(M0930-08-49) Multimodal Characterization of Protein–Protein Interactions in Monoclonal Antibodies Using k_D, B₂₂, and DLVO

Purpose: Reliable characterization of protein-protein interactions (PPI) is essential for identifying monoclonal antibody (mAb) candidates with desirable colloidal stability and favorable formulation behavior. However, relying on a single biophysical metric often yields an incomplete or misleading picture, especially for complex cases. This study aims to evaluate the combined application of three orthogonal approaches, the diffusion interaction parameter (k_D), the osmotic second virial coefficient (B₂₂), and DLVO (Derjaguin–Landau–Verwey–Overbeek) modeling, in elucidating PPI behavior in three IgG1 mAbs with distinct physicochemical properties. Each method captures different facets of intermolecular interactions: k_D accounts for the influence of thermodynamic and hydrodynamic interparticle interactions on the diffusion behavior [1], B₂₂ reveals non-ideal solution behavior arising from pairwise solute interactions [2], and DLVO decomposes total interaction energy into mechanistic components [3]. By integrating these complementary approaches, this work seeks to provide a more complete understanding of PPI and to assess the potential of this framework in guiding early-stage developability assessments for biotherapeutics.

Methods: Three IgG1 mAbs were extracted from commercial formulations, buffer-exchanged into 10 mM histidine (pH 6.0), and tested across a range of ionic strengths (7.5–160 mM) by adding NaCl to the formulation. Dynamic and static light scattering were performed simultaneously with an 830 nm laser wavelength and a scattering angle of 158° to determine k_D and B₂₂, respectively. A concentration series of the mAb samples were tested ranging from 1.0 to 10.0 mg/mL in 1.0 mg/mL increments at each ionic strength. DLVO modeling was performed by fitting B₂₂ values to a simplified interaction energy equation, enabling estimation of model parameters such as effective protein charge and Hamaker coefficient. Interaction energy profiles were analyzed as a function of intermolecular distance and ionic strength to explore mechanistic insights into PPI.

Results: At low ionic strength, mAb1 displayed strong repulsion (k_D = +54 mL/g), while mAb2 (k_D = -8.4 mL/g) and mAb3 (k_D = -16.6 mL/g) showed attraction-dominated interactions. The introduction of electrolyte into the formulations led to a marked decline in k_D values for all three mAbs. mAb1 showed the most substantial drop, exceeding 98%, followed by a 28% decrease for mAb3 and a modest 5.4% reduction for mAb2. These differences reflect the charge screening effect of the added ions, which diminish electrostatic repulsion between molecules. The magnitude of k_D reduction was found to align with the overall charge of each antibody, with the more highly charged mAb1 and mAb3 exhibiting the greatest sensitivity to ionic strength. Further increasing ionic strength revealed nonmonotonic trends in k_D and B₂₂, with decoupling observed in some cases indicating contributions from transient clustering and hydrodynamic effects. DLVO modeling offered an additional mechanistic layer by estimating the magnitude and range of van der Waals and double-layer interactions. The model revealed the presence of a secondary energy minimum in mAb3, associated with reversible self-association, which explained its unique diffusion behavior.

Conclusion: This study demonstrates the power of integrating k_D, B₂₂, and DLVO modeling as orthogonal tools for characterizing protein–protein interactions in therapeutic mAbs. Each parameter captures distinct physical phenomena, and their combined application allows for a more complete and nuanced understanding of colloidal behavior. This multiparametric strategy offers a more reliable basis for guiding formulation decisions, ultimately improving the predictability of risks such as aggregation, opalescence, and viscosity in biologics drug product development.

References: [1] J.S. Kingsbury, C.G. Starr, Y.R. Gokarn, A scaling relationship between thermodynamic and hydrodynamic interactions in protein solutions, BIOPHYS J, 123 (2024) 3871-3883.
[2] J.J. Janke, C.G. Starr, J.S. Kingsbury, N. Furtmann, C.J. Roberts, C. Calero-Rubio, Computational Screening for mAb Colloidal Stability with Coarse-Grained, Molecular-Scale Simulations, J PHYS CHEM B, 128 (2024) 1515-1526.
[3] J.N. Israelachvili, Intermolecular and surface forces, 3rd Edition ed., Burlington, MA: Academic Press, Burlington, MA, 2010.

Acknowledgements: This work is supported by the National Science Foundation under Award No.2242763.