Fig. 8: ZEB1-dependent vesicular trafficking facilitates the establishment of a front-rear polarity axis.

a, b Correlation of peripheral Tfn signal intensity with metastatic capacity (a) and Golgi area (b) in murine LUAD cell lines fixed 10 min after initiating endocytosis. c Percentages of cells on the wound front with Golgi organelles polarized toward the leading edge. n = 5 independent experiments. d Percentages of cells on the wound front with FAs polarized toward the leading edge. n = 10 independent experiments. e Merged confocal micrographs of siRNA-transfected 344SQ cells stained with anti-phospho-paxillin (p-Pxn) antibody to detect FAs. The direction of cell migration on advancing front (arrow). Scale bars: 20 µm. f, g Plot showing average FA number (f) (n = 25 cells from 3 independent experiments) and size g (n = 23 cells from 3 independent experiments) per cell in the siRNA-transfected 344SQ cells (e). h, i Cell speed and directionality for each cell (dot) measured from automated tracking of 393P_ZEB1 cells seeded at low density on the fibronectin-coated surface and transfected with siRNAs against KIF13A (h), AP1S2 (h), or ITGα5 (i). h n = 180 cells (speed) or 143 cells (directionality) from 3 independent experiments. i n = 250 cells (speed) or 225 cells (directionality) from 3 independent experiments. j, k Migration and invasion assays in Boyden chambers for 344SQ cells transfected with KIF13A (j) or ITGα5 (k) siRNAs. n = 3 independent experiments. l Invasive activities of siRNA-transfected 393P_ZEB1 cells seeded as multicellular aggregates in collagen gels. Invasions scored as single-cell or collectively invasive protrusions. n = 14 aggregates from 3 independent experiments. Scale bars: 100 µm. m Model illustrating that ZEB1 activates KIF13A-dependent endocytic recycling to promote front-rear axis polarity and motility. Data are presented as mean values ± SEM, or as mean values ± SD (j–l); P values, two-tailed Student’s t test, or one-way ANOVA (c, d, f–l). R values, Spearman (a, b).