Development of a Murine Model of Weak Kel: Similarities to Weak Rh(D)

Aspects of RBC antigens that define their immunogenicity are not fully understood. Multiple studies in humans have demonstrated Rh(D) to be the most immunogenic RBC antigen, yet the same antigen present in lower copy number (weak Rh(D)) is, with rare exception, considered non-immunogenic. A better understanding of factors influencing RBC alloimmunogenicity would be helpful in predicting antibody responses in transfusion and pregnancy situations alike, as well as in managing donor RBC inventory. Herein, we describe the generation of transgenic mice with different levels of RBC specific expression of the clinically significant human KEL2 antigen, and test the hypothesis that antigen density impacts recipient immune response post-transfusion.

Mice with RBC specific expression of KEL2 were generated utilizing constructs containing the human KEL2 sequence expressed behind a B-globin promoter, using a random integration approach. TER119+, CD45+, and CD41+ cells were evaluated by flow cytometry for KEL expression using monoclonal anti-Js^a and anti-Kp^b, and RBC antigen density was estimated utilizing QIFIKIT beads. MuMT recipients were transfused with RBCs labeled with a lipophilic dye, and post-transfusion RBC recovery and antigen expression were evaluated by flow cytometry. To determine the immunogenicity of KEL2 or weak KEL2 RBCs, RBCs were transfused into C57BL/6 recipients every 2–3 weeks in the presence or absence of poly (I:C) pre-treatment. Recipient serum was analyzed by flow cytometric crossmatch with KEL2 or C57BL/6 RBC targets, using IgM or IgG secondary antibodies. To determine the effect of recipient RBC expression of weak KEL2 on the immunogenicity of KEL2 RBCs, weak KEL2 animals were transfused with KEL2 RBCs, the clearance of lipophilic labeled RBCs was tracked, and anti-KEL was evaluated on the transfused RBCs and also in the serum.

KEL2 RBCs have approximately 1200 antigenic sites per cell, whereas weak KEL2 RBCs have fewer than 200 sites; flow cytometric studies of TER 119+, CD45+, and CD41+ cells suggest both strains have RBC specific KEL expression. Transfusion of KEL2 or weak KEL2 RBCs into muMT animals resulted in stable post-transfusion RBC recovery and antigen expression. In 3/3 experiments (n=30 animals), all C57BL/6 recipients of KEL2 RBCs generated detectable anti-KEL IgM and IgG, which boosted with subsequent transfusions and which was enhanced in the presence of recipient inflammation with poly (I:C). However, in 2/2 experiments (n=20 animals), weak KEL RBCs led to no detectable antibody (IgM or IgG) in C57BL/6 recipients following 3 transfusions, even in the presence of recipient pre-treatment with poly (I:C). Furthermore, weak KEL2 recipients of KEL2 RBCs generated no detectable IgG and demonstrated no clearance of KEL2 RBCs, though low levels of anti-KEL IgM were detectable on the transfused RBCs and in the serum from approximately 5–12 days post-transfusion.

As hypothesized, antigen density significantly impacts the immunogenicity of KEL RBCs in this reductionist murine alloimmunization model. C57BL/6 recipients, like Rh(D) negative recipients, lack the human antigen in question (KEL2 in this case), and recipient antibody responses to weak KEL2 are, like most recipient responses to weak Rh(D), undetectable. Strengths of the KEL2 and weak KEL2 system include the fact that these animals are, to the best of our knowledge, genetically identical except for RBC KEL antigen copy number. Thus, this system allows for detailed analyses of the immune response to KEL RBCs with different antigen densities, on both the donor and recipient side of the equation, without confounding factors encountered in human studies (such as HLA presentation issues or considerations of the molecular basis of a particular type of weak Rh(D)). A better understanding of primary, secondary, and other immune responses in the KEL2 and weak KEL2 system may lay the groundwork for strategies to induce non-responsiveness to RBCs in humans, not only in the setting of transfusion medicine but also potentially in the setting of hemolytic disease of the fetus and newborn. Furthermore, translation of these and future findings in the KEL and weak KEL systems to Rh(D), weak Rh(D), and other human antigen systems may ultimately allow for creative solutions in blood inventory management.

No relevant conflicts of interest to declare.