Reconceptualizing Cancer: The development of new models and frameworks

Abstract

Cancer research secures an enormous amount of money each year, and researchers are generally quite productive with that money. The trend over the recent years has been for cancer biologists to focus on understanding the early stages of cancer and through a molecular and mechanistic framework. This gets translated into clinical applications through implementation of preventive and screening practices and early treatments. This work is important and has undoubtedly saved many lives. However, metastasis (the spread of cancer throughout the body) is the pathogenic aspect of cancer, and the cause of most cancer-related deaths. We have little ability to successfully treat cancers once they have reached this stage. Thus, this dissertation offers three novel models (two conceptual models and one empirical model) that facilitate the shift of attention away from the elucidation of tumor initiation and early stage cancers to the investigation of metastatic cancers. This shift to metastatic cancers involves more than a temporal shift in the cancer progression. That is, part of my argument is that we cannot simply apply the same conceptual frameworks and methods used to study tumorigenesis but to metastatic cells. Metastasis is more complex than tumor growth, so we need different conceptual models to guide that research. This dissertation begins with a pair of conceptual models, that I argue engage with important causal factors not contained in the molecular and mechanistic frameworks used to study tumorigenesis. I then use a simple model (budding yeast, Saccharomyces cerevisiae) and experimental evolution to empirically validate one of the models that suggests changes in interactions between levels of organization are causally informative in understanding the evolution of a trait (e.g., the evolution of cancer). This work explores how mitochondrial defection is involved in the evolution of tolerance to heat shock, which I use as proof of concept that one needs to consider multiple levels of organization together to understand cancer. In other words, this interdisciplinary dissertation demonstrates how context matters both in interpreting data and setting up research questions. The models I present here may help interpret existing data but also encourage novel research programs such as the empirical work presented here.