Nicotinamide riboside restores cognition through an upregulation of proliferator-activated receptor-γ coactivator 1α regulated β-secretase 1 degradation and mitochondrial gene expression in Alzheimer's mouse models

Abstract

Nicotinamide adenine dinucleotide (NAD)(+), a coenzyme involved in redox activities in the mitochondrial electron transport chain, has been identified as a key regulator of the lifespan-extending effects, and the activation of NAD(+) expression has been linked with a decrease in beta-amyloid (Aβ) toxicity in Alzheimer's disease (AD). Nicotinamide riboside (NR) is a NAD(+) precursor, it promotes peroxisome proliferator-activated receptor-γ coactivator 1 (PGC)-1α expression in the brain. Evidence has shown that PGC-1α is a crucial regulator of Aβ generation because it affects β-secretase (BACE1) degradation. In this study we tested the hypothesis that NR treatment in an AD mouse model could attenuate Aβ toxicity through the activation of PGC-1α-mediated BACE1 degradation. Using the Tg2576 AD mouse model, using in vivo behavioral analyses, biochemistry assays, small hairpin RNA (shRNA) gene silencing and electrophysiological recording, we found (1) dietary treatment of Tg2576 mice with 250 mg/kg/day of NR for 3 months significantly attenuates cognitive deterioration in Tg2576 mice and coincides with an increase in the steady-state levels of NAD(+) in the cerebral cortex; (2) application of NR to hippocampal slices (10 μM) for 4 hours abolishes the deficits in long-term potentiation recorded in the CA1 region of Tg2576 mice; (3) NR treatment promotes PGC-1α expression in the brain coinciding with enhanced degradation of BACE1 and the reduction of Aβ production in Tg2576 mice. Further in vitro studies confirmed that BACE1 protein content is decreased by NR treatment in primary neuronal cultures derived from Tg2576 embryos, in which BACE1 degradation was prevented by PGC-1α-shRNA gene silencing; and (4) NR treatment and PGC-1α overexpression enhance BACE1 ubiquitination and proteasomal degradation. Our studies suggest that dietary treatment with NR might benefit AD cognitive function and synaptic plasticity, in part by promoting PGC-1α-mediated BACE1 ubiquitination and degradation, thus preventing Aβ production in the brain.

Fig. 1

NR improves cognitive function in Tg2576 mice via a promotion of NAD⁺ and PGC-1α levels. (A) Treatment with NR 250 mg/kg/day in Tg2576 mice for 3 months improves cognitive function. Object recognition memory test, performed as described by Bevins and Besheer (2006). Values are expressed as mean ± standard error of the mean, n = 10 mice per group. * p < 0.05, 2-tailed Student t test. (B) The NR treatment significantly increased the levels of NAD⁺ levels measured by NAD/NADH Assay Kit (Abcam). (C) PGC-1α mRNA levels in brain. Values are expressed as mean ± standard error of the mean, n = 8 mice per group. * p < 0.05, n = 5 mice per group, 2-tailed Student t test. (D) Enzyme-linked immunosorbent assay showed the levels of Aβ_1–42 levels in brains treated with NR comparing with placebo-treated brains in Tg2576 mice. n = 8 mice. Abbreviations: Aβ, beta-amyloid; CTL, control; mRNA, messenger RNA; NAD, nicotinamide adenine dinucleotide; NR, nicotinamide riboside; PGC-1α, peroxisome proliferator-activated receptor-γ coactivator 1α.

Fig. 2

NR-treated slices from Tg2576 mice improves synaptic function but has no effect on WT mice. (A) Application of NR at 20 µM for 4 hours in hippocampal slices from Tg2576 mice (12–14-month-old) has no significantly effects on basal synaptic transmission recorded in CA1 region (A) (p > 0.05), however, the deficits of LTP in the Tg2576 mice was greatly rescued, n = 8 slices, 2-way analysis of variance; p < 0.01 (B). Perfusion of NR has no effects on the basal synaptic transmission (C) nor on the LTP (D) in hippocampal slices from WT littermates (n = 8, 2-way analysis of variance; p > 0.05). Abbreviations: fEPSP, field excitatory postsynaptic potentials; LTP, long-term potentiation; NR, nicotinamide riboside; WT, wild type.

Fig. 3

PGC-1α deficiency promotes the generation of Aβ peptides in vivo and in vitro. (A) Increased Aβ_1–40 or Aβ_1–42 concentrations in approximately 4-month-old PGC-1α^−/−/APP mice relative to age- and sex-matched control APP (Tg2576) mice as assessed by enzyme-linked immunosorbent assay. (B) Silencing PGC-1α in primary corticohippocampal neurons derived from Tg2576 embryos resulted in elevation of Alzheimer’s disease-type Aβ levels in the cultures with conditioned medium relative to control neuron cultures. PGC-1α silencing was achieved by infecting neurons with a PGC-1α shRNA adenovirus (a gift from Dr Puigserver) at 10 MOI. Control cells were infected with a GFP adenovirus. Aβ accumulation in cultured media was assessed 24 hours after adenoviral infection. Values are expressed as mean ± standard error of the mean from 2–3 independent experiments, with n = 3–4 per group; * p < 0.05; ** p < 0.01; 2-tailed Student t test. Abbreviations: Aβ, beta-amyloid; GFP, green fluorescent protein; MOI, multiplicity of infection; PGC-1α, peroxisome proliferator-activated receptor-γ coactivator 1α; shRNA, small hairpin RNA.

Fig. 4

NR promotes PGC-1α responses and BACE1 degradation in the brain. (A) NR-treated primary neuron cultures derived from Tg2576 mice showed increased PGC-1α protein levels probed by Western blot analysis with anti-PGC-1α antibody (Sigma). Values are expressed as mean ± standard error of the mean; n = 3; * p < 0.05, Student t test. (B) BACE1 protein levels are decreased by NR treatment and the decrease was attenuated by PGC-1α-shRNA probed by Western blot analysis with anti-BACE1 antibody (Sigma). Values are expressed as mean ± standard error of the mean; n = 3; * p < 0.05, Student t test. Inset shows the protein of adeno-PGC-1α and PGC-1α-shRNA after infected in neurons. Abbreviations: BACE1, β-secretase; CTR, control; NR, nicotinamide riboside; PGC-1α, peroxisome proliferator-activated receptor-γ coactivator 1α; shRNA, small hairpin RNA.

Fig. 5

PGC-1α expression promotes BACE1 degradation. (A) Primary hippocampal-cortical neurons derived from Tg2576 embryos at 14 days in vitro were infected by adenoviral-GFP PGC-1α, scramble PGC-1α-shRNA, or PGC-1α-shRNA, respectively, 72 hours after infection, cell lysates were collected and analyzed via Western blot using anti-BACE1 antibodies. BACE1 levels were quantified and normalized against the level of β-tubulin and plotted as percentage of CTL. Data are expressed as mean±standard error of the mean (n=5). * p<0.05 compared with CTL group. Inset represents PGC-1α immunoreactive signals. (B) The degradation of BACE1 caused by PGC-1α expression on was blocked by lactacystin treatment (5 µM). Data are expressed as mean ± standard error of the mean (n = 5). * p < 0.05 compared with CTL group. Inset represents BACE1 immunoreactive signals. Abbreviations: BACE1, β-secretase; CTL, control; GFP, green fluorescent protein; PGC-1α, peroxisome proliferator-activated receptor-γ coactivator 1α; shRNA, small hairpin RNA.

Fig. 6

NR promotes BACE1 ubiquitination and degradation is linked with PGC-1α. (A) BACE1 stable HEK293 cells were infected with adeno-PGC-1α or adeno-PGC-1α-shRNA and scramble shRNA (B). The adeno-GFP viral constructs were used as control. Cells were treated with NR 200 µg and 5 µM lactacystin. After 48 hours of transfection, cell lysate was collected and BACE1 immunoprecipitated using anti-BACE1 antibody, and probed with anti-ubiquitin antibodies (Sigma). (C) Quantification of the square area in the left panel by ImageJ and represented in the graph. This area represents BACE1 proteins at different ubiquitinated levels. Data are mean ± SEM (error bar) of the results from 2 independent experiments. Student t test was performed; * p < 0.05 compared with GFP-lactacystin treated cells. Abbreviations: BACE1, β-secretase; CTL, control; GFP, green fluorescent protein; NR, nicotinamide riboside; PGC-1α, peroxisome proliferator-activated receptor-γ coactivator 1α; SEM, standard error of the mean; WB, Western blot analysis.

Fig. 7

Energy metabolism gene products are induced in response to short-term NR treatment in the brains of Tg2576 mice. In this study, 5-month-old Tg2576 mice were treated with 250 mg/kg/day of NR for approximately 3 months, and total RNA from the cerebral cortex was extracted using RNeasy Mini Kit (Qiagen) 24 hours after behavioral testing. Complementary DNA was synthesized using Superscript III First-Strand Synthesis SuperMix for quantitative RT-PCR (Invitrogen) with 1 µg of total RNA. Quantitative RT-PCR was performed using Maxima SYBR Green Master Mix (Fermentas). Gene targets examined here included (A) CS, (B) aconitase, (C) PGC-1α, (D) PDHA, (E) PDK3, (F) COX6C, (G) PGK1, (H) M-PFK, and (I) GPI1. Fold changes between groups were calculated using the ΔCt method. Statistical analysis was done in GraphPad Prism 4. Abbreviations: COX6C, cytochrome c subunit Vic; CS, citrate synthase; GPI1, glucose phosphate isomerase 1; M-PFK, muscle phosphofructokinase; NR, nicotinamide riboside; OXPHOS, oxidative phosphorylation system; PCR, polymerase chain reaction; PDHA, pyruvate dehydrogenase; PDK3, pyruvate dehydrogenase kinase 3; PGC-1α, peroxisome proliferator-activated receptor-γ coactivator 1α; PGK1, phosphoglycerate kinase; RT, reverse transcription; TCA, tricarboxylic acid cycle.