Food Chemistry 183 (2015) 8390

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Physical features, phenolic compounds, betalains and total antioxidant

capacity of coloured quinoa seeds (Chenopodium quinoa Willd.)
from Peruvian Altiplano
Fatima Abderrahim a, Elizabeth Huanatico b, Roger Segura b, Silvia Arribas a, M. Carmen Gonzalez a,
Luis Condezo-Hoyos a,
a
b

Universidad Autnoma de Madrid, Facultad de Medicina, Madrid, Spain

Universidad Nacional del Altiplano, Facultad de Agroindustria, Puno, Peru

a r t i c l e

i n f o

Article history:
Received 16 October 2014
Received in revised form 6 March 2015
Accepted 10 March 2015
Available online 17 March 2015
Keywords:
Quinoa (Chenopodium quinoa Willd.)
Phenolic compounds
Betalains
Total antioxidant capacity
QUENCHER procedure
Peruvian Altiplano

a b s t r a c t
Physical features, bioactive compounds and total antioxidant capacity (TAC) of coloured quinoa varieties
(Chenopodium quinoa Willd.) from Peruvian Altiplano were studied. Quinoa seeds did not show a pure red
colour, but a mixture which corresponded to different fractal colour values (51.071.8), and they varied
from small to large size. Regarding bioactive compounds, total phenolic (1.233.24 mg gallic acid equivalents/g) and avonol contents (0.472.55 mg quercetin equivalents/g) were highly correlated (r = 0.910).
Betalains content (0.156.10 mg/100 g) was correlated with L colour parameter (r = 0.569), total phenolics (r = 0.703) and avonols content (r = 0.718). Ratio of betaxanthins to betacyanins (0.01.41) was
negatively correlated with L value (r = 0.744). Whereas, high TAC values (119.8335.9 mmol Trolox
equivalents/kg) were negatively correlated with L value (r = 0.779), but positively with betalains
(r = 0.730), as well as with free (r = 0.639), bound (r = 0.558) and total phenolic compounds (r = 0.676).
Unexploited coloured quinoa seeds are proposed as a valuable natural source of phenolics and betalains
with high antioxidant capacity.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
Quinoa (Chenopodium quinoa Willd.) plants from the Andean
region in South America possess a huge genetic variability which
allows its adaptation and growth under adverse environmental
conditions, such as drought, hail, frost and high altitude (Aguilar
& Jacobsen, 2003; Jacobsen, 2003). Quinoa seed has been recognized as an extremely nutritious grain all over the world, due to
both its relatively high amount (compared to cereals) and the ne
quality of its proteins, as regards essential amino acids content;
quinoa seed also contains essential fatty acids and minerals
(Vega-Galvez, Miranda, Vergara, Uribe, Puente, & Martinez,
2010). Besides, its high total antioxidant capacity (Hirose, Fujita,
Ishii, & Ueno, 2010; Repo-Carrasco-Valencia, Hellstrm, Pihlava,
& Mattila, 2010; Tang et al., 2015) is related to its high phenolic
content which can vary depending on genetic variability and
environmental conditions (Repo-Carrasco-Valencia et al., 2010).
Corresponding author.
E-mail address: lcondezotm@hotmail.com (L. Condezo-Hoyos).
http://dx.doi.org/10.1016/j.foodchem.2015.03.029
0308-8146/ 2015 Elsevier Ltd. All rights reserved.

Thus, nutritional composition of quinoa seeds and their bioactive

compounds (Fischer, Wilckens, Jara, & Aranda, 2013), such as
phenolics (Hirose et al., 2010; Repo-Carrasco-Valencia et al.,
2010) and betalains (Tang et al., 2015), can differ among ecotypes
(groups of cultivars dened according to distributional, ecological,
agronomic and morphological criteria). Phenolic antioxidants in
quinoa seeds might be present as free but also as bound forms
attached to cell wall structures (Abderrahim et al., 2012; AcostaEstrada, Gutirrez-Uribe, & Serna-Saldvar, 2014; Serpen,
Gkmen, Pellegrini, & Fogliano, 2008). In this regard, it has been
pointed out that previous studies might not reect well the real
antioxidant capacity of quinoa seeds (Hirose et al., 2010;
Miranda et al., 2011; Yawadio Nsimba, Kikuzaki, & Konishi,
2008). Thus, our aim was to identify, classify and characterize thirteen unexploited red-coloured quinoa seeds from Peruvian
Altiplano, mainly based on their physical features (colour, size,
shape) and bioactive compounds (free and bound phenolics, betalain pigments), in connection with their total antioxidant capacity
measured by direct QUENCHER-CUPRAC procedure and as a potential source of natural antioxidants for the functional food market.

84

F. Abderrahim et al. / Food Chemistry 183 (2015) 8390

2. Materials and methods

2.1. Chemicals
2-aminoethyl diphenyl borate, 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox), bathocuproine disulfonic acid
disodium salt (BCS), copper sulphate, ethylenediaminetetraacetic
acid disodium salt dihydrate (EDTA-Na2), gallic acid, and quercetin
were obtained from SigmaAldrich (Spain). Acetone, Folin
Ciocalteau phenol reagent, methanol and sodium carbonate were
purchased from Merck (Spain). Sulphuric acid was purchased from
Probus S.A. (Spain).
2.2. Coloured quinoa samples
Thirteen Peruvian Altiplano coloured quinoa seeds (Chenopodium
quinoa Willd.) were harvested in March 2014 from cultivated plants
located at different provinces from Puno, Peru (Supplementary
Table 1). Common Peruvian Altiplano names of the plants are:
M1 = Cuchiwilla morado, M2 = Quinua rosada, M3 = Quinua chica,
M4 = Pasankalla dorado, M5 = Quinua real rosado, M6 = Pasankalla
rojo, M7 = Quinua rosado, M8 = Cuchiwilla roja, M9 = Ayrampo,
M10 = Panela, M11 = Witulla roja, M12 = Cuchiwilla puaquinua and
M13 = Vilacuyo. Whole quinoa seeds were used for colour measurement, and morphological characterization, meanwhile chemical
evaluation was performed on ground (commercial coffee blender,
Endecotts Ltd., London, 102 England) and sieved (100 mm mesh)
samples. Whole and nely powdered quinoa samples were packed
in a sealed container and were stored at 20 C until analysis.
2.3. Physical features of Peruvian Altiplano coloured quinoa
2.3.1. Image acquisition
Twenty-ve top images of coloured quinoa seeds were
simultaneously acquired with an HP PSC 1510 scanner at their
maximum resolution (HP, NY, USA) into an adapted black box to
avoid entering the external light in order to get the best image quality. A calibrated ruler was included in the image acquisition to help
calculate the colour and morphological features of the samples.
2.3.2. Colour features
Red (R), green (G) and blue (B) signals were obtained by image
analyses using the colour histogram tool of Image J programme
(http://rsb.info.nih.gov/ij/). RGB values were converted into hue
H-saturation S-lightness L-values using the following equation:
H = Arctan [(1.73/2) (GB)]/[R 0.5(G + B)], L = (R + G + B)/3, S = 1
lowest value of any ratio: R/L, G/L, or B/L (Condezo-Hoyos,
Mohanty, & Noratto, 2014; Medina, Skurtys, & Aguilera, 2010)
and colour of the quinoa samples were represented in HSL space
colour (polar graph). In addition, fractal colour parameter was
determined using harmonic and fractal image analyser software
HarFA
5.5.30
(http://www.fch.vutbr.cz/lectures/imagesci/).
Briey, red colour channel was split from each image and thresholded for colour values between 0 and 255. Then, the fractal
dimension (FD) was determined by the box counting method and
presented as a function of thresholding condition, which is called
fractal spectrum. After drawing the baseline (when FD = 1), area
under the curve from fractal spectra-fractal colour parameter
was calculated as previously reported (Lu, Zheng, Hu, Lou, &
Kong, 2011).
2.3.3. Morphological features
Morphological features of coloured quinoa seeds were analyzed
using Image J software (http://imagej.nih.gov/ij/). Firstly, RGB
images were transformed to binary format and dimensionally

calibrated by set scale option and thereafter, basic morphological

features area, perimeter, major and minor diameters- and shape
descriptors circularity, round and solidity- were estimated by
particle analyses procedure from Image J.
2.4. Free, bound and total phenolics and avonols
2.4.1. Extraction procedure
Free and bound phenolics and avonols from the coloured quinoa samples were extracted by the previously reported procedure
for pseudo-cereals (Abderrahim et al., 2012). Briey, samples were
successively extracted with acidic (HCl)-methanol/water (50:50,
v/v, pH 2) and acetone/water solutions (70:30, v/v) (40 g/L) on a
movil rod shaker (J.P. Selecta S.A, Barcelona, Spain) at maximum
speed for 1 h (room temperature). Supernatants, containing free
phenolics and avonols, were collected by centrifugation at
2100g for 10 min (4 C). Thereafter, the pellets from previous
step, were extracted with methanol/H2SO4 (90:10, v/v) at 85 C
for 10 h and then centrifuged at 2100g for 10 min (4 C) to obtain
the bound phenolics fraction. A solvent/solid ratio equal to 100 was
used for all extraction steps, and the extracts were protected from
light by covering them with aluminium foil. Aliquots of all extracts
were stored at 20 C until analysis.
2.4.2. Quantication of total phenolics
Total phenolics were assessed by the Folin Ciocalteu (FC) micromethod adapted to a microplate reader (Abderrahim et al., 2011).
Briey, 10 lL of the sample or gallic acid standard solution (0
0.2 mg/mL) was pipetted in triplicate into separate wells from a
96-well microplate, and 10 lL of the FC reagent was added to
each, thoroughly mixed, and equilibrated at room temperature
for 2 min. Thereafter, 80 lL of 5% (w/v) sodium carbonate was
added and the reaction was carried out in a water bath at 40 C
for 10 min. The 96-well microplate was cooled to room temperature and the absorbance read at 740 nm in a Synergy HT MultiMode Microplate Reader (Biotek, Rochester, VT, USA). Free, bound
and total phenolics content was expressed as mg Gallic Acid
Equivalents (GAE) per gram of wet sample (r2 = 0.9972).
2.4.3. Quantication of total avonols
Total avonols were assessed by a colorimetric assay adapted
from Oomah and Mazza (Oomah & Mazza, 1996). Briey, 10 lL
of extracts were mixed with 100 lL 2-aminoethyl diphenyl borate
(0.125% w/v in methanol) at room temperature for 10 min on a 96well microplate and the absorbance was measured at 405 nm in a
microplate reader (Abderrahim et al., 2012). Flavonols content
was estimated from a quercetin standard calibration curve
(00.25 mM) and expressed as mg of Quercetin Equivalent (QE)
per gram of wet sample.
2.5. Betalains
2.5.1. Extraction procedure
Betalains from the coloured quinoa samples were extracted
with water/methanol solution (80:20 v/v) (20 g/L) under constant
shaking at maximum speed on a movil rod shaker as above for
30 min (room temperature). The samples were centrifuged at
2100g for 10 min at 4 C to collect supernatant containing betalains (Swarna, Lokeswari, Smita, & Ravindhran, 2013). Aliquots
were stored at 20 C until analysis.
2.5.2. Quantication
Betalains content was determined as the sum of betacyanin and
betaxanthin pigments by the spectrophotometric multiple-component method of Nilsson (1970), adapted to a microplate reader. The
measurements of absorption spectrum from the extracts were

F. Abderrahim et al. / Food Chemistry 183 (2015) 8390

performed in triplicate at 400700 nm every 1 nm. A residuals

procedure was used for peak separation and analysis (betacyanins
and betaxanthins) using a linear baseline, FFT lter for smoothing and Gauss peak type model from PeakFit v.4.12 software
(Systat Software, Inc., USA). Betacyanin and betaxanthin contents
were quantied from estimated absorbance, after peak separation, by using molar extinction coefcients of 60,000 and
48,000 L mol 1 cm 1, respectively. Regarding chemical nature of
pigments from quinoa samples presence of betacyanins was
conrmed by reaction of extracts (200 lL) with 2 M NaOH
(12.5 lL). A change from pink to yellow colour indicated that
pigment was betacyanin (Harborne, 1998).

85

1:1 v/v PPE-), 50 lL PPE, and 250 lL 0.5 mM CuSO4 (dissolved

in PPE) under shaking on a movil rod shaker at maximum speed
for 30 min (room temperature). Thereafter, 250 lL of 10 mM
EDTA-Na2 (dissolved in water) was added to stop the reaction
and the samples were centrifuged at 2100g for 10 min at room
temperature. Two hundred microlitre of supernatant was transferred to a 96-well microplate and absorbance was read at
490 nm in a microplate reader. Total antioxidant capacity was
calculated from a standard curve of Trolox (0320 lM) and
expressed as mmol of Trolox Equivalent (TE) per kg of wet sample.

2.7. Statistical analysis

2.6. Total antioxidant capacity
A modied direct QUENCHER-CUPRAC procedure was used to
perform total antioxidant capacity of coloured quinoa samples
(Campos, Guzmn, Lpez-Fernndez, & Casado, 2009; Tufan,
elik, zyrek, Gl, & Apak, 2013). Briey, 2 mg of the sample
was put on each tube and reacted with 950 lL of 0.25 mM
BCS (dissolved in 10 mM phosphate buffer, pH 7.4/ethanol

Pearson correlation analysis and principal component analysis

(PCA), for automatically standardized physical or chemical data
(on mean-centred), were performed using IBM SPSS statistics 19
(SPSS Inc., Chicago, USA). Polar graph (HSL model colour) was
performed using SigmaPlot 11.0 software (Systat Software, Inc.,
CA, USA). Area under the curve from fractal spectra was calculated
by GraphPad (GraphPad Prism 6.1, San Diego, CA, USA).

Fig. 1. Colour parameters of coloured quinoa seeds from Peruvian Altiplano on HSL space colour. (A) Lightness (%) and (B) Saturation (%) and Hue (sexagesimal grade) were
calculated from RGB values obtained from 25 images by histogram color analysis. Common names of the samples are: M1 = Cuchiwilla morado, M2 = Quinua rosada,
M3 = Quinua chica, M4 = Pasankalla dorado, M5 = Quinua real rosado, M6 = Pasankalla rojo, M7 = Quinua rosado, M8 = Cuchiwilla roja, M9 = Ayrampo, M10 = Panela,
M11 = Witulla roja, M12 = Cuchiwilla puaquinua and M13 = Vilacuyo. (For interpretation of the references to colour in this gure legend, the reader is referred to the web
version of this article.)

86

F. Abderrahim et al. / Food Chemistry 183 (2015) 8390

3. Results and discussion

3.1. Physical features of coloured quinoa seeds
3.1.1. Colour features
Most coloured quinoa seeds showed H values ranging from +5
to 5, except for samples M7 and M13 that exhibited values of
17.5 and 10, respectively (Fig. 1). Previous studies on quinoas from
Peruvian Altiplano report a chestnut colour with H values comprised between 30 and 40 (Medina et al., 2010). However,
Nario-INIA-Pasto (NAR) variety shows an H value of 14, typically
of red color, associated with the presence of betalains in red and
dark quinoa seeds (Tang et al., 2015). The present data also conrmed that coloured quinoa samples are a source of betalains.
Most colored quinoa showed L values below or near 100
(Fig. 1A), which indicated that these samples were darker than
those from previous studies on quinoa seeds from Peruvian
Altiplano, with L values ranging from 150 to 200 (Medina et al.,
2010). Based on colour darkness the samples were classied into
four groups: very high M11 (L  25); high M1, M3, M6, M8, M9,
M12 (50 < L < 75); medium M2, M4, M5, M10 (L values between
85 and 110); and low M7, M13 (125 < L < 150). The fairly low
saturation degree (S < 40) of most coloured quinoa samples indicated that the principal colour was not pure (Fig. 1B). In addition,
an increase in darkness (lower L values) was accomplished by
enhancing at H (Pearson r = 0.713, p = 0.006) and S values
(Pearson r = 0.663, p = 0.013), which was related to an increase
in red colours intensity and impurity.
Recently, several fractal colour parameters have been applied as
classication criteria of food quality. For example in fruits, based
on whether they are healthy or bruised (Lu et al., 2011) or in biscuits, taking into account its acrylamide content (Lu & Zheng,
2012). These recent uses overcome traditional colour spaces such
as RGB. Coloured quinoa samples showed an area under the curve
from fractal spectra (fractal colour parameter) comprised between
51.0 and 71.8, with lower values for M7 and M13 samples, which
showed the highest values of lightness (Table 1 and Fig. 1). In order
to classify coloured quinoa from Peruvian Altiplano, the above
parameter could be considered as a new colour feature.
3.1.2. Morphological features
Area (2.314.71 mm2) and perimeter (5.698.17 mm) of
coloured quinoa seeds (Table 1) were closer to previous report
for quinoa varieties from Europe and Peruvian Altiplano region
(Area = 2.565.1 mm2; perimeter = 6.048.65 mm) (Medina et al.,
2010). In addition, AR (1.061.17) and circularity (0.890.90) from

Fig. 2. PCA score plot biplot for the discrimination of the coloured quinoa based on
morphological and colour features. Score plot (samples) and loading (variable
measured) are superimposed. (For interpretation of the references to colour in this
gure legend, the reader is referred to the web version of this article.)

top view showed that Peruvian Altiplanos coloured quinoa seeds

are circular in shape (Table 1), and similar to previous report on
quinoa seeds cultivated in Europe and South America (Medina
et al., 2010). This fact was conrmed by the estimated coefcient
of determination (R2 = 0.9201; p < 0.0001) from top minor and
major diameter plot (Supplementary Fig. 1). Previous study on
quinoa samples has reported that the plot of thickness from lateral
view versus major diameter from top view allows to classify them
according to their size (Medina et al., 2010). Similarly, minor and
major diameter (from view) plots have also allowed us to group
Peruvian Altiplanos coloured quinoa seeds into large, medium
and small-sized (Supplementary Fig. 1). This is consistent with
the fact that the volume of quinoa seeds (geometrically like a
cylinder) was more strongly inuenced by the major diameter than
by thickness (data not shown).
Principal component analysis (PCA) of geometrical parameters
(area, perimeter, minor and major diameter), traditional colour
model (H, S and L values) and fractal colour parameter allowed
us to classify coloured quinoa seeds into four main groups (I, II,

Table 1
Morphological and fractal colour features of coloured quinoa seeds (Chenopodium quinoa Willd.) from Peruvian Altiplano.
Sample

M1
M2
M3
M4
M5
M6
M7
M8
M9
M10
M11
M12
M13

Morphological features
Area (mm2)

Perimeter (mm)

Major diameter (mm)

Minor diameter (mm)

Circularity

AR

Fractal color parametera

2.52 0.32
2.94 0.44
2.85 0.34
4.71 0.59
3.55 0.40
3.53 0.36
3.72 0.54
2.31 0.20
2.41 0.25
4.21 0.47
2.46 0.32
3.15 0.37
3.16 0.40

5.95 0.40
6.50 0.49
6.33 0.39
8.17 0.52
7.15 0.44
7.07 0.39
7.61 0.91
5.69 0.27
5.83 0.30
7.80 0.45
5.90 0.41
6.70 0.40
6.75 0.43

1.89 0.13
2.03 0.16
1.97 0.11
2.52 0.17
2.19 0.14
2.22 0.13
2.24 0.17
1.78 0.08
1.81 0.10
2.45 0.13
1.87 0.13
2.16 0.13
2.16 0.13

1.69 0.11
1.83 0.13
1.83 0.13
2.37 0.15
2.06 0.12
2.02 0.10
2.10 0.15
1.65 0.07
1.69 0.09
2.18 0.14
1.66 0.11
1.85 0.13
1.86 0.15

0.89 0.01
0.87 0.01
0.89 0.01
0.88 0.01
0.87 0.02
0.89 0.01
0.82 0.12
0.90 0.01
0.89 0.01
0.87 0.03
0.88 0.01
0.88 0.01
0.87 0.01

1.12 0.05
1.11 0.04
1.08 0.04
1.06 0.04
1.06 0.04
1.10 0.04
1.07 0.03
1.08 0.03
1.07 0.04
1.12 0.05
1.13 0.05
1.17 0.06
1.16 0.09

56.9 1.5
68.9 3.3
59.7 1.5
64.9 4.4
51.0 0.5
65.9 2.4
70.8 4.2
60.6 4.0
60.4 4.1
62.9 5.2
57.8 6.7
61.3 3.3
71.8 7.8

Data are the mean SD (n = 25).

a
Represents the area under curve from fractal spectral for baseline fractal dimension = 1.

87

F. Abderrahim et al. / Food Chemistry 183 (2015) 8390

III, IV) with decreasing levels of both, L value and fractal colour,
respectively (Fig. 2). To the best of our knowledge, this is the rst
time that the above physical parameters have been successfully
applied to group coloured quinoa samples.
3.2. Free, bound and total phenolics and avonols in coloured quinoa
seeds
Coloured quinoa seeds from Peruvian Altiplano showed free
phenolic values comprised between 1.23 and 3.41 mg GAE/g sample (Table 2), which are comparable to those found for red quinoa
samples (4.2 mg GAE/g sample) grown in Ontario (Canada) (Tang
et al., 2015), canihua (2.5 mg GAE/g sample dry weight) another
Altiplano pseudo-cereal that belongs to the same genus as the
quinoa seed (Abderrahim et al., 2012) and different types of
millet (1.46.3 mg ferulic acid equivalent/g defatted sample)
(Chandrasekara & Shahidi, 2010). Regarding bound phenolics content, coloured quinoa seeds from Peruvian Altiplano showed values
ranging from 1.28 to 4.52 mg GAE/g (Table 2) and a ratio of free to
bound phenolics between 0.5 and 2.0 (calculated from data in
Table 2). The former values were similar to those found (0.61.6)
for other previously studied Altiplanos coloured quinoa seeds
(Pasankalla, Roja de Coporaque and Witulla) (Repo-CarrascoValencia et al., 2010). On the contrary, bound phenolics in coloured
quinoa were lower than those found for canihua seed (8.8 mg/g
sample dry weight) (Abderrahim et al., 2012) and whole millet
types (0.415.9 mg ferulic acid/g defatted sample) (Chandrasekara
& Shahidi, 2010). However, the hydrolysis conditions (2 M HCl at
85 C for 1 h) employed to release bound phenolics, could also
increase furan derivatives that at very high concentration
(>3000 mg/kg) could react with FolinCiocalteau reagent leading
to the overestimation of bound phenolic content. In the present
study an efcient hydrolysis with methanol and sulphuric acid
(instead of hydrochloric acid), avoided furan derivatives formation
and released bound phenolics by more than vefold, compared to a
previous report using alkaline hydrolysis for nonextractable
polyphenols in cereal samples (Arranz & Saura Calixto, 2010). In
fact, a positive correlation (r = 0.838; p < 0.0001) between bound
phenolics and bound avonols, conrmed the lack of interferences
for furan derivatives, as detected with 2-aminoethyldiphenyl
borate colorimetric assay.
Previous studies have described that quinoa seeds are an exceptionally rich source of avonols, such as quercetin and kaempferol
(Repo-Carrasco-Valencia et al., 2010; Tang et al., 2015). In the
present study, we have also demonstrated a very high content of
total avonols (0.472.55 mg QE/mg sample) in quinoa coloured

seeds from Peruvian Altiplano (Table 2). The above avonols content was comparable to that of total avonoids for quinoa seeds
(white, red and black) commercialized in the Canadian market
(1.55.0 mg catechin equivalent/g quinoa sample) (Tang et al.,
2015). However, avonols content found in coloured quinoa seeds
from Peruvian Altiplano was markedly higher than total avonoids
(36.2 to 144.3 mg/100 g quinoa sample) of some other coloured
quinoa seeds (Pasankalla, Roja de Coporaque and Witulla) (RepoCarrasco-Valencia et al., 2010). Differences in extraction, hydrolysis procedures and analytical techniques are likely to explain the
content of total avonols found in the present study, and total
avonoids reported by Repo-Carrasco-Valencia et al. (2010).
Similarly, total avonols of coloured quinoa seeds from Peruvian
Altiplano (1937% of total phenolics, as calculated from Table 2)
were similar to the free avonoids in white, red and black genotypes of quinoa seeds commercialized in the Canadian markets,
which reached only approximately 30% of total phenolics (Tang
et al., 2015). In contrast, a very high avonoids percentage regarding total phenolics was reported for some red quinoa seeds from
Peruvian Altiplano: roja de coporaque (69.3%), witulla (69.5%) and
03-21-0093 (54.6%) (Repo-Carrasco-Valencia et al., 2010). These
differences could be explained by the short time (30 min) and
low acid concentration (approximately 1 M HCl) employed by
Repo-Carrasco-Valencia et al. (2010), which did not allow the full
extraction of bound phenolics from quinoa samples, predominantly composed by phenolic acids in cereals (Acosta-Estrada
et al., 2014).
A very high positive association was found between total phenolics and total avonols (r = 0.91, p < 0.0001). Another interesting
negative correlation was found between phenolic compounds and
L values (value r = 0.619, p = 0.024), which reected well that
both phenolics and pigments from coloured quinoa samples
are simultaneously extracted in methanol, and that an improvement
in phenolic compounds was accompanied by an increase in pigments content. Finally, a signicant and positive association between
free avonols and free phenolics (r = 0.677, p = 0.011), bound avonols and bound phenolics (r = 0.838, p < 0.0001), and total avonols
and total phenolics (r = 0.910, p < 0.0001) was found.
3.3. Betalains content of coloured quinoa seeds
The presence of betalains in red quinoa seeds remains controversial, despite quinoa belong to the same family as amaranth,
which seed contains low levels of amaranthine type of betacyanins
(1.4 mg/100 g). Repo-Carrasco-Valencia et al. (2010) do not detect
the presence of betacyanins in some Peruvian Altiplanos red

Table 2
Free and bound phenolics, avonols and total antioxidant capacity of coloured quinoa seeds (Chenopodium quinoa Willd.) from Peruvian Altiplano.
Sample

M1
M2
M3
M4
M5
M6
M7
M8
M9
M10
M11
M12
M13

Phenolics (mg GAE/g)

Flavonols (mg QE/g)

Betalains (mg/100 g)

TAC (mmol TE/kg)

Free

Bound

Total

Free

Bound

Total

Bcy

Bx

Total

2.87 0.17
1.90 0.16
3.24 0.03
2.72 0.03
1.23 0.18
2.77 0.01
2.67 0.13
2.59 0.13
3.41 0.06
1.38 0.02
2.54 0.02
2.98 0.03
2.57 0.15

4.01 0.11
3.81 0.33
2.45 0.05
1.45 0.08
1.28 0.04
4.29 0.17
1.79 0.02
2.53 0.03
4.52 0.06
1.89 0.03
3.04 0.17
3.23 0.06
4.09 0.18

6.89 0.06
5.71 0.17
5.69 0.02
4.17 0.10
2.50 0.22
7.07 0.18
4.46 0.15
5.12 0.02
7.92 0.12
3.27 0.05
5.58 0.19
6.20 0.09
6.66 0.33

0.75 0.02
0.90 0.16
0.93 0.06
0.92 0.04
0.30 0.02
0.81 0.16
1.13 0.23
1.12 0.00
1.08 0.12
0.62 0.06
0.84 0.05
1.09 0.10
0.52 0.03

1.32 0.04
0.33 0.11
0.75 0.02
0.38 0.03
0.16 0.03
1.34 0.03
0.28 0.03
0.75 0.17
1.47 0.10
0.52 0.08
0.89 0.01
1.10 0.06
1.29 0.03

2.06 0.02
1.23 0.05
1.68 0.04
1.30 0.01
0.47 0.01
2.15 0.13
1.40 0.19
1.87 0.17
2.55 0.02
1.14 0.02
1.72 0.04
2.19 0.04
1.81 0.00

4.18 0.21
1.20 0.08
3.59 0.18
1.99 0.07
0.15 0.01
2.94 0.10
0.43 0.02
1.78 0.05
5.23 0.23
0.38 0.02
0.68 0.03
1.87 0.05
0.18 0.00

0.44 0.01
0.18 0.01
0.59 0.02
0.52 0.03
0.00 0.00
1.63 0.01
0.00 0.00
0.30 0.01
0.87 0.04
0.05 0.01
0.46 0.01
0.42 0.02
0.00 0.00

4.63 0.03
1.38 0.05
4.17 0.06
2.51 0.05
0.15 0.01
4.57 0.07
0.43 0.02
2.08 0.04
6.10 0.12
0.43 0.03
1.14 0.04
2.29 0.04
0.18 0.00

321.4 3.2
219.5 10.5
264.4 11.0
201.4 10.6
119.8 2.8
335.9 12.2
182.3 8.2
311.7 2.3
259.2 9.8
167.2 9.9
262.2 12.5
267.6 3.7
162.6 4.9

Data are the means SD (n = 3).

Bcy: betacyanins; Bx: betaxanthin; total betalains is the sum of Bcy and Bx; TAC: total antioxidant capacity measured by direct QUENCHER-CUPRAC approach.

88

F. Abderrahim et al. / Food Chemistry 183 (2015) 8390

Fig. 3. Correlation between the total antioxidant capacity measured by QUENCHER-CUPRAC procedure and L value as colour parameter, total betalains and free, bound and
total phenolics and avonols in coloured quinoa seed from Peruvian Altiplano. A parametric Pearson procedure was used and p < 0.05 denote a statistical signicance. (For
interpretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)

quinoa seeds (pasankalla, roja de coporaque and witulla). Contrarily,

UV/Vis spectra, retention time and MS data show that commercial
red and black quinoa cultivated in Canada contain betanin and isobetanin in similar doses to those found in beet root a rich source of
betalains, but no quantitative data are reported (Tang et al., 2015).
In the present study, variable betalain contents (0.156.10 mg/
100 g, expressed as the sum of betacyanins and betaxanthins)
(Table 2) have been found, for the rst time, in coloured quinoa
seeds from Peruvian Altiplano. Moreover, betaxanthin to betacyanin ratios (01.4) (Table 2) were lower than those of beet root
(0.23). The presence of betalains in all extracts from coloured
quinoa samples was qualitatively conrmed by the change in the
initial colour from red to yellow, after treatment with 2 M sodium
hydroxide (data not shown). It was also indicated by peaks at
479 5 nm and 537 4 nm in the absorption spectra, obtained
by peak separation analysis (Supplementary Fig. 2). Absorption
spectra also allowed coloured quinoa samples to be classied into
two groups by comparison with beet root extract: the rst one
with a similar spectra (M1, M2, M3, M4, M6, M8, M9, M10, M11,
M12) and the second one in which the peak related to betaxanthins
was lacking (M5, M7, M13) (Supplementary Fig. 3).
On the other hand, colour analysis has been suggested as an
efcient and practical technique to indirectly estimate the

presence of bioactive components based on a good correlation

(Fischer et al., 2013). For the rst time, this study described the
existence of negative associations between total betalains content
and L values (r = 0.569, p = 0.043), as well as between betaxanthins to betacyanins ratio and L values (r = 0.744, p = 0.004)
in Peruvian Altiplanos coloured quinoa seeds. A similar correlation
has been observed between betalains content and CIELAB parameters (r = 0.72) in fruits from different genotypes of Hylocereus sp.;
i.e., the increase in betalains content was accompanied by an
increase in the red colour of the samples. Moreover, a positive
and signicant correlation was found between betaxanthins to
betacyanins ratio and S values (r = 0.911, p < 0.001), which showed
that an increase in the betaxanthins or a decrease in betacyanins
contributed to saturate the colour of coloured quinoa samples. It
was demonstrated that coloured quinoa samples consisted of a
colour mixture related to the wide range of betaxanthins to betacyanins ratio values (S < 40) (Fig. 1B). Ultimately, a signicant
correlation between total betalains and total phenolics (r = 0.703,
p = 0.007), and total avonols (r = 0.718, p = 0.006) was also found
in coloured quinoa seeds. An increase in total phenolic index (sum
of all individual concentrations) with colour intensity has also
been found in white, red and black quinoa (Tang et al., 2015).
We hypothesize that an up-regulation of the shikimate pathway,

F. Abderrahim et al. / Food Chemistry 183 (2015) 8390

Fig. 4. PCA score plot biplot for the discrimination of the coloured quinoa based on
phenolics, betalains and total antioxidant capacity. Score plot (samples) and loading
(variable measured) are superimposed. The closer the samples are to a particular
variable the higher the positive correlation. (For interpretation of the references to
colour in this gure legend, the reader is referred to the web version of this article.)

implied in the phenolics and betalains biosynthesis, might occur

under the extreme stress conditions at the Altiplano region
(drought, cold and high altitude). Thus, an increase of the antioxidant capacity in seeds of quinoa plants subjected to drought stress
has been reported (Fischer et al., 2013).

89

antioxidants and ABTS detection probe (Tufan et al., 2013). Total

antioxidant capacity of amaranth seed measured by direct procedure has not been reported and therefore comparison with quinoa
samples is not possible. However, previous studies on antioxidant
activity of extracts have demonstrated that quinoa seeds exhibited
signicantly higher value than amaranth in different antioxidant
assays such as ABTS, DPPH (Pasko et al., 2009) and FRAP (ferric
ion reducing antioxidant power) (Alvarez-Jubete, Wijngaard,
Arendt, & Gallagher, 2010).
Interestingly, TAC of the coloured quinoa samples only showed
a signicant and negative correlation with L values (Fig. 3A), indicating that it is possible to associate the antioxidant activity with a
colour parameter obtained from a simple procedure of image
analysis. Similarly, TAC in the Peruvian Altiplanos quinoa samples
exhibited a positive association with betalains (Fig. 3B), as well as
with betacyanins (r = 0.6797, p = 0.0106) and betaxanthins content
(r = 0.7098, p = 0.0066). Furthermore, TAC of the quinoa samples
showed a signicant and positive correlation of free, bound and
total phenolic and avonols, except for free avonols that showed
a marginal correlation (Fig. 3CH). These data indicated that the
modied QUENCHER-CUPRAC using Cu (II) bathocuproine disulfonic acid disodium salt (BCS) was also able to directly measure
the antioxidant activity of both, free as well as bound phenolic
compounds, in quinoa samples. However, other antioxidants such
as vitamin E and ascorbic acid in quinoa samples (Dini, Tenore, &
Dini, 2010; Moncada, Gonzlez Martn, Escuredo, Fischer, &
Mguez, 2013) might affect the TAC obtained.
Finally, principal component analysis (PCA), based on phenolics
and betalains content as well as on TAC, allowed the establishment
of three groups of coloured quinoa samples with great potential as
antioxidant functional foods: group I, with very high values (M1,
M3, M6, M8, M9, M11 and M12; TAC = 259.2335.9 mmol TE/kg),
group II, with high values (M2, M4 and M7; TAC = 182.3
219.5 mmol TE/kg) and group III, with medium values (M5, M10
and M13; TAC = 119.8167.2 mmol TE/kg) (Fig. 4).

4. Conclusion
3.4. Total antioxidant capacity of coloured quinoa seeds
In cereal and pseudo-cereal samples, a high phenolics amount is
covalently bound to the insoluble fraction, so that they cannot be
totally extracted by solvents (Abderrahim et al., 2012; AcostaEstrada et al., 2014; Arranz, Silvan, & Saura-Calixto, 2010).
Consequently, the antioxidant capacity of cereal and pseudo-cereal
samples, as measured by based-extracts traditional procedures
might be underestimated (Serpen et al., 2008). For this reason, a
direct procedure which does not require a previous extraction step
of antioxidants and based on copper (II) oxidizing reagents (Tufan
et al., 2013) has been used to assess the antioxidant capacity of
Peruvian Altiplanos coloured quinoa samples. In agreement with
their higher phenolic and betalain content, a very high total
antioxidant capacity (TAC), comprised between 119.8 2.8 and
335.9 12.2 mmol Trolox equivalent/kg was found for coloured
quinoa samples (Table 2). TAC of quinoa samples was markedly
higher than values reported for cereals as barley (26.35 0.36),
rye (16.21 0.25), wheat (13.44 0.46) and oat (10.46 0.23),
which were also measured using a CUPRAC chromogenic reagent
QUENCHER-CUPRAC (Tufan et al., 2013). In addition, TAC for
Peruvian Altiplanos coloured quinoa seeds was even higher than
that measured for canihua seed (71.0 1.7 mmol Trolox equivalent/kg sample dry weight). Nevertheless, this value was assessed
by using ABTS method (Abderrahim et al., 2012), which can provide lower TAC values than CUPRAC chromogenic reagent, due to
steric hindrance that reduces the kinetic of the reaction between

This study demonstrated for the rst time that coloured quinoa
seeds from Peruvian Altiplano are a rich source of free and bound
phenolic compounds and betalains. Moreover, these quinoa samples showed a very high antioxidant capacity compared to cereals
and as measured by QUENCHER-CUPRAC direct procedure. Overall,
these results strongly suggest that coloured quinoa seeds grown
under extreme conditions in the Peruvian Altiplano region might
be interesting as a natural source of functional food ingredients.

Acknowledgements
The authors thank Dr. Oldrich Zmeskal, Tomas Bzatek and
Martin Nezadal from Brno University of Technology in Brno
(Czech Republic) for supplying HarFA 5.5.30 software and Dr.
Pilar Ruprez from Instituto de Ciencia y Tecnologa de Alimentos
y Nutricin (ICTAN), Consejo Superior de Investigaciones
Cientcas (CSIC), Madrid, Spain for her critical review of the
manuscript.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.foodchem.2015.
03.029.

90

F. Abderrahim et al. / Food Chemistry 183 (2015) 8390

References
Abderrahim, F., Estrella, S., Susin, C., Arribas, S. M., Gonzalez, M. C., & CondezoHoyos, L. (2011). The antioxidant activity and thermal stability of lemon
verbena (Aloysia triphylla) infusion. Journal of Medicinal Food, 14(5), 517527.
Abderrahim, F., Huanatico, E., Repo-Carrasco-Valencia, R., Arribas, S. M., Gonzalez,
M. C., & Condezo-Hoyos, L. (2012). Effect of germination on total phenolic
compounds, total antioxidant capacity, Maillard reaction products and
oxidative stress markers in canihua (Chenopodium pallidicaule). Journal of
Cereal Science, 56(2), 410417.
Acosta-Estrada, B. A., Gutirrez-Uribe, J. A., & Serna-Saldvar, S. O. (2014). Bound
phenolics in foods, a review. Food Chemistry, 152(1), 4655.
Aguilar, P. C., & Jacobsen, S.-E. (2003). Cultivation of Quinoa on the Peruvian
Altiplano. Food Reviews International, 19(12), 3141.
Alvarez-Jubete, L., Wijngaard, H., Arendt, E. K., & Gallagher, E. (2010). Polyphenol
composition and in vitro antioxidant activity of amaranth, quinoa buckwheat
and wheat as affected by sprouting and baking. Food Chemistry, 119(2),
770778.
Arranz, S., & Saura Calixto, F. (2010). Analysis of polyphenols in cereals may be
improved performing acidic hydrolysis: A study in wheat our and wheat bran
and cereals of the diet. Journal of Cereal Science, 51(3), 313318.
Arranz, S., Silvan, J. M., & Saura-Calixto, F. (2010). Nonextractable polyphenols,
usually ignored, are the major part of dietary polyphenols: A study on the
Spanish diet. Molecular Nutrition & Food Research, 54(11), 16461658.
Campos, C., Guzmn, R., Lpez-Fernndez, E., & Casado, . (2009). Evaluation of the
copper(II) reduction assay using bathocuproinedisulfonic acid disodium salt for
the total antioxidant capacity assessment: The CUPRACBCS assay. Analytical
Biochemistry, 392(1), 3744.
Chandrasekara, A., & Shahidi, F. (2010). Content of insoluble bound phenolics in
millets and their contribution to antioxidant capacity. Journal of Agriculture and
Food Chemistry, 58(11), 67066714.
Condezo-Hoyos, L., Mohanty, I. P., & Noratto, G. D. (2014). Assessing non-digestible
compounds in apple cultivars and their potential as modulators of obese faecal
microbiota in vitro. Food Chemistry, 161, 208215.
Dini, I., Tenore, G. C., & Dini, A. (2010). Antioxidant compound contents and
antioxidant activity before and after cooking in sweet and bitter Chenopodium
quinoa seeds. LWT Food Science and Technology, 43(3), 447451.
Fischer, S., Wilckens, R., Jara, J., & Aranda, M. (2013). Variation in antioxidant
capacity of quinoa (Chenopodium quinoa Will) subjected to drought stress.
Industrial Crops and Products, 46, 341349.
Harborne, J. B. (1998). Phytochemical methods: A guide to modern techniques of plant
analysis. London: Chapman A & Hall.
Hirose, Y., Fujita, T., Ishii, T., & Ueno, N. (2010). Antioxidative properties and
avonoid composition of Chenopodium quinoa seeds cultivated in Japan. Food
Chemistry, 119(4), 13001306.
Jacobsen, S.-E. (2003). The Worldwide Potential for Quinoa (Chenopodium quinoa
Willd.). Food Reviews International, 19(12), 167177.

Lu, H., & Zheng, H. (2012). Fractal colour: A new approach for evaluation of
acrylamide contents in biscuits. Food Chemistry, 134(4), 25212525.
Lu, H., Zheng, H., Hu, Y., Lou, H., & Kong, X. (2011). Bruise detection on red bayberry
(Myrica rubra Sieb. & Zucc.) using fractal analysis and support vector machine.
Journal of Food Engineering, 104(1), 149153.
Medina, W., Skurtys, O., & Aguilera, J. M. (2010). Study on image analysis application
for identication Quinoa seeds (Chenopodium quinoa Willd.) geographical
provenance. LWT Food Science and Technology, 43(2), 238246.
Miranda, M., Vega-Glvez, A., Uribe, E., Lpez, J., Martnez, E., Rodrguez, M. J., et al.
(2011). Physico-chemical analysis, antioxidant capacity and vitamins of six
ecotypes of Chilean quinoa (Chenopodium quinoa Willd.). Procedia Food Science,
1, 14391446.
Moncada, G. W., Gonzlez Martn, M. I., Escuredo, O., Fischer, S., & Mguez, M.
(2013). Multivariate calibration by near infrared spectroscopy for the
determination of the vitamin E and the antioxidant properties of quinoa.
Talanta, 116, 6570.
Nilsson, T. (1970). Studies into the pigments in beetroot (Beta vulgaris L. ssp. vulgaris
var. rubra L.). Lantbrukshogskolans Annaler, 36, 179219.
Oomah, B. D., & Mazza, G. (1996). Flavonoids and antioxidative activities in
buckwheat. Journal of Agricultural and Food Chemistry, 44(7), 17461750.
Pasko, P., Barton, H., Zagrodzki, P., Gorinstein, S., Fota, M., & Zachwieja, Z. (2009).
Anthocyanins, total polyphenols and antioxidant activity in amaranth and
quinoa seeds and sprouts during their growth. Food Chemistry, 115(3), 994998.
Repo-Carrasco-Valencia, R., Hellstrm, J. K., Pihlava, J.-M., & Mattila, P. H. (2010).
Flavonoids and other phenolic compounds in Andean indigenous grains: Quinoa
(Chenopodium quinoa), kaiwa (Chenopodium pallidicaule) and kiwicha
(Amaranthus caudatus). Food Chemistry, 120(1), 128133.
Serpen, A., Gkmen, V., Pellegrini, N., & Fogliano, V. (2008). Direct measurement of
the total antioxidant capacity of cereal products. Journal of Cereal Science, 48(3),
816820.
Swarna, J., Lokeswari, T. S., Smita, M., & Ravindhran, R. (2013). Characterisation and
determination of in vitro antioxidant potential of betalains from Talinum
triangulare (Jacq.) Willd. Food Chemistry, 141(4), 43824390.
Tang, Y., Li, X., Zhang, B., Chen, P. X., Liu, R., & Tsao, R. (2015). Characterisation of
phenolics, betanins and antioxidant activities in seeds of three Chenopodium
quinoa Willd. genotypes. Food Chemistry, 166, 380388.
Tufan, A. N., elik, S. E., zyrek, M., Gl, K., & Apak, R. (2013). Direct
measurement of total antioxidant capacity of cereals: QUENCHER-CUPRAC
method. Talanta, 108, 136142.
Vega-Galvez, A., Miranda, M., Vergara, J., Uribe, E., Puente, L., & Martinez, E. A.
(2010). Nutrition facts and functional potential of quinoa (Chenopodium quinoa
Will.), an ancient Andean grain: a review. Journal of the Science of Food and
Agriculture, 90(15), 25412547.
Yawadio Nsimba, R., Kikuzaki, H., & Konishi, Y. (2008). Antioxidant activity of
various extracts and fractions of Chenopodium quinoa and Amaranthus spp.
seeds. Food Chemistry, 106(2), 760766.