Synthetic cannabinoids identification in e-liquid and biological samples through molecular network approach

The detection of new psychoactive substances, including synthetic cannabinoids, remains highly challenging. A wide array of chemical structures must be structurally elucidated, along with their associated metabolites, in biological samples collected from poisoned patients. For this purpose, non-targeted screening using liquid chromatography hyphenated to high resolution tandem mass spectrometry (LC-HRMS/MS) is a gold standard approach. Nevertheless, the identification steps remain the bottleneck of such analyses. For this purpose, molecular network has been recently proposed as a suitable strategy to organize tandem mass spectrometry data, needed for a reliable identification of unknown compounds. To illustrate this statement, we report the identification of synthetic cannabinoids and their metabolites by a non-targeted approach using molecular network strategy, in an 18-year-old poisoned patient hospitalized for neuropsychiatric disorders related to recent synthetic cannabinoid withdrawal. Upon admission, this patient was suspected of using adulterated e-liquid containing synthetic cannabinoids. Sample preparation of the tainted e-liquid was based on a liquid-liquid extraction using acetonitrile and methyl tert-butyl-ether. The tainted e-liquid was subsequently analysed with LC-HRMS/MS using electrospray as ion source in polarity switching mode and a Q-Exactive Focus® Orbitrap mass spectrometer. Data dependant acquisition was used to acquire tandem mass spectra on the three most intense ions. Molecular network was then generated following data processing on MZmine. The e-liquid was also analysed using GC-MS. The urine samples were extracted using the same procedure as for the e-liquid and was then analysed by LC-HRMS/MS. A segmental hair analysis (2 × 2 cm-long hair segments) was carried out by the CHRU de Lille using LC-HRMS and LC-MS/MS. Molecules detected in the e-liquid included glycerol, propylene glycol and nicotine. As expected, the e-liquid was also adulterated with synthetic cannabinoids namely MDMB-4en-PINACA and MDMB-5F-PICA. These two molecules were identified thanks to molecular network generated using LC-HRMS/MS data as well as using GC-MS, through SWGDRUG 3.9 library querying. Interestingly, in the e-liquid, compounds corresponding to trans-esterification products of MDMB-4en-PINACA was identified. The trans-esterification reaction was performed between MDMB-4en-PINACA and glycerol or propylene glycol and led to the corresponding products. Regarding urine sample of the patient, additional metabolites of MDMB-4en-PINACA were identified, including MDMB-4en-PINACA butanoic acid, dihydroxylated MDMB-4en-PINACA butanoic acid and glucurono-conjugated MDMB-4en-PINACA butanoic acid. The hair analysis of the patient allowed detection of MDMB-4en-PINACA and MDMB-5F-PICA in the two investigated hair segments. Analysis of the e-liquid and biological samples from the poisoned patient provides information on the chemical structure of synthetic cannabinoids and confirmation of synthetic cannabinoid use, respectively. The complete characterization of synthetic cannabinoid-related molecules in adulterated e-liquid, as exemplified by the identification of trans-esterification products of MDMB-4en-PINACA, is a needed step for understanding the mechanisms of cannabinoid toxicity. In addition, the identification of several metabolites in the patient's urine was possible through molecular network and propagation of the annotation. Hair analysis was useful in confirming the exposure to the two synthetic cannabinoids in the two segments studied and suggested chronic consumption of these products by the patient. The non-targeted analysis approach used to study this case offers new insights into the identification of novel psychoactive substances, as demonstrated by the identification of synthetic cannabinoids belonging to the indazole and indole structural families and related structures in e-liquid, hair and urine. Molecular networks have proven essential for the identification of synthetic cannabinoids, their derivatives, and metabolites in numerous matrices, as shown in the example of adulterated e-liquid and patient urine samples.