Fig. 2.

DKK3 is a HSF1 target gene associated with CAF emergence. a Tukey boxplots showing z-score values of Dickkopft genes mRNA expression in murine mammary NFs and CAFs (NF, n = 6; CAF, n = 8). b Western blot showing protein levels of DKK3, DKK2, αSMA and tubulin in total lysates of two sets of murine mammary NFs and PyMT-CAFs. Levels of secreted DKK3 are also shown. c Venn diagram showing the overlap between transcription factors (TFs) known to bind the DKK3 promoter (left) and TFs whose perturbation modulates DKK3 expression levels (right). In red, TFs that negatively affect DKK3 expression; in green, TFs that positively affect DKK3 expression. d Graphs show Hsf1 and Dkk3 fold mRNA expression levels (relative to Rplp1) in murine PyMT-CAFs (CAF1 and CAF5) after transfection with control or Hsf1 siRNAs (smart-pool). Bars show mean value ± SEM (n = 3). e Western blots show Hsf1, Dkk3 and Gapdh expression in murine PyMT-CAFs (CAF1 and CAF5) after transfection with control or Hsf1 siRNAs (smart-pool). (f) Graphs show correlations between DKK3 gene expression and HSF1 activity as measured by the expression of HSF1 gene signature (z-score normalised) in stroma of breast, colorectal and ovarian cancers. Pearson correlation coefficient (r) is shown. Each dot represents z-score values from individual samples. g Graph shows binding of Hsf1 to regulatory elements in the Dkk3 gene (DKK3 P – promoter; DKK3 E – enhancer) in murine NF1 and CAF1. Primers targeting the bona fide Hsf1 target gene Rilpl were used as a positive control. Floating boxes: centre line, mean; box limits, min and max values (normalised to DNA amount)(n = 3). Where indicated, individual p values are shown; alternatively the following symbols were used to describe statistical significance: *P < 0.05; **P < 0.01; ***P < 0.001; #P < 0.0001; n.s., non-significant