关键词: 桑叶；发酵；超高效液相色谱-四极杆-飞行时间串联质谱；主成分分析；偏最小二乘判别分析

Abstract: The objective of this study is to compare the effects of different fermentation processes with lactic acid bacteria (LAB) on the chemical constituents of mulberry leaves and to select superior LAB. The chemical components in mulberry leaves were detected by ultra-high performance liquid chromatography-quadrupole-time of flight-mass spectrometry (UPLC-Q-TOF-MS) and were identified by comparison with reference substances based on MS fragmentation pathways and literature information. Principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) models for discriminating different fermented mulberry leaves were built with SIMCA 14.1 software. PCA score plots, PLS-DA loading plots and variable importance in projection (VIP) values were obtained and based on them the major differential substances between before and after fermentation were determined. As a result, 41 chemical components were identified. PCA indicated a clear separation of mulberry leaves subjected to different fermentation processes. By PLS-DA, 11 chemical components with VIP values > 1, cosmetin, isorhamnetin-3-O-neohespeidoside, hyperoside, 3,4-dicaffeoyl quinine, aurantio-obtusin, salicylic acid, genistine, cyanidin 3-O-glucoside, guanosine, isoacteoside and acteoside, were selected as differential markers between mulberry leaves before and after fermentation. Based on comprehensive weighted scores, it was found that mulberry leaves fermented by Lactobacillus rhamnosus had the highest content of bioactive components. In conclusion, the chemical components of mulberry leaves changed significantly after fermentation. Flavonoids were the most important compounds to distinguish mulberry leaves fermented by different strains, and L. rhamnosus was a superior strain for the fermentation of mulberry leaves.

Key words: mulberry leaves; fermentation; ultra-high performance liquid chromatography-quadrupole-time of flight-mass spectrometry; principal component analysis; partial least squares-discriminant analysis