Integrated strategy for the extraction and profiling of bioactive metabolites from Passiflora mollissima seeds combining pressurized-liquid extraction and gas/liquid chromatography–high resolution mass spectrometry

Abstract

An integrated analytical methodology based on pressurized-liquid extraction (PLE) in two steps, followed by in vitro assays and liquid chromatography/gas chromatography coupled to high-resolution mass spectrometry (HRMS), was developed and applied for the isolation and characterization of potential bioactive metabolites from Passiflora mollissima seeds. PLE was proposed in two sequential steps: 1) recovery of the lipidic fraction using nonpolar solvents, and 2) recovery of the phenolic fraction from the defatted seeds’ residue using polar solvents. Cyclohexane was selected as the most suitable extraction solvent for the seeds defatting process (20 min, 100 °C and 100 bar). PLE optimization by response surface methodology was carried out to obtain phenolics-rich extracts with the highest antioxidant activity. Optimal extraction yield (6.58%), total phenolic content (29.99 mg g−1), total flavonoids content (0.94 mg g−1) and antioxidant activity (6.94 mM trolox g-1 and EC50 of 2.66 μg mL−1) were obtained operating at 150 °C with EtOH (100%) as solvent. Untargeted and semi-targeted MS and MS/MS data-mining strategies were successfully implemented for the rapid and comprehensive profiling of the polar and lipidic PLE fractions analysed by UHPLC and GC, respectively, coupled to quadrupole time-of-flight mass spectrometry (q-TOF-MS/MS). Polyphenols-rich extracts from P. mollisima seeds were characterized for the first time applying this approach, showing a broad variety of flavonoids, genuine flavanols (e.g. (epi)fisetinidol) and abundant proanthocyanidins. This application can be considered a successful demonstration of the great potential of the proposed methodology to effectively obtain and characterize complex natural extracts with potential bioactivity, by making use of powerful integrated identification strategies to facilitate the challenging post-acquisition data processing of huge datasets generated by HRMS analysis.