Comparison of physicochemical properties, amino acids, mineral elements, total phenolic compounds, and antioxidant capacity of Cuban fruit and rice wines

Abstract Physicochemical characterization, amino acids contents, minerals composition, total phenolic compounds, and antioxidant capacity of Cuban wines from different raw materials were studied. The wines studied were grape wines, tropical fruit wines, and rice wines. Twenty‐one amino acids were identified and quantified, being Asp and Glu detected in all wines. The highest concentration of total amino acid content was found in wines elaborated from Cimarrona grape subjected to maceration with grape skins, while the raisined mixture grape wine presented the lowest values, probably caused by the amino acid degradation during the dehydration process by sun exposure. Minerals quantified were range amount limits of acceptable according to the OIV recommendation. Total phenolic compounds and antioxidant capacity showed the greatest values in wine from roasting rice. No statistical separation could be clearly observed by multivariate principal component analysis; however, 3 wine groups could be defined taking account the scores on the PC1.

be transferred into the wine, reducing the risks of certain diseases (Amidžić et al., 2016;Pantelic et al., 2014;Shahidi, 2009;Shen et al., 2010Shen et al., , 2011De Souza et al., 2018;Zhong et al., 2020). Among the different bioactive compounds studied in these type wines, phenolic compounds present an important antioxidant capacity by different ways: reducing agents, hydrogen donators, free radical scavengers, singlet oxygen quenchers, and so forth (Mundaragi & Thangadurai, 2018;De Souza et al., 2018). Also, daily consumption of wine in moderate quantities contributes significantly to the requirements of human organism for essential mineral elements. But, special attention must be given to trace elements for their potential toxicity (OIV, 2020;Plotka-Wasylkaa et al., 2017;Zuñiga et al., 2014).
Moreover, amino acids are important both as essential components of proteins and for their roles in energetic metabolism, neurotransmission, and lipid transport (Shen et al., 2010). Also, are precursors for aroma compounds and directly contribute to wine's aroma, taste, and appearance. For wines elaborated with different raw materials, the amino acid profiles vary according to origin, wine-making technique, and variety. In this sense, the studies about amino acid composition may be considered as a useful tool to ensure the authenticity of fruit and rice wines (Shen et al., 2010;Valero et al., 2003).
For a long time, Cuba has handcrafted fermented beverages from a wide variety of raw materials such as tropical fruits, pumpkin, grape, corn, and rice. In recent years, and mainly due to the creation of the National Coordinator of Wine Clubs of Cuba (NCWCC), Cuba has become one of the countries that produce wines from a greater range of raw materials reaching a great international interest. Although the number of publications about fruit wines and rice wines has increased in recent years (Chay et al., 2020;Kosseva et al., 2017;Płotka-Wasylkaa et al., 2018;Qian et al., 2019;Seixas et al., 2020;De Souza et al., 2018;Sun et al., 2020;Velic et al., 2018;Wu, Xu, Long, Zhang, et al., 2015;Zhong et al., 2020), no investigation has been conducted to appraise health-related major components of these Cuban wines. This study, therefore, evaluates the bioactive-nutritional compounds (mineral elements, amino acids, and total polyphenols content), antioxidant activities (DPPH and ABTS), and the correlation between the selected factors and the samples of Cuban wines with the objective to highlight these unknown wines for production of healthier wine options.

| Physicochemical properties of fruit wines
The pH, total acidity, volatile acidity, ethanol content, and absorbances at 420 and 520 nm were determined by OIV methods (2020).

| Determination of amino acids
Amino acids were quantified according to Gómez-Alonso et al., (2007) and by Garde-Cerdán et al., (2014) with slight modifications. In a 10 ml screw-cap tube was introduced 1.75 ml of borate buffer 1 M (pH 9), 0.75 ml of methanol, 1 ml of sample (wine, previously centrifuged), 20 µl of internal standard (L-2-aminoadipic acid at 1 g L −1 ), and 30 µl of diethyl ethoxymethylenemalonate (DEEMM) and subsequently were mixed. Tubes were incubated in an ultrasound bath over 30 min and heated at 70°C in a constant temperature heater over 2 hr. The samples were cooled at room temperature for 15 min, filtered using 0.45 µm PVDF syringe filters and introduced into 2 ml screw amber vials. The injection was made immediately after the filtration of the samples. The injected volume of derivatized samples was 50 µl for wines.
These compounds were identified according to the retention times and UV-Vis spectral characteristics of the derivatives of the corresponding standards. Quantification was done using the calibration graphs of the respective standards with the same process of derivatization as the samples. The analysis was carried out in triplicate.

| Determination of mineral contents
The mineral elements were determined using an atomic absorption spectrophotometer Shimadzu AA-6800, (Shimadzu corporation, Japan) according to international methods of wine and must anal-

| Determination of total phenolic content (TPC)
Total phenolic content was performed according to the adapted

| Determination of antioxidant capacity (AC) by DPPH assay
This was performed according to the method of Katalinic Antioxidant capacity was expressed as milligrams of Trolox equivalents per liter (mg TE L −1 ) using to calculate the inhibition percentage.

| Statistical analysis
Multiple comparisons between means were performed by one-way analysis of variance test (ANOVA). Homogeneous groups were calculated in order to establish significant differences between means at p < .05. Simple linear correlation was applied to check the relationship between ABTS and DPPH values. TPC and TAAC. In addition, principal component analysis (PCA) was used to identify the specific parameters most accurately reflecting the differences between wines. The software used was the Statgraphics Centurion (v.XVI StatPoint Technologies, Inc). Table 1 showed the physicochemical properties of Cuban wines elaborated from different raw materials. The pH values of the wines ranged from 3.10 (G5) to 3.86 (TF3), optimal values to maintain wine stability and preventing them from taking place oxidative and browning reactions. Moreover, in rice wines, the pH values have a great influence on the growth and propagation of microzymes, the activity of enzyme, and the decomposition of some nutrients in fermented mash or the decomposition of yeast mesostate (Wei, 2005).

| Physicochemical properties
In the current study, the pH values for rice wines were lower those reported in Chinese rice wines by Liu et al., (2007) and Chen and Xu (2012) Color is one of the main parameters of the quality of wines and has an important influence on the overall acceptability by consumers. Traditionally, the color of wines is evaluated by measuring wine absorbance at 420 nm (yellow) and 520 nm (red

| Amino acid composition
As shown in Table 2, twenty-one amino acids were identified and quantified in the studied Cuban wines; however, not all the amino acids are found in all wines. In this sense, aspartic acid (Asp) and glutamic acid (Glu) were detected in all wines, followed by asparagine (Asn), serine (Ser), arginine (Arg), alanine (Ala), and aminobutyric acid (GABA) detected in eight of the ten wines studied. The total amino acid content (TAAC) of grape wines ranged from 287 mg L −1 (G5) to 6,233 mg L −1 (G2) and showed significant differences at p < .05. The low value in G5 wine may be caused by the amino acid degradation during the process of grape dehydration by sun exposure; similar results were reported by Pereira et al., (2008) for Madeira fortified wines. In the remaining grape wines, the TAAC was higher than some reported values which ranged between 350 and 1,000 mg L −1 (Gutierrez-Gamboa et al., 2020;Hernandez-Orte et al., 2006). In general, the most abundant amino acids found in the current study for grape wines were GABA, Pro, and Tyr, accounting for 72.9% (G1), 94.3% (G2), 58% (G3), and 57% (G4) of total amino acids. Contrary, for G5 wine these amino acids were not detected. TAAC of the tropical fruit wines varied from 381 mg L −1 (TF2) to 1795 mg·L −1 (TF1) and showed significant differences at p < .05. The lowest TAAC was shown in Papaya wine (TF2) while "frutilla" wine had the highest content. Lee et al., (2011) reported that fresh papayas have relatively low amino acid content as compared to grape and other tropical fruits. Moreover, the amino acid profile varies significantly across the different fruits and may be influenced by a large variety of factors including variety, edaphoclimatic conditions, fertilization, management, and fermentation conditions (Clark et al., 1992;Gutierrez-Gamboa et al., 2020). Except for papaya wine, Pro and Tyr were the most abundant amino acids in fruits wines, representing between 80.0% and 85% of total amino acid content. Regarding rice wines, it could be found that TAAC in R1 wine (620 mg L −1 ) was significantly (p < .05) higher than obtained for R2 wine (418 mg L −1 ).
The decrease might be due to deamination and decarboxylation reactions during the roasting process (Feuillat & Charpentier, 1982).
Besides, these contents were lower than 1.0 to 5.0 g L −1 , reported in studies about Chinese rice wines (Cao et al., 2010;He et al., 2013;Shen et al., 2011). GABA and Tyr were found to be the most abundant amino acids in R1 wine, accounting 42.7% of the TAAC. In contrast, Asp and Glu were the major amino acids in R2 wine accounted for 73.4% of the TAAC.
Overall, Tyr was found to be the highest in G2, TF1, and TF3 wines; this amino acid has been known to play a role in the wine taste (acerbity) if it presents at high levels (Cao et al., 2010). Besides, Pro was the most abundant amino acid found in G3 and G4, consistent wine. This amino acid, catalyzed from glutamic acid by glutamic acid decarboxylase, is the most important inhibiting neurotransmitter in the brain with numerous health effects (Joye et al., 2011;. Likewise, Giovanni et al., (2015) reported high levels of Asn and GABA amino acids in Italian wines. Table 3 showed the mineral composition of the wines studied. A total of seven mineral elements (Ca, Mg, Fe, Zn, Pb, Cu, and Cd) were determined and the total mineral contents (TMC) were calculated.

| Mineral composition
As can be observed, the TMC ranged from 162 mg L −1 in TF3 wine (raisined fruits) to 424 mg L −1 in G5 wine (botrytised and raisined grapes) and exhibited significant differences (p < .05) for most of the Cuban wines them. Ca and Mg are defined as macroelements and in the current study were found to be the most abundant minerals, representing more than 95% of the TMC, in all wines (Paneque et al., 2010;Shen et al., 2013). Fe and Zn were among the minor mineral present, while Pb, Cu, and Cd were below the detection limits in all wine types and, therefore, below the maximum acceptable limits for these elements reported by OIV (OIV, 2020). These results in the current study are important because these last-mentioned metals can become highly toxic if they accumulate in biological system (Rupasinghe & Clegg, 2007). The TMC of different grape wines varied from 170 mg L −1 (G1) to 424 mg L −1 (G5), and there were no significant differences between G1 and G4 wines. The TMC ranged to 162 mg L −1 (TF3) to 308 mg L −1 (TF2) in tropical fruit wines showed significant differences among all wines (p < .05). For rice wines, TMC was significant different varied from 176 mg L −1 (R1) to 246 mg L −1 (R2), lower than that reported by Qian et al., (2019) in Chinese rice wines.
Regarding the macroelements minerals individually, Ca levels (369 mg L −1 ) were the highest in G5 wine (botrytised and raisined Aramon and Cimarrona grapes) and were significantly higher (p < .05) than in all other studied wines. Contrary, Ca content was the lower in G1, G4, TF3, and R2 wines without significant differences among them. In addition, Mg contents were the highest in the previous wine mentioned R2 (raw rice and roasted rice, 150 mg L −1 ), exhibiting significant differences with the wines in the current study (p < .05).
Moreover, this rice wine was the only one that showed a higher content of Mg than Ca. This is in accordance with the results pointed by Shen et al., (2013), for Chinese rice wines. On the contrary, this mineral presented the lowest content (53.0 mg L −1 ) in G5 wine. In the current study, the concentration of microelements Fe and Zn ranged between 0.55 (R1)-10.2 mg L −1 (G3) and 0.13 (G5)-1.21 mg L −1 (G2), respectively. It is known that Fe, Zn, and Cu are closely related to the winemaking process (Rupasinghe & Clegg, 2007;Shen et al., 2013), and this fact in addition with the different raw material used in the elaboration of Cuban wines, could explain the variability of the concentrations of these elements.

| Total phenolics content and antioxidant capacity
The TPC of different Cuban wines was between 200 mg GAE L −1 in TF1, TF2 and R1 (p < .05) to 2,250 mg GAE L −1 in R2 (Table 4). This wide range is due to concentration and composition of the phenolic compounds present in wines depends largely on the source of raw material and the winemaking processes. Based on the TPC and according to Rupasinghe and Clegg (2007), studied wines can be categorized into three major groups: high TPC (R2 wine, 2,250 mg GAE L −1 ); moderately high TPC (G3 and G4 wines, ranged from 817 to   Pantelic et al., 2014;Rupasinghe & Clegg, 2007). However, they were close to the values reported by Rupasinghe and Clegg (2007) for pear, peach, and apple wines, which were between 310 and 451 mg GAE L −1 . The TPC of TF3 sample was significantly higher (p < .05) than the other tropical fruit wines due to the enrichment with grape must during its elaboration.

TA B L E 3 Minerals contents (mg L −1 ) in Cuban wines
The highest values of TPC were found in wines R2, G4, and G3 (2,250, 904, and 817 mg GAE·L −1 , respectively), which were elaborated using fungus during their winemaking. Some authors observed a decrease in total phenolic compounds of botrytized grapes compared with healthy grapes; however, they reported an increase in flavan-3-ol family (Carbajal-Ida et al., 2016). The studied wines were elaborated from different raw materials and different winemaking procedures, for example, the use of chips, so the differences in the values of total phenolic compounds could be due to these facts. To establish a correlation between antioxidant capacity and the bioactive composition, the correlation coefficients were determined (Table 5). The antioxidant capacity measured by DPPH assay expresses a low correlation with TPC (R = 0.3473) and a moderate positive correlation was found for the values obtained by ABTS assay (R = 0.6200) correlated with phenolic compounds. In this sense, some authors found a high correlation between the phenolic content and the antioxidant capacity (Kuskoski et al., 2005;Reddy et al., 2010), while others found no relationship (Imeh & Khokhar, 2002;Thaipong et al., 2006). The results suggest that the phenolic compounds may be not the unique contributors to the antioxidant capacity in the Cuban wine studied in this research. In the other hand, the correlation of total amino acid content with the antioxidant capacity values showed a low correlation. In all case analyzed, the correlation was not significant at p < .05 level. Therefore, the antioxidant capacity in Cuban wines can be attributed to the actions of different antioxidant compounds, such as phenolic compounds, amino acids, or bioactive peptides among other substances such as isothiocyanates, vitamins, and flavonoids. whose effect can be antagonistic or synergistic.
To identify the specific parameters most accurately reflecting the differences between wines, the TPC (total phenolic content), TAAC (total amino acid contents), and TMC (total mineral contents) data for the different Cuban wines were subjected to multivariate principal component analysis. Figure 1 shows

| CON CLUS ION
In summary, Cuban wines elaborated from different raw materials as follows: grapes, tropical fruits, and rice and through different winemaking processes were found to be a source of essential amino acids, minerals, and phenolic compounds. The significantly highest values of total amino acid content and total mineral content were displayed in G2 wine and G5 wine, respectively. Ca and Mg were the principal mineral elements quantified in all wines. Excepted R2 wine, Ca presented higher levels than Mg. Regarding the total phenolic content, rice wines were significantly highest. Among all wines, grape wines (G1, G3, and G4) and R2 wine showed the highest value of antioxidant capacity. The multivariate principal component analysis carried out allowed define three wine groups influenced by PC1, in this variable total mineral content and total amino acid content had the highest statistical weight. Therefore, the results of the studied wines must be taken into consideration to improve the winemaking processes and nutritional quality of the Cuban wines. Cuba), for providing samples.

CO N FLI C T O F I NTE R E S T
The author declares that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.