Antioxidant activity of phenolic and phenylethanoid glycosides from Teucrium polium L

Article abstract of DOI:10.1016/j.foodchem.2011.12.054

In the present study, the chemical composition and antioxidant activities of Teucrium polium L. (Lamiaceae) were assessed. Fractionation of methanol extract obtained from the aerial parts of T. polium yielded one new phenylethanoid glycoside, named poliumoside B, together with four known flavonoids, two iridoid glycosides and a known poliumoside. The structures of all of these compounds were elucidated on the basis of spectroscopic data from 1D and 2D NMR experiments and MS spectral analyses. The antioxidant activities of the crude extracts and of the isolated compounds were evaluated through tests such as DPPH radical-scavenging capability, reducing power, xanthine oxidase inhibitory effect and inhibition of linoleic acid peroxidation. The highest antioxidant activity was found in the n-butanol extract, which also contained active polyphenols, thus suggesting that this plant could be used as a source of natural molecules, to provide safe antioxidant additives and nutraceuticals. The structure-activity relationships of the isolated compounds are also discussed.

Full text of DOI:10.1016/j.foodchem.2011.12.054

Food Chemistry 133 (2012) 21–28
Contents lists available at SciVerse ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Antioxidant activity of phenolic and phenylethanoid glycosides from
Teucrium polium L
Simona De Marino a, Carmen Festa a, Franco Zollo a, Filomena Incollingo b, Gennaro Raimo b,
Giovanna Evangelista b, Maria Iorizzi b,
⇑
a Dipartimento di Chimica delle Sostanze Naturali, Università degli Studi di Napoli ‘‘Federico II’’, Via D. Montesano 49, I-80131 Napoli, Italy
b Dipartimento di Scienze e Tecnologie per l’Ambiente e il Territorio, Università degli Studi del Molise, Contrada Fonte Lappone, I-86090 Pesche (Isernia), Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 June 2011
Received in revised form 28 October 2011
Accepted 17 December 2011
Available online 5 January 2012
In the present study, the chemical composition and antioxidant activities of Teucrium polium L. (Lamia-
ceae) were assessed. Fractionation of methanol extract obtained from the aerial parts of T. polium yielded
one new phenylethanoid glycoside, named poliumoside B, together with four known flavonoids, two irid-
oid glycosides and a known poliumoside. The structures of all of these compounds were elucidated on the
basis of spectroscopic data from 1D and 2D NMR experiments and MS spectral analyses. The antioxidant
activities of the crude extracts and of the isolated compounds were evaluated through tests such as DPPH
radical-scavenging capability, reducing power, xanthine oxidase inhibitory effect and inhibition of lino-
leic acid peroxidation. The highest antioxidant activity was found in the n-butanol extract, which also
contained active polyphenols, thus suggesting that this plant could be used as a source of natural mole-
cules, to provide safe antioxidant additives and nutraceuticals. The structure–activity relationships of the
isolated compounds are also discussed.
Keywords:
Lamiaceae
Teucrium polium
Phenylethanoid glycosides
Poliumoside B
Spectroscopic analysis
Antioxidant activity
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
evaluated through the use of several commonly accepted in vitro
assays, which included DPPH scavenging (DPPH), reducing power
Recently many antioxidant metabolites have been isolated from
edible and medicinal plants like oregano, sage and thyme that
grow in the Mediterranean area, of which several belong to the
Lamiaceae family.
(RP), xanthine oxidase inhibitor effect (XOI) and antioxidant activ-
ity in a linoleic acid system (ALP).
2. Materials and methods
The Teucrium genus (Lamiaceae) includes 300 species distrib-
uted all around the world. Decoctions or infusions of representa-
tives of this genus have been used in folk medicine as stimulants,
tonics, diaphoretics, diuretics and in the treatments of stomach
pain, asthma, amenorrhea, leucorrhoea, chronic bronchitis and gout
(Ulubelen, Topcu, & Sonmez, 2000). The Teucrium sp. contains a
number of compound classes that include phenylpropanoid glyco-
sides (Serrilli et al., 2007), iridoid glycosides (Ruhdorfer & Rimpler,
1981), and flavonoids (Rizk, Hammouda, Rimpler, & Kamel, 1986),
which have been reported in the literature to have antioxidant prop-
erties (Sharififar, Dehghn-Nudeh, & Mirtajaldini, 2009) and neocle-
rodane diterpenes (Bruno et al., 2000).
2.1. Reagent and chemicals
2,2-Diphenylpicrylhydrazyl radical (DPPHꢀ), linoleic acid, xan-
thine, 2,6-di-tert-butyl-4-methylphenol (BHT), xanthine oxidase
(XO), ascorbic acid (AA), 2-thiobarbituric acid (TBA), Trolox, Tween
20, potassium ferricyanide, Tris buffer, potassium phosphate and
EDTA were purchased from Sigma–Aldrich Chemical Co. (St. Louis,
MO).
2.2. General experimental procedure
The aim of the present study was the investigation of the anti-
oxidant power of the crude extracts of Teucrium polium followed by
the chemical characterisation of selected pure compounds isolated
from an n-butanol extract, which demonstrated the highest antiox-
idant properties. In particular, their antioxidant activities were
High-resolution ESI mass spectrometry (HRESI-MS) was carried
out on a Micromass QTOF spectrometer and electrospray ionisation
mass spectrometry (ESI–MS) experiments were performed on an API
2000 triple-quadrupole mass spectrometer (Applied Biosystems,
Foster City, CA). Optical rotations were determined on a P-2000
polarimeter (Jasco, Tokyo).
⇑
NMR spectra were obtained on a Varian (Palo Alto, CA) Inova
Corresponding author. Tel.: +39 0874 404100; fax: +39 0874 404123.
E-mail address: iorizzi@unimol.it (M. Iorizzi).
500 NMR spectrometer (1H at 500 MHz and 13C at 125 MHz)
0308-8146/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.foodchem.2011.12.054

22
S. De Marino et al. / Food Chemistry 133 (2012) 21–28
equipped with Sun hardware, d (ppm), J in Hz, spectra relative to
CD3OH (dH = 3.31) as internal standard. Heteronuclear single-quan-
tum coherence (HSQC) spectra were optimised for an average JCH
was separated with MeOH/H2O (4:6) as eluent to give mainly com-
pound (2) (6.4 mg).
1
Fractions D–F were separated with 35% of MeOH/H2O + 0.1%
trifluoroacetic acid (TFA) (flow rate 1.5 mL/min) to give luteolin 7-
O-rutinoside (5) (4.0 mg), luteolin 7-O-neohesperidoside (6)
(4.7 mg), luteolin 7-O-glucoside (7) (2.0 mg) and luteolin 40-O-glu-
coside (8) (2.2 mg).
of 140 Hz; the gradient-enhanced heteronuclear multiple bond
3
correlation (HMBC) experiment were optimised for a JCH of 8 Hz.
Nuclear Overhauser effect (NOE) measurements were performed
by 2D ROESY experiment. Through-space 1H connectivities were
evidenced using a ROESY experiment with mixing times of 300 ms.
HPLC was performed using a Waters 510 pump equipped with a
Rheodyne injector and a Waters 401 differential refractometer as
Leaves and flowers decoction of T. polium was also prepared.
Decoction was prepared following the traditional preparation by
adding 100 g of plant material to 1.0 L of water heated to 95 °C. After
the decoction has begun to boil, the volume was reduced by around
500 mL. The decoction was then cooled and filtered, and evapora-
tion of aqueous extract, under vacuum at 45 °C, afforded 22.5 g of
a glassy material. An aliquot (10 g) of aqueous extract was dissolved
at room temperature in MeOH (3 ꢀ 500 mL), concentrated under re-
duced pressure to give 2.5 g of the methanol extract that was then
fractionated by DCCC using CHCl3/MeOH/H2O (7:13:8) in the
ascending mode under the conditions discussed above. Four major
fractions (A1–D1) were obtained. Fraction A1 contained a mixture
of free sugars; fractions B1 and C1 contained mainly teulamifin B
(9) and teusalvin C (10), respectively, and were purified by HPLC un-
der the conditions described above.
detector, using a 30 cm ꢀ 3.9 mm; i.d., C18
l-Bondapak (Waters,
Milford, MA) column; flow rate was 1 mL minꢁ1
.
The GC/MS analysis was carried out with an Agilent Technolo-
gies 6890 N gas chromatograph coupled to an Agilent Technologies
5973 N quadrupole mass selective spectrometer and provided with
a split/splitless injection port (Agilent Technologies, Santa Clara,
CA). Helium was used as carrier gas at a linear velocity of 40 cm/s.
Droplet counter-current chromatography (DCCC) was per-
formed on a DCC-A apparatus (Tokyo Rikakikai Co., Tokyo, Japan)
equipped with 250 glass columns. The purity of compounds was
determined to be greater than 95% by HPLC, MS and NMR.
2.3. Plant material
2.5. Acid hydrolysis of 1
Selected samples of wild growing Teucrium polium L. (Lamia-
ceae) were collected during the blooming period in the mountain
areas of Isernia (Italy) in June 2009. Plants were identified at Dipar-
timento di Scienze e Tecnologie per l’Ambiente e il Territorio (Uni-
versity of Molise) and a voucher specimen was deposited under No.
TP-77–09 in the Herbarium of University of Molise (Pesche, Iser-
nia). Leaves and flowers were frozen at ꢁ80 °C under liquid nitro-
gen and then stored at ꢁ20 °C until analysed.
Compound 1 (0.5 mg) was hydrolysed with 2 N CF3CO2H (2 mL)
at 110 °C in a sealed tube for 8 h. After cooling, the solution was
diluted with H2O (5 mL) and extracted with AcOEt (3 ꢀ 2 mL). The
aqueous layer was evaporated to dryness under reduced pressure
and the residue was reacted with 0.1 M
hydrochloride in anhydrous pyridine (200
L
-cysteine methyl ester
lL) for 1 h at 60 °C (Hara,
Okabe, & Mihashi, 1987). 1-(Trimethylsilyl)imidazole in pyridine
was added and the thiazolidine derivatives analysed by GC–MS.
L-Rhamnose,
L-arabinose and D-glucose were confirmed in 1 by
2.4. Extraction and isolation
comparison of the retention times of their derivatives with those of
D
-rhamnose (tR = 12.60 min) and
-arabinose (tR = 12.03 min) and -arabinose (tR = 10.91 min);
-glucose (tR = 19.20 min).
L
-rhamnose (tR = 12.14 min);
An 80-g portion of the frozen aerial parts of Teucrium polium L.
was extracted with MeOH (3 ꢀ 700 ml) for 12 h at room tempera-
ture. The combined extracts (15 g) were concentrated and subjected
to a modified Kupchan’s (Kupchan, Britton, Ziegler, & Siegel, 1973)
partitioning procedure as follows. The MeOH extract was dissolved
in 10% aqueous methanol and partitioned against n-hexane. The
water content (% v/v) of the MeOH extract was adjusted to 40%
and partitioned against CHCl3, to furnish a CHCl3 extract (2 g). The
aqueous phase was concentrated to remove MeOH and then
extracted with n-BuOH, yielding 3.5 g of glassy material (n-BuOH
extract). The CHCl3 extract (1.0 g) was fractionated by DCCC using
CHCl3/MeOH/H2O (7:13:8) in the ascending mode (the lower phase
was the stationary phase), flow rate 8 mL/min; 4-mL fractions were
collected. Fractions were monitored by TLC on SiO2 with CHCl3/
MeOH/H2O (80:18:2) as eluent and combined on the basis of their
similar TLC retention factors. Three major fractions were obtained
and then separated by HPLC: fraction 2 was purified by HPLC with
MeOH/H2O (2:8) as eluent, to afford 1.2 mg of teulamifin B (9); frac-
tion 3 was purified by HPLC with MeOH/H2O (3:7) to give teusalvin
C (10) (5.5 mg).
D
L
D-glu-
cose (tR = 17.88 min) and
L
The hydrolysate of 1 was subjected to GC–MS for identification of
the sugar moieties on an HP-5MS capillary column (30 m ꢀ
0.25 mm, 0.25 lm film thickness; Agilent). GC oven temperature
was kept at 180 °C. The injector temperature was 230 °C. The tem-
peratures of the ion source and the transfer line were 250 and
280 °C, respectively. Mass spectra were taken at 70 eV and the mass
range was from 40 to 350 amu.
Compound 1:
½
a 2D5
ꢂ
þ 37:7 (c 0.3 MeOH); HR-ESI–MS m/z
925.2947 [M + Na]+. (Calcd. for C40H54O23 Na: 925.2954); NMR
data in Table 2. ESIMS spectrum (positive ion): pseudomolecular
ion peak at m/z 925 [M + Na]+.
2.6. Antioxidant activity assays
The antioxidant activities of extracts and of each isolated com-
pound were evaluated in vitro through the use of four different
methods: DPPH scavenging test, reducing power assay (RP), xan-
thine oxidase inhibition (XOI) and anti-lipid peroxidation test (ALP).
The n-BuOH extract (2.0 g) was submitted to DCCC with n-BuOH/
Me2CO/H2O (3:1:5) in the descending mode (the upper phase was
the stationary phase). The obtained fractions were monitored by
TLC on silica gel plates with n-BuOH/OHAc/H2O (12:3:5) and
CHCl3/MeOH/H2O (80:18:2) as eluents. Six fractions A–F were ob-
tained and purified by HPLC. Fraction A (67.8 mg) was separated
with MeOH/H2O (25:75) as eluent (flow rate 0.8 ml/min) and con-
tained 2.3 mg of pure compound (3) and 3.0 mg of compound (4);
fraction B (70 mg) was purified by HPLC with MeOH/H2O (3:7) as
eluent and contained compound (1) (4.8 mg); fraction C (152 mg)
2.6.1. DPPH decoloration assay
The quenching of free radicals by extracts, fractions and purified
compounds was evaluated spectrophotometrically at 517 nm by
the decoloration of a methanol solution of DPPH radical, a stable
organic free radical with an absorption band at 517 nm which loses
this absorption on accepting an electron or a free radical species
(Feresin et al., 2002). A freshly prepared DPPH solution (84 lM)
was used for the assays. Samples were dissolved in methanol and
the DPPH solution served as a control. The degree of decoloration

S. De Marino et al. / Food Chemistry 133 (2012) 21–28
23
indicates the free radical-scavenging efficiency of the samples (SA,
scavenging activity). Ascorbic acid was used as the reference free
radical scavenger. The extent of DPPH decoloration was evaluated
at different concentrations of each sample and scavenging activity
was then calculated as follows:
2.6.4. Inhibition of linoleic acid peroxidation
The capability of the extracts and of the isolated compounds to
limit the extent of non-enzymatic peroxidation of unsaturated
fatty acids was investigated by adapting the spectrophotometric
assay already described by Zarghami and Khosrowbeygi (2004).
One millilitre final volume of buffer B (200 mM phosphate, pH
7.8, 0.4 mM FeCl2, 0.7% v/v Tween 20), containing 23 mM linoleic
acid, was incubated at 37 °C in the presence of increasing amounts
of different compounds. In order to evaluate the amount of mal-
ondialdehyde (MDA), one of the final products of lipid peroxidation
SA ¼ ½ðAcompound ꢁ AblankÞ=ðAreference ꢁ Ablank
Þ
where Acompound and Ablank correspond to the absorbance of the
reaction mixture measured in the presence or absence of the sam-
ples, respectively; Areference indicates the absorbance measured in
reaction, 250-
24 h incubation time, supplemented with 100
50 L HCl (10% v/v), 50 L of 1% (w/v) TBA and then boiled. Twenty
minutes later the reaction mixtures were cooled on ice, supple-
mented with 900 L n-butanol and stirred for 5 min. These sam-
l
L aliquots of each mixture were withdrawn after
the presence of 10 lg/mL of ascorbic acid.
lL BHT (0.8 g/L),
The evaluation of the slope obtained by plotting the decoloration
extent, expressed as ln(1 – SA), versus the respective concentration
value of the sample tested allowed the calculation of the EC50 value
of each compound as follows:
l
l
l
ples were then centrifuged at 4 °C at 14,000 rpm for 5 min. The
absorbance at 532 nm of the n-butanolic phase was then
measured.
EC50 ¼ lnð0:5Þ=slope
The extent of linoleic acid peroxidation was expressed as:
2.6.2. Determination of reducing power
ALP ¼ ½ðAcompound ꢁ AreferenceÞ=ðAmaximum ꢁ AreferenceÞꢂ
The reducing power of all samples was determined adapting the
method described by Zhu, Lian, Guo, Peng, and Zhou (2011), using
a modified Fe3+ to Fe2+ reduction assay, whereby the yellow colour
of the test solution changes to various shades of green and blue,
depending on the reducing power of the samples. The presence
of antioxidants in the samples, which causes the reduction of the
Fe3+/ferricyanide complex to the Fe2+ form, can be monitored by
measurement of the formation of Perl’s Prussian blue at 700 nm.
Therefore, appropriate amounts of each sample were incubated
in 1 mL buffer A (50 mM phosphate buffer, pH 7.8, supplemented
with 0.1 mM EDTA), containing 0.5% (w/v) potassium ferricyanide
and 5 % (v/v) methanol. The mixtures were incubated at 50 °C for
where Acompound and Amaximum correspond to the absorbance mea-
sured in the presence or in the absence of the samples, respectively;
Areference indicates the absorbance measured in the presence of
0.25 mM Trolox used as positive control.
The evaluation of the slope observed by plotting lnALP versus
the respective concentration value of the sample tested allowed
the calculation of the IC50 by the following formula:
IC50 ¼ lnð0:5Þ=slope:
30 min and then 300 lL 10% (w/v) trichloroacetic acid were added
2.6.5. Statistical analyses
to each sample. After vigorous stirring, the mixtures were centri-
fuged at 14,000 rpm at 4 °C for 5 min. Four hundred microlitres
The dependence of the antioxidant activity on the sample con-
centration was evaluated using a standard curve-fitting program.
At each single concentration the activity was evaluated in tripli-
cate. The correlation coefficient r2 between the concentration of
the samples and the relative effect was always greater than
0.900. Correlation and regression analyses and statistical analysis
were made with Student’s test, using the software KaleidaGraph.
Values of p lower than 0.05 were considered to be significant.
of each sample were withdrawn and mixed with 400
lL water
and 100 L 0.1% (w/v) ferric chloride and the absorbance was mea-
l
sured at 700 nm. Ascorbic acid was used as positive control and as
reference. The mean absorbance values were plotted against sam-
ple concentration, and a linear regression analysis was carried out.
The reducing power (RP) was expressed in ascorbic acid equiva-
lents (AAE), which are defined as milligrams of ascorbic acid per
gram of extract. The higher the AAE value was, the higher the
reducing power was.
3. Results and discussion
3.1. Isolation and identification of secondary methabolites from T.
polium
2.6.3. Inhibition of XO activity
Using xanthine as substrate, the enzyme activity of XO was
measured spectrophotometrically by following uric acid formation,
adapting the method described by Marcocci, Suzuki, Tsuchiya, and
Packer (1994). The XO assay was carried out in a 500-
mixture, containing 10% (v/v) methanol, 40 M xanthine as the
substrate, 1.2 mg/mL xanthine oxidase and different amounts of
each compound as potential XO inhibitor. The assay was initiated
by adding the enzyme to the reaction mixture with or without
inhibitors and followed in kinetics by reading the absorbance at
295 nm for 2 min at 20 °C, thus calculating the respective rate of
the reaction. For each sample, the extent of the inhibitory effect
(i) was evaluated as follow: i = (V0–Vi)/V0, where V0 and Vi are the
rates measured in the absence or presence of each concentration
of the compound used [I], respectively. The IC50 value is calculated
from the slope of the double reciprocal plot between i and [I],
according to the following equation:
The methanolic crude extracts of the aerial parts of Teucrium
polium were submitted to extractive and chromatographic analy-
ses. Four extracts were obtained using n-hexane, CHCl3, n-BuOH
and water. The antioxidant capacity of each extract was assessed
by testing its DPPH radical scavenging capacity (DPPH), its reduc-
ing power (RP), its xanthine oxidase inhibitory effect (XOI) and
its lipid peroxidation inhibitory effect (ALP). As reported in Table 1,
among these extracts, the n-butanol extract was the most bioac-
tive, showing the highest DPPH radical-scavenging effect (EC50
lL reaction
l
7.8
(IC50 40.5
l
g/ml) as well as RP (424 AAE), XOI (IC50 35.4
g/mL).
lg/mL) and ALP
l
The n-BuOH extract was submitted to DCCC and HPLC purifica-
tion and yielded the new phenylethanoid glycoside (1), a poliumo-
side (2) (Andary et al., 1985), an 8-O-acetylharpagide (3)
(Hannedouche, Jacquemond-Collet, Fabre, Stanislas,
& Moulis,
1999), a teucardoside (4) (Ruhdorfer & Rimpler, 1981), along with
flavonoid glycosides: luteolin 7-O-rutinoside (5) (Wang et al.,
2003), luteolin 7-O-neohesperidoside (6) (Gattuso, Barreca, Caristi,
Gargiulli, & Lezzi, 2007), luteolin 7-O-glucoside (7) (Williams &
1=i ¼ 1 þ IC50=½Iꢂ
The IC50 values represent the extent of XO inhibition (XOI)
shown by each extract or compound analysed.

24
S. De Marino et al. / Food Chemistry 133 (2012) 21–28
Table 1
sides teulamifin B (9) (Malakov, Boneva, Papanov, & Spasov, 1988)
Antioxidant activity tests of T. polium leaves extracts.
and teusalvin C (10) (Bruno et al., 2000) were the most abundant
components and had been previously reported as typical of Teucri-
um spp. The chemical structures of all of these compounds were
established on the basis of their 1D and 2D NMR spectral proper-
ties, ESIMS data and comparison with values from the literature.
The structures of the isolated natural compounds are shown in
Fig. 1.
Extracts
DPPHa
RPb
(AAE)
XOIc
(IC50
ALPd
(IC50 lg/ml)
(EC50
l
g/ml)
l
g/ml)
n-Hexane
CHCl3
n-BuOH
Aqueous
>300
>300
7.8
26
22
424
307
152.5
n.a.
35.4
84.6
503.3
187.1
40.5
10.8
59.9
n.a. – no activity detected.
The new compound 1, named poliumoside B, in its HR–ESI–MS
showed a pseudomolecular ion at m/z 925.2947 ([M + Na]+) sug-
gesting the molecular formula C40H54O23, which was in agreement
with the observation of 40 carbon resonances in its 13C NMR spec-
trum (Table 2).
a
DPPH-scavenging activity (DPPH).
Reducing power (RP).
Xanthine oxidase inhibitory effect (XOI).
Anti-lipid peroxidation activity (ALP).
b
c
d
Also, in the 1H NMR spectrum of compound 1 the characteristic
signals belonging to (E)-caffeic acid (C9) and 3,4-dihydroxyphenyl-
ethyl (C8) moieties were observed. Additionally, four anomeric pro-
ton resonances were detected at dH 5.47 (d, J = 1.4 Hz, H-1000 Rha II),
4.61 (d, J = 1.4 Hz, H-100 Rha I), 4.37 (d, J = 7.9 Hz, H-10 Glc) and 4.30
(d, J = 7.3 Hz, H-10000 Ara), thus indicating its tetrasaccharidic struc-
ture. The results of COSY, HSQC, TOCSY and HMBC experiments (Ta-
ble 2) allowed us to assign all the 1H and 13C chemical shifts of the
tetrasaccharide unit and of the aromatic systems. Acid hydrolysis
with 2 N CF3CO2H of poliumoside B (1) produced rhamnose (ꢀ2),
Table 2
1H and 13C NMR data (CD3OD, 500 and 125 MHz) for compound 1.
a
a
Position
dH
dC
HMBC
Caffeoyl
1
2
3
4
5
6
7
8
9
–
127.6
114.9
147.1
150.0
116.2
122.9
147.8
114.3
167.6
7.05 s
–
–
6.77 d (8.1)
6.95 dd (1.2, 8.1)
7.62 d (15.7)
6.27 d (15.7)
-
3, 4, 7
1, 3
2, 4, 7
1, 2, 6, 9
1, 9
glucose and arabinose. The
the arabinose units and the
L-configuration of the rhamnose and
D-configuration of glucose was inferred
by the GC–MS analysis (Hara et al., 1987). The caffeoyl system was
located at C-40 of the glucose unit on the basis of the strong deshiel-
ding of H-40 signal at dH 4.99 (t, J = 9.9 Hz). In the HMBC experiment,
this system showed a cross-peak correlation with a carbonyl carbon
atom (167.6 ppm) confirming the acyl position (Fig. 2).
Phenylethylalcohol
1
2
3
4
5
6
a
–
131.1
116.9
146.9
144.8
116.1
120.9
72.2
6.70 s
–
–
6.67 d (8.0)
6.57 dd (1.2, 8.0)
3.98 m
4, 6, 7
In the same experiment, the anomeric proton of glucose (dH 4.37)
1, 3
2, 4, 7
1, 7, 10Glc
1, 7, 10Glc
1, 6, 8
showed a correlation with
a-C atom (dC 72.2) of the phenylethyl
alcohol moiety. Moreover, the HMBC experiment also allowed the
determination of all of the relevant interglycosidic linkages
(Fig. 2). More information on the spatial conformation of the mole-
cule, was obtained from a ROESY experiment that showed the dipo-
lar correlations between H-100of Rha I (dH 4.61) and H-60 of Glc (dH
3.47); between H-1000 of Rha II (dH 5.47) and H-30 of Glc (dH 3.76);
H-10000 of Ara (dH 4.30) with H-3000 of Rha II (dH 3.65); and anomeric
3.73 m
b
2.80 m
36.4
Glucose
10
20
30
40
50
60
4.37 d (7.9)
3.37b
104.0
75.7
82.1
70.0
74.3
67.3
3.76b
4.99 t (9.9)
30, 50, 9
3.69b
proton of Glc (dH 4.37) with a-H (dH 3.73) of phenylethanoid moiety.
3.47 dd (11.4, 5.6)
3.76b
In the ROESY experiment, significant cross peaks were also detected
in the caffeoyl system: H-7 (dH 7.62)/H-2 (dH 7.05) and H-7/H-6 (dH
6.95) and H-b (dH 2.80)/H-2 (dH 6.70) and H-b/H-6 (dH 6.57) in the
phenylethyl alcohol residue. Based on the above data, the structure
of compound 1 was identified as 3,4-dihydroxy-b-phenylethoxy-O-
Rhamnose I
100
200
300
400
500
600
4.61 d (1.4)
3.84b
101.9
71.6
71.8
73.7
69.8
17.7
500, 60Glc
3.68b
3.35b
[
a-
L
-arabinopyranosyl-(1 ? 2)-
a
-
L
-rhamnopyranosyl-(1 ? 3)]-[
a-
3.60b
L
-rhamnopyranosyl-(1 ? 6)]-4-O-caffeoyl-b-
D
-glucopyranoside, for
1.20 d (6.1)
400, 500
which the trivial name poliumoside B is proposed.
Rhamnose II
1000
2000
3000
4000
5000
6000
5.47 d (1.4)
3.95 dd (1.4, 3.2)
3.65b
101.7
82.6
71.9
73.8
70.1
18.2
5000, 30Glc
10000 Ara
3.2. Antioxidant activities of T. polium extracts and their isolated
compounds
3.28 t (8.6)
3.53b
Because different antioxidant compounds may act through dif-
ferent mechanisms, no single method can fully evaluate the total
antioxidant capacity or give a comprehensive prediction of the
antioxidant efficacy of different compounds Therefore, four differ-
ent assays were used to study in vitro antioxidant activity of ex-
tracts and isolated compounds from T. polium.
1.05 d (6.1)
4000, 5000
Arabinose
10000
4.30 d (7.3)
3.58b
107.1
72.5
74.1
69.5
67.2
2000 Rha II
20000
30000
3.50b
40000
3.77b
50000
3.53b
3.86b
3.2.1. DPPH scavenging activity
a
Coupling constants are in parentheses and given in hertz. 1H and 13C assign-
In the present study, the different extracts of T. polium showed a
dose-dependent DPPH-scavenging activity which varied among
the different extracts. As shown in Table 1, the n-butanol extract
possessed the strongest DPPH-scavenging activity, showing an
ments aided by COSY, TOCSY, HSQC and HMBC experiments.
b
Signal patterns are unclear due to overlapping.
Harborne, 1975), and luteolin 40-O-glucoside (8) (Kellam, Mitchell,
EC50 of 7.8
lg/mL. The results obtained led us to select this extract
Blunt, Munro, & Walker, 1993). Neo-clerodane dieterpenoid glyco-
for the isolation of pure antioxidants.

S. De Marino et al. / Food Chemistry 133 (2012) 21–28
25
Fig. 1. Natural compounds isolated from Teucrium polium L.
Fig. 3 shows the DPPH-scavenging activity for each class of
chemical compounds isolated from the n-butanol fraction: flavo-
noids (panel C), phenylethanoids (panel A), and iridoids (panel B).
Flavonoids represent a class of phenolic metabolites that have
significant antioxidant and chelating properties. Their propensity
to inhibit free radical mediated events is governed by their chem-
ical structure. This structure–activity relationship has been well
established in vitro and has been demonstrated by many studies
requirements include an ortho-dihydroxy substitution in the B
ring, a C2–C3 double bond and a C4 carbonyl group in the C ring.
The free hydroxyl groups on the B-ring donate hydrogen to a rad-
ical, thus stabilising it and giving rise to a relatively stable flavo-
noid radical. Flavones with the C2–C3 double bond in
conjugation with a C4 carbonyl group are planar; this structural
feature allows a charge delocalisation throughout the three rings
system. In flavonoids with the ortho-dihydroxy (catechol) group,
the formation of flavonoid phenoxy radicals may be stabilised by
´
(Heim, Tagliaferro, & Bobiya, 2002; Amic et al., 2007). The isolated
´
flavone glycosides 5–8, [luteolin 7-O-rutinoside (5), luteolin 7-O-
neohesperidoside (6), luteolin 7-O-glucoside (7), luteolin 40-O-glu-
coside (8)] have shown to be active antioxidants (Table 3). The
mechanism behind the radical-scavenging activity of flavonoids
is thought to be hydrogen atom donation and the structural
the mesomeric equilibrium to ortho-semiquinone structures (Amic
et al., 2007). Aglycones are more potent antioxidants than their
corresponding glycosides. O-glycosylation of the A or B ring de-
creases the antiradical power. All the tested glycosylated flavones
(5–8) from T. polium have the 30,40-catechol, the C2–C3 double

26
S. De Marino et al. / Food Chemistry 133 (2012) 21–28
Table 3
Antioxidant activity tests of pure compounds isolated from n-BuOH extract of T.
polium leaves.
Compound
DPPH (EC50
lM)
RP (AAE)
XOI (IC50
lM)
ALP (IC50 lM)
1
2
3
4
5
6
7
8
4.8
7.0
573
298
<20
<20
306
80
345.3
690.5
n.a.
16.6
107.1
15.0
18.6
36.3
72.8
149.2
54.6
24.9
23.3
23.7
61.6
13.4
72.7
n.a.
106.8
28.7
10.7
6.7
108
<20
For the abbreviations see Table 1.
n.a. – no activity detected.
Fig. 2. Key HMBC correlations for poliumoside B (1).
bond and C4-carbonyl functions. Luteolin 7-O-glucoside (7) exhib-
ited an EC50 of 13.4 M, while the antioxidant activity of luteolin
40-O-glucoside (8) (EC50 72.7
M), which has the same glycosyla-
tion at position 40 of the B-ring, was significantly decreased These
l
l
´
data were consistent with those reported by Amic et al. (2007),
which suggest that the loss of the catechol function, due to glyco-
sylation, is accompanied by a loss in the ability to form a stable
quinone, which is a structural requirement for the stabilisation of
the resulting radical. However, despite the presence of catechol
on the B ring, a disaccharide moiety bound to the C7 of the A-ring
weakens the antioxidant effect, as observed in luteolin 7-O-rutino-
side (5) and luteolin 7-O-neohesperidoside (6), which showed EC50
values of 23.7 and 61.6 lM, respectively, higher than that exhib-
ited by luteolin 7-O-glucoside (7) (see Fig. 4).
Fig. 4. Kinetic assays of xanthine oxidase inhibition by compound 8. A Lineweaver–
Burk double-reciprocal plot was constructed for the inhibition of xanthine oxidase
by 8. The plot is expressed as 1/
the following concentration of inhibitor in the reaction solution: 1
g/mL (---) and
g/mL (–ꢃ–ꢃ–). Values show means from experiments
performed in triplicate.
vi versus 1/[xanthine] (l
Mꢁ1) without (–––) or with
Worthy of note is the fact that both compounds 5 and 6 contain
glucose and rhamnose units which differ in their interglycosidic
linkage. Luteolin 7-O-neohesperidoside (6) has a rhamnose unit
linked to C-200 of glucose, which results in a reduction, probably
generated by a steric hindrance, of its antioxidant activity. The flav-
ones examined showed different solubility patterns and these dif-
ferences may have influenced the antioxidant activity.
The DPPH-scavenging activity of poliumoside B (1) and poli-
umoside (2) were also examined. They are phenylethanoid glyco-
sides with an (E)-caffeoyl and 3,4-dihydroxyphenylethyl moieties
linked at C-40 and C-10 of the glucosyl unit. The most significant dif-
ference between them was the presence of an additional arabino-
syl unit in the poliumoside B (1) which is a phenylethanoid
tetrasaccharide. A recent paper reported the significant antioxidant
activity of verbascoside, a phenylethanoid glycoside isolated from
a devil’s claw suspension culture (Georgiev et al., 2010). As re-
ported in Table 3, of all of the compounds purified from the T. poli-
um n-BuOH extract, compounds 1 and 2 displayed the highest
l
g/mL (–––),
2
l
4 l
DPPH radical-scavenging activities (EC50 of 4.8 and 7.0 lM, respec-
tively). In particular, these two poliumosides have several free hy-
droxyl groups which could act as hydrogen donors. Accordingly, 1
and 2 possess the (E)-caffeoyl residue, the C7–C8 unsaturation and
the C9-keto group responsible for the co-planarity and for the elec-
tronic delocalisation throughout the molecule, leading to the stabi-
lisation of the aryloxy radical. The greater antioxidant activity,
with respect to flavonoid glycosides, could be attributed to the
presence of a second phenolic ring with an ortho-dihydroxy group.
Of the flavonoids here reported, only compound 7 was endowed
with a scavenger activity comparable to that of 1 and 2 (Table 3).
Few iridoid glycosides have demonstrated antioxidant proper-
ties. Although iridoid glycosides 3 and 4 were less active than phe-
Fig. 3. DPPH-scavenging activity of T. polium isolated compounds: 1 (s), 2 (d), 3 (}) 4 (ꢁ), 5 (4), 6 (N), 7 (j) and 8 (h). Values show mean from experiments performed in
triplicate.

S. De Marino et al. / Food Chemistry 133 (2012) 21–28
27
nylethanoids, their EC50 values suggest the ability to act as antiox-
idants. The radical scavenging activity of 3 (EC50 24.9 M) could be
related to the presence of a C8-acetyl group which readily leads to
the formation of an -carbonyl radical stabilised by charge delocal-
isation. In the iridoid diglycoside 4, which has an EC50 of 23.3 M,
the activity is probably related to the presence of an ,b-unsatu-
the O-glycosylation at C7 leads to a reduction of the XO inhibitory
power (da Silva et al., 2004). Furthermore, the presence of a free
hydroxyl group at C40 position in compound 5, 6 and 7 makes them
more easily ionisable and therefore less able to interact with the
hydrophobic channel, which is the main access to XO active site.
Besides, the presence of a sugar residue also reduces the inhibitory
effect (Lespade & Bercion, 2010). Nevertheless, in the luteolin
group, compound 7 showed the highest activity because of the
lower steric hindrance of its sugar moiety, with respect to those
of compounds 5 and 6 (da Silva et al., 2004).
l
a
l
a
rated ketone, which generates radical delocalisation on the cyclo-
pentenone ring.
3.2.2. Reducing power
The reducing power of the T. polium extracts, as well as that of
the isolated compounds, in most cases reflects what has already
been described for their DPPH-scavenging activity. Poliumoside B
(1) showed a remarkable activity (573 AAE), which was higher
than that of the n-butanol extract (424 AAE, Table 3) while in the
luteolin group, only compound 5 showed a considerable reducing
activity (306 AAE). Iridoids 3 and 4 did not show any detectable
reducing power.
The main factor that influences the reducing activity of mole-
cules such as polyphenols seems to be the presence of a catechol
group in their structure (DeGraft-Johnson et al., 2007). Indeed,
the metabolites with the greatest biological activity were the phe-
nylethanoid glycosides 1 and 2, which are characterised by the
presence of two catechol units. Metabolites 6 and 7 possess a sin-
gle catechol group and, as expected, showed a reducing power low-
er than compounds 1 and 2. Flavone 8, in which a sugar unit is
present in the 40 position of the catechol moiety did not show
any appreciable reducing activity, when compared with 7. The RP
detected for compounds 5 and 6 was comparable to that observed
for the DPPH assay, compound 5 was more active than compound 6
(306 and 80 AAE, respectively).
The inhibitory effect of phenylethanoid glycosides 1 and 2 was
at least one order of magnitude lower than that of flavonoids 5, 6, 7
and 8 (Table 3). In general, for the phenylethanoid derivatives it
was observed that the XO inhibitory effect is governed by the pres-
ence of free ortho-dihydroxyl groups of caffeic acid which bind,
through hydrogen bonds, a specific amino acid residue of the
molybdopterine domain of the enzyme (Chang et al., 2007). Fur-
thermore, the phenylethanoid portion of these metabolites exerts
a hydrophobic interaction with the molybdopterine domain, thus
improving the inhibitory effect on XO activity (Chang et al.,
2007). The higher XOI activity of compound 1 (IC50 345.3
lM) with
respect to compound 2 (IC50 690.5 M) is probably related to the
l
different spatial arrangement of the oligosaccharide chain of com-
pound 1, which leads to a more efficient interaction with the XO
enzyme. Further studies regarding this observation should be car-
ried out.
3.2.4. Antioxidant activity in a linoleic acid system
The antioxidant activity of the isolated compounds was also
evaluated by assessing their ability to delay linoleic acid peroxida-
tion using the thiocyanate method (Zarghami & Khosrowbeygi,
2004). The iridoids 3 and 4 were characterised by the lowest IC50
3.2.3. Xanthine oxidase inhibitory effect
values (15.0 and 18.6 lM, respectively, Table 3) thus indicating
An important source of free radicals is the production, catalysed
by XO, of uric acid from xanthine (Fridovich, 1970). However, it has
been found that many natural compounds can inhibit XO, by
affecting its reaction kinetics and thus lowering the levels of free
radicals produced. In order to evaluate this potential antioxidant
activity, the rate of uric acid formation catalysed by XO was mea-
sured in the absence or in the presence of the extracts or the iso-
lated compounds obtained from T. polium. The most active
compounds were those belonging to the luteolin group (Table 3)
and compounds 7 and 8 displayed an inhibitory effect two orders
of magnitude higher than those observed for the phenylethanoid
an extraordinary ability to inhibit lipid peroxidation. This biologi-
cal activity has also been reported for iridoids from other source
(Pacifico et al., 2009). Regarding the structure–activity relation-
ship, the presence of a free hydroxyl group on C-5 of the iridoid
3 could be responsible for its higher antioxidant ability, with re-
spect to what was observed for compound 4 (Pacifico et al.,
2009). However, it should be noted that the process of linoleic acid
peroxidation inhibition could be due to the effect of more and/or
combined antioxidant mechanisms (DeGraft-Johnson et al.,
2007). Indeed, flavonoids are characterised by both radical-scav-
enging and iron-chelating activity. In addition, due to their molec-
ular structure, these compounds could potentially have
antioxidant activity in both hydrophilic and lipophilic media;
therefore, the different medium-solubility of compounds 5, 6, 7
and 8 should explain the different extent of their anti-lipid perox-
idation activity.
molecules (10.7 and 6.7 lM, respectively). The iridoids did not
show any significant effect. From kinetic assays aimed at exploring
the activity of XO in the absence or in the presence of different con-
centrations of selected inhibitor, it was observed that compound 8
acted through a competitive mechanism, as shown by the Linewe-
aver–Burk double-reciprocal plot reported in Fig. 3.
The enzyme XO, in its active form, is a homodimer consisting of
two subunits of 290 kDa. Each of these subunits contains two [2Fe–
2S] centres, one FAD cofactor and one molybdopterine cofactor. It
appears that the XO inhibition by phenylpropanoids derives from
their interaction with the molybdopterine cofactor and with spe-
cific amino acid residues present in the catalytic site of XO (Le-
spade & Bercion, 2010).
The activity of phenylethanoids
1 and 2 (IC50 16.6 and
107.1 M, respectively) followed the same trend observed in other
l
tests, where the activity of compound 1 was always higher than
that of compound 2.
4. Conclusions
In a previous study, the structure–activity relationship of flavo-
noids as inhibitors of XO was investigated and it was demonstrated
that these secondary metabolites are structural analogues of natu-
ral substrates of XO and that most of them act as competitive
inhibitors. In particular, it was shown that their inhibiting activity
was due to the planar flavonoid structure, the presence of a C2–C3
double bond and free hydroxyl groups at C-5 and C7 (Lespade &
Bercion, 2010; da Silva et al., 2004). T. polium flavonoids 5, 6 and
7 showed a lower XOI activity when compared with 8. Indeed,
The n-butanol extract of the aerial part of Teucrium polium L.
showed a high antioxidant activity, due to the presence of bioac-
tive phytochemicals, such as flavonoids, iridoids and phenyletha-
noids. These components could have a synergistic effect through
different antioxidant mechanisms, such as free radical scavenging,
binding metal catalysts, decomposing peroxides and breaking
chain reactions.
In order to explore the potential applications of these biocom-
pounds isolated from T. polium, we would like to assess if they

28
S. De Marino et al. / Food Chemistry 133 (2012) 21–28
Georgiev, M., Alipieva, K., Pashova, S., Denev, P., Angelova, M., Kerns, G., et al. (2010).
Antioxidant activity of devil’s claw cell biomass and its active constituents. Food
Chemistry, 121, 967–972.
Hannedouche, S., Jacquemond-Collet, I., Fabre, N., Stanislas, E., & Moulis, C. (1999).
Iridoid keto-glycosides from Caryopteris clandonensis. Phytochemistry, 51,
767–769.
Hara, S., Okabe, H., & Mihashi, K. (1987). Gas-liquid chromatographic separation of
aldose enantiomers as trimethylsilyl ethers of methyl 2-(polyhydroxyalkyl)-
thiazolidine-4-(R) carboxylates. Chemical and Pharmaceutical Bulletin, 35,
501–506.
can exert antioxidant protection even in cell systems, in order to
verify if the addition of these metabolites in the growth medium
can positively modulate the redox state of cells exposed to differ-
ent oxidative stress generating agents. Therefore their possible
use as natural antioxidant additives in various food products will
be also considered in place of synthetic antioxidants.
Acknowledgements
Heim, K. E., Tagliaferro, A. R.,
& Bobiya, D. J. (2002). Flavonoid antioxidants:
chemistry, metabolism and structure–activity relationships. Journal of
Nutritional Biochemistry, 13, 572–584.
Kellam, S. J., Mitchell, K. A., Blunt, J. W., Munro, M. H. G., & Walker, J. R. L. (1993).
Luteolin and 6-hydroxyluteolin glycosides from Hebe stricta. Phytochemistry, 33,
867–869.
Kupchan, S. M., Britton, R. W., Ziegler, M. F., & Siegel, C. W. (1973). Bruceantin, a new
potent antileukemic simaroubolide from Brucea antidysenterica. Journal of
Organic Chemistry, 38, 178–179.
Lespade, L., & Bercion, S. (2010). Theoretical study of the mechanism of inhibition of
xanthine oxidase by flavonoids and gallic acid derivatives. Journal of Physical
Chemistry B, 114, 921–928.
MS and NMR spectra were provided by Centro di Servizio Inter-
dipartimentale di Analisi Strumentale (CSIAS), Università di Napoli
‘‘Federico II’’, Napoli, Italy. We are grateful for the financial assis-
tance provided by Osservatorio Ambientale Permanente della Bio-
diversità (Convenzione - Colli al Volturno) and by grants from
MIUR (Prin 2009, Rome) awarded to G. Raimo.
Malakov, P. Y., Boneva, I., Papanov, G., & Spasov, S. (1988). Teulamifin B, a neo-
clerodane diterpenoid from Teucrium lamiifolium and T. polium. Phytochemistry,
27, 1141–1143.
Marcocci, L., Suzuki, Y. J., Tsuchiya, M., & Packer, L. (1994). Antioxidant activity of
nitecapone and its analog OR-1246: effect of structural modification on
antioxidant action. Methods in Enzymology, 234, 525–541.
Pacifico, S., D’Abrosca, B., Pascarella, M. T., Letizia, M., Uzzo, P., Piscopo, V., et al.
(2009). Antioxidant efficacy of iridoid and phenylethanoid glycosides from the
medicinal plant Teucrium chamaedris in cell-free systems. Bioorganic and
Medicinal Chemistry, 17, 6173–6179.
References
Amic´, D., Davidovic´-Amic´, D., Beslo, D., Rastija, V., Lucic´, B., & Trinajstic´, N. (2007).
SAR and QSAR of the antioxidant activity of flavonoids. Current Medicinal
Chemistry, 14, 827–845.
Andary, C., Wylde, R., Heitz, A., Rascol, J. P., Roussel, J. L., & Laffite, C. (1985).
Poliumoside, a caffeic glycoside ester from Teucrium belion. Phytochemistry, 24,
362–364.
Bruno, M., Bondi, M. L., Rosselli, S., Piozzi, F., Al-Hillo, M. R. Y., Lamara, K., et al.
(2000). Neoclerodane diterpenoids from Teucrium maghrebinum. Journal of
Natural Products, 63, 1029–1031.
Chang, Y. C., Lee, F. W., Chen, C. S., Huang, S. T., Tsai, S. H., Huang, S. H., et al. (2007).
Structure–activity relationship of C3–C6 phenylpropanoids on xanthine
oxidaseinhibiting and free radical-scavengig activities. Free Radical Biology and
Medicine, 43, 1541–1551.
da Silva, S. L., da Silva, A., Honorio, K. M., Marangoni, S., Toyama, M. H., & da Silva, A.
B. F. (2004). The influence of electronic, steric and hydrophobic properties of
flavonoid compounds in the inhibition of the xanthine oxidase. Journal of
Molecular Structure (Theochem), 684, 1–7.
DeGraft-Johnson, J., Koloddziejczyk, K., Krol, M., Krol, B., Nowak, P., & Nowak, D.
(2007). Ferric-Reducing ability power of selected plant poliphenols and their
metabolites: implications for clinical studies on the antioxidant effects of fruits
and vegetable consumption. Basic and Clinical Pharmacology and Toxicology, 100,
345–352.
Rizk, A. M., Hammouda, F. M., Rimpler, H.,
& Kamel, A. (1986). Iridoids and
flavonoids of Teucrium polium herb. Planta Medica, 2, 87–88.
Ruhdorfer, J., & Rimpler, H. (1981). Teucardoside, a new iridoid from Teucrium
arduini L. and Teucrium hircanicum L.. Tetrahedron Letters, 22, 839–842.
Serrilli, A. M., Ramunno, A., Rullo, R., Ballero, M., Serafini, M., & Bianco, A. (2007). The
occurrence of phenyl propanoid glycosides in endemic Teucrium species.
Natural Product Research, 21, 814–818.
Sharififar, F., Dehghn-Nudeh, G., & Mirtajaldini, M. (2009). Major flavonoids with
antioxidant activity from Teucrium polium L.. Food Chemistry, 112, 885–888.
Ulubelen, A., Topcu, G., & Sonmez, U. (2000). Chemical and biological evaluation of
genus Teucrium. Studies in Natural Products Chemistry (Bioactive natural
Products, Part D) (vol. 23, pp. 591–648).
Wang, M., Simon, J. E., Aviles, I. F., He, K., Zheng, Q. Y., & Tadmor, Y. (2003). Analysis
of antioxidative phenolic compounds in artichoke (Cynara scolymus L.). Journal
of Agricultural and Food Chemistry, 51, 601–608.
Williams, C. A., & Harborne, J. B. (1975). Luteolin and daphnetin derivatives in the
Juncaceae and their systematic significance. Biochemical Systematics and
Ecology, 3, 181–190.
Feresin, G. E., Tapia, A., Angel, G. R., Delporte, C., Erazo, N. B.,
& Schmeda-
Hirschmann, G. (2002). Free radical scavengers, anti-inflammatory and
analgesic activity of Acaena magellanica. Journal of Pharmacy and
Pharmacology, 54, 835–844.
Zarghami, N., & Khosrowbeygi, A. (2004). Evaluation of lipid peroxidation as an
indirect measure of oxidative stress in seminal plasma. Iranian Journal of
Reproductive Medicine, 2, 34–39.
Zhu, K. X., Lian, C. X., Guo, X. N., Peng, W., & Zhou, H. M. (2011). Antioxidant
activities and total phenolic contents of various extracts from defatted wheat
germ. Food Chemistry, 126, 1122–1126.
Fridovich, I. (1970). Quantitative aspects of the production of superoxide anion
radical by milk xanthine oxidase. Journal of Biological Chemistry, 245,
4053–4057.
Gattuso, G., Barreca, D., Caristi, C., Gargiulli, C., & Lezzi, U. (2007). Distribution of
flavonoids and furocoumarins in juice from cultivars of Citrus bergamia Risso.
Journal of Agricultural Food Chemistry, 55, 9921–9927.