Indispensable for characterizing structures, interactions, and functional processes
Instruct has 5 centres offering Solution NMR across Europe. Navigate the map and click on the pins to discover centres near you.
To get structural information on proteins by solution NMR, isotopically labelled samples (usually 15N and/or 15N/13C) are required. For proteins with MW > 20 kDa, deuteration is also needed. In general, the degree of deuteration required is largely dependent on spectral quality, the size and modularity of the system studied and the types of experiment performed. However, very high levels of deuteration (>90 %) are usually required for proteins with molecular weights of the order of 35 kDa or greater.
The protein sample is usually dissolved in an aqueous buffer (typically ranging from 10 to 50 mM buffer concentration) to avoid pH variations in the biological samples, which can otherwise drastically change chemical shifts during NMR spectra acquisition. Buffer content plays a critical role in protein sample stability. Buffer optimization may be used to improve sample stability and avoid the following issues: slow precipitation; mixture of folded and unfolded protein; aggregation problems; degradation. A screening with several buffers (with different pHs and containing chemicals such as detergents, protease inhibitor cocktails, arginine etc.) is therefore recommended to optimize NMR samples. These buffers usually contain salts (i.e. NaCl or KCl) which often increase protein solubility. Samples high in salt or ionic strength may however yield reduced signal to noise when using cold NMR probes. The use of 3 mm tubes in this NMR probe, instead of the standard 5 mm tube, can alleviate this problem. When available, use of deuterated buffer is preferred.
Samples should be clear of precipitate and particulates and therefore need to be filtered or centrifugated in this circumstance. Samples require deuterium solvent for a lock signal and therefore water protein samples contain approximately 5-10% D2O.
Samples should be slowly transferred into and out of NMR tubes using long pipettes that reach the bottom of the NMR tube to avoid losing liquid to the walls of the tube or adding air bubbles to the sample which can negatively affect the shimming profile.
Standard NMR tubes are 5mm, which should contain at least 500 microliters of protein sample. If a small volume is required, Shigemi tubes, 5mm, use volumes of 270-300 microliters and are made of glass that is matched to the magnetic susceptibility of the solvent to be used. In the high-field cold probes, Shigemi tubes should also be used for more efficient water suppression.
A concentrated solution of a biological macromolecule at high temperature and around neutral pH represents an ideal growth medium for bacteria. Sodium azide and fluoride at less than 50 mM can be good growth inhibitors. Metal chelators (EDTA or EGTA) can also be good suppressor of microbial growth.
Structure determination by NMR typically requires a protein concentration of 0.5 mM or greater, stable for several days at the desired temperature. NMR studies for ligand binding or protein-protein interaction studies require concentrations of at least 0.05 mM.
The acquisition of standard NMR spectra takes from a few minutes (1D spectra) to a maximum of 3/4 days (3D or 4D spectra) for each experiment. A full set of experiments for protein structure determination usually takes about 15-20 days. In the case of relatively small proteins, recent advanced multidimensional data acquisition schemes have been successfully used to reduce experimental acquisition time by up to an order of magnitude or more.
The first NMR experiments acquired on site will be focused on the investigation of protein folding state of user’s sample as well as its aggregation state at the selected NMR protein concentration, acquiring 1D 1H and/or 2D 15N HSQC spectra and 15N backbone relaxation properties. This will take about 5-6 days. These preliminary data will allow to evaluate the time and the experimental conditions (i.e. kind of labelled nuclei, optimal protein, buffer concentration and pH) needed to obtain a high resolution structural determination on the user’s protein. A large number of pulse sequences is routinely available for assignment and structural characterization, and local and global dynamics can be easily estimated.
NMR structure determination requires a set of NMR experiments to obtain structural restraints of different nature, i.e. distances, angles and bond orientations. Several approaches are available to be specifically exploited depending on the size, aggregation state and folding properties of the protein molecule. The standard approach for solving protein structures by NMR involves two stages. In the first, the chemical shifts of the NMR-active nuclei in the protein are identified (chemical shift assignment), and in the second the protein fold is determined by computing protein conformations that are in agreement with thousands of experimentally determined inter-proton contacts (NOE assignment and structure generation). These conformers are generated through molecular dynamics (MD) simulations coupled with a simulated annealing protocol; a simplified force field is used in the MD simulations. The determination of the inter-proton contacts and calculation of protein conformations are run in an integrated, iterative manner. This combination allows the automation of the NOE assignment and structure generation to be part of the workflow. In addition, the chemical shift assignment is also supported by automated approaches. These computational services are available via the WeNMR web portal (http://www.wenmr.eu), which hosts a number of programs: MARS and Auto Assign for automatic chemical shifts assignment of the backbone atoms; CYANA and XPlor-NIH programs for automated/semi-automated NOE assignment and structure calculations. After structure generation, an energy optimization step is typically performed with a state-of-the-art force field (structure refinement). For this, a web interface to the AMBERprogram is available.Finally, protein-protein adducts can be structurally characterized with the HADDOCK program, which implements an information-driven flexible docking approach.The use of web portals makes the programs accessible also to non-specialists. The automation of the various steps mentioned above, and also of the complete, integrated workflow that is underway e.g. with the program UNIO, allows the researchers’ time devoted to the analysis and interpretation of the NMR data to be dramatically reduced.
A recent approach for structure determination exploits the chemical shift assignment step only. Several software, such as Chemical-Shift-ROSETTA (CS-ROSETTA), CHESHIRE and CS23D, have been developed to generate 3D structural models of proteins, using only the 13CA, 13CB, 13C', 15N, 1HA and 1HN NMR chemical shifts which are generally available at the early stage of the traditional NMR structure determination procedure, prior to the collection and analysis of structural restraints. The structural info contained in the Chemical Shifts is exploited by comparison with appropriate databases. At the moment these methodologies provide a robust approach to determining accurate structures of only small (<~125-residue) proteins. Info on “invisible” conformations in equilibrium with more abundant conformations can be obtained with additional experiments.