Capacitative Ca2+ entry links the emptying of intracellular Ca2+ stores to the activation of store-operated Ca2+ channels in the plasma membrane. In the twenty years since the inception of the concept of capacitative Ca2+ entry, a number of activation mechanisms have been proposed, and there has been considerable interest in the possibility that TRP channels function as store-operated channels. However, in the past two years, two major players in both the signaling and permeation mechanisms for store-operated channels have been discovered: Stim1 and the Orai proteins. Stim1 is an endoplasmic reticulum Ca2+ sensor. It appears to act by redistributing within a small component of the endoplasmic reticulum, approaching the plasma membrane, but does not seem to translocate into the plasma membrane. Stim1 signals to plasma membrane Orai proteins, which constitute pore-forming subunits of store-operated channels.
In a variety of cell types, activation of phospholipase C by G-protein-coupled receptors results in release of intracellular Ca2+ stores and activation of Ca2+ entry across the plasma membrane. The intracellular release of Ca2+ is most commonly signaled by the second messenger inositol 1,4,5-trisphosphate [Ins(1,4,5)P3]. Early work on mechanisms of Ca2+ mobilization revealed that Ca2+ entry involves signaling from depleted intracellular stores to plasma membrane Ca2+ channels, a process referred to as capacitative Ca2+ entry or store-operated Ca2+ entry. In 2001, my colleagues and I reviewed the state of knowledge about capacitative Ca2+ entry for Journal of Cell Science, emphasizing theories on the activation mechanism and the nature of the channels (Putney et al., 2001). Embarrassingly, while the phenomenology we summarized at that time was correct, almost all of the favored ideas about activation mechanisms and channels have turned out to be either incorrect or, at the very least, incomplete. In just the past two years, major breakthroughs have revealed key players that initiate signaling from depleted stores and form store-operated channels themselves. In this Commentary, I provide a snapshot of current knowledge about the new players, Stim1 and the Orai, speculate on how they act, and make some guesses as to where this rapidly moving field is headed. I first, however, briefly review the concept of capacitative Ca2+ entry and describe some of the pharmacological and biophysical properties that are important for assessing the actions of the newly discovered components.
Capacitative Ca2+ entry
Capacitative Ca2+ entry, also called store-operated Ca2+ entry, simply refers to a phenomenon in which depletion of intracellular Ca2+ stores leads to activation of Ca2+-permeable channels in the plasma membrane. The Ca2+ entering through the store-operated channels can then be pumped into the stores and thus replenish them. Actions of relatively specific inhibitors of sarcoplasmic-endoplasmic reticulum Ca2+ (SERCA) pumps (discussed below) and clear localization of structures sensitive to the Ca2+-releasing action of Ins(1,4,5)P3 indicate that the intracellular store involved is the endoplasmic reticulum, or a component of it. The function of this pathway is, at the very least, to maintain sufficient Ca2+ in the endoplasmic reticulum, where it is required for proper protein synthesis and folding (Sambrook, 1990; Teasdale and Jackson, 1996). Indeed, prolonged depletion of endoplasmic Ca2+ leads to a stress response and ultimately to apoptosis (Tsukamoto and Kaneko, 1993; Bian et al., 1997). The pathway is also needed to support ligand-activated Ca2+ signaling, essential for a wide variety of regulated physiological pathways (Berridge et al., 2000).
Thapsigargin is a specific inhibitor of SERCA pumps on the endoplasmic reticulum. Its application to cells results in passive depletion of intracellular Ca2+ stores and thereby activation of the store-operated channels (Takemura et al., 1989). The ability to activate Ca2+ entry simply by blocking these pumps led to a number of important conclusions. First, entry can be fully activated with no discernible activation of phospholipase C. Second, additivity experiments showed that, at least in the parotid cells initially studied, there is no additional pathway activated by phospholipase-C-linked receptors. Third, because thapsigargin activates Ca2+ entry to the same extent that phospholipase C does, but without increasing Ins(1,4,5)P3 levels, passage of Ca2+ through the endoplasmic reticulum is not required for it to access the cytoplasm (i.e. the permeability of the endoplasmic reticulum membrane to Ca2+ does not determine the rate at which Ca2+ enters the cytoplasm). This and other arguments favoring a transmembrane flux of Ca2+ rather than transit through intracellular stores were summarized in an earlier review (Putney, Jr, 1990).
But perhaps the most significant outcome of the discovery of thapsigargin and other SERCA-inhibiting drugs was that it provided a functional and pharmacological diagnosis of capacitative entry: a thapsigargin-induced sustained elevation of intracellular Ca2+ that is dependent upon extracellular Ca2+ is generally attributed to capacitative Ca2+ entry – although additional pharmacological criteria may sometimes be considered (Putney, 2001). Notably, the ability of thapsigargin to specifically activate capacitative Ca2+ entry, while minimizing the roles of other upstream players in the pathway (receptors, G proteins, phospholipase C, etc.), formed the basis for the high-throughput assays that led to the discoveries of the functions of both Stim and Orai.
A decade and a half ago, Hoth and Penner discovered and characterized a very small but highly Ca2+-selective current, which they termed Icrac for Ca2+-release-activated Ca2+ (CRAC) current (Hoth and Penner, 1992; Hoth and Penner, 1993). This current was identified earlier by Lewis and Cahalan although it was not recognized as a store-operated current (Lewis and Cahalan, 1989). Icrac shares some properties with previously studied Ca2+-selective voltage-activated currents. For example, Icrac demonstrates an anomalous mole-fraction effect*; removing external Ca2+ leads to loss of current, unless all external divalent cations are removed by use of chelators, in which case the CRAC channels become permeable to monovalent cations and the current actually becomes larger. However, Icrac differs from other known currents in one important way: the single-channel conductance is too small to be measured directly and can only be estimated indirectly by noise analysis (Zweifach and Lewis, 1993). The molecular structure of CRAC channels might therefore be very different from those of other known Ca2+ channels (this has turned out to be the case). In addition, Icrac shows unusual regulation by Ca2+; intracellular Ca2+ inhibits CRAC channels by multiple mechanisms (Hoth and Penner, 1993; Zweifach and Lewis, 1995), whereas extracellular Ca2+ potentiates channel function (Zweifach and Lewis, 1996). These and a number of other findings have provided a rigorous fingerprint for the CRAC channel that eventually aided its correct molecular identification.
The anomalous mole fraction effect refers to a phenomenon whereby a mixture of two permeant species produces a current less than that obtained with either component of mixture alone (Almers and McCleskey, 1984). In the case of Ca2+ channels, the most striking instance of such behavior is seen with divalent-cation-selective channels that pass essentially no monovalent cations unless the concentration of divalent cations is very low, generally in the low μM range or less. The favored interpretation is that when divalent cations are present, they bind to and reside for relatively long times on a site within the channel known as a selectivity filter (Hess and Tsien, 1984). Ions are thought to permeate such channels in single-file, the rate for a particular species depending both on the permeability of the pore for that species as well as the affinity or off-rate of the selectivity filter. Thus, when Ca2+ concentration is reduced below its affinity for the selectivity filter, the site is no longer occupied for the majority of the time, and more permeable Na+ ions can freely permeate.
Activation mechanisms abound
Numerous reviews have summarized the ideas that have come in and out of vogue in the last twenty years regarding the signaling mechanism for capacitative Ca2+ entry (Putney and Bird, 1993; Sage et al., 1994; Berridge, 1995; Favre et al., 1996; Putney, 1997; Parekh and Penner, 1997; Putney and McKay, 1999; Barritt, 1999; Birnbaumer et al., 2000; Elliott, 2001; Gill and Patterson, 2004; Parekh and Putney, 2005; Bolotina and Csutora, 2005; Rosado et al., 2005). Two fundamental mechanisms have been proposed for transmitting the signal from intracellular stores to the plasma membrane: (1) a diffusible message (Randriamampita and Tsien, 1993; Bolotina and Csutora, 2005) and, (2) conformational coupling of proteins in the intracellular store membranes to plasma membrane channels (Irvine, 1990; Berridge, 1995). It has also been suggested that the activation mechanism might involve the insertion of active channels into the plasma membrane by vesicle secretion (Somasundaram et al., 1995; Yao et al., 1999; Scott et al., 2003). However, this idea does not address the more fundamental question of how the signaling is initiated from Ca2+-depleted stores. This concept is therefore not discussed further here.
Early ideas about capacitative Ca2+ entry envisioned specialized structures permitting direct access of external Ca2+ to the intracellular stores (Putney, 1977; Casteels and Droogmans, 1981). When it was later realized that Ca2+ enters the cell directly across the plasma membrane into the cytoplasm, there was no apparent connection between the stores and the channels, leading to the idea that a diffusible signal might be involved (Putney, 1990). However, there is precedent for direct plasma-membrane–sarcoplasmic-reticulum communication in the case of the very specialized skeletal muscle cells. In the process of excitation-contraction coupling in skeletal muscle, plasma membrane dihydropyridine receptors (voltage-dependent Ca2+ channels) interact with, and directly activate, intracellular ryanodine Ca2+ release channels. This led Irvine (Irvine, 1990) to propose an analagous conformational coupling model to explain regulated Ca2+ entry in cells that utilize the phospholipase-C–Ins(1,4,5)P3 signaling mechanism: endoplasmic reticulum Ins(1,4,5)P3 receptors would interact directly with plasma membrane capacitative Ca2+ entry channels (see also Berridge, 1995).
In the case of skeletal muscle, information flows from the t-tubule membrane to the sarcoplasmic reticulum; in the case of the Ins(1,4,5)P3 receptor and capacitative Ca2+ entry, a fall in luminal Ca2+ levels in the endoplasmic reticulum was proposed to induce a conformational change in the Ins(1,4,5)P3 receptor, and this would be conveyed directly to the plasma membrane channel by protein-protein interaction. This was an intriguing idea, but there was little direct evidence for it. There is, however, evidence for a requirement for close spatial association between endoplasmic reticulum and plasma membrane store-operated channels. Jaconi et al. used centrifugation to redistribute the organellar contents of oocytes and found that entry only occurred in regions with closely apposed endoplasmic reticulum (Jaconi et al., 1997). Patterson et al. (Patterson et al., 1999) used drugs to stimulate peripheral actin polymerization and disrupted communication between Ca2+ stores and plasma membrane store-operated channels (see also Bakowski et al., 2001). Golovina et al. demonstrated that in astrocytes, store-operated Ca2+ entry occurred at sites of close plasma membrane–endoplasmic-reticulum apposition (Golovina et al., 2005).
Numerous publications have argued that a diffusible signal couples depletion of intracellular Ca2+ stores to Ca2+ entry, and some evidence, albeit more controversial, supports the involvement of specific mediators. Two laboratories published findings suggesting cyclic GMP is involved (Pandol and Schoeffield-Payne, 1990; Xu et al., 1994); others have implicated arachidonic acid (or one of its metabolites) (Graier et al., 1995; Gailly, 1998; Rzigalinski et al., 1999). However, it now seems likely that these mediators act on channels distinct from the store-operated ones (Thompson, 1997; Mignen et al., 2005). The one candidate for a diffusible signal for store-operated channels that has withstood the test of time is one whose structure is not yet known: a Ca2+-entry activator partially purified from store-depleted cells called CIF (for Ca2+ influx factor) (Randriamampita and Tsien, 1993; Thomas and Hanley, 1995; Bolotina and Csutora, 2005). A role for such a messenger within the context of the new Stim1 and Orai story is certainly well within the realm of possibility.
The TRP story
Hardie and Minke first pointed out that the Drosophila photoreceptor Ca2+ channel, TRP, is activated downstream of phospholipase C and thus is a candidate for a store-operated channel (Hardie and Minke, 1992; Hardie and Minke, 1993). When its closest mammalian relatives were cloned (the seven canonical TRPs, TRPCs), a number of laboratories aggressively pursued this possibility. Early results were encouraging (Zhu et al., 1996; Zitt et al., 1996; Kiselyov et al., 1998). However, it soon became apparent that many of the early findings either resulted from misinterpretation of constitutive activity of overexpressed channels or could not be reproduced in other laboratories (Zhu et al., 1998; Trebak et al., 2003). In addition, subsequent work on the Drosophila TRP channel demonstrated that this is clearly not a store-operated channel (Acharya et al., 1997). Nonetheless, there are numerous reports of diminished store-operated entry when TRPC expression is supressed or eliminated (e.g. Mori et al., 2002), and others that are reviewed in Parekh and Putney (Parekh and Putney, 2005). Also, it is clear that, under some circumstances, TRPC channels can exhibit what appears to be store-operated activity when ectopically expressed (Vazquez et al., 2001; Vazquez et al., 2003; Lievremont et al., 2004; Zeng et al., 2004). The major problem is that, when experimentally introduced into cells, TRPC channels do not recapitulate the properties of Icrac; rather, they form non-selective cation channels or, at best, channels that show only modest Ca2+ selectivity. More significantly, they have single channel conductances of conventional size (tens of picosiemens). A number of laboratories have demonstrated that TRPC channels can function physiologically as non-store-operated channels activated downstream of phospholipase C, like the original Drosophila channel (Putney, 2005; Dietrich et al., 2005). This does not, however, rule out the possibility that TRPCs also function as store-operated channels in different contexts, for example, when assembled in signaling complexes of differing composition. There are a number of examples of store-operated Ca2+ entry that have properties distinct from Icrac (e.g. Krause et al., 1996; Trepakova et al., 2000), others are reviewed in Parekh and Putney (Parekh and Putney, 2005), and perhaps TRPCs function as store-operated channels in such instances.
Almost simultaneously, two laboratories discovered the role of Stim1 (initially Drosophila Stim) in capacitative Ca2+ entry by use of limited RNAi screens for modulators of thapsigargin-activated Ca2+ entry in Drosophila S2 cells (Roos et al., 2005) or in mammalian HeLa cells (Liou et al., 2005). Drosophila has a single Stim gene, whereas mammals have two, Stim1 and Stim2. Roos et al. found that knocking down Stim1, but not Stim2, reduces store-operated entry and Icrac in mammalian cells. However, Liou et al. reported a slight reduction in Ca2+ entry in HeLa cells in response to knocking down Stim2 as well. We have found no effect of knockdown of Stim2 in HEK293 cells (Wedel et al., 2007) and no effect of expression of Stim2 in HEK293 cells, even when co-expressed with Orai1 (Mercer et al., 2006). Soboloff et al. reported that Stim2 can act as an inhibitor of store-operated entry (Soboloff et al., 2006a); however, this action is only seen when Stim2 is expressed at very high levels. Thus, the true physiological function of Stim2 requires further investigation.
It seems clear that the function of Stim1 is to act as the initial sensor of Ca2+ levels in the endoplasmic reticulum or the component of it involved in regulation of Icrac. The Ca2+-sensing domain is an EF-hand that is N-terminal to the single transmembrane segment, and oriented in the lumen of the endoplasmic reticulum. Several laboratories have demonstrated that mutations in the EF-hand region, which presumably reduce Ca2+ affinity, result in constitutive activation of store-operated entry and Icrac. (Liou et al., 2005; Zhang et al., 2005; Spassova et al., 2006; Mercer et al., 2006).
Experiments examining the cellular distribution of Stim1 by histochemistry or fluorescently tagged Stim1 with real-time imaging techniques have provided intriguing information on the cell biology of this Ca2+ sensor. Liou et al. reported that Stim1 appeared to be located in the endoplasmic reticulum (Liou et al., 2005). However, we observed Stim1 with enhanced yellow fluorescent protein fused to its N-terminus (EYFP-Stim1) in fibrillar, or possibly tubular, patterns in HEK293 cells (Mercer et al., 2006); these patterns are not apparent from labeling of generic endoplasmic reticulum in the same cell type (Ribeiro et al., 2000). A similar conclusion was reached by Baba et al., who imaged Stim1 and endoplasmic reticulum markers within the same cell (Baba et al., 2006). In that study, the unique distribution of Stim1 was apparently dependent on the microtubular cytoskeleton (Baba et al., 2006). Thus, Stim1 may reside on a distinct organelle or, more likely, a distinct compartment within the endoplasmic reticulum. This would be consistent with a number of previous studies that concluded that the pool of Ca2+ that regulates store-operated channels is a small component of the total endoplasmic reticulum (Ribeiro and Putney, 1996; Parekh et al., 1997; Huang and Putney, 1998; Broad et al., 1999; Turner et al., 2003; Wisnoskey et al., 2003).
An even more exciting finding is that dissociation of Ca2+ causes Stim1 to redistribute into punctate structures and move closer to the plasma membrane (Liou et al., 2005; Zhang et al., 2005; Spassova et al., 2006; Mercer et al., 2006). However, a major point of controversy remains the subcellular localization of Stim1 when it redistributes into punctae. Liou et al. reported that EYFP-Stim1 moved close to the plasma membrane but, using an antibody directed against the EYFP on the N-terminus, they failed to detect it on the plasma membrane (Liou et al., 2005). By contrast, on the basis of surface biotinylation studies, Zhang et al. reported that Stim1 actually translocates into the plasma membrane (Zhang et al., 2005). A third position was taken by Spassova et al. (Spassova et al., 2006). They reported that a fraction of Stim1 is present in the plasma membrane, where it appears to function in store-operated entry (this conclusion was based on the inhibition of entry by extracellular application of an anti-Stim1 antibody), but they failed to see any change in the amount of plasma membrane Stim1 following Ca2+ store depletion.
Mercer et al. (Mercer et al., 2006) have repeated the experiments performed by Liou et al. (Liou et al., 2005), using the same EYFP-tagged Stim1. These authors confirmed that no EYFP-Stim1 can be detected on the plasma membrane of HEK293 cells by surface antibody staining and confocal microscopy or by flow cytometry (Mercer et al., 2006). Xu et al. reached a similar conclusion, on the basis of studies using a Stim1 N-terminally tagged with a pH-sensitive fluorescent indicator (Xu et al., 2006). Likewise, Wu et al. failed to observe plasma membrane Stim1 by immuno-electronmicroscopy, by using an N-terminally horsesradish-peroxidase-tagged Stim1 (Wu et al., 2006). However, Spassova et al. clearly observed surface labeling of Stim1 by flow cytometry, using an untagged construct and an antibody directed against the N-terminus (Spassova et al., 2006). Furthermore, the first report on Stim1, prior to the discovery that it acts as an initiator of capacitative Ca2+ entry, identified it as a surface protein on stromal cells involved in their interaction with B lymphocytes (Oritani and Kincade, 1996), and this surface localization was substantiated in subsequent studies (Manji et al., 2000; Williams et al., 2002).
A recent report by (Hauser and Tsien, 2007) provides an explanation for these disparate findings: the large additions, such as EYFP, to the N-terminus of Stim1 prevent its trafficking to the plasma membrane. This is actually fortuitous, since it allows us to assess the function of Stim1 by using a construct restricted to intracellular sites. EYFP-Stim1 is fully functional, both rescuing store-operated entry after RNAi-mediated knockdown of Stim1, and generating constitutive Ca2+ entry following mutation of the EF hand domain (Liou et al., 2005; Mercer et al., 2006). Similarly, Baba et al. found that N-terminally Flag-tagged Stim1 could not be detected in the plasma membrane of a B-cell line but still fully supports store-operated Ca2+ entry (Baba et al., 2006). In this case, they knocked out Stim1 by targeted gene disruption, leaving no doubt that the N-terminally tagged Stim1 can function in the absence of any residual wild-type protein. As I discuss below, co-expression of Stim1 with Orai1 greatly increases store-operated entry; however, even these very large Icrac-like currents are supported by EYFP-Stim1, and again no surface expression can be detected (Mercer et al., 2006). Finally, Huang et al. demonstrated that the cytoplasmic C-terminal region of Stim1, expressed in the cytoplasm, was sufficient to activate store-operated entry (Huang et al., 2006). Clearly this construct cannot produce any plasma membrane Stim1 in the correct orientation. My conclusion is that Stim1 does not translocate to the plasma membrane in response to Ca2+-store depletion, and that plasma membrane Stim1, although probably present, does not play an obligatory role in activating store-operated channels.
Orai1 was first reported (and named) by Feske et al. through a combination of gene mapping in a family with an immunodeficiency attributed to loss of Icrac and a whole-genome screen of Drosophila S2 cells (Feske et al., 2006). Almost simultaneously, two other groups, also using a similar whole-genome screen of S2 cells, reported that Orai [termed CRACM1 by Vig et al. (Vig et al., 2006a)] is essential for store-operated entry and for Icrac (Zhang et al., 2006; Vig et al., 2006a). Orai1 is located in the plasma membrane and appears to have four transmembrane domains. Unlike the case for Stim1, which was included in restricted screens based on its signaling domains, Orai1 has no recognizable signaling or channel-like domains and thus required the full genome screen for its detection. When Orai and Stim are co-expressed, they synergize to generate unusually large Icrac-like currents (Zhang et al., 2006; Mercer et al., 2006; Peinelt et al., 2006; Soboloff et al., 2006b), which are indistinguishable from native Icrac in both biophysical and pharmacological properties (Zhang et al., 2006; Mercer et al., 2006; Peinelt et al., 2006; Soboloff et al., 2006b). These two proteins can thus fully recapitulate the properties of Icrac. This does not necessarily mean that no other players are involved; however, other proteins functioning in a stoichiometric complex with Stim1 and/or Orai1 would have to be constitutively present in considerable excess.
The large Icrac-like currents observed upon co-expression of Stim1 and Orai1 immediately led to speculation that Orai1 is the CRAC channel itself or a subunit of it. However, there are no obvious channel-pore-like sequences in Orai1. Three laboratories focused on a string of acidic residues near the extracellular boundary of the first transmembrane domain (Prakriya et al., 2006; Yeromin et al., 2006; Vig et al., 2006b). The most interesting mutations affect glutamate 106 in mammalian Orai1, and the corresponding residue in Drosophila Orai. Mutation to alanine results in a non-functional channel; however, mutation to an aspartate (E106D; E180D in Drosophila Stim) results in a functional channel that has markedly reduced selectivity for Ca2+. Acidic residues just downstream from E180 in Drosophila Orai confer the well-known high sensitivity of Icrac to Gd3+ (Yeromin et al., 2006). These findings provide strong evidence that this region functions as part of the Ca2+ selectivity filter and indicate that Orai1 is indeed a pore-forming subunit of the CRAC channel.
Prakriya et al. (Prakriya et al., 2006) and Vig et al. (Vig et al., 2006b) have also investigated a glutamate residue at position 190. Mutation of this residue to aspartate or even alanine had no effect on channel function; however, alteration to glutamine (E190Q) resulted in diminished Ca2+ selectivity. Since the complete loss of charge from this amino acid (the mutation to alanine) had no effect on channel selectivity, the mutation to glutamine may alter the secondary or tertiary structure of the channel, rather than indicating a function as part of the Ca2+-binding site in the channel pore.
Mammalian cells have two additional proteins, Orai2 and Orai3. Orai2 appears to function similarly to Orai1, at least when expressed with Stim1 in HEK293 cells (Mercer et al., 2006). The Orai2 Ca2+ currents were somewhat smaller, however, and whether this is an intrinsic property of Orai2 channels or indicates a lower level of expression in this particular study is not known. Orai3 Ca2+ currents were even smaller; those resulting from co-expression of Stim1 and Orai3 in HEK293 cells were below the limits of detection. However, Orai3-dependent currents can be observed when Na+ carries the current (DeHaven et al., 2007), and Orai3 can rescue store-operated entry when Orai1 is knocked down in HEK293 cells (Mercer et al., 2006). Whether Orai2 and Orai3 form distinct store-operated channels in specific cell types, or function as subunits of heteromeric channels with Orai1 will be a topic of future investigation.
The evidence is very strong that the signaling pathway for capacitative Ca2+ entry begins with the Ca2+ sensor Stim1, and culminates in the activation of channels composed partly or wholly of Orai subunits. The obvious question is how Stim1 conveys information from depleted Ca2+ stores to Orai channels. The simplest answer is that, when Stim1 coalesces into punctae and approaches the plasma membrane, it directly interacts with Orai channels there. In support of this idea, Yeromin et al. reported that Drosophila Stim and Orai can be co-immunoprecipitated, and this association is increased by depletion of Ca2+ stores (Yeromin et al., 2006). However, Feske et al. failed to observe any co-immumoprecipitation of mammalian Stim1 and Orai1 (Feske et al., 2006). Vig et al. reported co-immunoprecipitation of transfected and tagged Stim1 and Orai1 but did not assess effects of store depletion (Vig et al., 2006b). Nonetheless, it is clear from the work of Luik et al. (Luik et al., 2006) and Xu et al. (Xu et al., 2006) that communication between Stim1 and Orai1 occurs over very short distances. These investigators observed that Orai1 is recruited to sites of Stim1 punctae formation, and Luik et al. (Luik et al., 2006) showed that Ca2+ entry is spatially restricted to these plasma membrane sites as well. Wu et al. estimated that Stim1 localizes in junctional endoplasmic reticulum within 10-25 nm of the plasma membrane (Wu et al., 2006). Finally, although a direct interaction between Stim1 and Orai seems the simplest mechanism for signaling from intracellular stores to the plasma membrane channels, there is nothing to rule out the generation of a secondary message downstream of Stim1 that acts on Orai – for example, the proposed CIF and iPLA2 pathway (Bolotina and Csutora, 2005).
In just the past two years, the discoveries of Stim and Orai proteins have revolutionized our thinking about capacitative Ca2+ entry and Icrac. It is remarkable how quickly different laboratories confirmed the dependence of Icrac on these two proteins and confirmed their ability to recapitulate the long-known properties of CRAC channels. In fact, it could be said that these are the only aspects of the Icrac story upon which all (or most) seem to agree. The current findings about Stim1 and Orai action are consistent with a number of earlier ideas about store-operated entry. For example, the presence of a small, specialized component of the endoplasmic reticulum dedicated to communicating with plasma membrane channels; the observation that biophysical properties of CRAC channels differ substantially from conventional ion channels, which indicated that they might also differ substantially in their structure; the general finding that Ca2+ refills the stores efficiently without causing a rise in global cytoplasmic Ca2+; and the suggested close physical association between sites of endoplasmic reticulum signaling and plasma membrane Ca2+ entry. One might argue that Irvine's original idea (Irvine, 1990) of conformational coupling is at least partially vindicated by the current model. However, it appears to be Stim1 rather than Ins(1,4,5)P3 receptors that are responsible for this coupling.
Fig. 1 summarizes the store-operated signaling pathway, incorporating the roles of Stim1 and Orai. Several general questions remain. (1) Are any other players or proteins necessary for activation of capacitative Ca2+ entry and/or Icrac? (2) What is the role, if any, of plasma membrane Stim1? (3) What is the composition of native CRAC channels – are these Orai homo- or heteromultimers or combinations with TRPs? (4) Are there mechanisms of store-operated entry involving other modes of activation and other store-operated channels? The availability of Stim1 and Orai cDNAs will lead to a detailed structural understanding of the various modes of regulation and activation of this pathway. The development of animal models will lead to a better understanding of the roles of these proteins in specific physiological and pathological processes. Such information may ultimately prove useful in design of novel pharmacological reagents to aid in the treatment of a number of diseases in which the store-operated entry pathway is thought to play a role.
David Miller and Stephen Shears read the manuscript and provided helpful comments. Research from the author's laboratory described in this review was supported by the Intramural Research Program of the NIH, National Institute of Environmental Health Sciences.