Electrostatically Mediated Protein-Protein Interactions for Monoclonal Antibodies: A Combined Experimental and Coarse-Grained Molecular Modeling Approach

Protein-Protein Interactions, Clustering, and Rheology for Bovine IgG up to High Concentrations Characterized by Small Angle X-Ray Scattering and Molecular Dynamics Simulations

“…[33][34][35] Similar models were also used to successfully predict S(0) behavior of a weakly attractive mAb at high concentrations determined via SLS by tuning the model fit parameters using low-concentration data. [36][37][38] Furthermore, a 12-bead model was used to study SANS data across multiple concentrations to predict cluster sizes and relate them to mAb solution viscosity and diffusivity. 39 In these studies, the interaction potential on each bead is determined from the theoretical charge calculated from the sequence and also an electrostatic scaling parameter to account for differences between the theoretical charge and the effective charge in solution.…”

Product-Specific Impact of Viscosity Modulating Formulation Excipients During Ultra-High Concentration Biotherapeutics Drug Product Development

“…Pro had no effect on viscosity at 100 mg/mL or 137 mg/ mL but increased viscosity at 230 mg/mL compared to the control formulation. Dilute solution interaction parameters such as k D and B 22 are often used to predict high protein concentration behavior [32][33][34][35][36] with the assumption that these inherent repulsive or attractive interactions will be maintained upon solution concentration. However, our findings with mAbN indicate that trends observed at relatively lower (100 mg/mL in this study) protein concentrations are not always representative or predictive of effects at higher (137 mg/mL in this study) protein concentrations.…”

“…Often in early development, there is limited protein material for formulation screening, and properties measured in dilute solutions such as the interaction parameter k D 32,33 or osmotic second virial coefficient B 22 [34][35][36] are used to predict protein behavior at high concentrations. These extrapolations and conclusions are challenging because dilute solution properties may not account for all the changes in molecular configurations and PPIs influencing viscosity as the protein concentration in solutions is increased to the high and ultra-high concentration regime.…”

Toward Drug-Like Multispecific Antibodies by Design

“…Coarse-grain molecular simulations at the one bead-per-amino-acid degree of coarse-graining have been used to evaluate protein-protein interactions computationally, as this amount of detail was found to best balance trade-offs between accuracy and computational load [ 169 ]. The results from these models have been correlated with light scattering data across a range of formulation conditions (pH 5, 6.5, and 8, and ionic strengths varying between 5 to 300 mM) [ 170 ]. These models were used to effectively predict the B 22 values of three monoclonal antibodies that exhibit a range of colloidal behaviors, demonstrating the utility of the potential generality of the approach.…”