Fig. 2
Inhibition of EZH2 reduces CC-IC frequency in vitro. a Tissue microarray samples processed for immunohistochemistry staining of EZH2 and H3K27me3. Pictures shown are representative of 32 patients (normal) and 251 patients (tumour). Scale bars are 200 µm (10 × ) and 50 µm (40 × ). b Intensity of IHC staining performed in (a) was scored manually as described in the Methods section, and plotted as median ± min and max, n = 32 (normal), n = 251 (tumour), Student’s t test. c Frequency of disease recurrence in patients within the top and bottom quintiles of EZH2 IHC expression as determined in (a, b) (EZH2-low IHC score < 2.4; EZH2-High IHC score > 9.6) (Chi-square test). d, e Pearson’s correlation value for EZH2 mRNA levels from TCGA CRC patient samples (COAD), compared with those of 257 genes from a colon crypt (stem cell) gene expression signature (d), and 388 genes from a colon top (differentiated) expression signature (e). Inverse correlations are plotted in red and positive correlations in blue. f TCF/LEF GFP Wnt reporter activity in POP92 and POP66 after 7 days of UNC1999 or UNC2400 treatment. Data are plotted normalised to DMSO control to the 10% brightest Wnt-High population (n = 4, ± SEM, two-way ANOVA). g–j Limiting dilution assay performed in vitro on UNC1999-pretreated cells at 3 µM for 7 days. Data shown are n = 4 (POP66, (h)), n = 6 (LS174T, (i)), n = 6 (POP92, (j)). Data are shown as mean ± 95% confidence interval, frequency and probability were computed using ELDA software) *P < 0.05, **P < 0.01, ***P < 0.001. Source data are provided as a Source Data file
