Induction of ferroptosis and mitochondrial dysfunction by oxidative stress in PC12 cells

Neurodegenerative diseases (NDD) are typically associated with neuron loss in nervous system areas. Interventions with related death mechanisms may ameliorate NDD progression. Oxidative stress plays an important role in NDD cell death routines. However, tert-butylhydroperoxide (t-BHP), a widely used oxidative stress stimulus, induces neural cell death through a mechanism that remains elusive. In our study, the ferroptosis marker events occurred after co-treatment with 100 μM t-BHP for 1 h, all of which were reversed in the presence of the ferroptosis inhibitor ferrostatin-1 (Fer-1) and the iron chelator deferoxamine, implying the occurrence of ferroptosis. Moreover, mitochondrial dysfunction accompanied by a decreased in membrane potential and ATP production, increased mitochondrial ROS generation. Furthermore, this mitochondrial dysfunction could be reversed by Fer-1. In addition, JNK1/2 and ERK1/2 were activated upstream of the ferroptosis and mitochondrial dysfunction. In summary, these data suggest that ferroptosis, coupled with mitochondrial dysfunction, was involved in t-BHP-induced PC12 death. JNK1/2 and ERK1/2 played important roles in t-BHP-induced cell death. Overall, this study might provide clues to the oxidative stress-based strategies for cell protection in NDD.

Cell death and proliferation control the homeostatic balance of metazoans. Precise control of the key nodes of death plays a pivotal role in the control of this homeostatic balance, which produces disease pathologies such as neurodegenerative diseases (NDD), once disturbed 1 . However, cell death induced by different environmental stresses or intracellular disorders exhibits diverse morphological and biochemical characteristics, which are defined as distinguished types of cell death. For several years, the first-discovered form of regulated cell death, called apoptosis, was considered the only therapeutically tractable mechanism 2 . However, this long-standing paradigm in the field has been challenged and revised in recognition of other regulated necrotic forms of cell death, such as programmed necrosis, ferroptosis, parthanatos and cyclophilin D-dependent necrosis 3 . In addition, several non-apoptotic forms of cell death lack a clear necrotic phenotype, such as the neutrophil extracellular-trap-associated cell death 4 , pyroptosis and autophagic cell death 5,6 . The identification of a specific type of cell death is crucial for prevention when homeostatic balance is disturbed.
Oxidative stress is a well-accepted paradigm that mediates neuronal dysfunction and death in age-related neurodegenerative diseases (NDD) such as Alzheimer's disease and Parkinson's disease. According to investigations of the underlying mechanisms involved in different paradigms of neurodegeneration, deregulated intracellular oxidative stress was identified as a common triggers of death signaling in neurons 7 .
Ferroptosis was identified more recently to be involved in oxidative stress-induced cell death 8 and it is introduced as an iron-dependent form of oxidative cell death in cancer cells and in neurons. The paradigm of ferroptosis involves the generation of soluble and lipid ROS through iron-dependent enzymatic reactions. Further, mechanisms of ferroptosis have been identified in neurons, in death signaling pathways induced by cerebral ischemia 9 , and in glutamate-induced neurotoxicity in organotypic hippocampal slice cultures 10 . The biochemical mechanisms linking oxidative stress to mitochondrial dysfunction in ferroptosis have been elucidated and identified that Bid plays an important role in this process 7 .

Results
Selection of the oxidative stress condition in PC12 cells. PC12 cells were treated with t-BHP (50 μM, 100 μM, 200 μM and 400 μM) for 0.5 h, 1 h, 2 h and 4 h. Then, the Annexin V/PI staining results revealed that PC12 cells that were co-treated with t-BHP (100 μM) for 1 h yielded ~15-16% cells death, and lower concentrations of t-BHP (50 μM to 100 μM) caused little apoptosis. However, when t-BHP was used at high concentrations (200 μM and 400 μM) and co-treated for longer than 1 h, the cell death percentage increased to as high as 85% ( Fig. 1A and B). The electron microscopy images revealed condensed nuclei and cytoplasm, accompanied by cell rounding, in PC12 cells that were treated with t-BHP at a concentration of 100 μM for 1 h (Fig. 1C). Fluorescence images of PC12 cells stained with PI and Calcein-AM, indicated that PC12 cells treated with t-BHP at a concentration of 100 μM for 1 h, showed a yellow merged color, which implied that the staining of these two dyes came to a contrast balance ( Fig. 2A and B). Considering that live cells remained suitable for rescue after treatment of t-BHP (100 μM) for 1 h, the same experimental conditions were selected for further study. Here, as shown in Fig. 1D, PC12 cell co-treatment with t-BHP at 100 μM for 1 h, caused a remarkable increase in ROS generation, and this could be reversed by the ROS scavenger N-acetyl-L -cysteine (NAC, 5 mM).

Evidence of ferroptosis in PC12 cells after co-treatment with t-BHP. Ferroptosis is a recently
coined type of cell death 10 . Lipid peroxidation 10 , down-regulated glutathione peroxidase 4 (Gpx4) 22 , and glutathione (GSH) depletion 23 are three crucial events that occur during ferroptosis. t-BHP increased the generation of lipid ROS ( Fig. 3A and B) but decreased the expression of Gpx4 (Fig. 3D-F). All of these effects could be reversed by the ferroptosis inhibitor, ferrostatin-1 (Fer-1) and an iron chelator deferoxamine (DFO), but were counteracted by the ferric ion provider ammonium ferric citrate (FAC). In addition, t-BHP could decrease the ratio of GSH/GSSG (Fig. 3C).

Evidence of mitochondrial dysfunction in PC12 cells after co-treatment with t-BHP. Biochemical
pathways of apoptosis activation can be extracellular or intracellular, as well as caspase-dependent or mitochondrial-dependent 24 . To explore the role of caspases, the activity of caspase-3 and caspase-7 was detected. t-BHP had no influence on caspase-3 or caspase-7 activity (Fig. 4F). Other caspase-related proteins were also detected, but their activity was not altered by t-BHP ( Figure S3). Given that caspases are not affected by t-BHP, the mitochondrial dysfunction was further detected. As shown in Fig. 4A, the JC-1 staining indicated that the mitochondrial membrane potential was significantly decreased after t-BHP treatment, as evidenced by the increased ratio of green to red fluorescence. Moreover, the mitochondrial ROS were also increased after co-treatment with t-BHP ( Fig. 4A and B), and ATP production was decreased (Fig. 4E). Furthermore, the expression of the mitochondria-dependent pro-apoptotic protein Bid was increased, and the expression of anti-apoptotic protein Bcl-2 was decreased. In addition, cytochrome c was translocated from the mitochondria to the cytoplasm ( Fig. 4C and D). Collectively, these results suggested that t-BHP induced mitochondrial dysfunction in PC12 cells and induced the release of apoptotic inducing protein, which led to cell death.

Evidence against t-BHP-induced autophagy and necroptosis in PC12 cells. Autophagy is a
non-apoptotic form of programmed cell death 25 . There are reports suggesting that ferroptosis is an autophagic inducible cell death 26 . Increased fluorescence was not observed after monodansylcadaverine (MDC) staining (Fig. 5A). In addition, the protein expression of LC-3II, which is a universal biomarker for autophagy, remained unaffected by t-BHP (Fig. 5B). Therefore, autophagy was not involved in the cell death triggered by t-BHP. Necroptosis is another form of caspase-independent regulated cell death that has been implicated in the development of a range of inflammatory, autoimmune, and neurodegenerative diseases 27 . In response to the activation of TNF receptor family members, RIPK1 is recruited to the cytosolic side of the receptor and its kinase activity is activated. RIPK1 then interacts with and phosphorylates a related kinase, RIPK3, leading to its activation 28,29 . Once active, RIPK3 then phosphorylates MLKL 30 . MLKL forms oligomers and translocates to the plasma membrane, resulting in necrotic cell death 3 . The role of necroptosis was further explored in this study. The protein expression levels of RIPK1, RIPK3, p-MLKL (Ser358), and MLKL were all increased after treatment with the necroptosis inducer PCB-95, but were not affected by t-BHP (Fig. 5C). Therefore, necroptosis might not be involved in t-BHP-induced PC12 death.

Ferroptosis inhibitor protected PC12 cells from mitochondrial dysfunction.
In line with earlier observations, the ferroptosis inhibitor, Fer-1, prevented t-BHP -induced cell death, decreased the generation of lipid peroxidation, induced the depletion of GSH, and increased the expression of Gpx4. Then, we analyzed the effects of Fer-1 on mitochondrial integrity in the t-BHP-induced PC12 cell model.
As shown in Fig. 6, Fer-1 increased the fluorescence ratio of red to green (Fig. 6A), as well as decreased the ATP generation (Fig. 6B) and the generation of mitochondrial ROS ( Fig. 6A and C), All of these results revealed that Fer-1 preserves mitochondrial integrity.

JNK1/2 and ERK1/2 were involved in t-BHP-induced ferroptosis and mitochondrial dysfunction.
The mitogen-activated protein kinases (MAPK) signaling pathway is an important signal transduction system in the organism and participates in many physiological and pathological processes 31 . Accumulated literature has reported that MAPK signaling is involved in cell functions from survival to death. In our study, we investigated the involvement of MAPKs in t-BHP-induced cell death. t-BHP treatment significantly increased the protein expression of p-JNK (Thr183/Tyr185) and p-ERK (Thr202/Tyr204) but had no effect on p-p38 mitogen-activated protein kinase (MAPK; Thr180/Tyr182) ( Figure S5). Furthermore, SP600125 and U0126 decreased the t-BHP-induced generation of lipid ROS (Fig. 7E), increased the ratio of GSH/GSSG (Fig. 7D), and restored the decreased expression of Gpx4 (Fig. 7B). Interestingly, SP600125 and U0126 also increased the fluorescence ratio of red to green (Fig. 7A), as well as decreased the generation of mitochondrial ROS ( Fig. 7A and F) and ATP generation (Fig. 7C).

Discussion
Cell death is an important biological process that shapes the development of multicellular organisms. Proper regulation of cell death is pivotal for the process of neurodegenerative diseases (NDD). Oxidative damage is a well-demonstrated pathogenic factor in NDD brains. Since the brain is rich in lipids containing polyunsaturated fatty acids, lipid peroxidation is the prominent type of oxidative damage. Lipid peroxidation is believed to be an important event in NDD pathogenesis 32 . Dysfunction of iron, which can augment the production of reactive oxygen species (ROS), is also evident in NDD brains. Interestingly, studies of lipid peroxidation and iron dysregulation revealed a new cell death mechanism, called ferroptosis, a cell death pathway that is genetically, morphologically, and biochemically distinct from other major forms of regulated cell death.
The phenomenon of ferroptosis is characterized by the overwhelming, iron-dependent accumulation of lethal lipid ROS, leading to lipid peroxidation 8,10,33 . Lipid peroxidation is the driving force of cell death in ferroptosis. Cystine/glutamate antiporter (also called system Xc − ) transports Glu, Cys and cysteine (Cys 2 ) into cells and plays important roles in GSH synthesis 34 . GSH depletion can directly activate lipoxygenases and suppress the activity of Gpx4 to trigger lipid peroxidation 35,36 . Gpx4 has been identified as an important regulator of ferroptosis in certain cancer cells 37,38 . Our results showed that the iron chelator DFO could restore t-BHP-induced cell morphology and viability. By contrast, the iron donor FAC further aggravated the cell morphology and decreased t-BHP-induced cell injury, thereby suggesting that cell death induced by t-BHP was iron-dependent. Moreover, the ferroptosis inhibitor Fer-1 could also restore the t-BHP-induced cell morphology and viability, suggesting the involvement of ferroptosis. Furthermore, t-BHP (100 μM) co-treatment with PC12 cells for 1 h increased the generation of lipid ROS, restored the expression of Gpx4, and decreased the ratio of GSH/GSSG. All of these results were reversed by Fer-1 and DFO, but aggravated by FAC, implying the involvement of ferroptosis in t-BHP-induced PC12 death accompanied by alteration of the system Xc − . In studies that investigated the underlying mechanisms in the different paradigms of NDD, oxidative stress was also usually involved in pathways of programmed cell death (PCD) converging at the mitochondria, the key organelle of energy metabolism that secures neuronal functions and survival under physiological conditions 39 .
In paradigms of programmed cell death, mitochondrial dysfunction is associated with the clash of the membrane potential and the release of pro-apoptotic proteins such as cytochrome c from the intermembrane space into the cytosol, where they orchestrate the final steps of PCD 40 . Here, mitochondrial dysfunction was also investigated. t-BHP decreased the mitochondrial membrane potential and the production of ATP, as well as increased the generation of mitochondrial ROS. Moreover, the mitochondrial membrane was disrupted, and cytochrome c was released from the mitochondria to the cytoplasm.
There are reports that suggest that ferroptosis was an autophagic cell death mechanism 26 . Autophagic cell death is characterized by the accumulation of large intracellular vesicles and the engagement of the autophagic machinery; however, cell death could be retarded by autophagy inhibitors, such as CQ 41 . The up-regulated LC-3II and down-regulated core autophagic-machinery components marked the occurrence of autophagy 42 . The role of autophagy in t-BHP-induced PC12 death remains elusive. Our results revealed that CQ had no effect on the t-BHP-induced alterations of cell morphology and viability. Furthermore, the number of autophagic vesicles and LC-3 II expression were not increased in response to t-BHP. Therefore, autophagy was not involved in t-BHP-induced cell death under our experimental conditions.
Considering the occurrence of ferroptosis and mitochondrial dysfunction, we further investigated the relationship between these two events. Interestingly, in our study, mitochondrial dysfunction was reversed by Fer-1, the ferroptosis inhibitor. Fer-1 increased the mitochondrial membrane potential and the production of ATP, as well as decreased the generation of mitochondrial ROS.
MAPKs are activated by environmental stressors to regulate multiple cell process, from cell survival to cell death 43,44 . MAPK activation has been implicated in t-BHP-induced cell death 45 . Consistent with a previous study in HepG2 cells, JNK1/2 and ERK1/2 were activated by t-BHP. JNK1/2 and p38 MAPK, but not ERK1/2, play pivotal roles in erastin-induced ferroptosis in leukemia cells 33 . U0126, an ERK inhibitor, could reverse erastin-induced ferroptosis in HT1080 cells 10 . In the present study, JNK1/2 and ERK1/2 were activated by t-BHP (100 μM) within 1 h in PC12 cells. SP600125 (JNK1/2 inhibitor) and U0126 (ERK1/2 inhibitor) increased the cell viability induced by t-BHP. Moreover, SP600125 and U0126 inhibited t-BHP-induced lipid ROS generation, increased the ratio of GSH/GSSG, and increased the expression of Gpx4. Furthermore, SP600125 and U0126 increased the mitochondrial membrane potential, restored ATP production, and decreased the generation of mitochondrial ROS. t-BHP 46 is a membrane-permeable oxidant compound that could induce different routes of cell death 47 . Low concentrations of t-BHP (2.5 μM) induced the phosphorylation of ERK1/2, JNK1/2 and p38 MAPK and activated apoptotic cascades in PC12 cells 45   the apoptosis effector enzyme caspase 3 in retinal pigment epithelial cells but did induce RIPK3-independent necrosis 50 . t-BHP co-treatment at 300 μM induced an ~ 80% cell viability decrease in PC12 cells 51 . Cell viability decreased to 40% when co-treated with 300 μM for 3 h in PC12 cells 37 . Here, in our study, t-BHP induced ferroptosis, and this mechanism of cell death was coupled with mitochondrial dysfunction during co-treatment of PC12 cells with 100 μM t-BHP for 1 h.
This study provided comprehensive evidence supporting that t-BHP-induced PC12 cell death involves ferroptosis and mitochondrial dysfunction. Fer-1 reversed mitochondrial dysfunction, and JNK1/2 and ERK1/2 were important regulators of this cell model and reversed the cell death induced by t-BHP (Fig. 8). Considering the close link between oxidative stress and NDD, this study might provide new insight into the prevention of cell loss in NDD.    biphenyl (PCB-95) (RPC-130AS) was purchased from Ultra Scientific (RI, USA). Ammonium ferric citrate (FAC) (A100170) and Z-VAD-FMK (C1202) were purchased from Aladdin (Shanghai, China). N-Acetyl-L -cysteine (NAC) (S0077), the JC-1 staining kit (C2006) and the lactate dehydrogenase (LDH) assay kit (C0016) were purchased from Beyotime Biotechnology (Haimen, China).

Materials. tert-Butylhydroperoxide
Cell culture. PC12 cells obtained from the American Type Culture Collection (ATCC; USA), were cultured in complete DMEM/F-12K medium containing 5% fetal bovine serum and 10% horse serum. Cells were incubated at 37 °C in a humidified atmosphere of 5% CO 2 in air. Electron microscopy observation for cell morphology changes. PC12 cells were seeded in 10-cm dishes. The following day, cells were treated with or without t-BHP (100 μM) for 1 h. Cells were collected in PBS, gently spun down, fixed in 2% paraformaldehyde, 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer before washing in 0.1 M sodium cacodylate buffer followed by post-fixing in 1% osmium tetroxide in 0.1 M sodium cacodylate buffer. Following fixation, cells were dehydrated through a graded series of alcohols and embedded in Spurrs Resin. Ultrathin sections were cut with a diamond knife using a Leica Ultracut S ultra-microtome and stained with both methanolic uranyl acetate and lead citrate before viewing in a LEO 906 transmission electron microscope operated at 60 kV. Determination of Caspase-3 and Caspase-7 activity. PC12 cells (5 × 10 4 /well) were cultured in 96-well white-walled multiwall luminometer plates. After treatment with t-BHP (100 μM) for 1 h, the activity of caspase-3 and caspase-7 was detected with a luminometer kit following the manufacturer's instructions.

Cell viability assay.
JC-1 staining. PC12 cells (5 × 10 4 /well) were cultured in 96-well black-walled multiwall plates. After treatment with t-BHP (100 μM) for 1 h, the mitochondrial membrane potential was detected by JC-1 staining with a commercial kit according to the manufacturer's instructions. Images were analyzed by Image J.
Annexin V-FITC/PI staining. To distinguish between dead cells and apoptotic cells, 10 6 cells/mL of each sample were co-stained with 5 μg/mL propidium iodide (PI) and fluorescein isothiocyanate (FITC)-conjugated annexin V. Cell counts were acquired with a FACS Calibur flow cytometer, and cell fractions in each quadrant were analyzed using Cell Quest Pro software. Cells in the lower right quadrant represented early apoptosis, while cells in the upper right quadrant represented late apoptosis. Dead cells were found in the upper left quadrant. Fluorescence compensation on the flow cytometer was adjusted to minimize overlap of the FITC and PI signals. A total of 10000 events in the gate were acquired for each sample from three different experiments.
Western Blot analysis. PC12 cells were collected, and the total protein was extracted after treatment with t-BHP or inhibitors. The protein concentrations were detected by the BCA protein assay according to the manufacturer's instructions. Thirty micrograms of cellular protein from each group was electro-blotted onto PVDF membrane following separation on 8% SDS-PAGE. The immuno-blot was incubated with the blocking solution SCIentIfIC REPORTS | (2018) 8:574 | DOI:10.1038/s41598-017-18935-1 (5% skim milk) at room temperature for 1 h, followed by incubation with a primary antibody overnight at 4 °C. After washing with Tween 20/Tris-buffered saline (TBST), the immune-blot was incubated with the respective secondary antibodies (1:10,000) for 2 h at room temperature. Blots were developed by the SuperSignal West Femto Chemi-luminescence Substrate and quantified by Image J.
Mitochondrial and cytoplasmic proteins were isolated with a commercial kit following the manufacturer's instructions.
Statistical analysis. The data were expressed as means ± SD and analyzed with one-way ANOVA by SPSS (17.0) for statistical significance using one-way analysis of variance 53 . P < 0.05 was considered statistically significant.