A High-Throughput Screen to Identify Compounds Preserving Primary Human AML Stem Cells Ex-Vivo,

Abstract 3587

Acute myeloid leukemia (AML) is the second most frequent leukemia in adults with still high relapse rates supposed to be due to a rare population of leukemia stem cells (LSC) within the leukemic bulk not responding to conventional anti-cancer drugs. Engraftment of human leukemic cells in immunocompromised mice mimicking the human disease is considered to prove the existence of stem cells in the sample; however, the engraftment potential of primary human AML cells is usually quickly lost upon ex-vivo culture indicating that LSC have undergone differentiation or apoptosis and thus lost their stem cell properties. This lack of appropriate culture conditions enabling ex-vivo maintenance of primary human LSC is a major obstacle in the development of LSC targeted therapies. Our aim is, therefore, to identify conditions which allow the maintenance and expansion of these rare cells ex-vivo to understand their unique properties and to further elucidate cellular pathways underlying the self-renewal program of human primary LSC. We established a flow cytometry based high-throughput screen in which we tested more than 6,000 selected compounds with regards to their potential to expand primary human AML cells while preserving their phenotypic and functional characteristics. We considered alteration of surface marker expression, i.e. loss of CD34 and acquisition of markers of myeloid maturation initially not expressed on the cells, as indicator of differentiation implicating at the same time a loss of stem cells in the culture. We hypothesized that compounds that would be able to prevent these alterations in surface marker expression could have the potential to promote self-renewal of primary human LSC ex vivo. A normal-karyotype AML representing the most frequent cytogenetic group of AML in adults with FAB M1 morphology was chosen to perform the screen. Moreover, the initial CD34+CD15- phenotype allowed us to monitor changes in surface marker expression. Cells were plated in 384 well plates with a density of 5,000 cells per well. 6,160 compounds were tested with and without an inhibitor to aryl hydrocarbon receptor (AhR) and the vehicle DMSO was used as negative control resulting in a final DMSO concentration of ∼0.1% in each well, which had been proven to be not toxic to the cells. After 5 days of incubation the cells were analyzed on a BD LSRII high-throughput flow cytometer with regards to CD34 and CD15 surface expression. Compounds showing an increase of ≥45% of CD34+ and CD34+CD15- cells over DMSO compared to the AhR antagonist were selected for further evaluation. In a secondary screen we retested 32 compounds in 5 different concentrations ranging from 1:3⁴ to 3x based on the initial screen concentration after 5 and 9 days of incubation and confirmed an increase ≥50% of relative and absolute CD34+CD15- leukemic cells after 5 days for 13 compounds and after 9 days for 7 compounds. We have initiated functional in-vitro and in-vivo experiments to validate these results and to prove the existence of LSC in cultures where compounds have been added. The ability of the newly identified compounds to preserve human LSC ex-vivo will be further evaluated using a collection of primary AML samples comprising different morphological and cytogenetic groups. Together, our observations document the feasibility of finding anti-differentiation (pro-self-renewal?) compounds of primary human AML cells that can be expanded ex-vivo. These studies represent a first key step in the identification of specific LSC targeting compounds.