The use of monoclonal antibodies for the treatment of diseases has increased in recent years. To date, 44 therapeutic monoclonal antibodies have been approved in the United States and Europe, with 14 currently under review (as of May 2015) and more in development. Pre-clinical and clinical development of antibody drugs requires intensive pharmacokinetic and immune response testing in non-human and human subjects. This typically involves the use of anti-idiotypic antibodies, which are essentially antibodies that bind to specific regions of other antibodies. In order to be effective, these anti-idiotypic antibodies must demonstrate high affinity and specificity. In addition, they should be easily manufactured and available at consistent quality throughout the drug development process to avoid having to source new reagents.
Phage display of large recombinant antibody libraries can be used to rapidly generate anti-idiotypic antibodies to any antibody paratope (Fig. 1). However, the affinity of the antibody to its target antigen is a critical aspect that determines the performance of these anti-idiotypic antibodies in pharmacokinetic and immune response assays. Accordingly, selection of antibodies based on binding strength can significantly improve assay sensitivity. This has been a challenge in antibody generation projects as several unique candidate antibodies are obtained after initial screening. For example, the high-throughput HuCAL® antibody generation process results in 100-200 unique antibodies of confirmed specificity by ELISA after the primary screening process. Only a small fraction of these antibodies move on for further characterization, and because they are selected based on specific binding in ELISA and not on other characteristics such as binding strength, high affinity antibodies that could perform better in certain assays could be overlooked.
To improve the antibody generation process in order to obtain unique antibodies based on antigen specificity as well as binding strength, Ylera et al. (2013) developed a secondary screening step to rank antibodies of confirmed specificity based on their kinetic koff-rate. The antibody drugs trastuzumab and cetuximab were used as antigen examples. Trastuzumab and cetuximab are monocloncal antibodies used for the treatment of Her-2/neu positive breast cancer and colorectal cancer, respectively. Anti-idiotypic antibodies were generated against the respective antibodies following the process shown in figure 1. The specificity of the generated antibodies was tested by ELISA in a primary screening step. In an additional secondary screening step their koff-rates were determined by biolayer interferometry. For the anti-cetuximab project, 53 unique antibody sequences were identified with affinities in the range of 0.5 – 174 nM. The off-rates of antibodies selected by the highest ELISA signal were compared to those ranked by the secondary koff-rate screening method (see Fig. 4 in Ylera et. al). The addition of this secondary screening step resulted in 8 unique antibodies with affinities of 0.5 – 47 nM, obtained from sequencing 20 clones with the lowest off rates. This indicates that the secondary screening step is suitable for choosing antibodies with higher affinities. Interestingly, the antibody clone with the best koff-rate would have been missed if ELISA signal strength was the only selection criteria.
A similar procedure was performed with antibodies against trastuzumab and one of the resulting anti-trastuzumab antibodies underwent an affinity maturation process, which improved the affinity range to 65 pM – 4 nM. In this case,the addition of the secondary screening step also allowed detection of antibodies with the lowest koff-rates.
This improved antibody generation process described by Ylera et al. can be used to achieve more sensitive assays for analysis and evaluation of therapeutic antibodies without compromising time to antibody selection.
HuCAL® is a registered trademark of MorphoSys AG.