PerspectiveCalcium signaling

L-Type Calcium Channels: On the Fast Track to Nuclear Signaling

See allHide authors and affiliations

Science Signaling  14 Aug 2012:
Vol. 5, Issue 237, pp. pe34
DOI: 10.1126/scisignal.2003355


Calcium signaling resulting from depolarization of neurons can trigger changes in transcription, and this response has been called excitation-transcription (E-T) coupling. In neurons, voltage-gated and ligand-gated calcium-permeable channels contribute to the increase in intracellular calcium. It appears that calcium signals mediated by specific voltage-gated calcium channels may have distinct roles in E-T coupling.

Excitation-transcription (E-T) coupling is the process, found in both neurons and muscle cells, that converts a rise in intracellular Ca2+, resulting from depolarization, into a change in transcription in the nucleus (1, 2). In neurons, Ca2+ entry may occur through multiple routes, including the opening of voltage-gated calcium channels (VGCCs) and activation of N-methyl-d-aspartate–type glutamate receptors, which are ligand-gated calcium-permeable channels (3). It has been known since 1986 that L-type Ca2+ channels, a type of VGCC, are involved in signaling to the nucleus and E-T coupling, which results in the transcription of immediate early genes, such as c-fos (4, 5). Wheeler et al. (6) have now investigated whether all Ca2+ entry through VGCCs is equal in this process and come to some surprising conclusions.

VGCCs can be divided functionally into three main classes: (i) the L-type channels, which are activated by medium to high voltages and are sensitive to 1,4 dihydropyridine (DHP) agonists and antagonists (7); (ii) the non–L-type, high-voltage-activated channels, which are subdivided into N, P/Q, and R (8); and (iii) the low-voltage-activated T-type channels (9). Their molecular counterparts are the CaV1 (corresponding to L-type channel class), CaV2 (corresponding to P/Q-, N-, and R-type channels), and Cav3 channels (corresponding to T-type channel class) (10) (Table 1). Neurons possess different combinations of these VGCCs. For example, in rat superior cervical ganglion (SCG) neurons, which are peripheral sympathetic neurons, N-type calcium channels contribute most of the voltage-dependent calcium influx and L-type channels contribute a smaller component (11). In rat SCG neurons, the L-type current is likely to consist mainly of CaV1.3 channels based on message abundance (12). By contrast, very little or no P/Q type current or message encoding Cav2.1 was found in rat sympathetic neurons (12, 13).

Table 1

Calcium channel nomenclature and pharmacology. [Compiled from multiple sources, including references (7–10)]

View this table:

One of the most extensively investigated mechanisms of E-T coupling in neurons is signaling to the transcriptional regulator CREB (cyclic AMP response element–binding protein), which is found mainly in the nucleus and which becomes rapidly phosphorylated within tens of seconds of Ca2+ entry (14, 15). Several pathways may converge to transfer information from the plasma membrane to the nucleus (2). These pathways include the nuclear translocation of the activated Ca2+ sensor, Ca2+-calmodulin (CaM), followed by activation of the nuclear protein kinase Ca2+-calmodulin kinase IV (CaMKIV) [for a review, see (2)]. Alternatively or additionally, cytoplasmic Ca2+-CaM may activate local cytosolic pools of CaMKII in the immediate vicinity of Ca2+-CaM (16), which then participate in the pathway(s) culminating in CREB phosphorylation. Also, the activation of other cytoplasmic intermediaries, such as mitogen-activated protein kinase (MAPK), may come into play. In this case, Ca2+-CaM stimulates nuclear translocation of MAPK through a mechanism that requires protein kinase A (PKA) activity (17, 18). It is still a matter of debate whether an increase in bulk cytoplasmic Ca2+ supports a sustained increase in nuclear Ca2+ to affect CREB phosphorylation (1) or whether an increase in bulk cytoplasmic Ca2+ prolongs the activation of cytoplasmic Ca2+-CaM to facilitate gene transcription (2).

In addition to activation of CREB, other transcriptional regulators can contribute to the E-T response. For example, another established route of E-T coupling involves the transcription factor NFAT (nuclear factor of activated T cells). In this pathway, Ca2+-CaM–mediated activation of the phosphatase calcineurin results in dephosphorylation of cytoplasmic NFAT, which then translocates to the nucleus (19). Another postulated mechanism for signaling to the nucleus involves the production of a cleaved C-terminal fragment of the L-type channel CaV1.2, which is present in the nucleus of cortical neurons and mediates changes in gene transcription, although this is found to exit the nucleus upon depolarization (20).

Wheeler et al. (6) investigated the specific contributions of L-type and N-type Ca2+ channels in mediating E-T coupling, taking advantage of their different pharmacological and biophysical properties (Table 1). They applied the straightforward technique of increasing extracellular K+ to trigger depolarization of cultured rat SCG neurons, employing varying concentrations of K+ to reach increasing states of depolarization. They also used pharmacological blockers of the different channels (including ω-conotoxin GVIA to block N-type channels, ω-agatoxin IVA to block P/Q-type channels, and nimodipine to block L-type channels) to tease out the roles of the different calcium channel subtypes in E-T coupling. Ca2+ signals and the localization of phosphorylated proteins involved in the E-T response were imaged to identify how Ca2+ entry mediated by specific channels contributed to E-T coupling. The principal findings of Wheeler et al. (6) are that Ca2+ entry through L-type channels results in more CaMKII phosphorylation and signaling to the nucleus (as measured by CREB phosphorylation or c-fos transcription) than Ca2+ entry through N-type calcium channels. Depolarization resulted in accumulation of cytoplasmic “hotspots” of phosphorylated CamKII coincident with or near to the L-type (CaV1.3) channel puncta. Whether L-type channels of other molecular composition (such as CaV1.2 channels) exhibit similar properties will be very interesting to pursue with imaging studies using CaV1.3 and other knockout mice (21, 22).

On the other hand, Ca2+ entering through N-type (CaV2.2) channels seems to act at a distance, because there was no local accumulation of phosphorylated CaMKII associated with CaV2 channels. Instead, Ca2+ entering through N-type channels was found to contribute to the phosphorylated aggregates of CaMKII associated with L-type channels. However, less phosphorylated CaMKII was produced in response to Ca2+ entering through N-type channels, because of the buffering of the calcium signaling by the endoplasmic reticulum (ER) (Fig. 1). One caveat to this conclusion is that the immunocytochemistry showing punctate staining was performed with CaV2.1 antibodies rather than CaV2.2 antibodies, whereas the functional data show that CaV2.2 is the main CaV2 channel involved in these neurons.

Fig. 1

The differential localization of L-type, compared to N- and P/Q-type, channels affects their capacity to signal to the nucleus in SCG neurons. ER and mitochondrial buffering of the calcium that enters through N-type channels reduces the ability of this calcium signal to mediate E-T coupling. In contrast, calcium entry through L-type channels produces a local calcium signal that efficiently stimulates CaMKII phosphorylation and activation and promotes E-T coupling by phosphorylation and activation of the transcription factor CREB.


Within neurons and other cell types, the ER represents a continuous network containing a releasable store of Ca2+. With respect to Ca2+ handling, the two main roles of the ER are to take up Ca2+ through the calcium pump SERCA (sarcoplasmic-endoplasmic reticulum Ca2+–adenosine triphosphatase), which maintains low cytoplasmic Ca2+ concentrations, and to participate in Ca2+ release mediated by inositol trisphosphate (IP3) receptors in response to production of the second messenger IP3 and by ryanodine receptors in response to increased cytosolic Ca2+, the latter of which is known as Ca2+-induced Ca2+ release. The mitochondria also serve as Ca2+ buffers, representing high-capacity, low-affinity Ca2+ sinks (23). This function of the mitochondria is intimately connected to the ER (24) (Fig. 1), although how Ca2+ transfers between the ER and mitochondria is still under debate (25).

As an explanation of why entry of Ca2+ through N-type (CaV2.2) channels is ~10-fold less effective than that through L-type (CaV1.2 and 1.3) channels at triggering E-T coupling, as measured by CREB phosphorylation, Wheeler et al. (6) show that Ca2+ entering through CaV2.2 channels is buffered by uptake into the ER and thence into mitochondria, suggesting a colocalization of these channels with specialized regions of the ER containing SERCA. In agreement with this, blockers of SERCA activity increased the Ca2+ signal from CaV2.2 channels, yielding increased size and intensity of phosphorylated CaMKII puncta. Although it has been shown previously that ryanodine receptors are present on the surface of the ER that is juxtaposed to the plasma membrane (26), their role in E-T coupling has not been investigated here. Although the ER is present throughout SCG neurons, Wheeler et al. (6) find that there are hotspots of SERCA situated beneath the plasma membrane, some of which are near P/Q-type (CaV2.1) channel puncta. It will be of future interest to determine the basis for this juxtaposition and whether N-type channels are also localized or directly tethered near SERCA pumps.

On the basis of their electrophysiological results, Wheeler et al. concluded that L-type channels also had a “gating advantage” for E-T coupling compared with non–L-type channels (6). CaV1.3 is activated at more negative potentials than are CaV1.2, CaV2.1, or CaV2.2, and this could explain the gating advantage, because L-type channels are likely to comprise mainly CaV1.3 in SCG neurons. However, this advantage will only apply to subthreshold excitation, not to E-T coupling generated by action potentials (27). Another possible caveat to the study of Wheeler et al. arises due to the use of the channel blocker nimodipine, which at the concentrations used may not block all of the L-type (CaV1.3) channels, because these channels exhibit a lower affinity for DHPs (28, 29). However, the same conclusion of more efficient E-T coupling by L-type channels was supported by their experiments using various toxins to block non–L-type channels (6).

In apparent contrast to the results of Wheeler et al. (6), two other studies showed that there was a close interaction between L-type Ca2+ channels and components of the ER, both in neuronal and non-neuronal cells (30, 31). Specifically, stromal interaction molecule 1 (STIM1), which is present on the ER and activates plasma membrane Orai channels that mediate store-operated Ca2+ entry when the ER Ca2+ stores are depleted, interacts with CaV1.2 channels (30, 31). This was found to occur in a wide range of cell types, including smooth muscle and embryonic human embryonic kidney (HEK) 293 cell lines (30), cortical and hippocampal neurons, and T lymphocytes (31). Whereas STIM1 interacts with Orai to activate this channel, STIM1 inhibited the opening of CaV1.2 channels by binding to the C terminus of CaV1.2, suggesting that there is a local apposition of these L-type channels and the ER in these cells. The interaction of CaV1.2 with STIM1 also resulted in internalization of the channel from the plasma membrane (31). It has not yet been studied whether or not the same interaction occurs with CaV1.3, which could explain why this mechanism was not observed by Wheeler et al. (6).

In conclusion, Ca2+ entry through L-type and non–L-type Ca2+ channels shows differential abilities to signal to the nucleus and also exhibit cell type–dependent interactions with the ER or proteins that reside in the ER. Wheeler et al. (6) propose that the differential signaling to the nucleus results from local CaMKII accumulation and phosphorylation only near the L-type channels (shown for CaV1.3) and selective sequestration of Ca2+ entering through non–L-type channels by ER and mitochondrial buffering (Fig. 1). How cells sort their VGCCs into different compartments of the plasma membrane and the impact that this compartmentalization may have on responses other than E-T coupling in neurons remain open questions.


View Abstract

Stay Connected to Science Signaling

Navigate This Article