Human urine as test material in 1H NMR-based metabonomics: recommendations for sample preparation and storage.

Metabonomic approaches are believed to have the capability of revolutionizing diagnosis of diseases and assessment of patient conditions after medical interventions. In order to ensure comparability of metabonomic 1H NMR data from different studies, we suggest validated sample preparation guidelines for human urine based on a stability study that evaluates effects of storage time and temperature, freeze-drying, and the presence of preservatives. The results indicated that human urine samples should be stored at or below -25 degrees C, as no changes in the 1H NMR fingerprints have been observed during storage at this temperature for 26 weeks. Formation of acetate, presumably due to microbial contamination, was occasionally observed in samples stored at 4 degrees C without addition of a preservative. Addition of a preserving agent is not mandatory provided that the samples are stored at -25 degrees C. Thus, no differences were observed between 1H NMR spectra of nonpreserved urines and urines with added sodium azide and stored at -25 degrees C, whereas the presence of sodium fluoride caused a shift of especially citrate resonances. Freeze-drying of urine and reconstitution in D2O at pH 7.4 resulted in the disappearance of the creatinine CH2 signal at delta 4.06 due to deuteration. A study evaluating the effects of phosphate buffer concentration on signal variability and assessment of the probability of citrate or creatinine resonances crossing bucket border (a boundary between adjacent integrated regions) led to the conclusion that a minimum buffer concentration of 0.3 M is adequate for normal urines used in this study. However, final buffer concentration of 1 M will be required for very concentrated urines.