STIM1 Deficiency Leads to Specific Down-Regulation of ITPR3 in SH-SY5Y Cells

STIM1 is an endoplasmic reticulum (ER) protein that modulates the activity of a number of Ca2+ transport systems. By direct physical interaction with ORAI1, a plasma membrane Ca2+ channel, STIM1 activates the ICRAC current, whereas the binding with the voltage-operated Ca2+ channel CaV1.2 inhibits the current through this latter channel. In this way, STIM1 is a key regulator of Ca2+ signaling in excitable and non-excitable cells, and altered STIM1 levels have been reported to underlie several pathologies, including immunodeficiency, neurodegenerative diseases, and cancer. In both sporadic and familial Alzheimer’s disease, a decrease of STIM1 protein levels accounts for the alteration of Ca2+ handling that compromises neuronal cell viability. Using SH-SY5Y cells edited by CRISPR/Cas9 to knockout STIM1 gene expression, this work evaluated the molecular mechanisms underlying the cell death triggered by the deficiency of STIM1, demonstrating that STIM1 is a positive regulator of ITPR3 gene expression. ITPR3 (or IP3R3) is a Ca2+ channel enriched at ER-mitochondria contact sites where it provides Ca2+ for transport into the mitochondria. Thus, STIM1 deficiency leads to a strong reduction of ITPR3 transcript and ITPR3 protein levels, a consequent decrease of the mitochondria free Ca2+ concentration ([Ca2+]mit), reduction of mitochondrial oxygen consumption rate, and decrease in ATP synthesis rate. All these values were normalized by ectopic expression of ITPR3 in STIM1-KO cells, providing strong evidence for a new mode of regulation of [Ca2+]mit mediated by the STIM1-ITPR3 axis.


Introduction
Calcium (Ca 2+ ) transport through biological membranes is essential to control a wide range of intracellular signaling pathways. This transport is tightly regulated to maintain a cytosolic free Ca 2+ concentration ([Ca 2+ ] i ) close to 100 nM in resting conditions, and the transient increase over this basal level acts as an effector for many stimuli. There are multiple signaling pathways that become deregulated during cytosolic Ca 2+ dyshomeostasis, which is the basis for the proposal of the Ca 2+

Expression of ITPR3 Is Modulated by STIM1 Protein Levels
As stated above, [Ca 2+ ] mit is critical to support mitochondrial function. We recently reported that human brain samples from SAD patients present significant decreases in the levels of STIM1 [24]. Using SH-SY5Y cells as an in vitro model system, we showed that this decline in STIM1 leads to reductions in [Ca 2+ ] mit and the inhibition of mitochondrial complex I (NADH-coenzyme Q oxidoreductase) activity. However, the molecular mechanism underlying these alterations is not clear.
Impairments in the transport of Ca 2+ into mitochondria could be caused by alterations in the levels or the activity of a wide range of Ca 2+ transporters and signaling molecules. To find the regulators responsible for this impairment, we monitored the expression of different transcripts included in the Human Intracellular Calcium Signaling TaqMan Gene Expression Assay (Supplemental File 1). The results revealed a significant drop in the expression of the ITPR3 gene in STIM1-KO SH-SY5Y cells compared to controls (Supplemental File 1 and Figure 1a).
It is well-known that the efficiency of Ca 2+ shuttling between ER and mitochondria relies on the activity of inositol 1,4,5-trisphosphate receptors (IP3Rs or ITPRs) [27][28][29]. We, therefore, studied in depth the possibility that the low [Ca 2+ ] mit found in STIM1-KO cells was caused by reductions in the expression levels of ITPR3. To confirm the TaqMan results, we analyzed mRNA levels for ITPR1, ITPR2, and ITPR3 genes by quantitative RT-PCR in non-differentiated SH-SY5Y cells. In agreement with the results shown above, we found~90% reduction in ITPR3 gene transcripts, whereas the products of ITPR1 and ITPR2 did not change significantly ( Figure 1b). Remarkably, overexpression of STIM1 in SH-SY5Y cells ( Figure A1, top panel) significantly increased the level of ITPR3 transcripts, without altering the level of ITPR1-2 ( Figure A1, bottom panel), which suggests a positive correlation between the expression of STIM1 and the levels of ITPR3. Immunoblot analyses confirmed that there was a significant decrease of ITPR3 protein in STIM1-KO cells, which was <30% of that found in the parental wild-type cell line (Figure 1c). A similar result was observed in differentiated SH-SY5Y cells ( Figure A2). On the contrary, the level of ITPR1 and ITPR2 proteins were not significantly altered by the absence of STIM1 (Figure 1d,e), as it was shown for ITPR1/2 transcripts (Figure 1b).
The decline of ITPR3 expression could have impact on the overall regulation of Ca 2+ signaling in the cell. However, the comparison of ITPR3 expression between SH-SY5Y and U2OS cells, where ITPR3 is the most abundant isoform, suggests that this specific variant is weakly expressed in SH-SY5Y cells (Figure 1f, left panel), as shown before [30]. In addition, the use of an antibody that recognizes the three isoforms did not show any difference between wild-type SH-SY5Y and STIM1-KO SH-SY5Y cells (Figure 1f, right panel), further confirming that the expression of ITPR3 represents a minor contribution to the total level of ITPRs in SH-SY5Y cells.
The importance of ITPR3 relies on the fact that this specific isoform was found to be enriched at the interactions between ER and mitochondria (mitochondria-associated ER membranes (MAMs)) in different cell lines [27][28][29], where it regulates the transfer of Ca 2+ between these organelles. This specific localization of ITPR3 in MAMs was also found in SH-SY5Y cells in this work ( Figure A3).

Functional Consequences of the ITPR3 Downregulation
To study the functional consequences of the ITPR3 downregulation in STIM1-deficient cells, we assessed the release of Ca 2+ from the ER in fura-2-loaded cells stimulated with 100 µM carbachol (CCh) in Ca 2+ -free medium. This stimulation, which activates the IP3 pathway, revealed a partial decrease in the amplitude of the Ca 2+ release from the ER in response to a 30 s pulse of CCh ( Figure 2a). Significantly, we were able to recapitulate these results in differentiated cells (Figure 2b). To study the functional consequences of the ITPR3 downregulation in STIM1-deficient cells, we assessed the release of Ca 2+ from the ER in fura-2-loaded cells stimulated with 100 μM carbachol (CCh) in Ca 2+ -free medium. This stimulation, which activates the IP3 pathway, revealed a partial decrease in the amplitude of the Ca 2+ release from the ER in response to a 30 s pulse of CCh ( Figure  2a). Significantly, we were able to recapitulate these results in differentiated cells (Figure 2b).  Alterations in the gradient of Ca 2+ between the intraluminal space of the ER and the cytosol could also result in reductions in the release of Ca 2+ from the ER. To test for this possibility, we measured the intraluminal free Ca 2+ concentration in the ER ([Ca 2+ ] ER ) in resting conditions by transfecting the genetically-encoded Ca 2+ sensor ER-GCaMP6-210 [31]. Our results showed there to be a statistically significant but slight decrease of the [Ca 2+ ] ER in STIM1-deficient cells (from 246 ± 10.8 µM in wild-type cells to 202 ± 11.7 µM in STIM1-KO cells) (Figure 3a). However, it is important to note here that the measured cytosolic Ca 2+ concentration ([Ca 2+ ] i ) in wild-type cells was greater than that found in STIM1-KO cells (77.4 ± 10 nM vs 44 ± 4.4 nM; see data in Reference [24]), so that the gradient of the [Ca 2+ ] across the ER membrane differed little between wild-type and STIM1-KO cells and cannot explain the diminished Ca 2+ release in STIM1-deficient cells in response to CCh. Therefore, we analyzed the speed of ER emptying by measuring the time constant tau (τ) of the kinetics of [Ca 2+ ] ER in response to CCh. This ER emptying follows a first-order exponential decay (Figure 3b), with the τ-value of this kinetics being significantly increased by the reduction in STIM1, rising from 10.5 ± 0.48 s (for wild-type cells) to 14.29 ± 0.94 s (for STIM1-KO cells, and n = 26 regions of interest, ROIs, in each experimental condition). In agreement with the reductions in ITPR3 levels in STIM1-KO cells shown in Figure 1, these data support a slower Ca 2+ release from the ER in STIM1-KO cells. These results, which illustrate a deficient fast release of Ca 2+ from the ER in STIM1-deficient cells, are of significant importance considering the localization of ITPR3, which were found highly enriched in MAMs ( Figure A3). kinetics of [Ca 2+ ]ER in response to CCh. This ER emptying follows a first-order exponential decay (Figure 3b), with the τ-value of this kinetics being significantly increased by the reduction in STIM1, rising from 10.5 ± 0.48 s (for wild-type cells) to 14.29 ± 0.94 s (for STIM1-KO cells, and n = 26 regions of interest, ROIs, in each experimental condition). In agreement with the reductions in ITPR3 levels in STIM1-KO cells shown in Figure 1, these data support a slower Ca 2+ release from the ER in STIM1-KO cells. These results, which illustrate a deficient fast release of Ca 2+ from the ER in STIM1-deficient cells, are of significant importance considering the localization of ITPR3, which were found highly enriched in MAMs ( Figure A3).

Overexpression of ITPR3 Restores [Ca 2+ ] mit in STIM1-KO Cells
To investigate whether deficiency in STIM1 and the consequent lowering of ITPR3 levels modified the regulation of the mitochondrial Ca 2+ buffering capacity, we measured the kinetics of the mitochondrial Ca 2+ concentration in SH-SY5Y cells, expressing the Ca 2+ sensor mito 4× -GCaMP6f, and stimulated with ionomycin following a previously published approach [32]. We stimulated wild-type and STIM1-KO cells, cultured in fetal bovine serum (FBS)-containing medium, with ionomycin and measured the extent of the increase in the fluorescence intensity relative to baseline (F/F0) of the GCaMP6-210, which then returned back to basal levels ( Figure 4). The dynamics of the kinetics of F/F0 was similar in both wild-type and STIM1-KO cells (Figure 4b,c), and the maximal F/F0 values did not show any alteration in the absence of STIM1 (Figure 4d). Because the speed of mitochondrial Ca 2+ import was not affected by the absence of STIM1 (Figure 4e), these data suggest that the mitochondrial maximal efficiency to import Ca 2+ was not significantly altered by the lack of STIM1. In addition, no significant differences were observed in total levels of the mitochondrial Ca 2+ uniporter (MCU) between wild-type and STIM1-KO cells, nor in the expression of the Grp75 (75 kDa glucose-regulated protein) and VDAC1 (voltage-dependent anion channel 1) (Figure 4f), two proteins that can associate with ITPR3 to facilitate the Ca 2+ transfer between ER and mitochondria.
However, the steady-state [Ca 2+ ] mit was markedly decreased in STIM1-KO (Figure 4g), also reported previously by our group [24]. Notably, the wild-type phenotype was rescued by the overexpression of ITPR3 tagged with mCherry in STIM1-KO cells ( Figure 4g). The level of ITPR3-mCherry compared to the endogenous ITPR3 is shown in Figure A4. As a control of this experiment, we assessed the [Ca 2+ ] mit in cells expressing the mCherry tag only (Figure 4h), an experiment that proved that the expression of this tag did not modify the rescue of the [Ca 2+ ] mit by the overexpression of ITPR3, as shown in Figure 4g.
We were able to recapitulate this phenotype in STIM1-KO cell lines stably overexpressing ITPR3-Myc (Figure 4i,j), where we assessed the [Ca 2+ ] mit of two positive clones (labeled as clones #29 and #57). Our data confirmed that stable overexpression of ITPR3 rescued the defects in [Ca 2+ ] mit triggered by the deletion of STIM1, which strongly suggests that ITPR3 deficiency was the cause of [Ca 2+ ] mit dysregulation in STIM1-KO cells.

Overexpression of ITPR3 Normalizes Mitochondrial Function
To understand whether the expression of ITPR3 could also rescue mitochondrial fitness in STIM1-KO cells, we measured the oxygen consumption rate (OCR) sensitive to oligomycin, carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP), and rotenone + antimycin A, as a readout of the mitochondrial functionality and metabolic status ( Figure 5). The metabolic assay indicated that basal respiration was lower in STIM1-KO cells, and that most of the O 2 consumption was non-mitochondrial ( Figure 5a). Indeed, the coupling efficiency of the electron transport and ATP synthesis was significantly lower in STIM1-deficient cells, as well as the overall ATP production.
Remarkably, the stable overexpression of ITPR3, which partially normalized [Ca 2+ ] mit in STIM1-KO cells, normalized basal oxygen consumption, coupling efficiency, and ATP production (Figure 5a), further confirming that ITPR3 is an essential regulator of mitochondrial function and suggesting that variations in ITPR3 levels constitute a major hallmark for STIM1-deficient cells. This latter conclusion was strengthened by the fact that the basal cell death observed in STIM1-KO cell cultures was prevented by the stable overexpression of ITPR3 in these cells (Figure 5b).  Remarkably, the stable overexpression of ITPR3, which partially normalized [Ca 2+ ]mit in STIM1-KO cells, normalized basal oxygen consumption, coupling efficiency, and ATP production ( Figure  5a), further confirming that ITPR3 is an essential regulator of mitochondrial function and suggesting that variations in ITPR3 levels constitute a major hallmark for STIM1-deficient cells. This latter

Discussion
As a major Ca 2+ signaling regulator, STIM1 is involved in a number of pathologies, including neurodegenerative diseases, severe immunodeficiency, and cancer [33][34][35][36]. In primary cancer samples, as well as in cancer cell lines, the total amount of STIM1 correlates with augmented migration and proliferation rates, and consequently with poor prognosis (reviewed in Reference [37]). Conversely, the decline in total STIM1 levels in brain tissue of Alzheimer's disease patients correlates with the Braak stage [24], again suggesting a role for STIM1 in cell survival.
In order to understand the molecular mechanism by which STIM1 regulates cell viability in excitable cells, we have examined here STIM1-KO SH-SY5Y cells as an in vitro cellular model. In this system, the deficiency in STIM1 does not affect differentiation to a neuronal-like cell type, but it does trigger a significant loss of cell viability [24]. Using this system, we searched for alterations in the expression of Ca 2+ regulators and transporters, finding a strong decrease in ITPR3 transcripts and ITPR3 protein expression in STIM1-deficient cells, as is reported here. ITPRs are most abundantly expressed in the ER, but they are also found in the nuclear envelope, Golgi apparatus, secretory vesicles, and plasma membrane (reviewed in References [38,39]). The significant decrease in ITPR3 is particularly important since an enrichment of ITPR3 in mitochondrial-associated membranes from SH-SY5Y cells is reported in this work. Similar distribution of ITPR3 has been reported for other cells lines [27,29,40], and it is assumed that this localization in MAMs promotes the shuttling of Ca 2+ between ER and mitochondria [41,42].
ER-mitochondria Ca 2+ transfer is essential for mitochondrial function because Ca 2+ positively regulates three enzymatic reactions of the Krebs cycle (isocitrate dehydrogenase, α-ketoglutarate dehydrogenase, and pyruvate dehydrogenase) [43], stimulating the supply of NADH, and therefore, normalizing the mitochondrial electron transport chain. In this regard, it has been reported that the deficiency of Ca 2+ transfer from the ER to the mitochondria triggers autophagy as a result of the failure in the energy production [44]. On the contrary, mitochondrial Ca 2+ overload has been related to the opening of the mitochondrial permeability transition pore, which ultimately leads to other types of cell death [45,46].
The need for a constitutive Ca 2+ transfer between ER and mitochondria has been reported in several human tumorigenic cell lines [47], and the relationship between ITPR3 expression or activity and cell survival is particularly critical in the case of cancer cells. For instance, in colon cancer, greater ITPR3 expression is associated with cancer cell proliferation and lower 5-year survival of patients. On the other hand, ITPR3 knock-down in Caco-2 colon cancer cells enhances apoptosis, while over-expression enhances cell survival [48]. Similarly, cholangiocarcinoma cells were shown to be particularly sensitive to the knock-out of ITPR3, a genetic manipulation that led to the reduction of cell migration, as well as to the reduction of mitochondrial oxygen consumption and proliferation [49]. The migration and invasion potential have been shown to be severely influenced by ITPR3 levels in other cancer cells. Whereas high levels of ITPR3 were observed in the highly migrating and invasive MDA-MB-231 and MDA-MB-435S breast cancer cell lines, the low-migrating MCF-7 showed low levels of ITPR3, but the stable overexpression of ITPR3 increased the migration capacity of this cell line [50]. In colorectal cancer DLD-1 cells, the silencing of ITPR3 expression led to a reduction of tumor volumes after injection of these cells in nude mice, increasing the apoptosis of these cells in hypoxic conditions and demonstrating the proliferative and anti-apoptotic role of ITPR3 [51]. Therefore, the direct correlation between STIM1 protein expression level and ITPR3 expression found here may explain the suggested correlation between STIM1 protein levels and proliferation and malignancy in cancer cells [37,52,53].
In the specific case of neuronal cells, some mutations in PSEN1/2 linked to Alzheimer's disease leads to the partial inhibition of presenilins/γ-secretase activity, a result that end up in upregulated MAM function and ER-mitochondria communication [54]. The functional consequence for this upregulation is the altered Ca 2+ homeostasis and altered cholesterol and phospholipid metabolism. However, in the absence of these mutations, like in the case reported here, the mitochondrial Ca 2+ homeostasis and bioenergetics were significantly altered by the absence of STIM1 and the subsequent reduction of ITPR3. It must be emphasized also that the ITPRs are positive regulators of MAMs and that the expression of any of these isoforms on triple KO cells (knock-out for the 3 ITPR variants) increased the number of ER-mitochondria contacts and that ITPR2 and ITPR3 were the two most effective isoforms engaging ER-mitochondria Ca 2+ transfer [55].
On the other hand, STIM2, a STIM1 paralogue, performs similar but not identical functions. For example, the dissociation constant of STIM2 for Ca 2+ is close to 0.5 mM [56], which is significantly higher than that of STIM1 which is around 0.25 mM [6]. Therefore, the function of STIM2 is closely related to the maintenance of a basal intraluminal Ca 2+ concentration, while STIM1 acts in response to more intense variations of this concentration. The transmembrane domain of both proteins is very similar, so it is possible that STIM2 is also a substrate of gamma-secretase activity, as suggested elsewhere [20]. In fact, it has been described that STIM2 abundance is reduced in cells expressing some FAD-associated variants of PSEN1 [23,57], and that STIM2 reduction destabilizes mature dendritic spines in mice bearing FAD-associated PSEN1 variants [57]. Therefore, it would be necessary to address the study of the role of STIM2 in this neuronal model (SH-SY5Y) by means of total gene suppression, as we have carried out here with CRISPR/Cas9-mediated gene editing for STIM1.
The present results in SH-SY5Y demonstrate that the reduced ITPR3 levels observed in STIM1-deficient cells are responsible for the reduced Ca 2+ release from the ER in response to the activation of the phosphoinositide pathway. Because the Ca 2+ import capability of mitochondria in STIM1-KO is not significantly deregulated, our results suggest that the decrease of ITPR3 is also responsible for the reduction in [Ca 2+ ] mit . This conclusion is supported by the normalization of [Ca 2+ ] mit , as well as the recovery of the basal OCR, ATP production, and cell viability, observed in STIM1-KO cells overexpressing ectopic ITPR3. For this reason, our results indicate that STIM1 has a key role in the normalization of [Ca 2+ ] mit by controlling ITPR3 protein levels. Interestingly, postmortem cerebella from patients with the PSEN1-E280A mutation, typical of FAD, showed much lower levels of ITPR3 compared to control patients, and the expression of PSEN1-E280A in a neuronal cell line altered ER-mitochondria tethering compared with that in cells expressing wild-type PSEN1 [58]. Because we also have reported a significant decrease of STIM1 in brain samples from SAD patients [24], the axis STIM1-ITPR3-[Ca 2+ ] mit must be considered to explain the modulation of neurodegeneration and survival.

DNA Constructs
The construct for the expression of ITPR3-mCherry was made by inserting the human ITPR3 cDNA (NM_002224.4) into the pmCherry-N1 vector using the cDNA from the Harvard plasmid clone HsCD00399229 as a template. The following primers were used to insert SmaI flanking sites (underlined): ITPR3-fwd: 5 -TCCCCCGGG-ATGAGTGAAATGTCCAGCTTTC, and ITPR3-rev: 5 -TCCCCCGGG-GGCGGCTAATGCAGTTCTGGAC. The construct for the expression of mouse Itpr3-Myc was purchased from OriGene (Rockville, MD, USA) (clone #MR225699).

Culture, Differentiation and Generation of Stable Cell Lines
Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% (v/v) heat-inactivated fetal bovine serum (FBS) (Ref. 10500-064 from ThermoFisher Scientific), 2 mM L-glutamine, 100 U/mL penicillin, and 0.1 mg/mL streptomycin in a humidified atmosphere of 95% air/5% CO 2 at 37 • C, as described elsewhere [60]. Cell culture dishes and glass coverslips were treated with 1.5 g/mL collagen type I solution (freshly diluted in Milli-Q water from the stock solution purchased from Sigma). Collagen treatment was extended for a minimum of 30 min at 37 • C in a humidified atmosphere to avoid drying, and then it was washed 3 times with PBS before cell plating. The differentiation of cells was performed as described previously [24]. Stable cell lines overexpressing ITPR3-Myc were generated by selection with G418 (0.5 mg/mL) for 7 days in culture, and subsequent analysis of individual clones.
The generation of the SH-SY5Y STIM1-KO cell line by genome editing with CRISPR/Cas9 is described elsewhere [24].

Lysis of Cells and Immunoblot
Cells were lysed in the following buffer: 50 mM Tris-HCl (pH 7.5), 1 mM EGTA, 1 mM EDTA, 1% (w/v) Nonidet P40, 1 mM sodium orthovanadate, 50 mM NaF, 5 mM sodium pyrophosphate, 0.27 M sucrose, 1 mM DTT, 1 mM benzamidine, and 0.1 mM phenylmethylsulfonyl fluoride. The last three reagents of this list were added from a stock solution just before the use of the lysis buffer. Clarification of samples was performed after lysis with 0.75-1 mL of ice-cold lysis buffer/dish and centrifugation at 4 • C for 15 min at 20,000× g. Samples were sonicated with 5 × 10 s pulses with a setting of 45% amplitude using a Branson Digital Sonifier (ThermoFisher Scientific). Protein concentration was determined using the Coomassie Protein Assay Reagent (ThermoFisher Scientific).

XF Cell Mito Stress Assay
Cell bioenergetics was evaluated in a Seahorse XFp analyzer (Agilent Technologies, Santa Clara, CA, USA). SH-SY5Y cells were plated on Seahorse XFp plates (55,000 cells/per well). After 24 h, cells were washed with XF DMEM medium followed by a 1 h incubation at 37 • C in a CO 2 -free incubator. Each plate contained two wells without cells to serve as blank controls. After monitoring basal respiration, cells were sequentially treated with 1.5 µM oligomycin, 1 µM FCCP, and 0.5 µM antimycin A + 0.5 µM rotenone (Seahorse XFp Cell Mito Stress kit, from Agilent Technologies). Oxygen consumption rate (OCR) data were normalized to the total cell amount per well estimated by Janus Green staining [61]. Briefly, after bioenergetics analysis, cells were fixed in 4% paraformaldehyde (PFA) and incubated with 0.2% Janus Green B in PBS for 3 min at room temperature. The excess of dye was removed by dipping into cold water and gentle shaking, and the bound dye was dissolved in 0.5 N HCl (0.1 mL/well) and the optical density at 595 nm was evaluated. The metabolic parameters of the assay were calculated with the Seahorse Wave software. In all cases, 3 technical replicates were recorded in every experiment, and the experiments were carried out with 3 biological replicates.

Isolation of Mitochondria-Associated ER Membranes
The isolation of MAMs was performed as described previously [62]. Briefly, cells from 50-60 10-cm dishes were scrapped in cold PBS and concentrated by centrifugation at 3000× g for 15 min. Cells were then washed in the following homogenization buffer (HB): 225 mM mannitol, 25 mM HEPES-KOH (pH = 7.4), 1 mM EGTA, and protease inhibitor cocktail. Cells were homogenized with 8 strokes in a Teflon-pestle grinder. The homogenate was clarified with 2 sequential centrifugations at 600× g for 10 min. The resulting supernatant was then centrifuged at 10,000× g to pellet the mitochondrial fraction. This pellet was resuspended in 1 mL of HB and loaded onto a freshly prepared 30% Percoll solution in HB and then centrifuged at 95,000× g for 30 min. The lower-density upper band was collected with the tip of a Pasteur pipette, diluted in 5 volumes of HB, vortexed for 15 s, and centrifuged twice at 8000× g for 10 min to remove the contaminant Percoll. The supernatant was centrifuged for 1 h at 100,000× g and the floating clear membrane pellet was collected as the pure MAMs fraction. All centrifugations were carried out at 4 • C. Other details are described in Reference [62].
To study maximal Ca 2+ uptake by mitochondria, [Ca 2+ ] mit dynamics were assessed in cells transfected with the genetically-encoded Ca 2+ sensor mito 4× -GCaMP6f, due to its improved dynamic range, which results in a much better signal-to-noise ratio when detecting peak responses, as described in Reference [68]. This probe was designed to express 4 consecutive copies of the signal peptide of COX8 (MSVLTPLLLRGLTGSARRLPVPRAKIHSLGDP) and a short linker (RSGSAKDPT) before the sequence of GCaMP6f [68]. The main advantage of this probe is that it shows increased response rate compared to 4mtD3cpv, allowing the fast monitoring of the Ca 2+ uptake by mitochondria. All measurements were performed in L15 medium + 10% FBS and temperature was controlled and set to 36 • C. Excitation fluorescence wavelengths were selected with a 485/10 filter (Semrock), and emission fluorescence with a 535/20 nm filter.
Endoplasmic reticulum free calcium concentration ([Ca 2+ ] ER ) was measured using the geneticallyencoded Ca 2+ sensor ER-GCaMP6-210 [31]. Transfected cells were monitored using the following settings: excitation fluorescence wavelengths were selected with 480/30 nm filters (Semrock), and emission fluorescence with a 535/40 nm filter. The emission of fluorescence was calibrated with the subsequent addition of 5 µM ionomycin + 5 mM EGTA followed by the addition of 5 µM ionomycin + 10 mM Ca 2+ .
All measurements were performed using an EM-CCD C9100 digital camera (Hamamatsu Photonics, Hamamatsu City, Japan) attached to a Nikon Eclipse Ti-E inverted microscope (Nikon Instruments Europe B.V., Amsterdam, The Netherlands). Illumination was performed with a xenon arc lamp. All measurements were performed in a DH-35iL culture dish incubator, and the temperature was set at 36 • C (Warner Instruments, Holliston, MA, USA). In all cases, excitation and emission conditions were controlled by the NIS-Elements AR software.

Cell Death Analysis
Cells were plated on 8-well µslide plates from Ibidi (Gräfelfing, Germany) at 40,000 cells/well density. Forty-eight hours after plating, cells were stained with the Live/Dead Cell Imaging kit (ThermoFisher Scientific), according to the manufacturer protocol. Ten images from randomly chosen fields were recorded for every experimental condition. Imaging was performed with a 20× objective on a Nikon Eclipse Ti-E inverted microscope.

Statistical Analysis of Data
Statistical analyses between pairs of data groups were done using the Mann-Whitney test of data (non-parametric unpaired t-test). Analyses were performed with the GraphPad software. Differences between groups of data were taken statistically significant for p < 0.05. The p-values are represented as follows: (*) p < 0.05, (**) p < 0.01, and (***) p < 0.001. Funding: This research was funded by the Spanish Ministerio de Ciencia, Innovación y Universidades (grant BFU2017-82716-P), and Junta de Extremadura (grants GR18084 and IB16088). All grants were co-financed by FEDER funds. CPC was supported by a predoctoral fellowship from the Spanish Ministerio de Educación, Cultura y Deporte (FPU13/03430). YOA and ISL were supported by predoctoral fellowships from the University of Extremadura.

Conflicts of Interest:
The authors declare no conflict of interest.     , and ITPR3 transcripts from differentiated STIM1-KO cells compared to wild-type cells. Data are mean ± S.D. from 2 different biological replicates with technical triplicates (n = 6). (b) ITPR3 protein expression in whole cell lysates was quantified in differentiated wild-type and STIM1-KO cells (betatubulin as a loading control). Data are mean ± S.D. from 2 independent experiments. Figure A3. The isolation of mitochondria-associated ER membranes (MAMs), ER, and mitochondria from undifferentiated SH-SY5Y cells revealed that ITPR3 was enriched in MAMs, compared to the bulk ER. As markers to assess the purity of the fractions we studied the level of ACSL4 (MAMs), Grp75, VDAC1, and MCU (mitochondria), and calreticulin (ER). Figure A3. The isolation of mitochondria-associated ER membranes (MAMs), ER, and mitochondria from undifferentiated SH-SY5Y cells revealed that ITPR3 was enriched in MAMs, compared to the bulk ER. As markers to assess the purity of the fractions we studied the level of ACSL4 (MAMs), Grp75, VDAC1, and MCU (mitochondria), and calreticulin (ER).  Figure A4. Undifferentiated wild-type and STIM1-KO cells were transfected for the transient expression of ITPR3-mCherry, which was evaluated with a specific antibody against ITPR3 and an antibody against mCherry (Red Fluorescent Protein, RFP).