Nova Biotechnol Chim (2018) 17(1): 74-85 DOI: 10.2478/nbec-2018-0008  Corresponding author: jalcam@unsa.edu.pe Nova Biotechnologica et Chimica Nutritional composition, total phenolic compounds and antioxidant activity of quinoa (Chenopodium quinoa Willd.) of different colours Yemina Karen Diaz-Valencia1, Juan José Alca1,, Maria Antonia Calori-Domingues2, Sonia Jackeline Zanabria-Galvez1 and Sandra Helena Da Cruz2 1Department of Food Industry Engineering, Universidad Nacional de San Agustín de Arequipa, Independencia Avenue s/n, Arequipa 04000, Peru 2Department of Agri-food Industry, Food and Nutrition, College of Agriculture “Luiz de Queiroz”, University of Sao Paulo, Pádua Dias Avenue 11, Piracicaba, Brazil Article info Article history: Received: 2nd January 2018 Accepted: 30th April 2018 Keywords: Antioxidant Minerals Nutrition Phenolics Starch Quinoa Abstract Quinoa (Chenopodium quinoa Willd.) has been nutritionally highlighted when compared to other grains. In recent years the research on this pseudocereal has increased. In this work, six quinoa samples were studied: three from Peru, one from Brazil and two commercial samples. The samples were physically and physicochemically characterized, including macro- and micronutrient analysis, phenolic compounds content and antioxidant activity. Black, red and white samples showed as main difference the size, weight, ashes and dietary fibre content. Black samples were the smallest and lightest and had the lowest starch content but presented the highest levels of ashes and dietary fibre. The protein content (16.9 %) in the white Brazilian variety was higher than the others. Red and black samples had the highest levels of most minerals analysed. The antioxidant capacity measured by the DPPH method was higher for black and red samples in comparison with the white ones. However, the white Brazilian variety showed a significantly higher antioxidant capacity measured by the ABTS assay. With regard to the phenolic content, a difference was found between the samples which ranged from 55.5 to 95.5 g GAE 100 g-1. The colour of the grain was found as not related to a higher content of phenolic compounds. Because their compositions are generally similar to light-coloured grains, and in some parameters such as dietary fibre and content of some micronutrients are superior, the grains of dark-coloured quinoa varieties (RPP, BCP) would have to be explored to develop foods that take advantage of this colour diversity.  University of SS. Cyril and Methodius in Trnava Introduction Quinoa (Chenopodium quinoa Willd.) has been cultivated by the inhabitants of the Andes for human consumption for centuries and in recent years it has increased its sowing and market value in part to its excellent nutritional characteristics (Navruz-Varli and Sanlier 2016). The light- coloured varieties had been the most preferred by the consumers and also the most studied. Several studies had already reported the nutritional profile of these varieties (Kozioł 1992; Ando et al. 2002; Aluwi et al. 2017) indicating also the presence of some anti-nutrients like phytates and saponins. Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 75 The relationship between the content of saponins and the colour of the quinoa grains has been studied showing that the whites are sweeter (lower saponins content) than the yellow ones (Souza et al. 2004). With regard to the phenolic acid content of quinoa of different colours, the published results were inconclusive. Some have reported that the content of phenolic acids and flavonoids did not differ much between samples of different colours, highlighting only some variety while others of similar colour presented similar values to the rest (Repo-Carrasco-Valencia et al. 2010). On the other hand, comparing one cream-coloured quinoa sample with other white ones, the first has been reported to have a higher content of total polyphenols and scavenging activity (Repo- Carrasco-Valencia et al. 2011). Other authors have pointed out that for red quinoa varieties, the ones that were redder would have greater antioxidant activity and higher content of phenolic compounds (Abderrahim et al. 2015). It is reported that the content and characteristics of starch and protein from different quinoa lines vary due to the environment and genetic and agronomic differences (Lindeboom 2005). In the same way, the variation in the mineral content of quinoa of different varieties is due to the interactions of genotype and environmental parameters (Prado et al. 2014). The content of certain minerals was higher in varieties that were of a determined colour. Physically, some coloured varieties were reported to be heavier and larger than the white ones (Apaza et al. 2013). Thus, our aim was to determine some characteristics of different quinoa varieties in order to confirm possible relationships between their values and the colour. Experimental Quinoa samples Three samples of Quinoa (Chenopodium quinoa Willd.) grains were obtained from the National Agricultural Research Institute (INIA) of Peru: INIA 415 “Pasankalla”, INIA 420 “Negra Collana” and INIA “Salcedo”, coded as RPP, BNP and WSaP for this study. RPP, BNP and WSaP were released in 2006, 2008 and 1995 respectively (Apaza et al. 2013). Two commercial samples from Arequipa, Peru were purchased for the study (coded as RCP and BCP for this study) and a Brazilian quinoa sample (BRS “Syetetuba”, coded as WSyB, released in 2006 according to Spehar et al. 2011) was obtained from the farm of the Federal University of Brasilia. RPP and BNP samples were sown in 2011 and harvested in 2012 (Puno, Peru), WSaP was sown in 2013 and harvested in 2014 (Puno, Peru), WSyB was harvested in 2014 (Brasilia, Brazil), the commercial varieties (RCP and BCP) were both acquired in 2014. The samples of quinoa grains are shown in Fig. 1. Fig. 1. Quinoa grains of different colors. RPP: INIA 415 “Pasankalla”, RCP: Red Peruvian commercial sample, BNP: INIA 420 “Negra Collana”, BCP: Black Peruvian commercial sample, WSyB: BRS “Syetetuba”, WSaP: INIA “Salcedo”. Saponin removal The saponin content in the samples was evaluated by a semiquantitative method described by Kozioł Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 76 (1990). This method is based on the tensioactive properties of the saponins, which form a stable foam whose height is related to the saponin content in the grains after being dissolved in water and shaken. The percentage of saponins was obtained with Eq. 1. (1) Preparation of quinoa samples The quinoa grains were washed with distilled water (at 37 °C) in the ratio 1:2 (quinoa: distilled water). Then the grains were shaken by hand for 20 min with distilled water and rinsed. This procedure was performed twice and then the grains were dried at 37 °C for 12 hours in a hot air oven. The dried grains were stored in polypropylene bags in a cold storage chamber (10 °C) for two months before use (August 2014). For the colour evaluation, four sub-samples were taken from these stored grains. For the determination of size distribution and weight of thousand grains eight sub-samples were taken. Three, two and six sub- samples were taken to the physical-chemical, elemental and antioxidant activity analysis respectively. For all cases, each of these sub- samples were processed according to the corresponding methodology described in this work. Colour evaluation The colour of the quinoa grains was evaluated using the colourimeter Chroma Meter 200b (Minolta Co., Osaka, Japan) by the CIE 1976 L*a*b* system obtaining the values of L* (lightness), a∗ and b∗ (opponent colour axes). The hue angle (H°) and chroma (C∗) were calculated according the Eq. 2 and 3. Size distribution and weight of thousand grains The dried quinoa grains after saponin removal were analyzed by size. For this purpose, Tyler/mesh sieves number 12, 14 and 20, corresponding to openings of 1.41 mm, 1.19 mm, and 0.841 mm respectively, were used. The amounts retained in each sieve were weighed and expressed in percentages. In addition, 100 whole grains without impurities were counted manually, in 8 replicates, for the determination of the weight of one thousand grains. After weighing using a Sartorius BL210S analytical balance (Sartorius, Germany), the mean, standard deviation and coefficient of variation of the measurements were calculated. The result of a thousand grains was calculated by multiplying by 10 the average weight of the repetitions if the coefficient of variation was less than 4 %. As the weight of a thousand grains of a sample varies according to the moisture content the results obtained were standardized to 10 % moisture, which was the average moisture obtained from the grains at the time of evaluation. Physical-chemical analysis The grains previously washed were ground in a Marconi MA-090 hammer mill (Marconi Ltd., Piracicaba, Brazil) with a 20 mesh screen (0.841 mm). The milled grains were evaluated for moisture using an A&D MX-50 Moisture Analyzer (A&D Co. Ltd., Tokyo, Japan). The ash content (calcination in muffle at 550 °C for two hours), lipids (extraction with ethyl ether using Soxhlet extractor) and crude protein (Kjeldhal digestion with the factor 6.25 for conversion of the total nitrogen content) were determined according to the methods 923.03, 920.39C and 979.09A described by AOAC (2005). The starch content of the milled grains of each sample was determined by the Megazyme Total Starch Kit (AA / AMG) K- TSTA 09 (Megazyme, Ireland). Macro- and microminerals content Macrominerals phosphorus, potassium, calcium, magnesium, sulfur and microminerals iron, zinc, copper, manganese, aluminum, sodium, and lithium were determined using a methodology based on the method 953.01 by AOAC (2005). For all elements, nitroperchloric digestion was used, except for boron, which was extracted by dry digestion in muffle according to Malavolta et al. (1997). The quantification of the elements was performed in an atomic absorption spectrophotometer, (2) (3) Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 77 the lithium was quantified in a coupled plasma mass spectrometer (iCAP 7400 ICP-OES Analyzer, Thermo Fisher Scientific, USA). The results were expressed as mg 100 g−1 of dry basis product. Determination of phenolic compounds The content of total phenolic compounds (TPC) was determined according to the Folin-Ciocalteu spectrophotometric method (Singleton et al. 1999) using gallic acid as standard. The hydroalcoholic extracts were obtained according to the methodology of Repo-Carrasco-Valencia et al. (2010), with modifications. A sample containing 5 g of milled grains was dissolved in 20 mL of 95 % ethanol. After homogenization for 1 min the samples were storaged at -4 °C for 18 hours. Then the extracts were centrifuged at 29,000 g for 20 min in an Eppendorf 5810 R refrigerated centrifuge at 4 °C (Eppendorf AG, Hamburg, Germany) and filtered using Whatman filter paper No. 2 and stored in amber bottles under refrigeration at 7 °C until the time of analysis. Aliquots of 0.5 mL of the hydroalcoholic extract were transferred to test tubes, where 4 mL of distilled water and 0.5 mL of the Folin-Ciocalteu solution were added. Then the tubes were shaken in a vortex mixer and put aside for 3 min. After the addition of 0.5 mL of a 4 % (w/v) sodium carbonate solution, the tubes were vortexed again and held for two hours at room temperature and protected from light. Immediately after, the absorbance reading was performed in a FEMTO 700S spectrophotometer at 740 nm (Sao Paulo, Brazil). The results were expressed in milligrams of gallic acid equivalent (GAE) per gram (dry basis). Evaluation of antioxidant capacity The evaluation of the antioxidant capacity of the hydroalcoholic extracts obtained for the determination of phenolic compounds was performed using the 2,2-diphenyl-1-picrylhydrazyl radical (DPPH) and 2,2'-azino-bis(3- ethylbenzthiazoline-6-sulphonic acid) (ABTS) assays. For the DPPH assay the methodology proposed by Brand-Williams et al. (1995) was used. An aliquot (500 μL) of solution was transferred to a test tube containing 3 mL of 95 % ethanol and 300 μL of the DPPH solution. The solution was held out of the light for 45 min previously to the absorbance reading at 515 nm using a spectrophotometer and compared with a blank. A standard curve was obtained using trolox solutions from 5 to 50 μM. The trolox equivalent antioxidant capacity (TEAC) was obtained with Eq. 4 and expressed as μmol TEAC g−1 db (dry basis). The methodology proposed by Re et al. (1999) was used to evaluate the reduction of the ABTS radical. An aliquot (30 μL) was transferred to a test tube containing 3.0 mL ABTS (7 mM). The solutions were kept in the dark for 6 min and the absorbance was read at 734 nm. The standard curve was obtained from 500 to 2000 μM using trolox solutions. The results were expressed in μmol TEAC g−1 db using a blank as a reference. Statistical analysis The results were subjected to analysis of variance (ANOVA) and the Tukey test for comparison of means (p < 0.05) using the SAS statistical software version 9.3 (SAS Institute Inc., Cary, NC, USA). Results and Discussion Saponin content The quinoa grains of the different samples were evaluated for saponin content before and after saponin removal. The grains presented low initial saponin content, from nearly zero (≤ 0.01 % db for RPP and BCP) to ≤ 0.04 % db (RCP, BNP and WSaP), except for WSyB which presented a high content (0.34% db). BNP and RPP were considered as free from saponins by Apaza et al. (2013) like WSyB according to Spehar et al. (2011), which differed from our results. Kozioł (1992) suggested, as possible explanation, the different environmental conditions of grain production or the different methodologies used to determine the saponin (4) Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 78 Table 1. Lightness (L*), opponent colour axes (a* and b*), chroma (C*) and hue angle (H°) of quinoa grains. Four replicates. Values followed by the same letter in the column do not differ (Tukey test, p < 0.05). Sample L* a* b* Chroma (C*) H° RPP 33.4±1.4c 16.2±1.2 25.4±2.8 30.2±2.9a 57.4±1.5b RCP 30.8±1.5d 18.9±0.9 22.1±2.1 29.1±2.1a 49.4±1.7 c BNP 14.1±1.8e 7.3±1.5 4.9±0.9 8.8±1.7c 34.2±4.4e BCP 13.2±2.6e 6.9±1.6 5.7±1.5 9.1±1.6c 39.4±9.3d WSyB 66.0±0.9b 2.9±2.7 23.1±0.7 23.2±0.7b 82.8±0.5a WSaP 76.1±1.2a 2.1±0.3 24.8±1.0 24.9±1.0b 85.3±0.7a content. Our results showed that after saponin removal procedure, the saponin content was 0.01 % (dry basis) for BSyB and 0 % for the other cultivars. According to this author the saponins are the main antinutritional factors of the quinoa grain but they can be removed by wet or abrasive methods, reaching levels not exceeding 0.01 %, as observed in our samples because such substances are concentrated in the outer layers of the grains. It has been reported that some plant- based saponins can form complexes with iron and zinc, which would reduce absorption in rats (Southon et al. 1988). However, saponins have very low toxicity to other mammals (Malinow et al. 1982) and do not affect protein availability in the case of quinoa (Ruales and Nair 1992). Color evaluation The studied quinoa samples presented different colour parameters as shown in Table 1. In the literature consulted the saponin content was related to the colour; reporting that the yellow cultivars had more saponin than the white ones (Souza et al. 2004). The values observed for H° (above 80) suggest that the samples considered white could be classified with a tendency to yellow. However, even though the white Brazilian sample (WSyB) had the highest content of saponins, the Peruvian white sample (WSaP) had a similar content to the samples of different colour, so it is not possible to confirm with our data the relationship between colour and content of saponins in quinoa grains of different colours. The differences observed for H°, among the grains of the same colour, RPP and RCP and BNP and BCP samples, may be due to the commercial samples (RCP and BCP) are not standardized and a mixture of different colours may have occurred. Medina et al. (2010) tried to use image analysis to determine the geographic origin of different quinoa grains but proved that it was impossible to use the colour of the grains for this purpose. However, it has been stated that the colour of grains of dark quinoa varieties is an indicator of the presence of betacyanins (low L* values), while the colour of lighter varieties is related to a higher betaxanthin content (Escribano et al. 2017). Thus, the main use of measuring the colour of quinoa grains has been as a characteristic associated with their composition. As regards manufactured products (based on quinoa flour), Table 2. Distribution of quinoa grains by size and weight of one thousand grains. For size distribution n = 4 whereas for the weight of one thousand grains n = 8. Values followed by the same letter in the row do not differ (Tukey test, p < 0.05). Tyler mesh No. (aperture in mm) Percentage with respect to the weight of one thousand grains [%] RPP RCP BNP BCP WSyB WSaP 12 (1.41) 86.2 83.9 12.6 19.6 78.2 85.7 14 (1.19) 13.4 13.9 81.0 74.7 21.2 13.7 20 (0,841) 0.4 2.2 6.4 5.4 0.4 0.4 Bottom (< 0.841) 0.0 0.0 0.0 0.3 0.2 0.2 One thousand grains weight [g] 3.71±0.06c 4.11±0.09a 2.30±0.07e 2.06±0.05f 3.88±0.06b 3.12±0.09d Coefficient of variation [%] 1.6 2.1 3.2 2.3 1.6 2.9 Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 79 Table 3. Physical-chemical composition of quinoa grains. Three replicates. Values followed by the same letter in the column do not differ (Tukey test, p < 0.05). Quinoa samples Ash [%] Protein [%] Lipid [%] Total Dietary Fibre [%] Starch [%] RPP 2.51±0.02b 14.0±0.3b 7.33±0.11a 14.30±0.01b 56.02±0.03 b RCP 2.58±0.02b 12.5±0.6b 6.70±0.14b 13.90±0.06b 54.56±0.01 d BNP 2.96±0.10a 14.1±0.1b 6.67±0.10b 18.20±0.00a 49.25±0.07 e BCP 3.01±0.09a 14.0±0.2b 7.34±0.11a 19.70±0.22a 47.22±0.09 f WSyB 2.32±0.01c 16.9±0.2a 6.84±0.08b 8.70±1.67c 55.39±0.05 c WSaP 2.36±0.02c 13.7±0.1b 7.29±0.03a 10.50±0.45c 59.72±0.05 a it has been observed that there is an effect on the colour of the final product in relation to the percentage of flour used, but it was not indicated whether the pigments of quinoa grains played a role in this characteristic (Demir and Kilinç 2017). Size distribution and weight of thousand grains The red and white samples showed larger grains than the black ones (Table 2). This is consistent with the data from Apaza et al. (2013) who reported the average diameter of BNP, RPP and WSaP as 1.60, 2.10 and 2.00 mm, respectively. The same authors also reported the weight of thousand grains for BNP, RPP and WSaP as 2.03, 3.51 to 3.72 and 3.10 to 3.70 g, respectively. This trend is followed by our data, which showed that black samples are lighter than the others. Values above 3.0 g per thousand grains are considered as large grains (Spehar et al. 2011) and considered more desirable in the growing quinoa market. Physical-chemical analysis All the samples had moisture levels below 12 % (from 8.3 to 11.2 %), being thus suitable for storing according to Spehar (2006). The black samples presented the highest values for the ash content (Table 3), around 3 % db, followed by the red and white samples. These values were similar to those found by Repo-Carrasco-Valencia et al. (2010) for the cultivars RPP (2.5 %) and WSaP (2.4 %). All the samples analysed had a similar protein content, with only one of the white samples (WSyB) standing out. With respect to lipid content, a sample of each colour (RPP, BCP, WSaP) was particularly noteworthy. The protein content of the samples was also similar to those observed by Miranda et al. (2012), who reported values from 11.32 to 16.10 % db. Repo-Carrasco-Valencia et al. (2010) reported values for lipids ranging from 5.2 to 6.8% for white, black and red cultivars. Meanwhile, Ando et al. (2002) found 6.5 % of lipids in the white cultivar Real, which was slightly lower than our results for samples of the same colour. The content of lipids of all samples was higher than other cereals such as rice, corn, wheat and barley, which present 2.2 %, 4.7 %, 2.3 % and 1.9 %, respectively (Kozioł 1992). The two white samples had the lowest values of dietary fibre, followed by the two red samples. The dietary fibre content of black samples is comparable to that of hulled barley and superior to that of raw white rice (17.3 % and 1.3 % respectively, according to USDA (2017). The variation in the ash, protein, lipid, and fibre contents among quinoa samples can be attributed to the production site (climate, soil, temperature, management) and to the genetic characteristics of the cultivars (Lindeboom 2005; Repo-Carrasco-Valencia et al. 2011). The main component found in the samples was starch (Table 3). The observed results for starch are close to the ranges already reported in previous studies. The composition of quinoa grains of different colours (red, yellow and white) was evaluated by Bruin (1964), reporting the starch content ranged from 58.1 to 64.2 % and the white samples showed the highest levels. On the other hand, Lindeboom (2005) evaluated 8 lines of quinoa grains founding that the starch content varied from 48.3 to 62.5 % db, but the colour of the grains was not specified. Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 80 Table 4. Contents of macrominerals and microminerals in mg 100 g⁻¹ db, except for Li. Two replicates. Values followed by the same letter in the row do not differ (Tukey test, p < 0.05). Element Quinoa samples RPP RCP BNP BCP WSyB WSaP P 578±13a 588±13a 512±14b 490±7b 510±12b 492±4b K 699±1b 593±5c 815±33a 693±1b 559±10c 663±1b Ca 73±3b 109±0a 71±7b 118±1a 69±5b 77±2b Mg 243±7bc 264±5a 226±6cd 246±0ab 239±5bcd 220±1d S 185±5a 174±5ab 158±14ab 144±1b 177±16ab 164±2ab Fe 5.63±0.15cd 5.25±0.03d 7.28±0.18bc 17.2±1.10a 8.59±0.09b 4.74±0.03d Zn 3.70±0.05b 3.79±0.04b 2.98±0.06d 3.36±0.04c 5.12±0.08a 3.17±0.03cd B 1.34±0.05a 1.16±0.03b 1.01±0.01cd 0.92±0.01d 1.11±0.02bc 0.80±0.02e Cu 0.47±0.01b 0.50±0.01b 0.47±0.04b 0.57±0.00a 0.48±0.01b 0.51±0.01ab Mn 6.90±0.15c 3.41±0.02e 9.48±0.11a 7.99±0.06b 2.53±0.07f 5.53±0.10d Al 2.21±0.05dc 2.82±0.25c 1.84±0.01d 5.87±0.20a 4.54±0.19b 1.79±0.07d Na 17.5±0.29c 36.9±0.56b 11.8±0.32d 49.3±0.17a 17.1±0.06c 16.6±0.38c Li (µg kg-1) 0.14±0.01b 0.58±0.02a 0.14±0.02b 0.65±0.08a 0.17±0.02b 0.19±0.00b Macro- and micromineral content A high content of P and K was observed, confirming that these two elements represent the greatest amount of minerals in quinoa grains (Bruin 1964; Mota et al. 2016). Red samples showed the highest phosphorus content with respect to the others samples. RPP and RCP had more P content that varieties from Chile (285.6 to 526.36 mg 100 g−1 db according to Miranda et al. 2012) or Bolivia (123.72 to 330.96 mg 100 g−1 db according to Prado et al. 2014). The content of P in quinoa is higher than that found in maize (210 mg 100 g−1), wheat (323 mg 100 g−1), rice (95 mg 100 g−1) and barley (221 mg 100 g−1) (USDA 2017). It was reported that the potassium content of quinoa milled grains was 639.3 mg 100 g−1 (Ando et al. 2002), whereas (Prado et al. 2014) showed a wider range (from 498.47 to 995.24 mg 100 g−1). Those results were consistent with ours although no colour effect has been observed and just BNP stood out from the rest. Quinoa has higher potassium contents than other cereals and pseudocereals of interest as corn, wheat, rice and barley (USDA 2017) and amaranth (Nascimento et al. 2014). It has been reported that even after quinoa grains were cooked, potassium and phosphorus remained the main elements (Mota et al. 2016). The commercial samples RCP and BCP had the higher calcium content, with no grain colour effect observed. Other studies had reported broader ranges in the concentration of this mineral. Varieties analyzed by Miranda et al. (2012) ranged from 25.15 to 116.60 mg 100 g−1, whereas Prado et al. (2014) reported a range from 77.10 to 211.29 mg 100 g−1. The saponin removal process could decreased the percentage of calcium because this mineral is more concentrated in the pericarp followed by the embryo and perisperm (Konishi et al. 2004). Magnesium was present in more concentration than reported for other quinoa varieties (Miranda et al. 2012; Prado et al. 2014) but close to 250 mg 100 g−1, informed by Kozioł (1992). However, it has been recorded values as high as 502.0 mg 100 g−1 (Konishi et al. 2004). In comparison, the magnesium content of hard red winter wheat and hulled barley was reported as 126 and 35 mg 100 g−1 respectively (USDA 2017). Although there is no recommended intake intake for sulphur, and that deficiencies of this mineral are not known (Sizer and Whitney 2003), our results were consistent with those of Bruin (1964), who reported a range from 150 to 220 mg 100 g−1. The genetic characteristics influence the mineral composition of the quinoa grains, as well as the type of soil where they are cultivated due to the chemical or mineral composition of the soils and fertilizers used (Vega-Gálvez et al. 2010; Nascimento et al. 2014). In relation to the Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 81 microminerals evaluated (Table 4), it was observed that the iron content was in the range previously reported by Miranda et al. (2012), who reported a range from 4.82 to 7.19 mg 100 g−1, except for BCP which stood out from the rest with 17.2 mg 100 g−1. However, Prado et al. (2014) reported a wider range from 4.76 to 24.04 mg 100 g−1, evidencing the great variation that can exist between the different samples of quinoa grains, as well as in this study. It was reported that the hard red winter wheat and hulled barley had just 3.19 and 3.60 mg 100 g−1 of iron respectively, whereas the raw medium-grain white rice had 0.80 mg 100 g−1 (USDA 2017). The zinc content in the samples was similar to the range reported by Miranda et al. (2012) (from 2.73 to 5.01 mg 100 g−1) and Prado et al. (2014) (from 1.65 to 4.22 mg 100 g−1). One white sample (WSyB) had more zinc content than the red and black samples although the other white sample (WSaP) show a similar value to the black ones, so it was not possible to establish a colour-based relationship with our data. Red samples (RPP and RCP) had the highest boron content. These results were similar to those of Bruin (1964), who reported a range from 0.92 to 1.29 mg 100 g−1, but lower than those of Karyotis et al. (2003) who reported a completely different range (from 3.4 to 4.7 mg 100 g−1). Boron is considered to be essential for animals, but the requirement for humans is still under study (Sizer and Whitney 2003). As far as the authors know, the presence of aluminium in quinoa was analysed previously only by Bruin (1964), who reported a range from 6 to 10 mg 100 g−1. Our results are lower than those but since aluminium is not known to have a biological function in animal or human organisms (Stahl et al. 2017), it is not considered an essential nutrient but an anti-nutrient. A black sample (BCP) had the higher copper content although the other black showed a similar value than the ones of different colour. Overall, our results were in the range reported by Prado et al. (2014) (from 0.35 to 1.12 mg 100 g−1) for varieties harvested in Bolivia. However, copper content of quinoa harvested in Chile showed a different and higher range: from 0.75 to 1.52 mg 100 g−1 (Miranda et al. 2012). Both black samples (BNP and BCP) presented higher content of manganese than the red ones and lastly by the white ones. Our results for black and red samples were higher than those reported by Prado et al. (2014) (from 1.55 to 3.85 mg 100 g−1) and the black samples even surpassed the levels reported by Miranda et al. (2012) (from 2.36 to 6.47 mg 100 g−1). With regard to sodium and lithium content, black and red samples showed the highest levels. In the case of sodium, our results were higher than those by Prado et al. (2014) (from 1.74 to 7.33 mg 100 g−1) and Karyotis et al. (2003) (0.34 to 2.13 mg 100 g−1) but lower than those reported by Bruin (1964) (from 11 to 22 mg 100 g−1). Regarding the lithium, both commercial samples (BCP and RCP) had the highest content. Nascimento et al. (2014) observed lithium contents of 7.95 g 100 g−1, being higher than those found in the present study. However, grains of Bolivia quinoa showed levels as high as 7.5 mg 100 g−1 (Figueroa et al. 2013). These authors have attributed the high lithium content to the Salar de Uyuni (Bolivia), which is considered to be among the areas with the highest lithium content in the world, so the dissimilar data found in the literature on the micromineral contents in quinoa grains could probably be explained by the type of soil and climate from which the cultivars come from. Table 5. Content of total phenolic compounds present in quinoa grains. Six replicates. Values followed by the same letter in the column do not differ (Tukey test, p < 0.05). Samples TPC [mg GAE 100 g-1] RPP 61.1±3.9bc RCP 65.4±4.0b BNP 95.9±3.6a BCP 55.5±3.7c WSyB 66.4±5.6b WSaP 59.6±4.4bc Content of total phenolic compounds One of the black samples (BNP) showed the higher content of phenolic compounds (Table 5). The RPP cultivar was analysed reporting a content of 53.8 mg GAE 100 g−1 (Repo de Carrasco and Zelada 2008), lower than that found in the present study. Fifteen quinoa cultivars evaluated by the same authors presented values ranging from 35.3 Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 82 to 139.9 mg GAE 100 g−1, a range that encompasses our results. Other studies presented values of phenolic compounds above 3.75 mg GAE g−1, using a different methodology for the extraction process (using methanol and then acetone as extraction solvents, as reported by Paśko et al. (2009). Repo-Carrasco-Valencia et al. (2011) evaluated four different quinoa (Blanca de Juli, Kcancolla, La Molina and Sajama) cultivars from Peru and observed phenolic compounds values higher than ours ranging from 142 to 197 mg GAE 100 g−1. It has been noted that the content of secondary plant metabolites, such as phenolic compounds, varies from generation to generation, depending on environmental factors (Verma and Shukla 2015). These factors include soil nutrients and rainfall (Borges et al. 2013). Because our samples were obtained from different locations in a different year, the effect of these factors on phenolic compounds should be considered, as some plant species have been reported to produce more phenolic compounds and antioxidant activity in highland and semi-arid climates (Kumar et al. 2017). It was expected that the black and red samples had a higher TPC value than the white ones since a previous study reported that an improvement in TPC was accompanied by an increase in the pigments content in quinoa grains (Abderrahim et al. 2015). The same authors reported a negative correlation between TPC and lightness (L) value (r = -0.619, p = 0.024). However, our results did not allow us to confirm this relationship because the other black sample (BCP) had the lowest value of TPC, so other factors besides pigmentation must have influence on the content of phenolic compounds in quinoa grains. In this regard, Escribano et al. (2017) quantified some pigments (amaranthin, iso-amaranthin, betanin, iso-betanin, dopaxanthin, dopamine-BX, proline-BX and one unknown-BX) of RPP, BNP, WSaP and 26 other quinoa varieties of different colors, detecting a concentration of amaranthin and iso-amaranthin of 0.8 mg kg-1 (for each pigment) for RPP only, while the pigment content of the WSaP and BNP varieties was undetectable, although the colour varieties showed higher antioxidant activity. However, Tang et al. (2015b) presented data showing that the TPC content in Canadian commercial samples of dark-coloured quinoa is greater than that of white samples, with ferulic acid being the main contributor to the TPC value of black samples. Therefore, it would be necessary to further quantify the content of ferulic acid and other phenolic compounds to elucidate whether the storage conditions or genotype of the present BCP sample are responsible for its low TCP value. Evaluation of antioxidant capacity The antioxidant capacity of the studied quinoa samples, evaluated by the DPPH method are shown in Table 6. It was previously reported that the antioxidant activity of the black quinoa grain samples was approximately 5.6 μM Trolox equivalent g−1, while for the white and red samples it was between 4.4 and 4.8 μM Trolox equivalent g−1 (Tang et al. 2015a), respectively. Our results shown a similar trend, being the coloured grains those with the highest antioxidant activities. However, an unidentified Bolivian quinoa sample presented 38.84 μM Trolox equivalent g−1 (Paśko et al. 2009), a value much higher than ours. A wide range of activity results was also reported by Repo de Carrasco and Zelada (2008), who evaluated the antioxidant capacity of 15 quinoa cultivars by the DPPH method, obtained values ranging from 117.5 to 2,400.5 μg Trolox g−1, although the grain colour was not specified. The same authors presented values for amaranth in the range of 556.49 to 660.90 μg Trolox g−1. For six varieties of barley, Ondrejovič et al. (2014) determined antioxidant activity of 200 to 1,400 µg Trolox g−1. Regarding the percentage of inhibition, the black and red samples had higher values than the white ones which also presented the lowest contents of phenolic compounds and antioxidant capacity. Analyzing the results of antioxidant capacity determined by the ABTS assay (Table 6), it can be observed that one white sample (WSyB) presented the highest antioxidant capacity, followed by one of the black samples (BCP). Repo-Carrasco- Valencia et al. (2011) studied four light-coloured quinoa cultivars (Blanca Juli, Kcancolla, La Molina 89 and Sajama) which presented higher values from 2,351.9 to 3,689.5 μg Trolox g−1, being the yellow cultivar (La Molina 89) the one with the highest value. Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 83 Overall the results of antioxidant capacity are lower than those presented in the literature for quinoa. These large differences with respect to our results may be partly explained by the different extraction Table 6. Antioxidant activity of quinoa grains using DPPH and ABTS and percentage of inhibition. Six replicates. Values followed by the same letter in the row do not differ (Tukey test, p < 0.05). Quinoa samples RPP RCP BNP BCP WSyB WSaP DPPH [µM TEAC g-1] 5.67±0.29b 4.35±0.20c 5.99±0.33ab 6.18±0.31a 3.33±0.39d 1.95±0.16e DPPH [µg TEAC g-1] 1,419.1 1,088.8 1,499.2 1,546.8 833.5 488.1 Inhibition [%] 45.0±2.4a 34.4±1.6b 47.6±2.7a 47.4±2.6a 23.5±3.3c 12.0±1.3d ABTS [µM TEAC g-1] 2.02±0.19c 1.79±0.07d 1.98±0.10c 2.25±0.08b 2.78±0.14a 1.95±0.12cd ABTS [µg TEAC g-1] 505.6 448.0 495.6 563.2 695.8 488.1 Inhibition [%] 12.92±1.85bc 10.59±0.71d 12.45±0.98c 14.59±0.39b 18.67±1.42a 10.32±1.25d methodologies, so we could not accurately estimate the effect of colour. Conclusions This study has determined several nutritional characteristics of quinoa varieties of different colours (black, red and white) obtained from different sites (Arequipa and Puno regions, in Peru) in different harvest years (2012 and 2014). Physically, the red samples were larger and heavier than the white ones, and these in turn were larger and heavier than the black ones, confirming previous reports. Regarding the main anti-nutrient, one of the white samples (WSyP) had the highest content of saponins while BCP and RPP had a content of less than 0.01 % db. The lipid and protein content were similar for all the samples but black samples presented the lowest starch content. Black and red samples had more content of ashes and dietary fibre than the white ones indicating a possible correlation just for this two components. Samples of red and black colour stood out for their macrominerals content, especially RCP for Mg and BNP for K. The microminerals analysis revealed the same trend, with BNP standing out in most results although it was matched in aluminium content by WSyB. This white sample had the highest zinc content. The black samples showed the highest and lowest levels of phenolic compounds, while the red and white samples had similar results. Regarding the antioxidant activity, black and red samples showed the highest activities, highlighting BCP using the DPPH method. Using the ABTS method there was no clear differentiation, although WSyB showed increased antioxidant activity and inhibition percentage. Thus, the relationship between the colour of the samples and the antioxidant activity could not be demonstrated with our results, although further studies are needed to assess the effect of environmental conditions on the concentration of secondary metabolites of interest, such as phenolic compounds. However, all the quinua samples were found to be a good source of nutrients. References Abderrahim F, Huanatico E, Segura R, Arribas S, Gonzales MC, Condezo-Hoyos L (2015) Physical features, phenolic compounds, betalains and total antioxidant capacity of coloured quinoa seeds (Chenopodium quinoa Willd.) from Peruvian Altiplano. Food Chem. 183: 83- 90. Aluwi NA, Murphy KM, Ganjyal GM (2017) Physicochemical characterization of different varieties of quinoa. Cereal Chem. 94: 847-856. Ando H, Chen Y-C, Tang H, Shimizu M, Watanabe K, Mitsunaga T (2002) Food components in fractions of quinoa seed. Food Sci. Technol. Res. 8: 80-84. AOAC (2005) Official methods of analysis of AOAC International. 18th edition: AOAC International. Gaithersburg, 1141 pp. Apaza V, Cáceres G, Estrada R, Pinedo R (2013) Catálogo de variedades comerciales de quinua en el Perú. FAO / INIA, Lima, 79 pp. Borges LL, Alves SF, Sampaio BL, Conceição EC, Bara MTF, Paula JR (2013) Environmental factors affecting the concentration of phenolic compounds in Myrcia tomentosa leaves. Braz. J. Pharmacogn. 23: 230-238. Brand-Williams W, Cuvelier M, Berset C (1995) Use of a free radical method to evaluate antioxidant activity. LWT - Food Sci. Tech. 28: 25-30. Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 84 Bruin AD (1964) Investigation of the food value of quinoa and Cañihua seed. J. Food Sci. 29: 872-876. Demir MK, Kilinç M (2017) Utilization of quinoa flour in cookie production. Int. Food Res. J. 24: 2394-2401. Escibrano J, Cabanes J, Jiménoz-Atiénzar M, Ibañez- Tremolada M, Gómez-Pando L R, García-Carmona F, Gandía-Herrero F (2017) Characterization of betalains, saponins and antioxidant power in differently colored quinoa (Chenopodium quinoa) varieties. Food Chem. 234: 285-294. Figueroa LT, Razmillic B, Zumeata O, Aranda GN, Barton SA, Schull WJ, Young AH, Kamiya YM, Hoskins JA, Ilgren EB (2013) Environmental lithium exposure in the North of Chile – II. Natural food sources. Biol. Trace Elem. Res. 15:122-131. Karyotis T, Iliadis C, Noulas C, Mitsibonas T (2003) Preliminary research on seed production and nutrient content for certain quinoa varieties in a saline-sodic soil. J. Agron. Crop Sci. 189: 402-408. Konishi Y, Hirano S, Tsuboi H, Wada M (2004) Distribution of minerals in quinoa (Chenopodium quinoa Willd.) Seeds. Biosci. Biotech. Bioch. 68: 231-234. Kozioł MJ (1990) Desarrollo del método para determinar el contenido de saponinas en la quinua. In Wahli C (Ed.), Quinua: Hacia su cultivo comercial. Latinreco, Quito, 175-185. Kozioł M (1992) Chemical composition and nutritional evaluation of quinoa (Chenopodium quinoa Willd.). J. Food Compos. Anal. 5: 35-68. Kumar S, Yadav A, Yadav M, Yadav JP (2017) Effect of climate change on phytochemical diversity, total phenolic content and in vitro antioxidant activity of Aloe vera (L.) Burm.f. BMC Res. Notes. 10: 1-12. Lindeboom N (2005) Studies on the characterization, biosynthesis and isolation of starch and protein from quinoa (Chenopodium quinua Willd.). PhD dissertation, University of Saskatchewan, Saskatchewan, 170 pp. Malavolta E, Vitti GC, Oliviera S A. D (1997) Avaliação do estado nutricional das plantas: principios e aplicações. 2nd edition, Associação Brasileira para Pesquisa da Potassa e do Fosfato, Piracicaba, 319 pp. Malinow MR, Mcnulty WP, Houghton DC, Kessler S, Stenzel P, Goodnight Jr. SH, Palotay JL, Mclaughlin P, Livingston AL (1982) Lack of toxicity of alfalfa saponins in cynomolgus macaques. J. Med. Primatol. 11: 106-118. Medina W, Skurtys O, Aguilera JM (2010) Study on image analysis application for identification quinoa seeds (Chenopodium quinoa Willd) geographical provenance. LWT - Food Sci. Technol. 43: 238-246. Miranda M, Vega-Gálvez A, Quispe-Fuentes I, Rodríguez, MJ, Maureira H, Martínez EA (2012) Nutritional aspects of six quinoa (Chenopodium quinoa Willd.) ecotypes from three geographical areas of Chile. Chil. J. Agr. Res. 72: 175-181. Mota C, Nascimento AC, Santos M, Delgado I, Coelho I, Rego A, Matos AS, Torres D, Castanheira I (2016) The effect of cooking methods on the mineral content of quinoa (Chenopodium quinoa), amaranth (Amaranthus sp.) and buckwheat (Fagopyrum esculentum). J. Food Compos. Anal. 49: 57-64. Nascimento AC, Mota C, Celaho I, Gueifão S, Santos M, Matos AS, Gimenez A, Lobo M, Samman N, Castanheira I (2014) Characterisation of nutrient profile of quinoa (Chenopodium quinoa), amaranth (Amaranthus caudatus), and purple corn (Zea mays L.) consumed in the North of Argentina: Proximates, minerals and trace elements. Food Chem. 148: 420-426. Navruz-Varli S, Sanlier N (2016) Nutritional and health benefits of quinoa (Chenopodium quinoa Willd.). J. Cereal Sci. 69: 371-376. Ondrejovič M, Chmelová D, Ivanišová E, Dráb Š, Psota V (2014) Evaluation of antioxidant activities of cereals and their malts. Nova Biotechnol. Chim. 13: 172-181. Paśko P, Bartoń H, Zagrodzki P, Gorinstein S, Fołta M, Zachwieja Z (2009) Anthocyanins, total polyphenols and antioxidant activity in amaranth and quinoa seeds and sprouts during their growth. Food Chem. 115: 994- 998. Prado FE, Fernández-Turiel JL, Tsarouchi M, Psaras GK, González JA (2014) Variation of seed mineral concentrations in seven quinoa cultivars grown in two agroecological sites. Cereal Chem. 91: 453-459. Re R, Pellegrini N, Proteggente A, Pannala A, Yang M, Rice- Evans C (1999) Antioxidant activity applying an improved ABTS radical cation decolorization assay. Free Radical Bio. Med. 26: 1231-1237. Repo De Carrasco R, Zelada CRE (2008) Determinación de la capacidad antioxidante y compuestos fenólicos de cereales andinos: quinua (Chenopodium quinoa), kañiwa (Chenopodium pallidicaule) y kiwicha (Amaranthus caudatus). Rev. Soc. Quím. Perú. 74: 85-99. Repo-Carrasco-Valencia R, Hellström JK, Pihlava J-M, Matilla PH (2010) Flavonoids and other phenolic compounds in Andean indigenous grains: Quinoa (Chenopodium quinoa), kañiwa (Chenopodium pallidicaule) and kiwicha (Amaranthus caudatus). Food Chem. 120: 128-133. Repo-Carrasco-Valencia RA-M, Sern LA (2011) Quinoa (Chenopodium quinoa, Willd.) as a source of dietary fiber and other functional components. Ciência Tecnol. Alime. 31:225-230. Ruales J, Nair BM (1992) Nutritional quality of the protein in quinoa (Chenopodium quinoa, Willd) seeds. Plant Food. Hum. Nutr. 42: 1-11. Singleton VL, Orthofer R, Lamuela-Raventós RM (1999) Analysis of total phenols and other oxidation substrates and antioxidants by means of folin-ciocalteu reagent. In Packer L (Ed.), Methods in Enzymology, Academic Press, Oxford., p. 152-178. Sizer F, Whitney E (2003) Nutrição: Conceitos e controversias. 8th edition, Manole, São Paulo, 567 pp. Southon S, Wright AJA, Price KR, Fairweather-Tait SJ, Fenwick GR (1988) The effect of three types of saponin on iron and zinc absorption from a single meal in the rat. Brit. J. Nutr. 59: 389-396. Souza LAC, Spehar CR, Santos RLB (2004) Análise de imagem para determinação do teor de saponina em Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC Nova Biotechnol Chim (2018) 17(1): 74-85 85 quinoa. Pesqui. Agropecu. Bras. 39: 397-401. Spehar CR (2006) Adaptação da quinoa (Chenopodium quinoa Willd.) para incrementar a diversidade agrícola e alimentar no Brasil. Cad. Ciência Tecnol. 23: 41-62. Spehar CR, Rocha JEDS, Santos RLDB (2011) Desempenho agronômico e recomendações para cultivo de quinoa (BRS Syetetuba) no cerrado. Pesqui. Agropecu. Trop. 41: 145-147. Stahl T, Falk S, Rohrbeck A, Georgii S, Herzog C, Wiegand A, Hotz S, Boschek B, Zorn H, Bunn H (2017) Migration of aluminum from food contact materials to food – a health risk for consumers? Part I of III: exposure to aluminum, release of aluminum, tolerable weekly intake (TWI), toxicological effects of aluminum, study design, and methods. Environ. Sci. Eur. 29: 19. Tang Y, Li X, Chen PX, Zhang B, Hernandez M, Zhang H, Marcone MF, Liu R, Tsao R (2015a) Characterisation of fatty acid, carotenoid, tocopherol/tocotrienol compositions and antioxidant activities in seeds of three Chenopodium quinoa Willd. Genotypes. Food Chem. 174: 502-508. Tang Y, Li X, Zhang B, Chen PX, Liu R, Tsao R (2015b) Characterisation of phenolics, betanins and antioxidant activities in seeds of three Chenopodium quinoa Willd. genotypes. Food Chem. 166: 380-388. USDA (2017) Food composition databases. United States Department of Agriculture, Washington D.C. Vega-Gálvez A, Miranda M, Vergara J, Uribe E, Puente L, Martínez EA (2010) Nutrition facts and functional potential of quinoa (Chenopodium quinoa willd.), an ancient Andean grain: A review. J. Sci. Food Agr. 90: 2541-2547. Verma N, Shukla S (2015) Impact of various factors responsible for fluctuation in plant secondary metabolites. J. Appl. Res. Med. Arom. Plants. 2: 105- 113. Bereitgestellt von Slovenská poľnohospodárska knižnica | Heruntergeladen 28.02.20 09:30 UTC