Volume 85 Issue 23 | p. 10 | News of The Week
Issue Date: June 4, 2007

Fast Route To Structures

Technique defines protein structures from NMR chemical shifts
Department: Science & Technology
The new technique predicts likely structures of protein fragments on the basis of chemical shift data and assembles the fragments into an overall protein structure, which is then refined with chemical shift data and molecular force fields.
8523notw6_fragmentcxd
 
The new technique predicts likely structures of protein fragments on the basis of chemical shift data and assembles the fragments into an overall protein structure, which is then refined with chemical shift data and molecular force fields.
Chemical-shift-based structure (blue) of the protein ubiquitin is a close match with its X-ray crystal structure (red).
Credit: © PNAS 2007 (both)
8523notw6_ubqcxd
 
Chemical-shift-based structure (blue) of the protein ubiquitin is a close match with its X-ray crystal structure (red).
Credit: © PNAS 2007 (both)

IN A DEVELOPMENT that promises to ease structural analysis of proteins, British researchers have found a way to determine protein structures by using only basic and easy-to-obtain data complemented by theoretical calculations.

The new approach, based on nuclear magnetic resonance spectroscopy (NMR) chemical shifts, is faster and simpler than conventional NMR methods.

Typically, NMR protein structures are based on interatomic distances, but these are difficult to measure. On the other hand, chemical shifts are the most readily obtained and accurate NMR parameters. They provide information about the molecular environment of atoms. They're used to determine structures of small molecules, and they're usually the first thing chemistry undergrads learn about NMR.

Researchers have wanted to use chemical shifts to determine protein structures because that sidesteps the need to make time-consuming nuclear Overhauser effect measurements. NOEs, which are pairwise distances between specific atoms, are currently the primary basis for most NMR protein structures.

Now, work carried out by structural biologist Michele Vendruscolo, chemistry professor Christopher M. Dobson, and coworkers at the University of Cambridge makes it possible to determine protein structures without NOEs (Proc. Natl. Acad. Sci. USA 2007, 104, 9615).

Vendruscolo, Dobson, and coworkers have devised a technique called CHESHIRE (short for "chemical shift restraints") that makes it possible to use chemical shifts instead. In this technique, a protein sequence is divided into fragments whose likely structures are predicted on the basis of experimental chemical shift data. The fragments' structures are assembled into a complete protein structure that is refined with chemical shift and molecular force field information and evaluated for reliability.

The researchers used the approach to accurately define the high-resolution (2-Å resolution or better) structures of 11 proteins of known structure. The proteins analyzed, including ubiquitin, contain up to 123 residues and are representative of major structural classes.

The technique "will enable the determination of a vast range of protein conformations for which chemical shifts are essentially the only NMR observables that are possible to obtain," Vendruscolo says. For example, he believes it could be used to study transient or excited states of proteins, for which chemical shifts are often the only type of NMR data available. In addition, "we are exploring the applicability of the method to membrane proteins, large biomolecular complexes, and amyloid fibrils," he says.

"It's a tremendous success story," says Lewis E. Kay, a specialist in NMR structure determination at the University of Toronto. "Ultimately, you'd like to be able to determine structures by NMR in the simplest sort of way, and NMR chemical shifts are the easiest thing to measure. So if you can come up with an algorithm that effectively allows you to predict structure accurately from chemical shifts, you're well ahead of the ball game. People have been trying to do this for a long time, but I think this paper is the first where there is a demonstrated success, not on one protein but on 11."

NMR specialist Ad Bax of the National Institute of Diabetes & Digestive & Kidney Diseases, Bethesda, Md., says, "The study represents a major step forward, and the results shown are quite impressive. Even though automated NOE analysis is also gaining traction, the fact that chemical shifts now suffice to generate reasonable quality structures can greatly expedite the structure determination process."

"It has become clear that structures of small proteins can almost be predicted just from their sequence," says NMR structural biologist David A. Case of Scripps Research Institute. "This has led to the hope that a modest amount of additional experimental information might be sufficient to find reliable structures." The new study, he adds, "is certainly a nice milestone along the way. It's an impressive achievement—probably harder to actually do in practice than an outsider might guess—and there is every reason to think that the present results are just the beginning."

 
Chemical & Engineering News
ISSN 0009-2347
Copyright © American Chemical Society

Leave A Comment

*Required to comment