Food Chemistry 141 (2013) 3480–3485

Contents lists available at SciVerse ScienceDirect

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

Influence of taro (Colocasia esculenta L. Shott) growth conditions

on the phenolic composition and biological properties
Rui F. Gonçalves a, Artur M.S. Silva b, Ana Margarida Silva a, Patrícia Valentão a, Federico Ferreres c,
Angel Gil-Izquierdo c, João B. Silva d, Delfim Santos e, Paula B. Andrade a,⇑
a
REQUIMTE/Laboratório de Farmacognosia, Departamento de Química, Faculdade de Farmácia, Universidade do Porto, Rua de Jorge Viterbo Ferreira, n° 228, 4050-313 Porto, Portugal
b
Department of Chemistry & QOPNA, University of Aveiro, Campo Universitário de Santiago, 3810-193 Aveiro, Portugal
c
CEBAS (CSIC) Research Group on Quality, Safety and Bioactivity of Plant Foods, Department of Food Science and Technology, P.O. Box 164, Campus University Espinardo, 30100
Murcia, Spain
d
Unidade de Investigação Geobiotec, Departamento de Geociências, Universidade de Aveiro, Campo Universitário de Santiago, 3810-193 Aveiro, Portugal
e
Centro de Investigação em Ciências Farmacêuticas/Laboratório de Tecnologia Farmacêutica, Departamento de Ciências do Medicamento, Faculdade de Farmácia, Universidade do
Porto, Rua de Jorge Viterbo Ferreira, n° 228, 4050-313 Porto, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: Colocasia esculenta (L.) Shott, commonly known as taro, is an essential food for millions of people. The
Received 10 April 2013 leaves are consumed in sauces, purees, stews, and soups, being also used in wound healing treatment.
Received in revised form 31 May 2013 Nowadays, the consumers’ demand for bioactive compounds from the diet led to the development of
Accepted 3 June 2013
new agricultural strategies for the production of health-promoting constituents in vegetables. In this
Available online 12 June 2013
work, two strategies (variety choice and irrigation conditions) were considered in the cultivation of C.
esculenta. The effect on the phenolic composition of the leaves was evaluated. Furthermore, a correlation
Keywords:
between the biological activity of the different varieties and their chemical composition was established.
Colocasia esculenta (L.) Shott
Wound repair
Qualitative and quantitative differences in the phenolic composition were observed between varieties;
Oxidative/nitrosative stress furthermore, the irrigation conditions also influenced the composition. C. esculenta varieties were able
Enzymatic inhibition to scavenge several oxidant species and to inhibit hyaluronidase, but data suggest that metabolites other
Phenolic compounds than phenolics are involved. The results show that cultivation strategies can effectively modulate the
accumulation of these types of bioactive compounds. Furthermore C. esculenta wound healing potential
can be attributed, at least in part, to the protection of the wound site against oxidative/nitrosative dam-
age and prevention of hyaluronic acid degradation.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction crops being produced today. New investigation has provided the
improvement of the vegetables properties and quality through
Nowadays, the consumers’ demand for a higher content of bio- pre-harvest factors (cultivar selection, grafting), optimisation of
active compounds in the diet has increased. Bioactive metabolites, environmental conditions (light, temperature, humidity), as well
like polyphenols, are not strictly required in the diet, but when as growing conditions (soilless culture, irrigation management,
present at sufficient levels promote good health and prevent dis- salinity and fertilisation) (Padula et al., 2012; Rouphael et al.,
eases of different origins (Padula et al., 2012; Rouphael et al., 2012).
2012). Numerous studies relate the ingestion of phenolic com- Colocasia esculenta (L.) Shott is an annual herbaceous plant
pounds from the diet to a lower risk of cardiovascular diseases belonging to the Araceae family, commonly known as taro (Praja-
and development of cancers. Furthermore, they have long been pati, Kalariya, Umbarkar, Parmar, & Sheth, 2011). C. esculenta is cul-
recognised to possess anti-inflammatory, antioxidant, antiallergic, tivated mainly due to its tuber, an essential food for millions of
hepatoprotective and antiviral activities (Abad-García, Berrueta, people, and it is considered the 14th cultivated vegetable/staple
Garmón-Lobato, Gallo, & Vicente, 2009; Padula et al., 2012). There- around the world (Oscarsson & Savage, 2007). Depending on the
fore, much attention has now been placed on the agricultural prac- varieties and local cultural traditions, other plant tissues, such as
tices which will enhance the nutritional content of horticultural leaves, flowers and stems, are also consumed, especially in sauces,
purees, stews and soups (Ejoh, Mbiapo, & Fokou, 1996; Ferreres
et al., 2012b). Apart from its nutritional value, C. esculenta leaves
⇑ Corresponding author. Tel.: +351 220428654; fax: +351 226093390. and tubers are also used in the treatment of cutaneous wounds
E-mail address: pandrade@ff.up.pt (P.B. Andrade). (Agyarea et al., 2009).

0308-8146/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodchem.2013.06.009
 R.F. Gonçalves et al. / Food Chemistry 141 (2013) 3480–3485 3481

Few studies have previously addressed the phenolic composi- ethylenediamine dihydrochloride were obtained from Merck
tion of C. esculenta, anthocyanins (cyanidin and pelargonidin deriv- (Darmstadt, Germany). 2,2-Diphenyl-1-picrylhydrazyl (DPPH),
atives) and flavones (apigenin and luteolin derivatives) being butylated hydroxytoluene [BHT, 2,6-bis(1,1-dimethylethyl)-4-
already described in different varieties (Ferreres et al., 2012b). methylphenol], sodium cromoglycate, b-nicotinamide adenine
Taro crop is one of the important crops in Azores archipelago, dinucleotide reduced form (NADH), phenazine methosulfate
mainly in São Jorge, Pico and São Miguel Island, were its produc- (PMS), nitrotetrazolium blue chloride (NBT), 5,50 -dithio-bis(2-
tion covers 55 hectares and is estimated in c.a. 1000 tons per year nitrobenzoic acid) (DTNB), sodium nitroprusside dehydrate
(INE, 2012). Therefore, seeking the valorisation of taro crop in the (SNP), linoleic acid, hyaluronidase (HAase) from bovine testes
Azores archipelago, this work aimed to study the influence of (type IV-S), L-ascorbic acid and sulphanilamide were purchased
two pre-harvest factors (variety and irrigation conditions) on the from Sigma (St. Louis, MO, USA). Sodium formate, trizma hydro-
phenolic composition of C. esculenta from Azores islands, as well chloride (Tris–HCl) and bovine serum albumin (BSA) were from
as to relate the phenolic profile of the different varieties with their Sigma–Aldrich (Steinheim, Germany). Hyaluronic acid (HA) so-
biological capacity. dium salt from Streptococus equi was from Sigma–Aldrich (Pra-
gue, Czech Republic). Water was deionised using a Milli-Q
water purification system (Millipore, Bedford, MA, USA).
2. Materials and methods
2.3. Extraction
2.1. Plant material
Extraction was performed by mixing dried powdered leaves of
All C. esculenta leaves were collected in São Miguel island
each sample with water (0.2 g/50 mL), as follows: 0.5 h of sonica-
(Azores archipelago) in June 2010, at the same stage of develop-
tion followed by 1 h of stirring (200 rpm) at room temperature.
ment (Table 1): sample WD was collected in Ajuda da Bretanha re-
The extract obtained was filtered under vacuum and concentrated
gion, while the remaining ones were from Furnas region, situated
to dryness, under reduced pressure (40 °C).
53 km away from the first location. The island has a typical climate,
so all samples were subjected to the same environmental condi-
2.4. HPLC-DAD qualitative and quantitative analyses
tions: temperate rainy climate with dry summer, with tempera-
tures ranging from 17.5 °C in February (minimum values) and
Analysis was performed following the procedure of our previ-
22 °C in August (maximum values) and with a monthly average
ous work (Ferreres et al., 2012b). Redissolved aqueous extract
80% of humidity. The soil in this island is terrigenous and pumice.
(ca. 50 mg/mL in water) was analysed on an analytical HPLC-DAD
The leaves from the different varieties come from local farmers’
unit (Gilson) using a Luna C18 column (250  4.6 mm, 5 lm parti-
traditional cultures, without fertilisers, differing just in the natural
cle size; Phenomenex, Macclesfield, UK). The mobile phase con-
soil watering, because they have distinct hydric conditions: inun-
sisted of two solvents: water–acetic acid (1%) (A) and methanol
dation with cold water without drip irrigation, inundation with
(B), starting with 20% B and using a gradient to obtain 35% B at
hot water without drip irrigation or soil without irrigation
30 min, 50% B at 40 min, 90% B at 42 min and 100% B at 50 min.
(Table 1).
The flow rate was 1 mL/min. Spectral data from all peaks were col-
The plant material was immediately transferred to the labora-
lected in the range of 200–400 nm and chromatograms were re-
tory and dried in an oven at 30 °C for two weeks. The dried mate-
corded at 340 nm for flavonoids and at 320 nm for phenolic
rial was powdered (mean particle size lower than 910 lm) and
acids. The data were processed on Unipoint System software.
stored in a desiccator. The analysed samples correspond to a mix-
Phenolic compounds quantification was achieved by the absor-
ture of the leaves of three different individuals from the same vari-
bance recorded in the chromatograms relative to external calibra-
ety. Voucher specimens were deposited at the Laboratory of
tion standards. Since there is no commercially available standard,
Pharmacognosy of the Faculty of Pharmacy of Porto University, un-
di-C-glycosides, mono-C-(O-glycosyl)glycosides, di-C-(O-glyco-
der the identification CEGWC-L-062010, CERC-L-062010, CEWC-L-
syl)glycosides of luteolin and apigenin were quantified as luteo-
062010, CEWD-L-062010 and CEWH-L-062010 (leaves from ‘‘giant
lin-30 ,7-di-O-glucoside and apigenin-6-C-glucoside-7-O-glucoside,
white’’, ‘‘red’’ and the ‘‘white’’ varieties samples from three differ-
respectively. Chrysoeriol-8-C-hexoside was quantified as chrysoe-
ent culture conditions, respectively).
riol. The other compounds were quantified as themselves.

2.2. Standards and reagents 2.5. Antioxidant activity

Standards of luteolin-30 ,7-di-O-glucoside, apigenin-8-C-gluco-


2.5.1. DPPH scavenging activity
side, apigenin-6-C-glucoside, luteolin-8-C-glucoside, luteolin-6- DPPH scavenging was monitored spectrophotometrically at
C-glucoside and apigenin-6-C-glucoside-7-O-glucoside were 515 nm on a Multiskan Ascent plate reader (Thermo, Electron Cor-
purchased from Extrasynthese (Genay, France). HPLC-grade poration), following the procedure described by Ferreres et al.
methanol, acetic acid, ethanol, potassium di-hydrogen phos- (2012a). For each extract, a dilution series (five concentrations)
phate, di-sodium tetraborate, iron(II) sulphate (FeSO47H2O), was prepared: the concentration in each well ranged from 0.041
4-dimethylaminobenzaldehyde (DMAB) and N-(1-naphthyl) to 2.622 mg/mL. IC50 was calculated from three independent

Table 1
Characterisation of C. esculenta samples.

Sample Variety Soil watering

GWC Giant white Inundation of the plots with cold water without drip irrigation
RC Red Inundation of the plots with cold water without drip irrigation
WD White Dry land
WH White Inundation of the plots with hot water without drip irrigation
WC White Inundation of the plots with cold water without drip irrigation
3482 R.F. Gonçalves et al. / Food Chemistry 141 (2013) 3480–3485

assays, performed in triplicate. Ascorbic acid was used as positive variety samples presented mono-C-glycosylated, di-C-glycosylated
control. and mono-C-(O-glycosyl)-glycosylated derivatives of luteolin, api-
genin and chrysoeriol (Fig. 1 and Table 2), while in ‘‘giant white’’
2.5.2. Superoxide radical (O2) scavenging activity and ‘‘red’’ varieties (samples GWC and RC, respectively) phenolic
The effect of the extract on the superoxide radical-induced acids, flavones di-C-(O-glycosyl)glycosides and flavones O-glyco-
reduction of NBT was monitored spectrophotometrically in a Mul- sides were also found (Ferreres et al., 2012b). The three ‘‘white’’
tiskan Ascent plate reader (Thermo, electron corporation), in ki- samples exhibit distinct flavones mono-C-glycosides diversity (Ta-
netic function, at 562 nm. Superoxide radicals were generated by ble 2): samples WD and WC presented five compounds from this
the NADH/PMS system, as previously reported (Ferreres et al., class (compounds 7, 9, 10, 13 and 14), while WH only contained
2012a). For each extract, five different concentrations were tested two of them (compounds 7 and 10).
(0.031–4.917 mg/mL final concentration). IC50 was calculated from The amount of phenolic compounds in the three ‘‘white’’ varie-
three independent assays, performed in triplicate. Ascorbic acid ties ranged from ca. 3 to 7 g/kg (Table 2), being lower than that
was used as positive control. found before in the other two varieties (Ferreres et al., 2012b).
Flavones di-C-glycosides were the main compounds, representing

2.5.3. Nitric oxide ( NO) radical scavenging activity from 84% to 97% of the determined phenolics (Table 2). In addition,
The anti-radical activity was determined spectrophotometri- the relative content of flavones mono-C-glycosides in WD and WC
cally in a Multiskan Ascent plate reader (Thermo, electron corpora- samples was similar (ca. 15% and 14%, respectively) (Table 2). Apige-
tion), according to a previously described procedure (Ferreres et al., nin derivatives represent ca. 53% of the phenolic profile of WD, ca.
2012a), with final extracts concentrations between 0.061 and 50% of the phenolic profile of WH and ca. 45% of WC. As for luteolin
1.052 mg/mL. IC50 was calculated from three independent assays, derivatives, they are found in greater amounts in WH and WC sam-
performed in triplicate. Ascorbic acid was used as positive control. ples (ca. 50% and 52% respectively), being ca. 45% of the phenolic pro-
file of WD. Chrysoeriol-8-C-hexoside (14), found only in WD and WC
samples, represented ca. 3% in both samples (Table 2).
2.5.4. Lipid peroxidation inhibition assay
The three major phenolic compounds are, in decreasing order,
Peroxidation of fatty acyl groups was determined following the
luteolin-6-C-pentoside-8-C-hexoside (6), apigenin-6-C-pentoside-
methodology proposed by Ferreres et al. (2012c), with slight mod-
8-C-hexoside (5) and apigenin-6-C-hexoside-8-C-hexoside (8).
ifications. The reaction mixture contained 250 lL of linoleic acid
These compounds represent more than half of the phenolics deter-
4 mM, 150 lL of Tris–HCl (100 mM, pH 7.5), 50 lL of FeSO47H2O
mined (ca. 80% in WH, ca. 70% in WD and ca. 65% in WC).
4 mM and 50 lL of serial dilutions of the five different extracts pre-
The differences observed among the three samples from the
pared in distiled H2O (final concentration: 0.036–2.273 mg/mL).
‘‘white’’ variety could be attributed to the distinct growth condi-
Linoleic acid peroxidation was initiated by the addition of 50 lL
tions. WD is the sample that contains the highest amounts of phe-
of ascorbic acid 5 mM, and the mixture was immediately incubated
nolic compounds. This sample is grown without irrigation, so the
for 1 h at 37 °C. After the incubation period, 3 mL of ethanol was
higher production of such metabolites, compared with the samples
added to each test tube. The mixtures were vortexed and the
WH and WC, would be expected, since a reduction in water supply
absorbance was immediately measured at 233 nm in a Helios a
usually increases the content of phytochemicals like phenolic com-
(Unicam) spectrophotometer, at room temperature. Three inde-
pounds (Sánchez-Rodríguez, Moreno, Ferreres, Rubio-Wilhelmi, &
pendent assays were performed in triplicate. BHT was used as po-
Ruiz, 2011).
sitive control.
Regarding sample WH, data suggest that the hot water treat-
ment not only resulted in lower biosynthesis of phenolic com-
2.6. Hyaluronidase (HAase) inhibition assay pounds, being also noticed lower diversity of compounds (Table 2).
Data from literature obtained with different C. esculenta varie-
HAase inhibition assay was performed as proposed by Ferreres ties reveal distinct phenolic composition (Ferreres et al., 2012b).
et al. (2012c). The reaction mixture contained 50 lL of HA stock Furthermore, sample WC studied herein shares the same geo-
solution, 100 lL of buffer (0.2 M sodium formate, 0.1 M NaCl and graphical origin and similar watering conditions of the ‘‘giant
0.2 mg/mL BSA, pH 3.68), 200 lL of water and 50 lL of serial dilu- white’’ and ‘‘red’’ varieties samples previously analysed by our
tions of the five different extracts prepared in distiled H2O (0.021– group (Ferreres et al., 2012b), but exhibited a distinct phenolic pro-
1.052 mg/mL on well). The reaction was started by the addition of file. Taken together, the results obtained in this study reinforce the
50 lL of HAase (600 U/mL) prepared in NaCl 0.9%. possibility of the occurrence of infraspecific chemical variability in
The enzymatic reaction was stopped by adding 25 lL of di-so- C. esculenta pointed before (Ferreres et al., 2012b).
dium tetraborate (0.8 M in water), and subsequent heating for
3 min in a boiling water bath. The test tubes were cooled at room 3.2. Biological activity
temperature and 750 lL of DMAB solution was added. The tubes
were incubated at 37 °C for 20 min. The absorbance of the coloured Wound healing consists of an orderly progression of events in
product was measured at 560 nm in a Multiskan Ascent plate read- order to re-establish the integrity of a damaged tissue. The healing
er. Three independent assays were performed in triplicate. Sodium process involves three phases that overlap in time and space:
cromoglycate was used as positive control. inflammation, tissue formation and tissue remodelling (Gupta
et al., 2006). During inflammation highly reactive species, such as
3. Results and discussion superoxide radical (O2), peroxynitrite (ONOO), hydrogen perox-

ide (H2O2), hypochlorous acid (HOCl), peroxyl radicals (ROO ) and
3.1. Phenolic composition nitric oxide radical (NO) are produced. Therefore, the wound site
is an oxidative/nitrosative environment, rich in reactive oxygen
The phenolic composition of the leaves from ‘‘giant white’’ and and nitrogen species (ROS and RNS, respectively), protecting the
‘‘red’’ varieties (samples GWC and RC, respectively) was described wound against pathogens growth (Khanna et al., 2002). However,
before by our group (Ferreres et al., 2012b). In the present work, when at high concentrations, ROS/RNS cause severe tissue damage,
three samples from another variety (samples WC, WH and WD) leading to neoplastic transformations that undermine the wound
were analysed, revealing a distinct qualitative profile: ‘‘white’’ healing process (Lundvig, Immenschuh, & Wagener, 2012).
 R.F. Gonçalves et al. / Food Chemistry 141 (2013) 3480–3485 3483

Fig. 1. HPLC-DAD (340 nm) phenolic profile of the aqueous extract from C. esculenta ‘‘white’’ variety (sample WD). (1) Luteolin-6-C-hexoside-8-C-pentoside; (2) apigenin-6,8-
di-C-hexoside; (3) luteolin-6-C-hexoside-8-C-pentoside; (4) luteolin-6-C-pentoside-8-C-hexoside; (5) apigenin-6-C-pentoside-8-C-hexoside; (6) luteolin-6-C-pentoside-8-C-
hexoside; (7) luteolin-8-C-glucoside (orientin); (8) apigenin-6-C-hexoside-8-C-pentoside; (9) luteolin-6-C-glucoside (isoorientin); (10) apigenin-8-C-glucoside (vitexin); (11)
apigenin-8-C-(2-O-pentosyl)hexoside; (12) apigenin-6-C-hexoside-8-C-pentoside; (13) apigenin-6-C-glucoside (isovitexin); (14) chrysoeriol-8-C-hexoside.

Table 2
Phenolic composition of C. esculenta samples (mg/kg dry basis).a

Compoundb Sample
WD WH WC
Flavones mono-C-glycosides
7 Lut 8-C-Gluc (orientin) 89.04 ± 4.79 46.31 ± 3.41 45.98 ± 1.04
9 Lut 6-C-Gluc (isoorientin) 76.61 ± 2.21 n.d. 103.66 ± 0.90
10 Api 8-C-Gluc (vitexin) 584.48 ± 15.31 20.89 ± 2.82 273.93 ± 12.70
13 Api 6-C-Gluc (isovitexin) 31.43 ± 9.27 n.d. 13.88 ± 1.02
14 Chrys 8-C-Hex 168.58 ± 11.09 n.d. 105.19 ± 11.68
Flavones di-C-glycosides
1 Lut 6-C-Hex-8-C-Pent 287.27 ± 25.24 83.58 ± 1.39 311.39 ± 29.07
2 Api 6,8-di-C-Hex n.q. n.q. n.q.
3 Lut 6-C-Hex-8-C-Pent 197.43 ± 4.19 70.41 ± 8.43 199.18 ± 11.03
4 Lut 6-C-Pent-8-C-Hex 49.95 ± 4.26 24.86 ± 0.14 69.15 ± 5.69
5 Api 6-C-Pent-8-C-Hex 1433.39 ± 22.64 666.66 ± 46.31 769.11 ± 41.73
6 Lut 6-C-Pent-8-C-Hex 2384.15 ± 207.01 1251.22 ± 71.39 1415.39 ± 81.65
8 Api 6-C-Hex-8-C-Pent 801.01 ± 36.14 420.99 ± 28.24 521.98 ± 9.75
12 Api 6-C-Hex-8-C-Pent 487.69 ± 5.41 257.48 ± 17.21 303.58 ± 20.12
Flavones mono-C-(O-glycosyl)glycosides
11 Api 8-C-(2-O-pent)Hex 84.69 ± 6.32 29.72 ± 1.44 16.90 ± 0.75
Total phenolic content (mg/kg) 6675.72 2938.45 4149.36
Flavones mono-C-glycosides (%) 14.63 ± 0.57 2.24 ± 0.18 13.84 ± 0.44
Flavones di-C-glycosides (%) 84.07 ± 0.57 96.74 ± 0.19 85.87 ± 0.51
Flavones mono-C-(O-glycosyl)glycosides (%) 1.30 ± 0.03 1.02 ± 0.01 0.44 ± 0.03
Luteolin derivatives (%) 44.55 ± 1.57 50.07 ± 0.38 52.27 ± 3.63
Apigenin derivatives (%) 52.71 ± 1.41 49.92 ± 0.38 44.80 ± 3.78
Chrysoeriol derivatives (%) 2.63 ± 0.18 n.d. 2.93 ± 0.20
a
Identity of samples as in Table 1. Results are expressed as mean ± standard deviation of three determinations; n.d.: not detected; n.q.: not quantified.
b
Lut: luteolin; Api: apigenin; Chrys: chrysoeriol; Diosmt: diosmetin; Gluc: glucoside; Hex: hexoside; Pent: pentoside; Rhmn: rhamnoside; Der: derivatives.

In order to evaluate the influence of the phenolic composition contents, the last being the richest. However, GWC contains a high-
on C. esculenta wound healing potential we analysed the ‘‘giant er amount of flavones mono-C-glycosylated, namely luteolin deriv-
white’’ (sample GWC) and ‘‘red’’ (sample RC) varieties which phe- atives [luteolin-8-C-glucoside (orientin) and luteolin-6-C-glucoside
nolic composition was already described by our group (Ferreres (isoorientin)] (Table 2). These two compounds have great DPPH
et al., 2012b) and the three different samples from the ‘‘white’’ scavenging ability, since they are dihydroxylated at positions 30
variety for which phenolic profile is provided in this study (sam- and 40 of the B ring, forming a highly reactive catechol group
ples WC, WH and WD). The antiradical capacity of the samples (Zhang, 2005). Their corresponding apigenin derivatives, also
was assessed against DPPH, O2, NO and lipid peroxidation identified in these varieties [apigenin-8-C-glucoside (vitexin) and

(LOO ) (Table 3). Ascorbic acid was used as positive control of apinenin-6-C-glucoside (isovitexin)] (Table 2), only have one hy-
the first three assays and BHT in the last. droxyl substituent at 40 , thus having lower activity (Zielińska &
The ‘‘giant white’’ (GWC) variety demonstrated the highest Zieliński, 2011).
DPPH scavenging capacity. The ‘‘white’’ varieties WC and WD Regarding flavones di-C-glycosides, the qualitative analysis has
have a similar activity, while WH displayed the lowest activity shown that these compounds represent the major constituents of
(Table 3). GWC and RC samples contained the highest phenolic the ‘‘white’’ variety phenolic composition. The C-glycosylation, in
3484 R.F. Gonçalves et al. / Food Chemistry 141 (2013) 3480–3485

Table 3
IC50 (mg/mL) found in the antioxidant and hyaluronidase inhibitory potential assays with the extracts from C. esculenta samples and with reference compounds.a

Samples IC50 (mg/mL) LOO HAase

 
DPPH NO O2
GWC 0.226 ± 0.006 1.247 ± 0.098 0.123 ± 0.002 0.491 ± 0.033 5.219 ± 0.017
RC 0.535 ± 0.032 0.521 ± 0.018 0.145 ± 0.005 0.570 ± 0.048 4.658 ± 0.005
WD 0.746 ± 0.040 1.181 ± 0.107 0.123 ± 0.006 0.788 ± 0.080 2.642 ± 0.106
WH 1.140 ± 0.051b 1.145 ± 0.038 0.106 ± 0.011 0.650 ± 0.087 0.333 ± 0.016
WC 0.695 ± 0.054 1.485 ± 0.102 0.076 ± 0.004 0.958 ± 0.043 4.002 ± 0.078
Ascorbic acid 7.100 ± 0.197c 0.250 ± 0.015 0.425 ± 0.053 – –
BHT – – – 4.000 ± 0.171c –
Sodium cromoglycate – – – – 1.890 ± 0.038
a
Identity of samples as in Table 1. Values represent mean ± SD of three assays.
b
Value corresponds to IC25.
c
Value expressed in lg/mL.

particular at position 8 of the A ring, decreases the activity of the composition (Tables 2 and 3, Fig. 1). These results lead us to sug-
phenolic compounds, as it induces a molecule distortion, affecting gest again that other compounds besides phenolics are involved
the co-planarity and making it difficult to stabilise the phenoxyl in the lipid peroxidation scavenging potential of C. esculenta.
radical (Zielińska & Zieliński, 2011). All flavones di-C-glycosides Although oxidative stress is a key factor in the inflammatory
identified in this work are substituted at C-8. process, hyaluronic acid (HA) also plays a significant role. This non-
In this way, it is possible to conjecture that the highest activity sulfated, linear glycosaminoglycan (GAG), is present in most living
of the ‘‘giant white’’ variety may be due to the higher concentration tissues as a high molecular mass polymer (>106 Da) and in signif-
of flavones mono-C-glycosides derived from luteolin. The absence icant quantities in the skin (dermis and epidermis), brain, and cen-
or low amount of these compounds in ‘‘white’’ taro can also explain tral nervous system. HA plays a crucial role in tissue repair,
the lower activity of WD, WC and WH samples, with particular rel- including wound healing, as it is involved in dynamic cellular pro-
evance to WH. Nevertheless, comparing with a known antioxidant cesses, such as cell migration and cell–cell recognition, during

compound, ascorbic acid, the DPPH scavenging activity of all taro wound healing and inflammation (David-Raoudi et al., 2008).
samples is lower. For normal tissue organisation and function, a balance between

The different varieties exhibit considerable O2  scavenging HA synthesis and degradation is important. High molecular weight
activity, being even more effective than ascorbic acid (Table 3). tissue HA is degraded extracellularlly by HAase. Non-enzymati-

However, unlike the results obtained in the DPPH assay, it is not cally, HA is degraded by reactive oxygen species (ROS). Hyaluro-
possible to speculate a relationship with the phenolic profile of nidases are endoglucosaminidases, whereas ROS degrade HA
the samples. Even though this activity may probably be due to randomly at internal glycosidic linkages (Girish & Kemparaju,
other compounds present in the analysed extracts, phenolic com- 2007). In an unbalanced healing process, where the inflammatory
pounds may have a synergistic effect, probably recycling radicals condition is increased by the accumulation of HA fragments, hyal-
of other antioxidant metabolites (Kang, 2007). uronidase inhibitors can be important to prevent the accumulation
The ‘‘red’’ variety showed a considerably higher antioxidant of these low molecular weight hyaluronic acid (LMWHA) frag-
capacity against nitric oxide radical than the other two varieties, ments and, therefore, avoid the prolonged inflammatory condition
which have a similar capacity (Table 3). Again, it is not possible (Mio & Stern, 2002).
to relate the content of phenolic compounds to the activity of the The WH sample exhibited the strongest potential to inhibit
different varieties. For example, WH is the sample with the second hyaluronidase (Table 3), followed by the other two samples from
strongest activity against this radical, though its phenolic profile is this variety (WD and WC). GWC and RC displayed the lower capac-
qualitatively the poorest and quantitatively the lowest. Hence, ity. Previous studies speculated that the inhibition of hyaluroni-
other unidentified compounds may probably be involved in the dase by flavonoids is higher with the increasing of hydroxyl
antioxidant activity of the different analysed varieties. As for groups in the flavonoid skeleton; furthermore glycosylation de-
DPPH, NO scavenging ability of these samples is significantly creases the activity (Hertel, Peschel, Ozegowski, & Müller, 2006).
lower than that of ascorbic acid. This structure–activity relationship is the same as for ROS/RNS
Oxygen radicals cause tissue damage by lipid peroxidation of scavenging. In this way, the samples with higher activity should
structural lipids and proteins from the skin. These agents not only be GWC and RC as referred above for DPPH scavenging activity.
damage the lipids, but also produce lipid hydroperoxides, second- This leads us to infer that the phenolic compounds present in the
ary intermediates that can lead to a lipid peroxidation chain reac- studied samples are not the main responsible for the HAase inhibi-
tion (Altavilla et al., 2001). Lipid peroxidation is considered tion, as the samples from the white variety showed the qualita-
responsible for the impairment of endothelial cells, keratinocyte tively poorest and lowest phenolics content (Table 2). Finally, it
capillary permeability, and fibroblast and collagen metabolism, is noteworthy that WH sample has higher potential to inhibit hyal-
thus affecting the skin homeostasis and leading to several skin uronidase than a commercialised HAase inhibitor, sodium cromo-
pathologies (Altavilla et al., 2001; Briganti & Picardo, 2003). glycate (Yingprasertchai, Bunyasrisawat, & Ratanabanangkoon,
The analysed C. esculenta extracts presented a moderate lipid 2003).
peroxide scavenging potential, that from GWC being the one with
greater scavenging potential (IC50 = 0.491 mg/mL) and WC the
variety with lower activity (IC50 = 0.958 mg/mL) (Table 3). The lipid 4. Conclusions
peroxide scavenging potential of WH is similar to that of RC (Ta-
ble 3), although their phenolic composition is quite different (Ta- In this study, we demonstrated the impact of the irrigation con-
ble 2 and Ferreres et al., 2012b). On the other hand, the WD and ditions in the phenolic composition of C. esculenta leaves. Dry land
WC samples from the ‘‘white’’ variety present a quite different lipid culture leads to the accumulation of higher contents of phenolic
peroxidation preventing capacity in spite of their similar phenolic compounds. Additionally, the speculation of a qualitative and
 R.F. Gonçalves et al. / Food Chemistry 141 (2013) 3480–3485 3485

quantitative infraspecific chemical variability in C. esculenta can be Ferreres, F., Gil-Izquierdo, A., Vinholes, J., Silva, S. T., Valentão, P., & Andrade, P. B.
(2012a). Bauhinia forficata link authenticity using flavonoids profile: Relation
reinforced by the results obtained in this study, ‘‘giant white’’ and
with their biological properties. Food Chemistry, 134, 894–904.
‘‘red’’ varieties providing higher diversity and quantity of phenolic Ferreres, F., Gonçalves, R. F., Gil-Izquierdo, A., Valentão, P., Silva, A. M., Silva, J. B.,
compounds. et al. (2012b). Further knowledge on the phenolic profile of Colocasia esculenta
The wound healing property attributed to C. esculenta can be re- (L.) Shott. Journal of Agricultural and Food Chemistry, 60, 7005–7015.
Ferreres, F., Lopes, G., Gil-Izquierdo, A., Andrade, P. B., Sousa, C., Mouga, T., et al.
lated to its antioxidant activity, namely to its superoxide radical (2012c). Phlorotannin extracts from Fucales characterised by HPLC-DAD-ESI-
scavenging potential, and to the inhibition of hyaluronidase, thus MSn: Approaches to hyaluronidase inhibitory capacity and antioxidant
protecting skin cells from oxidative damage and accelerating the properties. Marine Drugs, 10, 2766–2781.
Girish, K. S., & Kemparaju, K. (2007). The magic glue hyaluronan and its eraser
recovery of the wound in the inflammatory state. Phenolic com- hyaluronidase: A biological overview. Life Sciences, 80, 1921–1943.
pounds may contribute to these properties. Gupta, A., Kumar, R., Pal, K., Singh, V., Banerjee, P. K., & Sawhney, R. C. (2006). Herbal
The present study increased the knowledge about the variation medicines for wound healing among tribal people in Southern India:
Ethnobotanical and scientific evidences. Molecular and Cellular Biochemistry,
in the composition of C. esculenta leaves caused by abiotic and ge- 290, 193–198.
netic factors, providing useful information about the best cultiva- Hertel, W., Peschel, G., Ozegowski, J.-H., & Müller, P.-J. (2006). Inhibitory effects of
tion procedure to obtain health-promoting antioxidants and triterpenes and flavonoids on the enzymatic activity of hyaluronic acid-splitting
enzymes. Archiv der Pharmazie – Chemistry in Life Sciences, 339, 313–318.
hyaluronidase inhibitors. INE (Instituto Nacional de Estatística). Estatísticas Agrícolas 2011. URL http://
www.ine.pt/xportal/xmain?xpid=INE&xpgid=ine_publicacoes&PUBLICACOES
Acknowledgements pub_boui=142185148&PUBLICACOESmodo=2, accessed 25.02.2013.
Kang, E. M. S. (2007). Dietary flavonoids as protectors from ascorbate-induced
oxidative stress in vivo. PhD Thesis, College of Pharmacy and Nutrition
The authors thank Fundação para a Ciência e a Tecnologia (FCT) University of Saskatchewan Saskatoon, Saskatoon, Canada.
for Grant No. PEst-C/EQB/LA0006/2011, to ‘‘Programa PROCON- Khanna, S., Venojarvi, M., Roy, S., Sharma, N., Trikha, P., Bagchi, D., et al. (2002).
VERGÊNCIA’’ from Secretaria Regional da Economia da Região Dermal wound healing properties of redox-active grape seed
proanthocyanidins. Free Radical Biology & Medicine, 33, 1089–1096.
Autónoma dos Açores, to ‘‘TERMAZ’’ Project promoted by INOVA Lundvig, D. M. S., Immenschuh, S., & Wagener, F. A. D. T. G. (2012). Heme oxygenase,
(Instituto Inovação tecnológica dos Açores), to CYTED Programme inflammation, and fibrosis: The good, the bad, and the ugly? Frontiers in
(Ref. 112RT0460) CORNUCOPIA Thematic Network and project Pharmacology, 3. http://dx.doi.org/10.3389/fphar.2012.00081.
Mio, K., & Stern, R. (2002). Inhibitors of the hyaluronidases. Matrix Biology, 21,
AGL2011-23690 (CICYT). 31–37.
Oscarsson, K. V., & Savage, G. P. (2007). Composition and availability of soluble and
References insoluble oxalates in raw and cooked taro (Colocasia esculenta var. Schott)
leaves. Food Chemistry, 101, 559–562.
Padula, M. C., Lepore, L., Milella, L., Ovesna, J., Malafronte, N., Martelli, G., et al.
Abad-García, B., Berrueta, L. A., Garmón-Lobato, S., Gallo, B., & Vicente, F. A. (2009).
(2012). Cultivar based selection and genetic analysis of strawberry fruits with
General analytical strategy for the characterisation of phenolic compounds in
high levels of health promoting compounds. Food Chemistry. http://dx.doi.org/
fruit juices by high-performance liquid chromatography with diode array
10.1016/j.foodchem.2012.11.025. in press.
detection coupled to electrospray ionisation and triple quadrupole mass
Prajapati, R., Kalariya, M., Umbarkar, R., Parmar, S., & Sheth, N. (2011). Colocasia
spectrometry. Journal of Chromatography A, 1216, 5398–5415.
esculenta: A potent indigenous plant. International Journal of Nutrition,
Agyarea, C., Asaseb, A., Lechtenbergc, M., Niehuesc, M., Detersc, A., & Henselc, A.
Pharmacology, Neurological Diseases, 1, 90–96.
(2009). An ethnopharmacological survey and in vitro confirmation of
Rouphael, Y., Cardarelli, M., Bassal, A., Leonardi, C., Giuffrida, F., & Colla, G. (2012).
ethnopharmacological use of medicinal plants used for wound healing in
Vegetable quality as affected by genetic, agronomic and environmental factors.
Bosomtwi–Atwima–Kwanwoma area, Ghana. Journal of Ethnopharmacology,
Journal of Food, Agriculture & Environment, 10, 680–688.
125, 393–403.
Sánchez-Rodríguez, E., Moreno, D. A., Ferreres, F., Rubio-Wilhelmi, M. D. M., & Ruiz,
Altavilla, D., Saitta, A., Cucinotta, D., Galeano, M., Deodato, B., Colonna, M., et al.
J. M. (2011). Differential responses of five cherry tomato varieties to water
(2001). Inhibition of lipid peroxidation restores impaired vascular endothelial
stress: Changes on phenolic metabolites and related enzymes. Phytochemistry,
growth factor expression and stimulates wound healing and angiogenesis in the
72, 723–729.
genetically diabetic mouse. Diabetes, 50, 667–674.
Yingprasertchai, S., Bunyasrisawat, S., & Ratanabanangkoon, K. (2003).
Briganti, S., & Picardo, M. (2003). Antioxidant activity, lipid peroxidation and skin
Hyaluronidase inhibitors (sodium cromoglycate and sodium auro-thiomalate)
diseases. What’s new. Journal of the European Academy of Dermatology and
reduce the local tissue damage and prolong the survival time of mice injected
Venereology, 17, 663–669.
with Naja kaouthia and Calloselasma rhodostoma venoms. Toxicon, 42, 635–646.
David-Raoudi, M., Tranchepain, F., Deschrevel, B., Vincent, J.-C., Bogdanowicz, P.,
Zhang, H.-Y. (2005). Structure–activity relationships and rational design strategies
Boumediene, K., et al. (2008). Differential effects of hyaluronan and its
for radical-scavenging antioxidants. Current Computer-Aided Drug Design, 1,
fragments on fibroblasts: Relation to wound healing. Wound Repair and
257–273.
Regeneration, 16, 274–287.
Zielińska, D., & Zieliński, H. (2011). Antioxidant activity of flavone C-glucosides
Ejoh, A. R., Mbiapo, F. T., & Fokou, E. (1996). Nutrient composition of the leaves and
determined by updated analytical strategies. Food Chemistry, 124, 672–678.
flowers of Colocasia esculenta and the fruits of Solanum melongena. Plant Foods
for Human Nutrition, 49, 107–112.