A Nontargeted UHPLC-HRMS Metabolomics Pipeline for Metabolite Identification: Application to Cardiac Remote Ischemic Preconditioning

Judith Kouassi Nzoughet (1) , Cinzia Bocca (1) , Gilles Simard (2) , Delphine Prunier-Mirebeau (3) , Juan Manuel Chao de La Barca , Dominique Bonneau (4) , Vincent Procaccio (5) , Fabrice Prunier (6) , Guy Lenaers (4) , Pascal Reynier (7)

1 MITOVASC - MitoVasc - Physiopathologie Cardiovasculaire et Mitochondriale
2 SOPAM - Stress Oxydant et Pathologies Métaboliques
3 Service Biochimie et Génétique
4 BNMI - Biologie Neurovasculaire et Mitochondriale Intégrée
5 PREMMi - Pôle de Recherche et d’Enseignement en Médecine Mitochondriale
6 CRT - Cardioprotection, Remodelage et Thrombose
7 Mitochondrie : Régulations et Pathologie

In recent years, the number of investigations based on nontargeted metabolomics has increased, although often without a thorough assessment of analytical strategies applied to acquire data. Following published guidelines for metabolomics experiments, we report a validated nontargeted metabolomics strategy with pipeline for unequivocal identification of metabolites using the MSMLS molecule library. We achieved an in-house database containing accurate m/z values, retention times, isotopic patterns, full MS, and MS/MS spectra. A UHPLC-HRMS Q-Exactive method was developed, and experimental variations were determined within and between 3 experimental days. The extraction efficiency as well as the accuracy, precision, repeatability, and linearity of the method were assessed, the method demonstrating good performances. The methodology was further blindly applied to plasma from remote ischemic pre-conditioning (RIPC) rats. Samples, previously analyzed by targeted metabolomics using completely different protocol, analytical strategy, and platform, were submitted to our analytical pipeline. A combination of multivariate and univariate statistical analyses was employed. Selection of putative biomarkers from OPLS-DA model and S-plot was combined to jack-knife confidence intervals, metabolites' VIP values, and univariate statistics. Only variables with strong model contribution and highly statistical reliability were selected as discriminated metabolites. Three biomarkers identified by the previous targeted metabolomics study were found in the current work, in addition to three novel metabolites, emphasizing the efficiency of the current methodology and its ability to identify new biomarkers of clinical interest, in a single sequence. The biomarkers were identified to level 1 according to the metabolomics standard initiative and confirmed by both RPLC and HILIC-HRMS.

Domaines: Sciences du Vivant [q-bio]; Biologie cellulaire; Sciences pharmaceutiques; Alimentation et Nutrition; Biologie animale

Judith Kouassi Nzoughet, Cinzia Bocca, Gilles Simard, Delphine Prunier-Mirebeau, Juan Manuel Chao de La Barca, et al.. A Nontargeted UHPLC-HRMS Metabolomics Pipeline for Metabolite Identification: Application to Cardiac Remote Ischemic Preconditioning. Analytical Chemistry, 2017, 89 (3), pp.2138-2146. ⟨10.1021/acs.analchem.6b04912⟩. ⟨hal-02544712⟩