Antibody-mediated immune suppression is improved when blends of anti-RBC monoclonal antibodies are used in mice

The prevention of hemolytic disease of the fetus and newborn (HDFN) by polyclonal anti-D is a successful clinical application of the suppressive capability of anti-red blood cell (RBC) antibodies. Although anti-D has been highly successful in preventing HDFN, the development of a recombinant replacement for anti-D is highly desirable. A recombinant alternative to plasma-derived anti-D would ensure a renewable and well-defined source of anti-RhD antibodies without the potential risk of pathogen transmission.

Monoclonal antibodies could provide an alternative to polyclonal immunoglobulin G (IgG) anti-D. Several monoclonal and recombinant anti-D antibodies have been tested in humans for their ability to remove RBCs from the circulation and to prevent D immunization.¹  Unfortunately, these monoclonal anti-D antibodies have not been sufficiently effective for further clinical development¹  and some monoclonal antibodies even resulted in enhanced immune responses to D-positive RBCs instead of preventing immunity.² 

More recently, a new class of monoclonal antibodies (Rozrolimupab) comprising 25 recombinant human IgG1 antibodies recognizing multiple epitopes on the RhD antigen has been developed.³  Rozrolimupab was effective in the treatment of immune thrombocytopenia but to the best of our knowledge has not been used in the prevention of HDFN.⁴  We evaluate here the efficacy of single-monoclonal anti-RBC antibodies vs blends of antibodies in the induction of antibody-mediated immune suppression (AMIS) using a fully allogeneic mouse model of RBC immunization and demonstrate maximal suppression only when the monoclonal antibodies bound to different epitopes.

C57BL/6 mice were purchased from Charles River Laboratories (Kingston, NY). HOD mice on the FVB background consisted of transgenic animals expressing an RBC-specific fusion protein composed of hen egg lysozyme (HEL), ovalbumin, and human Duffy^b (donated by James C. Zimring, Bloodworks NW Research Institute, Seattle, WA).⁵  FcRγ chain–deficient (FcRγ^−/−; Model 583) mice were purchased from Taconic Farms, Inc (Germantown, NY). All animal studies were approved by the St. Michael’s Hospital animal care committee.

Polyclonal anti-HEL was produced and purified in our laboratory.⁶  Monoclonal HEL-specific antibodies 4B7, 2F4, 8E12, and 5B9 (all mouse IgG1 subtype) were made available to us by James C. Zimring.^7,8  The antibodies 4B7, 8E12, and 5B9 block each other’s binding to the HEL protein, whereas 2F4 and 4B7 bind to different epitopes.⁷  The ability of antibodies to bind HOD erythrocytes was assessed by flow cytometry.⁹ 

HOD-RBCs were washed 3 times with phosphate-buffered saline, and the concentration was adjusted to 10⁸ cells per milliliter. Mice were transfused (tail vein) with 10⁷ HOD-RBC, or 10⁷ HOD-RBC presensitized with the indicated concentrations of polyclonal anti-HEL, individual monoclonal antibodies, monoclonal mixtures, or phosphate-buffered saline without washing away the unbound antibody. Mice were bled for serum on days 0, 7, 14, and 21. AMIS was evaluated by the development of HEL-specific antibodies after transfusion. IgM antibodies specific for HEL were detected by enzyme-linked immunosorbent assay (ELISA).¹⁰ 

Data were expressed as the mean ± standard error of the mean (SEM) and analyzed by the Kruskal-Wallis nonparametric test with Dunn’s multiple comparison test.

The clinical success of polyclonal anti-D over monoclonal antibodies in the prevention of HDFN has suggested that polyclonal preparations may have an advantage as a therapeutic.¹  In this work, we evaluated the efficacy of single-monoclonal RBC antibodies, 2 antibody blends, and a polyclonal antibody preparation at suppressing the RBC-specific immune response (eg, AMIS) using the HOD mouse model of RBC immunization.

AMIS was first evaluated with 4 distinct monoclonal antibodies of the same subtype (clones 4B7, 8E12, 2F4, and 5B9) and a polyclonal IgG specific for the HEL portion of the HOD-RBC. All the HEL monoclonal antibodies significantly but partially suppressed the antibody response to the HEL protein (Figure 1; supplemental Figure 1, available on the Blood Web site). However, although essentially complete suppression of the IgM response was induced by saturating doses of polyclonal anti-HEL (1-5 µg), incomplete suppression was observed with saturating concentrations (0.1-5 µg) of any of the monoclonal antibodies. This result supports the contention that polyclonal formulations are superior to monoclonal antibodies at preventing RBC immunization.

We evaluated next if the blend of the monoclonal antibodies, 4B7 and 2F4, that bind to different epitopes on the HEL protein^7,8  can increase the AMIS effect compared with the individual monoclonal antibodies. A significantly better AMIS effect on the HEL-specific IgM response was observed by the antibody blend (2F4+4B7) compared with the individual antibodies (Figure 2B). The antibody blend reacted with the HOD-RBCs with a slightly higher fluorescence intensity of binding (Figure 2A). In sharp contrast, the blend of 4B7, 8E12, and 5B9 that blocks each other’s binding to the HEL protein did not increase the AMIS effect (Figure 2D) nor the total reactivity of the antibody blend with the HOD-RBCs (Figure 2C). It is important to note that interference of IgG anti-HEL on the detection of HEL-specific IgM antibody was not observed (supplemental Figure 2). These data thus demonstrate that the blend of anti-RBC antibodies binding to different epitopes is significantly more effective than any of the 4 single formulations at preventing anti-RBC immunization.

The binding of multiple antibodies to the RBC can induce some degree of antigen loss, where the RBCs lose the incompatible antigen and then circulate normally.^7,8  Antigen loss appears to require the function of activating FcγRs.⁸  To indirectly evaluate the possible role of FcγR-dependent antigen loss mechanisms on the increased AMIS effect by nonblocking monoclonal blends, we studied the suppressive capability of single vs mixtures of antibodies that bind to different epitopes in mice that do not express functional activating FcγR. The antibody mixture still significantly suppresses the IgM antibody response compared with the single-antibody formulations in the absence of FcγR function (Figure 2E). Although these results do not support FcγR-dependent antigen loss as the primary mechanism involved in the suppressive capability of these monoclonal antibody blends, antigen loss cannot be excluded as a possible mechanism until directly tested.

The literature also favors the epitope masking model for the efficacy of the monoclonal blends, which postulates that IgG could mask epitopes and prevent B cells from recognizing the antigen.^11-13  However, our previous experiments indicate that AMIS can occur through steric hindrance-independent mechanisms. Antibodies to the Duffy portion of the HOD molecule suppress the antibody response to the HEL antigen without interference in the binding between the anti-Duffy and anti-HEL antibodies,⁶  and also in the absence of activating FcγR function (L.B., unpublished data, May-November 2013). It could be possible that AMIS may be the result of multiple mechanisms, which can replace each other functionally depending on the immune settings.

Some of the implications of our findings are that maximized AMIS effects can be observed without the need for contaminants in the polyclonal preparation or the requirement of multiple IgG subtypes, which are known to have differential effector functions.¹⁴  Anti-D IgG contains higher quantities of IgG1 followed by IgG3 and less IgG2 and IgG4.¹⁵  However, the present study shows that maximal AMIS effects can be observed without multiple IgG subtypes as the monoclonal antibody blend used here contained only mouse IgG1 subtype antibodies.

In summary, our results bring forward a concept that monoclonal antibody blends can decrease RBC immunization when the antibodies bind to nonblocking antigen epitopes. Because new technologies have been developed to produce novel combinations of antibody mixtures, it may be possible to now rationally develop highly successful antibody mixtures to prevent RBC immunization in the context of HDFN.

The authors thank James Zimring (Bloodworks NW Research Institute, Seattle, WA) for kindly providing the HOD mice as well as access to the HEL-specific monoclonal antibodies. They also thank Andrew Crow, Joan Legarda, Melissa Menard, Xiaojie Yu, Wang Lin, and the St. Michael’s Hospital Research Vivarium staff.

This work was supported by a grant from Health Canada as part of the Canadian Blood Services/Canadian Institutes of Health Research partnership fund (grant CBS221511) (A.H.L.). L.B. was supported by postdoctoral scholarships from the Canadian Blood Services.

The views expressed herein do not necessarily represent the view of the federal government of Canada.

Contribution: L.B. designed the research, performed the experiments, analyzed the data, and wrote the manuscript; A.A. performed experiments; D.M. performed experiments and analyzed data; A.H.L. designed research, analyzed data, obtained grant funding, and wrote the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Alan H. Lazarus, The Keenan Research Centre, St. Michael's Hospital, 30 Bond St, Toronto, ON M5B 1W8, Canada; e-mail: lazarusa@smh.ca.