Figure 7: AXIN1 stabilization normalizes β-catenin and partially restores cell cycle control in RUNX1-depleted cells.

(a) MCF7 cells constitutively expressing shRx1_3′-UTR (shRx1) were treated for 36 h with either 5 μM IWR1 or its dimethyl sulphoxide vehicle followed by western blot analysis of the indicated proteins. MCF7 expressing a nonspecific shRNA (shNS) were analysed as a reference control. (b) Cells as in a were treated as indicated for 6 days and their growth rate was calculated based on MTT assays as in Fig. 2b. *P<0.05 by t-test. (c) AXIN1 and P-β-cat levels were assessed 8 h after the release of MCF7/shRx1_RUNT^dox cells from a G1/S double thymidine block as in Fig. 6h. Dox treatment (to silence RUNX1) initiated along with the release from the first thymidine block and IWR1 treatment (to stabilize AXIN1) initiated 17 h before harvest. (d) MCF7/shRx1_RUNT^dox cells were treated for 72 h with dox (to silence RUNX1) and 2 nM docetaxel for 48 h (to induce mitotic slippage) as in Fig. 6f,g and IWR1 was added for the last 24 h before FACS analysis. Data are mean±s.e.m. (n=3). *P<0.05 by t-test. (e) Working model for the tumour suppressor function of RUNX1 in ER⁺ breast cancer, whereby RUNX1 prevents E2-mediated AXIN1 suppression. Mechanisms linking the RUNX1/AXIN1/β–catenin axis to loss of cell cycle control in RUNX1-deficient ER⁺ mammary epithelial cells remain to be fully elucidated. They entail stimulation of neither LEF/TCF, nor c-MYC, nor CCND1, nor G1/S phase transition, but are associated instead with deregulated mitosis.