A “seed”-based molecular networking strategy for the screening and identification of unknown glucocorticoids in cosmetics

Abstract

The illegal addition of glucocorticoids in cosmetics has become a growing concern. However, due to the covert use of these additives, traditional targeted analytical methods have proven inadequate in addressing the evolving regulatory landscape. To tackle this issue, our study employed a “seed”-based molecular networking strategy for the non-targeted detection of glucocorticoids in cosmetics obtained through market surveillance. By utilizing 36 known glucocorticoids as “seed” nodes, we successfully constructed visualized molecular networking spectra for seven cosmetic products. Then, leveraging the data mining capabilities of MS-DIAL and MS-FINDER, 14 potentially risk substances were successfully identified, including newly discovered glucocorticoids, such as dexamethasone phosphate (Dex-P), prednylidene, and 7 alpha-thiospironolactone. To ensure the reliability of our findings, we proposed fragmentation pathways for the newly discovered glucocorticoids. Subsequent analyses involving molecular docking and molecular dynamics simulations indicated that these newly identified glucocorticoids could trigger skin atrophy and endocrine disorders, with Dex-P having the potential to exhibit the most potent impact. Furthermore, the accuracy of the Dex-P identification was validated through standard reference analysis, and its presence was confirmed in additional actual samples. This study presents an efficient methodology for regulating glucocorticoids in cosmetics and provides new insights into the scientific supervision of cosmetics.