Highly concentrated protein solutions are pharmaceutically relevant for subcutaneous delivery of biologics, however the high concentration also introduces significant effects from protein-protein interaction (PPI) in solution which can increase viscosity and decrease stability. Small angle X-ray scattering (SAXS) is an increasingly popular technique to understand PPI at high concentration because it can probe mAb structuring over a range of length scales. However, SAXS alone cannot describe how PPI affect viscosity because it does not give high-resolution, real-space information about microscopic short-range order in the sample. In this work, microscopic information is obtained using a coarse-grained “12-bead” molecular dynamics (MD) simulation model with anisotropic short-range attractions placed on specific beads, e.g., in the fragment antigen-binding (F_AB) and fragment crystallizable (F_C) region. Interactions for which simulated and experimental structure factors best match are determined, and simulated cluster-size distributions are used together with a phenomenological model which accounts for inefficient packing of clusters and cluster rigidity to predict zero-shear viscosity.

Here, we present two case studies in which the bead-specific attractions are assigned to capture interactions characteristic of different protein and solution conditions. The first considers attractions assigned to modulate the effect of different crystallizable fragment (F_C) domains on Vedolizumab (Vedo) on PPI and how the F_C domains impact shear viscosity at a protein concentration of 130 mg/mL. The second study considers bead attractions placed on beads associated with the complimentary determining region and the F_C region on a proprietary mAb known as “mAb 2” to understand the effects of pH, cosolute, and concentration on PPI and protein solution viscosity.

Together, these studies indicate that protein clustering and cluster rigidity are essential for understanding protein viscosity at high concentration, and that SAXS provides information critical for linking coarse-grained interactions and structure with experimental shear viscosity.

Richard Gillilan

Appearances