Discovery and Characterization of Phenolic Compounds in Bearberry (Arctostaphylos uva-ursi) Leaves Using Liquid Chromatography-Ion Mobility-High-Resolution Mass Spectrometry.

The characterization and quantification of phenolic compounds in bearberry leaves were performed using hyphenated ion mobility spectroscopy (IMS) and a quadrupole time-of-flight mass spectrometer. A higher identification confidence level was obtained by comparing the measured collision cross section (TWCCSN2) with predicted values using a machine learning algorithm. A total of 88 compounds were identified, including 14 arbutin derivatives, 33 hydrolyzable tannins, 6 flavanols, 26 flavonols, 9 saccharide derivatives, and glycosidic compounds. Those most reliably reproduced in all samples were quantified against respective standards. Arbutin (47-107 mg/g), 1,2,3,4,6-pentagalloylglucose (6.6-12.9 mg/g), and quercetin 3-galactoside/quercetin 3-glucoside (2.7-5.7 mg/g) were the most abundant phenolic components in the leaves. Quinic acid and ellagic acid were also detected at relatively high concentrations. The antioxidant activity of the most abundant compounds was evaluated. A critical view of the advantages and limitations of traveling wave IMS and CCS for the discovery of natural products is given.

Predicting the antioxidant capacity and total phenolic content of bearberry leaves by data fusion of UV-Vis spectroscopy and UHPLC/Q-TOF-MS.

The total phenolic content (TPC) and antioxidant capacity have been considered as important quality parameters for plant extracts. In this study, bearberry leaves were regarded as studied subject and a reliable method was established to predict the TPC and antioxidant capacity of bearberry leaves. Ultraviolet-visible spectrometry (UV-Vis) and ultra high pressure liquid chromatography coupled to time-of-flight mass spectrometry (UHPLC/Q-TOF-MS) were used to provide spectral fingerprinting and metabolomic profiling. The data obtained (separately and merged) were used to build partial least squares (PLS) regression model. The PLS model built by using ultraviolet-visible spectra provided a satisfactory prediction result. Mid-level data fusion using the scores significantly improved the performance of PLS regression model, the residual predictive deviations (RPDs) for TPC and α, α-diphenyl-ß-picrylhydrazyl (DPPH) were 6.258 and 6.699, respectively, showing an excellent predictive ability. This study proved the potential of combination of UV-Vis spectrometry and UHPLC/Q-TOF-MS in the prediction of TPC and antioxidant capacity of plant extracts.