Understanding protein folding is one of the grand challenges of modern biology.
It is also a critical test of our ability to accurately model interactions in protein systems.
The group participates in an EC funded
on protein (mis)folding (coordinated by Alan Mark) as well as having several projects on folding funded by the
Currently, it is not possible to directly simulate the folding of complete proteins using realistic conditions.
Nevertheless, dramatic progress is being made in the de novo folding of small peptides and even some proteins.
In several cases we have been able to reversibly fold small peptides (~10 aa) from a set of arbitrary starting structures under native conditions with experimental accuracy.
Research on folding is conducted at multiple levels.
The group has studied the folding and stability of Betanova and a series of related peptides designed and characterised by
Dr. Luis Serrano's group
(EMBL in Heidelberg).
Betanova was proposed to form a beta-hairpin motif in solution and shows evidence of co-operative folding.
Extensive simulations indicate, however, that the beta-hairpin motif is only transiently present.
Simulations of a series of mutants also suggest that the peptide is highly flexible.
This has led to a reanalysis of the experimental results and is an example of where the combination
of simulations and experiment can give a highly detailed description of the behaviour of molecules in solution.
In association with the protein (mis)folding network and with collaborators in a project funded by the Volkswagen Stiftung on the conformational control of biomolecules, the group has investigated the stability and assembly of a series of peptides that have been designed to form into coiled-coils (a structure in which two alpha-helical peptides wrap around each other forming a super helix). These simulations showed that it was possible to discriminate between both the stability of different sequences, which could be correlated with experiments, and the different orientations of the helices, which is very hard to determine experimentally. Recent work has also focused on attempts to simulate the aggregation of small beta-sheet forming peptides into extended structures as models of amyloid fibrils. The formation of amyloid fibrils is associated with a wide range of human diseases including Alzheimer's and the various spongiform encephalopathies. Little is known, however, regarding how amyloid formation is initiated at a molecular level.
One of the most challenging questions when simulating protein folding is how to efficiently explore the vast conformational space accessible to a typical system on a realistic time scale. Currently, the inability to sample conformational space efficiently places a severe limitation to the size of system (~10 aa) for which we can accurately predict quantities such as the free energy of folding. One promising new approach for sampling conformational space is the Replica Exchange method, in which several copies (replicas) of the protein or peptide are simulated simultaneously at different temperatures, with exchanges being made between temperatures in a correct thermodynamic manner. This approach has been shown to be very effective on small model systems. The group is currently using Replica Exchange to simulate the folding of a beta-heptapeptide in explicit solvent for which the time scale of folding is known to test the efficiency of the approach on a realistic system. Studies, in conjunction with Guy Lippens of the Pasteur Institute in Lille-France , are also being performed on a 30 amino-acid peptide, the WW domain, in explicit solvent in association with the protein (mis)folding network.
This page was last updated on August the 30th, 2016.