Skip to main content
SLU publication database (SLUpub)
Doctoral thesis, 2019

Automated quantification of plasma metabolites by NMR to study prostate cancer risk biomarkers

Eriksson Röhnisch, Hanna

Abstract

Targeted quantitative NMR-based metabolomics can be used to identify disease risk biomarkers. However, NMR-based metabolomics yields complex spectra with signals from many different metabolites. These signals are often located within close proximity and, therefore, signal interferences are observed. Such interferences must be accounted for to yield accurate metabolite concentrations. Quantifications can become very time-consuming, especially in large-scale studies. In response to this, an Automated Quantification Algorithm (AQuA) was designed as the final step of the NMR-based metabolomics workflow. Implementation and evaluation was done for quantification of human plasma metabolites in samples collected using heparin as anticoagulant. AQuA enabled the quantification of 67 metabolites in 1342 samples within one second on a standard personal computer. AQuA performed with equal accuracy as a manual procedure for targeted profiling performed using a software package dedicated to metabolite quantification by NMR. In contrast to using heparin as anticoagulant, the use of EDTA introduced additional interferences. With some modifications, AQuA also quantified human plasma metabolites despite the presence of the high intensity signals from EDTA, some of which displayed inter-spectral deviations in signal positions and line widths. To further demonstrate its usefulness, AQuA was utilised for risk biomarker discovery in a case-control study nested within the Northern Sweden Health and Disease Cohort. Plasma metabolites were quantified in samples from 1554 men, 777 whom were diagnosed with prostate cancer more than 5 years after sample collection (baseline), and 777 whom were matched controls. MS-based metabolomics was also employed to yield complementary information. Conditional logistic regression and correction for multiple testing were performed. Risk biomarkers for prostate cancer varied with baseline age and disease aggressiveness. For example, glycine and pyruvic acid were identified in younger subjects, while lipid species (e.g., lysophosphatidylcholines) associated with overall disease risk in older subjects and with risk of aggressive disease. A reverse cross-association could also be identified between risk of prostate cancer and type 2 diabetes at the metabolite level.

Keywords

Automated Quantification Algorithm (AQuA), mass spectrometry, Northern Sweden Health and Disease Cohort, nuclear magnetic resonance, prostate cancer, risk biomarkers, targeted metabolomics, type 2 diabetes

Published in

Acta Universitatis Agriculturae Sueciae
2019, number: 2019:2
ISBN: 978-91-7760-322-1, eISBN: 978-91-7760-323-8
Publisher: Department of Molecular Sciences, Swedish University of Agricultural Sciences