Identification And Prediction Exploratory Study Of Synthetic Cannabinoids In Vivo

Objective:1.To establish High Performance Liquid Chromatography-Mass Spectrometry(HPLC-MS/MS)method for the simultaneous detection of 87 Synthetic Cannabinoids and their metabolites in urine;2.Determine the positive samples in authentic cases,and explore the prevalence of Synthetic Cannabinoids and their metabolites and the characteristics of drug users.3.Establish a retention time prediction model for Synthetic Cannabinoids based on machine learning algorithms.To provide an effective tool for predicting the chromatographic properties of synthetic cannabinoids,and improve the efficiency of screening for the identification of unknown synthetic cannabinoids and their metabolic.Methods:1.Pretreatment and detectionTake 0.1m L of urine sample and add 10μLβ-glucuronidase and vortex it adequately.After that incubate the mixture at 55℃for 30 min.900μL acetonitrile mixed with internal standard was added to the sample for precipitates protein,and the extracted supernatant is detected by HPLC-MS/MS method.Qualitative ion pairs,retention times,and relative abundance ratios are used for qualitative analysis.Quantitative analysis is performed using standard curves(internal standard method).2.Study on the prevalence of synthetic cannabinoids and their metabolites in authentic samples and the characteristics of drug users.The established HPLC-MS/MS method was used to quantitatively detect the content of Synthetic Cannabinoids in real positive cases,and to explore the prevalence and characteristics of Synthetic Cannabinoids and their metabolites.3.Establishment of prediction model for the retention time of Synthetic Cannabinoids.Collect 232 Synthetic Cannabinoids under high resolution mass spectrometry liquid phase conditions(Waters Acquisition UPLC BEH C18(100×2.1mm,1.7μm)Columns and equivalent Van Guard pre-columns(2.1×5mm),35℃,gradient elution of 0.1%formic acid aqueous solution and 0.1%formic acid methanol solution,flow rate of 0.4m L/min),using Pa DEL software to calculate 13 molecular descriptors of 232 targets,and retention time prediction models were established based on five machine learning algorithms:Support Vector Regression,Random Forests,Gradient Boosting Regression,Ada Boost Regression,and e Xtreme Gradient Boosting.The model was validated using a 10-fold cross validation method,using 80%of the sample size as a training set,20%as a test set,and four Synthetic Cannabinoids out of model data as an external test set.Results:1.Detection of 87 Synthetic Cannabinoids and their metabolites in urineThe detection limit of Synthetic Cannabinoids and its metabolites in urine is0.02-2ng/m L,and the limit of quantitative is 0.05-5ng/m L.In the linear range,the correlation coefficient is greater than 0.995,the intra-day precision is 87.20-112.12%,and the intra-day precision is 91.51-110.42%.2.Application of analytical methods in authentic casesThis method has been successfully applied to 109 authentic forensic cases.The study involved 94 male and 15 female aged between 15 and 41 years(median 23 years).The detection of Synthetic Cannabinoids parents in urine is rare and low in content,with most of the detected metabolites being their metabolites.The most common metabolites are MDMB-4en-PINACA butanoic acid metabolite(86 cases)and ADB-BUTINACA acid metabolites(51 cases),with concentrations ranged from0.25-883.14 ng/m L and 0.34-46.38 ng/m L,respectively.3.Establishment of retention time predictive model for Synthetic Cannabinoids The best model selected through training is a Support Vector Regression(SVR)model that combines two types of fingerprints:Substructure Fingerprint Count(Sub FPC)and Fingerprint(FP).The R~2value of the verification set is 0.808,and the R~2value of the test set is 0.831.Using the optimal model FP+Sub FPC-SVR to predict four new Synthetic Cannabinoids,the prediction error is within 3%.Conclusion:1.This experiment has established a HPLC-MS/MS method for the simultaneous detection of 87 synthetic cannabinoids and their metabolites in urine.This method covers a variety of target substances,the sample pretreatment process is simple and rapid with high accuracy and sensitivity.It can be used for the screening and quantitative detection of 87 Synthetic Cannabinoids in forensic toxicology cases.2.The analysis of Synthetic Cannabinoids in 109 actual cases showed that Synthetic Cannabinoids mainly exist in the form of metabolites in urine,with almost no parent compounds in this work.Men aged 19-35 years are the main abusers of Synthetic Cannabinoids illegally used in this work.The drug combinations MDMB-4en-PINACA butanoic acid metabolite and ADB-BUTINACA acid metabolite are used with the highest frequency,which infers the trend of abuse of Synthetic Cannabinoids in 2021.Retrospective analysis of the detection results of Synthetic Cannabinoids in urine can help reveal the trends and characteristics of Synthetic Cannabinoid abuse,and the results are of great significance for the prevention and intervention of the abuse of Synthetic Cannabinoids.3.This study provides a model that can predict the retention time of Synthetic Cannabinoids.When used in combination with LC-HRMS,especially in the absence of reference standards,it can predict the retention time of Synthetic Cannabinoids.By comparing the predictive retention time with the experimental retention time of unknown samples,it can be used as a filter to improve the screening efficiency.