Roles of receptor‐interacting protein kinase 1 in SH‐SY5Y cells with beta amyloid‐induced neurotoxicity

Abstract Alzheimer's disease (AD), the major cause of dementia, affects the elderly population worldwide. Previous studies have shown that depletion of receptor‐interacting protein kinase 1 (RIPK1) expression reverted the AD phenotype in murine AD models. Necroptosis, executed by mixed lineage kinase domain‐like (MLKL) protein and activated by RIPK1 and RIPK3, has been shown to be involved in AD. However, the role of RIPK1 in beta‐amyloid (Aβ)‐induced necroptosis is not yet fully understood. In this study, we explored the role of RIPK1 in the SH‐SY5Y human neuroblastoma cells treated with Aβ 1–40 or Aβ 1–42. We showed that Aβ‐induced neuronal cell death was independent of apoptosis and autophagy pathways. Further analyses depicted that activation of RIPK1/MLKL‐dependant necroptosis pathway was observed in vitro. We demonstrated that inhibition of RIPK1 expression rescued the cells from Aβ‐induced neuronal cell death and ectopic expression of RIPK1 was found to enhance the stability of the endogenous APP. In summary, our findings demonstrated that Aβ can potentially drive necroptosis in an RIPK1‐MLKL‐dependent manner, proposing that RIPK1 plays an important role in the pathogenesis of AD.

cells in vitro. 7 Additionally, Necrostatin-1 (Nec-1), a RIPK1 inhibitor, drastically increased the survival of murine neuronal cells in vitro and promoted memory and learning retention as well as cognitive performance in AD models. 8,9 Nec-1 treatment was further demonstrated to significantly diminish the level of Aβ oligomers, plaques and hyperphosphorylated tau. 9 Moreover, Yang et al. proved that Nec-1 could disaggregate Aβ fibrils and oligomers and block Aβ aggregateinduced brain cell death. 10 In SH-SY5Y cells, RIPK1expression was considerably upregulated after glutamate treatment, suggesting that the protein may specifically regulate programmed necrosis in glutamate-mediated excitatory toxicity. 11 Taken together, RIPK1 showed to play an important role in AD. However, evidence from these studies is insufficient to elucidate the role of RIPK1 in mediating Aβ-induced cell death and APP processing in human AD models. Hence, potential RIPK1-mediated cell death mechanisms were investigated in SH-SY5Y cells treated with Aβ, as well as its role on Aβ-induced cell death and APP protein stability. The results of this study may prove beneficial to enhance targeted therapies effective for AD in the future.
Emerging documentations showed that human neuroblastoma SH-SY5Y cells have been used as a cell model of neurodegenerative diseases, including AD. SH-SY5Y cells in both undifferentiated and differentiated forms express a number of neuronal markers such as tyrosine hydroxylase, muscarinic and nicotinic acetylcholine receptors, that are frequently used in in vitro neurological experiments. 12 In order to establish a human AD cell model, the undifferentiated SH-SY5Y cells are often challenged with Aβ or stably overexpressing APP to mimic the AD phenotypes in vitro. 13 were cultured in an incubator with 5% carbon dioxide at 37°C and passaged upon reaching 80% confluence.

| CellTiter-Glo luminescent cell viability assay
Cells were seeded in 384-well plate at a density of 6 × 10 3 cells per well and treated with Aβ and/or inhibitors (Z-vad, 3-Methyladenine (3-MA), Nec-1, necrosulfonamide (NSA), GSK-872) as described in Table S3 and Table S4 for 72 h. Briefly, CellTiter-Glo ® substrate and buffer (Promega Corporation) were mixed to form a homogenous solution. Prior adding to the cells, the mixture was equilibrated to room temperature and the required amount was diluted at 1:1 ratio with phosphate-buffered saline (PBS). The mixture was then added at 1:1 ratio to each well and mixed on an orbital shaker. The plate was left in room temperature for 10 min to stabilise the luminescent signal, which was subsequently recorded using the SoftMax Pro 6.3 application on the SpectraMax microplate reader.

| Protein stability assay
Cells were seeded in 6-well plate at a density of 6 × 10 5 cells per well. Following overnight incubation, the cells were treated with cycloheximide (CHX) (Cayman Chemicals) at a concentration of 10 µg/ml. Protein lysate was harvested and quantified at 0, 30, 60 and 90 min after treatment. The expression of RIPK1 and APP was determined by immunoblotting.

| Immunoblotting assay
Protein lysates were extracted using ice-cold lysis buffer (1% NP-40, 1mM DTT, supplemented with protease and phosphatase inhibitors in PBS) and subjected to immunoblotting. Primary antibodies against APP and β-actin were obtained from Biolegend, and Santa Cruz Biotechnology Inc, respectively. RIP, phospho-RIP, MLKL, phospho-MLKL, RIP3 and phospho-RIP3 were purchased from Cell Signalling Technology Inc. The dilution ratio used for each antibody was shown in Table S5. All images were captured using the ChemiDocTM XRS+molecular imager (Bio-Rad Laboratories).

| Quantitative real-time PCR analysis
Total RNA extraction and conversion of RNA to cDNA was per-  Table S6.

| Statistical analysis
All data were expressed as mean ± standard deviation from at least three independent experiments. The analysis was performed in Student's t test with Microsoft Excel, to compare the differences between two groups and one-way ANOVA, followed by Tukey's multiple comparison test with SPSS software version 28.0 (IBM SPSS), to compare the differences in more than two groups. Data with p < 0.01 were considered statistically significant.

| Aβ 1-42 had a higher tendency to form oligomers
To investigate the tendency of Aβ 40 and Aβ 42 to form oligomer or aggregates, both Aβs were reconstituted and incubated under the same conditions. As shown in Figure 1, Aβ 1-42 tended to produce large oligomers and aggregates, whilst Aβ 1-40 was mainly in its monomer or dimer forms (4-8 kDa). On the contrary, their corresponding reverse control peptides (Aβ 40-1 and Aβ 42-1) did not form aggregates.

| Aβ dose-dependently inhibited cell viability in SH-SY5Y cells
To investigate the neurotoxic effect of Aβ in SH-SY5Y cells, concentrations from 0 µM to 80 µM were used. The cell viability assay showed that with Aβ 1-40 or Aβ 1-42 at 80 µM transformed from its bipolarelongated shape into an irregular-polygonal shape, whereas no F I G U R E 1 Aβ 1-42 tends to form large aggregate peptides compared to Aβ 1-40. Aβ peptides were reconstituted with acetonitrile-water (1:5) in 5 mg/ml. The reconstituted peptides were incubated at 37°C for 72 h and 4°C for two weeks. Immunoblotting was performed to validate the aggregation status of Aβ. Reverse control peptides were included change was observed in cells treated with reverse control peptides ( Figure 2B).

| Aβ induces caspase-independent neuronal cell death
In this study, cells treated with Aβ were subjected to caspase activity assay. Figure 3A shows that the activity of apoptosis initiator caspase-8 and caspase-9 was no different between the Aβ-treated groups, reverse control peptide groups and control group (cell with culture medium only). Surprisingly, the activity of executioner caspase-3 was significantly increased in both Aβ 1-40 and Aβ 1-42, compared with their corresponding reverse control peptide group and control group.
To further investigate whether the neuronal cell death was dependent on caspase activity, cells were co-treated with Aβ and 10 µM caspase inhibitor Z-Vad. Figure 3B shows that Z-Vad did Since autophagy has been reportedly involved in AD, the role of autophagy in Aβ-induced neuronal cell death was assessed. As shown in Figure 3C, cells co-treated with Aβ and 10 mM of 3-MA, an autophagy inhibitor, did not improve the cell viability as compared to cells treated with Aβ alone, implying that autophagy may not be involved in mediating Aβ-induced neuronal cell death.

| Aβ and APP overexpression activate necroptosis
The role of Aβ and APP in the activation of necroptosis was assessed by evaluating their protein expression in Aβ-treated cells through immunoblotting. As shown in Figure 4A To further validate the role of Aβ in the activation of necroptosis, a rescue experiment was performed by co-treating the Aβ-treated cells with necroptotic-inhibitors. Figure 4C shows that Nec-1, a RIPK1 inhibitor and necrosulfonamide, a MLKL inhibitor, markedly rescued the Aβ-treated cells from neurotoxicity compared to cells treated with Aβ alone and control group. However, RIPK3 inhibitor (GSK-872) did not show any neuroprotection in Aβ-induced neuronal cell death, instead causing further reduction of cell viability. However, the difference was not significant.

| RIPK-1 is required for Aβ-induced neuronal cell death
In the present study, the role of RIPK1 in mediating Aβ-induced neuronal cell death was demonstrated in a human cell model.
Briefly, inhibition of RIPK1 protein expression was achieved by two

| RIPK1 mediation of Aβ-induced neuronal cell death is independent of caspase activity
To confirm whether the RIPK1 is involved in mediating Aβ-induced neuronal cell death via apoptosis, RIPK1-deficient cells were treated with Aβ at 80 µM for 72 h and subjected to a caspase activity assay. Figure 6 showed that the activity of all three caspases was not significantly different between the RIPK1-deficient cells and its control (Vec), suggesting that RIPK1 mediates Aβ-induced neuronal cell death through a caspase-independent mechanism.

| Endogenous APP protein stability is dependent on RIPK1
In this study, cells overexpressing RIPK1 were generated. As shown in Figure 7A, protein expression of endogenous APP was increased in cells overexpressing RIPK1. Surprisingly, qPCR results showed that the expression of endogenous APP did not show any differences compared with vector control (Figure 7B), suggesting that induction of endogenous APP protein expression does not depend on its mRNA transcriptional level. Additionally, the cycloheximide (CHX) chase assay showed that RIPK1 noticeably increased the lifespan of endogenous APP ( Figure 7C), suggesting that RIPK1 stabilises endogenous APP rather than transcriptional activation.

| DISCUSS ION
In the present study, we demonstrated that RIPK1 expression was significantly upregulated in cells treated with Aβ or cells overexpressing APP, which may lead to the development of necroptosismediated neuronal cell death in AD. Furthermore, inhibition of RIPK1 appears to be one of the underlying contributors to cell survival against Aβ-induced neurotoxicity. We further demonstrated that upregulation of RIPK1 enhanced the stability of endogenous APP.
Our data suggest that modulation of RIPK1 expression would have a significant impact on future therapeutic strategies for AD treatment.
To further investigate the role of increased caspase-3 activity, we co-treated the cells with Aβ and caspase inhibitor (Z-Vad) and the results showed that co-treatment with Z-vad did not significantly improve cell viability. Thus, these findings indicate that the neurotoxic effect of Aβ is independent of caspase activity. Aside from apoptosis, activation of caspase-3 has been shown to process and activate pro-interleukin-16 and stimulate T-lymphocytes. 34,35 Perhaps, increased caspase-3 activity in the current study was involved in activating the inflammatory process rather than apoptosis.
On the contrary, autophagy is a bulk degradation pathway for large and aggregated proteins such as Aβ, and has also been implicated in AD. [36][37][38][39] However, current findings show that co-treatment of Aβtreated cells with autophagy inhibitor (3-MA) did not show any rescue effects on cell viability.
Necroptosis is a programmed form of necrosis in mediating cell death and was first documented as a consequence of inflammation. 40 Receptor-interacting protein kinase 3 (RIPK-3) and its substrate, the pseudokinase mixed lineage kinase domain-like protein (MLKL), are core components of the necroptotic signalling pathway. 41 RIPK1 is known to be involved in RIPK-3-MLKL-dependent necroptosis. 42,43 Recently, activated necrosomes consisting of p-RIPK1, p-RIPK3 and p-MLKL were found in granulovacuolar degeneration lesions in the degenerating neurons of preclinical AD and AD patients. 23 However, there is no concrete evidence to support that necroptosis-mediated neuronal cell death is induced by Aβ.
Hence, the expression level of necroptotic-related proteins was investigated in current study, in both Aβ-treated and APP-overexpressed AD cell models. Our findings revealed that the expression levels of RIPK1, p-RIPK1, MLKL and p-MLKL were increased after exposure to Aβ. A similar observation was attained from ectopic expression of APP. Further experiments showed that inhibition of RIPK1 and MLKL by pharmacological means markedly increased the viability of Aβtreated cells. A similar study reported that both Nec-1 and NSA protected human embryonic stem cell-derived motor neurons (hES-MNs) from human sporadic ALS astrocyte toxicity. 44 Furthermore, inhibition of MLKL by NSA was able to reverse the interleukin-1β-induced nucleus pulposus cell death, suggesting that NSA could protect intervertebral disc degeneration via necroptosis and apoptosis inhibition. 45 A study conducted by Qiu et al. showed that inhibition of RIPK1 expression could activate apoptosis by cleaving the caspase-3; however, these findings are contradicted to the current study whereby the expression of RIPK1 and activity of caspase-3 were upregulated after the cells treated with Aβ. 46 More efforts are needed to investigate the role of caspase-3 and RIPK1 in mediating the necroptosis.
Although RIPK1 is known to be a death receptor that activates RIPK3 and MLKL to mediate necroptosis, 47 contradictory findings indicated that RIPK1 could intrinsically suppress spontaneous RIPK3 activation in the cytosol by controlling RIPK3 oligomerisation. 48 Furthermore, p-MLKL has been reported to activate necroptosis in the absence of death stimuli, and this observation was most profound in MLKL and RIPK3 double knockdown cells, suggesting that RIPK3 is a suppressor of MLKL activation. 49 Zhang et al. 50 showed that the lipopolysaccharide (LPS)-induced acute kidney injury increased tubular epithelial cell apoptosis and RIPK3 expression in mice. Conversely, inhibition of RIPK3 was shown to reduce the F I G U R E 5 Inhibition of RIPK1 expression protects cells from Aβinduced cell death. (A) Efficient inhibition of RIPK1 expression was attained by two independent shRNA constructs. Vector (Vec) was included as control for validation of knockdown efficiency. (B) Cell viability of Aβ 1-40 and Aβ 1-42 treated cells at different concentrations at 72 h. All data represent the mean ± standard deviation. Asterisks (*) indicate statistical significance (p < 0.01), compared with control group. (C) Morphological images were captured in cells treated with respective Aβ at 80 µM at 72 h. Images were taken at 100x magnification apoptosis in tubular epithelial cells and improve renal function in mice with LPS-induced acute kidney injury. 50 In cancer biology, knockdown of RIPK3 was reported to trigger p53 signalling and mitotic defects in murine recurrent tumour cells. 51 Furthermore, silencing of RIPK3 in recurrent tumour cells caused depletion of YAP and TAZ (transcriptional coactivator with PDZ-binding motif), which are essential in mediating the expression levels of proliferation-promoting and cancer-causing genes, indicating that RIPK3 might be responsible to the proliferation and recurrence of tumour cells 51,52 In the present study, we showed that the expression of RIPK-3 was decreased in both of the Aβ-treated and APP-overexpressed AD cell models, and inhibition of its activity by pharmacological means in Aβ-treated cells did not 'rescue' cell viability. Thus, the mechanism underlying the Aβ-induced neuronal cell death may be far more complex than previously thought. Further efforts are indeed necessary to elucidate the underlying molecular mechanism of RIPK3 in activating necroptosis induced by Aβ in the presence of RIPK1.
Emerging evidence indicates that Nec-1 displays a neuroprotective effect in reversing the clinical hallmarks and symptoms in a murine AD model. 7-10 However, these beneficial outcomes have yet to be elucidated on a human AD model. In this study, RIPK1-deficient cells were resistant to the cytotoxicity induced by Aβ, and its effect was independent of caspase activity. In accordance with our data, Re et al. demonstrated that depletion of RIPK1 protected human embryonic stem cell-derived motor neurons against sporadic ALS astrocyte toxicity. 44 Furthermore, Yang et al. showed that co-treatment of Nec-1 in Aβ-treated HT22 cells did not affect the expression of cleaved caspase-3. 9 However, Nec-1 has been shown to reduce the expression of cleaved caspase-3 in non-neuronal cells such as mouse microglial cells (BV2) and primary astrocytes treated with Aβ. 9 These contradicting outcomes suggest that RIPK1-mediated necroptosis may be cell-type specific, with a tendency to affect neuronal cells.
Aβ monomers have the ability to form oligomers and fibrils during the progression of AD. A recent study conducted by Yang et al. showed that Nec-1 was found to disaggregate Aβ fibrils and F I G U R E 6 RIPK1 mediates Aβ-induced neuronal cell death is independent of caspase activity. Caspase-3, caspase-8 and caspase-9 activities were measured after treating the cells with respective Aβ at 80 µM at 72 h F I G U R E 7 RIPK1 stabilises endogenous APP protein independent of APP mRNA transcription. (A) RIPK1 was overexpressed in cells. Vector (Vec) was included as control for validation of overexpression efficiency. (B) mRNA expression of APP in cells overexpressing RIPK1. All data were normalised against GAPDH and represent the mean ± standard deviation. (C) RIPK1 increased the lifespan of endogenous APP. Cells were treated with 10 µg/mL CHX at 0, 30, 60 and 90 min oligomers and inhibit brain cell death. 10 In the mentioned study, thioflavin-T fluorescence assay was performed to quantify the aggregation of Aβ. 10 Results showed that Nec-1 significantly reduced Aβ fibrils. 10 Furthermore, cytotoxicitiy assay of pre-formed Aβ 1-42 aggregates in cells revealed that Nec-1 demonstrated neuroprotective effect against Aβ aggregates, proposing that such anti-neuronal cell death effects were contributed by Nec-1 in dissociating Aβ aggregates. 10 An in vivo study also showed that Nec-1 was able to diminish Aβ plaques in the brains of aged APP/PS1 mice. 10 However, data on the roles of RIPK1 in the Aβ precursor protein are highly limited. In accordance with a previous study, our findings indicate that ectopic expression of RIPK1 increases the expression and stability of endogenous APP protein independent of mRNA expression. Hence, we postulate that RIPK1 is involved in the regulation of endogenous APP stability and perhaps acts to accelerate AD progression.

| CON CLUS ION
In summary, we demonstrated that Aβ-induced neuronal cell death involves an RIPK1/MLKL-dependent necroptosis mechanism in an in vitro human AD model. Furthermore, the expression of RIPK1 was significantly increased in both Aβ-treated and APP-overexpressed AD cell models. Additionally, inhibition of RIPK1 expression rescued the cells from Aβ-induced neuronal cell death, whilst ectopic expression of RIPK1 enhanced the stability of endogenous APP. Contrary to the canonical necroptosis pathway, Aβ-induced RIPK1-MLKLdependent necroptosis is independent of RIPK3 activity. Hence, these findings suggest that RIPK1 could be a potential target in reversing the pathogenesis of AD.

| FUTURE PROS PEC TS
Although current findings showed that RIPK1 mediated Aβ-induced neuronal cell death in RIPK1-MLKL-depedent necroptosis and independent of RIPK3 activity, however, as forementioned the neuropathological hallmarks of AD include the deposition of extracellular Aβ and intracellular neurofibrillary tau proteins. Perhaps in the future study, the role of RIPK1 in tau protein-induced neuronal cell death needs to be investigated. In addition to neuronal cells, participating cells in the brain include astrocytes, oligodendrocytes and microglial are important in maintaining brain homeostasis and function.
Therefore, developing a 3D-culture or coculture system is needed to investigate the role of RIPK1 in the pathogenesis of AD. Lastly, further downstream assays are needed to explain the increased caspase-3 observed in current study and elucidate the interaction of RIPK1, RIPK3, MLKL and caspase-3 in mediating necroptosis in AD.

ACK N OWLED G EM ENTS
This work was supported by a research grant from the Ministry of Higher Education, Malaysia (FRGS/1/2016/SKK08/IMU/03/3).

CO N FLI C T O F I NTE R E S T
The author(s) declare no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available in the supplementary material of this article.