The Glc2Man2-fragment of the N-glycan precursor - A novel ligand for the glycan-binding protein malectin?

Article abstract of DOI:10.1039/c004502k

The Glcα(1→3)Glcα(1→3)Manα(1→2)Man tetrasaccharide (Glc₂Man₂-fragment), a substructure of the natural N-glycan precursor, was synthesized. The interaction of this fragment with the protein malectin, a carbohydrate binding protein localized in the endoplasmatic reticulum, was investigated by ¹H¹⁵N HSQC experiments and isothermal calorimetry. The chemical shift perturbations of nuclei in the protein's backbone caused by the binding of the Glc ₂Man₂-fragment to malectin suggest a binding mode like the known ligand nigerose. The Royal Society of Chemistry 2010.

Full text of DOI:10.1039/c004502k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
The Glc2Man2-fragment of the N-glycan precursor – a novel ligand for the
glycan-binding protein malectin?†‡
Lisa N. Mu¨ller,a Claudia Muhle-Gollb and Moritz B. Biskup*a
Received 19th March 2010, Accepted 13th May 2010
First published as an Advance Article on the web 26th May 2010
DOI: 10.1039/c004502k
The Glca(1→3)Glca(1→3)Mana(1→2)Man tetrasaccharide (Glc2Man2-fragment), a substructure of
the natural N-glycan precursor, was synthesized. The interaction of this fragment with the protein
malectin, a carbohydrate binding protein localized in the endoplasmatic reticulum, was investigated by
1H15N HSQC experiments and isothermal calorimetry. The chemical shift perturbations of nuclei in the
protein’s backbone caused by the binding of the Glc2Man2-fragment to malectin suggest a binding
mode like the known ligand nigerose.
Introduction
The majority of all proteins expressed in living organisms is
subjected to posttranslational modifications; among the most
widely spread of those modifications is the addition of carbohy-
drate moieties to the protein’s side chains to form glycoproteins.1
Over 90% of the glycan structures present in eukaryotic pro-
teins are glycosidically linked to the side chains of asparagines
(→ N-glycans). These N-glycan structures participate in a multi-
tude of functions e.g. as ligands in various recognition processes,
mediators in pathogen interactions, stabilizers against enzymatic
degradation, solubility enhancers, or modifiers of an enzyme’s
charge, structure or orientation. This plethora of adopted roles is
matched by an equally diverse multitude of N-glycan structures.2
While the glycan structures present in an organism are
highly diverse, the first steps of their biogenesis are highly
conserved. N-Glycan structures in eukaryotic organisms are
derived from a common lipid-bound, tetradecasaccharidic pre-
cursor (Glc3Man9GlcNAc2-glycan; A in Fig. 1A), that is co-
translationally transferred onto the nascent polypeptide chain
while still in the endoplasmic reticulum (ER).
Subsequently the glycan moiety is trimmed to a common core
region present in all N-glycans. The early stages of this trimming
process occur in the ER and the resultant N-glycan structures
are closely associated with the correct folding process of the
newly formed protein.3 Only recently malectin, a membrane-
bound ER protein, was identified to exhibit preferential binding
to an early stage form of N-glycan trimming. By carbohydrate
microarray experiments a selective binding of malectin to the
Glc2Man7GlcNAc2-form (Glc2-N-glycan; A without the residues
G1, D2 and D3) was detected. Furthermore, the structure of a
Fig. 1 (A) Lipid-bound Glc3Man9GlcNAc2-N-glycan precursor A (dol =
dolichol); (B) Glc2Man2-fragment, assumed binding region for the car-
bohydrate binding protein malectin; (C) known malectin ligand nigerose
(Glca1→3Glc).
complex of malectin with nigerose (Glca1-3Glc; Fig. 1C), was
determined by NMR experiments. Malectin’s location in the ER,
its remarkable selectivity for the Glc2-N-glycan and its wide
conservation across different species, has led to the assumption
that malectin is closely involved in the quality assurance of
N-glycan biogenesis.4 However, the exact mode of malectin’s
binding to the Glc2-N-glycan has not yet been determined. Despite
the fact that a binding of malectin to mannose could not be
observed, the interaction with the Glc2-N-glycan appeared to
be much stronger than only with nigerose (deduced from the
carbohydrate microarray data). This leads to the assumption that
a cooperative binding effect to one or several of the adjacent
mannose residues of the Glc2-N-glycan takes place.
While the synthesis of fully processed N-glycan structures has
been of interest for the last 30 years, the synthesis of early stage
N-glycan trimming products or substructures of these has not
received such widespread attention. Some syntheses in this field
have been reported, e.g. the group of Ito has recently reported
on their synthesis of the complete Glc3Man9GlcNAc2-glycan, as
well as the trimmed Glc2Man9GlcNAc2-glycan,5 and Fairbanks
et al. have recently accomplished a synthesis of the 1,3-arm’s
terminal Glc3Man moiety.6 However, to our knowledge, there is
no literature precedent for the preparation of small substructures
of the trimmed 1,3-arm.
aKarlsruher Institut fu¨r Technologie – KIT, Institut fu¨r Organische Chemie,
Fritz-Haber-Weg 6, 76131, Karlsruhe, Germany. E-mail: biskup@kit.edu;
Fax: +49 721 608-5637; Tel: +49 721 608-8795
bKarlsruher Institut fu¨r Technologie – KIT, Institut fu¨r Biologische Grenz-
fla¨chen (IBG-2), POB 3640, 76021, Karlsruhe, Germany
† Dedicated to Professor Dr Richard R. Schmidt on the occasion of his
75th birthday.
‡ Electronic supplementary information (ESI) available: 1H15N HSQC
spectra of malectin with/without 1, experimental details for compounds
2c, 4c and 6–11 , and NMR spectra for compounds 1–3, 4c and 6–12a. See
DOI: 10.1039/c004502k
3294 | Org. Biomol. Chem., 2010, 8, 3294–3299
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Table 1 Formation of disaccharide 8a/b from different donors
Donora
Acceptor
Conditions
Yieldb
a/b ratioc
1
4a
4b
4c
4c
5
5
5
5
TMSOTf, Et2O, RT, 25 min
NIS, AgOTf, 4 A MS, CH2Cl2, Et2O, RT, 30 min
~65%
~75%
70–80%
65–75%
2 : 1–3 : 1
1.2 : 1–1.5 : 1
1 : 1
2d
3d
4
˚
˚
NIS, AgOTf, 4 A MS, CH2Cl2, Et2O, RT, 30 min
n
˚
CuBr2, Bu4NBr, 4 A MS, DCE, DMF, RT, 3–5 d
6 : 1–8 : 1
a 4a: trichloroacetimidyl a-glucoside; 4b: phenyl b-selenoglucoside; 4c: methyl b-thioglucoside. b Determined from the isolated mass of epimers 8a/b.
c Estimated from 1H NMR spectra. d Substitution of Et2O by THF did not lead to an increase in diastereoselectivity.
Here, we report on our chemical synthesis of the G2-G3-D1-C
motif of the N-glycan precursor (Glc2Man2-fragment; highlighted
region in Fig. 1A or chemical structure depicted in Fig. 1B). Using
this defined carbohydrate structure in NMR-based chemical shift
perturbation experiments, we hope to gain further information on
the Glc2-N-glycan’s binding epitope with respect to malectin and
a possible participation of the adjacent mannose units.
readily accessible by making use of the participation of an O-
acyl residue located at the C2-position of the D1-building block
during the glycosylation. Thus, our approach to the tetrasac-
charide 1 was based on a block glycosylation strategy relying
on building blocks carrying a general O-benzyl based protecting
group scheme (Scheme 1). This leads to nigerose building block 28
and dimannose 3 as key intermediates in the synthesis. Retrosyn-
thetically these disaccharides can be traced to the correspond-
ing monosaccharide derivatives 4–7, which are accessible from
D-glucose and D-mannose, respectively. The preparation of the
building blocks 4–7 closely followed known procedures and is
detailed in the ESI.‡
Results and discussion
Synthesis of the Glc2Man2-fragment
Inspecting the Glc2Man2-fragment 1 (Scheme 1) it is evident that
the main challenge in preparing 1 resides in the glucose a1→3-
linkages between the G2- and G3-residues and the G3- and D1-
residues of the N-glycan precursor, respectively.
In our initial attempts to efficiently prepare the diglucose donor
2, we investigated a variety of glucose donors (4a–c; cf. Table 1)
in the coupling of 4 and 5 with regard to the yields and the
ratio of a/b-epimers obtained. The trichloroacetimidate donor 4a9
and the seleno glycoside 4b6,10 afforded the desired disaccharide
8 in useful yields of 65–75% upon rapid activation of the
glycoside donor. However, the selectivity with regard to the a/b-
configuration of the newly formed glycosidic bond was moderate
(at best a/b ratios of 3 : 1 were obtained); the use of different
coordinating solvents11 (e.g. diethylether or THF) did not increase
the diastereoselectivity of the reaction (Table 1 entries 1 and 2).
When thioglycoside 4c was used under reaction conditions
as employed for glycosylations with 4a or 4b, that is a rapid
activation of the donor and reliance on the solvent effect to
ensure a stereoselective reaction, disaccharide 8a/b was obtained
in good yields with no diastereoselectivity observed (Table 1
entry 3). Based on literature precedent8 we switched the mode of
activation to a metal bromide mediated variant12 of Lemieux’s in
situ anomerisation methodology,13 thus we were able to obtain
the desired disaccharide 8a/b in yields of 65–75% and with
diastereoselectivities of a/b > 7 : 1 (Table 1 entry 4). To our delight,
the epimers could easily be separated at this stage by conventional
chromatography on silica gel, allowing an efficient access of up to
several tens of grams of 8a.
The conversion of disaccharide 8a to the desired thioglycosyl
donor 2 has previously been described by Takeo et al.;8 however,
that route requires the handling of toxic tin species to introduce the
methyl thioglycoside. Thus we decided to investigate an alternative
approach to accomplish the conversion of the O-acetyl group at
the reducing end into an S-methyl group. Our approach (Scheme 2)
requires the installation of a uniform O-acetyl protecting group
Scheme
1
Retrosynthetic analysis of the N-glycan precursor’s
Glc2Man2-fragment 1, which can be traced to the monosaccharide building
blocks 4–7 (X = OC(NH)CCl3 (4a), SePh (4b), SMe (4c)).
The formation of these 1,2-cis-configured glycosides cannot
be aided by classical neighbouring group assistance.7 However,
the a-1→2-linkage between the mannose residues D1 and C is
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 3294–3299 | 3295
©

View Article Online
mixture of diastereoisomeric tetrasaccharides 12a/b was observed
under these reaction conditions in moderate yields of <50%
(ratio of the diastereoisomers was not determined), we were
not able to isolate the desired tetrasaccharide 12a as a uniform
compound. Instead we observed that the glycosyl acceptor 3 is
consumed slower than in the corresponding glycosylation of the
monosaccharide 5 with the S-methyl thioglycoside 4c to afford
8a/b (Scheme 2). After seven days some amounts of the acceptor
3 still remained, while the donor 2 had partially decomposed.
To overcome this mismatch of reactivities, we reconsidered the
activation of the S-methyl thioglycoside 2 with methyl triflate.17
These reactions were carried out in the presence of diethylether
and 1,2-dimethoxyethane as coordinating solvents11 to effect
preferential formation of the 1,2-cis glycosidic linkage. This more
SN1-like glycosylation reaction proceeded decidedly faster and
afforded a mixture of the diastereoisomers 12a/b (e.g. 61% yield,
a/b ratio 2.5 : 1; Scheme 4). Fortunately a separation of the
obtained mixture of isomers was possible by standard column
chromatography on silica gel, directly affording 12a in amounts
of several hundred milligrams.
Scheme 2 Preparation of the G2-G3-glycosyl donor 2 from D-Glucose
precursors 4c and 5; (a) CuBr2, Bu4NBr, 4 A MS, DCE, DMF, RT, 5 d,
n
˚
68%; (b) i. H2, Pd(OH)2/C, MeOH, RT, 12 h; ii. Ac2O, pyridine, RT, over
night, 88% over 2 steps; (c) i. HBr, HOAc, 0 ◦C → RT, 120 min; ii. thiourea,
acetone, reflux, 60 min; iii. Na2SO3, K2CO3, CHCl3–H2O, RT, 60 min;
iv. EtiPr2N, MeI, CH2Cl2, RT, 60 min, 90% over 4 steps; (d) i. NaOMe
(~0.02 M), MeOH, RT, 4 h; ii. BnBr, NaH, DMF, RT, over night, 71% over
two steps.
pattern as a first step. This was achieved by catalytic cleavage of the
O-benzyl groups by hydrogenolysis in the presence of Pearlman’s
catalyst14 and subsequent treatment with acetic anhydride in
pyridine to afford the fully O-acetylated 9 in 88% yield in two
synthetic steps.
An easily scalable route was chosen to form the S-methyl
thioglycoside 10 in a 4-step sequence15 from 9 in 90% yield (the
purification of intermediates was not required). This approach
compares favourably with Takeo’s tin-based methodology, for
which yields of ~85% were reported. At this stage the final
O-benzyl protecting groups in 2 are installed in 71% yield by
removing the ester groups in fully O-acetylated 10 under Zemple´n
conditions,16 followed by exhaustive O-benzylation.
The synthesis of the glycosyl acceptor, dimannose 3 (Scheme 3),
was achieved by glycosylating the D-mannose acceptor 7 carrying
a free 2-OH group with the trichloroacetimidate donor 6, which
carried a 2-O-acetyl group to ensure the formation of the desired
1,2-trans-configured glycosidic bond.
Scheme 4 Glycosylation towards the fully protected G2-G3-D1-C moi-
˚
ety and liberation of title tetrasaccharide 1; (a) MeOTf, 4 A MS,
Et2O/1,2-dimethoxyethane, 0 ◦C → RT, 3.5 h, 61% (a/b = 2.5 : 1); (b)
i. Na, NH3/THF, -78 ◦C, 120 min; ii. NH4Cl, NH3 -78 ◦C → RT; iii. gel
filtration, >90% from 12a.
Subjecting fully protected 12a to Birch-like conditions18 re-
moved the O-benzyl groups as well as the O-benzylidene group.
Under these reaction conditions the concomitant cleavage of the
3-O-acetyl group on the D1-mannose moiety allowed complete
deprotection in one step to afford the fully deprotected tetrasac-
charide. After purification by desalting and gel filtration of the
title compound 1, the Glc2Man2-fragment, was afforded in over
90% yield as a mixture of the anomers at the reducing end.
Scheme 3 Preparation of the D1-C-glycosyl acceptor 3 from the mannose
building blocks 6 and 7; (a) TMSOTf, CH2Cl2, 0 ◦C → RT, 30 min, 90%;
(b) (HF)x·pyr, pyridine, RT, overnight, 79%.
This glycosylation was effected by the action of trimethylsilyl
triflate as a promoter in dichloromethane to afford the disac-
charide 11 in excellent yield (90%). Subsequent cleavage of the
3-O-tert-butyldimethylsilyl protecting group from the terminal
mannose moiety with hydrogen fluoride·pyridine complex in
pyridine allowed an efficient access to the 3-OH free glycosyl
acceptor 3 in 79% yield.
Interaction studies
The glycosylation of the G2-G3 moiety 2 with the D1-C moiety
3 to afford the fully protected tetrasaccharide 12 was anticipated
to proceed under conditions similar to those reported for the
glycosylation of 4c and 5 (vide supra). While the formation of a
The chemical shifts of NMR-active nuclei are sensitive to
alterations of their electronic and/or conformational environ-
ments as e.g. induced by binding of a ligand molecule. Conse-
quently, we investigated the interaction between malectin and the
3296 | Org. Biomol. Chem., 2010, 8, 3294–3299
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
tetrasaccharide 1 by chemical shift perturbation in 1H15N HSQC
spectra. The pattern of the corresponding chemical shift changes
per residue is comparable to the one obtained for the known ligand
nigerose4 (Fig. 2). A superposition of the HSQC spectra of the free
and Glc2Man2-bound form of malectin is depicted in the ESI.‡
Conclusion
In conclusion, we have established a concise, preparative access to
the Glc2Man2-fragment 1 of the N-glycan precursor. Hallmark
features of our synthesis are the successful construction of
Glca(1→3)-linkages between the rings G2-G3 and G3-D1, and
the global deprotection of fourteen hydroxyl groups in one single
step to afford title compound 1 in high yield and purity.
At the current stage, without further knowledge of the structure
of the N-glycan-malectin-complex, we find no evidence for inter-
actions of the additional two D-mannose residues in 1 with the
protein malectin.
Experimental section
General methods
Solvents were purified by distillation and dried by standard
procedures. Solvents for flash chromatography were purchased
in technical quality and redistilled. Organic solvents for HPLC
(MeCN, MeOH) were purchased from Fischer Scientific. Thin
layer chromatography (TLC) was performed on E. Merck Silica
Gel 60 F254 plates (0.2 mm) and E. Merck RP-18 F254 plates
(0.2 mm). The plates were visualized by immersion in mostain
(200 mL 10% H2SO4, 10 g (NH4)6Mo7O24·4H2O, 200mg Ce(SO4)2),
10% H2SO4 or KMnO4 solution (1% in water, 1% NaHCO3)
Fig. 2 Comparison of the chemical shifts changes Dd induced by the
addition of (A) Glc2Man2-fragment 1 and (B) nigerose; abscissa labelled
with amino acid numbers in the malectin sequence.
It was expected that the additional mannose residues present at
the reducing end of the Glc2Man2-fragment would mainly affect
the region close to the nigerose binding site. According to the
determined malectin-nigerose complex structure (PBD accession
number 2K46)19 the region following Tyr89 or Tyr185 was most
strongly influenced by the binding of nigerose. Consequently,
we explicitly inspected the chemical shift changes in these areas
(Fig. 2A). However, significant chemical shift changes additional
to those already observed for nigerose (Fig. 2B) could not be
detected. These observations lead us to the assumption, that the
Glc2Man2-fragment is bound by malectin and it is bound by
the same region of malectin as nigerose is.
◦
followed by heating (165 C). Preparative flash chromatography
was carried out on Macherey-Nagel Silica Gel 60 (43–60 mm) at a
positive pressure of 0.2–0.4 bar. 1H NMR and 13C NMR spectra
were recorded on Bruker WM 250 Cryospec, AM 400, DRX 500
or Avance 600 instruments, respectively. Proton chemical shifts are
reported in ppm relative to signal from residual solvent protons or
Me4Si as internal standard. Assignments of protons and carbons
were carried out with the aid of 400 or 600 MHz spectra: COSY,
HMQC, nOESY, ROESY, TOCSY. Multiplicities are reported as
they appear in the spectra. They are designated as: s = single,
d = doublet, t = triplet, q = quartet, qu = quintet, dd = doublet
of doublets, m = multiplet, bs = broad single. Diastereotopic
geminal protons are designated with subscripts a and b; non-
distinguishable diastereotopic protons are designated n.d. Carbo-
hydrate residues in disaccharides and larger are differentiated with
superscripts (e.g. ¢, ¢¢, …); monosaccharides or the carbohydrate
residue at the reducing end carry no superscripts. Measurements of
optical rotations were performed on a Perkin-Elmer polarimeter
241 MC (1 dm cell). MALDI-MS were obtained on a Bruker
Biflex MALDI-TOF instrument with 2,5-dihydroxybenzoic acid
(DHB), 2¢,4¢,6¢-trihydroxyacetophenone monohydrate (THAP)
or a-cyano-4-hydroxycinnamic acid (CHCA) as matrix (positive
mode).
However, earlier results4 had shown that at least in the case
of the malectin-nigerose interaction, large chemical shift changes
cannot necessarily be interpreted as a direct interaction with the
bound ligand. For example, Phe117 had a very low chemical shift
change despite measurable nOE contacts to the glucose residue
at the reducing end of the nigerose, whereas Glu114 displayed
the highest observed shift change without any direct contact to
the bound carbohydrate as evidenced by nOE contacts. The lack
of additional chemical shift changes, thus offered no additional
information with regard to malectin’s binding epitope for the
Glc2Man2-fragment 1.
To substantiate the results obtained from NMR experiments the
binding affinity of malectin to tetrasaccharide 1 was determined
by isothermal calorimetry. The determined dissociation constant
(KD = 12.3 mM) is comparable to the value previously determined
for nigerose (KD = 26.3 mM).4 With observed binding affinities
that lie in the same order of magnitude the existence of significant,
additional interactions between the D1- and C-residues of the
Glc2Man2-fragment and malectin seems unlikely. This leads us to
the assumption, that malectin reliably recognizes only the nigerose
motif in the Glc2Man2-fragment and that the stronger interaction
observed with the Glc2-N-glycan, may be mediated by residues on
the 1,6-arms of the Glc2-N-glycan.
NMR experiments
15N-labelled His6-malectin was prepared as previously described.4
NMR spectra were acquired at 22 ◦C in 20 mM potassium
1
phosphate buffer pH 6.8, 150 mM KCl and 2 mM DTT. H15N
HSQC spectra of 250 mM His6-malectin in the absence or presence
of 1 mM Glc2Man2-fragment 1 were acquired on a Bruker Avance
II 600 spectrometer equipped with a broadband triple resonance
probe, with 8 scans per increment and a total of 128 increments
in the indirect dimension. Details on the experiments used for
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 3294–3299 | 3297
©

View Article Online
chemical shift assignment of the protein have been previously
published.4 Data were processed with NMRPIPE20 and analyzed
in pyridine, 1.09 mL, 8.16 mmol) was added and the mixture
was stirred overnight. Saturated aqueous NaHCO3 and Et2O
were added, the layers were separated and the aqueous layer
was extracted with Et2O (2 ¥ 30 mL). The combined organic
layers were dried over MgSO4, filtered, concentrated in vacuo
and purified by column chromatography (silica gel, Hx : EtOAc
33%). The disaccharide 3 (720 mg, 0.87 mmol, 79%) was obtained
as colourless oil (TLC: Rf = 0.49 (silica gel, Hx : EtOAc 50%);
using NMRVIEW.21 The difference in chemical shift Dd was
√
calculated with the formula Dd = ((DdHN)2 + (DdN/10)2).
Isothermal microcalorimetry
ITC measurements were carried out using a VP-ITC Micro-
calorimeter (Microcal, Northhampton, MA, USA) in 20 mM
phosphate buffer pH 6.8, 150 mM KCl and 1 mM TCEP. A
typical titration consisted of sequential injections (10 ml each)
of a Glc2Man2-solution (1.5 mM) into a malectin sample (50 mM),
at time intervals of five minutes, to ensure that the titration peak
returned to the baseline.
20
[a]D = + 58.8 (c = 1.0, CHCl3); dH(600 MHz; CDCl3) 7.47-7.44
(m, 2H, ar), 7.38-7.21 (m, 21H, ar), 7.16-7.13 (m, 2H, ar), 5.54 (s,
3
3
1H, benzylidene), 5.42 (dd, J2¢,3¢ = 3.5 Hz, J1¢,2¢ = 1.3 Hz, 1H,
H-2¢), 5.01 (d, 3J1¢,2¢ = 1.3 Hz, 1H, H-1¢), 4.92 (d, 3J1,2 = 1.5 Hz, 1H,
H-1), 4.82 (d, 2J = 10.7 Hz, 1H, CH2Ph), 4.71 (d, 2J = 11.8 Hz, 1H,
CH2Ph), 4.69 (d, 2J = 11.7 Hz, 1H, CH2Ph), 4.68 (d, 2J = 12.1 Hz,
2
2
1H, CH2Ph), 4.63 (d, J = 11.7 Hz, 1H, CH2Ph), 4.55 (d, J =
a-D-Glucopyranosyl-(1→3)-a-D-glucopyranosyl-(1→3)-a-D-
mannopyranosyl-(1→2)-D-mannopyranose (1)
2
12.1 Hz, 1H, CH2Ph), 4.51 (d, J = 10.7 Hz, 1H, CH2Ph), 4.47
2
3
3
(d, J = 11.8 Hz, 1H, CH2Ph), 4.25 (dt, J3¢,4¢ = 9.5 Hz, J2¢,3¢
ª
3J3¢,OH = 3.1 Hz, 1H, H-3¢), 4.07 (dd, 2J = 10.3 Hz, 3J6a’,5¢ = 4.7 Hz,
Into a two-neck Schlenk flask equipped with a glass stirbar and
a Dewar-type condenser ammonia (~20 mL) was condensed at
-78 ◦C. Sodium sand (50 mg, 2.18 mmol, 130 eq.) was added,
whereupon the solution turned a deep blue colour. It was stirred
for 10 min at -78 ◦C until all sodium was dissolved. Then a solution
of tetrasaccharide (30 mg, 17 mmol, 1 eq.) in anhydrous THF
(2.00 mL) was added dropwise via a syringe and the reaction was
stirred for 2 h at -78 ◦C. Solid NH4Cl (200 mg) was added and the
solution was stirred until the blue colour disappeared. The cooling
bath was removed so that the ammonia could volatilize. The
resulting white solid was dried under high vacuum for 1 h and then
purified by gel filtration on a Bio-Rad P-2 gel filtration column
(~15 ¥ 1.5 cm, H2O : MeOH 20%). The remaining salts were
removed by precipitating them with MeOH. After lyophilization
the unprotected tetrasaccharide 1 (11 mg, 16.5 mmol, 98%) was
obtained as a mixture of the anomers at the reducing end;
in aqueous solution the a-configured compound predominates
1H, H-6a’), 3.98 (t, 3J1,2 ª J2,3 ª 2 Hz, 1H, H-2), 3.95 (dd, 3J2,3
=
3
2.8 Hz, 3J3,4 = 9.2 Hz, 1H, H-3), 3.93 (t, 3J3,4 ª J4,5 = 9.2 Hz Hz,
3
1H, H-4), 3.90 (dd,3J5¢,4¢ = 9.5 Hz, 3J5¢,6a,b’ = 4.6 Hz, 1H, H-5¢), 3.85
(t, 3J3¢,4¢ ª J4¢,5¢ = 9.5 Hz, 1H, H-4¢), 3.81 (ddd, 3J4,5 = 9.3 Hz, 3J5,6a
=
=
3
4.6 Hz, 3J5,6b = 1,3 Hz, 1H, H-5), 3.78 (dd, 2J = 10.7 Hz, 3J5,6a
4.7 Hz, 1H, H-6a), 3.72 (d, 2J = 10.3 Hz, 1H, H-6b’), 3.69 (dd, 2J =
10.7 Hz, 3J5,6b = 1.3 Hz, 1H, H-6b), 2.23 (bs, 1H, OH), 2.12 (s, 3H,
CH3CO) ppm; dC(151 MHz; CDCl3) 170.2, 138.4, 138.2, 137.1,
137.0, 129.2, 128.4, 128.4, 128.3, 128.3, 128.3, 128.0, 127.9, 127.9,
127.7, 127.6, 127.5, 127.5, 127.5, 126.3, 102.2 (benzylidene), 100.0
(C1¢), 97.9 (C1), 79.7, 78.8, 75.2, 75.1, 74.6, 73.3, 72.3, 72.1, 71.8,
69.0, 68.9, 68.5, 67.2, 63.7, 21.0; non-decoupled H13C HMQC
1
(151 MHz, CDCl3): JC,Hbenzylidene = 161.5 Hz, JC1,H-1 = 170.1 Hz,
JC1¢,H-1¢ = 173.7 Hz; MALDI-MS (pos. mode, DHB): [M+Na]+
calcd.: 855.3, found: 855.1; [M+K]+ calcd.: 871.3, found: 871.0;
C49H52O12 (832.9)).
20
(TLC: Rf = 0.03 (silica gel, EtOAc : MeOH 20%); [a]D = + 60.9
3
(c = 1.2, H2O); dH(600 MHz; D2O) 5.43 (d, J1,2 = 1.9 Hz, 1H,
Benzyl (2,3,4,6-tetra-O-benzyl-a-D-glucopyranosyl)-(1→3)-(2,4,6-
tri-O-benzyl-a-D-glucopyranosyl)-(1→3)-(2-O-acetyl-4,6-O-
benzylidene-a-D-mannopyranosyl)-(1→2)-3,4,6-tri-O-benzyl-a-D-
mannopyranoside (12a)
H-1), 5.41 (d, 3J1,2 = 3.9 Hz, 1H, H-1¢’’), 5.33 (d, 3J1,2 = 3.9 Hz, 1H,
H-1¢’), 5.09 (d, 3J1,2 = 2.0 Hz, 1H, H-1¢), 4.30 (dd, 3J2,3 = 3.4 Hz,
3J1,2 = 1.9 Hz, 1H, H-2¢), 4.06 (ddd, 3J4,5 = 10.0 Hz,3J5,6a = 4.2 Hz,
3J5,6b = 2.0 Hz, 1H, H-5¢’’), 4.01 (1H, H-2), 4.01 (1H, H-3¢), 4.00
(n.d., 2H, H-6a,b), 3.96 (1H, H-3¢’), 3.95-3.83 (n.d., 2H, H-6¢a,b),
3.91 (1H, H-5¢’), 3.89 (1H, H-3), 3.88 (1H, H-4¢), 3.88-3.77 (n.d.,
2H, H-6¢’a,b), 3.86 (1H, H-4), 3.84 (n.d., 2H, H-6¢’’a,b), 3.82 (1H,
H-5¢), 3.80 (1H, H-3¢’’), 3.71 (1H, H-2¢’), 3.75 (1H, H-5), 3.68 (1H,
H-4¢’), 3.62 (dd, 3J2,3 = 10.0 Hz, 3J1,2 = 4.0 Hz, 1H, H-2¢’’), 3.50 (t,
Donor 2 (668 mg, 0.67 mmol, 1.15 eq.) and acceptor 3 (482 mg,
0.58 mmol, 1 eq.) were dried together under high vacuum. After
dissolving under argon atmosphere in anhydrous diethyl ether
˚
(20 mL) and anhydrous 1,2-dimethoxyethane (1.33 mL) 4 A
molecular sieve (3 g) was added and the mixture was stirred
◦
for 30 min. The solution was cooled to 0 C and methyl triflate
3
3J3,4 ª J4,5 = 9.7 Hz, 1H, H-4¢’’); dC(101 MHz; D2O) 102.5 (C1¢),
(459 mL, 4.05 mmol, 7 eq.) was added. The ice bath was removed
and the mixture was stirred for 3.5 h at room temperature. Et3N
(1 mL) was added and the solution was diluted with diethyl ether.
After stirring for a few minutes the mixture was filtered over a pad
of Celite. The solids were washed successively with ether and the
combined filtrates were concentrated in vacuo. The crude product
was purified by column chromatography on (silica gel, Hx : EtOAc
16%). Three fractions were collected: pure b-epimer (80 mg, 47
mmol), followed by a mixture of the epimers (120 mg, 67 mmol,
a/b ~ 1 : 7) and finally the pure a-epimer (430 mg, 240 mmol) as
colourless oils. In total the glycosylation reaction afforded 12a/b
(630 mg, 352 mmol, 61% a/b ~ 2.5 : 1). (TLC: 12a: Rf = 0.53;
101.0 (C1¢’), 99.7 (C1¢’’), 93.0 (C1), 80.2, 79.6, 78.8, 73.7, 73.4,
73.0, 72.6, 72.2, 72.2, 70.9, 70.7, 70.5, 70.3, 69.9, 67.6, 66.6, 61.5,
1
61.4, 61.0, 60.9 ppm; non-decoupled H13C HMQC (151 MHz;
CDCl3): JC1,H-1 = 172 Hz, JC1¢,H-1¢ = 172 Hz, JC1¢’,H-1¢’ = 173 Hz,
JC1¢’’,H-1¢’’ = 172 Hz; MALDI-MS (pos. mode, THAP): [M+Na]+
calcd.: 689.2, found: 689.3; [M+K]+ calcd.: 705.2, found: 705.2;
C24H42O21 (666.6)).
Benzyl (2-O-acetyl-4,6-O-benzylidene-a-D-mannopyranosyl)-
(1→2)-3,4,6-tri-O-benzyl-a-D-mannopyranoside (3)
In a Teflon flask disaccharide 11 (1.03 g, 1.09 mmol, 1 eq.)
was dissolved in dry pyridine (12 mL). (HF)x·pyridine (65–70%
20
12b: Rf = 0.56 (silica gel, Hx : EtOAc 33%); 12a: ([a]D = + 49.1
3298 | Org. Biomol. Chem., 2010, 8, 3294–3299
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
(c = 1.1, CHCl3); dH(600 MHz; CDCl3) 7.46-7.02 (m, 60H, ar),
Chemie – KIT, Karlsruhe) for access to the equipment required
for handling liquid ammonia.
3
3
5.58 (d, J1¢’’,2¢’’ = 3.3 Hz, 1H, H-1¢’’), 5.55 (d, J2¢,3¢ = 3.4 Hz,
3
3J1¢,2¢ = 1.1 Hz, 1H, H-2¢), 5.51 (dd, J1¢’,2¢’ = 3.4 Hz, 1H, H-1¢’),
5.30 (s, 1H, benzylidene), 5.13 (d, 3J1¢,2¢ = 1.1 Hz, 1H, H-1¢), 5.02
Notes and references
3
(d, J1,2 = 1.7 Hz, 1H, H-1), 4.96-4.90 (m, 3H, CH2Ph), 4.85 (d,
2J = 10.4 Hz, 1H, CH2Ph), 4.84 (d,2J = 10.6 Hz, 1H CH2Ph),
1 R. Dwek, Chem. Rev., 1996, 96, 683–720; A. Varki, Essentials of
Glycobiology, 2nd edn, Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, N.Y., 2009.
2 A. Helenius and M. Aebi, Science, 2001, 291, 2364–2369; A. Helenius
and M. Aebi, Annu. Rev. Biochem., 2004, 73, 1019–1049.
3 A. Petrescu, T. Butters, G. Reinkensmeier, S. Petrescu, F. Platt, R.
Dwek and M. Wormald, EMBO J., 1997, 16, 4302–4310; M. Aebi, R.
Bernasconi, S. Clerc and M. Molinari, Trends Biochem. Sci., 2010, 35,
74–82; P. Deprez, M. Gautschi and A. Helenius, Mol. Cell, 2005, 19,
183–195.
4 T. Schallus, C. Jaeckh, K. Feher, A. S. Palma, Y. Liu, J. C. Simpson, M.
Mackeen, G. Stier, T. J. Gibson, T. Feizi, T. Pieler and C. Muhle-Goll,
Mol. Biol. Cell, 2008, 19, 3404–3414.
5 I. Matsuo, T. Kashiwagi, K. Totani and Y. Ito, Tetrahedron Lett., 2005,
46, 4197–4200.
6 S. Ennis, I. Cumpstey, A. Fairbanks, T. Butters, M. Mackeen and M.
Wormald, Tetrahedron, 2002, 58, 9403–9411.
7 P. Garegg, Adv. Carbohydr. Chem. Biochem., 2004, 59, 69–134; X. Zhu
and R. R. Schmidt, Angew. Chem., 2009, 121, 1932–1967, (Angew.
Chem., Int. Ed., 2009, 48, 1900–1934); L. K. Mydock and A. V.
Demchenko, Org. Biomol. Chem., 2010, 8, 497.
2
2
4.79 (d, J = 12.0 Hz, 1H, CH2Ph), 4.79 (d, J = 11.2 Hz, 1H,
2
CH2Ph), 4.74-4.70 (m, 3H, CH2Ph), 4.67 (d, J = 12.5 Hz, 1H,
CH2Ph), 4.62 (m, 2H, CH2Ph), 4.55 (d, 2J = 11.6 Hz, 1H, CH2Ph),
2
2
4.55 (d, J = 10.6 Hz, 1H, CH2Ph), 4.54 (d, J = 12.0 Hz, 1H,
2
CH2Ph), 4.53 (d, J = 9.3 Hz, 1H, CH2Ph), 4.51-4.45 (m, 2H,
2
CH2Ph, H-3¢), 4.45-4.35 (m, 3H, CH2Ph, H-5¢’’), 4.30 (d, J =
2
12.2 Hz, 1H, CH2Ph), 4.25 (d, J = 11.6 Hz, 1H, CH2Ph), 4.24
2
3
3
(d, J = 11.2 Hz, 1H, CH2Ph), 4.19 (t, J2¢’,3¢’ ª J3¢’,4¢’ = 9.3 Hz,
1H, H-3¢’), 4.15 (t, 3J3¢,4¢ ª J4¢,5¢ = 9.9 Hz, 1H, H-4¢), 4.12 (m, 2H,
3
H-6a,b), 4.10 (d, 3J1,2 = 1.9 Hz, 1H, H-2), 4.03 (m, 1H, H-3), 4.02
(m, 1H, H-3¢’’), 4.00 (m, 1H, H-5¢), 3.93-3.89 (m, 2H, H-4, H-5¢’),
3
3
3.87 (dd, J4¢’,5¢’ = 10.1 Hz, J3¢’,4¢’ = 9.3 Hz, 1H, H-4¢’), 3.78 (m,
1H, H-5), 3.76 (m, 2H, H-6a,b’), 3.66 (t, 3J3¢’’,4¢’’ ª J4¢’’,5¢’’ = 9.4 Hz,
3
1H, H-4¢’’), 3.60 (dd, 2J = 10.9 Hz, 3J5¢’,6a’’ = 2.1 Hz, 1H, H-6a’’),
3.56 (m, 1H, H-6b’’), 3.53 (dd, 3J2¢’’,3¢’’ = 9.8 Hz, 3J1¢’’,2¢’’ = 3.7 Hz,
1H, H-2¢’’), 3.47 (dd,3J2¢’,3¢’ = 9.6 Hz, 3J1¢’,2¢’ = 3.5 Hz, 1H, H-2¢’),
8 K. Takeo, S. Kitamura and Y. Murata, Carbohydr. Res., 1992, 224,
111–122.
9 R. R. Schmidt and J. Michel, Angew. Chem., 1980, 92, 763–764, (Angew.
Chem., Int. Ed. Engl., 1980, 19, 731–732).
2
3
3.25 (dd, J = 11.0 Hz, J5¢’’,6a’’’ = 1.8 Hz, 1H, H-6a’’’), 3.23 (dd,
2J = 10.9 Hz, 3J5¢’’,6a’’’ = 2.5 Hz, 1H, H-6b’’’), 2.21 (s, 3H, CH3CO);
dC(151 MHz; CDCl3) 169.6, 138.8 (2C), 138,7, 138.2, 138.1, 138.1,
138.0, 137.9, 137.9, 137.8, 136.9 (2C), 129.1, 128.4, 128.3, 128.3,
128.3, 128.2, 128.1, 128.0, 128.0, 127.9, 127.9, 127.8, 127.7, 127.6,
127.6, 127.5, 127.5, 127.5, 127.4, 127.4, 127.3, 127.2, 127.2, 126.8,
126.3, 126.2, 102.2 (Cbenzylidene), 100.1 (C1¢), 97.8 (C1), 97.4 (C1¢’’),
96.3 (C1¢’), 82.2, 79.7, 79.4, 79.2, 78.0, 77.9, 76.7, 75.4, 75.2,
75.1, 74.8, 74.7, 74.4, 73.3, 73.3, 73.2, 72.6, 72.1, 71.9, 70.2, 70.2,
70.0, 69.6, 69.0, 68.9, 68.5, 68.0, 63.7, 20.8 ppm; non-decoupled
10 S. Mehta and B. M. Pinto, J. Org. Chem., 1993, 58, 3269–3276.
11 R. Eby and C. Schuerch, Carbohydr. Res., 1974, 34, 79–90; G. Wulff
and G. Rohle, Angew. Chem., 1974, 86, 173–187, (Angew. Chem., Int.
Ed. Engl., 1974, 13, 157–170).
12 S. Koto, N. Morishima, C. Kusuhara, S. Sekido, T. Yoshida and S. Zen,
Bull. Chem. Soc. Jpn., 1982, 55, 2995–2999; S. Sato, M. Mori, Y. Ito
and T. Ogawa, Carbohydr. Res., 1986, 155, C6–C10.
13 R. U. Lemieux, K. James and T. L. Nagabhus, Can. J. Chem., 1973,
51, 42–47; R. U. Lemieux, K. B. Hendriks, R. V. Stick and K. James,
J. Am. Chem. Soc., 1975, 97, 4056–4062.
14 W. M. Pearlman, Tetrahedron Lett., 1967, 8, 1663–1664.
15 K. Leontein, M. Nilsson and T. Norberg, Carbohydr. Res., 1985, 144,
231–240.
1H,13C HMQC (151 MHz, CDCl3): JC,Hbenzylidene = 163 Hz, JC1,H-1
=
172 Hz, JC1¢,H-1¢ = 175 Hz, JC1¢’,H-1¢’ = 173 Hz, JC1¢’’-H-1¢’’
=
175 Hz; MALDI-MS (pos. mode, DHB): [M+Na]+ calcd.: 1809.8,
found: 1809.8; [M+K]+ calcd.: 1825.7, found: 1825.8; C110H114O22
(1788.1)).
16 G. Zemple´n and E. Pacsu, Ber. Dtsch. Chem. Ges. B, 1929, 62, 1613–
1614.
17 H. Lo¨nn, Carbohydr. Res., 1985, 139, 115–121; H. Lo¨nn, Carbohydr.
Res., 1985, 139, 105–113.
18 Z. Wang, X. Zhang, M. Visser, D. Live, A. Zatorski, U. Iserloh, K. Lloyd
and S. Danishefsky, Angew. Chem., 2001, 113, 1778–1782, (Angew.
Chem., Int. Ed., 2001, 40, 1728–1732); U. Iserloh, V. Dudkin, Z. Wang
and S. Danishefsky, Tetrahedron Lett., 2002, 43, 7027–7030.
19 Structural data deposited: http://dx.doi.org/10.2210/pdb2k46/pdb.
20 F. Delaglio, S. Grzesiek, G. W. Vuister, G. Zhu, J. Pfeifer and A. Bax,
J. Biomol. NMR, 1995, 6, 277–293.
Acknowledgements
This work was supported by the KIT “Concept for the Future”
(RG49-1); funding was provided by the federal “Excellence Initia-
tive”. The authors thank Dr Vladimir Rybin (EMBL, Heidelberg)
for the determination of the ITC data; and Dipl.-Chem. Dominik
Nied and Professor Frank Breher (Institut fu¨r Anorganische
21 B. A. Johnson and R. A. Blevins, J. Biomol. NMR, 1994, 4, 603–614;
B. A. Johnson, Methods Mol. Biol., 2004, 278, 313–352.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 3294–3299 | 3299
©