For determination of tertiary structures by multidimensional NMR techniques two complementary types of information can be used: (i) torsion angles, and (ii) distances between spatially proximate nuclei.1 Accurate measurements of such structural constraints require high-resolution data with high signal to noise ratio. NMR spectral resolution is negatively affected by increasing molecular weight, which can be a notable problem in studies of macromolecules. This is due to decreased rotational diffusion rate and decreased signal intensity per peak in increasingly large molecules. It is also more likely that spectral overlap becomes a problem in large molecules. Spectral overlap is especially problematic in the ribose sugar region of RNA 1H-NMR spectra. To overcome these limitations uniform isotope labelling with 13C and 15N has recently been used.2 Preparation of the isotope labelled samples, however, is a costly procedure that does not always prevent line broadening or spectral overlap to be prohibitive as molecular weights increase.3
We have devised an alternative method to circumvent problems associated with high molecular weight. By using non-uniform deuterium labelling only the selected regions of the molecule are detected by NMR [Tetrahedron 48, 9033 (1992) [PDF 203], J. Biochem. Biophys. Methods 26, 1 (1993) [PDF 207] ]. Partially deuterated mononucleotides are incorporated in specific parts of a large DNA or RNA molecule. Structural information (coupling constants and NOE volumes) is obtained from the shorter, NMR-visible, non- deuterated or partially deuterated regions of the molecule (the 1H-NMR-window). This results narrower lines and considerably less crowded spectra. Selective deuteration also suppresses tertiary spin interactions giving more accurate NOE buildup rates and therefore more reliable proton-proton distances for use in structure refinement [Nucleic Acids Res. 22, 1404 (1994)].
The NMR-window approach can be used to study structure locally and by successively moving the window the structure of the entire molecule can be determined. The method has been successfully applied to solve structures of relatively large DNA and RNA-molecules (varying from 20-55mers [Nucleic Acids Res. 21, 5005 (1993), Nucleic Acids Res. 24, 1187 (1996) (PDF:256), Nucleic Acids Res. 24, 2022 (1996) (PDF:257), J. Biochem. Biophys. Methods 42, 153 (2000) (PDF:308)]), which were not amenable to NMR structural studies earlier. Synthetic chemistry necessary for implementation of the method on large RNAs like RNAse P and tRNAs, hammerhead and hairpin ribozymes (i.e 50-90 rsidues long) is currently under development.
We are also interested in understanding the role of bound water for
nucleic acid structure and conformational stability, and for nucleic acid-protein
interactions. To this end we have developed ROESY-NOESY methods that specifically
locates the water binding sites and gives information about the association
kinetics of water [Tetrahedron 51, 5501 (1995) (PDF:248),
Biomol. Struct. & Dynam. 16, 569 (1998)[PDF
279], Nucleosides & Nucleotides 17, 1617 (1998)[PDF
276]]. The NMR-window method can be applied to study the binding of
water to large molecules.
Nuclear Magnetic Resonance spectroscopy (NMR) has become a powerful tool to determine structures at high resolution as well as studying dynamic properties of macromolecules in solution. For determination of tertiary structure by multidimensional NMR two complementary types of information can be used: (i) torsional angles of each nucleotide or amino acid unit, and (ii) spatial proximities of these units in the DNA, RNA or protein molecule.1 With increasing molecular size two effects jointly reduce the resolution of NMR spectra. Firstly, the rotational diffusion rate decreases linearly with molecular size, giving rise to broader peaks in the spectrum. Secondly, the number of resonance lines to be assigned most likely increases with the size of the molecule, making spectral overlap increasingly problematic. In addition, the signal intensity per peak decreases as the inverse of the molecular weight. Two practical problems plague NMR spectroscopy as molecular weights increase: (1) the line broadening which is due to slow molecular tumbling rate which in turn gives shorter T1 and T2 relaxation times, and (2) the overlap of spectral lines. These problems often make it difficult to accurately determine coupling constants and to extract NOE information. To overcome these obstacles uniform isotope labelling with 13C and 15N has increasingly been used in studies of RNA and protein folding.2
However, also with such labelling, the line widths and spectral overlap can be prohibitive with increasing molecular weight.3 In addition, uniform 13C labeling of a protein increases the line width of its alpha protons because it reduces the T2 relaxation. Spectral overcrowding in 1H-NMR spectra especially affects the ribose sugar region of RNA and the a-protons of proteins. We have developed a novel approach to circumvent this problem and to enhance relaxation rates by using non-uniform deuterium isotope labelling. Partially deuterated nucleotide residues are incorporated in specific parts of DNA and RNA to make these regions invisible for NMR [Tetrahedron 48, 9033 (1992) [PDF 203], J. Biochem. Biophys. Methods 26, 1 (1993) [PDF 207] ]. Structural information (coupling constants, NOE intensities) is obtained from the NMR-visible, non-deuterated or partially-deuterated parts: the 1H-NMR-window. Protons with vicinal deuterons have about 5 times longer T1 and 10-30 times longer T2 relaxation rates, as compared to protons vicinal to protons [Nucleic Acids Res. 22, 1404 (1994) [PDF 235]]. Hence, by deuterating atoms vicinal to protons or to 13C [Magn. Reson. Chem. 36, 227 (1998) (PDF:275)], the relaxation properties of these protons can be modulated. We will implement the NMR-window in functional DNA or RNA molecules, by incorporating non-uniformly labelled (2H/1H) portions of the molecule along with a fully deuterated portions in conjunction with 13C/1H labelled windows [Tetrahedron 55, 6603 (1999) (PDF:293)]. For the identification of tertiary base-base hydrogen bonding in large RNA molecules, we will 15N label certain nucleotides [J. Labelled. Cpd. Radiopharm. 44, 763 (2001) (PDF:328)] or whole domains and conduct 15N NOESY-HMQC NMR experiments.
Similarly, specific non-uniform incorporation of alpha deuterated amino acids in a polypeptide, creating an NMR-window, will also increase, by 10-fold or more, the T2 relaxation rate of the vicinal 13C. Selective incorporation of stereospecifically a-2H- a-13C-15N-labelled amino acid residues in a small polypeptide or in a protein creates specific "NMR-windows" also in these molecules. This allows spectral editing in multidimensional NMR experiments, providing local structural and dynamic information from the window region. In such experiments the 2H frequency is decoupled to enhance sensitivity in the 13C and 15N domains. The "NMR-window" approach also reduces the multistep cross-relaxations between protons and vicinal carbons. The intensities of cross- peaks in NOE experiments will, in our approach, only depend on direct field interactions without spin-diffusion, which increases the NOE cross-peak intensities about 4-10 fold. Selective deuteration suppresses tertiary spin interactions allowing more accurate measurements of NOE buildup rates and therefore more reliable proton-proton distances, which in turn should allow more reliable structure determination.
By successively moving the window over small overlapping regions until the whole molecule has been covered, large DNA, RNA and protein molecules can be studied. The method has already produced a host of new results for small DNA and RNA-molecules. The necessary engineering techniques to implement the same idea for large RNAs like RNAse P, tRNAs and proteins are now being worked out (See below).
1. Wüthrich, K. NMR of Proteins and Nucleic Acids Wiley, New York, 1986.
2. (a) Nikonowicz, E. P.; Sirr, A.; Legault,
P.; Jucker, F. M.; Baer, L. M.; Pardi, A. Nucleic Acids Res. 1992,
(b) Batey, R. T.; Battiste, J. L.; Williamson, J. R. Methods Enzymol. 1995, 261, 300.
(c) Wyatt, J. R.; Chastain, M.; Puglisi, J. D. BioTechniques 1991, 11, 764.
(d) Zimmer, D. P.; Crothers, D. M. Proc. Natl. Acad. Sci. USA 1995, 92, 3091.
(e) Louis, J. M.; Martin, R. G.; Clore, G. M.; Gronenborn, A. M. J. Biol. Chem. 1998, 273, 2374.
(f) Masse, J. E.; Bortmann, P.; Dieckmann, T.; Feigon, J. Nucleic Acids Res. 1998, 26, 2618.
3. Sattler, M.; Fesik, W. S. Structure 1996,