Macromolecular Structure Determination: Comparison of X‐ray Crystallography and NMR Spectroscopy

VV Krishnan, University of California Davis, Sacramento, California, USA
B Rupp, k‐k Hofkristallamt, San Marcos, California, USA

Published online: June 2012

DOI: 10.1002/9780470015902.a0002716.pub2

X-ray crystallography and nuclear magnetic resonance (NMR) spectroscopy are the most powerful and predominant techniques used to experimentally determine the three-dimensional structures of biological macromolecules at near atomic resolution. X-ray diffraction (XRD) studies require a crystalisable protein, whereas NMR is suitable for macromolecules in solution. XRD has no size limitations and provides the most precise atomic detail, whereas information about the dynamics of the molecule may be limited. NMR excels in cases where no protein crystals can be obtained and it provides solution state dynamics, but in turn delivers lower resolution structures and is in general limited to molecular weights below approximately 50 kDa. The two techniques can deliver complementary information. Approximately 90% of the experimentally determined macromolecular structures deposited in Protein Data Bank are crystal structures, with NMR dominating the <10 kDa molecular weight range.

Key Concepts:

  • XRD and NMR are complimentary structure determination methods.
  • XRD has no size limitations and provides the most precise atomic detail, whereas information about the dynamics of the molecule may be limited.
  • NMR excels in cases where no protein crystals can be obtained and it provides solution state dynamics, but in turn delivers less detail and in general practice is limited to molecular weights below approximately 50 kDa.

Keywords: crystallography; NMR; X-ray; macromolecular structure; diffraction

Figure 1. The principle of X-ray structure determination. A crystal mounted on a goniostat with at least one rotatable axis is exposed to a finely collimated, intense X-ray beam in the 5–20 keV energy range (approximately 2.3–0.6 Å wavelength). Individual diffraction images are recorded during small rotation increments of the crystal and combined into a diffraction data set. Unfortunately, the diffraction images are not direct images of the molecule. The basic mathematical tool of back-transformation from reciprocal diffraction space into direct space is the Fourier transform (FT), which together with separately acquired phases for each diffraction spot allows synthesising or reconstructing the electron density (blue grid) of the molecules self-assembled into the diffracting crystal. An atomic model of the structure is then built into the three-dimensional electron density. From Biomolecular Crystallography: Principles, Practice, and Application to Structural Biology by Bernhard Rupp. Reproduced by permission of Garland Science/Taylor & Francis LLC © 2011.
Figure 2. Experimental basis of NMR structure determination. Biomolecules in solution at close to physiological conditions are inserted into a magnet. The radio frequency circuit detects the time domain signal corresponding response of the nuclear spins to resonance. This analogue time domain signal detected by the circuit is amplified and digitised prior to Fourier transformation into the spectral domain. A combination of 2D and 3D experiments are generally collected, processed and analysed to obtain NMR restraint parameter that are sensitive to determine both local structural relations and events (through chemical shifts and coupling constants) as well as the global fold (via NOEs) of a protein. Three-dimensional structural models generated by NMR methods also carry additional information on residue-specific dynamic motion.
Figure 3. Flowchart describing the major steps in NMR based structure determination. Four principal elements are combined in the NMR method for protein structure determination: (1) multidimensional NMR experiments that provide the data for all the following steps; (2) sequence-specific resonance assignment – matching each proton in the protein to respective peaks in the spectra; (3) the Nuclear Overhauser Effect (NOE) data – providing inter-proton distances to identify the global fold of the protein; (4) computational tools such as distance geometry (restraint) based approach for the structural interpretation of the NMR data and the evaluation of the resulting molecular structures and Each of these elements is critically important in obtaining a good quality NMR structure.
Figure 4. Sequential assignment strategy. The process starts from first identifying a C/C resonance in the CBCANH spectrum of a residue i, as these resonance frequencies vary depending on the protein fold and amino acid type. Comparing the chemical shifts side-by-side on strip plots (right) a new NH group is identified in the CBCA(CO)NH spectrum corresponding to the (i+1)th residue. The CBCANH spectrum is again used to identify the C/C resonances of this (i+1)th residue and the process continues until all the resonances are identified. In a CBCANNH spectrum the NH group correlates strongly with its own C and C, but the correlation is weak with the preceding residue, whereas the CBCA(CO)NH only correlates the NH group to the preceding C and C chemical shifts. This combination of information is used to differentiate the intra- or inter-residue correlations. In the strip plots on the right, the C atoms are shown by blue, C atoms in green with the intra-residue peaks repeated in red.
Figure 5. Data quality determines structural detail and accuracy. The qualitative relation between the extent of X-ray diffraction, the resulting amount of available diffraction data, and the quality and detail of the electron density reconstruction and protein structure model are evident from this figure: The crystals are labelled with the nominal resolution in Å as determined by the highest diffraction angle at which X-ray reflections are observed. Above the crystals is a sketch of the corresponding diffraction pattern, which contains significantly more data at higher resolution. As a consequence, both the reconstruction of the electron density (blue grid) and the resulting structure model (stick model) are much more detailed and accurate. The non-SI unit Å (10–8 cm or 0.1 nm=10–10 m) is generally used in the crystallographic literature, simply because it is of the same order of magnitude as atomic radii (approximately 0.77 Å for carbon) or bond lengths (approximately 1.54 Å for the C–C single bond). From Biomolecular Crystallography: Principles, Practice, and Application to Structural Biology by Bernhard Rupp. Reproduced by permission of Garland Science/Taylor & Francis LLC © 2011.
Figure 6. HSQC spectrum of ligand-free and ligand-bound protein. The 2-d 1H–15N heteronuclear single-quantum coherence (HSQC) NMR spectrum of bacterial methionine aminopeptidase (bMAP) with (right) and without (left) a tightly bound novel inhibitor (Evdokimov et al., 2007). Note the drastic improvement in the discrimination of the spectrum for the bMAP-ligand complex compared to the apo-protein. The crystals of the bMAP-ligand complex diffracted to 0.9 Å resolution. Image courtesy of Artem Evdokimov, Procter & Gamble Pharmaceuticals, Mason, OH. From Biomolecular Crystallography: Principles, Practice, and Application to Structural Biology by Bernhard Rupp. Reproduced by permission of Garland Science/Taylor & Francis LLC © 2011.
close
 References
    Berman H (2008) The protein data bank: a historical perspective. Acta Crystallographica Section A 64(1): 88–95.
    Bloch F, Hansen WW and Packard M (1946) Nuclear induction. Physical Review 69(3–4): 127.
    Brunger A T (2007) Version 1.2 of the Crystallography and NMR system. Nature Protocols 2(11): 2728–2733.
    Dickinson WC (1950) Dependence of the f-19 nuclear resonance position on chemical compound. Physical Review 77(5): 736–737.
    Ernst RR and Anderson WA (1966) Application of Fourier transform spectroscopy to magnetic resonance. Review of Scientific Instruments 37(1): 93–102
    Evdokimov AG, Pokross M, Walter RL et al. (2007) Serendipitous discovery of novel bacterial methionine aminopeptidase inhibitors. Proteins 66(3): 538–546.
    Fiaux J, Bertelsen EB, Horwich AL and Wüthrich K, (2002) NMR analysis of a 900 K GroEL–GroES complex. Nature 418: 207–211.
    book Grant DM and Harris RK (1996) Encyclopedia of Nuclear Magnetic Resonance. Chichester; New York: John Wiley.
    Griswold IJ and Dahlquist FW (2002) Bigger is better: megadalton protein NMR in solution. Nature Structural & Molecular Biology 9(8): 567–568.
    Gronwald W, Brunner K, Kirchhofer R et al. (2007) AUREMOL-RFAC-3D, combination of R-factors and their use for automated quality assessment of protein solution structures. Journal of Biomolecular NMR 37(1): 15–30.
    Grzesiek S and Bax A (1992) Correlating backbone amide and side chain resonances in larger proteins by multiple relayed triple resonance NMR. Journal of American Chemical Society 114(16): 6291–6293.
    Gutowsky HS and McCall DW (1951) Nuclear magnetic resonance fine structure in liquids. Physical Review 82(5): 748–749.
    Hahn EL and Maxwell DE (1951) Chemical shift and field independent frequency modulation of the spin echo envelope. Physical Review 84(6): 1246–1247.
    Kendrew JC, Bodo G, Dintzis HM et al. (1958) A three-dimensional model of the myoglobin molecule obtained by X-ray analysis. Nature 181: 662–666.
    Kleywegt GJ, Harris MR, Zou J-Y et al. (2004) The Uppsala electron-density server. Acta Crystallographica D60: 2240–2249.
    Knight WD (1949) Nuclear magnetic resonance shift in metals. Physical Review 76(8): 1259–1260.
    Kobayashi N, Iwahara J, Koshiba S et al. (2007) KUJIRA, a package of integrated modules for systematic and interactive analysis of NMR data directed to high-throughput NMR structure studies. Journal of Biomolecular NMR 39(1): 31–52.
    Kovacs H, Moskau D and Spraul M (2005) Cryogenically cooled probes – a leap in NMR technology. Progress in Nuclear Magnetic Resonance Spectroscopy 46(2–3): 131–155.
    Kumar A, Ernst RR and Wüthrich K (1980) A two-dimensional nuclear overhauser enhancement (2D NOE) experiment for the elucidation of complete proton–proton cross-relaxation networks in biological macromolecules. Biochemical and Biophysical Research Communications 95(1): 1–6.
    Merritt EA (2012) To B or not to B: a question of resolution? Acta Crystallographica D68 in press.
    Murshudov GN, Skubak P, Lebedev AA et al. (2011) REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallographica Section D D67(Pt 4): 355–367.
    Overhauser AW (1953a) Polarization of nuclei in metals. Physical Review 92(2): 411–415.
    Overhauser AW (1953b) Polarization of nuclei in metals. Physical Review 91(2): 476.
    Pound RV and Purcell EM (1946) Measurement of magnetic resonance absorption by nuclear moments in a solid. Physical Review 69(11–1): 681.
    Proctor WG and Yu FC (1950) The dependence of a nuclear magnetic resonance frequency upon chemical compound. Physical Review 77(5): 717.
    Proctor WG and Yu FC (1951) On the nuclear magnetic moments of several stable isotopes. Physical Review 81(1): 20–30.
    Purcell EM, Pound RV and Bloembergen N (1946) Nuclear magnetic resonance absorption in hydrogen gas. Physical Review 70(11-1): 986–987.
    Rabi II, Zacharias JR, Millman S et al. (1938) A new method of measuring nuclear magnetic moment. Physical Review 53(4): 318.
    Rieping W, Habeck M, Bardiaux B et al. (2007) ARIA2: automated NOE assignment and data integration in NMR structure calculation. Bioinformatics 23(3): 381–382.
    book Rupp B (2009) Biomolecular Crystallography: Principles, Practice, and Application to Structural Biology. New York: Garland Science.
    Saunders M, Wishnia A and Kirkwood JG (1957) The nuclear magnetic resonance spectrum of ribonuclease. Journal of the American Chemical Society 79(12): 3289–3290.
    Schroder GF, Levitt M and Brunger AT (2010) Super-resolution biomolecular crystallography with low-resolution data. Nature 464(7292): 1218–1222.
    Tickle I (2007) Experimental determination of optimal root-mean-square deviations of macromolecular bond lengths and angles from their restrained ideal values. Acta Crystallographica D63: 1274–1281.
    Tugarinov V, Choy WY, Orekhov VY et al. (2005) Solution NMR-derived global fold of a monomeric 82-kDa enzyme. Proceedings of the National Academy of Sciences of the USA 102(3): 622–627.
    Williamson MP, Havel TF and Wüthrich K (1985) Solution conformation of proteinase inhibitor-IIA from bull seminal plasma byH-1 nuclear magnetic-resonance and distance geometry. Journal of Molecular Biology 182(2): 295–315.
 Further Reading
    book Cavanagh J, Fairbrother WJ, Palmer AG et al. (2007) Protein NMR Spectroscopy, Second Edition: Principles and Practice. New York: Academic Press.
    book Doucleff M, Hatcher-Skeers M and Crane NJ (2011) Pocket Guide to Biomolecular NMR. New York: Springer.
    book Rhodes G (2006) Crystallography Made Crystal Clear. London, UK: Academic Press.

Related Sub-Topics:

X-Ray Crystallography | Nuclear Magnetic Resonance
Contact Editor close
Submit a note to the editor about this article by filling in the form below.

* Required Field

Article Title: Macromolecular Structure Determination: Comparison of X‐ray Crystallography and NMR Spectroscopy
 
How to Cite close
Krishnan, VV, and Rupp, B(Jun 2012) Macromolecular Structure Determination: Comparison of X‐ray Crystallography and NMR Spectroscopy. In: eLS. John Wiley & Sons Ltd, Chichester. http://www.els.net [doi: 10.1002/9780470015902.a0002716.pub2]