Simultaneous High-Performance Recovery and Extended Acid-Catalyzed Hydrolysis of Oleuropein and Flavonoid Glycosides of Olive (Olea europaea) Leaves: Hydrothermal versus Ethanol Organosolv Treatment

Abstract

Olive leaves (OLLs) are an exceptional bioresource of natural polyphenols with proven antioxidant activity, yet the applicability of OLL extracts is constrained by the relatively high polarity of the major polyphenols, which occur as glycosides. To overcome this limitation, OLLs were subjected to both hydrothermal and ethanol organosolv treatments, fostered by acid catalysis to solicit in parallel increased polyphenol recovery and polyphenol modification into simpler, lower-polarity substances. After an initial screening of natural organic acids, oxalic acid (OxAc) was found to be the highest-performing catalyst. The extraction behavior using OxAc-catalyzed hydrothermal and ethanol organosolv treatments was appraised using kinetics, while treatment optimization was accomplished by deploying response-surface methodology. The comparative assessment of the composition extracts produced under optimal conditions of residence time and temperature was performed with liquid chromatography-tandem mass spectrometry and revealed that OLLs treated with 50% ethanol/1.5% HCl suffered extensive oleuropein and flavone glycoside hydrolysis, affording almost 23.4 mg hydroxytyrosol and 2 mg luteolin per g dry weight. On the other hand, hydrothermal treatment with 5% OxAc provided 20.2 and 0.12 mg of hydroxytyrosol and luteolin, respectively. Apigenin was in all cases a minor extract constituent. The study presented herein demonstrated for the first time the usefulness of using a natural, food-grade organic acid to perform such a task, yet further investigation is needed to maximize the desired effect.

Keywords: acid hydrolysis; antioxidants; olive leaves; oxalic acid; polyphenols.

Figure 1

Bar plot depicting the effect of various acids and concentrations on the polyphenol recovery from OLLs. Extractions were accomplished at 90 °C, for 180 min. Bars assigned with different letters (a–d) represent statistically different values (p < 0.05).

Figure 2

Bar plot presenting the effect of water/ethanol proportion (v/v) on the polyphenol recovery from OLLs, using 1.5% HCl or 5% oxalic acid as acid catalysts. Bars denoted with different letters (a–c) are statistically different (p < 0.05).

Figure 3

Polyphenol extraction kinetics during treatment of OLLs with various solvent systems: (A) water; (B) 1.5% aqueous HCl; (C) 5% aqueous oxalic acid; (D) 50% ethanol; (E) 50% ethanol/1.5% HCl; (F) 50% ethanol/5% oxalic acid. Bars represent the standard deviation.

Figure 4

(A) Correlation between the actual and the predicted values of the response (Y_TP), as determined by response-surface methodology, implemented to model the effect of residence time (t) and temperature (T) on the polyphenol recovery from OLLs. (B) Desirability plot showing the codified values of the optimum conditions (t, T) and the optimum predicted response (Y_TP) under these conditions. Inset tables display the statistical information obtained after performing lack-of-fit and ANOVA tests. Colored values are statistically significant (p < 0.05).

Figure 5

Graphical 3D representation of the effect of the independent variables (t, T) on the response (Y_TP), as revealed by deploying response-surface methodology.

Figure 6

Selected ion chromatograms illustrating the transformation of oleuropein and its isomers after OLL treatment with 50% ethanol/5% oxalic acid.

Figure 7

Selected ion chromatograms illustrating the transformation of luteolin glycosides after OLL treatment with 50% ethanol/5% oxalic acid.

Figure 8

Selected ion chromatograms illustrating the transformation of apigenin glycosides after OLL treatment with 50% ethanol/5% oxalic acid.

Figure 9

Bar plots showing the antiradical activity (A) and the ferric-reducing power (B) of the extracts obtained by treating OLLs with various solvent systems, under optimized conditions. Bars assigned with different letters (a–f) represent statistically different values (p < 0.05).