Académique Documents
Professionnel Documents
Culture Documents
www.elsevier.com/locate/jelekin
Abstract
The independent force generator and the power-stroke cross-bridge model have dominated the thinking on mechanisms of muscu-
lar contraction for nearly the past five decades. Here, we review the evolution of the cross-bridge theory from its origins as a two-
state model to the current thinking of a multi-state mechanical model that is tightly coupled with the hydrolysis of ATP. Finally,
we emphasize the role of skeletal muscle myosin II as a molecular motor whose actions are greatly influenced by Brownian motion.
We briefly consider the conceptual idea of myosin II working as a ratchet rather than a power stroke model, an idea that is explored
in detail in the companion paper.
2002 Elsevier Science Ltd. All rights reserved.
Keywords: Skeletal muscle; Sliding filament theory; Cross-bridge theory; Brownian motion; Molecular motors; Mechanisms of contraction; ATP
hydrolysis; Muscle energetics; Molecular ratchet
1050-6411/02/$ - see front matter. 2002 Elsevier Science Ltd. All rights reserved.
PII: S 1 0 5 0 - 6 4 1 1 ( 0 2 ) 0 0 0 3 6 - 6
426 W. Herzog, R. Ait-Haddou / Journal of Electromyography and Kinesiology 12 (2002) 425–433
Therefore, on average, the actin filament would undergo (almost) uniformly over the range [⫺0.5la, 0.5la], where
a zero displacement relative to the myosin filament. la is the distance between actin attachment sites. The pro-
Obviously, such a system cannot produce directed mus- portion of attached cross-bridges, n(x, t), is now a func-
cle force and shortening. In order to produce a directed tion of x and t, and the rate of change in n(x, t) is
movement of actin relative to myosin (such that muscle obtained as the material derivative:
shortening and directed force production could occur),
Dn ∂n ∂n
Huxley [13] introduced the idea that the rate of attach- ⫽ ⫹ v (3)
ment (f) and detachment (g) were dependent on the x- Dt ∂t ∂x
location (Fig. 4) of the actin attachment site relative to where v is assumed to be negative for muscle shortening.
the cross-bridge equilibrium position. These rate con- The governing differential equation, assuming that the
stants had to be chosen asymmetrically in the sense that two myofilaments are perfectly rigid, becomes:
冉 冊
the attachment probability of a cross-bridge had to be
greater for positive x-values (thus, cross-bridge attach- ∂n ∂n 1
⫹ v ⫽ ⫺n f(x)⫺ng(x) (4)
ment would cause muscle shortening) than for negative ∂t ∂x Ia
x-values. Huxley’s asymmetrical choice is shown in Fig.
Eq. (4) may be solved using the method of characteristic
5. No explanation was given as to how this asymmetry
curves [5,35].
may be produced.
If we now consider, as Huxley [13] did, a great num-
ber of M–A pairs that have one and the same x-location
3. The 1969 formulation of the cross-bridge model
at each instant in time, the proportion of attached cross-
bridges, n(x), is a function of time exclusively. By defi-
So far, the cross-bridge head (M, Fig. 4) was thought
nition of f(x) and g(x), the rate of change of n(x) over
to be in a single configuration. However, Reedy et al.
time becomes:
[30] demonstrated that cross-bridges at rest were about
dn perpendicular to the filament axis, while cross-bridges
⫽ (1⫺n)f(x)⫺ng(x) (1) in rigor were tilted 45° to the perpendicular rest con-
dt
figuration. Based on this information, H. Huxley [18]
For a dynamic equilibrium state (i.e., dn/dt=0), we find proposed that the cross-bridge head (subfragment 1)
that the number of attached cross-bridges, neq, is given attaches to the thin filament, and force and filament
by the probability of attachment: transport would be generated by a tendency of the head
f(x) to rotate about its attachment point (Fig. 6). Therefore,
neq ⫽ (2) subfragment 1 was assumed to act as a crank, pulling
f(x) ⫹ g(x)
the thin relative to the thick filament. Subfragment 2,
So far, we have assumed that all cross-bridges have the that connects subfragment 1 to the backbone of the thick
same value x at each instant in time, t. However, for a
great number of cross-bridges, x(t) would be distributed
from the rigor conformation, Rayment et al. [29] sug- The binding of ATP catalyzes the dissociation of myosin
gested that the narrow cleft that splits the 50 kD seg- from actin [3,32]. When dissociated, ATP is hydrolyzed,
ments of the myosin heavy chain sequence into two and the cross-bridge head goes through a conformational
domains is closed (Fig. 11A, horizontal gap, perpendicu- change (the recovery stroke, Fig. 12). Myosin attaches
lar to the actin filament axis). Addition of ATP, and to actin, phosphate is released, and the power or working
initial binding of ATP to myosin at the active site causes stroke is initiated. ADP is released, establishing the rigor
an opening of the narrow cleft between the upper and conformation. ATP then attaches to the myosin head,
lower domains of the 50 kD segments. This event, in myosin can now dissociate from actin, and the cycle
turn, disrupts the “strong” binding between actin and may restart.
myosin, but still allows for a “weak” attachment (Fig. Obviously, there are variations from the above pro-
11B). The final ATP binding to myosin causes a closure posed theory of contraction, however Figs. 11 and 12,
of the nucleotide binding pocket and a corresponding and Table 1 represent the most accepted thinking at
change of the myosin molecule. Myosin now detaches present. An interesting variation of the above ideas was
from actin and ATP is hydrolyzed (Fig. 11C). Rebinding proposed by A. Huxley [14] who suggested the possi-
of myosin to actin can now occur, presumably in mul- bility of cross-bridge rotation about its attachment point
tiple steps. The gap between the upper and lower domain on the actin filament. Another important aspect of the
closes in this process to produce strong binding, and scheme discussed above is that the rate limiting step of
phosphate, P, is released. This event starts the power the cross-bridge cycle occurs in the attachment of myo-
stroke (Fig. 11D). During the power stroke, the myosin sin to actin (Table 1). The remaining steps in the
S1 reverses its conformational change induced by ATP hydrolysis cycle are considered much faster than the
binding, the active site pocket is reopened, ADP is attachment rate.
released, and the rigor conformation is established (Fig.
11E). The cross-bridge cycle can now be restarted by
ATP binding to myosin. 6. Limitations
The detailed reactions of myosin S1 with actin and
ATP are not known because not all of the rate constants The detailed mechanical steps of muscle contraction,
have been determined. However, combining the struc- as well as the details of the actin–myosin hydrolysis
tural evidence [28,29], the biochemical data measured cycle are not known. And to cite A. Huxley [14]:
in solution [2,8–10,22–24,31], and the mechanical data
[6,16], it is possible to infer the likely actin–myosin the whole history of muscular contraction during the
hydrolysis cycle (Table 1). Superimposing the biochemi- last half century shows that even when a set of ideas
cal cycle with a schematic structural model for the seems to be well established, there is a large chance
hydrolysis cycle (Fig. 12) gives the current thinking of that it will be overthrown by some unexpected dis-
how muscular contraction might occur. covery.
Starting from the top left drawing in Fig. 12, and fol-
lowing the arrows (the likely path of the actin–myosin With this quote in mind, we should continuously ask
hydrolysis cycle), we go through the following steps. ourselves, what are the limitations of the current thinking
Table 1
Actin–myosin hydrolysis cycle (rabbit skeletal muscle). (Adapted from Howard [12], with permission)
M, myosin; T, ATP; D, ADP; P, phosphate; A, actin. The most commonly followed path is shown in bold type. In parentheses are the approximate
free energies of the states assuming [ATP]=4 mM, [Pi]=2 mM, [ADP]=20 µM. The actin concentration is taken to equal 1 mM, its concentration
in skeletal muscle (this is fairly arbitrary because the effective concentration in a muscle depends on the accessibility of the actin site to the myosin
head). Overall ATPase: kcat=25 s⫺1, KM(ATP)=10 µM, KM(actin)⬵100 µM (low ionic strength) at 20 °C [2]. The rate constants are based on
experimental data taken from isolated S1 and/or myofibrils under roughly physiological ionic strength. The values vary by up to a factor of 10
between laboratories. The constants (K) are dissociation constants and k+(k⫺) is the rate constant in the clockwise (counterclockwise) direction
(from Howard [12], with permission).
W. Herzog, R. Ait-Haddou / Journal of Electromyography and Kinesiology 12 (2002) 425–433 431
Fig. 12. Schematic illustration of the current thinking of the cross- bridge cycle, including a best estimate of the rate constants of the reactions
(adapted from Howard [12], with permission).
on muscle contraction, and is the set of well established cannot be readily explained by a power stroke model of
ideas still the best in view of newly emerging experi- muscle contraction.
mental results? As a microscopic molecular motor, the myosin head
One of the things that has troubled us is the asym- operates at energy levels that are just a little greater than
metry of the rate constants in a seemingly perfectly sym- the random thermal energy of the surroundings. These
metrical model in the 1957 theory (Figs. 4 and 5). Of thermal baths are essential for the process of the chemi-
course, the asymmetry was required, otherwise there cal reactions. The myosin head is exposed to these ther-
would have been no net force and no net sliding of the mal forces that arise from collision with molecules in
actin relative to the myosin filament. However, no reason the surrounding fluid. Because of the random occurrence
was provided why, from a biological point of view, the of these collisions, in terms of direction and magnitude,
rate constants for attachment and detachment of the the cross-bridge head will undergo a random diffusion,
cross-bridges were asymmetrical. Although, in more or Brownian motion. In the detached state, the myosin
recent models, structural arguments for this asymmetry head moves more or less freely, and therefore, undergoes
have been proposed, none seems terribly convincing. a free Brownian motion. In the attached state, the ran-
Other questions arise from experiments on single mol- dom diffusion is limited by the force field of the actin
ecules. For example, if contraction occurs as sche- filament. The transition between the two diffusive modes
matically proposed in Figs. 11 and 12, one would expect is driven by the chemical reaction of ATP hydrolysis.
that the relative sliding of actin and myosin during a Therefore, the movement of the cross-bridge head has
single powerstroke could be calculated from the angular to be stochastic in nature and must be described by a
reorientation and the magnitude of the lever arm. How- reaction–diffusion equation. Note that thermal diffusion
ever, it is known that the myosin subfragment 1(S1) has was part of the cross-bridge theory from its conception
a shorter neck than the double-headed HMM construct. [13].
Assuming, as has been done, that the neck region of the At first glance, it may appear that Brownian motion
cross-bridge acts as the lever arm that determines the could be quite disruptive and may greatly affect the
step size of a single power-stroke, one would expect a efficiency of events such as muscular contraction. How-
smaller step size from S1 compared to HMM. However, ever, contrary to this intuitive thinking, it appears that
such a difference was not found [25]. Similarly, Yanag- the myosin head can extract work from the thermal bath
ida [33] found that step sizes remained about the same if the system is far from equilibrium and if there is a
for myosin II molecules with normal neck lengths, and structural asymmetry in the geometry of the system. A
molecules in which the neck region was reduced to 20% system that extracts work under these two conditions is
of its normal length. Furthermore, Yanagida [33,34] known as a thermal ratchet. We will investigate these
found that myosin II had a step reversal in about 10% of systems in much more detail in the following paper on
all steps observed. These results, although preliminary, ratchet-type mechanisms of molecular movement,
432 W. Herzog, R. Ait-Haddou / Journal of Electromyography and Kinesiology 12 (2002) 425–433
specifically as they apply to contraction in skeletal mechanics: piconewton forces and nanometre steps. Nature
muscles. Ratchet-type mechanisms can work without the 1994;368:113–9.
[7] Ford LE, Huxley AF, Simmons RM. Tension responses to sudden
assumption of asymmetry of the rate constants, an length change in stimulated frog muscle fibers near slack length.
assumption that is essential for the power-stroke cross- J Physiol 1977;269:441–515.
bridge model. In the following, we would like to illus- [8] Geeves MA. The dynamics of actin and myosin association and
trate the role of noise in a system that is far from equilib- the crossbridge model of muscle contraction. Biochem J
rium with a simple example. The goal of this example 1991;274:1–14.
[9] Greene LE, Eisenberg E. Dissociation of the actin subfragment
is to show the difference between a system in which 1 complex by adenyl-5⬘-yl imidodiphosphate, ADP, and PPi. J
force is produced by a power stroke and one in which Biol Chem 1980;255:543–8.
the system extracts work from the Brownian motion. [10] Highsmith S. Interactions of the actin and nucleotide binding sites
Imagine a microscopic “being” at a starting point A on myosin subfragment 1. J Biol Chem 1976;251:6170–2.
of a road, who wants to reach a point B further along [11] Hill AV. The heat of shortening and the dynamic constants of
muscle. Proc Royal Soc London 1938;126:136–95.
the road. Assume that the road has two parallel lanes : [12] Howard J. ATP hydrolysis. In: Howard J, editor. Mechanics of
One lane has wind locally, but the average wind along motor proteins and cytoskeleton. Sunderland, MA: Sinauer
the lane is zero. In the second lane, there is no wind. Associates, Inc.; 2001. p. 229–44.
The microscopic “being” is not allowed to walk on the [13] Huxley AF. Muscle structure and theories of contraction. Prog
lane without wind. There are two strategies to reach Biophys Biophys Chem 1957;7:255–318.
[14] Huxley AF. Cross-bridge action: present views, prospects, and
point B. The first strategy consists of walking forcefully unknowns. In: Herzog W, editor. Skeletal muscle mechanics:
in the windy lane. The second strategy consists of stay- from mechanisms to function. West Sussex, UK: John Wiley &
ing in the lane without wind and watching the windy Sons Ltd; 2000. p. 7–31.
lane. Every time the wind is in the required direction, [15] Huxley AF, Niedergerke R. Structural changes in muscle during
the microscopic “being” jumps into the windy lane and contraction. Interference microscopy of living muscle fibres. Nat-
ure 1954;173:971–3.
takes advantage of the direction of the wind. Every time [16] Huxley AF, Simmons RM. Proposed mechanism of force gener-
the wind is in a direction opposite to the one the “being” ation in striated muscle. Nature 1971;233:533–8.
wants to go, our macroscopic friend jumps to the quiet [17] Huxley HE. Electron microscope studies of the organisation of
lane. Loosely, the first strategy is analogous to the power the filaments in striated muscle. Biochim Biophys Acta
stroke theory in which the energy is used directly (“pay 1953;12:387–94.
[18] Huxley HE. The mechanism of muscular contraction. Science
as you go”); and the second strategy is analogous to the 1969;164:1356–66.
ratchet theory in which the energy is used for the switch- [19] Huxley HE, Hansen J. Changes in cross-striations of muscle dur-
ing strategy itself. ing contraction and stretch and their structural implications. Nat-
Ratchet-type mechanisms are thought to be part of ure 1954;173:973–6.
many molecular motors [1]. In the companion paper, we [20] Lohmann K. Ueber die Pyrophatfraktion im Muskel. Naturwis-
sensch 1929;17:624–5.
explore ideas about how a ratchet mechanism may [21] Lundsgaard EU. Untersuchungen über Muskelkontraktionen ohne
explain the workings of skeletal muscle myosin II and Milchsäurebildung. Biochemische Zeitschrift 1930;217:162–77.
actin, i.e. muscular contraction and force production. [22] Ma YZ, Taylor EW. Kinetic mechanism of myofibril ATPase.
Biophys J 1994;66:1542–53.
[23] Margossian SS, Lowey S. Interaction of myosin subfragments
with F-actin. Biochemistry 1978;17:5431–9.
Acknowledgement [24] Millar NC, Geeves MA. The limiting rate of the ATP-mediated
dissociation of actin from rabbit skeletal muscle myosin subfrag-
ment. FEBS Lett 1983;160:141–8.
NSERC of Canada. [25] Molloy JE, Burns JE, Kendrick-Jones J, Tregear RT, White DCS.
Movement and force produced by a single myosin head. Nature
1995;378:209–12.
[26] Needham DM. Myosin and adenosinetriphosphate in relation to
References muscle contraction. Biochim Biophys Acta 1950;4:42–9.
[27] Ramsey RW, Street SF. The isometric length-tension diagram of
[1] Astumian RD, Bier M. Fluctuation driven ratchets: molecular isolated skeletal muscle fibers of the frog. J Cell Comp
motors. Phys Rev Lett 1994;72:1766–9. 1940;15:11–34.
[2] Bagshaw CR. Muscle contraction. New York: Chapman & [28] Rayment I, Holden HM, Whittaker M, Yohn CB, Lorenz M,
Hall, 1993. Holmes KC, Milligan RA. Structure of the actin–myosin complex
[3] Eisenberg E, Moos C. The adenosine triphosphatase activity of and its implications for muscle contraction. Science
acto-heavy meromyosin. A kinetic analysis of actin activation. 1993;261:58–65.
Biochemistry 1968;7:1486–9. [29] Rayment I, Rypniewski WR, Schmidt-Bäse K, Smith R, Tomch-
[4] Engelhardt WA, Ljubimowa MN. Myosine and adenosinetriphos- ick DR, Benning MM, Winkelmann DA, Wesenberg G, Holden
phatase. Nature 1939;144:668. HM. Three-dimensional structure of myosin subfragment-1: a
[5] Epstein M, Herzog W. Theoretical models of skeletal muscle: molecular motor. Science 1993;261:50–8.
biological and mathematical considerations. New York: John [30] Reedy MK, Holmes KC, Tregear RT. Induced changes in orien-
Wiley & Sons Ltd, 1998. tation of the cross-bridges of glycerinated insect flight muscle.
[6] Finer JT, Simmons RM, Spudich JA. Single myosin molecule Nature 1965;207:1276–80.
W. Herzog, R. Ait-Haddou / Journal of Electromyography and Kinesiology 12 (2002) 425–433 433
[31] Siemankowski RF, Wiseman MO, White HD. ADP dissociation Rachid Ait-Haddou received his MSc in
from actomyosin subfragment 1 is sufficiently slow to limit the Applied Mathematics from the University
unloaded shortening velocity in vertebrate muscle. Proc Natl Mohamed V, Rabat, Morocco, in 1991, and his
Acad Sci U S A 1985;82:658–62. PhD in Applied Mathematics from the Univer-
sity Joseph Fourier, Grenoble, France, in 1996.
[32] Szent-Györgyi AG. Contraction of myosin threads. Stud Inst Med
He is currently a postdoctoral scientist wih Dr.
Chem Univ Szeged 1941;1:17–26. Herzog at the University of Calgary. His
[33] Yanagida T. Simultaneous observation of individual ATPase and research interests are in the area of biomolecular
mechanical events by a single myosin molecule during interaction motors and biological transport phenomena,
with actin. Canmore Symposium on Skeletal Muscle; Canmore, mechanisms of muscle contraction, and approxi-
Alberta; 1999. p. 22. mation theory.
[34] Yanagida T. Personal communication; August 2000.
[35] Zahalak GI, Ma SP. Muscle activation and contraction: constitut-
ive relations based directly on cross-bridge kinetics. J Biomech
Eng 1990;112:52–62.