Nature Structural Biology
8, 923 - 925 (2001)
doi:10.1038/nsb1101-923
The way to NMR structures of proteinsKurt WüthrichKurt Wüthrich is Professor of Biophysics at the Institute of Molecular Biology and Biophysics, ETH Zürich, CH-8093 Zürich, Switzerland, Fax: 41 1-633-1151, and Cecil H. and Ida M. Green Visiting Professor of Structural Biology at The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, USA, Fax: 1 858-784-8014. In 1998 Kurt Wüthrich was awarded the Kyoto Prize in Advanced Technology for having "developed a method of determining the conformations of proteins, nucleic acids and other biomacromolecules in solutions or biomembranes, where they exhibit their function"1.
Wüthrich has used nuclear magnetic resonance (NMR) techniques to study proteins and nucleic acids since 1967. In a series of four papers his group outlined a framework for NMR structure determination of proteins in 1982, and in 1984 the first de novo structure of a globular protein in solution was determined. The Wüthrich group went on to solve more than 60 protein structures in solution, including the Antennapedia homeodomain, the cyclophillin A−cyclosporin A complex, and the human and bovine prion proteins.
What follows is a personal recollection by Kurt Wüthrich of how he and his associates arrived at the first view of a protein structure through the NMR eye.
In the 1950s, magnetic resonance spectroscopy evolved into a useful tool in chemistry. During the period of 1962−1967, my graduate and postdoctoral research, with Professor Silvio Fallab at the University of Basel and Professor Robert E. Connick at the University of California, Berkeley, focused on the use of electron paramagnetic resonance (EPR) and nuclear magnetic resonance (NMR) spin relaxation measurements to study metal complexes in solution. With this background, I joined the Biophysics Department of Dr. Robert G. Shulman at Bell Telephone Laboratories in Murray Hill, New Jersey, where a superconducting high resolution 1H NMR spectrometer operating at 220 MHz was available for 'research on protein structure and function'. At that time I was aware of exactly 10 papers on NMR observations of proteins and nucleic acids, which had all been published during the period of 1957−19652. Prominent figures in the small community of spectroscopists that ventured into direct NMR observation of biological macromolecules were William D. Phillips3, Oleg Jardetzky4 and Robert G. Shulman5. Based on the observation of empirical correlations between protein unfolding and NMR spectra2,
3,
4, there was much enthusiasm about the future of NMR for de novo protein structure determination. Nonetheless, true to my background, I initially focused on the metal ion coordination in the active sites of hemoproteins and on the electronic structure of the heme groups6.
At the time, Swiss scientists who landed a job at the famous Bell Telephone Laboratories were automatically considered prime candidates for academic positions back home. In 1969, I moved to the Eidgenössiche Technische Hochschule (ETH) in Zürich, where my startup package included an EPR and three NMR spectrometers — all the instrumentation that had been available to me at Bell Telephone Laboratories. I assembled a small research group, and, with time, I was promoted to Professor of Biophysics, which is also my present position at ETH. During the first years at Zürich, my research continued to focus primarily on the metal ions in the active centers of hemoproteins2,
3,
4,
5,
6, and I developed a mild infatuation with polypeptide chains only in connection with the discovery of aromatic ring flipping2. My primary research interest changed in 1975, when I took some time to write a monograph on the early years of biomacromolecular NMR2. These reflections on the state of the field turned out to have been the starting point for our subsequent work on de novo protein structure determination by NMR7.
Four principal elements are combined in the NMR method for protein structure determination8,
9: (i) the nuclear Overhauser effect (NOE) as an experimentally accessible NMR parameter in proteins that can yield the information needed for de novo global fold determination of a polymer chain; (ii) sequence-specific assignment of the many hundred to several thousand NMR peaks from a protein; (iii) computational tools for the structural interpretation of the NMR data and the evaluation of the resulting molecular structures; and (iv) multidimensional NMR techniques for efficient data collection. During the period 1976−1980, my research group at the ETH Zurich had grown to more than 20 scientists, all of whom made great contributions toward the structure determination method. In particular, I worked with Regula M. Keller, Sidney L. Gordon and Gerhard Wagner on developing techniques to measure NOEs for the collection of conformational constraints in proteins, and with Martin Billeter, Werner Braun and Gerhard Wagner on the sequential resonance assignment strategy and algorithms for structure calculation from NMR data. This technology passed its initial tests when we obtained partial structure determinations of the bovine pancreatic trypsin inhibitor (BPTI), cytochrome b5 and the polypeptide hormone glucagon based on data collection with one-dimensional (1D) NMR experiments.
In a parallel project from 1976 to 1980, Richard R. Ernst (Nobel Prize in Chemistry, 1991), who also worked at the ETH Zürich, and I joined forces to develop two-dimensional (2D) NMR techniques for applications with biological macromolecules. Kuniaki Nagayama used 2D correlation spectroscopy for amino acid spin system identification in a protein, and Anil Kumar recorded the first 2D NOE spectra during the Christmas break 1979, when he was allotted two weeks of the precious measuring time on our highest-field spectrometer operating at a proton NMR frequency of 360 MHz10. By 1981 we routinely applied a group of four homonuclear 2D 1H NMR experiments, known under the acronyms COSY, SECSY, FOCSY and NOESY9, in the protein structure determination project. This resulted in complete resonance assignments of several small proteins in 1982 and 198311, and in the first de novo atomic resolution NMR structure determination of a globular protein, the bull seminal protease inhibitor (BUSI)12, by Timothy F. Havel and Michael P. Williamson in 1984.
The completion of the first protein NMR structure brought new, unexpected challenges. When I presented the structure of BUSI (Fig. 1a)12 in some lectures in the spring of 1984, the reaction was one of disbelief, and because of the close coincidence (Fig. 1b) with results from an independent crystallographic study of the homologous protein PSTI (porcine pancreatic secretory trypsin inhibitor)13 it was suggested that our structure must have been modeled after this crystal structure. In a discussion following a seminar in Munich on May 14, 1984, Robert Huber (Nobel Prize in Chemistry, 1988) proposed that we settle the matter by independently solving a new protein structure by X-ray crystallography and by NMR. For this purpose, each one of us received an ample supply of the  -amylase inhibitor tendamistat from scientists at the Hoechst company. Virtually identical three-dimensional structures of tendamistat were obtained in our laboratory by NMR in solution and in Robert Huber's laboratory by X-ray diffraction in single crystals.
 | | |
The refined tendamistat structure was published in Journal of Molecular Biology as a 50-page report14, and the addendum to that paper clearly illustrated the impact of structure determination by NMR. I quote: "Editor's Note: We have taken the step of publishing this paper with full supporting data since it is the first high resolution structure worked out in detail by 2D NMR. We therefore think that in this one instance everything should be published in full, but it does not set a precedent, since it is hoped that in the future, such supporting data can be deposited in a data bank, as is the practice in X-ray protein crystallography". Considering that over 2,000 NMR structures have since been deposited in the Protein Data Bank, the Editor should be commended for his vision.
At that time his kind comments were comforting in the context of our structure determinations of mammalian metallothioneins, which are a class of small, metal-rich proteins that we studied in collaboration with Jeremias H.R. Kägi at the University of Zürich. In June 1985 I presented the structure of rabbit metallothionein at Yale University, where I learned about a manuscript accepted for publication in Proc. Nat. Acad. Sci. USA, which described a completely different metallothionein 'NMR structure', and at the University of Pittsburgh, where I was confronted with a rat metallothionein crystal structure that was again very different from our NMR structure. In both instances the structural differences were very clearcut, since they involved different polypeptide folds as well as different coordinating ligands to the metals. Metallothionein had been a tough challenge for all of us involved15, and my initial reaction was to spend two nights on the phone in my US motel room rechecking step by step the sequential resonance assignments with Gerhard Wagner in Zürich. All the assignments were, of course, correct, and I am afraid that Gerhard still bears a grudge against me for ever having doubted his spectral analysis. The crystal structure, which included erroneous chain tracing and identification of 11 out of a total of 20 metal-coordinating amino acid residues, eventually appeared as a feature article in Science, whereas Nature rejected our NMR structure paper. In 1992, the crystal structure of rat metallothionein was redetermined, a correction of the first structure was published, and the correct crystal structure was found to be identical with the NMR structures of the rabbit, rat and human metallothioneins that we had solved from 1985 to 199016.
Over the years a variety of applications of the NMR structure determination method have been pursued in my laboratory. The following three examples may convey some of the excitement that was thus generated in our professional life and further indicates the wide range of NMR applications in structural biology. Studies on the structural foundations of transcriptional regulation in higher organisms pursued in collaboration with Walter J. Gehring at the Biocenter of the University of Basel, Switzerland, yielded the NMR structure of the Antennapedia homeodomain17, and provided entirely novel insights into the role of hydration water in protein−DNA recognition18. An NMR structure determination of the human cyclophilin A−cyclosporin A complex was obtained in collaboration with two of my former graduate students, Hans Senn and Hans Widmer, who had subsequently joined the Sandoz company in Basel, Switzerland. This structure determination not only introduced me to the field of immune suppression but also had an immediate practical impact on cyclosporin research, since the structure of the bound drug molecule was found to be turned inside-out when compared with the structure of free cyclosporin A19. Barely 10 days after the bovine spongiform encephalopathy (BSE) crisis in Great Britain had broken into the open in March 1996, we completed the NMR structure determination of the murine prion protein20 in a collaboration with Rudi Glockshuber, who had joined our institute at the ETH Zürich as an Assistant Professor in 1994. The observation of a long flexible tail in prion proteins21 presents on the one hand a striking illustration of the unique power of NMR to characterize partially structured polypeptide chains in physiological milieus, and on the other hand indicates novel possible avenues for the transition of the benign cellular form of prion proteins to the disease-related scrapie form. With the introduction of TROSY (transverse relaxation-optimized spectroscopy)22, the molecular weight limit for solution NMR spectroscopy has extended to  500 kDa, and we may soon be able to obtain information on the structure of the disease-related, aggregated form of the prion protein.
REFERENCES
- Kyoto Prizes and Inamori Grants 1998, 13 (The Inamori Foundation, Kyoto; 1999).
- Wüthrich, K. NMR in Biological Research: Peptides and Proteins (North Holland, Amsterdam; 1976).
- McDonald, C.C. & Phillips, W.D. J. Am. Chem. Soc. 89, 6332−6341 (1967). | Article | PubMed | ISI | ChemPort |
- Jardetzky, O. & Roberts, G.C.K. NMR in Molecular Biology (Academic Press, New York; 1981).
- Shulman. R.G. et al. Science 165, 251−257 (1969). | PubMed | ISI | ChemPort |
- Wüthrich, K. Structure and Bonding 8, 53−121 (1970). | ISI |
- Wüthrich, K. NMR in Structural Biology — A Collection of Papers by Kurt Wüthrich (World Scientific, Singapore; 1995).
- Wüthrich, K., Wider, G., Wagner, G. & Braun, W. J. Mol. Biol. 155, 311−319 (1982). | PubMed | ISI |
- Wüthrich, K. NMR of Proteins and Nucleic Acids (Wiley, New York; 1986).
- Anil-Kumar, Ernst, R.R. & Wüthrich, K. Biochem. Biophys. Res. Comm. 95, 1−6 (1980). | PubMed |
- Wagner, G. & Wüthrich, K. J. Mol. Biol. 155, 347−366 (1982). | Article | PubMed | ISI | ChemPort |
- Williamson, M.P., Havel, T.F. & Wüthrich, K. J. Mol. Biol. 182, 295−315 (1985). | Article | PubMed | ISI | ChemPort |
- Bolognesi, M. et al. J. Mol. Biol. 162, 839−868 (1992). | Article |
- Kline, A.D., Braun, W. & Wüthrich, K. J. Mol. Biol. 204, 675−724 (1988). | Article | PubMed | ISI | ChemPort |
- Braun, W. et al. J. Mol. Biol. 187, 125−129 (1986). | Article | PubMed | ISI | ChemPort |
- Braun, W. et al. Proc. Natl. Acad. Sci. USA 89, 10124−10128 (1992). | PubMed | ChemPort |
- Qian, Y.Q. et al. Cell 59, 573−580 (1989). | Article | PubMed | ISI | ChemPort |
- Billeter, M., Güntert, P., Luginbühl, P. & Wüthrich, K. Cell 85, 1057−1065 (1996). | Article | PubMed | ISI | ChemPort |
- Wüthrich, K. et al. Science 254, 953−954 (1991). | PubMed | ISI |
- Riek, R. et al. Nature 382, 180−182 (1996). | Article | PubMed | ISI | ChemPort |
- Riek, R., Hornemann, S., Wider, G., Glockshuber R. & Wüthrich, K. FEBS Lett. 413, 277−281 (1997). | Article | PubMed | ISI |
- Pervushin, K., Riek, R., Wider, G. & Wüthrich, K. Proc. Natl. Acad. Sci. USA 94, 12366−12371 (1997). | Article | PubMed | ChemPort |
|
