Deciphering Drug Action and Escape Pathways: An Example on Nasopharyngeal Carcinoma

Motivation: Nasopharyngeal carcinoma (NPC) is a malignant cancer in the head and neck region, with especially high incidence in South China, Southeastern Asia and North Africa. Recently, a cyclin dependent kinase (CDK) inhibitor, CYC202, is studied for its antitumor effect in human NPC cells in vitro and in vivo. Results show that both cell lines and patients in the study responded to the drug treatment differently. To further investigate the drug response, expression of selected genes for apoptosis, cell proliferation and cell cycle regulation were measured during the process of treatment. Our issue is how to identify the reason for the different responses in these NPC individuals using the gene expression data. Results: Biological pathway information has long been incorporated into gene expression analysis for the purpose of treatment response understanding. However, the conclusions are usually too general, and hardly sufficient for guiding further research. In our current study, we design a drug pathway identification system, the Drug Pathway Decipherer, which identifies genetic regulations in response to drug treatment that are consistent with respect to a given detailed signaling pathway structure. By applying our system to the NPC dataset, we discover that the status of ERK pathway and apoptosis pathway are differently regulated between responders and non-responders both in vitro and in vivo. Our results indicate that the dysregulation of Ras-ERK pathway and PI3K-Akt-NFκB pathway are probably the mechanisms for CYC202-insensitive NPC cells to resist the drug treatment. Availability: The Drug Pathway Decipherer is available at http://www.comp.nus.edu.sg/∼wongls/projects/drug-pathway/DPD-v1. It is implemented in JAVA. Contact: [email protected], [email protected], [email protected], and [email protected]