Therapeutic antibodies (TMabs) are part of the best successful therapeutic products of the last decades. They are currently used to treat many inflammatory diseases such as rheumatoid arthritis (RA) and bowel diseases and represent worldwide a market of several billions of dollars. However they have the major drawback to be potentially immunogenic and therefore might elicit anti-drug antibodies (ADA). ADA could dramatically reduce the efficacy of the drugs or might provoke allergic reactions. Because generally self-proteins are less immunogenic than foreign proteins, the sequence of therapeutic antibodies has been humanized. However humanization even at the highest level does not fully guarantee the lack of immune responses demonstrating the important need to know more about ADA response. Because T cells are known to initiate the ADA response, we are currently investigating the T cell response to immunogenic therapeutic antibodies. With the perspective of immunogenicity prediction, we quantified the number of very rare T cells specific for therapeutic antibodies in the blood of normal donors and found a good concordance between the number of T cells specific to them and their respective clinical immunogenicity level. We then identified the CD4 T cell epitopes of four immunogenic TMabs with different levels of humanization, namely the chimeric antibodies Infliximab (Ifx) and Rituximab (Rtx), the humanized antibody Natalizumab (Ntz) and the fully human Adalimumab (Adm). CD4 T lymphocytes were expanded by several weekly rounds of stimulation with autologous dendritic cells loaded with each of the investigated antibodies and the T cell specificity was assessed by IFNg ELISPOT using overlapping peptides encompassing the whole sequence of their variable parts. Nine epitopes were identified in the VL and VH chains of Rituximab and Infliximab. They overlap CDR or FR regions of both chimeric antibodies and some of them are shared by multiple donors. As inferred from binding experiments, T cell epitopes often exhibited a good affinity for HLA-DR molecules found in the responding donors. Nine CD4 T cell epitopes were found in the VH and VL parts of the humanized therapeutic antibody Natalizumab while the fully human antibody Adalimumab hosted 10 T cell epitopes. As a result, the number of T cell epitopes is very similar across the different therapeutic antibodies but their location is highly variable from one antibody to another one. Finally to assess the clinical relevance of the identified T cell epitopes, we evaluated the ability of Ifx and Rtx T cell epitopes to stimulate T cells of patients having developed ADA. Two third of the T cell epitopes identified from the healthy donors stimulated PBMCs from ADA+ patients and promoted the secretion of a diversity of cytokines. These data emphasize the predictive value of evaluating the T cell repertoire of healthy donors to anticipate and prevent immunogenicity of therapeutic antibodies. Together our data provide new insights on the origin of immunogenicity of chimeric, humanized and human therapeutic antibodies.
Disclosure of Interest None declared