PGK1 inhibitor CBR-470-1 protects neuronal cells from MPP+

The neurotoxin MPP+ (1-methyl-4-phenylpyridinium ion) disrupts mitochondrial function leading to oxidative stress and neuronal death. Here we examine whether activation of the Keap1-Nrf2 cascade can protect SH-SY5Y neuroblastoma cells from MPP+-induced cytotoxicity. Treatment of SH-SY5Y cells with CBR-470-1, an inhibitor of the glycolytic enzyme phosphoglycerate kinase 1 (PGK1), leads to methylglyoxal modification of Keap1, Keap1-Nrf2 disassociation, and increased expression of Nrf2 responsive genes. Pretreatment with CBR-470-1 potently attenuated MPP+-induced oxidative injury and SH-SY5Y cell apoptosis. CBR-470-1 neuroprotection is dependent upon Nrf2, as Nrf2 shRNA or CRISPR/Cas9-mediated Nrf2 knockout, abolished CBR-470-1-induced SH-SY5Y cytoprotection against MPP+. Consistent with these findings, PGK1 depletion or knockout mimicked CBR-470-1-induced actions and rendered SH-SY5Y cells resistant to MPP+-induced cytotoxicity. Furthermore, activation of the Nrf2 cascade by CRISPR/Cas9-induced Keap1 knockout protected SH-SY5Y cells from MPP+. In Keap1 or PGK1 knockout SH-SY5Y cells,CBR-470-1 failed to offer further cytoprotection against MPP+. Collectively PGK1 inhibition by CBR-470-1 protects SH-SY5Y cells from MPP+ via activation of the Keap1-Nrf2 cascade.

The transcription factor Nrf2 exerts a vital defensive activity against oxidative stress. Under resting conditions, Nrf2 targeted for degradation through ubiquitination by AGING binding Keap1, an adaptor protein for the Cul3 ubiquitin ligase complex [5,6]. Activated Nrf2 protein disassociates from Keap1 and translocates to cell nuclei. After binding to antioxidant response element (ARE) loci, Nrf2 mediates transcription of several key antioxidant enzymes and cytoprotective genes [7][8][9]. Studies have shown that forced activation of the Nrf2 cascade, through genetic or pharmacological strategies, can protect neuronal cells from oxidative stress [10][11][12]. Recently, novel Nrf2 activators been identified that can scavenge free radicals and efficiently protect neuronal cells from oxidative injury [13,14].

PGK1 shRNA or KO activates Nrf2 signaling and inhibits MPP + -induced cytotoxicity in SH-SY5Y cells
If the neuroprotective effects of CBR-470-1 are elicited through inhibition of PGK1,PGK1 silencing or depletion would be predicted to similarly induce Nrf2 cascade activation in SH-SY5Y neuronal cells. To test this, differentiated SH-SY5Y cells were transduced with the lentiviral PGK1 shRNA ("sh-PGK1", from Dr. Tan [18]), and cells selected by puromycin to establish stable cells ("sh-PGK1" cells). Additionally, a CRISPR/Cas9 470-1 (10 μM) or the vehicle control ("Veh"), cells were further cultured for the applied time periods, Keap1 immunoprecipitation with Nrf2 was tested by a Co-IP assay (A); Nrf2 and Keap1 protein expression in total cell lysates was tested as "Inputs" (A). SH-SY5Y neuronal cells were pre-treated with MG-132 (10 μM) for 24h, followed by CBR-470-1 (10 μM) stimulation and cultured for another 4-8h, Nrf2 and Tubulin protein expression in total cell lysates was tested (B). SH-SY5Y cells were pretreated with cycloheximide (CHX, 100 μg/mL) for 12h, followed by CBR-470-1 (10 μM) stimulation and cultured for another 4h, Nrf2 and Tubulin protein expression in total cell lysates was shown (C). SH-SY5Y neuronal cells were treated with CBR-470-1 (10 μM) or the vehicle control ("Veh"), cells were further cultured for the applied time periods, expression of indicated Keap1-Nrf2 pathway genes was tested by qPCR and Western blotting analyses (D, F, G); Alternatively, cells were harvested and relative ARE luciferase activity was tested, with results normalized to that of vehicle control (E). Expression of listed proteins was quantified and normalized to the loading control (A-D, G). Data were expressed as mean ± standard deviation (SD, n=5). * P< 0.05 vs. "Veh" cells (E, F). Experiments were repeated four times with similar results obtained.
AGING SH-SY5Y cells were protected from MPP + , presenting with significantly inhibited viability reduction ( Figure  4H), cell death ( Figure 4I) and apoptosis ( Figure 4J), when compared to parental control cells. Therefore, PGK1 silencing or KO inhibited MPP + -induced oxidative stress and cytotoxicity in SH-SY5Y cells.
Studies have shown that PGK1 inhibition or silencing leads to the accumulation of methylglyoxal and subsequent activation of the Keap1-Nrf2 cascade [17,18]. In the present study we show that CBR-470-1 activates the Nrf2 cascade in SH-SY5Y neuroblastoma cells, causing disassociation of the Keap1-Nrf2 complex, cytosol Nrf2 protein stabilization and nuclear translocation, followed by increased expression of Nrf2 pathway genes (HMOX1, NQO1 and SOD1).
Our results further support that PGK1 inhibition is the main mechanism of CBR-470-1-induced Nrf2 cascade activation and SH-SY5Y cell cytoprotection against MPP + . Mimicking CBR-470-1's activity, PGK1 depletion induced significant Nrf2 cascade activation. Functional studies demonstrated that SH-SY5Y cells with PGK1 silencing or KO were protected from MPP +induced cytotoxicity and apoptosis. In PGK1-depleted SH-SY5Y cells, CBR-470-1 failed to induce further Nrf2 cascade activation or offer additional cytoprotection AGING against MPP + . These results suggest that PGK1 inhibition by CBR-470-1 activated Nrf2 signaling cascade, thereby protectingSH-SY5Y neuronal cells from MPP + . To further support our hypothesis, we show that forced activation of the Nrf2 cascade by Keap1 KO similarly inhibited MPP + -inducedcytotoxicity in SH-SY5Y cells. Furthermore, CBR-470-1 was once again completely ineffective on MPP + -induced actions in Keap1 KO cells with sustained Nrf2 activation.

CONCLUSIONS
We discovered that CBR-470-1 activated the Nrf2 cascade to protect SH-SY5Y cells from MPP + -induced oxidative injury, suggesting that CBR-470-1 could be promising drug for neuron protection.

Chemicals, reagents and antibodies
CBR-470-1 was synthesized by Shanghai Rui-lu Biotech (Shanghai, China) based on a previously-described protocol [17]. MPP + was purchased from Sigma-Aldrich (St. Louis, Mo). All the antibodies utilized in the present study were purchased from Cell Signaling Tech (Danvers, MA). Fetal bovine serum (FBS), antibiotics and other cell culture reagents were from Gibco Co. (Suzhou, China). Primers and reagents for RNA studies were provided by Shanghai Genechem Co. (Shanghai, China).

SH-SY5Y cells
SH-SY5Y DA neuronal cells were provided by Dr. Zhang at Soochow University [2], cultured using a previously-described protocol [26]. For cell differentiation, SH-SY5Y cells were first cultured in complete DMEM (5% FBS) plus 10 µM all-trans retinoic acid (RA) for three days. Cells were then cultured in Neurobasal-A medium (NB) (minus phenol red, Invitrogen), supplemented with 1% L-glutamine (200 mM), 1% N-2 supplement (Invitrogen), and 1% P/S. Human BDNF (at 50 ng/mL) was added shortly. After another three days, SH-SY5Y cells were differentiated. The number of neurites in each differentiated SH-SY5Y cells is 4-5, and the average length of the neurites is close to 15 μm.

Lactate dehydrogenase (LDH) studies
The differentiated SH-SY5Y cells were cultured onto six well-tissue plates (at 1×10 5 cells per well). After the indicated MPP + treatment, medium LDH contents and total LDH contents were examined by a two-step LDH detection kit (Promega, Shanghai, China). The medium LDH was normalized (% vs. total LDH).

Cell viability
The differentiated SH-SY5Y cells were cultured onto six well-tissue plates (at 1×10 5 cells per well). Following the applied MPP + treatment, cell viability was quantified via a cell counting kit-8 (CCK-8) assay (Dojindo Molecular Technologies, Kumamoto, Japan), and its optical density (OD) values tested at 550 nm.

Western blotting and co-immunoprecipitation (co-IP)
The detailed protocols of Western blotting were previously reported [30,31]. In brief, lysate proteins were separated by SDS-PAGE gels [32], transferred to PVDF blots(Millipore, Shanghai, China). The blots were blocked and incubated with the designated primary and secondary antibodies. An ECL reagent kit (Pierce, Shanghai, China) was applied to detect the protein band under X-ray films. Data quantification was carried out by an ImageJ software (NIH). For the co-IP studies, the quantified protein lysates (1, 000 μg for each treatment) were pre-cleared and incubated with anti-Keap1 antibody [33]. Keap1-Nrf2 complex was captured by the G-Sepharose ("Beads", Sigma), tested by Western blotting. Testing nuclear fraction lysates was described in our previous studies [30,31] Caspase-3 activity SH-SY5Y cells were cultured onto six well-tissue plates (at 1×10 5 cells per well). Following the applied MPP + treatment, the caspase-3 activity was tested by a commercial fluorometric caspase-3 assay kit (Beyotime Biotechnology, Wuxi, China) [34], using the previouslydescribed protocol [31].

Cell apoptosis analyses
The detailed protocols for TUNEL [terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling] nuclear staining assay and Annexin V fluorescence-activated cell sorting (FACS) were described in detail previously [35,36]. For TUNEL assays, TUNEL percentage (% vs. DAPI staining) of 500 cells per treatment in five random views (1: 200 magnification) was calculated.

Lipid peroxidation quantification
As described elsewhere [37], the cellular lipid peroxidation levels were examined by using a thiobarbituric acid reactive substances (TBAR) activity assay kit [38].

Single strand DNA (ssDNA) ELISA
The differentiated SH-SY5Y cells were cultured onto six well-tissue plates (at 1×10 5 cells per well). After the indicated MPP + treatment, forty μg (40 μg) total cell lysates of each treatment were analyzed by a ssDNA ELISA kit (Roche, Basel, Switzerland) to quantify DNA fragmentations. The ssDNA ELISA absorbance was tested at 405 nm.

ARE reporter assay
The differentiated SH-SY5Y cells were cultured onto six well-tissue plates (at 1×10 5 cells per well), and transfected with an ARE-inducible firefly luciferase vector (from Dr. Jiang at Nanjing Medical University [39]). The transfected cells were subjected to applied treatment, and the luciferase activity tested by a luminescence instrument.

Mitochondrial depolarization
The differentiated SH-SY5Y cells were cultured onto six well-tissue plates (at 0.8×10 5 cells per well). Following the applied MPP + treatment for 24h, the fluorescence dye JC-1 (Thermo-Fisher Invitrogen, Shanghai, China) was added to detect mitochondrial depolarization ("∆Ψ"). JC-1 will aggregate in the mitochondria, forming green monomers with mitochondrial depolarization in cells with oxidative injury. JC-1 fluorescence absorbance was recorded at 550 nm.

The quantitative real-time reverse transcriptase polymerase chain reaction (qPCR)
The detailed protocols of qPCR were described previously [30,31]. Total cellular RNA was extracted by TRIzol reagents (Sigma), with reverse transcription and qPCR performed using a TOYOBO ReverTra Ace qPCR kit (Tokyo, Japan) and under the ABI Prism 7500H qPCR Instrument (Applied Biosystems, Foster City, CA). The melt curve analysis was performed, and a 2 −ΔΔCt method applied for quantification of targeted mRNAs using GAPDH as the reference gene. The primers for Nrf2 pathway gens, including heme oxygenase-1 (HMOX1),NADPH: quinone acceptor oxidoreductase 1 (NQO1), and superoxide dismutase (SOD), as well as Nrf2, Keap1 and PGK1, were provided by Dr. Tan at Sino-Japanese Friendship Hospital [18].
shRNA PGK1 shRNA-expressing lentivirus and Nrf2 shRNAexpressing lentivirus were provided by Dr. Tan at Sino-Japanese Friendship Hospital [18], that was individually transduced to SH-SY5Y cells (cultured in polybrenecontaining complete medium). Following selection by puromycin the stable cells were established, with PGK1 or Nrf2 silencing (over 95% knockdown efficiency) confirmed by qPCR and Western blotting analyses. Control SH-SY5Y cells were transduced with lentiviral scramble control shRNA ("sh-C").

CRISPR/Cas9-induced KO of Nrf2 or Keap1
The monoclonal SH-SY5Y cells with a lentiCRISPR-GFP-Nrf2-puro KO construct (Nrf2-KO SH-SY5Y cells), as well as the control cells with empty vector ("Vec"), were provided by Dr. Di at Wannan Medical College [40]. A lentiCRISPR-Keap1-KO-puro-GFP construct, provided by Dr. Zhen at Soochow University [24], was transduced to SH-SY5Y cells. Using FACS, GFP-positive cells were sorted and distributed to 24-well plates. Keap1-KO was screened and stable monoclonal cells established, with Keap1 KO verified by qPCR and Western blotting analyses.

Statistics
Data were presented as mean ±standard deviation (SD). Statistical differences were analyzed by one-way analysis of variance (ANOVA) followed by multiple comparisons performed with post hoc Bonferroni test (SPSS). A two-tailed unpaired T test (Excel 2007) was utilized to examine significance between two treatment groups. Values of P< 0.05 were considered statistically significant. Values of P> 0.05 stand for no significant difference.

AUTHOR CONTRIBUTIONS
All listed authors carried out the experiments, participated in the design of the study and performed the statistical analysis, participated in its design and coordination and helped to draft the manuscript.

CONFLICTS OF INTEREST
The listed authors have no conflicts of interest.