Carboxyl-terminal-truncated apolipoprotein E4 causes Alzheimer's disease-like neurodegeneration and behavioral deficits in transgenic mice

Abstract

Apolipoprotein (apo) E4 increases the risk and accelerates the onset of Alzheimer's disease (AD). However, the underlying mechanisms remain to be determined. We previously found that apoE undergoes proteolytic cleavage in AD brains and in cultured neuronal cells, resulting in the accumulation of carboxyl-terminal-truncated fragments of apoE that are neurotoxic. Here we show that this fragmentation is caused by proteolysis of apoE by a chymotrypsin-like serine protease that cleaves apoE4 more efficiently than apoE3. Transgenic mice expressing the carboxyl-terminal-cleaved product, apoE4(Delta272-299), at high levels in the brain died at 2-4 months of age. The cortex and hippocampus of these mice displayed AD-like neurodegenerative alterations, including abnormally phosphorylated tau (p-tau) and Gallyas silver-positive neurons that contained cytosolic straight filaments with diameters of 15-20 nm, resembling preneurofibrillary tangles. Transgenic mice expressing lower levels of the truncated apoE4 survived longer but showed impaired learning and memory at 6-7 months of age. Thus, carboxyl-terminal-truncated fragments of apoE4, which occur in AD brains, are sufficient to elicit AD-like neurodegeneration and behavioral deficits in vivo. Inhibiting their formation might inhibit apoE4-associated neuronal deficits.

Fig. 1.

Cleavage of apoE by a chymotrypsin-like serine protease. ApoE fragmentation in human brain lysates was detected by Western blotting with antibodies against full-length apoE (a) or the carboxyl-terminal 28 amino acids of apoE (b) and quantified by densitometry (c). Brain tissues were from 17 nondemented subjects (10 apoE3/3, mean age 72 ± 6 years; seven apoE4/3, 70 ± 5 years) and 19 AD cases (nine apoE3/3, 75 ± 7 years; 10 apoE4/3 and apoE4/4, 72 ± 6 years). (d) ApoE fragmentation was assessed in vitro by incubating recombinant human apoE3 or apoE4 (1 μg) with brain lysate (20 μl) from apoE-deficient mice at 37°C for 3 h followed by anti-apoE Western blotting. (e) Effects of protease inhibitors (1–20 mM) on apoE4 cleavage by partially purified AECE. (f) Determination of substrate specificity of AECE with synthetic peptides (50–100 μM). (g) Time-dependent cleavage of apoE4 (1 μg) by partially purified AECE (5 μl) in vitro.(h) Effect of the Leu-268 to Ala mutation on apoE4 cleavage by AECE in transfected Neuro-2a cells. Neuro-2a cells were transiently transfected with cDNA constructs encoding wild-type apoE4 or apoE4–Ala-268. ApoE fragmentation was determined by SDS/PAGE and anti-apoE Western blotting 24 h after transfection. (i) Effect of the Met-272 to Ala mutation on apoE4 cleavage by AECE in Neuro-2a cells transfected with wild-type apoE4 or apoE4–Ala-272. ApoE fragmentation was determined by SDS/PAGE and anti-apoE Western blotting 24 h after transfection.

Fig. 2.

Expression of human apoE4(Δ272–299) or apoE4(Δ241–299) in neurons of transgenic mice. (a) Thy-1-apoE cDNA constructs used for generating transgenic mice. (b) In situ hybridization of human apoE mRNA in the neocortex and hippocampus of transgenic mice expressing apoE4(Δ272–299) and nontransgenic wild-type (WT) controls. (c) Human apoE-specific Western blot analysis of brain homogenates from different founder mice. NSE, neuron-specific enolase.

Fig. 3.

Neurodegeneration in the brains of hemizygous transgenic mice expressing apoE4(Δ272–299) in neurons. Brain sections from 2- to 4-month-old transgenic mice (C57BL/6J background) expressing high levels of apoE4(Δ272–299) (A–E) or apoE4(Δ241–299) (F and G) at comparable levels were stained with anti-apoE (A–C and F) or hematoxylin and eosin (D, E, and G). Nontransgenic wild-type (Wt) mice (D and E) served as additional controls. Note the formation of truncated apoE4-containing inclusions in cortical (A), CA1 (B), and CA3 (C) neurons and degeneration of neurons in CA1 (D Upper) and CA3 (E Upper) in apoE4(Δ272–299) mice. Original magnifications: A, C, and F, ×600; B, D, E, and G, ×400.

Fig. 4.

Hyperphosphorylation of tau and formation of preNFT-like filaments in the brains of hemizygous transgenic mice expressing apoE4(Δ272–299). P-tau in the supernatants (a and b) and the solubilized pellets (c and d) of brain lysates of 2- to 4-month-old wild-type (n = 6) or high-expresser apoE4(Δ272–299) (n = 4) mice was detected by Western blotting with monoclonal antibody AT8 (a and c) and quantified by densitometry (b and d). (e) Cortical neurons positive for AT8 immunostaining in transgenic mice. (f) Cortical neurons positive for Gallyas silver staining in transgenic mice. (g) Intraneuronal straight filaments with diameters of 15–20 nm visualized by electron microscopy in transgenic mice. Original magnifications: e and f, ×600; g, ×60,000.

Fig. 5.

Learning and memory deficits of hemizygous transgenic mice expressing low levels of apoE4(Δ272–299). Five transgenic (Tg) founder mice (C57BL/6J) expressing low levels of apoE4(Δ272–299) on a wild-type mouse apoE background and five nontransgenic (Non Tg) wild-type littermates (all males) from the same founder generation were tested in the water maze at 6–7 months of age. Mice were tested in two sessions per day (three trials for each session). The y axis indicates the time to reach the target platform (mean ± SEM). For sessions 1–10, the platform was hidden under the opaque water in a constant location. For sessions 11–15, the platform was visible and changed in location between sessions. Repeated-measures ANOVA revealed significant differences in the learning curves of transgenic and nontransgenic mice for the hidden (P < 0.01), but not the visible (P > 0.05), component of the water maze test.