The functional interaction between DHPR voltage sensor charge movement and RyR mediated SR Ca2+ release serves as the fundamental trigger of skeletal muscle EC coupling, and ultimately force generation. Here, we demonstrate for the first time in mammalian muscle fibres two clearly and reproducibly observed components of intra-membrane charge movement, monitored simultaneously with high temporal resolution characterization of SR Ca2+ release. We show a clear correlation between Ca2+ release and the Qγ component of charge movement, and provide evidence that the Qγ component is a consequence of release. Furthermore, we demonstrate decreased amplitude of the rate of Ca2+ release in fibres lacking S100A1, but an essentially identical relative time course and voltage dependence of release, when compared to WT fibres. We propose that this depressed release is the underlying cause for the suppression of Qγ in KO fibres (Prosser et al. 2009). This hypothesis is strengthened by the finding that potentiating Ca2+ release with 4–CMC partially restores Qγ in KO fibres.
Qγ: cause vs. consequence of SR Ca2+ release
Based on studies conducted primarily in amphibian muscle fibres, Qγ has been argued to be a consequence (and perhaps an additional cause) of Ca2+ release from the SR. This was demonstrated by interfering with Ca2+ release through various methods, such as voltage-dependent inactivation of RyR1, depletion of SR Ca2+ stores, or pharmacological inhibition of RyR1, and monitoring the resulting effects on the hump component of charge movement currents in amphibian fibres (Csernoch et al. 1991; García et al. 1991; Pizarro et al. 1991; Gonzalez & Rios, 1993). These authors proposed that released Ca2+ ions act as a local positive feedback signal at the t-tubule membrane, driving further charge movement and possibly greater release. Therefore, interventions that suppress release also suppress a temporally delayed component of intra-membrane charge movement.
Our results in mammalian fibres generally support this model, as evidenced by the following. (1) S100A1 KO fibres demonstrate decreased peak Ca2+ release, but a remarkably similar time course and voltage dependence of release (Figs 4 and 5) compared to WT fibres. As Qγ is a steeply voltage dependent and temporally delayed component of charge movement, if Qγ were driving release we would expect to see alterations in the time course and voltage dependence of release in KO fibres that lack Qγ. In contrast, if the depressed amplitude of release in KO fibres is altering charge movement, we would expect to see differences in charge movement current kinetics and total charge moved only at voltages where there is substantial release, as is demonstrated in Figs 4D and 7. It should be noted that despite the very similar relative time course of release detected here, we cannot rule out the possibility that limitations in clamp speed or signal to noise may prevent the detection of very subtle changes in release time course between WT and KO fibres. (2) As illustrated in Fig. 7, there is considerable charge development prior to the rise of detectable release, and Qγ begins to develop in time with release. Additionally, there is a close temporal correlation between the peak of Ca2+ release and the peak of Qγ difference current. If Qγ were driving further Ca2+ release, KO fibres that lack Qγ may be expected to demonstrate an earlier release peak than WT fibres that show Qγ. Instead, KO and WT fibres show an essentially simultaneous peak of release, suggesting that the increased amplitude of Ca2+ release in WT fibres is driving this temporally delayed charge component. (3) Interventions that specifically target RyR1 mediated Ca2+ release and bypass the voltage sensor alter Qγ, supporting the concept of Qγ as a consequence of release. Specifically, potentiating release at intermediate voltages with 4–CMC restores a temporally delayed component of charge movement in KO fibres. As demonstrated in the accompanying paper (Prosser et al. 2009), the application of dantrolene, an RyR1 inhibitor, to WT fibres suppresses a similar temporally delayed component of charge movement current. Furthermore, S100A1 KO fibres exhibit depressed Ca2+ release, detailed here in voltage clamp studies, but originally S100A1 was shown to directly affect ligand activated RyR1-mediated Ca2+ release, in the absence of voltage sensor activation (Fano et al. 1989; Treves et al. 1997; Most et al. 2003). Therefore it seems likely that enhancement of release in the presence of S100A1 is driving the additional charge component seen in WT, but not in S100A1 KO, muscle fibres. Our findings in general, therefore, support the concept of local Ca2+ release driving additional intra-membrane charge movement, or, simply re-phrased, Qγ being a consequence of SR Ca2+ release.
However, there is a notable dissimilarity between the difference component isolated by pharmacological manipulations of release, specifically application of 4–CMC to KO fibres or application of dantrolene to WT fibres, and the difference component isolated between WT and KO fibres. 4–CMC produced a leftward shift in the voltage dependence of release, augmenting release and ‘rescuing’Qγ in KO fibres only at intermediate depolarizations. This suggests that 4–CMC sensitized the RyR Ca2+ release channels to activation by the T-tubule voltage sensor, and thereby potentiated release to a sufficient extent at these intermediate voltages to surpass the threshold of release required to drive Qγ. However, at maximal depolarizations (+20 mV), there was no significant enhancement of release by 4–CMC, and subsequently no additional charge moved, in these treated KO fibres. In contrast, WT fibres exhibited greater release than KO fibres at every voltage, and correspondingly moved additional charge at every voltage (Figs 8 and 9C). This suggests two different and possibly independent mechanisms for the enhancement of release by 4–CMC and S100A1. 4–CMC appears to increase the sensitivity of the RyR Ca2+ release channels to activation. Consequently, 4–CMC improves the efficiency of coupling between the voltage sensor Qβ component and release channel activation at sub-maximal activation in KO fibres, without altering the effectiveness of the Qβ component of charge movement at maximal depolarizations, where presumably all available RyRs are already activated in the absence of 4–CMC (Fig. 9A). S100A1, on the other hand, appears to augment release without affecting the coupling efficiency of the release channels to Qβ. One possible mechanism could be that S100A1 increases the amount of Ca2+ released by an active RyR channel, with no change in activation kinetics, by increasing channel open time (or increasing single channel Ca2+ current), so that global Ca2+ release is simply scaled up without changing kinetics in WT compared to S100A1 KO fibres. Additionally, we cannot rule out the possibility that S100A1 may have a secondary effect on the voltage sensor, besides simply potentiating RyR1 Ca2+ release, which could further alter charge movement. However, the fact that atypical S100A1 KO fibres, which demonstrate ‘WT’ rates of Ca2+ release, exhibit definite Qγ and a similar voltage dependence of charge movement, argues against this possibility.
Mechanisms for generation of Qγ and comparison with previous models
The fact that the total charge moved at large depolarizations in S100A1 containing fibres that exhibit greater Ca2+ release compared to KO fibres indicates that the extra release allows movement of a separate component of intra-membrane charge that is not moved in the S100A1 KO fibres. In contrast, adding 4–CMC to KO fibres causes an increase in charge movement at intermediate, but not at large, depolarizations. Thus, this extra charge movement at intermediate voltages could be generated by the same charged groups that move at larger depolarizations in KO fibres, but now moving at intermediate depolarizations in the presence of 4–CMC. This mechanism is similar to that proposed previously for Qγ in amphibian fibres (Csernoch et al. 1991; García et al. 1991; Pizarro et al. 1991; Gonzalez & Rios, 1993). However, our finding that the total charge moved at large depolarizations is greater in WT than KO fibres is not consistent with the model for generating Qγ by a shift in the voltage range of movement of a single group of charges. Instead, our results suggest that, in WT fibres, Qγ may arise from a separate set of voltage sensors. The function of this charge movement component remains to be determined.
Our results with 4–CMC potentiation of Ca2+ release in S100A1 KO fibres raise interesting questions about the requirements for the development of Qγ. In KO fibres, increasing release by treatment with 4–CMC leads to the generation of Qγ (Figs 8 and 9), but a similar increase in release produced by further depolarization of KO fibres in the absence of 4–CMC does not. If Qγ were purely a consequence of the overall rate of release, we might expect that the Qγ charge would just be right shifted along the voltage axis in KO compared to WT fibres, a prediction not supported by our results. This suggests that an additional factor, other than simply a necessary overall rate of release, is required for the generation of Qγ. This additional factor could be the release per activated channel, rather than the overall release from all activated channels, implying a possible regulation of Qγ by the local Ca2+ in close proximity of the activated RyR channels. However, in that case 4–CMC would have to increase the Ca2+ release per activated channel at the intermediate voltage range, but not increase single channel Ca2+ release at the larger depolarizations, in addition to sensitizing the RyR channel to activation by the DHPR. Another possibility, to be tested in future studies, is whether the peak rate of release must be sufficient to match the state or conformation of the voltage sensor in order to produce Qγ. In this scenario, the altered state of the voltage sensor with further depolarization (perhaps the increasing kinetics of charge movement with increasing voltage) continuously raises the threshold for the peak rate of release required to trigger Qγ. KO fibres with low rates of release simply never are able to catch up to this threshold.
Ca2+ transients and calculated rate of Ca2+ release: comparison with previous reports
In our studies we utilized an internal solution in the whole cell patch pipette containing 20 mm EGTA with no added Ca2+, which should have resulted in near zero free Ca2+ in the resting muscle fibres 20 min after establishing the whole cell recording configuration, when our voltage clamp data acquisition began. The high intra-fibre [EGTA] also buffers the released Ca2+ at a very low level. Under these conditions, we used fluo-4 (50 μm) in the pipette to record the fluorescence signals during voltage clamp depolarizations, calculated the Ca2+ transient from the ΔF signal assuming resting Ca2+ prior to each pulse to be zero, and then calculated the rate of Ca2+ release to equal the time course of the rate of change of [Ca-EGTA]. The Ca2+ release waveforms recorded under these conditions generally resembled release records calculated previously, exhibiting a rapid rise to peak, followed first by a rapid decline attributed to Ca2+ dependent inactivation of the RyR Ca2+ release channel, and then a slower decline (Melzer et al. 1984; Delbono & Stefani, 1993). However, the slower decline seen here was considerably more rapid than recently reported by another laboratory, also using high EGTA in voltage clamped mouse FDB fibres, but with resting Ca2+ buffered at about 0.1 μm (Royer et al. 2008). The ΔF records in that paper were generally maintained during a 100 ms depolarization, whereas here the ΔF signals clearly declined during the larger 80 ms depolarizing pulses, which in turn was reflected by the relatively rapid rate of the slow phase of decline of Ca2+ release flux during our pulses. The source of this faster decline in release during the latter part of the pulses here is not known, but might be due to partial depletion of the SR at the very low level of cytoplasmic Ca2+ prior to the pulses, or possibly to some unidentified secondary effect of low resting Ca2+ concentration.
An interesting observation from our fluorescence records was the consistent elevation of baseline fluorescence in the presence, but not absence, of Cd2+ and Co2+ (Fig. 2A). These divalent cations have been previously demonstrated to interact with fluorescent Ca2+ indicators and produce an artificial elevation in dye fluorescence in multiple cell types other than muscle (Hinkle et al. 1992; Shibuya & Douglas, 1992; Schaefer et al. 1994). Presumably there is some permeation of these ions into the cell through cationic channels, allowing the interaction with cytosolic indicator dyes. This is of specific importance to the field of excitation–contraction coupling, as these cations are commonly used for the blocking of L-type Ca2+ currents, often in conjunction with fluorescence recordings. While we consistently see this artificial elevation of fluorescence in mammalian fibres, it is unclear whether this permeation into the cell occurs in amphibian fibres. We have evaluated the effects of both Cd2+ and Co2+ separately on resting fluorescence in unstimulated fibres loaded with fluo-4-AM and found a similar elevation in fluorescence, apparently independent from a rise in resting Ca2+ concentration, for both divalent cations.
Model for the relationship between Ca2+ release and Qγ
The results presented here clearly demonstrate that although the amplitude of the rate of Ca2+ release was clearly depressed at all voltages in S100A1 KO fibres compared to fibres from WT littermates, the time course of the release waveform was essentially unaltered. Furthermore, the voltage dependence of Ca2+ release was the same, except for a scaling factor, in fibres from WT and KO fibres. In this case, movement of some Qβ in WT fibres during depolarization (Fig. 10, top centre) would activate sufficient SR Ca2+ release to generate Qγ, presumably by released Ca2+ interacting with the voltage sensor and thereby causing an additional component of charge movement (Fig. 10, right), as suggested earlier for frog fibres (Pizarro et al. 1991). Then the absence of Qγ in fibres from KO mice is simply explained by the lower rate of release observed at all voltages in KO fibres (Fig. 10, bottom). The lower rate of Ca2+ release simply does not reach the level required for the Ca2+-dependent step(s) needed for the appearance of Qγ in the KO fibres. Consistent with the link between release amplitude and appearance of Qγ, atypical KO fibres that exhibit unusually high release rates (for the KO fibres) also exhibited definite Qγ in their charge movement records. A straightforward effect of S100A1 on RyR channel gating that could account for these observations would be for the mean RyR channel open time to be longer in the presence of S100A1 then in its absence, as previously observed in bilayer studies (Treves et al. 1997). In this case, on average there would be a greater amount of Ca2+ released during each channel opening in WT compared to S100A1 KO fibres, and a consequent larger increase in local Ca2+ in the vicinity of a putative Ca2+ binding site regulating the generation of Qγ by intra-membrane voltage sensors.
Figure 10. Model for the effects of Ca2+ release on charge movement in WT and S100A1 KO fibres A, proposed effects of Ca2+ release on charge movement in WT fibres. Upon depolarization, charged residues in the α1s subunit of the DHPR (cylinders in DHPR cartoon) act as voltage sensors that transduce its movement (Qβ) through the II–III loop into conformational changes that are sensed by the RyR1 and initiate Ca2+ release (middle cartoon). In WT fibres, S100A1 binding to the cytoplasmic foot of RyR1 enhances channel open time and optimizes release. This optimized Ca2+ release results in an elevation of local Ca2+ in the vicinity of a putative Ca2+ binding site regulating the generation of Qγ by the DHPR (right cartoon). B, proposed effects of the absence of S100A1 on Ca2+ release and charge movement. In this scenario, the lower rate of Ca2+ release (middle cartoon) and the subsequently blunted local Ca2+ gradient, due to the lack of S100A1, simply does not reach the ‘optimal’ level required for the Ca2+-dependent step(s) needed for the generation of Qγ (right cartoon). Qγ may arise from the further movement of the same set of voltage sensors that generate Qβ, or from a parallel set of charges uncoupled to the activation pathway. The function of this additional component of charge movement remains to be determined.
Download figure to PowerPoint
In contrast to the simple uniform increase in release at all voltages seen in WT compared to S100A1 KO fibres, the addition of 4–CMC to KO fibres caused a shift of release activation to lower depolarizations, but with little or no change in the maximum rate of Ca2+ release. This is likely to have reflected a change in the coupling efficiency between charge moved and channel activation, and caused Qγ to appear for intermediate depolarizations, but not for the largest depolarizations where release was not altered.
The exact charged residues that contribute to Qγ, as well as their function, remain undetermined. Qγ may arise from the further movement of the same set of voltage sensors that generate Qβ, or from a parallel set of charges uncoupled to the activation pathway (Bezanilla, 2000). Recent work has demonstrated several non-conducting functions of voltage-gated ion channels, such as the activation of enzymatic pathways by α and β subunits of ion channels (reviewed in Kaczmarek, 2006). It is tempting to speculate that Qγ may contribute to some alternative cell signalling pathway in skeletal muscle fibres.