Introduction:Disease recurrence in patients with acute myeloid leukemia (AML) has been attributed to the self-renewal, self-protective, and quiescent characteristics of leukemia stem cells (LSC). There is currently an unmet clinical need in AML for novel targeted therapies capable of eradicating LSC to achieve long-term cure. Curaxins were identified during a cell-based screening as small molecules able to bind to DNA and alter the chromatin structure without directly causing DNA damage. Curaxin binding leads to functional inactivation of the facilitated chromatin remodeling complex (FACT) involved in transcription, replication, and DNA repair. FACT, which normally binds to DNA in a transient manner, gets trapped on the DNA following curaxin binding, leading to activation of the pro-apoptotic p53 pathway, suppression of anti-apoptotic NF-κB pathway, and FACT dependent cell death. Structurally, FACT is a histone chaperone composed of two subunits, structure specific recognition protein 1 (SSRP1) and suppressor of Ty16 (SPT16). FACT controls the assembly of nucleosomes only for cells with open dynamic chromatin states (such as embryonic stem cells and tumor cells) but not normal mature cells. FACT is an attractive candidate for leukemia therapeutics due to its high specificity for leukemia cells and LSC, with limited expression on normal hematopoietic stem cells. Here we evaluated the effects of the small molecule curaxin, CBL0137 (Cleveland Biolabs, Buffalo, NY) in vitro and in vivo in preclinical human AML models. We hypothesized that AML cells, specifically leukemia-initiating cells from primary samples, might be sensitive to apoptosis induced by curaxin treatment due to underlying FACT expression.Materials and Methods:Human AML cells (HL60 and a multi-drug resistant clone HL60/VCR) were treated in vitro with CBL0137 (doses 0.039-20μM) and assessed by WST-8 viability assays (Dojindo Molecular Technologies). Expression of both FACT subunits (SSRP1 and SPT16) was evaluated by Western blot (WB) analysis in 23 primary AML samples. A subset of primary AML cells was then plated in methocellulose culture with cytokine support. The number of leukemia colony forming unit (CFU-L) potentially reflecting leukemia initiating cell growth was quantified by microscopy after 14-15 days of incubation. In vivo treatment with CBL0137 was also assessed in SCID mice systematically engrafted with HL60/VCR-luciferase AML cells. Mice were treated with oral vehicle or CBL0137 dosed at 40 mg/kg (5 days/week for 2 weeks) or 70 mg/kg (twice a week for 5 doses). Effects on leukemia burden, toxicity, and survival were determined by whole animal bioluminescent imaging, total animal weights, and time to morbidity, respectively.Results:Expression of one or both FACT subunits was detected in 22 of the 23 primary AML samples and in HL60 cells by WB. CBL0137 resulted in dose-dependent in vitro inhibition of HL60 and HL60/VCR cell growth with IC50 values of 0.54-57 μM. CBL0137 also inhibited the colony formation ability of 9 primary AML samples at doses ranging from 0.625 - 10μM with estimated IC50 values of 1-3 μM. Of note, CBL0137 treatment in CFU-L assays resulted in similar anti-leukemic activity in patient samples characterized by low/undetectable or high/intermediate levels of FACT subunits (SSRP1 and SPT16) by WB analysis. In vivo, CBL0137 treatment in systemic HL60/VCRluciferase engrafted SCID mice was generally well tolerated with no significant weight loss. Curaxin treatment markedly decreased leukemic disease burden (as determined by bioluminescent imaging) and prolonged survival as compared with control-treated mice (p =0.0004). Animals receiving the higher dose CBL0137 regimen exhibited the greatest reduction in disease burden and longest overall survival (70 vs. 40mg/kg, p=0.04)Conclusions:Our results demonstrate that treatment with the curaxin, CBL0137, results in potent in vitro and in vivo dose-dependent anti-tumor activity in preclinical AML models. Inhibition of primary AML colony formation did not appear to correlate with baseline FACT expression in tumor cells. Taken together, these data support the future clinical development of this novel DNA interacting agent for AML therapy. Correlative studies using primary AML xenograft (PDX) models to clarify if differential FACT expression is a predictive marker for in vivo curaxin efficacy are underway. DisclosuresNo relevant conflicts of interest to declare.