Antiangiogenic activity of synthetic dihydrobenzofuran lignans

Article abstract of DOI:10.1021/np0103968

A series of synthetic dihydrobenzofuran lignans, obtained by biomimetic oxidative dimerization of caffeic or ferulic acid methyl ester followed by derivatization reactions, was tested for its antiangiogenic activity in the CAM (chorioallantoic membrane) a

Full text of DOI:10.1021/np0103968

7
18
J . Nat. Prod. 2002, 65, 718-720
Notes
An tia n giogen ic Activity of Syn th etic Dih yd r oben zofu r a n Lign a n s
†
‡
‡
†
§
§
Sandra Apers, Dietrich Paper, J utta B u¨ rgermeister, Slavka Baronikova, Stefaan Van Dyck, Guy Lemi e` re,
†
,†
Arnold Vlietinck, and Luc Pieters*
Department of Pharmaceutical Sciences, University of Antwerp, Universiteitsplein 1, B-2610 Antwerp, Belgium,
Department of Pharmacy, University of Regensburg, Universit a¨ tsstrasse 31, D-93040 Regensburg, Germany, and
Department of Chemistry, University of Antwerp, Groenenborgerlaan 171, B-2020 Antwerp, Belgium
Received August 10, 2001
A series of synthetic dihydrobenzofuran lignans, obtained by biomimetic oxidative dimerization of caffeic
or ferulic acid methyl ester followed by derivatization reactions, was tested for its antiangiogenic activity
in the CAM (chorioallantoic membrane) assay. The dimerization product of caffeic acid methyl ester (2a )
(methyl (E)-3-[2-(3,4-dihydroxyphenyl)-7-hydroxy-3-methoxycarbonyl-2,3-dihydro-1-benzofuran-5-yl]prop-
2
-enoate) showed a pronounced antiangiogenic activity, especially the 2R,3R-enantiomer.
Angiogenesis or neovascularization is a complex process
involving the activation, adhesion, proliferation, and trans-
migration of endothelial cells from preexisting blood ves-
sels. It plays a critical role in normal physiological pro-
cesses such as wound healing, but also in a number of
pathological processes, for instance, diabetic retinopathy,
arthritis, and the growth of solid tumors. Therefore angio-
genesis is considered as a potential target for antitumoral
therapy. Some endogenic angiogenesis inhibitors that are
currently being assessed in clinical trials include angiosta-
tin and endostatin, which both are fragments of larger
circulatory proteins (plasminogen and collagen XVIII,
respectively); inhibitors of vascular endothelial growth
factor (VEGF); platelet factor 4 (PF4), a protein present in
R-granules of platelets; and recombinant human PF4.
Nonendogenous angiogenesis inhibitors that have entered
clinical trials include CM 101, a group B streptococcus
polysaccharide toxin that is responsible for pulmonary
disease in infected human neonates; TNP-470, a derivative
of the microbial antibiotic fumagillin; genistein, an isofla-
vonoid with a selective tyrosine kinase inhibitory activity;
pentosan polysulfate, a xylan polysulfate; and tecogalan,
a sulfated polysaccharide-peptidoglycan complex isolated
from the cell wall of an Arthrobacter species.1 Charac-
terization of new antiangiogenic leads may yield new
therapeutic agents in this area. Among the known angio-
genesis inhibitors, natural compounds such as sulfated
Sch em e 1
-5
2
carbohydrates or triterpenoids play a prominent role. In
the present study the antiangiogenic activity of synthetic
3
′,4-di-O-methylcedrusin 4c, a dihydrobenzofuran lignan
6
originally isolated from South American Croton species,
and some synthetic precursors and analogues (2-4) was
evaluated in the CAM (chorioallantoic membrane) assay,
an in vivo model for angiogenesis.7
Synthesis of dihydrobenzofuran lignans 2-4 from caffeic
acid methyl ester 1a or ferulic acid methyl ester 1b has
been described before. Briefly, the monomeric precursors
are dimerized in the presence of silver oxide, generating
the dihydrobenzofuran skeleton with a 2,3-trans-configu-
ration (Scheme 1). The 4′-OH group of 2b is methylated
with methyl iodide in the presence of potassium carbonate
*
To whom correspondence should be addressed. Tel and fax: +32-3-
to yield 2c. The double bond in the C
hydrogenated in the presence of Pd-C, giving 3a and 3c.
LiAlH reduction of the ester functionalities to primary
alcohols yields 4a and 4c (Scheme 1). All compounds were
3
side chain is
8
20 27 09. E-mail: luc.pieters@ua.ac.be.
†
Department of Pharmaceutical Sciences, University of Antwerp.
Department of Pharmacy, University of Regensburg.
Department of Chemistry, University of Antwerp.
‡
4
§
1
0.1021/np0103968 CCC: $22.00
© 2002 American Chemical Society and American Society of Pharmacognosy
Published on Web 04/02/2002

Notes
J ournal of Natural Products, 2002, Vol. 65, No. 5 719
Ta ble 1. Antiangiogenic Activity of Compounds 2-4 in the
CAM Assay
60, F254, 0.2 mm) were used for TLC analysis. All products and
reagents were puchased from Acros, Belgium.
antiangiogenic scorea
Chiral separations were performed on a column (5 × 21 cm)
packed with 0.25 kg of Chiralpack AD with EtOH as the mobile
phase. Peak shaving and closed loop recycling techniques were
concentration
( sd
test
compound
(µg/pellet) (nmol/pellet)
(n ) no. of expt)
used during the process. Determination of their absolute
2
2
a (rac)
a (2R,3R)
20
20
5
20
5
20
20
20
20
20
20
51
51
13
51
13
48
47
47
51
60
53
1.1 ( 0.1 (n)2)
1.5 ( 0.1 (n)2)
0.8 ( 0.1 (n)3)
0.7 ( 0.2 (n)3)
0.2 ( 0.1 (n)2)
0.4 ( 0.1 (n)2)
0.3 ( 0.4 (n)2)
0.5 ( 0.1 (n)2)
0.5 ( 0.3 (n)2)
0.3 ( 0.4 (n)2)
0.6 ( 0.1 (n)2)
0.1 ( 0.1 (n>10)
configuration was based on their CD spectra.8 Analytical
HPLC using two independent systems was performed to check
the purity of the products. Column A: Alltech Econosil C8 (4.6
2
a (2S,3S)
×
250 mm); column B: BIO-RAD Bio-Sil C18 (4.6 × 150 mm).
Experimental details of the HPLC analysis are reported for
all test compounds in the following form: column; mobile
phase; retention time. The flow rate was 1 mL/min.
2
2
2
3
4
4
b (rac)
c (2S,3S)
c (2R,3R)
a (rac)
a (rac)
c (rac)
agarose pellet
blank)
â-1,4-galactan
sulfate
Meth yl cin n a m a tes (1a ,b) were prepared from a mixture
of the corresponding cinnamic acid (4 g) and Dowex 50 W ×
8200-400 (0.4 g) in 25 mL of absolute methanol. After heating
under reflux for 1 night the mixture was filtered and evapo-
rated under reduced pressure to afford the product as a solid
(100%), which was used without further purification.
(
50
2.5
1.4 ( 0.2 (n>10)
a : amorphous, mp 158 °C; 1H NMR, 13C NMR, DCI-MS
1
(
LuPS S5)
8
(
3
NH ).
a
1b: amorphous, mp 148 °C; H NMR, 13C NMR, DCI-MS
1
0
) no or weak effect, 1 ) medium effect, 2 ) strong effect.
8
(
3
NH ).
Meth yl (E)-3-[2-(3,4-dih ydr oxyh en yl)-7-h ydr oxy-3-m eth -
oxyca r bon yl-2,3-d ih yd r o-1-ben zofu r a n -5-yl]p r op -2-en o-
obtained as racemic mixtures. Compounds 2a and 2c were
resolved into their enantiomers by preparative-scale chiral
HPLC.8
a te (2a ) was prepared according to the method of Lemi e` re et
10
al. using 1.905 g (9.8 mmol) of methyl caffeate, 826 g (3.5
The antiangiogenic activity of the test compounds is
listed in Table 1. The dihydrobenzofuran lignans were
tested at a dose of 20 µg/pellet, corresponding to about 50
nmol/pellet, because at a dose of 40 µg some of the lignans
showed a toxic effect. Only 2a , the dimerization product of
caffeic acid methyl ester, showed an antiangiogenic score
of more than 1. When evaluating both enantiomers of 2a
separately, it turned out that especially the (2R,3R)-isomer
exhibited a pronounced antiangiogenic activity, with a score
of 1.5 ( 0.1. However, also a membrane-irritating effect
mmol) of silver(I) oxide, 40 mL of anhydrous benzene, and 20
mL of anhydrous acetone. The product was purified by column
chromatography (3.8 × 30 cm, silica gel 60, 0.040-0.063 mm)
with ethyl acetate-n-heptane, 1:1, as the eluent. After evapo-
ration and lyophilization a white foam was obtained (33%):
1
13
8
3
amorphous, mp 159 °C; H NMR, C NMR, DCI-MS (NH ).
Chiral HPLC analysis: Chiralpak AD, length 5 cm; n-hexane-
ethanol, 60:40. CD spectra of both enantiomers [2a -(2R,3R)
8
and 2a -(2S,3S)] have been recorded and published before.
Anal. (C20
18 6
H O ) C, H.
Meth yl (E)-3-[2-(4-h yd r oxy-3-m eth oxyp h en yl)-7-m eth -
oxy-3-m eth oxyca r bon yl-2,3-d ih yd r o-1-ben zofu r a n -5-yl]-
p r op -2-en oa te (2b) was prepared according to the method
of Lemi e` re et al.10 using an improved workup. After evapora-
tion the residual brown oil was dissolved in methanol. The
solution was left to stand overnight. Crystals (white, mp 151
°C) were formed, filtered, and washed with cold methanol
(about 50%) was observed at 20 µg/pellet. Compound 2a -
(
2R,3R) still showed antiangiogenic properties (score 0.8)
at 5 µg/pellet. Methylation and reduction of the parent
compound reduced the antiangiogenic activity, as observed
before for the antimitotic activity and the inhibition of
tubulin polymerization.8 Indeed, in the same series of
1
13
8
(
50%): H NMR, C NMR, DCI-MS (NH
column A, methanol-water; 65:35; 5.01 min; column B,
methanol-water, 50:20; 4.93 min. Anal. (C22 ) C, H.
2S,3S)- a n d (2R,3R)-Met h yl (E)-3-[2-(3,4-d im et h oxy-
3
). HPLC analysis:
dihydrobenzofuran lignans, also 2a -(2R,3R) was the most
potent antimitotic compound and inhibitor of tubulin
polymerization. Also other antimitotic agents and inhibi-
tors of tubulin polymerization, such as colchicine or vin-
22 8
H O
(
p h en yl)-7-m et h oxy-3-m et h oxyca r b on yl-2,3-d ih yd r o-1-
2
cristine, inhibited angiogenesis in the CAM-assay. How-
ben zofu r a n -5-yl]p r op -2-en oa te (2c) were prepared using
ever, they did not inhibit the FGF-2 (fibroblast growth
factor-2) or VEGF (vascular endothelial growth factor)
10
the method of Lemi e` re et al., including preparative scale
HPLC, using an improved workup. After the evaporation of
the reaction mixture, the resulting yellow oil was dissolved in
methanol. The solution was left to stand overnight. Crystals
induced corneal neovascularization in mice, in contrast to
9
2
-methoxyestradiol or paclitaxel, which also act on tubulin.
(
75%) were formed, filtered, and washed with cold methanol:
Since the proliferation of endothelial cells is involved in
angiogenesis, the antiangiogenic effect of antimitotic agents
may be related to their antiproliferative action on endo-
thelial cells. Dihydrobenzofuran 2a can be considered as a
new lead for antiangiogenic agents, which deserves further
exploration.
1
13
8
3
white crystals, mp 135 °C; H NMR, C NMR, DCI-MS (NH ).
HPLC analysis: column A, methanol-water, 65:35; 5.06 min;
column B, methanol-water, 50:20, 5.61 min. Anal. (C23
C, H.
24 8
H O )
Met h yl (E)-3-[2-(3,4-d ih yd r oxyp h en yl)-7-h yd r oxy-3-
m e t h oxyca r b on yl-2,3-d ih yd r o-1-b e n zofu r a n -5-yl]p r o-
p a n oa te (3a ) was prepared from 2a by dissolving 418 mg (1.18
mmol) of 2a in 30 mL of acetone, to which 220 mg of 5% Pd/C
was added. The mixture was put in a Parr apparatus under
Exp er im en ta l Section
Gen er a l Exp er im en ta l P r oced u r es. Melting points were
6
0 psi hydrogen pressure and was shaken for 20 min. After
1
determined on a B u¨ chi B-545 melting point apparatus.
H
filtration and evaporation the crude oil was purified by column
chromatography (3 × 20 cm, silica gel 60, 0.040-0.063 mm)
with ethyl acetate-n-heptane, 1:2, as the eluent, resulting in
1
3
NMR and C NMR spectra were measured on a Varian Unity
00 spectrometer. Additional HETCOR and long-range HET-
4
COR measurements to verify the proposed assignments were
performed on the same spectrometer. Chemical shifts are
reported in δ (ppm). DCI mass spectra were obtained on a
Ribermag R-10-10B mass spectrometer. Column chromatog-
raphy was performed on Merck silica gel 60, 0.040-0.063 mm,
1
13
8
3
a colorless oil (95%): H NMR, C NMR, DCI-MS (NH ).
HPLC analysis: column A, methanol-water, 65:35; 6.25 min;
column B, methanol-water, 60:40; 3.45 min.
Meth yl 3-[2-(3,4-d im eth oxyp h en yl)-7-m eth oxy-3-m eth -
oxyca r b on yl-2,3-d ih yd r o-1-b e n zofu r a n -5-yl]p r op a n -
oa te (3c) was prepared according to the method of Lemi e` re
2
30-400 mesh ASTM. Precoated silica gel plates (Kieselgel

7
20 J ournal of Natural Products, 2002, Vol. 65, No. 5
Notes
2
et al. The reaction was improved by using a Parr apparatus
at 60 psi hydrogen pressure for 20 min. Workup was performed
identically, resulting in a white powder (81%): mp 97-99 °C;
be applied to the CAM was taken by means of a micropipet
for viscous solutions. Therefore the agarose pellets do not have
a uniform size. The half-cone-shaped agarose pellets are fixed
because they slightly sink into the CAM. After 24 h the
antiangiogenic effect was measured after addition of cream
as a contrast fluid, by means of a stereomicroscope, by
observing the avascular zone surrounding the pellet. Anti-
angiogenic activity is expressed as a score where 0 ) no or
weak effect, 1 ) medium effect, and 2 ) strong effect (capillary
free zone is at least twice as large as the pellet). Also
membrane irritation and embryotoxicity can be evaluated.
â-1,4-Galactan sulfate (LuPS S5) with an average molecular
1
13
8
3
H NMR, C NMR, DCI-MS (NH ). HPLC analysis: column
A, methanol-water, 65:35; 4.50 min; column B, methanol-
water, 50:20; 4.26 min.
3
-[2-(3,4-Dih yd r oxyp h e n yl)-3-h yd r oxym e t h yl-7-h y-
dr oxy-2,3-dih ydr o-1-ben zofu r an -5-yl]pr opan -1-ol (4a) was
prepared from 3a according to the method of Lemi e` re et al.10
In the workup the crude oil was dissolved in 3 mL of ethyl
acetate and left to stand overnight at 0 °C, resulting in the
formation of white crystals (20%): mp 123 °C; H NMR,
NMR, DCI-MS (NH
1
13
C
8
11
3
). HPLC analysis: column A, methanol-
weight of 20 000 was used as positive control, and an agarose
water, 65:35; 3.01 min; column B, methanol-water, 60:40; 2.41
pellet as a blank.
min.
3
Ack n ow led gm en t. This work was financially supported
by the FWO (Fund for Scientific Research, Flanders, Belgium)
(Grant No. G.0119.96, and a postdoctoral fellowship for S.A.),
and by the Special Fund for Research of the University of
Antwerp (Concerted Research Project No. 99/3/34).
-[2-(3,4-Dim eth oxyp h en yl)-3-h yd r oxym eth yl-7-m eth -
oxy-2,3-d ih yd r o-1-ben zofu r a n -5-yl]p r op a n -1-ol (4c) was
prepared from 3c according to the method of Lemi e` re et al.
without further modification (56%). H NMR, C NMR, DCI-
MS (NH
5; 3.43 min; column B, methanol-water, 50:20; 2.01 min.
Ch or ioa lla n toic Mem br a n e (CAM) Assa y. This assay is
based upon the formation of a chorioallantoic membrane, in
which neovascularization takes place, in fertilized chicken eggs
at a certain stage of the development of the embryo. Agarose
pellets impregnated with the test compound are placed onto
the vascular membrane of opened eggs, and the influence on
1
0
1
13
8
3
). HPLC analysis: column A, methanol-water, 65:
3
Refer en ces a n d Notes
(1) Gibaldi, H. J . Clin. Pharmacol. 1998, 38, 898-903.
(
(
2) Paper, D. H. Planta Med. 1998, 64, 686-695.
3) Hayes, A. J .; Li, L. Y.; Lippman, M. E. Brit. Med. J . 1999, 318, 853-
8
56.
(4) Twardowski, P.; Gradishar, W. J . Curr. Opin. Oncol. 1997, 9, 584-
589.
7
(5) Wynendaele, W.; Van Oosterom, A. T.; Pawinski, A.; De Bruijn, E.
A.; Maes, R. A. Pharm. World Sci. 1998, 20, 225-235.
angiogenesis is evaluated. Twelve eggs were used per experi-
ment to test one compound at a given dose. The eggs were
fertilized at 37 °C and 80% relative humidity in ideal condi-
tions. The shells of the eggs were cleaned with 70% EtOH to
avoid infections (e.g., Salmonella). After 72 h, 8-10 mL of
albumin was removed with a syringe at the lower side of the
egg, and the hole was sealed with tape. Subsequently the
upper part of the shell was removed, and the eggs were covered
with a plastic film and incubated for another 72 h. At this point
of time, when the diameter of the CAM is between 1.8 and 2.6
cm, the pellets containing the test substances were placed on
the CAM. Test substances were dissolved or suspended in a
(
6) Pieters, L.; De Bruyne, T.; Claeys, M.; Vlietinck, A.; Calomme, M.;
Vanden Berghe, D. J . Nat. Prod. 1993, 56, 899-906.
(
7) Paper, D. H.; Vogl, H.; Franz, G. In Relevance of Tumor Models for
Anticancer Drug Development. Contributions to Oncology, Vol. 54;
Fiebig, H. H., Burger, A. M., Eds.; Karger-Verlag: Basel, 1999; pp
1
91-199.
(8) Pieters, L.; Van Dyck, S.; Gao, M.; Bai, R.; Hamel, E.; Vlietinck, A.;
Lemi e` re, G. J . Med. Chem. 1999, 42, 5475-5481.
9) Klauber, N.; Parangi, S.; Flynn, E.; Hamel, E.; D’Amato, R. J . Cancer
(
Res. 1997, 57, 81-86
(10) Lemi e` re, G.; Gao, M.; De Groot, A.; Dommisse, R.; Lepoivre, J .;
Pieters, L.; Buss, V. J . Chem. Soc., Perkin Trans. 1 1995, 1775-1779.
11) Hoffman, R.; Paper, D.; Donaldson, J .; Vogl, H. Br. J . Cancer 1996,
(
7
3, 1183-1186.
2
.5% agarose solution. After gel formation, the volume of
agarose gel corresponding to the dose of the test compound to
NP0103968