Supplementary Figure 2: Results showing TAG-1 deficiency–induced shortening of progenitors, contractility of the apical surface and slight retraction of the basal process by laser ablation at the basal surface

a) Stage-dependent disappearance of the progenitor's basal process (number of GFP+ VZ cells examined = 238 cells for E10 1 E12, 102 cells for E11 1 E13, 134 cells for E12 1 E14, and 125 cells for E13 1 E15). TAG-1 shRNA electroporation at E11, E12, and E13 revealed that the phenotype of basal-process disappearance was reproduced frequently in embryos electoroporated at E11, but rarely in those electoroporated at E13; E12-electroporated embryos exhibited an intermediate phenotype. This difference may be explained partially by the observation that TAG-1 distribution shifts from the most basal region at E11–E12 (Fig. 3a) to deeper regions at E13 and beyond (data not shown). Alternatively, progenitors' requirement for TAG-1 and/or the mechanisms of basal process morphogenesis might be regulated in a stage-dependent manner. b) The shortening (basal-disconnection) phenotype was reproduced by shRNA against a diffent TAG-1 sequence. c) The apically accumulated GFP+ cells were positive for Nestin. The overall pattern of the course of Nestin+ fibers was abnormal, and did not show a typical radial arrangement. d) Comparison of cleavage orientation of apically dividing cells between control and TAG-1–KD VZs at E12. e) Comparison of progenitors' morphology between normal and TAG-1–KD VZs. In contrast to the control progenitor cells (left), which have thin flexible apical processes after successful basalward nucleokinesis, TAG-1–deficient progenitor cells that could not send the nucleus sufficiently basally (right) were voluminous at the apical side, with much thicker processes than controls. f) When each cerebral wall (E14) was divided into apical and basal parts, the apical part still bent apically, indicating that apicalward bending of each intact wall is not secondary to expansion of the basal part. Bending/curling of cerebral wall slices occurred not only when slices were made along the ventral-to-dorsal axis (Fig. 4h and this panel), but also when sliced along the anterior-to-posterior axis (data not shown). Slices prepared at E12, E13, or E14 showed similar bending/curling. g) Apicalward bulging (arrow) and tangential expansion of a TAG-1–KD cerebral wall. Magenta is Ki67 immunoreactivity. Length between the dorsal (D) and ventral (V) border of the pallial wall part increased in TAG-1–KD henmispheres. h) Schematic illustration showing morphology and periventricular nuclear density of control and TAG-1-KD (shortened) progenitor cells. i) Laser ablation experiments on the basal surface. Cross-sectional live inspection of a cerebral wall that received a laser pulse onto the basal surface approximately 60 min before slice preparation. Basal processes (arrowhead) were retracted only slightly and cell bodies (arrow) were at normal positions during observation for 12 h.