Resumen de Estudi de l'evolució i el tractament del càncer des d'una perspectiva de la biologia de sistemes
Jordi Serra Musach
Cancer can be defined as a family of complex disease characterized by a cellular hyperproliferation and the capacity of the cell to invade surrounding tissue and metastasize to other organs in the body. The development and progression of cancer is a complex multistep process in which cells gradually acquire mutations altering the cellular mechanisms until the cell becomes malignant. Cancer develops and progresses through the acquisition of the above mutations or alterations across biological levels. Downstream of these alterations are expression changes in many genes at each stage of the disease. Sets of genes (also called signatures whose differential expression or profiles have prognostic or predictive (in terms of prediction of drug-response) values have been identified for almost every type of cancer. In some cases, several signatures have proved to be useful in independent evaluations, although, intriguingly, their overlap in gene identities was minimal. Then, integrative approaches using different types of gene and protein relationships have demonstrated the existence of biological convergence among apparently disparate gene sets. Moreover, integrating data from he network of known protein-protein interactions (hereafter interactome network) has been shown to improve the reproducibility and accuracy of prognostic signatures. However, in this scenario, the network topological patterns linked to the dynamic molecular alterations that characterize cancer development and progression, and treatment response, remain unknown. Several key genetic, transcriptomic and molecular determinants of cancer therapeutic response have been identified in recent years. Specific genetic alterations have been demonstrated to mediate the existence and/or promote the acquisition of therapeutic resistance. In addition, sets of transcripts have been identified whose profiles have predictive value for prognosis and/or treatment benefit. Successively, the integration and modeling of data from different types of molecular interactions has been shown to enhance understanding of the mechanisms of cancer progression and therapy response. In this scenario, however, cancer patients all too frequently show no or only modest benefit from a given therapy. The persistence of this fundamental clinical problem has been partially attributed to the lack of specific biomarkers. However, the identification of a comprehensive measure of cancer cell activity may support the interpretation of therapy efficacy. At a cancer system level, several studies have shown extensive molecular rewiring and increased signaling entropy, which may endorse the characteristic robustness of cancer. Given these observations, an a priori understanding of therapeutic response should be supported by an integrated measure of cancer network activity. In this study, firstly, we hypothesized that the features of dynamism and robustness intrinsic to cancer should also be present at different biological levels and, in particular, evident within the topology of he interactome network. To assess this hypothesis we analyzed the impact of cancer-related expression changes- including cancer development, progression, response to treatments, and targeted perturbation-in the interactome network using the concept of cascading failures. The results of these analyses associate robustness with cancer and identify autophagy as an opposite condition. Next, we hypothesized that a comprehensive measure of cancer cell status or function applicable to predicting therapy response should integrate and/or partially reflect the fundamental hallmarks of cancer. To address these questions, we have developed a novel weighted network score, which is used to show that cancer network activity is associated with therapeutic response and synergism.
