Food Chemistry 125 (2011) 193–200

Contents lists available at ScienceDirect

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

Antioxidant constituents from Lawsonia inermis leaves: Isolation, structure

elucidation and antioxidative capacity
Anis Ben Hsouna a, Mohamed Trigui a, Gérald Culioli b, Yves Blache b, Samir Jaoua a,c,⇑
a
Laboratory of Biopesticides, Centre of Biotechnology of Sfax, P.O. Box: ‘‘1177”, 3038 Sfax, Tunisia
b
Laboratoire MAPIEM (EA 4323), Equipe «Biofouling and Substances Naturelles Marines», Université du Sud Toulon-Var, Avenue de l’Université, BP 20132,
F-83957 La Garde Cedex, France
c
University of Qatar, College of Arts and Sciences, P.O. Box 2713, Doha, Qatar

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

Article history: The antioxidant potential of different fractions of Lawsonia inermis (Lythraceae) was investigated. The n-
Received 18 March 2010 butanolic fraction showed the highest yield of extraction; it also exhibited a strong antioxidant activity in
Received in revised form 23 August 2010 the DPPH assay and a potent capacity in preventing linoleic acid oxidation. Five phenolic glycosides were
Accepted 25 August 2010
identified in this fraction. The structure of a new compound was established as 1,2,4-trihydroxynaphtha-
lene-1-J-b-D-glucopyranoside. In addition, the known 2,4,6-trihydroxyacetophenone-2-J-b-D-glucopy-
ranoside was described for the first time in this species. The three other compounds, lalioside (2,3,4,
Keywords:
6-tetrahydroxyacetophenone-2-J-b-D-glucopyranoside), lawsoniaside (1,2,4-trihydroxynaphthalene-
Lawsonia inermis
Naphthalene glycosides
1,4-di-J-b-D-glucopyranoside) and luteolin-7-J-b-D-glucopyranoside, have been previously reported in
Acetophenone glycosides L. inermis. The antioxidant activity of these glycosides was evaluated by DPPH and b-carotene assays,
1,2,4-Trihydroxynaphthalene-1-O-b-D- and compared to those of commercial standards. 1,2,4-Trihydroxynaphthalene-1-J-b-D-glucopyranoside
glucopyranoside was the most active in the DPPH free-radical scavenging test (EC50 = 6.5 lg/ml) and showed a moderate
Antioxidant activities inhibition in the b-carotene bleaching assay. Chemical components of L. inermis have good antioxidant
capacities and this species could be used as a potential source of new natural antioxidants.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction or broadly lanceolate, entire, acute or obtuse, 2–3 cm long and 1–

2 cm wide (Muhammad & Muhammad, 2005). This plant is fre-
Antioxidants play an important role in preventing undesirable quently cultivated in India, Middle East and along the African
changes in flavour and nutritional quality of foods and may reduce coasts of the Mediterranean Sea. Besides its use in cosmetics for
the risk of a wide spectrum of chronic diseases, such as carcinogen- staining hands and as a hair dye, the leaves are used as a prophy-
esis, diabetes, or inflammation (Vasundhara, Vijay, & Jagan, 2008). lactic agent against skin diseases (Ahmed et al., 2000).
Synthetic antioxidants are widely used as food additives, but their Phytochemical investigations of L. inermis have shown predom-
application has been reassessed because of the formation of possi- inantly the presence of phenolic compounds (coumarins, flavo-
ble toxic or carcinogenic components during their degradation noids, naphthalene and gallic acid derivatives) which could be
(Barlow, 1990; Linderschmidt, Trylka, Good, & Witschi, 1986). glycosylated (Siddiqui, Kardar, Ali, & Khan, 2003). Other com-
Hence, identification and isolation of new antioxidants from natu- pounds, such as triterpenoids, steroids and aliphatic hydrocarbons
ral sources has become an active area of research. A number of nat- have been also isolated from this plant (Siddiqui & Kardar, 2001;
ural products, such as flavonoids, coumarins, curcuminoids or Siddiqui et al., 2003). However, no correlation has been proved be-
terpenes, isolated from plants have shown potent antioxidant tween the biochemical composition of this plant and its therapeu-
activity and low toxicity (Yang et al., 2007). tic uses. The screening of 20 plant species used by Yemeni
Tunisian flora is known for its diversity of medicinal plants such traditional healers to treat infectious diseases showed that the
as Lawsonia inermis (syn. Lawsonia alba), commonly known as ‘hen- ethyl acetate extract of L. inermis was found to be the most active
na’. This plant is a glabrous, much branched shrub or small tree against both Gram+ and Gram bacteria (Ali, Juelich, Kusnick, &
with greyish-brown bark. Leaves are opposite, sub-sessile, elliptic Lindequist, 2001). In addition, an ethanolic extract of the whole
plant possessed antifungal (Misra & Dixit, 1979) and antitubercular
⇑ Corresponding author. Address: Centre of Biotechnology of Sfax, Laboratory of activities (Bhatnagar et al., 1961). The decoction of bark and leaves
Biopesticides, P. O. Box ‘‘1177”, Sfax 3038, Tunisia. Tel./fax: +216 74874446. has been found to inhibit trypsin enzymes (Prasad & Gupta, 1967)
E-mail address: [email protected] (S. Jaoua). and showed anti-inflammatory activity (Singh, Shrivastava, Modi,

0308-8146/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodchem.2010.08.060
194 A.B. Hsouna et al. / Food Chemistry 125 (2011) 193–200

& Saifi, 1982). The bark was also known to alleviate serum trans- 63 lm, 37 g; Merck, Darmstadt, Germany) used for the fraction-
aminase levels and to restore normal bile flow. ation. The separation process was carried out using a ternary sol-
The aim of the present study was to evaluate the antioxidant vent system: water (solvent A), methanol (solvent B) and
and radical-scavenging activities of n-hexane, chloroform, ethyl dichloromethane (solvent C). After an initial isocratic step with
acetate, n-butanol and water fractions obtained from the hydroal- 90% of A and 10% of B from 0 to 2 min, a linear gradient up to
coholic extract of L. inermis leaves. The amounts of total phenolics 100% of B from 2 to 15 min, an isocratic step with 100% of B from
in these fractions were also determined. As the n-butanolic fraction 15 to 17 min, and a linear gradient up to 100% of C from 17 to
showed the strongest antioxidant activity, it was subjected to fur- 20 min, an isocratic step with 100% of C was finally used from 20
ther fractionation by flash chromatography (FC). The chemical to 25 min. Concerning the flow rate, an initial linear ramp was used
composition of the most active fractions was analysed by LC/ from 2 to 20 ml/min in 2 min and then this final flow rate was
DAD-ELSD-ESI/MSn and five phenolic glycosides were isolated by maintained during all the fractionation process. The separation
semi-preparative reversed-phase HPLC. NMR and MS analyses of was monitored by UV detection at 280 nm. All sub-fractions were
these compounds allowed the unambiguous characterisation of automatically collected, evaporated and then weighed.
their chemical structure. The purified molecules were the subject
of antioxidant activity evaluation by measuring their capacity to 2.3.3. LC-DAD-ELSD-ESI/MSn analysis
scavenge DPPH and to inhibit the bleaching of b-carotene. In order to check the purity of the constituents and to deter-
mine their molecular mass, all sub-fractions obtained after the FC
2. Materials and methods fractionation of the n-butanolic extract were analysed by LC-
DAD-ELSD-ESI/MSn. Their chromatographic profiles were com-
2.1. Chemicals pared with those of the crude n-butanolic fraction. For this exper-
iment, a LaChrom Elite HPLC (VWR-Hitachi, Fontenay-sous-Bois,
2,2-Diphenyl-1-picrylhydrazyl (DPPH), butylated hydroxytolu- France) was used, which comprised an L-2130 quaternary pump,
ene (BHT), b-carotene, linoleic acid, ascorbic acid, Folin–Ciocalteu an L-2200 autosampler and an L-2300 column oven. Detection
reagent, Tween 20 and sodium carbonate were purchased from was performed with an L2455 PDA detector and an evaporative
Sigma–Aldrich Co. (St. Louis, MO). All solutions were freshly pre- light scattering detector (ELSD; Chromachem model; Eurosep, Cer-
pared in distiled water. The solvents used for chromatographic gy-Saint Christophe, France) coupled to an ion trap mass spectrom-
studies were all of HPLC grade and were from Fisher Scientific eter fitted with an electrospray ionisation (ESI) interface (Esquire
(Illkirch, France) while extraction and partition were realised with 6000; Bruker Daltonics, Bremen, Germany). Operating conditions
solvents from Sigma–Aldrich. for MS analysis were: dry temperature, 350 °C; capillary voltage,
4000 V; nebuliser, 50 psi; dry gas, helium at 12 l/min. Ion trap
full-scan analysis was conducted from m/z 50 to 1200 with an
2.2. Plant material
upper fill time of 200 ms. One milligram of each fraction was dis-
solved in 1 ml of a mixture of methanol/water (1:1, v/v) and a
L. inermis leaves were collected from Sfax (Tunisia, 35.23°N and
10-ll sample volume was injected into the chromatographic
11.11°E) in May 2008. After the botanical identification of the spe-
system.
cies, a voucher specimen (LBPes 03) was deposited in the Herbar-
Compounds separation was achieved using an analytical re-
ium of the Laboratory of Biopesticides (Centre of Biotechnology
versed-phase column (Gemini C6-Phenyl, 250  3 mm, 5 lm, Phe-
of Sfax, Tunisia) for future reference.
nomenex) maintained at 30 °C. The gradient elution was optimised
and consisted of a binary mobile phase in which eluent A was 1%
2.3. Phytochemical study of L. inermis (v/v) formic acid in water and eluent B was 1% (v/v) formic acid
in acetonitrile. The elution was started with 30% of B and ascended
2.3.1. Preparation of plant extracts to 50% of B in 20 min (linear ramp) and 100% B in 5 min (linear
The L. inermis fresh leaves were dried and fine powdered. The ob- ramp). This composition was maintained for 5 min until re-equili-
tained powder (100 g) was extracted by maceration into 600 ml of bration of the system to the initial conditions (5 min). The flow rate
aqueous ethanol (ethanol/water, 4:1, v/v) with occasional shaking, was fixed at 0.5 ml/min.
at room temperature. After 3 days, the extract was filtered through
a 45-lm filter and was concentrated under vacuum. The resulted 2.3.4. Purification of phenolic glycosides by semi-preparative HPLC
residue (22 g) was suspended in 100 ml of distiled water and The main chemical components of selected sub-fractions ob-
sequentially partitioned into n-hexane (3  250 ml), chloroform tained from the n-butanolic fraction of henna were separated by
(3  250 ml), ethyl acetate (3  250 ml) and then n-butanol semi-preparative reversed-phase HPLC. This purification was rea-
(3  250 ml). The resulting five solutions were concentrated in vacuo lised on a Biotek-Kontron (Montigny-le-Bretonneux, France) HPLC
to dryness, to give n-hexane (1.92 g), chloroform (0.28 g), ethyl ace- system equipped with a ternary pump (Model 525), an auto sam-
tate (1.97 g), n-butanol (6.41 g) and water (8.84 g) soluble fractions. pler (Model 560) and a column oven (Model 582). The detection
These fractions were kept at 4 °C in the dark until further analysis. was made with a UV detector (LDC Analytical, Model 3100) at
280 nm. The separation of the main constituents of these fractions
2.3.2. Fractionation of the n-butanolic fraction using flash was achieved on a semi-preparative reversed-phase column (Gem-
chromatography ini C6-Phenyl, 250  10 mm, 5 lm, Phenomenex) with the same
Separation of phenolic constituents from the n-butanolic frac- gradient of elution previously used for the LC–MS analysis.
tion of L. inermis was carried out by FC on a Spot Flash (Armen
Instruments; Saint Ave, France) chromatographic system. For frac- 2.3.5. Identification of phenolic glycosides
tionation, 1.5 g of this extract were dissolved in 5 ml of methanol The structures of the phenolic glycosides isolated in this study
and mixed with 20 g of reversed-phase silica gel (Sepra C18e, were determined through extensive spectrometric analysis (ESI/
50 lm; Phenomenex, Torrance, CA). Methanol was removed by MS, 1D and 2D-NMR) and by comparison with data reported in
evaporation under reduced pressure at 35 °C. The obtained dry ex- the literature. High resolution mass spectra (HR-ESI/MS) were ob-
tract of henna was then loaded on a guard column on the top of the tained on an LTQ Orbitrap mass spectrometer (Thermo Finnigan,
reversed-phase cartridge (SuperVarioFlash D26-RP18 model, 40– San José, CA). 1D (1H, 13C and DEPT) and 2D (COSY 1H–1H, HSQC,
 A.B. Hsouna et al. / Food Chemistry 125 (2011) 193–200 195

HMBC and NOESY 1H–1H) NMR spectra were obtained at 400 and of the sample and the blank, respectively, measured after 2 h. All
100 MHz for 1H and 13C, respectively, on an Avance 400 MHz tests were carried out in triplicate.
NMR spectrometer (Bruker BioSpin, Rheinstetten, Germany) in
DMSO-d6, D2O or CD3OD. All chemical shifts were referenced to 2.6. Statistical analysis
the solvent peaks (dH 2.50 and dC 39.5 for DMSO-d6, dH 3.31 and
dC 49.0 for CD3OD and dH 4.79 for D2O). Experimental results from this study were expressed as
means ± standard deviation of three parallel measurements. The
2.4. Determination of the total phenolic content significance of difference was calculated by Student’s t-test, and
values p < 0.05 were considered to be significant. Correlation and
Total phenolic content (TPC) was determined using the Folin– regression analysis was carried out using Microsoft Excel.
Ciocalteu method (Waterman & Mole, 1994) adapted to a micro-
scale. Briefly, 10 ll of diluted sample solution were shaken for
3. Results and discussion
5 min with 50 ll of Folin–Ciocalteu reagent. Then 150 ll of 20%
Na2CO3 were added and the mixture was shaken once again for
3.1. Recovery percent and phenolic content of L. inermis fractions
1 min. Finally, the solution was brought up to 790 ll by the addi-
tion of distiled water. After 90 min, the absorbance at 760 nm
The crude aqueous ethanolic extract of L. inermis leaves com-
was evaluated using a spectrophotometer (SmartSpecTm3000;
pounds was fractionated by sequential extraction with solvents
Bio-Rad; Hercules, CA). Gallic acid was used as an internal standard
of increasing polarity: n-hexane, chloroform, ethyl acetate, n-buta-
for the calibration curve. The phenolic content was expressed as
nol and water. It is well known that phenolic substances contribute
mg of gallic acid equivalent (GAE)/gram of dry sample using the
directly to the antioxidant activity of plant extracts (Rice-Evans,
linear equation based on the calibration curve.
Miller, & Paganga, 1996). Therefore, it was quite important to eval-
uate the total phenolic content (TPC) of these five fractions. Extrac-
2.5. Antioxidant assays tion yield, TPC, EC50 of DPPH assay and b-carotene bleaching
method of the different solvent fractions obtained from L. inermis
2.5.1. Determination of DPPH radical-scavenging capacity hydroalcoholic extract are given in Table 1. The amount of extract-
Radical-scavenging activity of the different fractions was deter- able compounds, expressed as percentage by weight of dried pow-
mined using DPPH radical as a reagent, according to the method of der, ranged from 0.29% for the chloroform extract to 6.74% and
Kirby and Schmidt (1997), with some modifications. Briefly, 1 ml of 8.84% for the n-butanol and water extracts, respectively. The TPC
a 4% (w/v) solution of DPPH radical in ethanol was mixed with of each fraction was determined according to the Folin–Ciocalteu
500 ll of sample solution in ethanol (different concentrations). method and expressed as mg GAE/g of extract. As shown in Table
The mixture was incubated for 20 min in the dark at room temper- 1, the highest levels of TPC were found in the water extract, fol-
ature. Scavenging capacity was read spectrophotometrically by lowed by ethyl acetate and n-butanol extracts. In contrast, TPC of
monitoring the decrease of the absorbance at 517 nm. Lower absor- n-hexane fraction was the lowest. This result was in agreement
bance of the reaction mixture indicates higher free radical-scaveng- with recent findings of Tunalier, Kosar, Kupeli, Calis, and Baser
ing activity. Ascorbic acid was used as standard. The percent DPPH (2007) who demonstrated that the yield of aqueous extracts of Ly-
scavenging effect was calculated using the following equation: thrum salicaria (Lythraceae) was significantly higher than those ob-
tained with organic solvents. This variation might be due to the
DPPH scavenging effect ð%Þ ¼ ðAcontrol  Asample =Acontrol Þ  100: better solubility of phenolic constituents in water. Additionally, a
linear correlation between the extraction yield and the total ex-
where Acontrol is the absorbance of the control reaction, where the
tracted phenolics was observed (r2 = 0.923).
sample is replaced by 500 ll ethanol. Tests were carried out in
triplicate.
3.2. Antioxidant activity of L. inermis fractions

2.5.2. b-Carotene bleaching assay

3.2.1. DPPH free radical-scavenging activity of L. inermis fractions
The antioxidant activity was determined according to the b-car-
The imbalance between reactive oxygen species (ROS) and anti-
otene bleaching method described by Pratt (1980). A stock solution
oxidant defence mechanisms leads to oxidative modification in
of b-carotene/linoleic acid mixture was prepared by dissolving
cellular membrane or intracellular molecules. Phytochemicals like
0.5 mg of b-carotene in 1 ml of chloroform with 25 ll of linoleic
phenolics, commonly found in plants, could constitute strong nat-
acid and 200 mg of Tween 20. Chloroform was completely evapo-
ural antioxidants and could play an important role in health. The
rated, using a vacuum evaporator. Then, 100 ml of distiled water,
DPPH radical-scavenging assay was used to evaluate the ability
saturated with oxygen (30 min), were added and the obtained
of a compound or an extract to scavenge free radicals and to pro-
solution was vigorously shaken.
mote the formation of the non-radical form DPPH-H by hydro-
Four millilitres of this reaction mixture were dispensed into test
gen-donating action (Chung, Chien, Teng, & Chou, 2006). Free
tubes and 200 ll of each sample, prepared at different concentra-
radical-scavenging capacities of the different fractions, measured
tions, were added. The emulsion system was incubated for 2 h at
by DPPH assay, are shown in Fig. 1a.
50 °C. The same procedure was repeated with BHT as positive con-
The scavenging activity of the five fractions tested was com-
trol, and a blank as a negative control. After this incubation period,
pared to those of ascorbic acid, used as positive control, and was
the absorbance of each mixture was measured at 490 nm. Antiox-
found to be concentration-dependent. This activity can be evalu-
idant activity in b-carotene bleaching model as a percentage (A%)
ated by the determination of the EC50 values, corresponding to
was calculated with the following equation:
the amount of extract required to scavenge 50% of DPPH radicals
present in the reaction mixture. High EC50 values indicate low anti-
A% ¼ 1  ðA0  At =A00  A0t Þ  100; oxidant activity. As shown in Table 1, the ethyl acetate fraction was
the most potent radical scavenger, followed by water and n-butan-
where A0 and A00 are absorbances of the sample and the blank, olic fractions. Chloroform and n-hexane fractions did not show any
respectively, measured at zero time, and At and A0t are absorbances antioxidant activities.
196 A.B. Hsouna et al. / Food Chemistry 125 (2011) 193–200

Table 1
The yield extraction, the total phenolic content, the evaluation of the EC50 values in the DPPH free radical-scavenging and b-carotene bleaching assay of L. inermis leaves crude
fractions. Ascorbic acid was used as standard.

Yield (%) Total phenolic contenta (mg GAE/g)b DPPHa EC50 (lg/ml) b-Carotenea bleaching assay EC50 (lg/ml)
n-Hexane 1.92 50 ± 5.0 >200 11.8 ± 0.3
Chloroform 0.28 73 ± 4.0 >200 >200
Ethyl acetate 1.97 367 ± 19 4.8 ± 0.2 38.7 ± 0.7
n-Butanol 6.41 286 ± 8.0 9.0 ± 0.3 52.0 ± 3.0
Water 8.84 444 ± 38 7.6 ± 0.1 16.5 ± 1.5
Ascorbic acid – – 3.5 ± 0.2 –
a
Each value represents the mean ± SD of three experiments.
b
(mg GAE/g): mg of gallic acid equivalent per g of dry plant extract.

120

100
Scavenging activities (%)

80

60

40

20

0
H CHCL3 EA n-ButOH W A. acid
1 5 10 20 50 (μg/ml)

Fig. 1a. The DPPH free radical-scavenging activity of L. inermis leaves fractions at different concentrations (lg/ml). H: n-hexane, CHCl3: chloroform, EA: ethyl acetate, n-
BuOH: n-butanol, W: water, A. acid: Ascorbic acid. Each value represents the mean ± SD of three experiments.

It has been reported that the free radical-scavenging activity is activities was found to be n-hexane > water > ethyl acetate > n-
greatly influenced by the phenolic composition of the sample (Che- butanol (Table 1).
ung, Cheung, & Ooi, 2003). Thus, the antioxidant activity of the n-
butanol, water and ethyl acetate fractions may be attributed to 3.3. Purification and identification of compounds from L. inermis
their phenolic content. Although the TPC of the aqueous fraction leaves
was significantly higher (p = 0.034) than that of the ethyl acetate
fraction, its antioxidant activity was lower. Tunalier et al. (2007) LC-DAD-ELSD ESI/MSn analysis of the water fraction showed the
reported that the EC50 values from L. salicaria extracts (a Lythracea) presence of tannin compounds, widely known, from the literature,
were found between 0.1 and 2.7 mg/ml. Our results demonstrate to have antioxidant and other biological activities. Of the several
that L. inermis extracts, particularly ethyl acetate, n-butanol and fractions obtained from the hydroalcoholic extract of L. inermis
water fractions, have a better performance against DPPH radical leaves, the n-butanolic was found to have an excellent DPPH radi-
and therefore showed higher antioxidant activities. cal-scavenging activity and to inhibit lipid peroxidation. Active
compounds in the n-butanol extract were possibly different mole-
3.2.2. Antioxidant activity of L. inermis fractions measured by b- cules to those found in the water extract. Thus, the n-butanolic
carotene bleaching method fraction was chosen for the identification of potential compounds
The inhibitory effect on lipid peroxidation was determined by with antioxidant activity.
the b-carotene/linoleic acid bleaching test. This test simulates the After a preliminary survey of the chromatographic profile of this
oxidation of the membrane lipid components in the presence of fraction by LC-DAD-ELSD-ESI/MSn, its fractionation by FC was per-
antioxidants inside the cells. Lipid peroxidation is one of the major formed, in order to isolate its main components. From the n-butan-
causes of quality deterioration in lipid-containing foods. It affects olic extract (1.5 g), 41 fractions were generated. Fractions 7 and 10,
colour, flavour, texture, and nutritive value of foods. b-Carotene, eluted with 65% and 60% water in MeOH, respectively, were re-
in this model system, may be oxidised in the absence of an antiox- chromatographed by semi-preparative reversed-phase HPLC, in or-
idant and the test solution undergoes rapid discoloration. The pres- der to isolate their components. This procedure separated three
ence of extract with antioxidant activity can hinder the extent of b- compounds: 1 (5 mg), 2 (9 mg) and 4 (7 mg) from fraction 7, while
carotene bleaching by neutralising the linoleate-free radical fraction 10 afforded two compounds: 3 (17 mg) and 5 (15 mg). The
formed in the system (Jayaprakasha, Singh, & Sakariah, 2001). As chemical structures of these constituents of henna leaves were
shown in Fig. 1b, n-hexane, ethyl acetate, n-butanol and water identified on the basis of NMR spectroscopic and mass spectromet-
fractions exhibited antioxidant activity in a concentration-depen- ric analyses.
dent manner, with the exception of the chloroform fraction. This Compound 1 was obtained as a white powder and its molecular
fraction was inactive at concentrations less than 200 lg/ml. Based formula C14H18O10 was deduced from ESIMS (m/z 347.3 [M+H]+)
on the comparison of the EC50 values, the order of the antioxidant and 13C NMR spectra. These data together with a complete NMR
 A.B. Hsouna et al. / Food Chemistry 125 (2011) 193–200 197

120

Inhibition percentage (%)

100

80

60

40

20

0
H CHCL3 EA n-ButOH Water BHT

1 2,5 5 10 20 50 (μg/ml)

Fig. 1b. Antioxidant activities of different fractions of L. inermis and BHT at different concentrations (lg/ml). The values represent the percent of inhibition of autoxidation of
the linoleic acid/b-carotene emulsion. Measurements were carried out in triplicate. H: n-hexane, CHCl3: chloroform, EA: ethyl acetate, n-BuOH: n-butanol, W: water, A. acid:
Ascorbic acid. Each value represents the mean ± SD of three experiments.

analysis showed that this compound was lalioside (2,3,4,6-tetra- 361.08951 [M+Na]+, D 0.33 ppm). The 1H NMR spectrum of 5 in
hydroxyacetophenone-2-J-b-D-glucopyranoside). The confirma- CD3OD exhibited the presence of five aromatic protons ascribable
tion of spectral data was made by comparison with the literature to: (i) an isolated proton at dH 6.50 (s, H-3) and, (ii) four protons
(Takeda & Fatope, 1988). at dH 7.22 (ddd, J = 8.5, 7.5 and 1.0 Hz, H-6), 7.40 (ddd, J = 8.5, 7.5
Compound 2, a white amorphous material, possessed a molec- and 1.0 Hz, H-7), 8.03 (dd, J = 8.5 and 1.0, H-5) and 8.33 (dd, J =
ular formula of C22H28O13, which was established on the basis of 8.5 and 1.0, H-8) with an ABCD coupling pattern implying the
its HRESIMS spectrum (m/z 523.14227 [M+Na]+, D 0.12 ppm). occurrence of a 1,2-disubstituted benzene moiety. The 1H NMR
Based on NMR analysis and thanks to the comparison of the data spectrum showed also characteristic signals of a sugar moiety be-
with the reported literature values (Takeda & Fatope, 1988), com- tween dH 3.26 and 4.59 including one anomeric signal (dH 4.59, d,
pound 2 was identified as lawsoniaside (1,2,4-trihydroxynaphtha- J = 8.0 Hz, H-10 ). The sugar appeared to be b-D-glucopyranose
lene-1,4-di-J-b-D-glucopyranoside). according to 13C NMR data (Table 2), where signals belonging to
Compound 3 was isolated as a white powder with a molecular the sugar moiety were four methines at dC 71.0 (C-40 ), 75.6 (C-
formula of C21H20O11 (ESIMS; m/z 449.1 [M+H]+) and NMR data 20 ), 78.0 (C-30 ) and 78.3 (C-50 ), one methylene at dC 62.2 (C-60 )
consistent with those of luteolin-7-J-b-D-glucopyranoside (Lin, and one anomeric methine at dC 108.4 (C-10 ). In addition to the sig-
Chen, Tseng, Lee, & Chen, 2009). nals due to the glucosidic part, the 13C NMR and DEPT spectra of 5
Compound 4, a white powder, had a molecular formula of exhibited 10 sp2 carbon signals between dC 101.3 and 152.6,
C14H18O9 determined by ESIMS (m/z 331.1 [M+H]+) as well as 13C including five methine and five quaternary carbons. On the basis
NMR data. The comparison of its MS and NMR data with those of these data, 1H–1H COSY and HMBC experiments (Table 2) were
described in the literature (Lee, Hong, Kwak, Pyo, & Jee, 1996) al- used to establish the structure of the aglycone part as 1,2,4-tri-
lowed us to identify unambiguously 4 as 2,4,6-trihydroxyaceto- hydroxynaphthalene. The HMBC spectrum also showed the con-
phenone-2-J-b-D-glucopyranoside. nectivity between the sugar moiety and the aglycone part,
The new compound 5 was isolated as a white powder and its thanks to the correlation between C-1 and H-10 . A NOE cross-peak
molecular formula was defined as C16H18O8 by HRESIMS (m/z between H-8 and H-10 confirmed that glucose moiety was involved

Table 2
The NMR spectral data of compound 5.
13 a 1
Position C DEPT Hb HMBC COSY 1H–1H NOESY 1H–1H
1 132.4 C – – – –
2 147.7 C – – – –
3 101.3 CH 6.50 s C-1, C-2, C-4, C-4a – –
4 152.6 C – – – –
4a 121.6 C – – – –
5 123.1 CH 8.03 dd (8.5, 1.0) C-4, C-7, C-8a H-6 –
6 123.1 CH 7.22 ddd (8.5, 7.5, 1.0) C-4a, C-6 H-5, H-7 –
7 127.5 CH 7.40 ddd (8.5, 7.5, 1.0) C-5, C-8a H-6, H-8 –
8 122.3 CH 8.33 dd (8.5, 1.0) C-1, C-4a, C-6 H-7 H-10 ,
8a 131.0 C – – – –
10 108.4 CH 4.59 d (8.0) C-1, C-20 , C-3, C-50 H-20 H-8, H-30 , H-50
20 75.6 CH 3.61 dd (9.0, 8.0) C-10 , C-30 H-10 , H-30 –
30 78.0 CH 3.44 t (9.0) C-20 , C-40 H-20 , H-40 H-10 , H-50
40 71.0 CH 3.52 t (9.0) C-20 , C-30 , C-50 , C-60 H-30 , H-50 H-60
50 78.3 CH 3.26 ddd (9.0, 4.5, 2.5) C-30 H-40 , H-60 H-10 , H-30
60 62.2 CH2 a: 3.85 dd (12.0, 2.5)b: 3.77 dd (12.0, 4.5) C-40 , C-50 H-50 H-40
a
Chemical shifts are shown in the d scale (in ppm); measured at 100 MHz in CD3OD.
b
Chemical shifts are shown in the d scale (in ppm) with J values in parentheses (in Hz); measured at 400 MHz in CD3OD.
198 A.B. Hsouna et al. / Food Chemistry 125 (2011) 193–200

H OH 3.4. Antioxidative properties of the compounds isolated from the L.

HO
inermis leaves
H OH OH
HO
OH 3.4.1. DPPH radical-scavenging activity
H H
H It is well known that the mechanism of reaction between an
O
H H antioxidant and DPPH radical depends on the structural conforma-
H
O OH
tion of the antioxidant. For phenolic compounds, the free radical-
1'
R OH O H scavenging activity is dependent on the presence of free OH groups
8 1 especially at C-3 in flavonoid structures (Sharififar, Dehghn-Nudeh,
HO O H OH
& Mirtajaldini, 2009). In this study, the DPPH radical-scavenging
2
capacities of the five compounds isolated from L. inermis leaves
were evaluated and compared to that of ascorbic acid, used as po-
4 sitive control. The second aim of this part of the work was to con-
OH O OR firm possible structure–antioxidant activity relationships and the
hydrogen-donating ability of the isolated compounds (Cao, Sofic,
1 R=OH 2 R=β-D-glucopyranoside & Prior, 1997).
4 R=H 5 R=H As shown in Fig. 3, the compounds isolated from L. inermis dem-
onstrated a concentration-dependent antiradical activity, by
OH reducing the stable radical DPPH. Among the tested compounds,
H
HO OH
5, 2 and 1 showed potent antioxidant activities and the relative or-
der of DPPH scavenging capacity was found to be: 5 > 2 > 1 > 3 > 4
as depicted in Table 3. Our results showed a significant difference
H
H H in the activity of the naphthalene derivatives 5 and 2. This finding
O suggests that antioxidant activity may be influenced by the num-
HO OH ber of hydroxyls in the aromatic rings of the compound as previ-
H ously reported by Sroka and Cisowski (2003). The comparison of
O O
HO
the results obtained with compounds 5 and 2 showed that the anti-
oxidant capacities of phenolic compounds increased with the num-
ber of hydroxyl groups attached to the aromatic core. Singh,
O’Malley, and Popelier (2005) reported that ortho and para substi-
O OH tution of phenols by electron-donating groups reduced the bond
dissociation energies and increased the transfer rate of hydrogen
3 atom to peroxyl radicals.
Subsequently, it could be noticed that the O-glycosidation of the
Fig. 2. The chemical structures of the phenolic glucosides identified from L. inermis
naphthalene core contributed to the reduction of the antioxidant
leaves.
activity. In the case of 2, the substitution of the 4-hydroxyl group
with a glycosyl group caused a decrease of 23% of the antioxidant
in glycosidation with the naphthalene aglycone at the C-1 hydroxyl capacity when compared to compound 5. In addition, this struc-
group. Therefore, compound 5 was identified as 1,2,4-trihydroxy- ture–activity relationship could also be observed with the aceto-
naphthalene-1-J-b-D-glucopyranoside. phenone-derived compounds 1 and 4. Compound 1 which
Among these compounds, 4 was isolated from this plant species contains three hydroxyl groups showed a higher scavenging power
for the first time, while it is the first report, to our knowledge, of 5 than compound 4, which bears only two hydroxyl groups. There-
as a natural product. The chemical structures of the isolated and fore, an additional hydroxyl group confers to the molecule a better
identified compounds from the n-butanolic part of henna leaves radical-scavenging activity, due to the facts that: (i) more hydro-
hydroalcoholic extracts are depicted in Fig. 2. gen atoms of the phenolic hydroxyl groups could be donated to

120

100
Scavenging activities (%)

80

Compound 1
60
Compound 2
Compound 3
40 Compound 4
Compound 5
Ascorbic acid
20

0
0 10 20 30 40 50 60 70 80 90 100 110
Concentration (μg/ml)

Fig. 3. The DPPH radical-scavenging activity of compounds isolated from L. inermis, expressed as percentages of inhibition (%), versus the positive control (ascorbic acid). Data
represent the means ± SD of three experiments.
 A.B. Hsouna et al. / Food Chemistry 125 (2011) 193–200 199

Table 3 compound 5 showed a potent antioxidant activity in DPPH assay

The antioxidant activities of compounds 1–5 on DPPH and b-carotene assays. while compound 3 demonstrated the highest inhibition of oxida-
Compounds EC50 DPPH % Inhibitiona of lipid tion of the b-carotene/linoleic acid system.
assaya (lg/ peroxidation at Both the antioxidant activity and the phenolic contents found in
ml) 100 lg/ml L. inermis leaves make this plant species a suitable potential source
Lalioside (1) 10.0 ± 0.5 0 of antioxidative natural compounds.
Lawsoniaside (2) 8.0 ± 0.6 18.5 ± 0.1
Luteolin-7-J-b-D-glucopyranoside 17.0 ± 0.7 61.5 ± 0.4
(3) Acknowledgements
2,4,6-Trihydroxyacetophenone- 54.0 ± 1.2 14.0 ± 0.1
2-J-b-D-glucopyranoside (4)
This work was supported by grants from the Tunisian ‘‘Ministry
1,2,4-Trihydroxynaphthalene-1- 6.5 ± 0.8 12.0 ± 0.5
J-b-D-glucopyranoside (5) of Higher Education, Scientific Research and Technology” and from
Ascorbic acid 3.5 ± 0.2 n.d.b ARCUS-CERES MAEE program. The authors wish to thank D. Bonho-
BHT n.d.b 79.0 ± 0.5 mme and Dr. A. Ortalo-Magné (Laboratoire MAPIEM, USTV, France)
a
Each value represents the mean ± SD of three experiments. for their technical assistance during the isolation and characterisa-
b
Not determined. tion of pure compounds, and J.-M. Guigonis (IFR 50, Faculté de
Médecine, Nice, France) for recording the HRESIMS spectra. We
thank also, Dr. M. Siella Trigui and Dr. R. Jarraya-Mezghanni for
stabilise the free radicals, and (ii) the bond dissociation of the their suggestions and help to improve the project paper.
phenoxy group was easiest when the aromatic system was substi-
tuted by electron-donating groups. References

3.4.2. Protection of b-carotene from degradation Ahmed, S., Rahman, A., Alam, A., Saleem, M., Athar, M., & Sultana, S. (2000).
Evaluation of the efficacy of Lawsonia alba in the alleviation of carbon
The antioxidant activity of compounds 1–5 was also investi-
tetrachloride-induced oxidative stress. Journal of Ethnopharmacology, 69(2),
gated on the b-carotene/linoleic acid model system. This activity 157–164.
was compared to BHT, a synthetic antioxidant reference widely Ali, N. A. A., Juelich, W. D., Kusnick, C., & Lindequist, U. (2001). Screening of Yemeni
used in the protection of food. The effects of these five compounds medicinal plants for antibacterial and cytotoxic activities. Journal of
Ethnopharmacology, 74(2), 173–179.
on b-carotene consumption are shown in Table 3. In this assay, the Barlow, S. M. (1990). Toxicological aspects of antioxidants used as food additives. In
highest antioxidative effect was obtained with BHT, which com- Hudson, B. J. F. (Ed.), Food antioxidants (pp. 253–307). Amsterdam: Elsevier;
pletely inhibited b-carotene consumption during incubation. The 1990.
Bhatnagar, S. S., Santapau, H., Desa, J. D. H., Maniar, A. C., Ghadially, N. C., Solomon,
lowest activity against b-carotene consumption was observed for M. J., et al. (1961). Biological activity of Indian medicinal plants, part 1:
compounds 4 and 5 and the highest activity for compound 3. Antibacterial, antitubercular and antifungal action. Indian Journal of Medical
Compound 1 did not show any antioxidant activity in this assay. Research, 49(5), 799–813.
Cao, G. H., Sofic, E., & Prior, R. L. (1997). Antioxidant and prooxidant behavior of
Therefore, the order of b-carotene assay was found to be: flavonoids: Structure–activity relationships. Free Radical Biology and Medicine,
BHT > 3 > 2 > 4 > 5 > 1, as shown in Table 3. We conclude that the 22(5), 749–760.
highest activity for compound 3 could be due to its flavonoid struc- Cheung, L. M., Cheung, P. C. K., & Ooi, V. E. C. (2003). Antioxidant activity and total
phenolics of edible mushroom extracts. Food Chemistry, 81(2), 249–255.
ture together with the number of hydroxyl groups on the flavonoid Chung, Y. C., Chien, C. T., Teng, K. Y., & Chou, S. T. (2006). Antioxidative and
nucleus (Foti, Piattelli, Baratta, & Ruberto, 1996). Moreover, the mutagenic properties of Zanthoxylum ailanthoides Sieb and zucc. Food Chemistry,
solubility of a compound in the reaction environment may affect 97(3), 418–425.
Clarke, D. T., Jones, G. R., & Martin, M. M. (1986). The anti-sickling drug lawsone (2-
its activity. In this case, a hydrophobic compound is more available
OH-1,4-naphthoquinone) protects sickled cells against membrane damage.
to the b-carotene bleaching assay in water emulsion and may be Biochemical and Biophysical Research Communications, 139(2), 780–786.
more active. Thus, the solubility of aglycones in emulsion phase Foti, M., Piattelli, M., Baratta, M. T., & Ruberto, G. (1996). Flavonoids, coumarins, and
cinnamic acids as antioxidants in a micellar system: Structure–activity
is more important than those of glycosides, which naturally move
relationship. Journal of Agricultural and Food Chemistry, 44(2), 497–501.
within the aqueous phase of an oil–water emulsion (Weng & Jayaprakasha, G. K., Singh, R. P., & Sakariah, K. K. (2001). Antioxidant activity of
Wang, 2000). Hence, in comparison to BHT, lower antioxidant grape seed (Vitis vinifera) extracts on peroxidation models in vitro. Food
activities of glycosides isolated in our study may be due to the sol- Chemistry, 73(3), 285–290.
Kirby, A. J., & Schmidt, R. J. (1997). The antioxidant activity of Chinese herbs for
ubility effect and the number of the free hydroxyl groups in pheno- eczema and of placebo herbs. Journal of Ethnopharmacology, 56(2), 103–108.
lic compounds. Lee, K. R., Hong, S. W., Kwak, J. H., Pyo, S., & Jee, O. P. (1996). Phenolic constituents
The constituents of henna act as free radical scavengers that from the aerial parts of Artemisia stolonifera. Archives of Pharmacal Research,
19(3), 231–234.
may prevent phospholipid membrane peroxidation and protect Lin, Y. P., Chen, T. Y., Tseng, H. W., Lee, M. H., & Chen, S. T. (2009). Neural cell
immuno-compromised cells from free radical damage. This may protective compounds isolated from Phoenix hanceana var. Formosana.
explain the previously reported anti-sickling activity, through pro- Phytochemistry, 70(9), 1173–1181.
Linderschmidt, R., Trylka, A., Good, M., & Witschi, H. (1986). The effects of dietary
tecting the membrane of normal and sickle erythrocytes (Clarke, butylated hydroxytoluene on liver and colon tumor development in mice.
Jones, & Martin, 1986). Toxicology, 387(38), 151–160.
Misra, S. B., & Dixit, S. N. (1979). Antifungal activity of leaf extracts of some higher
plants. Acta Botanica Indica, 7(2), 147–150.
4. Conclusion Muhammad, H. S., & Muhammad, S. (2005). The use of Lawsonia inermis Linn.
(henna) in the management of burn wound infections. African Journal of
Biotechnology, 4(9), 934–937.
In the present study, it was found that fractions of the hydroal- Prasad, V., & Gupta, S. C. (1967). Inhibitory effect of bark and leaf decoctions on
coholic extract obtained from leaves of L. inermis possessed antiox- activity of pectic enzymes of Alternaria tenuis. Indian Journal of Experimental
idant and radical-scavenging activities. To the best of our Biology, 5(3), 192–193.
Pratt, D. E. (1980). Natural antioxidants of soybean and other oil-seeds. In M. G.
knowledge, this is the first report of such activities from this plant.
Simic & M. Karel (Eds.), Autoxidation in food and biological systems
These fractions were tested by two distinct assays, in order to (pp. 283–292). New York: Plenum Press.
determine which were the most active. Our results revealed that Rice-Evans, C. A., Miller, N. J., & Paganga, G. (1996). Structure-antioxidant activity
the n-butanolic fraction showed significant antioxidant and free relationships of flavonoids and phenolic acids. Free Radical Biology and Medicine,
20(7), 933–956.
radical-scavenging capacities. Furthermore, five compounds were Sharififar, F., Dehghn-Nudeh, G., & Mirtajaldini, M. (2009). Major flavonoids with
identified and isolated from this fraction. Among them, the new antioxidant activity from Teucrium polium L. Food Chemistry, 112, 885–888.
200 A.B. Hsouna et al. / Food Chemistry 125 (2011) 193–200

Siddiqui, B. S., & Kardar, M. N. (2001). Triterpenoids from Lawsonia alba. Tunalier, Z., Kosar, M., Kupeli, E., Calis, I., & Baser, K. H. C. (2007). Antioxidant, anti-
Phytochemistry, 58(8), 1195–1198. inflammatory, anti-nociceptive activities and composition of Lythrum salicaria L.
Siddiqui, B. S., Kardar, M. N., Ali, S. T., & Khan, S. (2003). Two new and a known extracts. Journal of Ethnopharmacology, 110(3), 539–547.
compound from Lawsonia inermis. Helvetica Chimica Acta, 86(6), 2164–2169. Vasundhara, S., Vijay, K. H., & Jagan, M. R. J. (2008). Influence of milk and sugar
Singh, N., O’Malley, P. J., & Popelier, P. L. A. (2005). Mechanistic aspects of hydrogen on antioxidant potential of black tea. Food Research International, 41(2),
abstraction for phenolic antioxidants. Electronic structure and topological 124–129.
electron density analysis. Physical Chemistry Chemical Physics, 7(4), 614–619. Waterman, P. G., & Mole, S. (1994). Analysis of phenolic plant metabolites. Oxford:
Singh, S., Shrivastava, N. M., Modi, N. T., & Saifi, A. Q. (1982). Anti-inflammatory Blackwell Scientific Publications.
activity of Lawsonia inermis. Current Science, 51(9), 470–471. Weng, X. C., & Wang, W. (2000). Antioxidant activity of compounds isolated from
Sroka, Z., & Cisowski, W. (2003). Hydrogen peroxide scavenging, antioxidant and Salvia plebeia. Food Chemistry, 71(4), 489–493.
anti-radical activity of some phenolic acids. Food and Chemical Toxicology, 41(6), Yang, L., Han, H., Nakamura, N., Hattori, M., Wang, Z., & Xu, L. (2007). Bio-guided
753–758. isolation of antioxidants from the stems of Dendrobium aurantiacum var.
Takeda, Y., & Fatope, M. O. (1988). New Phenolic Glucosides from Lawsonia inermis. Denneanum. Phytotherapy Research, 21(7), 696–698.
Journal of Natural Product, 51(4), 725–729.

The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate.