Nonsteroidal anti-inflammatory drugs such as sulindac inhibit Wnt signaling, which is critical to maintain cancer stem cell-like cells (CSC), but they also suppress the activity of 5-lipoxygenase (5-LO) at clinically feasible concentrations. Recently, 5-LO was shown to be critical to maintain CSC in a model of chronic myeloid leukemia. For these reasons, we hypothesized that 5-LO may offer a therapeutic target to improve the management of acute myeloid leukemia (AML), an aggressive disease driven by CSCs. Pharmacologic and genetic approaches were used to evaluate the effects of 5-LO blockade in a PML/RARa-positive model of AML. As CSC models, we used Sca-1 þ /lin À murine hematopoietic stem and progenitor cells (HSPC), which were retrovirally transduced with PML/RARa. We found that pharmacologic inhibition of 5-LO interfered strongly with the aberrant stem cell capacity of PML/RARa-expressing HSPCs. Through small-molecule inhibitor studies and genetic disruption of 5-LO, we also found that Wnt and CSC inhibition is mediated by the enzymatically inactive form of 5-LO, which hinders nuclear translocation of b-catenin. Overall, our findings revealed that 5-LO inhibitors also inhibit Wnt signaling, not due to the interruption of 5-LO-mediated lipid signaling but rather due to the generation of a catalytically inactive form of 5-LO, which assumes a new function. Given the evidence that CSCs mediate AML relapse after remission, eradication of CSCs in this setting by 5-LO inhibition may offer a new clinical approach for immediate evaluation in patients with AML. Cancer Res; 74(18); 5244-55. Ó2014 AACR.

Acute myeloid leukemia (AML) is characterized by an aberrant self-renewal of hematopoietic stem cells (HSC) and a block in differentiation. The major therapeutic challenge is the characterization of the leukemic stem cell as a target for the eradication of the disease. Until now the biology of AML-associated fusion proteins (AAFPs), such as the t(15;17)-PML/RARα, t(8;21)-RUNX1/RUNX1T1 and t(6;9)-DEK/NUP214, all able to induce AML in mice, was investigated in different models and genetic backgrounds, not directly comparable to each other. To avoid the bias of different techniques and models we expressed these three AML-inducing oncogenes in an identical genetic background and compared their influence on the HSC compartment in vitro and in vivo.These AAFPs exerted differential effects on HSCs and PML/RARα, similar to DEK/NUP214, induced a leukemic phenotype from a small subpopulation of HSCs with a surface marker pattern of long-term HSC and characterized by activated STAT3 and 5. In contrast the established AML occurred from mature populations in the bone marrow. The activation of STAT5 by PML/RARα and DEK/NUP214 was confirmed in t(15;17)(PML/RARα) and t(6;9)(DEK/NUP214)-positive patients as compared to normal CD34+ cells. The activation of STAT5 was reduced upon the exposure to Arsenic which was accompanied by apoptosis in both PML/RARα- and DEK/NUP214-positive leukemic cells. These findings indicate that in AML the activation of STATs plays a decisive role in the biology of the leukemic stem cell. Furthermore we establish exposure to arsenic as a novel concept for the treatment of this high risk t(6;9)-positive AML.

The repression of the androgen receptor (AR) activity is a major objective to inhibit prostate cancer growth. One underlying mechanism for efficient hormone therapy is based on corepressors that inactivate the AR. In line with this, castration-resistant prostate cancer is associated with malfunction or reduced corepressor action. To overcome this, the overexpression of endogenous corepressors, however, affects many other transcription factors. Therefore, an AR-specific corepressor could be of advantage. Using a yeast peptide aptamer two-hybrid screen with the full-length human AR, we identified a short amino acid-stretch that binds specifically to the human AR in yeast and in mammalian cells and not to the closely related progesterone or glucocorticoid receptors. Furthermore, fused to a silencing domain, this aptamer-based corepressor (AB-CoR) exhibits corepressor activity by inhibiting both the AR-mediated transactivation and expression of the AR target gene PSA. Furthermore, stable expression of the AB-CoR inhibits growth of human LNCaP prostate cancer cells. Moreover, we generated a cell-permeable AB-CoR by fusing a protein transduction domain to establish a vector-free transport system. Treatment of LNCaP cells with the bacterially expressed and affinity-purified cell-permeable AB-CoR peptide resulted in a significant inhibition of both AR-mediated transactivation and prostate cancer cell proliferation. Thus, generation of a novel AR-specific aptamer-based corepressor may present a vector-free inhibition of AR-dependent prostate cancer growth as a novel approach.

<p>Laser scan confocal analysis of 5-LO and RARα protein localization in U937-PR9 cells.</p>

<p>This figure depicts an additional IF analysis on NB4 cells showing the interactioin between b-catenin and 5-LO upon threatment with CJ13,610</p>