Appendix II K. Peptide Identification by Nuclear Magnetic Resonance Spectrometry
This general chapter is to be used in conjunction with general chapter 2.2.33. Nuclear magnetic resonance spectrometry in the context of peptide identification. The approach to be followed is qualitative and consists of comparing the nuclear magnetic resonance (NMR) spectrum of a test sample with that of a reference sample acquired under identical conditions.
This general chapter mainly applies to the use of proton NMR (1H NMR) spectrometry, to confirm the identity of small peptide products (up to approximately 15 amino acids). It is also applicable when using 13C NMR spectrometry with some modifications. The scope is restricted to the use of one-dimensional NMR spectrometry.
GENERAL PRINCIPLES
Equipment
Unless otherwise specified, an apparatus with a field strength giving an operating frequency for proton NMR of at least 300 MHz.
Spectral acquisition conditions and their optimisation
After introduction into the magnet, the sample is allowed to come to thermal equilibrium, especially if analysis is carried out at a temperature significantly different from room temperature: monitoring the lock signal is often a valuable visual guide to the progress of this process.
The spectral width must encompass the complete spectrum of the peptide, with an empty spectral region at each side. Typically, a spectral width of 12 ppm or 16 ppm is appropriate.
The following parameters may be optimised to improve resolution of characteristic peaks: temperature and/or pH primarily, buffer and peptide concentrations. Control of sample temperature is recommended but is not mandatory; if not used, the effect of small temperature changes on the appearance of the spectrum is validated.
The number of data points collected is such as to define peaks adequately.
Solvent suppression is not recommended but, if used, the intensities of peaks close to the solvent resonance may be affected and this has to be validated when comparing spectra.
Chemical shift referencing
For samples in aqueous solution, sodium 2,2-dimethyl-2-silapentane-5-sulfonate (DSS), sodium 3-(trimethylsilyl)propionate (TSP) or a deuterated analogue (TSP-d4) are appropriate, and the chemical shift of the methyl signals is often set to 0 ppm. Either the reference material is added at low amounts (10-100 ppm has been found to be appropriate) to the deuterated water used to dissolve the final sample, or an easily recognised internal resonance that is consistently present (such as acetate anion) can be used as a secondary reference. In this case, a validation spectrum obtained under the same spectral conditions is used to define the chemical shift of the secondary standard.
Sample size
Usually a few milligrams are used. If sample sizes are variable, the effects of this variation on the appearance of the spectrum are validated.
Sample preparation
The test and reference samples must be comparable in terms of concentration, pH and buffer composition. Typically, samples in solution are lyophilised, and the dried samples dissolved in deuterated water or a buffer in deuterated water. It may be worthwhile to lyophilise a solution in deuterated water one or more times (‘deuterium exchange’) as this reduces the intensity of strong solvent signals; volatile process impurities such as ethanol will also be lost. Use of buffer for the final sample preparation can reduce aggregation and improve spectral reproducibility by reducing batch-to-batch pH variation. Some probes are intolerant to high salt concentrations, but ionic strengths up to 200 mM sodium chloride are normally tolerated. High salt concentrations tend to increase 90° pulse length.
VERIFICATION OF IDENTITY
Determination of key spectral factors
Use of a qualitative approach does not entail stringent requirements on spectral parameters (for example, fast pulse repetition rates can be used, as full relaxation is not required). The use of short pulse widths (for example, a 30° pulse) and fast repetition rates will have no significantly deleterious effect on spectra, and will allow faster acquisition of acceptable signal-to-noise ratios. Variation in the pulse width and acquisition time within wide limits will not affect the ability to compare spectra. The number of scans collected must give appropriate signal-to-noise ratios for low intensity resonances and therefore a minimum signal-to-noise ratio of 50:1 is recommended.
Identification of characteristic resonances
It is possible to compare either the complete spectrum or a portion of it. Comparison of spectra of relevant samples will highlight regions of the spectrum that are distinctive, and comparison can be constrained to these regions. It is important to define resonances from impurities, such as residual solvents, which may be essentially irrelevant to product quality and which may vary in intensity between batches.
Spectral comparison
See the provisions of general chapter 2.2.33.