NMR-Informed Molecular Modeling Uncovers the Conformational Landscape of Chaperone Binding with Unfolded Substrate

Abstract

Molecular chaperones interact with unfolded proteins to prevent aberrant folding and aggregation in the crowded cellular milieu. Despite their critical role in maintaining protein homeostasis, achieving detailed models of the highly dynamic interaction between a chaperone and an unfolded substrate has proven to be a difficult task. We address this challenge by devising a methodology that fuses site-specific NMR information with coarse-grained molecular dynamics simulations to achieve a residue-level description of chaperone-substrate interaction. As a model system, we decided to consider the interaction of the small, ATP-independent chaperone Spy with its in vivo substrate Im7. Specifically, we develop an automated procedure that optimizes coarse-grained force fields for the individual partners against experimentally measured carbon chemical shifts. Additional NMR data is used to validate the models. We then combine the force fields of the two binding partners with a generic inter-protein potential to perform simulations of the dynamic chaperone-substrate complex. We observe an overall decrease in global dynamics of the Spy chaperone that is counter-balanced by an increase in the local dynamics in its flexible loop region. Increased loop flexibility appears to facilitate chaperone interaction with multiple partially folded states that are sampled by the Im7 substrate. Moreover, interaction with the Spy chaperone slows down conformational exchange in the Im7 substrate, while also shifting the substrate folding landscape toward more structured states. Our hybrid strategy provides an avenue to investigate other heterogeneous biomolecular complexes through the integration of NMR data with efficient computational models.