Patient-Derived iPSCs Faithfully Represent the Genetic Diversity and Cellular Architecture of Human Acute Myeloid Leukemia

The reprogramming of malignant cells to pluripotency has presented considerable challenges, hindering the application of induced pluripotent stem cell (iPSC) modeling technology to human leukemia. In addition, how well iPSC-derived cells resemble their primary counterparts is currently largely unknown. We developed a reprogramming method tailored to human malignancies, "Complete Capture of Mutational Burden" (CCoMB), that combines comprehensive genetic characterization of the starting sample and inference of its clonal architecture with large-scale screening of clonally reprogrammed colonies. Using this method we were able to generate a panel of iPSC lines for the modeling of human acute myeloid leukemia (AML) representing all major genetic groups of AML. Specifically, we derived iPSC lines from 15 patients representing all major genetic groups of AML - PML-RARA; chromatin-spliceosome; TP53-mutated/aneuploidy; AML1-ETO, MLL-rearranged; NPM1-mutated and others - collectively capturing 21 distinct genotypes and 24 driver genetic lesions (mutations, translocations, deletions). Matched normal iPSCs were derived from 7 of these patients. Reprogramming to iPSCs captured both preleukemic (CH/initiating mutation only) and fully leukemic clones (baring the full set of patient mutations). In almost all cases, reprogramming informed reconstruction of the evolutionary hierarchy of the AML, with unexpected hierarchies unveiled in 4 of the cases. These AML-iPSCs retain genetic fidelity and, upon in vitro hematopoietic differentiation, produce hematopoietic cells with hallmark phenotypic leukemic features, including serial engraftment of a lethal myeloid leukemia into immunocompromised mice and extended self-renewal in vitro. Transplantation of cells derived from iPSCs representing two distinct AML clones from each of 3 patients revealed that these mimic the clonal dynamics in the patients, with more advanced clones showing increased representation in the xenografts, compared to the earlier clones. To compare the iPSC-derived to the primary leukemias, we performed single-cell transcriptomics analyses in patient-matched iPSC-derived and primary leukemic cells from 3 patients, both ex vivo/in vitro and after transplantation into NSGS mice. Clustering analyses identified cell types corresponding to primitive hematopoietic stem cell (HSC)/ multipotent progenitor (MPP), hematopoietic progenitor cells (HPCs) and more mature myelomonocytic lineage cells in all samples from all patients, at varying frequencies. Leukemias derived through in vitro differentiation from iPSC lines exhibited both similarities, as well as differences, in their cellular composition and transcriptome, compared to the patient-matched ex vivo leukemias. However, upon transplantation of the same leukemias into NSG-3GS mice, iPSC-derived xenografts were strikingly similar to the patient-derived xenografts. iPSC-derived leukemic cells exhibited a more stem/progenitor cell phenotype in vitro, with progressive maturation along the myeloid lineage upon primary and, even more, upon secondary transplantation, mimicking primary xenografts. In summary, our results reveal very few true biological barriers to the reprogramming of AML cells (with NPM1 mutations the most notable) and show that AML-iPSC-derived leukemias faithfully mimic the primary patient leukemias upon xenotransplantation. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal