Preview	PDF, English - main document Download (72MB) | Terms of use
Preview	PDF, English (Errata) Download (119kB) | Terms of use

Abstract

The field of metabolism research evaluates the molecular basis of many important phenomena in biology and medicine, ranging from cellular function to systemic-level metabolic diseases. Spatial metabolomics investigates these phenomena in situ, by mapping metabolites in their native spatial context. Recent technical advances led to the development of matrix-assisted laser desorption/ionisation mass spectrometry (MALDI-MS) imaging, now established as a key technology for spatial metabolomics. MALDI-MS imaging allows for simultaneous and label-free detection of a wide range of analytes, such as metabolites, lipids, peptides and drugs at single-cell resolution. In this dissertation, I have looked into the capabilities this technology offers for untargeted metabolomics, and applied it to a scientific question that would be difficult to address with other technologies. The first aim of this dissertation was to create a systematic account of metabolite detectability by the common MALDI-MS imaging protocols. Despite the abundance of studies that focus on experimental or theoretical aspects of MALDI-MS, the scope of metabolite detectability by different protocols is not fully established, and choosing a suitable protocol for the detection of metabolites of interest remains a non-trivial task. To address this, I have developed experimental and computational tools for the preparation and analysis of a reference standard sample containing a wide selection of biologically-relevant metabolites. The comparison of 24 MALDI-MS protocols has shown the suitability of MS imaging for untargeted metabolomics and clarified which methods could be applicable for the analysis of individual chemical classes and biochemical pathways. The broad applicability of the obtained results was demonstrated through analyses of standard samples on other MS imaging technologies, as well as in comparison with biological tissue data sets. A community web resource was created to facilitate sharing of results and serve as an aid for the selection of the most suitable protocols for future spatial metabolomics experiments. The second aim of this dissertation was to study the associations between spatial differences in metabolism and composition of microbiota in the murine gut. This topic, important for the understanding of host-microbe interactions and their role in health and disease, remains underexplored in the field of microbiome research. By applying MALDI-MS imaging to this question, I discovered regions with distinct metabolic profiles in the whole faecal-matter filled intestine that previously were undetectable by bulk metabolomics. Next, I examined the contributions of individual metabolites, and found that both host physiology and bacterial metabolism contribute to the formation of the observed regions. Additionally, in a proof-of-principle experiment, I have shown that MALDI-MS imaging can serve as a method for determining bacterial localisation in high spatial resolution. Combining the information on the localisation of metabolites and bacteria is a way forward for obtaining direct functional insights into host-microbe interactions by a single technique, simultaneously, in high spatial resolution. Taken together, the computational developments and experimental results obtained in this dissertation help advance the field of spatially resolved untargeted metabolomics.