What types of biological questions can I answer with NMR?
There are too many uses for NMR of biological samples to list here, but here are some of the more common biological questions we routinely address:
• Determining the three-dimensional structure of proteins and protein complexes in solution
• Using NMR relaxation to probe the molecular motions at various timescales
• To determine the site of binding for small molecules, metals, other proteins, nucleic acids, etc.
• To monitor the protein stability for site directed mutants
• For studying conformational exchange in proteins and probing low populated intermediate states
See also the two review papers designed to give an introduction to Macromolecular NMR spectroscopy for the non-spectroscopist.
Why is there a size limit for biological samples?
As the molecular size of the sample goes up the NMR signal will decrease and eventually vanish. As the molecular size increases the tumbling rate of the molecule will slow. The slow tumbling will cause the NMR signal to de-phase rapidly causing the signal to broaden and decay quickly. In NMR pulse sequences there are a series of delays. When the relaxation or de-phasing of the NMR signal occurs faster than the delays needed to run the experiment, the experiment will fail.
In general well-behaved proteins below 20 kDa work well. Proteins between 20 – 30 kDa are more difficult, but can often be studied with conventional techniques under certain conditions. Proteins above 30 kDa are problematic with conventional NMR experiments. However, using high field NMR instruments, such as the NYSBC 800 MHz or 900 MHz instruments, coupled with deuteration of the protein sample, higher molecular weight samples may be studied but with an increased difficulty in interpretation of the data.
How long does it take to run an NMR experiment?
The answer to this question depends on the type of NMR experiment that is being run, the concentration of the sample, and the overall behavior of the system being studied. For a simple 1D spectrum of a small molecule data can typically be collected in 15 minutes. For very insensitive 3D and 4D NMR experiments a single experiment may take a week. To collect an entire set of experiments necessary to calculate a three-dimension structure data collection time can vary from a week to a few months.
Is there training for NMR data processing and/or data analysis?
Yes, the facility manager can help train new users on all steps necessary to utilize NMR in your research.
Can my protein by studied by NMR?
High-resolution NMR investigations are most feasible for proteins less than approximately 25 kDa in mass and soluble to around 0.5 - 1.0 mM. In certain cases it may be possible to investigate proteins or complexes of larger size or lower solubility. Membrane proteins are difficult to study using high-resolution methods, but new techniques are emerging that may be applicable. Because NMR studies usually require labeling the protein with the stable isotopes ¹⁵N, ¹³C, and sometimes ²H, an expression system suitable for growth in labeled media must be available. Proteins must be purified (typically >95% is required), folded, and at least marginally stable. Preliminary characterization by circular dichroism and thermal or solvent denaturation is recommended.
How much sample do I need?
This question is difficult to answer, as it is dependent on the questions being asked and the behavior of the sample. Protein concentrations for well-behaved systems should be above 300 µM for structural studies, but lower concentrations may be used for other non-structural studies. Whilst 300 µM is a rough estimate for the lowest concentration to use for structural studies it is advisable to make protein concentrations as high as possible and should be limited by solubility and protein behavior, not the amount of protein prepared. The amount of time running longer experiments to compensate for low concentrations and/or low sample volumes in small tubes and the increased time to interpret NMR spectra will almost certainly be longer then the time it takes to prepare additional sample. NMR tube configurations and typical sample volumes are listed below:
5mm NMR Tube volume: ~550 ul (used for typical buffer conditions like PBS and to get maximum sensitivity especially if protein quantity is not limiting or if protein solubility is low). For example, a ~20 kDa protein at ~500 uM concentration, you will need ~5-6 mg of protein.
5mm Shigemi NMR Tube volume: ~350 ul (used for nmr assignment production runs, especially when protein yields are limiting - avoid using at high temperatures because bubbles will often form within the sample window over time). For example, a ~20 kDa protein at ~500 uM concentration, you will need ~3-4 mg of protein.
3mm NMR Tube volume: ~160-180 ul (used for samples that are at elevated salt concentrations (>250 mM))
NOTE: sensitivity of 3mm is ~1/3x that of 5mm tube so the protein concentration and/or the number of scans must be increased accordingly . For example, a ~20 kDa protein at ~500 uM concentration, you will need ~1.5-2 mg of protein.
1.7mm SampleJet NMR Tube volume: ~50-60 ul (used only for screening conditions using 1D and 2D NMR; not practical for full nmr assignment production runs requiring multiple 3D experiments)
NOTE: sensitivity of 1.7mm is ~1/10x that of 5mm tube so the protein concentration and/or the number of scans must increased accordingly. For example, a ~20 kDa protein at 500 uM - 1 mM concentration, you will need ~0.5-1 mg of protein.
Please refer to our Sample Preparation Guide for details on tube types and volume requirements: NMR Sample Prep Info
What buffers can I use?
Traditionally non-protonated buffers such as sodium phosphate have been popular for NMR studies, as it does not have any detectable NMR signals to interfere. However, for proteins fully labeled with 15N and 13C just about any buffer will work. If buffer concentrations will be very high such that the 1% natural abundance of 13C is detectable the use of deuterated buffers may eliminate some artifacts.
See Kelly et al. (2002) "Low-Conductivity Buffers for High-Sensitity NMR Measurements"
What ionic strengths are appropriate?
The NMR signal strength will drop as ionic strength increases and this effect is exacerbated on high Q probes such as the cryogenically cooled probe on the Einstein Bruker 600 and at the NYSBC. Therefore, it is advisable to use the lowest ionic strength buffer as your sample will tolerate and still be well behaved. While low ionic strengths are preferred it is still possible to collect data with very high ionic strength albeit with reduced sensitivity. In general <50mM is best but up to 500mM salt can be used and above this the use of small diameter tubes might be useful. If high salt is required to keep a sample stable then an alternative is Arg+Glu salt, which has lower conductivity than NaCl and so better for use with cryoprobes.
See Hautbergue & Golovanov (2008) "Increasing the sensitivity of cryoprobe protein NMR experiments by using the sole low-conductivity arginine glutamate salt",
What pH should I use?
Check the theoretical pI of your protein with Expasy Protparam and then chose pHs at least 1 unit above/below this value. In general NMR spectra can be acquired at any pH. However, protons that are chemically labile (NH, OH, SH) can exchange with solvent protons and the rate of this exchange increases logarithmically above ~pH 2.6. Once the exchange becomes too fast the signal from the labile proton will no longer be observable. So practically pHs <7.5 is preferred.
Why do I need to add D2O to my sample?
The presence of deuterium in the solvent will allow the spectrometer to "lock" on the frequency of the D in the solvent allowing the instrument to compensate for any drift or changes in the magnetic field. For aqueous samples, at least 10% D2O is needed - we always have D2O in stock and can provide you with some.
Is there anything else I need to add to my nmr sample?
A reference compound for chemical shift referencing of the NMR spectra is useful to have especially if you are running your sample on multiple NMR instruments or are running spectra at temperature(s) different than 298 K. Two different reference compounds are available for use in aqueous solution: DSS and TSP-d4. We always have these in stock and can provide you with some - you only need enough to achieve a final concentration of <250 uM.
Another thing to remember is that since data acquisition can take many days to weeks on a protein NMR sample, you will want to limit both proteolysis and microbial growth that might occur during the NMR experiments. A common way to prevent this is to add EDTA (~100 uM) or a suitable protease inhibitor to prevent proteases from functioning and adding azide to 0.02% concentration to inhibit microbial growth. We have concentrated stock solutions of EDTA and azide if you need them for your sample prep.
Why do I need to 15N and/or 13C label my protein?
The experiments that are used to study proteins rely on observing the 15N and 13C chemical shifts in addition to protons. The natural abundance of 15N and 13C are 0.3% and 1.0% respectively. At such low natural abundance the sensitivity of the experiments is too low to detect. To resolve this issue proteins can be expressed on enriched media allowing full incorporation of these stable isotopes.
What temperature should I use for the NMR experiments?
The spectrometers have variable temperature capabilities, and the temperature of the sample can be adjusted between 0 C and 50 C to optimize the stability of your system. Raising the temperature above ambient typically gives higher quality spectra (because molecules tumble more rapidly) while lowering it may lead to lowered resolution. Obviously you will want to choose a temperature that gives high quality spectra without impacting the structure or function of the protein during the NMR data collection which can take days to weeks to complete.
NOTE: Once an NMR spectrum is obtained on your sample, the sample should be retested a week or other suitable time later to establish the stability of the sample. Obviously, the length of time that the sample must remain stable must be long enough to collect the data. Therefore, if you want to determine the structure of a large biomacromolecule, it must remain stable for a week or longer or you may have to prepare multiple samples so that a complete set of experiments can be obtained on it.
The UniProt Knowledgebase (UniProtKB) database contains protein sequences and information about the known biological functions: www.uniprot.org/
BMRB database: https://bmrb.io/
PDB database: https://www.rcsb.org/
Expasy - Swiss Bioinformatics Resource Portal: https://www.expasy.org/
Compute various physical and chemical parameters for a given protein sequence: https://www.expasy.org/resources/protparam