Therefore, we stimulated cortical neurons with BDNF at 15 days in vitro (DIV) (Figure S2A), a stage when FMRP, CYFIP1, and eIF4E are highly expressed and neurons are mature (Figure S2A). We stimulated neurons with BDNF, which induces translation (Aakalu et al., 2001, Schratt et al., 2004 and Takei et al., 2004) and actin remodeling (Bramham, 2008), and followed the subsequent changes in the colocalization of CYFIP1 with eIF4E or NCKAP1. Stimulation by BDNF significantly reduced the degree of CYFIP1-eIF4E colocalization, and concomitantly increased the number of CYFIP1-NCKAP1 puncta, suggesting that CYFIP1 distribution changes between these complexes upon TrkB receptor

activation (Figures 2A and S2B). The magnitude of these changes is similar to those observed with manipulations that alter interactions of eIF4E with canonical Fluorouracil price eIF4E-BPs (Costa-Mattioli et al., 2009, Richter and Klann, 2009 and Sonenberg and Hinnebusch, 2009). These changes were observed 15 min after BDNF stimulation (Figure S2C). Only a very small proportion of CYFIP1 remained not engaged

within these two complexes (∼15% according to the colocalization data). Consistently, blue native PAGE (BN-PAGE) revealed that the majority, if not all, of CYFIP1 is part of CX5461 high molecular weight complexes (Figure S2D). Based on these data, we infer that a “free” CYFIP1 pool is minor. We then aimed at identifying the factors regulating this equilibrium. A candidate is Rac1, because in its active form (GTP-Rac1), it interacts with CYFIP1 (Kobayashi et al., 1998) and favors WRC activation (Chen et al., 2010, Eden et al., 2002, Schenck et al., 2003 and Steffen over et al., 2004). To test this hypothesis, we used NSC23766,

a specific inhibitor of Rac1 activation (Gao et al., 2004) (Figure S2E). Addition of NSC23766 before BDNF stimulation prevented the redistribution of CYFIP1 (Figure 2A), indicating that active Rac1 is needed for the effect of BDNF on the CYFIP1 complexes. To further monitor the dynamics of CYFIP1 redistribution, we quantified the changes in fluorescence of EYFP-CYFIP1, Cerulean-NCKAP1, and eIF4E-mCherry in spines of BDNF-stimulated primary neurons over time (Figure S3). We observed that the ratio of Cerulean-NCKAP1 over EYFP-CYFIP1 steadily increases, indicating a build-up of WRC (Figure S3C). CYFIP1 redistribution between eIF4E- and NCKAP1-containing complexes was further corroborated by biochemical evidence in isolated synaptoneurosomes: BDNF stimulation increased the amount of CYFIP1 coprecipitating with NCKAP1, and conversely reduced its binding to eIF4E; the Rac1 inhibitor was able to prevent the CYFIP1 redistribution (Figure 2B). To investigate whether active Rac1 directly changes the ability of CYFIP1 to bind eIF4E, we used GTP-Rac1 as exogenous competitor in m7GTP chromatography on cortical lysates.