Figure 6. Inhibition of PAK3 prevents activity-mediated long-term stabilization of stimulated spines. A, TBS of slice cultures results in a transient enlargement of a fraction of spines that become more stable (arrowhead; scale bar, 1 μm). B, The fraction of spines enlarging as a result of TBS tended to be lower after treatment with IPA-3 (15 μm; n = 4; p = 0.1). C, The long-term stability of spines that enlarged 5 h after TBS (gray circles; n = 4 cell, 27 spines analyzed) is increased compared with non-enlarged spines (black circles; n = 4 cells, 111 spines analyzed; p < 0.001, two-way ANOVA). D, In contrast, the same stimulation protocol applied to IPA-3-treated slices failed to enhance the stability of enlarged (gray circles; n = 4 cells, 23 spines analyzed) versus non-enlarged spines (black circles; n = 4 cells, 153 spines analyzed). E, Dendritic segment of an mRFP-transfected neuron loaded with Fluo-4 AM and imaged at different times after TBS (scale bar, 2 μm). Line scans through the indicated spine (arrow) show an increase in fluorescence intensity of the green calcium signal at the time of stimulation (right panel, black arrow; calibration: 0.5 μm, 0.5 s). F, The fraction of stimulated spines that showed an enlargement 5 h after TBS was decreased in R419X–PAK3-transfected cells versus mRFP neurons (n = 8 cells, 81 and 64 spines analyzed; p < 0.05). G, Under control conditions (mRFP-transfected cells), the 48 h stability of stimulated spines was significantly enhanced versus nonstimulated spines (n = 8 cells, 37 and 44 spines analyzed; p < 0.001). H, In contrast, the stability of stimulated spines in R419X–PAK3-transfected cells was no longer enhanced by stimulation (n = 8 cells, 30 and 34 spines analyzed). *p < 0.05; ***p < 0.001.