Fig. 4

Sc polarity correlates with metastatic potential. a GFP+ CTCs isolated from MDA-MB-231-GFP tumour-bearing mice (Supplementary Fig. 6), quantified (as the percentage of circulating PBMCs) and analysed for sc polarity by phalloidin staining and imaging of actin polarisation. Lung tumour burden was assessed by the number of GFP+ cells/total lung cells. Number of CTCs is plotted against number of tumour cells in the lung (left) or fraction of polarised CTCs against logarithmised metastatic capacity (tumour cells in the lung normalised by tumour cells in the blood, middle) (Pearson’s correlation). Lower panel: representative density plots of mice with the lowest (left) and highest (right) tumour burden in the lung, showing 8000 events gated as intact and viable single cells by DAPI exclusion. b Fraction of polarised cells in suspension measured in parental A375P and the more metastatic daughter cell line A375-M2 expressing ezrin-GFP (n = 6, mean ± SD, paired t-test). c Fraction of polarised cells in suspension measured in parental TD and the more metastatic daughter cell lines P1 and P2 expressing ezrin-GFP (n = 6, mean ± SD, paired t-tests). d Fraction of polarised cells in suspension measured in parental B16-F0 and the more metastatic daughter cell lines B16-F1, B16-F10 and B16-BL6 expressing ezrin-GFP (n = 6, mean ± SD, paired t-tests). e Patient-derived ccRCC cell lines from M0 or M1 stage at early passages analysed in suspension for sc polarity by staining and imaging of actin polarisation. (n = 3, mean ± SD, repeated-measures ANOVA). f Patient-derived melanoma cell lines isolated either from primary melanoma (left) or metastases (right) analysed in suspension (spotted on a TMA) for fraction of polarised cells by staining and imaging of actin. Samples containing ≥3 polarised melanoma cells in a spot were graded as polarised (grey); samples containing 0–2 polarised melanoma cells were graded as unpolarised (white) (n = 13 for primary melanoma, n = 27 for metastases, Fisher’s exact test)