Abnormal activity of a subset of transcriptional factors (TF) is a nearly universal property of tumor cells. However, development of drugs targeting TF remains unresolved problem due to complex interface of TF/DNA interactions. Furthermore, many TFs have functionally redundant activities and modulation of several tumor-related TF is a highly desirable strategy with no obvious rational solution. Presuming the existence of presently unknown targetable mechanisms enabling simultaneous modulation of several cancer-related TFs, we used cell phenotype-based screening of small molecules capable of simultaneous activation of tumor suppressor p53 and inhibition of pro-survival TF, NF-κB. Identified molecules, curaxins (CXs), demonstrated remarkable anti-cancer activity in a broad set of preclinical models and a pharmacologically optimized CX compound, CBL0137, is now being evaluated in Phase I clinical trials (NCT01905228). Study of the mechanism of action of CX revealed that they inhibit activity of chromatin remodeling complex FACT (Facilitates Chromatin Transcription).We found that FACT is not needed for general transcription but cooperates with several specific TF that are normally involved in embryonic development, response to stresses and commonly deregulated in cancer. In line with this, we found that FACT is essential for the survival of tumor cells, including tumor initiating cells (TIC), but not for nrmal cells, including normal hematopoietic stem cells (HSC). Since CX possess FACT-modulating activity, we hypothesize that its anti-cancer effect is exerted through FACT inhibition that result in simultaneous modulation of activity of several TF, critically important for the tumor, but not normal cells. The goal of this study is to facilitate clinical advancement of CX through development of predictive and pharmacodynamics markers of curaxins efficacy and better understanding of selective toxicity of CX to tumor vs normal cells. We chose to focus this study on acute myeloid leukemia (AML) which is among major candidate indications for Phase II trial of CBL0137 due to: (i) unmet clinical need; (ii) strong efficacy of CX in preclinical testing, (iii mostly wild type p53 status, and (iv) evidence of powerful induction of p53 in PBMC in patients in Phase I trial. We propose to: 1. Explain the reasons of differential toxicity of CX to leukemia initiating cells and HSC; 2. Evaluate FACT as a prognostic marker of tumor sensitivity to CX treatment. 3. Develop pharmacodynamic (PD) marker of CX activity in vivo.