The murine microglia cell line, BV-2, has been used previously as a substitute for primary microglia cells, as it exhibits very similar behavior. The C17.2 NSC line is capable of self-renewal and differentiation and has been used as a model system or cell therapy for neurodegenerative diseases (Wang et al., 2015). In this study, we used murine C17.2 cells line and BV-2 cells line as replacements for primary NSCs and microglia (MG), respectively. C17.2 cells were cultured at 37°C, 5% CO₂ in high glucose DMEM (Gibco, Grand Island, NY, USA) supplemented with 10% (v/v) fetal bovine serum (Gibco, Grand Island, NY, USA), 5% (v/v) horse serum (Gibco, Grand Island, NY, USA) and 2 mM glutamine (Invitrogen, Carlsbad, CA, USA). BV-2 cells were propagated in flasks containing DMEM supplemented with 10% fetal bovine serum, at 37°C with 5% CO₂.

The microglia were cultured in transwell co-cultured system (Corning, NY, USA) that was placed above the NSC layer. Aβ peptide was added to a final concentration of 10 μM in microglia layer. The NSCs and microglia shared the same medium but no direct cell-cell interactions were possible due to the physical separation of the cells by a 0.4 μm polycarbonate membrane. We observed the response of the NSCs to the diffusible oxidative stress factors secreted by the stimulated microglia.

Experiments were performed on the following groups: (1) NSC: NSCs were cultured in the low chamber only for 48 h; (2) MG + NSC: NSCs were co-cultured with microglia for 48 h; (3) Aβ + NSC: NSCs were cultured in the low chamber with the addition of Aβ_1–42 (10 μM, Sigma, San Francisco, CA, USA) for 48 h; (4) Aβ + MG + NSC (Control): NSCs were co-cultured with microglia with the addition of Aβ_1–42 (10 μM) into the microglia chamber for 48 h, also known as the Control group in followed SIRT3 overexpression and interference tests; (5) Scrambled or siSIRT3: NSCs in the chamber were pretreated with negative control small interfering RNA (siRNA) or SIRT3 siRNA (Santa Cruz Biotechnology, Santa Cruz, CA, USA) for 6 h, respectively, then NSCs were co-cultured with microglia with the addition of Aβ_1–42 (10 μM) into the microglia chamber for 48 h; and (6) Vector or SIRT3: NSCs in the chamber were pretreated with empty vector or SIRT3 expression plasmid (Vigenebio, Rockville, MD, USA) for 6 h, respectively, then NSCs were co-cultured with microglia with the addition of Aβ_1–42 (10 μM) into the microglia chamber for 48 h.

The levels of TNF-α, IL-1β and IL-6 in the cell culture medium supernatants of co-culture system were detected with appropriate enzyme linked immunosorbent assay (ELISA) kits per manufacturer’s instructions. TNF-α kits was purchased from R&D System (Minneapolis, MN, USA), IL-1β and IL-6 kits were purchased from Genetimes Technology (Shanghai, China). The level of NO in supernatants was measured by Griess reagent. Griess reagent was obtained from Beyotime Biotechnology (Shanghai, China). The concentration of soluble cytokines in each sample was determined from a standard curve generated by using positive controls provided in the kits.

The recombinant adenoviral vector overexpressing SIRT3 (SIRT3 group) and the empty vector (Vector group) were produced by Vigenebio (Rockville, MD, USA). When the NSCs reached 50% confluence, recombinant adenovirus was added at a multiplicity of infection of 100 for 6 h. The SIRT3 and manganese superoxide dismutase (MnSOD) genes expression in NSCs were evaluated by quantitative reverse transcription-polymerase chain reaction (qRT-PCR) and western blotting. Negative control siRNA (Scrambled group) and SIRT3 siRNA (siSIRT3 group) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). NSCs were transiently transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions.

The fluorescent probe dihydroethidium (DHE; Beyotime Biotechnology, Shanghai, China) and hydrogen peroxide assay kit (Beyotime Biotechnology, Shanghai, China) were used to evaluate the intracellular accumulation of superoxide anion and hydrogen peroxide according to the manufacturer’s recommendations, respectively. Briefly, experiments were performed in 24-wells in which cells were incubated with 5 μM DHE for 30 min at 37°C, or hydrogen peroxide reaction solution for 30 min at room temperature. Cells were subsequently washed with PBS, lysed and transferred to a 96-well black plate for measurement of fluorescence intensity with a plate reader (Thermo Fisher Scientific, Waltham, MA, USA).

Total RNA was isolated with TRIzol regent (Life Technology, Carlsbad, CA, USA). The expression of mRNA level was analyzed by qRT-PCR analysis as described previously (Wang et al., 2015). Specific primers were synthesized by Sangon Biotech (Shanghai, China), and the sequences were described as follows:

Western blotting was performed as previously described (Wang et al., 2015). Briefly, cell protein extracts (15–20 μg) were separated by 10–15% SDS-PAGE and transferred to PVDF membranes. Proteins were detected with the following antibodies: SIRT3 (1:1000, Cell Signaling Technology, Boston, MA, USA), MnSOD (1:1000, Proteintech, Chicago, IL, USA), cyclophilin D (CypD; 1:1000, Abcam, Cambridge, UK), cytochrome C (Cyt C; 1:1000, Cell Signaling Technology), cleaved caspase-3 (1:1000, Cell Signaling Technology), Bax (1:1000, Cell Signaling Technology, Boston, MA, USA), B-cell lymphoma 2 (Bcl-2; 1:1000, Cell Signaling Technology, Boston, MA, USA), COX-IV (1:1000, Abcam, Cambridge, MA, USA) and GAPDH (1:2000, Proteintech, Chicago, IL, USA).

Cell apoptosis of NSCs was analyzed with FACSCalibur flow cytometer (BD, USA) using Annexin-V/propidium iodide (PI) assays according to the manufacturer’s instructions (BD Biosciences, Franklin Lakes, NJ, USA) as described previously (Wang et al., 2015). The NSCs were harvested, washed and then fixed with 70% ethanol. The cell fixation took overnight time in −20°C freezer. The next day, the fixed cells were centrifuged to collect pellet, washed with ice-cold PBS and resuspended with staining buffer containing 50 mg/mL PI, 0.1% Triton X-100, 0.1% sodium citrate and 100 mg/mL RNase. The cell suspension was incubated in the dark condition for 30 min at room temperature. The stained cells were then analyzed with flow cytometer.

The NSCs were incubated with JC-1 (Beyotime Biotechnology, Shanghai, China) at 37°C for 20 min and were then washed 2× with the dyeing buffer. The fluorescence level was measured immediately at a single excitation wavelength (488 nm) and dual emission wavelengths (shift from 530 nm to 590 nm) using a high-content screening platform (Thermo Fisher Scientific, Waltham, MA, USA).

NSCs were loaded with calcein-AM (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and CoCl₂ (Sigma-Aldrich, St Louis, MO, USA), incubated for 15 min at 37°C. Cells were centrifuged, washed, and cell pellets were resuspended in 0.4 ml PBS and analyzed for mitochondrial calcein-AM fluorescence by flow cytometry analysis. The excitation/emission fluorescence for calcein-AM is 494/517 nm. There is an inverse relationship between calcein-AM fluorescence intensity and the number of mitochondrial permeability transition pore (mPTP) opening.

Caspase-3 and caspase-9 enzyme activities were assayed by the spectrophotometric method according to the manufacturer’s instructions, the caspase-3 and caspase-9 assay kits were obtained from Beyotime (Shanghai, China).

Data were expressed as mean ± standard deviation and analyzed using SPSS version 20.0 (IBM, Armonk, NY, USA). The Kruskal-Wallis analysis followed by Dunn-Bonferroni post hoc comparisons were used to compare the difference in multiple groups. Values were considered significant at p < 0.05.

Our previous studies demonstrated that microglia activation-derived inflammatory cytokines could effectively inhibit proliferation and promote apoptosis of NSCs (Wang et al., 2015; Jiang et al., 2016). We want to see whether microglia activation brings about oxidative stress injury to NSCs. The effects of microglia activation on the intracellular ROS production, cell apoptosis, cell cycle distribution, mitochondrial ΔΨm and the opening of mPTP in NSCs were explored.

The levels of NO, TNF-α, IL-6 and IL-1β in cell culture medium supernatant of co-culture system were all significantly elevated after the addition of Aβ for 48 h (Figure 1). Results of assays using the fluorescent probes DHE and hydrogen peroxide reagent kit showed that microglia activation remarkably elevated the intracellular ROS levels in NSCs (Figure 2A). The results of the annexin V-FITC/PI double-staining assays indicated that microglia activation dramatically increased the apoptosis rate of NSCs (Figure 2B). As shown in Figure 2C, microglia activation resulted in the ascending proportion of G₀/G₁ phase cells significantly, and the proportion of S and G₂/M phase cells descended markedly, which tested by flow cytometry. Assays using the fluorescent probes JC-1 and calcein-AM verified that microglia activation promoted ΔΨm depolarization (Figure 2D) and mPTP opening in NSCs (Figure 2E).

To determine the effects of microglia activation-induced oxidative stress on the SIRT3 and MnSOD expression in NSCs. MnSOD, a downstream target gene of SIRT3, is a critical antioxidant enzyme that eliminate ROS in mitochondria. Western blotting and qRT-PCR assays were performed to assess the SIRT3 and MnSOD gene expression. Figures 2F,G displayed that microglia activation made a significant reduction of SIRT3 and MnSOD expression in both mRNA and protein levels. These results elucidated that microglia activation resulted in oxidative stress injury, mitochondrial dysfunction and decreased expression of SIRT3 and MnSOD in NSCs.

To evaluate whether SIRT3 knockdown in NSCs deteriorates microglia activation-induced cytotoxicity. After transfection with the SIRT3-siRNA (siSIRT3 group), the SIRT3 and MnSOD mRNA levels in NSCs significantly dropped by about 75% compared to that in untransfected cells (control group) or in cells transfected with the negative siRNA (scrambled group), and the SIRT3 and MnSOD protein levels in NSCs of SIRT3 group were also lower than that in control or scrambled group (Figures 3A,B).

The results depicted in Figure 3C manifested that SIRT3 knockdown further accelerated microglia activation-induced ROS production. The rate of NSCs apoptosis in siSIRT3 group was much higher than that in the control or scrambled group (Figure 3D). SIRT3 knockdown further increased the proportion of G₀/G₁ phase cells significantly, while S phase cells proportion declined dramatically compared to that in the control or scrambled group (Figure 3E). Assays using the fluorescent probes JC-1 and calcein-AM illuminated that SIRT3 knockdown promoted microglia activation-induced ΔΨm depolarization (Figure 3F) and mPTP opening (Figure 3G) in NSCs, consistent with enhanced levels of CypD protein expression (Figure 3H). In summary, these findings provided important evidence that inhibition of SIRT3 expression exacerbated microglia activation-induced cell damage in NSCs.

To confirm the protective role of SIRT3 overexpression in NSCs against microglia activation-induced cytotoxicity. After transfection with the SIRT3-expressing plasmid (SIRT3 group), the SIRT3 and MnSOD mRNA levels in NSCs were remarkably elevated by approximately 8-fold compared with that in untransfected cells (control group) or in cells transfected with the empty vector (vector group), the SIRT3 and MnSOD protein levels in NSCs of SIRT3 group were also higher than that in control or vector group (Figures 4A,B). SIRT3 overexpression significantly suppressed microglia activation-dependent ROS production (Figure 4C). The rate of cell apoptosis in SIRT3 group was lower than that in control or vector group (Figure 4D). SIRT3 overexpression also declined the proportion of G₀/G₁ phase cells significantly, while S and G₂/M phase cells proportion increased dramatically (Figure 4E). SIRT3 overexpression suppressed microglia activation-dependent ΔΨm depolarization (Figure 4F) and mPTP opening (Figure 4G) in NSCs, consistent with decreased protein levels of CypD expression (Figure 4H). These data illustrated that SIRT3 overexpression in NSCs could ameliorate microglia activation-induced cell injury.

In order to explore the molecular mechanism underlying the neuroprotective effect of SIRT3 in NSCs against microglia activation-induced cytotoxicity. We next sought to investigate mitochondrial apoptotic pathway in oxidative stress-mediated central nervous diseases, the levels of several key molecules in the mitochondrial apoptotic pathway were measured by western blotting assay. The results depicted in Figure 5A revealed that protein levels of total cleaved caspase-3, Bax and cytoplasmic Cyt C decreased significantly in NSCs of SIRT3 group, which correlated with increased protein levels of total Bcl-2 and mitochondrial Cyt C compared to the control or vector group. On the contrary, compared to the control or scrambled group, SIRT3 knockdown in NSCs of siSIRT3 group made the protein levels of total cleaved caspase-3, Bax and cytoplasmic Cyt C up-regulated, the protein levels of total Bcl-2 and mitochondrial Cyt C down-regulated (Figure 5C).

In addition, the effect of SIRT3 on microglia activation-induced oxidative stress was also assessed by measuring the enzymatic activities of caspase-3 and caspase-9 in NSCs, which are pivotal pro-apoptotic factors. As shown in Figures 5B,D, SIRT3 overexpression effectively weakened the two enzymatic activities of caspase-3/9, while SIRT3 knockdown markedly enhanced them. These findings suggested that SIRT3 in NSCs alleviated microglia activation-induced oxidative stress injury through mitochondrial pathway, including the reduction of mitochondrial Cyt C released into cytoplasm, the decrease of Bax protein expression and caspase-3/9 enzymatic activity, along with the increase of Bcl-2 protein expression.