Fig. 4: Activated MAPK signalling antagonizes zone 3 differentiation enabling WNT-driven cancer and transformation of Lgr5+ zone 3 hepatocytes.
From: Hepatic zonation determines tumorigenic potential of mutant β-catenin
a, Central vein Lgr5-specific model of WNT and BrafV600E activation. i.p., intraperitoneal. b–d, Representative macroscopic images from Lgr5CreER;Ctnnb1ex3/WT;R26LSL-MYC (n = 10; b), Gls2Cre-ER;Ctnnb1ex3/WT;R26LSL-MYC (n = 5; c), Lgr5CreER;BrafV600E/+ (n = 7; left in d) and Lgr5CreER;BrafV600E/+;Rnf43fl/fl;Znrf3fl/fl (n = 15; right in d) mouse livers. Scale bars, 1 cm. e, Representative images of CDH1 and GLUL IHC in Lgr5CreER;BrafV600E liver tumours (n = 4). Scale bars, 100 μm. f, Day 10 whole-liver RNA sequencing and liver-zonation GSEA. For biological replicates, n = 9 for A, n = 8 for Apcfl/fl;R26LSL-MYC, n = 5 for Br, n = 3 for Ctnnb1ex3/ex3, n = 3 for B, n = 5 for BM, n = 5 for BrRZ, n = 5 for M and n = 5 for RZ. The schematic of the lobule levels was created in BioRender. Raven, A. (2025) (https://BioRender.com/pv8v7yw). g, Schematic recombining WNT-mutant and Braf-mutant alleles in the hepatocyte epithelium. h, Tumour-free survival. For biological replicates, n = 17 for BrB, n = 15 for BrRZ, n = 12 for B, n = 9 for Br and n = 9 for RZ. A log-rank (Mantel–Cox) test was used. See Methods for the censoring criteria (censors are denoted by the vertical tick marks). i,j, LW:BW ratio (i; biological replicates: n = 15 for BrB, n = 15 for BrRZ, n = 10 for B, n = 7 for Br and n = 6 for RZ) and tumour scoring (j; biological replicates: n = 9 for BrB, n = 9 for BrRZ, n = 11 for B, n = 6 for Br and n = 6 for RZ) in mouse models of sporadic WNT-driven tumorigenesis. A one-way Kruskal–Wallis test and Dunn’s multiple comparisons test were used to determine significance. The bars are mean ± s.d. B and RZ samples are plotted in Fig. 1b,c and Extended Data Fig. 1c. k,l, Schematic describing LGK974 (k) or dabrafenib (l) treatment from day 30 in the BrRZ tumour model. The illustrations of the mouse and adenovirus in panels a,g,k,l were adapted from Medical Art Servier (https://servier.com) under a CC BY 4.0 licence. m–p, Tumour-free survival (m,o; log-rank (Mantel–Cox) test; see Methods for the censoring criteria (censors are denoted by the vertical tick marks)), tumour scoring (right in n,p) and LW:BW ratio (left in n,p; two-tailed Student’s t-test and two-tailed Mann–Whitney test) from BrRZ mice treated with either LGK974 (PORCN inhibitor; m,n) or dabrafenib (BRAF inhibitor; o,p). For biological replicates for the PORCN inhibitor experiments: for the LW:BW ratio, n = 11 for vehicle and n = 12 for LGK974; and for tumour number, n = 6 for vehicle and n = 9 for LGK974. For biological replicates for BRAF inhibitor experiments, for the LW:BW ratio, n = 11 for vehicle and n = 11 for dabrafenib; and for tumour number, n = 8 for vehicle and n = 12 for dabrafenib. The bars are mean ± s.d.
