Photodegradation of target oligosaccharides by light-activated small molecules

Article abstract of DOI:10.1002/anie.201005161

Shine on, shine on: The designed anthraquinone/boronic acid hybrid 1 selectively caused the degradation of oligosaccharides that have a β-D-galactofuranoside residue, which is a unique component in mycobacterial cell walls. Degradation was achieved using

Full text of DOI:10.1002/anie.201005161

Communications
DOI: 10.1002/anie.201005161
Oligosaccharides
Photodegradation of Target Oligosaccharides by Light-Activated Small
Molecules**
Daisuke Takahashi, Shingo Hirono, Chigusa Hayashi, Masayuki Igarashi, Yoshio Nishimura,
and Kazunobu Toshima*
Carbohydrates in living cells play important roles in many
biological events, including bacterial cell wall recognition,
viral and bacterial infection, cell signaling, tumor cell meta-
stasis and fertilization.[1] The development of innovative
methods for selectively controlling specific functions of
certain oligosaccharides has attracted much attention in the
fields of chemistry, biology, and medicine. One major
approach in the functional analysis of oligosaccharides is
genetic knockout experiments of oligosaccharide-processing
enzymes.[2] For example, in a 1994 study of the gene Mgat-1,
which encodes N-acetylglucosaminyl transferase I (GlcNAc-
TI), the absence of the gene in knockout mice was found to be
embryonically lethal because of the loss of complex and
hybrid N-glycan biosynthesis.[3,4] However, it is extremely
difficult to knockout multiple genes simultaneously to remove
a specific oligosaccharide. Because oligosaccharides share
sugar components and sometimes exhibit similar functions,
knocking out a gene that codes for an enzyme involved in
many early synthetic processes may affect downstream path-
ways. Therefore, the possibility of developing a chemical
agent that can selectively and directly degrade target oligo-
saccharides under mild conditions has attracted much atten-
tion. In general, however, it is difficult to achieve selective
degradation of a target oligosaccharide, even with a chemical
approach, because of the complexity of the oligosaccharide
structures compared to those of DNA and proteins. Herein
we report an innovative method for the degradation of target
oligosaccharides induced by light-activated small molecules
under mild conditions and without additives.
dextrins (CDs),[6] which show a high affinity with AQ
derivatives, upon irradiation with long wavelength UV light
and without further additives. In addition, it was found that
the designed and synthesized AQ/lectin (peanut agglutinin;
PNA) hybrid selectively degraded the target tumor-associated
disaccharide Gal(b-1,3)GalNAc.[6]
On the basis of our preliminary and fundamental findings,
we expected that a hybrid molecule consisting of an artificial
receptor, which selectively binds to a target oligosaccharide,
and a photodegrading AQ derivative could be used for
selective photodegradation of target oligosaccharides. To
investigate our hypothesis, we focused on the structure of d-
galactofuranoside, which is a major sugar component of d-
galactan in mycobacterial cell walls, as the target oligosac-
charide. It has been reported previously that d-galactan is a
central supporting structure of the cell wall of Mycobacterium
tuberculosis,[7] and essential for cell growth and survival in the
host. In addition, the main constituents (d-galactofuranose
residues) are not found in mammalian metabolism. There-
fore, d-galactans have been the subject of much attention as
drug targets for new antituberculosis drugs without delete-
rious side effects.[8,9] Arylboronic acid was chosen as a
recognition moiety. It is widely known that phenylboronic
acid can bind to 1,2- or 1,3-diols through reversible boronate
formation under physiological conditions.[10,11] Thus, the use
of boronic acids is considered to be a promising approach for
carbohydrate recognition.[12] Norrild and co-workers reported
that d-glucose binds to boronic acids in water in its weakly
populated furanose form but not in its pyranose form;[13]
therefore, we envisaged that arylboronic acids would bind
preferentially to an acyclic 1,2-diol on the exocyclic side chain
of d-galactofuranoside, compared to 1,2- or 1,3-diols on the
pyranosides, which are the majority of the sugar components
in mammalian cells.
In our previous studies, certain anthraquinone (AQ)
derivatives were found to be capable of degrading not only
DNA[5] but also carbohydrates including b- and g-cyclo-
[*] Dr. D. Takahashi, S. Hirono, Prof. Dr. K. Toshima
Department of Applied Chemistry
To confirm the difference in the affinity of phenylboronic
acid (1) for methyl b-d-galactofuranoside (3) and methyl b-d-
galactopyranoside (4; Scheme 1), we carried out a quantita-
Faculty of Science and Technology, Keio University
3-14-1 Hiyoshi, Kohoku-ku, Yokohama 223-8522 (Japan)
Fax: (+81)45-566-1576
E-mail: toshima@applc.keio.ac.jp
C. Hayashi, Dr. M. Igarashi, Dr. Y. Nishimura
Institute of Microbial Chemistry
3-14-23 Kamiosaki, Shinagawa-ku, Tokyo 141-0021 (Japan)
[**] This research was supported in part by High-Tech Research Center
Project for Private Universities (Matching Fund Subsidy, 2006–
2011) and by Grants-in-Aid for Young Scientists (B) (No. 22710220)
and Scientific Research (B) (No. 20310140) from the Ministry of
Education, Culture, Sports, Science and Technology of Japan
(MEXT).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201005161.
Scheme 1. Chemical structures of phenylboronic acid (1) and several
diols.
10096
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 10096 –10100

Angewandte
Chemie
tive three-component alizarin red S (ARS) assay[14] with
fluorescence measurements in neutral (buffered pH 7.4)
water. As a result, the Ka value of 3 with respect to 1 (Ka =
158) was found to be 30-fold greater than that of 4 with 1
(Ka = 5), and, interestingly, also greater than that of d-
mannitol (2) with 1 (Ka = 126).
On the basis of these favorable preliminary findings, we
attempted to design a small molecule in which the AQ
derivative was attached to an arylboronic acid that had a high
affinity for d-galactofuranoside, thereby creating a hybrid
consisting of both degradation and recognition sites for a
target oligosaccharide. Our designed hybrid molecules 5 and 6
are shown in Scheme 2a. We chose a pyridinium boronic acid
expected that the photodegradation ability of the hybrid
molecules would be influenced by the length of the spacer
(Scheme 2b).
After chemical synthesis of our designed AQ/boronic acid
hybrids 5 and 6 (see Scheme S1 in the Supporting Informa-
tion), we carried out ARS assays of 5 with several glycosides,
including methyl b-d-galactofuranoside (3), methyl b-d-
galactopyranoside (4), methyl a-d-glucopyranoside (8),
methyl a-d-mannopyranoside (9), methyl 6-O-(b-d-galacto-
furanosyl)b-d-galactofuranoside (10), methyl a-d-maltoside
(11) methyl b-d-lactoside (12), and N-acetylneuraminic acid
a-methylglycoside (Neu5Aca2Me) (13). One of the target
glycosides, 10, was synthesized as shown in Scheme S2 on the
Supporting Information.
The results of the ARS binding assay are summarized in
Table 1. It was found that 5 effectively bound to d-galacto-
furanosides 3 and 10 (entries 1 and 5); the Ka value obtained
for 5 with 3 was slightly greater than that obtained for 1 with 3.
In sharp contrast the other glycosides, including Neu5A-
ca2Me (13), which possesses an extended glycerol side chain,
showed low affinities with 5 (entries 2–4 and 6–8). These
results clearly indicate that our designed molecule 5 binds
Table 1: Association constant (Ka) for hybrid molecule 5 with different
diols using ARS-based fluorescent method at neutral pH.[a]
Scheme 2. One of the expected photodegradation pathways of d-
galactofuranoside by a designed AQ/boronic acid hybrid (only oxida-
tive cleavage of glycosidic bond is shown). a) Chemical structures of
AQ/boronic acid hybrids 5 and 6. b) One of the expected photo-
degradation pathways of d-galactofuranoside by a radical species
produced by photoexcitation of the AQ moiety of hybrid 5 and O2.
as a recognition moiety because a positive charge on the
pyridine ring was expected to increase both the Lewis acidity
of the boronic acid for tighter binding to the target
oligosaccharide, and the water solubility of the hybrid
molecule. In considering the length of the spacer moiety, we
took into account the distance between the d-galactofurano-
side ring and the AQ moiety after complexation of the hybrid
with the 1,2-diol on the target oligosaccharide. On the basis of
our previous work on the photodegradation of oligosacchar-
ides[6] using light-activated AQ derivatives, and work reported
by other groups on the oxidative scission of nucleotides[15] and
oligosaccharides[16] by radical species, it seems certain that the
abstraction of the anomeric and other protons from the
glycoside is the one of the major driving forces. Thus, we
Entry Diol
Ka [mÀ1
]
1
2
3
4
5
methyl b-d-galactofuranoside (3)
170
7
6
8
174
methyl b-d-galactopyranoside (4)
methyl b-d-glucopyranoside (8)
methyl a-d-mannopyranoside (9)
methyl 6-O-(b-d-galactofuranosyl)b-d-galactofurano-
side (10)
6
7
8
methyl a-d-maltoside (11)
methyl b-d-lactoside (12)
N-acetylneuraminic acid a-methylglycoside (13)
1
1
5
[a] The Ka values were determined in 10% MeCN/0.1m sodium
phosphate buffer (pH 7.4) and are the average of at least two
reproducible measurements.
Angew. Chem. Int. Ed. 2010, 49, 10096 –10100
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 10097

Communications
favorably to acyclic 1,2-diols on the target glycosides rather
These degradation phenomena were confirmed by ESI/
TOF MS analysis after the photoreaction and subsequent
acetylation of the resulting products. The MS peak corre-
sponding to the acetylated disaccharide 15 disappeared only
after incubation of 10 with 5 under photoirradiation. In
addition, the MS peak corresponding to the acetylated
monosaccharide 14, which resulted from the cleavage of 10
and subsequent acetylation, was detected as one of the major
peaks (see Figure S1 in the Supporting Information). HPLC
analysis also indicated the presence of 14, and the chemical
yield was calculated, based on the peak area, to be 13%.
Next, we conducted electron paramagnetic resonance
(EPR) spin trapping experiments using 5,5-dimethyl-1-pyrro-
line-N-oxide (DMPO) to confirm the generation of reactive
oxygen species from the reaction of photoexcited AQ
derivatives with O2. It was found that photoirradiation of 5
in the presence of DMPO gave a product having an EPR
spectrum that is characteristic of the DMPO hydroxyl radical
spin adduct DMPO/COH (Figure 2b), which results from the
than to 1,2- or 1,3-diols on other pyranosides and a glycerol
side chain on sialoside.[17]
Next, we examined the photoinduced oligosaccharide-
degrading activities of 5 and 6 (1.0 mm) against glycosides 3, 4,
8, 9, 10, 11, 12, and 13 (1.0 mm each) in a 10% MeCN/0.1m
phosphate buffer (pH 7.4) at 258C for 10 minutes using a
long-wavelength UV light (365 nm, 100 W) that was placed
10 cm from the sample. The progress of the photodegradation
reaction was monitored by HPLC/UV (210 nm or 254 nm)
analysis of the resulting photodegradation products after their
acetylation (Ac2O, pyridine; for 3, 4, and 8–12) or tert-
butyldiphenylsilylation (TBDPS) of
a primary alcohol
(TBDPSCl/imidazole in N,N-dimethylformamide; for 13)
and a subsequent short-pass silica gel column purification.
The percentage degradation was calculated based on the peak
area corresponding to each peracetylated glycoside or 9-O-
TBDPS-Neu5Aca2Me, and the results are summarized in
Figure 1. When 2-hydroxymethyl AQ (7) was used as a
Figure 1. Photodegradation of glycosides by AQ/boronic acid hybrids.
Each glycoside (1.0 mm) was incubated with AQ/boronic acid hybrid 5,
6, or 2-hydroxymethyl AQ (7) (1.0 mm) in 10% MeCN/0.1m phosphate
buffer (100 mL, pH 7.4) at 258C for 10 min under irradiation with a UV
lamp (365 nm, 100 W) placed 10 cm from the sample, and analyzed by
HPLC methods (Mightysil RP-18 GP 5 mm, 4.6ꢀ150 mm; 408C;
detection by UV (210 nm or 254 nm) after total acetylation or silylation
of a primary alcohol of the photodegradation products.
Figure 2. EPR spectrum obtained during photoirradiation of AQ/bor-
onic acid hybrids in the presence of DMPO. AQ/boronic acid hybrid 5
(500 mm) and DMPO (500 mm) were incubated in 10%MeCN/0.1m
phosphate buffer (pH 7.4) containing 1.0 mm DETAPAC under irradi-
ation with a UV lamp (365 nm, 100 W) placed 40 cm from a flat cell.
a) Before irradiation. b) After 1 min irradiation. c) Possible pathways
for the formation of DMPO/·OH. DETAPAC=diethylenetriaminepenta-
acetic acid.
control, less than 10% degradation of the glycosides took
place owing to its low affinity for glycosides. However, when 5
or 6 was exposed to glycoside 10 under photoirradiation,
significant degradation took place, and the degradation
activities of 5 and 6 were found to be almost the same.
Glycoside 3 also underwent degradation by 5 and 6 under
photoirradiation, although the efficiency was lower than that
for 10. These results clearly indicate that the selectivity of the
hybrids, and thus their degradation ability, is highly depen-
dent upon their affinity with the target glycosides. Moreover,
it should be noted that the degradation of disaccharide 10 by 5
and 6 was much more effective than for the monosaccharide
3. In addition, 5, which possesses a short spacer, showed
higher activity in the degradation of 3 than 6, which suggests
that the number of protons on the glycoside that can react
with the photoactivated AQ has a significant influence upon
the degradation ability of the hybrid molecule.
reaction of DMPO with COH or the decay of the DMPO
superoxide anion spin adduct DMPO–COOH (Figure 2c).[18,19]
In addition, it was confirmed that no peaks corresponding to
DMPO/COOH or DMPO/COH were detected either by treat-
ment of DMPO with 5 without photoirradiation or by
photoirradiation of DMPO in the absence of 5 (Figure 2a).
These results indicate that oligosaccharide degradation
(including oxidative cleavage of the glycosidic bond) arises
10098 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 10096 –10100

Angewandte
Chemie
from a radical species produced by the reaction of photo-
excited AQ derivatives with O2.
Finally, we examined the ability of our AQ/boronic acid
hybrids to inhibit growth in Mycobacterium bovis BCG
(ATCC 35734), which possesses the target oligosaccharides
on the cell wall, in the presence and absence of photo-
irradiation (368 nm, 30 W, 5 min). Bacterial growth was
determined based on the number of colony-forming units
(CFU) after 21 days of incubation at 378C. The results are
summarized in Figure 3a. When photoirradiation was carried
out in the absence of the AQ/boronic acid hybrid, the CFU
numbers were almost the same as those for the control
(Figure 3a , lanes 1 and 2). When the cells were exposed to 5,
6, and
7 without photoirradiation, the CFU numbers
decreased slightly in a dose-dependent manner as a result of
the cytotoxicity of the compounds (Figure 3a , lanes 3, 5, 7, 9,
11, and 13). However, when the bacterial cells were exposed
to photoirradiation in the presence of 5 or 6, the CFU
numbers decreased significantly in line with the concentration
of each compound (Figure 3a , lanes 4, 6, 8, and 10), and in the
presence of 7 the CFU numbers did not decrease in a
concentration-dependent manner, even with photoirradiation
(Figure 3a , lanes 12 and 14).
In the case of Staphylococcus aureus (FDA209P), which
does not possess the target oligosaccharides in its cell wall,[20]
it was confirmed that 5, at a concentration of both 10 and
100 mgmLÀ1, did not show bactericidal activity even under
photoirradiation (Figure 3b, lanes 2 and 4). These results
indicate that the AQ/boronic acid hybrids exhibited selective
bactericidal activities against BCG.
To reveal whether 5 degraded the mycobacterial cell wall
component with photoirradiation, the cell wall and DNA of
BCG were detected by applying a combination of two
fluorescent dyes (SynaptoRed for the cell wall and SYBR
green for DNA) to samples incubated in the presence of 5
with or without photoirradiation, and subsequently analyzed
using confocal laser scanning microscopy. The resulting
images are shown in Figure 3c and d. A comparison of
these figures clearly shows that 5 caused effective degradation
of the cell wall but not of the DNA under photoirradiation.
Although we have not yet obtained direct evidence of
photodegradation of the target oligosaccharides on the
mycobacterial cell wall, these results suggest that selective
photodegradation of the target oligosaccharides on the
bacterial cell wall by 5 took place and induced significant
whole-cell destruction.
In conclusion, we have developed a new and innovative
method for selective degradation of a target oligosaccharide
by photoirradiation using AQ/boronic acid hybrids under
neutral conditions. The results presented herein will contrib-
ute to the molecular design of novel artificial oligosaccharide
photodegradation agents. The development of oligosacchar-
ide-photodegrading agents having high specificity and diver-
sity is now under investigation in our laboratories. We hope
this method will provide a means of controlling the specific
functions of certain oligosaccharides.
Figure 3. Bactericidal assay of AQ/boronic acid hybrids against BCG
and S. aureus. a) Each compound (5–7, 10 and 100 mgmLÀ1) was
incubated with BCG under photoirradiation (368 nm, 30 W) for 0 and
5 min using a UV lamp placed 20 cm from the samples, and the
bactericidal activity of each compound was determined by CFU after
21 days of incubation at 378C. b) Hybrid 5 (10 and 100 mgmLÀ1) was
incubated with S. aureus under photoirradiation (368 nm, 30 W) for 0
and 5 min using a UV lamp placed 20 cm from the samples. The
bactericidal activity of 5 was determined by CFU after an 18 h
incubation at 378C. Visualization of cell wall and DNA of BCG with
different fluorescent dyes by confocal laser scanning microscopy after
treatment with 5 without photoirradiation (c) or with photoirradiation
(d). DNA is shown in green (left) and the cell wall is shown in red
(middle); a composite image is also displayed (right).
Received: August 18, 2010
Published online: November 25, 2010
Angew. Chem. Int. Ed. 2010, 49, 10096 –10100
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 10099

Communications
[10] J. P. Lorand, J. O. Edwards, J. Org. Chem. 1959, 24, 769 – 774.
[11] G. Springsteen, B. Wang, Tetrahedron 2002, 58, 5291 – 5300.
[12] T. D. James in Boronic acids in Organic Synthesis and Chemical
Biology (Ed.: D. G. Hall), Wiley-VCH, Weinheim, 2005,
pp. 441 – 479.
[13] J. C. Norrild, H. Eggert, J. Am. Chem. Soc. 1995, 117, 1479 –
1484.
[14] G. Springsteen, B. Wang, Chem. Commun. 2001, 1608 – 1609.
[15] W. K. Pogozelski, T. D. Tullius, Chem. Rev. 1998, 98, 1089 – 1107.
[16] B. C. Gilbert, D. M. King, B. Thomas, Carbohydr. Res. 1984, 125,
217 – 235.
[17] K. Djanashvili, L. Frullano, J. A. Peters, Chem. Eur. J. 2005, 11,
4010 – 4018.
[18] E. Wertz, B. Bolton in Electron Spin Resonance, McGraw-Hill,
New York, 1972.
[19] H. M. Swartz, J. R. Bolton, D. C. Borg in Biological Application
of Electron Spin Resonance, Wiley-Interscience, New York,
1972.
[20] W. W. Navarre, O. Schneewind, Microbiol. Mol. Biol. Rev. 1999,
63, 174 – 229.
Keywords: anthraquinones · boronic acids · carbohydrates ·
Mycobacterium tuberculosis · photochemistry
.
[1] Comprehensive Glycoscience (Eds: J. P. Kamerling, G.-J. Boons,
Y. C. Lee, A. Suzuki, N. Taniguchi, A. G. J. Voragen), Elsevier,
Oxford, 2007.
[2] J. B. Lowe, J. D. Marth, Annu. Rev. Biochem. 2003, 72, 643 – 691.
[3] E. Ioffe, P. Stanley, Proc. Natl. Acad. Sci. USA 1994, 91, 728 –
732.
[4] M. Metzler, A. Gertz, M. Sarkar, H. Schachter, J. W. Schrader,
J. D. Marth, EMBO J. 1994, 13, 2056 – 2065.
[5] K. Toshima, Y. Maeda, H. Ouchi, A. Asai, S. Matsumura,
Bioorg. Med. Chem. Lett. 2000, 10, 2163 – 2165.
[6] M. Ishii, S. Matsumura, K. Toshima, Angew. Chem. 2007, 119,
8548 – 8551; Angew. Chem. Int. Ed. 2007, 46, 8396 – 8399.
[7] M. Daffe, P. J. Brennan, M. McNeil, J. Biol. Chem. 1990, 265,
6734 – 6743.
ˇ
[8] K. Mikusovꢀ, T. Yagi, R. Stern, M. R. McNeil, G. S. Besra, D. C.
Crick, P. J Brennan, J. Biol. Chem. 2000, 275, 33890 – 33897.
[9] M. S. Scherman, K. A. Winans, R. J. Stern, V. Jones, C. R.
Bertozzi, M. R. McNeil, Antimicrob. Agents Chemother. 2003,
47, 378 – 382.
10100 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 10096 –10100