elocation-id: e3254
Roselle (Hibiscus sabdariffa L.) calyces have abundant phenolic compounds to which multiple therapeutic benefits are attributed. The study aimed to develop a nutraceutical beverage from Hs extracts of genotypes with different pigmentation and phytochemical composition, with sensory characteristics accepted by the consumer and with a high content of antioxidants. Calyces of three cultivars with intense red, light red, and non-pigmentation were used to prepare aqueous extracts, which were mixed in different proportions to achieve beverages with a standardized anthocyanin content (TAC). Four beverage formulations, denoted as B1, B2, B3, and B4, were obtained, which were subjected to global acceptance tests through different sensory variables. The variables determined in the beverages were color, pH, titratable acidity (TA), total soluble phenols (TSPs), TAC, proanthocyanidins (PAs), and antioxidant capacity (AC). The TAC in the beverages ranged from 23 to 24.9 mg anthocyanin per 240 ml portion. The color and TA of the beverages were statistically different according to the objective measurement, but sensory differences were not detected. Beverages B1 and B4, formulated with mixtures of the three cultivars, had the highest TSP content and AC and had the best aroma and flavor. Incorporating pigment-free calyx extracts in the formulation of roselle beverages improves their sensory characteristics and antioxidant content.
Hibiscus sabdariffa L., antioxidant capacity, calyces, sensory analysis.
Traditional Mexican medicine attributes therapeutic properties to extracts of roselle [Hibiscus sabdariffa L. (Hs)] calyces in conditions such as high blood pressure, kidney stones, and inflammation (González-Stuart, 2011). These properties are associated with calyx polyphenols, among which anthocyanins, phenolic acids, flavonols, and proanthocyanidins predominate (Riaz and Chopra, 2018), the main biological activity related to these compounds is their ability to neutralize or capture free radicals and reduce their oxidative action on cellular components.
Studies on the effectiveness of aqueous roselle extracts in treating conditions are numerous and have been conducted to different extents, including clinical trials (Mckay et al., 2009) and laboratory animal trials (El-Shiekh et al., 2020). In clinical studies, there is no consensus regarding the therapeutic dose of roselle-based teas or beverages. Nonetheless, the information available so far demonstrates the low toxicity of the extracts used at medium and low doses (Riaz and Chopra, 2018).
The phenolic composition of roselle genotypes’ calyces differs according to their pigmentation. Dark-colored calyces have higher anthocyanin content than light-colored calyces (Reyes-Luengas et al., 2015), and white calyces contain more flavonols than red calyces (Camelo-Méndez et al., 2016). The composition of volatile compounds, which are responsible for the aroma of calyces, also differs between genotypes, with those with light-colored calyces being less aromatic than those with dark-colored calyces (Musa et al., 2021).
Published scientific information on differences in the chemical composition of roselle calyces with different pigmentation and information related to therapeutic doses for disease prevention or control can be used to formulate nutraceutical beverages for consumers interested in obtaining health benefits from the consumption of roselle refreshing beverages. The present study aimed to develop a beverage from mixtures of roselle calyx extracts of genotypes with different pigmentation, sensory qualities accepted by consumers, and high antioxidant content.
The dehydrated roselle calyces used were from the genotypes Sudán (S) of dark red calyces, Alma Blanca (AB) of white calyces, and Criolla Nayarit (CN) of light red calyces. The first two were grown in Santa María Cortijo, Jamiltepec, in the Costa region of the state of Oaxaca, and the third in Nayarit, in the municipality of Xalisco.
Two point five grams of calyces (dry base) were weighed, and 100 ml of distilled water was added. The mixture was boiled for 15 min, performing two extraction cycles. The extracts from the first and second extraction were combined, made up to 200 ml with distilled water, and filtered with Whatman 4 paper. Different beverage formulations were made from these extracts. Information on the physical and chemical characterization of the calyces of roselle genotypes used in beverage formulations was published in a previous study (Reyes-Luengas et al., 2015).
Four formulations were prepared, which contained at least 21 mg of total anthocyanins/240 ml of beverage (Table 1), the amount of anthocyanins required per day to achieve benefits in aspects such as moderate blood pressure control (McKay et al., 2009) or hypercholesterolemia (Gurrola-Díaz et al., 2010). The sweetness level was adjusted to 7% with cane sugar. Each beverage formulation was prepared in duplicate.
pH. It was measured with a Denver Instrument UB10 potentiometer, calibrated with pH 4 and 7 buffers. Titratable acidity (TA). The AOAC (1984) method for slightly colored solutions was used. Acidity was expressed in milliequivalents of citric acid in % w/v. Color. A Hunter Lab Mini Scan XE Plus (model 45/0-L) colorimeter was used on the CIEL*a*b* scale, with D/65 illuminant, a 10° angle, and a quartz capsule into which 30 mL of the beverage was poured. The Hue and Chroma were calculated based on the parameters a* and b*. The ΔE between each of the different beverages was calculated using the expression ΔE= √ (L1-L2)2 + (a1-a2)2+(b1- b2)2 (Wrolstad and Smith, 2010).
The TAC was performed as described by Galicia-Flores et al. (2008). A standard curve of cyanidin-3-glucoside (Polyphenols, NW) with five curve points (5-25 ppm, R2= 0.9997) was generated and read on a Perkin-Elmer® lambda 25 spectrophotometer. The TAC was expressed in mg cyanidin-3-glucoside equivalents (CGE)/240 ml beverage.
TSPs were analyzed using the Folin-Ciocalteu method, as described by Galicia-Flores et al. (2008). Gallic acid was used as a standard to express the results in mg gallic acid equivalents (GAE) per 240 mL of beverage.
APs were assessed using Wallace and Giustis (2010) method. A standard curve of catechin (Sigma, MN) was prepared. Results were expressed in mg catechin equivalents (CE) per 240 ml of beverage.
A volume of 2 900 μ of a solution of DPPH (2,2-diphenyl-1-picrylhydrazyl) (Sigma-Aldrich®, Mexico), 100 μM in methanol at 80% v/v were mixed with 100 μl of beverage. After resting 30 min in darkness, absorbance was measured at 515 nm. The reference absorbance was the one obtained by mixing 100 μl of 80% methanol with 2 900 μl of DPPH. The percentage of DPPH reduced was calculated using the formula:
Where: Absr= reference absorbance of the DPPH solution; AbsT= absorbance of the formulation at 30 min (Prior et al., 2003).
To determine the IC50, dilutions of each formulated beverage were carried out based on the concentration of total soluble phenols (TSPs), and the percentage of reduced DPPH was measured. The concentration of total phenols was plotted against the percentage of reduced DPPH; then, a logarithmic regression was fitted to calculate the IC50. The IC50 is the concentration of the extract with which 50% of the DPPH is reduced.
A 20 μl aliquot of each formulation was deposited in the well of a dark 96-well flat-bottom plate along with 5 μl of buffer and 360 μl of 48 nM fluorescein (Sigma, St. Louis, MO). Samples were incubated for 30 s, and initial fluorescence was recorded. The plate was read on a microplate reader (Synergy HT, Biotekinstruments), and the reaction was initiated by adding AAPH (2-2’-Azobis amidinopropane, free radical generator) using the device’s auto dispenser (Ou et al., 2001).
Samples and control (buffer instead of extract) were analyzed in quadruplicate. The excitation and emission wavelengths were at 485 and 530 nm, respectively. Antioxidant capacity (AC) data were expressed as Trolox equivalents (TE) per 100 ml of beverage (μmol of TE 100 ml-1).
Discriminatory tests were carried out to determine differences between beverage formulations. For the global acceptability tests, a nine-point hedonic scale was used, where: 1 ‘I dislike it extremely’, 5 ‘I neither like nor dislike it’, and 9 ‘I like it extremely’. The variables evaluated were color, aroma, acidity, roselle flavor, residual flavor, sweetness, and global acceptance. The evaluation was conducted with 65 panelists.
The physical and chemical characterization data of the beverages and the attributes of the sensory evaluation were analyzed under a completely randomized design. The data from the physical and chemical characterization were used to perform a one-way analysis of variance and mean comparison tests (Tukey, p≤ 0.05) when appropriate, using the SAS® program version 9.0 (SAS Institute, Inc., 2009). Principal component (PC) analysis of sensory evaluation data was performed with the Unscrambler® program version 9.2 (Camo Process AS, Oslo, Norway). In this case, the means were compared using Duncan’s test (p≤ 0.05).
The pH differed between the beverages, with the highest corresponding to B2 and the lowest to B3 and B4, which contained a higher proportion of the extract of the CN variety, Table 2. In all cases, it was < 3, which favors the chemical form of the flavylium cation of anthocyanins (Brouillard, 1982), its stability, and its characteristic bright red color. The TA varied between 9.8 and 20.7%, with a statistical difference between the beverages: the highest TA occurred in the beverages with extract of the AB genotype (B1 and B4). The TAC of the beverages was around 24 mg CGE/240 ml of the beverage.
[i] LSD= least significant difference; TA= titratable acidity (% w/v); TAC= total anthocyanin content (mg cyanidin-3-glucoside equivalents 240 m-1 of beverage); TSPs= total soluble phenols (mg gallic acid equivalents 240 ml-1 of beverage); PAs= proanthocyanidins (mg catechin equivalents 240 ml-1 of beverage). Equal letters in columns are statistically equal (Tukey, α= 0.05).
The difference in TSPs between the beverages is related to the addition of the extract of the AB variety, with a high content of these compounds (Reyes-Luengas et al., 2015). However, all beverages kept a TSP content above 65.55 mg, the daily dose required to achieve beneficial effects (Mckay et al., 2009). Regarding PAs, the highest values were in B1 and B2, attributed to the fact that their formulation included a higher proportion of the extract of the S variety, with a higher concentration of these flavonoids than AB and CN (Reyes-Luengas et al., 2015).
Luminosity (L) was statistically different between the four beverages. The low values were related to their dark color, which was higher in B1 and lower in B4. All beverages had a purple-red hue (positive values of a*). The values of b* in the beverages were low (< 1), indicative of a slight yellow tint (Table 3).
Hue values correspond to a red-purple hue. No statistical difference was observed for this variable between the beverages analyzed. The color saturation or chroma index showed low values (2.33 - 2.86), which is associated with a high presence of gray tones.
Visual color differences between the four formulated beverages were expressed through ∆E. According to Obon et al. (2009), a difference of ∆E between 0 and 1.5 can be considered small and visually imperceptible; from 1.5 to 5, the color difference can be distinguished and is evident for a ΔE greater than 5. Therefore, it would be difficult for the consumer to notice color differences between the beverages B1 vs B2, B3 vs B2, and B3 vs B4; in contrast, the color differences between the beverages B1 vs B3, B1 vs B4, and B2 vs B4, would be easily noticed.
Antioxidant capacity, as a function of IC50 values, was statistically different between beverages. The IC50 of B2 and B3 were the same (p≤ 0.05) but higher than that of B1 and B4. A higher IC50 implies lower antioxidant power (Noreen et al., 2017) (Figure 1). The standardization of the concentration of anthocyanins in the beverages caused the IC50 values to be similar.
The lowest IC50 values corresponded to B1 and B4 beverages, which contained the extract of the AB genotype (Table 1), which is abundant in phenolic acids and other flavonoids with antioxidant capacity (Reyes-Luengas et al., 2015).
There was a statistical difference in antioxidant capacity (AC) between the four beverages (Figure 2). The ones with the highest AC were B1 and B4, with equal AC between them but statistically different from B3. In B1 and B4, the AC values were 1 080 and 1 193 μmoles TE per 100 ml of beverage, respectively.
All beverages had values higher than 621.8 ±1.7 μmol TE per 100 ml, the value reported in roselle water prepared with the traditional procedure used by housewives in Mexico (Sáyago-Ayerdi et al., 2007).
Only the aroma and sweetness attributes were different between the beverages. The beverage with the best roselle aroma was B4, which is attributed to the fact that it incorporates extracts from the three roselle cultivars; in contrast, B2 and B3 beverages had the lowest scores (Table 4), possibly due to the absence of the AB genotype extract in their formulation (Table 1). The characteristic aroma of roselle beverages is attributed to the quantitative balance of their volatile compounds and not to the predominance of any of them (Musa et al., 2021).
The panelists perceived the B2 and B3 beverages as the sweetest, even though they all contained 7% w/v of sugar, due to their lower acidity, as they contain more water than the rest. Notably, although the TA difference was more than 10 percentage points between the beverages (Table 2), the panelists did not perceive it. Bechoff et al. (2014) sensorily evaluated roselle beverages with differences in acidity up to three times without the panelists identifying them.
The evaluators did not identify color differences between the beverages evaluated either, which is attributed to the fact that the hue angle, which is the one most associated with visual color perception (Wrolstad and Smith, 2010), was equal among the beverage formulations. In this regard, Bechoff et al. (2014) pointed out a lack of correlation between the red color perceived by the panelists and the L*a*b* color parameters.
The ellipses represent 50 to 100% load correlations between the first and second (Figure 3). All the variables showed a high correlation, as they were located in the area of the two ellipses, except for the variable of color intensity (int color), which was located outside the area of the first ellipse, which means that it did not correlate with the rest of the variables. None of the variables analyzed showed opposite locations or negative weight.
Principal component 1 (PC1) was mainly explained by sweetness and acidity, attributes that are located to the right of the ellipse, which means that the sweeter and more acidic the beverage, the better it will be accepted by the panelists, a result that coincides with what was found by Bechoff et al. (2014). For principal component 2 (PC2), the representative variables were aroma and color intensity (Int Color). PC1 explained 52% of the total variation in the attributes determined in roselle beverages, while PC2 explained only 14% of this variability (Figure 3 B).
Mixing aqueous extracts of roselle (Hibiscus sabdariffa L.) genotypes with calyces of contrasting pigmentation enhances the beverage’s sensory attributes and consumer acceptance. Incorporating the extract of the non-pigmented variety (Alma Blanca) into the mixture favors the beverage’s flavor and aroma and increases its antioxidant capacity.
Camelo-Méndez, G. A.; Jara-Palacios, M. J.; Escudero-Gilete, M. L.; Gordillo, B.; Hernanz, D.; Paredes-López, O.; Vanegas-Espinoza, P. E.; Del Villar-Martínez, A. D. and Heredia, F. J. 2016. Comparative study of phenolic profile, antioxidant capacity, and color-composition relation of roselle cultivars with contrasting pigmentation. Plant Foods Hum. Nutr. 71(1):109-114.
Gurrola-Díaz, C. M.; García-López, P. M.; Sánchez-Enríquez, S; Troyo-Sanromán, R; Andrade-González, I. and Gómez-Leyva. J. F. 2010. Effects of Hibiscus sabdariffa extract powder and preventive treatment (diet) on the lipid profiles of patients with metabolic syndrome (MeSy). Phytomedicine. 17(7):500-505.
Prior, R. L.; Hoang, H.; Gu, L.; Wu, X.; Bacchiocca, M.; Howard, L.; Hampsch-Woodill, M.; Huang, D.; Ou, B. and Jacob, R. 2003. Assays for hydrophilic and lipophilic antioxidant capacity oxygen radical absorbance capacity (ORACFL) of plasma and other biological and food samples. J. Agr.Food Chem. 51(11):3273-3279.