The Twentieth Century was the ¿Century of the Gene.¿ Using mostly isolated cells as units of function, we have identified the mechanisms by which genetic defects contribute to cancer. We now know of a number of very important genes that can become ¿cancer genes¿ (known as ¿oncogenes¿) when overexpressed, and others that make the cell more cancer-prone when lost (referred to as ¿tumor suppressor genes¿). Many of the oncogenes and tumor suppressor genes are involved in regulating growth of cells, a vital function that goes awry when cells become cancerous. Breast cancer is considered a stepwise accumulation of genetic mistakes in a one-way street toward the demise of the host. Traditional breast cancer therapies, both chemotherapy and radiation therapy, act by attacking growing cancer cells only. This strategy is not perfect: these agents can damage normal growing cells, and the tumor cells can escape the therapy and become resistant.

We believe that the new century will be the ¿Century of the Organism¿ and will place increasing emphasis on intact tissues, organs, and the cellular microenvironment (or ¿context¿). We also believe that while the mutation paradigm is undoubtedly correct in that, while cancer would not initiate without mutations, the genes themselves are like the keys on the piano, and it is the context that makes the music. This means that despite mutations, breast cancer may not proceed if the microenvironment of the cells is normal. My laboratory has proposed, and has provided experimental evidence, that the unit of function in higher organisms is the cell plus its surroundings, and that signals from the cells¿ surroundings contribute to normal and malignant behaviors. We have shown that by changing the context outside the cell (or at the cell surface), we can cause breast cancer (and mutation) in experimental animals. Conversely, by using a three-dimensional physiological method (a 3D ¿assay¿) that allows us to distinguish normal and cancerous human breast cells, we have shown that we can reverse the behavior (¿phenotype¿) of cancer cells to make them look and act like normal cells do in the breast. These ¿normalized¿ cells, despite having the same cancerous genome, do not form tumors in experimental animals. We have succeeded recently to normalize the severity of even a very aggressive breast cancer cell line and have found a number of antibodies and other small molecule inhibitors that can perform this task. Our very preliminary data in animals with one antibody indicate that it is effective and is not toxic in animals.

Our hypothesis is that progression to breast cancer can be reversible if we normalize the regulatory circuits that have to function to form the breast tissue, and that we need much more ambitious approaches to studying this complex disease than one gene at a time in isolated tumor cells grown on two-dimensional support. We propose to analyze a large number of existing aggressive tumor cell lines and primary tumors in our 3D assay, and to find novel ways of normalizing them using the extensive markers we have developed. We want to classify these tumors in terms of the kind of combination treatments that can reverse the malignant behavior in each class, and then to test these in experimental animals, and ultimately, in clinical trials.

We have brought together a multidisciplinary team of investigators from Lawrence Berkeley National Laboratory and other laboratories and universities across the United States. We will analyze many of the molecules that make up the cells¿ behavior, including their signaling circuitry as well as the structural cues that are used by breast cells, in a physiological surrounding, to form a functional unit. We will analyze these data with the help of time-lapse movies, sophisticated image analysis, mathematical modeling, and computers to track cancer complexity and will devise combination therapies that minimize damage to the host and control the cancer. We will also organize yearly workshops for all the collaborators and consultants and share our information and data as soon as they become available.