Untargeted metabolomics in human health risk assessment of nanoparticles: data processing workflow challenges

Abstract

Consumers are potentially exposed to nanomaterials during the use of consumer products. Emerging approaches in the area of assessment of human health effects combine the use of in vitro cell systems and -omics approaches. Metabolomics is complementary to the other -omics disciplines and the combination of these approaches (eg. proteomics and metabolomics) may provide important information about the status of a cell and unique insights into metabolic dynamics (metabolite changes over time). Analysis of the metabolome is challenging, not least because of the chemical diversity of cellular metabolites, as well as the redundancy of cellular metabolic pathways that complicate data elaboration and interpretation. Progress in metabolomic data processing, however, lags considerably behind that of the other -omics disciplines. Our work aims at bringing metabolomics to a similar "standard" as the other large -omics disciplines by working on the workflow for LC-MS data processing as well as on data management. We have contributed to the development of R scripts that call XCMS to setup a workflow of LC-MS metabolic data processing. XCMS is a modular, open source and platform independent data processing pipeline (http://metlin.scripps.edu/xcms/). The flexibility of R-XCMS based programmes allows us to keep track of all operations in the workflow. This work was carried out as a feasibility study for implementation of LC/MS metabolomic workflows. We applied an "untargeted" LC-MS-based approach and chose here to investigate the modification in the metabolome of Caco-2 cell line (as a model of the intestinal barrier) after exposure to gold nanoparticles (AuNPs).