UV-ABC screens of luteolin derivatives compared to edelweiss extract

Article abstract of DOI:10.1016/j.jphotobiol.2011.01.005

Pure luteolin is a remarkably heat (200°C/6 days) and UV stable UV-A screen, however, native luteolin enriched to 37% in an edelweiss extract lost its UV-A screen properties upon UV irradiation (～4 MJ m^-2). This contrasting behavior led to the

Full text of DOI:10.1016/j.jphotobiol.2011.01.005

Journal of Photochemistry and Photobiology B: Biology 103 (2011) 8–15
Contents lists available at ScienceDirect
Journal of Photochemistry and Photobiology B: Biology
journal homepage: www.elsevier.com/locate/jphotobiol
UV-ABC screens of luteolin derivatives compared to edelweiss extract
a
b
a
a
Fabian Fischer a, , Evelyne Zufferey , Jean-Marc Bourgeois , Julien Héritier , Fabrice Micaux
⇑
a Institute of Life Technologies, HES-SO Valais, University of Applied Sciences Western Switzerland, Route du Rawyl 64, 1950 Sion, Switzerland
b Chemistry Department, HES-SO Fribourg, University of Applied Sciences Western Switzerland, Bd de Pérolles 80, CH-1705 Fribourg, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 20 July 2010
Received in revised form 3 November 2010
Accepted 10 January 2011
Available online 19 January 2011
Pure luteolin is a remarkably heat (200 °C/6 days) and UV stable UV-A screen, however, native luteolin
enriched to 37% in an edelweiss extract lost its UV-A screen properties upon UV irradiation (ꢀ4 MJ mꢁ2).
This contrasting behavior led to the examination of a series of purified luteolin derivatives as UV screen
candidates. 30,40,5,7-Tetralipoyloxyflavones were synthesized from luteolin (30,40,5,7-tetrahydroxyflavone)
and fatty acid chlorides. These acylated semi-biomolecules show a hypsochromic shift in UV–Vis spectra
of about
DkA?B = 58 nm and absorbed in the centre of the harmful UV-B band (kmax = 295 nm). Luteolin
Keywords:
Leontopodium alpinum
Edelweiss
Luteolin
UV-ABC absorption
Sunscreen
was also hydroxyethylated with Br(CH2)2OH. This substitution has no effect on the kmax = 330 nm absorp-
tion of luteolin (UV-A band). Finally the natural 40-O-b-glucosyl-30,5,7-trihydroxyflavone was extracted
from edelweiss and used as a purified natural benchmark. Glycosylated and hydroxyethylated luteolin
are both UV stable. Fully acylated luteolin derivatives degrade upon UV exposure to a stable UV-C screen
with a hypsochroic shift
DkB?C = 35 nm. All in all, three molecular structures based on luteolin with sun-
Acylation
screen properties were found, distinguishable in: UV-A, UV-B, and UV-C filters. The natural product based
UV-absorbers show promise as alternatives to synthetic molecules and nanoparticles in sunscreen
products.
Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction
cence leaves and pigments in these structures [15,16]. The UV-
active principles responsible for the sunscreen properties in this al-
The development of natural sun screens from edelweiss
(Leontopodium alpinum) is based on the inherent UV absorber prop-
erties of this plant. Natural sunscreens from plant extracts are
investigated because synthetic sunscreens lose their acceptance
with consumers. A prominent discussion deals with the probable
toxicity of nanoparticles in sunscreen lotions such as the often
used titanium dioxide [1]. The development of natural sunscreen
products pursued in many research groups [2,3] is therefore of
considerable interest.
Plants protect themselves in particular from UV-B radiation,
while the more energetic UV-C is absorbed by O2 (k < 240 nm)
and O3 (k < 310 nm) in the upper atmosphere [4,5]. Like plants,
the human skin needs to be protected from UV-A and UV-B radia-
tion to avoid sun burn, skin ageing and cancer [6]. Flavonoids are
well known to absorb UV-light in various plants [7,8]. Edelweiss
(L. alpinum Cass.) contains flavones with UV-absorption and anti-
radical properties, which are used in cosmetics and traditional
medical care [9,10]. This protected plant grows in middle and
southern Europe, in the Alps and is nowadays cultivated for com-
mercial use [11–14]. The UV protection in the edelweiss appears
to be a combination of absorbing nanostructures on the inflores-
pine plant are still not well identified. Studies with various plants
show that there is an altitudinal effect, for example in maize,
where flavone contents increases upon exposure to UV-B in higher
altitudes [17]. Another altitudinal study of Arnica montana L. shows
a positive correlation of vicinal hydroxyl groups in the B-ring of
flavonoids with higher altitudes [18]. Luteolin 1 contains such vic-
inal hydroxyl groups and in controlled experiments with Marchantia
polymorpha L. the absence of UV-B decreased the luteolin 1 content
whereas an increase of UV-B radiation above natural levels had a
slightly positive effect on its content [19]. Luteolin is in fact found
in various concentrations in the edelweiss inflorescence leaves
isolated as luteolin-7-O-b-D-glucopyranoside, luteolin-30-O-b-D-
glucopyranoside, luteolin-40-O-b-D-glucopyranoside, but most of
the time as the aglycon [13,20]. Luteolin 1 is also known to be an anti-
oxidant; it has been studied by scavenging experiments with 1,1-di-
phenyl-2-picrylhydrazyle (DPPH), superoxide dismutase (SOD) and
inhibited lipid peroxidation [21]. Because of its antioxidant properties
luteolin 1 eliminates radicals and prevents cell damage, which is an
additional valuable property for a sunscreen product. The 30,40,5,7-
tetrahydroxyflavone 1 is also known to be an antimutagen for the
dietary carcinogen Trp-P-2 and has been found to induce cell cycle ar-
rest and apoptosis in colon and prostate cancer cells [22–24]. In addi-
tion antiinflammatory properties were discovered in animal models
[25].
⇑
Corresponding author. Tel.: +41 (0) 27 606 86 58; fax: +41 (0) 27 606 86 15.
E-mail address: Fabian.Fischer@hevs.ch (F. Fischer).
1011-1344/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotobiol.2011.01.005

F. Fischer et al. / Journal of Photochemistry and Photobiology B: Biology 103 (2011) 8–15
9
In order to study luteolin’s UV absorber properties a series of
luteolin derivatives were synthesized and isolated from edelweiss
for the present study.
its UV–Vis properties and for tetracylated luteolin (5–7) MeCl2
was also employed. The mixtures were filled into UV quartz screw
cap cuvettes and exposed horizontally to the light of a UV lamp at
room temperature. The UV254 16.9 W bulb provided an irradiation
power of 1.0 MJ mꢁ2 dayꢁ1. UV–Vis spectra (200–850 nm) were re-
corded frequently during the experiments with an Agilent 8454
UV–Vis Spectrometer.
Acetylated flavonoids are in general divided into two groups: the
regioselective esterified glucopyranoside substituents and acety-
lated hydroxyl groups of luteolin 1. The acylation of the pyranosyl
residues in O-60 position mediated by lipases has been thoroughly
investigated [26]. The esterification on the ABC-ring of flavones
however has been studied less (Scheme 3). Natoli et al. [27] synthe-
sized polyacetylated flavones with vinyl acetate and then hydro-
lyzed them selectively employing lipases to isolate partly
acetylated products. Luteolin tetraacetates have also been produced
by Ac2O in pyridine [28]. 3-O-Acetylation of quercetin was realized
with several fatty acids and the product was examined as UV protec-
tive agent [29]. The absorption maxima of these mono-esterified
compounds remained largely in the UV-C region. In this work, lute-
2.1.2. Heat stability of luteolin derivatives
The heat stability of derivatives was tested with solutions as
prepared for UV–Vis analysis (Section 2.1.1). The mixtures were
incubated at 40 °C in the dark in screw cap UV-cuvettes and spec-
tra were recorded as above.
2.1.3. Heat stability of pure luteolin
The heat stability of pure luteolin 1 (aglycon) was verified by
preparing a dry sample on a watch glass and placed it in a pre-
heated oven (80–200 °C/1 atm). UV–Vis spectra were recorded
daily (6 days) and finally a 1H NMR spectrum was recorded to ver-
ify the integrity of the sample.
olin
1 was also hydroxyethylated employing Br(CH2)2OH 8
(Scheme 3) and the UV-absorption maxima were examined
(Fig. 1). Luteolin derivatives of this kind were synthesized from tri-
hydroxyrutin generating polyhydroxyalkylluteolin [30].
Luteolin 1 is a particularly useful candidate for sunscreens be-
cause of its absorption properties in the UV-A range
(kmax = 333 nm, Fig. 1A) [31]. Furthermore glycosylated 10 and
hydroxyethylated 9 luteolin show roughly the same absorption
maxima as the aglycon 1 (Fig. 1B). A shift of the absorption maxima
was found when all hydroxyl groups in luteolin 1 were acylated
with fatty acids, an ideal absorption in the UV-B band,
kmax ꢀ 295 nm, was observed (Fig. 1C). This absorption maximum
is close to the centre of the UV-B radiation (315 –280 nm) and rep-
2.2. Structure analysis of derivatives
Luteolin derivatives were analyzed by 1H NMR and 13C NMR on
a Bruker 400 MHz spectrometer. Mass Spectroscopy data were ob-
tained with a Hewlett Packard (Series 1100 MSD) and from an LC/
MS setup with an Agilent 1100 Series hyphenated to a Hewlett
Packard Series 1100 MSD.
resents a hypsochromic shift of about
DkA?B = 58 nm found with
30,40,5,7-tetralipoyloxyflavones 5–7. When in the latter derivatives
the acyl group in C(5) position was absent, an ideal UV-C screen
(kmax = 260 nm) representing and additional hypsochromic shift
2.3. Derivatizations
2.3.1. General procedure for 30,40,5,7-tetralipoyloxyflavones 5–7
One hundred milligram (0.35 mmol) luteolin 1 was added to
5 ml pyridine and an excess (2 mmol) fatty acid chloride 2–4 was
slowly added at 0 °C. After stirring, for 1 h at 0 °C and 24 h at room
temperature, the reaction mixture was diluted with 20 ml CH2Cl2
and washed with 20 ml H2O. The CH2Cl2 phase was acidified with
1 N HCl permitting pyridine extraction. Then 10 ml saturated NaH-
CO3 was added to the organic phase to extract the acids and finally
washed with 20 ml H2O. The obtained crude product 5–7 was final-
ly purified by chromatography on silica gel (60 Å) with an ether/
petrol ether (1:1) mixture.
D
kB?C = 35 nm was discovered.
In short, three distinct UV-A, UV-B, and UV-C screens based on
luteolin were found (Scheme 1).
2. Material and methods
2.1. Material
A luteolin 1 (aglycon) extract from dried L. alpinum Cass. (edel-
weiss) [Valplant] was obtained from the plant process laboratory
at HES-SO Valais. A second quantity of glycosylated luteolin 10
was isolated during the project. Luteolin 1 standard was purchased
from A&Z Food Additives Co. Solvents such as acetonitrile, acetone,
pyridine [Lab-Scan] were HPLC grade; DMF [Fluka] was distilled
before use. For ethoxylations 2-bromoethanol 8 [Fluka] and for
acylations dodecanoyl chloride 2, tetradecanoyl chloride 3 and
hexadecanoyl chloride 4 were bought from Fluka.
2.3.1.1. 30,40,5,7-Tetralauroyloxyflavone 5. 1H-RMN (CDCl3): 7.73
(dd, J = 8.22 Hz, J = 2.21 Hz, 1H), 7.69 (d, J = 2.21 Hz, 1H), 7.35(d,
J = 2.06 Hz), 7.34 (d, J = 8.22 Hz, 1H), 6.83 (d, J = 2.06 Hz, 1H), 6.62
(s, 1H), 2.28–2.76 (m, 8H), 1.27 (bm, 64H), 1.6–1.85 (m, 8H), 0.89
(t, 12H) ppm.
13C – RMN (DMSO-d6) d : 176.23, 160.81, 157.59, 154.05,
150.35, 145.01, 142.85, 129.75, 124.32, 124.29, 121.62, 114.95,
113.87, 108.92, 108.87 ppm. Lauroyloxy-substituents: 172.11,
170.90, 170.79, 170.61, 34.42, 34.22, 34.12, 34.05, 29.62, 29.59,
29.49, 29.43, 29.34, 29.32, 29.29, 29.25, 29.23, 29.21, 29.19,
29.14, 29.08, 29.05, 31.9, 24.97, 24.90, 24.88, 24.73, 22.67
2.1.1. UV stability of luteolin derivatives
Luteolin derivatives were dissolved as 65–90 lM solutions.
Most of the time, acetonitrile was used as the solvent because of
OH
OR
OR
OR
OR
HO
O
O
RO
O
RO
O
O
OH
OH
O
OR
OH
UV-A
UV-B
UV-C
Scheme 1. Three distinct sunscreens based on luteolin 1. UV-Amax, absorption is characteristic for the aglycon luteolin 1 including ethoxylated 9 and glycosylated 10
derivatives. UV-Bmax absorption is strong for fully acylated luteolin 5–7 and triacylated 5b–7b provides a selective UV-Cmax absorption.

10
F. Fischer et al. / Journal of Photochemistry and Photobiology B: Biology 103 (2011) 8–15
B
A
H
H
H
2.5
H
OGlu
OH
2.5
2
H
OH
OH
H
H
O
HO
H
O
O
HO
H
O
O
2
1.5
1
H
H
H
OH
1.5
1
H
0.5
0.5
0
0
220
220
270
320
370
420
270
320
370
420
C
D
H
H
3
3
H
OCOR
OCOR
H
OCOR
OCOR
H
H
2.5
2
ROCO
O
O
O
ROCO
H
2.5
2
H
H
H
H
H
O
O
ROCO
H
1.5
1
1.5
1
0.5
0.5
0
0
220
270
320
370
420
220
270
320
370
420
[nm]
Fig. 1. UV–Vis spectra of luteolin 1 and derivatives thereof show distinct UV-Amax, UV-Bmax, and UV-Cmax absorptions. 30,40,5,7-tetrahydroxyflavone 1 (A) and of luteolin-40-O-
b-D-glucopyranoside 10 (B) are UV-A screens, kmax ꢀ 330 nm. The four acyl groups in 30,40,5,7-tetralauroyloxyflavone 5 (C) provoke a hypsochromic shift yielding a UV-B
screen, kmax ꢀ 295 nm. The absence of an acyl group in 5–7 at C(5) position leads to 5-hydroxy-30,40,7-tripalmoyloxyflavone 7a–c (D) causing a further hypsochromic shift,
resulting in a UV-C screen, kmax ꢀ250 nm.
14.10 ppm. APCI-MS m/z 1016.8 (M+), UV–Vis kmax: 225, 257,
2.3.2. Procedure for 30,40,5,7-tetrahydroxyethoxyflavone 9
291 nm.
Two hundred and twenty milligram (1.76 mmol) 2-bromethanol 8
was added to a solution of 100 mg (0.35 mmol) luteolin 1 in DMF.
100 mg (1.78 mmol) KOH as pellets were added and stirred at
80 °C for 3 h. DMF was removed by distillation under reduced pres-
sure and the residue extracted with 20 ml AcOEt and 20 ml H2O. A
product mixture was isolated with mainly mono- and di-ethoxy-
lated luteolin 9 and traces of tri- and tetraacylated products.
(DMSO-d6): d: 6.16–6.22 (d, J = 2.06 Hz), 6.45–6.53 (d,
J = 2.06 Hz), 6.63–6.79 (s), 6.88–7.12 (d, J = 8.22 Hz), 7.35–7.63 (d,
J = 2.21 Hz and dd, J = 8.22 Hz, J = 2.21 Hz), APCI-MS m/z 329(100)
M+ 373(48) M+ LC/MS: Retention time/2503000 329(100) M+,
373 M+(%); 2403000 419 M+ 463 M+; UV–Vis kmax : 268, 334 nm.
2.3.1.2. 30,40,5,7-Tetramyristoyloxyflavone 6. 1H-RMN (CDCl3): 7.73
(dd, J = 8.22 Hz, J = 2.21 Hz, 1H), 7.69 (d, J = 2.21 Hz, 1H), 7.35(d,
J = 2.06 Hz), 7.34 (d, J = 8.22 Hz, 1H), 6.83 (d, J = 2.06 Hz, 1H), 6.62
(s, 1H), 2.28–2.76 (m, 8H), 1.6–1.85 (m, 8H), 1.27 (bs, ꢀ80H),
0.89 (t, 12H) ppm.
13C – RMN (DMSO-d6) d : 176.14, 160.74, 157.57, 154.17, 150.4,
144.99, 142.78, 129.55, 124.36, 124.27, 121.59, 114.95, 113.84,
108.92, 108.86 ppm. Myristoyloxy-substituents: 172.06, 170.87,
170.76, 170.57 34.40, 34.21, 34.11, 34.04, 31.92, 29.64, 29.62,
29.60, 29.49, 29.44, 29.42, 29.35, 29.32, 29.29, 29.25, 29.23,
29.21, 29.19, 29.16, 29.14, 29.05, 24.96, 24.90, 24.88, 24.81,
22.68, 14.09 ppm. APCI-MS m/z 1128.8 (M+), UV–Vis kmax : 226,
260, 294 nm.
2.4. Isolation 40-O-b-glucopyranosyl-30,5,7-trihydroxyflavone 10
Dry inflorescence leaves of edelweiss were macerated for 16 h
at 40 °C with slow stirring in a solution of water/EtOH/BuOH that
was tuned to be monophasic, filtrated (AF31H), the filter cake
was rinsed with a mixture of water/EtOH/BuOH. From the collected
aqueous liquids ethanol was evaporated and the H2O/BuOH mix-
ture became biphasic. The aqueous phase was concentrated and
fractionated by MPLC using MeOH/HCOOH (99.5:0.5) as eluent.
The MPLC consisted of a HPP-VPC column connected to a High
Performance Pump SEPAPRESS HPP-40/200 with gradient control
unit SEPACON GCU-311 [Kronwald].
2.3.1.3. 30,40,5,7-Tetrapalmitoyloxyflavone 7. 1H-RMN (CDCl3): 7.73
(dd, J = 8.22 Hz, J = 2.21 Hz, 1H), 7.69 (d, J = 2.21 Hz, 1H), 7.35(d,
J = 2.06 Hz), 7.34 (d, J = 8.22 Hz, 1H), 6.83 (d, J = 2.06 Hz, 1H), 6.62
(s, 1H), 2.28–2.76 (m, 8H), 1.27 (bs,ꢀ88H), 1.6–1.85 (m, 8H), 0.89
(t, 12H) ppm.
13C – RMN (DMSO-d6) d : 176.16, 160.75, 157.57, 154.17,
150.40, 144.99, 142.78, 129.54, 124.29, 124.27, 121.59, 114.94,
113.83, 108.91, 108.86 ppm. Palmytoyloxy substituents: 172.07,
170.86, 170.75, 170.56, 34.40, 34.21, 34.10, 34.04, 31.92, 29.70,
29.67, 29.64, 29.59, 29.50, 29.48, 29.45, 29.43, 29.38, 29.36,
29.35, 29.32, 29.29, 29.26, 29.23, 29.21, 29.19, 29.16, 29.05,
24.99, 24.96, 24.89, 24.88, 22.68, 14.09 ppm. APCI-MS m/z 1240.2
(M+).
For HPLC analyses an Agilent 1100 Series equipped with an UV
detector [Agilent] and a CC 155/3 Nucleosil 100-5 C18 column
[Macherey–Nagel] was used.
1H-RMN (DMSO-d6) 6.60(s, 1H), 6.20(d, J = 1.95, 1H), 6.44(d,
J = 1.83, 1H), 7.44(s, 1H), 7.31(d, J = 8.41, 1H), 7.44(d, J = 8.40, 1H),

F. Fischer et al. / Journal of Photochemistry and Photobiology B: Biology 103 (2011) 8–15
11
pyranosyl: 4.94(d, J = 12.8, 1H), 3.93(d, J = 10.1, 1H), 3.74(dd,
J = 12.1, J = 5.1, 1H), 3.64(m, 1H), 3.53(m, 2H), 3.44(m, 1H), ppm.
13C NMR (100 MHz) (DMSO-d6) 182.43, 164.74, 164.08, 161.92,
158.01, 148.59, 147.24, 125.82, 118.42, 116.59, 113.47, 104.03,
103.66, 101.82, 98.83, 93.67, 77.07, 76.14, 73.38, 69.9, 61.03 ppm.
APCI-MS: m/z 447.1 M+. UV–Vis kmax: 269, 328 nm.
during extraction. The extract contained in addition various gly-
cans, lipids and other non-identified compounds. The luteolin con-
taining mixture degraded under biological effective UV254
irradiation [32]. This irradiation (5 MJ mꢁ2) resulted in a complete
transformation of the UV-Amax into a shoulder (Fig. 2B). The ob-
served UV-resistance is higher than that of some commercial prod-
ucts that degrade upon 400 kJ mꢁ2 irradiation [33]. Heating the
extract (at 40 °C/1 day) the UV-A absorber property (kmax
=
3. Results and discussion
327 nm) equally disappears. The instability of enriched luteolin 1
in the edelweiss extract is in sharp contrast to the elevated stabil-
ity of pure luteolin 1 (Fig. 2A) (Section 3.1.2).
3.1. UV properties
3.1.1. UV-absorption properties of luteolin and derivatives
3.1.4. Stability of luteolin derivatives
The UV-absorption properties can be distinguished into substit-
The UV stability of luteolin derivatives is found to be lower (5–7)
and higher (5b–7b, 9, 10) than the luteolin 1 containing edelweiss
extract. Glycosylated luteolin 10 is a natural derivative and was ob-
tained from edelweissplant material (L. alpinum). The UVmax absorp-
tions of this natural compound 10 remained unchanged at
kmax = 328 and 269 nm with UV254 irradiation (7 MJ mꢁ2) (Fig. 2C).
This is in contrast to the crude edelweiss extract, which lost the
UV-A screen property after ꢀ5 MJ mꢁ2 of UV irradiation (Fig. 2B).
The hydroxyethoxylated luteolin 9 is, in comparison to 10, a
synthetic derivative that provides similar UV-A stability as glycos-
ylated luteolin (Fig. 2C, D). This was expected as both derivatives (9
and 10) belong to the sigma bond type derivatives, whose glyco-
side respectively ether bonds are not part of the photon absorbing
uents, which alter
ers, that are linked only by
on the -electrons. The UV-absorption spectra of the aglycon lute-
p-conjugation in the ABC-ring of 1, and by oth-
r
-bonds inducing less or no influence
p
olin 1 showed a maximum in the UV-A (kmax = 333 nm) and a sec-
ond one in the UV-C range (kmax = 268 nm) with a minimum in the
UV-B zone, kmin = 284 nm (Fig. 1A). The UV-absorption of natural
40-O-b-glucopyranosyl-30,5,7-trihydroxyflavone 10 is very similar
(UV-A kmax = 334 nm) (Fig. 1B), which is in line with earlier results
for the 7-O-glycoside [31]. Neither the mono- nor the dihydroxye-
thylated luteolin (9) changed the absorption spectra (UV-A,
kmax = 334 nm). This is due to the fact that the inductive effects
of alkyl and acetal substituents have no influence through the
bond linkage, it being too weak to influence the -ABC-ring system
of luteolin 1. In comparison, acylated luteolin (5–7) shows a differ-
ent absorption spectrum because -electrons of the added car-
r-
p
p-system.
30,40,5,7-Tetralipoyloxyflavones 5–7 behave very differently, as
p
they are
p-conjugated derivatives. It is noteworthy that these are
bonyl function are conjugated with the aromatic ABC-ring
system of the flavone 1. The 30,40,5,7-tetralipoyloxyflavones 5–7 in-
duce a hypsochromic shift of the maxima in the UV-A section
molecular structures not found in edelweiss nor in any other plant.
Nevertheless, they represent semi-biological derivatives 5–7, a
combination of biomolecules that degrade according to initial
experiments along known metabolic pathways. This is illustrated
by the selective decomposition upon UV irradiation of 1.5 MJ mꢁ2
(Fig. 2E). In this experiment, the 30,40,5,7-tetralipoyloxyflavones
5–7 lose regio-selectively one of the acylated fatty acids at C(5)
yielding 5-hydroxy-30,40,7-trilipoyloxyflavones 5b–7b (Scheme 2).
This site specific loss is facilitated by steric encumbrance caused
by the vicinal carbonyl at C(4) (Fig. 3). The instability of this substi-
tuent was equally noticed during acylation in HOAC(5) position,
which were not always successful, in particular with longer chain
fatty acid chlorides 3,4 (see Section 3.2.1). The resulting 5-hydro-
xy-30,40,7-trilipoyloxyflavones 5b–7b were also exposed to UV light
and no further substituent loss was observed (7 MJ mꢁ2) (Fig. 2F).
Purified luteolin derivatives appear in part to be photo-stable
(9,10) whereas the fully acylated derivatives (5b–7b) lose their ini-
tial activity partly.
(DkA?B = 58 nm) to the UV-B range with a first maximum at
kmax = 291 (5) respectively at kmax = 294 nm for 6 (Fig. 1C). These
maxima are in the centre of the UV-B region constituting an ideal
absorption property for a sunscreen since UV-B radiation causes
sun burns. They are followed by two additional maxima at shorter
wavelengths around kmax = 260 and kmax = 225 nm (5). In the ab-
sence of the acyl substituent in C(5) there is an additional hypso-
chromic shift
DkB?C = 30 nm with a new absorption maximum at
kmax = 261 nm (5) (Fig. 1D). A mixture of mono and diacylated lute-
olin shows two maxima at k = 315 and 269 nm with a flat mini-
mum in between, indicating that not all acylations induce the
same hypsochromic shift.
The three distinct UV-screens: UV-A, UV-B, and UV-C (Scheme 1,
Fig. 1) based on luteolin 1 represent the most important result of
the present study. Their sole or combined use appears to be useful
for natural sunscreen products.
3.1.5. Luteolin and the classical sunscreens
Most classical sunscreens are combination of UV-A and UV-B
filters, plus physical components, which work through reflection
in order to have broad spectrum protection. The luteolin deriva-
tives could replace some components. A new sunscreen replaces
existing products once it protects against cancer, photodermatoses
(polymorphic light eruption, chronic actinic dermatitis) and photo-
toxic respectively photoallergic reactions. In short sensitizing and
other photochemical effects demand further evaluation. Beside
superior efficiency, a new sunscreen should also be produced in a
sustainable manner. Classical sunscreens are not based on renew-
able resources. Luteolin is natural and fits therefore to the wide
consumer demand for green sun care products.
3.1.2. Stability of non-derivatized luteolin
Pure luteolin 1 is surprisingly heat stable and resists 200 °C for
more than 6 days (Fig. 2A). The week long heating transformed the
pale yellow luteolin into a slightly brown powder. Minor impuri-
ties were detected by 1H NMR spectroscopy, but a closer analysis
by EI mass spectroscopy showed no other compound. Neither does
it degrade under UV254 irradiation in a week long experiment. A
control with free fatty acids as used for luteolin 1 derivatisation
shows no loss to the UV-A absorption in luteolin 1. This stability
is in sharp contrast to a luteolin-enriched edelweiss extract (Sec-
tion 3.1.3), where non-derivatized luteolin 1 degrades upon UV
irradiation (ꢀ5 MJ mꢁ2) (Fig. 2B).
3.2. Synthesis of luteolin derivatives
3.1.3. UV stability of luteolin contained in an edelweiss extract
The edelweiss extract lost its UV-A screen property upon mod-
erate UV irradiation. The extract contained 37% luteolin present as
the aglycon because the glycoside bonds hydrolyzed quantitatively
3.2.1. Acylation of luteolin
The acylation of luteolin 1 was realized with a 1.5 excess of fatty
acid chlorides 2–4 in pyridine at room temperature yielding

12
F. Fischer et al. / Journal of Photochemistry and Photobiology B: Biology 103 (2011) 8–15
A
B
C
D
E
F
Fig. 2. (A) Unsubstituted purified luteolin 1 was heated as a dry powder in the air for 7 days at 200 °C and remained stable. (B) Unsubstituted luteolin enriched (37%) in an
edelweiss extract lost UV-A screen properties kmax = 285 nm upon UV254 irradiation. (C) 40-O-b-glucopyranosyl-30,5,7-trihydroxyflavone 10 isolated from Leontopodium
alpinum, whose UVmax k = 269 and 328 nm remained unchanged with UV irradiation. (D) The glycol substituents on 30,40,5,7 mono- and di-hydroxyethoxyluteolin 9 do not
change the UV-A absorber properties (5 MJ mꢁ2). (E) 30,40,5,7-Tetralauroyloxyflavone 5 lost its UV-B screen properties upon irradiation resulting in an effective UV-C screen,
5-hydroxy-30,40,7-trilauroyloxyflavone 5b. (F) 5-Hydroxy-30,40,7-tripalmoyloxyflavone 7b is a stable UV-C screen.
30,40,5,7-tetralipoyloxyflavones 5–7 (Scheme 3). The raw product
contained fatty acids from quenched fatty acid chlorides 2–4. Dur-
ing the purification CH2Cl2/H2O was used to extract the product 5–
7 into the organic phase. The collected organic phases were then
acidified by 1 N HCl to remove pyridine. The 30,40,5,7-tetralipoyl-
oxyflavones 5–7 were recovered with moderate to good yields
depending on the fatty acid chloride 2–4 used. A quantitative
tetra-acylation was possible with lauroylchloride 2. Acylation with
myristoylchloride 3 yields a product mixture of tetra(6):tri(6b)
acylated luteolin in a 9:1 ratio; palmitoylchloride 4 acylated lute-
olin in a 1:3 ratio, tetra(7):tri(7b).
in 1 is reduced due to steric hindrance caused by the vicinal car-
bonyl function in C(4) position. In addition, the phenolic proton
is retained by the hydroxyl and carbonyl function and resists
deprotonation somewhat longer. The 5-hydroxy-30,40,7-trilauroyl-
oxyflavone 5b is identified by a strong chemical shift change to-
ward higher field for the HAC(6) atom (7.35 ? 6.83 ppm) and to
a lesser extent of the HAC(8) atom (6.83 ? 6.57 ppm) (Fig. 3).
The regioselectivity for the other hydroxyls in 30,40,7 position (1)
appears to be equal; it was studied by reacting one molar equiva-
lent fatty acid chloride 2 with luteolin 1. 1H NMR and MS indicated
several products and therefore the acylation rates on the other hy-
droxyl groups at 30,40, and 7 position are rather close.
The variation in the yields of 5–7 correlates with the fatty acids’
chain length. Lauroyl (2) is a C12 chain, myristoyl (3) a C14 and pal-
mitoyl (4) a longer (C16) chain; the acylation efficiency decreased
for the later by a factor 3. The nucleophilicity of the C(5) hydroxyl
1H NMR spectra of 30,40,5,7-tetralipoyloxyflavones 5–7 were re-
corded on a 400 MHz Spectrometer. The structure elucidation of
30,40,5,7-tetralauroyloxyflavone 5 based on NMR data is discussed

F. Fischer et al. / Journal of Photochemistry and Photobiology B: Biology 103 (2011) 8–15
13
OCOR
OCOR
OCOR
OCOR
ROCO
O
O
ROCO
O
O
hν
HOOC(CH2)10CH3
+
OH
OCOR
6 (B-Screen)
6b (C-Screen)
Scheme 2. UV-B screen transformation. 30,40,5,7-Tetralipoyloxyflavones 6 degraded under UV254 irradiation (1.5 MJ mꢁ2) regio-selectively at C(5) into 5-hydroxy-30,40,7-
trilipoyloxyflavones 6b.
OCO(CH2)nCH3
O(CH2)2OH
O(CH2)2OH
OH
OH
CH3(CH2)nCOCl
H3C(CH2)nOCO
O
Br(CH2)2OH
OCO(CH2)nCH3
HO(CH2)2O
O
HO
O
O
2-4
8
O
OCO(CH2)nCH3
O
OH
O(CH2)2OH
9
1
5-7
n = 10,12 or 14
Scheme 3. Acylation and hydroxyethylation of luteolin 1 by fatty acid chlorides 2–4 and 2-bromoethanol 8.
in the following (Table 1). There are two sets of proton signals: at
low field the aromatic hydrogens of the ABC-ring system and at
higher field the alkyl protons of the fatty acid substituents. The
two aromatic protons at HAC(6) and HAC(8) on the A-ring couple
with a long range doublet (J = 2.06 Hz), in the B-ring HAC(60) cou-
ples to the vicinal HAC(50) (J = 8.22 Hz) and by long range coupling
to HAC(20) (J = 2.21). The C-ring contains a singlet proton HAC(30)
at dH = 6.62 ppm. The chemical shifts in 13C NMR spectra for the
hydrogenated C-atoms were assigned by a HETCOR experiment.
The quaternary C-atoms were determined from calculated values
with a Chem-Draw™ simulation.
C(100–300); for these first C-atoms there is three times a set of four
signals. This pattern may persist for further C-atoms such as
C(400) and C(500) but because the signals are very close together on
the ppm scale (dC = 29.05–29.62 ppm) they cannot be distin-
guished and assigned in detail. The number of carbon signals in
this section C(400–900) is 14, which represents an average of 2.3 sig-
nals per carbon. In view of obtained derivatives (5–7) it is shown
that luteolin derivatives are synthesized in yields that are higher
the smaller the acyl substituent is.
3.2.2. Hydroxyethylation of luteolin
The lauroyl substituents generate three chemical shifts (dH) for
the HAC(2’’), which was verified by HETCOR analysis. There are
two triplets at dH = 2.74 and 2.35 ppm, the multiplet in between
appears as combination of two triplets, which is confirmed by
the 4H-integral. The regio-specific chemical shifts are also ob-
served for the HAC(300) protons represented by two multiplets at
dH = 1.77 and 1.65 ppm. The other signals correspond to the typical
chemical shifts of the lauroyl chain. The 13C NMR of the lauroyl
substituents confirms the regio-dependent shift differences for
Hydroxyethylation of luteolin yields a flavone that was once or
twice modified with a glycol substituent (Scheme 3). In order to
hydroxyethylate the four hydroxyl functions in luteolin 1, a
30,40,5,7-tetrahydroxylated flavone intermediate was prepared by
deprotonation with a biphasic KOH/DMF mixture followed by
nucleophilic substitution on 2-bromoethanol 8 at 80 °C/3 h. Based
on a MS analysis one to two substitutions were realized with no
specific preference for any HOAC(position) according to a ROESY
analysis. Forcing the alkylation of all four hydroxyl functions in
30,40,5,7-tetrahydroxyflavone 1 with an excess of 2-bromoethanol
8 resulted also in the generation of by-products such as oligomeric
polyethyleneglycol. As before mainly mono- and di-hydroxyethylated
products 9 were obtained. An analysis by LC/MS revealed that trace
amounts of tri and tetrahydroxyethylated luteolin were formed.
This result was confirmed by 1H NMR showing many products re-
lated to various chemical shifts of aromatic protons of 1. The inter-
pretation of such spectra was realized under the hypothesis that
there are up to 24 products, which result from substitution varia-
tions on the aglycon 1, without taking into account that diethylen-
glycols and further chain elongations theoretically also can occur.
After all, the most surprising result consists in the almost complete
absence of tetrahydroxylated luteolin.
7.34
7.73
7.34
H
B
6.83
6.57
H
7.76
5'
H
OCOR
6'
1'
4'
3'
8
ROCO
O
9
10
H
7
6
2
3
OCOR
2'
A
C
H
7.73
7.69
H
H
6.86
7.35
5
6.69
6.62
O
O
3.3. Isolation of glycosylated luteolin
5b
The isolation of glycosylated luteolin 10 from plant material of
L. alpinum was realized by a procedure developed for flavones by
[9]. The obtained crude extracts were purified with MPLC yielding
the aglycon 1 and a monoglycosylated luteolin 10. According to a
ROESY experiment the pyranosyl is attached at the C(40) position.
Fig. 3. 1H NMR comparison of UV-B and UV-C screen. Chemical shifts from 1H NMR
(CDCl3) spectra of 5-hydroxy-30,40,7-trilauroyloxyflavone 5b (UV-C) compared to
values of the tretracylated 30,40,5,7-tetralauroyloxyflavone
values of 5 in italics). The presence of a conjugated carbonylester at C(5) shifts in
particular the proton signal of vicinal HAC(6) (6.86 ? 7.35 ppm) to lower field.
5 (UV-B) (reference

14
F. Fischer et al. / Journal of Photochemistry and Photobiology B: Biology 103 (2011) 8–15
Table 1
13
1H and C NMR spectral data for 30,40,5,7-tetralauroyloxyflavone 5.
Position H (ppm)/J (Hz)
C (ppm)
calcb
C (ppm)
2
3
4
5
6
7
8
9
10
10
20
30
40
50
60
–
163.6
104.5
176.1
151.5
114.0
158.2
107.9
156.8
115.4
129.7
122.6
142.5
143.2
123.8
125.0
160.81
108.87a
176.23
150.35
108.92a
157.59
113.87a
154.05
114.95
129.75
121.62a
142.85
145.01
124.29a
124.32a
H
6.62(s)
–
–
7.35(d), J = 2.06
–
6.83(d), J = 2.06
–
5'
H
2
OR
OR
6'
1'
H
4'
3'
8
RO
H
O
9
7
6
2'
H
3
10
H
5
–
–
OR
O
O
7.69(d), J = 2.21
–
–
7.34(d), J = 8.22
7.73(dd), J = 8.22;
J = 2.21
11''
5''
7''
3''
9''
R =
1''
12''
10''
6''
8''
4''
2''
5
100
200
–
172.3
33.5
172.11, 170.90, 170.79, 170.61
34.42, 34.22, 34.12, 34.05
2.74(t), 2.58(m),
2.35(t)
1.77(m), 1.65(m)
1.1(m)
300
4–900
25.0
24.97, 24.90, 24.88, 24.73
29.6, 29.3 29.62, 29.59, 29.49, 29.43, 29.34, 29.32, 29.29, 29.25, 29.23, 29.21,
29.19, 29.14, 29.08, 29.05
31.9
22.7
14.1
1000
1100
1200
1.1(m)
1.1(m)
0.4(t)
31.90
22.67
14.10
a
Verified by a hetero correlation (H ? C), HETCOR.
Values calculated using Chem-Draw™.
b
The anomeric hydrogen C(100) shows a coupling constant of
J = 12.1 Hz indicating a trans position to C(200) [34], and the chem-
ical shift dH = 4.94 ppm of HAC(10) an axial position, which proves
a b-glycosidic linkage to C(40). The 1H-signals of the flavone ring
system are typical for the 30,40,5,7-tetrahydroxyflavone 1 and in
one of the recorded 1H NMR spectra the coupling constants were
obtained. These data were compared to published luteolin-40-O-
screen products as UV-C is blocked off by ozone in the upper atmo-
sphere. The impressive stability of pure luteolin 1 alone shows
promise for use as a natural commercial sunscreen that has still
to prove as alternative to widely used titanium dioxide and other
synthetic sunscreens. Luteolin is not an approved sunscreen and
photobiological effects like sensitizing demand further evaluation.
In conclusion, three distinct UV-absorbers (UV-ABC) based on
luteolin were found:
a
-D-glucopyranoside, whose chemical shifts are in part different,
in particular the HAC(100), which helps to distinguish the
a/b-
glycosidic linkage [35]. An APCI-MS shows the molecular peak at
447.1 M+ of 40-O-b-glucopyranosyl-30,5,7-trihydroxyflavone 10
confirming the presence of a pyranosyl substituent.
(1) UV-A: Pure luteolin 1 and the sigma bond type glycosylated
10 and hydroxyethylated 9 luteolin, kmax ꢀ 330 nm.
(2) UV-B: The 30,40,5,7-tetralipoyloxyflavones 5-7 absorb at
kmax ꢀ 290 nm.
(3) UV-C: The 5-hydroxy-30,40,7-trilipoyloxyflavones 5b–7b
absorb at kmax ꢀ 260 nm.
4. Conclusions
Purified luteolin 1 is surprisingly heat stable (200 °C/7 days)
and provides good UV-A absorption and stability. In contrast, lute-
olin 1 enriched (37%) in an edelweiss extract loses the UV-A screen
activity upon moderate UV-irradiation and is neither heat stable.
Thus, for the manufacturing of a natural sun screen product puri-
fied luteolin and its derivatives are needed for predictable behav-
ior. A series of well defined luteolin derivatives with mostly
biomolecular structure were synthesized or isolated and compared
to native luteolin contained in an edelweiss extract. Fatty acids as
substituents were considered first as they enhance performance in
formulation and skin application, and shift UV-absorption toward
the interesting UV-B maxima. Some derivatives shift even further
to the UV-C band, which is not needed in typical sun screens. In
all purified derivatives no degradation of the luteolin ABC-ring
system upon UV irradiation was observed. Nevertheless 30,40,5,7-
tetralipoyloxyflavones 5–7 lose a fatty acid substituent in C(5)
position upon low UV-irradiation (ꢀ1.5 MJ) which resulted in a
Acknowledgement
The authors thank the HES-SO competence network RealTech
for financial support of Project No. 15576.
References
[1] T. Faunce, K. Murray, H. Nasu, D. Bowman, Sunscreen safety: the precautionary
principle, the australian therapeutic goods administration and nanoparticles in
sunscreens, Nanoethics 2 (2008) 231–240.
[2] Z.M.F. de Freitas, E.P. dos Santos, J.F. da Rocha, G.M. Dellamora-Ortiz, J.C.S.
Gonçalves, A new sunscreen of the cinnamate class: synthesis and enzymatic
hydrolysis evaluation of glyceryl esters of p-methoxycinnamic acid, Eur. J.
Pharm. Sci. 25 (2005) 67–72.
[3] T. Said, M. Dutot, C. Martin, J.-L. Beaudeux, C. Boucher, E. Enee, C. Baudouin, J.-
M. Warnet, P. Rat, Cytoprotective effect against UV-induced DNA damage and
oxidative stress: role of new biological UV filter, Eur. J. Pharm. Sci. 30 (2007)
203–210.
[4] M.A.K. Jansen, V. Gaba, B.M. Greenberg, Higher plants and UV-B radiation:
balancing damage, repair and acclimation, Trends Plant Sci. 3 (1998) 131–135.
[5] A. Heintz, G.A. Reinhardt, Chemie und Umwelt, Vieweg Verlag, Braunschweig/
Wiesbaden, 1993. pp. 59–64.
hypsochromic UVmax shift of
UV-C screen.
D
kB?C ꢀ 30 nm yielding a stable
UV-A screens are of interest for antiageing and cancer preven-
tion; UV-B screens are used for the same purpose and protect from
sunburn. In comparison, a UV-C screen is not really needed in sun-
[6] Y. Miyamura, S.G. Coelho, R. Wolber, S.A. Miller, K. Wakamatsu, B.Z. Zmudzka,
S. Ito, C. Smuda, T. Passeron, W. Choi, J. Batzer, Y. Yamaguchi, J.Z. Beer, V.J.

F. Fischer et al. / Journal of Photochemistry and Photobiology B: Biology 103 (2011) 8–15
15
Hearing, Regulation of human skin pigmentation and responses to ultraviolet
radiation, Pigment Cell Res. 1 (2007) 2–13.
[22] K. Samejima, K. Kanazawa, N. Ashida, G. Danno, Luteolin: a strong antimutagen
against dietary carcinogen Trp-P-2, in peppermint, sage, and thyme, J. Agric.
Food Chem. 43 (1995) 410–414.
[7] G. Liakopoulos, S. Stavrianakou, G. Karabourniotis, Trichome layers versus
dehaired lamina of Olea europaea leaves: differences in flavonoid distribution,
UV-absorbing capacity, and wax yield, Environ. Exp. Bot. 55 (2006) 294–304.
[8] P. Chicaro, E. Pinto, P. Colepicolo, J.L.C. Lopes, N.P. Lopes, Flavonoids from
Lychnophora passerina (Asteraceae): potential antioxidants and UV-protectants,
Biochem. Syst. Ecol. 32 (2004) 239–243.
[9] C.-A. Carron, L. Jaunin, A.F. Grogg, La monarde fistuleuse, source naturelle de
géraniol, d’ acide rosmarinique et de flavonoïdes, Revue Suisse Viticult.
Arboricult. Horticult. 39 (2007) 229–235.
[23] D.Y. Lim, Y. Jeong, A.L. Tyner, J.H.Y. Park, Induction of cell cycle arrest and
apoptosis in HT-29 human colon cancer cells by the dietary compound
luteolin, Am. J. Physiol.-Gastro. Liver Physiol. 292 (2007) G66–G75.
[24] J. Fang, Q. Zhou, X. Shi, B. Jiang, Luteolin inhibits insulin-like growth factor 1
receptor signaling in prostate cancer cells, Carcinogenesis 28 (2007) 713–723.
[25] L. Ziyan, Z. Yongmei, Z. Nan, T. Ning, L. Baolin, Evaluation of the anti-
inflammatory activity of luteolin in experimental animal models, Planta Med.
73 (2007) 221–226.
[10] A.C. Dweck, A review of Edelweiss, SÖFW-J. 130 (2004) 65–68.
[11] Convention of Bern, Convention on the Conservation of European Wildlife and
Natural Habitats, 19 September 1979.
[12] K. Lauber, G. Wagner, in: Flora Helvetica 2.0, Paul Haupt, Bern, 2001.
[13] C. Rey, I. Slacanin, Approche culturale et phytochimique de l’edelweiss, Revue
Suisse Viticult. Arboricult. Horticult. 31 (1999) 89–96.
[14] C.-A. Carron, Ch. Rey, S. Previdoli, C. Baroffio, Helvetia, une nouvelle variété
d’edelweiss issue d’hybrides de clones, Revue Suisse Viticult. Arboricult.
Horticult. 39 (2007) 125–130.
[26] L. Chebil, C. Humeau, A. Falcimaigne, J.-M. Engasser, M. Ghoul, Enzymatic
acylation of flavonoids, Proc. Biochem. 41 (2006) 2237–2251.
[27] M. Natoli, G. Nicolosi, M. Piatelli, Enzyme-catalyzed alcoholysis of flavone
acetates in organic-solvent, Tetrahedron Lett. 31 (1990) 7371–7374.
[28] M. Natoli, G. Nicolosi, M. Piatelli, Regioselective alcoholysis of flavonoid
acetates with lipase in an organic solvent, J. Org. Chem. 57 (1992) 5776–
5778.
[29] A. Saija, A. Tomaino, D. Trombetta, M.L. Pellegrino, B. Tita, C. Messina, F.P.
Bonina, C. Rocco, G. Nicolosi, F. Castelli, ‘In vitro’ antioxidant and
photoprotective properties and interaction with model membranes of three
new quercetin esters, Eur. J. Pharmaceut. Biopharm. 56 (2003) 167–174.
[30] H. Buchholz, R. Rosskopf, A. Lichtenberg, C. Kraus, Method for producing
luteolin and luteolin derivatives, 2003, US 6538021 B1.
[15] K. Kertész, Z. Bálint, Z. Vértesy, G.I. Márk, V. Lousse, J.-P. Vigneron, L.P. Biró,
Photonic crystal type structures of biological origin: structural and spectral
characterization, Curr. Appl. Phys. 6 (2006) 252–258.
9
[16] J.-P. Vigneron, M. Rassart, Z. Vértesy, K. Kertész, M. Sarrazin, L.P. Biro, D. Ertz, V.
Lousse, Optical structure and function of the white filamentary hair covering
the edelweiss bracts, Phys. Rev. E 71 (2005) 011906.
[31] S. Perruchon, Synthese und Struktur-Aktivitäts-Beziehungen von Flavonoiden,
2004, Thesis of the Technische Universität Darmstadt.
[17] P. Casati, V. Walbot, Differential accumulation of maysin and rhamnosylisoorientin
inleavesofhigh-altitudelandracesofmaizeafterUV-Bexposure,PlantCellEnviron.
28 (2005) 788–799.
[18] R. Spitaler, P.D. Schlorhaufer, E.P. Ellmerer, I. Merfort, S. Bortenschlager, H.
Stuppner, C. Zidorn, Altitudinal variation of secondary metabolite profiles in
flowering heads of Arnica Montana cv. ARBO, Phytochemistry 67 (2006) 409–
417.
[19] K.R. Markham, K.G. Ryan, S.J. Bloor, K.A. Mitchel, An increase in the luteolin:
apigenin ratio in Marchantia polymorpha on UV-B enhancement, Phytochemistry 48
(1998) 791–794.
[20] S. Schwaiger, C. Seger, B. Wiesbauer, P. Schneider, E.P. Ellmerer, S. Sturm, H.
Stuppner, Development of an HPLC-PAD-MS assay for the identification and
quantification of major phenolic edelweiss (Leontopodium alpium Cass.)
constituents, Phytochem. Anal. 17 (2006) 291–298.
[32] G. Horneck, P. Rettberg, E. Rabbow, W. Strauch, G. Seckmeyer, R. Facius, G.
Reitz, K. Strauch, J.-U. Schott, Biological dosimetry of solar radiation for
different simulated ozone column thicknesses, J. Photochem. Photobiol., B 32
(1996) 189–196.
[33] E. Damiani, L. Rosati, R. Castagna, P. Carloni, L. Greci, Changes in ultraviolet
absorbance and hence in protective efficacy against lipid peroxidation of
organic sunscreens after UVA irradiation, J. Photochem. Photobiol., B 82 (2006)
204–213.
[34] J.Ø. Duus, C.H. Gotfredsen, K. Bock, Carbohydrate structural determination by
NMR spectroscopy: modern methods and limitations, Chem. Rev. 100 (2000)
4589–4614.
[35] A. Bertrand, S. Morel, F. Lefoulon, Y. Rolland, P. Monsan, M. Remaud-Simeon,
Leuconostoc mesenteroides glucansucrase synthesis of flavonoid glucosides by
acceptor reactions in aqueous-organic solvents, Carbohydr. Res. 341 (2006)
855–863.
[21] M. Furusawa, T. Tanaka, T. Ito, A. Nishikawa, N. Yamazaki, K. Nakaya, N.
Matsuura, H. Tsuchiya, M. Nagayama, M. Iinuma, Antioxidant activity of
hydroxyflavonoids, J. Health Sci. 51 (2005) 376–378.