Computational Models of the Human Cell Cycle to Reveal Disease Mechanism and Inform Treatment

点击翻译按钮获取中文摘要

PROJECT SUMMARY / ABSTRACT The overall goal of this project is to develop computational models that predict how the human cell cycle responds to clinically-relevant perturbations such as radiotherapy, targeted therapy, oncogenic mutation, and directed differentiation. These models will fill a significant void in our understanding of the mechanisms underlying the initiation, progression, and treatment of diseases that involve abnormal cell proliferation. Our approach is to use quantitative single-cell imaging to measure the molecular states of proliferating cells and to integrate these data into predictive modeling frameworks. We have assembled a cross-institutional team comprising a computational biologist, two cell biologists, and a physician scientist with specialization in radiation oncology. The team has a strong and productive history of collaboration with six joint publications to date. Aim 1 investigates the mechanism by which retinal epithelial cells respond to radiation-induced DNA damage during S phase to execute G2 arrest. Time-lapse imaging and deterministic modeling will predict: how the response to DNA damage is delayed until the S/G2 transition; how a small-molecule inhibitor of DNA repair—currently involved in clinical trial—intensifies the arrest response; and how loss of the tumor suppressor p53 renders cells refractory to combination therapy. Aim 2 asks how pancreatic epithelial cells with mutations in KRAS escape permanent cell cycle arrest. We will use high-content imaging to profile multiple signaling activities in single cells expressing oncogenic KRAS. These data will be used to construct a manifold representation of cell cycle progression that spans a two-week time course of oncogenic KRAS-mediated transformation. Computational analysis of the manifold’s geometry will identify molecular branching points in G1 that govern the proliferation/arrest decision in pancreatic cells, and we will validate these predictions through small molecules and genetic manipulation. Aim 3 tests the hypothesis that human embryonic stem cells inherit cell-cycle-specific gene products (specifically, G1 regulators) from the previous G2 phase to promote pluripotency in daughter cells. We will combine mitosis-specific chromatin profiling with convolutional neural network-based image analysis to identify the mechanisms by which stem cells sustain rapid proliferation and pluripotency over multiple cell-cycle generations. Each aim yields both basic and applied knowledge, providing fundamental insights into cell cycle progression under perturbation and generating specific, molecular predictions to inform new treatment schemes. With an eye toward the future, predictive models of the human cell cycle will enable patient-specific treatments for diseases that are driven by abnormal cell proliferation.