β-Catenin and K-RAS synergize to form primitive renal epithelial tumors with features of epithelial Wilms' tumors

Abstract

Wilms' tumor (WT) is the most common childhood renal cancer. Although mutations in known tumor-associated genes (WT1, WTX, and CATNB) occur only in a third of tumors, many tumors show evidence of activated β-catenin-dependent Wnt signaling, but the molecular mechanism by which this occurs is unknown. A key obstacle to understanding the pathogenesis of WT is the paucity of mouse models that recapitulate its features in humans. Herein, we describe a transgenic mouse model of primitive renal epithelial neoplasms that have high penetrance and mimic the epithelial component of human WT. Introduction of a stabilizing β-catenin mutation restricted to the kidney is sufficient to induce primitive renal epithelial tumors; however, when compounded with activation of K-RAS, the mice develop large, bilateral, metastatic, multifocal primitive renal epithelial tumors that have the histologic and staining characteristics of the epithelial component of human WT. These highly malignant tumors have increased activation of the phosphatidylinositol 3-kinase-AKT and extracellular signal-regulated kinase pathways, increased expression of total and nuclear β-catenin, and increased downstream targets of this pathway, such as c-Myc and survivin. Thus, we developed a novel mouse model in which activated K-RAS synergizes with canonical Wnt/β-catenin signaling to form metastatic primitive renal epithelial tumors that mimic the epithelial component of human WT.

Figure 1

An activating mutation of β-catenin restricted to the kidney proximal tubule is sufficient to induce primitive renal epithelial tumors in mice. A: Kaplan-Meier graph showing a modest decrease in overall survival between tamoxifen-treated Cited1-CreER^T2/Catnb^+/lox(ex3) (C-Catnb^Δex3) and control mice (P <; 0.0001). B–D: H&E examination shows small tumors in C-Catnb^Δex3 mice with histologic findings consistent with a renal epithelial neoplasm with embryonal histologic features. Shown are a representative 2× power image of a kidney from a 62-week-old C-Catnb^Δex3 mouse (B) and ×20 (C) and ×40 (D) magnification images. E: Kaplan-Meier graph showing minimal decrease in overall survival between γGT-Cre/Catnb^+/lox(ex3) (G-Catnb^Δex3) and Catnb^+/lox(ex3) control mice (P = 0.043). F–H: H&E examination shows small tumors in G-Catnb^Δex3 mice with histologic findings consistent with a renal epithelial neoplasm with embryonal histologic features. Shown are a representative 2× power image of a kidney from a 28-week-old G-Catnb^Δex3 mouse (F) and ×20 (G) and ×40 (H) magnification images. Scale bars: 50 μm (C and G); 25 μm (D and H).

Figure 2

An activating mutation of Kras and β-catenin restricted to the kidney proximal tubule induces numerous large metastatic primitive renal neoplasms. Kaplan-Meier graphs showing significantly lower overall survival for C-Kras^G12D/Catnb^Δex3 mice compared with Catnb^Δex3 and control animals (A) and for G-Kras^G12D/Catnb^Δex3 mice compared with Kras^G12D, Catnb^Δex3, and control animals (B) (P <; 0.0001 for both). C: A gross picture of the kidneys from a 15-week-old G-Kras^G12D/Catnb^Δex3 mouse. D–H: H&E examination shows high-volume, multifocal tumors in Kras^G12D/Catnb^Δex3 mice consistent with epithelial-dominant WT with metastases to the lung. Shown are a 2× power image of a kidney from a different 15-week-old G-Kras^G12D/Catnb^Δex3 mouse (D); ×10 (E), ×20 (F), and ×40 (G) magnification images of the kidney from a 13-week-old G-Kras^G12D/Catnb^Δex3 mouse; and a ×20 magnification image of the lung from the same 13-week-old G-Kras^G12D/Catnb^Δex3 mouse (H). Scale bars: 100 μm (E); 50 μm (F and H); 25 μm (G).

Figure 3

Kras^G12D/Catnb^Δex3 tumors show staining consistent with the epithelial component of some human WTs. The tumors, marked by an asterisk, from Kras^G12D/Catnb^Δex3 mice were stained for the following well-defined markers of WT: Pax-2 (A), Pax-8 (B), S-100 (C, arrowheads indicate examples of positively staining normal renal tubules), epithelial membrane antigen (EMA) (D, arrowhead indicates normal glomerulus), WT-1 (E, arrowhead indicates normal glomerulus with positively staining podocytes), CD56/neural cell adhesion molecule (NCAM) (F), and SALL4 (G). Scale bars: 50 μm.

Figure 4

Kras^G12D/Catnb^Δex3 tumors exhibit AKT and ERK activation. Kidneys from Kras^G12D/Catnb^Δex3, Kras^G12D, Catnb^Δex3, and control mice were stained for phosphorylated (p) AKT (A), ERK (B), and P38 MAPK (C). Tumors are denoted by asterisks. Scale bars: 50 μm.

Figure 5

Kras^G12D/Catnb^Δex3 tumors have increased levels of total and intranuclear β-catenin, increased levels of c-Myc and survivin, increased markers of cell proliferation, and decreased apoptosis. A: Low-power magnification images (×5) of a kidney from a Kras^G12D/Catnb^Δex3 mouse stained for β-catenin. Tumors are denoted by asterisks. Scale bar = 200 μm. Evidence of nuclear β-catenin (arrows) is seen on a ×40 magnification image (inset). Scale bar = 25 μm. B: Western blot analysis of whole kidney lysates for β-catenin and actin loading controls. Wild-type β-catenin (top band) and Δ exon 3 β-catenin (bottom band) are visualized. C: Western blot analysis of nuclear fractionated protein lysates of the kidney stained for β-catenin and histone (nuclear fraction) loading control. A representative example of individual mice is shown. Kidneys from Catnb^Δex3 (D) and Kras^G12D/Catnb^Δex3 (E) mice were stained for c-Myc. Tumors are denoted by asterisks. Scale bars: 100 μm. F: Western blot analysis of whole kidney protein for c-Myc and survivin and corresponding actin loading controls. A representative example of individual mice is shown. The tumors were stained for the proliferative markers Ki-67 (G) and cyclin D1 (H) and for a marker of apoptosis, cleaved caspase 3 (I). Tumors are denoted by asterisks. Scale bars: 100 μm.