Scientists continue to be puzzled by how proteins fold into their three-dimensional structures. Small single-domain proteins may hold the key to solving this puzzle. These proteins often fold into their three-dimensional structures by crossing only a single barrier. The barrier consists of an ensemble of extremely short-lived transition state structures which cannot be observed directly. However, mutations that slightly shift the folding barrier may provide indirect access to transition states. Researchers from the Max Planck Institute of Colloids and Interfaces and the University of California, San Francisco have suggested a novel method to construct transition state structures from mutational data (PNAS, July, 2005).
Proteins are chain molecules assembled from amino acids. The precise sequence of the twenty different types of amino acids in a protein chain is what determines which structure a protein folds into. The three-dimensional structures in turn specify the functions of proteins, which range from the transport of oxygen in our blood, to the conversion of energy in our muscles, and the strengthening of our hair. During evolution, the protein sequences encoded in our DNA have been optimised for these functions.
The reliable folding of proteins is a prerequisite for them to function robustly. Mis-folding can lead to protein aggregates that cause severe diseases, such as Alzheimer’s, Parkinson’s, or the variant Creutzfeldt-Jakob disease. To understand protein folding, research has long focused on metastable folding intermediates, which were thought to guide the unfolded protein chain into its folded structure. It came as a surprise about a decade ago that certain small proteins fold without any detectable intermediates. This astonishingly direct folding from the unfolded state into the folded state has been termed ‘two-state folding’. In the past few years, scientists have shown that the majority of small single-domain proteins are ‘two-state folders’, which are now a new paradigm in protein folding.
The characteristic event of two-state folding is the crossing of a barrier between the unfolded and folded state. This folding barrier is thought to consist of a large number of extremely short-lived transition state structures. Each of these structures is partially folded and will either complete the folding process, or will unfold again, with equal probability. Transition state structures are thus similar to a ball on a saddle point, which has the same probability, 0.5, of rolling to either side of the saddle. Since transition state structures are highly instable, they cannot be observed directly. To explore two-state folding, experimentalists instead create mutants of a protein. The mutants typically differ from the original protein — the wild type — in just a single amino acid. The majority of these mutants still fold into the same structure, however the mutations may slightly change the transition state barrier and, thus the folding time; that is, the time an unfolding protein chain on average needs to cross the folding barrier.
The central question is: can we reconstruct the transition state from the observed changes in the folding times? Such a reconstruction clearly requires experimental data on a large number of mutants. In the traditional interpretation, the structural information is extracted for each mutation, independent of the other mutations. If a mutation does not change the folding time, then the mutated amino acid traditionally is interpreted to be still unstructured in the transition state. In contrast, if a mutation changes the folding time, the mutated amino acid is interpreted to be partially or fully structured in the transition state, depending on the magnitude of the change.
This traditional interpretation is however often not consistent. For example, twenty single-residue mutations in the a-helix of the protein Chymotrypsin Inhibitor 2 (CI2) have very different effects on the folding time. Naïvely interpreted, these differences seem to indicate that some of the helical residues are unstructured in the transition state, while other residues, often direct neighbours, are highly structured. This naïve interpretation contradicts the fact that the folding of helices is co-operative, and can only occur if several consecutive helical turns are structured, stabilizing each other.
In a recent article in PNAS, a research team from the Max Planck Institute of Colloids and Interfaces and the University of California, San Francisco has suggested a novel interpretation of the mutational data. Instead of considering each mutation on its own, the new interpretation collectively considers all mutations within a cooperative substructure, such as a helix. In case of the a-helix of the protein CI2, this leads to a structurally consistent picture, in which the helix is fully formed in the transition state, but has not yet formed significant interactions with the ß-sheet. In the future, the Max Planck researchers hope to construct complete transition states from mutational data. An important step is to identify the cooperative subunits of proteins, which requires molecular modelling. In a similar way to how a mountain pass shows us how to cross the landscape, the transition states eventually may help us to understand how proteins navigate from the unfolded into the folded structure.
November 8, 2005
Original web page at Science Daily