Probing a sialyltransferase's recognition domain to prepare α(2,8)-linked oligosialosides and analogs

Article abstract of DOI:10.1039/b908933k

We report the first observation that a monosialyl residue is the essential structural element recognized by the enzyme CST-II; this has resulted in an attractive route to synthesize a series of α(2,8)-linked oligosialic acids and their thioanalogs in one

Full text of DOI:10.1039/b908933k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Probing a sialyltransferase’s recognition domain to prepare a(2,8)-linked
oligosialosides and analogsw
Ping Zhang, Amir J. Zuccolo, Wenling Li, Ruixiang Blake Zheng and Chang-Chun Ling*
Received (in College Park, MD, USA) 5th May 2009, Accepted 1st June 2009
First published as an Advance Article on the web 19th June 2009
DOI: 10.1039/b908933k
We report the first observation that a monosialyl residue is the
essential structural element recognized by the enzyme CST-II;
this has resulted in an attractive route to synthesize a series of
a(2,8)-linked oligosialic acids and their thioanalogs in one step.
desirable because functional groups can be incorporated into
the sialoside a priori. This approach represents one of the
most challenging tasks in carbohydrate chemistry and has
encountered considerable difficulties.11 Many strategies have
been reported in the literature to overcome these challenges
with varying degrees of success.12 To date, the longest oligomer
synthezised is an a(2,8)-linked tetramer which was obtained in
a stepwise manner.12h
a(2,8)-Linked oligosialosides (OSAs, Fig. 1) containing up
to four sialyl residues are important structural elements com-
monly found in gangliosides.1 Recently, a(2,8)-OSAs with up
to seven sialic acid residues have been found in a large number
of neural glycoproteins.2 These sialyl residues, mainly consist-
ing of N-acetylneuraminic acid (NeuNAc), are known to play
important roles in cell adhesion, differentiation and signal
transduction.3 a(2,8)-Linked polysialic acids4 (PSAs) are also
naturally occurring and play crucial roles in various bio-
logical mechanisms such as human neural development and
regeneration,5 bacterial virulence6 and cancer metastasis.7
a(2,8)-OSAs/PSAs with longer chain lengths (48 residues)
form helical conformations which are linked to their biological
functions.2b Carbohydrate antibodies specific to a(2,8)-OSAs/PSAs
have widespread utility in cancer therapies and prevention of
bacterial infection.8
Herein we wish to report an efficient, one-step chemo-
enzymatic synthesis of a(2,8)-linked OSAs. The success of this
approach has been demonstrated by the synthesis of a series of
a(2,8)-linked S- and O-oligosialosides. In 2004 Wakarchuk’s
group13 showed that CST-II, an a(2,8)-sialyltransferase cloned
from a gene extracted from Campylobacter jejuni OH4384, is
capable of carrying out sequential a(2,8)-sialylations on
GM3-based substrates to produce
a series of complex
ganglioside structures, GD3, GT3 and a tetrasialosyl analog.14
GM3 is a trisaccharide that has a NeuNAc residue linked to
the O-30 position of lactose. CST-II’s ability to perform
sequential sialylations intrigued us. When analyzing the
enzyme’s structural requirement for substrates, we considered
the possibility that CST-II might not necessarily require the
entire GM3 trisaccharide as its substrate, since the product
formed following the first sialyl transfer immediately served as
a substrate for the second sialyl transfer. Thus it is possible
that the structural element recognized by the enzyme might be
a small portion of the GM3 molecule and possibly only its
sialyl residue. If this hypothesis held true, CST-II would be
able to accept much simpler monosialosides as substrates
and still perform sequential a(2,8)-sialylations, leading to an
attractive route to synthesize a(2,8)-linked OSAs and PSAs
(Fig. 2). Transfer of more than three sialyl residues by the
enyzme had not previously been observed and it was uncertain
if this was due to its substrate requirements or its overall
efficiency. Despite this we were interested in finding conditions
that would permit the synthesis of higher oligomers.
Fig. 1 Structure of a(2,8)-oligo/polysialosides.
a(2,8)-Linked OSAs and PSAs are difficult molecules to
handle because of their ionic nature, high polarity, and fragile
a(2,8)-linkages which can lead to intramolecular lactonization.
a(2,8)-Linked OSAs, up to the pentamer, have been obtained
with a top-down approach through the controlled hydrolysis
of readily available colominic acid polymer.9 The resulting
mixture of unprotected oligosaccharides required tedious
separation; in order to introduce functional groups at the
reducing end, multiple chemical steps were required.10
Strategies based on bond formation between C-2 of one sialic
acid monomer and O-8 of another, a(2,8)-sialylation, are more
Two a-monosialosides 1 and 2 (Fig. 2) were prepared to
investigate CST-II’s ability to recognize them as substrates
and perform a(2,8)-sialylations. A long dodecyl chain was
chosen as the aglycone to facilitate the subsequent purification
by reverse-phase chromatography. Aglycones longer than
12 carbons have the potential to be more advantageous in
easing the purification of the products formed; however, the
solubility of the starting material in aqueous media was
also a requirement. Although a-monosialoside 2 bears more
similarity to the enzyme’s natural substrate than 1 because
it is also a 2-O-glycoside, 1 is an attractive substrate because
it is synthetically easier to prepare. Additionally if the thio-
glycoside 1 was found to be accepted by CST-II, this would
Alberta Ingenuity Centre for Carbohydrate Science, Department of
Chemistry, University of Calgary, 2500, University Drive NW,
Calgary, Alberta, Canada T2N 1N4. E-mail: ccling@ucalgary.ca;
Fax: +1 403-289-9488; Tel: +1 403-220-2768
w Electronic supplementary information (ESI) available: Experimental
procedures, NMR data, and NMR spectra of synthesized compounds.
See DOI: 10.1039/b908933k
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 4233–4235 | 4233

View Article Online
CST-II smoothly converted 1 to a series of more polar
compounds (Scheme 3) easily detected by TLC. A combina-
tion of gel filtration chromatography on Sephadex LH-20 and
reverse-phase HPLC on C18 silica gel were found to be
optimal conditions for purification. Compounds 8 through
10 were isolated with yields varying from 5 to 12%.
Scheme 3 Enzymatic a(2,8)-sialylation of a-thiosialoside 1.
When O-sialoside 2 was subjected to the same enzymatic
reaction conditions (Scheme 4), the experiments proved to be
equally successful. a(2,8)-Linked sialyl dimer 11, trimer 12 and
tetramer 13 were obtained in similar yields, 15%, 7.5% and
5%, respectively.
Fig.
2
Hypothesis and potential substrates of CST-II for
a(2,8)-sialylations.
expand its synthetic scope allowing access to a wide range of
thioanalogs related to a(2,8)-OSAs.
Thiosialoside 1 was prepared in large quantities with high
stereoselectivity using established methodology.15 The syn-
thesis (Scheme 1) began with 316 which when treated with
1-iodododecane and diethylamine in DMF afforded thioglycoside
4 in 70% yield. A Zemplen transesterification followed by an
´
in situ saponification of the methyl ester at C-1 afforded
compound 1 in 75% yield.
Scheme 4 Enzymatic a(2,8)-sialylation of a-sialoside 2.
After obtaining the lower oligomers 11–13, we were inter-
ested in preparing higher oligomers (44 residues, Scheme 4).
We attempted to vary the reaction conditions in a systematic
manner by adjusting the amounts of CMP-NeuNAc, enzyme
units or concentrations of the substrate; the concentration of
the substrates in the reaction media was found to be critically
important and needed to be increased. This was effectively
achieved by decreasing the volume of the CST-II enzyme
(0.3 U per mL) used; although this resulted in a decrease of
the amount of CST-II used for the reaction, we observed a
significant increase in the formation of higher oligomers. The
best result was obtained when CST-II was added in a portion-
wise fashion.
Scheme 1 Synthesis of a-thiosialoside 1.
The synthesis of O-sialoside 2 began with b-thioglycoside
donor 6 which was prepared from 5 in 80% yield (Scheme 2).17
N-Iodosuccinimide–triflic acid promoted glycosylation of 6
with n-dodecanol afforded the desired a-anomer 7 in 59%
yield which could be easily separated from the b-anomer (12%
yield, not shown). Utilising the same deprotection sequences
as used for 4 on 7, a-sialoside 2 was obtained in 91% yield.
An experiment with 42 mg of 2 was carried out by adding
both CST-II (0.3 U per mL, 20 mL total) and CMP-NeuNAc
(4 equiv. total) in two equal portions with a 16 h interval—
allowing the reaction to begin at a higher substrate concentra-
tion (4.2 mg mLꢁ1); with this strategy all compounds from
dimer 11 to octamer 17 were formed and could be isolated
(see ESIw) in milligram quantities following one round of
purifications as before (Scheme 4); the yields of the lower
oligomers 11, 12 and 13 were 9.6%, 7.1% and 6.9%, respec-
tively, while the pentamer 14, hexamer 15, heptamer 16 and
Scheme 2 Synthesis of a-O-sialoside 2.
Thioglycoside 1 was then subjected to an enzymatic reaction
(concentration 0.8 mg mLꢁ1) with 2.8 equiv. of CMP-NeuNAc
under the catalysis of crude CST-II isolated from lysed
bacterial culture (0.3 U per mL, 10.5 U). To our delight,
ꢀc
This journal is The Royal Society of Chemistry 2009
4234 | Chem. Commun., 2009, 4233–4235

View Article Online
octamer 18 were isolated in 3.5%, 3.9%, 1.5% and 1.7%
yields, respectively. Trace amounts of nonamer and decamer
were also confirmed by mass spectrometry. It is noteworthy
that higher oligomers from pentamer to octamer have not been
synthesized before and when compared with the chemical
synthesis of a(2,8)-tetrasialoside,12h which required time con-
suming multi-step procedures, this new route is advantageous
because of its overall efficiency and simplicity.
It has also been demonstrated for the first time that CST-II
recognizes 2-S-thiosialosides as substrates. As thiosialosides
have widespread utility as mimics and enzyme inhibitors, this
could prompt the synthesis of other functionalized analogs of
a(2,8)-oligosialosides for various applications including the
use as biological probes and vaccine components.
We thank Dr W. Wakarchuk and Dr M. Gilbert from the
Institute for Biological Sciences, NRCC, for providing clones
of the enzymes and Prof. D. Bundle and Dr A. Morales-
Izquierdo from the Department of Chemistry, University of
Alberta, for acquiring mass spectra. The financial support
from the Alberta Ingenuity, the Natural Sciences and
Engineering Research Council of Canada and the University
of Calgary is acknowledged.
The structural difference between GM3 and monosialosides
1 and 2 is the replacement of the lactose unit with a dodecyl
group. The ability of CST-II to accept all compounds and
their sequentially sialylated congeners for a(2,8)-sialylations
supports our initial hypothesis that the sialyl unit is the
fragment recognized by the enzyme. For futher confirmation,
disaccharide 18 (Scheme 5), which is structurally distinct
from both GM3 and substrates 1 and 2 because it has an
a(2,6)-linked galactose unit, was designed for testing in the
Notes and references
1 T. Angat and A. Varki, Chem. Rev., 2002, 102, 439.
2 (a) C. Sato, H. Fukuoka, K. Ohta, T. Matsuda, R. Koshino,
K. Kobayashi, F. A. Troy II and K. Kitajima, J. Biol. Chem., 2000,
275, 15422; (b) C. Sato and K. Kitajima, Trends Glycosci.
Glycotechnol., 1999, 11, 371.
reaction. The synthesis began with b-thiogalactoside 19.18
A
selective silylation at the O-6 position followed by a perbenzoyl-
ation gave intermediate 20, which was subjected to the selective
removal of the silyl group to afford 21 in 75% yield over three
steps; after a triflation, activated ester 22 was coupled with
thioacetate 3 to give disaccharide 23 in 30% yield. Sequential
deprotections gave the disaccharide 18 in 75% yield.
3 A. Varki, Trends Mol. Med., 2008, 14, 351.
4 H. J. Jennings, in Carbohydrate-Based Drug Discovery, ed. C.-H.
Wong, Wiley-VCH, Weinheim, 2003, p. 357.
5 F. A. Troy, Glycobiology, 1992, 2, 5.
´
6 L. Jodar, I. M. Feavers, D. Salisbury and D. M. Granoff, Lancet,
2002, 359, 1499.
7 (a) S. Hakomori, Cancer Res., 1996, 56, 5309; (b) M. Suzuki,
M. Suzuki, J. Nakayama, A. Suzuki, K. Angata, S. Chen,
K. Sakai, K. Hagihara, Y. Yamaguchi and M. Fukuda, Glycobiology,
2005, 15, 887.
8 V. S. Joshia, I. B. Bajaja, S. A. Survasea, R. S. Singhala and
J. F. Kennedy, Carbohydr. Polym., 2009, 75, 553.
9 R. Roy and R. A. Pon, Glycoconjugate J., 1990, 7, 3.
10 H. Ando, H. Ishida, M. Kiso and A. Hasegawa, Carbohydr. Res.,
1997, 300, 207.
11 G.-J. Boons and A. V. Demchenko, Chem. Rev., 2000, 100, 4539.
12 (a) Y. Ito, M. Numata, M. Sugimoto and T. Ogawa, J. Am. Chem.
Soc., 1989, 111, 8508; (b) J. C. Castro-Palomino, Y. E. Tsvetkov
and R. R. Schmidt, J. Am. Chem. Soc., 1998, 120, 5434; (c) C. De
Meo, A. V. Demchenko and G.-J. Boons, J. Org. Chem., 2001, 66,
5490; (d) A. V. Demchenko and G.-J. Boons, Tetrahedron Lett.,
1998, 39, 3065; (e) H. Tanaka, M. Adachi and T. Takahashi,
Chem.–Eur. J., 2005, 11, 849; (f) K. Tanaka, T. Goi and K. Fukase,
Synlett, 2005, 2958; (g) C.-S. Yu, K. Niikura, C.-C. Lin and
C.-H. Wong, Angew. Chem., Int. Ed., 2001, 40, 2900;
(h) H. Tanaka, Y. Nishiura and T. Takahashi, J. Am. Chem.
Soc., 2006, 128, 7124.
Scheme 5 Synthesis of disaccharide 18.
Compound 18 was subjected to a(2,8)-sialylation by CST-II.
The result showed that the enzyme recognized the disaccharide
and smoothly converted it to trisaccharide 24 (48% yield)
under similar conditions19 (Scheme 6).
13 C. P. Chiu, A. G. Watts, L. L. Lairson, M. Gilbert, D. Lim,
W. W. Wakarchuk, S. G. Withers and N. C. Strynadka,
Nat. Struct. Mol. Biol., 2004, 11, 163, and references therein.
14 (a) O. Blixt, D. Vasiliu, K. Allin, N. Jacobsen, D. Warnock,
N. Razi, J. C. Paulson, S. Bernatchez, M. Gilbert and
W. W. Wakarchuk, Carbohydr. Res., 2005, 340, 1963;
(b) S. Jacques, J. R. Rich, C.-C. Ling and D. R. Bundle, Org.
Biomol. Chem., 2006, 4, 142.
15 (a) J. R. Rich and D. R. Bundle, Org. Lett., 2004, 6, 897;
(b) H. N. Yu, C.-C. Ling and D. R. Bundle, J. Chem. Soc., Perkin
Trans. 1, 2001, 832.
16 A. Hasegawa, J. Nakamura and M. Kiso, J. Carbohydr. Chem.,
1986, 5, 11.
Scheme 6 Enzymatic a(2,8)-sialylation of disaccharide 18.
In conclusion, convincing evidence has been provided that a
monosialyl residue is the key structural element recognized by
CST-II for a(2,8)-sialylations. This has resulted in an efficient
route to synthesize a class of a(2,8)-linked OSAs which have
been previously difficult to obtain using convential chemical
syntheses; the efficiency of the new route has been demon-
strated by the synthesis of a series of OSAs up to the octamer.
17 A. Marra and P. Sinay, Carbohydr. Res., 1989, 187, 35.
¨
18 S. Sugiyama and J. M. Diakur, Org. Lett., 2000, 2,
2713.
19 Formation of higher oligomers was observed by TLC, but they
were not isolated.
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 4233–4235 | 4235