The series elastic component and the force-velocity relation in the anterior byssal retractor muscle of Mytilus edulis during active and catch contractions

The anterior byssal retractor muscle (ABRM) of a common mussel, Mytilus edulis, is known to exhibit ‘catch’ contraction which continues long after stimulation has ended with a small rate of energy expenditure (Nauss & Davies, 1966; Baguet & Gillis, 1968). The ABRM can shorten or develop tension actively during the early phase of contraction in response to acetylcholine (ACh) or long cathodal pulses. It barely shortens, however, after an isotonic release and barely redevelops tension after a quick release during catch contraction attained after the removal of ACh or stimulating currents (Abbott & Lowy, 1958; Lowy & Millman, 1963). The mechanism of catch is still a matter of controversy. According to the linkage hypothesis (Lowy & Millman, 1963), both active and catch contractions are produced by the same crosslinkages between the thick and thin filaments, whereas the parallel hypothesis (or the independent catch hypothesis, Johnson, 1962; Rüegg, 1963, 1968, 1971) supposes that active and catch contractions are due to the actomyosin and the paramyosin systems respectively.

The classical concept of the series elastic component (SEC) proposed by Hill (1938) is that the SEC in series with the contractile component has a definite force-extension relation and, therefore, its characteristics do not change during contraction. Recently, Huxley & Simmons (1971a, b, 1973) have demonstrated that the SEC in vertebrate skeletal muscle is a function of isometric tension immediately before the time of release. This fact indicates that most of the SEC may reside in the cross-linkages between the thick and thin filaments rather than in series with the contractile component. Thus, the magnitude of isometric tension may correspond directly to the number of cross-linkages attached (Huxley & Simmons, 1973). The present experiments were undertaken to examine the dependence of the SEC on the isometric tension in the ABRM during both active and catch contractions to give information about the mechanism of active and catch contractions. The opportunity was also taken to obtain the force-velocity relations at various phases of contraction induced by ACh.

Specimens of Mytilus edulis were collected at the Misaki Marine Biological Station) and kept in tanks of aerated sea water at 18 °C. The ABRM was isolated with a rectangular piece of shell attached to one end and the byssal organ left at the other, and was carefully teased under a binocular microscope to obtain a small bundle of fibres of about 1 mm diameter. The preparation was mounted horizontally in an Acrylic plastic chamber (10 ml) filled with the experimental solution (Fig. 1). The rectangular piece of shell was held securely by an acrylic holder with stainless-steel wire (diameter, 1 mm) and connected to the tension transducer, while the byssal end was hooked to the lever of the displacement transducer by means of a stainless-steel wire connector (o-z mm in diameter and 5 mm in length, Sugi, 1972). The standard experimental solution (artificial sea water) had the following composition (mm): NaCl, 497; KC1, 10; CaCl₂, 20; MgCl₂, 54 (pH adjusted to 7·O–7·2 by NaHCO₃). Solutions were flowed at a rate of 4–5 ml/min by a water vacuum suction tube.

The displacement transducer was a light beam-phototube system. The moving element was an aluminium lever pivoted on the bearings. The distal part of the lever dipped into the experimental solution, and the preparation was connected at a point 3 cm distant from the pivot. Movement of the lever varied the amount of light transmitted to the phototube, the output of which was displayed on the upper trace of an ink-writing oscillograph. The transducer was linear over a range of 2 mm. The compliance of the lever at the point of attachment of the preparation was 0̲5 μm/g-

A strain gauge (U-gauge, Shinko Tsushin Co.) was used as the tension transducer. Its compliance was I μm/g and its natural frequency of oscillation about 150/s. The output of the tension transducer was displayed on the lower trace of the oscillograph.

Stray compliance of the whole recording system, including the connexions between the preparation and the transducers, was estimated by substituting a length of stainless-steel wire for the preparation, and was found to be about 2 μm/g. Since the tension generated by the preparation was mostly below 15 g and the length of the preparation was I6 · 2 cm, the stray compliance did not exceed 0 · 2% of the length of the preparation.

The resting length Lo of the preparation, at which the resting tension was just barely detectable in the presence of 10 −6 M 5-hydroxytryptamine (5– HT), was determined to perform the experiments within the range of lengths where the resting tension was negligible (o · 8-1 · oL0), thus avoiding complications arising from the development of resting tension. As illustrated in Fig. 1 the preparation was first kept isometric by the stops, the length of the preparation being varied by a micromanipulator carrying the tension transducer.

The preparation was stimulated to contract maximally by the application of a supramaximal concentration of ACh (10 − 3M). At any chosen moment during the resulting tension response the lever was released by removing one of the stops with an electromagnetic device (not shown) so as to change the load on the preparation quickly from the isometric value to a new level determined by the loading spring. Since the isometric tension in response to ACh was not constant but changed with time (Fig. i, inset), the tension just before the moment of the step change in load was designated as Pt while Pl represented the load under which the isotonic shortening or lengthening of the preparation took place. The right stop was removed for the shortening steps (Pl < Pt) and the left stop for the lengthening steps (Pl > Pt).

After each isotonic release the preparation was made to relax by 10−6 M 5-HT. Contractions produced by ACh at intervals of more than 10 min were fully reproducible. All experiments were performed at room temperature (18–25 °C).

The ABRM fibres showed active tension development in response to ACh. The peak tension was reached within 30 sec after the beginning of the application of ACh. The rising phase of the tension developed in the presence of ACh will hereafter be designated as active contraction. The preparation was then returned to the standard solution to produce catch tension. The tension normally decayed to 70–50 % of the peak tension within a few minutes after the removal of ACh, and then continued to relax with a much slower rate (Fig. 1 inset). The transition from active to catch contractions was preliminarily examined by recording the redevelopment of tension after a quick release (Jewell, 1959). It was found that catch contraction was established within 3–5 min after the removal of ACh.

Fig. 2 A shows the results of a typical experiment in which the load on the preparation was quickly decreased from the same isometric level of Pt to three different values of Pl < Pt during active contraction. The step change in load was accompanied by a rapid shortening of the preparation followed by the subsequent slower, isotonic, shortening with a constant velocity depending on the value of Pl. The response of the ABRM to the shortening steps was qualitatively similar to that of active frog skeletal muscle (Jewell & Wilkie, 1958).

When similar step changes in load were made during catch contraction, the initial rapid change in length occurred in the same manner as during active contraction, while the preparation barely shortened isotonically as shown in Fig. 2B. These results of shortening steps are in agreement with those reported by Jewell (1959) on the whole ABRM.

Fig. 3 A shows typical records of length changes when the load on the preparation was quickly increased from the same isometric level Pt to four different values of Pl > Pt during active contraction. When Pl was not more than about i-j-Pf, the initial rapid lengthening was followed by a slower steady lengthening (Fig. 3 Al, A2). If Pl was further increased, the velocity of isotonic lengthening was not constant but decreased with time (Fig. 3 A4). The above mode of isotonic lengthening in the ABRM also seems to be qualitatively similar to that in frog skeletal muscle; Katz (1939) reported that frog muscle showed various amounts of ‘give’ intervening between the immediate lengthening of the SEC and the steady isotonic lengthening when a suddenly applied load was larger than 1·2– 1·4Po- On this basis, the non-steady isotonic lengthening with Pl > 1·3Pt might result from partial ‘give’ of the ABRM fibres.

Fig. 3B shows the records of lengthening steps during catch contraction. The velocity of isotonic lengthening after the initial rapid lengthening was very much smaller during catch contraction than during active contraction, and the two phases of lengthening were clearly separated by a corner even if Pt was increased to 2·oPt (Fig. 3B3).

The length change of the SEC in response to quick changes in load was measured by extrapolating the first 100 ms of the isotonic phase of length change back to the moment of release (Jewell & Wilkie, 1958), on the assumption that the initial rapid length change is due to the response of the SEC to the step change in load. The load-extension curve of the SEC was obtained by plotting the initial rapid length changes against the corresponding step changes in load (Jewell & Wilkie, 1958).

To compare the load-extension curves of the SEC between active and catch contractions, shortening steps from the same level of isometric tension were applied to the same preparation during both active and catch contractions. As shown in Fig. 4 all the points obtained fell on the same load-extension curve, indicating that the SEC in the ABRM remains the same during both active and catch contractions as has been reported by Jewell (1959) and Lowy & Millman (1963).

Examples of the effect of changing the level of the isometric tension immediately before step changes in load on the load-extension curves of the SEC are shown in Figs. 5 and 6. In Fig. 5 A, one series of shortening steps (open circles) was applied during active contraction, and the other (closed circles) during catch contraction. The value of Pt was larger during active contraction than during catch contraction. The magnitude of initial rapid shortening for a given amount of quick decrease in load varied in proportion to the value of Pt, so that the load-extension curve was scaled down in proportion to the isometric tension immediately before step decrease in load. Thus, if the initial rapid shortening was plotted against the relative load Pl/Pt instead of Pl, all the points obtained from different levels of Pt fell on the same load-extension curve (Fig. 5B). Similar results were obtained when two series of shortening steps were applied at two different levels of Pt during active contraction (Fig. 6).

In some experiments, the initial rapid lengthening in response to lengthening steps was also estimated during active contraction in the range of Pl < 1·3Pt where the steady isotonic lengthening was observed, and up to 1·5Pt or more during catch contraction. When the initial rapid lengthening was plotted against Pl/Pt, the points obtained both during active and catch contractions fell on the same straight line (Fig. 6). These results indicate that the load-extension curve of the SEC in the ABRM is dependent on the isometric tension immediately before step changes in load, but remains unchanged irrespective of whether the preparation is in active or in catch contraction.

The amount of shortening of the SEC corresponding to the shortening step from Pt to zero appeared to be independent of the level of Pt, and was estimated to be 2–4% Lo. This value was somewhat smaller than those obtained by previous investigators on the ABRM (3–4 % Lo, Jewell, 1959 ; 4–5 % LQ, Lowy & Millman, 1963).

The effect of hypertonic solution on the SEC was also examined by soaking the preparation in a hypertonic solution prepared by adding 0·5 M sucrose to the standard solution. The osmotic strength of the hypertonic solution was 1·4 times that of the standard solution as measured with a freezing point osmometer (Advanced Instruments Co.). Since the peak height of tension produced by ACh was decreased by 50 % or more by soaking the preparation in the hypertonic solution for 30 min, the initial rapid length changes for shortening steps were plotted against the relative load. In the hypertonic solution, the compliance of the SEC in the ABRM was reduced to about one half its normal value (Fig. 7) as is the case in frog skeletal muscle (Jewell & Wilkie, 1958).

Fig. 8 is a typical example of the force-velocity relation obtained by applying step changes in load at the same level of Pt during active contraction. When the velocity of isotonic shortening was plotted against the corresponding relative load, the curve obtained for the shortening steps approximated closely to the hyperbolic relation which can be expressed by the Hill equation (Hill, 1938) as

For the lengthening steps the velocity of isotonic lengthening was measured only in the range of Pl < 1·3Pt since the velocity of isotonic lengthening was not steady for Pl >1·3Pt (Fig. 3 A). It was noticed that the observed values for the velocity of isotonic lengthening were appreciably larger than the values predicted by the Hill equation using constants a and b obtained from the shortening steps (Fig. 8). This contrasts with frog skeletal muscle in which the actual velocities of isotonic lengthening are several times smaller than those calculated from the force-velocity relation of shortening muscle (Katz, 1939). In cat soleus muscle the velocities of isotonic lengthening are also larger than those expected from the force-velocity relation in isotonic shortening (Joyce & Rack, 1969).

During catch contraction, the steady velocities of isotonic shortening and lengthening could be measured over a wide range of Pl. The velocities of isotonic shortening and lengthening were many times smaller than the corresponding values for active contraction (Fig. 9).

The present experiments have clearly demonstrated that the load-extension curves of the SEC in the ABRM are dependent on the isometric tension immediately before step changes in load, but remained unchanged irrespective of whether the ABRM is in active or in catch contraction (Figs. 4-6). Such a dependence of the SEC on the isometric tension may not be explained by the classical concept of the SEC, which is in series with the contractile component and has a definite force-extension relation (Hill, 1938). A similar dependence of the SEC on the isometric tension has been reported with vertebrate skeletal muscle fibres. In these fibres, the isometric tension is changed by varying the amount of overlap between the thick and thin filaments (Huxley & Simmons, 1971a, 1973), by varying the time of release during tension development in a tetanus (Huxley & Simmons, 1973; Bressler & Clinch, 1974) or by varying calcium ion concentration (Podolsky & Teichholz, 1970; Wise, Rondinone & Briggs, 1973). These observations have been taken as evidence that the level of isometric tension may correspond directly to the number of the cross-linkages in which the SEC may largely reside (Huxley & Simmons, 1973).

In the ABRM, active tension is generally believed to be due to the interaction between the thick and thin filaments (Hanson & Lowy, 1959, 1961 ; Lowy & Hanson, 1962; Szent-Gyôrgyi, Cohen & Kendrick-Jones, 1971). The thin filaments are composed predominantly of actin, while the thick filaments bear projections of myosin heads on the surface of the paramyosin core (Sovieszek, 1973; Elliott, 1974;

Nonomura, 1974). On this basis, if one assumes that active tension in the ABRM is proportional to the number of cross-linkages between the two filaments, a large part of the tension-dependent SEC may reside in each elementary site of the contractile mechanism.

Vertebrate skeletal muscles bathed in hypertonic solutions produce less contractile tension than in normal tonicity solution. Though the hypertonic solution may cause some excitation-contraction uncoupling, its major effect is that the resultant increase in ionic strength inside the fibre directly affects the ability of the contractile proteins to shorten (Podolsky & Sugi, 1967) and to generate tension (Gordon, Godt, Donaldson & Harris, 1973 ; Homsher, Briggs & Wise, 1974). Jewell & Wilkie (1958) have shown that doubling the tonicity of bathing solution reduces the compliance of the SEC to about one-half of its normal value in whole sartorius muscle, in which one-half of the SEC is distributed along the muscle fibre in the normal tonicity solution. This fact has been taken as evidence that the hypertonic solution has the effect of damping any parts of the SEC located within the fibres. As in vertebrate skeletal muscle, the compliance of the SEC in the ABRM was reduced to about onehalf of its normal value in a hypertonic solution (Fig. 7) with a reduction of the isometric tension. The effect of hypertonic solution on the SEC, together with the tension dependence of the SEC, could indicate that at least 50 % of the SEC observed in the normal tonicity solution resides in the contractile component. The contribution of the connective tissue to the SEC may be small, though the individual fibres of the ABRM (I-2-I-8 mm in length and 5 μm in diameter) are connected to one another via connective tissue components (Twarog, Dewey & Hidaka, 1973). At present, little information is available about the mode of action of hypertonicity on the SEC and the isometric tension in the ABRM.

According to Sobieszek (1973), the length of the contractile unit equivalent to a half-sarcomere of striated muscle cannot be smaller than 10 μm in the ABRM if the length of the thick filaments is taken into consideration. In the present study, the maximal extension of the SEC was 2–4% L0. Therefore, a conservative estimate of the maximal extension of the SEC located in the contractile component is 1–2 % Lo. If one assumes that the SEC inside the fibre resides entirely in the cross-linkages, then the cross-linkages in each half-sarcomere may be elongated by 100-200 nm. This amount of extension is much larger than the maximal extension supposed for the cross-linkages of frog skeletal muscle fibres during isometric contraction (8 nm), Huxley & Simmons, 1971a, 1973), and is unlikely to arise only from the elongation of the cross-linkages between the thick and thin filaments. It seems possible that the SEC in the ABRM may reside not only in the cross-linkages, but also along the thick or thin filaments ; if the compliance of the filaments is altered in proportion to the number of the cross-linkages formed on them, then the change in compliance of the filaments during contraction would also contribute to the tension-dependent SEC. In glycerinated rabbit skeletal muscle, the maximal extension of the SEC has been estimated to be 24–78 nm in each half-sarcomere. This also seems too large to be ascribed to the extension of the cross-linkages, which are believed to be formed by the S–I segment of myosin of 57 nm in length (Wise et al. 1973). Further knowledge of the structures and mechanical properties of the filaments is needed to decide the actual cause of the SEC in the ABRM. However, the identity of the SEC between active and catch contractions implies that the undamped elasticity of the contractile component is the same irrespective of the state of the ABRM. This favours the linkage rather than the parallel hypothesis, for if active and catch contractions are due to two independent contractile systems in parallel, then an identical elasticity in each system seems to be somewhat surprising.

The results on the force-velocity relation have shown that the value of varies with the phase of active contraction, being 0-12-0-25 L0/8 during the rising phase of tension (Fig. 8) and 0-05-0-09 L0/8 at the peak tension (Fig. 9). The latter values are in agreement with V_max of about 0-05 L0/8 obtained from the force velocity relation when shortening steps were applied to the whole ABRM at the peak tension produced by electrical stimulation (at 14 °C, Abbott & Lowy, 1958). Lowy & Millman (1963) determined V_max of 0 · 25 L0/8 by releasing the ABRM by about 10% Lo during phasic or tonic stimulation and then measuring the time that the muscle had been shortening at zero load from the tension records (18-25 °C)-This value is also compatible with the value of V_max obtained during the rising phase of tension since their experiments seem to be performed during the early phase of active contraction. According to Brady (1966), the velocity of shortening following step changes in load is larger at the early phase of tension development than at peak tension in mammalian cardiac muscle.

In vertebrate skeletal muscle, V_max decreases with decreasing concentration of free calcium ions in the myoplasm (Julian, 1971 ; Wise, Rondinone & Briggs, 1971). It is of interest, therefore, that the peak of the calcium transient coincides with the time of the maximum rate of rise of tension in barnacle muscle fibres (Ashley & Ridgway, 1970). This suggests that the difference in the value of V_max according to the phase of active contraction in the ABRM reflects the change in free calcium ion concentration in the myoplasm by which the turnover rate of the cross-linkages is controlled. This accords with existing evidence which indicates that the transition from active to catch contraction may be related to a decrease in the free calcium ion concentration of the myoplasm (Atsumi, Sugi & Aikawa, 1974; Atsumi & Sugi, 1975).