Application of diffusion-edited NMR spectroscopy for the structural characterization of drug metabolites in mixtures.

Diffusion-edited NMR spectroscopy is used to enable the structural characterization of low level metabolites in the presence of endogenous compounds, and organic solvents. We compared data from standard one-dimensional (1D) (1)H, 1D NOESY-presaturation, and 1D diffusion-edited experiments run on 20 microg and 100 microg samples of ethacrynic acid glutathione thioether (EASG) and a previously unreported metabolite of mefenamic acid, mefenamic acid glutathione thioester (MSG). The 1D NOESY-presaturation technique gave spectra with the best signal-to-noise (S/N) ratio, approximately three times that observed with the standard (1)H experiment, with respect to the metabolite signals. However, it was not selective for solvent signals as overlapping metabolite signals were also suppressed by this technique. In some cases, these signals were key to determining the site of glutathione attachment on the parent molecule. 1D NOESY-presaturation spectra also produced baseline distortions and inconsistent integration values. By comparison, 1D diffusion-edited experiments were found to selectively and simultaneously remove multiple solvent signals, resolve overlapping metabolite signals, and provide more uniform integration for metabolite signals overlapping with or proximal to solvent peaks, without producing baseline distortions. However, the diffusion-edited experiments caused significant signal attenuation of the metabolite signals when compared with a standard (1)H spectrum. Partially purified metabolites isolated from biological matrices were also characterized by using two-dimensional diffusion-ordered spectroscopy (DOSY). DOSY spectra acquired on a sample of EASG purified from rat bile proved useful in 'separating' the signals of EASG, from those of a co-eluting bile acid and parent drug ethacrynic acid (EA) in the diffusion-dimension in regions where there was no spectral overlap. In the low-field regions of high overlap, the DOSY experiment did not effectively separate the signals from the individual components. Diffusion based experiments provide a way to determine the total number of components that are present in a metabolite sample as well as an ability to identify them based on the chemical shift information, without the need for laborious chromatography on small samples.