Functional differences in agonist‐induced plasma membrane expression of Orai1α and Orai1β

Orai1 is the pore‐forming subunit of the store‐operated Ca2+ release‐activated Ca2+ (CRAC) channels involved in a variety of cellular functions. Two Orai1 variants have been identified, the long form, Orai1α, containing 301 amino acids, and the short form, Orai1β, which arises from alternative translation initiation from methionines 64 or 71, in Orai1α. Orai1 is mostly expressed in the plasma membrane, but a subset of Orai1 is located in intracellular compartments. Here we show that Ca2+ store depletion leads to trafficking and insertion of compartmentalized Orai1α in the plasma membrane via a mechanism that is independent on changes in cytosolic free‐Ca2+ concentration, as demonstrated by cell loading with the fast intracellular Ca2+ chelator dimethyl BAPTA in the absence of extracellular Ca2+. Interestingly, thapsigargin (TG) was found to be unable to induce translocation of Orai1β to the plasma membrane when expressed individually; by contrast, when Orai1β is co‐expressed with Orai1α, cell treatment with TG induced rapid trafficking and insertion of compartmentalized Orai1β in the plasma membrane. Translocation of Orai1 forms to the plasma membrane was found to require the integrity of the actin cytoskeleton. Finally, expression of a dominant negative mutant of the small GTPase ARF6, and ARF6‐T27N, abolished the translocation of compartmentalized Orai1 variants to the plasma membrane upon store depletion. These findings provide new insights into the mechanism that regulate the plasma membrane abundance of Orai1 variants after Ca2+ store depletion.

change of the STIM proteins, which, in turn, leads to store-dependent activation of Orai channels in the plasma membrane (Lunz et al., 2019;Yen & Lewis, 2019;Zhou et al., 2017).Orai1 and its mammalian paralogs, Orai2 and Orai3, have been suggested to heteromultimerize to match the extent of Ca 2+ influx to the strength of agonist stimulation (Emrich et al., 2021;Yoast et al., 2020).
Orai1 recycles dynamically and rapidly at the plasma membrane (Hodeify et al., 2015).The chaperonin-containing T-complex protein 1 (CCT) has been reported to modulate Orai1 recycling by modulating Orai1 endocytosis (Hodeify et al., 2018).Ca 2+ store depletion leads to the accumulation of Orai1 in the plasma membrane by a mechanism that requires the participation of proteins of the exocytotic machinery, such as the synaptosomal-associated protein-25 (SNAP-25) (Woodard et al., 2008).The accumulation of Orai1 in the plasma membrane upon store depletion is then facilitated by a STIM1-dependent Orai1 "trafficking trap" mechanism that retains Orai1 in the cell surface, reducing the recycling or compartmentalized pool (Hodeify et al., 2015).Two Orai1 variants have been identified in mammalian cells, the long form, Orai1α, which contains 301 amino acids, and the short variant, Orai1β, which arises from the same transcript by alternative translation initiation from a methionine at position 64, and perhaps 71, in the long form (Fukushima et al., 2012).Despite both variants supporting the store-operated and highly Ca 2+ selective current I CRAC with similar efficacy (Desai et al., 2015), several functional and biophysical differences between the two variants have been shown.
Concerning the functional differences, in addition to I CRAC , Orai1α participates in the less Ca 2+ selective store-operated current I SOC , which involves the participation of STIM1 and the transient receptor potential channel family member TRPC1 (Cheng et al., 2011;Jardin et al., 2008;Kim et al., 2009), as well as in the arachidonate-regulated current, I ARC (Desai et al., 2015).By contrast, Orai1β does not support I ARC , and its participation in I SOC is cell specific (Sanchez-Collado et al., 2022).Moreover, Orai1α but not Orai1β is required for the activation of the transcription factor NF-κB (Nieto-Felipe et al., 2023).Some differences in the biophysical properties have also been reported, including that Orai1α is more sensitive to fast Ca 2+ -dependent inactivation than Orai1β (Desai et al., 2015).Here we report significant differences in the efficacy of Orai1α and Orai1β to accumulate at the plasma membrane following store depletion.
While Ca 2+ store depletion induces plasma membrane enrichment of Orai1α when expressed either singly or together with Orai1β, the latter behaves as a defective protein that requires the presence of Orai1α to accumulate in the plasma membrane after store depletion.
Biotinylated fraction was loaded and separated in 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and analyzed by western blot analysis using anti-PMCA antibody, as control, or anti-Orai1 (catalog number: O8264, epitope: amino acids 288-301 of human Orai1), as described in Section 2.4.The specificity of the anti-Orai1 antibody was tested in whole-cell lysates from WT HEK-293 cells and O1KO HEK-293 cells (Figure S1).As expected, Orai1 was detected exclusively in WT HEK-293 cells as multiple bands corresponding to the long and short Orai1 variants with different posttranslational modifications, that is, N-linked glycosylation and phosphorylation, among others (Fukushima et al., 2012;Sanchez-Collado et al., 2019).

| Western blot analysis
Western blot analysis was performed as described previously (Sanchez-Collado et al., 2022).Briefly, cells cultured on 100-mm petri dish (8 × 10 6 cells) were stimulated with 2 μM TG or with vehicle and subsequently lysed with ice-cold NP-40 buffer pH 8 containing 137 mM of NaCl, 20 mM of Tris, 2 mM of EDTA, 10% glycerol, 1% Nonidet P-40, 1 mM of Na 3 VO 4 , and complete EDTA-free protease inhibitor tablets.After 20-30 min at 4°C, cell lysates were centrifugated at 10,000 g for 10 min.Next, supernatant was transferred to a fresh tube, the protein concentration was determined by using the BCA Protein Assay kit (23227; ThermoFisher Scientific), and subsequently, equal volume 2X Laemmli sample buffer was added.
Cell lysates were resolved by 10% or 12% SDS-PAGE, and separated proteins were electrophoretically transferred onto nitrocellulose membranes for subsequent probing.Blots were incubated overnight with 10% (w/v) BSA in Tris-buffered saline with 0.1% Tween-20 (TBST) to block residual protein binding sites.Immunodetection of Orai1 and PMCA was achieved by incubation for 1 h with anti-Orai1 antibody diluted 1:1000 in TBST and overnight with anti-PMCA antibody diluted 1:500 in TBST.The primary antibody was removed, and blots were washed six times for 5 min each with TBST.To detect the primary antibody, blots were incubated for 1 h with horseradish peroxidase-conjugated goat anti-mouse IgG antibody (115-035-003; Jackson), horseradish peroxidase-conjugated goat anti-rabbit IgG antibody (111-035-003; Jackson) diluted 1:10000 in TBST and then, exposed to enhanced chemiluminiscence reagents for 5 min.The antibody binding was assessed with a ChemiDoc Imaging System (Bio-Rad) and the density of bands was measured using ImageJ software v.1.8.0_172 (NIH).Data were normalized to the amount PMCA from the same gel.

| Fluorescence recovery after photobleaching (FRAP)
FRAP measurements were carried out on an LSM900 confocal microscope (Carl Zeiss) using a Plan-Apochromat 63 × 1.4 Oil objective at a zoom of 1.3X.eGFP was excited with the 488 nm line of 10 mW Diode laser at 1%-4% power to minimalize the bleaching of the sample during monitoring.Fluorescence emission was detected between 492 and 602 nm by an Airyscan detector.The pinhole was set at 1 Airy Unit (23.5099 pixels per micron resolution), and no line averaging was used.The detection gain 700 V was chosen such that the Orai1-eGFP pool at the plasma membrane was almost saturating the dynamic level of the detectors.The FRAP bleaching regions were randomly selected from areas with fluorescence in the dynamic range of the detector.Bleaching area (ROI) was 2.6 µm wide square and not more than 5% of the whole plasma membrane of the targeted cell.Ten bleach iterations, 0.5 s of bleaching time, at 100% transmission of a 10 mW 488 nm laser were sufficient to bleach 80%-90% of eGFP.Three pre-bleach images were taken to assess the noise.Image size, 76.98 µm × 76.98 µm, was chosen to be able to Zeiss (Fukushima et al., 2012).

| Statistical analysis
All data are presented as the mean ± standard error of mean (SEM).
Analysis of statistical significance was performed using GraphPad Prism v.8.4.3 (GraphPad Software).Kruskal-Wallis test combined with Dunn's post hoc test were used to compare the different experimental groups.For comparison between two groups, the Mann-Whitney U test was used.All data with p < 0.05 was deemed significant; "ns" = nonsignificant.

| Translocation of Orai1 to the plasma membrane upon Ca 2+ -store depletion
Previous studies from our lab and others have demonstrated that there is a pool of compartmentalized Orai1 that translocates to the plasma membrane in response to Ca 2+ -store discharge (Hodeify et al., 2018;Jardin et al., 2018;Woodard et al., 2008;Yu et al., 2010).By biotinylation of surface proteins, we have analyzed the role of TG, an inhibitor of the sarco/endoplasmic reticulum Ca 2+ -ATPase (SERCA), on the plasma membrane expression of Orai1 using CRISPR/Cas9-generated Orai1knockout (Orai1-KO) HEK-293 cells expressing wild-type (WT) Orai1, which is expected to yield a mixture of the two Orai1 variants.As shown in Figure 1a, plasma membrane expression of Orai1 was detected in resting cells, and stimulation with 2 µM TG significantly enhanced the amount of Orai1 resident in the plasma membrane (n = 6).Expression of functional Orai1 was confirmed by the determination of SOCE in these cells.Treatment of Orai1-KO HEK-293 cells expressing WT-Orai1, perfused with a Ca 2+ -free medium, with 2 µM TG resulted in a transient increase in the fura-2 fluorescence ratio as a result of Ca 2+ release from the intracellular Ca 2+ stores.Subsequent addition of extracellular Ca 2+ leads to a rise in the fura-2 fluorescence ratio that is indicative of SOCE (n = 10-15 cells/day/4 days).
Chaperonin-containing T-complex protein 1 (CCT) has been reported to regulate Orai1 trafficking and residence in the plasma membrane in Chinese hamster ovary (CHO) cells by its interaction with the Orai1-CCT binding domain located in the intracellular loop between residues 157 and 167 (Hodeify et al., 2018).Hence, we have further explored whether this mechanism also regulates the plasma membrane expression of Orai1 in HEK-293 cells.Cells were transfected with expression plasmid of a mutant Orai1 bearing the scrambled sequence 157NEKPHRSLVES167 (Orai1 157-167sc), which prevents CCT-Orai1 interaction (Hodeify et al., 2018).As shown in Figure 1a, cells expressing the Orai1 157-167sc mutant exhibited a significantly greater expression of Orai1 in the plasma membrane at rest than cells expressing WT-Orai1 (p < 0.05; n = 6), which is consistent with the crucial role of CCT in Orai1 endocytosis reported by Hodeify and coworkers (Hodeify et al., 2018).
Under these conditions, cell stimulation with TG did not significantly modify the amount of plasma membrane resident Orai1.In cells expressing the Orai1 157-167sc mutant, SOCE was significantly greater than in cells expressing WT-Orai1, which is consistent with the greater amount of Orai1 expressed in the plasma membrane (Figure 1b; p < 0.05; n = 10-15 cells/day/4 days).

| Plasma membrane location of Orai1α and
Orai1β upon Ca 2+ -store depletion Previous studies have provided evidence supporting that Orai1β is more efficient than Orai1α supporting SOCE in mouse embryonic fibroblasts (Desai et al., 2015).These findings are consistent with the lower sensitivity of Orai1β to Ca 2+ -dependent inactivation demonstrated in HEK-293 cells (Desai et al., 2015).We have now explored the efficacy of Orai1α and Orai1β supporting SOCE in HEK-293 cells.For this study, Orai1-KO HEK-293 cells were used to have a clean background.To identify potential differences between both Orai1 variants Orai1-KO cells were transfected with expression plasmids for Orai1α, Orai1β or both using thymidine kinase (TK) promoter, which results in a considerably smaller expression level than plasmids incorporating a CMV promoter (X.Zhang et al., 2019).As shown in Figure 2a, treatment of wild-type HEK-293 cells, perfused with a Ca 2+ -free medium, with 2 µM TG induced a transient increase in the fura-2 fluorescence ratio due to Ca 2+ release from the intracellular Ca 2+ stores.Subsequent addition of extracellular Ca 2+ results in a rise in the fura-2 fluorescence ratio indicative of SOCE.
As expected, in Orai1-KO cells, TG-evoked Ca 2+ entry was essentially abrogated, and the expression of either Orai1α or Orai1β rescued SOCE in these cells (Figure 2a).Interestingly, we found that in Orai1-KO cells expressing Orai1α TG-induced SOCE was greater than in Orai1-KO cells expressing Orai1β, thus suggesting that the efficacy of Orai1α supporting SOCE in Orai1-KO HEK-293 cells was greater than that of Orai1β (Figure 2a); therefore, we analyzed the plasma membrane expression of both variants under these experimental conditions.Orai1-KO cells and Orai1-KO cells expressing either Orai1α, Orai1β or both variants were stimulated for 1 min in a medium containing 1.8 mM Ca 2+ with TG or left untreated, and the plasma membrane location of the Orai1 variants was analyzed by surface biotinylation.As shown in Figure 2b, Orai1α and Orai1β were detected in the plasma membrane in resting conditions when they were expressed alone or in combination.In Orai1-KO cells expressing Orai1α, Ca 2+ store depletion by TG significantly increased the plasma membrane expression of Orai1α by 36 ± 1% as compared to nonstimulated cells.By contrast, TG was unable to enhance the surface expression of Orai1β over the pretreatment level when expressed alone in Orai1-KO cells.Interestingly, in cells expressing both variants TG significantly enhanced over the pretreatment level the plasma membrane expression of Orai1α and Orai1β by 39 ± 8 and 81 ± 2%, respectively (Figure 2b), thus suggesting that agonist-induced plasma membrane location of the intracellular pool of Orai1β requires co-expression of Orai1α.These findings are consistent with the analysis of SOCE in Orai1-KO cells expressing both Orai1α and Orai1β, which was significantly greater than that observed in WT and Orai1-KO cells expressing either Orai1α or Orai1β alone (Figure 2a).3.4 | Orai1α and Orai1β trafficking to the plasma membrane upon Ca 2+ -store depletion requires functional ARF6 ARF6 regulates endosomal plasma membrane trafficking in many cell types (Aikawa & Martin, 2003).ARF6 has been reported to mediate constitutive recycling of Orai1 in Xenopus oocytes and HEK-293 cells F I G U R E 2 Distribution of Orai1α and Orai1β in the plasma membrane at rest and upon stimulation with thapsigargin.(a) Representative Ca 2+ mobilization in response to 2 µM thapsigargin (TG) measured using fura-2 in wild-type HEK-293 cells (HEK-293 WT), Orai1-KO HEK-293 cells (HEK-293 O1KO) and Orai1-KO HEK-293 cells expressing either Orai1α (HEK-293O1KO(O1α)), Orai1β (HEK-293O1KO(O1β)) or both (HEK-293O1KO(O1α + O1β)), as described.Cells were superfused with a Ca 2+ -free HBS (100 µM EGTA added) and stimulated with 2 µM TG followed by re-addition of CaCl 2 (1.8 mM) to estimate Ca 2+ influx.Quantification of TG-evoked Ca 2+ release from the intracellular stores and entry is shown in the bar graphs.Traces are representative of 10-15 cells/day/4 days.Data are represented as mean ± SEM and were statistically analyzed using Kruskal-Wallis test with multiple comparisons (Dunn's test).*p < 0.05 and $$$ p < 0.001.(b) Orai1-KO HEK-293 cells (O1KO), and Orai1-KO HEK-293 cells expressing either Orai1α (O1KO(O1α)), Orai1β (O1KO(O1β)) or both (O1KO(O1α + O1β)) were suspended in a medium containing 1.8 mM CaCl 2 , and then stimulated for 1 min with 2 µM TG or left untreated.Samples were then mixed with biotinylation buffer containing EZ-Link sulfo-NHS-LC-biotin, and cell surface proteins were labeled by biotinylation.Labeled proteins were pulled down with streptavidin-coated agarose beads.The pellet (containing the plasma membrane fraction) and the supernatant were analyzed by SDS-PAGE and western blot analysis using anti-Orai1 antibody, as indicated (top (pellet) and bottom (supernatant) panels).Membranes were probed with anti-PMCA antibody.Molecular masses indicated on the right were determined using molecular-mass markers run in the same gel.These results are representative of 6 separate experiments.Quantification of Orai1 plasma membrane expression under the different experimental conditions normalized to the PMCA expression is depicted in the bar graph.Data are represented as mean ± SEM and were statistically analyzed using Kruskal-Wallis test with multiple comparisons (Dunn's test).*p < 0.05 as compared with their respective control.ɸ p < 0.05 as compared with TGstimulated O1KO(O1β) cells.(c) HEK-293 O1KO cells expressing either Orai1α-GFP (O1KO(O1α)), Orai1β-GFP (O1KO(O1β)) or Orai1α-CFP and Orai1β-GFP (O1KO(O1α + O1β)) were mounted on an inverted LSM900 confocal microscope (Carl Zeiss).Cells were suspended in a medium containing 1.8 mM CaCl 2 , and then stimulated with 2 µM TG or left untreated, as indicated.The mobility of Orai1α (left panel) or Orai1β (middle and right panels) was determined by using fluorescence recovery after photobleaching (FRAP) as described in Section 2. The graphs represent the fit normalized recovery curves of both Orai1 variants in resting and upon stimulation.SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis SEM, standard error of mean; WT, wild-type.(Yeh et al., 2020;Yu et al., 2010).We have investigated the involvement of ARF6 in TG-induced trafficking of Orai1 variants to the plasma membrane by overexpressing either wild-type ARF6 (ARF6wt) or a dominant negative ARF6 mutant (ARF6-T27N; ARF6dn).Orai1-KO cells expressing Orai1α alone or in combination with Orai1β were transfected with expression plasmids for ARF6wt or ARF6dn or empty vector.Forty-eight hours after transfection cells were stimulated for 1 min in a medium containing 1.8 mM Ca 2+ in the absence or presence of TG and the plasma membrane location of the Orai1 variants was analyzed by surface biotinylation.As shown in Figure 5a, in cells expressing ARF6wt TG-induced translocation of Orai1α and Orai1β (when co-expressed with Orai1α) to the plasma membrane was not significantly modified; however, expression of ARF6dn essentially abolished TG-evoked trafficking of Orai1α and Orai1β to the plasma membrane, which indicates that this process requires functional ARF6.
F I G U R E 3 Role of intracellular Ca 2+ in the plasma membrane location of Orai1α and Orai1β at rest and upon stimulation with thapsigargin.Orai1-KO HEK-293 cells (O1KO), and Orai1-KO HEK-293 cells expressing either Orai1α (O1KO(O1α)), Orai1β (O1KO(O1β)) or both (O1KO (O1α + O1β)) were either loaded with dimethyl BAPTA (b) or treated with vehicle (a) and suspended in a Ca 2+ -free medium (100 µM EGTA added).Cells were then stimulated for 1 min with 2 µM TG or left untreated.Samples were then mixed with biotinylation buffer containing EZ-Link sulfo-NHS-LC-biotin, and cell surface proteins were labeled by biotinylation.Labeled proteins were pulled down with streptavidin-coated agarose beads.The pellet (containing the plasma membrane fraction) and the supernatant were analyzed by SDS-PAGE and western blot analysis using anti-Orai1 antibody, as indicated (top (pellet) and bottom (supernatant) panels).Membranes were probed with anti-PMCA antibody.Molecular masses indicated on the right were determined using molecular-mass markers run in the same gel.These results are representative of six separate experiments.Quantification of Orai1 plasma membrane expression under the different experimental conditions normalized to the PMCA expression is depicted in the bar graph.Data are represented as mean ± SEM and were statistically analyzed using Kruskal-Wallis test with multiple comparisons (Dunn's test).*p < 0.05 as compared with their respective control.ɸ p < 0.05 as compared with TGstimulated O1KO(O1β) cells.SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis SEM, standard error of mean; TG, thapsigargin; WT, wild-type.Analysis of TG-stimulated Ca 2+ mobilization in Orai1-KO cells expressing Orai1α alone or in combination with Orai1β revealed that expression of ARF6wt did not significantly alter either TG-induced Ca 2+ release or entry; however, in cells expressing ARF6dn, TGevoked SOCE was significantly attenuated, without having any effect on Ca 2+ release from the intracellular stores (Figure 5b).These findings further confirm the functional role of ARF6 in TG-induced trafficking of Orai1 variants to the plasma membrane and supports an important role for this mechanism in TG-evoked SOCE.

| DISCUSSION
Orai1α and Orai1β were identified in 2012 as two Orai1 variants generated by alternative translation initiation from the same transcript (Fukushima et al., 2012).Both Orai1 forms have been reported to support I CRAC with similar efficiency (Desai et al., 2015).However, the participation of both variants in I SOC has been reported to be cell specific (Sanchez-Collado et al., 2022), and only Orai1α is involved in the arachidonate-regulated Ca 2+ current I ARC (Desai et al., 2015).The 63 amino acids located in the N-terminal region of Orai1α mediates the association of this variant with a number of interacting partners, such as caveolin, that might lead to different plasma membrane location and mobilities between the two Orai1 forms.As a result, Orai1α and Orai1β have been shown to exhibit distinct plasma membrane mobilities in resting cells as detected by FRAP, where Orai1β shows a faster diffusion than the majority of Orai1α subunits (Fukushima et al., 2012).These, and other observations (Desai et al., 2015;Sanchez-Collado et al., 2022), suggest that both Orai1 variants do not form hetero-multimeric channels or, at least, that is not the predominant presentation.Here, we show that Orai1α and Orai1β exhibit differences in the trafficking and plasma membrane expression upon Ca 2+ store depletion.It has been reported that, in addition to the plasma membrane pool of Orai1, there is a subset of Orai1 located in an endosomal compartment that is inserted in the plasma membrane upon Ca 2+ store depletion in an exocytotic manner that involves proteins such as SNAP25 (synaptosomal-associated protein-25) and Rho (Woodard et al., 2008;Yu et al., 2010).Our results indicate that when Orai1α and Orai1β are expressed either singly or together in Orai1-KO HEK-293 cells a subset of the proteins is constitutively expressed in the plasma membrane.When the Orai1 variants are individually expressed, store depletion induces a significant increase in the plasma membrane expression of Orai1α, while the expression of the compartmentalized Orai1β requires the presence of Orai1α.These findings suggest that there is some form of functional interaction between both variants, so the trafficking of both Orai1 forms is associated with Orai1α, but this is unknown at present.

Store depletion-associated trafficking of compartmentalized
Orai1 variants is independent of Ca 2+ influx and intracellular Ca 2+ mobilization and, according to the FRAP analysis, occurs within the first few seconds after store depletion.Furthermore, this mechanism is strongly dependent on the integrity of the actin filament network, which suggests that the actin cytoskeleton supports the translocation of the Orai1 forms to the plasma membrane.This observation is consistent with the regulation of plasma membrane abundance of other Ca 2+ channels, such as Ca v 1.2 by exocytic and post-endocytic trafficking to the cell surface (Conrad et al., 2018).
Orai1 recycling at the plasma membrane has been reported to involve the CCT chaperonin complex, required for Orai1 endocytosis (Hodeify et al., 2018), as well as a constitutive exocytosis with a rate constant of approximately 0.1 min −1 (Hodeify et al., 2015).Furthermore, Orai1 plasma membrane abundance is enhanced by different mechanisms, including Ca 2+ store depletion (Woodard et al., 2008) or the secretory pathway Ca 2+ -ATPase (SPCA) in specialized mammary epithelial cells associated to store-independent Ca 2+ entry (Cross et al., 2013).Concerning the trafficking of Orai1 to the plasma membrane upon store depletion, we previously reported that the mechanism involves the participation of SNAP-25 (Woodard et al., 2008), and here, we present evidence for the involvement of the adenosine diphosphate ribosylation factor ARF6, as expression of the dominant negative ARF6-T27N mutant abolished store depletion-evoked translocation of Orai1 to the plasma membrane.
We noticed that while expression of the dominant negative ARF6 mutant attenuated Orai1 variants expression in the plasma membrane and SOCE, expression of WT-ARF6 did not significantly enhance these processes.We attribute this observation to the fact that ARF6 is necessary for the translocation of Orai1 to the plasma membrane, but the native expression of ARF6 might be sufficient to support this mechanism, so that overexpression of the protein does not have an increasing effect on Orai1 expression at the plasma membrane.ARF6 has been reported to participate in internalization of TRPC1 during the recycling of this channel (de Souza et al., 2015).Furthermore, ARF6 modulates SOCE, as a downstream effector of RAS association domain family 4 (RASSF4), by regulating the plasma membrane PI(4,5)P 2 levels (Chen et al., 2017).More recently, in patients with atopic dermatitis, ARF6 has been shown to mediate the exocytic insertion of the Orai1 single-nucleotide polymorphisms S218G-Orai1 and N223S-Orai1 in the plasma membrane, thus escaping from degradation, which leads to a mismatch in the Orai1-STIM stoichiometry and, consequently, to attenuation of SOCE and the associated gene expression (Yeh et al., 2020).Summarizing, our results indicate that Ca 2+ store depletion modulates the cell surface abundance of Orai1α and Orai1β by a mechanism dependent on ARF6 and the actin filament network but independent on intracellular Ca 2+ mobilization.Interestingly, while Orai1β constitutively translocates to the plasma membrane in resting cells with a similar efficiency as Orai1α, the former acts as a defective form that requires the presence of Orai1α for the trafficking of compartmentalized subunits to the plasma membrane upon store depletion.Current evidence indicates that the Orai1α/Orai1β expression ratio ranges from 0.3 in nontumoral breast epithelial MCF10A cells to slightly over 1 in Jurkat cells (Fukushima et al., 2012;Sanchez-Collado et al., 2019) due to still unknown mechanism.In this context, the mechanism reported here might facilitate that cells express both variants evenly on the cell surface independently on the Orai1α/Orai1β expression ratio.
monitor two neighboring cells to calculate the overall photobleaching and evaluate focal plane drifts during postbleaching data acquisition.The monitored fluorescence recovery plateau was between 100 and 150 s at 23°C with mobile fraction >80%.Sample rate was 120-280 images at a rate of 2.1 image/s.20-25 cells (40-50 FRAP curves) of Orai1α Orai1β were obtained from 12 to 15 independent experiments.To compare recovery curves for different cells, we normalized the data to correct for variations in protein expression levels, background fluorescence, and loss of fluorescence during bleaching.The percentage of mobile fractions of the two Orai1 forms, background photobleaching correction, photofading factor, fit formula (mono exponential), fit range, and fit normalized recovery curve were calculated by the FRAP analysis module included in the ZenBlue 3.4 acquisition and analysis software from Figure 4a, in cells pretreated with Cyt D TG was unable to induce

F
I G U R E 4 Effect of cytochalasin D on the plasma membrane location of Orai1α and Orai1β at rest and thapsigargin-treated cells.(a) Orai1-KO HEK-293 cells (O1KO), and Orai1-KO HEK-293 cells expressing either Orai1α (O1KO(O1α)), Orai1β (O1KO(O1β)) or both (O1KO (O1α + O1β)) were pretreated with 10 µM cytochalasin D or the vehicle for 40 min, as indicated.Cells were suspended in a medium containing 1. 8 mM CaCl 2 , and then stimulated for 1 min with 2 µM TG or left untreated.Samples were then mixed with biotinylation buffer containing EZ-Link sulfo-NHS-LC-biotin, and cell surface proteins were labeled by biotinylation.Labeled proteins were pulled down with streptavidin-coated agarose beads.The pellet (containing the plasma membrane fraction) and the supernatant were analyzed by SDS-PAGE and western blot analysis using anti-Orai1 antibody, as indicated (top (pellet) and bottom (supernatant) panels).Membranes were probed with anti-PMCA antibody.Molecular masses indicated on the right were determined using molecular-mass markers run in the same gel.These results are representative of six separate experiments.Quantification of Orai1 plasma membrane expression under the different experimental conditions normalized to the PMCA expression is depicted in the bar graph.Data are represented as mean ± SEM and were statistically analyzed using Kruskal-Wallis test with multiple comparisons (Dunn's test).*p < 0.05 as compared with their respective control.ɸ p < 0.05 as compared with the corresponding TG-stimulated cells not pretreated with cytochalasin D. (b) HEK-293 O1KO expressing either Orai1α (O1KO(O1α)), Orai1β (O1KO(O1β)) or Orai1α-CFP and Orai1β-GFP (O1KO(O1α + O1β)) were pretreated with 10 µM cytochalasin D or the vehicle for 40 min, as indicated, and mounted on an inverted LSM900 confocal microscope (Carl Zeiss).Cells were suspended in a medium containing 1.8 mM CaCl 2 , and then stimulated with 2 µM TG, or left untreated.The mobility of Orai1α (left panel) or Orai1β (middle and right panels) was determined by using fluorescence recovery after photobleaching (FRAP) as described in Section 2. The graphs represent the fit normalized recovery curves of both Orai1 variants in resting and upon stimulation.

F
I G U R E 5 Role of ARF6 in the plasma membrane expression of an Orai1 upon stimulation with thapsigargin.(a) Orai1-KO HEK-293 cells expressing either Orai1α (O1KO(O1α)) or Orai1α and Orai1β (O1KO(O1α + O1β)) were transfected with expression plasmid for wild-type ARF6 (ARF6wt), a dominant negative mutant of ARF6 (ARF6dn) or empty vector.Forty-eight hours later cells were suspended in a medium containing 1.8 mM CaCl 2 , and then stimulated for 1 min with 2 µM TG or left untreated.Samples were then mixed with biotinylation buffer containing EZ-Link sulfo-NHS-LC-biotin, and cell surface proteins were labeled by biotinylation.Labeled proteins were pulled down with streptavidin-coated agarose beads.The pellet (containing the plasma membrane fraction) and the supernatant were analyzed by SDS-PAGE and western blot analysis using anti-Orai1 antibody, as indicated (top (pellet) and bottom (supernatant) panels).Membranes were probed with anti-PMCA antibody.Molecular masses indicated on the right were determined using molecular-mass markers run in the same gel.Membranes were probed with anti-PMCA antibody.These results are representative of six separate experiments.Quantification of Orai1 plasma membrane expression under the different experimental conditions normalized to the PMCA expression is depicted in the bar graph.Data are represented as mean ± SEM and were statistically analyzed using Kruskal-Wallis test with multiple comparisons (Dunn's test).*p < 0.05.(b and c) Representative Ca 2+ mobilization in response to 2 µM thapsigargin (TG) measured using fura-2 in Orai1-KO HEK-293 cells expressing Orai1α or Orai1α and Orai1β alone or in combination with wild-type ARF6 (ARF6wt) or a dominant negative mutant of ARF6 (ARF6dn), as described.Cells were superfused with a Ca 2+ -free HBS (100 µM EGTA added) and stimulated with 2 µM TG, followed by re-addition of CaCl 2 (1. 8 mM) to estimate Ca 2+ influx.(d) Quantification of TG-evoked Ca 2+ release from the intracellular stores and entry is shown in the bar graphs.Data are represented as mean ± SEM and were statistically analyzed using Kruskal-Wallis test with multiple comparisons (Dunn's test).***p < 0.001 and ****p < 0. 0001.SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis SEM, standard error of mean; WT, wild-type.