Detection of Rh Antibodies in Patient Plasma Using Genome Engineered Induced Pluripotent Stem Cell-Derived Red Cells

Background: Despite transfusion of Rh matched red cells for patients with sickle cell disease, Rh alloimmunization remains a persistent challenge. Rh specificities can be complex, resulting from RH genetic diversity found in patients and donors. Antibody identification is hampered by the lack of appropriate reagent red cells, especially those that can identify antibodies against high prevalence or low prevalence Rh antigens. We used human induced pluripotent stem cells (iPSCs) with the goal of producing renewable red cell reagents to both screen for Rh alloimmunization and to aid complex antibody identification.

Methods: We generated a panel of iPSCs that include Rh null, D--, lack the high prevalence antigens hr ^S or hr ^B, or express uncommon Rh antigens such as V, VS, Go ^a, or DAK. For the Rh null line, we used CRISPR/Cas9 genetic engineering to disrupt RHCE via a large deletion in a D- iPSC. For D--, RHD was inserted into the AAVS1 safe harbor locus of an Rh null iPSC line using zinc finger nucleases resulting in a line that constitutively expresses RhD but no RhCE. iPSCs with uncommon variants were reprogrammed from RH genotyped donors or engineered similar to the generation of the D-- line. Hematopoietic differentiation by embryoid body formation was used to generate hematopoietic progenitors that were subsequently cultured towards the erythroid lineage. Mature iRBCs were ficin treated and tested with patient plasma with previously identified Rh antibodies using gel agglutination assays.

Results: Rh null iPSC-derived RBCs (iRBCs) showed complete absence of cell surface Rh protein by flow cytometry, while D-- iRBCs showed Rh protein expression levels comparable to D-ce+ iRBCs using an anti-D/CE antibody. We assessed RBC agglutination of Rh null, D--, hr ^S-, hr ^B-, VVS+, Go ^a+, and DAK+ iRBCs using standard Rh typing reagents (Ortho). The reprogrammed uncommon donor iRBCs agglutinated with monoclonal anti-Rh antibodies as predicted by RH genotype, while the Rh null iRBCs showed no agglutination with all 5 common Rh antibodies and D-- iRBCs showed agglutination with anti-D reagents only. Rh null iRBCs showed no agglutination against patient plasma containing anti-D, while D-- iRBCs agglutinated. While D- RHCE*ce homozygous iRBCs showed strong agglutination against patient plasma containing anti-hr ^S, Rh null, D--, and hr ^S- iRBCs did not agglutinate. No iRBCs showed agglutination by plasma containing anti-V/VS while VVS+ iRBCs showed strong agglutination. Similarly, no iRBCs showed agglutination by plasma containing anti-Go ^a while Go ^a+ iRBCs showed strong agglutination. Detection of most antibodies against Rhce on iRBCs was enhanced by ficin treatment whereas antibodies with D specificity did not require ficin treated cells for detection.

Conclusion: We suggest that genetically engineered iPSCs expressing uncommon Rh antigen phenotypes that are difficult or impossible to obtain from red cell donors can expedite antibody identification. Rh null and D-- iRBCs could be useful to discriminate antibodies against RhD versus RhCE. Customized iPSCs that lack high prevalence or express low prevalence Rh antigens could potentially standardize antibody evaluation in patients with complex Rh specificities.