Thermodynamics and Kinetics of Multi-Protein Binding in Crowded Environments

Abstract

Protein-protein interactions play an essential role in many biological processes inside a cell. The cellular medium is crowded with an ensemble of macromolecules, e.g., proteins, nucleic acids, sugars, lipids, etc. These macromolecules occupy as much as 30% of the cell volume, thereby affecting the stability and rate of multi-protein binding. We investigate the effect of macromolecular crowding on the protein complex formation, using coarse-grained simulation models for proteins and macromolecular crowders. Proteins are represented by a residue-level coarse-grained model that has been shown to yield binding affinities and native structures of various weakly binding protein complexes, in good agreement with experimental data. Macromolecular crowders are modeled as spherical particles or polymeric chains that interact with proteins via repulsive as well as attractive interactions. Repulsive crowders, interacting via excluded volume interactions, of various types stabilize the formation of the protein complex, but the attractive protein-crowder interactions are shown to destabilize the protein complex above moderate attraction strengths. We find that the translational and rotational diffusion for both proteins are slowed down with both the repulsive and attractive crowders, although the anisotropy of the rotational diffusion coefficient increases for both cases. Consequently, the dissociation rate decreases with increasing crowder volume fraction. But the protein association rate is found to increase as a function of the crowder volume fraction in the presence of repulsive crowders, while attractive protein-crowder interactions decrease the association rate. Interestingly, we find that the polymeric crowders can change the protein binding behavior in a complex manner depending on the degree of crowder polymerization and conformational flexibility. We develop a theory, with physically meaningful parameters, that can describe the simulation data very well and provide further insights into the observed results.