Cohen David P A, Martignetti Loredana, Robine Sylvie, Barillot Emmanuel, Zinovyev Andrei, Calzone Laurence
Institut Curie, Paris, France; INSERM, U900, Paris, France; Mines ParisTech, Fontainebleau, Paris, France.
Institut Curie, Paris, France; CNRS UMR144, Paris, France.
PLoS Comput Biol. 2015 Nov 3;11(11):e1004571. doi: 10.1371/journal.pcbi.1004571. eCollection 2015 Nov.
Understanding the etiology of metastasis is very important in clinical perspective, since it is estimated that metastasis accounts for 90% of cancer patient mortality. Metastasis results from a sequence of multiple steps including invasion and migration. The early stages of metastasis are tightly controlled in normal cells and can be drastically affected by malignant mutations; therefore, they might constitute the principal determinants of the overall metastatic rate even if the later stages take long to occur. To elucidate the role of individual mutations or their combinations affecting the metastatic development, a logical model has been constructed that recapitulates published experimental results of known gene perturbations on local invasion and migration processes, and predict the effect of not yet experimentally assessed mutations. The model has been validated using experimental data on transcriptome dynamics following TGF-β-dependent induction of Epithelial to Mesenchymal Transition in lung cancer cell lines. A method to associate gene expression profiles with different stable state solutions of the logical model has been developed for that purpose. In addition, we have systematically predicted alleviating (masking) and synergistic pairwise genetic interactions between the genes composing the model with respect to the probability of acquiring the metastatic phenotype. We focused on several unexpected synergistic genetic interactions leading to theoretically very high metastasis probability. Among them, the synergistic combination of Notch overexpression and p53 deletion shows one of the strongest effects, which is in agreement with a recent published experiment in a mouse model of gut cancer. The mathematical model can recapitulate experimental mutations in both cell line and mouse models. Furthermore, the model predicts new gene perturbations that affect the early steps of metastasis underlying potential intervention points for innovative therapeutic strategies in oncology.
从临床角度来看，了解转移的病因非常重要，因为据估计转移占癌症患者死亡率的90%。转移是由包括侵袭和迁移在内的一系列多个步骤导致的。在正常细胞中，转移的早期阶段受到严格控制，并且可能会受到恶性突变的严重影响；因此，即使后期阶段发生时间较长，它们也可能构成总体转移率的主要决定因素。为了阐明影响转移发展的单个突变或其组合的作用，构建了一个逻辑模型，该模型概括了已知基因扰动对局部侵袭和迁移过程的已发表实验结果，并预测尚未经过实验评估的突变的影响。该模型已使用肺癌细胞系中TGF-β依赖性诱导上皮-间质转化后转录组动力学的实验数据进行了验证。为此，开发了一种将基因表达谱与逻辑模型的不同稳定状态解相关联的方法。此外，我们系统地预测了组成模型的基因之间关于获得转移表型概率的缓解（掩盖）和协同成对遗传相互作用。我们关注了几种导致理论上转移概率非常高的意外协同遗传相互作用。其中，Notch过表达和p53缺失的协同组合显示出最强的效应之一，这与最近在肠道癌小鼠模型中发表的实验结果一致。该数学模型可以概括细胞系和小鼠模型中的实验突变。此外，该模型预测了影响转移早期步骤的新基因扰动，为肿瘤学创新治疗策略提供了潜在的干预点。
