Neurotoxicity and memory deficits induced by soluble low-molecular-weight amyloid-β1-42 oligomers are revealed in vivo by using a novel animal model

Abstract

Neuronal and synaptic degeneration are the best pathological correlates for memory decline in Alzheimer's disease (AD). Although the accumulation of soluble low-molecular-weight amyloid-β (Aβ) oligomers has been suggested to trigger neurodegeneration in AD, animal models overexpressing or infused with Aβ lack neuronal loss at the onset of memory deficits. Using a novel in vivo approach, we found that repeated hippocampal injections of small soluble Aβ(1-42) oligomers in awake, freely moving mice were able to induce marked neuronal loss, tau hyperphosphorylation, and deficits in hippocampus-dependent memory. The neurotoxicity of small Aβ(1-42) species was observed in vivo as well as in vitro in association with increased caspase-3 activity and reduced levels of the NMDA receptor subunit NR2B. We found that the sequestering agent transthyretin is able to bind the toxic Aβ(1-42) species and attenuated the loss of neurons and memory deficits. Our novel mouse model provides evidence that small, soluble Aβ(1-42) oligomers are able to induce extensive neuronal loss in vivo and initiate a cascade of events that mimic the key neuropathological hallmarks of AD.

Figure 1.

General profiles of Aβ_1–42 oligomers injected in vivo and oligomerization of Aβ_1–42 over time. A, Representative immunoblot of the Aβ_1–42 solution used in our mouse model. The Aβ_1–42 preparation is kept at 25°C for 1 h before administration and is almost exclusively composed of low-molecular-weight Aβ_1–42 oligomers as revealed by electrophoresis on an SDS-NuPAGE Novex 4–12% Bis-Tris gel system using the Aβ antibodies 6E10 and 3D6. B, Representative 6E10-immunoblots of Aβ_1–42 oligomerization. Soluble Aβ_1–42 forms higher molecular weight (MW) oligomers over time and its concentration decreases in favor of insoluble Aβ_1–42 species, which were recovered from the side of low-adhesion tubes with formic acid treatment. Scrambled Aβ_1–42 (Aβscr) is not recognized by the 6E10 antibody (n = 3 for each gel in 3 independent experiments). C, The Aβ_1–42 solution was verified in nondenaturing condition by TEM 1 h (i–iii) and 24 h (iv–vi) after preparation (i and iv, ×30000, scale bars 0, 5 μm; ii and v, ×85000, scale bars 0, 1 μm; iii and vi, ×140000, scale bars 0, 1 μm).

Figure 2.

Neuronal deaths induced by Aβ_1–42-oligomer deposition and tau phosphorylation following repeated intrahippocampal injections. A, Bilateral cannulae were stereotaxically implanted in the DG. One week following surgery, Aβ_1–42 oligomers (Aβ; 0.2 μg/μl; 2 μl) and vehicle (Ctl) were simultaneously injected collaterally in awake, freely moving mice once a day for 6 consecutive days and killed 24 h following the last injection. B, Representative accumulation of Aβ_1–42 oligomers in the DG on a section immediately next to the cannulae insertion site. Bottom, Representative staining with Fluoro-Jade B, cresyl violet, cleaved caspase-3 antibody, and NR2B antibody. C, Quantification of Fluoro-Jade B staining within the DG 24 h following the last injection. Marked surface (MS) versus total surface (TS) counted (n = 4, ***p < 0.001). D, Representative immunoblots of Aβ_1–42 oligomer profiles 24 h following the last injection. E, Representative 6E10-immunostained section showing cannulae insertion. F, Representative accumulation of Aβ_1–42 oligomers and cell death when the Aβ_1–42 preparation is injected in the more ventral part of the hippocampus (−3.4 mm AP, ±2.0 mm ML, −2.4 mm DV from bregma) in sections stained with the 6E10 antibody and cresyl violet. G, Representative DG sections stained with 6E10 antibody and cresyl violet following collateral injections (1 per day for 6 d) of vehicle (Ctl) and scrambled Aβ_1–42 (Aβscr; 0.2 μg/μl; 2 μl). H, Representative immunoblots of tau phosphorylation in the dorsal hippocampus of mice injected collaterally with vehicle (Ctl) and Aβ_1–42 oligomers for 6 d. Bottom, Densitometric quantification of changes expressed as the mean ratio of phospho-tau to total tau antibody (Tau N-ter) staining. Error bars indicate ± SEM. **p < 0.01. I, Representative AT8-immunostained section 24 h and 5 d after the last injection (n = 4, two independent experiments). Scale bars: 50 μm.

Figure 3.

Time-dependent deposition of Aβ_1–42 oligomers and neuronal death. Representative DG sections following single (A) or multiple (B) injections of Aβ_1–42 oligomers (Aβ) and vehicle (Ctl) at the time points indicated and stained with the 6E10 antibody and cresyl violet. Scale bars: 50 μm (n = 4 for each time point, two independent experiments).

Figure 4.

Neurotoxicity induced by soluble Aβ_1–42 oligomers in primary hippocampal culture is accompanied by tau phosphorylation, increased caspase-3 activity, and decreased levels of spinophilin and NR2B subunit. A, MTS and resazurin assays following 72 h treatment with vehicle (Veh), 2 μm of scrambled Aβ_1–42 oligomers (Aβscr), or 2 μm of soluble Aβ_1–42 oligomers (Aβ) kept at 25°C for 1 h before administration. The histograms are mean values expressed as a percentage of control (cultures without any treatment). B, C, Representative immunoblots of tau phosphorylation (B) as well as spinophilin, procaspase-3 and cleaved caspase-3, NR2B, and GAPDH (C) from neurons treated with vehicle (Ctl) or soluble Aβ_1–42 oligomers (2 μm). For tau phosphorylation, densitometric quantification of changes is expressed as the mean ratio of phospho-tau to total tau antibody (Tau N-ter) staining. Results are all expressed normalized to GAPDH. Error bars indicate SEM. *p < 0.05, **p < 0.01, ***p < 0.001 (3 independent experiments performed in duplicate).

Figure 5.

Neurotoxicity induced by soluble Aβ_1–42 oligomers prepared 1 or 24 h before administration. A, Profiles of Aβ_1–42 oligomers kept at 25°C for 1 h (Aβ 1 h) or 37°C for 24 h (Aβ 24 h) before administration. B, Representative DG sections 24 h following the last injection (6 injections, 1 per day) of Aβ_1–42 oligomers (1 and 24 h; 0.2 μg/μl; 2 μl) or vehicle (Ctl) stained with the 6E10 antibody and cresyl violet. Scale bars: 50 μm (n = 4, two independent experiments). Bottom, Resazurin assays on neuronal cultures treated with vehicle (Veh) or 2 μm of Aβ 1 h or Aβ 24 h. Histograms show mean values expressed as a percentage of controls (cultures without any treatment) ± SEM. *p < 0.05, **p < 0.01 (3 independent experiments performed in duplicate). C, Hippocampal Aβ_1–42 oligomer profiles 24 h following the last injection (6 injections, 1 per day) using the 6E10 antibody. Bottom, The membrane was incubated with the anti-mouse secondary antibody (AB) without the 6E10 primary mouse monoclonal antibody (n = 4, two independent experiments).

Figure 6.

Sequestration of soluble Aβ_1–42 oligomers with TTR. A, Representative 6E10-immunoblot of soluble Aβ_1–42 oligomers 1 or 24 h following Aβ_1–42 preparation with TTR (TTR-Aβ, molar (M) ratio of 2:1) or Aβ_1–42 alone (Aβ) at the temperature indicated. B, Increasing doses of TTR mixed with Aβ_1–42 at 37°C during 24 h promoted the sequestration and retrieval of soluble Aβ_1–42 oligomers on an SDS-NuPAGE gel system using the 6E10 antibody. Lower levels of insoluble Aβ_1–42 stuck to the side of low-adhesion tubes were recovered with formic acid treatment while increasing the TTR concentration in Aβ_1–42 solutions (right blot) (n = 3 for each gel, 3 independent experiments).

Figure 7.

Neuronal death, tau phosphorylation, and memory deficits induced by Aβ_1–42 oligomers are attenuated by TTR. A, Representative DG section 24 h after the last injection (1 per day during 6 d) with Aβ_1–42 oligomers alone (0.2 μg/μl, 2 μl) or with TTR (TTR-Aβ, molar ratio of 2:1), TTR alone, or vehicle (Ctl), kept at 37°C for 24 h before administration. Scale bars: 50 μm. B, Representative immunoblots of hippocampal tau phosphorylation in mice injected collaterally with Aβ_1–42 oligomers and TTR-Aβ. Densitometric quantification of changes is expressed as the mean ratio of AT8 antibody to total Tau N-ter. Error bars indicate ± SEM. *p < 0.05 (n = 4, 2 independent experiments). C, D, Mice bilaterally injected in the DG with control vehicle, Aβ_1–42 oligomers, or TTR-Aβ were tested in the passive avoidance memory task (C) and the Y maze (D). During the passive avoidance conditioning, latency of entry into the dark compartment was evaluated during the training phase and the testing phase 24 h later. For the Y maze, we analyzed the percentage of time spent in the arm (other) already explored during the training phase and the new arm available during the testing phase. Values are mean latency ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 (n = 8 for each experimental group for the passive avoidance task, n = 4 for each experimental group for the Y maze).

Figure 8.

TTR prevented the deleterious effects of soluble Aβ_1–42 oligomers in primary hippocampal cultures. A, Resazurin assays following a 72 h treatment with vehicle (Veh), 2 μm of scramble Aβ_1–42 oligomers (Aβscr), or 2 μm of soluble Aβ_1–42 oligomers alone (Aβ) or with TTR (TTR-Aβ; molar ratio of 2:1) kept at 37°C for 24 h before administration. B, C, Representative immunoblots of tau phosphorylation (B) as well as spinophilin, caspase-3, NR2B, and GAPDH (C) for neurons treated with vehicle (Ctl), Aβ, or TTR-Aβ (molar ratio of 2:1). For tau phosphorylation, densitometric quantification of changes is expressed as the mean ratio of AT8 to Tau N-ter antibody. The histograms are mean values expressed as a percentage of control ± SEM. * or ** indicates a statistical significant variation between Ctl and Aβ, and † or ‡ indicates a difference between Aβ and TTR-Aβ. *^,†p < 0.05, **^,‡p < 0.01, ***p < 0.001 (3 independent experiments performed in duplicate).