An automated software pipeline for shotgun lipidomics using direct-infusion, high-resolution mass spectrometry

In the last ten years, lipidomics has attracted increasing attention as a research tool in a wide range of disciplines including physiology, lipid biochemistry, clinical biomarker discovery and pathology. Lipid metabolism is found to be critically aberrant in several different human diseases such as diabetes, obesity, atherosclerosis and Alzheimer’s disease. All these characteristics make lipids profiling an essential tool not only for investigation of many pathological processes but also in identifying potential biomarkers for establishing preventive or therapeutic approaches for human health. Here we present a direct-infusion mass spectrometry approach (shotgun lipidomics) in order to identify and quantify at least ten lipid species classes using a two-step extraction procedure, enabling both lipidiomics as well as polar metabolite analysis via GC-MS. We developed a new automated data analysis pipeline, which allows for the fast and robust quantification and identification of lipid species from high-resolution MS data. The software is based on the open-source C++ library OpenMS and R scripts, and supports automated ion-mode and adduct detection, isotope- assembly and correction, non-linear mass calibration, an in-house lipid database combining LipidMaps and HMDB for mass-based identification. Quality control plots are created for all major processing steps for each sample, allowing the operator to quickly judge data quality. Spike-in internal standards serve as abundance normalization for their respective lipid class. Additionally, a new robot-based sample preparation method is introduced, which allows standardized sample handling, ensures rapid sample processing, and minimizes potential variations in pipetting or weighing. Furthermore, due to use of glass vials, we show that common background signal (e.g. from tris(ditert-butylphenyl) phosphate) are significantly reduced. To study the effects of lifestyle factors, we investigate the lipidome changes in a mouse model considering four tissue types (WAT, serum, muscle and liver) under normal vs. high-fat diet.