J~URNALOF NEUROPHYS~OLOGY Vol. 48, No. 1, July 1982, Printed

in U.S.A.

Architectural, Histochemical, and Contractile Characteristics of a Unique Biarticular Muscle: the Cat Semitendinosus SUE C. BODINE, ROLAND R. ROY, DEBRA A. MEADOWS, RONALD F. ZERNICKE, ROBERT D. SACKS, MARIO FOURNIER, V. REGGIE EDGERTON

AND

Department of Kinesiology, Neuromuscular Research Laboratory, University of Califurnia, Los Angeles, California 90024

SUMMARY

AND

CONCLUSIONS

I. The semitendinosus (ST) muscle of the cat consists of two anatomically distinct sets of fibers connected in series by a dense connective tissue band that divides the “muscle” into a proximal and distal compartment, each having separate innervation. 2. Isometric and isotonic contractile properties of the ST were studied for three stimulation conditions: 1) proximal activated (STp), 2) distal-activated (STd), 3) whole muscle activated (ST). In addition, the architectural and histochemical profiles of each compartment were determined. 3. The fiber type composition was similar in the STp and STd, with the majority of the fibers staining darkly for alkaline myofibrillar adenosine triphosphatase ( ATPase). The muscle fibers of the STd were twice as long as the fibers of the STp. The cross-sectional areas of the two compartments were similar. 4. The in situ maximum isometric tension (P,) was identical for all three stimulation conditions. However, the maximum twitch tension (P,) was greatest when ST was activated and least when only the STp was activated. The time to peak tension (TPT) and half relaxation times (95 RT) were the same for all three conditions. The absolute maximal shortening velocity (V,,,) revealed an orderly progression of speeds in relation to the fiber length. When the V,,, of the STd (424 t 26 mm/s) and STp (224 t 23

mm/s) were added, the sum was approximately the same as the V,,, obtained for the ST (624 t 30 mm/s). The intrinsic speed of shortening (millimeters per second per 1,000 sarcomeres) recorded at the common tendon of insertion was the same for all three conditions. 5. The identical nature of the P, values demonstrated the significance of the architectural design, specifically, the physiological cross-sectional area, in determining the potential force production of a muscle. In contrast, the P, values were significantly different for each stimulation condition. These differences were probably influenced in part by the “stiffness-compliance” properties of the muscle-connective tissue unit, which varied with different activation states of the ST muscle. 6. The similarity in TPT and 1/2RT values for all three conditions illustrated the independence of the isometric speed-related parameters from the architecture of the muscle. The differences in the absolute V,,, data, however, revealed a positive relationship between the maximum speed of shortening and the fiber length. 7. These data demonstrate the necessity of the neuromotor command being matched with the architectural as well as the biochemical properties of each sarcomere (e.g., myosin ATPase and sarcoplasmic reticulum) of the muscles in the control of forces and velocities in vivo.

192

Copyright 0 1982 The American Physiological Society

0022-3077/82/0000-0000$01.25

CHARACTERISTICS

OF

INTRODUCTION

Although a large variety of architectural arrangements of muscle fibers are found in mammalian skeletal muscle ( 1 I, 14, 17), the relative significance of this characteristic in determining the physiological properties and functional mechanics of muscle is not well known. Both theoretical (11, 17) and experimental results (33), have emphasized the close relationship belween morphology and function in synergistic muscles as well as within individual muscles. Intrinsic factors, such as sarcoplasmic reticular properties (6) and myosin ATPase (2) can mediate the rate of tension developed in an isometric contraction. However, during lengthening and shortening contractions, muscle architecture assumes an equivalent if not greater importance than intrinsic fiber properties. From previous studies of force-velocity characteristics of muscle, it is apparent that the number of myofibrillar cross-links aligned in parallel determines the maximum force production, while the maximum velocity of shortening is influenced by the number of sarcomeres arranged in series (8). In the cat, the semitendinosus (ST) is a biarticular muscle with a parallel fiber arrangement. It is located posteriorly in the thigh and is active during knee flexion and hip extension (9, 32). The ST has an interesting anatomical adaptation, which complicates this otherwise “straightforward” parallel-fibered muscle. A band of dense fibruus connective tissue divides the muscle into distinct proximal and distal parts, arranged in series and with each part having a separate innervation (7, 3 1). This unusual architectural design provides an interesting model fur examining the functional significance of morphology on the contractile dynamics of skeletal muscle. The in series design of the ST combined with the dual innervation provides the potential for rather complex mechanical function (5). In addition to the compound architecture, complex electromyographic (EMG) activity has been reported for the ST. During unrestrained locomotion, Engberg and Lundberg (9) observed a burst of activity in both the swing and stance phase of the cat step cycle in the ST. It is unclear,

CAT

193

SEMITENDINOSUS

however, whether a single burst can be associated with one or both compartments of the muscle. The existence of an endogenous in series muscle arrangement provides an intriguing experimental preparation for examining the passive and dynamic mechanical properties of skeletal muscle. The purpose of this study was to detail the architectural and histochemical profiles of the cat semitendinosus and to examine its compound mechanical properties during isometric and isotonic contractions. No significant difference exists between the histochemical properties or the cross-sectional areas of the proximal and distal portions of the ST. The contractile experiments nevertheless emphasize the multiple combinations of tensions and velocities that are possible, although not necessarily utilized, as a consequence of the serial architecture and separate innervation of the ST. Preliminary results were published elsewhere (4). METHODS

Architectural

characteristics

Six adult cats weighing between 2.0 and 6.0 kg were used to determine the histochemical, architectural, and contractile properties of the ST. Previous work (33) showed a strong correlation between the right and left architectural profiles of mammalian hindlimb musculature. Thus, the paired hindlimbs were similar and the architectural characteristics of the right side could be directly related to the contractile properties of the corresponding left side. A detailed description of the procedures used to determine muscle architecture has been reported (30, 33). Briefly, the lower extremity was skinned and fixed in Formalin with the knee and hip joints positioned at approximately 90°. Goslow et al. (12) indicated that the average joint angles adapted in quiet standing for the hip (iliofemoral), knee, and ankle are 95, 100, and 1 loo, respectively, and therefore our method of fixation allowed muscle lengths to be measured at a functional standing length. Individual muscles were dissected, removed, and placed in a 15-20% sulfuric acid solution, and subsequently stored in 50% glycerol. By measuring the muscle length before and after fixation, the amount of shrinkage or elongation was found to be less than *5%. Small bindles of fibers ( lo- 15) were teased out from each portion of the muscle and measured

194

BODINE

with calipers. Samples were taken from several regions along the length of the muscle at varying depths. Average sarcomere lengths were determined from individual fibers under a light microscope (x400) using a calibrated eyepiece micrometer. Sarcomere lengths from different regions within a muscle were within -t5%. Measurements of muscle lengths and sarcomere lengths established the number of sarcomeres in series. Although shrinkage occurs with fixation, there will be proportional changes occurring in both whole muscle and sarcomere lengths. Any length change will not affect the absolute number of sarcomeres in series. Physiological cross-sectional areas were calculated using the following formula: cross sectional area = ((muscle mass) (cos@))/((fiber length) (density)). Muscle mass (g) was determined by weighing fresh muscles from cats of similar weights, 8 was equal to the approximate angle of pinnation, absolute fiber length (cm) was calculated from the architectural data, and density of the muscle was assumed to be 1.0564 g/ cm3 (22).

Histochemistry Fiber-type distributions were determined for the ST on three adult cats. On the average, 125 fibers were analyzed from serial sections of the superficial and deep portions of both the STp and STd. Fibers were classified as slow oxidative (SO), fast oxidative-glycolytic (FOG), and fast glycolytic (FG) using reduced nicotinamide adenine dinucleotide diaphorase, adenosine triphosphatase (acid and base preincubation), and a-glycerophosphate dehydrogenase activity as described by Peter et al. (26).

Contractile properties The in situ isometric and isotonic contractile properties of the proximal-activated (STp), distalactivated (STd), and whole muscle-activated (ST) semitendinosus muscle were determined from the contralateral hindlimbs of six adult cats.

Animal preparation Animals were anesthetized with sodium pentobarbital (35 mg/kg ip) and the hindlimb musculature was exposed with a posterior sagittal incision extending from the ischial tuberosity to just inferior to the knee joint. The ST was carefully isolated with the normal blood supply left intact. The ST was freed from surrounding musculature to allow for the full range of motion and to assure an unrestricted view during filming of some contractions. This film data will be used to assess movement within each compartment. The nerves innervating the ST originate from distinct branches of the sciatic nerve. As the sciatic nerve passes over the obturator internus muscle in the gluteal region, there is a large mus-

ET AL.

cular branch that supplies the ST, semimembranosus, and biceps femoris. The nerve branch going to the ST bifurcates and sends one branch to the proximal and the other to the distal compartment of the ST. These two branches were isolated, separated to the point of bifurcation, tied with suture, and cut. Surrounding muscles were denervated and fine-wire electrodes (50 pm in diameter) were placed in the STp and STd to monitor electromyographic (EMG) activity. The distal tendon of the ST was detached from its insertion site at the medial side of the tibia1 crest and tied with nylon ligature (2-O) to a steel wire attached directly to an isotonic pneumatic lever (22). The cat was cradled in a frame in a prone position and the hindlimb was rigidly secured with metal clamps. A mineral-oil bath was formed using the retracted skin of the hindlimb and maintained at a temperature of 36 ? l°C. The ST goes through rather large displacements when shortening and to insure the optimal line of pull for the muscle and the largest range of motion, the ST was positioned horizontally and suspended above, rather than submerged, in the oil pool. Isometric and isotonic contractions for each stimulation condition were filmed. To reduce glare, it was necessary to have the muscle suspended above the pool. Due to these experimental restraints, the temperature of the ST was maintained at 31 k 3OC.

Testing protocol Isometric and isotonic contractions were recorded on a modified pneumatic lever system (22) for each of three activation conditions: I) STd stimulated and STp passive, 2) STp stimulated and STd passive, and 3) both STd and STp stimulated simultaneously. The order for testing the three activation conditions was alternated to avoid any potential bias. The distal tendon of the ST was attached in situ to a magnesium lever on which force and displacement transducers were mounted. The muscle length (L,) at which maximum isometric twitch tension could be produced was determined for each of the three activation conditions. The length of the muscle at each L, was measured from film taken during an isometric twitch. All contractions were initiated at the L, appropriate for that activation condition. For isotonic contractions, the mechanical restraint was removed and the resistance to shortening was supplied with a variable air pressure system in series with the lever arm (33). All force and displacement data from the contractions were recorded on a polygraph (Grass Instruments 7B) and FM tape recorder (Hewlett-Packard 3968) and were monitored on a storage oscilloscope. Muscular contractions were elicited through stimulation of the isolated nerve branches with

CHARACTERISTICS

OF CAT

bipolar silver electrodes. Stimulating electrodes were placed on each of the branches to permit activation of each compartment independently or both simultaneously. The stimulus delivered was a 0.02-ms squarewave pulse with a strength approximately twice the voltage required to obtain a maximal isometric twitch response. The twitch response for each stimulation condition was obtained at the beginning of an experiment prior to testing for the frequency-tension and force-velocity properties. Isometric twitch data yielded measures of maximum twitch tension (P,), time to peak tension (TPT), and half relaxation time (Y2 RT). Peak isometric tension (P,) was determined from the tension response to stimulation at frequencies of 5, 10, 15, 20, 25, 30, 40, 50, 75, 100, and 200 Hz for 500 ms duration. Isotonic properties were obtained for 15-20 afterloaded contractions at varying loads less than P, and stimulated at a frequency of 200 Hz for 330 ms. A 2-min interval between each contraction was allowed, during which the muscle was stimulated at 1 Hz. P, values were obtained and compared at several points during each experimental condition. These values deviated less than 8%. Analog FM tape records were digitally sampled ( 1,000 Hz) with a minicomputer (HP 21 MX). Calibrated voltage fluctuations were converted to force, rate of tension development (dP/dr), and velocity (dL/dt) via methods previously described (28). To calculate a maximal velocity of shortening (V,,,) for each stimulation condition, the linear form of Hill’s equation ((P, - P)/v = m (P) + b) was used (15). For each isotonic contraction, peak velocity (V) and corresponding force (P) were plotted as ((P, - P)/V) versus P. A linear regression equation was fitted to these data with Hill’s constants a and b corresponding to the slope and the y intercept of the regression line, respectively. An averaged estimated forcevelocity curve was derived for each activation conSciatic

PST FIG. I. Schematic drawing ST into a proximal (STp) and and are connected in series at the proximal fibers. Each end

SEMITENDINOSUS

195

dition (STp, STd, and ST) by averaging Hill’s constants for all muscles tested for each activation condition. The values listed in the text are means k SE. RESULTS

Architecture Both the STp and STd have a parallel muscle fiber arrangement; the angle of pinnation was 0’ with respect to the long axis of the muscle-tendon unit (Fig. 1). The STp and STd, respectively, constitute approximately one-third and two-thirds of the total ST length and their muscle fibers are continuous from the respective tendinous insertions to the fibrous connective tissue septum. Specifically, the length of the fibers for the STd was 3.93 t 0.1 cm, and the length of the STp was 2.12 t 0.1 cm. The estimated wet weight of the STp was 5.14 t 0.13 g or approximately 40% of the ST wet weight, while the STd was 7.69 k 0.47 g. The wet weights, however, were difficult to determine precisely because of the uneven alignment of the surface of the septum. The physiological cross-sectional area of the STp (2.29 cm’) was not significantly different from the cross-sectional area of the STd ( 1.85 cm2). The average sarcomere length in both the STp and STd was 2.19 k 0.10 pm.

Histochemistry The muscle fiber type composition is similar in both the STp and STd (Table 1). The deep portions of the STp and STd are nearly uniform in composition, with the majority of the muscle fibers being fast twitch. In the superficial portion, the STd has 27% oxi-

n.

Con&t tissue

ive sept urn

DST

of the cat semitendinosus (ST) muscle. A dense connective tissue band divides the distal (STd) end. The muscle fibers of each end are arranged in parallel(0 = O”) the connective tissue band, with distal fibers being approximately twice as long as is innervated by a separate branch from the sciatic nerve.

196

ET AL.

BODINE

TABLE 1. Cat semitendinosus fiber type composition

FG

muscle

FOG

so

Proximal Deep Superficial

48 zk 1 80 * 3

29 * 2 18 k 3

22 k 2 l&l

Distal Deep Superficial

46 k 3 73 -t- 3

29 -t 2 24 k 2

24 t 2 3-+2

Values are means k SE, expressed as percentages. II = 3 cats, with an average of 125 fibers per area. Deep, one-fourth the area of the muscle that was closest to the femur; superficial, one-fourth the area of the muscle that was farthest from the femur,

dative fibers (FOG + SO), while the STp has somewhat fewer oxidative fibers (19%) There was no difference in the SO populations of the superficial or the deep portions of either compartment. Contractile properties To establish the potential for independent activation, electromyographic (EMG) signals were studied from both the STp and STd in response to stimulation to each of the separate nerve branches. Only the stimulated compartment showed an EMG signal suggesting that each compartment could be electrically isolated by the connective tissue septum that separates proximal from distal muscle fibers. Consequently, since differential activation was possible, the separate and compound contractile properties of the ST could be determined. FORCE RELATED. Isometric and isotonic contractile properties were obtained from the three stimulation conditions: STp, STd, and ST (Table 2). All contractions were elicTABLE

2.

Contractile properties of cat semidendinosus

TPT

Y2RT h P 20 PO V max V max

ited at the L, that was appropriate for each stimulation condition. The length of the whole muscle at L, was 142 t 2, 138 t 2, and 134 t 3 mm for STp, STd, and ST, respectively. The length-tension relationship of all three conditions revealed that active tension was remarkably constant over a wide range of lengths. A relatively flat length-tension curve would be expected since the ST is a parallel-fibered muscle (11, 12, 27). According to Goslow et al, (12), during locomotion, the ST is generally used at lengths shorter than L, (the length associated with quiet standing) and goes through a range of approximately t 10 mm from L,. Within this physiological range, the ST is able to actively produce greater than 90% of its maximum isometric twitch tension. The maximum twitch tension was greatest when STp and STd were stimulated simultaneously and was least when only the STp was stimulated. In contrast, the maximum tetanic tension (P,) was identical across all three conditions. The frequency-tension response revealed a difference in all three stimulation conditions at frequencies lower than 20 Hz; however, at frequencies between 20 and 75 Hz, STp produced lesstension than both ST and STd (Fig. 2). The force produced at 20 Hz (PzO) was 40 and 37% of P, for the ST and STd, respectively, but only 24% of P, for STp (Table 2). SPEED RELATED. The TPT and ‘/2 RT were similar for all three stimulation conditions (Table 2). However, when the maximal velocity of shortening (V,,,) was expressed in absolute terms (millimeters per second), there was an orderly V,,, relationship across the three stimulation conditions based on the differences in fiber lengths of the STd and

mm

l

Units

Distal

ms ms

47 37 3.53 37 20.6 424 23.6

k P, n mm/s s-I/ 1,000 sarcomeres

Values are means k SE. n = 6 cats. Significant * STd vs, STp; t STd vs. ST; $ STp vs. ST.

(STd) k k Ik IL * k t

differences

Proximal

3 5 0.46? 3* 2 26” 1.4 (P < 0.05)

41 31 2.74 24 18.6 224 23.2

(STp)

t t5 4 tk t t

Both

3

41 37 5.29 40 19.6 624 22.6

0.48$ 4 3 23$ 2.3

for the following

paired

(ST) * 2 z?z 5 & 0.80 t 65 t 2 do 307 k

1.1

comparisons:

CHARACTERISTICS

OF CAT

SEMITENDINOSUS 600

t

h

500

0 0

25

50 FREQ"ENCY7~"z

FIG, 2, force for activated STp-STd expressed

100

-'

200

)

Frequency of stimulation is plotted versus each of three stimulation conditions: proximal (STp) (A), distal activated (STd) (w), and activated simultaneously (ST) (a). Force is as a percent of maximum isometric tension.

STp. Stimulation of the whole muscle (ST) produced the highest absolute V,,,. When the Lx of the STd (424 t 26 mm/s) and the STp (224 & 23 mm/s) were added, the resulting V,,, was similar to that observed for the ST (624 t 30 mm/s). The intrinsic maximum velocities of shortening (millimeters per second per 1,000 sarcomeres) for the three stimulation conditions were identical (Table 2), reflecting the similarities in fiber type composition. The characteristic force-velocity relationships for the three stimulation conditions are illustrated as a function of absolute velocity (millimeters per second) versus force (% P,) in Fig. 3, and relative velocity (millimeters per second per 1,000 sarcomeres) versus force (% P,) in Fig. 4.

197

40 FORCE

60 (%Po )

FIG. 3. Force-velocity relationships are plotted for the three stimulation conditions: STp (A), STd (w), and ST (a). Velocities are expressed in absolute units of millimeters per second and forces are expressed as percent of maximum isometric tension. These curves ilIustrate the direct relationship between maximum speed of shortening and fiber length.

is not unique to the cat. It is also found in dogs (20), monkeys ( 1), and humans (3,20). It appears that the in series arrangement of the ST is the selected mode of linkage phylogenetically. In other mammals, this muscle is even more complex structurally. The ST in rats (1, 29), mice (18), rabbits ( I), and guinea pigs ( 1) has two proximal heads of origin with a tendinous inscription at the more distal junction of the two heads. The dorsal head has its origin from the caudal

DISCUSSION

Architectural considerations The semitendinosus has an unusual architectural design for cat hindlimb musculature. A band of dense fibrous connective tissue divides the muscle into two distinct compartments innervated by separate nerve branches from the sciatic nerve, Glycogendepletion studies (10) have shown that there are two distinct populations of muscle fibers proximally and distally. However, there is extensive topographical overlap among the motoneurons supplying each end of the muscle (5, 19). The in series fiber arrangement of the ST

FIG. 4. Force-velocity relationships STd (w),. and ST (a). The relative pressed as millimeters per second per and forces are expressed as percent metric tension. The intrinsic velocities not significantly different for the three ditions reflecting the similarity in fiber of the proximal and distal ends.

of the STp (A), velocities are ex1,000 sarcomeres of maximum isoof shortening are stimulation contype composition

198

BODINE

vertebrae, whereas the ventral head attaches to the ischial tuberosity. According to Appleton (l), the tendinous inscription appears to be a new structure found in mammals. It appears that two heads of origin is an ancestral condition, whereas the absence of the dorsal head or an ischial head as found in some rodents is a developed state (18). Although the in series architecture of the ST is unique in the hindlimb, there are other muscles with a similar design. The human rectus abdominus ( 13), human digastric ( 13), and rat plantaris-flexor digitorum brevis complex (29) are some examples that demonstrate sets of fibers attached in series. However, each of these examples have an in series arrangement that differs with respect to the force vectors and/or probably the stiffness of the connecting elements. Mechanical

properties

The maximum isometric tension (P,) that can be produced by a muscle is proportional to the number of contractile elements or myofibrillar cross-links in parallel ( 14). The fact that the PO’s of the STp, STd, and ST were identical further demonstrates the significance of the architectural design, i.e., the importance of the physiological cross-sectional area in determining the potential P, of a muscle. When the STp and STd are stimulated simultaneously, the P, should be proportional to the smallest total cross-sectional area of all the muscle fibers along the proximodistal axis. Since the physiological cross-sectional areas of both compartments were similar, the number of myofibrillar cross-links in parallel would be expected to be similar in each end, and consequently the P, of all three conditions would also be expected to be similar. The maximum values of twitch tension, in contrast to the PO values, were significantly different. Stimulation of the STp or STd produced P, values which, when summed, approximated the P, value for simultaneous stimulation of the ST. The stiffness-compliance properties of the muscleconnective tissue unit undoubtedly influenced the values measured for P, conditions. When only one compartment was stimulated, the compliance occurred primarily in the passive compartment of the muscle. As

ET AL.

demonstrated by Hill (16), the maximum twitch tension decreased as the series compliance increased. Because of the differences in lengths (STp/ STd = 2.12/3.93 cm) there will be a substantial difference in the stiffness when one is stimulated independently of the other. The combined stiffness of the STd and STp would be expected to be the same as if the semitendinosus were a single muscle with muscle fibers equal to 6.05 cm (assuming similar material composition and cross-sectional area). It would seem that the mechanical significance of the in series fiber arrangement in the ST would be important only if each compartment were activated independently in situ. The frequency-tension relationships (Fig. 2) indicated that the compliance effects of a passive coupled with an active contractile compartment were most evident at frequencies at 20 Hz or less. The STp had the lowest PzO (Table 2), the STd was next in order of increased tension output, and the ST was highest in tension. At frequencies between 20 and 75 Hz, only STp differed in its tension response. The order effect was consistent with the compliance characteristics of the ST components, but a more detailed intracontraction assessment of the force-deformation characteristics of the STd, STp, and full muscle is needed before the definitive mechanics of the in series arrangement can be elucidated. The absolute velocity of shortening is determined by the biochemical properties (2) and the length of the fibers (17). Hence, the V,,, (millimeters per second) was lowest for the STp at a value approximately one-third the speed of the ST, with the STd at approximately two-thirds the ST V,,,. These differences were simply a function of the differences in fiber lengths, since with a normalization of V,,, values as a function of the number of sarcomeres in series (millimeters per second per 1,000 sarcomeres) there was no difference between the three conditions. The normalized V,,, values were consistent with the histochemical findings that showed each compartment to have identical “slow-fast” fiber proportions. The data from these experiments emphasize the importance of architectural design in determining functional force and velocity

CHARACTERl:STICS

OF

characteristics of skeletal muscle. The relationship between fiber length and V,,, indicates that the longer the fiber, the faster the absolute speed of shortening, i.e., speed is a function of the number of sarcomeres in series. For this and other parallel fibers in general, this arrangement allows the maximum possible displacement with the least loss of force during the contraction (11, 27). The parallel arrangement may permit the muscle to operate at lengths close to L, with only small fractions of sarcomere length changes to maintain force production at effective levels. The P, values from the three conditions of stimulation reinforce the known relationship between cross-sectional area and force production in muscle. For an in series arrangement of two muscles, or parts of the same muscle, force is not additive but is dependent on cross-sectional area and will be a reflection of the weakest link in the series. Unlike the isotonic contractions, these data further illustrate that isometric speed-related parameters (TPT and ‘/2 RT) are independent of muscle architecture (30). Implications for motor control The ST is a biarticular muscle attaching proximally on the ischial tuberosity and disl tally on the medial side of the tibia1 crest. Because of its attachments, the ST functions as both a hip extensor and knee flexor. Studies (9,24,25,34) have shown the ST to have a double burst of EMG activity during each locomotor step cycle. During the step cycle, the initial burst of activity occurred during the E3 and F phases (28), while the second burst occurred after a brief pause in the El and E2 phases. Perret and Cabelguen (24, 25) have also shown a double burst in the decorticate preparation in cats. However, “burst 2” could be completely eliminated during free locomotion after complete deafferentation of the ST. Nevertheless, if the limb was fixed to counteract flexion or given constant tactile stimulation, burst 2 occurred alone during the locomotor cycle. This suggests that some constant inflow of afferent information from the limb is needed to bring about the second burst of activity in the ST. In fictive locomotion (34) the ST can also display two bursts of activity in each step cycle, one during the flexor activity and one

CAT

SEMITENDINOSUS

199

during the extensor activity. Various discharge patterns can be seen with different “excitation levels” and afferent input (34). During locomotion, the length changes in the ST are markedly influenced by the relative movements of the hip and knee joints (12). The ST, according to Goslow et al. (12) is generally used in stepping at muscle lengths shorter than a length (L,) that refers to a hindlimb configuration associated with a quiet standing posture: hip (iliofemoral), knee, and ankle angles at 95, 100, and 1 lo”, respectively. Time of the “active” (EMG activity present) phases of the ST during locomotion was determined (12) with a comparison of kinematic data with previously reported EMG data (9). In the F phase (28) of the step cycle, Goslow et al. ( 12) reported that in walking (0.67 m/s) the ST undergoes active shortening then lengthening, while in the F phase of the trot (1.61 m/s) and gallop (7.29 m/s), the ST only undergoes active shortening contraction. From the data presented on the average contraction velocities of the ST during locomotion (Table 5 in Ref. 12), the highest average shortening velocity (154 mm/s) of the ST occurred during the F phase of the trot. Because the velocity of 154 mm/s was an average over the complete F phase, it is likely that instantaneous active velocity may be higher during portions of the F phase (e.g., see their Fig. 12). It is appropriate to consider this in vivo average velocity in light of the range of differences in the in situ force-velocity characteristics of the three stimulation conditions: STp, STd, and ST. For example, if only the STp was recruited with a velocity of shortening of 154 mm/s, the maximum force potential would be about 14% of P, (Fig. 3). If, however, the STd was recruited at the same speed of shortening, the potential force would now be approximately 29% of PO. In comparison, if the ST was active during a shortening velocity of 154 mm/s, the potential force production would be 4 1% of P,. From Fig. 3, it is evident that at any given velocity of shortening at the ST tendon, the amount of force that can be produced is greatest when the ST is activated. Our results to date (2 1) have shown that a double burst of activity occurs in the STp and STd and that they can occur simultaneously during treadmill locomotion. This might suggest that al-

200

BODINE

though anatomically the STp and STd could function independently, they in fact are closely coordinated during locomotion. Interestingly, both have the same number of muscle spindle capsules (7). Nelson and Mendell (23) have also indicated that single Ia-fibers from the STp and STd project similarly to motoneurons of both compartments. These results are consistent with the hypothesis that the spinal cord is organized to treat the ST as a single muscle and that the afferent systems of both have similar properties. Even though the in situ preparation demonstrates the possibility of independently activating the STp and STd, these data show that the effectiveness of the system to produce tension and velocity may be compromised if one is activated independently of the other, particularly at frequencies of 20 Hz or less. However, the existence of the

ET AL.

separate innervations coupled with the compound muscle arrangement in the cat ST provides the central nervous system with a significantly greater number of potential combinations of forces and velocities than if it were a single muscle structure. Whether all of these combinations are used in normal movement remain to be determined, but regardless, the means of neurally controlling these “added” variables seem to exist. ACKNOWLEDGMENTS We thank Doug Stuart for bringing to our attention the unusual arrangement of the semitendinosus muscle and for his critical review of the manuscript. We also thank Tom Hamm and Ziaul Hasan for their helpful criticism of the manuscript. This study was supported by National Institutes of Health Grant N516333. Received 17 February

8 September 1982.

198 1; accepted

in final

form

REFERENCES 1. APPLETON, A, B. The muscles and nerves of the post-axial region of the tetrapod thigh, J. Anat. 62: 364-438, 1928. 2. BARANY, M. ATPase activity of myosin correlated with speed of muscle shortening, J. Gen. PhysioE. 50: 197-218, 1967. 3. BARDEEN, C. R. Development and variation of the nerves and the musculature of the inferior extremity and of neighboring regions of the trunk in man. Am. J. Anat. 6: 259-390, 1906. 4. BoDINE, S. C., MEADOWS, D. A., ROY, R. R., SACKS, R,, AND EDGERTON, V. R. Contractile properties of the cat semitendinosus muscle. Med. Sci. Sports Exercise I 3: 127, 198 1. 5. BOTTERMAN, B. R,, HAMM, J. M., REINKING, R. M., AND STUART, D. G. Intramuscular reflex localization in relation to muscle design and action. Sot. Neurosci. Abstr. 7: 557, 198 1 a 6. BRODY, J. A. Regulation of isometric contraction in skeletal muscle. Exp. Neural. 50: 673-683, 1976, 7. CHIN, N. K., COPE, M., AND PANG, M. Number and distribution of spindle capsules in seven hindlimb muscles of the cat. In: Symposium on Muscle Receptors, edited by D. Barker. Hong Kong: Hong Kong University Press, 1962, p. 241-248, 8, CLOSE, R. 1. Dynamic properties of mammalian skeletal muscles. Physiol. Rev. 52: 129-197, 1972. 9. ENGBERG, I. AND LUNDBERG, A. An electromyographic analysis of muscular activity in the hindlimb of the cat during unrestrained locomotion. Acta Physiol. &and. 75: 614-630, 1969. 10. ENGLISH, A, W. AND LETBETTER, W. D. Intramuscular “compartmentalization” of the cat biceps femoris and semitendinosus muscles: anatcimy and EMG patterns. Sot. Neurosci. Abstr. 7: 557, 19#S1. I 1. GANS. signif~ -I C. AND BOCK. r W. J. The functional V

12.

13. 14.

15.

16.

17,

18.

19.

20,

21.

icance of muscle architecture: a theoretical analysis. Ergeb. Anat. Entwicklungsgesch, 38: 1 f. 5- 142, 1965. GOSLOW, G. E., REINKING, R. M., AND STUART, D. S. The cat step cycle: hindlimb joint angles and muscle lengths during unrestrained locomotion. J. Morphol. 141: l-42, 1973. GRAY, H. Gray’s Anatomy, edited by T. P. Pick and R. Howden. New York: Bounty, 1977. HENNEMAN, E. AND OLSON, C. B. Relations between structure and function in the design of skeletal muscle. J. Neurophysiol. 28: 85-99, 1965. HILL, A. V. The heat of shortening and the dynamic constants of muscle. Proc. R. Sot. London Ser. B 126: 139495, 1938. HILL, A. V. The effect of series compliance on the tension developed in a muscle twitch. Proc. R. Sot. London 138: 325-329, 1951. JOSEPHSON, R, K. Extensive and intensive factors determining the performance of striated muscle. J. Exp. 2001. 194: 134-154, 1975. LANCE-JONES, C. The morphogenesis of the thigh of the mouse with special reference to tetrapod muscle homologies, J. Morphol. 162: 275-3 10, 1979. LETBETTER, W.D. ANDENGLISH, A. W.The relationship between peripheral intramuscular %ompartments” and spatial arrangement of biceps femoris and semitendinosus motor nuclei in the cat lumbar spinal cord. Sot. Neurosci. Abstr. 7: 557, 1981. MARKEE, J. E,, L~GUE, J. T,, WILLIAMS, M., STANTON, W. B., WRENN, R. N., AND WALKER, L. B. Two-joint muscles of the thigh. J. Bone Jt. Surg. Am. Vol. 37: 125-142, 1955. MURPHY, K,, ROY, R. R., ANDBODINE,S.C. Re-

CHARACTERISTICS

22.

23.

24.

25.

26.

27.

28.

OF CAT

cruitment of the proximal and distal portions of the cat semitendinosus during running and jumping. Med. Sci. Sports Exercise 13: 127-128, 198 1. MURPHY, R. A. AND BEARDSLEY, A. C. Mechanical properties of the cat soleus muscle in situ. Am. J. Physiol. 227: 1008-1013, 1974. NELSON, S, C. AND MENDELL, L, M. Projection of single knee flexor Ia fibers to homonymous and heteronymous motoneurons. J. Newophysiol. 4 1: 77&-787, 1978, PERRET, C. AND CABELGUEN, J. M. Central and reflex participation in the timing of locomotor activations of a bifunctional muscle, the semitendinosus, in the cat. Brain Res. 106: 390-395, 1976. PERRET, C. AND CABELGUEN, J. M. Main characteristics of the hindlimb locomotor cycle in the decorticate cat with special reference to bifunctional muscles, Brain Res. 187: 333-352, 1980. PETER, J. B., BARNARD, R, J., EDGERTON, V. R,, GILLESPIE, C. A,, AND STEMPEL, K. E. Metabolic profiles of three fiber types of skeletal muscle in guinea pigs and rabbits. Biochemistry 11: 26272633, 1972. PETERS, S, AND RICK, C. The actions of the three hamstring muscles of the cat: a mechanical analysis. J. Morphol. 152: 315-328, 1977. PHILIPPSON, M. L’autonomie et la centralization dans le nerveux des animaux. Trav. Lab. Physiol. Inst. Solvay 7: l-208, 1905,

SEMITENDINOSUS

201

29. RINKER, G. C. The comparative myology of the mammalian genera Sigmodon, Oryzomys, Neotoma, and Peromyscus (Cricetinae), with remarks on their intergeneric relationships. Misc. P&l. MUX 2001. Univ. Mich. 83: 1-124, 1954. R. R,, MEADOWS, I. D., BALDWIN, 30. ROY, K. M., AND EDGERTON, V. R. Functional significance of compensatory overloaded rat fast muscle. J, Appl. Physiol.: Respirat. Environ. Exercise Physiol. 52: 473-478, 1982. 31. SACKS, R. D. AND ROY, R. R. Architecture of cat hindlimb muscle: implications for functional significance. Sot, Neurosci. Abstr. 7: 685, 198 1. 32. SHERRINGTON, C. S. Flexion-reflex of the limb, crossed extension-reflex, and reflex stepping and standing. J. Physiol. London 40: 28-121, 19 10. 33. SPECTOR, S. A., GARDINER, P. F., ZERNICKE, R. F., ROY, R. R., AND EDGERTON, V. R. Muscle architecture and force-velocity characteristics of cat soleus and medial gastrocnemius: implications for motor control. J. Neurophysiol. 44: 95 l-960, 1980. 34. VIDAL, C., VIALA, D., AND BUSER, P. Central locomotor programming in the rabbit. Brain Res. 168: 57-73, 1979. B, AND PROSKE, U. Comparison of 35. WALMSLEY, stiffness of soleus and medial gastrocnemius muscles in cats. J. Neurophysiol. 46: 250-259, 1981.