1. Mouse mutants of the Wnt signaling components Pygo and Bcl9

Bcl9 and Pygopus (Pygo) were discovered in Drosophila as essential components of the Wnt canonical signaling pathway. Bcl9 binds to ß‑Catenin and to Pygo and tethers Pygo to the ß‑Catenin‑Tcf activation complex (Kramps et al., Cell 109, 47-60 (2002)). To elucidate the role of Bcl9 and Pygo during embryonic development and adult tissue homeostasis in the mouse, we generated loxP-flanked alleles of the four relevant genes (Bcl9, Bcl9l, Pygo1, Pygo2) and derived several conditionally mutant strains.

Our recent work was dedicated to characterizing the developmental anomalies observed in Pygo1/2 mutant embryos, to investigating the regenerative deficiency observed when Bcl9/Bcl9l genes were inactivated in the adult gastrointestinal epithelium, and to exploring the role of Bcl9/Bcl9l and Pygo1/2 in colon tumorigenesis.

Role of Pygo in heart development and lens induction

Ablating both Pygo orthologs in the mouse caused abnormalities including a striking defect in lens development, and a deficiency in cardiac neural crest cell expansion that resulted in lethal cardiac outflow tract and valve anomalies. Surprisingly, compound Pygo1/2 mutant embryos presented no anomalies in early Wnt mediated developmental processes. Accordingly, expression of known Wnt target genes was not significantly affected. While it cannot be ruled out that Pygo proteins may be required for discrete spatial and/or temporal activation of Wnt-regulated genes, their role as obligatory transcriptional co-activators of Wnt target genes in Drosophila does not appear to be conserved in the mouse.

Role of Bcl9/Bcl9l in homeostasis and regeneration of adult gastrointestinal epithelium

Canonical Wnt signaling through ß‑catenin/Tcf complexes is known to regulate development and differentiation of the gastrointestinal epithelium. Locally confined Wnt signals regulate and compartmentalize cell proliferation and differentiation along the crypt villus axis. Bcl9 and Bcl9l are expressed throughout the gastrointestinal epithelium. Induced deletion of Bcl9 and Bcl9l in the adult gastrointestinal epithelium proved efficient and stable, indicating that deletion occurred in the stem cell compartment. Unexpectedly, ablation of these genes did not result in any detectable anomaly and had no effect on cell proliferation and lineage determination. Analysis of transcriptional profiles in micro-dissected epithelium from wild-type and Bcl9/Bcl9l mutant mice indicated that the expression levels of known Wnt-regulated target genes was largely unaffected. Mice with inactivated Bcl9/Bcl9l in their gastrointestinal epithelium proved highly susceptible to the irritant dextrane sulfate sodium (DSS) and displayed epithelial lesions that were substantially more extended as compared to wild-type mice (see Figure 1). We observed that epithelial cells at the borders of ulcers caused by DSS treatment have up-regulated expression of Bcl9/Bcl9l as assessed by in situ hybridization. These cells also stain for nuclear ß‑catenin and Axin-2, which is indicative of activated canonical Wnt signaling. While it remains uncertain whether the critical role Bcl9/Bcl9l appear to play in this wound healing process is related to Wnt signaling, we currently explore to what extent cell migration and epithelial-mesenchymal transition may be affected in the absence of Bcl9/Bcl9l.

Figure 1: Regeneration deficiency in Bcl9/Bcl9l mutant mice exposed to the gastrointestinal irritant DSS. HE stained sections of colon epithelium of A) wt and B) compound Bcl9/Bcl9l mutant mice 8 days after recovery from DSS treatment.

Role of Bcl9 and Pygo proteins in tumor development

Constitutive activation of Wnt/ß‑catenin signaling is an initiating event in the development of colorectal carcinomas in humans. To assess the role of Pygo and Bcl9 proteins in colorectal tumorigenesis, we utilized a tumor model in which colon tumors are induced in mice through exposure to the mutagen N,N'-dimethylhydrazine (DMH) and subsequent challenge with DSS. The tumors were analyzed with regards to elevated Wnt signaling activity and correlation with expression of Bcl9/Bcl9l and Pygo1/2. We observed no difference in colon tumor development between wild-type and Bcl9/Bcl9l or Pygo1/2 deleted animals. Wnt signaling, as assessed by expression level of the Wnt target gene Axin2, was activated in colon tumors of Bcl9/Bcl9l or Pygo1/2 deleted mice. These results indicate that neither Bcl9/Bcl9l nor Pygo1/2 seem to be essential for tumorigenesis in this colon cancer model.

2. Exploring the microenvironment of human colon cancer

There is increasing evidence that tumor cells actively recruit stromal cells, such as inflammatory cells, vascular cells, and fibroblasts, into the tumor. The resulting microenvironment seems to play an important role in tumor proliferation and progression. Although many factors have been identified as stromal signals, the intracellular cross-talk is still only partly understood. In recent years, various groups have focused mainly on prostate and breast cancers. The major objectives of this project are to identify genes that are differentially expressed within the reactive stroma in human colon carcinoma, and the establishment of appropriate in vitro co-culture systems and in vivo models (genetically engineered mouse and/or xenograft  models) to validate the functional relevance of candidate genes.

Using laser capture microdissection of colon carcinoma tissue from freshly operated patients, we have established a database of genes differentially expressed in differentiated tumor epithelium versus invasive tumor areas, and from quiescent versus reactive tumor stroma. Using cell-based assays, we aim to evaluate the role of these newly discovered factors in processes that are relevant to tumor progression, including notably cell motility and invasion.

We are in the process of establishing mouse xenograft platform to characterize the function of candidate genes in vivo and notably monitor tumor growth, invasion and dissemination. We make use of a two-vector lentiviral system for doxycycline-regulated induction of any shRNA of interest in colon carcinoma cell lines (Wiznerowicz and Trono, J. Virol. 77, 8957–8961, (2003)). Transduced cells are marked so that, upon orthotopic implantation, tumors can be monitored by non-invasive, luciferase-based imaging and areas expressing the relevant shRNA readily identified by immunohistochemistry. The approach was validated in proof of concept experiments, whereby doxycycline-regulated shRNA-mediated knock-down of ß-catenin in established tumor xenografts resulted in massive cell differentiation in areas corresponding to high shRNA efficacy. This technology is currently applied to further investigate the role of the Pygo and Bcl9 genes in tumor progression and, in the context of a FP6 collaborative network, for the in vivo validation of shRNAs identified in cell-based assays to interfere with c-Myc.

Collaborations

These projects are carried out in part as collaborations with the groups of Prof. Konrad Basler, Institute of Molecular Biology, University of Zürich, Prof. Ivan Stamenkovic, Department of Experimental Pathology (IUP), CHUV Lausanne, and Prof. Martin Eilers, Institut für Molekularbiologie und Tumorforschung, Philipps-Universität Marburg.

Keywords

Mouse mutants of the Wnt signaling components
Developmental pathways and cancer
Tumor cell invasion