Contractile and cellular remodeling in rabbit skeletal...

9
Contractile and cellular remodeling in rabbit skeletal muscle after cyclic eccentric contractions RICHARD L. LIEBER, MARY C. SCHMITZ, DEV K. MISHRA, AND JAN FRIDfiN Departments of Orthopaedics and Applied Mathematics and Engineering ScienceIBioengineering, Biomedical Sciences Graduate Group, University of California and Veterans Affairs Medical Centers, San Diego, California 92161; and Departments of Anatomy and Hand Surgery, University of Umeb, S-901 87 Urned, Sweden Lieber, Richard L., Mary C. Schmitz, Dev K. Mishra, and Jan Frid6n. Contractile and cellular remodeling in rabbit skeletal muscle after cyclic eccentric contractions. J. A&. Physiol. 77(4): 1926-1934, 1994.-The time course of muscle contractile and cellular properties was studied in rabbit ankle flexor muscles after injury producedby eccentric exercise.Cy- clic eccentric exercisewasproduced by increasingthe tibiotar- sal angle of the rabbit while activating the peroneal nerve by useof transcutaneous electrodes.Muscle properties were mea- sured 1,2,3,7,14, and 28 days after exercise to define the time course of muscle changes after injury. A control group receiving only isometric contraction wasused to study the effect of cyclic activation itself. The magnitude of the torque decline after 1 day was the same with use of isometric or eccentric exercise, but eccentric exerciseresulted in a further decrease in torque after 2 days, at which time isometrically exercisedmuscles had fully recovered. The most prominent morphological changes in the injured muscle fibers were the lossof antibody staining for the desmin cytoskeletal protein and deposition of intracellular fibronectin, evenwhen the injured muscle fibers retained their normal complement of contractile and enzymatic proteins. The presence of fibronectin inside the myofibers indicated a loss of cellular integrity. Invasion by inflammatory cellswas apparent on the basis of localization of embryonic myosin. Thus eccen- tric exercise initiates a series of events that results in disrup- tion of the cytoskeletal network and an inflammatory response that could be the mechanismfor further deterioration of the contractile response. eccentric exercise; desmin;cytoskeleton; inflammation IT IS WIDELY AGREED that skeletal muscle injury and muscle soreness are associated with forced lengthening of an activated skeletal muscle, i.e., eccentric contraction (for reviews see Refs. 7 and 21). Recently we demon- strated that acute skeletal muscle injury can be fiber type specific (14) and that injury results primarily from the strain imposed on the muscle as opposed to the stress (15). Analogous experiments have been performed on isolated muscle groups in humans (6,18). Recent reports (4, 5, 10) demonstrated that, in addition to direct me- chanical damage, postexercise inflammation can also cause muscle damage in excess of that caused by the ini- tial injury, as indicated by elevated serum enzymes and circulating leukocyte levels. In a previous study of the time course of muscle injury, using the mouse extensor digitorum longus (EDL) mus- cle, McCully and Faulkner (17) measured the time course of muscle injury and recovery in a surgical model. They found that muscle force transiently decreased to a mini- mum after -3 days and then recovered after -14 days. However, the surgical procedure itself caused a 20% de- crease in muscle force and probably also induced an in- flammatory reaction, which complicated interpretation of their results. Faulkner et al. (9) then developed a non- invasive method of controlled ankle plantarflexion that documented a similar time course of injury and recovery and also demonstrated differential injury to the different dorsiflexor muscles, with the mouse EDL decreasing in maximum tetanic tension by -50% and the tibialis ante- rior (TA) by -40% 3 days after the initial injury. Given the evidence of differential injury to the TA and EDL, the putative role of the inflammatory response (18, 25), and our ability to localize specific intracellular pro- teins, we developed a noninvasive model of the rabbit TA similar to that used by Ashton-Miller et al. (2). We used monoclonal antibodies to localize muscle cytoskeletal and contractile proteins that permitted determination of the muscle fiber’s response to eccentric contraction-in- duced muscle injury. We tested the hypothesis that mus- cle cytoskeletal dissolution and loss of myofiber integrity are associated with the development of exercise-induced muscle damage and the drop in muscle force production. METHODS Animal cure. The TA and EDL muscles of the New Zealand White rabbit (3.2 t 0.4 kg) were chosenfor this study because of their accessibility and architectural differences. Animal care adhered to the “National Institutes of Health Guide for the Care and Useof Laboratory Animals” and wasapproved by the University of California San DiegoCommittee on Care and Use of Laboratory Animals. After terminal experiments, all ani- mals were killed by intravenous injection of pentobarbital so- dium (Euthanol). Experimental design. Ankle dorsiflexion torque was mea- sured before exercise treatment and at one recovery time pe- riod 1 (n = 8), 2 (n = 7), 3 (n = 7), 7 (n = 8), 14 (n = 6), and 28 days later (n = 6). After initial torque values were measured, animals were eccentrically or isometrically exercised for 30 min. An isometric group wastested 1,2, and 7 days after stimu- lation (n = G/time period) to provide an estimate of the torque decline due to 30 min of repetitive stimulation alone. After the specifiedrecovery period, contractile properties were again re- corded, followed by measurement of isolated TA and EDL forces. In vivo noninvasive treatment. Rabbits were anesthetized with a subcutaneous injection of a ketamine-xylazine-acepro- mazine cocktail (50, 5, and 1 mg/kg body mass, respectively) and maintained on isoflurane anesthesia (2%, 1 l/min). Heart and respiratory rates were manually monitored throughout testing. A dual-mode servomotor (model 6400, CambridgeTechnol- ogy, Cambridge,MA), with an adjustablefoot plate attached to the motor arm, measureddorsiflexion torque during muscle 1926 0161~7567/94 $3.00 Copyright 0 1994 the American Physiological Society

Transcript of Contractile and cellular remodeling in rabbit skeletal...

Page 1: Contractile and cellular remodeling in rabbit skeletal ...muscle.ucsd.edu/More_HTML/papers/pdf/Lieber_JAP_1994.pdfskeletal muscle after cyclic eccentric contractions. J. A&. Physiol.

Contractile and cellular remodeling in rabbit skeletal muscle after cyclic eccentric contractions

RICHARD L. LIEBER, MARY C. SCHMITZ, DEV K. MISHRA, AND JAN FRIDfiN Departments of Orthopaedics and Applied Mathematics and Engineering ScienceIBioengineering, Biomedical Sciences Graduate Group, University of California and Veterans Affairs Medical Centers, San Diego, California 92161; and Departments of Anatomy and Hand Surgery, University of Umeb, S-901 87 Urned, Sweden

Lieber, Richard L., Mary C. Schmitz, Dev K. Mishra, and Jan Frid6n. Contractile and cellular remodeling in rabbit skeletal muscle after cyclic eccentric contractions. J. A&. Physiol. 77(4): 1926-1934, 1994.-The time course of muscle contractile and cellular properties was studied in rabbit ankle flexor muscles after injury produced by eccentric exercise. Cy- clic eccentric exercise was produced by increasing the tibiotar- sal angle of the rabbit while activating the peroneal nerve by use of transcutaneous electrodes. Muscle properties were mea- sured 1,2,3,7,14, and 28 days after exercise to define the time course of muscle changes after injury. A control group receiving only isometric contraction was used to study the effect of cyclic activation itself. The magnitude of the torque decline after 1 day was the same with use of isometric or eccentric exercise, but eccentric exercise resulted in a further decrease in torque after 2 days, at which time isometrically exercised muscles had fully recovered. The most prominent morphological changes in the injured muscle fibers were the loss of antibody staining for the desmin cytoskeletal protein and deposition of intracellular fibronectin, even when the injured muscle fibers retained their normal complement of contractile and enzymatic proteins. The presence of fibronectin inside the myofibers indicated a loss of cellular integrity. Invasion by inflammatory cells was apparent on the basis of localization of embryonic myosin. Thus eccen- tric exercise initiates a series of events that results in disrup- tion of the cytoskeletal network and an inflammatory response that could be the mechanism for further deterioration of the contractile response.

eccentric exercise; desmin; cytoskeleton; inflammation

IT IS WIDELY AGREED that skeletal muscle injury and muscle soreness are associated with forced lengthening of an activated skeletal muscle, i.e., eccentric contraction (for reviews see Refs. 7 and 21). Recently we demon- strated that acute skeletal muscle injury can be fiber type specific (14) and that injury results primarily from the strain imposed on the muscle as opposed to the stress (15). Analogous experiments have been performed on isolated muscle groups in humans (6,18). Recent reports (4, 5, 10) demonstrated that, in addition to direct me- chanical damage, postexercise inflammation can also cause muscle damage in excess of that caused by the ini- tial injury, as indicated by elevated serum enzymes and circulating leukocyte levels.

In a previous study of the time course of muscle injury, using the mouse extensor digitorum longus (EDL) mus- cle, McCully and Faulkner (17) measured the time course of muscle injury and recovery in a surgical model. They found that muscle force transiently decreased to a mini- mum after -3 days and then recovered after -14 days. However, the surgical procedure itself caused a 20% de-

crease in muscle force and probably also induced an in- flammatory reaction, which complicated interpretation of their results. Faulkner et al. (9) then developed a non- invasive method of controlled ankle plantarflexion that documented a similar time course of injury and recovery and also demonstrated differential injury to the different dorsiflexor muscles, with the mouse EDL decreasing in maximum tetanic tension by -50% and the tibialis ante- rior (TA) by -40% 3 days after the initial injury.

Given the evidence of differential injury to the TA and EDL, the putative role of the inflammatory response (18, 25), and our ability to localize specific intracellular pro- teins, we developed a noninvasive model of the rabbit TA similar to that used by Ashton-Miller et al. (2). We used monoclonal antibodies to localize muscle cytoskeletal and contractile proteins that permitted determination of the muscle fiber’s response to eccentric contraction-in- duced muscle injury. We tested the hypothesis that mus- cle cytoskeletal dissolution and loss of myofiber integrity are associated with the development of exercise-induced muscle damage and the drop in muscle force production.

METHODS

Animal cure. The TA and EDL muscles of the New Zealand White rabbit (3.2 t 0.4 kg) were chosen for this study because of their accessibility and architectural differences. Animal care adhered to the “National Institutes of Health Guide for the Care and Use of Laboratory Animals” and was approved by the University of California San Diego Committee on Care and Use of Laboratory Animals. After terminal experiments, all ani- mals were killed by intravenous injection of pentobarbital so- dium (Euthanol).

Experimental design. Ankle dorsiflexion torque was mea- sured before exercise treatment and at one recovery time pe- riod 1 (n = 8), 2 (n = 7), 3 (n = 7), 7 (n = 8), 14 (n = 6), and 28 days later (n = 6). After initial torque values were measured, animals were eccentrically or isometrically exercised for 30 min. An isometric group was tested 1,2, and 7 days after stimu- lation (n = G/time period) to provide an estimate of the torque decline due to 30 min of repetitive stimulation alone. After the specified recovery period, contractile properties were again re- corded, followed by measurement of isolated TA and EDL forces.

In vivo noninvasive treatment. Rabbits were anesthetized with a subcutaneous injection of a ketamine-xylazine-acepro- mazine cocktail (50, 5, and 1 mg/kg body mass, respectively) and maintained on isoflurane anesthesia (2%, 1 l/min). Heart and respiratory rates were manually monitored throughout testing.

A dual-mode servomotor (model 6400, Cambridge Technol- ogy, Cambridge, MA), with an adjustable foot plate attached to the motor arm, measured dorsiflexion torque during muscle

1926 0161~7567/94 $3.00 Copyright 0 1994 the American Physiological Society

Page 2: Contractile and cellular remodeling in rabbit skeletal ...muscle.ucsd.edu/More_HTML/papers/pdf/Lieber_JAP_1994.pdfskeletal muscle after cyclic eccentric contractions. J. A&. Physiol.

F

RABBIT HINDLIMB

i FROM \f

I srlMULATOR AXIS OF

ROTATION

FIG. 1. Method for noninvasive creation of eccentric contraction-induced injury. Percutaneous electrodes are used to stimulate peroneal nerve during forced ankle plantarflexion (heavy arrow). This results in eccentric contraction of dorsiflexors. N, nerve; Ant, anterior; Ext, extensor.

activation (Fig. 1). The motor was calibrated by hanging known masses on the foot plate at measured distances from the motor arm axis of rotation (calibration factor 2.5 V l N-’ l m-‘, ? = 0.99). Accuracy was +2%, and repeated-measurement error was 0.65 t 0.20%. The rabbit ankle joint center of rotation was aligned with the motor arm axis of rotation. Tibiotarsal and femorotibial angles were set to 100 and 90”, respectively, with use of a goniometer. Twitch threshold current was determined by activating the peroneal nerve via subcutaneous needle elec- trodes placed over the peroneal nerve in the region of the fibu- lar head (Fig. 1). Threshold currents below -3.5 mA indicated accurate electrode placement. If threshold current was too high, electrodes were repositioned to achieve optimal stimula- tion. Stimulation current was then increased until peak twitch torque was reached. Subsequent treatments and all dynamic torque measurements were made at two to three times this level to guarantee complete activation of the dorsiflexor mus-

Passive Stretch Isometric

cles. (Activation of antagonist muscles was easily detected, be- cause it resulted in torque waveforms that were in the wrong direction or showed abrupt changes during contraction. If this occurred, electrodes were repositioned.) For noninvasive ec- centric exercise treatment, the ankle was moved from -100 to 70° over a 400-ms period (stretch) and then returned to the starting position (shorten). The stretching movement during activation provided the eccentric exercise for TA and EDL muscles. This pattern was repeated every 2 s for 30 min, result- ing in 900 cyclic eccentric contractions. It was important to estimate the in vivo tensions experienced by the muscles during treatment to determine muscle forces occurring during eccen- tric exercise. We thus used the moment arm calculations (see below) to estimate the deformation magnitude for the TA and EDL and then imposed the identical length changes on the isolated muscles (Fig. 2). During eccentric contraction, the EDL experienced - 140% maximal isometric tetanic force (P,),

Eccentric

- T B-e----- 1: : b : . : . . . i :

L

FIG. 2. Muscle forces calculated to occur during in vivo plantarflexion. Muscles were deformed by 10 mm to mimic muscle deformations during ec- centric exercise. Because of differential fiber length, extensor digitorum longus (EDL) experiences a greater tension [ -140% of maximal isometric tetanic force (P,)] than tibialis anterior (TA; - 115% of P,). Dashed line, isometric P,.

Page 3: Contractile and cellular remodeling in rabbit skeletal ...muscle.ucsd.edu/More_HTML/papers/pdf/Lieber_JAP_1994.pdfskeletal muscle after cyclic eccentric contractions. J. A&. Physiol.

1928 CONTRACTILE AND MORPHOLOGICAL CHANGES IN MUSCLE

whereas the TA experienced only -115% P,. Therefore the “treatment tension” was greater for the EDL than for the TA.

For determination of joint torque, twitch and tetanic contrac- tions were measured. Stimulation frequencies of 5, 10, 15, 20, 40,60,80,100, and 200 Hz with a train duration of 1,000 ms (for frequencies <20 Hz) or 750 ms (for frequencies >20 Hz) were used to generate the force-frequency relationship. All tetanic records demonstrated a clear tension maximum (Fig. 2). Maxi- mum tetanic torque was defined as the peak of the force-fre- quency relationship, which occurred between 100 and 200 Hz for all rabbits tested.

In situ muscle contractile testing. The TA and EDL tendons were each clamped to a separate motor arm (models 360 and 6400, Cambridge Technology). We used methods described by Lieber et al. (16). Briefly, an incision was made from the ankle to the midthigh, and the peroneal nerve was located. The leg was immobilized using Steinmann pins at the distal femur and midtibia, which were secured in a rigid fixture, and distal TA and EDL tendons were clamped and aligned with their respec- tive motor measurement axes. A small cuff electrode was placed around the nerve. Muscle temperature was maintained at 37°C with use of radiant heat, mineral oil, and a servo-tem- perature controller (model 73A, Yellow Springs Instrument, Yellow Springs, OH). TA and EDL muscle length were ad- justed until twitch tension was maximal, and contractile proper- ties were measured during twitch and tetanic contractions. Stimulation frequencies were the same as those used in the noninvasive procedure. Half-fusion frequency (HFF) was de- fined as the frequency at which the tension was 50% of PO. This was determined by linear interpolation of the force-frequency curve with use of only the data points immediately above and below 50% of P, (i.e., in the linear region of the fusion-fre- quency curve). Finally, muscles were repetitively stimulated at 40 Hz, 330 ms on and 670 ms off, for 2 min, as described by Burke et al. (3). This test was designed to fatigue the muscle fibers but not the neuromuscular junction or the motor nerve. Fatigue index was defined as the tension after 2 min of stimula- tion divided by initial tension. Superficial biopsies from the TA and EDL midbellies were prepared for electron microscopy, as described by Lieber et al. (16). Remaining portions of the mus- cles were oriented in OCT embedding medium (Lab-Tek Prod- ucts, Naperville, IL) on a piece of paper, frozen in isopentane cooled by liquid nitrogen (-159”C), and stored at -80°C for histochemical processing.

Hematology and blood chemistry. Blood was collected for anal- ysis at three points during the course of treatment. For the initial noninvasive exercise session, blood was collected immedi- ately after subcutaneous sedation was performed and before isoflurane administration. Blood was again collected immedi- ately after the exercise session, usually within 5 min of comple- tion. Final blood was collected just before the terminal testing, again just after subcutaneous sedation was given.

Whole blood for complete blood count and hemoglobin was collected in EDTA-lined Microtainer tubes and analyzed on a Serono-Baker System 9000 device (Serono-Baker, Allentown, PA). White blood cell differential was manually counted by a technician experienced in this method and blinded to the treat- ment protocol.

Blood for serum chemistry analysis was collected in an inert gel serum separator Microtainer, immediately centrifuged, and analyzed by a Kodak Ektachem System DT60 machine (East- man Kodak, Rochester, NY). Serum creatinine, creatine ki- nase, and lactate dehydrogenase concentrations were mea- sured.

Immunohistochemistry methods. Tissue analysis was per- formed on six EDL and six TA muscles per time period, al- though these were not necessarily the identical muscles on which contractile data were obtained. (Twenty-four of 42 mus- cles for which contractile data were obtained were also sub- jected to tissue analysis. The remaining muscles were gener-

ated using the identical protocol but harvested without con- tractile testing.) Serial transverse sections were cut using a cryostat microtome at -20°C and mounted on glass slides. Sec- tions were stained with Gomori Trichrome to visualize fascicle outlines, NADH, and adenosinetriphosphatase (pH 9.4). Cryo- sections were stained with antibodies against laminin to visual- ize the fiber’s basal lamina (19); fibronectin (24) to detect intra- cellular deposits, thus indicating membrane lesions; desmin for evaluation of the structural integrity of the cytoskeletal net- work (22); and embryonic myosin as an index of fiber regenera- tion.

To quantify fiber size and distribution, morphometry was performed on laminin-stained sections after fiber typing with use of an interactive image analysis system connected to a Zeiss . Axiophot light microscope that was equipped with a camera (DAGE-MTI, Michigan City, IN). Measurements were per- formed using a calibrated digitizing pad in conjunction with a computer-controlled morphometry program (IBAS, Kontron, Eching Munich, Germany). All observers were blinded with re- spect to sample identity.

For each muscle, two sections free of large cracks or obvious sectioning artifacts were obtained from the central muscle por- tion. Within this section, nine fascicles were sampled. About 60 fibers were measured per fascicle (by stereological point count- ing), corresponding to -10% of the entire muscle, which is well above that recommended by Weibel (26). Point counting was used to define the area fraction of muscle fibers demonstrating the absence of the desmin protein or the expression of embry- onic myosin. All specimens were coded, and counting was per- formed by observers who were blinded to the code.

Muscle moment arm determination. After the muscles were excised and the animals were killed, TA and EDL tendons were exposed and the foot was partially skinned, with the ankle reti- naculum remaining intact. Silk suture (3-O) was tied to the proximal TA tendon stump, and the EDL tendons were grouped together and tied as one tendon. The suture-tendon junctions remained distal to the retinaculum during joint rota- tion while both sutures coursed beneath the retinaculum. The tibia was immobilized with two Steinmann pins, one in the proximal one-third and one in the distal one-third. The foot was then positioned in a custom jig, and the suture was connected to a toothed cable (model 33GBF, Berg, East Rocka- way, NY).

A 200-g mass was tied to the cable end, and the cable was placed over a toothed nonbacklash gear (model 32D416, Berg) mounted to a potentiometer used to measure tendon excursion (potentiometer calibration = 3.86 mm/V, accuracy = 0.57 t 0.13%). An electrogoniometer (model 150, Penny and Giles, Gwent, UK) was attached to the foot and enabled ankle angle measurement as the ankle was passed from 90’ dorsiflexion to full extension (180”). Joint angle and tendon excursion were recorded by a Macintosh IIcx computer running the Super- scope data acquisition program (G. W. Instruments, Cam- bridge, MA).

Muscle moment arms were then calculated as the slope of the tendon excursion-foot angle curve (Fig. 3) by use of the equation described by An et al. (1)

ds; _ lri

=C

de (I)

where ri represents the moment arm of the TA or EDL and &i/d6 represents the slope of the TA or EDL tendon excursion (si)-joint angle (0) relationship over 1OO-128”, because this was the same range of motion experienced during eccentric exer- cise. The data closely approximated a line (? > 0.98) for each foot.

TA or EDL muscle fiber strain was then calculated for each specimen as follows

Page 4: Contractile and cellular remodeling in rabbit skeletal ...muscle.ucsd.edu/More_HTML/papers/pdf/Lieber_JAP_1994.pdfskeletal muscle after cyclic eccentric contractions. J. A&. Physiol.

CONTRACTILE AND MORPHOLOGICAL CHANGES IN MUSCLE 1929

n

E

E

;20 0 .- 2

Slope = -10.1 mmlrad

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Tibiotarsal Angle (radians) FIG. 3. Sample raw data used to calculate moment arm. Slope of

tendon excursion-joint angle relationship, as determined by linear re- gression, is defined as moment arm. Moment arm equals 10.1 mm.

ri l A0 Ei = -

k

(2)

where ci represents TA or EDL strain, li represents TA or EDL muscle fiber length [calculated from TA or EDL muscle length multiplied by TA or EDL fiber length-to-muscle length ratio (l3)], and A0 represents the range over which the muscle was activated.

Statistical analysis. Contractile properties at each time pe- riod O-28 days after exercise were compared using one-way analysis of variance (SuperANOVA, Abacus Concepts, Berke-

0.8

IA

ley, CA). Paired comparisons of torque and tension values for eccentric vs. isometric exercise were then made using Fisher’s least-squared difference post hoc paired comparisons. Sample size was calculated based on the following equation

n = 2 l (016)~ l ( i t , , , + t2p,v)2 (3)

where n is required sample size, u is population standard devia- tion, 6 is difference desired to detect, Q is desired significance level, u is degrees of freedom, and ,8 is desired type II error rate. b,” and t,,,” are the t statistics for the selected cy, ,8, and u.

In these experiments, significance was set at 0.05 (i.e., cy = 0.05) and the statistical power at 280% (i.e., ,8 = 0.20). Because maximum tetanic tension has a standard deviation of ~300 g, to resolve a change in muscle tetanic tension of approximately this degree (i.e., 6 = 300 g), a sample size of five to six was required. Values are means t SE.

RESULTS

Contractile torque and muscle tension after treatment. A significant amount of growth occurred in these young rabbits throughout the experimental period. Whereas the average preexperiment rabbit mass was 2.6 t 0.1 kg, the average rabbit mass after 28 days was 3.4 t 0.2 kg. Because a strong linear relationship between rabbit mass and dorsiflexion torque was observed in all animals, dor- siflexion torque was corrected for animal mass with use of the following regression relationship: torque (N l m) = mass (kg) l 0.231 (N l m l kg-‘) - 0.33 (N l m) (P < 0.01, ? = 0.77). In practice, this affected only the values obtained 14 and 28 days after exercise.

Torque decreased from -0.7 No m to a minimum of 0.55 N l m 2-3 days after eccentric contraction and recov-

65, B

00 1

FIG. 4. A: time course of torque after noninvasive eccentric contraction (0) or noninvasive isometric contraction (0).

Stippled bar, rabbit dorsiflexion torque before exercise. B: time course of half-fusion frequency after noninvasive eccen- tric contraction (0) or noninvasive isometric contraction (0). Stippled bar, normal rabbit dorsiflexor half-fusion frequency. Values are means + SE (n = 6 animals/time point). * P < 0.05 between eccentric and isometric exercise.

5 10 15 20 Time After Exercise (Days)

Page 5: Contractile and cellular remodeling in rabbit skeletal ...muscle.ucsd.edu/More_HTML/papers/pdf/Lieber_JAP_1994.pdfskeletal muscle after cyclic eccentric contractions. J. A&. Physiol.

CONTRACTILE AND MORPHOLOGICAL CHANGES IN MUSCLE 1930

5ooo- A 5 . 5 4ooo-

.-

5 I- 0 3000-

.-

ii 3 I- A 2000-

n ul .

looo! .,.,.,., ',V ,'I - I = 1 0 1 2 3 4 5 6 7 8

Time Following Exercise (Days)

ered to nearly control levels after 7 days (Fig. 4A). Torque levels at 14 and 28 days after eccentric exercise were not significantly different from those observed be- fore exercise (P > 0.5). Interestingly, in the isometric contraction group, torque decreased significantly after I day to the level for the eccentric contraction group and then recovered to control levels only 2 days after isomet- ric exercise. These data suggest that the eccentric exer- cise itself triggered a series of events that further de- creased muscle force over time. In terms of contractile

0.7

0.6 3 1

x 0.5 $ = 0.4 J d* f* +*

2 l 1 *

.p 0.3-

% * f

) :::~~:~:~~~:~i::~~.:~~~i~~~:~~~~:~~~~~~~~~~~~~~~~,~~~:~~~~:~~~~~::~~~.:.:.:~:~~ ;.;:::;:.:::<;:;? ,.., :: ; :T ,., ,. .y,:;.::i. T .: ;, ,:, ,,,.

lL 0.2- .& ..,.,. v.h.w A. \.~....>A>~.....

f

.,,. > ,:,.,. > .) ,.,:,.,, . . . . y.:$.'.y:&*<<.: . . . . . . . . . . . . . . . . :.:.:.:.:.:.:.~~.:.:.:.:,:.:.:.:.:,:.:.:.:...:.,,:,:.: :,_ .,:.:.:.:...:,: : : ..:.:., . .b $. : :.;\: :. : :+: :.;,:A.v A.... .A. . . . .vE.:.:.:.:.;.. ./. ..: :'._..,.,,. ..,p.,i:,:~~i,,;..~~.~~~:.~~~ i,~~?~:.~~~:~~~~~~a:~~::~~~~::~~~~:~::~~~,~~,~:~.:~:::~ ;: j : /..; ,.:;jJ&;j;,: ,, _, ?; .: ,. ,.,...+.,.j ,y. .: : c .: ;,

. . . . . , . . . . . . &" '.~ . . :, . . . . . .% b. . . . . . . ..A.... . . . .A.. .A...... f :: .; : : :_:;, :;:I. i,::;.'; $:, ,:,r:' ,::<:,z,: : :, ,:, ~~~~::~~:~::::~~~~~~~::~::~~~:~:~~~:~~~~~:~:~~:~...~,:~:~;,:~~ ,:,_; :,:: .;:::': . ,.,.,.,. .A ..p.wp\ , ,, . . . . . . ..y$$.:.<* . . . . :" . . . . . . . : .::;..;y; .. : i',> ,c .'. .'::'. . . . L .,.,.,.,.,...,, .,.,., <A:.~.:~. . . 4 . . . . . . . . A .,. _.,,_ :&&.*.:: :.... :,..:.:.: . . . . . . :::~~:+w..:.. :- .> .A . . . . .Q. I.. . . . . :.::.:.:.:~.~:.:.:.::::::::':.:.:.,.:.:.:::~:::::: ~~~~~~:':.i'.,...."' x ,.,., ,., ,,, : _, v . . A.. ,.,. ._. .,.,.. _., _, . .,. .,. . . . . .A'::;::'. .: .P. ..:::; ;; ';;:'!!,.,yj,,.: :,: ,: ';; ':,:::: : '! .:, ; : ._ ; ,, ,, ,,, ,,. ,, ,. . . C". .' .'$',' ,., . . . . <,.:.. ,,,,i > :: ,::., ".'. .,. qpw..:.>:<+ .,.. ".'.y<... \_. .,../. :...: .,.,.,..,.,... ,,_. ‘,,,;*~,~: :x.x :.:..:: .i... ,. ,, _. ..A _i..,.... % '.:.'. .Y....... ,$ . . . . :,:.:.~.~~.:.:.~:'...;.:.~.~~~.~~,.~:...~ :....:..,...,.,.,.,.,., ,...., ;;...:,r.~.~3:8:.? y ..,.. :,:. .:, :, ..> ,.,. i .'.i : *. b.~~..~............... . .A... .v :.:.:...~ . . ..c...........,.....r,.,.,.,.,.,. 0 ,...,.,.,.,.,.,,,,,,,,_,,,,,~,,,,~~,, ,,,,,, . . . .,.,. x .,.,.: .,.,. x.,.&: .,.,.,,..,.,.,,.,..,.,.,,_,,,, ., >.': ;, '. .: '. ': 1 .: ;, :. ::, .':':.,: :: ', :., ,_..,.:, '. :j:;,:._. ,..... . . ..A.._... '..)" . . ..I. ,,: ,, ,.,: ,:. ,:. : ,,:_: ,,,, ,. ,, ._.. . . ..,:.i 0.1 - J

0.0 I - I - I - I - I - I . 1

-5 0 25 30

FIG. 6. Time course of change in fatigue index of TA muscles over 28 days after eccentric contraction. Fatigue index was the only contrac- tile parameter to remain significantly elevated 7-14 days after eccen- tric exercise. Isometric fatigue indexes were elevated for only 1-2 days (data not shown). Stippled bar, normal values for rabbit TA muscles. * P < 0.05 relative to baseline.

FIG. 5. Time course of muscle force change during 7 days after eccentric contraction. A: EDL muscles; B: TA muscles. l , Muscle forces measured after eccentric contraction; 0, muscle forces after isometric contraction. Magnitude of force change for EDL is significantly greater than that for TA. Stippled bar, normal values for rabbit EDL and TA muscles (means k SE). * P < 0.05 between eccentric and isometric exercise.

speed, the stimulation frequency at which the contractile record was half-fused, i.e., HFF, significantly increased in the eccentrically exercised muscles from an initial value of -48 Hz to -60 Hz 1 and 2 days after the initial exercise (Fig. 4B). HFF was only slightly changed in iso- metrically exercised muscles and, in all cases, returned to control levels after 7 days.

The isolated TA and the EDL demonstrated a differ- ential response to eccentric contraction. Both muscles demonstrated a transient decrease and then recovery of force throughout the experimental period (Fig. 5). How- ever, the magnitude of the force decline was much greater for the EDL (Fig. 5A) than for the TA (Fig. 5B). The EDL decreased in maximum tetanic tension by -75%, whereas the TA decreased only -40%. Maxi- mum tetanic tension after isometric contraction was equivalent to that observed after eccentric contraction at 1 day but was significantly greater than that observed after eccentric contraction at 2 days (Fig. 5; P < 0.01).

MuscZe moment arms. Kinematic measurements re- vealed that the TA and EDL muscle moment arms were nearly identical. Average TA moment arm acting at the ankle was 10.6 t 0.4 mm, whereas the EDL moment arm was 9.5 t 0.4 mm (n = 6 measurements/tendon). Given that the TA muscle fibers were, on average, 53.3 mm long, whereas the EDL fibers were only 17.7 mm long (14), we calculated that EDL muscle fibers strained 27%, whereas TA muscle fibers strained only 10%.

Fatigability of muscle. Interestingly, the only contrac- tile parameter that had not resolved to control levels by 7

Page 6: Contractile and cellular remodeling in rabbit skeletal ...muscle.ucsd.edu/More_HTML/papers/pdf/Lieber_JAP_1994.pdfskeletal muscle after cyclic eccentric contractions. J. A&. Physiol.

CONTRACTILE AND MORPHOLOGICAL CHANGES IN MUSCLE 1931

5000 2 1 A

2 6

t

FIG. 7. A: time course of serum creatine kinase after eccentric exercise. 0, Average of all experimental animals before exercise (stippled bar). Isometric exercise did not significantly change serum creatine kinase concentra- tion. B: time course of rabbit white blood cell (WBC) count. 0, Value obtained immediately before exercise. Note large decrease in WBC concentration immediately after exercise, which then returns to normal level after 1 day. No significant difference between isometric and ec- centric exercise was observed. * P < 0.05 relative to base- line.

2f .,,,., I - 1 - ’ - 1 0 5 10 15 20 25 30

Time After Exercise (Days)

days after eccentric exercise was the fatigue index (Fig. 6). The fatigue index remained significantly elevated for 14 days after eccentric exercise (P < 0.05) in the TA and for 7 days in the EDL. Fatigue index of both muscles returned to control levels only after 28 days. Muscles activated isometrically demonstrated the identical in- crease in fatigue index for 1 day, but most muscles (14 of 18) returned to control levels after 2 days.

Blood chemistry and hematology. Serum creatine ki- nase was significantly elevated 1 and 2 days after eccen- tric exercise but resolved after -7 days (Fig. 7). This represents a much faster transient response to eccentric exercise than is observed in humans (6,8) (Fig. 7A). Cir- culating white blood cell count significantly decreased immediately after eccentric exercise but then was re- stored to normal after only 1 day postexercise (Fig. 7B). However, because a similar effect was observed in the isometric contraction group, this probably simply repre- sents white blood cell trapping by the tissue.

Alterations in myofibrillar cytoskeleton. The most strik- ing change observed after eccentric exercise was the loss of desmin staining in various fibers across the muscle (Fig. 8A). Desmin is a cytoskeletal protein that is located in the muscle Z disk. Because the sections were -8 pm thick, loss of desmin staining indicated Z-disk disruption extending across several sarcomeres, which are each -2.5 pm long. As expected on the basis of contractile data, the percentage of “desmin-negative” fibers within a section was much greater in the EDL than in the TA. For

example, 3 days after eccentric exercise, >30% of the EDL fibers had lost desmin staining, whereas only -15% of TA fibers were desmin negative (Fig. 9A). This percentage rapidly decreased in both muscles, so that, by 7 days postinjury, the percentage was - 10% for the EDL and only 5% for the TA (P < 0.05). By 28 days postinjury, ~1% of desmin-negative fibers could be found. Interest- ingly, most, but not all, desmin-negative fibers were also fibronectin positive, indicating loss of cellular integrity accompanying cytoskeletal disruption (Fig. 8B). Desmin staining was lost even though the muscle fiber main- tained contractile and oxidative enzymes (Fig. 8C). In such micrographs, infiltration of inflammatory cells was apparent.

In the 18 muscles for which contractile and immunohis- tochemical analysis was performed 1 and 2 days after eccentric exercise, there was a significant negative corre- lation between maximum tetanic tension and the percent- age of desmin-negative fibers (Fig. 10). For the TA, this relationship was described as follows

tetanic tension (g)

= -24.6 (%desmin-negative fibers/g) (4)

+ 1,962 (g) (P < 0.001, ? = 0.50)

whereas that describing the relationship between EDL tetanic tension and percent desmin-negative fibers was described as follows

Page 7: Contractile and cellular remodeling in rabbit skeletal ...muscle.ucsd.edu/More_HTML/papers/pdf/Lieber_JAP_1994.pdfskeletal muscle after cyclic eccentric contractions. J. A&. Physiol.

1932 CONTRACTILE AND MORPHOLOGICAL CHANGES IN MUSCLE

FIG. 8. Serial sections of injured EDL muscle 3 days after eccentric exercise. A: desmin antibody; B: fibronectin antibody; C: oxidative en- zyme histochemistry for NADH enzyme. Note injured fibers in which desmin staining is lost (1 and 2) completely or partially. Fiber 1 is more severely damaged, because it also permits entrance of extracellular ti- bronectin into cell, whereas fiber 2 still excludes tibronectin. Fibers 1 and 2 have low oxidative capacity, as shown in C. Two normal fibers are also identified (3 and 4). Fibers 3 and 4 have higher oxidative capacity, as shown in C. Calibration bar, 250 pm.

tetanic tension (g)

= -72 (%desmin-negative fibers/g) (5)

+ 4,547 (g) (P < 0.0001, 3 = 0.82)

suggesting that the loss of the cytoskeletal protein des- min is a causal factor in muscle force decline for the TA and EDL.

Signs of muscle fiber regeneration. A specific marker for the myofiber regeneration process is the expression of embryonic myosin, a form of the myosin contractile pro- tein that is normally not present in adult skeletal muscle

fibers. We thus used the antiembryonic myosin antibody to detect the presence of regenerating fibers. Again, not surprisingly, a larger proportion of EDL than TA muscle fibers entered the degeneration-regeneration cycle. How- ever, the timing of myosin expression was different for the two muscles (Fig. 9B). The EDL demonstrated -10% regenerating fibers 3 days after injury and >15% regenerating fibers 7 days after injury. The TA, on the other hand, while also showing -10% of the fibers re- generating 3 days after injury, showed a decreased per- centage of -5% regenerating fibers 7 days after injury (P > 0.4). This probably represented a more rapid recovery of the TA than of the EDL in response to a less severe injury.

DISCUSSION

The purpose of this study was to define the time course and nature of eccentric contraction-induced injury to rabbit hindlimb muscle fibers. Our main finding was that severe cellular disruption occurred in both muscles, which consisted of loss of normal staining for the cyto- skeletal protein desmin; deposition of intracellular fibro-

T

Day 1

i! Day 3 Day 7 Day 28

B

T h

T L

-

Day 1 Day 3 Day 7

Time

, Day 28

FIG. 9. A: area fraction of fibers demonstrating absence of cytoskele- tal protein desmin on immunohistochemical sections. Damage is greater in EDL than in TA, which is consistent with contractile results (cf. Fig. 5). B: area fraction of fibers demonstrating a regenerative re- sponse, as indicated by expression of embryonic myosin on immunohis- tochemical sections. Regenerative response is greater in EDL than in TA, which is also consistent with contractile results.

Page 8: Contractile and cellular remodeling in rabbit skeletal ...muscle.ucsd.edu/More_HTML/papers/pdf/Lieber_JAP_1994.pdfskeletal muscle after cyclic eccentric contractions. J. A&. Physiol.

CONTRACTILE AND MORPHOLOGICAL CHANGES IN MUSCLE 1933

0 0

:

0

‘* 1

% 0 a 0

l

0 10 20 30 40 50 Percent Desmin Negative Fibers

FIG. 10. Relationship between maximum tetanic tension and per- centage of desmin-negative fibers for TA (0) and EDL (0). In both cases, there was a significant correlation between the 2 variables: te- tanic tension (g) = -24.6 (%desmin-negative fibers/g) + 1,962 (g) (P < 0.001, ? = 0.50) for TA, and tetanic tension (g) = -72 (%desmin-nega- tive fibers/g) + 4,547 (g) (P < 0.0001, 4 = 0.82) for EDL. Data were from animals 1 or 2 days after eccentric exercise.

nectin, indicating loss of cellular integrity; expression of embryonic myosin, signaling initiation of the regenera- tion process; and cellular infiltration, indicating initia- tion of an inflammatory response. In addition, differen- tial injury to the TA and EDL occurred, which was con- sistent with the magnitude of the strain experienced by the muscles during eccentric exercise (15).

Role of inflammation in muscle injury. The contractile data obtained generally mimic the results from the inva- sive study of McCully and Faulkner (17), in which muscle force decreased to a minimum after 2-3 days and almost fully recovered after 1 wk. We also observed the presence of intramuscular inflammatory cells, which was consis- tent with the time course of force decline after the initial exercise bout. Cellular infiltration was uniquely asso- ciated with the eccentric exercise itself in that isometri- cally exercised muscles were devoid of infiltrating cells and the same force decrement was not observed after isometric exercise of the same duration. Thus we hy- pothesize that eccentric exercise initiates events that re- sult in extravasation of leukocytes and monocytes from the bloodstream and infiltration by these cells into the tissue. This cellular infiltration results in further tissue degradation, probably by release of proteolytic enzymes, which are designed to enable tissue remodeling. A similar scenario has been proposed by Cannon et al. (5) in hu- man exercise studies. In the current study, the magni- tude of the force drop 1 day after eccentric exercise was -50% of the entire response, which then occurred ov the next 2-3 days. If the hypothetical damage scheme

‘er is

correct, postexercise cellular infiltration and the asso- ciated release of proteolytic enzymes were responsible for as much injury as the initial mechanical and meta- bolic insults.

Fiber cytoskeletal disruption. The fact that the muscle fiber cytoskeleton reveals complete disruption within 1 day o f eccentric exercise further emphasizes its dynamic nature and ability to respond to severe mechanical events. We observed loss of desmin staining in cells where other contractile and metabolic enzymes appeared

completely normal and in cells where the extracellular matrix protein fibronectin was still excluded. Thus cellu- lar disruption is not required for cytoskeletal derange- ment and may even represent a deliberate proteolytic event that permits rapid intracellular remodeling. The dynamic nature of the cytoskeleton, which has proven important in regulating other cellular synthetic and deg- radative processes, may also be involved in sending mes- sages to muscle cells to respond to an exercise stimulus. Perhaps the desmin loss has the functional purpose of detaching the myofibrillar apparatus from the extracel- lular matrix, in effect putting the muscle in “neutral” during the remodeling process.

The differential time course of embryonic myosin ex- pression between the TA and the EDL may simply indi- cate that the TA response, which appeared to be less severe in magnitude, is nearing resolution by 7 days, whereas the more severe EDL response is still in the process of fully developing.

Fiber type specificity in muscle injury. Finally, these data lead to the conclusion that different muscle fibers within a muscle and different muscles within synergistic groups can be differentially injured in response to eccen- tric exercise. Whereas our previous acute experiments revealed that these injured fibers were exclusively of the fast-twitch glycolytic fiber type, no such clear-cut dis- tinction could be made in the present study. Thus the elevated fatigue index is probably not a reflection of fiber type-specific damage.

Our inability to demonstrate a fiber type-specific effect was in large part due to the fact that the injury process itself, which often resulted in fiber regeneration, made traditional histochemical fiber typing impossible. How- ever, even in the cases where fiber typing was possible, a number of type IIa and even some type I fibers demon- strated morphological abnormalities. Thus, whereas fiber oxidative enzyme potential may be one factor that predisposes a fiber to injury, it is clearly not the only factor. For example, a type IIa fiber in the EDL, which experiences ~20% strain, is much more likely to be in- jured than a type IIa fiber in the TA, which experiences only -10% strain. What cannot be answered presently, however, is the relative risk of injury to a type IIa fiber in the EDL, which experiences 20% strain compared with a type IIb fiber in the TA, which experiences 10% strain. Clearly, future experiments are required to quantify the relative role of metabolism and mechanics in muscle fiber injury.

The prolonged elevation in fatigue index for the TA and EDL (Fig. 6) is difficult to interpret. First, it is not clear exactly what part of the neuromuscular system is responsible for the force decline measured during the 2- min fatiguing protocol. Whereas the method was devel- oped to classify relatively homogeneous fibers within a motor unit (3), it is often applied to the study of whole muscle properties. Fatigue indexes are useful for discrim- inating between motor unit types but may not be as ef- fective in discriminating between muscles with mixed fiber type populations. We measured a significant in- crease in fatigue index over the first 2 days, even in iso- metrically exercised muscles, so that such an elevation cannot be uniquely attributed to tissue injury. Regener-

Page 9: Contractile and cellular remodeling in rabbit skeletal ...muscle.ucsd.edu/More_HTML/papers/pdf/Lieber_JAP_1994.pdfskeletal muscle after cyclic eccentric contractions. J. A&. Physiol.

1934 CONTRACTILE AND MORPHOLOGICAL CHANGES IN MUSCLE

ating muscle fibers tend to be smaller than normal, and it is thus possible that this decreases fatigability by de- creasing oxygen diffusion distance. However, because the magnitude of the regenerative response was greater for the EDL than for the TA, especially after 7 days, and their fatigue indexes were not significantly different (P > 0 6) . 9 this explanation is unlikely.

In summary, these studies have revealed that eccentric exercise initiates a series of events that results in loss of contractile force. A most significant finding is the loss of the desmin protein after cyclic eccentric exercise, which is associated with prolonged loss in contractile function and the infiltration of inflammatory cells. Desmin loss probably precedes inflammation on the basis of the ob- servation that not all desmin-negative fibers demon- strate loss of sarcolemmal integrity. Clearly, future stud- ies are needed to define the very early events responsible for the loss of this cytoskeletal protein.

We thank Taby Ahsan, Chris Giangreco, Ulla Hedlund, Bridget Ma- taban, and Tina Pate1 for technical assistance and Prof. Lars-Eric Thornell for fruitful discussions.

This work was supported by National Institute of Arthritis and Mus- culoskeletal and Skin Diseases Grant AR-40050, the Veterans Affairs, and the Swedish Medical Research Council.

A portion of this work has been presented elsewhere (20). Address for reprint requests: R. L. Lieber, Dept. of Orthopaedics

(V-151), University of California San Diego School of Medicine and VA Medical Center, 3350 La Jolla Village Dr., San Diego, CA 92161.

Received 10 May 1993; accepted in final form 26 April 1994.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

1.

2.

3.

4.

5

6.

7.

An, K. N., Y. Ueba, E. Y. Chao, W. P. Cooney, and R. L. Linscheid. Tendon excursion and moment arm of index finger muscles. J. Biomech. 16: 419-425, 1983. A&on-Miller, J. A., Y. He, V. A. Kadhiresan, D. A. McCub- brey, and J. A. Faulkner. An apparatus to measure in vivo bio- mechanical behavior of dorsi- and plantarflexors of mouse ankle. J. Appl. physiol. 72: 1205-1211, 1992. Burke, R. E., D. N. Levine, P. Tsairis, and F. E. Zajac. Physi- ological types and histochemical profiles in motor units of the cat gastrocnemius. J. Physiol. Lond. 234: 723-748, 1973. Cannon, J. G., S. N. Meydani, R. A. Fielding, M. A. Fiatar- one, M. Meydani, M. Farhangmehr, S. F. Orencole, J. B. Blumberg, and W. J. Evans. Acute phase response in exercise. II. Associations between vitamin E, cytokines, and muscle proteoly- sis. Am. J. Physiol. 260 (Regulatory Integrative Comp. Physiol. 29): Rl235-Rl240,1991. Cannon, J. G., S. F. Orencole, R. A. Fielding, M. Meydani, S. N. Meydani, M. A. Fiatarone, J. B. Blumberg, and W. J. Evans. Acute phase response in exercise: interaction of age and vitamin E on neutrophils and muscle enzyme release. Am. J. Phys- iol. 259 (Regulatory Integrative Camp. Physiol. 28): R1214-Rl219, 1990. Clarkson, P. M., and I. Tremblay. Exercise-induced muscle damage, repair and adaptation in humans. J. Appl. Physiol. 65: l-6, 1988. Evans, W. J., and J. G. Cannon. The metabolic effects of exer-

19.

20.

21.

22.

23.

24.

25.

26.

cise-induced muscle damage. In: Exercise and Sport Science Re- views. Baltimore, MD: Williams 8~ Wilkins, 1991, pp. 99-125. Evans, W. J., C. N. Meredith, J. G. Cannon, C. A. Dinarello, W. R. Frontera, V. A. Hughes, B. H. Jones, and H. G. Knutt- gen. Metabolic changes following eccentric exercise in trained and untrained men. J. Appl. Physiol. 61: 1864-1868, 1985. Faulkner, J. A., D. A. Jones, and J. M. Round. Injury to skele- tal muscles of mice by forced lengthening during contractions. Q. J. Exp. Physiol. 74: 661-670, 1989. Fielding, R. A., T. J. Manfredi, W. Ding, M. A. Fiatarone, W. J. Evans, and J. G. Cannon. Acute phase response in exer- cise. III. Neutrophil and IL-l@ accumulation in skeletal muscle. Am. J. Physiol. 265 (Regulatory Integrative Comp. Physiol. 34): Rl66-Rl72,1993. Friden, J., R. L. Lieber, and L.-E. Thornell. Subtle indications of muscle damage following eccentric contractions. Acta Physiol. Stand. 142: 523-524,199l. Howell, J. N., G. Chleboun, and R. Conatser. Muscle stiffness, strength loss, swelling and soreness following exercise-induced in- jury in humans. J. Physiol. Lond. 464: 183-196,1993. Lieber, R. L., and F. T. Blevins. Skeletal muscle architecture of the rabbit hindlimb: functional implications of muscle design. J. Morphol. 199: 93-101, 1989. Lieber, R. L., and J. Friden. Selective damage of fast glycolytic muscle fibers with eccentric contraction of the rabbit tibialis ante- rior. Acta Physiol. Stand. 133: 587-588, 1988. Lieber, R. L., and J. Friden. Muscle damage is not a function of muscle force but active muscle strain. J. Appl. Physiol. 74: 520-526, 1993. Lieber, R. L., J. 0. Friden, andT. G. McKee-Woodburn. Dam- age of the rabbit tibialis anterior induced by eccentric contractions of twenty-five percent strain. J. Appl. Physiol. 70: 2498-2507,199l. McCully, K. K., and J. A. Faulkner. Injury to skeletal muscle fibers of mice following lengthening contractions. J. Appl. Physiol. 59: 119-126,1985. Newham, D. J., G. McPhail, K. R. Mills, and R. H. T. Ed- wards. Ultrastructural changes after concentric and eccentric con- tractions. J. NeuroZ. Sci. 61: 109-122, 1983. Sanes, J. R., and J. M. Cheney. Laminin, fibronectin and colla- gen in synaptic and extrasynaptic portions of muscle fiber base- ment membrane. J. Cell Biol. 93: 442-451, 1988. Schmitz, M. C., D. K. Mishra, C. Giangreco, J. Fridbn, and R. L. Lieber. Differential injury to tibialis anterior and extensor cligitorum longus muscles during eccentric exercise in rabbits (Ab- stract). Trans. 39th Orthop. Res. Sot. 39: 374, 1993. Stauber, W. T. Eccentric action of muscles: physiology, injury, and adaptation. Exercise Sport Sci. Rev. 17: 157-185, 1989. Thornell, L.-E., A. Eriksson, B. Johansson, U. Kjorell, W. W. Franke, I. Virtanen, and V. P. Lehto. Intermediate fila- ment and associated proteins in heart Purkinje fibers: a mem- brane-myofibril anchored cytoskeletal system. Ann. NY Acad. Sci. 455: 213-240,1985. Thornell, L.-E., and M. G. Price. The cytoskeleton in muscle cells in relation to function. Biochem. Sot. Trans. 19: 1116-1120, 1991. Vartio, T., L. Laitinen, 0. Narviinen, M. Cutolo, L.-E. Thor- nell, L. Zardi, and I. Virtanen. Differential expression of the ED sequence-containing form of cellular fibronectin in embryonic and adult human tissues. J. Cell Sci. 88: 419-430, 1987. Vihko, V., A. Salminen, and J. Rantamaki. Acid hydrolase activity in red and white skeletal muscle of mice during a two-week period following exhausting exercise. Pfluegers Arch. 378: 99-106, 1978. Weibel, E. R. Stereological Methods. Practical Methods for Biologi- cal Morwhometry. New York: Academic. 1980. vol. 1.