APPLICATION OF OMICS APPROACHES IN DRUG DISCOVERY AND BIOMARKER IDENTIFICATION

Recent advances in omics sciences have represented a revolution in many fields of biological research. Among these, lipidomics and metabolomics are gaining increasing interest due to their close connection with the organism’s phenotype. As the main analytical goal of omics approaches is to identify and quantify as many compounds as possible within a studied system, powerful analytical techniques and bioinformatics tools are required. Mass spectrometry (MS) coupled with liquid chromatography (LC) is one of the most powerful analytical platforms for the analysis of the lipidome in complex biological matrices and offers promising new insights in this field. Lipids are a large class of biomolecules involved in many biological processes with signalling, biophysical, and metabolic functions. The lipidome is complex, consisting of many species that share the same elemental composition but have different structural and physicochemical properties. Its comprehensive analysis is, therefore, technically challenging, and advanced LC-MS-based lipidomics workflows aim to better address this complexity. Lipidomics approaches can potentially be applied in all therapeutic areas, including cardiovascular, metabolic and inflammatory diseases. This is because an increasing number of human diseases are associated with significant lipidome remodelling, and changes in lipid profile have been identified as a major risk factor for many of them. With this thesis, we aimed to unravel the complexity of the lipidome and to assemble this knowledge in the description of biological systems to highlight lipid changes in pathology or response to drug treatment. To this end, we have optimized MS-based untargeted analytical workflows for the comprehensive profiling of lipids in different biological matrices and under various experimental conditions (in vitro and in vivo models) using advanced high-resolution LC-MS approaches. In Chapter 2, we integrated omics approaches into phenotypic screening drug discovery to support the identification of novel bioactive compounds. These studies aimed to apply untargeted multi-omics approaches to infer the mechanism of action of phytocomplex with potential lipid-lowering properties. Once better understood, lipid metabolism can be targeted pharmacologically. Phytochemicals from natural extracts represent an important source of bioactive compounds potentially useful in treating metabolic diseases characterized by major lipidome remodelling. However, the heterogeneity of natural matrices emphasizes the need for advanced analytical workflows to explore their beneficial effects, elucidating their mechanism of action and potential molecular targets. Using untargeted MS-based measurement of lipids and proteins allows a comprehensive investigation of the molecular mechanism of Scutellaria baicalensis (Study I) across different cellular states to help define the molecular changes responsible for its beneficial effect against liver disorders. In Study II, we established an analytical workflow to identify bioactive compounds in a polyphenolic fraction of bergamot leaves (Citrus bergamia). The experimental approach, which included a prefractionation step followed by LC-MS/MS analysis and phenotypic screening, revealed the great potential of leaves, usually considered as only waste products, as a source of polyphenols with positive effects on liver lipid metabolism. In Chapters 3 and 4, we applied UHPLC-MS analysis to study lipidome signatures in in vivo models focusing on specific key objects. Our primary goal was to determine the applicability of untargeted LC-MS/MS methods for the comprehensive analysis of lipid profiles in complex biological samples and to ensure that the analytical method was effective in capturing lipidome diversity and possible alterations with respect to environmental damage or pathological features. This helped us to derive some knowledge about how such stimuli could affect lipid metabolism. In particular, for both studies, we also investigated whether drug treatment could influence the lipidomic profile of the different biological samples analyzed (skin, liver and plasma). In Chapter 5, we aimed to accurately annotate oxidized lipid molecular species in complex biological matrices. Technically, the discovery and structural elucidation of oxidized lipids are still lacking, as their identification in biological samples is hampered by their low natural abundance and structural diversity. To deal with this complexity, advanced analytical and computational tools are required. We optimized a workflow combining a rapid untargeted lipidomic analysis with our newly assembled pipeline to support high-throughput detection and annotation of oxidized lipid molecular species in liver and plasma samples. The proposed workflow demonstrates its potential to address changes in oxidized lipidome in an untargeted manner and to guide future research on the role of oxidized lipids in biological systems. I.E. modified lipid species, which are known to be massively involved in the regulation of physiological and pathological conditions. In Chapter 6, we used the lipidomic dataset generated in Chapter 5 to infer deeper structural information about oxidized lipids putatively annotated. Currently, no single analytical method can fully characterize a lipid, and even in the field of MS-based lipidomics, more than one MS experiment is usually required to obtain deeper structural information. To identify the exact localization of unsaturation and oxidation positions within the acyl chains of oxidized lipids, we applied a different type of fragmentation mechanism, such as electron-activated dissociation (EAD). With EAD, we could generate accurate annotations of all selected oxidized lipids, providing information on chain length, location of oxygen, double bonds, and regioisomerism. Overall, this thesis highlights the utility of omics studies, particularly lipidomics, to the comprehensive description of biological systems. The project underlines the necessity of including omics analysis in drug discovery screening to explore molecular patterns responsible for observed phenotypic effects and generate hypotheses about the molecular targets. Moreover, our omics datasets helped us uncover molecular patterns defining pathological states and identify lipid species involved in the response of the organism to external stimuli. This also extends to deciphering the regulatory capacity of oxidized lipids. We believe that our efforts in the field of omics sciences have the potential to significantly advance our understanding of biological processes, improve drug discovery and development, and ultimately contribute to better therapeutic strategies.