Cytokine Independent Production of Enucleated Red Blood Cells from Induced Pluripotent Stem Cells

Induced-Pluripotent Stem cells (iPSCs) are an attractive source of cells to produce cultured Red Blood Cells (cRBCs) because they are immortal, karyotypically stable and easy to produce and genetically modify. We have developed a chemically-defined Robust Erythroid Differentiation (RED) protocol to differentiate iPSCs into enucleated erythroid cells. The RED protocol yield more than 100,000 cRBCs per iPSCs, do not require any albumin and only small amounts of transferrin (Tf) because supplementation with an iron chelator, allows Tf recycling to take place in cell culture.

Mutations in the c-kit gene (the SCF receptor) that are often found in leukemias and mastocytosis are known to render the SCF receptor constitutively active. Similarly mutations in the Jak2 gene, which transduced the erythropoietin (EPo) signal, are associated with Polycytemia Vera and erythropoietin independence. Since Stem Cell Factor (SCF) and Epo in combination with glucocorticoids are sufficient to drive erythropoiesis and since the SCF and Epo receptors can hetero-dimerize, we have introduced into iPSCs derived from two group O negative blood donors a D816V mutation in the kit gene and a V617F in the Jak2 gene using CRSPR Cas9 and assessed their differentiation using the RED protocol. Unedited cells grown in the absence of cytokines died rapidly. By contrast edited iPSCs could be differentiated into late erythroid progenitors in the complete absence of SCF and Epo. Hemizygous and heterozygous clones for the kit mutation yielded more that 200,000 cells per iPSCs, a yield that was identical to that of the control cells grown with cytokines. Homozygous clones also grew exponentially but died massively at the end of the culture. Fifteen markers spectral flow cytometry analysis and dimensionality reduction analysis revealed that the evolution of the cells in culture could be recapitulated by dividing the cells into 8 major populations defined by expression of CD43, CD34, CD31, CD45, CD235a, CD71 and CD117. At day 10, the phenotypes of the cultures of the unmodified and of the mutated lines were similar and composed mostly of populations of CD43+CD45-CD34+ HPSCs which differed by expression of CD235a. In the control cells, a population of definitive CD45+ HSPCs became predominant at days 17 and 24 and eventually differentiated into mature erythrocytes. In the modified lines, the CD45+ HSPCs were present but did not amplify as much. Instead, a pro-erythroblast population became pro-eminent and differentiated into mature erythroblasts in the case of the hemizygous clone, but not in the case of the homozygous clone. We concluded that it is possible to generate mature erythrocytes from iPSCs without the use of any SCF and Epo but that excessive stimulation by the constitutive SCF receptor led to rapid cell death at a late stage of erythroid differentiation.

Determination of the rate of enucleation using DRAQ5, a cell permeant DNA dye, or Romanovsky staining revealed that the control iPSCs and the hemizygous and heterozygous clones enucleated at a similar rate of about 20% and that addition of dasatinib, an inhibitor of SCF signaling, accelerated the differentiation and increased the enucleation rate to about 25%. We concluded from these experiments that genetically modified iPSCs carrying mutations can expand and enucleate in the absence of SCF and Epo at about the same rate as control cells.

As expected, HPLC analysis revealed that control cRBCs expressed about 94% gamma-globin and 6% adult globin. By contrast, RBCs derived from the genetically modified cells expressed about 75% gamma-globin, 22% epsilon-globin and 3% beta-globin. We conclude that erythroblasts derived from genetically modified iPSCs are a mixture of primitive cells which cannot enucleate and of definitive cells that can. Low-pass sequencing analysis revealed that the genetically modified clones were karyotypically stable. Immunophenotyping of the cRBCs using the Immucor platform revealed that the cRBCs expressed the same antigens as the original blood donors.

In conclusion, introduction of mutations in the kit and jak2 genes allows the production of cRBCs from iPSCs without any SCF or Epo. This considerably reduces the cost of production of cRBCs since these cytokines are the most expensive components in the RED protocol. In turn this should facilitated the implementation of the many applications of iPSC-derived cRBCs.

No relevant conflicts of interest to declare.