Using human genetic variation to improve red blood cell production from stem cells

Multipotent and pluripotent stem cells have significant potential as sources for cell replacement therapies. However, the low yield and quality of in vitro differentiated cells produced from various stem cell sources presents a significant limitation for therapeutic applications. The most mature use of these stem cell products is in the field of transfusion medicine, where stem cell-derived red blood cells (RBCs) have clinically-proven potential as alternative transfusion products. To improve upon current approaches for RBC production, we used insight from both common and rare human genetic variation of blood counts to focus on the SH2B3 gene. By producing loss of function of SH2B3 using targeted knockdown and genome editing approaches in human hematopoietic stem and progenitor cells, as well as human pluripotent stem cells, we are able to significantly improve both the quality and yield of in vitro derived RBCs. We illustrate how insight from human genetic variation can assist in the development of broadly applicable approaches that have tremendous value for regenerative medicine. CD34+ cells from G-CSF mobilized peripheral blood was purified by positive magnetic selection using an Ultrapure Microbead Kit (Miltenyi Biotech) according to the manufacturer's protocol following mononuclear cell purification on a Ficoll Density Gradient. At least 95% purity was reached as assessed by post-purification flow cytometry with a PE conjugated anti-human CD34 antibody (8012-0349, eBioscience). shRNA constructs targeting SH2B3 were used with the following sequences: sh83 (CCGGCCTGACAACCTTTACACCTTTCTCGAGAAAGGTGTAAAGGTTGTCAGGTTTTTG) and sh84 (CCGGGCCTGACAACCTTTACACCTTCTCGAGAAGGTGTAAAGGTTGTCAGGCTTTTTG). Both constructs were in the pLKO.1-puro lentiviral vector. The lentiviral vectors pLKO-GFP and pLKO.1-puro Luciferase served as controls (the RNAi consortium of the Broad Institute of MIT and Harvard). At day 7 post infection, cells were isolated and Human Gene 2.0 ST microarrays were performed on RNA.