L-type Ca2+ channels play a key role in many excitable cells, e.g. cardiomyocytes, skeletal myocytes, pancreatic β-cells, retinal cells and inner hair-cells. They are heteromeric protein complexes composed of a pore-forming CaV1 subunit and auxiliary α2–δ, β and γ subunits. As well, a tightly associated calmodulin may be considered as an ‘honorary’ Ca2+ channel subunit. L-type Ca2+ channels are subject to voltage-dependent inactivation (VDI), but a hallmark feature of L-type Ca2+ channels is the Ca2+-dependent inactivation (CDI). This type of inactivation can be triggered by global enhancement of intracellular Ca2+ levels, but elevated Ca2+ concentration in close vicinity to the channel is sufficient to induce CDI (Tadross et al. 2008). It is clear that the interaction of Ca2+ with calmodulin (Alseikhan et al. 2002), and of calmodulin with the C-terminus of the pore-forming subunit Cav1 (with an additional role of the N-terminus in regard to NSCaTE in CaV1.2 and CaV1.3) are necessary and sufficient for CDI. However, CDI has not yet been understood in detail, in particular regarding the alterations of single-channel gating. In an outstanding study Imredy & Yue (1994) proposed that the introduction of a Ca2+-dependent mode (‘mode Ca’) of L-type Ca2+ channel gating can serve as a basis to ‘enable successful separation of Ca2+- and voltage-sensitive forms of inactivation’. Their study thus added one more mode to those proposed by Hess et al. (1984) 10 years earlier (‘mode 1’, ‘mode 2’ and ‘mode 0’). In the later years, much of the biophysics of CDI has been investigated using whole-cell current recordings. Along this path, Ian Findlay cast remarkable doubt upon the – by then stipulated – separation of fast L-type Ca2+ current (ICaL) decay as the Ca2+-dependent inactivation and the delayed ICaL decay as the voltage-dependent inactivation (reviewed in Findlay, 2004). Furthermore his data show that β-adrenergic stimulation – used in many studies to yield sufficient currents – affects the relative impact of fast (or ‘Ca2+-dependent’) versus slow (or ‘voltage-dependent’) inactivation.
The study by Josephson et al. (2010) published in this issue of The Journal of Physiology is an impressive – and successful – attempt to extend the findings of the above-mentioned studies on inactivation and overcome some of their limitations. They analyse single-channel recordings, and thus engage with phenomena intimately related to the molecular level of mechanisms underlying inactivation processes. Data were obtained with Ca2+ as charge carrier in a nearly physiological concentration, thus improving the relevance of these findings. They avoided using channel activating substances, thus eliminating possible drug effects on inactivation in general and channel modes in particular. Furthermore the authors withstood the temptation to use genetically engineered recombinant systems and investigated native primary cells. Josephson and colleagues conclude – based on distinct kinetics – that inactivation of single channels results from two Ca2+-sensitive mechanisms: (1) a relatively fast reduction of frequency of channel re-opening and (2) a slower decay of mean open time. Of note the first mechanism was clearly weaker but still present when Ba2+ was used as charge carrier. By extension of the model proposed by Imredy & Yue, this may reflect a transition between ‘mode 1’, a ‘mode BaCa’ and ‘mode Ca’ (Fig. 1). How voltage feeds into this model and whether extension of the model to give a complete 3 × 3 modal matrix will be required remain to be settled.
Figure 1
How classical models of modal L-type Ca2+ channel gating could be expanded
On a more technical note, the demanding signal-to-noise ratio and the resolution bandwidth achieved by Josephson et al. should cause more missed events and a greater overestimation of mean open time towards the end of pulses. In other words, the time course of ‘frequency of re-opening’ may in fact be shallower, but the time course of open times be steeper than they appear (based on detected events). Also, safe exclusion of multi-channel patches (which is hard to achieve at low open probability) is mandatory to discriminate classical ‘inactivation’ (i.e. eventually a channel will enter an absorbing non-open state during depolarisation) from true gradual changes of the frequency of re-openings. In these respects, the very existence of ‘mode BaCa’ awaits confirmation under well-resolved recording conditions. Therefore, it seems quite straightforward to use ‘non-physiological’ conditions (channel modulators, charge carrier, engineered channels), but see above!
Yet, the study by Josephson et al. is an encouraging example of how less can be more: using a simple but innovative approach they show that a countable number of calcium ions seen by a single channel is enough to refine our understanding of CDI. This paves the way for further in-depth analysis (not only) at the single-channel level, e.g. to elucidate the interplay between calcium- and voltage-dependent processes by quantitative modelling, or to address the role of other players like auxiliary β-subunits that have been shown to differentially affect fast and slow inactivation (Cens et al. 1999).