HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (280) Citing Articles via Google Scholar Google Scholar Articles by Parekh, A. B. Articles by Putney, J. W. Search for Related Content PubMed PubMed Citation Articles by Parekh, A. B. Articles by Putney, J. W., Jr.

Store-Operated Calcium Channels

Department of Physiology, University of Oxford, Oxford, United Kingdom; and National Institute of Environmental Health Sciences-NIH, Research Triangle Park, North Carolina

ABSTRACT

I. INTRODUCTION

II. STORE-OPERATED CALCIUM ENTRY: REFILLING THE STORES AND MORE

III. THE FUNDAMENTAL PROPERTY OF STORE-OPERATED CALCIUM CHANNELS IS ACTIVATION BY STORE DEPLETION

IV. MEASURING STORE-OPERATED CALCIUM INFLUX

V. ELECTROPHYSIOLOGICAL PROPERTIES OF STORE-OPERATED CALCIUM CURRENTS

A. ICRAC

1. Current-voltage relationship

2. Voltage-dependent CRAC conductance

3. Channel selectivity

4. Anomalous mole fraction

5. Monovalent permeation through CRAC channels: sizing the pore

6. Single CRAC channel conductance

B. Non-CRAC Store-Operated Currents

1. Store-operated channels in A431 epidermal cells

2. Store-operated channels in endothelia

3. Store-operated channels in vascular smooth muscle

4. Store-operated calcium channels in skeletal muscle

5. Neurons and neuroendocrine cells

VI. PHARMACOLOGY OF STORE-OPERATED CHANNELS

VII. ACTIVATION OF STORE-OPERATED CALCIUM ENTRY: NEED FOR A RETROGRADE SIGNAL

VIII. THE CALCIUM SENSOR

IX. ACTIVATION MECHANISM

A. Diffusible Messenger

1. Ca2+ influx factor

B. Conformational Coupling and Secretion-like Coupling

1. Store-operated channels potentially gated by InsP3 receptors

2. Movement of the ER

C. Vesicular Fusion

D. Removal of Ca2+ Inhibition

X. DO STORE-OPERATED CHANNELS DEACTIVATE BY REVERSAL OF ACTIVATION?

XI. REGULATION OF STORE-OPERATED CALCIUM ENTRY

A. Ca2+-Dependent Inactivation

1. Rapid inactivation

2. Store refilling

3. Slow inactivation

B. Sphingosine

C. cGMP and Protein Kinase G

D. Protein Kinase C

E. Arachidonic Acid

F. InsP4

XII. CHASING THE STORE-OPERATED CHANNEL GENE(S)

A. TRPCs

1. TRPC1

2. TRPC3, TRPC6, and TRPC7

3. TRPC4 and TRPC5

4. TRPC2

B. TRPV6 (CaT1)

C. Summary of TRPs

XIII. MITOCHONDRIAL REGULATION OF STORE-OPERATED CALCIUM ENTRY

XIV. QUANTITATIVE RELATIONSHIP BETWEEN STORE DEPLETION AND STORE-OPERATED ENTRY AND EVIDENCE FOR A SPECIALIZED CALCIUM STORE INVOLVED IN REGULATING ENTRY

XV. PHYSIOLOGICAL FUNCTIONS OF STORE-OPERATED CALCIUM INFLUX: SHORT-TERM RESPONSES

A. General Functions of Store-Operated Ca2+ Channels

B. Regulated Exocytosis

C. Regulation of Enzymatic Activity

D. Ca2+ Oscillations

E. Muscle Contraction

F. Sperm Chemotaxis and the Acrosome Reaction

XVI. PHYSIOLOGICAL FUNCTIONS OF STORE-OPERATED CALCIUM ENTRY: LONG-TERM RESPONSES

A. Gene Transcription

B. Cell Cycle

C. Apoptosis

XVII. PATHOPHYSIOLOGY

A. Severe Combined Immunodeficiency

B. Acute Pancreatitis

C. Alzheimer's Disease

D. Toxicology

XVIII. CONCLUSIONS

ACKNOWLEDGMENTS

REFERENCES

	ABSTRACT

Top Next References

	I. INTRODUCTION

Top Previous Next References

 
An increase in cytoplasmic Ca²⁺ concentration is used as a key signaling messenger in virtually every cell through the phylogenetic tree, where it regulates a broad spectrum of kinetically distinct processes (24, 53). Intracellular Ca²⁺ is often considered a "life-giving" signal because it is essential both for sperm motility and the acrosome reaction and, in the form of periodic Ca²⁺ oscillations, for fertilization of the egg. Ca²⁺ is also vital for sustaining the life of the organism. Distinct spatial and temporal patterns of cytoplasmic Ca²⁺ increases drive processes as diverse as exocytosis, gene transcription, and cell motility. Having given life, Ca²⁺ can also take it away. This dark side of the Ca²⁺ signal can be executed through apoptosis or the more destructive and less specific necrosis.

Eukaryotic cells can increase their cytoplasmic Ca²⁺ concentration in one of two ways: release from intracellular stores or Ca²⁺ influx into the cell. We now have a good understanding of the organelles that function as Ca²⁺stores and how Ca²⁺ can be released from them into the cytosol (27). Although the importance of the endoplasmic/sarcoplasmic reticulum (ER/SR) has been firmly established, growing evidence indicates that functional compartmentalization exists within the reticulum such that its Ca²⁺-releasing capabilities are not homogeneously distributed throughout the organelle. Instead, some subcompartments play a disproportionally greater role in Ca²⁺ release (290). Use of genetically targeted Ca²⁺ reporter proteins like aequorin and the cameleons together with detailed immunocytochemical mapping and functional studies have identified contributions from additional organelles like the Golgi apparatus, lysosomes, nuclear envelope (continguous with the ER), and possibly secretory granules (290, 339). In addition, mitochondria play a central role in intracellular Ca²⁺ dynamics under physiological conditions (53).

A relatively small complement of second messengers is thought to release Ca²⁺ from the stores. In addition to the ubiquitous second messengers Ca²⁺ and inositol 1,4,5-trisphosphate (InsP₃), roles for cyclic ADP ribose and nicotinic acid adenine dinucleotide phosphate (NAADP) have been described in some cell types (51). Just how a limited number of second messengers can generate the vast array of diverse intracellular Ca²⁺ release patterns consequent to receptor stimulation is unclear, but receptor-specific recruitment of different combinations of second messengers together with mobilization of distinct Ca²⁺ stores is likely to be of major significance.

In spite of its importance, the Ca²⁺ release phase is transient, sometimes fully deactivating within a few tens of seconds. This is in part due to Ca²⁺- and/or ligand-dependent inactivation of the release channels themselves as well as clearance of Ca²⁺ from the cytosol by resequestration into other organelles (notably mictochondria and ER) as well as extrusion from the cell by Na⁺/Ca²⁺ exchangers and Ca²⁺-ATPases in the plasma membrane. However, many key processes require sustained increases in intracellular Ca²⁺, and this is accomplished through Ca²⁺ entry into the cell.

The 10,000-fold concentration gradient for Ca²⁺ across the plasma membrane of resting cells coupled with a hyperpolarized resting membrane potential results in a huge electrochemical driving force in favor of Ca²⁺ influx. Resting cells generally have a low membrane permeability to Ca²⁺, but even modest increases in permeability result in large Ca²⁺ influx. An increase in membrane permeability to Ca²⁺ can be achieved by opening Ca²⁺-permeable ion channels in the plasma membrane.

A variety of different Ca²⁺-permeable channels have been found to coexist in the plasma membrane (134), and the major ones are depicted in Figure 1. Voltage-gated Ca²⁺ channels are found in excitable cells like nerve and muscle but are largely excluded from nonexcitable cells. Receptor-operated channels, which open rapidly upon binding an external ligand that is usually a neurotransmitter, are also preponderate in excitable cells. Second messenger-operated channels are less widely distributed and are found in some excitable and nonexcitable cells. Store-operated Ca²⁺-permeable channels on the other hand appear to be widespread, apparently existing in all eukaryotes from yeast (201) to humans (275). Hence, the argument could be advanced that store-operated Ca²⁺ channels represent the primordial Ca²⁺ entry pathway.

View larger version (33K): [in this window] [in a new window]   FIG. 1. Modes of regulated Ca2+ entry across the plasma membrane. Calcium can enter cells by any of several general classes of channels, including voltage-operated channels (VOC), second messenger-operated channels (SMOC), store-operated channels (SOC), and receptor-operated channels (ROC). VOCs are activated by membrane depolarization, and SMOCs are activated by any of a number of small messenger molecules, the most common being inositol phosphates, cyclic nucleotides, and lipid-derived messengers (diacylglycerol and arachidonic acid and its metabolites). SOCs are activated by depletion of intracellular Ca2+ stores, and ROCs are activated by direct binding of a neurotransmitter or hormone agonist (Ag). In addition, under some conditions, Ca2+ can enter cells via the Na+-Ca2+ exchanger (NCX) operating in reverse mode.

	II. STORE-OPERATED CALCIUM ENTRY: REFILLING THE STORES AND MORE

Top Previous Next References

	III. THE FUNDAMENTAL PROPERTY OF STORE-OPERATED CALCIUM CHANNELS IS ACTIVATION BY STORE DEPLETION

Top Previous Next References

 
The concept of store-operated Ca²⁺ entry was proposed in 1986 (297). This idea originated from a series of experiments in parotid acinar cells investigating the relationship between Ca²⁺ release from internal stores, Ca²⁺ entry, and store refilling. On the basis of this work, and a few eclectic observations in the literature, it was suggested that the amount of Ca²⁺ in the stores controlled the extent of Ca²⁺ influx in nonexcitable cells, a process originally called capacitative calcium entry. When stores were full, Ca²⁺ influx did not occur but, as the stores emptied, Ca²⁺ entry developed. Subsequently, more straightforward evidence for capacitative or store-operated entry was obtained through the use of thapsigargin, a specific inhibitor of the endoplasmic reticulum Ca²⁺ pump, discussed in more detail in a subsequent section. Direct evidence in support of the basic tenet of this model, namely, that store depletion activates Ca²⁺ influx, was provided by electrophysiological studies which established that the process of emptying the stores activated a Ca²⁺ current in mast cells called Ca²⁺ release-activated Ca²⁺ current or I_CRAC (143). I_CRAC is non-voltage activated, inwardly rectifying, and remarkably selective for Ca²⁺ (270, 417). It is found in several cell types mainly of hemapoietic origin. However, as Table 1 indicates, I_CRAC is not the only store-operated current, and it is now apparent that store-operated influx encompasses a family of Ca²⁺-permeable channels, with different properties in different cell types. Nevertheless, I_CRAC was the first store-operated Ca²⁺current to be described and remains a popular model for studying store-operated influx. As a consequence, our understanding of store-operated Ca²⁺ currents is heavily influenced by I_CRAC.

These methods of emptying stores are not devoid of potential problems. The key feature of store-operated Ca²⁺ entry is that it is the fall in Ca²⁺ content within the stores and not the subsequent rise in cytoplasmic Ca²⁺ concentration that activates the channels (270). However, ionomycin and SERCA pump blockers generally cause a rise in cytoplasmic Ca²⁺ concentration as a consequence of store depletion, and such a rise in Ca²⁺ could open Ca²⁺-activated cation channels permeable to Ca²⁺. One way to avoid such problems is to use agents under conditions where cytoplasmic Ca²⁺ has been strongly buffered with high concentrations of Ca²⁺ chelator such as EGTA or BAPTA. The recent discovery that TRPV6 channels are expressed in many nonexcitable cells has added further complication (72, 379). TRPV6 channels are Ca²⁺ selective but are inactivated at resting levels of Ca²⁺. Lowering cytoplasmic Ca²⁺ concentration, for example following dialysis with EGTA or BAPTA, removes the Ca²⁺ inhibition and the TRPV6 channels now open (379). Ca²⁺influx follows, but this is not store-operated. Clearly therefore, one needs to carefully dissect the dependence on lowering intraluminal Ca²⁺ concentration from both the subsequent rise in cytoplasmic Ca²⁺ as well as removal of Ca²⁺-dependent inhibition of other channels following dialysis with exogenous buffer. Hence, store-operated Ca²⁺ entry is best demonstrated using protocols that empty stores under conditions where cytoplasmic Ca²⁺ has been strongly buffered at close to resting levels (100 nM).

In some cell types, store-operated single-channel currents have been reported in the cell-attached configuration (2, 364, 414). In these experiments, after formation of a cell-attached patch, stores are depleted by either thapsigargin or receptor stimulation, and single-channel events are seen. A potential concern with this approach is that it may be the rise in Ca²⁺ itself or a Ca²⁺-dependent second messenger but not store depletion that gates the channels. Preincubating the cells with BAPTA-AM would eliminate a Ca²⁺-dependent current provided the cytosol accumulated enough free BAPTA to prevent a rise in cytoplasmic Ca²⁺ following store emptying. An alternative approach is to carry out whole cell and cell-attached recordings using two pipettes simultaneously, with the whole cell pipette being used to dialyze the cytosol with a high concentration of Ca²⁺ chelator.

Can receptor-evoked Ca²⁺ influx be entirely explained by a store-operated mechanism? In some nonexcitable cells, a solid body of evidence indicates that a variety of Ca²⁺ entry pathways exist. Inositol polyphosphate, cyclic nucleotide, Ca²⁺, and arachidonic acid-gated Ca²⁺-permeable channels have all been described and may contribute to agonist-evoked Ca²⁺ influx. In one instance, it has been concluded that agonist-activated entry did not occur by a store-operated mechanism at all, but rather by an entirely distinct mechanism, which the authors termed "ACE" (agonist-activated calcium entry). Reducing expression of either the phospholipase C (PLC)-1 or PLC-2 isoforms resulted in a pronounced decrease in agonist-evoked Ca²⁺ entry, but Ca²⁺ influx in response to thapsigargin was unaffected. Hence, PLC- was not required for the activation of store-operated Ca²⁺ entry, but was essential for the ability of receptor stimulation to evoke Ca²⁺ influx (276). The requirement for phospholipase C- was independent of its enzymatic activity because a lipase-deficient mutant was equally effective. On the other hand, it was subsequently suggested that PLC- might have a structural role requiring its SH3 domains and which perhaps involved localization of specialized Ca²⁺stores to the receptors in the plasma membrane (300). In a more recent report, Nishida et al. (258) provided evidence that PLC- acts to amplify the agonist-activated PLC signal, necessary for store depletion and activation of store-operated channels (258). It is not clear then whether PLC- simply provides additional InsP₃ giving adequate release to activate the store-operated channels, or whether PLC- provides InsP₃ that is localized to regulate specialized stores that in turn are coupled to the entry channels.

As discussed above, the original concept of store regulation of Ca²⁺ entry into cells was termed "capacitative calcium entry," sometimes referred to as CCE. In the current literature, the more descriptive term store-operated Ca²⁺ entry (SOCE) is now more commonly used. The channels through which Ca²⁺ enters the cells are often called store-operated channels (SOC). These generic terms should be used when the detailed nature of a particular channel or current is not known, and more specific terms such as I_CRAC or CRAC channels should be reserved for instances in which the specific electrophysiological properties are those originally described for I_CRAC. Finally, there are a number of even more general terms in the literature that imply even more limited knowledge of mechanism, for example, noncapacitative calcium entry (NCCE) or agonist-activated calcium entry (ACE).

Research continues on the relative roles of store-operated and non-store-operated mechanisms in various physiological situations. Discussion of some of these issues, and other aspects of store-operated channels, can be found in the electronically published proceedings of a recent E-conference (104 and references therein).

	IV. MEASURING STORE-OPERATED CALCIUM INFLUX

Top Previous Next References

 
The issue of how best to measure Ca²⁺ influx is a recurring theme in the store-operated channel field. Several different methods have been employed, and these have been discussed in some depth in a previous review (270). The most popular method is to use fluorescent dyes like fura 2, which are easily loaded into cells in a noninvasive manner. In many studies, the cytoplasmic Ca²⁺ concentration is taken as a direct indication of store-operated Ca²⁺ influx. Such an approach is fraught with dangers. First, the membrane potential is not controlled and hence is free to fluctuate. Changes in membrane potential, should they occur, will alter the extent of store-operated influx, through changes in the driving force for Ca²⁺ entry. Second, the cytoplasmic Ca²⁺ signal is not a reliable indicator of Ca²⁺ influx because the former will be determined by the balance between Ca²⁺ influx, cytoplasmic Ca²⁺ buffering, and Ca²⁺ removal. Changes in the activity of SERCA pumps, mitochondrial Ca²⁺ uptake, plasmalemmal Na⁺-Ca²⁺ exchange, or Ca²⁺-ATPases will all alter the size of a cytoplasmic Ca²⁺ signal, if Ca²⁺ influx stays constant. In T lymphocytes, for example, plasmalemmal Ca²⁺-ATPases increase their activity substantially following Ca²⁺ entry through CRAC channels and so extrude Ca²⁺ more efficiently (19). Inhibiting the pumps would increase cytoplasmic Ca²⁺, but this would not reflect an increase in Ca²⁺entry. Third, the nature of the Ca²⁺ influx pathway is not clear. Store-operated Ca²⁺ entry is not the only Ca²⁺ influx mechanism in nonexcitable cells, and simply measuring cytoplasmic Ca²⁺ levels does not enable one to discriminate between the various possibilities. Even applying thapsigargin in Ca²⁺-free solution then observing a Ca²⁺ signal on readmission of external Ca²⁺ is not unequivocal evidence for store-operated entry. Thapsigargin-evoked Ca²⁺ release can activate Ca²⁺-dependent channels permeable to Ca²⁺ and, in some cases, the channels are activated by a Ca²⁺/calmodulin-dependent protein kinase-mediated phosphorylation that can maintain channel activity for some time after cytoplasmic Ca²⁺ levels have returned to resting levels (41). There are a number of ways to reduce or eliminate these problems when using fluorescent indicators. Membrane potential can be controlled by voltage clamping with patch pipettes in the whole cell configuration. Alternatively, substitution of extracellular Na⁺ with K⁺ provides essentially a chemical clamp of membrane potential at K⁺ equilibrium potential (E_K), which will be close to 0 mV. It is possible to deplete intracellular stores without increasing cytoplasmic Ca²⁺, by using low concentrations of thapsigargin for prolonged periods (298). Finally, problems of Ca²⁺ buffering, Ca²⁺-mediated inactivation mechanisms, and Ca²⁺-activated channels can be reduced somewhat by the use of Ca²⁺ surrogates, such as Ba²⁺, which will pass through all known store-operated channels, but is not a substrate for Ca²⁺ transporters and does not activate most Ca²⁺-dependent processes to the extent that Ca²⁺ does. A direct way to eliminate these problems is to use the patch-clamp technique to monitor the store-operated Ca²⁺ current (Fig. 2). Using this method, complications from changes in membrane potential and Ca²⁺ clearance are eliminated. Moreover, by strongly buffering the pipette solution, changes in cytoplasmic Ca²⁺ can be suppressed. However, patch clamping is not without its own complications. In the whole cell configuration, which is the most popular way to study store-operated entry at the present juncture since the single-channel conductance of many, but not all, store-operated channels is well beyond the bandwidth of current patch-clamp amplifiers, important but often ill-defined, cytoplasmic factors diffuse into the pipette and are hence lost from the cell. For example, mitochondria are not in a highly energized state due to washout of oxidizable substrates, and therefore, these substrates need to be included in the pipette solution to support I_CRAC under physiological conditions of weak intracellular Ca²⁺ buffering (102, 103, 142). Moreover, isolation of I_CRACrequires inhibition of other currents, and hence, the ionic composition of the pipette solution is hardly physiological. Some of these problems can be avoided using perforated patch recordings. Finally, in some cell types, the store-operated currents appear to be very small, limiting the usefulness of direct current measurement in such instances. The one advantage of fluorescent indicators in these cases is their greater sensitivity to small Ca²⁺ fluxes compared with direct current measurement. Nevertheless, patch clamping remains the most direct and unambiguous method for studying store-operated influx. It is probably no exaggeration to say that a lot of the discordant results and controversies in the store-operated Ca²⁺ influx field emanate from differences in the techniques used. Ideally, it is probably best to use a combination of experimental approaches.

View larger version (19K): [in this window] [in a new window]   FIG. 2. Fundamental features of the store-operated Ca2+ current ICRAC. A: time course of development of ICRAC is depicted following store depletion in RBL-1 (rat basophilic leukemia) cells by including inositol 1,4,5-trisphosphate (InsP3; ) in the patch pipette or by applying ionomycin () onto the cell. In both cases, the whole cell patch-clamp technique has been used and Ca2+ has been strongly buffered at 120 nM (5 mM free EGTA) in the pipette solution. The voltage-ramp protocol for measuring ICRAC is shown at the top. The bottom panel in A shows the current-voltage relationship for ICRAC, taken when the current has reached steady-state. The current is inwardly rectifying and has a very positive reversal potential (>+70 mV), indicating high selectivity for Ca2+. The current-voltage relationship has the same features irrespective of how stores are emptied. B: a general feature of CRAC channels is that they show fast Ca2+-dependent inactivation. Permeating Ca2+ feedback locally (within a few nanometers of the pore) to reduce channel activity. Fast inactivation can be revealed by applying hyperpolarizing pulses, as shown in the top panel. After depleting store, stepping the voltage from 0 to –120 mV results in an initially large ICRAC (due to the large electrical driving force), but the current then inactivates over a few milliseconds during the hyperpolarizing pulse. The extent of inactivation is 60% when the slow Ca2+ chelator EGTA is included in the pipette, but it is considerably less when the fast Ca2+ chelator BAPTA is used instead. The bottom panel shows the extent of inactivation at different voltages in the presence of EGTA () and BAPTA ().

	V. ELECTROPHYSIOLOGICAL PROPERTIES OF STORE-OPERATED CALCIUM CURRENTS

Top Previous Next References

1. Current-voltage relationship

Ca²⁺ release-activated Ca²⁺ current is a non-voltage-gated current in that, unlike Ca²⁺ currents through the Cav family, it is not opened by membrane depolarization. Once activated, I_CRAC has a characteristic current-voltage relationship (Fig. 2). The current amplitude is large at negative potentials and approaches the zero current level at very positive potentials (>+60 mV). In the standard experimental paradigm to monitor I_CRAC(voltage ramps spanning –100 to +100 mV in 50 ms), the current-voltage relationship reveals a prominent inward rectification at negative voltages (143, 270). Part of this rectification is a consequence of the asymmetric Ca²⁺ concentrations used to measure the current (usually 10 mM outside and a few nanomolar Ca²⁺ inside), which would give rise to some rectification as predicted by the Goldman-Hodgkin-Katz theory. In addition, rectification is accentuated during the ramp protocol by Ca²⁺-dependent inactivation of the CRAC channels, which leads to a modest steepening of the current-voltage curve at negative potentials. Consistent with this is the finding that, in divalent-free external solution, Na⁺ readily permeate through CRAC channels (see sect. VA5) and now rectification is less steep (see Fig. 1B in Ref. 17 for superimposition of current-voltage plots with Ca²⁺ and Na⁺ as the charge carriers).

The non-voltage-gated behavior and inward rectification are sometimes taken as unequivocal evidence for the presence of I_CRAC. However, inward rectification and non-voltage-gated behavior are shared by other Ca²⁺-selective channels including both TRPV5 and TRPV6, neither of which appear to be store-operated (72, 381). Inward rectification and non-voltage-dependent gating are not unique to I_CRAC and therefore should not be considered as diagnostic of the current.

2. Voltage-dependent CRAC conductance

Although CRAC channels are not opened by membrane depolarization, they nevertheless exhibit a slow voltage dependence at least in RBL-1 cells (14, 16). Whole cell CRAC conductance appears to be, directly or indirectly, voltage dependent in that hyperpolarizing holding potentials reduce the size of the current, whereas depolarization increases it, when the current amplitude is measured at –80 mV. Voltage jump relaxation experiments reveal that the voltage-dependent conductance changes develop and reverse slowly, with time constants of several seconds (14). This voltage dependence of the whole cell CRAC conductance can be seen in divalent-free external solution where Na⁺ is the charge carrier, although it is less prominent than when Ca²⁺ is the charge carrier. Hence, the slow voltage dependence cannot be wholly explained by Ca²⁺ entry-dependent inactivation of I_CRAC. The physiological relevance of voltage-dependent I_CRAC remains to be established. Quite large fluctuations in membrane potential have been reported in RBL-1 cells following store depletion (88, 222). Depolarization of the membrane potential would reduce the electrical driving force for Ca²⁺ entry through CRAC channels, but this would be compensated somewhat by the increased macroscopic conductance imparted by the voltage dependence. Hence, different cell-surface receptors may evoke distinct patterns of intracellular Ca²⁺ signals depending on their effects on the membrane potential.

3. Channel selectivity

With Ca²⁺ as the charge carrier, I_CRAC approaches zero at very positive voltages (>+60 mV), indicative of a high selectivity for Ca²⁺(270). Indeed, if extracellular Na⁺is replaced by large organic cations like N-methyl-D-glucamine (NMDG⁺), then neither the extent nor apparent reversal potential of I_CRAC is altered (144), even in the presence of physiological levels of external Ca²⁺ (1–2 mM; Ref. 90). Moreover, removal of external Ca²⁺ in the continuous presence of external Na⁺ and Mg²⁺ abolishes the current completely (90, 143, 417). Unlike current through voltage-operated Ca²⁺ channels, I_CRAC does not seem to support any detectable outward currents carried by K⁺ or Cs⁺, and this renders it difficult to establish a clear reversal potential for the current (270). Hence, the permeability ratio of Ca²⁺ to other ions has been hard to quantify. One way to assess the Ca²⁺ permeability of a Ca²⁺ channel is to relate the amount of Ca²⁺ entering per unit time (integral of the Ca²⁺ current) to the change in the Ca²⁺-dependent wavelength of fura 2, when this dye is the dominant Ca²⁺ buffer in the cell (252). The assumption is that all incoming Ca²⁺ through the channels is captured by the fura 2. With the use of this approach, it has been concluded that I_CRACin mast and RBL cells is more selective for Ca²⁺ than Cav channels (141). As the latter have a Ca²⁺:Na⁺ permeability ratio of 1,000:1 and Na⁺ outnumber Ca²⁺ by more than 70:1 under physiological conditions, CRAC channels are remarkably selective for Ca²⁺.

Permeability studies of other divalent cations like Ba²⁺ and Sr²⁺ through CRAC channels have been hampered by the fact that CRAC channels seem to require external Ca²⁺ to maintain their maximal activity, a process called calcium-dependent potentiation (60, 420). Because whole cell dialysis with high concentrations of BAPTA fail to affect calcium-dependent potentiation and the nonpermeating cation Ni²⁺can replace external Ca²⁺ in supporting channel activity, the Ca²⁺ binding site is thought to be extracellular. Ba²⁺ and Sr²⁺ do not support potentiation. Therefore, replacing external Ca²⁺ with either Ba²⁺ or Sr²⁺ results in a decline of channel activity, and steady-state current measurements lead to a significant underestimate of permeability of divalent cations that do not support potentiation. Rapidly replacing Ca²⁺with Ba²⁺ results in a transiently larger peak Ba²⁺ current (before depotentiation develops), indicating that CRAC channels actually conduct Ba²⁺ better than Ca²⁺.

4. Anomalous mole fraction

CRAC channels, like Cav channels, are thought to achieve high Ca²⁺ selectivity by high-affinity binding of Ca²⁺ within the channel pore, which prevents Na⁺ from permeating (193, 270). When external Ca²⁺ concentration is very low (submicromolar range), large Na⁺ currents readily flow through CRAC channels. As Ca²⁺ concentration increases to the micromolar range however, Na⁺ permeation is reduced as a Ca²⁺ occupies a high-affinity site within the channel (16, 192). The apparent dissociation constant (K_D) for Ca²⁺ block of the Na⁺ current is close to 10 µM in RBL-1 cells (16) and 4 µM in Jurkats (192). As external Ca²⁺ increases further to the millimolar range, mutual repulsion between two Ca²⁺ provides the driving force for selective Ca²⁺ permeation, with apparent K_D of 0.8 mM in RBL-1 cells (90), 2.1 mM in Jurkat T lymphocytes (293), and 3.3 mM in mast cells (144).

Similar behavior is also seen in mixtures of divalent cations Ca²⁺ and Ba²⁺ (141). The conductance of CRAC channels is lower in 10 mM external Ca²⁺ than equimolar Ba²⁺ solutions, but with mixtures of the two ions, the conductance falls to a level less than that seen in either pure Ca²⁺ or Ba²⁺ solutions. Such concentration-dependent permeability ratios are indicative of multi-ion pores and support the idea that CRAC channels select for Ca²⁺ over Na⁺ by high-affinity binding of Ca²⁺ to the selectivity filter.

5. Monovalent permeation through CRAC channels: sizing the pore

Like voltage-operated Ca²⁺ channels and TRPV5/6 channels, CRAC channels lose their selectivity in divalent-free external solution (17, 144, 192, 292). Now, Na⁺ can readily permeate the channels, resulting in whole cell currents that are five- to eightfold larger than the corresponding Ca²⁺ currents. In divalent-free bath solution and Mg²⁺/Mg²⁺-ATP in the pipette, the Na⁺ current develops with a similar time course to that of Ca²⁺ following store depletion, and the current is still inwardly rectifying, although to a slightly lesser extent than with Ca²⁺ (16, 17, 144, 292). Unlike the situation with Ca²⁺, with Na⁺as the permeating species a clear reversal potential can be discerned because of a small but resolvable outward current. Hence, it has become possible to study the selectivity of CRAC channels and hence estimate the minimum pore diameter (17, 292). In RBL cells, ion substitution experiments revealed that the permeability profile for monovalent cations through CRAC channels was Na⁺ = Li⁺ > Rb⁺ (0.67) > Cs⁺ (0.10), and relatively large organic cations like trimethylamine, tetraethylammonium, NMDG, and Tris were essentially impermeant (16). A similarly low P_Cs/P_Na has been found for CRAC channels in RBL-2H3 (381) and Jurkat T lymphocytes (292). This monovalent permeation profile corresponds to an Eisenmann sequence type X, indicative of a strong-field strength site (17). Such selectivity studies reveal some interesting differences between CRAC and Cav channels (17, 292). Cav1.1–1.4 channels are around six times more permeable to Cs⁺ than CRAC channels, and trimethylamine is conducted through Cavs but not CRAC channels. Trimethylamine has a molecular diameter of 0.55 nm, placing the estimated minimum pore size of Cav to be >0.6 nm. The corresponding minimum diameter of the CRAC channel is slightly larger than 0.32 nm (diameter of a Cs⁺) but <0.55 nm. CRAC channels, like Cavs, are multi-ion pores, and their selectivity is likely to be achieved by binding of Ca²⁺ to sites (aspartate or glutamate) lining the Ca²⁺-selective pore. In addition, the relatively small size of the CRAC channels suggests that molecular sieving may play an auxiliary role in determining channel selectivity and might explain why Cs⁺ is relatively impermeable (17).

6. Single CRAC channel conductance

With Ca²⁺ as the charge carrier, single CRAC channel openings have not been seen. Over a wide range of voltages, Hoth and Penner (144) failed to detect any increase in whole cell variance as I_CRAC developed in mast cells. They estimated the single-channel conductance to be significantly lower than 1 pS. Stationary noise analysis in Jurkat cells revealed a unitary chord conductance of 24 fS in isotonic Ca²⁺ solution, well beyond the typical bandwidth of a patch-clamp experiment (417). This conductance is almost 1,000-fold smaller than the single-channel conductance of most ion channels. Hoth and Penner (144) observed a small increase in the current variance when Na⁺ permeated the CRAC channels. Using this method in lymphocytes, Lepple-Wienhues and Cahalan (192) reported a unitary CRAC conductance of 2 pS. Kerschbaum and Cahalan (166) subsequently found that removal of Mg²⁺ from the pipette solution dramatically increased the size and duration of the monovalent current. Under these conditions of divalent-free solution on both sides of the Jurkat cell membrane, they detected single-channel events of 35–40 pS which developed with a time course that corresponded with passive store depletion and which was inhibited by extracellular divalent and trivalent cations (Ca²⁺, Mg²⁺, Ni²⁺, and Gd³⁺). The ability to record single CRAC channel activity opened up the possibility for accreting molecular details of the channels as well as directly investigating the activation mechanism. Indeed, recording single-channel events in the absence of external divalents, which was attributed to CRAC channels, has been used to probe both the regulation and gating mechanisms of these channels (42, 92, 323). Furthermore, the single-channel conductance formed one of the central pieces of evidence that the TRPV6 gene encoded the CRAC channel pore (403; see sect. XIIB). However, subsequent studies by several laboratories have now established that the 35- to 40-pS conductance channels are not CRAC channels (17, 132, 177, 292). Instead, they represent nonselective cation channels that are opened following the removal of intracellular Mg²⁺/Mg-ATP. These channels, called MagNuM or MIC, are quite widespread, being found in several cell lines including RBL-1, Jurkat, and HEK 293 cells, all of which are popular systems for studying I_CRAC (251). MagNuM/MIC is most likely encoded by TRPM7 gene (251, 322). The current through MagNum/MIC channels, unlike CRAC, is not regulated by store depletion, is much more permeable to Cs⁺, has a very different current-voltage relationship in both divalent-containing and divalent-free solution dominated, and has a different pharmacological profile (17, 130, 177, 292). A crucial point is that MagNuM/MIC is suppressed by millimolar levels of Mg²⁺/Mg-ATP in the recording pipette (251). Under these conditions, I_CRAC can be studied in relative isolation. In the presence of intracellular Mg²⁺, fluctuation analysis has revealed that the single-channel CRAC conductance in divalent-free external solution is 0.2 pS (292). This is an important result for several reasons. First, it establishes a biophysical hallmark of the CRAC channel that can be used to assess the validity of putative CRAC channel genes. Second, it raises intriguing questions concerning channel permeation. Cavs achieve high selectivity by binding Ca²⁺ within the pore. When the pores lack Ca²⁺, selectivity is compromised, and Na⁺, Cs⁺, and even large cations like trimethylammonium (TMA⁺) permeate with a high throughput rate. Even in the absence of Ca²⁺ and Mg²⁺ however, CRAC channels retain some selectivity by discriminating between monovalent cations, and the throughput rate of Na⁺ is relatively low. How this is achieved will require detailed structure-function studies once the channel has been cloned.

B. Non-CRAC Store-Operated Currents

These channels have not been as well studied as CRAC channels. Therefore, their biophysical features are sometimes sketchy. Basic features of these channels are compared with those of CRAC channels in Table 1.

1. Store-operated channels in A431 epidermal cells

Mozhayeva and colleagues (171, 405, 414) have described single store-operated channels from human A431 carcinoma cells. In cell-attached patches, these channels could be activated by stimulation of cell-surface receptors, by thapsigargin or, less frequently, by incubating cells in BAPTA-AM. In inside-out patches, channel activity could be induced by InsP₃ applied to the cytoplasmic side. With 105 mM BaCl₂ or CaCl₂ in the pipette solution, the reversal potential was estimated to be around +65 mV, indicating selectivity for divalent cations. Ba²⁺ and Ca²⁺ were equally permeable but 1,000 times more so than K⁺. The channels were blocked by SK&F-96365 and had a resolvable conductance of 1 pS, which increased to 6 pS with Na⁺ as the charge carrier. Channel mean open time was 7.7 ms, and the channels were voltage dependent in that open probability increased with hyperpolarizations beyond –40 mV. The channels have been referred to as I_min or I_CRACL (CRAC-like). The gating of these channels will be discussed in section IXB.

Lueckhoff and Clapham (207) have also reported a store-operated channel in A431 cells. These authors used a double patch approach in which one pipette was used to deplete the stores in the whole cell configuration and the second pipette was in the cell-attached mode. Following store depletion with either thapsigargin and high BAPTA or high BAPTA alone, 2-pS channels were opened in the cell-attached patch in the presence of 160 mM CaCl₂, and the conductance increased to 20 pS in the presence of Ba²⁺. Channel activity was transient, decaying within 4 min. Excision of the patch resulted in rapid rundown of channel activity, and this could not be recovered by InsP₃. There are some striking differences between this store-operated channel and I_min (171, 405), also reported in A431 cells. The channels differ in single-channel conductance, voltage dependence, and gating by InsP₃. The reason for these discrepancies is unclear.

2. Store-operated channels in endothelia

In bovine aortic endothelial cells, application of either the receptor agonists bradykinin or ATP or the SERCA pump blocker, di-tert-butylhydroquinone, activated Ca²⁺-permeable channels in cell-attached patches (369). The Ca²⁺:Na⁺ permeability ratio was estimated to be >10:1, and anomalous mole fraction was seen in mixtures of Na⁺ and Ca²⁺. In the absence of external Ca²⁺, the single-channel conductance was 5 pS and fell to 2.5 pS in 1 mM Ca²⁺. Raising Ca²⁺ to 10 mM increased the conductance to 11 pS. Channel activity was lost quite rapidly on excising the patch to the inside-out configuration but was less resistant to run down in the outside-out mode.

In calf pulmonary artery endothelial cells, a Ca²⁺-permeable current was described that could be activated by store depletion with InsP₃, the SERCA pump blocker di-tert-butylhydroquinone or ionomycin (83). The current was small, being only 20% that of I_CRAC in Jurkats at –80 mV. It was inwardly rectifying with a positive reversal potential and was blocked by micromolar concentrations of La³⁺. Perifusing cells with divalent-free solution increased the size of the current severalfold, and rectification was maintained somewhat. The authors concluded that calf pulmonary artery endothelia expressed a current very similar to I_CRAC.

In mouse aortic endothelial cells, Nilius and colleagues (96) have described an inwardly rectifying store-operated Ca²⁺-permeable current. On switching to divalent-free solution, the current amplitude increased threefold and the channels were quite permeable to Cs⁺ as well as Na⁺. Another interesting feature of this endothelial current was that Ca²⁺ blocked the monovalent flux with 20-fold higher affinity than that seen for CRAC channels.

Store-operated currents have been reported in several different types of endothelial cells, and the reader is referred to the excellent discussion of this issue in a recent review (256).

3. Store-operated channels in vascular smooth muscle

In mouse and rabbit aorta, 3-pS store-operated channels have been described (364). In cell-attached patches, the channels were activated by thapsigargin even after the cells had been exposed to BAPTA-AM, suggesting that they were not activated by the thapsigargin-evoked rise in cytoplasmic Ca²⁺ that accompanies store emptying. In 30% of the cells, incubation with BAPTA-AM alone was able to activate the channels. These channels were cation selective but did not discriminate between Na⁺, K⁺, Cs⁺, Ca²⁺, Ba²⁺, or Sr²⁺. In excised patches, the conductance did not change when Ca²⁺ (1 or 10 mM) was added to the Na⁺-containing pipette solution, indicating that the channels do not prefer Ca²⁺ over Na⁺ when both cations are present. However, with 90 mM Ca²⁺ in the pipette (and no Na⁺), a slope conductance of 2.7 pS was found. Hence, these channels are permeable to Ca²⁺, but it would appear that much of the current is carried by Na⁺(P_Ca:P_Na = 1). These channels were voltage dependent in that channel activity increased more than threefold at potentials positive to +50 mV. The channels were activated by Ca²⁺ influx factor (363), and this is discussed further in section IXA.

In myocytes from rabbit portal vein, single store-operated channels have also been described (1). In cell-attached recordings, the channels could be activated by the SERCA pump blocker cyclopiazonic acid, caffeine, incubating cells with BAPTA-AM, or the calmodulin antagonist W-7. In addition, spontaneous openings of the channels were also observed in the absence of any stimulation. For all cases, the current-voltage relationship was linear over the range –40 to –120 mV, with a slope conductance of 2–3 pS. No single-channel events were observed at positive potentials, so the reversal potential was estimated by linear interpolation to be +20 mV (126 Na⁺, 1.5 mM Ca²⁺). The open lifetime distributions could be fitted by the sum of at least two exponentials, yielding time constants of 5 and 30 ms. Unlike the channels reported in aortic myocytes (364), those in portal vein were affected by external Ca²⁺. In Ca²⁺-free solution, the slope conductance increased to 7 pS, and the reversal potential shifted to –4 mV. In isotonic CaCl₂, the conductance fell to 1.3 pS, and the reversal potential was estimated to be +80 mV. P_Ca:P_Na was calculated to be 50:1.

In proliferating pulmonary artery smooth muscle cells, cyclopiazonic acid activated single channels with a slope conductance of 5.4 pS in the presence of 120 mM Na⁺ and 20 mM Ca²⁺ (112). However, the selectivity of these channels was not explored. In human glomerular mesangial cells, which have contractile properties similar to smooth muscle cells, recordings from cell-attached patches revealed spontaneously active channels that were considered to be store-operated (212, 213). These channels had a slope conductance of 0.7 pS in 90 mM CaCl₂ and 2.1 pS in 90 mM BaCl₂, with estimated reversal potentials of +123 and +63 mV, respectively. Channel activity was not voltage dependent over the range 0 to –80 mV.

It is not always clear whether the cell-attached single-channel events described above are indeed due to store-operated channels as opposed to another Ca²⁺ influx pathway. The best evidence that these channels are store-operated is that channel activity is still seen after cells have been loaded with BAPTA-AM. However, it has not always been shown that sufficient BAPTA has accumulated in the cytosol to suppress the rise in Ca²⁺ following store depletion. Furthermore, for those channels with a P_Ca:P_Na of 1, most of the current will be carried by Na⁺ under physiological conditions. Very large inward currents would be required to elevate Ca²⁺ appreciably. The role of these nonselective channels in muscle, for example, might simply be to provide the depolarization that is necessary for the more Ca²⁺-selective voltage-gated Ca²⁺ channels to open.

4. Store-operated calcium channels in skeletal muscle

In mouse skeletal muscle, Ca²⁺ leak channels have been observed in resting cells (140). These channels were non-voltage-gated; exhibited a single-channel conductance of between 7 and 14 pS; did not distinguish between Ca²⁺, Ba²⁺, and Mn²⁺; and were inhibited by novel dihydropyridines like AN406 and AN1043 which only weakly affect voltage-operated Ca²⁺ channels. In cell-attached patches, channel open probability was increased by the SERCA pump blocker cyclopiazonic acid. Inhibition of the Ca²⁺ leak channels with AN406 reduced the extent of store refilling following store emptying, indicating that the Ca²⁺ leak channels could contribute to the reloading of the stores. Interestingly, the open probability of these Ca²⁺ leak channels was greater in resting dystrophic muscle cells than in resting normal myocytes. It has been suggested that this higher resting permeability to Ca²⁺ in dystrophic cells results in stimulation of the Ca²⁺-dependent protease calpain (4). One idea is that calpain might then alter the activity of the Ca²⁺ leak channels resulting in further Ca²⁺ influx. This kind of positive feedback could result in a gradual loss of Ca²⁺ homeostasis leading ultimately to cell death.

5. Neurons and neuroendocrine cells

In bovine adrenal chromaffin cells, store depletion with either thapsigargin or dialysis with 10 mM BAPTA activated a small, inward current at negative potentials (93). The current was nonselective, being carried by both Ca²⁺ and Na⁺. Because of the presence of voltage-gated Ca²⁺ channels, the store-operated current could only be measured at potentials more negative than –60 mV. The estimated reversal potential of the current was quite negative (–20 mV), although linear extrapolations do not take into account rectification and hence may underestimate the zero current potential. The small Ca²⁺-permeable current induced by thapsigargin was able to stimulate exocytosis at negative potentials as well as potentiate secretory responses following activation of voltage-operated Ca²⁺ channels. Although the rate of secretion to thapsigargin was 30- to 60-fold slower than that elicited by depolarizing pulses which open voltage-operated Ca²⁺ channels, nevertheless the total secretory response to thapsigargin was substantial, amounting to the fusion of 300–400 large dense-core vesicles. In chromaffin cells, movement of vesicles from the reserve pool to the ready-releasable pool is dependent on cytoplasmic Ca²⁺ and protein kinase C (106). Perhaps the main role for store-operated influx in exocytosis is to maintain the size of the ready-releasable pool that would be severely depleted following a train of action potentials.

The existence of store-operated entry in neurons has been much harder to establish. This reflects the complex architecture of neurons rendering it hard to reliably measure small currents that are not located exclusively on the soma, the presence of many ionic conductances which need to be eliminated to dissect out the store-operated pathway, and the fact that neurons coexpress a multitude of Ca²⁺-permeable channels (voltage operated, ligand gated, second messenger operated). Nevertheless, studies using SERCA pump blockers to deplete stores and fluorescent dyes to monitor Ca²⁺ influx have been interpreted as evidence for store-operated entry (301). A major concern with these sorts of experiments is that the Ca²⁺ influx pathway is not known. It could easily be a second messenger-operated one or, if membrane potential changes, even voltage-operated. The combination of patch-clamp recordings with microfluorimetry in dorsal root ganglion neurons has revealed that depletion of the caffeine-sensitive store evokes Ca²⁺ influx at hyperpolarized potentials (367). Ca²⁺ influx following store emptying was blocked by 2 mM Ni²⁺, was insensitive to antagonists of voltage-gated Ca²⁺ channels, and was facilitated by hyperpolarizing the membrane potential from –55 to –80 mV, a maneuver that increases Ca²⁺ influx due to the larger driving force. This Ca²⁺ influx pathway was required for refilling the caffeine-sensitive stores. Whether store-operated influx is more widespread in the nervous system and the nature of the underlying channels remain to be determined.

	VI. PHARMACOLOGY OF STORE-OPERATED CHANNELS

Top Previous Next References

 
Progress in the store-operated Ca²⁺ influx field has been severely hindered by the lack of relatively specific inhibitors of the underlying Ca²⁺ channels. Although several agents (econazole, SK&F-96365, 2-aminoethoxydiphenylborane) are known to inhibit store-operated channels, they also can inhibit other channels over similar concentration ranges (95). Hence, they cannot be considered diagnostic of a store-operated mechanism.

Like all Ca²⁺ influx pathways, store-operated channels are inhibited by divalent and trivalent cations, probably via a block by slow permeation. Trivalent cations like La³⁺ and Gd³⁺ are particularly effective, blocking the channels fully in the low micromolar concentration range (144). In experiments employing fluorescent dyes to study Ca²⁺ influx, Gd³⁺is often used to separate endogenous store-operated channels from recombinant TRPs, since the endogenous pathway is effectively blocked by Gd³⁺ at concentrations that fail to interfere with TRP channel activity (360).

Another pharmacological agent that has become popular for probing store-operated Ca²⁺ entry is 2-aminoethoxydiphenylborane (2-APB). 2-APB is a membrane-permeable inhibitor of InsP₃ receptor function, but it exploded onto the store-operated Ca²⁺ entry scene following the report by Ma et al. (211) that it rapidly inhibited thapsigargin-evoked Ca²⁺ influx even when applied after Ca²⁺ entry had developed. This suggested that InsP₃ receptors were required for sustaining store-operated influx and was therefore considered compelling evidence in support of the conformational-coupling mechanism for activation of store-operated entry. Although electrophysiological experiments subsequently confirmed that 2-APB inhibited I_CRACactivation (13, 291, 381), several observations cast doubt on the conclusion that the effects of the drug arose simply from inhibition of InsP₃ receptors. Instead, the most parsimonious explanation was that 2-APB inhibited the store-operated channels directly, most likely on an external site. First, 2-APB was much less effective in inhibiting I_CRAC when included in the pipette solution (13, 42), even when the pipette concentration was 20-fold higher than an external concentration which caused full block. However, for hydrophobic drugs that freely cross membranes, the rate of diffusion across the pipette tip constitutes the rate-limiting step. Hence the steady-state concentration of the drug in the cytosol may be significantly lower than the pipette concentration. Nevertheless, 2-APB still blocked I_CRAC when applied externally in acidified medium, a condition which presumably protonates 2-APB and thus reduces its membrane permeability (291). Second, following full activation of I_CRAC, external application of 2-APB rapidly inhibited the current with a time constant only slightly longer than that seen with the direct channel blocker La³⁺ (13). Third, 2-APB inhibited I_CRAC and store-operated entry in the mutant DT40 cell line in which InsP₃ receptors are not expressed (45, 210, 291). Hence, InsP₃ receptors are not required for 2-APB block of store-operated entry.

Because 2-APB seems to block CRAC channels directly and rapidly, it is becoming a popular tool to probe functional consequences of inhibiting store-operated entry. A caveat here is that the drug is now known to interfere with a variety of transport processes including SERCA pumps (237), K⁺ channels (385), MagNuM/MIC channels (130), and mitochondrial Ca²⁺ efflux (291). Furthermore, 2-APB has been found to activate the heat-gated recombinant TRPV1, TRPV2, and TRPV3 channels in HEK 293 cells (61, 146) as well as in native keratinocytes (61). The concentration range over which 2-APB activates TRPV3 is similar to that with which it affects store-operated entry (61). Hence, great care is needed in interpreting results based on the use of 2-APB.

At low concentrations (1–5 µM), 2-APB potentiates I_CRACup to fivefold in Jurkat lymphocytes (291). At higher concentrations, the drug first enhances I_CRAC but then the inhibitory effect dominates (291). Low concentrations of 2-APB actually accelerate Ca²⁺-depedent fast inactivation, whereas higher concentrations reduce it. In RBL-1 cells on the other hand, the potentiation is either much weaker (291) or absent (13, 381). In some systems then, 2-APB seems to potentiate I_CRAC even after maximal store depletion. A similar potentiating effect has been reported by the antidiarrheal agent loperamide. This drug increased store-operated influx, measured using fura 2, in a variety of cell types (124), following store emptying with thapsigargin, ionomycin, or receptor-induced elevation of InsP₃. Loperamide did not enhance Ca²⁺ signals to sphingosine. Loperamide, like 2-APB, is a promiscuous drug and hence should not be considered a specific tool for modulating store-operated influx. Nevertheless, understanding how these drugs enhance I_CRAC could provide new insight into CRAC channel gating.

Another inhibitor of I_CRACis the vanilloid capsaicin, the piquant component of red chili peppers. Capsaicin inhibited I_CRACrapidly and reversibly in Jurkats, with an IC₅₀ of 30 µM (91). The block was not voltage-dependent. Capasaicin also inhibited the voltage-dependent K⁺current (Kv1.3) in these cells with a similar concentration dependence. Because capsaicin activates vanilloid type I (VR1) receptors as well as affecting other channels, the block should not be considered specific for I_CRAC<