Modeling Chronic Myelomonocytic Leukemia Through Patient-Derived Induced Pluripotent Stem Cells

Chronic myelomonocytic leukemia (CMML), the most common entity among myelodysplastic syndrome/myeloproliferative neoplasm, is characterized by monocytosis, morphologic dysplasia, and progression to acute myeloid leukemia. Despite its relatively high incidence, the pathogenesis of CMML remains elusive, mainly due to the paucity of suitable animal models and the difficulties in the establishment of CMML cell lines. As reprogramming technology has been established as a new powerful tool for disease modeling, here we developed induced pluripotent stem cells (iPSC) from CMML leukemic cells. By introducing episomal vectors expressing OCT3/4, SOX2, KLF4, L-MYC, LIN28 combined with shRNA for p53, several lines of iPSC were generated from CD34⁺cells of a healthy donor (wild-type: WT) and a CMML patient with der (1; 7) (q10; p10) translocation. All iPSC expressed pluripotent surface markers (TRA1-60 and SSEA4) and stemness-related genes (NANOG, OCT3/4, SOX2, KLF4, C-MYC, and REL). The der (1; 7) (q10; p10) translocation were detected in all CMML iPSC.

When co-cultured with 10T1/2 stromal cells in the presence of VEGF, CMML and WT iPSC generated a comparable frequency of CD34⁺ CD43⁺ hematopoietic progenitor cells (HPC). However, when cultured in cytokine-supplemented semisolid medium, CMML iPSC-derived CD34⁺ CD43⁺ HPC yielded an increased number of hematopoietic colonies with larger sizes, especially CFU-GM and CFU-GEMM, compared with WT iPSC-derived cells. Importantly, CMML iPSC-derived hematopoietic colonies mainly consisted of monoblasts with a high nucleus/cytoplasm ratio, while those of WT iPSC predominantly composed of macrophages. Flow cytometric analysis showed marked increases of CD34⁺ hematopoietic progenitors and CD13⁺ myeloid cells in CMML iPSC-derived hematopoietic colonies. Of note, among myeloid lineages, there were remarkable increases in CD14⁺ monocytic cells and CD24⁺ CD14^- immature granulocytes, which were unique characteristics of human CMML. In addition, CMML iPSC-derived CD13⁺myeloid cells exhibited a weak expression of CD56, which was never detected in WT iPSC-derived cells, suggesting CMML iPSC-derived hematopoietic cells recapitulate the original phenotype of CMML. Notably, we found that CMML iPSC-derived HPC retained the ability to serially replate and generate colonies even after the fourth plating, although essentially no WT iPSC-derived colonies were detected after the second plating, suggesting enhanced self-renewal capacity of CMML iPSC-derived HPC. In addition, when cultured in methylcellulose without cytokines, CMML iPSC-derived HPC were able to form spontaneous hematopoietic colonies, in contrast to WT iPSC-derived cells that gave rise to almost no detectable colonies. Taken together, these results suggest CMML iPSC-derived HPC possess the multiple biologic properties of CMML leukemic blasts.

Then, we conducted a comprehensive gene expression and DNA methylation profiling of WT and CMML parental CD34⁺ cells, iPSC, and iPSC-derived CD34⁺ CD43⁺ HPC. Strikingly, although gene expression and DNA methylation status were quite different between WT and CMML parental CD34⁺ cells (R² = 0.72 for gene expression and 0.90 for DNA methylation), WT and CMML iPSC-derived HPC exhibited similar gene expression and DNA methylation pattern (R²= 0.92 and 0.96), indicating reprogramming followed by redifferentiation may enable to obtain more homogenous population of normal and CMML cells that reside in almost the same differentiation stage. Using these multi-omics platforms, we searched differentially expressed and methylated genes between WT and CMML iPSC-derived HPC to identify molecular abnormalities that contribute to the pathophysiology of CMML. Indeed, gene set enrichment analysis revealed that embryonic stem cell (ESC)-related gene sets were enriched in CMML iPSC-derived HPC, which is consistent with a previous report that showed murine MLL leukemia stem cells employ a transcriptional program shared with ESC. Functional analysis of candidate genes is underway to further define the molecular aberrations involved in the CMML pathogenesis.

In summary, we have established a novel CMML model of patient-derived iPSC and revealed their derived HPC recapitulate the disease’s primary features. These findings highlight the iPSC as an attractive platform to investigate the cellular and molecular pathophysiology of CMML.