Targeted Inactivation of the c-Myb Oncogene through Integrated Network Analysis

Major progress has been made in the treatment of Acute Myeloid Leukaemia (AML). However, drugs that have been developed to treat leukaemia still fail in about half the patients. New treatments are urgently needed. In the last couple of years we have been working to understand the exact molecular changes in the blood forming cells that cause leukaemia. We, and others, have identified key proteins that these cells need to survive. These are known as the driving oncogenes. Here we describe a pipeline, implemented by us, to inactivate key driving transcription factors through drug repositioning. In this study we focus on the oncogene c-MYB, a transcription factor that has a critical role in AML.

To achieve the goal of discovering drugs that pharmacologically target the driving transcription factors for inactivation we set up a technique for integrated analysis of gene expression datasets and transcription factor binding to DNA and combined this with pattern-matching software to mine the connectivity map database. Briefly, inducible expression/knockdown systems and next generation profiling allowed us to integrate cellular output (RNA-Seq or Genechip) and transcription factor binding (ChIP-Seq or ChIPip-on-Chip) of the driving c-MYB oncoprotein in leukaemia. This profile was used to mine the connectivity database for perturbagens of this profile.

The top hit, of this integrated network-based analysis of transcription factor behaviour and perturbing agents, was taken forward for analysis. It was found to cause an acute reduction in the protein levels of MYB, preceding changes in the level of its mRNA. This reduction in c-MYB proteins level could be blocked with proteasome inhibitor MG132. Meaning we induced the proteolysis of the c-MYB oncogene.

The reduction in c-MYB levels was accompanied by a significant anti-AML activity in an in vitro colony forming assay of both AML cell lines and primary patient samples. Interestingly, the drug had no significant effect on colony formation of CD34 positive cord blood cells.

To investigate the specificity we tried to rescue the function of MYB. Through overexpressing a "stabilized" MYB mutant, lacking the negative regulatory domain we could, in large part, rescue the block in leukemic self-renewal. Meaning, its main effect on the block in AML stem cells is through targeting c-MYB oncogene.

It is vital to know if this drug would be active in patients and we therefore required a model that reflected, as closely as possible, the disease progression in vivo. In order to achieve this, we established a robust xenograft model with human AML cell lines. Next, we genetically modified these cells with a luciferase-expressing lentiviral vector, so that they could be detected within the xenograft host using the IVIS III pre-clinical imaging system. Using this approach, we have shown that oral administration, even when used alone as a monotherapy, causes a strong block in the leukaemia progression (over a 300 fold difference in leukemic burden at day 17 of treatment vs control), resulting in prolonged survival of xenograft hosts.

Here we describe a pipeline to discover drugs that inactivate key transcription factors important in AML. By combining our knowledge of transcription factor behaviour and drug repositioning, we discovered candidate drugs that inactivate key oncogenes. It's important to note that repositioning candidates have been through several stages of clinical development and therefore have well-known safety and pharmacokinetic profiles. Therefore, we hope that these candidates can be taken forward rapidly into clinical trials to improve the patient outcome in AML.