Fig. 4

Cholesterol-enriched domains are more abundant at the ECM-contact side of gelatin-degrading cell lines. A X–Z reconstructions of confocal images of 9 breast cancer cell lines plated on fibronectin-coated coverslips, serum-starved combined or not with 2mM mβCD (chol depletion) for 2 h and labeled at 4 °C with the mCherry-Theta toxin fragment specific to endogenous chol. Green and red arrowheads, chol-enriched domains at ECM-free and ECM-contact cell sides respectively. B, D Quantification of the Theta ECM-free side fluorescence intensity of the 9 cell lines treated (D) or not (B) with mβCD. Each data point represents the average fluorescence intensity value of 1 coverslip (120–464 cells from n = 23–62 images from 3–5 independent experiments). C, E Theta fluorescence intensity at ECM-free side of cell lines treated (E) or not (C) with mβCD and divided in 3 groups based on aggressiveness determined in Additional file 1: Table S2: low (score ranging from 0 to 4), intermediate (int; score ranging from 5 to 8) and high aggressiveness (score ranging from 9 to 13). Each data point represents the mean fluorescence intensity value per cell line. Ordinary one-way ANOVA followed by Tukey post-hoc comparison test (B, D) and one sample t test (C, E). F Theta fluorescence score at the ECM-contact side attributed to the 9 cell lines based on X–Z reconstructions of Theta-labeled cells (see Additional file 1: Materials and methods section). Score ranged from 0 to 2. Higher scores were attributed to cells with higher chol-enriched domains at the ECM-contact side. G Inverse correlation between the Theta fluorescence score at the cell ECM-contact side and oriented invasion upon mβCD