Directed migration of dendritic cells toward a CCL19 gradient in 3D collagen gels. (A) Three-dimensional in vitro setup to study DC tissue invasion (in panels B,C). DCs were layered on top of a polymerized 1.6 mg/mL collagen gel containing 2.5 μg/mL CCL19 and incubated at 37°C, 5% CO₂. After 8 hours, DC invasion into the gel was analyzed. (B) Images of WT and Cdc42^−/− DC invasion were obtained by bright-field microscopy (inverted Axiovert 40; Zeiss). Scale bar represents 200 μm. Objective: A-Plan 10×/0.25 Ph1 (Zeiss). (C) Cells were derived from 2 independent BM DC cultures (of both WT and Cdc42^−/− mice). Single DC cultures were applied on top of 6 collagen gels and the distance of DC migration into the gel was measured (mean ± SD, t test, P < .001, WT: n = 12, Cdc42^−/−: n = 12). ***P < .001. (D) Three-dimensional in vitro setup to study DC interstitial migration. DCs were added to 1.6 mg/mL collagen, followed by fiber assembly for 30 minutes at 37°C. Polymerized gels were overlaid with 0.6 μg/mL CCL19. DC migration at 37°C, 5% CO₂ was recorded for 4 hours by time-lapse videomicroscopy and analyzed with ImageJ software (in panels E,F). (E) Tracks of single DCs migrating toward a CCL19 gradient in 3D (n = 40 each). (F) Comparison of velocities (line: median, U test, P < .001) and directionality (line: mean, t test, P < .001) of DCs chemotaxing in 3D collagen matrices. Cells were derived from 2 independent BM DC cultures (of both WT and Cdc42^−/− mice) and applied to individual experiments. Fifty single cells (dots) per experiment were tracked (WT: n = 100, Cdc42^−/−: n = 100). ***P < .001.