Non-Hemolytic Antigen-Loss during Transfusion of Crossmatch Incompatible Blood Requires Cooperative Binding of Multiple Antibodies Recognizing Different Epitopes.

Background: Hemolysis is not the inevitable result of transfusing crossmatch incompatible blood, even for clinically significant antigens. Less-well appreciated is the phenomenon of “antigen-loss” where the offending antigen is stripped off of RBCs, leaving the RBCs intact with a normal circulatory lifespan. Recently, we described a novel murine model of antigen-loss during transfusion of crossmatch incompatible RBC. mHEL mice express the model antigen Hen Egg Lysozyme (HEL) as a RBC cell surface protein. Transfusion of mHEL RBC into recipients immunized against HEL results in only minimal hemolysis, with the majority of RBC undergoing non-hemolytic loss of the mHEL antigen. In the current report, we isolated a panel of novel monoclonal anti-HEL antibodies to investigate the properties of antibodies that can cause antigen-loss.

Methods: BALB/c mice were hyper-immunized with HEL, resulting in a polyclonal IgG anti-HEL response. Splenic fusions with myeloma cells were performed and monoclonal antibodies were isolated. To test the ability of the monoclonal antibodies to induce antigen-loss, recipient mice were infused with the antibodies and then transfused with mHEL RBC labeled with a fluorescent dye. RBC survival, antibody binding to the transfused RBC and mHEL surface antigen levels were monitored by flow cytometry.

Results: A panel of 7 monoclonal antibodies that showed strong binding to mHEL RBC were tested. No antigen-loss was observed when monoclonal antibodies were infused alone, despite strong binding of the antibodies to transfused RBC. This was not due to slowed kinetics, as no antigen-loss was observed even after 21 days. In contrast, infusing a mixture of the 7 monoclonal antibodies caused complete antigen-loss within 4 days, similar to polyclonal antiserum. Each combination of 2 antibodies was tested (21 conditions). (14/21) combinations caused greater than 99% antigen-loss by 4 days, whereas no antigen-loss was observed in the other 7 combinations. To determine the relative epitope specificity of the monoclonal antibodies, binding competition studies were carried out. Of the 7 antibodies, 4 completely blocked each other (Group 1 – 5B9.A1, 4B7.B1, 8E12.D10, and GD7). In contrast, 2 antibodies (Group 2 – 2F4 and 6F7.D8) did not block any antibodies from group 1. Preincubation with 2F4 blocked binding of 6F7.D8 but 6F7.D8 did not block 2F4 (defined as partial identity). The seventh antibody, 6F7.D8, had partial identity to group 1, but did not block antibodies from group 2. A comparison of the groups revealed that antigen-loss occurred in all cases where two antibodies bound to different epitopes (including partial identity), but not when antibodies bound the same epitope.

Discussion: The current findings demonstrate that antigen-loss of mHEL only occurs when at least two antibodies, which recognize different epitopes, simultaneously bind the mHEL antigen. These data suggest mechanisms of lattice formation, involving crosslinking of multiple antigens by different antibodies. Ongoing mechanistic studies are assessing this hypothesis. A practical implication of these findings is that the likelihood of antigen-loss in a given patient may correlate to the multiplicity of epitopes of a blood group antigen recognized by a given patient’s antibody response.