macromolecular assemblies. Fiber Diffraction Review 11, 28–35. Squire, J. M., Knupp, C., Roessle, M., Al‐Khayat, H. A., Irving, T. C., Eakins, F., Mok, N. S., Harford, J. J., and Reedy, M. K. (2005a). X‐ray diffraction studies of striated muscles. In ‘‘Mysteries About the Sliding Filament Mechanism: Fifty Years After Its Proposal’’ (H. Sugi, Ed.). Kluwer: Adv. Exp. Med. Biol. In press. Squire, J. M., Luther, P. K., and Knupp, C. (2003d). Structural evidence for the interaction of C‐ protein (MyBP‐C) with actin and sequence identification of a possible actin‐binding domain. J. Mol. Biol. 331, 713–724. Squire, J. M., Roessle, M., and Knupp, C. (2004). New X‐ray diffraction observations on vertebrate muscle: Organisation of C‐protein (MyBP‐C) and troponin and evidence for unknown structures in the vertebrate A‐band. J. Mol. Biol. 343, 1345–1363. Steffen, W., Smith, D., Simmons, R., and Sleep, J. (2001). Mapping the actin filament with myosin. Proc. Natl. Acad. Sci. USA 98, 14949–14954. Tregear, R. T., Edwards, R. J., Irving, T. C., Poole, K. J., Reedy, M. C., Schmitz, H., Towns‐ Andrews, E., and Reedy, M. K. (1998). X‐ray diffraction indicates that active crossbridges bind to actin target zones in insect flight muscle. Biophys. J. 74, 1439–1451. Tsaturyan, A. K., Bershitsky, S. Y., Burns, R., and Ferenczi, M. A. (1999a). Structural changes in the actin‐myosin crossbridges associated with force generation induced by temperature jump in permeabilized frog muscle fibers. Biophys. J. 77, 354–372. Tsaturyan, A. K., Bershitsky, S. Y., Burns, R., He, Z.‐H., and Ferenczi, M. A. (1999b). Structural responses to the photolytic release of ATP in frog muscle fibers observed by time‐resolved X‐ray diffraction. J. Physiol. 520, 681–696. Vibert, P. J., Craig, R., and Lehman, W. (1997). Steric‐model for activation of muscle thin filaments. J. Molec. Biol. 266, 8–14. Wakabayashi, K., Sugimoto, Y., Tanaka, H., Ueno, Y., Takezawa, Y., and Amemiya, Y. (1994). X‐ray diffraction evidence for the extensibility of actin and myosin filaments during muscle contraction. Biophys. J. 67, 2422–2435. Wong, W. W., Doyle, T. C., and Reisler, E. (1999). Nonspecific weak actomyosin interactions: Relocation of charged residues in subdomain 1 of actin does not alter actomyosin function. Biochemistry 38, 1365–1370. Xu, S., Gu, J., Melvin, G., and Yu, L. C. (2002). Structural characterization of weakly attached crossbridges in the A*M*ATP state in permeabilized rabbit psoas muscle. Biophys. J. 82, 2111–2122. Xu, S., Gu, J., Rhodes, T., Belknap, B., Rosenbaum, G., Offer, G., White, H., and Yu, L. C. (1999). The M.ADP.P(i) state is required for helical order in the thick filaments of skeletal muscle. Biophys. J. 77, 2665–2676. Yagi, N. (1996). Labelling of thin filaments by myosin heads in contracting and rigor vertebrate skeletal muscles. Acta Cryst. D. 52, 1169–1173.

STUDIES OF MUSCLE AND THE CROSSBRIDGE CYCLE

255

Yagi, N. (1992). Effects of N‐ethylmaleimide on the structure of skinned frog skeletal muscles. J. Muscle Res. Cell Motil. 13, 457–463. Yagi, N. (2003). An X‐ray diffraction study on early structural changes in skeletal muscle contraction. Biophys. J. 84, 1093–1102. Yagi, N., Horiuti, K., and Takemori, S. (1998). A pre‐active attached state of myosin heads in rat skeletal muscles. J. Muscle Res. Cell Motil. 19, 75–86. Yu, L. C., and Brenner, B. (1989). Structures of actomyosin crossbridges in relaxed and rigor muscle fibers. Biophys. J. 55, 441–453. Yu, L. C., Lymn, R. W., and Podolsky, R. J. (1977). Characterization of a non‐indexible equatorial X‐ray reflection from frog sartorius muscle. J. Mol. Biol. 115, 455–464.

MICROTUBULES AND MAPS By LINDA A. AMOS AND DANIEL SCHLIEPER MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 2QH, United Kingdom

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Microtubule Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Subunit Lattices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Atomic Structure of the Protofilament . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Atomic Structure of a Microtubule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. GTP‐Binding Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Effect of GTP Hydrolysis on Tubulin Structure. . . . . . . . . . . . . . . . . . . . . . . F. The Curved Protofilament Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Dynamic Instability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Dynamic Behavior and GTP Hydrolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Effects of Assembly‐Inhibiting Drugs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Mechanisms of Stabilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Structural MAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. The Tau Family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. MAP1A and MAP1B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. STOPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Doublecortin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Microtubule Destabilizers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Stathmin (Op18). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Katanin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. EMAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Kinesins as Regulators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. MINUS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Proteins That Control Microtubule Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. g‐Tubulin Ring Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. XMAP215, ch‐TOG, Msps Family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. CLIPs/þTIPs and CLASPs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. The Dynactin Complex. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. EB1 and APC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Hook Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. Centrosomal Coiled‐Coil Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

258 258 259 259 262 265 267 268 268 269 270 271 272 273 275 275 275 279 279 279 280 280 282 282 282 283 285 286 288 288 289 290 291

Abstract Microtubules are very dynamic polymers whose assembly and disassembly is determined by whether their heterodimeric tubulin subunits are in a straight or curved conformation. Curvature is introduced by bending at ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71007-4

257

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

258

AMOS AND SCHLIEPER

the interfaces between monomers. Assembly and disassembly are primarily controlled by the hydrolysis of guanosine triphosphate (GTP) in a site that is completed by the association of two heterodimers. However, a multitude of associated proteins are able to fine‐tune these dynamics so that microtubules are assembled and disassembled where and when they are required by the cell. We review the recent progress that has been made in obtaining a glimpse of the structural interactions involved.

I. Introduction When microtubules were visualized by electron microscopy (EM), after the improvement of methods of fixation, it was realized that they formed the structural basis of flagellar axonemes and of so‐called spindle fibers, as well as occurring as individual filaments in the cytoplasm. Their designation as part of the ‘‘cytoskeleton’’ suggested that they acted mainly as fixed structural supports. Subsequent research has focused more and more on their dynamic behavior and on their role as tracks for motor proteins, which may, for example, transport chromosomes during cell division. Microtubules are found in all eukaryotic cells and are essential for many cellular functions, such as motility, morphogenesis, intracellular transport, and cell division. It is that dynamic behavior that allows microtubules to fulfill all of these functions in specific places and at appropriate times in the cell cycle.

II.

Microtubule Structure

Unpolymerized tubulin exists as a tight ab‐tubulin heterodimer with binding sites for two molecules of GTP, one exchangeable and the other not. A microtubule is a tube constructed from parallel linear polymers (protofilaments) in which the heterodimers are assembled head‐to‐tail in a polar fashion. This polarity is reflected by the distinction between the so‐called plus and minus ends of protofilaments. It can readily be seen by EM that each protofilament consists of globular, 4‐nm subunits (Figs. 1A and 2); however, a‐ and b‐tubulin are so similar in structure that the two kinds of monomer subunits cannot be distinguished except at very high resolution. In vivo, microtubules usually have 13 protofilaments, although the number may differ in particular locations. In vitro, it is possible for purified tubulin to assemble with a range of diameters and to contain between 9 and 16 protofilaments (Figs. 1A and 2). This variation reveals a degree of flexibility in the bonds between adjacent heterodimers, at least in the direction running around the microtubule.

MICROTUBULES AND MAPS

259

When the number of protofilaments is 13, they run straight, allowing microtubule‐associated motor proteins, such as dynein and kinesin (Hirokawa, 1998; Vale, 2003), to run for long distances along a microtubule without switching lanes. The transport of vesicles and mitochondria along axons, for example, would be difficult if they needed to rotate around microtubules. If microtubules have more or fewer than 13 protofilaments, the lattice must rotate slightly, so that the protofilaments wind slowly around the microtubule axis (Chre´ tien and Wade, 1991; Wade et al., 1990). The amount and direction of rotation is such as to allow the two‐ dimensional (2D) lattice of subunits to close up neatly into tubes with larger or smaller than the standard diameter. In vivo, the occurrence of microtubules with protofilament numbers other than 13 appears to be determined by isoforms with specific changes in the amino acid sequences of a‐ and b‐tubulin, as for example in the case of some specialized 15‐ protofilament microtubules in neurons of the nematode Caenorhabditis elegans (Fukushige et al., 1999; Savage et al., 1989). These changes presumably allow the formation of a slightly deformed lattice in which a larger number of protofilaments can still be straight.

A.

Subunit Lattices

Monomers in adjacent protofilaments are slightly staggered (Fig. 2) so that they form a set of shallow helices. In a 13‐protofilament microtubule, three shallow helices run in parallel, and each makes a complete turn over an axial distance of 12 nm. For a smaller or larger number of protofilaments, there may be two or four shallow helices. It is possible to deduce the number of protofilaments and hence the number of shallow helices in a particular microtubule in an EM image by looking at the Moire´ pattern due to the superposition of the front and back of the tube (Fig. 1A). However, the most reliable way of determining the subunit lattice of tubulin monomers is to calculate a Fourier transform (diffraction pattern) of the image (Fig. 1B) and to analyze the positions of the diffraction spots. The data from the Fourier transform can also be used to reconstruct a 3D image of a helically symmetrical specimen.

B.

Atomic Structure of the Protofilament

The atomic structure of tubulin protofilaments (Figs. 3 and 4) is known from 3.5 A˚ resolution maps, the first of which came from electron crystallography of zinc‐induced 2D sheets (Nogales et al., 1998b), subsequently refined (Lo¨ we et al., 2001). Recent data, including important information

260

AMOS AND SCHLIEPER

Fig. 1. (A) Electron microscopy of microtubules assembled in vitro. The microtubules shown were assembled from pure pig brain tubulin and rapidly frozen in a very thin layer of ice. Since frozen‐hydrated specimens are unstained, all the contrast comes from the difference between protein and ice. Microtubules with different numbers of

MICROTUBULES AND MAPS

261

Fig. 2. Microtubules with varying numbers (12–16) of longitudinal protofilaments (pf). When 13 protofilaments make up the cylinder, they run straight, but larger or smaller numbers must wind slowly around the axis if the monomer subunits are to line up correctly at the seam. The tilting of the subunit lattice is most obvious in the 15‐ or 16‐ pf structure. In this lattice, four separate helical lines run through laterally adjacent monomers. For 13‐ or 14‐pf tubes, only three shallow helices run in parallel, forming the standard ‘‘3‐start’’ set. Narrower tubes have two shallow helices. In this drawing, monomer subunits are represented as darker and lighter spheres to distinguish between a‐ and b‐tubulin. The lattice of subunits, known as the B‐lattice, is one in which all or most lateral interactions are a‐a or b‐b. Perfect helical symmetry (with all lateral interactions alike) is possible for B‐lattice microtubules with some pf numbers, such as 12, 15, or 16, but a standard 13‐pf microtubule can only close with a seam in which each a‐tubulin monomer makes lateral contact with a b‐tubulin subunit. The so‐called A‐lattice is one in which all lateral contacts would be like this (Amos and Klug, 1974). Most microtubules assembled in vitro from pure tubulin have a B‐lattice. Because of their instability during isolation, the lattices of truly native microtubules are difficult to investigate.

protofilaments (pf) vary in diameter. Also, because their protofilaments run at slightly different angles to the microtubule axis, the Moire´ patterns created by superposition of the front and back layers of the tubes have different appearances. Protofilaments in a 13‐pf segment of a microtubule run straight (white arrow). White arrowheads indicate repeats in the Moire´ pattern of a 15‐pf microtubule. (Image provided by J. Fan). (B) Diffraction patterns were obtained by calculating the 2D Fourier transforms of individual microtubule images and displaying their amplitudes. For well‐ordered specimens, the patterns show a ‘‘layer‐line’’ of peaks at a reciprocal height above the origin of (4 nm)–1, arising from the 4‐nm longitudinal spacing of the tubulin monomers. Along the equatorial line, there are peaks arising from the ~5‐nm lateral separation of the protofilaments. The precise pattern of peaks provides information about how much the tubulin lattice is rotated and, hence, about the number of protofilaments.

262

AMOS AND SCHLIEPER

about conformational changes, have been obtained by X‐ray crystallography (Ravelli et al., 2004). Each monomer has a pair of globular domains set on either side of a central (core) helix H7. The larger globular domain, composing the N‐terminal half of the polypeptide, has the same fold as a family of dinucleotide binding proteins with the so‐called Rossmann fold (Nogales et al., 1998a). There is a binding site for the guanosine nucleotide on the plus end surface of this domain, where contact is made with the next subunit in the protofilament. The position of the nucleotide at the center of the polymerization interface prevents its exchange as long as the subunits remain assembled. The second globular domain of b‐tubulin has a binding site for Taxol, which also makes contact with the core helix, on the opposite side from its contact with the nucleotide base. The C‐ terminal end of each tubulin polypeptide forms two long helices (H11 and H12) connected by a U‐turn, whereas the final 13 residues of a‐tubulin and 9 residues of b‐tubulin are also disordered in the 2D crystals to show up as electron density, but they are assumed to project out into the solution. These C‐terminal residues would be suitable for isoform recognition by tubulin binding proteins, for example, katanin. They are acidic and negatively charged in physiological conditions.

C. Atomic Structure of a Microtubule Several long coils loop out from the globular domains, and some of these are involved in lateral contacts between the protofilaments in a microtubule. A number of groups have docked protofilaments from the zinc‐sheet structure solved by electron crystallography to near‐atomic resolution (Lo¨ we et al., 2001) into lower‐resolution helical microtubule maps imaged by electron cryomicroscopy (Amos, 2000; Li et al., 2002; Meurer‐Grob et al., 2001; Nogales et al., 1999). There is general agreement that the ‘‘M‐loops’’ of one protofilament make contact with the GTPase domains of the next one, in the region between helix H3 and the b‐sheet (Fig. 3). Although the zinc‐sheet structure fits into the low‐resolution envelopes, there are structural differences between tubulin in zinc‐sheets and in microtubules. It was found that protofilaments in zinc‐induced sheets are rotated around their axes by about 20 degrees compared with protofilaments in a normal sheet or opened‐out microtubule. As a result, the M‐loops make different contacts with the adjacent protofilament, covering part of the surface that corresponds to the outside of a microtubule. Kinesin is, therefore, unable to bind to tubulin in these sheets. The 0.8‐nm resolution microtubule map of Li et al. (2002) shows that the M‐ loop is also in a slightly different conformation in a microtubule compared with that in a zinc‐induced sheet.

MICROTUBULES AND MAPS

263

Fig. 3. Ribbon diagram of an ab‐tubulin heterodimer. The structure solved by electron crystallography using sheets of bovine brain tubulin in complex with Taxol (Lo¨ we et al., 2001; Nogales et al., 1998b) is shown in an orientation corresponding to the inside view of a microtubule. The GTPase domains are colored pink and the activation domains blue. The core helix that connects the two globular domains in each monomer is colored yellow, and the C‐terminal domain on the external surface is green. GTP is sandwiched between the a‐ and b‐tubulin subunits of each heterodimer, being bound to a‐tubulin by loops T1–T6, and also makes contact with loop T7 of b‐tubulin. The nucleotide bound to b‐tubulin has been hydrolyzed to GDP through contact with helix H8 and loop T7 of the activation domain of another a‐tubulin subunit. Taxol sits in the pocket of b‐tubulin on the inside face of microtubules. In a‐tubulin, this pocket is occupied by the extended L‐loop (prepared using Molscript; Kraulis, 1991).

264

AMOS AND SCHLIEPER

Fig. 4. Atomic models of tubulin in straight and curved conformations. Ribbon diagrams similar to that in Figure 3, with the same color scheme. (A and B) Outside and side views of the straight heterodimer (Lo¨ we et al., 2001). (C) Crystal structure of the

MICROTUBULES AND MAPS

265

When building these docking models, it was already known that the plus end of each protofilament ends with a b‐tubulin subunit (Fan et al., 1996; Mitchison, 1993). The orientation of low‐resolution microtubule maps was also known, from experiments in which sheets growing from the ends of axoneme fragments had been decorated with kinesin motor domains (Hirose et al., 1995a,b). The latter bind stoichiometrically, one motor domain per tubulin heterodimer, and the shape of this complex is sufficiently asymmetric to define which of the two tubulin subunits is a and which is b (Fig. 5A). The orientation of tubulin heterodimers was confirmed by the docking exercises, which also confirmed that the surface consisting of a longitudinal ridge lies on the outside of a microtubule, whereas the surface exhibiting lateral ridges and containing the Taxol binding site is on the inside (Nogales et al., 1999). The longitudinal ridges seen in 3D EM images correspond to the green helices shown in Figs. 3 and 4.

D.

GTP‐Binding Sites

GTP is in direct contact with loops T1 to T6 of the GTPase domain. During assembly into protofilaments, loop T7 and helix H8, in the activation domain of the next subunit, are brought close to the phosphates of the nucleotide and promote hydrolysis to guanosine diphosphate (GDP) (Nogales et al., 1998a). The idea that the smaller globular domain of tubulin or of FtsZ, the bacterial monomeric homologue of tubulin, acts as a GTP hydrolysis‐activating protein (GAP; Erickson, 1998) to the GTPase domain has been confirmed experimentally by mutagenesis of FtsZ (Mukherjee et al., 2001; Scheffers et al., 2002). The protofilaments of both tubulin and FtsZ thus consist of alternating GTPase domains and activation domains (Figs. 3 and 4; Oliva et al., 2004). After assembly and hydrolysis, the position of the nucleotide at the center of the interface prevents its exchange from b‐tubulin until the subunit disassembles. GTP bound to a‐tubulin is nonexchangeable, being permanently trapped tubulin‐stathmin complex (Ravelli et al., 2004). Stathmin, shown in gray, induces curvature of two tubulin heterodimers. Its N‐terminal domain caps one end binding to an a‐tubulin subunit. The depolymerizing drugs colchicine (CH) and podophyllotoxin (POD) bind to similar sites on b‐tubulin. Pironetin binds to a lysine (K352) on a‐tubulin (Usui et al., 2004). Vinblastine (VB), with which pironetin competes, binds to the GTPase domain of b‐tubulin, to loop T5, or to the loop between H6 and H7 (residues 177–215; Rai and Wolff, 1996). The curvature of the stathmin helix indicates the degree of the curvature of this complex (12 degrees/monomer). Figure prepared using Molscript (Kraulis, 1991).

266

AMOS AND SCHLIEPER

Fig. 5. 3D EM shows how kinesin and tau bind to microtubules. (A) Reconstruction of a microtubule decorated with kinesin heads (ochre). One kinesin head binds per ab‐ tubulin heterodimer (grey) and, due to its asymmetric form, can be used to distinguish between the subunits. (B) Inside view of a microtubule that was coassembled with gold‐ labeled tau and decorated with kinesin heads. The kinesin heads can be seen on the outside through the holes between the protofilaments. The labeled repeat motif of tau binds to the inside face of microtubule. The averaged nanogold density (yellow), which is attached to a repeat motif of tau through a linker, can only be seen near the Taxol binding site of b‐tubulin, but not on the a subunit (Kar et al., 2003a). The ribbon diagram of the refined zinc‐sheet structure is also shown for reference (see Figure 3).

between the two monomers of the heterodimer, and is never hydrolyzed. b‐tubulin has lysine at the lower end of helix H8, in place of glutamic acid (as in a‐tubulin E254) or aspartic acid (as in FtsZ), which would be needed for hydrolysis (Nogales et al., 1998a). Because GTP is never hydrolyzed in a‐ tubulin, this interface remains tightly bound when the heterodimer becomes soluble. As first noted by Nogales et al. (1998b, 1999), the other

MICROTUBULES AND MAPS

267

significant difference between a‐ and b‐tubulin is found in the loop L in the activation domain. a‐tubulin has a longer L‐loop occupying a pocket that provides a binding site in b‐tubulin for Taxol and other drugs. This pocket is located on the inner surface of microtubules, but, in a‐tubulin, it is closed by the conserved sequence TVVPGGDL of the extended L‐loop. Therefore, each ab‐tubulin heterodimer has one nonexchangeable GTP site and one open Taxol‐binding pocket. Loops T1 through T7 surrounding the nucleotide are regions of highest sequence homology between tubulin and FtsZ, the bacterial homolog of tubulin (Lo¨ we and Amos, 1998; Nogales et al., 1998a). The high affinity of this site for nucleotide was demonstrated by a remarkable experiment with FtsZ (Andreu et al., 2002). Unlike tubulin, which is very unstable and loses all activity, the FtsZ of Methanococcus jannaschii could be refolded after being unfolded with guanidinium hydrochloride. Nucleotide was released during unfolding of the protein, but as much as 80% rebound when the denaturant was diluted 50‐fold in fresh buffer and the FtsZ refolded.

E. Effect of GTP Hydrolysis on Tubulin Structure When a microtubule disassembles, after GTP hydrolysis, its protofilaments roll up to form rings or pieces of a ring (Nogales et al., 2003; Watts et al., 2002). This is because GTP hydrolysis promotes bends in protofilaments. However, whereas GDP‐bound protofilaments are still associated together as a microtubule or 2D sheet, the contacts between neighboring subunits constrain them to remain in a straight form. The resulting tension is proposed to store conformational energy that is released during depolymerization. Thus, the structure solved by Nogales and colleagues corresponds to the ‘‘strained’’ straight state. In microtubules, nucleotide hydrolysis produces a small conformational change that shows up as a 2–4% reduction in the length of the tubulin dimer. This was discovered by comparing microtubules assembled with GTP with those assembled with guanylyl‐(a,b)‐methylene‐diphosphonate (GMPCPP; Hyman et al., 1995). While the microtubules hydrolyze GTP quickly and thus have GDP bound to most of their b‐subunits, they hydrolyze GMPCPP very slowly and the microtubules can be seen in a GTP‐like state. The difference in longitudinal spacing can be measured in diffraction patterns (Fig. 1B). Microtubules assembled with GMPCPP are relatively stable. The nucleotide‐dependent change in spacing is indicative of a conformational change that puts the protofilaments into the straight, but strained, condition responsible for the dynamic instability of microtubules. The change in subunit length presumably involves movement of some of the loops that are involved in longitudinal bonds. Also, the core helix H7 (connected at one end to loop

268

AMOS AND SCHLIEPER

T7, and at the other end to loop N, which is involved in both longitudinal and lateral inter‐subunit contacts) may also shift, tilt, or even shorten. Such a change might explain how hydrolysis of the nucleotide bound to b‐tubulin may lead to changes in a‐tubulin as well.

F. The Curved Protofilament Structure A surprising observation is that when a protofilament forms a ring, it appears to bend at all the interfaces between monomers, both between and within heterodimers. The bending at all interfaces was first seen by EM (Nogales et al., 2003; Watts et al., 2002) and has been confirmed in co‐crystals of tubulin and the tubulin‐sequestering protein stathmin (Gigant et al., 2000; Ravelli et al., 2004). It is unexpected because there is still GTP bound in the intra‐dimer interface, so there is no nucleotide hydrolysis driving a conformational change here. The bending occurs even in the absence of destabilizing agents such as stathmin or drugs such as colchicine. The core helix is a likely means of communication from the top to the bottom of the b‐subunit. This cooperative mechanism is presumably advantageous during the rapid disassembly phase of microtubule dynamics. It remains to be seen how communication along the protofilament axis is actually achieved. As well as the observed variation in the number of protofilaments in microtubules, and hence variations in their curvature, whole microtubules can bend and twist without snapping or coming under great elastic strain. This is apparent from images of fluorescently labeled microtubules growing and shrinking in living cells (e.g., Rodionov et al., 1999). When a microtubule bends, individual protofilaments are bent in a variety of directions. It is likely, therefore, that there are multiple ‘‘bent’’ states for dimers and protofilaments and there is no certainty that the curved conformations induced by different agents of disassembly are identical.

III.

Dynamic Instability

Microtubules, especially those that make up the mitotic spindle, are in a delicate state of balance between assembly and disassembly. This is because both the formation of the spindle and the movement of chromosomes to opposite spindle poles depend on carefully coordinated extension or shrinkage at both ends of the microtubules in the spindle. The end of a microtubule that terminates with b‐tubulin is more dynamic than the other end, which has an a‐tubulin monomer as its final subunit. In cells, microtubules usually grow out from some sort of organizing

MICROTUBULES AND MAPS

269

center from which the more dynamic end, known as the plus end, is able to grow and shrink, whereas the minus end may not be able to change. If both ends are free, as in vitro, assembly and disassembly can occur from either end, although at different rates. Microtubules can continue to grow as long as the free tubulin concentration is above a critical level. The critical concentration at a minus end is somewhat higher than at a plus end and thus minus ends tend to stop growing first.

A. Dynamic Behavior and GTP Hydrolysis Even when the tubulin concentration is above the critical level, it is observed that any microtubule end may suddenly stop growing and begin to shrink rapidly. The switch is a stochastic process; individual plus ends shrink rapidly while others are still growing (see Desai and Mitchison, 1997). The change from growth to shrinkage has been termed a ‘‘catastrophe.’’ After a while, a microtubule end shrinking in vivo may ‘‘pause’’ in a state intermediate between the growing and shrinking states, although this does not occur in the case of microtubules assembled in vitro from pure tubulin. In either situation, a microtubule that has shrunk may begin to grow again; the latter process is known as a ‘‘rescue.’’ Microtubules tend to disassemble when cells are cooled below their normal temperature and to reassemble when they are rewarmed, but they show dynamic instability even under constant warm conditions. The hydrolysis rate of GTP by unpolymerized tubulin dimer is very low (0.054 min–1 at most; David‐Pfeuty et al., 1979; Erickson, 1995), but it dramatically increases during the polymerization into microtubules (21 min–1; Melki et al., 1996). The GTP bound to a‐tubulin is nonexchangeable, being trapped between the two monomers of the heterodimer (the so‐called N‐site), and is apparently never hydrolyzed. However, when heterodimers associate to form protofilaments, GTP on b‐tubulin is hydrolyzed to GDP as a consequence of the interaction with a‐tubulin in the next dimer. It is not clear how soon this exchangeable GTP (on the so‐called E‐site) is hydrolyzed after the addition of a new dimer to the plus end, but it may be very quick. The ‘‘cap’’ of dimers containing GTP on a microtubule end may be as little as one subunit deep (Walker et al., 1991). Unpolymerized tubulin dimer with GDP bound has a curved conformation. Double rings of 12–16 GDP‐bound dimers have an average bend per monomer of 11 to 15 degrees. However, the GDP protofilament structure that exists throughout most of the microtubule is constrained to form straight protofilaments by contacts between neighboring subunits in the lattice. This has been proposed to store conformational energy that is released during depolymerization.

270

AMOS AND SCHLIEPER

Assembly of microtubules appears to take place mainly through the addition of individual heterodimers to the ends of protofilaments. During rapid assembly at the plus end, there is usually a group of protofilaments in part of the microtubule wall that takes the lead (Chre´ tien et al., 1995) and may extend as a narrow sheet for some distance beyond the end of the closed tube. Such sheets have not been seen at minus ends, where growth is more even. Disassembly at either end appears to be a cooperative process, since the ends of disassembling microtubules have been seen splaying apart and bending outward into a curved conformation. In some circumstances, long segments of protofilaments are shed as spirals, and 30‐ to 40‐nm diameter rings can form. It has been postulated that catastrophes, pauses, and rescues at the plus end are caused by random loss or restoration of the GTP tubulin cap. This would happen by hydrolysis of the nucleotide or attachment of fresh GTP‐bound tubulin heterodimers, respectively. Thus, the stochastic events responsible for changes in behavior may be spontaneous conformational changes that are propagated along either individual protofilaments or small groups of protofilaments. The bending of protofilaments during disassembly is a manifestation of a cooperative change.

B. Effects of Assembly‐Inhibiting Drugs The role of microtubules in the process of segregating duplicated chromosomes before cell division makes them an important target for antimitotic drugs. A variety of drugs are known to inhibit microtubule assembly and thereby stall cells in mitosis, when microtubules are most dynamic and least stable. It seems likely that most of these compounds, if not all of them, favor the curved conformation of tubulin. The crystal structure of tubulin in complex with stathmin and colchicine (Gigant et al., 2000; Ravelli et al., 2004) confirmed earlier indications that colchicine binds to b‐tubulin near to the interface between monomers (Fig. 4). Podophyllotoxin was seen to bind to a similar site. This binding location necessarily requires a distortion within the dimer structure that would inhibit its polymerization into straight protofilaments. In contrast, vinblastine and pironetin bind to regions that form contacts between dimers. For vinblastine, which can turn protofilaments into fairly tightly wound helices, cross‐linking experiments identified a binding region somewhere on residues 175–213 of b‐tubulin (Rai and Wolff, 1996). This peptide includes regions that are involved in longitudinal polymerization contacts between dimers (Fig. 4). A surprising feature of the recently discovered compound pironetin is that its binding to a‐tubulin inhibits the binding of vinblastine to b‐tubulin

MICROTUBULES AND MAPS

271

(Kondoh et al., 1999; Usui et al., 2004). Such competition suggests that the two drugs occupy overlapping sites in the interface between tubulin heterodimers, where they must each bind without actually disassembling the protofilaments. In contrast, colchicine binds more readily to the intra‐ dimer site if either vinblastine or pironetin is already bound to the inter‐dimer interface. The binding of either drug apparently forces protofilaments to be curved, with the intra‐dimer interfaces partially opened up. Another microtubule disassembler, cryptophycin‐1, induces the formation of 24‐nm diameter rings, each containing eight tubulin dimers (Watts et al., 2002). Digitally processed EM images of these rings showed 13‐degree bends at intra‐dimer contacts and 32‐degree bends at inter‐dimer contacts. Although the drug binds only to the b‐subunit, it protects both a‐ and b‐tubulin against proteolysis by trypsin, indicating that conformational changes were induced in specific regions of both subunits.

C.

Mechanisms of Stabilization

Since the structure of Taxol has been solved to high resolution repeatedly, by X‐ray crystallography and nuclear magnetic resonance (NMR), Snyder et al. (2001) were able to correlate the electron density in the Taxol‐ binding pocket in the map of tubulin sheets with all the known Taxol conformations and thus identify the one most likely to be present. This is a T‐shaped or butterfly‐like structure, opened up to expose a hydrophobic surface that interacts with a hydrophobic patch on the surface of b‐tubulin. A recent study of tubulin sheets complexed with epothilone B, a compound that competes with Taxol and has similar microtubule‐stabilizing effects (Bollag et al., 1995; Giannakakou et al., 2000), found it to bind in the same pocket as Taxol, but it interacts with the protein via quite different groups of atoms in the polypeptide (Nettles et al., 2004). Other stabilizing drugs such as discodermolide (ter Haar et al., 1996), eleutherobin (Hamel et al., 1999; Long et al., 1998), and the sarcodictyins (Hamel et al., 1999) are structurally diverse, yet all compete with Taxol for binding to microtubules, apparently because they bind to the same pocket on b‐tubulin. It is clear that this common binding site is of immense importance for controlling the assembly of tubulin polymers. This pocket in b‐tubulin, where microtubule‐stabilizing drugs from different organisms can bind, also binds the assembly‐promoting repeat motifs of tau protein (and other microtubule‐associated proteins [MAPs]). It lies above the b‐sheet of the second domain and next to the core helix (Figs. 3 and 4), so that anything that fills it is in contact with the core helix and with the M‐loop. When assembled in microtubules, the pocket is located on the inside face. In a‐tubulin, the corresponding

272

AMOS AND SCHLIEPER

pocket is closed by the extended L‐loop. One proposed mechanism for stabilizing the assembled state is that structures that hold the M‐loop in place stabilize lateral contacts between protofilaments (Li et al., 2002; Nogales et al., 1999). An alternative possibility is that these structures hold the GTPase and second domains in a relative orientation that favors the straight protofilament conformation, which would otherwise be compromised by GTP hydrolysis (Amos and Lo¨ we, 1999). The two mechanisms may be combined. The effect of the GTP analogue GMPCPP (Hyman et al., 1992), which tubulin can hydrolyze only very slowly, shows that the GTP‐bound state of b‐tubulin makes microtubules very stable. The conformation of a‐tubulin is permanently stable because of the nonexchangeable, nonhydrolyzed GTP as well as the presence of the extended loop. It is not really clear why it is advantageous to have a permanently stable a‐subunit, but it may be because it makes the properties of the two ends of a microtubule more distinct. GTP binds on the other side of the core helix from Taxol, and the contact may exert a similar influence on the relative orientations of the GTPase domain, the core helix, and the second domain. The presence of GTP’s g‐phosphate will also have a direct effect on the interface between a‐ and b‐tubulin, strengthening the bond between heterodimers. It is notable that Taxol‐stabilized microtubules are relatively flexible and brittle (Dye et al., 1993; Venier et al., 1994), suggesting that the inter‐dimer bonds are not strengthened in this case.

IV.

Structural MAPs

In interphase, microtubules are stabilized by several kinds of proteins that are found all along microtubules and are called MAPs. They tend to have repeating domains, which allow each MAP molecule to associate with more than one tubulin dimer. This produces a doubly effective method of controlling assembly, in that the conformations of several tubulin dimers may be individually stabilized and the stabilized subunits are also cross‐ linked. The binding of these structural MAPs is in turn controlled by kinases and phosphatases (Cassimeris and Spittle, 2001). During mitosis they are phosphorylated and detach from tubulin, whose assembly and disassembly comes under the control of proteins that operate more at the ends of microtubules. Differentiated cells, such as neurons, do not divide. However, as microtubules and MAPs are slowly transported along axons (Baas and Buster, 2004), the MAPs may be phosphorylated in particular places, at times when structural plasticity is required for making synapses or other contacts.

MICROTUBULES AND MAPS

A.

273

The Tau Family

The mammalian neuronal proteins MAP2 and tau belong to a widespread family of MAPs. MAP2 is concentrated in dendrites and the taus, a set of smaller polypeptides, are found in axons. In non‐neuronal mammalian cells, there is a high molecular weight protein called MAP4. In Xenopus species a similar protein is called XMAP230. Homologous proteins have also been found in invertebrates (Goedert et al., 1996; Heidary and Fortini, 2001). All are extended molecules, which are lacking in detectable secondary structure and unusually heat stable. Each consists of a microtubule‐ binding domain, containing from one to five copies of a semiconserved motif, preceded by an N‐terminal ‘‘projection’’ domain that extends from the outer surface of the microtubule. The side‐to‐side spacing between microtubules containing MAPs depends on the size of the projection domains, which seem to repel rather than crosslink neighboring microtubules. They do, however, contain sites that interact with other proteins, such as the cortical layer, including actin, on the cell membrane (see Cassimeris and Spittle, 2001 for further references). When overexpressed in vivo or added in excessive amounts to microtubules in vitro, tau accumulates on microtubule surfaces and interferes with the movement of motor proteins along them (Ackmann et al., 2000; Stamer et al., 2002). But at normal levels, tau and related MAPs clearly assist motor transport by creating space around microtubules (Chen et al., 1992). Since Taxol and other microtubule‐stabilizing drugs are each found only in a specific group of organisms, the highly conserved site on the inside surface of b‐tubulin in which they sit presumably evolved to bind some more widespread natural stabilizing agents, such as the structural MAPs. There is, indeed, evidence that at least part of the repeating motif in tau binds to a site on b‐tubulin that overlaps with the Taxol‐binding site (Kar et al., 2003a). This was shown by labeling one of the motifs with a nanogold particle and localizing the gold by 3D analysis of electron micrographs. The specimens were microtubules assembled with tubulin and tau and then decorated with kinesin motor domains which, as mentioned above, bind to tubulin dimers with a characteristic polarity and allow the distinction between a‐ and b‐tubulin. Difference maps between images of microtubules containing labeled and unlabeled tau gave a peak on the inside surface of b‐tubulin (Fig. 5). Supporting evidence that Taxol and discodermolide both compete with tau for overlapping binding sites came from binding assays in which pelleted microtubules contained fewer tau in the presence of these drugs (Kar et al., 2003a,b). Sequence comparisons suggest that the tau repeat motif binds to b‐tubulin in a similar way as the extended L‐loop closes (and stabilizes) the pocket in a‐tubulin. The

274

AMOS AND SCHLIEPER

conserved sequence (TVVPGGDL) here closely resembles the PGGG signature in the sequence THVPGGGN from tau and in the other conserved repeats found in tau and related MAPs. These repeat motifs could be the natural substrates to bind to the Taxol‐binding pocket on the inside of the microtubules. It is also possible that successive repeat domains bind along the shallow helices of tubulin monomers and link three or four adjacent protofilaments together. This would explain rapidly growing triplets of protofilaments in some conditions when tubulin and MAPs are mixed (Mandelkow et al., 1984). Regions of tau outside the repeat region almost certainly bind to sites on the outer surface and support the N‐terminal domain that projects out. This projection domain is thought to have important roles in determining the spacing between microtubules in axons and possibly in binding to other structures such as the axonal membrane (Baas and Buster, 2004). The link between the N‐terminal projection domain of tau on the outside of the microtubules and the repeat motifs on the inside surface could be provided by tau’s proline‐rich flanking region; it would thread through one of the holes between protofilaments. A model of the arrangement of a complete tau molecule in a microtubule is shown in Fig. 6A. Clearly this sort of arrangement can only be reached by coassembling tubulin and tau, which would happen naturally in vivo. Other models in which tau binds only to the outer surface (e.g., Al‐Bassam et al., 2002), apparently overlapping with the kinesin binding site, are based on experiments in which preassembled microtubules were stabilized with Taxol before the addition of tau. The difference between the two modes of binding has been nicely highlighted by a stopped‐flow kinetic analysis: Makrides et al. (2004) showed that labeled tau bound to preassembled microtubules could exchange freely with freshly added unlabeled protein, whereas coassembled tau is nonexchangeably bound to the polymer. Besides filling the pocket on b‐tubulin that seems to be a primary control center for assembly, tau and related MAPs provide two additional forms of stabilization. First, the loops that occupy the pockets are interconnected in the repeat domain, thus cross‐linking three or four dimers, probably in adjacent protofilaments (Kar et al., 2003a). Second, the molecules have other domains that almost certainly bind well to the outer surface of a microtubule and probably run along a protofilament covering several tubulin dimers (Kar et al., 2003b). Figure 6B shows the charge distribution of tau. The three different domains have characteristic net charges. The negatively charged N‐terminal projection domain is repelled by the outer microtubule surface, which has the same charge. The positive proline‐rich region binds well to this surface, whereas the less positive repeat region could bind at the inside of the microtubules. MAPs

MICROTUBULES AND MAPS

275

are known to have a stiffening effect on microtubules (Dye et al., 1993; Felgner et al., 1997). Overall, their binding should both favor the straight heterodimer conformation and hold the protofilaments together.

B.

MAP1A and MAP1B

The two closely related neuronal proteins MAP1A and MAP1B are also highly extended and apparently unstructured in solution. Their sequences contain many short repetitive motifs, especially in the N‐terminal microtubule‐binding domain where there is a predominance of basic residues, and it is likely that they bind to the negatively charged outer surface of the microtubule, although little has been published about the structure of MAP1. The repeats bear no obvious relationship to any part of the microtubule binding domains of tau or MAP2. In vitro, MAP1A has been shown to increase nucleation and to stimulate microtubule elongation (Pedrotti and Islam, 1996), but it has less power than MAP2 to stabilize the polymer (Vaillant et al., 1998). However, MAP1B can apparently substitute for tau in vivo, since knockout mice are viable without one or other of these proteins but die if both are missing (Takei et al., 2000).

C.

STOPs

In vertebrates, resistance of microtubules to cold is largely due to their association with the class of MAPs known as stable tubule only polypeptides (STOPs). These are calmodulin‐binding and calmodulin‐regulated proteins, and their microtubule‐stabilizing activity has been ascribed to microtubule‐binding motifs that also bind calmodulin. The repeating sequence is different from that in tau and MAP2, but includes an EGGP motif that may fold up as a sharp turn like the extended L‐loop of a‐tubulin. It is, therefore, conceivable that part of each STOP repeat binds to the pocket on b‐tubulin. The proteins seem to have more specific functions in nerve cells than simply to stabilize microtubules, since STOP‐ null knockout mice show synaptic defects and abnormal behavior (Bosc et al., 2003).

D. Doublecortin Doublecortin is so named because its mutation leads to brain developmental disorders that can include the formation in the brain of a ‘‘double cortex’’ due to an additional band of aberrantly placed neurons. The 30‐kD N‐terminal domain of the protein binds to and stabilizes microtubules

276

AMOS AND SCHLIEPER

Fig. 6. Model of tau molecules and DCX domains interacting with a microtubule. (A) Diagram of a microtubule with the inside surface exposed. The acidic C‐terminal segment of each tubulin monomer projects from the outer surface (short black stubs). The longer projections represent the N‐terminal domains of tau molecules whose C‐ terminal segments, including the repeat motifs, bind strongly to the inside of the microtubule (Kar et al., 2003a,b). Tau’s proline‐rich region binds to the outer surface along the protofilaments; the connecting piece of polypeptide is sandwiched between two protofilaments during assembly. Kinesin motor domains are shown bound to the

MICROTUBULES AND MAPS

277

in vitro. It enhances polymerization and can also bundle microtubules because the tubulin binding part of the molecule consists of a tandem repeat of ‘‘DCX domains’’ that may cross‐link two tubulin subunits either in the same or different microtubules (Fig. 6C). Another protein, LIS1, appears to have a similar function in the brain but targets microtubule ends rather than binding all along their length (Tanaka et al., 2004). The structure of the N‐terminal DCX domains has been solved to high resolution by NMR and X‐ray crystallography (Fig. 7A). The domains have also been imaged bound to microtubules (Moores et al., 2004), making doublecortin the only MAP, apart from tau and MAP2c (Al‐Bassam et al., 2002; Kar et al., 2003a) to have been studied so far by 3D analysis of EM images. When microtubules were decorated with molecules consisting of tandem domains, single globular domains were seen binding in the longitudinal groove between protofilaments, at intervals of 8 nm, the same as the spacing of ab‐heterodimers (Fig. 6A). The most significant observation from the EM analysis was that the molecules selectively bound to and stabilized 13‐protofilament microtubules, presumably because each domain binds in a stereospecific way to two protofilaments. This result shows, for the first time, that a basal template is not necessary to specify 13‐protofilament microtubules, but that their diameter can be controlled by proteins that bind to microtubule sides. It is not clear whether both domains from each DCX molecule were bound identically, linking adjacent sites, or whether only one was bound and the other was tethered to it, but was too mobile to appear in the averaged image of many molecules. The two domains are known to have slightly different properties: the

protofilament on the left (light gray dropshapes). They can probably move past the tau projections without difficulty. The binding site for doublecortin DCX domains (Moores et al., 2004) is also indicated (white ovals). It is close to the tau/Taxol binding site but on the external, rather than the internal, side of the M‐loop. Its position would also not interfere with kinesin binding. (B) Distribution of basic and acidic amino acid residues in human 4‐repeat tau (modified from Goedert et al., 1994). The fairly acidic N‐terminal segment forms a projection (shown in A), that is repelled by the negatively charged tubulin surface. The proline‐rich region, with a net positive charge, interacts quite strongly with the microtubule surface. The net positive charge of the repeat region is much less than the proline‐rich region. Its sequence consists of three or four semi‐ conserved repeats with motifs similar to the extended loop in a‐tubulin (Kar et al., 2003a). Each motif binds to b‐tubulin in the pocket that corresponds to the extended loop in a‐tubulin (i.e., where Taxol has been seen to bind). (C) Scheme for the doublecortin molecule. The N‐terminal DCX domain (N‐DC) binds only to microtubules. Its structure is shown in Figure 7A. The C‐terminal DCX domain (C‐DC) binds to both microtubules and soluble tubulin dimers.

278

AMOS AND SCHLIEPER

Fig. 7. Atomic models of MAP domains. Three globular domains all bind to microtubules but do not appear to be homologous. (A) Human doublecortin N‐terminal DCX domain (Protein Data Bank [PDB] ID code: 1MJD), solved by NMR, shows a ubiquitin‐ like fold (Kim et al., 2003). The domain has been docked into a 3D map of the protein bound to microtubules, as reconstructed from EM images (Moores et al., 2004). The regions marked PF are thought to make contact with two adjacent tubulin protofilaments. (B) N‐terminal CAP‐Gly domain, as found in CLIPs and the dynactin heavy chain. The structure shown is from C. elegans (PDB‐ID: 1LPL), solved by X‐ray crystallography (Li et al., 2002), and contains three b‐strands. (C) The N‐terminal domain of human EB1 (PDB‐ID: 1PA7), solved by X‐ray crystallography (Hayashi and Ikura, 2003), contains a single calponin‐homology (CH) domain consisting of several a‐helices. Tandem CH domains are often found in actin‐binding proteins; this is the first example of a tubulin‐ binding CH domain.

N‐terminal domain binds only to microtubules, whereas the C‐terminal domain binds to both microtubules and soluble tubulin dimers (Kim et al., 2003). The second domain might, therefore, be active only at a microtubule end.

MICROTUBULES AND MAPS

279

V. Microtubule Destabilizers

A. Stathmin (Op18) The tubulin‐sequestering protein called stathmin has been crystallized with tubulin (Gigant et al., 2000; Ravelli et al., 2004). Each molecule runs alongside a pair of tubulin heterodimers and its amino‐terminal domain caps one of the a‐tubulin subunits, preventing any further interactions (Fig. 4C). In addition to approximately 12‐degree bends between the tubulin monomers, there are changes within each subunit when compared with the straight conformation (Fig. 4A and B), including a relative rotation between the GTPase and activation domains. Contacts between subunits are preserved by local movements of helices H6, H7, and H8 and loop T5.

B.

Katanin

Katanin, a member of the AAA superfamily (ATPases associated with different cellular activities), is a heterodimer of 60‐ and 80‐kDa subunits. In the presence of ATP and microtubules, it forms a transient hexadimer, which can be seen in electron micrographs. It is capable of disrupting contacts between ab‐tubulin heterodimers in the wall of microtubules. This results in severing the microtubules into short pieces (reviewed by Quarmby, 2000, and Vale, 2000). One important role is in preparing free lengths of microtubule for transport along axons (Baas and Buster, 2004). Davis et al. (2002) proposed that katanin binds preferentially to lattice defects in microtubules, for example where an ab‐tubulin building block is missing, rather than to random locations. These defects do occur, at least in microtubules that are assembled in vitro, as shown by scanning force microscopy with single‐protein resolution (Schaap et al., 2004). The severing mechanism of microtubule dynamics modulation is intriguing, and is consistent with in vivo results from two members of the katanin family: Human spastin promotes microtubule disassembly (Errico et al., 2002) and the MEI–1/MEI–2 complex is needed for C. elegans meiosis (Srayko et al., 2000). As MEI–1/MEI–2 interacts differently with microtubules containing different b‐tubulin isoforms, it could directly regulate meiotic spindle microtubule function. The specific binding site was mapped to the extreme C‐terminus of b‐tubulin (Lu et al., 2004). Katanin might be regulated by several mechanisms. One way would involve MAPs interfering with katanin’s access to the microtubule lattice (McNally et al., 2002).

280

AMOS AND SCHLIEPER

C. EMAP A 77‐kDa WD repeat protein, called EMAP, from sea urchins reduces the frequency of rescues by eightfold without producing a change in the frequency of catastrophe (Hamill et al., 1998). The mechanism for this activity is still to be investigated. EMAP‐like proteins have also been identified in starfish, sand dollars, and mammals.

D. Kinesins as Regulators Many members of the kinesin family are involved in regulating microtubule dynamics. The kinesin‐13 (Kin‐I) subfamily (see Chapter 8) do not move along microtubules, but instead use energy provided by ATP to depolymerize microtubules at their ends (Desai et al., 1999; Homma et al., 2003; Walczak et al., 1996). Binding of ATP to a kinesin‐13 motor domain attached to a tubulin heterodimer causes the latter to bend and the protofilament to roll up. The activity is catalytic since the motor domain dissociates from the bent tubulin after ATP hydrolysis. Two crystal structures of kinesin‐13 motor domains show relatively small differences from conventional kinesins (Ogawa et al., 2004; Shipley et al., 2004). The two groups agree that an extended L2‐loop is of major importance. Whereas most of the loops that interact with tubulin lie on the plus‐end half of the motor domain and are thought to interact with b‐tubulin, the L2‐loop extends towards the minus end, in a suitable position to interact with a‐ tubulin. When attached to microtubules, motor proteins in the strongly binding states (with ATP or a nonhydrolyzable analogue bound to the motor domain or with an empty nucleotide binding site) produce a small bend in the tubulin heterodimer (Fig. 8; Hirose et al., 1999). If this activity is exaggerated by more strongly binding versions of L2 and other loops, it can presumably cause depolymerization. Ogawa et al. (2004) also found the class‐specific N‐terminal neck to be a long, rigid helical structure that would extend into the inter‐protofilament groove and thereby inhibit the motor from binding except at the ends of microtubule. Most kinesin-13 molecules also have an additional N‐terminal domain compared with conventional kinesin, and this may also have a role in targeting microtubule ends. Even some kinesins that act as normal motor proteins appear to be involved in controlling the length of microtubules in cells (Hunter and Wordeman, 2000). The minus‐end‐directed motor Ncd has been seen to shorten microtubules in vitro (Endow et al., 1994). It is also required for efficient transport of the end‐controlling Msps protein (see section VI.B) to spindle poles (Ohkura et al., 2001). Other, nondepolymerizing kinesins may act solely by transporting other molecules to the ends of microtubules.

MICROTUBULES AND MAPS

281

Fig. 8. 3D EM shows the effect of motor binding on tubulin structure. Sections through 3D EM maps of microtubules decorated with conventional kinesin (K) or Ncd (N) motor domains: the sections pass through the microtubule axis (left‐hand edge of each box), through a tubulin protofilament (tub) and then through the middle of the attached heads. The weak interaction (with ADP in the nucleotide‐binding site) appears similar for both kinesin and Ncd, although Ncd may make an additional bond to a‐tubulin via the L2 loop, which is very short in kinesin. Strong binding of either motor (without nucleotide or with the nonhydrolyzable analogue AMPPNP (adenosine–50 ‐[(b,g)‐imido]triphosphate) bound) appears to induce the same change in the tubulin dimer, as indicated by dashed lines through the tubulin monomers; compared with the ADP state (or with an undecorated microtubule, not shown), in which both monomers appear equal, the outer tip of the b‐tubulin subunit tilts closer to that of the a‐tubulin subunit (Hirose et al., 1999).

Several kinesin members have microtubule binding sites that are separate from the motor domain and lead to cross‐bridging of microtubules. MKLP1 is an interesting example. In vitro it can bundle and slide anti‐ parallel microtubules apart. Microtubule binding involves a highly basic N‐terminal extension of the motor, which may interact with the C‐terminal acidic tail of tubulin (Mishima et al., 2004). It is thought to help elongate the anaphase spindle by cross‐bridging microtubules that overlap in

282

AMOS AND SCHLIEPER

the midzone. However, it also has a role in triggering cytokinesis by cross‐ bridging astral microtubules from opposite poles (Inoue et al., 2004).

E. MINUS The small polypeptide MINUS is a novel type of microtubule regulator that has been proposed to act as a nucleation suppressor (Shahani et al., 2001).

VI.

Proteins That Control Microtubule Location

Apart from the structural MAPs discussed above, there are many proteins that affect microtubule behavior. In cells, many factors affect plus‐end dynamics and also play roles in the assembly of higher‐order structures. Growing microtubules use dynamic instability to search with their tips for structures that are capable of capturing them. Thus, microtubules are ‘‘guided’’ toward specialized membrane domains, chromosomes, and other components of the cell. A wide variety of assembly regulators and adapters that form links with these structures are found at microtubule ends, especially the plus end. Their distributions and activities are reviewed in detail elsewhere (Cassimeris, 1999; Cassimeris and Spittle, 2001; Kline‐Smith and Walczak, 2004; Miyamoto et al., 2003; Schroer, 2001). Plus‐end tracking proteins that actively remain at the tips of growing microtubules are reviewed by Schuyler and Pellman (2001), Galjart and Perez (2003), and Vaughan (2004). Here, we try to summarize what is known from a structural point of view. The minus end is less active but is also subject to stabilizing and destabilizing activity (Dammermann et al., 2003).

A.

g‐Tubulin Ring Complexes

The protein complexes that make up microtubule‐organizing centers usually include a third kind of tubulin known as g‐tubulin, which binds, but probably does not hydrolyze, GTP. There is tentative evidence that it may be capable of assembling into protofilaments (Incla´ n and Nogales, 2001; Llanos et al., 1999). g‐tubulin complexes have been detected in the form of ~25‐nm diameter rings in centrosomes (g‐TuRCs; Moritz and Agard, 2001), but cells also contain smaller g‐tubulin–containing complexes ( Job et al., 2003; Leguy et al., 2000). Also present in each type of complex are various special MAPs, known as Dgrips in the case of the Drosophila complexes that have been studied in detail. g‐TuRCs have been

MICROTUBULES AND MAPS

283

proposed to provide a basal template for the 13‐protofilament microtubule lattice (Moritz and Agard, 2001). In an alternative model, based on the observation that both interfaces involved in the formation of protofilaments are conserved in g‐tubulin, the ring is proposed to be a stored form of a g‐tubulin protofilament that can straighten to stabilize an ab‐ tubulin microtubule lattice by lateral interactions (Erickson, 2000). It is even possible that g‐tubulin merely serves as a seed for one ab‐tubulin protofilament that initiates microtubule assembly (Leguy et al., 2000).

B. XMAP215, ch‐TOG, Msps Family This very widespread family of proteins is important in controling the local interactions of microtubules with many other structures (see, e.g., Ohkura et al., 2001). XMAP215 is an assembly promoter that can increase microtubule plus‐end growth by seven‐ to ten‐fold but also appears to stabilize minus ends. Its usual effect is to increase the average length of microtubules (Kinoshita et al., 2002). According to Popov et al. (2002), it is the N‐terminal domain (residues 1–560) of XMAP215 that protects plus ends from catastrophe promoters such as the KinI kinesin XKCM1. These workers found that the C‐terminal domain (residues 1168–2065) also bound to tubulin, even as a separate fragment, but did not stabilize assembly. However, there seems to be no general agreement about which parts of the molecules are important for microtubule stabilization (Ohkura et al., 2001). Two hypotheses have been considered for the mechanism by which XMAP215 promotes polymerization of microtubule plus ends (Spittle et al., 2000). By oligomerizing soluble tubulin dimers, it might catalyze the addition of several dimers per association event. Alternatively, it might promote a defined structure at the growing end that either adds dimers more rapidly or is less likely to go into the paused state. Cassimeris et al. (2001) showed that the XMAP215 molecule from Xenopus is long, thin, and flexible and can bind to protofilaments along their long axis. Its length, 60  18 nm, is sufficient to span seven or eight tubulin dimers. Incubation of XMAP215 and tubulin at 4 C resulted in the assembly of curved protofilaments. A major part of the polypeptide (Fig. 9A) is made up of HEAT motifs (Huntingtin/Elongation factor 3/ protein phosphatase 2A/TOR1 repeats, which are related to armadillo repeats and typically form protein–protein interaction surfaces). The atomic structure of a series of HEAT motifs (Fig. 9B) shows a curved row of a‐helical hairpins. Proteins in the XMAP215 family may, therefore, consist of concertinas of short helices with more extended connecting regions. The smaller proteins Stu2p from Saccharomyces cerevisiae and Alp14 and Dis1 from Schizosaccharomyces pombe, which are homologous to XMAP215,

284

AMOS AND SCHLIEPER

Fig. 9. XMAP215/Stu2p polypeptides. (A) A plan showing the arrangement of domains. Xenopus XMAP215 has five TOG domains, each containing up to five HEAT repeats. Similar proteins are found in humans (ch‐TOG), Drosophila (Msps), and C. elegans (Zyg-9). Homologous proteins from S. cerevisiae (Stu2p) and S. pombe (Alp14 and Dis1/Mtc1) are less than half the length of those from higher organisms. (B) The structure of part of a typical HEAT‐motif protein (Groves et al., 1999), showing what one of the TOG repeats may look like. If the whole molecule is long enough to cover 7–8 tubulin dimers (Cassimeris et al., 2001), the inter‐repeats must have a more stretched‐out structure; each TOG repeat plus an inter‐repeat may bind to an 8‐nm dimer.

have fewer repeats in the N‐terminal domain (Fig. 9A), and there is no obvious sequence homology in the C‐terminal domain. Another difference is that Stu2p slows polymerization rates in vitro rather than increasing them (van Breugel et al., 2003). Stu2p depolymerized microtubules in vitro by binding directly to plus ends, hindering tubulin dimer addition and increasing catastrophe rates. A stu2 mutant had longer microtubules than normal in interphase, which may correspond to the in vitro conditions, whereas during mitosis microtubules were shorter than normal. XMAP215 can also have a destabilizing effect under certain conditions (Shirasu‐Hiza et al., 2003). EM images of microtubules assembled in vitro with GMPCPP and incubated with XMAP215 revealed a rolled‐up structure characteristic of shrinkage. This observation suggests that XMAP215 can operate a protofilament peeling mechanism similar to that of KinI kinesins. Shirasu‐Hiza et al. proposed that the plus ends of microtubules stabilized by GMPCPP, previously thought to resemble the GTP caps on growing microtubules,

MICROTUBULES AND MAPS

285

may instead mimic the paused state. Since Chre´ tien et al. (1995) found that loss of the sheetlike protofilament extensions of growing microtubule plus ends correlated with slower growth, it is possible that the paused state corresponds to a blunt‐ended, closed tube structure. It seems likely that XMAP215 and Stu2p both disrupt the paused state and either increase polymerization or cause depolymerization, depending on conditions such as free tubulin concentration. In vivo, their activity would vary with the cell cycle or the location within the cell. At the start of mitosis, this family of proteins is important in organizing spindle poles. In Drosophila, for example, Msps (mini‐spindles) interacts via its C‐terminal with D‐TACC (the Drosophila ‘‘transforming acidic coiled‐coil–containing’’) protein (Lee et al., 2001). This and other members of the family interact, directly or indirectly, with a wide variety of other proteins (Ohkura et al., 2001).

C. CLIPs/þTIPs and CLASPs Proteins that bind specifically to microtubule plus ends have been named plus‐end tracking proteins (þTIPs; Carvalho et al., 2003). They are a subset of cytoplasmic linker proteins (CLIPs), some of which appear to be capable of binding along the length of microtubules but associate preferentially with the ends. Green fluorescent protein‐labeled CLIP170 association with growing tips has been observed directly (Perez et al., 1999). Its behavior has been explained either as surfing (sliding along with the growing end) or as ‘‘treadmilling.’’ Specific binding to growing plus ends may occur because these proteins bind preferentially to narrow sheets rather than closed microtubules; possibly, part of the binding site lies on the side of a protofilament or even on the inside surface. Alternatively, they may be able to distinguish between tubulin containing GTP rather than GDP in the exchangeable site. A final possibility, which could include aspects of the previous two, is that they bind to soluble tubulin, making it more likely to add to a microtubule and then dissociate from the polymer after the conformational change in tubulin due to GTP hydrolysis. In fission yeast, growing microtubules are stable until they reach the cell end, provided Tip1p is bound. In its absence, catastrophes are no longer restricted to times when the microtubule ends contact the membrane at the end of the cell (Brunner and Nurse, 2000). The CAP‐Gly motif is a conserved glycine‐rich sequence of about 42 residues found in the microtubule‐binding domains of the various CLIPs, in the heavy chain (P150) of the dynactin complex, and in the Golgi protein GMAP210 (Infante et al., 1999; Pernet‐Gallay et al., 2002). The structure of a globular domain (Fig. 7B) containing this motif has been

286

AMOS AND SCHLIEPER

solved by X‐ray crystallography (Li et al., 2002). CLIP170 (also called ‘‘restin’’) has two such domains near its N‐terminus and binds specifically to the plus ends of growing microtubules (Perez et al., 1999). CLIP115, a neuronal CLIP, has a single CAP‐Gly domain, as does the P150 chain of dynactin. The CAP‐Gly domain(s) are typically followed by a stretch of sequence that is able to dimerize in the form of a coiled-coil. Finally, a smaller globular C‐terminal region of each protein is thought to be involved in linking to membranous structures. Orbit/MAST proteins, also known as CLIP‐associated proteins (CLASPs), are involved in the regulation of microtubule dynamics and bind to microtubule plus ends via CLIP115 or CLIP170. Active CLASP suppresses microtubule assembly and axon outgrowth (Lee et al., 2004), whereas activated adenomatous polyposis coli protein (APC; see below) promotes microtubule assembly and axon outgrowth.

D. The Dynactin Complex Dynactin is a 1200‐kD protein complex that is usually found in association with the motor protein, cytoplasmic dynein, an equally large complex. Together, they are active in a wide variety of processes such as axonal transport, organelle movement, nuclear localization, and chromosome separation (reviewed by Allan, 2000; Holleran et al., 1998). Although dynein is able to bind to membranes and membranous organelles by itself, it apparently cannot move them for long distances along microtubules without help from dynactin (King et al., 2003). It remains to be seen how the two complexes interact to produce movement. Dynactin itself can bind all along microtubules but shows a preference for the ends of microtubules in cells as a result of association with CLIP170 (Valetti et al., 1999; Vaughan et al., 2002). The dynactin complex contains at least 10 distinct subunits (Schroer, 1996). The dominant structure seen by EM (Eckley et al., 1999; Schafer et al., 1994) is a 37‐nm actin‐like filament (consisting mainly of monomers of the actin‐related protein, Arp1), from which a sidearm projects (Fig. 9). A dimer of dynactin heavy chains (P150glued) is the main component of this sidearm. Each chain has a globular CAP‐Gly domain at its N‐terminus, followed by a long coiled‐coil segment that effects the dimerization. Some other components of the complex have been identified by antibody labeling (Schafer et al., 1994). A central region of P150 is linked to the dynein intermediate chain by a protein called dynamitin (Valetti et al., 1999; Vaughan and Vallee, 1995), which forms the stem of the dynein complex (Fig. 10A and C).

MICROTUBULES AND MAPS

287

Fig. 10. The dynactin complex and cytoplasmic dynein. (A) Model of the 3D structure of dynactin based on EM images of shadowed molecules (redrawn from Schroer, 1996). The cargo‐binding domain is a short filament assembled from Arp1 (centractin) plus some standard actin, with capping proteins at each end. The P50 (dynamitin) component helps to link the filament to the heavy chain P150glued. Globular CAP‐Gly domains (see Figure 7B) at the end of the thin sidearm bind to tubulin. (B) The binding regions along the dynactin heavy chain. (C) Hypothetical model of the interaction between dynactin and cytoplasmic dynein. Each large complex can bind independently with both microtubules and cargo, but they also associate with each other. Each of two dynein heavy chains (dyn HC) forms a multidomain globular head from which extends a long thin ‘‘stalk.’’ The stalks of freshly isolated active dimers are closely associated with each other (Fan and Amos, 2001). A small globular domain at the end of each stalk has a microtubule (MT)‐binding site. Dynein intermediate chains (dyn IC) form part of the ‘‘stem’’ of the molecule, used for associating with dynactin and with cargo. What is known about the atomic structures of dynein subunits has been reviewed recently by Sakato and King (2004).

288

AMOS AND SCHLIEPER

E. EB1 and APC End‐binding protein 1 (EB1) is involved in a variety of macromolecular complexes at microtubule plus ends, including those that make contact with the cell membrane (Bu and Su, 2003). It cooperates with another plus‐end protein, the tumor suppressor, APC (not to be confused with APC/C, the anaphase‐promoting complex/cyclosome, which is a ubiquitinase). Loss of this interaction, due to mutations in APC, may lead to colon cancer. EB1 and APC both localize to kinetochores (containing plus ends) during mitosis (Fodde et al., 2001). In centrosomes (where minus ends are located), they are assumed to be associated with the plus ends of very short, newly initiated microtubules (Rehberg and Graf, 2002). The globular N‐terminal CH domain of EB1, for which a crystal structure is known and which shows no obvious homology to the globular CAP‐ Gly domain of the CLIPs (Fig. 7C), binds preferentially to the plus ends of microtubules. The extended C‐terminal domain dimerizes via a stretch of coiled-coil and also binds to APC (Fig. 11). Thus, APC is targeted to plus ends too, where it promotes microtubule polymerization. Neither of these proteins alone has this effect, which is regulated by phosphorylation of APC by Cdc2, disrupting EB1 binding to the N‐terminal region of APC. This N‐ terminal domain forms a weakly bonded coiled‐coil dimer (Day and Alber, 2000), while the middle of the polypeptide includes a cysteine‐rich region, with armadillo repeats that interact with b‐catenin. It is not yet known how APC promotes tubulin assembly. However, the C‐terminal domain contains a microtubule‐binding site with a high proportion of positively charged amino acids and shows some sequence homology to the proline‐rich region of tau (Deka et al., 1998). The EB1 binding domain is at the C‐ terminus of APC and includes a binding site for a PDZ protein–protein interaction domain. PDZ domain proteins are frequently associated with the plasma membrane.

F.

Hook Proteins

The Hook family of proteins resembles CLIPs in that they consist predominantly of coiled-coil. However, the N‐terminal domains, which attach to microtubules, differ in sequence from other known microtubule‐ binding domains. The C‐terminal domains are adapted to bind specifically to particular organelles. Hook3, for example, appears to play a role in defining the architecture and localization of the mammalian Golgi complex (Walenta et al., 2001). This helps to explain why the integrity of the Golgi complex is completely dependent on the presence of the microtubule network.

MICROTUBULES AND MAPS

289

Fig. 11. Model of EB1/APC complex. Summary of what is known about different regions of the two polypeptides EB1 and APC and their interactions. The globular N‐ terminal CH‐domain of EB1 is shown in Figure 7C. The complete molecule, which binds preferentially to microtubule plus ends, dimerizes via a stretch of coiled-coil. This region also binds to APC, thus targeting APC to the plus ends, too. The shape of APC is purely hypothetical.

G. Centrosomal Coiled‐Coil Proteins Animal spindle poles usually consist of one mature and one developing centriole, supported by a meshwork of material whose basic components now appear to be coiled‐coil proteins. The spindle poles of non‐animal cells apparently consist of similar matrices, but without the centrioles. A picture is beginning to emerge of the dynamic and complex roles of some of these matrix proteins. TACC (mentioned above as interacting with members of the XMAP215/ch‐TOG family), NuMA (nuclear and mitotic apparatus protein; Dionne et al., 1999), and TPX2 (targeting protein for Xenopus kinesin‐like protein 2, Xklp2; Garrett et al., 2002; Wittmann et al., 2000) are all dimeric coiled-coils with a tendency to form filamentous polymers. They are found concentrated in centrosomes where they are thought to help in organizing centrosomal microtubules, including direct interaction with tubulin. When the C‐terminal domain of TACC was overexpressed in cells, large polymers that formed in the cytoplasm interacted with both tubulin and microtubules. Full‐length TACC proteins formed similar polymers, but their interaction with tubulin was regulated by the cell cycle (Gergely et al., 2000). NuMA is also known to have microtubule‐ binding capacity near its C‐terminus. NuMA and TPX2 are sequestered in the nucleus before breakdown of the nuclear membrane but are then carried by interaction with dynein/dynactin (Gaglio et al., 1995; Merdes et al., 1996) to the poles of the developing spindle. There, TPX2 has a role of targeting Xklp2 to microtubule minus ends, but in late anaphase TPX2 and Xklp2 become relocalized from the spindle poles to the midbody

290

AMOS AND SCHLIEPER

(Wittmann et al., 2000). Another family of coiled‐coil proteins that includes MAP65 in plants (Mu¨ ller et al., 2004), Ase1 in budding yeast (Schuyler et al., 2003), Mto1p in fission yeast, and PRC1 in animals (Kurasawa et al., 2004) stabilizes central spindle microtubules in the late stages (anaphase/telophase) of cell division. MAP65 appears to be required to maintain the structures of phragmoplast microtubules that are involved in constructing a new cell wall between two daughter cells.

VII.

Conclusions

Protein domains that bind to tubulin and to microtubules come in a variety of forms: single long a‐helices (stathmin), helical coiled-coils (APC), helical hairpins (XMAP215/ch‐TOG), extended random coils (tau), and three unrelated globular folds (EB1, CAP‐Gly, and doublecortin). Most of these have not yet been observed by EM or x‐ray crystallography when bound to tubulin, but it seems likely that each type of structure binds differently, possibly without competing with the others or with the motor proteins that move along microtubules. Some proteins, such as tau, MAP1, and MAP2, bind equally well at any position along a microtubule, stabilizing the tubulin lattice directly. This is done by a combination of cross‐linking different tubulin dimers and of binding to specific sites that control either the conformation of the whole subunit or that of loops involved in binding to other subunits. The binding of DCX domains to sites in the grooves between protofilaments strongly favors the formation of 13‐protofilament microtubules, without requiring a basal template. The doublecortins seem to occur only in mammalian nerve cells, but it seems likely that molecules that control microtubule curvature in this way are widespread. Other proteins, such as XMAP215, CLIP170, or EB1, control microtubule dynamics by binding to the growing or shrinking ends. At present, we have a very limited understanding of how the ends are targeted or how their dynamics are influenced, but the current research in this area is likely to produce more information soon. Growth may be stimulated by stabilizing tubulin in its straight (GTP‐bound) conformation, shrinkage by promoting the curved (GDP‐bound) conformation. End‐binding proteins often work together to provide sophisticated control of the behavior of the ends.

Acknowledgments D. S. is supported by an award from the Leverhulme Trust to Dr Jan Lo¨ we, whom we also thank for inspiring discussions.

MICROTUBULES AND MAPS

291

References Ackmann, M., Wiech, H., and Mandelkow, E. (2000). Nonsaturable binding indicates clustering of tau on the microtubule surface in a paired helical filament‐like conformation. J. Biol. Chem. 275, 30335–30343. Al‐Bassam, J., Ozer, R. S., Safer, D., Halpain, S., and Milligan, R. A. (2002). MAP2 and tau bind longitudinally along the outer ridges of microtubule protofilaments. J. Cell Biol. 157, 1187–1196. Allan, V. (2000). Dynactin. Curr. Biol. 10, R432. Amos, L. A. (2000). Focusing‐in on microtubules. Curr. Op. Struct. Biol. 10, 236–241. Amos, L., and Klug, A. (1974). Arrangement of subunits in flagellar microtubules. J. Cell Sci. 14, 523–549. Amos, L. A., and Lo¨ we, J. (1999). How taxol stabilises microtubule structure. Chem. Biol. 6, R65–R69. Andreu, J. M., Oliva, M. A., and Monasterio, O. (2002). Reversible unfolding of FtsZ cell division proteins from archaea and bacteria: Comparison with eukaryotic tubulin folding and assembly. J. Biol. Chem. 277, 43262–43270. Baas, P. W., and Buster, D. W. (2004). Slow axonal transport and the genesis of neuronal morphology. J. Neurobiol. 58, 3–17. Bollag, D. M., McQueney, P. A., Zhu, J., Hensens, O., Koupal, L., Liesch, J., Goetz, M., Lazarides, E., and Woods, C. M. (1995). Epothilones, a new class of microtubule‐ stabilizing agents with a taxol‐like mechanism of action. Cancer Res. 55, 2325–2333. Bosc, C., Andrieux, A., and Job, D. (2003). STOP proteins. Biochemistry 42, 12125–12132. Brunner, D., and Nurse, P. (2000). CLIP170‐like Tip1p spatially organizes microtubular dynamics in fission yeast. Cell 102, 695–704. Bu, W., and Su, L. K. (2003). Characterization of functional domains of human EB1 family proteins. J. Biol. Chem. 278, 49721–49731. Carvalho, P., Tirnauer, J. S., and Pellman, D. (2003). Surfing on microtubule ends. Trends Cell Biol. 13, 229–237. Cassimeris, L. (1999). Accessory protein regulation of microtubule dynamics throughout the cell cycle. Curr. Opin. Cell Biol. 11, 134–141. Cassimeris, A., Gard, D., Tran, P. T., and Erickson, H. P. (2001). XMAP215 is a long thin molecule that does not increase microtubule stiffness. J. Cell Sci. 114, 3025–3033. Cassimeris, L., and Spittle, C. (2001). Regulation of microtubule‐associated proteins. Int. Rev. Cytol. 210, 163–226. Chen, J., Kanai, Y., Cowan, N. J., and Hirokawa, N. (1992). Projection domains of MAP2 and tau determine spacings between microtubules in dendrites and axons. Nature 360, 674–676. Chre´ tien, D., Fuller, S. D., and Karsenti, E. (1995). Structure of growing microtubule ends: Two‐dimensional sheets close into tubes at variable rates. J. Cell Biol. 129, 1311–1328. Chre´ tien, D., and Wade, R. H. (1991). New data on the microtubule surface lattice. Biol. Cell 71, 161–174. Dammermann, A., Desai, A., and Oegema, K. (2003). The minus end in sight. Curr. Biol. 13, R614–R624. David‐Pfeuty, T., Simon, C., and Pantaloni, D. (1979). Effect of antimitotic drugs on tubulin GTPase activity and self‐assembly. J. Biol. Chem. 254, 11696–11702. Davis, L. J., Odde, D. J., Block, S. M., and Gross, S. P. (2002). The importance of lattice defects in katanin‐mediated microtubule severing in vitro. Biophys. J. 82, 2916–2927.

292

AMOS AND SCHLIEPER

Day, C. L., and Alber, T. (2000). Crystal structure of the amino‐terminal coiled‐coil domain of the APC tumor suppressor. J. Mol. Biol. 301, 147–156. Deka, J., Kuhlmann, J., and Mu¨ ller, O. (1998). A domain within the tumor suppressor protein APC shows very similar biochemical properties as the microtubule‐associated protein tau. Eur. J. Biochem. 253, 591–597. Desai, A., and Mitchison, T. J. (1997). Microtubule polymerization dynamics. Ann. Rev. Cell Dev. Biol. 13, 83–117. Desai, A., Verma, S., Mitchison, T. J., and Walczak, C. E. (1999). Kin I kinesins are microtubule‐destabilizing enzymes. Cell 96, 69–78. Dionne, M. A., Howard, L., and Compton, D. A. (1999). NuMA is a component of an insoluble matrix at mitotic spindle poles. Cell Motil Cytoskel. 42, 189–203. Dye, R. B., Fink, S. P., and Williams, R. C. J. (1993). Taxol‐induced flexibility of microtubules and its reversal by MAP2 and Tau. J. Biol. Chem. 268, 6847–6850. Eckley, D. M., Gill, S. R., Melkonian, K. A., Bingham, J. B., Goodson, H. V., Heuser, J. E., and Schroer, T. A. (1999). Analysis of dynactin subcomplexes reveals a novel actin‐ related protein associated with the Arp1 minifilament pointed end. J Cell Biol. 147, 307–320. Endow, S. A., Kang, S. J., Satterwhite, L. L., Rose, M. D., Skeen, V. P., and Salmon, E. D. (1994). Yeast Kar3 is a minus‐end microtubule motor protein that destabilizes microtubules preferentially at the minus ends. EMBO J. 13, 2708–2713. Erickson, H. P. (1995). FtsZ, a prokaryotic homolog of tubulin? Cell 80, 367–370. Erickson, H. P. (1998). Atomic structures of tubulin and FtsZ. Trends Cell Biol. 8, 133–137. Erickson, H. P. (2000). g‐tubulin nucleation: Template or protofilament? Nat. Cell Biol. 2, E93–E96. Errico, A., Ballabio, A., and Rugarli, E. I. (2002). Spastin, the protein mutated in autosomal dominant hereditary spastic paraplegia, is involved in microtubule dynamics. Hum. Mol. Genet. 11, 153–163. Fan, J., and Amos, L. A. (2001). Antibodies to cytoplasmic dynein heavy chain map the surface and inhibit motility. J. Mol. Biol. 307, 1317–1327. Fan, J., Griffiths, A., Lockhart, D. A., Cross, R. A., and Amos, L. A. (1996). Microtubule minus ends can be labelled with a phage display antibody specific to alpha‐tubulin. J. Mol. Biol. 259, 325–330. Felgner, H., Frank, R., Biernat, J., Mandelkow, E.‐M., Mandelkow, E., Ludin, B., Matus, A., and Schliwa, M. (1997). Domains of neuronal microtubule‐associated proteins and flexural rigidity of microtubules. J. Cell Biol. 138, 1067–1075. Fodde, R., Kuipers, J., Rosenberg, C., Smits, R., Kielman, M., Gaspar, C., van Es, J. H., Breukel, C., Wiegant, J., Giles, R. H., and Clevers, H. (2001). Mutations in the APC tumour suppressor gene cause chromosomal instability. Nat. Cell Biol. 3, 433–438. Fukushige, T., Siddiqui, Z. K., Chou, M., Culotti, J. G., Gogonea, C. B., Siddiqui, S. S., and Hamelin, M. (1999). MEC–12, an alpha‐tubulin required for touch sensitivity in C. elegans. J. Cell Sci. 112, 395–403. Gaglio, T., Saredi, A., and Compton, D. A. (1995). NuMA is required for the organization of microtubules into aster‐like mitotic arrays. J. Cell Biol. 131, 693–708. Galjart, N., and Perez, F. (2003). A plus‐end raft to control microtubule dynamics and function. Curr. Opin. Cell Biol. 15, 48–53. Garrett, S., Auer, K., Compton, D. A., and Kapoor, T. M. (2002). hTPX2 is required for normal spindle morphology and centrosome integrity during vertebrate cell division. Curr. Biol. 12, 2055–2059.

MICROTUBULES AND MAPS

293

Gergely, F., Karlsson, C., Still, I., Cowell, J., Kilmartin, J., and Raff, J. W. (2000). The TACC domain identifies a family of centrosomal proteins that can interact with microtubules. Proc. Natl. Acad. Sci. USA 97, 14352–14357. Giannakakou, P., Gussio, R., Nogales, E., Downing, K. H., Zaharevitz, D., Bollbuck, B., Poy, G., Sackett, D., Nicolaou, K. C., and Fojo, T. (2000). A common pharmacophore for epothilone and taxanes: molecular basis for drug resistance conferred by tubulin mutations in human cancer cells. Proc. Natl. Acad. Sci. USA 97, 2904–2909. Gigant, B., Curmi, P. A., Martin‐Barbey, C., Charbaut, E., Lachkar, S., Lebeau, L., Siavoshian, S., Sobel, A., and Knossow, M. (2000). The 4 A˚ X‐ray structure of a tubulin: Stathmin‐like domain complex. Cell 102, 809–816. Goedert, M., Baur, C. P., Ahringer, J., Jakes, R., Hasegawa, M., Spillantini, M. G., Smith, M. J., and Hill, F. (1996). PTL–1, a microtubule‐associated protein with tau‐like repeats from the nematode Caenorhabditis elegans. J. Cell Sci. 109, 2661–2672. Goedert, M., Jakes, R., Spillantini, M. G., and Crowther, R. A. (1994). Tau protein and Alzheimer’s disease. In ‘‘Microtubules’’ (J. Hyams and C. Lloyd, Eds.), Wiley, New York. Groves, M. R., Hanlon, N., Turowski, P., Hemmings, B. A., and Barford, D. (1999). The structure of the protein phosphatase 2A PR65/A subunit reveals the conformation of its 15 tandemly repeated HEAT motifs. Cell 8, 99–110. Hamel, E., Sackett, D. L., Vourloumis, D., and Nicolaou, K. C. (1999). The coral‐derived natural products eleutherobin and sarcodictyins A and B: Effects on the assembly of purified tubulin with and without microtubule‐associated proteins and binding at the polymer taxoid site. Biochemistry 38, 3490–3498. Hamill, D. R., Howell, B., Cassimeris, L., and Suprenant, K. A. (1998). Purification of a WD repeat protein, EMAP, that promotes microtubule dynamics through an inhibition of rescue. J. Biol. Chem. 273, 9285–9291. Hayashi, I., and Ikura, M. (2003). Crystal structure of the amino‐terminal microtubule‐ binding domain of end‐binding protein 1 (EB1). J. Biol. Chem. 278, 36430–36434. Heidary, G., and Fortini, M. E. (2001). Identification and characterization of the Drosophila tau homolog. Mech. Dev. 108, 171–178. Hirokawa, N. (1998). Kinesin and dynein superfamily proteins and the mechanism of organelle transport. Science 279, 519–526. Hirose, K., Fan, J., and Amos, L. A. (1995a). Re‐examination of the polarity of microtubules and sheets decorated with kinesin motor domain. J. Mol. Biol. 251, 329–333. Hirose, K., Lockhart, A., Cross, R. A., and Amos, L. A. (1995b). Nucleotide‐dependent angular change in kinesin motor domain bound to tubulin. Nature 376, 277–279. Hirose, K., Lo¨ we, J., Alonso, M., Cross, R. A., and Amos, L. A. (1999). 3D electron microscopy of the interaction of kinesin with tubulin. Cell Struct. Func. 24, 277–284. Holleran, E. A., Karki, S., and Holzbaur, E. L. F. (1998). The role of the dynactin complex in intracellular motility. Int. Rev. Cytol. 182, 69–109. Homma, N., Takei, Y., Tanaka, Y., Nakata, T., Terada, S., Kikkawa, M., Noda, Y., and Hirokawa, N. (2003). Kinesin superfamily protein 2A (KIF2A) functions in suppression of collateral branch extension. Cell 114, 229–239. Hunter, A. W., and Wordeman, L. (2000). How motor proteins influence microtubule polymerization dynamics. J. Cell Sci. 113, 4379–4389. Hyman, A. A., Chre´ tien, D., Arnal, I., and Wade, R. H. (1995). Structural changes accompanying GTP hydrolysis in microtubules: Information from a slowly hydrolyzable analogue guanylyl‐(a,b)‐methylene‐diphosphonate. J. Cell Biol. 128, 117–125.

294

AMOS AND SCHLIEPER

Hyman, A. A., Salser, S., Drechsel, D. N., Unwin, N., and Mitchison, T. J. (1992). Role of GTP hydrolysis in microtubule dynamics: Information from a slowly hydrolyzable analogue, GMPCPP. Mol. Biol. Cell 3, 1155–1167. Incla´ n, Y. F., and Nogales, E. (2001). Structural models for the self‐assembly and microtubule interactions of g‐, d‐ and E‐tubulin. J. Cell Sci. 114, 413–422. Infante, C., Ramos‐Morales, F., Fedriani, C., Bornens, M., and Rios, R. M. (1999). GMAP–210, a cis‐Golgi network‐associated protein, is a minus end microtubule‐ binding protein. J. Cell Biol. 145, 83–98. Inoue, Y. H., Savoian, R. M., Suzuki, T., Mathe, E., Yamamoto, M. T., and Glover, D. M. (2004). Mutations in orbit/mast reveal that the central spindle is comprised of two microtubule populations, those that initiate cleavage and those that propagate furrow ingression. J. Cell Biol. 166, 49–60. Job, D., Valiron, O., and Oakley, B. (2003). Microtubule nucleation. Curr. Opin. Cell Biol. 15, 111–117. Kar, S., Fan, J., Smith, M. J., Goedert, M., and Amos, L. A. (2003a). Repeat motifs of tau bind to the insides of microtubules in the absence of taxol. EMBO J. 22, 1–8. Kar, S., Florence, G. J., Paterson, I., and Amos, L. A. (2003b). Discodermolide interferes with the binding of tau protein to microtubules. FEBS Lett. 539, 34–36. Kim, M. H., Cierpicki, T., Derewenda, U., Krowarsch, D., Feng, Y., Devedjiev, Y., Dauter, Z., Walsh, C. A., Otlewski, J., Bushweller, J. H., and Derewenda, Z. S. (2003). The DCX‐domain tandems of doublecortin and doublecortin‐like kinase. Nat. Struct. Biol. 10, 324–333. King, S. J., Brown, C. L., Maier, K. C., Quintyne, N. J., and Schroer, T. A. (2003). Analysis of the dynein‐dynactin interaction in vitro and in vivo. Mol. Biol. Cell 14, 5089–5097. Kinoshita, K., Habermann, B., and Hyman, A. A. (2002). XMAP215: A key component of the dynamic microtubule cytoskeleton. Trends Cell Biol. 12, 267–273. Kline‐Smith, S. L., and Walczak, C. E. (2004). Mitotic spindle assembly and chromosome segregation: Refocusing on microtubule dynamics. Mol. Cell. 15, 317–327. Kondoh, M., Usui, T., Nishikiori, T., Mayumi, T., and Osada, H. (1999). Apoptosis induction via microtubule disassembly by an antitumour compound, pironetin. Biochem. J. 340, 411–416. Kraulis, P. J. (1991). Molscript: A program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24, 946–950. Kurasawa, Y., Earnshaw, W. C., Mochizuki, Y., Dohmae, N., and Todokoro, K. (2004). Essential roles of KIF4 and its binding partner PRC1 in organized central spindle midzone formation. EMBO J. 23, 3237–3248. Lee, H., Engel, U., Rusch, J., Scherrer, S., Sheard, K., and Van Vactor, D. (2004). The microtubule plus end tracking protein Orbit/MAST/CLASP acts downstream of the tyrosine kinase Abl in mediating axon guidance. Neuron 42, 913–926. Lee, M. J., Gergely, F., Jeffers, K., Peak‐Chew, S. Y., and Raff, J. W. (2001). Msps/ XMAP215 interacts with the centrosomal protein D‐TACC to regulate microtubule behaviour. Nat. Cell Biol. 3, 643–649. Leguy, R., Melki, R., Pantaloni, D., and Carlier, M. F. (2000). Monomeric g‐tubulin nucleates microtubules. J. Biol. Chem. 275, 21975–21980. Li, H., De Rosier, D., Nicholson, W., Nogales, E., and Downing, K. (2002). Microtubule structure at 8A˚ resolution. Structure 10, 1317–1328. Li, S., Finley, J., Liu, Z. J., Qiu, S. H., Chen, H., Luan, C. H., Carson, M., Tsao, J., Johnson, D., Lin, G., Zhao, J., Thomas, W., Nagy, L. A., Sha, B., DeLucas, L. J., Wang, B. C., and Luo, M. (2002). Crystal structure of the cytoskeleton‐associated protein glycine‐ rich (CAP‐Gly) domain. J. Biol. Chem. 277, 48596–48601.

MICROTUBULES AND MAPS

295

Llanos, R., Chevrier, V., Ronjat, M., Meurer‐Grob, P., Martinez, P., Frank, R., Bornens, M., Wade, R. H., Wehland, J., and Job, D. (1999). Tubulin binding sites on g‐tubulin: Identification and molecular characterization. Biochemistry 38, 15712–15720. Long, B. H., Carboni, J. M., Wasserman, A. J., Cornell, L. A., Casazza, A. M., Jensen, P. R., Lindel, T., Fenical, W., and Fairchild, C. R. (1998). Eleutherobin, a novel cytotoxic agent that induces tubulin polymerization, is similar to paclitaxel (Taxol). Cancer Res. 58, 1111–1115. Lo¨ we, J., and Amos, L. A. (1998). Crystal structure of the bacterial cell‐division protein FtsZ. Nature 391, 203–206. Lo¨ we, J., Li, H., Downing, K. H., and Nogales, E. (2001). Refined structure of ab tubulin at 3.5 A˚ resolution. J. Mol. Biol. 313, 1045–1057. Lu, C., Srayko, M., and Mains, P. E. (2004). The Caenorhabditis elegans microtubule‐ severing complex MEI–1/MEI–2 katanin interacts differently with two superficially redundant b‐tubulin isotypes. Mol. Biol. Cell 15, 142–150. Makrides, V., Massie, M. R., Feinstein, S. C., and Lew, J. (2004). Evidence for two distinct binding sites for tau on microtubules. Proc. Natl. Acad. Sci. USA 101, 6746–6751. Mandelkow, E., Schultheiss, R., and Mandelkow, E.‐M. (1984). Assembly and 3‐dimensional image reconstruction of tubulin hoops. J. Mol. Biol. 177, 507–529. McNally, K. P., Buster, D. W., and McNally, F. J. (2002). Katanin‐mediated microtubule severing can be regulated by multiple mechanisms. Cell Motil. Cytoskel. 53, 337–349. Melki, R., Fievez, S., and Carlier, M.‐F. (1996). Continuous monitoring of Pi release following nucleotide hydrolysis in actin or tubulin assembly using 2‐amino‐6‐mercapto‐7‐methylpurine ribonucleoside and purine‐nucleoside phosphorylase as an enzyme‐linked assay. Biochemistry 35, 12038–12045. Merdes, A., Ramyar, K., Vechio, J. D., and Cleveland, D.W (1996). A complex of NuMA and cytoplasmic dynein is essential for mitotic spindle assembly. Cell 87, 447–458. Merdes, A., Heald, R. Samejima,K., Earnshaw, W. C., and Cleveland, D. W. (2000). Formation of spindle poles by dynein/dynactin‐dependent transport of NuMA. J. Cell Biol. 149, 851–861. Meurer‐Grob, P., Kasparian, J., and Wade, R. H. (2001). Microtubule structure at improved resolution. Biochemistry 40, 8000–8008. Mishima, M., Pavicic, V., Gruneberg, U., Nigg, E. A., and Glotzer, M. (2004). Cell cycle regulation of central spindle assembly. Nature 430, 908–913. Mitchison, T. J. (1993). Localization of an exchangeable GTP‐binding site at the plus end of microtubules. Science 261, 1044–1047. Miyamoto, D. T., Perlman, Z. E., Mitchison, T. J., and Shirasu‐Hiza, M. (2003). Dynamics of the mitotic spindle‐potential therapeutic targets. Prog. Cell Cycle Res. 5, 349–360. Moores, C. A., Perderiset, M., Francis, F., Chelly, J., Houdusse, A., and Milligan, R. A. (2004). Mechanism of microtubule stabilization by doublecortin. Mol. Cell 14, 833–839. Moritz, M., and Agard, D. A. (2001). g‐Tubulin complexes and microtubule nucleation. Curr. Op. Struct. Biol. 11, 174–181. Mukherjee, A., Saez, C., and Lutkenhaus, J. (2001). Assembly of an FtsZ mutant deficient in GTPase activity has implications for FtsZ assembly and the role of the Z ring in cell division. J. Bacteriol. 183, 190–197. Mu¨ ller, S., Smertenko, A., Wagner, V., Heinrich, M., Hussey, P. J., and Hauser, M. T. (2004). The plant microtubule‐associated protein AtMAP65–3/PLE is essential for cytokinetic phragmoplast function. Curr. Biol. 14, 412–417.

296

AMOS AND SCHLIEPER

Nettles, J. H., Li, H., Cornett, B., Krahn, J. M., Snyder, J. P., and Downing, K. H. (2004). The binding mode of epothilone A on b‐tubulin by electron crystallography. Science 305, 866–869. Nogales, E., Downing, K. H., Amos, L. A., and Lo¨ we, J. (1998a). Tubulin and FtsZ form a distinct family of GTPases. Nat. Struct. Biol. 5, 451–458. Nogales, E., Wang, H.‐W., and Niederstrasser, H. (2003). Tubulin rings: Which way do they curve? Curr. Op. Struct. Biol. 13, 256–261. Nogales, E., Whittaker, M., Milligan, R. A., and Downing, K. H. (1999). High resolution model of the microtubule. Cell 96, 79–88. Nogales, E., Wolf, S., and Downing, K. H. (1998b). Structure of the ab‐tubulin dimer by electron crystallography. Nature 391, 199–203. Ogawa, T., Nitta, R., Okada, Y., and Hirokawa, N. (2004). A common mechanism for microtubule destabilizers‐M type kinesins stabilize curling of the protofilament using the class‐specific neck and loops. Cell, 116, 591–602. Ohkura, H., Garcia, M. A., and Toda, T. (2001). Dis1/TOG universal microtubule adaptors—one MAP for all? J. Cell Sci. 114, 3805–3812. Oliva, M. A., Cordell, S. C., and Lo¨ we, J. (2004). Structural insights into FtsZ protofilament formation. Nat. Struct. Mol. Biol. 11, 1243–1250. Pedrotti, B., and Islam, K. (1996). Dephosphorylated but not phosphorylated microtubule associated protein MAP1B binds to microfilaments. FEBS Lett. 388, 131–133. Perez, F., Diamantopoulos, G. S., Stalder, R., and Kreis, T. E. (1999). CLIP–170 highlights growing microtubule ends in vivo. Cell 96, 517–527. Pernet‐Gallay, K., Antony, C., Johannes, L., Bornens, M., Goud, B., and Rios, R. M. (2002). The overexpression of GMAP–210 blocks anterograde and retrograde transport between the ER and the Golgi apparatus. Traffic 3, 822–832. Popov, A. V., Severin, F., and Karsenti, E. (2002). XMAP215 is required for the microtubule‐nucleating activity of centrosomes. Curr. Biol. 12, 1326–1330. Quarmby, L. (2000). Cellular samurai: Katanin and the severing of microtubules. J. Cell Sci. 113, 2821–2827. Rai, S. S., and Wolff, J. (1996). Localization of the vinblastine‐binding site on b‐tubulin. J. Biol. Chem. 271, 14707–14711. Ravelli, R. B. G., Gigant, B., Curmi, P. A., Jourdain, I., Lachkar, S., Sobel, A., and Knossow, M. (2004). Insight into tubulin regulation from a complex with colchicine and a stathmin‐like domain. Nature 428, 198–202. Rehberg, M., and Graf, R. (2002). Dictyostelium EB1 is a genuine centrosomal component required for proper spindle formation. Mol. Biol. Cell 13, 2301–2310. Rodionov, V., Nadezhdina, E., and Borisy, G. (1999). Centrosomal control of microtubule dynamics. Proc. Natl. Acad. Sci. USA 96, 115–120. Sakato, M., and King, S. M. (2004). Design and regulation of the AAAþ microtubule motor dynein. J. Struct. Biol. 146, 58–71. Savage, C., Hamelin, M., Culotti, J. G., Coulson, A., Albertson, D. G., and Chalfie, M. (1989). Mec–7 is a b‐tubulin gene required for the production of 15‐protofilament microtubules in Caenorhabditis elegans. Genes Dev. 3, 870–881. Schaap, I. A. T., de Pablo, P. J., and Schmidt, C. F. (2004). Resolving the molecular structure of microtubules under physiological conditions with scanning force microscopy. Eur. Biophys. J. 33, 462–467. Schafer, D. A., Gill, S. R., Cooper, J. A., Heuser, J. E., and Schroer, T. A. (1994). Ultrastructural analysis of the dynactin complex: An actin‐related protein is a component of a filament that resembles F‐actin. J. Cell Biol. 126, 403–412.

MICROTUBULES AND MAPS

297

Scheffers, D. J., de Wit, J. G., den Blaauwen, T., and Driessen, A. J. (2002). GTP hydrolysis of cell division protein FtsZ: Evidence that the active site is formed by the association of monomers. Biochemistry 41, 521–529. Schroer, T. A. (1996). Structure and function of dynactin. Semin. Cell Dev. Biol. 7, 321–328. Schroer, T. A. (2001). Microtubules don and doff their caps: Dynamic attachments at plus and minus ends. Curr. Opin. Cell Biol. 13, 92–96. Schuyler, S. C., Liu, J. Y., and Pellman, D. (2003). The molecular function of Ase1p: Evidence for a MAP‐dependent midzone‐specific spindle matrix. J. Cell Biol. 160, 517–528. Schuyler, S. C., and Pellman, D. (2001). Microtubule ‘‘plus‐end‐tracking proteins’’: The end is just the beginning. Cell 105, 421–424. Shahani, N., Fath, T., and Brandt, R. (2001). MINUS: A negative regulator of microtubule nucleation. Mol. Biol. Cell 12, 937. Shipley, K., Hekmat‐Nejad, M., Turner, J., Moores, C., Anderson, R., Milligan, R., Sakowicz, R., and Fletterick, R. (2004). Structure of a kinesin microtubule depolymerization machine. EMBO J. 23, 1422–1432. Shirasu‐Hiza, M., Coughlin, P., and Mitchison, T. (2003). Identification of XMAP215 as a microtubule‐destabilizing factor in Xenopus egg extract by biochemical purification. J. Cell Biol. 161, 349–358. Snyder, J. P., Nettles, J. H., Cornett, B., Downing, K. H., and Nogales, E. (2001). The binding conformation of Taxol in b‐tubulin: A model based on electron crystallographic density. Proc. Natl. Acad. Sci. USA 98, 5312–5316. Spittle, C., Charrasse, S., Larroque, C., and Cassimeris, L. (2000). The interaction of TOGp with microtubules and tubulin. J. Biol. Chem. 275, 20748–20753. Srayko, M., Buster, D. W., Bazirgan, O. A., McNally, F. J., and Mains, P. E. (2000). MEI-1/ MEI-2 katanin‐like microtubule severing activity is required for Caenorhabditis elegans meiosis. Gens. Dev. 14, 1072–1084. Stamer, K., Vogel, R., Thies, E., Mandelkow, E., and Mandelkow, E.‐M. (2002). Tau blocks traffic of organelles, neurofilaments and APP vesicles in neurons and enhances oxidative stress. J. Cell Biol. 156, 1051–1063. Takei, Y., Teng, J., Harada, A., and Hirokawa, N. (2000). Defects in axonal elongation and neuronal migration in mice with disrupted tau and map1b genes. J. Cell Biol. 150, 989–1000. Tanaka, T., Serneo, F. F., Higgins, C., Gambello, M. J., Wynshaw‐Boris, A., and Gleeson, J. G. (2004). Lis1 and doublecortin function with dynein to mediate coupling of the nucleus to the centrosome in neuronal migration. J. Cell Biol. 165, 709–721. ter Haar, E., Kowalski, R. J., Hamel, E., Lin, C. M., Longley, R. E., Gunasekera, S. P., Rosenkranz, H. S., and Day, B. W. (1996). Discodermolide, a cytotoxic marine agent that stabilizes microtubules more potently than Taxol. Biochemistry 35, 243–250. Usui, T., Watanabe, H., Nakayama, H., Tada, Y., Kanoh, N., Kondoh, M., Asao, T., Takio, K., Watanabe, H., Nishikawa, K., Kitahara, T., and Osada, H. (2004). The anti‐cancer natural product pironetin selectively targets Lys352 of a‐tubulin. Chem. Biol. 11, 799–806. Vaillant, A. R., Mu¨ ller, R., Langkopf, A., and Brown, D. L. (1998). Characterization of the microtubule‐binding domain of microtubule‐associated protein 1A and its effects on microtubule dynamics. J. Biol. Chem. 273, 13973–13981. Vale, R. D. (2000). AAA proteins: Lords of the ring. J. Cell Biol. 150, F13–F19. Vale, R. D. (2003). The molecular motor toolbox for intracellular transport. Cell 112, 467–480.

298

AMOS AND SCHLIEPER

Valetti, C., Wetzel, D. M., Schrader, M., Hasbani, M. J., Gill, S. R., Kreis, T. E., and Schroer, T. A. (1999). Role of dynactin in endocytic traffic: Effects of dynamitin overexpression and colocalization with CLIP–170. Mol. Biol. Cell 10, 4107–4120. van Breugel, M., Drechsel, D., and Hyman, A. (2003). Stu2p, the budding yeast member of the conserved Dis1/XMAP215 family of microtubule‐associated proteins is a plus end‐binding microtubule destabilizer. J. Cell Biol. 161, 359–369. Vaughan, K. T. (2004). Surfing, regulating and capturing: Are all microtubule‐tip‐ tracking proteins created equal? Trends Cell Biol. 14, 491–496. Vaughan, K. T., and Vallee, R. B. (1995). Cytoplasmic dynein binds dynactin through a direct interaction between the intermediate chains and p150Glued. J. Cell Biol. 131, 1507–1516. Vaughan, P. S., Miura, P., Henderson, M., Byrne, B., and Vaughan, K. T. (2002). A role for regulated binding of p150Glued to microtubule plus ends in organelle transport. J. Cell Biol. 158, 305–319. Venier, P., Maggs, A. C., Carlier, M. F., and Pantaloni, D. (1994). Analysis of microtubule rigidity using hydrodynamic flow and thermal fluctuations. J. Biol. Chem. 269, 13353–13360. Wade, R. H., Chre´ tien, D., and Job, D. (1990). Characterization of microtubule protofilament numbers: How does the surface lattice accommodate? J. Mol. Biol. 212, 775–786. Walczak, C. E., Mitchison, T. J., and Desai, A. (1996). XKCM1: A Xenopus kinesin‐related protein that regulates microtubule dynamics during mitotic spindle assembly. Cell 84, 37–47. Walenta, J. H., Didier, A. J., Liu, X., and Kramer, H. (2001). The Golgi‐associated hook3 protein is a member of a novel family of microtubule‐binding proteins. J. Cell Biol. 152, 923–934. Walker, R., Pryer, N., and Salmon, E. (1991). Dilution of individual microtubules observed in real time in vitro: Evidence that cap size is small and independent of elongation rate. J. Cell Biol. 114, 73–81. Watts, N. R., Cheng, N., West, W., Steven, A. C., and Sackett, D. L. (2002). The cryptophycin‐tubulin ring structure indicates two points of curvature in the tubulin dimer. Biochemistry 41, 12662–12669. Wittmann, T., Wilm, M., Karsenti, E., and Vernos, I. (2000). TPX2, a novel Xenopus MAP involved in spindle pole organization. J. Cell Biol. 149, 1405–1418.

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS ¨ LLER, AND E. MANDELKOW By A. MARX, J. MU Max‐Planck‐Unit for Structural Molecular Biology Notkestrasse 85, 22607 Hamburg, Germany

I. Kinesin Classes, Domain Structure, and Nomenclature . . . . . . . . . . . . . . . . . . . II. Kinesin‐1 as Prototypical Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Comparison of Kinesin Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Human and Rat Kinesin‐1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Fungal Kinesin‐1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Other N‐Type Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Kinesin‐14 (C‐Type Motors) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Kinesin‐13 (M‐Type Motors) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Conformational Switching in Kinesin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Comparison with Myosin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Nucleotide Binding, Switch‐I and ‐II, and Conformational Relays. . . . . V. Structures of Kinesin‐Related Domains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Dynein Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

299 301 310 310 313 314 320 326 328 328 332 333 334 338 340

Abstract Microtubules are the intracellular tracks for two classes of motor proteins: kinesins and dyneins. During the past few years, the motor domain structures of several kinesins from different organisms have been determined by X‐ray crystallography. Compared with kinesins, dyneins are much larger proteins and attempts to crystallize them have failed so far. Structural information about these proteins comes mostly from electron microscopy. In this chapter, we mainly focus on the crystal structures of kinesin motor domains.

I.

Kinesin Classes, Domain Structure, and Nomenclature

Kinesins constitute a large protein family that realizes a wide range of functions within eukaryotic cells, including the transport of different cargoes (vesicles, organelles, protein complexes, chromosomes) and the regulation of microtubule dynamics. The superfamily of kinesins currently includes more than 600 sequences from a variety of species. This large number of proteins led to a confusing variety of names and classifications. To overcome these problems, the kinesin research community just ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71008-6

299

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

300

MARX ET AL.

proposed a new standardized nomenclature (Lawrence et al., 2004) that subdivides the kinesin superfamily into 14 families. Each family bears the name ‘‘kinesin’’ and is followed by an Arabic number (e.g., the founding member of the protein superfamily, the conventional kinesin or kinesin heavy chain, KHC, is now named as kinesin‐1). Kinesin‐1 comprises three major domains: the N‐terminal motor domain that can be subdivided into the core motor domain and the adjacent neck linker and neck region, the central stalk domain, and the C‐terminal tail or light chain‐binding domain (Fig. 1A). The core motor domain has a length of about 325 amino acids and contains both the microtubule and the nucleotide binding elements. In different kinesin families, this motor

Fig. 1. Domain structures of typical members of the kinesin superfamily. (A) Bar diagram of the kinesin heavy chain (KHC) of conventional kinesin (kinesin‐1 family) as a typical representative of N‐type motors (motor domain at the N‐terminus, red); the cartoon model beneath the bar diagram shows the tetrameric complex of two heavy and two light chains. (B) M‐type kinesin like MCAK of the kinesin‐13 family. (C) C‐type kinesin like Ncd of the kinesin‐14 family.

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

301

domain can be located at various places within the molecule, depending on the function of the specific kinesin family as a plus‐end (N‐terminal location) or minus‐end (C‐terminal location) directed motor or as microtubule depolymerizing machines (internal location). The different possibilities for the location of the motor domain (correlating with different functions; Fig. 1) led to the classification into N‐type, M‐type, and C‐type motors (Vale and Fletterick, 1997). M‐type motors were also referred to as Kin‐I (for ‘‘internal’’). The motor core domain produces force in concert with the adjacent neck; this region is found on either the N‐ (in N‐type kinesins) or the C‐terminus (in C‐ and M‐type kinesins). Within cells, most kinesins are not monomeric; they carry out their functions at least pairwise. For example, kinesin‐1 forms dimers through a coiled‐coil interaction of the stalk domain. The ‘‘kinesin heavy chains’’ bind to ‘‘light chains,’’ thus forming a tetrameric complex. The light chains can dock onto cargo receptors or adaptors, linking kinesin to its different cargoes. Other kinesins fulfill their functions as homodimers (without light chains, e.g., Neurospora kinesin‐1 or kinesin‐7), heterotrimers (kinesin‐2, where two different motor molecules are associated with a nonmotor subunit), or homotetramers (kinesin‐5). The first structure of a kinesin motor domain was published in 1996. After 8 years of kinesin crystallization and structure determination, there are today 26 X‐ray structures of motor domains from kinesins and mutants of different sources deposited in the Protein Data Bank (PDB; October 2004; http://www.rcsb.org/pdb/; Berman et al., 2000). These structures are from 11 different proteins which belong to six different kinesin families including kinesin‐1 (conventional kinesin), kinesin‐3 (Unc104/Kif1a‐family), kinesin‐5 (bimC family), kinesin‐7 (CENP‐E family), kinesin‐13 (MCAK family), and kinesin‐14 (C‐terminal motors). The crystallized proteins come from eight different organisms including mammals, Drosophila, yeast, Plasmodium, Neurospora, and a plant (potato). Two of the proteins, a mouse kinesin‐3 (Kif1a) and mouse kinesin‐13 (Kif2C), have been solved in different nucleotide states. One structure (human Eg5) has been solved together with a class‐specific inhibitor, and another structure (KCBP) contains a short nonmotor domain. Referring to the position of the motor domain, members of all three groups (N‐type, M‐type, and C‐type) have been crystallized.

II.

Kinesin‐1 as Prototypical Motor

The first structure of a kinesin motor domain—that of human kinesin‐1 (formerly named ‘‘KHC’’ or ‘‘conventional kinesin’’)—was determined by Kull and coworkers (1996). This is still the structure of highest resolution (1.8 A˚ ; PDB code 1BG2) among all structures of conventional kinesins

302

MARX ET AL.

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

303

about which literature has been published so far. Even so, it contains several regions of unstructured and therefore invisible amino acids. Shortly after that, Sack et al. (1997) published an article on the structure of a similar construct from rat kinesin‐1. Although the resolution is lower (2.0 A˚ ; PDB code 2KIN) than that of human kinesin‐1, the rat kinesin‐1 structure includes some of the structural elements that are invisible in the Kull structure. More recently, another crystal structure of the human kinesin‐1 motor domain was obtained, using crystallization conditions similar to those used for the crystallization of monomeric rat kinesin‐1 (Sindelar et al., 2002). This structure (PDB code 1MKJ; 2.7‐A˚ resolution) is very similar to that of rat kinesin‐1, demonstrating that the differences between the human and the rat structures are largely due to the crystallization conditions. Nevertheless, these differences point to ‘‘hot spots’’ of conformational variability that might be significant for the dynamic behavior of the motor domain. The more complete structure of rat kinesin‐1 is chosen as a paradigm for the discussion of similarities and differences to other kinesin structures (Fig. 2). The core domain is an a/b class protein with a three‐layer (aba) sandwich architecture. The central b‐sheet consists of eight strands, named b1 to b8 according to their succession in amino acid sequence (spatial sequence: 2‐1‐8‐3‐7‐6‐4‐5; Fig. 2B). These strands are all parallel with the exception of b5 and b6. Strands b6 and b7 form a b‐hairpin connected by L10 (a short loop containing a b‐turn of type I). b5 consists of two short stretches of three and four amino acids with a 26–amino acid insert (loop L8) in between. The b5 strands serve as hinges that anchor loop L8 to the core b‐sheet. b4 is the longest b‐strand of the motor domain (14 amino acids), the length of the other strands falling off toward the other edge of the b‐sheet (b2 with 3 amino acids is the shortest strand of the core sheet). As the strands at one end are more or less lined up, the

Fig. 2. Structure of the motor domain of rat kinesin‐1. All panels display ribbon diagrams of the crystal structure of monomeric RnKHC (PDB code 2KIN; Sack et al., 1997) in two standard orientations that differ by a rotation of 180 degrees about the vertical axis within the drawing plane. (A) Overview with all secondary structure elements shown in intensive colors (blue: b‐strands, red: helices). (B, C, D, E) The same with selected structural elements highlighted and labeled (B, central b‐sheet; C, helices surrounding the central b‐sheet; D, extra lobes anchored at the left and right edge of the core structure; E: neck linker and neck helix). The orientation at the right side presents the contact surface with the structural elements that interact with tubulin subunits when the motor is attached to the microtubule (proximal surface). The view shown on the left side is toward the outer (distal) surface. In the orientations selected, the microtubule runs roughly parallel to the drawing plane, plus end down. The figure has been prepared using Deep View Swiss‐PDB Viewer (Guex and Peitsch, 1997).

304

MARX ET AL.

central b‐sheet assumes roughly the shape of a right triangle (disregarding the b5‐appendix), which is distorted in space. The central b‐sheet is the supporting structure for all the other structural elements and moving parts of the motor domain. Most of the central b‐sheet is covered with helices and loops: a1, a2a‐L5‐a2b, a3‐L9‐a3a on one face, a4‐L12‐a5 and a6 on the opposite face (Fig. 2C). The overall shape of the core domain is conelike with the exposed N‐terminus of b4 and the loop between b6 and b7 forming the tip of the cone. In addition to the aba‐ sandwich, there are two lobes (Fig. 2D); one lobe of about 30 amino acids at the b2‐vertex of the central b‐sheet, consisting of a short helix (a0) and a small, three‐stranded, antiparallel b‐sheet (b1a,b,c), and the second lobe consisting of loop L8 (already mentioned) attached to b4 of the central b‐sheet via the split strand b5. The latter is sometimes called the ‘‘b5‐L8 lobe.’’ Besides loop regions of undefined secondary structure, this lobe contains two b‐strands (b5a and b5b) in antiparallel conformation. The monomeric rat kinesin construct (amino acids 1–354) comprises the head domain (including the core motor domain, amino acids 2–325, and the neck linker, amino acids 326–338) and the first half of the neck domain. In the crystal structure, the neck linker consists of two strands, b9 and b10, that form hydrogen bonds with strands b8 and b7 of the core b‐sheet (Fig. 2E). The neck linker ends close to loop L10 at the tip of the core domain where the (a‐helical neck domain (helix a7) is attached to the core motor domain. Its orientation is roughly in the ‘‘plane’’ of the core b‐sheet and perpendicular to the strands. The nucleotide binding pocket consists of four motifs, N1 to N4, also found in other ATP‐binding proteins, such as myosins and G‐proteins (Sack et al., 1997). N1 is a Walker A motif at the end of b3 (86GQTSSGKT93, rat kinesin‐1 sequence, consensus motif underlined) that forms a ‘‘P‐loop,’’ a common fold of b‐loop‐helix type that binds oxygen atoms of the b‐ and g‐phosphates. N2 (199NEHSSR204, located in a3a at the N‐terminus of b6) and N3 (232DLAGSE237 at the C‐terminus of b7) are known as switch‐1 and switch‐2 in analogy to G‐proteins and myosins. The switches may function as g‐phosphate sensors that are thought to be first to respond to ATP‐ hydrolysis. By changing position and conformation of the switches, adjacent elements are forced to move and, thus, the local adjustment in nucleotide coordination will be transduced and amplified. N4 at the C‐terminus of b1 (14RFRP17) is involved in nucleotide binding by interaction with the adenine moiety. Histidine H94 at the end of the P‐loop also interacts with the base. Structural elements that interact with the microtubule surface have been identified by the effect of point mutations (Woehlke et al., 1997) and by fitting crystal structures of kinesin motor domains to low‐resolution electron density maps obtained by cryo‐electron microscopy of microtubules

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

305

saturated with kinesin (Hoenger et al., 1998, 2000; Kikkawa et al., 2000; Song et al., 2001). According to these experiments, the structural elements that contribute most to the interaction with microtubules are (1) the b5‐L8 lobe; (2) the cluster formed by helix a4, loop L12, and a5; and (3) loop L11, a loop approximately 15 amino acids long between switch2 and the a4‐ L12‐a5 cluster, which is largely disordered in rat kinesin‐1 and in most other kinesin structures known so far (Table I). Exceptions are the structures of Neurospora crassa kinesin‐1 (NcKin; PDB code 1GOJ; Song et al., 2001), the monomeric kinesin‐3 motor domain complex with adenosine diphosphate (ADP) and vanadate (MmKIF1a, PDB code 1VFZ, Nitta et al., 2004), the R598A mutant of ScKar3 protein, a member of the kinesin‐14 family (PDB code 1F9V; Yun et al., 2001), and the M‐type kinesin‐13 PfMCAK (PDB code 1RY6; Shipley et al., 2004). Loop L11 is supposed to be disordered in free kinesin and to adopt a rigid structure in the kinesin‐ microtubule complex, thus establishing a linkage between the g‐phosphate sensor switch‐2 and the microtubule‐binding cluster a4‐L12‐a5, which is also named the ‘‘switch‐2 cluster.’’ In addition to these elements, neck helix a7 may also contribute to microtubule binding via interaction with the negatively charged C‐terminus of b‐tubulin. The kinesin construct shown in Fig. 2 ends with K354 (last visible amino acid E351) in the middle of a sequence that is predicted by the program Paircoil (Berger et al., 1995) to form a continuous coiled‐coil (amino acids A334–R371). A longer construct comprising the complete coiled‐coil domain (amino acids 2–379) was crystallized in the form of dimers as shown in Fig. 3B (Kozielski et al., 1997). The structures of the two heads are very similar to the 2KIN structure of the monomeric construct. Surprisingly, there is no proper symmetry relation between the heads of the dimer, although the region responsible for dimerization (the neck coiled‐coil) has almost perfect twofold symmetry. The heads are related to each other by rotation of approximately 120 degrees about an axis that is inclined to the coiled‐coil axis. There is only a little direct contact between the heads via interaction of K160 in the microtubule‐binding loop L8 (between b5a and b5b) and the tip of head B (E221 in loop L10 between b6 and b7). It has been suggested on the basis of solution scattering experiments that the asymmetric conformation of the heads in the crystal structure may be representative for the conformation of the isolated dimer (Kozielski et al., 2001). Thus, the special arrangement of the heads in the crystal may not be an artifact caused by crystal packing, but rather the effect of direct and indirect interactions (between residues of the individual heads and residues at the N‐termini of the neck coiled‐coil). In any case, the asymmetric conformation must be of limited stability because both heads can separate during their working cycle and bind simultaneously to adjacent sites on the microtubule (e.g., Asenjo et al., 2003; Hoenger et al., 2000; Skiniotis et al.,

306

MARX ET AL.

Table I Structural Alignment of the Switch‐1 and Switch‐2 Regions of Kinesin Motor Domains with Secondary Structure Assignments and Classification of the Switch‐2 Cluster and Neck/Neck Linker Conformations

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

307

308

MARX ET AL.

2003). This is illustrated in Fig. 4, which shows schematically how a kinesin dimer could walk along a microtubule protofilament. A detailed view of the kinesin‐microtubule complex has been obtained by combining ‘‘high‐resolution’’ structures of the individual components from X‐ray crystallography (kinesin) and electron diffraction (tubulin; Lowe et al., 2001; Nogales et al., 1998) with ‘‘low‐resolution’’ models of kinesin‐ decorated microtubules obtained by cryoelectron microscopy and image reconstruction (Hirose et al., 1999; Hoenger et al., 2000; Kikkawa et al., 2001; Kozielski et al., 1998; Rice et al., 1999; Skiniotis et al., 2003; Wendt

a PDB code for the coordinates made available by the Research Collaboratory for Structural Bioinformatics Protein Data Bank (http://www.rcsb.org/pdb/; Berman et al., 2000). If the crystal contains several independent molecules per asymmetric unit, the corresponding chain codes are indicated in parentheses. b Limits for the linker between helix a3 and strand b6 are defined by the last residue of the core of helix a3 (see code assignments below) and the first residue of b6 that participates in a b‐ladder. In most cases, this is the conserved histidine corresponding to H206 in RnKHC. c Limits for the region marked as ‘‘loop 11’’ are defined by the last residue of strand b7 participating in a b‐ladder and the first residue of the core of helix a4 (see code assignments below). In most cases, the residue immediately preceding the ‘‘loop 11 region’’ is the leading aspartate of the switch‐2 motif (DLAGSE in RnKHC). d perm ¼ permissive conformation, compatible with docking of the neck linker or neck mimic (Vinogradowa et al., 2004); obstr ¼ obstructive conformation (prohibits docking). e For C‐type motors, additional information about the C‐terminal residues following helix a6 is given in parentheses. f Although the PDB file contains coordinates for all residues of loop L11, residues V238–I254 are not part of the model (Kull et al., 1996). g The neck linker is roughly perpendicular to a6 and parallel to the central b‐sheet. h Residues preceding strand b1 form a continuous a‐helix. i Construct starts with b1. j The PfKinI construct contains an engineering artifact of 30 amino acids C‐terminal to helix a6, which is disordered. k The neck helix is roughly perpendicular to the central b‐sheet. The construct ends with helix a6. l aa: amino acids. Code used for secondary structure assignments determined by PROCHECK v.3.5. (Laskowski et al., 1993) using the extended Kabsch/Sander classification (Kabsch and Sander, 1983): NNN (boldface) core region of a‐helix. NNN (boldface, italic) core region of 3/10 helix. NNN (normal face, thin underline) residues in core b- or allowed b-region of the Ramachandran plot. NNN (normal face, thick underline) residues participating in b-ladder or isolated b‐bridge.

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

309

Fig. 3. Conformation of the switch–2 cluster and neck linker/neck region in various members of the kinesin superfamily. The upper four panels (A, B, E, F) show crystal structures of N‐type kinesins with their motor domain at the N‐terminus and the neck at the C‐terminus. (C), (D), (G), and (H) show C‐ and M‐type kinesins with their neck N‐terminal to the motor domain, except for PoKCBP (G) where the C‐terminal neck mimic is shown instead of the N‐terminal neck (which is not included in the crystal structure). Each structure is shown in two orientations that differ by a rotation of 90 degrees. Rat conventional kinesin (RnKHC [A]) has been chosen to define standard orientations with the neck helix a7 parallel/perpendicular to the drawing area. Orientations for the other structures have been determined by least‐squares superposition of their P‐loop regions with that of RnKHC (using 11 Ca‐atoms of residues F83–T93 in RnKHC). (B), (C), and (D) show the structures of dimeric constructs with the second motor domain in pale colors. The Ncd structure in (C) is 180‐degree symmetric; the symmetry axis is oblique to the drawing plane and coincides with the axis of the coiled‐ coil that is formed by the two neck helices. In the asymmetric structure of the Ncd N600K mutant (D), the second motor domain (pale) is rotated by about 75 degrees around an axis perpendicular to the coiled‐coil. The structures shown in (A), (B), (F), and (G) have their switch‐2 cluster in ‘‘permissive’’ conformation, whereas the conformation of structures (C), (D), (E), and (H) is ‘‘obstructive,’’ as can be told by observing the slope of the extended switch‐2 helix a4. Color code: red, switch‐2 cluster including the extended

310

MARX ET AL.

et al., 2002). Figure 5 shows the binding geometry for the motor domain of N. crassa conventional kinesin (NcKin; Song et al., 2001) which is similar to that of human and rat kinesin‐1. Helix a4 is almost perpendicular to the protofilament and indents into the cleft between the a‐ and b‐subunits of the tubulin dimer. The major binding regions in kinesin (loop L8/b5a,b and the switch‐2 cluster including loop L11) are close to the H12 helices at the outer rim of the microtubule surface, with loop L8 and L12 approaching H12 of b‐tubulin, whereas loop L11 projects to the adjacent a‐subunit.

III.

Comparison of Kinesin Structures A.

Human and Rat Kinesin‐1

A comparison of the first human kinesin‐1 structure (PDB code 1BG2; Kull et al., 1996) and the rat kinesin‐1 structure (PDB code 2KIN) has been published previously (Sack et al., 1999). The overall fold of the motor

Fig. 4. Dimeric kinesin moving along a microtubule protofilament. Four ‘‘snapshots’’ taken from an animated cartoon that illustrates how a kinesin dimer could walk along a microtubule. The pictures shown here are all from one half‐cycle: one head (green/blue) is fixed while the other one (yellow/red) orbits around the common neck and stalk (stalk not shown). During the second half‐cycle, the motor domains change their roles. (The animated cartoon is available from http://www. mpasmb‐hamburg.mpg.de/.) helix a4; magenta, neck region and neck helix, if present; yellow, ADP; orange, AMPPNP; green, monastrol. The figure has been prepared using Deep View Swiss‐PDB Viewer (Guex and Peitsch, 1997).

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

311

Fig. 5. Binding of the motor domain of Neurospora conventional kinesin to the microtubule surface. The microtubule is represented by three tubulin subunits (ab‐a) of a single protofilament, plus end at the top. Tubulin subunits according to coordinates determined by electron crystallography (Nogales et al., 1998) and refined to 3.5 A˚ (PDB code 1JFF; Lowe et al., 2001) are colored in green (helices) and magenta (b‐strands), except for helices H11 and H12 (olive), the loops formed by residues 35–60 in both a‐tubulins (orange; adopted from PDB code 1TUB; Nogales et al., 1998; this loop is not part of the refined structure 1JFF), and the C‐terminal residues not determined by electron diffraction (red; arbitrary random conformation). Small molecule ligands are drawn as space‐filling models (taxol in red, nucleotides in yellow). The motor domain of NcKin (PDB code 1GOJ; Song et al., 2001) is colored mostly in pink (helices) and cyan (b‐strands), except for the major microtubule binding regions b5a,b/L8 and L11–a4– L12–a5. The b‐hairpin b6–L10–b7 at the tip of the motor domain points to the plus end. Helix a4 protrudes into the groove between a‐ and b‐subunits of a tubulin dimer. Loops L8 and L12 are close to the C‐terminal helix H12 of b‐tubulin, while loop L11 approaches helix H12 of the adjacent a‐tubulin. The figure has been prepared using Deep View Swiss‐PDB Viewer (Guex and Peitsch, 1997) and POVray for Windows (Persistence of Vision Pty. Ltd. 2004, Persistence of Vision Raytracer Version 3.5, retrieved from http://www.povray.org/).

domains is the same, yet the structures differ in significant shifts of surface elements. At that time it was not clear whether the differences that had been spotted were due to variances in the primary structures of human and rat kinesin‐1 (86.4% identity, 93.8% with conserved substitutions) or if the observed structures represent two possible conformations of the motor

312

MARX ET AL.

domain that are selectively stabilized by different conditions used for crystallization. The human structure was determined with crystals grown in polyethylene glycol (PEG) at pH 4.6, whereas rat kinesin was crystallized at pH 7.5 with lithium sulphate as precipitant. This uncertainty was resolved by a new structure of the human kinesin‐1 motor domain (PDB code 1MKJ; Sindelar et al., 2002). Using the same construct of the human kinesin‐1 motor domain (HsKHC, amino acids 2–349), crystals were grown under the conditions used for rat kinesin. The new structure of human kinesin‐1 turned out to be very similar to that of rat kinesin‐1. The root‐mean‐square (rms) distance of the Ca atoms after structural alignment is 0.78 A˚ (all amino acids included except for the N‐terminal alanine; calculated with DeepView Swiss‐Pdb Viewer 3.7, Guex and Peitsch, 1997). Thus, the differences between rat and human kinesin‐1 described earlier are not species specific, but reveal two conformational states of the motor domain that may be of physiological relevance. The most obvious difference between the PEG‐grown crystal structure of human kinesin‐1 and the crystal structures of human and rat kinesin‐1 obtained with lithium sulphate is that the neck linker and neck helix are disordered and thus invisible in the ‘‘PEG structure.’’ This correlates with a significant displacement of the switch‐2 cluster. In the ‘‘PEG structure,’’ the C‐termini of helices a4 and a5 in the switch‐2 cluster occlude binding sites for the neck linker at the surface of the core motor domain. This explains why the neck linker is ‘‘undocked’’ (i.e., mostly detached from the motor core) and disordered, therefore making it invisible in this crystal structure. In the ‘‘lithium sulphate structures,’’ the switch‐2 cluster assumes a different position and orientation, allowing the binding of the neck linker to the motor core. The displacement of the switch‐2 cluster can be described to a large extent by a rigid body movement. The rms distance of the Ca atoms in the switch‐2 clusters of the human kinesin‐1 structures 1MKJ and 1BG2 (K256‐G291, human kinesin‐1 numbering) is 3.25 A˚ , if the superposition is calculated with the core structures. The maximum displacement is 5.8 A˚ at amino acid Thr273 in loop L12, close to the neck linker binding site. Rigid body superposition of the clusters alone reduces the rms distance to 0.77 A˚ . The movement consists of an approximately 2‐A˚ translation toward the nucleotide binding site combined with a tilt away from the neck linker. How does the ‘‘rigid body’’ movement of the switch‐2 cluster comply with its interactions with the rest of the motor domain? Regarding the main chain connectivity, such a movement requires flexible adaptors at both ends of the cluster. At the N‐terminus, loop L11 obviously fulfills this function. At the other end, loop L13 (between a5 and strand b8 of the central b‐sheet) may provide sufficient flexibility. Loop L13 contains two

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

313

glycines (G293G294 in rat kinesin‐1; glycine is less restricted in conformational space than any other amino acid) that are conserved throughout the kinesin family with the major exception of the kinesin‐13 (formerly MCAK/Kif4) family, which has only one glycine in loop L13 (according to an alignment of 143 motor domains available on the Kinesin Home Page, http://www.proweb.org/kinesin//KinesinAlign.html). Interestingly, the 2KIN structure of rat kinesin‐1 contains a point mutation (G293D) of one of these glycines. This may reduce the flexibility of loop L13 and favor the conformation of the switch‐2 cluster that allows docking of the neck linker. This could explain why the rat kinesin‐1 mutant produced crystals of rather high quality that could be solved to a resolution of 2.0 A˚ (compared with 2.7 A˚ for human kinesin‐1 structure 1MKJ which was crystallized in conditions similar to that used to crystallize monomeric rat kinesin). In addition to the flexible main chain connections, the ability to slide over the central b‐sheet may also impose restrictions on the side‐chain interactions between the switch‐2 cluster and its supporting structure. In fact, the cluster seems to be especially suited for gliding because of a hydrophobic patch at its inner surface that faces an extended region of predominantly hydrophobic residues at the central b‐sheet.

B.

Fungal Kinesin‐1

Compared with animal kinesin‐1, conventional kinesins of filamentous fungi are about four times faster and show greater processivity (Steinberg and Schliwa, 1996; Xiang and Plamann, 2003). The motor domain of N. crassa kinesin‐1 (NcKin355, amino acids 1–355) differs from that of rat and human kinesin‐1 in several distinct features (Song et al., 2001). Most remarkably, loop L11 is ordered (although the B‐factors are high) and visible. It comprises a helical part, a4a, that looks like an imperfect extension of the switch‐2 helix a4. The conformation of the switch‐2 cluster resembles that of rat and human kinesin crystallized with lithium sulphate, although the NcKin construct has been crystallized using PEGMME 2000 as the precipitant. Thus, it seems that stiffening of loop L11 favors the ‘‘permissive’’ state, that is, the conformation that allows docking of the neck linker to the core. However, the neck linker (amino acids 329–342 in NcKin) is only ‘‘semi‐docked’’: as expected, the binding pockets that are under control of the switch‐2 cluster are occupied by residues Ile330 and Asp332 of the N‐terminal half of the neck linker (b9), yet this does not result in binding of the second half, as it does in rat and human kinesin‐1. Instead, the C‐terminal half of the neck linker as well as the neck region that follows assume a random coil conformation without defined secondary structure and without much direct contact to the core.

314

MARX ET AL.

Other conformational differences occur in the switch‐1 region (a3‐L9‐a 3a): helix a3 of the NcKin structure is extended by five amino acids at its C‐terminus compared with rat kinesin‐1 (2KIN). This is at the expense of loop L9 and helix a3a, the structural elements that link the end of helix a3 to strand b6 of the central b‐sheet (Table I). In rat kinesin‐1 these elements comprise 12–14 amino acids in total, and a3a is a‐helical with two full turns (seven amino acids, MNEHSSR; here, bold letters highlight the switch‐1 motif). In the NcKin structure, the amino acids that correspond to loop L9 in the 2KIN structure are all incorporated into helix a3, and the remaining peptide chain (MNQESSR in NcKin) is stretched to span the distance between a3 and b6. Consequently, L9 of the NcKin structure consists of amino acids MNQE, while the remaining residues (SSR) form a short 310‐helix (a helix type that is more extended than the usual a‐helix). Thus, it seems that the peptide stretch that connects helix a3 to the distant and roughly antiparallel strand b6 contains a short sequence of amino acids that is prone to form an extension of helix a3, and another short sequence that may form an extension of a3a, but only one of these two possibilities can be realized at any time because of geometric constraints. This could result in a bistable, ‘‘switch‐like’’ behavior of the switch‐1 region. In the 2KIN conformation of the switch‐1 region, the nucleotide binding pocket is partially occluded by L9‐a3a residues, whereas it is easily accessible in the NcKin conformation. Thus it appears that the NcKin conformation facilitates nucleotide exchange and speeds up the ATPase cycle.

C. 1.

Other N‐Type Motors

Monomeric Kinesin‐3

Mouse Kif1A is a member of the kinesin‐3 (formerly Unc104/KIF1) family of kinesins. Crystal structures of the Kif1A motor domain (Kikkawa et al., 2001) have received much attention for two reasons. First, the Kif1A construct (a chimera of the Kif1A head with an engineered, short neck linker where six residues around b9 have been replaced by corresponding residues of mouse kinesin‐1) has been crystallized in several forms mimicking various intermediate states of the ATPase cycle. Second, Kif1A is a member of the kinesin‐3 family of ‘‘monomeric’’ kinesin‐like proteins. This raised the question about a variant mechanism of processive movement. Most models for kinesin‐1 movement assume a close coordination of the activities of two heads. At any time, at least one head is tightly bound to the microtubule, preventing rapid detachment and diffusion off the track. In the case of Kif1A, a positively charged insert of 11 amino acids in loop

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

315

L12 with a polylysine motif (‘‘K‐loop’’) may anchor the motor to the microtubule by electrostatic interaction with the negatively charged C‐termini of a‐ and b‐tubulin (‘‘E‐hook’’). This could compensate for the lack of a second motor domain (Okada and Hirokawa, 2000; Tomishige and Vale, 2000). On the other hand, there is growing evidence for reversible dimerization of Kif1A in situ (Klopfenstein et al., 2002; Tomishige et al., 2002). It appears now that Kif1A uses a mechanism for processive movement that is similar to that used by kinesin‐1. A reversible monomer‐dimer transition seems to be used as a method to regulate Kif1A’s motor activity. This has been confirmed by cryo‐electron microscopy showing that the neck of Unc104 protein (the C. elegans homologue of Kif1A) consists of two helical segments connected by a flexible hinge region that form an intramolecular coiled‐coil in the monomer. Under certain conditions, the Unc104 neck switches from the self‐folded state of the monomer to a true dimeric state by formation of an intermolecular coiled‐coil (Al‐Bassam et al., 2003). Superposition of the structures with AMPPCP and ADP bound to the active site (Kikkawa et al., 2001) shows little change in the catalytic core. The conserved serine S215 in switch‐1 (corresponding to RnKHC‐S203) has moved by 1.2 A˚ around the g‐phosphate, and the conserved glycine G251 of the switch‐2 motif (corresponding to RnKHC‐G235) has moved 0.6 A˚ towards the nucleotide (AMPPCP form compared with ADP form). In rat kinesin‐1 (2KIN), the positions of the corresponding amino acids are nearer to the ADP form than to the AMPPCP form of Kif1A, in accordance with the fact that rat kinesin—like most of the other kinesin structures— was crystallized with ADP. Although the two crystal forms do not differ very much in the immediate vicinity of the nucleotide, there are considerable changes in more distant regions, especially in the switch‐1 and switch‐2 regions. In the ADP form, the peptide stretch between helix a3 and the switch‐1 motif at the entrance to b6 is roughly similar to the corresponding region (loop L9 and a3a) of rat kinesin‐1; the secondary structure in Kif1A is less well defined but still predominantly helical. In the AMPPCP form, this region transforms into a b‐hairpin. The switch‐2 cluster adopts two conformations that closely resemble the conformations found in rat and human kinesin‐1 crystallized with lithium sulphate (permissive for docking of the neck linker; AMPPCP form) and with PEG (obstructive to neck linker docking; ADP form). Accordingly, the engineered neck linker of the Kif1A construct is undocked and disordered in the ADP form, and partially docked in the AMPPCP form. In the ADP form, the switch‐2 helix is elongated by two turns at the N‐terminal end. This leads to substantial shortening of loop L11 compared with the AMPPNP form and to a considerable shift

316

MARX ET AL.

of the helix toward the neck linker binding site, thus preventing docking of the neck linker. This is remarkable in view of the Nkin structure, in which stiffening of loop L11 and elongation of the switch‐2 helix by a4a go along with the ‘‘permissive’’ conformation of the switch‐2 cluster (Table I). Recently, three more crystal structures of the same Kif1A motor construct have been determined (Nitta et al., 2004), representing different intermediate states of the ATPase cycle: the motor domain in the complex with AMPPNP (PDB code 1VFV and 1VFW, these two structures of different resolution are virtually identical), ADP and aluminium fluoride (PDB code 1VFX), or ADP and vanadate (PDB code 1VFZ). In the light of the new structures, the AMPPCP structure (Kikkawa et al., 2001) is interpreted as an early, ‘‘preisomerization’’ or ‘‘collision complex’’ (Nitta et al., 2004), in accordance with the observation of only minor changes in the nucleotide‐ binding pocket. A better representation of the ‘‘prehydrolysis’’ state is the structure of the AMPPNP complex. Before isomerization (AMPPCP state), the linker between helix a3 and b6 (including the switch‐1 motif) assumes a tight b‐hairpin conformation and the side‐chain of the conserved serine S215 points away from the g‐phosphate. In the prehydrolysis state (AMPPNP), the switch‐1 region is partially melted (Table I) allowing the conserved serine to rotate and to approach the g‐phosphate. Likewise, the conserved glycine G251 of the switch‐2 motif approaches the g‐phosphate by another 0.6 A˚ , compared with the AMPPCP structure. The complex with ADP and aluminium fluoride is thought to resemble the early ADP.Pi state immediately after hydrolysis, whereas the complex with ADP and vanadate may represent a late state in which the phosphate (mimicked by vanadate) has moved quite a long distance (15 A˚ ) from the active center to the surface of the motor domain. There it is fixed by two hydrogen bonds to the solvent exposed tips of the switch‐1 loop region (L9) at one side and the switch‐2 loop (L11) at the other side. In the early posthydrolysis state (with ADP and aluminium fluoride), the switch‐1 region folds in a way that resembles that of the preisomerization state (AMPPCP), although the details are a bit different (Table I). Furthermore, there is an overall shift of the switch‐1 region by about 1 A˚ away from the nucleotide. Similarly, the switch‐2 motif and the switch‐2 cluster resemble the corresponding structures in both the AMPPCP and the AMPPNP state. The major change is a relative displacement along the axis of helix a4, with the position of the cluster being nearest to the nucleotide in the prehydrolysis (AMPPNP) state and farthest from it in the posthydrolysis (ADP.AlFx) state. In the preisomerization (AMPPCP) state, the position is intermediate. The displacement between the two extreme states is about 1.5 A˚ .

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

317

The ‘‘late ADP phosphate’’ state, as represented by the complex with ADP and vanadate (PDB code 1VFZ), is remarkable because it is the structure with one of the longest a4 helices in all known kinesin structures, and the only structure of Kif1A with loop L11 completely resolved. This is even more notable as the residues that are disordered and invisible in the other Kif1A complexes are surrounded by a structure similar to that observed in the complex with ADP (PDB code 1I5S; Kikkawa et al., 2001). In fact, the entire visible part of the ADP structure fits almost perfectly to the ADP. Vi structure (rms deviation for 324 Ca atoms: 0.36 A˚ ). Thus, it seems unlikely that the structure observed with ADP and vanadate can be really unique for the ‘‘late ADP phosphate complex’’ and different from that of the structure with only ADP. Rather, the same conformation of loop L11 and helix a4 could also prevail in the ADP state, but with a certain fraction of the molecules in the crystal assuming one or several other conformations, thus reducing the electron density for the dominant conformation below visibility in the most variable regions. The presence of the vanadate ion at the surface of the molecule could decrease this type of disorder and increase the number of molecules in the predominant conformation, thus raising the electron density to above the level of discrimination. Whether trapping of the cleaved g‐phosphate at the surface of the molecule (similar to vanadate) is of physiological relevance is hard to decide solely on the basis of the crystal structures. The main argument for the ‘‘late ADP phosphate state’’ being functionally significant draws on measurements of the (apparent) equilibrium dissociation constants for Kif1A binding to microtubules in the presence of different nucleotides (Nitta et al., 2004). These experiments suggest that the Kif1A cycle includes a state of very low affinity to microtubules. According to the authors, the ‘‘late ADP phosphate’’ state could be a good candidate for this ‘‘actively detaching state.’’

2. Tetrameric Kinesin‐5 The motor domain of human Eg5 (HsKSP), a member of the kinesin‐5 (formerly BimC) family, shares more than 40% identity with the kinesin‐1 motor domain. The overall structure of an HsKSP construct of the first 368 amino acids (including 10 amino acids of the class‐specific neck linker) complexed with ADP is very similar to the structure of kinesin‐1 (PDB code 1II6; Turner et al., 2001). The major differences are (1) an extension of the b‐hairpin b1b‐L2‐b1c in the N‐terminal lobe (‘‘L2 finger’’) due to an insert of eight amino acids, (2) an enlargement of loop L5 between a2a and a2b by another insert of eight amino acids, (3) an elongation of loop L10 between b6 and b7 at the tip of the core domain, and, most remarkably,

318

MARX ET AL.

(4) a novel conformation of the neck linker. Although the switch‐2 cluster is in the ‘‘obstructive state’’ (very similar to human kinesin‐1 crystallized with PEG), and docking of the neck linker is prohibited, the HsKSP neck linker is ordered and well defined in the crystal structure. It is almost straight, but without special secondary structure, and it extends roughly perpendicular to helix a6. This unusual conformation is stabilized by interactions of conserved residues within the neck linker and within the N‐terminal lobe (Turner et al., 2001). Family‐specific residues were also identified in the region that is involved in regular docking of the neck linker and, therefore, it was anticipated that the neck linker adopts the normal ‘‘docked’’ conformation (roughly in line with a6) when HsKSP switches to the ‘‘permissive’’ state. Information about the plasticity of the motor domain can be obtained by comparing the two crystallographically nonequivalent molecules of the HsKSP crystal (PDB code 1II6). The most remarkable difference pertains to the length of the switch‐2 helix a4, which differs by 10 amino acids (approximately three turns), most of them at the N‐terminal end (9 amino acids). This goes along with a considerable shortening of the disordered (invisible) loop L11. Interestingly, in the case of HsKSP, the variability of a4 has little effect on the position and orientation of the common part of the switch‐2 helix and the entire switch‐2 cluster. Other, fairly moderate, differences are restricted to the L2 finger and the loop L5 at the surface. The rms difference of the Ca positions after superposition of 318 amino acids (of a total of ~340 residues located in the crystal structure, disregarding the variable regions) is 0.65 A˚ . It should be noted that the high similarity of the two molecules includes the neck linker, which is well defined and structured in both molecules. Although the conformation of the switch‐2 cluster (apart from the variable length of a4) of human KSP corresponds to that of human kinesin‐1 in the ‘‘obstructive state’’ (PDB code 1BG2), the switch‐1 region (a3 and the linker between a3 and b6, including the switch‐1 motif) is quite different from human and rat but very similar to that of the fungal kinesin‐1 NcKin (PDB code 1GOJ). Helix a3 is longer by 1–2 turns at the C‐terminal end at the cost of the a3‐b6 linker. The shortened linker is stretched into a rather straight conformation with helix a3a restricted to three amino acids and transformed into a short 310‐helix (SSR of the switch‐1 motif). A comparison of the sequences and the secondary structure assignments (Table I) suggests that the conformations of the switch‐1 region in human KSP and fungal kinesin‐1 on the one hand, and human kinesin‐1 on the other hand represent two possible and thermodynamically significant (i.e., not singular) states. These conclusions are further substantiated by a recent structure of human Eg5 complexed with the small,

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

319

class‐specific, antimitotic compound monastrol (Yan et al., 2004). The ligand binds to a pocket formed by loop L5 and the N‐terminus of a3, close to the nucleotide binding site. Loop L5, which is one of the loops with the highest B‐factors in the structure without monastrol, adopts a rigid conformation by binding of monastrol (induced fit). In spite of the vicinity to the P‐loop, the drug has little effect on the nucleotide and the core b‐ structure. Nevertheless, helix a3, which is involved in ligand binding, moves about 1 A˚ in an axial direction. This is accompanied by a rearrangement of the linker between a3 and b6 toward a more a‐helical conformation (YSSR) of the switch‐1 motif. The main difference between the structure of the ternary complex of KSP with ADP and monastrol (Yan et al., 2004) and the complex with ADP alone (Turner et al., 2001) is that switch‐2 in the complex with monastrol adopts a permissive conformation, similar to rat and human kinesin‐1 (PDB code 2KIN, 1MKJ). Consequently, the neck linker binds to the docking site. As for the switch‐1 region, the conformation of the switch‐2 cluster is even more similar to that of the fungal kinesin‐1 (PDB code 1GOJ). It is remarkable that binding of monastrol to a site antipodal to the switch‐ 2 cluster has such a strong effect on distant structural elements at the periphery, whereas the central b‐sheet and the nucleotide binding site remain virtually unaffected. Another difference between the ternary and the binary complex is a bent conformation of the L2 finger in the complex with monastrol. Furthermore, the pointed tip of the core structure (b6–L10–b7) has a variable conformation in the structure with monastrol. In the two molecules of the asymmetric unit, the tip bends and moves by about 7 A˚ . The conformation observed without monastrol is intermediate between the two conformations of the ternary complex. The double conformation of L10 is clearly a crystal packing effect. However, it shows that the tip of the core domain is rather flexible. Binding of monastrol to loop L5 induces a conformational change of this loop and makes it more rigid. This allows two molecules to form a close‐packed dimer of twofold noncrystallographic symmetry (NCS) with loops L5 of both molecules at the common interface. This NCS dimer would probably not be stable without monastrol because of the intrinsic flexibility of loop L5. As a consequence, crystal packing of the ternary complex is totally different from that of the binary complex. This is most obvious in a curved, four‐stranded, intermolecular b‐sheet that is formed in the monastrol structure by antiparallel interaction of the L2 fingers of two molecules. Thus, bending of the L2 finger is also a crystal packing effect. It is, however, not clear to what extent changes in the switch regions and the neck linker may also be ascribed to crystal packing effects.

320 3.

MARX ET AL.

Kinesin‐7

The centromere‐associated protein CENP‐E (a member of the kinesin‐7 family) is another kinesin‐related molecule with an N‐terminal motor domain. CENP‐E from Xenopus species has been reported to be essential for the alignment of chromosomes during metaphase and to support slow plus end–directed motion in vitro, suggesting that its function is to tether chromosomes to the ends of dynamically growing and shrinking spindle microtubules (Wood et al., 1997). The crystal structure of a human CENP‐E construct including the motor domain and the neck linker (amino acids 2–342; PDB code 1T5C, to be released May 4, 2005; Garcia‐Saez et al., 2004) shows structural features characteristic for plus end–directed motors. In fact, the CENP‐E motor domain is largely superimposable with the human conventional kinesin motor domain. Remarkably, the switch‐2 cluster is in the permissive conformation as observed for rat kinesin‐1 (PDB code 2KIN) and human kinesin‐1 crystallized with lithium sulphate (PDB code 1MKJ), although the CENP‐E motor construct has been crystallized with PEG as the precipitant. Accordingly, the neck linker adopts a docked conformation and forms two short b‐strands that interact with the central b‐sheet in the same way that they do in the case of rat conventional kinesin. Major differences between CENP‐E and conventional kinesin are found in the N‐terminal lobe and at loop L10 between b6 and b7, at the tip of the motor domain. Human CENP‐E has a five‐residue insert in loop L10 compared with conventional kinesin. Due to the additional residues, the tip of the CENP‐E motor domain is more flexible and, thus, invisible in the crystal structure. At the N‐terminal lobe, helix a0 of conventional kinesin is replaced by an extended loop without special secondary structure assignments.

D.

Kinesin‐14 (C‐Type Motors)

Structures of the motor domains of three different kinesin‐14 proteins, kinesins with a C‐terminal motor domain, have been determined so far: (1) the Drosophila nonclaret disjunctional gene product (DmNcd), (2) the yeast kinesin‐like nuclear fusion protein (ScKar3), and (3) the kinesin‐like calmodulin binding protein from potato (PoKCBP).

1.

Ncd

C‐type kinesins have a class‐specific neck at the N‐terminal side of their motor domain. In the case of DmNcd, the neck forms a continuous a‐helix with the less conserved stalk. Constructs of the motor domain with a sufficiently long part of the neck dimerize by formation of a coiled‐coil.

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

321

The first structure of an Ncd motor domain (Sablin et al., 1996) was that of a monomeric construct (amino acids 335–700) including only part of the neck (amino acids 328–348). Later, several crystal structures of dimeric constructs were published. These constructs dimerize by coiled‐coil interaction of the neck/stalk helices. The PDB database now contains three crystal structures of dimeric DmNcd constructs that differ mainly by the overall conformation (symmetry) of the dimers, whereas the individual motor domains are very similar in all known structures. The first structure of a dimeric DmNcd construct (PDB code 2NCD; amino acids 281–700; Sablin et al., 1998) turned out to be perfectly symmetric (by contrast to dimers of rat kinesin‐1): the two molecules of a dimer are related by a crystallographic twofold axis. The symmetry axis coincides with the axis of the coiled‐coil (Fig. 3C). A similar construct (PDB code 1CZ7; amino acids 295–700; Kozielski et al., 1999) crystallized in a different space group with two dimers per asymmetric unit. Although none of the dimers has a proper twofold symmetry, their conformation is not far from that. The deviation from perfect symmetry can be described by 2‐ and 10‐ degree torsions, respectively. Recently, a novel conformation of dimeric Ncd has been found in crystals of a point mutant (DmNcd‐N600K, PDB code 1N6M; amino acids 293–700; Yun et al., 2003). As crystals of the same type could be produced from the corresponding wild‐type construct—although at lower resolution—it is presumed that the new conformation represents a state that plays a significant role in the mechanochemical cycle of Ncd. The new conformation can be obtained from the symmetric conformation by a 75‐ degree rotation of one head, leaving the other head and the neck coiled‐ coil untouched (Fig. 3D). The pivot point is G347 at the transition from the neck to the globular motor domain. The axis of rotation is perpendicular to the coiled‐coil (i.e., the former symmetry axis). In spite of the gross conformational differences between the Ncd dimers, there are only minor differences between the individual motor domains. The overall fold of the motor domain is very similar to that of kinesin‐1 and other N‐type motors. Major differences are: (1) The N‐ terminal lobe of Ncd is enlarged (þ9 amino acids) compared with rat kinesin‐1. The additional residues are located between b1b and b1c (the ‘‘L2 finger’’). This, however, does not result in a simple elongation of the b‐hairpin as in HsKSP and in M‐type motors (see below). In fact, the ‘‘tip’’ of the ‘‘L2 finger’’ is rather broadened and forms a short a‐helix. (2) Loop L5, the insert in the P‐loop helix a2, is quite short (approximately eight residues compared with 12 in rat kinesin‐1), due to three residues that are missing in the primary structure of DmNcd. (3) Switch‐1 helix a3 is short and loop L9, the linker between a3 and b6 that includes the switch‐1

322

MARX ET AL.

motif, is rather long, but without any defined secondary structure. In a superposition of all known Ncd motor domains, loop L9 is the region with the highest variability. (4) Helix a4 and the entire switch‐2 cluster is in the ‘‘obstructive’’ conformation (similar to human kinesin‐1 crystallized with PEG). The adjacent b5‐L8 lobe follows the movement of the switch‐2 cluster. (5) According to the obstructive conformation, the C‐terminal residues beginning with the end of helix a6 (about 30 amino acids) are disordered and invisible. So far, the significance of this is unclear since the neck linker is N‐terminal to b1, and it is ordered and ‘‘docked’’ to the core via multiple interactions with helix a1 and loops L6 and L10 at the tip of the core domain. The neck also contacts K640 at the start of b8, the strand adjacent to b1. K640 is next to the flexible hinge (loop L13) of the switch‐2 cluster and to the principal docking site for the neck linker in N‐type motors. It seems that simultaneous binding of Ncd’s N‐terminal neck and the motor domain’s C‐terminal extension (which may be considered a ‘‘pseudoneck’’; disordered amino acids 671–700) to the docking site is excluded by steric hindrance. Thus, it might be hypothesized, that the real neck and the C‐terminal pseudoneck compete for binding to the core domain. Then, switch‐2 movement could play a critical role in driving the conformation between a state with the neck ‘‘docked’’ (as observed in the Ncd structures) and another state with the C‐terminal peptide docked and the real neck displaced. This hypothesis is strongly supported by the structure of PoKCBP that has recently been solved (Vinogradova et al., 2004; see below). In the N600K mutant, the switch‐2 helix a4 of the rotated head moves toward the switch‐2 motif by attraction of the mutated residue at the proximal end of a4 (N600K) toward the conserved switch‐1 arginine R552, but without substantial changes at the distal end next to the presumed docking site. (Interestingly, attraction towards the switch‐1 motif leads to a partial unwinding of helix a4.) Thus, both heads of the asymmetric dimer retain an obstructive conformation, and the C‐terminal residues beyond helix a6 remain disordered and invisible. Nevertheless, the small change in the switch‐2 conformation is accompanied by a large‐ scale (approximately 75‐degree) rotation of the neck helix relative to the head. Another consequence of the rather limited rearrangements in the switch regions is that the nucleotide (ADP) seems to be less tightly bound to the rotated head compared with the other head, as indicated by a reduced electron density of the adenosine moiety.

2.

Kar3

In contrast to Ncd, native Kar3 is a heterodimer with a single motor domain associated to either Cik1 or Vik1 (Mackey and Gilbert, 2003). Kar3 is involved in spindle assembly and integrity. It is a slow, minus

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

323

end‐directed motor and seems to have microtubule destabilizing activity. The structure of the ScKar3 motor domain has been solved in two variants that differ in length by 11 amino acids at the N‐terminus (PDB code 3KAR, amino acids 383–729; Gulick et al., 1998; and PDB code 1F9T, amino acids 372–729; Yun et al., 2001). In addition, structures of three point mutants of the shorter construct have been determined (PDB code 1F9U, 1F9V, and 1F9W; Yun et al., 2001). The Kar3 motor domain is very similar to that of other kinesins and especially to Ncd. Like Ncd, Kar3 has an N‐terminal lobe considerably larger (þ12 amino acids) than that of kinesin‐1. The predominant effect is that the b1a,b,c sheet appears uniformly elongated. Neither the loop between b1b and b1c nor the loop before the b1a,b,c sheet is helical as in Ncd and kinesin‐1. The loop L5 insert between a2a and a2b is even shorter than that of Ncd (–5 amino acids). Another difference at the sequence level consists in a 10– to 11–amino acid increase in length of the b5–L8 lobe compared with Ncd and kinesin‐1. On the conformational level, this is accompanied by a significant relocation of the lobe toward the tip of the motor domain, associated with a 90‐degree rotation (untwisting) of the b5a–b5b hairpin. This leads to a considerable approach of b5b to the central b‐sheet, closing the gap between the b5–L8 lobe and the core domain. Remarkably, in all Kar3 structures (wild‐type and mutants), the root of the b5–L8 lobe is partially disordered (residues 532–545, next to the site where the lobe is attached to the central b‐sheet). Other conformational differences between Kar3 and Ncd, which vary to a certain extent between the Kar3 structures presently available, are found in the switch regions. In the wild‐type structures, helix a3 moves and tilts away from the central b‐sheet and bends toward the nucleotide binding site. The linker between helix a3 and strand b6 containing the switch‐1 motif is partially a‐helical (a3a), similar to kinesin‐1. Compared with kinesin‐1, the tilting of helix a3 is even more pronounced, whereas the position and orientation of the short linker helix a3a are virtually unchanged. To accommodate the large tilt angle, loop L9 between a3 and a3a moves by 18 A˚ (Ca distance between ScKar3‐Thr587 and RnKHC‐ Ala194). The switch‐2 cluster has a conformation similar to Ncd (‘‘obstructive’’). However, the switch‐2 helix a4 is longer by nine (3KAR) or six (1F9T) amino acids at the end proximal to the switch motifs. The variation of length of a4 in the wild‐type structures differing by 11 amino acids at the N‐terminus suggests that residues at the proximal end of the switch‐2 helix and the adjacent loop L11 are prone to reversible melting, and the exact conformation may depend on subtle details that cannot be controlled easily. Among the Kar3 point mutants, the N650K mutant (within helix a4, close to the invariable core of the switch‐2 helix; PDB code 1F9U) is almost

324

MARX ET AL.

identical to the wild‐type construct. Only small, local differences are observed at the site of the amino acid exchange. Nevertheless, the mutation has marked effects on the kinetics of the construct (Yun et al., 2001): the mutant binds tightly to microtubules, independent of the nucleotide state, it displays no motor activity in the microtubule gliding assay, and its ATPase activity is not stimulated by microtubules. It has been suggested that the N650K exchange may disrupt the communication pathway between the microtubule binding site (switch‐2 cluster) and the nucleotide binding site (switch motifs, primarily switch‐2). This mutant has, therefore, been called a ‘‘decoupling mutant.’’ The ‘‘salt bridge mutants’’ R598A (conserved arginine of the switch‐1 motif; PDB code 1F9V) and E631A (conserved glutamine of the switch‐ 2 motif; PDB code 1F9W) both have marked functional and structural effects. In some of the kinesin structures, a salt bridge between the conserved arginine in switch‐1 and the conserved glutamine in switch‐ 2 has been found, and it has been suggested that this salt bridge plays an important role in the coordination of the switch regions. Disruption of the salt bridge by either mutant leads to the loss of microtubule stimulated ATPase activity similar to the decoupling mutant. Interestingly, the switch‐ 2 mutant (E631A) binds tightly to microtubules irrespective of the nucleotide, whereas the switch‐1 mutant (R598A) has only weak affinity to microtubules. Surprisingly, the main structural effect of the switch‐2 mutant is in loop L9 of the switch‐1 region, which is largely disordered. The short helix a3a is further reduced to a 310‐helix of minimal size (three residues, SSR of the switch‐1 motif). By contrast, the switch‐1 mutant displays strong effects in both switch regions. The switch‐1 region is largely disordered. Helix a3 is shortened (at the C‐terminal end) and partially distorted, and its orientation/position is more similar to Ncd and kinesin‐1 than to the other Kar3 structures. The linker between a3 and b6 is invisible except for the residues of the switch‐1 motif immediately N‐terminal to b6. The destabilizing effect on the switch‐1 region is accompanied by stabilization of loop L11 between the switch‐2 motif and switch‐2 cluster. The R598A mutant of ScKar3 is one of three structures so far with loop L11 ordered and visible. It should be noted that stabilization of loop L11 has virtually no effect on the conformation of the switch‐2 cluster. Helix a4 is extremely long, and its conformation relative to the core is ‘‘obstructive’’ as in the Kar3 wild‐type structures. Accordingly, the loop is restricted to nine amino acids (from the end of the switch‐2 motif to the first amino acid of helix a4). Three of them in the center of the loop (V635‐S636‐Q637) form a short 310‐helix antiparallel to helix a4. This is remarkably similar to the structure of PfMCAK (Shipley et al., 2004; see below), where loop L11 and a4 form a full‐fledged helix‐loop‐helix motif.

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

325

The low affinity of the R598A mutant to microtubules has been explained by the rigidity of loop L11 (Yun et al., 2001): strong binding may require a flexible loop that is able to accommodate to the microtubule surface. If this assumption is true, the helix‐loop‐helix conformation of L11‐a4 seen in the Kar3 mutant and in PfMCAK should be different from the conformation induced by strong binding to microtubules (in the presence of ATP). According to the model proposed by Rice and coworkers (1999), this unknown conformation should be accompanied by the transition of the entire switch‐2 cluster from the obstructive to the permissive state, leading to docking of the neck linker. How does the NcKin structure fit these assumptions? In NcKin, the loop folds in a way faintly similar to the Kar3 mutant. It is much larger (approximately 15 amino acids), meaning that the switch‐2 helix is relatively short, and it displays no well‐defined secondary structure, if one neglects a4a, the short 310‐helix defined in the NcKin structure. Helix a4a should not be confused with the short antiparallel helix seen in the Kar3 mutant. It should rather be considered as a part of the switch‐2 helix. Thus, the switch‐2 helix in NcKin is rather short (compared with Kar3) and kinked, yet it is in the permissive orientation and allows docking of the neck linker. It seems that loop L11 of the NcKin structure is in an intermediate state that shares features observed in the structure of the Kar3 mutant as well as features anticipated for the tightly bound microtubule complex.

3. KCBP Recently, the motor domain structure of a kinesin‐like calmodulin‐ binding protein from potato (PoKCBP), another member of the kinesin‐14 family, has been reported (residues 884–1252; PDB code 1SDM; Vinogradova et al., 2004). The construct has been crystallized with PEG in the presence of 200 mM sodium phosphate. It is the first structure of a minus end‐directed motor domain that shows most of the C‐terminal residues beyond helix a6 (amino acids 1209–1252) in a well‐defined conformation. The C‐terminal extension of the motor domain contains a calmodulin‐binding motif (amino acids 1209–1252) that forms an a‐helix, a short peptide sequence (the ‘‘neck mimic,’’ according to Vinogradova et al.) that connects this helix to the end of a6, and a negatively charged sequence at the C‐terminus, part of which binds to the microtubule binding surface of the core domain while the rest is disordered. Although the construct has ADP bound to the active site, the switch‐2 cluster resembles that of Kif1A complexed with AMPPCP (presumed ATP state, ‘‘permissive’’ conformation). Strikingly, the neck mimic and the calmodulin‐binding helix that follows assume a conformation very similar to the neck linker

326

MARX ET AL.

and neck helix in rat and human conventional kinesin crystallized with lithium sulphate: the neck mimic is docked and runs parallel to the edge of the core structure to the tip, where the calmodulin‐binding helix is attached at a right angle to the neck mimic. The only difference is that the calmodulin‐binding helix points in the opposite direction compared with the neck helix of rat conventional kinesin. Unfortunately, the present KCBP structure cannot tell anything about the true neck because it was not part of the construct.

E. Kinesin‐13 (M‐Type Motors) The characteristic feature of the kinesin‐13 family is that members of this kinesin family have their ‘‘motor’’ domain surrounded by N‐ and C‐terminal domains. Therefore, they are also named internal kinesins (KinI) or M‐(middle)‐type kinesins. The main function of this class of kinesin‐related proteins is to target to the ends of microtubules and to induce depolymerization. Although it is not clear whether these proteins display motor activity in a strict sense, it seems appropriate to include them into this review of motor proteins.

1.

Kif 2C

Ogawa and colleagues (2004) have determined the crystal structure of the minimal construct of mouse Kif 2C that preserves full microtubule destabilizing activity. The construct (amino acids S183–S585 of MmKif2C, þ 7 His) comprises the catalytic core (amino acid R254–S585, from b1 to a6, including three additional residues at the C‐terminus) as well as 70 residues of the family‐specific neck that is N‐terminal to the core domain. This construct has been crystallized and solved with ADP (PDB code 1V8J) and with AMPPNP (PDB code 1V8K). There is no significant difference between the two crystal forms. The catalytic domain of MmKif2C is not much different from motor domains of other kinesins. Variations from the structure of kinesin‐1 are within the range covered by N‐ and C‐terminal kinesin motors. Loop L9 in the switch‐1 region is partially disordered. Switch‐2 loop L11 is also disordered as in most other structures, yet helix a4 is quite long and straight. The switch‐2 cluster adopts an ‘‘obstructive’’ conformation that would prevent docking of a C‐terminal peptide stretch, if present. The most remarkable feature of the catalytic domain is an extension of the L2 finger by 13 additional residues between b1b and b1c, similar to Eg5, but even more pronounced. As the tip of this long b‐hairpin contains three KinI‐conserved residues (K293‐V294‐D295) that are essential for

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

327

microtubule depolymerization by Kif 2C (Ogawa et al., 2004), this structure has been termed the ‘‘KVD‐finger.’’ The Kif 2C structure also reveals structural features of the N‐terminal neck. Although partially disordered (only ~30 of 70 neck residues are visible), the crystal structures show a sequence of 14 amino acids forming an a‐helix surrounded by flexible regions that are not visible. Interestingly, this helix is roughly perpendicular to the microtubule‐binding surface of the core domain. It has been proposed (Ogawa et al., 2004) that the neck helix serves a double function. It may prevent strong binding to the microtubule side‐wall, thus facilitating one‐dimensional diffusion. Once the molecule has reached the end of the microtubule, the neck helix may strengthen the interaction with a terminal tubulin subunit.

2. MCAK The catalytic domain of the MCAK homolog of Plasmodium falciparum (named pKinI or PfMCAK; PDB code 1RY6; Shipley et al., 2004) has been solved at high resolution (1.6 A˚ ). The crystallographic model corresponds to the core motor domain (residues 68–396). Although this construct lacks the N‐terminal neck, it still has microtubule depolymerizing activity. The construct was crystallized with high concentrations of sulphate. Interestingly, no nucleotide was found in the nucleotide binding site. Instead of a nucleotide, a sulphate ion is bound to the P‐loop at exactly the same position that is usually occupied by the b‐phosphate. Notwithstanding the absence of a nucleotide, the nucleotide binding site shows little change, if any, compared with kinesin structures with ADP in the active site. There is a 1‐A˚ shift in switch‐1 and switch‐2, indicating a slight opening of the nucleotide‐binding pocket. As a consequence, a hydrogen bond between D236 in the switch‐2 motif (DLAGSE) and the conserved P‐loop threonine (T99) is broken. However, all conserved amino acids that normally interact with the nucleotide have their side‐chains in similar positions as ADP‐bound motor domains. By contrast to Kif2C, loop L8 is partially disordered and seems to point in a direction opposite to the normal direction. On the other hand, both switch regions are ordered and completely resolved, including the ‘‘loop’’ between the switch‐2 motif and helix a4, which forms a two‐turn 310‐helix in PfMCAK. This short helix is stabilized by hydrogen bonds between residues of the switch‐1 motif and residues of the switch‐ 2 loop that are conserved in members of the kinesin‐13 family (S210‐R242 and R211‐D245). These two hydrogen bonds replace the salt bridge between switch‐1 and switch‐2 that is observed in other kinesin structures and that is considered essential for the communication between the

328

MARX ET AL.

microtubule binding site and the catalytic center (cf Kar3 mutants, salt bridge between R598 and E631, corresponding to R211 and E241 in PfMCAK).

IV.

Conformational Switching in Kinesin A.

Comparison with Myosin

Myosins are actin‐based motors (see Chapter 5). The reason for considering myosin in the context of microtubule motors is that the catalytic domains of myosin and kinesin share structural similarities indicating that both families use a similar mechanism for energy conversion. The structural relationship between these families suggests that both descend from a common ancestor, a primordial nucleotide binding protein (Kull et al., 1998). Like kinesins, myosins form a superfamily of proteins with a growing number of myosin classes. Myosins are present in most, if not all eukaryotic cells, and they serve functions in muscle contraction, cytokinesis, cellular locomotion, and actin‐based, short‐range transport of vesicles and organelles. Myosins are characterized by a heavy chain with a highly homologous, globular, ~80‐kDa catalytic domain (motor domain). Most myosin heavy chains consist of an N‐terminal motor domain followed by an a‐helical light‐chain binding domain (LCBD) and a C‐terminal tail. For historical reasons, class II myosins, comprising striated muscle myosin‐II as well as smooth‐muscle and non‐muscle myosin‐II, are referred to as conventional myosins. Myosin‐II consists of two heavy chains, each complemented by two light chains. With its coiled‐coil tail domain, myosin‐II forms bipolar spindles or filaments as the active part of a contractile system (see Chapter 2). More than 18 classes of myosin have been identified in different organisms so far. Myosins of class V and class XI (the plant class V) are most akin to kinesin because they are dimeric, plus end (barbed end)‐directed, processive motors used for membrane and particle transport along actin filaments. There is also a class of myosins with reversed motility (class VI). It is not yet clear whether these are monomeric or dimeric. Myosin‐I, the first class of ‘‘unconventional’’ myosins to be identified, comprises monomeric motors with a basic tail that interacts by electrostatic interaction with the cargo. The structure of the myosin‐II motor domain has been determined in various nucleotide states by crystal structure analysis using constructs originating from diverse sources (chicken skeletal and smooth muscle

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

329

myosin‐II, as well as myosin‐II from scallop and non‐muscle myosin‐II of the slime mold Dictyostelium discoideum). In addition, structures of unconventional chicken myosin‐V (Coureux et al., 2003) and of the monomeric D. discoideum myosin‐IE (Kollmar et al., 2002) have been solved recently. The motor domains of all myosins determined so far are structurally very similar. The motor domain has an elongated shape. The actin‐binding site is located at one end and is split in two subdomains (the 50K upper and lower domains) with a marked cleft in between. Some of the solved structures include part of the a‐helical LCBD originating at the opposite end of the motor domain. The orientation of the LCBD differs by 70 degrees and more in various structures, thus underpinning the hypothesis that the a‐helical extension of the motor domain, stiffened by the light chains, serves as a lever arm that amplifies conformational changes powered by nucleotide processing in the core of the motor domain (Holmes and Geeves, 2000; Chapter 5). The swinging lever arm hypothesis for conventional myosin has been strongly supported by functional analysis of myosin heads with genetically modified LCBD (Uyeda et al., 1996) and constructs with artificial a‐helical extension of the motor domain (Anson et al., 1996), showing that the velocity (i.e., the step size) is proportional to the length of the engineered lever arm. Furthermore, it is even possible to reverse the direction of motility by redirecting the lever arm by 180 degrees, which is formally equivalent to a motor with a lever arm of negative length (Tsiavaliaris et al., 2004). As in the case of kinesin, the folding of the motor domain can be described as a central b‐sheet that constitutes a supporting structure for the surrounding elements. However, these elements greatly exceed the peripheral structural elements of kinesin in size and they are mostly a‐helical subdomains. The nucleotide binding site is buried in the middle of the motor domain. It is structurally and topologically homologous to that of kinesin and comprises four characteristic motifs that are conserved in all myosins and that are similar to the nucleotide binding motifs described for kinesin. These are the adenine binding motif, the phosphate binding P‐loop, and the two switch motifs. Comparison of myosin structures in different nucleotide states showed that the switch‐2 conformations fall roughly into two classes, open and closed, with the open conformation supposed to occur predominantly in the ADP or apo state (i.e., in the absence of a g‐phosphate), and the closed conformation occurring preferentially in structures with ATP or transition state analogues. Similarly, the lever arm (or the ‘‘converter domain’’ at the base of the lever arm, which is indicative of the lever arm position in structures of truncated motor domains) is either up or down (or more descriptively, in

330

MARX ET AL.

pre‐power stroke or post‐power stroke conformation). Intermediate states can be defined both for the switch‐2 conformation (e.g., ‘‘half open’’) and for the orientation of the lever arm (e.g., stepwise rotation accompanied by successive release of Pi and ADP; Houdusse et al., 2000; Veigel et al., 1999). In any case, switch‐2 and the lever arm are tightly coupled. The translation of the switch‐2 movement into lever arm rotation is ascribed to a rotation of the converter domain that is driven by a long (~40 A˚ ) helix between switch‐2 and the converter domain (‘‘relay helix’’ or ‘‘switch‐2 helix’’). It has long been surmised that switch‐2 movement and the concomitant swinging of the lever arm must be controlled by binding to and detachment from the actin filament to avoid futile consumption of ATP. However, direct evidence was lacking because near‐atomic resolution crystal structures are necessarily obtained in the absence of the filament. Now, crystal structures of Dictyostelium myosin II (Reubold et al., 2003) and chicken myosin‐V (Coureux et al., 2003) have revealed that the switch‐1 motif can also exist in open and closed conformations. It has been inferred that switch‐1 opening may be coupled to cleft closure and tight binding of the myosin head to the actin filament. This conclusion is supported by electron microscopy (Holmes et al., 2003) and fluorescence spectroscopy (Conibear et al., 2003) studies of the acto‐myosin complex, which show that the concepts derived from crystal structures of isolated myosin heads are indeed valid for the functional complex. These observations show that there are striking similarities between myosin and kinesin motors, suggesting that both use a similar if not the same mechanism for transforming ATP’s free energy into directed motion. There are, however, also notable differences, both in structure and kinetics, which may reflect the diverse functions of the motors. The myosin motor domain seems to be composed of structurally and functionally well‐separated building blocks. The actin binding site and the main mechanical actuators (converter domain and lever arm) are located at opposite ends of the head domain, with the catalytic center in the middle between them. Binding to the actin filament competes with nucleotide binding: the nucleotide‐free myosin head binds strongly to F‐actin (rigor state) while binding of ATP to the catalytic center leads to detachment from the filament. This competition is due to a mechanical linkage of actin binding site and switch‐1. Cleft closure is linked to switch‐1 opening and vice versa. On the other hand, switch‐2 movement is coupled to rotation of the converter domain and swinging of the lever arm. Communication between the actin binding site and the lever arm is both mediated and controlled by the catalytic center. It seems that proper

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

331

coordination of the two poles of the head domain is based on a cooperative behavior of the switches, most probably due to the formation of a salt bridge between switch‐1 and switch‐2 when both are open or closed, and breakage of the salt bridge in mixed states (Reubold et al., 2003). The important role of this salt bridge and the analogous salt bridge in kinesin (R204‐E237 in rat kinesin‐1) has been demonstrated by kinetic and structural studies of single and double mutants (Klumpp et al., 2003; Onishi et al., 1998; Yun et al., 2001). In contrast to myosin, attempts to solve kinesin structures in different nucleotide states have had rather limited success so far, and in the few crystal structures with an ATP analogue or without nucleotide, the observed effects on the catalytic center are small. Nevertheless, it is generally assumed, by analogy with myosin, that nucleotide processing, microtubule binding, and force generation are coordinated by means of some conformational changes in switch‐1 and switch‐2. However, there are obvious differences from myosin: (1) Because of the limited size of the kinesin motor domain, the filament binding site and the mechanical actuator are not well separated from each other. This suggests the possibility of direct interactions, which may not easily and exclusively be controlled by the catalytic core. (2) Main chain connectivity and the spatial vicinity of switch‐2 and the main microtubule binding site (the ‘‘switch‐2 cluster’’) strongly indicate that filament binding is directly linked to switch‐2 movement, not to switch‐1 movement as in myosin. (3) The switch‐2 cluster with helix a4 serves two distinct functions: one in microtubule binding and another in controlling kinesin’s supposed mechanical actuator (the neck linker). In contrast to myosin, these two functions seem to be coupled in a single chain of actions. (4) The role of switch‐1 in kinesin is quite elusive. It seems that in kinesin, switch‐2 plays a dominant role in controlling both filament binding and force generation. However, there is growing evidence that switch‐1 movement and the conformational variability of the switch‐1 region (a3‐loop9‐a3a) are coupled to the b5‐ L8 lobe, which is also involved in microtubule binding (Ogawa et al., 2004). It seems plausible that rigid body movement of a single binding site (the switch‐2 cluster) would not change the affinity too much, but would only result in repositioning of the motor domain (cf Kikkawa et al., 2001). To change the affinity, it would be more effective to change the spatial arrangement of two (or more) binding sites. (5) Perhaps the most important difference between kinesin and myosin is that, in kinesin, the main microtubule binding site together with the mechanical actuator and the switch‐2 that should control them are largely decoupled according to the crystal structures, as seen in the occurrence of permissive and obstructive conformations without noticeable changes in the catalytic

332

MARX ET AL.

center. This is due to loop L11, which is disordered in most crystal structures but may assume a rigid conformation when the motor domain attaches to the microtubule surface. This means that kinesin assumes a fully functional form only while it is bound to the filament.

B.

Nucleotide Binding, Switch‐I and ‐II, and Conformational Relays

According to the ‘‘walking model’’ for processive motion of kinesin‐1, switch‐2–induced transition from the ‘‘obstructive’’ to the ‘‘permissive’’ state in one head allows binding of the neck linker to this head in a zipperlike manner. By this, the second (trailing) head is repositioned toward the plus end of the microtubule, thus increasing the probability for the second, now leading head to bind in the forward direction. By contrast to the swinging lever arm model of myosin, the movement of kinesin can be considered a diffusive process that is biased by nucleotide‐ controlled docking of the neck linker (thermal ratchet). At any time, at least one head is attached to the microtubule, which enables single molecules of conventional kinesin to function as ‘‘porters’’ (Leibler and Huse, 1993). Communication between the two heads is probably mediated by mechanical strain that is introduced by simultaneous binding of both heads to adjacent binding sites. Conventional myosin, on the other hand, has a low duty ratio; it is mostly detached from the actin filament in active muscle. Productive motion is the result of a large number of myosins (assembled, for instance, in the muscle’s thick filament) working in cooperation. The heads of a single myosin molecule seem to act independently; the main advantage of having two heads probably consists in increasing the chances to find a binding site on the actin filament. For unconventional myosins and kinesins, the mode of action may be different. Class V and class VI myosins, and other myosins that function as single‐molecule transporters of organelles, may use a mechanism more similar to that of kinesin‐1. Conversely, the minus end–directed kinesin Ncd seems to use its a‐helical neck coiled‐coil, stabilized by interaction with one of its heads, as a rigid lever arm that performs a large‐scale rotation relative to the other head (Yun et al., 2003). Electron microscopy of microtubules decorated with Ncd indicates that only one head of this dimeric, but unprocessive motor functions as an active, force‐ generating ATPase, suggesting a hopping‐type motion that combines a deterministic lever arm movement with a diffusive component (Wendt et al., 2002). As a general picture resulting from this synopsis, it appears that the different members of the myosin and kinesin protein families all combine

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

333

a common switch‐based ATPase mechanism with various transducing elements to convert chemical free energy into mechanical work. A characteristic feature of this type of chemomechanical nanomachine is the concerted action of conformational changes on extremely different scales in time and space: local (subnanometer scale) rearrangements in the catalytic center induced by virtually instantaneous (probably subpicosecond) cleavage of ATP into ADP and phosphate drive large‐scale conformational changes (of the order of tens of nanometers) on a timescale of milliseconds. The combination of small‐ and large‐scale movements is intuitively described by the concept of switches. To account for the different timescales, the existence of elastic storage elements that can match conformational changes occurring at different timescales has to be assumed. These elements should be able to store the energy set free by ATP hydrolysis and to release it slowly during the process of motion like the spring of a clock. The nature of these storage elements may be as diverse as the types of motor proteins are. Comparison of the molecular structures, however, suggests that reversible melting of helices plays an important role in energy storage and conversion. In myosins, transitions of the relay helix and the SH1–SH2 helix between straight and bent conformations should be mentioned in this respect (Houdusse et al., 2000). For kinesins, reversible melting of helical regions is most evident in the observed length changes of the switch‐2 helix, in the variability of the switch‐1 region (helix a3 and the linker between this helix and the switch‐1 motif), and possibly in reversible unwinding of the neck coiled‐coil.

V.

Structures of Kinesin‐Related Domains

Compared with the structural knowledge about the kinesin core motor domain, less is known about its nonmotor parts or associated proteins. One exception is the a‐helix (neck helix) preceding (as in kinesin‐13 or ‐14) or following (as in kinesin‐1) the motor domain. In rat kinesin‐1, a short sequence of amino acids (T326–T338) links the neck helix to the motor core. The neck helix (amino acids A339–W370) is separated from the coiled‐coil stalk by a stretch of amino acids with predicted random conformation (hinge, approximately amino acids 371–410; Tripet et al., 1997), allowing the stalk to kink and swivel. The NMR structure of the peptide K357–D386, comprising the second part of the neck helix and the first part of the hinge, supports these predictions. The X‐ray and NMR structures agree well in the range between residues K357 and W370. Beyond this tryptophan, there is no electron density in the X‐ray structure, although the crystallized construct comprises nine more residues. The NMR structure shows many possible orientations in this region (Seeberger

334

MARX ET AL.

et al., 2000). In C‐type kinesins (kinesin‐14 family), the conserved neck is entirely helical and connects to the less conserved coiled‐coil of the stalk without any hinge region (Sablin et al., 1998). The neck helix of the middle‐type motor MmKif2C (kinesin‐13 family) has a conformation very different from other kinesins as it points toward the microtubule surface in a direction almost perpendicular to the neck helices found in N‐ and C‐type motors (Ogawa et al., 2004). This may reflect its function as a microtubule depolymerizing machine. In contrast, the neck helices of N‐ and C‐type motors are more or less parallel to the microtubule surface when the motor binds to the microtubule lattice. The structure of the globular tail domain of kinesin‐1 is still unknown. It contains determinants necessary for folding into an inactive conformation (Stock et al., 1999). The structure of kinesin light chains (KLC, KAP) have not yet been determined crystallographically either. However, local sequence homologies to other proteins of known structure allow some predictions (Fig. 6; Mandelkow and Mandelkow, 2002). The C‐terminal domain of the kinesin light chains (~320 amino acids) contains six tetratricopeptide repeats (TPR) probably involved in protein–protein interactions. The structure of these repeats can be modeled according to structures of other TPR‐containing proteins like protein phosphatase PP5 (PDB code 1A17; Das et al., 1998). Each TPR domain comprises about 34 residues presumably folded into two antiparallel helices. The N‐terminal part of the light chains (~250 amino acids) has regions predicted to be engaged in coiled‐coil interactions linking the light chains to the heavy chains. In the case of heterotrimeric kinesin‐2, the two motor molecules are slightly different and they are associated with only one nonmotor subunit (KAP). It is a largely a‐helical protein containing 11 armadillo repeats, a motif first described in Drosophila armadillo protein but also found in many other proteins like b‐catenin.

VI.

Dynein Structure

Dyneins are microtubule‐based motors responsible for minus end–directed transport in eukaryotes. Unlike the other motor proteins, kinesins and myosins, dyneins are huge complexes consisting of one to three heavy chains of >500 kDa and several intermediate, light intermediate, and light chains, some of them specific to dynein subclasses (Vallee et al., 2004). Only a limited number of dyneins are found in eukaryotes. Most of them are integral parts of axonemes and cause bending of eukaryotic cilia and flagella by sliding of adjacent outer doublet microtubules. Cytoplasmic dyneins are involved in retrograde vesicle transport, mitosis, cell migration, maintenance of the Golgi apparatus, and many other processes. The heavy

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

335

Fig. 6. Kinesin domain structure and associated proteins. Conventional kinesin (kinesin‐1 family) with two heavy and two light chains attached to a microtubule protofilament (schematically) and a single light chain (KAP3) of the heterotrimeric kinesin KIF3A/B‐KAP3 (kinesin‐2 family). Head, neck, and neck linker are modeled according to the crystal structure of RnKHC (PDB code 3KIN), the mostly a‐helical central part of the heavy chain is depicted schematically as a coiled‐coil (red). The stalk is interrupted by several hinges (not shown). Most of the stalk is suppressed; only small pieces at both ends are shown. Regions of the light chains (yellow) with homology to tetratricopeptide repeats (TPR) are modeled according to the known TPR domains of protein phosphatase 5 (PP5, PDB code 1A17; Das et al., 1998). Other parts of unknown structure are represented as spheres. In the case of the heterotrimeric kinesin, the two heavy chains are distinct but similar to each other and to the KHCs of kinesin‐1, and there is only one light chain, which is predicted to be mostly a‐helical. Picture reproduced from Mandelkow and Mandelkow, 2002, with kind permission of the publisher.

336

MARX ET AL.

chains are responsible for ATPase and motor activity, whereas regulation and specificity for diverse cargoes are mainly due to the combination of heavy chains with accessory proteins. Notably, single‐dynein heavy chains are capable of processive motion along microtubules (Sakakibara et al., 1999; Sale and Fox, 1988; Vale and Toyoshima, 1989). The dynein heavy chain consists of an N‐terminal domain that contains intermediate chain interaction and heavy chain oligomerization sites and a motor domain. The motor domain has a chain of six so‐called AAA modules of 35–40 kDa each (ATPases Associated with different cellular Activities; Mocz and Gibbons, 2001; Neuwald et al., 1999) and another, supposedly globular domain of 150–450 amino acids (depending on the isoform) at the C‐terminus (Fig. 7A). Between AAA modules 4 and 5, an insert of variable length appears that is predicted to form an antiparallel coiled‐coil with a short microtubule binding motif between the helices. Dynein has four ATP binding sites: four of the six AAA modules contain Walker A and B motifs and bind ATP; modules 5 and 6 are degenerate and have lost their P‐loop. ATP binding and hydrolysis by module 1 is absolutely essential for dynein’s motor activity, as shown by mutagenesis of the Walker A motif (Eshel, 1995) and domain‐specific photolysis by ultraviolet irradiation in the presence of vanadate (Lee‐Eiford et al., 1986). The role of the other nucleotide binding sites is less clear. Binding, but not hydrolysis of ATP to modules 2 to 4, also seems important and probably has regulatory functions. By electron microscopy imaging, the dynein heavy chain appears as a tripartite molecule, an elongated stem corresponding to the N‐terminal part, a ringlike structure or hollow sphere (head) consisting of several globular subdomains, and a stalk about 10–15 nm in length, corresponding to the insert between AAA modules 4 and 5 (Vallee et al., 2004). There is no high‐resolution structure of dynein available so far. Present atomic‐level structural information is restricted to homology modeling based on the sequence similarity with the AAA class of chaperone‐like ATPases (Mocz and Gibbons, 2001). According to this prediction, the AAA modules are a/b type structures with a core b‐sheet of five strands surrounded by about eight a‐helices (Fig. 7B). Three or four of them form a C‐terminal subdomain. The nucleotide binds in a cleft between the N‐ and the C‐terminal domain. Comparison with other ATPases of the AAA family suggests that the modules form a hexameric ring with the N‐ and C‐terminal subdomains of one module contacting the N‐terminal subdomain of the adjacent module. The attractive feature of this model is the cooperativity that suggests itself by the geometry of interaction between neighboring modules. A change of the angle between the subdomains of module 1 due to nucleotide processing should propagate throughout the ring and produce a substantial conformational change of the overall structure, thus

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

337

Fig. 7. Domain structure and homology model of the dynein heavy chain. (A) Bar diagram of the sequence of the b heavy chain of outer‐arm axonemal dynein with six AAA modules highlighted by different colors. The stalk domain between modules 4 and 5 contains a microtubule binding site flanked by two segments that are predicted to form a coiled‐coil. (B) Homology model of the globular part of the dynein heavy chain based on the structures of three AAA proteins (PDB code 1HN5; Mocz and Gibbons, 2001). Module 1 is shown with an ATP in the catalytic cleft. Hydrolysis of this ATP is essential for dynein’s motor activity. The figure has been prepared using Deep View Swiss‐PDB Viewer (Guex and Peitsch, 1997) and POVray for Windows (Persistence of Vision Pty. Ltd. 2004, Persistence of Vision Raytracer Version 3.5, retrieved from http://www. povray.org/).

explaining the communication between the ATPase site at one end of the ring and the change in microtubule affinity of the stalk at the opposite side. The model derived by homology with other AAA type ATPases certainly needs some elaboration. In electron micrographs, the head often seems to consist of seven or eight globular domains. It has been suggested that the C‐terminal domain following the last AAA module (King, 2000) or part of

338

MARX ET AL.

the stem sequence preceding the first module (Fan and Amos, 2001) account for the additional density. According to a recent electron microscopy study (Burgess et al., 2004), the seventh subdomain is most probably due to the C‐terminal amino acids. A large rearrangement of the head following release of hydrolysis products has not been observed. However, the analysis shows further details of the stem structure and motility. The stem consists of four sections: linker, neck, shaft, and base. On product release, the linker performs a swinging movement from a position roughly diametrical to the ring to a position rather tangential. This movement can be likened to the docking of the neck linker proposed for kinesin. Docking of dynein’s linker could be induced by relatively small conformational changes in the structure of the AAA modules resulting from nucleotide processing in module 1 and propagated throughout the ring. As in the other motor proteins, flexible elements must be present that can match the rearrangements occurring at different length‐ and timescales. In the case of dynein, the neck within the stem as well as the stalk seem to provide sufficient flexibility.

VII.

Summary and Outlook

In this review, we have mainly focused on the structural comparison of microtubule‐based motor proteins. Twenty years after the first identification of kinesin as a force‐generating protein (Brady, 1985; Vale et al., 1985), many details are known about kinesin and various kinesin‐related proteins due to X‐ray structure analyses of their motor domains, whereas the structure of dyneins, as well as the structure of the nonmotor domains of kinesin, are still elusive. A comparison of the known structures, including additional information gained from myosin and G‐proteins, may help to elucidate the mechanisms that are at work in these various motor molecules, which function as highly efficient chemomechanical energy converters. These efforts have already led to a certain understanding of the working principles at the level of plausibility, usually expressed in the form of ‘‘cartoon models’’ using switches, pistons, linkers, springs, and lever arms. Of course, our present understanding of motor proteins is not only based on structural data, but also on data obtained by many other methods like single‐molecule microscopy, biochemical and kinetic experiments, various types of spectroscopy, electron microscopy, and so forth, although these methods have not been covered in this review to the extent they would deserve. Substantial progress in our understanding of motor proteins can be expected from the extension of the presently available data about the

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

339

structure of the isolated motor domain to the full length protein and its complexes with the filamentous track, the cargo adapter molecules, and any other cofactors; and from a combination of these structural (i.e., mostly static) data with molecular dynamics. X‐ray crystallography seems rather restricted with regard to the analysis of large complexes, although this may change with the availability of new technologies, only to mention high‐throughput methods and the advent of new X‐ray sources like the free electron X‐ray laser. At present, structural analysis of larger complexes is limited to low‐ or medium‐resolution methods like FRET, NMR, and electron microscopy. Another approach to a better understanding of the mechanisms used by motor proteins like kinesin consists of using computational methods to simulate the dynamics of the motor molecules. Ideally, molecular dynamics simulation would start from first principles, using quantum mechanics to describe the whole system. However, this is far out of reach because of the sheer size of the problem: the relevant scales cover several orders of magnitude in space and many orders of magnitude in time, from the small‐scale events involved in hydrolysis of ATP to the large‐scale conformational changes occurring during the movement along microtubules. The popular cartoon models essentially use metaphors (like switch, spring, lever arm) to describe the inter‐relationship of the processes occurring at the different scales. A possible solution to this problem could be to use ab initio molecular dynamics for the active center and the nucleotide, and to switch to a coarse‐grained model to simulate large‐scale effects (Lattanzi and Maritan, 2004; Zheng and Doniach, 2003). The general problem for gaining a thorough understanding of motor proteins seems to be related to other big problems of structural molecular biology: the prediction of protein folding from first principles using only the sequence of amino acids and the prediction of the interaction of proteins with other proteins or with small ligands like cofactors and inhibitors. Since motor proteins are involved in many physiological processes, better models can be expected to be helpful in the search for new drugs (cf specific inhibition of the mitotic motor Eg5 by monastrol and other small organic compounds that could serve as potential antitumor drugs; Mayer et al., 1999; Sakowicz et al., 2004). Because of their ability to exert molecular control at the nanometer scale, motor proteins and related biomechanical proteins lend themselves as natural models for applications in nanotechnology. Biomolecular motors have been manipulated and used for artificial tasks on nano‐ and micro‐scales, and they serve as models for engineered systems and biomimetic devices (Schmidt and Montemagno, 2004). Progress in bionanotechnology will certainly benefit from future advances in molecular biology of motor proteins.

340

MARX ET AL.

References Al‐Bassam, J., Cui, Y., Klopfenstein, D., Carragher, B. O., Vale, R. D., and Milligan, R. A. (2003). Distinct conformations of the kinesin Unc104 neck regulate a monomer to dimer motor transition. J. Cell Biol. 163, 743–753. Anson, M., Geeves, M. A., Kurzawa, S. E., and Manstein, D. J. (1996). Myosin motors with artificial lever arms. EMBO J. 15, 6069–6074. Asenjo, A. B., Krohn, N., and Sosa, H. (2003). Configuration of the two kinesin motor domains during ATP hydrolysis. Nat. Struct. Biol. 10, 836–842. Berger, B., Wilson, D. B., Wolf, E., Tonchev, T., Milla, M., and Kim, P. S. (1995). Predicting coiled‐coils by use of pairwise residue correlations. Proc. Natl. Acad. Sci. USA 92, 8259–8263. Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T. N., Weissig, H., Shindyalov, I. N., and Bourne, P. E. (2000). The Protein Data Bank. Nucleic Acids Res. 28, 235–242. Brady, S. T. (1985). A novel brain ATPase with properties expected for the fast axonal transport motor. Nature 317, 73–75. Burgess, S. A., Walker, M. L., Sakakibara, H., Oiwa, K., and Knight, P. J. (2004). The structure of dynein‐c by negative stain electron microscopy. J. Struct. Biol. 146, 205–216. Conibear, P. B., Bagshaw, C. R., Fajer, P. G., Kovacs, M., and Malnasi‐Csizmadia, A. (2003). Myosin cleft movement and its coupling to actomyosin dissociation. Nat. Struct. Biol. 10, 831–835. Coureux, P. D., Wells, A. L., Menetrey, J., Yengo, C. M., Morris, C. A., Sweeney, H. L., and Houdusse, A. (2003). A structural state of the myosin V motor without bound nucleotide. Nature 425, 419–423. Das, A. K., Cohen, P. W., and Barford, D. (1998). The structure of the tetratricopeptide repeats of protein phosphatase 5: Implications for TPR‐mediated protein‐protein interactions. EMBO J. 17, 1192–1199. Eshel, D. (1995). Functional dissection of the dynein motor domain. Cell Motil. Cytoskeleton 32, 133–135. Fan, J., and Amos, L. A. (2001). Antibodies to cytoplasmic dynein heavy chain map the surface and inhibit motility. J. Mol. Biol. 307, 1317–1327. Garcia‐Saez, I., Yen, T., Wade, R. H., and Kozielski, F. (2004). Crystal structure of the motor domain of the human kinetochore protein CENP‐E. J. Mol. Biol. 340, 1107–1116. Guex, N., and Peitsch, M. C. (1997). SWISS‐MODEL and the Swiss‐PdbViewer: An environment for comparative protein modeling. Electrophoresis 18, 2714–2723. Gulick, A. M., Song, H., Endow, S. A., and Rayment, I. (1998). X‐ray crystal structure of the yeast Kar3 motor domain complexed with Mg.ADP to 2.3 A resolution. Biochemistry 37, 1769–1776. Hirose, K., Lowe, J., Alonso, M., Cross, R. A., and Amos, L. A. (1999). Congruent docking of dimeric kinesin and ncd into three‐dimensional electron cryomicroscopy maps of microtubule‐motor ADP complexes. Mol. Biol. Cell 10, 2063–2074. Hoenger, A., Sack, S., Thormahlen, M., Marx, A., Muller, J., Gross, H., and Mandelkow, E. (1998). Image reconstructions of microtubules decorated with monomeric and dimeric kinesins: Comparison with x‐ray structure and implications for motility. J. Cell Biol. 141, 419–430. Hoenger, A., Thormahlen, M., Diaz‐Avalos, R., Doerhoefer, M., Goldie, K. N., Muller, J., and Mandelkow, E. (2000). A new look at the microtubule binding patterns of dimeric kinesins. J. Mol. Biol. 297, 1087–1103.

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

341

Holmes, K. C., Angert, I., Kull, F. J., Jahn, W., and Schroder, R. R. (2003). Electron cryo‐ microscopy shows how strong binding of myosin to actin releases nucleotide. Nature 425, 423–427. Holmes, K. C., and Geeves, M. A. (2000). The structural basis of muscle contraction. Philos. Trans. R. Soc. Lond. B Biol. Sci. 355, 419–431. Houdusse, A., Szent‐Gyorgyi, A. G., and Cohen, C. (2000). Three conformational states of scallop myosin S1. Proc. Natl. Acad. Sci. USA 97, 11238–11243. Kabsch, W., and Sander, C. (1983). Dictionary of protein secondary structure: Pattern recognition of hydrogen‐bonded and geometrical features. Biopolymers 22, 2577–2637. Kikkawa, M., Okada, Y., and Hirokawa, N. (2000). 15 A resolution model of the monomeric kinesin motor, KIF1A. Cell 100, 241–252. Kikkawa, M., Sablin, E. P., Okada, Y., Yajima, H., Fletterick, R. J., and Hirokawa, N. (2001). Switch‐based mechanism of kinesin motors. Nature 411, 439–445. King, S. M. (2000). AAA domains and organisation of the dynein motor unit. J. Cell Sci. 113(Pt 14), 2521–2526. Klopfenstein, D. R., Tomishige, M., Stuurman, N., and Vale, R. D. (2002). Role of phosphatidylinositol(4,5)bisphosphate organisation in membrane transport by the Unc104 kinesin motor. Cell 109, 347–358. Klumpp, L. M., Mackey, A. T., Farrell, C. M., Rosenberg, J. M., and Gilbert, S. P. (2003). A kinesin switch I arginine to lysine mutation rescues microtubule function. J. Biol. Chem. 278, 39059–39067. Kollmar, M., Durrwang, U., Kliche, W., Manstein, D. J., and Kull, F. J. (2002). Crystal structure of the motor domain of a class‐I myosin. EMBO J. 21, 2517–2525. Kozielski, F., Arnal, I., and Wade, R. H. (1998). A model of the microtubule‐kinesin complex based on electron cryomicroscopy and X‐ray crystallograhpy. Curr Biol. 8, 191–198. Kozielski, F., De Bonis, S., Burmeister, W. P., Cohen‐Addad, C., and Wade, R. H. (1999). The crystal structure of the minus‐end‐directed microtubule motor protein ncd reveals variable dimer conformations. Structure Fold. Des. 7, 1407–1416. Kozielski, F., Sack, S., Marx, A., Thorma¨ hlen, M., Scho¨ nbrunn, E., Biou, V., Thompson, A., Mandelkow, E. M., and Mandelkow, E. (1997). The crystal structure of dimeric kinesin and implications for microtubule‐dependent motility. Cell 91, 985–994. Kozielski, F., Svergun, D., Zaccai, G., Wade, R. H., and Koch, M. H. (2001). The overall conformation of conventional kinesins studied by small angle x‐ray and neutron scattering. J. Biol. Chem. 276, 1267–1275. Kull, F. J., Sablin, E. P., Lau, R., Fletterick, R. J., and Vale, R. D. (1996). Crystal structure of the kinesin motor domain reveals a structural similarity to myosin. Nature 380, 550–555. Kull, F. J., Vale, R. D., and Fletterick, R. J. (1998). The case for a common ancestor: Kinesin and myosin motor proteins and G proteins. J. Muscle Res. Cell Motil. 19, 877–886. Laskowski, R. A., Mac Arthur, M. W., Moss, D. S., and Thornton, J. M. (1993). PROCHECK: A program to check the stereochemical quality of protein structures. J. Appl. Cryst. 26, 283–291. Lattanzi, G., and Maritan, A. (2004). Coarse grained models: The kinetics of motor proteins. Comput. Mater. Sci. 30, 172–179. Lawrence, C. J., Dawe, R. K., Christie, K. R., Cleveland, D. W., Dawson, S. C., Endow, S. A., Goldstein, L. S., Goodson, H. V., Hirokawa, N., Howard, J., Malmberg, R. L., McIntosh, J. R., Miki, H., Mitchison, T. J., Okada, Y., Reddy, A. S., Saxton, W. M., Schliwa, M., Scholey, J. M., Vale, R. D., Walczak, C. E., and Wordeman, L. (2004). A standardised kinesin nomenclature. J. Cell Biol. 167, 19–22.

342

MARX ET AL.

Lee‐Eiford, A., Ow, R. A., and Gibbons, I. R. (1986). Specific cleavage of dynein heavy chains by ultraviolet irradiation in the presence of ATP and vanadate. J. Biol. Chem. 261, 2337–2342. Leibler, S., and Huse, D. A. (1993). Porters versus rowers: A unified stochastic model of motor proteins. J. Cell Biol. 121, 1357–1368. Lowe, J., Li, H., Downing, K. H., and Nogales, E. (2001). Refined structure of ab‐tubulin at 3.5 A resolution. J. Mol. Biol. 313, 1045–1057. Mackey, A. T., and Gilbert, S. P. (2003). The ATPase cross‐bridge cycle of the Kar3 motor domain: implications for single head motility. J. Biol. Chem. 278, 3527–3535. Mandelkow, E., and Mandelkow, E. M. (2002). Kinesin motors and disease. Trends Cell Biol. 12, 585–591. Mayer, T. U., Kapoor, T. M., Haggarty, S. J., King, R. W., Schreiber, S. L., and Mitchison, T. J. (1999). Small molecule inhibitor of mitotic spindle bipolarity identified in a phenotype‐based screen. Science 286, 971–974. Mocz, G., and Gibbons, I. R. (2001). Model for the motor component of dynein heavy chain based on homology to the AAA family of oligomeric ATPases. Structure (Camb.) 9, 93–103. Neuwald, A. F., Aravind, L., Spouge, J. L., and Koonin, E. V. (1999). AAAþ: A class of chaperone‐like ATPases associated with the assembly, operation, and disassembly of protein complexes. Genome Res. 9, 27–43. Nitta, R., Kikkawa, M., Okada, Y., and Hirokawa, N. (2004). KIF1A alternately uses two loops to bind microtubules. Science 305, 678–683. Nogales, E., Wolf, S. G., and Downing, K. H. (1998). Structure of the ab tubulin dimer by electron crystallography. Nature 391, 199–203. Ogawa, T., Nitta, R., Okada, Y., and Hirokawa, N. (2004). A common mechanism for microtubule destabilizers‐M type kinesins stabilize curling of the protofilament using the class‐specific neck and loops. Cell 116, 591–602. Okada, Y., and Hirokawa, N. (2000). Mechanism of the single‐headed processivity: Diffusional anchoring between the K‐loop of kinesin and the C terminus of tubulin. Proc. Natl. Acad. Sci. USA 97, 640–645. Onishi, H., Kojima, S., Katoh, K., Fujiwara, K., Martinez, H. M., and Morales, M. F. (1998). Functional transitions in myosin: Formation of a critical salt‐bridge and transmission of effect to the sensitive tryptophan. Proc. Natl. Acad. Sci. USA 95, 6653–6658. Reubold, T. F., Eschenburg, S., Becker, A., Kull, F. J., and Manstein, D. J. (2003). A structural model for actin‐induced nucleotide release in myosin. Nat. Struct. Biol. 10, 826–830. Rice, S., Lin, A. W., Safer, D., Hart, C. L., Naber, N., Carragher, B. O., Cain, S. M., Pechatnikova, E., Wilson‐Kubalek, E. M., Whittaker, M., Pate, E., Cooke, R., Taylor, E. W., Milligan, R. A., and Vale, R. D. (1999). A structural change in the kinesin motor protein that drives motility. Nature 402, 778–784. Sablin, E. P., Case, R. B., Dai, S. C., Hart, C. L., Ruby, A., Vale, R. D., and Fletterick, R. J. (1998). Direction determination in the minus‐end‐directed kinesin motor ncd. Nature 395, 813–816. Sablin, E. P., Kull, F. J., Cooke, R., Vale, R. D., and Fletterick, R. J. (1996). Crystal structure of the motor domain of the kinesin‐related motor ncd. Nature 380, 555–559. Sack, S., Kull, F. J., and Mandelkow, E. (1999). Motor proteins of the kinesin family: Structures, variations, and nucleotide binding sites. Eur. J. Biochem. 262, 1–11.

THE STRUCTURE OF MICROTUBULE MOTOR PROTEINS

343

Sack, S., Mu¨ ller, J., Marx, A., Thorma¨ hlen, M., Mandelkow, E. M., Brady, S. T., and Mandelkow, E. (1997). X‐ray structure of motor and neck domains from rat brain kinesin. Biochemistry 36, 16155–16165. Sakakibara, H., Kojima, H., Sakai, Y., Katayama, E., and Oiwa, K. (1999). Inner‐arm dynein c of Chlamydomonas flagella is a single‐headed processive motor. Nature 400, 586–590. Sakowicz, R., Finer, J. T., Beraud, C., Crompton, A., Lewis, E., Fritsch, A., Lee, Y., Mak, J., Moody, R., Turincio, R., Chabala, J. C., Gonzales, P., Roth, S., Weitman, S., and Wood, K. W. (2004). Antitumor activity of a kinesin inhibitor. Cancer Res. 64, 3276–3280. Sale, W. S., and Fox, L. A. (1988). Isolated b‐heavy chain subunit of dynein translocates microtubules in vitro. J. Cell Biol. 107, 1793–1797. Schmidt, J. J., and Montemagno, C. D. (2004). Bionanomechanical systems. Annual Rev. Materials Res. 34, 315–337. Seeberger, C., Mandelkow, E., and Meyer, B. (2000). Conformational preferences of a synthetic 30mer peptide from the interface between the neck and stalk regions of kinesin. Biochemistry 39, 12558–12567. Shipley, K., Hekmat‐Nejad, M., Turner, J., Moores, C., Anderson, R., Milligan, R., Sakowicz, R., and Fletterick, R. (2004). Structure of a kinesin microtubule depolymerization machine. EMBO J. 23, 1422–1432. Sindelar, C. V., Budny, M. J., Rice, S., Naber, N., Fletterick, R., and Cooke, R. (2002). Two conformations in the human kinesin power stroke defined by x‐ray crystallography and EPR spectroscopy. Nat. Struct. Biol. 9, 844–848. Skiniotis, G., Surrey, T., Altmann, S., Gross, H., Song, Y. H., Mandelkow, E., and Hoenger, A. (2003). Nucleotide‐induced conformations in the neck region of dimeric kinesin. EMBO J. 22, 1518–1528. Song, Y. H., Marx, A., Muller, J., Woehlke, G., Schliwa, M., Krebs, A., Hoenger, A., and Mandelkow, E. (2001). Structure of a fast kinesin: Implications for ATPase mechanism and interactions with microtubules. EMBO J. 20, 6213–6225. Steinberg, G., and Schliwa, M. (1996). Characterisation of the biophysical and motility properties of kinesin from the fungus. Neurospora crassa. J. Biol. Chem. 271, 7516–7521. Stock, M. F., Guerrero, J., Cobb, B., Eggers, C. T., Huang, T. G., Li, X., and Hackney, D. D. (1999). Formation of the compact conformer of kinesin requires a COOH‐ terminal heavy chain domain and inhibits microtubule‐stimulated ATPase activity. J. Biol. Chem. 274, 14617–14623. Tomishige, M., Klopfenstein, D. R., and Vale, R. D. (2002). Conversion of Unc104/ KIF1A kinesin into a processive motor after dimerisation. Science 297, 2263–2267. Tomishige, M., and Vale, R. D. (2000). Controlling kinesin by reversible disulfide cross‐ linking: Identifying the motility‐producing conformational change. J. Cell Biol. 151, 1081–1092. Tripet, B., Vale, R. D., and Hodges, R. S. (1997). Demonstration of coiled‐coil interactions within the kinesin neck region using synthetic peptides. J. Biol. Chem. 272, 8946–8956. Tsiavaliaris, G., Fujita‐Becker, S., and Manstein, D. J. (2004). Molecular engineering of a backwards‐moving myosin motor. Nature 427, 558–561. Turner, J., Anderson, R., Guo, J., Beraud, C., Fletterick, R., and Sakowicz, R. (2001). Crystal structure of the mitotic spindle kinesin Eg5 reveals a novel conformation of the neck‐linker. J. Biol. Chem. 276, 25496–25502.

344

MARX ET AL.

Uyeda, T. Q., Abramson, P. D., and Spudich, J. A. (1996). The neck region of the myosin motor domain acts as a lever arm to generate movement. Proc. Natl. Acad. Sci. USA 93, 4459–4464. Vale, R. D., and Fletterick, R. J. (1997). The design plan of kinesin motors. Ann. Rev. Cell Dev. Biol. 13, 745–777. Vale, R. D., Reese, T. S., and Sheetz, M. P. (1985). Identification of a novel force‐ generating protein, kinesin, involved in microtubule‐based motility. Cell 42, 39–50. Vale, R. D., and Toyoshima, Y. Y. (1989). Microtubule translocation properties of intact and proteolytically digested dyneins from Tetrahymena cilia. J. Cell Biol. 108, 2327–2334. Vallee, R. B., Williams, J. C., Varma, D., and Barnhart, L. E. (2004). Dynein: An ancient motor protein involved in multiple modes of transport. J. Neurobiol. 58, 189–200. Veigel, C., Coluccio, L. M., Jontes, J. D., Sparrow, J. C., Milligan, R. A., and Molloy, J. E. (1999). The motor protein myosin‐I produces its working stroke in two steps. Nature 398, 530–533. Vinogradova, M. V., Reddy, V. S., Reddy, A. S., Sablin, E. P., and Fletterick, R. J. (2004). Crystal structure of kinesin regulated by Ca2þ/calmodulin. J. Biol. Chem. 279, 23504–23509. Wendt, T. G., Volkmann, N., Skiniotis, G., Goldie, K. N., Muller, J., Mandelkow, E., and Hoenger, A. (2002). Microscopic evidence for a minus‐end‐directed power stroke in the kinesin motor ncd. EMBO J. 21, 5969–5978. Woehlke, G., Ruby, A. K., Hart, C. L., Ly, B., Hom‐Booher, N., and Vale, R. D. (1997). Microtubule interaction site of the kinesin motor. Cell 90, 207–216. Wood, K. W., Sakowicz, R., Goldstein, L. S., and Cleveland, D. W. (1997). CENP‐E is a plus end‐directed kinetochore motor required for metaphase chromosome alignment. Cell 91, 357–366. Xiang, X., and Plamann, M. (2003). Cytoskeleton and motor proteins in filamentous fungi. Curr. Opin. Microbiol. 6, 628–633. Yan, Y., Sardana, V., Xu, B., Homnick, C., Halczenko, W., Buser, C. A., Schaber, M., Hartman, G. D., Huber, H. E., and Kuo, L. C. (2004). Inhibition of a mitotic motor protein: Where, how, and conformational consequences. J. Mol. Biol. 335, 547–554. Yun, M., Bronner, C. E., Park, C. G., Cha, S. S., Park, H. W., and Endow, S. A. (2003). Rotation of the stalk/neck and one head in a new crystal structure of the kinesin motor protein, Ncd. EMBO J. 22, 5382–5389. Yun, M., Zhang, X., Park, C. G., Park, H. W., and Endow, S. A. (2001). A structural pathway for activation of the kinesin motor ATPase. EMBO J. 20, 2611–2618. Zheng, W., and Doniach, S. (2003). A comparative study of motor‐protein motions by using a simple elastic‐network model. Proc. Natl. Acad. Sci. USA 100, 13253–13258.

ROTARY MOLECULAR MOTORS By STEPHAN WILKENS Department of Biochemistry, University of California, Riverside, Riverside, California 92521

I.

II.

III.

IV.

Introduction: The F‐, V‐, and A‐ATPases and Their Function in the Cell. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. F‐ATPase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. V‐ATPase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. A‐ATPase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overall Structural Features of the F‐, V‐, and A‐ATPases . . . . . . . . . . . . . . . . . . . A. Structure of the F‐ATPase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Structure of the V‐ATPase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Structure of the A‐ATPase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanistic Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Rotational Catalysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. The Classic Rotation Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. The Functional Elements: Rotor and Stator . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Torque Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Motor Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

346 348 350 352 352 352 357 361 362 362 363 363 368 371 372 372

Abstract The F‐, V‐, and A‐adenosine triphosphatases (ATPases) represent a family of evolutionarily related ion pumps found in every living cell. They either function to synthesize adenosine triphosphate (ATP) at the expense of an ion gradient or they act as primary ion pumps establishing transmembrane ion motive force at the expense of ATP hydrolysis. The A‐, F‐, and V‐ATPases are rotary motor enzymes. Synthesis or hydrolysis of ATP taking place in the three catalytic sites of the membrane extrinsic domain is coupled to ion translocation across the single ion channel in the membrane‐bound domain via rotation of a central part of the complex with respect to a static portion of the enzyme. This chapter reviews recent progress in the structure determination of several members of the family of F‐, A‐, and V‐ATPases and our current understanding of the rotary mechanism of energy coupling.

ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71009-8

345

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

346

WILKENS

I.

Introduction: The F‐, V‐, and A‐ATPases and Their Function in the Cell

The F‐, V‐, and A‐ATPases constitute a family of ATP hydrolysis‐driven ion pumps which are found in Archaea, eubacteria, simple eukaryotes such as yeast, and higher eukaryotes including plants and mammals. The family of ion pumps is divided into three subfamilies: the F‐ATPases (which function mainly as ATP synthases), the vacuolar ATPases (which function solely as ATP hydrolysis‐driven ion pumps) and the Archaeal A‐ type ATPases (whose function can be either in the direction of ATP synthesis or hydrolysis). All three members of the family are evolutionarily related, and it is believed that the three subfamilies have arisen from a common ancestor. The F/V/A‐ATPases are multisubunit complexes containing between eight (for the bacterial F‐ATPase) and 14 (for the eukaryotic V‐ATPase) different subunits, some of which are present in multiple copies. The subunits are organized in two parts: a water‐soluble domain called F1, V1, or A1 and a membrane bound domain called F0, V0, or A 0. Consequently, the holoenzymes are called F1F0, V1V0, or A1A 0. Based on current knowledge, all members of the family share a common mechanism of energy coupling. Hydrolysis or synthesis of ATP takes place on the membrane extrinsic F1, V1, or A1 domains. This process is coupled over a distance of more than 100 A˚ to ion translocation through the membrane‐ bound F0, V0, or A 0 domains. This long‐range energy coupling occurs via the so‐called stalk domain, an assembly of several polypeptides that constitute the structural and functional connection between the membrane‐ and water‐soluble domains. Table I summarizes the subunit composition of the F1F0‐ATP synthase from Escherichia coli, the vacuolar proton pumping ATPase from yeast, and the archaeal A‐ATPase from Thermoplasma acidophilum. Structural models for these three enzyme complexes are shown in Fig. 1. A unique feature of the F/V/A‐ATPases is that they are rotary molecular motor enzymes. This has been shown by experiment for members of the F‐ and V‐ATPase subfamilies and is generally assumed to be true for the closely related A‐ATPases as well. The two enzymatic processes, ATP synthesis/hydrolysis and ion translocation, are coupled via a rotational motion of a central domain of the complex (the rotor) relative to a static domain (the stator). The A‐, F‐, and V‐ATPases represent the smallest rotary motors found in the living cell so far. Most of what we know about the structure and mechanism of these microscopic energy converters comes from studies conducted with the F‐ATPase. In the following review, current structural knowledge for all three members of the family of F‐, V‐,

347

ROTARY MOLECULAR MOTORS

Table I Subunit Composition of the F1F0‐ATP Synthase, the Eukaryotic Vacuolar ATPase and the Archaeal A‐ATPase Molecular Weight (and Stoichiometry)

Subunit

Yeast Gene

Yeast

Coated Vesicleb

A‐ATPasea Thermoplasma acidophilum

A B C D Ec F Gd He a c c 0f c 00 f,g M9, eh d Ac45

VMA1 VMA2 VMA5 VMA8 VMA4 VMA7 VMA10 VMA13 VPH1/STV1 VMA3 VMA11 VMA16 VMA9 VMA6

68 58 44 29 26 13 13 54 96 16 17 23 9 40 —

68 (3) 57 (3) 44 (1) 28 (1) 26 (1) 13 (1) 14 (2) 54/56 (2) 96 (1) 16 (4–5) 17 19 (1) ? 38 (1) 45 (1)

66 53 — 25 22 12 12a — 72a 8a — — — 41a —

V‐ATPase

F‐ATPase Escherichia coli b (3) a (3) — g (1) d (1)  (1) b (2) — a (1) c (10) — — — — —

a The subunit nomenclature for the A‐ATPase is K for the proteolipid, I for the V‐ and F‐ATPase a‐subunit, and C for the V‐ATPase subunit d. The small polypeptide called H in the A‐ATPase is probably the homologue of the V‐ATPase G‐subunit. b The subunit stoichiometry for the coated vesicle enzyme has been determined by quantitative amino acid analysis (Arai et al., 1988; Xu et al., 1999). c A stoichiometry of three copies of subunit E has been reported for the bovine brain– coated vesicle enzyme based on densitometry of Coomassie‐stained SDS polyacrylamide gels (Xie, 1996). d Three copies of G have been reported for the yeast enzyme (Supekova et al., 1996) and suggested for the complex from Manduca sexta (Wieczorek et al., 2000). e Isoforms have been reported for the mammalian V‐ATPase for subunit H (Xie et al., 1994), G (Crider et al., 1997), and most other peripheral, noncatalytic subunits (see Smith et al., 2003). f The presence of each one proteolipid c0 and c 00 (Powell et al., 2000) and two c 00 (Gibson et al., 2002) have been reported for the yeast enzyme. g It is assumed that this subunit, formerly called the 19‐kDa subunit, is the homologue to Vma16p. h The subunit has been confirmed for the insect (Merzendorfer et al., 1999) and chromaffin granule enzyme (Ludwig et al., 1998) and has recently been found in the yeast enzyme (Sambade and Kane, 2004). The subunit compositions of the F‐ATPase from the bacterium Escherichia coli, the vacuolar ATPase from yeast and bovine brain clathrin‐coated vesicles, and the A‐ATPase from the Archaeon Thermoplasma acidophilum are listed. Molecular masses are calculated from the amino acid sequence where available.

348

WILKENS

Fig. 1. Working models of the F‐, V‐, and A‐ATPases. Model of the subunit arrangement in the (A) F1F0‐ATP synthase from Escherichia coli, (B) vacuolar ATPase from bovine brain clathrin‐ coated vesicles, and (C) A1A0‐ATPase from Thermoplasma acidophilum. The catalytic domain is in blue, the rotor domain is in green, and the stator domain is in orange.

and A‐ATPases will be presented, but discussion of the general rotary mechanism will be based on the findings obtained for the F‐ATPase.

A.

F‐ATPase

The F‐ATPase or F1F0‐ATP synthase is found in the inner membrane of mitochondria, the thylakoid membrane of chloroplasts and the plasma membrane of bacteria where it is responsible for synthesis of ATP (Boyer, 1997; Fillingame, 1997; Junge et al., 1997; Penefsky and Cross, 1991; Senior, 1988; Yoshida et al., 2001). Under physiological conditions, synthesis of ATP from ADP and inorganic phosphate is an endergonic process. The required energy, as first postulated in the chemiosmotic theory by Peter Mitchell (1961), is provided by the electrochemical gradient across the lipid bilayer, which is generated and maintained during electron transport by the enzymes of the respiratory chain or by photosynthesis. In the F‐ ATPase, chemiosmotic coupling is reversible, and in certain bacteria (such as E. coli) the F1F0‐ATP synthase can function as an ion‐pumping ATPase to establish an electrochemical gradient across the plasma membrane, which in turn can be used by the organism to drive secondary transport processes. Membrane‐bound F1F0‐ATP synthase is easily dissociated into a water‐ soluble ‘‘coupling factor’’ F1 and a membrane‐bound F0. The isolated F1

ROTARY MOLECULAR MOTORS

349

can be a highly active MgATPase, and the membrane‐embedded F0 catalyzes passive translocation of protons across the membrane along a chemical gradient. For example, the dissociation of F1F0 into F1 and F0 can be induced by treatment of bacterial or mitochondrial inner membranes or thylakoid membranes with low ionic strength buffer in the presence of magnesium chelators such as ethylenediamine tetraactic acid (EDTA). The dissociation of F1F0 is not supposed to occur under physiological conditions because the free F1 would quickly hydrolyze the existing pool of ATP to ADP and inorganic phosphate and the free F0 would lead to a rapid collapse of the existing proton motive force. In turn, this would prevent the remaining intact ATP synthase molecules from synthesizing ATP. In bacteria, the activity of membrane‐dissociated F1 is partly inhibited by the ‐ subunit and in mitochondria, an inhibitor protein (IF1) protects against the damaging effects of dissociated F1 (Cabezon et al., 2002). The structure and catalytic mechanism of the F‐ATPase has been under intense investigation for more than 40 years, and much of what we are now learning about the vacuolar‐ and archaeal ATPase is based on experiments originally designed for the study of the related F‐type ATPase. Over the past 10 years, several atomic resolution crystal structures for F1‐ ATPase domains from a number of species have been determined, and these structures have helped to establish the catalytic mechanism of the ATPase domain. What is still missing is a high‐resolution structure of the intact ATP synthase, and this deficiency limits our current understanding of the working of the membrane domain of the complex. Synthesis or hydrolysis of ATP is catalyzed in the nucleotide binding sites that are located mostly on the b‐subunits, at the interface to the a‐subunits (Abrahams et al., 1994). The a‐subunits contain so‐called noncatalytic nucleotide binding sites, which are normally filled with nucleotide under physiological conditions. The function of the noncatalytic nucleotide binding sites is not fully understood, but it is thought that they might be important for the assembly and stability of the complex. Each nucleotide binding site is composed of an adenine binding pocket and a phosphate binding loop with the well‐known P‐loop or Walker A motif GXXXXGKT/S. According to the ‘‘binding change’’ model of cooperative catalysis (reviewed in Boyer, 1993), synthesis or hydrolysis of ATP occurs on the three catalytic sites in a sequential fashion. All three sites undergo a series of conformational changes that are out of phase by 120 degrees. The nucleotide binding sites can be either in a tight, loose, or open (empty) state, and interconversion of these states is coupled to translocation of protons (or sodium ions) through the membrane‐bound portion of the enzyme.

350

WILKENS

It is now well established that the F‐ATPase is a rotary motor enzyme (Diez et al., 2004; Duncan et al., 1995; Nishio et al., 2002; Noji et al., 1997, 1999; Sabbert et al., 1996; Zhou et al., 1997). Translocation of protons along the interface of subunits c and a drives rotation of the ring of c‐subunits together with the g‐ and ‐subunits against the a3b3 catalytic core. The proton gradient‐induced rotation of the g‐subunit inside the a3b3 hexamer allows binding of MgADP and inorganic phosphate at the catalytic sites and subsequent release of product ATP. The peripheral stalk composed of the d‐ and b‐subunits holds the F1 and F0 in the correct spatial arrangement for energy coupling to occur.

B. þ

V‐ATPase

A vacuolar H ‐ATPase (V1V0‐ or V‐ATPase) is found in the membranes of a wide variety of intracellular compartments such as clathrin‐coated vesicles, chromaffin granules, endosomes, lysosomes, synaptic vesicles, Golgi‐derived vesicles, the tonoplast of plants, and the yeast vacuole. The proton‐pumping action of the vacuolar ATPase plays a vital role in a large number of intra‐ and intercellular processes, including receptor‐mediated endocytosis, protein trafficking, pH maintenance, storage of metabolites, and neurotransmitter release (Finbow and Harrison, 1997; Futai et al., 2000; Nelson and Harvey, 1999; Nishi and Forgac, 2002; Stevens and Forgac, 1997). In higher eukaryotes, V‐type ATPases are also found in the plasma membrane of polarized cells like osteoclasts, renal epithelial cells, insect epithelial cells, and amphibian skin cells. Here, the V‐ATPase pumps protons into the extracellular space. Acidification of the enclosed space between the ruffled membrane of osteoclasts and the bone surface plays an important role in bone resorption and remodeling and, in the kidney, a V‐ATPase in the plasma membrane of renal intercalated cells functions in blood deacidification. Defects in the human osteoclast or kidney V‐ATPases can lead to osteoporosis or renal tubular acidosis, respectively (reviewed in, e.g., Alper, 2002). In the cell, the vacuolar ATPase functions exclusively as an ATP hydrolysis‐driven proton pump. This functional preference of the vacuolar ATPase seems not to be due to a fundamental structural difference compared with the ATP synthase because it has been shown that the V‐ATPase can be reversed to synthesize ATP in vitro, albeit with very low efficiency (Hirata et al., 2000). F‐ and V‐ATPase are evolutionarily related (Gogarten et al., 1989; Nelson and Taiz, 1989). This relationship between F‐ and V‐ATPases was first described based on the similarity of the amino acid sequence of the V‐ATPase A‐ and B‐subunits and the F‐ATPase b‐ and a‐subunits,

ROTARY MOLECULAR MOTORS

351

respectively (Bowman et al., 1988; Zimniak et al., 1988). The proteolipids are the most conserved subunits between F‐ and V‐ATPases, even though the proteolipids of the V‐ATPase contain four predicted membrane‐ spanning helices compared with the two helix F‐ATPase subunit c. It is believed that the V‐ATPase subunit c has evolved from a common ancestor proteolipid by gene duplication (Hilario and Gogarten, 1998; Mandel et al., 1988). The yeast V‐ATPase A‐subunit is expressed as an intein‐containing polypeptide (Kane et al., 1990). After translation, the intein is excised by an autocatalytic process resulting in the mature subunit A and the excised polypeptide. Inteins from the eukaryotic V‐ATPase A‐subunit can function as nucleases. The larger size of the A subunit compared with the F‐ATPase b‐subunit is due to an approximately 100–amino acid insertion near the N‐terminus of the polypeptide. This so‐called nonhomologous region is highly conserved among V‐ATPases, and it has been suggested that this part of the A‐subunit might be involved in energy coupling (Shao and Forgac, 2004; Shao et al., 2003). Enzymes that are structurally related to the eukaryotic V‐ATPase are also found in certain eubacteria (Speelmans et al., 1994; Takase et al., 1994; Yokoyama et al., 1990). Based on nucleotide sequence analysis, it is believed that these bacterial V‐like ATPases have been introduced into the eubacteria via horizontal gene transfer from Archaea (Hilario and Gogarten, 1993, 1998). The subunit composition of the bacterial V‐like ATPase is indeed more similar to the archaeal A‐ATPase than to the eukaryotic V‐ATPase, and we will therefore treat the bacterial V‐ATPase–like enzyme together with the archaeal A‐ATPase (see below). In the following, we will use the name V‐ATPase only for the eukaryotic enzyme, and we will call the bacterial enzyme the A/V‐type ATPase as suggested by Hilario and Gogarten (1998). A unique feature of the vacuolar ATPase is its nutrient‐dependent dissociation and reassociation in vivo. The process has been described for the yeast (Kane, 1995) and insect (Sumner et al., 1995) enzymes. It was shown that the dissociation is reversible, that the dissociated V1 cannot hydrolyze MgATP (Gra¨ f et al., 1996; Parra et al., 2000; Xie and Stone, 1988; Zhang et al., 2003), and that the free V0 is impermeable to protons (Zhang et al., 1992). Dissociation of the F1F0‐ATP synthase, on the other hand, results in a potentially active F1‐ATPase and a proton‐conducting F0. The dissociation of the V‐ATPase is employed by the organism to preserve ATP and at the same time conserve the existing transmembrane proton gradient generated by the intact enzyme in times when nutrient supply is limited. Both V0 (Zhang et al., 1992) and V1 (Gra¨ f et al., 1996; Parra et al., 2000; Zhang et al., 2003) domains have been purified and characterized.

352

WILKENS

Besides its function as a proton pump, the V‐ATPase has been shown to interact with a variety of other cellular protein components such as ectoapyrase (Zhang et al., 2000), the yeast RAVE complex (Smardon et al., 2002), and HIV Nef and AP–2 (Geyer et al., 2002; Lu et al., 1998).

C. A‐ATPase The A‐ATPases, which are found exclusively in Archaea, are chimeric enzymes. Although members of this subfamily are structurally more similar to the eukaryotic vacuolar ATPases (Denda et al., 1988), they probably functions as (reversible) ATP synthases much like the F‐ATPases (Gru¨ ber et al., 2001; Mu¨ ller and Gru¨ ber, 2003; Mu¨ ller et al., 1999). Due to the difficulty of obtaining sufficient quantities of the protein for biochemical and structural studies and due to the enzyme’s limited stability, few members of the A‐ATPase subfamily have been characterized with respect to their enzymatic and structural properties. Subunits of the A1, when expressed from their native operon in E. coli, allowed isolation of an assembled A3B3DF subcomplex with MgATPase activity (Lemker et al., 2001). Although not shown yet for any member of the A‐ATPase subfamily, it is generally assumed that the A‐ATPases, much like the other members of the family, function as rotary motor enzymes. Rotation of the domain containing subunits D and F and the ring of proteolipids with respect to the A3B3 catalytic core has been shown for the related A/V‐type ATPases from the eubacterium Thermus thermophilus (Imamura et al., 2003; Yokoyama et al., 2003b).

II.

Overall Structural Features of the F‐, V‐, and A‐ATPases A.

1.

Structure of the F‐ATPase

Structure of the F1

Crystals diffracting x rays to high resolution were first reported for the F1‐ATPase from rat liver (Amzel and Pedersen, 1978), but the first atomic resolution structure was obtained for the enzyme from bovine heart (Abrahams et al., 1994), followed by structures for the bacterial (Shirakihara et al., 1997) rat liver (Bianchet et al., 1998), and chloroplast complexes (Groth and Pohl, 2001). Only a 4.4‐A˚ resolution structure is available for the F1‐ATPase from E. coli (Hausrath et al., 1999). All structures show a hexagonal barrel of alternating a‐ and b‐subunits about 100‐A˚ long and 120‐A˚ wide. The a‐ and b‐subunits display significant similarity at

ROTARY MOLECULAR MOTORS

353

the level of the primary structures, and the crystal structure showed that this is also true for the tertiary structures of the subunits. Each a‐ and b‐subunit is folded in three domains: an N‐terminal b barrel, a middle domain containing a‐helices, and b sheet and a C‐terminal domain containing only a‐helix. In the assembled F1, the N‐terminal domains interact to form a ring of b barrels, and it is this interaction that is mostly responsible for the stability of the complex. The hexagon of a‐ and b‐subunits contains a central cavity within which is located the a‐helical N‐ and C‐terminus of the g‐subunit. The remaining middle portion of g, together with the ‐subunit (subunit nomenclature of the bacterial enzyme), are protruding about 35 A˚ from the bottom of the a3b3 hexagon to form the central stalk. (Note that in the first high‐resolution MF1 structure, the middle portion of g and the ‐subunit were not resolved.) The ‐subunit (subunit nomenclature of the bacterial enzyme; Table I) is bound to the g‐subunit, and both subunits make contact with the membrane‐embedded ring of c‐subunits. A peripheral stalk made of the b‐ and d‐subunits connects the membrane embedded a‐subunit with the membrane extrinsic a3b3 domain (Dunn et al., 2000; Ogilvie et al., 1997). Features of the crystal structure of the F1‐ ATPase from bovine heart mitochondria, inhibited with dicyclohexylcarbodiimide (DCCD; Gibbons et al., 2000) are summarized in Fig. 2. Note that subunit d (OSCP in the mitochondrial enzyme) was not present in the F1‐ATPase preparation used for crystallization. As mentioned earlier, the F1 has six nucleotide binding sites, all located at the interfaces of the a‐ and b‐subunits. All six nucleotide binding sites are formed by amino acid residues from both a‐ and b‐subunits. Although the catalytic nucleotide binding sites are mostly located on the b‐subunits, with some contribution from the a‐subunits, the noncatalytic ones are mostly formed from a‐subunit residues with some contribution from the adjacent b‐subunits. In the crystal structures of the mitochondrial F1, all three noncatalytic nucleotide binding sites are filled with magnesium‐ nucleotide (e.g., the nonhydrolyzable ATP analogue, AMP‐PNP, or ADP) and the structure of these noncatalytic a/b interfaces shows little variation among the three sites. The three catalytic nucleotide binding sites, on the other hand, can exist in different states. In the first high‐resolution structure (and most subsequent structures), one catalytic site was filled with MgAMP‐PNP (called bTP), one with MgADP (bDP), and one was empty (bE). Figure 2B shows a 30‐A˚ ‐thick cross section of the F1 at the level of the nucleotide binding sites and in Fig. 2C a comparison of the nucleotide filled and empty a‐ and b‐subunits is shown. It can be seen that although there is little difference between the DP (ADP‐) and TP (AMP‐PNP‐) bound b‐subunits, the empty b‐subunit is quite different in that the nucleotide binding pocket is collapsed and the entire C‐terminal domain is

354

WILKENS

Fig. 2. Crystal structure of the mitochondrial F1‐ATPase. Crystal structure of the beef heart mitochondrial F1‐ATPase, inhibited with DCCD (Gibbons et al., 2000; 1E79.pdb). (A) View parallel to the membrane. (B) Cross section at the level of the nucleotide binding sites, viewed perpendicular to the membrane. The nucleotide occupancy is indicated. (C) Comparison of the different a‐ and b‐subunit pairs.

rotated and shifted downward by about 20A˚ . The structure of the a‐subunits shows little variation among the three copies in the complex. The nucleotide occupancy of the catalytic sites observed in the first crystal structure was exactly what Paul Boyer had predicted earlier in his binding‐change model of cooperative catalysis (Boyer, 1993). Consequently, this first high‐resolution structure of the F1‐ATPase immediately initiated a number of studies that ultimately led to the elucidation of the F1‐ATPase’s rotational mechanism of cooperative catalysis. At the time, the F1‐ATPase structure represented the largest asymmetric structure solved to atomic resolution by x‐ray crystallography, and this accomplishment, together with the visionary prediction of rotary catalysis, was subsequently awarded the 1997 Nobel prize in chemistry (to John Walker for the structure and Paul Boyer for the catalytic mechanism). However, whether the first (and many subsequent) structure(s) represented physiologically

ROTARY MOLECULAR MOTORS

355

relevant species was subsequently questioned by experimental evidence that showed that the maximum turnover of ATP hydrolysis in F1‐ATPase occurs only with all three catalytic sites filled with nucleotide (Weber and Senior, 1997). Despite the presence of three copies of each of the a‐ and b‐subunits, the F1 displays a significant degree of asymmetry which is probably due to the differential interaction of the three ab pairs with the single copy g‐subunit and the different occupancy of the nucleotide binding sites. This structural asymmetry is not only an interesting consequence of the sequential catalysis, but it probably also provides a mechanism for interaction with only one copy of the peripheral stalk‐forming d‐subunit (see Section II.A.3). This asymmetry, however, is only seen in structures in which Mg‐nucleotides were included in the crystallization mix. In the absence of nucleotides or magnesium, perfect threefold symmetry is observed as for the structure of the nucleotide‐free a3b3 catalytic core of the thermophilic bacterium PS3 (Shirakihara et al., 1997) and the F1‐ATPase from rat liver mitochondria (Bianchet et al., 1998), which was crystallized with ATP, but no magnesium.

2. Structure of the F0 At the time of this writing, there is no high‐resolution structural model available for the intact membrane domain of the F‐ATPase. The atomic resolution structure of the monomeric c‐subunit dissolved in organic solvent has been determined by nuclear magnetic resonance (NMR) spectroscopy at neutral and acidic pH (Girvin et al., 1998; Rastogi and Girvin, 1999). What is known is that the c‐subunits from different species are able to form rings of between 10 and 14 proteolipids. For example, the x‐ray crystal structure of the yeast F‐ATPase (Stock et al., 1999) shows that there are 10 c‐subunits, whereas the chloroplast enzyme has 14 (Seelert et al., 2000) and the complex from the bacterium Ilyobacter tartaricus has 11 (Stahlberg et al., 2001). None of these structures, however, contains the a‐subunit, and the majority of what we know at this point about the interface of this subunit and the c‐subunit ring comes from disulfide‐ mediated cross‐linking studies and molecular modeling (Fillingame and Dmitriev, 2002). The proteolipid ring alone does not catalyze passive proton translocation in the absence of subunit a and the N‐terminal domains of the b‐subunits (Greie et al., 2004; Schneider and Altendorf, 1985). The electron microscopy–derived structural models of the F0 domain alone (Birkenha¨ ger et al., 1995) and as part of F1F0, (Bo¨ ttcher et al., 1998; Gogol et al., 1987; Karrasch and Walker, 1999; Mellwig and Bo¨ ttcher, 2003; Wilkens and Capaldi, 1998a,b) display a limited amount of structural

356

WILKENS

detail, possibly due to bound lipid and/or detergent and conformational flexibility. What can be seen in the electron microscopy–derived structures is the asymmetry of the membrane domain, consistent with the 30‐kDa a‐subunit binding at the outside of the proteolipid ring. This asymmetry is more pronounced in the model of the mitochondrial F1F0, which contains a number of additional membrane‐bound polypeptides that have no homologues in the bacterial enzyme (Chen et al., 2004; Rubinstein et al., 2003). Subunit a has been modeled as a transmembrane protein containing either five (Vik et al., 2000; Yamada et al., 1996) or six ( Ja¨ ger et al., 1998) transmembrane a‐helices. The secondary structure of E. coli subunit a dissolved in organic solvent has been determined by NMR spectroscopy (Dmitriev et al., 2004a,b).

3.

Structure of the F1F0

The only crystal structure showing the F1 and part of the F0 (the proteolipid ring containing 10 c ‐subunits) has been obtained for the enzyme from yeast mitochondria at a resolution of 3.8 A˚ (Stock et al., 1999). The majority of structural information for the intact ATP synthase has been obtained by electron microscopy and single‐particle image analysis. The earliest projection images of stained specimens revealed the two‐ domain nature of the complex and the presence of the central stalk (McEnery et al., 1984; Tsuprun et al., 1989), and later on, the presence of the peripheral stalk (Bo¨ ttcher et al., 1998; Karrasch and Walker, 1999; Wilkens and Capaldi, 1998a,b; Wilkens et al., 2000). Although both central (Gogol et al., 1987) and peripheral (Mellwig and Bottcher, 2003; Rubinstein et al., 2003) stalks were subsequently confirmed by cryoelectron microscopy of unstained protein, the inherent difficulty of cryoelectron microscopy of detergent‐solubilized protein complexes has limited the resolution of the images to between 20 and 30 A˚ , thus revealing only gross features of the complex. Not resolved so far in any high‐resolution crystal structure are subunits d, b, and a. High‐resolution structures for domains of subunits d (Wilkens et al., 1997) and b (Del Rizzo et al., 2002; Dmitriev et al., 1999) are available, but the question remains how these domains interact with each other and with the remainder of the complex. Figure 3A shows electron microscopy of the intact E. coli F‐ATPase. Combining the low‐resolution electron density maps derived by electron microscopy with the high‐resolution structural information available for various subunits and their domains allows us to construct a high‐ resolution structural model of the intact E. coli F1F0‐ATP synthase as shown in Fig. 3B.

ROTARY MOLECULAR MOTORS

357

B. Structure of the V‐ATPase The majority of the biochemical and genetic information available for the vacuolar ATPase has been obtained from studies with the enzymes from yeast, plant, bovine brain clathrin‐coated vesicles, and insect. The only high‐resolution structural information reported for the vacuolar ATPase is for isolated subunits H (Sagermann et al., 2001) and C (Drory et al., 2004), both from the yeast enzyme. As for the intact F1F0‐ATP synthase, virtually all of the overall structural information for the vacuolar ATPase has been obtained by transmission electron microscopy in combination with single‐particle image analysis. The majority of the data for the eukaryotic V‐ATPase has been reported for the enzymes from bovine brain (Wilkens, 2001; Wilkens and Forgac, 2001; Wilkens et al., 1999, 2004), insect (Gru¨ ber et al., 2000; Radermacher et al., 1999; Rizzo et al., 2003), plant (Domgall et al., 2002; Li and Zhang, 2004), and yeast (Wilkens et al., 2005; Zhang et al., 2003). Figures 3C and D show an electron microscopic projection image and the three‐dimensional (3D) reconstruction of the V‐ATPase from bovine brain clathrin‐coated vesicles, respectively (Wilkens et al., 1999, 2004).

1. Structure of the V1 As indicated in Table I, the eukaryotic V1‐ATPase is composed of subunits AB(C)DEFGH. Subunit C is considered a V1 subunit, although it has little affinity for the V1 when it is dissociated from the V0. Although it is well established that there are three pairs of A and B subunits, the stoichiometry of the stalk subunits has not been settled and the ratio of these subunits might even be different depending on the source of the complex. In the bovine V‐ATPase, the ratio of subunits ABCDEFGH has been measured to be 3:3:1:1:1:1:2:2 (Arai et al., 1988; Xu et al., 1999), but the presence of as many as three copies of subunits E (Xie, 1996) and G (Supekova et al., 1996; Wieczorek et al., 2000) has been suggested based on Coomassie staining intensities of the subunit bands in denaturing polyacrylamide gels (Table I). The structure of the V1‐ATPase from insect (Gru¨ ber et al., 2000) and yeast (Zhang et al., 2003) has been analyzed by electron microscopy and 3D image reconstruction. Much like the F1‐ATPase structure, the electron microscopy–derived models of the V1‐ATPase show the alternating arrangement of the A and B subunits around a central cavity that contains the central stalk. Additional densities seen in the models of the vacuolar ATPase may account for the some of the additional subunits and subunit domains for which there are no homologues in the F‐ATPase such as subunit H and the nonhomologous region in the A‐subunit (Gru¨ ber et al., 2000; Zhang et al., 2003).

358

WILKENS

Fig. 3. Electron microscopy of the Escherichia coli F1F0‐ATP synthase and the bovine brain vacuolar ATPase: (A) Projection image of the E. coli F‐ATPase (Wilkens, 2000). The positions of subunit a (large black arrowhead), the peripheral stalk (white arrow), the C‐ terminal domain of the b‐subunits (white arrowhead), and the d‐subunit (small black arrowhead) are indicated. (B) Model of the E. coli F1F0‐ATP synthase, constructed from the crystal structures of the bovine F1‐ATPase (1E79.pdb) and the yeast F1‐c10 complex (1QO1.pdb). (C) Projection image of the bovine V‐ATPase (Wilkens et al., 1999). The

ROTARY MOLECULAR MOTORS

359

Subunits A and B, which are the counterparts to the F1‐ATPase b‐ and a‐subunits, respectively, are both involved in nucleotide binding (Vasilyeva and Forgac, 1996; Zhang et al., 1995). Evidence from chemical labeling and site‐directed mutagenesis experiments indicates that only the nucleotide binding sites on the A‐subunits are catalytic, whereas the binding sites on the B‐subunits are termed noncatalytic or nonexchangeable (reviewed in Forgac, 1999, 2000), just like the catalytic and noncatalytic sites on the b‐ and a‐subunits in the F1. The site‐directed mutagenesis data for the two types of nucleotide binding sites on the A‐ and B‐subunits also indicate that these subunits have a similar 3D fold to their F1‐ATPase counterparts b and a in that mutations in the yeast A‐ and B‐subunits, which were selected based on sequence alignment and the crystal structure of the mitochondrial F1 (Abrahams et al., 1994), have a differential inhibitory effect depending on whether residues in the putative catalytic or noncatalytic sites were changed (Liu et al., 1996, 1997; MacLeod et al., 1998; reviewed in Forgac, 1999, 2000). Additional evidence that both the vacuolar A‐ and the F1 b‐subunit exhibit the same tertiary fold comes from disulfide cross‐linking studies. Residues Cys254 and Cys532 in the V‐ATPase A‐subunit can form a disulfide bond under native conditions (Feng and Forgac, 1994), which puts these positions close together in 3D space. The homologous positions in the mitochondrial F1‐ATPase b‐subunit are also close in space based on the crystal structure of the MF1 (Abrahams et al., 1994), again suggesting a similar 3D structure for the A and b polypeptides. High‐resolution crystal structures have been reported for the H‐ and C‐subunits of the yeast V‐ATPase (Drory et al., 2004; Sagermann et al., 2001).

2. Structure of the V0 The V0 is made of subunits acc 0 c 00 de. Subunits c, c 0 , and c 00 are proteolipid isoforms each containing one essential lipid‐exposed glutamate residue. The proteolipids in the V‐ATPase are twice the size compared with the F‐ATPase proteolipids, and it is assumed that the four transmembrane a helix c‐ and c 0 ‐subunits are a result of an early gene fusion event. Subunit c 00 has two lipid‐exposed glutamates, but only one of them is essential for proton pumping. The 100‐kDa subunit a is a two‐domain protein with a

positions of the C‐subunit (black arrowhead), one of the peripheral stalks (white arrow and arrowhead), and the nonhomologous region in the A‐subunit (small black arrowhead) are indicated. (D) 3D reconstruction of the bovine V‐ATPase. In the stalk region of the complex, several protein densities can be identified that can be assigned to subunits C, H, H’, and the N‐terminal domain of subunit a. The central rotating stalk (subunits D/F and d) is hidden from view by the H’ subunit.

360

WILKENS

membrane‐integral C‐terminal domain and a cytoplasmic N‐terminal domain. Both domains are about the same size, and it is the membrane‐ spanning region that is the functional homologue to the F‐type a‐subunit. The topology of the V0 a‐subunit has been determined for the yeast enzyme (Leng et al., 1999). The data suggest that there are nine transmembrane a‐helices in the C‐terminal half of the a‐subunit and that the hydrophilic N‐terminal portion is exposed to the cytosolic side of the membrane. As for the V1, the ratio of the V0 subunits is still under discussion. In one study, evidence had been presented for there being one copy of c 00 per yeast V‐ATPase complex (Powell et al., 2000) and, in another report, it was proposed that there are two copies of the same subunit in the complex from the same source (Gibson et al., 2002). Despite this uncertainty, it can be assumed that in the V0, the total number of lipid‐exposed carboxyl groups involved in proton transport is between five and seven (Arai et al., 1988; Murata et al., 2003; Pa´ li et al., 1995), which means that there are about half as many proton binding sites in the V0‐ compared with the F0‐ proteolipid ring. Subunit d, although not a transmembrane protein, is tightly associated with the proteolipid ring. The crystal structure of the bacterial homologue of the bovine d‐subunit (subunit C) has recently been reported (Iwata et al., 2004), and it has been shown that this subunit binds to the cytoplasmic loops of the proteolipids (Yokoyama et al., 2003a). The subunit thus forms a ‘‘spacer’’ connecting the V1 and V0 rotor domains and, since there is no subunit d homologue in the F‐ATPase, this would explain the substantially longer central stalk in the V‐ compared with the F‐ATPase. A structural model of the V0 domain from the bovine brain V‐ATPase has been calculated from electron microscopic images of the negatively stained complex at a resolution of 21 A˚ (Wilkens and Forgac, 2001).

3.

Structure of the V1V0

The first electron microscopic images of the eukaryotic V‐ATPase revealed its structural similarity to the mitochondrial F‐ATPase (Bowman et al., 1989; Dschida and Bowman, 1992). Subsequent studies using electron microscopy in combination with image analysis showed the more complex nature of the V‐ATPase (Domgall et al., 2002; Li and Zhang, 2004; Wilkens et al., 1999, 2004, 2005). The more complex structure of the vacuolar ATPase compared with that of the F‐ATPase had to be expected based on the subunit composition and the V‐ATPase’s relative molecular mass (~900 kD), which is almost twice as large as the bacterial F‐ATPase’s (~530 kD). The larger degree of complexity of the V‐ATPase might be explained by the fact that the activity of the vacuolar ATPase is subject to a

ROTARY MOLECULAR MOTORS

361

high degree of regulation, often in a tissue‐specific manner. It is likely that the V‐ATPase subunits or subunit domains that have no homologue in the F‐ATPase are, for example, involved in the process of reversible dissociation, a regulation mechanism that is unique to the V‐ATPases.

C.

Structure of the A‐ATPase

Only very limited structural information is available for the A‐ATPase found in Archaea. Although the structure of the A‐ATPase is more closely related to the eukaryotic V‐ATPase, the A‐ATPase probably functions as an ATP synthase much like the F‐ATPases (Mu¨ ller and Gru¨ ber, 2003; Scha¨ fer and Meyering‐Vos, 1992). Structural analysis of this A1 subcomplex by electron microscopy and image reconstruction revealed a close similarity to the structure of the eukaryotic V1 domain (Coskun et al., 2004b). Recent electron microscopic images of the intact A‐ATPase from Methanococcus jannaschii (Coskun et al., 2004a) have revealed that the overall structure of the A‐ATPase is very similar to the eubacterial sodium pumping A/V‐type ATPase from Caloramator fervidus (Boekema et al., 1997) and T. thermophilus (Iwata et al., 2004). The number of transmembrane‐spanning a‐helices found in the proteolipids of A‐ATPases ranges from two to six, and it is thought that this variation is a result of multiple gene fusion events starting from the two‐transmembrane helix ancestral proteolipid subunit (Mu¨ ller, 2004). In the methanogen Methanopyrus kandleri, several fusion events lead to a membrane domain in which the proteolipid ring is believed to be composed of a single polypeptide containing 13 predicted transmembrane‐spanning a‐helical hairpins (Lolkema and Boekema, 2003). Much like the yeast V‐ATPase A‐subunit, the A‐ATPase A‐subunit contains an intein, though significantly smaller in size. Based on sequence comparison, the eubacterial V‐type ATPase is a product of horizontal gene transfer from the Archaea. Much of the information about the subunit arrangement in the eubacterial A/V‐type–like ATPase comes from studies with the enzymes from C. fervidus and T. thermophilus (Boekema et al., 1997, 1999; Ubbink‐Kok et al., 2000; Yokoyama et al., 2003a). A recent 3D reconstruction of the A/V‐ ATPase from T. thermophilus (Bernal and Stock, 2004) revealed the structural similarity to both the eukaryotic V‐ATPase (Domgall et al., 2002; Wilkens et al., 2004, 2005) and the archaeal A‐ATPase (Coskun et al., 2004a,b). The x‐ray crystal structure for the C‐subunit of the bacterial A/V‐like ATPase has been reported recently (Iwata et al., 2004). The C‐subunit of the bacterial V‐like ATPase shows limited but significant sequence homology to the d‐subunit of the eukaryotic V‐ATPase.

362

WILKENS

III.

Mechanistic Aspects

A. Rotational Catalysis The mismatch between the three nucleotide binding sites on the F1 as opposed to only a single ion channel in the F0 was recognized long before detailed structural information for any portion of the F1F0‐ATP synthase was available. An elegant solution for overcoming the mismatch problem was first offered by Paul Boyer and colleagues in the early 1980s (Boyer, 1993, 1997). In Boyer’s model, the three catalytic sites on the F1 were coupled to the single ion channel by a rotational movement of the g‐subunit inside the catalytic core of the complex. The model of rotational catalysis, together with the same author’s ‘‘binding‐change’’ model of cooperative catalysis in the F1, set the tone for the F‐ATPase research from that time on. Over the following years, several laboratories were involved in trying to provide the ultimate experimental proof that rotation was indeed part of the catalytic mechanism of F‐ATPase. By using a combination of experimental approaches, including protease digestion, chemical cross‐linking, electron microscopy, and fluorescence spectroscopy, Capaldi and coworkers (1992, 1994) provided compelling evidence that motion of the central stalk subunits g and  with respect to the (ab)3 catalytic core takes place during catalysis. The cross‐linking experiments especially defined the subunits of the rotor and stator early on in that cross‐links between g and  and g and c did not inhibit enzymatic activity, whereas cross‐links between g or  and a or b did (Capaldi and Aggeler, 2002). Although these experiments provided a great deal of structural and mechanistic information for the F‐ATPase, they fell short of providing a full 360‐degree rotational movement of the g‐subunit with respect to a or b . Using the site‐directed disulfide cross‐link approach applied earlier by Aggeler and Capaldi, Cross and coworkers were the first to show that the g‐subunit changes position from one b‐subunit to another during ATP hydrolysis (Duncan et al., 1995). Junge and colleagues, on the other hand, using a photo‐bleaching fluorescence polarization relaxation approach, showed that the g‐subunit was rotating at least 270 degrees with respect to the immobilized (ab)3 hexagon during turnover (Sabbert et al., 1996), and while still short of showing full rotation, the experiment was the first to be conducted in real time. However, all these experiments had in common that the information from a large number of ATPase molecules had to be ‘‘averaged’’ to obtain enough ‘‘signal,’’ be it by gel electrophoresis, fluorescence, or image analysis of electron microscopic projections, and it became clear that in order to provide the ultimate proof of

ROTARY MOLECULAR MOTORS

363

continuous rotation, a single molecule had to be observed. The design and realization of this experiment was first accomplished by the groups of Yoshida and Kinosita in 1997, and their experimental setup can now be considered the classic rotation experiment.

B.

The Classic Rotation Experiment

The setup for the rotation experiment used by Yoshida and Kinosita (Noji et al., 1997) is illustrated in Fig. 4. F1‐ATPase ((ab)3g) from the thermophilic bacillus PS3 was immobilized on a NiNTA‐modified glass surface via polyhistidine tags introduced into the N‐termini of the b‐subunits. A cysteine in the g‐subunit was linked via maleimide biotin to streptavidin‐labeled actin filaments that were up to several micrometers in length. The actin filaments were decorated with fluorescent dye molecules. Rotation of the actin filaments was observed under the fluorescence microscope only when MgATP was added to the medium. In the experimental setup used, the observed rotation was always counter‐clockwise and the rate of the rotation was related to the length of the actin filament linked to the g‐subunit. This rate dependency on the length of the actin filament allowed an estimate of the ATP hydrolysis‐generated torque, which was calculated to be approximately 40–50 pN
nm. This setup, or modifications thereof, was then used in experiments to show rotation of the ‐subunit (Kato‐ Yamada et al., 1998), proteolipid ring (Pa¨ nke et al., 2000; Sambongi et al., 1999; Tsunoda et al., 2001a; Ueno et al., 2005), and other subunits with respect to an immobilized a3b3 or proteolipid domain (Nishio et al., 2002). Rotation has also been shown for the yeast V‐ATPase (Hirata et al., 2003) and the eubacterial A/V‐like ATPase (Imamura et al., 2003; Yokoyama et al., 2003b).

C. The Functional Elements: Rotor and Stator Any macroscopic motor is divided into a static portion, the stator, and a portion that is moving with respect to the static portion, called the rotor. To understand the mechanical properties of the molecular motor, we have to look at the properties of the motor’s rotor and stator and their interactions on a molecular level. In the case of the A/V/F‐ATPases, deciding which portion to call the rotor and which the stator is based on the relative molecular masses of the two functional elements. In the E. coli F‐ATPase, for example, the stator is composed of a3b3dab2 with a relative molecular mass of approximately 380 kD, whereas the rotor is composed of gc10 with a relative molecular mass of 150 kD. In the related vacuolar ATPase, it is

364

WILKENS

Fig. 4. The classic rotation experiment: (A) F1 (a3b3g) is immobilized on a NiNTA modified glass surface via polyhistidine tags introduced into the N‐termini of the b‐ subunits. A fluorescently labeled actin filament is attached to the g‐subunit via biotin/

ROTARY MOLECULAR MOTORS

365

assumed that the rotor contains subunits DFd(c,c 0 ,c 00 )5–6 whereas the stator is made of subunits A3B3CEG2H2a (subunit stoichiometry of the bovine enzyme; Arai et al., 1988; Xu et al., 1999).

1. Rotor The rotor in the F‐ATPase is composed of F1 subunits g and  and the ring of proteolipids from the F0. As mentioned before, high‐resolution structural information is available for the F1‐ATPase from bovine heart, rat liver, and spinach chloroplasts. Only the crystallographic model of the bovine enzyme, however, contains complete structural data for the rotor subunits g and  ( is called d in the mitochondrial F‐ATPase). For the ring of proteolipids, only a 3.8‐A˚ resolution crystal structure is available (Stock et al., 1999). The portion of the g‐subunit located inside the ring of a‐ and b‐subunits consists of the N‐ and C‐terminus of the polypeptide, which are folded as an a‐helical coiled-coil. The coiled‐coil domain of g interacts only weakly with the interior of the a3b3 hexagon (Abrahams et al., 1994). The weakness of this interaction can be explained by the fact that the interaction has to be broken and reformed three times during one revolution of the motor. The middle portion of g is a more or less globular domain that contains the binding site to the ‐subunit and a short a‐helix that is in contact with one of the C‐terminal domains of the b‐subunits (bTP in the MF1 crystal structure; Fig. 2). The structure of the ‐subunit, first determined by solution NMR spectroscopy and x‐ray crystallography of the isolated protein (Uhlin et al., 1997; Wilkens and Capaldi, 1998c; Wilkens et al., 1995), consists of an N‐terminal 10‐stranded b sandwich and a C‐terminal a‐helical hairpin. The ‐subunit possesses a high affinity for the g‐subunit with a binding constant in the nanomolar range (Dunn, 1997). The interaction between the two subunits is mediated by the b sandwich domain of the ‐subunit and the middle domain of the g‐subunit (Gibbons et al., 2000; Rodgers and Wilce, 2000). The two subunits can be cross‐linked without affecting the function of the ATP synthase (Capaldi and Aggeler, 2002). The bottom of the g‐subunit middle domain and the bottom opening of the ‐subunit b sandwich sit on the cytoplasmic opening of the proteolipid ring to couple rotational movements of the F1 and F0 parts of the rotor. The majority of the binding energy between the proteolipid ring and the g complex, however, is most likely provided by the

streptavidin and rotation of the actin filament is observed under the light microscope (Noji et al., 1997). (B) F1F0‐ATP synthase is immobilized on the glass surface via polyhistidine tags in the proteolipids (c‐subunits) and rotation of the a3b3ab2d domain with respect to the c10g domain is observed (Nishio et al., 2002).

366

WILKENS

interaction of the bottom of the ‐subunit with one or more of the cytoplasmic loops of the c‐subunits. In fact, the first evidence that the F1 ‐ subunit and the F0 c‐subunits have to be in close contact came from mutagenesis studies showing that uncoupling mutations in the ‐subunit could be suppressed by mutations in the c‐subunit cytoplasmic loop (Zhang et al., 1994). The asymmetric interaction of g with the c‐subunit ring means that the axis of the F1 rotor does not coincide with the axis of the F0 rotor, giving the power coupling an eccentric appearance. This eccentric interaction is clearly seen in structural models of the F1F0‐ATP synthase from x‐ray crystallography (Stock et al., 1999) and electron microscopy (Chen et al., 2004; Mellwig and Bo¨ ttcher 2003; Rubinstein et al., 2003; Wilkens, 2000). The ‐subunit can exist in two distinct conformations: one in which the N‐ terminal b sheet and C‐terminal a‐helical domains are close together, and one in which the a‐helical domain is pointing away from the b sheet domain to make contact with the bottom of the a3b3 hexamer. The first conformation is seen in the isolated ‐subunit (Uhlin et al., 1997; Wilkens et al., 1995) and when  is bound to F1 with MgADP at the catalytic sites (Wilkens and Capaldi, 1998c). The second conformation is seen in the isolated g complex (Rodgers and Wilce, 2000), and it has been speculated that the up‐and‐down conformation of the ‐subunit is involved in regulation of the enzyme during ATP synthesis/hydrolysis (Tsunoda et al., 2001b), possibly induced by nucleotide binding to the subunit (Suzuki et al., 2003). The structure of the E. coli c‐subunit has been determined by solution NMR spectroscopy in organic solvent (Girvin et al., 1998). The c‐subunit contains the proton binding residue in the second a‐helix (Asp61 in E. coli). Interestingly, the structure of the c‐subunit in the deprotonated form is significantly different to the protonated form, suggesting a possible mechanism of torque generation in the F0 upon protonation and deprotonation of neighboring c‐subunits in the interface to the a‐subunit (Fillingame et al., 2002; Rastogi and Girvin, 1999). In the E. coli enzyme, there are 10 c‐subunits that form the proteolipid ring. The number of c‐subunits seems to be species‐dependent, with the number being 10 in E. coli, 11 in I. tartaricus, and 14 in spinach chloroplasts. The different numbers of c‐subunits must result in a different ratio of ATP molecules synthesized/hydrolyzed per protons translocated in the different species. It is possible that this difference in ratios is due to the differences in average membrane potential in the various organisms, but the possibility that the difference seen is an artifact generated by the polymerization of the isolated c‐subunits used for structural studies cannot be entirely excluded at this point. A recent study in which c‐subunit fusions were generated showed that a proteolipid ring containing 10 c‐subunits resulted in active

ROTARY MOLECULAR MOTORS

367

ATP synthase, whereas fusions of 9, 11, or 12 proteolipids were essentially inactive in ATP synthesis and ATP hydrolysis‐driven proton pumping, but active in ATP hydrolysis (Mitome et al., 2004). Currently, there is no reliable high‐resolution structure of subunit a. Solution NMR studies in organic solvent (Dmitriev et al., 2004a,b) have allowed assignment of most of the backbone and many side‐chain resonances, but a 3D structure has not been reported yet. Consequently, most of the current structural information for subunit a and its interaction with subunits c and b has been obtained by disulfide‐mediated cross‐linking from cysteines introduced into the subunits by site‐directed mutagenesis (Fillingame and Dmitriev, 2002). Functional studies have revealed early on that arginine 210 (Arg210) of the E. coli a‐subunit is essential for proton translocation activity. Together with aspartate 61 (Asp61) of the c‐subunit, these two amino acids appear to play the key role in proton translocation through the F0. Cysteines introduced by site‐directed mutagenesis were used to probe accessibility of subunit a and c side‐chains to water molecules. These studies suggested the presence of water‐accessible half channels in the a‐subunit at the putative interface to the c‐subunits (Angevine et al., 2003), just as had been predicted earlier (Vik and Antonio, 1994). Although it is known that the transmembrane domains of subunit b are also required for proton transport (Greie et al., 2004; Schneider and Altendorf, 1985), it is not clear at this point whether the two N‐terminal a‐helices of the b‐subunits are directly involved in the transport mechanism or whether they have only a structural role. Models for the dynamic mechanism of the proton channel have been presented based on mathematical modeling, assuming static structures (Elston et al., 1998; Oster and Wang, 2003) and helix swiveling (Fillingame et al., 2002) of the c‐subunits. These models provide interesting insights into how the electrostatic interactions between the negatively charged carboxylate of Asp61 and the positively charged guanidinium moiety of Arg210 might drive rotation of the proteolipid ring. However, in the absence of a high‐resolution structure of the interface between the two proton channel subunits, these models remain somewhat speculative.

2. Stator The stator in the F‐ATPase consists of F1 subunits a, b, d, and F0 subunits a and b. When describing the second or peripheral stalk on the other hand, only the b‐ and d‐ subunits are usually included. Structurally, the peripheral stalk is much less well defined compared with the rotating central stalk, possibly due to its flexible nature. There is no high‐resolution structure at this point for the membrane‐bound a‐subunit, the intact b‐subunit and the

368

WILKENS

C‐terminal domain of the d‐subunit. Structures are available for the membrane domain of the b‐subunits (Dmitriev et al., 1999), the dimerization domain of the b‐subunit (Del Rizzo et al., 2002), and the N‐terminal domain of the d‐subunit (Wilkens et al., 1997). Based on biochemical and genetic data, models for the interaction of these subunits have been proposed. According to the data, the N‐terminus of one a‐subunit binds to the N‐terminal domain of the d‐subunit (Weber et al., 2003), and the C‐terminal domain of the d‐subunit binds to the C‐terminal domain of one of the two b‐subunits (Grabar and Cain, 2003; McLachlin et al., 1998). This interaction seems only to occur with high affinity in the intact ATP synthase because it is not seen with the isolated d‐ and b‐subunits alone (Rodgers et al., 1997). The two b‐subunit N‐terminal transmembrane a‐helices bind to each other and to the a‐subunit, and they alone are sufficient to reconstitute a passive proton translocating pore together with the a‐ and c‐subunits (Greie et al., 2004). It has been shown that only one of the C‐termini of the b‐subunits is required for the formation of an intact and functional ATP synthase (Grabar and Cain, 2003), and it is possible that the second copy of the b‐subunit functions to stabilize the peripheral stalk for much of its length. Different‐length b‐subunits (here called subunits I and II) are found in F1F0‐ATP synthase from chloroplasts. In the V‐ATPase, the presence of at least two peripheral stators has been reported (Boekema et al., 1999; Domgall et al., 2002; Wilkens et al., 1999, 2004). Candidates for the stator proteins in the V‐ATPase are subunits E and G, and support for this assignment comes from photochemical cross‐ linking studies that place these subunits on the outside of the V1 domain for much of its length (Arata et al., 2002a,b). Another subunit that probably plays the role of a stator in the V‐ATPase is the a‐subunit of the V0. The V0 a‐subunit has a large cytoplasmic domain, and it has been shown that this domain interacts with the A‐subunit of the V1 (Landolt‐Marticorena et al., 2000).

D. Torque Generation F‐ATPase is a reversible enzyme that can work in the direction of proton gradient‐driven ATP synthesis and in the direction of ATP hydrolysis‐ driven proton pumping. Reversibility has recently been shown by elegantly turning the rotor in the F‐ATPase mechanically in the direction of ATP synthesis. When ADP and inorganic phosphate were present during the forced rotation, ATP was synthesized and released from the enzyme (Itoh et al., 2004). This means that the complex contains two molecular motors: the F1, which is an ATP hydrolysis‐driven motor, and the F0, which is a

ROTARY MOLECULAR MOTORS

369

proton gradient‐driven motor. In the vacuolar ATPase, only the ATP hydrolysis‐driven motor in the V1 domain has physiological significance, but, as mentioned earlier, V‐ATPase can be reversed to synthesize ATP (Hirata et al., 2000).

1. Torque Generation in the ATPase domain In the F1 and V1 (and probably the A1, although not yet shown directly for the A‐ATPase), rotational torque is generated by the sequential binding of MgATP to the three catalytic sites, g phosphate bond cleavage, and subsequent release of the products MgADP and Pi. It has been shown that each 120‐degree rotation of the g‐subunit within the F1 occurs in two substeps—an 80‐degree and a 40‐degree rotational movement (Yasuda et al., 2001). This means that possibly more than one of the processes involved in ATP hydrolysis are able to generate rotational movement, or torque. The exact molecular mechanism by which catalytic events on the a‐ and b‐subunits drive the rotational movement of the g‐subunit is still poorly understood. Catalysis, be it ATP hydrolysis or ATP synthesis, is taking place on b‐subunits with contributions from a‐subunit residues, but there is no evidence that any part of the g‐subunit participates directly in the chemistry of phosphoester bond formation or cleavage. That means that in the hydrolysis direction, events in the catalytic sites are moving the g‐subunit, and in the direction of ATP synthesis, the moving g‐subunit will induce similar changes in the catalytic sites, ultimately leading to the binding of ADP and Pi. In the hydrolysis mode, the steps that could cause rotation are MgATP binding to the empty site and/or the hydrolysis reaction. Release of product ADP most likely requires energy input. Based on the recently described three‐nucleotide structure of MF1 (Menz et al., 2001), it appears that MgATP binding to bE in the two‐nucleotide structure results in a partially closed b‐subunit and a g‐subunit rotation of about 20–30 degrees, the value of the angle depending on which part of the structure is chosen as a reference point. In this model, the hydrolysis step immediately following MgATP binding to the empty site would provide the greatest rotational push on the g‐subunit. Angle‐resolved simultaneous observation of nucleotide binding and rotation suggests that each 120‐degree rotational step is, indeed, preceded by nucleotide binding to an empty catalytic site, although the resolution in these observations was not sufficient to provide the ultimate proof for that order of events (Nishizaka et al., 2004). What are the contact points between ab and g that provide the mechanical link? Targeted molecular dynamics has been used to model a probable pathway of the g‐subunit when rotating between catalytic sites (Ma et al., 2002). The data provide a detailed model of how the g‐subunit twists while it rotates,

370

WILKENS

what the contact points are within the a3b3 hexamer during the motion, and how the forced rotation of g might induce the different conformations in the catalytic sites, thus enabling binding of ADP and phosphate in one catalytic site and release of ATP from another.

2.

Torque Generation in the Ion Channel

Elucidating the torque generation in the ion channel has been hindered by the fact that there is no high‐resolution structure for the intact membrane domain of any of the F‐, V‐, and A‐ATPases. What is known is that rotation of the c‐subunit ring against the a/b subunit domain is required for passive proton translocation (Suzuki et al., 2002). Mechanistic models of the ion conduction in the F0 have been proposed based on the structure of the protonated and deprotonated E. coli F‐ATPase c‐subunit (Rastogi and Girvin, 1999), and based on what is known about the a‐ subunit from genetic and biochemical experiments. Disulfide cross‐linking via cysteines introduced into the a‐ and c‐subunits and the membrane domain of the b‐subunit has led to a detailed picture of how these subunits interact to form the proton‐permeable membrane domain (see above). Sophisticated mathematical models have been presented that are based on electrostatic attraction/repulsion between the a‐subunit arginine and c‐ subunit aspartate side‐chains as well as (biased) Brownian motion that is needed to overcome local potential minima (Oster and Wang, 2003). These models, however, have only been partially satisfactory due to fact that they are based on low‐resolution structural information that leaves too much freedom to accommodate the limited number of experimental constraints.

3.

Elasticity of the Stator and Rotor

There is now good evidence that there are 10 c‐subunits in the functional E. coli ATP synthase ( Jiang et al., 2001; Stock et al., 1999). How, then, are ten 36‐degree rotations in the F0 coupled to three 120‐degree rotations (or more accurately, to three each of 80‐ and 40‐degree substep rotations) in the F1? It has been proposed that both the central and/or peripheral stalks are able to store elastic energy that is built up by the stepwise translocation of protons. The elastic energy is then used portion by portion for ATP synthesis. Experimental evidence has shown that the part of the stator consisting of the two b‐subunits can be shortened or extended without significant loss of function (Sorgen et al., 1998, 1999). This observation seems to rule out that the peripheral stalk formed by the highly a‐helical b‐subunits provides a completely rigid connection between F1 and F0. Molecular modeling for the g‐subunit, on the other hand, suggests that

ROTARY MOLECULAR MOTORS

371

its coiled‐coil region consisting of the N‐ and C‐terminal a‐helices can act as a rotational spring that might be able to store part of the energy released in the F1 or F0 motor during power coupling (Mitome et al., 2004). By superimposing the energy landscapes of the rotating g in F1 and the c‐ring rotation, Mitome and coworkers could model that the rotation of g is not a smooth angular motion, but rather a motion in which either a rotating c‐ring twists g without much overall g rotation in F1, or that a twisted g rotates in the F1 without much overall c‐ring rotation. Occasionally, a mixture of the two occurs in which both c‐ring rotation and g rotation in F1 are seen at the same time. The extent to which g rotation in F1, c‐ring rotation induced g twisting, and coupled rotation occur varies over one full revolution, thus allowing the transport of, say, 10 protons per synthesis or hydrolysis of three molecules of ATP. This way, an average of 3.3 protons is used per ATP synthesized (or in other words, sometimes three protons are used, sometimes four). As of now, only c‐subunit ring stoichiometries have been found that are not multiples of three. It remains to be seen whether this is a common principle in all F0‐, V0‐, and A0‐motors.

E. Motor Efficiency Motor efficiency in the F‐ATPase has been estimated by comparing the work done per step and the free energy change under the conditions of the experiment. It has been shown that one ATP molecule hydrolyzed causes the rotor to rotate one step of 120 degrees. The work per step is 2p/3 (120 degrees) times the torque of ~40–45 pN
nm, which is approximately 80–90 pN
nm. This work has been measured under experimental conditions in which the free energy change (
G) of ATP hydrolysis is approximately 90 pN
nm, indicating that the ATP hydrolysis‐driven motor can work with nearly 100% efficiency. This means that F‐ATPase is able to convert the chemical energy stored in ATP (or better, the intracellular or experimental ATP to ADP/Pi ratio) to mechanical work almost without any irreversible loss of energy in form of heat. The question why the F‐ATPase motor can work with such high efficiency has been investigated by theoretical modeling, which suggests that the binding and release of nucleotide has to occur in several small steps, each accompanied with only small changes in free energy (Oster and Wang, 2000, 2003). The small steps represent, for example, the formation or breaking of individual hydrogen bonds and salt bridges during binding of substrate or release of product. By breaking up the total free energy change into several small steps, the likelihood of diverting the free energy change to heat‐generating processes, such as exciting vibrational energy levels, is minimized.

372

WILKENS

IV.

Conclusions

The F‐, V‐, (and A)‐ATPases represent a unique family of highly efficient energy converters. They function as reversible rotary motors, either converting ion motive force into the chemical energy currency ATP or chemical energy in the form of ATP into potential energy in the form of an electrochemical gradient. From an engineering standpoint, these tiny rotary motors are truly unique with respect to size, efficiency, and adaptability. Over the past 10 years, since the publication of the first high‐ resolution crystal structure of the F1‐ATPase, we have learned a great deal about how the ATPase motor works. Many questions have been answered, but other issues remain poorly understood. These include the nature of the elastic power coupling, and whether a ‘‘bi‐site’’ mechanism is sufficient for rotation or ‘‘tri‐site’’ catalysis is required. Another big question concerns the atomic structure of the membrane domains of the F‐, V‐, and A‐ATPases. Molecular modeling based on genetic, cross‐linking, and low‐ resolution structural data has provided valuable insight here, but, in the end, only high‐resolution structures of all the intermediates of the intact F‐, V‐, and A‐ATPase will give us the opportunity to address all the questions on a molecular level.

References Abrahams, J. P., Leslie, A. G., Lutter, R., and Walker, J. E. (1994). Structure at 2.8 A˚ resolution of F1‐ATPase from bovine heart mitochondria. Nature 370, 621–628. Alper, S. L. (2002). Genetic diseases of acid‐base transporters. Annu. Rev. Physiol. 64, 899–923. Amzel, L. M., and Pedersen, P. L. (1978). Adenosine triphosphatase from rat liver mitochondria: Crystallization and x‐ray diffraction studies of the F1‐component of the enzyme. J. Biol. Chem. 253, 2067–2069. Angevine, C. M., Herold, K. A., and Fillingame, R. H. (2003). Aqueous access pathways in subunit a of rotary ATP synthase extend to both sides of the membrane. Proc. Natl. Acad. Sci. USA 100, 13179–13183. Arai, H., Terres, G., Pink, S., and Forgac, M. (1988). Topography and subunit stoichiometry of the coated vesicle proton pump. J. Biol. Chem. 263, 8796–8802. Arata, Y., Baleja, J. D., and Forgac, M. (2002a). Localization of subunits D, E, and G in the yeast V‐ATPAse complex using cysteine mediated cross‐linking to subunit B. Biochemistry 41, 11301–11307. Arata, Y., Baleja, J. D., and Forgac, M. (2002b). Cysteine‐directed cross‐linking to subunit B suggests that subunit E forms part of the peripheral stalk of the vacuolar Hþ‐ ATPase. J. Biol. Chem. 277, 3357–3363. Bernal, R. A., and Stock, D. (2004). Three‐dimensional structure of the intact Thermus thermophilus Hþ‐ATPase/synthase by electron microscopy. Structure 12, 1789–1798. Bianchet, M., Hullihen, J., Pedersen, P. L., and Amzel, M. (1998). The 2.8‐A˚ structure of rat liver F1‐ATPase: Configuration of a critical intermediate in ATP synthesis/ hydrolysis. Proc. Natl. Acad. Sci. USA 95, 11065–11070.

ROTARY MOLECULAR MOTORS

373

Birkenha¨ ger, R., Hoppert, M., Deckers‐Hebestreit, G., Mayer, F., and Altendorf, K. (1995). The F0 complex of the Escherichia coli ATP synthase: Investigation by electron spectroscopic imaging and immunoelectron microscopy. Eur. J. Biochem. 230, 58–67. Boekema, E. J., Ubbink‐Kok, T., Lolkema, J. S., Brisson, A., and Konings, W. N. (1997). Visualization of a peripheral stalk in V‐type ATPase: Evidence for the stator structure essential to rotational catalysis. Proc. Natl Acad. Sci. USA 94, 14291–14293. Boekema, E. J., van Breemen, J. F. L., Brisson, A., Ubbink‐Kok, T., Konings, W. N., and Lolkema, J. S. (1999). Connecting stalks in V‐ATPase. Nature 401, 37–38. Bo¨ ttcher, B., Schwarz, L., and Gra¨ ber, P. (1998). Direct indication for the existence of a double stalk in CF0F1. J. Mol. Biol. 281, 757–762. Bowman, B. J., Dschida, W. J., Harris, T., and Bowman, E. J. (1989). The vacuolar ATPase of Neurospora crassa contains an F1‐like structure. J. Biol. Chem. 264, 15606–15612. Bowman, E. J., Tenney, K., and Bowman, B. J. (1988). Isolation of genes encoding the Neurospora vacuolar ATPase: Analysis of vma–1 encoding the 67‐kDa subunit reveals homology to other ATPases. J. Biol. Chem. 263, 13994–14001. Boyer, P. D. (1993). The binding change mechanism for ATP synthase: Some probabilities and possibilities. Biochim. Biophys. Acta 1140, 215–250. Boyer, P. D. (1997). The ATP synthase–a splendid molecular machine. Ann. Rev. Biochem. 66, 717–749. Cabezon, E., Butler, P. J., Runswick, M. J., Carbajo, R. J., and Walker, J. E. (2002). Homologous and heterologous inhibitory effects of ATPase inhibitor proteins on F‐ATPases. J. Biol. Chem. 277, 41334–41341. Capaldi, R. A., and Aggeler, R. (2002). Mechanism of the F(1)F(0)‐type ATP synthase, a biological rotary motor. Trends Biochem. Sci. 27, 154–160. Capaldi, R. A., Aggeler, R., Gogol, E. P., and Wilkens, S. (1992). Structure of the Escherichia coli ATPsynthase and role of the g and  subunits in coupling catalytic sites and proton channeling functions. J. Bioenerg. Biomembr. 24, 435–439. Capaldi, R. A., Aggeler, R., Turina, P., and Wilkens, S. (1994). Coupling between the catalytic sites and the proton channel in F1F0‐type ATPases. TIBS 219, 284–289. Chen, C., Ko, Y., Delannoy, M., Ludtke, S. J., Chiu, W., and Pedersen, P. L. (2004). Mitochondrial ATP synthasome: Three‐dimensional structure by electron microscopy of the ATP synthase in complex formation with carriers for Pi and ADP/ATP. J. Biol. Chem. 279, 31761–31768. Coskun, U., Chaban, Y. L., Lingl, A., Mu¨ ller, V., Keegstra, W., Boekema, E. J., and Gru¨ ber, G. (2004a). Structure and subunit arrangement of the A‐type ATP synthase complex from the archaeon Methanococcus jannaschii visualized by electron microscopy. J. Biol. Chem. 279, 38644–38648. Coskun, U., Radermacher, M., Mu¨ ller, V., Ruiz, T., and Gru¨ ber, G. (2004b). Three‐ dimensional organization of the archaeal A1‐ATPase from Methanosarcina mazei Go¨ 1. J. Biol. Chem. 279, 22759–22764. Crider, B. P., Andersen, P., White, A. E., Zhou, Z., Li, X., Mattsson, J. P., Lundberg, L., Keeling, D. J., Xie, X. S., Stone, D. K., and Peng, S. B. (1997). Subunit G of the vacuolar proton pump: Molecular characterization and functional expression. J. Biol. Chem. 272, 10721–10728. Del Rizzo, P. A., Bi, Y., Dunn, S. D., and Shilton, B. H. (2002). The ‘‘second stalk’’ of Escherichia coli ATP synthase: Structure of the isolated dimerization domain. Biochemistry 41, 6875–6884. Denda, K., Konishi, J., Oshima, T., Date, T., and Yoshida, M. (1988). Molecular cloning of the b‐subunit of a possible non‐F0F1 type ATPsynthase from the acidothermophilic archaebacterium, Sulfolobus acidocaldarius. J. Biol. Chem. 263, 17251–17254.

374

WILKENS

Diez, M., Zimmermann, B., Bo¨ rsch, M., Konig, M., Schweinberger, E., Steigmu¨ ller, S., Reuter, R., Felekyan, S., Kudryavtsev, V., Seidel, C. A., and Gra¨ ber, P. (2004). Proton‐ powered subunit rotation in single membrane‐bound F0F1‐ATP synthase. Nat. Struct. Mol. Biol. 11, 135–141. Dmitriev, O. Y., Abildgaard, F., Markley, J. L., and Fillingame, R. H. (2004a). Backbone 1 H, 15N and 13C assignments for the subunit a of the E. coli ATP synthase. J. Biomol. NMR 29, 439–440. Dmitriev, O. Y., Altendorf, K., and Fillingame, R. H. (2004b). Subunit a of the E. coli ATP synthase: Reconstitution and high resolution NMR with protein purified in a mixed polarity solvent. FEBS Lett. 556, 35–38. Dmitriev, O., Jones, P. C., Jiang, W., and Fillingame, R. H. (1999). Structure of the membrane domain of subunit b of the Escherichia coli F0F1 ATP synthase. J. Biol. Chem. 274, 15598–15604. Domgall, I., Venzke, D., Lu¨ ttge, U., Ratajczak, R., and Bo¨ ttcher, B. (2002). Three‐ dimensional map of a plant V‐ATPase. J. Biol. Chem. 277, 13115–13121. Drory, O., Frolow, F., and Nelson, N. (2004). Crystal structure of yeast V‐ATPase subunit C reveals its stator function. EMBO Rep. 5, 1148–1152. Dschida, W. J., and Bowman, B. J. (1992). Structure of the vacuolar ATPase from Neurospora crassa as determined by electron microscopy. J. Biol. Chem. 267, 18783–18789. Duncan, T. M., Bulygin, V. V., Zhou, Y., Hutcheon, M. L., and Cross, R. L. (1995). Rotation of subunits during catalysis by Escherichia coli F1‐ATPase. Proc. Natl. Acad. Sci. USA 92, 10964–10968. Dunn, S. D. (1997). Epsilon‐binding regions of the g subunit of Escherichia coli ATP synthase. Biochim. Biophys. Acta 1319, 177–184. Dunn, S. D., McLachlin, D. T., and Revington, M. (2000). The second stalk of Escherichia coli ATP synthase. Biochim. Biophys. Acta 1458, 356–363. Elston, T., Wang, H., and Oster, G. (1998). Energy transduction in ATP synthase. Nature 391, 510–513. Feng, Y., and Forgac, M. (1994). Inhibition of vacuolar H(þ)‐ATPase by disulfide bond formation between cysteine 254 and cysteine 532 in subunit A. J. Biol. Chem. 269, 13224–13230. Fillingame, R. H. (1997). Coupling Hþ transport and ATP synthesis in F1F0‐ATP synthases: Glimpses of interacting parts in a dynamic molecular machine. J. Exp. Biol. 200, 217–224. Fillingame, R. H., Angevine, C. M., and Dmitriev, O. Y. (2002). Coupling proton movements to c‐ring rotation in F(1)F(0) ATP synthase: Aqueous access channels and helix rotations at the a‐c interface. Biochim. Biophys. Acta. 1555, 29–36. Fillingame, R. H., and Dmitriev, O. Y. (2002). Structural model of the transmembrane F0 rotary sector of Hþ‐transporting ATP synthase derived by solution NMR and inter‐ subunit cross‐linking in situ. Biochim. Biophys. Acta 1565, 232–245. Finbow, M. E., and Harrison, M. A. (1997). The vacuolar Hþ‐ATPase: A universal proton pump of eukaryotes. Biochem. J. 324, 697–712. Forgac, M. (1999). Structure and properties of the clathrin‐coated vesicle and yeast vacuolar V‐ATPases. J. Bioenerg. Biomembr. 31, 57–65. Forgac, M. (2000). Structure, mechanism and regulation of the clathrin‐coated and yeast vacuolar Hþ‐ATPases. J. Exp. Biol. 203, 71–80. Futai, M., Oka, T., Sun‐Wada, G.‐H., Moriyama, Y., Kanazawa, H., and Wada, Y. (2000). Luminal acidification of diverse organelles by V‐ATPase in animal cells. J. Exp. Biol. 203, 107–116.

ROTARY MOLECULAR MOTORS

375

Geyer, M., Yu, H., Mandic, R., Linnemann, T., Zheng, Y. H., Fackler, O. T., and Peterlin, B. M. (2002). Subunit H of the V‐ATPase binds to the medium chain of adaptor protein complex 2 and connects Nef to the endocytic machinery. J. Biol. Chem. 277, 28521–28529. Gibbons, C., Montgomery, M. G., Leslie, A. G., and Walker, J. E. (2000). The structure of the central stalk in bovine F(1)‐ATPase at 2.4 A˚ resolution. Nat. Struct. Biol. 7, 1055–1061. Gibson, L. C. D., Cadwallader, G., and Finbow, M. E. (2002). Evidence that there are two copies of subunit c 00 in V0 complexes in the vacuolar HþATPase. Biochem. J. 366, 911–919. Girvin, M. E., Rastogi, V. K., Abildgaard, F., Markley, J. L., and Fillingame, R. H. (1998). Solution structure of the transmembrane Hþ‐transporting subunit c of the F1F0 ATP synthase. Biochemistry 37, 8817–8824. Gogarten, J. P., Kibak, H., Dittrich, P., Taiz, L., Bowman, E. J., Bowman, B. J., Manolson, M. F., Poole, R. J., Date, T., Oshima, T, Konishi, J., Denda, K., and Yoshida, M. (1989). Evolution of the vacuolar Hþ‐ATPase: Implications for the origin of eukaryotes. Proc. Natl. Acad. Sci. USA 86, 6661–6665. Gogol, E. P., Lu¨ cken, U., and Capaldi, R. A. (1987). The stalk connecting the F1 and F0 domains of ATP synthase visualized by electron microscopy of unstained specimens. FEBS Lett. 219, 274–278. Grabar, T. B., and Cain, B. D. (2003). Integration of b subunits of unequal lengths into F1F0‐ATP synthase. J. Biol. Chem. 278, 34751–34756. Gra¨ f, R., Harvey, W. R., and Wieczorek, H. (1996). Purification and properties of a cytosolic V1‐ATPase. J. Biol. Chem. 271, 20908–20913. Greie, J. C., Heitkamp, T., and Altendorf, K. (2004). The transmembrane domain of subunit b of the Escherichia coli F1F0 ATP synthase is sufficient for H(þ)‐translocating activity together with subunits a and c. Eur. J. Biochem. 271, 3036–3042. Groth, G., and Pohl, E. (2001). The structure of the chloroplast F1‐ATPase at 3.2 A˚ resolution. J. Biol. Chem. 276, 1345–1352. Gru¨ ber, G., Radermacher, M., Ruiz, T., Godovac‐Zimmermann, J., Canas, B., Kleine‐ Kohlbrecher, D., Huss, M., Harvey, W. R., and Wieczorek, H. (2000). Three‐dimensional structure and subunit topology of the V1 ATPase from Manduca sexta midgut. Biochemistry 39, 8609–8616. Gru¨ ber, G., Wieczorek, H., Harvey, W. R., and Mu¨ ller, V. (2001). Structure‐function relationships of A‐, F‐ and V‐ATPases. J. Exp. Biol. 204, 2597–2605. Hausrath, A. C., Gru¨ ber, G., Matthews, B. W., and Capaldi, R. A. (1999). Structural features of the gamma subunit of the Escherichia coli F(1) ATPase revealed by a 4.4‐A˚ resolution map obtained by X‐ray crystallography. Proc. Natl. Acad. Sci. USA 96, 13697–13702. Hilario, E., and Gogarten, J. P. (1993). Horizontal transfer of ATPase genes: The tree of life becomes a net of life. Biosystems 31, 111–119. Hilario, E., and Gogarten, J. P. (1998). The prokaryote‐to‐eukaryote transition reflected in the evolution of the V/F/A‐ATPase catalytic and proteolipid subunits. J. Mol. Evol. 46, 703–715. Hirata, T., Iwamotot‐Kihara, A., Sun‐Wada, G.‐H., Okajima, T., Wada, Y., and Futai, M. (2003). Subunit rotation of vacuolar‐type proton pumping ATPase: Relative rotation of the G as to the c subunit. J. Biol. Chem. 278, 23714–23719. Hirata, T., Nakamura, N., Omote, H., Wada, Y., and Futai, M. (2000). Regulation and reversibility of vacuolar Hþ‐ATPase. J. Biol. Chem. 275, 386–389.

376

WILKENS

Imamura, H., Nakano, M., Noji, H., Muneyuki, E., Ohkuma, S., Yoshida, M., and Yokoyama, K. (2003). Evidence for rotation of V‐ATPase. Proc. Natl. Acad. Sci. USA 100, 2312–2315. Itoh, H., Takahashi, A., Adachi, K., Noji, H., Yasuda, R., Yoshida, M., and Kinosita, K. (2004). Mechanically driven ATP synthesis by F1‐ATPase. Nature 427, 465–468. Iwata, M., Imamura, H., Stambouli, E., Ikeda, C., Tamakoshi, M., Nagata, K., Makyio, H., Hankamer, B., Barber, J, Yoshida, M., Yokoyama, K., and Iwata, S. (2004). Crystal structure of a central stalk subunit C and reversible association/dissociation of vacuole‐type ATPase. Proc. Natl. Acad. Sci. USA 101, 59–64. Ja¨ ger, H., Birkenha¨ ger, R., Stalz, W. D., Altendorf, K., and Deckers‐Hebestreit, G. (1998). Topology of subunit a of the Escherichia coli ATP synthase. Eur. J. Biochem. 251, 122–132. Jiang, W., Hermolin, J., and Fillingame, R. H. (2001). The preferred stoichiometry of c subunits in the rotary motor sector of Escherichia coli ATP synthase is 10. Proc. Natl. Acad. Sci. USA 98, 4966–4971. Junge, W., Lill, H., and Engelbrecht, S. (1997). ATP synthase: An electrochemical transducer with rotatory mechanics. Trends Biochem. Sci. 22, 420–423. Kane, P. M. (1995). Disassembly and reassembly of the yeast vacuolar Hþ‐ATPase in vivo. J. Biol. Chem. 270, 17025–17032. Kane, P. M., Yamashiro, C. T., Wolczyk, D. F., Neff, N., Goebl, M., and Stevens, T. H. (1990). Protein splicing converts the yeast TFP1 gene product to the 69‐kD subunit of the vacuolar H(þ)‐adenosine triphosphatase. Science 250, 651–657. Karrasch, S., and Walker, J. E. (1999). Novel features in the structure of bovine ATP synthase. J. Mol. Biol. 290, 379–384. Kato‐Yamada, Y., Noji, H., Yasuda, R., Kinosita, K., Jr, and Yoshida, M. (1998). Direct observation of the rotation of  subunit in F1‐ATPase. J. Biol. Chem. 273, 19375–19377. Landolt‐Marticorena, C., Williams, K. M., Correa, J., Chen, W., and Manolson, M. F. (2000). Evidence that the NH2‐terminus of Vph1p, an integral subunit of the V0 sector of the yeast V‐ATPase, interacts directly with the Vma1p and Vma13p subunits of the V1 sector. J. Biol. Chem. 275, 15449–15457. Lemker, T., Ruppert, C., Stoger, H., Wimmers, S., and Mu¨ ller, V. (2001). Overproduction of a functional A1 ATPase from the archaeon Methanosarcina mazei Go¨ 1 in Escherichia coli. Eur. J. Biochem. 268, 3744–3750. Leng, X. H., Nishi, T., and Forgac, M. (1999). Transmembrane topography of the 100‐ kDa a subunit (Vph1p) of the yeast vacuolar proton‐translocating ATPase. J. Biol. Chem. 274, 14655–14661. Li, Z., and Zhang, X. (2004). Electron‐microscopic structure of the V‐ATPase from mung bean. Planta. 219, 948–954. Liu, Q., Kane., P. M., Newman, P. R., and Forgac, M. (1996). Site‐directed mutagenesis of the yeast V‐ATPase B subunit (Vma2p). J. Biol. Chem. 269, 31592–31597. Liu, Q., Leng, X.‐H., Newman, P. R., Vasilyeva, E., Kane, P. M., and Forgac, M. (1997). Site‐directed mutagenesis of the yeast V‐ATPase A subunit. J. Biol. Chem. 272, 11750–11756. Lolkema, J. S., and Boekema, E. J. (2003). The A‐type ATP synthase subunit K of Methanopyrus kandleri is deduced from its sequence to form a monomeric rotor comprising 13 hairpin domains. FEBS Lett. 543, 47–50. Lu, X., Yu, H., Liu, S. H., Brodsky, F. M., and Peterlin, B. M. (1998). Interactions between HIV1 Nef and vacuolar ATPase facilitate the internalization of CD4. Immunity 8, 647–656.

ROTARY MOLECULAR MOTORS

377

Ludwig, J., Kerscher, S., Brandt, U., Pfeiffer, K., Getlawi, F., Apps, D. K., and Scha¨ gger, H. (1998). Identification and characterization of a novel 9.2‐kDa membrane sector‐ associated protein of vacuolar proton ATPase from chromaffin granules. J. Biol. Chem. 273, 10939–10947. Ma, J., Flynn, T. C., Cui, Q., Leslie, A. G., Walker, J. E., and Karplus, M. (2002). A dynamic analysis of the rotation mechanism for conformational change in F(1)‐ATPase. Structure 10, 921–931. Mac Leod, K. J., Vasilyeva, E., Baleja, J. D., and Forgac, M. (1998). Mutational analysis of the nucleotide binding sites of the yeast vacuolar proton‐translocating ATPase. JBiolChem. 273, 150–156. Mandel, M., Moriyama, Y., Hulmes, J. D., Pan, Y. C., Nelson, H., and Nelson, N. (1988). cDNA sequence encoding the 16‐kDa proteolipid of chromaffin granules implies gene duplication in the evolution of Hþ‐ATPases. Proc. Natl. Acad. Sci. USA 85, 5521–5524. McEnery, M. W., Buhle, E. L., Jr, Aebi, U., and Pedersen, P. L. (1984). Proton ATPase of rat liver mitochondria: Preparation and visualization of a functional complex using the novel zwitterionic detergent 3‐[(3‐cholamidopropyl)dimethylammonio]‐1‐ propanesulfonate. J. Biol. Chem. 259, 4642–4651. McLachlin, D. T., Bestard, J. A., and Dunn, S. D. (1998). The b and d subunits of the Escherichia coli ATP synthase interact via residues in their C‐terminal regions. J. Biol. Chem. 273, 15162–15168. Mellwig, C., and Bo¨ ttcher, B. (2003). A unique resting position of the ATP‐synthase from chloroplasts. J. Biol. Chem. 278, 18544–18549. Menz, R. I., Walker, J. E., and Leslie, A. G. (2001). Structure of bovine mitochondrial F(1)‐ATPase with nucleotide bound to all three catalytic sites: Implications for the mechanism of rotary catalysis. Cell 106, 331–341. Merzendorfer, H., Huss, M., Schmid, R., Harvey, W. R., and Wieczorek, H. (1999). A novel insect V‐ATPase subunit M9.7 is glycosylated extensively. J. Biol. Chem. 274, 17372–17378. Mitchell, P. (1961). Coupling of phophorylation to electron and hydrogen transfer by a chemi‐ osmotic type of mechanism. Nature 191, 144–148. Mitome, N., Suzuki, T., Hayashi, S., and Yoshida, M. (2004). Thermophilic ATP synthase has a decamer c‐ring: Indication of noninteger 10:3 Hþ/ATP ratio and permissive elastic coupling. Proc. Natl. Acad. Sci. USA 101, 12159–12164. Mu¨ ller, V. (2004). An exceptional variability in the motor of archael A1A0 ATPases: From multimeric to monomeric rotors comprising 6–13 ion binding sites. J. Bioenerg. Biomembr. 36, 115–125. Mu¨ ller, V., and Gru¨ ber, G. (2003). ATP synthases: Structure, function and evolution of unique energy converters. Cell. Mol. Life Sci. 60, 474–494. Mu¨ ller, V., Ruppert, C., and Lemker, T. (1999). Structure and function of the A1A0‐ ATPases from methanogenic Archaea. J. Bioenerg. Biomembr. 31, 15–27. Murata, T., Arechaga, I., Fearnley, I. M., Kakinuma, Y., Yamato, I., and Walker, J. E. (2003). The membrane domain of the Naþ‐motive V‐ATPase from Enterococcus hirae contains a heptameric rotor. J. Biol. Chem. 278, 21162–21167. Nelson, N., and Harvey, W. R. (1999). Vacuolar and plasma membrane V‐ATPases. Physiol. Rev. 79, 361–385. Nelson, N., and Taiz, L. (1989). The evolution of Hþ‐ATPases. Trends. Biochem. Sci. 14, 113–116. Nishi, T., and Forgac, M. (2002). The vacuolar (Hþ)‐ATPases: Nature’s most versatile proton pumps. Nat. Rev. Mol. Cell Biol. 3, 94–103.

378

WILKENS

Nishio, K., Iwamoto‐Kihara, A., Yamamoto, A., Wada, Y., and Futai, M. (2002). Subunit rotation of ATP synthase embedded in membranes: a or b subunit rotation relative to the c subunit ring. Proc. Natl. Acad. Sci. USA 99, 13448–13452. Nishizaka, T., Oiwa, K., Noji, H., Kimura, S., Muneyuki, E., Yoshida, M., and Kinosita, K, Jr. (2004). Chemomechanical coupling in F1‐ATPase revealed by simultaneous observation of nucleotide kinetics and rotation. Nat. Struct. Mol. Biol. 11, 142–148. Noji, H., Hasler, K., Junge, W., Kinosita, K., Jr, Yoshida, M., and Engelbrecht, S. (1999). Rotation of Escherichia coli F(1)‐ATPase. Biochem. Biophys. Res. Commun. 260, 597–599. Noji, H., Yasuda, R., Yoshida, M., and Kinosita, K., Jr. (1997). Direct observation of the rotation of F1‐ATPase. Nature 386, 299–302. Ogilvie, I., Wilkens, S., Rodgers, A. J., Aggeler, R., and Capaldi, R. A. (1997). The second stalk: The delta‐b subunit connection in ECF1F0. Acta Physiol. Scand. Suppl. 643, 169–175. Oster, G., and Wang, H. (2000). Why is the mechanical efficiency of F1‐ATPase so high? J. Bioenerg. Biomembr. 32, 459–469. Oster, G., and Wang, H. (2003). Rotary protein motors. Trends Cell Biol. 13, 114–121. Pa´ li, T., Finbow, M. E., Holzenburg, A., Findlay, J. B., and Marsh, D. (1995). Lipid‐ protein interactions and assembly of the 16‐kDa channel polypeptide from Nephrops norvegicus: Studies with spin‐label electron spin resonance spectroscopy and electron microscopy. Biochemistry 34, 9211–9218. Pa¨ nke, O., Gumbiowski, K., Junge, W., and Engelbrecht, S. (2000). F‐ATPase: Specific observation of the rotating c subunit oligomer of EF0EF1. FEBS Lett. 472, 34–38. Parra, K. J., Keenan, K. L., and Kane, P. M. (2000). The H subunit (Vma13p) of the yeast V‐ATPase inhibits the ATPase activity of cytosolic V1 complexes. J. Biol. Chem. 275, 21761–21767. Penefsky, H. S., and Cross, R. L. (1991). Structure and Mechanism of F1F0 ATP synthases and ATPases. Adv. Enzymol. 64, 173–214. Powell, B., Graham, L. A., and Stevens, T. H. (2000). Molecular characterization of the yeast vacuolar Hþ‐ATPase proton pore. J. Biol. Chem. 275, 23654–23660. Radermacher, M., Ruiz, T., Harvey, W. R., Wieczorek, H., and Gru¨ ber, G. (1999). Molecular architecture of Manduca sexta midgut V1 ATPase visualized by electron microscopy. FEBS Lett. 453, 383–386. Rastogi, V. K., and Girvin, M. E. (1999). Structural changes linked to proton translocation by subunit c of the ATP synthase. Nature 402, 263–268. Rizzo, V. F., Coskun, U., Radermacher, M., Ruiz, T., Armbru¨ ster, A., and Gru¨ ber, G. (2003). Resolution of the V1 ATPase from Manduca sexta into subcomplexes and visualization of an ATPase‐active A3B3EG complex by electron microscopy. J. Biol. Chem. 278, 270–275. Rodgers, A. J. W., and Wilce, M. C. (2000). Structure of the g‐ complex of ATP synthase. Nat. Struct. Biol. 7, 1051–1054. Rodgers, A. J. W., Wilkens, S., Aggeler, R., Morris, M. B., Howitt, S. M., and Capaldi, R. A. (1997). The subunit d‐subunit b domain of the Escherichia coli F1F0 ATPase: The b subunits interact with F1 as a dimer and through the d subunit. J. Biol. Chem. 272, 31058–31064. Rubinstein, J. L., Walker, J. E., and Henderson, R. (2003). Structure of the mitochondrial ATP synthase by electron cryomicroscopy. EMBO J. 22, 6182–6192. Sabbert, D., Engelbrecht, S., and Junge, W. (1996). Intersubunit rotation in active F‐ATPase. Nature 381, 623–626.

ROTARY MOLECULAR MOTORS

379

Sagermann, M., Stevens, T. H., and Matthews, B. W. (2001). Crystal structure of the regulatory subunit H of the V‐type ATPase of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 98, 7134–7139. Sambade, M., and Kane, P. M. (2004). The yeast V‐ATPase contains a subunit homologous to the M. sexta and bovine e subunits that is essential for function. J. Biol. Chem. 279, 17361–17365. Sambongi, Y., Iko, Y., Tanabe, M., Omote, H., Iwamoto‐Kihara, A., Ueda, I., Yanagida, T., Wada, Y., and Futai, M. (1999). Mechanical rotation of the c subunit oligomer in ATP synthase (F0F1): Direct observation. Science 286, 1722–1724. Scha¨ fer, G., and Meyering‐Vos, M. (1992). F‐type or V‐type? The chimeric nature of the archaebacterial ATP synthase. Biochim. Biophys. Acta 1101, 232–235. Schneider, E., and Altendorf, K. (1985). All three subunits are required for the reconstitution of an active proton channel (F0) of Escherichia coli ATP synthase (F1F0). EMBO J. 4, 515–518. Seelert, H., Poetsch, A., Dencher, N. A., Engel, A., Stahlberg, H., and Mu¨ ller, D. J. (2000). Structural biology: Proton‐powered turbine of a plant motor. Nature 405, 418–419. Senior, A. E. (1988). ATP synthesis by oxidative phosphorylation. Physiol. Rev. 68, 177–231. Shao, E., and Forgac, M. (2004). Involvement of the nonhomologous region of subunit A of the yeast V‐ATPase in coupling and in vivo dissociation. J. Biol. Chem. 279, 48663–48670. Shao, E., Nishi, T., Kawasaki‐Nishi, S., and Forgac, M. (2003). Mutational analysis of the non‐homologous region of subunit A of the yeast V‐ATPase. J. Biol. Chem. 278, 12985–12991. Shirakihara, Y., Leslie, A. G., Abrahams, J. P., Walker, J. E., Ueda, T., Sekimoto, Y., Kambara, M., Saika, K., Kagawa, Y., and Yoshida, M. (1997). The crystal structure of the nucleotide‐free a3b3 subcomplex of F1‐ATPase from the thermophilic Bacillus PS3 is a symmetric trimer. Structure 5, 825–836. Smardon, A. M., Tarsio, M., and Kane, P. M. (2002). The RAVE complex is essential for stable assembly of the yeast V‐ATPase. J. Biol. Chem. 277, 13831–13839. Smith, A. N., Lovering, R. C., Futai, M., Takeda, J., Brown, D., and Karet, F. E. (2003). Revised nomenclature for mammalian vacuolar‐type Hþ‐ATPase subunit genes. Mol. Cell 12, 801–803. Sorgen, P. L., Bubb, M. R., and Cain, B. D. (1999). Lengthening the second stalk of F(1) F(0) ATP synthase in Escherichia coli. J. Biol. Chem. 274, 36261–36266. Sorgen, P. L., Caviston, T. L., Perry, R. C., and Cain, B. D. (1998). Deletions in the second stalk of F1F0‐ATP synthase in Escherichia coli. J. Biol. Chem. 273, 27873–27878. Speelmans, G., Poolman, B., Abee, T., and Konings., W. N. (1994). The F‐ or V‐type Na (þ)‐ATPase of the thermophilic bacterium Clostridium fervidus. J. Bacteriol. 176, 5160–5162. Stahlberg, H., Mu¨ ller, D. J., Suda, K., Fotiadis, D., Engel, A., Meier, T., Matthey, U., and Dimroth, P. (2001). Bacterial Na(þ)‐ATP synthase has an undecameric rotor. EMBO Reports 2, 229–233. Stevens, T. H., and Forgac, M. (1997). Structure, function and regulation of the vacuolar (Hþ)‐ATPase. Annu. Rev. Cell. Dev. Biol. 13, 779–808. Stock, D., Leslie, A. G. W., and Walker, J. E. (1999). Molecular architecture of the rotary motor in ATP synthase. Science 286, 1700–1705.

380

WILKENS

Sumner, J. P., Dow, J. A., Earley, F. G., Klein, U., Ja¨ ger, D., and Wieczorek, H. (1995). Regulation of plasma membrane V‐ATPase activity by dissociation of peripheral subunits. J. Biol. Chem. 270, 5649–5653. Supekova, L., Sbia, M., Supek, F., Ma, Y., and Nelson, N. (1996). A novel subunit of vacuolar H(þ)‐ATPase related to the b subunit of F‐ATPases. J. Exp. Biol. 199, 1147–1156. Suzuki, T., Murakami, T., Iino, R., Suzuki, J., Ono, S., Shirakihara, Y., and Yoshida, M. (2003). F0F1‐ATPase/synthase is geared to the synthesis mode by conformational rearrangement of  subunit in response to proton motive force and ADP/ATP balance. J. Biol. Chem. 278, 46840–46846. Suzuki, T., Ueno, H., Mitome, N., Suzuki, J., and Yoshida, M. (2002). F(0) of ATP synthase is a rotary proton channel: Obligatory coupling of proton translocation with rotation of c‐subunit ring. J. Biol. Chem. 277, 13281–13285. Takase, K., Kakinuma, S., Yamato, I., Konishi, K., Igarashi, K., and Kakinuma, Y. (1994). Sequencing and characterization of the ntp gene cluster for vacuolar‐type Na(þ)‐ translocating ATPase of Enterococcus hirae. J. Biol. Chem. 269, 11037–11044. Tsunoda, S. P., Aggeler, R., Yoshida, M., and Capaldi, R. A. (2001a). Rotation of the c subunit oligomer in fully functional F1F0 ATP synthase. Proc. Natl. Acad. Sci. USA 98, 898–902. Tsunoda, S. P., Rodgers, A. J., Aggeler, R., Wilce, M. C., Yoshida, M., and Capaldi, R. A. (2001b). Large conformational changes of the  subunit in the bacterial F1F0 ATP synthase provide a ratchet action to regulate this rotary motor enzyme. Proc. Natl. Acad. Sci. USA 98, 6560–6564. Tsuprun, V. L., Orlova, E. V., and Mesyanzhinova, I. V. (1989). Structure of the ATP‐ synthase studied by electron microscopy and image processing. FEBS Lett. 244, 279–282. Ubbink‐Kok, T., Boekema, E. J., van Breemen, J. F. L., Brisson, A., Konings, W. N., and Lolkema, J. S. (2000). Stator structure and subunit composition of the V1/V0 Naþ‐ ATPase of the thermophilic bacterium Caloramator fervidus. J. Mol. Biol. 296, 311–321. Ueno, H., Suzuki, T., Kinosita, K., Jr, and Yoshida, M. (2005). ATP‐driven stepwise rotation of F0F1‐ATP synthase. Proc. Natl. Acad. Sci. USA 102, 1333–1338. Uhlin, U., Cox, G. B., and Guss, J. M. (1997). Crystal structure of the  subunit of the proton‐translocating ATP synthase from Escherichia coli. Structure 15, 1219–1230. Vasilyeva, E., and Forgac, M. (1996). 3’‐O‐(4‐Benzoyl)benzoyladenosine 5’‐triphosphate inhibits activity of the vacuolar (Hþ)‐ATPase from bovine brain clathrin‐coated vesicles by modification of a rapidly exchangeable, non‐catalytic nucleotide binding site on the B subunit. J. Biol. Chem. 271, 12775–12782. Vik, S. B., and Antonio, B. J. (1994). A mechanism of proton translocation by F1F0 ATP synthases suggested by double mutants of the a subunit. J. Biol. Chem. 269, 30364–30369. Vik, S. B., Long, J. C., Wada, T., and Zhang, D. (2000). A model for the structure of subunit a of the Escherichia coli ATP synthase and its role in proton translocation. Biochim. Biophys. Acta 1458, 457–466. Weber, J., Muharemagic, A., Wilke‐Mounts, S., and Senior, A. E. (2003). F1F0‐ATP synthase: Binding of d subunit to a 22‐residue peptide mimicking the N‐terminal region of a subunit. J. Biol. Chem. 278, 13623–13626. Weber, J., and Senior, A. E. (1997). Catalytic mechanism of F1‐ATPase. Biochem. Biophys. Acta 1319, 19–58.

ROTARY MOLECULAR MOTORS

381

Wieczorek, H., Gru¨ ber, G., Harvey, W. R., Huss, M., Merzendorfer, H., and Zeiske, W. (2000). Structure and regulation of insect plasma membrane Hþ‐V‐ATPase. J. Exp. Biol. 203, 127–135. Wilkens, S. (2000). F1F0‐ATPsynthase: Stalking mind and imagination. J. Bioenerg. Biomembr. 32, 333–339. Wilkens, S. (2001). Structure of the vacuolar adenosine triphosphatases. Cell Biochem. Biophys. 34, 191–208. Wilkens, S., and Capaldi, R. A. (1998a). Electron microscopic evidence for two stalks linking the F1 and F0 in the ATPsynthase from Escherichia coli. Biochem. Biophys. Acta 1365, 93–97. Wilkens, S., and Capaldi, R. A. (1998b). ATPsynthase’s second stalk comes into focus. Nature 393, 29. Wilkens, S., and Capaldi, R. A. (1998c). Solution structure of the  subunit of the F1‐ATPase from Escherichia coli and interactions of this subunit with b subunits in the complex. J. Biol. Chem. 273, 26645–26651. Wilkens, S., Dahlquist, F., McIntosh, L., Donaldson, L., and Capaldi, R. A. (1995). Structural features of the  subunit of the Escherichia coli ATP synthase (ECF1F0) from NMR spectroscopy. Nature Struct. Biol. 2, 961–967. Wilkens, S., Dunn, S. D., Chandler, J., Dahlquist, F. W., and Capaldi, R. A. (1997). Solution structure of the N‐terminal domain (residues 1–134) of the d subunit of the Escherichia coli F1F0 ATPsynthase. Nature Struct. Biol. 4, 198–201. Wilkens, S., and Forgac, M. (2001). Three‐dimensional structure of the vacuolar ATPase proton channel by electron microscopy. J. Biol. Chem. 276, 44064–44068. Wilkens, S., Inue, T., and Forgac, M. (2004). Three‐dimensional structure of the vacuolar ATPase – Localization of subunit H by difference imaging and chemical cross‐ linking. J. Biol. Chem. 279, 41942–41949. Wilkens, S., Vasilyeva, E., and Forgac, M. (1999). Structure of the vacuolar ATPase by electron microscopy. J. Biol. Chem. 274, 31804–31810. Wilkens, S., Zhang, Z., and Zheng, Y. (2005). A structural model of the vacuolar ATPase from transmission electron microscopy. Micron 36, 109–126. Wilkens, S., Zhou, J., Nakayama, R., Dunn, S. D., and Capaldi, R. A. (2000). Localization of the d subunit in the Escherichia coli F1F0‐ATPsynthase by immuno‐electron microscopy: The d subunit binds on top of the F1. J. Mol. Biol. 295, 387–391. Xie, X. S. (1996). Reconstitution of ATPase activity from individual subunits of the clathrin‐coated vesicle proton pump. J. Biol. Chem. 271, 30980–30985. Xie, X. S., Crider, B. P., Ma, Y. M., and Stone, D. K. (1994). Role of a 50–57‐kDa polypeptide heterodimer in the function of the clathrin‐coated vesicle proton pump. J. Biol. Chem. 269, 25809–25815. Xie, X. S., and Stone, D. K. (1988). Partial resolution and reconstitution of the subunits of the clathrin‐coated vesicle proton ATPase responsible for Ca2þ‐activated ATP hydrolysis. J. Biol. Chem. 263, 9859–9867. Xu, T., Vasilyeva, E., and Forgac, M. (1999). Subunit interactions in the clathrin‐coated vesicle V‐ATPase complex. J. Biol. Chem. 274, 28909–28915. Yamada, H., Moriyama, Y., Maeda, M., and Futai, M. (1996). Transmembrane topology of Escherichia coli H(þ)‐ATPase (ATP synthase) subunit a. FEBS Lett. 390, 34–38. Yasuda, R., Noji, H., Yoshida, M., Kinosita, K., Jr, and Itoh, H. (2001). Resolution of distinct rotational substeps by submillisecond kinetic analysis of F1‐ATPase. Nature 410, 898–904.

382

WILKENS

Yokoyama, K., Nagata, K., Imamura, H., Ohkuma, S., Yoshida, M., and Tamakoshi, M. (2003a). Subunit arrangement in V‐ATPase from Thermus thermophilus. J. Biol. Chem. 278, 42686–42691. Yokoyama, K., Nakano, M., Imamura, H., Yoshida, M., and Tamakoshi, M. (2003b). Rotation of the proteolipid ring in the V‐ATPase. J. Biol. Chem. 278, 24255–24258. Yokoyama, K., Oshima, T., and Yoshida, M. (1990). Thermus thermophilus membrane‐ associated ATPase: Indication of a eubacterial V‐type ATPase. J. Biol. Chem. 265, 21946–21950. Yoshida, M., Muneyuki, E., and Hisabori, T. (2001). ATP synthase: A marvellous rotary engine of the cell. Nat. Rev. Mol. Cell Biol. 2, 669–677. Zhang, J., Myers, M., and Forgac, M. (1992). Characterization of the V0 domain of the coated vesicle (Hþ)‐ATPase. J. Biol. Chem. 267, 9773–9778. Zhang, J., Vasilyeva, E., Feng, Y., and Forgac, M. (1995). Inhibition and labelling of the coated vesicle V‐ATPase by 2‐azido‐[32P]ATP. J. Biol. Chem. 270, 15494–15500. Zhang, X., Malhotra, R., and Guidotti, G. (2000). Regulation of yeast ecto‐apyrase ynd1p by activating subunit Vma13p of the vacuolar Hþ‐ATPase. J. Biol. Chem. 275, 35592–35599. Zhang, Y., Oldenburg, M., and Fillingame, R. H. (1994). Suppressor mutations in F1 subunit  recouple ATP‐driven Hþ translocation in uncoupled Q42E subunit c mutant of Escherichia coli F1F0 ATP synthase. J. Biol. Chem. 269, 10221–10224. Zhang, Z., Charsky, C., Kane, P. M., and Wilkens, S. (2003). Yeast V1‐ATPase: Affinity purification and structural features by electron microscopy. J. Biol. Chem. 278, 47299–47306. Zhou, Y., Duncan, T. M., and Cross, R. L. (1997). Subunit rotation in Escherichia coli F0F1‐ATP synthase during oxidative phosphorylation. Proc. Natl. Acad. Sci. USA 94, 10583–10587. Zimniak, L., Dittrich, P., Gogarten, J. P., Kibak, H., and Taiz, L. (1988). The cDNA sequence of the 69‐kDa subunit of the carrot vacuolar Hþ‐ATPase. Homology to the b‐chain of F0F1‐ATPases. J. Biol. Chem. 263, 9102–9112.

CYTOSKELETON DYNAMICS POWERS NEMATODE SPERM MOTILITY By MURRAY STEWART* AND THOMAS M. ROBERTS{ {

*MRC Laboratory of Molecular Biology, Hills Rd, Cambridge CB2 2QH, England; Department of Biological Science, Florida State University, Tallahassee, Florida 32306

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Comparison Between MSP and Actin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Locomotion in Nematode Sperm is Coupled to Assembly and Disassembly of the Cytoskeleton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Reconstitution of Lamellipodial Protrusion In Vitro . . . . . . . . . . . . . . . . . . . . . . . . . V. Retraction is Also Required for Crawling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Reconstitution of Retraction In Vitro. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. A Push‐Pull Model for Nematode Sperm Amoeboid Motility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

384 385 387 389 391 393 395 397 397

ABSTRACT Nematode sperm provide a simple and specialized system for studying the molecular mechanism of amoeboid cell motility. Locomotion is generated by the assembly dynamics of their cytoskeleton, which is based on the major sperm protein (MSP). Protrusive force is generated at the leading edge of the lamellipod by MSP filament formation and bundling, whereas the contractile force that drags the rearward cell body forward is generated by cytoskeleton disassembly. The dynamics of the system can be reconstituted in vitro using cell‐free extracts of Ascaris sperm, in which vesicles derived from the leading edge of the cell can be either pushed or pulled. The addition of ATP to the cell‐free extract initiates MSP filament polymerization and bundling immediately behind the vesicle, and the expansion of the resulting gel pushes the vesicle at rates comparable to those seen in living cells. In contrast, the addition of Yersinia tyrosine phosphatase generates depolymerization and gel contraction that pulls the vesicles. Overall, nematode sperm motility illustrates that cell locomotion can be generated by cytoskeletal dynamics alone without the use of myosin‐like motor proteins.

ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71010-4

383

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

384

STEWART AND ROBERTS

I.

INTRODUCTION

Many eukaryotic cells move by crawling over a substrate. This movement requires the establishment of cell polarity, extension of the leading edge of the lamellipod, attachment to the substratum, and finally, retraction that drags the cell body forward. Each of these events depends on the dynamics of the actin cytoskeleton that, in turn, are orchestrated by a large number of actin‐binding proteins and signaling molecules. Although myosinlike motors were originally thought to be central components of amoeboid cell motility, a consensus is now emerging that emphasizes primarily the contribution made by the vectorial assembly and organization of cytoskeletal filaments to the generation of the forces involved in both protrusion and retraction (Miao et al., 2003; Pantaloni et al., 2001; Pollard and Borisy, 2003). However, because amoeboid cell motility involves so many components, it has been difficult to formulate models to account completely for the generation of protrusion and retraction at the molecular level and how they are coordinated with adhesion and integrated to generate locomotion. The amoeboid sperm of nematodes are a simple, more specialized system in which the role usually played by actin has been taken over by the 14‐kDa MSP. Although nematode sperm do not contain appreciable amounts of F‐actin, the cells display the characteristic features of amoeboid locomotion. For example, Ascaris sperm extend a persistent flattened lamellipod that attaches to the substrate and pulls along a trailing, organelle‐packed cell body. The lamellipod is packed with filaments that assemble along the leading edge and flow rearward as the cell progresses in a manner analogous to that observed for the actin cytoskeleton in a number of other crawling cells (reviewed in Mogilner and Oster, 2003c; Pollard and Borisy, 2003; Roberts and Stewart, 2000; Theriot, 1996). Indeed, MSP‐based and actin‐based cell crawling resemble each other so closely that, although they employ different sets of molecular components to generate movement, the physical and mechanical principles that generate force must be essentially the same. This model system has given a number of insights into the general mechanism of cell motility and has indicated that, at least in nematode sperm, locomotion appears to be produced primarily by a push‐pull mechanism based on MSP assembly dynamics (Miao et al., 2003; Mogilner and Oster, 2003c; Roberts and Stewart, 2000). Moreover, in addition to providing a novel molecular perspective for studying amoeboid cell motility, nematode sperm also offer several advantages as an experimental system that complements those provided by actin‐containing cells. For example, now that many of the molecules that organize and regulate the actin cytoskeleton have been identified, attention is shifting

AMOEBOID CELL MOTILITY WITHOUT MOTORS

385

to understanding how those molecules interact to produce movement (reviewed in Borisy and Svitkina, 2000; Pantaloni et al., 2001; Pollard and Borisy, 2003). However, this task is complicated by the central role played by actin in a range of cellular functions in addition to locomotion, including shape determination, establishment of cell polarity, endocytosis, movement of organelles, rearrangement of surface components, and cytokinesis. Nematode sperm, by contrast, use their MSP‐based motility system exclusively for locomotion. Furthermore, the organization of the cytoskeleton can be observed directly in crawling Ascaris sperm cells. These advantages have been exploited, together with the ability to dissect and reconstitute the MSP machinery, to compare its operation to that present in actin‐based cells and thereby identify some of the fundamental principles governing the generation of amoeboid cell motility. Remarkably, it seems that, at least in this system, protrusion of the leading edge and the retraction that pulls the cell body forward are both generated primarily by the assembly dynamics of the cytoskeleton without the need for molecular motor proteins. Thus, amoeboid cell motility appears to be generated primarily by a polymerization motor (see Mogilner and Oster, 2003a,b).

II.

COMPARISON BETWEEN MSP AND ACTIN

Although MSP and actin lie at the core of similar motile systems, the two proteins have surprisingly little in common. Both are abundant cellular components (Ascaris sperm contain
4 mM MSP) that can self‐assemble into filaments and, in turn, into multifilament arrays. However, the proteins themselves have no sequence homology, no structural similarity, and form filaments with quite different structural and polymerization properties (Fig. 1). MSP contains only 126 residues and is based on an immunoglobulin fold that has no counterpart in the four‐domain structure of actin (Bullock et al., 1996; King et al., 1992). In contrast with actin, MSP does not bind nucleotides and the polymerizing unit is a dimer rather than a monomer (Haaf et al., 1996; Italiano et al., 1996). Both proteins assemble into two‐stranded polymers, but in actin filaments the subunits in each stand are arranged like beads on a string, whereas MSP filaments are constructed from two loosely connected helical subfilaments (Stewart et al., 1994). X‐ray crystallography (Baker et al., 2002; Bullock et al., 1998) has established that each subfilament is constructed from a helix of MSP dimers and these helices then wrap around one another (like a telephone cord) to generate a filament. The most striking difference between MSP and actin, from the standpoint of the mechanism of motility, lies in the polarity of the filaments

386

STEWART AND ROBERTS

FIG. 1. Ribbon diagrams of Dictyostelium discoideum G‐actin (Matsuura et al., 2000) and the Ascaris suum a˜MSP dimer (Bullock et al., 1996) at the same magnification. Actin consists of four subdomains that surround a nucleotide‐binding cleft. The G‐actin molecule is asymmetric, so that when it polymerizes, the filament it forms has a characteristic polarity and its two ends differ structurally. By contrast, MSP does not contain a nucleotide binding site and the polymerizing unit is a dimer in which the two MSP molecules are related by twofold rotational symmetry. Polymerization produces filaments composed of two helical subfilaments in which the dimers’ twofold axes are oriented perpendicular to the helix axis. Consequently, the MSP helices have no polarity and the subfilament ends are identical structurally (Bullock et al., 1998). Reproduced from The Journal of Cell Biology, 2000, vol. 149, pp. 7–12 by copyright permission of the Rockefeller University Press.

AMOEBOID CELL MOTILITY WITHOUT MOTORS

387

they form. Actin filaments have a characteristic structural polarity based on their having ‘‘barbed’’ and ‘‘pointed’’ ends. This intrinsic actin filament polarity influences the pattern and regulation of cytoskeletal assembly and also defines the direction in which filaments are moved by myosin family molecular motors. MSP filaments do not have any overall polarity. The two polypeptide chains in the dimers from which MSP filaments are constructed are related by twofold rotational symmetry (Bullock et al., 1996). In filaments, the dimer twofold axes are parallel to the subfilament helix axis (Bullock et al., 1998). Thus, the two ends of each subfilament are identical structurally and so the filaments formed from these subfilaments are nonpolar. Therefore, in contrast to F‐actin, in which structural differences between the two ends of the polymer modulate the rates of subunit addition and loss, MSP filament polymerization must be controlled entirely by external factors. Moreover, it is unlikely that molecular motor proteins analogous to myosin could function on an apolar MSP filament because the motor could exert force in either direction along the filament axis. This observation has focused attention on the continuous remodeling of the cytoskeleton that accompanies sperm locomotion as the principal source of the forces required for motility.

III.

LOCOMOTION IN NEMATODE SPERM IS COUPLED TO ASSEMBLY AND DISASSEMBLY of THE CYTOSKELETON

The cytoskeleton in Ascaris sperm can be imaged directly in live cells without resorting to the labeled probes that are often needed to detect actin filaments (Fig. 2). Filaments are assembled and organized into meshworks along the advancing front, and then flow rearward to the region occupied by the cell body where they disassemble and release subunits that can be recycled to the leading edge for reassembly. In crawling Ascaris sperm, MSP filaments are arranged into long branched meshworks, called fiber complexes, which span the entire length of the lamellipodium. Filaments extend radially from each fiber complex and are able to interact with similar filaments from adjacent complexes so that the entire cytoskeleton functions as an interconnected unit (Sepsenwol et al., 1989). Filaments are assembled and begin to incorporate into fiber complexes at small protrusions along the leading edge of the lamellipod, and, as the cell moves forward, the fiber complexes treadmill from front to rear through the lamellipodium without major changes in their shape or filament packing density. The rate of cytoskeletal treadmilling is tightly coupled to the speed of sperm locomotion (Fig. 2). Not all crawling cells exhibit such a close correlation between cytoskeletal dynamics and locomotion. Often, as in fibroblasts (Wang, 1985) and

388

STEWART AND ROBERTS

FIG. 2. Cytoskeletal dynamics can be observed directly by light microscopy in crawling Ascaris. The figure shows two images, taken 10 sec apart, of a sperm crawling on a glass coverslip. The MSP fiber complexes are readily visible in the lamellipodium, so their dynamics can be followed in real time. The rate of cytoskeletal assembly along the leading edge and disassembly at the base of the lamellipodium is tightly coupled to locomotion. Thus, fiduciary markers such as branches in the fiber complexes (arrow) flow centripetally through the lamellipodium but remain nearly stationary with respect to the substratum. Bar, 5 mm. Reproduced from Current Opinion in Cell Biology, vol. 7, pp. 13–17 (1995), copyright Current Biology Ltd.

Aplysia neuronal growth cones (Forscher and Smith, 1988), the rate of cytoskeletal treadmilling outpaces the speed of translocation. In fish epithelial keratocytes, the rate of localized actin cytoskeletal assembly matches that of leading edge protrusion, but cytoskeletal disassembly occurs throughout the lamellipodium (Theriot and Mitchison, 1991). In Ascaris sperm, cytoskeletal assembly and disassembly occur at opposite ends of the lamellipodium, 15–20 mm apart, but at the same rate. Thus, elongation of the fiber complexes appears to push the plasma membrane forward, allowing the leading edge to advance, while simultaneously the cell body is pulled forward as the cytoskeleton disassembles at the base of the lamellipodium. Methods have now been developed to uncouple MSP cytoskeletal assembly and disassembly and explore their independent contributions to sperm locomotion as well as to reconstitute both protrusion and retraction in vitro.

AMOEBOID CELL MOTILITY WITHOUT MOTORS

IV.

389

RECONSTITUTION OF LAMELLIPODIAL PROTRUSION IN VITRO

In actin‐based cells, the localized cytoskeletal assembly that occurs at the leading edge suggested that this pattern of filament polymerization process itself may be the primary mechanism by which protrusive force is generated (Pollard and Borisy, 2003). Reconstitution of MSP‐based lamellipodial protrusion in cell‐free extracts of Ascaris sperm has provided direct confirmation of this hypothesis (Italiano et al., 1996). Addition of ATP to sperm extract induces the formation of discrete meshworks of MSP filaments, called fibers, each of which has a membrane vesicle at one end (Fig. 3). Growth of these fibers is due to the assembly of filaments at the vesicle‐bearing end, which produces vectorial movement of the vesicle. Immunolabeling indicates that the vesicles that build fibers derive from the plasma membrane at the leading edge of the lamellipodium. Thus, simply adding ATP to a crude cell extract can reconstitute lamellipodial

FIG. 3. Leading edge dynamics can be reconstituted in vitro using cell‐free extracts of Ascaris sperm in which vesicles derived from the plasma membrane induce the assembly of MSP filament meshworks called fibers that push the vesicles forward as they elongate. The two images were taken 10 sec apart. Bar, 2.5 mm. Reproduced from The Journal of Cell Biology, 1999, vol. 146, pp. 1087–1095 by copyright permission of the Rockefeller University Press.

390

STEWART AND ROBERTS

protrusion: a fragment of the leading edge membrane triggers polymerization and bundling of a meshwork of filaments that moves a vesicle in the same way as elongation of the fiber complexes appears to push the lamellipodial membrane forward in crawling cells. MSP does not bind ATP and is not phosphorylated. Thus, ATP appears to be used indirectly, but its exact role in protrusion still needs to be defined. The in vitro motility system has also been used to identify additional components of the nematode cell motility apparatus. Analysis of the effects of pressure (Roberts et al., 1998) and of dilution of the sperm extract (Italiano et al., 1996) indicated that, in addition to MSP, motility required additional cytosolic components as well as at least one membrane protein from the leading edge of the lamellipod. Two of these cytosolic proteins, designated MSP fiber proteins (MFPs) 1 and 2, have now been purified and characterized (Buttery et al., 2003). Immunolabeling demonstrated that these proteins are located throughout the MSP cytoskeleton in cells and in fibers. Both of these cytosolic proteins influence the rate of MSP fiber assembly in vitro: MFP1 decreased the rate and MFP2 increased the rate. In addition, a cell‐membrane phosphoprotein has been identified that stimulates MSP polymerization (LeClaire et al., 2003) and is a strong candidate for the membrane component that ensures that MSP polymerization occurs primarily close to the vesicles in the in vitro motility system and adjacent to the leading edge of the lamellipod in vivo. Comparison to actin‐based motility indicated that there may be conserved mechanisms, albeit involving different biochemical components, to control the rate of polymerization in both MSP and actin‐based movement. MSP‐driven vesicle motility resembles a number of specialized actin‐ based motile systems typified by the movement of Listeria monocytogenes (reviewed in Carlier et al., 2002). This intracellular bacterial pathogen commandeers proteins from its host cell to build a columnar meshwork of actin filaments. Elongation of this column pushes the bacterium forward in the same way as growth of an MSP fiber moves its associated vesicle (Fig. 3). Like the MSP in vitro system, movement of Listeria organisms is thought to be a simplified version of leading edge dynamics in crawling cells, and identification of properties shared by these two systems reveals important clues about the mechanism of lamellipodial protrusion. For example, both use the same general mechanism to build their motile apparatus. In Listeria species, a membrane protein, ActA, recruits soluble proteins to the bacterial surface to initiate localized filament assembly (reviewed in Carlier et al., 2002), whereas in the MSP system, an integral membrane phosphoprotein interacts with at least one cytosolic protein other than MSP to trigger filament assembly at the vesicle surface (LeClaire et al., 2003). In both systems, filaments appear to be assembled de novo by

AMOEBOID CELL MOTILITY WITHOUT MOTORS

391

a nucleation‐elongation reaction rather than by the addition of subunits to the ends of existing filaments. The Arp2/3 protein complex, which is a nucleator of actin polymerization (Mullins et al., 1998), is a key component for Listeria movement (Welch et al., 1997) that is required for reconstitution of motility from purified components (Loisel et al., 1999). Assays of the effects of hydrostatic pressure on MSP fiber growth have shown that increased pressure reduces both the number of filaments assembled at the vesicle surface and their rate of polymerization (Roberts et al., 1998). Thus, MSP filament assembly also involves a site‐directed nucleation–elongation reaction. Moreover, in both systems, the newly formed filaments are rapidly enmeshed with other filaments in the fiber and remain essentially stationary as assembly proceeds and the vesicle or bacterium moves away. Even in these simple, reconstituted systems, the precise mechanism of propulsion remains controversial. Listeria motility can be reconstituted in vitro without myosins (Loisel et al., 1999), so neither system seems to require motor proteins. Instead, movement appears to be associated with polymerization and bundling of filaments. Mogilner and Oster (2003a,b) have proposed an elastic Brownian ratchet mechanism to account for this movement, whereby the thermal writhing of a filament allows it to move away from an object sufficiently to add a subunit, then the elastic restoring force of the lengthened filament pushes the object forward. Other models have proposed a role for mechanical stresses that develop during polymerization (Gerbal et al., 2000a,b). Further comparison of MSP‐ and actin‐ based motility promises to facilitate testing of the models and lead to a fuller understanding of the mechanism of protrusion.

V. RETRACTION IS ALSO REQUIRED FOR CRAWLING Although the generation of a protrusive force at the leading edge is necessary for cell crawling, it is not sufficient. A second force, independent of that involved in protrusion, is required to pull the cell body forward as the cell advances. Analyses of cell crawling on flexible substrates have shown that traction forces are produced well behind the leading edge (Harris et al., 1980; Lee et al., 1994; Pelham and Wang, 1997). Moreover, when actin polymerization in the lamellipodium of fish epithelial keratocytes is blocked by treatment with cytochalasin, the trailing cell body continues to retract (Anderson et al., 1996). Evidence for a specific retraction force in Ascaris sperm was obtained by exploiting the sensitivity of the MSP cytoskeleton to changes in intracellular pH to uncouple protrusion of the leading edge from retraction of the cell body (Italiano et al., 1999). Lowering intracellular pH in sperm below

392

STEWART AND ROBERTS

FIG. 4. (LH column) Treatment of sperm with acetate buffer, pH 6.75, uncouples protrusion from retraction. The positions of the cell body and the leading edge when the cell was perfused with acetate buffer are outlined in white. During this sequence (elapsed time, 30 sec) part of the leading edge protruded slowly (3 mm/min) while an adjacent portion, toward the top of the frame, retracted slightly. Thus, treatment with acetate buffer slowed cytoskeletal assembly dramatically. However, cytoskeletal disassembly was not inhibited, so the lamellipodium and the fiber complexes within shortened and the cell body continued to move forward at 15 mm/min. The black arrow indicates a kink in a fiber complex that moved rearward with respect to the leading edge during the sequence. This indicates that cytoskeletal treadmilling persists even when the rate of protrusion slowed. Note that the distance between the cell body and this kink decreases throughout the sequence due to continued disassembly at the base of the

AMOEBOID CELL MOTILITY WITHOUT MOTORS

393

6.0 causes a complete but fully reversible disassembly of the MSP cytoskeleton. By fine‐tuning this pH effect, cytoskeletal assembly can be uncoupled from disassembly and so the role of each in sperm motility can be studied independently (Fig. 4). For example, at pH 6.35, filament assembly along the leading edge stops and the tips of the fiber complexes detach from the lamellipodial membrane. Localized disassembly at the base of the lamellipodium continues and the fiber complexes are pulled toward the cell body as they shorten. At slightly higher pH (6.75), assembly at the leading edge again stops, but the fiber complexes remain attached to the lamellipodial membrane. In this case, disassembly at the base of the lamellipodium continues but, instead of pulling the fiber complexes rearward, the cell body is pulled forward. These observations indicate that, at the base of the lamellipodium, a force is generated that is associated with cytoskeletal disassembly but which is independent of the protrusive force at the leading edge. This second force places the MSP cytoskeleton under tension, as illustrated by the movement of the fiber complexes toward the site of disassembly at pH 6.35. When the fiber complexes maintain their attachment at the leading edge, as in cells at pH 6.75 and in crawling sperm, this tension powers the retraction of the cell body.

VI.

RECONSTITUTION OF RETRACTION IN VITRO

Miao et al. (2003) have been able to reconstitute the essential features of retraction in vitro by adding Yersinia phosphatase (YOP) to the MSP‐based motility apparatus assembled from cell‐free extracts from Ascaris sperm lamellipodium. The white arrows indicate refringent spots that maintained their position in the cell body during retraction. This shows that the cell body moves forward without rolling. Interval between frames, 10 sec. Bar, 10 mm. (RH column) Treatment of sperm with acetate buffer at pH 6.35 causes the cytoskeleton to release from the membrane at the leading edge and recede as disassembly continues at the base of the lamellipodium. Numerals indicate elapsed time in seconds after perfusion with acetate buffer (a). The black arrowhead indicates the forward margin of the cytoskeleton in each frame. Disassembly occurred primarily at the base of the lamellipodium so that the tips of the fiber complexes (white arrowhead), initially in the surface protrusions along the leading edge, remained clearly visible as the cytoskeleton receded and the gap between the cytoskeleton and the leading edge widened (b–e). The black arrow indicates a branch in a fiber complex that moved progressively rearward toward the site of cytoskeletal disassembly (a–d) until it reached the base of the lamellipodium and disappeared as this part of the cytoskeleton depolymerized. The way in which the cytoskeleton moved, with the fiber complexes maintaining their shape as they shortened, showed that the rearward movement was due to local depolymerization of MSP near the cell body rather than general depolymerization along the length of the cytoskeleton. Bar, 10 mm. Reproduced from The Journal of Cell Biology, 1999, vol. 146, pp. 1087–1095 by copyright permission of the Rockefeller University Press.

394

STEWART AND ROBERTS

that also reconstitutes protrusion. Thus, when a preparation containing actively extending fibers was perfused with cell‐free extract (S-100) containing YOP, the fibers ceased to grow and instead began to shorten. Retractions to less than 10% of the initial fiber length were often observed. The optical density of retracting fibers decreased progressively (Fig. 5),

FIG. 5. Retraction requires Yersinia phosphatase (YOP) plus an additional component in S‐100. (A) Time‐lapse sequence of phase contrast micrographs of a fiber assembled in S‐100 and then perfused with S‐100 containing YOP. The vesicle‐bearing end of the fiber is at the top. With time, the fiber shortened and its optical density decreased. Numbers in each frame are elapsed time in minutes. (B) Fiber perfused with YOP in KPM assembly buffer. Without S‐100 present, the fiber lost optical density but exhibited very little shortening. (C) A similar sequence showing a fiber perfused with S‐100 containing 1 mM ATP, but without added YOP. The fiber continued to grow at its vesicle bearing end and did not retract. Bars, 5 mm. Reproduced from Science, 2003, vol. 302, pp. 1405–1407.

AMOEBOID CELL MOTILITY WITHOUT MOTORS

395

indicating that the retraction was accompanied by depolymerization of MSP filaments. Moreover, contracting vesicles were able to move beads and vesicles that were attached to them, confirming that the shortening of the fiber was able to do work and demonstrating that the analogous shortening of fiber complexes that takes place at the base of the lamellipod in crawling sperm could generate the force for cell body retraction. Depolymerization of filaments itself was insufficient to power fiber retraction. Perfusion of actively growing fibers with buffer alone stopped fiber growth and resulted in a slow decrease in optical density, but these fibers did not exhibit the same dramatic shortening observed after YOP was added. Thus, loss of filament mass cannot account for fiber shortening. Although the precise mechanism by which YOP generates fiber retraction remains to be established, it most likely operates in the dephosphorylation of a key component of the system that then allows the MSP filaments to rearrange within the fiber and thereby generate tension.

VII.

A PUSH‐PULL MODEL FOR NEMATODE SPERM AMOEBOID MOTILITY

Ascaris sperm motility suggests a simple ‘‘push‐pull’’ mechanism for locomotion (Fig. 6) in which two separate and distinct forces are required for movement: a protrusive force along the leading edge that pushes against the membrane, and a traction force at the base of the lamellipodium that pulls the cell body forward. This model suggests that substrate attachments, which provide the traction needed to convert forces generated within the cytoskeleton into movement, also have another role in facilitating the mechanical separation of the forces for protrusion and retraction. The organization of the motility apparatus in Ascaris sperm, where the forces are generated at the opposite ends of the fiber complexes, illustrates the need for such separation. The protrusive force at the leading edge would place a fiber complex under compression while the force generated at the rear places that same fiber complex under tension. Ordinarily, these two forces would tend to cancel each other. However, between the regions of polymerization and depolymerization there is a region where the membrane (and the cytoskeleton) is attached to the substrate. Without this attachment, directional movement would not be possible. Bottino et al. (2002) generated a physical model of nematode sperm motility based on these concepts and experimentally determined properties of the system. By simulating the addition of new MSP filaments at the leading edge of the cell coupled to MSP filament depolymerization near

396

STEWART AND ROBERTS

FIG. 6. Proposed push‐pull model for nematode sperm locomotion. Assembly and bundling of MSP filaments into fiber complexes (dark band spanning the lamellipodium) pushes the membrane at the leading edge forward. At the same time a second force, associated with disassembly of the fiber complexes at the base of the lamellipodium, pulls the cell body forward. In this model, attachments where the cytoskeleton is linked to the membrane and the membrane anchored to the substratum establish traction and separate mechanically the forces produced at opposite ends of the fiber complexes. Thus, rather than canceling each other, these forces can be exerted independently against the substratum. Reproduced from The Journal of Cell Biology, 2000, vol. 149, pp. 7–12 by copyright permission of the Rockefeller University Press.

the cell body, it was possible to generate cell movement in silico that closely paralleled that seen in vivo. Thus, the model simulated not only the protrusion of the lamellipod, but also its characteristic shape, and generated an overall cell velocity of the order of the 30 mm/min observed in vivo. The ability to generate a physical model that so closely replicated the movement of nematode sperm observed in vivo further supports the hypothesis that amoeboid cell motility can be powered by cytoskeleton dynamics alone without having to invoke a role for molecular motor proteins analogous to myosin or kinesin. The principles of the push‐pull model probably apply generally to amoeboid cell motility. Indeed, a consensus is developing that in both sperm and actin‐based crawling cells the force for protrusion is derived from localized cytoskeletal assembly (reviewed by Pollard and Borisy, 2003). However, as applied to nematode sperm locomotion, the model envisions that lamellipod extension and cell body retraction are linked reciprocally to the polymerization state of the cytoskeleton. The lack of structural polarity of MSP filaments, the precise localization of cytoskeletal polymerization and depolymerization at opposite ends of the fiber complexes, and insights gained from reconstitution of cytoskeletal dynamics and motility in vitro and in vivo all support the conclusion that nematode sperm move without using motor proteins and that, instead, they rely on

AMOEBOID CELL MOTILITY WITHOUT MOTORS

397

spatially controlled assembly and rearrangement of filaments. Naturally, we cannot rule out that myosins play a role in actin‐based cell crawling and indeed may be required for cell body retraction in some cell types. However, our results suggest that it is plausible that a push‐pull polymerization/ depolymerization–based mechanism may contribute to the locomotion of at least some actin‐based crawling cells.

VIII. FUTURE DIRECTIONS Comparison of the MSP‐based and actin‐based locomotory machinery has yielded a number of insights into the basic mechanism of cell crawling and, for example, has emphasized the importance of vectorial assembly and filament bundling in protrusion. For the MSP system, now that many of the molecules that orchestrate the assembly and disassembly of the motility machinery in cells have been identified, a key goal is to reconstitute motility from purified components. This will enable us to capitalize on the simplicity of nematode sperm and the ability to reconstitute both protrusion and retraction in the same in vitro motility system to address several key questions about this fascinating motility system: how the leading edge membrane directs MSP polymerization, how ATP is used to produce movement, how disassembly generates tension, and how cytoskeletal assembly and disassembly are coupled. The answers to these questions should help pave the way to understanding how reciprocal cytoskeletal events at opposite ends of the lamellipodium are coordinated to drive the migration of the cell and to illuminate the shared physical properties of MSP and actin cytoskeletons that are responsible for crawling movement.

References Anderson, K. I., Wang, Y.‐L., and Small, J. V. (1996). Coordination of protrusion and translocation of the keratocyte involves rolling of the cell body. J. Cell Biol. 134, 1209–1218. Baker, A. M. E., Robers, T. M., and Stewart, M. (2002). 2.5 A˚ crystal structure of helices of the motile major sperm protein (MSP ) of C. elegans. J. Mol. Biol. 319, 491–499. Borisy, G. G., and Svitkina, T. M. (2000). Actin machinery: Pushing the envelope. Curr. Opin. Cell Biol. 12, 104 –112. Bottino, D., Mogilner, A., Roberts, T. M., Stewart, M., and Oster, G. (2002). How nematode sperm crawl. J. Cell Sci. 115, 367–384. Bullock, T. L., McCoy, A. J., Kent, H. M., Roberts, T. M., and Stewart, M. (1998). Structural basis for amoeboid motility in nematode sperm. Nature Struct. Biol. 5, 184–189. Bullock, T. L., Roberts, T. M., and Stewart, M. (1996). 2.5 A˚ resolution crystal structure of the motile major sperm protein (MSP) of Ascaris suum. J. Mol. Biol. 263, 284–296.

398

STEWART AND ROBERTS

Buttery, S. M., Ekman, G. C., Seavy, M., Italiano, J. E., Stewart, M., and Roberts, T. M. (2003). Dissection of the Ascaris sperm motility machinery identifies novel proteins involved in MSP‐based amoeboid locomotion. Mol. Biol. Cell 14, 5082–5088. Carlier, M.‐F., Le Clainche, C., Wiesner, S., and Pantaloni, D. (2002). Actin‐based motility: From Molecules to movement. Bioessays 25, 336–345. Forscher, P., and Smith, S. J. (1988). Actions of cytochalasins on the organization of actin filaments and microtubules in a neuronal growth cone. J. Cell Biol. 107, 1505–1516. Gerbal, F., Chaikin, P., Rabin, Y., and Prost, J. (2000a). An elastic analysis of Listeria monocytogenes propulsion. Biophys. J. 79, 2259–2275. Gerbal, F., Laurent, V., Ott, A., Carlier, M. F., Chaikin, P., and Prost, J. (2000b). Measurement of the elasticity of the actin tail of. Listeria monocytogenes. Eur. Biophys. J. 29, 134–140. Haaf, A., Butler, P. J. G., Kent, H. M., Fearnley, I. M., Roberts, T. M., Neuhaus, D., and Stewart, M. (1996). The motile major sperm protein (MSP) from Ascaris suum is a symmetric dimer in solution. J. Mol. Biol. 260, 251–260. Harris, A. K., Wild, P., and Stopak, D. (1980). Silicone rubber substrata: A new wrinkle in the study of cell locomotion. Science 208, 177–179. Italiano, J. E., Jr., Roberts, T. M., Stewart, M., and Fontana, C. A. (1996). Reconstitution in vitro of the motile apparatus from the amoeboid sperm of Ascaris shows that filament assembly and bundling move membranes. Cell 84, 105–114. Italiano, J. E., Jr., Stewart, M., and Roberts, T. M. (1999). Localized depolymerization of the major sperm protein cytoskeleton correlates with the forward movement of the cell body in the amoeboid movement of nematode sperm. J. Cell Biol. 146, 1087–1095. King, K. L., Stewart, M., Roberts, T. M., and Seavy, M. (1992). Structure and macromolecular assembly of two isoforms of the major sperm protein (MSP) from the amoeboid sperm of the nematode, Ascaris suum.. J. Cell Sci. 101, 847–857. Le Claire, L. L., III, Roberts, T. M., and Stewart, M. (2003). A 48 kDa integral membrane phosphoprotein orchestrates the cytoskeletal dynamics that generate amoeboid cell motility in Ascaris sperm. J. Cell Sci. 116, 2655–2663. Lee, J., Leonard, M., Oliver, T., Ishihara, A., and Jacobson, K. (1994). Traction forces generated by locomoting keratocytes. J. Cell Biol. 127, 1957–1964. Loisel, T. P., Boujemaa, R., Pantaloni, D., and Carlier, M. F. (1999). Reconstitution of actin‐based motility of Listeria and Shigella using pure proteins. Nature 401, 13–16. Matsuura, Y., Stewart, M., Kawamoto, M., Kamiya, N., Saeki, K., Yasunaga, Y., and Wakabayashi, T. (2000). Structural basis for the higher Ca2þ activation of the regulated actin‐activated myosin ATPase observed in Dictyostelium/Tetrahymena chimeras. J. Mol. Biol. 296, 579–595. Miao, L., Vanderlinde, O., Stewart, M., and Roberts, T. M. (2003). Retraction in amoeboid cell motility powered by cytoskeletal dynamics. Science 302, 1405–1407. Mogilner, A., and Oster, G. (2003a). Polymer motors: Pushing out the front and pulling up the back. Curr. Biol. 13, R721–R733. Mogilner, A., and Oster, G. (2003b). Force generation by actin polymerization. II The elastic ratchet and tethered filaments. Biophys. J. 84, 1591–1605. Mogilner, A., and Oster, G. (2003c). Shrinking gels pull cells. Science 302, 1340–1341. Mullins, R. D., Heuser, J. A., and Pollard, T. D. (1998). The interaction of Arp2/3 complex with actin: Nucleation, high affinity pointed end capping, and formation of branching networks of filaments. Proc. Nat. Acad. Sci. USA 95, 6181–6186. Pantaloni, D., Le Clainche, C., and Carlier, M.‐F. (2001). Mechanism of actin‐based motility. Science 292, 1502–1506.

AMOEBOID CELL MOTILITY WITHOUT MOTORS

399

Pelham, R. J., and Wang, Y.‐L. (1997). Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc. Nat. Acad. Sci. USA 94, 13661–13665. Pollard, T. D., and Borisy, G. G. (2003). Cellular motility driven by assembly and disassembly of actin filaments. Cell 112, 453–465. Roberts, T. M., Salmon, E. D., and Stewart, M. (1998). Hydrostatic pressure shows that lamellipodial motility in Ascaris sperm requires membrane‐associated major sperm protein filament nucleation and elongation. J. Cell Biol. 140, 367–375. Roberts, T. M., and Stewart, M. (2000). Acting like actin: The dynamics of the nematode major sperm protein (MSP) cytoskeleton indicate a push‐pull mechanism for amoeboid cell motility. J. Cell Biol. 149, 7–12. Sepsenwol, S., Ris, H., and Roberts, T. M. (1989). A unique cytoskeleton associated with crawling in the amoeboid sperm of the nematode. Ascaris suum. J. Cell Biol. 108, 55–66. Stewart, M., King, K. L., and Roberts, T. M. (1994). The motile major sperm protein (MSP) from Ascaris suum forms filaments constructed from two helical subfilaments. J. Mol. Biol. 243, 60–71. Theriot, J. A. (1996). Worm sperm and advances in cell locomotion. Cell 84, 1–4. Theriot, J. A., and Mitchison, T. J. (1991). Actin microfilament dynamics in locomoting cells. Nature 352, 126–131. Wang, Y.‐L. (1985). Exchange of actin subunits at the leading edge of motile fibroblasts: Possible role of treadmilling. J. Cell Biol. 101, 597–602. Welch, M. D., Iwamatsu, A., and Mitchison, T. J. (1997). Actin polymerization is induced by Arp2/3 complex at the surface of Listeria monocytogenes. Nature 385, 265–269.

STRUCTURE AND MECHANISM OF DNA POLYMERASES By PAUL J. ROTHWELL AND GABRIEL WAKSMAN Institute of Structural Molecular Biology, University College London and Birkbeck College, Malet Street, London WC1E 7HX, United Kingdom

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Biological Diversity of DNA Polymerases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Family A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Family B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Family C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Family D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Family X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Family Y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G. RT Family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. The Nucleotide Incorporation Pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. General Theme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Variation on a Theme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. The E State: Basic Architecture of DNA Polymerases . . . . . . . . . . . . . . . . . . . . V. Primer/Template DNA Binding and Recognition . . . . . . . . . . . . . . . . . . . . . . . VI. Formation of the E:p/t:dNTP Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Conformational Transition to a Catalytically Active Ternary Complex: The E’:p/t:dNTP Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Structures of E’:p/t:ddNTP Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Molecular Basis for the Rate‐Limiting Step. . . . . . . . . . . . . . . . . . . . . . . . . . C. Discrimination Between Correct Versus Incorrect Nucleotide . . . . . . . . VIII. Phosphoryl Transfer Reaction, Product Release, and Translocation of the Primer/Template DNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Phosphoryl Transfer Reaction and Product Release . . . . . . . . . . . . . . . . . B. Translocation of Primer/Template DNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Force Generation: DNA Polymerases as Molecular Motors . . . . . . . . . . . IX. Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

402 403 403 403 404 404 404 405 406 406 406 408 409 414 417 419 420 423 426 428 428 429 429 430 431

Abstract DNA polymerases are molecular motors directing the synthesis of DNA from nucleotides. All polymerases have a common architectural framework consisting of three canonical subdomains termed the fingers, palm, and thumb subdomains. Kinetically, they cycle through various states corresponding to conformational transitions, which may or may not generate force. In this review, we present and discuss the kinetic, structural, and single‐molecule works that have contributed to our understanding of DNA polymerase function. ADVANCES IN PROTEIN CHEMISTRY, Vol. 71 DOI: 10.1016/S0065-3233(04)71011-6

401

Copyright 2005, Elsevier Inc. All rights reserved. 0065-3233/05 $35.00

402

ROTHWELL AND WAKSMAN

I. Introduction The viability of an organism is dependent on the accurate replication of its genome. In general, this is performed with only one error generated for every 109–1010 bases replicated (Echols and Goodman, 1991). This remarkable accuracy is achieved by a combination of different mechanisms working in unison. The initial discrimination is carried out during the nucleotide incorporation stage in which a DNA polymerase accurately selects a nucleotide (2’‐deoxyribonucleoside-5’‐triphosphate (dNTP)) to be added to a primer strand of a duplex DNA, based on its complementarity to a template base provided by a template strand of DNA. In the event of the wrong nucleotide being inserted, this can be removed by the actions of 3’-5’ exonucleases. In the case of an error being ‘‘sealed’’ into the genome by incorporation of dNTPs past the site of an error, replicative excision/repair pathways exist to remove the offending base. The highest contribution to accuracy is provided by the DNA polymerase activity. Various models have been proposed by which DNA polymerases select the correct nucleotide from a pool of structurally similar nucleotides. The initial idea, proposed by Watson and Crick (1953a,b) based on the structure of DNA, was that selection could be determined by the A‐T and G‐C hydrogen bond–mediated base pairing. However, replicative polymerases have error rates for nucleotide insertion in the range of 10–3–10–6/base replicated and the difference in energy between correct versus incorrect base pairs would, at best, only account for an error frequency of 2
10–2 (Loeb and Kunkel, 1982). The polymerase, therefore, must be performing a more active role, rather than providing a platform for zipping DNA. A second model proposed that DNA polymerases select correct over incorrect nucleotides due to base pair geometry (Bruskov and Poltev, 1979; Engel and von Hippel, 1978; Sloane et al., 1988). In this case, the shape of the DNA polymerase active site is such that a correct Watson‐Crick (W‐C) base pair could be accommodated, whereas a non‐W‐C base pair would be rejected. It has also been proposed that free energy differences between correct versus incorrect base pairs could be amplified due to exclusion of water molecules in the polymerase active site (Petruska et al., 1986). This chapter will review the current understanding of the nucleotide incorporation cycle by DNA polymerases and the mechanisms employed by DNA polymerases to replicate DNA accurately. It includes a review of more recent and stimulating work that explores the mechanochemistry of DNA polymerases and their role as force generators and molecular motors. Although additional activities are present on many polymerases (e.g., 5’-3’ exonuclease [family A], a 3’-5’ exonuclease [family A and B],

STRUCTURE AND MECHANISM OF DNA POLYMERASES

403

lyase [family X], or RNAseH (RT family) activities), we will discuss here only advances in our knowledge of the polymerase activity.

II. Biological Diversity of DNA Polymerases Many different polymerases have been discovered and, based on the primary sequence homologies (Braithwaite and Ito, 1993; Delarue et al., 1990; Ito and Braithwaite, 1991) and crystal structure analyses ( Joyce and Steitz, 1994), the different polymerases have been classified into seven different families: A, B, C, D, X, Y, and RT.

A.

Family A

The family A polymerases can be grouped into replicative and repair enzymes. Enzymes from the bacteriophages T3, T5, and T7 and the eukaryotic mitochondrial DNA polymerase g are replicative polymerases and interact with other proteins for accurate DNA replication. T7, for example, interacts with bacterial thioredoxin, which acts as a processitivity factor increasing the number of nucleotides added to the DNA chain from 1–15 to several thousand before dissociation of the polymerase (Tabor et al., 1987). Additionally, the T7 polymerase interacts with the T7 DNA primase‐helicase and a single‐stranded DNA‐binding protein (Kornberg and Baker, 1992). Together, this T7 replisome coordinates the synthesis of leading and lagging strand synthesis (Kornberg and Baker, 1992). The family A repair enzymes include Escherichia coli polymerase I (pol I), Thermus aquaticus (Taq) pol I, and Bacillus stearothermophilus pol I. They are involved in nucleotide excision repair and in processing Okazaki fragments that are generated during lagging strand synthesis (Kornberg, 1980). Most pol I enzymes contain a 5’-3’ exonuclease activity and a 3’-5’ proofreading activity. Only the 5’-3’ exonuclease is required for viability because it is necessary for the removal of RNA primers from Okazaki fragments generated during replicative DNA synthesis. The DNA polymerase activity is used to fill in the resulting gap. During repair, pol I enzymes also fill in DNA gaps that result from the removal of a variety of DNA lesions (Kornberg and Baker, 1992).

B.

Family B

Eukaryotic replicative DNA polymerase a, d, E, archaebacterial DNA polymerases, viral DNA polymerases, DNA polymerases encoded by mitochondrial plasmids of various fungi and plants, and some bacteriophage

404

ROTHWELL AND WAKSMAN

polymerases (T4 and RB69) all belong to family B. Family B polymerases are predominantly involved in DNA replication and carry out processive replication of chromosomal DNA during cell division in eukaryotes. The enzyme participates in leading and lagging strand synthesis and is tethered to the DNA by a processity factor. It is also stimulated by a single stranded DNA‐binding protein (Kornberg, 1980). The family B polymerases contain a 3’-5’ exonuclease activity that corrects errors during DNA replication. The 3’-5’ exonuclease activity of family B polymerases is very strong, being over a 1000 times higher than that of E. coli pol I (Capson et al., 1992; Lin et al., 1994).

C.

Family C

Bacterial family C polymerases are the major chromosomal replicative enzyme (Kornberg and Baker, 1992). Like other replicative polymerases, the holoenzyme interacts with other proteins and forms a large multisubunit complex consisting of at least 10 subunits (Kornberg and Baker, 1992). The a‐subunit contains the DNA polymerase activity that is tightly associated with the E‐subunit, which contains a 3’-5’ exonuclease activity (Kelman and O’Donnell, 1995).

D.

Family D

Family D polymerases are found in the Euryarchaeota subdomain of Archaea. Although characterization of this family is at an early stage, it is known that the enzyme is heterodimeric (Uemori et al., 1997). The smallest subunit shows low but significant homology to the eukaryotic DNA polymerase d (Cann et al., 1998), whereas the large subunit is thought to harbor the catalytic region. Characterization of Pyrococcus furiosus DNA polymerase II of family D has shown that the enzyme contains both a DNA polymerase activity and a 3’-5’ exonuclease activity, and it has been suggested to be a replicative polymerase (Uemori et al., 1997).

E. Family X Known members of the family X polymerases include eukaryotic DNA polymerase b (pol b) (Abbotts et al., 1988), polymerase s (Burgers et al., 2001), polymerase m (Dominguez et al., 2000), polymerase l (Garcia‐Diaz et al., 2000), yeast polymerase IV (Prasad et al., 1993), and the African swine fever virus polymerase X (Martins et al., 1994). Pol b is known to be involved in the base excision repair (BER) pathway, which is important for repairing

STRUCTURE AND MECHANISM OF DNA POLYMERASES

405

abasic sites (Matsumoto and Kim, 1995). The enzyme has a modular organization with an 8‐kDa amino‐terminal domain connected to the carboxy‐terminal domain (31 kDa) by a protease‐hypersensitive hinge region. The 8‐kDa domain contains a 5’‐deoxyribose phosphate (dRP) lyase activity that is needed for the process of BER, whereas the large domain contains the DNA polymerase. During BER, the phosphodiester backbone of an abasic site is cleaved 5’ to the sugar moiety by an abasic site endonuclease (Matsumoto and Bogenhagen, 1991). DNA pol b cleaves the remaining 5’‐deoxyribose phosphate using its dRP lyase activity and fills in the resulting gap using its polymerase activity. The resulting nick is sealed by the action of a DNA ligase (Dianov et al., 2003). The dRP lyase domain also seems to play a role in directing the polymerase to a 5’‐phosphate and has single‐stranded DNA binding affinity (Kumar et al., 1990). Other members of this family contain additional structural elements, which may be important for function. For example, polymerase l contains a nuclear localization signal and a breast cancer susceptibility gene (BRCT) domain. The BRCT domain is thought to mediate protein–protein and protein–DNA interactions upon DNA damage (Bork et al., 1997).

F.

Family Y

Ultraviolet (UV) irradiation and other mutagenic agents often cause damage to cellular DNA. This can result in physical damage (e.g., base loss creating an abasic site) or modification (e.g., UV crosslinks). Due to the high selectivity and fidelity of replicative DNA polymerases, such damage would stall the replication complex. Recently, a new family of DNA polymerases, family Y, has been identified, the members of which are able to recognize and bypass different classes of lesions on DNA (Friedberg and Gerlach, 1999; Friedberg et al., 2000; Goodman and Tippin, 2000; Johnson et al., 1999; Woodgate, 1999). Family Y polymerases are found in eubacteria, eukaryotes, and archae. The polymerase requires a relatively low specificity to deal with DNA damage. This is reflected in the low fidelity (in the range of 10–2–10–4 errors/base replicated) of the polymerase on undamaged DNA and in the lack of an intrinsic 3’-5’ proofreading ability (Zhou et al., 2001). Due to their low selectivity, Y polymerases must also function in a distributive manner; they do not remain bound to the DNA during multiple cycles of nucleotide addition, so as not to cause mutagenic incorporation events after lesion bypass is completed. The distributive mode of synthesis should also allow their displacement by the more processive replicative polymerases, thereby reducing their potential mutagenic activity. The family Y polymerases include the DinB (damage‐induced) and

406

ROTHWELL AND WAKSMAN

UmuC polymerases, which are also known as DNA polymerase IV and V in E. coli, respectively, and from eukaryotes, the Rev1 (pol z) and Rad30 (pol ) polymerases (Yang, 2003; Zhou et al., 2001).

G. RT Family The reverse transcriptase (RT) family includes RTs from retroviruses as well as the eukaryotic telomerases. During the course of reverse transcription, retroviral RTs interact with a variety of different nucleic acid substrates (RNA/RNA, DNA/RNA, RNA/DNA, DNA/DNA) to convert a single‐stranded viral RNA genome into double‐stranded proviral DNA (Gotte et al., 1999). Some of the retroviral RTs function as dimers, such as those from the human immunodeficiency virus (HIV) 1 and 2, whereas others, like Moloney murine leukemia virus (MULV) RT, are monomeric. However, both types contain a polymerase domain as well as an RNAseH domain to cleave viral RNA during DNA synthesis. Although they will not be discussed in the context of this review, telomerases also belong to the RT family of polymerases: they use an integral RNA component as a template for synthesis of dGT‐rich strands of telomeres (Greider and Blackburn, 1985, 1989). This review focuses on the polymerases of the A, B, X, Y, and RT families due to the fact that they have been extensively characterized both kinetically and structurally.

III.

The Nucleotide Incorporation Pathway A. General Theme

Various studies of DNA polymerases have established a minimal model of nucleotide incorporation for family A (Dahlberg and Benkovic, 1991; Donlin et al., 1991; Eger and Benkovic, 1992; Kuchta et al., 1987, 1988; Patel et al., 1991; Wong et al., 1991), family B (Capson et al., 1992; Lin et al., 1994), family X (Ahn et al., 1997; Kraynov et al., 1997; Werneburg et al., 1996; Zhong et al., 1997), family Y (Fiala and Suo, 2004b; Washington et al., 2001), and RT (Hsieh et al., 1993; Kati et al., 1992; Rittinger et al., 1995; Wo¨ hrl et al., 1999) polymerases. This model is largely common to all polymerases and will be discussed first. Variations on this general theme have been observed and will be discussed next and also in sections below. The basic model for the nucleotide incorporation by all DNA polymerases is shown in Fig. 1. DNA polymerases catalyze the synthesis of

STRUCTURE AND MECHANISM OF DNA POLYMERASES

407

Fig. 1. Kinetic pathway of nucleotide incorporation. The various complexes are indicated as mentioned in the text. ‘‘kpol,’’ the rate constant of the rate‐limiting step, is indicated.

DNA via an ordered mechanism in which the primer template DNA (p/t) binds prior to the dNTP. Polymerization begins with the binding of a p/t to the unliganded enzyme (E) to form the enzyme‐p/t complex (E:p/t) (step 1). Nucleotide incorporation into the enzyme‐p/t complex is initiated by the binding of a dNTP to the E:p/t complex to form the enzyme‐p/t‐dNTP complex (E:p/t:dNTP) (step 2). The rate‐limiting step of polymerization is the conversion of the E:p/t:dNTP complex to an activated complex, E’:p/t: dNTP, which is competent to undergo chemistry (step 3). This rate‐limiting step is thought to be caused by a conformational change. The cycle is completed by the nucleophilic attack by the 3’‐OH primer terminus on the a‐phosphate of the dNTP that results in the formation of a phosphodiester bond (step 4). This is followed by a second conformational change, which allows the release of the pyrophosphate (PPi) product (step 5). The enzyme can then either dissociate from the p/t (distributive synthesis) or translocate the substrate to form a new 3’ terminus for a new round of incorporation (processive). If an incorrect nucleotide is incorporated, the new terminus can be partitioned to the 3’-5’ exonuclease domain, if present in the polymerase architecture, to be removed. Alternatively, the misincorporated nucleoside can be removed directly by pyrophosphorolysis (the reverse reaction of DNA synthesis), or the polymerase can extend past the incorrect nucleotide, ‘‘sealing’’ the misincorporated nucleotide within the elongated strand. Because step 3 will be discussed in great detail in the sections below, it is appropriate to provide a brief summary of how it was discovered, using the T7 DNA polymerase as an example (Donlin et al., 1991; Patel et al., 1991; Wong et al., 1991). The rate‐limiting step was observed when the dependence of the burst amplitude (corresponding to the incorporation of the first nucleotide) on nucleotide concentration was examined. This experiment provided kpol of step 3 (Fig. 1), which was lower than all the other

408

ROTHWELL AND WAKSMAN

rates for the preceding chemistry steps. The next question was whether the observed rate was a direct measure of the rate of phosphodiester bond formation. It has been shown that a rate‐limiting step involving the making or breaking of a phosphate bond shows a phosphothiate elemental effect. Thus, kpol was measured using both dTTP and dTTP(aS) (the phosphothioate analogue of dTTP) and both values of kpol were compared. These experiments demonstrated a weak elemental effect, indicating that the chemical step is not rate limiting. Thus, the rate‐limiting step was ascribed to a conformational change required to form the catalytically competent E’:p/t:dNTP complex from the E:p/t:dNTP complex. Direct observation of the E’:p/t:dNTP complex was obtained using pulse‐chase experiments. In such experiments, incorporation of labeled nucleotide to an E:p/t complex is either quenched by the addition of HCl or allowed to proceed after the addition of a large excess of ‘‘cold’’ unlabeled dNTP (the ‘‘chase’’ step) followed by acid quench. In the HCl quench experiments, the acid quenches all the enzyme‐bound species. On the other hand, when the reaction is chased with cold dNTP, each of the enzyme‐bound species is allowed to partition both in the forward and reverse directions. The amount of partitioning in the forward direction is observed as an excess of labeled product, compared with the acid quench experiment, while the dNTP that partitions in the reverse direction is diluted and remains unobservable. As an excess was observed and because the binding of dNTP to the E:p/t complex is rapid, the observed flux or excess is mainly due to the E’:p/t:dNTP complex.

B. Variation on a Theme Although most polymerases conform to the general kinetic scheme, some polymerases have different mechanisms with regard to p/t binding and selection. Other aspects in the polymerase cycle such as dNTP binding, chemistry, and the conformational change will be discussed in later sections. Spectroscopic techniques have been applied to monitor nucleic acid binding by both pol b and HIV-1 RT, and the data indicate a more complex binding process than indicated by the general model. For pol b, it has been shown that binding of gapped DNA is a three‐step process ( Jezewska et al., 2002). The initial event is thought to involve binding of the nucleic acid by the 8‐kDa lyase domain, followed by two docking steps that position the substrate for catalysis ( Jezewska et al., 2002). Quantitative studies on pol b have shown that the single‐stranded DNA can be bound in two modes that differ in the length of DNA buried by the protein ( Jezewska et al., 2001a,b; Rajendran et al., 1998, 2001). This is

STRUCTURE AND MECHANISM OF DNA POLYMERASES

409

thought to be due to the fact that both the lyase domain and the polymerase domain can bind single‐stranded DNA. In one binding mode, both the lyase and polymerase domains interact with the DNA. In the second binding mode, interactions with the DNA are mediated exclusively by the lyase domain. Although the lyase domain should not be required for the polymerization mechanism, its removal reduces the efficiency in which polymerization occurs, which may be due to its role in substrate binding (Kumar et al., 1990). For HIV-1 RT, binding of nucleic acid is a two‐step process that is thought to involve the formation of an initial collision complex followed by a conformational change in either the nucleic acid or the protein that ‘‘locks’’ the substrate in place (Rittinger et al., 1995; Wo¨hrl et al., 1999). It has also been shown for HIV-1 RT that binding of the nucleic acid substrate leads to the formation of three types of nucleic acid/protein complexes (Wo¨ hrl et al., 1999). The DNA in one complex is bound in a productive mode and is able to incorporate nucleotides. For the second complex, the enzyme or DNA must undergo a conformational change for nucleotide incorporation to occur. For the third complex, the primer template must first dissociate and reassociate before nucleotide incorporation can occur (Wo¨ hrl et al., 1999). Single‐molecule solution‐based studies using Fluorescence Resonance Energy Transfer (FRET) confirmed the existence of all three complexes and provided structural information on each (Rothwell et al., 2003). In the first complex (productive complex in product state [PP]), the DNA is indeed bound to the protein in such a way that would allow nucleotide incorporation in the mode observed crystallographically ( Jacobo‐Molina et al., 1993). In the second complex, the primer terminus now occupies the dNTP‐binding pocket (productive complex in educt state [PE]). This would correspond to a state in which dNTP incorporation has occurred, but before translocation of the p/t substrate. The third complex appears to be bound by the epitope recognized by the Fab fragment (PJR unpublished) used to crystallize the binary RT:p/t complex ( Jacobo‐Molina et al., 1993). The role played by the two latter complexes is not fully understood.

IV.

The E State: Basic Architecture of DNA Polymerases

The first DNA polymerase activity was identified in 1956 in E. coli (Kornberg et al., 1956; Lehman et al., 1958). The enzyme was subsequently named DNA polymerase I (pol I). E. coli pol I is a 109‐kDa enzyme that supports a multidomain architecture containing a polymerase activity, a 5’-3’ exonuclease activity, and a 3’-5’ exonuclease activity. The C‐terminal portion of E. coli pol I, called the Klenow fragment, which lacks the 5’-3’

410

ROTHWELL AND WAKSMAN

function (Klenow and Overgaard‐Hansen, 1970), was the first DNA polymerase structure to be solved crystallographically (Ollis et al., 1985). The structure of the Klenow fragment has subsequently been solved with DNA bound in the exonuclease site (Beese et al., 1993a; Freemont et al., 1988), and with dNTP and pyrophosphate (Beese et al., 1993b). From the initial structural studies of the Klenow fragment, the overall shape of the polymerase domain was likened to a right hand, with subdomains referred to as fingers, palm, and thumb (Fig. 2A), an organization that has been seen for nearly all classes of polymerases solved so far. The polymerase active site, which contains catalytically essential amino acids, is located within the palm subdomain that forms the base of the crevice formed by the fingers and thumb subdomains. The fingers subdomain is important for nucleotide recognition/binding. The thumb subdomain is important for binding the DNA substrate. Subsequent structures of family A polymerases (T7 DNA polymerase, Taq DNA polymerase, the Klenow fragment of the Taq polymerase [Klentaq1]) and B. stearothermophilus DNA polymerase I fragment (BF) show that this topology is conserved (Doublie et al., 1998; Kiefer et al., 1998; Kim et al., 1995; Korolev et al., 1995). For both Taq polymerase and the Bacillus fragment, the 3’-5’ exonuclease is inactive due to both lacking key catalytic residues in this region. Several structures have been solved for family B enzymes, including those from the bacteriophage RB69 (Franklin et al., 2001; Hogg et al., 2004; Wang et al., 1997b) and from several archaeal organisms (Hashimoto et al., 2001; Hopfner et al., 1999; Rodriguez et al., 2000; Zhao et al., 1999). Like family A polymerases, the general architecture of the polymerase domain is conserved despite very little sequence homology (Fig. 2B). RB69 DNA polymerase is a circular 103‐kDa polypeptide with a central cavity. Five separate subdomains surround the cavity. One half of the enzyme constitutes the polymerase domain, containing the fingers, palm, and thumb subdomains named in accordance with the Klenow fragment structure. As with all polymerases, the catalytically important residues are located in the palm subdomain, which can be superimposed with polymerases from different classes. The circle is completed by an N‐terminal domain and a 3’-5’ exonuclease domain that is homologous to that seen in the Klenow fragment. The N‐terminal domain shares homology with RNA‐ binding domains. Both T4 and RB69 polymerases are able to bind specifically to the ribosome binding sites of their own mRNA to repress their translation; thus, the N‐terminal domain is thought to be involved in this process (Pavlov and Karam, 1994; Tuerk et al., 1990; Wang et al., 1997a). The best characterized family X member is DNA pol b, the smallest eukaryotic polymerase. Although family X polymerases seem to have evolved separately from other classes of DNA polymerases, they share

STRUCTURE AND MECHANISM OF DNA POLYMERASES

411

Fig. 2. Structures of family A, B, X, Y, and RT polymerases. The proteins are in ribbon representation. The fingers, palm, and thumb subdomains are color‐coded in gold, red, and green, respectively. (A) Structure of apo Klentaq1 (family A). The 3’-5’ vestigial exonuclease domain is indicated in silver. (B) Structure of apo RB69 DNA polymerase (family B). The 3’-5’ exonuclease domain and the N‐terminal domain are indicated in grey and silver, respectively. (C) Structure of apo pol b DNA polymerase (family X). The lyase domain is indicated grey. (D) Structure of the Dpo4 DNA polymerase (family Y ). The little finger subdomain is indicated in silver. (E) Structure of the p66 subunit of reverse transcriptase (RT family). The RNAseH and connection subdomains are indicated in grey and silver, respectively.

412

ROTHWELL AND WAKSMAN

many common structural features (Fig. 2C). Mammalian pol b is composed of two distinct domains. The C‐terminal domain has the overall shape of a polymerase domain and has structures analogous to the fingers, palm, and thumb subdomains, although with little homology (Sawaya et al., 1994). The subdomains were named according to their spatial relationship to conserved features in the palm (Steitz, 1994). The 8‐kDa N‐terminal domain is tethered to the thumb by a flexible loop. Pol b was the first structure to be solved as an active ternary complex (Pelletier et al., 1994), and the structure has also been solved in the presence of DNA (Pelletier et al., 1994, 1996; Sawaya et al., 1997). The structure of a ternary complex in which the terminal base pair of the nucleic acid substrate contains the DNA lesion 8‐oxodeoxyguanine and an incoming nucleotide (Krahn et al., 2003) has also been determined. Although only identified recently as DNA polymerases, many crystal structures of Y family DNA polymerases have been solved (Ling et al., 2001, 2003, 2004; Nair et al., 2004; Silvian et al., 2001; Trincao et al., 2001; Uljon et al., 2004; Zhou et al., 2001). The structures of the unliganded DinB polymerase from Sulfolobus solfataricus (Silvian et al., 2001), of the liganded Dpo4 polymerase from the same organism (Ling et al., 2001), of the unliganded pol  polymerase from Saccharomyces cerevisiae (Trincao et al., 2001), and of the liganded human polymerase i (Nair et al., 2004) reveal the overall architecture of the polymerase domain of family Y polymerases (Fig. 2D). Family Y polymerases share the polymerase domain architecture of family A polymerases. In addition an extra‐subdomain is observed, at the C‐terminus that is referred to as little finger, polymerase associated domain (PAD), or wrist. This subdomain is tethered to the thumb subdomain but is physically located next to the fingers subdomain. The binding groove is thus made up of fingers, little finger, palm, and thumb (Ling et al., 2001). The finger and thumb subdomains are both unusually small and lead to a more open and solvent‐accessible active site than observed in family A or B polymerases. In contrast to other polymerase families, the structures of some family Y polymerases show that the fingers subdomain appears to be in a closed conformation in the absence of substrate. Also, the enzyme appears to lack any equivalent of the O‐helix, which, in family A polymerases, is located in the fingers subdomain and is an important structural feature for nucleotide binding. Instead, it has a b strand and an adjacent extended loop, which is more similar to the arrangement seen in HIV-1 RT (see below). The structures of MULV (Das and Georgiadis, 2004), HIV-1 (Esnouf et al., 1995; Hsiou et al., 1996; Rodgers et al., 1995), and HIV-2 (Ren et al., 2002) RTs reveal that RT family polymerases share a common architecture of the polymerase domain with family A enzymes. In addition to the

STRUCTURE AND MECHANISM OF DNA POLYMERASES

413

polymerase domain, RTs contain an RNAseH domain, which is used to cleave viral RNA during DNA synthesis. The enzyme also has a connection subdomain, which connects the RNAseH domain to the polymerase domain. Both HIV enzymes function as heterodimers (Divita et al., 1995a,b) composed of two related chains. For example, HIV-1 RT consists of a 66‐kDa subunit (p66) and a 51‐kDa subunit (p51) derived from p66 by proteolytic cleavage. MULV RT is monomeric (Das and Georgiadis, 2004; Roth et al., 1985). HIV-1 RT is the best structurally characterized of this family (Fig. 2E). High‐resolution structures are available of HIV-1 RT in unliganded form (Esnouf et al., 1995; Hsiou et al., 1996; Rodgers et al., 1995), bound to non‐ nucleoside RT inhibitors (NNRTIs) (Ding et al., 1995a,b; Kohlstaedt et al., 1992; Pata et al., 2004; Ren et al., 1995, 1998, 2004; Smerdon et al., 1994), in complex with nucleic acid substrates (Ding et al., 1998; Huang et al., 1998; Jacobo‐Molina et al., 1993; Jaeger et al., 1998; Sarafianos et al., 2001, 2002), or in a ternary complex with p/t DNA and nucleotide (Huang et al., 1998). Both subunits of this heterodimeric enzyme, p66 and p51, contain four subdomains (fingers, palm, thumb, connection) (Kohlstaedt et al., 1992). In addition, the p66 subunit contains the RNAseH domain. Although the structure of the subdomains within p66 and p51 are similar, the relative arrangement of the subdomains within the subunits is different. The p51 subunit, despite containing the components of a polymerase active site, has no catalytic activity and is thought to stabilize the dimer. Perhaps the most obvious difference in the HIV-1 RT, compared with family A, B, and X polymerases, is in the fingers subdomain. Family A, family B, and pol b, all have an a‐helical region that forms part of the dNTP‐binding site and is important for nucleotide selection. For HIV-1 RT, a region of antiparallel b‐ribbon (b3–b4) fulfills this role. The apo form of HIV-1 RT was solved crystallographically in both ‘‘open’’ and ‘‘closed’’ forms, corresponding to a different position of the thumb subdomain, either closer to the opposite fingers (closed) or farther away (open). Subsequently, EPR measurements showed the enzyme to be predominantly in a closed conformation (Kensch et al., 2000). Molecular dynamics simulations also suggest that the closed conformation is also favored for the unliganded enzyme (Madrid et al., 1999). This closed form should not be confused with the closed structure of the ternary complex bound to DNA and nucleotide (see Section VII). However, the thumb subdomain can also adopt an even more open form: NNRTI binding induces an additional hinge movement of the p66 thumb subdomain near the thumb’s knuckle, causing the p66 thumb subdomain to adopt a configuration that is even more extended than in the open apo and DNA‐bound RT structures (see below) (Hsiou et al., 1996; Smerdon et al., 1994).

414

ROTHWELL AND WAKSMAN

V. Primer/Template DNA Binding and Recognition The initial event in the nucleotide incorporation cycle is binding of the double‐stranded p/t DNA to the polymerase to form the E:p/t complex (Step 1; Fig. 1). Many structures now exist, in which DNA is bound at the polymerase active site. The structure of the E:p/t complex for the Klenow fragment of the Taq polymerase, Klentaq1, shows the overall binding mode of DNA for family A polymerases (Li et al., 1998b; Fig. 3A). When compared with the apo form of the enzyme, this structure shows that DNA‐binding results in structural changes mostly localized in the thumb subdomain (Fig. 3B). The movement corresponds to an initial opening of the thumb followed by a rotation in the opposite direction that brings residues in the tip of the thumb closer to the DNA (Li et al., 1998b). The movements in the thumb are mostly localized to a helix‐loop‐helix motif at the tip of the thumb, which, for structures not containing bound DNA, is disordered, suggesting a high degree of flexibility. The net result of the thumb movement is to form a cylinder that completely surrounds the DNA. The wrapping of the domain around the DNA may hold the DNA during processive polymerization. Residues within the palm subdomain interact with the template strand along the minor groove. The 3’ terminus of the primer is held near the polymerase active site in the palm subdomain. Neither the duplex DNA nor the single‐stranded template passes through the crevice between the fingers and the thumb subdomains; rather, the single‐stranded part of the DNA template is flipped out of stacking arrangement with the duplex by a sharp angle in the template sugar‐phosphate backbone, which positions the single‐stranded template on the same side of the crevice as the duplex DNA (Fig. 3C). A Tyr (671 for Klentaq, 714 for BF, 766 for Klenow) at the C‐terminus of the O‐helix is inserted into the stacking arrangement of the template bases, lying directly on top of the first base pair of the duplex part of the DNA (Fig. 3C). This residue may act as a positioning device for the p/t such that the first base pair of the duplex can register itself against the active site. Interestingly, in the complex of Klentaq1 bound to a nucleotide (see Section VI), although the conformation of the fingers subdomain is the same (i.e., open), Tyr671 is not seen, suggesting a high degree of flexibility in the absence of DNA (Li et al., 1998a). The bound DNA is mostly in the B‐form, except for the last three bases closest to the polymerase active site, which are in the A‐form. The DNA is distorted to assume an S‐shape: the first bend is caused by interactions within the palm subdomain and the second bend is caused by interactions with the thumb. A similar arrangement is seen for other family A polymerases with bound DNA (Doublie et al., 1998; Eom et al., 1996; Kiefer et al., 1998).

STRUCTURE AND MECHANISM OF DNA POLYMERASES

415

Fig. 3. Structure of the p/t DNA bound Klentaq1. (A) Overall structure of the Klentaq1 bound to a p/t DNA. The protein is in ribbon representation, whereas the p/t DNA is in both ribbon and ball‐and‐stick representation. The fingers, palm, and thumb subdomains are color‐coded as in Fig. 2, whereas the vestigial 3’-5’ exonuclease is colored grey. The primer strand ribbon is in cyan, and the template strand ribbon is in magenta. (B) Superimposition of the apo (red) and p/t‐bound (cyan) Klentaq1. Only the protein is shown and is in ribbon representation. The orientation is the same as in Fig. 3A. The DNA‐bound structure differs from that of the apo form by a conformational change in the thumb subdomain. (C) Location of Tyr 671 and the templating base in the E:p/t complex. The O‐helix is labeled and is shown in red ribbon representation. The primer and template strands are in both ribbon and ball‐and‐stick representation. The primer and template strands are in silver and cyan, respectively. The first single‐ stranded template base (the templating base) is in deep blue. Tyr 671 is seen stacked on top of the template base of the first duplex base pair, and the templating base is flipped out. (D) Model of a dNTP bound to the O‐helix of Klentaq1 in the E:dNTP complex and its position relative to active site. The subdomains are color‐coded as in Fig. 3A. The various subdomains are in ribbon representation with lined instead of filled‐in ribbons for the fingers subdomain. This allows the visualization the O‐helix in pale green in the fingers domain. The distance between the base of the dNTP and the active site residues is indicated.

416

ROTHWELL AND WAKSMAN

Recently, a binary complex of RB69 DNA polymerase (family B) was solved bound to a furan containing p/t DNA (Hogg et al., 2004). Although the bound DNA represents a state in which the DNA has not translocated after incorporation, it does give an overall idea of how the DNA is bound by family B polymerases. Like the apo form of the enzyme, the fingers subdomain is in an upright conformation (Wang et al., 1997b). In contrast to family A polymerases, the DNA is bound in the B‐form throughout. However, like family A polymerase, the single‐stranded region of the template is flipped out of the stacking arrangement with the duplex part of the DNA. Binding of a p/t DNA with a four‐base template overhang to pol b (family X) results in an inward movement of the 8‐kDa N‐terminal domain to form a more compact structure. Although the 8‐kDa domain does not interact with the DNA, it approaches close to where a longer template overhang would presumably locate (Pelletier et al., 1994). The pol b:DNA complex structure has also been determined with a gapped DNA substrate (Sawaya et al., 1997). In this form, the DNA is held and the gap opened by the concerted action of the 8‐kDa lyase domain, which grips the DNA downstream of the gap and interacts with the 5’ phosphate and the polymerase domain that holds the substrate upstream of the gap. This introduces a 90‐degree kink into the DNA, which has B‐form characteristics on both sides of the kink (Pelletier et al., 1996; Sawaya et al., 1997). Compared with the recessed p/t substrate‐bound model of pol b, the 8‐kDa domain is in an even more compact closed form and interacts with the fingers subdomain. From the structure of the ternary E’:p/t:ddNTP complex for the Dpo4 polymerase (family Y; see also Section VII), it can be seen that the DNA bound by the polymerase domain of family Y polymerases is in the B‐form throughout (Ling et al., 2001). The DNA bound to family Y polymerases is more solvent exposed than DNA bound to family A or B polymerases. The Dpo4 polymerase makes relatively few contacts with the DNA outside the active site, and the bound DNA buries less than 600 A˚ of the molecular surface of the catalytic core (fingers, palm, and thumb) compared with 1000 A˚ in family A and B polymerases (Ling et al., 2001). However, the little finger seems to increase the overall interaction of the DNA with the protein. It contains a four‐stranded b‐sheet and two parallel a‐helices. The b‐sheet interacts with the DNA to increase the contact surface area. The thumb and the little finger grip the nucleic acid across the minor groove. Family Y polymerases lack an equivalent to the Tyr residue of family A polymerases (Tyr 671 of Klentaq1 for example), which is seen to insert into the stacking arrangement of the template bases, lying directly on top of the duplex part of the DNA (Ling et al., 2001). Interestingly, mutations of this tyrosine residue to smaller side

STRUCTURE AND MECHANISM OF DNA POLYMERASES

417

chains in family A polymerases result in a ‘‘family Y–like’’ mutator phenotype (Carroll et al., 1991). For the DNA‐bound forms of HIV-1 RT (RT family), it can be seen that the thumb is rotated outward when compared with the unliganded structure (Ding et al., 1998; Jacobo‐Molina et al., 1993). In the RT:p/t complex, the p66 thumb subdomain adopts an upright position, as opposed to the ‘‘closed’’ unliganded structure (Hsiou et al., 1996; Rodgers et al., 1995). This movement results in an opening of the structure, generating a large DNA‐binding cleft that extends over 60 A˚ from the N‐terminus of the polymerase domain to the C‐terminal RNAseH domain. The position of the other subdomains remains largely unchanged. For HIV-1 RT, the first five base pairs of the p/t substrate are in A‐form, followed by a four–base pair region that introduces a sharp kink of about 40–45 degrees that then leads to a B‐form DNA. This large region of A‐form DNA may be important in binding RNA/RNA and RNA/DNA duplexes, which are expected to adopt an A‐form (Kornberg and Baker, 1992). The 5’‐template (resolved for three bases in the ternary E’:p/t:ddNTP complex structure by Huang et al. (1998; see Section VII) bends away from the duplex by a sharp angle and packs against residues on the surface of the fingers subdomain.

VI.

Formation of the E:p/t:dNTP Complex

Nucleotide incorporation into the enzyme‐p/t complex is initiated by the binding of a dNTP to the E:p/t complex to form the E:p/t:dNTP complex (Step 2; Fig. 1). Step 2 in the nucleotide incorporation cycle is an important step at which the enzyme is able to discriminate between correct versus incorrect nucleotides. However, the efficiency with which different polymerases bind correct or incorrect dNTPs at this step varies greatly. The replicative polymerases discriminate with high efficiency against incorrect nucleotides during Step 2. The KD differences between correct and incorrect (KD(dNTPcorrect)/KD(dNTPincorrect)) range from 390‐fold for T7 DNA polymerase (family A) to 263‐fold for T4 DNA polymerase (family B) and 250‐ fold for HIV-1 RT (Gillin and Nossal, 1976; Kati et al., 1992; Topal et al., 1980; Wong et al., 1991). For the repair enzymes, the differences between correct and incorrect nucleotide binding are much smaller. The Klenow fragment (family A) only selects against incorrect nucleotide incorporation on average by a factor of 3.4, yeast polymerase  (family Y ) by a factor of 4, and polymerase b (family X) by a factor of 20 (Ahn et al., 1997; Kuchta et al., 1987, 1988; Washington et al., 2001; Werneburg et al., 1996). This suggests that initial dNTP recognition at Step 2 is defined by function (replicative vs. repair) rather than family.

418

ROTHWELL AND WAKSMAN

Although the E:p/t:dNTP complex has not been captured structurally with a native p/t DNA‐bound substrate, both the Klenow fragment and Klentaq1 have been crystallized in the presence of all four nucleotides (Beese et al., 1993b; Li et al., 1998a). In this form, all four nucleotides are bound by the N‐terminal end of the O‐helix in the fingers subdomain. The triphosphate group of the dNTP is within the range of electrostatic interactions with the positively charged side chains of Arg659 (754 in E. coli pol I), Lys 633 (758 in E. coli pol I), and Arg587 (682 in E. coli pol I), with the triphosphate moiety running parallel to the O‐helix (Fig. 3D). The base of each dNTP points toward the DNA‐binding cleft, although the electron density indicates in each case that there is disorder. The position of the triphosphate moiety of the nucleotide is nearly identical for all bound nucleotides, suggesting that the initial recognition of the incoming nucleotide is through the triphosphate. Characterization of mutations of O‐helix residues Arg754 and Lys758 in E. coli pol I shows that they both greatly affect the KM(dNTP) and KM(PPi) and have little or no effect on DNA binding (Astatke et al., 1995; Suzuki et al., 1996). Lys758 of E. coli pol I was also identified as the site of labeling with pyridoxal 5’‐phosphate (PLP), a compound that binds specifically to triphosphate‐binding sites (Basu and Modak, 1987). When bound to the O‐helix, the nucleotide is far away (10–15 A˚ ) from the active site located in the palm (Fig. 3D) and, thus, a large conformational change affecting the fingers subdomain must be invoked to ‘‘deliver’’ the nucleotide to the active site of the enzyme (see next section). Recently, an E:p/t:dNTP complex was captured for the T7 (family A) DNA polymerase, a high‐fidelity DNA polymerase, in the presence of a p/t DNA containing a cis‐syn thymine dimer (Li et al., 2004). Although this type of lesion may impede the closing of the fingers subdomain in the presence of dNTP, it does add evidence that binding within the O‐helix is the initial event of the nucleotide incorporation cycle for family A polymerases. However, this triphosphate‐mediated binding would not provide the high‐level discrimination between correct and incorrect nucleotides displayed by T7 DNA polymerase. It is therefore unlikely that this complex represents the E:p/t:dNTP complex observed kinetically. However, it cannot be ruled out that, in the presence of a ‘‘conventional’’ p/t DNA substrate, additional constraints are imposed on dNTP binding. Recent kinetic work on RB69 polymerase (family B) and structural comparison between the RB69 polymerase and other polymerase families led to the postulation of a different initial binding event for dNTP (Yang et al., 2002a). The crystal structure of the closed ternary E’:p/t:ddNTP complex of the RB69 polymerase (see Section VII) shows interactions between the fingers subdomain and the ddNTP’s triphosphate moiety

STRUCTURE AND MECHANISM OF DNA POLYMERASES

419

similar to those observed in family A polymerases (Doublie et al., 1998; Franklin et al., 2001; Li et al., 1998b). However, it was suggested that, due to the greater discrimination at Step 2 between correct and incorrect nucleotides displayed by family B members compared with family A repair enzymes, the nucleotide would diffuse directly into the polymerase active site (Yang et al., 2002a). This would allow the RB69 polymerase to probe correct versus incorrect base pairing directly. The difference in affinity between correct and incorrect nucleotides would result in only correct nucleotides remaining bound long enough for the formation of the closed ternary E’:p/t:dNTP complex, a conformation that seems to be required for catalysis (see next section). Mutational work on human DNA polymerase a also supports a different initial recognition of nucleotide by family B members. Mutations within the conserved motif II, which contains one of the catalytically important aspartates, reduce the affinity of correct nucleotides (Dong et al., 1993), suggesting that residues in motif II form both the dNTP‐binding region and the catalytic site. Based on these results, one could argue that the dNTP‐binding site is located in close proximity to the active site, possibly coinciding. Thus, DNA‐binding would position the templating base (the first base of the template’s single‐stranded overhang) within the dNTP‐ binding site, providing a powerful readout of correct versus incorrect nucleotide binding through W‐C pairing. Consistent with this hypothesis, while wild‐type human DNA polymerase a (like other family B polymerases) strongly discriminates at the dNTP‐ground state binding stage with KMdTTP(correct)/KMdATP(incorrect) being over 2000, mutations of Tyr865 within motif II (a mutation shown to reduce dNTP‐binding but not DNA‐ binding) to Ser reduces discrimination by a 20‐fold, whereas mutation to Phe has little effect on activity and discrimination (Dong et al., 1993).

VII.

Conformational Transition to a Catalytically Active Ternary Complex: The E’:p/t:dNTP Complex

The rate‐limiting step in the kinetic pathway of nucleotide incorporation is the conversion of the E:p/t:dNTP complex to the activated complex, E’:p/t:dNTP (Step 3 in Fig. 1). This step is crucial in many respects. First, it is essential for the phosphoryl transfer reaction to occur. During the E:p/t:dNTP to E’:p/t:dNTP transition, all the components of the active site are assembled and organized in a topological and geometrical arrangement that allows the enzyme to proceed with the chemical step (Step 4). Second, Step 3 plays a major role in the mechanism of discrimination between correct versus incorrect nucleotides. Interpretation of the kinetic measurements has led to the hypothesis that the E:p/t:dNTP

420

ROTHWELL AND WAKSMAN

to E’:p/t:dNTP rate‐limiting step is caused by a slow conformational change in the binding partners. In this section, we will first describe the various attempts that structural biologists have made to elucidate the molecular basis of this kinetically defined conformational change. We will next describe the mutational, kinetic, and computational work that has challenged the conclusions of the structural biology work. Finally, in a third section, we will discuss the mechanism of nucleotide discrimination at Step 3.

A. Structures of E’:p/t:ddNTP Complexes The E’:p/t:dNTP complex is a transient complex and thus cannot be captured crystallographically as is. However, by using dideoxyribonucleoside triphosphates (ddNTP) in a reaction that involves a E:p/t binary complex where the primer is terminated by a ddNMP, Kraut and colleagues have shown that a ternary complex of pol b can be obtained where the catalytic residues, the Mg ions, the p/t DNA, and the ddNTP appear to be ‘‘trapped’’ in a chemistry‐competent state (Pelletier et al., 1994). Thus, in the presence of a terminated primer, the incoming ddNTP appears to be able to trigger the conformational changes required to form a catalytically competent complex. However, due to the fact that ddNTP lack the 3’ OH group, chemistry cannot occur. The difference in the structures on going from the binary E:p/t complex to the ternary E’:p/t:ddNTP complex shows large conformational changes that are described below. Formation of the ddNTP‐trapped ternary complex (E’:p/t:ddNTP) of Klentaq1 (family A) results in a large reorientation of the fingers subdomain (Fig. 4; Li et al., 1998b). The effect of the conformational transition affecting the fingers subdomain is to position the O‐helix in two different conformations (Fig. 4B). In the first conformation (termed open), the O‐helix is in the configuration of the apo and DNA‐bound enzyme. Tyr671 is inserted into the stacking arrangement of the template base and lies on top of the first base pair of the duplex DNA (Fig. 3C). In the second orientation (termed closed), seen in the ddNTP‐trapped ternary complex, the O‐helix has moved inward by 46 degrees and is now much closer to the active site formed by the three carboxylates located in the palm domain. During this transition, the side chain of Tyr671 is released from the stacking arrangement with the template base (Fig. 4C). This allows the first single‐stranded DNA base of the template (the templating base) to position itself in front of the incoming nucleotide (Fig. 4C). The ddNTP is bound to the O‐helix and is stacked onto the 3’ base of the primer strand. Two metal ions (Mg2þ) are bound to the catalytic aspartate residues (Asp 610 and Asp 785) and to the nucleotide (see mechanism in

STRUCTURE AND MECHANISM OF DNA POLYMERASES

421

Fig. 4. The open‐to‐closed conformational transition affecting the fingers subdomain. (A) Structure of the Klentaq1 E’:p/t:ddNTP complex. See legend of Fig. 3A for key to color coding. The ddNTP is indicated in black but is hardly visible in this orientation. (B) Superimposition of the E:p/t (cyan) and E’:p/t:ddNTP (gold) complexes. Only the protein is shown in ribbon representation. The orientation is the same as in (A). The O‐helix is indicated in both conformations. The ternary complex differs from the binary complex by a large conformational change affecting the O‐helix. (C) Location of Tyr 671 and the templating base in the E’:p/t:ddNTP complex. See legend of Fig. 3A for key to color coding. The incoming ddNTP nucleotide is colored black and is located in front of the templating base in deep blue. The Mg2þ ions are indicated in gold balls.

Section VIII). In this closed form of the enzyme, the complex appears to be poised for chemistry. One of the striking features of the ternary complex described above is the close fit of the protein around the nascent base pair in the closed ternary complex (Li and Waksman, 2001a; Li et al., 1998b).

422

ROTHWELL AND WAKSMAN

Such a snug fit of the protein and DNA around the nascent base pair provides a most efficient steric readout for appropriate base pairing and is thought to play a fundamental role in the mechanism of discrimination between correct versus incorrect nucleotide (see below). In the active ternary complex of the RB69 polymerase, the fingers subdomain is in a closed conformation as seen in family A polymerases (Franklin et al., 2001). The rotation of the fingers subdomain inward brings conserved residues of the finger, which contact the triphosphate moiety, close to the active site. One interesting consequence of the fingers subdomain movement in RB69 polymerase is that the finger domain now makes direct contact with the exonuclease site (Franklin et al., 2001). This may be important in transmitting structural changes caused by misincorporation to the exonuclease site. Several structures of ternary E’:p/t:ddNTP complexes of a family Y polymerase, Dpo4, have also been determined (Ling et al., 2001, 2004). Two forms of the ternary E’:p/t:ddNTP complex have been solved, one in which the correct nucleotide is present for incorporation and the other in which an incorrect nucleotide is provided (Ling et al., 2001). For the correctly base paired E’:p/t:ddNTP complex, the newly formed base pair is less protected than in the replicative polymerases due to a more solvent exposed active site. In the second form, the first templating base is a G followed by a C, and Ling et al. supplied ddGTP. A G:G mismatch has been shown to be disfavored by this enzyme, with a mismatch frequency of 3.5
10–4 (Boudsocq et al., 2001). Interestingly, the enzyme skips the G base by translocating the p/t substrate so that the ddGTP can base pair with the C. This mechanism of incorporation may be important in translocating through certain lesions, such as cys‐syn cyclobutane pyrimidine dimers (CPD). From the structures of the ternary complexes, a mechanism by which the enzyme can bypass cys‐syn CPD was suggested (Ling et al., 2001) and was later refined by the determination of two structures of Dpo4 in a complex with a p/t DNA carrying a CPD and ddATP (Ling et al., 2004). In these latter structures, the 3’ thymine of the CPD forms a W‐C base pair with the incoming ddATP, but the 5’ thymine forms a Hoogsteen base pair with the ddATP in syn conformation. Dpo4 retains a similar tertiary structure, but each unusual DNA structure is individually fitted into the active site for catalysis. A rate‐limiting step corresponding to a conformational change has been observed in family Y polymerases (Fiala and Suo, 2004a,b; Washington et al., 2001). However, it has been difficult to ascertain the nature of the conformational change for family Y members due to the fact that no complete set of structures (e.g., E, E:p/t, E’:p/t:ddNTP) is available for a single polymerase from this family. However, it has been suggested, based

STRUCTURE AND MECHANISM OF DNA POLYMERASES

423

on superimposition of the structure of yeast pol  apo form (Trincao et al., 2001) and those of the Dpo4 ternary complexes (Ling et al., 2001, 2004), that a 48‐degree inward rotation of the fingers and little finger subdomains occurs (Fiala and Suo, 2004a). However, in the case of Dbh, the apo form of the enzyme (Silvian et al., 2001) appears to be in a closed conformation, and superimposition of the structure with the ternary complexes of Dpo4 (Ling et al., 2001) shows no obvious conformational change (Fiala and Suo, 2004a). Therefore, the nature of the conformational change detected by pre–steady state kinetic analysis (Fiala and Suo, 2004a,b; Washington et al., 2001) remains unclear for Y family DNA polymerases. For HIV-1 RT, binding of the incoming nucleotide substrate, in the presence of bound p/t, also results in a structural rearrangement in the fingers subdomain (Huang et al., 1998). This structural rearrangement results in a 20‐degree inward bending of the outer part of the fingers subdomain towards the palm subdomain. It is not yet clear whether the dNTP is ‘‘delivered’’ to the active site by initial binding of the triphosphate moiety to the fingers subdomain, as would seem to be the case for family A polymerases. Nevertheless, like Klentaq1, the closure of the fingers subdomain results in the formation of a pocket that accommodates the nascent base pair. Residues in the fingers subdomain coordinate the triphosphate moiety of the incoming nucleotide. The 3’‐OH of the incoming nucleotide projects into a pocket referred to as the 3’‐OH pocket. This pocket is important and has been exploited in the design of nucleoside RT inhibitors (NRTIs) used in retroviral therapy (for review, see Sluis‐Cremer et al., 2000). The pol b:p/t:ddCTP structure shows that, in the absence of a downstream DNA fragment, the lyase domain does not interact with the DNA and is positioned some distance from the active site (Pelletier et al., 1994). However, in a structure solved with gapped DNA, the lyase domain binds to the 5’‐phosphate in the DNA gap and interacts with its own carboxy‐ terminus in the thumb (Sawaya et al., 1997). In common with all polymerases, the fingers subdomain closes down around the correct nucleotide and its complementary template base. This motion corresponds to a 30‐ degree rotation of a‐helix N (the dNTP‐binding site) around a‐helix M and thus brings helix N and its bound nucleotide into position to probe correct W‐C base pairing.

B.

Molecular Basis for the Rate‐Limiting Step

With the crystallographic data of different states in the nucleotide incorporation cycle now available, it is tempting to believe that the conformational change identified by kinetic experiments at step 3 could be

424

ROTHWELL AND WAKSMAN

due to the closing of the fingers subdomain observed structurally on binding of the ddNTP to the E:p/t complex. However, recent mutational, kinetic, and computational studies have challenged these conclusions. Mutational work on pol b led to the suggestion that the open‐to‐closed transition affecting the fingers subdomain (and in particular, helix N) observed crystallographically is not the rate‐limiting conformational change (Pelletier et al., 1994). Asp 276 in a‐helix N makes contact with the incoming nucleotide only in the closed conformation of pol b. Remarkably, a mutation of this residue to Val results in an increase in free energy in dNTP ground‐state binding (Vande Berg et al., 2001). This result strongly suggests that the rate‐limiting conformational change is not the open‐to‐ closed structural transition, but instead is triggered in the closed polymerase conformation. Vande Berg et al. suggested a three‐step binding model in which the dNTP initially binds in a nonspecific manner through either the triphosphate or the sugar moiety. This step is associated with weak binding so as to facilitate rapid sampling of the nucleotide pools and is not kinetically resolvable. They further suggested that this rapid‐sampling step is associated with the open‐to‐closed transition seen crystallographically when going from the binary E:p/t complex to the ternary E’:p/t: ddNTP complex. During the second step, bases are selected due to base pairing and steric complementarity at the polymerase active site. In the presence of a correct nucleotide, steric complementarity in the closed form of the ternary complex triggers the kinetically defined rate‐limiting conformational change, which results in the final alignment of all components participating in the reaction (Vande Berg et al., 2001). Similar conclusions were drawn from inspection of the rate‐limiting kinetics of wild‐type pol b and a pol b variant mutated at Tyr265, a residue that does not make any contact with the DNA or the dNTP (details on the wild‐type enzyme are provided in Section VII.C; Shah et al., 2003; Zhong et al., 1997). This mechanism would also apply to family A polymerases, in which a nucleotide is bound in the O‐helix in the open conformation. A fast conformational change of the O‐helix would allow the bound nucleotide to be delivered to the active site. At this stage, correct base‐pairing would trigger a conformational change, whereas incorrect base‐pairing would result in the nucleotide leaving the active site (Vande Berg et al., 2001). Conformational transitions in the Klenow fragment have been examined by stopped‐flow fluorescence using DNA substrates containing the fluorescent reporter 2‐aminopurine (2‐AP) on the template strand, either at the templating position opposite the incoming nucleotide or 5’ to the templating base (Purohit et al., 2003). It was shown that the 2‐AP reporter at the templating position undergoes a sizable and very rapid decrease in fluorescence associated with dNTP binding. Comparison of

STRUCTURE AND MECHANISM OF DNA POLYMERASES

425

the structures of the E:p/t and E’:p/t:ddNTP complexes in Klentaq1 shows that the templating base must undergo a very large conformational change to position itself in the stacking arrangement with the duplex part of the p/t DNA (Figs. 3C and 4C). However, in the open form, this position is occupied by Tyr 671 and thus Tyr 671 must move. This is achieved by the rotation of the O‐helix during the open‐to‐closed transition, which takes Tyr 671 with it, thereby liberating the space into which the templating base can come (Figs. 3C and 4C). As the motion of the templating base is fast, the open‐to‐closed conformational transition affecting the O‐helix must be at least as fast. These data appear to indicate that the open‐to‐closed transition affecting the fingers subdomain is very fast and not rate limiting. Note that this evidence is all indirect and that the rate of the open‐to‐closed conformational transition has not been measured. However, all data are indicative of a fast motion. This is counterintuitive because the fingers subdomain is relatively big and one may have difficulties envisaging that such a large domain motion could be fast. However, direct evidence for a rapid motion of the fingers subdomain has been provided. To probe the microenvironment and dynamics of a‐helix N in the polymerase domain of pol b, the single native tryptophan (Trp 325) was removed, and a tryptophan was strategically placed near the end of a‐helix N (Kim et al., 2003). Influences of substrates on the fluorescence anisotropy decay of this single tryptophan form of pol b were determined. The results revealed that the segmental motion of a‐helix N was rapid (~1 nsec) and far more rapid than the step that limits chemistry. Binding of a correct dNTP significantly limited the angular range of the nanosecond motion within a‐helix N, whereas binding of a p/t or gapped DNAs had minor effects. These results again argue that the rate‐limiting step is not the conformational transition observed crystallographically (Pelletier et al., 1994). Although the experiments described above are instructive, they still leave the question open as to the nature of the rate‐limiting conformational change. This process was investigated in more detail using computer simulations (Yang et al., 2002b). Again, the suggestion was that the closing of a‐helix N was rapid and that the rate‐limiting conformational change may correspond to more subtle movements of side‐chains within the active site. Further simulations of pol b substrate‐induced dynamics showed a role for Mg2þ in the formation of the E’:p/t:dNTP complex (Yang et al., 2004). The results showed that the closing of the fingers subdomain is favored in the presence of Mg2þ, and removal of Mg2þ favors the reopening of the subdomain. The sequence of events proposed to occur during the nucleotide incorporation cycle is as follows: (1) binding of the dNTP: Mg2þ to the open pol b:p/t complex; (2) binding of the catalytic Mg2þ;

426

ROTHWELL AND WAKSMAN

(3) relatively fast conformational transition of the fingers subdomain from an open to closed state; (4) slow, possibly rate‐limiting assembly of the key amino acids residues, template bases, Mg2þ ions, and the primer 3’‐OH group; (5) slow and possibly rate‐limiting chemical step of the nucleotidyl transfer reaction; (6) release of the catalytic Mg2þ; (7) relatively fast conformational transition from the closed‐to‐open complex state again involving subtle residue motions; (8) release of the product PPi/Mg2þ (Yang et al., 2004). Computer simulations must be met with a degree of skepticism. However, these will no doubt suggest experiments aimed at confirming or perhaps disproving their conclusions.

C.

Discrimination Between Correct Versus Incorrect Nucleotide

How do the processes described above contribute to fidelity? Steps 2 and 3 of the nucleotide incorporation cycle are the most important steps in the mechanism of nucleotide discrimination. Discrimination at Step 2 has been discussed in Section VI. Here we discuss the role of Step 3 in that process. Much of the discrimination against the incorrect nucleotide occurs at Step 3, although here again the degree to which the enzyme is able to discriminate against incorrect nucleotide incorporation varies, as does the extent to which the rate‐limiting step contributes to it. Quench flow studies on both T7 and Klenow fragment (family A) show that both strongly discourage nucleotide misincorporation, as measured by the ratio of kpol(correct)/kpol(incorrect), by slowing Step 3 by a factor of 2000 and 5000, respectively (Kuchta et al., 1987, 1988; Patel et al., 1991; Wong et al., 1991). Due to observed elemental effects, it was thought that for T7 DNA polymerase, the rate‐limiting step was due to a conformational change irrespective of whether correct or incorrect nucleotides are incorporated (Patel et al., 1991). For the Klenow fragment, however, nucleotide incorporation of the correct nucleotide is limited by the conformational change, while, in the case of misincorporation, Step 4 (chemistry) becomes rate‐limiting (Patel et al., 1991). Nucleotide incorporation into a (2‐AP)‐containing p/t DNA substrate by the Klenow fragment indicated that a conformational change occurs during misincorporation, but the chemical step determines the slowest rate (Frey et al., 1995). HIV-1 RT has an extremely low discrimination at Step 3, varying between 7‐ and 90‐fold (Kati et al., 1992). This low discrimination at Step 3 results in HIV-1 RT having one of the lowest fidelities among replicative polymerases. The level of discrimination against incorrect nucleotides by HIV-1 RT is lower even than family Y polymerases, for example, yeast pol , which

STRUCTURE AND MECHANISM OF DNA POLYMERASES

427

discriminates by a factor of 150, and also pol b, which discriminates between 85‐ and 2800‐fold at Step 3 (Ahn et al., 1997; Washington et al., 2001). For both pol b (family X) and yeast pol  (family Y ), elemental effects on nucleotide incorporation would indicate that the conformational change is rate limiting for both correct and incorrect nucleotide incorporation, although yeast pol  has a higher elemental effect during dNTP incorporation opposite certain DNA lesions (Washington et al., 2001; Werneburg et al., 1996), which may indicate that chemistry becomes rate limiting in certain cases. However, it has been suggested for the Dpo4 polymerase, another family Y member, that chemistry is rate limiting for incorrect nucleotide incorporation (Fiala and Suo, 2004a,b). Experiments studying dNTP incorporation using 2‐AP as a signal have expanded the details of nucleotide selection by pol b. Correct nucleotide incorporation by this enzyme shows two changes in fluorescence, a fast change followed by a slow one. The slow change corresponds to the quench‐flow derived kpol (Step 3), the rate‐limiting step. The faster change is thought to correspond to an event occurring after dNTP‐binding but before the conformational change has occurred (likely the crystallographically observed conformational transition affecting the fingers subdomain). The two fluorescent changes were also observed upon incorporation of incorrect nucleotides. The rate of the fastest change was of a similar value to that observed for correct nucleotide incorporation. However, the second phase is slowed dramatically (600‐fold) compared with correct nucleotide incorporation (Zhong et al., 1997). These experiments suggest that correct and incorrect nucleotide binding both induce the fast open‐to‐closed conformational transition affecting the fingers subdomain, and that discrimination occurs at a subsequent rate‐limiting step, the molecular basis of which is still unclear. In summary, all evidence so far points to the fact that discrimination takes place at a step occurring after the closing of the fingers domain. Kinetically, this step is rate limiting. Structurally, it is likely to correspond to the setting of the molecular stage leading to chemistry, which may include side‐chain rearrangements, positioning of the various groups involved in catalysis, or other requirements to reach the transition state. When a correct nucleotide is provided, the setting of the chemical stage proceeds harmoniously, albeit at a rate‐limiting pace. When an incorrect nucleotide is provided, the proper geometry of all active site components required for chemistry cannot be attained with the same degree of perfection, and thus the chemical step may become rate limiting. In any case, the setting of the stage is slower because all interactions required for chemistry are either prevented from being made or are made suboptimally.

428

ROTHWELL AND WAKSMAN

VIII. Phosphoryl Transfer Reaction, Product Release, and Translocation of the Primer/Template DNA A. Phosphoryl Transfer Reaction and Product Release The mechanism of nucleotide addition by DNA polymerases was proposed on the basis of the nearly identical mechanism seen for the 3’-5’ exonuclease of DNA polymerase I (Steitz, 1993, 1998). Briefly (because it has been reviewed extensively, e.g., in Steitz, 1998, and Li and Waksman, 2001b), the mechanism is proposed to occur as follows. Two metal ions (Mg2þ) are octahedrally coordinated by the triphosphate of an incoming nucleotide and side‐chains in the active site. One metal (metal B) is ligated in the base of the octahedral plane by four oxygen atoms, contributed by the b‐ and g‐phosphates and two universally conserved carboxylate groups. The coordination sphere of the metal ion is completed on each side of the octahedral plane by interactions with oxygen atoms in the a‐phosphate and the carbonyl oxygen of a Tyr residue. The other metal ion (metal A) is coordinated to the octahedral plane by oxygen atoms from the carboxylate of Asp 882 (Klenow fragment), the a‐phosphate and two water molecules. On one side of the octahedral plane, metal A is ligated by an oxygen atom from the carboxylate of Asp 705 (Klenow fragment) and on the other side by the 3’‐OH group of the ribose moiety on the primer strand. In the proposed mechanism, metal ion A lowers the affinity of the 3’OH for the hydrogen, facilitating the 3’ O– attack on the a‐phosphate. Metal ion B assists the leaving of the PPi, and both metal ions stabilize the structure and charge of the expected pentacovalent transition state. The phosphoryl transfer reaction is followed by a second conformational change, which allows the release of the PPi product (Step 5). Studying the reverse reaction, that is, pyrophosphorolysis for pol b with 2‐AP fluorescence, showed three distinct fluorescence changes. The slowest phase corresponded to the rate of formation of dNTP, the product of pyrophosphorolysis, whereas the other two phases were thought to report on events happening before chemistry (Dunlap and Tsai, 2002; Zhong et al., 1997). It has been shown for the Klenow fragment that the PPi product has only a fivefold lower affinity for the E:p/t complex than dNTPs, suggesting that the product of the reaction could compete for binding of dNTPs (Kuchta et al., 1987). However, for T7 DNA polymerase, the affinity of PPi is extremely low and is nowhere near comparable to affinities for correct nucleotide binding, differing by nearly a factor of 1000 (Patel et al., 1991). T4 has similarly reduced affinities for PPi compared with Klenow, being in the low millimolar range (Capson et al., 1992).

STRUCTURE AND MECHANISM OF DNA POLYMERASES

429

B. Translocation of Primer/Template DNA In family A polymerases, translocation of the p/t DNA must occur coincidentally with the opening of the fingers subdomain. The Klentaq1 system provides an appropriate model to illustrate the motion and its consequences. We have mentioned the role of Tyr 671 which, in the open form of the fingers subdomain, stacks against the first base pair of the duplex part of the p/t DNA (Fig. 3C). In that position, Tyr 671 appears to position the p/t DNA in the appropriate register against the active site of the enzyme. In the closed form, Tyr 671 moves out to make room for the first templating base (Fig. 4C). Once a nucleotide has reacted to the 3’‐OH primer strand, the opening of the fingers subdomain would bring Tyr 671 back and thus would risk a steric clash with the newly formed base pair unless the latter translocates. How this occurs is not known. However, it has been hypothesized that in the open state, the DNA is able to move along the ‘‘electrostatic’’ tunnel or cylinder formed by the DNA‐binding site (mostly formed by the palm subdomain and the wrapping around of the thumb subdomain; see Section V; Guajardo and Sousa, 1997). Indeed, the DNA‐binding site is ‘‘coated’’ by a strongly positive electrostatic field emanating from a large number of positively charged residues. In such an environment, interactions with the DNA may be interchangeable and thus, perhaps paradoxically, the DNA may be relatively free to move. Thus, at any given time, the newly formed base pair is present only a fraction of the time in the active site of the open complex. Thus, Tyr 671 can probe the environment around the newly formed base pair and insert itself on top of it when space allows, that is, when the p/t DNA has slid away from the active site. Packing of Tyr 671 against the newly formed base pair stabilizes the DNA in the proper register, and another cycle of nucleotide incorporation could be initiated.

C.

Force Generation: DNA Polymerases as Molecular Motors

One of the most exciting developments in the field of DNA polymerases has been the use of single‐molecule optical and magnetic traps and atomic force microscopes to study the behavior of DNA polymerases under the constraints of forces applied to the primer/template substrate. The T7 DNA polymerase was studied with an optical trap (Wuite et al., 2000) while the Klenow fragment of E. coli, its 3’-5’ exonuclease deficient mutant, and 3’-5’ exonuclease deficient mutant of T7 polymerase were studied using a magnetic trap (Maier et al., 2000). More recently, the behavior of single molecules of HIV1 RT was studied using atomic force microscopy (Lu et al., 2004). These studies take advantage of progress made in single‐molecule

430

ROTHWELL AND WAKSMAN

manipulation of DNAs and use single‐molecule assays based on the differential elasticity of single‐stranded and double‐stranded DNAs. The three sets of experiments showed similar behavior. These studies showed that the rate of the replication reaction catalyzed by DNA polymerases is altered when a force is applied to the template strand (Lu et al., 2004; Maier et al., 2000; Wuite et al., 2000). It was observed that the replication rate decreases at high forces and appears to increase at low force. The ‘‘stall’’ force, the mechanical force that is necessary to apply to the stretched DNA to block DNA polymerization, varies but is within a range of 15–40 pN, a force that is considered ‘‘large’’ for molecular motors. These studies also point to the fact that the force load acts on the rate‐limiting step identified kinetically, thereby indicating that the limiting (rate‐determining) step involves work by the DNA polymerase and therefore motion against an external force. The molecular basis of this force‐generating step is unknown, although most studies have argued (but not demonstrated) that the step affected by the force load applied to the DNA is the closing of the fingers domain. Such a suggestion has been borne out by a recent in silico study using molecular dynamics simulations (Andricioaei et al., 2004). In this study, the Klentaq1 ternary complex of Li et al. (1998b) was subjected to an external force applied to the template DNA. It was shown that at a medium‐to‐high force regimen, the closing of the fingers domain appears to be most affected, thus arguing for a role of this conformational transition in the force‐generating step. It appears therefore that, although the open‐to‐closed conformational transition affecting the fingers domain is unlikely to be the kinetically observed rate‐limiting step (see Section VII), it may play an active role in motioning the DNA polymerase along the DNA. However, it is important to keep in mind that not everybody agrees with such a conclusion and that it has been argued there is no need to involve a force generation step to move DNA polymerases along the DNA as a fraying mechanism as described in Section VIII.B may be sufficient (Guajardo and Sousa, 1997). In that context, the force generated by the DNA polymerase would correspond to the indirect effect of a tightening of the polymerase around the DNA during the rate‐limiting step following the closing of the fingers domain.

IX. Conclusion The study of DNA polymerase function has experienced remarkable progress reflected in three phases. The first phase is marked by the discovery of DNA polymerases by Kornberg and colleagues (Kornberg et al., 1956; Lehman et al., 1958). This pioneering phase saw rapid advances in our understanding of DNA replication. The second phase was initiated by the structure of the Klenow fragment by Steitz and colleagues (Ollis

STRUCTURE AND MECHANISM OF DNA POLYMERASES

431

et al., 1985). This structure provided the basis for understanding the molecular basis of DNA polymerization and informed the conduct of research in this area. The third phase provided the details of the kinetic pathway of nucleotide incorporation by Benkovic and Johnson and their colleagues (Capson et al., 1992; Dahlberg and Benkovic, 1991; Donlin et al., 1991; Eger and Benkovic, 1992; Kati et al., 1992; Kuchta et al., 1987, 1988; Patel et al., 1991; Wong et al., 1991), demonstrating that the enzyme must cycle through a succession of states, some of which provide the readout necessary to discriminate against incorrect nucleotide incorporation. These papers have formed the basis for research in the polymerase field to this day. Notably, looming large in the horizon of the structural biology landscape was the capture of the ‘‘rate‐limiting’’ step. This holy grail of structural biology appeared to have been attained in 1994 and 1998 with the remarkable succession of structural characterization of the E’:p/t: ddNTP complexes for several DNA polymerases: these structures suggested that the rate‐limiting step is caused by the closure of the fingers subdomain (Doublie et al., 1998; Huang et al., 1998; Kiefer et al., 1998; Li et al., 1998b; Pelletier et al., 1994). However, this suggestion does not seem to hold very well over the passage of time. Evidence, albeit all indirect, has accumulated to suggest that, on the contrary, the captured E’:p/t:ddNTP state may be that of a fast step preceding the rate‐limiting conformational step, but not the rate‐limiting step itself. In the next few years, research will head toward elucidating the nature of the rate‐limiting step and providing the lacking details of the nucleotide incorporation cycle.

Acknowledgments This work was funded by grant #067879 from the Wellcome Trust to GW.

References Abbotts, J., Sen Gupta, D. N., Zmudzka, B., Widen, S. G., Notario, V., and Wilson, S. H. (1988). Expression of human DNA polymerase b in Escherichia coli and characterization of the recombinant enzyme. Biochemistry 27, 901–909. Ahn, J., Werneburg, B. G., and Tsai, M. D. (1997). DNA polymerase b: Structure‐fidelity relationship from pre‐steady‐state kinetic analyses of all possible correct and incorrect base pairs for wild type and R283A mutant. Biochemistry 36, 1100–1107. Andricioaei, I., Goel, A., Herschbach, D., and Karplus, M. (2004). Dependence of DNA polymerase replication rate on external forces: A model based on molecular dynamics simulations. Biophys. J. 87, 1478–1497. Astatke, M., Grindley, N. D., and Joyce, C. M. (1995). Deoxynucleoside triphosphate and pyrophosphate binding sites in the catalytically competent ternary complex for the polymerase reaction catalyzed by DNA polymerase I (Klenow fragment). J. Biol. Chem. 270, 1945–1954.

432

ROTHWELL AND WAKSMAN

Basu, A., and Modak, M. J. (1987). Identification and amino acid sequence of the deoxynucleoside triphosphate binding site in Escherichia coli DNA polymerase I. Biochemistry 26, 1704–1709. Beese, L. S., Derbyshire, V., and Steitz, T. A. (1993a). Structure of DNA polymerase I Klenow fragment bound to duplex DNA. Science 260, 352–355. Beese, L. S., Friedman, J. M., and Steitz, T. A. (1993b). Crystal structures of the Klenow fragment of DNA polymerase I complexed with deoxynucleoside triphosphate and pyrophosphate. Biochemistry 32, 14095–14101. Bork, P., Hofmann, K., Bucher, P., Neuwald, A., Altschul, S., and Koonin, E. (1997). A superFamily of conserved domains in DNA damage‐responsive cell cycle checkpoint proteins. FASEB J. 11, 68–76. Boudsocq, F., Iwai, S., Hanaoka, F., and Woodgate, R. (2001). Sulfolobus solfataricus P2 DNA polymerase IV (Dpo4): An archaeal DNA polymerase with lesion‐bypass properties akin to eukaryotic pol . Nucleic Acids Res. 15, 4607–4613. Braithwaite, D. K., and Ito, J. (1993). Compilation, alignment, and phylogenetic relationships of DNA polymerases. Nucleic Acids Res. 21, 787–802. Bruskov, V. I., and Poltev, V. I. (1979). On molecular mechanisms of nucleic acid synthesis. Fidelity aspects: 2. Contribution of protein‐nucleotide recognition. J. Theor. Biol. 78, 29–41. Burgers, P. M., Koonin, E. V., Bruford, E., Blanco, L., Burtis, K. C., Christman, M. F., Copeland, W. C., Friedberg, E. C., Hanaoka, F., Hinkle, D. C., Lawrence, C. W., Nakanishi, M., Ohmori, H., Prakash, L., Prakash, S., Reynaud, C. A., Sugino, A., Todo, T., Wang, Z., Weill, J. C., and Woodgate, R. (2001). Eukaryotic DNA polymerases: Proposal for a revised nomenclature. J. Biol. Chem. 276, 43487–43490. Cann, I. K., Komori, K., Toh, H., Kanai, S., and Ishino, Y. (1998). A heterodimeric DNA polymerase: Evidence that members of Euryarchaeota possess a distinct DNA polymerase. Proc. Natl. Acad. Sci. USA 95, 14250–14255. Capson, T. L., Peliska, J. A., Kaboord, B. F., Frey, M. W., Lively, C., Dahlberg, M., and Benkovic, S. J. (1992). Kinetic characterization of the polymerase and exonuclease activities of the gene 43 protein of bacteriophage T4. Biochemistry 31, 10984–10994. Carroll, S. S., Cowaet, M., and Benkovic, S. J. (1991). A mutant of DNA polymerase I (Klenow Fragment) with reduced fidelity. Biochemistry 30, 804–813. Dahlberg, M. E., and Benkovic, S. J. (1991). Kinetic mechanism of DNA polymerase I (Klenow fragment): Identification of a second conformational change and evaluation of the internal equilibrium constant. Biochemistry 30, 4835–4843. Das, D., and Georgiadis, M. M. (2004). The crystal structure of the monomeric reverse transcriptase from Moloney murine leukemia virus. Structure 12, 819–829. Delarue, M., Poch, O., Tordo, N., Moras, D., and Argos, P. (1990). An attempt to unify the structure of polymerases. Protein Eng. 3, 461–467. Dianov, G. L., Sleeth, K. M., Dianova, II, and Allinson, S. L. (2003). Repair of abasic sites in DNA. Mutat. Res. 531, 157–163. Ding, J., Das, K., Hsiou, Y., Sarafianos, S. G., Clark, A. D., Jr., Jacobo‐Molina, A., Tantillo, C., Hughes, S. H., and Arnold, E. (1998). Structure and functional implications of the polymerase active site region in a complex of HIV-1 RT with a double‐stranded DNA template‐primer and an antibody Fab fragment at 2.8 A˚ resolution. J. Mol. Biol. 284, 1095–1111. Ding, J., Das, K., Moereels, H., Koymans, L., Andries, K., Janssen, P. A., Hughes, S. H., and Arnold, E. (1995a). Structure of HIV-1 RT/TIBO R 86183 complex reveals similarity in the binding of diverse nonnucleoside inhibitors. Nat. Struct. Biol. 2, 407–415.

STRUCTURE AND MECHANISM OF DNA POLYMERASES

433

Ding, J., Das, K., Tantillo, C., Zhang, W., Clark, A. D., Jr., Jessen, S., Lu, X., Hsiou, Y., Jacobo‐Molina, A., Andries, K., Pauwels, R., Moereels, H., Koymans, L., Janssen, P. A. J., Smith, R. H., Jr., Kroeger Koepke, M., Michejda, C. J., Hughes, S. H., and Arnold, E. (1995b). Structure of HIV-1 reverse transcriptase in a complex with the non‐nucleoside inhibitor alpha‐APA R 95845 at 2.8 A˚ resolution. Structure 3, 365–379. Divita, G., Rittinger, K., Geourjon, C., Deleage, G., and Goody, R. S. (1995a). Dimerization kinetics of HIV-1 and HIV-2 reverse transcriptase: A two step process. J. Mol. Biol. 245, 508–521. Divita, G., Rittinger, K., Restle, T., Immendorfer, U., and Goody, R. S. (1995b). Conformational stability of dimeric HIV-1 and HIV-2 reverse transcriptases. Biochemistry 34, 16337–16346. Dominguez, O., Ruiz, J. F., Lain de Lera, T., Garcia‐Diaz, M., Gonzalez, M. A., Kirchhoff, T., Martinez, A. C., Bernad, A., and Blanco, L. (2000). DNA polymerase m (Pol m), homologous to TdT, could act as a DNA mutator in eukaryotic cells. EMBO J. 19, 1731–1742. Dong, Q., Copeland, W. C., and Wang, T. S. (1993). Mutational studies of human DNA polymerase a: Identification of residues critical for deoxynucleotide binding and misinsertion fidelity of DNA synthesis. J. Biol. Chem. 268, 24163–24174. Donlin, M. J., Patel, S. S., and Johnson, K. A. (1991). Kinetic partitioning between the exonuclease and polymerase sites in DNA error correction. Biochemistry 30, 538–546. Doublie, S., Tabor, S., Long, A. M., Richardson, C. C., and Ellenberger, T. (1998). Crystal structure of a bacteriophage T7 DNA replication complex at 2.2 A resolution. Nature 391, 251–258. Dunlap, C. A., and Tsai, M. D. (2002). Use of 2‐aminopurine and tryptophan fluorescence as probes in kinetic analyses of DNA polymerase b. Biochemistry 41, 11226–11235. Echols, H., and Goodman, M. F. (1991). Fidelity mechanism in DNA replication. Annu. Rev. Biochem. 60, 477–511. Eger, B. T., and Benkovic, S. J. (1992). Minimal kinetic mechanism for misincorporation by DNA polymerase I (Klenow fragment). Biochemistry 31, 9227–9236. Engel, J. D., and von Hippel, P. H. (1978). D(M6ATP) as a probe of the fidelity of base incorporation into polynucleotides by Escherichia coli DNA polymerase I. J. Biol. Chem. 253, 935–939. Eom, S. H., Wang, J., and Steitz, T. A. (1996). Structure of Taq ploymerase with DNA at the polymerase active site. Nature 382, 278–281. Esnouf, R., Ren, J., Ross, C., Jones, Y., Stammers, D., and Stuart, D. (1995). Mechanism of inhibition of HIV-1 reverse transcriptase by non‐nucleoside inhibitors. Nat. Struct. Biol. 2, 303–308. Fiala, K. A., and Suo, Z. (2004a). Mechanism of DNA polymerization catalyzed by Sulfolobus solfataricus P2 DNA polymerase IV. Biochemistry 43, 2116–2125. Fiala, K. A., and Suo, Z. (2004b). Pre‐steady‐state kinetic studies of the fidelity of Sulfolobus solfataricus P2 DNA polymerase IV. Biochemistry 43, 2106–2115. Franklin, M. C., Wang, J., and Steitz, T. A. (2001). Structure of the replicating complex of a pol a Family DNA polymerase. Cell 105, 657–667. Freemont, P. S., Friedman, J. M., Beese, L. S., Sanderson, M. R., and Steitz, T. A. (1988). Cocrystal structure of an editing complex of Klenow fragment with DNA. Proc. Natl. Acad. Sci. USA 85, 8924–8928.

434

ROTHWELL AND WAKSMAN

Frey, M. W., Sowers, L. C., Millar, D. P., and Benkovic, S. J. (1995). The nucleotide analog 2‐aminopurine as a spectroscopic probe of nucleotide incorporation by the Klenow fragment of Escherichia coli polymerase I and bacteriophage T4 DNA polymerase. Biochemistry 34, 9185–9192. Friedberg, E. C., Feaver, W. J., and Gerlach, V. L. (2000). The many faces of DNA polymerases: Strategies for mutagenesis and for mutational avoidance. Proc. Natl. Acad. Sci. USA 97, 5681–5683. Friedberg, E. C., and Gerlach, V. L. (1999). Novel DNA polymerases offer clues to the molecular basis of mutagenesis. Cell 98, 413–416. Garcia‐Diaz, M., Dominguez, O., Lopez‐Fernandez, L. A., de Lera, L. T., Saniger, M. L., Ruiz, J. F., Parraga, M., Garcia‐Ortiz, M. J., Kirchhoff, T., and del Mazo, J. (2000). DNA polymerase l (Pol l), a novel eukaryotic DNA polymerase with a potential role in meiosis. J. Mol. Biol. 301, 851–867. Gillin, F. D., and Nossal, N. G. (1976). Control of mutation frequency by bacteriophage T4 DNA polymerase: II. Accuracy of nucleotide selection by the L88 mutator, CB120 antimutator, and wild type phage T4 DNA polymerases. J. Biol. Chem. 251, 5225–5232. Goodman, M. F., and Tippin, B. (2000). Sloppier copier DNA polymerases involved in genome repair. Curr. Opin. Genet. Dev. 10, 162–168. Gotte, M., Li, X., and Wainberg, M. A. (1999). HIV-1 reverse transcription: A brief overview focused on structure‐function relationships among molecules involved in initiation of the reaction. Arch. Biochem. Biophys. 365, 199–210. Greider, C. W., and Blackburn, E. H. (1985). Identification of a specific telomere terminal transferase activity in Tetrahymena extracts. Cell 43, 405–413. Greider, C. W., and Blackburn, E. H. (1989). A telomeric sequence in the RNA of Tetrahymena telomerase required for telomere repeat synthesis. Nature 337, 331–337. Guajardo, R., and Sousa, R. (1997). A model for the mechanism of polymerase translocation. J. Mol. Biol. 265, 8–19. Hashimoto, H., Nishioka, M., Fujiwara, S., Takagi, M., Imanaka, T., Inoue, T., and Kai, Y. (2001). Crystal structure of DNA polymerase from hyperthermophilic archaeon Pyrococcus kodakaraensis KOD1. J. Mol. Biol. 306, 469–477. Hogg, M., Wallace, S. S., and Doublie, S. (2004). Crystallographic snapshots of a replicative DNA polymerase encountering an abasic site. EMBO J. 23, 1483–1493. Hopfner, K. P., Eichinger, A., Engh, R. A., Laue, F., Ankenbauer, W., Huber, R., and Angerer, B. (1999). Crystal structure of a thermostable type B DNA polymerase from Thermococcus gorgonarius. Proc. Natl. Acad. Sci. USA 96, 3600–3605. Hsieh, J. C., Zinnen, S., and Modrich, P. (1993). Kinetic mechanism of the DNA‐ dependent DNA polymerase activity of human immunodeficiency virus reverse transcriptase. J. Biol. Chem. 268, 24607–24613. Hsiou, Y., Ding, J., Das, K., Clark, A. D., Jr., Hughes, S. H., and Arnold, E. (1996). Structure of unliganded HIV-1 reverse transcriptase at 2.7 A˚ resolution: Implications of conformational changes for polymerization and inhibition mechanisms. Structure 4, 853–860. Huang, H., Chopra, R., Verdine, G. L., and Harrison, S. C. (1998). Structure of a covalently trapped catalytic complex of HIV-1 reverse transcriptase: Implications for drug resistance. Science 282, 1669–1675. Ito, J., and Braithwaite, D. K. (1991). Compilation and alignment of DNA polymerase sequences. Nucl. Acids Res. 19, 4045–4057.

STRUCTURE AND MECHANISM OF DNA POLYMERASES

435

Jacobo‐Molina, A., Ding, J., Nanni, R. G., Clark, A. D., Jr., Lu, X., Tantillo, C., Williams, R. L., Kamer, G., Ferris, A. L., Clark, P., Hizi, A., Hughes, S. H., and Arnold, E. (1993). Crystal structure of human immunodeficiency virus type 1 reverse transcriptase complexed with double‐stranded DNA at 3.0 A˚ resolution shows bent DNA. Proc. Natl. Acad. Sci. USA 90, 6320–6324. Jaeger, J., Restle, T., and Steitz, T. A. (1998). The structure of HIV-1 reverse transcriptase complexed with an RNA pseudoknot inhibitor. EMBO J. 17, 4535–4542. Jezewska, M. J., Galletto, R., and Bujalowski, W. (2002). Dynamics of gapped DNA recognition by human polymerase b. J. Biol. Chem. 277, 20316–20327. Jezewska, M. J., Rajendran, S., and Bujalowski, W. (2001a). Energetics and specificity of Rat DNA polymerase b interactions with template‐primer and gapped DNA substrates. J. Biol. Chem. 276, 16123–16136. Jezewska, M. J., Rajendran, S., and Bujalowski, W. (2001b). Interactions of the 8‐kDa domain of rat DNA polymerase b with DNA. Biochemistry 40, 3295–3307. Johnson, R. E., Washington, M. T., Prakash, S., and Prakash, L. (1999). Bridging the gap: A family of novel DNA polymerases that replicate faulty DNA. Proc. Natl. Acad. Sci. USA 96, 12224–12226. Joyce, C. M., and Steitz, T. A. (1994). Function and structure relationships in DNA polymerases. Annu. Rev. Biochem. 63, 777–822. Kati, W. M., Johnson, K. A., Jerva, L. F., and Anderson, K. S. (1992). Mechanism and fidelity of HIV reverse transcriptase. J. Biol. Chem. 267, 25988–25997. Kelman, Z., and O’Donnell, M. (1995). DNA polymerase III holoenzyme: Structure and function of a chromosomal replicating machine. Ann. Rev. Biochem. 64, 171–200. Kensch, O., Restle, T., Wo¨ hrl, B. M., Goody, R. S., and Steinhoff, H. J. (2000). Temperature‐dependent equilibrium between the open and closed conformation of the p66 subunit of HIV-1 reverse transcriptase revealed by site‐directed spin labelling. J. Mol. Biol. 301, 1029–1039. Kiefer, J. R., Mao, C., Braman, J. C., and Beese, L. S. (1998). Visualizing DNA replication in a catalytically active Bacillus DNA polymerase crystal. Nature 391, 304–307. Kim, S. J., Beard, W. A., Harvey, J., Shock, D. D., Knutson, J. R., and Wilson, S. H. (2003). Rapid segmental and subdomain motions of DNA polymerase b. J. Biol. Chem. 278, 5072–5081. Kim, Y., Eom, S. H., Wang, J., Lee, D. S., Suh, S. W., and Steitz, T. A. (1995). Crystal structure of Thermus aquaticus DNA polymerase. Nature 376, 612–616. Klenow, H., and Overgaard‐Hansen, K. (1970). Proteolytic cleavage of DNA polymerase from Escherichia coli B into an exonuclease unit and a polymerase unit. FEBS Lett. 6, 25–27. Kohlstaedt, L. A., Wang, J., Friedman, J. M., Rice, P. A., and Steitz, T. A. (1992). Crystal structure at 3.5 A resolution of HIV-1 reverse transcriptase complexed with an inhibitor. Science 256, 1783–1790. Kornberg, A. (1980). ‘‘DNA Replication.’’ W. H. Freeman, San Francisco. Kornberg, A., and Baker, T. A. (1992). ‘‘DNA Replication.’’ W. H. Freeman, New York. Kornberg, A., Lehman, I. R., Bessman, M. J., and Simms, E. S. (1956). Enzymic synthesis of deoxyribonucleic acid. Biochim. Biophys. Acta 21, 197–198. Korolev, S., Nayal, M., Barnes, W. M., Di Cera, E., and Waksman, G. (1995). Crystal structure of the large fragment of Thermus aquaticus DNA polymerase I at 2.5 A resolution: Structural basis for thermostability. Proc. Natl. Acad. Sci. USA 92, 9264–9268.

436

ROTHWELL AND WAKSMAN

Krahn, J. M., Beard, W. A., Miller, H., Grollman, A. P., and Wilson, S. H. (2003). Structure of DNA polymerase b with the mutagenic DNA lesion 8‐oxodeoxyguanine reveals structural insights into its coding potential. Structure 11, 121–127. Kraynov, V. S., Werneburg, B. G., Zhong, X., Lee, H., Ahn, J., and Tsai, M. D. (1997). DNA polymerase b: Analysis of the contributions of tyrosine-271 and asparagine-279 to substrate specificity and fidelity of DNA replication by pre‐steady‐state kinetics. Biochem. J. 323, 103–111. Kuchta, R. D., Benkovic, P., and Benkovic, S. J. (1988). Kinetic mechanism whereby DNA polymerase I (Klenow) replicates DNA with high fidelity. Biochemistry 27, 6716–6725. Kuchta, R. D., Mizrahi, V., Benkovic, P. A., Johnson, K. A., and Benkovic, S. J. (1987). Kinetic mechanism of DNA polymerase I (Klenow). Biochemistry 26, 8410–8417. Kumar, A., Abbotts, J., Karawya, E. M., and Wilson, S. H. (1990). Identification and properties of the catalytic domain of mammalian DNA polymerase b. Biochemistry 29, 7156–7159. Lehman, I. R., Zimmerman, S. B., Adler, J., Bessman, M. J., Simms, E. S., and Kornberg, A. (1958). Enzymatic synthesis of deoxyribonucleic acid: V. Chemical composition of enzymatically synthesized deoxyribonucleic acid. Proc. Natl. Acad. Sci. USA 44, 1191–1196. Li, Y., Dutta, S., Doublie, S., Bdour, H. M., Taylor, J. S., and Ellenberger, T. (2004). Nucleotide insertion opposite a cis‐syn thymine dimer by a replicative DNA polymerase from bacteriophage T7. Nat. Struct. Mol. Biol. 11, 784–790. Li, Y., Kong, Y., Korolev, S., and Waksman, G. (1998a). Crystal structures of the Klenow fragment of Thermus aquaticus DNA polymerase I complexed with deoxyribonucleoside triphosphates. Protein Sci. 7, 1116–1123. Li, Y., Korolev, S., and Waksman, G. (1998b). Crystal structures of open and closed forms of binary and ternary complexes of the large fragment of Thermus aquaticus DNA polymerase I: Structural basis for nucleotide incorporation. EMBO J. 17, 7514–7525. Li, Y., and Waksman, G. (2001a). Crystal structures of a ddATP‐, ddTTP‐, ddCTP‐, and ddGTP‐trapped ternary complex of Klentaq1: Insights into the nucleotide incorporation and selectivity. Protein Sci. 10, 1225–1233. Li, Y., and Waksman, G. (2001b). Structural studies of the Klentaq1 DNA polymerase. Curr. Organic Chem. 5, 871–883. Lin, T. C., Karam, G., and Konigsberg, W. H. (1994). Isolation, characterization, and kinetic properties of truncated forms of T4 DNA polymerase that exhibit 3’-5’ exonuclease activity. J. Biol. Chem. 269, 19286–19294. Ling, H., Boudsocq, F., Plosky, B. S., Woodgate, R., and Yang, W. (2003). Replication of a cis‐syn thymine dimer at atomic resolution. Nature 424, 1083–1087. Ling, H., Boudsocq, F., Woodgate, R., and Yang, W. (2001). Crystal structure of a Y‐Family DNA polymerase in action: A mechanism for error‐prone and lesion‐bypass replication. Cell 107, 91–102. Ling, H., Sayer, J. M., Plosky, B. S., Yagi, H., Boudsocq, F., Woodgate, R., Jerina, D. M., and Yang, W. (2004). Crystal structure of a benzo[a]pyrene diol epoxide adduct in a ternary complex with a DNA polymerase. Proc. Natl. Acad. Sci. USA 101, 2265–2269. Loeb, L. A., and Kunkel, T. A. (1982). Fidelity of DNA synthesis. Ann. Rev. Biochem. 52, 429–447. Lu, H., Macosko, J., Habel‐Rodriguez, D., Keller, R. W., Brozik, J. A., and Keller, D. J. (2004). Closing of the fingers domain generates motor forces in the HIV reverse transcriptase. J. Biol. Chem., 279, 54529–54532.

STRUCTURE AND MECHANISM OF DNA POLYMERASES

437

Madrid, M., Jacobo‐Molina, A., Ding, J., and Arnold, E. (1999). Major subdomain rearrangement in HIV-1 reverse transcriptase simulated by molecular dynamics. Proteins 35, 332–337. Maier, D., Bensimon, D., and Croquette, V. (2000). Replication by a single DNA polymerase of a stretched single‐stranded DNA. Proc. Natl. Acad. Sci. USA 97, 12002–12007. Martins, A., Ribeiro, G., Marques, M. I., and Costa, J. V. (1994). Genetic identification and nucleotide sequence of the DNA polymerase gene of African swine fever virus. Nucleic Acids Res. 22, 208–213. Matsumoto, Y., and Bogenhagen, D. F. (1991). Repair of a synthetic abasic site involves concerted reactions of DNA synthesis followed by excision and ligation. Mol. Cell. Biol. 11, 4441–4447. Matsumoto, Y., and Kim, K. (1995). Excision of deoxyribose phosphate residues by DNA polymerase b during DNA repair. Science 269, 699–702. Nair, D. T., Johnson, R. E., Prakash, S., Prakash, L., and Aggarwal, A. K. (2004). Replication by human DNA polymerase‐i occurs by Hoogsteen base‐pairing. Nature 430, 377–380. Ollis, D. L., Brick, P., Hamlin, R., Xuong, N. G., and Steitz, T. A. (1985). Structure of large fragment of Escherichia coli DNA polymerase I complexed with dTMP. Nature 313, 762–766. Pata, J. D., Stirtan, W. G., Goldstein, S. W., and Steitz, T. A. (2004). Structure of HIV-1 reverse transcriptase bound to an inhibitor active against mutant reverse transcriptases resistant to other nonnucleoside inhibitors. Proc. Natl. Acad. Sci. USA 101, 10548–10553. Patel, S. S., Wong, I., and Johnson, K. A. (1991). Pre‐steady‐state kinetic analysis of processive DNA replication including complete characterization of an exonuclease‐deficient mutant. Biochemistry 30, 511–525. Pavlov, A. R., and Karam, J. D. (1994). Binding specificity of T4 DNA polymerase to RNA. J. Biol. Chem. 269, 12968–12972. Pelletier, H., Sawaya, M. R., Kumar, A., Wilson, S. H., and Kraut, J. (1994). Structures of ternary complexes of rat DNA polymerase b, a DNA template‐primer, and ddCTP. Science 264, 1891–1903. Pelletier, H., Sawaya, M. R., Wolfle, W., Wilson, S. H., and Kraut, J. (1996). Crystal structures of human DNA polymerase b complexed with DNA: Implications for catalytic mechanism, processivity, and fidelity. Biochemistry 35, 12742–12761. Petruska, J., Sowers, L. C., and Goodman, M. F. (1986). Comparison of nucleotide interactions in water, proteins, and vacuum: Model for DNA polymerase fidelity. Proc. Natl. Acad. Sci. USA 83, 1559–1562. Prasad, R., Widen, S., Singhal, R., Watkins, J., Prakash, L., and Wilson, S. (1993). Yeast open reading frame YCR14C encodes a DNA b‐polymerase‐like enzyme. Nucl. Acids Res. 21, 5301–5307. Purohit, V., Grindley, N. D., and Joyce, C. M. (2003). Use of 2‐aminopurine fluorescence to examine conformational changes during nucleotide incorporation by DNA polymerase I (Klenow fragment). Biochemistry 42, 10200–10211. Rajendran, S., Jezewska, M. J., and Bujalowski, W. (1998). Human DNA polymerase b recognizes single‐stranded DNA using two different binding modes. J. Biol. Chem. 273, 31021–31031. Rajendran, S., Jezewska, M. J., and Bujalowski, W. (2001). Recognition of template‐ primer and gapped DNA substrates by the human DNA polymerase b. J. Mol. Biol. 308, 477–500.

438

ROTHWELL AND WAKSMAN

Ren, J., Bird, L. E., Chamberlain, P. P., Stewart‐Jones, G. B., Stuart, D. I., and Stammers, D. K. (2002). Structure of HIV-2 reverse transcriptase at 2.35‐A˚ resolution and the mechanism of resistance to non‐nucleoside inhibitors. Proc. Natl. Acad. Sci. USA 99, 14410–14415. Ren, J., Esnouf, R., Garman, E., Somers, D., Ross, C., Kirby, I., Keeling, J., Darby, G., Jones, Y., and Stuart, D. (1995). High resolution structures of HIV-1 RT from four RT‐inhibitor complexes. Nat. Struct. Biol. 2, 293–302. Ren, J., Esnouf, R. M., Hopkins, A. L., Warren, J., Balzarini, J., Stuart, D. I., and Stammers, D. K. (1998). Crystal structures of HIV-1 reverse transcriptase in complex with carboxanilide derivatives. Biochemistry 37, 14394–14403. Ren, J., Nichols, C. E., Chamberlain, P. P., Weaver, K. L., Short, S. A., and Stammers, D. K. (2004). Crystal structures of HIV-1 reverse transcriptases mutated at codons 100, 106 and 108 and mechanisms of resistance to non‐nucleoside inhibitors. J. Mol. Biol. 336, 569–578. Rittinger, K., Divita, G., and Goody, R. S. (1995). Human immunodeficiency virus reverse transcriptase substrate‐induced conformational changes and the mechanism of inhibition by nonnucleoside inhibitors. Proc. Natl. Acad. Sci. USA 92, 8046–8049. Rodgers, D. W., Gamblin, S. J., Harris, B. A., Ray, S., Culp, J. S., Hellmig, B., Woolf, D. J., Debouck, C., and Harrison, S. C. (1995). The structure of unliganded reverse transcriptase from the human immunodeficiency virus type 1. Proc. Natl. Acad. Sci. USA 92, 1222–1226. Rodriguez, A. C., Park, H. W., Mao, C., and Beese, L. S. (2000). Crystal structure of a pol a Family DNA polymerase from the hyperthermophilic archaeon Thermococcus sp. 9 degrees N-7. J. Mol. Biol. 299, 447–462. Roth, M. J., Tanese, N., and Goff, S. P. (1985). Purification and characterization of murine retroviral reverse transcriptase expressed in Escherichia coli. J. Biol. Chem. 260, 9326–9335. Rothwell, P. J., Berger, S., Kensch, O., Felekyan, S., Antonik, M., Wohrl, B. M., Restle, T., Goody, R. S., and Seidel, C. A. (2003). Multiparameter single‐molecule fluorescence spectroscopy reveals heterogeneity of HIV-1 reverse transcriptase:primer/template complexes. Proc. Natl. Acad. Sci. USA 100, 1655–1660. Sarafianos, S. G., Clark, A. D., Jr., Das, K., Tuske, S., Birktoft, J. J., Ilankumaran, P., Ramesha, A. R., Sayer, J. M., Jerina, D. M., Boyer, P. L., Hughes, S. H., and Arnold, E. (2002). Structures of HIV-1 reverse transcriptase with pre‐ and post‐translocation AZTMP‐terminated DNA. EMBO J. 21, 6614–6624. Sarafianos, S. G., Das, K., Tantillo, C., Clark, A. D., Jr., Ding, J., Whitcomb, J. M., Boyer, P. L., Hughes, S. H., and Arnold, E. (2001). Crystal structure of HIV-1 reverse transcriptase in complex with a polypurine tract RNA:DNA. EMBO J. 20, 1449–1461. Sawaya, M. R., Pelletier, H., Kumar, A., Wilson, S. H., and Kraut, J. (1994). Crystal structure of rat DNA polymerase b: Evidence for a common polymerase mechanism. Science 264, 1930–1935. Sawaya, M. R., Prasad, R., Wilson, S. H., Kraut, J., and Pelletier, H. (1997). Crystal structures of human DNA polymerase b complexed with gapped and nicked DNA: Evidence for an induced fit mechanism. Biochemistry 36, 11205–11215. Shah, A. M., Maitra, M., and Sweasy, J. B. (2003). Variants of DNA polymerase b extend mispaired DNA due to increased affinity for nucleotide substrate. Biochemistry 42, 10709–10717. Silvian, L. F., Toth, E. A., Pham, P., Goodman, M. F., and Ellenberger, T. (2001). Crystal structure of a DinB Family error‐prone DNA polymerase from Sulfolobus solfataricus. Nat. Struct. Biol. 8, 984–989.

STRUCTURE AND MECHANISM OF DNA POLYMERASES

439

Sloane, D. L., Goodman, M. F., and Echols, H. (1988). The fidelity of base selection by the polymerase subunit of DNA polymerase III holoenzyme. Nucleic Acids Res. 16, 6465–6475. Sluis‐Cremer, N., Arion, D., and Parniak, M. A. (2000). Molecular mechanisms of HIV-1 resistance to nucleoside reverse transcriptase inhibitors (NRTIs). Cell Mol. Life Sci. 57, 1408–1422. Smerdon, S. J., Jager, J., Wang, J., Kohlstaedt, L. A., Chirino, A. J., Friedman, J. M., Rice, P. A., and Steitz, T. A. (1994). Structure of the binding site for nonnucleoside inhibitors of the reverse transcriptase of human immunodeficiency virus type 1. Proc. Natl. Acad. Sci. USA 91, 3911–3915. Steitz, T. A. (1993). DNA‐ and RNA‐dependent DNA polymerases. Curr. Opin. Struct. Biol. 3, 31–38. Steitz, T. A. (1994). A unified polymerase mechanism for nonhomologous DNA and RNA polymerases. Science 266, 2022–2026. Steitz, T. A. (1998). A mechanism for all polymerases. Nature 391, 231–232. Suzuki, M., Baskin, D., Hood, L., and Loeb, L. A. (1996). Random mutagenesis of Thermus aquaticus DNA polymerase I: Concordance of immutable sites in vivo with the crystal structure. Proc. Natl. Acad. Sci. USA 93, 9670–9675. Tabor, S., Huber, H. E., and Richardson, C. C. (1987). Escherichia coli thioredoxin confers processivity on the DNA polymerase activity of the gene 5 protein of bacteriophage T7. J. Biol. Chem. 262, 16212–16223. Topal, M. D., Di Guiseppi, S. R., and Sinha, N. K. (1980). Molecular basis for substitution mutations: Effect of primer terminal and template residues on nucleotide selection by phage T4 DNA polymerase in vitro. J. Biol. Chem. 255, 11717–11724. Trincao, J., Johnson, R. E., Escalante, C. R., Prakash, S., Prakash, L., and Aggarwal, A. K. (2001). Structure of the catalytic core of S. cerevisiae DNA polymerase : Implications for translesion DNA synthesis. Mol. Cell 8, 417–426. Tuerk, C., Eddy, S., Parma, D., and Gold, L. (1990). Autogenous translational operator recognized by bacteriophage T4 DNA polymerase. J. Mol. Biol. 213, 749–761. Uemori, T., Sato, Y., Doi, H., and Ishino, Y. (1997). A novel DNA polymerase in the hyperthermophilic acrhaeon, Pyrococcus furiosus: Gene cloning, expression and characterization. Genes to Cell 2, 499–512. Uljon, S. N., Johnson, R. E., Edwards, T. A., Prakash, S., Prakash, L., and Aggarwal, A. K. (2004). Crystal structure of the catalytic core of human DNA polymerase k. Structure 12, 1395–1404. Vande Berg, B. J., Beard, W. A., and Wilson, S. H. (2001). DNA structure and aspartate 276 influence nucleotide binding to human DNA polymerase b. Implication for the identity of the rate‐limiting conformational change. J. Biol. Chem. 276, 3408–3416. Wang, C. C., Pavlov, A., and Karam, J. D. (1997a). Evolution of RNA‐binding specificity in T4 DNA polymerase. J. Biol. Chem. 272, 17703–17710. Wang, J., Sattar, A. K., Wang, C. C., Karam, J. D., Konigsberg, W. H., and Steitz, T. A. (1997b). Crystal structure of a pol a Family replication DNA polymerase from bacteriophage RB69. Cell 89, 1087–1099. Washington, M. T., Prakash, L., and Prakash, S. (2001). Yeast DNA polymerase  utilizes an induced‐fit mechanism of nucleotide incorporation. Cell 107, 917–927. Watson, J. D., and Crick, F. H. (1953a). Genetical implications of the structure of deoxyribonucleic acid. Nature 171, 964–967. Watson, J. D., and Crick, F. H. (1953b). Molecular structure of nucleic acids: A structure for deoxyribose nucleic acid. Nature 171, 737–738.

440

ROTHWELL AND WAKSMAN

Werneburg, B. G., Ahn, J., Zhong, X., Hondal, R. J., Kraynov, V. S., and Tsai, M. D. (1996). DNA polymerase b: Pre‐steady‐state kinetic analysis and roles of arginine–283 in catalysis and fidelity. Biochemistry 35, 7041–7050. Wo¨ hrl, B. M., Krebs, R., Goody, R. S., and Restle, T. (1999). Refined model for primer/ template binding by HIV-1 reverse transcriptase: Pre‐steady‐state kinetic analyses of primer/template binding and nucleotide incorporation events distinguish between different binding modes depending on the nature of the nucleic acid substrate. J. Mol. Biol. 292, 333–344. Wong, I., Patel, S. S., and Johnson, K. A. (1991). An induced‐fit kinetic mechanism for DNA replication fidelity: Direct measurement by single‐turnover kinetics. Biochemistry 30, 526–537. Woodgate, R. (1999). A plethora of lesion‐replicating DNA polymerases. Genes Dev. 13, 2191–2195. Wuite, G. J. L., Smith, S. B., Young, M., Keller, D., and Bustamante, C. (2000). Single‐ molecule studies of the effect of template tension on T7 DNA polymerase activity. Nature 404, 103–106. Yang, G., Franklin, M., Li, J., Lin, T. C., and Konigsberg, W. (2002a). Correlation of the kinetics of finger domain mutants in RB69 DNA polymerase with its structure. Biochemistry 41, 2526–2534. Yang, L., Arora, K., Beard, W. A., Wilson, S. H., and Schlick, T. (2004). Critical role of magnesium ions in DNA polymerase b’s closing and active site assembly. J. Am. Chem. Soc. 126, 8441–8453. Yang, L., Beard, W. A., Wilson, S. H., Broyde, S., and Schlick, T. (2002b). Polymerase b simulations suggest that Arg258 rotation is a slow step rather than large subdomain motions per se. J. Mol. Biol. 317, 651–671. Yang, W. (2003). Damage repair DNA polymerases Y. Curr. Opin. Struct. Biol. 13, 23–30. Zhao, Y., Jeruzalmi, D., Moarefi, I., Leighton, L., Lasken, R., and Kuriyan, J. (1999). Crystal structure of an archaebacterial DNA polymerase. Structure Fold Des. 7, 1189–1199. Zhong, X., Patel, S. S., Werneburg, B. G., and Tsai, M. D. (1997). DNA polymerase b: Multiple conformational changes in the mechanism of catalysis. Biochemistry 36, 11891–11900. Zhou, B. L., Pata, J. D., and Steitz, T. A. (2001). Crystal structure of a DinB lesion bypass DNA polymerase catalytic fragment reveals a classic polymerase catalytic domain. Mol. Cell 8, 427–437.

AUTHOR INDEX

A Abbotts, J., 404, 405, 409 Abee, T., 351 Abildgaard, F., 355, 356, 366, 367 Abrahams, J. P., 349, 352, 355, 359, 365 Abramson, P. D., 163, 329 Ackmann, M., 273 Adelstein, R. S., 165 Adhikari, B. B., 217 Adler, J., 409, 430 Aebi, U., 127, 129, 356 Agard, D. A., 282, 283 Aggarwal, A. K., 412, 423 Aggeler, R., 353, 362, 363, 365, 366, 368 Ahn, J., 406, 417, 427 Ahringer, J., 273 Al-Bassam, J., 274, 277, 315 Al-Khayat, H. A., 1, 2, 4, 36, 40, 52, 58, 59, 60, 61, 63, 67, 71, 72, 73, 74, 78, 79, 87, 110, 162, 195, 196, 197, 213, 215, 216, 221, 225, 226, 227, 229, 231, 234, 237, 247 Alamo, L., 42, 71, 74 Alber, T., 125, 126, 130, 288 Albertson, D. G., 259 Allan, V., 286 Allinson, S. L., 405 Alonso, M., 280, 281, 308 Alper, S. L., 350 Altendorf, K., 355, 356, 367, 368 Altmann, S., 305, 308 Altschul, S., 405 Amemiya, Y., 245 Amenitsch, H., 231, 239, 240, 245 Amos, L. A., 1, 3, 261, 262, 265, 266, 267, 272, 273, 276, 277, 280, 281, 287, 308, 338

Amzel, L. M., 352 Amzel, M., 352, 355 Andersen, P., 347 Anderson, K. I., 391 Anderson, K. S., 406, 417, 426, 431 Anderson, R., 280, 305, 307, 317, 318, 319, 324, 327 Andreu, J. M., 267 Andricioaei, I., 430 Andries, K., 413 Andrieux, A., 275 Angerer, B., 410 Angert, I., 36, 145, 171, 223, 330 Angevine, C. M., 366, 367 Ankenbauer, W., 410 Ansari, S., 148 Anson, M., 329 Antonik, M., 409 Antonio, B. J., 367 Antony, C., 285 Apps, D. K., 347 April, E. W., 40 Arai, H., 347, 357, 360, 365 Arata, Y., 368 Aravind, L., 7, 336 Arechaga, I., 360 Argos, P., 126, 129, 403 Arion, D., 423 Armbru¨ster, A., 357 Arnal, I., 267, 308 Arnold, E., 412, 413, 417 Arora, K., 425, 426 Asao, T., 265, 271 Asenjo, A. B., 305 Ashley, C. C., 231, 239, 240, 245 Astatke, M., 418 Astier, C., 45 Atherton, J., 112, 113 Atkinson, R. A., 45 Auer, K., 289

441

442

AUTHOR INDEX

B Baas, P. W., 272, 274, 279 Bacchiocchi, C., 149 Bagni, M. A., 108, 231, 239, 240, 245 Bagshaw, C. R., 179, 181, 182, 330 Bagshaw, R. J., 165 B€ahler, M., 163 Bailey, K., 131, 150 Baker, A. M. E., 385 Baker, J. E., 163 Baker, J. P., 3 Baker, T. A., 403, 404, 417 Baldwin, P. R., 59 Baleja, J. D., 359, 368 Ballabio, A., 279 Balzarini, J., 413 Bang, M. L., 93, 94, 95, 110, 112, 113 Barany, M., 133 Barber, J., 360, 361 Barford, D., 284, 334, 335 Barlow, D. P., 90, 93, 113 Barnes, W. M., 410 Barnett, V. A., 182 Barnhart, L. E., 334, 336 Barry, J. S., 45, 47 Baskin, D., 418 Bass, A. H., 217 Bassford, M. L., 231 Basu, A., 418 Batra, R., 187 Bauer, C. B., 9, 165 Baumann, B. A., 228 Baur, C. P., 273 Bazirgan, O. A., 279 Bdour, H. M., 418 Beard, W. A., 412, 424, 425, 426 Becker, A., 172, 175, 330, 331 Beese, L. S., 143, 410, 414, 418, 431 Bekarova, T., 239 Belknap, B., 185, 229, 230 Bell, S. P., 94, 106 Bement, W. M., 180 ben-Avraham, D., 145 Benkovic, P. A., 406, 417, 426, 428, 431 Benkovic, S. J., 404, 406, 417, 426, 428, 431 Bennett, P. M., 29, 63 Benning, M. M., 23, 51, 145, 165, 166, 167, 223 Bensimon, D., 429, 430

Benyamin, Y., 45 Beraud, C., 307, 317, 318, 319, 339 Berger, B., 305 Berger, S., 409 Berman, H. M., 301, 308 Bernad, A., 404 Bernal, R. A., 361 Bernstor, S., 231, 239, 240, 245 Berri, M., 91, 108, 109 Bershitskaya, O. N., 223, 231 Bershitsky, S. Y., 223, 230, 231, 245 Bessman, M. J., 409, 430 Bestard, J. A., 368 Besterman, J. M., 64 Bhat, T. N., 301, 308 Bi, Y., 356, 368 Bianchet, M., 352, 355 Biernat, J., 275 Bingham, J. B., 286 Biou, V., 305, 307 Bird, L. E., 412 Birkenh€ager, R., 355, 356 Birktoft, J. J., 413 Birolo, L., 45 Birren, B., 93 Bisher, M. E., 91 Blackburn, E. H., 406 Blair, D. A., 11 Blanco, L., 404 Blazsek, A., 127 Block, S. M., 279 Blow, D. M., 204, 206 Boekema, E. J., 361, 368 Boesecke, P., 231, 238, 245 Bogenhagen, D. F., 405 Bollag, D. M., 271 Bollbuck, B., 271 Bonnal, C., 45 Bordas, J., 231, 239 Borisy, G., 268, 384, 385, 389 Bork, P., 405 Bornens, M., 282, 285 Bo¨ rsch, M., 350 Bosc, C., 275 Botchway, S. W., 179, 181, 182 Bo¨ ttcher, B., 355, 356, 357, 360, 361, 366, 368 Bottinelli, R., 187 Bottino, D., 395 Boudsocq, F., 412, 416, 422, 423

443

AUTHOR INDEX

Boujemaa, R., 391 Bourenkov, G. P., 129 Bourne, P. E., 301, 308 Bowman, B. J., 350, 351, 360 Bowman, E. J., 350, 351, 360 Boyer, P. D., 348, 349, 354, 362 Boyer, P. L., 413 Brady, S. T., 303, 304, 307, 338 Braithwaite, D. K., 403 Braman, J. C., 410, 414, 431 Brandmeier, B., 231 Brandt, P. W., 40 Brandt, R., 282 Brandt, U., 347 Bremel, R. D., 142 Brenner, B., 219, 229, 230, 231, 234 Brent, R. P., 56 Breukel, C., 288 Brick, P., 410, 430 Brisson, A., 361, 368 Broderick, M. J., 19, 29, 39 Brodsky, F. M., 352 Bronner, C. E., 307, 321, 332 Brown, D., 347 Brown, D. L., 275 Brown, F., 165, 187 Brown, J. H., 1, 123, 126, 127, 128, 129, 131, 132, 133, 139, 149, 162, 195, 196, 213, 231, 247 Brown, J. R., 4 Brown, P., 223, 231 Brown, W., 25, 30, 212, 230 Broyde, S., 425 Brozik, J. A., 429, 430 Bruford, E., 404 Brunner, D., 285 Bruskov, V. I., 402 Bryan, J. T., 127 Brzeska, H., 180 Bu, W., 288 Bubb, M. R., 370 Bucher, P., 405 Budny, M. J., 303, 307, 312 Buhle, E. L., Jr., 356 Bujalowski, W., 408 Bullard, B., 45, 91, 94, 112 Bullock, T. L., 385, 386, 387 Bulygin, V. V., 350, 362 Burgers, P. M., 404 Burgess, S. A., 12, 338

Burkhard, P., 127, 129 Burmeister, W. P., 307, 321 Burns, R., 231, 245 Burtis, K. C., 404 Buser, C. A., 307, 319 Bushweller, J. H., 278 Bustamante, C., 99, 101, 429, 430 Buster, D. W., 272, 274, 279 Butler, P. J., 349, 385 Butters, C. A., 146, 148 Buttery, S. M., 390 Byrne, B., 286

C Cabezon, E., 349 Cadwallader, G., 347, 360 Cain, B. D., 368, 370 Cain, S. M., 308, 325 Call, D. R., 97, 106, 107 Campbell, K. S., 106 Campmany, J., 231 Canas, B., 357 Cann, I. K., 404 Cantino, M. E., 63, 222, 230, 247 Capaldi, R. A., 352, 353, 355, 356, 362, 363, 365, 366, 368 Capson, T. L., 404, 406, 428, 431 Carbajo, R. J., 349 Carboni, J. M., 271 Carlier, M.-F., 269, 272, 282, 283, 384, 385, 390, 391 Carlsson, E., 66 Carragher, B. O., 308, 315, 325 Carrier, L., 64 Carrion-Vazquez, M., 99 Carroll, S. S., 417 Carson, M., 262, 272, 278 Carvalho, P., 285 Casazza, A. M., 271 Case, R. B., 307, 321, 334 Caspar, D. L., 141 Cassimeris, A., 283, 284 Cassimeris, L., 272, 273, 280, 282, 283 Caviston, T. L., 370 Cazorla, O., 110 Cecchi, G., 108, 231, 239, 240, 245 Centner, T., 93, 94, 95, 101, 103, 110, 112, 113

444

AUTHOR INDEX

Cha, S. S., 307, 321, 332 Chabala, J. C., 339 Chaban, Y. L., 361 Chacko, S., 128 Chaikin, P., 391 Chalfie, M., 259 Chamberlain, P. P., 412, 413 Chandler, J., 356, 368 Chandrasekaran, R., 27, 206 Charbaut, E., 268, 270, 279 Charrasse, S., 283 Charsky, C., 351, 357 Chelly, J., 277, 278 Chen, C., 356, 366 Chen, H., 262, 272, 278 Chen, J., 45, 273 Chen, L. F., 228 Chen, V., 217 Chen, W., 368 Cheney, R. E., 187 Cheng, N., 267, 268, 271 Cherney, M., 135, 140 Chevrier, V., 282 Chew, M. W., 63, 76, 78, 221, 222, 230, 231, 247 Chik, J. K., 35, 145, 146, 147 Chirino, A. J., 413 Chiu, W., 37, 59, 356, 366 Cho, Y. J., 131 Chong, P. C., 138, 140 Chopra, R., 413, 417, 423, 431 Chou, M., 259 Chre´tien, D., 259, 267, 270, 285 Christie, K. R., 300 Christman, M. F., 404 Cierpicki, T., 278 Clark, A. D., Jr., 409, 412, 413, 417 Clark, P., 409, 413, 417 Clark, R. J., 180, 187 Clayton, J., 36 Cleveland, D. W., 289, 300, 320 Clevers, H., 288 Cliord, P., 231 Cobb, B., 334 Cohen, C., 1, 19, 29, 39, 42, 52, 79, 123, 124, 126, 127, 128, 129, 131, 132, 133, 135, 137, 138, 139, 140, 141, 142, 143, 144, 145, 147, 148, 149, 150, 162, 165, 195, 196, 213, 231, 247, 330, 333 Cohen, D., 59

Cohen, P. W., 334, 335 Cohen-Addad, C., 307, 321 Colombini, B., 108, 231, 239, 240, 245 Colomo, F., 108, 186 Coluccio, L. M., 181, 187, 330 Compton, D. A., 289 Conibear, P. B., 179, 181, 183, 330 Conway, J. F., 128 Cooke, R., 9, 164, 303, 307, 308, 312, 321, 325 Cooper, J. A., 45, 286 Copeland, W. C., 404, 419 Cordell, S. C., 265 Corfield, V., 63, 64, 217 Cornell, L. A., 271 Cornett, B., 271 Correa, F., 149 Correa, J., 368 Corrie, J., 231 Coskun, U., 357, 361 Costa, J. V., 404 Coughlin, P., 284 Coulson, A., 259 Coupland, M. E., 186 Coureux, P. D., 165, 172, 173, 175, 177, 179, 187, 329, 330 Cowaert, M., 417 Cowan, N. J., 273 Cowell, J., 289 Cowie, R., 179, 181 Cox, G. B., 365, 366 Craig, R., 29, 42, 43, 47, 71, 74, 123, 143, 146, 147, 148, 213, 230 Craik, J. S., 231 Cremo, C. R., 187 Crick, F. H. C., 17, 123, 124, 402 Criddle, A. H., 179 Crider, B. P., 347 Crompton, A., 339 Croquette, V., 429, 430 Cross, R. A., 265, 280, 281, 308 Cross, R. L., 348, 350, 362 Crowther, R. A., 31, 42, 66, 71, 277 Cui, Q., 369 Cui, Y., 315 Culotti, J. G., 259 Culp, J. S., 412, 413, 417 Curmi, P. A., 262, 265, 268, 270, 279 Curmi, P. M. G., 39

AUTHOR INDEX

D da Silva, A. C., 140 Dahlberg, M. E., 404, 406, 428, 431 Dahlquist, F. W., 356, 365, 366, 368 Dai, S. C., 307, 321, 334 Dal Piaz, F., 45 Dammermann, A., 282 Dancker, P., 40 Dantzig, J. A., 186 Darby, G., 413 Das, A. K., 334, 335 Das, D., 412, 413 Das, K., 412, 413, 417 Date, T., 350, 352 Dauter, Z., 278 David-Pfeuty, T., 269 Davis, L. J., 279 Dawe, R. K., 300 Dawson, S. C., 300 Day, B. W., 271 Day, C. L., 126, 130, 288 De Bonis, S., 307, 321 De La Cruz, E. M., 180 De Lange, W., 63, 64, 217 de Lera, L. T., 404 de Pablo, P. J., 279 De Rosier, D. J., 38, 59, 142, 143, 210, 262, 272, 278, 286 de Wit, J. G., 265 Debouck, C., 412, 413, 417 Deckers-Hebestreit, G., 355, 356 Deka, J., 288 del Mazo, J., 404 del Monte, F., 106 Del Rizzo, P. A., 356, 368 Delacruz, J., 4 Delannoy, M., 356, 366 Delarue, M., 403 Deleage, G., 413 DeLucas, L. J., 262, 272, 278 den Blaauwen, T., 265 Dencher, N. A., 355 Denda, K., 350, 352 Denny, R. C., 31, 55, 56, 57, 61, 71, 78, 247 Denny, R. J., 217, 219, 231, 247 Derbyshire, V., 410 Derewenda, U., 278 Derewenda, Z. S., 278 Desai, A., 269, 280, 282

445

Devedjiev, Y., 278 Di Cera, E., 410 Di Guiseppi, S. R., 417 Diakun, G. P., 231 Diamantopoulos, G. S., 285, 286 Dianov, G. L., 405 Dianova, II, 405 Diat, O., 231, 238, 239, 240, 243 Diaz, F. G., 231 Diaz-Avalos, R., 305, 308 Dickens, J., 179, 181 Dickinson, M., 239 Didier, A., 288 Diez, M., 350 Dimroth, P., 355 Ding, J., 409, 412, 413, 417 Dionne, M. A., 289 Dittrich, P., 350, 351 Divita, G., 406, 409, 413 Djinovic-Carugo, K., 43 Dmitriev, O. Y., 355, 356, 366, 367, 368 Dobbie, I., 231, 238, 239, 240, 243 Doerhoefer, M., 305, 308 Dohmae, N., 290 Doi, H., 404 Domgill, I., 357, 360, 361, 368 Dominguez, O., 404 Dominguez, R., 35, 52, 79, 123, 126, 128, 129, 131, 132, 135, 140, 145, 146, 148, 149, 165 Donaldson, L., 365, 366 Dong, Q., 419 Doniach, S., 339 Donlin, M. J., 406, 407, 431 Doublie, S., 410, 414, 416, 418, 419, 431 Dow, J. A., 351 Downing, K. H., 259, 262, 263, 264, 265, 266, 267, 271, 272, 278, 286, 308, 311 Doyle, T. C., 229 Drabikowski, W., 135 Drechsel, D. N., 272, 284 Drendel, W., 135 Driessen, A. J., 265 Drory, O., 357, 359 Dschida, W. J., 360 Dubin, J. H., 64 Dubord, P., 133 Duncan, T. M., 350, 362 Dunlap, C. A., 428 Dunn, S. D., 353, 356, 365, 368

446

AUTHOR INDEX

Durrwang, U., 180, 187, 329 Dye, R. B., 272, 275

E Eakins, F., 59, 73, 195, 196, 197, 215, 216, 221, 234 Earley, F. G., 351 Earnshaw, W. C., 289, 290 Ebashi, F., 150 Ebashi, S., 29, 135, 150 Eccleston, J. F., 181, 182 Echols, H., 402 Eckley, D. M., 286 Eddy, S., 410 Edman, A. C., 27, 38, 42, 64, 74, 75, 76, 78, 217, 236, 237 Edwards, R. J., 215 Edwards, T. A., 412 Egelman, E. H., 38, 145, 148, 210 Eger, B. T., 406, 431 Eggers, C. T., 334 Ehler, E., 94 Eichinger, A., 410 Eisenhaber, F., 126, 129 Ekman, G. C., 390 El-Mezgueldi, M., 148 Ellenberger, T., 410, 412, 414, 418, 419, 423, 431 Elliott, G. F., 40, 230 Elston, T., 367 Endo, M., 29, 135 Endow, S. A., 7, 280, 300, 305, 307, 321, 323, 324, 325, 331, 332 Engel, A., 355 Engel, J., 127 Engel, J. D., 402 Engel, U., 286 Engelbrecht, S., 348, 350, 362, 363 Engh, R. A., 410 Eom, S. H., 410, 414 Eppenberger, H. M., 66 Erent, M., 180, 187 Erickson, H. P., 265, 269, 283, 284 Errico, A., 279 Escalante, C. R., 412, 423 Eschenburg, F. S., 172, 175 Eschenburg, S., 330, 331 Eshel, D., 336

Esnouf, R., 412, 413 Estes, J. E., 40

F Fackler, O. T., 352 Faiella, A., 93 Fairchild, C. R., 271 Faix, J., 127 Fajer, P. G., 330 Fan, J., 265, 266, 273, 276, 277, 287, 338 Farah, C. S., 131, 133, 139, 140, 149 Farid, R. S., 131, 132 Farkas, L., 127 Farman, G., 239 Farrell, C. M., 331 Faruqi, A. R., 231, 233, 239 Fath, T., 282 Faust, L., 165, 187 Fearnley, I. M., 360, 385 Feaver, W. J., 405 Fedriani, C., 285 Feinstein, S. C., 274 Felekyan, S., 350, 409 Felgner, H., 275 Feng, I. N., 64 Feng, Y., 278, 359 Feng, Z., 301, 308 Fenical, W., 271 Fenn, W. O., 187 Ferenczi, M. A., 223, 230, 231, 239, 245 Ferguson, C., 112 Ferris, A. L., 409, 413, 417 Ferro, J. A., 132 Fiala, K. A., 406, 422, 423, 427 Fievez, S., 269 Fillers, J. P., 123, 128, 144, 145, 147, 148 Fillingame, R. H., 348, 355, 356, 366, 367, 368, 370 Finbow, M. E., 347, 350, 360 Findlay, J. B., 360 Finer, J. T., 339 Fink, R. H., 64 Fink, S. P., 272, 275 Finley, J., 262, 272, 278 Finn, R., 59 Fischetti, R. F., 231, 245 Fisher, T. E., 99 Flashman, E., 63, 64, 217

447

AUTHOR INDEX

Flavigny, J., 64 Fletterick, R. J., 280, 301, 303, 305, 307, 308, 310, 312, 314, 315, 317, 318, 319, 321, 322, 324, 325, 327, 328, 331, 334 Flicker, P. F., 135, 137, 138, 143 Florence, G. J., 273, 276 Flynn, T. C., 369 Fodde, R., 288 Fojo, T., 271 Fontana, C. A., 385, 389, 390 Fore, N. S., 231 Forey, P. L., 31 Forgac, M., 347, 350, 351, 357, 358, 359, 360, 361, 365, 368 Forno, F., 93, 94, 95, 110, 112 Forscher, P., 388 Fortini, M. E., 273 Fortune, N. S., 186 Fotiadis, D., 355 Fowler, S. B., 99 Fowler, V. M., 39 Fox, L. A., 336 Francis, F., 277, 278 Francis, N., 38, 210 Frank, J., 59 Frank, R., 275, 282 Franklin, M. C., 410, 418, 419, 422 Franks-Skiba, K., 9 Franzini-Armstrong, C., 20, 228 Freiburg, A., 94, 95, 110, 112 French, G., 93, 101 Freundlich, A., 42 Frey, M. W., 404, 406, 426, 428, 431 Freyzon, Y., 52, 79, 163, 165 Friedberg, E. C., 404, 405 Friedman, J. M., 410, 413, 418 Fritsch, A., 339 Frolow, F., 357, 359 Frye, M., 239 Fujimori, K., 140 Fujita, H., 108 Fujita-Becker, S., 108, 180, 187, 329 Fujiwara, K., 165, 331 Fujiwara, S., 410 Fukuda, N., 109 Fukushige, T., 259 Fuller, G. W., 71 Fuller, S. D., 270, 285 Furch, M., 183, 231 Furst, D. O., 45, 69, 71, 90

Furukawa, T., 112 Futai, M., 347, 350, 356, 363, 365, 369

G Gaglio, T., 289 Gagne, S. M., 135 Galjart, N., 282 Galkin, V. E., 145, 148 Galletto, R., 408 Gambello, M. J., 277 Gamblin, S. J., 412, 413, 417 Garcia, M. A., 280, 283, 285 Garcia-Diaz, M., 404 Garcia-Ortiz, M. J., 404 Garcia-Saez, I., 320 Gard, D., 283, 284 Garman, E., 413 Garrett, S., 289 Gaspar, C., 288 Gautel, M., 43, 45, 64, 69, 71, 90, 91, 93, 94, 98, 99, 100, 101, 103, 104, 110, 112, 113 Gautel, N., 69 Gay, N., 9, 11 Gebhard, W., 36, 37, 38, 145, 149, 208, 212, 221 Geeves, M. A., 1, 2, 9, 12, 19, 23, 25, 38, 51, 77, 133, 147, 169, 179, 180, 181, 182, 183, 186, 187, 195, 197, 223, 225, 228, 229, 246, 247, 329 Geisterfer-Lowrance, A. A., 180 Georgiadis, M. M., 412, 413 Geourjon, C., 413 Gerbal, F., 391 Gergeley, J., 135, 140 Gergely, F., 285, 289 Gerisch, G., 127 Gerlach, V. L., 405 Gernert, K. M., 126, 129 Gershman, L. C., 40 Gerull, B., 108, 112, 113 Getlawi, F., 347 Geyer, M., 352 Giannakakou, P., 271 Gibbons, C., 353, 354, 365 Gibbons, I. R., 3, 336, 337 Gibson, L. C. D., 347, 360 Gigant, B., 262, 265, 268, 270, 279 Gilbert, S. P., 322, 331

448

AUTHOR INDEX

Giles, R. H., 288 Gill, S. R., 286 Gilliland, G., 301, 308 Gillin, F. D., 417 Girvin, M. E., 355, 366, 370 Gleeson, J. G., 277 Glotzer, M., 281 Glover, D. M., 282 Godovac-Zimmermann, J., 357 Goebl, M., 351 Goedert, M., 266, 273, 276, 277 Goel, A., 430 Goetz, M., 271 Go, S. P., 413 Gogarten, J. P., 350, 351 Gogol, E. P., 355, 356, 362 Gogonea, C. B., 259 Gold, L., 410 Goldie, K. N., 305, 308, 332 Goldman, Y. E., 186, 228, 231 Goldstein, L. S., 300, 320 Goldstein, M. A., 43 Goldstein, S. W., 413 Golitsina, N. L., 147 Gonzales, P., 339 Gonzalez, M. A., 404 Goodman, M. F., 402, 405, 412, 423 Goodson, H. V., 7, 286, 300 Goody, R. S., 406, 409, 413 Gordon, A. M., 33, 34 Gore, D., 239 Gotte, M., 406 Goud, B., 285 Goulding, D., 45, 91, 94, 112 Gowen, B., 59 Grabar, T. B., 368 Grabarek, Z., 135, 140 Gr€aber, P., 350, 355, 356 Graboski, S., 131, 132, 133, 139 Gracea, P., 35, 133, 145, 146, 149 Gr€af, R., 288, 351 Graham, L. A., 347, 360 Gramlich, M., 112, 113 Granzier, H. L., 19, 29, 45, 49, 63, 91, 98, 99, 101, 109, 195, 196, 221, 247 Greaser, M., 93, 101, 102, 104, 135, 187 Greaser, M. L., 91, 97, 108 Greenfield, N. J., 123, 126, 128, 129, 131, 132, 133, 139, 149 Gregorio, C. C., 94, 112, 113

Greider, C. W., 406 Greie, J. C., 355, 367, 368 GriYths, A., 265 GriYths, P. J., 231, 239, 240, 245 Grindley, N. D., 418, 424 Grollman, A. P., 412 Gross, H., 305, 308 Gross, S. P., 279 Groth, G., 352 Grove, B. K., 66 Groves, M. R., 284 Gru¨ ber, G., 347, 352, 357, 361 Gruen, M., 64 Gruneberg, U., 281 Gu, J., 229, 230 Guajardo, R., 429, 430 Guerrero, J., 334 Guez, N., 303, 310, 311, 312, 337 Guidotti, G., 352 Guilford, W. H., 163 Gulick, A. M., 9, 307, 323 Gumbiowski, K., 363 Gunasekera, S. P., 271 Guo, J., 307, 317, 318, 319 Guss, J. M., 365, 366 Gussio, R., 271

H Haaf, A., 385 Habel-Rodriguez, D., 429, 430 Habermann, B., 283 Hackman, P., 113 Hackney, D. D., 334 Haggarty, S. J., 339 Halczenko, W., 307, 319 Hall, C., 231 Hall, C. J., 231 Halpain, S., 274, 277 Halvorson, H. R., 186 Hamel, E., 271 Hamelin, M., 259 Hamill, D. R., 280 Hamlin, R., 410, 430 Hammell, R. L., 138 Hanaoka, F., 404, 422 Hankamer, B., 360, 361 Hanlon, N., 284 Hanson, J., 2, 29, 31, 36, 89, 142

AUTHOR INDEX

Harada, A., 275 Harauz, G., 59 Haravuori, H., 113 Harbury, P. B., 126 Harford, J. J., 27, 31, 38, 42, 55, 56, 57, 61, 64, 71, 73, 74, 75, 76, 78, 195, 196, 197, 206, 212, 215, 216, 217, 219, 220, 221, 223, 225, 226, 227, 229, 230, 231, 234, 236, 237, 247 Harries, J. E., 231 Harrington, W. F., 226 Harris, A. K., 391 Harris, B. A., 412, 413, 417 Harris, T., 360 Harrison, M. A., 350 Harrison, S. C., 412, 413, 417, 423, 431 Hart, C. L., 304, 307, 308, 321, 325, 334 Hartman, G. D., 307, 319 Hartmann, H., 45 Hartshorne, D. J., 101 Harvey, E. V., 165 Harvey, J., 425 Harvey, W. R., 347, 350, 351, 352, 357 Hasbani, M. J., 286 Hasegawa, M., 273 Haselgrove, J. C., 39, 64, 213, 223, 234, 235 Hashimoto, H., 410 Hasler, K., 350 Hasselbach, W., 40 Hatch, V., 146, 148 Hattori, A., 108 Haus, U., 45 Hauser, M. T., 290 Hausrath, A. C., 352 Hayashi, H., 140 Hayashi, I., 278 Hayashi, S., 367, 371 Hayess, K., 69 Hays, T. S., 9 He, Z.-H., 231 Head, J. G., 147 Heald, R. W., 133, 289 Heeley, D. H., 133 Heidary, G., 273 Heinrich, M., 290 Heitkamp, T., 355, 367, 368 Hekmat-Nejad, M., 280, 305, 307, 324, 327 Helfman, D. M., 139 Hell, W., 95 Heller, M., 146, 148

449

Hellmig, B., 412, 413, 417 Helmes, M., 103, 110, 112 Helsby, W. I., 230, 231, 245 Hemelin, M., 259 Hemmings, B. A., 284 Henderson, M., 286 Henderson, R., 356, 366 Hendrix, J., 233 Hensens, O., 271 Hermolin, J., 370 Herold, K. A., 367 Herschbach, D., 430 Herzberg, O., 37, 38, 134, 135, 140 Herzog, W., 108 Hess, L., 40 Heuser, J. A., 391 Heuser, J. E., 286 Higgins, C., 277 Higuchi, H., 101 Hilario, E., 351 Hill, F., 273 Hillen, M., 231 Himmel, D., 52, 165 Himmel, M., 69 Hinkle, D. C., 404 Hirabayashi, T., 235 Hirata, T., 350, 363, 369 Hirokawa, N., 3, 259, 273, 275, 280, 300, 305, 307, 308, 314, 315, 316, 317, 326, 327, 331, 334 Hirose, K., 265, 280, 281, 308 Hisabori, T., 348 Hitchcock-DeGregorio, S. E., 123, 126, 128, 129, 130, 131, 132, 133, 138, 139, 149 Hodges, R. S., 38, 126, 127, 128, 129, 138, 140, 144, 333 Hoenger, A., 305, 307, 308, 310, 311, 313, 332 Homan, H. M., 112 Hofmann, K., 405 Hogg, M., 410, 416 Holden, H. M., 23, 51, 52, 78, 101, 145, 165, 166, 167, 223 Holleran, E. A., 286 Holmes, D., 49, 63, 110 Holmes, K. C., 1, 2, 9, 12, 19, 23, 25, 31, 33, 34, 35, 36, 37, 38, 40, 51, 52, 77, 78, 127, 145, 149, 166, 169, 171, 172, 173, 174, 195, 197, 206, 208, 212, 221, 223, 225, 228, 229, 230, 246, 247, 329, 330

450

AUTHOR INDEX

Holt, J., 110 Holtauzen, L. M., 149 Holzbaur, E. L., 3, 286 Holzenburg, A., 360 Hom-Booher, N., 304 Homes, K. C., 169, 181 Homma, N., 280 Homnick, C., 307, 319 Homsher, E., 146, 148, 186, 187 Hondal, R. J., 406, 417, 427 Hood, L., 418 Hopfner, K. P., 410 Hopkins, A. L., 413 Hoppert, M., 355 Horiuti, K., 230 Hornemann, T., 69 Horowits, R., 39, 91 Hoshijima, M., 112 Houdusse, A., 52, 79, 135, 140, 165, 169, 171, 172, 173, 174, 175, 177, 179, 187, 223, 277, 278, 329, 330, 333 Houmeida, A., 110 Howard, J., 300 Howard, L., 289 Howell, B., 280 Howitt, S. M., 368 Hsieh, J. C., 406 Hsiou, Y., 412, 413, 417 Hu, D. H., 94 Huang, H., 413, 417, 423, 431 Huang, T. G., 334 Huang, Y., 131, 132, 139 Huber, H. E., 307, 319, 403 Huber, R., 410 Hudson, L., 31, 40, 52, 55, 56, 57, 61, 71, 72, 73, 78, 79, 217, 219, 231, 237, 247 Hughes, S. H., 412, 413, 417 Hullihen, J., 352, 355 Hulmes, J. D., 351 Hunter, A. W., 280 Huse, D. A., 332 Huss, M., 347, 357 Hussey, P. J., 290 Hutcheon, M. L., 350, 362 Hutton, R. L., 186 Huxley, A. F., 2, 19, 25, 30, 31, 33, 34, 39, 89, 90 Huxley, H. E., 2, 59, 87, 89, 124, 142, 143, 197, 212, 213, 223, 230, 231, 233, 234, 235, 238, 239, 240, 245 Hyman, A. A., 267, 272, 283, 284

I Igarashi, K., 351 Iino, R., 366 Ikeda, C., 360, 361 Ikura, M., 135, 278 Ilankumaran, P., 413 Imamura, H., 352, 360, 361, 363 Imanaka, T., 410 Immendorfer, U., 413 Improta, S., 101 Incla´ n, Y. F., 282 Infante, C., 285 Inoue, A., 165 Inoue, T., 410 Inoue, Y. H., 282 Inue, T., 357, 360, 361, 368 Irving, M., 231, 238, 239, 240, 243, 245 Irving, T. C., 40, 52, 71, 72, 73, 78, 79, 195, 196, 197, 215, 216, 221, 223, 228, 231, 234, 237, 238, 239, 240, 245, 247 Ishihara, A., 391 Ishijima, A., 133 Ishino, Y., 404 Islam, K., 275 Italiano, J. E., 390 Italiano, J. E., Jr., 385, 389, 390, 391 Ito, J., 403 Ito, M., 101 Itoh, H., 369 Itoh, Y., 90 Ivaninskii, S., 127 Iwai, S., 422 Iwamatsu, A., 391 Iwamoto-Kihara, A., 350, 363, 365 Iwata, M., 360, 361 Iwata, S., 360, 361 Izumimoto, M., 43

J Jacobo-Molina, A., 409, 413, 417 Jacobson, K., 391 Jaeger, J., 413 J€ager, D., 351 J€ager, H., 356 Jager, J., 413 Jahn, W., 145, 223, 330 Jakes, R., 273, 277

AUTHOR INDEX

James, M. N., 37, 38, 134, 135, 140 Jancso, A., 133 Jang, S. I., 127 Janssen, P. A., 413 Jeers, K., 285 Jeries, T. E., 179, 180 Jeng, C. J., 90 Jensen, P. R., 271 Jerina, D. M., 412, 413, 422, 423 Jeruzalmi, D., 410 Jerva, L. F., 406, 417, 426, 431 Jessen, S., 413 Jezewska, M. J., 408 Jha, P. K., 135 Jiang, W., 356, 367, 368, 370 Job, D., 259, 275, 282 Johannes, L., 285 John, W., 171 Johnson, D., 262, 272, 278 Johnson, K. A., 406, 407, 417, 426, 428, 431 Johnson, P., 123, 124 Johnson, R. E., 405, 412, 423 Jones, A., 231 Jones, A. O., 231 Jones, K. L., 66, 67, 69 Jones, P. C., 356, 367, 368 Jones, R. W. M., 231 Jones, Y., 412, 413 Jontes, J. D., 187, 330 Jordan, M. C., 106 Joseph, C., 45 Jourdain, I., 262, 265, 268, 270, 279 Joyce, C. M., 403, 418, 424 Juanhuix, J., 231 Julian, F. J., 33, 34 Jun, G., 123, 126, 128, 129, 131, 132, 149 Junge, W., 348, 350, 362, 363 Jurasek, L., 38, 140, 144

K Kaboord, B. F., 404, 406, 428, 431 Kabsch, W., 34, 35, 36, 37, 38, 145, 149, 208, 212, 221, 308 Kagawa, Y., 352, 355 Kaharevitz, D., 271 Kahn, W., 36 Kai Y., 410 Kakinuma, S., 351

Kakinuma, Y., 351, 360 Kalabokis, V. N., 52, 165 Kambara, M., 352, 355 Kamer, G., 409, 413, 417 Kamiya, N., 386 Kammerer, R. A., 127, 129 Kan, L., 102 Kanai, S., 404 Kanai, Y., 273 Kanazawa, H., 350 Kane, P. M., 347, 351, 352, 357, 359 Kang, S. J., 7, 280 Kanoh, N., 265, 271 Kapoor, T. M., 289, 339 Kar, S., 266, 273, 276, 277 Karam, G., 404, 406 Karam, J. D., 410, 416 Karawya, E. M., 405, 409 Karet, F. E., 347 Karki, S., 3, 286 Karlsson, C., 289 Karplus, M., 369, 430 Karrasch, S., 355, 356 Karsenti, E., 270, 283, 285, 289, 290 Kasparian, J., 262 Kass, S., 180 Katayama, E., 336 Kati, W. M., 406, 417, 426, 431 Kato-Yamada, Y., 363 Katoh, K., 331 Katsuyama, A. M., 131, 133 Kaupmann, K., 93 Kawai, M., 186 Kawamoto, M., 386 Kawasaki-Nishi, S., 351 Kay, C. M., 131 Keegstra, W., 361 Keeling, D. J., 347 Keeling, J., 413 Keenan, K. L., 351 Keep, N. H., 43 Keilman, M., 288 Keller, D., 429, 430 Keller, R. W., 429, 430 Kellermayer, M. S., 98, 99, 100, 101 Kelman, Z., 404 Kemp, 113 Kempa, S., 69 Kempner, E. S., 91 Kendrew, J. C., 17

451

452

AUTHOR INDEX

Kendrick-Jones, J., 43, 187, 231 Kensch, O., 409, 413 Kensler, R. W., 42, 58, 59, 60, 61, 63, 67, 71, 74, 78, 87 Kent, H. M., 385, 386, 387 Kerscher, S., 347 Kibak, H., 350, 351 Kiefer, J. R., 410, 414, 431 Kikkawa, M., 280, 305, 307, 308, 314, 315, 316, 317, 331 Kilmartin, J., 289 Kim, E., 145, 148 Kim, K., 405 Kim, K.-H., 123, 126, 128, 129, 131, 132, 149 Kim, M. H., 278 Kim, P. S., 125, 126, 305 Kim, S. J., 425 Kim, Y., 410 Kimura, E., 113 Kimura, S., 90, 94, 97, 112, 113, 369 Kinbara, K., 97, 112, 113 King, K. L., 385 King, R. W., 339 King, S. M., 286, 287, 337 King, W. A., 39 Kinoshita, K., 283 Kinosita, K., Jr., 350, 363, 365, 369 Kirby, I., 413 Kirchho, T., 404 Kitahara, T., 265, 271 Kiyotoshi, M., 63, 238 Klein, U., 351 Kleine-Kohlbrecher, D., 357 Klemm, J. D., 125 Klenow, H., 409 Kliche, W., 329 Kline-Smith, S. L., 282 Klopfenstein, D. R., 315 Klug, A., 261 Klumpp, L. M., 331 Knight, P. J., 12, 49, 63, 110, 226, 338 Knoll, R., 112 Knossow, M., 262, 265, 268, 270, 279 Knupp, C., 1, 2, 4, 19, 25, 27, 49, 51, 52, 64, 65, 73, 77, 110, 162, 195, 196, 197, 205, 208, 214, 215, 216, 217, 218, 219, 221, 225, 226, 227, 229, 231, 232, 233, 234, 237, 238, 241, 242, 243, 244, 246, 247 Knutson, J. R., 425

Ko, Y., 356, 366 Kobayashi, T., 140 Koch, M., 231 Koch, M. H., 231, 239, 305 Ko¨ hler, D., 163 Kohlstaedt, L. A., 413 Kojima, H., 133, 336 Kojima, S., 331 Kollmar, M., 329 Kolmerer, B., 45, 90, 93, 94, 97, 102, 108, 112, 113, 222 Komori, K., 404 Kondoh, M., 265, 271 Konig, M., 350 Konigsberg, W. H., 404, 406, 410, 416, 418, 419 Konings, W. N., 351, 361, 368 Konishi, J., 350, 352 Konishi, K., 351 Konrad, M., 133 Koonin, E. V., 7, 336, 404, 405 Korn, E. D., 180 Kornberg, A., 403, 404, 409, 417, 430 Korolev, S., 410, 414, 418, 419, 420, 421, 430, 431 Kostin, S., 45 Kostyukova, A., 39 Koubassova, N., 225, 230, 231, 238, 239, 240, 243, 245 Koupal, L., 271 Kovacs, M., 165, 179, 181, 182, 330 Kowalski, R. J., 271 Koymans, L., 413 Kozielski, F., 305, 307, 308, 320, 321 Kraft, T., 231 Krahn, J. M., 271, 412 Kramer, H., 288 Kraulis, P. J., 263, 265 Kraut, J., 412, 416, 420, 423, 424, 425, 431 Kraynov, V. S., 406, 417, 427 Krebs, A., 305, 307, 310, 311, 313 Krebs, R., 406, 409 Kreis, T. E., 285, 286 Krementsova, E., 163 Kress, K. R., 64 Kress, M., 231, 233, 239 Krieger, I., 39 Krohn, N., 305 Krowarsch, D., 278 Krueger, J. K., 101

AUTHOR INDEX

Krylova, I., 39 Krzesinski, P. R., 97, 106 Kuchta, R. D., 406, 417, 426, 428, 431 Kudryavtsev, V., 350 Kuhlmann, J., 288 Kuipers, J., 288 Kulikovskaya, I., 64 Kulke, M., 106, 108 Kull, F. J., 7, 36, 145, 171, 172, 175, 223, 301, 307, 308, 310, 321, 328, 329, 330, 331 Kumar, A., 405, 409, 412, 416, 423, 424, 425, 431 Kunkel, T. A., 402 Kunst, G., 64 Kuo, L. C., 307, 319 Kurasawa, Y., 290 Kuriyan, J., 410 Kurzawa, S. E., 180, 181, 329 Kurzawa-Goertz, S. E., 187 Kuznicki, J., 135 Kwiatek, O., 45 Kwok, S. C., 126, 128, 129

L Labean, T. H., 126, 129 Labeit, D., 95, 99, 100, 102, 103, 104, 108, 112 Labeit, S., 19, 29, 45, 49, 63, 90, 91, 93, 101, 102, 104, 108, 110, 112, 113, 195, 196, 221, 222, 247 Lachkar, S., 262, 265, 268, 270, 279 Lacktis, J., 186 Ladjadj, M., 59 Lahmers, S., 97, 106, 107 Lain de Lera, T., 404 Laing, N. G., 49 Lakey, A., 110, 113 Lanar, D. E., 42 Lander, E. S., 93 Landolt-Marticorena, C., 368 Landwehr, R., 127 Langford, G. M., 4 Langkopf, A., 275 Langley, B., 112 Langsetmo, K., 140 Larroque, C., 283 Lasken, R., 410 Laskowski, R. A., 308

453

Lataro, R. C., 132 Lattanzi, G., 339 Lau, R., 301, 307, 308, 310 Laue, F., 410 Laurent, V., 391 Lavigne, P., 127, 128 Lawrence, C. J., 300 Lazarides, E., 271 Le Clainche, C., 384, 385, 390 Le Claire, L. L. III, 390 Leake, M. C., 100, 103, 104, 106 Lebart, M. C., 45 Lebeau, L., 268, 270, 279 Lee, D. S., 410 Lee, H., 286, 406 Lee, J. H., 127, 391 Lee, M. J., 285 Lee, Y., 339 Lee-Eiford, A., 336 Lees-Miller, J. P., 139 Leet, N. G., 90 Leguy, R., 282, 283 Lehman, I. R., 409, 430 Lehman, W., 123, 143, 146, 147, 148, 213 Lehrer, S. S., 133, 147, 149 Leibler, S., 332 Leighton, L., 410 Leipe, D. D., 7 Leith, A., 59 Lemker, T., 352 Leng, X. H., 360 Leonard, M., 391 Leonard, T. R., 108 Leslie, A. G., 349, 352, 353, 354, 355, 356, 359, 365, 366, 369, 370 Levine, R. J., 42, 71, 74 Lew, J., 274 Lewis, E., 339 Lewis, M. K., 217 Lewis, R. A., 231 Li, B., 140 Li, H., 102, 103, 259, 262, 263, 264, 271, 272, 278, 286, 308, 311 Li, J., 418, 419 Li, M. X., 135, 137, 140 Li, M.-G., 9 Li, S., 262, 272, 278 Li, X., 334, 347, 406 Li, Y., 59, 127, 131, 132, 133, 135, 139, 414, 418, 419, 420, 421, 428, 430, 431

454

AUTHOR INDEX

Li, Z., 63, 64, 140, 217, 357, 360 Liesch, J., 271 Lill, H., 348 Lin, A. W., 308, 325 Lin, C. M., 271 Lin, G., 262, 272, 278 Lin, T. C., 404, 406, 418, 419 Linari, M., 231, 238, 239, 240, 243, 245 Lindberg, U., 35, 145, 146, 147 Lindel, T., 271 Lindhout, D., 39 Ling, H., 412, 416, 422, 423 Linke, W. A., 102, 103, 112 Linnemann, T., 352 Linton, L. M., 93 Littlefield, R., 39 Liu, J., 228 Liu, J. Y., 290 Liu, Q., 359 Liu, S. H., 352 Liu, W., 135 Liu, X., 288 Liu, Z. J., 262, 272, 278 Lively, C., 404, 406, 428, 431 Liversage, A. D., 49, 63, 110 Llanos, R., 282 Lloyd, S. A., 11 Locker, R. H., 90 Lockhart, A., 265 Lockhart, D. A., 265 Loeb, L. A., 402, 418 Loisel, T. P., 391 Lolkema, J. S., 361, 368 Lombardi, V., 231, 238, 239, 240, 243, 245 Long, A. M., 410, 414, 419, 431 Long, B. H., 271 Long, J. C., 356 Longley, R. E., 271 Lopez-Fernandez, L. A., 404 Lorenz, M., 36, 40, 51, 52, 78, 145, 166, 208, 223 Love, M. L., 135, 140 Lovell, S. J., 226 Lovering, R. C., 347 Low, I., 40 Lo¨ we, J., 259, 262, 263, 264, 265, 266, 267, 272, 280, 281, 308, 311 Lowey, S., 162 Lowy, J., 29, 36, 75, 142, 230, 231 Lu, C., 279

Lu, H., 429, 430 Lu, R. C., 135 Lu, S. M., 126, 128, 129 Lu, X., 352, 409, 413, 417 Luan, C. H., 262, 272, 278 Lucas, R. M., 141 Lucaveche, C., 42, 228 Lucii, L., 231, 245 Lu¨ cken, U., 355, 356 Ludin, B., 275 Ludtke, S. J., 59, 356, 366 Ludwig, J., 347 Lundberg, L., 347 Lungl, A., 361 Luo, M., 262, 272, 278 Luo, Y., 140 Lupas, A., 126 Lustig, A., 127 Luther, P. K., 1, 2, 4, 27, 30, 31, 43, 45, 47, 48, 49, 51, 63, 64, 65, 66, 67, 68, 69, 94, 110, 162, 208, 217, 218, 219, 220, 222, 223, 230, 231, 247 Lutkenhaus, J., 265 Lutter, R., 349, 352, 359, 365 Lu¨ ttge, U., 357, 360, 361, 368 Ly, B., 304 Lymn, R. W., 33, 162, 163, 164, 221, 223, 229, 230

M Ma, J., 369 Ma, K., 93, 102 Ma, Y. M., 347, 357 Mac Arthur, M. W., 308 Mac Leod, K. J., 359 Mackey, A. T., 322, 331 Macosko, J., 429, 430 Madrid, M., 413 Maeda, K., 39, 45, 108, 124, 133, 135, 137, 140, 141, 150, 165 Maeda, M., 356 Maeda, Y., 36, 39, 40, 45, 124, 133, 135, 137, 140, 141, 149, 150, 165, 231 Maggs, A. C., 272 Maier, D., 429, 430 Mains, P. E., 279 Maio, L., 384, 393 Maitra, M., 424

AUTHOR INDEX

Mak, A. S., 133, 138 Mak, J., 339 Makarenko, I., 106 Makino, K., 36, 40, 149 Makrides, V., 274 Makyio, H., 360, 361 Malhotra, R., 352 Malmberg, R. L., 300 Malnasi-Csizmadia, A., 165, 179, 181, 182, 330 Mandel, M., 351 Mandelkow, E., 1, 2, 3, 4, 7, 9, 12, 273, 274, 275, 303, 304, 305, 307, 308, 310, 311, 313, 332, 333, 334, 335 Mandic, R., 352 Mannherz, H. G., 34, 35, 145 Manolson, M. F., 350, 368 Manstein, D. J., 165, 172, 175, 180, 183, 187, 329, 330, 331 Mant, C. T., 127, 128 Mant, G. R., 231 Mao, C., 410, 414, 431 Marekov, L. N., 127 Margossian, S. S., 162 Maritan, A., 339 Markley, J. L., 355, 356, 366 Marko, J. F., 99 Marques, M. I., 404 Marsh, D., 360 Marston, S., 148 Marszalek, P. E., 99 Martin-Barbey, C., 268, 270, 279 Martin-Fernandez, M. L., 231 Martinez, A. C., 404 Martinez, H. M. M. M. F., 331 Martinez, P., 282 Martins, A., 404 Martone, M. E., 45 Maruyama, K., 90, 94, 101 Marx, A., 1, 2, 3, 4, 9, 12, 303, 304, 305, 307, 310, 311, 313 Mashanov, G. I., 223, 231 Mason, C. M., 64 Massie, M. R., 274 Matadeen, R., 59 Mathe, E., 282 Matsumoto, Y., 405 Matsumura, M., 139 Matsuura, Y., 386 Mattei, T., 231 Matthews, B. W., 352, 357, 359

455

Matthey, U., 355 Mattsson, J. P., 347 Matus, A., 275 Maughan, D., 239 Mayans, O., 69, 113 Mayer, F., 355 Mayer, T. U., 339 Maytum, R., 133 Mayumi, T., 271 McCarter, R., 94 McClellan, G., 64 McClure, J., 90 McComas, A. J., 22 McCoy, A. J., 385, 386, 387 McEnery, M. W., 356 McGough, A., 37 McIntosh, J. R., 300 McIntosh, L., 365, 366 McKenna, W., 180 McKillop, D. F., 147 McLachlan, A. D., 38, 122, 123, 125, 126, 128, 131, 133, 144, 149 McLachlin, D. T., 353, 365, 368 McNabb, M., 99, 100, 108, 109 McNally, F. J., 279 McNally, K. P., 279 McQueney, P. A., 271 Mead, R., 56 Mehta, A. D., 187 Meier, M., 127 Meier, T., 355 Melki, R., 269, 282, 283 Melkonian, K. A., 286 Mellwig, C., 355, 356, 366 Melvin, G., 229, 230 Mendelson, R. A., 39 Menetrey, J., 165, 172, 175, 187, 329, 330 Menz, R. I., 369 Merdes, A., 289 Merzendorfer, H., 347, 357 Mesyanzhinova, I. V., 356 Metropolis, N., 56 Meurer-Grob, P., 262, 282 Meyer, B., 333 Meyering-Vos, M., 361 Meyho¨ fer, E., 163 Michael, L. H., 43 Miki, H., 300 Milch, J. R., 231 Milla, M., 305

456

AUTHOR INDEX

Millar, D. P., 426 Millar, N. C., 180, 181, 186 Miller, H., 412 Miller, M. K., 112, 113 Millevoi, S., 45 Milligan, R. A., 51, 52, 78, 166, 187, 223, 262, 265, 266, 272, 274, 277, 278, 280, 305, 307, 308, 315, 324, 325, 327, 330 Milligan, V. N., 187 Millman, B. M., 223, 230 Minajeva, A., 112 Mische, S., 9 Mishima, M., 281 Mitchell, P., 348 Mitchison, T. J., 265, 269, 272, 280, 282, 284, 300, 339, 384, 391 Mitome, N., 367, 370, 371 Miura, P., 286 Miyamoto, C. A., 140 Miyamoto, D. T., 282 Miyamoto, S., 139 Mizrahi, V., 406, 417, 426, 428, 431 Moarefi, I., 410 Mochizuki, Y., 290 Mocz, G., 336, 337 Modak, M. J., 418 Modrich, P., 406 Moereels, H., 413 Mogilner, A., 384, 385, 391, 395 Mok, N. S., 73, 195, 196, 197, 215, 216, 221, 231, 232, 233, 234, 238 Molloy, J. E., 187, 330 Monasterio, O., 267 Monteiro, P. B., 132 Montelione, G. T., 131, 132, 139 Montemagno, C. D., 339 Montgomery, M. G., 353, 354, 365 Moody, R., 339 Moolman-Smook, J., 63, 64, 217 Moore, P. B., 59, 142, 143 Moores, C. A., 43, 277, 278, 280, 305, 307, 324, 327 Moos, C., 29, 64 Mooseker, M. S., 180, 187 Moras, D., 403 Morasso, M. I., 127 Moritz, M., 282, 283 Moriyama, Y., 350, 351, 356 Morris, C. A., 163, 165, 172, 175, 187, 329, 330

Morris, E. P., 37, 40, 45, 58, 59, 60, 61, 63, 67, 71, 72, 74, 78, 87, 147, 213, 222, 230 Morris, M. B., 368 Moss, D. S., 308 Moss, R.-L., 187 Moult, J., 140 Mues, A., 45, 69 Muharemagic, A., 355, 368 Muhle-Goll, C., 99, 100 Mui, S., 131, 132, 133, 139 Muirhead, H., 17 Mukherjee, A., 265 Mu¨ ller, D. J., 355 Muller, J., 1, 2, 3, 4, 9, 12, 303, 304, 305, 307, 308, 310, 311, 313, 332 Mu¨ ller, O., 288 Mu¨ ller, R., 275 Mu¨ ller, S., 290 Mu¨ ller, V., 352, 361 Muller-Seitz, M., 93 Mullins, R. D., 391 Muneyuki, E., 348, 352, 363, 369 Munro, P. M., 27, 69, 208, 231 Murakami, T., 366 Murata, T., 360 Murgich, J., 42, 74 Murphy, C. T., 187 Murray, J. M., 142, 144, 145, 147 Mutungi, G., 186 Myers, M., 351

N Naber, N., 303, 307, 308, 312, 325 Nadal-Ginard, B., 123 Nadezhdina, E., 268 Nagano, K., 139 Nagata, K., 360, 361 Nagueh, S. F., 106, 107 Nagy, L. A., 262, 272, 278 Nahirney, P. C., 217 Nair, D. T., 412 Nair, P., 99, 100, 102, 103, 104 Nakaie, C. R., 139 Nakamura, N., 350, 369 Nakano, M., 352, 363 Nakata, T., 280 Nakauchi, Y., 101 Nakayama, H., 265, 271

457

AUTHOR INDEX

Namba, K., 1, 4, 36, 40, 149 Nanni, R. G., 409, 413, 417 Narayan, T., 231, 238, 239, 240 Narayanan, T., 39, 230, 231, 238, 240, 243, 245 Nave, R., 90 Nayal, M., 410 Neagoe, C., 106 Ne, N., 351 Nelder, J. A., 56 Nelson, H., 351 Nelson, N., 347, 350, 351, 357, 359 Nencini, S., 186 Nettles, J. H., 271 Neuhaus, D., 385 Neuwald, A. F., 336, 405 Newman, P. R., 359 Nicholas, G., 112 Nichols, C. E., 413 Nicholson, W., 262, 272, 278, 286 Nicolaou, K. C., 271 Niedergerke, R., 2, 31, 89 Niederstrasser, H., 267, 268 Nigg, E. A., 281 Nihli, M., 146, 148 Nili, M., 187 Nishi, T., 350, 351, 360 Nishikawa, K., 265, 271 Nishikiori, T., 271 Nishio, K., 350, 363, 365 Nishioka, M., 410 Nishizaka, T., 369 Nitanai, Y., 39 Nitta, R., 280, 305, 307, 316, 317, 326, 327, 331, 334 Nneji, G., 40, 45 Nocula, C., 231 Noda, Y., 3, 280 Noegel, A. A., 45 Nogales, E., 259, 262, 263, 264, 265, 266, 267, 268, 271, 272, 278, 282, 286, 308, 311 Noji, H., 350, 352, 363, 365, 369 Norwood, F. L., 43 Nossal, N. G., 417 Notario, V., 404 Nurse, P., 285 Nyitrai, M., 187 Nyitray, L., 127 Nyland, L., 106

O O’Donnell, M., 404 O’Shea, E. K., 125 Oakley, B., 282 Oberhauser, A. F., 99, 102, 103 Obermann, W. M., 69, 71, 113 Oda, T., 36, 40, 133, 149 Odde, D. J., 279 Oegema, K., 282 Oenarier, E., 275 Oesterhelt, F., 98, 99 Oer, G., 29, 217, 229, 230, 235 Ogawa, T., 280, 307, 326, 327, 331, 334 Ogilvie, I., 353 Ohashi, K., 90 Ohkuma, S., 352, 360, 361, 363 Ohkura, H., 280, 283, 285 Ohtsuki, I., 138, 139, 140 Oiwa, K., 12, 336, 338, 369 Oka, T., 350 Okada, Y., 3, 280, 300, 305, 307, 308, 314, 315, 316, 317, 326, 327, 331, 334 Okajima, T., 363 Oldenburg, M., 366 Oliva, M. A., 265, 267 Oliveira, D. M., 139 Oliver, T., 391 Olivieri, N., 97, 101, 113 Ollis, D. L., 410, 430 Omote, H., 350, 369 Onishi, H., 165, 331 Ono, A., 138, 140 Ono, S., 366 Ono, Y., 112 Opitz, C. A., 106 Orlova, A., 145, 148 Orlova, E. V., 59, 356 Osada, H., 265, 271 Osborn, M., 90 Oshima, K., 63, 238 Oshima, T., 350, 351, 352 Ostap, E. M., 180, 182 Oster, G., 367, 370, 371, 384, 385, 391, 395 Otlewski, J., 278 Otsuki, I., 29, 135 Ott, A., 391 Otterbein, L. R., 35, 145, 146 Overgaard-Hansen, K., 409

458

AUTHOR INDEX

Ow, R. A., 336 Ozer, R. S., 274, 277

P Padron, R., 42, 43, 47, 71, 74, 230 Pai, E. F., 34, 35, 145 Pa´ li, T., 360 Palm, T., 131, 132, 133, 139 Palmiter, K. A., 163 Pan, Y. C., 351 Panine, P., 230, 231, 245 P€anke, O., 363 Pantaloni, D., 269, 272, 282, 283, 384, 385, 390, 391 Papa, I., 45 Pape, T., 59 Park, C. G., 305, 307, 321, 323, 324, 325, 331, 332 Park, H. W., 305, 307, 321, 323, 324, 325, 331, 332, 410 Parker, B., 231 Parma, D., 410 Parniak, M. A., 423 Parra, K. J., 351 Parraga, M., 404 Parry, D. A. D., 23, 38, 39, 42, 123, 124, 125, 127, 128, 135, 141, 142, 144, 148, 149, 163, 208, 213, 234 Pask, H. T., 66, 67, 69 Pastore, A., 45 Pata, J. D., 405, 406, 412, 413 Pate, E., 9, 308, 325 Patel, S. S., 406, 407, 417, 424, 426, 427, 428, 431 Paterson, I., 273, 276 Pato, M. D., 165 Patwardhan, A., 59 Paul, D., 59 Paulucci, A. A., 131, 133 Pavicic, V., 281 Pavlov, A. R., 410 Pavlov, D., 187 Peachey, L. D., 90 Peak-Chew, S. Y., 285 Pearlstone, J. R., 135, 138 Pearson, D. S., 179, 181, 182 Pechatnikova, E., 308, 325 Pedersen, P. L., 352, 355, 356, 366

Pedrotti, B., 275 Peitsch, M. C., 303, 310, 311, 312, 337 Pelham, R. J., 391 Pelin, K., 45 Peliska, J. A., 404, 406, 428, 431 Pellegrino, M. A., 187 Pelletier, H., 412, 416, 420, 423, 424, 425, 431 Pellman, D., 282, 285, 290 Penczek, P., 59 Penefsky, H. S., 348 Peng, S. B., 347 Pepe, F. A., 217, 235 Perderiset, M., 277, 278 Pereira, J. S., 187 Perez, F., 282, 285, 286 Perlman, Z. E., 282 Pernet-Gallay, K., 285 Perreault-Micale, C. L., 187 Perry, R. C., 370 Perry, S. V., 124, 134, 148, 235 Perutz, M. F., 17 Peterlin, B. M., 352 Petruska, J., 402 Pfeier, K., 347 Pham, P., 412, 423 Phillips, G. N., 123, 128, 131, 144, 145, 147, 148 Phillips, G. N., Jr., 135, 137, 138 Piazzesi, G., 231, 238, 239, 240, 243, 245 Piazzesi, P., 231 Piep, B., 231 Pink, S., 347, 357, 360, 365 Piroddi, N., 186 Plamann, M., 313 Plosky, B. S., 412, 422, 423 Poch, O., 403 Podolsky, R. J., 118, 221, 223, 229, 230, 234 Poetsch, A., 355 Poggesi, C., 186 Pohl, E., 352 Pollack, G. H., 90 Pollard, T. D., 121, 384, 385, 389, 391 Poltev, V. I., 402 Poole, K. J., 215 Poole, R. J., 350 Poolman, B., 351 Pope, B., 37 Popov, A. V., 283

AUTHOR INDEX

Popp, D., 36, 37, 38, 40, 145, 149, 208, 212, 221 Potter, J. D., 135 Poulsen, F. R., 75 Powell, A. S., 58, 60, 61, 63, 78 Powell, B., 347, 360 Poy, G., 271 Prakash, L., 404, 405, 406, 412, 417, 422, 423, 427 Prakash, S., 405, 406, 412, 417, 422, 423, 427 Prasad, R., 404, 412, 416, 423 Preston, Y. A., 165 Pringle, J. W. S., 40, 49 Prost, J., 391 Pryer, N., 269 Pucci, P., 45 Purcell, T. J., 163, 172 Purohit, V., 424 Putkey, J. A., 135 Pyzalska, D., 135

Q Qiu, S. H., 262, 272, 278 Quaggio, R. B., 140 Quarmby, L., 279

R Rabin, Y., 391 Radermacher, M., 59, 357, 361 Radocaj, A., 231 Ra, J. W., 285, 289 Rai, S. S., 265, 270 Rajendran, S., 408 Ramesha, A. R., 413 Ramos, C. H., 39, 140 Ramos-Morales, F., 285 Ranatunga, K. W., 186 Rao, S. T., 135 Rapp, G., 231, 239 Rastogi, V. K., 355, 366, 370 Ratajczak, R., 357, 360, 361, 368 Ravelli, R. B. G., 262, 265, 268, 270, 279 Ray, S., 412, 413, 417 Rayment, I., 7, 9, 23, 51, 52, 78, 145, 165, 166, 167, 170, 223, 307, 323 Raynaud, F., 45

459

Reconditi, M., 231, 238, 239, 240, 243, 245 Reddy, A. S., 300 Reddy, C. S., 307, 308, 322, 325 Redick, S. D., 102 Redkar, A., 103 Redwood, C., 63, 64, 217 Reedy, M. C., 148, 215, 228 Reedy, M. K., 31, 33, 40, 42, 52, 71, 72, 73, 78, 79, 148, 195, 196, 197, 215, 216, 217, 221, 225, 228, 231, 234, 237, 247 Reese, T. S., 338 Rehberg, M., 288 Reinach, F. C., 39, 139, 140 Reinach, F. D., 132 Reisler, E., 145, 148, 229 Remmel, B., 183 Ren, J., 412, 413 Reshetnikova, L., 127, 131, 132, 133, 139 Restle, T., 406, 409, 413 Reubold, T. F., 172, 175, 330, 331 Reuter, R., 350 Revington, M., 353, 365 Rhodes, T., 229, 230 Ribeiro, G., 404 Rice, P. A., 413 Rice, S., 303, 307, 308, 312, 325 Richardson, C. C., 403, 410, 414, 419, 431 Richardson, D. C., 126, 129 Richardson, J. S., 126, 129 Rief, M., 98, 99, 187 Rios, R. M., 285 Ris, H., 387 Ritchie, M. D., 147 Ritter, S., 43, 47 Rittinger, K., 406, 409, 413 Rizzo, V. F., 357 Roberts, K., 231 Roberts, T. M., 2, 4, 11, 384, 385, 386, 387, 389, 390, 391, 393, 395 Robson, R. M., 43 Rock, R. S., 187 Rodgers, A. J. W., 353, 365, 366, 368 Rodgers, D. W., 412, 413, 417 Rodionov, V., 268 Rodriguez, A. C., 410 Roessle, M., 64, 73, 195, 196, 197, 214, 215, 216, 221, 230, 231, 234, 245, 246 Rome, E., 217, 235 Ronjat, M., 282 Roos, K. P., 106

460

AUTHOR INDEX

Rose, M. D., 280 Rosenbaum, G., 229, 230 Rosenberg, C., 288 Rosenberg, J. M., 331 Rosenbluth, A., 56 Rosenbluth, M. N., 56 Rosenfeld, S. S., 165, 187 Rosenkranz, H. S., 271 Rosol, M., 123 Ross, C., 412, 413 Rossi, E., 93 Rossi, R., 187 Rostkova, E., 108 Roth, M. J., 413 Roth, S., 339 Rothwell, P. J., 2, 409 Roustan, C., 45 Rubinstein, J. L., 356, 366 Ruby, A., 304, 307, 321, 334 Rudy, D. E., 103, 112 Ru, C., 163 Rugarli, E. I., 279 Ruiz, J. F., 404 Ruiz, T., 357, 361 Ruiz-Lozano, P., 45 Ruiz-Opazo, N., 123 Runswick, M. J., 9, 11, 349 Ruppel, K. M., 9, 165 Ruppert, C., 352 Rusch, J., 286 Rypniewski, W. R., 23, 51, 145, 165, 166, 167, 223

S Sabbert, D., 350, 362 Sabido-David, C., 231 Sablin, E. P., 301, 307, 308, 310, 314, 315, 317, 321, 322, 325, 331, 334 Sack, S., 7, 303, 304, 305, 307, 310 Sackett, D. L., 267, 268, 271 Saeki, K., 386 Saez, C., 265 Safer, D., 274, 277, 308, 325 Sagermann, M., 357, 359 Saika, K., 352, 355 Sakai, Y., 336 Sakakibara, H., 12, 336, 338 Sakato, M., 287

Sakowicz, R., 280, 305, 307, 317, 318, 319, 320, 324, 327, 339 Sale, W. S., 336 Salmon, E. D., 269, 280, 384, 390, 391 Salser, S., 272 Sambade, M., 347 Sambongi, 363 Samejima, K., 289 Sander, C., 308 Sanders, C., 133 Sanderson, M. R., 410 Saniger, M. L., 404 Sano, K.-I., 133 Sarafianos, S. G., 413, 417 Saraste, M., 7, 9, 11, 43 Sardana, V., 307, 319 Saredi, A., 289 Sarkar, S., 135 Sasaki, H., 228 Sass, R. L., 43 Sato, Y., 404 Sattar, A. K., 410, 416 Satterwhite, L. L., 280 Satyshur, K. A., 135 Savage, C., 259 Savoian, R. M., 282 Sawaya, M. R., 412, 416, 420, 423, 424, 425, 431 Saxton, W. M., 300 Sayer, J. M., 412, 413, 422, 423 Sbia, M., 347, 357 Schaap, I. A. T., 279 Schaber, M., 307, 319 Schachar, R., 108 Schachat, F., 148 Schafer, D. A., 286 Sch€afer, G., 361 Sch€agger, H., 347 Schaper, J., 45 Schatz, M., 59 Scheers, D. J., 265 Schezek, K., 43 Scherrer, S., 286 Schleicher, M., 45 Schlick, T., 425, 426 Schlieper, D., 1, 3 Schliwa, M., 275, 300, 305, 307, 310, 311, 313 Schmid, R., 347 Schmidt, C. F., 279 Schmidt, J. J., 339

AUTHOR INDEX

Schmidt, R., 59 Schmidt-Base, K., 23, 51, 145, 165, 166, 167, 223 Schmitz, H., 42, 215 Schmitz, S., 187 Schneider, D. K., 39 Schneider, E., 355, 367 Scholey, J. M., 300 Scho¨ nbrunn, E., 305, 307 Schrader, M., 286 Schreiber, S. L., 339 Schroder, R. R., 36, 145, 169, 171, 172, 173, 174, 223, 330 Schroer, T. A., 282, 286, 287 Schroeter, J. P., 43 Schultheiss, R., 274 Schulthess, T., 127 Schutt, C. E., 35, 145, 146, 147 Schuyler, S. C., 282, 290 Schwarz, L., 355, 356 Schweinberger, E., 350 Seavy, M., 385, 390 Seeberger, C., 333 Seelert, H., 355 Seidel, C. A., 350, 409 Seidman, C. E., 180 Seidman, J. G., 180 Sekimoto, Y., 352, 355 Selden, L. A., 40 Sellers, J. R., 7, 165, 187 Sen Gupta, D. N., 404 Senior, A. E., 348, 355, 368 Sepsenwol, S., 387 Serneo, F. F., 277 Serr, M., 9 Severin, F., 283 Sha, B., 262, 272, 278 Shah, A. M., 424 Shah, G., 106, 107 Shahani, N., 282 Shao, E., 351 Sheard, K., 286 Sheetz, M. P., 338 Sheldon, J., 231 Sheterline, P., 36 Shilton, B. H., 356, 368 Shindyalov, I. N., 301, 308 Shipley, K., 280, 305, 307, 324, 327 Shirakihara, Y., 352, 355, 366 Shirasu-Hiza, M., 282, 284

461

Shock, D. D., 425 Short, S. A., 413 Shvetsov, A., 145, 148 Sia, S. K., 135 Siavoshian, S., 268, 270, 279 Sibbald, P. R., 7 Siddiqui, S. S., 259 Siddiqui, Z. K., 259 Sielecki, A. R., 135, 140 Siemankowski, R. F., 187 Siggia, E. D., 99 Silvanovich, A., 9 Silvian, L. F., 412, 423 Simmons, R. M., 34, 225, 231, 239 Simms, E. S., 409, 430 Simon, C., 269 Sindelar, C. V., 303, 307, 312 Singh, A., 126, 128, 130 Singhal, R., 404 Sinha, N. K., 417 Sithanandan, V., 230, 231, 245 Sjostrom, M., 27, 29, 38, 42, 64, 66, 74, 75, 76, 78, 217, 236, 237 Skeen, V. P., 280 Skiniotis, G., 305, 308, 332 Sleep, J. A., 98, 99, 101, 184, 186, 225 Sleeth, K. M., 405 Sloane, D. L., 402 Sluis-Cremer, N., 423 Slupsky, C. M., 135 Small, J. V., 75, 391 Smardon, A. M., 352 Smerdon, S. J., 413 Smertenko, A., 290 Smillie, L. B., 38, 123, 124, 133, 135, 138, 140, 144 Smith, A. N., 347 Smith, C. A., 7, 165, 170 Smith, D. A., 184, 225 Smith, J. E., 165, 187 Smith, M. J., 266, 273, 276, 277 Smith, R., 23, 51, 145, 165, 166, 167, 223 Smith, S. B., 98, 99, 429, 430 Smith, S. J., 388 Smits, R., 288 Snyder, J. P., 271 Sobel, A., 262, 265, 268, 270, 279 Sodek, J., 38, 123, 124, 140, 144 Somers, D., 413 Song, H., 307, 323

462

AUTHOR INDEX

Song, Y. H., 305, 307, 308, 310, 311, 313 Sorgen, P. L., 370 Sorimachi, H., 94, 97, 112, 113 Sosa, H., 245, 305 Sousa, A. D., 131, 133, 139 Sousa, R., 429, 430 Sowers, L. C., 402, 426 Sparrow, J. C., 36, 187, 330 Speelmans, G., 351 Spillantini, M. G., 273, 277 Spittle, C., 272, 273, 282, 283 Spouge, J. L., 336 Spudich, J. A., 9, 163, 165, 172, 187, 329 Spyracopoulos, L., 135, 137, 140 Squire, J. M., 1, 2, 3, 4, 5, 6, 19, 23, 25, 27, 29, 30, 31, 36, 37, 38, 39, 40, 42, 43, 45, 47, 48, 49, 51, 52, 55, 56, 57, 58, 59, 60, 61, 63, 64, 65, 66, 67, 68, 69, 71, 72, 73, 74, 75, 76, 77, 78, 79, 87, 94, 110, 124, 142, 144, 147, 148, 162, 163, 195, 196, 197, 203, 205, 206, 208, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 241, 242, 243, 244, 246, 247 Srayko, M., 279 St. Claire Allen, T., 231 Stahlberg, H., 355 Stalder, R., 285, 286 Stalz, W. D., 356 Stambouli, E., 360, 361 Stamer, K., 273 Stammers, D. K., 412, 413 Stark, H., 59 Starr, R., 29 Stefancsik, R., 135 Steen, W., 225 Steigmu¨ ller, S., 350 Stein, L. A., 165, 187 Steinberg, G., 313 Steiner, F., 69 Steinert, P. M., 127 Steinho, H. J., 413 Steinmetz, M. O., 127, 129 Steitz, T. A., 6, 403, 405, 406, 410, 412, 413, 414, 416, 418, 419, 422, 428, 430 Steven, A. C., 267, 268, 271 Stevens, F. C., 135 Stevens, T. H., 347, 350, 351, 357, 359, 360 Stewart, A., 231, 238, 239, 240, 245

Stewart, M., 2, 4, 11, 38, 42, 71, 74, 122, 123, 125, 126, 128, 131, 133, 144, 149, 384, 385, 386, 387, 389, 390, 391, 393, 395 Stewart-Jones, G. B., 412 Stier, G., 45 Still, I., 289 Stirtan, W. G., 413 Stock, A., 127 Stock, D., 355, 356, 361, 365, 366, 370 Stock, M. F., 334 Stoger, H., 352 Stokes, A. R., 17 Stone, D., 38, 39, 123, 124 Stone, D. B., 39 Stone, D. K., 347, 351 Stopak, D., 391 Strand, J., 131, 132, 133, 139 Stromer, M. H., 43 Strynadka, N. C., 135, 140 Stuart, D. I., 412, 413 Stubbs, G., 27, 206 Stuurman, N., 315 Su, L. K., 288 Suck, D., 34, 35, 145 Suda, K., 355 Sugimoto, Y., 63, 238, 245 Suh, S. W., 410 Sumner, I., 231 Sumner, J. P., 351 Sun, M. C., 90 Sun, Y.-B., 231, 245 Sun-Wada, G.-H., 350, 363 Sundaralingam, M., 135 Suo, Z., 406, 422, 423, 427 Supek, F., 347, 357 Supekova, L., 347, 357 Suprenant, K. A., 280 Surles, M. C., 126, 129 Surrey, T., 305, 308 Suzuki, H., 1, 4 Suzuki, J., 366, 370 Suzuki, M., 418 Suzuki, T., 90, 282, 363, 366, 367, 370, 371 Svennson, A., 231 Svensson, A., 231 Svergun, D., 305 Svitkina, T. M., 385 Swapna, G. V., 131, 132, 139 Sweasy, J. B., 424 Sweeney, E. M., 180

AUTHOR INDEX

Sweeney, H. L., 163, 165, 169, 171, 172, 173, 174, 175, 177, 179, 187, 223, 329, 330 Sykes, B. D., 39, 135, 137, 140 Szent-Gyorgyi, A. G., 52, 79, 143, 165, 187, 330, 333

T Tabor, S., 403, 410, 414, 419, 431 Tada, Y., 265, 271 Taiz, L., 350, 351 Takagi, M., 410 Takase, K., 351 Takeda, J., 347 Takeda, S., 39, 124, 135, 137, 140, 141, 150 Takei, Y., 275, 280 Takemori, S., 230 Takezawa, Y., 63, 238, 245 Takio, K., 265, 271 Talbot, J. A., 140 Tamakoshi, M., 352, 360, 361, 363 Tanaka, H., 245 Tanaka, T., 135, 277 Tanaka, Y., 280 Tanese, N., 413 Tanigawa, G., 180 Taniguchi, H., 140 Tanokura, M., 138, 140 Tantillo, C., 409, 413, 417 Tao, T., 140 Tarsio, M., 352 Tatsumi, R., 108 Tawada, Y., 138, 140 Taylor, E. W., 33, 162, 163, 164, 183, 308, 325 Taylor, J. S., 418 Taylor, K. A., 42, 228 Teller, A. H., 56 Teller, E., 56 Teng, J., 275 Tenney, K., 351 ter Haar, E., 271 Terada, S., 280 Terres, G., 347, 357, 360, 365 Tesi, C., 186 Theriot, J. A., 384 Thies, E., 273 Thirlwell, H., 231 Thoden, J. B., 9, 165 Thomas, D. D., 182

463

Thomas, M., 112 Thomas, W., 262, 272, 278 Thompson, A., 305, 307 Thorm€ahlen, M., 303, 304, 305, 307, 308 Thornell, L. E., 66 Thornton, J. M., 308 Tigyi-Sebes, A., 90 Timmins, P. A., 39 Timson, D. J., 64 Tippin, B., 405 Tirion, M. M., 145 Tirnauer, J. S., 285 Tischler, M. D., 106 Titus, M. A., 3 Tobacman, L. S., 123, 131, 132, 133, 139, 146, 148 Toda, T., 280, 283, 285 Todokoro, K., 290 Toh, H., 404 Tomchick, D. R., 23, 51, 145, 165, 166, 167, 223 Tomishige, M., 315 Tonchev, T., 305 Topal, M. D., 417 Tordo, N., 403 Toth, E. A., 412, 423 Towns-Andrews, E., 215, 231 Toyoshima, Y. Y., 336 Tran, P. T., 283, 284 Trayer, H. R., 64 Trayer, I. P., 64 Tregear, R. T., 31, 33, 215, 228 Trentham, D. R., 181, 182, 231 Trincao, J., 412, 423 Trinick, J. A., 63, 90, 98, 99, 101 Trinick, T., 49, 63 Tripet, B., 127, 128, 333 Trombitas, K., 45, 90, 93, 94, 95, 99, 100, 101, 102, 103, 104, 106, 112, 113 Trommler, P., 45 Trybus, K. M., 52, 79, 163, 165, 187 Tsai, M. D., 406, 417, 424, 427, 428 Tsao, J., 262, 272, 278 Tsaturyan, A. K., 223, 225, 230, 231, 245 Tsiavaliaris, G., 329 Tskhovrebova, L., 49, 63, 98, 99, 101, 110 Tsuchiya, H., 112 Tsuda, S., 135 Tsunoda, S. P., 363, 366 Tsuprun, V. L., 356

464

AUTHOR INDEX

Tu, A., 90 Tuerk, C., 410 Turina, P., 362 Turincio, R., 339 Turner, J., 280, 305, 307, 317, 318, 319, 324, 327 Turowski, P., 284 Tuske, S., 413 Tyska, M. J., 163

U Ubbink-Kok, T., 361, 368 Ueda, T., 352, 355 Uemori, T., 404 Ueno, H., 363, 370 Ueno, Y., 245 Uhlin, U., 365, 366 Uljon, S. N., 412 Uman, P., 147 Unwin, N., 272 Urbanke, C., 181 Usui, T., 265, 271 Uttenweiler, D., 64 Uyeda, T. Q., 163, 329

V Vaillant, A. R., 275 Vale, R. D., 259, 279, 300, 301, 304, 307, 308, 310, 315, 321, 325, 328, 333, 334, 336, 338 Valetti, C., 286 Valiron, O., 282 Vallee, R. B., 286, 334, 336 van Breemen, J. F. L., 361, 368 van Breugel, M., 284 van der Ven, P. F., 45, 69, 113 van Es, J. H., 288 van Heel, M., 59 Van Loock, M. S., 145, 148 Van Vactor, D., 286 Vande Berg, B. J., 424 Vanderlinde, O., 384, 393 Varma, D., 334, 336 Vasilyeva, E., 347, 357, 358, 359, 360, 365, 368 Vassylyev, D. G., 137 Vaughan, K. T., 282, 286 Vaughan, P. S., 286

Veigel, C., 187, 330 Venier, P., 272 Venzke, D., 357, 360, 361, 368 Verdine, G. L., 413, 417, 423, 431 Verma, S., 280 Vernos, I., 289, 290 Vibert, P. J., 42, 71, 74, 75, 77, 142, 143, 147, 213 Vihola, A., 113 Vik, S. B., 356, 367 Vinogradova, M. V., 124, 135, 138, 140, 141, 150, 307, 308, 322, 325 Vinokurov, L., 135 Vogel, R., 273 Volkmann, N., 123, 126, 128, 129, 131, 132, 149, 308, 332 von Brasch, A. A. M., 39 von Hippel, P. H., 402 Vosberg, H. P., 180 Vourloumis, D., 271

W Wada, T., 356 Wada, Y., 350, 363, 365, 369 Wade, R. H., 259, 262, 267, 282, 305, 307, 308, 320, 321 Wagner, V., 290 Wagschal, K., 127, 128 Wainberg, M. A., 406 Wakabayashi, K., 63, 238, 245 Wakabayashi, T., 386 Wakatsuki, S., 137 Wakelin, S., 179, 181 Waksman, G., 2, 410, 414, 418, 419, 420, 421, 428, 430, 431 Walczak, C. E., 280, 282, 300 Walenta, J. H., 288 Walker, J. E., 9, 11, 349, 352, 353, 354, 355, 356, 359, 360, 365, 366, 369, 370 Walker, M. L., 12, 338 Walker, R., 269 Wallace, S. S., 410, 416 Wallgren-Petterson, C., 49 Wallimann, T., 66, 69 Walsh, C. A., 278 Walsh, M., 135 Walther, D., 126, 129 Wang, B. B., 262, 272, 278 Wang, C. C., 410, 416

AUTHOR INDEX

Wang, H., 367, 370, 371 Wang, H.-W., 267, 268 Wang, J., 410, 413, 414, 416, 419, 422 Wang, K., 90, 93, 94, 102, 217 Wang, S. M., 90 Wang, T. S., 419 Wang, Y.-L., 387, 391 Wardale, J., 90 Warren, C. M., 91, 97, 106, 108 Warren, J., 413 Warshaw, D. M., 163 Washington, M. T., 405, 406, 417, 422, 423, 427 Wasserman, A. J., 271 Watanabe, H., 265, 271 Watanabe, K., 99, 100, 102, 103, 104, 108, 112 Watkins, H., 63, 64, 217 Watkins, J., 404 Watson, J. D., 17, 402 Watson, M. H., 133 Watts, N. R., 267, 268, 271 Wear, M. A., 45 Weaver, K. L., 413 Webb, M. R., 185, 223, 231 Weber, A., 142, 144, 145, 147 Weber, J., 355, 368 Weber, K., 45, 69, 71, 113 Weeds, W., 37 Wegner, A., 133 Wehland, J., 282 Weisberg, A., 39 Weiss, S., 187 Weissig, H., 301, 308 Weitman, S., 339 Welch, M. D., 391 Wells, A. L., 165, 172, 175, 187, 329, 330 Wendt, T. G., 308, 332 Werneburg, B. G., 406, 417, 424, 427, 428 Wesenberg, G., 23, 51, 145, 165, 166, 167, 223 West, W., 267, 268, 271 Westbrook, J., 301, 308 Wetzel, D. M., 286 Whitby, F. G., 131 Whitcomb, J. M., 413 White, A. E., 347 White, H. D., 9, 183, 185, 229, 230 White, H. W., 187 White, S. P., 138 Whiting, A., 90

465

Whittaker, M., 51, 52, 78, 166, 187, 223, 262, 265, 266, 272, 308, 325 Widen, S. G., 404 Wiech, H., 273 Wieczorek, H., 347, 351, 352, 357 Wiegant, J., 288 Wieland, T., 40 Wiesner, S., 390 Wilce, M. C., 365, 366 Wild, P., 391 Wilke-Mounts, S., 355, 368 Wilkens, S., 351, 353, 355, 356, 357, 358, 360, 361, 365, 366, 368 Wilkins, M. H. F., 17 Wilkins, S., 362 Williams, J. C., 334, 336 Williams, K. M., 368 Williams, R. C. J., 272, 275 Williams, R. L., 409, 413, 417 Wilm, M., 69, 113, 289, 290 Wilmanns, M., 69 Wilson, D., 100, 103, 104 Wilson, D. B., 305 Wilson, R. H., 17 Wilson, S. H., 404, 405, 409, 412, 416, 420, 423, 424, 425, 426, 431 Wilson-Kubalek, E. M., 187, 308, 325 Wimmers, S., 352 Winder, S. J., 43 Winegrad, S., 64 Winkelmann, D. A., 23, 51, 145, 165, 166, 167, 223 Winkler, H., 228 Wiseman, M. O., 187 Witt, C. C., 97, 101, 103, 108, 112, 113 Wittinghofer, A., 7 Wittmann, T., 289, 290 Woehlke, G., 304, 305, 307, 310, 311, 313 Wo¨ hrl, B. M., 406, 409, 413 Wolczyk, D. F., 351 Wolf, E., 305 Wolf, S., 259, 263, 266 Wolf, S. G., 308, 311 Wolf, Y. I., 7 Wol, J., 265, 270 Wong, I., 406, 407, 417, 426, 428, 431 Wong, W. W., 229 Wood, K. W., 320, 339 Woodgate, R., 405, 412, 416, 422, 423 Woods, C. M., 271

466

AUTHOR INDEX

Woodward, S. K., 181 Woolf, D. J., 412, 413, 417 Woolfson, D. N., 126 Woolley, R. J., 165, 179, 181, 182 Wordeman, L., 280, 300 Wray, J., 181 Wright, J., 94, 217 Wu, J. L., 140 Wu, K. C., 127 Wu, Y., 94, 95, 97, 99, 100, 106, 107, 108, 109 Wuite, G. J. L., 429, 430 Wynshaw-Boris, A., 277

X Xiang, X., 313 Xie, X. S., 347, 351, 357 Xing, J., 165, 187 Xu, B., 307, 319 Xu, S., 229, 230, 231 Xu, T., 347, 357, 365 Xuong, N. G., 410, 430

Y Yagi, H., 412, 422, 423 Yagi, N., 36, 213, 223, 225, 230, 234 Yajima, H., 307, 314, 315, 317, 331 Yamada, H., 356 Yamaguchi, M., 43 Yamamoto, A., 350, 363, 365 Yamamoto, M. T., 282 Yamasaki, R., 108, 109 Yamashiro, C. T., 351 Yamashita, A., 39, 45, 124, 135, 137, 140, 141, 150 Yamashita, I., 36, 40, 149 Yamato, I., 351, 360 Yan, Y., 307, 319 Yanagida, T., 133 Yang, G., 418, 419 Yang, J. M., 127 Yang, L., 425, 426 Yang, W., 406, 412, 416, 422, 423 Yano, J., 113 Yasuda, R., 350, 363, 365, 369 Yasunaga, Y., 386

Yen, T., 320 Yengo, C. M., 165, 172, 175, 187, 329, 330 Ylanne, J., 43 Yohn, C. B., 51, 52, 78, 166, 223 Yokoyama, K., 351, 352, 360, 361, 363 Yonekura, K., 1, 4 Yoshida, M., 348, 350, 351, 352, 355, 360, 361, 363, 365, 366, 367, 369, 370, 371 Young, M., 429, 430 Young, P., 43, 45, 69, 94 Yu, H., 352 Yu, L. C., 219, 221, 223, 229, 230, 231, 234 Yun, M., 305, 307, 321, 323, 324, 325, 331, 332

Z Zaccai, G., 305 Zeiske, W., 347, 357 Zeng, W., 179, 181 Zeviani, M., 93 Zhang, D., 356 Zhang, J., 351, 359 Zhang, J. Q., 39 Zhang, M., 135 Zhang, T., 126 Zhang, W., 413 Zhang, X., 305, 307, 323, 324, 325, 331, 352, 357, 360 Zhang, Y., 366 Zhang, Z., 351, 357, 360, 361 Zhao, J., 262, 272, 278 Zhao, Y., 410 Zheng, W., 339 Zheng, Y., 357, 360, 361 Zheng, Y. H., 352 Zhong, X., 406, 417, 424, 427, 428 Zhou, B. L., 405, 406, 412 Zhou, J., 11 Zhou, Q., 45 Zhou, Y., 350, 362 Zhou, Z., 347 Zhu, J., 59, 271 Zimmerman, S. B., 409, 430 Zimmermann, B., 350 Zimniak, L., 351 Zinnen, S., 406 Zmudzka, B., 404

SUBJECT INDEX

A A-ATPase function of, 352 introduction to, 346–348 structure of, 361 working model of, 348 AB10, 232, 234 AB11, 232, 234 A-band analysis of, 61–71 filament lattices in bony fish muscle, 219 filament lattices in striated muscle, 41 of sarcomere, 27 vertebrate, lattices, 30–31 Acetate buffer, treatment of sperm with, 392 Actin filaments, 1, 10 50K cleft and, 171 binding of, and powerstroke state, 182–185 binding of, and pre-powerstroke state, 181–182 binding of, and rigor states, 179–181 composition of, 4 decorated, 171 diffraction from, 206–215 diffraction patterns of different, 211 electron micrograph image of, 58 F-actin and, 36–38 I-band and, 49 of IFM, 208–209, 215 illustration of, 24 mass transfer between, and myosins, 223 monomer, 34–36 MSP and, 385–387 myosin heads on, 226, 227 myosin interactions with, 165 organization of, in contractile units, 40–42 polymerization, 391 pyrene label on, 180–181

radial net of, 65, 214 regulation of, 148–149 representation of, 210 in rigor muscle, 225–228 of sarcomere, 29 in striated muscles, 196 strong-binding complex between, and myosin, 174 structural features of, 208 subdomains of, 212–213 SW1 opening and, 176 thin, 38–40 troponin and, 38–40 twisting and untwisting of, 210–211 Z-band and, 34–50 Actin monomer, 34–36 illustration of, 35 in interference calculations, 241 subdomains of, 37 Actin trimer, 149 Actin-binding cleft, 172 -actinin molecules, 47 structure of, 44 in Z-band, 43 Acto-S1, in powerstroke, 185–186 Adenosine diphosphate (ADP), 2, 305, 315 ATP and, 23, 333 phosphate complex, 317 P-loop structure of, 7 rigor state and, 187–188 Adenosine triphosphate (ATP), 1, 147, 187 ADP and, 23, 333 binding of, 33 binding of, and powerstroke state, 182–185 binding of, and pre-powerstroke state, 181–182 binding of, and rigor states, 179–181 breakdown of, in myosin, 162 structure of, 7 synthesis of, 349

467

468

SUBJECT INDEX

Adenosine triphosphate synthase (ATPase) mitocondrial, 10 motor type of, 11 movement produced by, 12–14 ADP. See Adenosine diphosphate Alanines, N-terminal and, 129–130 Aluminium fluoride, 316 AM.ADP.Pi state, 229–230 A.M.D. state, 186 AMPPCP, 315, 316, 325 AMPPNP, 315, 316, 326 Animal cells, illustrations of, 3 APC, microtubules and, 288 Array lengths, diffraction peaks and, 205 Ascaris cytoskeleton of, 387, 388 leading edge dynamics and, 389 MSP dimer, 386 retraction force in, 391–392 sperm of, 384 Asp, 8–9, 180, 313, 367 Assembly-inhibiting drugs, microtubules and, 270–271 Asymmetric unit, 204 Atomic force microscopy (AFM), 98–99 schematic of, 100 single-molecule experiments with, 102 Atomic structure of microtubules, 262–625 of tubulin, 264 ATP. See Adenosine triphosphate ATPase domains, torque generation in, 369–370 ATPase. See Adenosine triphosphate synthase Axial structure, of sarcomeres, 50

B Bacterial flagella, 4 proton gradients in, 11 Bessel function, 208 Binding change model, 349–350 Binding sites, in relation to steric blocking, 144–145 Bony fish muscles A-band filament lattices in, 219 active and relaxed, 232 diffraction patterns from, 216, 220, 231–232

intensity profiles of, 220 LAD of, 216, 220 preferences for, 231 tetanic contractions in, 233 Bragg planes, 205 Bragg’s law, 212 BRCT. See Breast cancer susceptibility gene Breast cancer susceptibility gene (BRCT), 405

C C10, 11 CAD. See Coronary artery disease Calcium in elasticity regulation, 108–109 in TnC, 140–141 troponin and, 141 CAP-GLY domain, 285–286 atomic model of, 278 Cell division, 4–5 schematic illustration of, 6 CENP-E, 320 Centrioles, 6 Centromere-associated proteins, 320 Centrosomal coiled-coil proteins, microtubules and, 289–290 ch-TOG, microtubules and, 283–285 CLASPs. See CLIP-associated proteins CLIP-associated proteins (CLASPs), microtubules and, 285–286 CLIPs. See Cytoplasmic linker proteins Coiled-coil structure, 365 bending, 126 of tropomyosin, 124–127 Colchicine, 270 binding of, 271 Common-line projection theorem, 59 Conformational relays, in kinesins, 332–333 Conformational switching in kinesin, 328–333 in myosins, 328–332 Conformational transition in Klenow fragment, 424–425 open-to-closed, 421 Conformers on crossbridge, 165 definition of main, 168

469

SUBJECT INDEX

Convolution theorem, 209 of diffraction peaks, 202–204 illustration of, 203 Cooperative catalysis, 349–350 Core residues, in tropomyosin bending, 127–130 Coronary artery disease (CAD), 106 CPD. See Cyclobutane pyrimidine dimers C-proteins, 56 in A-bands, 61–71 arrangement of, on myosin filament, 218 myosin and, 64 peaks in C-zones, 236 substructure of, 63–64 Crossbridge configurations classification of, 75 in muscle contraction, 238 in myosin filaments, 74 –77 sarcomere lengths and, 243 Crossbridge cycle general ideas about, 196–197 illustration of, 32, 164 parts of, 165–166 in sarcomeres, 31–34 weak-binding state and, 228–229 Crossbridge polymorphisms, primary conformations of, 166–169 Crystallization, expression and, 164–165 Crystals 3D, 201 diffraction from, 199–201 lattice shapes in, 212 c-subunits, E. coli, 366 C-terminal of kinesins, 300 myosin binding to, 216 N-terminal and, 131–132 structure of, 132 of TnT, 139 C-type motors, 320–326 Cyclobutane pyrimidine dimers (CPD), 422 Cys residues, 359 Cytokineis, 6 Cytoplasmic linker proteins (CLIPs), microtubules and, 285–286 Cytoskeleton assembly of, 387–388 dynamics, 388 C-zone

C-protein peaks in, 236 interference, 235–236

D DCCD. See Dicyclohexylcarbodiimide DCX domains atomic model of, 278 model of, 276 N-terminal, 277 20 -deoxyribonucleoside-50 -triphosphate (dNTP), 402, 407, 418–419 incorporation, 427 50 -deoxyribose phosphate (dRP), 405 Dictyostelium G -actin, 386 myosin II, 7–8, 172, 330 studies in, 164–165 Dicyclohexylcarbodiimide (DCCD), 353 Difference intensity map, of diffraction patterns, 232 Differential splicing, in elasticity regulation, 104–107 Diffraction pattern from actin in rigor muscles, 225–228 in bony fish muscle, 216, 220, 231–232 difference intensity maps of, 232 from myosin filament backbone, 221–222 from myosin head array, 215–217 of myosin heads on actin, 227 of tropomyosin, 213 Z-band in, 217–221 Diffraction patterns of actin filaments, 211 from crystals, 199–201 extents of lattice planes in, 204–206 general ideas about, 197–198 helical, 209 of helical structures, 207 lattice planes in, 201–202 products of, 203 reciprocal nature of, 198 from relaxed fish, 57 scattering in, 199 of troponin, 213–214 Diffraction peaks array lengths and, 205 convolution theorem of, 202–204

470

SUBJECT INDEX

Diffusion-limited complex formation, 183–184 DNA, 17–18 DNA binding, primer/template, 414–417 DNA polymerases basic architecture of, 409–414 in E. coli, 409–410 family A, 403 family B, 403–404, 410 family C, 404 family D, 404 family X, 404–405, 410–411 family Y, 405–406, 412, 416, 422–423 as molecular motors, 429–430 nucleotide selection of, 402, 426–427 RT family, 406 structures of, 411 dNTP. See 20 -deoxyribonucleoside-50 triphosphate Doublecortin, 275–276 Dpo4, 416 Drosophila, 239, 240 dRP. See 50 -deoxyribose phosphate Dynactin complex microtubules and, 286 model of, 287 Dynein heavy chain, 336 domain structure of, 337 Dyneins, 259 electron microscopy of, 336 general structures of, 13 model of, 287 motor domains of, 12 N-terminal of, 336 structure of, 334–338

E E631A mutants, 324 EB1 domain atomic model of, 278 microtubules and, 288 EB1/APC complex, model of, 289 EDTA. See Ethylenediamine tetraactic acid Elasticity regulation calcium in, 108–109 differential splicing and, 104–107 phosphorylation in, 109 ELC. See Essential light chain

Electron micrograph of C-proteins, 64 cross-sectional view of crowns in, 60 of dynein, 336 of E. Coli F-ATPase, 358 history of, 90 of M-bands, 68 of microtubule binding, 266 microtubules visualized by, 258, 260 of motor binding on tubulin structures, 281 EMAN, 59 EMAP, as microtubule destabilizer, 280 E:p/t complex, structure of, 414 E:p/t:dNTP complex conformational transition of, 419–428 formation of, 417–419 structures of, 420–423 Equator, 206 changes in, reflections in muscle contraction, 230–234 Escherichia coli, 370 c -subunit structure, 366 DNA polymerases in, 409–410 F-ATPase of, 358 Essential light chain (ELC), 23 Ethylenediamine tetraactic acid (EDTA), 349 Eubacteria, enzymes found in, 351 Eukaryotic cells, movement of, 384 Exon microarray, titin, 97–98 Exon shuffling pathways, titin and, 94–96

F F-actin, 36–38 crystal structure of, 149 negatively stained, 146 regulation of, 148–149 structure of, 148 F-actin helix, 133 visualizing, 142 F-ATPase crystal structure of mitochondrial, 354 E. Coli, 358 F0 structure in, 355–356 F1 structure in, 352–355 F1F0 structure in, 356 functions of, 348–350

471

SUBJECT INDEX

introduction to, 346–348 motor efficiency in, 371 in rotation experiments rotational catalysis and, 362–363 rotor in, 365–367 stator, 367–358 structure of, 352–356 subunit composition of, 347 torque generation in, 368–371 working model of, 348 Fiber complexes, 387 50K domain, 172, 173 actin binding closing, 171 in powerstroke, 185 stereo-specific interactions of, 183 Fn3-like domains, 63–64 Fourier transform, radial nets of crossbridge lattices and, 73 FtsZ, 267

G G347, 321 G-actin, 34–35, 36, 145 Dictyostelium, 386 GAP. See GTP hydrolysis-activating protein GDP. See Guanosine diphosphate Globular particles, analysis of, 59 Globular proteins, 4 Glutamic acid, 266 Glutamine, 180 Glycogen, 22 GMPCPP, 272 Grating equation, 198 GTP hydrolysis-activating protein (GAP), 265 GTP. See Guanosine triphosphate GTPases, 7 domain, 265 Guanosine diphosphate (GDP), 9–10 microtubules and, 269 Guanosine triphosphate (GTP), 9–10, 258, 263 binding sites of, 265–267 hydrolysis, 266, 267–268 microtubule dynamic behavior and hydrolysis of, 269–270 in microtubule stabilization, 272

H HAC muscle. See Human adult cardiac muscle HAS muscle. See Human adult skeletal muscle Head-to-tail overlap, in tropomyosin, 131–133 Helical diffraction, 206 generation of, 209 Helical structure, geometry of, 207 -helical coiled coil, heptad repeats of, 125 HELIX, 27 Heptad repeats of -helical coiled coil, 125 of tropomyoscin, 124–127 Heterodimeres, 270 High-angle diffraction, of myosin rods, 222 HIV-1 RT, 409, 412–413, 417, 423 discrimination of, 426–427 Hook proteins, microtubules and, 288 HsKSP, 317 Human adult cardiac (HAC) muscle, analysis of, 96 Human adult skeletal (HAS) muscle, analysis of, 96

I I-band Exons 49-224 and, 95 filament organization in, 49 of titin, 63 IFM. See Insect fibrillar flight muscles Ig segments, 90–91 in titin, 99–101 Ig-like domains, 63–64 Ile, 313 IMAGIC, 59 Insect fibrillar flight muscles (IFM) actin filaments of, 208–209, 215 filament organization in, 40–41 X-ray crystallography of, 71, 72 Insect leg muscle, myosin filaments in, 42 Intein, 351 Intensity profiles of bony fish muscle, 220 from frog sartorius muscle, 235

472

SUBJECT INDEX

Intercalated disks, 19–20 Interference calculating, 241–242 C-zone, 235–236 experiments, 240–245 interpretations of, 237–239 introduction to, 234–237 in muscle transients, 239–240 origin of, 236 scattering objects and, 242 summarized, 198 Ion channel, torque generation in, 370 IQ motifs, of myosin, 166 Isometric case, 186–187 Isotonic contractions, muscle contraction and, 245–246

K K354, 305 Kar3, 322–325 motor domain, 323 Ncd and, 323 point mutants, 323–324 Katanin, 262 as microtubule destabilizer, 279 KCBP. See Kinesin-like calmodulin-binding protein KHC. See Kinesin heavy chain Kif1A, 314, 315 Kif2C, 326–327 Kinesin heavy chain (KHC), 300 Rn, 309 Kinesin light chains (KLC), 334, 335 Kinesin-like calmodulin-binding protein (KCBP), 325–326 Kinesin-related domains, structures of, 333–334 Kinesins, 1–2, 259 classes, 299–301 comparing structures, 310–328 conformational switching in, 328–333 C-type, 320–326 in different nucleotide states, 331 dimetric, and microtubule protofilaments, 310 domains of, 300–301 fungal, 313–314 general structures of, 13

globular tail domain of, 334 human v. rat, 310–313 microtubule binding sites of, 281–282 microtubule complex, 308–309 as microtubule destabilizer, 280–282 microtubules and, 266 monomeric, 314–317 in motor binding, 281 motor domain binding of, 311 motor domain structure of rat, 302 motor domains of, 301 motor types of, 12 M-type, 326–328 nucleotide binding, 332–333 organelle transplant and, 3–4 PEG -grown crystal structure of, 312 proteins associated with, 335 as prototypical motor, 301–310 seven, 320 structural alignment of Switch regions of, 306–307 SW2-clusters in, 309 switch 1 and 2 conformational relays, 332–333 switch 1 in, 331 tetrameric, 317–319 Kinetochores, 6 KLC. See Kinesin light chains Klenow fragment, 410, 414 conformational transitions in, 424 –425 crystallizing, 418 Klentaq1 crystallizing, 418 structure of, 415, 421 ternary complex formation in, 420

L LAD. See Low-angle diffraction Lamellipodium, reconstitution of, 389–391 Laser-tweezer experiments, 100–101 Lattice planes description of, 201–202 extents of, 204–206 LCBD. See Light-chain binding domain Leading edge dynamics, reconstituting, 389 Lethocerus, 208 Ligands, titin, 110–114 Light, 197–198

SUBJECT INDEX

Light-chain binding domain (LCBD), 328, 329 Listeria, 390 Low-angle diffraction (LAD), 212 from bony fish muscle, 216 of bony fish muscle, 220 of myosin heads, 221 from permeabilized bony fish muscle, 224 Lynn-Taylor cycle, structural correlates of, 178–179 Lysine, 266

M M3 intensity, 240 M3 layer lines, 227–228 M3 peaks, 238 sarcomere lengths and, 243, 244 M6 reflections, 240–241 M.ADP.Pi state, 229–230 Major sperm protein (MSP), 383 actin and, 385–387 Ascaris, 386 vesicle motility and, 390–391 MAP1A, 275 MAP1B, 275 MAP2, 273 MAP4, 273 MAP65, 290 MAPS. See Microtubule-associated proteins Mass transfer, between myosin and actin in rigor muscle, 223 M-band, 27–28, 64–65 A-bands and, 61–71 cardiac, 70 densities of, 66 electron microscopy of, 68 fish muscle, 67 fish simple lattice, 66, 68 myosin filament structure and, 51–79 MCAK, 327–328 Meridian, 206 MFPs. See MSP fiber proteins MgATPase, 349 Microtubule-associated proteins (MAPS), 5, 259, 271 atomic models of domains of, 278 structural, 272–278

473

Microtubules APC and, 288 assembly of, 270 assembly-inhibiting drugs and, 270–271 atomic structure of, 262–265 atomic structures of protofilament in, 259–262 centrosomal coiled-coil proteins and, 289–290 ch-TOG and, 283–285 CLASPs and, 285–286 CLIPs/þTIPs and, 285–286 composition of, 4 curved protofilament structure and, 268 DCX domain interacting with, 276 destabilizers, 279–282 dimetric kinesin and, 310 dynamic instability of, 268–272 dynctin complex and, 286 EB1 and, 288 EM visualization of, 258, 260 EMAP and, 280 GTP hydrolysis and, 269–270

-tubulin ring complexes and, 282–283 hook proteins and, 288 katanin and, 279 kinesin and tau binding to, 266 kinesin binding sites on, 281–282 kinesins and, 280–281, 308–309 MINUS and, 282 Msps and, 283–285 proteins controlling location of, 282–290 protofilament numbers in, 261 stabilization mechanisms of, 271–272 stahmin and, 279 structure of, 5, 258–268 subunit lattices of, 259 tandem domain molecules and, 277–278 tau interacting with, 276 XMAP 215 and, 283–285 Miller indices, 201 demonstration of, 202 MINUS, microtubules and, 282 Mitochondria, 2 M-loops, 262, 271 Molecular biology, early days of, 17–18 Molecular rulers, in striated muscle sarcomeres, 39–40 Monastrol, 319–320

474

SUBJECT INDEX

Motifs, 204 Motor domains plasticity of, 318 post rigor state of myosin, 167 Motor efficiency, in F-ATPase, 371 Motor nerves, 20–21 Motor proteins common features of, 7–11 locations and roles of, 2–6 Motor units, fibers in, 21 M-proteins, 69 MSP fiber proteins (MFPs), 390 MSP. See Major sperm protein Msps, microtubules and, 283–285 M-type motors, 326–328 MULV, 412–413 Muscle. See also Skeletal muscle; Striated muscle contractile units in, 40–42 electrical stimulation of, 22 hierarchy of, 19–34 organelles found in, 3 schematic illustration of hierarchy of, 18 speeds of, 22 structure of, 196–197 Muscle contraction crossbridge events in, 238–239 equatorial reflections in, 230–234 general overview of, 161–162 isotonic, 245–246 myosin in, 51–52 temperature jump experiments on, 245–246 theories of, 2–3 time-resolved events in, 230–234 in voluntary muscles, 20 Muscle transients, interference changes during, 239–240 MusLabel, 226, 241, 243 Myofibrils, 20 T-tubular network and, 21 Myomesin, 69 Myosins, 1–2 actin interactions with, 165 arrangement of C-proteins on, filament, 218 arrangement of heads on, 54 ATP breakdown in, 162 classification of crossbridge configurations in, 75

conformational switching in, 328–332 C-proteins, 64 crossbridge arrangements on, 74 –77, 162 crossbridge of, in pre-powerstroke state, 176 diffraction from, filament backbone, 221–222 diffraction from, head array, 215–217 domains of, 52 electron micrograph image of, 58 expression of, 164–165 fiber axis, 76 filament long axis, 76 in fish and insects, 77 folding motor domains, 329–330 general overview of, 161–162 general structures of, 13 head attachment, 32, 33–34 head locations in, 215 head organization in relaxed vertebrate myosin filaments, 56–61 heads of, on actin, 226, 227 helical ordering of, 229 high-angle diffraction, 222 illustration of, 24 in insect leg muscle, 42 invertebrate, 71–74 IQ motifs of, 166 mass transfer between, and actin, 223 M-band and structure of, 51–79 modeling, 55 motor domains of, 12, 330–331 in muscle contraction, 51–52 organelle transplant and, 3–4 P-loop structure of, 7–8 post-rigor and pre-powerstroke, 169–171 post-rigor structure of motor domain in, 167 regions in, 8 rigor state models and, 172–173 in sarcomere, 23–29 in scallop striated abductor muscle, 74 simple lattice arrangements of, 30 stereo views of heads of, 52 in striated muscles, 25, 196 strong-binding complex between, and actin, 174 structure of, head, 26 swinging lever arm of, 163–165 X-ray diffraction and structure of, 51–56

475

SUBJECT INDEX

N N2B-unique sequence, 108–109 titin and, 103–104 N600K mutants, 322 Ncd, 320–322 dimeric, 321 Kar3 and, 323 in motor binding, 281 NcKin, 311, 314, 325 NCS. See Noncrystallographic symmetry Nebulin, 39 in Z-band, 45 Neck helix, 333 Neck linkers, 313 Nemaline myopathy, Z-band structure in, 45–46 Neurospora crassa kinesin-1, 305 motor domain binding of, 311 NMR. See Nuclear magnetic resonance NNRTIs. See Non-nucleoside RT inhibitors Noncrystallographic symmetry (NCS), 319 Non-nucleoside RT inhibitors (NNRTIs), 413 Novx-3 pathway, Exon 48 and, 95 N-terminal actin binding to, 216 core alanines and, 129–130 crystal structure of, 128–129 C-terminal and, 131–132 DCX domains of, 277 of dynein, 336 of kinesins, 300 main chain, 131 in post rigor myosin motor domains, 167 structure of, 132 of tau, 274 N-type motors, 314–320 Nuclear magnetic resonance (NMR), 124, 271, 367 of F-ATPase, 355 Nucleotide binding in F1, 353 in kinesins, 332–333 Nucleotide binding pockets, domains of, 304 Nucleotide incorporation pathway correct v. incorrect discrimination and, 426–427 general theme of, 406–408

kinetic pathway of, 407 steps of, 417 variations on, 408–409 Nucleotides, DNA polymerase selection of, 402 NuMA, 289

O OH groups, 8 Organic solvents, 367

P PAD. See Polymerase associated domain Paircoil, 305 Paramyosin, 42 Passive stiffness, 106 PEVK repeats, 91, 97 PEVK segment, 90–91 elasticity of, 104 in titin, 101–103 Phalloidin, 39–40 Phosphate in powerstroke, 185–186 release of, 189 Phosphoryl transfer reaction, product release and, 428 Phosphorylation, 109 Pironetin, tubulin binding of, 270–271 PKA. See Protein kinase A Plasmodium falciparum, 327 P-loops characteristic sequences of, 9 constraining, 170 of myosin II, 7, 8 Plus-end tracking proteins (þTIPS), microtubules and, 285–286 Podophyllotoxin, 270 Point mutations, 304–305, 321 Kar3, 323–324 Pol , 424, 425 Polymerase associated domain (PAD), 412 Polymerization, 269 Polypeptides, 262 Post-rigor state, 166 actin binding and, 179–181 ATP binding and, 179–181

476

SUBJECT INDEX

Post-rigor state (cont.) classifying, 177–178 myosin, and pre-powerstroke, 169–171 of myosin motor domains, 167 pre-powerstroke state and, 181–182 Potentiated state, tropomyosin in, 147–148 Powerstroke, 177 actin binding and, 182–185, 188 ATP binding and, 182–185 defined, 169 structured origin of, 171 unloaded, 185–186 Pre-powerstroke actin binding and, 181–182 ATP binding and, 181–182 classifying, 177–178 myosin, 169–171 rigor like state v., 175 strongly bound, 175–177 truncated myosin crossbridge in, 176 Primer/template DNA (p/t DNA), 414–417 binding of, 416 Klentaq1 and, 415 translocation of, 429 Pro-Ala domains, 64 Product release, phosphoryl transfer reaction and, 428 Protein Data Bank, 301 Protein kinase A (PKA), 109 Protofilament, 270 atomic structures of, 259–262 curved structure of, 268 dimetric kinesin moving along, 310 numbers of, in microtubules, 261 p/t DNA. See Primer/template DNA Push-pull model, 396 Pyrene label, on actin, 180–181 Pyrophosphate products, 407

R R598A mutants, 324 Radial nets, 55 of actin filaments, 65, 214 of crossbridge lattices, 73 Random variable twist, 38 Rate-limiting step, molecular basis for, 423–426 RB69, 410, 418–419

ternary complex, 422 Recombinant DNA technology, 163 Region C, 228 Regulatory light chain (RLC), 23 Relay helix -sheet and, 173–175 kink in, 188 Relay/converter conformation, 178 Retraction force, 391–393 reconstitution of, 393–395 YOP in, 394 Reverse transcriptase polymerase chain reaction (RT-PCR), 97 R-factor, 54 Ribbon diagram of  -tubulin heterodimer, 263 of Ascaris MSP dimer, 386 of Dictyostelium G-actin, 386 of rat kinesin-1 motor domains, 302 of tubulin, 264 Rigid body movement, 312–313 Rigor muscle actin labeling in, 225–228 introduction to, 222–223 Rigor state, 31 actin binding and, 179–181 actin binding in, 171 ADP and, 187–188 ATP binding and, 179–181 classifying, 177–178 myosin V and, 172–173 pre-powerstroke state v., 175 RLC. See Regulatory light chain RNKHC-S203, 315 Rotation experiments F-ATPase in, 363 illustration of, 364 Rotational catalysis, F-ATPase and, 362–363 Rotors elasticity of, 370–371 F-ATPase, 367 RT-PCR. See Reverse transcriptase polymerase chain reaction

S SAD. See Small-angle diffraction Sarcomeres, 20, 204–205 A-band of, 27–28

477

SUBJECT INDEX

actin filaments of, 29 band lattices in, 217 crossbridge cycle in, 31–34 electron micrograph of, 28 force extension curves of titin in, 105 introduction to, 23 lengths of, and crossbridge arrays, 243 M3 reflection and, 244 molecular rulers in, of striated muscles, 39–40 myosin filaments in, 23–29 schematic diagram of, 28 sliding filament model in, 31 in striated muscles, 196 titin filament system in, 110 titin in cardiac, 111 titin layout in, 92 transverse and axial structure of, 50 two-filament model of, 89–90 vertebrate A-band lattices in, 30–31 Scattering in diffraction, 199 interference and, 242 of X-ray beams, 200 -sheet distortion of, 184 relay helix and, 173–175 seven-stranded, 170 as supporting structure, 304 twisting of, 175 Sinusoidal length changes, 239 Skeletal muscle, 20 Sliding filament model, of sarcomeres, 31–34 Small-angle diffraction (SAD), 212 Spastin, 279 Sperm, 384 acetate buffer treatment, 392 locomotion in, 387–388 push-pull model for, 396 SPIDER, 59 Stable tubule only polypeptides (STOP), 275 Stahmin, as microtubule destabilizer, 279 Stathmin, 270 Stator elasticity of, 370–371 of F-ATPase, 367–368 Steric blocking model, 39 atomic structures and molecular models in, 145–147

binding sites in relation to, 144 –145 explained, 142 thin filaments and, 143 STOP. See Stable tubule only polypeptides Strained complex, 184 Strained state, 186–187 Striated muscle A-band filament lattices in, 41 assembly of, 195–196 defined, 19 molecular rulers in sarcomeres of, 39–40 myosin molecules in, 25 structure of, 20, 196 transverse structure and axial structure of sarcomeres in, 50 tropomyosin terminal structures, 132 troponin in, 214 Stu2p domains of, 283–284 polypeptides, 284 Surface residues, tropomyosin, 130 SW1. See Switch 1 SW2. See Switch 2 Swinging lever arm, of myosin, 163–165 Switch 1 (SW1), 9–10, 172 actin binding and, 176 conformation of, 318 conformational relays in kinesins, 332–333 in kinesins, 331 movement of, 330 mutants, 324 in powerstroke state, 184 structural alignment of, 306–307 Switch 2 (SW2), 9–10, 169–170, 173 closure of, 182 conformation of, 309, 318 conformational relays in kinesins, 332–333 movement of, 330 in powerstroke state, 184 in pre-powerstroke state, 177–178 rigid body movement of, 312–313 structural alignment of, 306–307

T TACC, 289 Tandem domains, microtubules and, 277 Taq, 414

478

SUBJECT INDEX

Tau, 273–275 binding of, 271–272 microtubules and, 266 model of, 276 regions of, 274 Taxol, 267 binding of, 273 tubulin and, 271 TEDS site, 180 Temperature jump experiments, muscle contraction and, 245–246 Tension records, 21, 34 Tension responses, 21 Terminal web (TW), 5 Ternary complex, 319 of RB69, 422 Tetanic contraction, 22 in bony fish muscles, 233 Tetanus, 22 Tetratricopeptide repeats (TPR), 334, 335 Thermoplasma acidophilum, 346 Thin filaments of actin, 38–40 atomic structures and molecular models of, 145–147 regulation of, 146 steric blocking and, 143 turning on, 142–149 Titin, 39 in A-bands, 61–71 in cardiac sarcomere, 111 differential splicing of, 94–97 elasticity of, 98–109 elasticity regulation in, 104–109 exon microarray, 97–98 exon shuffling pathways, 94–96 filament system, 110 force extension curves of, 105 functional genomics of, 91–98 in heart, 107 in heart disease, 106 history of research on, 89–91 Ig segments, 99–101 layout of, in sarcomeres, 92 ligands, 110–114 mutations in, exons, 112–113 N2B-unique sequence and, 103–104 organization of, at genomic level, 93–94 organization of, at protein level, 91–93 PEVK extension, 101–103

repeating patterns in A-band of, 62 as wormlike chain, 98–99 Z-repeats of, 48 TnC. See Troponin C TnI. See Troponin I TnT. See Troponin T Top-of-powerstroke state, 179 Torque generation in ATPase domains, 369–370 in F-ATPase, 368–371 in ion channel, 370 TPR. See Tetratricopeptide repeats TPX2, 289 Transverse structure, of sarcomeres, 50 Tropomyosin, 34–35, 38 amino acid sequences of, 123 aperiodic features of, 130–139 atomic structures of, 145–147 binding sites, 144–145 core residues in bending of, 127–130 diffraction pattern of, 213 heptad repeat and coiled-coil structure of, 124–127 local bending of, 126 periodic features of, 122 in potentiated state, 147–148 role of, 123–124 structure of, 37 surface residues, 130 terminal structures of striated muscle, 132 TnT and, 134–139, 141 troponin bound to, 136 Troponin, 34–35 actin filaments and, 38–40 atomic structures of, 145–147 binding sites, 144–145 calcium and, 141 diffraction pattern of, 213–214 head structure, 140 intereference and, 237 role of, 123–124 in striated muscles, 214 structure of, 134–138 thin filament linkage to, 136 tropomyosin binding with, 136 turning on, 139–141 Troponin C (TnC), 123 calcium and, 140–141 crystal structure of, 37

479

SUBJECT INDEX

ribbon diagram of, 134 subunit, 135 Troponin complex, 134–139 Troponin I (TnI), 123 inhibitory region of, 141 structure of, 135–136 Troponin T (TnT), 123 C-terminal of, 139 interaction sites of, 138–139 structure of, 135–136 tropomyosin and, 134–139, 141 Trypsin, 271 Tryptophan, 425 T-tubular network, 21 Tubulin atomic models of, 264 GTP hydrolysis and, 267–268 heterodimers, 263 motor binding on, 281 pironetin and, 270–271 ribbon diagram of, 264 Taxol and, 271 unpolymerized, 258  -tubulin heterodimer, ribbon diagram of, 263 -tubulin, 267–268 dynamics of, 268–269 stabilizing drugs and, 271–272 -tubulin, 265 dynamics of, 268–269 stabilizing drugs and, 271 tau and, 274–275

-tubulin ring complexes, microtubules and, 282–283 TW. See Terminal web Twitches, 22 Tyr, 9, 425, 429

U Unit cells, 201 contents, 204

V Vanadate, 317 V-ATPase functions of, 350–352 introduction to, 346–348

stators in, 368 structure of, 357–361 subunit composition of, 347 V0 structure, 359–360 V1 structure in, 357–359 V1V0 structure, 360–361 working model of, 348 Vesicle motility, MSP driven, 390–391 Vinblastine, 271

W Walker A regions, 11 Walker B regions, 11 Watson-Crick (W-C) base, 402 W-C base. See Watson-Crick base Weak-binding state, 228–229 WLC. See Wormlike chains Wormlike chains (WLC), titin as, 98–99

X Xenopus, 273 Xklp2, 289 XMAP215 microtubules and, 283–285 polypeptides, 284 XMAP230, 273 X-ray beams, 199–200 scattering of, 200 X-ray crystallography actin monomer structure solved by, 35 myosin head structure and, 26 X-ray diffraction, 17–18, 25 of C-proteins, 64 cross-sectional view of crowns in, 60 of IFM, 71, 72 muscle structure and, 196–197 myosin structure and, 51–56 patterns from relaxed fish, 57 X-ray interference measurements, implications of, 234–246

Y Yersinia phosphatase (YOP), 393 in retraction, 394

480

SUBJECT INDEX

Z Z-band actin filament structure and, 34–50 -actinin molecules in, 43 assembly model, 48 in diffraction patterns, 217–221 modular structures of, 46 nebulin in, 45

in nemaline myopathy patients, 45 structure of, lattices, 44 thickness of, 43–44 Z-crystals and, 47 Z-crystals, 45 vertebrate muscle Z-bands and, 47 Z-repeats Exons 8-14 and, 94 of titin, 48