Amyloid beta induces Fmr1-dependent translational suppression and hyposynchrony of neural activity via phosphorylation of eIF2α and eEF2

Abstract

Alzheimer's disease (AD) is the most common cause of dementia, with the accumulation of amyloid beta peptide (Aβ) being one of the main causes of the disease. Fragile X mental retardation protein (FMRP), encoded by fragile X mental retardation 1 (Fmr1), is an RNA-binding protein that represses translation of its bound mRNAs or exerts other indirect mechanisms that result in translational suppression. Because the accumulation of Aβ has been shown to cause translational suppression resulting from the elevated cellular stress response, in this study we asked whether and how Fmr1 is involved in Aβ-induced translational regulation. Our data first showed that the application of synthetic Aβ peptide induces the expression of Fmr1 in cultured primary neurons. We followed by showing that Fmr1 is required for Aβ-induced translational suppression, hyposynchrony of neuronal firing activity, and loss of excitatory synapses. Mechanistically, we revealed that Fmr1 functions to repress the expression of phosphatases including protein phosphatase 2A (PP2A) and protein phosphatase 1 (PP1), leading to elevated phosphorylation of eukaryotic initiation factor 2-α (eIF2α) and eukaryotic elongation factor 2 (eEF2), and subsequent translational suppression. Finally, our data suggest that such translational suppression is critical to Aβ-induced hyposynchrony of firing activity, but not the loss of synapses. Altogether, our study uncovers a novel mechanism by which Aβ triggers translational suppression and we reveal the participation of Fmr1 in altered neural plasticity associated with Aβ pathology. Our study may also provide information for a better understanding of Aβ-induced cellular stress responses in AD.

Keywords: Fmr1; PP1; PP2A; eEF2; eIF2α; translation.

Figure 1

Amyloid beta inducesFmr1 andFmr1‐dependent translational suppression. (a) Representative western blots and quantifications of FMRP and GAPDH from WT cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 2, 4, 8, or 24 h at DIV 12−14. (n = 5−6 from three independent cultures). (b) Quantitative real‐time RT‐PCR of Fmr1 mRNA normalized to Actin mRNA from WT cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h at DIV 12−14. (n = 6 from three independent cultures). (c) Representative western blots and quantifications of puromycin and GAPDH from WT and Fmr1 KO cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h at DIV 12−14 (n = 8 from three independent cultures). No data points were removed after the Grubbs’ outlier test. Student's t test was used. Data are represented as mean ± SEM with *p < 0.05, ns, nonsignificant.

Figure 2

Amyloid beta inducesFmr1‐dependent hyposynchrony of neural network activity and reduction of synapse number. (a) Representative raster plots of spontaneous spikes from WT and Fmr1 cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h at DIV 12−14. Quantification of spontaneous spike rate and synchrony index by comparing “after treatment” to “before treatment,” from the same culture was shown on the right. (n = 6−8 independent cultures after removing one culture from spike rate and synchrony index analyses in WT cultures treated with amyloid beta, and one culture from spike rate and synchrony index analyses in Fmr1 cultures treated with scrambled peptide following the Grubbs' outlier test.) (b) Immunocytochemistry showing postsynaptic marker PSD‐95 (red), presynaptic marker synapsin‐I (green), dendritic marker Map2 (blue), and colocalization of PSD‐95 and synapsin‐I from dissociated WT and Fmr1 KO cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h at DIV 12−14. Representative secondary dendrites are displayed and quantification of colocalized synapses as relative percentage of number of synapses was shown. (For WT, data were collected from two independent cultures with n = 10 and 7 cells treated with amyloid beta and n = 11 and 7 cells treated with scrambled peptide. For Fmr1 KO, data were collected from two independent cultures with n = 8 and 9 cells treated with amyloid beta and n = 8 and 9 cells treated with scrambled peptide. One cell in WT cultures treated with amyloid beta was removed following the Grubbs' outlier test.) Student's t test was used. Scale bar: 10 µm. Data are represented as mean ± SEM with *p < 0.05, ns, nonsignificant.

Figure 3

Amyloid beta induces phosphorylation of eIF2α in anFmr1‐dependent manner. (a) Representative western blots and quantifications of eIF2α, p‐eIF2α, PDI, and GAPDH from WT and Fmr1 KO cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h at DIV 12−14. (n = 5−6 independent cultures. No data points were removed after the Grubbs' outlier test.) (b) Representative western blots and quantifications of PP2A‐A, PP2A‐B, PP2A‐C, PP1, and GAPDH from WT and Fmr1 KO cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h at DIV 12−14. (n = 6−9 from three independent cultures. One Fmr1 KO culture treated with amyloid beta was removed from analyses for PP2A‐A, PP2A‐B, and PP1 following the Grubbs' outlier test.) Student's t test was used. Data are represented as mean ± SEM with *p < 0.05; ns, nonsignificant.

Figure 4

PP2A and PP1 differentially regulate Aβ42‐induced eIF2α phosphorylation and translational suppression. (a) Representative western blots and quantifications of puromycin and GAPDH from Fmr1 KO cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h and with vehicle (DMSO) or okadaic acid (2 nM or 100 nM) for the last hour at DIV 12−14. (n = 7−9 independent cultures after removing one culture treated with scrambled peptide + 2 nM okadaic acid following the Grubbs' outlier test.) Two‐way ANOVA with Tukey test were used. (For the left panel, interaction: F _1,35 = 0.724, p = 0.400; drug effect: F _1,35 = 9.624, p = 0.004; peptide effect: F _1,35 = 2.426, p = 0.128. For the right panel, interaction: F _1,28 = 5.969, p = 0.021; drug effect: F _1,28 = 22.689, p = 0.00005; peptide effect: F _1,28 = 0.335, p = 0.567.) (b) Representative western blots and quantifications of eIF2α, p‐eIF2α, and GAPDH from Fmr1 KO cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h and with vehicle (DMSO) or okadaic acid (2 or 100 nM) for the last hour at DIV 12−14. (n = 7−10 independent cultures after removing one culture treated with amyloid beta + 100 nM okadaic acid for the analyses of eIF2α and p‐eIF2α following the Grubbs' outlier test.) Two‐way ANOVA with Tukey test were used. (For OA at 2 nM, the left panel, interaction: F _1,24 = 0.041, p = 0.842; drug effect: F _1,24 = 0.010, p = 0.920; peptide effect: F _1,24 = 3.093, p = 0.091. For OA at 2 nM, the right panel, interaction: F _1,24 = 0.189, p = 0.668; drug effect: F _1,24 = 0.471, p = 0.499; peptide effect: F _1,24 = 1.673, p = 0.208. For OA at 100 nM, the left panel, interaction: F _1,35 = 0.274, p = 0.604; drug effect: F _1,35 = 0.005, p = 0.941; peptide effect: F _1,35 = 0.122, p = 0.729. For OA at 100 nM, the right panel, interaction: F _1,35 = 0.152, p = 0.698; drug effect: F _1,35 = 14.641, p = 0.0005; peptide effect: F _1,35 = 1.467, p = 0.234.) Data are represented as mean ± SEM with *p < 0.05, ns, nonsignificant.

Figure 5

Fmr1 mediates amyloid beta‐induced phosphorylation of eEF2 through PP2A. (a) Representative western blots and quantifications of p‐eEF2, eEF2, and GAPDH from WT and Fmr1 KO cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h at DIV 12−14. (n = 6−7 independent cultures after removing one Fmr1 KO culture treated with amyloid beta following the Grubbs' outlier test.) (b) Representative western blots and quantifications of eEF2, p‐eEF2, and GAPDH from Fmr1 KO cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h and with vehicle (DMSO) or okadaic acid (2 or 100 nM) for the last hour at DIV 12−14. (n = 5−9 independent cultures after removing one culture treated with amyloid beta + DMSO for the analysis of p‐eEF2 following the Grubbs' outlier test.) Student's t test was used in (a) and two‐way ANOVA with Tukey test was used in (b). (For OA at 2 nM, the left panel, interaction: F _1,32 = 0.438, p = 0.513; drug effect: F _1,32 = 0.982, p = 0.329; peptide effect: F _1,32 = 2.437, p = 0.128. For OA at 2 nM, the right panel, interaction: F _1,31 = 2.689, p = 0.111; drug effect: F _1,31 = 6.920, p = 0.013; peptide effect: F _1,31 = 1.430, p = 0.241. For OA at 100 nM, the left panel, interaction: F _1,16 = 1.727, p = 0.207; drug effect: F _1,16 = 0.068, p = 0.797; peptide effect: F _1,16 = 1.766, p = 0.202. For OA at 100 nM, the right panel, interaction: F _1,16 = 1.671, p = 0.215; drug effect: F _1,16 = 12.825, p = 0.003; peptide effect: F _1,16 = 0.057, p = 0.814.) Data are represented as mean ± SEM with *p < 0.05, ns, nonsignificant.

Figure 6

Inhibition of PP2A restores amyloid beta‐induced hyposynchrony of firing activity in Fmr1KO neurons. (a,b) Representative raster plots of spontaneous spikes from WT (a) and Fmr1 KO (b) cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h and with vehicle (DMSO) or okadaic acid (2 nM) for the last hour at DIV 12−14. Quantification of spontaneous spike rate and synchrony index by comparing “after treatment” to “before treatment,” from the same culture was shown on the right. (n = 6−8 independent cultures after removing one Fmr1 KO culture treated with scrambled peptide + DMSO for analyses of spike rate and synchrony index, one Fmr1 KO culture treated with amyloid beta + DMSO for analyses of spike rate, and one Fmr1 KO culture treated with scrambled peptide + okadaic acid for analyses of spike rate following the Grubbs' outlier test.) Two‐way ANOVA with Tukey test were used; for (a), top panel, interaction: F _1,24 = 0.315, p = 0.580; drug effect: F _1,24 = 0.000, p = 0.990; peptide effect: F _1,24 = 22.904, p = 0.00007. For (a), bottom panel, interaction: F _1,24 = 0.147, p = 0.704; drug effect: F _1,24 = 1.222, p = 0.280; peptide effect: F _1,24 = 25.825, p = 0.00003. For (b), top panel, interaction: F _1,28 = 0.188, p = 0.668; drug effect: F _1,28 = 0.131, p = 0.720; peptide effect: F _1,28 = 25.948, p = 0.00002. For (a), bottom panel, interaction: F _1,27 = 3.376, p = 0.077; drug effect: F _1,27 = 5.660, p = 0.025; peptide effect: F _1,27 = 4.037, p = 0.055. Data are represented as mean ± SEM with *p < 0.05; **p < 0.01, ns, nonsignificant.

Figure 7

Inhibition of PP2A does not restore amyloid beta‐induced reduction of number of synapses in Fmr1KO neurons. (a,b) Immunocytochemistry showing postsynaptic marker PSD‐95 (red), presynaptic marker synapsin‐I (green), dendritic marker Map2 (blue), and colocalization of PSD‐95 and synapsin‐I from dissociated WT (a) or Fmr1 KO (b) cortical neuron cultures treated with amyloid beta 1−42 (Aβ; 1 µM) or scrambled Aβ peptide (Ctrl, 1 µM) for 24 h and with vehicle (DMSO) or okadaic acid (2 nM) for the last hour at DIV 12−14. Representative secondary dendrites are displayed and quantification of colocalized synapses as relative percentage of number of synapses was shown. (For WT, data were collected from two independent cultures with n = 10 and 16 cells treated with scrambled peptide, n = 9 and 15 cells treated with s amyloid beta, n = 13 and 14 cells treated with scrambled peptide + okadaic acid, and n = 9 and 16 cells treated with s amyloid beta + okadaic acid. For Fmr1 KO, data were collected from two independent cultures with n = 16 and 14 cells treated with scrambled peptide, n = 13 and 12 cells treated with s amyloid beta, n = 15 and 15 cells treated with scrambled peptide + okadaic acid, and n = 15 and 11 cells treated with s amyloid beta + okadaic acid. One cell in WT cultures treated with scrambled peptide + DMSO and one cell in WT cultures treated with amyloid beta + okadaic acid was removed following the Grubbs' outlier test.) Two‐way ANOVA with Tukey test were used; for (a), interaction: F _1,97 = 0.040, p = 0.842; drug effect: F _1,97 = 1.822, p = 0.180; peptide effect: F _1,97 = 20.995, p = 0.00001. For (b), interaction: F _1,106 = 3.306, p = 0.072; drug effect: F _1,106 = 1.267, p = 0.263; peptide effect: F _1,106 = 0.654, p = 0.421. Data are represented as mean ± SEM with *p < 0.05; **p < 0.01, ns, nonsignificant.

Figure 8

A working model describing the role of Fmr1 in amyloid beta (Aβ) induced‐translational suppression.