ABSTRACT
Screening of primary patient acute myeloid leukemia (AML) cells is challenging based on intrinsic characteristics of human AML disease and patient-specific conditions required to sustain AML cells in culture. This is further complicated by inter- and intra-patient heterogeneity, and "contaminating" normal cells devoid of molecular AML mutations. Derivation of induced pluripotent stem cells (iPSCs) from human somatic cells has provided approaches for the development of patient-specific models of disease biology and has recently included AML. Although reprogramming patient-derived cancer cells to pluripotency allows for aspects of disease modeling, the major limitation preventing applications and deeper insights using AML-iPSCs is the rarity of success and limited subtypes of AML disease that can be captured by reprogramming to date. Here, we tested and refined methods including de novo, xenografting, naïve versus prime states and prospective isolation for reprogramming AML cells using a total of 22 AML patient samples representing the wide variety of cytogenetic abnormalities. These efforts allowed us to derive genetically matched healthy control (isogenic) lines and capture clones found originally in patients with AML. Using fluorescently activated cell sorting, we revealed that AML reprogramming is linked to the differentiation state of diseased tissue, where use of myeloid marker CD33 compared to the stem cell marker, CD34, reduces reprogramming capture of AML+ clones. Our efforts provide a platform for further optimization of AML-iPSC generation, and a unique library of iPSC derived from patients with AML for detailed cellular and molecular study.
