Low Levels of Brain-Derived Neurotrophic Factor Trigger Self-aggregated Amyloid β-Induced Neuronal Cell Death in an Alzheimer’s Cell Model

Abstract

Alzheimer’s disease (AD) is pathologically characterized by accumulation of amyloid β (Aβ) and hyperphosphorylated tau, and thereby induction of neuronal cell death. The Aβ-induced neuronal cell death has been shown to occur by several modes, such as apoptosis, necrosis, and necroptosis. Interestingly, in AD patients, the brain and serum levels of brain-derived neurotrophic factor (BDNF) have been reported to be significantly decreased. However, the relationship between Aβ and BDNF in the onset of AD remains to be fully understood. Here, we used neuron-like differentiated human neuroblastoma SH-SY5Y (ndSH-SY5Y) cells to study the neurotoxicity of self-aggregated Aβ_1–42 peptide under different concentrations of BDNF in the culture medium. Importantly, decreasing levels of BDNF caused a considerable suppression in the extension of neurite length. Furthermore, only under low levels of BDNF, the aggregated Aβ was revealed to induce neurite fragmentation and neuronal cell death in ndSH-SY5Y cells. Notably, the aggregated Aβ and low levels of BDNF-induced neuronal cell death was characterized at least as caspase-6 dependent cell death and necroptosis. These results indicate that our ndSH-SY5Y cell system, cultured under decreasing levels of BDNF and aggregated Aβ, has the potential to be applied in the analysis of the molecular mechanisms of the progressive neurodegenerative processes of AD and the discovery of neuroprotective drug candidates.

INTRODUCTION

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder with accompanying cognitive dysfunction.^1,2) Brains affected by AD are characterized by the accumulation of extracellular deposits of aggregated amyloid β (Aβ) and intracellular neurofibrillary tangles of hyperphosphorylated tau.^3–5) These abnormal misfolded proteins have been reported to induce chronic inflammation and several types of cell death in neurons, such as apoptosis (caspase-dependent or -independent, and regenerative cell death), apobiosis (caspase-dependent or -independent, and non-regenerative cell death), necrosis (non-regulative cell death), and necroptosis (regulated necrotic cell death).^4,6–8)

Meanwhile, brain-derived neurotrophic factor (BDNF), which is a neurotrophin that is widely distributed in the adult brain, is known to play an important role in the survival of neurons, modulating synaptic plasticity, cognition, and memory.^9,10) The BDNF gene is expressed in the cortex, hippocampus, and basal brain region, which are indispensable for memory, learning, and higher cognitive functions. Therefore, alterations in BDNF levels and expression may induce synapse loss and cognitive dysfunction. Interestingly, brain¹¹⁾ and serum¹²⁾ levels of BDNF have been shown to be significantly decreased in AD patients. Furthermore, it is known that high peripheral BDNF levels protect older adults against AD.¹³⁾

Regarding the interaction between Aβ and BDNF in AD, many studies have shown opposite and conflicting effects on neuronal functions.¹⁴⁾ Aβ has been suggested to be involved in the down-regulation of BDNF expression and production in AD.¹⁵⁾ However, in 3 × Tg AD model mice, a reduction in BDNF altered neither Aβ nor tau pathology.¹⁶⁾ This implies that a decrease in BDNF level is not a cause, but rather a consequence of abnormal Aβ deposition. On the other hand, the transcriptional levels of BDNF have been proven to reduce the amyloidogenic processing of amyloid precursor protein (APP), and thereby downregulate Aβ formation and accumulation.¹⁷⁾ It is possible that application of BDNF will decrease the abnormal production of Aβ and prevent inflammation and synaptic dysfunction in AD.^17–19) Although these observations indicate that BDNF plays an important role in the pathogenesis of Aβ-induced AD, it has not been established whether BDNF is directly involved in aggregated Aβ induced neuronal cell death in AD.

In this study, we investigated the neuronal response to self-aggregated Aβ_1–42 peptide in the culture medium by using cultured neuron-like differentiated SH-SY5Y (ndSH-SY5Y) cells under different levels of BDNF. Importantly, decreasing levels of BDNF suppressed neurite extension, and combinational treatments of aggregated Aβ with low levels BDNF could induce potentiated neurite fragmentation and at least two types of neuronal cell death, caspase-6-dependent cell death and necroptosis. This is the first report demonstrating that decreasing BDNF could trigger two types of aggregated Aβ-induced cell death in an Alzheimer’s neuronal cell model. These findings suggest that the cooperative effects of aggregated Aβ and low levels of BDNF on the induction of neuronal cell death might cause the neurodegenerative symptoms of AD.

MATERIALS AND METHODS

Reagents

All-trans retinoic acid (RA) and human-recombinant animal-free BDNF were obtained from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). Amyloid β-peptide 1–42, O-acyl isopeptide (Aβ), which self-aggregates in the culture medium, was obtained from Peptide Institute Inc. (Osaka, Japan). NucBlue Live Cell Stain Ready Probes Reagent and 7-aminoactinomysin D (7-AAD) were obtained from Thermo Fisher and Life Technology (Waltham, MA, U.S.A.), respectively. A pan-caspase inhibitor, Z-VAD-fmk, and a caspse-6 specific inhibitor, Z-VEID-fmk, were obtained from BioVision (Milipitas, CA, U.S.A.) and R&D systems (Minneapolis, MN, U.S.A.), respectively. The necroptosis inhibitors Necrostatin-1 (Nec-1) and necrostatin-1 inactive (Nec-1i) were obtained from Cayman Chemical (Ann Arbor, MI, U.S.A.).

Cell Culture

The human neuroblastoma cell line SH-SY5Y (ATC C^® CRL-2266™) was obtained from the American Type Culture Collection (Manassas, VA, U.S.A.). Cells were cultured in Dulbecco’s modified Eagle’s Medium (D-MEM)/Ham’s F-12 containing 10% heat-inactivated fetal bovine serum, 100 units/mL penicillin, and 100 µg/mL streptomycin in a 37°C incubator under an atmosphere of 5% CO₂ and 100% relative humidity.

Cell Differentiation

SH-SY5Y cells were inoculated at an initial density of 1.0 × 10⁵ cells/dish in a collagen I-coated ϕ3.5 cm dish. RA was added 2 d after plating at a final concentration of 10 µM in high-glucose D-MEM with 15% heat-inactivated fetal bovine serum. After 5 d in the presence of RA, cells were washed with high-glucose D-MEM and incubated with 100 ng/mL BDNF in high-glucose D-MEM without L-glutamine and phenol red containing 4 mM sodium pyruvate and 1 mM L-glutamine for 2 d.

Cell Viability

Neuron-like differentiated SH-SY5Y cells (ndSH-SY5Y) were treated with various inhibitors under a low concentration of BDNF (10 ng/mL) in the presence or absence of 3 or 10 µM self-aggregated Aβ for 96 h, and then stained with 7-AAD and Hoechst33342 stain. Cell viability was analyzed using LAS AF with a DMI 6000B-AFC microscope (Leica, Wetzlar, Germany) at 200× magnification and calculated according to the proportion of 7-AAD positive cells.

Morphological Observation

Cell morphology was observed under a DMil microscope (Leica) using LAS V4.12 at 200× magnification. Measurements of neurite length and neurite fragmentation were analyzed by the image processing software ImageJ using NeuronJ and Particle Analysis. Neurite fragments were measured as follows: image binarization followed by automatic calculation by Particle Analysis using appropriate parameters (size: 0.00005–0.0005; circularity: 0.40–1.00).

Western Blotting

Western blot analysis was performed as described previously.^20,21) Intensity of total protein bands was used as an internal control. Total proteins bands were detected by 12% Mini-PROTEAN TGX Stain-Free gel (Bio-Rad, Hercules, CA, U.S.A.), then transfer to polyvinylidene difluoride membrane and immunoblotting. The following antibodies were used: rabbit anti-GAP43 (1 : 2000, Abcam, Cambridge, U.K.), horseradish peroxidase-linked anti-rabbit immunoglobulin G (1 : 20000, GE Healthcare, Pittsburgh, PA, U.S.A.).

Statistical Analysis

Data are presented as the means ± standard error. The significance of differences among groups was evaluated using the Student’s t-test or one-way ANOVA followed by the Turkey–Kramer test; p < 0.05 was considered to indicate a statistically significant difference.

RESULTS

Characteristics of the Neuronal Differentiated SH-SY5Y Cell System

In order to investigate the relationship between the abnormal production of Aβ and decreased levels of BDNF observed in the brain and serum levels of AD patients and the induction of neuronal cell death, we established an Alzheimer’s mimic cell model using neuron-like differentiated SH-SY5Y (ndSH-SY5Y) cells. Then, we analyzed self-aggregated Aβ-induced neuronal dysfunction and cell death modes under decreasing concentrations of BDNF.

We first prepared ndSH-SY5Y cells using a differentiation protocol modified from previously described protocols.^22–24) In order to differentiate SH-SY5Y cells into neuron-like cells, they were treated for 5 d (days 2–7) with 10 µM RA before treatment for 2 d (days 7–9) with 100 ng/mL BDNF (Fig. 1A). Based on morphological observations, ndSH-SY5Y cells ceased to proliferate on day 5 and their morphologies became neuron-like, including the development of long neurites, on days 5–9 (Fig. 1B). The average length of neurites of ndSH-SY5Y cells on day 9 was approximately 87 µm, which was longer than that in undifferentiated control SH-SY5Y cells on day 2 (neurite length approximately 29 µm) (Fig. 1C). The ndSH-SY5Y cells expressed a neuron-specific marker protein, GAP43, on day 9, which was detected by immunoblot analysis (Fig. 1D). Thus, this protocol successfully produced the ndSH-SY5Y cell system, a neuronal cell model.

Fig. 1. Characteristics of Neuron-Like Differentiation of Human Neuroblastoma SH-SY5Y Cells

(A) Schedule of neuron-like differentiation of SH-SY5Y cells, and (B) the cell morphologies of neuronal differentiating neuroblastoma SH-SY5Y cells or neuron-like differentiated SH-SY5Y (ndSH-SY5Y) cells. Scale bar = 100 µm. Day 2, cells exhibited neuroblastoma cell morphology; Day 5, after medium change, the cells were treated with 10 µM RA for 3 d; Day 7, after medium change, the cells were treated with 10 µM RA for 2 d; Day 9, after medium change, the cells were treated with 100 ng/mL BDNF for 2 d and differentiated to neuron-like cells (ndSH-SY5Y). (C) The neurite length of neuroblastoma SH-SY5Y cells on day 2 or ndSH-SY5Y cells on day 9. (D) The protein expression of a neuron-marker GAP43 in neuroblastoma SH-SY5Y cells on day 2 or ndSH-SY5Y cells on day 9. Total protein bands are shown as an internal control. (E) Cell viability and (F) neurite length of ndSH-SY5Y cells on day 13 under the culture conditions in the presence of 100 or 10 ng/mL BDNF for 4 d from day 9. * p < 0.05, ** p < 0.01, *** p < 0.001. (Color figure can be accessed in the online version.)

Effects of Decreasing BDNF Levels and Aggregated Aβ on the Activities of ndSH-SY5Y Cells

Using the ndSH-SY5Y cell system, we next investigated the effect of decreasing BDNF concentrations (100–10 ng/mL for 4 d, days 9–13) on the activities of ndSH-SY5Y cells. The viability of ndSH-SY5Y cells was decreased by lowering concentrations of BDNF in a dose-dependent manner. The viability of ndSH-SY5Y cells cultured for 4 d in the presence of 10 ng/mL BDNF (on day 13) was decreased by approximately 15% compared with that of cells maintained under normal culture conditions (100 ng/mL BDNF for 4 d) (Fig. 1E). Surprisingly, the average length of the neurite of ndSH-SY5Y cells on day 13 extended at most 105 µm from the control average length of 87 µm on day 9 (Fig. 1C) under decreasing levels of BDNF (10 ng/mL) compared with 140 µm under normal culture conditions (100 ng/mL BDNF) (Fig. 1F) (an extension suppression of approximately 70%). These data indicate that decreasing levels of BDNF caused a considerable suppression in the extension of neurite length.

Then, we next examined the effect of a self-aggregated form of Aβ_1–42 peptide on neurite fragmentation and neuronal cell death in ndSH-SY5Y cells under decreasing levels of BDNF (100–10 ng/mL) conditions for 4 d (day 9–13). We found that neurite fragmentation was induced by self-aggregated Aβ in a dose and a time-dependent manner (Figs. 2A, B). Further, the proportion of 7-AAD-positive dead cells was increased in a dose-dependent manner by aggregated Aβ treatment (Figs. 2C, D). The neurite fragmentation and cell death in ndSHSY-5Y cells were increased by the treatment with Aβ under BDNF decreasing condition more than the BDNF decreasing condition alone.

Fig. 2. Neurotoxic Effects of Self-aggregated Aβ_1–42 Peptide on ndSH-SY5Y Cells under the Low Level BDNF Conditions

(A) Cell morphologies of ndSH-SY5Y cells treated with 3 µM Aβ under the decreasing (10 ng/mL) or healthy-high (100 ng/mL) BDNF conditions for 4 d (on day 13). Scale bar = 100 µm. (B) Neurite fragmentation of ndSH-SY5Y cells under the indicated conditions for 96 h (from day 9 to day 13). (C) The images of Hoechst and 7-AAD stained ndSH-SY5Y cells under the indicated conditions on for 4 d (on day 13). Scale bar = 50 µm. B10 + Aβ3, the cells were treated with 10 ng/mL BDNF and 10 µM Aβ; B10 + Aβ10, the cells were treated with 10 ng/mL BDNF and 10 µM Aβ. (D) The 7-AAD positive dead cells of ndSH-SY5Y cells under the indicated condition at for 4 d (on day 13). (Color figure can be accessed in the online version.)

We investigated the effects of BDNF supplementation on the neurite fragmentation and neuronal cell death induced by aggregated Aβ and low levels of BDNF in ndSH-SY5Y cells. Neurite fragmentation (Figs. 3A, B) and cell death (Figs. 3C, D) were suppressed by supplementation with 100 ng/mL BDNF. These observations confirm that low levels of BDNF trigger the expression of the neurodegenerative activities of aggregated Aβ. This finding suggests that a reduction in BDNF levels could trigger the manifestation of the neurotoxic effects of aggregated Aβ, such as neurite fragmentation and neuronal cell death, in ndSH-SY5Y cells.

Fig. 3. Effects of Supplementation of Normal (High) Level BDNF on Neurotoxicity of Self-aggregated Aβ_1–42 Peptide on ndSH-SY5Y Cells

(A) Cell morphologies of ndSH-SY5Y cells treated with 3 µM Aβ under the BDNF-low (10 ng/mL) or healthy-high (100 ng/mL) conditions for 4 d (on day 13). Scale bar = 100 µm. (B) Neurite fragmentation of ndSH-SY5Y cells under the indicated conditions for 4 d (on day 13). (C) The images of Hoechst and 7-AAD stained ndSH-SY5Y cells under the indicated conditions on for 4 d (on day 13). Scale bar = 50 µm. B10 + Aβ, the cells were treated with 10 ng/mL BDNF and 3 µM Aβ; B100 + Aβ, the cells were treated with 100 ng/mL BDNF and 3 µM Aβ. (D) The 7-AAD positive dead cells of ndSH-SY5Y cells under the indicated condition at for 4 d (on day 13). * p < 0.05, ** p < 0.01, *** p < 0.001. (Color figure can be accessed in the online version.)

Neuronal Cell Death Modes in ndSH-SY5Y Cells Induced by Treatment with Aggregated Aβ under Low Levels of BDNF

To elucidate the types of neuronal cell death induced by treatment with aggregated Aβ under low levels of BDNF, we investigated the activation of caspase in the ndSH-SY5Y cells. As shown in Fig. 4A, the proportion of 7-AAD-positive dead cells was drastically increased by treatment with 3 µM Aβ for 96 h (day 13) in the presence of 10 ng/mL BDNF. Interestingly, treatment with a pan-caspase inhibitor, Z-VAD-fmk, resulted in a decrease in the proportion of 7-AAD-positive dead cells, even under low levels of BDNF (Fig. 4B).

Fig. 4. Association of Aggregated Aβ-Induced Cell Death and Caspase-6 in ndSH-SY5Y Cells under the Decreasing BDNF Conditions

(A) Hoechst and 7-AAD staining of ndSH-SY5Y cells after treatment with or without 100 µM Z-VAD-fmk under the indicated conditions for 4 d (on day 13). Scale bar = 50 µm. B10 + Aβ, the cells were cotreated with 10 ng/mL BDNF and 3 µM Aβ; B10 + Aβ + Z-VAD, the cells were cotreated with 10 ng/mL BDNF, 3 µM Aβ, and 100 µM Z-VAD-fmk. (B) The 7-AAD positive dead cells of ndSH-SY5Y cells under the indicated conditions for 4 d (on day 13). (C) The images of Cell morphologies and (D) the neurite fragmentation of ndSH-SY5Y cells under the indicated conditions for 4 d (on day 13). Scale bar = 100 µm. (E) Hoechst and 7-AAD staining of ndSH-SY5Y cells after treatment with or without Z-VEID-fmk under the indicated conditions for 4 d (on day 13). Scale bar = 50 µm. B10 + Aβ, the cells were cotreated with 10 ng/mL BDNF and 3 µM Aβ; B10 + Aβ + Z-VEID, the cells were cotreated with 10 ng/mL BDNF, 3 µM Aβ and 100 µM Z-VEID-fmk. (F) The 7-AAD positive dead cells (%) of ndSH-SY5Y cells under the indicated conditions on day 13. * p < 0.05, ** p < 0.01, *** p < 0.001. (Color figure can be accessed in the online version.)

To determine which caspase is involved in Aβ-induced cell death under low levels of BDNF, we analyzed the effect of a caspase-6-specific inhibitor on the induced cell death, since caspase-6-dependent neuronal cell death has previously been observed in AD patients.^25,26) As we expected, neurite fragmentation (Figs. 4C, D) and the proportion of 7-AAD-positive dead cells (Figs. 4E, F) in the aggregated Aβ-treated ndSH-SY5Y cells under low levels of BDNF were significantly suppressed by treatment with the caspase-6 specific inhibitor Z-VEID-fmk.

Furthermore, we investigated the occurrence of necroptosis as a regulated necrosis in neurite fragmentation and cell death induced in ndSH-SY5Y cells by aggregated Aβ and low levels of BDNF using the necroptosis inhibitors necrostatin-1 (Nec-1) and necrostatin-1 inactive (Nec-1i).²⁷⁾ We found that neurite fragmentation was suppressed by treatment with Nec-1 (45%) or Nec-1i (71%) (Figs. 5A, B). The percentage of neurite fragment by treatment with Nec-1 or Nec-1i was shown relative to Aβ-treated ndSHSY-5Y cells under low levels of BDNF condition (100%). The proportion of 7-AAD-positive dead cells in the treated ndSH-SY5Y cells was reduced by approximately 25 and 15% by treatment with Nec-1 and Nec-1i, respectively (Figs. 5C, D). Several researchers have previously reported that necroptotic cell death is observed in AD patients.^7,28) Taken together, these observations suggest that aggregated Aβ and low levels of BDNF could induce at least two cell death modes, caspase-6-dependent cell death and necroptosis.

Fig. 5. Effects of Necroptosis Inhibitors on Induction of Cell Death after Aggregated Aβ-Treatment in ndSH-SY5Y Cells under the Decreasing BDNF Conditions

(A) Cell morphologies of ndSH-SY5Y cells after treatment with vehicle (solvent alone), Nec-1, and Nec-1i under the indicated conditions for 4 d (on day 13). Scale bar = 100 µm. (B) The neurite fragmentation of ndSH-SY5Y cells under the indicated conditions for 4 d (on day 13). (C) The images of Hoechst and 7-AAD stained ndSH-SY5Y cells after treatment with Nec-1 or Nec-1i under the indicated conditions for 4 d (on day 13). Scale bar = 50 µm. B10 + Aβ, the cells were cotreated with 10 ng/mL BDNF and 3 µM Aβ; B10 + Aβ + Nec-1, the cells were cotreated with 10 ng/mL BDNF, 3 µM Aβ and 20 µM Nec-1; B10 + Aβ + Nec-1i, the cells were cotreated with 10 ng/mL BDNF, 3 µM Aβ and 20 µM Nec-1i. (D) The 7-AAD positive dead cells (%) of ndSH-SY5Y cells under the indicated conditions on day 13. * p < 0.05, ** p < 0.01, *** p < 0.001. (Color figure can be accessed in the online version.)

DISCUSSION

Neurons exhibit robustness and plasticity maintained by high levels of neurotrophic signals, aerobic glycolysis, and proteolytic degradation systems for abnormal and misfolded proteins.^29,30) The pathological hallmarks of AD are defined as proteinopathy, characterized by the accumulation of Aβ and hyperphosphorylated tau, which are extensively found in brains affected by AD.^3–5) Indeed, the accumulation of condensed plaques and neurofibrillary tangles composed of aggregated Aβ peptide and hyperphosphorylated tau, respectively, are known to induce neuronal cell death in AD.^4,6,7) Recently, the effects of degradation of senile plaques and neurofibrillary tangles through biochemical and biomedical strategies, including treatments with neprilysin and anti-Aβ antibodies, on AD progression are examining against AD in model mice.^31,32)

In the present study, we showed that the neurotoxic effect of aggregated Aβ promotes neuronal cell death in ndSH-SY5Y cells, but only under decreasing BDNF conditions. This finding suggests that decreasing levels of BDNF could trigger the manifestation of the neurotoxic effects of aggregated Aβ, and thereby induce neuronal cell death. Our observations are in agreement with those reported in neurons during the onset of AD.^11–13) In light of this, we focused on delineating the features of neuronal cell death events induced by aggregated Aβ under decreasing levels of BDNF in our AD mimic cultured cell system. The cell death modes that occurred in our system under these conditions were revealed to include at least caspase-6-dependent cell death and necroptosis, as was the case in previous in vivo studies of AD brain neurons.³³⁾ Interestingly, the aggregated Aβ-induced neurite fragmentation and cell death suppressed by treatment with Nec-1 and Nec-1i (Fig. 5). These finding suggest that the AD neuronal cell model occurs necroptosis and other non-programmed cell death, i.e., necrosis. Of note, it is known that the necroptosis inhibitor Nec-1 inhibits the receptor interacting serine/threonine kinase 1 (RIPK1) and indoleamine 2,3-dioxygenase (IDO).^34,35) In addition, Nec-1 inactive analogue Nec-1i inhibits IDO, but lacks RIPK1 inhibitory activity.^34,35) Souza et al., reported that the Aβ increases IDO activity in brain of mice.³⁶⁾ IDO is the first and rate-limiting enzyme in the kynurenine pathway, its major route for the metabolism of the tryptophan. Several metabolites of this pathway, such as 3-hydroxykynurenine and quinolinic acid, are known to have neurotoxic effects.³⁷⁾ These findings raise the possibility that the AD neuronal cell model partially occurs cell death mediated the accumulation of the neurotoxic tryptophan metabolites. We consider that the neurotoxic metabolites-induced cell death suppresses by treatment with Nec-1 and Nec-1i.

Our findings, therefore, could provide a framework for the development of therapies for AD. Up-regulation of BDNF levels could be an effective therapeutic strategy to prevent neuronal cell death and mitigate synaptic dysfunction, and thereby slow multiple AD pathologies. This may be achieved by regulating BDNF production by controlling both transcription and post-transcriptional BDNF expression. Thus, the molecular mechanisms by which the down-regulation of BDNF occurs should be extensively investigated to identify potential targets for the development of effective therapies for AD. Furthermore, the molecular mechanisms of crosstalk between BDNF and Aβ signaling transduction systems in neuronal cell death or survival should be precisely elucidated using in vitro neuronal cellular systems. Our AD neuronal cell mimetic system will be very significant in the analysis of the molecular mechanisms of this crosstalk to determine neuronal cell fates. Also, this system will aid in the discovery of neuroprotective drug candidates. However, although our results provide important clues that can be applied in the development of innovative chemical agents for the regulation of signal transduction systems, such as BDNF/tyrosine kinase B (Trk B) (neuron functions and maintenance) and Aβ/receptor for advanced glycation end products (RAGE) (chronic inflammation), additional neuronal cell-based and in vivo studies are needed to verify the efficacies of such chemotherapies. The molecular mechanisms how BDNF regulated neuronal cell death remains to be determined. These are our subject of ongoing investigation. In particular, we would like to further investigating the relationship with effects of BDNF and Trk B signaling in Aβ-induced neurite fragmentation and cell death.

Acknowledgments

This work was supported by a Research Education Fund for Tokyo University of Science. We thank Dr. Takahiro Oyama, and Dr. Hideaki Abe (Hinoki Shinyaku Co., Ltd.) for helpful discussions.

Author Contributions

Conceived and designed the experiments: NT, AT, AS, and ST. Performed the experiments: NT and AT. Analyzed the data: NT, AT, AS, FU, and ST. Wrote the paper: NT, AS, and ST.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) Scheff SW, Price DA, Schmitt FA, Mufson EJ. Hippocampal synaptic loss in early Alzheimer’s disease and mild cognitive impairment. Neurobiol. Aging, 27, 1372–1384 (2006).
• 2) Liu-Seifert H, Siemers E, Sundell K, Price K, Han B, Selzler K, Aisen P, Cummings J, Raskin J, Mohs R. Cognitive and functional decline and their relationship in patients with mild Alzheimer’s dementia. J. Alzheimers Dis., 43, 949–955 (2014).
• 3) Duyckaerts C, Delatour B, Potier MC. Classification and basic pathology of Alzheimer disease. Acta Neuropathol., 118, 5–36 (2009).
• 4) Fricker M, Tolkovsky AM, Borutaite V, Coleman M, Brown GC. Neuronal cell death. Physiol. Rev., 98, 813–880 (2018).
• 5) Scheltens P, Blennow K, Breteler MM, de Strooper B, Frisoni GB, Salloway S, Van der Flier WM. Alzheimer’s disease. Lancet, 388, 505–517 (2016).
• 6) Obulesu M, Lakshmi MJ. Apoptosis in Alzheimer’s disease: an understanding of the physiology, pathology and therapeutic avenues. Neurochem. Res., 39, 2301–2312 (2014).
• 7) Caccamo A, Branca C, Piras IS, Ferreira E, Huentelman MJ, Liang WS, Readhead B, Dudley JT, Spangenberg EE, Green KN, Belfiore R, Winslow W, Oddo S. Necroptosis activation in Alzheimer’s disease. Nat. Neurosci., 20, 1236–1246 (2017).
• 8) Tanuma S. Biological implications of apoptosis/apobiosis in senescence and life span. Nagoya Med. J., 43, 73–80 (2000).
• 9) Acheson A, Conover JC, Fandl JP, DeChiara TM, Russell M, Thadani A, Squinto SP, Yancopoulos GD, Lindsay RM. A BDNF autocrine loop in adult sensory neurons prevents cell death. Nature, 374, 450–453 (1995).
• 10) Arancio O, Chao MV. Neurotrophins, synaptic plasticity and dementia. Curr. Opin. Neurobiol., 17, 325–330 (2007).
• 11) Connor B, Young D, Yan Q, Faull RL, Synek B, Dragunow M. Brain-derived neurotrophic factor is reduced in Alzheimer’s disease. Brain Res. Mol. Brain Res., 49, 71–81 (1997).
• 12) Ng TKS, Ho CSH, Tam WWS, Kua EH, Ho RC. Decreased serum brain-derived neurotrophic factor (BDNF) levels in patients with Alzheimer’s disease (AD): a systematic review and meta-analysis. Int. J. Mol. Sci., 20, 257 (2019).
• 13) Weinstein G, Beiser AS, Choi SH, Preis SR, Chen TC, Vorgas D, Au R, Pikula A, Wolf PA, DeStefano AL, Vasan RS, Seshadri S. Serum brain-derived neurotrophic factor and the risk for dementia: the Framingham Heart Study. JAMA Neurol., 71, 55–61 (2014).
• 14) Song JH, Yu JT, Tan L. Brain-derived neurotrophic factor in Alzheimer’s disease: risk, mechanisms, and therapy. Mol. Neurobiol., 52, 1477–1493 (2015).
• 15) Francis BM, Kim J, Barakat ME, Fraenkl S, Yücel YH, Peng S, Michalski B, Fahnestock M, McLaurin J, Mount HT. Object recognition memory and BDNF expression are reduced in young TgCRND8 mice. Neurobiol. Aging, 33, 555–563 (2012).
• 16) Castello NA, Green KN, LaFerla FM. Genetic knockdown of brain-derived neurotrophic factor in 3xTg-AD mice does not alter Aβ or tau pathology. PLOS ONE, 7, e39566 (2012).
• 17) Rohe M, Synowitz M, Glass R, Paul SM, Nykjaer A, Willnow TE. Brain-derived neurotrophic factor reduces amyloidogenic processing through control of SORLA gene expression. J. Neurosci., 29, 15472–15478 (2009).
• 18) Li QS, Yang W, Pan YF, Min J, Zhang Z, Gao HZ, Qi JS. Brain-derived neurotrophic factor prevents against amyloid beta protein-induced impairment of hippocampal in vivo long-term potentiation in rats. Zhongguo Ying Yong Sheng Li Xue Za Zhi, 28, 425–429 (2012).
• 19) Peng S, Garzon DJ, Marchese M, Klein W, Ginsberg SD, Francis BM, Mount HT, Mufson EJ, Salehi A, Fahnestock M. Decreased brain-derived neurotrophic factor depends on amyloid aggregation state in transgenic mouse models of Alzheimer’s disease. J. Neurosci., 29, 9321–9329 (2009).
• 20) Sato A, Satake A, Hiramoto A, Wataya Y, Kim HS. Protein expression profiles of necrosis and apoptosis induced by 5-fluoro-2′-deoxyuridine in mouse cancer cells. J. Proteome Res., 9, 2329–2338 (2010).
• 21) Sato A, Nakama K, Watanabe H, Satake A, Yamamoto A, Omi T, Hiramoto A, Masutani M, Wataya Y, Kim HS. Role of activating transcription factor 3 protein ATF3 in necrosis and apoptosis induced by 5-fluoro-2′-deoxyuridine. FEBS J., 281, 1892–1900 (2014).
• 22) Encinas M, Iglesias M, Liu Y, Wang H, Muhaisen A, Ceña V, Gallego C, Comella JX. Sequential treatment of SH-SY5Y cells with retinoic acid and brain-derived neurotrophic factor gives rise to fully differentiated, neurotrophic factor-dependent, human neuron-like cells. J. Neurochem., 75, 991–1003 (2000).
• 23) Agholme L, Lindstrom T, Kagedal K, Marcusson J, Hallbeck M. An in vitro model for neuroscience: differentiation of SH-SY5Y cells into cells with morphological and biochemical characteristics of mature neurons. J. Alzheimers Dis., 20, 1069–1082 (2010).
• 24) Krishtal J, Bragina O, Metsla K, Palumaa P, Tõugu V. In situ fibrillizing amyloid-beta 1–42 induces neurite degeneration and apoptosis of differentiated SH-SY5Y cells. PLOS ONE, 12, e0186636 (2017).
• 25) LeBlanc A, Liu H, Goodyer C, Bergeron C, Hammond J. Caspase-6 role in apoptosis of human neurons, amyloidogenesis, and Alzheimer’s disease. J. Biol. Chem., 274, 23426–23436 (1999).
• 26) Wang XJ, Cao Q, Zhang Y, Su XD. Activation and regulation of caspase-6 and its role in neurodegenerative diseases. Annu. Rev. Pharmacol. Toxicol., 55, 553–572 (2015).
• 27) Degterev A, Huang Z, Boyce M, Li Y, Jagtap P, Mizushima N, Cuny GD, Mitchison TJ, Moskowitz MA, Yuan J. Chemical inhibitor of nonapoptotic cell death with therapeutic potential for ischemic brain injury. Nat. Chem. Biol., 1, 112–119 (2005).
• 28) Yang SH, Lee DK, Shin J, Lee S, Baek S, Kim J, Jung H, Hah JM, Kim Y. Nec-1 alleviates cognitive impairment with reduction of Abeta and tau abnormalities in APP/PS1 mice. EMBO Mol. Med., 9, 61–77 (2017).
• 29) Zhang W, Linden DJ. The other side of the engram: experience-driven changes in neuronal intrinsic excitability. Nat. Rev. Neurosci., 4, 885–900 (2003).
• 30) Kim DY, Abdelwahab MG, Lee SH, O’Neill D, Thompson RJ, Duff HJ, Sullivan PG, Rho JM. Ketones prevent oxidative impairment of hippocampal synaptic integrity through KATP channels. PLoS ONE, 10, e0119316 (2015).
• 31) Iwata N, Sekiguchi M, Hattori Y, Takahashi A, Asai M, Ji B, Higuchi M, Staufenbiel M, Muramatsu S, Saido TC. Global brain delivery of neprilysin gene by intravascular administration of AAV vector in mice. Sci. Rep., 3, 1472 (2013).
• 32) Brendza RP, Bacskai BJ, Cirrito JR, Simmons KA, Skoch JM, Klunk WE, Mathis CA, Bales KR, Paul SM, Hyman BT, Holtzman DM. Anti-Abeta antibody treatment promotes the rapid recovery of amyloid-associated neuritic dystrophy in PDAPP transgenic mice. J. Clin. Invest., 115, 428–433 (2005).
• 33) Arancibia S, Silhol M, Moulière F, Meffre J, Höllinger I, Maurice T, Tapia-Arancibia L. Protective effect of BDNF against beta-amyloid induced neurotoxicity in vitro and in vivo in rats. Neurobiol. Dis., 31, 316–326 (2008).
• 34) Takahashi N, Duprez L, Grootjans S, Cauwels A, Nerinckx W, DuHadaway JB, Goossens V, Roelandt R, Van Hauwermeiren F, Libert C, Declercq W, Callewaert N, Prendergast GC, Degterev A, Yuan J, Vandenabeele P. Necrostatin-1 analogues: critical issues on the specificity, activity and in vivo use in experimental disease models. Cell Death Dis., 3, e437 (2012).
• 35) Vandenabeele P, Grootjans S, Callewaert N, Takahashi N. Necrostatin-1 blocks both RIPK1 and IDO: consequences for the study of cell death in experimental disease models. Cell Death Differ., 20, 185–187 (2013).
• 36) Souza LC, Jesse CR, Antunes MS, Ruff JR, de Oliveira Espinosa D, Gomes NS, Donato F, Giacomeli R, Boeira SP. Indoleamine-2,3-dioxygenase mediates neurobehavioral alterations induced by an intracerebroventricular injection of amyloid-beta 1–42 peptide in mice. Brain Behav. Immun., 56, 363–377 (2016).
• 37) Kincses ZT, Toldi J, Vecsei L. Kynurenines, neurodegeneration and Alzheimer’s disease. J. Cell. Mol. Med., 14, 2045–2054 (2010).