Functional Early Hematopoietic Progenitor Cells Derived From Mouse Embryonic Stem Cells and Induced Pluripotent Stem Cells

We have previously reported a method to produce early hematopoietic progenitor cells from C57BL/6J-derived mouse embryonic stem cells (mESCs). After co-culture on OP9 stromal cells for one week, four different C57Bl/6 mESC lines consistently differentiated into hematopoietic progenitors, as determined by immunophenotyping; we detected cells that mark as KLS (Lin^- Kit⁺ Sca⁺), CMPs, GMPs, and MEPs (but not SLAMs) from all four lines. In addition, functional progenitors for erythrocytes, monocytes, and mast cells (by morphology and immunophenotyping) were detected after another week of culture in methylcellulose with hematopoietic cytokines (SCF, IL-3, IL-6, and Epo). These findings were replicated using four different lots of fetal bovine serum, and with three different lots of OP9 cells from ATCC. We injected 1x10⁶ “OP9-induced” progenitor cells retroorbitally into unconditioned NSG mice, and detected multilineage hematopoietic engraftment (myeloid compartments marked by CD34, CD11b, Kit, and Gr-1, lymphoid compartments marked by CD3 and B220, and erythroid compartments marked by Ter119) in the bone marrow and/or spleens of 10 out of 19 recipients at 3 months.

Using the OP9 co-culture system, we have differentiated miPSC clones from three independent iPSC experiments, using an integrating polycistronic lentivirus expressing OCT4, SOX2, and KLF4 as the reprogramming vector. One set of miPSC clones was produced from mouse embryonic fibroblasts (MEFs) from pooled C57BL/6J embryos, and two sets were made from adult mouse fibroblasts derived from a single animal, producing 6, 12, and 12 independent iPS clones for analysis, respectively. All thirty clones had pluripotent features, as determined by alkaline phosphatase staining and immunophenotyping (SSEA1, Oct4, and Nanog). We have injected the OP9-induced progenitors derived from one miPSC clone into NSG mice; thus far, 2 out of 14 recipients have demonstrated engraftment in the peripheral blood. However, the efficiency of hematopoietic progenitor generation with OP9 induction (based on the immunophenotyping and progenitor assays noted above) was highly variable for miPSCs from all three experiments. Among all three sets of miPSC clones, 18/30 exhibited differentiation efficiencies comparable to wild-type B6 ESCs, 5/30 clones exhibited moderately reduced differentiation efficiencies, 5/30 clones exhibited markedly reduced differentiation efficiencies, and 2/30 clones (from two different iPSC experiments) did not produce any detectable hematopoietic progenitors with OP9 induction. These phenotypes were stable and highly reproducible. The 2 clones that did not yield any hematopoietic progenitors had robust pluripotency marks, and one that was injected into the hindflank of NSG mice produced cystic teratomas. We found that 2% DMSO pretreatment of mESCs for 24 hours prior to OP9 co-culture improved the differentiation efficiency of wild-type B6 ESCs by 50% (Chetty et al. Nature Methods 10(6):553-6, 2013), but it did not rescue the phenotype of miPSC clones that did not produce hematopoietic progenitors. We are currently performing exome sequencing on the 24 miPSC clones from the adult fibroblast reprogramming experiments to determine whether phenotypic heterogeneity is due to specific mutations in the iPSC clones (Young et al. Cell Stem Cell 10(5):570-82, 2012).

In summary, we have developed a simple system to derive functional early hematopoietic progenitor cells from mouse embryonic stem cells and/or induced pluripotent stem cells. OP9-induced progenitor cells engraft into NSG mice without the need for forced expression of HoxB4 (Wang et al. Proc Natl Acad Sci USA 102(52):19081-6, 2005). We have detected functional heterogeneity in miPSC clones derived from the same parental cells, which could be due to genetic variation in the founding cell from which each clone was derived, to different integration sites of the OSK lentivirus in each clone, or to as yet undefined epigenetic mechanisms. Exome sequencing may help to resolve this issue. Regardless, this approach could be a valuable tool for studying the hematopoietic development of a variety of mESC lines and/or miPSC lines derived from genetically altered mice.