Antioxidant Activity of Co-Products from Guava, Mango and Barbados Cherry Produced in the Brazilian Northeast

Co-products from the juice processing of guava (CG), mango (CM) and barbados cherry (CB) were investigated with a view to their exploitation as a potential source of natural antioxidants. The ethanolic extracts were analyzed for total extractable phenolic content (TEP), DPPH radical scavenging activity (RSA-DPPH), ferric reducing antioxidant power (FRAP) and antioxidant activity in relation to the β-carotene/linoleic acid system. The TEP levels in the CG, CM and CB extracts were (24.15 ± 1.59), (44.18 ± 1.73) and (49.21 ± 3.70) mg GAE/g extract, respectively. The CM extract showed higher DPPH, FRAP and antioxidant activity in the β-carotene/linoleic acid system. The data revealed a positive linear correlation between TEP, RSA-DPPH and FRAP (r2 = 0.85 − 0.98); however, the β-carotene/linoleic acid system (r2 = 0.01 − 0.26) shows low correlation with the TEP levels and other assessment systems. The results suggest that co-products generated from the juice processing of the studied fruit have promising use as a natural source of antioxidants.


Introduction
Brazil is the sixth world's largest producer of fruits, with an annual production exceeding 37 million tons, which represents 5% of global production [1]. Of the total production in Brazil, around 53% is aimed at the processed fruit market and 47% is fresh fruit [2]. In response to the high productivity of the processing sector and the favorable business results in the Brazilian fruit growing activity, there has been a considerable increase in fruit processing industries. However, these industries generate large amounts of co-products such as peels, seeds, and bagasse during fruit processing, which are usually inappropriately discarded in the environment, leading to waste accumulation and negative environmental impacts [3,4]. Like fruits, co-products contain antioxidant compounds, including polyphenols, carotenoids, ascorbic acid and tocopherols of great physiological importance [5]. In this sense, an effective option for adding value to these co-products is to extract any bioactive constituents that may be used in the pharmaceutical, food or cosmetics industries [3,6].
The number of studies using agroindustrial co-products as sources of antioxidants has increased considerably in recent years. Ajila et al. [7] have studied the bioactive compounds and the antioxidant potential of mango peel extracts using different measurement systems and reported high antioxidant activity. Due to this important property, these researchers suggested the use of mango peel as a nutraceutical functional food. Oliveira et al. [3] evaluated the antioxidant activity of methanolic extracts obtained from barbados cherry, passion fruit and pineapple industrial co-products and concluded that the barbados cherry flour has a high content of phenolic compounds and antioxidant activity against the DPPH radical, while passion fruit flour provides good protection against membrane lipid peroxidation. Babbar et al. [4] assessed the antioxidant activity and total phenolic compounds of powders from six co-products, and grape seed extract showed the highest content of phenolic compounds and highest antioxidant potential, and banana peel extract showed the lowest content of both.
Several methods with different principles have been used in the in vitro antioxidant activity determination, among which the DPPH radical scavenging activity assay, which provides information on the overall antioxidant capacity of the test system [8] and the FRAP assay (ferric reducing antioxidant power) that relies on the ability of an anti-oxidant in reducing Fe(III) into Fe(II) and should be used combined with other methods because it cannot measure all antioxidants of complex matrices [9] stand out.. The β-carotene/linoleic acid coupled oxidation test is also used, where the model system is submitted to oxidation conditions that generate a free radical from the oxidation of linoleic acid, abstracting hydrogen from the unsaturated β-carotene molecule [10]. According to literature data, tests that evaluate antioxidant properties should employ more than one methodology in order to infer with greater certainty the antioxidant activity results [3].
Thus, the objective of this study was to quantify the total extractable phenolic compounds and to assess the in vitro antioxidant activity of co-products generated during industrial processing of guava Psidium Guayaba L.), mango (Mangifera indica L.) and barbados cherry (Malpighia glabra L.).

Yield and Total Extractable Phenolic Content (TEP)
The highest ethanolic extract yield of co-products from fruit processing was observed for the mango (CM; 11% ± 0.31%) and the lowest for the barbados cherry (CB; 5.5% ± 0.21%) ( Table 1). However, the result found for CB was higher than that reported by Oliveira et al. [3] assessing methanolic extracts of barbados cherry (2.5%), pineapple (7.1%), and passion fruit (3.02%) co-products. Phenolic compounds, present in large amounts in fruits, are reported as important constituents due to their antioxidant activity [11,12]. This power of scavenging radical structures results from the presence of at least one aromatic ring with hydroxyl groups on the chemical structure of phenolic compounds. The total extractable phenolic content (Table 1) ranged from 24.15 ± 1.59 to 49.21 ± 3.70 mg GAE/g of extracts and from 188.40 ± 12.38 to 485.93 ± 19.08 mg GAE/100 g in the dry co-products. The total extractable phenolic content in CB and CM extracts showed no significant difference between each other, reaching values two times higher than the extract CG. On the other hand, in dry fruit co-product , the TEP differed (p ˂ 0.05) and the highest values were observed in CM, followed by CB and CG. The results found in the CM extracts were lower than those reported by Kim et al. [13] for green and ripe mango peels (92.6 and 70.1 mg GAE/g each g of extract, respectively). However, like the results of this study, Ajila et al. [7] reported total extractable phenolic content values between 33.31 to 73.88 mg GAE/g of alcoholic extract, analyzing peel from two Indian varieties of green and ripe mangoes. The CB extract (49.21 ± 3.70 mg GAE/g) also showed differences in the total extractable phenolic content compared to the study by Oliveira et al. [3] (94.6 mg GAE/g) in methanolic extracts. However, Sousa et al. [2], observed total extractable phenolic contents from 46.77 to 279.99 mg GAE/100 g in wet co-product of fruit pulp extracted by a hydroethanolic process for guava and barbados cherry, respectively, which is equivalent to CB and lower to those reported for CG in this study. An interesting finding is that the value obtained for CG (188.40 ± 12.38 mg GAE/100g) was higher than that described for wet guava pulp (83.1 mg GAE/100 g) by Kuskoski, et al. [14], suggesting the potential use of co-products discarded during fruit processing. It is noteworthy that differences may also be due to factors such as climate, soil, stage of fruit ripening, drying method, type of solvent and extraction time, and part of the fruit that is present in the co-product [15]. Table 2 shows the antioxidant activity of each extract and BHT (used as reference), where it can be observed that the CM extract has the highest antioxidant activity among the analyzed extracts, followed by CB and CG extracts. The CM and CB extracts showed higher antioxidant activity (86% and 55%, respectively) when compared to BHT. Sousa et al. [2] have reported EC 50 (µg/mL) of 142.89 and 308.07 for wet guava and barbados cherry co-products, respectively, and the value for barbados cherry is lower than that shown in this study. However, in dry products, the amount of phenolic compounds is concentrated.   In a study by Babbar et al. [4] on methanolic extracts of lychee, tangerine, grapefruit and banana co-products at concentration of 5 mg/mL, it was observed that the co-products had antioxidant activity close to that of BHT at the same concentration (83%), except for banana peel, which was lower.

Ferric Reducing Antioxidant Power (FRAP)
Through the FRAP method, the CM extract also showed higher antioxidant activity, followed by CB and CG (Table 2). Muller et al. [16] reported FRAP values in barbados cherry of 17.23 mmol of Fe(II)/100 g of fresh fruit, and Rufino et al. [17] of 199.60 mMol Fe(II)/100 g of fruit on a dry basis. The value reported in this work for CB is higher than that found by these authors. This difference suggests that the content of antioxidant compounds present in seeds, the major component of co-products, is greater, which may contain compounds with higher antioxidant activity.

β-Carotene Bleaching Method
All extracts at a concentration of 1 mg/mL inhibited β-carotene oxidation ( Table 2). The CM extract, using both DPPH and FRAP methods, obtained the best result and the CG extract inhibited β-carotene oxidation more than the CB extract. The kinetics of β-carotene degradation ( Figure 2) show that the extracts inhibit the β-carotene oxidation throughout the degradation curve both at the beginning of the oxidation process (15-45 min) as at the end of it (75-105 min). The CM extract gave better antioxidant activity when compared to CG and CB extracts. Importantly, in this system, barbados cherry co-products had a different behavior from that observed for FRAP and DPPH, which agrees with previous reports by Rufino et al. [17] and Alves et al. [18]. This change in behavior occurs due to the action of vitamin C, which acts as a pro-oxidant factor in the system with the formation of ascorbyl radicals during oxidation [17,19]. Another factor that may affect this result is the use of high temperature resulting in the decomposition of phenolic compounds [20].

Correlation between Total Phenolic Compounds and Antioxidant Activity using DPPH, FRAP and β-Carotene/Linoleic Acid Methods
The correlation between concentration of total phenolic compounds and antioxidant activity has been widely studied, including with fruits and fruit co-products [4]. In the present study, a high linear correlation (r > 0.92) was observed between the anti-radical efficiency in relation to DPPH and FRAP and total phenolic content present in the extracts tested, as reported by Oliveira et al. [3]. The antioxidant activity assessed by the β-carotene/linoleic acid system showed no linear correlation with the other methods tested or with the total phenolic content, as reported in the study by Rufino et al. [17]. It is known that not only phenolic compounds, but other components such as ascorbate, reducing carbohydrates, tocopherols, carotenoids, terpenes and pigments are factors that affect the antioxidant activity. In addition, each phenolic compound has distinct antioxidant potential, which in synergistic or antagonistic interactions, including with non-phenolic compounds, may affect the results [21].

Sample Preparation and Extraction
Co-products from the processing of guava (CG), mango (CM) and barbados cherry (CB) were obtained from a fruit pulp industry located in the city of João Pessoa/PB, Brazil. The fruit co-products collected directly from the production line were dried at 45 °C in an oven with air circulation (Model MA 035, Marconi, Piracicaba, Brazil) for 48 h. Then, the dried co-products were ground in a knife mill and vacuum packed in polyethylene bags and kept in a freezer (−18 °C) until the experiments. The extracts were obtained by stirring dry co-product (10 g) with ethanol (50 mL) in thermostated bath for 2 h at 25 °C. Then, the extract was centrifuged at 5,000 rpm for 10 min and the supernatant obtained was filtered with ethanol and diluted in a 100 mL flask. The extracts were stored in amber vials under an inert atmosphere until used. To determine the dry matter yield, 1 mL of extract was added to beakers previously dried in an oven and dried to constant weight at a temperature of 105 °C and the result expressed as percentage of extract/100 g of dry co-product (dc).

Determination of Total Extractable Phenolic Content
The levels of total extractable phenolic content (TEP) were determined colorimetrically by the Folin-Ciocalteau method [22] with volume reductions. Aliquots (4 mL) of ethanolic extracts diluted in water were added to Folin-Ciocalteau reagent (0.25 mL) followed by addition of 20% sodium carbonate (0.75 mL). The mixture was stirred and maintained in the dark for two hours. The absorbance at 765 nm was measured for UV-Vis spectrophotometer (model UV-2550, Shimadzu, Tokyo, Japan), together with a control containing only water and reagents. The concentration of phenolic compounds was estimated using a gallic acid curve calibration (50-500 mg/L). The results were expressed as mean ± standard deviation of mg gallic acid equivalent (GAE)/g of extract and GAE/100 g of dry co-product (dc).

Radical-Scavenging Activity of the 2,2'-Diphenyl-β-Picrylhydrazyl Radical (DPPH) Assay
The capacity of ethanolic extracts obtained from dry extracts to scavenge the DPPH radical was compared to the synthetic antioxidant BHT by the RSA-DPPH method proposed by Brand-Williams et al. [23]. Aliquots of ethanolic extract of each co-product and BHT (3.0 mL) at dilutions of 20, 40 and 80 µg/mL were added to DPPH ethanolic solution (0.1 mL, 0.1 mol/L). The control sample consisted each extract (3.0 mL) and ethanol (0.1 mL) and the negative control of ethanol (3.0 mL) and DPPH (0.1 mL). The decrease in absorbance at 515 nm was measured after 120 min. The RSA-DPPH was calculated according Rufino et al. [17]. From the results obtained, a graph for the % of DPPH radical scavenging activity in each extract concentration (mg/mL) was constructed. For the calculation of the EC 50 (concentration of extract with the capacity to reduce 50% of the initial DPPH), the equation of a line was used, substituting the value of y by 50. The anti-radical efficiency was defined as the inverse relation of EC 50 , i.e., (1/EC 50 (mg/mL) ).

Ferric Reducing Antioxidant Power (FRAP) Assay
The antioxidant activity of each sample was estimated by the ferric reducing antioxidant power (FRAP) assay [24]. Briefly, FRAP reagent (2.7 mL) prepared immediately before use (TPTZ 10 mM FeCl 3 and 20 mM acetate buffer) was homogenized with each extract (90 µL) and distilled water (270 µL) in a thermostated bath at 37 °C for 30 min. Then, the absorbance at 595 nm was measured using the FRAP reagent as negative control to calibrate the spectrophotometer. Concentrations of 500-2000 µmol/L of ferrous sulfate (FeSO 4 ·7H 2 O) were used to determine the standard curve. The results were expressed as mean ± standard deviation of mM ferrous sulfate per g of co-product extracts of each fruit and mM ferrous sulfate per 100 g dry co-product (dc).

β-Carotene Bleaching Method
The antioxidant activity measured by the β-carotene/linoleic acid system model system co-product extracts was determined as method described by Miller [25]. This assay is based on the oxidation of β-caroteno/linoleic acid induced by oxidative degradation products of the linoleic acid. The solutions were prepared by mixing β-carotene/linoleic acid system solution (5 mL) and extract of each fruit (0.4 mL). After an initial absorbance reading at 470 nm, the mixture was kept in a thermostated bath at 40 °C, and then absorbance was measured at intervals of 15 min to 120 min. The negative control sample consisted of β-carotene/linoleic acid system (5 mL) and ethanol (0.4 mL).

Statistical Analysis
All assays analysis was performed in triplicate in three different occasions. Results were expressed as mean ± standard deviation, compared by analysis of variance (ANOVA) and Tukey's test, considering p < 0.05, using the Statistica 7.0 software (Statsoft ® , Boston, MA, USA). The Pearson's correlation coefficient (r) was calculated between total phenolic content and the methods used that asses the antioxidant activity.

Conclusions
Industrial co-products resulting from the processing of mango, barbados cherry and guava pulp are rich in total phenolic compounds, especially CM and CB. All ethanolic extracts tested showed effective antioxidant activity in the DPPH and FRAP assay systems and antioxidant activity in lipid systems, especially the CM extract. The results suggest that co-products from the processing of the three fruit pulps have promising use as sources of antioxidants.