Dr Peter M. Abuja
Title of Presentation
Metabolomic analysis of plasma and serum of healthy donors from several European biobanks – need for standardization and quality management
Date and Place
Peter M. Abuja studied Biochemistry and Physical Chemistry in Graz. He worked several years on the structure of biological macromolecules in solution (X-ray scattering), and subsequently on the effect of antimicrobial peptides on conformation and stability of biological membranes. Before joining the biobanking community, a considerable proportion of his scientific work was done on oxidative stress and the activity of antioxidants, including modelling and simulation of lipid peroxidation reactions in human low-density lipoprotein.
Peter M. Abuja joined the Institute of Pathology at the Medical University of Graz in 2005 and has been involved since then in various projects in biobanking, metabolic disease and quality management in biomedicine. His present work focuses on the one hand on investigations of the stress response and mitochondrial function in mouse models for metabolic liver disease, and on the other hand on quality assessment in pre-analytical processing of tissue, serum and plasma with emphasis on metabolites. Besides the large proficiency testing study presented at the EBW, he was involved in an extensive study on the stability of the metabolome in cryopreserved liver tissue.
Multicenter studies involving large prospective cohorts require harmonized SOPs for processing biological samples for DNA, RNA, protein, metabolites or cells, to provide high quality samples and maintain high quality during storage. Only such samples allow state-of-the-art high throughput analytics, next-generation sequencing, mRNA expression analysis, epigenetic screening, as well as proteomic and metabolomic analyses.
To establish an overview on the current status of sample quality we performed proficiency testing in a ring-trial on archived serum and plasma samples from biobanks within the BBMRI-LPC consortium. Included were samples from 30 healthy donors per biobank, with equal share of male and female donors. Inclusion criteria were: overnight fast, age between 18 and 60 years, self-reportedly healthy, no medication, BMI between 18 and 30, no intensive training during the last 2 weeks before sample collection. Together with donors’ data, SOPs for the preparation of the samples were collected from each biobank.
Sample metabolomes were analysed by 1H-NMR profiling and by GC-MS and LC-MS/MS analysis. Principal component analysis of data from both metabolomics technologies revealed distinct variations between individual biobanks, which were much larger than the variation within donors from the same biobank. Several metabolites of higher concentrations could be quantified and were significantly different between the biobanks. Univariate analysis of MS data from plasma showed clear differences for 3 out of 8 biobanks compared to the other five, and one biobank revealed indications of variations of preanalytical sample processing.
These results again substantiate the need of standardization and implementation of quality control and management.