Vulnerability of primary neurons derived from Tg2576 Alzheimer mice to oxygen and glucose deprivation: role of intraneuronal amyloid-β accumulation and astrocytes

Abstract

Microvascular dysfunction is considered an integral part of Alzheimer disease (AD) pathogenesis, but the possible relationship between amyloid pathology, microvascular dysfunction and cell death is still unclear. In order to investigate the influence of intraneuronal amyloid-β (Aβ) accumulation on vulnerability to hypoxia, we isolated primary cortical neurons from Tg2576 (carrying the amyloid precursor protein APPSwe mutation) and wild-type fetal mice. We first demonstrated that neurons isolated from Tg2576 newborn mice show an increase in VEGFa mRNA expression and a decrease in the expression of the two VEGF receptors, Flt1 and Kdr, compared with wild-type cells. Moreover, APPSwe primary neurons displayed higher spontaneous and glutamate-induced cell death. We then deprived the cultures of oxygen and glucose (OGD) as an in vitro model of hypoxia. After OGD, APPSwe neurons display higher levels of cell death in terms of percentage of pyknotic/fragmented nuclei and mitochondrial depolarization, accompanied by an increase in the intraneuronal Aβ content. To explore the influence of intraneuronal Aβ peptide accumulation, we used the γ-secretase inhibitor LY450139, which showed that the reduction of the intracellular amyloid fully protects APPSwe neurons from OGD-induced degeneration. Conditioned medium from OGD-exposed APPSwe or wild-type astrocytes protected APPswe neurons but not wild-type neurons, during OGD. In conclusion, the presence of the mutated human APP gene, leading to the intracellular accumulation of APP and Aβ fragments, worsens OGD toxicity. Protection of APPSwe neurons can be obtained either using a γ-secretase inhibitor or astrocyte conditioned medium.

Keywords: Alzheimer's disease; Glutamate; Intraneuronal amyloid; Neurovascular coupling; Oxygen glucose deprivation; Primary neurons.

Fig. 1.

Culture characterization. (A) Representative images of double-stained cells showing β-III-tubulin-positive neurons and GFAP-positive astrocytes. Scale bar: 50 µm. (B) Representative images of 6E10-stained cells showing intracellular accumulation of the human APP protein/β-amyloid fragments. Scale bar: 10 µm. Nuclei are stained blue with Hoechst 33258. (C) mRNA expression level of VEGF and VEGF receptors FLT-1 and KDR (VEGFA, Wt n=15, APPswe n=11; FLT1, Wt n=13, APPswe n=11; KDR, Wt n=13, APPswe n=9). Bars represent mean+s.e.m. Statistical analysis: Student's t-test between genotypes (*P<0.05; **P<0.01; ***P<0.001).

Fig. 2.

Vulnerability of Wt and APPswe neurons to culture condition, glutamate excitotoxicity and OGD. (A) Experimental design. Primary neurons isolated from Wt and Tg2576 mice were exposed at 7 DIV to the challenge stimulus (42 µM glutamate or 3 h OGD). Cells were then exposed to the original culture medium for 24 h. (B) Cell viability analysis of Wt and APPswe neurons exposed to vehicle and glutamate, as established by nuclear morphology (Wt vehicle, n=4; Wt glutamate, n=5; APPswe vehicle, n=5; APPswe glutamate, n=5). (C,D) Cell viability analysis of Wt and APPswe neurons exposed to normoxia and OGD, as established by mitochondrial function (C; Wt normoxia, n=10; Wt OGD, n=10; APPswe normoxia, n=8; APPswe OGD, n=9) and nuclear morphology (D; Wt normoxia, n=6; Wt OGD, n=5; APPswe normoxia, n=5; APPswe OGD, n=5). (E-L) Representative images of MitoTracker-stained cells (E,F,I,J) and Hoechst-stained nuclei (G,H,K,L) isolated from Wt (E,I,G,K) and APPswe (F,J,H,L) mice and exposed to normoxia (E-H) or OGD (I-L). Scale bar: 50 µm (E,F,I,J) and 80 µm (G,H,K,L). G and K include high-magnification images of a normal nucleus (G) or pyknotic/fragmented nuclei (K); scale bar: 10 µm. Bars represent mean±s.e.m. Statistical analysis: Two-way ANOVA, followed by Sidak's multiple comparison test. Asterisks represent differences between vehicle- and glutamate-treated groups (B; ****P<0.0001) or between normoxia- and OGD-exposed groups (C,D; *P<0.05; **P<0.01; ***P<0.001); letters represent differences between genotypes (a, P<0.05; c, P<0.001; d, P<0.0001).

Fig. 3.

Effect of γ-secretase inhibition on the vulnerability of primary neurons to OGD. (A) Experimental design. Primary neurons isolated from Wt and Tg2576 mice were treated with LY450139 from 2 DIV to the end of the experiment. At 7 DIV, cells were exposed to 3 h of OGD and 24 h of reperfusion in the previous culture medium. (B) Representative images of 6E10 immunostaining of APPswe neurons, showing the intracellular accumulation of human APP protein/Aβ fragments. Scale bar: 50 µm. (C) Quantification of 6E10 in Tg2576 cells exposed to normoxia and treated or not with LY450139 (APPswe vehicle, n=5; APPswe LY450139, n=5). (D,E) Cell viability analysis of Wt and APPswe neurons exposed to normoxia and OGD, treated or not with LY450139, as established by mitochondrial function (E; Wt normoxia, n=10; Wt normoxia, LY450139, n=6; Wt OGD, n=10; Wt OGD LY450139, n=4; APPswe normoxia, n=7; APPswe normoxia LY450139, n=6; APPswe OGD, n=8; APPswe OGD LY450139, n=4) and nuclear morphology (F; Wt normoxia, n=6; Wt normoxia LY450139, n=4; Wt OGD, n=5; Wt OGD LY450139, n=4; APPswe normoxia, n=5; APPswe normoxia LY450139, n=4; APPswe OGD, n=5; APPswe OGD LY450139, n=5). Bars represent mean±s.e.m. Statistical analysis: one-way ANOVA followed by Tukey's multiple comparisons test inside the same genotype. Asterisks represent differences between LY450139- and vehicle-treated groups (*P<0.05; ***P<0.001; ****P<0.0001).

Fig. 4.

Effect of astrocyte conditioned medium on the vulnerability of primary neurons to OGD. (A) Experimental design. Primary astrocytes isolated from Wt and Tg2576 mice were exposed to 3 h OGD, and conditioned medium collected. Wt and APPswe primary neurons were then exposed to Wt and APPswe ACM during normoxia or OGD and reperfusion. (B) 6E10 immunostaining of APPswe astrocytes, showing the intracellular accumulation of human APP protein/Aβ fragments in GFAP-positive cells. Scale bar: 50 µm. (C,D) Cell viability analysis of Wt and APPswe neurons exposed to normoxia and OGD, treated or not with ACM, as established by mitochondrial function (C; Wt OGD, n=10; Wt OGD ACM-Wt, n=5; Wt OGD ACM-APPswe, n=6; APPswe OGD, n=9; APPswe OGD ACM-Wt, n=5; APPswe OGD ACM-APPswe, n=6) and nuclear morphology (D; Wt OGD, n=5; Wt OGD ACM-Wt, n=5; Wt OGD ACM-APPswe, n=4; APPswe OGD, n=5; APPswe OGD ACM-Wt, n=5; APPswe OGD ACM-APPswe, n=5). The yellow horizontal bars represent normoxia values (the height of the bar represent the range of the mean value±s.e.m.). Bars represent mean+s.e.m. Statistical analysis: one-way ANOVA followed by Dunnett's multiple comparison test inside the same genotype. Asterisks represent differences between ACM-treated groups and groups exposed to OGD only (*P<0.05; **P<0.01; ****P<0.0001).