In NOD mice, immune targeting of insulin/proinsulin is essential for diabetes development. Insulin B chain amino acids 9-23 (B:9-23) are necessary for NOD diabetes as a single amino acid mutation (B16Tyr to Ala) abrogates diabetes onset. CD4 T cell recognition occurs when the insulin B: 9-23 peptide is bound in a low affinity register (termed register 3) to the murine MHC class II molecule, I-Ag7. Targeting the tri-molecular complex of I-Ag7–insulin B:9-23 peptide-T cell receptor is a rationale target for immune intervention. Our previous data showed immunization with recombinant I-Ag7–insulin B:9-23 protein induced specific antibodies and delayed NOD diabetes onset. We hypothesize that antibodies bound to the insulin B:9-23 in the peptide binding groove of MHC class II will modulate autoimmune diabetes.
To test our hypothesis we generated a MHCII/insulin B:9-23 in register 3 monoclonal antibody, named mAb287. MAb287 binds specifically to I-Ag7-B:9-23 register 3 complexes but not I-Ag7 with alternative peptides, I-Ab linked peptide complexes, or free insulin B:9-23 peptide. Furthermore, mAb287 selectively bound insulin B:9-23 to IAg7 in register 3 but not in other registers. The crystal structure of mAb287 and I-Ag7-insulin register 3 defined the interactions between mAb287, insulin peptide, and I-Ag7. The mAb287 inhibited T cells in responding to soluble B:9-23 peptide presented by antigen presenting cells in vitro. In vivo, mAb287 significantly decreased the number of insulin specific T cells in islets measured with insulin tetramers and delayed the development of diabetes administrated at early stage or pre-diabetic stage.
Our data provides proof of concept that a monoclonal antibody targeting a primary auto-antigen in a defined register of binding to MHC class II is able to modulate disease. This may be applicable to a wide range of autoimmune disorders in which the auto-antigen epitope binding register is known.