NMR-Based Method for Intramolecular 13C Distribution at Natural Abundance Adapted to Small Amounts of Glucose.

Quantitative nuclear magnetic resonance (NMR) for isotopic measurements, known as irm-NMR (isotope ratio measured by NMR), is well suited for the quantitation of 13C-isotopomers in position-specific isotope analysis and thus for measuring the carbon isotope composition (δ13C, mUr) in C-atom positions. Irm-NMR has already been used with glucose after derivatization to study sugar metabolism in plants. However, up to now, irm-NMR has exploited a "single-pulse" sequence and requires a relatively large amount of material and long experimental time, precluding many applications with biological tissues or extracts. To reduce the required amount of sample, we investigated the use of 2D-NMR analysis. We adapted and optimized the NMR sequence so as to be able to analyze a small amount (10 mg) of a glucose derivative (diacetonide glucofuranose, DAGF) with a precision better than 1 mUr at each C-atom position. We also set up a method to correct raw data and express 13C abundance on the usual δ13C scale (δ-scale). In fact, due to the distortion associated with polarization transfer and spin manipulation during 2D-NMR analyses, raw 13C abundance is found to be on an unusual scale. This was compensated for by a correction factor obtained via comparative analysis of a reference material (commercial DAGF) using both previous (single-pulse) and new (2D) sequences. Glucose from different biological origins (CO2 assimilation metabolisms of plants, namely, C3, C4, and CAM) was analyzed with the two sequences and compared. Validation criteria such as selectivity, limit of quantification, precision, trueness, and robustness are discussed, including in the framework of green analytical chemistry.

Radio-frequency pulse calibration using the MISSTEC sequence.

NMR sequences are composed of multiple radio-frequency pulses. Probe adjustment, sample concentration and solvent influence the loading factor, therefore these parameters also impact the validity of flip angles. The commonly used method to calibrate RF pulses is to measure a nutation curve by varying the pulse duration. However, this method is impacted by off-resonance effects, radiation damping and B1 and B0 inhomogeneities. Furthermore, it is important to avoid partial saturation. In this work, the MISSTEC sequence is proposed for pulse calibration. This sequence takes only 8 s or 2 min for 1H or 13C calibration, respectively. High accuracy (with an error below 1%) was obtained for both nuclei. Therefore, the calibrations can be done rapidly and accurately. Furthermore, the MISSTEC measurement could be performed on each sample - in an automated way- before acquisitions, after which the calibration found could be automatically used.

A precise and rapid isotopomic analysis of small quantities of cholesterol at natural abundance by optimized 1H-13C 2D NMR.

Cholesterol, the principal zoosterol, is a key metabolite linked to several health complications. Studies have shown its potential as a metabolic biomarker for predicting various diseases and determining food origin. However, the existing INEPT (insensitive nuclei enhanced by polarization transfer) 13C position-specific isotope analysis method of cholesterol by NMR was not suitable for very precise analysis of small quantities due to its long acquisition time and therefore is restricted to products rich in cholesterol. In this work, a symmetric and adiabatic heteronuclear single quantum coherence (HSQC) 2D NMR sequence was developed for the high-precision (few permil) analysis of small quantities of cholesterol. Adiabatic pulses were incremented for improving precision and sensitivity. Moreover, several strategies such as the use of non-uniform sampling, linear prediction, and variable recycling time were optimized to reduce the acquisition time. The number of increments and spectral range were also adjusted. The method was developed on a system with a cryogenically cooled probe and was not tested on a room-temperature system. Our new approach allowed analyzing as low as 5 mg of cholesterol in 31 min with a long-term repeatability lower than 2‰ on the 24 non-quaternary carbon atoms of the molecule comparing to 16.2 h for the same quantity using the existing INEPT method. This result makes conceivable the isotope analysis of matrices low in cholesterol. Graphical abstract.