Extended Data Fig. 7: COL17A1 regulates dormancy in LGR5+p27+ CSCs.
From: Cell–matrix interface regulates dormancy in human colon cancer stem cells
a. Whole-mount immunostaining of COL17A1 (cyan), and endogenous expression of p27-mVenus (green) and LGR5-tdTomato (red) in CCO32 (left) and CCO7 (right) organoids with the indicated COL17A1 genotypes. b. The expression of LGR5-tdTomato (red) and p27-mVenus (green) reporters before (left) and after COL17A1 KO (centre, right) in CCO12 organoids. c, d. The number (c) and average size (d) of organoid colonies formed from COL17A1 WT or two independent COL17A1 KO clones of CCO32 organoids. Each dot shows one well. Data are shown as mean ± s.e.m. Statistics, two-sided Welch’s t-test. e. Representative confocal images of Ki67 or EdU staining (green) in COL17A1 WT and KO CCO32 organoids (left). The proportion of Ki67 or EdU+ per all cells (mean ± s.e.m.) in COL17A1 WT and KO CCO32 organoids (right). Each dot represents one organoid. From the left, p = 8.2 x 10–11 and 7.0 x 10–4, two-sided Welch’s t-test. f. Representative confocal images of cleaved caspase-3 staining (green) in COL17A1 WT and KO, CCO7 and CCO32 organoids (left). The ratio of cleaved caspase-3-positive area to nuclei area (mean ± s.e.m.) in COL17A1 WT and KO, CCO7 and CCO32 organoids. Each dot shows one organoid. Statistics, two-sided Welch’s t-test. g. The construct of Dox-inducible COL17A1. h. Confirmation of COL17A1 overexpression using capillary-based immunoassay. COL17A1 expression was induced in CCO7 and CCO32 organoids by a 2-day treatment with 500 ng/ml Dox. i. The viability of CCO32 organoids with the indicated genotypes following CPT treatment. The viability was measured using an ATP luminescence assay and shown in relative light unit (RLU). Data are presented as mean ± s.e.m. j. The schedule for treating COL17A1 WT and KO CCO7 xenografts with IRI. k. The growth of vehicle- or IRI-treated COL17A1 WT and KO CCO7 xenografts. Tumour sizes are shown as the relative values compared to the day 0 volume and as mean ± s.e.m. *p < 0.05 in a mixed effect model with repeated measurement, followed by Šidák correction. Thick lines show model-predicted values, and the areas show 95% upper and lower prediction intervals. Scale bars: 50 μm (a, e, f), 100 μm (b)
