The tissue-specific lipidomes of Arabidopsis thaliana

© 2021 Cheka Lumbini KehelpannalaLipids are important constituents of plants, performing many fundamental functions in growth and development. To study lipid structures and their biological functions robust analytical techniques are necessary. Lipidomics is a powerful tool, which can be used to measure the abundance of lipids in plant tissues and thus learn about their roles in plants, as well as to explore how plants respond to environmental perturbations and conduct an in-depth analysis of lipid biosynthetic pathways. A mass spectrometer is the predominant analytical instrument used in the field of plant lipidomics and two main approaches: the targeted and the untargeted approach are commonly followed. The untargeted approach seeks to analyse all lipids in a sample. By contrast, the targeted approach aims at identifying and quantifying a set of specific lipids. The identification of the individual chemical lipid structures represents a major limitation and challenge of the untargeted approach. Many lipids, which are structurally different, have identical molecular masses thus proving it difficult to identify them based on their molecular mass alone. As a result, biochemical pathways of specific lipid species are often difficult to decipher, which makes it challenging to interpret the data in a biological context and to integrate them with other ‘omics data. Tandem mass spectrometry allows a lipid to be assigned to a class of compounds and provides information on the total length of fatty acid chains and the number of double bonds. Although there are several in silico generated lipid mass spectral libraries for the annotation of lipids based on tandem mass spectrometric data, confident lipid identification remains challenging. This is mostly due to an over annotation of mass spectral features, and as a consequence demands manual corrections. Furthermore, plant lipids are significantly underrepresented in currently available lipid libraries. My PhD project aimed to address the above-mentioned challenges in plant lipidomics, using Arabidopsis thaliana (Arabidopsis) as the model plant. At the start of my PhD, a comprehensive untargeted lipid analysis workflow was established for the extraction, detection and identification of plant lipids. Next, this untargeted lipid analysis workflow was applied to explore the lipid profiles of different Arabidopsis tissues and developmental stages. In collaboration with the group of Prof Provart (University of Toronto) the resultant data of this study were integrated into an open-source in silico lipid map and database. In addition, a manually curated Arabidopsis specific lipid library, which represents the first plant-specific lipid library, was generated. This library can be used to confirm lipid annotations obtained through computer-generated lipid libraries and significantly reduce the number of false positives. It can also be used to annotate lipids when tandem mass spectrometric data are not available using retention time and accurate mass as identification parameters. Finally, the developed untargeted lipid analysis workflow was applied to explore the lipidomes of Arabidopsis loss-of-function mutants. Collectively, the research tools developed during my PhD and the data generated using the untargeted lipid analysis workflow will facilitate future studies into plant lipids and improve our understanding of lipids in plant functions