Decrease in GLI1 expression inhibits cell migration and anchorage-independent growth. (A) Panel I: Cartoon depicting experimental setup. Cells (yellow) were plated in the top of a Boyden chamber in 1% FBS above a lower-chamber of media containing 10% FBS. After 16 hours the number of cells that have migrated to the bottom side of the filter were stained and counted. II: Light microscopy images of the bottom side of the migration filter after staining. BT549 cells were infected with the indicated retroviruses and selected prior to experimentation. Scale bar indicates 100 μm. III: Migration of BT549 and MDA.MB.436 cells infected with the indicated retroviruses and assayed as described in I. (B) Clonogenic colony formation assay in BT549 (I) and MDA.MB.436 (II) cells. (C) Panel I: Representative 12 day spheres formed in MDA.MB.436 cells infected with the indicated retroviruses. Scale bar indicates 100 μm. II: Relative sphere formation by MDA.MB.436 cells. (D) Panel I: GLI1 mRNA expression obtained from real-time RT-PCR analysis of mRNA from adherent cells incubated in mammosphere medium (white bars) or sphere cells grown under non-adherent conditions (gray bars). II: Western blot detailing GLI1 levels in HMLE-shEcad or HMLE-Snail adherent and sphere cells. For all: * = P ≤0.05, t test. ** = P ≤0.005, t test. (E) (I) Plot of tumor volume over time arising from orthotopic injection of MDA.MB.436 cells infected with either non-targeting (NT) or GLI1 knockdown constructs. n = six animals. (II) Final averaged tumor weights at seven weeks post-injection for n = six animals. (III) Final averaged tumor weights from an independent biological replicate of the orthotopic xenograft study, conducted on four animals. FBS, fetal bovine serum; GLI1, glioma-associated oncogene 1.