Promiscuous T cell binding epitopes in two osteoporosis medications lead to different immunogenicity profiles: A Tale of Two Peptides
Brian Roberts
Director of Preclinical Immunology, EpiVax, Inc.
Here we provide a case study to illustrate the concept that understanding the inherent immunogenicity of the active pharmaceutical ingredient (API) is critical to estimating the potential immunogenicity of its impurities. Using in silico prediction models and in vitro validation assays, we have evaluated two common generic peptide drugs, salmon calcitonin, and teriparatide, used in the treatment of post-menopausal osteoporosis. Salmon calcitonin is a foreign peptide, sharing only 50% sequence homology with its human homolog, leading to the development of anti-drug antibodies (ADA) in 35-60% of treated patients, with up to 45% developing neutralizing ADA. Conversely, teriparatide, derived from the n-terminal 34 amino acids of human parathyroid hormone, is a self-peptide considered non-immunogenic with only 2-4% of patients developing ADA following treatment. In silico analysis of both peptides revealed a 9-mer frame predicted to be a promiscuous HLA DR binding epitope referred to as an “EpiBar”. The salmon calcitonin EpiBar found within the center of the peptide is amongst the most foreign regions of the peptide sharing only 33% sequence homology. Modifications to salmon calcitonin derived from synthetic peptide synthesis were not shown to enhance the immunogenicity of salmon calcitonin. In vitro, HLA Binding and ex vivo T cell activation assays indicate that this region is driving the immune response to salmon calcitonin, further supported by reports in the literature. The n-terminal EpiBar associated with teriparatide was shown to have a high degree of homology with several endogenous proteins including human parathyroid hormone (as expected) and β-tubulin, a highly ubiquitous protein found within all cells in the human body. Preliminary studies further indicate that this region of teriparatide functions as a regulatory T cell epitope, or “Tregitope” which when presented to T cells via HLA molecules located on the surface of antigen-presenting cells, are capable of suppressing the immune response. Modifications to this epitope found within synthetic peptide impurities that make this region of teriparatide appear less human led to a dramatic increase in number of responding donors when compared to native teriparatide.
Brian Roberts is the director of Preclinical Immunology at EpiVax, Inc. in Providence, RI. Brian earned his Ph.D. in Cell and Molecular Biology at the University of Vermont where he focused on the role of sex hormones on viral-induced autoimmunity. Since joining EpiVax he has focused on understanding the immunogenic risk potential of synthetic peptide impurities on generic peptide drug products. Brian has lead several FDA-funded projects related to understanding the role of both in silico and in vitro risk assessment methods, including the development of a novel in silico tool (the What if Machine) that prospectively identifies potential peptide impurities from generic drug peptides and scores them based on predicted immunogenic risk. In addition, Brian leads the PANDA generic drug immunogenicity risk assessment program at EpiVax that provides risk assessments for generic drug peptides that are included in FDA ANDA filings.