PPARδ Activation Mitigates 6-OHDA-Induced Neuronal Damage by Regulating Intracellular Iron Levels

Intracellular iron accumulation in dopaminergic neurons contributes to neuronal cell death in progressive neurodegenerative disorders such as Parkinson’s disease. However, the mechanisms of iron homeostasis in this context remain incompletely understood. In the present study, we assessed the role of the nuclear receptor peroxisome proliferator-activated receptor δ (PPARδ) in cellular iron homeostasis. We identified that PPARδ inhibited 6-hydroxydopamine (6-OHDA)-triggered neurotoxicity in SH-SY5Y neuroblastoma cells. PPARδ activation with GW501516, a specific PPARδ agonist, mitigated 6-OHDA-induced neuronal damage. Further, PPARδ activation also suppressed iron accumulation, which contributes to 6-OHDA-induced neuronal damage. PPARδ activation attenuated 6-OHDA-induced neuronal damage in a similar manner to that of the iron chelator deferoxamine. We further elucidated that PPARδ modulated cellular iron homeostasis by regulating expression of divalent metal transporter 1, ferroportin 1, and ferritin, but not transferrin receptor 1, through iron regulatory protein 1 in 6-OHDA-treated cells. Interestingly, PPARδ activation suppressed 6-OHDA-triggered generation of reactive oxygen species and lipid peroxides. The effects of GW501516 were abrogated by shRNA knockdown of PPARδ, indicating that the effects of GW501516 were PPARδ-dependent. Taken together, these findings suggest that PPARδ attenuates 6-OHDA-induced neurotoxicity by preventing intracellular iron accumulation, thereby suppressing iron overload-associated generation of reactive oxygen species and lipid peroxides, key mediators of ferroptotic cell death.


Introduction
Parkinson's disease (PD), a progressive neurodegenerative disorder, is defined by loss of dopaminergic neurons in the substantia nigra, accompanied by prominent neuropathological symptoms such as tremor and bradykinesia [1]. Although multiple approaches to PD treatment have been evaluated in experimental and clinical trials, the only clinically available PD therapy is replacement of dopamine (DA) with the DA precursor levodopa [2,3]. Because continuous degeneration of dopaminergic neurons does not allow conversion of the precursor to dopamine in advanced PD [3], developing a therapeutic that can prevent degeneration of dopaminergic neurons would address a substantial unmet clinical need. 6-hydroxydopamine (6-OHDA) is a neurotoxic, hydroxylated form of DA that contributes to the degeneration of both noradrenergic and dopaminergic neurons in PD [4]. 6-OHDA is transported into neurons by binding to cell membrane transporters expressed in both 2.3. Gene Silencing SH-SY5Y cells stably expressing shRNA targeting non-specific control or PPARδ (target sequence: 5 -CCGCAAACCCTTCAGTGATAT-3 ) were generated by transducing lentiviral particles expressing each shRNA. For transduction, SH-SY5Y cells were seeded into 6-well plates as 1.6 × 10 5 cells/well, cultured for 16 h, and subsequently transduced lentiviral particles with 8 µg/mL of hexadimethrine bromide (Sigma, St. Louis, MO, USA, Cat# H9268) in growth medium. The usage of lentiviral particles per well was the same number of seeded cell number (1 MOI). Transduced cells were selected by incubating the cells in culture medium containing 2 µg/mL puromycin for 7 days. PPARD silencing was verified by immunoblot analysis.

Cytotoxicity Assay
To assess cytotoxicity, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma-Aldrich, Cat# M5655) and lactose dehydrogenase (LDH) release assays were conducted. For the MTT assay, SH-SY5Y cells were plated at 4 × 10 4 cells/well in 24-well plates. Following incubation of the cells with the indicated reagents for 24 h, MTT solution was added to the culture medium and cells were incubated for an additional 2 h. After medium removal, formazan crystal formed in living cells was dissolved in acidified isopropanol. Absorbance was then measured at 570 nm using a Multiskan™ GO microplate spectrophotometer (Thermo Scientific, Waltham, MA, USA). To measure secreted LDH, culture media from cells treated with the specified reagents was collected. The level of released LDH was determined using a CytoTox 96 non-radioactive cytotoxicity assay kit (Promega, Madison, WI, USA, Cat# G1780). Absorbance was measured at 490 nm using a Multiskan™ GO microplate spectrophotometer (Thermo Scientific).

Measurement of Intracellular Iron Levels
Intracellular iron level was determined by two methods. First, a calcein fluorescent probe, which was inversely proportional to fluorescence intensity, was used to determine intercellular iron level as described previously [33]. Briefly, SH-SY5Y cells were plated at 1.6 × 10 5 cells/well in 6-well plates (Corning Inc., Corning, NY, USA) and subsequently treated with the specified reagents. After incubation for the indicated duration, 500 nM calcein-AM (BD Biosciences, San Diego, CA, USA, Cat# 564061) was added to the culture medium. After treatment for 30 min at 37 • C, cells were then washed using phosphatebuffered saline (PBS) and green fluorescence was detected using an Eclipse Ti2 fluorescence microscope (Nikon, Minato, Tokyo, Japan). Second, a ferene-based colorimetric method was performed as described previously [34]. Labile iron concentrations were determined in cell lysates treated as above. Briefly, 100 µL of ammonium acetate buffer (2.5 M, pH 4.5) and 120 µL of labile iron working solution (5 mM ferene and 10 mM ascorbic acid prepared in 2.5 M ammonium acetate buffer, pH 4.5) were added to cell lysates. This mixture was vortexed and left overnight at room temperature. Then, the absorbance was measured at 595 nm in a Multiskan™ GO microplate spectrophotometer (Thermo Scientific). Iron concentrations were determined with a curve calibrated on iron standards and normalized to the amount of protein.

Western Blot Analysis
SH-SY5Y cells treated with the specified reagents for indicated durations were washed with ice-cold PBS and lysed in PRO-PREP™ Protein Extraction Solution (iNtRON Biotechnology, Seongnam, Korea, Cat# 17081). Cell lysate aliquots were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (7.5% polyacrylamide) and transferred to Immobilon-P polyvinylidene difluoride membranes (Merck, Darmstadt, Germany, Cat# IPVH00010). Membranes were blocked for 2 h at ambient temperature with 5% non-fat milk suspended in Tris-buffered saline (TBS) containing 0.1% Tween-20 and subsequently incubated overnight at 4 • C with the indicated primary antibodies diluted in TBS containing 0.1% Tween-20. Membranes were subsequently incubated with a peroxidase-conjugated secondary antibody for 1 h at room temperature. After thorough washing of the membranes with TBS containing 0.1% Tween-20, chemiluminescence was visualized using WesternBright ECL (Advansta Inc., Menlo Park, CA, USA, Cat# K-12045-D50).

Measurement of Intracellular ROS
Intracellular ROS was measured by two methods. First, a H2DCF-DA fluorescent probe was used to determine ROS level as described previously [35]. Briefly, SH-SY5Y cells were plated at 1.6 × 10 5 cells/well in 6-well plates (Corning Inc., Corning, NY, USA) and treated with the specified reagents. After incubation for the indicated duration, cells were treated with 50 µM 2 ,7 -dichlorofluorescin diacetate (DCF-DA, Merck, Cat# 287810). Following incubation for 30 min at 37 • C, cells were washed with PBS and green fluorescence was then detected using a Ti2 fluorescence microscope (Nikon). Fluorescence intensity was quantified using Image J software (NIH, Bethesda, MD, USA). Second, an Amplex™ Red hydrogen peroxide assay kit was used to detect intracellular hydrogen peroxide according to the manufacturer's instructions (Invitrogen, Carlsbad, CA, USA, Cat# A22188).

Measurement of Lipid Peroxidation
Lipid peroxidation was determined by two methods. First, a fluorescent probe BODIPY (581/591) (Invitrogen, Cat# D3861) was used to measure lipid peroxides as described previously [36]. Briefly, SH-SY5Y cells were plated at 1.6 × 10 5 cells/well in 6-well plates (Corning) and treated with indicated reagents. After incubation for the indicated durations, cells were treated with 1 µM C11-BODIPY (581.591) for 30 min at 37 • C. C11-BODIPY fluorescence corresponding to oxidized or non-oxidized lipids was detected using an Eclipse Ti2 fluorescence microscope (Nikon). Fluorescence intensity was quantified using Image J software (NIH). Second, an EZ-lipid peroxidation (TBARS) assay kit was used to detect intracellular lipid peroxides according to the manufacturer's instructions (DOGEN, Seoul, Korea, Cat# DG-TBA200).

Statistical Analysis
Data were analyzed using SigmaPlot 10 software (Systat, Chicago, IL, USA) and are expressed as means ± standard error (SE). Statistical significance was determined by oneway ANOVA with Tukey's post hoc test or unpaired t-test with Welch's correction using Prism 5 software (GraphPad Software, San Diego, CA, USA). p < 0.05 was considered statistically significant [37].

GW501516 Activation of PPARδ Attenuates 6-OHDA-Triggered Cellular Damage in SH-SY5Y Cells
Because 6-OHDA induces degeneration of both noradrenergic and dopaminergic neurons and is known to contribute to PD pathology [4], we used 6-OHDA treatment as an in vitro model of PD. Cells were treated with various concentrations of 6-OHDA for 24 h. The lowest neurotoxic concentration was 20 µM 6-OHDA (Supplementary Figure S1). Accordingly, we used 20 µM 6-OHDA for subsequent experiments.
Subsequently, we evaluated the effects of GW501516, a PPARδ-specific agonist against 6-OHDA neurotoxicity. When SH-SY5Y cells were treated with 6-OHDA for 24 h, LDH release was significantly increased. However, 6-OHDA-induced LDH release was dosedependently decreased by GW501516 treatment ( Figure 1A). Similarly, 6-OHDA-induced decrease in cell viability (MTT assay) was also reversed by GW501516 in a concentrationdependent manner ( Figure 1B). SY5Y cells transduced with shRNA targeting PPARδ, but not scrambled sequence-targeting shRNA (Supplementary Figure S2). The neuroprotective effects of GW501516 against 6-OHDA were ablated by PPARδ-targeting shRNA, as demonstrated by both LDH release and MTT assay ( Figure 1C,D). These results indicated that GW501516 attenuates 6-OHDA-induced neurotoxicity in a PPARδ-dependent manner. (A,B) Cells pretreated with increasing concentrations of GW501516 for 8 h were exposed to 6-OHDA. After incubation for 24 h, LDH release (A) and MTT assays (B) were performed to evaluate 6-OHDA-induced cellular damage. (C,D) Cells stably expressing shRNA targeting scrambled sequences or PPARδ were pretreated with vehicle (DMSO) or GW501516 for 8 h and subsequently exposed to 6-OHDA. Following incubation for 24 h, cells were subjected to LDH release (C) and MTT assays (D). Results are expressed as means ± SE (n = 3). ** p < 0.01 relative to untreated group; # p < 0.05, ## p < 0.01 relative to 6-OHDA-treated group.

GW501516 Activation of PPARδ Decreases 6-OHDA-Induced Iron Accumulation
Iron accumulation in dopaminergic neurons has been implicated in the pathogenesis of PD [15,16]. Moreover, 6-OHDA-induced reduction in viability is attenuated by ferroptosis inhibitors such as ferrostatin-1 and liproxstatin-1 (Supplementary Figure S3). Thus, we determined if the neuroprotective role of PPARδ in 6-OHDA-induced neurotoxicity was associated with the regulation of intracellular iron. When SH-SY5Y cells were exposed to 6-OHDA for 24 h, intracellular iron levels were markedly increased. However, (A,B) Cells pretreated with increasing concentrations of GW501516 for 8 h were exposed to 6-OHDA. After incubation for 24 h, LDH release (A) and MTT assays (B) were performed to evaluate 6-OHDAinduced cellular damage. (C,D) Cells stably expressing shRNA targeting scrambled sequences or PPARδ were pretreated with vehicle (DMSO) or GW501516 for 8 h and subsequently exposed to 6-OHDA. Following incubation for 24 h, cells were subjected to LDH release (C) and MTT assays (D). Results are expressed as means ± SE (n = 3). ** p < 0.01 relative to untreated group; # p < 0.05, ## p < 0.01 relative to 6-OHDA-treated group.
To determine if the effects of GW501516 were PPARδ-dependent, the protective effects of GW501516 were evaluated in SH-SY5Y cells stably expressing shRNA targeting PPARδ or scrambled sequences. PPARδ protein levels were dramatically decreased in SH-SY5Y cells transduced with shRNA targeting PPARδ, but not scrambled sequence-targeting shRNA (Supplementary Figure S2). The neuroprotective effects of GW501516 against 6-OHDA were ablated by PPARδ-targeting shRNA, as demonstrated by both LDH release and MTT assay ( Figure 1C,D). These results indicated that GW501516 attenuates 6-OHDA-induced neurotoxicity in a PPARδ-dependent manner.

GW501516 Activation of PPARδ Decreases 6-OHDA-Induced Iron Accumulation
Iron accumulation in dopaminergic neurons has been implicated in the pathogenesis of PD [15,16]. Moreover, 6-OHDA-induced reduction in viability is attenuated by ferroptosis inhibitors such as ferrostatin-1 and liproxstatin-1 (Supplementary Figure S3). Thus, we determined if the neuroprotective role of PPARδ in 6-OHDA-induced neurotoxicity was associated with the regulation of intracellular iron. When SH-SY5Y cells were exposed to 6-OHDA for 24 h, intracellular iron levels were markedly increased. However, 6-OHDAinduced iron accumulation was significantly decreased by GW501516 (Figure 2A,B).
Antioxidants 2022, 11, x FOR PEER REVIEW 7 of 18 6-OHDA-induced iron accumulation was significantly decreased by GW501516 ( Figure  2A,B). To further examine the actions of PPARδ on the intracellular iron levels, we evaluated the effect of deferoxamine (DFO), an iron chelator. 6-OHDA-mediated increase in intracellular iron was also significantly reduced by DFO as did GW501516 ( Figure 3A,B). Furthermore, DFO and GW501516 had similar protective effects against 6-OHDA-induced neuronal damage, as assessed by both LDH release ( Figure 3C) and MTT assays ( Figure 3D). Combined treatment of both GW501516 and DFO, however, did not have increased protective effects relative to GW501516 treatment alone in 6-OHDA-induced iron accumulation or neuronal damage. These results suggested that the neuroprotective effects of PPARδ against 6-OHDA are directly related to modulation of intracellular iron levels. To further examine the actions of PPARδ on the intracellular iron levels, we evaluated the effect of deferoxamine (DFO), an iron chelator. 6-OHDA-mediated increase in intracellular iron was also significantly reduced by DFO as did GW501516 ( Figure 3A,B). Furthermore, DFO and GW501516 had similar protective effects against 6-OHDA-induced neuronal damage, as assessed by both LDH release ( Figure 3C) and MTT assays ( Figure 3D). Combined treatment of both GW501516 and DFO, however, did not have increased protective effects relative to GW501516 treatment alone in 6-OHDA-induced iron accumulation or neuronal damage. These results suggested that the neuroprotective effects of PPARδ against 6-OHDA are directly related to modulation of intracellular iron levels.
To assess the regulatory mechanisms of PPARδ-mediated iron homeostasis, we examined the effects of GW501516 on the expression of key regulators of iron homeostasis, DMT1 and FPN1. Treatment of SH-SY5Y cells with 6-OHDA time-dependently increased mRNA levels of DMT1, an iron importer, which was prevented by pre-treatment with GW501516 ( Figure 4A,B). Similarly, 6-OHDA increased DMT1 protein levels, which was reduced by GW501516 pre-treatment ( Figure 4C).
Subsequently, we evaluated the effects of 6-OHDA and GW501516 on the expression of FPN1, an iron exporter. Time-dependent exposure of SH-SY5Y cells to GW501516 did not affect FPN1 mRNA levels ( Figure 4D). Although 6-OHDA increased FPN1 mRNA in SH-SY5Y cells, GW501516 co-treatment did not reverse this effect ( Figure 4E). Contrastingly, 6-OHDA decreased protein levels of FPN1 in SH-SY5Y cells, which was prevented by GW501516 pre-treatment, implying that PPARδ modulates FPN1 protein levels, but not mRNA levels, potentially by regulating FPN1 translation ( Figure 4F).  To assess the regulatory mechanisms of PPARδ-mediated iron homeostasis, we examined the effects of GW501516 on the expression of key regulators of iron homeostasis, DMT1 and FPN1. Treatment of SH-SY5Y cells with 6-OHDA time-dependently increased mRNA levels of DMT1, an iron importer, which was prevented by pre-treatment with GW501516 ( Figure 4A,B). Similarly, 6-OHDA increased DMT1 protein levels, which was reduced by GW501516 pre-treatment ( Figure 4C).
Subsequently, we evaluated the effects of 6-OHDA and GW501516 on the expression of FPN1, an iron exporter. Time-dependent exposure of SH-SY5Y cells to GW501516 did not affect FPN1 mRNA levels ( Figure 4D). Although 6-OHDA increased FPN1 mRNA in SH-SY5Y cells, GW501516 co-treatment did not reverse this effect ( Figure 4E). Contrastingly, 6-OHDA decreased protein levels of FPN1 in SH-SY5Y cells, which was prevented by GW501516 pre-treatment, implying that PPARδ modulates FPN1 protein levels, but not mRNA levels, potentially by regulating FPN1 translation ( Figure 4F).  Western blot (C,F), respectively. RPS18 and α-tubulin were used as internal controls for real-time PCR and Western blot, respectively. Results are expressed as means ± SE (n = 3). An image analyzer was used to quantify band intensity of Western blot, and the ratio of protein to α-tubulin is indicated above each lane. * p < 0.05, ** p < 0.01 relative to the untreated group; # p < 0.05, ## p < 0.01 relative to the 6-OHDA-treated group.
To directly assess the role of PPARδ in GW501516-mediated expression of key ironregulatory proteins including DMT1, FPN1, ferritin, and TfR1, the effect of GW501516 was examined in SH-SY5Y cells stably expressing shRNA targeting PPARδ or scrambled sequences. GW501516 suppression of DMT1 mRNA and protein levels in 6-OHDA-challenged cells was significantly attenuated in SH-SY5Y cells stably expressing PPARδ-targeting shRNA, but not control shRNA ( Figure 5A,F). Contrastingly, GW501516 did not alter mRNA levels of FPN1 and ferritin in 6-OHDA-treated cells stably expressing shRNA targeting either PPARδ or scrambled sequences ( Figure 5B,D,E). However, GW501516 reversed the 6-OHDA-triggered decrease in FPN1 and ferritin protein in SH-SY5Y cells expressing scrambled shRNA, but not in cells stably expressing PPARδ-targeting shRNA, indicating that PPARδ regulates translation of FPN1 and ferritin rather than transcription ( Figure 5F). The expression of TfR1 mRNA and protein was not significantly affected by either 6-OHDA or GW501516 ( Figure 5C,F). (E,F) Cells pretreated with DMSO or GW501516 for 8 h were incubated with or without 6-ODHA for 16 h. Total RNA and protein were extracted, and mRNA and protein levels were analyzed by realtime PCR (A,B,D,E) and Western blot (C,F), respectively. RPS18 and α-tubulin were used as internal controls for real-time PCR and Western blot, respectively. Results are expressed as means ± SE (n = 3). An image analyzer was used to quantify band intensity of Western blot, and the ratio of protein to α-tubulin is indicated above each lane. * p < 0.05, ** p < 0.01 relative to the untreated group; # p < 0.05, ## p < 0.01 relative to the 6-OHDA-treated group.
To directly assess the role of PPARδ in GW501516-mediated expression of key ironregulatory proteins including DMT1, FPN1, ferritin, and TfR1, the effect of GW501516 was examined in SH-SY5Y cells stably expressing shRNA targeting PPARδ or scrambled sequences. GW501516 suppression of DMT1 mRNA and protein levels in 6-OHDAchallenged cells was significantly attenuated in SH-SY5Y cells stably expressing PPARδtargeting shRNA, but not control shRNA ( Figure 5A,F). Contrastingly, GW501516 did not alter mRNA levels of FPN1 and ferritin in 6-OHDA-treated cells stably expressing shRNA targeting either PPARδ or scrambled sequences ( Figure 5B,D,E). However, GW501516 reversed the 6-OHDA-triggered decrease in FPN1 and ferritin protein in SH-SY5Y cells expressing scrambled shRNA, but not in cells stably expressing PPARδ-targeting shRNA, indicating that PPARδ regulates translation of FPN1 and ferritin rather than transcription ( Figure 5F). The expression of TfR1 mRNA and protein was not significantly affected by either 6-OHDA or GW501516 ( Figure 5C,F).

Figure 5.
PPARδ knockdown abrogated the effects of GW501516 on 6-OHDA-triggered protein levels of DMT1, FPN1, and Ferritin, and mRNA expression of DMT1. Cells stably expressing shRNA targeting scrambled sequences or PPARδ were pretreated with DMSO or GW501516 for 8 h, and subsequently incubated with or without 6-OHDA for 16 h. Total RNA and protein were extracted, and levels of mRNA and protein were analyzed by real-time PCR (A-E) and Western blot (F). RPS18 and α-tubulin were used as internal controls for real-time PCR and Western blot, respectively. Results are expressed as means of triplicate ± SE (A-D). * p < 0.05, ** p < 0.01 relative to untreated group; # p < 0.05 relative to 6-OHDA-treated group.

GW501516 Activation of PPARδ Suppresses IRP1 Expression in SH-SY5Y Cells
To further elucidate the molecular mechanism of PPARδ-mediated modulation of key iron-regulatory proteins in SH-SY5Y cells, we determined the effect of GW501516 on the expression of IRP1, which is known to regulate the expression of these key proteins by binding IREs in the 3′ or 5′ UTRs of these genes [38]. When SH-SY5Y cells were exposed to 6-OHDA, IRP1 mRNA was increased, reaching a maximum of two-fold upregulation after 16 h and declining to basal levels after 24 h ( Figure 6A). Contrastingly, GW501516 monotreatment downregulated IRP1 mRNA in a time-dependent manner ( Figure 6B). Consistent with these results, shRNA-targeted knockdown of PPARδ suppressed the effects of GW501516 on 6-OHDA-induced IRP1 upregulation ( Figure 6C). Consistently, the pre-treatment with GW501516 attenuated 6-OHDA-induced upregulation of IRP1 mRNA ( Figure 6D). This PPARδ-mediated downregulation of IRP1 was correlated with expression of target proteins, in which GW501516 markedly reversed the effects of 6-OHDA on the protein expression of DMT1, FPN1, and ferritin, but not TfR1 ( Figure 6E). In contrast, GW501516 did not alter markedly the mRNA and protein expression of IRP2 (Supplementary Figure S4), which has high sequence homology and similar biochemical activities to that of IRP1 [39]. This result may explain the effect of GW501516 on the expression of TfR1, which is more affected by IRP2 than IRP1 [40]. Additionally, GPX4 or FSP1, important effectors in anti-ferroptotic signaling [41,42], were not regulated in presence of GW501516 (Supplementary Figure S5). and α-tubulin were used as internal controls for real-time PCR and Western blot, respectively. Results are expressed as means of triplicate ± SE (A-D). * p < 0.05, ** p < 0.01 relative to untreated group; # p < 0.05 relative to 6-OHDA-treated group.

GW501516 Activation of PPARδ Suppresses IRP1 Expression in SH-SY5Y Cells
To further elucidate the molecular mechanism of PPARδ-mediated modulation of key iron-regulatory proteins in SH-SY5Y cells, we determined the effect of GW501516 on the expression of IRP1, which is known to regulate the expression of these key proteins by binding IREs in the 3 or 5 UTRs of these genes [38]. When SH-SY5Y cells were exposed to 6-OHDA, IRP1 mRNA was increased, reaching a maximum of two-fold upregulation after 16 h and declining to basal levels after 24 h ( Figure 6A). Contrastingly, GW501516 monotreatment downregulated IRP1 mRNA in a time-dependent manner ( Figure 6B). Consistent with these results, shRNA-targeted knockdown of PPARδ suppressed the effects of GW501516 on 6-OHDA-induced IRP1 upregulation ( Figure 6C). Consistently, the pre-treatment with GW501516 attenuated 6-OHDA-induced upregulation of IRP1 mRNA ( Figure 6D). This PPARδ-mediated downregulation of IRP1 was correlated with expression of target proteins, in which GW501516 markedly reversed the effects of 6-OHDA on the protein expression of DMT1, FPN1, and ferritin, but not TfR1 ( Figure 6E). In contrast, GW501516 did not alter markedly the mRNA and protein expression of IRP2 (Supplementary Figure S4), which has high sequence homology and similar biochemical activities to that of IRP1 [39]. This result may explain the effect of GW501516 on the expression of TfR1, which is more affected by IRP2 than IRP1 [40]. Additionally, GPX4 or FSP1, important effectors in anti-ferroptotic signaling [41,42], were not regulated in presence of GW501516 (Supplementary Figure S5).  Figure 4C,F. RPS18 and α-tubulin were used as internal controls for real-time PCR and Western blot, respectively. Results are expressed as means of triplicate ± SE (A-D). * p < 0.05, ** p < 0.01 relative to untreated group; # p < 0.05 relative to 6-OHDA-treated group.

GW501516 Activation of PPARδ Decreases 6-OHDA-Triggered Accumulation of ROS and Lipid Peroxides
Because pathologically increased intracellular iron levels trigger generation of ROS and lipid peroxides [7], we determined if 6-OHDA affected intracellular ROS and lipid peroxides. Exposure of SH-SY5Y cells to 6-OHDA for 24 h dramatically increased the levels of intracellular ROS and hydrogen peroxide ( Figure 7A-C). Consistent with increased intracellular ROS levels, lipid peroxides were significantly increased in 6-OHDA-treated cells ( Figure 7D-F). However, 6-OHDA-triggered accumulation of intracellular ROS and lipid peroxides was almost completely abolished by GW501516 pre-treatment (Figure 7). Furthermore, the beneficial effects of GW501516 were ablated in SH-SY5Y cells stably expressing PPARδ-targeting shRNA, but not control shRNA (Figure 8). These results indicated that PPARδ elicits its neuroprotective effects against 6-OHDA-induced toxicity in part through ferroptotic signaling.  Figure 4C,F. RPS18 and α-tubulin were used as internal controls for real-time PCR and Western blot, respectively. Results are expressed as means of triplicate ± SE (A-D). * p < 0.05, ** p < 0.01 relative to untreated group; # p < 0.05 relative to 6-OHDA-treated group.

GW501516 Activation of PPARδ Decreases 6-OHDA-Triggered Accumulation of ROS and Lipid Peroxides
Because pathologically increased intracellular iron levels trigger generation of ROS and lipid peroxides [7], we determined if 6-OHDA affected intracellular ROS and lipid peroxides. Exposure of SH-SY5Y cells to 6-OHDA for 24 h dramatically increased the levels of intracellular ROS and hydrogen peroxide ( Figure 7A-C). Consistent with increased intracellular ROS levels, lipid peroxides were significantly increased in 6-OHDA-treated cells ( Figure 7D-F). However, 6-OHDA-triggered accumulation of intracellular ROS and lipid peroxides was almost completely abolished by GW501516 pre-treatment (Figure 7). Furthermore, the beneficial effects of GW501516 were ablated in SH-SY5Y cells stably expressing PPARδ-targeting shRNA, but not control shRNA (Figure 8). These results indicated that PPARδ elicits its neuroprotective effects against 6-OHDA-induced toxicity in part through ferroptotic signaling.  Results are expressed as means ± SE (n = 3). Scale bar, 100 μm. * p < 0.05, ** p < 0.01 relative to untreated group; # p < 0.05, ## p < 0.01 relative to 6-OHDA-treated group.

Discussion
Among the PPAR subtypes, PPARδ is most abundantly expressed in the brain, suggesting its importance in neuronal function [43]. Although the anti-inflammatory and metabolic roles of PPARδ have been extensively studied [22], the potential protective roles of PPARδ against neuronal cell death in the dopaminergic system, including against ferroptosis, have not been evaluated. The present study demonstrated that activation of PPARδ with a synthetic PPARδ-specific ligand, GW501516, attenuates 6-OHDA-induced neuronal damage, a major pathological event in PD, by suppressing intracellular iron accumulation and lipid peroxidation, the primary phenotypes of ferroptosis. The anti-ferroptotic effects of PPARδ were associated with the key iron regulatory protein IRP1, which functions as a translational regulator of the cellular iron transporters DMT1 and FPN1.
The anti-ferroptotic effects of PPARδ are related directly to blockade of neuronal damage induced by 6-OHDA. A previous study demonstrated that pathologically increased intracellular iron levels are a critical initiator of ferroptosis, and that multiple iron metabolism-associated genes such as HAMP, FTH1, HSPB1, TF, SLC40A1, TFRC, and STEAP3, are involved in this event [10,44]. In addition, iron chelation reversed ferroptotic Fluorescence signals corresponding to intracellular iron (B) and oxidized C11-BODIPY (B) were detected using fluorescence microscopy and fluorescence intensities were quantified. Results are expressed as means ± SE (n = 3). Scale bar, 100 µm. * p < 0.05, ** p < 0.01 relative to untreated group; # p < 0.05, ## p < 0.01 relative to 6-OHDA-treated group.

Discussion
Among the PPAR subtypes, PPARδ is most abundantly expressed in the brain, suggesting its importance in neuronal function [43]. Although the anti-inflammatory and metabolic roles of PPARδ have been extensively studied [22], the potential protective roles of PPARδ against neuronal cell death in the dopaminergic system, including against ferroptosis, have not been evaluated. The present study demonstrated that activation of PPARδ with a synthetic PPARδ-specific ligand, GW501516, attenuates 6-OHDA-induced neuronal damage, a major pathological event in PD, by suppressing intracellular iron accumulation and lipid peroxidation, the primary phenotypes of ferroptosis. The anti-ferroptotic effects of PPARδ were associated with the key iron regulatory protein IRP1, which functions as a translational regulator of the cellular iron transporters DMT1 and FPN1.
The anti-ferroptotic effects of PPARδ are related directly to blockade of neuronal damage induced by 6-OHDA. A previous study demonstrated that pathologically increased intracellular iron levels are a critical initiator of ferroptosis, and that multiple iron metabolism-associated genes such as HAMP, FTH1, HSPB1, TF, SLC40A1, TFRC, and STEAP3, are involved in this event [10,44]. In addition, iron chelation reversed ferroptotic cell death triggered by the ferroptosis inducers erastin, sulfasalazine, and RSL3, suggesting a critical role for iron accumulation in ferroptotic cell death [7]. Consistent with previous studies, activation of PPARδ by the specific ligand GW501516 attenuated 6-OHDA-induced intracellular iron accumulation. This effect of PPARδ was mediated in part by modulation of the key iron-regulatory proteins DMT1, FPN1, and ferritin. Our findings suggest that PPARδ affects DMT1, FPN1, and ferritin, but not TfR1, by downregulating IRP1, which modulates expression of these iron transporters.
Although calcium-independent phospholipase A2 beta and protein kinase C are implicated in regulation of IRP1 expression [45,46], 6-OHDA treatment did not affect either of these pathways in SH-SY5Y cells (data not shown). The molecular mechanisms underlying PPARδ-mediated downregulation of IRP1 are unclear. However, the present findings demonstrated that activation of PPARδ by GW501516 dramatically attenuated 6-OHDA-induced intracellular iron accumulation by regulating iron transporter expression. Accordingly, further studies will determine the precise molecular mechanism for PPARδmediated downregulation of IRP1.
Although the precise molecular mechanisms of PPARδ-mediated neuroprotection from 6-OHDA are lesser known than its roles in cellular inflammation and metabolism [22], the present observations are consistent with a previous report demonstrating that PPARδ improves motor impairment and dopaminergic neurodegeneration by inhibiting neuroinflammation in the MPTP mouse model of PD [28]. In addition, the PPARδ agonists GW501516 and L-165041 significantly suppressed MPTP-triggered depletion of striatal dopamine in the murine brain [27]. The other PPARδ agonist GW0742 also improved MPTPinduced cognitive impairment by suppressing oxidative damage and DNA fragmentation in a rat model of PD [47]. Although the function of the PPARδ in the pathogenesis of PD is not fully elucidated, the present findings clearly indicate that activation of PPARδ with GW501516 attenuates 6-OHDA-induced neuronal damage in SH-SY5Y neuroblastoma cells, indicating the potential therapeutic potential of PPARδ as a novel approach to treatment for neurodegenerative diseases, particularly PD.
Ferroptosis is a potential mechanism for degeneration of dopaminergic neurons in the context of PD. Excess iron accumulation is present in the substantia nigra of autosomal recessive juvenile parkinsonism patients, suggesting that regulation of iron levels and ironmediated oxidative stress could suppress parkinsonism [16]. Consistent with this notion, oxidated cholesterol metabolites are elevated in a dopaminergic cell line, suggesting that oxidative stress exacerbates neurodegeneration via generation of cholesterol aldehydes [48]. In fact, iron-dependent cell death, termed ferroptosis, includes induction of oxidative stress and is implicated in the pathogenesis of PD [7,49]. Linkage between ferroptosis and neuronal atrophy in dopaminergic and noradrenergic neurons has been mainstream opinion since it was discovered [50,51]. Moreover, ferritinophagy accompanying ferroptosis is a form of autophage involving iron-mediated ferroptosis, and blockade of ferritinophagy may be an important strategy for blocking ferroptosis-mediated cell death [52]. In this context, our result indicates that the PPARδ might associated with ferrototic cells death by regulating the expression of IRP1, which is known as a ferroptotic effector that acts by modulating the level of iron homeostasis-associated proteins such as DMT1, FPN1 and Ferritin.
Multiple antioxidants such as coenzyme Q10, vitamin E, and creatine were ineffective in clinical trials for PD [53][54][55]. However, the iron chelator deferiprone improved motor function while raising striatal dopamine by reducing labile iron levels and oxidative stressinduced damages in mice [56]. Similar improvements in substantia nigra iron deposits and motor function were also observed in a pilot clinical trial for PD patients [56].

Conclusions
Consistent with these previous studies, we demonstrated in the present report that activation of PPARδ by GW501516 attenuated 6-OHDA-triggered intracellular iron accumulation by regulating the expression of iron transport proteins. PPARδ modulation of intracellular iron levels was accompanied by decreased ROS and lipid peroxide levels in 6-OHDA-treated cells pretreated with GW501516. These observations provide insight into the potential protective and anti-ferroptotic roles of PPARδ in the context of PD.