The effects of Pi on sarcoplasmic reticulum (SR) Ca2+ regulation were studied in mechanically skinned rat skeletal muscle fibers. Brief application of caffeine was used to assess the SR Ca2+ content, and changes in concentration of Ca2+([Ca2+]) within the cytosol were detected with fura 2 fluorescence. Introduction of Pi (1–40 mM) induced a concentration-dependent Ca2+ efflux from the SR. In solutions lacking creatine phosphate (CP), the amplitude of the Pi-induced Ca2+ transient approximately doubled. A similar potentiation of Pi-induced Ca2+ release occurred after inhibition of creatine kinase (CK) with 2,4-dinitrofluorobenzene. In the presence of ruthenium red or ryanodine, caffeine-induced Ca2+ release was almost abolished, whereas Pi-induced Ca2+ release was unaffected. However, introduction of the SR Ca2+ ATPase inhibitor cyclopiazonic acid effectively abolished Pi-induced Ca2+ release. These data suggest that Pi induces Ca2+ release from the SR by reversal of the SR Ca2+ pump but not via the SR Ca2+ channel under these conditions. If this occurs in intact skeletal muscle during fatigue, activation of a Ca2+efflux pathway by Pi may contribute to the reported decrease in net Ca2+ uptake and increase in resting [Ca2+].
- sarcoplasmic reticulum
- Ca2+ pump
previous studies on isolated skeletal muscle fibers have revealed characteristic changes in intracellular concentration of Ca2+([Ca2+]i) regulation during fatigue induced by intermittent tetanic stimulation [for reviews see Fitts (11) and Allen et. al (2)]. In the early stages of fatiguing stimulation, tetanic [Ca2+]i increases transiently. This is followed by prolongation of the [Ca2+]i transient and a progressive increase in resting [Ca2+]i. These effects are consistent with a reduction in the rate of net Ca2+ accumulation by the sarcoplasmic reticulum (SR). In the final stages, tetanic [Ca2+]iand force decline markedly due to failure of the sarcoplasmic reticulum (SR) Ca2+ release mechanism (26). During this phase, Ca2+ release can be restored by a rapid increase in cytosolic [ATP], or by application of caffeine, which increases the opening probability of the SR Ca2+ channel (1, 27). Hence, release failure may reflect desensitization of the SR Ca2+ channel, possibly due to a localized fall in [ATP]i in the triads. However, the mechanism underlying the reduction in net SR Ca2+ uptake and the progressive rise in resting [Ca2+]iobserved earlier in the fatigue process remains uncertain.
It has been shown recently that depletion of creatine phosphate (CP), which occurs rapidly after the onset of fatiguing stimulation, impairs SR Ca2+ uptake (9). This factor may contribute to the slowing of the descending phase of the Ca2+ transient and the rise in resting [Ca2+]i. Another possibility considered in previous studies is that the progressive increase in [Pi]i that accompanies CP breakdown may inhibit SR Ca2+ uptake. However, in intact mouse skeletal muscle fibers, intracellular injection of Pi increased the rate of relaxation and decreased resting [Ca2+]i (28). These effects appear to result from stimulation of the SR Ca2+ pump due to calcium phosphate (Ca-Pi) precipitation within the SR lumen. Precipitation of Ca-Piand increased Ca2+ uptake has also been reported in mechanically skinned skeletal muscle fibers (13).
Although the current experimental evidence suggests that Piis unlikely to impair SR Ca2+ regulation in fatigue, previous studies have generally involved introduction of Piin the presence of millimolar levels of cytosolic CP. However, CP can reach undetectable levels within minutes of fatiguing stimulation in single skeletal muscle fibers (18). Hence, the highest levels of Pi within the cell occur in the presence of very low [CP]. This may be of significance, because experiments on cardiac muscle have shown that Pi can induce Ca2+ efflux from the SR, which increases markedly as [CP] decreases below 2 mM (24). The possible interdependent actions of Pi and CP on skeletal muscle SR have not yet been addressed.
In this study, we have investigated the effects of Pi on SR Ca2+ regulation in mechanically skinned rat skeletal muscle fibers. The results show that Pi induces a Ca2+efflux from the SR that is independent of the Ca2+ release channel but abolished by the pump inhibitor cyclopiazonic acid (CPA). The efflux was strongly dependent on cytosolic [CP] or the activity of creatine kinase (CK). These results suggest that during fatigue as [CP] is depleted and [Pi]i increases, activation of a Ca2+ efflux pathway by Pi may contribute to the decrease in net Ca2+ uptake and the rise in resting [Ca2+].
Wistar rats (250–300 g) were killed by a blow to the head and cervical dislocation in accordance with standard UK Schedule 1 procedures. The extensor digitorium longus (EDL) was removed rapidly and placed in a relaxing solution that approximated the intracellular milieu (see Solution composition). Single muscle fibers or pairs of adjacent fibers were mechanically skinned with fine forceps and then attached between an isometric tension transducer (SensoNor, Norway) and a fixed support by means of monofilament snares (30 μm diameter) within stainless steel tubes (Goodfellow Metals, UK). In most experiments, two muscle fibers were attached in parallel to increase the amount of light collected by the objective and to improve the signal-to-noise ratio on the fluorescence records. However, qualitatively similar results were obtained on single fibers. In this study, fibers were mechanically skinned in the presence of a low [Ca2+] (40nM), which is likely to inactivate the t-tubule depolarization-induced Ca2+-release mechanism (4, 16). Consistent with this, substitution of K+with Na+ failed to release Ca2+ in these preparations.
Apparatus. The apparatus for simultaneous measurement of isometric force and SR Ca2+ release is described in detail elsewhere (7). Briefly, the mounted preparation was lowered close to the bottom of a shallow bath with a coverslip base. A Perspex column (5 mm diameter) was lowered to within a few micrometers of the muscle to minimize the volume of solution above the preparation. Throughout the experimental protocols, preparations were perfused by pumping solution at 0.8 ml/min via a narrow duct (200 μm diameter) passing through the center of the column. Waste solution was collected continuously at the column edge. The volume of solution between the coverslip and the base of the column (i.e., the effective bath volume) was ∼6 μl. The perfusing solution was exchanged with a series of valves positioned above the column. With this method, the solution within the bath could be exchanged within 10–15 s. The comparatively slow solution exchange reflects mixing of solutions in the tubing between the valves and the column. Solutions containing caffeine were rapidly applied (20 ml/min) for a 1.5-s duration via a narrow plastic tube connected to the base of the column. The higher flow rate and the smaller dead space allowed a more rapid exchange of solutions within the bath. Previous measurements based on quench of Indo-1 fluorescence by caffeine (19) under similar conditions have shown that the caffeine concentration within the bath typically increased to 50% of the concentration injected within 8–10 ms (23).
The bath was placed on the stage of a S200 Nikon Diaphot inverted microscope. The muscle was viewed via a 40× objective (Fluor, oil immersion; Nikon), and the length was increased to ∼20% above slack length. In control experiments, we have found that length does not have a direct effect on SR Ca2+ regulation. The preparation was alternately illuminated with light of wavelengths of 340 and 380 nm at 30 Hz frequency by means of a spinning wheel spectrophotometer (Cairn Research; Faversham, Kent, UK). The average [Ca2+] within the visual field containing the preparation was indicated by the ratio of light intensities emitted at >500 nm. Light emitted from areas of the field not occupied by the muscle image was reduced with a variable rectangular diaphragm on the side port of the microscope.
A detailed discussion of technical difficulties and limitations of Ca2+ measurement in skinned fibers is published elsewhere (7). One significant problem is that the dye is present in the solution surrounding the preparation as well as in the cytosol, and it is difficult to ensure that contribution of the signal from the surrounding solution remains constant from one experiment to the next. Because of this, we have opted to present the fluorescence ratio rather than converting our measurements to absolute free [Ca2+]. However, our previous measurements suggest that the peak [Ca2+] reached during the largest caffeine-induced transient is ∼700–800 nM. We have shown that the peak of the fluorescence ratio of the transient after caffeine application can be increased by rapid inhibition of the SR Ca2+ pump in amphibian and mammalian skeletal muscle fibers (7, 9). This confirms that fura 2 is not saturated during a response to caffeine under these conditions.
Solution composition. All chemicals were purchased from Sigma Chemical unless otherwise stated. The ionic composition of the solution was adjusted to maintain the [Ca2+], [Mg2+], [Na+], [K+], and pH constant. In brief, for most experiments, a basic solution was prepared containing (in mM) 100 KCl, 25 HEPES, 0.2 EGTA, 10 CP, 5 ATP, and 1.5 μM fura 2. MgCl2 and CaCl2 were added [1 M calcium chloride (BDH) stock] to produce free concentrations of 1.2 mM and 100 nM, respectively.
Both ATP and CP were added as disodium salts. In experiments where the [CP] was reduced from 10 to 0 mM, 20 mM NaCl was added to maintain the [Na+] at 30 mM. This means that the [Cl−] increased by ∼20 mM from 111.6 to 131.2 mM. However, in control experiments, increasing the [Cl−] from 112 to 132 mM at a constant [CP] had no significant effect on the caffeine-induced Ca2+ transients or the response to Pi. The total [Mg2+] in the basic solution was 6 mM. The MgCl2 was reduced by ∼0.2 mM in CP-free solution to maintain the free Mg2+ at a constant level. In solutions containing Pi, the Mg2+ was increased to compensate for binding. We used direct measurement of [Mg2+] with furaptra (4 μM) to find that an extra 0.5 mM Mg2+ was required to maintain a constant free [Mg2+] in the presence of 20 mM Pi. In all solutions, Pi was added as a K+ salt, and KCl was therefore reduced to maintain the [K+]. The KCl was further reduced to compensate for the fact that Pi-containing solutions required addition of more KOH to obtain a pH of 7.0. As a consequence, the [Cl−] was lower in Pi-containing solutions. For example, in control solutions the [Cl−] was 112 and 84 mM in solutions containing 20 mM Pi. Again, in control experiments we found that changes in Cl− with the range that occurred in this study had no significant effect on the reported results.
In further control experiments, the effects of CP withdrawal and Pi addition were studied when potassium propionate was used in place of KCl. However, the effects of Pi were the same when propionate replaced Cl− as the principal anion. Corrections for ionic strength, details of pH measurement, and the principles of the calculations are described elsewhere (10, 22). The total concentration of Na+ and K+ was 30 and 130 mM, respectively. The pH was adjusted to 7.0 by addition of KOH. In some experiments 45 U/ml CK was added as indicated. Unless otherwise stated, the free [Ca2+] concentration was adjusted to 100 nM by addition of BDH. Stock solutions of CPA (20 mM), ruthenium red (Aldrich, 10 mM) and ryanodine (Calbiochem, 100 mM) were prepared in DMSO. The final concentration of DMSO in solutions containing 20 μm CPA, 10 μM ruthenium red, and 100 μM ryanodine was 0.1%. We added 2 mM azide routinely to inhibit mitochondrial Ca2+ regulation. However, 2 mM azide or 0.1% DMSO had no detectable effect on the results obtained. All experiments were done at room temperature (22–24°C).
Data recording and analysis. Unless otherwise stated, the ratio and individual wavelength intensities and the isometric tension signals were low-pass filtered (−3 dB at 30 Hz) and digitized for later analysis with an IBM compatible 80486 computer with a Data Translation 2801A card and in-house software. Statistical analysis was done with a paired t-test.
Characteristics of Pi-induced SR Ca2+ release and dependence on CP. Figure1 A shows the 340:380 fluorescence ratio recorded from a mechanically skinned skeletal muscle fiber preparation. The preparation was continuously perfused with a weak Ca2+-buffered solution containing 1.5 μM fura 2. In this and subsequent protocols, the resting fluorescence ratio corresponds to a free-bathing [Ca2+] of 100 nM. Brief (1.5 s) application of 40 mM caffeine at 2-min intervals caused a transient increase in the fluorescence ratio due to Ca2+ release from the SR (Ca2+ transient). As in previous studies (8), increasing the loading period beyond 2 min had no significant effect on the caffeine response (data not shown). This indicates that at 100 nM [Ca2+], the SR Ca2+ content had reached a steady state within 2 min. Brief caffeine application was sufficient to produce a maximal response; i.e., a more prolonged exposure to caffeine did not increase the amplitude of the Ca2+ transient further (data not shown). Caffeine transients were highly reproducible under these conditions, and no significant decline occurred during the course of these experiments. The amplitude of caffeine-evoked Ca2+ transient was used as an index of the SR Ca2+ content under control conditions. Unless otherwise indicated, responses to Pihave been expressed relative to the caffeine-induced transient amplitude in the presence of CP. Solutions containing Piwere added after a 2-min loading period to ensure that Piwas introduced at the same SR Ca2+ content.
In the presence of 10 mM CP, addition of 5 mM Pi caused a slow transient increase in the fluorescence ratio due to Ca2+ release from the SR. Conversely, withdrawal of Pi resulted in a transient decrease in cytosolic [Ca2+]. In the presence of 10 mM CP, the mean amplitude of the Pi-induced Ca2+ release was 8.73 ± 1.03% (n = 9 ± SE) of that induced by 40 mM caffeine. However, after equilibration with solutions lacking CP, the amplitude of the Pi-induced fluorescence transient was markedly increased to 15.39 ± 1.09% (n = 14 ± SE) of the caffeine-induced response. The increased amplitude of the Pi-induced transient occurred despite the fact that the steady-state Ca2+ content of the SR is lower in the absence of CP (9). The Ca2+ transients induced by 1–40 mM Pi did not cause measurable Ca2+ activated force. This in part reflects the fact that Pi directly inhibits force production by the myofilaments (14).
Figure 1 B shows cumulative data obtained by means of this protocol. The ordinate represents the mean amplitude of the Pi-induced Ca2+ efflux expressed as a percentage of the steady-state caffeine response obtained under control conditions in the presence and absence of CP. Increasing the [Pi] from 1 to 40 mM resulted in a concentration-dependent increase in the amplitude of the Pi-induced fluorescence transient in the presence or absence of CP. However, in the presence of 10 mM CP the transient amplitude was reduced by ∼50% at all [Pi] tested. These results suggest that the mechanism of Pi-induced Ca2+ efflux is dependent on the cytosolic [CP]. Also shown is the amplitude of the fluorescence transient in the presence of 10 mM CP and 45 U/ml CK added to the bathing solution. In the presence of CK, the efflux induced by 1–20 mM Pi was further reduced compared with that in the presence of 10 mM CP alone.
Superimposed fluorescence transients induced by addition of 2, 10, and 40 mM Pi in the absence of CP are shown in Fig.2 A. All responses were from the same preparation. The time to the peak of the fluorescence transient decreased slightly as the [Pi] increased. However, unexpectedly, the descending phase of the transient was markedly abbreviated as the [Pi] increased (seediscussion). Cumulative data showing the time to peak of the Pi-induced Ca2+ transient at a range of [Pi] is shown in Fig. 2 B. Cumulative data showing the time for the descending phase of the transient efflux to decay by 50% at a range of [Pi] is shown in Fig. 2 C.
Effects of 1,2-dinitrofluorobenzene on Pi-induced Ca2+ release. The effect of CP removal was compared with inhibition of CK by 1,2-dinitrofluorobenzene (DNFB; Fig.3). In the presence of 10 mM CP, steady-state caffeine-induced responses were obtained at 2-min intervals (3A). Repeated caffeine application was then stopped, and 10 mM Pi was introduced after a further 2-min loading period. This resulted in a characteristic Ca2+ efflux, which peaked at 8.32 ± 0.50% (n =5 ± SE) of the caffeine response. As shown previously, in the continuous presence of 10 mM CP, exposure to 20 μM DNFB results in prolongation of the caffeine-induced Ca2+ transient and a reduction in amplitude (9). These effects also occur when CP is withdrawn, suggesting that the efficiency of the SR Ca2+ pump is dependent on local regeneration of ATP via CK (9). After ∼10 min exposure to 20 μM DNFB, 10 mM Pi was reapplied. This resulted in a larger efflux of Ca2+ from the SR, which peaked at 13.3 ± 1.0% (n = 5 ± SE) of the caffeine response obtained under control conditions, in the absence of DNFB. In control experiments, removal of CP had no significant effect on Pi-induced Ca2+ efflux after 20 μM DNFB treatment (n = 3). Furthermore, DNFB did not alter the Pi-induced efflux in the absence of CP (n = 3). As previously reported, the effects of DNFB were irreversible within 30 min (9).
Effects of ruthenium red and ryanodine on Pi-induced Ca2+ release. Previous studies have suggested that Pi may directly activate the SR Ca2+ channel (12). This possibility was investigated by exposing muscle fibers to the SR Ca2+ channel blockers ruthenium red and ryanodine. Figure4 A shows Pi-induced Ca2+ efflux and caffeine-induced Ca2+transients obtained in a skinned EDL preparation under control conditions. After exposure to 20 μM ruthenium for ∼10 min, 40 mM caffeine failed to elicit a Ca2+ transient, whereas Pi-induced Ca2+ efflux was apparently unaffected. Figure 4 B shows data from another preparation, where caffeine-induced Ca2+ release was almost completely abolished by 100 μM ryanodine, but Pi-induced Ca2+ release was again unaffected. Similar results were obtained in eight other preparations (4 with ryanodine, 4 with ruthenium red). In both experimental protocols, subsequent exposure to the nonionic detergent Triton X-100 resulted in a large prolonged Ca2+ release due to disruption of the SR membrane. After Triton treatment, introduction of Pi failed to influence the fluorescence ratio (data not shown). These results suggest that Pi releases Ca2+ from the SR via a mechanism that does not involve ruthenium red or ryanodine-sensitive SR Ca2+ channels.
Effects of CPA on Pi-induced Ca2+ efflux. The results shown in Fig. 4suggest that Pi does not induce Ca2+ efflux via the SR Ca2+ channel. An alternative possibility is that Pi-induced Ca2+ efflux occurs via the SR Ca2+ pump (3). Involvement of pump reversal in Pi-induced Ca2+ efflux was investigated by exposing the preparations to the pump inhibitor CPA, which blocks forward and reverse modes of the SR Ca2+ pump (7, 20). Figure 5 shows control caffeine-induced transients obtained at 2-min intervals (Fig. 5, left) and a transient induced by addition of 20mM Pi (Fig. 5,middle) 2 min after the preceding caffeine response. Pi was then washed out for 10 min before introduction of 20 μM CPA (Fig. 5, right). As shown previously, introduction of CPA was associated with a slow transient increase in the fluorescence ratio due to net Ca2+ release from the SR (7). The transient rise in Ca2+ probably occurs because pump inhibition reveals Ca2+ efflux via a Ca2+ leak pathway. After 8 min exposure to CPA, introduction of 20 mM Pi failed to induce a detectable release of Ca2+. However, reapplication of caffeine induced a substantial Ca2+ transient that was slightly greater than conditions under control. This suggests that failure of Pito induce Ca2+ efflux was not due to full depletion of the SR Ca2+ content. After CPA treatment, the descending phase of the caffeine-induced Ca2+ transient was markedly slowed, suggesting that the SR Ca2+ pump contributes to the decline in Ca2+ after brief caffeine application (7). After a further 2 min, reapplication of caffeine failed to release Ca2+. This was presumably due to the reduced ability of the SR to reaccumulate Ca2+ in the intervening period.
One difficulty in interpreting the results shown in Fig. 5 is that CPA induces some loss of Ca2+ from the SR, and it is possible that this factor alone might reduce the sensitivity to Pi. Therefore, the effect of CPA on Pi-induced Ca2+release was investigated further by exposing the fibers to a lower [CPA], which did not result in any detectable loss of Ca2+ from the SR. Figure6 A shows a Pi-induced Ca2+ efflux and caffeine-induced Ca2+transients obtained under control conditions. The amplitude of the transient induced by 10 mM Pi was ∼20% of the caffeine response obtained in the absence of CP. Control responses to caffeine applied at 2-min intervals are shown in Fig. 6 A, top middle. In the presence of 0.5 μM CPA, there was no detectable rise in resting fluorescence (data not shown, n = 13), and the amplitude of steady-state caffeine-induced Ca2+ transients did not alter significantly (Fig.6 A, top right). This suggests that the SR Ca2+ content was not significantly affected by 0.5 μM CPA. However, the amplitude of the Pi-induced Ca efflux was reduced by almost 50% in the presence of 0.5 μM CPA (Fig.6 A, bottom left). Pi-induced fluorescence transients in the presence and absence of 0.5 μM CPA are shown superimposed on an expanded scale (Fig. 6 A, bottom right).
The accumulated data (Fig. 6 B) shows the effect of 0.5 μM CPA on release induced by 5 and 10 mM Pi. The amplitude of the transient induced by 10 mM Pi was 20.40 ± 0.92% (n = 7 ± SE) of the caffeine response obtained in the absence of CP. The amplitude of the Pi-induced Ca efflux was markedly reduced to 11.00 ± 0.61% (n = 7 ± SE) of the caffeine response in the presence of 0.5 μM CPA. The amplitude of the Ca2+ transient induced by 5 mM Pi was smaller than that after addition of 10 mM Pi. However, the proportional reduction in the amplitude of the transient in the presence of CPA was similar.
Effects of ADP on SR Ca2+ regulation.If Pi induces Ca2+ release via pump reversal, then this effect should be mimicked by ADP, which is also known to reverse the SR Ca2+ pump (5, 6). Figure7 A shows control caffeine-induced transients obtained at 2-min intervals. Introduction of 2 mM ADP was associated with a release of Ca2+ from the SR, which peaked at ∼25% of the caffeine-induced Ca2+ transient. Again, introduction of a high concentration of CPA (20 μM) caused a slow prolonged Ca2+ efflux from the SR. As with Pi(Fig. 5), ADP-induced Ca2+ efflux was abolished after exposure to 20 μM CPA, whereas subsequent application of caffeine induced a large Ca2+ transient. The efflux caused by ADP was also unaffected by ryanodine or ruthenium red (data not shown), suggesting the ADP-induced Ca2+ efflux is similarly independent of the SR Ca2+ channel. No effect of ADP was observed after disruption of the SR membrane with Triton X-100 (data not shown). Similar results were obtained in four other preparations.
Figure 7 B shows superimposed fluorescence transients obtained after addition of 2 and 10 mM ADP. As with Pi (Fig. 2), both the rising and falling phases of the transients were more rapid at 10 mM ADP.
Characteristics of Pi-induced Ca2+ release and dependence on CP or CK activity. The present study has shown that Pi induces a net efflux of Ca2+ from the SR in rat EDL muscle fibers (Fig. 1). Pi-induced Ca2+ efflux was markedly increased in the absence of CP (Fig. 1 A). This occurs despite the fact that the SR Ca2+ content decreases in the absence of CP due to reduced efficiency of the SR Ca2+ pump (9). The effect of CP withdrawal was mimicked by treatment with the CK inhibitor DNFB (Fig. 3). The dependence of Pi-induced Ca2+ release on CP and CK activity is similar to that reported in permeabilized cardiac trabeculae (24).
A likely explanation for the effects reported in this study is that Pi induces Ca2+ efflux by reversal of the SR Ca2+ ATPase. This phenomenon was first characterized in isolated SR vesicles, where millimolar levels of Pi can induce a rapid Ca2+ efflux from the SR by reversal of the Ca pump (3), which is accompanied by synthesis of ATP from ADP (5, 17). In SR vesicles, pump reversal requires both micromolar levels of ADP and a [Ca2+] gradient across the SR membrane, which provides the energy necessary for ATP synthesis (15).
In skinned muscle fibers, therefore, the transient increase in [Ca2+] on addition of Pi may reflect the dual action of Pi and ADP, which together induce a net Ca2+ efflux via the SR Ca pump. ADP is continually produced by cellular ATPase, and micromolar levels may also be present in the ATP added to the solution. The requirement for ADP may explain the dependence of Pi-induced Ca2+release on cytosolic CP or the activity of CK. In the presence of CP, local ADP levels will be low due to the efficient rephosphorylation of ADP to ATP at sites where endogenous CK is bound within the cell. In the absence of CP, the local [ADP] will equal the contamination level, plus that produced by cellular ATPase. When exogenous CK is added to the solution, Pi-induced Ca2+ release was further reduced (Fig. 1 B). This may be explained if exogenous CK reduces the contamination levels of ADP in the bulk solution and increases the ability of CP to buffer ATP within the preparation.
Time course of Pi-induced Ca2+release. As shown in Fig. 2, the time course of the Pi-induced Ca2+ transient is dependent on the [Pi]. Increasing the [Pi] from 2 to 40 mM was associated with a progressive increase in the amplitude of the resulting Ca2+transient. The time to peak of the Pi-induced Ca2+ efflux also decreased as the [Pi] increased. On removal of Pi, a characteristic undershoot in [Ca2+] occurred within the muscle (e.g., Figs. 1 A and 3 A).
The fact that the amplitude increases and peaks earlier as the [Pi] increases may suggest a simple concentration dependence of the Pi-induced Ca2+efflux or a more rapid rise to a threshold [Pi]. The undershoot may be explained if addition of Pi results in maintained activation of a Ca2+ efflux pathway. In the constant presence of Pi, the SR Ca2+ content will fall until a new steady state is reached and the total Ca2+ efflux is balanced by Ca2+ reuptake via the pump. On withdrawal of Pi, the Pi-induced Ca2+ efflux pathway will inactivate, resulting in net Ca2+ uptake. As this happens rapidly, the SR is transiently able to reduce the [Ca2+] inside the preparation, resulting in the characteristic undershoot. This interpretation is supported by previous work on cardiac muscle, where we have shown that after complete abolition of SR Ca2+ uptake by ATP withdrawal, Pi-induced Ca2+ efflux persists, but the undershoot is absent on Pi withdrawal (25).
Another consistent feature was that the descending phase of the transient was more rapid as the [Pi] increased (Fig. 2). One possible explanation for this is that release is abbreviated by Pi entry into the SR, followed by precipitation of Ca-Pi (13). This would result in a rapid decline in the free luminal [Ca2+] to a level dictated by the Ca-Pi solubility product. Such an effect should also be more pronounced at higher [Pi]. However, the declining phase of the transient was also more rapid after addition of 10 mM ADP than with 2 mM ADP (Fig. 7 B). Although this does not exclude the possibility that precipitation may contribute to the more rapid response at 40 mM Pi, it does suggest that other factors contribute to this effect. One possibility is that application of a higher [Pi] may induce a rapid release of Ca2+ and a rapid establishment of a steady state, whereby the increased Ca2+ leak is balanced by reuptake. In contrast, lower levels of Pi may release a similar amount of Ca2+, but over a longer period of time. However, a more complete understanding of the kinetics of the response requires detailed knowledge of 1) the influence of SR Ca2+ reuptake on the transient and 2) the absolute decrease in SR Ca2+ content that occurs in the presence of Pi. This is complicated by the fact that pump inhibitors such as CPA inhibit the response and by the possible effects of Pi on the SR Ca2+ release mechanism (seeLack of evidence for Pi-induced SR Ca2+ channel activation).
Effects of CPA on Pi-induced Ca2+ efflux. The suggestion that Pi-induced Ca2+ efflux may reflect reversal of the Ca2+ pump is further supported by the finding that release could be abolished completely by pretreatment with 20 μM CPA (Fig. 5), which has been shown to inhibit both forward and reverse modes of the ATPase (6, 20).
Introduction of 20 μM CPA was itself associated with loss of Ca2+ from the SR. This appears to reflect a passive efflux of Ca2+ via a leak pathway, which is revealed after rapid inhibition of the Ca2+ pump by CPA (7). Despite this Ca2+ loss, subsequent application of caffeine induced a large Ca2+ transient of similar amplitude to that obtained under control conditions (Fig. 5). This confirms that the ability of CPA to block Pi-induced Ca2+ efflux does not result from complete depletion of SR Ca2+. However, the apparently similar amplitude of the SR caffeine-induced Ca2+ transient after CPA treatment occurs despite the fact that the SR Ca2+ content has decreased. This is because under control conditions the amplitude of the caffeine-induced Ca2+ transient is attenuated by ∼20–30% due to rapid Ca2+ reuptake by the SR Ca2+ pump (7). Thus when the SR Ca2+ pump is inhibited by high levels of CPA, any given caffeine-induced Ca2+ transient amplitude reflects a lower Ca2+ content.
The loss of Ca2+ after introduction of CPA complicates interpretation of the data. Recent work suggests that Ca2+within the SR lumen affects the sensitivity of the Ca2+channel (21). It is therefore possible that CPA might inhibit Pi-induced Ca2+ release indirectly by desensitization of the SR Ca2+ channel, rather than by a direct effect on pump reversal. However, in further experiments it was found that lower concentrations of CPA (0.5 μM), which did not induce a detectable efflux of Ca2+ from the SR or influence the caffeine-induced Ca2+ transient, markedly inhibited Pi-induced Ca2+ release (Fig. 6). This suggests that CPA inhibits Pi-induced Ca2+ release by preventing pump reversal, rather than by causing an indirect effect via changes in SR Ca2+ content. It also suggests that the forward and reverse reactions of the SR Ca2+ pump may differ in sensitivity to CPA. This is an interesting finding, because it may be possible to use low levels of CPA or other drugs to prevent pump reversal occurring in intact cells under conditions of fatigue without major effects on Ca2+ uptake.
Lack of evidence for Pi-induced SR Ca2+ channel activation. It has been reported that Pi (3–30 mM) produces a concentration-dependent increase in ryanodine binding in SR membrane fractions from porcine skeletal muscle and increases the open probability of the isolated channel (12). However, in the present study in mechanically skinned skeletal muscle fibers, the efflux of Ca2+ observed on addition of Pi was not inhibited by the Ca2+ channel blockers ruthenium red or ryanodine, whereas caffeine-induced Ca2+ release was effectively abolished (Fig. 4). This strongly suggests that the Pi-activated Ca2+ efflux pathway reported in the present study is independent of the SR Ca2+ channel.
The apparent differences between results obtained on skinned fibers and isolated channels or membrane vesicles may reflect the experimental conditions. For example, the study by Fruen et al. (12) appears to have been carried out in the absence of ATP, which markedly increases open probability of the SR Ca2+ channel. It is possible that the relative influence of Pi may be reduced in the presence of millimolar levels of ATP or other physiological channel activators. Alternatively, the lack of effect of Pi on the Ca2+ channel in skinned fibers may result from the Ca2+-dependence of channel activation by Pi. In isolated vesicles, stimulation of ryanodine binding was small at 100 nM Ca2+, which corresponds to the resting [Ca2+] in this study (12). Ryanodine binding was increased most by Pi when [Ca2+] was in the micromolar range. Therefore, Pi might have an enhancing effect on release when the [Ca2+] rises markedly after caffeine application or t-tubule depolarization, but not at resting [Ca2+]. However, in preliminary work we have found evidence that Pi inhibits the SR Ca2+release process in skinned fibers (data not shown). This unexpected effect also complicates assessment of the absolute reduction in SR content in the presence of Pi and is the subject of work in progress.
Relevance to events during skeletal muscle fatigue. Previous studies on intact or skinned skeletal muscle fibers have not provided evidence that Pi impairs Ca2+ transport by the SR. Indeed, it has been reported that increased cytosolic levels of Pi 1) accelerate the rate of decline of the Ca2+ transient after tetanus in intact fibers and2) increase the Ca2+ capacity of the SR in skinned cells (13, 28). This enhancing effect on uptake has been ascribed to precipitation of Ca-Pi within the SR lumen and a consequent increase in Ca2+ pump activity. However, previous studies have generally been carried out in the presence of millimolar levels of cytosolic CP. As shown in Fig. 1, Pi-induced Ca2+ efflux increases markedly as the [CP] declines. Therefore, during fatigue as [CP]i is depleted and [Pi]i rises, the associated Ca2+ efflux may progressively reduce the maximum Ca2+ capacity of the SR, thereby contributing to the reported increase in resting [Ca2+] (2). The Pi-induced reduction in SR Ca2+ content that occurs at low levels of CP will also reduce the probability of Ca-Pi precipitation within the SR.
In summary, the present study shows that millimolar Piinduces a marked concentration-dependent Ca2+ efflux from the SR. Pi-induced Ca2+ efflux was inhibited by CP, insensitive to ryanodine or ruthenium red, and effectively abolished after pretreatment with CPA. These effects are consistent with reversal of the SR Ca2+ pump by Pi. In intact fibers, activation of a Ca2+ efflux pathway by Pi may occur during fatiguing stimulation as CP is depleted. This would be expected to decrease the maximum Ca2+ capacity of the SR and may increase resting [Ca2+]i.
Financial support for this work was provided by the Welcome Trust and the British Heart Foundation.
Address for reprint requests and other correspondence: D. S. Steele, School of Biology, Univ. of Leeds, Leeds LS2 9JT, United Kingdom (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society