AUTHOR=Mojsak Patrycja , Maliszewska Katarzyna , Klimaszewska Paulina , Miniewska Katarzyna , Godzien Joanna , Sieminska Julia , Kretowski Adam , Ciborowski Michal TITLE=Optimization of a GC-MS method for the profiling of microbiota-dependent metabolites in blood samples: An application to type 2 diabetes and prediabetes JOURNAL=Frontiers in Molecular Biosciences VOLUME=9 YEAR=2022 URL=https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2022.982672 DOI=10.3389/fmolb.2022.982672 ISSN=2296-889X ABSTRACT=

Changes in serum or plasma metabolome may reflect gut microbiota dysbiosis, which is also known to occur in patients with prediabetes and type 2 diabetes (T2DM). Thus, developing a robust method for the analysis of microbiota-dependent metabolites (MDMs) is an important issue. Gas chromatography with mass spectrometry (GC–MS) is a powerful approach enabling detection of a wide range of MDMs in biofluid samples with good repeatability and reproducibility, but requires selection of a suitable solvents and conditions. For this reason, we conducted for the first time the study in which, we demonstrated an optimisation of samples preparation steps for the measurement of 75 MDMs in two matrices. Different solvents or mixtures of solvents for MDMs extraction, various concentrations and volumes of derivatizing reagents as well as temperature programs at methoxymation and silylation step, were tested. The stability, repeatability and reproducibility of the 75 MDMs measurement were assessed by determining the relative standard deviation (RSD). Finally, we used the developed method to analyse serum samples from 18 prediabetic (PreDiab group) and 24 T2DM patients (T2DM group) from our 1000PLUS cohort. The study groups were homogeneous and did not differ in age and body mass index. To select statistically significant metabolites, T2DM vs. PreDiab comparison was performed using multivariate statistics. Our experiment revealed changes in 18 MDMs belonging to different classes of compounds, and seven of them, based on the SVM classification model, were selected as a panel of potential biomarkers, able to distinguish between patients with T2DM and prediabetes.