Abstract
purpose. To gain insights into the functional significance of myosin heavy-chain
(MyHC) heterogeneity by comparing the mechanical kinetic properties of
single rabbit extraocular muscle (EOM) fibers with those of limb
fibers. EOMs are known to contain developmental and EOM-specific MyHCs
in addition to those present in limb muscles, and MyHCs profoundly
influence muscle mechanics.
methods. Isometric cross-bridge kinetics were analyzed in
Ca2+-activated single glycerinated fibers from rabbit EOM
and limb fast and slow muscles at 15°C by means of mechanical
perturbation analysis. The plots of stiffness and phase against
frequency display a characteristic frequency, f min, at which stiffness is minimum, and
phase shift is zero. The value of f min is
independent of Ca2+ or force level but reflects the
kinetics of cross-bridge cycling.
results. Analysis of 121 limb fast fibers gave f min values ranging from 10 to 26 Hz. f min for
the 10 slow soleus fibers was 0.5 Hz. Analysis of 170 EOM fibers gave f min values in the range for fast limb
fibers, but in addition yielded f min values
below (4–9 Hz) and above (27–33 Hz) this range.
conclusions. The wider range of mechanical kinetic characteristics in EOM fibers
compared with limb fibers is likely due to the expression of
developmental (low f min) and EOM-specific
(high f min) MyHCs in addition to isoforms
present in adult limb muscles. The considerable diversity of functional
characteristics in EOM fibers is likely to be important for rotating
the eyeball at various speeds during tracking and for executing
saccades over a wide range of angles.
Extraocular muscles (EOMs) have a diverse repertoire of functions
including steady eyeball fixation, slow vergence movements, pursuit
movements at various speeds, and high-speed saccades over a wide range
of angles.
1 These functional intricacies are reflected in
the complex fiber types seen in EOMs. EOM fibers have been classified
into six types on the basis of histochemistry, ultrastructure, and
innervation.
2 This system of classification differs
markedly from that used to classify functionally different limb muscle
fibers, the latter subserving locomotion and posture. Limb muscle
fibers are classified into four types: type I (or slow) and three
subtypes (IIA, IIX, IIB) of type II or fast fibers, each expressing a
different isoform of myosin heavy chain (MyHC).
3 An
indicator of the complexity of EOMs is seen in the fact that nine
different myosin heavy chains (MyHCs) have been identified in these
muscles in adult animals. These include isoforms found in adult limb
fast (IIA-, IIX-, IIB-MyHC) and slow (type I-MyHC, synonymous with
cardiac β-MyHC) fibers,
4 5 but in addition, MyHCs found
in developing (embryonic- and fetal/neonatal-MyHCs) but not in mature
limb muscles,
4 5 the cardiac-specificα
-MyHC
6 7 as well as two EO-specific isoforms: the
EO-MyHC
8 9 and the slow-tonic MyHC.
10
Through its cyclic interaction with actin during muscle contraction,
myosin controls the kinetics of energy transduction from ATP into
mechanical work. In limb muscle fibers, the speed of contraction and
thus the power and efficiency of muscle fibers are controlled
principally by the type of MyHC.
11 The complexity of MyHC
types found in EOMs suggests that mechanical properties of single
fibers in these muscle should be correspondingly complex. There is
little information in the literature on mechanical properties of single
EOM fibers. Published works on contractile properties of EOMs are
limited to analyses of isometric and isotonic contractile
characteristics of whole EOMs.
12 These show that isometric
contraction times are very short while the twitch:tetanus tension ratio
is low, compared with fast limb muscles of the same species. The
maximal speed of shortening (
V max) is
generally higher than that of the fastest limb muscle in the same
animal. For the rabbit,
V max of EOM is
42% higher than that of the extensor digitorum longus, a limb fast
muscle.
12 Whole muscle mechanical data give little insight
into the functional significance of MyHC complexity. Single-fiber
analysis of EOM has been limited to the measurement of isometric force
at various Ca
2+ and Sr
2+ concentrations.
13 Studies on cross-bridge cycling kinetics
have not been reported for single EOM fibers.
The mechanical characteristics of single muscle fibers can be analyzed
using low amplitude length oscillations at various frequencies to probe
the dynamic stiffness of active fibers.
14 This analysis
yields a parameter,
f min, the
frequency at which dynamic stiffness of the fiber is a minimum. The
value of this parameter reflects the kinetics of cross-bridge cycling
in the fiber.
14 15 In this article, we compared the
mechanical characteristics of single rabbit EOM fibers with those of
limb fibers using this method, to gain insights into the functional
significance of MyHC heterogeneity of EOMs.
Muscle fibers from four adult female and two male New Zealand
White rabbits were used in these experiments. The use of these animals
was approved by the Animal Ethics Committee of our institution and
adhered to the ARVO Statement for use of animals. The animals
were killed by sodium pentobarbitone overdose. Muscle bundles were
dissected from EOMs, vastus lateralis (IIA, IIX, IIB fibers), psoas
(predominantly 2X fibers), extensor digitorum longus (predominantly 2A,
2X fibers), and soleus (slow fibers) muscles. The dissected muscle
bundles were pinned out on a flat wooden stick and kept in skinning
solution and agitated continuously for 24 hours at 4°C. After 24
hours the muscles were placed in storage solution and kept at −20°C
and used within 4 weeks.
Skinning solution contained (in mM): 5 MgCl2, 60
HEPES (pH 6.7), 108 Na acetate, 5
NaH2PO4, 2.5 EGTA (stock
solution adjusted with NaOH to pH 7), 0.01% (g/ml)
NaN3, 0.01% (g/ml) DTT, 50% (v/v) glycerol.
Storage solution contained (in mM): 5 MgCl2 60
HEPES (pH 6.7), 108 K acetate, 5
KH2PO4, 2.5 EGTA (stock
solution adjusted with KOH to pH 7), 0.01% (g/ml)
NaN3, 0.01% (g/ml) DTT, 50% (v/v) glycerol.
Both relaxing and activation solutions contained (in mM): 7 EGTA
(stock solution adjusted with KOH to pH 7.0), 5.26
MgCl
2, 20 imidazole, 5
KH
2PO
4, 5 ATP, 20 creatine
phosphate. In addition, activation solution contained 7.36 mM
CaCl
2. Because of the relatively short shelf life
of creatine phosphokinase, it was kept at −20°C and added directly
to the activation solution in the fiber bath as required (1 mg/ml). The
pH of both solutions was adjusted to 7.0 by adding either KOH or HCl at
room temperature. The ionic strength of the solution was determined by
the algorithm of Perrin and Sayce.
16 KCl was used to bring
the ionic strength to the required level. In this study, the ionic
strength was 173 mM in relaxing solution and 168 mM in activation
solution. Both solutions were stored at −20°C.
Single muscle fibers, typically 2 to 3 mm in length, were randomly
teased from small glycerinated bundles at 0°C by means of jeweler’s
forceps. One end of the fiber was glued onto a force gauge (AE801;
SensoNor, Horten, Norway) and the other end onto a length
driver (P-841.10; Physik Instruments, Waldbronn, Germany). The fiber
was viewed through an inverted microscope (Fluorovert; Leitz, Wetzlar,
Germany) at a magnification of ×400, and the sarcomere length of the
fiber was adjusted to 2.5 μm by means of a calibrated eyepiece
micrometer. The temperature of the fiber bath solution was maintained
by means of a peltier module (KSM-0617; Komatsu Electronics, Tokyo,
Japan) and a temperature sensor (AD 590; Analog Devices,
Norwood, MA), which provided a feedback signal for a
custom-built proportional-integral controller. A digital thermometer
(KM-330; Kane Instruments, Bedford, MA) was located
in the fiber bath solution to provide an independent and continued
read-out of temperature.
Fiber length was perturbed by a signal that was software-generated and
introduced to the fiber via D/A conversion and the length driver. The
form of the signal was pseudorandom binary noise (PRBN), which enabled
the calculation of stiffness and phase spectra with greater resolution
and in less time compared with the sinusoidal technique.
14 The amplitude of the length signal was typically 0.05% of the fiber
length.
Single fibers were incubated for approximately 5 minutes in the muscle
bath filled with relaxing solution. Activation of the fiber was
achieved by changing from the well with relaxing solution to one
containing activation solution. The Ca2+ concentration of activation solution was pCa 4.0, which ensured maximal
activation of the fiber. The temperature of the bath solution was set
at 15°C. When steady tension had been attained, fiber length was
perturbed with the PRBN signal. The length changes together with the
resulting force responses of the fiber were sampled by the control
computer via A/D conversion. Fast Fourier transforms of the length and
force data yielded the stiffness and phase values. The stiffness and
phase data were smoothed, using a three-point convolution procedure,
and displayed on a digital plotter (7225A; Hewlett Packard, Palo Alto,
CA). f min was evaluated from
the dynamic stiffness plots by noting the frequency at which stiffness
was at the minimum.
The dynamic stiffness and phase plots of single fibers from EOM
displayed the characteristic minimum in stiffness and the accompanying
mini-max feature in the phase curve.
Figure 1 show respectively the dynamic stiffness
(Fig. 1A) and phase
(Fig. 1B) plots of two EOM single fibers whose
f min values were located at the two
extremes of the range for the 170 fibers investigated. For comparison,
Figures 1C and 1D shows stiffness and phase plots of representative
single limb fast fibers, respectively.
Figure 2 shows the distribution of
f min values
for 121 limb and 170 EOM fibers. The
f min values for the single soleus
fibers
(Fig. 2A) were 0.56 ± 0.05 Hz (mean ± SD,
n = 10). Values for single psoas fibers
(Fig. 2B) ranged from 11 to 14 Hz (12.3 ± 1.2 Hz,
n = 33) and those for the EDL
(Fig. 2C) ranged from 10 to 16 Hz, (12.8 ± 1.6 Hz,
n = 24). Values for the vastus
lateralis fibers
(Fig. 2D) showed a wider range, from 10 to 26 Hz
(
n = 54), and showed two distinct populations with
f min ranges 10 to 17 Hz (14.0 ±
1.3 Hz,
n = 36) and 21 to 26 Hz (23.8 ± 1.6 Hz,
n = 18).
EOM fibers
(Fig. 2E) have
f min values
that not only span the full range of values found in limb fast fibers
but also extended below and above this range at 4 to 9 and 27 to 35 Hz,
respectively. Approximately 5% of extraocular fiber investigated in
this study showed no minimum in their dynamic stiffness at 0.1 Hz or
higher.
The use of frequency domain analysis of muscle mechanics was
introduced as an alternative to classical force:velocity measurements
for characterizing muscle mechanics in dealing with insect flight
muscles, which are intrinsically oscillatory. This approach is
particularly advantageous for insect flight muscles because their
sarcomeres are unable to undergo the long-range sliding characteristic
of vertebrate striated muscles necessary for force:velocity
measurements.
17 The frequency domain method for
characterizing isometric cross-bridge mechanics was found to be
generally applicable to vertebrate skeletal
14 18 and
cardiac
19 20 muscles.
In the original method of analyzing the dynamic stiffness of
contracting muscle, sinusoidal length changes over a range of
frequencies were applied to the muscle sequentially, and the resulting
near sinusoidal force and phase changes at each frequency were used to
derive the dynamic stiffness characteristics. These dynamic stiffness
characteristics can be related to the time courses of tension
transients in responses to rectangular changes of muscle
length.
21 For very small amplitudes, these two
methodologies are approximately related through the Fourier transform.
In terms of this relationship, it can be deduced that
1/
f min correlates with the time
t 2 taken to complete Phase 2 of such a
tension step transient.
The method used in this article differed from the classical method in
that the applied oscillations of muscle length took the form of PRBN.
This signal has encoded within itself the full range of frequencies of
interest. Fourier analysis of the PRBN length oscillations and the
resulting interrupted tension transients were used to derive the
dynamic stiffness values.
14 The advantages of this method
are that it gives a high-resolution plot of the data as a function of
frequency and reduces the data acquisition time. The method has been
shown to give comparable results to the classical sinusoidal
method.
14
It is well established that dynamic stiffness and phase parameters
of contracting muscle reflect cross-bridge cycling
rates.
15 21 22 23 In a three-state (detached, attached but
not force-generating, attached and force-generating), cross-bridge
model,
f min is sensitive to rate
constants for the power-stroke and the cross-bridge detachment rate
during isometric contraction.
15 At 25°C,
f min values range from 60 Hz for
insect flight muscles
22 to 1 to 2 Hz for mammalian cardiac
muscle.
20 The value is correlated with the MyHC structure
and ATPase activity of the myosin.
20 This is well
illustrated in rat cardiac muscle. Rat ventricular muscle may contain
V1 or V3 myosin, depending on the thyroid state. V1 is composed
of two α-MyHCs and has a high ATPase activity, whereas V3 myosin is
composed of two β-MyHCs and has low ATPase activity. The ratio of
f min values for V1 and V3 is 2:1, the
same as the ratio of their myosin ATPase activities
24 and
V max.
25 f min is independent of the level of
muscle force
20 but is sensitive to the temperature at
which measurements are made.
26
The correlation between complex stiffness and tension transients is
further supported from simulation studies, where
t 2 was also demonstrated to share the
sensitivity of
f min to changes in the
rate constant of the power stroke and to changes in the isometric
detachment rate.
15 In contrast,
V max is principally driven by the
cross-bridge detachment rate encountered during large-scale filament
sliding.
15 This detachment rate constant is faster than
the corresponding rate constant governing detachment from the isometric
state and is not encountered during small-amplitude perturbations of
the isometric state. Therefore, it is at least possible for
V max and the small-amplitude isometric
mechanics to diverge in their sensitivities to changes in myosin
isoforms.
Slow fibers from the soleus muscle that express β-MyHC gave an
f min value of approximately 0.5 Hz at
15°C, consistent with the value of 1.3 Hz at 24°C found for rabbit
ventricular muscle having the same MyHC.
27
Fast limb muscle fibers express 2A, 2X, and 2B MyHC isoforms. Fibers
containing these isoforms are associated with different force:velocity
relations,
3 consistent with the influence of MyHC on
muscle mechanics. Our results in limb fast muscle showed
f min values between 10 and 26 Hz.
Rabbit psoas muscle has 96% IIX MyHC, and the predominant MyHCs in the
extensor digitorum longus are IIA and IIX, whereas the vastus
lateralis has significant amounts of all three fast
MyHCs.
28 Comparing the fiber type predominance
of individual muscles with the
f min values obtained from single fibers from these muscles, it is likely
that 2A/2X fibers have
f min values
between 10 and 17 Hz, whereas 2B fibers have values between 19 and 26
Hz. The correlation of
f min with MyHC
isoforms is further supported by the correlation of MyHC isoforms with
the tension step transient parameter
t 2.
29 30 31 Because IIB
fibers have the highest and IIA fibers the lowest
V max values, there is a broad
correlation between
f min and
V max in skeletal muscle fibers, as
found for cardiac muscle discussed above.
In this study, the distributions of
f min values in 170 single fibers were
wider than that of limb fast fibers. In addition to
f min values in the range 10 to 26 Hz
found in limb fast fibers, there were
f min values below (4–9 Hz) and above
this range (27–36 Hz). Values within the range found for limb fast
muscle fibers are most likely due to the presence of fibers containing
2A, 2X, and 2B MyHCs. Coexpression of multiple isoforms of limb MyHC in
EOM fibers is likely to produce only
f min values within the range found in
limb fibers. Values above and below this range found in EOM fibers are
likely due to MyHCs found in EOM fibers only, namely, the developmental
and EOM-specific isoforms. Although myosin light chains and other
myofibrillar protein are able to affect mechanical properties of muscle
fibers, rabbit EOMs are known to express fast isoforms of light chains,
which are indistinguishable from those in limb fast
muscle.
32 Rabbit EOMs also express the same isoforms of
tropomyosin and troponin T as limb muscle fibers, though the dominant
isoform of TnT is TnT
3f, which is a minor
component in limb fast fibers.
33
During early postnatal life, limb muscles express embryonic and
fetal/neonatal myosins. These myosins are progressively replaced by the
appropriate MyHCs during the first few weeks of life.
34 35 Functionally, all limb muscles in newborn animals are slow contracting.
During the first few weeks of postnatal life, the developing fast
muscle increases in speed of shortening by a factor or 2 to 3, whereas
the speed of shortening of slow muscles remain
unchanged.
36 37 The developmental isoforms of MyHC are
therefore associated with slower speed of contraction relative to fast
isoforms. We suggest that
f min values
of 4 to 9 Hz are due to the persistence of embryonic and fetal/neonatal
isoforms of MyHC in adult EOM fibers.
The
V max of rabbit EOM is 42% faster
than limb fast muscle, and this has been attributed to the presence of
EO-MyHC.
12 In view of the correlation between
f min and
V max, we may attribute the presence of
f min values above 26 Hz to the
presence of this EO-MyHC.
We found that 5% of EOM fibers did not display a stiffness minimum in
the range of 0.1 to 100 Hz. These fibers are probably slow-tonic fibers
with very slow cross-bridge kinetics.
Our results show that single EOM fibers have an extremely wide
range of mechanical characteristics. This result represents a
considerable advance in our understanding of EOM mechanics. The
force:velocity relation of rat EOM is complex, deviating from the
classical Hill’s hyperbolic relation, but can be fitted well by an
exponential function with three constants.
38 In the light
of the current work, it is likely that the complexity of whole EOM
mechanics is due to fibers with different intrinsic speeds contracting
in parallel.
The diversity of f min values of single
EOM fibers is no doubt generated by the expression in single fibers of
myosins with different kinetic properties. Clearly, to provide the
means of generating such functional diversity is likely to be the
explanation for the well known but puzzling complexity of MyHC
composition of EOMs. Presumably such functional diversity is found also
at the motor unit level. The oculomotor system would thus have at its
disposal motor units with a diverse range of speeds and thus have the
potential for recruiting appropriate motor units for the task at hand.
We suggest that for tracking movements at various angular speeds, motor
units of matching speeds are recruited. The work involved in rotating
the eyeball during a saccade will differ depending on the angle through
which the eyeball has to rotate. We suggest that slower units are
recruited for low-angle saccades, whereas faster and thus more powerful
units are recruited for wide-angle saccades.