Critical Amino Acids in a Conserved Region of the Spacer Domain Contributing to the Binding of Anti-ADAMTS13 Antibodies

Abstract

Thrombotic thrombocytopenic purpura (TTP) is a severe disorder characterized by the absence or dysfunction of the von Willebrand factor cleaving protease ADAMTS13. In plasma of the majority of patients with TTP autoantibodies directed towards the metalloprotease ADAMTS13 are present. We have previously shown that the spacer domain of ADAMTS13 contains a major binding site for antibodies that develop in patients with TTP. More detailed analysis revealed that residues Tyr657-Tyr666 within the spacer domain comprise part of a common antigenic core that is critical for binding of anti-ADAMTS13 antibodies. Here, we determined individual amino acids within region Tyr657-Tyr666 are involved in binding of a panel of human monoclonal and polyclonal anti-ADAMTS13 antibodies. A series of human monoclonal antibodies reactive with the spacer domain has been isolated from the immunoglobulin repertoire of patients with acquired TTP using phage display. Two human monoclonal antibodies, designated I-9 and II-1 that were both directed against the spacer domain were included in this analysis. Individual amino acids within region Tyr657-Tyr666 were introduced and expressed in the context of an ADAMTS13 variant truncated after the spacer domain. Reactivity of human monoclonal antibodies I-9 and II-1 with the spacer domain variants was determined by immunoprecipation using conditioned medium of stably transfected High Five cells expressing the different variants. Replacement of Arg660, Tyr661 or Tyr665 resulted in a loss of reactivity of the corresponding spacer domain variants with antibody I-9 and II-1. Reactivity of variants in which Gly662 or Glu664 were replaced by an alanine resulted in a reduced binding to antibody I-9 whereas binding to antibody II-1 was not affected by these amino acid replacements. We subsequently addressed the reactivity of polyclonal IgG present in plasma of six patients with acquired TTP with the spacer domain variants described above. First, we showed that polyclonal IgG derived from these patients did react exclusively with the spacer domain and not with other domains on ADAMTS13. Polyclonal antibodies derived from plasma of patients with acquired TTP were subsequently evaluated for binding to the spacer domain variants. For only one of six patients, binding of polyclonal IgG to ADAMTS13 was lost when either Arg660, Tyr661 or Tyr665 was replaced by an alanine. For the other patient plasma's a reduced reactivity with some of the spacer domain variants was observed; however binding was never completely abolished. These findings suggest that multiple residues within Tyr658-Tyr665 are involved in the binding of polyclonal anti-ADAMTS13 antibodies. To further explore this possibility we generated variants in which double and triple combinations of residues Arg660, Tyr661 or Tyr665 were replaced by an alanine. The resulting variants designated RY1 (both Arg660 and Tyr661 replaced for Ala), RY2 (both Arg660 and Tyr665 replaced for an Ala), YY (both Tyr661 and Tyr665 replaced for an Ala) and RYY (both Arg660, Tyr661 and Tyr665 replaced for an Ala) were evaluated for binding to polyclonal IgG present in plasma. Plasma of two patients showed a lack of binding to variants RY1 and RY2. Three patients showed a lack of reactivity with the YY and the RYY variants. A reduced binding of the three other patients to the RYY variant was observed. These data suggest that Arg660, Tyr661 and Tyr665 contribute to an antigenic surface present on the spacer domain of ADAMTS13. Based on our findings we speculate that an oligoclonal population of anti-ADAMTS13 antibodies targets residues present within this antigenic surface in the spacer domain.