Evaluation of antioxidant extracts from Uruguayan native plants: importance of sensory characteristics
Evaluación de extractos antioxidantes de plantas nativas uruguayas: importancia de las características sensoriales

Abstract

There is increasing interest in the development of antioxidant extracts as functional ingredients. The aim of the present work was to evaluate the sensory profile, the polyphenolic (PP) content and 2,2-diphenyl-1-picryl-hydrazyl (DPPH) radical scavenging capacity of solvent extracts of three native Uruguayan plants: Achyrocline satureoides, Baccharis trimera, and Mikania guaco. Extraction solvent and native plant remarkably affected the PP content, the radical scavenging capacity, and the sensory profile of the extracts. Water extracts were the most effective radical scavengers and showed the highest polyphenolic concentration and the lowest intensity of bitter, astringent, and characteristic flavor. The bitterness, astringency and characteristic flavor of the extracts could limit their use as functional ingredients. Thus, when optimizing the extraction of antioxidant compounds from plants, apart from studying the influence of extraction solvents on yield, PP content and antioxidant capacity, the sensory characteristics of the antioxidant extracts should also be taken into account.

Existe creciente interés en el desarrollo de extractos antioxidantes como ingredientes funcionales. El objetivo del presente trabajo fue evaluar el perfil sensorial, el contenido de polifenoles y la capacidad de secuestrar radicales DPPH de extractos de tres plantas nativas uruguayas: Achyrocline satureoides, Baccharis trimera y Mikania guaco. Tanto el solvente como la planta nativa tuvieron un efecto significativo en el contenido de polifenoles, la capacidad de secuestrar radicales DPPH y el perfil sensorial de los extractos. Los extractos acuosos fueron los más eficientes secuestrantes de radicales DPPH y presentaron el mayor contenido de polifenoles y la menor intensidad de amargor, astringencia y sabor característico. El amargor, astringencia y sabor característico de los extractos podrían limitar su aplicación como ingredientes funcionales. Estos resultados sugieren que durante la optimización de las condiciones de extracción de compuestos antioxidantes de plantas, además de la influencia de la capacidad antioxidante y el rendimiento, deben tenerse en cuenta las características sensoriales de los extractos.

Palabras clave:

• extractos antioxidantes
• Achyrocline satureoides
• Baccharis trimera
• Mikania guaco
• extracción con solventes
• evaluación sensorial
• alimentos funcionales

Keywords:

• antioxidant extracts
• Achyrocline satureoides
• Baccharis trimera
• Mikania guaco
• solvent extraction
• sensory evaluation
• functional foods

Introduction

Polyphenolic constituents of plants are known to have nutritional as well as medicinal functions in the human body (Ames, Shigenaga, & Hagen, 1990; Nair, Li, & Kong, 2007; Walton, Rhodes, Michael, & Parr, 1999). Polyphenols appear as a group of compounds potentially beneficial to human health, and have been claim to be responsible for the overall physiological health effects of polyphenols-rich plants and foods. In this context, there is increasing interest in the development and use of polyphenolic-rich antioxidant extracts from different plants as functional ingredients.

Achyrocline satureioides (“Marcela”), Baccharis trimera (“Carqueja”), and Mikania guaco (“Guaco”) are native South-American plants. They are widespread in Argentina, Brazil, Paraguay, and Uruguay. Infusions, decoctions, and tinctures of these plants have been used for many years in natural and popular medicine (Filot Da Silva & Langeloh, 1994). These plants have been reported to have a high concentration of polyphenolic compounds (De Oliveira et al., 2003; Desmarchelier, Coussio, & Ciccia, 1998), which make them an interesting source of these functional ingredients.

Solvent extraction is frequently used for isolation of polyphenolic compounds (Pinelo, Rubilar, Sineiro, & Nunez, 2004; Spigno, Tramelli, & De Faveri, 2007; Sun & Ho, 2005). The polyphenolic (PP) content and antioxidant activity (AA) of the extracts depends on the solvent and method of extraction used, due to the different polarities, solubilities and antioxidant potentials of polyphenolic. For this reason, different solvents, such as water, ethanol, methanol, acetone and their mixtures, have been commonly used to obtain antioxidant extracts of plants (Khokhar & Magnusdottı, 2002; Miliauskas, Venskutonis, & van Beek, 2004; Spigno et al., 2007; Sun & Ho, 2005).

The aim of an extraction process to obtain an antioxidant extract is usually to provide extracts with the maximum concentration of target compounds and antioxidant capacity (Spigno et al., 2007). However, the suitability of an antioxidant extract as functional ingredient in food products does not only depend on its antioxidant capacity. The sensory characteristics should also be taken into consideration for commercial applications of antioxidant extracts as a functional ingredient in food products. Different extraction methods could yield antioxidant extracts with different sensory characteristics that might affect the sensory profile of the final product. Polyphenolic compounds are the main responsible for the bitterness and astringency of tea, red wine, olive oil, and several types of fruits (Lesschaeve & Noble, 2005). Therefore, the addition of these compounds to food products could change their sensory characteristics, which may lead to a decrease in consumer acceptability and willingness to purchase the product (Tuorila & Cardello, 2002; Urala & Lähteenmäki, 2003). This may limit the use of polyphenolic extracts from natural plants as functional ingredients in food products. Most published work investigated the influence of solvents on PP content and AA but no study has been found reporting the influence of solvent on the sensory properties of antioxidant extracts.

The aim of the present work was to evaluate the sensory profile, the PP content, and AA of solvent extracts (water, ethanol and acetone) of three native Uruguayan plants: Achyrocline satureoides, Baccharis trímera, and Mikania guaco.

Materials and methods

Reagents

2,2-diphenyl-1-picryl-hydrazyl (DPPH), Folin-Cioucalteu reagent and gallic acid were obtained from Sigma-Aldrich Co. (Steinheim, Germany). All reagents used were of analytical grade.

Plant material

Three native Uruguayan plants were considered: Achiyrocline satureoides, Baccharis trimera, and Mikania guaco. Dried leaves from these three native plants were obtained from a local retailer (La Botica del Señor, Montevideo, Uruguay). Three different samples of each plant were considered in the present work. Dried native plants were ground in a domestic blender immediately before the extractions.

Solvent extraction

Three solvents were used: acetone, ethanol, and distilled water.

Ethanol and acetone extracts were obtained by extraction of each ground dried native plant (10 g) in a Soxhlet extractor for 4 h, using a 15:1 solvent-plant material ratio (v/w). The extract was evaporated to dryness under vacuum at 50 °C in a rotary evaporator to dryness. To consider both yield and PP content simultaneously, the dried extracts were redissolved in 20 mL distilled water and filtered. The volume was adjusted to 20 mL after filtration.

For obtaining the aqueous extracts, 10 g of ground dried leaves of each plant was put into a glass jar containing 150 mL distilled water and kept at 90 °C for 1 h with occasional smooth stirring. After this extraction time, the extracts were filtered using Whatman number 1 filter paper. The clear filtrate was concentrated to 20 mL at under vacuum at 70 °C in a rotary evaporator.

Extracts were kept at 5 °C until the evaluations were performed.

Beverages

The PP content and the radical scavenging capacity of the Uruguayan native plants under study were compared to those of beverages with well-known antioxidant capacity, such as tea and wine.

Tea (Camellia sinensis) bags of finely ground black (Darjeeling Tea, R. Twining and Company Limited, London, U.K.), red (Instituto Botánico La Selva, Montevideo, Uruguay), and green (Instituto Botánico La Selva, Montevideo, Uruguay) tea, as well as Cabernet Sauvignon red wine (Frontera, Concha y Toro, Santiago de Chile, Chile) and Torrontés white wine (Bodegas Pisano, Progreso, Uruguay) were purchased in local supermarkets in Montevideo (Uruguay).

Tea infusions were prepared as specified in the packages. Tea bags (containing 3 g of tea leaves) were brewed for 3 min in 250 mL of boiling water. After this time, tea bags were removed and final volume adjusted to 250mL.

Polyphenolic content

The total polyphenolic content (TPP) content of beverages and extracts was spectrophotometrically determined at 750 nm using the Folin-Ciocalteau reagent (Singleton, Joseph, & Rossi, 1965; Taga, Miller, & Pratt, 1984) and gallic acid as standard. Although there are several compounds that could interfere with the Folin-Ciocalteau reagent, this methodology has been extensively used in the literature to determine the TPP of beverages and antioxidant extracts from different plants.

Briefly, 100 μL of an appropriate dilution of the extracts was added to 2 mL of 2% Na₂CO₃. After 2 min, Folin-Ciocalteu reagent (100 μL) was added to the mixture which was left to stand for 30 min at 20 °C. The absorbance at 750 nm was measured using a Spectronic Genesys 2 spectrophotometer (Spectronic Instruments, Rochester, NY), and compared to gallic acid calibration curves. Results were expressed as gallic acid equivalents (mg/L) using a gallic acid standard curve. All determinations were performed in triplicate for each sample of each of the three native plants and mean values and their corresponding standard deviations are given.

DPPH radical scavenging capacity

The radical scavenging activity of the extracts was estimated in terms of their hydrogen donating or radical scavenging ability, using the DPPH radical scavenging method (Brand-Williams, Cuvellier, & Berset, 1995). This method is based on the reduction of the stable radical DPPH (2,2-diphenyl-1-picrylhydrazyl) in the presence of antioxidant active substances.

For the analysis, 100 μL of different dilutions of the aqueous solutions of the extracts were placed in a cuvette containing 3.9 mL of a methanolic 60 μM DPPH solution. The reduction of the DPPH radical was measured by monitoring the decrease of absorption at 515 nm using a Spectronic Genesys 2 spectrophotometer (Spectronic Instruments, Rochester, NY), until stable extinction values were obtained (∼60 min). The exact DPPH concentration was calculated using a calibration curve. The radical scavenging activity of the extracts was calculated as the amount of DPPH inhibited by the sample as compared to a blank control. DPPH radical scavenging capacity was expressed as the volume of extract necessary to decrease 50% the initial DPPH concentration (EC50). All determinations were performed in triplicate for each sample of each of the three native plants and mean values and their corresponding standard deviations are given.

Sensory profile

The sensory panel consisted of eight assessors, ages ranging from 20 to 45 years old. Assessors were selected and trained following the guidelines of the ISO (1993) standard. They all had a minimum of 200 h of experience in discrimination and descriptive tests of different food products.

To generate sensory descriptors two aqueous dilutions (1/10 and 1/80) of three extracts were presented to assessors. First, assessors were asked to generate their individual descriptors using a modified grid method (Damasio & Costell, 1991). By open discussion with the panel leader, assessors agreed on the best descriptors to describe the samples. The descriptors used for the sensory profile of the antioxidant extracts were: bitter, astringent and characteristic flavor (it differed according to the native plant considered). Assessors were trained in the descriptors' evaluate on using different aqueous dilutions of the extracts. In a first session, assessors were asked to score the bitterness, astringency, and characteristic flavor of these samples using 10-cm unstructured scales. Characteristic flavor was defined as the herb flavor, characteristic of each native plant. Through open discussion with the panel leader assessors agreed on the scores for each of the samples. In successive sessions, the assessors received the samples, coded with 3-digit numbers. The assessors were asked to evaluate their bitterness, astringency and characteristic flavor, and to score them according to the consensus score. A variation in the scale of ±1 cm was considered acceptable. A total of eight 30 min sessions were used to train the panel.

A dilution flavor profile (Jellinek, 1985) was performed on the antioxidant extracts of the three evaluated native plants. As suggested by this methodology, seven samples were prepared from each antioxidant extract. Sample 7 was prepared by diluting 1/10 the aqueous dispersion of the extracts with tap water. This dilution was selected based on their PP content and preliminary studies. The other six samples were sequentially diluted with tap water, in a one to one ratio. Table 1 shows dilution of the aqueous dispersion of the antioxidant extracts corresponding to each of the seven samples. Samples were numbered according to their degree of dilution, i.e. the first sample (Sample 1) was the most diluted extract whereas the seventh sample (Sample 7) was the most concentrated.

Table 1. Dilutions of the seven evaluated samples of each solvent extract of the three evaluated native plants.
Tabla 1. Diluciones de las siete muestras evaluadas para los extractos de cada una de las tres plantas nativas evaluadas.

Sample	Dilution
1	1/640
2	1/320
3	1/160
4	1/80
5	1/40
6	1/20
7	1/10

Samples from each sample of each native plant were evaluated in duplicate by assessors. The evaluations were carried out in 54 sessions. In each session, the assessors evaluated the seven samples corresponding to a solvent extract of a sample of a native plant. Thirty milliliters of extracts were served in 100 mL plastic glasses at room temperature, coded with 3-digit random numbers. Unstructured 10-cm-long scales anchored with “nil” and “high” were used to describe attribute intensity. The testing was carried out in a sensory laboratory that was designed in accordance with ISO 8589 (1988). The evaluations were performed under red light, temperature control (between 22 and 24 °C) and air circulation.

Data analysis

Analysis of variance

Analysis of variance (ANOVA) was performed on data from PP content and radical scavenging capacity of extract and beverages considering sample as fixed source of variation. Also, an ANOVA was performed considering native plant, solvent and their interaction as fixed sources of variation.

An ANOVA was performed on data from the extracts' sensory profiles, considering native plant, solvent, concentration, and the first level interactions: native plant *solvent, native plant *concentration and concentration *solvent.

When differences were significant, honestly significant differences were calculated using Tukey's test. A 5% significance level (p < 0.05) was considered.

These analyses were performed using Genstat Discovery Edition 2 (VSN International, Oxford, UK).

Results and discussion

Total polyphenolic content and DPPH radical scavenging capacity

The total polypehnolic (TPP) content of the nine evaluated native plants extracts ranged from 7155 to 15,290 mg/L, as shown in Table 2.

Table 2. Total polyphenolic content and radical scavenging activity of solvent extracts of native uruguayan plants, and common beverages
Tabla 2. Contenido de polifenoles y capacidad de secuestrar radicales DPPH de extractos de plantas nativas uruguayas y bebidas comúnmente consumidas.

Sample	Polyphenols (mg/L)^a	DPPH radical scavenging capacity EC50 (μL)
Water extract of Achyrocline satureoides	11017 ± 95^c	1.21 ± 0.03^k
Ethanol extract of Achyrocline satureoides	9858 ± 87^e	1.89 ± 0.05^j
Acetone extract of Achyrocline satureoides	9078 ± 72^f	2.63 ± 0.07^h
Water extract of Baccharis trimera	15290 ± 150^a	1.10 ± 0.02^l
Ethanol extract of Baccharis trimera	14535 ± 110^b	1.24 ± 0.07^k
Acetone extract of Baccharis trimera	9840 ± 82^e	2.77 ± 0.09^g
Water extract of Mikania guaco	10155 ± 99^d	2.31 ± 0.05^i
Ethanol extract of Mikania guaco	7155 ± 58^g	15.9 ± 0.8^d
Acetone extract of Mikania guaco	9858 ± 82^e	19.0 ± 0.5^c
Black tea	942 ± 12^j	16.1 ± 0.5^d
Green tea	1538 ± 41^i	13.3 ± 1.1^e
Red tea	737 ± 42^k	50.1 ± 2.3^b
Red wine	2055 ± 52^h	5.0 ± 0.3^f
White wine	393 ± 42^l	62.2 ± 3.1^a
Note: ^aTotal polyphenolic content was expressed as gallic acid equivalents. Results are expressed as mean ± standard deviation from triplicate determinations. Values within a column with different superscript are significantly different according to Tukey's test (p ≤ 0.05). Nota: ^aEl contenido total de polifenoles se expresa como equivalentes de ácido gálico. Los resultados se expresan como promedio ± desviación estándar de determinaciones realizadas por triplicado. Valores en una columna con diferentes superíndices son significativamente diferentes de acuerdo al test de Tukey (p ≤ 0,05).

There was a highly significant effect of the native plant (p < 0.0001), extraction solvent (p < 0.0001) and their interaction (p < 0.0001) on the PP content. Baccharis trimera extracts showed the highest PP content, followed by Achyrocline satureoides and finally Mikania guaco. Water extraction yielded significantly more polyphenolic compounds than acetone and ethanol, for all the evaluated native plants. Ethanol extracts of Baccharis trimera and Achyrocline satureoides showed a higher TPP content than that of acetone. However, the opposite trend was found when Mikania guaco was considered. This suggests that the efficiency of an extraction solvent might be dependent on the type and polarity of PP compounds in the plants.

The free radical scavenging activity of the extracts was determined by the DPPH method. Percent DPPH radical scavenging activity was dependent of the concentration of the extracts. The extract volume necessary to reduce 50% the initial DPPH concentration is shown in Table 2.

There was a highly significant effect of the native plant (p < 0.0001), extraction solvent (p < 0.0001) and their interaction (p < 0.0001) on the radical scavenging activity of the extracts. The highest DPPH scavenging activities were shown by the Baccharis trimera extracts, in agreement with their highest TPP content. Baccharis trimera extracts were followed by Achyrocline satureoides regarding their radical scavenging capacity; whereas Mikania guaco extracts were considerably less effective radical scavengers. For the three native plants considered, water extracts contained remarkably higher amounts of radical scavenging compounds, followed by ethanol and finally acetone.

The radical scavenging activity of the extracts could be related to the content and nature of phenolics. Considering the evaluated antioxidant extracts, there was not a significant correlation between the TPP content of the extracts and their radical scavenging capacity, expressed as EC50 (R = −0.514, p = 0.154). Ethanol and acetone extracts from Mikania guaco showed a very weak radical scavenging activity, suggesting that their polyphenolic compounds were not very effective radical scavengers. Therefore, the correlation between TPP and EC50 was also calculated after removing these extracts. Without considering these extracts, the relationship became significant (R = −0.81, p = 0.029). This indicates that ethanol and acetone extracts of Mikania guaco showed different types of TPP than the rest of the extracts. In the case of extracts from Baccharis trimera and Achyrocline satureoides and water extracts of Mikania guaco, their radical scavenging capacity could be mainly attributed to their content of polyphenolic compounds.

To determine at which concentrations the evaluated native plants extracts should be used in food products, the TPP content and radical scavenging capacity of wines and teas was determined. TPP content and DPPH radical scavenging capacity of teas and wines are shown in Table 2. The values encountered are similar to those reported in literature (Astill, Birch, Dacombe, Humphrey, & Martin, 2001; Bravo, Goya, & Lecumberri, 2007; German & Walzem, 2000). Red wine showed the highest TPP content, ∼2 g/L, an average value for this type of wine (Bravo et al., 2007; German & Walzem, 2000). Black tea showed a lower PP content than green tea, in agreement with results reported by Astill et al. (2001) and Bravo et al. (2007). Radical scavenging capacity of these beverages followed the same trend than TPP content (c.f. Table 2).

The concentration at which the antioxidant extracts of Uruguayan native plants should be included in foods could be selected considering the TPP content of the evaluated beverages. Thus, if the antioxidant extracts were diluted 10–40 times they would show a similar TPP content than the evaluated beverages. For this reason, a 1/10 dilution of the extracts was the first dilution considered for sensory evaluation.

Sensory profile of aqueous solutions of solvent extracts of Uruguayan native plants

Bitter, astringent, and characteristic flavor intensities significantly (p < 0.001) increased with increasing antioxidant extract concentration, until it reached a plateau. Figure 1 shows the dilution profile of ethanol extract of Baccharis trimera. A similar trend was found for all the evaluated extracts. All the evaluated extracts showed very intense bitter, astringent and characteristic flavor, which could compromise their suitability as antioxidant extracts. Thus, these results suggest that the sensory characteristics of antioxidant extracts should be taken into account when studying its suitability as functional ingredient in food products.

Figure 1. Flavor dilution profile of ethanol extract from Baccharis trimera. Vertical bars represent standard deviations.

Figura 1. Perfil de dilución del extracto etanólico de Baccharis trimera. Las barras verticales representan la desviación estándar.

Extraction solvent and native plant significantly affected (p < 0.001) the intensity of bitter, astringent and characteristic flavor of the aqueous dilutions of the extracts. On the other hand, none of the evaluated interactions showed a significant effect, showing that the variables could be independently considered.

The influence of extraction solvent on the sensory profile of the extracts suggests that apart from affecting the yield of polyphenolic compounds, extraction solvent also influences the sensory characteristics of an antioxidant extract. As shown in Table 3, water extracts were less bitter and astringent than acetone and ethanol extracts. This shows that solvents different from water could also extract compounds with bitter, astringent or metallic flavors, according to their polarities. Several authors have recommended the use of methanol, ethanol, acetone or mixtures of these solvents to obtain antioxidant extracts of different plant materials (Miliauskas et al., 2004; Pinelo et al., 2004; Spigno et al., 2007; Sun and Ho, 2005), without considering the influence of these solvents on the sensory characteristics of the extracts.

Table 3. Mean ratings throughout concentrations for astringent, bitter and characteristic flavor of aqueous dilutions of solvent extracts of native Uruguayan plants.
Tabla 3. Promedio de intensidad de astringencia, amargor y sabor característico entre concentraciones para las diluciones acuosas de los extractos de plantas nativas uruguayas.

Sample	Mean rating
	Astringent	Bitter	Characteristic
Solvent
Water	4.0^a	4.1^a	4.9^a
Ethanol	5.3^b	5.6^b	6.2^b
Acetone	5.2^b	5.8^b	5.9^b
Native plant
Achyrocline satureoides	4.0^a	4.2^a	4.8^a
Baccharis trimera	5.4^c	5.8^c	5.6^b
Mikania guaco	4.9^b	5.3^b	6.5^c
Note: Mean values with different superscript are significantly different according to Tukey's test (p < 0.05). Nota: Valores con distinto superíndice son significativamente diferentes de acuerdo al test de Tukey (p < 0,05).

Extracts from Achyrocline satureoides showed the lowest intensity of bitter, astringent and characteristic flavor, as shown in Table 3. On the other hand, extracts from Baccharis trimera were the most bitter and astringent. Although water extracts from this native plant showed the highest TPP content and were the most effective radical scavengers, their higher bitterness and astringency could limit their application in food products.

Polyphenolic compounds are the main responsible for the bitterness and astringency of vegetables (Lesschaeve & Noble, 2005). However, their bitterness and astringency does not depend only on their concentration but also on their chemical structure. Differences in the degree of polymerization, the identity of the monomeric units, the bond location and the degree of esterification of polyphenolic compounds have been reported to markedly affect their bitterness and astringency (Lesschaeve & Noble, 2005). In the present study, TPP of the extracts was not significantly correlated with the average bitterness (R = 0.367) and astringency (R = 0.356) intensities throughout concentrations of their aqueous dilutions. Therefore, differences between the bitterness and astringency of the extracts could not be explained throughout differences in their TPP. This suggests that differences in the sensory characteristics of the extracts might be explained by differences in the composition of polyphenolic compounds instead of TPP. Moreover, apart from causing differences in TPP, the evaluated solvents might have also introduced differences in the composition of polyphenolic compounds of the extracts or in the concentration of other compounds. Further research is necessary to relate the sensory characteristics of the extracts with their composition of polyphenolic compounds.

On the other hand, the characteristic flavor of the extracts was not significantly correlated to their TPP (R = 0.05), as expected because polyphenolic compounds are not responsible for the characteristic flavor of the evaluated native plants.

These results suggest that the selection of antioxidant extracts to be used as functional ingredients in food products should based on a compromise between its antioxidant capacity and sensory characteristics.

Conclusions

Water extracts of Baccharis trimera and Achyrocline saturoides showed the highest PP content and DPPH radical scavenging capacity. Although these extracts appear as potential functional ingredients, their high bitterness and astringency could limit their application in food products. Further research is needed to evaluate their application possibilities and to study alternatives to mask the undesired flavors of the extracts.

Baccharis trimera extracts showed the highest radical scavenging capacity but they also were the most bitter and astringent. Thus, when optimizing the extraction of antioxidant compounds from plants, apart from studying the influence of extraction solvents on yield, PP content and antioxidant capacity, the sensory characteristics of the antioxidant extracts should also be taken into account. Results from the present work suggest that the sensory characteristics of antioxidant extracts could be a key issue that could limit the application of antioxidant extracts as functional ingredients in food products.

Further research is needed to characterize the polyphenolic compounds of the extracts and to better explain the differences in their sensory characteristics.

Acknowledgements

The authors are indebted to CSIC (Comisión Sectorial de Investigación Científica – Programa Iniciación a la Investigación Modalidad 1, Uruguay) for financial support, and to the Sensory Science Scholarship Fund and GlaxoSmithKline Consumer Healthcare for the Rose Marie Pangborn Sensory Science Scholarship granted to Gastón Ares. The authors are also grateful to TEPIVE, particularly to Raúl Estrada for providing the plants used in the study.