A novel analytical workflow was developed and applied for the detection and identification of unknown xenobiotics in biological samples. High-resolution mass spectrometry (HRMS)-based data-independent MSE acquisition was employed to record full scan MS and fragment spectral datasets of test and control samples. Then, an untargeted data-mining technique, background subtraction, was utilized to find xenobiotics present only in test samples. Structural elucidation of the detected xenobiotics was accomplished by database search, spectral interpretation, and/or comparison with reference standards. Application of the workflow to analysis of unknown xenobiotics in plasma samples collected from four poisoned patients led to generation of xenobiotic profiles, which were regarded as xenobiotic fingerprints of the individual samples. Among 19 xenobiotics detected, 11 xenobiotics existed in a majority of the patients' plasma samples, thus were considered as potential toxins. The follow-up database search led to the tentative identification of azithromycin (X5), α-chaconine (X9) and penfluridol (X12). The identity of X12 was further confirmed with its reference standard. In addition, one xenobiotic component (Y5) was tentatively identified as a penfluridol metabolite. The remaining unidentified xenobiotics listed in the xenobiotic fingerprints can be further characterized or identified in retrospective analyses after their spectral data and/or reference compounds are available. This HRMS-based workflow may have broad applications in the detection and identification of unknown xenobiotics in individual biological samples, such as forensic and toxicological analysis and sport enhancement drug screening.
Keywords: Background subtraction; Database search; High-resolution mass spectrometry; Toxic chemicals; Unknown xenobiotics; Untargeted data-mining.
Copyright © 2016 Elsevier B.V. All rights reserved.