Final Report Summary - GUT GENE REGULATION (TCF4 transcriptional program in crypt stem cells and resulting differentiated cells)
Traditionally, chemotherapeutic agents have many specific effects causing severe side effects in patients. This is because most drug therapies target pleiotropic factors involved in multiple cellular pathways. During the last decade, we have seen the development of very specific chemotherapeutics, usually small molecules or antibodies directed at molecular targets that were identified by basic research as key regulators of growth in specific tumours. Such new therapies have dramatically changed the clinical outcome of the cancers for which they are used, demonstrating the importance of understanding the underlying molecular mechanisms of cancer initiation and progression in distinct tumours.
Colorectal cancer is the second most prevalent cancer in humans. Because physical symptoms of the patient often arise late in progression of the disease, colon cancer is frequently detected too late for successful treatment by surgical removal and / or chemotherapy, underscoring the importance of finding curative chemotherapeutic targets. Normal intestinal development and progression to colon cancer critically depend on the canonical Wnt signalling pathway, a key regulatory pathway controlling intestinal cell proliferation, differentiation, and stem cell maintenance. Deregulation of the Wnt pathway leads to malignancies in the mammalian gut epithelium; over 80% of colorectal cancers arise from mutations in Wnt pathway components.
There is extensive current knowledge of the physiology of the intestine and its cell types, the signalling pathways and the Wnt target genes involved. However, it remains incompletely understood how a Wnt target gene is differentially regulated in development or in its progression to colorectal cancer. For example, all cell types in the small intestinal crypt; stem cells, TA cells and paneth cells, as well as colorectal cancer cells, have active Wnt signalling. However, they display differential Wnt transcription programs.
The transcription factors TCF4 and beta-catenin are the molecular effectors of Wnt signalling in the intestinal epithelium and drive the expression of Wnt target genes. In the absence of the Wnt signal, cytoplasmic beta-catenin is phosphorylated by the beta-catenin destruction complex, a complex mainly containing APC, Axin, and the kinases GSK3 and CK1 followed by its ubiquitination and proteasomal degradation. Without nuclear beta-catenin, TCF4 represses Wnt target genes in complex with putative co-repressors. Upon Wnt signalling, the beta-catenin destruction complex is functionally inactivated, beta-catenin accumulates in the cytoplasm, translocates into the nucleus and complexes with TCF4 to activate the Wnt transcription program in stem or progenitor cells.
The beta-catenin/TCF4 complex is thought to drive TCF4 target gene expression by recruitment of activating cofactors. At present, although several potential cofactors for TCF4/beta-catenin have been implicated in Wnt-dependent transcription, few transcription factors that are essential and unique to this transcription program have been found. The identification of the full repertoire of the TCF4/beta-catenin co-activator complex dedicated to this transcription program in each of the above cell types is required to unveil the key differential mechanistic events responsible for normal development and for initiation and progression of cancer.
The identification of new players in the Wnt pathway, and by inference, players involved in colon cancer will allow for the development of new specific therapies. Another factor contributing to the differential Wnt transcription programs observed is long distance gene regulation. Genes are regulated not only by their cognate upstream promoters, but also by long distance regulatory elements, which can be tens or hundreds of kilobases away from their target promoters when thought of linearly. The presence of distinct long-range elements and long distance gene regulation in the different crypt cell types and in CRC under Wnt signalling provides an important added level of complexity to regulation of Wnt targets that would result in the observed differential transcription programs.
In this project, based on the premise that target identification is a promising approach towards development of specific therapies, we used an unbiased approach to identify novel components of the endogenous TCF4 complex in the murine small intestinal epithelium. We applied the combination of affinity purification and mass spectrometry to purify components of the TCF4 transcriptional complex from proliferative crypts where Wnt is 'on' and differentiated villi where Wnt is 'off'. The first results of the proposed study immediately demonstrated the strength of the experimental approach by identifying a novel TCF4 interacting factor specific for the Wnt pathway: TNIK (Mahmoudi et al., 2009).
We found that the kinase TNIK is localised to the nuclei of Wnt-active intestinal crypts, phosphorylates TCF4 and is recruited to promoters of Wnt target genes in mouse crypts and in colorectal cancer cells in a beta-catenin dependent manner. Importantly, gene expression profiling following depletion of TNIK in cell lines indicated that in contrast to other published coactivators of TCF4/beta-catenin such as BRG1 and p300, the transcriptional function of TNIK was not only essential but also exclusive to Wnt target gene regulation. Therefore, TNIK is a kinase potentially amenable to inhibition by small molecules and presents an attractive candidate for drug development against colorectal cancer. This study was the highlight article in EMBO J and later the subject of a SciBX (Science Business Exchange) news and views article.
Because of the potential of the kinase TNIK as a drug target in colorectal cancer, Wintherix LLC, a company which is specialised in developing small molecules for therapeutic purposes, has initiated a study for the inhibition of this kinase.Similar to TNIK, we identified MLLT10/AF10, an MLL fusion protein implicated in leukemia, together with its partner the histone methyltransferase enzyme DOT1L in the crypt TCF4 complex and found them to be recruited to Wnt target genes in vivo in a beta-catenin dependent manner (Mahmoudi et al., 2010).
Gene expression profiling in MLLT10/AF10 or DOT1L-depleted cells indicated that, although not exclusive to regulation of the Wnt transcription program, they are more specific to Wnt target gene regulation than other published beta-catenin coactivators. Using zebrafish as a model system, we demonstrated a genetic link between the Wnt pathway and MLLT10/AF10-DOT1L and delineated their essential role in Wnt-driven endogenous gene expression. Finally, we demonstrated the physiological role of MLLT10/AF10-DOT1L in Wnt-driven intestinal development and homeostasis; depletion of MLLT10/AF10-DOT1L in zebrafish embryos mimicked the TCF4 depleted phenotype and displayed significant intestinal proliferation defects accompanied by decreased total number of intestinal cells.
We postulate that similar to TNIK the enzyme DOT1L may present an attractive candidate for drug targeting in colorectal cancer.We extended the immunoprecipitation coupled to Mass Spec approach in order to study the identity of putative novel components and composition of the beta-catenin destruction complex in the presence and absence of the Wnt signal. Axin1, the least abundant component of the complex serves as a scaffold within this complex and is critical for rapid turnover of b-catenin.
To examine the mechanism by which Wnt signalling disables the destruction complex, we generated specific antibodies against Axin1 (Ng et al., 2009) and used an immunoprecipitation-coupled proteomics approach to identify novel endogenous binding partners of Axin1. We found MAP3K1 as an Axin1 interactor in colorectal cancer cells. Importantly, we found that the Axin1-MAP3K1 interaction is induced and modulated by Wnt stimulation. siRNA depletion and further functional studies dissecting the enzymatic activities of MAP3K1 demonstrated the essential role of the MAP3K1 E3 ubiquitin ligase activity in the positive regulation of the Wnt/b-catenin pathway and Wnt target gene expression (Ng et al., 2010).
Finally, to examine the role of long-distance transcriptional activation in regulation of Wnt target genes, we determined the genome wide TCF4 binding profile in the mouse small intestinal epithelium (Mahmoudi et al, in preparation). We find that a large majority, > 70 % of TCF4 binding sites are not in the vicinity of (more than 10 kilobases away from) known genes or transcription start sites. Thus they likely function as long distance regulatory elements.
We also find that binding of Tcf4 to its target sites is not only a mark, which is present irrespective of activity of the target gene in question. In fact Tcf4 was found to bind differentially to certain target genes in the proliferative crypt versus the differentiated villus intestinal compartment (Mahmoudi et al., in preparation).
This project addressed novel questions in the field of development and cancer biology. Focusing on the underlying molecular mechanisms that determine stem cell maintenance, proliferation, cell fate determination and differentiation, and progression into cancer in vivo, we set out to identify the DNA elements and protein complexes involved in Wnt/TCF4 target gene regulation in mammals.
This study used high through-put and state-of-the-art recently developed screening methods and techniques such as affinity purification of complexes followed by proteomics, and ChIP-chip, to address Wnt target gene regulation in an in vivo relevant model. By combining these experimental techniques and different fields of research, this project has identified novel players in and contributed to the development of novel concepts of transcription regulation of Wnt target genes over long distance.
These studies have also led to the identification of both novel DNA elements and other novel enzymes specifically associating with the transcriptionally active TCF4/beta-catenin crypt complex, whose functions we are actively characterising in Wnt signalling.
Colorectal cancer is the second most prevalent cancer in humans. Because physical symptoms of the patient often arise late in progression of the disease, colon cancer is frequently detected too late for successful treatment by surgical removal and / or chemotherapy, underscoring the importance of finding curative chemotherapeutic targets. Normal intestinal development and progression to colon cancer critically depend on the canonical Wnt signalling pathway, a key regulatory pathway controlling intestinal cell proliferation, differentiation, and stem cell maintenance. Deregulation of the Wnt pathway leads to malignancies in the mammalian gut epithelium; over 80% of colorectal cancers arise from mutations in Wnt pathway components.
There is extensive current knowledge of the physiology of the intestine and its cell types, the signalling pathways and the Wnt target genes involved. However, it remains incompletely understood how a Wnt target gene is differentially regulated in development or in its progression to colorectal cancer. For example, all cell types in the small intestinal crypt; stem cells, TA cells and paneth cells, as well as colorectal cancer cells, have active Wnt signalling. However, they display differential Wnt transcription programs.
The transcription factors TCF4 and beta-catenin are the molecular effectors of Wnt signalling in the intestinal epithelium and drive the expression of Wnt target genes. In the absence of the Wnt signal, cytoplasmic beta-catenin is phosphorylated by the beta-catenin destruction complex, a complex mainly containing APC, Axin, and the kinases GSK3 and CK1 followed by its ubiquitination and proteasomal degradation. Without nuclear beta-catenin, TCF4 represses Wnt target genes in complex with putative co-repressors. Upon Wnt signalling, the beta-catenin destruction complex is functionally inactivated, beta-catenin accumulates in the cytoplasm, translocates into the nucleus and complexes with TCF4 to activate the Wnt transcription program in stem or progenitor cells.
The beta-catenin/TCF4 complex is thought to drive TCF4 target gene expression by recruitment of activating cofactors. At present, although several potential cofactors for TCF4/beta-catenin have been implicated in Wnt-dependent transcription, few transcription factors that are essential and unique to this transcription program have been found. The identification of the full repertoire of the TCF4/beta-catenin co-activator complex dedicated to this transcription program in each of the above cell types is required to unveil the key differential mechanistic events responsible for normal development and for initiation and progression of cancer.
The identification of new players in the Wnt pathway, and by inference, players involved in colon cancer will allow for the development of new specific therapies. Another factor contributing to the differential Wnt transcription programs observed is long distance gene regulation. Genes are regulated not only by their cognate upstream promoters, but also by long distance regulatory elements, which can be tens or hundreds of kilobases away from their target promoters when thought of linearly. The presence of distinct long-range elements and long distance gene regulation in the different crypt cell types and in CRC under Wnt signalling provides an important added level of complexity to regulation of Wnt targets that would result in the observed differential transcription programs.
In this project, based on the premise that target identification is a promising approach towards development of specific therapies, we used an unbiased approach to identify novel components of the endogenous TCF4 complex in the murine small intestinal epithelium. We applied the combination of affinity purification and mass spectrometry to purify components of the TCF4 transcriptional complex from proliferative crypts where Wnt is 'on' and differentiated villi where Wnt is 'off'. The first results of the proposed study immediately demonstrated the strength of the experimental approach by identifying a novel TCF4 interacting factor specific for the Wnt pathway: TNIK (Mahmoudi et al., 2009).
We found that the kinase TNIK is localised to the nuclei of Wnt-active intestinal crypts, phosphorylates TCF4 and is recruited to promoters of Wnt target genes in mouse crypts and in colorectal cancer cells in a beta-catenin dependent manner. Importantly, gene expression profiling following depletion of TNIK in cell lines indicated that in contrast to other published coactivators of TCF4/beta-catenin such as BRG1 and p300, the transcriptional function of TNIK was not only essential but also exclusive to Wnt target gene regulation. Therefore, TNIK is a kinase potentially amenable to inhibition by small molecules and presents an attractive candidate for drug development against colorectal cancer. This study was the highlight article in EMBO J and later the subject of a SciBX (Science Business Exchange) news and views article.
Because of the potential of the kinase TNIK as a drug target in colorectal cancer, Wintherix LLC, a company which is specialised in developing small molecules for therapeutic purposes, has initiated a study for the inhibition of this kinase.Similar to TNIK, we identified MLLT10/AF10, an MLL fusion protein implicated in leukemia, together with its partner the histone methyltransferase enzyme DOT1L in the crypt TCF4 complex and found them to be recruited to Wnt target genes in vivo in a beta-catenin dependent manner (Mahmoudi et al., 2010).
Gene expression profiling in MLLT10/AF10 or DOT1L-depleted cells indicated that, although not exclusive to regulation of the Wnt transcription program, they are more specific to Wnt target gene regulation than other published beta-catenin coactivators. Using zebrafish as a model system, we demonstrated a genetic link between the Wnt pathway and MLLT10/AF10-DOT1L and delineated their essential role in Wnt-driven endogenous gene expression. Finally, we demonstrated the physiological role of MLLT10/AF10-DOT1L in Wnt-driven intestinal development and homeostasis; depletion of MLLT10/AF10-DOT1L in zebrafish embryos mimicked the TCF4 depleted phenotype and displayed significant intestinal proliferation defects accompanied by decreased total number of intestinal cells.
We postulate that similar to TNIK the enzyme DOT1L may present an attractive candidate for drug targeting in colorectal cancer.We extended the immunoprecipitation coupled to Mass Spec approach in order to study the identity of putative novel components and composition of the beta-catenin destruction complex in the presence and absence of the Wnt signal. Axin1, the least abundant component of the complex serves as a scaffold within this complex and is critical for rapid turnover of b-catenin.
To examine the mechanism by which Wnt signalling disables the destruction complex, we generated specific antibodies against Axin1 (Ng et al., 2009) and used an immunoprecipitation-coupled proteomics approach to identify novel endogenous binding partners of Axin1. We found MAP3K1 as an Axin1 interactor in colorectal cancer cells. Importantly, we found that the Axin1-MAP3K1 interaction is induced and modulated by Wnt stimulation. siRNA depletion and further functional studies dissecting the enzymatic activities of MAP3K1 demonstrated the essential role of the MAP3K1 E3 ubiquitin ligase activity in the positive regulation of the Wnt/b-catenin pathway and Wnt target gene expression (Ng et al., 2010).
Finally, to examine the role of long-distance transcriptional activation in regulation of Wnt target genes, we determined the genome wide TCF4 binding profile in the mouse small intestinal epithelium (Mahmoudi et al, in preparation). We find that a large majority, > 70 % of TCF4 binding sites are not in the vicinity of (more than 10 kilobases away from) known genes or transcription start sites. Thus they likely function as long distance regulatory elements.
We also find that binding of Tcf4 to its target sites is not only a mark, which is present irrespective of activity of the target gene in question. In fact Tcf4 was found to bind differentially to certain target genes in the proliferative crypt versus the differentiated villus intestinal compartment (Mahmoudi et al., in preparation).
This project addressed novel questions in the field of development and cancer biology. Focusing on the underlying molecular mechanisms that determine stem cell maintenance, proliferation, cell fate determination and differentiation, and progression into cancer in vivo, we set out to identify the DNA elements and protein complexes involved in Wnt/TCF4 target gene regulation in mammals.
This study used high through-put and state-of-the-art recently developed screening methods and techniques such as affinity purification of complexes followed by proteomics, and ChIP-chip, to address Wnt target gene regulation in an in vivo relevant model. By combining these experimental techniques and different fields of research, this project has identified novel players in and contributed to the development of novel concepts of transcription regulation of Wnt target genes over long distance.
These studies have also led to the identification of both novel DNA elements and other novel enzymes specifically associating with the transcriptionally active TCF4/beta-catenin crypt complex, whose functions we are actively characterising in Wnt signalling.