High-throughput characterization of colloidal properties of monoclonal antibodies

Abstract

Monoclonal antibodies (mAbs) are currently the focus of intense interest for therapeutic applications due to their potent bioactivities. However, the process of identifying and developing mAbs into approved therapeutics is an extremely challenging task. One of the key challenges is to identify mAb candidates with excellent biophysical properties, including high solubility (colloidal stability) and low viscosity at elevated antibody concentrations required for subcutaneous delivery. However, it is difficult to characterize the biophysical properties of mAbs early in the discovery process due to the large number of antibody candidates (hundreds to thousands) as well as their low concentrations (micrograms per mL) and purities (cell culture supernatants). We aim to address some of these challenges by developing a high-throughput assay (affinity-capture self-interaction nanoparticle spectroscopy, AC-SINS) capable of measuring weak antibody colloidal interactions that is compatible with extremely dilute and unpurified mAb solutions. AC-SINS uses gold nanoparticles coated with polyclonal antibodies to capture the target human mAbs on the surface of the particles, which amplifies antibody self-interactions due to polyvalency. Attractive mAb self-interactions reduce interparticle separation distances between conjugates, which leads to red shifts in the plasmon wavelength (wavelength of maximum absorbance).

In this work, we have significantly improved the AC-SINS method and used it for differentiating between mAbs with different levels of self-association in serum as well as in other formulation solutions. We find that the AC-SINS measurements are well correlated with several conventional antibody biophysical properties (solubility, aggregation) and non-conventional biophysical properties (non-specific interactions with non-adsorptive chromatography columns). Moreover, we have identified sequence features of antibodies that are responsible for their variable biophysical properties, and find strong correlation between our self-interaction measurements and computational predictions based on the sequences of the antibody binding loops. These findings highlight the potential of using high-throughput biophysical methods for rapid identification and optimization of mAbs for high concentration therapeutic applications.