Comprehensive lipid analysis of green Arabica coffee beans by LC-HRMS/MS

Abstract

Lipids play an important role in coffee bean development, coffee brew and in the effects of coffee on human health. They account for around 17% of the dry bean weight and encompass different classes and subclasses, mostly triacylglycerols (TAG) and a minor quantity of phospholipids (PL) and ^βN-alkanoyl-5-hydroxytryptamides (C-5HT). To comprehensive profile these different lipids, it is important to evaluate extraction methods that provide high lipid coverage and to analyze the lipids in high-resolution techniques. In this work, liquid chromatography-high resolution tandem mass spectrometry (LC-HRMS/MS) was employed to comprehensive profile lipids from green Arabica coffee beans and to evaluate the extraction efficiency and lipid coverage of three methods: Bligh-Dyer (BD), Folch (FO), and Matyash (MA). The MA method yielded the greatest number of annotated compounds (131 lipids) compared to the other methods. In the positive electrospray ionization (ESI) mode, the main difference among extraction methods was observed for TAG and diacylglycerols, whereas for the negative ESI it was observed differences for phosphatidylinositol (PI), lysophosphatidylinositol and phosphatidic acid (p < 0.05). The analysis of coffees from different maturation stages and/or post-harvest processes were also performed using the MA method. Immature beans were discriminated from mature and overripe beans by its lower levels of C-5HT, PI, phosphatidylcholine, lysophosphatidylcholine, phosphatidylethanolamine, and lysophosphatidylethanolamine. These results can help to better understand the coffee lipid composition and its association with coffee quality.

Keywords: Coffee; LC-MS; Lipidomics; Liquid-liquid extraction; MTBE; Orbitrap.