Homozygous nNOS-deficient mice (B6; 129S4-Nos1^tm1Plh, nNOS^-/-) and their wild-type controls of similar genetic background (B6129SF2, WT) (both from Jackson Laboratories) were maintained in Model Animal Research Center of Nanjing University. Adult male (2-month-old) and embryonic ICR mice (Shanghai Silaike Laboratory Animal) were also used in this study. All experimental protocols using animals were approved by the Institutional Animal Care and Use Committee of Nanjing Medical University.

Embryonic NSCs were isolated from embryonic day 14 (E14) mouse cortex as we previously described (Luo et al., 2010), with minor modifications. Cells were floating cultured in proliferation medium, DMEM/F12 medium (1:1; Invitrogen) containing 20 ng/ml basic fibroblast growth factor (bFGF; Sigma-Aldrich), 20 ng/ml epidermal growth factor (EGF; Sigma-Aldrich), and 2% B27 supplement (Invitrogen), and passaged every 4–6 days. Embryonic NSCs of the second to fifth passage were used in this study.

Adult NSCs were isolated and cultured as previously described (Luo et al., 2010), with some modifications. In brief, the dentate gyri of 2-month-old male mice (10 mice were used in each primary isolation experiment) were dissected and digested with 0.125% trypsin (Invitrogen) at 37°C for 15 min. Then the digestion was terminated with DMEM/F12 medium containing 10% fetal bovine serum (FBS; TBD), and serum-free DMEM/F12 medium was used to wash remanent FBS. Finally, cells were centrifuged and resuspended in proliferation medium, seeded at 1 × 10⁶ cells/ml, and maintained at 37°C with 5% CO₂. Adult NSCs of the second to fourth passage were used in this study.

Monolayer-cultured NSCs were allowed to differentiate in differentiation medium (DMEM/F12 medium containing 2% B27 and 0.5% FBS). Cells underwent different treatments and were fixed at different time points for immunofluorescence label.

For HDAC2 expression and enzymatic activity assay, neurospheres were planted in differentiation medium in cell culture dishes coated with polyornithine (10 μg/ml; Sigma-Aldrich), and were immediately treated with vehicle or 100 μM N⁵-(1-imino-3-butenyl)-L-ornithine (L-VNIO; Alexis Biochemicals) for 24 h. Then the protein was extracted.

Cell proliferation was assessed by cell counting for the neurosphere-cultured NSCs and by bromodeoxyuridine (BrdU) incorporation for the monolayer-cultured NSCs. For cell counting, single-cell suspension was seeded at 2 × 10⁴ cells/ml. Four days later, new-generated neurospheres were dissociated to single cells and the number of cells was counted on a hemocytometer. Data were normalized to the percentage of control.

For BrdU incorporation, NSCs were plated on coverslips coated with polyornithine and laminin (5 μg/ml; Invitrogen) and cultured as a monolayer. The cells were treated with 2.5 μM BrdU (Sigma-Aldrich) during the later 2 days of the 4-day differentiation, or 10 μM BrdU during the last 2 h of the 24-h proliferation culture, or 10 μM BrdU for 24 h for identification of cultured adult NSCs. Then the cultures were fixed in phosphate-buffered solution (PBS) containing 4% paraformaldehyde for 15 min at room temperature and BrdU⁺ cells were visualized by immunofluorescence label.

Fixed cultures were blocked in PBS with 3% goat serum, 0.3% Triton X-100, and 0.1% bovine serum albumin at room temperature for 1 h, followed by incubation with primary antibody at 4°C overnight and then secondary antibody for 2 h at room temperature. The primary antibodies used were as follows: mouse anti-nestin (1:100; sc-33677; Santa Cruz Biotechnology), mouse anti-β-III-Tubulin (1:200; MAB1637; Millipore Bioscience Research Reagents) or mouse anti-glial fibrillary acidic protein (GFAP; 1:1000; MAB360; Millipore Bioscience Research Reagents). The fluorescent secondary antibody used was goat anti-mouse DyLight488 (1:400; 515-485-062; Jackson ImmunoReasch). Finally, Hoechst 33258 (Sigma-Aldrich) was used to label the nuclei.

For BrdU immunofluorescence, fixed membranes were first ruptured by 0.2% Triton X-100. Then, the cells were denatured in 2 M HCl (20°C for 10 min) and rinsed in 0.1 M boric acid (pH 8.5) for 10 min before blocking. The next steps were the same to those mentioned above in immunofluorescence. The primary antibody was mouse anti-BrdU (1:1000; MAB4072; Millipore Bioscience Research Reagents), and the secondary antibody was goat anti-mouse Cy3 (1:200; 115-165-003; Jackson ImmunoReasch).

All these images of immunostained cells were captured with a fluorescence microscope (Axio Imager; Zeiss). The percentages of neurons, astrocytes and BrdU-labeled dividing cells were calculated in 40 high-power fields systematically across the coverslip. The cells were counted using Image-Pro (Media Cybernetics).

Cell apoptosis and necrosis assay kit (Beyotime) was used to detect dead cells during NSC differentiation. Living cultures were stained with propidium iodide (PI) which stains dead cells and Hoechst 33342 which stains live and dead cells for 20 min at 37°C and imaged with a fluorescence microscope at 40×.

A mathematical description of neurogenesis was used to analyze the effect of nNOS from NSCs on neuronal fate as previously described (Song et al., 2002; Hsieh et al., 2004a), with some modifications. PI_j is the number of dead cells present at the start of the j day in culture, N_j is the total number of cells present at that time, δ_j is the death rate of all cells. δ_j is given by the equation δ_j N_j = PI_j+1 – PI_j. In addition, n_j+1 is the total number of neurons present at the start of the (j+1) day, and it is given by the equation n_j+1 = n_j + β_j (N_j-n_j) – δ_jⁿn_j, where β_j is the conversion rate of progenitors to neurons for that day and δ_jⁿ is the death rate of neurons. The second term on the right is the number of neurons generated by conversion of neural progenitors, and the third term is the number of dead neurons. If δ_jⁿn_j = δ_j N_j, that is all dead cells are neurons, we can obtain β_j.

Western bolt analysis was performed as described previously (Luo et al., 2007). The primary antibodies were as follows: mouse anti-nNOS (1:600; 610308; BD Biosciences), mouse anti-HDAC2 (1:2000; ab51832; Abcam) or mouse anti-GAPDH (1:2500; KC-5G4; KangChen Bio-tech). Appropriate horseradish peroxidase-linked secondary antibodies were used for detection by enhanced chemiluminescence (Pierce).

Histone deacetylase activity was assessed using a fluorometric HDAC assay kit (EMD Millipore). For HDAC2-specific activity, immunoprecipitation with the specific antibody was performed before the assay as described previously (Guan et al., 2009; Luo et al., 2014). Briefly, 0.5 μg of mouse anti-HDAC2 antibody (Abcam) was added to the cellular lysates and incubated for 12 h at 4°C. Then 20 μl of protein G-Agarose (Sigma-Aldrich) was added and incubated on a tube rotator for 4 h at 4°C. Beads were centrifuged at 2500 × g and washed five times in immunoprecipitation buffer (50 mM Tris-HCl, 150 mM NaCl, 5 mM EDTA, 0.5% NP40, pH 8.0, supplemented with 1 mM PMSF). Finally, HDAC assay substrate was added to the beads and incubated at 30°C for 45 min, after which the reaction was stopped and fluorescence was measured according to the instructions of HDAC assay kit.

Comparisons were made with one-way ANOVA followed by Scheffe’s post hoc test. Data were presented as mean ± SEM. Differences were considered significant when p < 0.05.

To investigate the role of endogenous nNOS in neuronal differentiation, we cultured embryonic NSCs from nNOS^-/- and WT mice. Monolayer-cultured NSCs proliferated for 24 h and then were allowed to differentiate for 4 days. We found that although nNOS gene knockout had much weak effect on the percentage of astrocytes (GFAP⁺; Figures 1A,B), it substantially reduced neuronal differentiation (β-III-Tubulin⁺) (WT 20.39 ± 1.20% vs. nNOS^-/- 11.21 ± 2.69%, p < 0.05; Figures 1C,D), indicating that nNOS in NSCs is necessary for neuronal differentiation in vitro.

The reduction of neuronal differentiation by nNOS gene knockout could be due to some combination of decreased proliferation of NSCs or progenitors, attenuated neuronal survival, and/or inhibition of fate commitment by NSCs to a neuronal lineage. First, we examined whether a decrease in cell proliferation or survival might explain the reduced number of neurons in nNOS-repression cultures. L-VNIO, a highly selective inhibitor of nNOS (Babu and Griffith, 1998), was used to inhibit nNOS enzymatic activity. Cultures were treated with 100 μM L-VNIO or vehicle for the later 2 days during the 4-day differentiation of monolayer-cultured embryonic NSCs, immediately followed by fixation and staining with antibody against β-III-Tubulin. Consistent with results of nNOS gene knockout, we observed that L-VNIO did not affect glial differentiation (Supplementary Figure S1), but remarkably reduced the percentage of neurons (β-III-Tubulin⁺) at day 4 after differentiation (Control 20.83 ± 0.31% vs. L-VNIO 6.23 ± 0.40%, p < 0.001; Figures 2A,B). The reduction of neuronal differentiation was not due to decreased proliferation of progenitors, since L-VNIO did not inhibit cell proliferation (Figures 2C,D). However, PI staining revealed that L-VNIO significantly increased the apoptotic ratio of cells (PI⁺) (Control 3.48 ± 0.07% vs. L-VNIO 4.91 ± 0.15%, p < 0.01; Figures 2E,F). Thus, cell apoptosis may contribute to the effect of L-VNIO on decreased neurogenesis.

Does the repression of nNOS also inhibit progenitors to adopt a neuronal fate? To address this question, we used a mathematical description of neurogenesis, which is developed by Song et al. (2002). During the process of quantitative description, we defined δ_j as the death rate of all cells for the jth day in culture, and β_j as the conversion rate of progenitors (Tuj1^-) to neurons (Tuj1⁺) during that day. The death rate, δ_j, was measured by using PI dye in live cultures and was given by the equation δ_j N_j = PI_j+1 – PI_j. PI_j is the number of dead cells present at the start of the j day in culture and N_j is the total number of cells present at that time. Consistent with results of Figures 2E,F, we found that the death rate (δ₄) is significantly higher in L-VNIO-treated cultures (Control 0.024 ± 0.0007 vs. L-VNIO 0.054 ± 0.0033, p < 0.001; Figure 2G). The rate of conversion of progenitors to neurons, β_j, was given by the equation n _j+1 = n_j + β_j (N_j-n_j) – δ_jⁿn_j. n_j is the total number of neurons present at the start of the j day and δ_jⁿ is the death rate of neurons for the jth day. However, we could not determine what fractions of the dead cells were neurons. If no neurons die, we can say L-VNIO decreases the rate of neuronal fate commitment of NSCs. If all cells that die are neurons, that is, δ_jⁿn_j = δ_j N_j, we can obtain β_j. As shown in Figure 2H, L-VNIO leads to notably lower value of β_j (j = 4) (Control 0.426 ± 0.05 vs. L-VNIO 0.136 ± 0.06, p < 0.05). In addition, nestin labeling experiments showed that there are still neural progenitor cells at day 4 after differentiation (Figure 2I). Taken together, these results demonstrate that repression of nNOS negatively regulates neuronal fate commitment of embryonic NSCs.

To examine the effect of nNOS upregulation on neuronal differentiation, we incubated the cultures with 50 mM KCl during the first 12 or 24 h of NSCs differentiation. Sasaki et al. (2000) demonstrated that 50 mM KCl induces nNOS expression in cortical neurons. Our results showed that 50 mM KCl for 24 h clearly increased nNOS expression at day 1 after differentiation (Figure 3A), when almost all cells are nestin⁺ neural progenitors (Figure 2I). Treatment with 50 mM KCl for the first 24 h significantly promoted neuronal differentiation at day 4 after differentiation. The ratio of β-III-Tubulin⁺ neurons in KCl-treated cultures was 24.95 ± 2.07%, compared with 14.07 ± 1.03% in control (Figures 3B,C, p < 0.01). Moreover, 50 mM KCl markedly induced cell apoptosis (Figures 3D,E). The apoptotic effect was not mediated by nNOS since nNOS gene knockout did not abolish the effect (Supplementary Figure S2). If all dead cells are astrocytes, the data of increased neuronal ratio in KCl-treated cultures might be superficial. We then counted the number of β-III-Tubulin⁺ cells, which was proved to be much higher in KCl-treated cultures (Figure 3F). To confirm that nNOS mediates the effect of KCl and L-VNIO on neuronal differentiation, we treated nNOS^-/- cells with 50 mM KCl for the first 24 h or 100 μM L-VNIO for the later 2 days of 4-day differentiation. As shown in Figures 3G,H, nNOS gene knockout abolished the effect of KCl and L-VNIO on neuronal differentiation. These findings collectively indicate that upregulation of endogenous nNOS promotes neuronal differentiation.

To explore the molecular mechanisms underlying the regulation of NSC fate by nNOS, we investigated the role of HDAC2. To regulate HDAC2 specifically, we generated a lentiviral vector that contains shRNA of HDAC2 and named it LV-HDAC2 shRNA. LV-HDAC2 shRNA effectively infected NSCs and significantly decreased HDAC2 expression (Figure 4A). LV-HDAC2 shRNA-infected NSCs gave rise to significantly increased numbers of neurons (Figures 4B,C) but exhibited a decrease in their own proliferation (Figures 4D–F), suggesting that HDAC2 down-regulation instructs NSCs to exit from the cell cycle and adopt a neuronal fate. To further confirm the effect of HDAC2 in NSCs on neurogenesis, we infected NSCs with an adenovirus vector selectively expressing HDAC2 (AD-HDAC2) for 12 h, then AD-HDAC2 increased HDAC2 expression at day 4 after proliferation (Figure 5A, top) or differentiation (Figure 5A, bottom). Unexpectedly, AD-HDAC2 had no effect on NSC proliferation (Figure 5B), but significantly decreased neuronal differentiation (Figures 5C,D). Both LV-HDAC2 shRNA and AD-HDAC2 did not affect cell apoptosis (data not shown). Thus, HDAC2 negatively regulates neuronal fate commitment of embryonic NSCs.

Next, we asked whether nNOS could regulate HDAC2, thereby affecting the fate of NSCs. We planted neurospheres in polyornithine-coated cell culture dishes in differentiation medium and immediately treated cells with 100 μM L-VNIO or vehicle for 24 h. We found that L-VNIO significantly increased HDAC2 expression and enzymatic activity at day 1 after differentiation (Figures 6A,B). To determine whether HDAC2 upregulation mediates the role of L-VNIO in inhibition of fate commitment by NSCs to a neuronal lineage, we treated cultures with 100 μM L-VNIO or vehicle for the later 2 days during the 4-day differentiation of LV-HDAC2 shRNA- or LV-Control shRNA-infected NSCs. HDAC2 down-regulation rescued L-VNIO-induced neuronal differentiation reduction (Figures 6C–E). Together, nNOS repression negatively regulates neuronal fate commitment by upregulation of HDAC2 in progenitor cells.

To determine whether HDAC2 is a mediator for the regulation of NSC fate by nNOS, we cultured adult NSCs isolated from the hippocampus of 2-month-old male mice. These adult NSCs express nestin--- a marker of neural stem/progenitor cells, actively proliferate and differentiate into neurons (β-III-Tubulin⁺) and astrocytes (GFAP⁺) (Figure 7A). L-VNIO (100 μM) was added to the cultures at the start of day 3 during differentiation of monolayer-cultured adult NSCs. After 48 h, L-VNIO significantly inhibited neuronal differentiation (Figure 7B). Furthermore, 100 μM L-VNIO added for the first 24 h during differentiation dramatically increased HDAC2 expression and enzymatic activity at day 1 after differentiation (Figures 7C,D). LV-HDAC2 shRNA-infected adult NSCs differentiated into more neurons, and the negative effect of L-VNIO on neuronal differentiation was abolished in LV-HDAC2 shRNA-infected adult NSCs (Figures 7E,F), suggesting that HDAC2 upregulation mediates the effect of nNOS repression on inhibiting differentiation of adult NSCs into neurons.