S-SCAM inhibits Axin-dependent synaptic function of GSK3β in a sex-dependent manner

Abstract

S-SCAM/MAGI-2 gene duplication is associated with schizophrenia (SCZ). S-SCAM overexpression in the forebrain induces SCZ-like phenotypes in a transgenic (Tg) mouse model. Interestingly, S-SCAM Tg mice show male-specific impairments in synaptic plasticity and working memory. However, mechanisms underlying the sex-specific deficits remain unknown. Here we report that S-SCAM Tg mice have male-specific deficits in synaptic GSK3β functions, as shown by reduced synaptic protein levels and increased inhibitory phosphorylation of GSK3β. This GSK3β hyper-phosphorylation was associated with increased CaMKII activities. Notably, synaptic levels of Axin1, to which GSK3β binds in competition with S-SCAM, were also reduced in male S-SCAM Tg mice. We demonstrated that Axin-binding is required for the S-SCAM overexpression-induced synaptic GSK3β reduction. Axin stabilization using XAV939 rescued the GSK3β deficits and restored the temporal activation of GSK3β during long-term depression in S-SCAM overexpressing neurons. Interestingly, synaptic Axin2 levels were increased in female S-SCAM Tg mice. Female sex hormone 17β-estradiol increased Axin2 expression and increased synaptic GSK3β levels in S-SCAM overexpressing neurons. These results reveal the role of S-SCAM in controlling Axin-dependent synaptic localization of GSK3β. Moreover, our studies point out the pathological relevance of GSK3β hypofunction found in humans and contribute to understanding the molecular underpinnings of sex differences in SCZ.

Figure 1

Reduced synaptic GSK3β protein levels and hyper-phosphorylation of GSK3β in the forebrain tissues of 3-month-old male S-SCAM Tg mice. (a) Schematic diagram showing the competitive binding (↔) of S-SCAM with GSK3β for the GID domain of Axin. (b–d) GSK3β protein levels in the synaptosomal fraction (P2) and total homogenate (H) of male S-SCAM Tg mice. Representative results (b) and quantification (c, d). n = 5–8 mice per group. N-ethylmaleimide-sensitive factor (NSF) was used as loading control. (e–g) pS9-GSK3β protein levels in the P2 and H fractions. Representative results (e) and quantification of relative pS9-GSK3β levels normalized to GSK3β protein levels (f, g). n = 5–7 mice per group. ***p < 0.001, unpaired t-test with Welch’s correction. n.s., not significant.

Figure 2

Enhanced CaMKII activity, not Akt, is associated with the hyper-phosphorylation of GSK3β in male S-SCAM Tg mice. (a) Representative immunoblots showing pCaMKIIs, CaMKIIα, CaMKIIβ, pAkt, and Akt protein levels in the P2 (left) and H fractions (right) of male S-SCAM Tg mice. (b–c) Quantification of P2 (b) and H data (c). n = 4 mice per group. ***p < 0.001, **p < 0.01, unpaired t-test.

Figure 3

Altered synaptic Axin1 levels are responsible for the reduction of synaptic GSK3β. (a–c) Reduced Axin1 levels in the P2 fraction of male S-SCAM Tg mice. Representative immunoblots (a), quantification of Axin1 (b) and Axin2 protein levels (c). n = 5 mice per group. **p < 0.01, unpaired t-test with Welch’s correction. (d, e) Removal of the GK domain in S-SCAM prevents the loss of synaptic GSK3β in cultured hippocampal neurons. Representative images (d) and quantification of synaptic GSK3β immunofluorescent intensity (e). n = 12 neurons per group. One-way ANOVA, F _(2,33) = 11.62, p < 0.001. Tukey’s multiple comparison test: ***p < 0.001, *p < 0.05. Scale bars represent 5 mm.

Figure 4

Axin stabilization restores synaptic GSK3β levels and its temporal regulation during Chem-LTD. (a–c) Effect of XAV939 (5 mM) on the total protein levels of Axin1 and Axin2 in GFP- or S-SCAM-overexpressing cultured hippocampal neurons. Representative immunoblots (a), and quantification of the data for Axin1 (b) and Axin2 (c). n = 4 per group. *p < 0.05, unpaired t-test. (d, e) Effect of XAV939 on synaptic GSK3β levels in myc-S-SCAM overexpressing neurons. Representative images (d), and quantification of the data (e). n = 15 per group. One-way ANOVA, F _(2,42) = 29.91, p < 0.001. Tukey’s multiple comparison test: ***p < 0.0001, *p < 0.05. (f–g) Effect of XAV939 on the GSK3β activation during Chem-LTD. Representative immunoblots (f) and quantification of the data (g). n = 3 per group. *p < 0.05, unpaired t-test with Welch’s correction.

Figure 5

Sex-specific reduction of synaptic GSK3β is mediated by estradiol-mediated increase of Axin2. (a, b) Female S-SCAM Tg mice have normal synaptosomal GSK3β protein levels. Representative images (a) and quantification of GSK3β levels in the P2 fraction (b). n = 5 per group. (c, d) Increased Axin2 protein levels in the P2 fraction of female S-SCAM Tg mice. Representative images (c) and quantification of Axin1 and Axin2 levels in the P2 fraction (d). n = 6 per group. ***p < 0.0001, unpaired t-test with Welch’s correction. (e–g) E2 (10 nM) increases Axin2 protein levels in cultured hippocampal neurons. Representative images (e) and quantification of Axin1 (f) and Axin2 levels (g). n = 4 per group. One-way ANOVA, F _(3,12) = 7.856, p < 0.01. Tukey’s multiple comparison test: *p < 0.05. (h, i) Effect of E2 on synaptic GSK3β levels in myc-S-SCAM overexpressing neurons. Representative images (h) and quantification of the data (i). n = 15 per group. ***p < 0.001, unpaired t-test with Welch’s correction.

Figure 6

Total and synaptic β-catenin protein levels in S-SCAM male Tg mice. (a) Representative immunoblots showing β-catenin and GAPDH protein levels in the P2 (left) and H fractions (right) of male S-SCAM Tg mice. (b, c) Quantification of P2 (b) and H data (c). n = 4 mice per group.