The significant effect of the carbohydrate structures on the DNA photocleavage of the quinoxaline-carbohydrate hybrids

Article abstract of DOI:10.1016/j.bmcl.2004.03.065

The novel DNA interactive quinoxaline-carbohydrate hybrids possessing disaccharides as the carbohydrate moieties were designed and synthesized, and their DNA photocleaving abilities were evaluated in order to examine the effect of the disaccharide structu

Full text of DOI:10.1016/j.bmcl.2004.03.065

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2777–2779
The significant effect of the carbohydrate structures on the DNA
photocleavage of the quinoxaline–carbohydrate hybrids
Kazunobu Toshima,* Tomohiro Ozawa, Tomonori Kimura 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 5 March 2004; revised 23 March 2004; accepted 23 March 2004
Abstract—The novel DNA interactive quinoxaline–carbohydrate hybrids possessing disaccharides as the carbohydrate moieties
were designed and synthesized, and their DNA photocleaving abilities were evaluated in order to examine the effect of the disac-
charide structures. The configurations of the glycosidic bonds in the disaccharide strongly affected the DNA photocleaving ability,
and two b-glycosidic bond linkages were very effective for the DNA photocleavage. Furthermore, the quinoxaline–disaccharide
hybrids exhibited selective cytotoxicity against cancer cells with photoirradiation.
Ó 2004 Elsevier Ltd. All rights reserved.
Studies of the interaction between small molecules and
DNA, especially the effects of the structural character-
istics of the small molecules on the DNA interaction, are
very important in the design of new DNA targeting
drugs.1 A number of DNA interactive small molecules
including DNA oligomers and polyamides have already
been evaluated. In contrast, although carbohydrates
play a key role in various biological processes, the
interaction between carbohydrates and DNA, however,
are still not yet well understood. Only limited studies on
the interaction between certain carbohydrates present in
DNA interactive antibiotics and DNA have been re-
ported.2 On the other hand, the development of pho-
tochemical 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 a significant potential in medicine especially in
the post-genome era.3 In this context, we have previ-
ously reported several novel DNA interactive and
photocleaving small molecules possessing carbohy-
drates.4 Among them, it has been found that a quinox-
affinity to DNA and cleaved the double-stranded DNA
upon irradiation with UV light with a long wavelength
and without any additive.4e;f Furthermore, it has been
confirmed that the quinoxaline–carbohydrate hybrid
system was very effective for the DNA cleavage. Based
on our previous results, we had additional questions on
the DNA cleaving activity–carbohydrate structure rela-
tionship. First, how do oligosaccharides work in the
DNA binding and DNA photocleaving events? Second,
which structure of the oligosaccharide is effective in such
events? Therefore, in this study, several quinoxaline–
carbohydrate hybrids possessing disaccharides were
newly designed and synthesized, and their DNA
photocleaving abilities were evaluated in order to
examine the effect of the disaccharide structures. We
now report the molecular design, chemical synthesis,
DNA photocleaving, and cytotoxic properties of novel
and artificial DNA interactive agents, that include the
quinoxaline–disaccharide hybrids 1–4 (Fig. 1).
In this preliminary study, we designed the quinoxaline–
disaccharide hybrids 1–4. They were commonly con-
structed of a quinoxaline as the DNA intercalative
photocleaving moiety and a disaccharide consisting of a
2,3,6-trideoxy-3-amino sugar linked 2,6-dideoxy sugar
as a DNA groove binder. We selected two types of 2,6-
dideoxy sugars as the carbohydrate part because several
kinds of 2,6-dideoxy sugars are present in many natu-
rally occurring anticancer antibiotics, which bind to
DNA,5 and the hydrophobicity of the 2,6-dideoxy sugar
would be suitable for the DNA groove binding.6
aline–carbohydrate hybrid possessing
deoxygenated monosaccharide showed a high binding
a
suitably
Keywords: Carbohydrate; Oligosaccharide; Quinoxaline; DNA photo-
cleavage; Cytotoxicity.
* Corresponding author. Tel./fax: +81-45-566-1576; e-mail: toshima@
applc.keio.ac.jp
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.03.065

2778
K. Toshima et al. / Bioorg. Med. Chem. Lett. 14(2004) 2777–2779
a mixture of a- and b-anomers. The esterification of 8
with 2-quinoxaloyl chloride (9) in the presence of Et3N
yielded the a-hybrid 10 after separation from the b-
anomer. In these three steps, it was found that no pro-
tection of the C3-hydroxy groupin the sugar moiety was
necessary to carry out these regioselective reactions. The
second glycosidation of 10 and the another phenyl
thioglycoside 114a was effectively performed using NBS
and TfOH10 in EtCN at low temperature to furnish the
a-disaccharide 12 after separation from the b-anomer.
After the azide groupin 12 was converted into the free
amino groupusing Ph 3P, the TBS groups were removed
employing HFꢀPy to afford the targeted quinoxaline–
disaccharide hybrid 1. The other hybrids 2–4 were also
synthesized using the isomers produced in the glycosi-
dation steps by very similar protocols.
Figure 1. Molecular structures of quinoxaline–disaccharide hybrids
1–4.
With all the designed quinoxaline–disaccharide hybrids
1–4 in hand, the photo-induced DNA cleaving activities
of these hybrids were next assayed using supercoiled
UX174 DNA in concentrations of 3–500 lM. Based on
the results shown in Figure 2, the quinoxaline–disac-
charide hybrids 1–4 caused DNA cleavage during irra-
diation with a long wavelength UV light (365 nm). It was
confirmed that no DNA cleavage by 1–4 was observed
in the absence of light (lanes 3 in Fig. 2). Thus, the UV
light functioned as the trigger to initiate these quinox-
aline–disaccharide hybrids for the DNA strand scission.
Furthermore, interestingly, the DNA cleaving abilities
of the quinoxaline–disaccharide hybrids 1–4 were quite
different and significantly increased in that order. The
Actually, based on this concept, we have previously
demonstrated that a suitably deoxygenated monosac-
charide showed a high affinity to DNA, and significantly
enhanced the intercalating ability of certain intercal-
ators.4;7 These hybrids, however, had different glycosidic
bonds to each other in terms of their stereochemistry.
Thus, the hybrids 1–4 possessed a,a-, a,b-, b,a-, and b,b-
glycosidic linkages, respectively.
The synthesis of the representative quinoxaline–disac-
charide hybrid 1 is summarized in Scheme 1. The syn-
thesis of 1 commenced with the conversion of the known
methyl glycoside 58 into the phenyl thioglycoside 6 using
BF3ꢀEt2O in the presence of PhSH. The glycosidation of
6 and ethylene glycol (7) by Nicolaou et al.’s method9
using NBS proceeded smoothly to give the glycoside 8 as
Figure 2. Photocleavage of supercoiled UX174 DNA. UX174 DNA
(50 lM per base pair) was incubated with various compounds in 20%
acetonitrile in Tris–HCl buffer (pH 7.5, 50 mM) at 25 °C 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, ethi-
dium bromide stain): (a)–(d) for the compounds 1, 2, 3, and 4,
respectively: lane 1, DNA alone; lane 2, DNA with UV; lane 3,
DNA+compound (500 lM) without UV; lanes 4–9, compound (500),
compound (300), compound (100), compound (30), compound (10),
and compound (3 lM), respectively, following UV irradiation.
Scheme 1. Synthesis of 1. Reagents and conditions: (a) PhSH,
BF3ꢀEt2O, )10 °C, 30 min, 80%; (b) NBS, MS 4A, THF, 0 °C, 10 min,
74% (a=b ¼ 1:1=1); (c) Et3N, CH2Cl2, rt, 10 min, 80%; (d) NBS, TfOH,
MS 4A, EtCN, )78 °C, 20 min, 84% (a=b ¼ 2:2=1); (e) Ph3P, H2O,
THF, 60 °C, 13 h, 79%; (f) HFꢀPy, Py, 40 °C, 14 h, 71%.

K. Toshima et al. / Bioorg. Med. Chem. Lett. 14(2004) 2777–2779
2779
CC50 values, which were the concentrations at which
50% Form I DNA was converted into Form II DNA, of
1–4 were 160, 65, 43, and 25 lM, respectively. The
strongest DNA cleaving hybrid 4 possessing b; b-glyco-
sidic linkages was more than six times stronger than the
weakest hybrid 1 possessing a; a-glycosidic linkages, and
cleaved DNA in concentrations over 3 lM, and caused a
100% DNA break at concentrations over 100 lM ((d) in
Fig. 2). Since only the quinoxaline had the DNA
photocleaving activity and the carbohydrates themselves
showed no DNA photocleaving ability, the difference in
the DNA photocleaving ability of these hybrids came
from the structure difference of the disaccharide moieties
in the hybrids. It was also confirmed that no difference
was observed in the absorbance of each hybrid at
365 nm. Thus, it was noteworthy that the DNA cleaving
activity was highly dependent on the structure of the
carbohydrate, especially the configurations of the two
glycosidic bonds, and the two b-glycosidic bond linkages
were very effective for the DNA photocleavage.11 Fur-
thermore, these results also strongly suggest that the
disaccharide containing the suitable glycosidic linkages
works well as the DNA groove binder and significantly
enhances the intercalating ability of the quinoxaline.
photocleaving and cytotoxic agents including carbohy-
drate(s), especially oligosaccharide(s).
Acknowledgements
This research was partially supported by Grant-in-Aid
for the 21st Century COE program ‘KEIO Life Conju-
gate Chemistry’ and for General Scientific Research
from the Ministry of Education, Culture, Sports, Sci-
ence, and Technology, Japan.
References and notes
1. Demeunynck, M.; Bailly, C.; Wilson, W. D. DNA and
RNA Binders; Wiley-Vch: Weinheim, 2003.
2. Moser, H. E. In Carbohydrates in Chemistry and Biology;
€
Ernst, B., Hart, G. W., Sinay, P., Eds.; Wiley-Vch:
Weinheim, 2000; Vol. 2, pp 1095–1124.
3. Armitage, B. Chem. Rev. 1998, 98, 1171.
4. (a) Toshima, K.; Takano, R.; Maeda, Y.; Suzuki, M.;
Asai, A.; Matsumura, S. Angew. Chem., Int. Ed. Engl.
1999, 38, 3733; (b) Toshima, K.; Maeda, Y.; Ouchi, H.;
Asai, A.; Matsumura, S. Bioorg. Med. Chem. Lett. 2000,
10, 2163; (c) Toshima, K.; Takai, S.; Maeda, Y.; Takano,
R.; Matsumura, S. Angew. Chem., Int. Ed. 2000, 39, 3656;
(d) Toshima, K.; Okuno, Y.; Nakajima, Y.; Matsumura,
S. Bioorg. Med. Chem. Lett. 2002, 12, 671; (e) Toshima,
K.; Takano, R.; Ozawa, T.; Matsumura, S. Chem.
Commun. 2002, 23, 212; (f) Toshima, K.; Kimura, T.;
Takano, R.; Ozawa, T.; Ariga, A.; Shima, Y.; Umezawa,
K.; Matsumura, S. Tetrahedron 2003, 59, 7057.
The cytotoxicity of the potent hybrids 3 and 4 was next
examined using HeLa S3 cells exposed to each agent for
72 h with or without 1 h of photoirradiation (Table 1).12
It was found that the IC50 values of 3 and 4 against the
HeLa S3 cells without photoirradiation were >100 lM,
and those with photoirradiation were 6.0 and 5.6 lM,
respectively. These results clearly indicate that the qui-
noxaline–disaccharide hybrids 3 and 4 are themselves
nontoxic while they show high cytotoxic activities with
photoirradiation. Furthermore, these results also dem-
onstrate that the DNA-cleaving activity induced by
photoirradiation significantly affects the cytotoxicity of
the hybrid, and the life of cancer cells can be controlled
by treatment with an appropriate amount of the qui-
noxaline–disaccharide hybrid with or without the
photoirradiation.
€
5. Jutten, P.; Greven, R. In Polysaccharides in Medicinal
Applications; Dumitriu, S., Ed.; Marcel Dekker: New
York, 1996; pp 339–410.
6. (a) Walker, S.; Valentine, K. G.; Kahne, D. J. Am. Chem.
Soc. 1990, 112, 6428; (b) Ding, W.; Ellestad, G. A. J. Am.
Chem. Soc. 1991, 113, 6617; (c) Uesugi, M.; Sugiura, Y.
Biochemistry 1993, 32, 4622.
7. Toshima, K.; Ouchi, H.; Okazaki, Y.; Kano, T.; Morig-
uchi, M.; Asai, A.; Matsumura, S. Angew. Chem., Int. Ed.
Engl. 1997, 36, 2748.
8. Juergen, D.; Alexander, K.; Jan, R.; Dieter, S. H. Liebigs
Ann. Chem. 1985, 1838.
Table 1. Cytotoxicity of 3 and 4 against HeLa S3 cells with or without
photoirradiation
Compounds
3
4
9. Nicolaou, K. C.; Seitz, S. P.; Papahatjis, D. P. J. Am.
Chem. Soc. 1983, 105, 2430.
IC50 [lM] HeLa S3
Without UV
With UV
>100
>100
6.0
5.6
10. (a) Veeneman, G. H.; van Leeuwen, S. H.; van Boom, J.
H. Tetrahedron Lett. 1990, 31, 1331; (b) Konradsson, P.;
Udodong, U. K.; Fraser-Reid, B. Tetrahedron Lett. 1990,
31, 4313.
11. The site selectivities of the DNA cleavages using 1–4 were
quite similar and identical with that of the corresponding
monosaccharide derivative previously reported4e;f due to
the electron transfer mechanism of the DNA photoclea-
vage of quinoxaline.
In summary, the present study demonstrates not only
the molecular design and chemical synthesis of novel
quinoxaline–disaccharide hybrids, but also their DNA
photocleavage and cytotoxic profiles. The effect of the
carbohydrate structures on the DNA binding and
cleaving abilities of the DNA photocleaving agents was
also demonstrated. The described chemistry and bio-
logical evaluation provided significant information
about the molecular design of novel and selective DNA
12. 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.