MS 2799, pp. 689-708

Journal of Physiology (1994), 481.3

689

Storage and release of mechanical energy by contracting frog muscle fibres G. A. Cavagna, N. C. Heglund, J. D. Harry and M. Mantovani Istituto di Fisiologia Umana, Universita di Milano, 20133 Milan, Italy 1.

2.

3.

4.

5.

6. 7.

8. 9.

Stretching a contracting muscle leads to greater mechanical work being done during subsequent shortening by its contractile component; the mechanism of this enhancement is not known. This mechanism has been investigated here by subjecting tetanized frog muscle fibres to ramp stretches followed by an isotonic release against a load equal to the maximum isometric tension, To. Shortening against To was taken as direct evidence of an absolute increase in the ability to do work as a consequence of the previous stretch. Ramp stretches (0-5-8-6% sarcomere strain, confined to the plateau of the isometric tension-length relationship) were given at different velocities of lengthening (0f03-1f8 sarcomere lengths s-1). Isotonic release to T. took place immediately after the end of the ramp, or 5-800 ms after the end of the largest ramp stretches. The length changes taking place after release were measured both at the fibre end and on a tendon-free segment of the fibre. The experiments were carried out at 4 and 14 'C. After the elastic recoil of the undamped elastic elements, taking place during the fall in tension at the instant of the isotonic release, a well-defined shortening took place against T. (transient shortening against T.). The amplitude and time course of transient shortening against To were similar at the fibre end and in the segment, indicating that it is due to a property of the sarcomeres and not due to stress relaxation of the tendons. Transient shortening against To increased with sarcomere stretch amplitude up to about 8 nm per half-sarcomere independent of stretch velocity. When a short delay (5-20 ms) was introduced between the end of the stretch and the isotonic release, the transient shortening against To did not change; after longer time delays, the transient shortening against T. decreased in amplitude. The velocity of transient shortening against To increased with temperature with a temperature coefficient, Qlo, of - 2-5. It is suggested that transient shortening against To results from the release of mechanical energy stored within the damped element of the cross-bridges. The crossbridges are brought into a state of greater potential energy not only during the ramp stretch, but also immediately afterwards, during the fast phase of stress relaxation.

It is well known that an active muscle resists stretching with a force greater than that exerted during an isometric contraction (Katz, 1939). The greater force developed during stretching involves a greater storage of mechanical energy in the undamped elastic elements of muscle, and a greater elastic work is done when the force falls during a subsequent shortening. The force enhancement above the isometric value does not imply per se a storage of mechanical energy within the contractile component of muscle, and

does not explain the greater amount of work done by a previously stretched muscle when it shortens against a constant force (Cavagna & Citterio, 1974; Cavagna, Citterio & Jacini, 1975; Edman, Elzinga & Noble, 1978; Sugi & Tsuchiya, 1981; Cavagna, Mazzanti, Heglund & Citterio, 1986). In fact, if the force is held constant during shortening, the greater amount of work cannot be derived in a simple way from the release of elastic energy stored in the stretched undamped elastic elements. Possible origins

690

G. A. Cavagna and others

of this enhanced work capacity are (1) the recoil of a viscoelastic system charged during stretching; (2) a greater number of attached cross-bridges due, for example, to a more favourable overlap between filaments induced by the previous stretching; or (3) a modification of the state of the individual cross-bridges. Data reported in the literature are contradictory and do not allow these possibilities to be distinguished. A greater energy output after stretching has been reported for sarcomere lengths greater than 2-3 ,um, i.e. on the descending limb of the tension-length relationship, but it was not found on the plateau of the tension-length relationship (Edman et al. 1978; Edman, Elzinga & Noble, 1981). However, a transient shortening against the maximum isometric tension (To) resulted from ramp stretches confined to the plateau of the tension-length relationship in both the whole muscle and single fibres (Cavagna et al. 1975, 1986). On the other hand, no sign of a transient character to the increased load-sustaining ability was found after quick decreases in load applied during isotonic lengthening (Sugi & Tsuchiya, 1981). In this paper we investigate the effect of previous stretching on the ability of tetanically stimulated frog muscle fibres to do positive work during subsequent shortening against a tension equal to the maximum isometric tension, To. Shortening against such a maximum tension was taken as expression of an absolute energy gain due to the previous stretching (not a relative energy gain) as that described on the descending limb of the tension-length relationship; Edman et al. 1978). In order to determine if this energy is released directly by the contractile component, the length changes of the sarcomeres during shortening were measured in a tendonfree segment of the fibre by means of a striation follower (Huxley, Lombardi & Peachey, 1981). In addition, shortening was simultaneously measured at the fibre end in order to determine the contribution of stress relaxation of tendons and other end compliance. Fibre and sarcomere shortening against To were measured after ramp stretches of different velocity and amplitude; all motion was confined to the plateau of the tension-length relationship. In some experiments, measurements were made at two different temperatures (4 and 14 °C), and a variable duration isometric pause was left between the end of the stretch and the release to To. The results suggest that previous stretching increases the ability of the individual cross-bridges to do work; they seem to reach, and retain for an appreciable time interval after stretching, a level of potential energy unattainable during an isometric contraction.

METHODS Several components of the apparatus used (fibre chamber, stimulator, force transducer and motor system) and the procedure followed to dissect, mount and stimulate the fibre have been described in detail by Cavagna (1993).

J. Physiol. 481.3

Muscle fibres Frogs (Irish Rana temporaria) were killed by decapitation followed by destruction of the spinal cord. Fibres of the caput laterale of the tibialis anterior muscle were used. (Maximum isometric force = 4-64 + 1-54 mN, developed at a length of 5.93 + 045 mm, with a contracting sarcomeres length of 2 14 + 0 04 /sm and a cross-section of 22796 + 5530 Zsm2; means + S.D. of all the fibres studied at 4 °C, n = 37.) Fibres were used only if the isometric tension measured at the optimum fibre length was greater than 100 kN m-2; the decrease in isometric tension during an experiment averaged

5.3%. Experiments lasted up to 16 h and were stopped when (1) all of the desired information was obtained (23% of experiments); (2) the striation follower signal deteriorated due to interfering optical impurities, blurring of sarcomeres or lateral movement of the fibre (50%); (3) the fibre no longer responded to stimulation (23%); or (4) force feedback failed (4%). All data obtained were used unless (1) fibre inhomogeneity (see below) was greater than 30%; (2) an aberrant shortening against To was present after the elastic recoil (see the section Abnormal tracings, and Fig. 4); (4) large vibrations of the motor were initiated by the isotonic release; (5) the average force applied during transient shortening against To was smaller than 1 000 T. or larger than 1-045 T. (except for Fig. 13); (5) discontinuities (jumps) in the striation follower tracing led to more than a 10% error in the measured length change; or (6) the end of transient shortening against To was undefined. These requirements reduced the number of data included in Figs 5, 7, 9 and 10 (157 tetani recorded on 44 fibres, 33 at 4 °C, 7 at 14 °C and 4 at both temperatures) to about onethird of the tetani initially recorded.

Fibre and sarcomere force and length change measurements The fibre was suspended in Ringer solution between a force transducer (Huxley & Lombardi, 1980) and a motor capable of ramp displacements up to + 0-2 mm and step displacements complete in about 100 i,s. The motor position was used to monitor length changes at the fibre end. Simultaneously with measuring the length changes of the whole fibre, the average length change of sarcomeres in a tendon-free segment of the fibre was measured using a striation follower similar to that described by Huxley et al. (1981). The output of this apparatus (A V) is the difference between two voltages which are proportional to the number of sarcomeres moving past two laser spots delimiting a segment of the fibre. In each laser spot the image of ten contiguous sarcomeres is optically averaged and the resulting single-cycle sinusoidal light intensity pattern is exactly imposed upon five contiguous photodiodes. Because the image of the averaged sarcomeres looses registration with the five photodiodes as the sarcomeres change length, the maximum strain which can be measured is limited to approximately 45 nm per halfsarcomere (Fig. 2). The average sarcomeres strain within the segment was calculated from the striation follower output as follows. The initial number of sarcomeres between the laser spots is N1 = tseg/li where 1seg is the segment length and li is the initial sarcomere length. Since the signal corresponding to the movement of one sarcomere past a laser spot is 0-03922 volts (10 V/255 steps, lengthening results in a positive A V), the final

J. Physiol. 481.3

Stretch effect on work output by muscle

number of sarcomeres between the spots, Nf, is given by Nf = Ni - (A V/0 03922). The percentage sarcomere strain is:

OO(final length initial length) -

initial length or

loci {(

segINf) (seg/Ni))

691

preferred so that fibre motion under the laser spots was minimized. Twitches and short tetani were given to the fibre during the above operations; the exact stimulus threshold was measured at the experimental temperature. Sarcomere length changes during passive fibre strain and isometric tetanic contraction at If,O were recorded (Cavagna, Heglund, Harry & Passerini, 1990).

-

Experimental procedure

( lseg/Ni)

or

Ni-)Nf Nf

1oo(

Hence:

(IOOA

(0-03922 Iseg

-

A V4)

(1)

Sarcomere strain is not proportional to the striation follower output due to the term A VI, in the denominator of eqn (1); this, however, has a negligible effect, being at a maximum at about 4% of 0-03922 lseg. The maximum sarcomere count capacity

127 for stretching and 128 for shortening. A microcomputer controlled the timing (01 ms resolution) and most other aspects of the experiment, including the oscilloscope memory and data acquisition, triggering the stimulator, the striation follower and the channels of the oscilloscopes, determining and recording the value of the isometric force plateau, controlling the appropriate length or force level of the ergometer, reading the data from the memory of the oscilloscopes and doing the required measurements and calculations on the data.

was

-

Pre-experimental procedure Each experiment was preceded by a fixed succession of operations during which the fibre data required for the experiment were collected and the striation follower was calibrated. The motor coil was first moved 0-2 mm from its neutral position to the position reached at the end of the ramp stretch; the fibre length corresponding to this position, lf o, was the same for all stretch amplitudes. (This procedure was not followed for the first six of the forty-four fibres studied, with the consequence that for two of them the length reached at the end of the largest stretches was 0-125 and 0-2 mm greater than that reached at the end of the smallest stretches.) The relaxed sarcomere length, 18sr' at the fibre length lf,o was set at 2'19 ,sm (2'09-2-24 in two of the first six fibres; no appreciable difference found in the results) by adjusting the position of the whole motor. Sarcomere length was determined by averaging counts of twenty or more contiguous sarcomeres at different points along the fibre using a microscope with a x32 objective and a x25 micrometre eyepiece. Sarcomeres near the tendons were usually slightly shorter than 2-19 ,um. The cross-sectional area of the fibre was measured as described by Ford, Huxley & Simmons (1977). The relaxed fibre length (excluding tendons and attachments), If oX was then measured by means of a micrometer eyepiece fitted to a stereomicroscope (x 16). A suitable segment of fibre (800-1280 ,um long) was found by moving the two laser spots along the fibre and by changing their relative position. In general, the segment choice was based upon the absence of connective tissue or other impurities near the laser spots; the half of the fibre towards the stationary force transducer was

Two 4-channel digital oscilloscopes (Nicolet 4094, Madison, WI, USA), labelled END and SEG, were used to record the force and length changes of the fibre and sarcomeres, respectively. Sample experimental tracings are shown in Figs 2 and 3. The records taken by each oscilloscope were stored on disk (Nicolet XF-44). The relaxed fibre was first shortened from 1f,o by the amount of the planned subsequent stretching (Alstr). At the end of shortening, the fibre was given a few stimuli to take up its slack. The slow trace of the END scope was triggered and the fibre was stimulated to tetanus isometrically. When a force plateau was detected by the computer, a ramp of a preset amplitude and duration was sent to the ergometer. If the entire stretch length was within the range of the striation follower, the slow trace of the SEG oscilloscope and the striation follower were triggered 2 ms before the beginning of the ramp; otherwise, for large stretches, the slow SEG trace and the striation follower were triggered 3 ms before the end of the ramp. In either case, the fast traces of both oscilloscopes were triggered 3 ms before the end of the ramp. At the end of the ramp, or after a preset time interval during which the fibre was held active at the stretched length, a signal proportional to the isometric tension, measured just before the start of the ramp (To), was sent to the ergometer, and the ergometer was switched to force feedback mode. After a preset time interval, stimulation was stopped and the fibre was brought back to its initial length under position control. At the beginning of each experiment the fibre was released, without previous stretching, from a state of isometric contraction to a force about 0 9 times the isometric value; release was made from the final length reached by the fibre

during stretching, If,.. Motor vibrations (%-442 kHz lasting about 2 ms), initiated by the isotonic release, disturbed both the force and length records (Figs 2 and 3); in addition, the force trace was affected after release by an upward drift (-2 5% of T.). As shown in Figs 2 and 3, these artifacts do not prevent determination of the transition from the end of the elastic recoil (lasting about 100 ,as) to the beginning of transient shortening against To (lasting on average about 60 ms at 4 °). Measurements were made directly by the microcomputer from the oscilloscope tracings; points used in the calculations were selected by eye using cursor controls. The beginning and the end of transient shortening against T. were determined respectively on the fast length tracing (at the end of the elastic recoil) and on the slow length tracing; any offset between the two oscilloscope traces (fast and slow) was taken into account.

Measurement of sarcomere strain with the striation follower Sarcomere strain during the shorter ramp stretches (normally less than 30 nm per half-sarcomere, exceptionally 45 nm per half-sarcomere; Figs 2, 6 and 11) and during the transient shortening against To was measured from the striation

692

J. Physiol. 481.3

G. A. Cavagna and others

follower output, the segment length and the initial sarcomere length within the segment (eqn (1)). The initial length of sarcomeres contracting against T. at the beginning of the transient shortening against T. has been taken as: 182 =

sO(f

Ale8f)

(2)

1f,o

where 18O is the length of the contracting sarcomeres developing the tension To at the maximal fibre length If O; Is2 is the length of the contracting sarcomeres developing tension To at the fibre length 1f,o - Alef after the fibre elastic recoil, Ale,f. The initial length of the sarcomeres within the segment at start of the ramp, 18br, was calculated from 182 using the striation follower outputs during the elastic recoil and the ramp stretch, respectively. Since these outputs are proportional to the fractional number of sarcomeres moving in and out the segment, the calculation is easily done.

Estimation of sarcomere strain during large ramp stretches When the ramp stretch was too large to be entirely recorded by the striation follower, strain of the sarcomeres within the segment during the entire ramp stretch had to be calculated assuming a uniform distribution of fibre length changes among all sarcomeres. The length change imposed by the motor on the fibre, Al tr, may be divided into the length change of the sarcomeres, Al8, and the tendons, Alt: A = A/str AlI. (3) As a contracting fibre is stretched, the force initially rises abruptly, but after a length change of approximately 2% of the initial sarcomere length, the force settles to a roughly steady value. It follows that Al. is smaller than Alstr during the tension rise, whereas it approaches A l8tr at the end of large stretches. The effect of the tendon length change on the measurement of sarcomere length during stretching was estimated as follows. The average length of sarcomeres contracting isometrically within the segment at the beginning of the ramp, 18 br, was calculated as: |s

br = ls, ot

Atr

(4)

assuming (1) the isometric tension To, and therefore A lt, are the same at the maximum fibre length, 1f,OX and at the

beginning of the ramp, i.e. at the length If 0 - Alstr (this is true if Alstr occurs within the plateau of the tension length diagram; see below); and (2) the length of the sarcomeres within the segment changes in proportion to the length change of the relaxed fibre when it is shortened by Alstr (after rearranging the sarcomere lengths with a few stimuli). The calculated length change of contracting sarcomeres during the ramp is therefore: Al80

= 18,

-4,brX

where 1I,1 is the sarcomere length at the end of the ramp, i.e. at start of the elastic recoil, determined as described in the preceding section. In the largest ramp stretches used (0 4 mm imposed to the fibre end), Isbr was 1X989 + 0-027 /am (mean + S.D., n = 77); the largest sarcomere length 1,,, was 2-136 + 0 034 sum (mean + S.D., n = 44); these sarcomere lengths correspond to isometric tensions of 0-996 T. at 18 br and 0-992 T, at 184, (interpolated from the values in Table 1 of Bagni, Cecchi, Colomo & Tesi (1988)).

Evaluation of fibre inhomogeneity during lengthening As expected, the calculated sarcomere stretch was a fraction of the length change imposed by the motor, and this fraction increased with stretch amplitude as follows: 0-76 + 0413 (mean + S.D., n = 5) when A1str = 0 05 mm; 0-88 + 0414 (n = 17) when Alasr=O01 mm; 0 95+0-06 (n= 22) when Alstr = 0-2-0-3 mm and 0 99 + 0-02 (n = 77) when Alstr = 0 4 mm. In the experiments where sarcomere strain during stretching was not measured with the striation follower, a calculated sarcomere strain had to be used as the abscissa in Fig. 5. In the fibres where the sarcomere strain was both calculated as described above and measured by the striation follower, an index of inhomogeneity was determined from the ratio between measured and calculated sarcomere strain during the ramp. Only data showing a ratio included between 0 7 and 1-3, i.e. an inhomogeneity