Acidosis or inorganic phosphate enhances the length dependence of tension in rat skinned cardiac muscle

Abstract

1. We investigated the effect of acidosis on the sarcomere length (SL) dependence of tension generation, in comparison with the effect of inorganic phosphate (P_i), in rat skinned ventricular trabeculae. The shift of the mid-point of the pCa-tension relationship associated with an increase in SL from 1.9 to 2.3 μm (ΔpCa₅₀) was studied.

2. Decreasing pH from 7.0 to 6.2 lowered maximal and submaximal Ca²⁺-activated tension and increased ΔpCa₅₀ in a pH-dependent manner (from 0.21 ± 0.01 to 0.30 ± 0.01 pCa units). The addition of P_i (20 mm) decreased maximal tension and enhanced the SL dependence, both to a similar degree as observed when decreasing pH to 6.2 (ΔpCa₅₀ increased from 0.20 ± 0.01 to 0.29 ± 0.01 pCa units).

3. Further experiments were performed using 6 % (w/v) Dextran T-500 (molecular weight ∼500 000) to osmotically reduce interfilament lattice spacing (SL, 1.9 μm). Compared with that at pH 7.0, in the absence of P_i the increase in the Ca²⁺ sensitivity of tension induced by osmotic compression was enhanced at pH 6.2 (0.18 ± 0.01 vs. 0.25 ± 0.01 pCa units) or in the presence of 20 mm P_i (0.17 ± 0.01 vs. 0.24 ± 0.01 pCa units).

4. H⁺, as well as P_i, has been reported to decrease the number of strongly binding cross-bridges, which reduces the co-operative activation of the thin filament and increases the pool of detached cross-bridges available for interaction with actin. It is therefore considered that during acidosis, the degree of increase in the number of force-generating cross-bridges upon reduction of interfilament lattice spacing is enhanced, resulting in greater SL dependence of tension generation.

5. Our results suggest that the Frank-Starling mechanism may be enhanced when tension development is suppressed due to increased H⁺ and/or P_i under conditions of myocardial ischaemia or hypoxia.

It is well known that twitch tension in myocardium is enhanced as muscle length (i.e. sarcomere length, SL) is increased within the physiological range (SL from ∼1.8 to ∼2.3 μm), and is accompanied by an increase in the Ca²⁺ sensitivity of tension (Allen & Kurihara, 1982; Allen & Kentish, 1985; Lakatta, 1991). This, in part, is the basis for the Frank-Starling law of the heart. At the myofilament level, there is an increasing amount of evidence suggesting that length dependence is primarily caused by a change in interfilament lattice spacing (Harrison et al. 1988; McDonald & Moss, 1995; Fuchs & Wang, 1996; Fitzsimons & Moss, 1998; Fukuda et al. 2000). A possible consequence of decreased lattice spacing is an increase in the probability of myosin attaching to the thin filament, resulting in an increase in the number of force-generating cross-bridges (Ishiwata & Oosawa, 1974; McDonald & Moss, 1995; Fukuda et al. 2000).

An alternative hypothesis suggested that SL-dependent activation of the myofilament was caused by cardiac troponin C (TnC), which possibily acts as a ‘length sensor’ (Babu et al. 1988; Gulati et al. 1990). However, McDonald et al. (1995) demonstrated that the expression of skeletal TnC in ventricular myocytes of transgenic mice did not alter the SL dependence of the Ca²⁺ sensitivity of tension in skinned myocytes. Hence making it unlikely that cardiac TnC acts as a length sensor (Moss et al. 1991; McDonald et al. 1995).

It is well known that contractile dysfunction is one of the earliest consequences of myocardial ischaemia/hypoxia, during which intracellular H⁺ and inorganic phosphate (P_i) are markedly increased (for review see Allen & Orchard, 1987). The effects of acidosis on the contractile functions of cardiac muscle have been extensively studied (for review see Orchard & Kentish, 1990). Acidosis decreases both maximal Ca²⁺-activated tension and the Ca²⁺ sensitivity of tension (Godt & Nosek, 1989; Orchard & Kentish, 1990; Komukai et al. 1998; Fujita & Ishiwata, 1999; Fukuda & Ishiwata, 1999). Acidosis also decreases maximum shortening velocity and lowers maximum power output (Ricciardi et al. 1994). In intact ventricular preparations, it has been reported that a reduction in extracellular pH results in a change in the length-tension relationship which is similar to that seen with a decrease in the extracellular Ca²⁺ concentration (Ricciardi et al. 1986; Orchard et al. 1991), exerting little influence on the slope of the length-tension relationship when normalized to peak tension (Orchard et al. 1991). In those studies, however, twitch contraction was used to examine the effects of acidosis at limited extracellular Ca²⁺ concentrations, and the degree of presumed pH reduction inside the cell was limited to within a narrow range. Therefore, a more systematic study is needed to clarify the effect of acidosis on the SL dependence of tension generation using skinned muscle preparations.

In the present report, we studied the SL-dependent shift of the pCa-tension relationship using skinned ventricular trabeculae over a large pH range to investigate the effects of acidosis on the SL dependence of tension generation. We compared the effect of acidosis with that of inorganic phosphate (P_i), which acts on the actin-myosin interaction, but not directly on TnC, resulting in a decrease in the number of attached cross-bridges (Hibberd et al. 1985; Kentish, 1986, 1991; Kawai et al. 1988; Fukuda et al. 1998). Our results demonstrate that acidosis, as well as P_i, enhances the SL dependence of tension generation.

METHODS

Experimental procedure

Male Wistar rats (250–300 g) were anaesthetized with sodium pentobarbitone (50 mg kg⁻¹i.p.) and the hearts were removed. The rats were supplied by Saitama Experimental Animals Supply (Saitama, Japan), and the present study conforms with the Guiding Principles for the Care and Use of Animals approved by the Council of the Physiological Society of Japan. Thin trabecula muscles (diameter 100–150 μm) were dissected from the right ventricle in oxygenated Tyrode solution without Ca²⁺ at 30 °C (135 mm Na⁺, 5 mm K⁺, 1 mm Mg²⁺, 98 mm Cl⁻, 20 mm HCO₃⁻, 1 mm HPO₄^2-, 1 mm SO₄^2-, 20 mm acetate, 10 mm glucose, 5 i.u. l⁻¹ insulin, pH 7.35 when equilibrated with 5 % CO₂-95 % O₂) (Komukai et al. 1998; Fukuda et al. 2000). The preparations were skinned by superfusion with 1 % (v/v) Triton X-100 in relaxing solution (composition (mm): MgATP 4, Mops 10, EGTA 10, free Mg²⁺ 1, ionic strength 180 (pH 7.0)) for 60 min at about 2 °C. The ionic strength (IS) was adjusted with KCl. The preparations were stored at −20 °C in relaxing solution containing 50 % (v/v) glycerol and 2 mm leupeptin for 1 week or less. Both ends of the preparation were tied to thin tungsten wires with silk threads (Fukuda et al. 2000). One end was attached to a tension transducer (BG-10; Kulite Semiconductor Products, Inc., Leonia, NJ, USA) and the other to a micromanipulator (Narishige, Tokyo). The SL was adjusted to either 1.9 or 2.3 μm by measuring laser light diffraction in the relaxing solution. Isometric tension was measured in solutions containing 4 mm MgATP, 10 mm Mops (or 10 mm Mes for pH 6.2), 1 mm free Mg²⁺, various concentrations of free Ca²⁺ (adjusted with CaCl₂), 10 mm EGTA, 0.1 mm P¹,P⁵-di(adenosine-5′)pentaphosphate (AP₅A), 15 mm creatine phosphate (CP), 15 IU ml⁻¹ creatine phosphokinase (CPK) and 180 mm IS (adjusted with KCl). The pH of each solution was finally adjusted to within error of 0.02 of the desired pH.

The control pCa-tension relationship (at pH 7.0 without P_i) was first obtained at SL 1.9 μm and then at 2.3 μm. Subsequently, the pCa-tension relationships at a lower pH (6.8, 6.6 or 6.2) or in the presence of 20 mm P_i (pH 7.0) were obtained at the two SLs using the same preparation. In another series of experiments, the effect of acidosis or P_i on maximal Ca²⁺-activated tension was measured at SLs of 1.9 and 2.3 μm according to a previously described procedure (Fukuda et al. 1998; Fukuda & Ishiwata, 1999). At the end of the experiments, maximal Ca²⁺-activated tension was measured at the two SLs under control conditions (pH 7.0, no P_i). We only used data in which the final tension values were greater than 90 % of those measured at the beginning of the experiment.

The muscle width was measured under a microscope (SMZ645, Nikon, Tokyo) at a magnification of × 225. The concentrations of ionic species in solutions were estimated by computer calculation (Horiuti, 1986). All experiments were carried out at 20 ± 0.2 °C.

Data and statistical analyses

The pCa-tension relationship was fitted to the Hill equation: log(P/(100 −P)) =n_H (pCa₅₀− pCa), where P is the relative tension expressed as a percentage of the maximum, n_H is the Hill coefficient and pCa₅₀ is -log[Ca²⁺] at P = 50. All data are expressed as means ±s.e.m. Student's paired t test was used and significance verified at P < 0.05.

RESULTS

Effect of acidosis on the length dependence of the Ca²⁺ sensitivity of tension

The effect of acidosis on the SL dependence of the Ca²⁺ sensitivity of tension in the absence of P_i is shown in Fig. 1. The pH value was decreased from 7.0 to 6.8 (A), 6.6 (B) or 6.2 (C). At pH 7.0, maximal Ca²⁺-activated tension values were 50.8 ± 4.0 and 83.1 ± 6.7 mg (P < 0.001, n = 21) at SL 1.9 and 2.3 μm, respectively, and the SL-dependent shift of pCa₅₀ (ΔpCa₅₀) was ∼0.2 pCa units (see Fukuda et al. 2000). The muscle width was decreased from 120.4 ± 7.2 to 107.1 ± 6.0 μm (P < 0.001, n = 5) when SL was increased from 1.9 to 2.3 μm during relaxation (i.e. ∼11 % decrease in the width). Acidosis did not change the muscle width (119.4 ± 7.3 and 105.8 ± 6.3 μm (P < 0.001, n = 5) at SL 1.9 and 2.3 μm, respectively, at pH 6.2). Consistent with previous studies using skinned cardiac muscle preparations (Godt & Nosek, 1989; Orchard & Kentish, 1990; Fujita & Ishiwata, 1999; Fukuda & Ishiwata, 1999), acidosis shifted the pCa-tension relationship to the right (Fig. 1). It was also found that the SL-dependent shift of the pCa-tension relationship was enhanced with acidosis in a pH-dependent manner (P < 0.05 and P < 0.01 for pH 6.6 and 6.2, respectively). At pH 6.8, ΔpCa₅₀ was not significantly different from that at pH 7.0 (Fig. 1A, inset); however, it increased from 0.21 ± 0.01 to 0.30 ± 0.01 pCa units at pH 6.2 (Fig. 1C, inset).

Figure 1. Effects of acidosis on pCa-tension relationships at different SLs.

Solvent conditions: 4 mm MgATP, 10 mm Mops (or 10 mm Mes for pH 6.2), 1 mm free Mg²⁺, various concentrations of free Ca²⁺ (pCa adjusted using Ca-EGTA), 0.1 mm AP₅A, 15 mm CP, 15 IU ml⁻¹ CPK and IS maintained at 180 ± 1 mm. Temperature, 20 ± 0.2 °C. A–C show the effect of lowering pH from 7.0 to 6.8, 6.6 and 6.2, respectively, on the pCa-tension relationships at SLs of 1.9 μm (circles) and 2.3 μm (squares). Note that different preparations were used in A, B and C. Continuous lines and open symbols, pCa-tension relationships at pH 7.0; dotted lines and filled symbols, those at a lower pH. Data obtained for each preparation were fitted to the Hill equation, and the results were simulated by the Hill equation with the mean values of pCa₅₀ and n_H. Each inset represents ΔpCa₅₀ (difference between the values of pCa₅₀ at long and short SLs) at pH 7.0 and at a lower pH. pCa-tension relationships were normalized to the maximal tension at pCa 4.8 for pH 7.0 and at 4.4, 4.1 and 3.8, respectively, for pH 6.8, 6.6 and 6.2. Error bars, s.e.m. of 5–6 data points. N.S., not significant compared with pH 7.0.

Table 1 summarizes the n_H values of the pCa-tension relationships of Fig. 1. Decreasing pH to 6.2 significantly increased n_H at SL 2.3 μm. However, there was no definitive correlation between the magnitude of n_H at either SL and that of ΔpCa₅₀ (i.e. at pH 6.6, ΔpCa₅₀ was significantly increased with no significant change in n_H at either SL, Fig. 1B and Table 1) (see also Fukuda et al. 2000). The n_H values obtained here are relatively high in comparison with those reported in previous studies using rat skinned ventricular muscle (e.g. Kentish, 1986; Harrison et al. 1988; McDonald & Moss, 1995). One reason for this may lie in the different experimental conditions. It is also possible that the high n_H values came about because of our measurement protocol for the pCa-tension relationship. In accordance with our previous report (Fukuda et al. 2000), we measured the pCa-tension relationship by cumulatively raising the free Ca²⁺ concentration from the relaxing condition. It is thus possible that the thin filament is readily activated upon solution change, because of pre-activation at the preceding lower Ca²⁺. This may result in enhanced co-operative activation of the thin filament (i.e. high n_H).

Table 1.

Change in the Hill coefficient with acidosis

	Fig. 1A		Fig. 1B		Fig. 1C
SL	pH 7.0	pH 6.8	pH 7.0	pH 6.6	pH 7.0	pH 6.2
1.9 μm	5.81 ± 0.21	5.79 ± 0.28	5.49 ± 0.28	4.86 ± 0.30	5.48 ± 0.38	6.02 ± 0.52
2.3 μm	4.18 ± 0.29	4.38 ± 0.31	4.70 ± 0.19	4.80 ± 0.18	4.59 ± 0.21	5.70 ± 0.43 *

Values were obtained from Fig. 1A–C (mean ±s.e.m.; n = 5–6)

^*P < 0.05 compared with the corresponding value of pH 7.0.

Effect of P_i on the length dependence of the Ca²⁺ sensitivity of tension

Figure 2 shows the effect of P_i (20 mm) on the SL dependence of the Ca²⁺ sensitivity of tension at pH 7.0. Consistent with previous studies using skinned cardiac muscle (Godt & Nosek, 1989; Kentish, 1986; Fukuda et al. 1998), P_i shifted the pCa-tension relationship to the right. It was found that ΔpCa₅₀ was significantly increased in the presence of P_i (0.20 ± 0.01 vs. 0.29 ± 0.01 pCa units), and the magnitude of increase was comparable to that obtained by lowering pH from 7.0 to 6.2 (see Fig. 1C), although the rightward shift of the pCa curve was smaller. The n_H values in the absence and presence of P_i were 6.26 ± 0.24 and 6.94 ± 0.36 (P > 0.05, n = 5), respectively, at SL 1.9 μm, and 5.08 ± 0.33 and 6.40 ± 0.43 (P > 0.05, n = 5), respectively, at SL 2.3 μm. Consistent with the work of Kentish (1986) using rat ventricular muscles, n_H did not change significantly with P_i.

Figure 2. Effects of P_i on pCa-tension relationships at different SLs.

Solvent conditions: the same as in Fig. 1 except that pH was maintained at 7.0 and 20 mm P_i was added, maintaining IS at 180 mm. Continuous lines and open symbols, pCa-tension relationships in the absence of P_i; dotted lines and filled symbols, those in the presence of P_i. Circles, SL 1.9 μm; squares, SL 2.3 μm. Inset represents ΔpCa₅₀ in the absence (-P_i) and presence (+P_i) of P_i. pCa-tension relationships were normalized to the maximal tension at pCa 4.8 in the absence of P_i and at 4.6 in the presence of P_i. Error bars, s.e.m. of 5 data points.

Effects of acidosis and P_i on tension induced by osmotic compression

Dextran T-500 (6 % w/v) reduced the muscle width from 129.3 ± 8.1 to 111.6 ± 6.8 μm (P < 0.001, n = 10) under relaxing conditions at SL 1.9 μm. The degree of reduction (∼13.5 %) was similar to that observed when increasing SL to 2.3 μm without Dextran T-500 (∼11 %). Osmotic compression shifted the pCa-tension relationship leftward (0.17–0.18 pCa units), similar to that observed when increasing SL to 2.3 μm (∼0.2 pCa units) (Fukuda et al. 2000). Maximal tension was also increased by about 15 % with compression, but this change was smaller than that obtained by lengthening (McDonald & Moss, 1995; Fukuda et al. 2000); in contrast, increasing SL from 1.9 to 2.3 μm in the absence of Dextran T-500 increased maximal tension by about 60 % (see above). It was found that the Dextran-induced increase in the apparent Ca²⁺ sensitivity was significantly increased at pH 6.2 (0.18 ± 0.01 vs. 0.25 ± 0.01 pCa units) or in the presence of 20 mm P_i (0.17 ± 0.01 vs. 0.24 ± 0.01 pCa units) (Fig. 3). The n_H values at pH 7.0 and 6.2 were 6.12 ± 0.36 and 6.36 ± 0.40 (P > 0.1, n = 5), respectively, without Dextran T-500 and 5.63 ± 0.26 and 6.08 ± 0.54 (P > 0.1, n = 5), respectively, with Dextran T-500. Those in the absence and presence of P_i were 6.18 ± 0.54 and 6.43 ± 0.39 (P > 0.1, n = 5), respectively, without Dextran T-500 and 6.18 ± 0.38 and 7.08 ± 0.80 (P > 0.1, n = 5), respectively, with Dextran T-500.

Figure 3. Effects of acidosis or P_i on pCa-tension relationships in the absence and presence of Dextran T-500.

Solvent conditions: the same as in Figs 1C and 2 for acidosis and P_i, respectively, except that 6 % (w/v) Dextran T-500 (molecular weight ∼500 000, Amersham Pharmacia Biotech AB, Sweden) was used instead of increasing SL to 2.3 μm. Circles, in the absence of Dextran T-500; squares, in the presence of Dextran T-500. SL was fixed at 1.9 μm. A, effect of acidosis (pH 6.2). P_i was absent. Continuous lines and open symbols, pCa-tension relationships at pH 7.0; dotted lines and filled symbols, those at pH 6.2. Inset represents ΔpCa₅₀ (difference between the values of pCa₅₀ with and without Dextran T-500) at pH 7.0 and 6.2. B, effect of P_i (20 mm). pH was held at 7.0. Continuous lines and open symbols, pCa-tension relationships in the absence of P_i; dotted lines and filled symbols, those in the presence of P_i. Inset represents ΔpCa₅₀ in the absence (-P_i) and presence (+P_i) of P_i. pCa-tension relationships in A and B were normalized to the maximal tension at the values of pCa shown in Figs 1C and 2, respectively. Error bars, s.e.m. of 5 data points.

Effects of acidosis and P_i on maximal Ca²⁺-activated tension

Finally, we studied maximal Ca²⁺-activated tension as functions of pH and P_i at SLs of 1.9 and 2.3 μm. As reported previously in skinned skeletal and cardiac muscle preparations (Godt & Nosek, 1989; Kentish, 1991; Fujita & Ishiwata, 1999; Fukuda & Ishiwata, 1999), maximal tension decreased almost linearly with the reduction of pH (Fig. 4A). We found that the inhibitory effect was SL dependent, and that the effect was weaker at the longer SL. The inhibitory effect of P_i on maximal tension was also SL dependent; the effect was less at the longer SL (Fig. 4B).

Figure 4. Effects of acidosis or P_i on maximal Ca²⁺-activated tension at different SLs.

Conditions: the same as in Figs 1 and 2 for acidosis and P_i, respectively. Circles, SL 1.9 μm; squares, SL 2.3 μm. A, effect of acidosis in the absence of P_i. The values of pCa at pH 7.0, 6.8, 6.6 and 6.2 for obtaining maximal Ca²⁺-activated tension were set based on Fig. 1 (i.e. 4.8, 4.4, 4.1 and 3.8 for pH 7.0, 6.8, 6.6 and 6.2, respectively) and estimated to be 4.0 for pH 6.4 by drawing the pCa curve. Tension was normalized with respect to pH 7.0. B, effect of P_i. pH was held at 7.0. Tension was normalized with respect to 0 mm P_i. The value of pCa was held at 4.8 in all conditions, because there was little difference between the tension at pCa 4.8 and that at pCa < 4.8 in the presence of 30 mm P_i. Error bars, s.e.m. of 3–4 data points. *P < 0.05 compared with the corresponding values for SL 1.9 μm.

DISCUSSION

We found that in rat skinned ventricular trabeculae, both acidosis and P_i exhibited a similar inhibitory effect on maximal Ca²⁺-activated tension and enhanced the magnitude of SL-dependent changes of the pCa-tension relationship. We discuss these results of the effects of interfilament lattice spacing, acidosis and P_i in relation to the formation of force-generating cross-bridges.

First, an increase in SL results in a decrease in the lateral separation between the thick and thin filaments both in living (Matsubara & Millman, 1974) and skinned (Irving et al. 2000) myocardium. A recent X-ray study using rat skinned ventricular trabeculae confirmed that interfilament lattice spacing (d₁₀) decreases with an increase in SL, maintaining a nearly constant lattice volume over the physiological range of SL (i.e. from ∼1.9 to ∼2.4 μm) (Irving et al. 2000). Consistent with our previous study (Fukuda et al. 2000), an increase in SL from 1.9 to 2.3 μm produced a similar muscle width reduction as osmotic compression using 6 % (w/v) Dextran T-500, and both procedures increased the Ca²⁺ sensitivity of isometric tension to a similar degree (Figs 1–3). Thus, we consider that the change in the lattice spacing is the primary cause of the modulation of the SL-dependent Ca²⁺ sensitivity of tension in the cardiac myofilament (McDonald & Moss, 1995; Fuchs & Wang, 1996; Fukuda et al. 2000).

A decrease in pH towards the isoelectric point of the myofilaments decreases the charge distribution on the filaments, causing a reduction of electrostatic repulsion force between filaments, which in turn causes interfilament lattice spacing to shrink (Millman, 1998). In the present study, a decrease in pH from 7.0 to 6.2 did not significantly alter the muscle width at a SL of 1.9 or 2.3 μm. It is thus reasonable to consider that interfilament lattice spacing is not directly influenced by the change in pH in the range examined here, presumably because the isoelectric point is far below 6.2 (i.e. < 5.0; see Kawai et al. 1988; Millman, 1998).

The inhibitory effect of acidosis on Ca²⁺-activated tension is in part attributable to a decrease in the number of force-generating cross-bridges (Metzger & Moss, 1988, 1990; Kentish, 1991; Fujita & Ishiwata, 1999; Fukuda & Ishiwata, 1999). Kentish (1991) postulated that H⁺ is released during the transition between the non-force-generating states preceding the P_i release (force-generating) step. An increased concentration of H⁺ will thus reverse the H⁺ release step and decrease the number of force-generating cross-bridges (Kentish, 1991). P_i is believed to reverse the P_i release step in the cross-bridge cycle by shifting the distribution of cross-bridges toward a state with a full complement of bound products (ADP and P_i) (i.e. non-force-generating state) (Hibberd et al. 1985; Kentish, 1986, 1991; Kawai et al. 1988). It is thus realized that both acidosis and P_i reduce the co-operative activation of the thin filament by decreasing the number of strongly binding cross-bridges. In the present study, lowering pH to 6.2 and the addition of 20 mm P_i reduced maximal tension and enhanced the SL-dependent increase in the Ca²⁺ sensitivity of tension to a similar degree (Figs 1C, 2 and 4). However, the rightward shift of the pCa₅₀-tension curve was substantially larger for acidosis (Figs 1C and 2), which is reasonable because P_i decreases the number of force-generating cross-bridges with no direct effect on TnC, whereas acidosis reduces the affinity of TnC for Ca²⁺ in addition to its inhibitory effect on cross-bridges (Orchard & Kentish, 1990). It is therefore suggested that a change in the cross-bridge distribution, shown by a decrease in maximal tension, but not a reduced affinity of TnC for Ca²⁺, underlies the mechanisms of enhanced SL-dependent tension generation by acidosis (see Fukuda et al. 2000). Once the number of strongly binding cross-bridges is decreased by either H⁺ or P_i, co-operative activation of the thin filament will be reduced, resulting in an increase of detached cross-bridges that are available for interaction with actin. Consequently, the fraction of cross-bridges recruited to the force-generating state on lengthening (i.e. lattice shrinkage) will be increased, resulting in a greater SL dependence of isometric tension (Ishiwata & Oosawa, 1974). This interpretation is supported by the fact that both acidosis and P_i enhance the increase in the Ca²⁺ sensitivity of tension due to osmotic compression to a degree similar to that observed in the study of varying SL (Fig. 3). Also, this interpretation is consistent with the recent finding that SL-dependent tension generation is attenuated by an enhanced co-operative activation of the thin filament by strongly binding cross-bridges produced by the addition of N-ethylmaleimide-modified myosin subfragment 1 (Fitzsimons & Moss, 1998) or ADP (Fukuda et al. 2000).

We observed SL dependence in the inhibitory effects of acidosis and P_i (Fig. 4). Assuming that the formation of the force-generating cross-bridges is accelerated with an increase in SL, it is expected that the inhibitory effect of H⁺ or P_i becomes less pronounced at a longer SL. Since maximal tension was dramatically increased by lengthening compared with osmotic compression (i.e. ∼60 %vs.∼15 %) (see also McDonald & Moss, 1995; Fukuda et al. 2000), the reduction of interfilament lattice spacing may not fully account for the increase in maximal tension due to an increase in SL. There may be additional mechanisms modulating the SL-dependent activation in the cardiac myofilament, such as extension of titin/connectin (Fukuda et al. 2001), in addition to the reduction of the lattice spacing, and therefore the reduction of the lattice spacing alone may not be sufficient to produce force-generating cross-bridges. Alternatively, it is possible that osmotic compression and lengthening may not be precisely equivalent such that changes in lattice packing may differ in the two cases; the shape of the muscle may be altered differently by mechanical stretch compared with osmotic compression (McDonald & Moss, 1995; Fukuda et al. 2000).

Alterations of SL-dependent shift of the pCa-tension relationship are predicted to affect the relation between muscle length and tension production (Fuchs, 1995). The present finding that acidosis enhances the SL dependence of the Ca²⁺ sensitivity of tension (Fig. 1) leads us to conclude that as pH falls, the slope of the length-tension relationship at physiologically submaximal levels of Ca²⁺ is increased. This is in disagreement with an earlier finding that acidosis does not essentially change the slope of the length-tension relationship in intact myocardium (Ricciardi et al. 1986; Orchard et al. 1991). The difference in the effect of acidosis on length-dependent tension generation may lie in the buffering capacity of pH by the cell membrane in intact preparations (Allen & Orchard, 1987; Ricciardi et al. 1994), resulting in values of intracellular pH between 6.6 and 6.8 in those studies (Ricciardi et al. 1986; Orchard et al. 1991), which are higher than that used in the present study (6.2). Other differences include ionic conditions and experimental protocols.

The manifestation of a change in the pCa-tension relationship in the ventricular volume-end-systolic pressure (ESP) relation of the intact ventricle is complex. However, a simplified prediction of how a change in SL-dependent activation may affect the slope of the volume-ESP relation can be made (Arteaga et al. 2000). Arteaga et al. (2000) reported that the slope of the volume-ESP relation may depend on the degree of activation of the myofilament on SL. We have reported that the volume-ESP relation may be impaired during the late phase of myocardial ischaemia/hypoxia (Fukuda et al. 2000), in which the ratio of the intracellular concentration of ADP to that of ATP is dramatically increased and subsequently augments tension development (Kammermeier et al. 1982; Allen & Orchard, 1987; Ventura-Clapier & Veksler, 1994; Fukuda et al. 1998). In contrast, the present results indicate that during the early phase of myocardial ischaemia/hypoxia, in which accumulation of ADP is minimal and tension development is suppressed by the increased H⁺ and/or P_i (Allen & Orchard, 1987; Ventura-Clapier & Veksler, 1994), the SL-dependent activation is enhanced. It can therefore be predicted that in the beating heart, the slope of the volume-ESP relation may be increased during the initial phase of myocardial ischaemia/hypoxia, thereby enhancing the Frank-Starling mechanism.