Toxicity of Necrostatin-1 in Parkinson’s Disease Models

Parkinson’s disease (PD) is a neurodegenerative disorder that is characterized by the loss of dopaminergic neurons in the substantia nigra pars compacta. This neuronal loss, inherent to age, is related to exposure to environmental toxins and/or a genetic predisposition. PD-induced cell death has been studied thoroughly, but its characterization remains elusive. To date, several types of cell death, including apoptosis, autophagy-induced cell death, and necrosis, have been implicated in PD progression. In this study, we evaluated necroptosis, which is a programmed type of necrosis, in primary fibroblasts from PD patients with and without the G2019S leucine-rich repeat kinase 2 (LRRK2) mutation and in rotenone-treated cells (SH-SY5Y and fibroblasts). The results showed that programmed necrosis was not activated in the cells of PD patients, but it was activated in cells exposed to rotenone. Necrostatin-1 (Nec-1), an inhibitor of the necroptosis pathway, prevented rotenone-induced necroptosis in PD models. However, Nec-1 affected mitochondrial morphology and failed to protect mitochondria against rotenone toxicity. Therefore, despite the inhibition of rotenone-mediated necroptosis, PD models were susceptible to the effects of both Nec-1 and rotenone.


Introduction
Parkinson's disease (PD) is a neurodegenerative disorder characterized by progressive neuronal loss in various regions of the central nervous system. The most prominent aspect is the loss of dopaminergic neurons in the substantia nigra pars compacta [1]. This extensive neuronal death triggers motor and non-motor clinical features which uncover the PD pathogenesis that has probably been underway for a long time. The pathogenesis of PD is very complex and has been reported to be related to gene mutations and/or to long-term exposure to stressor agents throughout an individual's life.
Necroptosis is a programmed type of necrosis that develops in response to the activation of two members of the receptor-interacting protein (RIP) family, RIP1 and RIP3. Both are serine/threonine kinase proteins [11] that interact with each other via their RIP homotypic interaction motifs (RHIM). Indeed, RIP1 activation requires autophosphorylation, and thereafter, phosphorylates the RIP3 protein [12]. Beyond phosphorylation, RIP1 and RIP3 form a complex called a necrosome [13] which activates the mixed lineage kinase domain-like (MLKL) protein, a RIP3 substrate that is critical for plasma membrane rupture [14]. In fact, upon phosphorylation by RIP3, MLKL forms oligomers that translocate to the plasma membrane and promote either ion influx (calcium and sodium) or pore formation, which is responsible for a loss in membrane integrity [13]. Such membrane disruption triggers inflammation through the release of cell contents and, eventually, leads to cell death. The involvement of necroptosis in PD has been demonstrated in animals injected with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) [15], with 6-OHDA and in postmortem PD midbrain samples [16] through an increase in MLKL phosphorylation. In addition, necrostatin-1 (Nec-1), an inhibitor of the necroptotic pathway, exerts a neuroprotective effect on 6-OHDA-treated pheochromocytoma (PC12) cells [17], on MPTP-treated mice [15,18], and prevents neurite degeneration in 6-OHDA mesencephalic neurons [16]. Although non-apoptotic death induced by MPP + was shown to be inhibited by Nec-1 in differentiated SH-SY5Y cells, this was not classified as necroptosis [10].
In this study, we used primary fibroblasts from PD patients with or without the G2019S LRRK2 mutation to characterize the necroptosis pathway. Moreover, we assessed the oxidative stress-inducing effect of rotenone in PD models. Interestingly, protein executioners of necroptosis were expressed in PD patients, but were only activated under rotenone treatment. Nec-1 totally abolished rotenone-induced necroptosis but did not prevent rotenone toxicity.

Immunofluorescence Microscopy
Cells were plated on pretreated coverslips with poly-L-lysine. After rotenone treatment, cells were washed with annexin buffer 1X and incubated with annexin V-FITC (1X) for 15 min at room temperature (RT). Subsequently, 5 µL of PI was directly added to cells. Thereafter, cells were washed and incubated with annexin buffer 1X to wipe away excess dye. Annexin/PI staining was observed by in vivo immunofluorescence. Nuclei were stained with 300 nM of 4 , 6-diamidino-2-phenylindole (DAPI) (Thermo Fisher, D1306). For mitochondrial morphology, cells were fixed with 4% paraformaldehyde (PFA) and permeabilized with 0.1% Triton X-100 (Sigma-Aldrich, T9284). Once permeabilized, cells were incubated with bovine serum albumin (BSA)/PBS solution (1 mg/mL) for 1 h. Thereafter, cells were incubated for 1 h with TOMM20 (F10) (Santa-Cruz, sc-17764, dilution 1:200) and, subsequently, with Thermo Fisher Alexa Fluor ® 568 (A11004)-conjugated secondary antibodies for another hour at RT. Results were analyzed, and represented the proportion of cells with filamentous and damaged mitochondria for each condition. Images were visualized using an Olympus IX51 inverted microscope equipped with a DP71 camera.

Statistics
Data were collected and analyzed using Microsoft Excel and/or SPSS software. Statistical analyses were done using Student t and X 2 tests to establish the significant differences (p values * < 0.05, ** ≤ 0.01, *** ≤ 0.001) between control and PD cell lines under basal and treated conditions.

Characterization of Necroptosis in PD Models
Apoptotic [5] and necrotic [19] cell death have been described in cells from patients with genetic or sporadic PD, respectively. In this study, we characterized the necroptosis pathway in primary fibroblasts from PD patients harboring or not harboring the G2019S LRRK2 mutation (Figure 1). Based on the phosphorylation levels of RIP1 (p-RIP1) ( Figure 1A,B) and RIP3 (p-RIP3) ( Figure 1C,D), the necroptosis pathway did not seem to be activated in cells from PD patients. Although there were no differences in p-RIP1 levels among the groups, p-RIP3, the substrate of p-RIP1 was significantly decreased in IPD cells and not in GS cells. Therefore, we did not observe any significant protein phosphorylation levels that could show that the necroptosis pathway was activated in PD patient cells. However, proteins responsible for necroptosis activation were found to be well-expressed in human fibroblasts.

Characterization of Necroptosis in PD Models
Apoptotic [5] and necrotic [19] cell death have been described in cells from patients with genetic or sporadic PD, respectively. In this study, we characterized the necroptosis pathway in primary fibroblasts from PD patients harboring or not harboring the G2019S LRRK2 mutation (Figure 1). Based on the phosphorylation levels of RIP1 (p-RIP1) ( Figure 1A,B) and RIP3 (p-RIP3) ( Figure 1C,D), the necroptosis pathway did not seem to be activated in cells from PD patients. Although there were no differences in p-RIP1 levels among the groups, p-RIP3, the substrate of p-RIP1 was significantly decreased in IPD cells and not in GS cells. Therefore, we did not observe any significant protein phosphorylation levels that could show that the necroptosis pathway was activated in PD patient cells. However, proteins responsible for necroptosis activation were found to be well-expressed in human fibroblasts.

Study of Necroptosis in Rotenone-Induced Models
Cells were treated with the mitochondrial complex I inhibitor rotenone and/or Nec-1, an RIP1 inhibitor [10], for 24 h. In rotenone-treated SH-SY5Y cells, we observed an increase in the levels of p-RIP3 (Figure 2A,B) and its substrate, p-MLKL ( Figure 2C,D). Similar data were mentioned for p-RIP3 in a study involving primary culture of mesencephalic neurons with much lower doses of rotenone [20]. However, while Nec-1 treatment failed to decrease the level of p-RIP3, the level of p-MLKL was reduced. As a consequence, Nec-1 inhibited rotenone-induced necroptosis. We wondered whether

Study of Necroptosis in Rotenone-Induced Models
Cells were treated with the mitochondrial complex I inhibitor rotenone and/or Nec-1, an RIP1 inhibitor [10], for 24 h. In rotenone-treated SH-SY5Y cells, we observed an increase in the levels of p-RIP3 (Figure 2A,B) and its substrate, p-MLKL ( Figure 2C,D). Similar data were mentioned for p-RIP3 in a study involving primary culture of mesencephalic neurons with much lower doses of rotenone [20]. However, while Nec-1 treatment failed to decrease the level of p-RIP3, the level of p-MLKL was reduced. As a consequence, Nec-1 inhibited rotenone-induced necroptosis. We wondered whether rotenone could activate necroptosis in the cells of PD patients. Interestingly, we noticed an enhanced level of p-MLKL with rotenone treatment ( Figure 2E), which reduced with Nec-1. We inferred then that Nec-1 efficiently inhibited the activation of necroptosis by rotenone in both neuroblastoma cells and primary fibroblasts. Of note, the expression level of MLKL in the cells of PD patients was not affected by Nec-1. However, a previous study has related the upregulation of MLKL to necroptosis activation in human neurons with OPA1 mutations [18]. In our study, a different variation of MLKL was also observed during its characterization in primary fibroblasts ( Figure S1), but was not significant.
Antioxidants 2020, 9, x FOR PEER REVIEW 5 of 13 rotenone could activate necroptosis in the cells of PD patients. Interestingly, we noticed an enhanced level of p-MLKL with rotenone treatment ( Figure 2E), which reduced with Nec-1. We inferred then that Nec-1 efficiently inhibited the activation of necroptosis by rotenone in both neuroblastoma cells and primary fibroblasts. Of note, the expression level of MLKL in the cells of PD patients was not affected by Nec-1. However, a previous study has related the upregulation of MLKL to necroptosis activation in human neurons with OPA1 mutations [18]. In our study, a different variation of MLKL was also observed during its characterization in primary fibroblasts ( Figure S1), but was not significant.

Cellular Toxicity of Necrostatin-1
Rotenone induces mitochondrial dysfunction, resulting in the generation of mitochondrial ROS and the loss of MMP [21]. A previous study showed that oxidative stress leads to necroptosis activation [22]. Given that Nec-1 abrogated rotenone-induced necroptosis, we decided to study MMP by loading cells with TMRM. While healthy mitochondria retained the dye, the damaged one did not. The percentage of TMRM-negative (-) cells was remarkably augmented following Nec-1 treatment with or without rotenone incubation ( Figure 3A). Rotenone decreased MMP by 10% in the cells of PD patients. Unexpectedly, it did not affect MMP in the control line, but it exacerbated the number of annexin + /PI + cells (data not shown) observed with fluorescence microscopy. Although Nec-1 inhibited the induction of necroptosis in rotenone-treated SH-SY5Y cells, it was not able to decrease the percentage of annexin + cells ( Figure 3B) and/or PI + cells ( Figure 3C), nor did alter the level of mitochondrial ROS produced with rotenone ( Figure 3D). Taken together, rotenone induced not only necroptosis but also apoptosis [21] and/or necrosis [20] in cell cultures, as previously reported. Surprisingly, Nec-1 treatment alone resulted in an increase in annexin/PI + cells and a decrease in MMP.

Cellular Toxicity of Necrostatin-1
Rotenone induces mitochondrial dysfunction, resulting in the generation of mitochondrial ROS and the loss of MMP [21]. A previous study showed that oxidative stress leads to necroptosis activation [22]. Given that Nec-1 abrogated rotenone-induced necroptosis, we decided to study MMP by loading cells with TMRM. While healthy mitochondria retained the dye, the damaged one did not. The percentage of TMRM-negative (-) cells was remarkably augmented following Nec-1 treatment with or without rotenone incubation ( Figure 3A). Rotenone decreased MMP by 10% in the cells of PD patients. Unexpectedly, it did not affect MMP in the control line, but it exacerbated the number of annexin + /PI + cells (data not shown) observed with fluorescence microscopy. Although Nec-1 inhibited the induction of necroptosis in rotenone-treated SH-SY5Y cells, it was not able to decrease the percentage of annexin + cells ( Figure 3B) and/or PI + cells ( Figure 3C), nor did alter the level of mitochondrial ROS produced with rotenone ( Figure 3D). Taken together, rotenone induced not only necroptosis but also apoptosis [21] and/or necrosis [20] in cell cultures, as previously reported. Surprisingly, Nec-1 treatment alone resulted in an increase in annexin/PI + cells and a decrease in MMP.  The results represent the percentage ± SD of annexin + or PI + cells detected by flow cytometry (** p < 0.01, *** p < 0.001 compared to basal conditions), n = 10,000 events. (D) SH-SY5Y cells were loaded with MitoSOX (2 µM). The results represent the percentage ± SD of MitoSOX + cells detected by flow cytometry (** p < 0.01, compared to basal conditions), n = 10,000 events. Experiments were done three times.

Mitochondrial Morphological Changes with Necrostatin-1
The parkinsonian toxin rotenone causes mitochondrial fragmentation [23]. Thus, PD is related to abnormal mitochondrial morphology and oxidative stress [9]. To better understand the noxiousness of Nec-1, we questioned whether the mitochondrial injuries triggered by rotenone could be prevented with Nec-1 treatment. In this study, to evaluate mitochondrial morphology, we used TOMM20 staining and classified mitochondria into two groups: filamentous and damaged mitochondria. Damaged mitochondria were considered to be diffused or fragmented ( Figure 4A, right panel). Then, we observed that the percentage of cells with damaged mitochondria increased remarkably in all cell lines with rotenone and/or Nec-1 treatments ( Figure 4A,B). The inefficiency of Nec-1 in rotenone-treated cells was confirmed by the maintained level of damaged mitochondria and ROS generation ( Figure 4C). In fact, the percentage of ethidium + cells enhanced with rotenone was augmented with Nec-1. Moreover, Nec-1 alone affects mitochondrial morphology and increases ROS production, which could exacerbate the effect already caused by rotenone. Thus, we deduced that the toxicity of Nec-1 is related to mitochondrial dysfunction.
Antioxidants 2020, 9, x FOR PEER REVIEW 7 of 13 cells detected by flow cytometry (** p < 0.01, compared to basal conditions), n = 10,000 events. Experiments were done three times.

Mitochondrial Morphological Changes with Necrostatin-1
The parkinsonian toxin rotenone causes mitochondrial fragmentation [23]. Thus, PD is related to abnormal mitochondrial morphology and oxidative stress [9]. To better understand the noxiousness of Nec-1, we questioned whether the mitochondrial injuries triggered by rotenone could be prevented with Nec-1 treatment. In this study, to evaluate mitochondrial morphology, we used TOMM20 staining and classified mitochondria into two groups: filamentous and damaged mitochondria. Damaged mitochondria were considered to be diffused or fragmented ( Figure 4A, right panel). Then, we observed that the percentage of cells with damaged mitochondria increased remarkably in all cell lines with rotenone and/or Nec-1 treatments ( Figure 4A,B). The inefficiency of Nec-1 in rotenone-treated cells was confirmed by the maintained level of damaged mitochondria and ROS generation ( Figure 4C). In fact, the percentage of ethidium + cells enhanced with rotenone was augmented with Nec-1. Moreover, Nec-1 alone affects mitochondrial morphology and increases ROS production, which could exacerbate the effect already caused by rotenone. Thus, we deduced that the toxicity of Nec-1 is related to mitochondrial dysfunction. The results represent the percentage ± SD of ethidium + cells detected by flow cytometry (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.001, compared to basal conditions), n = 10,000 events. Experiments were done two times. The results represent the percentage ± SD of ethidium + cells detected by flow cytometry (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.001, compared to basal conditions), n = 10,000 events. Experiments were done two times.

Necrostatin-1 Impairs Mitochondrial Clearance
In fibroblasts treated four-hourly, rotenone as well as Nec-1 increased TOMM20 and PHB1 protein levels ( Figure 5A). This increase was accompanied by an enhancement of LONP1, a mitochondrial protease involved in the PINK1-dependent mitophagy pathway [24]. After 24 h of rotenone and/or Nec-1 exposure, the expression level of TOMM20, but not PHB1, was consistent with that observed at 4 h ( Figure 5B). We suggest that this could be due to inhibition of mitophagy, a selective mitochondrial clearance. In fact, Nec-1 has been reported to inhibit autophagosome formation in neurons treated with 6-OHDA [17]. To further elucidate the role of Nec-1 in mitochondrial degradation, we treated fibroblasts with CCCP ( Figure 5C), a potent mitophagy inducer. Remarkably, treatment with Nec-1 alone did not lead to a significant increase in MTG + fluorescence compared to Co cells, but it did inhibit CCCP-induced mitophagy in all cell lines. Although we did not observe mitophagy induction with rotenone in fibroblasts ( Figure 5C), we did notice a significant reduction in MTG + fluorescence in SH-SY5Y cells ( Figure 5D), as previously reported [25], which was prevented by Nec-1. Overall, neuroblastoma cells were found to be more sensitive than fibroblasts to rotenone. Despite the effect of rotenone on mitochondrial turnover in both cell models, Nec-1 accumulated mitochondrial mass.
Antioxidants 2020, 9, x FOR PEER REVIEW 8 of 13 In fibroblasts treated four-hourly, rotenone as well as Nec-1 increased TOMM20 and PHB1 protein levels ( Figure 5A). This increase was accompanied by an enhancement of LONP1, a mitochondrial protease involved in the PINK1-dependent mitophagy pathway [24]. After 24 h of rotenone and/or Nec-1 exposure, the expression level of TOMM20, but not PHB1, was consistent with that observed at 4 h ( Figure 5B). We suggest that this could be due to inhibition of mitophagy, a selective mitochondrial clearance. In fact, Nec-1 has been reported to inhibit autophagosome formation in neurons treated with 6-OHDA [17]. To further elucidate the role of Nec-1 in mitochondrial degradation, we treated fibroblasts with CCCP ( Figure 5C), a potent mitophagy inducer. Remarkably, treatment with Nec-1 alone did not lead to a significant increase in MTG + fluorescence compared to Co cells, but it did inhibit CCCP-induced mitophagy in all cell lines. Although we did not observe mitophagy induction with rotenone in fibroblasts ( Figure 5C), we did notice a significant reduction in MTG + fluorescence in SH-SY5Y cells ( Figure 5D), as previously reported [25], which was prevented by Nec-1. Overall, neuroblastoma cells were found to be more sensitive than fibroblasts to rotenone. Despite the effect of rotenone on mitochondrial turnover in both cell models, Nec-1 accumulated mitochondrial mass.

Discussion
Events that lead to neuronal death in PD are numerous and interconnected [26]. Thorough studies in PD cell death indicate that mitochondrial dysfunction, oxidative stress, and proteolysis deregulation together trigger cell death years before the appearance of the first clinical hallmarks [1]. Several cell death types can be found in postmortem brains of PD patients in in vitro and in vivo models of PD [26]. In this study, we explored necrotic programmed cell death (necroptosis) in primary fibroblasts from PD patients and used rotenone to model PD in cell culture. Necroptosis could be induced by ROS generation [27] and mitochondrial depolarization. Of note, the cells of PD patients displayed all of these previously cited characteristics [19]; however, necroptosis was not activated (Figure 1). Interestingly, rotenone induced necroptosis in SH-SY5Y cells ( Figure 2C,D) and in primary fibroblasts ( Figure 2E) by increasing the phosphorylation level of MLKL at ser358. If necroptosis was activated in postmortem brains of PD patients, it was not clearly activated in the primary fibroblasts. However, rotenone treatment made the cells of PD patients more susceptible to necroptosis activation. Rotenone-induced necroptosis was inhibited through the inactivation of RIP1, but other forms of cell death were activated or initiated in a RIP1-independent manner ( Figure 3B,C). In fact, Nec-1 was found to inhibit RIP1 kinase activity and prevent the phosphorylation of MLKL and the plasma membrane rupture related to p-MLKL [14]. Even though necroptosis was inhibited in MLKL-deficient cells, the latter could undergo apoptosis [28]. In addition, 6-OHDA induced necroptosis in mesencephalic and cortical neurons. Nevertheless, inhibition through RIP1 or MLKL did not reduce nuclei condensation but did protect against axonal degeneration [16]. Most certainly, Nec-1 doses beneath 30 µM showed a neuroprotective effect on 6-OHDA-treated PC12 cells [17]. The doses used in our study promoted apoptotic and necrotic cell death in both rotenone-treated cell models. These data are consistent with Jie's previous report [29].
Rotenone is a mitochondrial complex I inhibitor [30], that acts as one of the necroptosis initiators. It has become crucial to determine why cells are still more susceptible to rotenone treatment despite the inhibition of necroptosis by Nec-1 at 20 µM in SH-SY5Y cells or 30 µM in fibroblasts. As a matter of fact, Nec-1 was found to decrease MMP ( Figure 3A) and increase ROS production ( Figure 3D or Figure 4C). We hypothesized that the inhibition of necroptosis by Nec-1 is mitochondria-independent and that Nec-1 failed to reverse rotenone-induced mitochondrial alterations and oxidative stress ( Figure 4) by acting directly on the necrosome. It is noteworthy that Nec-1 downregulates autophagosome formation and lysosomal enzyme activity [17], illustrating that RIP1 kinase activity is not needed for autophagy inhibition [31]. Consistent with these reports, we think that Nec-1 inhibits mitophagy and fosters the accumulation of rotenone-induced mitochondrial damage which, in turn, induces oxidative stress, apoptosis, and necrosis. Swollen lysosomes (data not shown) may reveal the need to remove damaged mitochondria with Nec-1 treatment. Rotenone positively regulates mitophagy in SH-SY5Y cells via an externalized cardiolipin to the outer mitochondrial membrane, which binds to LC3 [25]. However, human fibroblasts showed lower sensitivity to rotenone ( Figure 5D) than SH-SY5Y cells ( Figure 5E), as reflected by the mitochondrial clearance response degree. Thus, the concentration of rotenone used was sufficient to trigger mitochondrial inclusion, but a higher concentration may be needed to remove accumulated mitochondria in fibroblasts. Nec-1 displayed different effects on internal and external mitochondrial membrane proteins at 24 h by Western blotting but increased the mitochondrial content with MTG staining by flow cytometry assay. Nec-1 was not shown to be effective for preventing rotenone-induced non-necroptotic cell death because of accumulation of mitochondrial damage and the activation of alternative mechanisms of cell death ( Figure 6). Moreover, when Nec-1 showed limited effects on the survival of neurons, RIP3 ablation protected the cells treated with MPP + against apoptosis [32]. However, divergent results suggest that the neuroprotective effect of RIP3 deletion is independent of apoptosis [15].
( Figure 6). Moreover, when Nec-1 showed limited effects on the survival of neurons, RIP3 ablation protected the cells treated with MPP + against apoptosis [32]. However, divergent results suggest that the neuroprotective effect of RIP3 deletion is independent of apoptosis [15]. In fact, it was found that Nec-1 does not have the ability to reverse the inhibition of mitochondrial complex I by rotenone. Therefore, Nec-1 exerts a dual effect on cells through the inhibition of RIP1 kinase activity. By downregulating mitophagy, Nec-1 prevents the turnover of rotenone-injured mitochondria, maintaining the production of ROS and the activation of other nonnecroptotic cell death pathways ( Figure 6). In spite of necroptosis inhibition, mitophagy deregulation contributes to Nec-1 neurotoxicity. It would be interesting to investigate the use of a combined treatment of Nec-1 and/or autophagy modulators and even antioxidants to inhibit necroptosis and ROS production in PD cell models.  In fact, it was found that Nec-1 does not have the ability to reverse the inhibition of mitochondrial complex I by rotenone. Therefore, Nec-1 exerts a dual effect on cells through the inhibition of RIP1 kinase activity. By downregulating mitophagy, Nec-1 prevents the turnover of rotenone-injured mitochondria, maintaining the production of ROS and the activation of other non-necroptotic cell death pathways ( Figure 6). In spite of necroptosis inhibition, mitophagy deregulation contributes to Nec-1 neurotoxicity. It would be interesting to investigate the use of a combined treatment of Nec-1 and/or autophagy modulators and even antioxidants to inhibit necroptosis and ROS production in PD cell models.
Supplementary Materials: The following are available online at http://www.mdpi.com/2076-3921/9/6/524/s1, Figure S1: Expression level of the MLKL protein in primary human fibroblasts under basal condition. and Isabel Gemio Foundation. This work was also partially supported by "Fondo Europeo de Desarrollo Regional" (FEDER) from the European Union.