Similarly, loss of FMRP increased ARC basal expression

( Figure 3C). Furthermore, ARC synthesis triggered by BDNF was much lower in Fmr1 KO neurons compared with wild-type (WT); inhibition of Rac1 activation before BDNF stimulation blocked ARC synthesis in WT as well as the residual synthesis in Fmr1 KO neurons, whereas no effect was observed in Cyfip1-silenced neurons ( Figure 3C). Fmr1 KO neurons silenced for Cyfip1 phenocopied CYFIP1-deficient neurons, further confirming that FMRP and CYFIP1 act in the same pathway ( Figure 3C). We also investigated ARC levels in mice where CYFIP1 expression was genetically reduced. Because click here Cyfip1 KO animals are embryonic-lethal (our observation and Bozdagi et al., 2012), we used heterozygous animals where CYFIP1 levels are reduced by 40% ( Figure 3D). We examined ARC expression in both total brain cortex and cortical synaptoneurosomes and found that Cyfip1+/− mice have elevated ARC levels at synapses ( Figure 3D). These data support the hypothesis that FMRP and CYFIP1 regulate protein synthesis downstream of Rac1 learn more activation. Activated Rac1 reshapes the CYFIP1-eIF4E complex through a conformational change, so that when translation inhibition is lifted, more CYFIP1 becomes

available for the WRC. Our results suggest that CYFIP1 complexes have a specific function in synaptic protein synthesis and actin polymerization. As proof of principle, we aimed at uncoupling the two complexes and studying their contribution to protein translation and actin polymerization. For this purpose, we designed specific CYFIP1 mutants impairing the interactions with either eIF4E or NCKAP1. To reduce the CYFIP1-eIF4E interaction, we used a mutant replacing Lys743 with a Glu (mutant E), which has been shown Rutecarpine to reduce the interaction with eIF4E (Napoli et al., 2008). To interfere with the CYFIP1-NCKAP1 complex, we studied the large surface of interaction between the two proteins (Chen et al., 2010), and found two hydrophobic patches on CYFIP1 that fit to corresponding sites on NCKAP1

(Figure S5B). The second patch shows a higher complementarity to NCKAP1, in particular in a stretch of eight consecutive hydrophobic amino acids (Ala1003–Ile1010), which was predicted as an essential binding site for NCKAP1. We therefore designed two mutants: mutant Δ, lacking the C-terminal domain that harbors the hydrophobic patch (aa 922–1251), and mutant H, in which the eight hydrophobic residues were replaced by glycines. WT and mutant proteins tagged with the yellow fluorescent protein (EYFP) were expressed in HEK293T cells (Figure S5C) and displayed correct cytoplasmic localization (data not shown). To promote the incorporation of the exogenous proteins into functional complexes, we silenced the endogenous Cyfip1 with siRNAs directed against its 3′UTR ( Figures 4A and S5D).