Medizinische Universität Graz - Research portal
Selected Publication:
Zuellig, T.
Lipidomic analysis of human serum samples from different gut microbial background and different health conditions
Doktoratsstudium der Medizinischen Wissenschaft; Humanmedizin; [ Dissertation ] Medical University of Graz; 2021. pp. 149
[OPEN ACCESS]
FullText
- Authors Med Uni Graz:
- Advisor:
- Köfeler Harald
- Stojakovic Tatjana
- Trötzmüller Martin
- Altmetrics:
- Abstract:
- Lipids are molecules with great structural diversity and are responsible for a large number of biological functions. Dysregulation of lipid metabolism can have significant effects on various diseases and pathogenic conditions such as lung cancer, diabetes, and Alzheimer’s disease. Clinical lipidomics is a relatively new area of research and may show a lack of sensitivity to changes in response to certain diseases caused by problems of insufficient standardization in sample preparation, sample measurement, and data analysis. However, if lipid profiles and their role in the interaction of the metabolic pathway are better understood, it has the potential of a powerful tool for disease-specific diagnosis and therapy. In this thesis we present a global lipidomic workflow from sample preparation and data processing to statistical data analysis including batch drift normalization. With this workflow, we were able to process 3 large human serum sample cohorts with different phenotype backgrounds and an intervention study on obesity patients on a low-calorie diet with a total of over 1850 serum samples and an analysis time of over 185 days measured with the Q Exactive Focus mass spectrometer. The intervention study showed a reversible effect in the lipid profile between the six-week low-calorie diet and the normal food intake for a further six weeks. We were also able to determine differences in the phenotype-specific lipid profile in Tomorrow's cohort of the healthy group compared to the diabetes group. We identified 235 significantly different lipids (adj. P-value ≤ 0.05) across all lipid subclasses with multiple triacylglycerol species with more than twofold changes compared to the healthy group. We could also find differences between other phenotype-specific groups, e.g. cardiovascular disease and rheumatoid arthritis. The data were also transmitted to our project partner in order to compare them with metabolomics and microbiome data.