Transcriptome-wide identification of key regulatory networks in drug resistant cancer cells

2015 Fiscal Year Research-status Report

• Project/Area Number: 15K21634
• Research Institution: Institute of Physical and Chemical Research
• Principal Investigator: フゥァーン イー 国立研究開発法人理化学研究所, ライフサイエンス技術基盤研究センター, 特別研究員 (60750289)
• Project Period (FY): 2015-04-01 – 2017-03-31
• Keywords: Cancer / Drug resistance / Transcriptome / CAGE

Outline of Annual Research Achievements

In order to achieve the goal to identify the key regulatory networks in drug resistant cancer cells, an integrative framework for identifying key regulators for specific pathways that induced by drug treatment was developed in this fiscal year. Besides, this framework could euclidate the unexpected effects of drugs. We first used the in-house existed CAGE (Cap Analysis of Gene Expression) data as pilot dataset, which was generated from four different kinds of statins treated with three cell lines, including HepG2, MCF7 and THP1 cells , to examine the feasibility of pipeline. The inhibitors of 3-hydroxy-3-methyl-glutaryl-coenzyme A (HMG- CoA) reductase, statins, are commonly used for reducing blood levels of low-density lipoprotein (LDL) cholesterol. In recent years, not only re-positioning statins as anticancer drugs has been reported but also some serious side-effects have been revealed. However, the underlying mechanisms of unexpected effects are not well studied. According to previous studies, our pipeline can successfully identify unreported transcriptional factors for known unexpected pathways of statins For example, statins have been proved that can induce apoptosis in MCF7 and THP1 cells but the key regulator of statin-indcued apoptosis is still unknonw. By our pipeline, we identify that SP1, which has been reported as one of key regulators of apoptosis, acts key regulator of statin-induced apoptosis in MCF7 cells.

Current Status of Research Progress

2: Research has progressed on the whole more than it was originally planned.

Reason

To process the raw data retreiving from CCLE and LINCS database, we need to develop an appropriate normalisation approaches unless we cannot perform unbiased meta-analyses from public dataset. In addition, the drug sensitiviy information by the prediciting model of CCLE cannot be downloaded. Instead, we attempt to used API provided by LINCS database to download the drug response data. Therefore, we decide to use the pre-existed data as pilot dataset to test our pipeline in this fiscal yesr.

Strategy for Future Research Activity

Retreving the expression profiles and drug reponse information from CCLE and LINCS database apply to our framework identify the drug-resitance-related TF-pathways pairs. If we can identify some key regulators for drug-resistant states, we will try to design experimental experiments to validate our findings in vitro. For example, knock-down the key transcriptional factors and then treat with drug to examine if it can revert the drug resistance.

Causes of Carryover

Bioinformatics software and hardware are out of our expectation.

Expenditure Plan for Carryover Budget

Purchase the software license.