Modeling CML Development and Drug Resistance Using iPSC Technology,

Abstract 3767

Reprogramming of neoplastic cells to pluripotency provides a unique tool to personalize the exploration of tumor pathogenic mechanisms and drug resistance using iPSCs with patient-specific chromosomal abnormalities. We have developed a technology to generate transgene-free iPSCs from bone marrow and cord blood cells employing episomal vectors. Using this approach we created transgene-free iPSCs from a patient with CML in the chronic phase. CMLiPSCs showed a unique complex chromosomal translocation identified in the patinet's marrow sample while displaying typical embryonic stem cell phenotype and pluripotent differentiation potential. Importantly, these CMLiPSCs are devoid of genomic integration and expression of reprogramming factors, which are incompatible for modeling tumor development and drug response (Hu et al. Blood 117:e109). We have also shown that these CMLiPSCs contain the BCR-ABL oncogene without any detectable mutations in its kinase domain. By coculture with OP9, we generated APLNR⁺ mesodermal cells, MSCs, and lin-CD34+CD45+ hematopoietic progenitors from CMLiPSCs, and control BMiPSCs from a normal subject and analyzed the levels of BCR-ABL protein and tyrosine-phosphorylated (pTyr) cellular proteins in the different cell populations. The highest level of BCR-ABL protein expression was found in the in undifferentiated iPSCs, however, the overall cellular pTyr levels was lower than the control BMiPSCs, suggesting that BCR-ABL kinase activity was suppressed in the CMLiPScs. Consistent with these findings, imatinib does not inhibit the growth and survival of these CMLiPSCs. The levels of BCR-ABL protein decreased upon differentiation with a major reduction observed when cells became mesoderm. Following differentiation of CMLiPSC-derived mesoderm into the MSCs and lin-CD34+CD45+ hematopoietic progenitors, the levels of BCR-ABL protein did not change significantly, indicating that the major epigenetic regulation of BCR-ABL expression occurs during the transition to mesoderm. In spite of the decrease in BCR-ABL expression, the total pTyr levels significantly increased following transition of CMLiPSCs to mesoderm and blood cells, suggesting recovery of BCR-ABL kinase activity during differentiation. Interestingly, we found that imatinib had no effect on CFC potential of the most primitive lin-CD34+CD45+ hematopoietic progenitors derived from CMLiPSCs, while significant inhibition in hematopoietic CFC potential was observed when we used the patient's bone marrow cells. Following expansion of lin-CD34+CD45+ progenitors in serum-free medium with cytokines, we found that more differentiated hematopoietic cells became imatinib sensitive. The differential response of progenitors versus more differentiated cells to imatinib recapitulate the clinical observation that CML stem cells display innate resistance to imatinib but their differentiated progenies become sensitive to this BCR-ABL kinase inhibitor. The iPSC-based models provide several advantages for the study of CML pathogenesis. iPSCs can provide an unlimited supply of hematopoietic cells carrying patient-specific genetic abnormalities. Using well-defined temporal windows and surface markers, distinct cell subsets with tumor-initiating/tumor-propagating potential after transplantation in immunodeficient mice could be identified and used for drug screening. iPSC models make it possible to address CML stem-cell potential at various stages of differentiation for which it may be difficult to obtain samples from the patient, for example, at the hemangioblast stage. They also provide a unique opportunity to explore the interplays between epigenetics and oncogene function, as we have demonstrated using the CMLiPSCs. The major unsolved question is why CML stem cells are naturally resistant to imatinib, and this question can be addressed using the iPS system.