Mechanisms underlying insulin deficiency-induced acceleration of β-amyloidosis in a mouse model of Alzheimer's disease

Abstract

Although evidence is accumulating that diabetes mellitus is an important risk factor for sporadic Alzheimer's disease (AD), the mechanisms by which defects in insulin signaling may lead to the acceleration of AD progression remain unclear. In this study, we applied streptozotocin (STZ) to induce experimental diabetes in AD transgenic mice (5XFAD model) and investigated how insulin deficiency affects the β-amyloidogenic processing of amyloid precursor protein (APP). Two and half months after 5XFAD mice were treated with STZ (90 mg/kg, i.p., once daily for two consecutive days), they showed significant reductions in brain insulin levels without changes in insulin receptor expression. Concentrations of cerebral amyloid-β peptides (Aβ40 and Aβ42) were significantly increased in STZ-treated 5XFAD mice as compared with vehicle-treated 5XFAD controls. Importantly, STZ-induced insulin deficiency upregulated levels of both β-site APP cleaving enzyme 1 (BACE1) and full-length APP in 5XFAD mouse brains, which was accompanied by dramatic elevations in the β-cleaved C-terminal fragment (C99). Interestingly, BACE1 mRNA levels were not affected, whereas phosphorylation of the translation initiation factor eIF2α, a mechanism proposed to mediate the post-transcriptional upregulation of BACE1, was significantly elevated in STZ-treated 5XFAD mice. Meanwhile, levels of GGA3, an adapter protein responsible for sorting BACE1 to lysosomal degradation, are indistinguishable between STZ- and vehicle-treated 5XFAD mice. Moreover, STZ treatments did not affect levels of Aβ-degrading enzymes such as neprilysin and insulin-degrading enzyme (IDE) in 5XFAD brains. Taken together, our findings provide a mechanistic foundation for a link between diabetes and AD by demonstrating that insulin deficiency may change APP processing to favor β-amyloidogenesis via the translational upregulation of BACE1 in combination with elevations in its substrate, APP.

Figure 1. Effects of STZ-induced insulin deficiency on Aβ accumulation in 5XFAD mice.

(A) Western blot analysis of hemibrain lysates from vehicle- and STZ-treated 5XFAD mice. (B) Immunoreactive bands for insulin and insulin receptor (IR) were quantified and expressed as the percentage of vehicle-treated 5XFAD levels (n = 4–7 mice per group). STZ treatments significantly reduced cerebral insulin levels without affecting IR in 5XFAD mice (*p<0.05 vs. vehicle). (C) Levels of total Aβ40 and Aβ42 were quantified by sandwich ELISAs of guanidine extracts of hemibrain samples and expressed as the percentage of vehicle-treated 5XFAD levels (n = 3–5 mice per group). Aβ40 and Aβ42 levels were significantly higher in brains of STZ-treated 5XFAD mice (*p<0.05 vs. vehicle). All data are presented as mean ± SEM.

Figure 2. Effects of STZ-induced insulin deficiency on APP processing in 5XFAD mice.

(A) Western blot analysis of hemibrain lysates from vehicle- and STZ-treated 5XFAD mice. Immunoreactive bands for secretases involved in the APP cleavage (B), full-length APP (fl-APP) and its metabolites (C) were quantified and expressed as the percentage of vehicle-treated 5XFAD levels (n = 4–7 mice per group). BACE1, fl-APP and C99 levels were significantly increased, while sAPPα levels were significantly reduced in STZ-treated 5XFAD mice (*p<0.05 vs. vehicle). All data are presented as mean ± SEM.

Figure 3. Mechanisms by which STZ-induced insulin deficiency elevates BACE1 levels in 5XFAD mice.

(A) Real-time qPCR revealed no difference in BACE1 mRNA levels between STZ- and vehicle-treated 5XFAD mouse brains (n = 4–5 mice per group). (B, E) Western blot analysis of hemibrain lysates from vehicle- and STZ-treated 5XFAD mice. Immunoreactive bands for phosphorylated eIF2α (p-eIF2α) and total eIF2α (C), phosphorylated PERK (p-PERK) (D), and the 17-kDa fragment of activated caspase-3 and GGA3 (F) were quantified and expressed as the percentage of vehicle-treated 5XFAD levels (n = 4–7 mice per group). Levels of p-eIF2α, but not those of total eIF2α, were significantly elevated in STZ-treated 5XFAD mice (*p<0.05 vs. vehicle). STZ treatments also dramatically increased p-PERK levels in 5XFAD mice (p = 0.05), while cleaved caspase-3 and GGA3 levels were indistinguishable between STZ- and vehicle-treated subjects. All data are presented as mean ± SEM.

Figure 4. Effects of STZ-induced insulin deficiency on neprilysin and IDE levels in 5XFAD mice.

(A) Western blot analysis of hemibrain lysates from vehicle- and STZ-treated 5XFAD mice. (B) Immunoreactive bands for neprilysin and IDE were quantified and expressed as the percentage of vehicle-treated 5XFAD levels (n = 4–7 mice per group). There was no difference in neprilysin or IDE levels between STZ- and vehicle-treated 5XFAD mice. All data are presented as mean ± SEM.