Reprogramming Breast Cancer Epithelial Cells

Institution: The Burnham Institute for Medical Research
Investigator(s): Steven Frisch, Ph.D. -
Award Cycle: 1997 (Cycle III) Grant #: 3IB-0081 Award: $97,500
Award Type: IDEA
Research Priorities
Biology of the Breast Cell>Pathogenesis: understanding the disease



Initial Award Abstract (1997)
Most breast cancers originate from previously normal breast epithelial cells. During this transformation into cancer cells, the epithelial cells often lose the expression of cell adhesion molecules. These transformed breast cancer cells are no longer regulated by normal cell-cell contact and environmental factors. Also accompanying this is the loss both of cell polarity and of a sensitivity to being detached from the extracellular matrix (anoikis-sensitivity), that in normal cells would cause cell death (apoptosis). We believe that at certain initiation (and perhaps advanced) stages of breast cancer there exist opportunities to reverse the cellular changes associated with malignant transformation. It is the focus of our research to identify cellular mechanisms to reprogram cancer cells back into their previously normal phenotype. In some respects this is analogous to reversing an epithelial-to-mesenchymal differentiation that occurs in development.

Recently, our laboratory has identified a viral gene product that converts human carcinoma cells and mesenchymal tumor cells into normal epithelial cells. This gene product, called the adenovirus ‘Ela’ 243 amino acid protein, is a nuclear DNA (gene)-regulatory protein. Ela does not bind directly to DNA, but acts as a cofactor with other DNA ‘adapter’ proteins, such as p300 and CBP. It is thought that the ‘signal’ for such DNA regulation relates to the critical adhesion events that are lost when cells become cancerous. Thus, Ela serves to selectively repress mesenchymal genes, thereby maintaining cells in the normal epithelial state. We think that the normal epithelial cell represents a default state. Moreover, it suggests that carcinoma cells and mesenchymal tumor cells can be induced to revert into this default state simply by inactivating the p300/CBP adapter proteins. The ‘default-phenotype’ hypothesis will be tested by determining if epithelial conversion can be achieved by inactivating the adapter proteins without the use of Ela. Our experiments on Ela and the target adapter proteins will use tumor cell lines and the primary method will be microinjection of antibodies and other compounds.

Our hypothesis of reprogramming transformed cancer cells with Ela opens several novel avenues to treatment. An added feature of this approach is that it could be useful even if only a fraction of the tumor cells are ‘reprogrammed’. Ela-converted cells become targets for immune response in the host, which can attack neighboring tumor cells. Further, the Ela protein could serve as a ‘template’ for designing drugs that mimic its activity.


Final Report (1998)
Many genetic alterations contribute to causing breast cancer. The primary aim of molecular biology is to understand these alterations so that they might be individually fixed in tumor cells. Our research has identified a gene that can apparently convert most or all kinds of lung cancer cells into normal epithelial cells, regardless of their genetic alterations. This gene is called E1a, and is normally carried by human adenoviruses. Our primary goal has been to analyze the mechanism by which E1a works. Hopefully, understanding this mechanism will help us to design new drugs that work based on the E1a principle, drugs that should suppress most or all breast cancer.

As a first step toward this goal, we have shown that the E1a protein's known interaction with a cellular protein called p300 is likely to be critical for tumor suppression. However, at the onset of this UC-BCRP project, more experiments were needed to show that this was really the case. With the help of the UC-BCRP funding, we have now clearly demonstrated that perturbations of p300 can lead to generalized breast cancer suppression. We have accomplished this by showing first that only one the two regions of p300 that interacts with E1a is relevant. We then designed "dominant-negative" proteins that interacted with this region of p300, proteins that were unrelated to E1a itself. Remarkably, these proteins also converted tumor cells into normal epithelial cells.

These results were consistent with our original hypothesis that we could revert tumor cells to a more normal phenotype by better understanding of the adenovirus E1a gene and its interactions with p300. In future work we will be able to relate this to adhesive properties of cells and to identify the detailed mechanism involved.