INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

IT IS COMMON EXPERIENCE that use of muscles in situations in which they are repetitively forced to lengthen while generating active tension, known as "eccentric" contractions, leads to pain and tenderness that peak 1-2 days after the exercise. This "delayed-onset muscle soreness" is accompanied by a series of changes to the mechanics, histology, and biochemistry of the muscle, collectively referred to as "damage" from eccentric contractions (reviewed in Refs. 1, 3, 8). Similar changes can also be shown in whole animals and in isolated muscles subjected to a series of forced active lengthenings. Such a series of contractions, whether voluntary in intact conscious animals or stimulated in isolated muscles, is referred to as "eccentric exercise."

Schwane and Armstrong (26) showed that a prolonged bout of decline running by untrained rats produced biochemical changes in muscles, particularly the vastus intermedius, the deep red layer of the knee extensors, indicative of cell damage. They concluded that these postural muscles undergo eccentric contractions while lowering the animal down the decline, acting as brakes, but undergo concentric contractions during incline running. They also showed that training, consisting of 30 min of running on a declined treadmill daily for 5 days before a prolonged test bout of decline running, greatly reduced the apparent damage from the test bout, compared with otherwise similar incline training. Neither they nor others were able to explain the mechanism by which the different training exercises produced the different levels of protection (1, 3, 8).

Lynn and Morgan (19) showed that rats similarly trained on a declined treadmill averaged 12% more sarcomeres in series than did rats that had done the same period and speed of running on an incline, as shown by measuring the fiber lengths and sarcomere lengths of fibers from fixed, acid-digested muscles. They postulated that the greater number of sarcomeres reduced the sarcomere length at a given knee angle, confining operation during decline running, assuming unchanged gait, to the ascending limb of the sarcomere length-tension curve and avoiding the plateau and descending limb with its associated instability of a near-uniform distribution of sarcomere lengths. These instabilities have been proposed as the first step in the process by which damage occurs (21). A previous study (27) has shown that the amount of damage is strongly dependent on muscle length. Increased shortening velocity after training that involved lengthening contractions (see Fig. 1 of Ref. 9) is compatible with an increased number of sarcomeres, although many other factors may be involved. Because the acid digestion removed the tendinous material, Lynn and Morgan presented no evidence concerning possible differences in the tendinous attachments of the muscle fibers after the different exercises.

Such differences in the number of sarcomeres in series would be expected to result in differences in the length-tension relationships of the muscles, which is the prediction to be tested in the present paper. Difficulties in accurately measuring muscle length and relating it to the physiological range can be overcome by leaving the muscle with its insertion and origin intact and by measuring the torque generated as a function of joint, in this case knee, angle. Because the moment arm of the knee is relatively independent of knee angle, the angle-torque curve should be very similar to a length-tension curve. This method is also potentially able to measure adaptation of the effective tendon length.

The training effect measured as extra sarcomeres is predicted to be correlated with the effects of a test bout of eccentric contractions. These effects can be assessed by many different parameters: histological, perceptual, and biochemical as well as mechanical. A simple and widely used mechanical measurement is the fall in tension measured at the length that was optimum before the contractions (see references in DISCUSSION, Measurement of damage). However, because eccentric contractions lead to a shift in the length-tension curve (4, 16, 29), the drop in tension at constant length is at least partly due to the shift in the length-tension curve and not a real decrease in tension-generating capability. In this study, the comparison of shift in optimum angle, the drop in torque at the pre-eccentric contraction testing (ECT) optimum length, and the drop in torque when the length is readjusted to the post-ECT optimum was undertaken.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Twenty mature male Long Evans rats (Rattus norvegicus), mostly littermates, were selected for uniform age (all animals were born within a 3-day period) and body weight at the time training began. Brief treadmill-running tests were used to select the 16 most willing runners. These were divided into four batches of four rats, two for incline training and two for decline training per batch. The batches were trained and tested in succeeding weeks. The age at surgery thus varied from 156 to 184 days, but incline- and decline-trained groups had equal numbers of rats of each age, with near-equal average body weights.

At surgery, the optimum knee angle for torque generation was determined for both vastus intermedius muscles from each rat (see Surgery for details). Within each batch, one incline-trained and one decline-trained animal were then used for fixation and fiber length determination (10 of 16 muscles completed the protocol; the other 6 produced estimates of sarcomere count but not of optimum angle). The other animals were used for ECT (see ECT; 14 of 16 muscles completed the protocol). The losses were due to premature death of the anesthetized animals.

Training. The method for training the rats has been described previously (19). Briefly, rats were trained over 5 days. The only difference between the training groups was the slope of the treadmill, either inclined or declined by 16° to horizontal. This ensures that differences are due to the slope of the treadmill and not to other exercise parameters. Training consisted of the rats running at a velocity slightly above the walk-trot gait transition speed, ~16 m/min, for a period of 15 min the first day, building by 5 min/day up to 35 min on the final days, with a 1.5-min rest break included every 5 min. Each animal was started at a slower speed on the first day to accustom it to the procedure. No stimuli were used to encourage running except for a jet of air applied to the tails of animals on the first days if necessary. Incline- and decline-trained animals were housed together, with water and food freely available but with no exercise facilities.

Surgery. All data collection occurred 3 days after the last training session. The animals were anesthetized with intraperitoneal urethan (~2.0 g/kg), with supplementary doses (1 g/ml) as required to maintain deep anesthesia. For each leg, the femoral nerve was freed to enable stimulation. The vastus intermedius was mechanically isolated by dividing away the overlying vasti and cutting their insertions on the patella tendon. The femoral nerve had all branches not innervating the vastus intermedius cut. Thus the vastus intermedius was the only muscle generating torque about the knee. Connective tissue was removed from both sides of the muscle, with care being taken to maintain blood supply. For testing, the animal was in a supine position on a heating blanket, with its pelvic girdle and the base of the tail clamped while the lower leg (tibia and fibula) was secured to the equipment (see Angle-torque curves). The femur was positioned so that its axis was vertical and the lower limb was set initially at a knee angle of 90°. The muscle and nerve were immersed in paraffin while retained in skin flaps.

Stimulation. Muscle activation was accomplished by stimulation of the femoral nerve proximal to the femur at 90 pulses/s, which was determined to be supramaximal stimulation during preliminary work. Bipolar platinum electrodes driven by a Devices type 2533 isolated stimulator were used with a voltage that was just supramaximal, between 1 and 1.2 V. Pulse duration was 0.2 ms, and the period of activation was 350 ms or just enough to produce a maximal response. The rest period between contractions during measurement of length tension curves was 2 min.

Angle-torque curves. All results were measured as knee angle and torque generated by the muscle while it was still attached to the bones, rather than as length and tension of an isolated muscle. This avoids difficulties that we have experienced in attaching tendons to equipment, particularly slippage and the difficulty of preserving the in vivo muscle length with sufficient accuracy to show differences in optimum. Provided that the lever arm of the muscle about the joint does not change significantly with angle, these measurements are equivalent. Our unpublished observations support this assumption.

Sarcomere counting. Both vastus intermedius muscles from four incline-trained and four decline-trained rats were used for sarcomere counting. The methods used and the tests undertaken to validate them are described by Lynn and Morgan (19). The muscles were fixed, while still attached to the femur and the lower leg, in formaldehyde-saline buffered at pH 6.9, at maximum flexion of the knee for at least 1 wk. The fixed legs were imaged by using a video camera, and the femur length of each leg and the included angle of each knee were then calculated by using the image-processing program IPlab (Signal Analytics, Vienna, VA). The femur length measurements were used to ensure that relevant body dimensions were well matched between training groups. The muscles were then digested in 70% nitric acid and dispersed in an ultrasonic bath, and the suspension was serially diluted to provide a truly random sample of 24 fibers per muscle by elimination of any choosing of fibers by the experimenter. The use of ultrasound dispersion reduced the duration of digestion needed and greatly reduced the number of broken fibers seen. Fiber lengths were measured by computer imaging and sarcomere lengths were measured by laser diffraction to produce an estimate of the average number of sarcomeres in series in the muscle fibers.

ECT. Both vastus intermedius muscles from eight animals were used for ECT. The ECT consisted of 20 tetanic contractions, spaced 1 min apart, that included the muscle's being lengthened due to a 27° decrease in included knee angle in 33 ms, starting from 90°.

Data processing. Statistical processing used Stat-View (Abacus Concepts, Berkeley, CA). All values are means ± SE. Incline-decline comparisons are unpaired two-tailed t-tests. Igor (Wavemetrics, Lake Oswego, OR) was used for data processing and fitting of Gaussian curves for optimum angle estimation.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Length-tension relationship. Figure 1 shows two length-tension plots, one for a muscle from an incline-trained rat and one for a muscle from a decline-trained rat. The fitted Gaussian curves, used to find the optimum angles, are also shown. Smaller included angles, corresponding to greater muscle lengths, are shown in Fig. 1, right. Note that the optimum occurred at longer muscle length (smaller included angle) in the muscle from the decline-trained animal.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Angle-torque curves for 1 vastus intermedius muscle from a decline-trained rat () and 1 muscle from an incline-trained rat (). Torque is normalized to maximum torque. Note that smaller included angles, corresponding to longer muscle lengths, are shown to right here and on Figs. 2-5. Optimum angle was taken as the peak of the fitted Gaussian curve.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Histograms of optimum knee angle for torque generation of vastus intermedius muscles of rats trained on a declined or inclined treadmill. Muscles for decline-trained rats have their optimum at a smaller knee angle, corresponding to a longer muscle length. ECT, eccentric contraction test.

Sarcomere counts. The number of sarcomeres in series per fiber was estimated for 24 fibers from each of 16 muscles, 8 each from decline and incline-trained rats, as described by Lynn and Morgan (19). The overall means, 3,475 ± 27 for the 192 fibers from eight decline-trained rats and 3,193 ± 27 for the 192 fibers from eight incline-trained rats, were significantly different, with P < 0.0001. Means for individual animals are shown in Fig. 3.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Mean number of sarcomeres in series in muscle fibers compared with optimum knee angle for torque generation. Vertical error bars are SE for the 24 fibers measured from each muscle. Horizontal error bars are estimated error (SD) of position of peak of fitted Gaussian curve for the same muscle, as supplied by the fitting algorithm in Igor (see Data processing). Sarcomere counts are clearly dependent on training and strongly correlated with the optimum angle. Line has slope predicted if all muscles had the same tendon lengths and skeletal dimensions. Vertical position of the line is arbitrary.

ECT. Six muscles from incline-trained rats and eight from decline-trained rats completed the protocol comprising initial measurement of optimum angle, ECT, and post-ECT measurement of optimum angle. Several measures were made of the effect of the ECT. The first was the shift in optimum from before to after the series of eccentric contractions (Fig. 4). Negative angles of shift imply smaller included knee angles and longer muscle lengths for maximum torque after ECT, as predicted. The mean shift in optimum angle resulting from ECT was significantly less (P < 0.0001) for the muscles from decline-trained rats (1.4 ± 0.25°; n = 8) than for those from incline-trained rats (12.1 ± 0.75°; n = 6).

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Effect of eccentric contractions on optimum knee angle as a function of pre-ECT optimum angle. Each muscle was subjected to 20 stretches (27° decrease of knee angle from 90°).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Effect of a series of eccentric contractions on post-ECT tension as a function of pre-ECT optimum angle. Tension is expressed as percentage of pre-ECT tension. Muscles with their pre-ECT optima at smaller angles (longer muscle lengths) were less affected by eccentric contractions than were those with larger optimum angles. A: post-ECT tension measured at post-ECT optimum angle. B: post-ECT tension measured at pre-ECT optimum angle.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Measurement of damage. While the most direct evidence of muscle damage is probably histological evidence, this is very difficult to quantify because of the extreme sampling difficulties and the continued development of histological changes for days and even weeks after eccentric contraction exercise (10, 14). A fiber that has a single half sarcomere of a single myofibril disrupted immediately after the exercise may be necrotic along its whole length some hours later. For this reason, most studies use other parameters, including drop in tension (see Mechanism of training), lack of fusion at low stimulation frequencies (5, 7, 23) and shift in the length-tension relationship (16, 29), as well as nonmechanical measures such as pain and tenderness levels of muscle and circulating proteins and other measures (14), to quantify damage. The shift in length-tension relationship has the advantages of being fully apparent at the end of the exercise, simply measured, largely independent of fatigue, and readily understood in terms of our theory (29) as evidence of the first stage of the process that leads to histologically observable damage.

Choice of controls. The choice of controls for experiments of this type has been considered previously (19). The conclusion was that the best comparison is between incline- and decline-exercised rats, rather than between unexercised and decline-exercised rats, to control for the effects of exercise as such and the probable variability in the amount of exercise undertaken spontaneously by nominally unexercised animals. This decision does make it impossible to answer the question, "Which exercise group changed?" Previous sarcomere counts indicated considerable variability, but on average both incline- and decline-trained groups did (19).

Mechanism of training. The existence of a rapid adaptation or training of muscle to suit eccentric or concentric exercise, such as that which angled treadmill running provides for the vasti, has been shown by many workers in many different animals, muscles, and exercises (3, 5, 6, 9, 26). They all showed reduced evidence of damage from a test bout of eccentric contractions after training involving eccentric contractions, but none showed any structural differences in the muscles that might explain the difference in susceptibility to damage. Morgan (21) suggested that the structural difference could be the number of sarcomeres in series in the fibers, increased by eccentric contraction training and/or decreased by concentric contraction training. To take the rat example, this would cause the contractions involved in the test bout of decline running to occur at shorter sarcomere lengths in muscles from decline-trained rats than in the muscles from incline-trained rats. This would avoid operation on the plateau and descending limb of the sarcomere length-tension curve and hence the sarcomere length instabilities that are to be expected there during lengthening and are postulated to be the first stage in the process leading to damage from such exercise (21). A previous study by Talbot and Morgan (27) confirmed this. It was found that more damage occurred to isolated toad muscles when the stretch occurred on the descending limb of the length-tension curve as opposed to the ascending limb. The reduced strain on sarcomeres for a given joint movement would also provide a contribution to protection. Such a difference in the number of sarcomeres in series has been reported (19) by using fixation and acid digestion. The present experiments confirm those results by the completely independent method of measuring optimum knee angle as well as by repeating those experiments and showing a strong correlation between the two methods when all the muscles are considered. There would seem to be no possible source of error that could produce such strongly correlated differences in such independently measured parameters.

Changes in tendon. The finding of different sarcomere numbers raises the question of whether the tendon length changes to partially compensate for differences in the number of sarcomeres. If it did not, then tension generation at short muscle lengths (extreme knee extension) would be compromised after eccentric contraction training. The results (Fig. 3) suggest that some adaptation of the tendon does occur within the 1-wk time span investigated, but the magnitude of the change is subject to considerable uncertainty.

Conclusion. We conclude that eccentric contraction training of muscles, compared with concentric contraction training, leads to muscles with more sarcomeres connected in series and more resistant to damage from eccentric test contractions. This rapid adaptation of the number of sarcomeres to the pattern of use of the muscle is in accord with the observations of Goldspink (11) and the predictions of Morgan (21), adding further credibility to the underlying proposal that lengthening of active muscle at long length is distributed very nonuniformly, with a few sarcomeres that are stretched to beyond filament overlap accounting for most of the length change in each myofibril.