Reconstructing Kinase Network Dynamics to Predict Stochastic Cell Cycle Fate

摘要-重构激酶网络动力学预测随机细胞周期命运。 在遗传上相同的克隆来源的细胞群体中，随机基因表达导致 蛋白质表达水平的自然细胞间差异1 - 8，导致单个细胞表现出不同的细胞命运 当用同样的刺激物处理时，这可能潜在地产生耐药细胞群体2，9， 阻碍癌症治疗。这种现象被称为自然表型趋异（NPD），它 蛋白质表达噪音如何影响相互作用的非线性信号的随机动力学 networks2，9 - 11.蛋白质表达噪音和信号动力学的非线性性质使得难以 预测单细胞将如何响应扰动，如化疗药物或促分裂剂。聚焦细胞 增殖反应，我们假设我们可以预测细胞增殖的时间和概率 通过推断ERK、JNK和Akt信号网络的动态连接结构， 普遍控制细胞周期进入。通过了解这些通路的连接结构， 因果计算网络模型可以被开发，它可以预测细胞的时间和概率 在单细胞水平上增殖。通过将该网络模型与活细胞成像实验结合， 不同的乳腺癌亚型，我们可以评估激酶网络如何控制细胞周期进入的一般性， 以及细胞转化如何影响这些控制系统;这可以提供翻译的见解， 癌症中的新信号靶标，预测转化细胞如何对化疗药物作出反应， 短暂耐药性的发展。 为了实现这一目标，提出了以下目标：目标1。产生扰动成像时间过程 ERK、JNK、Akt和S期进入动力学数据用于动态网络模型重建。目标2.构建体 预测S期进入概率动态的ERK-JNK-Akt网络模型。目标3。实验测试 在一组不同临床表现的乳腺癌细胞系中基于模型的S期进入反应预测 亚型活细胞成像将用于获取ERK JNK和Akt动态，沿着S期进入 分别使用激酶易位报告基因（KTR）24，25和mCherry-双生蛋白S-期探针26检测应答。 该动态数据将作为动态模块化响应分析（DMRA）的输入数据， 用于构建一个因果网络模型，该模型由ERK、JNK 和Akt，沿着S期进入。该网络模型沿着不同乳腺的活细胞成像实验 癌症亚型将用于生成和测试模型预测，这些预测提供了对细胞的普遍存在的洞察力。 哺乳动物细胞中的周期进入控制系统以及细胞转化如何影响该控制。该模型可以 提供对新型癌症信号通路靶点的翻译见解，并预测单细胞反应 化疗和靶向药物，并因此短暂的耐药机制。

Abstract- Reconstructing Kinase Network Dynamics to Predict Stochastic Cell Cycle Fate. In a genetically identical and clonally-derived population of cells, stochastic gene expression causes natural cell-to-cell variations in protein expression levels1–8, which causes single cells to exhibit different cell fate when treated with the same stimuli, which can potentially give rise to a population of drug resistant cells2,9, impeding cancer treatment. This phenomenon is referred to as natural phenotypic divergence (NPD), and it arises from how protein expression noise influences the stochastic dynamics of interacting non-linear signaling networks2,9–11. Both protein expression noise and the non-linear nature of signaling dynamics makes it difficult to predict how single cells will respond to a perturbation such as, chemotherapeutics or mitogens. Focusing on cell proliferation responses, we hypothesize that we can predict the timing and probability of cell proliferation by inferring the dynamic connection architecture of the ERK, JNK and Akt signaling networks, which ubiquitously control cell cycle entry. By understanding the connection architecture of these pathways, a causal computational network model can be developed, which can predict the timing and probability of cell proliferation at the single cell level. By combining this network model with live cell imaging experiments spanning different breast cancer subtypes, we can evaluate the generality of how kinase networks control cell cycle entry as well as how cell transformation affects these control systems; which can provide translational insight into novel signaling targets in cancer, predict how transformed cells respond to chemotherapeutics, and the development of transient drug resistance. To achieve this goal, the following aims are proposed: Aim 1. Generate perturbation imaging time course data of ERK, JNK, Akt and S-phase entry dynamics for dynamic network model reconstruction. Aim 2. Construct an ERK-JNK-Akt network model predictive of S-phase entry probability dynamics. Aim 3. Experimentally test model-based predictions of S-phase entry response in a panel of breast cancer cell lines of varying clinical subtypes. Live cell imaging will be used to acquire ERK JNK and Akt dynamics, along with S-phase entry response using kinase translocation reporters (KTR)24,25 and the mCherry-geminin S-phase probe26 respectively. This dynamic data will serve as input data for dynamic modular responses analysis22,23 (DMRA) which will be used to construct a causal network model consisting of the empirical interaction strengths between ERK, JNK and Akt, along with S-phase entry. This network model along with live cell imaging experiments in different breast cancer subtypes will be used to generate and test model predictions, which provide insight the ubiquity of cell cycle entry control systems in mammalian cells and how cell transformation affects that control. The model can provide translational insight into novel cancer signaling pathway targets, as well as predict single cell response to chemotherapeutic and targeted drugs, and consequently transient drug resistance mechanisms.