Quantitative measurement and modeling of the DNA damage signaling network : DNA double-strand breaks

Abstract

DNA double-strand breaks (DSB) are one of the major mediators of chemotherapy-induced cytotoxicity in tumors. Cells that experience DNA damage can initiate a DNA damage-mediated cell-cycle arrest, attempt to repair the damage and, if successful, resume the cell-cycle (arrest/repair/resume). Cells can also initiate an active cell-death program known as apoptosis. However, it is not known what "formula" a cell uses to integrate protein signaling molecule activities to determine which of these paths it will take, or what protein signaling-molecules are essential to the execution of that decision. A better understanding of how these cellular decisions are made and mediated on a molecular level is essential to the improvement of existing combination and targeted chemotherapies, and to the development of novel targeted and personalized therapies. Our goal has been to gain an understanding of how cells responding to DSB integrate protein signaling-molecule activities across distinct signaling networks to make and execute binary cell-fate decisions, under conditions relevant to tumor physiology and treatment. We created a quantitative signal-response dataset, measuring signals that widely sample the response of signaling networks activated by the induction of DSB, and the associated cellular phenotypic responses, that together reflect the dynamic cellular responses that follow the induction of DSB. We made use of mathematical modeling approaches to systematically discover signal-response relationships within the DSB-responsive protein signaling network. The structure and content of the signal-response dataset is described, and the use of mathematical modeling approaches to analyze the dataset and discover specific signal-response relationships is illustrated. As a specific example, we selected a particularly strong set of identified signal-response correlations between ERK1/2 activity and S phase cell-cycle phenotype, identified in the mathematical data analysis, to posit a causal relationship between ERK1/2 and S phase cell cycle phenotype. We translated this posited causal relationship into an experimental hypothesis and experimentally test this hypothesis. We describe the validation of an experimental hypothesis based upon model-derived signal response relationships, and demonstrate a dual role for ERK1/2 in mediating cell-cycle arrest and apoptosis following DNA damage. Directions for the extension of the signal-response dataset and mathematical modeling approaches are outlined.