STIM/Orai signalling complexes in vascular smooth muscle

Abstract

Stromal interaction molecules (STIM1 and STIM2) are single pass transmembrane proteins located mainly in the endoplasmic reticulum (ER). STIM proteins contain an EF-hand in their N-termini that faces the lumen side of the ER allowing them to act as ER calcium (Ca²⁺) sensors. STIM1 has been recognized as central to the activation of the highly Ca²⁺ selective store-operated Ca²⁺ (SOC) entry current mediated by the Ca²⁺ release-activated Ca²⁺ (CRAC) channel; CRAC channels are formed by tetramers of the plasma membrane (PM) protein Orai1. Physiologically, the production of inositol 1,4,5-trisphosphate (IP₃) upon stimulation of phospholipase C-coupled receptors and the subsequent emptying of IP₃-sensitive ER Ca²⁺ stores are sensed by STIM1 molecules which aggregate and move closer to the PM to interact physically with Orai1 channels and activate Ca²⁺ entry. Orai1 has two homologous proteins encoded by separate genes, Orai2 and Orai3. Other modes of receptor-regulated Ca²⁺ entry into cells are store-independent; for example, arachidonic acid activates a highly Ca²⁺ selective store-independent channel formed by heteropentamers of Orai1 and Orai3 and regulated by the PM pool of STIM1. Here, I will discuss results pertaining to the roles of STIM and Orai proteins in smooth muscle Ca²⁺ entry pathways and their role in vascular remodelling.

Mohamed Trebak obtained an MSc in Biochemistry and a PhD in Biochemistry from Université de Liège, Belgium, and was a postdoctoral fellow at the Wistar Institute in Philadelphia, PA, USA, and subsequently in the laboratory of Jim Putney at the National Institute of Environmental Health Sciences (NIEHS/NIH) in Research Triangle Park, North Carolina. He moved in late 2006 to the Albany Medical College, New York as an independent investigator where he is currently an Associate Professor of Cardiovascular Sciences. Initially an immunologist, his interest in store-operated Ca²⁺ signalling led him to investigate the activation mechanisms of transient receptor potential canonical (TRPC) channels and more recently the role of STIM/Orai channels in vascular function and dysfunction. His current research is focused on ion channel regulation and its role in driving pathological remodelling in the vasculature and airways.

STIM and Orai proteins, and Ca²⁺ entry

Increased cytosolic calcium (Ca²⁺) concentrations control a plethora of cell functions ranging from immediate responses such as contraction and secretion to long-term effects on gene regulation, proliferation, migration and differentiation (Berridge et al. 2000). In smooth muscle cells, the two most important Ca²⁺ signalling elements that are central to excitation–contraction coupling are the plasma membrane (PM) voltage-gated L-type Ca²⁺ channels and the ryanodine receptor Ca²⁺ release channels located in the sarcoplasmic reticulum. Smooth muscle cells also express receptor-evoked Ca²⁺ signalling pathways, typically found in non-excitable cells. Ligation of PM receptors that couple to isoforms of phospholipase C (PLC) generally induces Ca²⁺ release from inositol 1,4,5-trisphosphate (IP₃)-sensitive internal stores (Berridge, 1993), and activation, by various means, of voltage-independent Ca²⁺ influx channels at the PM (Bird et al. 2004). These receptor-activated Ca²⁺ entry pathways that are better characterized in non-excitable cells, can be divided into two major classes: (i) store-operated Ca²⁺ (SOC) entry channels activated by the depletion of the Ca²⁺ content in the endoplasmic reticulum (ER) as a result of IP₃-induced Ca²⁺ release through IP₃ receptors (IP₃Rs) (Putney, 1986, 1990; Parekh & Putney, 2005; Potier & Trebak, 2008); and (ii) store-independent Ca²⁺ channels that are activated independently of the content of Ca²⁺ stores by various means, including second messengers produced during downstream PLC-mediated phosphatidylinositol 4,5-bisphosphate (PIP₂) breakdown such as diacylglycerol (DAG) and arachidonic acid (AA) (Barritt, 1999; Trebak et al. 2003; Bird et al. 2004; Shuttleworth, 2009).

Significant progress has been achieved in the past few years regarding the molecular composition and the signalling mechanisms controlling the activation of SOC channels and the electrophysiological current they mediate, the Ca²⁺ release-activated Ca²⁺ current (CRAC) (Hoth & Penner, 1992). The protein STIM1 is the Ca²⁺ store sensor located in the membrane of the ER (Liou et al. 2005; Roos et al. 2005), and the PM protein Orai1 is the SOC channel (Feske et al. 2006; Vig et al. 2006). STIM1 comprises a single transmembrane domain and a low affinity N-terminal EF-hand facing the lumen of the ER. Ca²⁺ store depletion causes STIM1 aggregation and translocation to junctional areas of close ER–PM contacts where a STIM1 C-terminal 100 amino acid SOAR/CAD (STIM-Orai activating region/CRAC activating domain) domain physically interacts with Orai1 C- and N-termini to cause Ca²⁺ entry (Park et al. 2009; Yuan et al. 2009). STIM1 has a homologue, STIM2, that appears to play a role in maintaining Ca²⁺ levels in the ER under basal conditions in the absence of agonist stimulation (Brandman et al. 2007), while Orai1 has two homologues, Orai2 and Orai3, that also form SOC channels when co-expressed with STIM1 in HEK293 cells (DeHaven et al. 2007). As pointed out elsewhere (Trebak, 2011), studies using either knockdown strategies in vitro or knockout mice have clearly established Orai1 as the native SOC channel mediating CRAC currents in a number of cell types. However, with the exception of one example of Orai3 mediating native CRAC currents (Motiani et al. 2010), the role of Orai2 and Orai3 in encoding native SOC channels is largely unknown. This raises the interesting possibility that homo- and hetero-multimeric channels composed of Orai2 and Orai3 (along with Orai1) might encode a different class of Ca²⁺ channels with alternative activation mechanisms whose raison d’être is to enhance the diversity of Ca²⁺ selective channels for the purpose of selective signalling. The extensively characterized AA-regulated Ca²⁺ (ARC) entry channel is a perfect example of a store-independent highly Ca²⁺ selective channel (Shuttleworth, 2009). ARC channels are encoded by heteropentamers of Orai1 and Orai3 (three Orai1 and two Orai3 molecules; Mignen et al. 2008) and are regulated by the pool of STIM1 that is located in the PM (Mignen et al. 2007). The N-terminus of Orai3 appears to determine the selectivity for activation of ARC channels by AA (Thompson et al. 2010). The existence of native Ca²⁺ channels encoded by different Orai stoichiometric assemblies and regulated by alternative mechanisms is currently unknown.

STIM/Orai and smooth muscle contractility

Smooth muscles are structurally and functionally different from either skeletal or cardiac muscles and display extensive heterogeneity; depending on their target organ (vascular tree, lungs, airways, gastrointestinal tract and reproductive organs, etc.) or their location within a specific vascular bed, smooth muscles display differences in the expression of receptors, signalling proteins and ion channels. Some types of smooth muscle show spontaneous rhythmic contractility, but in general smooth muscle contractility is fundamentally different from skeletal and cardiac muscle. For instance, in the type of smooth muscle that will be exclusively addressed below, the vascular smooth muscle (VSM), contractility differs substantially from that of skeletal and cardiac muscle. VSM cells are partially constricted at rest and their increased contractility, typically in response to neuronal, humoral, and endothelial factors acting on membrane receptors, is relatively slow to develop and, in some cases, can be sustained and tonic. Store-operated Calcium entry (SOCE) does not appear to play a significant role in smooth muscle or cardiac contractility. In contrast, skeletal muscle contractility appears to depend on SOCE, although the mechanism by which SOCE makes such contribution remains contentious. In line with these differences, skeletal muscle displays higher levels of STIM1 and Orai1 expression compared to cardiac muscle (Vig et al. 2008; Kiviluoto et al. 2011). Adult cardiomyocytes and contractile smooth muscle appear to express almost undetectable levels of STIM1 and Orai1 (Berra-Romani et al. 2008; Potier et al. 2009; Bisaillon et al. 2010; Luo et al. 2012). STIM1 and Orai1 (as well as Orai2 and Orai3) protein expression in primary contractile aortic smooth muscle cells is very low or undetectable, when compared with rat basophilic leukemia (RBL) mast cells (Fig. 1; Berra-Romani et al. 2008; Potier et al. 2009; Bisaillon et al. 2010). A recent report showed that SOC activity is robust in neonatal cardiomyocytes while absent from adult cardiomyocytes (Luo et al. 2012), and an earlier report described marginal SOC activity in adult cardiomyocytes (Hulot et al. 2011). These observations are consistent with a negligible role, if any, of SOCE in cardiac or smooth muscle contractility. In fact, STIM1/Orai1-deficient patients and mice do not have any obvious cardiac or vascular phenotypes while displaying skeletal muscle hypotonia (McCarl et al. 2009; Feske, 2010).

Figure 1.

Protein expression of STIM1(A), Orai1 and Orai3 (B) assessed by Western blots in freshly isolated contractile quiescent rat aortic VSM cells (Fresh) and synthetic cultured VSM cells (synth.) that are highly proliferative and migratory and are known to recapitulate VSM features found in disease states. Rat basophilic leukemia (RBL) cells are used as control. VSMC, vascular smooth muscle cells. C, Ca²⁺ imaging experiment showing thapsigargin-induced Ca²⁺ release and entry (after restoration of 2 mm external Ca²⁺) in freshly isolated rat VSM cells (quiescent) and cultured synthetic (highly proliferative and migratory) VSM cells. Data taken from Potier et al. (2009; with permission from The FASEB Journal), and Bisaillon et al. (2010; © 2010 the American Physiological Society).

STIM/Orai channels in smooth muscle remodelling

There is ample evidence to suggest that STIM/Orai signalling could become more relevant during long-term adaptive changes, which occur during the processes of physiological or pathological smooth muscle remodelling, where activation of the SOC pathway via G protein coupled receptor (GPCR)-mediated Ca²⁺ store depletion through IP₃ receptors leads to subsequent activation of gene transcription programmes. It is now clearly established that SOC is not only important for the replenishment of internal Ca²⁺ stores but is also essential for driving transcriptional programmes in the nucleus, with nuclear factor for activated T cells (NFAT) being the archetypical transcription factor requiring CRAC channel function for its nuclear translocation and activity in haematopoietic cells (Gwack et al. 2007). In fact, the major phenotype of STIM1- and Orai1-deficient patients and mice is severe combined immunodeficiency due to lack of NFAT-dependent interleukin-2 production by lymphocytes and subsequent clonal expansion (Feske, 2007). As pointed out earlier, receptor-activated Ca²⁺ entry in mammalian cells also involves alternative store-independent pathways and as argued elsewhere (Trebak, 2011), Orai proteins (as homomers or heteromers) might encode alternative store-independent Ca²⁺ selective channels for the purpose of enhancing the diversity and efficacy of Ca²⁺ signalling in complex mammalian organisms; the evolutionary appearance of Orai3 in mammals, and possibly that of Orai3-encoded ARC channels (and potentially other unidentified channels), would offer a means to functionally meet the demands of this mammalian complexity. Whether different types of GPCR or receptor tyrosine kinase would couple preferentially to ARC and other Orai-dependent store-independent Ca²⁺ entry pathways and whether these store-independent Ca²⁺ entry pathways would activate alternative downstream transcription factors and distinct sets of gene programmes is an interesting possibility that warrants exploration.

Unlike skeletal or cardiac muscles that are terminally differentiated, VSM retains the unique ability to dedifferentiate, perhaps as an evolutionarily acquired trait necessary for tissue repair and wound healing. While skeletal and cardiac muscle will undergo hypertrophy in response to physiological or pathological stimuli, VSM cells will go a step further: they will proliferate and migrate (House et al. 2008). This smooth muscle remodelling process, which is central to various vascular and respiratory diseases, is characterized by the loss of contractile proteins, acquisition of pro-proliferative and pro-migratory proteins and remodelling of plasma membrane receptors and ion channels expression (House et al. 2008; Beech, 2012). As we argued elsewhere, during remodelling VSM cells acquire an ion channel repertoire and a structural phenotype that is more reminiscent of non-excitable cells (House et al. 2008). While contractile quiescent VSM cells barely show any expression of STIM1 and Orai1, STIM1 and Orai1 levels and SOC activity are increased in proliferative VSM cells in comparison to contractile cells isolated from the same vascular bed (Fig. 1; Berra-Romani et al. 2008; Potier et al. 2009; Bisaillon et al. 2010). STIM1 and Orai1 mediate CRAC currents in proliferative VSM and are important for VSM proliferation and migration in vitro (Potier et al. 2009; Bisaillon et al. 2010). In studies conducted in our laboratory, we found no evidence for the expression of the longer version of STIM1, STIM1L (Darbellay et al. 2011) in either contractile or proliferative smooth muscle. It is worth mentioning that STIM1/Orai1 protein levels (Fig. 1) and CRAC currents in primary proliferative cultured VSM or endothelial cells remain 4- to 5-fold lower than those of lymphocytes or RBL mast cells (Abdullaev et al. 2008; Potier et al. 2009). The balloon injury model of rat carotids that induces in vivo VSM proliferation and migration and subsequent vessel occlusion due to neointimal growth was used to assess the potential upregulation of STIM1 and Orai1 in vivo. STIM1 and Orai1 mRNA and protein levels are indeed upregulated in injured carotid arteries compared to sham control arteries (Fig. 2; Bisaillon et al. 2010). Significantly, in vivo STIM1 knockdown caused inhibition of neointimal growth (Fig. 3; Aubart et al. 2009; Guo et al. 2009; Zhang et al. 2011). A marked inhibition of neointimal growth was observed in balloon-injured rat carotids where Orai1 upregulation was prevented in vivo by lentiviruses encoding short hairpin RNA (shRNA; Fig. 3; Zhang et al. 2011). Significantly, when Orai1 upregulation was prevented in vivo by shRNA-encoding lentiviruses, it also prevented upregulation of STIM1, the proliferation marker Proliferating cell nuclear antigen (PCNA) and the δ isoform of the enzyme effector Calcium/Calmodulin Kinase II (CaMKII) (Fig. 3; Zhang et al. 2011); CaMKIIδ is associated with neointima formation in vivo (House & Singer, 2008). VSM cells treated with either STIM1 or Orai1 knockdown showed abrogated NFAT nuclear translocation and activity (Zhang et al. 2011).

Figure 2.

Immunofluorescence staining on carotid artery sections from rats subject to balloon injury, an animal vascular disease model characterized by in vivo smooth muscle proliferation and migration (C and F; 14 days post injury) and control non-injured vessels from sham-operated animals (A, B, D and E) using anti-Orai1 (B and C) and anti-STIM1 (E and F) antibodies followed by a secondary antibody coupled to FITC or secondary antibody alone (A and D). Sections were imaged using confocal microscopy. The brackets indicate the neointima in injured vessels. M, media; NI, neointima; L, lumen. Data taken from Bisaillon et al. (2010; © 2010 the American Physiological Society). (1mm on picture equals 10 micron)

Figure 3. The prevention of STIM1 or Orai1 upregulation in the balloon injury model using shRNA-encoding lentiviruses inhibits neointima formation.

A, control shLuciferase-, shOrai1-, and shSTIM1-encoding lentiviruses efficiently infected carotid vessels as evidenced with expression of green fluorescent protein (GFP driven by a promoter separate from shRNA) in medial and neointimal protein extracts from left carotid arteries that are balloon injured (14 days after injury) and infected with viral particles; right non-infected carotids show no GFP expression. shOrai1 inhibited upregulation of Orai1 protein as well as that of STIM1 and CaMKIIδ2, 14 days post-injury. Similarly, shSTIM1 inhibited upregulation of STIM1 protein as well as that of Orai1 and CaMKIIδ2. shOrai1 and shSTIM1 inhibited VSMC proliferation as evidenced by reduced protein expression of proliferating cell nuclear antigen (PCNA) in medial and neointimal protein extracts from balloon-injured left carotids treated with shOrai1 and shSTIM1 lentiviruses in comparison with injured left carotids treated with shLuciferase lentiviruses. B, statistical analyses on Orai1, STIM1, CaMKIIδ2 and PCNA protein expression data on extracts of medial and neointimal VSMC from balloon-injured left carotids and treated with shLuciferase, shOrai1, or shSTIM1 lentiviruses. Data represent densitometry on protein bands with average ± SEM from 6 rats per condition, determined using Image J and normalized to β actin expression. C, shOrai1 and shSTIM1 lentiviral infection after balloon-injury inhibited neointima formation. H&E staining on left carotid artery sections from either sham-operated (intact) or balloon-injured rats treated with shLuciferase (shLuc), shOrai1 (shO1), or shSTIM1 (shS1) lentiviruses 14 days post injury and lentiviral treatment. Neointimal regions are highlighted in green. Data taken from Zhang et al. (2011; © 2011 American Heart Association, Inc.).

It is worth emphasizing a couple of important points. First, caution should be exercised when extrapolating functions inferred from experiments with STIM1 knockdown or knockout to Orai1. STIM1 regulates the function of an increasing number of channels and transporters; all Orai channels (DeHaven et al. 2007), ARC channels (Mignen et al. 2007), Cav1.2 L-type Ca²⁺ channels (Park et al. 2010; Wang et al. 2010), transient receptor potential canonical (TRPC) channels (Yuan et al. 2007) and Plasma membrane Calcium ATPase (PMCA) (Ritchie et al. 2012) are all regulated by STIM1. Similarly, Orai1 might serve functions that are independent of its Ca²⁺ conducting properties or might contribute subunits to channels that are regulated in a STIM1-independent manner. In fact, Orai2 and Orai3 are also upregulated during smooth muscle remodelling (Berra-Romani et al. 2008) (also see Fig. 1 for Orai3), along with TRPC1 and TRPC6 channel isoforms (Yu et al. 2004; Kumar et al. 2006). TRPC channel upregulation was shown to contribute to pathological cardiac remodelling in mice (Eder & Molkentin, 2011). Second, not all functions mediated by STIM1 and Orai1 are necessarily through SOC activity. As pointed out earlier, STIM and Orai isoforms are likely to contribute to additional store-independent channels such as ARC channels. These store-independent channels would create a diverse array of Ca²⁺ signalling microdomains generated through alternative activation mechanisms to fine tune signalling specificity in response to a variety of growth factors and agonists.

Conclusions

While perhaps its original intent was to refill internal Ca²⁺ stores, the role of the SOC pathway in mammals has evolved beyond Ca²⁺ store refilling to active spatiotemporal Ca²⁺ signalling, as it is probable that any Ca²⁺-conducting channel can fulfil the function of store repletion in the absence of SOC channels. The role of discrete Ca²⁺ entry through receptor-activated SOC channels in driving transcription in haematopoietic cells and other non-excitable cells is well established (Parekh, 2011). As discussed above, emerging evidence suggests an important role for Ca²⁺ entry through SOC channels in coupling to transcriptional programmes in all types of muscle. The long-term and adaptive smooth muscle responses to increased physiological demands or enhanced pathophysiological stimuli involves remodelling of the smooth muscle ion channels repertoire, including upregulation of STIM and Orai proteins and activation of downstream transcriptional programmes. The appearance of Orai2 in vertebrates and that of Orai3 later in mammals (Cai, 2007; Shuttleworth, 2012) would contribute to alternative store-independent Ca²⁺ channels that would match the complexity of higher organisms. These additional channels would serve to enhance the diversity of Ca²⁺ signalling microdomains for the purpose of selective coupling to specific transcription factors in response to stimulation of diverse GPCRs and receptor tyrosine kinases. These additional Orai-containing channels would have selective coupling to specific PM receptors, discrete PM localization, alternative activation mechanisms and diverse Orai subunit requirements. ARC channels activated by AA, contributed by heteromultimers of Orai1 and Orai3, and regulated by STIM1 in a store-independent manner (Shuttleworth, 2009) are one example of such Ca²⁺-selective conductances; as we look more actively into native Ca²⁺ signalling pathways in muscle and non-muscle cell systems, and as we generate mice lacking either Orai2 or Orai3 specifically in different tissues, other examples might just emerge.

Glossary

AA

arachidonic acid

ARC channel

arachidonate-regulated Ca²⁺ channel

CRAC channel

Ca²⁺ release-activated Ca²⁺ channel

DAG

diacylglycerol

ER

endoplasmic reticulum

GPCR

G protein coupled receptor

IP₃

inositol 1,4,5-trisphosphate

NFAT

nuclear factor for activated T cells

PLC

phospholipase C

PM

plasma membrane

shRNA

short hairpin RNA

SOC channel

store-operated Ca²⁺ channel

STIM1

stromal interacting molecule 1

VSM

vascular smooth muscle