Ultrafast multidimensionnal NMR

Multidimensional NMR in its conventional form requires extended experimental durations, which make it unsuitable to characterise fast processes and study large numbers of samples. We develop ultrafast NMR pulse sequences, which exploit a spatial encoding of the NMR interactions to record multidimensional spectra in less than one second. We have, for example, developed an ultrafast version of multiple-quantum experiments, as well as pulse sequences for ultrafast diffusion-ordered NMR spectroscopy (DOSY). Both methods are particularly well suited to analyse mixtures. We also develop a simulation framework to analyse ultrafast NMR experiments.


  • L. Guduff, I. Kuprov, C. van Heijenoort and J.-N. Dumez, Spatially encoded 2D and 3D diffusion- ordered NMR spectroscopy, Chem. Commun., 53, 701 (2017). http://dx.doi.org/10.1039/C6CC09028A
  • A. Le Guennec, P. Giraudeau, S. Caldarelli and J.-N. Dumez, Ultrafast double-quantum NMR spectroscopy, Chem. Commun. 51, 354 (2015). http://dx.doi.org/10.1039/c4cc07232d
  • L. Rouger, B. Gouilleux, M. Pourchet-Gellez, J.-N. Dumez and P. Giraudeau, Ultrafast double-quantum NMR spectroscopy with optimized sensitivity for the analysis of mixtures, Analyst 141, 1686 (2015) http://dx.doi.org/10.1039/c6an00089d
  • L. Guduff, A.J. Allami, C. van Heijenoort, J.-N. Dumez and I. Kuprov, Efficient simulation of ultrafast magnetic resonance experiments, Phys. Chem. Chem. Phys. 19, 17577 (2017). http://dx.doi.org/10.1039/C7CP03074F
  • B. Gouilleux, L. Rouger, B. Charrier, I. Kuprov, S. Akoka, J.-N. Dumez and P. Giraudeau, Understanding J-modulation during spatial encoding for sensitivity-optimized ultrafast NMR, ChemPhysChem, 16, 3093 (2015). http://dx.doi.org/10.1002/cphc.201500514

Hyperpolarisation: DNP and para-hydrogen

Hyperpolarisation methods provide a significant increase of NMR sensitivity. We exploit the potential of this approach to analyse mixtures, through collaboration with two groups that have implemented hyperpolarisation methods.

Para-hydrogen does provide intense polarisation in a very short time and is applicable to a range of small molecules. In collaboration with Patrick Berthault and Gaspard Huber, at CEA Saclay, we have shown that 2D spectra can be obtained in a single scan for mixtures with sub-millimolar concentrations, using para-hydrogen-enhanced ultrafast NMR.

Dynamic Nuclear Polarisation (DNP) is a very general hyperpolarisation method. In collaboration with Geoffrey Bodenhausen, Sami Jannin and Patrick Giraudeau, at EPFL, we have shown that 1D and ultrafast 2D spectra can be obtained quickly from extracts of biological samples (cancer cells and fruit), using DNP. Together with D. Abergel we have also shown that 2D DOSY data can be collected in a single scan from substrates hyperpolarised with D-DNP.


  • L. Guduff, D. Kurzbach, C. van Heijenoort, D. Abergel and J.-N. Dumez, Single-scan 13C diffusion-ordered NMR spectroscopy of DNP-hyperpolarised substrates, Chem. Eur. J., in press. http://dx.doi.org/10.1002/chem.201703300
  • J.-N. Dumez, J. Milani, B. Vuichoud, A. Bornet, J. Lalande-Martin, I. Tea, M. Yon, M. Maucourt, C. Deborde, A. Moing, L. Frydman, G. Bodenhausen, S. Jannin and P. Giraudeau, Hyperpolarized NMR of plant and cancer cell extracts at natural abundance, Analyst, 140, 5860 (2015). http://dx.doi.org/10.1039/c5an01203a
  • V. Daniele, F.-X. Legrand, P. Berthault, J.-N. Dumez and G. Huber, Single-Scan Multidimensional NMR Analysis of Mixtures at Sub-Millimolar Concentrations by using SABRE Hyperpolarization, ChemPhysChem, in press. http://dx.doi.org/10.1002/cphc.201500535

Long-lived states and their applications

Longitudinal nuclear relaxation sets a stringent limit on the range of information that can be obtained from magnetic resonance experiments. Long-lived nuclear spin states (LLS) provide a possibility to extend the timescale over which information can be encoded in NMR, and particularly in hyperpolarised experiments. We exploit the properties of long-lived-states in multi-spin systems, which may extend the scope of application of the LLS concept and of hyperpolarisation. We have notably shown that long-lived states exist in methyl group and that they may be accessed using dynamic nuclear polarisation.

Publications :

  • S.S. Roy, J.-N. Dumez, G. Stevanato, B. Meier, J.T. Hill-Cousins, R.C.D. Brown, G. Pileio and M.H. Levitt, Enhancement of quantum rotor NMR signals by frequency-selective pulses, J. Magn Reson. 250, 25 (2015). http://dx.doi.org/10.1016/j.jmr.2014.11.004
  • J.-N. Dumez, P. Hakansson, S. Mamone, B. Meier, G. Stevanato, J.T. Hill-Cousins, S.S. Roy R.C.D. Brown, G. Pileio and M.H. Levitt, Theory of long-lived nuclear spin states in methyl groups and quantum-rotor-induced polarization, J. Chem. Phys. 142, 044506 (2015). http://dx.doi.org/10.1063/1.4906273
  • D. Mammoli, B. Vuichoud, A. Bornet, J. Milani J.-N. Dumez, S. Jannin and G. Bodenhausen, Hyperpolarized para-ethanol, J. Phys. Chem. B. 119, 4048 (2015). http://dx.doi.org/10.1021/jp512128c
  • S. Elliott, L.J. Brown, J.-N. Dumez, M.H. Levitt, Long-lived nuclear spin states in monodeuterated methyl groups, Phys. Chem. Chem. Phys., 18, 17965 (2016). http://dx.doi.org/10.1039/c6cp03619h
  • J.-N. Dumez, B. Vuichoud, D. Mammoli, A. Bornet, A.C. Pinon, G. Stevanato, B. Meier, G. Bodenhausen, S. Jannin and M.H. Levitt, Dynamic nuclear polarisation of long-lived nuclear spin states in methyl groups, J. Phys. Chem. Lett., 8, 3549 (2017). http://dx.doi.org/10.1021/acs.jpclett.7b01512

Insect cell production of labeled proteins for NMR applications

The studies of protein structural and functional properties by NMR rely on the possibility to enrich them in stable isotopes, either in 15N, 13C and sometime 2H for intermediate size proteins (up to several tens of kDa) or in 1H, 13C-methyls in the context of a fully 2H,12C-labeled protein for large objects (several hundreds of kDa). These labeling are rather easy and not to expensive to obtain when the protein can be overexpressed in Escherichia coli. Unfortunately, many proteins cannot be produced this way, either because they need specific chaperons to be correctly folded or specific post-translational modifications to be active. It is sometime possible to over-express these proteins in yeast but in general higher organisms cells are required, the most commonly used being insect cells.

Insect cells are heterotrophic organisms that only grow on complex medium, the most efficient being commercial medium of unknown compositions. Some of them are available for NMR applications but at prohibitive prices. We began to develop, for our own applications but also to provide the NMR community a new tool, a set of labeling solutions relying on the use of commercial standard media depleted in amino acids and supplemented with labeled amino acid rich yeast extracts.

Publications :

  • Meola A, Deville C, Jeffers SA, Guardado-Calvo P, Vasiliauskaite I, Sizun C, Girard-Blanc C, Malosse C, van Heijenoort C, Chamot-Rooke J, Krey T, Guittet E, Pêtres S, Rey FA, Bontems F. Robust and low cost uniform 15N-labeling of proteins expressed in Drosophila S2 cells and Spodoptera frugiperda Sf9 cells for NMR applications. J Struct Biol. 2014 188(1):71-8.
  • Deville C, Girard-Blanc C, Assrir N, Nhiri N, Jacquet E, Bontems F, Renault L, Petres S, van Heijenoort C. FEBS Lett. 2016 590(20):3690-3699.

DADIMODO: a program for refining the structure of multidomain proteins and complexes against small-angle scattering data and NMR-derived restraints

DADIMODO is a program for refining atomic models of multidomain proteins or complexes against small-angle X-ray scattering data which can be combined with inter-domain distance and orientational restraints (e.g. residual dipolar couplings), derived from NMR measurements. While domain structures are mainly kept rigid during the optimization process, flexible regions can be user defined. Stepwise generic conformational changes, specified by the user, are applied cyclically in a stochastic optimization algorithm that performs a search in the protein conformation space. The convergence for this genetic algorithm is driven by an adaptable selection pressure. The algorithmic structure guarantees that a physically acceptable full atomic model of the structure is present at all stages of the optimization. A graphical interface was designed for user-friendly handling. A software package including the Python code, documentation, examples and additional scripts is available from the authors on request.

Publications :

  • Evrard G, Mareuil F, Bontems F, Sizun C, Perez J (2007) ” DADIMODO : a program for refining the structure of multidomain proteins and complexes against small-angle scattering data and NMR-derived restraints”. J. Appl. Crystallogr., 44(6):1264-1271.
  • Macheboeuf P, Piuzzi M, Finet S, Bontems F, Pérez J, Dessen A, Vachette P (2011) “Solution X-ray scattering study of a full-length class A penicillin-binding protein”. Biochem Biophys Res Commun, 405(1):107-11.
  • Mareuil F, Sizun C, Perez J, Schoenauer M, Lallemand JY, Bontems F (2007) “A simple genetic algorithm for the optimization of multidomain protein homology models driven by NMR residual dipolar coupling and small angle X-ray scattering data”. Eur Biophys J, 37 : 95-104.

NOEnet – Assignment of protein backbone resonances on the basis of the 3D structure and the NOE network

NOENet is a program to obtain the NMR resonance assignment of the protein backbone by exploiting an available 3D structure of the protein. Compared to the standard NMR assignment approach which is based on sequential J-coupling connectivities involving 13C, 15N and 1H nuclei, the structure-based assignment approach of NOEnet uses a completely independent data set, based on spatial NOE-connectivities among HN-HN nuclei. The comparison of the NOE-network with the 3D structure yields already satisfactory assignment results, which are improved by the addition of chemical shift (CS) and/or residual dipolar coupling (RDC) data. One step further, the combination of the two orthogonal data sources – J-coupling and NOE – gives very good assignment results, even for large proteins or difficult cases. The structure-based assignment approach of NOEnet is therefore a promising alternative to the standard NMR assignment approach and at the same time a helpful complement to it.

Publications :

  • Stratmann D, Guittet E, van Heijenoort C. Robust structure-based resonance assignment for functional protein studies by NMR. J Biomol NMR 2010 Feb ;46(2):157-73
  • Stratmann D, van Heijenoort C, Guittet E. NOEnet—use of NOE networks for NMR resonance assignment of proteins with known 3D structure. Bioinformatics 2009 25(4):474-481

New methods for fast protein resonance resonance assignment

The backbone resonance assignment of protein takes several days or weeks, even in favorable cases, such as small (<100 amino acids) proteins and cryoprobes, and in spite of the development of new methods to collect the data and sofwares for automatic assignment. The protein backbone resonance assignment is usually based on the protein sequence, a set of triple resonance experiments connecting pairs of H-N frequencies and the amino-acid type coded in 13C chemical shifts. We developped a new strategy that is based on several new experiments and original processing schemes, allowing to obtain the assignment in a few hours for favorable cases:

The BEST principle (Schanda, Van Melckebeke & Brutscher JACS2006/ Lescop, Schanda & Brutscher, JMR 2007) allows to pulse faster due to reduced proton relaxation between two repetitions of the pulse sequence; triple resonance H-N-C experiments can now be recorded in few tens of minutes instead of few hours
The spectral compression method ASCOM (Lescop E, Schanda P, Rasia R & Brutscher, B, JACS 2007) allows to significantly reduce the number of repetitions of the pulsesequence by optimizing the spectral width without loss of information.
The COBRA technique (Lescop E & Brutscher, B. JACS 2007) automatically extracts the connectivity information from sequential H-N frequency pairs from a pair of triple resonance experiment, such HN(CO)CA and intraresidual HNCA. COBRA is based on the computation of a correlation coefficient from two complex time series.
The new HADAMAC experiment (Lescop, E., Rasia, R., & Brutscher, B. JACS 2008) provide a very discriminant method for the amino-acid type identification for a H-N frequency pair.

In order to optimize the whole assignment process, we developped the BATCH software that works as a plateform for spectral processing, peak picking, fully automated backbone resonance assignment and chemical shifts extraction. In case of a small protein (<100 amino acids) at a mM protein concentration, the resonance assignment can now be obtained in a few hours.


Publications :

  • Brutscher B, Lescop E. “Fast protein backbone NMR resonance assignment using the BATCH strategy.” Methods Mol Biol. (2012), 831:407-28.
  • Lescop E, Kern T, Brutscher B. “Guidelines for the use of band-selective radiofrequency pulses in hetero-nuclear NMR : example of longitudinal-relaxation-enhanced BEST-type 1H-15N correlation experiments.” J Magn Reson. (2010) 203, 190-8
  • BATCH : Lescop E, & Brutscher B, “Highly automated protein backbone resonance assignment within a few hours : the BATCH strategy and software package” (2009) J. Biomol. NMR, 44(1):43-57