Store-Operated Ca2+ Entry Channels: Still Elusive!

See allHide authors and affiliations

Science's STKE  27 Jul 2004:
Vol. 2004, Issue 243, pp. pe36
DOI: 10.1126/stke.2432004pe36


Depletion of intracellular Ca2+ stores is believed to trigger Ca2+ entry through a Ca2+- permeable channel in the plasma membrane called the store-operated Ca2+ channel (SOC). This type of Ca2+ entry is thought to play a pivotal role in a plethora of cell functions ranging from gene expression to sensory signal transduction. However, the molecular nature of these channels is still elusive. Many molecular candidates have been described as SOCs, but the candidacy of each has been countered by reports indicating that they are not SOCs. Most of the suggested candidates are members of the recently discovered superfamily of transient receptor potential cation channels (TRPCs). However, no TRP-family channel has yet been incontestably identified as an SOC. Even for the electrophysiologically best-described SOC, the highly Ca2+-permeable channel Ca2+ release-activated Ca2+ channel (CRAC), an acceptable candidate has yet to emerge. This perspective summarizes problems in identifying SOCs, suggests approaches to solving these problems, and discusses concerns over the current tendency to focus exclusively on the hypothesis that TRPCs comprise SOCs.

Twenty-two years ago, Casteels and Droogmans showed that depletion of agonist-sensitive intracellular Ca2+ stores increased the rate of Ca2+ uptake from the extracellular solution in vascular smooth muscle cells (1). On the basis of this initial description, a model was developed that explained most of the observed Ca2+ signals induced by agonist stimulation of membrane receptors. The model described an initial emptying of the intracellular Ca2+ stores by inositol 1,4,5-trisphosphate (IP3), followed by entry of Ca2+ into the cytosol and refilling of the stores (2, 3). In subsequent studies, emptying of the stores in response to different protocols was considered to act as the trigger for Ca2+ entry. This type of Ca2+ entry is referred to as store-operated Ca2+ entry (SOCE). The first electrical measurement of current through a store-operated channel (SOC) was achieved in mast cells (4). This current was termed Ca2+ release activated Ca2+ channel (CRAC) current. Later on, much evidence was published indicating that SOC-related calcium signals control a plethora of essential cellular functions, ranging from secretion to gene expression, which in turn control cell growth, proliferation, and cell death and are also able to trigger a wide range of sensory signaling cascades (5, 6). However, even given all of the progress in this field, two fundamental problems remain unsolved: (i) What is the molecular nature of SOCs? and (ii) What is the mechanism of SOC activation? This Perspective discusses some of the factors that have impeded our ability to solve these two questions and suggests future research directions that might help solve the first problem.

The Properties of Store-Operated Calcium Channels

One has to take into account that "SOCE" describes the phenomenon of Ca2+ entry into a cell measured under conditions that lead to depletion of intracellular Ca2+ stores. It has not yet been proved directly that the associated Ca2+-entry channels are really SOCs, because the protocols normally used to elicit Ca2+ entry do not allow a direct comparison between the content of calcium in the store and the activation of a current. Clearly, researchers in the field have yet to reach a complete consensus on what is required to define a "store-operated" channel. This lack of consensus depends in part on the lack of comparable and well-defined protocols used to identify the mechanisms that underlie channel activation. Before a particular entity can be accepted as an SOC, it must be shown that the "channel" is activated by Ca2+ depletion of a store in a native system at a physiologic temperature. Store depletion must be achieved in different ways, and each method should result in activation of the same channel. Preferably, this analysis should be done using a patch-clamp approach. However, manipulating "global" or "domain" free intracellular Ca2+ concentration using Ca2+ buffering or Ca2+ release protocols will probably affect many other store-dependent processes. These other processes may modulate the content of Ca2+ stores and might be able to induce positive or negative feedback mechanisms that act on Ca2+ entry mechanisms. It will be necessary to carefully define a system that allows a clear correlation between depletion of Ca2+ stores and activation of plasma membrane channels to rigorously define an SOC.

The most straightforward approach would be to show activation of a whole-cell current--preferably measured by means of a perforated patch approach--caused by store depletion and to identify a single-channel equivalent (as has been done for all voltage-operated Ca2+ channels). If a channel is very Ca2+-selective, we have to expect tiny currents. Such currents might be difficult to distinguish from background currents. In this case, nonstationary noise analysis, a method that allows the discrimination of currents through channels (and possibly transporters) with femtosiemens (fS) conductance, should be used (7). For channels with a low permeability for Ca2+, like most of the "canonical" transient receptor potential channel (TRPC) candidates, the whole-cell approach (preferably in the perforated patch configuration, which prevents leakage of intracellular regulatory compounds) should be most straightforward, and it is surprising that this approach is widely neglected. Single-channel data alone are often worthless if they are not accompanied by a careful kinetic analysis.

To definitively identify a particular molecular candidate, the critically defined "SOC" phenomenon should disappear when the candidate protein or gene is eliminated in vivo or in a native cell.

Thus, three criteria must be met before a candidate can be unequivocally identified as an SOC: (i) It must be activated by various forms of store depletion, such as "active" and "passive"; (ii) there must be a close correlation between current activation/deactivation and changes in the Ca2+ content of the store; and (iii) the phenomenon must disappear when the candidate protein or gene is eliminated in a transgenic model, a controlled antisense, or an siRNA protocol. These criteria have not yet been met for any identified channel.

What should the biophysical properties of an SOC be? From the biophysical point of view, it should be Ca2+-permeable. The best-characterized SOC from this perspective is CRAC, which has been identified in several cell types, particularly in blood cells (5). CRAC is an SOC with properties differing from other less consistently described SOCs (such as several nonselective cation channels from the TRPC family). CRAC is characterized by a permeability ratio for Ca2+ and Na+ (PCa/PNa) of about 1000, but has a very low single-channel conductance for monovalent cations (~0.1 pS), which drops into the fS range in the presence of Ca2+ (8, 9). So far, even for this best-described SOC, no clear molecular candidate exists.

Although CRAC is the best-characterized SOC, we should not exclude less Ca2+-selective cation channels as potential candidates for mediating SOCE. Their mode of operation will vary from that of CRAC. First, Ca2+ entry will represent a fraction of the total current and will probably be less precisely regulated than Ca2+ entry though a highly selective channel. Second, the cell will depolarize, causing a negative feedback for Ca2+ entry. Third, a nonselective cation channel will enable Na+ entry. Such a channel could promote Ca2+ entry; for instance, if it is tightly coupled to Na+-Ca2+ exchange (10). Possible candidates could therefore be sought in the whole range of cation channels, Ca2+-permeable or even Ca2+-impermeable nonselective cation channels.

Why focus on only Ca2+-permeable store-operated channels? A "store-operated" K+ channel could be very helpful in clarifying the driving force of Ca2+ entry into the cell. For instance, turning on a K+ channel through store depletion could stabilize the negative potential required for driving Ca2+ through the SOC, and turning off a K+ channel or activating a Cl channel in a store-dependent manner could depolarize a cell, thereby activating Ca2+ influx through voltage dependent Ca2+ channels. Why do we exclude such possibilities from consideration?

The Role of Known Channel Proteins

Most intriguingly, TRPCs have been overwhelmingly welcomed as the missing molecular candidates for SOCs (6, 11). But are TRPCs really SOCs? The evidence that depletion of intracellular Ca2+ stores initiates or modulates the activation of various TRPCs is overwhelming. Such experiments have been shown for all TRPCs (9) and also for TRP-vanilloid 6 (TRPV6), a member of one TRPC subfamily (12, 13). Most TRPCs have a low Ca2+ selectivity with a PCa/PNa between 0.1 and 10 (6, 1315) and undoubtedly contribute to Ca2+ entry. The only Ca2+-impermeable TRPCs so far identified are members of another TRPC subfamily, TRP-melastatin 4 (TRPM4) and TRPM5 (16, 17). Very likely they contribute to the regulation of Ca2+ entry through SOCE, but are not themselves store-operated.

Most of these studies were done in heterologous expression systems. Needless to say, results from cell systems in which the signaling cascade between store and plasma membrane might be altered, the correct stoichiometry might be violated, or the correct players (subunits) might be missing, are of dubious value. In addition, in any overexpression system, the effects of local changes in Ca2+ concentration in a domain around the channel might be dramatic, considering the extraordinarily high Ca2+ sensitivity of nearly all TRPCs. The question remains, do any of the TRPCs fulfill the above-described criteria for SOCs? Referring to the last criterion listed above (regarding disappearance of critical SOC phenomena when a candidate channel has been eliminated), only a few studies have been performed. In a TRPC4-deficient mouse model, an SOC current disappeared (18). However, the biophysical properties of heterogeneously expressed TRPC4 channels do not match the "lost" currents (19, 20). It is therefore not clear whether TRPC4 is an SOC or a regulator thereof. TRPC1 has been described in a knockdown approach as an essential component of SOCs (21). Recently, TRPC1, TRPC3, and TRPV6 have also been described in native cells, and by a knockdown approach they were all found to act as SOCs, although with different modes of activation (13, 15). Surprisingly, except for TRPC4, none of these studies was performed on native cells from the available TRPC knockout mice. None of the studies, which are only a selection from many reports, clearly defines a "store-operated channel" according to the above-mentioned criteria. Thus, we cannot adequately answer yet the simple question, "Are some TRPs SOCs?"

The same is true for the more specific question, "Are there TRPs that fulfill the criteria of CRAC?" The only highly Ca2+-selective channels in the TRP family described so far are TRPV5 and TRPV6, which have PCa/PNa > 100 (6, 13, 22). Several of their features are identical with CRAC. However, single-channel conductance, open-pore block by intracellular Mg2+, and permeability for Cs+, which all reflect pore properties, and several pharmacological properties, differ substantially between TRPV6 and CRAC (22). Therefore, TRPV6 is very likely not CRAC. However, endogenous CRAC was markedly depressed by expression of N-terminal TRPV6 fragments, indicating a possible modulatory role of TRPV6 on CRAC. However, these findings also underline the fact that TRPV6 is not CRAC (23). So far, no CRAC channel seems to be present in the TRP family, at least judging from our knowledge of expressed channels including heteromers. In any case, the superficial identification of TRPs as SOCs must be avoided!

In an alternative approach, however, we should consider that SOC activity could be attributed to transporters. In support of this notion, recall that most of the adenosine triphosphate (ATP)-driven primary transporters, many ion exchangers, and even ClCec1, the prokaryotic predecessor of ClC channels, are electrogenic (24, 25). Moreover, the conductance of CRAC for monovalent cations is very small (8). Thus, we reach a transfer rate for ion entry that is in the range of a transporter rather than a channel. Two other exciting examples of a possible involvement of transporters in Ca2+ entry have recently been published. The divalent metal transporter-1 (DMT1 ) can easily be converted into a Ca2+ channel by a single mutation, G185R, which creates a constitutively open, highly Ca2+-permeable channel (26). Also, the mitochondrial Ca2+ uniporter is a highly Ca2+-permeable channel (27). Permeation pathways are also coupled to transporters. Should these examples not immediately draw our attention to novel candidates for SOC?

The Assembly and Organization of Channels

One of the reasons for the elusive nature of SOCs could be that a functional channel requires the formation of heteromers (which are not formed in an overexpression system), assembly with unknown subunits, or the formation of signaling complexes (signalplexes) with regulatory proteins. Those necessary interactions might be not present in the exploited experimental system. The question remains whether forming of signalplexes is always a surplus value. Anchoring diverse subunits in a signalplex might even be functionally undesirable; for instance, coupling guanine triphosphate (GTP)-binding proteins in a signalplex could reduce their amplifier function.

In addition, when using proteomic approaches to search for channel binding partners, it is worth remembering that functional couplings between channels and transporters exist that do not require physical interaction. Thus, it is worth considering that activation of K+ or Cl channels can tune the driving force for Ca2+ entry, or that modulating the forward versus the reversed mode for Na+-Ca2+ exchange can dramatically affect SOC.

Focusing on CRAC, which is the only reliably described SOC so far, the crucial experiment will determine whether the correct pore properties of CRAC will emerge in a heteromer or a signalplex. This is not yet known. There is no doubt that multimerization occurs for many TRPs; these heteromers include the complexes TRPC1-TRPC4-TRPC5, TRPC3-TRPC6-TRPC7, TRPV5-TRPC6, TRPM4-TRPC5, and TRPM6-TRPC7. Further, it is clear that heteromer formation changes the permeation and kinetic properties of these channels (28). Partnership with other proteins acting in a signaling network has been most impressively shown for Drosophila TRPCs (29). So far, many modulators of mammalian TRPCs have been identified. These include calmodulin, the IP3 receptor (IP3R), ankyrin, the Na+-H+ exchanger regulator factor (NHERF), phosphoinositide 3-kinase (PI3K), annexin 2, the Ca2+-binding protein S100A10, caveolin-1, 80K-H [a protein kinase C (PKC) substrate with as-yet unclear biological function], phospholipase C-gamma (PLC-γ), the neurotrophin nerve growth factor receptor TrkA, recombinase gene activator (RGA), and microtubule-associated protein 7 (MAP7), all of which probably act through direct binding to TRPCs [some are reviewed in (6)]. The most intriguing assembling function might be credited to the adaptor protein Homer, which couples TRPC1 to the IP3R. Disassembly of the TRPC1-homer-IP3R complex parallels TRPC1 activation (30). However, even this example provides more evidence for TRPC1 modulation by protein-protein interaction than for TRPC1 being an SOC.

The bottom line is that none of the many interactions described so far in the TRPC family has clearly identified an SOC. Thus, in my opinion, SOC as a molecular entity is still elusive. We must keep open minds and look beyond the now somewhat absorbing TRP hypothesis.

Editor's Note: This Perspective is part of a series related to an E-Conference held at Science's STKE. Invited participants were asked to provide Perspectives that summarized the ideas, conclusions, areas of controversy, and challenges for future work that emerged in the online discussion at;14.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
View Abstract

Stay Connected to Science Signaling

Navigate This Article