Synthesis and antiangiogenic activity of thioacetal artemisinin derivatives

Article abstract of DOI:10.1016/j.bmc.2004.05.013

Various thioacetal artemisinin derivatives can inhibit the angiogenesis and might be angiogenesis inhibitors. In particular, 10α- phenylthiodihydroartemisinins (5), 10β-benzenesulfonyl-9-epi- dihydroartemisinin (11) and 10α-mercaptodihydroartemisinin (13)

Full text of DOI:10.1016/j.bmc.2004.05.013

Bioorganic & Medicinal Chemistry 12 (2004) 3783–3790
Synthesis and antiangiogenic activity of thioacetal artemisinin
derivatives
Sangtae Oh,a In Howa Jeong,a Chan Mug Ahn,b Woon-Seob Shinc and Seokjoon Leed,*
aDepartment of Chemistry, Yonsei University, Wonju 220-710, South Korea
bDepartment of Basic Science, Wonju College of Medicine, Yonsei University, Wonju 220-710, South Korea
cDepartment of Microbiology, Kwandong University College of Medicine, Gangneung 210-701, South Korea
dDepartment of Basic Science, Kwandong University College of Medicine, Gangneung 210-701, South Korea
Received 9 April 2004; accepted 10 May 2004
Available online 7 June 2004
Abstract—Various thioacetal artemisinin derivatives can inhibit the angiogenesis and might be angiogenesis inhibitors. In particular,
10a-phenylthiodihydroartemisinins (5), 10b-benzenesulfonyl-9-epi-dihydroartemisinin (11) and 10a-mercaptodihydroartemisinin
(13) exhibit strong growth inhibition activity against HUVEC proliferation. Compound 11 have a good inhibitiory activity upon
HUVEC tube formation, and 5 and 11 show a strong inhibitory effect on angiogenesis using CAM assay at 5 lg/egg by 90%.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
and identification of the endogenous inhibitor,8 as well
as gene9 and antibody therapy.10 However, because of
bioavailability, biostability, and effectiveness, it is very
important to discover the antiangiogenic small mole-
cules that might be suitable as clinical therapies.
Angiogenesis is the formation of new vascular capillaries
from preexisting host vessels by various biological
stimulators.1 In normal system, except during wound
healing2 and embryonic development,3 an elaborate
balance between positive and negative regulators tightly
controls angiogenesis. On the other hands, in abnormal
conditions, angiogenesis occurs in the course of tumor
growth,4 diabetic retinopathy,5 and rheumatoid arthri-
tis.6 In particular, tumor angiogenesis plays a key role in
the growth, invasion, and metastasis of tumors. There-
fore, the control of angiogenesis may be a promising
therapeutic strategy for the related diseases.
Although angiogenesis passes through complex process,
basically, ECs’ function, such as the migration, differ-
entiation, tube formation of ECs, is the most important.
Therefore, in order to search for a novel small angio-
genesis inhibitor, the inhibitory effects of promising
molecules against the proliferation of human umbilical
vein endothelial cells (HUVEC) in response to the var-
ious growth factors should be tested.
Angiogenesis is a complex process: endothelial cells
(ECs) of existing blood vessels must degrade the sur-
rounding matrix, activated, and proliferating ECs mi-
grate into stroma of the tissues, and then form tubes.
Subsequently, these new vessels maturate to form a
vascular network.7
In our previous report,11 thioacetal artemisinin deriva-
tives (1) as shown in Figure 1 exhibit good growth
inhibition activity against HUVEC proliferation. How-
ever, their growth inhibition activity against HUVEC
certainly does not imply antiangiogenic activity in tumor
angiogenic model. Therefore, based on the further
screening, such as formation assay of HUVEC tube on
Matrigel and Chorioallantomic membrane (CAM) assay
as well as HUVEC proliferation assay, we must confirm
that the selected molecules from the thioacetal artemis-
inin derivatives might be antiangiogenic inhibitors.
Strategies for regulating angiogenesis have been carried
out mainly in molecular biology, such as the isolation
Keywords: Angiogenesis; Antiangiogenesis; Artemisinin; Thioacetal
artemisinin.
The natural sesquiterpene endoperoxide artemisinin (2),
which was isolated from Artemisia annua L.,12 has
become a potential lead compound in the development
* Corresponding author. Tel.: +82-33-649-7454; fax: +82-33-641-1074;
e-mail: sjlee@kwandong.ac.kr
0968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.05.013

3784
S. Oh et al. / Bioorg. Med. Chem. 12 (2004) 3783–3790
15
2. Results and discussion
H
4
5 H
2
6
3
O
O
O
7
8
5a
8a
1
As seen in Scheme 1, separable diastereomeric mixture
of 10a-(5) and 10b-phenylthiodihydroartemisinins (6)
were prepared by reacting known dihydroartemisinin
(3)18 with thiophenol (2 equiv) under the catalysis of
BF3Et2O (1 equiv) at room temperature during 10 min.19
Under this condition, the reaction of S-acetalization of 3
was highly stereoselectively as the ratio of a:b ¼ 10:1.
The stereochemistry of S-phenyl thioartemisinin 5 and 6
was determined by chemical shift of H-10 and coupling
constant between H-9 and H-10. Major product 5 is the
a-isomer as indicated by a chemical shift (4.74 ppm) and
a large coupling constant (J ¼ 11:0 Hz), while minor
product 6 is the b-isomer (d ¼ 5:6 ppm and J ¼ 5:3 Hz).
Unlike the O-acetalization of dihydroartemisinin (2)
using appropriate alcohols under the same condition,
the S-acetalization of 3 diastereoselectively produced an
a-anomer (5). The thioacetal products (5 and 6) were
transformed to produce 10a-(7) and 10b-benz-
enesulfonyldihydroartemisinins (8), respectively, by
using oxidation with H2O2/urea complex (UHP), trifluo-
roacetic anhydride (TFAA) and NaHCO3 in good
yields.20 In addition, desoxy-isomer of 5 (9) was formed
from the thioacetal reaction as the side product because
of the reductive property of thiol group, and successive
oxidation of 9 with H2O2/urea complex gave 10a-benz-
enesulfonyl-desoxydihydroartemisinins (10). It was
conformed by elemental analysis that 9 and 10 are des-
oxy artemisinin compounds.
14
O
O
O
12
H
H
9
O
O
10
O
2
n= 0, 2
16
n(O)S
R
1
H
O
H
O
O
O
O
O
H
H
O
O
OH
OR
3
4
Figure 1.
of antimalarial13 and recently anticancer agents.14 The
semisynthetic acetal-type artemisinin derivatives (4),
ether, and ester derivatives of trioxane lactol dihydro-
artemisinin (3), were developed for their higher anti-
malarial efficacy and are now widely used to treat
malarial patients (Fig. 1).15
Particularly, Posner et al. reported that various sulfone
endoperoxides have a highly antimalarial activity.16
Recently, Chen et al. reported that artemisinin (2), di-
hydroartemisinin (3) and artesunate have the antiangio-
genic activity as well as the antitumor activity on in vitro
models of angiogenesis.17
Dihydroartemisinin (3) with benzenesulfinic acid under
the condition shown in Scheme 1 was directly converted
10b-benzenesulfonyl-9-epi-dihydroartemisinin (11) in
75% yield. The stereochemistry of 9-epi-intermediate
Herein, based on the screening of formation assay of
HUVEC tube on Matrigel and Chorioallantomic
membrane (CAM) assay as well as growth inhibition
activity against HUVEC, we report that the details of
the synthesis of the thioacetal artemisinin derivatives
and their antiangiogenic activity, which have proved
that the thioacetal artemisinin derivatives might be
angiogenic inhibitors.
1
(11) synthesized in this reaction was determined by H
and 13C NMR of 9-epi-artemisinin series as reported by
21
ꢀ
ꢀ
Begue et al. and comparison with its diastereomeric
sulfone intermediates 7 and 8. The chemical shift of H-
10 and 16-methyl carbon at C-9 in 7 having natural
H
H
H
S
H
O
O
O
O
O
O
a)
+
O
O
O
O
O
H
H
H
O
O
O
H
O
S
S
OH
3
5
6
9
c)
b)
b)
b)
H
H
H
H
O
O
O
O
O
O
O
O
O
O
O
H
H
H
H
O
O
O
O
O2S
O2S
O2S
O2S
11
7
10
8
Scheme 1. Reagents and conditions: (a) Thiophenol (2 equiv), BF3Et2O (1 equiv), CH2Cl2, reflux, 10 min (95%, 5:6 ¼ 10:1); (b) UHP (3 equiv), TFAA
(3 equiv), NaHCO3 (5 equiv), CH3CN, )30 ꢁC (95% for 7 and 93% for 8); (c) Benzenesulfinic acid (2 equiv), BF3Et2O (1 equiv), CH2Cl2, rt (75%).

S. Oh et al. / Bioorg. Med. Chem. 12 (2004) 3783–3790
3785
stereochemistry were appeared at 4.41 ppm
(J ¼ 10:98 Hz) and 14.0 ppm, while its epimer (11) were
at 5.13 ppm (J ¼ 10:24 Hz) and 20.0 ppm, respectively,
which signals indicate all trans configuration and dia-
stereomeric relationshipbetween sulfonyl groupand 16-
methyl at C-9 in 7 and 11.
MTT colorimetric method.23 The results are listed in
Table 1.
The natural artemisinin (2) with endoperoxide ring and
desoxyartemisinin (17) with no peroxide did not show
inhibitory effect at the concentration of 50 lM, while
dihydroartemisinin (3) was more effective than 2 in
inhibiting the HUVEC growth (IC50 ¼ 8.91 lM). As the
Chen’s report, we could also make sure that acetal-type
derivatives might be a lead compound for use as an
antiangiogenic inhibitor.
The reaction dihydroartemisinin (3) with thiolacetate
under the catalysis of BF3Et2O shown in Scheme 2,
afforded a 10b-thioacetoxydihydroartemisinin (12) in
84% yield, and then by the basic catalytic deacetylation,
12 was transformed to 10a-mercaptodihydroartemisinin
(13), S-acetal analog of 3. Comparing the coupling
constant of H9 and H10 between 12 and 13 (J ¼ 5:13 Hz
for 12 and J ¼ 10:80 Hz for 13), we confirmed that the
stereochemistry of C-10 position was converted from b
to a caused by basic catalyst.18
On the basis of our elementary result, we assumed that
all acetal-type artemisinin derivatives (4) would have an
inhibitory effect on HUVEC growth, but, disappoint-
ingly, 10b-(14) and 10a-phenoxydihydroartemisinin (15)
with endoperoxide and aromatic phenyl group were less
active than 3.
To investigate the antiangiogenic activity of the struc-
ture activity relationshipbetween S-acetal and O-acetal
type artemisinin derivatives, we have synthesized the
two a-(14) and b-isomer (15) of C-10 phenoxy dihydro-
artemisinin and b-artemether (16) using known meth-
od.22 Desoxyartemisinin (17) was synthesized to
compare the angiogenesis property of peroxide group
(Fig. 2).
So, as the new synthetic approach, we next synthesized a
series of S-acetal compounds of 3 to determine the
inhibitory activity on HUVEC. 10a-Phenylthiodihyd-
roartemisinin (5) and 10b-phenylthiodihydroartemisinin
(6) with endoperoxide and thiophenoxy moiety showed
a very strong inhibitory activity with an IC50 ¼ 0.93 and
5.24 lM, respectively. Among the synthetic sulfonyl
compounds, such as 10a-(7) and 10b-benz-
enesulfonyldihydroartemisinin (8) and 10b-benz-
enesulfonyl-9-epi-dihydroartemisinin (11), the inhibitory
potency of 9-epimer (11) was over 7–10 times higher
than that of 7 and 8. It was very interesting that those
compounds bearing a same molecular structure but only
different C-9 stereochemistry showed a lot of difference
in inhibitory effect. Even if desoxy compounds (9 and
10) had the thioacetal functionality and b-artemether
(16) had the endoperoxide, they have no inhibitory
activity for lack of endoperoxide or aromatic functional
group.
The growth inhibition effect of thioacetal-type artemis-
inin derivatives, known acetal type artemisinin com-
pounds and some desoxy-artemisinin derivatives were
examined on a HUVEC proliferation assay using the
H
H
O
O
O
O
a)
b)
O
O
3
H
H
O
O
S
SH
O
In particular, a comparison between 10a-merca-
ptodihydroartemisinin (13) and dihydroartemisinin (3)
13
12
Scheme 2. Reagents and conditions: (a) Thiolacetic acid (1 equiv),
BF3Et2O (1 equiv), CH2Cl2, rt, 10 min (84%); (b) NaOEt (5 equiv),
EtOH, rt, 1 h (62%).
Table 1. HUVEC proliferation inhibition assay results for artemisinin
derivatives
Compounds
Growth inhibition IC50 (lM)
H
O
H
O
O
O
O
2
>50
O
3
8.91
0.93
5.24
12.77
18.21
>50
O
O
H
5
H
O
O
6
7
8
9
14
15
10
11
12
13
14
15
16
17
>50
1.74
4.56
1.29
33.81
31.03
>50
H
H
O
O
O
O
O
H
H
O
O
OEt
O
17
>50
16
IC50 was calculated from nonlinear regression by GraphPad Prism
software (r2 > 0:9).
Figure 2.

3786
S. Oh et al. / Bioorg. Med. Chem. 12 (2004) 3783–3790
suggests that the sulfur atom at the C-10 position was
critical for inhibition activity on HUVEC. According as
change oxygen to sulfur, the biological activity im-
proved much better as seven times.
Next, the activity to suppress the growth factor induced
tube formation by HUVEC on Matrigel was assessed at
the concentration of 10 lg/mL.24 As shown in Table 2,
artemisinin (2), dihydroartemisinin (3) and some desoxy
artemisinin derivatives (9, 10, and 17) showed no tube
formation inhibitory activity or very weak activity.
So when assuming the HUVEC proliferation inhibitory
effect listed in Table 1, it was postulated that the thio-
acetal dihydroartemisinin derivatives bearing sulfur
acetal linkage at C-10 position and aromatic functional
groupas well as endoperoxide ring showed significantly
efficient activity.
10a-Phenylthiodihydroartemisinin (5) and 10b-phenyl-
thiodihydroartemisinin (6) with high growth inhibitory
activity upon HUVEC showed a weak activity, but 10a-
(7) and 10b-benzenesulfonyldihydroartemisinin (8) with
relatively weak growth inhibition activity than 5 and 6
had
a moderate activity. Particularly, 10b-benz-
enesulfonyl-9-epi-dihydroartemisinin (11) with potent
growth inhibition activity effectively inhibited the tube
formation by 74% at the concentration of 10 lg/mL
(Fig. 3).
Table 2. HUVEC tube formation inhibition assay results for thio-
acetal artemisinin derivatives
Compounds
Inhibition percentage at
10 lg/mL (%)
When compared to thioacetal compounds (5 and 6) and
sulfonyl derivaitves (7, 8, and 11), sulfone substitution at
the C-10 position of tricyclic artemisinin improved the
tube formation inhibitory activity.
2
0
19
20
16
41
52
0
3
5
6
7
8
Interestingly, 10b-benzenesulfonyl-9-epi-desoxyartemis-
inin (10) and b-artemether (16), which have no inhibi-
tory activity against HUVEC growth, exhibited
moderate tube formation inhibitory activity by 32% and
48%, respectively. So, detail study on the structure
activity relationshipabout various functional groups is
needed.
9
10
11
12
13
14
15
16
17
32
74
0
7
3
16
48
0
Third, in vivo inhibitory effect of thioacetal artemisinin
derivatives on angiogenesis was examined using CAM
assay at the concentration of 5 lg/egg.25 As shown in
The inhibition percentages were determined with a image analysis
software, Image-Pro plus version 3.0.
Figure 3. Effect of selected thioacetal artemisinin derivatives on tube formation by human umbilical vein endothelial cells (HUVEC) on Matrigel.
(a) Control, (b) compound 8 (10 lg/mL), (c) compound 11 (10 lg/mL), (d) compound 16 (10 lg/mL).

S. Oh et al. / Bioorg. Med. Chem. 12 (2004) 3783–3790
Table 3. Inhibitory effect on angiogenesis using CAM assay at 5 lg/egg
3787
m, Ph), 5.35 (1H, s, H-12), 4.73 (1H, d, J ¼ 11:0 Hz, H-
10), 2.57 (1H, m, H-9), 2.37 (1H, td, J ¼ 14:5, 4.0 Hz, H-
4a), 1.47 (3H, s, H-14), 0.95 (3H, d, J ¼ 6:2 Hz, H-16),
0.90 (3H, d, J ¼ 7:1 Hz, H-15) ppm; 13C NMR (75 MHz,
CDCl3) d 132.9, 132.5, 128.6, 127.3, 104.3, 92.2, 83.5,
80.3, 51.7, 46.0, 37.4, 36.2, 34.1, 31.1, 26.0, 24.8, 21.4,
20.2, 15.1 ppm; GC/MSD (m=z) retention time 23.4
(min) 376 (Mþ), 330, 316, 287, 267, 225, 207, 138 (100);
Anal. Calcd for C21H28O4S: C, 66.99; H, 7.50; S, 8.52.
Compounds
Inhibited eggs/tested
eggs
Inhibition percentage
(%)
5
18/20
11/16
6/16
90
69
38
40
89
6
7
8
11
6/15
16/18
Found: C, 67.34; H, 7.73; S, 8.17; for 6: white crystal,
20
D
mp98–99 ꢁC; ½aŠ +276.7 (c 0.300, CHCl3); IR (KBr
Table 3, selected thioacetal molecules (5 and 6) and 9-
epi-sulfonyl artemisinin 11 strongly inhibited the for-
mation of new blood vessel on CAM. Particularly, 5 and
11 have a potent activity by 90%. The other analogues (7
and 8) were mildly potent.
1
pellet) mmax 2926, 1551, 1384, 1253, 1097, 1021 cmÀ1; H
NMR (300 MHz, CDCl3) d 7.54 (2H, dd, J ¼ 8:6,
1.5 Hz, Ph), 7.26 (3H, m, Ph), 5.74 (1H, s, H-12), 5.57
(1H, d, J ¼ 5:3 Hz, H-10), 3.12 (1H, m, H-9), 2.39 (1H,
td, J ¼ 14:3, 3.8 Hz, H-4a), 1.44 (3H, s, H-14), 1.05 (3H,
d, J ¼ 7:3 Hz, H-16), 0.98 (3H, d, J ¼ 6:4 Hz, H-15)
ppm; 13C NMR (75 MHz, CDCl3) d 136.9, 130.9, 128.8,
126.7, 104.2, 90.2, 88.3, 81.0, 52.6, 45.1, 37.2, 36.4, 34.4,
32.7, 26.0, 24.6, 24.3, 20.3, 15.0 ppm; GC/MSD (m=z)
retention time 23.8 (min) 376 (Mþ), 316, 287, 267, 225,
207, 138 (100); Anal. Calcd for C21H28O4S: C, 66.99; H,
7.50; S, 8.52. Found: C, 67.54; H, 7.73; S, 8.18.
In conclusion, with the various screening methods, such
as growth inhibition activity against HUVEC, forma-
tion assay of HUVEC tube on Matrigel and Chorio-
allantomic membrane (CAM) assay, we concluded that
the thioacetal deoxoartemisinin derivatives can inhibit
the angiogenesis and might be angiogenesis inhibitors.
3. Experimental
3.1. General
3.3. 10a-Benzenesulfonyldihydroartemisinin (7)
Trifluoroacetic anhydride (1.13 mL, 7.97 mmol) was
added to a stirred suspension of ureahydrogenperoxide
(750 mg, 7.97 mmol) in acetonitrile (50 mL) for 10 min at
room temperature. The solution was added dropwise to
a stirred suspension of 5 (1 g 2.66 mmol) and NaHCO3
(1.11 g 13.3 mmol) in acetonitrile (50 mL) at )40 ꢁC for
10 min. The suspension was stirred for 20 min, after the
solution was quenched with water (100 mL), extracted
with EtOAc (3 · 50 mL) and washed with brine (50 mL).
The organic layer was separated, dried with MgSO4,
filtered, and evaporated to dryness. The crude product
was purified by flash chromatography (hexane:EtOAc
Melting points were determined on Gallenkamp appa-
ratus and were not corrected. Specific rotations were
determined on optical activity AA-5 polarimeter at the
given temperatures. IR spectra were recorded as thin
films for solid and neat state for liquid on Mattson
1
FTIR spectrometer. H and 13C NMR spectra recorded
on Jeol Lambda 300 spectrometer at 300 and 75 MHz,
respectively, in the indicated solvent using TMS as
internal standard. Chemical shifts are expressed in ppm
(d) and coupling constants (J) in Hz. Mass spectrum
were determined on HP 5973 MSD system. Thin-layer
chromatography (TLC) has been performed on pre-
coated Merck silica gel 60 F254 plates. Elemental
analysis was carried out on CE instruments EA1110
elemental analyzer. All other reagents were commer-
cially available.
5:1) to give a desired product 7 (1.03 g, 95%). For 7;
20
D
white crystal, mp135–137 ꢁC; ½aŠ +62.5 (c 0.160,
CHCl3); IR (KBr pellet) mmax 3051, 2927, 1551, 1383,
1
1252, 1096, 1012 cmÀ1; H NMR (300 MHz, CDCl3) d
7.98 (2H, d, J ¼ 7:9 Hz, Ph), 7.63 (1H, t, J ¼ 7:1 Hz,
Ph), 7.50 (2H, t, J ¼ 7:9 Hz, Ph), 5.26 (1H, s, H-12), 4.42
(1H, d, J ¼ 11:0 Hz, H-10), 2.42 (1H, m), 2.26 (1H, td,
J ¼ 14:5, 4.0 Hz, H-4a), 1.35 (3H, s, H-14), 1.12 (3H, d,
J ¼ 7:0 Hz, H-16), 0.90 (3H, d, J ¼ 6:0 Hz, H-15) ppm;
13C NMR (75 MHz, CDCl3) d 135.9, 133.8, 130.0, 128.4,
104.3, 91.9, 90.7, 79.6, 51.2, 46.5, 37.3, 35.9, 33.9, 28.3,
25.6, 24.6, 21.3, 20.0, 14.0 ppm; GC/MSD (m=z) reten-
tion time 25.8 (min) 408 (Mþ), 392, 364, 327, 251 (100),
233, 205; Anal. Calcd for C21H28O6S: C, 61.74; H, 6.91;
S, 7.85. Found: C, 61.81; H, 7.01; S, 7.40.
3.2. 10a-Phenylthiodihydroartemisinin (5) and 10b-phen-
ylthiodihydroartemisinin (6)
Thiophenol (698 mg, 6.34 mmol) and BF3Et2O (500 mg,
3.52 mmol) were added to a stirred solution of dihyd-
roartemisinin 2 (1 g, 3.52 mmol) in CH2Cl2 (50 mL) at
room temperature. The solution was stirred for 10 min,
after which it was diluted with CH2Cl2 (100 mL), wa-
shed with sat-NaHCO3 and brine. The organic layer was
separated, dried with MgSO4, filtered and evaporated to
dryness. The crude product was purified by flash chro-
matography (hexane:EtOAc 15:1) to give 1.13 g (85%) of
3.4. 10b-Benzenesulfonyldihydroartemisinin (8)
Same method as shown in the synthesis of 7. Yield 93%.
20
D
5 and 140 mg (11%) of 6 for 5: white crystal, mp115–
For 8; white crystal, mp94–96 ꢁC; ½aŠ +145.8 (c 0.240,
20
D
117 ꢁC; ½aŠ +108.3 (c 0.120, CHCl3); IR (KBr pellet)
CHCl3); IR (KBr pellet) mmax 3054, 2986, 1422, 1384,
1
1
mmax 2922, 1592, 1384, 1042 cmÀ1; H NMR (300 MHz,
1265, 1048 cmÀ1; H NMR (300 MHz, CDCl3) d 7.94
CDCl3) d 7.68 (2H, dd, J ¼ 8:3, 1.8 Hz, Ph), 7.26 (3H,
(2H, d, J ¼ 6:9 Hz, Ph), 7.60 (1H, t, J ¼ 7:2 Hz, Ph),

3788
S. Oh et al. / Bioorg. Med. Chem. 12 (2004) 3783–3790
7.52 (2H, t, J ¼ 7:5 Hz, Ph), 5.98 (1H, s, H-12), 5.01
(1H, d, J ¼ 6:3 Hz, H-10), 3.17 (1H, m), 2.28 (1H, td,
J ¼ 14:4, 3.9 Hz, H-4a), 1.35 (3H, d, J ¼ 7:7 Hz, H-14),
1.18 (3H, s, H-16), 0.94 (3H, d, J ¼ 4:5 Hz, H-15) ppm;
13C NMR (75 MHz, CDCl3) d 133.4, 129.1, 128.8, 128.7,
103.8, 92.0, 90.3, 80.9, 52.1, 43.8, 37.0, 36.1, 34.1, 31.6,
25.5, 24.5, 23.4, 20.1, 13.2 ppm; GC/MSD (m=z) reten-
tion time 26.5 (min) 408 (Mþ), 392, 364, 327, 251 (100),
233, 205; Anal. Calcd for C21H28O6S: C, 61.74; H, 6.91;
S, 7.85. Found: C, 62.02; H, 7.26; S, 7.55.
(30 mL) at room temperature. The solution was stirred
for 10 min, after which it was diluted with CH2Cl2
(50 mL), washed with sat NaHCO3 and brine. The or-
ganic layer was separated, dried with MgSO4, filtered,
and evaporated to dryness. The crude product was
purified by flash chromatography (hexane:EtOAc ¼ 5:1)
to give a product 11 (1.07 g, 75%). For 11; white crystal,
20
mp107–108 ꢁC; ½aŠ +71.4 (c 0.210, CHCl3); IR (KBr
D
pellet) mmax 2926, 1552, 1384, 1252, 1210, 1101,
1
1020 cmÀ1 ; H NMR (300 MHz, CDCl3) d 7.99 (2H, d,
J ¼ 7:0 Hz, Ph), 7.64 (1H, t, J ¼ 7:3 Hz, Ph), 7.54 (2H, t,
J ¼ 7:7 Hz, Ph), 5.41 (1H, s, H-12), 5.14 (1H, d,
J ¼ 10:3 Hz, H-10), 1.43 (3H, d, J ¼ 7:0 Hz, H-16), 1.05
(3H, s, H-14), 0.92 (3H, d, J ¼ 5:7 Hz, H-15) ppm; 13C
NMR (75 MHz, CDCl3) d 137.5, 133.4, 129.2, 128.5,
102.3, 90.6, 89.7, 81.9, 50.7, 48.5, 37.2, 35.9, 34.8, 33.8,
30.9, 25.1, 24.6, 20.7, 19.7 ppm; GC/MSD (m=z) reten-
tion time 27.2 (min) 408 (Mþ), 390, 375, 350, 221, 163
(100); Anal. Calcd for C21H28O6S: C, 61.74; H, 6.91; S,
7.85. Found: C, 61.57; H, 7.25; S, 7.65.
3.5. 10b-Phenylthio-desoxy-9-epi-dihydroartemisinin (9)
Thiophenol (350 mg, 3.17 mmol) and BF3Et2O (250 mg,
1.25 mmol) were added to a stirred solution of dihydro-
artemisinin 2 (500 mg, 1.25 mmol) in CH2Cl2 (30 mL) at
room temperature. The solution was stirred for 12 h,
after which it was diluted with CH2Cl2 (50 mL), washed
with satd NaHCO3 and brine. The organic layer was
separated, dried with MgSO4, filtered and evaporated to
dryness. The crude product was purified by flash chro-
matography (hexane:EtOAc ¼ 20:1) to give a desoxy
product 9 (321 mg, 50%). For 9; white crystal, mp
3.8. 10b-Thioacetoxydihydroartemisinin (12)
102–104 ꢁC; ½aŠ20 )62.5 (c 0.064, CHCl3); IR (KBr pellet)
Thioacetic acid (320 mg, 4.2 mmol) and BF3Et2O
(300 mg, 2.11 mmol) were added to a stirred solution of
dihydroartemisinin 2 (600 mg, 2.11 mmol) in CH2Cl2
(30 mL) at room temperature. The solution was stirred
for 10 min, after which it was diluted with CH2Cl2
(50 mL), washed with satd NaHCO3 and brine. The
organic layer was separated, dried with MgSO4, filtered,
and evaporated to dryness. The crude product was
purified by flash chromatography (hexane:EtOAc ¼ 5:1)
D
mmax 3049, 2927, 1552, 1384, 1252, 1209, 1097,
1
1039 cmÀ1; H NMR (300 MHz, CDCl3) d 7.52 (2H, d,
J ¼ 8:5 Hz, Ph), 7.23 (1H, t, J ¼ 8:0 Hz, Ph), 7.18 (2H, t,
J ¼ 7:3 Hz, Ph), 5.45 (1H, s, H-12), 5.11 (1H,
d, J ¼ 10:5 Hz, H-10), 1.49 (3H, s, H-14), 1.20 (3H, d,
J ¼ 7:0 Hz, H-16), 0.89 (3H, d, J ¼ 5:7 Hz, H-15) ppm;
13C NMR (75 MHz, CDCl3) d 136.2, 128.7, 128.6, 126.0,
107.2, 97.5, 82.7, 82.4, 45.2, 43.7, 39.3, 35.1, 34.3, 32.3,
23.5, 22.0, 20.8, 18.6 ppm; GC/MSD (m=z) retention
time 22.3 (min) 360 (Mþ), 331, 314, 299, 281, 269, 251
(100); Anal. Calcd for C21H28O3S: C, 69.96; H, 7.83; S,
8.89. Found: C, 68.69; H, 7.77; S, 8.91.
to give 606 mg (84%) of thioacetoxy product 14. For
20
14; white crystal, mp114–119 ꢁC; ½aŠ +305.6 (c 0.180,
D
CHCl3); IR (KBr pellet) mmax 2933, 2873, 1707, 1550,
1382, 1252, 1058, 1009 cmÀ1
;
1H NMR (300 MHz,
CDCl3) d 6.12 (1H, d, J ¼ 5:1 Hz, H-10), 5.33 (1H, s, H-
12), 3.17 (1H, m, H-9), 2.37 (3H, s, –SCOCH3), 2.41–
2.31 (1H, m, H-4a), 1.35 (3H, s, H-14), 0.87 (3H, d,
J ¼ 6:0 Hz, H-16), 0.88 (3H, d, J ¼ 7:4 Hz, H-15) ppm;
13C NMR (75 MHz, CDCl3) d 193.20, 104.37, 89.34,
82.69, 80.46, 52.40, 44.80, 37.25, 36.12, 34.21, 31.25,
31.19, 25.90, 24.48, 23.54, 20.19, 14.19 ppm; GC/MSD
(m=z) retention time 21.3 (min) 324 (Mþ)16), 296, 282,
263, 237, 221, 180, 162 (100); Anal. Calcd for
C17H26O5S: C, 59.62; H, 7.65; S, 9.36. Found: C, 60.29;
H, 7.37; S, 9.48.
3.6. 10b-Benzenesulfonyl-desoxy-9-epi-dihydroartemi-
sinin (10)
Same method as shown in the synthesis of 7. Yield 84%.
20
D
For 10; white crystal, mp129–131 ꢁC; ½aŠ )86.5 (c
0.104, CHCl3); IR (KBr pellet) mmax 2927, 1551, 1384,
1
1251, 1101, 1010 cmÀ1; H NMR(300 MHz, CDCl3) d
7.93 (2H, d, J ¼ 7:1 Hz, Ph), 7.64 (1H, t, J ¼ 8:1 Hz,
Ph), 7.55 (2H, t, J ¼ 6:3 Hz, Ph), 5.30 (1H, s, H-12), 4.60
(1H, d, J ¼ 9:9 Hz, H-10), 1.40 (3H, s, H-14), 1.39 (3H,
d, J ¼ 6:5 Hz, H-16), 0.86 (3H, d, J ¼ 5:4 Hz, H-
15) ppm; 13C NMR (75 MHz, CDCl3) d 136.9, 133.7,
129.3, 128.7, 107.6, 96.7, 88.4, 82.7, 44.8, 44.7, 35.1,
34.1, 33.9, 32.9, 31.7, 23.6, 21.9, 21.6, 18.5 ppm; GC/
MSD (m=z) retention time 26.8 (min) 392 (Mþ), 374,
346, 331, 301, 283 (100); Anal. Calcd for C21H28O5S: C,
64.26; H, 7.19; S, 8.17. Found: C, 63.96; H, 7.08; S, 8.34.
3.9. 10a-Mercaptodihydroartemisinin (13)
Sodium ethoxide (300 mg 4.41 mmol) was added to a
stirred solution of 14 (300 mg 0.88 mmol) in ethanol
(20 mL). The suspension was stirred for 1 day at room
temperature. The solution was then added to aqueous
HCl (10 mL, 1 M), extracted with EtOAc (3 · 30 mL)
and washed with brine. The organic layer was separated,
dried with MgSO4, filtered, and evaporated to dryness.
The crude product was purified by flash chromatogra-
3.7. 10b-Benzenesulfonyl-9-epi-dihydroartemisinin (11)
Benzenesulfinic acid (500 mg, 3.5 mmol) and BF3Et2O
(250 mg, 1.76 mmol) were added to a stirred solution of
dihydroartemisinin 2 (500 mg, 1.76 mmol) in CH2Cl2
phy (hexane:EtOAc ¼ 5:2) to give a mercapto prod-
20
uct 15 (163 mg, 62%). For 15; colorless oil; ½aŠ )183.9
D
(c 0.018, CHCl3); IR (neat) mmax 2929, 2873, 1550, 1379,

S. Oh et al. / Bioorg. Med. Chem. 12 (2004) 3783–3790
3789
1
1242, 1036 cmÀ1; H NMR (300 MHz, CDCl3) d 5.29
(1H, s, H-12), 4.73 (1H, d, J ¼ 10:8 Hz, H-10), 2.71 (1H,
m, H-9), 2.36 (1H, td, J ¼ 14:3, 3.8 Hz, H-4a), 1.41(3H,
s, H-14), 0.99 (3H, d, J ¼ 7:1 Hz, H-16), 0.94 (3H, d,
J ¼ 6:0 Hz, H-15) ppm; 13C NMR (75 MHz, CDCl3) d
104.17, 92.57, 86.98, 80.51, 51.82, 46.12, 37.34, 36.28,
33.96, 31.88, 26.07, 24.68, 21.89, 20.27, 14.87 ppm; GC/
MSD (m=z) retention time 15.6 (min) 300 (Mþ), 268,
250, 236, 211, 178 (100); Anal. Calcd for C15H24O4S: C,
59.97; H, 8.05; S, 10.67. Found: C, 60.37; H, 7.87; S,
9.84.
Acknowledgements
We are grateful for the financial support from research
grants (R05-2002-000-00961-0) from the Korea Science
and Engineering Foundation.
References and notes
1. Folkman, J. New Engl. J. Med. 1971, 285, 1182.
2. Arnold, F.; West, D. C. Pharmaco. Therapeut. 1991, 52,
407.
3. Hyder, S. M.; Stancel, G. M. Mol. Endocrinol. 1999, 13,
806.
4. Hanahan, D. Nat. Med. 1998, 5, 1359.
5. Klein, R.; Klein, B. E.; Moss, S. E. Diabetes Metab. Rev.
1989, 5, 559.
6. Koch, A. E. Arthritis Rheum. 1998, 41, 951.
7. (a) Hanahan, D.; Folkman, J. Cell 1996, 86, 353; (b)
Conway, E. M.; Collen, D.; Carmeliet, P. Cardiovasc. Res.
2001, 49, 507.
8. (a) O’Reilly, M. S.; Boehm, T.; Shing, Y.; Fukai, N.;
Vasios, G.; Lane, W. S.; Flynn, E.; Birkhead, J. R.; Olsen,
B. R.; Folkman, J. Cell 1997, 88, 277; (b) Black, W. R.;
Agner, R. C. Nature 1998, 391, 450.
3.10. Assay method for HUVEC proliferation inhibition
assay
HUVEC in EGM-2 medium containing the growth
factor (Cambrex, Walkersville, MD) was plated in a 48-
well plate (10,000 cells/well) for 4 h to allow cells to
adhere to the plate. The test compounds, diluted in the
medium, were added to the appropriate wells and
incubated for 72 h at 37 in a 5% CO2 humidified atmo-
sphere. After incubation, the viability of the HUVEC
was assayed using a 3-(4,5-dimethylthiazol-2-y1)-2,5-
diphenyltetrazolium bromide (MTT) colorimetric pro-
liferation assay.
9. Kong, H. L.; Crystal, R. G. J. Nat. Cancer Inst. 1998, 90,
273.
10. Hicklin, D. J.; Witte, L.; Zhu, Z.; Liao, F.; Wu, Y.; Li, Y.;
Bohlen, P. Drug Discovery Today 2001, 6, 517.
11. Oh, S.; Jeong, I. W.; Shin, W.-S.; Lee, S. Bioorg. Med.
Chem. Lett., 2003, 13, 3665.
3.11. Assay method for HUVEC tube formation inhibition
assay
12. Klayman, D. L. Science 1985, 228, 1049.
13. (a) Luo, X.-D.; Shen, C.-C. Med. Res. Rev. 1987, 7, 29; (b)
Jung, M. Curr. Med. Chem. 1994, 1, 35; (c) Haynes, R. K.;
Vonwiller, S. C. Acc. Chem. Res. 1997, 30, 73; (d) Vroman,
J. A.; Alvim-Gaston, M.; Avery, M. A. Curr. Pharm.
Design 1999, 5, 101.
Formation of HUVEC tube on Matrigel (BD Bio-
sciences, Bedford, MA) was assayed in a 96-well plate.
HUVEC was plated (5000 cells/well) on 50 lL of Ma-
trigel solidified, and incubated for 16 h at 37 in a 5%
CO2 humidified atmosphere. After incubation, the cells
were stained with 0.1% toluidin blue in 4% parafor-
maldehyde, and photographed with microscope (Olym-
pus, Tokyo, Japan). The total tube length in three
randomly chosen microscopic field per well was deter-
mined with a image analysis software, Image-Pro plus
version 3.0 (Media Cybernetics, MD, USA).
14. (a) Beekman, A. C.; Barentsen, A. R. W.; Woerdenbag,
H. J.; Uden, W. V.; Pras, N.; Konings, A. W. T.; El-
€
Feraly, F. S.; Galal, A. M.; Wikstrom, H. V. J. Nat. Prod.
1997, 60, 325; (b) Jung, M. Bioorg. Med. Chem. Lett. 1997,
7, 1091; (c) Posner, G. H.; Ploypradith, P.; Parker, M. H.;
O’Dowd, H.; Woo, S.-H.; Northrop, J.; Krasavin, M.;
Dolan, P.; Kensler, T. W.; Xie, S.; Shapiro, T. A. J. Med.
Chem. 1999, 42, 4275; (d) Li, Y.; Shan, F.; Wu, J.-M.; Wu,
G.-S.; Ding, J.; Xiao, D.; Yang, W.-Y.; Atassi, G.;
ꢀ
Leonce, S.; Caignard, D.-H.; Renard, P. Bioorg. Med.
3.12. Assay method for CAM assay
Chem. Lett. 2001, 11, 5; (e) Posner, G. H.; Northrop, J.;
Paik, I.-H.; Borstnik, K.; Dolan, P.; Kensler, T. W.; Xie,
S.; Shapiro, T. A. Bioorg. Med. Chem. 2002, 10, 227; (f)
Jung, M.; Lee, S.; Ham, J.; Lee, K.; Kim, H.; Kim, S. K.
J. Med. Chem. 2003, 46, 987; (g) Posner, G. H.; Paik, I.-H.;
Sur, S.; McRiner, A. J.; Borstnik, K.; Xie, S.; Shapiro,
T. A. J. Med. Chem. 2003, 46, 1060.
Fertilized chicken eggs were incubated horizontally at
37 ꢁC in a humidified incubator. After 3 days incubation,
about 2 mL of albumin was aspirated from the eggs with
18-gauge needle. On day 6, the shell covering the air sac
was punched out and membrane of air sac was removed
by forceps. Methanol solution of the compounds were
applied on polycarbonate membrane (Nuclepore, CA,
USA) coated with 1% methycellulose. Then, the disks
were placed on the CAM, and the eggs were sealed with
adhesive tape. After reincubation for 2 days, an appro-
priate volume of 20% fat emulsion was injected into the
chorioallantois to improve visualization of the vascular
network. Inhibition of angiogenesis was determined by
measuring the avascular zone in CAM. A positive anti-
angiogenic response was assessed as an avascular zone
of 3 mm in diameter and results were expressed as fre-
quency (%) of avascular zone.
15. (a) Li, Y.; Yu, P.-L.; Chen, Y.-X.; Li, L.-Q.; Gai, Y.-Z.;
Wang, D.-S.; Zheng, Y.-P. Kexue Tongbao 1979, 24, 667;
(b) Brossi, A.; Venugopalan, B.; Gerpe, L.; Yeh, H. J. C.;
Flippen-Anderson, J. L.; Buchs, P.; Luo, X. D.; Milhous,
W.; Peters, W. J. Med. Chem. 1988, 31, 645.
16. (a) Posner, G. H.; O’Dowd, H.; Caferro, T.; Cumming,
J. N.; Ploypradith, P.; Xie, S.; Shapiro, T. A. Tetrahedron
Lett. 1998, 39, 2273; (b) Bachi, M. D.; Korshin, E. E.;
Ploypradith, P.; Cumming, J. N.; Xie, J.; Shapiro, T. A.
Bioorg. Med. Chem. Lett. 1998, 8, 903.
17. (a) Chen, H.-H.; Zhou, H.-J.; Fang, X. Pharmcol. Res.
2003, 48, 231; (b) Chen, H.-H.; You L.-L.; Shi, M.-F.
Pharmcol. Res., in press.

3790
S. Oh et al. / Bioorg. Med. Chem. 12 (2004) 3783–3790
ꢀ
Ourevitch, M.; Bonnet-Deipon, D.; Begue, J. P. J. Org.
Chem. 2001, 66, 7858.
ꢀ ꢀ
18. Lin, A. J.; Klayman, D. L.; Milhous, W. K. J. Med. Chem.
1987, 30, 2147.
19. Venugopalan, B.; Karnik, P. J.; Bapat, C. P.; Chatterjee, D.
K.; Iyer, N.; Lepcha, D. Eur. J. Med. Chem. 1995, 30, 697.
20. (a) Varma, R. S.; Naicker, K. P. Org. Lett. 1999, 1, 189;
(b) Caron, S.; Do, N. M.; Sieser, J. E. Tetrahedron Lett.
2000, 41, 2299.
22. O’Neill, P. M.; Miller, A.; Ward, S. A.; Park, B. K.;
Scheinmann, F.; Stachulski, A. V. Tetrahedron Lett. 1999,
40, 9129.
23. Mosmann, T. J. Immunol. Meth. 1983, 65, 55.
24. Kubota, Y.; Kleinman, H. K.; Martin, G. R.; Lawley,
T. J. J. Cell. Biol. 1988, 107, 1589.
ꢀ
ꢀ
21. (a) Chorki, F.; Crousse, B.; Bonnet-Deipon, D.; Begue, J.
P.; Brigaud, T.; Portella, C. Tetrahedeon Lett. 2001, 42,
1487; (b) Chorki, F.; Grellepois, F.; Crousse, B.;
25. Ribatti, D.; Vacca, A.; Roncali, L.; Dammacco, F. In. J.
Dev. Biol. 1996, 40, 1189.