Phospho-eIF2α level is important for determining abilities of BACE1 reduction to rescue cholinergic neurodegeneration and memory defects in 5XFAD mice

Abstract

β-Site APP-cleaving enzyme 1 (BACE1) initiates amyloid-β (Aβ) generation and thus represents a prime therapeutic target in treating Alzheimer's disease (AD). Notably, increasing evidence indicates that BACE1 levels become elevated in AD brains as disease progresses; however, it remains unclear how the BACE1 upregulation may affect efficacies of therapeutic interventions including BACE1-inhibiting approaches. Here, we crossed heterozygous BACE1 knockout mice with AD transgenic mice (5XFAD model) and compared the abilities of partial BACE1 reduction to rescue AD-like phenotypes at earlier (6-month-old) and advanced (15-18-month-old) stages of disease, which expressed normal (∼100%) and elevated (∼200%) levels of BACE1, respectively. BACE1(+/-) deletion rescued memory deficits as tested by the spontaneous alternation Y-maze task in 5XFAD mice at the earlier stage and prevented their septohippocampal cholinergic deficits associated with significant neuronal loss. Importantly, BACE1(+/-) deletion was no longer able to rescue memory deficits or cholinergic neurodegeneration in 5XFAD mice at the advanced stage. Moreover, BACE1(+/-) deletion significantly reduced levels of Aβ42 and the β-secretase-cleaved C-terminal fragment (C99) in 6-month-old 5XFAD mouse brains, while these neurotoxic β-cleavage products dramatically elevated with age and were not affected by BACE1(+/-) deletion in 15-18-month-old 5XFAD brains. Interestingly, although BACE1(+/-) deletion lowered BACE1 expression by ∼50% in 5XFAD mice irrespective of age in concordance with the reduction in gene copy number, BACE1 equivalent to wild-type controls remained in BACE1(+/-)·5XFAD mice at the advanced age. In accord, phosphorylation of the translation initiation factor eIF2α, an important mediator of BACE1 elevation, was dramatically increased (∼9-fold) in 15-18-month-old 5XFAD mice and remained highly upregulated (∼6-fold) in age-matched BACE1(+/-)·5XFAD mice. Together, our results indicate that partial reduction of BACE1 is not sufficient to block the phospho-eIF2α-dependent BACE1 elevation during the progression of AD, thus limiting its abilities to reduce cerebral Aβ/C99 levels and rescue memory deficits and cholinergic neurodegeneration.

Figure 1. Age-related effects of BACE1^+/− deletion on memory deficits in 5XFAD mice.

Memory function of mice at 6 months (A, C) and 15–18 months (B, D) of age was tested using the spontaneous alternation Y-maze task. (A, B) Spatial working memory, as assessed by the spontaneous alternation performance, is significantly impaired (around 50% chance levels) in 5XFAD mice irrespective of age compared to wild-type controls (* p<0.05). Note that BACE1^+/−·5XFAD are rescued completely back to wild-type levels of alternation performance at 6 months but not at 15–18 months of age (^# p<0.05 versus age-matched 5XFAD). n = 5–12 mice per group. (C, D) Total number of arm entries reflecting exploratory activities of mice in the Y-maze does not differ between the four groups irrespective of age. n = 5–12 mice per group. All data are presented as mean ± SEM.

Figure 2. Age-related effects of BACE1^+/− deletion on cholinergic neurodegeneration in 5XFAD mice.

(A, B) Immunoblot analysis of ChAT in hippocampal lysates from wild-type mice and 5XFAD mice with BACE1^+/+ or BACE1^+/− genotype. Intensities of ChAT-immunoreactive bands of mice at 6 months (A) and 15–18 months (B) of age were quantified by phosphorimaging and expressed as percentage of wild-type levels (n = 6–9 mice per group). (C, D) Brain sections from wild-type mice and 5XFAD mice with BACE1^+/+ or BACE1^+/− genotype were immunostained for ChAT. Shown are representative photomicrographs of ChAT-immunoreactive neurons in the medial septum of mice at 6 months (C) and 15–18 months (D) of age. Scale bar  = 200 µm. The number of ChAT-positive neurons in the medial septum and the vertical limb of the diagonal band (Ch1/2), which provide the cholinergic innervation to the hippocampus, was counted for quantification (n = 4–7 mice per group). Both ChAT measures are significantly reduced in 5XFAD mice irrespective of age as compared with wild-type controls (* p<0.05). Note that ChAT is restored completely back to wild-type levels in BACE1^+/−·5XFAD mice at 6 months but not at 15–18 months of age (^# p<0.05 versus age-matched 5XFAD). All data are presented as mean ± SEM.

Figure 3. Age-related effects of BACE1^+/− deletion on APP processing and Aβ levels in 5XFAD mice.

(A) Immunoblot analysis of hemibrain lysates from 5XFAD mice with BACE1^+/+ or BACE1^+/− genotype at 6 months and 15–18 months of age. (B, C) Intensities of immunoreactive bands for full-length APP (B) and C99 (C) were quantified by phosphorimaging and expressed as percentage of 6-month-old 5XFAD mouse levels (n = 5–7 mice per group). (D) Levels of total Aβ42 were quantified by sandwich ELISA of guanidine extracts of hemibrain samples and expressed in nanograms per milligram of total protein (n = 4 mice per group). APP overexpression levels are not changed in BACE1^+/−·5XFAD mice irrespective of age as compared to 5XFAD mice. Levels of C99 and Aβ42 are significantly lower in BACE1^+/−·5XFAD mice at 6 months of age as compared with 5XFAD wild-type controls (^# p<0.05). Note that C99 and Aβ42 levels are significantly elevated in 15–18-month-old 5XFAD mouse brains compared to 6-month-old 5XFAD brains (* p<0.05), while BACE1^+/− deletion is no longer able to reduce levels of the β-cleavage products in 5XFAD mice at the advanced age. All data are presented as mean ± SEM.

Figure 4. Age-related changes in BACE1 levels and the modification by BACE1^+/− deletion in 5XFAD mice.

(A, B) Immunoblot analysis of BACE1 in hemibrain lysates from wild-type mice and 5XFAD mice with BACE1^+/+ or BACE1^+/− genotype. Intensities of BACE1-immunoreactive bands of mice at 6 months (A) and 15–18 months (B) of age were quantified by phosphorimaging and expressed as percentage of wild-type levels (n = 4–8 mice per group). BACE1^+/− deletion reduced BACE1 expression by ∼50% in 5XFAD mice irrespective age as compared with age-matched 5XFAD mice (^# p<0.05). However, since BACE1 expression is significantly elevated up to ∼200% in 15–18-month-old 5XFAD mouse brains (* p<0.05 versus wild-type), levels of BACE1 equivalent to wild-type controls remain in BACE1^+/−·5XFAD brains at the advanced age. All data are presented as mean ± SEM.

Figure 5. Age-related changes in phospho-eIF2α levels and the modification by BACE1^+/− deletion in 5XFAD mice.

(A, B) Immunoblot analysis of phospho-eIF2α (p-eIF2α) in hemibrain lysates from wild-type mice and 5XFAD mice with BACE1^+/+ or BACE1^+/− genotype. Intensities of p-eIF2α-immunoreactive bands of mice at 6 months (A) and 15–18 months (B) of age were quantified by phosphorimaging and expressed as percentage of wild-type levels (n = 4–9 mice per group). While changes in p-eIF2α levels are marginal at 6 months of age, p-eIF2α levels are dramatically elevated in 15–18-month-old 5XFAD (∼9-fold relative to wild-type controls) and BACE1^+/−·5XFAD (∼6-fold) mouse brains. * p<0.05 versus wild-type, ^# p<0.05 versus 5XFAD. All data are presented as mean ± SEM. (C) After brain slices from 6-month-old 5XFAD mice were exposed to normal aCSF or Sal 003-containing aCSF (30 µM), they were lysed for immunoblot analysis. The eIF2α phosphatase inhibitor Sal 003-induced increase in p-eIF2α level caused the elevation of BACE1 in 5XFAD mouse brains.