Research ArticleChannels

Multiple types of calcium channels arising from alternative translation initiation of the Orai1 message

See allHide authors and affiliations

Sci. Signal.  28 Jul 2015:
Vol. 8, Issue 387, pp. ra74
DOI: 10.1126/scisignal.aaa8323


In mammals exclusively, the pore-forming Ca2+ release–activated Ca2+ (CRAC) channel subunit Orai1 occurs in two forms because of alternative translation initiation. The longer, mammal-specific Orai1α contains an additional 63 amino acids upstream of the conserved start site for Orai1β, which occurs at methionine 64 in Orai1α. Orai1 participates in the generation of three distinct Ca2+ currents, including two store-operated currents: Icrac, which involves activation of Orai1 channels by the Ca2+-sensing protein STIM1 (stromal interaction molecule 1), and Isoc, which involves an interaction among Orai1, the transient receptor potential (TRP) family member TRPC1 (TRP canonical 1), and STIM1. Orai1 is also a pore-forming subunit of an arachidonic acid (or leukotriene C4)–regulated current Iarc that involves interactions among Orai1, Orai3, and STIM1. We evaluated the roles of the two Orai1 forms in the Ca2+ currents Icrac, Isoc, and Iarc. We found that Orai1α and Orai1β were largely interchangeable for Icrac and Isoc, although Orai1α exhibited stronger inhibition by Ca2+. Only the mammalian-specific Orai1α functioned in the arachidonic acid–regulated current Iarc. Thus, alternative translation initiation of the Orai1 message produces at least three types of Ca2+ channels with distinct signaling and regulatory properties.


Store-operated Ca2+ entry (SOCE), mediated by store-operated Ca2+ channels, is among the most widely encountered signaling mechanism in the animal kingdom, and the most extensively characterized store-operated channel is the Ca2+ release–activated Ca2+ (CRAC) channel. CRAC channels are composed of the pore-forming subunit Orai1, Orai2, or Orai3 and are gated by an interaction with the endoplasmic reticulum–localized Ca2+ sensor STIM1 (stromal interaction molecule 1) or STIM2 (1). These components are also necessary for a Ca2+ entry pathway involving a complex interaction between Orai1, STIM1, and channels of the canonical transient receptor potential (TRPC) family, and this complex produces a current that is less selective than Icrac, termed Isoc. There are two views regarding the molecular underpinnings of Isoc: one holding that the current represents the sum of inwardly rectifying Icrac and a less selective current through TRPC channels (2), whereas another view suggests that the current involves distinct channels composed of both TRPC and Orai pore-forming subunits (3). In addition, Orai1 is a constituent of a non–store-operated current, Iarc, involving Orai1, Orai3, and STIM1, which is gated by arachidonic acid (4) or a metabolite of arachidonic acid, leukotriene C4 (5). Recent studies indicate that currents previously shown to be mediated by arachidonic acid, Iarc (6), as well as currents mediated by leukotriene C4 (5, 7), arise from the same channel. In a recent study, their activation was inhibited when metabolism of exogenous arachidonic acid to leukotriene C4 is prevented (8), whereas in earlier studies, that was not the case (9, 10). For the sake of simplicity, we will adhere to the original term for this current, Iarc, although we believe that it may be regulated by a metabolite of arachidonic acid. Not surprisingly, Iarc is similar to Icrac in many ways, because its pore is also composed of Orai subunits. However, the arachidonate-regulated Ca2+ (ARC) channels appear to be heteromers of Orai1 and Orai3 (11), whereas the CRAC channels are thought to be homomers of Orai1, Orai2, or Orai3 (12, 13). Like Icrac, Iarc is a small current and strongly inwardly rectifying. Whereas Icrac is activated by Ca2+ store depletion detected by STIM1 (14) and is inhibited by 2-aminoethyldiphenyl borate (2-APB), Iarc is activated by a ligand, has a different pH sensitivity (15), exhibits reduced or lacks fast Ca2+-dependent inactivation (CDI) (15), does not rapidly depotentiate (5, 16), and is not inhibited by 2-APB (5, 10).

Orai1 is expressed in two forms, Orai1α (long) and Orai1β (short), because of alternative translation initiation (17). Here, we have examined the role of these distinct forms of Orai1 in the three Orai1 currents: Icrac, Isoc, and Iarc. We show that both Orai forms can support Icrac and Isoc. However, the longer form, Orai1α, undergoes much stronger CDI than does Orai1β. Additionally, only the longer Orai1α formed channels underlying Iarc. Thus, alternative forms of an ion channel encoded by a single gene can give rise to three distinctly different channels. This finding adds considerable breadth to the roles of Orai1 subunits in signaling and changes the way experimental work with Orai1 mouse models can be interpreted. It also provides a molecular framework for separating and analyzing the distinct functions of Orai1-containing channels in mammalian cell signaling.


Generation of Orai1α- and Orai1β-optimized constructs

Here, we used a construct in which the Kozak sequence for the first methionine is optimized, such that the ribosomal complex does not skip the first methionine, resulting in efficient formation of only the Orai1α form because the ribosomal complex now starts essentially every time at the first strengthened Kozak sequence (Orai1α). In a second construct, the ribosomal complex continues past a nonconsensus start site at the first methionine and starts essentially every time at the second Kozak sequence at methionine 64 (Orai1β). Thus, introduction of the Orai1α-optimized construct or the Orai1β-optimized construct results in the production of about equal amounts of either Orai1α or Orai1β (17).

Rescue of SOCE in Orai1-KO MEF cells

We investigated the efficacy of Orai1α and Orai1β in supporting SOCE with a line of mouse embryonic fibroblasts (MEFs) from an Orai1-knockout mouse (Orai1-KO MEFs) (18). To minimize overexpression of Orai1, we subcloned wild-type Orai1, Orai1α, and Orai1β into expression plasmids carrying a thymidine kinase (TK) promoter, which gives considerably less expression than in previous studies using the stronger cytomegalovirus (CMV) promoter (19). All constructs were tagged at the C terminus with enhanced green fluorescent protein (EGFP). With the strong CMV promoter, it has not generally been possible to study Orai1 expressed on its own because these constructs actually had a partial inhibitory effect, presumably due to inappropriate stoichiometry between Orai1 and endogenous STIM1 (20). We determined from the fluorescence of the EGFP-tagged proteins that the three constructs were expressed at similar amounts, although there was a small but significantly greater abundance of the wild-type and Orai1β constructs compared to that of Orai1α (Fig. 1A). We then measured SOCE from wild-type MEFs and Orai1-KO MEFs stably expressing wild-type Orai1, Orai1α, or Orai1β. SOCE in response to the passive Ca2+-depleting agent thapsigargin was undetectable in Orai1-KO MEFs but was restored by transfection with wild-type Orai1, Orai1α, or Orai1β (Fig. 1B). There were no significant differences in the thapsigargin-induced release of Ca2+ in the various cell lines (Fig. 1C). All three constructs rescued SOCE, whereas stable transfection with GFP did not. However, the Ca2+ entry in the wild-type Orai1– and Orai1β-expressing MEFs was significantly greater than in the Orai1α-expressing cells (Fig. 1D).

Fig. 1 Expression and rescue of Ca2+ entry in Orai1-KO MEFs by Orai1 constructs.

(A) Fluorescence intensity (arbitrary units) as an indicator of expression of Orai1-EGFP forms was measured and calculated as described in Materials and Methods. Data are means ± SEM [n = 421 for KO + WT (wild type), 490 for KO + Orai1α, and 838 for KO + Orai1β]. WT and KO cells are negative controls and give threshold fluorescence levels in the absence of Orai1-EGFP (n = 540 for KO, 456 for WT). Fluorescence unit is an arbitrary value in 8-bit scale. ***P < 0.001, one-way analysis of variance (ANOVA), followed by Tukey test. (B) Traces showing real-time measurements of intracellular Ca2+ in MEFs. Thapsigargin was added as indicated initially in the absence of extracellular Ca2+, and then as indicated, Ca2+ (3 mM) followed by Gd3+ (5 μM) was added. (C) Summary of the extent of thapsigargin-induced Ca2+ release, taken as the peak Ca2+ (ratio) rise after thapsigargin addition minus baseline. (D) Summary of the magnitude of thapsigargin-induced Ca2+ entry, taken as the peak Ca2+ (ratio) rise after restoration of extracellular Ca2+ minus baseline. *P < 0.05 versus WT; #P < 0.05 versus KO + WT and KO + Orai1β.

Rescue of Icrac in Orai1-KO MEF cells

We assessed Icrac in these same MEF cell lines. Because Icrac is extremely small in MEFs, we used the strategy of observing store-operated currents that were amplified in solutions lacking divalent cations [divalent-free (DVF)] (21, 22). In wild-type MEFs dialyzed with 20 mM BAPTA [1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid], a small inward current developed that was amplified by a transient switch to a DVF external solution (Fig. 2A). Shortly after breaking in with the patch pipette, a switch to the DVF solution resulted in a symmetrical (equal inward and outward) increase in current that we presume reflects an increase in leak at the cell membrane or pipette seal due to the absence of divalent cations. At the second addition, when the DVF inward current was transiently increased, this outward current was unchanged such that the leak-subtracted DVF current showed the typical inward rectification expected of Icrac (Fig. 2A). This current was completely blocked by subsequent addition of 5 μM Gd3+, again as expected for Icrac (20). In Orai1-KO MEFs, no such current was observed (Fig. 2B). Stable expression of any of the three constructs—wild-type Orai1 (Fig. 2C), Orai1α (Fig. 2D), or Orai1β (Fig. 2E)—in the Orai1-KO MEF cells rescued the inwardly rectifying inward current and produced similar leak-subtracted peak inward Na+ currents to those in wild-type MEFs (Fig. 2F).

Fig. 2 Na+-carried Icrac in WT and Orai1-KO MEFs.

A pipette solution containing 20 mM BAPTA was dialyzed through the patch pipette to activate CRAC currents. Representative time course of CRAC current recordings is shown. (A) CRAC currents amplified by DVF bath solution were measured in WT MEF cells. (B) Absence of CRAC currents in Orai1-KO MEF cells. (C to E) CRAC currents in Orai1-KO MEF cells stably expressing the indicated Orai1 constructs. Below each time course trace are CRAC sodium current I-V relationships, which are taken from traces in (A) to (E) where indicated by asterisks. (F) Summary of Na+-carried CRAC currents under the five different conditions. *P < 0.05 versus WT, ANOVA followed by Tukey test. n values represent the number of independently determined currents.

Wild-type Orai1, expected to yield a mixture of the two forms, as well as the Orai1α- and Orai1β-specific constructs, all restored Icrac to levels similar to the wild-type cells. This was somewhat unexpected because, in the Ca2+ experiments, Orai1α resulted in a significantly lower Ca2+ entry signal. Because in the Icrac measurements Ca2+ feedback on the channels is abrogated by the use of the fast buffer BAPTA, we considered that the smaller Ca2+ entry in the Orai1α cells, as measured in the fluorescence assays (Fig. 1), might reflect a greater degree of Ca2+ feedback, also known as CDI (23). We thus carried out experiments specifically designed to reveal the presence and extent of CDI by comparing the kinetics of inactivation of currents with the fast calcium chelator BAPTA or the slower chelator EGTA in the pipette. CDI is revealed by holding membrane potential at a positive value, where CDI is prevented, then rapidly jumping to a negative value, allowing Ca2+ to enter and then rapidly inactivate Icrac. For Orai1α, we observed rapid and considerable inactivation with the slower Ca2+ buffer EGTA, and this inactivation was diminished with the faster buffer BAPTA (Fig. 3A), whereas for Orai1β, this difference was absent (Fig. 3B). Quantification of the extent of inactivation data for different final membrane potentials showed that Orai1α and Orai1β behaved similarly in the presence of BAPTA (Fig. 3C) and that Orai1α produced a smaller current at all potentials tested in the presence of EGTA (Fig. 3D). Thus, we concluded that although both Orai1α and Orai1β can support Icrac, the two forms differ in that Orai1α exhibits stronger CDI.

Fig. 3 CDI of Icrac in HEK293 cells expressing STIM1 and Orai1α or Orai1β.

(A and B) CRAC currents were recorded in human embryonic kidney (HEK) 293 cells cotransfected with 1.0 μg of plasmid coding for EYFP (enhanced yellow fluorescent protein)–STIM1 and plasmids coding for 2.0 μg of either Orai1α (A) or Orai1β (B). As indicated, either 20 mM BAPTA or 10 mM EGTA was used in patch pipette. (C and D) For quantification of CDI, extent of inactivation at 146 ms (R146 ms), the residual current remaining at the end of the test pulse at 146 ms relative to the peak current at 3 ms, was plotted against test potential [as in (23)]. n values represent the number of independent current measurements.

Function of Orai1 forms in Isoc

We considered the possibility that one of the forms of Orai1 might specifically function in the less selective store-operated current Isoc (24). These currents are believed to result from a complex mechanism whereby Ca2+ entering through classical Orai-containing CRAC channels recruits TRPC channels to the plasma membrane where they are activated by STIM1 (2).

We transfected HEK293 cells with either STIM1 and wild-type Orai1, which produces large Icrac, or STIM1, wild-type Orai1, and TRPC1, which produces a larger current with a more significant outward component, resembling Isoc (Fig. 4A). We performed similar transfections using either the Orai1α-optimized construct (Fig. 4B) or the Orai1β-optimized construct (Fig. 4C). The Icrac that we observed was somewhat smaller than that in earlier publications (20), possibly due to variations in lines of HEK293 cells. We routinely observed small outward currents associated with the Icrac protocol in the cells not transfected with TRPC1, which could result from endogenous TRPC1 in HEK293 cells (25); this was not further investigated. Both Orai1α and Orai1β supported Isoc equally well, but unlike the findings in MEFs, the single forms of Orai gave greater Icrac currents than did the wild-type Orai1 (Fig. 4D). Inward Isoc was also greater with the specific constructs than with wild-type Orai1, perhaps in part because the amount of TRPC1 recruited should depend on the amount of Ca2+ entering through CRAC channels (2). However, for the outward Isoc, this difference was substantially less and was statistically significant only for Orai1β. This might support the model in which Isoc is composed of separate CRAC and TRPC1 channels (2, 26). In confirmation of a previous report (3) using the same TRPC1 construct and recording conditions, we failed to see currents in cells transfected with only STIM1 and TRPC1 (Fig. 4, A and D).

Fig. 4 Orai1α and Orai1β in Icrac and Isoc in HEK293 cells.

HEK293 cells were transiently transfected with the indicated combinations of constructs. Successfully transfected cells were identified by the fluorescence from EYFP-STIM1. (A to C) Whole-cell recordings of CRAC (solid line from cells expressing STIM1 + Orai1 constructs) or SOC (dashed line from cells expressing STIM1 + Orai1 constructs + TRPC1) currents by dialysis of BAPTA (10 mM) and IP3 (inositol 1,4,5-trisphosphate) (25 μM) through the patch pipette. Note in (A) that the current measured in the STIM1 + TRPC1 cells is barely detectable (dotted line). (D) Summary data for current density at +100 and –100 mV in cells transfected with STIM1 and the indicated Orai1 constructs with (Isoc) or without (Icrac) TRPC1. *P < 0.05, two-way ANOVA followed by Tukey test. n values are the number of independently determined currents.

We noted that the larger Icrac seen in HEK293 cells with the Orai1α and Orai1β constructs, compared to cells with wild-type Orai1 (Fig. 4D), was not seen in Icrac measurements in MEFs (Fig. 2F). One difference in the two protocols (aside from the different cell types) is that in the HEK293 experiments, exogenous STIM1 was supplied, whereas it was not in the MEFs, raising the possibility that STIM1 was limiting in the MEFs. Indeed, when we transiently transfected the stable MEF lines with STIM1, compared with the Icrac in the previous experiment (Fig. 2F), Icrac was significantly amplified in all three lines on the basis of two-way ANOVA (Fig. 5B) but to a significantly greater extent in the cell lines expressing either Orai1α or Orai1β alone rather than in the cells expressing wild-type Orai1 cells (Fig. 5B). The basis for this small but apparently reproducible difference is currently under investigation.

Fig. 5 Effect of expression of STIM1 on Icrac in MEFs transfected with Orai1α or Orai1β constructs.

(A) CRAC currents in Orai1-KO MEF cells stably expressing the indicated Orai1 constructs and transiently expressing EYFP-STIM1. Below each time course trace are CRAC sodium current I-V relationships, which are taken from traces where indicated by asterisks. (B) A summary of the Na+-carried inward currents, including currents from Fig. 2 (Orai1-KO cells transfected to express the indicated Orai1 protein without overexpressing STIM1). Two-way ANOVA indicated highly significant increase with all three constructs in the presence of overexpressed STIM1 (P < 0.001). Post hoc Tukey test indicated that in the presence of STIM1, both Orai1α and Orai1β have significantly larger currents than Orai1-WT (P < 0.05). n represents the number of independent current measurements.

The data to this point indicated that the two forms support Icrac equally well. In our previous study, we found that in unstimulated cells, the mobility of Orai1α was significantly less than that for Orai1β (17). Thus, we examined the rates at which Orai1α and Orai1β associate with STIM1 in cells after Ca2+ store depletion [see, for example, (17)]. By monitoring the area of puncta containing fluorescently tagged STIM1 and fluorescently tagged Orai1α or Orai1β, we determined that the rate of association of the two forms with STIM1 was similar for the two Orai1 forms (Fig. 6).

Fig. 6 Orai1 puncta formation in HEK293 cells in response to store depletion by activation of muscarinic acetylcholine receptors.

Orai1-positive puncta over time were measured in HEK293 cells expressing either Orai1α-EGFP and mCherry-STIM1 or Orai1β-EGFP and mCherry-STIM1 and that were exposed to 0.5 mM carbachol. Puncta size was calculated by averaging the area of each individual puncta in the single cells at each time point of 30 s. Puncta size was defined as the area of colocalization of Orai1 and STIM1. Maximal puncta size was set at 100%. Six cells were analyzed in six independent experiments, means + SEM.

Orai1 forms and Iarc

Finally, we investigated the possible role of a specific form of Orai1 in the non-SOCE pathway activated by arachidonic acid or its metabolite leukotriene C4 (57, 27). We used HEK293 cells for these studies because they exhibit a small endogenous Iarc (8, 16) and because the function of Orai1 in Icrac can be compared in the same cells. We used exogenous arachidonic acid to activate Iarc currents. Because of the small size of the currents, we examined Icrac- and Iarc-associated Na+ currents. In wild-type HEK293 cells, dialysis with an intracellular solution containing 20 mM BAPTA resulted in the development of an inwardly rectifying Na+ current that rapidly diminished by a process known as depotentiation (28) and was inhibited by 2-APB (Fig. 7A), which are characteristics of Icrac. This current was absent or substantially diminished after small interfering RNA (siRNA)–mediated knockdown of Orai1 (Fig. 7B). In cells in which Orai1 had been knocked down and subsequently transfected with plasmids for either the Orai1α (Fig. 7C) or Orai1β (Fig. 7D) constructs, both forms of Orai1 were equally efficacious in rescuing Icrac Na+ current (Fig. 7E). Note that both the Na+ and Ca2+ currents for Icrac and Iarc were substantially smaller in the current study than in earlier published studies [for example, (16, 22, 29)]. The reason for this difference is not known but, as pointed out previously, may be due to differences in lines of HEK293 cells.

Fig. 7 Rescue of Icrac by Orai1α or Orai1β after knockdown of Orai1 in HEK293 cells.

CRAC currents amplified in DVF bath solution were measured in HEK293 cells. In each case, Na+ current was obtained by brief substitution with the DVF solution. The first substitution of DVF occurs before substantial store depletion and provides a measure of leak current, which was subtracted by the current that developed later. (A) CRAC currents in HEK293 cells transfected with nontargeting control siRNA. (B) Effect of Orai1 knockdown with siRNA on CRAC currents. (C and D) Rescue of CRAC currents by Orai1α or Orai1β expression in Orai1-silenced cells. For each trace, sodium CRAC current I-V relationships are shown below, which are taken from traces in (A) to (D) where indicated by asterisks. (E) Summary of leak-subtracted Na+-carried CRAC currents. *P < 0.05 versus siControl, ANOVA followed by Tukey test. n values represent the number of independently determined current measurements.

With an intracellular patch pipette solution containing Ca2+ buffered to 150 nM to prevent development of Icrac, HEK293 cells developed inwardly rectifying Na+ currents in response to extracellular addition of 8 μM arachidonic acid (Fig. 8A). Unlike Icrac (Fig. 7), this Na+ current did not depotentiate and was not blocked by 2-APB, consistent with the properties of Iarc (5, 7, 8, 16). Knockdown of Orai1 substantially reduced the Na+ current (Fig. 8B). Subsequent transfection with a plasmid carrying a construct for Orai1α fully rescued Na+ Iarc (Fig. 8C), whereas Orai1β was completely ineffective (Fig. 8D; summarized in Fig. 8E). To ensure that this striking distinction in the function of the two Orai1 forms also applied to the channels when the physiological ion was carrying the current, we examined Ca2+ currents as well. These currents are extremely small, and so we show the results as a scatter plot of individual leak-subtracted currents (Fig. 8F). Despite the very small current values, only Orai1α rescued the Ca2+ Iarc.

Fig. 8 Rescue of Iarc by Orai1α, but not Orai1β, after knockdown of Orai1 in HEK293 cells.

Arachidonic acid (AA) (8 μM) was added to bath solution to activate ARC currents (15) in HEK293 cells. In each case, Na+ current was obtained by brief substitution with the DVF solution. The first substitution of DVF occurs before addition of arachidonic acid and provides a measure of leak current, which was subtracted by the current developing later. (A) ARC currents in HEK293 cells transfected with nontargeting control siRNA. (B) Effect of Orai1 knockdown on ARC currents. (C and D) Rescue of ARC currents by Orai1α or Orai1β expression in Orai1-silenced cells. Below each trace are Na+-carried ARC current I-V relationships taken from traces in (A) to (D) where indicated by asterisks. (E) Summary of Na+-carried ARC current under the four experimental conditions in (A) to (D). *P < 0.05 versus siControl; P = not significant compared to siOrai, ANOVA followed by Tukey test. n values represent the number of independently determined current measurements. (F) Summary of Ca2+-carried ARC currents in HEK293 cells under the four experimental conditions.


In a previous study reporting Icrac measurements (17), a construct that produced only Orai1α was made by mutating the start site at methionine 64, and a construct that produced only Orai1β was made by mutating the first methionine start site. Alternative translation initiation usually occurs through a process of ribosomal scanning. The two forms of Orai1 arise because the Kozak sequence at the first methionine is weak, thus will be skipped about half of the time. Therefore, with the Orai1α construct used in the previous study (17), the ribosomal complex would still fail to initiate much of the time at the first methionine, resulting in poorer expression than for the Orai1β construct. This likely explains why the Orai1α construct gave considerably less Icrac than did the Orai1β construct [Fig. 4 in (17)].

In this same study (17), we provided evidence that Orai1α and Orai1β could not form heteromeric channels. Nonetheless, on the basis of the current findings, we concluded that channels composed of either form are equally capable of forming CRAC channels or participating in the generation of Isoc. In addition, despite the previously documented difference in plasma membrane mobility under unstimulated conditions (17), we found that after Ca2+ store depletion, the two forms associated with STIM1 in near plasma membrane puncta at similar rates. We did observe a difference in Ca2+ regulation of the two forms, because only Orai1α exhibited CDI. However, more strikingly, it appeared that Orai1α is specifically required for the channels underlying Iarc. Only mammals express the N-terminal–extended Orai1α, and only mammals express Orai3, which associates with Orai1 to form ARC channels (5, 11). Thus, mammalian evolution has exploited the process of alternative translation initiation to produce from the same gene calcium channels with distinct modes of regulation and activation. Our previous work demonstrated that the two forms of Orai1 do not assemble to form channels that are heteromultimeric with respect to Orai1 form (17). Thus, cells may use Orai1 in three functionally distinct channel structures: homomeric Orai1α and homomeric Orai1β CRAC channels, with differing sensitivities to CDI, as well as heteromeric Orai1α-Orai3 ARC channels. We expect the presence of at least some Orai1α CRAC channels because the N-terminal extension contains the site at which the Ca2+-activated adenylyl cyclase 8 is tethered, permitting specific linkage between Ca2+ activation of cAMP (adenosine 3′,5′-monophosphate) formation and store-operated CRAC channels (30).

A number of mouse models for studying the function of Orai1 are now in use, and the general presumption is that a deficit in Orai1 means a loss of Icrac. However, it is clear that Orai1 functions in other Ca2+ channels and that the loss of Isoc or Iarc or both may underlie or contribute to the various phenotypes that these KO mice exhibit. Understanding distinctions at the molecular level between these different Ca2+ signaling mechanisms will permit the generation of other mouse models to shed light on the specific roles of Orai1 in different physiological and pathological contexts.



Arachidonic acid, Cs-methanesulfonate, and Na-methanesulfonate were purchased from Sigma. Fura-5F, AM was purchased from Setareh Biotech, LLC. 2-APB was purchased from Calbiochem. Cs-BAPTA was purchased from Invitrogen. GdCl3 was from Acros Organics. All Dicer substrate siRNAs (dsiRNAs) were purchased from Integrated DNA Technologies. The transfection kit (VCA-1003) for HEK293 cells was from Lonza. All other chemicals were from Fisher.

Cell culture

MEFs were cultured in Dulbecco’s minimum essential medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 U/ml) and maintained in a humidified 95% air and 5% CO2 incubator at 37°C. In preparation for Ca2+ measurements, cells were cultured to about 70% confluence and then transferred onto 30-mm round glass coverslips (#1.5 thickness). The cells were allowed to attach for a period of 6 hours, after which additional DMEM was added to the coverslip, and the cells were maintained in culture for an additional 24 hours before use in Ca2+ measurements.

Complementary DNA constructs

Human Orai1 complementary DNA (cDNA) fused to EGFP under the control of a TK promoter was provided by T. Balla [National Institute of Child Health and Human Development, National Institutes of Health (NIH)]. This wild-type Orai1–EGFP cDNA with the TK promoter was subcloned into CD510B-1 Lentivector (System Biosciences) for transfections and making lentivirus. The CMV promoter in CD510B-1 Lentivector was deleted during subcloning. Site-directed mutagenesis of the first methionine to alanine or to a strong Kozak sequence [as in (17)] was performed with QuikChange II XL (Stratagene) according to the manufacturer’s instructions. These mutations produce constructs that give essentially 100% Orai1β and 100% Orai1α, respectively (17). For transient transfections, STIM1, EYFP-STIM1, mCherry-STIM1, wild-type Orai1–EGFP, Orai1α-EGFP, and Orai1β-EGFP constructs were as described previously (17). The EYFP-STIM1 plasmid was obtained from T. Meyer (Stanford University). TRPC1 construct was obtained from L. Tsiokas (University of Oklahoma) (3).

Production of lentiviruses and stable Orai1-expressing MEF cells

All lentiviruses were packaged in HEK293T/17 cells [American Type Culture Collection (ATCC) #CRL-11268] (31). Briefly, 293 T cells were transiently transfected with pMD2G, psPAX2, and transfer vector containing wild-type Orai1–EGFP, Orai1α-EGFP, or Orai1β-EGFP using Lipofectamine 2000. Supernatant was collected 48 hours after transfection and concentrated by centrifugation at 50,000g for 2 hours. Pellets were resuspended in phosphate-buffered saline (pH 7.4) and used for infection. All titers were determined by performing quantitative polymerase chain reaction (PCR) to measure the number of lentiviral particles that integrated into the host genome. In addition to quantitative PCR, biological titration of viruses that coexpressed fluorescent moieties was determined by flow cytometry. Multiplicity of infection ranging from 3 to 1 was used for infection of MEF cells. The stable Orai1-expressing MEF cells were established by selecting with puromycin (4 μg/ml) in the medium.

Transient transfections

For experiments examining Orai1α and Orai1β in Icrac and Isoc, the HEK293 cells were transfected with the Amaxa electroporation system (Amaxa Inc., a Lonza Cologne company) following the guidelines set forth by the company for the cell line using the HEK293 (ATCC) cell line setting and the Nucleofector kit V buffer. HEK293 cells were transfected with the following amounts of cDNA: EYFP-STIM1 cDNA (0.1 μg) and wild-type Orai1–EGFP cDNA (0.1 μg) with or without TRPC1 cDNA (0.1 μg). To compare the α and β forms of Orai1, HEK293 cells were transfected with 0.1 μg of cDNA encoding either of these forms with or without 0.1 μg of TRPC1 cDNA. After a 6- to 8-hour incubation period, the medium bathing the cells was replaced with complete DMEM, and the cells were maintained in culture. The following day, cDNA-transfected cells were transferred to 30-mm glass coverslips in preparation for electrophysiological studies. For experiments examining the roles of Orai1α and Orai1β in Icrac and Iarc, and for examining CDI of Orai1α- and Orai1β-mediated Icrac, transfections were done using the Nucleofector Device II (Amaxa Biosystems) with program Q-001 according to the manufacturer’s instructions. For the CDI studies, cells were also transfected with EYFP-STIM1 as in (17). After transfection, HEK293 cells were seeded on glass coverslips and maintained in DMEM supplemented with 10% fetal bovine serum (HyClone Laboratories), 1% l-glutamine, and 1% antibiotic-antimycotic (Invitrogen) in a humidified, 5% CO2 incubator at 37°C.

For knockdown experiments, we used dsiRNA at 1 nmol dsiRNA per 1 million cells. Sequences are listed as follows: siOrai1: 5′-GGGAGGUUGAGACGGACAGGACAGGAG-3′ and 3′-CCCUCCAACUCUGCCUGUCCUGUCC-5′; negative control: 5′-CGUUAAUCGCGUAUAAUACGCGUAT-3′ and 3′-AUACGCGUAUUAUACGCGAUUAACGAC-5′.

Calcium measurements on single MEF cells

Fluorescence measurements were made with MEFs loaded with the ratiometric calcium-sensitive dye Fura-5F. Briefly, coverslips with attached MEFs were mounted in a Teflon chamber and incubated in DMEM with 4 μM Fura-5F, AM at 37°C in the dark for 25 min. Before [Ca2+]i (intracellular calcium concentration) measurements were made, cells were washed three times and incubated for 10 to 15 min at room temperature (25°C) in a Hepes-buffered salt solution (HBSS; 120 mM NaCl, 5.4 mM KCl, 0.8 mM MgCl2, 20 mM Hepes, 1.8 mM CaCl2, and 10 mM glucose, with pH 7.4 adjusted by NaOH). In these experiments, nominally Ca2+-free solutions were HBSS with no added CaCl2.

Coverslips with attached MEFs were mounted onto a Nikon TS100 inverted microscope equipped with a 20× Fluor objective (0.75 numerical aperture), and fluorescence images were recorded and analyzed with a digital fluorescence imaging system (InCyt Im2, Intracellular Imaging Inc.) equipped with a light-sensitive charge-coupled device camera (Cooke PixelFly, Applied Scientific Instrumentation). Fura-5F fluorescence was monitored by alternately exciting the dye at 340 and 380 nm and collecting the emission at 510 nm. Changes in intracellular calcium are represented as the ratio of Fura-5F fluorescence due to excitation at 340 nm to that due to excitation at 380 nm (ratio, F340/F380). We used the lower affinity indicator Fura-5F [Kd (dissociation constant) = 400 nM] to ensure that the data are well below dye saturation; most of the ratios were <1.0, well below the average observed Rmax of 5.6. The ratio changes in fields of Fura-5F–loaded cells were collected from multiple regions of interest (ROIs), with each ROI representing an individual cell. Typically, 40 to 50 ROIs were monitored per experiment. In all cases, ratio values were corrected for contributions of autofluorescence, which was measured after treating cells with 10 μM ionomycin and 20 mM MnCl2.

Assessment of expression of EGFP-tagged Orai1 forms in MEF cells

Orai1-EGFP expression in MEF cells was quantitated by recording cell fluorescence on an LSM 780 confocal microscope (Carl Zeiss) using a Plan-Apochromat 40×/1.4 Oil DIC M27 objective at a zoom of 1×. EGFP was excited with the 488-nm laser line. Fluorescence emission was detected using spectral imaging between 491 and 695 nm with a 9-nm bandwidth. Digital photon-counting mode was used for detection, with pinhole set at 180 μm and 4× line averaging. Linear unmixing was used to deconvolute the fluorescence spectral overlap of Orai1-EGFP and autofluorescence. After elimination of autofluorescence, total GFP signal intensity was calculated in individual cells using the ImageJ “Particle Analysis” method.

Confocal imaging of puncta formation of Orai1 forms

Fluorescence time-lapse images were taken with a Zeiss (LSM 710) confocal laser scanning microscope using the 488-nm line of an argon laser for excitation and a 493- to 558-nm emission for EGFP and 561-nm line of a HeNe laser for excitation and 578 to 700 nm for emission for mCherry. The specimens were viewed at a zoom of 2.4× using Plan oil objectives (Plan-Apochromat 63×/1.4 Oil DIC). Images were taken every 30 s for 20 min. Focal plane was maintained by using Perfect Focus System sampling before each time point. The puncta area of colocalization of Orai1 and STIM1 was calculated using Zen 2012 (Zeiss Inc.) and ImageJ (NIH) software. All data were analyzed using a laboratory-written macro (available upon request) of SigmaPlot (Systat Software).

Patch clamp electrophysiology

Patch clamp electrophysiological recordings were carried out using an Axopatch 200B and Digidata 1440A (Molecular Devices) as previously published (7, 8). Pipettes were pulled from borosilicate glass capillaries (World Precision Instruments) with a P-1000 Flaming/Brown micropipette puller (Sutter Instrument Company) and polished using DMF1000 (World Precision Instruments). Resistances of filled pipettes were 2 to 4 megohms. The liquid-junction potential offset due to different internal and external saline composition was around −4.5 mV and was corrected. Series resistances were between 5 and 10 megohms. Under whole-cell configuration, only cells with tight seals (>13 gigohms) were selected to perform recordings. Cells were maintained at a +30-mV holding potential during experiments. Clampfit 10.1 software was used for data analysis.

Store depletion–activated CRAC currents (MEFs)

Bath solution: 115 mM Na-methanesulfonate, 10 mM CsCl, 1.2 mM MgSO4, 10 mM Hepes, 20 mM CaCl2, and 10 mM glucose (pH 7.4 with NaOH).

Pipette solution: 115 mM Cs-methanesulfonate, 20 mM Cs-BAPTA, 8 mM MgCl2, and 10 mM Hepes (pH 7.2 with CsOH).

DVF solution: 155 mM Na-methanesulfonate, 10 mM HEDTA (hydroxyethyl ethylenediamine triacetic acid), 1 mM EDTA, and 10 mM Hepes (pH 7.4 with NaOH).

Protocol: Reverse voltage ramps from +100 to −140 mV lasting 250 ms every 2 s.

Icrac and Isoc in HEK293 cells

The solutions and recording protocol were those described by (3).

Bath solution: 140 mM NaCl, 1.2 mM MgCl2, 10 mM CaCl2, 5 mM CsCl, 30 mM d-glucose, and 10 mM Hepes (pH 7.4).

Pipette solution: 145 mM cesium methanesulfonate, 20 mM BAPTA, 10 mM Hepes, 8 mM MgCl2, and 25 μM IP3, pH to 7.2 with CsOH.

Protocol: Reverse voltage ramps from −100 to +100 mV of 250 ms were recorded every 2 s.

Iarc currents in HEK293 cells

Bath solution: 115 mM Na-methanesulfonate, 10 mM CsCl, 1.2 mM MgSO4, 10 mM Hepes, 20 mM CaCl2, and 10 mM glucose (pH 7.4 with NaOH). Arachidonic acid (8 μM) was added to the bath where indicated in the figures.

Pipette solution: 115 mM Cs-methanesulfonate, 10 mM Cs-BAPTA, 5 mM CaCl2, 8 mM MgCl2, and 10 mM Hepes (pH 7.2 with CsOH). Calculated free Ca2+ was 150 nM using Maxchelator software (

DVF solution: 155 mM Na-methanesulfonate, 10 mM HEDTA, 1 mM EDTA, and 10 mM Hepes (pH 7.4 with NaOH).

Protocol: Reverse voltage ramps from +100 to −140 mV lasting 250 ms every 2 s.

CDI experiments

Bath solution: 115 mM Na-methanesulfonate, 10 mM CsCl, 1.2 mM MgSO4, 10 mM Hepes, 20 mM CaCl2, and 10 mM glucose (pH 7.4 with NaOH).

Pipette solution 1: 115 mM Cs-methanesulfonate, 20 mM Cs-BAPTA, 8 mM MgCl2, and 10 mM Hepes (pH 7.2 with CsOH).

Pipette solution 2: 135 mM Cs-methanesulfonate, 10 mM EGTA, 8 mM MgCl2, and 10 mM Hepes (pH 7.2 with CsOH).

Recording method and protocols of voltage stimulation: After break-in, we ran a stimulus protocol consisting of families of 150-ms voltage steps (from +30-mV holding potential to −120, −100, −80, and −60 mV). There is a 5-s interval between steps. The first recording was used for leak subtraction. Next, a 250-ms voltage ramp from +100 to −140 mV was administered every 2 s until CRAC currents reached a steady-state level (usually 150 to 300 s). Once steady state for CRAC current activation was achieved, we performed a second voltage step protocol similar to the first one.


All values are expressed as means ± SEM. Statistical analyses comparing two or more experimental groups were performed using ANOVA, followed by Tukey test for differences between rows, columns, or cells as appropriate, or two-tailed t test, with Origin 8.1 software (OriginLab) or GraphPad Prism 6.0. Throughout the figures, *, **, and *** indicate P < 0.05, 0.01, and 0.001, respectively. Differences were considered significant when P < 0.05.


Acknowledgments: We thank L. Birnbaumer for the Orai1-null MEFs and T. Balla for the thymidine kinase promoter constructs. An EYFP-STIM1 plasmid was obtained from T. Meyer (Stanford University). TRPC1 construct was obtained from L. Tsiokas (University of Oklahoma). We thank C. Romeo and N. Martin (Viral Vector Core, National Institute of Environmental Health Sciences) for technical assistance in lentivirus preparations. S. Dudek and C. Erxleben read the manuscript and provided helpful comments. G. Kissling advised us on statistical evaluation of our data. Funding: Work described in this publication was supported in part by the Intramural Research Program, NIH, and by grants R01HL097111 and R01HL123364 from the NIH and American Heart Association grant 14GRNT18880008 to M.T. Author contributions: P.N.D., X.Z., A.J., S.B., and S.W. carried out experiments, the results of which are included in this article. J.W.P. and M.T. designed the study and wrote sections of the article. Competing interests: The authors declare that they have no competing interests. Data and materials availability: The laboratory-written macro for SigmaPlot and constructs generated as part of this study are available from the authors.
View Abstract

Navigate This Article