DOCK8 controls Cdc42 activation spatially during DC migration in 3D environments. (A) Migration of BW5147α⁻β⁻ cells stably expressing DOCK8, DOCK8 ΔDHR-2, and/or DOCK2 in 3D collagen gels was compared (n = 101-128 per group). Each box plot exhibits the median (central line within each box), the 25th and 75th percentile values (box ends), and the 10th and 90th percentile values (error bars). **P < .01. (B) Effect of DOCK8 deficiency on CCL21-induced activation of Cdc42, Rac1, and RhoA in DCs. (C) Localization of HA-tagged DOCK8 and βPIX in BW5147α⁻β⁻ cells. Concanavalin A (green) and DAPI (blue) were used to stain plasma membrane and nuclei, respectively. Scale bar indicates 5 μm. (D) Subcellular localization of DOCK8 and βPIX in DCs. Membrane (M) and cytosol (C) fractions were analyzed by immunoblot. TCL indicates total cell lysates. (E) Effect of DOCK8 deficiency on localization of activated Cdc42 in DCs migrating in 3D collagen gels. Images were taken every 1 minute and the emission ratio of 527 nm/475 nm (FRET/CFP ratio) was used to represent the FRET efficiency. The FRET/CFP ratios at the plasma membrane (PM) and the cytoplasm (CP) were normalized by dividing by the lowest value in the cells and were compared between Dock8^+/− and Dock8^−/− DCs (n = 10 cells per group). Scale bar indicates 5 μm. Data are expressed as means ± SD. Data are representative of 3 independent experiments (A-D) or are from 10 separate experiments (E).