HSYA (purity >98%, lot number Z01010BA13) was obtained from Shanghai Yuanye Biotechnology Co., Ltd. (Shanghai, China). AHSYB (purity >98%) was made at the Institute of Chinese Medicine Chemistry in Zhejiang Chinese Medical University (Xu et al., 2017). Nimodipine (Nim) injection was obtained from Jumpcan Pharmaceutical Group Co., Ltd. (Jiangsu, China). The antibodies used in this study were anti-SIRT1 (orb19330, Biorbyt), rabbit anti-FOXO1 (ab70382, Abcam), rabbit anti-PGC1α (ab191838, Abcam), rabbit anti-Bax (ab32503, Abcam), rabbit anti-Bcl-2 (ab194583, Abcam), rabbit anti-GAPDH (ab181602, Abcam), goat anti-mouse IgG (H + L) secondary antibody (31,160, Thermo), and goat anti-rabbit IgG (H + L) secondary antibody (31,210, Thermo).

Primary hippocampal neuronal cells were obtained from neonatal SD rats, and the culture method was based on Yu’s previous description with some slight modifications (L. Yu et al., 2018). In brief, the hippocampus was separated from the brains and minced into small pieces and then dissociated with 0.25% trypsin-EDTA for 10 min at 37°C. Following the termination of digestion, the cell suspension was centrifuged at 1,000 rpm for 6 min and resuspended in Dulbecco’s modified Eagle medium with Ham’s F12 medium (DMEM/F12) which contained 10% fetal bovine serum (FBS). The cell suspension was plated onto 0.1 mg/ml poly-l-lysine (P-1399; Sigma)–coated 6-well plates or 96-well plates at a density of 2 × 10⁵/ml. After incubation at 37°C for 4.5 h, the medium was removed and completely replaced with the neuronal culture medium containing Neurobasal-A medium (10,888–022; Gibco) supplemented with 2% B27 (17,504–044; Gibco) and 1% glutamine. After that, half of the medium was changed every 2 days. The cells were maintained in an incubator with 5% CO₂ at 37°C for 7 days for the following experiments.

After 7 days, the medium was removed, and the cells were washed with phosphate-buffered saline (PBS) three times. Then, the primary hippocampal neurons were incubated in a three-gas incubator (94% N₂, 5% CO₂, and 1% O₂) with glucose-free Eagle’s medium at 37°C for 6 h to mimic the oxygen-glucose deprivation injury. Thereafter, the cells were incubated with a normal medium under normoxic conditions for 24 h to ensure reoxygenation. The primary-cultured hippocampal neuronal cells were divided into 11 groups: control group; OGD/R group; Nim (10 μM) group; HSYA (40 μM) group; HSYA (60 μM) group; HSYA (80 μM) group; AHSYB (40 μM) group; AHSYB (60 μM) group; AHSYB (80 μM) group; and HSYA/AHSYB + EX527 (10 μM) groups. EX527, a SIRT1–specific inhibitor, was purchased from APE × BIO Technology (Houston, United States). The control group was cultured under normal conditions. All drugs were dissolved in the culture medium and applied to the cell during the period of OGD/R injury. Also, we pretreated hippocampal cells with EX527 for 1 h before initiating HSYA or AHSYB treatment.

After exposure to OGD/R and different treatments, cell viability was evaluated using the cell counting kit-8 (CCK-8) assay (Beijing Zoman Biotechnology Co., Ltd., Beijing, China) according to the manufacturer’s protocol. The cells seeded into 96-well plates were treated with the CCK-8 solution and cultured at 37°C for 2 h after treatment as mentioned above. Also, the optical density (OD) values were measured at a wavelength of 450 nm.

The integrity of the cell membrane and neuronal apoptosis are related to the release of lactate dehydrogenase (LDH) (Y. Yu et al., 2018). Also, the activity of LDH in the medium was measured by ELISA in accordance with the manufacturer’s instructions.

We used Hoechst 33342, a DNA–binding dye, to examine hippocampal cell apoptosis (Y. Liu et al., 2018). After being processed under OGD/R conditions, the cells were fixed for 10 min in 4% paraformaldehyde at room temperature and incubated with 10 μg/ml Hoechst 33342 (Solarbio, Beijing, China) at 37°C for 25 min in the dark. Then, the cells were washed three times with PBS and observed under a fluorescence microscope (Leica, Heidelberg, Germany) immediately. The cells which exhibited intense blue fluorescence and nuclear condensation were scored as apoptotic cells (Zeng et al., 2016). Also, the percentage of the apoptotic neuronal cells was calculated by the ratio of apoptotic cells to the total cells counted.

Sprague–Dawley rats (250–300 g) were obtained from the Animal Experimental Center of Zhejiang Chinese Medical University (license no. SYXK 2018–0,012). These rats were housed in a climate-controlled room (humidity of 50 ± 5% and temperature of 22 ± 2°C).

The MCAO/R model was operated with minor modifications according to Zea Longa’s method (Longa et al., 1989). After fixing the rat in the supine position, the left common carotid artery (CCA), internal carotid artery (ICA), and external carotid artery (ECA) were separated and ligated. A small hole was cut in the CCA to insert a silicone-coated 4-0 nylon monofilament (diameter of 0.28 mm, Beijing Sunbio Biotech Co., Ltd., China) into the ICA to maintain a state of ischemia. After MCAO for 1 h, the monofilament was slowly pulled to perform reperfusion and then the incision was sutured. During the surgery, all rats were positioned on a heating pad to maintain the core temperature at 37 ± 0.5°C. In the sham group, the middle cerebral artery was not occluded.

Rats were divided into nine groups, including the sham group, MCAO/R group, MCAO/R + HSYA groups (2, 4, and 8 mg/kg), MCAO/R + AHSYB groups (2, 4, and 8 mg/kg), and MCAO/R + Nim group (10 mg/kg). The drugs were administered through the tail vein immediately after MCAO/R for three consecutive days. Also, the doses of HSYA and AHSYB were determined by the clinical dosage of safflower, as given in the literature (Ramagiri and Taliyan, 2016; X.; Yang et al., 2020) and preliminary experiments in our research. For the sham and MCAO/R groups, rats were given normal saline (NS) using the same procedures. The schematic diagram of the animal experiments is shown in Supplementary Figure S2.

A neurological deficit assessment was conducted as per Longa’s method (Longa et al., 1989) by an investigator who was blinded to the experiment design using a five-point scale as described previously. The observations were as follows : 0, no symptoms of nerve damage; 1, rats cannot extend the ischemia contralateral forelimb; 2, rats turned to the contralateral side when crawling forward; 3, rats collapsed to the opposite side at rest; and 4, rats lost consciousness and could not be revived.

On the 3rd day after MCAO/R injury, the brain tissues were separated quickly and frozen at −20°C. After 12 min, the brains were taken and sliced into 2 mm thick coronal slices quickly. Then, the slices were stained with 2% 2,3,5-triphenyl tetrazolium chloride (TTC) at 37°C for 15 min. The normal brain tissue was rose-red, whereas the infarct area remained unstained. The brain slices were photographed using a digital camera, and the images were analyzed using Image J software.

In brief, on the 3rd day after MCAO/R, the rats were perfused with physiological saline and 4% paraformaldehyde solution and the brains were fixed in 4% paraformaldehyde (4°C). Then, the samples were embedded in paraffin and cut into 5 μm coronal sections for a series of Hematoxylin and Eosin (HE) staining.

The apoptotic cells in the brain tissues were detected by TUNEL staining, and the tissue sections were treated using the In Situ Apoptosis Detection Kit (TaKaRa DMK500). In brief, the paraffin-embedded sections of the rat brain tissue were deparaffinized and washed with PBS and permeabilized with proteinase K (20 μg/ml) for 30 min at 37°C. After that, the tissues were incubated in a TUNEL reaction mixture for 60 min. Then, the anti-FITC HRP was added into the tissue sections for 60 min at 37°C. Images were obtained using fluorescence microscopy (Olympus BX53, Olympus, Japan).

The levels of superoxide dismutase (SOD), malondialdehyde (MDA), ROS, and glutathione peroxidase (GSH-Px) in primary-cultured hippocampal neuronal cells and the rat serum were measured by ELISA. In brief, cells were seeded in 6-well plates, and after exposure to OGD/R, the cells were harvested to a sterile container, sonicated, and centrifuged at 12,000 rpm for 10 min and then the supernatants were collected. As for rat serum samples, they were obtained from the heart and collected by centrifugation at 4,000 rpm for 15 min at 4°C. All operations were determined with the respective ELISA kit according to the manufacturer’s instructions.

The total RNA of primary-cultured hippocampal cells and ipsilateral hemisphere brain tissues of rats were isolated using the TRIzol® reagent (Invitrogen). And then, the total mRNA in each group was reverse-transcribed into cDNA with SuperScript™ III First-Strand Synthesis SuperMix (Invitrogen, United States) following the manufacturer’s instructions. Real-time PCR was performed on Power SYBR® Green PCR Master Mix (Applied Biosystems, United States). Finally, the relative mRNA expression levels of SIRT1, FOXO1, PGC1α, Bax, and Bcl-2 were quantified by the CFX384 Real-time PCR system (Bio-Rad, United States) and analyzed by the 2^−ΔΔCt method. The primer sequences were obtained from Shanghai Sangon Biological Engineering Technology (Shanghai, China), and their sequences are listed in Table 1.

The total protein was extracted from primary-cultured hippocampal neuronal cells and rat brain tissues using the total-protein extraction kit with Protease Inhibitor Cocktail (Thermo Fisher, Waltham MA, United States). Protein samples containing 60 μg proteins were separated with 8–12% sodium dodecyl sulphate (SDS)-PAGE gel and then transferred onto PVDF membranes (Millipore, Billerica MA, United States). The membranes were then incubated for 1 h with 5% fat-free milk and incubated with the primary antibodies, rabbit anti-SIRT1 (1:200, Biorbyt), rabbit anti-FOXO1 (1:4,000, Abcam), rabbit anti-PGC1α (1:1,000, Abcam), rabbit anti-Bax (1:1,000, Abcam), rabbit anti-Bcl-2 (1:500, Abcam), and rabbit anti-GAPDH (1:10,000, Abcam), overnight at 4°C. Then, the membranes were incubated with appropriate secondary antibodies (1:5,000, Thermo) for 1 h at room temperature. Ultimately, the blots were visualized by ECL chemiluminescence and the protein quantification was measured using Image J software. The relative protein levels were normalized with GAPDH.

We took the paraffin-embedded brain tissues to carry out immunohistochemical assessment. In brief, the paraffin sections were deparaffinized and immersed in sodium citrate buffer (10 mmol/L, pH 6.0) and subjected to microwave antigen retrieval for 15 min. Subsequently, the sections were incubated in 0.3% H₂O₂ for 15 min and blocked with 10% serum for 15 min. The sections were treated with one of the following primary antibodies, rabbit anti-SIRT1 (1:200, Abcam), rabbit anti-FOXO1 (1:100, CST), and rabbit anti-PGC1α (1:300, Abcam), and incubated overnight at 4°C. Then, the sections were incubated with secondary antiserum for 40 min at room temperature. Finally, the sections were stained using DAB developer kits. The expressions of SIRT1, FOXO1, and PGC1α were measured by the integrated optical density (IOD) using Image-Proplus 6.0 software.

All results were expressed as means ± standard deviation (SD) and differences between the groups were evaluated using one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test. Data handling and statistical processing were performed using GraphPad Prism 8.0 (GraphPad Software, San Diego, CA, United States). A value of p less than 0.05 was regarded as statistically significant.

We first incubated hippocampal cells with different concentrations of HSYA and AHSYB under normal conditions to determine the optimal concentrations of HSYA and AHSYB. As shown in Supplementary Figures S3A,B, there is no significant difference between the control group and HSYA with different doses (0–320 μM). As for AHSYB, it did not affect cell viability until the concentration up to 160 μM and so 5–120 μM was selected for the following experiments. Then, hippocampal neuronal cells were subjected to OGD/R injury to study the neuroprotective properties of HSYA and AHSYB against cerebral I/R injury. As shown in Supplementary Figures S3C,D, no significant effect was observed when HSYA was given at the doses of 10 and 20 μM and AHSYB improved cell viability significantly at a concentration range of 10–120 μM. At the dose of 80 μM, both HSYA and AHSYB exhibited more remarkable effects. Thus, three concentrations of HSYA and AHSYB (40, 60, and 80 μM) were selected for subsequent experiments.

Cell viability and the release of LDH were examined to reflect the neuroprotective effects of HSYA and AHSYB on hippocampal cells after OGD/R. Compared to the control group, cell viability was markedly inhibited in the OGD/R group (Figure 1A). However, cell viability was increased by HSYA (40, 60, and 80 μM) and AHSYB (40, 60, and 80 μM) in a dose-dependent manner when compared with the OGD/R group. Meanwhile, the effects of HSYA and AHYSB were similar to those of the positive control drug (Nim group). In addition, once suffering from cerebral I/R injury occurs, a large amount of LDH was released to the outside of the cells. As shown in Figure 1B, after exposure to OGD/R injury, the cells released a higher amount of LDH than in the control group, indicating that the cytomembrane was broken by cerebral I/R injury. In HSYA and AHSYB treatment groups, the LDH leakage was significantly decreased showing that HSYA and AHSYB suppressed the OGD/R–induced cytotoxicity.

We also investigated the effects of HSYA and AHSYB on OGD/R–induced oxidative stress in hippocampal neurons by detecting the markers of oxidative stress, including ROS, MDA, SOD, and GSH-Px. Figures 2A,B show that ROS and MDA production significantly increased in the OGD/R group. Conversely, treatment with HSYA and AHSYB (40, 60, and 80 μM) decreased the levels of ROS and MDA in a dose-dependent manner. Interestingly, there was no significant difference among HSYA₈₀, AHSYB₈₀, and Nim groups. In addition, the activities of GSH-Px and SOD were markedly decreased in the OGD/R group relative to the control cells, whereas HSYA and AHSYB elevated the levels of GSH-Px and SOD (Figures 2C,D). In brief, these findings demonstrated that HSYA and AHSYB distinctly ameliorated OGD/R–induced oxidative stress in hippocampal neuronal cells.

Transient cerebral I/R injury leads to neuronal damage/death, and especially, the hippocampal cells in the brain are vulnerable to the attack of cerebral I/R injury. Herein, we use Hoechst 33342 to stain apoptotic nuclei to evaluate the antiapoptotic effect of HSYA and AHSYB in the hippocampus. As shown in Figure 3A, the typical morphological features of apoptosis were observed in the OGD/R group, including nuclear shrinkage, chromatin condensation, and more cells shown by bright staining. Treatment with HSYA and AHSYB significantly decreased the number of apoptotic cells characterized by less chromatin condensation. At the same time, the neuron apoptotic rate in the OGD/R group was higher than in drug-treated groups, through quantitative analysis (Figure 3B).

In addition, we also used Western blot and real-time RT-PCR to detect the expressions of apoptosis-related proteins and mRNA, including Bcl-2 and Bax. As shown in Figure 4, consistent with Hoechst 33342 staining, OGD/R induction caused significant increases in the expression of Bax, while leading to a reduction in the level of Bcl-2. Also, HSYA and AHSYB treatment changed the trends of Bax and Bcl-2. For example, HSYA and AHSYB reduced the decrease of Bcl-2 and the increase of Bax in OGD/R–induced hippocampal cells. Interestingly, HSYA and AHSYB exhibited similar pharmacodynamic action. In general, these results indicated that HSYA and AHSYB significantly inhibited the OGD/R–induced neuronal apoptosis.

It was vital for us to understand the underlying mechanisms of HSYA and AHSYB on anticerebral I/R injury and thus RT-PCR and Western blot technology were used to explore the expressions of SIRT1, FOXO1, and PGC1α. As shown in Figures 5A–C, OGD/R exposure dramatically suppressed the mRNA expressions of SIRT1, FOXO1, and PGC1α in primary hippocampal cells. However, these inhibitions were markedly counteracted by HSYA and AHSYB treatment. In addition, Western blot technology was applied to estimate whether HSYA and AHSYB increased the expression of the SIRT1 pathway–related proteins. As shown in Figures 5D–G, the detection result was in accordance with that of the above RT-PCR and the levels of SIRT1, FOXO1, and PGC1α in the HSYA and AHSYB group were significantly increased compared with those in the OGD/R group.

In this part, EX527, a specific inhibitor of SIRT1, was used to inhibit the SIRT1 pathway, which further used to assess whether the neuroprotective effects of HSYA and AHSYB were dependent on the SIRT1 pathway activation. After OGD/R treatment, cells were coated with HSYA/AHSYB and EX527 and then cell viability; the release of LDH; Hoechst 33342 staining; the levels of oxidative stress markers, including ROS, MDA, SOD, and GSH-Px; and the expressions of Bax, Bcl-2, and the SIRT1 pathway–related proteins and genes were measured. As shown in Figure 6A, the SIRT1 pathway inhibitor EX527 reversed the increase in cell viability induced by HSYA and AHSYB. Also, the levels of LDH, ROS, and MDA were increased by EX527 (Figures 6B–D). GSH-Px and SOD levels were decreased by EX527 (Figures 6E,F).Similar results were shown by Hoechst 33342 staining (Figures 6G,H). Besides, Western blot analysis and RT-PCR showed that the expressions of SIRT1, FOXO1, PGC-1α, and Bcl-2 were decreased after treatment with HSYA/AHSYB and EX527 (Figure 7), which meant that the activation of the SIRT1 pathway was indeed related to the neuroprotective effects of HSYA and AHSYB.

The neurological deficit scores were assessed to determine the effects of HSYA and AHSYB. Also, the differences among the sham group, MCAO/R group, and treatment groups are as illustrated in Figure 8C. Rats in the MCAO/R group showed more severe neurological injuries than in the sham group. Just like the positive drug (Nim group), treatment with HSYA and AHSYB decreased the neurological deficit scores in a dose-dependent manner. Also, there were no significant differences between HSYA and AHSYB.

The extent of brain infarct volume was evaluated by TTC staining. As shown in Figures 8A,B, there was no infarction in the sham group, but the cerebral infarct area of rats in the MCAO/R group was significantly increased when compared with the sham group. Of note, treatment with HSYA, AHSYB, and Nim significantly inhibited this phenotype, which indicated that HSYA and AHSYB could protect against cerebral I/R–induced injury. The synthesis results of neurological deficit and TTC staining indicated that both HSYA and AHSYB exerted a protective effect at the dosages of 4 and 8 mg/kg and the dosage of 8 mg/kg showed more powerful effects; thus, follow-up studies were performed with this dosage.

We used HE staining on sections from the ischemic hippocampus to further assess the pathological damage in rats. Compared with the sham group, the structure of hippocampal neurons in the MCAO/R group was severely damaged displaying characteristics such as nuclear deformation, shrinkage, and hyperchromia. Besides, the number of cells in the MCAO/R group was decreased and they were loosely arranged. However, it is worth noting that HSYA and AHSYB treatment improved the structural damage of hippocampal neurons, which indicates that HSYA and AHSYB protected the brain from MCAO/R injury (Figure 9).

The effects of HSYA and AHSYB on cell apoptosis in vivo were determined by TUNEL staining. As shown in Figure 10, there were nearly no apoptotic cells found in the sham group. However, a large number of apoptotic cells expressed as brown-yellow stained cells in the ischemic hippocampus of the MCAO/R group of rats were observed. Just as shown in the bar chart, the apoptosis rate in the MCAO/R group was significantly higher than in the sham group, and after HSYA and AHSYB administration, this percentage was significantly lower than in the MCAO/R group. In addition, we also detected the expressions of main apoptosis-related mRNA and proteins, including Bax and Bcl-2. According to the results shown in Figure 11 that were in line with TUNEL staining, HSYA and AHSYB reduced the increase of Bax and the decrease of Bcl-2. Meanwhile, the results were consistent with that found in vitro, which further confirmed that HSYA and AHSYB treatment prevented ischemia-induced neuronal apoptosis.

The expressions of ROS, MDA, GSH-Px, and SOD in the rat serum were measured by ELISA. As shown in Figure 12, the levels of ROS and MDA were markedly increased in the MCAO/R group when compared with the sham group. However, the treatment with different doses of HSYA and AHSYB dose-dependently downregulated the amount of ROS and MDA, while GSH-Px and SOD were upregulated in drug-administered groups.

Next, we sought to further clarify whether SIRT1 is involved in the regulation of the FOXO1 and PGC1α expression in vivo. As shown in Figure 13, MCAO/R induction significantly suppressed the mRNA and protein expressions of SIRT1, FOXO1, and PGC1α in ischemic brain tissues. Treatment with HSYA and AHSYB markedly increased the levels of the SIRT1 pathway–related mRNA and proteins when compared with the MCAO/R group.

The results of immunohistochemical assessment in MCAO rats also revealed that SIRT1–positive staining was decreased in the brain of MCAO/R–induced rats compared with the sham group and treatment with HSYA and AHSYB dramatically increased the IOD of SIRT1. As the most important downstream targets of SIRT1, the trends of FOXO1 and PGC1α were consistent with those of SIRT1 (Figure 14). On the basis of the abovementioned results, HSYA and AHSYB exerted their anti-ischemic effect mainly through the SIRT1 pathway.