Department of Pharmaceutical Analysis, National Institute of Pharmaceutical Education and Research (NIPER), Guwahati, Changsari 781101, India.
DBT Centre for Molecular Biology and Cancer Research, Dr. Bhubaneswar Borooah Cancer Institute, Guwahati 781016, India.
Chem Res Toxicol. 2023 Apr 17;36(4):669-684. doi: 10.1021/acs.chemrestox.2c00401. Epub 2023 Mar 28.
Gutka, a form of smokeless tobacco, is widely used in the Indian subcontinent and in other regions of South Asia. Smokeless tobacco exposure is most likely to increase the incidence of oral cancer in the Indian population, and metabolic changes are a hallmark of cancer. The development of biomarkers for early detection and better prevention measures for smokeless tobacco users at risk of oral cancer can be aided by studying urinary metabolomics and offering insight into altered metabolic profiles. This study aimed to investigate urine metabolic alterations among smokeless tobacco users using targeted LC-ESI-MS/MS metabolomics approaches to better understand the effects of smokeless tobacco on human metabolism. Smokeless tobacco users' specific urinary metabolomics signatures were extracted using univariate, multivariate analysis and machine learning methods. Statistical analysis identified 30 urine metabolites significantly associated with metabolomic alterations in humans who chew smokeless tobacco. Receiver operator characteristic (ROC) curve analysis evidenced the 5 most discriminatory metabolites from each approach that could differentiate between smokeless tobacco users and controls with higher sensitivity and specificity. An analysis of multiple-metabolite machine learning models and single-metabolite ROC curves revealed discriminatory metabolites capable of distinguishing smokeless tobacco users from nonusers more effectively with higher sensitivity and specificity. Furthermore, metabolic pathway analysis depicted several dysregulated pathways in smokeless tobacco users, including arginine biosynthesis, beta-alanine metabolism, TCA cycle, etc. This study devised a novel strategy to identify exposure biomarkers among smokeless tobacco users by combining metabolomics and machine learning algorithms.
古特卡，一种无烟气烟草，在印度次大陆和南亚其他地区广泛使用。无烟气烟草的暴露最有可能增加印度人口口腔癌的发病率，而代谢变化是癌症的一个标志。通过研究尿液代谢组学并深入了解有患口腔癌风险的无烟烟草使用者的代谢变化特征，可以为开发用于早期检测和改善无烟烟草使用者预防措施的生物标志物提供帮助。本研究旨在使用靶向 LC-ESI-MS/MS 代谢组学方法调查无烟烟草使用者的尿液代谢变化，以更好地了解无烟烟草对人体代谢的影响。使用单变量、多变量分析和机器学习方法提取无烟烟草使用者的特定尿液代谢组学特征。统计分析确定了 30 种与人类咀嚼无烟烟草时代谢变化相关的尿液代谢物。接收者操作特征 (ROC) 曲线分析证明了来自每种方法的 5 种最具区分性的代谢物，这些代谢物能够以更高的灵敏度和特异性区分无烟烟草使用者和对照者。对多代谢物机器学习模型和单代谢物 ROC 曲线的分析揭示了具有更高灵敏度和特异性的区分无烟烟草使用者和非使用者的代谢物。此外，代谢途径分析描绘了无烟烟草使用者中几个失调的途径，包括精氨酸生物合成、β-丙氨酸代谢、TCA 循环等。本研究通过结合代谢组学和机器学习算法，设计了一种识别无烟烟草使用者暴露生物标志物的新策略。
