STIM1 phosphorylation at Y³¹⁶ modulates its interaction with SARAF and the activation of SOCE and I_CRAC

Store-operated Ca²⁺ entry (SOCE) participates in the intracellular Ca²⁺ homeostasis of both excitable and non-excitable cells. Physiological agonists evoke the depletion of intracellular Ca²⁺ reservoirs leading to an interaction between the endoplasmic reticulum (ER) Ca²⁺ sensor, STIM1 and Orai1, the Ca²⁺-permeable channel located at ER–plasma membrane (PM) junctions, thus allowing Ca²⁺ entry. STIM1 contains an ER luminal N-terminal portion that includes a canonical EF-hand (amino acids 63–98) responsible for the Ca²⁺ binding, and further, a sterile alpha motif (SAM; amino acids 131–200) that contributes to STIM1 multimerization. After this motif, STIM1 contains its single transmembrane domain, and then, the cytosolic STIM1 region, which exhibits three coiled-coil domains (CC1, CC2 and CC3) the latter two of which contain the STIM1–Orai1-activating region (SOAR). Additionally, an auto-inhibitory region, including an acidic domain (E³¹⁸–E³²²) located in the CC1 region and a polybasic domain (K³⁸²–K³⁸⁷) in the CC2 can be found, which occlude the SOAR maintaining STIM1 in a quiescent state (Fahrner et al., 2014; Korzeniowski et al., 2010; Yuan et al., 2009; Zhou et al., 2013). Finally, the C-terminal domain of STIM1 exhibits the STIM1 C-terminal inhibitory domain (CTID; amino acids 448–530), which is responsible for the interaction with the Ca²⁺ entry regulatory protein SOCE-associated regulatory factor (SARAF), which regulates the slow Ca²⁺-dependent inactivation of Orai1 (SCDI) (Derler et al., 2016; Haniu et al., 2006; Perni et al., 2015). Structural and molecular analyses have revealed that in order to regulate the STIM1–Orai1 interaction, SARAF interacts with the CTID downstream of the SOAR region (Jha et al., 2013). CTID consists of two lobes (STIM1 448–490 and STIM1 490–530). The STIM1 (490–530) lobe directs SARAF to the SOAR; meanwhile the STIM1 (448–490) lobe wraps around the SOAR to restrict access of SARAF to the SOAR (Jha et al., 2013).

Recent studies have reported that STIM1 remains in a tightly packed conformation in resting cells (Ma et al., 2015; Muik et al., 2011). According to this hypothesis, after intracellular Ca²⁺ store depletion, STIM1 undergoes a conformational change resulting in the full extended and active configuration, thus, accommodating the C-terminal region of Orai1. The CC1 inhibitory domain of STIM1 has been reported to maintain STIM1 in a quiescent state by hiding the SOAR domain in non-stimulated cells (Haniu et al., 2006; Jha et al., 2013; Palty et al., 2012). In addition, recent studies have demonstrated that Ca²⁺-mediated dissociation of the STIM1 EF-hand domain during store depletion triggers the rearrangement of STIM1 EF-hand, SAM and transmembrane domains, which allows the reorganization of the cytosolic portion of STIM1, and subsequently, releases SOAR from the CC1 domain in order to interact with and activate Orai1 (Ma et al., 2015).

STIM1 phosphorylation has been reported to play a relevant role during its function. Phosphorylation of STIM1 at S⁴⁸⁶ and S⁶⁶⁸ has been shown to lead to SOCE suppression during mitosis (Smyth et al., 2009). Furthermore, STIM1 has been reported to be a target of ERK1 and ERK2 (ERK1/2, also known as MAPK3 and MAPK1, respectively), which phosphorylate STIM1 at S⁵⁷⁵, S⁶⁰⁸ and S⁶²¹ upon Ca²⁺ store depletion; this leads to the dissociation from the microtubule plus-end-binding protein EB1 and the activation of Orai1 and SOCE (Pozo-Guisado et al., 2010, 2013). Phosphorylation of STIM1 at S⁵⁷⁵ by ERK1/2 is essential for myoblast differentiation (Lee et al., 2012) and is inhibited by 17β-estradiol in airway epithelia, thus reducing SOCE and increasing the risk of lung diseases (Sheridan et al., 2013). Moreover, EGF-stimulated phosphorylation of STIM1 at S⁵⁷⁵, S⁶⁰⁸ and S⁶²¹ is required for human endometrial adenocarcinoma cell migration (Casas-Rua et al., 2015). In addition, phosphorylation of STIM1 at T³⁸⁹ by protein kinase A (PKA) induces selective activation of acid-regulated Ca²⁺-selective (ARC) channels while reducing the ability of STIM1 to activate Ca²⁺ release-activated Ca²⁺ current (CRAC) channels (Thompson and Shuttleworth, 2015). The role of tyrosine phosphorylation in STIM1 function has been scarcely investigated. Tyrosine kinase activation by intracellular Ca²⁺ store depletion, and subsequent activation of Ca²⁺ entry, has previously been reported (Mills et al., 2015; Sage et al., 1992; Zuo et al., 2011). Interestingly, Ca²⁺ store depletion by thapsigargin (TG) has been reported to evoke protein tyrosine phosphorylation in cells loaded with the Ca²⁺ chelator BAPTA (Sargeant et al., 1994), indicating that Ca²⁺ store depletion itself, and not the subsequent rise in the cytosolic free Ca²⁺ concentration ([Ca²⁺]_c), is sufficient for tyrosine kinase activation. We have previously reported that, in BAPTA-loaded cells, STIM1 is phosphorylated at tyrosine residues upon Ca²⁺ store depletion, an event that is relevant for STIM1–Orai1 interaction (López et al., 2012). Btk activation and subsequent regulation of SOCE in human platelets and murine B cells have been described (Fluckiger et al., 1998; Redondo et al., 2005), and we have reported interaction between Btk and STIM1 as well as Btk-dependent tyrosine phosphorylation of STIM1 in human platelets (López et al., 2012). In addition, a recent study has revealed that phosphorylation of STIM1 at Y³⁶¹ is important for the recruitment of Orai1 into STIM1 puncta and the activation of SOCE (Yazbeck et al., 2017). Furthermore, the Y³¹⁶ residue has been found to play a relevant role in maintaining STIM1 in a quiescent state in resting cells (Yu et al., 2013). Here, we show that phosphorylation of STIM1 at Y³¹⁶ plays an essential role in the STIM1 and Orai1 colocalization at ER–PM junctions, as well as for the activation of I_CRAC and SOCE. We also describe for the first time that phosphorylation of STIM1 at Y³¹⁶ modulates the interaction between STIM1 and its regulator SARAF, hence, regulating SCDI. Taken together, these findings might shed new light on the molecular mechanism of SOCE activation and modulation.

Computational analysis using the NetPhos 2.0 Server revealed six alignment sequences that are susceptible to phosphorylation (http://www.cbs.dtu.dk/services/NetPhos/). Among them, the score of three sequences (leading to phosphorylation of Y²³¹, Y³¹⁶ and Y³⁶¹) was sufficiently high for them to be considered as targets of tyrosine kinases; however, only two of them are present in the cytosolic region of STIM1, Y³¹⁶ and Y³⁶¹, which are located in the CC1 and CC2 (SOAR) domains, respectively (Table 1). Y³⁶¹ phosphorylation was recently described as a relevant posttranslational modification required for Orai1–STIM1 interaction during SOCE in human pulmonary arterial endothelial (HPAE) cells (Yazbeck et al., 2017). In these cells, STIM1 phosphorylation at Y³¹⁶ was found to have a minor effect on SOCE. However, these findings disagree with a previous study where the authors proposed a relevant role for STIM1 Y³¹⁶ phosphorylation in the maintenance of STIM1 in a closed conformation in quiescent cells (Yu et al., 2013). Taking into account this controversy, we explored the role of Ca²⁺ store depletion in the phosphorylation of both residues (Y³¹⁶ and Y³⁶¹) in HEK293 cells expressing either wild-type (STIM1 WT) or STIM1 mutants where we replaced the target tyrosine residues with phenylalanine, another aromatic amino acid that avoids the alteration of the structure of the STIM1 auto-inhibitory domain (Yu et al., 2013). As depicted in the Fig. 1A, treatment of HEK293 cells expressing STIM1 WT with TG evoked an increase in STIM1 tyrosine phosphorylation, as previously described in platelets (López et al., 2012). Overexpression of STIM1 Y316F impaired TG-evoked STIM1 phosphorylation (Fig. 1A,C; n=4). We have also found a significant reduction in TG-induced STIM1 tyrosine phosphorylation in HEK293 expressing the STIM1 Y361F mutant, as previously described in HPAE cells (Yazbeck et al., 2017). STIM1 tyrosine phosphorylation induced by TG was reduced to 39.3±9.0% (mean±s.e.m.) in HEK293 cells expressing STIM1 Y316F compared to cells expressing STIM1 WT, while less inhibition was observed in cells transfected with Y361F (Fig. 1A,C; n=3–5; P<0.05).

To confirm these results, we analyzed STIM1 phosphorylation at Y³¹⁶ and Y³⁶¹ in the STIM1-deficient NG115-401L cell line (Albarran et al., 2016; Zhang et al., 2015; Zhang and Thomas, 2016). Expression of the STIM1 Y316F mutant in NG115-401L cells attenuated the STIM1 phosphotyrosine content at resting conditions and upon stimulation with TG as compared with those cells expressing STIM1 WT (Fig. 1B,C; n=5). In addition, our results indicate that mutation of the tyrosine residue to a phenylalanine residue is an efficient mechanism to analyze STIM1 phosphorylation during SOCE, since this change might avoid the constitutive activation of STIM1 observed when Y³¹⁶ is replaced by an alanine or aspartate residue as previously demonstrated (Yu et al., 2013). Similarly, a reduction in the STIM1 phosphotyrosine level was observed in cells expressing YFP–STIM1-Y361F.

A previous study has reported that phosphorylation of STIM1 at Y³¹⁶ plays a relevant role in maintaining STIM1 in an inactive state in quiescent cells (Yu et al., 2013). Hence, we have further explored the role of the Y³¹⁶ and Y³⁶¹ residues in the STIM1 function during the activation of SOCE. Expression of STIM1 WT restored TG-evoked Ca²⁺ entry in NG115-401L cells as previously reported (Albarran et al., 2016). As shown in Fig. 2A, we observed an increase in TG-evoked SOCE of 91±24% (mean±s.e.m.) in the area under the curves for cells expressing YFP–STIM1 compared to cells expressing the empty vector (mock cells) (Fig. 2A; P<0.01, n=6). The analysis of the rate of increase in [Ca²⁺]_c evoked by TG, revealed a K₁ of 3.6×10⁻⁵±1.5×10⁻⁶ in mock cells and of 7.6×10⁻⁵±1.9×10⁻⁶ in YFP–STIM-WT cells. YFP–STIM1-WT expression also enhanced SOCE by 47±26% in MEG-01 cells (see Fig. S1; n=4–6; P<0.001). Interestingly, expression of the STIM1 Y316F mutant significantly reduced SOCE both in NG115-401L and MEG-01 cells (Fig. 2A and Fig. S1, respectively; P<0.001; n=8). NG115-401L cells expressing YFP–STIM1-Y316F presented a reduction in the area under the curves of 79% with respect to cells expressing the YFP–STIM1-WT, resulting in similar values to those found in mock cells, and subsequently the observed K₁ was also similar to that found in mock cells (3.8×10⁻⁶±1.1×10⁻⁶). As previously reported (Yazbeck et al., 2017), we found that expression of STIM1 Y361F attenuated TG-evoked SOCE in NG115-401L cells (Fig. 2A), which confirms that phosphorylation of STIM1 at Y³⁶¹ plays an important role in SOCE.

According to the inhibitory role evidenced by the Y316F or Y361F mutations of STIM1, we consider that Y316F might be more relevant for STIM1 function in the cell models investigated, and hence should be studied further. To this end, we further evaluated whether Y³¹⁶ phosphorylation is relevant for the activation of I_CRAC by employing the patch-clamp technique in the whole-cell configuration. HEK293 cells were co-transfected with YFP–STIM1-WT or YFP–STIM1-Y316F, and CFP–Orai1, and we measured the inward currents upon passive store-depletion evoked by employing 20 mM EGTA in the pipette solution (following whole-cell formation) in response to repetitive voltage-ramps from −90 mV to +90 mV applied from a holding potential of 0 mV. Current values were taken from each trace at −74 mV to generate the time-course depicted in Fig. 2B. Store-operated activation of HEK293 cells transfected with YFP–STIM1-WT and CFP–Orai1 reached a maximal inward current at ∼50 s (Fig. 2B, gray trace). By contrast, HEK293 cells containing YFP–STIM1-Y316F displayed a ∼50 s delayed activation to maximal current levels and significantly reduced current density (Fig. 2B, black trace).

Orai1 opening is driven by its interaction with STIM1 at ER–PM junctions (Lee et al., 2010). Because phosphorylation of STIM1 at Y³¹⁶ plays an important role in SOCE and I_CRAC, we analyzed whether this event is relevant for the interaction between STIM1 and Orai1 by looking for colocalization between STIM1 and Orai1 in HEK293 cells co-transfected with YFP–STIM1-WT or STIM1 mutants, and CFP–Orai1. As shown in Fig. 3A, a significant increase in the colocalization between STIM1 and Orai1 is observed upon stimulation with TG (see Fig. 3A.1). Interestingly, expression of the STIM1 Y316F mutant significantly attenuated TG-induced STIM1–Orai1 colocalization (Fig. 3A.2 and bar graph). We found that TG was able to induce colocalization of STIM1 and the Orai1 R91W mutant, which has been reported to be unable to induce Ca²⁺ entry (Liao et al., 2008), thus suggesting that the effect observed with the Y316F mutant reveals a specific role of Y³¹⁶ phosphorylation and is not due to other non-specific effects (Fig. 3A.3).

Co-immunoprecipitation was performed in NG115-401L cells expressing GFP–Orai1 and YFP–STIM1 to prevent changes in the Orai1:STIM1 stoichiometry (Hoda et al., 2006). As shown in the left two lanes of Fig. 3B, Orai1 was detected in STIM1 immunoprecipitates of resting cells and this interaction was enhanced after store depletion. In contrast, the expression of the STIM1 Y316F mutant significantly reduced the interaction by 70% between both proteins upon cell stimulation with TG (Fig. 3B, right two lanes; P<0.05; n=4). Taken together, these findings suggest that phosphorylation of STIM1 at Y³¹⁶ is required for the association with Orai1, as previously suggested in human platelets (López et al., 2012). This mechanism might underlie the role of STIM1 Y³¹⁶ phosphorylation in the activation of I_CRAC and SOCE.

It has been reported that Ca²⁺ store depletion leads to a conformational change of STIM1 from a resting close state to a fully extended conformation, required for the interaction with and activation of Orai1, as demonstrated by using double-labeled YFP–STIM1–233-474–CFP (YFP–OASF–CFP; OASF, Orai1-activating small fragment) as a sensor of the intramolecular transition that leads to an extended conformation when binding to Orai1 (Muik et al., 2011). We explored whether STIM1 phosphorylation at Y³¹⁶ is relevant for the conformational change in HEK293 cells co-expressing Orai1 and either YFP–OASF–CFP or the YFP–OASF-Y316F–CFP mutant, following a previously reported procedure (Muik et al., 2011). As depicted in Fig. 4A, FRET analysis revealed no changes in the YFP fluorescence for either in resting cells or cells where Ca²⁺ store depletion was induced by treatment with 2 µM TG. Considering that OASF may directly interact with Orai1 in the absence of stimuli, we repeated the experiments using a longer fragment of STIM1 known as OASF extended (amino acids 233–535), which has been found to lack the ability to induce constitutive CRAC currents in the absence of stimuli (Muik et al., 2009). Similar to the observations when expressing OASF, no FRET changes were detected upon stimulation with TG (Fig. 4B). Furthermore, no differences in FRET were detected either in the cytosol or near the plasma membrane, where Orai1 is located (Fig. 4B, bar graphs). These findings indicate that Y³¹⁶ phosphorylation in itself does not induce the intramolecular transition into an extended conformation.

We further evaluated whether STIM1 phosphorylation at Y³¹⁶ is important for the interaction with STIM1 regulatory proteins. SARAF has emerged as a physiological inhibitor of STIM1 (Palty et al., 2012) and interacts with the CTID region of STIM1. It has been reported that the four conserved glutamate residues (E³¹⁸–E³²²) of the auto-inhibitory domain of CC1 interact with the STIM1 (448–490) lobe of the CTID (Jha et al., 2013). Interestingly, in cells with a low expression of STIM1, like NG115-401L cells, SARAF interacts with and activates Orai1 (Albarran et al., 2016; Palty et al., 2012). Considering that Y³¹⁶ is very close to the four glutamate residues in CC1, we explored whether phosphorylation of STIM1 at Y³¹⁶ is relevant for SARAF–STIM1 interaction by assessing co-immunoprecipitation between SARAF and STIM1 in NG115-401L cells expressing STIM1 WT or the STIM1 Y316F mutant. As shown in Fig. 5, and as previously reported (Albarran et al., 2016), treatment with TG for 1 min did not significantly modify the interaction between SARAF and STIM1-WT as compared to that seen in resting cells. Interestingly, impairment of STIM1 phosphorylation at Y³¹⁶ by expression of the STIM1 Y316F mutant significantly increased the TG-induced SARAF–STIM1 interaction by 38% (see Fig. 5A lane 4 versus lane 2; n=4; P<0.05). These results strongly indicate that phosphorylation of STIM1 at Y³¹⁶ is required for the dissociation from SARAF, and thus, impairment of STIM1 phosphorylation at Y³¹⁶ might lead to an enhanced association with SARAF, which, in turn, would restrict the interaction of STIM1 with Orai1 attenuating SOCE and I_CRAC.

To further explore this possibility, we analyzed TG-induced Ca²⁺ influx in NG115-401L cells transfected either with STIM1 WT or the STIM1 Y316F mutant, as well as plasmids expressing shRNA targeting SARAF (shSARAF) or scramble shRNA. As depicted in Fig. 6A, and as expected, in NG115-401L cells overexpressing STIM1 WT reconstitutes SOCE. The increase in SOCE is revealed by analyzing the area under the curves (see histogram in Fig. 6), and the K₁ of the Ca²⁺ traces (3.4×10⁻⁵±1.3×10⁻⁶ versus 6.8×10⁻⁵±2.5×10⁻⁶ in cells transfected with the empty vector or with the STIM1 WT). Co-transfection of STIM1 WT together with shSARAF resulted in a further increase in SOCE (K₁, 8.8×10⁻⁵±1.7×10⁻⁶), which reveals the inhibitory role of SARAF previously reported (Albarran et al., 2016, Palty et al., 2012).

Expression of the STIM1 Y316F mutant in NG115-401L cells marginally reconstitutes SOCE in these cells, as shown above (Fig. 6B and Fig. 2). Interestingly, SARAF knockdown using shRNA technology in cells expressing the STIM1 Y316F mutant leads to a significant increase in SOCE, thus indicating that SARAF is involved in the inhibition of SOCE observed in cells expressing the STIM1 Y316F mutant (Fig. 6B). SOCE in cells transfected with shSARAF was found to be significantly greater in cells expressing the STIM1 Y316F mutant than in cells expressing STIM1 WT (Fig. 6B and bar graph), which indicates that, in addition to the role of STIM1 phosphorylation at Y³¹⁶ in the regulation of the STIM1–SARAF interaction, this posttranslational modification might regulate other SARAF-independent processes. Finally, we explored the effect of the STIM1 Y316F mutation in SCDI, which has been reported to be regulated by SARAF. As depicted in Fig. 6C,D, we performed patch-clamp experiments in HEK293 cells according to experimental conditions previously described (Jha et al., 2013). A greater SCDI was observed in cells transfected with the STIM1 Y316F and Orai1 as compared with cells expressing STIM1 WT and Orai1 (Fig. 6C,D, black traces), which is consistent with the greater SARAF–STIM1-Y316F interaction reported above (Fig. 5). These data are also consistent with the inhibition in SOCE and I_CRAC found in cells transfected with STIM1 Y316F (Fig. 2). As shown in Fig. 6C,D, SARAF knockdown attenuated SCDI in cells transfected with STIM1 WT and the STIM1 Y316F mutant, leading to a similar inactivation in both experimental conditions, and thus indicating that the greater SCDI induced by transfection of the STIM1 mutant is dependent on SARAF.

There is a growing body of evidence supporting a role for STIM1 phosphorylation in the regulation of SOCE. Although most studies have been focused on phosphorylation of STIM1 at serine or threonine residues, a recent study has revealed that Pyk-2-dependent phosphorylation of STIM1 at Y³⁶¹ plays an essential role in the recruitment of Orai1 into STIM1 puncta during the activation of SOCE in endothelial cells, which has been associated with an increase in endothelial barrier permeability (Yazbeck et al., 2017). Although in the aforementioned study the authors presented evidenced against Y³¹⁶ as the target residue of kinases in favor of Y³⁶¹; Yu and coworkers envisaged that the Y³¹⁶ residue plays a relevant role in maintaining STIM1 in a quiescent state in resting cells, as the expression of the STIM1 Y316A mutant in itself resulted in STIM1 oligomerization. The auto-inhibitory role of STIM Y³¹⁶ has been attributed to either the intramolecular interaction of the Y³¹⁶ with the SOAR domain, through hydrogen and/or hydrophobic bonds, or an intermolecular interaction mediated through repulsive forces, which would maintain STIM1 in a quiescent state (Yu et al., 2013). Here, we provide evidence for a role of phosphorylation of STIM1 at Y³¹⁶ in the STIM1 and Orai1 colocalization at ER–PM junctions as well as for the activation of I_CRAC and SOCE. Furthermore, we described for the first time that phosphorylation of STIM1 at Y³¹⁶ modulates the interaction between STIM1 and its regulator SARAF. SARAF has been reported to interact with STIM1 in a dynamic manner. According to this model SARAF is associated with STIM1 in resting cells but, upon Ca²⁺ store depletion, SARAF dissociates from STIM1, allowing STIM1 oligomerization and interaction with Orai1 channels in the PM, followed by re-interaction with STIM1 to activate slow Ca²⁺-dependent inactivation of Orai1 (Albarran et al., 2016; Jha et al., 2013; Palty et al., 2012). The role of the phosphorylation of STIM1 at Y³¹⁶ in the interaction between STIM1 and SARAF might provide a molecular basis for the findings reported by Yu and coworkers (Yu et al., 2013). Discrepancies in the tyrosine residue that is or could be phosphorylated in STIM1 might be indicative of the presence of different STIM1 regulatory pathways that would involve different tyrosine kinases, or perhaps, could indicate that these kinases are also sequentially activated. In fact, a recent study reported that SOCE activation by STIM1 is required for Pyk-2 activation in human lung microvascular endothelial cells, which then resulted in Pyk-2-dependent Src family protein activation (Soni et al., 2017). In platelets, we have reported that pp60^Src (Src itself) interacts with and phosphorylates STIM1, which requires changes in the [Ca²⁺]_c, thus regulating SOCE (López et al., 2012). In addition, STIM1 phosphorylation by Btk in platelets does not require changes in [Ca²⁺]_c. Hence, Pyk-2-mediated STIM1 phosphorylation might contribute to the regulation of STIM1 function due to the phosphorylation of Y³⁶¹, which might require p60^Src activation, while Y³¹⁶ would be targeted by Btk. Nonetheless, future investigation will be required in order to confirm this hypothesis.

In addition, previous proximity ligation assay and co-immunoprecipitation studies, have shown that there is detectable interaction between SARAF and STIM1 under resting conditions that is significantly reduced after Ca²⁺ store depletion using TG, reaching a minimum 30 s after stimulation and then this interaction increased, and exceeding the resting level between ∼60 and ∼120 s after stimulation (Albarran et al., 2016). Here, we show that 60 s after stimulation with TG the interaction between SARAF and STIM1 is significantly greater in cells expressing the STIM1 Y316F mutant, which indicates that phosphorylation of STIM1 at Y³¹⁶ plays a negative role in the association of STIM1 with its negative regulator SARAF. Based on previous findings (Albarran et al., 2016; Jha et al., 2013; Palty et al., 2012) and our observations (see Fig. 6), the enhanced association of SARAF with STIM1 might explain the attenuation in SOCE and I_CRAC, and the colocalization of STIM1 with Orai1 observed in cells expressing the STIM1 Y316F mutant as compared to what is seen in cells expressing STIM1 WT.

The reduction in I_CRAC and SOCE observed in cells with impaired phosphorylation at Y³¹⁶ cannot be attributed to inhibition of STIM1 puncta formation, as evidenced by confocal microscopy (see Fig. 3) and previously reported (Yu et al., 2013). This effect is more likely to occur through attenuation of the association between STIM1 and Orai1 (as shown in Fig. 3) that induced by the association of STIM1 with SARAF. Hence, here, we present for the first time, evidence that phosphorylation at Y³¹⁶ might finetune the formation of the CRAC signaling complex, which would contribute to dissociation of SARAF from STIM1 and regulation of SCDI, as depicted in Fig. 7. In addition, our results indicate that mutation of the tyrosine residue into phenylalanine is an efficient mechanism to analyze STIM1 phosphorylation during SOCE, since this change might avoid the constitutive activation of STIM1 observed when Y³¹⁶ is replaced by alanine or aspartate residues, as previously shown (Yu et al., 2013).

A role for Ca²⁺ store depletion-dependent tyrosine phosphorylation in the activation of SOCE has long been proposed in different cell types (López et al., 2012; Redondo et al., 2005; Rosado et al., 2000; Sargeant et al., 1993; Yu et al., 2013). The identification of two key residues (Y³¹⁶ and Y³⁶¹) that act to modulate STIM1 function, and subsequently, alter activation of I_CRAC and SOCE, provides compelling evidence for a relevant role of tyrosine phosphorylation in the regulation of Ca²⁺ influx, which might represent a novel target for the study of the molecular basis of disease.

Lipofectamine^® was from Thermo Fisher Scientific, transfectin was from Bio-Rad, kit C for AMAXA^® was provided by Lonza. Fura 2/AM, mouse anti-phosphotyrosine (4G10) antibody (cat. no. 05-321), rabbit polyclonal anti-SARAF (TMEM66) antibody (epitope amino acids 33–62 of the N-terminal region of human TMEM66; cat. no. PA5-24237) and mouse monoclonal anti-STIM1 antibody (clone 44 GOK⁻¹, epitope: amino acids 25–139 of human STIM1; cat. no. 610954) were supplied by Merck Millipore. Horseradish peroxidase (HRP)-conjugated secondary antibodies were ordered from Jackson ImmunoResearch laboratories. TG and anti-Orai1 (C-terminal; cat. no. SAB4200273) antibody as well as other reagents used, of analytical grade, were provided by Sigma-Aldrich.

MEG-01, HEK293 and NG115-401L (NG115) cells were purchased from the ATCC, and cultured following the manufacturer's recommendations. NG115-401L and HEK293 cells were transfected using Lipofectamine, while MEG-01 cell transfection was performed using kit C for the AMAXA nucleofector device. Human STIM1 (accession number NM_003156) N-terminally tagged with EYFP was kindly provided by the Meyer's laboratory, Stanford University, Stanford, CA. For double-tagged STIM1 constructs, CFP was cloned into pEYFP-C2 via the SacII and XbaI sites, and the OASF STIM1 fragment (233–474) and OASF extended STIM1 fragment (233–535) were introduced via the EcoRI and SacII sites. Point mutations (Y316F and Y361F) were generated by using the QuikChange XL site-directed mutagenesis kit (Stratagene). The integrity of all resulting clones was confirmed by sequence analysis (Eurofins). Following transfection protocols, the percentage of transfected cells was confirmed by visualizing the YFP emission under a fluorescence microscopy. In addition, we used shSARAF generated by our research group (Albarran et al., 2016) in order to reduce SARAF expression in NG115-401L cells as previously described.

Cells were suspended in HEPES-buffered saline (HBS; containing in mM: 145 NaCl, 10 HEPES pH 7.40, 10 D-glucose, 5 KCl, 1 Mg₂SO₄) and supplemented with 50 µM CaCl₂. Cells were stimulated with 200 nM TG in a Ca²⁺-free HBS (75 µM of EGTA added) and then lysed using NP40 buffer (containing in mM: 20 Tris-HCl pH 8.0, 1.37 NaCl, 2 EDTA, 10% glycerol and 1% nodidet P-40) and supplemented with a protease cocktail (Roche) and Na₃VO₄. Cell lysates were immunoprecipitated with anti-STIM1 (2 μM) or anti-SARAF (TMEM66) antibody (2 μM). Immunoprecipitated proteins were separated by 10% SDS-PAGE and subsequently transferred onto a nitrocellulose membrane. Membranes were probed with anti-phosphotyrosine (4G10) antibody for 1 h at room temperature, anti-Orai1 antibody overnight 4°C and anti-SARAF antibody for 1 h at room temperature (these antibodies were diluted in Tris-buffered saline with 0.1% Tween 20 at 1:1000, 1:250 and 1:1000, respectively). Upon membranes being exposed to enhanced chemiluminescence reagent for 4 min, blot densitometry was estimated using the C-digit chemiluminescent western blot scanner (Licor). Data were normalized to the amount of protein recovered by the antibody used for the immunoprecipitation.

Scramble- and STIM1-transfected cells were loaded with fura 2 by incubation with 2 µM fura 2-AM for 30 min at room temperature. Cells were then suspended in HBS containing 50 µM CaCl₂. Coverslips were placed in a perfusion chamber of an inverted microscope and excited alternatively at 340 and 380 nm. Fura 2 fluorescence was detected at 515 nm using a cooled digital CCD camera (Hisca CCD C-6790, Hamamatsu, Japan) and recorded using Aquacosmos 2.5 software (Hamamatsu Photonics, Hamamatsu, Japan). Resulting traces were normalized to the fluorescence emitted by the cells under resting conditions (F_n/F₀). SOCE was measured by determining the integral of the rise in [Ca²⁺]_c for 1.5–2 min after the addition of 0.3–1 mM CaCl₂ to TG-treated cells (0.2–1 µM), taking a sample every second, and is expressed as nM s (López et al., 2005). Additionally, we determined the K₁ the constant of the two-phase exponential decay equation [y=A(1−e^−K1t)e^−K2t], which represents the exponential increase constant of the Ca²⁺ rise during SOCE activation (López et al., 2005).

HEK293 cells were co-transfected with CFP–Orai1 and either YFP–STIM1-WT or YFP–STIM1-Y316F plasmids. An additional set of experiments were performed using a CFP–Orai1(R91W) overexpression plasmid, which was considered as an internal experimental control, evidencing that puntual modification of the proteins do not necessarily involve changes in their interaction pattern. Cells were observed under the confocal microscope in either the resting condition or upon stimulation with TG (1 µM) for 1 min. Fluorescence of both YFP and CFP were recorded in order to analyze protein localization as evidenced by merge images.

Electrophysiological experiments were performed at 20–24°C after 12–48 h of cell transfection with transfectin (Bio-Rad) and 1 µg of the Orai1 and STIM1-WT or STIM1-Y316F constructs. We used the patch-clamp technique in the whole-cell recording configuration for current measurements, with voltage ramps applied every 5 s from a holding potential of 0 mV, covering a range of −90 to +90 mV over 1 s. The internal pipette solution for passive store depletion contained (in mM): 3.5 MgCl₂, 145 cesium methanesulfonate, 8 NaCl, 10 HEPES pH 7.4, 20 EGTA, pH 7.2. Extracellular solution consisted of (in mM): 145 NaCl, 5 CsCl, 1 MgCl₂, 10 HEPES, 10 glucose, 10 CaCl₂. To measure SCDI, the pipette solution contained (in mM, adjusted to pH 7.2): 3.5 MgCl₂, 145 cesium methanesulfonate, 8 NaCl, 10 HEPES, 1,2 EGTA. After establishing the whole-cell configuration, the cells were kept in a Ca²⁺-free solution [(in mM): 140 NaCl, 5 KCl, 1 MgCl, 10 HEPES and 10 glucose] for 5 min to allow store depletion before exposing the cells to the extracellular solution containing 10 mM Ca²⁺ (Jha et al., 2013). All currents were leak corrected by subtracting either the value from the initial voltage ramps obtained shortly after break-in with no visible current activation or the remaining currents after 10 mM La³⁺ application at the end of the experiment, with both yielding identical results.

Confocal Förster resonance energy transfer microscopy (FRET) was performed on HEK293 cells, as previously described (Derler et al., 2006). In brief, for recording fluorescence images we used a QLC100 Real-Time Confocal System (VisiTech Int.) connected to two Photometrics CoolSNAPHQ monochrome cameras (Roper Scientific) and a dual port adapter (dichroic: 505lp; cyan emission filter: 485-30; yellow emission filter: 535-50; Chroma Technology Corp.). This system was implemented with an Axiovert 200M microscope (Zeiss, Germany) in conjunction with two diode lasers (445 nm and 515 nm) (Visitron Systems). Visiview 2.1.1 software (Visitron Systems) was used for image acquisition and controlling the confocal system. As required, we performed an image correction, to modulate for cross-talk and cross-excitation, before to the calculation; appropriate cross-talk calibration factors were determined for each construct on the raw data from the FRET experiments. After threshold determination and background subtraction, the corrected FRET (E_app) was calculated on a pixel-to-pixel basis with custom-made software (Zal and Gascoigne, 2004) integrated in MatLab 7.0.4 according to the method published previously (Singh et al., 2006), with a microscope-specific constant G value of 2.0. All experiments were performed at room temperature.

Analysis of statistical significance was performed with a unpaired Student's t-test or one-way analysis of variance (ANOVA) combined with the Dunnett's test. The Pearson correlation coefficient (R-value) was used to measure the strength of the linear association/colocalization between STIM1 and Orai1 variants. Overall, only values with P<0.05 were accepted as significant throughout the present research.

We are grateful to Dr A. Quesada and Prof. X. Pingyong (University of Extremadura and Institute of Biophysic and Chinese Academy of Sciences, respectively) for their technical support. Furthermore, we like also thank to Dr T. Meyer for providing the YFP-STIM1 overexpression plasmid.