Atomic-detailed milestones along the folding trajectory of protein G

The high computational cost of carrying out molecular dynamics simulations of even small-size proteins is a major obstacle in the study, at atomic detail and in explicit solvent, of the physical mechanism which is at the basis of the folding of proteins. Making use of a biasing algorithm, based on the principle of the ratchet-and-pawl, we have been able to calculate eight folding trajectories (to an RMSD between 1.2A and 2.5A) of the B1 domain of protein G in explicit solvent without the need of high-performance computing. The simulations show that in the denatured state there is a complex network of cause-effect relationships among contacts, which results in a rather hierarchical folding mechanism. The network displays few local and nonlocal native contacts which are cause of most of the others, in agreement with the NOE signals obtained in mildly-denatured conditions. Also nonnative contacts play an active role in the folding kinetics. The set of conformations corresponding to the transition state display phi-values with a correlation coefficient of 0.69 with the experimental ones. They are structurally quite homogeneous and topologically native-like, although some of the side chains and most of the hydrogen bonds are not in place.

💡 Research Summary

The authors address the long‑standing computational bottleneck that hampers atomistic molecular‑dynamics (MD) studies of protein folding in explicit solvent. By implementing a biasing scheme inspired by the ratchet‑and‑pawl principle, they are able to accelerate the sampling of folding pathways without artificially distorting the underlying physical trajectory. The method locks already‑formed native contacts, preventing their reversal, while allowing new contacts to form freely; this selective bias dramatically reduces the time required to reach near‑native conformations.

Using this algorithm, eight independent folding trajectories of the B1 domain of protein G (≈56 residues) were generated on modest computational resources. All final structures lie within 1.2 Å–2.5 Å root‑mean‑square deviation (RMSD) of the experimentally determined native state, demonstrating that high‑performance supercomputers are not a prerequisite for achieving atomic‑level accuracy in explicit water.

Analysis of the trajectories reveals a hierarchical network of cause‑effect relationships among contacts in the denatured ensemble. A small subset of both local and non‑local native contacts act as “drivers,” triggering the formation of many other contacts. These driver contacts correspond to NOE signals observed under mildly denaturing conditions, confirming that certain native interactions are pre‑organized even before the protein reaches its folded basin.

Importantly, the simulations also highlight an active role for non‑native contacts. Transient non‑native interactions appear early in the folding process, providing scaffolding that facilitates the correct alignment of driver contacts. As folding proceeds, these non‑native contacts dissolve, suggesting that they function as kinetic way‑points rather than mere errors. This observation challenges the traditional view that non‑native contacts are simply off‑pathway obstacles.

The set of structures identified as the transition state (TS) exhibits a Pearson correlation of 0.69 with experimental φ‑values, indicating a substantial agreement between simulation and experiment. Structurally, TS ensembles are topologically native‑like: the overall secondary‑structure topology (α‑helix and β‑sheet) is already established, yet many side‑chain orientations and hydrogen‑bond networks remain incomplete. This partial ordering mirrors experimental findings that the TS is “partially formed” and supports the notion that the TS is defined more by kinetic bottlenecks than by full structural perfection.

Overall, the study demonstrates that a ratchet‑and‑pawl bias can produce realistic folding pathways for a small protein in explicit solvent with modest computational effort. The hierarchical contact network and the functional contribution of non‑native contacts provide fresh insights into the folding mechanism, suggesting that folding proceeds through a small number of key native contacts guided by transient, supportive non‑native interactions. The methodology offers a scalable route to explore larger proteins or complexes, and its quantitative agreement with experimental φ‑values validates its potential as a predictive tool for protein‑folding studies.