Ligands leaching from purification columns continue to be a significant and costly problem for producers of therapeutic proteins. Leachates can lead to toxicity and regulatory sanction. With the increasing use of Affibodies in affinity columns (Ronnmark et al. 2002 and Ramstrom et al. 2009), the present study was conducted to assess the relative utility of an anti-IgG Affibody versus Protein A by comparing the affinity of each to two monoclonal antibodies and two Fc-fusion proteins. Functionally, the Affibody should mimic Protein A with similar binding affinities; this study was designed to determine if this is indeed the case.
The two monoclonal antibodies (Mab-A and Mab-B) and two Fc-fusion protein (Fc-A and Fc-B) chosen for this study have different molecular weights and unique molecular structures. Mab-A and Mab-B are IgG1 monoclonal antibodies (fully human and chimeric, respectively) with molecular weights of ~150 kDa. In contrast, the fusion proteins are comprised of an extracellular receptor domain and the hinge and Fc domains of human IgG. The molecular weights of Fc-A and Fc-B are 150 kDa and 92 kDa, respectively.
Materials and Methods
The recombinant monoclonal antibodies and Fc-fusion proteins were reconstituted according to the instructions of the manufacturer. The biotinylated Anti-IgG Affibody molecule (ab31901) and biotinylated Protein A were purchased from Abcam (Cambridge,UK) and Sigma-Aldrich (St. Louis MO), respectively. Surface plasmon-based measurements were obtained at 20ºC with a BiOptix 404Pi™ using CMD-150 sensor chips.
A CMD-150 sensor chip was first activated by a 7 min exposure to EDC/NHS reagent (EDC: 0.23 M 1-ethyl-3-(3-dimethylpropyl)-carbodiimide + NHS: 0.275 M N hydroxysuccinimide in water). The chip was washed with ultrapure water and dried with clean nitrogen.
The chip was next loaded into the instrument and NeutrAvidin (0.2 mg/mL in 30 mM sodium acetate pH 4.9) was loaded onto channels 1 and 3. Reference channels 2 and 4 were treated with BSA (1 mg/mL). The immobilization was allowed to proceed for 60 min. The chip was then removed from the instrument and treated with 1 M ethanolamine to neutralize any remaining active groups.
The chip was then inserted into the BiOptix 404Pi that was previous equilibrated with the assay running buffer (10 mM HEPES, pH 7.3, 150 mM NaCl, 3 mM EDTA, and 0.15% Tween 20). Biotinylated anti-IgG Affibody (0.2 μg per mL) or biotinylated Protein A (0.2 μg per mL) was loaded onto the slide while in the instrument for 60-90 sec with a target loading of 100-150 RUs. In subsequent experiments, Mab-A, Mab-B or Fc-fusion proteins at concentrations ranging from approximately 1 nM to 5 nM were injected at a flow rate of 130 μL/min for 10 min followed by running buffer for 10 min.
After each injection, the slide’s surface was regenerated with 50 mM NaOH, 1 M NaCl. Kinetic analyses were performed using the proprietary BiOptix software package. The protein analyte concentrations and flow rates that were used for the binding studies were prescreened to determine a range where mass transport artifacts would be minimal (data not shown).
The sensorgrams for Mab-A, Mab-B and the Fc-fusion proteins binding to the anti-IgG affibody or Protein A are shown in Figures 1 and 2. Calculated kinetic constants for a global fit of the biomolecular protein-protein interaction based on three concentrations of analyte are shown in Tables 1 and 2.