12. Tsien, R. Y. New calcium indicators and buffers with high selectivity
against magnesium and protons: design, synthesis, and properties of prototype structures. Biochemistry 19: 2396–2404, 1980.
13. Tsien, R. Y. Intracellular signal transduction in four dimensions: from molecular design to physiology. Am. J. Physiol. Cell. Physiol. 263: C723–C728, 1992.
14. Tsugorka, A., E. Rios, and L. A. Blatter. Imaging elementary events of calcium release in skeletal muscle cells. Science 269: 1723–1726, 1995.
15. Zhao, M., S. Hollingworth, and S. M. Baylor. Properties of tri- and tetracarboxylate Ca2+ indicators in frog skeletal muscle fibers. Biophys. J. 70:
896–916, 1996.
Sarcomeric Myosin Isoforms: Fine Tuning of a
Molecular Motor
Carlo Reggiani, Roberto Bottinelli, and Ger J. M. Stienen
M
yosin is the best-known molecular motor. Molecular
motors are proteins that are able to convert chemical
energy into mechanical energy (chemomechanical transduction) through a structural change (working stroke). The working
stroke, amplified by the ordered organization of an ensemble
of motors, can give rise to microscopically, or even macroscopically, visible movements of cells or parts of cells.
At present, three molecular motors are known: myosin,
kinesin, and dynein. The molecular motor that satisfies the
energy need for muscular contraction is myosin, or more precisely, class II myosin, which is one of the 15 distinct classes of
myosins presently identified. After many years as the only
known molecular motor, class II myosin still represents the
motor whose fine tuning is best understood. In multicellular
organisms, class II myosins are present not only in muscle cells
but also in platelets, epithelial gastroenteric cells, and neurons.
In unicellular organisms, such as the amoeba and dictyostelium, they play a central role in cytokinesis.
Class II myosins are hexamers composed of two myosin
heavy chains (MHC), two essential myosin light chains (MLC),
and two regulatory MLC (see Fig. 1). Each MHC (molecular
mass 220 kDa) is formed by a globular head and a long tail.
The globular head (~900 amino acids) contains the catalytic
site for ATP hydrolysis and the surface for actin binding. Its
detailed structure has been recently resolved (6) and is shown
in Fig. 1A. The head domain can be divided into three parts by
papain digestion: 25-kDa, 50-kDa, and 20-kDa domains. Flexible loops (loop 1 and loop 2) sensitive to enzymatic digestion
connect the three domains. The 20-kDa domain is largely
formed by a 8.5-nm-long a-helix that protrudes from the head,
forming a shaft around which one essential and one regulatory
MLC are wrapped. This shaft continues, after a sharp bend,
with a very long a-helical rod (150 nm long, ~1000 amino
acids), indicated as the myosin tail. Pairs of MHC dimerize by
C. Reggiani and R. Bottinelli are in the Institute of Human Physiology at the
University of Pavia, I-27100 Pavia, Italy. G. J. M. Stienen is in the Laboratory
for Physiology at the Free University, 1081 BT Amsterdam, the Netherlands.
26
News Physiol. Sci. • Volume 15 • February 2000
forming coiled structures with their tails, which are the backbone of the thick filament.
Thick filaments of sarcomeric or striated muscles are ~1.6 mm
long and are composed of two symmetrical portions, each
containing ~300 myosin molecules. Force and/or shortening
are generated by cyclic interactions of the myosin heads protruding from the thick filament with the adjacent actin thin filaments. Energy is provided by ATP hydrolysis. ATP binds to the
catalytic site on the myosin head and is quickly hydrolyzed.
After ATP hydrolysis, myosin attaches to the actin filament,
undergoes the structural changes (working stroke), thus producing an elementary displacement (Duni) of the thin filament,
and then releases the hydrolysis products (ADP and inorganic
phosphate). Finally, myosin detaches from the actin filament.
The repeated working strokes associated with cyclic attachment and detachment of myosin to actin drive the axial sliding
of the filaments in the direction of shortening. The speed of
shortening can be calculated as V = Duni/ton, where ton is the
time spent attached to actin. The ratio ton/ttotal, where ttotal is the
total duration of the ATP hydrolysis cycle, is called the duty
ratio and corresponds to the fraction of myosin heads attached
at each instant to the thin filament.
Myosin isoforms
The range of rates at which sarcomeric myosins are able to
convert chemical energy into work is surprisingly large when
various animal species and, in each species, different types of
muscles are compared. This is related to the functions that are
accomplished by striated muscles, from maintaining posture to
moving the body in space, from chewing food to pumping
blood. The requirements of these functions in terms of mechanical power are very different within each animal species and
vary further from species to species.
The secret of the extraordinary ability of the myosin motor to
comply with very different functional requirements resides in the
existence of multiple isoforms of both MHC and MLC. Isoforms
are proteins very similar to each other and able to replace each
0886-1714/99 5.00 © 2000 Int. Union Physiol. Sci./Am.Physiol. Soc.
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.5 on June 18, 2017
Sarcomeric or striated muscle myosins are the molecular motors whose fine tuning is best known.
Sarcomeric myosin isoforms convert chemical into mechanical energy at very different rates
without losing efficiency. Availability of the amino acid sequences offers for the first time the
chance to understand the molecular basis of the versatility of these molecular motors.
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.5 on June 18, 2017
FIGURE 1. Structure of myosin molecule and location of clusters of molecular diversity. A: tertiary structure of S1 myosin fragment with a schematic reference to
complete molecule and to filament lattice. Functionally relevant regions are indicated. Essential and regulatory myosin light chains (MLC) are blue. Myosin heavy
chain (MHC) is drawn with a smooth spectrum of colors from blue (NH2 terminus and the 25-kDa domain) to green and yellow (50-kDa domain) and red (20-kDa
domain—including the shaft—and COOH terminus). Figure was prepared using Rasmol and coordinates of myosin S1 from Ref. 6. B: schematic representation
of primary structure of complete MHC molecule, with indication of functionally relevant sequences and, below, localizations of regions where diversity among
isoforms tends to cluster (black bands) in paired comparison between MHCa and MHCb/slow, from Ref. 15.
other. Slight diversity in their amino acid sequences gives rise to
different structural and functional properties. In the genoma of
each species, a number of genes codify for MHC and MLC isoforms; further variations are produced during mRNA maturation
by alternative splicing (for review, see Ref. 8).
Information about the sarcomeric muscle myosin isoforms and
their relationship with contractile muscle performance are available for several species, from invertebrates (Drosophila) to fishes
(carp, cod), amphibians (toad, frog), birds (chicken), and several
mammals (mouse, rabbit, rat, human, horse). However, the
News Physiol. Sci. • Volume 15 • February 2000
27
“The expression of MHC and MLC genes during
development and adult life is controlled by
several factors....”
three fast isoforms expressed in fast skeletal muscle fibers;
MHCexoc, a fast isoform expressed only in extraocular muscles;
and MHCemb and MHCneo, which are two isoforms expressed
during embryonic and perinatal development, respectively. As a
general rule, in fast muscle fibers a fast MHC isoform associates
with a fast regulatory isoform (MLC2f) and with one of the two
fast essential MLC isoforms (MLC1f or MLC3f). Interestingly,
these two isoforms originate from the same gene through a
process of alternative splicing. In slow fibers, the slow MHCb
binds a slow regulatory isoform (MLC2s) and a slow essential
MLC, either MLC1sa or MLC1sb. In atrial myocardium, a specific atrial isoform of regulatory MLC (MLC2a) and a developmental form of essential MLC (MLC1a-emb) are associated with
MHCa. Finally, in ventricular myocardium, MHCb or MHCa is
expressed together with the slow isoforms MLC2s and MLC1sb.
Unfortunately, complete sequences are available only for
few sarcomeric MHC isoforms. For many other MHC isoforms
only portions of the sequences are known. On the basis of
available data, the degree of homology between corresponding MHC isoforms in different species (orthologous MHC isoforms) is between 90 and 95%; in other words, only 95–190
out of 1900 amino acids are different. If only nonconservative
changes, i.e., replacements of one amino acid with another
one belonging to a different group, are considered, the homology arrives at 98–99%; only 20–30 amino acids are different.
The degree of homology between different MHC isoforms in
one species (paralogous MHC isoforms) is lower; the percentage of nonidentical amino acids can reach 20% (15). The
percentages of nonidentical amino acids are not substantially
different in the tail and in the head domain. Most of the nonidentical amino acids are clustered in few regions (15): the two
flexible loops, the MLC binding regions, the S2 subfragment,
28
News Physiol. Sci. • Volume 15 • February 2000
the hinge region, and the tail (Fig. 1). Some other regions, like
the ATP-binding pocket or the actin-binding surface, are virtually identical in all isoforms. Particularly interesting are the two
flexible loops connecting the 25-kDa and 50-kDa domains
and the 50-kDa and 20-kDa domains, respectively. The lack of
conservation in the two loops might simply reflect lack of functional constraints. It is, however, also possible that these two
loops are capable of functional modulation (see below). Loop 1
is located near the entrance of the ATP-binding site, and
loop 2 is located on the actin-binding surface (see Fig. 1A).
The expression of MHC and MLC genes during development
and adult life is controlled by several factors, including
motoneuron discharge patterns, loading conditions, and hormones. Although coexpression of two or more MHC genes is
possible and even frequent during development or changes in
activity patterns, it is a general rule that each cell expresses only
one MHC gene associated with two MLC genes (8). This creates
a heterogeneous population of muscle fibers, each containing a
specific combination of MHC isoforms and MLC isoforms.
These conditions offer us the chance to study the chemomechanical transduction features of various isoforms by simply
analyzing the functional behavior of individual muscle fibers.
Single-fiber studies
In the last decade, several studies have analyzed the correlations between myosin isoform composition and functional properties of mammalian skeletal muscle fibers (for review, see
Ref. 8). The novelty of these studies has been represented by the
determination of the mechanical and energetic parameters
related to chemomechanical energy transduction and of the protein isoform composition in single muscle fibers. This has been
made possible by a more powerful resolution of electrophoretical and immunological identification of protein isoforms and by
the technical improvement of functional characterization of
demembranated fibers, i.e., fibers in which sarcolemma and
other membranes are solubilized or removed and only the
myofibrils with the contractile proteins are functionally intact.
In muscle fibers, the conversion of chemical into mechanical energy is defined by 1) the amount of chemical energy
released (G), which is obtained by multiplying the number of
ATP moles consumed by the energy released from each mole
(~50 kJ/mol, under the conditions present in muscle cells) and
2) the work generated as measured from the product of force
and movement. Only part of the chemical energy can be actually converted into work (W); another part is converted into
heat (H) and a third part is absorbed by the reaction itself. The
efficiency of the chemomechanical transduction is defined as
thermodynamic efficiency = W/G or as mechanical efficiency
= W/(W+H). Instead of work and the amount of energy
released, their respective time derivatives are more often used
because they are direct expressions of the instantaneous properties of the motor function: power (W’), rate of energy
released (E’), estimated from the rate of ATP hydrolysis multiplied by the ATP energy equivalent, and the sum of W’ and the
rate of heat production (H’).
The power output (W’) of a muscle shortening against a load
equals the product of force developed and velocity of shortening.
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.5 on June 18, 2017
largest body of information presently available on myosin isoforms and their properties related to chemomechanical transduction concerns sarcomeric myosins of mammals and particularly
of the rat, rabbit, and human. In these species, the sequences of
large portions and, in some cases, of whole MHC and MLC genes
have been provided by clone isolation and sequencing (for
review, see Ref. 15), whereas detailed functional characterization
of biochemical, energetic, and mechanical parameters have been
obtained from studies on single-fiber preparations and on purified myosin preparations (for review, see Ref. 8).
In the above-mentioned mammals, at least eight distinct isoforms of class II MHC encoded by eight separate genes are
expressed in sarcomeric or striated muscle cells. They combine
with several MLC isoforms to generate myosin molecule isoforms (isomyosins) (8). The eight known sarcomeric MHCs are as
follows: MHC1 or MHCb/slow, expressed in ventricular
myocardium and in slow skeletal muscle fibers; MHCa,
expressed in atrial myocardium and occasionally in skeletal
muscle fibers; MHC2A, MHC2X, and MHC2B, which are the
News Physiol. Sci. • Volume 15 • February 2000
29
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.5 on June 18, 2017
FIGURE 2. Mechanical and energetic parameters of 2 sets of rat fibers containing distinct MHC isoforms. Closed circles refer to a set of fibers containing
MHC2B. Open circles refer to a set of fibers containing MHC1 or b/slow.
Force (A), mechanical power (in Watts per liter) (B), total energy liberation rate
expressed relative to isometric value (C), and thermodynamic efficiency (D)
are expressed as functions of shortening velocity and are determined in
demembranated preparations during maximal activation (pCa = 4.6) at optimal sarcomere length (see Table 1), pH 7, ionic strength 0.2 M, and 12ºC
(modified from Ref. 7). Note the change in abscissa from A and B to C and D.
Examples of force-velocity curves of two fiber populations, one
expressing the slow MHCb and the other expressing the fast
MHC2B, are shown in Fig. 2A. Maximum shortening velocity
(V0), i.e., the shortening velocity at zero load, is about three
times higher in the fast than in the slow fibers. Power-velocity
curves are shown in Fig. 2B. The peak power, W’max, is reached
at a velocity (Vopt) approximately corresponding to 1/3 of V0.
Figure 2C shows the rates of energy release as a function of
shortening velocity. During shortening against a load, ATP consumption increases in direct relation to velocity of shortening.
The increase is closely proportional to the rate of ATP consumption during isometric contraction (Fig. 2C), being therefore much higher in muscle fibers that contain fast MHC than
in muscle fibers that contain slow MHC isoforms (7). However,
because fast fibers also develop much higher power than slow
fibers (see Fig. 2B), thermodynamic efficiency is only slightly
greater for slow than for fast fibers (see Fig. 2D). The optimum
efficiency is reached at approximately the same speed of shortening as power (Vopt). Vopt varies widely among muscle fibers
and depends on MHC isoforms. Recruitment of muscle fibers
containing distinct MHC isoforms enables muscles to optimize
power production and efficiency when movements are performed at different speeds. In other words, muscles in vivo
optimize their power and efficiency by shortening at Vopt. Thus
Vopt is possibly one of the most physiologically relevant contractile parameters of muscle fibers.
In Table 1, mechanical and energetic parameters of muscle
fibers from three mammalian species are reported. Fibers are
identified on the basis of their MHC isoform composition.
Data come from previously published work of our group and
refer to demembranated fiber segments maximally activated
at 12ºC. The MLC isoforms that are most frequently associated
with the MHC are indicated in some cases. The variations in
contractile and biochemical properties among muscle fibers
are extremely large, both comparing various isoforms in the
same species (paralogous) and corresponding isoforms in different species (orthologous). MHC isoform diversity generates
large variations in V0 (2, 3, 11), W’max, Vopt, rate of ATP splitting in isometric conditions (A0) (4, 5, 9), and in the rate of
tension rise (not shown). On the contrary, the variability in
isometric tension (P0) is only loosely related to the presence of
different MHC isoforms. Because ATP consumption in isometric conditions is substantially higher in muscle fibers containing fast MHC isoforms than in muscle fibers containing
the slow MHC (4, 9), the amount of ATP needed to maintain
a given amount of tension (tension cost) is much higher in fast
than in slow muscle fibers.
MLC isoforms also contribute significantly to modulating
motor properties. Fast MHC isoforms are associated with two
essential MLC isoforms: MLC1f and MLC3f. In the rat and the
rabbit, the relative proportion of these two isoforms appears to
significantly affect V0 through a complex interplay with MHC
isoforms: V0 increases in proportion to the amount of MLC3f
present (2, 11). ATPase activity and other parameters are not
affected (4). Furthermore, MLCs provide the only myosinbased posttranslational regulatory mechanism identified until
now: phosphorylation of regulatory MLC enhances force
development of submaximally activated muscle fibers.
Table 1.
Mechanical and energetic parameters determined in maximally activated demembranated skeletal muscle fibers
containing various myosin isoforms in three different species
P0, kN/m2
A0, mM/s
Tension Cost
W’max, W/l
1.13 ± 0.12
2.46 ± 0.29
3.32 ± 0.41
3.56 ± 0.22
68 ± 4
111 ± 15
95 ± 11
99 ± 10
45 ± 6
168 ± 26
178 ± 23
230 ± 10
0.66 ± 0.10
1.52 ± 0.13
1.89 ± 0.22
2.30 ± 0.11
1.96 ± 0.18
0.60 ± 0.07
2.83 ± 0.28
85 ± 15
106 ± 12
42 ± 3
323 ± 26
0.49 ± 0.2
3.04 ± 0.65
0.26 ± 0.18
1.12 ± 0.08
2.42 ± 0.06
64 ± 4
84 ± 3
98 ± 5
43 ± 4
107 ± 7
166 ± 12
0.79 ± 0.13
1.25 ± 0.06
1.68 ± 0.10
9.21 ± 0.75
0.34 ± 0.04
1.62 ± 0.09
3.03 ± 0.24
All parameters were determined under same conditions [12ºC, optimal sarcomere length (rat, 2.7 mm; rabbit, 2.4 mm; human,
2.5 mm), ionic strength 0.2 M, pH 7]. Fibers are grouped on the basis of their MHC isoform composition. When a strict association is known to exist, MLC isoforms are also indicated. Human fibers containing MHC2X are indicated as MHC2B/X because
they are generally known as human 2B fibers. V0, maximum (unloaded) shortening velocity, expressed in length per second. P0,
isometric tension, measured in kilonewtons per square meter. A0, ATP hydrolysis rate in isometric conditions, measured in micromoles per second. Rate of ATP hydrolysis per myosin head per second can be easily obtained assuming a concentration of
myosin head of 200 mM. Tension cost is equal to A0/P0 and is expressed in micromoles per second per kilonewton per square
meter. W’max, maximal power output measured in Watts per liter (W/l). Data for rat fibers come from Refs. 4 and 7; data for rabbit fibers come from Ref. 5 with a correction for temperature; data for human fibers come from Refs. 3 and 9.
Inspection of Table 1 shows that orthologous isoforms in
different species exhibit different values of the kinetic parameters. As a general rule, the values of the kinetic parameters
are higher in the rat than in the rabbit and in the human. This
has offered a new and definitive explanation of a classic
observation of A. V. Hill on limb muscle performance of different animal species. To achieve a comparable speed of
locomotion, small animals need faster muscles than large
animals. This goal can be reached in two alternative ways,
either with a high proportion of fast fibers or with a higher
speed of shortening of both fast and slow fibers. The precise
identification of MHC isoforms has shown that both mechanisms cooperate to create the difference in speed of shortening between muscles of different animal species. For example,
mouse muscles are richer in fast MHC2B than corresponding
muscles of rat, and these in turn express fast MHC2B more
than corresponding muscles of the rabbit. This fast and
expensive MHC isoform is virtually not expressed in human
muscles (8). Moreover, it is true that orthologous MHC have
different functional properties in different species. Each MHC
isoform is associated with a higher speed of shortening and
power output in small animals than in large animals (Table 1
and Fig. 3A). The same holds true for energetic properties:
each MHC isoform hydrolyzes more ATP and consumes
more energy for tension development in small than in large
animals. This fits well with the high general metabolic activity typical of small mammals and their large body surface-tovolume ratio, which promotes heat loss.
30
News Physiol. Sci. • Volume 15 • February 2000
Shortening velocity and ATP hydrolysis rate appear to vary in
proportion, comparing both paralogous MHC isoforms in one
species and orthologous MHC isoforms in different species. In
his influential study, Barany (1) showed, 30 years ago, that in
16 different muscles of various animal species there was a correlation between shortening velocity and ATPase activity of
purified myosin activated by actin or by calcium over a range
of variation of 200 times. The recent data obtained in single
fibers confirm that a correlation exists between V0 and ATP utilization rate during isometric contraction (A0; see Fig. 3B). In
the single-fiber studies, both kinetic parameters are measured
on the same preparations under identical experimental conditions. Furthermore, molecular characterization allows precise
definition of the fiber type, i.e., which MHC isoform is present.
The general correlation between A0 and V0 is in accordance
with Barany’s findings, but some interesting deviations occur.
For example, the orthologous b/slow or type 1 MHC isoforms
in rat, rabbit, and human, respectively, are associated with different speeds of shortening but similar isometric ATPase activity values. Among fast fibers, fibers containing MHC2X and
MHC2B of rat and rabbit have similar shortening velocities but
differ significantly in ATP-splitting rate. Interestingly, the correlation, which can be seen when groups formed with fibers all
containing the same MHC isoforms are compared (Fig. 3B),
does not always show up when individual fibers within each
group are considered. For example, in a population of rat fibers
containing MHC2B, V0 and A0 are not correlated. This can be
explained by assuming that MLC exerts a powerful influence
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.5 on June 18, 2017
Rat
MHC1+MLC1s+MLC2s
MHC2A+MLC1f/3f+MLC2f
MHC2X+MLC1f/3f+MLC2f
MHC2B+MLC1f/3f+MLC2f
Rabbit
MHC1+MLC1s+MLC2s
MHC2X+MLC1f/3f+MLC2f
Human
MHC1
MHC2A
MHC2X/B
V0, L/s
MHCb/slow purified from rat myocardium indicates that the
size of the unitary displacement (9–10 nm) as well as of unitary force (~2 pN) is similar for both isoforms. The durations of
both force and movement events are about four times longer
for MHCb/slow than for MHCa. Interestingly, for both isoforms
the duration of the displacement events is only 60% of that of
force events. This suggests that the two isoforms also share a
similar strain sensitivity (or load sensitivity) of the kinetic rate
constants for attachment to and detachment from actin.
Because force developed in elementary actin-myosin interaction is similar, as is the force developed by maximally activated
skinned preparations containing either MHCa or MHCb/slow,
no difference in the duty ratio (or in the instantaneous fraction
of attached cross bridges) can be expected.
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.5 on June 18, 2017
on V0 that is not accompanied by a corresponding effect on A0
(4). Finally, a dissociation between V0 and A0 appears to occur
when temperature is changed, because V0 shows a higher thermal sensitivity than A0 (3, 9). It is worth noting that the correlation between V0 and A0 is consistent with the observation
that efficiency is virtually constant. Actually, the slope of the
correlation between V0 and A0 (or the ratio V0/A0) is comparable to efficiency. Efficiency is the ratio between power output
and energy released by ATP hydrolysis during loaded shortening. As discussed above, ATPase rate during shortening
increases in close proportion to its value in isometric conditions (A0), and V0 is a major determinant of W’max.
The picture emerging from the results obtained in skinned
fiber studies points to an important conclusion. The diversity
between myosin isoforms produces large variations in several
kinetic parameters, such as V0, W’max, and tension cost,
whereas other parameters, such as efficiency, ratio of V0 to A0,
and P0 show much smaller variations. This can be explained
by assuming that all sarcomeric myosins share a common
conserved basic mechanism of chemomechanical transduction on which the mechanisms of modulation of the kinetic
properties are superimposed.
New insights with in vitro motility assay
Single-fiber studies find their limits in that they cannot
unambiguously distinguish whether the diversity of parameters
such as V0 or W’max arise from differences in rate constants of
the actomyosin interaction cycle or from differences in the elementary mechanical events. For example, V0 is equal to Duni/ton
under conditions of zero load. As stated above, Duni is the elementary displacement produced by myosin when attached to
actin and ton is the time spent by a myosin in the attached state.
Mechanical analysis of single fibers cannot distinguish whether
the variations in V0 are caused by variations of Duni or ton. The
in vitro motility assay has recently offered an exciting approach
to extend the knowledge of functional differences between
myosin isoforms to the molecular level (14).
Purified myosin isoforms translocate actin filaments at
speeds proportional to their actin-activated ATPase activity.
Mixtures of two myosin isoforms move actin filaments at
speeds intermediately between those produced by the faster
and the slower myosin alone (see Ref. 14). Analysis of the
motion shows that the two myosins in the mixture interact
mechanically, and their behavior can be predicted either by
simple force-velocity analysis or by the use of a kinetic model
of the actomyosin interaction. Not only MHC isoforms but also
essential MLC isoforms are able to influence speed of actin
translocation. In particular, the speed of actin movement
increases in relation to the amount of MLC3f, in agreement
with the results obtained in muscle fibers (reviewed in Ref. 14).
A major step toward understanding whether the diversity
between myosin isoforms must be attributed to the rate constants or to the elementary mechanical events has been provided by the optical laser traps. With this new technique, the
first measurements of elementary mechanical events of single myosin molecules were obtained. Recent comparison
(10) of unitary force and displacements between MHCa and
FIGURE 3. Maximum shortening velocity (V0) and rate of ATP consumption in
isometric contraction (A0) in several sets of muscle fibers containing paralogous and orthologous MHC isoforms. Two parameters are determined under
conditions reported in Fig. 2 and Table 1. A: V0 is plotted vs. average size of
animal species to which fibers belong. A negative correlation exists between
maximum speed of shortening and animal size for orthologous MHC variants
(connected by lines). B: A0 is plotted vs. V0. Lines connect values obtained in
paralogous MHC. A general trend toward positive correlation is visible. L/s,
length per second. Data come from Refs. 2, 3, 4, 7, and 9.
News Physiol. Sci. • Volume 15 • February 2000
31
A clear conclusion seems to emerge from the results obtained
in experiments on single myosin molecules. In fast and slow
MHC isoforms, the size of the elementary mechanical events
(both of force and displacement) are similar, and the diversity
resides in attachment and/or detachment rate constants. This conclusion is also in agreement with the results obtained by comparing skeletal muscle myosin with smooth muscle myosin (14).
Molecular diversity between myosin isoforms
“The road to achieving a full understanding of
the structure-function relationships of molecular
motors is still long.....”
ventional myosins. The actin-activated ATPase activities of the
chimeric myosins were found to correspond to those of the
donor myosins. The role of loop 1 is still more uncertain; in
smooth muscle, an insert of seven amino acids in loop 1 seems
sufficient to influence speed of shortening. In a very recent study
(12), loop 1 of smooth muscle myosin was replaced with the
corresponding sequences of a number of other myosins (including three sarcomeric MHC). The changes of length and charge in
loop 1 were found to modify the rates of ADP release, ATPase
activity, and actin filament translation in vitro. However, no clear
evidence of a direct relation between loop 1 features and kinetic
properties of donor myosin were found. This suggests that loop
1 alone cannot determine any of the kinetic properties of the
myosin. The diversity between isoforms is not restricted to the
two loops but is also present in other parts of the molecule. For
example, it is known that clusters of amino acid diversity are also
located in the myosin tail (see Fig. 1B). No information, however, is available on their possible functional impact.
Although most of the work aimed at establishing the molecular basis of the functional differences among isoforms is
focused on MHCs, the observation that essential MLC isoforms
also affect V0 should not be overlooked. The two essential
MLCs only differ in the NH2-terminal region. Compared with
MLC3f, MLC1f has a long NH2-terminal extension that allows
contact with actin. Through this contact with actin, MLC1f
might regulate V0 by strengthening actomyosin interaction. In
agreement with this, S1-MLC1f shows a greater affinity for
actin than S1-MLC3f. Alternately, it has been suggested that the
MLC isoforms that are wrapped around the a-helical shaft
extending from the globular portion of S1 might modulate its
flexibility (for further discussion, see Ref. 8).
32
News Physiol. Sci. • Volume 15 • February 2000
Skeletal muscle fibers have their cytosol filled with myosin
molecules (almost 50% of the muscle fiber proteins) arranged
in a highly ordered way along the thick filaments in the
myofibrils. For this reason, they offer a unique opportunity for
investigating the operation of molecular motors by studying
energetic and mechanical properties of the whole cell.
The study of the relationship between myosin variants and
diversity of energetic and mechanical performance has found
a convenient model in demembranated segments of skeletal
muscle fibers. Limb muscles of rabbit, rat, mouse, and human
have been carefully studied. The results have provided a
rather complete picture of how mechanical parameters, such
as shortening velocity, power, and energy release rate, vary
largely in relation to myosin and particularly MHC isoforms.
Other parameters, such as efficiency of chemomechanical
conversion, remain relatively constant. More recently, the
methods of molecular mechanics have been applied to isolated myosin motors. The first results indicate that isoform
diversity influences the kinetics but not the amplitude of the
elementary force and movement events. In parallel, the
understanding of molecular diversity between myosin isoforms is growing. The sequences of MHC and MLC isoforms
have become available and have shown that nonidentical
amino acids tend to cluster in defined regions of the molecule. These regions are likely endowed with regulatory properties. Further support to the identification of the functionally
relevant amino acid sequences has been provided by mutational analysis.
The road to achieving a full understanding of the structurefunction relationships of molecular motors is still long. Important steps forward can be made by determining complete
amino acid sequences of many orthologous and paralogous
MHC isoforms and by analyzing the correlations with functional properties of muscle fibers and isolated myosin molecules. Actually, for many myosin variants even in the restricted
field of the mammalian muscles, little or no functional data are
available. No less important, however, is the solution to many
still-open issues regarding the basic force generation mechanism, such as the communication between the ATP-binding
site and the actin-binding surface, the changes in myosin head
configuration during the power stroke, and the coordination
between the two myosin heads.
This work was supported by the European Union (Human Capital and
Mobility Grant ERBCHRXCT 940606) and by MURST-Italy (Cofin-1998). The
authors regret that many important contributions could not be cited due to
space restrictions.
REFERENCES
1. Barany, M. ATPase activity of myosin correlated with speed of muscle
shortening. J. Gen. Physiol. 50: 197–216, 1967.
2. Bottinelli, R., R. Betto, S. Schiaffino, and C. Reggiani. Unloaded shortening velocity and myosin heavy chain and light chain isoform composition
in rat skeletal muscle fibres. J. Physiol. (Lond.) 478: 341–349, 1994.
3. Bottinelli, R., M. Canepari, M. A. Pellegrino, and C. Reggiani. Force-velocity properties of human skeletal muscle fibres: myosin heavy chain isoform
and temperature dependence. J. Physiol. (Lond.) 495: 573–586, 1996.
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.5 on June 18, 2017
The molecular basis of the kinetic diversity must be localized
in those amino acids that form the structural difference between
isoforms. As stated above, a large fraction of the nonidentical
amino acids is clustered in few regions. Among them are the
two flexible loops of the MHC head. The hypothesis that these
two loops represent hypervariable regions that regulate ATPase
activity and shortening velocity was introduced a few years ago.
The role of loop 2 in regulating ATPase activity has received
support from the experimental work of Uyeda et al. (13). Loop 2
of dictyostelium myosin II was replaced with loops of other con-
Conclusion and future perspectives
10. Sugiura, S., N. Kobayakawa, H. Fujita, H. Yamashita, S. Momomura, S.
Chaen, M. Omata, and H. Sugi. Comparison of unitary displacements and
forces between two cardiac myosin isoforms by the optical trap technique.
Circ. Res. 82: 1029–1034, 1998.
11. Sweeney, H. L., M. J. Kushmerick, K. Mabuchi, F. A. Sreter, and J. Gergely.
Myosin alkali light chain and heavy chain variations correlate with altered
shortening velocity of isolated skeletal muscle fibres. J. Biol. Chem. 263:
9034–9039, 1988.
12. Sweeney, H. L., S. S. Rosenfeld, F. Brown, L. Faust, J. Smith, J. Xing,
L. A. Stein, and J. R. Sellers. Kinetic tuning of myosin via flexible
loop adjacent to the nucleotide binding pocket. J. Biol. Chem. 273:
6262–6270, 1998.
13. Uyeda, T. Q. P., K. M. Ruppel, and J. A. Spudich. Enzymatic activities correlate with chimaeric substitutions at the actin-binding face of myosin.
Nature 368: 567–569, 1994.
14. Warshaw, D. M. The in vitro motility assay: a window into the myosin
molecular motor. News Physiol. Sci. 11: 1–7, 1996.
15. Weiss, A., and L. Leiwand. The mammalian myosin heavy chain gene family. Annu. Rev. Cell Dev. Biol. 12: 417–439, 1996.
Images of Action Potential Propagation in Heart
Guy Salama and Bum-Rak Choi
Activation and repolarization across mammalian hearts follow complex three-dimensional pathways
that are governed by fiber structure, intercellular coupling, and action potentials (APs) with spatially
heterogeneous properties. Voltage-sensitive dyes and imaging techniques offer new insights on
how spatiotemporal heterogeneities of APs govern propagation, repolarization, and AV node
conduction and help us visualize arrhythmias with previously unattainable details.
T
he electrical potential across the cell membrane plays a
pivotal role in cellular function. In excitable cells, rapid membrane potential changes are the signaling mechanism for cell-tocell communication and the trigger for important cellular
functions, as in excitation-contraction or excitation-secretion coupling. Our knowledge of membrane potentials is based primarily
on impalements of cells with glass pipette microelectrodes. In
many experimental conditions, the difficulties in maintaining the
microelectrode tip in the intracellular space and damage to the
cells limit the use of the method. Microelectrode techniques are
not applicable to small cells (<3 µm in diameter) nor are they
practical in situations requiring the simultaneous monitoring of
membrane potential from multiple neighboring cells. Hence,
attempts were made in the late 1960s to develop techniques that
transduce membrane potential to an optical signal.
Voltage-sensitive dyes
Initial attempts to measure voltage-dependent optical signals
in biological systems were based on intrinsic signals (light scattering, birefringence) from isolated nerve cells and skeletal muscle fibers. Intrinsic signals appeared to mimic the time course of
the action potential (AP), but their signal-to-noise ratio (S/N) was
very poor. The next step was to stain excitable cells with a dye
G. Salama and B.-R. Choi are in the Department of Cell Biology and Physiology at the University of Pittsburgh School of Medicine, 3500 Terrace
Street, Pittsburgh, PA 15261.
0886-1714/99 5.00 © 2000 Int. Union Physiol. Sci./Am.Physiol. Soc.
and show that extrinsic optical signals (changes in dye absorption and/or fluorescence) followed changes in membrane potential. Extrinsic signals provided a major improvement in S/N and
bolstered the search for voltage-sensitive dyes (10).
The first measurements of optical APs from heart muscle
were obtained with merocyanine 540 (M-540) (13). Dyes with
larger voltage-dependent responses and less phototoxic damage have now replaced M-540. Figure 1A illustrates an instrument to measure light from a 1-mm light spot focused on a
sheet of ventricular muscle. Electrical stimulation of the
unstained preparation produces a change in light scattering
(∆I/I) due to the muscle contraction (Fig. 1B, top). After staining the muscle with M-540, the stimulus produced a rapid
increase in fluorescence followed by a slower response caused
by the movement artifact (Fig. 1B, middle). After bathing the
muscle in a Ca2+-free solution to suppress contractions, the fluorescence signal (∆F/F) had the time course of the cardiac AP
(Fig. 1B, bottom). The fluorescence intensity depends on the
amount of bound dye (after washout of excess dye), the optical
apparatus (i.e., intensity of excitation, filters, and so forth), and
the loss of dye over time. In general, “fast” dyes bind tightly to
the myocyte membrane and do not seem to be internalized in
the cytosol, making it possible to record optical APs for 2–4 h.
Their spectral properties depend on the local environment, and
only dye molecules bound to excitable membranes can transduce potential changes to an optical response. The larger the
fractional absorption and/or fluorescence (∆F/F) change per AP
the greater the voltage “sensitivity” of the dye. These dyes
News Physiol. Sci. • Volume 15 • February 2000
33
Downloaded from http://physiologyonline.physiology.org/ by 10.220.33.5 on June 18, 2017
4. Bottinelli, R., M. Canepari, C. Reggiani, and G. J. M. Stienen. Myofibrillar
ATPase activity during isometric contraction and isomyosin composition in
rat single skinned muscle fibres. J. Physiol. (Lond.) 481: 663–675, 1994.
5. Potma, E., I. A. Van Graas, and G. J. M. Stienen. Influence of inorganic
phosphate and pH on ATP utilization in fast and slow skeletal muscle
fibres. Biophys. J. 69: 2580–2589, 1995.
6. Rayment, I., W. R. Rypniewski, K. Schmidt-Base, R. Smith, D. R. Tomchick,
M. M. Benning, D. A. Winkelmann, G. Wesenberg, and H. M. Holden.
Three dimensional structure of myosin subfragment-1: a molecular motor.
Science 261: 50–58, 1993.
7. Reggiani, C., E. J. Potma, R. Bottinelli, M. Canepari, M. A. Pellegrino, and
G. J. M. Stienen. Chemo-mechanical energy transduction in relation to
myosin isoform composition in skeletal muscle fibres of the rat. J. Physiol.
(Lond.) 502: 449–460, 1997.
8. Schiaffino, S., and C. Reggiani. Molecular diversity of myofibrillar proteins:
gene regulation and functional significance. Physiol. Rev. 76: 371–423, 1996.
9. Stienen, G. J. M., J. L. Kiers, R. Bottinelli, and C. Reggiani. Myofibrillar
ATPase activity in skinned human skeletal muscle fibres: fibre type and
temperature dependence. J. Physiol. (Lond.) 493: 299–307, 1996.