Several mechanisms are known to cause monomeric protein misfolding. Coarse-grained simulations have predicted an additional mechanism exists involving off-pathway, noncovalent lasso entanglements, which are long-lived kinetic traps and structurally resemble the native state. Here, we examine whether such misfolded states occur in long-timescale, all-atom folding simulations of ubiquitin and λ-repressor. We find that these entangled misfolded states are populated in higher-resolution models. However, because of the small size of ubiquitin and λ-repressor, these states are short-lived. In contrast, coarse-grained simulations of a larger protein, IspE, predict that it populates long-lived misfolded states. Using an Arrhenius extrapolation applied to all-atom simulations, we estimate that these IspE misfolded states have lifetimes similar to the native state while remaining soluble. We further show that these misfolded states are consistent with the structural changes inferred from limited proteolysis and cross-linking mass spectrometry experiments. Our results indicate that misfolded states composed of non-native entanglements can persist for long timescales in both all-atom simulations and experiments.

Stretched-exponential protein refolding kinetics, first observed decades ago, were attributed to a nonnative ensemble of structures with parallel, non-interconverting folding pathways. However, the structural origin of the large energy barriers preventing interconversion between these folding pathways is unknown. Here, we combine simulations with limited proteolysis (LiP) and cross-linking (XL) mass spectrometry (MS) to study the protein phosphoglycerate kinase (PGK). Simulations recapitulate its stretched-exponential folding kinetics and reveal that misfolded states involving changes of entanglement underlie this behavior: either formation of a nonnative, noncovalent lasso entanglement or failure to form a native entanglement. These misfolded states act as kinetic traps, requiring extensive unfolding to escape, which results in a distribution of free energy barriers and pathway partitioning. Using LiP-MS and XL-MS, we propose heterogeneous structural ensembles consistent with these data that represent the potential long-lived misfolded states PGK populates. This structural and energetic heterogeneity creates a hierarchy of refolding timescales, explaining stretched-exponential kinetics. Copyright © 2025 The Authors, some rights reserved.

The concomitant folding of a protein with its synthesis on the ribosome is influenced by a number of different timescales including the translation rate. Here we present a kinetic formalism to describe cotranslational folding and predict the effects of variable translation rates on this process. Our approach, which utilizes equilibrium data from arrested ribosome nascent chain complexes, provides domain folding probabilities in quantitative agreement with molecular simulations of folding at different translation rates. We show that the effects of single codon mutations in messenger RNA that alter the translation rate can lead to a dramatic increase in the extent of folding under specific conditions. The kinetic formalism that we discuss can describe the cotranslational folding process occurring on a single ribosome molecule as well as for a collection of stochastically translating ribosomes. © 2012 Macmillan Publishers Limited. All rights reserved.

An emerging paradigm in the field of in vivo protein biophysics is that nascent-protein behavior is a type of nonequilibrium phenomenon, where translation-elongation kinetics can be more important in determining nascent-protein behavior than the thermodynamic properties of the protein. Synonymous codon substitutions, which change the translation rate at select codon positions along a transcript, have been shown to alter cotranslational protein folding, suggesting that evolution may have shaped synonymous codon usage in the genomes of organisms in part to increase the amount of folded and functional nascent protein. Here, we develop a Monte Carlo-master-equation method that allows for the control of nascent-chain folding during translation through the rational design of mRNA sequences to guide the cotranslational folding process. We test this framework using coarse-grained molecular dynamics simulations and find it provides optimal mRNA sequences to control the simulated, cotranslational folding of a protein in a user-prescribed manner. With this approach we discover that some codon positions in a transcript can have a much greater impact on nascent-protein folding than others because they tend to be positions where the nascent chain populates states that are far from equilibrium, as well as being dependent on a complex ratio of time scales. As a consequence, different cotranslational profiles of the same protein can have different critical codon positions and different numbers of synonymous mRNA sequences that encode for them. These findings explain that there is a fundamental connection between the nonequilibrium nature of cotranslational processes, nascent-protein behavior, and synonymous codon usage. © 2015 American Chemical Society.