The epithelial to mesenchymal transition (EMT) and mesenchymal to epithelial transition (MET) are crucial for proper embryonic development but are also exploited by tumor cells. Tumor cells of epithelial origin can benefit from an EMT event and lose their epithelial and polarized morphology and gain mesenchymal unpolarized cell morphology. These cells can now cross the blood barrier, relocate and colonize sites in distant organs. Eventually, they will initiate MET to become epithelial-like and successfully form secondary tumors. Induction of EMT is integrated by many different signals but shares the activation of intracellular EMT-inducers like Snail, Slug, Zeb1 and others, which are transcription factors acting as repressors of the Ca2+-dependent adhesion molecule (E-cadherin), which is at the center of EMT-MET, its expression must be successfully downregulated for EMT to proceed, and must be rapidly restored for proper MET.
To date, MET is considered as a reverse process of EMT. However, MET is not simply the reverse process of EMT. It is a complicated and dynamic process. There must be a more complex regulatory mechanisms ensuring the establishment of an epithelial phenotype. This dynamic MET process can only in part be explained by the upregulation and downregulation of EMT inducers. It suggests the presence of other transcriptional activators for E-cadherin, which could also be regulators of MET.RESEARCH INTERESTS
We are interested in understanding the genetic and molecular mechanisms of tumorigenesis, with a focus on dissecting the transcriptional networks controlling the mesenchymal to epithelial transition.
Transcriptional networks have been described before; examples include, but are not limited to, the transcriptional network of EMT, the core pluripotency network of embryonic stem cells, and the transcriptional network of blood stem cells. There are different kinds (or motifs) of transcriptional regulatory networks. These are sets of transcription factors that control the initiation and maintenance of certain transcriptional loops or chains to augment or stabilize gene expression. Examples of the network motifs include for example autoregulatory loops, where a transcription factor would be regulating its own expression, such as the Nanog loop in the core pluripotency network. In feed forward loops, transcription factor X will regulate both Gene Y and Gene Z, where Gene Y encodes a transcription factor that also regulates Gene-Z, an example of such regulatory loop comes from our recent study describing the initiation of MET, in which Grhl3 regulates the expression of both Cdh1 and Hnf4a, Hnf4a is then recruited to the Cdh1 chromatin to boost up the Grhl3-dependent regulation. In a multi-component loop, transcription factors regulate the expression of their regulators in a positive feedback loop; a very well-known example is the closed circuit of Oct4/Sox2 in embryonic stem cells regulatory network. In a regulatory chain, three or more transcription factors are ordered in a series, the chain ends when the transcription factor has no target (i.e. no transcription factor target) or is autoregulatory, an example of such regulatory chains comes from the hematopoietic stem cells, where b-catenin regulates HoxB4, which in turn activates the expression of c-myc. Common to most transcriptional network studies is the methodology used; most share similar sets of experimental and knowledge-based approaches, combining data mining, gene expression profiling and gene silencing, all integrated into a powerful computational analysis to correctly decipher transcriptional nodes and network motifs.
In order to validate the existence of an MET network, we will investigate in detail the expression profiles of several EMT-MET cellular models; including NMuMG cells as well as human cancer cell line models that are known to undergo mesenchymal to epithelial transition. We will use an integrated approach, including data mining, high throughput expression profiling, gene silencing, and validation by protein-protein and protein-DNA interaction techniques. As a result, we anticipate defining an epithelial state specific transcription factor gene set. Using this gene set we will continue to define the relationships between the members and their function within a network. Our preliminary data suggest the existence of several overlapping transcriptional nodes having Grhl3 in the center of a larger network and E-cadherin as the major outcome. Our main goal is to confirm the transcriptional relationships between the identified transcription factors and to draw a more comprehensive picture of the MET network. As a result, we aim to identify a core MET network, which, in the long run, may lead to the identification of therapeutic targets for the control of MET during metastasis. MET is also a critical step during the reprogramming of iPSCs, thus identifying a core MET network could be utilized to improve the efficiency and quality of the reprogramming process.
To date, there is a substantial amount of data in the literature analyzing the EMT program both during embryonic development and also during tumorigenesis, but a similar understanding of the MET program is still lagging. Recently we have identified a feed forward loop, which is composed of the transcription factors Grhl3 and Hnf4a that is essential for the progression of MET. In fact, Grhl3 emerged as an absolute requirement for the initiation of MET. Cells lacking Grhl3 as a result of RNAi-mediated knockdown failed to initiate the MET program. This is in part due to their inability to upregulate E-cad expression. Furthermore, we found that Grhl3 contributes to the regulation of Hnf4α by binding to a well-conserved motif present within the Hnf4α promoter P2. We believe that there is a need to gather more credible scientific evidence that answers key questions regarding how the process of MET is regulated, how it is initiated and the transcription factors involved in this regulation.