Am J Physiol Cell Physiol 282: C835–C844, 2002.
First published December 5, 2001; 10.1152/ajpcell.00482.2001.
Influence of myosin isoforms on contractile
properties of intact muscle fibers from Rana pipiens
GORDON J. LUTZ, SHASHANK R. SIRSI, SARAH A. SHAPARD-PALMER,
SHANNON N. BREMNER, AND RICHARD L. LIEBER
Biomedical Sciences Graduate Group, Departments of Orthopaedics and
Bioengineering, University of California, Veterans Affairs Medical Center,
and Veterans Medical Research Foundation, San Diego, California 92161
Received 10 October 2001; accepted in final form 14 November 2001
Lutz, Gordon J., Shashank R. Sirsi, Sarah A. Shapard-Palmer, Shannon N. Bremner, and Richard L.
Lieber. Influence of myosin isoforms on contractile properties of intact muscle fibers from Rana pipiens. Am J Physiol
Cell Physiol 282: C835–C844, 2002. First published December 5, 2001; 10.1152/ajpcell.00482.2001.—The myosin heavy
chain (MHC) and myosin light chain (MLC) isoforms in
skeletal muscle of Rana pipiens have been well characterized. We measured the force-velocity (F-V) properties of single intact fast-twitch fibers from R. pipiens that contained
MHC types 1 or 2 (MHC1 or MHC2) or coexpressed MHC1
and MHC2 isoforms. Velocities were measured between two
surface markers that spanned most of the fiber length. MHC
and MLC isoform content was quantified after mechanics
analysis by SDS-PAGE. Maximal shortening velocity (Vmax)
and velocity at half-maximal tension (VP 50) increased with
percentage of MHC1 (%MHC1). Maximal specific tension
(Po/CSA, where Po is isometric tension and CSA is fiber
cross-sectional area) and maximal mechanical power (Wmax)
also increased with %MHC1. MHC concentration was not
significantly correlated with %MHC1, indicating that the
influence of %MHC1 on Po/CSA and Wmax was due to intrinsic differences between MHC isoforms and not to concentration. The MLC3-to-MLC1 ratio was not significantly correlated with Vmax, VP 50, Po/CSA, or Wmax. These data
demonstrate the powerful relationship between MHC isoforms and F-V properties of the two most common R. pipiens
fiber types.
population of muscle
cells, with a wide range of contractile properties, is a
hallmark of vertebrate skeletal muscle. The differences
in force-velocity (F-V) properties among muscle fiber
types permit muscles to perform a wide range of mechanical tasks. Functional differences among fiber
types have been correlated with myosin heavy chain
(MHC) and myosin light chain (MLC) isoforms. Thus
differential expression of MHC and MLC isoforms is an
important determinant of functional diversity in mus-
cular systems and is matched to the specific motor
requirements of an organism.
The greatest emphasis in single-fiber mechanical
studies has been on determining the influence of MHC
and MLC isoforms on maximal shortening velocity
(Vmax). Mechanical studies of single skinned fibers (sarcolemma permeabilized or removed) have demonstrated
a correlation between MHC isoforms and Vmax (2, 3, 16,
24, 39, 40, 44; for reviews see Refs. 5 and 42). With the
use of the in vitro motility assay, where the potential
influences of thin-filament regulatory proteins and muscle activation are avoided, the correlation between speed
and MHC isoforms has been corroborated (6, 18).
The influence of MLC isoforms on Vmax is more
controversial and may be variable among vertebrates.
Several studies have established a significant effect of
the molar ratio of essential (alkali) light chains [MLC3to-MLC1 ratio (MLC3/MLC1)] on Vmax (2, 14, 25, 37,
44; for reviews see Refs. 5, 36, and 42). Independent
support for the role of MLC3/MLC1 in regulating myosin kinetics has been documented in a genetic model
of fish (7) and with the in vitro motility assay (26).
However, in human skinned fast-twitch fibers, MLC3/
MLC1 had no influence on Vmax (23, 24, 46). In any
case, the influence of MLC3/MLC1 appears to be restricted to Vmax, inasmuch as there has been no demonstration of an effect at higher relative forces or on
maximum specific tension.
The influence of MHC isoforms on the remainder of the
F-V relation (i.e., velocities below Vmax) and on maximum
specific tension remains controversial. Studies on a variety of species representing several vertebrate classes
have documented a correlation between MHC isoform
and maximum specific tension. However, other studies
have reported no clear relationship (for reviews see Refs.
5, 36, and 42). A complicating factor in these studies is
that differences in myosin density may exist between
fiber types, and this may obscure the effect of MHC
isoforms on maximum specific tension (12, 45).
A limitation to the aforementioned contractile studies is that the experiments have been performed al-
Address for reprint requests and other correspondence: G. J. Lutz,
Dept. of Orthopaedics, University of California, San Diego, School of
Medicine, Veterans Affairs Medical Centers (9151), 3350 La Jolla
Village Dr., San Diego, CA 92161 (E-mail: [email protected]).
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
mechanics; maximal shortening velocity; fiber type; maximal
power; specific tension
EXPRESSION OF A HETEROGENEOUS
http://www.ajpcell.org
C835
C836
MYOSIN ISOFORM PROPERTIES
most exclusively on skinned muscle fibers. In contrast,
relatively few studies have assessed the influence of
MHC or MLC isoforms on F-V properties of intact
single fibers. Although amphibian muscle has long
been favored for intact single-fiber contractile studies,
its use in studies of myosin isoform properties has been
limited by a lack of clear definition of MHC and MLC
isoforms. By far, the most comprehensive studies of the
influence of myosin isoforms on F-V properties of intact
amphibian fibers is the detailed work on Xenopus laevis by Lannergren and coworkers (19–21). They demonstrated a nearly 10-fold range in Vmax across the five
major fiber types in the following order: type 1 ⬎ type
2 ⬎ type 3 ⬎ type 4 ⬎ type 5. Fiber types were
classified primarily by their nondenatured myosin
banding pattern on native-polyacrylamide gels, along
with very limited analysis of MHC and MLC isoforms
on SDS-polyacrylamide gels (19). The five major fiber
types identified in these experiments are likely to be
comprised of as many as five different MHC isoforms,
although the specific relationship between fiber type
and MHC isoform remains unresolved. Thus, although
direct evidence is limited, differences in MHC isoforms
appear to be correlated with the differences in Vmax in
X. laevis. The influence of the ratio of essential MLCs
(MLC3/MLC1) on F-V properties was less clear, because only the velocity at half-maximal tension, rather
than Vmax, was considered. Significant differences in
maximal specific tension and maximal mechanical
power (Wmax) were also reported in the following order:
type 1 ⬎ type 2 ⬎ type 3 ⬎ type 4 ⬎ type 5 (19). These
data indicate that MHC isoforms have a significant
influence on maximal isometric tension and Wmax.
We recently performed a detailed analysis of MHC
isoforms in the full range of fiber types of Rana pipiens
at the protein and mRNA levels (29, 31; for review see
Ref. 30). We showed that four major fiber types (type 1,
type 2, type 3, and tonic) distinguished by myofibrillar
ATPase reactivity and immunohistochemistry contained four unique MHC isoforms. Single-fiber analysis showed that MHC1, MHC2, and MHC3 migrated as
separate bands on SDS-polyacrylamide gels, while the
tonic type comigrated with MHC1. The MLC isoforms
in these fiber types were also characterized, and their
migration pattern on SDS-polyacrylamide gels was established (27). These definitions of MHC and MLC
isoforms provide a foundation for the determination of
their influence on contractile function in intact single
fibers.
In this report, we measured the relationship between
MHC and MLC isoform content and F-V properties of
intact R. pipiens single fibers. Analysis was restricted
to fibers that contained type 1 or 2 or coexpressed type
1 and 2 MHCs, inasmuch as these are the predominant
MHCs expressed in the hindlimb musculature of R.
pipiens (28). A spot-follower device was used to measure the precise F-V relation of the central 78% of
single fibers, and myosin isoforms were determined
after mechanics analysis by quantitative SDS-PAGE.
We found that differences in MHC isoforms were correlated with differences in F-V properties across the
AJP-Cell Physiol • VOL
full extent of the F-V relation, from maximal isometric
specific tension (Po/CSA, where Po is isometric tension
and CSA is cross-sectional area) to Vmax. There was no
significant independent influence of MLC3/MLC1 on
the F-V relation. Increased percentage of MHC1
(%MHC1) resulted in higher shortening velocity and
specific tension, thus yielding a significant influence on
Wmax production. These differences in contractile properties at the single-fiber level are discussed in the
context of the functional importance of MHC isoform
expression on muscle performance during jumping.
METHODS
Animals. Adult male R. pipiens frogs (Charles Sullivan,
Nashville, TN) were housed at room temperature (⬃20°C) in
20-gallon aquaria containing dry, dark surfaces and circulating filtered water. Live crickets were provided twice weekly,
and the frogs were kept for up to 2 mo before experimentation.
Single-fiber dissection. Frogs were killed by double pithing,
and both anterior tibialis (AT) muscles were removed and
pinned in Sylgard-lined glass dishes containing Ringer solution. The Ringer solution contained (in mM) 120.0 NaCl, 3.0
KCl, 4.0 NaPO4, 3.0 MgSO4, and 4.0 CaCl2, at pH 7.1.
Single-fiber dissections were carried out at room temperature
under a stereomicroscope (model MZ12.5, Leica, Deerfield,
IL) equipped with a transmitted light base for dark-field
imaging. The medial portion of the AT toward its origin was
thinned to a shallow layer of cells that ran between the
central and medial aponeuroses. This portion of the AT was
chosen, because it contained predominantly type 1, type 1-2,
and type 2 fibers and because the tendons were relatively
robust. For single-fiber isolation, small bundles (2–4 fibers)
were taken from the thinned muscle and transferred to a
shallow glass-bottomed dish. Individual intact fibers were
isolated and cleaned of debris using microscissors, fine forceps, and 30-gauge needles. Fibers were tested periodically
throughout the dissection with point stimulation. Fibers
were considered viable for mechanical analysis if they responded to point stimulations along their full length. Small
holes were punched in both tendons for mounting in the
mechanics chamber.
Single-fiber mechanical recording system. Single fibers
were placed in a Ringer solution-filled anodized aluminum
chamber with a glass bottom insert. The chamber was
mounted on a translation stage under a stereomicroscope
(model MZ8, Leica). Ringer solution was constantly refreshed
with a miniperistaltic pump (Harvard Apparatus, Holliston,
MA). Fibers were mounted between a force transducer (model 405A, Aurora Scientific, Ontario, Canada) and a highspeed length controller (model 318B, Aurora Scientific) by
securing the tendons to fine titanium wires (127 ␮m; Aldrich,
Milwaukee, WI) protruding from the force transducer and
motor arm using 9-0 black monofilament nylon (Ashaway
Line and Twine, Ashaway, NJ). The force transducer and
motor were mounted independently on xyz-translation stages
to permit precise fiber alignment. Chamber temperature was
regulated by direct contact between the chamber plate and a
flow-through aluminum bar with temperature-controlled circulating water (model RTE101, Thermo Neslab, Portsmouth,
NH). The microscope was equipped with a charge-coupled
device videocamera (Techni-Quip, Livermore, CA) mounted
in the phototube. The video system included a videocassette
recorder, video micrometer (Techni-Quip), monitor (Sony),
and video capture card (model RS170, Scion, Frederick MD).
282 • APRIL 2002 •
www.ajpcell.org
C837
MYOSIN ISOFORM PROPERTIES
Fibers were activated by direct bipolar electrical stimulation
through a pair of platinum plate electrodes that straddled
the fiber along its full length. Stimulus pulses were generated with a bipolar stimulator (Pulsar 6bp, FHC, Bedowin,
NJ) and amplified with a power amplifier (model 6826A,
Hewlett-Packard, Palo Alto, CA).
To improve the accuracy of the mechanical measurements,
a spot-follower system was used to measure length changes
within a defined region of the fiber that constituted 78% of
the overall fiber length. Two surface markers consisting of
black cat hair (⬃200 ⫻ 50 ␮m) were adhered to the upper
surface of the fiber within ⬃0.75 mm of each myotendonous
junction. The markers were coated with silicon grease to
promote adherence to the fiber and were oriented perpendicular to the long axis of the fiber. Light was directed onto the
upper surface of the fiber, and the position of the markers
was detected with a 2,048-element linear photodiode array
(PDA; model S2048, Reticon) oriented parallel to the long
fiber axis in the phototube of the microscope. The position of
each marker was detected as a sharp decrease in light intensity from the video output signal of the PDA. Marker positions were monitored continuously during contractions and
converted to fractional length change (see below). To improve
signal-to-noise ratio, a removable white background was inserted below the fiber. On the basis of the dimensions of the
markers and their deviation from perpendicular, we calculated systematic error in fractional change in distance between the markers during F-V measurements to be ⬍2%.
We also attempted to measure the length transients in
smaller segments along the fiber length by placing additional
markers between the end-most markers. Typically, four additional markers were applied at ⬃1-mm increments. Unfortunately, in most cases, we could not apply a full set of
markers that were each perfectly perpendicular to the fiber
axis. We determined that, during contractions, slight angulation of markers, coupled with lateral fiber movement, could
result in relatively large systematic errors in our measurement of the change in distance between successive markers
and, hence, in determination of segment shortening velocity.
Therefore, all F-V measurements were based on the velocity
between the end-most markers (see F-V measurements).
However, the remaining internal markers served as useful
landmarks for dividing the fiber into segments for morphological and sarcomere length (SL) measurements before the
mechanics experiments and for measurements of myosin
content after the mechanics experiments (see below).
Computer control and software. Data were acquired and
mechanical experiments were controlled with a four-channel,
5-MHz analog-to-digital/digital-to-analog board (model PCI
6110E, National Instruments, Austin, TX) interfaced to a
Pentium personal computer using LabVIEW 5.11 and NIDAQ 6.7 software (National Instruments). Output from the
surface marker PDA was acquired at 4 MHz, yielding a
complete scan every 0.52 ms. Output signals from the force
transducer, motor, and stimulator were acquired on the remaining input channels. Custom LabVIEW software was
used to trigger the stimulator and control output to the
motor.
Resting SL, fiber length, and diameter measurements. After the fiber was mounted and aligned in the chamber,
surface markers were applied as described above. Passive SL
was measured with laser diffraction in each of the ⬃1-mm
segments. A helium-neon laser beam (Melles Griot, Irvine,
CA) was directed through the fiber from below, and the
diffraction lines were imaged directly onto a PDA (model
S2048, Reticon) placed 30 mm above the fiber. The PDA was
mounted so that it could be rotated away from the fiber when
AJP-Cell Physiol • VOL
not in use. SL was calculated from the distance between the
two first-order diffraction lines using the standard diffraction
equation. Fiber length was adjusted to yield an average
resting SL (SLo) of ⬃2.35 ␮m. Because the laser beam had a
diameter of ⬃0.8 mm, SLo provided an accurate estimate of
the average SL within the region between the end-most
markers. The length of the entire fiber (Lf) was measured
using the video system. Only a rough estimate of the exact
insertion points of the fiber on each tendon was possible,
because, in many cases, fibers curved slightly around the
connective tissue at the insertion, and the view of the insertion was obstructed by its tendon. Average diameter of each
segment was measured in two orthogonal views (a 45° mirror
was used for the side view) from digitized video images. The
entire length of each segment was imaged, and diameters
were measured along the full length of the segments with
cursors, and CSA was calculated from the two diameters with
the assumption of an elliptical shape. The position of each
marker relative to the fiber ends was measured using the
video system.
F-V measurements. All experiments were performed at
25°C. Fibers were activated with supramaximal 0.3-ms voltage pulses (25% above twitch threshold) with frequency of
200 Hz. A series of isovelocity contractions were performed to
obtain F-V curves, with contractions separated by 4 min.
Examples of typical contractions are shown in Fig. 1. For
each contraction, the fiber was stimulated and held isometric
at Lf for 100 ms, while tetanic force reached a plateau (Po).
The fiber was then given a high-velocity [20 fiber lengths/s
(L/s)] small release step to unload the series elastic component, which was followed immediately by a constant-velocity
shortening ramp. The ramp amplitude was 12–15% of Lf. The
amplitude of the rapid release step was adjusted in each contraction to cause force to drop to a level that could be maintained during the ensuing ramp. Force typically stabilized
within 5–10 ms after the step. Contractions were performed in
increments of 1.0 L/s over most of the velocity range and
increments of 0.5 L/s near the highest and lowest velocities.
Fractional length change was calculated as the change in
distance between the peaks corresponding to the end-most
surface markers on the PDA (units of pixels) divided by the
pixel distance under static conditions with the fiber held at
Lf. Shortening velocity was taken as fractional length change
divided by time, normalized to an SL of 2.35 ␮m (in L/s).
Force and length records were read over the shortening
period, beginning ⬃5–10 ms after the high-velocity step
(when force had stabilized), for 5–20 ms (depending on ramp
velocity; Fig. 1). When ramp velocity was high, the records
were read for a shorter duration. This corresponded to a
shortening range of 2–7.5% for the lowest and highest velocities, respectively. Velocity and force were nearly constant
over this shortening range in all contractions. Force (P) was
taken as the average value over the same time period used to
calculate velocity. SL change during contractions was calculated directly from the fractional length change and SLo.
Analysis and curve fitting of F-V data. F-V data were fit
with the standard Hill equation (Eq. 1) by least-squares
analysis over the region 0.05 ⬍ P/Po ⬍ 0.78 as
V ⫽ 共P *o ⫺ P兲b/共P ⫹ a兲
(1)
where P/Po is relative tension, a and b are constants with
dimensions of force and velocity, respectively, and P*o is the
intercept on the force axis (8, 17). The parameter a/P*o describes the degree of curvature of the F-V relation, where
higher values indicate less curvature. Velocity at P/Po ⫽ 0.05
and P/Po ⫽ 0.50 were calculated from Eq. 1 and designated
VP 05 and VP 50, respectively. Maximal specific tension was
282 • APRIL 2002 •
www.ajpcell.org
C838
MYOSIN ISOFORM PROPERTIES
Fig. 1. Sarcomere length transients
(A) and force production (B) of an intact single frog muscle fiber during a
series of isovelocity contractions over a
range of shortening velocities. Sarcomere length transients were obtained
by measuring the fractional length
change between two surface markers,
one near each end of the fiber, throughout the contractions. Sarcomere length
was calculated directly from the fractional length change and the average
initial sarcomere length between the
markers (see METHODS). Fiber shortening velocities were 3.0, 6.0, and 12.5
lengths/s (L/s) for traces a, b, and c,
respectively. The period over which
length and force were measured for
force-velocity (F-V) curves is indicated
by thickened lines in the respective
traces. In practice, ⬃13–15 such contractions were used to construct the
F-V relationship (cf. Fig. 2). Stimulus
began at time 0 and continued
throughout the period shown. C and D:
traces from A and B on a zoomed scale
to demonstrate that force and velocity
were quite constant over the period
used in the analysis of F-V curves. Experiments were performed at 25.0°C.
calculated as Po/CSA (in kN/m2), where CSA was taken as
the weighted average of the segments between the first and
last surface markers. Mechanical power was calculated directly from the F-V fit as the product of specific tension
(P/CSA) and velocity (in W/kg). Vmax was calculated by fitting
the data with Eq. 1 over the region 0 ⬍ P/Po ⬍ 0.78 and
calculating the intercept at zero force. Vmax was also calculated using linear regression of the F-V data in the region 0 ⬍
P/Po ⬍ 0.075.
SDS-PAGE analysis of MHC and MLC isoforms. After
mechanical experiments, fibers were pinned at ⬃Lf to pieces
of wax while in the chamber and then removed from the
chamber and stored at ⫺80°C. Fibers were freeze-dried and
cut into the segments delineated by the surface markers. In
most cases, one or more of the markers had fallen off during
freezing. The position of the lost markers on the freeze-dried
fibers was determined from previous video images of the fiber
in the mechanics chamber. Freeze-dried fiber segments were
placed into microfuge tubes and suspended in 10 ␮l of SDSPAGE sample buffer consisting of 100 mM dithiothreitol, 2%
SDS, 80 mM Tris-base (pH 6.8), 10% glycerol, and 0.012%
(wt/vol) bromphenol blue. Samples were boiled (2 min) and
stored at ⫺80°C for up to 3 wk before they were loaded onto
gels.
MHC and MLC isoforms were measured in each segment
using SDS-PAGE, with 10% of the sample volume used for
analysis of MHCs and the remaining 90% for MLCs. SDSPAGE of MHC isoforms was carried out as previously described (29). Gels were silver stained, and MHC bands were
quantified by densitometry (PDI, Bio-Rad, Hercules, CA).
The total amount of MHC in each segment was determined
by comparison with a standard prepared from semimembranosus (SM) muscle on the same gels. The SM standard was
prepared by isolating the myofibrillar fraction from SM muscle containing no external tendon, as previously described
(28). Protein content in the myofibrillar fraction was determined (Pierce, Rockford, IL) and suspended in SDS-PAGE
AJP-Cell Physiol • VOL
sample buffer to a concentration of 0.125 mg protein/ml.
Because myosin represents ⬃43% of myofibrillar protein (47)
and MHC ⬃85% of myosin mass, the SM standard contained
⬃0.045 mg MHC/ml. Ten microliters of a fivefold dilution of
the SM standard were loaded per lane, which was equivalent
to 0.091 ␮g MHC/lane. Molar concentration of MHC in each
segment was calculated as follows: M/V 䡠 MW, where M is the
mass of MHC estimated from the gel, V is segment volume
taken from its dimensions in the mechanical chamber (normalized to SL ⫽ 2.35 ␮m), and MW is the molecular weight
of MHC (205,000 g/mol).
SDS-PAGE of MLC isoforms was carried out as previously
described (27). Gels were silver stained, and MLC3/MLC1 in
each segment was calculated from the band densities and
molecular mass values using densitometry and NIH Image
software (version 1.62).
Statistical analysis. To quantify the relative influence of
various parameters on peak power production, multiple linear regression was implemented. Independent variables
were screened for normality and skew to justify the use of the
parametric analysis. Two separate regression analyses were
performed. First, the three mechanical parameters used to
calculate maximal power were allowed to enter a multiple
regression model. This was performed to understand the
extent to which each of these parameters was related to
maximum power. In an independent analysis, 15 mechanical
and structural parameters (see RESULTS) were allowed to
enter the multiple regression model. This was performed to
understand which parameters were overall more highly correlated with maximum power production. Significance level
(␣) was set to 0.05.
RESULTS
F-V properties of intact single fibers (n ⫽ 12) were
measured with a spot-follower device that continuously
tracked the distance between two surface markers that
282 • APRIL 2002 •
www.ajpcell.org
MYOSIN ISOFORM PROPERTIES
were placed within 0.80 ⫾ 0.05 and 0.71 ⫾ 0.09 mm of
the moving and fixed ends of the fiber, respectively.
Resting fiber length was 6.97 ⫾ 0.12 mm. Thus markers spanned 78.7 ⫾ 0.9% of the fiber length. Sample
records of a series of contractions from one fiber are
shown in Fig. 1. Experiments were designed so that
shortening ramps were performed over the plateau
region of the length-tension curve (optimal myofilament overlap), which, in frog muscle, occurs between
SL of 2.25 and 2.05 ␮m (13). On average, contractions
started from an SLo of 2.35 ⫾ 0.01 ␮m. During the
isometric phase, SL decreased from 2.35 to 2.23 ⫾ 0.07
␮m. Further sarcomere shortening occurred during the
high-velocity step and before force stabilization. Force
and velocity records were read immediately after force
stabilized. On average, sarcomeres shortened during
the period used to construct F-V curves from 2.13 ⫾
0.04 to 2.05 ⫾ 0.04 ␮m, and thus shortening velocity
was indeed measured over the plateau of the lengthtension curve. All fibers maintained excellent integrity
throughout the F-V protocol, inasmuch as isometric
tension (Po) declined by only 3.1 ⫾ 0.6% from the first
to the last contraction (range 0.9–8.4%).
Representative F-V curves for a relatively fast- and
slow-twitch fiber are shown in Fig. 2A. F-V data for all
fibers were well fit by the standard Hill hyperbolic
equation (Eq. 1) over the range 0.05 ⬍ P/Po ⬍ 0.78.
Vmax values obtained by the Hill fit over the range 0 ⬍
P/Po ⬍ 0.78 and linear regression over the range 0 ⬍
P/Po ⬍ 0.075 were not significantly different (P ⫽ 0.51;
Fig. 2A, inset). Thus only Vmax from the Hill fit is
reported. Power-velocity curves, calculated directly
from the F-V fits, produced reliable fits of the powervelocity data (Fig. 2B).
After mechanical experiments, MHC and MLC isoform contents were determined using quantitative
SDS-PAGE. Representative MHC and MLC gels are
shown in Fig. 3. For each fiber, MHC and MLC isoforms were measured in each segment between the
end-most markers, and data were averaged. Fibers
contained only type 1, 2, or 1 and 2 (coexpressed)
MHCs. For MLCs, we measured the molar ratio of
essential light chains, MLC3/MLC1, inasmuch as this
ratio has been shown in previous studies to correlate
with Vmax. The relationship between %MHC1 and
MLC3/MLC1 is shown in Fig. 4. Linear regression
showed that although MLC3/MLC1 increased with
%MHC1, the relationship was not highly correlated
(r2 ⫽ 0.097) and did not reach statistical significance
(P ⫽ 0.32). This weak correlation agreed with previous
examination of a larger data set from 35 single AT
fibers (27). The extreme variability in MLC3/MLC1
between fibers was not due to intrinsic variability in
the quantification technique, which was shown to have
a coefficient of variation ⬍15% (27). Also, we typically
found the coefficient of variation of %MHC1 to be
⬍1.0% over a 10-fold range of sample concentration
(including the range of band densities of single fibers in
this study; data not shown).
Analysis of the F-V data revealed that Vmax increased significantly with %MHC1 but was not signifAJP-Cell Physiol • VOL
C839
Fig. 2. F-V and power-velocity relationships for a relatively fast (F)
and a slow (E) intact frog muscle fiber. A: F-V curves fit with the
standard Hill equation (Eq. 1) over the range 0.05 ⬍ P/Po ⬍ 0.78. As
expected, the Hill fit deviates significantly from the data points in
the high-force region of the curves (dashed lines). Inset: high-velocity
region of the F-V data in A. Two methods were used to calculate
maximal shortening velocity (Vmax): Hill fit over the range 0 ⬍ P/Po ⬍
0.78 (where P/Po is relative tension; solid line) and linear regression
fit over the region P/Po ⬍ 0.075 (dashed line). Vmax was taken as the
intercept of the respective lines on the velocity axis. Force is shown
in relative units (P/Po). B: power-velocity curves calculated directly
from the F-V fits in A. Data points were calculated as the product of
force and velocity in A. Experiments were performed at 25.0°C.
icantly correlated with MLC3/MLC1 (Fig. 5, Table 1).
Similarly, VP 05 and VP 50 increased with %MHC1, but
we found no correlation between MLC3/MLC1 and
either VP 05 or VP 50.
Po/CSA also increased significantly with %MHC1
(Fig. 6, Table 1), while MLC3/MLC1 was not significantly correlated with Po/CSA (Table 1). From the
regression analysis in Fig. 6, Po/CSA increased by
22.3% as %MHC1 increased from 0 to 100%. Measurements of myosin concentration were used to determine
whether the significant increase in Po/CSA with
%MHC1 could be attributed to a greater amount of
myosin per unit volume, rather than actual differences
in MHC isoform properties. The amount of MHC in
each segment between the end-most fiber markers was
measured using quantitative SDS-PAGE, and segment
volumes were calculated as the product of CSA and
segment length (see METHODS). For each fiber, the
282 • APRIL 2002 •
www.ajpcell.org
C840
MYOSIN ISOFORM PROPERTIES
eters (Vmax, VP 50, VP 05, P*o, a/P*o, %MHC1, MLC3/
MLC1, Po, CSA, Po/CSA, velocity at Wmax, P/Po at
Wmax, MHC mass, volume, and MHC concentration)
showed that %MHC1 had the highest correlation with
Wmax (r2 ⫽ 0.846).
DISCUSSION
Fig. 3. Myosin heavy and light chain (MHC and MLC) isoforms
determined by SDS-PAGE. A: MHC isoforms in segments of one fiber
determined on an SDS-polyacrylamide gel (silver stained) after mechanical analysis. Each lane corresponds to a different segment
along the length of the fiber (a–f), and the segments together comprised the entire length between the end-most surface markers. Type
1 MHC (MHC1), calculated as a percentage of total MHC, is indicated below each lane. Total MHC in each segment was quantified by
comparison with a semimembranosus (SM) muscle standard that
contained a known amount of MHC (0.091 ␮g/lane). B: SDS-PAGE of
MLCs in each of the fiber segments in A. Ratio of MLC3 to MLC1
(MLC3/MLC1) calculated from this gel (silver stained) is indicated
below each sample lane. Molecular masses of gel markers (left-most
lane) were 45, 31, 21.5, and 14.5 kDa.
amount of MHC per unit volume was taken as the
weighted average of the segments. Linear regression
showed that the amount of MHC was directly proportional to fiber volume (r2 ⫽ 0.973; Fig. 7A), providing
confidence that MHC mass and fiber volume measurements were accurately determined. On average, fiber
MHC concentration was 197.7 ⫾ 10.9 ␮M, which was in
close agreement with previous estimates of MHC concentration in single muscle fibers (12, 45). The relationship
between %MHC1 and MHC concentration (Fig. 7B) was
poorly correlated (r2 ⫽ 0.113) and did not reach statistical
significance (P ⫽ 0.28). Thus the increase in Po/CSA with
%MHC1 was not explained by greater myosin concentration, indicating that the effect of MHC isoform on Po/CSA
was more likely to result directly from differences in
isoform properties (see DISCUSSION).
Wmax increased significantly with %MHC1 (Fig. 8,
Table 1), while no significant correlation was found
between MLC3/MLC1 and Wmax (Table 1). From the
regression analysis in Fig. 8, Wmax increased by 61.3%
as %MHC1 increased from 0 to 100%. Wmax is the
product of Po/CSA, P/Po, and velocity at the point of
peak power. Multiple regression analysis showed that
Po/CSA accounted for 70.4% of the variability in Wmax,
while velocity and P/Po explained an additional 17.1
and 11.6% of the variability, respectively. A more general multiple regression analysis of 15 different paramAJP-Cell Physiol • VOL
Because of their robust stability when isolated from
whole muscle, single intact frog fibers continue to serve
as a valuable experimental model for studies of contractile function and high-resolution myosin crossbridge properties. However, the lack of precise definition and detection of myosin isoforms in anuran
skeletal muscle have placed a significant limitation on
their use in determining the influence of myosin isoforms on contractile properties in intact cells. We took
advantage of the recent detailed description of MHC
and MLC isoforms in the various fiber types of R.
pipiens (27, 29) to determine their influence on the
dynamic properties of isolated intact single fibers. Contractile properties were measured in fibers that contained type 1, 2, or 1 and 2 MHCs, which are the
predominant MHC isoforms expressed in hindlimb
muscles of R. pipiens (28). Overall, the mechanical
analysis showed that Vmax, VP 05, VP 50, Po/CSA, and
Wmax were strongly correlated with %MHC1. We found
no significant influence of MLC3/MLC1 on any mechanical parameters.
Mechanical experiments were performed using a
spot-follower device that permitted F-V properties to
be precisely measured between two surface markers
that spanned the central 78% of the fiber length. The
surface marker technique has significant advantages
over whole fiber F-V measurements. First, systematic
errors in estimating fiber length are eliminated. Fiber
length can be difficult to measure accurately because of
inherent ambiguity in determining exactly where fibers insert onto tendons. Also, measurement of fiber
CSA is improved, because fiber diameter can be accurately measured throughout the region of interest,
whereas morphology at the fiber ends may be variable
Fig. 4. Relationship between percentage of MHC1 (%MHC1) and
MLC3/MLC1 in single fibers. Data are shown for each of the fibers
used in mechanical analysis (n ⫽ 12).
282 • APRIL 2002 •
www.ajpcell.org
MYOSIN ISOFORM PROPERTIES
C841
Fig. 5. Shortening velocity as a function of %MHC1 and MLC3/MLC1. A–C:
linear regression analysis showed that
Vmax, velocity at 5% of maximum isometric tension (VP 05), and velocity at 50% of
maximum isometric tension (VP 50) increased with %MHC1. D–F: Vmax, VP0 5,
and VP 50 were not significantly correlated with MLC3/MLC1. Data points are
individual fibers (n ⫽ 12), and regression coefficients are provided in Table 1.
compared with the central portion. Also, the end-most
regions of fibers are known to have mechanical properties that are significantly different from the rest of
the fiber (9, 11). Exclusion of the end regions should
produce more reliable SL and velocity measurements
and, thus, more meaningful correlations between contractile behavior and myosin isoforms.
For all contractions combined, fiber velocity measured
from the servomotor and velocity measured between the
surface markers were highly correlated (r2 ⫽ 0.998).
However, fiber velocity was 4.6 ⫾ 0.7% greater (P ⬍
Table 1. Linear regression variables describing
influence of %MHC1 and MLC3/MLC1 on contractile
function of intact single muscle fibers
%MHC1
Vmax
VP 05
VP 50
Po/CSA
Wmax
r2
0.373
0.603
0.753
0.595
0.860
MLC3/MLC1
P
r2
P
0.035
0.003
0.0003
0.003
⬍ 0.0001
0.072
0.184
0.016
0.139
0.075
0.400
0.164
0.694
0.239
0.390
MLC3/MLC1, ratio of myosin light chain (MLC) 3 to MLC 1;
MHC1, myosin heavy chain type 1; Vmax, maximal shortening velocity; VP 05 and VP 50, velocity at P/Po of 0.05 and 50 (where P is force
and Po is tetanic force); CSA, cross-sectional area; Wmax, maximum
mechanical power.
AJP-Cell Physiol • VOL
Fig. 6. Maximal specific isometric tension as a function of %MHC1 in
intact single frog fibers. Data points are individual fibers (n ⫽ 12),
and linear regression coefficients are provided in Table 1.
282 • APRIL 2002 •
www.ajpcell.org
C842
MYOSIN ISOFORM PROPERTIES
Fig. 7. MHC content in single frog fibers. A: MHC mass was directly
proportional to fiber volume (n ⫽ 11). B: MHC concentration increased slightly with %MHC1, but the correlation was weak and did
not reach statistical significance.
0.0001) than velocity between the surface markers (data
not shown). One explanation for the difference is that
fiber ends are thought to be stronger than the central
part of the fiber (9). Thus, for a given force level, velocity
is likely to be faster at the fiber ends than in the central
region. To account for the full 4.6% difference, the fiber
ends (which constitute ⬃22% of fiber length) would have
to shorten 22% faster than the central portion. Differences in velocity of this magnitude along fibers are reasonable, inasmuch as it has been shown that differences
in Vmax of up to 40% (at 2–3°C) occur between successive
segments of fibers (10, 11). However, it is also possible
that fiber length was systematically underestimated because of difficulty in determining fiber insertions.
The lack of an influence of MLC3/MLC1 on Vmax (Fig.
5) is in contrast to several previous studies of skinned
single fibers (for reviews see Refs. 5, 36, and 42). For
example, in rodent fast-twitch fibers that contained
only MHCIIB, MLC3/MLC1 was strongly correlated
with increased Vmax, and the differences were as large
as the differences in Vmax between fibers containing
MHCIIA, MHCIIX, and MHCIIB (2). In contrast, our
results agree with those of human fast-twitch fibers,
which showed no discernable influence of MLC3/MLC1
on Vmax (23, 24, 46). However, the MHCIIB isoform
AJP-Cell Physiol • VOL
was not studied in human fibers (it is not expressed),
but MHCIIB was the isoform that showed the greatest
effect with MLC3/MLC1 function in the rodent (5). The
lack of a modulatory effect of MLC3/MLC1 on Vmax in
R. pipiens is somewhat surprising, given the fact that
MLC3/MLC1 varied nearly fivefold (0.24–1.19), which
was nearly identical to the magnitude and variability
within rodent MHCIIB fibers (2). It remains to be seen
whether the somewhat greater MLC3/MLC1 values in
fibers containing only MHC1 from other muscles in R.
pipiens (27) are correlated with increased Vmax. We are
not aware of any previous studies of MLC3/MLC1 on
Vmax in intact single fibers; thus the extent to which
skinning may cause differences between our results
and those from skinned fibers remains unanswered. It
will be useful to see if the steep relationship between
Vmax and MLC3 holds up in intact fibers of rodents.
There appears to be a consensus that the effect of
MLC3/MLC1 is restricted to Vmax and does not extend
to loaded fibers (4, 5, 42). Thus the functional significance of variability in MLC3/MLC1 is questionable,
since muscles do not generate either force or power
when shortening at Vmax.
Whether MHC isoforms affect Po/CSA in striated
muscle remains controversial (5, 36, 42). We found that
Po/CSA increased by 22.3% as %MHC1 increased from
0 to 100% (Fig. 6). The MHC isoform effect did not
appear to result from an increased number of cross
bridges per half sarcomere, inasmuch as MHC concentration was unrelated to %MHC1 (Fig. 7B). Thus it
appears that intrinsic differences in cross-bridge function between type 1 and 2 MHC isoforms resulted in
differences in Po/CSA. Further experiments are required to assess whether the specific influence of MHC
isoforms on Po/CSA was due to differences in force per
cross bridge or the fraction of cross bridges in the
strongly bound force-generating state (duty cycle) (5,
12, 42, 45). Kinetic measurements of single myosin
molecules suggest that differences in force-generating
capacity between myosin isoforms are the result of
different duty cycles, rather than differences in unitary
force production (15).
Fig. 8. Maximal mechanical power (Wmax) as a function of %MHC1
in intact single frog fibers. Data points are individual fibers (n ⫽ 12),
and linear regression coefficients are provided in Table 1.
282 • APRIL 2002 •
www.ajpcell.org
MYOSIN ISOFORM PROPERTIES
The relationship between MHC isoforms and Po/
CSA, including factors that complicate the interpretation of results in single skinned fibers, has been comprehensively discussed (5, 12, 36, 42, 45). There is
general consensus that, in skinned fibers of mammals,
Po/CSA is greater in fast-twitch fibers (types IIA, IIX,
and IIB) than in slow-twitch fibers (type I). However,
because of significant swelling of fibers and other factors specific to skinned fibers, measurements of Po/CSA
in skinned fibers are unreliable, especially when optical measurements are used to determine CSA (see
Table 7 in Ref. 42 and Table 3 in Ref. 5). Importantly,
by measuring myosin density and fiber mass in single
fibers after mechanics experiments, the direct dependence of isometric force on MHC isoform can be determined (12, 45). Significant differences in Po/CSA
among mammalian fiber types have also been documented in intact fibers using motor unit (1) and whole
muscle mechanical measurements (38).
In a comprehensive series of studies, Lannergren and
colleagues (19–21) measured the contractile properties of
single intact fibers representing the full range of fiber
types from X. laevis at 20°C. Fiber types were classified
after mechanics analysis on the basis of native isomyosin
bands and limited analysis of MHC and MLC isoforms
with SDS-PAGE (19). Five major fiber types and several
subtypes were identified in X. laevis and appeared to
contain as many as five MHC isoforms (20, 22, 41, 43).
However, a one-to-one correspondence between fiber type
and MHC isoform is not yet certain. Importantly, it was
shown with SDS-polyacrylamide gels that type 1 and 2
fibers (subclassified as types 1n and 2n) in X. laevis
contained unique MHC isoforms (19).
The differences in contractile properties between type
1 and 2 fibers in R. pipiens are similar to the differences
between type 1n and 2n fibers in X. laevis. From linear
regression analysis, we calculate that Vmax, VP 50, Po/
CSA, and Wmax increased by 21, 34, 22, and 61% as
%MHC1 increased from 0 to 100% in R. pipiens fibers
(Figs. 5, 6, and 8). Lannergren (19) showed an increase of
41, 50, 32, and 93% for these same mechanical parameters in type 1n compared with type 2n fibers. Because
Lannergren’s studies on X. laevis were performed at 20°C
and our experiments on R. pipiens at 25°C, direct comparison of absolute values is problematic. We used Q10
values of 1.16, 1.63, and 1.88 for Po/CSA, Vmax, and Wmax,
respectively, to extrapolate the mechanics of type 1 fibers
in R. pipiens to 20°C (34). With the use of these Q10
values and the linear regression analysis stated above,
Po/CSA and Wmax of R. pipiens were 15.7 and 12.3% less,
respectively, for type 1 than type 1n fibers in X. laevis. In
contrast, Vmax was only 2.8% greater in R. pipiens than in
X. laevis type 1n fibers. Thus, overall, the contractile
properties of the fastest fiber types in these two anuran
species are very similar. More importantly, both studies
demonstrate that MHC isoforms have a significant influence on the full F-V relationship and maximal mechanical power production in intact fibers.
Because of the functional significance of mechanical
power production on in vivo motor performance, it was of
interest to determine in more detail how MHC isoforms
AJP-Cell Physiol • VOL
C843
influenced power production. From a strict mechanical
point of view, the largest influence on Wmax was Po/CSA,
which accounted for 70% of the variability in the data.
Surprisingly, when the influence of 15 mechanical and
structural parameters on Wmax was determined, %MHC1
was by far the best predictor of Wmax. Thus the MHC
isoform has a profound influence on Wmax in R. pipiens
single fibers, even more so than Vmax or Po/CSA, parameters traditionally associated with power production in
muscle. A potential problem with regression models is
the preferential entry of parameters that are measured
with greater accuracy and resolution than others. On the
basis of difficulties associated with quantitative measurements of %MHC1 in single fibers, we believe that the
%MHC1 parameter actually represents one of the more
difficult parameters to quantify. This makes its dominance in the regression analysis even more impressive.
Importance of MHC isoforms on muscle performance
during jumping. During jumping, frogs generate high
levels of mechanical power in their hindlimb extensor
muscles, and jump distance is directly proportional to
the peak power produced (33, 35). Fiber shortening
velocity has been measured in the muscle fibers of the
SM muscle (a large hindlimb extensor muscle) during
maximal jumps at 25°C in R. pipiens (32–34). On
average, shortening velocity during jumping was 3.94
L/s (normalized to SL ⫽ 2.35 ␮m), and fibers shortened
mostly over the plateau of the length-tension curve
(33). From the power-velocity curves of single fibers
(such as those in Fig. 2), we calculated that at 3.94 L/s
fibers generated ⬃100% of their maximal power (range
96.1–100%; data not shown). Accordingly, power produced at the in vivo jump velocity of 3.94 L/s increased
by 57% as %MHC1 increased from 0 to 100%.
The strong influence of MHC isoforms on mechanical
power production is reflected in their overall distribution
in the various hindlimb muscles of R. pipiens. We previously compared the MHC isoform content in the large
extensor muscles involved in jumping with other muscles
in the hindlimb (28). We found that the large extensor
muscles contained ⬃90% MHC1, while the other muscles
not involved in jumping were predominantly composed of
MHC2. The contractile data in this study, along with this
MHC isoform distribution data, demonstrate the importance of MHC isoform expression in muscular system
design. In the case of R. pipiens, the muscular system is
designed to generate high levels of mechanical power in
the hindlimb extensor muscles during jumping.
We are grateful to Dustin Robinson and Haiyan Yu for excellent
technical assistance and the University of California, San Diego,
Physics Department Electronics Shop for valuable assistance with
PDA design.
This study was supported by National Institute of Arthritis and
Musculoskeletal and Skin Diseases Grants AR-40050, AR-45631,
and AR-46469 and a grant from the Department of Veterans Affairs.
REFERENCES
1. Bodine SC, Roy RR, Eldred E, and Edgerton VR. Maximal
force as a function of anatomical features of motor units in the
cat tibialis anterior. J Neurophysiol 57: 1730–1745, 1987.
2. Bottinelli R, Betto R, Schiaffino S, and Reggiani C. Unloaded shortening velocity and myosin heavy chain and alkali
282 • APRIL 2002 •
www.ajpcell.org
C844
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
MYOSIN ISOFORM PROPERTIES
light chain isoform composition in rat skeletal muscle fibres.
J Physiol (Lond) 478: 341–349, 1994.
Bottinelli R, Canepari M, Pellegrino MA, and Reggiani C.
Force-velocity properties of human skeletal muscle fibres: myosin heavy chain isoform and temperature dependence. J Physiol
(Lond) 495: 573–586, 1996.
Bottinelli R and Reggiani C. Force-velocity properties and
myosin light chain isoform composition of an identified type of
skinned fibres from rat skeletal muscle. Pflügers Arch 429: 592–
594, 1995.
Bottinelli R and Reggiani C. Human skeletal muscle fibres:
molecular and functional diversity. Prog Biophys Mol Biol 73:
195–262, 2000.
Canepari M, Rossi R, Pellegrino MA, Reggiani C, and
Bottinelli R. Speeds of actin translocation in vitro by myosins
extracted from single rat muscle fibres of different types. Exp
Physiol 84: 803–806, 1999.
Crockford T, Johnston IA, and McAndrew BJ. Functionally
significant allelic variation in myosin light chain composition in
a tropical cichlid. J Exp Biol 198: 2501–2508, 1995.
Edman KAP. Double-hyperbolic force-velocity relation in frog
muscle fibres. J Physiol (Lond) 404: 301–321, 1988.
Edman KAP and Reggiani C. Redistribution of sarcomere
length during isometric contraction of frog muscle fibres and its
relation to tension creep. J Physiol (Lond) 351: 169–198, 1984.
Edman KAP, Reggiani C, Schiaffino S, and teKronnie G.
Maximum velocity of shortening related to myosin isoform composition in frog skeletal muscle fibres. J Physiol (Lond) 395:
679–694, 1988.
Edman KAP, Reggiani C, and teKronnie G. Differences in
maximum velocity of shortening along single muscle fibres of the
frog. J Physiol (Lond) 365: 147–163, 1985.
Geiger PC, Cody MJ, Macken RL, and Sieck GC. Maximum
specific force depends on myosin heavy chain content in rat
diaphragm muscle fibers. J Appl Physiol 89: 695–703, 2000.
Gordon AM, Huxley AF, and Julian FJ. The variation in
isometric tension with sarcomere length in vertebrate muscle
fibers. J Physiol (Lond) 184: 170–192, 1966.
Greaser ML, Moss RL, and Reiser PJ. Variations in contractile properties of rabbit single muscle fibres in relation to troponin T isoforms and myosin light chains. J Physiol (Lond) 406:
85–98, 1988.
Guilford WH, Dupuis DE, Kennedy G, Wu J, Patlak JB,
and Warshaw DM. Smooth muscle and skeletal muscle myosins produce similar unitary forces and displacements in the
laser trap. Biophys J 72: 984–986, 1997.
Hilber K and Galler S. Mechanical properties and myosin heavy
chain isoform composition of skinned skeletal muscle fibres from a
human biopsy sample. Pflügers Arch 434: 551–558, 1997.
Hill AV. The heat of shortening and the dynamic constants of
muscle. Proc R Soc Lond B Biol Sci 126: 136–195, 1938.
Hook P and Larsson L. Actomyosin interactions in a novel
single muscle fiber in vitro motility assay. J Muscle Res Cell
Motil 21: 357–365, 2000.
Lannergren J. Contractile properties and myosin isoenzymes
of various kinds of Xenopus twitch muscle fibres. J Muscle Res
Cell Motil 8: 260–273, 1987.
Lannergren J. Fibre types in Xenopus muscle and their functional properties. In: Muscular Contraction, edited by Simmons
RM. London: Cambridge University Press, 1992, p. 181–188.
Lannergren J and Hoh JFY. Myosin isoenzymes in single
muscle fibres of Xenopus laevis: analysis of five different functional types. Proc R Soc Lond B Biol Sci 222: 401–408, 1984.
Lannergren J and Smith RS. Types of muscle fibres in toad
skeletal muscle. Acta Physiol Scand 68: 263–274, 1966.
Larsson L, Li X, and Frontera WR. Effects of aging on shortening velocity and myosin isoform composition in human single
skeletal muscle cells. Am J Physiol Cell Physiol 272: C638–
C649, 1997.
Larsson L and Moss RL. Maximum velocity of shortening in
relation to myosin isoform composition in single fibres of human
skeletal muscle. J Physiol (Lond) 472: 595–614, 1993.
AJP-Cell Physiol • VOL
25. Li X and Larsson L. Maximum shortening velocity and myosin
isoforms in single muscle fibers from young and old rats. Am J
Physiol Cell Physiol 270: C352–C360, 1996.
26. Lowey S, Waller GS, and Trybus KM. Function of skeletal
muscle myosin heavy chain and light chain isoforms by an in
vitro motility assay. J Biol Chem 268: 20414–20418, 1993.
27. Lutz GJ, Bremner SN, Bade MJ, and Lieber RL. Identification
of myosin light chains in Rana pipiens skeletal muscle and their expression patterns along single fibres. J Exp Biol 204: 4237–4248, 2001.
28. Lutz GJ, Bremner SN, Lajevardi N, Lieber RL, and Rome
LC. Quantitative analysis of muscle fiber type and myosin heavy
chain distribution in the frog hindlimb: implications for locomotory design. J Muscle Res Cell Motil 19: 717–731, 1998.
29. Lutz GJ, Cuizon DB, Ryan AF, and Lieber RL. Four novel
myosin heavy chain transcripts define a molecular basis for muscle
fibre types in Rana pipiens. J Physiol (Lond) 508: 667–680, 1998.
30. Lutz GJ and Lieber RL. Myosin isoforms in anuran skeletal
muscle: their influence on contractile properties and in vivo
muscle function. Microsc Res Tech 50: 443–457, 2000.
31. Lutz GJ, Razzaghi S, and Lieber RL. Cloning and characterization of the S1 domain of four myosin isoforms from functionally divergent fiber types in adult Rana pipiens skeletal muscle.
Gene 250: 97–107, 2000.
32. Lutz GJ and Rome LC. Built for jumping: the design of frog
muscular system. Science 263: 370–372, 1994.
33. Lutz GJ and Rome LC. Muscle function during jumping in
frogs. I. Sarcomere length change, EMG pattern, and jumping
performance. Am J Physiol Cell Physiol 271: C563–C570, 1996.
34. Lutz GJ and Rome LC. Muscle function during jumping in
frogs. II. Mechanical properties of muscle: implications for system design. Am J Physiol Cell Physiol 271: C571–C578, 1996.
35. Marsh RL. Jumping ability of anuran amphibians. In: Comparative Vertebrate Exercise Physiology, edited by Jones JH. New
York: Academic, 1994, p. 51–111.
36. Moss RL, Diffee GM, and Greaser ML. Contractile properties
of skeletal muscle fibers in relation to myofibrillar protein isoforms. Rev Physiol Biochem Pharmacol 126: 1–63, 1995.
37. Moss RL, Giulian GG, and Greaser ML. Physiological effects
accompanying the removal of myosin LC2 from skinned muscle
fibers. J Biol Chem 257: 8588–8591, 1982.
38. Powell PL, Roy RR, Kanim P, Bello MA, and Edgerton VR.
Predictability of skeletal muscle tension from architectural determinations in guinea pig hindlimbs. J Appl Physiol 57: 1715–1721,
1984.
39. Reiser PJ, Greaser ML, and Moss RL. Myosin heavy chain
composition of single cells from avian slow skeletal muscle is
strongly correlated with velocity of shortening during development. Dev Biol 129: 400–407, 1988.
40. Reiser PJ, Greaser ML, and Moss RL. Contractile properties
and protein isoforms of singe fibres from the chicken pectoralis
red strip muscle. J Physiol (Lond) 493: 553–562, 1996.
41. Rowlerson AM and Spurway NC. Histochemical and immunohistochemical properties of skeletal muscle fibres from Rana
and Xenopus. Histochem J 20: 657–673, 1988.
42. Schiaffino S and Reggiani C. Molecular diversity of myofibrillar proteins: gene regulation and functional significance. Physiol
Rev 76: 371–423, 1996.
43. Smith RS and Ovalle WK Jr. Varieties of fast and slow
extrafusal muscle fibres in amphibian hindlimb muscles. J Anat
116: 1–24, 1973.
44. Sweeney HL, Kushmerick MJ, Mabuchi K, Sreter FA, and
Gergely J. Myosin alkali light chain and heavy chain variations
correlate with altered shortening velocity of isolated skeletal
muscle fibers. J Biol Chem 263: 9034–9039, 1988.
45. Tikunov BA, Sweeney HL, and Rome LC. Quantitative electrophoretic analysis of myosin heavy chains in single muscle
fibers. J Appl Physiol 90: 1927–1935, 2001.
46. Widrick JJ, Romatowski JG, Bain JL, Trappe SW, Trappe
TA, Thompson JL, Costill DL, Riley DA, and Fitts RH.
Effect of 17 days of bed rest on peak isometric force and unloaded
shortening velocity of human soleus fibers. Am J Physiol Cell
Physiol 273: C1690–C1699, 1997.
47. Yates LD and Greaser ML. Quantitative determination of myosin and actin in rabbit skeletal muscle. J Mol Biol 168: 123–141, 1983.
282 • APRIL 2002 •
www.ajpcell.org