Paul Kaijen

IgG subclasses of anti-FVIII antibodies during immune tolerance induction in patients with hemophilia A

The eradication of inhibitory antibodies in patients with haemophilia A can be accomplished by frequent administration of high or intermediate doses of factor VIII (FVIII), so‐called immune tolerance induction (ITI). This study monitored the distribution of IgG subclasses of anti‐FVIII antibodies during ITI. FVIII‐specific antibodies of subclass IgG1 were detected in all inhibitor patients tested, anti‐FVIII IgG4 in 16, IgG2 in 10 and IgG3 in one of 20 patients analysed. Levels of anti‐FVIII IgG1 and IgG4 correlated well with inhibitor titres as measured by Bethesda assay. In low‐titre inhibitor patients, anti‐FVIII antibodies consisted primarily of subclass IgG1 whereas, anti‐FVIII antibodies of subclass IgG4 were more prominent in patients with high titre inhibitors who needed prolonged treatment or who failed ITI. Longitudinal analysis of 14 patients undergoing ITI revealed that the relative contribution of IgG subclasses was constant for most of the patients analysed. In two patients, the relative contribution of IgG4 increased during ITI. Overall, our findings document the distribution and dynamics of anti‐FVIII IgG subclasses during ITI. Future studies will need to address whether monitoring the relative contribution of anti‐FVIII subclasses IgG1 and IgG4 may be useful for the identification of patients who are at risk of failing ITI.

Multiple B-cell clones producing antibodies directed to the spacer and disintegrin/thrombospondin type-1 repeat 1 (TSP1) of ADAMTS13 in a patient with acquired thrombotic thrombocytopenic purpura

Summary.  Background: The cysteine‐rich/spacer domains of ADAMTS13 contain a major binding site for antibodies in patients with acquired thrombotic thrombocytopenic purpura (TTP). Objective: To study the heterogeneity of the antibody response towards these domains an immunoglobulin V‐gene phage‐display library was constructed to isolate monoclonal anti‐ADAMTS13 antibodies from the immunoglobulin repertoire of a patient with acquired TTP. Methods: Combined variable heavy chain (VH) and variable light chain (VL) segments, expressed as single‐chain Fv fragments (scFv), were selected for binding to an ADAMTS13 fragment consisting of the disintegrin/thrombospondin type‐1 repeat 1 (TSP1)/cysteine‐rich/spacer domains. Results: Seven different scFv antibody clones were identified that were assigned to four groups based on their homology to VH germline gene segments. Epitope‐mapping revealed that scFv I‐9 (VH1‐69), I‐26 (VH1‐02), and I‐41 (VH3‐09) bind to an overlapping binding site in the ADAMTS13 spacer domain, whereas scFv I‐16 (VH3‐07) binds to the disintegrin/TSP1 domains. The affinity of scFv for the disintegrin/TSP1/cysteine‐rich/spacer domain was determined by surface plasmon resonance analysis and the dissociation constants ranged from 3 to 254 nm. The scFv partially inhibited ADAMTS13 activity. However, full‐length IgG prepared from the variable domains of scFv I‐9 inhibited ADAMTS13 activity more profoundly. Plasma of six patients with acquired TTP competed for binding of scFv I‐9 to ADAMTS13. Conclusion: Our data indicate that multiple B‐cell clones producing antibodies directed against the spacer domain are present in the patient analyzed in this study. Our findings also suggest that antibodies with a similar epitope specificity as scFv I‐9 are present in plasma of other patients with acquired TTP.

Residues Arg568 and Phe592 contribute to an antigenic surface for anti-ADAMTS13 antibodies in the spacer domain

Background The majority of patients diagnosed with thrombotic thrombocytopenic purpura have autoantibodies directed towards the spacer domain of ADAMTS13. Design and Methods In this study we explored the epitope specificity and immunoglobulin class and immunoglobulin G subclass distribution of anti-ADAMTS13 antibodies. The epitope specificity of anti-spacer domain antibodies was examined using plasma from 48 patients with acute acquired thrombotic thrombocytopenic purpura by means of immunoprecipitation of ADAMTS13 variants containing single or multiple alanine substitutions. Using similar methods, we also determined the presence of anti-TSP2-8 and CUB1-2 domain antibodies in this cohort of patients. Results Antibody profiling revealed that anti-ADAMTS13 immunoglobulin G1 and immunoglobulin G4 predominate in plasma of patients with acquired thrombotic thrombocytopenic purpura. Analysis of anti-spacer domain antibodies revealed that Arg568 and Phe592, in addition to residues Arg660, Tyr661, and Tyr665, also contribute to an antigenic surface in the spacer domain. The majority of patients (90%) lost reactivity towards the spacer domain following introduction of multiple alanine substitutions at Arg568, Phe592, Arg660, Tyr661 and Tyr665. Anti-TSP2-8 and anti-CUB1-2 domain-directed antibodies were present in, respectively, 17% and 35% of the patients’ samples analyzed. Conclusions Immunoglobulin G directed towards a single antigenic surface comprising residues Arg568, Phe592, Arg660, Tyr661 and Tyr665 predominates in the plasma of patients with acquired thrombotic thrombocytopenic purpura.

Amino acid regions 572-579 and 657-666 of the spacer domain of ADAMTS13 provide a common antigenic core required for binding of antibodies in patients with acquired TTP.

Antibodies directed against ADAMTS13 have been detected in the majority of patients with acquired thrombotic thrombocytopenic purpura (TTP). We have previously localized a major antigenic determinant within the spacer domain of ADAMTS13. To identify the amino acid residues of the spacer domain that are involved in binding of anti-ADAMTS13 antibodies, we constructed a series of fifteen hybrids (designated A-O) in which 5-10 amino acids of the spacer domain were exchanged for the corresponding region of ADAMTS1. Plasma from six patients with antibodies directed against the spacer domain was analyzed for reactivity with the ADAMTS13/ADAMTS1 chimeras. Exchange of amino acid residues 572-579 (hybrid C) and 657-666 (hybrid M) completely abolished the binding of antibodies from all six patients analyzed. Regions 580-587 (D), 602-620 (G, H), 629-638 (J), and 667-767 (N) contributed to binding of antibodies from patients 2, 4, and 5 (epitope present within regions CDGHJMN). Antibodies derived from patient 1 required region 602-620 (G, H) for binding (CGHM-epitope). For antibodies of patient 3, residues 564-571 (B), 580-587 (D), and 629-638 (J) were required (BCDJM-epitope), whereas replacement of residues 602-610 (G) and 629-638 (J) greatly diminished binding of antibodies from patient 6 (CGJM-epitope). Despite the presumably polyclonal origin of the antibodies present in plasma of patients, our results suggest that residues 572-579 (C) and 657-666 (M) comprise a common antigenic core region that is crucial for binding of anti-ADAMTS13 antibodies. Other regions that spatially surround this antigenic core further modulate binding of antibodies to the spacer domain.

Preferential HLA-DRB1*11 dependent presentation of CUB2 derived peptides by ADAMTS13 pulsed dendritic cells

Autoantibodies directed against ADAMTS13 prohibit the processing of von Willebrand factor multimers, initiating a rare and life-threatening disorder called acquired thrombotic thrombocytopenic purpura (TTP). Recently, HLA-DRB1*11 has been identified as a risk factor for the development of acquired TTP. Here, we identified ADAMTS13-derived peptides presented on MHC class II alleles from 17 healthy donors. Dendritic cells from a panel of both HLA-DRB1*11-positive and -negative donors were pulsed with ADAMTS13, and the HLA-DR-presented peptide repertoire was analyzed by mass spectrometry. Interestingly, at low antigen concentrations, HLA-DRB1*11- or DRB1*03-positive donors presented a limited number of CUB2-derived peptides. Pulsing of dendritic cells using higher concentrations of ADAMTS13 resulted in the presentation of larger numbers of ADAMTS13-derived peptides by both HLA-DRB1*11-positive and -negative donors. Although the presented peptides were derived from several ADAMTS13 domains, inspection of the peptide profiles revealed that CUB2 domain-derived peptides were presented with a higher efficiency when compared with other peptides. Remarkably, dendritic cells from DRB1*11 donors pulsed with higher concentrations of ADAMTS13-present derivatives of a single CUB2-derived peptide. We hypothesize that functional presentation of CUB2-derived peptides on HLA-DRB1*11 contributes to the onset of acquired TTP by stimulating low-affinity, self-reactive CD4+ T cells.

Modification of an exposed loop in the C1 domain reduces immune responses to factor VIII in hemophilia A mice

Development of neutralizing Abs to blood coagulation factor VIII (FVIII) provides a major complication in hemophilia care. In this study we explored whether modulation of the uptake of FVIII by APCs can reduce its intrinsic immunogenicity. Endocytosis of FVIII by professional APCs is significantly blocked by mAb KM33, directed toward the C1 domain of FVIII. We created a C1 domain variant (FVIII-R2090A/K2092A/F2093A), which showed only minimal binding to KM33 and retained its activity as measured by chromogenic assay. FVIII-R2090A/K2092A/F2093A displayed a strongly reduced internalization by human monocyte-derived dendritic cells and macrophages, as well as murine BM-derived dendritic cells. We subsequently investigated the ability of this variant to induce an immune response in FVIII-deficient mice. We show that mice treated with FVIII-R2090A/K2092A/F2093A have significantly lower anti-FVIII Ab titers and FVIII-specific CD4(+) T-cell responses compared with mice treated with wild-type FVIII. These data show that alanine substitutions at positions 2090, 2092, and 2093 reduce the immunogenicity of FVIII. According to our findings we hypothesize that FVIII variants displaying a reduced uptake by APCs provide a novel therapeutic approach to reduce inhibitor development in hemophilia A.

Factor VIII-specific memory B cells in patients with hemophilia A

Antibody levels in serum, asmaintained by plasma cells, are the first line of defense against incoming pathogens, whereas memory B cells are indispensable for long-lasting protection [1]. Upon re-exposure to antigen, memory B cells differentiate into antibody-secreting cells (ASCs), thereby enabling rapid eradication of pathogens that are not cleared by pre-existing antibodies. Additionally, there is evidence that memory B cells contribute to replenishing the pool of long-living plasma cells residing in the bone marrow [2]. Persisting levels of peripheral memory B cells have been identified in the context of vaccination [3] and viral infections [4]. Humoral responses against therapeutic or self-proteins are potentially harmful and in this setting development and persistence of antigen-specific memory B cells is undesirable. Thus far, circulating levels of memory B cells in patients with pathogenic antibodies to therapeutic or self-proteins have not been defined. Here, we assessed the frequency of antigen-specific memory B cells in hemophilia A patients [5] who developed inhibitory antibodies to factor (F) VIII. We used the EL4.B5 system [6] to analyze FVIII-specific memory B cells in the circulation of hemophilia A patients. Previously, it has been shown that CD40 L stimulation, provided by EL4.B5 thymoma cells in combination with a T-cell supernatant, results in activation and clonal expansion of the majority of B cells [6]. Subdivision of the CD19 fraction revealed that the IgG producing ASCs developed solely from the IgG B-cell population (data not shown). Frozen peripheral blood mononuclear cells were carefully thawed and stained for the B-cell marker CD19. CD19 cells were sorted onto irradiated EL4.B5 cells and cultured for 9–10 days. Culture supernatants were subsequently analyzed for the presence of FVIII-specific IgG using ELISA. Cells were then washed and incubated on ELISpot plates, coated with 5 lg mL of either FVIII or an anti-IgG monoclonal antibody, in order to visualize antibody-secreting cells. Sixteen previously treated adult hemophilia A patients were included from twoDutchHemophilia Treatment Centers. This study was approved by the institutional review board and informed consent was obtained from all participants. They were divided in three groups: (i) inhibitor patients, (ii) noninhibitor patients that were multi-transfused (>50 exposures) with FVIII without an apparent immune response and (iii) patients successfully treated with immune tolerance induction (ITI) as defined by a negative inhibitor titer. One patient suffering from acquired hemophilia A (A1) and one patient with mild hemophilia A (A2) were included in the group of inhibitor patients. All other patients had severe hemophilia A. In Fig. 1A, results of FVIII ELISpot are depicted for blood samples of five hemophilia A patients with inhibitors. For all five inhibitor patients FVIII-specific ASCs could be detected by ELISpot, although for patient A5 only one well contained spots corresponding to anti-FVIII IgG-producing B cells (Fig. 1A). For all patients the number of IgG-producing ASCs was also determined. These values were used to calculate the percentage of B cells that secreted FVIII-specific IgG. This analysis revealed that for patient A1–A4 0.07 to 0.35% of circulating B cells produces antibodies directed towards FVIII. For patient A5, the percentage of IgG-producing B cells secreting anti-FVIII IgG was 0.01%. To further substantiate our findings, we also determined the frequency of FVIIIspecific memory B cells using an ELISA as a read-out (Fig. 1B). The number of positive wells was used to calculate the frequency of FVIII-specific memory B cells in peripheral blood. A positive well indicates that at least one of the 1000 CD19 B cells in the well produces anti-FVIII IgG. The number of positive wells divided by the total number of sorted B cells allows for the estimation of the frequency of circulating memory B cells. For patient A1–A4, 5–24 anti-FVIII IgGproducing cells per 10 CD19 B cells were detected. It has been established that IgG memory B cells comprise 10–15% of the peripheral B cell compartment [7]. Assuming a similar frequency in our patients, the percentage of FVIII-specific memory B cells in patients A1–A4 ranges from 0.05 to 0.24% of that of total IgG producingmemory B cells. These values are similar to that obtained using ELISpot as a read-out. No Correspondence: Jan Voorberg, Department of Plasma Proteins, Sanquin Research, Plesmanlaan 125, 1066 CX Amsterdam, The Netherlands. Tel.: +31 20 5123120; fax: +31 20 5123680; e-mail: j.voorberg@sanquin.nl

Domain specificity of factor VIII inhibitors during immune tolerance induction in patients with haemophilia A

Summary.  Introduction‐Frequent administration of high dosages factor VIII (FVIII), so‐called immune tolerance induction (ITI), provides an efficient strategy to eradicate inhibitory antibodies in patients with haemophilia A. At present, our knowledge on the characteristics of inhibitory antibodies in patients undergoing ITI is limited. Aim‐In this study we characterized the domain specificity of FVIII inhibitors in 11 haemophilia A patients during ITI. Results‐In three of six patients who were successfully tolerized, inhibitory antibodies were directed predominantly against the FVIII light chain. In two other patients within this group, a significant contribution of A2 antibodies was observed which did not change during treatment. In the sixth patient the relative contribution of A2 inhibitors declined which coincided with an increase in antilight chain antibodies. In four of five patients who failed ITI, A2 inhibitors were observed. In two patients the contribution of A2 inhibitors increased during treatment, while in two other patients the contribution of A2 inhibitor remained constant. The fifth patient had inhibitory antibodies predominantly directed against the FVIII light chain. Conclusion‐Overall, our findings revealed changes in domain specificity of FVIII antibodies in five of 11 patients analysed. Remarkably, antibodies exclusively directed towards the light chain of FVIII were predominantly observed in patients who were successfully tolerized.

Analysis of factor VIII inhibitors in a haemophilia A patient with an Arg593-->Cys mutation using phage display.

Summary. We characterized anti‐factor VIII antibodies in a mild haemophilia A patient with an Arg593→Cys mutation in the A2 domain, using V gene phage‐display technology. All isolated single‐chain variable‐domain antibody fragments were directed against residues Arg484‐Ile508, a binding site for factor VIII inhibitors in the A2 domain. After a further period of replacement therapy, a transient rise in inhibitor titre was observed. These antibodies were directed against the A2 domain. Activation of a pre‐existing pool of B cells, which express antibodies against residues Arg484‐Ile508, could explain the rapid anamnestic response.

Identification of N‐linked glycosylation and putative O‐fucosylation, C‐mannosylation sites in plasma derived ADAMTS13

Acquired deficiency of ADAMTS13 causes a rare and life‐threatening disorder called thrombotic thrombocytopenic purpura (TTP). Several studies have shown that aberrant glycosylation can play an important role in the pathogenesis of autoimmune diseases. N‐linked glycosylation and putative O‐fucosylation sites have been predicted or identified in recombinant ADAMTS13. However, it is not known which of these sites are glycosylated in plasma derived ADAMTS13.