MKT077 (cat. no. HY-15096) was purchased from MedChemExpress. Penicillin-streptomycin mixed solution (cat. no. G4003), pancreatic enzyme (cat. no. G4001), primary antibody diluent (cat. no. G2025), phenylmethylsulphonyl fluoride (PMSF; cat. no. G2008), protein phosphatase inhibitors (cat. no. G2007), DMSO, minimal essential medium (MEM; cat. no. G4553) and 5X loading buffer (cat. no. G2013) were purchased from Wuhan Servicebio Technology Co., Ltd. High-glucose DMEM (cat. no. 10569010), glucose-free DMEM (cat. no. 10966025), Opti-MEM (cat. no. 31985062) and fetal bovine serum (FBS; cat. no. 10091148) were obtained from Gibco (Thermo Fisher Scientific, Inc.). The enhanced chemiluminescence (ECL) kit (cat. no. P0018FS) and bicinchoninic acid (BCA) assay kit (cat. no. P0010S) were purchased from Beyotime Institute of Biotechnology. Primary antibodies against β-actin (cat. no. AC026), GRP75 (cat. no. A112560), mitochondrial fission factor (MFF; cat. no. A8700), mitofusin (MFN)1 (cat. no. A9880), MFN2 (cat. no. A19678), dynamin-1-like protein (DRP1; cat. no. A17069), phosphorylated (p)-DRP1 (cat. no. AP0812) and hypoxia-inducible factor 1-alpha (HIF1α; cat. no. A11945) were purchased from ABclonal Biotech Co., Ltd. Horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (cat. no. AS014) and HRP-conjugated goat anti-mouse (cat. no. AS003) were purchased from ABclonal Biotech Co., Ltd. Lipofectamine^® 3000 transfection reagent (cat. no. L3000015) and MitoTracker^® Deep Red FM kits (cat. no. M7531) were from Thermo Fisher Scientific, Inc.

Male C57BL/6 mice (age, 10 weeks; weight, 25–30 g) were purchased from the Animal Experiment Center of Tongji Medical College, Huazhong University of Science and Technology (Wuhan, China). The mice were housed in a specific pathogen-free room with free access to rodent chow and water. The experimental protocol was reviewed and approved by the Animal Research and Care Committee of Tongji Medical College, Huazhong University of Science and Technology [Wuhan, China; no. SYXK(E)2016-0057].

A mouse model of MCAO was generated as reported previously with minor modifications (14,15). In brief, 14 mice were randomized into the control and MCAO groups (n=7 per group). Individual mice were intraperitoneally injected with 10% chloral hydrate (350 mg/kg) and no peritonitis was present. When the mice were anesthetized, an incision was made in the neck of the mice. After careful exposure of the right external carotid artery (ECA), common carotid artery (CCA) and internal carotid artery (ICA), the CCA was clamped and the ECA was tied, followed by insertion of a 6–0 monofilament nylon suture with a rounded tip from the CCA to the ICA to occlude the left MCA at its origin (~10 mm) and the skin was sutured. The mice were kept on a heating pad to maintain their temperature at 37°C. After recovery from anesthesia, the mice were returned into their cages and they were euthanized at 8 h post-ischemia. The mice were decapitated without anesthesia and four brains were used for 2,3,5-triphenyltetrazolium chloride (TTC) staining and three were analyzed using transmission electron microscopy (TEM).

The mice were euthanized by decapitation at 8 h of ischemia and their brains were rapidly removed and frozen. The frozen brains were cut into 2-mm sections. The brain slices were stained with 2% TTC solution with incubation at 37°C for 15 min. Subsequently, the brain slices were fixed with 10% paraformaldehyde for 10 min. The percentage of infarct volume of the total brain volume in individual mice was measured using Image J software (v1.51j8) [National Institutes of Health (NIH)] in a blinded manner.

The mouse brain tissues were immersed in 2.5% glutaraldehyde (4°C) for 24 h and fixed with 1% osmotic acid-1.5% potassium ferrocyanide for 1.5 h at room temperature. An alcohol gradient was used for dehydration: 75% ethanol for 4 h, 85% ethanol for 2 h, 90% ethanol for 1 h, 100% ethanol for 1 h and another 100% ethanol for 1 h. Next, samples were immersed in 100% xylene for 30 min and 100% xylene for 1 h. Finally, samples were paraffin-embedded for 90 min. Ultrathin tissue sections were stained with 3% uranyl acetate and lead citrate (50 nm) and the mitochondrial morphology was observed by TEM. The length/diameter ratio of the mitochondria was calculated and those with a ratio of >2 were considered as long mitochondria (long-mito) (16).

HT22 mouse hippocampus neuronal cells (Procell Life Science & Technology Co., Ltd.) and N2a neuroblastoma cells (Wuhan Servicebio Technology Co., Ltd) were cultured in 10% FBS high-glucose DMEM and MEM, respectively, with 1% penicillin and streptomycin (complete medium) in a 5% CO₂ incubator at 37°C.

Primary mouse neuron extraction was performed to investigate mitochondrial function. In brief, one pregnant (E15-17) C57 mouse which was purchased from the Animal Experiment Center of Tongji Medical College, Huazhong University of Science and Technology, was anesthetized using 10% chloral hydrate (350 mg/kg) and the abdominal cavity was quickly opened and fetal mice (six mice) were removed. The pregnant and fetal mice were subsequently decapitated and the cortical regions were isolated and digested with 0.25% trypsin (cat. no. G4001; Wuhan Servicebio Technology Co., Ltd.), followed by termination of digestion with medium (5% FBS neurobasal medium; cat. no. H430717; Shanghai Basalmedia Technologies Co., Ltd.). After centrifugation, the cells were rapidly resuspended in medium. The prepared single cells were cultured in petri dishes coated with polylysine (Gibco; Thermo Fisher Scientific, Inc.). After 12 h of culture, the cells were placed in 1% B27 (FBS-free) medium for subsequent experiments. The experimental protocol was reviewed and approved by the Animal Research and Care Committee of Tongji Medical College, Huazhong University of Science and Technology [approval no. SYXK(E)2017-0012].

The cells were transfected when they had grown to ~70% confluency. For this, 2 µl transfection reagent, 2 µg of probe plasmid or 5 pM of specific small inhibitory (si)RNA were added to 100 µl of Opti-MEM and incubated for 5 min. Opti-MEM with transfection reagents was then added to Opti-MEM with plasmid or siRNA. After resting for 15 min, the mixed suspension was added to the cell culture medium. In addition, a negative control group remained untransfected. The medium was replaced 6 h after transfection, the cells were cultured for another 24 h and then used in the following experiments. The sequences of negative control siRNA (si-NC) and GRP75-specific siGRP75 were as follows: si-NC forward, 5′-UUCUCCGAACGUGUCACGUTT-3′ and reverse, 5′-ACGUGACACGUUCGGAGAATT-3′; siGRP75 forward, 5-GCGUCUUUACCAAACUUAUTT-3′ and reverse, 5′-AUAAGUUUGGUAAAGACGCTT-3′.

For mitotracker staining, MitoTracker^® Deep Red FM kits were used. In brief, cells were stained in the dark using mitotracker (500 nM) for 15 min and then observed with a fluorescence confocal microscope. ImageJ software v.1.51j8 (NIH) was used for the quantification of fluorescence.

The cells were cultured in 10% FBS glucose-free DMEM at 37°C in an incubator with 91% N_2, 5% CO₂ and 4% O₂ as the OGD conditions for varying time periods. Control cells were cultured under normal cell culture conditions in 95% air and 5% CO₂; N2a cells were cultured in MEM with 10% FBS and 1% penicillin-streptomycin, while HT22 cells were cultured in DMEM with 10% FBS and 1% penicillin-streptomycin. MKT077 was added to the medium at a final concentration of 5 µM with incubation for 8 h in a 37°C cell incubator.

The different groups of cells were lysed in RIPA buffer containing 1 mM PMSF and 1 mM phosphatase inhibitor cocktail and centrifuged at 14,000 × g for 12 min at 4°C. After determining the protein concentration of each lysate sample using the BCA kit, the lysates (30 µg/lane) were separated by SDS-PAGE on 10–12% gels and transferred onto PVDF membranes (cat. no. WGPVDF45; Wuhan Servicebio Technology Co., Ltd.). The membranes were incubated with 5% dry skimmed milk in 0.1% Tween-20 in Tris-buffered saline for 1 h at room temperature and probed at 4°C overnight with primary antibodies (1:1,000 dilution for all). Subsequently, the membranes were incubated with HRP-conjugated goat anti-rabbit IgG and (1:10,000 dilution) and HRP-goat anti-mouse IgG and visualized with the ECL kit using a Bio-Rad exposure system (Bio-Rad Gel Doc XR; Bio-Rad Laboratories, Inc.) and the Image Lab 5.1 software (Bio-Rad Laboratories, Inc.). Quantitative results were obtained by densitometric scanning using ImageJ v.1.51j8 software (NIH).

The levels of mitochondrial Ca²⁺ in N2a cells were determined after transfection with mitochondrial Ca²⁺ probe plasmids which had been used previously (17). In each confocal dish, 5×10⁴ N2a cells were cultured for 24 h. The calcium plasmid was transfected into N2a cells for 24 h using the cell transfection method described above. After washing the cells with pre-warmed PBS solution, they were examined under a laser confocal microscope (A1R; Nikon Corporation) using Nikon Imaging Software Elements Viewer 4.20 (Nikon Corporation). The mitochondrial Ca²⁺ levels were determined by measuring average intensity levels using ImageJ software 1.51j8 (NIH).

The levels of intracellular ATP in individual groups of cells were measured using a specific kit (cat. no. S0026B; Beyotime Institute of Biotechnology) following the manufacturer's protocol. In brief, the same numbers of cells in each group were lysed on ice and centrifuged. The supernatants were collected for the measurement of ATP levels using a luminometer (Promega Corporation) and the levels of ATP were calculated using the standard curve.

The MMP of each group of cells was measured using the specific kit (cat. no. C2006; Beyotime Institute of Biotechnology) following the manufacturer's protocol. In brief, after treatment, the different groups of cells were stained with JC-1 solution at 37°C for 20 min. After being washed, the cells were imaged under a Nikon confocal microscope (A1R; Nikon Corporation) and the red and green fluorescent intensity of 100 cells from 6 randomly selected fields were analyzed using ImageJ software v.1.51j8 (NIH).

The levels of intracellular ROS were measured using a specific kit (cat. no. S0033; Beyotime Institute of Biotechnology) following the manufacturer's protocol. In brief, the dichloro-dihydro-fluorescein diacetate (DCFH-DA) probe was diluted to 10 mM with FBS-free DMEM. The different groups of cells were stained with 10 mM DCFH-DA in FBS-free DMEM. The fluorescent signals in each group of cells were imaged under a Nikon confocal microscope (A1R; Nikon Corporation) and the green, fluorescent intensity of ~100 cells from 10 randomly selected fields were measured using ImageJ software v.1.51j8 (NIH).

Data were analyzed by GraphPad Prism 5 software (GraphPad Software, Inc.) and are expressed as the mean ± standard error of the mean. Statistical significance was assessed by one-way ANOVA followed by Tukey's post-hoc test. P<0.05 was considered to indicate a statistically significant difference.

First, male C57BL/6 mice were subjected to MCAO as a classical in vivo model of ischemic stroke (18). TTC staining indicated obvious infraction in the brain of MCAO model mice (Fig. 1A). Acute hypoxia significantly influenced mitochondrial function; hence, TEM was used to confirm the morphology changes of mitochondria in ischemic mouse brains. As indicated in Fig. 1B, swollen and fragmented mitochondria were apparent in the MCAO group and the cristae of mitochondria were broken and fractured. However, the morphology of mitochondria was normal in the sham mice. Since mitochondrial function is tightly associated with the morphology of mitochondria (19), several indexes reflecting mitochondrial function, including ATP content, ROS production and the MMP were also determined and it was indicated that mitochondrial function was significantly impaired by OGD (Figs. 1C, S1 and S2).

In order to further confirm the above results in an in vitro model, OGD was implemented to resemble the effects of MCAO in the HT22 and N2a cell lines. After 8 h of OGD, cellular mitochondria were observed by mitotracker staining (Fig. 2A). The mitochondria became fragmented after OGD, as the length of mitochondria decreased while the number of mitochondria per cell increased after OGD, which led to a decreased mitochondrial ratio. Similarly, OGD also impeded mitochondrial function (Fig. 2B) and decreased the MMP and ATP levels but increased ROS levels. In N2a cells, western blot analysis was used to detect mitochondrial dynamic-related proteins. The results indicated that under OGD conditions, mitochondrial fusion proteins MFN1 and MFN2 were downregulated, while mitochondrial fission proteins MFF and p-DRP1 increased significantly, indicating that the mitochondria became fragmented (Fig. 2C).

GRP75 acts as a bridge, linking the ER-anchored inositol 1,4,5-trisphosphate receptor type 1 (IP3R) and mitochondria-located voltage-dependent anion-selective channel protein (VDAC) to form a protein complex and induce calcium transfer between ER and mitochondria (13). In the present study, a mitochondrial tagged calcium sensor vector was used to monitor calcium levels in mitochondria and it was indicated that mitochondrial calcium levels were significantly elevated after OGD (Fig. 3A). It was then examined whether mitochondrial overload was related to abnormal GRP75 expression. Consistent with the present hypothesis, GRP75 was significantly elevated after OGD (Fig. 3B and C). This indicated that GRP75 may be related to OGD-induced mitochondrial calcium overload. The expression of GRP75 in the infarction area of the MCAO model was also examined, suggesting that GRP75 was also elevated in vivo (Fig. 3D). HIF1α is one of the most important modulators of hypoxia and the expression of HIF1α was thus determined; the results indicated that HIF1α was upregulated (Fig. 3E). Therefore, it may be hypothesized that the increase of HIF1α induced the upregulation of GRP75.

After confirming abnormal elevation of GRP75 after OGD, it was investigated whether interfering with GRP75 is an effective strategy to protect mitochondrial function after OGD. MKT077, a GRP75-specific inhibitor (20), which is able to bind to GRP75 and abrogate its activity (21), was applied and it was observed that, although MKT077 did not downregulate GRP75 expression after OGD (Fig. 4A), mitochondrial calcium overload was obviously inhibited after MKT077 treatment (Fig. 4B), which indicated that inhibiting GRP75 function prevented mitochondrial calcium overload. As expected, mitochondrial morphology was well-preserved by MKT077 even after OGD (Fig. 4C). This further supported the notion that mitochondrial calcium overload was able to trigger mitochondrial fragmentation and induce mitochondrial dysfunction (Figs. 4D, S1 and S2).

In order to exclude any possible off-target effects of the small molecule antagonist, siRNA targeting at GRP75 was used to further confirm its effect on mitochondrial overload. siGRP75 efficiently reduced GRP75 expression levels after OGD, while si-NC did not influence the expression of GRP75 (Figs. 5A and S3). Similar to the effect of MKT077, siGRP75 was also able to reduce mitochondrial calcium overload (Fig. 5B) and preserve mitochondrial morphology and function after OGD (Figs. 5C and D, S1 and S2), which further confirmed that inhibiting GRP75 preserved mitochondrial morphology and function.