Nuclear magnetic resonance-based metabolomics of OCT-embedded frozen kidney samples in mouse and man following standardized pre-analytics

Introduction: Pre-analytical processing significantly affects tissue metabolomes. Since most frozen kidney samples are stored after embedding, standardization of cryoprotective medium removal before metabolomics is essential. Objectives: We used rodent and human kidney samples to develop an easy and robust pre-analytical procedure compatible with 1H-nuclear magnetic resonance (NMR)-based metabolomics. Methods: In mice, renal ischemia was induced for 30 min, followed by 48-h reperfusion (I/R, n = 6). Right kidneys were transversally cut in two fragments, and snap-frozen in liquid nitrogen (LN2) or in Optimal Cutting Temperature® (OCT) fixative. In man, double kidney biopsies were simultaneously obtained before transplantation (n = 15), and snap-frozen in LN2 or OCT. Results: 1H-NMR spectrum of pure OCT highlighted two major peaks, i.e. from 3.4 to 4.2 ppm (47.2%) and from 1.2 to 2.2 ppm (42.5%). 1H-NMR spectra of mouse OCT kidneys were biased at 3.7. By contrast, 1H-NMR analyses of mouse OCT kidneys iteratively rinsed in saline significantly discriminated sham versus I/R groups, with Q² at 0.695 (to be compared with Q² at 0.866 for LN2 sham vs. I/R kidneys). Discriminant metabolites were analogous in both OCT and LN2 kidneys, with a correlation coefficient of 0.83. In man, iteratively rinsing OCT kidneys in saline eliminated the spectral 3.7-peak, thereby making metabolomes of OCT kidneys interpretable and similar to LN2 samples, with a correlation coefficient of 0.73. Conclusion: NMR metabolomics using OCT-frozen kidney samples is valuable in mouse and man, following standardized OCT removal. This may help use residual biobanked human tissues to better understand renal pathophysiology. © 2017, Springer Science+Business Media, LLC.