Fig. 5. Co-transfection with constitutively active PAK3, but not wild-type PAK3 rescues the phenotype induced by ARHGEF6 knockdown. (A) Illustration of dendritic segments obtained from pyramidal neurons transfected with EGFP (Ctrl), a silencing siR1 construct (siR1), a siR1 construct together with constitutively active PAK3 (siR1 + caPAK3), or a siR1 construct together with wild-type PAK3 (siR1 + wtPAK3). Note the reversal of the ARHGEF6 phenotype by caPAK3, but not wtPAK3. Bar, 3 µm. (B) Blind quantitative analysis of spine parameters show that the decrease in large mushroom-type spines induced by siR1 transfection (dashed bar) is reversed by co-transfection with constitutively active PAK3 (black bar), but not with wild-type PAK3 (grey bar). Data are mean ± s.e.m., n=8-20; ^*P<0.05). (C,D) Same as in B, but with regard to proportion of filopodia and elongated spines (1.8 µm; n=8-20; ^*P<0.05). (E) Cumulative distribution plot of the spine lengths and spine head diameters measured under control conditions (open circles), in siR1 transfected cells (filled circles), in cells cotransfected with siR1 and constitutively active PAK3 (open squares) and in cells co-transfected with siR1 and wild-type PAK3 (filled squares).