Top-down chemoenzymatic approach to a high-mannose-type glycan library: Synthesis of a common precursor and its enzymatic trimming

Article abstract of DOI:10.1002/anie.201301613

From the stacks: A novel method for construction of a high-mannose-type glycan library by systematic enzymatic trimming of a single synthetic Man ₉-based precursor was developed. Efficient chemical synthesis of the tetradecasaccharide common precursor and orthogonal enzymatic trimming to obtain all M_8-9 and G₁M_8-9 derivatives was demonstrated. G=glucose, M=mannose. Copyright

Full text of DOI:10.1002/anie.201301613

.
Angewandte
Communications
DOI: 10.1002/anie.201301613
Carbohydrates
Top-Down Chemoenzymatic Approach to a High-Mannose-Type
Glycan Library: Synthesis of a Common Precursor and Its Enzymatic
Trimming**
Akihiko Koizumi, Ichiro Matsuo, Maki Takatani, Akira Seko, Masakazu Hachisu,
Yoichi Takeda,* and Yukishige Ito*
Processing of asparagine (Asn)-linked (N-linked) glycans
creates a variety of glycoforms which play important roles in
a variety of intracellular events.[1] In the endoplasmic
reticulum (ER) a combination of various enzymes, chaper-
ones, lectins, and cargo receptors constitutes the glycoprotein
quality control (GQC) system.[2]
N-linked glycans are co-translationally introduced to
nascent polypeptides initially as a triglucosylated tetradeca-
saccharide Glc3Man9GlcNAc2 (G3M9)[3] by oligosaccharyl-
transferase (OST; Figure 1).[4] Subsequently, trimming by
glucosidase I (G-I) and glucosidase II (G-II) produces the
monoglucosylated glycoform Glc1Man9GlcNAc2 (G1M9).[5]
Further removal of the remaining Glc by G-II forms the
nonglucosylated glycan Man9GlcNAc2 (M9). All the above-
mentioned glycans are potentially susceptible to trimming by
ER mannosidases,[6] which may generate glucosylated as well
as nonglucosylated glycans with a variable number of Man
residues (e.g. G1M8, G1M7, M8, M7, etc.).
CRT cycle).[10] Simultaneously, misfolded glycoproteins are
gradually demannosylated to Man8GlcNAc2 (M8),
Man7GlcNAc2 (M7), and Man5-6GlcNAc2 (M5-6) by ER
mannosidase and delivered to the cytosol for ER-associated
degradation (ERAD).[11]
Our previous effort has established synthetic routes to
most of the ER-related high-mannose-type glycans,[12] which
have proven valuable in clarifying specificities of CRT,[8b,13]
Fbs1,[14] UGGT,[15] and G-II.[16] To facilitate more compre-
hensive analysis of glycan functions in the GQC, we initiated
the study to construct a high-mannose-type glycan library
through top-down diversification of a common precursor.
Although our plan was hinted at by the biosynthetic
pathway of N-linked glycans,[17] it is obviously unrealistic to
reconstitute the biosynthetic pathway to generate the high-
mannose glycan library. For instance, both G1M9 and M9
have multiple a1,2-mannosyl linkages, thus prohibiting con-
trolled trimming of one of these Man residues (Figure 2a). To
Folding of glycoproteins is aided by a variety of chaper-
ones in the ER.[7] Among them, calnexin (CNX) and
calreticulin (CRT) occupy a privileged position as lectin
chaperones, which are able to specifically recognize Glca1!
3Man-containing glycans, such as G1M9.[8]
In the ER, UDP-glucose:glycoprotein glucosyltransferase
(UGGT) operates as the folding sensor, by which folding
defective glycoproteins are re-glucosylated.[9] A combination
of CNX/CRT, G-II, and UGGT constitutes a cycle (CNX/
[*] Dr. A. Koizumi, M. Takatani, Dr. A. Seko, Dr. M. Hachisu,
Dr. Y. Takeda, Dr. Y. Ito
ERATO
Ito Glycotrilogy Project (Japan) Science and Technology Agency
2-1 Hirosawa, Wako, Saitama 351-0198 (Japan)
E-mail: yotakeda@riken.jp
yukito@riken.jp
Prof. I. Matsuo
Department of Chemistry and Chemical Biology, Gunma University
1-5-1 Tenjin-cho, Kiryu, Gunma 376-8515 (Japan)
Dr. Y. Ito
Scheme 1. Synthesis of Glc1Man4GlcNAc2 heptasaccharide acceptor.
Reagents and conditions: a) [Cp2HfCl2], AgOTf, 4 ꢀ M.S., CH2Cl2,
ꢀ108C, 3.5 h, 85%. b) HF-pyridine/pyridine (1:5), 08C to RT, 21 h,
75%. c) MeOTf in ClCH2CH2Cl, 4 ꢀ M.S., CH2Cl2, ꢀ408C to RT, 12 h,
73%. d) HF-pyridine, pyridine, 1 GPa, RT, 21 h, quant. e) [Cp2HfCl2],
AgOTf, 4 ꢀ M.S., toluene/Et2O (10:1), ꢀ40 to ꢀ108C, 21 h, 86%.
f) TsOH·H2O, MeCN, RT, 3 h, 73%. Bz=benzoyl,
[Cp2HfCl2]=bis(cyclopentadienyl)hafnium dichloride, Phth=phthaloyl,
M.S.=molecular sieves, Tf=trifluoromethanesulfonyl, Ts=4-toluene-
sulfonyl.
Synthetic Cellular Chemistry Laboratory, RIKEN
2-1 Hirosawa, Wako, Saitama 351-0198 (Japan)
[**] We thank Dr. Akihiro Ishiwata for his valuable advice as wells as
Akemi Takahashi and Satoko Shirahata for their technical assis-
tance. We are also grateful to Dr. Yoshio Sakaguchi and his staff of
the Materials Characterization Team in RIKEN Advanced Support
Division for elemental analyses.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201301613.
7426
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 7426 –7431

Angewandte
Chemie
Figure 1. Glycoprotein quality control. OST=oligosaccharyltransferase, G-I=glucosidase I, G-II=glucosidase II, CNX=calnexin, CRT=calreticu-
lin, ERManI=ER mannosidase I, UGGT=UDP-glucose:glycoprotein glucosyltransferase, EDEM=ER degradation-enhancing a-mannosidase-like
protein, ERAD=ER-associated degradation.
overcome this problem, our strategy adopted the non-natural
precursor I, each branch of which was capped by a distinct
sugar residue (Figure 2b).[18] We anticipated that, if proper
selection of the terminal sugar residues A, B, and C has been
made, liberation of three arms can be executed indepen-
dently, thus giving the second-generation precursors I-A, I-B,
and I-C. Digestion with logically defined sets of glycosidases
will give a variety of high-mannose-type glycans.
To pursue this scheme, we designed the non-natural
tetradecasaccharide 1 as the precursor (Figure 3), in which A,
B, and C branches were capped by a-Glc, b-Gal, and b-
GlcNAc, respectively. We additionally planned to convert
1 into the BODIPY-labeled derivative 2 (BODIPY= boron-
dipyrromethene). Our plan required the fully protected
tetradecasaccharide 3. As depicted in Figure 3, 3 was dis-
sected into the three fragments 4, 5, and 6.
Namely, the disaccharide (Manb1-4GlcN) donor 7,[12a] which
was synthesized by intramolecular aglycon delivery,[19] was
coupled with 8 to give the trisaccharide 9, which was
desilylated to provide 10. Subsequent reaction with the
thioglycoside 11[12a] led to the hexasaccharide 12, whose tert-
butyldiphenylsilyl (TBDPS) group was removed[20] to give 13.
Its stereochemistry was confirmed by three signals, corre-
sponding to a-mannosyl linkages, at d > 5.0 ppm (d =
5.33 ppm, 5.25 ppm, 5.19 ppm). Subsequent glycosylation
was conducted with the fluoride 14.[12] The reaction was
promoted by [Cp2HfCl2]/AgOTf[21] exclusively to give the
desired isomer 15, acid treatment of which gave the A-arm-
containing heptasaccharide acceptor 4.
For the construction of the C- and B-arm fragments, 6 and
5, respectively, we opted to bifurcate the compound 16.[22]
[23]
Thus, treatment of 16 with LiAlH4/AlCl3 provided the 6-
To begin with, construction of the heptasaccharide
acceptor 4, which comprises the A arm and the core
trisaccharide, was carried out as depicted in Scheme 1.
OH product 17 (Scheme 2), which was protected by a penta-
fluoropropionyl (PFP) group[24] to give 18. The latter was used
for glycosylation with the mannose-derived acceptor 19,[25]
Angew. Chem. Int. Ed. 2013, 52, 7426 –7431
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7427

.
Angewandte
Communications
Figure 2. Overview of chemoenzymatic synthesis of high-mannose-type glycans. a) a1,2-Mannosidase-reactive sites of natural G1M9 and M9
glycans. b) Orthogonal liberation of A, B, and C arms of non-natural tetradecasaccharide for the construction of a high-mannose glycan library.
Figure 3. Retrosynthesis for the tetradecasaccharide 1 and BODIPY-labeled derivative 2.
which gave the coupled product 20. Removal of the PFP
group was conducted under weakly basic conditions (pyri-
dine/EtOH). The resultant 21 was subjected to coupling with
the glucosaminyl donor 22,[26] thus giving the trisaccharide 23.
Selective deacetylation gave 24, which in turn was converted
into the trichloroacetimidate 6.
To construct the B arm, reductive opening of 16 was
conducted by NaBH3CN/HCl/1,4-dioxane[27] to provide 25,
which was again protected by PFP to give 26. Glycosylation
with 19 promoted by MeOTf[28] gave 27, from which the PFP
group was removed quantitatively. Thus the obtained com-
pound 28 was glycosylated with the galactosyl chloride 29[29]
to give the trisaccharide 30. The anomeric acetate 30 was
converted into the trichloroacetimidate 32 via the hemiacetal
31, which was coupled with the 6-O-chloroacetylated thio-
glycoside 33,[30] thus giving the tetrasaccharide 34. Its
chloroacetyl group was removed[31] to give the tetrasaccharide
acceptor 5.
The compound 5 was then glycosylated with the C-arm
donor 6 to give the heptasaccharide 35 (Scheme 3). Based on
our previous experience,[12] the thioglycoside 35 was con-
verted into the fluoride 36. Coupling with the diol 4 gave 3,
which was fully deprotected to give 1. Having the designed
tetradecasaccharide in hand, fluorophoric modification was
conducted to facilitate the screening of enzymatic reactions
and analysis of glycan–protein interactions. The tetradeca-
7428 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 7426 –7431

Angewandte
Chemie
Scheme 2. Synthesis of the B-arm tetrasaccharide donor and the C-arm
trisaccharide acceptor. Reagents and conditions: a) LiAlH4, AlCl3, Et2O/
CH2Cl2 (1:1), 508C, 1.5 h, 93%. b) (PFP)2O, pyridine, CH2Cl2, RT, 1.5 h,
93%; c) DTBMP, MeOTf, 4 ꢀ M.S., ClCH2CH2Cl, 508C, 15 h, 87%.
d) Pyridine, EtOH, RT, overnight, quant. e) [Cp2HfCl2], AgOTf, 4 ꢀ
M.S., CH2Cl2, ꢀ408C, 12 h, 84%. f) NH2NH2·AcOH, DMF, RT, 2 h,
69%; g) CCl3CN, DBU, CH2Cl2, RT, 4 h, 84%. h) NaBH3CN, 4m HCl in
1,4-dioxane, THF, 4 ꢀ M.S., 08C, 9 h, 80%. i) (PFP)2O, pyridine,
CH2Cl2, RT, 1.5 h, 96%. j) DTBMP, MeOTf, 4 ꢀ M.S., ClCH2CH2Cl,
508C, 15 h, 82%. k) Pyridine, EtOH, 508C, overnight, quant. l) AgOTf
in ClCH2CH2Cl, 4 ꢀ M.S., toluene, 1008C, 18 h, 71%.
m) NH2NH2·AcOH, DMF, RT, 1.5 h, 85%. n) CCl3CN, DBU, CH2Cl2,
RT, 2 h, 87%. o) TMSOTf, 4 ꢀ M.S., CH2Cl2, ꢀ408C, overnight, 76%.
p) DABCO, EtOH/THF (2:1), RT, overnight, 91%. DBU=1,8-
diazabicyclo[5.4.0]undec-7-ene, DMF=N,N-dimethylformamide,
DTBMP=2,6-di-tert-butyl-4-methylpyridine, THF=tetrahydrofuran,
TMS=trimethylsilyl.
Figure 4. a) Enzymatic removal of GlcNAc, Gal, and Glc. b) Enzymatic
digestion by Golgi endo-a-mannosidase.
To remove the GlcNAc residue, Jack bean b-N-acetylhex-
osaminidase digestion was highly satisfactory, thus giving the
tridecasaccharide 37 smoothly (Figure 4a). However, the
b1,4-linked Gal residue was resistant to b-galactosidases
derived from bovine liver, Jack bean, Kluyveromyces lactis,
and Escherichia coli, presumably because of steric hindrance.
Only Aspergillus-oryzae-derived enzyme had substantial
activity to provide the B-arm-unblocked product 38. Sim-
ilarly, rat liver a-glucosidase II uniquely gave 39 by cleaving
the Glc residue.
Although removal of the a-Glc from the A arm of 2 was
straightforward, controlled removal of the terminal Man
would be difficult. Consequently, digestion with Golgi endo-
a-mannosidase was conducted, and cleanly gave 40 (Fig-
ure 4b). Importantly, the presence of terminal a-Glc con-
ferred susceptibility to endomannosidase.[34]
Scheme 3. Synthesis of the tetradecasaccharide 1 and labeling with
BODIPY. Reagents and conditions: a) TMSOTf, 4 ꢀ M.S., CH2Cl2,
ꢀ408C, overnight, 90%. b) DAST, NBS, CH2Cl2, ꢀ40 to ꢀ208C, 3 h,
93%. c) [Cp2HfCl2], AgOTf, 4 ꢀ M.S., toluene, ꢀ408C to RT, overnight,
54%. d) 1. ethylenediamine, nBuOH, 908C, 14 h; 2. Ac2O, pyridine, RT,
19 h; 3. NaOMe, MeOH, 408C, 21 h; 4. Pd(OH)2/C, H2 (gas)/MeOH-
H2O, RT, 24 h, 59%. e) 1. Sat. NH4HCO3 aq., 408C, overnight,
2. Fmoc-Gly-OPfp, pyridine, DMSO, 48C, 24 h; 3. piperidine, DMF, RT,
1 h; 4. BODIPY-FL-SE, DIPEA, DMF, RT, overnight, 55% (4 steps).
DAST=N,N-diethylaminosulfurtrifluoride, DMSO=dimethylsulfoxide,
DIPEA=diisopropylethylamine, Fmoc=9-fluorenylmethoxycarbonyl.
To demonstrate the feasibility of our strategy, we
attempted systematic preparation of mono- as well as non-
glucosylated M8 glycans (Figure 5), because of their impor-
tant roles in GQC.[2] Production of M8A was achieved by
a three-step enzymatic trimming with Jack bean b-N-acetyl-
hexosaminidase, Golgi endo-a-mannosidase, and A. oryzae b-
galactosidase. HPLC profiles for each step are shown in
Figure S1 of the Supporting Information. G1M8(B) was given
saccharide 1 was treated with aq. NH4HCO3,[32] thus giving the
glycosylamine, which was reacted with Fmoc glycine penta-
fluorophenyl ester (Fmoc-Gly-OPfp) in DMSO at temper-
atures below freezing.[33] Removal of the Fmoc group and
treatment with succinimide ester of BODIPY-FL gave the
fluorescently-labeled tetradecasaccharide 2 (Scheme 3).
Angew. Chem. Int. Ed. 2013, 52, 7426 –7431
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7429

.
Angewandte
Communications
Keywords: carbohydrates · enzymes · glycopeptides ·
.
protein folding · synthesis design
[1] a) D. N. Hebert, S. C. Garman, M. Molinari, Trends Cell Biol.
2005, 15, 364 – 370; b) A. J. Parodi, Biochem. J. 2000, 348, 1 – 13.
[2] a) D. N. Hebert, M. Molinari, Physiol. Rev. 2007, 87, 1377 – 1408;
b) R. Benyair, E. Ron, G. Z. Lederkremer, Int. Rev. Cell Mol.
Biol. 2011, 292, 197 – 280.
[3] a) S. J. Turco, B. Stetson, P. W. Robbins, Proc. Natl. Acad. Sci.
USA 1977, 74, 4411 – 4414; b) S. J. Turco, P. W. Robbins, J. Biol.
Chem. 1979, 254, 4560 – 4567.
[4] a) E. Bause, Biochem. J. 1983, 209, 331 – 336; b) P. Whitley, I. M.
Nilsson, G. von Heijne, J. Biol. Chem. 1996, 271, 6241 – 6244;
c) A. X. Yan, W. J. Lennarz, J. Biol. Chem. 2005, 280, 3121 –
3124.
[5] E. S. Trombetta, Glycobiology 2003, 13, 77R – 91R.
[6] S. W. Mast, K. W. Moremen, Methods Enzymol. 2006, 415, 31 –
46.
[7] a) P. Maattanen, K. Gehring, J. J. M. Bergeron, D. Y. Thomas,
Semin. Cell Dev. Biol. 2010, 21, 500 – 511; b) L. A. Rutkevich,
D. B. Williams, Curr. Opin. Cell Biol. 2011, 23, 157 – 166.
[8] a) A. R. Patil, C. J. Thomas, A. Surolia, J. Biol. Chem. 2000, 275,
24348 – 24356; b) A. Tatami, Y. S. Hon, I. Matsuo, M. Takatani,
H. Koshino, Y. Ito, Biochem. Biophys. Res. Commun. 2007, 364,
332 – 337; c) F. E. Ware, A. Vassilakos, P. A. Peterson, M. R.
Jackson, M. A. Lehrman, D. B. Williams, J. Biol. Chem. 1995,
270, 4697 – 4704.
[9] a) J. J. Caramelo, O. A. Castro, L. G. Alonso, G. De Prat-Gay,
A. J. Parodi, Proc. Natl. Acad. Sci. USA 2003, 100, 86 – 91; b) J. J.
Caramelo, O. A. Castro, G. de Prat-Gay, A. J. Parodi, J. Biol.
Chem. 2004, 279, 46280 – 46285; c) S. C. Taylor, A. D. Ferguson,
J. J. Bergeron, D. Y. Thomas, Nat. Struct. Mol. Biol. 2004, 11,
128 – 134.
Figure 5. Preparation of G1M9, G1M8(B), G1M8(C), M9, M8A, M8B,
and M8C from the tetradecasaccharide. Golgi endo-a-mannosidase
(endo-Man’ase), Jack bean b-N-acetylhexosaminidase (b-HexNAc’ase),
Aspergillus oryzae b-galactosidase (b-Gal’ase), Aspergillus saitoi a1,2-
mannosidase (a-Man’ase), and rat liver a-glucosidase-II (G-II) were
used. GN=N-acetylglucosamine, M=mannose, Ga=galactose,
Gc=glucose.
by A. oryzae b-galactosidase digestion, followed by sequential
treatment with Aspergillus saitoi a1,2-mannosidase and Jack
bean b-N-acetylhexosaminidase. Similarly, its regioisomer,
G1M8(C), was obtained by sequential digestion with b-N-
acetylhexosaminidase, a1,2-mannosidase, and b-galactosi-
dase. Glucosidase II treatment of these undecasaccharides
smoothly gave the nonglucosylated congeners M8B and M8C.
In support of their structures, M8A, M8B, and M8C gave the
same molecular mass (MW= 2074.5), but exhibited different
retention times (Figure S2). Further evidence of their struc-
tures was obtained by A. saitoi a1,2-mannosidase digestion,
which gave identical Man5 derivative (MW= 1589.0) from all
three isomers (Figure S3). Similarly, after the a1,2-mannosi-
dase digestion, G1M8(B), G1M8(C), and G1M9 gave the
same G1M7 derivative (MW= 2074.5).[35] Although these
enzymatic digestions were performed on the pmolmLꢀ1 scale,
this method can be easily extended to the mmolmLꢀ1 scale,
which is sufficient for various bioassays.
[10] a) J. J. Caramelo, A. J. Parodi, J. Biol. Chem. 2008, 283, 10221 –
10225; b) L. Ellgaard, E. M. Frickel, Cell Biochem. Biophys.
2003, 39, 223 – 247.
[11] J. C. Christianson, T. A. Shaler, R. E. Tyler, R. R. Kopito, Nat.
Cell Biol. 2008, 10, 272 – 282.
[12] a) I. Matsuo, M. Wada, S. Manabe, Y. Yamaguchi, K. Otake, K.
Kato, Y. Ito, J. Am. Chem. Soc. 2003, 125, 3402 – 3403; b) I.
Matsuo, K. Totani, A. Tatami, Y. Ito, Tetrahedron 2006, 62,
8262 – 8277.
[13] M. A. Arai, I. Matsuo, S. Hagihara, K. Totani, J. Maruyama, K.
Kitamoto, Y. Ito, ChemBioChem 2005, 6, 2281 – 2289.
[14] S. Hagihara, K. Totani, I. Matsuo, Y. Ito, J. Med. Chem. 2005, 48,
3126 – 3129.
[15] a) K. Totani, Y. Ihara, K. Matsuo, H. Koshino, Y. Ito, Angew.
Chem. 2005, 117, 8164 – 8168; Angew. Chem. Int. Ed. 2005, 44,
7950 – 7954; b) K. Totani, Y. Ihara, T. Tsujimoto, I. Matsuo, Y.
Ito, Biochemistry 2009, 48, 2933 – 2940; c) A. Miyagawa, K.
Totani, I. Matsuo, Y. Ito, Biochem. Biophys. Res. Commun. 2010,
403, 322 – 328.
[16] a) K. Totani, Y. Ihara, I. Matsuo, Y. Ito, J. Biol. Chem. 2006, 281,
31502 – 31508; b) K. Totani, Y. Ihara, I. Matsuo, Y. Ito, J. Am.
Chem. Soc. 2008, 130, 2101 – 2107; c) Y. Takeda, K. Totani, I.
Matsuo, Y. Ito, Bioorg. Med. Chem. Lett. 2010, 20, 5357 – 5359.
[17] a) A. Helenius, M. Aebi, Annu. Rev. Biochem. 2004, 73, 1019 –
1049; b) L. W. Ruddock, M. Molinari, J. Cell Sci. 2006, 119,
4373 – 4380.
In summary, we achieved efficient synthesis of the
tetradecasaccharide 1, which was shown to be appropriate
as the common precursor of top-down, chemoenzymatic
synthesis of high-mannose glycans. This method will provide
a powerful tool in studying biological events related to high-
mannose-type glycans.
[18] Kajihara et al. developed an elegant chemoenzymatic approach
toward complex-type glycans, in which an egg-yolk-derived
glycopeptide was employed as a common precursor:Y. Kajihara,
Y. Suzuki, N. Yamamoto, K. Sasaki, T. Sakakibara, L. R. Juneja,
Chem. Eur. J. 2004, 10, 971 – 985.
Received: February 25, 2013
Revised: May 1, 2013
Published online: June 6, 2013
[19] a) F. Barresi, O. Hindsgaul, J. Am. Chem. Soc. 1991, 113, 9376 –
9377; b) Y. Ito, Y. Ohnishi, T. Ogawa, Y. Nakahara, Synlett 1998,
7430 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 7426 –7431

Angewandte
Chemie
1102 – 1104; c) A. Ishiwata, Y. J. Lee, Y. Ito, Org. Biomol. Chem.
2010, 8, 3596 – 3608; d) A. Ishiwata, A. Sakurai, Y. Nishimiya, S.
Tsuda, Y. Ito, J. Am. Chem. Soc. 2011, 133, 19524 – 19535.
[20] I. Matsuo, M. Wada, Y. Ito, Tetrahedron Lett. 2002, 43, 3273 –
3275.
[21] K. Suzuki, H. Maeta, T. Matsumoto, G. Tsuchihashi, Tetrahedron
Lett. 1988, 29, 3571 – 3574.
[22] A. Ishiwata, A. Sakurai, K. Durr, Y. Ito, Bioorg. Med. Chem.
2010, 18, 3687 – 3695.
[23] A. Liptꢀk, I. Jodal, P. Nanasi, Carbohydr. Res. 1975, 44, 1 – 11.
[24] M. Takatani, I. Matsuo, Y. Ito, Carbohydr. Res. 2003, 338, 1073 –
1081.
[25] J. O. Deferrari, E. G. Gros, I. O. Mastronardi, Carbohydr. Res.
1967, 4, 432 – 434.
[26] Y. Ohnishi, H. Ando, T. Kawai, Y. Nakahara, Y. Ito, Carbohydr.
Res. 2000, 328, 263 – 276.
[27] P. J. Garegg, H. Hultberg, S. Wallin, Carbohydr. Res. 1982, 108,
97 – 101.
[29] C. W. Chang, S. S. Chang, C. S. Chao, K. K. T. Mong, Tetrahe-
dron Lett. 2009, 50, 4536 – 4540.
[30] I. Matsuo, T. Miyazaki, M. Isomura, T. Sakakibara, K. Ajisaka, J.
Carbohydr. Chem. 1998, 17, 1249 – 1258.
[31] D. J. Lefeber, J. P. Kamerling, J. F. G. Vliegenthart, Org. Lett.
2000, 2, 701 – 703.
[32] L. M. Likhosherstov, O. S. Novikova, V. A. Derevitskaja, N. K.
Kochetkov, Carbohydr. Res. 1986, 146, C1 – C5.
[33] a) R. E. Pincock, T. E. Kiovsky, J. Am. Chem. Soc. 1966, 88, 51 –
55; b) N. H. Grant, H. E. Alburn, Science 1965, 150, 1589 – 1590;
c) B. N. Naidu, W. Y. Li, M. E. Sorenson, T. P. Connolly, J. A.
Wichtowski, Y. H. Zhang, O. K. Kim, J. D. Matiskella, K. S. Lam,
J. J. Bronson, Y. Ueda, Tetrahedron Lett. 2004, 45, 1059 – 1063;
d) M. Takatani, J. Nakano, M. A. Arai, A. Ishiwata, H. Ohta, Y.
Ito, Tetrahedron Lett. 2004, 45, 3929 – 3932.
[34] W. A. Lubas, R. G. Spiro, J. Biol. Chem. 1988, 263, 3990 – 3998.
[35] HPLC profiles and MALDI TOF-MS data of all glycans are
provided in the Supporting Information (Table S1).
[28] H. Lonn, Carbohydr. Res. 1985, 139, 105 – 113.
Angew. Chem. Int. Ed. 2013, 52, 7426 –7431
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7431