Engineering of Induced Pluripotent Stem Cell-Derived Hemogenic Endothelium

Background: Autologous cellular therapy of hematological malignancies and primary immunodeficiencies has been limited by the inability to produce functional hematopoietic stem cells (HSCs) from patient-derived induced pluripotent stem cells (iPS). Previous approaches have failed to demonstrate multi-lineage potential or secondary bone marrow engraftment in murine hosts - the gold standards for demonstrating HSC self-renewal and differentiation. Building upon recent evidence that HSCs derive from the definitive hemogenic endothelium (HE) of the embryonic lateral plate mesoderm, we set out to engineer iPS-derived HSCs by directed differentiation, and to investigate potential applications for modeling genetic blood diseases in vitro .

Methods: Following embryoid body formation, human iPS derived from bone marrow mesenchymal stem cells underwent stepwise, morphogen-directed differentiation into HE, followed by polycistronic lentiviral transduction with five transcription factors (RUNX1, ERG, LCOR, HOXA5, HOXA9) on day 3 of endothelial-to-hematopoietic transition (EHT) induction, to enable further conversion to HSC-like cells with lymphoid potential in vitro and known multi-lineage engraftment capacity in vivo . A candidate screen of 17 cytokines, epigenetic modifiers and small molecules aimed at inducing HE proliferation was conducted during EHT induction in a RUNX1c gene reporter assay. To model human disease, dermal fibroblast-derived iPS from a hypomorphic RAG2 Omenn syndrome patient were differentiated into HSCs, and cocultured with MS5 or OP9-DL1 stromal cells to induce B or T cell development. Fluorescence-activated cell sorting (FACS) was performed weekly for 6 weeks after initiating lymphoid differentiation.

Results: In the candidate screen, IFN-γ enhanced RUNX1c reporter 2.4-fold during EHT, which was confirmed by flow cytometry showing increased hematopoietic progenitor cells (CD34⁺CD45⁺). Several compounds were suppressive to EHT, including IFN-α, which abrogated EHT enhancement when combined with IFN-γ. Despite 1 week of lymphoid-directed differentiation, cell populations derived from mutant RAG2 iPS cells showed a predominance of myeloid progenitors (CD38+CD45+CD45RA+), with only 5% of progenitors manifesting a lymphoid phenotype (CD10+CD34+CD45RA+). In contrast, wildtype iPS-derived HSCs produced CD19+IgM- B cell progenitors as well as mature IgM+ B cells (8% versus .06% of CD19+CD21+ B cells), which were comparable to cord blood controls; CD3+CD4+CD8+ T cells were also identified after 3 weeks. Ultimately, we determined that RAG2 -mutant iPS were unable to generate CD19+ B cells and arrested at the early T cell stage (CD5+CD4-CD8-).

Conclusion: We determined that transient exposure to IFN-γ greatly improved the EHT of HE, which may bolster the typically low yield of iPSC-derived HSCs. In a lymphoid differentiation assay, wildtype iPS-derived HE were comparable to CB in the formation of mature and immature B/T cells, whereas RAG2 -derived HEs arrested at the pro-thymocyte stage. This finding suggests that our approach to creating engraftable HSCs holds significant potential for modeling hematopoietic disease, both in animal models and as a "disease-on-a-dish," and for developing therapeutic strategies in genetic blood disorders.