Biomedical Engineering Ph.D. Thesis Defense
Date: Friday, February 1, 2019
Location: CHOA Seminar Room in EBB Krone, Room 1005
Dr. Eberhard O. Voit, Department of Biomedical Engineering
Dr. Bernd Bodenmiller, University of Zurich
Dr. Barbara D. Boyan, Virginia Commonwealth University
Dr. Melissa L. Kemp, Department of Biomedical Engineering
Dr. John F. McDonald, School of Biology
Dr. Peng Qiu, Department of Biomedical Engineering
Computational Modeling and Analysis of Single-cell Signaling Dynamics in Heterogeneous Cell Populations
Cell signaling pathways are complex biochemical systems at the core of cellular information processing. The dynamics of these signaling systems in response to internal and extracellular cues plays a critical role for proper cell functioning. While we have learned much about signaling at the cell population level, no two cells are the same, and cell-to-cell variability can have complex and important consequences for signaling in both individual cells and the cell population as a whole. In many contexts, cells perform essentially identical functions despite their differences, whereas in other contexts, especially in cancer, cell-cell differences in state propagate to differences in function.
The overall goal of this dissertation was the creation of mathematical and computational tools for the study of cell-to-cell variation in signaling and to use these tools to increase our understanding of when single cell differences do, or do not, make a meaningful difference. To address this goal we designed new methods of single-cell analysis, including a computational framework termed 'single-cell ordinary differential equation modeling' (SCODEM) that overcomes the prior experimental trade-off between continuous and multiplexed single-cell measurements of signaling. We tested SCODEM against increasingly demanding datasets, which were all represented in a satisfactory fashion. After the initial analysis of cell-to-cell variability, we analyzed targeted inhibition, protein overexpression and an epithelial-mesenchymal transition. Throughout this process, we provided illustrative examples of how our modeling framework may be used to identify operating principles and limits of signaling systems, which is a first step toward proposing novel therapeutic targets.
The work presented here provides new tools for analyzing cellular heterogeneity and increases our understanding of how differences in cell state effect function by showing intracellular signaling is primarily deterministic at the single cell level. The application of these tools to the dramatic phenotype shift during an epithelial-mesenchymal transition in murine breast cancer cells confirmed that stochasticity plays a much smaller role than had been assumed and that cells modulate signaling without the need of rewiring their signaling network.