Male rats were divided into control and weight-trained (WT) groups. WT rats performed squat-type exercises twice daily, 5 days/wk, for 14 wk. They averaged 36 lifts/day, with an average weight of 555 g. Muscle-to-body weight ratio (mg/g) of the soleus (Sol) was not different from control, but it increased 11 and 6% in the gastrocnemius (Gast) and plantaris, respectively (P < 0.05). The normalized twitch tension of the in situ Sol was elevated by 21%, whereas single-skinned type I fibers from the Sol showed an increased rate constant of tension redevelopment (Ktr) but no other contractile adaptations to WT. In contrast, the Gast type I fibers showed an increase (P < 0.05) in maximal velocity of shortening (25%), peak power (15%), Ktr (18%), and normalized tension (7%). The Ktr and normalized tension of the Gast type IIa fibers increased by 24% (P < 0.05) and 12% (P < 0.05), respectively, whereas velocity and power showed a tendency to increase. Fiber size, determined by myosin ATPase histochemistry, was not different for any fiber type from the Gast or Sol. These results indicate that isotonic resistance exercise of the calf targets the Gast (type I and type IIa fibers) and has little effect on the Sol.
- weight training
- single-fiber physiology
despite the widespread use of resistive exercise and its apparent effectiveness in increasing skeletal muscle size and strength and reducing muscle atrophy associated with aging and bed rest, there is a paucity of data concerning the contractile adaptations at either the whole muscle or cell level (12). Several animal models have been developed to study the effects of resistive exercise on skeletal muscle to include “squat”-type exercises (13, 15, 21, 34), weighted ladder climbing (8, 38), isometric training (10, 22), plyometric training (1), and electrically evoked contractions (3, 4, 42). A general finding is that resistance exercise increases the muscle-to-body weight ratio of the soleus (Sol) by 31–45% (21, 22), the plantaris (Plant) by 26–34% (21, 22, 34), and the gastrocnemius (Gast) by 7–18% (21, 22, 32, 34, 42). Few studies have investigated the contractile changes that occur as a result of resistance exercise. In rat studies, primarily slow-twitch muscles, such as the adductor longus and Sol, have shown no change in peak isometric tension (1, 8, 10, 32). Fast-twitch muscles may be more responsive to weight training (WT) as greater gains in peak forces have been reported for the rectus femoris and medial Gast (10).
Human resistance exercise training programs routinely produce increases in muscle cross-sectional area (CSA), strength, and power, with gains occurring in both the slow type I and fast type II fiber types (7, 12, 35, 36, 39). The increase in force and power appears to be primarily caused by the exercise-induced muscle hypertrophy (12). The effects of WT on maximal shortening velocity (Vmax) are not well understood (12). Both endurance and resistance exercise-training programs reduce the percentage of type IIb and increase the number of IIx and IIa fibers, a change that should reduce the velocity of fast skeletal muscles (12). Controversy exists as to whether or not the velocity of individual fiber types is altered by resistive exercise programs. WT in young men had no effect on the maximal unloaded shortening velocity (V0) of the slow type I or fast type IIa fibers (39). In older men but not women, Trappe et al. (35, 36) observed a high-resistance exercise program to increased V0 in both the type I and IIa fiber types.
Muscle atrophy and deleterious changes in cell function are known to be a serious problem during normal aging and in the microgravity environment of space. In both cases, regular exercise training has been utilized as a countermeasure to slow and reduce the deleterious alterations in limb skeletal muscle. It is our hypothesis that the optimal exercise countermeasure will require the combination of both isometric and isotonic resistance exercises, with the former best protecting tonic slow-twitch muscles and the latter phasic fast-twitch muscles. This hypothesis is supported by the data of Hurst and Fitts (20), who observed isometric exercise to better protect the rat slow Sol than the fast calf muscles from hindlimb unloading-induced atrophy. To test this hypothesis, it is necessary to first develop isometric and isotonic exercise protocols that can be quantified as to the work performed and tested independently and then in combination. We have recently developed an isometric testing device and studied the effects of isometric exercise-training on rat calf muscles (20). The purpose of this study was to use a modification of a “squat”-type exercise first reported by Klitgaard (21) to determine the effects of isotonic WT exercise on the following: 1) the mass of the plantar and dorsiflexors of the foot; and 2) the contractile properties of the intact Sol as well as individual slow- and fast-twitch fibers isolated from both the Sol and the Gast. This is the first study examining both whole muscle and cellular contractile properties of rat muscle in response to WT resistance exercise.
Animal characteristics, exercise program, and diet.
All methods were approved by the Marquette University Institutional Animal Care and Use Committee. Due to the time involved with training, this study was carried out in five phases with four rats comprising each phase. Twenty male Sprague-Dawley rats were obtained from the same vendor (Sasco, Madison, WI) and weighed ∼250 g on arrival. Rats were handled for 3 days to minimize contact stress. During this period, all rats were individually exposed to the cage in which WT took place. Those rats that readily entered the exercise tube and took food from the holder (see below) were assigned to the WT group that performed resistance exercises, twice daily, 5 days/wk, for 14 wk (n = 11). The remaining rats were assigned to a sedentary control (C) group that was restricted to cage activity (n = 9). Rats were housed individually in standard rat cages and maintained on a 12:12-h light-dark photoperiod. They had free access to water and were pair fed, as described below.
The model of isotonic WT used in this study was a modification of that originally reported by Klitgaard (21). The training apparatus (Fig. 1) consisted of a Lucite cylinder, 30 cm high by 10 cm diameter, mounted vertically at one end of a lucite platform. A door 10 cm high was cut into the front of the cylinder to allow rats to enter. A doughnut-shaped disk (or “collar”) was suspended from the top of the cylinder and could be easily moved vertically within the cylinder. The diameter of the hole in the collar through which the rat passed its head could be shimmed to a diameter of 3, 3.5, or 4 cm. Collars were chosen to ensure that the weight rested on the shoulder girdle of the rat and that only the head of the rat could pass through. A sliding indicator and ruler on the side of the tube indicated the vertical displacement of the collar during lifts. Weights could be added to horizontal bars that screwed into the collar to provide resistance (Fig. 1). A small food pellet (120–150 mg) was held in place near the top of the cylinder by means of rubber tubing attached to a 20-cm rod. The rod was passed through a rubber stopper that fit into a hole in the top of the cylinder. The position of the food pellet and collar could be adjusted to accommodate different sized rats and to maximize the range of motion. A successful lift consisted of the rat entering the tube, assuming a crouched posture on its hindlimbs, placing its head through the collar, and lifting the collar and attached weights by plantar flexion to receive a food pellet. The rat would usually withdraw from the tube to eat the pellet and then return to perform the next lift. Rats were trained to perform the lifting motion in the following four stages. In stage 1, small pellets of Purina Rodent Chow were placed on the floor within the walls of the cylinder and in the food holder ∼15 cm from the floor of the cage. During this stage, rats were trained to associate food with the cylinder and were encouraged to enter the cylinder and take food from the holder. This stage lasted ∼1–2 days. In stage 2, the food pellet was gradually moved upward so that rats had to place their entire body within the tube. The collar was progressively lowered so that the rats had to put their head through the collar, plantar flex against the load of the unweighted collar (200 g), and move it upward to get the food pellet. This stage lasted ∼3 days. In stage 3, the collar was lowered further so that resistance would be met soon after the rat began to raise its body from a crouched position. The food pellet was raised to a height that required full range of motion at the ankle, as determined by visual inspection, to receive a food pellet. Weight was added to the collar as rapidly as would be tolerated. Pilot studies indicated that rats could lift ∼200% of body weight after the first 2 wk of lifting. In stage 4, after the initial 2 wk of training, the weight lifted was kept near 200% of body weight for the remaining 12 wk of the study. Occasionally, rats were unwilling to lift this load, and the weight had to be reduced. The weight was returned to 200% of body weight as rapidly as possible.
WT rats were trained twice daily, for 30 min/session, 5 days/wk, for 14 wk. During each training session, rats performed as many repetitions as possible. The first training session occurred in the morning, approximately 1 h into the light cycle. Six hours later the rats were trained again.
To initiate training, all rats were fasted for 18 h. WT rats received small food pellets when they completed a lift successfully. After the morning training session, the amount of food eaten by the WT rat was measured, and its corresponding C rat was fed an equal amount. This was repeated for the afternoon training session approximately 6 h later. Typically, the amount of food eaten during lifting was between 3 and 8 g/day. In addition to the food received while lifting, all rats were fed an additional bolus of 11 g of food after the second training session. After the 5th day of training, rats received a bolus of 50–55 g, which lasted until the first training session of the next week. In pilot studies, this amount of food left the rats motivated to perform the exercise but was sufficient to allow an increase in body weight over the course of the study. The average number of calories ingested exceeded the minimal daily requirements for the maintenance of body size as determined by the National Research Council (29). This value was determined according to the following equation: kcal/day = 110 kcal/kg body mass0.75. Purina Rodent Chow contains 3.04 kcal/g of metabolizable energy and 4.00 kcal/g of gross energy. Using these values, a 300-g rat would require 14.7 g of food per day.
Calculation of work.
Work performed during lifting was calculated as the product of mass lifted, vertical displacement, and number of repetitions. A sliding indicator attached to the side of the lifting apparatus measured vertical displacement. To calculate more precisely the amount of work performed, a force plate on which the rat would stand was mounted below the lifting apparatus, and a photodiode array (Positron Development) was attached to the collar to measure vertical displacement (Fig. 1). Force and position traces and the corresponding electromyogram (EMG) traces for the Sol and lateral Gast during a representative lift are presented in Fig. 2. The signals from the force plate and photodiode were digitized and later analyzed by using custom software. The software calculated peak force, peak displacement, and the duration of the lift. Data were collected by using this system during one session, once a week for two of the WT rats.
Determination of whole muscle contractile properties.
To avoid experimenter bias, all contractile tests were performed blind. The identities of the rats were kept by a third party and not revealed until the completion of the study. Due to the force limitations of the servomotor force transducer employed (300 g maximal), we were only able to test the contractile properties of the Sol and not the Plant or Gast. In situ contractile studies were performed as previously described (24). On completing the 14-wk training period, rats were anesthetized to a surgical plane (pentobarbital sodium, 50 mg/kg ip). An incision was made along the centerline of the dorsal side of the left leg from the ankle to the hip. The hamstrings and medial and lateral heads of the Gast and the Plant were dissected to expose the underlying Sol, with care taken not to disrupt the blood or nerve supply. The surrounding muscles were denervated, and a suction electrode was placed on the sciatic nerve to apply stimulation. The distal tendon of the Sol was cut and attached, by using a loop of 2–0 suture, to a servo-controlled transducer (model 305, Cambridge Technologies, Cambridge, MA) capable of measuring force and displacement. The preparation was covered with gauze and kept moist with rat Ringer solution. Temperature probes were inserted into the rectum and the lateral head of the Gast to monitor core and muscle temperatures, respectively. Core and muscle temperatures were maintained at 37 and 35°C, respectively, by carrying out the experiments in a temperature-controlled chamber.
Isometric contractile properties.
The muscle was stimulated by using supramaximal, constant-current, square-wave pulses lasting 0.5 ms (Positron Development). The analog signals for force and length were displayed on an oscilloscope, amplified, converted to a digital signal, and sent to a personal computer for data analysis. The optimal length (LO) for force development was determined, and then twitch contractions were elicited. From each contraction, peak isometric twitch tension (Pt), time to peak tension (TPT), one-half relaxation time, and the peak rate of force development (+dP/dt) and decline (−dP/dt) were measured. TPT and one-half relaxation time are defined as the time elapsed from the onset of force production to Pt and the time required for force to decline from Pt to one-half of Pt, respectively. To measure peak tetanic force (PO), the Sol was stimulated with supramaximal, square-wave, biphasic pulses at 100 Hz for 1,000 ms. From each contraction, PO, +dP/dt, and −dP/dt were measured. Each contraction was separated by a minimum of 1 min.
Maximal V0 of the muscle was measured by the slack test method (26, 38). The muscle was stimulated to contract tetanically for 750 ms at which point the computer triggered shortening of the muscle by a predetermined distance, followed by an additional 200 ms of stimulation. The tension in the shortened muscle dropped to zero until cross-bridge cycling could take up the induced slack and begin to restore force. The period of unloaded shortening, which was defined as the time from when the length change (ΔL) was completed until the force was discernibly above baseline, was measured for each slack length. LO was reset, and the process was repeated to give a total of five different ΔL values. The duration of unloaded shortening was plotted vs. the slack step distance. The slope of this relationship represents V0. To compensate for muscles of different lengths, absolute V0 (mm/s) was divided by muscle length (ML), and V0 was expressed as ML per second.
To construct a force-velocity curve, the muscle was maximally activated, and, after the muscle had generated peak tension, the computer triggered a series of isotonic shortenings (23, 40). The position motor was integrated into a feedback servomechanism, whereby force was kept constant at a submaximal load by shortening the muscle at a constant rate necessary to maintain the selected force. During each contraction, the muscle was stepped through three submaximal loads. The velocity of shortening and relative force were measured during the final one-half of each load step. This process was repeated four to five times for each muscle at different loads to generate 12–15 data points per curve. The Vmax was determined for each muscle by fitting the data to the Hill equation: (P + a)(V + b) = (PO + a)b, where P is the isometric force, V is the velocity, a is a constant in units of force, and b is a constant with units of velocity (19). The data were fit by using an iterative nonlinear curve-fitting procedure (Marquardt-Levenberg algorithm), as previously described (40). The maximal velocity of shortening was normalized to ML. The parameters describing the force-velocity relationship, Vmax (the velocity axis intercept), a/PO (a unitless parameter describing the curvature of the relationship), and PO were used to calculate peak power (40).
Skinned fiber contractile properties.
To determine the fiber-type-specific effects of isotonic resistance exercise on contractile function, single-fiber studies were conducted. These studies were particularly important in determining the exercise training-induced changes in the heterogeneous Gast, where intact muscle studies of function are limited to isometric measurements. The direct Ca2+ activation of the individual skinned fiber allowed the effects of exercise-training on PO to be assessed independently of the excitation-contraction coupling process. Additionally, the measurements of the rate constant of tension redevelopment (Ktr) and fiber elastic modulus (E0) provide important information about the cross-bridge binding rate and force per cross bridge that cannot be obtained from whole muscle studies.
Fiber PO, Ktr, maximal V0, Vmax, peak power, and E0 were determined for single-fiber segments isolated from the Sol and Gast. All contractile tests were performed on each fiber segment. If fiber deterioration occurred (PO decline of >10%), the data for that fiber were discarded. The measurement of fiber V0 (slack test), Vmax (determined from force-velocity relationship), and peak power were exactly as described for the intact Sol. The relaxing (pCa 9.0) and activating (pCa 4.5) solutions, fiber preparation, measurement of PO, and stiffness were performed as previously described (23, 25, 26).
Briefly, on completing the in situ contractile tests, the contralateral Sol and Gast muscles were removed, trimmed of excess fat and connective tissue, and placed in cold relaxing solution (pCa < 9.0). The muscles were divided longitudinally, and one-half were used for single-fiber contractile studies and the other for ATPase histochemistry. The deep red portion of the lateral Gast was used as it contains a mix of fast- and slow-fiber types. The section for single-fiber analysis was dissected into small bundles (∼150 fibers) that were tied to capillary tubes and placed in skinning solution and stored at −20°C for up to 4 wk (26, 38). Individual fibers were isolated and transferred to the experimental chamber filled with relaxing solution (pCa 9.0 and 15°C). The fiber was suspended between a force transducer (Cambridge Technology, model 400A) and the arm of a position controller (Cambridge Technology, model 300B), as described previously (26, 38). The sarcomere length was set to 2.5 μm, and fiber diameter was determined by a Polaroid photograph taken while the fiber was briefly suspended in air. The fiber CSA was calculated based on the assumption that the fiber forms a circular cross section when suspended in air (38). Fibers were activated by transferring the fiber to a chamber containing a high calcium content (pCa 4.5). Peak force (PO) was measured, and V0 and Vmax were determined. Peak power for each fiber was calculated from PO, Vmax, and a/PO (40). Composite force-velocity and force-power curves were constructed by calculating individual fiber velocity and power values at 1% intervals between 0 and 100% of PO and then plotting the mean velocity and mean power obtained at each interval (40).
Fiber stiffness was measured by triggering a sinusoidal ΔL equal to 0.05% of fiber length at 1.5 kHz and measuring the resulting force change (ΔF). Peak fiber stiffness (ΔF/ΔL) was determined by computer subtraction of the stiffness recorded in relaxing solution from that measured in activating solution. The E0 of each fiber was calculated as follows: E0 = ΔF/ΔL × fiber length/fiber CSA.
Ktr was measured by using the methods of Brenner and Eisenberg (2). The fiber was activated at pCa 4.5, and, after the development of PO, the fiber was slacked 400 μm. After a brief pause, the fiber was reextended to its original length. The pause duration was 45 and 25 ms for fibers with V0 less than or equal to and greater than 2.0 FL/s, respectively. Based on fiber force, these values produced the greatest degree of cross-bridge detachment on reextension of the fiber to LO. The recovery of force after reextension was fit by using the monoexponential equation to calculate Ktr (28).
Electrophoretic determination of myosin heayy chain composition.
After the contractile test was performed, the fiber was removed from the chamber and placed in 10 μl of SDS sample buffer and stored at −80°C. Sample buffer contained 6 mg/ml EDTA, 0.06 M tris(hydroxymethyl)aminomethane, 1% SDS, 2 mg/ml bromophenol blue, 15% glycerol, and 5% β-mercaptoethanol. To determine the myosin heavy chain composition, each fiber was analyzed by SDS PAGE, as described previously (38). Fibers were identified as fast (type IIb, IIx, IIa) or slow (type I), based on their mobility compared with known standards.
Myosin ATPase histochemistry.
One-half of the unstimulated Sol was pinned to an index card at LO. The deep, red portion of the lateral head of the Gast was separated and also pinned to an index card at a slightly stretched length. The tissues were rapidly frozen in a slurry of Freon-22 cooled with liquid nitrogen and stored under liquid nitrogen until staining. Before staining, the muscle was mounted in Tissue Tek and sectioned into 10-μm sections, at −20°C, by using a Reichert-Jung cryostat. Muscle sections from each group were placed on each slide to avoid differences due to variations in the staining procedure. All slides were run in duplicate. Muscle sections were stained for acid and alkaline myofibrillar ATPase activity by using a modification of the methods of Guth and Samaha (16). Fibers types were distinguished by either a 15-min preincubation at pH 10.4 or an 8-min preincubation at pH 4.35. Fast fibers are alkaline stable, and slow fibers are acid stable.
After staining, sections were analyzed by using a light microscope interfaced to a personal computer. Fibers were classified as slow type I, fast type IIa, or fast type IIb. CSAs were measured by computerized digitizing morphometry (Bioquant II, R&M Biometrics, Nashville, TN). Before each digitization session, one fiber was digitized five times to determine the coefficient of variation. A minimum of 50 fibers of each fiber type were measured when possible (the Sol contains no type IIx or IIb fibers). To determine fiber-type distribution, fiber-type percentages were determined from four visual fields per muscle (∼200 fibers total).
Comparisons between groups were made by using a one-tailed, Student’s t-test for independent samples. All values are reported as means ± SE. The level of significance was established as either P ≤ 0.10 or P ≤ 0.05.
Eleven rats were successfully trained to perform the isotonic resistance exercise. After the first 2 wk of training, the average maximal weight lifted was 419 ± 5 g. This corresponded to 155 ± 10% of body mass. The maximum weight lifted was increased and maintained as close to 200% of body mass as possible for the remainder of the study. Due to an increase in body mass throughout the study, the average weight lifted increased weekly and was 631 ± 10 g or 192 ± 5% of body mass at the end of the study. Typically, rats began each training session with five repetitions at a weight between 150 and 175% of body weight, after which the weight was increased to the final amount. Although WT rats were fairly consistent in their performance, occasionally they were unwilling to lift, and the maximum weight lifted had to be reduced to facilitate exercise. Rats averaged 36 ± 1 repetitions/day over the course of the study. Each lift lasted an average of 1.00 ± 0.02 s, as measured from the photodiode trace (Fig. 2), and did not change over the course of the study. Vertical displacement increased from 2.93 ± 0.04 cm in the first week of lifting to 3.93 ± 0.01 cm in the last week. There was a <5% difference in vertical displacement and weight lifted between measurements made using the photodiode and force plate and measurements made by the sliding indicator and summing the added weights. The average work performed was 7.4 ± 1.3 J/day. The daily average increased from 5.9 ± 0.5 J/day during the second week of lifting to 9.0 ± 0.6 J/day in the final week. The increased amount of work is the result of an increase in the absolute amount of weight lifted and increased vertical displacement, because the number of lifts performed did not change over the course of the study.
Muscle and body weights.
Body and muscle weights are summarized in Table 1. At the onset of the study, the average body weight was not different between groups (C = 269 ± 11 g; WT = 262 ± 12 g), whereas, following the 14 wk of exercise-training, the body weights of C rats were significantly greater than those of WT (P < 0.10; Table 1). Muscle weights (mg) showed no significant differences between groups for any of the muscles studied. However, when normalized to body mass, the Gast and Plant muscles of the WT rats were 11 and 6% greater, respectively, than those of C (P < 0.05). The Sol, tibialis anterior, extensor digitorum longus, and heart showed no adaptations either in absolute weight or when normalized to body weight. In an attempt to determine whether there was a correlation between muscle weights and work performed, the muscle weights and muscle-to-body weight ratios for the Sol and Gast of the individual WT rats were correlated with the average daily work performed by each rat. Surprisingly, we observed a negative (albeit nonsignificant) correlation between the amount of work performed and the muscle wet weights for both the Sol and Gast. There was a stronger negative correlation between the amount of work performed and the muscle-to-body weight ratios, and for the Gast this correlation (r2 = 0.474) was significant (P < 0.05). For the Sol, ML was also measured, and it was unaffected by the resistance exercise (28.5 ± 1.0 vs. 27.1 ± 1.3 mm for WT and C, respectively).
Whole muscle contractile properties.
Twitch and tetanic contractile properties of the Sol are presented in Table 2. Training produced a significantly greater normalized twitch force (C: 53 ± 2 kN/m2; WT: 64 ± 4 kN/m2, P < 0.05). Representative twitches from C and WT Sol muscles are shown in Fig. 3. WT decreased TPT and increased absolute twitch force (N) and +dP/dt (P < 0.10). Following WT, the Sol PO normalized for muscle CSA (kN/m2) was higher than C (P < 0.10), whereas no differences were observed in tetanic +dP/dt or −dP/dt between groups. For the C group, the Sol Vmax determined by the slack test (V0) was 2.35 ± 0.19 ML/s, and this parameter was unaltered by WT. The force-velocity characteristics of the Sol (studied in situ at 35°C) are shown in Table 3. WT had no effect on Vmax, a/Po, %Po at peak power, or peak power (absolute or relative).
Single-fiber diameter and force.
Results from slow type I fibers from the Sol and Gast and type IIa fibers from the Gast are presented in Table 4. Due to insufficient numbers, those fibers expressing more than one MHC or other fast isoforms were excluded from analysis. There were no differences between groups in mean fiber diameter for any fiber type. Nor were there differences for absolute and normalized peak force of type I Sol fibers (Table 4). Absolute peak force from WT type I and type IIa Gast fibers was not significantly different than C. However, when normalized for fiber CSA, the tension of type I and type IIa fibers was 7 and 14% greater, respectively, than C (Table 4). Fiber stiffness and force-to-stiffness ratio were not different for any of the fiber types (Table 4).
Maximal V0 and Ktr.
WT had no significant effect on V0 for any fiber type (Table 5). In contrast, the Ktr of Sol type I and Gast type I and IIa fibers were 11, 18, and 24% greater, respectively, than C (Table 4). Figure 4 shows superimposed records of the recovery of force after the slack-reextension protocol for three different fibers. The Ktr of the WT type I fibers was clearly elevated above C but still less than that of type IIa fibers.
Table 5 summarizes the force-velocity-power characteristics. The force-power relationships for Sol type I and Gast type I and IIa fibers are shown in Fig. 5. Vmax determined by the Hill equation was not different between groups in the Sol. In the Gast type I and IIa fibers, Vmax of WT fibers was 16% (P < 0.05) and 13% (P < 0.1) greater, respectively, than C.
The a/Po, which describes the curvature of the force-velocity relationship, was not different between WT and C for the Sol type I or Gast type IIa fibers. In WT rats, the a/Po of the Gast type I fibers was significantly less, indicating that there was a greater curvature to the force-velocity profile for this fiber type after resistance exercise.
Peak power was generated at the same percentage of peak force in the Sol type I and Gast type IIa fibers. In the type I fibers of the Gast, peak power was generated at a lesser relative force after WT. Peak power was not different in the type I fibers of the Sol, either in absolute terms or when normalized for fiber CSA (Table 5). In the type I fibers of the Gast, the higher velocity of shortening and higher forces compared with C resulted in greater peak power (Table 5 and Fig. 5). Absolute peak power and normalized peak power were both 15% greater than C (Table 5). In the type IIa fibers of the Gast, there was a trend for greater absolute peak power in the WT group (17% higher than C with P = 0.12). In contrast, normalized power was 20% higher in WT compared with C, and this difference was significant at P < 0.05 (Table 5).
Results of the histochemical analysis are presented in Table 6. Resistance exercise produced no significant differences in fiber CSA for any fiber type or in the distribution of fiber types in either the Sol or Gast.
Effectiveness of the model.
A primary hypothesis of this study was that high-resistance, squat-type exercise would cause relatively greater adaptations in the fast Gast compared with the slow Sol. To test this hypothesis, it was necessary to successfully train rats to perform a squat-type resistance exercise with moderate to heavy loads. A potential criticism of this work is that the imposed load (∼200% of body weight) was too light to elicit optimal adaptations in muscle mass or function. Studies using electrically evoked contractions in anesthetized rats show that one leg can move in excess of 1,500 g (4), and it has been demonstrated that rats can volitionally produce forces upwards of 350–400% of body weight during activity (6). Roy et al. (32) were able to get rats to lift 300% of body weight. We attempted to increase the %load throughout the study, but, when the load was increased to ∼225% of body mass, the number of repetitions fell to less than five per 30-min session, and sometimes the rats were completely unwilling to lift the greater load. However, it is important to realize that, although the weight lifted remained at 200% of body weight throughout the study, the absolute load lifted increased weekly due to gains in body weight. This plus a 1-cm improvement in vertical displacement allowed for a 1.5-fold increase in the work performed from the beginning to the end of the training period. The significant inverse relationship between work performed and the Gast-to-body mass ratio suggests that the rats may have overtrained. Multiple training sessions were selected because Klitgaard (21) reported that three training sessions per day maximized the amount of work performed. Also, there are reports from human studies that indicate that greater strength gains are achieved when the same amount of work is performed in multiple sessions (17). Future studies are needed to test the hypothesis that overtraining reduced the functional adaptations.
Our hypothesis that WT would increase the mass of fast- but not slow-twitch muscle was not supported by the data, as the exercise-training had no effect on the wet weight of any of the hindlimb muscles studied (Table 1). However, the muscle-to-body weight ratios of the Gast and Plant, but not the Sol, of WT rats were greater than C (Table 1). Because fiber hypertrophy was not observed in either the Sol or the Gast (Tables 4 and 6), the increased Gast muscle-to-body weight ratio was likely primarily due to the reduced body weight in the WT group.
Muscle-specific adaptations were noted in single-fiber function, where slow and fast single fibers from the Gast showed contractile alterations, but slow fibers from the Sol were unaffected except for Ktr. Several factors may have contributed to the preferential effect of the WT on Gast compared with Sol fibers. First, the movement during lifting was very rapid and short in duration (∼1 s). The Sol, which is composed of 80–90% type I fibers, may become unloaded during the rapid contractions of the fast Gast and Plant muscles (37). Second, the Sol spans only the ankle joint, whereas the Gast spans both the ankle and the knee. This affects the extent and velocity of shortening during lifting. During a lift, there is both extension at the knee and plantar flexion at the ankle. The Sol is undergoing shortening as a result of plantar flexion and thus operating at continually shorter sarcomere lengths. The Gast shortens with plantar flexion, but also lengthens as a result of knee extension. Thus the overall change in ML is not as great for the Gast as it is for the Sol. Thus the Gast operates closer to optimal filament overlap for a greater portion of the lift. Muscle fiber architecture may also play a role in force development. The muscle fibers of the Sol are arranged parallel to the axis of force generation, and so a change in overall ML has a greater effect on fiber and hence sarcomere length than that observed for the pinnate-arranged Gast. The high-velocity movement and extensive shortening experienced by the Sol both act to limit the force generated by the Sol during lifting. During a high-velocity isotonic contraction, the Sol would be shortening at a relatively high percentage of its Vmax, and, consequently, the force developed would be low. Prilutsky et al. (30) used EMG and force records in cats to estimate that, during rapid (1.8 m/s) treadmill locomotion, the Sol would only be capable of generating 6–20% of its PO. Additionally, the Sol has been shown to be near maximally recruited during even low-intensity activities such as standing (37). So, imposing a high external load may not increase the activation of the Sol. The additional load may be imparted to the Gast and Plant muscles, which can increase EMG activity severalfold. To determine whether this was the case, stainless steel bipolar electrodes were surgically implanted to record EMG activity from the Sol and lateral head of the Gast of a rat (20). The rat was then trained to perform the same type of resistance exercise presented in this study. Figure 2 shows the displacement, force, and EMG activity measured during the lift. The results indicate that the Gast was inactive except during the lift. In contrast, the Sol was always active except for brief quiescent periods immediately before and after the lift when the rat was stepping to and from the force plate. Thus the increase in EMG activity caused by the lift was highest in the Gast. All of these factors may potentially limit the amount of force generated by the Sol during high-velocity movements.
Whole muscle contractile properties.
Despite the lack of increase in Sol wet weight, there were changes in the whole muscle contractile properties. Twitch properties were more rapid in WT rats (Table 2). TPT was less and +dP/dt was greater than C (P < 0.10). The former may reflect a change in the kinetics of calcium release and reuptake, whereas the latter could reflect an increased rate of SR calcium release or myosin binding to actin. The latter possibility is supported by the finding that single-fiber Ktr was greater in the WT group (Table 4). The duration of muscle activation, which varies with the mode of exercise, may be an important determinant of contractile adaptations. Periods of activation lasting several minutes, such as that performed during isometric training, have been found to prolong contraction time (9). However, contraction time was reduced in the Sol after plyometric training where the duration of muscle activation was very short (1). Twenty-six weeks of ladder climbing produced no differences in contraction time or relaxation time in the Sol or extensor digitorum longus (8). Resistance exercise did not result in substantially greater force production of the Sol. However, when normalized for muscle CSA, there were increases in peak twitch (P = 0.02) and tetanic (P = 0.06) tension. The latter effect was small and not observed in single-fiber analyses where Sol type I fiber PO was not different between groups. Gardiner et al. (15) found peak force in the Sol to be 25% lower than control after weight lifting. This, however, could be explained by the lower muscle to body weights of the trained rats because normalized tension was not different. In contrast, Klitgaard et al. (21, 22) found peak forces of the Sol and Plant to be higher in very old WT rats relative to age-matched C. The effects of WT on shortening velocity are controversial. This study found no change in V0 or Vmax in the intact Sol. The only other accounts of shortening velocity in rats after WT report a reduced Vmax in the adductor longus (32) and an increase in the Sol (1). Heavy resistance training has been shown to decrease the percentage of fast-type IIb and increase IIx and IIa fibers in fast-twitch muscles of both humans and rodents (3, 4, 12). However, no changes have been observed in either the percentage of type I fibers or myosin heavy chain composition (3, 4, 12). Thus one would not expect resistance training to alter the V0 of a primary slow-type I muscle such as the Sol.
Single-fiber contractile properties.
This is the first report on the effects of resistance exercise training on the contractile properties of single, skinned muscle fibers in rats. Consistent with the change observed in the muscle-to-body weight ratio, the fibers from the Gast but not the Sol showed adaptation in contractile function. The type I Gast fibers displayed the greatest adaptation with increased normalized force, Vmax, absolute and relative power, and Ktr.
Perhaps the most novel finding of this study is that rapid (<1 s), moderate-intensity resistance exercise increased Ktr in all fiber types. Ktr, first reported by Brenner and Eisenberg (2), is thought to reflect the rate of transition from the weakly bound low-force state (AM·ADP·Pi) to the strongly bound high force state (AM·ADP) of the cross bridge. Ktr has been shown to vary with myosin heavy chain isoform, free Ca2+, ionic strength, phosphate, and internal shortening of sarcomeres (2, 5, 25, 26). In the present study, the differences in Ktr were not likely due to differences in MHC content, because gel electrophoresis was used to ensure that all fibers expressed the same MHC isoform. It is possible that slow fibers from the exercise-trained muscles may have expressed an alternate form of the type I MHC with different kinetics that was undetectable using our gel protocol (11). Also undetectable amounts of fast MHC isoforms may have been present, but these would account for <2.5% of total fiber protein (26). The differences were not due to Ca2+ concentration, ionic strength, pH, or phosphate, as all of these were controlled in the solutions used for skinned fiber contractile tests. Remaining possible explanations include 1) differences in the stability of the sarcomeres during the slack-restretch protocol, 2) posttranslational alterations in myosin (or actin) that affect binding rate, 3) other unknown modulators of cross-bridge kinetics, or 4) a combination of the aforementioned. Sarcomere stability is a critical issue in the determination of Ktr. Brenner and Eisenberg (2) reported that Ktr is reduced by 50% if there is internal shortening of sarcomeres by only 2.5% at 15°C. Chase et al. (5) found that, when fiber length, as opposed to sarcomere length, was controlled, Ktr was 19% less. However, they also reported that there was a very high correlation (r2 = 0.97) between Ktr values obtained using these two different techniques. Therefore, even though the present study controlled overall fiber length, and not sarcomere length, the data should qualitatively reflect differences in Ktr. Also, our values of Ktr for rat type I Sol fibers are similar to those published by other laboratories in which sarcomere length was held constant (27, 28).
The functional consequences of the increased Ktr are uncertain. Hakkinen et al. (18) observed an increased +dP/dt in humans after resistance training, and, in this study, we found a 31% higher (P = 0.08) normalized twitch +dP/dt in the WT Sol (Table 2). Because Ktr is higher in skinned fast fibers than slow fibers, it is thought that this may partially explain the higher +dP/dt seen in intact fast-twitch muscle. While this probably holds true between muscles of different muscle fiber compositions, we found no correlation between +dP/dt normalized for peak force (during twitches or tetanic contractions) and the Ktr of fibers isolated from the contralateral muscle of the same subject. In fact, the correlation was fairly strong and negative (r2 = 0.57), suggesting that the calcium transient may be more important than Ktr in dictating whole muscle +dP/dt in muscles of similar fiber-type composition.
In the skinned fiber preparation, the increased peak force of the Gast fibers post-WT (Table 4) may have resulted from an increased number of cross bridges formed or an increase in the amount of force generated per cross bridge. It has been shown that fiber stiffness varies directly with the amount of filament overlap (14) and normalized force (2), which has lead to the conclusion that stiffness is an indication of the number of cross bridges in the strong bound state. The increase in normalized force of the type I and type IIa fibers of the Gast was mirrored by nonsignificant increases in stiffness of approximately the same magnitude. Because fiber diameter was not different between groups, this suggests that the increase in force is the result of an increased number of cross bridges per cross section of muscle fiber. This would be consistent with an increase in myofibrillar protein concentration as a result of resistance exercise, a response that has not been documented in rats, despite increases in Sol muscle wet weight (42). Also, there was no change in the PO/E0, which has traditionally been used as an indication of the force generated per cross bridge, again suggesting that the increased force was the result of an increase in the number of cross bridges per CSA. It should be noted that it is theoretically possible that other factors, such as thin filament compliance, may be responsible for the change in stiffness. The end compliance of the fiber, as determined from the y-intercept of the slack test, was similar between groups, indicating that this is unlikely.
Widrick et al. (39) and Trappe et al. (35, 36) recently studied the effects of a 12-wk WT exercise program on single-fiber function in humans. Unlike our rat data, in humans the resistance exercise increased fiber CSA and absolute force (N), whereas force per CSA (kN/m2) was unaltered. This suggests that, in these studies, rats and humans increased force output in response to resistance exercise in distinctively different ways. The humans relied on fiber hypertrophy, whereas the rats in our study depended on an increase in force per CSA (kN/m2). Although muscle hypertrophy has not been observed in rats trained with volitional exercise paradigms, it is apparent that it can occur, as Caiozzo et al. (3) observed an 11% increase in rat Gast muscle mass in response to resistance training that employed direct electrical stimulation of the muscle at high frequency. Additionally, Widrick et al. (39), studying young men, and Trappe et al. (35), studying older women, found WT in humans to have no effect on the Vmax of either slow- or fast-twitch fibers, whereas we observed an increased Vmax in the Gast type I (P = 0.002) and IIa (P = 0.06) fiber. This finding was similar to that observed in older men in whom Trappe et al. (36) found resistance exercise to increase Vmax in both the slow type I and fast type IIa fiber.
The mechanisms for the increased Vmax in the Gast fibers following WT are unknown. In an earlier study, Schluter and Fitts (33) observed a similar 23% increase in V0 in the type I fiber of the Sol after a regular program of endurance training. The authors attributed the change in V0 to an increased myosin-ATPase activity caused by a shift in the myosin light chain (MLC) composition of the fibers. Contrary to the findings presented here, endurance training caused an 11% decrease in the V0 of type IIa fibers from the Gast (33). It was thought that a lower V0 might provide for greater efficiency of fast muscle during slow speeds of locomotion because peak power is generated at velocities <50% of V0. In support of this hypothesis, Reggiani et al. (31) found that efficiency was reduced in fast muscle fibers that expressed greater amounts of MLC3f. Consequently, the muscles from endurance-trained animals may sacrifice power in favor of a more efficient performance. The increased peak power of the Gast type I fibers following WT suggests that resistive exercise training favors greater peak power over efficiency. This is consistent with the unique functional demands of a short-duration, power-type activity, where efficiency is not a limiting factor in performance. The increased velocity of the slow type I and fast type IIa fibers does not necessarily translate into a faster Gast. Both endurance and strength exercise training programs have been shown to reduce the percentage of fast type IIb fibers by transformation into the somewhat slower fast type IIx fiber (3, 4, 12). Thus whole muscle Vmax might stay the same or even decrease. What the WT adaptation does do is allow the slow type I and fast type IIa fibers to maintain a higher velocity of shortening and power when contracting against light to moderate loads (Fig. 5). Additionally, these fibers are less likely to produce significant internal drag to the faster contracting type IIx and IIb fibers.
Resistance exercise training did not result in fiber hypertrophy or any changes in fiber-type distribution in either the Sol or the Gast. Except for a small increase in whole muscle Pt and PO and an increased single-fiber Ktr, the Sol muscle was unaltered by isotonic WT. The primary adaptations to WT occurred in the Gast, especially the type I fibers, which showed significantly greater forces per CSA, higher absolute and normalized power, Vmax, and Ktr compared with the control fibers. These results indicate that isotonic resistance exercise training in rats selectively affected the Gast, improving single-fiber peak force and power without inducing fiber hypertrophy.
This research was supported by National Aeronautics and Space Administration Grant NAG 9–1156 to R. H. Fitts.
The authors extend thanks to Janell Romatowski for expert assistance with gel electrophoresis and Danny Riley and Kalpana Vijayan for assistance with the histochemical procedures.
Present address of K. M. Norenberg: Dept. of Biology, Xavier University of Louisiana, I Drexel Dr., Box 85B, New Orleans, LA 70125.
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.
- Copyright © 2004 the American Physiological Society