Phenomic Screen in Vivo and in Vitro to Explore Novel Pathogenesis of AML1-ETO-Positive Leukemia Using PSC-Derived Hematopoietic Cells

Onset of acute myeloid leukemia (AML) has been accounted for by cooperation between multiple genetic alterations. Among the previously listed leukemogenic lesions, AML1-ETO fusion (AE) generated by translocation (8;21) (q22;q22) is one of the common mutations observed in 20-40% of patients. AE affects transcriptional regulation associated with hematopoietic differentiation, while 60% of AE-positive AML cases are shown to have together other types of mutation of genes involved in cell proliferation, such as receptor tyrosine kinase (RTK) c-kit and FLT3. Those data are compatible with so-called “multi-step leukemogenesis” model. At the same time, variety in clinical phenotypes even among patients harboring same sets of mutation strongly indicates the yet unknown “cooperative ” cues in genetics or epigenetics. From this viewpoint, finely elucidating “What kinds of such mechanisms” give “What kinds of impact associating with AE fusion” on “What kinds of target cells” during leukemogenesis is important.

In order to address those problems in reproducible manner, we developed the novel phenomic screen system by combining reverse and forward genetic modifications with pluripotent stem cell (PSC)-derived hematopoietic culture. We induced the PiggyBac-based random knockdown motifs into PSC-derived hematopoietic cells that harbour artificial expression cassettes for AE fusion gene, then applied them for in vitro and in vivo assays. We performed these platform-based trials in parallel to the work with positive control cells carrying AE in addition to c-kit or Flt3-ITD mutations, recently identified poor prognosis conveyers among AE-positive AML. First, serial replanting assays in methylcellulose-containing semisolid media as well as liquid culture revealed that a few numbers of clones acquired significantly reinforced potential in cell growth and colony forming efficacy. As for positive control cells, they showed reproducibly strong tendency toward both increased colony forming efficacy and suppressed differentiation. Those results suggested the successful recapitulation of pathogenic cooperation between AE and RTK mutations in our PSC-derived hematopoietic cells, and confirmed the feasibility of our screening system. Interestingly, AE-positive myeloid lineage-committed progenitors as well as immature multipotent hematopoietic stem and progenitor cells (HSPCs) showed higher replating colony forming efficacy, which may indicate the representation of higher incidence of myeloblastic leukemia in AE-positive AML. We are identifying the knocked down genes and evaluating the perturbation of expression patterns of genes involved in differentiation, proliferation and survival as well as epigenetics associating genes within affected cells.

In addition to in vitro assays, we next evaluated the feasibility of in vivo phenomic screening by performing intra bone marrow transplantation of gene modified PSC-derived progenitors into immunodeficient NOD/Shi-scid,IL-2RγKO Jic(NOG) mice. Also in these experiments, we successfully observed the cooperation between AE and RTK mutations for prolonged engraftment in case HSPCs and myeloid lineage-committed progenitors were transplanted. After that, we are now performing these screenings to select the phenotype acquiring clones.

In conclusion, we successfully established a novel phenomic screen in vivo and in vitro to explore novel pathogenesis of AE-positive leukemia using PSC-derived hematopoietic cells. After confirming the recapitulation of already known cooperation between AE and RTK mutation, we are currently extending the screening using this platform. We believe that our model must allow us to better match treatment to prognosis across the disease spectrum via comprehensive understanding of pathogenesis.