Modeling the fragmentation patterns of triacylglycerides in mass spectrometry allows the quantification of the regioisomers with a minimal number of standards.

Affiliation

UR1268 BIA (Biopolymères Interactions Assemblages), INRA, 44316, Nantes, France; Laberca, Oniris, INRA, 44307, Nantes, France. Electronic address: [Email]

Abstract

Mass spectrometry allows the relative quantification of the regioisomers of triacylglycerides by the calibration of their fragmentation patterns. However, due to the plethora of regioisomers of triacylglycerides, calibration with every standard is not feasible. An analytical challenge in the field is the prediction of the fragmentation patterns of triacylglycerides to quantify their regioisomers. Thus, we aimed to model these fragmentation patterns to quantify the regioisomeric composition, even for those without commercially available standards. In a first step, we modeled the fragmentation patterns of the regioisomers of triacylglycerides obtained from different published datasets. We found the same qualitative trends of fragmentation beyond differences in the type of adduct in these datasets (both [M+NH4]+ and [M+H]+), and the type of instrument (orbitrap, Q-ToF, ion-trap, single quadrupole, and triple quadrupole). However, the quantitative trends of fragmentation were adduct and instrument specific. From these observations, we modeled quantitatively the common trends of fragmentation of triacylglycerides in every dataset. In a second step, we applied this methodology on a Synapt G2S Q-ToF to quantify the regioisomers of triacylglycerides in sunflower and olive oils. The results of our quantification were in good agreement with previous published quantifications of triacylglycerides, even for regioisomers that were not present in the training dataset. The species with more than two highly unsaturated fatty acids (arachidonic, eicosapentaenoic, and docosahexaenoic acids) showed a complex behavior and lower predictability of their fragmentation patterns. However, this framework presents the capacity to model this behavior when more data are available. It would be also applicable to standardize the quantification of the regioisomers of triacylglycerides in an inter-laboratory ring study.

Keywords

Lipidomics,Modeling,Positional isomers,Regioisomers,Triacylglycerols,