Introduction

Alcoholic beverages obtained by plant maceration are traditional foods, rich in natural compounds, with pleasant sensorial characteristics and physiological effects [1]. Specifically, berry liqueurs could be a source of bioactive compounds, because of the significant level of antioxidant phenolics in these fruits [2]. Myrtle liqueur is a traditional Sardinian beverage obtained by room temperature maceration of purple myrtle berries (Myrtus communis L.) in ethanol followed by dilution with water and sucrose [3, 4]. Commercially, two different myrtle liqueurs are produced: a purple-red liqueur, produced from purple myrtle berries (M. communis var. melanocarpa DC), and a white liqueur, produced by alcoholic maceration of young myrtle leaves. Less often, white liqueur is domestically produced from the yellowish berries of M. communis var. leucocarpa DC. Of these liqueurs, only the purple-red has been recognized with geographical designations, including the European Union “spirit drinks with geographical designations” [5] and the local Sardinian IGT [6]. At the moment, no liqueur from M. communis var. leucocarpa DC berries is commercially available, and several producers want to widen the choice of myrtle berry products through its promotion. A number of studies have been performed on the chemical composition of yellowish-white myrtle berries. The essential oil, fatty-acid composition, phenolic contents, and antioxidant activities were studied in samples from Tunisia [7]. Yellowish-white berries from Turkey were investigated for their total soluble solids, acidity, pH, tannic acid, ascorbic acid, phenolic and fatty acid composition [8]. Also, ash, crude protein, crude oil, water and alcohol soluble extracts, tartaric, malic and citric acids and minerals were studied in yellowish-white Turkish berries [9]. To the best of our knowledge, no studies on commercial liqueur from yellowish-white myrtle berries have been published so far. In the present study, the chemical composition of white myrtle berry liqueur was investigated, and results were compared with the macerate obtained from raw berries, before their final dilution with water and sucrose. The total phenolic (TP) content was measured with a modified Folin-Ciocalteu’s method and phenolic compounds were identified and dosed by LC-MS/MS and HPLC-DAD. The volatiles were identified by GC-MS/FID after HS-SPME or LLE. In addition, the antioxidant capacity of both macerate and liqueur was evaluated by FRAP, CUPRAC, DPPH•, and ABTS•+ assays.

Material and Methods

A more detailed description of the material and methods can be found as supplementary material.

Samples

White myrtle berries (three batches of 90 kg) from wild growing plants were randomly harvested in December 2015 in Southern Sardinia (Monte Arcosu, Uta, CA, Italy) by professional pickers. The specimens were identified by Prof. Andrea Maxia (University of Cagliari, Italy) according to Pignatti [10] and Conti et al. [11] and different references for the right taxonomic status of taxa [12,13,14]. Voucher number DISVA.ALI.04.2015 was deposited at the Department of Life and Environmental Sciences of the University of Cagliari (Italy). After collection, the berries were cleaned and each batch was separately placed in an ethanol-water mixture and left for four months. The macerates were separated from the berries in April 2016, and the liqueurs were produced by adding sucrose and water to obtain a final percentage of 28% v/v (alcohol) and 32% w/v (sugar). Before bottling, the liqueur was filtered through IF350 cellulose acetate cardboard filter (Industrialfiltro srl, Cologno Monzese, MI, Italy).

Results and Discussion

Table 1 reports CIE chromaticity coordinates of the samples. The values describe a yellowish-amber product, which becomes darker from the macerate to final liqueur (lower L* and h°ab, higher C* ab). This darkening could be a consequence of polyphenol oxidation, by polyphenol oxidase and other enzymes that create melanins and benzoquinones from natural phenols, resulting in a brown colour [15], or the consequence of non-enzymatic browning related to pigments degradation (e.g., chlorophylls) [16].

Table 1 CIE L*C*abh°ab chromaticity coordinates and in vitro antioxidant capacity of myrtle macerate and liqueur
Full size table

Total phenolic (TP) content dosed by Folin-Ciocalteu’s assay and phenolic compounds investigated by HPLC-MS and dosed by HPLC-DAD are reported in Table 2. TP content was ca. 2–3 folds lower than the average in purple berry liqueur [4]. This confirms previous observations regarding lower amounts of phenolic compounds in yellowish-white myrtle berries compared to the purple-red ones [7]. The TP content of this liqueur is comparable to that of strawberry, raspberry, and blackcurrant liqueurs [17] and to that of bitter herbal liqueurs [18], whereas it was more concentrated than in cherry and cranberry liqueurs [19]. HPLC-DAD was applied to investigate the phenolic fraction and LC-MS/MS analysis was used to confirm peaks’ attribution (Table 1). It can be observed that all detected polyphenols were also found in the purple berries as has been previously reported [3, 4]. Large amounts of hydroxybenzoic acids were found (408.2 ± 19.9 mg/L), mainly gallic and ellagic acids (294.2 ± 14.2 and 55.8 ± 2.6 mg/L, respectively). Myrtle liqueur also contained large amounts of flavonols, mainly myricetin-3-O-galactoside and myricetin-3-O-ramnoside. Recently, these compounds have been recognized as effective in limiting postprandial hyperglycemia, which is typical of the type 2 diabetes mellitus [20]. Myrtle macerate contained traces of malvidin-3-O-glucoside that were not detectable in the liqueur, confirming the observations of other authors regarding the small amounts of anthocyanins found in white-yellow berries [7]. This finding could be visually verified by observing small dark stains present on the skin of ripened white berries.

Table 2 Phenolic composition of myrtle white berries macerate and liqueur (mg/L)
Full size table

All results of antioxidant activity assessed with FRAP, CUPRAC, DPPH•, and ABTS•+ assays are reported in Table 1. The antioxidant capacities (total and antiradical) were comparable to those of purple myrtle berry liqueur, despite the lower amount of total phenols [4]. This can be explained by high concentrations of gallic and ellagic acids, as it is well-known that these acids possess the highest antioxidant activity among phenolic compounds [21,22,23]. In particular, gallic acid has been strongly correlated with the antioxidant activity of blackcurrant liqueurs (r = 1.00), unlike the TP content, which correlation was weak (r = 0.40) [24]. Besides, myricetin and its glycosides (myricetin-3-O-galactoside and myricetin-3-O-rhamnoside) and gallic acid derivates strongly inhibit free radical and lipid peroxidation [25, 26]. The macerate’s antioxidant activity was similar to that of the liqueur, although it exhibited different composition of phenolics: the macerate composition was richer in ellagic and in gallic acid derivatives, while the liqueur contained larger amounts of gallic acid. It can therefore be assumed that degradation of both ellagic and gallic acid derivatives occurred during the processing of the macerate to obtain the liqueur, leading to the increased gallic acid concentration. The antioxidant capacity of the white myrtle liqueur proved to be higher than that of walnut [27], cherry, and raspberry liqueurs [19], reaching levels similar to those of strawberry and blackcurrant liqueur [17].

Table 3 reports the volatiles composition of the myrtle macerate and liqueur determined by GC-MS/FID analyses, assessed after two different extractions (HS-SPME and LLE). The analyses allowed us to highlight several differences between the two samples, which were probably caused by the manufacturing techniques. The major headspace compounds of the macerate were the monoterpenes α-pinene (38.5%), limonene (21.3%), 1,8-cineole (16.7%) and trans-caryophyllene (9.7%). Beyond them, the importance of the presence of limonene in other ethanolic extracts has been highlighted because of its functional properties as an antioxidant [28]. On the other hand, higher aliphatic compounds (such as ethyl palmitate (21.7%), linoleic acid (12.3%) and ethyl linoleate (32.5%)) dominated in the macerate extract. Monoterpenes were less abundant in the liquid-liquid extract in comparison with the macerate headspace, and the major ones were 1,8-cineole (7.1%), α-pinene (0.9%), trans-caryophyllene (1.2%) and limonene (0.5%). The headspace of the liqueur contained 1,8-cineole (26.5%) and linalool (23.3%) as the major compounds. Terpinen-4-ol, α-terpineol and trans-anethole appeared only in the liqueur headspace. Monoterpenes were present among the minor constituents of the liqueur extract; the major ones were linalool (4.0%), 1,8-cineole (9.6%) and α-terpineol (4.4%), in very different percentages from purple-red myrtle [29]. (Z)-Octadec-9-en-1-ol (12.0%) was the most abundant among the higher aliphatic compounds. However, shikimic acid pathway derivatives were found in the liqueur extracts, such as: 4-hydroxybenzyl alcohol (10.8%), ethyl 4-hydroxybenzoate (8.4%), 4-hydroxybenzoic acid (2.0%), vanillic acid (2.4%) and ethyl vanillate (1.2%). Variability in the volatile composition from macerate to liqueur can be explained by long storage in the hydro-alcoholic solution. During four months of extraction, these compounds could have been extracted from harder parts as the seeds and different reactions could have occurred. Moreover, dilution of the macerate, and the addition of sucrose, could have modified the native chemical composition. The similarities between the macerate and liqueur headspace composition and the essential oil (EO) of M. communis var. leucocarpa DC plant was expected, as was an abundance of 1,8-cineole, α-pinene and limonene [30]. It should be noticed that in the referenced study, large amounts of myrtenyl acetate was found in the EOs from areal parts of M. communis var. leucocarpa DC, but this compound was not detected in the macerate and liqueur from white berries. Such differences could be related to the different parts of myrtle used (leaves or berries), or to the existence of different chemotypes of M. communis var. leucocarpa DC. This last observation also raises the problem that the actual botanical classification of the variety of M. communis with white berries is based solely on the colour of berries [14]. A more appropriate morphological and genetic investigation of M. communis var. leucocarpa DC is needed.

Table 3 The volatiles composition of the myrtle macerate and liqueur determined by GC-MS/FID analysis after HS-SPME (headspace solid-phase microextraction) and LLE (liquid-liquid extraction)
Full size table

Conclusions

This research represents the first investigation on liqueur from white myrtle berries, a beverage that remains almost unexplored and unexploited. The preliminary characterization of this product revealed an interesting content of phenolic compounds and good antioxidant activity, higher than that of purple myrtle berry liqueur. The volatile fraction is similar to purple myrtle berry liqueur, and is rich in oxygenated compounds that contribute to the pleasant flavour of this liqueur. This research could well help producers to both protect and improve this typical product, obtaining a legal recognition that, at present, is still limited to the purple berry liqueur.