Interplay between cyclin-dependent kinase 5 and glycogen synthase kinase 3 beta mediated by neuregulin signaling leads to differential effects on tau phosphorylation and amyloid precursor protein processing

Abstract

Cyclin-dependent kinase 5 (cdk5) and glycogen synthase kinase 3beta (GSK3beta) have been implicated in pathogenic processes associated with Alzheimer's disease because both kinases regulate tau hyperphosphorylation and enhance amyloid precursor protein (APP) processing leading to an increase in amyloid beta (Abeta) production. Here we show that young p25 overexpressing mice have enhanced cdk5 activity but reduced GSK3beta activity attributable to phosphorylation at the inhibitory GSK3beta-serine 9 (GSK3beta-S9) site. Phosphorylation at this site was mediated by enhanced activity of the neuregulin receptor complex, ErbB, and activation of the downstream phosphatidylinositol 3 kinase/Akt pathway. Young p25 mice had elevated Abeta peptide levels, but phospho-tau levels were decreased overall. Thus, cdk5 appears to play a dominant role in the regulation of amyloidogenic APP processing, whereas GSK3beta plays a dominant role in overall tau phosphorylation. In older mice, GSK3beta inhibitory phosphorylation at S9 was reduced relative to young mice. Abeta peptide levels were still elevated but phospho-tau levels were either unchanged or showed a trend to increase, suggesting that GSK3beta activity increases with aging. Inhibition of cdk5 by a specific inhibitor reduced cdk5 activity in p25 mice, leading to reduced Abeta production in both young and old mice. However, in young mice, cdk5 inhibition reversed GSK3beta inhibition, leading to an increase in overall tau phosphorylation. When cdk5 inhibitor was administered to very old, nontransgenic mice, inhibition of cdk5 reduced Abeta levels, and phospho-tau levels showed a trend to increase. Thus, cdk5 inhibitors may not be effective in targeting tau phosphorylation in the elderly.

Figure 1.

p25/cdk5 activation inhibits GSK3β through the NRG receptor pathway. A, Representative immunoblots of cdk5, p35/p25, pGSK3β–S9, GSK3β, β-catenin, and pβ-catenin on brain extracts from Ntg (n = 7) and p25 (n = 7) mice at postnatal day 4. pGSK3β–S9 levels normalized to GSK3β total protein were significantly increased in p25 mice compared with Ntg mice (p < 0.01). Four different sets of animals were examined with similar results; the most representative set is shown. p25 mice showed significantly increased levels of β-catenin and significantly decreased levels of the GSK3β target pβ-catenin. B, Representative immunoblots showing levels of neuregulin signaling components, pAkt-S473, Akt, pPI3K–Y199 (p55 subunit), and pErbB2–Y877. β-Tubulin was used for loading control; there was a significant increase of pAkt–S473 (p < 0.05), pPI3K–Y199 (p < 0.05), and pErbB2–Y877(p < 0.05) in p25 mice compared with Ntg mice. C, Correlation analysis of pErbB2 and pGSK3β–S9 in Ntg and p25 mice showed a significant correlation between pErbB2–Y877 and pGSK3β–S9 (p < 0.01).

Figure 2.

Inhibition of ErbB receptor activation reverses GSK3β inhibition in vitro and in vivo. A, Representative blots showing the dose–response effects of the ErbB inhibitor CI-1033 in blocking EGF-induced activity of ErbB in SH-SY5Y cells. EGF-induced phosphorylation of the inhibitory S9 site on GSK3β and the activation site S473 on Akt. Administration of CI-1033 reversed the effects of EGF at doses above 10 nm. B, Representative blots showing the effect of CI-1033 (I) compared with vehicle (V) in nontransgenic mice. C, Representative blots showing that p25 overexpression increased the level of pGSK3β–S9 and pAkt–S473 levels in 5-d-old mice (Ntg/Veh vs p25/Veh) and the ErbB inhibitor CI-1033 reversed the effects. Quantitative analysis of densitometry data are shown on the right.

Figure 3.

Analysis of ErbB receptor and PI3K/Akt pathway in cdk5-deficient mice. Representative immunoblots showing GSK3β, pGSKβ3–S9, Akt, pAkt–Ser473, pPI3K–Y199 (p55 subunit), and pErbB2–Y877 in brain extracts from fetal WT and cdk5-deficient (KO) mice. pGSKβ3–S9, pAkt–Ser473, pPI3K–Y199, and pErbB2–Y877 levels were all reduced in the cdk5 KO mice.

Figure 4.

Phosphorylation of tau in p25 mice. A, Representative immunoblots for total tau and tau pS231, pT181 (AT270), pS202 (CP13), and pS396/S404 (PHF-1). The level of phosphorylation at these sites did not change significantly between Ntg and p25 mice as they are targeted by both cdk5 and GSK3β. B, Representative immunoblots for pS235 (MC6) (p < 0.01) and pS202/pT205 (AT8) (p < 0.01). Phosphorylation at these sites was significantly increased, and they are thus considered primarily cdk5 responsive. C, Representative immunoblots for pS199 and Tau1 (dephosphorylated tau). Phosphorylation at the pS199 site was significantly decreased (p < 0.01), and it is thus considered GSK3β responsive. Results with Tau1 showed that, overall, tau in p25 mice was dephosphorylated compared with tau in Ntg mice.

Figure 5.

Effect of p25/cdk5 activation on APP processing and Aβ production in vivo. A, APP processing and Aβ production were examined in Ntg and p25 mice. In p25 mice, there was a significant increase in BACE levels (p < 0.05), accompanied by significantly increased β processing indicated by enhanced sAPPβ (p < 0.01) and increased level of APP C99 (β-CTF). FL, Full length. B, ELISA analysis of soluble Aβ levels in Ntg and p25 mice. There was a significant increase in Aβ₄₀ and Aβ₄₂ in p25 mice compared with Ntg mice.

Figure 6.

Effect of p25/cdk5 activation on GSK3β regulation and tau phosphorylation in adult mice. A, Representative blots and statistical analysis showing GSK3β, GSK3β–S9, and ErbB signaling components including pAktS473, total Akt, pPI3KY199, pErbB2Y877, and p35/p25 in Ntg (n = 5) and p25 (n = 6) mice at 3 months of age. B, Representative blots showing phospho-tau at sites pS396/pS404 (PHF-1), pS199, dephosphorylated tau (Tau1), total tau, and actin levels (loading control). Although there was a trend to decrease for S199, overall levels of tau phosphorylation level indicated by Tau1 showed no change at this age. C, Comparison between the GSK3β–S9 levels for 5-d-old and 3-month-old mice (3 mice each shown). Inhibitory phosphorylation at the S9 site was significantly reduced in 3-month-old mice.

Figure 7.

Effect of cdk5 inhibitor CP-681301 on GSK3β regulation and tau phosphorylation in neonatal p25 mice. A, Representative blots showing the effect of CP-681301 on GSK3β, GSK3β–S9, and tau phosphorylated at sites pS199 and pS235, relative to total tau, and dephosphorylated tau (Tau1) in neonatal Ntg mice treated with vehicle (Ntg/Veh; n = 7), p25 mice treated with vehicle (p25/Veh; n = 7), and p25 mice treated with CP-681301 (p25/CP; n = 6). Actin was used as a loading control for this study. B, Statistical analysis of GSK3β–S9 phosphorylation, pS199, and total dephosphorylated tau (Tau1) in Ntg mice treated with vehicle (Ntg/Veh), p25 mice treated with vehicle (p25/Veh), and p25 mice treated with cdk5 inhibitor CP-681301 (p25/CP). The cdk5 inhibitor reversed the effects of overactive cdk5 in the p25 mice, leading to reduced inhibitory phosphorylation of GSK3β and increased phosphorylation of tau at relevant sites.

Figure 8.

Effect of cdk5 inhibitor CP-681301 on GSK3β regulation, tau phosphorylation, and Aβ production in vivo in aged mice. A, Representative blots showing the effect of CP-681301 treatment on GSK3β, GSK3β–S9, dephosphorylated tau (Tau1), phosphorylated tau at sites pS199 and pS235, and actin in Ntg mice treated with vehicle (Veh; n = 6) or with CP-681301 (CP; n = 6). B, Analysis of soluble Aβ_1–40 and Aβ_1–42 levels in Ntg mice treated with vehicle or with cdk5 inhibitor CP-681301. Mice used in this study were between 24 and 28 months of age.

Figure 9.

Schematic illustration of cdk5 and GSK3β interaction and its effects on AD-related pathways including tau phosphorylation and Aβ production.