Characterizing protein-protein interactions in mAb formulations: A comparative study of kD, B22, and DLVO framework

Understanding protein-protein interactions (PPI) is essential for ensuring the developability of monoclonal antibody (mAb) therapeutics among other applications. This study investigated PPI in three IgG1 mAbs with distinct charge and molecular weight profiles using three approaches: the diffusion interaction parameter (k_D), the osmotic second virial coefficient (B₂₂), and Derjaguin-Landau-Verwey-Overbeek (DLVO) modeling. Light scattering measurements were performed to obtain k_D and B₂₂ values across a range of total ionic strengths (7.5–160 mM) and DLVO modeling was applied to decompose the total interaction energy into van der Waals attraction and electrostatic double-layer repulsion. While each method provided unique insight into PPI, results revealed distinct interaction profiles for each mAb and highlighted the limitations of using a single metric to interpret complex PPI behavior. k_D captured the combined effects of thermodynamic interactions and hydrodynamic drag, while B₂₂ quantified net thermodynamic interactions independent of hydrodynamics. DLVO modeling further enabled mechanistic interpretation by estimating the magnitude and range of attractive and repulsive forces, as well as identifying energy barriers and secondary energy minima indicative of reversible interaction. The orthogonal use of these tools revealed discrepancies between hydrodynamic and thermodynamic interactions, particularly at high ionic strengths, and helped explain nonlinear diffusion trends. Taken together, our findings demonstrate that combining k_D, B₂₂, and DLVO modeling as accessible biophysical characterization tools enables a more robust assessment of colloidal behavior, enhancing early-stage screening of mAbs for developability risks such as aggregation, high viscosity, or self-association.