Marginal Zone B Cells Mediate Alloantibody Formation to a Clinically Significant Human RBC Antigen in a Murine Model

Alloantibodies generated against red blood cell (RBC) antigens can lead to severe and potentially fatal outcomes such as hemolytic transfusion reactions (HTRs) or hemolytic disease of the fetus and newborn (HDFN). Several mouse models have been generated in efforts to study the alloimmune response to RBC antigens; however, few express clinically significant human RBC antigens. As a result, we developed a novel transgenic mouse model with RBC specific expression of the human KEL antigen, one of the most common non-ABO(H) antigens implicated in HTRs and HDFN. Using this model, we examined the rate, magnitude and consequence of alloimmunization following transfusion of RBCs expressing the human KEL2 antigen in moderate copy number (approximately 1200/RBC). In addition, we evaluated the potential involvement of marginal zone B (MZ B) cells, a B cell population involved in antibody-mediated immunity toward blood borne antigens, in the RBC alloimmunization process.

C57BL/6 or KEL2 transgenic control recipient mice were transfused with KEL2 and C57BL/6 RBCs labeled with distinct lipophilic dyes to facilitate detection of cells following transfusion. To determine the role of MZ B cells, an additional group of C57BL/6 mice was treated with a-LFA-1 and a- CD49d blocking antibodies, four days prior to transfusion, in order to selectively deplete B cells from the marginal zone; additional animals were treated with an isotype matched control antibody or were surgically splenectomized. Post-transfusion survival and RBC bound antibody levels were determined by flow cytometry at specific timepoints after transfusion utilizing the fluorescent lypophilic dyes, anti-IgM, and anti-IgG. Serum antibodies were also analyzed by indirect immunofluorescence using flow cytometry with KEL2 and control C57BL/6 RBCs as targets. All experiments were completed at least three times with 3–5 recipients per group per experiment.

In C57BL/6 recipient mice, anti-KEL glycoprotein IgM was detectable within 2 days after KEL2 RBC transfusion, and peaked at day 5 post transfusion (compilation data of 3 experiments: adjusted MFI of IgM bound to KEL2 RBCs was 24.5 +/− 8.2 (mean +/− SD) on D5). Anti-KEL glycoprotein IgG was detectable on the transfused KEL2 RBCS and in the serum by day 5 and peaked between 14 and 21 days post transfusion (compilation data of 3 experiments: adjusted MFI of IgG bound to KEL2 RBCs was 19.7 +/− 16.7 on D14). Clearance of KEL2 RBCs began as early as 2 days following transfusion and continued to largely parallel the formation of anti-KEL alloantibodies. In contrast, KEL2 recipients failed to make anti-KEL alloantibodies (adjusted MFI of IgM bound to KEL2 RBCs was 2.1 +/− 2.7 on D5; adjusted MFI of IgG bound to KEL2 RBCs was 1.2 +/− 0.6 on D14, p<0.001 (IgM) and <0.05 (IgG) compared to KEL2 RBCs into C57BL/6 recipients). Furthermore, KEL2 recipients failed to clear transfused cells, strongly suggesting that KEL2 RBC clearance in C57BL/6 recipient mice was an antibody-mediated process. Splenectomy abrogated the alloantibody response to KEL2 RBCs, and selective depletion of MZ B cells greatly reduced alloimmunization (adjusted MFI of IgM bound to KEL2 RBCs was 9.3 +/− 6.1 on D5, adjusted MFI of IgG bound to KEL2 RBCs was 1.7 +/− 3.2 on D14, p<0.001 (IgM) and p<0.01 (IgG) compared to KEL2 RBCs into C57BL/6 recipients); furthermore, MZ B cell depletion also prevented KEL2 RBC clearance.

Transfusion of RBCs expressing the KEL2 antigen elicits a robust alloimmune response in recipient mice lacking expression of the human KEL glycoprotein. Given the complete lack of the human KEL glycoprotein antigen on recipient RBCs, the described system more closely models Rh(D) than KEL1/KEL2 (which involves a single amino acid polymorphism). The anti-KEL alloantibody response observed during primary exposure in the described model system leads to clearance of the transfused KEL2 RBCs, and suggests that initial alloantibody formation may reduce therapeutic efficacy when antigenic differences exist on transfused RBCs. In addition, these results suggest a key role for splenocytes in RBC alloimmunization, specifically MZ B-cells. Ongoing studies are further examining alloimmune responses and clearance patterns within this model system, with a focus on manipulation of MZ B cells as a potential method to prevent RBC alloimmunization.

No relevant conflicts of interest to declare.