Molecular design, chemical synthesis and biological evaluation of quinoxaline-carbohydrate hybrids as novel and selective photo-induced DNA cleaving and cytotoxic agents

Article abstract of DOI:10.1016/S0040-4020(03)00820-2

The quinoxaline moiety in antitumor quinoxaline antibiotics cleaved double stranded DNA at the 5′ side guanine of the 5′-GG-3′ site upon irradiation with UV light with a long wavelength and without any additive. The quinoxaline-carbohydrate hybrid system

Full text of DOI:10.1016/S0040-4020(03)00820-2

Tetrahedron 59 (2003) 7057–7066
Molecular design, chemical synthesis and biological evaluation of
quinoxaline–carbohydrate hybrids as novel and selective
photo-induced DNA cleaving and cytotoxic agents
*
Kazunobu Toshima, Tomonori Kimura, Ryusuke Takano, Tomohiro Ozawa, Akiko Ariga,
Yutaka Shima, Kazuo Umezawa and Shuichi Matsumura
Department of Applied Chemistry, Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku,
Yokohama 223-8522, Japan
Received 2 April 2003; revised 29 April 2003; accepted 12 May 2003
This paper is dedicated to Professor K. C. Nicolaou on the occasion of his receiving the Tetrahedron Prize
Abstract—The quinoxaline moiety in antitumor quinoxaline antibiotics cleaved double stranded DNA at the 50 side guanine of the 50-GG-30
site upon irradiation with UV light with a long wavelength and without any additive. The quinoxaline–carbohydrate hybrid system was very
effective for the DNA cleavage. Furthermore, the quinoxaline–carbohydrate hybrids exhibited strong and selective cytotoxicity against
cancer cells with photoirradiation.
q 2003 Elsevier Science Ltd. All rights reserved.
1. Introduction
antitumor quinoxaline antibiotics5 such as echinomycin and
triostin A (Fig. 1). Although quinoxaline is known only as a
DNA intercalator,6 we expected that quinoxaline would be a
novel DNA photocleaving agent, because the conjugated
CvN bond in quinoxaline was expected to generate
the photo-excited3 (n–pp) and/or3 (p-pp) state(s) by
Studies of the interaction between the small molecules and
DNA, especially the effects of the structural characteristics
of the small molecules on the DNA interaction, are very
important for the design of DNA targeting new antitumor
drugs.1 In this context, the development of photochemical
DNA cleaving agents, which selectively cleave DNA by
irradiation with light of a specific wavelength under mild
conditions and without any additives such as metals and
reducing agents, is very interesting from chemical and
biological standpoints and offers considerable potential in
medicine especially in the post-genome era.2 Furthermore,
photodynamic therapy using a photosensitizing drug has
recently emerged as a promising modality against cancer
and allied diseases.3 In this paper, we discuss the molecular
design, chemical synthesis, DNA photocleaving property
and cytotoxicity of such novel and artificial light activatable
DNA cleaving and cytotoxic agents, that include the
quinoxaline–carbohydrate hybrids.4
2. Results and discussion
In our first approach to create such novel DNA photo-
cleaving molecules, we noted the quinoxaline structure,
because quinoxaline was found in DNA interactive and
Keywords: quinoxaline; carbohydrate; DNA photocleavage; cytotoxicity.
*
Corresponding author. Tel./fax: þ81-45-566-1576;
Figure 1. Molecular structures of quinoxaline antibiotics, echinomycin and
triostin A.
e-mail: toshima@applc.keio.ac.jp
0040–4020/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4020(03)00820-2

7058
K. Toshima et al. / Tetrahedron 59 (2003) 7057–7066
(365 nm) without any further additives, leading to the
nicked open circular DNA (Form II) (lanes 3 and 4 in
Figure 3). On the other hand, the quinoline derivatives 3 and
4 showed little DNA cleaving activity under similar
conditions (lanes 5 and 6 in Figure 3). These results clearly
demonstrate, for the first time, that the quinoxaline moiety,
present in certain antitumor antibiotics, is able to cleave
DNA upon irradiation with UV light with a long wavelength
without any additive and that the two nitrogen atoms in the
quinoxaline are essential for the DNA cleavage. Further-
more, it was found that the ester function at the C-2 position
of quinoxaline improved the DNA cleaving activity. To
further improve the DNA cleaving ability of the quinoxaline
derivatives, we designed the quinoxaline–carbohydrate
hybrids 5–9 (Fig. 4), because we have previously demon-
strated that a suitably deoxygenated amino sugar showed a
high affinity to DNA, and significantly enhances the
intercalating ability of certain intercalators.7 The quinoxa-
line–carbohydrate hybrids 5 and 6 are anomers of each
other, and 7–9 are bis(quinoxaline–carbohydrate) hybrids.
The syntheses of the representative quinoxaline–carbo-
hydrate hybrid 5 and bis(quinoxaline–carbohydrate) hybrid
7 are summarized in Schemes 1 and 2, respectively. The
synthesis of 5 commenced with the conversion of the known
methyl glycoside 108 into the phenyl thioglycoside 11 using
PhSH in the presence of BF3·Et2O. The protection of 11
with a t-butyldimethylsilyl (TBS) group gave the glycosyl
donor 12, which was subjected to glycosidation with
ethylene glycol (13) by Nicolaou’s method9 using N-bromo-
succinimide (NBS) to afford the glycosides 14 as a mixture
of the a- and b-anomers. 0After the conversion of the azide
group of 14 into the N,N -dimethylamino group using H2
and Pd–C in the presence of 35% HCHO, the resultant
anomers were easily separated by column chromatography
at this stage to isolate the a-anomer 15 and its b-anomer.
Esterification of the a-glycoside 15 with 2-quinoxaloyl
chloride (16) in the presence of Et3N furnished the hybrid
17. Finally, deprotection of the TBS group of 17 employing
HF/Py yielded the desired quinoxaline–carbohydrate
Figure 2. Quinoxaline and quinoline derivatives.
Figure 3. Photocleavage of supercoiled FX174 DNA. FX174 DNA
(50 mM per base pair) was incubated with various compounds in 20%
acetonitrile in Tris-HCl buffer (pH 7.5, 50 mM) at 258C for 1 h under
irradiation of the UV lamp (365 nm, 15 W) placed at 10 cm from the
mixture, and analyzed by gel electrophoresis (0.9% agarose gel, ethidium
bromide stain): lane 1, DNA alone; lane 2, DNA with UV; lanes 3–6,
compounds 1–4 (500 mM), respectively, following UV irradiation. Form I:
covalently closed supercoiled DNA, Form II: open circular DNA, and Form
III: linear DNA.
photoirradiation which could be capable of cleaving DNA
by H-abstraction, electron-transfer, and/or singlet oxygen
oxidation pathway(s).2 Therefore, we first examined the
photo-induced DNA cleaving activities of the quinoxaline
derivatives and the related compounds 1–4 (Fig. 2) using
supercoiled FX174 DNA (Form I). As is evident from
Figure 3, the quinoxaline derivatives 1 (500 mM) and 2
(500 mM) caused a significant single-strand scission of
DNA by photoirradiation using long wavelength UV light
Figure 4. Quinoxaline–carbohydrate hybrids.

K. Toshima et al. / Tetrahedron 59 (2003) 7057–7066
7059
hybrid 5. The b-anomer 6 was also obtained from the
b-anomer of 15 using a similar procedure. On the other
hand, the synthesis of 7 began with the glycosidation of 11
and ethylene glycol (13). Glycosidation of 11 and 13 using
NBS smoothly proceeded to give the glycoside 18 as a
mixture of the a- and b-anomers. The primary alcohol of 18
was regioselectively protected with a TBS group to afford
the a-anomer 19 and its b-anomer, which were easily
separated by column chromatography. The reduction of the
azide group of the a-glycoside 19, followed by protection of
the resultant amine 20 with p-nitrobenzenesulfonyl (Ns)10
and TBS groups, furnished 22 via 21. The coupling reaction
between 22 and p-dibromoxylene (23) in the presence of
K2CO3 smoothly proceeded to yield the dimer 24, the TBS
groups of which at the primary position were regio-
selectively deprotected to give the alcohol 25. Esterification
of 25 with 2-quinoxaloyl chloride (16), followed by the
deprotections of the Ns and TBS groups of the resultant 26,
gave the targeted bis(quinoxaline–carbohydrate) hybrid 7.
The other isomers 8 and 9 were also prepared using the a-
and b-anomers of 19 and the b-anomer of 19, respectively,
Scheme 1. Synthesis of 5. Reagents and conditions: (a) PhSH, BF3·Et2O,
08C, 50 min, 99%; (b) TBSCl, imidazole, CH2Cl2, 408C, 24 h, 94%;
(c) NBS, MS 4A, MeCN, 08C, 1 h, 79% (a/b¼1.4/1); (d) H2, Pd–C, 35%
HCHO aq. MeOH, 258C, 12 h, 81%; (e) Et3N, CH2Cl2, 258C, 15 h, 94%;
(f) HF/Py, Py, 258C, 12 h, 79%.
Scheme 2. Synthesis of 7. Reagents and conditions: (a) NBS, MS 4A, THF, 08C, 15 min, 80%; (b) TBSCl, imidazole, CH2Cl2, 258C, 10 min, 84% (a/b¼1/1.1);
(c) H2, Pd/CaCO3, EtOH, 258C, 16 h, 99%; (d) NsCl, K2CO3, CH2Cl2, 258C, 2 h, 72%; (e) TBSOTf, 2,6-lutidine, CH2Cl2, 408C, 16 h, 100%; (f) K2CO3, DMF,
258C, 3 h, 41%; (g) HF/Py, Py, 258C, 2 h, 95%; (h) Et3N, 258C, 20 min, 93%; (i) PhSH, K2CO3, DMF, 258C, 4 h, 54%; (j) HF/Py, Py, 258C, 16 h, 89%.

7060
K. Toshima et al. / Tetrahedron 59 (2003) 7057–7066
by a similar protocol. With all the designed quinoxaline–
carbohydrate hybrids 5–9 in hand, the photo-induced DNA
cleaving activities of these hybrids along with the reference
compound 1 were next assayed using supercoiled FX174
DNA in concentrations of 500–3 mM. Based on the results
shown in Figure 5, the quinoxaline–carbohydrate hybrids
5–9 caused effective DNA cleavage during irradiation with
a long wavelength UV light. It was confirmed that no DNA
cleavage by 1 and 5–9 was observed in the absence of light
(lanes 3 in Figure 5). Thus, the UV light functioned as a
trigger to initiate these quinoxaline derivatives for the DNA
strand scission. Furthermore, the DNA cleaving abilities of
the quinoxaline–carbohydrate hybrids 5 and 6 were found
to be stronger than that of 1, and the a-anomer 5 had a
higher DNA cleaving ability than the b-anomer 6 ((a)–(c)
in Figure 5). Thus, the DNA cleaving activity was
dependent on the configuration of the glycosidic bond of
the sugar moiety in the hybrids. These results strongly
suggest that the suitably deoxygenated amino sugar in these
hybrids works well as the DNA groove binder and
significantly enhances the intercalating ability of the
quinoxaline. Moreover, the bis(quinoxaline–carbohydrate)
hybrids 7–9 were the most effective photo-induced DNA
cleaving agents among them, and the DNA cleaving
activities of 9, 8 and 7 increased in that order ((d)–(f) in
Figure 5). Thus, the strongest DNA cleaving hybrid 7
cleaved DNA in concentrations over 3 mM, and caused a
100% DNA break at concentrations over 30 mM ((d) in
Figure 5). These results clearly show that the dimerization
of the quinoxaline–carbohydrate hybrids 5 and 6 via the
p-xylene linker is very effective for the DNA cleavage as
suggested from the observations by Chaires and
co-workers,11 and the configurations of the glycosidic
bonds of the sugars in the bis(quinoxaline–carbohydrate)
hybrids also influence their DNA cleaving activities. In
addition, it was interesting to note that the DNA
photocleaving abilities of these hybrids were much higher
than that of the natural quinoxaline antibiotic, echinomycin
((g) in Figure 5). These results demonstrate that the artificial
quinoxaline–carbohydrate hybrid is superior to the natural
product possessing quinoxaline as a DNA photocleaving
agent. The DNA cleaving site specificity of the representa-
tive quinoxaline derivatives 1, 5, 7 and 9 was next analyzed
according to the Sanger protocol.12 Since the Sanger
sequencing reactions result in base incorporation, cleavage
at the nucleotide N (sequencing) represents a cleaving site
by the agent or the Maxam–Gilbert reaction at Nþ1.13 The
results shown in Figure 6 clearly indicated that all the
quinoxaline derivatives selectively cleaved DNA at the 50
side guanine of the 50-GG-30 sites and the site-selective
DNA cleavage was enhanced upon treatment with hot
piperidine. Since both the free radical and a singlet oxygen
scavengers, dimethyl sulfoxide and 2,2,6,6-tetramethyl-
piperidine, did not inhibit the DNA cleavage, it is very
likely that an electron transfer from the electron-rich 50-GG-
30 site to the photo-excited quinoxaline is the initial step for
the photo-induced destruction of the guanine base on the 50
side of the 50-GG-30 sites.2,14 In addition, confirmation that
these quinoxaline derivatives bind to DNA with inter-
calation was obtained from a DNA unwinding assay using
supercoiled pBR322 DNA and topoisomerase I.15 With the
favorable results on the interactive and photocleaving
ability of the quinoxaline–carbohydrate hybrids against
Figure 5. Photocleavage of supercoiled FX174 DNA. A same protocol as
that mentioned in Figure 3 was carried out: (a), (b), (c), (d), (e), (f) and (g)
for the compounds 1, 5, 6, 7, 8, 9 and echinomycin, respectively: lane 1,
DNA alone; lane 2, DNA with UV; lane 3, DNAþcompound (500 mM)
without UV; lanes 4–9, compound (500), compound (300), compound
(100), compound (30), compound (10) and compound (3 mM), respectively,
following UV irradiation.

K. Toshima et al. / Tetrahedron 59 (2003) 7057–7066
7061
Figure 7. DNA ladders in HeLa S3 cells. HeLa S3 cells were treated with 5
(10 mM) and 6 (10 mM) with or without photoirradiation at 258C for 1 h
with a UV lamp (365 nm, 30 W) placed at 25 cm from the samples. After
cells were cultured for 72 h at 378C, the DNA was extracted from cells and
analyzed by gel electrohoresis in 1.5% agarose gel. The gel was stained
with ethidium bromide and photographed to observe fluoresence. lane 1,
DNA from HeLa S3 cells; lane 2, DNA from HeLa S3 cells with UV; lane
3, DNA from HeLa S3 cellsþ5 without UV; lane 4, DNA from HeLa S3
cellsþ6 without UV; lane 5, DNA from HeLa S3 cellsþ5 with UV; lane 6,
DNA from HeLa S3 cellsþ6 with UV.
MDA-MB-231 cells without photoirradiation were also
.100 mM, and those with photoirradiation were 6.0, 7.0,
7.0, 6.5 and 5.0 mM, respectively. These results indicate that
the quinoxaline–carbohydrate hybrids 5–9 are themselves
non-toxic while they show high cytotoxic activity with
photoirradiation. Furthermore, these results also demon-
strate that the DNA-cleaving activity induced by photo-
irradiation significantly affects the cytotoxicity of the
hybrids, and the life of cancer cells can be controlled by
treatment with an appropriate amount of the quinoxaline–
carbohydrate hybrid with or without the photoirradiation.
Figure 6. Autoradiogram of 12% polyacrylamide–8 M urea slab gel
electrophoresis for sequence analysis. The 50-end-labeled M13 mp18 DNA
at the primer site was cleaved by the compound at pH 7.5 and 258C for 1 h
under irradiation of the UV lamp (365 nm, 15 W) placed at 10 cm from the
mixture (bases 50–77 are shown): lanes A, G, C and T; Sanger A, G, C and
T reactions, respectively; lanes 1, 2, 3 and 4; the compounds 7, 9, 5 and 1
(500 mM), respectively, following UV irradiation: DNAs for lanes 1,4
were treated with hot piperidine prior to gel electrophoresis.
Table 1. Cytotoxicity of quinoxaline–carbohydrate hybrids against HeLa
S3 and MDA-MB-231 cells
Compounds
5
6
7
8
9
IC50(mM)
Hela S3
IC50(mM)
Without UV .100 .100 .100 .100 .100
With UV 1.5 1.3 1.0 4.0 4.0
Without UV .100 .100 .100 .100 .100
6.0 7.0 7.0 6.0 5.0
MDA-MB-231 With UV
the naked DNA, we further examined their DNA photo-
cleaving ability in cancer cells using the hybrids 5 and 6.
These results are summarized in Figure 7. After the HeLa S3
cells were exposed to each hybrid 5 (10 mM) or 6 (10 mM)
for 72 h with or without 1 h of photoirradiation, the DNA in
the cells was extracted and analyzed by gel electrophoresis.
It was found that the DNA was effectively cleaved by the
hybrids 5 and 6 under photoirradiation while 5 and 6 without
photoirradiation caused no DNA cleavage. These results
indicate that the quinoxaline–carbohydrate hybrid cleaves
DNA only with photoirradiation even in the cells. These
phenomena are quite consistent with those observed using
naked DNA. The cytotoxicity of the hybrids 5–7 was next
examined using HeLa S3 and MDA-MB-231 cells exposed
to each agent for 72 h with or without 1 h of photoirradiation
(Table 1).16 The IC50 values of 5–9 against the HeLa S3
cells without photoirradiation were .100 mM, and those
with photoirradiation were 1.5, 1.3, 1.0, 4.0 and 4.0 mM,
respectively. In addition, the IC50 values of 5–9 against the
3. Conclusions
In summary, the present study demonstrates not only the
molecular design and chemical synthesis of novel quinoxa-
line–carbohydrate hybrids, but also their DNA photo-
cleavage and cytotoxic profiles. The described chemistry
and biological evaluation provided significant information
about the molecular design of novel and selective DNA
photocleaving and cytotoxic agents.
4. Experimental
4.1. General synthetic methods
Melting points are uncorrected. 1H NMR spectra were

7062
K. Toshima et al. / Tetrahedron 59 (2003) 7057–7066
measured in CDCl3 using TMS as internal standard unless
otherwise noted. Silica gel TLC and column chromato-
graphy were performed on Merck TLC 60F-254 (0.25 mm)
and Merck Kieselgel 60 or Fuji-Davison BW-820MH,
respectively. Air- and/or moisture-sensitive reactions were
carried out under an atmosphere of argon with oven-dried
glassware. In general, organic solvents were purified and
dried by the appropriate procedure, and evaporation and
concentration were carried out under reduced pressure
below 308C, unless otherwise noted.
MeCN (12.6 mL) was added NBS (0.336 g, 1.89 mmol) at
08C. The reaction mixture was stirred at 08C for 1 h and
poured into ice-cold 10% NaHSO4 aq. (20 mL). The
resultant mixture was extracted with EtOAc (20 mL£3)
and the extracts were washed with brine (20 mL), dried over
anhydrous Na2SO4 and concentrated in vacuo. Purification
of the residue by column chromatography (20 g of silica gel,
1/1 n-hexane/EtOAc) gave 14 (0.325 g, 79%, a/b¼1.4/1) as
a colorless syrup. Rf 0.46 and 0.47 (4/1 n-hexane/EtOAc);
1H NMR d 0.11 (3/5£6H, s), 0.20 (2/5£6H, s), 0.91
(2/5£9H, s), 0.92 (3/5£9H, s), 1.23 (3/5£3H, d, J¼6.2 Hz),
1.28 (2/5£3H, d, J¼6.0 Hz), 1.62–1.80 (1H, m), 2.14–2.33
(8/5H, m), 2.74 (2/5£1H, br t, J¼6.0 Hz), 3.07 (3/5£1H, dd,
J¼9.0, 9.0 Hz), 3.08 (2/5£1H, dd, J¼9.0, 9.0 Hz), 3.27–
3.38 (2/5£2H, m), 3.53–3.85 (26/5H, m), 4.52 (2/5£1H,
dd, J¼9.6, 1.8 Hz), 4.86 (3/5£1H, dd, J¼3.6, 0.4 Hz);
HRMS (EI) m/z 331.1949 (331.1927 calcd for
C14H29N3O4Si, Mþ).
4.1.1. Phenyl 3-Azide-2,3,6-trideoxy-1-thio-D-arabino-
hexopyranoside (11). To a stirred solution of 10 (0.120 g,
0.641 mmol) in dry PhSH (0.36 mL) was added dropwise
BF3·Et2O (0.122 mL, 0.961 mmol) at 08C. The reaction
mixture was stirred at 08C for 50 min before the addition of
Et3N (1.15 mL). The resultant mixture was concentrated in
vacuo. Purification of the residue by column chromato-
graphy (58 g of silica gel, 5/2 n-hexane/EtOAc) gave 11
(0.168 g, 99%) as a pale yellow syrup. Rf 0.35 (3/1
4.1.4. 2-Hydroxyethyl 4-O-tert-Butyldimethylsilyl-3-
N,N0-dimethylamino-2,3,6-trideoxy-a-D-arabino-hexo-
pyranoside (15). To a solution of 14 (0.303 g, 0.914 mol)
and 35% HCHO aq. (1.57 mL, 18.3 mmol) in MeOH
(30.3 mL) was added 10% Pd–C (0.152 g). After the
mixture was vigorously stirred at 258C for 12 h under H2,
the mixture was filtered and the catalyst was washed with
MeOH. The combined filtrate and washings were concen-
trated in vacuo. Purification of the residue by column
chromatography (40 g of silica gel, 1/1 n-hexane/EtOAc)
gave 15 (0.145 g, 47%) as a colorless syrup and its
b-anomer (0.103 g, 34%) as a colorless syrup.
n-hexane/EtOAc); 1H NMR
d
1.31 (4/5£3H, d,
J¼6.0 Hz), 1.37 (1/5£3H, d, J¼6.0 Hz), 1.81 (1/5£1H,
ddd, J¼12.0, 12.0, 12.0 Hz), 2.12 (4/5£1H, ddd, J¼13.8,
12.6, 6.0 Hz), 2.21 (1H, d, J¼4.0 Hz), 2.37 (1/5£1H, ddd,
J¼12.0, 5.2, 2.0 Hz), 2.38 (4/5£1H, ddd, J¼13.8, 5.0,
1.0 Hz), 3.16 (1/5£1H, ddd, J¼9.0, 9.0, 4.0 Hz), 3.19
(4/5£1H, ddd, J¼9.0, 9.0, 4.0 Hz), 3.38 (1/5£1H, dq,
J¼9.0, 6.0 Hz), 3.48 (1/5£1H, ddd, J¼12.0, 9.0, 5.2 Hz),
3.77 (4/5£1H, ddd, J¼12.6, 9.0, 5.0 Hz), 4.20 (4/5£1H, dq,
J¼9.0, 6.0 Hz), 4.48 (1/5£1H, dd, J¼12.0, 2,0 Hz), 5.57
(4/5£1H, dd, J¼6.0, 1.0 Hz), 7.24–7.34 (3H, m), 7.42–7.52
(2H, m); HRMS (EI) m/z 265.0883 (265.0885 calcd for
C12H15N3O2S, Mþ).
Compound 15. Rf 0.45 (1/1 n-hexane/EtOAc);
1
[a]2D8¼þ49.38 (c 1.24, CHCl3); H NMR d 0.04 (3H, s),
4.1.2. Phenyl 3-Azide-4-O-tert-butyldimethylsilyl-2,3,6-
trideoxy-1-thio-D-arabino-hexopyranoside (12). To a
stirred solution of 11 (1.12 g, 4.22 mmol) in dry CH2Cl2
(22.4 mL) were added imidazole (1.43 g, 21.1 mmol)
followed by TBSCl (2.11 g, 14.0 mmol) at 08C. The
reaction mixture was stirred at 408C for 24 h before
the addition of cold saturated aqueous NH4Cl (30 mL).
The resultant mixture was extracted with EtOAc (30 mL£3)
and the extracts were washed with brine (30 mL), dried over
anhydrous Na2SO4 and concentrated in vacuo. Purification
of the residue by column chromatography (64 g of silica gel,
10/1 n-hexane/EtOAc) gave 12 (1.51 g, 94%) as a pale
0.07 (3H, s), 0.86 (9H, s), 1.25 (3H, d, J¼6.0 Hz), 1.59 (1H,
ddd, J¼13.0, 13.0, 3.9 Hz), 1.90 (1H, br dd, J¼13.0,
4.4 Hz), 2.04 (1H, 6H, s), 2.76 (1H, ddd, J¼13.0, 9.0,
4.4 Hz), 3.14 (1H, dd, J¼9.0, 9.0 Hz), 3.55–3.75 (5H, m),
4.88 (1H, br d, J¼3.9 Hz); HRMS (EI) m/z 333.2319
(333.2335 calcd for C16H35NO4Si, Mþ). The b-anomer of
15: Rf 0.50 (1/1 n-hexane/EtOAc); [a]2D8¼219.18 (c 3.15,
CHCl3); 1H NMR d 0.04 (3H, s), 0.06 (3H, s), 0.86 (9H, s),
1.32 (3H, d, J¼6.0 Hz), 1.49 (1H, ddd, J¼12.6, 12.6,
9.8 Hz), 1.99 (1H, ddd, J¼12.6, 4.4, 2.0 Hz), 2.20 (6H, s),
2.45 (1H, ddd, J¼12.6, 9.0, 4.4 Hz), 3.00 (1H, br t),
3.10 (1H, dd, J¼9.0, 9.0 Hz), 3.27 (1H, dq, J¼9.0,
6.0 Hz), 3.60–3.90 (4H, m), 4.45 (1H, dd, J¼9.8, 2.0 Hz);
HRMS (EI) m/z 333.2339 (333.2335 calcd for
C16H35NO4Si, Mþ).
1
yellow syrup. Rf 0.60 (4/1 n-hexane/EtOAc); H NMR d
0.10 (1/5£3H, s), 0.12 (4/5£3H, s), 0.18 (1/5£3H, s), 0.21
(4/5£3H, s), 0.90 (1/5£9H, s), 0.94 (4/5£9H, s), 1.27
(4/5£3H, d, J¼6.0 Hz), 1.31 (1/5£3H, d, J¼6.0 Hz), 1.82
(1/5£1H, ddd, J¼12.0, 12.0, 12.0 Hz), 2.13 (4/5£1H, ddd,
J¼13.8, 12.6, 5.9 Hz), 2.33–2.44 (1H, m), 3.07 (1/5£1H,
dd, J¼9.0, 9.0 Hz), 3.11 (4/5£1H, dd, J¼9.0, 9.0 Hz),
3.25–3.42 (1/5£2H, m), 3.64 (4/5£1H, ddd, J¼12.6, 9.0,
5.0 Hz), 4.14 (4/5£1H, dq, J¼9.0, 6.0 Hz), 4.78 (1/4£1H,
dd, J¼12.0, 2.0 Hz), 5.55 (4/5£1H, dd, J¼5.9, 1.0 Hz),
7.23–7.35 (3H, m), 7.40–7.52 (2H, m); HRMS (EI) m/z
379.1773 (379.1750 calcd for C18H29N3O2SSi, Mþ).
4.1.5. [2-(2-Quinoxaloyloxy)ethyl] 4-O-tert-Butyl-
dimethylsilyl-3-N,N0-dimethylamino-2,3,6-trideoxy-a-D-
arabino-hexopyranoside (17). To a stirred solution of 15
(0.193 g, 0.579 mmol) in dry CH2Cl2 (7.7 mL) were added
Et3N (0.323 mL, 2.32 mmol) and 2-quinoxaloyl chloride
(0.334 g, 1.74 mmol) at 08C. The reaction mixture was
stirred at 258C for 15 h before the addition of ice-cold
saturated aqueous NaHCO3 (15 mL). The resultant mixture
was extracted with Et2O (15 mL£3) and the extracts were
washed with brine (15 mL), dried over anhydrous Na2SO4
and concentrated in vacuo. Purification of the residue by
column chromatography (30 g of silica gel, 2/1 n-hexane/
EtOAc) gave 17 (0.270 g, 94%) as a colorless syrup. Rf 0.58
4.1.3. 2-Hydroxyethyl 3-Azide-4-O-tert-butyldimethyl-
silyl-2,3,6-trideoxy-D-arabino-hexopyranoside (14). To a
stirred mixture of 12 (0.480 g, 1.26 mmol), ethylene glycol
(13) (0.352 mL, 6.32 mmol) and MS 4A (0.480 g) in dry

K. Toshima et al. / Tetrahedron 59 (2003) 7057–7066
7063
1
(1/1 n-hexane/EtOAc); [a]2D8¼þ36.18 (c 1.10, CHCl3); H
NMR d 0.01 (3H, s), 0.03 (3H, s), 0.78 (9H, s), 1.24 (3H, d,
J¼6.0 Hz), 1.48–1.56 (1H, m), 1.88 (1H, br dd, J¼12.4,
4.0 Hz), 2.17 (3H, br s), 2.70–2.84 (1H, m), 3.10 (1H, dd,
J¼9.0, 9.0 Hz), 3.63 (1H, dq, J¼9.0, 6.0 Hz), 3.80–3.90
(1H, m), 3.98–4.08 (1H, m), 4.56–4.70 (1H, m), 4.72–4.82
(1H, m), 4.97 (1H, br d, J¼2.8 Hz), 7.86 (1H, ddd, J¼8.0,
8.0, 1.6 Hz), 7.92 (1H, ddd, J¼8.0, 8.0, 1.6 Hz), 8.20 (1H,
dd, J¼8.0, 1.6 Hz), 8.32 (1H, dd, J¼8.0, 1.6 Hz), 9.54 (1H,
s); HRMS (EI) m/z 489.2657 (489.2659 calcd for
C25H39N3O5Si, Mþ).
(99.6 mg, 0.662 mmol) at 08C. The reaction mixture was
stirred at 08C for 10 min and poured into ice-cold water
(5 mL). The resultant mixture was extracted with EtOAc
(5 mL£3) and the extracts were washed with brine (5 mL),
dried over anhydrous Na2SO4 and concentrated in vacuo.
Purification of the residue by column chromatography (10 g
of silica gel, 3/1 n-hexane/EtOAc) gave 19 (73.2 mg, 40%)
as a colorless syrup and the b-anomer (80.6 mg, 44%) as a
colorless syrup. 19: Rf 0.60 (3/1 n-hexane/EtOAc);
1
[a]2D9¼þ82.28 (c 1.91, CHCl3); H NMR d 0.07 (6H, s),
0.89 (9H, s), 1.28 (3H, d, J¼6.0 Hz), 1.92 (1H, ddd, J¼12.0,
12.0, 3.8 Hz), 2.18 (1H, ddd, J¼12.0, 4.4, 1.0 Hz), 3.14 (1H,
dd, J¼9.0, 9.0 Hz), 3.48 (1H, ddd, J¼10.4, 5.2, 5.2 Hz),
3.64–3.82 (5H, m), 4.88 (1H, br d, J¼3.8 Hz); HRMS (EI)
m/z 331.1949 (331.1927 calcd for C14H29N3O4Si, Mþ). The
b-anomer of 19: Rf 0.51 (3/1 n-hexane/EtOAc);
4.1.6. [2-(2-Quinoxaloyloxy)ethyl] 3-N,N0-Dimethyl-
amino-2,3,6-trideoxy-a-D-arabino-hexopyranoside (5).
To a stirred solution of 17 (27. 0 mg, 0.0547 mmol) in dry
pyridine (1.35 mL) was added 70% HF/Py (0.243 mL)
dropwise at 08C. The reaction mixture was stirred at 258C
for 12 h and poured into ice-cold saturated aqueous
NaHCO3 (6 mL). The resultant mixture was extracted
with EtOAc (5 mL£3) and the extracts were washed with
brine (5 mL), dried over anhydrous Na2SO4 and concen-
trated in vacuo. Purification of the residue by column
chromatography (2 g of silica gel, 5/1 CHCl3/MeOH) gave
5 (16.2 mg, 79%) as a colorless syrup. Rf 0.25 (5/1 CHCl3/
[a]2D9¼243.68 (c 1.48, CHCl3); H NMR d 0.06 (6H, s),
1
0.89 (9H, s), 1.34 (3H, d, J¼6.0 Hz), 1.64 (1H, ddd, J¼12.4,
12.4, 9.8 Hz), 2.20 (1H, d, J¼4.0 Hz), 2.25 (1H, ddd,
J¼12.4, 5.0, 2.0 Hz), 3.14 (1H, ddd, J¼9.0, 9.0, 4.0 Hz),
3.24–3.46 (2H, m), 3.56–3.66 (1H, m), 3.70–3.92 (3H, m),
4.59 (1H, dd, J¼9.8, 2.0 Hz); HRMS (EI) m/z 331.1932
(331.1927 calcd for C14H29N3O4Si, Mþ).
MeOH); [a]2D8¼þ38.48 (c 0.94, CHCl3); H NMR d 1.29
4.1.9. 2-tert-Butyldimethysilyloxyethyl 3-Amino-2,3,6-
trideoxy-a-D-arabino-hexopyranoside (20). To a solution
of 19 (92.2 mg, 0.278 mmol) in EtOH (9.2 mL) was added
Pd/CaCO3 (46.1 mg). After the mixture was vigorously
stirred at 258C for 16 h under H2, the mixture was filtered
and the catalyst was washed with EtOH. The combined
filtrate and washings were concentrated in vacuo. Purifi-
cation of the residue by column chromatography (1 g of
silica gel, 3/1 CHCl3/MeOH) gave 20 (84.9 mg, 99%) as a
colorless syrup. Rf 0.10 (2/1 CHCl3/MeOH); [a]2D7¼þ50.28
(c 2.19, CHCl3); 1H NMR d 0.06 (6H, s), 0.88 (9H, s), 1.26
(3H, ddd, J¼12.4, 12.4, 3.6 Hz), 2.00 (1H, br dd, J¼12.4,
4.4 Hz), 2.06 (3H, br s), 2.87 (1H, dd, J¼9.0, 9.0 Hz), 3.02
(1H, ddd, J¼12.4, 9.0, 4.4 Hz), 3.45 (1H, ddd, J¼10.0, 5.6,
5.6 Hz), 3.60–3.78 (4H, m), 4.81 (1H, br d, J¼3.6 Hz);
HRMS (EI) m/z 305.2017 (305.2022 calcd for
C14H31NO4Si, Mþ).
1
(3H, d, J¼6.0 Hz), 1.64 (1H, ddd, J¼12.4, 12.4, 3.8 Hz),
1.92 (1H, ddd, J¼12.4, 4.0, 1.0 Hz), 2.32 (6H, s), 2.98 (1H,
ddd, J¼12.4, 9.0, 4.0 Hz), 3.15 (1H, dd, J¼9.0, 9.0 Hz),
3.78 (1H, dq, J¼9.0, 6.0 Hz), 3.82–3.94 (1H, m), 4.00–4.10
(1H, m), 4.61–4.79 (2H, m), 5.05 (1H, br d, J¼3.8 Hz),
7.85–7.97 (2H, m), 8.17–8.23 (1H, m), 8.30–8.36 (1H, m),
9.55 (1H, s); HRMS (EI) m/z 375.1820 (375.1794 calcd for
C19H25N3O5, Mþ).
4.1.7. [2-(2-Quinoxaloyloxy)ethyl] 3-N,N0-Dimethyl-
amino-2,3,6-trideoxy-b-D-arabino-hexopyranoside (6).
The b-anomer 6 was prepared from the b-anomer of 15 in
a way similar to that for 5 from 15. Rf 0.38 (5/1 CHCl3/
MeOH); [a]2D8¼236.68 (c 1.33, CHCl3); H NMR d 1.35
1
(3H, d, J¼6.0 Hz), 1.50 (1H, ddd, J¼12.2, 12.2, 9.8 Hz),
1.99 (1H, ddd, J¼12.2, 4.0, 2.0 Hz), 2.25 (6H, s), 2.46 (1H,
ddd, J¼12.2, 9.0, 4.0 Hz), 3.05 (1H, dd, J¼9.0, 9.0 Hz),
3.32 (1H, dq, J¼9.0, 6.0 Hz), 3.92–4.02 (1H, m), 4.22–4.32
(1H, m), 4.62 (1H, dd, J¼9.8, 2.0 Hz), 4.64–4.78 (2H, m),
7.84–7.96 (2H, m), 8.16–8.23 (1H, m), 8.29–8.34 (1H, m),
9.55 (1H, s); HRMS (EI) m/z 375.1803 (375.1794 calcd for
C19H25N3O5, Mþ).
4.1.10. 2-tert-Butyldimethysilyloxyethyl 2,3,6-Trideoxy-
3-N-p-nitrobenzenesulfonylamino-a-D-arabino-hexo-
pyranoside (21). To a solution of 20 (0.417 g, 1.36 mmol)
in dry CH2Cl2 (8.3 mL) were added K2CO3 (0.576 g,
4.17 mmol) and p-nitrobenzensulfonyl chloride (0.646 g,
3.41 mmol) at 08C. The reaction mixture was stirred at 258C
for 2 h and poured into ice-cold water (15 mL). The
resultant mixture was extracted with CHCl3 (15 mL£3)
and the extracts were washed with brine (15 mL), dried over
anhydrous Na2SO4 and concentrated in vacuo. Purification
of the residue by column chromatography (25 g of silica gel,
3/1 CHCl3/acetone) gave 5 (0.479 g, 72%) as a colorless
syrup. Rf 0.90 (2/1 CHCl3/MeOH); [a]2D8¼þ64.48 (c 1.33,
CHCl3); 1H NMR d 0.02 (3H, s), 0.04 (3H, s), 0.87 (9H, s),
1.24 (3H, d, J¼6.0 Hz), 1.62 (1H, ddd, J¼12.4, 12.4,
3.8 Hz), 1.93 (1H, br dd, J¼12.4, 4.4 Hz), 2.51 (1H, br s),
3.07 (1H, br dd, J¼9.0, 9.0 Hz), 3.42 (1H, ddd, J¼10.0, 5.4,
5.4 Hz), 3.50–3.76 (5H, m), 4.75 (1H, br d, J¼3.8 Hz), 5.10
(1H, br d, J¼7.2 Hz), 8.07–8.13 (2H), 8.34–8.40 (2H);
HRMS (EI) m/z 490.1824 (490.1805 calcd for C20H34N2-
O8SSi, Mþ).
4.1.8. 2-tert-Butyldimethysilyloxyethyl 3-Azide-2,3,6-tri-
deoxy-a-D-arabino-hexopyranoside (19). To a stirred
mixture of 11 (0.183 g, 0.688 mmol), MS 4A (0.183 g)
and ethylene glycol (13) (0.192 mL, 3.44 mol) in dry MeCN
(4.6 mL) was added NBS (0.184 g, 1.03 mmol) at 08C. The
reaction mixture was stirred at 08C for 15 min and poured
into ice-cold 10% NaHSO4 aq. (15 mL). The resultant
mixture was extracted with EtOAc (15 mL£3) and the
extracts were washed with brine (15 mL), dried over
anhydrous Na2SO4 and concentrated in vacuo. Purification
of the residue by column chromatography (10 g of silica gel,
1/5 n-hexane/EtOAc) gave 18 (0.120 g, 80%, a/b¼1/1.1) as
a mixture of the a- and b-anomers. To a stirred solution of
18 (0.120 g, 0.552 mmol) in dry CH2Cl2 (2.4 mL) were
added imidazole (56.2 mg, 0.828 mmol) followed by TBSCl

7064
K. Toshima et al. / Tetrahedron 59 (2003) 7057–7066
4.1.11. 2-tert-Butyldimethysilyloxyethyl 4-O-tert-Butyl-
dimethylsilyl-2,3,6-trideoxy-3-N-p-nitrobenzenesul-
d, J¼6.0 Hz), 1.51 (2H, br s), 1.62 (2H, ddd, J¼12.4, 4.0,
0.8 Hz), 1.98 (2H, ddd, J¼12.4, 12.4, 3.6 Hz), 3.50 (2H, dq,
J¼9.0, 6.0 Hz), 3.55–3.70 (8H, m), 3.75 (2H, dd, J¼9.0,
9.0 Hz), 4.03 (2H, ddd, J¼12.4, 9.0, 4.0 Hz), 4.58, 4.65
(each 2H, ABq, J¼16.0 Hz), 4.72 (2H, br d, J¼3.6 Hz), 7.08
(4H), 7.78–7.85 (2H), 8.17–8.24 (2H); HRMS (EI) m/z
1082.4087 (1082.4080 calcd for C48H74N4O16S2Si2, Mþ).
fonylamino-a-D-arabino-hexopyranoside (22). To
a
stirred solution of 21 (0.479 g, 0.975 mmol) in dry
CH2Cl2 (9.6 mL) were added dropwise 2,6-lutidine
(0.454 mL, 3.90 mmol) followed by TBSOTf (0.672 mL,
2.93 mmol) at 08C. The reaction mixture was stirred at 408C
for 16 h and poured into ice-cold water (10 mL). The
resultant mixture was extracted with CHCl3 (10 mL£3) and
the extracts were washed with brine (10 mL), dried over
anhydrous Na2SO4 and concentrated in vacuo. Purification
of the residue by column chromatography (15 g of silica gel,
2/1 n-hexane/EtOAc) gave 22 (0.590 g, 100%) as a colorless
syrup. Rf 0.80 (3/1 CHCl3/acetone); [a]2D9¼þ53.68 (c 1.32,
CHCl3); 1H NMR d 0.05 (3H, s), 0.06 (3H, s), 0.10 (3H, s),
0.18 (3H, s), 0.85 (9H, s), 0.88 (9H, s), 1.18 (3H, d,
J¼6.0 Hz), 1.58 (1H, ddd, J¼12.4, 12.4, 3.8 Hz), 2.02 (1H,
ddd, J¼12.4, 4.4, 1.0 Hz), 3.12 (1H, dd, J¼9.0, 9.0 Hz),
3.40 (1H, ddd, J¼10.0, 5.6, 5.4 Hz), 3.46–3.77 (5H, m),
4.56 (1H, d, J¼6.2 Hz), 4.71 (1H, dd, J¼3.8, 1.0 Hz), 8.02–
8.10 (2H), 8.32–8.38 (2H); HRMS (EI) m/z 604.2650
(604.2670 calcd for C26H48N2O8SSi2, Mþ).
4.1.14. 1,4-Bis[2-(2-quinoxaloyloxy)ethyl 4-O-tert-butyl-
dimethylsilyl-2,3,6-trideoxy-3-N-p-nitrobenzenesul-
fonylamino-a-D-arabino-hexopyranoside-3-ylmethyl]-
benzene (26). To a stirred solution of 25 (23.6 mg,
0.0218 mmol) in dry CH2Cl2 (0.472 mL) were added Et3N
(0.0152 mL, 0.109 mmol) and 2-quinoxaloyl chloride
(16.8 mg, 0.0871 mmol) at 08C. The reaction mixture was
stirred at 258C for 20 min and poured into ice-cold water
(3 mL). The resultant mixture was extracted with CHCl3
(3 mL£3) and the extracts were washed with brine (3 mL),
dried over anhydrous Na2SO4 and concentrated in vacuo.
Purification of the residue by column chromatography (2 g
of silica gel, 2/1 CHCl3/EtOAc) gave 26 (28.3 mg, 93%) as
pale yellow solids. Rf 0.75 (1/1 n-hexane/acetone);
1
[a]2D9¼þ31.08 (c 0.91, CHCl3); H NMR d 0.11 (6H, s),
4.1.12. 1,4-Bis(2-tert-butyldimethysilyloxyethyl 4-O-tert-
butyldimethylsilyl-2,3,6-trideoxy-3-N-p-nitrobenzene-
sulfonylamino-a-D -arabino-hexopyranoside)-3-
ylmethyl)benzene (24). To a stirred solution of 22 (0.376 g,
0.621 mmol) and p-dibromoxylene (82. 0 mg, 0.310 mmol)
in dry DMF (7.5 mL) was added K2CO3 (0.343 g,
2.48 mmol) at 08C. The reaction mixture was stirred at
258C for 3 h and poured into ice-cold water (10 mL). The
resultant mixture was extracted with a mixture of n-hexane
and EtOAc (1/1) (10 mL£3) and the extracts were washed
with brine (10 mL), dried over anhydrous Na2SO4 and
concentrated in vacuo. Purification of the residue by column
chromatography (15 g of silica gel, 50/1 CHCl3/EtOAc)
gave 24 (0.165 g, 41%) as white solids. Rf 0.60 (5/1
n-hexane/EtOAc). Mp 178–1808C; [a]2D9¼þ28.58 (c 1.40,
CHCl3); 1H NMR d 0.04 (6H, s), 0.06 (6H, s), 0.17 (6H, s),
0.21 (6H, s), 0.88 (18H, s), 0.90 (18H, s), 1.25 (6H, d,
J¼6.0 Hz), 1.65 (2H, ddd, J¼12.4, 4.0, 1.0 Hz), 1.87 (2H,
ddd, J¼12.4, 12.4, 3.6 Hz), 3.43 (2H, ddd, J¼10.0, 5.2,
5.2 Hz), 3.58–3.76 (10H, m), 4.19 (2H, ddd, J¼12.4, 9.0,
4.0 Hz), 4.48, 4.63 (each 2H, ABq, J¼16.0 Hz), 4.72 (2H, br
d, J¼3.6 Hz), 7.10 (4H, s), 7.72–7.79 (4H), 8.12–8.19
(4H); HRMS (EI) m/z 1310.5825 (1310.5809 calcd for
C60H102N4O16S2Si4, Mþ).
0.16 (6H, s), 0.75 (18H, s), 1.23 (6H, d, J¼6.0 Hz), 1.63
(2H, br dd, J¼12.4, 4.0 Hz), 1.86 (2H, ddd, J¼12.4, 12.4,
3.6 Hz), 3.55–3.70 (4H, m), 3.80 (2H, ddd, J¼10.0, 6.0,
4.0 Hz), 3.99 (2H, ddd, J¼10.0, 6.2, 3.9 Hz), 4.06–4.20
(2H, m), 4.41, 4.58 (each 2H, ABq, J¼16.0 Hz), 4.55–4.65
(2H, m), 4.73 (2H, ddd, J¼10.0, 6.2, 3.9 Hz), 4.78 (2H, d,
J¼3.6 Hz), 7.08 (4H, s), 7.75–7.93 (8H, m), 8.10–8.27
(8H, m), 9.46 (2H, s); HRMS (EI) m/z 1394.4710
(1394.4727 calcd for C66H82N8O18S2Si2, Mþ).
4.1.15. 1,4-Bis[2-(2-quinoxaloyloxy)ethyl 3-amino-4-O-
tert-butyldimethylsilyl-2,3,6-trideoxy-a-D-arabino-hexo-
pyranoside-3-ylmethyl]benzene (27). To a stirred solution
of 26 (28.3 mg, 0.0203 mmol) in dry DMF (0.566 mL) were
added K2CO3 (22.4 mg, 0.162 mmol) and PhSH
(0.0125 mL, 0.122 mmol) at 08C. The reaction mixture
was stirred at 258C for 4 h and poured into ice-cold saturated
aqueous NaHCO3 (3 mL). The resultant mixture was
extracted with a mixture of n-hexane and EtOAc (1/1)
(3 mL£3) and the extracts were washed with brine (3 mL),
dried over anhydrous Na2SO4 and concentrated in vacuo.
Purification of the residue by column chromatography (2 g
of silica gel, 1/1 CHCl3/acetone) gave 27 (11.8 mg, 54%) as
a
colorless syrup. Rf 0.10 (2/5 CHCl3/EtOAc);
1
[a]2D9¼þ14.68 (c 1.17, CHCl3); H NMR d 20.03 (6H, s),
0.03 (6H, s), 0.78 (18H, s), 1.18 (6H, d, J¼6.0 Hz), 1.51
(2H, ddd, J¼12.4, 12.4, 3.6 Hz), 2.23 (2H, br dd, J¼12.4,
4.2 Hz), 2.98 (2H, ddd, J¼12.4, 9.0, 4.2 Hz), 3.16 (2H, dd,
J¼9.0, 9.0 Hz), 3.54, 3.77 (each 2H, ABq, J¼13.0 Hz), 3.69
(2H, dq, J¼9.0, 6.0 Hz), 3,80–3.90 (2H, m), 3.98–4.10
(2H, m), 4.60–4.82 (4H, m), 4.94 (2H, d, J¼3.6 Hz), 7.20
(4H, s), 7.82–7.95 (4H, m), 8.14–8.22 (2H), 8.26–8.34
(2H), 9.55 (2H, s); HRMS (EI) m/z 1024.5139 (1024.5161
calcd for C54H76N6O10Si2, Mþ).
4.1.13. 1,4-Bis(2-tert-butyldimethysilyloxyethyl 4-O-tert-
butyldimethylsilyl-2,3,6-trideoxy-3-N-p-nitrobenzene-
sulfonylamino-a-D -arabino-hexopyranoside)-3-
ylmethyl)benzene (25). To a stirred solution of 24
(87.5 mg, 0.0667 mmol) in dry pyridine (4.4 mL) was
added 70% HF/Py (0.788 mL) dropwise at 08C. The reaction
mixture was stirred at 258C for 2 h and poured into ice-cold
saturated aqueous NaHCO3 (6 mL). The resultant mixture
was extracted with CHCl3 (5 mL£3) and the extracts were
washed with brine (5 mL), dried over anhydrous Na2SO4
and concentrated in vacuo. Purification of the residue by
column chromatography (5 g of silica gel, 1/1 n-hexane/
acetone) gave 25 (68.6 mg, 95%) as a colorless syrup. Rf
0.20 (2/1 n-hexane/acetone); [a]2D9¼þ16.18 (c 1.18,
4.1.16. 1,4-Bis[2-(2-quinoxaloyloxy)ethyl 3-amino-2,3,6-
trideoxy-a-D-arabino-hexopyranoside-3-ylmethyl]ben-
zene (7). To a stirred solution of 27 (11.8 mg, 0.0115 mmol)
in dry pyridine (0.590 mL) was added 70% HF/Py
(0.106 mL) dropwise at 08C. The reaction mixture was
1
CHCl3); H NMR d 0.19 (6H, s), 0.25 (6H, s), 1.26 (6H,

K. Toshima et al. / Tetrahedron 59 (2003) 7057–7066
7065
stirred at 258C for 16 h and poured into ice-cold saturated
aqueous NaHCO3 (3 mL). The resultant mixture was
extracted with CHCl3 (3 mL£3) and the extracts were
washed with brine (3 mL), dried over anhydrous Na2SO4
and concentrated in vacuo. Purification of the residue by
column chromatography (5 g of silica gel, 3/1 CHCl3/
MeOH) gave 7 (7.9 mg, 89%) as a colorless syrup. Rf 0.50
(3/1 CHCl3/MeOH); [a]2D9¼þ15.08 (c 1.55, CHCl3); 1H
NMR d 1.27 (6H, d, J¼6.0 Hz), 1.48 (2H, ddd, J¼12.8,
12.8, 4.0 Hz), 2.29 (2H, br dd, J¼12.8, 3.4 Hz), 2.86–3.05
(4H, m), 3.62, 3.83 (each 2H, ABq, J¼13.0 Hz), 3.68–3.92
(4H, m), 4.00–4.12 (2H, m), 4.59–4.80 (4H, m), 4.98 (2H,
br d, J¼3.4 Hz), 7.20 (4H, s), 7.83–7.97 (4H, m), 8.16–8.25
(2H), 8.28–8.37 (2H), 9.52 (2H, s); HRMS (EI) m/z
796.3450 (796.3432 calcd for C42H48N6O10, Mþ).
varied as indicated in the figure captions. The results were
analyzed using 0.9% agarose gel electrophoresis and
detection with ethidium bromide fluorescence. The electro-
phoresis gels were immediately visualized on a UV
transilluminator and photographed using black and white
instant film. Figures 3 and 5 show the pictures of the agarose
gel electrophoresis results.
4.2.2. Identification of DNA cleavage sites. The reaction
samples contained the compounds (500 mM) and the 50-end-
labelled M13mp18 DNA (40 ng) in a volume of 30 mL
containing 20% acetonitrile in TE buffer (10 mM Tris–HCl,
1 mM EDTA, pH 8.5). The cleavage reactions were allowed
to proceed under the same conditions described above. To
stop the reactions, each reaction sample was washed with a
solution of TE buffer-saturated phenol–chloroform–iso-
amyl alcohol (25:24:1) and the resulting aqueous layer was
lyophilized. After each lyophilized sample was dissolved in
1 M piperidine–water (20 mL) and then heated at 908C for
30 min, each sample was again lyophilized. Each lyo-
philized sample was dissolved in a loading buffer containing
distilled water, 95% deionized formamide, 10 mM EDTA,
0.05% xylene cyanole FF, and 0.05% bromophenol blue and
then the mixture was loaded onto a 12% polyacrylamide gel
containing 8 M urea in TBE buffer. DNA sequencing was
carried out by the Sanger method. Figure 6 shows a picture
of the autoradiogram.
4.1.17. 1-[2-(2-Quinoxaloyloxy)ethyl 3-amino-2,3,6-tri-
deoxy-a-D-arabino-hexopyranoside-3-ylmethyl]-4-[2-(2-
quinoxaloyloxy)ethyl 3-amino-2,3,6-trideoxy-b-D-ara-
bino-hexopyranoside-3-ylmethyl]benzene (8). 8 was pre-
pared from 19 and the b-anomer in a way similar to that for
7 from 19. Rf 0.45 (3/1 CHCl3/MeOH); [a]2D8¼217.48 (c
1.72, CHCl3); 1H NMR d 1.27 (3H, d, J¼6.0 Hz), 1.32 (3H,
d, J¼6.0 Hz), 1.39 (1H, ddd, J¼12.2, 12.2, 9.4 Hz), 1.48
(1H, ddd, J¼12.8, 12.8, 4.0 Hz), 2.29 (1H, br dd, J¼12.8,
3.4 Hz), 2.36 (1H, ddd, J¼12.2, 4.0, 2.0 Hz), 2.55 (1H, ddd,
J¼12.2, 9.0, 4.0 Hz), 2.86–3.05 (3H, m), 3.32 (1H, dq,
J¼9.0, 6.0 Hz), 3.62, 3.83 (each 1H, ABq, J¼13.0 Hz),
3.64, 3.89 (each 1H, ABq, J¼12.4 Hz), 3.68–4.12 (3H, m),
4.19–4.30 (1H, m), 4.59–4.80 (4H, m), 4.60 (1H, dd,
J¼9.4, 2.0 Hz), 4.98 (2H, br d, J¼3.4 Hz), 7.22 (4H, s),
7.83–7.97 (4H), 8.16–8.25 (2H), 8.27–8.37 (2H), 9.52
(1H, s), 9.55 (1H, s); HRMS (EI) m/z 796.3411 (796.3432
calcd for C42H48N6O10, Mþ).
4.2.3. Assay for DNA cleavage in cells. HeLa S3 cells were
treated with the compounds (10 mM) with or without
photoirradiation at 258C for 1 h. The UV lamp (365 nm,
30 W) was placed at 25 cm from the mixture. After the cells
were cultured for 72 h at 378C, the DNA was extracted from
cells and analyzed by gel electrophoresis in 1.5% agarose
gel. The gel was stained with ethidium bromide, visualized
on a UV transilluminator and photographed using black
and white instant film. Figure 7 shows the pictures of the
agarose gel electrophoresis results. In this experiment, it
was confirmed that UV irradiation alone under these
conditions did not influence the cell survival, and the
DNA cleaving activity of each compound under these
conditions is similar to that under the conditions mentioned
in Section 4.2.1.
4.1.18. 1,4-Bis-[2-(2-quinoxaloyloxy)ethyl 3-amino-2,3,6-
trideoxy-b-D-arabino-hexopyranoside-3-ylmethyl]ben-
zene (9). Compound 9 was prepared from the b-anomer of
19 in a way similar to that for 7 from 19. Rf 0.50 (3/1 CHCl3/
1
MeOH); [a]3D0¼259.58 (c 0.95, CHCl3); H NMR d 1.32
(6H, d, J¼6.0 Hz), 1.39 (2H, ddd, J¼12.2, 12.2, 9.4 Hz),
2.36 (2H, ddd, J¼12.2, 4.0, 2.0 Hz), 2.55 (2H, ddd, J¼12.2,
9.0, 4.0 Hz), 2.95 (2H, dd, J¼9.0, 9.0 Hz), 3.32 (2H, dq,
J¼9.0, 6.0 Hz), 3.64, 3.89 (each 2H, ABq, J¼12.4 Hz),
3.85–4.03 (2H, m), 4.19–4.30 (2H, m), 4.60 (2H, dd,
J¼9.4, 2.0 Hz), 4.65–4.80 (4H, m), 7.25 (4H, s), 7.80–7.97
(4H, m), 8.17–8.23 (2H, m), 8.27–8.35 (2H, m), 9.55 (2H,
s); HRMS (EI) m/z 796.3427 (796.3432 calcd for
C42H48N6O10, Mþ).
4.3. Cytotoxicity assay
The cells were treated with each compound at the
concentrations of 300, 100, 30, 10, 3, 1, 0.3, 0.1, 0.03 and
0.01 mM with photoirradiation at 258C for 1 h with a UV
lamp (365 nm, 30 W) placed 25 cm from the mixture, and
incubated for 72 h at 378C. The cell viability was
determined using the XTT-tetrazolium assay described
by Scudiero. Table 1 indicated the IC50 values of
each compound against the HeLa S3 and MDA-MB-231
cells.
4.2. DNA cleavage studies
FX174 DNA and M13mp18 ss DNA were purchased from
Nippon Gene Co., Ltd., and TaKaRa Bio Inc., respectively.
XX-15 BLB (UVP Inc.) was used as a UV lamp (365 nm, 15
or 30 W) for photoirradiation.
4.2.1. Assay for damage to DNA. All the DNA cleavage
experiments were performed with FX174 DNA
(50 mM/base pair) in a volume of 6 mL containing 20%
acetonitrile in 50 mM Tris–HCl buffer (pH 7.5) at 258C for
1 h under irradiation of the UV lamp (365 nm, 15 W) placed
10 cm from the mixture. The DNA-sample levels were
Acknowledgements
This research was partially supported by Grant-in-Aid for
the 21st Century COE program ‘KEIO Life Conjugate
Chemistry’ and for General Scientific Research from the

7066
K. Toshima et al. / Tetrahedron 59 (2003) 7057–7066
3656–3658. (e) Toshima, K.; Okuno, Y.; Nakajima, Y.;
Matsumura, S. Bioorg. Med. Chem. Lett. 2002, 12, 671–673.
8. Bartner, P.; Boxier, D. L.; Brambilla, R.; Mallams, A. K.;
Morton, J. B.; Reichert, P.; Sancilio, F. D.; Surprenant, H.;
Tomalesky, G.; Lukacs, G.; Olesker, A.; Thang, T. T.;
Valente, L.; Omura, S. J. Chem. Soc. Perkin Trans. 1 1979,
1600–1624.
Ministry of Education, Culture, Sports, Science, and
Technology, Japan, and Takeda Science Foundation.
References
1. Demeunynck, M.; Bailly, C.; Wilson, W. D. DNA and RNA
Binders. Wiley-VCH: Weinheim, 2003.
2. Armitage, B. Chem. Rev. 1998, 98, 1171–1200.
9. Nicolaou, K. C.; Seitz, S. P.; Papahatjis, D. P. J. Am. Chem.
Soc. 1983, 105, 2430–2434.
10. Fukuyama, T.; Jow, C.-K.; Cheung, M. Tetrahedron Lett.
1995, 36, 6373–6374.
3. (a) Brown, J. E.; Brown, S. B.; Vernon, D. I. J. Soc. Dyers
Colour 1999, 115, 249–353. (b) Morgan, A. R. Curr. Med.
Chem. 1995, 2, 604–615. (c) Bonnett, R. Chem. Soc. Rev.
1995, 19–33. (d) Henderson, R. W.; Dougherty, T. J.
Photodynamic Therapy: Basic Principles and Clinical
Applications; Marcel Dekker: New York, 1992.
11. (a) Chaires, J. B.; Leng, F.; Prewloka, T.; Fokt, I.; Ling, Y.-H.;
Perez-Soler, R.; Priebe, W. J. Med. Chem. 1997, 40, 261–266.
(b) Hu, G. G.; Shui, X.; Leng, F.; Priebe, W.; Chaires, J. B.;
William, L. D. Biochemistry 1997, 36, 5940–5946.
12. Sanger, F.; Nicklen, S.; Coulsen, A. R. Proc. Natl Acad. Sci.
USA 1977, 74, 5463–5467.
4. A preliminary communication of this work Toshima, K.;
Takano, R.; Ozawa, T.; Matsumura, S. Chem. Commun. 2002,
212–213.
13. (a) Boger, D. L.; Munk, S. A.; Zarrinmayeh, H.; Ishizaki, T.;
Haught, J.; Bina, M. Tetrahedron 1991, 47, 2661–2682.
(b) Toshima, K.; Ohta, K.; Ohashi, A.; Nakamura, T.; Nakata,
M.; Tatsuta, K.; Matsumura, S. J. Am. Chem. Soc. 1995, 117,
4822–4831.
5. Waring, M. J.; Fox, K. R. In Molecular Aspects of Anti-cancer
Drug Action; Niedle, S., Waring, M. J., Eds.; Macmillan: New
York, 1983; pp 127–156.
6. Waring, M. J.; Wakelin, L. P. G. Nature 1974, 252, 653–657.
7. (a) Toshima, K.; Ouchi, H.; Okazaki, Y.; Kano, T.; Moriguchi,
M.; Asai, A.; Matsumura, S. Angew. Chem. Int. Ed. Engl.
1997, 36, 2748–2750. (b) Toshima, K.; Takano, R.; Maeda,
Y.; Suzuki, M.; Asai, A.; Matsumura, S. Angew. Chem. Int.
Ed. Engl. 1999, 38, 3733–3735. (c) Toshima, K.; Maeda, Y.;
Ouchi, H.; Asai, A.; Matsumura, S. Bioorg. Med. Chem. Lett.
2000, 10, 2163–2165. (d) Toshima, K.; Takai, S.; Maeda, Y.;
Takano, R.; Matsumura, S. Angew. Chem. Int. Ed. 2000, 39,
14. (a) Saito, I.; Takayama, M.; Sugiyama, H.; Nakatani, K.;
Tsuchida, A.; Yamamoto, M. J. Am. Chem. Soc. 1995, 117,
6406–6407. (b) Sugiyama, H.; Saito, I. J. Am. Chem. Soc.
1996, 118, 7063–7068.
15. DNA Unwinding Kit available from TopoGEN, Inc. was used.
16. Scudieero, D. A.; Shoemaker, R. H.; Paull, K. D.; Monks, A.;
Tierney, S.; Nofziger, T. H.; Currens, M. J.; Seniff, D.; Boyd,
M. R. Cancer Res. 1988, 48, 4827–4833.