ORAI2 Down-Regulation Potentiates SOCE and Decreases Aβ42 Accumulation in Human Neuroglioma Cells

Senile plaques, the hallmarks of Alzheimer’s Disease (AD), are generated by the deposition of amyloid-beta (Aβ), the proteolytic product of amyloid precursor protein (APP), by β and γ-secretase. A large body of evidence points towards a role for Ca2+ imbalances in the pathophysiology of both sporadic and familial forms of AD (FAD). A reduction in store-operated Ca2+ entry (SOCE) is shared by numerous FAD-linked mutations, and SOCE is involved in Aβ accumulation in different model cells. In neurons, both the role and components of SOCE remain quite obscure, whereas in astrocytes, SOCE controls their Ca2+-based excitability and communication to neurons. Glial cells are also directly involved in Aβ production and clearance. Here, we focus on the role of ORAI2, a key SOCE component, in modulating SOCE in the human neuroglioma cell line H4. We show that ORAI2 overexpression reduces both SOCE level and stores Ca2+ content, while ORAI2 downregulation significantly increases SOCE amplitude without affecting store Ca2+ handling. In Aβ-secreting H4-APPswe cells, SOCE inhibition by BTP2 and SOCE augmentation by ORAI2 downregulation respectively increases and decreases Aβ42 accumulation. Based on these findings, we suggest ORAI2 downregulation as a potential tool to rescue defective SOCE in AD, while preventing plaque formation.

In HEK293T cells, ORAI2 was also suggested to be part of the elusive ER Ca 2+ leak channel [52], therefore offering a possible link between ER and SOCE alterations found in different FAD models ranging from cell lines, fibroblasts and induced pluripotent stem cells from FAD patients [19,20,53] to neurons and astrocytes from AD mouse models [9][10][11]14,46,54].
In this study, we focus on the human neuroglioma cell line H4 and its clone H4-APPswe that stably expresses the FAD-linked APP Swedish mutation. These cells are a good model to modulate glial SOCE components and, at the same time, to verify how this modulation affects Aβ accumulation. Of note, glial cells, among the other relevant features also directly participate in Aβ production [55][56][57][58], especially upon increased cellular stress caused by different environmental factors [59] and neuroinflammation [60].
In neuroglioma cells, we have characterized the effect of increased and decreased ORAI2 expression on SOCE level and ER Ca 2+ content and found that ORAI2 downregulation significantly increases SOCE amplitude while leading to a marked reduction in the Aβ42/Aβ40 ratio in the extracellular environment. We suggest ORAI2 as a novel therapeutic target in AD because its downregulation allows for the rescue of SOCE reduction, and at the same time, it reduces Aβ42 secretion by glial cells.
Altogether, we can state that the overexpression of ORAI2 together with STIM1 decreases both ER Ca 2+ release and SOCE induced by store depletion. These findings expand previous observations obtained in HEK293T cells, but in the absence of STIM1 overexpression [52]. They also suggest that addition of a C-terminal fluorescent-tag mVenus to ORAI2-myc does not interfere with its functionality, since its expression mimics the Ca 2+ -related phenotype obtained with the expression of ORAI2-myc. . Upon shifting to a Ca 2+ -free mKRB containing EGTA (0.6 mM), cells were challenged with histamine (Hist, 100 μM) and CPA (20 μM) in the same medium. After 8 min, cells were bathed in mKRB containing CaCl2 and CPA to detect maximal SOCE amplitude. Boxplots show Ca 2+ release (B) and SOCE (C) peak height and AUC, upon baseline subtraction. Data are expressed as percentage of control (CTRL) cells expressing only STIM1 and cyt-AEQ. Coverslips were from at least three independent cell batches. Wilcoxon-Mann-Whitney test (* p < 0.05; ** p < 0.01; *** p < 0.001).  (1 mM). Upon shifting to a Ca 2+ -free mKRB containing EGTA (0.6 mM), cells were challenged with histamine (Hist, 100 µM) and CPA (20 µM) in the same medium. After 8 min, cells were bathed in mKRB containing CaCl 2 and CPA to detect maximal SOCE amplitude. Boxplots show Ca 2+ release (B) and SOCE (C) peak height and AUC, upon baseline subtraction. Data are expressed as percentage of control (CTRL) cells expressing only STIM1 and cyt-AEQ. Coverslips were from at least three independent cell batches. Wilcoxon-Mann-Whitney test (* p < 0.05; ** p < 0.01; *** p < 0.001).
Altogether, we can state that the overexpression of ORAI2 together with STIM1 decreases both ER Ca 2+ release and SOCE induced by store depletion. These findings expand previous observations obtained in HEK293T cells, but in the absence of STIM1 overexpression [52]. They also suggest that addition of a C-terminal fluorescent-tag mVenus to ORAI2-myc does not interfere with its functionality, since its expression mimics the Ca 2+ -related phenotype obtained with the expression of ORAI2-myc. Based on siRNA perturbation experiments and computational data, it was also suggested that, in HEK293T cells, ORAI2 contributes to ER Ca 2+ leak [52]. Our data indicate that, in these cells, endogenous ORAI2 localizes primarily in the early endosomes, as estimated by immunofluorescence studies (Supplementary Figure S1). The Pearson's coefficient was in fact larger for ORAI2 colocalization with markers of the early endosomes (Supplementary Figure Figure S1E). However, only with this latter marker was there a statistically significant difference (p < 0.05 compared to ER-GFP and p < 0.01 compared to LPBA).
To better define ORAI2 localization in H4 cells, we exploited the overexpression of ORAI2-myc-mVenus with the mCherry protein, targeted to different organelles. In intact H4 cells, ORAI2-myc-mVenus was spread over a large vesicular network (Figure 3), which was however clearly distinct from ER membranes ( Figure 3A) and showed only a minor overlap with the endosomal compartment ( Figure 3B). ORAI2 better localized at the PM upon live staining with wheat germ agglutinin (WGA) ( Figure 3C). Of note, the PM localization of ORAI2-mVenus was clearly visible upon cell permeabilization ( Figure 3D, see also Figure 2D). Based on siRNA perturbation experiments and computational data, it was also suggested that, in HEK293T cells, ORAI2 contributes to ER Ca 2+ leak [52]. Our data indicate that, in these cells, endogenous ORAI2 localizes primarily in the early endosomes, as estimated by immunofluorescence studies ( Figure S1). The Pearson's coefficient was in fact larger for ORAI2 colocalization with markers of the early endosomes ( Figure S1B,C) with respect to a marker of ER ( Figure S1A) or late endosomes ( Figure S1D), being respectively: 0.26 ± 0.07 (ER-GFP), 0.28 ± 0.04 (LBPA), 0.44 ± 0.09 (rab5) and 0.61 ± 0.07 (EEA1) ( Figure S1E). However, only with this latter marker was there a statistically significant difference (p < 0.05 compared to ER-GFP and p < 0.01 compared to LPBA).
To better define ORAI2 localization in H4 cells, we exploited the overexpression of ORAI2-myc-mVenus with the mCherry protein, targeted to different organelles. In intact H4 cells, ORAI2-myc-mVenus was spread over a large vesicular network (Figure 3), which was however clearly distinct from ER membranes ( Figure 3A) and showed only a minor overlap with the endosomal compartment ( Figure 3B). ORAI2 better localized at the PM upon live staining with wheat germ agglutinin (WGA) ( Figure 3C). Of note, the PM localization of ORAI2-mVenus was clearly visible upon cell permeabilization ( Figure 3D, see also Figure 2D). From a functional point of view, we investigated whether the reduction in ER Ca 2+ release, observed in ORAI2-overexpressing cells, was the consequence of a decrease in SOCE amplitude or IP3R signalling, or even a direct effect of ORAI2 on ER Ca 2+ handling. We directly evaluated the effect of ORAI2 at the ER Ca 2+ level by expressing ORAI2 together with G-CEPIA1er, a genetically encoded, ER-targeted Ca 2+ probe with very low Ca 2+ affinity and good dynamic range [63]. Surprisingly, G-CEPIA1er expression in H4 cells was extremely variable being also hampered by a high level of instability and photobleaching. We thus decided to take advantage of HeLa cells for this type of experiment because of a much more stable and reliable G-CEPIA1er signal. In particular, with respect to control (CTRL), void-vector transfected cells, ORAI2-myc overexpression decreased the resting ER From a functional point of view, we investigated whether the reduction in ER Ca 2+ release, observed in ORAI2-overexpressing cells, was the consequence of a decrease in SOCE amplitude or IP 3 R signalling, or even a direct effect of ORAI2 on ER Ca 2+ handling. We directly evaluated the effect of ORAI2 at the ER Ca 2+ level by expressing ORAI2 together with G-CEPIA1er, a genetically encoded, ER-targeted Ca 2+ probe with very low Ca 2+ affinity and good dynamic range [63]. Surprisingly, G-CEPIA1er expression in H4 cells was extremely variable being also hampered by a high level of instability and photobleaching. We thus decided to take advantage of HeLa cells for this type of experiment because of a much more stable and reliable G-CEPIA1er signal. In particular, with respect to control (CTRL), void-vector transfected cells, ORAI2-myc overexpression decreased the resting ER fluorescence ( Figure 4A,B). Since G-CEPIA1er is a non-ratiometric Ca 2+ indicator, we obtained an indirect estimation of the ER Ca 2+ content, calculated as the maximal change obtained after CPA and ionomycin addition ( Figure 4A). Upon ORAI2-myc overexpression, the maximal ER Ca 2+ decrease was significantly reduced by about 40% in ORAI2-myc-expressing cells ( Figure 4B). Int. J. Mol. Sci. 2020, 21, x 6 of 18 fluorescence ( Figure 4A,B). Since G-CEPIA1er is a non-ratiometric Ca 2+ indicator, we obtained an indirect estimation of the ER Ca 2+ content, calculated as the maximal change obtained after CPA and ionomycin addition ( Figure 4A). Upon ORAI2-myc overexpression, the maximal ER Ca 2+ decrease was significantly reduced by about 40% in ORAI2-myc-expressing cells ( Figure 4B). We also estimated the ER Ca 2+ refilling process in permeabilized HeLa cells expressing G-CEPIA1er ( Figure 4C). In permeabilized cells, ORAI2 overexpression significantly impaired the ER Ca 2+ refilling process by more than 62 ± 7% as estimated by maximal CaCl2 (1 mM) addition ( Figure  4D). Of note, in ORAI2 overexpressing cells, the rate of Ca 2+ uptake was markedly different from control cells even at the initial stage of the refilling process ( Figure 4C), suggesting an impairment of the SERCA pump activity rather than an effect on ER Ca 2+ leakage. This latter should in fact become . Upon CPA washing, cells were exposed to an intracellular solution containing EGTA (50 µM). Cells were permeabilized for 20 s with digitonin (100 µM) and store refilling was induced by CaCl 2 addition in the presence of ATP (100 µM). (C) Representative traces of the maximal fluorescence changes (∆F) normalized to the initial basal level (F 0 ). (D) Boxplots of ∆F/F 0 measured upon CaCl 2 addition, a rough estimate of the ER Ca 2+ concentration (n = 11, number of coverslips from three independent cell batches, Wilcoxon-Mann-Whitney test, ** p < 0.01; *** p < 0.001).
We also estimated the ER Ca 2+ refilling process in permeabilized HeLa cells expressing G-CEPIA1er ( Figure 4C). In permeabilized cells, ORAI2 overexpression significantly impaired the ER Ca 2+ refilling process by more than 62 ± 7% as estimated by maximal CaCl 2 (1 mM) addition ( Figure 4D). Of note, in ORAI2 overexpressing cells, the rate of Ca 2+ uptake was markedly different from control cells even at the initial stage of the refilling process ( Figure 4C), suggesting an impairment of the SERCA pump activity rather than an effect on ER Ca 2+ leakage. This latter should in fact become apparent only after a certain level of store replenishment. Altogether these findings support the idea that upregulation of ORAI2 impairs the SERCA activity independently of its effect on SOCE level.

SOCE Levels and Aβ-Secretion in Neuroglioma Cells
Different groups have suggested a close relationship between cytosolic Ca 2+ levels and Aβ accumulation, however with contrasting results [7,24,25,27]. To investigate this issue, we employed the H4-APPswe cell line, that stably express the human APP Swedish mutation (K670/M671L) and accumulates Aβ in the culture medium [64].
It was demonstrated that ORAI2 knockout increases SOCE in mouse T lymphocytes [35], we thus hypothesized that augmenting SOCE by downregulating ORAI2 would reduce Aβ accumulation by H4-APPswe cells. We firstly checked how ORAI2 modulates SOCE in this cell line. To detect SOCE we thus employed the Ca 2+ -addback protocol we adopted for H4 cells, as described in Figure 1. Unfortunately, very heterogeneous Ca 2+ increases were observed following CaCl 2 addition to H4-APPswe cells upon store depletion with histamine and CPA in Ca 2+ -free, EGTA-containing mKRB ( Figure S3A). Large Ca 2+ rises were observed also in the absence of store depletion that were insensitive to a two-minute-treatment with different Ca 2+ channel inhibitors, such as GdCl 3 (1 µM), nimodipine (Nim, 1 µM) plus verapamil (Ver, 10 µM), or carbenexolone (CBX, 50 µM) ( Figure S3B). These Ca 2+ rises were likely due to Ca 2+ entry through other channels, possibly activated by removal of extracellular Ca 2+ . To overcome this problem, cells were challenged with CPA (20 µM) in mKRB in the continuous presence of extracellular CaCl 2 (1 mM) ( Figure S3C). The contribution of Ca 2+ release to Ca 2+ rises was evaluated in parallel experiments by adding CPA in a Ca 2+ -free, EGTA-containing, mKRB ( Figure S3D). Under these conditions, the Ca 2+ release, induced by CPA, was rather modest and returned to the resting Ca 2+ level in a couple of minutes. Furthermore, given the strong dependence of SOCE on membrane hyperpolarization [31], to reduce the variability due to membrane potential, additional experiments were carried out in KCl-mKRB, upon substitution of extracellular NaCl with equimolar KCl (100 mM). In this condition, to compensate for the reduced driving force for Ca 2+ entry, the CaCl 2 concentration was increased from 1 to 10 mM [62]. Similarly to experiments in H4 cells, STIM1 was expressed in the absence (CTRL) or presence of ORAI2. It is worth mentioning that, when store depletion occurs in a Ca 2+ -containing medium, the initial rate of Ca 2+ entry following CPA addition is not indicative solely of SOCE kinetics, since the signal is contaminated by the Ca 2+ rise due to Ca 2+ release from intracellular stores. We thus estimated differences in SOCE amplitude by Ca 2+ peak and AUC, measured at two minutes, when Ca 2+ release was practically exhausted ( Figure S3D).
In H4-APPswe cells, bathed in Ca 2+ -containing mKRB, overexpression of ORAI2-myc, together with STIM1, significantly decreased SOCE by about 50%, in terms of both peak and AUC ( Figure  S4A); similar results were obtained when using KCl-mKRB ( Figure S4B). Thus, as shown in H4 cells, SOCE reached lower levels than in control cells upon ORAI2 upregulation. Conversely, ORAI2 downregulation by siRNA (siORAI2) increased SOCE by about 80%, when measured in mKRB ( Figure 5A,B) and 70% in KCl-mKRB ( Figure 5C,D), both in terms of peak amplitude and AUC. As expected, ORAI1 downregulation reduced SOCE by more than 60%, in both standard-and KCl-mKRB ( Figure S5). To investigate the effect of ORAI2 (plus STIM1) overexpression on the store Ca 2+ content of H4-APPswe cells, we induced maximal Ca 2+ release by histamine and CPA in a Ca 2+ -free, EGTAcontaining medium. Similarly, to what observed in H4 cells ( Figure 1A,B), ORAI2 overexpression significantly reduced ER Ca 2+ release by 32 ± 11% in terms of area ( Figure 6A,B). Instead, upon ORAI2 downregulation by siRNA, no difference was found in Ca 2+ released from the IP3-sensitive stores ( Figure 6C,D). To investigate the effect of ORAI2 (plus STIM1) overexpression on the store Ca 2+ content of H4-APPswe cells, we induced maximal Ca 2+ release by histamine and CPA in a Ca 2+ -free, EGTA-containing medium. Similarly, to what observed in H4 cells ( Figure 1A,B), ORAI2 overexpression significantly reduced ER Ca 2+ release by 32 ± 11% in terms of area ( Figure 6A,B). Instead, upon ORAI2 downregulation by siRNA, no difference was found in Ca 2+ released from the IP 3 -sensitive stores ( Figure 6C,D).
In summary, ORAI2 overexpression modifies SOCE, as well as the store Ca 2+ content and resting cytosolic levels. In contrast, ORAI2 downregulation selectively increases SOCE without altering the store Ca 2+ content or basal Ca 2+ levels. Since our data show that the SOCE inhibitor BTP2 increases Aβ42 accumulation ( Figure S2), it is expected that upregulation of SOCE will negatively affect this parameter. ELISA experiments were performed on conditioned media of H4-APPswe cells collected at 48 h upon transfection with siORAI2 or control siRNA. Under these conditions, the Aβ42 level was reduced by 44 ± 10%, while the Aβ40 level was increased by 86 ± 13%, resulting in an approximately 70% decrease of the Aβ42/Aβ40 ratio (Figure 7). In summary, ORAI2 overexpression modifies SOCE, as well as the store Ca 2+ content and resting cytosolic levels. In contrast, ORAI2 downregulation selectively increases SOCE without altering the store Ca 2+ content or basal Ca 2+ levels. Since our data show that the SOCE inhibitor BTP2 increases Aβ42 accumulation (Supplementary Figure S2), it is expected that upregulation of SOCE will negatively affect this parameter. ELISA experiments were performed on conditioned media of H4-APPswe cells collected at 48 h upon transfection with siORAI2 or control siRNA. Under these conditions, the Aβ42 level was reduced by 44 ± 10%, while the Aβ40 level was increased by 86 ± 13%, resulting in an approximately 70% decrease of the Aβ42/Aβ40 ratio (Figure 7).

Discussion
According to recent models, the store-operated Ca 2+ entry (SOCE) based on ICRAC is due to Ca 2+ permeation across hexameric channels formed by ORAI (1/2/3) subunits [28,30]. While the role of ORAI1 has widely been clarified, the contribution of the other subunits to ICRAC and SOCE emerged only later [28,33]. Thanks to in-depth studies on the immune system, it has elegantly been demonstrated that, in mouse T cells, ORAI2 forms heteromeric channels with ORAI1 and contributes to a reduced Ca 2+ permeation through ICRAC [35,39]. A negative role for ORAI2 on Ca 2+ entry was already postulated following overexpression studies in HEK293 cells [67]. In addition, Meyer and coworkers demonstrated that, in HEK293T cells, overexpression of ORAI2 decreases Ca 2+ release and content, whereas its downregulation exerts an opposite effect [52].
In this work, we addressed two aspects of ORAI2 modulation of Ca 2+ homeostasis: (i) by using three types of model cells, we investigated the effects of ORAI2 levels (overexpression or downregulation) on the Ca 2+ content of intracellular stores and the modulation of SOCE; (ii) in the neuroglioma cell line H4-APPswe, we also investigated the effects of SOCE modulation on Aβ production.
Concerning the first aspect, we here provided evidence that modulation of ORAI2 levels tunes the amplitude of SOCE, with an inverse relationship between ORAI2 expression and Ca 2+ entry induced by store depletion. Furthermore, by using different approaches in intact and permeabilized cells, we demonstrated that ORAI2 overexpression decreases the store Ca 2+ content, independently of SOCE reduction or IP3-signalling. These findings are in part consistent with data previously reported in HEK-293T cells [52]. However, the co-expression of ORAI2 with an excess of STIM1, as reported in this study, allows us to exclude a dominant negative effect of ORAI2 on endogenous ORA1/STIM1 subunits. At high levels, ORAI2 could in fact sequester a substantial amount of STIM1, preventing effective STIM1-ORAI1 interactions at the PM [30,61].

Discussion
According to recent models, the store-operated Ca 2+ entry (SOCE) based on I CRAC is due to Ca 2+ permeation across hexameric channels formed by ORAI (1/2/3) subunits [28,30]. While the role of ORAI1 has widely been clarified, the contribution of the other subunits to I CRAC and SOCE emerged only later [28,33]. Thanks to in-depth studies on the immune system, it has elegantly been demonstrated that, in mouse T cells, ORAI2 forms heteromeric channels with ORAI1 and contributes to a reduced Ca 2+ permeation through I CRAC [35,39]. A negative role for ORAI2 on Ca 2+ entry was already postulated following overexpression studies in HEK293 cells [67]. In addition, Meyer and co-workers demonstrated that, in HEK293T cells, overexpression of ORAI2 decreases Ca 2+ release and content, whereas its downregulation exerts an opposite effect [52].
In this work, we addressed two aspects of ORAI2 modulation of Ca 2+ homeostasis: (i) by using three types of model cells, we investigated the effects of ORAI2 levels (overexpression or downregulation) on the Ca 2+ content of intracellular stores and the modulation of SOCE; (ii) in the neuroglioma cell line H4-APPswe, we also investigated the effects of SOCE modulation on Aβ production.
Concerning the first aspect, we here provided evidence that modulation of ORAI2 levels tunes the amplitude of SOCE, with an inverse relationship between ORAI2 expression and Ca 2+ entry induced by store depletion. Furthermore, by using different approaches in intact and permeabilized cells, we demonstrated that ORAI2 overexpression decreases the store Ca 2+ content, independently of SOCE reduction or IP 3 -signalling. These findings are in part consistent with data previously reported in HEK-293T cells [52]. However, the co-expression of ORAI2 with an excess of STIM1, as reported in this study, allows us to exclude a dominant negative effect of ORAI2 on endogenous ORA1/STIM1 subunits. At high levels, ORAI2 could in fact sequester a substantial amount of STIM1, preventing effective STIM1-ORAI1 interactions at the PM [30,61].
It should be noted that, at variance with HEK-293T cells [52], in neuroglioma cells, ORAI2 downregulation by siRNA leaves unchanged the release of Ca 2+ from intracellular stores. This last result is not consistent with the hypothesis that endogenous ORAI2 controls the store Ca 2+ content by modulating the ER Ca 2+ leak.
In order to try to solve this discrepancy we investigated the subcellular distribution of overexpressed ORAI2 using either a version containing only a myc-tag or one with mVenus-tag fused at the C-terminal. No difference between the two constructs was observed. The overexpression of ORAI2 in H4 cells resulted in the accumulation of the tagged proteins primarily at the PM level and early endosomes, with no significant overlapping with ER marker. Thus while in H4 and HeLa the overexpression of ORAI2 produced Ca 2+ effects similar to those reported in HEK293T cells, i.e., it reduced the store Ca 2+ content and increased the resting Ca 2+ level [52], downregulation of the protein left both parameters unchanged. At the moment, the reason for these discrepancies remains unexplained. Yet, the reduction in store Ca 2+ level and the slowness in Ca 2+ re-uptake, observed in permeabilized cells upon ORAI2 overexpression, together with the lack of ORAI2 colocalization with ER markers suggest that the effect of ORAI2 is likely indirect and not consistent with ORAI2 as an ER leak channel. Furthermore, upon ORAI2 downregulation, the lack of effect on the basal Ca 2+ level excludes a direct role of endogenous ORAI2 in controlling the resting Ca 2+ concentration, whereas the increase in resting cytosolic Ca 2+ level, observed upon overexpression, is consistent with the capability of this subunit to form homomeric functional PM Ca 2+ channels [35,67], as well as with SOCE activation by chronic reduction in the store Ca 2+ content, as originally suggested [52]. SOCE dysregulation has been widely reported among the complex defects that characterize Ca 2+ dyshomeostasis in AD [6][7][8][10][11][12][13][14][15][16][17][18][19][20][21][22]46,68,69]. In particular, whereas the role of FAD-linked PS1 and PS2 mutations at the store level was debated [18], the large majority of data converge towards SOCE downregulation, when studied either in cell lines expressing the PS1/2 mutants or in fibroblasts from FAD patients [6,[9][10][11][12][13]16,[19][20][21]46] as well as in SAD [10,70]. However, SOCE upregulation was also reported in 3xTg AD mice [14].
Over the last two decades, different groups have investigated the relationship between SOCE and Aβ production, with different and often divergent results [7,[23][24][25][26][27]. In this work, we took advantage of ORAI2 downregulation to study the relationship between SOCE amplitude and Aβ accumulation while avoiding alterations at the store Ca 2+ level. When this relationship was studied by employing SERCA pump inhibitors [25][26][27] or SERCA2b-siRNA [23] in the presence of extracellular Ca 2+ , a direct relationship between SOCE activation and Aβ42 accumulation was observed. Yet, it was also reported that pharmacological inhibition of SOCE increases Aβ42 accumulation [7]. Furthermore, SOCE activation by overexpression of ORAI1 or STIM1-D76A, a constitutively active form of STIM1, reduces Aβ42 production [24]. We here show that, in Aβ-secreting neuroglioma cells, overnight incubation with CPA or BTP2 (to activate or inhibit SOCE respectively) decreases and increases Aβ42 accumulation, being consistent with an inverse relationship between the two pathways [7,23,24].
The conflicting results obtained with SERCA pump inhibitors by different groups could have different explanations: (i) the use of different cell models; (ii) the different contribution to Ca 2+ rises played by Ca 2+ release and Ca 2+ entry in those cells and (iii) the different roles played by Ca 2+ on amyloid processing, Aβ production and secretion.
ORAI2 downregulation increases SOCE amplitude without modifying the store Ca 2+ content thus it permits to test how an increased Ca 2+ entry through SOCE influences Aβ secretion.
In H4-APPswe cells, downregulating ORAI2 by siRNA decreases Aβ42 but increases Aβ40 accumulation, significantly reducing the Aβ42/Aβ40 ratio. Altogether, these findings suggest that increased Ca 2+ entry through ORAI channels favours the accumulation of the less amyloidogenic peptide. It is worth noting that the APP Swedish mutation, by itself, does not change the Aβ42/Aβ40 ratio but simply increases the number of secreted peptides, because of its enhancement of β-secretase cleavage. For this reason, we can consider H4-APPswe cells an acceptable model of Aβ accumulation in SAD.
It is known that γ-secretase exists in a dynamic equilibrium of conformational states, with the "closed" conformation being associated with the shift of the enzyme cleavage towards the production of longer, neurotoxic Aβ species; of note, the shift to the pathogenic closed conformation is regulated by Ca 2+ [74,75]. In neurons, the Aβ42/Aβ40 ratio is strictly dependent on the pattern of spiking activity [74]. In particular, an increase in burst activity, increases cytosolic Ca 2+ and shifts PS1 towards the open conformation thus augmenting the Aβ40 production and reducing the Aβ42/Aβ40 ratio [74]. In this work, we provide evidence that increasing Ca 2+ levels by SOCE potentiation similarly favours Aβ40 with respect to Aβ42 secretion; a model explaining the possible linkage between SOCE potentiation, PS conformation, and Aβ production is presented in Figure 8. peptide. It is worth noting that the APP Swedish mutation, by itself, does not change the Aβ42/Aβ40 ratio but simply increases the number of secreted peptides, because of its enhancement of β-secretase cleavage. For this reason, we can consider H4-APPswe cells an acceptable model of Aβ accumulation in SAD. It is known that γ-secretase exists in a dynamic equilibrium of conformational states, with the "closed" conformation being associated with the shift of the enzyme cleavage towards the production of longer, neurotoxic Aβ species; of note, the shift to the pathogenic closed conformation is regulated by Ca 2+ [74,75]. In neurons, the Aβ42/Aβ40 ratio is strictly dependent on the pattern of spiking activity [74]. In particular, an increase in burst activity, increases cytosolic Ca 2+ and shifts PS1 towards the open conformation thus augmenting the Aβ40 production and reducing the Aβ42/Aβ40 ratio [74]. In this work, we provide evidence that increasing Ca 2+ levels by SOCE potentiation similarly favours Aβ40 with respect to Aβ42 secretion; a model explaining the possible linkage between SOCE potentiation, PS conformation, and Aβ production is presented in Figure 8. A sustained increased in Ca 2+ entry might also affect the endocytic pathway that controls PS maturation and APP processing through γ-secretase. It is worth noting that, in addition to its PM localization, ORAI2 is widely distributed in a vesicular network, partially overlapping with the endocytic pathway, thus suggesting a possible role of this protein in the regulation of APP processing at the endosomal level. Here, we demonstrated that ORAI2-myc and ORAI2-myc-mVenus are indistinguishable with respect to Ca 2+ handling by neuroglioma cells, thus providing a useful tool to investigate the relationship between APP processing and SOCE in intact cells.
In conclusion, here, we showed that, in Aβ-secreting neuroglioma cells, downregulation of ORAI2 potentiates SOCE without altering the store Ca 2+ content and decreases the Aβ42/Aβ40 ratio thus being an interesting tool to restore defective Ca 2+ entry associated to AD [9,10,54,76], while preventing amyloid seeding, given the direct correlation of this latter with the Aβ42/Aβ40 ratio. A sustained increased in Ca 2+ entry might also affect the endocytic pathway that controls PS maturation and APP processing through γ-secretase. It is worth noting that, in addition to its PM localization, ORAI2 is widely distributed in a vesicular network, partially overlapping with the endocytic pathway, thus suggesting a possible role of this protein in the regulation of APP processing at the endosomal level. Here, we demonstrated that ORAI2-myc and ORAI2-myc-mVenus are indistinguishable with respect to Ca 2+ handling by neuroglioma cells, thus providing a useful tool to investigate the relationship between APP processing and SOCE in intact cells.
In conclusion, here, we showed that, in Aβ-secreting neuroglioma cells, downregulation of ORAI2 potentiates SOCE without altering the store Ca 2+ content and decreases the Aβ42/Aβ40 ratio thus being an interesting tool to restore defective Ca 2+ entry associated to AD [9,10,54,76], while preventing amyloid seeding, given the direct correlation of this latter with the Aβ42/Aβ40 ratio. Altogether, our data support the idea that SOCE, and particularly ORAI2, could be a potential therapeutic target in AD.

Ca 2+ Imaging
HeLa cells expressing G-CEPIA1er fluorescent probe and H4-APPswe cells expressing H2B-GCaMP6 fluorescent probe were bathed in mKRB containing CaCl 2 (1 mM) at 37 • C. When indicated, cells were bathed in a Ca 2+ -free mKRB containing EGTA (0.6 mM) or in a Ca 2+ -free intracellular solution also containing EGTA (0.05 mM) of the following composition in mM (130 KCl, 10 NaCl, 1 MgCl 2 , 2 succinic acid and 20 HEPES, pH 7.0 at 37 • C). Cells were analyzed using an inverted microscope (Zeiss Axiovert 100, Jena, Germany) with a Fluar 40X oil objective (NA 1.30). Excitation light (480 nm for G-CEPIA1er, 410/10 nm and 475/10 nm for H2B-GCaMP6) was produced by a monochromator (Polychrome V; TILL Photonics, Graefelting, Germany) and passed through a dichroic mirror DRLP (505 ext XF73). The emitted fluorescence was collected by a bandpass filter (500-530 nm). Images were acquired using a cooled CCD camera (SensiCam QE, PCO, Kelheim, Germany). All filters and dichroic mirrors were from Chroma Technologies (Bellow Falls, VT, USA). Images were collected at 0.2 Hz with 200 ms exposure time. Cells were mounted into an open-topped chamber and maintained in mKRB medium. Additions were made in the same or different medium, as specified in the figures. G-CEPIA1er experiments: data are presented as ∆F / F 0 or ∆F / F min values (where ∆F are fluorescence changes at any time, F min is the fluorescence value upon ionomycin addition and F 0 the baseline value). H2B-GCaMP6 experiments: data are presented as ∆R values (where ∆R is the change of the F 470 / F 410 emission intensity ratio at any time.