Our Research Interests

An overview of the research that goes on in the Anderson Lab

It is generally well accepted that cancer is a genetic disease driven by mutations in key genes that lead to uncontrolled growth and abnormal cell behavior. However, the fact that the tumor is an evolving system and therefore subject to selection pressure and adaptation is largely ignored. Theoretical models tell us that these evolutionary dynamics are what drive tumor progression and treatment resistance and are a major focus of our research. As soon as we think of cancer as an evolving system we are then forced to consider what are the selection pressures that drive this evolution and how does the tumor population adapt. Adaptability will certainly be influenced by heterogeneity within the tumor cell population be it genetic or phenotypic or both (so called intratumor heterogeneity). Selection pressures are imposed from microenvironmental factors as well as other cellular populations that ultimately define the ecology of the tissue in which the tumor is developing. In real ecological systems such as the lake shown on the right, the different individual components (e.g. birds, fish, trees, plants, rain, water etc) work in symphony to maintain a homeostatic balance that maintains the life of the system. This same perspective in relation to an organ, means that we should consider normal tissue homeostasis as an equilibrium that cancer cells can disrupt. Disruption of this equilibrium is often one of the first events in cancer development, as the normal control mechanisms of the tissue are damaged or ignored. Understanding the interplay between homeostasis, heterogeneity, evolution and ecology in cancer progression is a unifying theme for all the work being done in the lab and we further elaborate on some these below.

Mathematical and Computational Models

During the last few years we have been developing a number of mathematical and computational approaches that consider different spatial and temporal scales of tumor progression. The resolution and scope of each model is driven by the specific aspect of tumor progression we are considering. These models can be broken into discrete continuous components although most often these components operate together in hybrid multiscale models. We have previously developed purely continuous models that consider both temporal and spatio-temporal (Reaction Diffusion) dynamics but we believe that individual based models which operate at the cellular scale and are ideally suited to connect with in vitro experimental approaches. They can represent physical cell structure, phenotypic heterogeneity and capture cell-cell interactions with ease (IBcell, CPM, HDC, EHCA). However, they are computationally expensive, as increasing numbers of cells are considered, and are therefore limited in terms of the size of tumors they can reproduce. Continuous models don’t have such a limitation, but loose resolution at the cell scale, and can recapitulate clinical scale tumors as well as the long term evolutionary dynamics of the tumor (EGT).

Targeted Therapy, Intratumor Heterogeneity and Drug Resistance

Tumor phenotype space Molecularly targeted therapies are becoming more prevalent in our treatment of cancer but are they the future of cancer treatment or are they destined to fail precisely because they target a subset of the tumor population? Before we can address this question it is worth noting the relationship between a cells genotype and its phenotype is still very much a fundamental open biological question. The so called genotype-phenotype mapping held early promise when we still believed that individual genetic mutations drove individual phenotypic changes. From our current understanding we now know that multiple genotypes can produce the same phenotype and conversely multiple phenotypes can be produced from a single genotype. Evolution selects phenotypes and not genotypes so its hardly surprising that a growing evolving tumor contains massive amounts of molecular heterogeneity but in contrast it only appears to made up of a few convergent phenotypes. The figure on the right highlights the phenotypic diversity in a growing tumor (lower left) with the colors representing the levels of two key traits and the points represent the phenotype of every cell in the tumor (thanks to Dr. Mark Tessi for this figure). We can clearly see a large proportion of the tumor has high values in one trait and intermediate in another. So how does this diversity change treament - it makes logical sense that if evolution selects phenotypes then so must treatment and treatment is a very strong form of selection pressure, a life or death pressure. So if you have the right phenotype you live otherwise you die, leaving behind phenotypes that are resistant. We are currently developing mathematical models that will allow us to better understand the relationship between genotype and phenotype and how phenotypic diversity (including drug resistant phenotypes) might be overcome with the right combination therapies. Further details on this work can be found here, here and here.

Stomal Activation and Its Role in Tumor Initiation and Growth

tumor_stroma The role of tumour-stromal interactions in progression is generally well accepted but their role in initiation or treatment is less well understood. It is now generally agreed that rather than just consisting of malignant cells, tumors instead consist of a complex dynamic mixture of cancer cells, host fibroblasts, endothelial cells, and immune cells that interact with each other and microenvironmental factors to drive tumor progression (figure on the left shows fibroblasts [green] infiltrating a melanoma spheroid, courtesy of Dr. Keiran Smalley). We are particularly interested in stromal cells (e.g. fibroblasts) and stromal factors (e.g. fibronectin) as important players in tumour initiation, progression and since they have also been implicated in drug resistance. The dialogue between epithelial cells (or tumour precursors) and stromal cells does not fully begin until the epithelial cells break through their basement membrane (this is true for ductal cancers and for cancers such as melanoma). However, diffusible factors between the epithelium and the stroma are producing a constant dialogue that normally helps regulate tissue homeostasis. This dialogue can be altered from either epithelial or stromal side of the basement membrane and can tip the balance of homeostasis in favour of abnormal growth. Thus creating an environment ripe of tumour initiation and promoting tumour progression. To further add to the complexity of this dialogue, stromal cells and the factors they produce have been shown to aid tumour cell drug resistance. Further details on this work can be found here, here and here.

Tumor Invasion: The Roles of Heterogeneity and the Microenvironment

We have previously investigated the role of the tumor microenvironment (mE) in driving the evolution of aggressive tumor phenotypes. Specifically we considered a tumor to be a heterogeneous population containing potentially many distinct cellular phenotypes (defined in terms of variations of cell specific traits such as, cell-cell adhesion, migration speed, and proliferation rate) and via proliferation each cell had a small chance to mutate to one of these phenotypes in a random manner. By growing in silico heterogeneous tumors in several distinct mEs (e.g. nutrient rich/poor, uniform/grainy ECM) we were able to show that harsher mEs (e.g. low nutrient, grainy ECM) selected for fewer and more aggressive phenotypes. This change in cellular phenotypes was also accompanied by a morphological change, with the more aggressive clones producing more fingered structures. The Figure on the right shows an example of both the tumor morphology and graphs of the abundance of cellular phenotypes that evolved over time in the two distinct mEs. These simulations lead to a novel hypothesis: invasion is an emergent property of the collective behaviour of the cell population under selective pressure from the mE. This is alternative to the current invasive phenotype paradigm, in which invasion is the culmination of a linear cancer progression process. Further details on this work can be found here, here and here.

Tumor Ecology and Evolution

Figure by Dr. Omar Franco, see publication [3]. Ecology and evolution are intimately linked as one provides the players and the field and the rules of the game and the means to change them. The idea of viewing cancer from an ecological perspective fundamentally means that we cannot just consider cancer as a collection of mutated cells but as part of a complex balance of many interacting cellular and microenvironmental elements. Interactions are really at the heart of this research and it is commonly assumed that in an ecosystem with limited resources, nutrients and space for growth, the only relevant type of interaction between individuals would be competition. But there are potentially many more ecological interactions, defined according to the fitness benefit to the participants, such as: competition, predation, parasitism, mutualism and commensalism. All these types of interactions can be observed, to some degree, in a real tumor see figure on the left in relation to prostate cancer. In this example the recruitment of new blood vessels by tumor cells could be considered mutualistic since both benefit equally from the exchange; the competition for resources among various tumor phenotypes, which results in the environmental selection of the fittest cell types; and, the predatory digestion of tumor cells and the parasitic consumption of tumor cell resources by infiltrating immune cells, which represent two more examples of emergent relationships during tumorigenesis. Further details on this can be found here, here and here.