Ca_V1.2 calcium channels: Just cut out to be regulated?

Abstract

Tight regulation of calcium entry through the L-type calcium channel Ca_V1.2 ensures optimal excitation-response coupling. In this issue of Neuron, Michailidis and colleagues demonstrate that Ca_V1.2 activity triggers negative feedback regulation through proteolytic cleavage of the channel within the core of the pore-forming subunit.

The L-type calcium channel Ca_V1.2 is an integral cell membrane protein complex that contributes to the influx of calcium into excitable cells. This influx occurs in response to membrane depolarization and can trigger a wide range of cellular processes, including cardiac muscle contraction, endocrine hormone secretion, and neuronal gene expression (Catterall, 2000; Simms and Zamponi, 2014; Wheeler et al. 2012). These examples of excitation-response coupling depend critically on a cell's ability to maintain precise control over intracellular calcium levels. Therefore, it is not surprising that an arsenal of sophisticated mechanisms exist to regulate calcium channel activity itself. Indeed, feedback control over the entry of calcium through voltage-dependent calcium channels occurs through the regulation of channel activity, expression, and trafficking to and from the plasma membrane.

These disparate modes of regulation may be arrayed to provide negative feedback over multiple time scales. For example, in response to brief depolarization, calcium entry via L-type channels activates calmodulin that is already prebound to the channel. This starts rapid, calcium-dependent inactivation of the channel within milliseconds (Peterson et al., 1999; Zühlke et al., 1999), curtailing calcium entry while the membrane remains depolarized. On the other hand, prolonged depolarization results in the removal of Ca_V1.2 channels from the membrane, restraining calcium entry over a much slower timescale (Green et al., 2007).

In their provocative study featured in this issue of Neuron, Michailidis, Yang and colleagues (2014) provide evidence for yet another form of activity-dependent feedback inhibition of voltage-dependent calcium channels: calcium-dependent proteolysis of the main body of the Ca_V1.2 channel. The concept of triggering negative feedback regulation of the Ca_V1.2 channel through proteolytic processing is not unique in itself. Indeed, Hulme et al. (2006) found that proteolysis of the C-terminal domain of the Ca_V1.2 channel produces a non-covalently associated and potent auto inhibitory domain. Michailidis et al. have uncovered an equally striking example of Ca_V1.2 proteolytic processing; channel activity-dependent cleavage within the core of the Ca_V1.2 pore-forming subunit. This mid-channel proteolysis generates fragments in the plasma membrane that do not form functional channels on their own, but that seem to display distinct biophysical properties when paired with a complementary fragment.

In this study, the authorsbiotinylated surface channels and performed western blots using antibodies directed against distinct regions of the Ca_V1.2 channel. In so doing, they identified the full-length channel (240 kDa) and a prominent fragment that contains part of the II-III loop, repeats III and IV, and the C terminus (150 kDa). A complementary N-terminal fragment (90 kDa), including the N-terminus and repeats I and II, was also evident. Michailidis and colleagues went on to show that the cleavage of the full-length channel was the handiwork of calcium-dependent processes -- in part, the protease calpain -- and could be bidirectionally manipulated. They also found evidence for involvement of PEST sequences, which serve as signals for rapid proteolytic degradation through the cell quality control system (Rechsteinerand Rogers, 1996). PEST-mediated protein degradation plays a major role in modulating neuronal calcium channel function through regulation of theCa_vα1 subunit (Catalucci et al., 2009) and the Ca_vβ3 accessory subunit (Sandoval et al., 2006).

Whatever the detailed mechanism, mid-body regulation is intriguing for multiple reasons. It enzymatically severs the tandem linkage of four individual motifs -- each with Shaker K channel-like structure -- a hallmark feature of calcium and sodium channels that took eons to evolve; this is a more radical change in VGCC architecture than abbreviation of the long C-terminal tail. The mid-channel cleavage appears to leave two pieces that can remain together in a partially functional state, based on evidence from engineered complementary fragments. Resulting current intensities, at levels<40% of those generated by intact pore forming subunits, were consistently found regardless of where the split was imposed. Thus, the cleaved channel appears to lie somewhere between fully functional and nonfunctional, a convenient step down for auto regulation of calcium influx.

Biotinylation of intact tissue allowed Michailidis et al. to focus on proteins unambiguously localized to the surface membrane. In a complementary pre-labeling approach, they tackled the question of whether the Ca_V1.2 fragments always cling tightly to each other or sometimes drift away. After pre-labeling the N-terminal fragment with a green cytoplasmic GFP tag, and the C-terminal fragment with an HA-epitope tag later immunostainedin red, they checked whether red and green intensities had identical spatial distributions across the surface of the cell. If no cleavage took place, the red C-terminal and green N-terminal fragments would remain perfectly matched up (or would at least be equally abundant within the limits of optical analysis; pixel width, 0.2 μm) and their non-colocalization index (NCI=red/green ratio) would be unity throughout. Instead, the red/green ratio varied widely from unity over the surface of cultured neurons, resembling a patchwork quilt of yellow, red and green. This provided the authors with cell biological evidence for the existence of mid-body cleavage, but also raised new questions about what pulls the N and C-terminal fragments apart, and what their ultimate functions might be. As it stands, the evidence suggests that neurons display a dispersion of VGCC molecules in different states, reflecting the neuron's previous history of activity: full-length subunits, fragments that might modulate full-length subunits, paired-up fragments, and even isolated clusters of C-terminal fragments. The array of Ca_V1.2 components may even vary dramatically with age.

Preview authors have the privilege of blithely advocating for future experiments. We think that it will be important to study the kinetics of the mid-body cleavage and the cell biology of fragment anchoring and turnover. We note that differential regulation of internalization rather than regulation of cleavage itself would also show up as a calcium-dependent shift in the relative surface abundance of full-length and C-terminal fragments that could account for reduced channel current density.

It would also be worthwhile to determine if mid-body cleavage is generalizable to other kinds of VGCC's, as the preliminary findings of Michailidis et al. with Ca_V2.1 would suggest.

Finally, it would be interesting to test whether the isolated C-terminal fragments physiologically modulate full-length channel subunits or serve a unique neuronal function of their own. Our hunch is that they are not just proteinaceous detritus merely awaiting removal for further proteolysis. Might they serve as membrane anchors for ancillary proteins even when their usefulness as flux-generating devices is over? Many cytoplasmic proteins interact with the Ca_V1.2 C-terminal tail. Could they act as pore-less voltage sensors since they still contain structural components for sensing voltage? Voltage-dependent conformational changes may serve a signaling role in neurons (Tadross et al., 2013), by analogy to the function of Ca_V1.1, the classical voltage-sensor of skeletal muscle. Gating current measurements would indicate if conformational changes are intact in isolated C-terminal fragments and in complementary fragment pairs, both for generation of gated calcium flux, and for conveying information about neuronal depolarization per se. The voltage-dependent gating of fragment pairs is significantly different from the full-length channel, raising the possibility that gating conformational changes are somehow different.

In conclusion, Michailidis have added both potency and complexity to our picture of how calcium channel activity, so important for cellular homeostasis, is itself regulated.

Figure 1.

Mid-body proteolytic cleavage of the basic Ca_V1.2 pore-forming unit. This Ca²⁺-dependent cleavage, here illustrated for the II-III linker (left), gives rise to fragments that can remain together in a functional state (upper right). The fragments can also dissociate and form a mosaic pattern in the surface membrane (lower right), raising questions about their possible functional significance.