Measurement of Ca2+ stores during fatigue
Kabbara & Allen (1999) showed that under control conditions brief applications of 4-CmC or caffeine caused a large rise in [Ca2+]i which did not appear to damage the muscle and was repeatable. They argued that the pool from which this Ca2+ was being released was completely depleted on the grounds that application of a SR Ca2+ pump blocker did not increase the magnitude of the myoplasmic [Ca2+]i. However, calculation suggested that this pool was less than the total SR Ca2+. It is generally accepted that most of the Ca2+ within the SR (total Ca2+ 30–50 mM) is bound to calsequestrin, with only ∼1 mM remaining free (Somlyo et al. 1981). Since the Ca2+ released from the SR by 4-CmC approximates to half the Ca2+ within the SR (Somlyo et al. 1981; Kabbara & Allen, 1998), this implies that the Ca2+ bound to calsequestrin must be capable of dissociating relatively quickly. Note that the opposite is also true; if Ca2+ bound to some species within the SR from which it did not dissociate during the time course of a tetanus or a 4-CmC exposure, then this component of Ca2+ would not be measured by this test. For these reasons we refer to the 4-CmC-induced Ca2+ rise as measuring the rapidly releasable SR Ca2+.
A preliminary question is whether 4-CmC is still capable of opening the SR Ca2+ channels sufficiently during fatigue to enable the rapidly releasable SR Ca2+ to be measured reliably. An important observation by Kanaya et al. (1983) was that the sensitivity of contracture tension to caffeine was reduced considerably in fatigued muscles (the half-maximal sensitivity increased from ∼3 mM to ∼7 mM). This could arise either because the SR Ca2+ release channels in fatigued muscles had become less sensitive to caffeine or because the contractile proteins had a reduced Ca2+ sensitivity so that a greater Ca2+ release was required to produce the same tension. We previously showed that under control conditions 2 mM 4-CmC gave a just-maximal Ca2+ release and that 5 mM 4-CmC gave the same maximal release (Kabbara & Allen, 1999). In the present study we have shown that in fatigued conditions the rapidly releasable SR Ca2+ as a fraction of the control value was 53 ± 2 % for 2 mM 4-CmC and 45 ± 5 % for 5-CmC. Since these values are not significantly different it follows that the sensitivity of the SR release channels to 4-CmC is not reduced in fatigue. Furthermore it is known that the SR Ca2+ pump is inhibited during fatigue (Byrd, 1992; Westerblad & Allen, 1993) so that for a given increase in SR Ca2+ channel permeability the fraction of Ca2+ in the myoplasm compared with the SR will be greater. We conclude that rapid application of 4-CmC can effectively measure the rapidly releasable SR Ca2+ during fatigue as well as under control conditions.
The cause of the reduction of rapidly releasable SR Ca2+ during fatigue
It is clear that the rise of [Ca2+]i caused by 4-CmC declines during fatigue and returns to control levels during the period when the tetanic [Ca2+]i and tension recover. We argue above that this is not because the Ca2+ release channels become less sensitive to 4-CmC during fatigue and therefore this observation can be interpreted either as a reduction of the total amount of Ca2+ stored in the SR or, alternatively, Ca2+ binds to some substance within the SR to form a product from which Ca2+ cannot be released during a short 4-CmC exposure or a tetanus.
To test the possibility that Ca2+ stored in the SR is reduced we made use of an observation by Balnave & Allen (1998). They showed that during a prolonged period (20–30 min) of elevated [Ca2+]i, Ca2+ was removed from the cell on the Na+-Ca2+ exchanger. At the end of this period, the SR Ca2+ store was depleted and could be replenished by a 30 min period in normal extracellular [Ca2+]. However, if extracellular [Ca2+] was removed during the 30 min recovery period then the store was not replenished. The present results show that the reduced 4-CmC-releasable Ca2+ observed in fatigue recovers even in the absence of extracellular calcium. This result suggests that the SR Ca2+ store is not depleted, but instead that the Ca2+ is present but not able to be released and that this process can reverse during recovery of the muscle.
Our experiments provide strong evidence that the reduction in releasable Ca2+ is in some way related to the metabolic changes associated with fatigue. First, when the same number of tetani was performed with longer intervals between tetani so that no fatigue occurs, then no such reduction in releasable Ca2+ was observed. Second, the releasable Ca2+ recovered during a 20 min rest period, which is sufficient for many of the metabolic changes associated with fatigue to be reversed (Kushmerick & Meyer, 1985; Cady et al. 1989). Third, when oxidative phosphorylation was blocked during the recovery period, no such recovery of releasable Ca2+ occurred. It is known that preventing oxidative phosphorylation greatly reduces the degree of metabolic recovery in a resting muscle (Dawson et al. 1980). In particular the myoplasmic Pi remains elevated when the muscles are kept anaerobic during recovery.
The CaPi hypothesis
The present experiments provide no direct evidence as to the nature of the metabolic change which leads to a reduction of rapidly releasable SR Ca2+. However, our data are consistent with the hypothesis that Ca2+ precipitates with Pi within the SR, forming a species of CaPi which redissolves slowly.
This hypothesis was first proposed by Fryer et al. (1995) based on studies of Ca2+ release in skinned rat and cane toad fibres with intact SR (Owen et al. 1992). They showed that, in the presence of high Pi concentrations and a finite pool of Ca2+, Ca2+ release from the SR was reduced in a manner consistent with precipitation of CaPi in the SR. In subsequent studies Fryer et al. (1997) investigated the mechanism of Pi uptake by the SR and showed a delay of several minutes between the rise of Pi and the development of CaPi in the SR. Posterino & Fryer (1998) measured the time course of recovery of SR Ca2+ release when [Pi] was reduced from a high to a low level. This process had a half-time of ∼35 s which they attributed to Pi removal from the SR. All the above studies were in skinned fibres but further support for the development of CaPi in the SR came from experiments in which NaPi was injected into intact fibres. This produced a substantial reduction in Ca2+ release and tension production consistent with the above hypothesis (Westerblad & Allen, 1996). The above studies strongly suggest that precipitation of CaPi in the SR can occur under some conditions but they do not show whether this mechanism operates during fatigue.
The present experiments make the case that this mechanism operates during fatigue much stronger. The key items of evidence, discussed above, are as follows. (i) Releasable SR Ca2+ is reduced in fatigue and returns to control values during recovery. (ii) This phenomenon is closely associated with the metabolic change occurring during fatigue.
Interestingly, the early electron probe studies by Gonzalez-Serratos et al. (1978) showed not only a rise in the SR total Ca2+ but also a small rise in the phosphorus content which would be consistent with increased CaPi. However, it is important to note that there is, as yet, no direct evidence of precipitation of CaPi in the SR of fatigued muscles to our knowledge.
Westerblad & Allen (1996) pointed out two weaknesses of the CaPi hypothesis as an explanation for the reduced Ca2+ release during fatigue. (i). The reduction of Ca2+ release occurs mainly in the last phase of fatigue and correlates better with the fall in [ATP] and the rise of [Mg2+]i which occur at about that time (Westerblad & Allen, 1992). (ii). In the final phase of fatigue there are at least three changes in Ca2+ handling; reduced tetanic [Ca2+]i, elevated resting [Ca2+]i, and slowing of the SR Ca2+ pump. The CaPi hypothesis only explains the first of these. How can these points be addressed?
The time course issue remains a difficult and complex one. First, the time course of the decline of tetanic [Ca2+]i in amphibia (present experiments; Allen et al. 1989) is somewhat different to those reported in mouse (Westerblad & Allen, 1991, 1993) in showing a more or less monotonic decline after the initial small rise, which is more simply consistent with the CaPi hypothesis. Second, a fuller understanding of the time course requires knowledge of the time course of the rise of myoplasmic Pi, the time course of the entry of Pi into the SR and the time course of precipitation of CaPi. We need more information on all these processes to decide whether the time course of decline of releasable SR Ca2+ is consistent with the theory. Third, Posterino & Fryer (1998) make the interesting proposal that there may be an interaction between elevated Pi and reduced ATP. Similar considerations apply to the time course of recovery of releasable SR Ca2+.
It remains true that the CaPi hypothesis does not explain the elevated resting Ca2+ and the slowing of the SR pump (see Westerblad & Allen, 1996, for more detailed discussion). These two effects can be explained by metabolic inhibition of the SR pump which we suggest must be an independent contributor to fatigue, though possibly also triggered by raised myoplasmic Pi.
These findings are all consistent with the decline of Ca2+ release during fatigue being caused by precipitation of CaPi in the SR which reverses during recovery. A major attraction of this hypothesis is that it leads to a simple explanation of the correlation between failure of activation and the metabolic changes in fatigue. It is not yet clear whether other mechanisms, such as failure of action potentials or reduced opening of SR Ca2+ channels, also contribute to the decline in Ca2+ release which occurs in fatigue. The differences in time course of fatigue and the sensitivity to glucose during recovery in mammalian muscle may indicate that additional mechanisms operate in mammalian tissues.