Topology-based potentials and the study of the competition between protein folding and aggregation

Abstract

Topology-based or Go-type potentials have been shown to be very useful in the understanding of the relations between the structure of the native state of a protein and some of its folding characteristics. A different question is whether they can also make such a contribution when the aggregation process of misfolded or partially folded structures is under study. In this work, in spite of the obvious trend of these simulation models toward the native state, we show that there are some aspects about aggregation that can be addressed by topology-based potentials: the role of the thermodynamic characteristics of the transition on preventing the aggregation process, or the larger propensity of highly symmetric protein structures to form domain swapped dimers. In a second part of this work, we use the possibilities of computer simulation as a design of numerical experiments to analyze the fundamental role of intermediate states in the aggregation process of globular proteins.

Sketching protein aggregation with a physics-based toy model

“…Yet, it is the one which, according to our model, results more aggregation-prone. Although the presence of intermediates may favor aggregation, 47 something that had been also checked with simple simulation models, 23 our results show additional evidence that aggregation is also possible for two-state proteins. According to our model, this fact is related to the protein sequence and the compatibility between the hydrophobic interactions and the backbone hydrogen bonds, not only in the folded conformation but also in the domain-swapped configurations and in the β-type aggregates.…”

“…21 For this very reason, these models are not suitable for the study of non-native configurations such as aggregates, although some strategies like symmetrized or "colored" Gō potentials have been suggested. [22][23][24][25] The possibility of using only realistic driving forces (that is, a complete removal of a reference folded structure) is tackled by the purely physics-based coarse-grained potentials, which nevertheless have just proved their applicability in the case of sequenceless peptides. 26,27 An intermediate approach lies in the Sorenson-like potentials; 28,29 they use physics-based interactions for the calculation of hydrophobic interactions, but the local geometry (either helical, turn-like, or extended) is set a priori according to the desired secondary structure elements.…”

Prediction of the structure of a silk-like protein in oligomeric states using explicit and implicit solvent models

“…5h) similar to those that have been observed in the work of Prieto and coworkers for a different type of protein. 63 In this gure b-roll strands from different monomers interact and form b-sheets. This also occurs for the case of simulations with explicit solvent but in these simulations b-strands are less mobile than in the simulations with implicit solvent.…”

How determinant is N-terminal to C-terminal coupling for protein folding?

“…34 Monte Carlo simulations of coarse-grained models have a long history in the folding literature where they have been used to explore many aspects of protein folding, 20,[35][36][37][38][39][40][41][42][43][44][45][46][47] and, more recently, of protein aggregation. [48][49][50][51][52][53][54] A particularly relevant contribution of lattice models was the prediction of the nucleationcondensation mechanism for small, two-state proteins. [55][56][57][58] An important advantage of these models is that their computational feasibility allows accessing very long timescales resulting into accurate measures of thermodynamics and kinetics of folding and aggregation.…”