Profiling of Drug Metabolites and Detection of Unknown Controlled Substances Using Tandem Mass Spectrometry Molecular Networking

Abstract

Molecular networking (MN) has become a popular data analysis method for untargeted mass spectrometry (MS)/MS-based metabolomics. Thus, it has the potential to be used for research in which unknown substance screening is important, such as metabolite identification study or new psychoactive substances (NPSs) screening. However, the practical effectiveness or its detailed application method has not been studied. In this study, two utilization methods of MN were suggested. For drug metabolite screening study, the performance of using MN, mass defect filtering, Agilent MassHunter Metabolite ID, and the Agilent Mass Profiler Professional workflow was compared to annotate metabolites of sildenafil generated in an in vitro liver microsome-based metabolic study. In total, 28 known metabolites were found in this study, including 15 additional unknown isomers and 25 unknown metabolites. As a result of the comparison, MN outperformed the existing ones in terms of the number of detected metabolites (27 known and 22 unknown metabolites), the false positive rate, and the time and effort required for human labor-based postprocessing. It has been demonstrated that the performance is equal to or better than that of other tools. For controlled substance screening study, MN was proposed as a screening method for controlled substance by the example of NBOMe, a family of NPSs. The fragmentation patterns of NBOMe derivatives were analyzed using liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QTOF/MS). MS/MS spectral data were used to generate MN maps for NBOMe derivatives. The fragmentation patterns of nine NBOMe derivatives were interpreted based on the product ion spectrum data. NBOMe derivatives generally exhibited similar product ion spectral patterns. Among them, halogen-substituted methoxybenzylethanamine-type derivatives exhibited characteristic product ions of radical cations. MN analysis of MS/MS data revealed that all NBOMe derivatives formed one integrated network cluster, distinguishing them from other types of NPSs. NBOMe derivatives were spiked into human urine and identified by connection to the NBOMe database network. In addition, NBOMe compounds not registered in the database were also recognized by MN as NBOMe-related substances. These studies demonstrated that MN has the potential to become an efficient and powerful platform for the detection and identification of drug metabolites at all stages of drug development, including drug discovery, preclinical development and early clinical trials. Furthermore, MN-based Unknown screening methods could be promising tools for open screening of designer drugs for forensic or doping analyses.