To study the contribution of the two CYFIP1 complexes on ARC synthesis

and actin cytoskeleton at synapses, primary cortical neurons (DIV9) were transfected selleck compound with scrambled or Cyfip1 (sh315) shRNA, in combination with CYFIP1 WT, mutant H (affecting actin polymerization), or mutant E (affecting mRNA translation). ARC and F-actin were detected by immunolabeling in neurons at DIV14 with or without BDNF treatment, and the immunosignal was quantified in spines outlined by the membrane-targeted farnesylated GFP (F-GFP) carried by the shRNA construct. We found that CYFIP1 downregulation caused augmented ARC synthesis and reduced F-actin levels in spines ( Figure 4B). Moreover, ARC and F-actin were enhanced after BDNF treatment, but not in Cyfip1-silenced neurons ( Figure 4B). Cotransfection SB431542 purchase of the construct carrying CYFIP1 WT rescued all defects, both basal and BDNF-induced. As predicted, basal and inducible ARC expression was restored by mutant H, but not by mutant E. F-actin levels, in contrast, were rescued by mutant E, but not by mutant H. The fact that mutant E rescued F-actin expression but remains insensitive to BDNF stimulation ( Figure 4B) might

suggest that this pathway requires local translation in addition to WRC activation. In conclusion, the data demonstrate that the CYFIP1 mutants are valuable in separating the two functions found of CYFIP1 in the regulation of local protein translation and the control of actin cytoskeleton at synapses. Alterations in factors controlling protein synthesis (e.g., FMRP) or actin remodeling (e.g., WAVE1) cause dendritic spine defects (Irwin et al., 2001 and Kim et al., 2006). Therefore, we addressed the question of whether CYFIP1 plays a role in dendritic spine formation by studying mice deficient in Cyfip1. Brain slices were isolated from Cyfip1+/– and WT littermates, and individual neurons were labeled diolistically. Dendritic

spines were measured and assigned to four morphological classes, namely mature (stubby and mushroom) and immature types (long thin and filopodia). Neurons displayed a spine distribution in agreement with previous ex vivo studies ( Galvez and Greenough, 2005 and Irwin et al., 2002). Neurons from Cyfip1+/– mice, despite the mild reduction in CYFIP1, showed an increased population of filopodia ( Figures 5A–5C), but no defects in spine density and head width (data not shown). To reduce CYFIP1 expression more drastically, primary cortical neurons (DIV9) were silenced for Cyfip1 (sh319 and sh315, or scrambled shRNA), and spine density and morphology were examined at DIV14. Neuronal morphology was outlined by a farnesylated GFP (F-GFP) carried by the shRNA construct ( Figures 5A–5D and S5E). Spines were classified as above.