Line‐shape analysis of J‐resolved NMR spectra: application to metabolomics and quantification of intensity errors from signal processing and high signal congestion

Abstract

NMR spectroscopy remains one of the primary analytical approaches in metabolomics. Although 1D (1)H NMR spectroscopy is versatile, highly reproducible and currently the most widely used technique in NMR metabolomics, analysis of complex biological samples typically yields highly congested spectra with severely overlapping signals making unambiguous metabolite identification and quantification almost impossible. Consequently there is a growing use of 2D NMR methods, in particular (1)H J-resolved (JRES) spectroscopy, which spreads the high signal density into a second dimension. One potentially powerful method to deconvolute these JRES spectra, facilitating metabolite quantification, is via line-shape fitting. However, the mathematical functions describing the JRES NMR line-shape, in particular after applying apodisation functions and JRES specific processing, including tilting and symmetrisation, remain uncharacterised. Furthermore, possible quantitation errors arising from processing JRES spectra have not been evaluated, nor have the potentially adverse quantitative effects of overlapping dispersive tails of closely spaced signals in the 2D spectrum. Here we address these issues and evaluate the suitability of the JRES experiment for accurate complex mixture analysis. Specifically, we have examined changes in NMR line-shape and signal intensity after application of different apodisation functions (SINE and SEM) and JRES specific processing (tilting and symmetrising), comparing simulated and experimental data. We also report a significant quantitation error of up to 33%, dependent upon apodisation, due to overlap of the dispersive tails of closely spaced resonances. Finally, we have validated the use of these mathematical line-shape functions for metabolite quantitation of 2D JRES spectra, by comparison to corresponding 1D NMR datasets, using both gravimetrically-prepared chemically defined mixtures as well as biological tissue extracts.

Birmingham Metabolite Library: a publicly accessible database of 1-D 1H and 2-D 1H J-resolved NMR spectra of authentic metabolite standards (BML-NMR)

“…In regard to quantification, the JRES metabolic fingerprinting studies reported to date have an implicit assumption that the integrated signal areas are a measure of the relative concentrations of metabolites. This assumption has recently been confirmed following an evaluation of the suitability of the JRES experiment for quantitative metabolomics (Parsons et al 2009). …”

“…To date, the majority of 2-D JRES studies have simplified the spectra by calculating 1-D projections, followed by multivariate statistical analysis. Realising that considerable metabolic information was discarded by this approach, Parsons et al (2007) implemented the principal component analysis of multiple intact 2-D JRES spectra, yielding 3-D loadings plots that retained the advantages of signal dispersion into two frequency dimensions.…”

13C NMR Metabolomic Evaluation of Immediate and Delayed Mild Hypothermia in Cerebrocortical Slices after Oxygen–Glucose Deprivation

“…6C). It might be possible to improve the AMIX results by introducing more sophisticated preprocessing with better pH corrections or logarithmic scaling, 43 or other methods. 44–46 However, such would not circumvent the need for many more spectral data sets.…”

Metabolomics of Oxidative Stress in Recent Studies of Endogenous and Exogenously Administered Intermediate Metabolites

“…Of special note is the refined technique of Two-Dimensional 1 H J-resolved (JRES) NMR Spectroscopy [77,78]. J-coupling, which arises from the magnetic field property of particles that have spin, is the indirect coupling between two chemically bound nuclear spins that occurs via the electrons of their chemical bond.…”