Semisynthesis of natural flavones inhibiting tubulin polymerization, from hesperidin

Article abstract of DOI:10.1021/np100065v

Semisynthesis of 5,3'-dihydroxy-3,6,7,8,4'-pentamethoxyflavone (1), a natural flavone that binds with high affinity to tubulin, was performed from hesperidin, the very abundant Citrus flavanone, by a five-step sequence. The last step of the synthesis also gave rise to 5,3'-dihydroxy-3,6,7,4'- tetramethoxyflavone (= casticin or vitexicarpin) (10), 5,3'-dihydroxy-3,7,8,4'- tetramethoxyflavone (= gossypetin 3,7,8,4'-tetramethyl ether) (11), and, unexpectedly, 5,7,3'-trihydroxy-3,6,8,4'-tetramethoxyflavone (12) and 5,3'-dihydroxy-8-dimethylamino-3,6,7,4'-tetramethoxyflavone (= 8-dimethylaminocasticin) (13). Cytotoxicity and antitubulin activity of these five flavones, as well as 5,3'-dihydroxy-3,7,4'-trimethoxyflavone (= ayanin) (14) and intermediate 6,8-dibromo-ayanin (8), were evaluated. Comparison of the responses confirmed and clarified the influence of the A-ring substitution pattern on the biological activity.

Full text of DOI:10.1021/np100065v

702
J. Nat. Prod. 2010, 73, 702–706
Semisynthesis of Natural Flavones Inhibiting Tubulin Polymerization, from Hesperidin
Guy Lewin,*,† Alexandre Maciuk,† Sylviane Thoret,‡ Genevie`ve Aubert,‡ Joe¨lle Dubois,‡ and Thierry Cresteil‡
Laboratoire de Pharmacognosie, Faculte´ de Pharmacie, UniVersite´ Paris-Sud 11 BIOCIS UMR-8076 CNRS, AVenue J.B. Cle´ment,
92296 Chaˆtenay-Malabry Cedex, France, and ICSN-CNRS-UPR 2301, AVenue de la Terrasse, 91190 Gif-sur-YVette, France
ReceiVed January 29, 2010
Semisynthesis of 5,3′-dihydroxy-3,6,7,8,4′-pentamethoxyflavone (1), a natural flavone that binds with high affinity to
tubulin, was performed from hesperidin, the very abundant Citrus flavanone, by a five-step sequence. The last step of
the synthesis also gave rise to 5,3′-dihydroxy-3,6,7,4′-tetramethoxyflavone () casticin or vitexicarpin) (10), 5,3′-dihydroxy-
3,7,8,4′-tetramethoxyflavone () gossypetin 3,7,8,4′-tetramethyl ether) (11), and, unexpectedly, 5,7,3′-trihydroxy-3,6,8,4′-
tetramethoxyflavone (12) and 5,3′-dihydroxy-8-dimethylamino-3,6,7,4′-tetramethoxyflavone () 8-dimethylaminocasticin)
(13). Cytotoxicity and antitubulin activity of these five flavones, as well as 5,3′-dihydroxy-3,7,4′-trimethoxyflavone
() ayanin) (14) and intermediate 6,8-dibromo-ayanin (8), were evaluated. Comparison of the responses confirmed and
clarified the influence of the A-ring substitution pattern on the biological activity.
In 1998, Beutler et al. reported results of a comparative
cytotoxicity screening and subsequent studies on tubulin binding
carried out with a series of 79 natural and synthetic flavones.1 Some
structure-activity relationships for cytotoxicity and associated
inhibitory effects on tubulin polymerization were drawn from these
results. Maximum potencies for cytotoxicity and tubulin interaction
were found only for compounds bearing an OH group at C-5 on
the A ring, 3′-hydroxy-4′-methoxy groups on the B ring, and an
OCH3 at C-3 on the C ring. The most potent compound in both
tests was 5,3′-dihydroxy-3,6,7,8,4′-pentamethoxyflavone (1), a
natural flavone first isolated by Mabry et al. in 1986 from
Gutierrezia microcephala.2 Its remarkable antimitotic and cytotoxic
properties were reported in 1994 and in 1995 by the Se´venet3 and
Lee4 groups, respectively. To the best of our knowledge, 1 remains
the most antimitotic natural flavone isolated to date, thus appearing
as a reference compound within this phytochemical class.
Results and Discussion
Access to 1 from 4 was envisioned by the following five-step
sequence: (a) oxidation to the 3-flavonol; (b) cleavage of the
glycosidic bond; (c and d) selective methylation of phenol groups
at C-3 and C-7, then 6,8-dibromination (or in reverse order); (e)
nucleophilic aromatic substitution of bromide by methoxy groups
(Scheme 1). The first step, using the Algar-Flynn-Oyamada
(AFO) method8 (oxidation with alkaline hydrogen peroxide of
hesperidin chalcone, acyclic isomer of 4), has been described by
Pacheco et al.9 and provided crude 7-O-rutinosyltamarixetin (5)
(19%) by direct crystallization from the medium. This low yield,
in accordance with previously reported results, can be explained
mainly by the well-known competing oxidative cyclization of 6′-
substituted chalcones (such as hesperidin chalcone) to aurone and
coumaran-3-one skeletons.10 The second step, hydrolysis of the 7-O-
rutinosyl bond of 5, was achieved under conditions (11 N HCl,
1.5 h at 55 °C) previously used in our laboratory with diosmin
(the flavone analogue of hesperidin) derivatives,11,12 providing
tamarixetin () 4′-O-methylquercetin) (6) in 73% yield. The
methylation of only two (at C-3 and C-7) out of four OH groups
was a crucial step since both free phenol groups at C-5 and C-3′
are known to be necessary for strong inhibition of tubulin
polymerization.1
A recent study on selective methylations of quercetin by Rolando
et al. gave the order of reactivity as 7 > 3 > 3′ > 5 for the phenol
groups in tamarixetin (6).13 We decided to confirm this theoretical
order by a preliminary methylation study, carried out on a small
scale (0.1 mM) with 6 and its 6,8-dibromo analogue (7). According
to our previous studies14,15 in the same field, attempts at bromi-
nation of 6 were undertaken with N-bromosuccinimide (2 equiv,
rt) in a mixture of CH2Cl2-MeOH (2:1). Poor results, however,
led to changing the solvent to trifluoroacetic acid, which gave 6,8-
dibromotamarixetin (7) as the major product (72% yield from 6).
Our interest in antitumor natural compounds, especially fla-
vonoids, led us to look for a ready access to 1. Preparation of 1
could have been planned by total synthesis according to methods
previously described,5,6 but these multistep processes are exces-
sively tedious, partly due to the tetramethoxylation pattern of ring
A. Thus, we turned to semisynthetic methods starting from readily
available natural sources. Tangeretin (2) appeared of interest, since
its alkaline degradation is known to provide in a single step 2′-
hydroxy-3′,4′,5′,6′-tetramethoxyacetophenone (3), a key intermedi-
ate compound toward the synthesis of 1.7 However, although 2
occurs in high concentration in the peel of various Citrus species
such as sweet orange [Citrus sinensis (L.) Pers.] and mandarin
(Citrus reticulata Blanco), commercial sources are scarce and the
product is expensive. Thus we opted for hesperidin (4), the
predominant flavonoid in sweet oranges, a very abundant and
inexpensive starting material, which has the same B-ring substitution
pattern as 1.
* To whom correspondence should be addressed. Tel: 33 1 46 83 55 93.
Fax: 33 1 46 83 53 99. E-mail: guy.lewin@u-psud.fr.
† Univ. Paris-Sud 11 BIOCIS UMR-8076 CNRS.
‡ ICSN-CNRS-UPR 2301, Gif-sur-Yvette.
10.1021/np100065v  2010 American Chemical Society and American Society of Pharmacognosy
Published on Web 03/31/2010

Natural FlaVones from Hesperidin
Journal of Natural Products, 2010, Vol. 73, No. 4 703
Scheme 1. Synthesis of 1 and Analogues from Hesperidin (4)a
a (a) NaOH 2 N, H2O2 30%, 0 °C, 48 h; AcOH to pH 6, then 0 °C, 24 h; sodium metabisulfite, reflux, 2 h; (b) HCl 11 N, 55 °C, 1.5 h, 14% (6 from 4); (c) TFA,
NBS, rt, 4.5 h, 72% (7); (d) DMF, KHCO3, MeI, rt, 2 h, then K2CO3, MeI, rt, 2.5 h, 39% (8), 6% (9); (e) DMF, MeOH, CuBr, MeONa, 130 °C, 22 h, 20% (1), 19%
(10), 2% (11), 8% (12), 4% (13).
1
Brominations at both C-6 and C-8 were confirmed by H NMR
spectroscopy (loss of H-6 and H-8 signals, unchanged signals for
the ABX spin system of the B ring). Methylations of tamarixetin
(6) and its 6,8-dibromo derivative (7) in dimethylformamide with
excess iodomethane and successive additions of KHCO3 (2.2 equiv)
and K2CO3 (1 equiv) gave best results in the case of 7, which led
to 6,8-dibromo-3,7-O-dimethyltamarixetin () 6,8-dibromoayanin)
(8) in 39% (the monomethylation product 9 was also isolated in
6% yield).
at C-8, is well known with aryl halogens ortho to a methoxy
group.23 The natural flavones 1,3,4 10,16,24 and 1125–27 were
identified unambiguously by comparison of their melting points
and spectroscopic characteristics with literature values. Formation
of the last two flavones, 12 and 13, was rather unexpected.
Compound 12 probably originates from 1 by nucleophilic attack
at the C-7 methoxy group (nucleophilic cleavage of this methyl
ether is facilitated by the para carbonyl function and the two
adjacent methoxy groups), while 13 results from a nucleophilic
substitution of the bromide at C-8 by dimethylamine. These two
reactions can be explained by the reported property of DMF to
decompose to dimethylamine by extended heating under some
conditions.28,29 The structure of 12, a natural flavone, first isolated
The expected reactivity of 7- and 3-hydroxy groups was proved
by NMR experiments. Methylation of 7-OH was deduced from the
observed HMBC correlation between signals of a methoxy group
at δ 3.90 and C-7 at δ 159.0. Methylation of the 3-OH group was
ascertained from comparison of the spectra with literature NMR
data for 3,4′-dimethoxy-3′-hydroxyflavones3,16,17 and 3′,4′-dimethoxy-
3-hydroxyflavones.18,19 Particularly significant in the 13C NMR
spectrum of 8 were the signals of C-2, C-2′, and 3-OMe at δ 156.2,
115.1, and 59.7, respectively. The remaining step was substitution
of the bromines at C-6 and C-8 by methoxy groups, a reaction that
has been described as a MeO-/CuBr-promoted methanolysis in a
mixture of MeOH-DMF.20 In this paper, Bovicelli et al. reported
excellent yields (89% and 80%) with two 6,8-dibromoflavones (1
h reaction times), but they stressed the versatility and the high
dependence of the reaction on the experimental protocol, i.e., the
quality of the CuBr catalyst and the composition of the solvent,
which must contain a small percentage of water.20,21 Unfortunately,
in our hands, a first attempt with 6,8-dibromodiosmetin (used as a
model) proved to be unsuccessful, since the dibromoflavone was
fully recovered. Therefore, we decided to use the same reagents
and solvent, but under harsher, more classical conditions such as
those described by Guillaumet et al.22 Methanolysis of 6,8-
dibromoayanin (8) by heating it at 130 °C in DMF-MeOH with
MeONa (10 equiv) and CuBr (0.026 equiv) for 22 h led to SNAr
at both C-6 and C-8, giving the expected flavone (1) as the major
compound (20%). Four other flavones, 5,3′-dihydroxy-3,6,7,4′-
tetramethoxyflavone () casticin or vitexicarpin) (10), 5,3′-dihy-
droxy-3,7,8,4′-tetramethoxyflavone () gossypetin 3,7,8,4′-tetram-
ethyl ether) (11), 5,7,3′-trihydroxy-3,6,8,4′-tetramethoxyflavone
(12), and 5,3′-dihydroxy-8-dimethylamino-3,6,7,4′-tetramethoxy-
flavone () 8-dimethylaminocasticin) (13), were also isolated in
19%, 2%, 8%, and 4% yields, respectively.
at the same time as 1,2 was established by mass, 1H NMR, and 13
C
NMR spectrometry and confirmed by comparison of these spec-
troscopic data with literature values.30 Lastly, the presence of the
dimethylamino group in 13 was deduced from the mass (pseudo-
1
molecular ion at [M + H]+ 418) and H NMR spectra (s, 6H at δ
2.96), and its position at C-8 was inferred from NOESY experiments
(significant NOE correlations were observed between signals of
the dimethylamino group and H-2′ and 6′ on one hand and signals
of OH at C-5 and OCH3 at C-6 on the other hand).
This chemical study led to the isolation of six flavones having
in common OH groups at C-5 and C-3′ and OCH3 groups at C-3
and C-4′, favorable structural requirements for cytotoxicity and
inhibition of tubulin polymerization (ITP).1,31 A seventh flavone
with the same substituents, ayanin (14), was also prepared by
hydrogenolysis of 8 (H2, 10% Pd-C, rt, 48 h in DMF). As these
seven flavones differ only in C-6, C-7, and C-8 substituents, we
decided to measure their antiproliferative and ITP activities. We
thought indeed that such a comparative study, in the same set of
experiments, would improve knowledge about the influence of the
A-ring substitution pattern on these biological properties. It is
noteworthy that no comparison between activities of flavone 1 and
casticin(10),awell-knowncytotoxicandtubulin-bindingagent,24,32–35
had been reported until now. Furthermore, cytotoxicity and ITP
have not been investigated for the two other naturally occurring
flavones, gossypetin 3,7,8,4′-tetramethyl ether (11) and 5,7,3′-
trihydroxy-3,6,8,4′-tetramethoxyflavone (12). The antiproliferative
effect of flavones was assayed on KB human buccal carcinoma
cells as well as the activation of caspase 3 with DEVD-AMC as
substrate in HL60 human leukemia cells. Inhibition of tubulin
polymerization was determined according to Zavala and Guenard’s
method.36 Compounds were tested at 1 mg/mL (≈ 2.5 × 10-3 M),
Formation of flavones 10 and 11 is consistent with the combina-
tion of a SNAr reaction at C-6 or C-8 and a reductive debromination
on the other carbon. The side reaction, observed in this case mainly

704 Journal of Natural Products, 2010, Vol. 73, No. 4
Lewin et al.
water, and dried with P2O5 under vacuum to provide crude 7-O-
rutinosyltamarixetin (5) (9.6 g, 19%). Hydrolysis of crude 5 (9.2 g)
was achieved by treatment with 11 N HCl (120 mL) at 55 °C for 1.5 h.
The mixture was diluted with water (3 L), then extracted with EtOAc
(3 × 500 mL). The organic phase was made clear by addition of MeOH
(100 mL), dried with Na2SO4, filtered, and evaporated to dryness. The
dried residue (5 g) was purified by flash chromatography (silica gel,
CH2CH2-MeOH, 96:4) to provide pure tamarixetin (6) (3.37 g, 73%).
Tamarixetin (6): pale yellow crystals, mp 254-257 °C (lit.13,37
252-254 °C; 253-256 °C); 1H NMR (DMSO-d6) δ 3.82 (3H, s, OCH3-
4′), 6.18 (1H, d, J ) 2 Hz, H-6), 6.42 (1H, d, J ) 2 Hz, H-8), 7.04
(1H, d, J ) 8.8 Hz, H-5′), 7.61 (2H, m, H-2′ and H-6′), 12.40 (1H, s,
OH-5).
Table 1. Antiproliferative, Proapoptotic, and Antitubulin Activities
of Synthesized Flavones
activation of
cytotoxicity
apoptosis
compd
on KB cellsa
in HL60b
ITP activityc
Natural Flavones
1
10
11
12
14
IC50 ) 8 nM
IC50 ) 97 nM
9%
10%
3%
50 nM (×6.9)
1.3 µM (0.46)d
19 µM (6.8)d
0% inhibition
14 µM (5)d
500 nM (×5.3)
n.d.
n.d.
n.d.
0% inhibition
Non-natural Flavones
Bromination of Tamarixetin (6). NBS (3.47 g, 19.3 mmol) was
added to a solution of 6 (3.05 g, 9.65 mmol) in TFA (100 mL), then
stirred for 4.5 h at room temperature. The reaction mixture was diluted
with iced water, and the resulting precipitate was filtered, washed with
water, and dried with KOH pellets under vacuum overnight. The yellow
crystallized residue (3.3 g, 72%) gave a homogeneous spot [TLC on
silica gel (CH2Cl2-MeOH, 96:4)] and was identified as pure 7
according to NMR spectra.
8
13
IC50 ) 404 nM
40%
500 nM (×3.2)
36% inhibition
0% inhibition
n.d.
a As measured by the MTS assay after 72 h incubation of cells with
drug: results are expressed as the percentage of inhibition of cell growth
with 10-6 M flavone concentration or as IC50 (nM), calculated only for
the three most active compounds. b Activation of caspases 3/7: optimal
concentration of compound and fold-activation over control value after a
48 h exposure. c Results are expressed as the percentage of ITP at ∼2.5
× 10-3 M or as IC50 (µM). d IC50 compound/IC50 DPPT. n.d.: not determined.
6,8-Dibromotamarixetin (7): yellow crystals, mp 269-272 °C; 1H
NMR (DMSO-d6) δ 3.87 (3H, s, OCH3-4′), 7.14 (1H, d, J ) 8.8 Hz,
H-5′), 7.78 (1H, dd, J ) 8.8 and 1.6 Hz, H-6′), 7.83 (1H, d, J ) 1.6
Hz, H-2′), 9.39 and 9.90 (2H, 2s, OH-3 and OH-3′), 13.39 (1H, s, OH-
5); 13C NMR (DMSO-d6) δ 55.2 (CH3, OCH3-4′), 87.4 (C, C-8), 93.1
(C, C-6), 103.8 (C, C-10), 111.4 (CH, C-5′), 114.3 (CH, C-2′), 119.5
(CH, C-6′), 122.7 (C, C-1′), 136.0 (C, C-3), 145.9 (C, C-3′), 146.8 (C,
C-2), 149.3 (C, C-4′), 150.3 (C, C-9), 155.8 and 156.3 (2C, C-5 and
C-7), 175.0 (C, C-4); ESIMS (-) m/z [M - H]- 471-473-475.
Methylation of 6,8-Dibromotamarixetin (7). To a solution of 7
(3.18 g, 6.71 mmol) in DMF (100 mL) at room temperature were added
KHCO3 (1.55 g, 15.5 mmol) and iodomethane (5 mL, 80 mmol), and
the reaction mixture was stirred for 2 h at room temperature. K2CO3
(0.98 g, 7.1 mmol) and a new amount of iodomethane (5 mL) were
added, and the medium was stirred for a further 2.5 h. The reaction
mixture was filtered and evaporated to dryness. Crystallization of the
dried residue from CH2Cl2-MeOH afforded 2.37 g of a mixture of
two main compounds [TLC on silica gel (CH2CH2-MeOH, 99:1): two
blue spots with FeCl3]. Purification of this by flash chromatography
(silica gel, CH2CH2-MeOH, 99:1), then crystallization from
CH2Cl2-MeOH (8) or THF-EtOH (9), provided pure 6,8-dibromoaya-
nin (8) (1.32 g, 39%) and 9 (0.205 g, 6%).
and results were given as the percentage of ITP or as IC50, calculated
for the most active compounds, and also expressed in relation to
deoxypodophyllotoxin (DPPT) in terms of the IC50/IC50 DPPT ratio.
As depicted in Table 1, only flavone 1 and casticin (10) possess
antiproliferative, proapoptotic, and ITP activities. The 10-fold higher
activity of 1 versus 10 highlights the significant importance of an
additional methoxy group at C-8 in 1. Flavone 11, an isomer of
casticin, is devoid of both activities, like ayanin (14), previously
reported1 to be inactive. Comparison of isomeric flavones 10 (6-
methoxylated) and 11 (8-methoxylated) corroborates the critical
influence of the substitution pattern on the activity, while flavone
12, the 7-O-demethyl analogue of 1, displays a discrepancy between
a very weak cytotoxicity and an ITP activity in the range of casticin.
Lastly, responses of 8 and 13, the two non-natural evaluated
flavones, were rather unexpected: 8-dimethylaminocasticin (13),
which differs from 1 and 10 only in the substituent at C-8, was
weakly cytotoxic and completely inactive on ITP, whereas 8, the
6,8-dibromo analogue of ayanin, showed significant antiproliferative
and moderate ITP effects.
6,8-Dibromo-3,7-O-dimethyltamarixetin (6,8-dibromoayanin) (8):
1
pale yellow crystals, mp 234-236 °C; H NMR (DMSO-d6) δ 3.83
(3H, s, OCH3-3), 3.87 (3H, s, OCH3-4′), 3.90 (3H, s, OCH3-7), 7.14
(1H, d, J ) 8.8 Hz, H-5′), 7.69 (1H, dd, J ) 8.8 and 1.6 Hz, H-6′),
7.70 (1H, d, J ) 1.6 Hz, H-2′), 9.53 (1H, s, OH-3′), 13.51 (1H, s,
OH-5); 13C NMR (DMSO-d6) δ 55.7 (CH3, OCH3-4′), 59.7 (CH3,
OCH3-3), 61.0 (CH3, OCH3-7), 94.4 (C, C-8), 99.6 (C, C-6), 108.7 (C,
C-10), 112.1 (CH, C-5′), 115.1 (CH, C-2′), 120.7 (CH, C-6′), 121.7
(C, C-1′), 138.5 (C, C-3), 146.5 (C, C-3′), 150.9 (C, C-4′), 151.0 (C,
C-9), 156.2 (C, C-2), 157.2 (C, C-5), 159.0 (C, C-7), 177.7 (C, C-4);
ESIMS (-) m/z [M - H]- 499-501-503.
In conclusion, this study refines structure-activity relationships
previously predicted.1,31 Our biological results confirm that the most
favorable A-ring substitution pattern for strong antiproliferative and
ITP activities consists of an OH group at C-5 and OCH3 groups at
C-6, C-7, and C-8, as in 1.
Experimental Section
General Experimental Procedures. Melting points were determined
with a micro-Koffler apparatus and are uncorrected. NMR spectra,
including NOESY and 1H-13C (HMQC and HMBC) experiments, were
recorded on Bruker AC-300 (300 MHz) or Bruker AM-400 (400 MHz)
spectrometers. ESIMS were recorded on a Navigator Aqua thermoquest
spectrometer or an Agilent HP 1100 MSD spectrometer (ESI source).
Flash chromatography was performed with silica gel 60 (9385 Merck)
and, for final purifications, with a Spot 1 flash chromatography
integrated system and Reveleris silica 40 µm, 12 g cartridges.
Preparative TLC was performed using 60 F 254 silica gel (5715 Merck).
Hesperidin was purchased from Acros Organics.
Tamarixetin (6) from Hesperidin (4). The AFO oxidation of 4
was accomplished exactly under conditions described previously.9 A
solution of 4 (50 g, 82 mmol) in aqueous 2 N NaOH (1 L) was cooled
at 0 °C, aqueous H2O2 30% (50 mL) was added, and the mixture was
left for 24 h at 0 °C. To the mixture was added the same volume of
aqueous H2O2 30% (50 mL), and the solution was left for 24 h more
at 0 °C, adjusted to pH 6 with glacial acetic acid, then kept again for
24 h at 0 °C. Sodium metabisulfite (120 g) was then added, and the
mixture was heated and stirred under reflux for 2 h. The abundant
resulting yellow precipitate was recovered by filtration, washed with
6,8-Dibromo-7-O-methyltamarixetin (9): yellow crystals, mp
257-260 °C; 1H NMR (DMSO-d6) δ 3.86 (3H, s, OCH3-4′), 3.90 (3H,
s, OCH3-7), 7.13 (1H, d, J ) 8.8 Hz, H-5′), 7.79 (1H, dd, J ) 8.8 and
1.6 Hz, H-6′), 7.83 (1H, d, J ) 1.6 Hz, H-2′); 13C NMR (DMSO-d6)
δ 55.6 (CH3, OCH3-4′), 61.0 (CH3, OCH3-7), 94.8 (C, C-8), 98.6 (C,
C-6), 107.4 (C, C-10), 111.8 (CH, C-5′), 114.8 (CH, C-2′), 120.1 (CH,
C-6′), 122.9 (C, C-1′), 136.9 (C, C-3), 146.3 (C, C-3′), 148.0 (C, C-2),
149.9 and 150.4 (2C, C-4′ and C-9), 156.5 (C, C-5), 158.6 (C, C-7),
175.6 (C, C-4); ESIMS (-) m/z [M - H]- 485-487-489.
Methanolysis of 6,8-Dibromoayanin (8). To a solution of 8 (0.5
g, 1 mmol) in DMF (3.6 mL) in a test tube were added CuBr (0.038
g, 0.26 mmol), MeOH (1.7 mL), and 30% NaOMe in MeOH (1.9 mL,
10 mmol). The test tube was sealed, and the reaction mixture was stirred
at 130 °C for 22 h. The cooled reaction mixture was taken up in water
and extracted at pH 6 with CH2Cl2. Standard workup of the organic
layer provided an amorphous residue (0.390 g). Purification of the
residue by flash chromatography (silica gel, CH2CH2-MeOH, 99:1)
provided a mixture of 1, 10, 11, and 13 and pure compound 12 (0.03
g, 8%). A second flash chromatography (gradient elution from 100%
cyclohexane to 50:50 cyclohexane-acetone in 30 min) led to isolation
of pure 1 (0.081 g, 20%), 10 (0.07 g, 19%), and 13 (0.017 g, 4%), and

Natural FlaVones from Hesperidin
Journal of Natural Products, 2010, Vol. 73, No. 4 705
11 in a crude fraction. Separation of the latter by preparative TLC (silica
gel, CH2CH2-MeOH, 99.5:0.5) gave pure 11 (0.007 g, 2%).
5,3′-Dihydroxy-3,6,7,8,4′-pentamethoxyflavone (1): yellow crystals,
mp 169-171 °C (lit.3 170 °C); 1H NMR (CDCl3) δ 3.87 (3H, s, OCH3),
3.96 (6H, s, 2 OCH3), 3.99 (3H, s, OCH3), 4.11 (3H, s, OCH3), 5.73
(1H, s, OH-3′), 7.00 (1H, d, J ) 8.8 Hz, H-5′), 7.78 (2H, m, H-2′ and
H-6′), 12.38 (1H, s, OH-5); ESIMS (+) m/z [M + Na]+ 427.
Casticin (10): pale yellow crystals, mp 187-189 °C (lit.24 188-190
with compounds dissolved in DMSO. Control cells received the vehicle
only, and positive control cells were treated with 1 µM doxorubicine.
Cells were lysed with a buffer consisting of 25 mM HEPES, pH 7.5,
0.5 mM EDTA, 0.05% NP40, 0.001% SDS, and 5 mM DTT containing
50 µM Ac-DEVD-AMC (Biomol). Fluorescence was monitored (λex
) 360 nm, λem ) 465 nm) over a 3 h period.
Inhibition of Tubulin Polymerization Assay. Sheep brain micro-
tubule proteins were purified by two cycles of assembly/disassembly
at 37 °C/0 °C in MES buffer: 100 mM MES (2-[N-morpholino]et-
hanesulfonic acid, pH 6.6), 1 mM EGTA (ethyleneglycolbis[ꢀ-
aminoethyl ether]-N,N,N′,N′-tetraacetic acid), 0.5 mM MgCl2. All
samples were dissolved in DMSO. The evaluated compound (1 µL)
was added to a microtubular solution (150 µL) that was incubated
at 37 °C for 10 min and at 0 °C for 5 min. The tubulin
polymerization rate was measured by turbidimetry at 350 nm
according to Zavala and Gue´nard’s protocol36 using deoxypodo-
phyllotoxin as reference compound.
1
°C); H NMR (CDCl3) δ 3.88 (3H, s, OCH3), 3.93 (3H, s, OCH3),
3.97 (3H, s, OCH3), 4.00 (3H, s, OCH3), 5.71 (1H, s, OH-3′), 6.51
(1H, s, H-8), 6.96 (1H, d, J ) 8.8 Hz, H-5′), 7.68 (1H, d, J ) 1.6 Hz,
H-2′), 7.72 (1H, dd, J ) 8.8 and 1.6 Hz, H-6′), 12.57 (1H, s, OH-5);
1H NMR (DMSO-d6) δ 3.75 (3H, s, OCH3), 3.82 (3H, s, OCH3), 3.89
(3H, s, OCH3), 3.94 (3H, s, OCH3), 6.89 (1H, s, H-8), 7.13 (1H, d, J
) 8.8 Hz, H-5′), 7.61 (2H, m, H-2′ and H-6′), 9.45 (1H, br s, OH-3′),
12.60 (1H, br s, OH-5); ESIMS (+) m/z [M + H]+ 375, [M + Na]+
397.
Gossypetin 3,7,8,4′-tetramethyl ether (11): yellow crystals, mp
185-187 °C (lit.25,26 184-185 °C; 185.5-186.5 °C); 1H NMR
(DMSO-d6) δ 3.82 (3H, s, OCH3), 3.83 (3H, s, OCH3), 3.88 (3H, s,
OCH3), 3.92 (3H, s, OCH3), 6.60 (1H, s, H-6), 7.16 (1H, d, J ) 8.8
Hz, H-5′), 7.61 (2H, m, H-2′ and H-6′), 9.56 (1H, br s, OH-3′), 12.48
(1H, br s, OH-5); ESIMS (+) m/z [M + Na]+ 397.
Acknowledgment. The authors thank J.-C. Jullian, M. Danet, and
K. Leblanc for NMR and mass spectra measurements.
References and Notes
(1) Beutler, J. A.; Hamel, E.; Vlietinck, A. J.; Haemers, A.; Rajan, P.;
Roitman, J. N.; Cardellina, J. H., II; Boyd, M. R. J. Med. Chem. 1998,
41, 2333–2338.
5,7,3′-Trihydroxy-3,6,8,4′-tetramethoxyflavone (12): yellow crys-
tals, mp 184-186 °C (lit.30 182-184 °C); 1H NMR (DMSO-d6) δ 3.78
(3H, s, OCH3-6), 3.80 (3H, s, OCH3-3), 3.85 (3H, s, OCH3-8), 3.87
(3H, s, OCH3-4′), 7.13 (1H, d, J ) 8.8 Hz, H-5′), 7.58 (2H, m, H-2′
and H-6′), 9.55 and 10.4 (2H, 2 br s, OH- 7 and 3′), 12.48 (1H, s,
OH-5); 13C NMR (DMSO-d6) δ 55.6 (CH3, OCH3-4′), 59.7 and 61.0
(2 CH3, OCH3-3 and -7), 61.2 (CH3, OCH3-8), 103.4 (C, C-10), 112.0
(CH, C-5′), 114.8 (CH, C-2′), 120.2 (CH, C-6′), 122.4 (C, C-1′), 128.0
(C, C-8), 131.4 (C, C-6), 137.7 (C, C-3), 144.5 (C, C-9), 146.4 (C,
C-3′), 147.9 (C, C-5), 150.3 (C, C-4′), 151.0 (C, C-7), 155.1 (C, C-2),
178.4 (C, C-4); ESIMS (+) m/z [M + Na]+ 413.
(2) Fang, N.; Leidig, M.; Mabry, T. J. Phytochemistry 1986, 25, 927–
934.
(3) Lichius, J. J.; Thoison, O.; Montagnac, A.; Pa¨ıs, M.; Gue´ritte-
Voegelein, F.; Se´venet, T. J. Nat. Prod. 1994, 57, 1012–1016.
(4) Shi, Q.; Li, L.; Chang, J.-J.; Autry, C.; Kozuka, M.; Konoshima, T.;
Estes, J. R.; Lin, C. M.; Hamel, E.; McPhail, A. T.; McPhail, D. R.;
Lee, K.-H. J. Nat. Prod. 1995, 58, 475–482.
(5) Horie, T.; Kawamura, Y.; Yamamoto, H.; Yamashita, K. Chem. Pharm.
Bull. 1995, 43, 2054–2063.
(6) Fougerousse, A.; Gonzalez, E.; Brouillard, R. J. Org. Chem. 2000,
65, 583–586.
(7) Burnham, W. S.; Sidwell, R. W.; Tolman, R. L.; Stout, M. G. J. Med.
Chem. 1972, 15, 1075–1076.
(8) Cummins, B.; Donnelly, D. M. X.; Eades, J. F.; Fletcher, H.; Cinneide,
F. O’.; Philbin, E. M.; Swirski, J.; Wheeler, T. S.; Wilson, R. K.
Tetrahedron 1963, 19, 499–512.
(9) (a) Pacheco, H.; Grouiller, A. Bull. Soc. Chim. Fr. 1966, 3212–3217.
(b) French Patent (Brevet Special de Me´dicament) 3.794 M, 1965.
(10) Adams, C. J.; Main, L. Tetrahedron 1991, 47, 4979–4990.
(11) Quintin, J.; Roullier, C.; Thoret, S.; Lewin, G. Tetrahedron 2006, 62,
4038–4051.
(12) Ferte´, J.; Ku¨hnel, J.-M.; Chapuis, G.; Rolland, Y.; Lewin, G.;
Schwaller, M. J. Med. Chem. 1999, 42, 478–489.
(13) Bouktaib, M.; Lebrun, S.; Atmani, A.; Rolando, C. Tetrahedron 2002,
58, 10001–10009.
(14) Quintin, J.; Lewin, G. Tetrahedron Lett. 2005, 46, 4341–4343.
(15) Lewin, G.; Jullian, J.-C.; Rodrigo De Losada, J. Tetrahedron Lett.
2010, 51, 1145–1148.
(16) Barbera, O.; Marco, J. A.; Sanz, J. F.; Sanchez-Parareda, J. Phy-
tochemistry 1986, 25, 2357–2360.
(17) Wang, Y.; Hamburger, M.; Gueho, J.; Hostettmann, K. Phytochemistry
1989, 28, 2323–2327.
(18) Dong, H.; Gou, Y.-L.; Cao, S.-G.; Chen, S.-X.; Sim, K.-Y.; Goh, S.-
H.; Kini, R. M. Phytochemistry 1999, 50, 899–902.
(19) Xiong, Q.; Shi, D.; Mizuno, M. Phytochemistry 1995, 39, 723–725.
(20) Bovicelli, P.; D’Angelo, V.; Collalto, D.; Verzina, A.; D’Antona, N.;
Lambusta, D. J. Pharm. Pharmacol. 2007, 59, 1697–1701.
(21) Bovicelli, P.; Antonioletti, R.; Barontini, M.; Borioni, G.; Bernini,
R.; Mincione, E. Tetrahedron Lett. 2005, 46, 1255–1257.
(22) Besson, T.; Ruiz, N.; Coudert, G.; Guillaumet, G. Tetrahedron 1995,
51, 3197–3204.
(23) Lindley, J. Tetrahedron 1984, 40, 1433–1456.
(24) Diaz, F.; Chavez, D.; Lee, D.; Mi, Q.; Chai, H.-B.; Tan, G. T.;
Kardono, L. B. S.; Riswan, S.; Fairchild, C. R.; Wild, R.; Farnsworth,
N. R.; Cordell, G. A.; Pezzuto, J. M.; Kinghorn, A. D. J. Nat. Prod
2003, 66, 865–867, and references therein.
5,3′-Dihydroxy-8-dimethylamino-3,6,7,4′-tetramethoxyflavone (13):
1
yellow crystals, mp 160-162 °C; H NMR (CDCl3) δ 2.96 (6H, s,
N(CH3)2), 3.87 (3H, s, OCH3-3), 3.93 (3H, s, OCH3-6), 3.99 (3H, s,
OCH3-4′), 4.12 (3H, s, OCH3-7), 5.75 (1H, br s, OH-3′), 7.00 (1H, d,
J ) 8.8 Hz, H-5′), 7.77 (2H, m, H-2′ and H-6′), 12.64 (1H, s, OH-5);
13C NMR (CDCl3) main characteristic signals at δ 44.1 (2 CH3,
N(CH3)2), 56.0 (CH3, OCH3-4′), 60.1 (CH3, OCH3-3), 60.9 (CH3, OCH3-
6), 61.7 (CH3, OCH3-7), 110.5 (CH, C-5′), 114.6 (CH, C-2′), 121.7
(CH, C-6′), 123.5 (C, C-1′), 124.9 (C, C-8), 136.4 (C, C-6), 138.6 (C,
C-3), 145.6 (C, C-3′), 148.9 (C, C-4′), 157.0 (C, C-7); ESIMS (+) m/z
[M + H]+ 418.
Hydrogenolysis of 6,8-Dibromoayanin (8). A solution of 8 (20 mg,
0.04 mmol) in DMF (3 mL) was hydrogenated under 1 atm of hydrogen
with 10% Pd-C (20 mg) at room temperature for 48 h. The catalyst
was separated and the filtrate concentrated to dryness. Crystallization
of the dried residue (MeOH) afforded pure ayanin (14) (0.010 g, 73%).
Ayanin (14): pale yellow crystals, mp 169-171 °C (lit.38 169 °C);
1H NMR (CDCl3) δ 3.88 (6H, 2s, OCH3-3 and -4′), 3.99 (3H, s, OCH3-
7), 5.68 (1H, s, OH-3′), 6.35 (1H, d, J ) 2 Hz, H-6), 6.46 (1H, d, J )
2 Hz, H-8), 6.96 (1H, d, J ) 8.8 Hz, H-5′), 7.68-7.70 (2H, m, H-2′
and H-6′), 12.63 (1H, s, OH-5).
Cell Culture. The human cell lines KB (nasopharyngeal epidermoid
carcinoma) and HL60 (promyelocytic leukemia) were purchased from
ATCC. KB cells were cultured in D-MEM medium supplemented with
10% fetal calf serum, in the presence of penicillin, streptomycin, and
fungizone in a 75 cm2 flask under 5% CO2, whereas HL60 cells were
cultured in complete RPMI medium.
Cell Proliferation Assay. Cells (600 cells/well) were plated in 96-
well tissue culture microplates in 200 µL of medium and treated 24 h
later with compounds dissolved in DMSO at concentrations that ranged
0.5 nM to 10 µM with a Biomek 3000 automation workstation
(Beckman-Coulter). Control cells received the same volume of DMSO
(1% final volume). After 72 h exposure to the drug, MTS reagent
(Promega) was added and incubated for 3 h at 37 °C. Experiments
were performed in triplicate: the absorbance was monitored at 490 nm,
and results were expressed as the inhibition of cell proliferation
calculated as the ratio [(1 - (OD490 treated/OD490 control)) × 100].
For IC50 determinations (50% inhibition of cell proliferation) experi-
ments were performed in duplicate.
(25) Krishnamurti, M.; Seshadri, T. R.; Shankaran, P. R. Indian J. Chem.
1967, 5, 137–139.
(26) Farkas, L.; Nogradi, M.; Muszaki, E.; Szerves-Kem, T. Magy. Kem.
Foly. 1967, 73, 472–474.
(27) Wollenweber, E.; Fischer, R.; Doerr, M.; Irvine, K.; Pereira, C.;
Stevens, J. F. Z. Naturforsch., C: Biosci. 2008, 63, 731–739.
(28) Cho, Y. H.; Park, J. C. Tetrahedron Lett. 1997, 38, 8331–8334.
(29) Watanabe, T.; Tanaka, Y.; Sekiya, K.; Akita, Y.; Ohta, A. Synthesis
1980, 3, 9–41.
Activation of Caspase 3. HL60 cells (20 000 cells/well) were plated
in 96-well black tissue culture microplates and treated for 24 and 48 h

706 Journal of Natural Products, 2010, Vol. 73, No. 4
Lewin et al.
(30) Horie, T.; Kitou, T.; Kawamura, Y.; Yamashita, K. Bull. Chem. Soc.
Jpn. 1996, 69, 1033–1041.
(31) Lopez-Lazaro, M.; Galvez, M.; Martin-Cordero, C.; Ayuso, M. J. Stud.
Nat. Prod. Chem. 2002, 27, 891–932.
(35) Shen, J.-K.; Du, H.-P.; Yang, M.; Wang, Y.-G.; Jin, J. Ann. Hematol.
2009, 88, 743–752.
(36) Zavala, F.; Gue´nard, D.; Robin, J.-P.; Brown, E. J. Med. Chem. 1980,
23, 546–549.
(32) Kobayakawa, J.; Sato-Nishimori, F.; Moriyasu, M.; Matsukawa, Y.
Cancer Lett. 2004, 208, 59–64.
(33) Li, W.-X.; Cui, C.-B.; Cai, B.; Wang, H.-Y.; Yao, X.-S. J. Asian Nat.
Prod. Res. 2005, 7, 615–626.
(37) Le Quesne, P. W.; Levery, S. B.; Menachery, M. D.; Brennau, T. F.;
Raffauf, R. F. J. Chem. Soc., Perkin Trans. 1 1978, 1572–1580.
(38) Jurd, L. J. Org. Chem. 1962, 27, 1294–1297.
(34) Haidara, K.; Zamir, L.; Shi, Q.-W.; Batist, G. Cancer Lett. 2006, 242,
180–190.
NP100065V