Profiling and isomer recognition of phenylethanoid glycosides from Magnolia officinalis based on diagnostic/holistic fragment ions analysis coupled with chemometrics

Isomers derived from natural products are promising candidates for drug discovery. However, characterization of isomers by mass spectrometry, especially stereoisomers and positional isomers, remains a large challenge due to insufficient reference standards and isomers’ highly similar fragmentation pathways or nondistinctive ion abundances. Herein, this report presents the first proposal of a method, combining multiple diagnostic ion/neutral loss (DINL) postanalysis and especially untargeted fingerprint analysis of all fragment ions (FAAFI) by means of a home-made program and chemometrics, to profile chemical components and recognize their isomers derived from medicinal plants. As a proof-of-concept, the chemical profiling of phenylethanoid glycosides (PhGs) and their isomers, which showed remarkable neuroprotective, anti-inflammatory and immunomodulatory effects, was performed. Using DINL to extract PhGs and FAAFI to distinguish their stereoisomers and positional isomers, as many as 87 PhGs, including 14 isomers, were tentatively identified from Magnolia officinalis; in addition, 17 PhGs were unambiguously identified by comparing the retention time and MS/MS data with those of reference compounds. Under the theory of Big Data analysis, the untargeted fingerprint analysis concerning unbiased ions was sufficient to contribute to discrimination of isomers even without evident distinction occurred for main ions.