Human retinal endothelial cells (HRECs) were obtained from Cell Systems (Kirkland, WA, USA). The culture medium was EGM-2 Bulletkit medium (Lonza, Basel Switzerland) with 100 U/ml penicillin/100 µg/ml streptomycin (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA), containing 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.). Cells were maintained at 37°C and 5% CO₂. HRECs of passages 6–10 were used for experiments.

After an initial 24 h of culture in serum-free medium, HRECs were treatment with various concentration of gastrodin (0.1, 1, 10, 100 µM) or 100 nM gastrodin in normal medium for various time (6, 12, 24 and 48 h) to detect the optimal concentration and time of intervention. Cells were subjected to medium containing 5 mM glucose (control group) or high glucose medium containing 30 mM glucose (HG group). HRECs cultured in 6-well plates were transfected with recombinant plasmid pcDNA3.1 or pcDNA3.1-SIRT1 (Invitrogen; Thermo Fisher Scientific, Inc.) to overexpress SIRT1 using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) in accordance with the manufacturer's protocol.

The Cell Counting Kit-8 (CCK-8, Biosharp, Hefei, China) was performed to determine cell viability of HRECs. In brief, 5×10³ of cells were cultured in a 96-well plate at 37°C and allowed to attach for 24 h. After starvation for 12 h, cells were stimulated with different concentrations of gastrodin for different time in normal median. Then, 10 µl CCK-8 solution was added to each well, plates were incubated at 37°C for another 2 h. The absorbance of cells was then measured at 450 nm.

The intracellular ROS level was measured by dihydroethidium (DHE; Beyotime Institute of Biotechnology, Haimen, China) staining. After various treatment, HRECs were incubated with 5 µM DHE at 37°C for 30 min. Following incubation, cells were fed with normal growth medium without DHE for 1 h, and rinsed with PBS. Fluorescence images were observed using a fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany). The fluorescence intensity was measured using ImageJ software (NIH, Bethesda, MD, USA).

RT-qPCR analyses were carried out as described previously (14). Total RNA was extracted from cells in groups using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) and quantified by measuring the absorbance at 260 nm. Subsequently, 2 µl of total RNA was used for the preparation of cDNA by reverse transcription using the PrimeScript RT reagent kit (Takara Bio, Inc., Otsu, Japan) according to the manufacturer's instructions. The expression of HO-1, NQO1, NRF2 and GCLM mRNA were determined on the Applied Biosystems StepOne Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.). The reaction condition was 95°C at 10 min for a hot start, then 40 cycles at 95°C for 15 sec, 60°C for 60 sec and 72°C for 60 sec. The primer sequences used in this study were listed in Table I. The gene expression level was calculated by 2^−ΔΔCq methods and normalized to 18S RNA. All experiments were repeated in three times.

HRECs cell apoptosis rates were measured by flow cytometric analysis using Annexin V-FITC-PI Apoptosis Detection kit (Vazyme Biotech, Nanjing, China). Briefly, cells were trypsinized after treatment and rinsed with PBS to achieve the final concentration of 5×10³/ml. Then, 195 µl cell suspension and 5 µl Annexin V-FITC were mixed under dark for 15 min followed by stained with 10 µl of propidium iodide (PI) for another 5 min. The apoptosis rate was assayed by flow cytometry (FACSCalibur; Bio-Rad Laboratories, Inc., Hercules, CA, USA).

After treatment, HRECs lysates were collected in RIPA lysis buffer supplemented with protease/phosphatase inhibitor cocktail (Merck KGaA, Darmstadt, Germany) and total protein concentrations were measured using the BCA assay (Beyotime Institute of Biotechnology). Protein samples were loaded on 6–12% SDS-PAGE, transferred to polyvinylidene fluoride membranes (Bio-Rad) and were blocked with 5% skim milk for 1 h at room temperature. The membranes were incubated with the corresponding primary antibodies against SIRT1 (1:400; Santa Cruz Biotechnology, Inc., Dallas, TX, USA), Bcl-2 (1:1,000), Bax (1:1,000), cleaved caspase 3 (1:400), cytochrome C (1:1,000), TLR4 (1:500), NF-κBp65 (1:1,000), p-NF-κBp65 (1:1,000), and GAPDH (1:1,000; all from Cell Signaling Technology, Inc., Danvers, MA, USA) at 4°C overnight. Subsequently, the blots were washed with PBST and incubated with a peroxidase conjugated immunoglobulin G secondary antibody (Santa Cruz Biotechnology, Inc.) for 1 h at room temperature. Signals were visualized using an enhanced chemiluminescence kit (GE Healthcare, Chicago, IL, USA).

All data are expressed as the mean ± standard deviation, and analyzed with GraphPad Prism 5 (GraphPad Software, Inc., La Jolla, CA, USA). Statistical significance was tested using one-way analysis of variance (ANOVA) with Tukey's post hoc test, Kruskal-Wallis test with Dunn's post hoc test and two-way ANOVA. <P 0.05 was considered to indicate a statistically significant difference. All experiments were performed at least three times.

Firstly, to determine whether gastrodin (Fig. 1A) itself affects the cell viability of HRECs in normal median, cells were treated with various concentration of gastrodin. The results revealed that gastrodin had no effects on cell viability even with high concentration at 100 µM (P>0.05; Fig. 1B). In addition, 100 µM gastrodin had no significant effect on cell viability within 48 h (P>0.05; Fig. 1C). Then, we found HRECs treated with HG for 24 h showed significant decrease in cell viability compared to the control group, but 100 µM gastrodin markedly ameliorated the inhibitory effect caused by HG (Fig. 1D).

HRECs apoptosis rates was measured and Fig. 2A demonstrated that, comparised with the control group, treating of HRECs with HG for 24 h significantly increased apoptosis, which could be reversed by addition of 100 µM gastrodin (Fig. 2A). Additionally, we examined the expression levels of mitochondrial apoptotic markers by Western blot analysis. Expectedly, the ratio of Bcl-2/Bax was decreased, whereas the expressions of cytochrome C and cleaved caspase 3 were increased with the treatment of HG. When the cells were treated with both HG and 100 µM gastrodin, above changes were at least partly abolished (Fig. 2B-D).

Sincere the key role of oxidative stress in HG-triggered cell apoptosis, we tested the levels of ROS production in HRECs. Intriguing, HG treatment caused increase in ROS generation compared to the control and 100 µM gastrodin alleviated ROS generation significantly (Fig. 3A). Subsequently, we tested the expression levels of oxidative stress related genes including hemeoxygenase-1 (HO-1), NAD(P)H dehydrogenase quinone 1 (NQO1), NF-E2-related factor 2 (NRF2), and γ-glutamate-cysteine ligase modifier (GCLM) via RT-qPCR. HG dramatically increased the mRNA levels of above four genes compared to the control group. However, co-treatment with HG and gastrodin resulted in a decrease in the expression of oxidative stress related genes (Fig. 3B-E).

Activation of NF-κB was shown to promote expression of various pro-apoptotic regulators in HRECs (15) and TLR-mediated NF-κB activation were commonly shown in HRECs under HG conditions (16). We thereby investigated whether TLR4/NF-κBp65 signaling pathway was involved in protective of gastrodin on HRECs. As shown in Fig. 4A, HG inhibited the increase in TLR4 protein expression level and phosphorylation of NF-κB in a concentration dependent manner. Whereas, 100 µM gastrodin inhibited the activation of TLR4/NF-κBP65 induced by HG (Fig. 4B). In addition, emerging evidence suggested that SIRT1 appears to target numerous cellular factors to regulate oxidative stress and apoptosis resulting in improving diabetic retinopathy. Fig. 4C revealed that the expression level of SIRT1 protein was decreased in a HG concentration-dependent manner. However, when the HRECs were co-treated with gastrodin with HG, SIRT1 protein expression was significantly increased as compared with that treated with HG along (Fig. 4D). Moreover, overexpression of SIRT1 was then conducted to further validate the pathway involved. As a result, SIRT1 was increased by pcDNA3.1-SIRT1 but no significant change with HG or pcDNA3.1-vector treatment were observed (Fig. 4E). Compared with vector group with no treatment, TLR4/NF-κBP65 pathway was inhibited after SIRT1 overexpressed. Besides, overexpression of SIRT1 at least partly reversed TLR4/NF-κBP65 activation induced by HG in HRECs in the vector group (Fig. 4F).