New Advances in the Phenolic Composition of Tiger Nut (Cyperus esculentus L.) by-Products

“Horchata” is a well-known Spanish beverage obtained from pressing tiger nuts. Its by-product is a potential source of sugar and fiber but also contains polyphenols; thus, it could be used as a new ingredient in the food industry. The aim of this work is to determine the phenolic compounds and compare the phenolic profile of two tiger nut by-products. A Box–Behnken design has been carried out to optimize the extraction of phenolic compounds from tiger nut by-products by ultrasound technology. The independent factors were time (min), ethanol/water (% v/v), and solvent/sample ratio (v/w). The model was validated and confirmed by ANOVA. A Protected Designation of Origin (PDO) of Valencia and a non-Protected Designation of Origin (n-PDO) tiger nut by-products were extracted under the optimal conditions and were characterized by HPLC-DAD-ESI-TOF-MS (High Performance Liquid Chropatography coupled to a photodiode array time-of-flight mass detector). Moreover, their antioxidant activities measured by three different methods (DPPH (2,2-diphenyl-1-picrylhydrazyl), ABTS (2,2′-Azinobis [3-ethylbenzothiazoline-6-sulfonic acid]-diammonium salt) and FRAP (ferric reducing antioxidant power)) were compared. A total of 45 polar compounds were identified, and the phenolic ones were quantified, some of them for the first time. PDO tiger nut by-product has been demonstrated to be richer in phenolic acids and other polyphenols and has higher antioxidant activity; meanwhile, n-PDO tiger nut by-product is richer in phenol precursors.


Introduction
Tiger nut (Cyperus esculentus L.) is a tuber mainly used to obtain a tiger nut beverage with a milky appearance, called "horchata". It is typically from Spain, where the horchata production industry supposes €60 million per year. The by-products generated from this industry are up to 60% of the tiger nut used, which equals 1.8 million kg per year. Tiger nut by-products are mainly used for animal feed or as organic matter for combustion [1], as a carbon source for the growth of probiotic bacteria [2], or as a sugar source for microalgae [3]. Besides, there are several studies that have evaluated the potential of this by-product as a source of fiber to enrich meat products such as pork [4], beef [5] burgers, and pork sausages [6], to enhance cooking performance, moisture and fat retention. However, there are only a few studies concerning the potential use of this by-product for food and nutraceutical industries. Some studies have evaluated enzyme pretreatments combined with high pressure [7] or different mixtures of solvents at different temperatures [8] to extract phenolic compounds from tiger nut by-product. Nevertheless, they only used spectrophotometric methods to quantify total phenolic compounds, but individual components were not quantified nor chemically characterized. More recent studies have characterized the phenolic profile of the tiger nut by-product extracted by conventional extraction [9] or of the tiger nut by-product oil by supercritical fluid extraction with CO 2 [10]. Clemente-Villalba et al. (2021) [11] have recently published a comparison of the sensory profile, volatile composition, and consumer's acceptance of PDO and n-PDO tiger nut milks. Razola Díaz et al. (2020) [3] compared the chemical composition and sugar content in PDO and n-PDO tiger nut by-products. Thus, the objective of the present study is to compare, for the first time, the phenolic profiles of two tiger nut by-products from different origins (PDO and n-PDO) that were obtained during production at an industrial scale. The extraction of phenolic compounds was optimized using ultrasound technology with a Box-Behnken design. All the polar compounds were tentatively identified by HPLC-ESI-TOF-MS, and the phenolics were quantified. The antioxidant activity of both by-products was measured by three different methods (DPPH, ABTS, and FRAP).

Chemicals and Samples
HPLC (High Performance Liquid Chromatography)-grade water and other reagents and solvents were purchased from Merck KGaA (Darmstadt, Germany).
Tiger nut by-product samples were provided by Puleva company located in Granada, Spain, in May 2019. Two types of samples were collected, one from Protected Designation of Origin (PDO) of Valencia and another one without appellation of origin (non-Protected Designation of Origin (n-PDO)) from Ivory Coast. The PDO by-product was obtained after three successive pressings of the tiger nut, while the n-PDO by-product was obtained after only two presses due to its original composition and the final target product of the company as the non-PDO tiger nut was richer in fat. The samples were dried in a ventilated oven at 40 • C until obtaining 5.57 ± 0.29 and 5.46 ± 0.67 % of humidity in the PDO and n-PDO tiger nut by-product, respectively, and milled and sieved to 100 µm; after that, they were frozen at −18 • C until the analyses.

Experimental Design
A Box-Behnken design composed of 15 experiments in three levels (−1,0,1) was established to optimize the extraction of phenolic compounds from tiger nut by-products. All the parameters of the model were established according to the previous experience of the group and previous trials. The independent variables were time (5, 45 and 85 min), ratio ethanol/water (0, 50 and 100% v/v), and ratio solvent/sample (20,40 and 100 v/w), and the response was the total phenolic content measured using the Folin-Ciocalteu method. The model was fitted to a second-order polynomial equation, and its adjustment was evaluated and confirmed by ANOVA. To select the optimal conditions, response surface methodology (RSM) was used. All the data was processed using STATISTICA 7.0 (2002, StatSoft, Tulsa, OK, USA).

Ultrasound Bath Extraction
Briefly, 0.125 g of tiger nut by-product powder was dissolved in a 10 mL solution of ethanol/water 40/60, v/v. The mixture was placed in an ultrasonic bath for 50 min, and then it was centrifuged for 10 min at 9000 rpm. The supernatant was collected, evaporated, and reconstituted in 1 mL of methanol/water (50:50, v/v). The final extracts were filtered with regenerated cellulose filters 0.2 µm (Millipore, Bedford, MA, USA) and stored at −18 • C until the analyses.

Determination of Total Phenolic Content (TPC)
Folin-Ciocalteu spectrophotometric method was used to determine the TPC in tiger nut by-product [12]. Thus, 100 µL of the extract was added to 500 µL of the Folin-Ciocalteu reagents and 6 mL of bi-distilled water. The flask was agitated for a minute. After that, 2 mL of 15% (w/v) Na 2 CO 3 and made up to 10 mL with bi-distilled water. The flasks were kept in darkness. After 2 h, the measures were carried out at 750 nm and 25 • C with a UV-visible spectrophotometer (Spectrophotometer 300 Array, UV-Vis, single beam, Shimadzu, Duisburg, Germany). The calibration curve was developed with Gallic acid from 1 to 100 ppm, and the obtained equation was y = 0.0012x + 0.0227 (R 2 = 0.9995). Results are expressed as µg gallic acid equivalents (GAE)/g dry weight (d.w.).

Determination of Polar Compounds by HPLC-ESI-QTOF-MS
The phenolic profile characterization and quantification of the two tiger nut byproducts extracted by the optimal conditions were performed according to a previously described method [13]. The analyses were carried out in duplicate on an ACQUITY Ultra Performance LC system (Waters Corporation, Milford, MA, USA) coupled to an electrospray ionization (ESI) source operating in the negative mode and a time-of-flight (TOF) mass detector (Waters Corporation, Milford, MA, USA

Antioxidant Assays in Tiger Nut By-Products
The antioxidant capacity in the two tiger nut by-products has been evaluated in the extract obtained by the optimal ultrasound bath conditions using three different methods.
The ABTS method was carried out according to Re et al. (1999) [14]. The monocation ABTS •+ is generated by oxidation of the ABTS with potassium persulfate in the dark at room temperature for 12-24 h. For each extract, 1 mL of this ABTS solution was added to 0.01 mL of the extract and the detriment of absorbance during 30 min at 734 nm was measured.
The DPPH radical scavenging activity was assayed with a method proposed by several authors [15,16]. In total, 100 µL of each extract was added to 2.9 mL of DPPH, and after rapid stirring, the bleaching power of the extract was observed in a time interval from 0 to 30 min at 517 nm.
The FRAP assay was carried out following the procedure developed by Pulido et al. (2000) [17]. It is based on the reduction of Fe 3+ to Fe 2+ by the antioxidant substances. A total of 30 µL of each extract was added to 90 µL of distilled water and 900 µL of the FRAP reagent. It was kept for 30 min at 37 • C and measured in the spectrophotometer at 595 nm.

Fitting the Model
The extraction step is the most important to reach the highest amount of the target compounds, in this case, phenolic compounds. A Box-Behnken design coupled to RSM was used to find the optimal conditions of time (X 1 ), ethanol/water ratio (X 2 ), and solvent/sample ratio (X 3 ) to extract phenolic compounds from the tiger nut by-product using ultrasound bath technology. The experimental values of TPC obtained for each run are shown in Table 1.
The lowest recovery (42.80 ± 1.08 µg GAE/g d.w.) was at 45 min, with the ethanol/water 100% and ratio solvent/sample 20 v/w, and the highest (383.11 ± 1.76 µg GAE/g d.w.) at 85 min, was ethanol/water 50% and ratio solvent/sample 100 v/w. This by-product has a low relation weight/volume (<1); therefore, those experiments carried out with the lower value of ratio solvent/sample led to the lowest results due to the saturation of the solvent and consequently the reduced contact surface between sample and solvent. Similar results were found when the tendency using ethanol was 100%; therefore, it demonstrates that some water is needed as surfactant for the extraction of phenolic compounds to satisfactorily take place. Table 1. Box-Behnken design with natural and coded values (parenthesis) of the conditions of extraction and the experimental results obtained for total phenolic content (TPC) expressed with the average and the standard deviation.

Run
Independent Factors Response X 1 X 2 X 3 TPC (µg GAE/g d.w.) The experimental data were adjusted to a second-order polynomial equation, and all the estimated regression effects are shown in Table 2. The model was analyzed with a significance level of p < 0.05, and all the lineal terms (β 1 , β 2 , and β 3 ), all the quadratic terms (β 11 , β 22 and β 33 ), and the crossed values between time (X 1 ) and ratio solvent/sample (X 3 ), had significance, but the rest of the terms were discarded. Therefore, the model was recalculated, and the ANOVA test was performed. As shown in Table 2 the model revealed a high correlation coefficient (R 2 = 0.9891), a significant regression model (p < 0.05), and a non-significant lack of fit (p > 0.05). According to Bezerra et al. [18], the adequacy of the model was confirmed.
Optimal conditions were selected using RSM concerning the three-dimensional graphs presented in Figure 1. A compromise has been made between the independent factors to establish them at the minimal possible values. The optimal conditions chosen were: 50 min, 40% ethanol/water, and ratio solvent/sample 80 v/w that gave a predicted value of 403.61 ± 49.54 µg GAE/g d.w. The coefficient of variation between the obtained (400.43 ± 4.63) and predicted values was lower than 1%, so the model was validated.
The optimal result obtained was better than the one reported by Roselló-Soto et al. [8] that used conventional extraction with a similar solvent (ethanol 25%) but with differences in terms of time (3 h) and temperature (60 • C). Therefore, ultrasound technology has been demonstrated to be a non-thermal and lower time-consuming technique that allows the satisfactorily recovery of phenolic compounds from tiger nut by-products. However, the authors reported slightly higher results using 50% ethanol, 35 • C, and pH 2.5 for 3 h [9]. In our case, the pH has not been changed in order to avoid the hydrolysis of bound compounds.

Identification of Polar Compounds by HPLC-ESI-TOF-MS
The samples of PDO and n-PDO tiger nut by-products were characterized by HPLC-MS, and 44 compounds were identified, 18 were phenolic compounds and 26 other polar compounds were identified as hydroxyl fatty acids. To our knowledge, a total of 27 polar compounds were identified for the first time in tiger nut by-products.
All the identified compounds are described in Table 3 with their retention time, molecular formula, experimental and calculated m/z fragments, and score and error (ppm). To ensure mass accuracy, the tolerances chosen had a score higher than 90% and an error lower than 5 ppm (part per million) between the experimental and calculated m/z. To identify the compounds, the generated molecular formula and some in source fragments were checked and studied also comparing with different databases such as PubChem, Phenol-Explorer, and the literature.
Sinapyl alcohol and benzoic acid were identified at 11.64 and 16.13 min, corresponding with peaks 16 and 22. Sinapyl alcogol has been described as a precursor to various stilbenes and coumarins, and benzoic acid as a precursor of phenolic acids [25].
The phenolic composition, mainly based on phenolic acids, justified the optimal conditions of extraction obtained in this work. In fact, according to Waszkowiak and Gliszczynska-Swigło, the phenolic acid content decreases when ethanol concentration is higher than 60 to 70 % [30]. Similar data were reported by Roselló-Soto et al. [9], obtaining the highest recovery of phenolic acids in tiger nuts using 41.4 % ethanol.
Compared with other authors, we have achieved a higher amount of cyanidin, 4vinylphenol, ethyl vanillin, and p-coumaric acid than Roselló-Soto et al. [9]. This could be because they used a conventional extraction while we used ultrasound bath assisted extraction; however, the difference between the samples must also be considered. No other references were found in the bibliography regarding phenolic compounds in tiger nut by-products. Table 4. Phenolic compounds and their precursors quantified in PDO and n-PDO tiger nut by-product expressed as mean ± standard deviation (µg/g dry weight (d.w.)).

No.
Compound PDO by-Product (µg/g d.w.) n-PDO by-Product (µg/g d.w.) In addition, PDO and n-PDO tiger nut by-products were revealed to have antioxidant activity (Table 5), being around 15% higher for the PDO by-products with the three performed methods. Antioxidant activity has been reported previously in tiger nut oils [31] and tiger nut beverages [32]. However, no references were found in tiger nut by-products apart from those reported by Roselló-Soto et al. (2019) [9], but the method used was different (total antioxidant capacity). Table 5. Antioxidant activity of PDO and n-PDO tiger nut by-product expressed as mean ± standard deviation (µg Trolox equivalent (TE)/g dry weight (d.w.)).  [33] reported a total of 30 phenolic compounds that were previously found and quantified in tiger nut tubers. Parts of these phenols naturally remian in the horchata after the pressing steps; therefore, in the by-product, fewer compounds remain and in lower amounts. This corroborates the results obtained in this work as 18 phenolic compounds were identified and quantified in tiger nut industrial by-products. Besides, tiger nut by-products have a lower amount of phenolic compounds with lower antioxidant activity than tiger nut oils [31,34] and tiger nut by-product oils [10]. This is mainly because of the oil extraction techniques that hydrolysate the compounds. However, the use of ultrasound technology seems to be a promising technique to apply to tiger nut by-products that allow obtaining phenolic compounds with antioxidant activity. This makes it clear that the PDO tiger nut by-product contains higher amounts of phenolic compounds than the n-PDO tiger nut by-product. This result was realized as tiger nuts cultivated in Valencia (Spain) contain higher amounts of these antioxidants compared to the samples from Ivory Coast that contain more precursors. This difference could be justified with the growth differentiation balance framework. According to this hypothesis, when there are high levels of nitrogen and good environmental conditions, the growth of the plant is favored. On the contrary, if the environmental conditions are not favorable and the availability of essential components is low, the secondary metabolism is favored increasing the phenolic content. However, to confirm this trend, further analyses on the most robust sampling are needed.

Conclusions
The ultrasound-assisted extraction of phenolic compounds from tiger nut by-products has been established using a Box-Behnken design combined with RSM. The optimal selected conditions were 50 min, 40% ethanol, and ratio solvent/sample 80 v/w. The results highlighted that ultrasound technology permits a high recovery of polyphenols from tiger nut by-product. Moreover, PDO and n-PDO tiger nut by-products were compared for the first time according to their phenolic composition and antioxidant activity. Both were characterized by HPLC-MS, and a total of 45 free polar compounds were identified, from which 27 polar compounds were identified for the first time in tiger nut by-products. The predominant compounds were phenolic acids, the major ones being 2-O-galloyl-1,4galactarolactone and vanillic acid derivatives. The quality attributed to the Protected Designation of Origin of Valencia has been confirmed in the PDO tiger nut by-product as it showed higher amounts of phenolic acids and other polyphenols also exhibiting higher antioxidant activity than the n-PDO tiger nut by-product. Further research, with a more robust sampling, could confirm the added value of Protected Designation of Origin of Valencia tiger nut by-products compared to n-PDO options.
To conclude, taking into account that the use of tiger nut flour for gluten-free products was studied, tiger nut by-products could be used as cheaper ingredients/flours for the formulation of bakery products.