Orai1, STIM1, and their associating partners

Abstract

Store-operated Ca²⁺ (SOC) entry is one of the major mechanisms to raise intracellular Ca²⁺ concentration in non-excitable cells. Ca²⁺-release-activated Ca²⁺ (CRAC) channels are a subtype of SOC channels that are extensively characterized in immune cells. Identification of STIM1 as an endoplasmic reticulum Ca²⁺ sensor and Orai1 as the pore subunit has dramatically advanced the molecular understanding of CRAC channels. Recent efforts have focused on understanding the physiological aspects of CRAC channels at an organism level using transgenic animal models and at a molecular level using electrophysiological and biochemical tools. In this review, we summarize our current understanding of the interacting partners of Orai and STIM proteins in the regulation of CRAC channel activity and other non-CRAC channel-related functions.

Sonal Srikanth received her PhD in Life Sciences from the National Center for Biological Sciences, Bangalore, working with Dr Gaiti Hasan. She did her postdoctoral research with Drs Anjana Rao at Harvard Medical School and Yousang Gwack at UCLA. She is currently an Assistant Research Physiologist in the Department of Physiology at UCLA. Yousang Gwack received his MS and PhD in Biological Science from the Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea, working with Dr Joonho Choe. He did his postdoctoral research with Drs Jae U. Jung and Anjana Rao at Harvard Medical School before moving to the Department of Physiology, UCLA, as an Assistant Professor.

Introduction

In non-excitable cells, Ca²⁺ entry via store-operated Ca²⁺ (SOC) channels is a predominant mechanism to increase the intracellular Ca²⁺ concentration ([Ca²⁺]_i) (Cahalan & Chandy, 2009; Putney, 2009; Hogan et al. 2010; Lewis, 2011). SOC channels were so named because they are activated by depletion of intracellular Ca²⁺ stores (Putney, 1986, 2009). The Ca²⁺-release-activated Ca²⁺ (CRAC) channel is a specialized class of SOC channel in immune cells. RNA interference (RNAi) screening approaches greatly facilitated identification of the molecular components of CRAC channels. Initial RNAi screens identified STIM1, a Ca²⁺-binding protein localized predominantly in the endoplasmic reticulum (ER) membrane as an important component of CRAC channels (Liou et al. 2005; Roos et al. 2005; Zhang et al. 2005). Subsequent genome-wide RNAi screens identified Orai1 as a pore subunit of the CRAC channels (Feske et al. 2006; Vig et al. 2006; Zhang et al. 2006; Gwack et al. 2007).

Identification of CRAC channel components has greatly advanced the understanding of the Ca²⁺ signalling pathway in T cells. Antigen engagement of the T cell receptor (TCR) triggers phospholipase C-mediated generation of inositol 1,4,5 trisphosphate (IP₃). IP₃ binds to the IP₃ receptor (IP₃R) on the ER membrane and releases Ca²⁺ from the ER (Fig. 1, Phase I). Upon ER Ca²⁺ depletion, STIM1 loses its bound Ca²⁺, multimerizes, translocates to regions of ER that are proximal to the plasma membrane (PM), mediates clustering of Orai proteins on the PM, and stimulates Ca²⁺ entry (Phase II) (Liou et al. 2005; Roos et al. 2005; Zhang et al. 2005). Opening of Orai1 raises [Ca²⁺]_i, followed by increased Ca²⁺ accumulation in the mitochondria. Increased [Ca²⁺]_i also affects gene transcription by activating the Ca²⁺–calmodulin/calcineurin–NFAT (nuclear factor of activated T cells) pathway (Phase III) (Hogan et al. 2003). Ca²⁺-bound calmodulin forms a complex with the protein phosphatase calcineurin, which in turn dephosphorylates the heavily phosphorylated, cytoplasmic NFAT. Dephosphorylation of NFAT exposes its nuclear localization signal sequence (NLS) and induces its translocation into the nucleus. Nuclear NFAT forms a multimeric protein complex with itself or with other transcription factors to induce gene transcription involved either in cytokine production, cell proliferation, growth arrest, or cell death, depending on the amplitude and duration of [Ca²⁺]_i elevation (Macian et al. 2002; Kim et al. 2011).

Figure 1. Molecular components of CRAC channels in T cells.

Antigen engagement of T cell receptor stimulates phospholipase C (PLC) that hydrolyses PIP₂ (phosphatidylinositol 4,5-bisphosphate) into IP₃ (inositol 1,4,5 trisphosphate) and diacyl glycerol. Generated IP₃ binds to the IP₃ receptor (IP₃R) on the ER (endoplasmic reticulum) membrane to empty the ER Ca²⁺ store (Phase I, store depletion). ER Ca²⁺ depletion is sensed by STIM1, an EF-hand protein localized in the ER membrane, which translocates into the junctions of plasma membrane and ER membrane. STIM1 couples with Orai1 at the junctions by protein interaction and induces opening of CRAC channels (Phase II, store-operated Ca²⁺ entry). Opening of Orai1 raises the intracellular Ca²⁺ concentration and enhances mitochondrial Ca²⁺ uptake. The increased Ca²⁺ ions trigger a broad range of downstream signalling pathways including the Ca²⁺–calmodulin/calcineurin-NFAT (nuclear factor of activated T cells) pathway (Phase III, transcriptional events). Ca²⁺-bound calmodulin (CaM) forms a complex with a protein phosphatase, calcineurin (Cn), and dephosphorylates the heavily phosphorylated, cytoplasmic NFAT leading to its nuclear translocation. Nuclear NFAT forms a multimeric protein complex of itself or with other transcription factors (e.g. AP-1) to induce gene transcription involved in cytokine production, cell proliferation and cell death depending on the amplitude or duration of [Ca²⁺]_i and the associating partners.

CRAC channel activation involves multiple steps including STIM1 oligomerization, co-clustering of Orai1 and STIM1 at the ER–PM junctions, and STIM1-mediated gating of Orai1 (Liou et al. 2007; Muik et al. 2008, 2009; Navarro-Borelly et al. 2008; Park et al. 2009; Yuan et al. 2009). While the role of Orai1–STIM1 interaction is emphasized in the multiple activation and inactivation steps of CRAC channels, an understanding of the cellular machinery modulating these processes is just gaining momentum. In the current review, we summarize the recent progress in identification of interacting partners of Orai1 and STIM1 that highlight their CRAC channel-related and non-related functions.

Macromolecular Orai1–STIM1 protein complex

Amplified CRAC currents have been observed upon co-expression of Orai1 and STIM1 (Mercer et al. 2006; Peinelt et al. 2006; Soboloff et al. 2006). Furthermore, truncation analysis identified a cytoplasmic region of STIM1, termed the CRAC activation domain (CAD)/STIM1 Orai1 activating region (SOAR) to be sufficient to activate Orai1 (Kawasaki et al. 2009; Muik et al. 2009; Park et al. 2009; Yuan et al. 2009). In addition, in vitro reconstitution using membrane vesicles purified from yeast expressing Orai1 and recombinant STIM1 fragments purified from bacteria showed Orai1-mediated Ca²⁺ efflux (Zhou et al. 2010). These results suggested that Orai1 and STIM1 are sufficient to form functional CRAC channels. However, several studies suggest that numerous molecules are involved in regulating CRAC channel activity under physiological conditions. First, genome-wide RNAi screens performed for identification of the CRAC channel components yielded hundreds of positive hits (Feske et al. 2006; Gwack et al. 2006; Vig et al. 2006; Zhang et al. 2006; Gwack et al. 2007). Second, studies using chemically inducible bridge formation with linkers of variable lengths between the PM and ER membranes showed that Orai1 exists in a macromolecular complex with 11–14 nm protrusion into the cytoplasm (Varnai et al. 2007). Third, biochemical analyses using chemical cross-linking and glycerol gradient fractionation identified Orai1 and STIM1 in a macromolecular protein complex (Srikanth et al. 2010a). Accordingly, numerous interacting partners of Orai1 and STIM1 that regulate various stages of CRAC channel activation have been identified.

Cytoplasmic Ca²⁺ sensors of CRAC channels: calmodulin and CRACR2A

The cytoplasmic regions of Orai1 are predominantly involved in protein interactions (Fig. 2A). The cytoplasmic N and C termini of Orai1 mediate channel opening by interaction with STIM1. CRAC channels are also negatively regulated by Ca²⁺ influx, resulting in Ca²⁺-dependent inactivation (CDI) (Hoth & Penner, 1992, 1993; Zweifach & Lewis, 1995). Mutational studies showed that all the cytoplasmic regions of Orai1 are involved in CDI (Lee et al. 2009; Mullins et al. 2009; Srikanth et al. 2010b). Recent studies have shown calmodulin (CaM) binds to the N terminus of Orai1 at elevated [Ca²⁺]_i to mediate CDI (Mullins et al. 2009).

Figure 2. Summary of associating partners of Orai1 and STIM1.

A, schematic diagram of Orai1 structure and its interacting partners. Orai1 has four transmembrane segments (TM1–TM4). It has two extracellular domains and the second extracellular domain between TM3 and TM4 contains the asparagine (N²²³) residue involved in glycosylation. The TM1 lines the pore. Orai1 contains three intracellular domains including the N terminus, intracellular loop and C-terminal coiled-coil domain. The interacting partners for each domain such as STIM1, SPCA2, CaM, CRACR2A, and caveolin are indicated. Arginine at position 91 was mutated in patients (R91W) with non-functional CRAC channels suffering from severe combined immunodeficiency (indicated in red). B, associating partners of STIM1. STIM1 contains an ER-luminal N terminus, a single transmembrane segment, and a cytoplasmic region. The cytoplasmic region has three coiled-coil domains: a serine/threonine-rich domain, a proline/glutamate/serine/threonine-rich (PEST) domain, and a poly-basic region at the C terminus. The protein interaction domains of each partner determined by functional analyses or pulldown studies are indicated. Golli protein was shown to interact with the cytoplasmic domain of STIM1, but the exact binding site information is not known (dashed line).

Srikanth et al. identified a novel cytoplasmic EF-hand-containing protein, CRACR2A as a regulator of Orai1–STIM1 interaction (Fig. 2A) (Srikanth et al. 2010a). CRACR2A forms a ternary complex by direct interaction with Orai1 and STIM1 in a [Ca²⁺]_i-sensitive manner, with low [Ca²⁺]_i favouring association of CRACR2A and high [Ca²⁺]_i favouring its dissociation (Fig. 2A and B). CaM-binding sites on Orai1 partially overlap with those of CRACR2A. However, CaM binds to the N terminus of Orai1 at high [Ca²⁺], contrary to CRACR2A. These results suggest an interesting scenario where CRACR2A interaction favours an active CRAC channel complex upon store depletion while CaM binding facilitates inactivation of CRAC channels at high [Ca²⁺]. Future studies using transgenic animals will provide a better understanding of the physiological role of CRACR2A.

Golli as a negative regulator of CRAC channels

Golli proteins are products of alternatively spliced isoforms of the myelin basic protein (mbp) gene and contain a unique N-terminal golli domain. Of the three isoforms, golli-BG21 is most extensively investigated. Golli-BG21 is widely expressed and abundant in thymocytes and splenocytes (Feng et al. 2000). Overexpression of golli-BG21 in T cells reduces store-operated Ca²⁺ entry (SOCE) and this inhibitory effect requires myristoylation of its N terminus to allow for PM association (Feng et al. 2006). Furthermore, golli-deficient T cells showed enhanced sensitivity to TCR stimulation, increased SOCE and hyperproliferation in vitro. Unexpectedly, in vivo immunization studies showed reduced symptoms of MOG_35-55 peptide-induced experimental autoimmune encephalomyelitis (EAE) in golli-deficient mice. The authors suggested that in vivo, stimulated golli-deficient T cells undergo growth arrest or anergy due to excessive Ca²⁺–NFAT signalling, resulting in resistance to EAE. A recent study showed binding of golli-BG21 to the cytoplasmic region of STIM1 (Fig. 2B) and its co-localization with Orai1 and STIM1 after store depletion (Walsh et al. 2010b). It was proposed that golli might compete with Orai1 to interact with STIM1 thereby reducing Ca²⁺ entry.

Store-independent Orai1 modulator: secretory pathway Ca²⁺-ATPase 2

The role of CRAC channels has been predominantly emphasized in the immune system. However, identification of STIM1 and Orai1 helped in uncovering their function in various pathological conditions including cancer. Recent studies have shown that inhibition of Orai1 and STIM1 activity greatly diminishes cervical and mammary tumour cell migration and metastasis (Yang et al. 2009; Feng et al. 2010; Chen et al. 2011). Interestingly, an isoform of the secretory pathway Ca²⁺-ATPase, SPCA2 was identified to enhance mammary tumour cell growth by raising [Ca²⁺]_i (Feng et al. 2010). Overexpression of SPCA2 increased [Ca²⁺]_i by directly binding to the N and C termini of Orai1 in a STIM1- and store-independent manner (Fig. 2A). This study for the first time points towards a STIM1-independent mechanism of Orai1 activation. Further studies measuring store-independent Ca²⁺ entry in cells overexpressing SPCA2 and Orai1, and characterization of SPCA2-gated Orai1 currents, similar to that performed for Orai1 and STIM1, would provide important insights into this novel mechanism of regulation of Orai1.

STIM1 and plasma membrane Ca²⁺-ATPase (PMCA)

In previous studies, Bautista et al. showed enhanced Ca²⁺ extrusion by PMCA, which was positively regulated by increased [Ca²⁺]_i upon store depletion (Bautista et al. 2002; Bautista & Lewis, 2004). Recently, the relationship between PMCA activity and SOCE has been addressed in multiple studies. Quintana et al. used anti-CD3 antibody-coated coverslips to induce formation of the immunological synapse (IS), and measured SOCE and subsequently Ca²⁺ extrusion by PMCA (Quintana et al. 2011). The authors observed accumulation of Orai1, STIM1 and PMCA, as well as mitochondria, at the IS, and reduced PMCA activity at sites in close proximity to the mitochondria. Krapivinsky et al. identified a novel 10-transmembrane segment-containing protein, POST (partner of STIM1, TMEM20) as a STIM1-interacting protein (Krapivinsky et al. 2011). While POST overexpression did not significantly alter CRAC currents, it negatively regulated PMCA activity. The authors suggested that the STIM1–POST complex interacts with PMCA after store depletion and lowers Ca²⁺ extrusion, resulting in sustained [Ca²⁺]_i elevation. Using biochemical techniques, Ritchie et al. showed a protein interaction between PMCA and STIM1 that results in reduced Ca²⁺ clearance from the cytoplasm after stimulation with phytohaemagglutinin (PHA) (Fig. 2B) (Ritchie et al. 2012). The authors did not exclude the possibility of involvement of POST in these events.

While Quintana et al. measured Ca²⁺ extrusion rate in the context of the IS, Bautista et al. used thapsigargin to deplete the intracellular Ca²⁺ stores, which does not induce IS formation or local accumulation of mitochondria (Bautista et al. 2002; Bautista & Lewis, 2004; Quintana et al. 2011). Based on these results, Quintana et al. suggested that local, indirect inhibition of PMCA activity at the IS by mitochondrial Ca²⁺ buffering is important for generation of a sustained, global increase in [Ca²⁺]_i necessary for prolonged NFAT activation. An extensive analysis using different stimuli (e.g. thapsigargin vs. anti-CD3) and blockers of mitochondrial Ca²⁺ uptake will provide a better understanding of the relationship between STIM1, POST and mitochondria in the regulation of PMCA and CRAC channels.

Junctate, a structural component of the ER–PM junctions

Orai1 and STIM1 translocate into pre-existing junctional areas, a space of 10–25 nm between the PM and ER membranes (Wu et al. 2006; Varnai et al. 2007). However, the molecular components of these junctions remained poorly understood. In excitable cells (e.g. neurons and muscle cells), proteins localized to the junctions between the PM and ER/sarcoplasmic reticulum (SR) membranes form a structural foundation for regulating the intracellular Ca²⁺ stores and Ca²⁺ entry (Berridge et al. 2003; Carrasco & Meyer, 2011). Various screening approaches have identified junctophilins, mitsugumins, sarcalumenin, junctin, and junctate as important components of these junctional regions (Takeshima et al. 2000; Berridge et al. 2003; Weisleder et al. 2008; Carrasco & Meyer, 2011). It was shown that junctate, a ubiquitously expressed transcriptional isoform of junctin modulates SOCE via interaction with the IP₃R and transient receptor potential type C (TrpC) channels in non-excitable cells (Treves et al. 2000, 2004, 2010). However, its role in the regulation of CRAC channels remained unknown. Recent studies from our group showed that junctate is a structural component of the ER–PM junctions in T cells and recruits STIM1 into these junctions after store depletion (S. Srikanth and Y. Gwack, unpublished observations). These results suggest a conserved function of the components of the ER–PM junctions in excitable and non-excitable cells.

Other associating partners of STIM1 and Orai1

A surface plasmon resonance screen between the immobilized ER-luminal N terminus of STIM1 and various ER-resident proteins identified ERp57 as a binding partner of STIM1 (Fig. 2B) (Prins et al. 2011). Increased SOCE in ERp57-deficient cells suggested its negative role in STIM1 function. Polycystin-1 fragment P100 and stanniocalcin 2 (STC2) have also been shown to interact with STIM1 and regulate its translocation (Woodward et al. 2010; Zeiger et al. 2011). STIM1 is known to associate with and regulate other Ca²⁺ channels including the ARC channels (arachidonate-regulated Ca²⁺ channels), Cav1.2 and TrpC channels (Huang et al. 2006; DeHaven et al. 2007; Mignen et al. 2007; Ong et al. 2007; Worley et al. 2007; Park et al. 2010; Wang et al. 2010). A novel function of STIM1 in ER tubulation by association with EB1, a microtubule tip-binding protein has also been identified (Fig. 2B) (Grigoriev et al. 2008; Honnappa et al. 2009). In addition, a caveolin and dynamin-dependent pathway has been shown to regulate Orai1 internalization in oocytes via protein interaction between caveolin and the N terminus of Orai1 (Yu et al. 2010) (Fig. 2A). These are new and interesting aspects of the function and regulation of STIM1 and Orai1 that need further studies.

Working mechanisms of Orai1- and STIM1-associating proteins

The interacting partners of Orai and STIM proteins during different stages of CRAC channel activation are summarized in Fig. 3. Under resting conditions, Orai1 and STIM1 are distributed at the PM and the ER membrane, respectively. Ca²⁺-bound STIM1 exists as a folded structure mediated by intramolecular interaction between the CAD/SOAR domain and the autoinhibitory region within the coiled-coil domain (Korzeniowski et al. 2010). Upon store depletion, STIM1 unfolds itself, oligomerizes, and translocates to form clusters at the junctional regions.

Figure 3. Molecular mechanism of CRAC channel regulation by multiple Ca²⁺-sensing molecules.

Schematic diagram showing binding partners of CRAC channels during different stages of activation. Under resting conditions, Orai1 and STIM1 are distributed at the PM and the ER membrane. The subunit stoichiometry of Orai1 under resting conditions is currently contentious (indicated with a question mark). Upon store depletion, STIM1 oligomerizes and translocates to form clusters at the ER–PM junctions. Some junctional proteins well studied in excitable cells (e.g. junctophilins and junctate) may play a role in determining the ER–PM junctions (sites of Orai and STIM clustering) in non-excitable cells. The interaction between the polybasic residues of STIM1 and PM phosphoinositides as well as protein interaction between Orai1 and STIM1 plays an important role in their translocation. Several protein interactors including ERp57, P100, golli and CRACR2A have been identified to modulate STIM1 translocation. By physical interactions with Orai1 through the CAD/SOAR domain (coloured in red), clustered STIM1 recruits and activates Orai1 in the PM–ER junctions. After [Ca²⁺]_i increases, Ca²⁺-bound calmodulin interacts with the N terminus of Orai1 and inactivates CRAC channels (bottom). When ER Ca²⁺ is re-filled by sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA), the protein complex dissociates, and Orai1 and STIM1 redistribute to their respective membrane.

Orai1 redistribution into the ER–PM junctions primarily depends on interaction with STIM1 (Luik et al. 2006; Xu et al. 2006). However, STIM1 redistribution into the junctions seems more complex. Several protein interactors including ERp57, P100, golli, and CRACR2A are known to modulate STIM1 translocation. The polybasic residues in the C terminus of STIM1 play a crucial role in STIM1 clustering at the ER–PM junctions by interaction with PM phosphoinositides (Liou et al. 2007; Korzeniowski et al. 2009; Walsh et al. 2010a; Calloway et al. 2011). However, a STIM1 mutant truncated in its polybasic residues was able to translocate when co-expressed with Orai1 (Park et al. 2009). These results suggest that in addition to phosphoinositide interaction, alternative mechanisms mediated by protein interactions may be important for STIM1 translocation. The positive regulators of SOCE (e.g. CRACR2A) also form a complex with Orai1 and STIM1 to stabilize their interaction. After [Ca²⁺]_i increases, negative regulators of SOCE such as calmodulin interact with Orai1 to inactivate the CRAC channels. When ER Ca²⁺ is re-filled by SERCA, the protein complex of Orai1 and STIM1 dissociates.

Future directions

After identification of Orai1 and STIM1, numerous interacting partners have been described. Some of them show clear roles in regulation of the gating properties of CRAC channels or translocation of Orai1 and STIM1. Without doubt, Orai1 and STIM1 are the critical subunits of CRAC channels, and depletion or co-expression of these proteins show a profound effect on CRAC channel activity. However, the effect of gain-of-function or loss-of-function of regulators can be mild, especially in the commonly used overexpression system of Orai1 and STIM1. Therefore, measuring the cumulative outcomes of Ca²⁺ signalling over a long period (e.g. NFAT translocation or cell death) in a physiological condition can be important to examine the functions of novel regulators. A more technically challenging question arises from the identification methods of regulators using biochemical tools because Orai and STIM are membrane proteins, and tend to pull down other proteins non-specifically. To resolve these issues, together with development of enhanced biochemical tools, other functional analysis methods including knockdown and gene manipulation should be combined to cull out the genuine regulators. Because various patterns of Ca²⁺ influx (e.g. amplitude, duration, or frequency) are observed depending on cell types and agonists, another interesting aspect is to identify the regulators governing such activities. Finally, studies aimed at identification of associating partners of Orai and STIM proteins in non-CRAC channel-related functions will further our understanding of their physiological roles.