More than 150 protein biopharmaceuticals have been approved for clinical use, but hundreds more have fallen by the wayside. In clinical trials, many therapeutic candidate proteins are frequently found to be immunotoxic and induce anti-drug antibody responses. These responses can lead to allergic reactions, reduction or neutralization of drug activity, and/or cross-reactive immune responses, which can lead to serious adverse events. To avoid clinical-stage therapy failure, there is a growing need for better-informed drug design and testing that includes robust evaluation of candidate protein drug immunogenicity at a preclinical stage.
Currently, assessment of protein drug immunogenicity is not standardized and there is no stand-alone or comprehensive solution. This article outlines the pros and cons of some available techniques and assays, including the REVEAL Immunogenicity System (REVEAL-IS) from ProImmune, Oxford, England.
Immunogenicity is the inherent ability of biopharmaceuticals to induce humoral and/or cell-mediated immune responses. This is influenced by many variables, such as the intrinsic antigenicity of T-cell epitopes in a drug or extrinsic factors, including the disease setting, which can pre-dispose individuals to a higher immune response (e.g. autoimmune diseases) or lower immune response (e.g. compromized immune system in cancers). This is exemplified by the anti-B cell monoclonal antibody Rituxan (rituximab), which is strongly immunogenic in patients with rheumatoid arthritis, but not in cancer patients, where it is primarily applied. Furthermore, certain drugs are designed to be immunomodulatory, and the amount, frequency, and route of drug administration can affect their efficacy more than intrinsic antigenicity.
In order for a protein drug to cause a significant anti-drug-antibody response in a patient, the drug must contain relevant helper T-cell epitopes, which are typically peptides with a specific sequence of 11-20 amino acids. HLA class II molecules present those epitopes to CD4+ T cells and activate the T cell’s immune responses against the epitopes. Therefore, identifying all potential CD4+ T cell epitopes in the protein content of a drug is important in the preclinical assessment of its immunogenicity. In addition, unwanted immunogenicity can be attenuated by mutating key amino acid residues in T-cell epitopes that do not affect drug potency.
HLA class II-binding algorithms have been published in order to try to predict which peptides in a protein sequence can bind HLA class II molecules. This same information could then be used to predict which peptides could be presented to CD4+ T cells. Unfortunately, these algorithms tend to correlate poorly with physical binding data and lack consistency. This is especially problematic when applied across groups of HLA alleles that span a population.1
Alternatively, animal models of disease, autoimmunity, or immunosuppression may be available for assaying and validating therapeutics. The issue of correlating rodent data with their likely effects in humans still remains. This is particularly true when the drugs in development are humanized or generated from a fully human genetic background, making their assessment in animal models particularly inadequate.2,3 Overall, these standard techniques for assessing immunogenicity risk for biological drugs at a preclinical stage present serious shortfalls.
Detailing helper T-cell immune response
The REVEAL-IS system from ProImmune comprises two complementary assays to identify CD4+ T-cell epitopes in a protein drug sequence quickly and confidently. The system first examines antigen responses ex vivo, and then quantifies the MHC-binding properties of peptide epitopes.
The first component of REVEAL-IS is a highly sensitive flow cytometry based T-cell proliferation assay that uses intra cellular CFSE-dye labeling. Unlike traditional assays, which are based on radioactive thymidine corporation, ProImmune’s assay enables the accurate and sensitive determination of the percentage of proliferating CD4+ cells for quantification of drug response (Figure 1). Whole protein drug assessment is carried out using dendritic cells, while fine peptide epitope mapping is performed with naïve T cells. Both proliferation assay formats employ pre-banked naïve donor cells that are HLA-genotyped on DQ, DP, and DR loci, enabling correlation of responses with HLA sub-type. The use of pre-typed donors has the advantage that donors can be selected to mirror the population of interest for a particular disease or geographic area. When evaluating immunogenicity, both the strength and frequency of a response should be considered; this is achieved through scoring the number of donors and also the degree to which they respond.
The second component of REVEAL-IS monitors the HLA class II-binding specificity of putative epitope sequences. This is measured using high-throughput in vitro HLA class II-peptide binding and stability assays that cover a population-spanning group of class II alleles. These assays can determine the exact binding sequence in a binding peptide and the HLA restriction of that sequence. The assay results provide key information for epitope ranking and possible modification (Figure 2).
Either alone or in combination, the two components of REVEAL-IS are used to define the sequence and HLA restriction of epitopes, and rank their potential immunogenicity. This suite of assays provides a means of comparing drug candidates in the lead optimization phase of development. While the assessment of a drug’s T-cell epitopes can never be absolutely predictive of the immunogenicity observed in the clinic, it is always better to obtain more information at an earlier stage. Improvements to the immunogenicity screening framework surrounding preclinical drug development will facilitate the rapid development of effective and safe treatments at reduced costs.
1. Gowthaman U, Agrewala JN, In Silico Tools for Predicting Peptides Binding to HLA-Class II Molecules: More Confusion than Conclusion. J Proteome Res. 2008; 7(1): 154–163.
2. Nielsen M, Lund O, Buus S, Lundegaard C. MHC class II epitope predictive algorithms. Immunology. 2010; 130(3):319-28.
3. Bugelski PJ, Treacy G. Predictive power of preclinical studies in animals for the immunogenicity of recombinant therapeutic proteins in humans. Curr Opin Mol Ther. 2004; 6(1):10-6.