The induction of anti-drug antibodies (ADA) in patients is a major concern in the development of biological therapeutics. The detection and quantification of ADAs is of great importance to the pharmaceutical and biotechnology industry because any immune response against a therapeutic can affect the safety, pharmacokinetic profile, and efficacy of the drug. In fact, the U.S. Food and Drug Administration strongly recommends that immunogenicity assays that detect, quantitate, and characterize ADAs be implemented and that any observed toxicity or pharmacologic peculiarities be correlated to the presence of ADAs (1). In this application note, we demonstrate how immunogenicity screens can be performed on the BiOptix 404pi biosensor using monoclonal (mAb) or polyclonal antibodies (pAb) in various concentrations of human serum. Specifically, in this study a rabbit antimouse pAb (RAM) served as the “biologic drug” and a mouse anti-digoxin mAb (Fitzgerald Industries International) served as the “anti-drug antibody (ADA).”
Four flow cells (Fc) on four CMD200m sensor chips were immobilized with rabbit anti-mouse (RAM) pAb (GE Healthcare) in 10 mM sodium acetate buffer pH 5.0 using standard EDC/NHS coupling chemistry and blocking by ethanolamine. The running buffer for immobilization was HEPES-buffered saline with 0.05% polysorbate
20 pH 7.4 (HBS-P). Immobilization levels achieved on each Fc are shown in Table 1.
All immunogenicity data were collected on the BiOptix 404pi biosensor in 4×1 mode. The running and sample buffer for the immunogenicity experiments was HBS-P with 1 mg/mL soluble carboxymethyldextran added (HBS-PCM). Serum samples were injected for ten minutes and dissociation was followed for three minutes at a flow rate of 5 μL/min. Five buffer blanks were injected to “warmup” the surface before commencement of each immunogenicity screen. Buffer blanks were also injected every sixth injection so systematic noise on each flow cell could be subtracted out during data processing. Sensor surfaces were regenerated using a three min injection of 10 mM glycine pH 1.7. Serum from 30 human donors was purchased from BioreclamationIVT. The cut point and LOD were determined for the anti-digoxin mAb in four independent experiments with sensor chips 1-4 in human serum concentrations ranging from 10% -100%.
Results and Discussion
Thirty individual, blank human samples at 10%, 25%, 50%, and 100% serum concentrations were injected in 4×1 mode across four independent flow cells with four serum samples being interrogated each cycle. The serum concentrations were prepared using concentrated HBS-PCM buffer to always arrive at 1x HBS-PCM upon dilution (except for the 100% serum samples) to the desired serum concentration. Seven or eight different concentrations of mouse anti-digoxin mAb (ADA) were prepared in pooled human serum from the thirty human donors at the same serum concentrations as the blanks. Standard solutions were prepared by spiking in the highest ADA concentration and then performing six or seven three-fold serial dilutions to generate a calibration curve at a given serum concentration. Octuplicates of each concentration standard only required two cycles on the instrument to be completed owing to the higher throughput afforded by the 4×1 mode. The whole immunogenicity screen, depending on the number of concentration standards used, only took 32-34 cycles (including five buffer blank injections to “warm-up” the sensor surfaces) for a total elapsed time of ~14 hours for 128 to 136 injections.
Sensorgrams were normalized to the amount of RAM immobilized on a specific flow cell and multiplied by 10000 before statistical analysis. This multiplication allowed the normalized sensorgrams to achieve roughly the same magnitude in response units (RU) present before normalization and made for easier visualization of the data. Figure 1 shows normalized sensorgrams for the immunogenicity screen in 50% serum that are representative of the type of sensorgrams that are collected in immunogenicity screening.