Local is as local does: the unitary nature of SR Ca2+ release in cardiac ventricular myocytes

Excitation–contraction (E–C) coupling in cardiac muscle is the transduction mechanism linking the action potential to cell shortening which occurs in response to a transient rise in intracellular calcium (Ca2+i). Activation of the Ca2+ current (ICa,L, the substrate of which is the L-type Ca2+ channel (LTCC) or dihydropyridine receptor (DHPR)) triggers a larger release of Ca2+ from the sarcoplasmic reticulum (SR) store via Ca2+-induced Ca2+ release (CICR) due to activation of the SR Ca2+ release channel, the ryanodine receptor (RyR), on the SR membrane. Experimental evidence strongly supports ‘local control’ of Ca2+ release where a cluster of RyRs are juxtaposed with DHPRs forming a couplon or Ca2+ release unit (CRU) that functions autonomously. The voltage sensitivity of DHPR activity and the progressive recruitment of couplons is the cornerstone of graded Ca2+ release and contractility in cardiac tissue. Although the relationship between trigger Ca2+ and SR Ca2+ release has received much attention, critical questions still remain about the nature of the coupling between LTCC and RyR. In particular, elementary Ca2+ release events (Ca2+ sparks) can be triggered with very high probability during an action potential (AP) but at the peak of the AP the probability of DHPR openings is very much lower than that of Ca2+ sparks (Inoue & Bridge, 2003). Therefore, more than one DHPR must be involved in triggering SR Ca2+ release but it is not clear exactly how many DHPR channels reside in a couplon and how the numbers of channels and their activity modulates SR Ca2+ release. In a recent paper in The Journal of Physiology, Polakova et al. (2008) significantly advance our understanding of the local control process in a complex but powerful study, by demonstrating that a significant number of DHPRs are present in a couplon (20–40) and that at least eight of these channels must be open simultaneously in order to trigger SR Ca2+ release Spontaneous and triggered local release events have been intensively studied since the first description of Ca2+ sparks by Cheng et al. in 1993 (Cheng et al. 1993). An advance on the Ca2+ spark methodology was demonstrated by Song et al. (1998) where a low-affinity Ca2+-sensitive fluorescent probe was combined with high concentrations of EGTA which reported SR Ca2+ release from a couplon. This release activity was termed a Ca2+ spike. Ca2+ spikes are a more appropriate approach where SR Ca2+ release flux is important than measuring Ca2+ sparks, since the Ca2+ spark signal also contains components not directly associated with SR Ca2+ release flux. This is especially evident in the rising and decay phases of Ca2+ sparks, which are intrinsically much slower than Ca2+ spikes. Ca2+ spikes are therefore able to report the latency of Ca2+ release quite faithfully (Zahradnikova et al. 2007). The Ca2+ spike approach involves the use of EGTA in the cell (introduced via the patch pipette), which prevents diffusion of the Ca2+ signal, thereby filtering out global cytosolic changes in Ca2+ whilst the low-affinity indicator reports the large Ca2+ release flux. Polakova and co-workers apply this methodology to rat ventricular myocytes where the Ca2+ spike signal consists of two components: a dominant component attributable to SR Ca2+ release flux, which is manifest as a rapid and transient spike, and a minor component representing the integral of Ca2+ release from the SR. Song et al. (1998) describe this as a ‘pedestal’ since during a voltage pulse this component of the signal remains elevated above baseline. In addition to this powerful methodology, Polakova and colleagues utilize a clever voltage clamp protocol that permits control not only of the number of channels triggered to open but also the driving force acting on those channels. Channels were activated with minimal Ca2+ flux by a depolarizing prepulse to Erev of ICa,L (+60 mV). Rapid repolarization increased the driving force for Ca2+ entry in a temporally synchronized manner and in this aspect the protocol is superior to traditional step depolarizations since channel open probability and driving force would otherwise change in opposite directions. The number of channels activated was controlled by the time spent at +60 mV (1.5 or 5 ms). Line scanning confocal microscopy monitored the occurrence of Ca2+ spikes. A strong component of the research paper was the use of mathematical modelling to describe the stochastic nature of Ca2+ release events (probability, latency) and it is from this computational analysis that many of the significant conclusions of the paper are drawn. The experimental procedure described above was used by the group to demonstrate that the number of DHPRs activated in the couplon is a significant factor in triggering Ca2+ release. Varying the number of channels activated (controlled by the prepulse to +60 mV) affected spike probability and latency. Spike probability was higher and latency was shorter with larger numbers of activated channels. Where maximal numbers of channels were activated compared to only a fraction of channels (5 ms versus 1.5 ms prepulse), the difference between the spike probability was smallest at –40 mV. This repolarization voltage does not inhibit channel reopenings during the tail current since this potential is not sufficiently hyperpolarized to fully inactivate the channels. Therefore, the differences in latency between the two fractions of activated DHPRs were largest at –40 mV since larger numbers of channel reopenings contribute to increasing the synchrony of SR release. Given that these results demonstrate the importance of the number of DHPR channels for SR release, an examination of lengthening the channel open time with the agonist BayK 8644 was performed. This rendered spike properties independent of the repolarization potential, suggesting that, in addition to the numbers of channels open, the channel open time may also be an important factor in the coupling of SR release, a situation that has physiological relevance since it mimics modulation by β-adrenergic agonists in cardiac tissue. However, this result also describes a paradox in the data reported by this group. The analysis carried out by Polakova and colleagues was applied to the distribution of spike latencies. Latency may be interpreted simply as the synchrony of Ca2+ release from the couplon and this synchrony is regulated by the open probability and number of DHPRs in the couplon. As more channels are recruited (5 ms versus 1.5 ms prepulse) synchrony increases, i.e. more spikes occur with shorter latencies. The paradox rests in that prolonging channel openings with BayK 8664, synchrony (latency) was not increased and in fact it was longer than control. The interpretation of these data is that the number of active channels is a dominant feature in the control of SR Ca2+ release flux and not the length of channel openings per se. However, these data are inconsistent with other reports that demonstrate increased synchrony with β-adrenergic stimulation (which also has the effect of increasing channel open times (Song et al. 2001)). Both the application of BayK 8664 and β-adrenergic stimulation shift the voltage dependence of DHPRs in favour of activation at more negative voltages. This may have the effect of (1) recruiting more channels in the couplon which would explain the increased synchrony observed by other groups or (2) the effects of lengthening DHPR open times may not be discernable during an activating prepulse to +60 mV, and explains the paradox reported by the Polakova group. SR load has a significant impact of SR release flux which may also have contributed to this result. The analysis used by Polakova and co-workers enabled description of the coupling fidelity of SR release (i.e. the ability of a single DHPR to trigger SR Ca2+ release). The equation used to explain this includes the number of DHPRs in a couplon, channel open probability and the number of channels that open sequentially (together) (see eqn 4 in the paper). The probability and latency distribution were modelled to derive estimations of these parameters. Significantly, this demonstrated that the coupling fidelity is very low (0.15 at –120 mV) and thus the only way that SR release can occur with high probability (i.e. to mimic the experimental data) is if multiple channels open simultaneously in a couplon. It is from this analysis that the group surmise that at least eight channels must be open in a couplon during activation to trigger SR release and that a couplon contains between 20 and 40 DHPRs. This represents the key advance to our understanding of local control of E–C coupling. The power of this analysis lies in that it quantitatively describes the numbers of DHPRs of a total couplon population that are required to be active in order to trigger SR release and provides a paradigm for explaining the responses to adrenergic stimulation (Song et al. 2001), during pathological remodelling and variable E–C coupling gain. Each of these situations may result in the numbers and activity of DHPRs being modified and explains the capacity for physiological regulation by the local control mechanism. In summary the paper by Polakova and colleagues represents a significant advance in our understanding of the local control mechanism of E–C coupling. The number of channels in a couplon and the minimum number required to open to trigger SR flux are likely to be modified by adrenergic stimulation, regional location, species and during cardiac disease. Application of their approach to these situations is likely to provide a fruitful area for future investigation.