CDK5 Participates in Amyloid-β Production by Regulating PPARγ Phosphorylation in Primary Rat Hippocampal Neurons

Abstract

Cyclin-dependent kinase 5 (CDK5) in adipose tissue mediates peroxisome proliferator-activated receptor γ (PPARγ) phosphorylation at Ser273 to inhibit its activity, causing PPARγ target gene expression changes. Among these, insulin-degrading enzyme (IDE) degrades amyloid-β peptide (Aβ), the core pathological product of Alzheimer’s disease (AD), whereas β-amyloid cleavage enzyme 1 (BACE1) hydrolyzes amyloid-β protein precursor (AβPP). Therefore, we speculated that CDK5 activity in the brain might participate in Aβ production, thereby functioning as a key molecule in AD pathogenesis. To confirm this hypothesis, we transduced primary rat hippocampal neurons using CDK5-expressing lentiviral vectors. CDK5 overexpression increased PPARγ Ser273 phosphorylation, decreased IDE expression, increased BACE1 and AβPP expression, increased Aβ levels, and induced neuronal apoptosis. The CDK5 inhibitor roscovitine effectively reversed these CDK5 overexpression-mediated effects. Moreover, silencing of the Cdk5 gene via CDK5 shRNA-expressing lentiviral vectors in primary hippocampal neurons did not exert any protective effect against normal neuronal apoptosis, nor were significant effects observed on Aβ levels, PPARγ phosphorylation, or PPARγ target gene expression in the cells. However, Cdk5 gene silencing exhibited a neuroprotective effect in the Aβ-induced AD neuron model by effectively inhibiting the Aβ-induced neuronal apoptosis, PPARγ phosphorylation, PPARγ expression downregulation, and PPARγ target gene expression changes, and reducing Aβ levels. In conclusion, this study demonstrated that CDK5 played an important role in the pathogenesis of AD. Specifically, CDK5 participated in Aβ production by regulating PPARγ phosphorylation. Targeted therapy against CDK5 could effectively reduce and reverse the neurotoxic effects of Aβ and may represent a novel approach for AD treatment.

Keywords

Alzheimer’s disease amyloid cyclin-dependent kinase 5 peroxisome proliferator-activated receptors phosphorylation

INTRODUCTION

Alzheimer’s disease (AD) constitutes a degenerative disease of the central nervous system. The initial symptom is often memory loss, followed by continuous deterioration of cognitive function that may be accompanied by a series of mental or neurological symptoms and behavioral disorders. In addition, the ability to perform daily activities is severely reduced; thus, this disorder exerts significant negative impact on the quality of life and lifespan of patients [1]. However, although various theories regarding the pathogenesis of AD have been proposed, no consensus has been reached.

The excessive accumulation of amyloid-β (Aβ) in the brain has been shown to serve as a critical pathological change in AD [2], which can lead to neuronal apoptosis [3], neurofibrillary tangles [4], synaptic dysfunction [5], inflammation [6], and oxidative stress [7]. Levels of Aβ are, in part, regulated by the ligand-activated nuclear transcription factor peroxisome proliferator-activated receptor γ (PPARγ). Upon activation by its ligand, PPARγ binds to the peroxisome proliferator responsive element (PPRE) in the promoter of its target genes to regulate their expression [5]. In particular, PPARγ can bind to PPRE in the promoter of the insulin-degrading enzyme (IDE) gene to regulate its transcription and promote expression of the IDE protein [8], which can degrade Aβ [9]. In contrast, PPARγ binding to PPRE in the β-amyloid cleavage enzyme 1 (BACE1) promoter inhibits BACE1 expression [10]. BACE1 constitutes a key enzyme for the initiation of Aβ production and can hydrolyze the amyloid-β protein precursor (AβPP) to produce Aβ [11].

Phosphorylation serves as an important posttranslational modification of PPARγ that can lead to expression changes and dysfunction in a series of downstream target genes. It has been demonstrated in adipose tissue that cyclin-dependent kinase 5 (CDK5) can mediate the phosphorylation of PPARγ at Ser273 to inhibit its activity, leading to dysfunction and expression changes of PPARγ target genes [12]. Therefore, we speculated that CDK5 may constitute a key molecule in the pathogenesis of AD. In the brain, CDK5 may mediate the phosphorylation of PPARγ and regulate the expressions of IDE and BACE1, resulting in decreased Aβ clearance and increased Aβ production. The excessive accumulation of Aβ in the brain would in turn lead to a series of neuronal toxicities. The aim of this study was to confirm our hypothesis using lentiviral vector-mediated RNA interference and overexpression of CDK5 in primary rat hippocampal neurons.

MATERIALS AND METHODS

Ethics statement

All experimental procedures in this study were approved by the Ethics Committee of Xi’an Jiaotong University Health Science Center (No. 2018-2025).

Reagents

Aβ_1–42 and roscovitine were purchased from Sigma, St. Louis, MO, USA. Polyclonal rabbit anti-rat antibodies against IDE (ab133561), BACE1 (ab2077), and AβPP (ab15272) were purchased from Abcam, Cambridge, UK. Polyclonal rabbit anti-rat antibodies against CDK5 (WL01673), PPARγ (WL0269), and Aβ_1–42 (WL01427) were purchased from Wanleibio, Shenyang, China. The polyclonal rabbit anti-rat antibody against phosphorylated PPARγ-Ser273 (p-PPARγ-Ser273) (bs-4888R) was purchased from Bioss Antibodies, Beijing, China. Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay kit, total protein extraction kit, BCA protein concentration assay kit, β-actin antibody, and ECL chemiluminescence reagent were purchased from Wanleibio. The rat Aβ_1–42 enzyme-linked immunosorbent assay (ELISA) Kit was purchased from Milbio, Shanghai, China. Dulbecco’s modified Eagle medium (DMEM) was purchased from Gibco, Gaithersburg, MD, USA. Fetal bovine serum was purchased from Biological Industries, Cromwell, CT, USA. Cy3-labeled goat anti-rabbit IgG, trypsin, and DAPI were purchased from Beyotime, Shanghai, China. Multifunctional DNA purification and recovery kit, TRIzol, 2× Power Taq PCR MasterMix reaction amplification kit, plasmid maxiprep kit, and pUM-T simple vector were purchased from BioTeke, Beijing, China. SYBR GREEN Master Mix, Escherichia coli JM109 competent cells, and E. coli DH5α competent cells were purchased from Solarbio, Beijing, China.

Isolation and culture of rat hippocampal neurons

According to previous report [13], Sprague-Dawley rats at 2 d after birth were selected. After decapitation, the hippocampus tissue was excised and digested with trypsin. Then, DMEM containing 10% fetal bovine serum and penicillin/streptomycin dual antibiotics were added to neutralize the digestive enzyme. After filtration and centrifugation, the hippocampal neurons were inoculated into a poly-L-lysine-coated 6-well culture plate at a density of 5×10⁵/ml. After incubation for 8 h at 37°C with 5% CO₂ under saturated humidity, the medium was replaced with Neurobasal Medium containing 2% B27, 0.5 mmol/l glutamine, 100 U/ml penicillin, and 100 U/ml streptomycin. After 48 h, 10μmol/l cytarabine was added to inhibit glial cell growth; medium was changed every 3 d. After 15 d culture, the cells were used subsequent experiments.

CDK5 overexpression

The rat Cdk5 cDNA was provided by Shenyang Wanleibio Co., Ltd. and subjected to polymerase chain reaction (PCR) amplification using Cdk5 forward primer 5^′CGCGGATCCATGCAGAAATACGAGAAACT3^′, and reverse primer 5^′CAAGAATTCCTACGGG GGACAGAAGTCAG^′, 897 bp. The PCR products were purified and the target genes were obtained using a multifunctional DNA purification and recovery kit. Then, TA cloning was performed to ligate the recovered target genes to the pUM-T simple vector. After ligation, the ligated recombinant products were added to the E. coli JM109 competent cells for transformation. Then, positive clones with target genes inserted were selected using PCR screening, sequenced, and validated, after which the target plasmids were extracted using a plasmid maxiprep kit. Next, genic recombination was carried out. The target genes in the target plasmids and the pUM-T simple vector were subjected to double digestion with restriction enzymes. The resulting digested fragments were recovered, respectively, and used for ligation and recombination. Then, the ligated recombinant products were added to E. coli JM109 competent cells for transformation. After the sequences were validated, the fragments were packaged into the lentiviral vector expressing CDK5 (designated pUM-T-CDK5). The pUM-T simple vector without DNA fragment insertion was selected for use as a negative control. Rat hippocampal neurons that had been properly cultured were divided into four groups: control, pUM-T simple, pUM-T-CDK5, and pUM-T-CDK5 + roscovitine groups. The control group did not receive any intervention. pUM-T-CDK5 was used to infect the primary hippocampal neurons of the CDK5 overexpression group and the CDK5 overexpression + roscovitine group. The neurons of the pUM-T-simple group were infected with pUM-T simple. After 44 h of infection, the CDK5 overexpression + roscovitine group was treated with the CDK5 inhibitor roscovitine at a final concentration of 25μM [14] for 4 h. After 48 h of total infection time, the cells were collected for experiments.

CDK5 silencing

The target sequence for RNA interference was designed based on the sequence of rat Cdk5 (GenBank: NM_080885.1). A sequence without interference effect was selected as the control. The sequences were used to synthesize oligonucleotide primers for CDK5 shRNA and scramble shRNA. The primer sequences were as follows: CDK5 shRNA forward primer 5^′ccggcccGGAGAGACCTGTTGCAGAAttcaagaga TTCTGCAACAGGTCTCTCCttttt3^′, reverse primer 5^′aattaaaaaGGAGAGACCTGTTGCAGAAtctcttgaa TTCTGCAACAGGTCTCTCCggg3^′; scramble shRNA forward primer 5^′ccggcccTTCTCCGAA CGTGTCACGTttcaagagaACGT GACACGTTCGGAGAAttttt 3^′, reverse primer 5^′aattaaaaaTTC TCCGAACGTGTCACGTtctcttgaa ACGTGACACGTTCGGAG AAggg3^′. The corresponding double-stranded DNA was obtained after annealing of the above primers. The resulting double-stranded DNA was ligated to the enzyme-digested lentiviral vector. Then, the recombinant vectors were added into E. coli DH5α competent cells for transformation. PCR reactions were used to obtain positive clones with target sequences inserted. After the positive clones were sequenced and validated, the target plasmids were extracted, and lentiviral packaging was performed to construct the lentiviral vector expressing CDK5 shRNA and the negative control vector, termed LV-CDK5-shRNA and LV-scramble-shRNA, respectively. Rat hippocampal neurons that had been properly cultured were divided into six groups: control, scramble, CDK5 shRNA, Aβ_1–42, scramble + Aβ_1–42, and CDK5 shRNA + Aβ_1–42 groups. The control group did not receive any intervention. LV-CDK5-shRNA was used to infect the primary hippocampal neurons of the CDK5 shRNA and CDK5 shRNA + Aβ_1–42 group. The neurons of the scramble and scramble + Aβ_1–42 group were infected with LV-scramble-shRNA. After 24 h of culture, the Aβ_1–42, scramble+Aβ_1–42 and CDK5 shRNA+Aβ_1–42 groups were treated with Aβ_1–42 at a final concentration of 8μM [15] for 24 h. After 48 h of total treatment time, the cells were collected for experiments.

TUNEL staining

TUNEL staining was performed according to the manufacturer’s instructions (Wanleibio). Once the cells had grown onto the slides, TUNEL reaction solution was added in the dark at 37°C for 60 min. Slides were then rinsed with phosphate buffered saline and counterstained with DAPI for 5 min in the dark. After the slides had been sealed, neurons were counted by two pathologists blinded to the grouping under a×400-fold fluorescence microscope (BX53, Olympus, Japan). The average count of four random fields of view was taken as the final result. The percent of TUNEL-positive neurons was calculated. All neuronal nuclei could be stained by DAPI and produced blue fluorescence. Apoptotic neuronal nuclei were stained with TUNEL and produced green fluorescence.

Immunofluorescence staining

After fixation using 4% paraformaldehyde, the slides on which the cells had grown were permeabilized using 0.1% Triton X–100 and subsequently being blocked using lowlenthal serum. Then polyclonal rabbit anti-rat antibody against Aβ_1–42 was added (concentration: 1:200) and the slices were incubated overnight at 4°C. After washing with PBS, Cy3-labeled goat anti-rabbit IgG (concentration 1:200) was added and the slices were incubated in the dark at 25°C for 1 h. After being washed with PBS, the nuclei of the neurons were stained using DAPI in the dark as a counterstain. Finally, a fluorescence quencher was added and slides were sealed by coverslips. Observation (×400) was performed under a fluorescence microscope (BX53, Olympus, Japan).

Flow cytometry

After centrifuging each group of cells at 300×g for 5 min and washing with phosphate buffered saline, the cells were resuspended by adding 500μl of binding buffer. Then, 5μl of Annexin V-FITC and 10μl of propidium iodide were added, and the cells were incubated at 37°C for 15 min in the dark. The cells were then collected and rinsed, whereupon apoptosis was measured using a flow cytometer (Accuri C6, Becton Dickinson, Bedford, MA, USA). Apoptosis percent = [Annexin V (+) propidium iodide (–) cell number+Annexin V (+) propidium iodide (+) cell number] / total cell number×100%.

ELISA

The culture medium in each group was collected and centrifuged at 3000×g, 4∘C for 10 min to remove cellular material. Then, the supernatant was used to quantify the extracellular Aβ_1–42 concentration using a rat Aβ_1–42 ELISA Kit according to the manufacturer’s protocol (Milbio).

Western blot

The total protein of each group of neurons was extracted according to the instructions of the total protein extraction kit (Wanleibio). The protein was denatured by heating at 95°C after concentration determination was performed using a BCA protein concentration assay kit. Thereafter, 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis was utilized to separate proteins. After separation, the proteins were transferred to a nitrocellulose membrane. After washing with Tris-buffered saline containing Tween-20 (TBST), 5% skim milk powder (diluted in TBST) was added to block the antibodies at 37°C for 1 h. Polyclonal rabbit anti-rat antibodies against CDK5, p-PPARγ-Ser273, PPARγ, IDE, BACE1, AβPP, and Aβ_1–42 were added (concentrations: 1:500, 1:400, 1:500, 1:1000, 1:500, 1:1000, 1:400), and the membrane was incubated overnight at 4°C. After washing with TBST, horse radish peroxidase-labeled secondary antibody (goat anti-rabbit, 1:5000) was added and the membrane was incubated at 37°C for 45 min. Then, the membrane was washed with TBST and developed with ECL chemiluminescence reagent. After scanning the film, the optical density values of the target bands were analyzed using Gel-Pro-Analyzer 4.0 software (Media Cybernetics, Rockville, MD, USA).

Real-time PCR

Total RNA was extracted from each group of neurons using TRIzol, and the concentration of RNA was determined by ultraviolet spectrophotometry. Reverse transcription reaction was performed to obtain cDNA according to the instructions (BioTeke). The GenBank database was searched for mRNA sequences of rat CDK5, PPARγ, IDE, BACE1, and AβPP. The corresponding primers were designed using Primer Premier 5.0 software (Premier Biosoft, Palo Alto, CA, USA) and synthesized by Sangon Biotech (Shanghai) Co., Ltd. (China) The sequences of primers are as follows: Cdk5 forward primer 5^′GGACACCGACTGAGGAAC3^′, reverse primer 5^′TTGGGCACGACATTCAC3^′, 103 bp; Pparg forward primer 5^′TACCACGGTTGATTTCTC 3^′, reverse primer 5^′AATAATAAGGCGGGGACG 3^′, 155 bp; Ide forward primer 5^′TCCCGTGAAGCGACTGT3^′, reverse primer 5^′GACTTGTCCGTGGTGGG 3^′, 180 bp; Bace1 forward primer 5^′TCCGCATCACCATCCTT 3^′, reverse primer 5^′TGACCGCTCCCATAACG 3^′, 123 bp; Aβpp forward primer 5^′ACTCTGTGCCAGCCAATA3^′, reverse primer 5^′TGAATCATGTCCGAACTCC3^′, 158 bp; and β-actin forward primer 5^′GGAGATTACTGCCCTGGCTCCTAGC3^′, and reverse primer 5^′GGCCGGACT CATCGTACTCCTGCTT 3^′, 155 bp. Real-time fluorescent quantitative PCR reaction was carried out with reference to the instructions of the 2×Power Taq PCR MasterMix reaction amplification kit (BioTeke). Reaction system: 1μl of cDNA, 0.5μl of forward primer (10μM), 0.5μl of reverse primer (10μM), and 10μl of SYBR GREEN Master Mix, supplemented to 20μl with ddH₂O. Reaction conditions: Pre-denatured at 94°C for 5 min, then denatured at 94°C for 10 s, annealed at 60°C for 20 s, and extended at 72°C for 30 s, for a total of 40 cycles. The above reaction was carried out on an Exicycler^TM 96 Real Time PCR system (Exicycler 96, Bioneer, Korea). The relative mRNA expression levels of the tested genes were expressed by 2^£-ΔΔCt values [16].

Data analyses

All data are expressed as the means±S.E.M, and statistical analysis was performed using SPSS 19.0 statistical software (IBM, Armonk, NY, USA). One-way analysis of variance was used for inter-group comparison. A post-hoc least significant difference (LSD)-t-test was used for pair-wise comparison between groups after the analysis of variance. p < 0.05 was considered as indicating statistically significant difference.

RESULTS

Effects of CDK5 overexpression and silencing on CDK5 expression in hippocampal neurons

The expression levels of both CDK5 protein and mRNA significantly increased after transduction of hippocampal neurons with lentiviral vectors expressing CDK5 (control group versus pUM-T-CDK5 group, p < 0.05). These effects could be effectively inhibited by the CDK5 inhibitor roscovitine (pUM-T-CDK5 group versus pUM-T-CDK5 + roscovitine group, p < 0.05), which confirmed the transduction efficiency of the CDK5-expressing lentiviral vectors (Fig. 1). Conversely, transduction of primary rat hippocampal neurons with lentiviral vectors expressing CDK5 shRNA effectively interfered with CDK5 expression. The expression levels of both CDK5 protein and mRNA significantly decreased compared with those of the control group (p < 0.05), which confirmed the knockdown efficiency of the CDK5 shRNA lentiviral vectors (Fig. 2).

Fig.1

Effects of CDK5 overexpression on CDK5 expression in hippocampal neurons. A) Western blot image showing the CDK5 band. B) Comparison of protein levels of CDK5 in different groups. C) Comparison of mRNA levels of CDK5 in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5, with or without the CDK5 inhibitor roscovitine. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus, the pUM-T-CDK5 group.

Fig.2

Effects of CDK5 silencing and exogenous Aβ_1–42 on CDK5 expression in hippocampal neurons. A) Western blot image showing the CDK5 band. B) Comparison of protein levels of CDK5 in different groups. C) Comparison of mRNA levels of CDK5 in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5 shRNA, with or without Aβ_1–42. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the Aβ_1–42 group.

In turn, we also found that treatment of neurons with Aβ_1–42 resulted in a significant increase in CDK5 expression (control group versus Aβ_1–42 group, p < 0.05). Conversely, CDK5 shRNA significantly inhibited the increase of CDK5 expression induced by Aβ_1–42 (Aβ_1–42 group versus CDK5 shRNA+Aβ_1–42 group, p < 0.05 (Fig. 2).

Effects of CDK5 overexpression and silencing on apoptosis of hippocampal neurons

The TUNEL-positive percent from TUNEL staining and the apoptotic percent from flow cytometry of normal hippocampal neurons significantly increased after transduction with lentiviral vectors expressing CDK5 (control group versus pUM-T-CDK5 group, p < 0.05), whereas the addition of the CDK5 inhibitor roscovitine effectively inhibited the neuronal apoptosis induced by CDK5 overexpression (pUM-T-CDK5 group versus pUM-T-CDK5 + roscovitine group, p < 0.05), as shown in Figs. 3 and 4. To further investigate the role of CDK5 in the pathogenesis of AD, we also performed experiments to evaluate the effects of CDK5 silencing. The results from TUNEL staining showed that the TUNEL-positive percent of normal primary rat hippocampal neurons transduced with lentiviral vectors expressing CDK5 shRNA did not change significantly (control group versus CDK5 shRNA group, p > 0.05). However, the results from flow cytometry showed that transduction with lentiviral vectors expressing CDK5 shRNA reduced the apoptotic percent of normal primary rat hippocampal neurons (control group versus CDK5 shRNA group, p < 0.05). Treatment of hippocampal neurons with Aβ_1–42 significantly increased neuronal apoptosis (as shown by TUNEL staining and flow cytometry; control group versus Aβ_1–42 group, p < 0.05 for both), whereas transduction with lentiviral vectors expressing CDK5 shRNA significantly inhibited the Aβ_1–42-induced neuronal apoptosis. Both TUNEL-positive percent of the neurons evaluated by TUNEL staining and neuronal apoptotic percent evaluated by flow cytometry were significantly reduced (Aβ_1–42 group versus CDK5 shRNA+Aβ_1–42 group, p < 0.05), as shown in Figs. 5 and 6.

Fig.3

TUNEL staining of hippocampal neurons after overexpression of CDK5. A) TUNEL staining. Green represents nuclei with positive TUNEL staining and blue represents nuclei counterstained with DAPI; Bar = 50μm. B) Comparison of neuronal TUNEL-positive percentages in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5, with or without the CDK5 inhibitor roscovitine. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the pUM-T-CDK5 group.

Fig.4

Flow cytometry results after overexpression of CDK5. A) Flow cytometry results. B) Comparison of neuronal apoptotic percentages in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5, with or without the CDK5 inhibitor roscovitine. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the pUM-T-CDK5 group.

Fig.5

TUNEL staining of hippocampal neurons after silencing of CDK5. A) TUNEL staining. Green represents nuclei with positive TUNEL staining and blue represents nuclei counterstained with DAPI; Bar = 50μm. B) Comparison of neuronal TUNEL-positive percentages in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5 shRNA, with or without Aβ_1–42. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the Aβ_1–42 group.

Fig.6

Flow cytometry results after silencing of CDK5. A) Flow cytometry results. B) Comparison of neuronal apoptotic percentages in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5 shRNA, with or without Aβ_1–42. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the Aβ_1–42 group.

Effects of CDK5 overexpression and silencing on Aβ levels in hippocampal neurons

Both intracellular and extracellular Aβ can comprise a critical pathological product in the pathogenesis of AD that can lead to neuronal apoptosis [17, 18], with Aβ_1–42 being one of the main components of Aβ [19]. Therefore, we further tested the effects of overexpression and silencing of CDK5 on Aβ_1–42 levels in hippocampal neurons. The levels of intracellular Aβ_1–42 in the hippocampal neurons significantly increased after transduction with lentiviral vectors expressing CDK5. Immunofluorescence staining showed that Aβ_1–42 positive neurons in hippocampal neurons in the pUM-T-CDK5 group were significantly increased compared with those in the control group, with Aβ _1–42 being mainly expressed in the cytoplasm and cell membrane, as shown in Fig. 7A. In addition, no Aβ_1–42 positive staining was observed outside the neurons, which indicated that there was no residual exogenous added Aβ_1–42 in the detected neurons. Western blot quantitative analysis further confirmed that the level of intracellular Aβ_1–42 in hippocampal neurons in the pUM-T-CDK5 group was significantly higher than that in the control group (p < 0.05) (Fig. 7B, C). Next, we detected the concentration of extracellular Aβ_1–42 in the culture medium by ELISA. The results showed that the concentration of extracellular Aβ_1–42 in the pUM-T-CDK5 group was significantly higher than that in the control group (p < 0.05) (Fig. 7D). The CDK5 inhibitor roscovitine effectively inhibited the CDK5 overexpression-mediated increase in both intracellular and extracellular Aβ_1–42 levels (pUM-T-CDK5 group versus pUM-T-CDK5 + roscovitine group, p < 0.05 for both), which confirmed the effects of increased CDK5 expression on Aβ_1–42 levels. Notably, transduction with lentiviral vectors expressing CDK5 shRNA had no significant effects on either the endogenous or extracellular Aβ_1–42 levels in normal primary hippocampal neurons (as shown by western blot and ELISA, control group versus CDK5 shRNA group, p > 0.05 for each) but significantly reduced the elevated intracellular and extracellular Aβ_1–42 levels caused by exogenous Aβ_1–42 introduced to the neurons (Aβ_1–42 group versus CDK5 shRNA+Aβ_1–42 group, p < 0.05; Fig. 8C, D).

Fig.7

Effects of CDK5 overexpression on Aβ levels in hippocampal neurons. A) Immunofluorescence staining of Aβ_1–42. Red represents Aβ _1–42 positive staining and blue represents nuclei counterstained with DAPI; Bar = 50μm. B) Western blot image of the intracellular Aβ_1–42 band. C) Comparison of protein levels of intracellular Aβ_1–42 in different groups detected by western blot. D) Comparison of protein levels of extracellular Aβ_1–42 in different groups detected by ELISA. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5, with or without the CDK5 inhibitor roscovitine. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the pUM-T-CDK5 group.

Fig.8

Effects of CDK5 silencing on Aβ levels in hippocampal neurons. A) Immunofluorescence staining of Aβ_1–42. Red represents Aβ_1–42 positive staining and blue represents nuclei counterstained with DAPI; Bar = 50μm. B) Western blot image of the intracellular Aβ_1–42 band. C) Comparison of protein levels of intracellular Aβ_1–42 in different groups detected by western blot. D) Comparison of protein levels of extracellular Aβ_1–42 in different groups detected by ELISA. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5 shRNA, with or without Aβ_1–42. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the Aβ_1–42 group.

Effects of CDK5 overexpression and silencing on PPARγ phosphorylation in hippocampal neurons

Similar to its activity in adipose tissue [12], this study demonstrated that CDK5 could also mediate the phosphorylation of PPARγ in neurons. Transduction of hippocampal neurons with lentiviral vectors expressing CDK5 increased the phosphorylation of PPARγ at Ser273 (control group versus pUM-T-CDK5 group, p < 0.05), which increased the p-PPARγ/PPARγ ratio (control group versus pUM-T-CDK5 group, p < 0.05). The expression levels of PPARγ protein and mRNA were both reduced (control group versus pUM-T-CDK5 group, p < 0.05). Treatment with the CDK5 inhibitor roscovitine after transduction with lentiviral vectors expressing CDK5 reversed the above effects, as shown in Fig. 9. In contrast, silencing of CDK5 by transduction lentiviral vectors expressing CDK5 shRNA merely upregulated the expression levels of PPARγ protein in the primary rat hippocampal neurons (control group versus CDK5 shRNA group, p < 0.05) (Fig. 10D) but had no significant effect on the phosphorylation of PPARγ at Ser273. This was exemplified by the nonsignificant differences in the ratio of p-PPARγ/PPARγ and the expression of p-PPARγ between the control and CDK5 shRNA groups (p > 0.05), as shown in Fig. 10B and C.

Fig.9

Effects of CDK5 overexpression on phosphorylation of PPARγ at Ser273 in hippocampal neurons. A) Western blot image showing the p-PPARγ and PPARγ bands. B) Comparison of p-PPARγ/PPARγ ratios in different groups. C) Comparison of protein levels of p-PPARγ in different groups. D, E) Comparisons of protein levels and mRNA levels of PPARγ in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5, with or without the CDK5 inhibitor roscovitine. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the pUM-T-CDK5 group.

Fig.10

Effects of CDK5 silencing and exogenous Aβ_1–42 on phosphorylation of PPARγ at Ser273 in hippocampal neurons. A) Western blot image showing the p-PPARγ and PPARγ bands. B) Comparison of p-PPARγ/PPARγ ratios in different groups. C) Comparison of protein levels of p-PPARγ in different groups. D, E) Comparisons of protein levels and mRNA levels of PPARγ in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5 shRNA, with or without Aβ_1–42. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the Aβ_1–42 group.

Treatment of primary hippocampal neurons with Aβ_1–42 could increase the phosphorylation of PPARγ at Ser273, increase the ratio of p-PPARγ/PPARγ, and decrease the expression of PPARγ protein and mRNA (control group versus Aβ_1–42 group, p < 0.05). However, transduction of primary rat hippocampal neurons with lentiviral vectors expressing CDK5 shRNA prior to Aβ_1–42 treatment had a significant inhibitory effect on Aβ_1–42-induced PPARγ phosphorylation and downregulation of PPARγ expression (Aβ_1–42 group versus CDK5 shRNA+Aβ_1–42 group, p < 0.05), as shown in Fig. 10.

Effects of CDK5 overexpression and silencing on the expression of PPARγ target genes in hippocampal neurons

We next evaluated the effects of CDK5 expression on expression of the PPARγ target genes Ide and Bace1. Transduction of hippocampal neurons with lentiviral vectors expressing CDK5 reduced the protein and mRNA expression levels of IDE (control group versus pUM-T-CDK5 group, p < 0.05; Fig. 11A-C) and upregulated those of BACE1 and AβPP (control group versus pUM-T-CDK5 group, p < 0.05; Fig. 11D-I). Treatment with the CDK5 inhibitor roscovitine reversed the above effects. CDK5 silencing following transduction of primary rat hippocampal neurons with lentiviral vectors expressing CDK5 shRNA yielded no significant changes in the protein and mRNA levels of IDE, BACE1, or AβPP (control group versus CDK5 shRNA group, p > 0.05). Treatment of hippocampal neurons with Aβ_1–42 downregulated the protein and mRNA levels of IDE compared with those in control neurons (p < 0.05), whereas the protein and mRNA levels of BACE1 and AβPP were upregulated (p < 0.05). Transduction of the neurons with lentiviral vectors expressing CDK5 shRNA prior to Aβ_1–42 treatment effectively prevented the Aβ_1–42-induced effects, resulting in upregulated IDE protein and mRNA levels and downregulated BACE1 and AβPP protein and mRNA levels (Aβ_1–42 group versus CDK5 shRNA+Aβ_1–42 group, p < 0.05) (Fig. 12).

Fig.11

Effects of CDK5 overexpression on the expression of PPARγ target genes in hippocampal neurons. A, D, G) Western blot images showing the IDE, BACE1, and AβPP bands. B, E, H) Comparisons of protein levels of IDE, BACE1, and AβPP in different groups. C, F, I) Comparisons of mRNA levels of IDE, BACE1, and AβPP in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5, with or without the CDK5 inhibitor roscovitine. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the pUM-T-CDK5 group.

Fig.12

Effects of CDK5 silencing and exogenous Aβ_1–42 on the expressions of PPARγ target genes in hippocampal neurons. A, D, G) Western blot images showing the IDE, BACE1, and AβPP bands. B, E, H) Comparisons of protein levels of IDE, BACE1, and AβPP in different groups. C, F, I) Comparisons of mRNA levels of IDE, BACE1, and AβPP in different groups. Hippocampal neurons were transduced with lentiviral vectors expressing control or CDK5 shRNA, with or without Aβ_1–42. n = 6. *p < 0.05 versus the control group; ^#p < 0.05 versus the Aβ_1–42 group.

DISCUSSION

This study demonstrated that in primary rat hippocampal neurons, CDK5 could regulate the expression of IDE and BACE1 by mediating the phosphorylation of PPARγ, resulting in decreased Aβ clearance and increased Aβ production. CDK5 constitutes a member of the cyclin-dependent protein kinase family that functions as a serine/threonine protein kinase. Under physiological conditions, CDK5 can phosphorylate certain serine and threonine sites in proteins to play an important physiological role in neurodevelopment and neurological functions [20 –23]. CDK5 has also been shown to be closely associated with the development of AD. In particular, CDK5 is activated upon binding to P25, a truncated form of P35. The activated CDK5 can lead to highly phosphorylated tau, neurofilaments, and other cytoskeletal proteins, thereby forming neurofibrillary tangles and phosphorylated protein aggregates, which are hallmarks of certain neurodegenerative diseases such as AD and Parkinson’s disease [24 –26]. Presenilin-1 (PS1), the mutation of which represents a major factor involved in familial AD with autosomal dominant inheritance, is also phosphorylated (at Thr354) upon CDK5 activation, along with upregulated PS1 expression [27]. CDK5 can also mediate the phosphorylation of Drp1 at Ser 579, which in turn regulates the mitochondrial division and neuronal toxicity induced by Aβ_1–42 [28]. In addition, CDK5 activation promotes the production of the AD core pathological product, Aβ by phosphorylating AβPP at Thr668 [29, 30].

This study demonstrated another mechanism by which CDK5 could promote Aβ production. Specifically, CDK5 could mediate the phosphorylation of PPARγ at Ser273, thereby affecting the expression of PPARγ target genes Ide and Bace1, reducing Aβ clearance and promoting Aβ production. It has been demonstrated that both intracellular and extracellular Aβ play an important role in the pathogenesis of AD [31]. Intracellular Aβ mainly exists in endosomes, the endoplasmic reticulum, and the Golgi apparatus [32]. Intracellular Aβ originates from AβPP proteolysis mediated by BACE1 in the above organelles [33 –35]. Extracellular Aβ derives from γ-secretase-mediated proteolytic cleavage of the of 99 amino acid AβPP C-terminal fragment, which is produced by cleavage of transmembrane AβPP by BACE1 and is released into the extracellular matrix [36]. IDE, a zinc metallo-eptidase, is predominant in the cytoplasm but is also present in the cell surface, eroxisomes, endosomes, and other parts of the cell [37, 38]. The results show that IDE in the cytoplasm of neurons can degrade the intracellular Aβ [39], whereas the enzymes attached to the neuronal membrane and those released from microglia can degrade the extracellular Aβ [40]. The present study confirmed that transduction of primary rat hippocampal neurons with lentiviral vectors expressing CDK5 could increase PPARγ Ser273 phosphorylation, decrease IDE expression, increase BACE1 and AβPP expression, increase intracellular and extracellular Aβ levels, and induce neuronal apoptosis. Conversely, the CDK5 inhibitor roscovitine effectively reversed these CDK5 overexpression-mediated effects. This suggested that CDK5 is involved in Aβ production by regulating PPARγ phosphorylation. In addition, our results demonstrated that exogenous added Aβ could in turn promote the increase of CDK5 expression, aggravate PPARγ phosphorylation, and affect the expression of the PPARγ downstream target genes. These findings indicated that in the pathogenesis of AD, the relationship between CDK5 and Aβ may be reciprocal with each mutually influencing the other.

Notably, we also observed that CDK5 overexpression led to the increase of AβPP expression. A study confirmed Aβ can cause the accumulation of AβPP in neurons by inducing mouse primary neurons to become swollen and dystrophic [41]. Therefore, we speculated that the increase of AβPP expression induced by CDK5 overexpression in this study may be due to the increase of Aβ level induced by PPARγ phosphorylation, which induced the accumulation of AβPP in neurons. This hypothesis is also supported by the other results of the present study; i.e., exogenous Aβ_1–42 treatment of primary cultured hippocampal neurons led to increased AβPP expression. In addition, it has been shown that IDE can also degrade the AβPP intracellular domain [42]. Therefore, we further speculated that the increase of AβPP expression induced by CDK5 overexpression might also be due to the downregulation of IDE expression induced by PPARγ phosphorylation, thereby reducing the degradation of AβPP. These conjectures require experimental validation in future studies.

This study also confirmed that silencing of the Cdk5 gene had a significant protective effect in an Aβ-induced AD neuron model. CDK5 silencing effectively inhibited the neuronal apoptosis and increased CDK5 expression and PPARγ phosphorylation induced by Aβ. Further, it increased PPARγ expression, upregulated the expression of the PPARγ target gene Ide, downregulated BACE1 expression, promoted Aβ degradation, and reduced Aβ production, thus reducing Aβ levels. These results suggested that targeted therapy against CDK5 may serve as an effective approach for the treatment of AD. Therefore, CDK5 shRNA interference technology may have a promising prospect in the application for AD treatment, warranting further extensive research.

In addition, therapeutic drugs that inhibit CDK5 expression may also be beneficial for AD. For example, PPARγ agonists may inhibit PPARγ phosphorylation by suppressing the CDK5 pathway in the brain, as has been shown in adipose tissue [12]. This in turn would affect the expression of PPARγ target genes and decreases Aβ levels, thereby exerting a neuroprotective effect on AD. Notably, some studies have already adopted PPARγ agonists as interventions for AD. For example, pioglitazone can reduce Aβ levels in the brain of APP/PS1 transgenic mice and improve their learning and memory abilities [43, 44], as well as reduce tau phosphorylation and neuronal apoptosisin AD cellular models [45, 46], whereas rosiglitazone can reduce the levels of Aβ_1–40 and Aβ_1–42 and reduce tau phosphorylation in the cerebral cortex of AD transgenic mice, improving their memory impairment [47]. Rosiglitazone also prevents Aβ-induced oxidative stress [48], inflammatory response [49], and mitochondrial dysfunction [50]. Moreover, a clinical trial has confirmed that oral administration of rosiglitazone in patients with mild-to-moderate AD can improve their cognitive function [51]. However, whether these neuroprotective effects on AD exerted by PPARγ agonists act through the inhibition of the CDK5 pathway and PPARγ phosphorylation remains to be elucidated.

Moreover, the present study also found that silencing of the Cdk5 gene had minimal protective effect on the apoptosis of normal neurons and had no significant effects on intracellular and extracellular Aβ levels, PPARγ phosphorylation, or PPARγ target gene expression in the cells. Such limited effects may have occurred because under physiological conditions, the activation of CDK5 is crucial for neuronal development and maturation through various functions [29] including neuronal differentiation [52], neuronal migration [53, 54], gliogenesis [55], neurite growth [21], cerebellar development [23], motor coordination and cerebellar plasticity [22], and cerebral cortex development [56]. Therefore, in theory, silencing of the Cdk5 gene in normal neurons should have led to adverse effects, rather than neuroprotection. As neurodevelopmental and neurological changes were not specifically monitored in this study, the presence of such adverse outcomes is limited to theoretical reasoning and remains to be confirmed in future studies.

In conclusion, this study demonstrated that CDK5 plays an important role in the pathogenesis of AD and is involved in Aβ production by mediating the phosphorylation of PPARγ. Thus, targeted therapy against CDK5 may constitute a novel approach for the treatment of AD.

Footnotes

ACKNOWLEDGMENTS

We would like to thank Editage () for English language editing.

This work was supported by the China Postdoctoral Science Foundation [grant number 2017M623191], the Natural Science Foundation of Shaanxi Province [grant number 2017JQ8039], the Foundation of the Second Affiliated Hospital of Xi’an Jiaotong University [grant number YJ(ZD)201517], and the Natural Science Foundation of China [grant number 81571251].

Authors’ disclosures available online ().

References

Alzheimer’s Association (2018) 2018 Alzheimer’s disease facts and figures. Alzheimers Dement 14, 367–425.

Barage

, Sonawane

(2015) Amyloid cascade hypothesis: Pathogenesis and therapeutic strategies in Alzheimer’s disease. Neuropeptides 52, 1–18.

Islam

, Sharoar

, Ryu

, Park

(2017) Limited activation of the intrinsic apoptotic pathway plays a main role in amyloid-β-induced apoptosis without eliciting the activation of the extrinsic apoptotic pathway. Int J Mol Med 40, 1971–1982.

Bloom

(2014) Amyloid-β and tau: The trigger and bullet in Alzheimer disease pathogenesis. JAMA Neurol 71, 505–508.

Skaper

, Facci

, Zusso

, Giusti

(2017) Synaptic plasticity, dementia and Alzheimer disease. CNS Neurol Disord Drug Targets 16, 220–233.

Marottoli

, Katsumata

, Koster

, Thomas

, Fardo

, Tai

(2017) Peripheral inflammation, apolipoprotein E4, and amyloid-β interact to induce cognitive and cerebrovascular dysfunction. ASN Neuro 9, 1759091417719201.

Singh

, Bissoyi

, Kashyap

, Patra

, Rizvi

(2017) Autophagy activation alleviates amyloid-β-induced oxidative stress, apoptosis and neurotoxicity in human neuroblastoma SH-SY5Y cells. Neurotox Res 32, 351–361.

, Zhang

, Liu

, Zhang

, Huang

, Li

, Zhao

, Wang

(2009) PPARγ transcriptionally regulates the expression of insulin-degrading enzyme in primary neurons. Biochem Biophys Res Commun 383, 485–490.

, Kim

, Yoon

, Zhou

(2015) Common pesticide, dichlorodiphenyltrichloroethane (DDT), increases amyloid-β levels by impairing the function of ABCA1 and IDE: Implication for Alzheimer’s disease. J Alzheimers Dis 46, 109–122.

10.

Lin

, Chen

, Pan

, Zhu

, Zhang

, Shi

, Chen

(2016) Tripchlorolide attenuates β-amyloid generation via suppressing PPARγ-regulated BACE1 activity in N2a/APP695 cells. Mol Neurobiol 53, 6397–6406.

11.

Sadleir

, Eimer

, Cole

, Vassar

(2015) Aβ reduction in BACE1 heterozygous null 5XFAD mice is associated with transgenic APP level. Mol Neurodegener 10, 1.

12.

Choi

, Banks

, Estall

, Kajimura

, Boström

, Laznik

, Ruas

, Chalmers

, Kamenecka

, Blüher

, Griffin

, Spiegelman

(2010) Anti-diabetic drugs inhibit obesity-linked phosphorylation of PPARγ by cdk5. Nature 466, 451–456.

13.

Vadukul

, Gbajumo

, Marshall

, Serpell

(2017) Amyloidogenicity and toxicity of the reverse and scrambled variants of amyloid-β 1-42. FEBS Lett 591, 822–830.

14.

Manser

, Vagnoni

, Guillot

, Davies

, Miller

(2012) Cdk5/p35 phosphorylates lemur tyrosine kinase-2 to regulate protein phosphatase-1C phosphorylation and activity. J Neurochem 121, 343–348.

15.

Yang

, Ahn

, Ryu

, Choi

, Chong

, Hyun

, Chang

, Kim

(2015) Phloroglucinol attenuates the cognitive deficits of the 5XFAD mouse model of Alzheimer’s disease. PLoS One 10, e0135686.

16.

Mandrekar-Colucci

, Karlo

, Landreth

(2012) Mechanisms underlying the rapid peroxisome proliferator-activated receptor-γ-mediated amyloid clearance and reversal of cognitive deficits in a murine model of Alzheimer’s disease. J Neurosci 32, 10117–10128.

17.

Farajdokht

, Amani

, Mirzaei Bavil

, Alihemmati

, Mohaddes

, Babri

(2017) Troxerutin protects hippocampal neurons against amyloid beta-induced oxidative stress and apoptosis. EXCLI J 16, 1081–1089.

18.

Umeda

, Tomiyama

, Sakama

, Tanaka

, Lambert

, Klein

, Mori

(2011) Intraneuronal amyloid β oligomers cause cell death via endoplasmic reticulum stress, endosomal/lysosomal leakage, and mitochondrial dysfunction in vivo. J Neurosci Res 89, 1031–1042.

19.

Selkoe

(2001) Alzheimer’s disease: Genes, proteins, and therapy. Physiol Rev 81, 741–766.

20.

Zheng

, Zhang

, Fu

, Guo

, Shukla

, Amin

, E

, Bao

, Luo

, Li

, Lu

, Gao

(2016) Knockdown of expression of Cdk5 or p35 (a Cdk5 activator) results in podocyte apoptosis. PLoS One 11, e0160252.

21.

Duhr

, Déléris

, Raynaud

, Séveno

, Morisset-Lopez

, Mannoury la Cour

, Millan

, Bockaert

, Marin

, Chaumont-Dubel

(2014) Cdk5 induces constitutive activation of 5-HT6 receptors to promote neurite growth. Nat Chem Biol 10, 590–597.

22.

, Ishizeki

, Mita

, Wada

, Araki

, Ogura

, Abe

, Yamazaki

, Sakimura

, Mikoshiba

, Inoue

, Ohshima

(2014) Cdk5/p35 is required for motor coordination and cerebellar plasticity. J Neurochem 131, 53–64.

23.

Kumazawa

, Mita

, Hirasawa

, Adachi

, Suzuki

, Shafeghat

, Kulkarni

, Mikoshiba

, Inoue

, Ohshima

(2013) Cyclin-dependent kinase 5 is required for normal cerebellar development. Mol Cell Neurosci 52, 97–105.

24.

Kanungo

, Zheng

, Amin

, Pant

(2009) Targeting Cdk5 activity in neuronal degeneration and regeneration. Cell Mol Neurobiol 29, 1073–1080.

25.

Patrick

, Zukerberg

, Nikolic

, de la Monte

, Dikkes

, Tsai

(1999) Conversion of p35 to p25 deregulates Cdk5 activity and promotes neurodegeneration. Nature 402, 615–622.

26.

Posada-Duque

, López-Tobón

, Piedrahita

, González-Billault

, Cardona-Gomez

(2015) p35 and Rac1 underlie the neuroprotection and cognitive improvement induced by CDK5 silencing. J Neurochem 134, 354–370.

27.

Lau

, Howlett

, Kesavapany

, Standen

, Dingwall

, McLoughlin

, Miller

(2002) Cyclin-dependent kinase-5/p35 phosphorylates Presenilin 1 to regulate carboxy-terminal fragment stability. Mol Cell Neurosci 20, 13–20.

28.

Guo

, Shang

, Hu

, Jiang

, Wan

, Zhou

, Zhang

, Liao

, Yi

, Han

(2018) The role of Cdk5-mediated Drp1 phosphorylation in Aβ_1 - 42 induced mitochondrial fission and neuronal apoptosis. J Cell Biochem 119, 4815–4825.

29.

Liu

, Wang

, Jiang

, Tan

, Xing

, Yu

(2016) The role of Cdk5 in Alzheimer’s disease. Mol Neurobiol 53, 4328–4342.

30.

Liu

, Su

, Li

, Zhou

, Ryder

, Gonzalez-DeWhitt

, May

, Ni

(2003) Regulation of amyloid precursor protein (APP) phosphorylation and processing by p35/Cdk5 and p25/Cdk5. FEBS Lett 547, 193–196.

31.

LaFerla

, Green

, Oddo

(2007) Intracellular amyloid-β in Alzheimer’s disease. Nat Rev Neurosci 8, 499–509.

32.

Grant

, Ducatenzeiler

, Szyf

, Cuello

(2000) Aβ immunoreactive material is present in several intracellular compartments in transfected, neuronally differentiated, P19 cells expressing the human amyloid beta-protein precursor. J Alzheimers Dis 2, 207–222.

33.

Jang

, Park

, Lee

, Ko

, Kang

, Kim

(2017) Selenoprotein S is required for clearance of C99 through endoplasmic reticulum-associated degradation. Biochem Biophys Res Commun 486, 444–450.

34.

Saadipour

, Mañucat-Tan

, Lim

, Keating

, Smith

, Zhong

, Liao

, Bobrovskaya

, Wang

, Chao

, Zhou

(2018) p75 neurotrophin receptor interacts with and promotes BACE1 localization in endosomes aggravating amyloidogenesis. J Neurochem 144, 302–317.

35.

Toh

, Tan

, Zulkefli

, Houghton

, Gleeson

(2017) Amyloid precursor protein traffics from the Golgi directly to early endosomes in an Arl5b- and AP4-dependent pathway. Traffic 18, 159–175.

36.

Vassar

, Bennett

, Babu-Khan

, Kahn

, Mendiaz

, Denis

, Teplow

, Ross

, Amarante

, Loeloff

, Luo

, Fisher

, Fuller

, Edenson

, Lile

, Jarosinski

, Biere

, Curran

, Burgess

, Louis

, Collins

, Treanor

, Rogers

, Citron

(1999) β-secretase cleavage of Alzheimer’s amyloid precursor protein by the transmembrane aspartic protease BACE. Science 286, 735–741.

37.

Leissring

, Farris

, Wu

, Christodoulou

, Haigis

, Guarente

, Selkoe

(2004) Alternative translation initiation generates a novel isoform of insulin-degrading enzyme targeted to mitochondria. Biochem J 383, 439–446.

38.

Duckworth

, Bennett

, Hamel

(1998) Insulin degradation: Progress and potential. Endocr Rev 19, 608–624.

39.

Stargardt

, Gillis

, Kamphuis

, Wiemhoefer

, Kooijman

, Raspe

, Benckhuijsen

, Drijfhout

, Hol

, Reits

(2013) Reduced amyloid-β degradation in early Alzheimer’s disease but not in the APPswePS1dE9 and 3xTg-AD mouse models. Aging Cell 12, 499–507.

40.

Vekrellis

, Ye

, Qiu

, Walsh

, Hartley

, Chesneau

, Rosner

, Selkoe

(2000) Neurons regulate extracellular levels of amyloid β-protein via proteolysis by insulin-degrading enzyme. J Neurosci 20, 1657–1665.

41.

Sadleir

, Eimer

, Kaufman

, Osten

, Vassar

(2014) Genetic inhibition of phosphorylation of the translation initiation factor eIF2α does not block Aβ-dependent elevation of BACE1 and APP levels or reduce amyloid pathology in a mouse model of Alzheimer’s disease. PLoS One 9, e101643.

42.

Edbauer

, Willem

, Lammich

, Steiner

, Haass

(2002) Insulin-degrading enzyme rapidly removes the β-amyloid precursor protein intracellular domain (AICD). J Biol Chem 277, 13389–13393.

43.

Silva-Abreu

, Calpena

, Andrés-Benito

, Aso

, Romero

, Roig-Carles

, Gromnicova

, Espina

, Ferrer

, García

, Male

(2018) PPARγ agonist-loaded PLGA-PEG nanocarriers as a potential treatment for Alzheimer’s disease: in vitro and in vivo studies. Int J Nanomedicine 13, 5577–5590.

44.

Searcy

, Phelps

, Pancani

, Kadish

, Popovic

, Anderson

, Beckett

, Murphy

, Chen

, Blalock

, Landfield

, Porter

, Thibault

(2012) Long-term pioglitazone treatment improves learning and attenuates pathological markers in a mouse model of Alzheimer’s disease. J Alzheimers Dis 30, 943–961.

45.

Hamano

, Shirafuji

, Makino

, Yen

, Kanaan

, Ueno

, Suzuki

, Ikawa

, Matsunaga

, Yamamura

, Kuriyama

, Nakamoto

(2016) Pioglitazone prevents tau oligomerization. Biochem Biophys Res Commun 478, 1035–1042.

46.

Dehghani

, Meamar

, Askari

, Khorvash

, Shaygannejad

, Pour

, Javanmard

(2013) The effect of pioglitazone on the Alzheimer’s disease-induced apoptosis in human umbilical vein endothelial cells.S. Int J Prev Med 4, 205–210.

47.

Escribano

, Simón

, Gimeno

, Cuadrado-Tejedor

, López de Maturana

, García-Osta

, Ricobaraza

, Pérez-Mediavilla

, Del Río

, Frechilla

(2010) Rosiglitazone rescues memory impairment in Alzheimer’s transgenic mice: Mechanisms involving a reduced amyloid and tau pathology. Neuropsychopharmacology 35, 1593–1604.

48.

Chiang

, Nicol

, Cheng

, Lin

, Yen

, Lin

(2016) Rosiglitazone activation of PPARγ-dependent pathways is neuroprotective in human neural stem cells against amyloid-beta-induced mitochondrial dysfunction and oxidative stress. Neurobiol Aging 40, 181–190.

49.

Toledo

, Inestrosa

(2010) Activation of Wnt signaling by lithium and rosiglitazone reduced spatial memory impairment and neurodegeneration in brains of an APPswe/PSEN1DeltaE9 mouse model of Alzheimer’s disease.272-285. Mol Psychiatry 15, 228.

50.

Strum

, Shehee

, Virley

, Richardson

, Mattie

, Selley

, Ghosh

, Nock

, Saunders

, Roses

(2007) Rosiglitazone induces mitochondrial biogenesis in mouse brain. J Alzheimers Dis 11, 45–51.

51.

Risner

, Saunders

, Altman

, Ormandy

, Craft

, Foley

, Zvartau-Hind

, Hosford

, Roses

, Rosiglitazone in Alzheimer’s Disease Study Group (2006) Efficacy of rosiglitazone in a genetically defined population with mild-to-moderate Alzheimer’s disease. Pharmacogenomics J 6, 246–254.

52.

, Allen

, Rui

, Ku

, Liu

, Bankston

, Zheng

, Feng

(2016) p39 is responsible for increasing Cdk5 activity during postnatal neuron differentiation and governs neuronal network formation and epileptic responses. J Neurosci 36, 11283–11294.

53.

, Chen

, Shu

, Bu

, Shao

, Zhang

, Jiao

, Shi

, Tian

(2016) Phosphorylation of connexin 43 by Cdk5 modulates neuronal migration during embryonic brain development. Mol Neurobiol 53, 2969–2982.

54.

, Ip

, Fu

, Ip

(2014) Cdk5-mediated phosphorylation of RapGEF2 controls neuronal migration in the developing cerebral cortex. Nat Commun 5, 4826.

55.

Petrik

, Yun

, Latchney

, Kamrudin

, LeBlanc

, Bibb

, Eisch

(2013) Early postnatal in vivo gliogenesis from nestin-lineage progenitors requires cdk5. PLoS One 8, e72819.

56.

Fang

, Chen

, Fu

, Ip

(2013) Axin directs the amplification and differentiation of intermediate progenitors in the developing cerebral cortex. Neuron 79, 665–679.