Back

Surface interaction patches link non-specific binding and phase separation of antibodies

Ausserwöger, H.; Krainer, G.; Welsh, T. J.; Sneideris, T.; Schneider, M. M.; Invernizzi, G.; Herling, T. W.; Lorenzen, N.; Knowles, T. P. J.

2022-03-21 biophysics
10.1101/2022.03.07.483238 bioRxiv
Show abstract

Non-specificity is a key challenge in the successful development of therapeutic antibodies. The tendency for non-specific binding in antibodies is often difficult to reduce via judicious design and, instead, it is necessary to rely on comprehensive screening campaigns. A better understanding of the molecular origins that drive antibody non-specificity is therefore highly desirable in order to prevent non-specific off-target binding. Here, we perform a systematic analysis of the impact of surface patch properties on antibody non-specificity using a designer antibody library as a model system and DNA as a non-specificity ligand. Using an in solution microfluidics approach, we discover patches of surface hydrogen bonding to be causative of the observed non-specificity under physiological salt conditions and suggest them to be a vital addition to the molecular origins of non-specificity. Moreover, we find that a change in formulation conditions leads to DNA-induced antibody liquid-liquid phase separation as a manifestation of antibody non-specificity. We show that this behaviour is driven by a cooperative electrostatic network assembly mechanism enabled by mutations that yield a positively charged surface patch. Together, our study provides a direct link between molecular binding events and macroscopic liquid-liquid phase separation. These findings highlight a delicate balance between surface interaction patches and their crucial role in conferring antibody non-specificity.

Matching journals

The top 10 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.