Red blood cell transfusion is an essential therapy for patients with sickle cell disease (SCD). A major problem is the complexity of antibodies found in patients with SCD, in large part due to RH gene variation. Rh antibodies are common despite provision of serologic Rh-matched transfusions. The antibodies may be mistaken for autoantibodies when in fact they are alloantibodies to high prevalence Rh antigens that are absent or have altered Rh epitopes on the patient's RBCs. Finding compatible red cell units is often complicated by a lack of reagent red cells to properly identify the fine antibody specificity. This can delay patient care, increase costs, and make transfusion therapy impossible for some patients. We hypothesized that human induced pluripotent stem cells (iPSCs) could be reprogrammed from rare donors or genetically engineered to produce standard and reliable red cell reagents to aid identification for patients with complex antibodies. However, most human iPSC-derived red blood cells (iRBCs) are developmentally primitive red cells that differ from donor-derived red cells with high levels of embryonic globin expression, are larger in size and fail to enucleate in culture. We aimed to improve iPSC hematopoietic differentiation protocols to generate iRBCs compatible with standard blood bank assays.

We designed a panel of customized iPSCs that included 1) Rh null cells, 2) cells lacking the high prevalence Rh antigen termed hrS, 3) cells expressing a partial C antigen and lacking the high prevalence Rh antigen termed hrB and, 4) cells expressing the low prevalence Rh antigens V, VS and lacking hrB. For the Rh null line, we identified wild type iPSC lines that were group O, RhD- by homozygous RHD deletion. We used CRISPR/Cas9 genetic engineering to disrupt the RHCE alleles either by a large deletion from exon 1 to exon 2 and/or via insertion-deletion mutations resulting in a frameshift and early stop codon. For lines lacking high prevalence Rh antigens, we reprogrammed group O donor cells whose genotypes were heterozygous for RHD*DAU0/RHCE*ceMO and RHD*DOL/RHCE*ceBI, homozygous for RHD*DIIIa-CE(4-7)-D/RHCE*ceS, and homozygous for RHD/RHCE*ce733G, respectively. At least 2 independent clones from each customized line were characterized for pluripotency by standard flow cytometry and gene expression markers, and a normal karyotype confirmed.

We performed hematopoietic differentiation of customized iPSCs by a 3-step embryoid body protocol in defined, serum-free media with combinations of hematopoietic cytokines. We obtained CD41+235+ primitive hematopoietic progenitors on day 8 and cultured these cells in liquid culture with erythroid specific cytokines to generate iRBCs. We assessed erythroid cell maturation by morphology on cytospin preparations and by flow cytometry. We found an almost pure population of CD71+235+ iRBCs with erythroblast morphology by day 6 of culture, and progressive maturation to Band 3 high, alpha integrin low iRBCs that were smaller and had condensed nuclei that resemble erythroid cultures of primary CD34+ cells. Comparison of total Rh protein expression by flow cytometry on untargeted group O, D+ iRBCs compared to donor-derived RBCs demonstrated comparable cell surface Rh expression. As proof of principal, we assessed RBC agglutination by gel card assay of untargeted group O, D+, E-, e+ and the genetically engineered Rh null iRBCs using standard anti-D, -E, and -e reagents for Rh typing (Ortho Bioclone). As expected, the untargeted O, D+ iRBCs agglutinated with anti-D and anti-e while the engineered Rh-null iRBCs showed no agglutination with all 3 antibodies. Only 0.5 e6 iRBCs were required per gel assay.

We have designed a panel of customized iPSCs reprogrammed from rare donors or genetically engineered to express rare blood group antigen phenotypes or combinations that are difficult or impossible to find as donor red cells. Any number of combinations not found in natural populations can be produced and generated in quantities sufficient for reagents. iRBC produced from these customized iPSCs can be used with standard blood bank assays and potentially provide the means to streamline and standardize antibody identification in alloimmunized patients with complex antibody specificities. In the future, when technology for scale-up is available, Rh null iRBCs could be used as "universal" donor cells for future therapeutic applications.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.