GB1 Dimerization in Crowders: A Multiple Resolution Approach

Abstract

In-cell protein–protein association, which is crucial in enzyme catalysis and polymerization, occurs in an environment that is highly heterogeneous and crowded. The crowder molecules exclude the reactant molecules from occupying certain regions of the cell, resulting in changes in the reaction thermodynamics and kinetics. Recent studies, both experiment and simulations, revealed that the nature of the interaction between crowder and protein species, in particular the soft interactions, plays an important role in crowder induced effects on protein association. To this end, from a simulation perspective, it is important to decipher the level of structural resolution in a protein-crowder model that can faithfully capture the influence of crowding on protein association. Here, we investigate the dimerization of model system GB1 in the presence of lysozyme crowders at two structural resolutions. The lower resolution model assumes both protein and crowder species as spherical beads, similar to the analytical scaled particle theory model, whereas the higher resolution model retains residue specific structural details for protein and crowder species. From the higher resolution model, it is found that GB1 dimer formation is destabilized in the presence of lysozyme crowders, and the destabilization is more for the side-by-side dimer compared to the domain-swapped dimer, in qualitative agreement with experimental findings. However, the low resolution CG model predicts stabilization of the dimers in the presence of the lysozyme crowder, similar to the SPT model. Our results indicate a nontrivial role of the choice of model resolution in computer simulation studies investigating crowder induced effects.

Computational Approaches to Predict Protein–Protein Interactions in Crowded Cellular Environments

“…Some of them treat at atomistic resolution the biomolecule’s functionally relevant portion, while the rest of the system is described with a CG model, which results in a lower computational cost compared with AA model . One of the most used force fields is MARTINI. , As a downsize, these methods often require lengthy reference all-atom simulations and/or the usage of off-shelf coarse-grained force fields to parametrize the CG-modeled part.…”

“…279 One of the most used force fields is MARTINI. 280,281 As a downsize, these methods often require lengthy reference all-atom simulations and/or the usage of offshelf coarse-grained force fields to parametrize the CGmodeled part.…”

“…However, adding attractive interactions has resulted in different observations have observed that protein crowders can destabilize protein–peptide complexes because of weak nonspecific interactions between test proteins and crowders.…”

“…However, adding attractive interactions has resulted in different observations. 281 In vitro experiments 486 have observed that protein crowders can destabilize protein−peptide complexes because of weak nonspecific interactions between test proteins and crowders. The sticking determined by transient weak interactions can both hinder, via nonfunctional competitive binding, 54 or promote complex formation.…”

“…Ando et al recognized that the stabilization of the native or complex states of macromolecules predicted by their model is in contrast with what was observed by other experimental and computational studies. 43,85,414,418 The fact that considering only volume exclusion results in a wrong prediction of complex stabilization has been recently investigated by Pradhan et al 281 They simulated the dimerization of model system GB1 in the presence of lysozyme crowders at two different resolutions by assuming both protein and crowders as spherical beads and then retaining residuespecific structural details. With the former, they found stabilization of the dimers in the presence of the lysozyme crowder, in accordance with the SPT model.…”