Yin and yang in transfusion medicine

Comment on Zimring et al, page 1105

A new mouse model of transfusion of incompatible RBCs sheds light on the mysterious process of antibody-induced antigen suppression.

Antibody binding to red blood cell (RBC) antigens classically results in hemolysis. The converse is the phenomenon of “antigen suppression,” which has been described in only a few case reports. These cases typically occurred in patients with autoimmune hemolytic anemia (AIHA) in which the autoantibody led to suppressed expression of its cognate antigen. Suppression was transient with antigen re-expression when the antibody disappeared. In addition, 2 antigens were suppressed sequentially in 1 patient by the appearance and disappearance of autoantibodies with different specificities.¹  Antigen suppression occurred most often with antibodies recognizing Kell blood group system antigens, but other specificities were also described. These antibodies can hemolyze autologous and transfused RBCs,²  and antigen suppression protects RBCs from hemolysis. Thus, unraveling the mechanism of antigen suppression may lead to new ways of treating AIHA and preventing transfusion reactions. In this issue of Blood, Zimring and colleagues present a novel mouse model that may provide this understanding.

Various mechanisms have been proposed to explain antigen suppression, but it is still not understood. For example, since these antibodies may bind antigens on RBC progenitors, they may prevent antigen synthesis; however, thisdoes not explain antigen suppression of transfused RBCs.²  Antibody binding to the RBC antigen may also interfere with in vitro serological tests to detect that antigen; however, immunoblots of antigen-suppressed RBCs demonstrated the absence of the relevant antigen.¹  Finally, the RBC antigen may be destroyed, with circulating microbial enzymes the likely culprits; however, infections in these patients were rarely described. Similarly, Williamson and colleagues postulated the “shedding” of immune complexes containing the relevant RBC antigen and its bound autoantibody.¹ 

Although mice have at least 10 RBC alloantigen systems (ie, Ea1-Ea10 [erythrocyte alloantigens 1-10]), and several mouse strains express transgenic RBC antigens,³  these have not yet been exploited to study incompatible RBC transfusions. To this end, Zimring et al transfused RBCs from transgenic mice expressing a transmembrane form of hen egg lysozyme (HEL) into recipients with high titers of immunoglobulin G (IgG) anti-HEL. Surprisingly, the transfused RBCs were not hemolyzed; rather, HEL was specifically removed from the transfused RBCs, which then survived normally. This Fc receptor–mediated process did not require the spleen and probably occurred in the liver. Given that extravascular hemolysis, and not antigen suppression, is found in both a mouse model of AIHA⁴  and following transfusion of glycophorin A transgenic mouse RBCs into alloimmunized recipients (David A. Schirmer, Shuh-Chung Song, and S.L.S., unpublished observations, May 2005), it will be interesting to compare and contrast these models in future studies.

Finally, the model of Zimring et al is reminiscent of the “transfer reaction” involving immune complexes bound to complement receptor 1 (ie, CR1) on primate RBCs, in which macrophages ingest both the immune complex and CR1 without producing hemolysis.⁵  In particular, the Fc region of the relevant antibody and intact Fc receptor function are required for the transfer reaction.^5,6  In addition, the liver primarily removes the CR1-bound immune complexes in vivo.⁶  Although mouse RBCs do not express CR1, this process may not be limited to CR1,⁷  thereby suggesting additional ways of studying antigen suppression in vivo using the mouse model of Zimring et al and in vitro using antibody and RBC samples from human patients. ▪