Lipidomics from sample preparation to data analysis: a primer

Abstract

Lipids are amongst the most important organic compounds in living organisms, where they serve as building blocks for cellular membranes as well as energy storage and signaling molecules. Lipidomics is the science of the large-scale determination of individual lipid species, and the underlying analytical technology that is used to identify and quantify the lipidome is generally mass spectrometry (MS). This review article provides an overview of the crucial steps in MS-based lipidomics workflows, including sample preparation, either liquid-liquid or solid-phase extraction, derivatization, chromatography, ion-mobility spectrometry, MS, and data processing by various software packages. The associated concepts are discussed from a technical perspective as well as in terms of their application. Furthermore, this article sheds light on recent advances in the technology used in this field and its current limitations. Particular emphasis is placed on data quality assurance and adequate data reporting; some of the most common pitfalls in lipidomics are discussed, along with how to circumvent them.

Keywords: Chromatography; LC-MS; Lipidomics; Mass spectrometry; Shotgun lipidomics.

Fig. 1

Lipid categories according to the International Lipids Classification and Nomenclature Committee, with one representative structure shown for each category

Fig. 2

The lipidomics workflow, including all essential steps from sample to biological outcome

Fig. 3

The upper panel shows an extracted-ion chromatogram of glycerophosphoinositol, PI 37:4 (m/z 871.534 ± 5 ppm), acquired by an Orbitrap instrument running at a resolution of 100,000 (m/z 400). The lower panel shows the corresponding MS/MS spectrum at a retention time of 19.82 min (CID fragmented). Neutral loss (NL) of carboxy and inositol (Ino) or ketene and inositol can be discerned. Blue fragments of 20:4 acyl chains, red fragments of 17:0 acyl chains, green fragments of PI head groups

Fig. 4

Overview of data processing and lipid annotation software grouped by functionality. Group A encompasses all-in-one software that can handle the full process pipeline: conversion of vendor-specific raw files (1); data processing, including peak detection and alignment as well as peak filtering options (2); and lipid annotation at the precursor MS level using web-based databases or precursor m/z lists and MS/MS derived lipid annotation at the rules-based or spectral matching level (3). Group B are software that are only used in lipidomics up to the precursor annotation stage. Group B software are used in combination with the highly specialized lipid annotation software that comprise group C. * Commercial software intended for DDA^a or DIA^b workflows

Fig. 5

Levels of lipid identification derived from mass spectrometric data, as exemplified by the inherent ambiguities of a phosphatidylethanolamine (PE) species containing an odd-carbon-numbered fatty acyl chain. Various mass spectrometric techniques yield different levels of certainty, which in turn should be reflected in the annotation