Human SH-SY5Y neuroblastoma cells were obtained from the Chinese Academy of Sciences Committee Type Culture Collection cell bank and were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM/F12, Hyclone, Logan, UT, USA) supplemented with 10% fetal bovine serum (Gibco, Grand Island, NY, USA), 100 U/ml penicillin (Hyclone, Logan, UT, USA) and 100 U/ml streptomycin (Hyclone, Logan, UT, USA; complete media, CM). The cell line was cultured in 100 mm tissue culture plates at 37°C in a humidified incubator (Model No. 3130, Forma Scientific, OH, USA) containing 5% CO₂. When the cell density reached 80%–90%, the cells were harvested and dispersed. We replaced the culture medium every 2 days. The cells in CM were treated with 1 mM MPP⁺ (D048, Sigma-Aldrich, St. Louis, MO, USA) for 24 h (The antibodies and abbreviations lists see Supplement Materials 1, 2).

Human SH-SY5Y neuroblastoma cells were infected through incubation in high titer Adenovirus-Green Fluorescent Protein (Ad-GFP) diluted in a small volume of DMEM/F12 at 37°C for 2 h with gentle swaying every 20 min. The infected cells were maintained for 24 h in fresh CM and were treated with 1 mM MPP⁺ for 24 h. Briefly, 5.0 × 10³ cells/well in 100 ul of culture medium were seeded into a 96-well plate and incubated at 37°C in 5% CO₂ for 24 h to allow cells to grow to 50%–60% confluency. The culture medium was replaced by 100 ul of serum-free medium. The different amount of viruses, 1.25 × 10⁵ pfu/well, 2.5 × 10⁵ pfu/well and 5 × 10⁵ pfu/well, according to multiplicity of infection (MOI; 25, 50, 100) values were applied for infection. The plate was shaken one time every 20 min to increase infection efficiency. After 2 h incubation, the medium was replaced by 100 ul of 5% FBS DMEM/F12 medium. The expression of GFP was observed by fluorescence microscopy (Leica, Germany) 24 h after infection. The transfer efficiency of adenovirus to the SH-SY5Y cells was relatively high, and a MOI of 50 was found to be the most suitable. This MOI was predicted to infect 90%–100% of the SH-SY5Y cells. Then SH-SY5Y cells were also infected with adenovirus in 6-well plates. The number of cells per well was 1.0 × 10⁵ cells/well in 2 ml of culture medium, and the corresponding amount of viruses was 5 × 10⁶ pfu/well (MOI = 50). Adenoviral vector delivery of small interfering RNA (siRNA) targeting PGC-1α (GeneBank accession number: NM_013261) and nonsense control (Ad) were purchased from SBO Medical Biotechnology Co., Ltd (Shanghai, China). The sequences of siRNAs were as follows:

3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide (MTT; Solarbio, Beijing, China) is absorbed into cells and then converted to formazan by mitochondrial succinate dehydrogenase. The accumulation of formazan directly reflects the activity of mitochondria and indirectly reflects cell viability. Cells were plated at a density of 1 × 10⁴ cells/well in 96-well plates and were cultured, differentiated and treated according to the above described methods. Twenty microliter of 0.5 mg/ml MTT was added to each well. After 4 h of incubation at 37°C, the initial 220 μl of solution was removed from each well, and then 100 μl of dimethyl sulfoxide (DMSO) was added to each well. Optical density (OD) was evaluated at 570 nm on an ELISA plate reader (Bio-Rad, Hercules, CA, USA) after the precipitate in the well was dissolved on a microplate mixer for 10 min. All results were normalized against the OD measured in a well under the same conditions without cell culture.

Human SH-SY5Y neuroblastoma cells were permeabilized and fixed with 0.5% Triton X-100 and 4% paraformaldehyde. Slides were blocked with 1% normal donkey serum (Merck, Darmstadt, Germany) in phosphate buffered saline (PBS) for 60 min at room temperature. Cells were washed with 0.1% bovine serum albumin (BSA, Beyotime Institute of Biotechnology, Shanghai, China) in PBS three times with gentle shaking, then incubated with the primary antibody Tyrosine Hydroxylase (TH; 1:100, Santa Cruz, CA, USA) diluted in 0.1% BSA/PBS at 4°C overnight. Labeled donkey anti-rabbit IgG (1:1000 Invitrogen, Paisley, UK) was used as the secondary antibody and was incubated in the dark for 2 h at room temperature. The samples were subjected to chromogenic diaminobenzidine (DAB) staining. In general, one drop of A, B, C reagents was respectively added into 1 ml of distilled water. The mixture was used for cell staining. The color development was monitored under the microscope at room temperature. The reaction stopped by adding distilled water when the ideal color was developed. Hematoxylin was used as a counterstain. The samples mounted by mounting medium were visualized by an inverted microscope under 200 times and 400 times magnification.

Human SH-SY5Y neuroblastoma cells, plated at a density of 1 × 10⁵ cells per 6-well dish, were treated according to the aforementioned methods. The cells were washed with ice-cold PBS three times, then the PBS was removed, and the cells were harvested in RIPA Lysis Buffer [50 mM Tris pH 7.4, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mM sodium orthovanadate, 50 mM sodium fluoride, 1 mM Ethylenediaminetetraacetic acid (EDTA)], and 0.5 mM phenylmethanesulfonylfluoride (PMSF; Beyotime Institute of Biotechnology, Shanghai, China). The lysates were incubated for 10 min on ice and centrifuged at 12,000× g for 10 min at 4°C. The supernatant containing the cell lysates was collected. The protein concentration was measured using BCA Protein Assay Kit (Beyotime Institute of Biotechnology, Shanghai, China). Thirty micrograms of proteins from total cell lysates were denatured by boiling in 1× SDS sample buffer (P0015, Beyotime Institute of Biotechnology, Shanghai, China). Thirty micrograms of denatured protein by boiling were loaded per lane and resolved by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) for 90 min at 80 V. The separated proteins were transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore, Carrigtwohill, Ireland) for 2 h at 200 mA with Bradford reagent (Bio-Rad, Hercules, CA, USA). The membranes were blocked with 5% skim milk in 1× PBS containing 0.05% Tween 20 (PBST) for 4 h at room temperature. The following primary antibodies were incubated with: anti-PGC-1α (1:1500 EMD Millipore Billerica, MA, USA); anti-ERRα (1:2500 EMD Millipore Billerica, MA, USA); anti-NRF1 (1:800 Abcam Cambridge, MA, USA); anti-NRF2 (1:1000 Abcam Cambridge, MA, USA), anti-PPARγ (1:1000 Abcam Cambridge, MA, USA), anti-Actin (1:2000 Beyotime Company of Biotechnology Shanghai, China), anti-GAPDH (1:1000 Beyotime Company of Biotechnology Shanghai, China) in PBST at 4°C overnight, the membranes were washed three times in PBST for 10 min. Subsequently, the membranes were incubated for 1.5 h in PBST containing secondary antibody conjugated to horseradish peroxidase (HRP; anti-mouse IgG 1:2000 and anti-rabbit IgG 1:2000, Beyotime Institute of Biotechnology, Shanghai, China). The immunoreactive bands were visualized and quantified using the enhanced chemiluminescence (ECL) detection kit (Millipore, USA). Protein levels were normalized to the housekeeping protein β-actin or GAPDH to adjust for variability of protein loading and expressed as a percentage of the vehicle control (deemed to be 100%).

Total RNA from human SH-SY5Y neuroblastoma cells was isolated according to the manufacturer’s protocol using TRizol reagent (Invitrogen, Carlsbad, CA, USA). Total RNA purity and integrity was confirmed using an ND-1000 NanoDrop (NanoDrop Technologies, Wilmington, NC, USA) and 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA). RNA (1 μg) was reverse-transcribed into cDNA in a total volume of 20 μl using the RevertAid^TM First Strand cDNA Synthesis Kit (k1621, Fermentas, St. Leon-Rot, Germany). The cDNA (2 μl) was amplified using a sequence detection system (ABI Prism 7500) in a total volume of 20 μl containing 10 μl of Fast Start Universal SYBR Green Master Mix (ROX; Roche, Penzberg, Germany) and each primer at 0.3 μM. Quantitative real-time PCR was performed using the ABI prism 7500 HT sequence detection system (Applied Biosystems, Forster City, CA, USA) based on the 59-nuclease assay for the indicated genes and the housekeeping gene GAPDH. Relative quantification of mRNAs was calculated with the ^∆∆Ct method. Relative expression was calculated using the ^∆∆Ct method. The absolute value of the slope of ∆CT vs. a log input <0.1 was considered as a criterion for passing the validation experiment. PCR amplification was carried out on cDNA equivalent to 10 ng of starting mRNA with the following specific oligonucleotide primers:

The conditions for PCR were as follows: initial denaturation for 3 min followed by 40 cycles of two steps: 1st: 95°C for 1 min, 2nd: 72°C (annealing) for 1 min 30 s. Followed by a final incubation of 72°C for 10 min. The results are expressed as the average of triplicate samples from at least three independent experiments for both control and treated cells.

Loss of mitochondrial membrane potential was assessed using Rhodamine 123 (Sigma, USA). Human SH-SY5Y neuroblastoma cells were seeded in 6-well plates at a concentration of 10⁵ cells per well. After the cells were treated with viral infection and MPP⁺, the medium was removed, and the cells were washed three times with DMEM/F12. The cells were incubated with Rhodamine 123 (Rh123) at a final concentration of 10 μg/ml in FBS-free DMEM/F12 for 30 min at 37°C. The fluorescence signal was measured using a flow cytometer (FACScalibur, San Jose, CA, USA) with excitation and emission wavelengths set at 530 and 590 nm, respectively. For each analysis, 10,000 events were recorded. The value for each treatment group was converted to a percentage of the control value.

ATP levels were assessed using a luciferin/luciferase-based ATP assay kit (no. 11699709001, Roche). Briefly, SH-SY5Y cells were treated with 1 mM MPP⁺ for 24 h. Cells were harvested, centrifuged and diluted to a concentration of 1 × 10⁶ cells/ml. The cells were plated at 25 μl/well in black 96-well plates. An ATP standard was serially diluted in dilution buffer to the range of 10⁻⁶ to 10⁻¹² M ATP. The same volume of cell lysis reagent was added to the samples, which were then incubated for 5 min at 25°C. Appropriate volumes of luciferase reagents were added to the samples and readings were recorded between 1 s and 10 s at an interval of 1 s using a multifunctional microplate reader (SpectraMax M3, Sunnyvale, CA, USA).

Intracellular production of H₂O₂ was fluorometrically assayed according to the manufacturer’s instructions (Amplex® Red, Molecular Probes, A22188, Invitrogen, Eugene, OR, USA). The Amplex® Red reagent reacts with H₂O₂ in a 1:1 stoichiometry to produce the red-fluorescent oxidation product, resorufin. The resorufin was measured with excitation and emission wavelengths at 530 and 590 nm. SH-SY5Y cells in which the PGC-1α gene was silenced were treated with 1 mM MPP⁺ for 24 h. Cells were harvested, centrifuged, and diluted to a concentration of 1 × 10⁶ cells/ml. An H₂O₂ standard curve was prepared, and the cells were plated on black 96-well plates to be tested. Cultures were exposed to a working solution containing 50 μM Amplex Red reagent and 0.1 U/ml HRP for 30 min and were assayed using a multifunctional microplate reader (SpectraMax M3, Sunnyvale, CA, USA) equipped for excitation at 540 nm and fluorescence emission at 590 nm.

Mitochondria were extracted according to the Mitochondria Fractionation Kit (Active Motif, Cat.40015, Carlsbad, CA 92008, USA) instructions. A cyt c standard was serially diluted with dilution buffer to 5.0, 2.5, 1.25, 0.625, 0.31, 0.155, 0.078 and 0 ng/ml according to the instructions accompanying the cyt c Human ELISA Kit (Abcam, Cambridge, MA, USA). The sample and standard specimens were added to 96-well plates filled with equal volumes (100 μl/well), specimen diluent was added to the blank wells, and biotinylated antibody solution (50 μl/well) was incubated for 2 h in the dark at room temperature. Then, the enzyme conjugate working solution was added (100 μl/well) after washing the plates three times for 1 h in the dark at room temperature. Next, the reaction was terminated after the addition of chromogenic substrate for 25 min, and the results were read using a multifunctional microplate reader (SpectraMax M3, Sunnyvale, CA, USA) at OD 450.

All quantitative data were collected from at least three independent experiments. The final data are expressed as the mean ± SEM, and analyzed using SPSS 17.0 statistical software (SPSS, Inc., Chicago, IL, USA) by means of one-way analysis of variance (ANOVA), followed by Tukey’s multiple comparison post hoc test. Real Time PCR data (Ct) were translated into the 2^−∆∆Ct format for statistical analysis; differences between mean values were analyzed by one-way ANOVA, P < 0.05 and P < 0.01 were considered as significant.

To establish an MPP⁺-induced cell model of PD for investigating the potential pathogenic effects of PGC-1α on PD, we first examined whether SH-SY5Y cell could be eligible for a dopaminergic neuronal-like cell model of PD by expressing TH. As shown in Figure 1A, the brown particles staining in cytoplasm of SH-SY5Y cells but not in nuclei suggested that SH-SY5Y cells could be used as a dopaminergic neuronal-like PD cell model due to their expression of TH. Next, to evaluate the viability of SH-SY5Y cells after exposure to oxidative injury, cells were treated with several concentrations of MPP⁺ (0 μM–2 mM) for 24 h. Cell viability was measured using the MTT assay. MPP⁺ significantly decreased MTT levels, indicating that cell loss is concentration-dependent. In addition, 1 mM MPP⁺ was selected as an optimal concentration when cell viability was decreased by 35.24% (P < 0.01; Figure 1B).

Real-time PCR and Western blot were used to detect the expression of PGC-1α mRNA and PGC-1α protein, respectively, in each group. In order to clarify the infection level of adenovirus to SH-SY5Y cells, SH-SY5Y cells were infected by adenovirus expressing GFP with MOI 25, 50 and 100. The expression level of GFP was observed under fluorescence microscope and the cells were counted (Figure 2A). The results showed that SH-SY5Y cells could be infected by adenovirus efficiently with MOI 50 and 100. However, SH-SY5Y cells infected with MOI 100 showed severe toxicity when comparing to those infected with MOI 50. Therefore, for follow-up experiments we chose MOI 50. The PGC-1α mRNA and protein levels of groups transfected with siRNA PGC-1, siRNA PGC-2, siRNA PGC-3 and siRNA PGC-4 were lower than those of the control group (P < 0.01), among which the PGC-1 group was the most marked: gene and protein expression levels were 30.74% (Figure 2B) and 15.56% (Figures 2C,D) of the wild type, respectively. There was no difference between the control group and the nonsense control group. This indicated that PGC-1 was the most efficient siRNA specific to PGC-1α in SH-SY5Y cells.

The real-time PCR showed that (Figure 3) the PGC-1α mRNA level was increased in MPP⁺ treated cells compared with the control group. Compared with the control viral (Ad) group, the expression of PGC-1α mRNA was decreased by 49.67% (P < 0.05) in the PGC-1α silencing group. In PD model groups, which further silenced the PGC-1α gene, PGC-1α mRNA expression decreased by 41.34% (P < 0.05) compared with the Ad+MPP⁺ groups.

The immunoreactive bands specific to PGC-1α, NRF-1, PPARγ, ERRα, NRF-2, β-actin and GAPDH were present by Western blot (Figure 4A). Compared with the control group, PGC-1α protein expression in the MPP⁺ group was decreased by 23.41% (P < 0.05), while PGC-1α protein expression decreased by 57.31% (P < 0.01) when the PGC-1α gene was silenced. Compared with the nonsense control group, PGC-1α protein expression in the PGC-1α gene silencing groups decreased by 31.84% (P < 0.05). Further treated with MPP⁺ produced a more significant decrease in PGC-1α protein expression of 66.81% (P < 0.01), compared with Ad+MPP⁺ group (Figure 4B). As shown in Figure 4C, NRF-1 protein expression decreased by 38.25% (P < 0.05) following treatment with MPP⁺ when compared with the nonsense control group, while NRF-1 protein expression decreased by 65.81% (P < 0.05) in the PGC-1α gene silencing groups. Compared with the control viral (Ad) groups, NRF-1 protein expression decreased 32.35% (P < 0.05) when the PGC-1α gene was silenced. In the PD model groups, upon silencing the PGC-1α gene, NRF-1 protein expression decreased by 69.39% (P < 0.05), and compared with the Ad+MPP⁺ groups, NRF-1 protein expression decreased by 36.12% (P < 0.05) in the PGC-1α+MPP⁺ groups. At the same time (in Figure 4D), PPARγ protein expression in the MPP⁺ groups decreased by 21.69% (P < 0.05) compared with the control group, while PPARγ protein expression decreased more significantly, reaching 78.87% (P < 0.05) when the PGC-1α gene was silenced. In PD model groups, in which the PGC-1α gene was further silenced, PPARγ protein expression decreased by 75.32% (P < 0.05); compared with the control viral group, PPARγ protein expression in the PGC-1α gene silencing groups decreased by 50.87% (P < 0.05). As for ERRα protein (in Figure 4E), compared with the control group, ERRα protein expression decreased by 46.17% (P < 0.05) when treated with MPP⁺ only, while ERRα protein expression decreased by 40.74% (P < 0.01) in the PGC-1α gene silencing groups. Compared with the nonsense control group, ERRα protein expression decreased 15.10% (P < 0.05) when silencing the PGC-1α gene. In PD model groups, which further silenced the PGC-1α gene, ERRα protein expression decreased by 78.66% (P < 0.01); compared with the Ad+MPP⁺ groups, ERRα protein expression decreased by 36.12% (P < 0.05) in the siRNA PGC-1+MPP⁺ groups. As shown in Figure 4F, NRF-2 protein expression in the MPP⁺ groups decreased by 30.02% (P < 0.05) compared with the control group, while NRF-2 protein expression decreased by 73.31% (P < 0.01) when silencing the PGC-1α gene. Compared with the control viral (Ad) groups, NRF-2 protein expression decreased by 45.47% (P < 0.05) when silencing the PGC-1α gene. Compared with the only MPP⁺ group, which further silenced the PGC-1α gene, NRF-2 protein expression decreased by 55.20% (P < 0.05); compared with the control viral group, NRF-2 protein expression in the PGC-1α gene silencing groups decreased by 31.87% (P < 0.05).

PGC-1α expression is associated with mitochondrial respiration in regulation of energy metabolism and oxidative stress. To analyze the effects of silencing PGC-1α genes on the mitochondrial function of SH-SY5Y cells, we first evaluated whether silencing PGC-1α would affect the mitochondrial membrane potential (∆ψM) using Rh123 as an indicator, as well as cell viability (Figures 5, 6A). Upon MPP⁺ exposure, SH-SY5Y cells showed a statistically significant decrease in ∆ψM and cell viability when compared to the control. However, pre-treatment with PGC-1α silencing further decreased 41.19% (P < 0.01) of ∆ψM and 47.15% (P < 0.05) of cell viability when compared to SH-SY5Y cells treated with MPP⁺ alone, indicating that expression of PGC-1α rendered dysfuction of mitochondrial respiration induced by MPP⁺ through inhibiting mitochondrial complex I. We then asked whether knockdown of PGC-1α would result in further depletion of ATP upon MPP⁺ treatment. To this end, we measured ATP levels. As shown in Figure 6B, ATP generation in SH-SY5Y cells treated with MPP⁺ was decreased when compared to the control group. However, silencing the PGC-1α gene resulted in a further decrease (63.94%, P < 0.05) in ATP level when compared to SH-SY5Y cells treated with MPP⁺ alone. In agreement with the data from mitochondrial potential assay, decreases in ATP levels were correlated with the change in ∆ψM. In fact, decrease of ATP and change of mitochondrial ∆ψM can lead to an increase of reactive oxygen species (ROS). Given that mitochondria generate and accumulate the majority of ROS and H₂O₂ and that mitochondrial dysfunction causes an increase in radical production, we next assessed the effect of MPP⁺ on H₂O₂ production. As shown in Figure 6C, H₂O₂ was significantly increased (43.38%, P < 0.01) in SH-SY5Y cells exposed to MPP⁺. However, pre-treatment to silence the PGC-1α gene resulted in a more marked increase H₂O₂ level (93.44%, P < 0.01) compared to SH-SY5Y cells treated with MPP⁺ alone. In general, overloading of ROS causes cells undergoing apoptosis. We, therefore, detected changes in mitochondrial cyt c using ELISA (Figures 6D,E). Cyt c is an important component of the mitochondrial electron transport chain and is located in the mitochondria, but it can be released into the cytoplasm upon apoptosis. We found that the mitochondrial cyt c level was significantly decreased after SH-SY5Y cells were treated with MPP⁺, and it was reduced further (11.35%, P < 0.01) following PGC-1α gene silencing (Figure 6D). Correspondingly, the cyt c in the cytoplasm tended to increase (Figure 6E). In comparison with the control group, cyt c in the mitochondria decreased by 16.02% (P < 0.05) after cells were treated with MPP⁺ only. In PD model groups, compared with the control viral groups, the mitochondrial cyt c decreased by 16.22% (P < 0.05) in PGC-1α gene silencing groups, and the difference was statistically significant. Meanwhile, cyt c in the cytoplasm increased by 32.66% (P < 0.05). Thus, we can conclude that silencing the PGC-1α gene caused mitochondrial damage in the PD cell model and led to mitochondrial cyt c release into the cytoplasm.