The introduction of human monoclonal anti-D for therapeutic use

Immune regulation by B cells and antibodies a view towards the clinic

B lymphocytes contribute to immunity in multiple ways, including production of antibodies, presentation of antigen to T cells, organogenesis of secondary lymphoid organs, and secretion of cytokines. Recent clinical trials have shown that depleting B cells can be highly beneficial for patients with autoimmune diseases, implicating B cells and antibodies as key drivers of pathology. However, it should be kept in mind that B cell responses and antibodies also have important regulatory roles in limiting autoimmune pathology. Here, we analyze clinical examples illustrating the potential of antibodies as treatment for immune-mediated disorders and discuss the underlying mechanisms. Furthermore, we examine the regulatory functions of activated B cells, their involvement in the termination of some experimental autoimmune diseases, and their use in cell-based therapy for such pathologies. These suppressive functions of B cells and antibodies do not only open new ways for harnessing autoimmune illnesses, but they also should be taken into account when designing new strategies for vaccination against microbes and tumors.

Individuality: the barrier to optimal immunosuppression

Immunosuppressive therapy aims to protect transplanted organs from host responses. Individuals have unique repertoires of responses to foreign antigens and toxic reactions to immunosuppressants; the former determining the type or intensity of rejection reactions and the latter influencing the severity of iatrogenic effects. Because existing agents target molecules that are widely distributed in tissues, new strategies must selectively block lymphoid cells only, disrupt alloresponses but not innate immune responses, interact synergistically with other agents, facilitate the homeostatic process that naturally leads to graft acceptance and ideally only interrupt donor-specific responses. Approaches presently under investigation aim to alter cell trafficking, or selectively deviate the maturation of antigen-presenting cells or inhibit lymphocyte-activation cascades - events that are crucial to rejection responses.

Establishment of human heterohybridoma and lymphoblastoid cell lines specific for the Rh D and C antigens

Human monoclonal antibodies specific for the D antigen of the Rh system are valuable tools for blood group typing and prevention of erythroblastosis. In this study, peripheral blood lymphocytes obtained from an Rh-negative woman immunized with Rh-positive fetuses were immortalized with Epstein-Barr virus (EBV), and transformed lymphoblastoid cell lines (LCLs) secreting antibodies to Rh antigens were generated. The presence of specific antibody was assessed by direct haemagglutination using Rh-positive, papain-treated red blood cells (RBCs), and the production of human antibody was assayed by enzyme-linked immunosorbent assay (ELISA). Specificities of the antibodies were determined by a panel of RBCs of known Rh phenotypes. Five LCLs produced antibody specific for the D antigen, and one LCL showed specificity towards the C antigen of the Rh blood group system. High-titre anti-Rh antibody-producing LCLs were subsequently selected and fused with a human x mouse heteromyeloma cell line. A hybridoma line producing human antibody of the immunoglobulin M (IgM) isotype, which strongly reacted with the D antigen, was established. The hybridoma was cloned, and the monoclone has been stable for growth and antibody production during 8 months of continuous culture, with a mean antibody concentration of 11.5 microg mL-1 and haemagglutination titre of 1/20 480. This antibody was not able to agglutinate a sample of native weak D RBCs (Du); however, agglutination was achieved with papain-treated Du RBCs. Immunoprecipitation of the D antigen by this antibody, followed by Western blot analysis, did not reveal any immobilized D-specific polypeptide. As this human antibody readily agglutinates D+ RBCs in saline, it has the potential to be used as an efficient reagent in routine blood group typing.

Expression and functional analysis of recombinant scFv and diabody fragments with specificity for human RhD

In an attempt to generate recombinant anti-D reagents for possible diagnostic and therapeutic use we cloned the genes encoding the variable (V) domains of a human anti-D antibody secreted by the lymphoblastoid cell line BTSN4. A single-chain Fv (scFv) fragment was constructed using a 21 amino acid linker to join the genes encoding the variable domains of the BTSN4 heavy (VH) and light chains (VL). A diabody construct was also generated by reducing the length of the scFv linker from 21 to 10 residues. The scFv and diabody constructs were cloned into the pFLAG-CTS vector, expressed in E. coli host cells and the recombinant proteins were affinity-isolated from bacterial culture medium. Analysis of the recombinant proteins indicated that they retained the D antigen binding specificity of the parental BTSN4 IgG. Furthermore, both fragments mediated agglutination of papain-treated D positive erythrocytes in the absence of a cross-linking second antibody. While the agglutinating property of BTSN4 diabody was readily explained by the non-covalent association of this protein as a bivalent dimer, oligomeric forms of BTSN4 scFv were not detected when the protein was analysed by size exclusion chromatography. Thus, the agglutinating property of the scFv is not the result of the formation of non-covalently associated multimeric forms of the antibody fragment.

Monoclonal anti-D development programme

Administration of anti-D immunoglobulin to D- women after delivery of a D+ infant has dramatically reduced the number of immunised women and cases of haemolytic disease of the fetus and newborn. The use of monoclonal anti-D might alleviate some of the pressures on maintaining adequate supplies of plasma sourced anti-D. Two human monoclonal antibodies, BRAD-3 (IgG1) and BRAD-5 (IgG3), with proven activity in in vitro functional (immunological) assays with cells bearing IgG Fc receptors (Fc gammaR) were selected for clinical studies. They were prepared by purification of IgG secreted by culture of the Epstein-Barr virus-transformed B cell lines in hollow fibre bioreactors. The mean half-lives of BRAD-3 and BRAD-5 in D- subjects were 10.2 and 22.2 days, respectively. The clearance of D+ red cells injected into D- subjects was accelerated by prior injection of the monoclonal antibodies, both individually and blended (3:1, BRAD-5/BRAD-3). The clearance rate was related to the amount of anti-D on the red cells. Clearance of the D+ red cells coated with BRAD-3 was more rapid in subjects homozygous for Fc gammaRIIIa-F/F158 than in those expressing the Fc gammaRIIIa-V158 allele. The subjects were protected from Rh D immunisation. A large multi-centre study evaluated the BRAD-3/5 blend for its ability to prevent Rh D immunisation in 95 D- subjects given 400 microg i.m. 24 h after injection of 5 ml D+ red cells. Challenge injections of D+ red cells alone were given 24 and 36 weeks later, and blood samples were taken every 4 weeks from the subjects throughout the study for detection of anti-D responses. There was one definite and one possible failure of protection; in one subject the plasma anti-D level rose from week 12 onwards, and in another individual rapid seroconversion was observed at week 28. Considering the relatively large dose of red cells and the number of subjects studied, it was concluded that the failure rate was much lower than in routine Rh D prophylaxis. The responder rate was 13% by week 36 and 24% by week 48. There was no relationship between HLA haplotype and Rh D immunisation. The low percentage of responders and the modest levels of endogenous anti-D produced suggested that administration of monoclonal anti-D had induced long-term specific suppression of anti-D responses in these subjects. The most likely mechanism of action was considered to be inhibition of B cells resulting from co-cross-linking antigen receptors with inhibitory Fc gammaR when the B cells contacted red cells that had bound passive anti-D.

On the mechanism of tolerance to the Rh D antigen mediated by passive anti-D (Rh D prophylaxis)

Anti-D prophylaxis is the most successful clinical application of antibody-mediated immune suppression. Passive IgG anti-D is given to Rh D-negative women to prevent immunisation to foetal Rh D-positive red blood cells (RBC) and subsequent haemolytic disease of the newborn. Despite its widespread use and efficacy, the mechanism of action of this therapy is unproven. The known facts about the antigen, antibody response, dose of anti-D, RBC clearance and effects of the passive anti-D on subsequent primary and secondary immune responses are discussed in relation to recent information on ways by which immune responses may be suppressed. Most Rh D antigen sites on RBC are not bound by passive anti-D, and thus epitope masking (which may occur in experimental murine models using xenogeneic RBC) is not the reason why anti-D responses are prevented by administration of prophylactic anti-D. It is hypothesised that although clearance and destruction of the antigenic RBC may be a contributing factor in preventing immunisation, down-regulation of antigen-specific B cells through co-ligation of B cell receptors and inhibitory IgG Fc receptors must also occur.

Functional in vivo characterization of human monoclonal anti-D in NOD-scid mice

The prophylaxis of the hemolytic disease of the newborn requires significant amounts of plasma-derived polyclonal human anti-D. Because of procurement problems, there is a growing interest in replacing plasma-derived anti-D by in vitro-produced human monoclonal anti-D. Hundreds of monoclonal anti-D have been prepared, but the selection of the most potent for in vivo use is difficult because it cannot be predicted by in vitro characterization. This study evaluated the possibility of using nonobese diabetic/severe combined immunodeficient (NOD-scid) mice for the in vivo evaluation of human monoclonal anti-D. Human red blood cells (RBCs) were found to circulate normally in the blood of NOD-scid mice previously injected with a physiologic amount of human immunoglobulin G (10 mg). The addition of a small amount of anti-D (1 to 5 microg) resulted in the clearance of Rh D(+) RBCs within 4 hours. The comparative testing of 8 monoclonal anti-Ds showed a wide range of potency (15% to 87%) relative to plasma-derived polyclonal anti-D. There was no strong correlation between the in vivo potency index and the immunoglobulin G isotype, affinity, or fine specificity of the antibodies. These results show the usefulness of NOD-scid mice for the initial in vivo screening of human monoclonal anti-D before testing the most active antibodies in clinical trials done in human volunteers.

The role of immunoglobulins in neonatal Rhesus haemolytic disease

Rhesus (Rh) isoimmunisation is the most common form of severe haemolytic disease of the newborn (HDN). The introduction of prophylaxis with anti-D Rh0 immunoglobulin (anti-D) has resulted in a marked reduction in the sensitisation of Rh-negative women and deaths attributable to Rh HDN. The sensitisation rate could be further decreased if there was strict adherence to the guidelines for administration of anti-D prophylaxis. Whether additional prophylaxis at 28 and 34 weeks of gestation would be cost effective is controversial. Intrauterine transfusions to treat fetal anaemia, postnatal exchange transfusions and phototherapy are all part of the standard management of affected individuals. Intravenous immunoglobulin given to pregnant women can reduce fetal haemolysis, and when administered to neonates with Rh isoimmunisation has been associated with a reduction in the requirement for exchange transfusion. There are, however, potential risks of immunoglobulin administration, including haemolysis due to the presence of anti-A or anti-B antibodies, allergy and the transmission of disease.

Mechanism of anti-D-mediated immune suppression--a paradox awaiting resolution?

During pregnancy, women can be immunized by fetal red blood cells (RBCs) of an incompatible blood group. Subsequent transplacental passage of the antibodies can result in fetal morbidity or mortality due to RBC destruction. The administration of anti-D antibodies to D(-) women after delivery of a D(+) infant, and subsequent prevention of Rhesus (Rh) D haemolytic disease of the fetus and newborn, is the most successful clinical use of antibody-mediated immune suppression. The passive IgG anti-D might prevent immunization to D(+) RBCs by an IgG Fcgamma receptor (Fcgamma R)-dependent mechanism such as crosslinking the D-specific B-cell receptor and inhibitory FcgammaRIIb. However, recent murine studies demonstrate that the suppressive effects of antibodies to heterologous RBCs can be Fcgamma R-independent, suggesting other mechanisms might contribute.

Evaluation of a panel of human monoclonal antibodies to D and exploration of the synergistic effects of blending IgG1 and IgG3 antibodies on their in vitro biologic function

BACKGROUND

The D immunoprophylaxis program has successfully reduced the incidence of Rh hemolytic disease of the newborn (HDN), but it has also reduced the availability of plasma-derived polyclonal anti-D, which constitutes the current therapeutic product. Human monoclonal anti-D from hybridoma cell lines may be an acceptable alternative, and clinical efficacy of each anti-D is being evaluated in several centers.

STUDY DESIGN AND METHODS

This study represents the largest assessment (outside of the International Workshops) of human D monoclonal antibodies for potential therapeutic use. The in vitro biologic activity and immunologic and serologic reactivity of a coded panel of 20 D antibodies (THERAD) was investigated. The bioassays used were lymphocyte (K-cell) antibody-dependent cell-mediated cytotoxicity (ADCC), monocyte ADCC, and monocyte chemiluminescence, which together reflect the processes involved in antibody-coated red cell destruction in vivo. From this panel, six antibodies (THERADs 14, 19, 22, 23, 27, and 28, comprising 3 IgG1 and 3 IgG3 D monoclonal antibodies) were further selected to investigate the effects of blending in the three bioassays.

RESULTS

Several THERAD blends displayed greater activity than their component parts, in the range of 6 to 124 percent. There was no evidence to suggest functional blocking effects with this restricted panel of antibodies.

CONCLUSION

The THERAD blends containing both IgG1 and IgG3 anti-D appeared to be the most functionally active, as did blends containing antibodies to two distinct D epitopes. This in vitro evidence has important implications for the future formulation of an effective monoclonal preparation for the prevention of Rh HDN.