Stromal Interaction Molecule 1-Mediated Store-Operated Calcium Entry Promotes Autophagy Through AKT/Mammalian Target of Rapamycin Pathway in Hippocampal Neurons After Ischemic Stroke

—The pathophysiological process of neuronal injury due to cerebral ischemia is complex among which disturbance of calcium homeostasis and autophagy are two major pathogenesis. However, it remains ambiguous whether the two factors are independent. Stromal interaction molecule 1 (STIM1) is the most important Ca 2+ sensor mediating the store-operated Ca 2+ entry (SOCE) through interacting with Orai1 and has recently been proven to participate in autophagy in multiple cells. In this study, we aimed to investigate the potential role of STIM1-induced SOCE on autophagy and whether its regulator function contributes to neuronal injury under hypoxic conditions using in vivo transient middle cerebral artery occlusion (tMCAO) model and in vitro oxygen and glucose deprivation (OGD) primary cultured neuron model respectively. The present data indicated that STIM1 induces autophagic ﬂux impairment in neurons through promoting SOCE and inhibiting AKT/mTOR signaling pathway. Pharmacological inhibition of SOCE or downregulation of STIM1 with siRNA suppressed the autophagic activity in neurons. Moreover, stim1 knockdown attenuated neurological deﬁcits and brain damage after tMCAO, which could be reversed by AKT/mTOR pathway inhibitor AZD5363. Together, the modulation of STIM1 on autophagic activation indicated the potential link between Ca 2+ homeostasis and autophagy which provided evidence that STIM1 could be a promising therapeutic target for ischemic stroke. (cid:1) 2023 The Author(s). Published by Elsevier Ltd on behalf of IBRO. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).


INTRODUCTION
Ischemic stroke is the most common type of stroke and remains the second leading cause of mortality and morbidity worldwide (Seiffge et al., 2019). Recent studies suggested that autophagy is activated during ischemic stroke (Ajoolabady et al., 2021). Autophagy is critical for the degradation and recycling of cellular components, and is important in maintaining cellular homeostasis and cell survival in response to stress (Balduini et al., 2009). However, the role of autophagy in neuronal death remains controversial, and some studies suggested that excessive autophagy may aggravate ischemic brain damage . Therefore, further studies are needed to investigate how autophagy is activated in ischemic stroke and its role in neuronal injuries.
STIM1 is a single-pass transmembrane protein, which is localized predominantly in the endoplasmic reticulum (ER)  homeostasis has been proved to be associated with excessive activation of autophagy and cell death in dendritic cells , colorectal cancer cells , muscle cells, and neuron (Lemmer et al., 2021). STIM1 serves as the sensor and triggers SOCE through the plasma membrane channel ORAI, which plays an important role in multiple cellular pathophysiological processes including ER stress, inflammation, and apoptosis (Lunz et al., 2019;Di Buduo et al., 2020). It has been recently shown that blocking STIM1 with an ER-associated protein STING (stimulator of interferon genes) reduces autophagy in the mouse lupus model (Prabakaran et al., 2021). Moreover, in acute pancreatitis the activation of SOCE upregulated autophagy-related genes, including transcription factor EB (TFEB) which is a master transcriptional regulator of autophagy (Zhu et al., 2018). In addition, STIM1 deficiency attenuated oxygenated low-density lipoprotein (ox-LDL)-induced autophagy flux through the calcium/calmodulin dependent protein kinase 2 (CAMKK2)-mammalian target of rapamycin (mTOR) pathway in endothelial progenitor cells and inhibited their proliferation . Together, these studies supported a critical role of STIM1mediated SOCE in inducing autophagy. Our recent studies have demonstrated that the activation of autophagy attenuates neuronal apoptosis after ischemic stroke (Dai et al., 2017), however, it remains unknown whether STIM1 plays a role in ischemic neuronal autophagy.
In the present study, we investigated the role of STIM1 in the hippocampal regions which are known to be highly vulnerable to ischemic damage and may be salvageable in the early phase of stroke (Sanderson and Wider, 2013;Lu et al., 2022). We examined its role in regulating autophagy and whether they are involved in hypoxia-induced neuronal injury both in vivo and in vitro. Our data demonstrated that STIM1 induces autophagic flux in neurons through SOCE. We further showed that this effect was mediated by the AKT/mTOR signaling pathway. Our results suggested that the STIM1-regulated autophagy pathway may be a potential therapeutic target for ischemic stroke.

EXPERIMENTAL PROCEDURES
All the animal experiments were approved by the Institutional Animal Care and Use Committee of the Fourth Military Medical University (Xi'an, China, No. 2014-81371447). All experimental protocols and animal handling procedures were performed according to the National Institutes of Health (NIH) guidelines for the use of experimental animals.

Mouse model of transient middle cerebral artery occlusion (tMCAO)
Seventy-two C57BL/6J mice (male, 8-12 week-old) were housed under controlled environmental conditions with a consistent temperature of 22-25°C, a relative humidity of 65%, and 12 h light/dark cycle. All the mice had free access to food and water.
tMCAO model was created according to previously published methods (Denorme et al., 2022). Mice were anesthetized with 5% isoflurane and maintained at 2% isoflurane for the duration of the surgery. The right external carotid artery was isolated and inserted with a standardized nylon filament (Doccol Corporation, US) to the origin of the right MCA through the right internal carotid artery. After 90 min, the filament was removed to allow reperfusion. Cortical cerebral blood flow (CBF) was recorded by laser Doppler flowmetry (Moor Instruments, UK), and the reduction to 20% of the baseline was considered successful occlusion. Mice in the Sham group underwent the same procedure except for filament insertion. Carprofen (5 mg/kg) was subcutaneously injected right after surgery. Body temperature was maintained at 37°C with a heating pad during the surgery and resuscitation. All the mice were sacrificed at 24 h after tMCAO.

Recombinant adeno-associated virus (AAV) and siRNAs
According to the previously published articles, adenoassociated virus (AAV, serotype 9) was the most effective serotype for the transduction of RNA into neurons, therefore we used the virus to genetically modify the expression of STIM1 (Lu et al., 2021). The interference short hairpin RNA (shRNA) for mouse Stim1 (AGAAGGAGCTAGAATCTCAC) and non-silencing negative control (NC, GACGCTGAAGACTTGGC) were packaged into AAV9 (Genechem Corp., Ltd, Shanghai, China) and used for in vivo studies. siRNA for mouse Stim1 (GCAGGTAGGAGAGTGTACAGGATGCCTT) and negative control siRNA (UUCUCCGAACGUGUCAC-GUTT) were synthesized by Genechem Corp. Ltd. For overexpression of stim1, the primer for cloning mouse stim1 gene was designed based on NCBI gene sequence (NM_003156.3), synthesized and packed in a AAV9 plasmid by Sangon Biotech Co., Ltd (Shanghai, China).

In vivo and in vitro knockdown of Stim1
To knockdown Stim1 in moue brain, AAV-shStim-1 or AAV-NC (as control) (2 Â 10 12 v.g./ml) were injected into the bilateral hippocampus at a rate of 0.2 ll/min with a total volume of 1 ll according to the following coordinates: 2.1 mm anterior, ±1.8 mm lateral and 1.8 mm ventral. For in vitro study, cultured neurons were infected with AAV-NC or AAV-Stim1 at the dosage of 10 6 v.g./cell number in serum-free DMEM for 24 h, and or cells were transfected with siRNA using Lipofectamine 3000 (Thermo Fisher Scientific) according to the manufacturer's protocol.

Neurological deficit evaluation
Clark score was performed to evaluate the neurological deficits (Sardari et al., 2021). The score consisted of focal and general parts. The general part contained 6 items with the total score ranging from 0-28 points. The focal part contained 7 items with 0-4 points for each item.
Higher scores indicate greater neurological deficits.
Infarct volume and brain edema measurement 2,3,5-Triphenyltetrazolium hydrochloride (TTC) staining was performed to evaluate the infarct volume. Mouse brains were removed immediately after sacrifice and sectioned into five coronal slices at a 2-mm thickness (model No. 68707, RWD Life Science). The brain slices were stained with 2% TTC (Sigma-Aldrich, US) at 37°C in the dark for 30 min. Stained brain sections were fixed in 2% paraformaldehyde and photographed. The percentage of infarct volume (white) was analyzed using Image J software. The brain tissues were weighed to obtain the wet weight (WW) which was followed by drying at 110°C for 24 h to determine their dry weight (DW). Brain edema was calculated using the following formula: (WW À DW)/ WW Â 100%.

Primary culture of hippocampal neurons
Hippocampal neurons were prepared from C57BL/6 mice according to a previously published method (Picon et al., 2021). Cerebral cortices were collected from 16-18-day-old mouse embryos, placed in ice-cold medium (HBSS plus 10 mM HEPES, pH = 7.3), stripped of meninges and blood vessels, and minced with scissors. The tissues were digested with 0.25% trypsin at 37°C for 15 min with gentle trituration. Neurons were resuspended in a neurobasal medium containing 10% FBS (Invitrogen), 2% B27 supplement (Invitrogen), 1% penicillin/streptomycin (Invitrogen), and 0.5 mM L-Glutamine (Invitrogen), and then plated at a density of 3 Â 10 5 cells/cm 2 in Poly-Dlysine (10 mg/ml, Sigma-Aldrich)coated 6-well-plates. Cells were maintained at 37°C in a humidified incubator (model No. 371, Thermo Scientific) filled with 5% CO 2 /air. Half of the culture medium was replaced with fresh medium every other day. Cells were maintained in culture for 10 days before they were used for experiments. Quantification of STIM1 levels relative to b-actin (n = 3, Student t' test). Immunohistostaining of STIM1 (Red) and NeuN (Green) in CA1 (C) and CA3 (E) regions respectively, Scale bar = 50 lm. Quantification of STIM1 intensity in CA1 (D, n = 6, Student t' test) and CA3 (F, n = 6, Student t' test) regions. (G) TTC staining indicated the infarct tissue (white) and noninfarct tissue (red) of the brains at 24 h after tMCAO. The percentage of infarct volume (H, n = 6, one-way ANOVA, F 3,20 = 190.8) and brain edema (I, n = 6, one-way ANOVA, F 3,20 = 113.7) was calculated. Statistical analysis was performed using one-way ANOVA followed by post hoc Tukey's test for multiple comparisons and Student t' test between two groups. *P < 0.05 compared with Sham group, # P < 0.05 compared with tMCAO group.

Cellular model of oxygen and glucose deprivation (OGD)
The culture medium was removed, and cells were rinsed with phosphate-buffered saline (PBS) for three times. Cell cultures were then added with glucose-free DMEM, which was pre-gassed with 95%N 2 /5%CO 2 to remove residual oxygen, and placed in a specialized, humidified chamber filled with 95%N 2 /5%CO 2 at 37°C. At different timepoints (2, 4, 8 12, and 24 h), the cells were removed from the anaerobic chamber and the culture medium was replaced with a neurobasal medium containing a 2% B27 supplement and 0.5 mM L-Glutamine. Neurons were further incubated in a humidified 5% CO 2 /air incubator for 24 h at 37°C to mimic the reperfusion insult. Quantification of LC3 puncta in neurons (n = 6, one-way ANOVA, F 3,20 = 28.28). (C) Western blot analysis of SQSTIM1/P62 and LC3 protein levels in hippocampus. Quantification of relative gray value relative of SQSTIM1/P62 to b-actin (D, n = 3, one-way ANOVA, F 3,8 = 23.41) and LC3-II/LC3-I ratio (E, n = 3, one-way ANOVA, F 3,8 = 20.93). Statistical analysis was performed using one-way ANOVA followed by post hoc Tukey's test for multiple comparisons. *P < 0.05 compared with Sham group, # P < 0.05 compared with tMCAO group.

Immunohistochemistry and immunofluorescence
Deparaffinized 4 lm-thick coronal sections or cultured neurons fixed with 4% paraformaldehyde were incubated with 5% bovine serum albumin (BSA)/PBS for 30 min at room temperature and rinsed twice in PBS. Sections or cells were then permeabilized in 0.5% Triton X-100 for 20 min and incubated with the primary antibodies including rabbit anti-LC3 (1:100, Cell Signaling Technology, Cat# 41080), rabbit anti-STIM1(1:100, Cell Signaling Technology, Cat# 5668), rabbit anti-Orai1 (1:100, ThermoFisher Scientific, Cat# PA5-109270), rabbit anti-NeuN (1:100, Cell Signaling Technology, Cat# 24307) at 4°C overnight. Sections or cells were then rinsed twice in PBS and incubated with secondary antibody for 2 h at room temperature in the darkness. Slices were covered with a mounting medium containing 4,6-diamidino-2phenylindole (DAPI) to counterstain nuclei and imaged by confocal microscopy (Olympus iX80). Five regions of each sample were imaged and analyzed with ImageJ software.

Autophagic flux measurement
Neurons were infected with adenovirus expressing GFP-LC3 or GFP-mRFP-LC3 at a multiplicity of infection (MOI) of 50 (HanBio Technology, China). In the acidic lysosomes, the GFP signals were quenched while the mRFP remained stable. Accordingly, the formation of autophagosomes was indicated by yellow puncta while lysosomes remained red. Autophagy flux was determined by evaluating the number of GFP and mRFP puncta and the colocalization of the two signals with ImageJ software.

Cellular Ca 2+ measurement
Cytosolic free Ca 2+ was measured with Fluo-4 AM (Beyotime, China) according to manufacturers' instructions. Cultured neurons were incubated with 5 lM Fluo-4 AM for 30 min in normal Tyrode's solution, then washed three times with PBS and incubated for an additional 15 min in the absence of Fluo-4 AM to de-esterification AM esters. Cells were imaged using an inverted fluorescence microscope (Olympus, Germany) by monitoring the fluorescence intensity at Ex/Em = 490/5 25 nm. After obtaining a stable baseline recording in Ca 2+ -free Tyrode solution for 2 min, 4 lM thapsigargin (TG) was added to the solution to induce intracellular (D) Western blot of SQSTM1/P62 and LC3 after STIM1 silencing with siRNA under the condition of OGD for 12 h and quantification of relative gray value relative of SQSTIM1/P62 to b-actin (E, n = 6, one-way ANOVA, F 3,20 = 32.16), LC3-II/LC3-I ratio (F, n = 6, one-way ANOVA, F 3,20 = 34.52) and LC3-II/b-actin (G, n = 6, one-way ANOVA, F 3,20 = 47.01). Statistical analysis was performed using one-way ANOVA followed by post hoc Tukey's test for multiple comparisons. *P < 0.05 compared with Ctrl group, # P < 0.05 compared with OGD group.
Ca 2+ store depletion. After another 5 min, 2.5 mM Ca 2+ was added into the bath to induce SOCE. The fluorescent calcium images were recorded every 2 s for a total duration of 10 min. The data was analyzed with MetaFluor imaging software (Molecular Devices, US).

Statistical analysis
All the experiments were performed for a minimum of three times and data were presented as mean ± SD. In box-and-whisker plots, horizontal lines within boxes represent mean values, whiskers represent data outside 25th to 75th percentile range. Statistical analyses were carried out with GraphPad Prism version 9.0 software. Significant differences were assessed by univariate ANOVA among three or more groups, followed with post hoc Tukey's test for multiple comparisons. Differences between two groups were tested by unpaired t-test. P < 0.05 was considered as statically significant.

OGD enhances SOCE in hippocampal neurons
Since STIM1 is a key component of the store-operated Ca 2+ entry (SOCE) pathway, we further investigated whether SOCE plays a role in OGD-induced neuronal autophagy. Confocal intracellular Ca 2+ imaging showed that OGD enhanced SOCE, which was abolished by knockdown of Stim1 (Fig. 4A, B). In addition, OGD exposure also promoted co-localization of STIM1 and Orai1 (Fig. 4C), consisting with the activation of SOCE. Western blot also showed upregulation of Orai1 protein Quantification of relative gray value relative of p-AKT (E) and p-mTOR (F) to b-actin (n = 6, one-way ANOVA, F 7,40 = 162.8). (G) Western blot of LC3-II and SQSTM1/P62 protein levels after neurons were pretreated with 2 lM AZD5363 for 2 h prior to OGD. Quantification of relative gray value relative of SQSTM1/P62 to b-actin (H, n = 6, one-way ANOVA, F 7,40 = 162.8) and LC3-II/LC3-I ratio (I, n = 6, one-way ANOVA, F 7,40 = 331.3). Statistical analysis was performed using one-way ANOVA followed by post hoc Tukey's test for multiple comparisons. *P < 0.05. by OGD, but this effect was not affected by Stim1 knockdown ( Supplementary  Fig.  S5). These observations suggest that OGD increases SOCE which is associated with upregulation of both STIM1 and Orai1.

DISCUSSION
The present study demonstrated for the first time that STIM1-mediated SOCE plays an important role in regulating autophagy after ischemic stroke both in vivo and in vitro. The upregulation of STIM1 is a key reason for enhanced SOCE and autophagy, and neuronal injury after experimental ischemic insults in both mouse model and primary culture of hippocampal neurons. STIM1, a key component of SOCE, has been shwon to regulate multiple pathophysiological processes including proliferation, migration, inflammation, apoptosis, and autophagy (Henke et al., 2012;Somasundaram et al., 2014;Zhu et al., 2014;La Rovere et al., 2016;Volz et al., 2020). Accumulating evidence has proved that STIM1-mediated Ca 2+ influx is an important Ca 2+ entry pathway in both the neurons and nonexcitable cells (Serwach and Gruszczynska-Biegala, 2019). We have previously shown that cytoplasmic Ca 2+ overload is essential for neuronal apoptosis after ischemic stroke (Dai et al., 2014). In the present study, we found that STIM1 protein levels and its colocalization with Orai1 were increased, suggesting tMCAO/ OGD-induced Ca 2+ overload is likely driven via STIM1-related Ca 2+ influx pathway. Consistent with this concept, siRNA-mediated knockdown of STIM1 inhibited SOCE and attenuated neuronal injury after OGD.
Previous studies have shown that dysfunction of STIM1 may contribute to central nervous system (CNS) pathology, such as Alzheimerś disease, Huntington's disease, diffuse axonal injury, and traumatic brain injury (Hou et al., 2015;Kuang et al., 2016;Li et al., 2013;Serwach and Gruszczynska-Biegala, 2020). However, its role in ischemic stroke has not been clearly understood. In the present study, we found that knockdown of STIM1 protects against ischemic neuronal injury via regulating autophagic activity. Autophagy is an endogenous catabolic mechanism that depends on the fusion of doublemembrane autophagosomes with lysosomes to degrade misfolded proteins and impaired organelles . Although the role of autophagy in ischemic stroke is still controversial, it has been suggested that appropriate autophagy exerted a protective effect on ischemic cerebral damage, while excessive autophagy may cause nerve cell death (Wei et al., 2012). In both primary cortical neurons and SH-SY5Y neuronal cells, the specific autophagy inhibitor, 3-methyladenine (3-MA), reduces OGDinduced neuronal death (Shi et al., 2012). However, another study found that 3-MA or knockdown of autophagy-related protein 7 (Atg7) enhanced the ischemia-reperfusion-induced release of cytochrome c and the downstream activation of neuronal apoptosis Fig. 8. Schematic illustration of STIM1-mediated regulation of SOCE in hippocampal neurons after ischemic stroke. IS increases STIM1 protein levels and triggers SOCE by inducing STIM1 interaction with Oria1. STIM1-mediated SOCE reduces phosphorylation of AKT and mTOR, resulting in increased expression of LC3, P62, and impaired autophagic flux. . In our in vitro neuron model, OGD increased autophagic initiation but the autophagic flux was impaired indicated by the accumulation of SQSTM1/P62. However, in the previously published article, autophagic flux was enhanced during the reoxygenation phase after OGD (Thiebaut et al., 2022), the reason may be the different model conditions. It needs further to clarify which in vitro model can truly simulate the ischemic process of cerebral stroke in vivo. Therefore, it should be taken into consideration of choosing the model when doing the research on cerebral stroke.
In this study, we found that pharmacological inhibition of SOCE or siRNA-mediated knockdown of STIM1 markedly reduced OGD-induced autophagy, which had a neuroprotective role as suggested by the reduced LDH release. Interestingly, we used primarily cultured neurons infected with tandem mRFP-GFP-LC3 adenovirus to directly show the autophagic flux, when STIM1 expression was downregulated, the formation of autophagosome decreased while autolysosome didn't change significantly. It may indicate that STIM1 had a more important role in the early phase of autophagic activity. When considering the role of autophagy in stroke, age should be noticed since nearly-threequarters of all strokes occur in people over the age of 65 (Chen et al., 2019). The impairment of autophagic flux was associated with aging and neurodegenerative diseases (Popa-Wagner et al., 2020; La Barbera et al., 2022). In this context, the strategies regulating autophagy may be less effective in the aged than in the younger. Therefore, more research on aged animals or patients is needed to reveal the exact role of autophagic activity and develop effective treating strategies for ischemic stroke.
mTOR plays a crucial role in regulating autophagic activity in mammal cells which abundantly expresses in the brain (de Abreu et al., 2023). The present study suggests that upregulation of STIM1 promotes autophagy through inhibition of the AKT/mTOR signaling pathway. This is evidenced by the observations that the effects of stim1 knockdown on autophagic flux were reversed by the AKT inhibitor, AZD5363, in OGD-challenged neurons. In vivo studies using AZD5363 also suggested the involvement of the AKT/mTOR signaling pathway in the neuroprotective effects of shSTIM1 after tMCAO.
The present study demonstrated that upregulation of STIM1 enhances SOCE and suppresses AKT/mTOR signaling pathway, promoting the autophagic flux and ischemic neuronal injury (Fig. 8).

AVAILABILITY OF SUPPORTING DATA
The datasets used and/or analyzed during the current study are available from the corresponding authors upon reasonable request.