Human biomonitoring is the assessment of actual internal contamination of chemicals by measuring exposure markers, chemicals or their metabolites, in human urine, blood, serum, and other body fluids. However, the metabolism of chemicals within an organism is extremely complex. Therefore, the identification of metabolites is often difficult and laborious. Several untargeted metabolomics methods have been developed to perform objective searching/filtering of accurate-mass-based LC-MS data to facilitate metabolite identification. In this study, three metabolomics data processing approaches were used for chemical exposure marker discovery in urine with an LTQ-Orbitrap high-resolution mass spectrometry (HRMS) dataset; di-isononyl phthalate (DINP) was used as an example. The data processing techniques included the SMAIT, mass defect filtering (MDF), and XCMS Online. Sixteen, 83, and 139 probable DINP metabolite signals were obtained using the SMAIT, MDF, and XCMS procedures, respectively. Fourteen probable metabolite signals mined simultaneously by the three metabolomics approaches were confirmed as DINP metabolites by structural information provided by LC-MS/MS. Among them, 13 probable metabolite signals were validated as exposure-related markers in a rat model. Six (m/z 319.155, 361.127, 373.126, 389.157, 437.112 and 443.130) of the 13 exposure-related DINP metabolite signals have not previously been reported in the literature. Our data indicate that SMAIT provided an efficient method to discover effectively and systematically urinary exposure markers of toxicant. The DINP metabolism information can provide valuable information for further investigations of DINP toxicity, toxicokinetics, exposure assessment, and human health effects.
Keywords: Biomonitoring; Mass defect filtering (MDF); Metabolomics; Orbitrap; Signal mining algorithm with isotope tracing (SMAIT); XCMS online.
Copyright © 2016 Elsevier B.V. All rights reserved.