| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Pharmacology, Division of Neuroscience, University of Alberta, Edmonton T6G 2S2, Canada
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Gordon, T., N. Tyreman, V. F. Rafuse, and J. B. Munson. Fast-to-slow conversion following chronic low-frequency activation of medial gastrocnemius muscle in cats. I. Muscle and motor unit properties. J. Neurophysiol. 77: 2585-2604, 1997. This study of cat medial gastrocnemius (MG) muscle and motor unit (MU) properties tests the hypothesis that the normal ranges of MU contractile force, endurance, and speed are directly associated with the amount of neuromuscular activity normally experienced by each MU. We synchronously activated all MUs in the MG muscle with the same activity (20 Hz in a 50% duty cycle) and asked whether conversion of whole muscle contractile properties is associated with loss of the normal heterogeneity in MU properties. Chronically implanted cuff electrodes on the nerve to MG muscle were used for 24-h/day stimulation and for monitoring progressive changes in contractile force, endurance, and speed by periodic recording of maximal isometric twitch and tetanic contractions under halothane anesthesia. Chronic low-frequency stimulation slowed muscle contractions and made them weaker, and increased muscle endurance. The most rapid and least variable response to stimulation was a decline in force output of the muscle and constituent MUs. Fatigue resistance increased more slowly, whereas the increase in time to peak force varied most widely between animals and occurred with a longer time course than either force or endurance. Changes in contractile force, endurance, and speed of the whole MG muscle accurately reflected changes in the properties of the constituent MUs both in extent and time course. Normally there is a 100-fold range in tetanic force and a 10-fold range in fatigue indexes and twitch time to peak force. After chronic stimulation, the range in these properties was significantly reduced and, even in MU samples from single animals, the range was shown to correspond with the slow (type S) MUs of the normal MG. In no case was the range reduced to less than the type S range. The same results were obtained when the same chronic stimulation pattern of 20 Hz/50% duty cycle was imposed on paralyzed muscles after hemisection and unilateral deafferentation. The findings that the properties of MUs still varied within the normal range of type S MUs and were still heterogeneous despite a decline in the variance in any one property indicate that the neuromuscular activity can account only in part for the wide range of muscle properties. It is concluded that the normal range of properties within MU types reflects an intrinsic regulation of properties in the multinucleated muscle fibers.
Although phenotypic differences between mammalian slow- and fast-twitch skeletal muscles were recognized more than a hundred years ago (Grutzner 1884 Surgery, electrodes, and chronic stimulation
Nerve cuff electrodes were implanted bilaterally in 14 adult cats (3-4 kg body wt) for chronic stimulation of the MG muscle nerve and monitoring of muscle properties. In 4 of the 14 animals, the MG muscle to be stimulated was paralyzed by hemisection of the spinal cord at T13 and unilateral deafferentation (L2-S2) at the time of nerve cuff implantation. In seven other cats, intramuscular (IM) wire electrodes were used for chronic stimulation to control for any possible damage inflicted by the nerve cuff electrodes. These wire electrodes were implanted bilaterally on either side of the nerve as it enters the MG muscle for chronic stimulation. A portable, lightweight (80 g), locally made stimulator was carried on the back of the cat on a specially fabricated basket. The basket was secured through the musculature immediately in front of and behind the L7 spinal process with two Prolene sutures (Kernell et al. 1987a
Chronic muscle recordings
Strength, speed, and fatigability of MG muscle contraction were assessed at intervals of ~1 wk. This was accomplished in the halothane-anesthetized cats by recording the force or torque of ankle extension in response to supramaximal stimulation of the MG muscle nerve as described previously (Davis et al. 1978 Acute experiments: muscle and MU recording
Stimulated muscle and MU properties were analyzed in terminal acute experiments after 37-240 days of chronic stimulation and compared with MG data from 10 age-matched normal cats. Of the 21 experimental animals, ventral root splitting in 11 cats was carried out at the University of Alberta, Edmonton, to isolate single MUs for recording of conduction velocity and contractile properties. In another six cats, microelectrode impalement of MG motoneurons was carried out at the University of Florida, Gainesville, to record membrane electrical properties simultaneously with MU contractile properties (Table 1, experiments ADA, ADB, ADR, ADQ, ACY, and ACZ; Munson et al. 1997
Muscle histochemistry
At the end of the experiments, the MG muscles were removed, cut into five sections in the longitudinal axis, secured to corks with Tissue-Tek mounting media and frozen in frozen isopentane. The tissue blocks were stored at Statistics
Regression lines were fitted to the data with the use of least-mean-squares criteria and the Pearson product-moment correlation coefficients were calculated. X and Y variables were considered to be correlated if the slopes of the regression lines were significantly different from zero with P < 0.05. The Kolmogorov-Smirnov test was used to determine whether MU populations were statistically different with the use of cumulative frequency distributions. One-way analysis of variance (ANOVA) was used to determine statistical significance between mean values. Differences were considered as highly significant when P < 0.01 and significant when P < 0.05.
Chronic recordings of evoked MG muscle force and acute recordings of MU properties were made in a total of 21 experimental and 10 control cats (Table 1). In 14 cats, the MG muscles were stimulated via cuff electrodes placed around the nerve. Stimulation at 20 Hz in either a 50% or 5-6.3% duty cycle was initiated 15.9 ± 4.0 (SE) days after electrode implantation, when stable prestimulation values had been obtained for later comparison with poststimulation values. The cuff electrodes were also used at regular intervals to evoke twitch and tetanic contractions under halothane anesthesia to monitor progressive changes in muscle properties (Fig. 1). In 4 of the 14 animals, the MG muscle was inactivated by hemisection and unilateral deafferentation before chronic stimulation. In seven cats, MG muscles were stimulated via IM wire electrodes, which were placed against the MG nerve as it enters the muscle to control for any damage inflicted on the nerves by the cuff electrodes. Four to 57 MUs were sampled in a final acute experiment in 18 animals.
Chronic muscle stimulation and isometric force recordings in intact cats
Typical recordings of ankle extensor force are shown in Fig. 2 for one cat (Table 1, cat 10) to illustrate the changes in amplitude and time course of chronically stimulated MG muscles (24-h/day, 20-Hz stimulation in a 50% duty cycle). Absolute forces of twitch and tetanic contractions were reduced by 50% 76 days after commencement of stimulation; rates of rise and fall of twitch force were slower (Fig. 2, A and C). The typical sag of the unfused tetanus of the normal MG muscle before daily stimulation was replaced by the slow pattern of a staircase increase in force in the chronically stimulated muscle (Fig. 2B).
Muscle and MU properties in stimulated muscles: acute recording
Muscle and MU properties were studied at different time points after initiation of chronic stimulation. Acute recording of these properties was undertaken to determine whether the magnitude and time course of fast-to-slow conversion of muscle contractile properties accurately reflects a corresponding alteration in the MU population of the stimulated muscles, as suggested by the work of Kernell and colleagues (Eerbeek et al. 1984
Variability in fast-to-slow conversion
Changes in whole muscle properties recorded chronically (Fig. 5) and in the final acute experiment (Fig. 6, A and B) accurately reflected changes in mean values of the same parameters in the constituent MUs (Fig. 6, C and D). The similarities were particularly striking during the first 100 days, when the same variability in mean MU twitch force and TTP was observed between cats as the variability in whole muscle twitch force and TTP. This variability was also reflected in the variable increase in the proportion of type S MUs in the stimulated muscles as defined by their absence of sag during unfused tetanic contractions (Fig. 6F). Normally 20-25% of MG MUs are non-sagging type S MUs (Burke et al. 1973
Remaining variability in MU properties in converted muscles
The full extent of fast-to-slow conversion of MU properties is shown in Fig. 9, where MU data are collated from all the stimulated muscles in which there was complete fiber type II to I conversion. The data are displayed as cumulative sum histograms to compare the distribution of all the MUs in the normal and long-term stimulated muscles with the type S population in control muscles. These comparisons show that MUs in the stimulated muscles develop the same tetanic forces as the normal type S MU population and are almost as fatigue resistant (FR). They are, however, significantly slower: the cumulative sum histograms for TTP and half-relaxation times of the twitches of MUs from the stimulated muscles were shifted far to the right of the normal type S MU population (Kolmogorov-Smirnov test, P < 0.001; see also Fig. 4, A and C, in Munson et al. 1997
Chronic stimulation in hemisected and deafferented hindlimbs
Because MG nerves were chronically stimulated in otherwise normal hindlimbs, it is possible that normal activity during the 2.5-s OFF periods of the 50% duty cycle could account for the remaining heterogeneity in the stimulated MUs. To test for this possibility, the same pattern of stimulation was used in a cat in which voluntary and reflex recruitment were eliminated by hemisection and unilateral deafferentation of the hindlimb.
Correlation of MU force, MU TTP, and conduction velocity
Normally the continuous range of MU force is directly correlated with conduction velocity and inversely correlated with TTP, as shown for a sample of MUs recorded in a control MG muscle in Fig. 12. When MUs underwent progressive fast-to-slow conversion after chronic low-frequency stimulation and MUs become less forceful and slower, the same relationships were maintained. However, the smaller range of values in long-term stimulated muscles reduces the statistical significance of the correlations.
We have described a complete conversion of a mixed MU population in cat MG muscle to type S within 80 days of stimulation at 20 Hz in a 50% duty cycle. This pattern of stimulation corresponds to the "high" activity pattern shown previously to change the muscle properties of the fast-twitch cat peroneus longus to resemble a slow-twitch muscle (Eerbeek et al. 1984 Fast-to-slow conversion of MG muscle
Our finding that low-frequency stimulation of cat MG muscles in a 50% duty cycle leads to slowing of the fast-twitch muscle contraction is consistent with the slowing of the cat peroneus longus muscle when the same pattern of stimulation is used (Eerbeek et al. 1984 FORCE.
The decline in whole muscle twitch and tetanic forces is well explained by a parallel decline in the average force of the composite MUs; the wide spectrum of MU forces was reduced to the range normally seen for the type S MUs in unstimulated MG muscles (Fig. 9). The restriction of MU forces to the type S range was complete by 56 days even though a substantial proportion of the MUs still exhibited sag during unfused tetani and the time course of contraction was in the fast range (see also Rafuse et al. 1997 ENDURANCE.
In contrast to the rapid decline in muscle and MU forces, endurance index of the whole muscle and average FI of the constituent MUs increased more slowly to reach slow values (Figs. 5 and 6). Muscle endurance begins to increase within the first weeks of stimulation (see also Hudlicka et al. 1977 SPEED.
The effect of stimulation on muscle contraction time was the most variable among animals. Muscle stimulated at 20 Hz in a 50% duty cycle became progressively slower to attain the contraction speed of the synergistic slow soleus muscle (see Fig. 1 and Table 1 in Munson et al. 1997 Matching properties of nerve and muscle
The effects of chronic stimulation on the matching properties are analyzed and discussed in the companion paper (Munson et al. 1997
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Ranvier 1874), elucidation of the striking heterogeneity of muscle fibers and motor units (MUs) within muscles has had a major impact on our current understanding of muscle function (Burke 1981; Henneman and Mendell 1981; Vrbova et al. 1995). The slow and fast muscle fibers first distinguished by different pH sensitivities of mATPase (Engel 1962) have more recently been subdivided into four types on the basis of constituent myosin heavy chain isoforms and oxidativeglycolytic potential (Pette and Staron 1990). Muscle fiber classification corresponds well with physiological subdivisions of MU types (Burke et al. 1973; Totosy de Zepetnek et al. 1992b). Nevertheless, muscle fibers and MUs form a continuum with respect to any one property. For example, MUs form a continuum with respect to either contractile force or speed (Henneman and Mendell 1981). Motoneurons have also been subclassified on the basis of characteristic groupings of different membrane electrical properties (Zengel et al. 1985).
). The relationship between afterhyperpolarization duration of the motoneurons' action potential and MU contractile speed is the basis for the gradation of MU force by the firing rate of motoneurons (Kernell 1992). The relationships between motoneuron excitability and firing patterns, MU force, and fatigability underlie the ability of most muscles to sustain low forces without fatigue and to develop maximal forces for only short periods of time (Kernell 1992).
,b; Gordon and Stein 1982a,b; Gordon et al. 1986, 1988). Because reinnervated MU muscle fibers are initially heterogeneous in composition and mismatched to their innervating motor nerve, the rematching involves considerable adaptation within the MU.
to explain their findings of fast-to-slow and slow-to-fast conversion of muscle properties after cross-reinnervation. The first was that different nerves exert trophic influences on muscle fibers and the second was that reinnervated muscle fibers respond to the novel activation by the foreign innervation. Findings that chronic stimulation of fast-twitch muscles with low-frequency tonic stimulation mimics the effects of cross-reinnervation of fast muscles favored the second explanation (Ausoni et al. 1990; Salmons and Sreter 1976; Salmons and Vrbova 1969). The low-frequency pattern was chosen as typical of the firing patterns of the slow postural soleus muscle. Indeed, the extensive phenotypic changes described in chronically stimulated muscles which include conversion of contractile, regulatory, and sarcoplasmic reticular proteins to slow isoforms, upregulation of oxidative enzymes, and concurrent downregulation of glycolytic enzymes, are consistent with this explanation (Pette and Vrbova 1992; Vrbova et al. 1995). However, these experiments did not fully consider the normal heterogeneity of muscles and the range in properties of their constituent MUs.
observed to their surprise that, irrespective of frequency, chronically stimulated cat peroneus longus muscles became slowly contracting if the muscles were active for 50% of each day. Because the normal order of recruitment of MUs is associated with a progressive decline in the total daily neuromuscular activity from the smallest and most endurant slow (type S) MUs to the largest and most fatigable fast (type F) MUs, the authors reasoned that the normal range of MU properties and the four MU types are closely correlated with the neuromuscular activity that the MUs normally experience (see Kernell 1992). In support of this suggestion, the authors observed that the chronically stimulated muscles exhibited the same contractile properties as type F fatigue-resistant (FR) MUs when they were activated for 5% of each day and the properties of fast-fatigable (FF) MUs when activation was 0.5% (Kernell et al. 1987a,b). Although these findings were not verified at the single-MU level, the observation that the muscles that exhibited the contractile properties of type S MUs contained only slow oxidative muscle fibers provided supporting evidence (Donselaar et al. 1987).
).
), which permits the comparison of chronically stimulated MU properties with those of type S MUs in the same muscle. In addition, MG is one of three synergistic muscles of the triceps surae in which soleus muscle is composed only of type S MUs and lateral gastrocnemius is composed primarily of type F MUs (85-95%; Foehring et al. 1987; Gordon et al. 1986, 1988). Thus comparisons can be made between experimental stimulated MG and its synergistic unstimulated slow- and fast-twitch muscles in the same animals (see Munson et al. 1997).
; Rafuse et al. 1991) or summarized briefly in recent reviews (Gordon 1995; Gordon and Mao 1994; Gordon et al. 1993).
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
,b).
,b). Two of the three electrodes were used to stimulate the MG nerve in a bipolar configuration.
View larger version (15K):
[in a new window]
FIG. 1.
Diagrammatic representation of the indwelling cuff electrodes around the medial gastrocnemius (MG) nerve for chronic stimulation of MG motor units (MUs) and monitoring of progressive changes in MG muscle properties. The cuff electrodes were externalized via the skin overlying the lumbosacral cord and attached to an external stimulator that was mounted in a hexelite basket for 24-h stimulation in the awake cat. To monitor the changes in MG muscle properties, the cat was anesthetized with halothane and force or torque of ankle extension was recorded in response to supramaximal nerve stimulation. The knee joint was stabilized between 2 fixed support posts. The foot was coupled to a strain gauge via a boot that was free to rotate about a line coaxial with the ankle joint. The ankle was rotated to a fixed point at which nerve stimulation evoked maximum twitch and tetanic forces. (For further details, see text.)
) via the cuff or wire electrodes commenced within 1 mo of electrode implantation, allowing time for recovery from effects of surgery. The cats were housed in large cages with fleecy linings. Cats fed and drank normally and were not perturbed by the stimulation. The known sensation of vibration and/or movement evoked by stimulation of muscle afferents (Goodwin et al. 1972) did not appear to disturb the cats. In fact, the cats became increasingly attached to the experimenters during the months of stimulation as a result of the frequent handling during the regular changing of batteries and chronic muscle force recordings.
; Gordon and Stein 1982a). Briefly, the MG nerve was stimulated at 2 times threshold and ankle extensor torque was measured with the use of a specially designed boot to secure the foot and to couple it to an FT10 Grass strain gauge. The length of the muscle was adjusted for maximal force by rotating the ankle about a line coaxial with the ankle joint to the appropriate angle and fixing the knee with a support (Fig. 1). Evoked electromyogram responses were also monitored with the use of an implanted bipolar electromyogram pad electrode placed on the surface of the MG muscle (Davis et al. 1978).
).
). The endurance index at 2 min was used in all the experiments for direct comparison with the fatigue indexes (FIs) of the MUs.
). In two of these (Table 1, cats ADA, experiment 5, and ACY, experiment 10) ventral root splitting was also carried out to sample a larger number of muscle units. In four cats, motoneuron properties were recorded without muscle and muscle unit properties (Table 1).
View this table:
TABLE 1.
Properties of whole muscle contractions and MUs
; Rafuse et al. 1992a,b). It was used in this study to obtain a large sample of MUs to adequately represent the MU population for each cat. In brief, the cats were deeply anesthetized with pentobarbital sodium and the spinal cord and appropriate leg musculature was exposed surgically. Whole muscle isometric twitch and tetanic forces were recorded from stimulated and contralateral unstimulated MG muscles. L7 and S1 ventral roots were cut centrally for teasing to stimulate axons to single MUs in the stimulated MG muscles. Filaments were split until stimulation evoked an all-or-none extracellular action potential recorded on the isolated MG nerve, electromyogram potential, and twitch response from the muscle. The filament was then stimulated at 2 times threshold to record twitch and tetanic contractions.
). At regular intervals the muscle twitch in response to stimulation of MG nerve was monitored.
) and 2) to use a property of the MUs that was not one of the variables that were the object of the investigation and subject to parametric rather than nonparametric change. The latter variables include TTP and fatigability. The alternative classification, which uses TTP of the FF units to delineate the type F from the type S MUs (e.g., Foehring et al. 1986b; Gordon et al. 1988), is more difficult to apply when both time to peak force and fatigability are changing as a result of the chronic stimulation. When both the sag and TTP criteria are used together, there are many MUs that are "unclassifiable" under experimental conditions of plasticity (Gordon et al. 1988).
70°C. Sections (10-15 µm) were cut from the middle two blocks. The muscle sections were stained for myosin ATPase (pH 4.5 and 10.5), 
-glycerophosphate dehydrogenase, and 
-nicotin-amide adenine dinucleotide (NADH) to determine the muscle fiber type composition. Muscle fibers were classified as type I on the basis of acid stability and alkali lability of mATPase, high oxidative and low glycolytic staining. Muscle fibers were classified as type IIa when fibers showed alkali stability. After preincubation at acid pH, the type IIa fibers are lighter than those classified as type IIb fibers. Type IIa fibers were further distinguished from type IIb fibers by their high oxidative and glycolytic staining in contrast to the low oxidative, high glycolytic enzyme staining of the type IIb fibers. The type I/IIa/IIb classification according to pH sensitivity of mATPase combined with the distinction on the basis of metabolic profile corresponds well with the classifications of slow oxidative, fast oxidative glycolytic, and fast glycolytic, and the two classification schemes are used interchangeably here (Gordon and Pattullo 1993; Gordon et al. 1988; Totosy de Zepetnek et al. 1992b).
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
View larger version (16K):
[in a new window]
FIG. 2.
Chronic recordings of isometric twitch (A), unfused (B), and fused (C) tetanic contractions in the same MG muscle before and 73 days after initiation of chronic low-frequency, high-duration stimulation (20 Hz in a 50% duty cycle; Table 1, cat 10). A: twitch force and the rate of rise and fall of force was reduced after stimulation [as measured by the time to peak twitch force (TTP; 43 and 81 ms) and half-relaxation time of the twitch (29 and 110 ms) for before and after stimulation, respectively]. B: normal "sag" of the unfused tetanic contraction was replaced by a facilitation of the unfused tetani. Unfused tetani were elicited with pulse intervals of 1.25 times the TTP values of 43 and 81 ms, respectively. C: tetanic force and rate of rise and fall of force declined after chronic stimulation.
View larger version (33K):
[in a new window]
FIG. 3.
Force-frequency relationships in the same MG muscle illustrated in Fig. 2. Isometric force was elicited by 2(A-C) and 5 (D-F) stimulus pulses at progressively longer interpulse intervals in MG muscles before (A and D) and 76 days after (B and E) initiation of chronic stimulation. The slowing of the contraction times of the chronically stimulated muscles was accompanied by a shift of the force-frequency curves to the left to lower frequencies (C and F).
View larger version (33K):
[in a new window]
FIG. 4.
Progressive increase in endurance of a chronically stimulated MG muscle (Table 1, cat 17) demonstrated in a fatigue test recorded before chronic stimulation (A and D) and 21 days (B and E) and 78 days (C and F) after stimulation had commenced. For the fatigue test, the muscle was stimulated at 40 Hz for 330 ms/s for 4 min. Normally, the amplitude of the tetani declines and a contracture develops (A) because relaxation rate declines with fatigue as shown for tetani at 0, 1, and 2 min of the fatigue test in D. Twenty-one days after stimulation, the tetanic force was better maintained (B) and the relaxation rate declined less (E). Seventy-eight days after chronic stimulation, tetanic force was well maintained (C) during the duration of the test with little change in the amplitude or rate of relaxation of the tetani (F).
View larger version (23K):
[in a new window]
FIG. 5.
Whole muscle endurance index (A), TTP (B), and isometric tetanic force (C), which were recorded chronically at regular intervals from 10 different MG muscles (nerve cuff animals in Table 1), plotted as a function of duration of daily chronic low-frequency, high-duration electrical stimulation (20 Hz in a 50% duty cycle). The increase in endurance and TTP and the decline in tetanic force were fitted with single-exponential curves. Asymptotic values were 1.1 (A), 82 ms (B), and 0.40 (C), and time constants (time to reach 33% of asymptotic values) were 42 days (A), 49 days (B), and 13 days (C). The curves fitted to all the data provide a reasonable fit to the data obtained from a single MG muscle (cat 17), shown with the use of filled circles. Comparison of the filled and open circles indicates that the variability, particularly in the time course of the twitch contraction (B), is due to interanimal variability rather than intra-animal variation in recordings.
; Kernell et al. 1987a,b).
).
View larger version (20K):
[in a new window]
FIG. 6.
Muscle twitch force (A), muscle TTP (B), and mean ± SE values of MU twitch force (C), MU TTP (D), MU fatigue indexes (FIs; E), and proportion of MUs not demonstrating sag (F) plotted as a function of time of daily chronic stimulation at 20 Hz in a 50% duty cycle. Data from normal MG muscles from 10 control animals (Con) are compared with data from MG muscles in which the muscles were either stimulated via a nerve cuff electrode (N stim) or via intramuscular (IM) wire electrodes (IM stim). Note that the changes in MU parameters parallel the whole muscle values and that the proportion of non-sagging MUs shows good correspondence with changes in mean MU twitch TTP. Muscle and MU properties stabilized within 100 days after initiation of stimulation.
; Foehring et al. 1986a,b; Gordon et al. 1988). In the control animals in which type S MUs were preferentially sampled (>40%) and the animals in which MG muscles were stimulated, there was a weak correspondence between the proportion of non-sagging MUs and MU TTP (Fig. 6, D and F).
View larger version (29K):
[in a new window]
FIG. 7.
Frequency histograms of MU FIs and twitch TTP in normal MG control muscles (A and B) and muscles stimulated for 100 days: mean ± SE = 64 ± 7 days (n = 8) with nerve cuff electrodes (C and D) and 55 ± 7 days (n = 6) with IM wire electrodes (E and F). [Duration of stimulation was not statistically different: analysis of variance (ANOVA), P > 0.05.] Slow (type S), fatigue-resistant (FR), and fast-fatigable (FF) + fatigue-intermediate (FI) MUs are indicated by different shadings. Mean MU values ( 
) for cuff (61 ± 1.8 ms and 0.94 ± 0.02, n = 92) and IM (58 ± 1.5 ms and 0.95 ± 0.02, n = 101) stimulation were significantly greater than those for unstimulated control muscles (44 ± 1.1 ms and 0.67 ± 0.03, n = 175; ANOVA, P < 0.01) but not significantly different from each other (ANOVA, P > 0.05). G and H: comparisons of cumulative frequency histograms show that frequency distributions for the cuff- and IM-stimulated muscles were not statistically different (Kolmogorov-Smirnov test, P > 0.05). However, there were significantly more type S MUs in the cuff-stimulated (85%) than IM-stimulated (60%) muscles.
View this table:
TABLE 2.
Proportions of muscle fiber and MU types
View larger version (32K):
[in a new window]
FIG. 10.
Comparison of frequency histograms of MU FIs in chronically stimulated MG muscles in intact (A and B) and hemisected and unilaterally deafferented cats (C-H) for different patterns of stimulation: 20 Hz in a 50% duty cycle for 24 h (A-D), 20 Hz in a 6.3% duty cycle for 3 h (E and F; Table 1, cat 20), and 100 Hz in a 5% duty cycle for 24 h (G and H; Table 1, cat 19). All muscles were stimulated for >100 days (see Table 1). Mean values for the stimulated muscles (solid arrows) are compared with means for normal control muscles (dotted arrows). MU types are distinguished by different shading. Fast-to-slow conversion only occurred with 50% duty cycles. Duty cycles of 5-6.3% and 50% were both effective in increasing muscle endurance.
). The progressive conversion of muscle fiber histochemistry is consistent with biochemical and physiological studies at the single-fiber level showing that muscle fibers are converted to the S fiber type (Pette and Vrbova 1992). With time, all muscle fibers acquire the slow phenotype, as judged histochemically (Table 2). Yet, the very strong mATPase staining normally seen in type I fibers after acid preincubation is not seen in the chronically stimulated muscles.
View larger version (105K):
[in a new window]
FIG. 8.
Photomicrographs of serial cross-sections of MG muscles from a normal cat and MG muscles chronically stimulated at 20 Hz in a 50% duty cycle for 56 and 76 days (Table 1, cats 5 and 11). Three fiber types are seen in control muscles: type I (slow oxidative; 
), IIa (fast oxidative glycolytic; 
) and IIb (fast glycolytic; 
) according to the criteria described in METHODS. After 56 days of stimulation, most of the fibers are type I, with a few type IIa fibers visible. By 76 days of stimulation, all muscle fibers were type 1. Note that, during fast-to-slow conversion of muscle fibers after 56 days of stimulation, there are fibers that demonstrate an abnormal acid- and alkali-labile mATPase and have uniform staining with -nicotin-amide adenine dinucleotide (NADH) and -glycerophosphate dehydrogenase.
). Note that the ranges of the three MU parameters in the fully converted experimental muscles were the same as those for the normal type S MU population (Fig. 9). If MU activity were the primary factor determining muscle properties, it might be expected that all MUs that experience the same synchronous activation should not show the variation typically seen within any one MU type. This was not the case, because the range in the stimulated muscles was the same as and not smaller than the normal type S MU range for FI, force, TTP, and half-relaxation values in both the collated data for several animals and the data from individual animals.
View larger version (27K):
[in a new window]
FIG. 9.
MU cumulative sum distributions of (A) FIs, (B) tetanic forces, (C) twitch TTP, and (D) twitch half-relaxation times in chronically stimulated and normal control muscles compared with the distributions for the type S MU population in normal control muscles. Data for the stimulated muscles were collated from 6 muscles in which there was complete type II to I conversion and from 10 nonstimulated normal MG muscles (see Table 2 and text). Mean ± SE (n = 53-208) of the values for all control MUs, control type S MUs, and all stimulated MUs, respectively: 0.67 ± 0.03, 0.99 ± 0.03, and 0.92 ± 0.01 (A); 304.4 ± 18.1 mN, 92.3 ± 10.2 mN, and 87.4 ± 5.3 mN (B); 43.9 ± 1.1 ms, 57.5 ± 2.3 ms, and78.7 ± 1.4 ms (C); and 37.6 ± 1.4 ms, 47.8 ± 3.2 ms, and 76.6 ± 2.3 ms (D). Values for the stimulated MUs were significantly different from the control values and the same as the control type S MUs for FI and tetanic force. The stimulated muscles were, however, significantly slower than the control type S MU population: TTP and half-relaxation values of the stimulated MUs were significantly longer than those of the control type S MUs (ANOVA: P < 0.001).
View larger version (27K):
[in a new window]
FIG. 11.
Muscle endurance index (A), TTP (B), and isometric tetanic force (C) recorded at regular intervals in 4 deafferented and hemisected cats in which muscles were chronically stimulated with 1 of 4 different stimulation patterns: 20 Hz in a 50% duty cycle over a 24-h period ( 
), 20 Hz in a 5% duty cycle over a 24-h period (), 20 Hz in a 6.3% duty cycle over a 3-h period per day (
), and 100 Hz in a 5% duty cycle over a 24-h period (
). A: endurance index increased exponentially (time constant = 32 days) for the 20 Hz/50% and 100 Hz/5% patterns of stimulation. There was no significant change in endurance for the 20 Hz/5% and 20 Hz/6.3% patterns (slope of the regression line, 0.001, was not significantly different from 0). B: TTP (ms) increased for the 20 Hz/50% pattern of stimulation [pearson product-moment correlation (RO) = 0.96 for a 4th-order regression], in contrast to the other stimulation patterns. C: muscle tetanic force, normalized to day 0 of stimulation, declined exponentially with a time constant of 18 days for all stimulation patterns except 20 Hz/6.3% (3 h), where tetanic force remained constant (slope of the linear regression line, 0.002, was not significantly different from 0). MU data were obtained after the completion of the chronic recordings in 3 of the 4 cats (see Fig. 10).
,b). In agreement with the findings of Kernell et al. (1987b) that the slowing of chronically stimulated muscle contractions depended on the proportion of each day that the muscles are active and not the pulse rate of stimulation, there was a small and equal degree of slowing of the MUs and the muscles when stimulated for 5-6.3% of each day at 20 and 100 Hz [Figs. 10, E (20 Hz) and F (100 Hz), and 11B]. On the other hand, our data for the FIs suggested that the increase in MU FIs and muscle endurance depended on both the daily amount of stimulation and the pulse rate (Figs. 10, F and H, and 11A). These data can be reconciled with previous findings that the daily time of stimulation and not pulse rate affected muscle fatigue resistance (Kernell et al. 1987a) because the total number of pulses was 5 times higher at 100 Hz than at 20 Hz in our experiments, in contrast to experiments of Kernell et al. (1987a), in which the number of pulses was the same for the two frequencies. Our data showing that the slowing of twitch TTP depends on total daily activity (more slowing with 50% than with 5-6.3%) and not pulse rate are in agreement with findings of Eerbeek et al. (1984) in experiments in which the number of pulses was not equalized for the different frequencies.
) and high fatigability. In addition, the slowing of the relaxation rate is associated with a contracture that may contribute to the underestimation of the fatigue endurance of the whole muscle (see Fig. 4, B and E). The 5% duty cycle of the daily 24-h stimulation period was associated with a decline in muscle force for both 20- and 100-Hz stimulation frequencies (Fig. 11C).
View larger version (26K):
[in a new window]
FIG. 12.
Motor axon conduction velocity and twitch TTP plotted as a function of tetanic force in MUs from individual MG muscles stimulated daily with 20 Hz in a 50% duty cycle for (A) 42 days, (B) 119 days, and (C) 240 days ( ). These are compared with MUs from a normal MG muscle (). Regression lines are drawn when the slopes (means ± SE) of the lines are significantly different from 0 at the 5% level of confidence. Slopes of the regression lines for conduction velocity and tetanic force are 0.06 ± 0.01 (normal MG; - - -), 0.07 ± 0.01 (42 days), 0.06 ± 0.02 (119 days), and 0.06 ± 0.03 (240 days). The corresponding RO values are 0.67, 0.59, 0.47, and 0.34. Negative slopes of the regression lines for TTP and tetanic force are 0.13 ± 0.03 (normal MG; - - -), 0.20 ± 0.04 (42 days), 0.07 ± 0.02 (119 days), and 0.01 ± 0.04 (240 days), with RO values of 0.65, 0.59, 0.55, and 0.05.
).
View larger version (12K):
[in a new window]
FIG. 13.
Mean ± SE of the motor axon conduction velocity as a function of the days of stimulation at 20 Hz in a 50% duty cycle. Muscles stimulated via cuff and IM electrodes are distinguished by different symbols.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
; Kernell et al. 1987a,b). Although the previously reported change was interpreted to reflect a corresponding conversion of MUs to type S, MU conversion had not been demonstrated previously. The results are consistent with the hypotheses outlined in the INTRODUCTION that 1) daily amount of neuromuscular activity determines MU properties and 2) the presumed order of recruitment of MUs after muscle reinnervation accounts for rematching of nerve and muscle properties. However, our results also show that synchronous activation of all MUs with the same daily electrical stimulation in normal and paralyzed muscles does not make all the MUs the same with respect to their fatigability or contractile speed, as might be expected if MU properties were precisely graded with neuromuscular activity. Therefore small differences in neuromuscular activity are unlikely to explain the continuum of MU properties that is normally characteristic of most skeletal muscles. It is more likely that neuromuscular activity is a contributing but not the sole factor that determines the range in the properties of MUs and their constituent muscle fibers.
; Kernell et al. 1987a,b). The reduced speed of contraction was accompanied by a decline in twitch and tetanic forces and an increase in muscle endurance in both muscles. In the present study, implanted electrodes on the nerve to the MG muscle and regular chronic recording of muscle torque before and during the course of stimulation allowed us to make a detailed analysis of the time course and extent of the changes that has not previously been possible. Under these experimental conditions, each animal served as its own control to permit muscle conversion to be documented on an animal-to-animal basis. The most rapid and least variable response to stimulation was a decline in the force output. Fatigue resistance increased more slowly, whereas the increase in contraction time varied most widely between animals and occurred with a longer time course than either force or endurance (Figs. 5 and 6).
). Because innervation ratio is unlikely to be changed by chronic stimulation, and because the contribution of specific force to the range of MU force is normally small (Totosy de Zepetnek et al. 1992a), it is likely that the decline in muscle and MU force is due to the reduced cross-sectional area of the stimulated muscle fibers (Donselaar et al. 1987; Pette et al. 1976; Rafuse et al. 1997). Because force varies as the second power of diameter, a twofold decline in mean fiber diameter can account for the fourfold reduction in mean MU tetanic force. Furthermore, glycogen depletion studies of single MUs after chronic stimulation showed that the range of fiber cross-sectional area within MUs was the same as the range for all MUs (Gordon et al. 1993; Rafuse et al. 1997). Therefore the remaining range in MU force in the stimulated muscles corresponds to the range in innervation ratio of the normal MG muscle.
; Kwong and Vrbova 1981; Simoneau et al. 1993), when there are detectable increases in the ratio of oxidative to glycolytic enzyme activities and capillary density (Brown et al. 1976; Henriksson et al. 1986; Hudlicka et al. 1977; Pette et al. 1973, 1976; Reichmann et al. 1985; Simoneau et al. 1993). The rapid decline in fiber diameter, which accompanies the decline in muscle and MU forces, may contribute to the increase in oxidative potential of the muscle fibers by increasing the efficiency of oxygen delivery to the muscle fibers (see also Pette and Vrbova 1992).
; Pette et al. 1973; Salmons et al. 1978). There was a corresponding conversion of FF MUs to FR and type S MUs (Fig. 7, Tables 1 and 2). The increase in oxidative potential was accompanied by loss of the high glycolytic enzyme reactivity in the former type II fibers (Fig. 8). Attainment of maximal whole muscle and MU endurance coincided with complete fiber type II to type I conversion in muscles stimulated for >80 days (Table 2).
Fint FR S with increased fatigue resistance and reduced force output of the constituent MUs. During the early stages of conversion, when there were relatively large numbers of Fint units, the whole muscle fatigue reflects their disproportionate contribution to the fall in muscle force and slowing of muscle contractions during the fatigue test. With time, progressive decline in the force and increase in endurance resulted in generation of less absolute force during the fatigue test, but it is maintained with minimal decrement.
). A similar argument can be made to account for a remaining range in fatigability of stimulated MUs, i.e., that high neuromuscular activity upregulates metabolic enzymes from different basal levels in type I, IIa, and IIb muscle fibers. This line of reasoning is more consistent with a role of the nerve and neuromuscular activity in modulating rather than determining muscle endurance. Similar arguments explain the higher-than-expected variance in metabolic enzyme activities in muscle fibers belonging to a single MU in normal (Martin et al. 1988) and reinnervated muscles (Sesodia et al. 1993).
). Many of these sites are subject to change after chronic stimulation, for example Ca2+ release and uptake (Pette and Vrbova 1992). However, incomplete and/or different magnitudes of change at the different sites might conceivably explain the remaining heterogeneity in the fatigability of the MUs. However, strictly speaking, if muscle endurance were totally controlled by neuromuscular activity and adult muscles displayed complete plasticity, the same neuromuscular activity experienced by all MUs would result in uniform changes in the many sites of fatigue. That these uniform changes do not come about argues for a more limited plasticity of adult muscles in which neuromuscular activity modulates MU properties within an adaptive range that is either preset during myogenesis or that arises during the maturation of the muscle fibers (Ausoni et al. 1990; Condon et al. 1990a,b). This conclusion is supported by experiments in which neuromuscular activity has been experimentally reduced in several "disuse" models. These experiments demonstrate a much wider range of enzyme activities in inactive muscles than expected if neuromuscular activity were the only determining factor (reviewed by Gordon 1995; Gordon and Pattullo 1993). Even when lumbar motoneurons are completely silenced by spinal isolation, there remains a wide range of oxidative enzyme activities in soleus, gastrocnemius, and tibialis anterior muscles (Graham et al. 1992; Jiang et al. 1991; Pierotti et al. 1991). The soleus muscle in particular maintains a high oxidative potential under all conditions in which neuromuscular activity is reduced or eliminated (Gordon and Pattullo 1993).
). Some variability during the first 80 days of stimulation was accounted for by differences in muscles that were stimulated via wire electrodes around the MG nerve where it enters the muscle as compared with muscles stimulated via cuff electrodes around the MG nerve. However, these small differences could not account for the considerable interanimal variability. It is not clear whether differences in the time course of fast-to-slow replacement of multiple isoforms of regulatory, contractile, and sarcoplasmic proteins (reviewed by Pette and Vrbova 1992) contributes to this variability. We observed that muscles that were stimulated for 60 days contained muscle fibers that were classified as type IIa (fast oxidative glycolytic) on the basis of their acid-labile mATPase and high oxidative and glycolytic enzyme activities. Thereafter the muscle fibers all contained acid stable mATPase. During the transition, the mATPase was atypical in the sense that fibers that showed acid lability were not always stable at alkaline pH as they are normally (Fig. 8). This finding is consistent with the observations that stimulated muscles coexpress fast and slow isoforms of the heavy myosin chains that are responsible for the pH sensitivity of the mATPase histochemistry (Termin et al. 1989). This coexpression of fast and slow isoforms for several contractile and regulatory proteins is the normal pattern during conversion (Pette and Vrbova 1992). As was the case for MU muscle fatigability, the low-frequency, high-duration chronic stimulation did not restrict TTP values to a very narrow range, as would be expected if neuromuscular activity controlled the normal continuum of contractile speed of the MUs. Interestingly, long-term stimulation slowed some MU TTPs beyond the normal type S MUs in the muscle. The long-term stimulated MUs were soleus-like in attaining the very slow contractile speed of soleus muscle, consistent with previous accounts of a "superslow" phenotype of chronically stimulated fast muscles (Pette and Vrbova 1992). Although upregulation of slow isoforms and downregulation of fast isoforms can account for the transition to type S phenotype, the mechanism by which the type S MUs in MG are transformed to soleus type S MUs is still uncertain, particularly because the former are known to express the same slow isoforms of regulatory and contractile proteins as the latter (Pette and Staron 1990). The nature of the factors that set the range of contractile speeds within the MU type is not known, but is likely to include inherent differences between fibers in slow isoforms of contractile, regulatory proteins and proteins concerned with Ca2+ release, uptake, and storage as well as differences in the anatomic substrates for slow contraction, including the relative amount of t tubular membranes and calcium binding proteins (Pette and Vrbova 1992). There is evidence to support the idea that MU subtypes are established by clonal selection during development or in the course of maturation (reviewed by Gordon and Pattullo 1993). The considerable adaptation of the chronically stimulated adult muscle to increased neuromuscular activity indicates, nevertheless, that the multiple regulation of contractile speed of muscle fibers has considerable adaptive capacity in the adult.
) following the observations that chronic stimulation leads to changes in conduction velocity toward type S motoneurons (Fig. 13) in conjunction with fast-to-slow conversion of muscle units and type II to type I conversion of muscle fibers.
| |
ACKNOWLEDGEMENTS |
|---|
This research was supported by a grant from the Medical Research Council of Canada to T. Gordon and National Institute of Neurological Disorders and Stroke Grants RO1 NS-15913 (Javits Neuroscience Award) and PO1 NS-27511 to J. B. Munson. T. Gordon is an Alberta Heritage Foundation for Medical Research (AHFMR) scientist and V. Rafuse was an AHFMR predoctoral fellow.
Permanent addresses: V. F. Rafuse, Dept. of Neuroscience, Case Western Reserve University, 10900 Euclid Ave., Cleveland, OH 44106-4975; J. B. Munson, Dept. of Neuroscience, University of Florida College of Medicine, Gainesville, FL 32610-0244.
| |
FOOTNOTES |
|---|
Address for reprint requests: T. Gordon, Division of Neuroscience, 525 Heritage Medical Research Centre, University of Alberta, Edmonton, Alberta T6G 2S2, Canada.
Received 21 May 1996; accepted in final form 6 January 1997.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
a single fiber study.
Eur. J. Biochem.
186: 749-754, 1989.[Medline]This article has been cited by other articles:
|
|
 |
|
  C. K. Thomas, C. K. Hager-Ross, and C. S. Klein Effects of baclofen on motor units paralysed by chronic cervical spinal cord injury Brain, November 10, 2009; (2009) awp285v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Kudryashova, J. Wu, L. A. Havton, and M. J. Spencer Deficiency of the E3 ubiquitin ligase TRIM32 in mice leads to a myopathy with a neurogenic component Hum. Mol. Genet., April 1, 2009; 18(7): 1353 - 1367. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. C. Yohn, G. B. Miles, V. F. Rafuse, and R. M. Brownstone Transplanted Mouse Embryonic Stem-Cell-Derived Motoneurons Form Functional Motor Units and Reduce Muscle Atrophy J. Neurosci., November 19, 2008; 28(47): 12409 - 12418. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Hegedus, C. T. Putman, N. Tyreman, and T. Gordon Preferential motor unit loss in the SOD1G93A transgenic mouse model of amyotrophic lateral sclerosis J. Physiol., July 15, 2008; 586(14): 3337 - 3351. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Hager-Ross, C. S. Klein, and C. K. Thomas Twitch and Tetanic Properties of Human Thenar Motor Units Paralyzed by Chronic Spinal Cord Injury J Neurophysiol, July 1, 2006; 96(1): 165 - 174. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. S. Klein, C. K. Hager-Ross, and C. K. Thomas Fatigue properties of human thenar motor units paralysed by chronic spinal cord injury J. Physiol., May 15, 2006; 573(1): 161 - 171. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gallo, T. Gordon, N. Tyreman, Y. Shu, and C. T. Putman Reliability of isolated isometric function measures in rat muscles composed of different fibre types Exp Physiol, September 1, 2004; 89(5): 583 - 592. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Gonzalez-Forero, R. R. de la Cruz, J. M. Delgado-Garcia, F. J. Alvarez, and A. M. Pastor Functional Alterations of Cat Abducens Neurons After Peripheral Tetanus Neurotoxin Injection J Neurophysiol, April 1, 2003; 89(4): 1878 - 1890. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Hudlicka, L. Graciotti, G. Fulgenzi, M. D Brown, S Egginton, M. Milkiewicz, and A.-L. Granata The effect of chronic skeletal muscle stimulation on capillary growth in the rat: are sensory nerve fibres involved? J. Physiol., February 1, 2003; 546(3): 813 - 822. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Beaumont and P. Gardiner Effects of daily spontaneous running on the electrophysiological properties of hindlimb motoneurones in rats J. Physiol., April 1, 2002; 540(1): 129 - 138. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Gordon and V. F Rafuse Size of myelinated nerve fibres is not increased by expansion of the peripheral field in cats J. Physiol., May 1, 2001; 532(3): 835 - 849. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. D. Milner, V. F. Rafuse, and L. T. Landmesser Selective Fasciculation and Divergent Pathfinding Decisions of Embryonic Chick Motor Axons Projecting to Fast and Slow Muscle Regions J. Neurosci., May 1, 1998; 18(9): 3297 - 3313. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Beaumont and P. Gardiner Effects of daily spontaneous running on the electrophysiological properties of hindlimb motoneurones in rats J. Physiol., April 1, 2002; 540(1): 129 - 138. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
