Assessment of lipid turnover (i.e. rates of synthesis, breakdown, and remodeling) is crucial in understanding lipid metabolism. The flux is typically monitored by relating the amount of newly synthesized and endogenous lipids by pulse-chase isotopic labeling experiment. In contrast to common 13C labeling of lipids by feeding 13C-glucose or 13C-glycerol, 15N labeling offers simple isotopic profile and distinguishes the lipid turnover from fatty acid synthesis. However, the molecular peaks of 15N labeled lipids are not distinguishable from 13C peaks of endogenous lipids by conventional means of mass spectrometry.
We report a workflow that combines in vivo organism-wide 15N metabolic labeling and shotgun ultra-high resolution mass spectrometry (sUHRMS) to measure absolute abundance and estimate the turnover rates of membrane lipids and major lipid precursors in body fluids and tissues of mice. Metabolic labeling was carried out by feeding young- and old-aged mice with a 15N-enriched SILAM diet over five time points. To achieve ultra-high resolution (~1.5M @ m/z 200), we coupled Q Exactive Orbitrap MS with an external data acquisition system (Booster X2) and accessed time-domain signals (transients). The transients were then processed using Peak-by-Peak software and lipids were identified and quantified by LipidXplorer software.
With sUHRMS workflow, we resolved 13C isotopes of unlabeled and monoisotopic peaks of 15N labeled lipid species (Δm = 0.00633 Da). We determined the molar abundance and turnover rates of over 120 nitrogen-containing species covering major classes of membrane lipids in mouse plasma, whole blood, four distinct regions of brain and liver. Furthermore, a broad-spectrum of specimen-specific, lipid class and molecular species-characteristic turnover kinetics were observed across ages. Notably, ethanolamine- and serine-containing lipids showed rapid turnover and relatively higher (ca. 2-fold) rate in comparison to the choline-containing lipids. The kinetics of turnover strongly differed among diacyl glycerophospholipid species in contrast to lysolipids. Moreover, brain regions (cortex, cerebellum, striatum, and hippocampus) exhibited much delayed and lower (ca. 4-5 fold) lipid turnover rates and discrete kinetic profiles than body fluids and liver. In the end, to check the reliability of our measures, we computed the abundance of 15N in the intracellular precursor (e.g. choline, serine) and intermediate (e.g. P-Chol, P-Etn, P-Ser, CDP-Chol, CDP-Etn) molecules that are available in the organism for lipid synthesis using shotgun metabolomics.
Taken together, for the first time we monitored lipid flux rates at the full organism level across ages, which will serve as a useful resource for a better understanding of the dynamics (metabolism and transport) of lipid species in mammals.