Dulbecco's modified Eagle's medium (DMEM), Neurobasal medium, L-Glutamine, B-27 Supplement, geneticin, penicillin, streptomycin, Lipofectamine^® 2000, Lipofectamine^® 3000 and MitoSOX Red Mitochondrial Superoxide Indicator were purchased from Thermo Fisher Scientific, Inc. Fetal bovine serum (FBS) and 0.05% trypsin-EDTA were purchased from Gibco (Thermo Fisher Scientific, Inc.). The N2a cell line was purchased from the American Type Culture Collection. The chicken anti-Ngb primary antibody was purchased from BioVendor. AKT and phosphorylated CREB (Ser133; p-CREB) antibodies were purchased from Cell Signaling Technology, Inc. The MAP2 antibody was purchased from Novus Biologicals, LLC. β-actin antibody was purchased from Sigma-Aldrich (Merck KGaA). Mouse Ngb short interfering (si)RNA1, siRNA2 and siRNA3 were purchased from OriGene Technologies, Inc. Negative control (NC) siRNA was purchased from Santa Cruz Biotechnology, Inc. The pCMV-HA vector and CREB Dominant-Negative Vector Set were purchased from Takara Bio USA, Inc. The Dual-Luciferase Reporter Assay System was purchased from Promega Corporation. The pGL3-Basic vector, the pGEM-T Easy Vector System II and the Dual Luciferase Reporter Assay System were purchased from Promega Corporation. Reporter plasmids including the P2033(−2027/+6) and P2033(CRE2del) plasmids (mutated Ngb promoter plasmid with deletion of the specific CRE site at −854) were constructed as described previously (31). PD, with a purity of 98.95%, was purchased from MedChemExpress, LLC and DMSO was used as the solvent.

The animal protocol (approval number 2016^#0101) was approved by the ethical committee of Hunan Normal University. The 10 week-old male and female C57BL/6 mice (weight, 25–30 g) were purchased from Hunan SJA Laboratory Animal Co., LTD, and bred in animal cages in the controlled environment with 12/12-h light/dark cycle, room temperature of 21–23°C and humidity of 55±10%. A total of four pregnant female mice (E16) were used for the present study. They were be anesthetized with isoflurane (5%) for ~3 min. Anesthetic adequacy was checked continuously by observing the respiration rate and checking tailpinches. Then they were sacrificed via immediate decapitation. Once it was confirmed that the mouse was dead, the fetuses were surgically remove. The fetal mouse pups were immediately decapitated, as immediate decapitation is the fastest, most humane and standard procedure to sacrifice fetal mice for purposes of extracting brains and growing primary neuron culture. Animals were removed from the study and sacrificed if any of the following clinical signs/conditions were found: Persistent recumbence; inability to rise; loss of righting reflex; pain or distress that cannot be alleviated by analgesics; difficulty with ambulation (paralysis, fractures and trauma); severe central nervous system signs (including circling, rolling, persistent seizures or convulsions); abnormal breathing (dyspnea) and cyanosis; body condition score (32) of ≥2 (out of 5); excessive weight loss; vomiting/diarrhea resulting in severe dehydration; and tumor production specific endpoints (maximum tumor diameter, ≤20 mm) (33). As described previously (29), primary mouse cortical neuronal cultures were prepared from the cortices of E16 C57/BL6 female mouse embryos and plated onto 12-, 24-, or 48-well plates precoated with poly-D-lysine (Sigma-Aldrich; Merck KGaA) at a density of 3.5×10⁵ cells/ml in neurobasal medium (NBM) plus 2% B27, 0.8 mM L-glutamine and 100 µg/ml penicillin/streptomycin (Thermo Fisher Scientific, Inc.). Half of the medium was changed every 3 days. The purity of primary mouse cortical neurons on day 8 was determined by calculating the ratio of MAP2-positive cells to the total number of viable cells (Fig. S1). These data indicated that >95% of the cells in the cultures were neurons.

Mouse N2a neuronal cells were cultured in complete DMEM (DMEM containing 10% FBS and 100 µg/ml penicillin/streptomycin) and maintained in a humidified (5% CO₂, 37°C) incubator. For transfection, N2a cells at ~80% confluence in 12 wells plates were transfected with 30 pmol NC siRNA or Ngb siRNAs (siNgb1 Target sequence: GCCAGGCTCTTCGCCCTGGAA, siNgb2 Target sequence: GCCAAGCTGTTTACTTTCCAA, siNgb3 Target sequence: GGTGATGCTAGTGATTGAT) with 3 µl Lipofectamine^® 3000 or 3 µg plasmids [mouse Ngb promoter plasmid [P2033(−2027/+6)], pCMV-CREB or pCMV-CREBs133a or pCMV-KCREB] with 3 µl Lipofectamine^® 2000 at 37°C according to the manufacturer's instructions for 6 h, followed by replacement of medium with fresh complete DMEM and culture for another 18 or 42 h. The transfected cells were treated with 200 µM H₂O₂ in the presence or absence of 10 µM PD at 37°C for 16 h for MTT assays, or for 4 h for JC-1 mitochondrial membrane potential and ROS production assay.

An MTT assay was carried out as described previously (29). After different treatments, N2a cells were incubated with fresh DMEM containing 0.5 mg/ml MTT for 4 h in a 37°C cell culture incubator. Next, the medium was aspirated and DMSO was added to each well to dissolve the MTT formazan crystals. The absorbance of each sample at 570 nm was measured with a Molecualr Device Spectramax M5 plater reader. Percentages of live cell counts were used for assay normalization.

LDH assays were used to measure neurotoxicity as previously described (27). Following treatment, the medium was removed and the cells were lysed with PBS containing 0.1% Triton at room temperature for 5 min. Subsequently, the lysate was collected and centrifuged at 380 × g for 10 min at 4°C. The supernatant was used for LDH activity measurement using a cytotoxicity detection kit (LDH) following the manufacturer's instructions (Roche Applied Science).

∆Ψm measurements were performed as described previously (29). Following treatment, primary mouse cortical neurons or N2a cells were stained with 2.0 µg/ml 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolyl-carbocyanine iodide (JC-1) at 37°C for 20 min, followed by washing with Hanks' Balanced Salt Solution (HBSS) three times. Then, the cellular fluorescence was detected with a microplate reader (green fluorescence, excitation 489 nm and emission 529 nm; red fluorescence, excitation 570 nm and emission 600 nm) and imaged with a fluorescence microscope with 200× magnifications. Four random microscopic fields in each group were selected. The ∆Ψm was calculated as the fluorescence ratio of red (JC-1 aggregates) to green (JC-1 monomers).

Mitochondrial ROS production was assessed with MitoSOX Red Mitochondrial Superoxide Indicator according to the manufacturer's instructions. Briefly, following treatment, the medium was aspirated and HBSS was added to primary mouse cortical neurons or N2a cells, which were diluted with MitoSOX Red dye (10 µM) and incubated in a cell incubator at 37°C in the dark for 30 min. After loading the dye, the cells were washed twice with HBSS and the cellular fluorescence was imaged with a fluorescence microscope with 200× magnifications. Four random fields in each group were selected. The mitochondrial ROS was also measured at 510 nm excitation and 580 nm emission with a microplate reader.

N2a cells were seeded into 24-well plates and transfected with the indicated plasmids, including the mouse Ngb promoter plasmid P2033(−2027/+6) or mutated Ngb promoter plasmid P2033(CRE2del) plasmid (200 ng/well) and pRL-TK (10 ng/well). The wild-type and mutated mouse Ngb promoter were cloned into the pGL3-Basic vector (Promega Corporation) as previously described (31). After 24 h, cells were treated with H₂O₂ or PD as described. Reporter assays were performed with Dual Luciferase Reporter Assay System (Promega Corporation). All assays were carried out in a Veritas Microplate Luminometer (Turner BioSystems). The activity was normalized and defined as the Firefly/Renilla luciferase ratio. Each experiment was performed at least three times.

Cells were lysed in RIPA buffer and centrifuged at 13,500 g for 10 min at 4°C. Supernatants were measured with a Bradford assay (Bio-Rad Laboratories, Inc.) and 30 µg of total protein were separated using a 10% SDS-PAGE gel and then transferred to a polyvinylidene difluoride membrane. The membrane was blocked in 5% nonfat dried milk in TBST (TBS containing 0.1% Tween-20) for 1 h at room temperature, followed by immunoblotting and incubation overnight at 4°C with the primary antibodies as follows: CREB (1:1,000, Cell Signaling Technology, cat. no. #4820), phospho-CREB antibody (1:1,000; Cell Signaling Technology, #9198), Ngb (1:500; Abnova, 2B5-A7), β-actin antibody (1:3,000; Sigma, A5441). Subsequently, the membrane was incubated with the appropriate horseradish peroxidase-conjugated secondary antibody (both 1:5,000; cat. nos. #31430 and #31460; Thermo Fisher Scientific, Inc.) for 1 h at room temperature. After washing with TBST, immunoreactive proteins were determined by an enhanced chemiluminescence substrate (Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Images were captured with ChemiDoc Imaging Systems (Bio-Rad Laboratories). Quantification of protein band intensity obtained by Western blotting analysis was analyzed with ImageJ 1.52a (http://rsb.info.nih.gov/ij/) and normalized to the actin band intensity.

An EMSA was performed as described previously (29,34). N2a cell nuclear extracts were prepared with a NE-PER Nuclear and Cytoplasmic Extraction kit (Thermo Fisher Scientific, Inc.). Oligonucleotides containing wild-type CRE2 (5′-TGGGGGCTGAGGTGATCCAAC-3′) were synthesized and labeled with biotin using a biotin 3-end DNA-labeling kit (Thermo Fisher Scientific, Inc.) and annealed to obtain double-stranded DNA probes. DNA/nuclear protein/antibody binding reactions for the supershift assay were performed using the LightShift EMSA Optimization and Control kit (Thermo Fisher Scientific, Inc.). The reaction mixtures were loaded onto a 6% nondenaturing polyacrylamide gel and electrophoresed in 0.5× TBE buffer at 100 V at 4°C. Following electrophoresis, the DNA/nuclear protein/antibody complexes were transferred to a nylon membrane (Ambion; Thermo Fisher Scientific, Inc.), followed by cross-linking of the membrane at 120 mJ/cm² for 60 sec using a UV cross-linker (Stratagene; Agilent Sumitomo Dainippon Pharma Co., Ltd.). Biotin-labeled DNA was further detected using the Chemiluminescent Nucleic Acid Detection Module (Thermo Fisher Scientific, Inc.) and exposed to film (X-Omat; Kodak).

All data are expressed as the mean ± standard deviation of ≥3 independently repeated experiments. Differences between groups were analyzed by one-way ANOVA followed by Tukey-Kramer's test using GraphPad Prism 7 software (GraphPad Software, Inc.). P<0.05 was considered to indicate a statistically significant difference.

Primary cultured mouse cortical neurons and mouse N2a cells were treated with H₂O₂ at different doses and neuronal cell viability was measured by MTT assay. The data showed that treatment with H₂O₂ for 16 h significantly reduced cell viability of primary mouse cortical neurons and N2a cells in a dose-dependent manner (Fig. 1A and B). It was found the two cells have a different sensitivity to PD. Primary mouse cortical neurons are more sensitive to PD than N2a cells. It was found that ~30–40% neuronal cell death was induced by 50 and 200 µM H₂O₂ in primary mouse cortical neurons and N2a cells, respectively (Fig. 1A and 1B) and these H₂O₂ concentrations were selected for further experiments. To investigate whether PD can protect neuronal cells from oxidative stress, MTT assay and LDH assay were used for measurement of cytotoxicity. The primary mouse cortical neurons are non-dividing cells, but N2a cells are proliferated cells. If PD has a promoting effect on the proliferation of N2a cells, there would be more intracellular LDH in N2a cells, which may be accompanied by more LDH release. The present study thus performed an LDH assay with primary mouse cortical neurons and MTT assay for N2a cells, respectively. Primary mouse cortical neurons were treated with H₂O₂ (50 µM) plus PD (0, 5, 10 and 20 µM) for 16 h, followed by the MTT assay. The results showed that PD dose-dependently inhibited H₂O₂ (50 µM)-induced neurotoxicity in primary mouse cortical neurons. Compared with the control, 10 and 20 µM PD could inhibit H₂O₂-induced neurotoxicity. However, the cell viability of these two concentrations was significantly different compared with the control. Considering that the primary mouse neurons are more sensitive to PD treatment, there may be more reverse effects with a high dose of PD in primary mouse neurons; 10 µM PD exhibited the maximal protective effects among the tested doses (Fig. 1C). Moreover, LDH assays showed that 10 µM PD significantly prevented H₂O₂ (50 µM)-induced LDH release in primary mouse cortical neurons (Fig. 1D). Similarly, treatment with PD (10 µM) significantly attenuated neurotoxicity in N2a cells, which was induced by H₂O₂ at a dose of ≤400 µM (Fig. 1E). These results indicated the potent neuroprotective effects of PD.

Mitochondrial dysfunction has been proposed as a mediator of neurodegeneration (35,36). To investigate whether PD attenuates H₂O₂-induced Δψm, a JC-1 assay was performed to determine the Δψm in primary mouse cortical neurons treated with H₂O₂ (50 µM) or H₂O₂ (50 µM) plus PD (10 µM). It has been reported that the mitochondrial dysfunction that occurs as early as 4 h can affect neuronal cell viability (35). Δψm was thus detected at the early time point 4 h. The results showed that the JC-1 red/green fluorescence ratio in neurons significantly decreased after H₂O₂ (50 µM) treatment compared with the untreated control, while this decrease was significantly attenuated by cotreatment with PD (10 µM; Fig. 2A). Similar observations were also found in N2a cells, as PD (10 µM) treatment significantly preserved the Δψm, which was dramatically reduced by H₂O₂ (200 µM) treatment (Fig. 2C). To further investigate whether PD inhibits H₂O₂-induced mitochondrial oxidative stress, MitoSOX dye was used. The results showed that PD treatment significantly attenuated H₂O₂-induced mitochondrial ROS production in primary mouse cortical neurons (Fig. 2B) and N2a cells (Fig. 2D). Taken together, these results indicated that PD attenuates H₂O₂-induced mitochondrial dysfunction and oxidative stress in primary mouse cortical neurons and N2a cells.

To investigate the molecular mechanism underlying the neuroprotective effects of PD, western blotting analysis was performed. The results showed that PD significantly induced the phosphorylation of AKT and CREB and Ngb protein expression in a dose-dependent manner and that there was no effect on the AKT and CREB protein expression (Fig. 3A). H₂O₂ induces neuronal cell death by reducing the levels of CREB and p-CREB (37). In agreement with this study, the results of the present study showed that H₂O₂ (200 µM) treatment led to a significant reduction in p-AKT, p-CREB and Ngb expression in N2a cells and that there was no effect on the AKT and CREB protein expression. Nevertheless, this decrease was rescued by cotreatment with PD (10 µM) (Fig. 3B). Taken together, these results suggested that PD attenuates H₂O₂-reduced p-AKT, p-CREB and Ngb expression in N2a cells.

To investigate whether CREB is required for PD-induced Ngb upregulation, a mouse Ngb promoter plasmid [P2033(−2027/+6)] and wild-type CREB or mutant CREB (CREBs133a or KCREB) were cotransfected into N2a cells for 24 h, followed by PD (10 µM) treatment for another 16 h. The results showed that mutant CREB significantly attenuated PD-induced Ngb promoter activity (Fig. 4A). Moreover, it was observed that KG501, a potent inhibitor of CREB, significantly attenuated PD-induced Ngb promoter activity (Fig. 4B). These results indicated that CREB is responsible for PD-induced Ngb promoter activity. To determine whether PD promotes the binding of CREB to CRE in the mouse Ngb promoter, EMSA was performed. The results showed that compared with DMSO-treated N2a cell nuclear extracts, PD-treated N2a cell nuclear extracts produced a stronger supershift band after incubation with p-CREB antibody and specific Ngb CRE oligos, indicating that PD significantly induces the binding of p-CREB to the specific CRE site located at −854 in the promoter region of the mouse Ngb gene (Fig. 4C). To further investigate whether the specific CRE site in the Ngb promoter region is responsible for PD-induced Ngb promoter activity, P2033(−2027/+6) or P2033(CRE2del, mutated Ngb promoter plasmid with deletion of the specific CRE site at −854) plasmids were transfected into N2a cells for 24 h followed by PD treatment for another 16 h. The results showed that deletion of the specific CRE site significantly attenuated basal Ngb promoter activity and PD-induced Ngb promoter activity (Fig. 4D). Taken together, these results suggested that PD upregulates Ngb via CREB in N2a cells.

Next, the role of Ngb in mediating PD protection against H₂O₂ (200 µM) was determined. Transfection of Ngb-specific siRNA (siNgb-2 and siNgb-3) into N2a cells for 48 h successfully led to the knockdown of endogenous Ngb (Fig. 5A) and siNgb-2 was chosen for further experiments. It was found that knockdown of Ngb significantly attenuated the protection of PD against H₂O₂ via an MTT assay (Fig. 5B). Next, the role of endogenous Ngb in mediating PD-preserved mitochondrial membrane potential in N2a cells was determined. It was found that the JC-1 ratio was significantly reduced in Ngb siRNA-transfected N2a cells compared with NC siRNA-transfected N2a cells under PD plus H₂O₂ treatment conditions (Fig. 5C). Furthermore, it was observed that there was a significant increase in ROS production in N2a cells treated with siNgb-2 transfection plus PD and H₂O₂ compared with NC siRNA transfection plus PD and H₂O₂ (Fig. 5D). These results indicate that PD inhibits neuronal cell death, maintains mitochondrial function and reduces ROS production via Ngb.