Critical examination of the inherent-structure-landscape analysis of two-state folding proteins

Recent studies attracted the attention on the inherent structure landscape (ISL) approach as a reduced description of proteins allowing to map their full thermodynamic properties. However, the analysis has been so far limited to a single topology of a two-state folding protein, and the simplifying assumptions of the method have not been examined. In this work, we construct the thermodynamics of four two-state folding proteins of different sizes and secondary structure by MD simulations using the ISL method, and critically examine possible limitations of the method. Our results show that the ISL approach correctly describes the thermodynamics function, such as the specific heat, on a qualitative level. Using both analytical and numerical methods, we show that some quantitative limitations cannot be overcome with enhanced sampling or the inclusion of harmonic corrections.

💡 Research Summary

The paper presents a systematic evaluation of the Inherent Structure Landscape (ISL) approach as a reduced‑dimensional framework for describing the thermodynamics of two‑state folding proteins. While previous work applied ISL to a single protein topology, the authors extend the analysis to four proteins of varying size and secondary‑structure composition: a small α‑helical protein (CI2), a β‑sheet WW domain, an α/β mixed Protein G B1 domain, and a larger α‑helical/β‑sheet λ‑repressor. For each system, extensive atomistic molecular‑dynamics (MD) simulations were performed over a temperature range that spans the folding transition. All sampled configurations were quenched to their nearest potential‑energy minima, generating a comprehensive set of inherent structures. The ISL method then reconstructs the full partition function by assigning each inherent structure a Boltzmann weight exp(−βE) and, optionally, a harmonic vibrational contribution derived from the Hessian eigenvalues at the minimum.

Two central assumptions underlie the ISL formalism: (i) the density of inherent‑structure energies, ρ(E), is temperature‑independent, and (ii) the free‑energy contribution of each basin can be captured by a harmonic (second‑order) approximation. The authors test these assumptions analytically and numerically. By comparing ρ(E) obtained at different temperatures, they find that near the melting temperature (Tm) the population of high‑energy unfolded basins and low‑energy folded basins shifts dramatically, causing a subtle temperature dependence of ρ(E). This violates assumption (i) in the critical transition region.

Regarding assumption (ii), the inclusion of harmonic vibrational corrections improves the qualitative shape of the specific‑heat (Cp) curves but does not eliminate quantitative discrepancies. The reconstructed Cp peaks are systematically lower and narrower than those obtained directly from the MD trajectories or from experimental calorimetry. The authors attribute this residual error to non‑harmonic motions within basins and to the neglect of free‑energy barriers separating distinct basins. To probe whether enhanced sampling could remedy the problem, they augment the data set with metadynamics and replica‑exchange simulations, thereby expanding the inherent‑structure catalogue. Although the enlarged dataset reduces statistical noise, the quantitative gap persists, indicating that the limitation is intrinsic to the ISL’s basin‑wise additive formulation.

Overall, the ISL method successfully captures the qualitative thermodynamic signatures of two‑state folding: the location of the folding transition, the overall trend of Cp versus temperature, and the relative stability of folded versus unfolded states. However, it falls short when precise free‑energy differences or absolute Cp values are required. The paper concludes by suggesting possible extensions: (1) explicit calculation of inter‑basin transition barriers to incorporate non‑additive effects, (2) higher‑order anharmonic corrections or quasi‑harmonic analyses, and (3) machine‑learning‑driven clustering of inherent structures to achieve a more faithful representation of the underlying energy landscape.

In summary, this work validates ISL as a valuable, computationally efficient tool for gaining insight into protein folding thermodynamics, while simultaneously highlighting its current theoretical constraints and outlining concrete avenues for future methodological improvements.