Mast cell degranulation represents one of the earliest described examples of a Ca²⁺-dependent biological process in non-excitable cells (Foreman et al., 1973; Kanno et al., 1973). The capacity for IgE-mediated activation of this process via crosslinking of IgE receptor (FcεRI) complexes at the mast cell surface was established in the 1970s (Metzger, 1978). Since biochemical identification (Metzger et al., 1986) and cloning (Ravetch and Kinet, 1991) of the three different subunits of FcεRI in the 1980s, much has been learned about the signaling pathways activated via this prototypic multichain immune recognition receptor and their roles in stimulated Ca²⁺ responses leading to granule exocytosis (Siraganian et al., 2002; Kraft and Kinet, 2007). In the present review we focus on the mechanisms by which FcεRI signaling mediates elevation of cytoplasmic Ca²⁺ and the consequences of this process in mast cell functional responses, including granule exocytosis, chemotaxis to antigen, and stimulated endosomal trafficking, including cytokine secretion and recycling endosomal exocytosis. We will emphasize contributions to this area from our own laboratory, and we will focus on subcellular spatiotemporal relationships where this information is available.

Although the importance of stimulated Ca²⁺ influx for mast cell degranulation has been known for decades, the mechanisms by which FcεRI-mediated Ca²⁺ entry occurs and the role of this process in stimulating granule exocytosis are incompletely understood. Mast cells do not exhibit voltage gated Ca²⁺ influx, a property of excitable cells (Jarvis and Zamponi, 2007). However, release of Ca²⁺ from endoplasmic reticulum (ER) stores by antigen or thapsigargin (an irreversible inhibitor of the SERCA ATPase, which normally maintains high Ca²⁺ levels in the ER) is sufficient to activate Ca²⁺ influx necessary for degranulation from these cells (Wolfe et al., 1996). This Ca²⁺ influx mechanism is commonly referred to as store-operated Ca²⁺ entry (SOCE), and it was originally shown by Putney and colleagues to depend critically on depletion of luminal Ca²⁺ from the ER (Takemura et al., 1989). The Ca²⁺ channel that mediates SOCE, known as the calcium release activated calcium (CRAC) channel, was defined electrophysiologically about two decades ago by Hoth and Penner (1992) in mast cells and by Lewis and Cahalan (1989) in T cells. However, its molecular identity remained unknown until 2006, when three papers from independent groups described cloning and functional reconstitution of the tetraspan channel protein Orai1/CRACM1 (Feske et al., 2006; Vig et al., 2006; Zhang et al., 2006).

In 2005, studies using large-scale siRNA screens to identify proteins important for SOCE discovered another protein, STIM1, that is critical for activation of Orai1 (Liou et al., 2005; Zhang et al., 2005). Unlike the CRAC channel protein, which is localized to the plasma membrane, STIM1 is an ER-localized Type I membrane protein that redistributes to discrete regions of the ER proximal to the plasma membrane upon Ca²⁺ depletion from ER stores (Liou et al., 2005; Wu et al., 2006). The mechanism of this redistribution, which appears to be critical for activation of Orai1/CRACM1, involves oligomerization of STIM1 that results from the loss of Ca²⁺ binding to EF-hand motifs in a luminal domain of this protein (Liou et al., 2007). Recent knockout studies demonstrated essential roles for STIM1 (Baba et al., 2008) and Orai1/CRACM1 (Vig et al., 2008) in mast cell degranulation and other processes in the allergic response (reviewed in Di Capite et al., 2011).

Our laboratory developed an imaging-based method to monitor the time-dependence of stimulated association of Orai1 with STIM1 that utilizes fluorescence resonance energy transfer (FRET) from monomeric AcGFP-Orai1 to STIM1-mRFP co-expressed in RBL mast cells (Calloway et al., 2009). Using this method, we found that thapsigargin stimulates much more extensive association of Orai1 with STIM1 than does antigen-mediated crosslinking of IgE receptors in the presence of extracellular Ca²⁺. However, removal of extracellular Ca²⁺ results in antigen-stimulated FRET that is comparable to that stimulated by thapsigargin in the presence of Ca²⁺, indicating that antigen-stimulated association of Orai1 with STIM1 is highly regulated in these cells (Calloway et al., 2009). Never the less, this more highly regulated coupling can still achieve robust SOCE in response to antigen (Parekh et al., 1997; Vasudevan et al., 2009). Interestingly, G protein-coupled receptors typically activate transient Ca²⁺ responses in mast cells in response to ligands such as adenosine or prostaglandin E2, but they fail to activate SOCE (Gilfillan and Beaven, 2011).

In our initial study, we defined an acidic amino acid segment in a putative coiled-coil sequence at the C-terminus of Orai1 that is important for stimulated SOCE (Calloway et al., 2009; Figure 1). We subsequently showed that this acidic segment (E272–D291) couples to a short basic sequence (K382–R387) in the C-terminal segment of STIM1 to activate Ca²⁺ entry under conditions in which STIM1/Orai1 association is promoted (Calloway et al., 2010). This apparent electrostatic interaction, while necessary for Ca²⁺ entry, is not necessary for stimulated association of STIM1 with Orai1, which depends on a larger, ∼110 amino acid sequence called the “CAD domain” that can constitutively activate CRAC channels when expressed as a cytoplasmic protein (Kawasaki et al., 2009; Muik et al., 2009; Park et al., 2009; Yuan et al., 2009). More recent studies have shown that the sequence comprising five basic amino acids within the CAD domain is inaccessible to Orai1 prior to ER store depletion because of a conformational restriction in STIM1 that sequesters it by intramolecular association with an acidic sequence (E302–D322) that is N-terminal to the CAD domain in the STIM1 cytoplasmic tail (Korzeniowski et al., 2010; Muik et al., 2011). By a mechanism that is still unclear, homo-clustering of STIM1 following depletion of ER luminal Ca²⁺ results in a conformational transition to a more extended C-terminus that readily couples to Orai1 to activate Ca²⁺ influx into the cytoplasm through a productive Orai1 channel.

Many ion channels are regulated by phosphoinositides in the plasma membrane, but the roles for these lipids in SOCE are controversial (Huang, 2007). Korzeniowski et al. (2009) demonstrated that activation of SOCE is sensitive to inhibition of phosphatidylinositol 4-kinase, but is not prevented by acute depletion of phosphatidylinositol 4,5-bisphosphate (PIP₂) at the plasma membrane of COS-7 cells, suggesting a role for PI4P but not for PIP₂. Walsh et al. (2010) showed that inhibition of multiple pathways of PIP₂ generation is necessary to prevent thapsigargin-mediated translocation of STIM1 to the plasma membrane in HeLa cells, but that co-expression of Orai1 permits STIM1 to concentrate at ER–PM junctions even in the absence of PIP₂. A previous study in our laboratory provided evidence that different PI5-kinase family members synthesize functionally different pools of PIP₂ (Vasudevan et al., 2009). To explore whether these functionally different pools of PIP₂ might be synthesized in detergent-resistant ordered lipid domains and detergent-sensitive disordered lipid domains (Lingwood and Simons, 2010), we used transient expression of an inositol 5′-phosphatase that is targeted to either ordered or disordered lipids in RBL mast cells, and we found that the thapsigargin-stimulated association of STIM1 with Orai1 is regulated by the balance between PIP₂ associated with ordered lipids and PIP₂ associated with disordered lipids, such that PIP₂ in ordered lipid domains promotes this association and PIP₂ in disordered lipid domains reduces it, with corresponding alterations in SOCE (Calloway et al., 2011). Furthermore, we found that over-expression of PIP5-kinase Iβ, which enhances PIP₂ levels in both ordered and disordered lipid domains, enhances the stimulated association of STIM1 and Orai1, whereas over-expression of PIP5-kinase Iγ, which enhances PIP₂ levels only in the disordered lipid domains, has an inhibitory effect on this association. This regulation of STIM1/Orai1 coupling by PIP₂ could be attributed to a basic sequence in the N-terminal segment of Orai1 and a basic sequence at the C-terminus of STIM1 because deletion of either of these prevents the effects of both PI5-kinase over-expression and targeted inositol 5-phosphatase expression on stimulated FRET (Calloway et al., 2011). A simple model for these effects is illustrated in Figure 2, in which ER store depletion causes STIM1 to bind initially to PIP₂ in ordered domains at the plasma membrane, followed by redistribution of Orai1 from disordered regions of the plasma membrane to its association with STIM1 in ordered lipid domains. We are currently testing this hypothesis.

It is well-established that antigen-stimulated Ca²⁺ mobilization in mast cells is initiated by activation of phospholipase C (PLC) γ-1 and γ-2, which depends on tyrosine phosphorylation of specific residues in these proteins by Syk kinase and/or by the Tec kinase BTK (Gilfillan and Rivera, 2009; Ma and Beaven, 2009). These lipases hydrolyze PIP₂ to produce inositol 1,4,5-trisphosphate (IP₃) and diacylglycerol, and it is IP₃-mediated release of Ca²⁺ from ER stores that activates the association of STIM1 with Orai1 to form productive CRAC channels as discussed above. Using fast confocal imaging of the genetically encoded Ca²⁺ sensor, GCaMP2, and proximal antigen delivery via a micropipette, we found that antigen-stimulated Ca²⁺ responses usually initiate as a fast Ca²⁺ wave that begins at the tips of extended membrane protrusions, which are common on well-attached cells (Figure 3; Cohen et al., 2009). Not surprisingly, this Ca²⁺ wave initiation depends on PLC activity, but it does not utilize CRAC channels because it is not inhibited by Gd³⁺, which selectively blocks these channels in RBL mast cells (Cohen et al., 2009; Calloway et al., 2011). In the absence of extracellular Ca²⁺, wave initiation occurs more frequently at other sites along the cell body, suggesting that Ca²⁺ entry is important for its initiation from protrusions. A previous study provided evidence that certain six transmembrane-structured TRPC cation channel proteins can contribute to antigen-stimulated Ca²⁺ entry in RBL mast cells (Ma et al., 2008). Our results from shRNA knockdown experiments indicated that wave initiation from protrusions depends on Ca²⁺ entry via two particular TRPC channels, TRPC1 and TRPC3, particularly at limiting concentrations of antigen (Cohen et al., 2009).

Mast cells are receiving increasing attention as mediators of innate immune responses in addition to their better-known role in the adaptive allergic response (Metz et al., 2007; Brown et al., 2008). Although mast cells are recruited to the gut where they undergo dramatic proliferation in response to certain parasitic infections such as with Trichinella spiralis (Friend et al., 1996), a recent study has shown that they do not contribute significantly to parasite clearance during the secondary immune response (Blum et al., 2009). This suggests that mast cells participate in the initiation of the Th2 immune response, rather than in the effector phase. Our studies summarized above suggest that antigen-specific recruitment of mast cells to small intestinal villi during such immune responses may be driven by Ca²⁺ mobilization in a process that utilizes STIM1-mediated activation of CRAC channels that are formed by Orai1. Understanding the mechanism by which these proteins and the Ca²⁺ response they elicit provide a cell polarization cue in response to antigen will be a challenging goal for future studies.

Mast cells also respond to other chemotactic cues, such as the lipophilic ligand for G protein-coupled receptors, sphingosine-1-phosphate (Jolly et al., 2004; Olivera et al., 2007). We found that RBL cells chemotax toward this ligand with a robust response, but in this case, chemotaxis is not dependent on extracellular Ca²⁺ (J. Lee et al., submitted). The mechanism of signaling involved in this process remains to be determined.

RBL mast cells undergo robust phagocytosis of beads coated with ligands for IgE receptors, and, under these conditions, granule exocytosis is severely restricted (Pierini et al., 1996). The mechanism for this differential regulation is unclear, and recent experiments indicate that Ca²⁺ mobilization and PKC recruitment to the plasma membrane are not limiting under these conditions (K. Corwith, unpublished results). In macrophages, trafficking of recycling endsomes to the forming phagosome has been shown to contribute to this process (Cox et al., 2000), and we are interested in the potential role of this trafficking in mast cell phagocytosis and in the signaling consequences. Recent studies have revealed a role for recycling endosomal trafficking in the maintenance of the parasitic vacuole that is formed by invading Toxoplasma gondii in mast cells, as well as disruption of normal FcεRI signaling by this process (N. Smith et al., in preparation). Together, these studies reveal dynamic alterations of IgE receptor signaling by phagocytosis or parasite invasion, but the molecular mechanisms of these alterations remain to be characterized.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.