Optimierung und Charakterisierung eines Lipidomics-Workflows – von der Probenvorbereitung zur Datenanalyse

Lipidomics, the systematic analysis and characterization of all lipids in a biological sample, is a rapidly evolving field that provides deeper insights into the underlying biology and health status of an individual. This work addresses the challenges associated with the enormous diversity and different physico-chemical properties of lipids. A comprehensive lipidomics workflow was developed, optimized and characterized to study the lipidome of various biological samples. In this context, data analysis is a critical point. Typically, feature analyses of complex data sets are performed to determine differences between experimental and control groups. However, the results can depend heavily on the selected parameters. The workflow consists of three main components: the lipid extraction, the measurement method and the data processing and analysis. A multistep liquid-liquid extraction method with different organic solvents was chosen for the lipid extraction. A propane-2-ol/n-hexane extraction was performed in the first step followed by an optimized extraction according to Bligh and Dyer in the second step. This method ensures a high recovery of lipids from different classes, as evidenced by recovery rates of 60 to 110 % for eleven deuterated lipid standards in human blood plasma. An LC-IM-qTOF-MS-based method was established as the measuring method. The combination of liquid chromatography (LC) and ion mobility spectrometry (IMS) with a high-resolution quadrupole time-of-flight mass spectrometer (qTOF-MS) enables the separation and detection of complex lipid samples. IMS additionally separates the lipids according to their size-to-charge ratio, which increases the separation performance and enables the determination of collision cross-section (CCS) values as an additional identification parameter. The alternating frame mode enables a data-independent acquisition (DIA) of fragmentation data. Due to the simultaneous acquisition of fragment spectra, the structural elucidation of the lipids is further improved. This work compares different reversed phase liquid chromatography (RPLC) methods and a supercritical fluid chromatography (SFC) method for the separation of lipids. An RPLC method from the literature proved to be advantageous due to the shorter analysis time and better comparability with other studies. The SFC method, which separates lipids according to their head groups, can be used as a supplementary method for preliminary investigation and plausibility checks. In addition, acquisition sampling rate was optimized to reduce both peak variation and background noise, which improved the following data analysis. The developed LC-IM-qTOF-MS method is characterized by low detection and quantification limits, which are sufficient for most physiologically relevant lipids in human blood plasma. Data processing and analysis were optimized to improve the quality of the feature analysis. To this end, various strategies for reducing background noise and improving peak shapes were investigated. The combination of a threshold value of 20 counts and the smoothing of the data in the drift and chromatography dimension over three points proved to be optimal. This pretreatment effectively reduced background noise and improved peak shapes without discriminating low abundant lipids. The optimized alignment parameters and filtering of single-ion features during feature analysis also contributed to the improvement of data quality. Lipids were annotated using m/z ratios, drift times and fragmentation patterns. The method was validated on lipid extracts from HepG2 cells, human blood plasma and serum. About 200 lipids were annotated in positive and about 100 lipids in negative ESI mode. The results show that the developed method is robust and reproducible and allows a comprehensive analysis of the lipidome. In summary, this work provides an important contribution to the establishment of a robust and reproducible lipidomics workflow, which is important for health applications and the development of new diagnostic methods. The workflow enables the analysis of the lipidome and offers the potential to identify new health-promoting substances.