Induced Pluripotent Stem Cells to Model Blood Diseases

Our group is developing induced pluripotent stem cell (iPSC) models of myeloid malignancies, including MDS, MPN and AML. We are generating iPSCs from the bone marrow or blood of patients, which can be maintained indefinitely as pluripotent cell lines and, upon in vitro differentiation along hematopoietic lineages, exhibit hallmark features of these diseases. By integrating mutational analyses with cell reprogramming we can derive iPSCs capturing dominant clones, subclones and normal cells from the same patient and thus have established a collection of iPSC lines representing distinct disease stages along the spectrum of myeloid transformation: predisposition syndromes/preleukemic cells/clonal hematopoiesis; low risk MDS; high risk MDS; and MDS/AML. In parallel, we are using the CRISPR/Cas9 system to introduce or correct mutations in normal or malignant iPSCs, respectively, in isogenic settings and sequential CRISPR gene editing to model mutational cooperation. We recently reported that iPSC lines derived from patients with AML re-establish upon differentiation a leukemic phenotype characterized by extensive proliferation of immature myeloid cells that serially transplant a lethal leukemia into NSG mice (Kotini et al. Cell Stem Cell 2017). Strikingly, we observed that the AML-iPSC-derived hematopoietic stem/progenitor cells (HSPCs) contain two morphologically and immunophenotypically distinct cell subpopulations: a cell fraction (adherent, A) exhibiting adherent growth and containing immature cells with an HSC immunophenotype (CD34⁺/CD38^-/CD90⁺/CD45RA^-/CD49f⁺); and a non-adherent fraction (suspension, S) of more differentiated cells. Fate-tracking experiments revealed a hierarchical organization, with the A cells renewing themselves and continuously giving rise to the S cells through symmetric and asymmetric divisions. The NSG engraftment potential was largely contained within the adherent cell fraction. Thus, AML-iPSCs exhibit the hallmarks of a leukemia stem cell (LSC) model, namely phenotypic and functional heterogeneity and hierarchical organization, with the A fraction containing LSCs that serially transplant leukemia and give rise to more differentiated cells (S fraction) without engraftment potential. LSCs are believed to be a prominent source of AML relapse, but their rarity and the unavailability of universal and specific immunophenotypic markers prohibits their prospective isolation and makes the study of their properties challenging. This new iPSC-based AML-LSC model enables us for the first time to prospectively obtain large numbers of genetically clonal human LSCs and perform genome-wide integrative molecular analyses and large-scale screening to identify key molecular mechanisms sustaining the properties of LSCs as potential new therapeutic targets. Using this model we characterized the effects of previously proposed compounds with LSC selectivity in self-renewal vs differentiation of LSCs. We also screened a small molecule library of 1280 compounds in the A and S cells in parallel to identify compounds with selectivity for the former as candidates for LSC-specific targeting.