Comprehensive identification of sphingolipid species by in silico retention time and tandem mass spectral library

Hiroshi Tsugawa,Masanori Arita,Makoto Arita,Kazutaka Ikeda,Wataru Tanaka,Yuya Senoo

doi:10.1186/s13321-017-0205-3

Abstract

BackgroundLiquid chromatography coupled with electrospray ionization tandem mass spectrometry (LC–ESI–MS/MS) is used for comprehensive metabolome and lipidome analyses. Compound identification relies on similarity matching of the retention time (RT), precursor m/z, isotopic ratio, and MS/MS spectrum with reference compounds. For sphingolipids, however, little information on the RT and MS/MS references is available.ResultsNegative-ion ESI–MS/MS is a useful method for the structural characterization of sphingolipids. We created theoretical MS/MS spectra for 21 sphingolipid classes in human and mouse (109,448 molecules), with substructure-level annotation of unique fragment ions by MS-FINDER software. The existence of ceramides with β-hydroxy fatty acids was confirmed in mouse tissues based on cheminformatic- and quantum chemical evidences. The RT of sphingo- and glycerolipid species was also predicted for our LC condition. With this information, MS-DIAL software for untargeted metabolome profiling could identify 415 unique structures including 282 glycerolipids and 133 sphingolipids from human cells (HEK and HeLa) and mouse tissues (ear and liver).ConclusionsMS-DIAL and MS-FINDER software programs can identify 42 lipid classes (21 sphingo- and 21 glycerolipids) with the in silico RT and MS/MS library. The library is freely available as Microsoft Excel files at the software section of our RIKEN PRIMe website (http://prime.psc.riken.jp/).

Highlights

Liquid chromatography coupled with electrospray ionization tandem mass spectrometry (LC–ESI–MS/ MS) is used for comprehensive metabolome and lipidome analyses
The use of the enriched in silico retention time (RT) and MS/MS database was demonstrated as the application of MS-DIAL software
42 lipid classes, including 21 glycerolipids and 21 sphingolipids, can be identified in MSDIAL. Their diagnostic ions were theoretically assigned by hydrogen rearrangement rules in combination with manual curation (Additional file 1: Fig. S1, Additional file 3: Fig. S2, Additional file 4: Fig. S3). Their ion abundances were empirically optimized in the Sciex QTOFMS instrument at the collision energy of 40 with a 15 spread, the MS/MS library has been successfully utilized on other machines such as Waters, Brucker, and Agilent quadrupole time-of-flight (QTOF)-MS, and on Thermo Q-Exactive with the optimization of their MS conditions

Summary

Results

Negative-ion ESI–MS/MS is a useful method for the structural characterization of sphingolipids. We created theoretical MS/MS spectra for 21 sphingolipid classes in human and mouse (109,448 molecules), with substructurelevel annotation of unique fragment ions by MS-FINDER software. The existence of ceramides with β-hydroxy fatty acids was confirmed in mouse tissues based on cheminformatic- and quantum chemical evidences. The RT of sphingo- and glycerolipid species was predicted for our LC condition. With this information, MS-DIAL software for untargeted metabolome profiling could identify 415 unique structures including 282 glycerolipids and 133 sphingolipids from human cells (HEK and HeLa) and mouse tissues (ear and liver)

Conclusions

Background

Results and discussion