Discovery of exposure markers in urine for Brassica-containing meals served with different protein sources by UPLC-qTOF-MS untargeted metabolomics

Maj-Britt Schmidt Andersen, Helene Christine Reinbach, Åsmund Rinnan, Thaer Barri, Charlotte Elisabeth Mithril, Lars Ove Dragsted

30 Citationer (Scopus)

Abstract

An untargeted metabolomics approach has been applied to discover and identify exposure markers in urine for nine Nordic meals. A cross-over meal study was carried out in 17 subjects. The meals included a Pie, a Soup and a Barleyotto (pearl barley based risotto), each prepared with three protein sources; meat, fish or vegetarian. Urine samples were collected in different time intervals before and after intake of the test meals, covering a total of 24 h. The samples were analyzed by UPLC-qTOF-MS. Discriminating features for meals and protein sources were selected by use of double cross-validated partial least squares discriminant analysis and two additional validation steps: (1) time-course of excretion and (2) analysis of sensitivity and specificity. In addition, eight meal studies with single foods were carried out to investigate the food sources of the markers. In total 31 potential exposure markers (PEMs) of foods were found for the meals and protein sources. Fifteen of the 31 PEMs were also found in studies with single foods. Ten PEMs were identified or putatively annotated. Among the PEMs were a range of conjugated isothiocyanates from the Brassica oleracea species. Trimethylamine N-oxide was found as a fish marker. Additional unknown PEMs were found for chicory salad, parsley and fava beans, while other PEMs were dependent on the meal matrix rather than individual foods. The study demonstrates that it is possible to find PEMs in 24 h urine samples even when foods are given as part of a complex meal.

OriginalsprogEngelsk
TidsskriftMetabolomics
Vol/bind9
Udgave nummer5
Sider (fra-til)984-997
Antal sider14
ISSN1573-3882
DOI
StatusUdgivet - okt. 2013

Fingeraftryk

Dyk ned i forskningsemnerne om 'Discovery of exposure markers in urine for Brassica-containing meals served with different protein sources by UPLC-qTOF-MS untargeted metabolomics'. Sammen danner de et unikt fingeraftryk.

Citationsformater