3,6’-Disinapoyl sucrose and TSFA were isolated from Radix Polygalae, identified by a combination of spectral methods (UV, IR, MS, and NMR) and purified by HPLC (purity >90%). U0126 and LY294002 were purchased from Selleck Chemicals LLC (Houston, TX, USA). Dulbecco’s modified Eagle’s medium (nutrient mixture F-12; DMEM/F-12) was obtained from Gibco (Carlsbad, CA, USA). Fetal bovine serum (FBS) was supplied by Hyclone (Logan, UT, USA). Polyclonal rabbit antibodies against BDNF, CREB, pCREB, CRTC1, and β-actin were all provided by Bioworld Technology (Atlanta, GA, USA) or Abcam (Burlingame, CA, USA).

SH-SY5Y human neuroblastoma cells were obtained from the American Type Culture Collection (Rockville, MD, USA) and cultured in DMEM/F-12 supplemented with 10% heat-inactivated FBS and 1% penicillin-steptomycin (50 U/mL). The cells were maintained in a humidified atmosphere with 5% CO₂ at 37°C. The culture medium was refreshed every other day.

Cells were seeded into 96-well plates at a density of approximately 2 × 10³ cells/well and then grown for 24 h. Subsequently, cells were cultured with serum-free medium (0.1% BSA in growth medium without FBS) for another 24 h and then treated with DISS or TFSA at desired concentrations (15.625–2,000 μM). After 48-h treatment, cell viability was determined by Cell Counting Kit-8 (CCK-8; Keygen Biotech., Nanjing, China). Results were from six independent experiments, and each experiment was run in triplicate. The data were presented as percentage relative to untreated cells. The drug response curve was generated based on the determined relative cell viability, and 50% effective concentration (EC₅₀) was measured using software Graphpad Prism5 (Roscilli et al., 2016). EC₅₀ of combination was tested on TFSA at a fixed concentration (40 μM) plus DISS at various concentrations (600, 300, 150, 75, 37.5, 18.75 μmol/L). Combination effects of DISS and TFSA treatment were evaluated by calculating the combination index (CI) value as previously reported by Chou (Chou, 2006). CI = EC_{50(DISSplusTFSA)}/EC_{50 (DISSalone)} + C_TFSA/EC_{50 (TFSAalone)}. C_TFSA was the TFSA concentration used in combination. In this analysis, synergism was defined as follows: CI > 1.1, antagonism; 1.1 < CI < 0.9, additive effect; 0.9 < CI < 0.3, synergic effect; and CI < 0.3, strong synergism.

Briefly, cells were seeded into a 96-well plate or 6-well plate at a density of approximately 3 × 10³ cells/well or 1 × 10⁴ cells/well with full growth medium. After 24 h, medium was refreshed by serum-free medium and incubated 24 h. Cells were exposed to 8 mM glutamate (Glu) for 24 h, followed by treatment with DISS (75/150 μM), TFSA (25/50 μM) alone or combinations of DISS and TFSA at desired concentrations with or without antagonists (U0126/LY294002) for 24 or 48 h (assay BNDF mRNA, western blot), and each concentration of the inhibitor was established according to a previous study with minor modifications (Hu et al., 2014).

Cells were seeded into 6-well plates at a density of approximately 1 × 10⁴ cells/well cells and grown for 24 h. Subsequently, cells were exposed to 8 mM Glu for another 24 h. After different treatments for 48 h, cells were collected and re-suspended in PBS. Cell suspension was centrifuged, and the supernatant was collected into EP tubes and then stored at low temperature (0°C). NOS and iNOS activities as well as NO content were determined using a reagent kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) using an ultraviolet spectrophotometer at a wavelength of 540 nm.

Total RNA was extracted from cells using Trizol reagent. RNA concentration was determined using an Eppendorf BioPhotometer plus (Eppendorf, Germany). Subsequently, purified RNA was reversely transcribed into cDNA using iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA, USA) according to the manufacture’s instructions. BDNF expression at the mRNA level was assessed by qPCR using SsO Fast Eva Green Supermix (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s instructions. GAPDH was selected as the housekeeping gene. The following sequences were used in the study:

Briefly, after an initial denaturation step at 95°C for 30 sec, amplifications were carried out with 40 cycle at a melting temperature of 95°C for 5 sec, an annealing temperature of 60°C for 5 sec, and an extension temperature of 65°C for 5 sec. Standard curves were generated for each gene, and the gene expression at the mRNA level was calculated relative to serial dilution of cDNA as described in Bio-Rad iQ5 System (California, USA). All experiments were conducted in triplicates. Relative expression was estimated using the 2^-ΔΔCt method.

After 48-h treatment, cells were collected and washed twice with ice-cold PBS. Subsequently, cell pellet was lyzed in lysis buffer (1% Triton, 20 mM Tris–HCl, 150 mM NaCl, 1 mM EDTA, pH 7.5) supplemented with one tablet of complete protease inhibitor cocktail (Roche, Laval, QC, Canada) and phosphatase inhibitor cocktail (Sigma, St. Louis, MO, USA), and the cell lysates were centrifuged at 1,000 g for 10 min at 4°C. Total amounts of proteins in each sample were determined by BCA kit, and the protein concentrations of all samples were adjusted to be the same. Supernatants containing 20 μg proteins were separated on 10% SDS-PAGE and electrotransferred onto polyvinylidenedifluoride (PVDF) membranes. Blots were blocked in Tris buffered saline containing 0.1% Tween-20 (TBST) and 5% non-fat-dried milk at room temperature for 1 h. Next, the membranes were incubated with the respective primary antibodies (anti-CREB 1:500, anti-phospho-CREB 1:500, anti-BDNF 1:2,000, and CRTC1 1:10,000) overnight at 4°C. Following three washes with TBST, the blots were incubated with the secondary horseradish peroxidase-conjugated antibody (1:6,000) at room temperature for 1 h, and optical density of protein bands was determined using an image analysis system (Bio-Imaging Analyzer, UVP). All the results of western blot consisted of three independent experiments, and each independent experiment was performed in triplicate.

Data were presented as mean ± SD. Comparisons were carried out using one way analysis of variance (ANOVA) followed by Tukey-Kramer’s test for post hoc analysis. All statistical analyses were performed using ANOVA by the PRISM software (GraphPad software, San Diego, CA, USA). P < 0.05 was considered as statistically significant.

Figure 1 shows that the EC₅₀ values of DISS and TFSA were 606.4 ± 23.3 μM and 237.8 ± 13.3 μM, respectively. The fixed concentration of 40 μM TFSA in combination was according to EC₂₀ of TFSA (40.71 ± 2.47 μM). The EC₅₀ of combination of DISS and TFSA was 83.86 ± 1.06 μM. CI value of co-treatment was 0.31, suggesting that combination of DISS and TFSA possessed a synergistic effect. Next, we investigated whether the combination of DISS and TFSA could have an additive or synergistic effect in preventing neuronal damage in SH-SY5Y cells. In order to lower the possible side effects of the compounds according to our previous experience, the concentrations of DISS and TFSA were set to 150/75 μM and 75/50 μM, respectively. The EC₂₀ of DISS alone was 214.2 ± 1.23 μM. In combination, we chose 150 or 75 μM DISS that lower than EC₂₀ of DISS and increased slightly TFSA concentrations based on EC₂₀ of TFSA alone.

Based on this Glu-induced decrease in cell activity, the neuronal cytotoxicity of Glu could be partially inhibited by the pretreatment of DISS or TFSA in a dose-dependent manner. More importantly, the co-treatment of TFSA and DISS at two testing doses (75 μM + 25 μM and 150 μM + 25 μM) could produce a significant neuroprotective activity as compared with individual compounds alone, exhibiting increased cell viability from 48% to more than 85%. Figures 2A,B shows that such a protective effect of DISS at 150 and 75 μM could be blocked by the MAPK/ERK1/2 inhibitor, U0126 (10 μM), but not by the PI3-K inhibitor, LY290042 (20 μM). In contrast, the LY290042, but not U0126, inhibited the neuroprotective effect of TFSA. Moreover, the inhibitory effect of combination of U0126 and LY290042 became even more obvious, showing reduced cell viability from 85% to less than 60% (Figure 2C).

In order to further understand the mechanism of the combination effect of DISS and TFSA on the neuroprotection, we investigated the expressions of CREB, CRTC1, pCREB, and BDNF. DISS or TFSA alone could induce the protein expression of pCREB/CREB at both doses, and additive effect was observed for the combination of DISS and TFSA (Figure 3A). In addition, TFSA at 25 μM had no effect on the CRTC1 expression. However, combinations of 25 μM TFSA and 150 or 75 μM DISS both significantly increased the CRTC1 expression (Figure 3B). Meanwhile, the combinations of DISS and TFSA at different ratios all significantly enhanced the BDNF expression at the mRNA and protein levels (Figures 3C,D). Moreover, BDNF expression at the mRNA level was induced for more than threefold by combinations of 150 μM DISS and 25 μM TFSA; 75 μM DISS and 50 μM TFSA; or 75 μM DISS and 25 μM TFSA compared with compounds alone (Figure 3D).

Considering the various CREB pathways activated by DISS and TFSA, we assessed the signaling mechanisms involved in CRTC1 and BDNF regulation using two inhibitors. Figure 4 reveals that treatment of DISS or TFSA increased the expression of BDNF and CRTC1, but such an increase could be partially attenuated by either U0126 or LY294002. PI3-K inhibitor LY294003 showed a similar but weaker inhibitory effect on DISS-induced CRTC1 expression compared with U0126 (Figure 4A). The LY294002 decreased the DISS- or TFSA-induced BDNF expression by 47 and 25%, respectively. The ERK1/2 inhibitor U0126 decreased the DISS- or TFSA-induced BDNF expression by 25 and 21%, respectively. (Figure 4B). A bigger decrease in DISS- or TFSA-induced BDNF expression was caused by LY294003 compared with U0126 (Figure 4B).

To further investigate the additive neuroprotection of DISS and TFSA against the Glu-induced cell injury, we examined the levels of tNOS and iNOS, which play an important role in cell apoptosis or neurodegenerative disorders (Chen et al., 2016; Zhu et al., 2016). Figure 5 shows that Glu significantly increased the levels of tNOS and iNOS. DISS or TFSA treatment alone decreased the intracellular level of tNOS by 4–8%, but showed no effect on iNOS level. However, co-treatment of DISS and TFSA significantly decreased such an elevation of tNOS and iNOS levels by 29.3 and 40.5%, respectively. These results suggested that the combination of DISS and TFSA could more vigorously inhibit the Glu-induced NO overproduction.