Combining Untargeted Lipidomics Analysis and Chemometrics to Identify the Edible and Poisonous Mushrooms (Pleurotus cornucopiae vs Omphalotus japonicus)

The global phenomenon of eating poisonous mushrooms by mistake occurs every year. Untargeted lipidomics analysis combined with chemometrics was used to identify mushroom varieties. Two kinds of mushrooms with similar appearance, namely, Pleurotus cornucopiae (P. cornucopiae) and Omphalotus japonicus (O. japonicus) were selected as models, where O. japonicus was a poisonous mushroom and P. cornucopiae was an edible mushroom. First, the lipid extraction efficiency of eight solvents was compared. The methyl tert-butyl ether/methanol (2:1, v/v) had higher lipid extraction efficiency of extracting mushroom lipids than other solvents, in terms of the lipid coverage, response intensity, and solvent safety. Afterward, the comprehensive lipidomics analysis of the two mushrooms was conducted. A total of 21 lipid classes and 267 molecular species were identified in O. japonicus, whereas 22 lipid classes and 266 molecular species in P. cornucopiae. The principal component analysis demonstrated that 37 characteristic metabolites, including TAG 18:1_18:2_18:0;1O, TAG 18:1_18:1_18:2, TAG 16:2_18:2_18:2, etc., could be used to distinguish the two mushrooms. These differential lipids were able to identify P. cornucopiae blended with 5% (w/w) O. japonicus. This study explored a novel method for identifying poisonous mushrooms from edible mushrooms and provided a reference for food safety of consumers.