Research ArticleCalcium signaling

Activation of STIM1-Orai1 Involves an Intramolecular Switching Mechanism

See allHide authors and affiliations

Science Signaling  16 Nov 2010:
Vol. 3, Issue 148, pp. ra82
DOI: 10.1126/scisignal.2001122

Abstract

Stromal interaction molecule 1 (STIM1) stimulates calcium ion (Ca2+) entry through plasma membrane Orai1 channels in response to decreased Ca2+ concentrations in the endoplasmic reticulum lumen. We identified an acidic motif within the STIM1 coiled-coil region that keeps its Ca2+ activation domain [Ca2+ release–activated Ca2+ (CRAC) activation domain/STIM1-Orai activating region (CAD/SOAR)]—a cytoplasmic region required for its activation of Orai1—inactive. The sequence of the STIM1 acidic motif shows substantial similarity to that of the carboxyl-terminal coiled-coil segment of Orai1, which is the postulated site of interaction with STIM1. Mutations within this acidic region rendered STIM1 constitutively active, whereas mutations within a short basic segment of CAD/SOAR prevented Orai1 activation. We propose that the CAD/SOAR domain is released from an intramolecular clamp during STIM1 activation, allowing the basic segment to activate Orai1 channels. This evolutionarily conserved mechanism of STIM1 activation resembles the regulation of protein kinases by intramolecular silencing through pseudosubstrate binding.

Introduction

Store-operated Ca2+ entry (SOCE) was postulated more than 20 years ago as a Ca2+ entry pathway triggered by a decrease in the endoplasmic reticulum (ER) luminal Ca2+ concentration (1). Although patch-clamp analyses had long identified a current, named ICRAC [Ca2+ release–activated Ca2+ (CRAC) current], corresponding to SOCE in T cells and mast cells (2, 3), the molecular mechanisms underlying ICRAC or SOCE had until recently been elusive. Small interfering RNA screens identified the stromal interaction molecules (STIM1 and STIM2), acting as ER Ca2+ sensors (4, 5), and the Orai [also known as CRAC modulator (CRACM)] proteins, acting as plasma membrane (PM) Ca2+ channels, as working together to give rise to ICRAC (68). Although best known for their role in T cell activation, where defects in STIM1 or Orai1 cause severe immunodeficiency (6, 9), STIM1 and Orai1 are also important for platelet aggregation (10) and for skeletal muscle development (11).

Since their identification, rapid progress has been made in analyzing the molecular details of the STIM1 and Orai1 activation process [reviewed in (12)]. The N-terminal luminal segment of the ER-localized, membrane-spanning STIM1 protein contains EF-hand and SAM (sterile α motif) domains, which are responsible for luminal Ca2+ sensing and oligomerization, respectively (13). The C-terminal segment of STIM1 (STIM-ct), which faces the cytosol, contains two putative coiled-coil domains (CC1 and CC2), as well as several acidic, serine-proline–rich, and basic segments [reviewed in (14)] (Fig. 1A). The STIM-ct is capable of activating Orai1 channels (1517), and oligomerization alone is sufficient to activate STIM1 (18). Several studies have identified small (~100 amino acids) partly overlapping segments within the cytosolic aspect of STIM1 as the minimal Orai1 activating domain, which have been variably called CAD (CRAC activation domain), SOAR (STIM1-Orai activating region), or OASF (Orai1-activating small fragment) (1921) (identified here as the CAD/SOAR domain). In addition, the C-terminal putative coiled-coil domain of Orai1 has been shown to be critical for Orai1 activation by STIM1 (2224). However, it is not clear how oligomerization and the subsequent clustering of STIM1 facilitate the interaction of the STIM1 CAD/SOAR domain with Orai1 channels.

Fig. 1

(A) Schematics of STIM1 structure and the constructs used in the present study. The numbering corresponds to the human STIM1 protein. EF, Ca2+-binding EF-hand motif; SAM, sterile α motif; TM, transmembrane segment; CC1 and CC2, coiled-coil domains; CAD/SOAR, minimal Orai1 activation domain; D, acidic region; S/P, proline-, serine-threonine–rich segment; K, polybasic domain [after (43)]. The blue and yellow asterisks indicate the positions of the basic and acidic regions, respectively, identified in the present study. (B) Schematics of clustering by rapamycin-induced heterodimerization of the FKBP12-fused cytosolic STIM1 (FK-STIM1-ct) and the FRB construct targeted to the ER surface. The blue oval represents the C-terminal polybasic domain of STIM1, which is important for its PM localization.

Here, we used a strategy based on the heterodimerization of FKBP12 and the FKBP12-binding region (FRB) of mammalian target of rapamycin (mTOR) (25) to oligomerize the isolated soluble STIM-ct and show that, although the full cytoplasmic segment of STIM1 activates Orai1 only poorly, it becomes highly effective upon STIM1 clustering. In contrast, the CAD/SOAR domain effectively activates Orai1 channels even without oligomerization, which suggests that it is kept inactive when in the context of the whole STIM1 cytosolic segment. In searching for an intramolecular silencing mechanism, we identified a short acidic segment within the first coiled-coil domain of STIM1 that formed an intramolecular interaction possibly with a basic sequence within CAD/SOAR that has recently been identified as key for Orai1 activation (26). We found that mutations within the acidic stretch render STIM1 constitutively active and confirmed that the basic sequence within CAD/SOAR is essential for STIM1-mediated Orai1 activation. Intriguingly, the region around the acidic stretch within STIM1 shows substantial sequence similarity with the C terminus of Orai1 and could act as a decoy to interfere with activation of Orai1 channels by STIM1, when presented at the PM as part of the juxtamembrane first coiled-coil domain of STIM1. We propose that oligomerization of STIM1 unmasks the CAD/SOAR domain by breaking the intramolecular interaction that keeps it inactive in the quiescent STIM1 molecule.

Results

STIM1 clustering is required for activation of the cytosolic segment

The STIM1-ct has been shown to activate Orai1 channels (15), a finding confirmed in subsequent studies [reviewed in (14)]. However, as previously shown (1921), when expressed (together with Orai1) in COS-7 cells in the form of a monomeric red fluorescent protein (mRFP) fusion protein, STIM1-ct elicited only a moderate increase in basal cytosolic Ca2+. Very few cells showed substantially increased basal Ca2+ [above a Fura-2 fluorescence ratio at 340 and 380 nm (F340/F380) of 0.3] despite prominent PM localization of STIM1-ct (see below), presumably because the expressed protein was mostly in an inactive conformation. We devised a strategy to cluster STIM1-ct, in which we placed an FKBP12 module between the mRFP and the STIM1-ct (FK-STIM1-ct) and expressed this protein together with a recruiter construct consisting of an ER-targeted FRB-CFP (fragment of mTOR that binds FKBP12 fused to cyan fluorescent protein) (27) (Fig. 1B). We anticipated that rapamycin would connect PM-associated FK-STIM1-ct to the ER-bound FRB, thereby enriching FK-STIM1-ct and FRB-CFP at the ER-PM contact zones (Fig. 1B). Rapamycin indeed induced the formation of ER-PM contact zones enriched in both binding partners and also recruited Orai1 proteins to these zones (Fig. 2A, arrows). This manipulation led to increased Förster resonance energy transfer (FRET) between STIM1-ct molecules (Fig. 2B) and elicited a substantial increase in cytoplasmic Ca2+, indicating that the oligomerized STIM1-ct construct activated Orai1 channels (Fig. 2C).

Fig. 2

Clustering of the STIM1 cytosolic domain activates Orai1 channels. (A) Localization of the indicated proteins expressed in COS-7 cells before (top row) and 5 min after (bottom row) rapamycin addition. Confocal images were taken in live cells 1 day after transfection. The areas in the white boxes are shown enlarged in the images labeled “before” and “after.” Note the substantial membrane localization of the cytoplasmic STIM1 segment and its co-clustering with the Orai1 at the ER-PM contact zones (arrows) after addition of rapamycin (100 nM). (B) Rapamycin-induced clustering increases FRET between YFP- and mRFP-tagged recruitable STIM1-ct, indicating that clustering brings them within FRET distance. (C) Cytosolic Ca2+ increases evoked by rapamycin-induced clustering of FK-STIM1-ct. COS-7 cells were transfected with the ER-targeted CFP-FRB, the mRFP-FKBP12-STIM1-ct, and untagged Orai1. Cytosolic Ca2+ changes were visualized with Fura-2 (red trace). Blue trace represents cells with no visible transfection. Means ± SEM are shown (n = 51 and 58 cells for red and blue, respectively, from two separate experiments).

This suggested that the cytosolic segment of STIM1 contains the information necessary for switching from an inactive to an active state. These data thus extend previous work showing that STIM1-ct is sufficient to activate Orai1 channels and that artificial clustering of STIM1 molecules through a rapamycin-inducible multimerization system can activate Orai1 [reviewed in (14)] by showing that an isolated soluble STIM1-ct can be switched from an inactive to an active state. To exclude the possibility that STIM1-ct activation was due to recruitment of an unidentified ER-localized protein to the vicinity of the PM, we performed experiments with an FRB domain targeted to the mitochondrial surface. Here, rapamycin caused clustering of FK-STIM1-ct at contact zones formed between the PM and the adjacent mitochondria (fig. S1A), as well as a substantial increase in cytoplasmic Ca2+ (fig. S1B). AP1510 binds and dimerizes FKBP12 molecules, an association that can be reversed by FK506 (28). No visible clustering was observed when FK-STIM1-ct molecules were dimerized with this compound, although FRET analysis confirmed that AP1510 induced dimerization of STIM1-ct and that this was reversed by FK506 (fig. S1D). Dimerization with AP1510 caused a small cytosolic Ca2+ increase that was reversed by FK506, but this effect was small relative to that of clustering through the ER-membrane–anchored FRB domain (fig. S1C). No Ca2+ influx was apparent when Orai1 itself was directly tagged with a C-terminal FKBP12 module and clustered with the ER-targeted FRB by rapamycin (fig. S2, A and B). The FKBP12-tagged Orai1 construct could be activated through STIM1 by ER Ca2+ depletion (fig. S2C), indicating that it was not the clustering of Orai1 per se but clustering of STIM1-ct that led to Orai1 activation. Note that the large Ca2+ responses shown here depended on Orai1 overexpression; responses were substantially smaller in cells expressing the recruitable form of STIM1 with endogenous Orai1. Furthermore, 50 μM 2-aminoethoxydiphenyl borate (2-APB), which inhibits ICRAC (29), prevented or rapidly reversed the Ca2+ increase, indicating that these responses were due to STIM1-Orai1 activation and not ER Ca2+ release or other Ca2+ signaling mechanisms (fig. S3A).

An inhibitory domain within the first predicted coiled-coil domain keeps STIM1 inactive

As previously shown [see (12) and (14)], STIM1 mutants containing C-terminal truncations to residue 463 activated Orai1 in response to ER Ca2+ depletion (Fig. 3A). Forms of the recruitable FK-STIM1-ct protein with the same truncations showed notable differences in their ability to increase Ca2+ influx after rapamycin-induced clustering (Fig. 3B); thus, successive C-terminal deletions gradually increased the basal Ca2+ concentration, especially after removal of residues 462 to 502. These results confirmed that C-terminal regions in the STIM1 molecule, consistent with those found previously (30, 31), are important for Orai1 inactivation; however, our data indicated that these regions were only modulatory, because truncation to STIM1 residue 462 yielded a protein (TK-STIM1-463stop) that, upon clustering, was capable of further activating Orai1 (Fig. 3B). When TK-STIM1-463stop was truncated from the N-terminal direction to residue 315, the resulting fragment was constitutively active; its ability to activate Orai1 no longer depended on clustering (Fig. 3C). This activating fragment (315 to 462) closely corresponded to the recently described CAD (342 to 448) and SOAR (344 to 442) domains (19, 20), and its activity was indistinguishable from that of CAD (Fig. 3C). Addition of rapamycin to the recruitable versions of these constitutively active fragments decreased cytoplasmic Ca2+ concentration in a fraction of cells (not plotted separately), an observation consistent with a decrease in the free concentration of activating fragments in the cytosol after their recruitment to the ER surface and the inability of these domains to efficiently bridge the ER and PM after rapamycin addition because they lack the C-terminal polybasic tail, which substantially helps their PM localization (32). Cells were kept in a low-Ca2+ (0.2 mM) medium during transfection with the constitutively active constructs (see Materials and Methods) to avoid Ca2+ toxicity. This spared more cells with activated Orai1 and increased the number of cells showing high basal Ca2+ once external Ca2+ was restored to normal levels during the experiments. Even with these modifications, CAD was more toxic than was the residue 315 to 462 STIM1 fragment.

Fig. 3

Cytoplasmic Ca2+ responses evoked by various STIM1 constructs truncated from the C terminus. (A) C-terminal truncations of full-length STIM1 (with intact N terminus) protein to residue 463 have only minor effects on the thapsigargin (Tg, 200 nM)–induced Ca2+ increases. Cells were cotransfected with untagged Orai1 and the indicated mRFP-tagged STIM1 constructs driven by the TK promoter (TK-STIM1). The full-length STIM1 has 685 residues. Means ± SEM are shown (n = 89, 37, 32, and 198 cells for black, green, red, and blue traces, respectively, obtained in two to five separate experiments). (B) The same truncations affect the ability of cytoplasmic STIM1 (STIM1-ct) fragments to activate Orai1 channels after rapamycin-induced clustering. Here, the indicated STIM1-ct fragments were fused with the mRFP-FKBP12 module (FK-STIM-ct) and coexpressed with untagged Orai1 and ER-targeted CFP-FRB. Means ± SEM are shown (n = 356, 254, 259, and 639 cells for black, green, red, and blue traces, respectively, obtained in 9 to 10 separate experiments). (C) N-terminal truncation of the cytoplasmic STIM1 fragments renders them constitutively active. Cells were transfected with the indicated constructs as described in (B). Means ± SEM are shown (n = 208, 82, 124, and 305 cells for black, green, red, and blue traces, respectively, obtained in four to six separate experiments).

Together, these results suggest that the juxtamembrane coiled-coil segment of STIM1 (located between amino acid residues 238 and 343) keeps the CAD/SOAR domain inactive and that STIM1 oligomerization is required for unmasking the latter. We hypothesized that the “inhibitory” coiled-coil segment makes an intramolecular contact with parts of the CAD/SOAR domain required for Orai1 activation and that the interaction site within the STIM1 first coiled-coil domain that keeps CAD/SOAR inactive might show sequence similarity with the Orai1 C-terminal tail. The first coiled-coil domain of STIM1 has several acidic residues with spacing that suggests that they form a highly acidic surface, as does the C-terminal putative coiled-coil domain of Orai1 (Fig. 4A). Indeed, alignment of the two domains revealed a sequence similarity, with identical spacing of the acidic residues. This STIM1 segment also had a highly acidic cluster a few amino acids downstream (Fig. 4A).

Fig. 4

An acidic segment within the coiled-coil domain of STIM1 resembles the Orai1 C-terminal tail and is essential to keep STIM1 in an inactive state. (A) Sequence alignment between STIM1 orthologs and the Orai1 C terminus identifies similarity within the coiled-coil domain as also demonstrated in a helical wheel diagram (hs, Homo sapiens; gg, Gallus gallus; xl, Xenopus laevis). A Leu-to-Ser substitution of the Orai1 residue (indicated by the red arrow) prevents its activation by STIM1 (22). Mutations within this segment of STIM1 are labeled by asterisks (AGAA, or 3EA). Abbreviations for the amino acids are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr. (B) The 3EA mutation makes STIM1-ct(238–685) partially active without rapamycin-mediated clustering (red). Mutation of the four E residues in the distal part of the acidic motif (4EA) renders STIM1-ct(238–685) constitutively fully active. Means ± SEM are shown (n = 143 to 802 cells from three to nine experiments). (C) The same mutations within the full-length STIM1 protein also show partial or full constitutive activity similarly to the D76A mutant (4). Here, the TK-driven mRFP-STIM1 constructs were expressed together with untagged Orai1. Means ± SEM are shown (n = 80 to 785 cells from three to eight experiments). (D) Unlike wild-type STIM1 (upper two rows), the 4EA mutant mRFP-STIM1 forms constitutive clusters together with Orai1-YFP (lower row). (E) Plasma membrane recruitment of STIM1-ct(238–343) inhibits Orai1-mediated Ca2+ influx activated by the constitutively active STIM1-ct(315–462). Here, COS-7 cells were transfected with nonrecruitable mRFP-fused STIM1-ct(315–462), untagged Orai1, and the PM recruiter (PM-FRB-CFP) together with either FK-STIM1-ct(238–343) (red) or mRFP-FKBP12only (FK-only, black). Rapamycin decreased the high basal Ca2+ in a fraction of cells (~16%) in the STIM1-ct(238–343)–expressing group but had no effect on cells in the FKBP-only group. Means ± SEM are shown (n = 80 to 247 cells from 5 to 10 experiments).

An acidic region of the first STIM1 coiled-coil domain similar to a segment of the Orai1 C-terminal helix is an autoinhibitory domain

The acidic segment within the first STIM1 coiled-coil domain contains three Glu residues that show a good alignment with the Orai1 C-terminal helix, and an additional adjacent acidic stretch with four Glu residues (Fig. 4A). To determine whether this segment keeps the basic domain from activating Orai1, we mutated the acidic residues. First, we substituted three of the glutamate residues with alanines (3EA mutation) both in FK-STIM1-ct and in the full-length STIM1 molecule [these constructs also had a mutation in the Leu residue that corresponded to the Leu residue of Orai1 necessary for activation (22), but mutation of the first Glu and Leu to AA in STIM1 was without effect (see fig. S3B)]. The 3EA mutation rendered FK-STIM1-ct partially active even in the absence of rapamycin, although rapamycin-induced clustering elicited further activation (Fig. 4B). Intriguingly, strong constitutive activation was observed when the distal four Glu residues were mutated to alanines (4EA mutant), and this mutant showed no response to rapamycin, indicating that the CAD/SOAR domain was fully exposed (Fig. 4B). Similarly, full-length STIM1 molecules carrying the 3EA mutation showed partial constitutive activity when expressed with Orai1, whereas the 4EA mutant was fully active (Fig. 4C). Combination of the two mutations (7EA) did not enhance the activity of the constructs above that of the 4EA mutants. Moreover, all of the constitutively active mutant full-length STIM1 molecules showed clustering and colocalization with expressed Orai1 without depletion of the Ca2+ stores (Fig. 4D shows the 4EA mutant). These results support the hypothesis that the acidic segment acts as a negative regulator of the STIM1 molecule with the negative cluster of the four Glu residues playing a critical role in this process.

Intriguingly, 4EA mutant STIM1 molecules did not show the massive clustering characteristic of the D76A mutant, which contains a mutation within the luminal calcium-binding EF-hand domain (4) when expressed without exogenous Orai1 molecules. Like wild-type STIM1, the 4EA mutant construct was predominantly associated with microtubules, but differed from wild-type STIM1 in that it had a rosary-like appearance, appearing in smaller bead-like structures distributed along the microtubules. These structures turned to larger and more intense clusters after store depletion (fig. S4), which suggests that the 4EA mutation did not cause a gross alteration in the STIM1 molecule that led to its aggregation and activation but that a subtle conformational change produced by this mutation was sufficient to promote activation of the endogenous Orai1 molecules and to create visible clusters when Orai1 molecules were overexpressed (Fig. 4D).

The acidic segment within the STIM1 first coiled-coil domain can act as a decoy to inhibit Orai1 activation

If the acidic segment of STIM1 is indeed similar to the C-terminal Orai1 activation domain, then it would be expected to act as a competitor when recruited to the PM. To test this hypothesis, we coexpressed Orai1, the isolated acidic segment of STIM1 [STIM1-ct(238–343)], and the constitutively active CAD/SOAR domain (315 to 462) to see whether STIM1-ct(238–343) could inhibit CAD/SOAR. We assumed that the affinity of the two interacting STIM1 fragments was relatively low (as are many intramolecular interactions that are reversibly interrupted by conformational changes) and used rapamycin-induced recruitment of the acidic segment to the PM to increase its concentration in the proximity of Orai1. Using this manipulation, we found that a substantial fraction of cells showed a decrease in cytoplasmic Ca2+ concentration after rapamycin addition. No such inhibition was found in any cells in which the mRFP-FKBP12only construct was used instead of the acidic domain (Fig. 4E). These data suggest that the acidic segment within the coiled-coil domain can act as a decoy to bind and inactivate some of the CAD/SOAR domain of STIM1, decreasing the amount available for Orai1 activation.

A polybasic domain within CAD/SOAR is required for Orai1 activation

The importance of the acidic stretch in keeping STIM1 in an inactive state raised the possibility that the segment within the CAD/SOAR domain important for both the intramolecular interaction and the Orai1 activation has a basic character. We identified such a basic sequence in the middle of the CAD/SOAR domain (382 to 386) that is highly conserved in various STIM orthologs (Fig. 5A). Mutant forms of recruitable FK-STIM1-ct(238–685) or the full-length STIM1 molecule in which four of these basic residues were mutated to AGAG were unable to activate Ca2+ influx through the Orai1 proteins (Fig. 5B). Moreover, a constitutively active wild-type STIM1-ct(315–462) fragment that causes a strong constitutive activation of overexpressed Orai1 (19, 21) was unable to do so when subjected to the 4K (AGAG) mutation (Fig. 5C). These experiments indicate that the basic region within the CAD/SOAR domain is indeed necessary for Orai1 activation. While this manuscript was in preparation, these same basic residues were identified by Calloway et al. as critical for Orai1 activation (26).

Fig. 5

A polybasic segment within the CAD/SOAR domain is responsible for Orai1 activation. (A) Sequence alignment within the CAD/SOAR domain for various STIM1 orthologs and STIM2 (hs, H. sapiens; gg, G. gallus; xl, X. laevis; dr, Danio rerio; dm, Drosophila melanogaster). The mutated Lys residues are labeled with an asterisk. (B) The 4K mutation renders full-length STIM1 proteins inactive. COS-7 cells were transfected with TK-driven STIM1 (black) or its 4K mutant (red) together with untagged Orai1 and cytoplasmic Ca2+ monitored with Fura-2. Thapsigargin (Tg, 200 nM) was added to deplete Ca2+ stores. Means ± SEM are shown (n = 132 to 160 cells, obtained in four to five separate experiments). (C) The 4K mutation makes the STIM1-ct fragment unresponsive to rapamycin (red). Means ± SEM are shown (n = 170 to 239 cells obtained in six to seven separate experiments). (D) The 4K (AGAG) mutation inactivates the otherwise constitutively active STIM1-ct(315–462) segment. Means ± SEM are shown (n = 163 to 244 cells obtained in five to seven separate experiments). (E) The constitutively active STIM1(315–462) piece causes clustering and colocalization with Orai1 in COS-7 cells (white arrows). However, the 4K mutation prevents clustering of Orai1 and the STIM1 construct remains largely in the cytosol. Confocal images of live cells are shown 1 day after transfection.

Discussion

We identified an acidic segment within the first coiled-coil domain of the STIM1 molecule as an autoinhibitory domain that maintains the STIM1-ct in a quiescent state. Mutation of the acidic residues within this domain rendered STIM1 constitutively active both in its native full-length form and as a STIM1-ct tail. Whereas the constitutively active D76A mutant triggers the native STIM1 activation process by decreasing the Ca2+ affinity of the luminal EF-hand domain (4), the 4EA mutant bypasses this step and becomes active even when its luminal domain is Ca2+-bound.

Our attention was drawn to this region of STIM1 by its resemblance to the Orai1 C-terminal helical tail. This similarity suggested that the STIM1 acidic segment might mediate an intramolecular interaction with the CAD/SOAR domain, which interacts with and activates Orai1 channels (19, 20). We propose that STIM1 molecules with neutralizing mutations within this region (4EA) are unable to keep the STIM1-ct in an inactive state, thus allowing unimpeded access of the CAD/SOAR domain to Orai1. These findings, combined with the importance in Orai1 activation of the basic stretch in the CAD/SOAR domain, are consistent with a mechanistic model in which, under basal (quiescent) conditions, the CAD/SOAR-positive residues form an intramolecular interaction with the acidic segment. This intramolecular silencing is interrupted during normal STIM1 activation and is disrupted in the 4EA mutant form of STIM1. This model is reminiscent of how certain protein kinases are kept inactive by their own pseudosubstrate sequences (33). Direct evidence to support this model, however, would require demonstration of an interaction between the acidic and the basic segments of STIM1, and experiments are in progress to address this question using separately expressed domains. It is likely that this interaction depends on a conformation that is regulated by luminal Ca2+ binding of STIM1 (likely transduced by oligomerization) and has substantial structural constraints. On the other hand, the isolated CAD and SOAR domains form multimers (CAD forms a tetramer in solution) (19), and it is possible that CAD interactions with the acidic region of the first coiled-coil of STIM1 could also affect the ability of CAD to multimerize and, reciprocally, CAD multimerization could impair its ability to reassociate with the inhibitory STIM1 segment.

Alternatively, the constitutive activity of the 4EA STIM1 mutants could result from a simple disruption of the structure of the first coiled-coil domain that unmasks the CAD/SOAR domain. Theoretically, the 4EA mutation could even cause clustering of STIM1 and thereby trigger the activation process. However, unlike the D76A mutant, the 4EA or 7EA mutant STIM1 molecules did not show massive clustering without Orai1 overexpression and underwent clustering only after store depletion; these observations argue against constitutive clustering, but a structural distortion within the coiled-coil region cannot be ruled out. However, a major structural defect would more likely render STIM1 inactive than constitutively active. We also considered the possibility that mechanical stretching of STIM1-ct during formation of the ER-PM bridge contributes to its activation. Although such a mechanism cannot be discounted (even in the natural activation process), the fact that we observed similar activation with the mitochondrial anchor, which probably does not provide the same pulling force, and with a substantially longer [9× helical linker; see (27)] ER-targeted FRB construct (see fig. S6), makes mechanical pulling less likely.

One remaining question concerns how the luminal ER Ca2+ decrease leads to the unmasking of the CAD domain. A major conformational change within the isolated luminal EF-hand and SAM domains has been demonstrated upon Ca2+ unbinding (13), and oligomerization of full-length STIM1 in response to Ca2+ depletion has been well documented by FRET analysis (34). The role of the various STIM domains in this process was investigated in a recent study by Covington et al. (35). These authors showed that overexpressed STIM1 molecules could self-associate even in the resting state but only when they contained the STIM-ct. A STIM1 mutant lacking the entire cytoplasmic domain showed no basal oligomers but still responded to ER Ca2+ depletion with oligomerization. Covington et al. also found that STIM1 truncated after the first coiled-coil (CC1) domain promoted constitutive self-association that was unresponsive to Ca2+ depletion unless the CAD domain was added. This study established an important role of the CAD domain in regulating oligomerization, with data suggesting that communication exists between the CC1 and the CAD domains.

Our rapamycin-inducible activation of the STIM1-ct suggests that forcing STIM1 molecules to tightly line up in an orderly fashion (their N termini being fixed) mimics the clustered state. However, in our studies, dimerization of the STIM1-ct was inefficient at activating Ca2+ influx. Given the uncertainty regarding the “basal” oligomerization state of expressed STIM1-ct, we do not know how many STIM1-ct molecules are associated after dimerization via their FKBP12 modules. This raises the question of how many STIM1-ct molecules are required to activate Orai1 channels. This was recently addressed in a study using Orai1 channels fused with STIM1-ct(336–485) with various stoichiometries (36). These studies showed that maximal currents could be achieved by four Orai1 subunits each fused with a tandem STIM1-ct(336–485). Although these data suggest that two SOAR domains may be sufficient to activate one Orai1 molecule when tethered to the channel, they may not be able to do so efficiently when the molecules are separated as in their natural state.

The similarity between the STIM1 acidic domain and the Orai1 C-tail drew our attention to the importance of the acidic residues in the latter during Orai1 activation. It has been reported that removal of all Glu and Asp residues within this cassette is required to eliminate thapsigargin-induced Ca2+ influx (24). FRET studies with polybasic domain mutant STIM1 and acidic mutant Orai1 molecules have indicated that the putative electrostatic interaction is important for Orai1 activation, although the two molecules still show FRET when these charges are neutralized (26). More studies are needed to understand the molecular sequences that transmit the ER luminal Ca2+ change to the Orai1 channels.

Autoinhibitory segments in the cytosolic STIM1 different from the one identified in this study have been described in three separate studies (19, 21, 37). One such region was located between residues 470 and 490 of STIM1 in the two former studies and between 445 and 475 in the latter, and there is a highly acidic sequence between residues 475 and 483. We confirmed that, as previously described (19, 21), deletion of this acidic region in the STIM1-ct constructs (as in the 238 to 463 STIM1 piece) yielded higher constitutive activity in a number of cells. A recent elegant study, analyzing direct interactions between recombinant cytosolic STIM1 pieces and Orai1 channels expressed in yeast, found that the 233 to 463 piece of STIM1 interacted with and activated Orai1 poorly (38), which is consistent with the transient activation of Ca2+ influx we observed with this fragment, and suggests that sequences between 463 and 502 contain sites that stabilize STIM1-Orai1 interactions. Recently, the same acidic segment found within STIM1-ct(475–483) was implicated in the fast Ca2+-dependent inactivation of Orai1 (30, 31). Together, these data suggest that the 475 to 483 region is involved in Ca2+-dependent inactivation rather than in keeping the full-length STIM1 molecule inactive in quiescent cells.

Several modifiers are superimposed on the basic activation mechanism of Orai1 by STIM1 suggested here. Drosophila STIM roughly corresponds to the minimal segment of mammalian STIM1(1–499) (39) required to elicit almost full Orai1 activation. Nevertheless, STIM1 truncations in the cytoplasmic region that extends beyond residue 502 have identified several other STIM1 regions that modify the ICRAC activation and inactivation pattern [reviewed in (12)]. Furthermore, phosphorylation of STIM1 in regions beyond residue 482 can prevent STIM1 activation and is responsible for the inability of thapsigargin to induce Ca2+ influx during the cell cycle (40).

In summary, our data suggest the existence of an autoinhibitory intramolecular interaction within the cytoplasmic segment of STIM1 molecules. We identified an acidic region within the STIM1 first coiled-coil domain that keeps STIM1 in an inactive state and a short basic region within the activation domain of STIM1 that appears to be important for Orai1 activation and could be silenced by the acidic region. The similarity between the acidic inhibitory STIM1 segment and the C-terminal helical segment of Orai1 presumed to be the site of activating interaction with the STIM1 molecule further supports this model. This mechanism of intramolecular silencing resembles the regulation of kinases by intramolecular pseudosubstrate binding and may provide an efficient means for reversible activation of Orai1-mediated Ca2+ influx.

Materials and Methods

Materials

Rapamycin and thapsigargin were purchased from Calbiochem. Adenosine 5′-triphosphate (ATP) was obtained from Sigma. Fura-2 AM was from Molecular Probes (Invitrogen). All other chemicals were of the highest analytical grade.

DNA constructs

Constructs and the primers used to generate them are listed in table S1. All STIM1 constructs were based on previously described cytomegalovirus (CMV)– and thymidine kinase (TK)–driven mRFP- or yellow fluorescent protein (YFP)–tagged (luminally) human STIM1 protein (27). In all experiments using STIM1 molecules with intact ER luminal domains, we used TK-driven versions to keep abundance low. For the recruitable cytosolic fragments, the various STIM1 pieces obtained by polymerase chain reaction (PCR) (using Pfu polymerase) were inserted in place of the polyphosphoinositide 5-phosphatase enzyme between Pvu I and Kpn I restriction sites in the mRFP-FKBP12-5-ptase construct described in (41). Mutations were introduced with the QuikChange mutagenesis kit (Stratagene). The Orai1 constructs used in this study have been described (27), as were the PM- and ER-targeted FRB constructs. The mitochondrial recruiter contained the small sequence from AKAP1 (34 to 63) (MAIQLRSLFPLALPGLLALLGWWWFFSRKK) described in (42). Orai1-mRFP-FKBP12 was obtained by cloning a PCR product of FKBP12 flanked with Xba I and Sgr AI sites. PCR product for this cloning was generated with mRFP-FKBP12-5-phosphatase as a template [previously described in (27)]. To cut the Orai1-mRFP constructs with Xba I, we amplified and purified plasmids from a Dam(−) bacterial strain. All constructs were verified by dideoxy sequencing.

Cytoplasmic Ca2+ measurements and confocal microscopy

COS-7 cells were cultured on glass coverslips (3 × 105 cells per 35-mm dish) and transfected with the indicated constructs (0.5 μg of DNA per dish, each) with Lipofectamine 2000 for 24 hours as previously described (27). For calcium experiments, cells were loaded with 3 μM Fura-2 AM for 45 min in Hepes-buffered M199, Hanks’ salt solution containing 0.1% bovine serum albumin, 0.06% pluronic acid, and 200 μM sulfinpyrazone at room temperature. Because of the Ca2+ toxicity caused by expression of the constitutively active constructs, cells were kept in a low-Ca2+ (0.2 mM) medium during transfection, which greatly increased the number of cells showing the high basal Ca2+ during the experiments. Calcium measurements with Fura-2 were performed in modified Krebs-Ringer solution containing 120 mM NaCl, 4.7 mM KCl, 1.2 mM CaCl2, 0.7 mM MgSO4, 10 mM glucose, and 10 mM Na-Hepes (pH 7.4) at room temperature with an Olympus IX70 inverted microscope equipped with a Lambda DG4 illuminator and a MicroMAX-1024BFT digital camera and the appropriate filter sets to record F340/F380 fluorescence ratios. The MetaFluor (Molecular Devices) software was used for data acquisition and analysis. For confocal analysis, the coverslips with the transfected cells were mounted on the heated (35°C) stage of a Zeiss 510Meta scanning confocal microscope and images were taken at selected times before and after stimulation with the appropriate chemicals. Some experiments were performed at room temperature, which allows a better assessment of the clustering process of full-length STIM1. Images were analyzed after acquisition with Adobe Photoshop, but only linear transformations were allowed to use the full dynamic range of the 12-bit images.

FRET measurements

COS-7 cells were cotransfected with the YFP- and mRFP-tagged versions of the various STIM1 constructs. Cells were examined with the same microscope used for Ca2+ measurements, but a different dual dichroic mirror was used in the microscope along with a Dual-View beam splitter (565D, 535/30, and 630/75; Photometrics) to separate the YFP and mRFP signals. The ratio of the signals obtained in the respective channels (red/green) by 488 excitation was used as an indication of a FRET change after stimulation. Some FRET experiments were done in cell suspension by means of a PTI fluorescence spectrophotometer.

Western analysis

To determine the abundance of the individual constructs, we transfected cells as described above, washed them with phosphate-buffered saline, and lysed them in Laemmli buffer the following day. Lysates were boiled, briefly sonicated, and then subjected to SDS–polyacrylamide gel electrophoresis with 8 to 12% precast gradient gels (Novex) and transferred to nitrocellulose membranes. A rat monoclonal anti-mRFP antibody (5F8) (ChromoTek GmbH) was used to visualize the expressed proteins with the Odyssey infrared detection system (fig. S5).

Acknowledgments

Acknowledgments: Confocal imaging was performed at the Microscopy and Imaging Core of the National Institute of Child Health and Human Development (NICHD), NIH, with the assistance of V. Schram and J. T. Russell. Funding: This research was supported in part by the Intramural Research Program of the NICHD, NIH. P.V. was supported by the Hungarian Scientific Research fund (OTKA NF-68563) and the Medical Research Council (ETT 440/2006). I.M.M. was supported by a Fellowship from the Spanish Personnel Research Training Program from the Ministry of Science and Education of Spain. Author contributions: M.K.K. generated most of the STIM1 constructs and performed all Ca2+ measurements, analyzed the data, and participated in the confocal studies and the planning of the experiments. I.M.M. created the first STIM1-ct constructs and obtained the initial observations. P.V. made all of the FRB-containing recruiting constructs and built the station for Ca2+ measurements. T.B. planned the experiments, participated in confocal microscopy experiments, and wrote the manuscript. Competing interests: The authors declare that they have no competing interests.

Supplementary Materials

www.sciencesignaling.org/cgi/content/full/3/148/ra82/DC1

Fig. S1. Clustering of STIM1 cytosolic fragment in rapamycin-induced mito-PM contact zones activates Ca2+ entry.

Fig. S2. Clustering of Orai1 is not sufficient to increase cytosolic Ca2+ concentration.

Fig. S3. Inhibition of STIM1-ct–induced Ca2+ influx by 2-APB in COS-7 cells.

Fig. S4. Morphology of TK-driven YFP-tagged STIM1 mutant proteins in COS-7 cells without expression of Orai1.

Fig. S5. Western analysis showing the expression and integrity of the various STIM1-ct forms in both their recruitable (FKBP12-fused) and nonrecruitable versions.

Fig. S6. Effect of linker length on the ability of STIM1-ct to activate Orai1-mediated Ca2+ influx.

Table S1. PCR primers used to amplify the segments cloned in the various constructs.

References and Notes

View Abstract

Navigate This Article