Peptide templated glycosylation reactions

Article abstract of DOI:10.1016/S0957-4166(99)00494-2

Glycosyl donors and acceptors may be covalently linked to aspartic acid residues via OH-6 esters. Peptide elaboration allows glycosylation reactions to be performed between donors and acceptors linked to this peptide template. These reactions display increased regio- and stereoselectivities, which are dependent on the nature of the peptide. Simple molecular modelling is used to rationalise the differing product distributions obtained by variation of the linking amino acid sequence. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(99)00494-2

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 11 (2000) 231–243
Peptide templated glycosylation reactions
Richard J. Tennant-Eyles,a Benjamin G. Davis b and Antony J. Fairbanks a,∗
aDyson Perrins Laboratory, Oxford University, South Parks Road, Oxford OX1 3QY, UK
bDepartment of Chemistry, University of Durham, Science Laboratories, South Road, Durham DH1 3LE, UK
Received 28 October 1999; accepted 5 November 1999
Abstract
Glycosyl donors and acceptors may be covalently linked to aspartic acid residues via OH-6 esters. Peptide
elaboration allows glycosylation reactions to be performed between donors and acceptors linked to this peptide
template. These reactions display increased regio- and stereoselectivities, which are dependent on the nature of the
peptide. Simple molecular modelling is used to rationalise the differing product distributions obtained by variation
of the linking amino acid sequence. © 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
The ultimate goal for chemists interested in oligosaccharide synthesis is the development of a reliable,
automatable, and generally applicable solid phase process, a feat which completely revolutionalised the
synthesis of oligonucleotides and oligopeptides some 30 years ago. The importance of oligosaccharides
in a plethora of biological processes is now extremely well documented1 and it is universally acknow-
ledged that the further unravelling of the biological roles that many oligosaccharide structures play
would be greatly facilitated by more efficient synthetic approaches to these materials. However, the extra
complications inherent in the construction of glycosidic bonds, namely those of control of regio- and
stereochemistry, have been continually hampering this development. Although regiochemical control
can be achieved via selective protecting group manipulations, such sequences are often protracted and
time consuming. A much more difficult task is the ability to completely control the stereochemistry of
the newly created interglycosidic linkage. Despite extremely significant advances in this area, there is
still no completely general method for the stereospecific formation of glycosidic linkages. Considering
the vast amount of time and effort that has gone into this avenue of research,2 perhaps a single solution
to such a complex problem simply does not exist.
As a potential solution to the problem of generality we recently disclosed3 some preliminary in-
vestigations into a new approach to the construction of glycosidic linkages. Therein, the strategy is to
∗
Corresponding author. E-mail: antony.fairbanks@chem.ox.ac.uk
0957-4166/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(99)00494-2
tetasy 3129

232
R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
abandon the classical philosophy of the search for a ‘general glycosylation reaction’, and to attempt
to use combinatorial techniques to search for separate synthetic solutions to the construction of each
individual glycosidic bond.
The idea was inspired by the technique of molecular tethering, so elegantly developed for the
synthesis of β-mannosides by Stork,4 Hindsgaul5 and Ogawa.6 Subsequent to these investigations several
other groups have disclosed the use of more extended tethers for the stereo- and even regioselective
construction of glycosidic linkages.7 As a solution to the problems of glycosidic bond formation an
approach was envisaged which combines the use of tethering and combinatorial chemistry. In theory
this will facilitate rapid searches through libraries of tethers, screening their potential to promote the
formation of particular glycosidic linkages in stereo- and regioselective fashions.
Since tethers were required in which structural diversity can be readily introduced by solid phase
synthesis, initial investigations have focused on peptides. One particularly attractive feature of this
approach is that it may allow the indirect use of existing automated methodology for solid phase peptide
synthesis for oligosaccharide construction. However, before the development of such a technology can
be contemplated, two questions must be answered. Firstly, can glycosylation reactions be performed
between donors and acceptors that are linked to a peptide template? Secondly, can variation of the peptide
amino acid sequence influence the stereo- and regiochemical outcome of this glycosylation reaction?
Both of these two preliminary issues are addressed herein.
2. Results and discussion
2.1. Synthesis of glycosyl amino acid building blocks
For the initial studies mannose was chosen as both glycosyl donor and acceptor, since it was thought
that any particular influence of the peptide template on the glycosylation reaction would be well
demonstrated if large amounts of β-mannoside formation were observed. Initial investigations focused
on a thioethylglycoside as the glycosyl donor, and it was decided to link both acceptor and donor to
the peptide template via the OH-6 hydroxyl. Firstly, the differentially protected glycosyl acceptor 4 was
synthesised in three steps from methyl mannopyranoside 1. Tritylation of the primary hydroxyl yielded
the triol 2, which was then completely protected by treatment with excess para-methoxybenzyl (PMB)
chloride and sodium hydride in DMF to yield 3. Acid catalysed removal of the trityl group then yielded
the glycosyl acceptor 4 in an overall yield of 41% (Scheme 1). Subsequently, the thioglycoside donor 8
was synthesised in four steps, again from methyl mannopyranoside 1 (Scheme 1). Thus, reaction of 1
with benzyl bromide and sodium hydride produced the tetrabenzyl methylpyranoside 5. Acetolysis8 of
5 then yielded the diacetate 6. Reaction with ethanethiol and BF3·etherate produced the thioglycoside
7, which was simply deacetylated with sodium methoxide to yield the glycosyl donor 8 (44% yield over
four steps).
Both donor 8 and acceptor 4 were then coupled to the aspartic acid derivative 9 to yield the glycosyl
amino acids 10 and 11, in acceptable 72% and 75% yields, respectively, following existing protocol.9
Subsequent protecting group manipulations, namely cleavage of the tert-butyl ester of 10 to yield the acid
12, and removal of the PMB protecting groups from the glycosyl acceptor 12 yielding the deprotected
donor 13, were again performed as previously described9 (Scheme 2).

R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
233
Scheme 1. Reagents: (i) Ph3CCl, DMAP, pyridine, 80°C, 75%; (ii) PMBBr, NaH, DMF, 76%; (iii) EtOH, AcOH, 80°C, 71%;
(iv) BnBr, NaH, DMF, 83%; (v) Ac2O, AcOH, H2SO4, 0°C, 90%; (vi) EtSH, BF3·Et2O, CH2Cl2, 0°C, 65%; (vii) Na, MeOH,
0°C, 90%
2.2. Peptide elaboration
The aspartate linked glycosyl acceptor 13 was coupled to a series of amino acids. Peptide formation
was achieved via a two-step coupling of 13, involving initial piperidine-induced Fmoc removal and
subsequent EEDQ mediated coupling with the required Fmoc-protected amino acid (Scheme 3). Once
the desired intermediate amino acids had been added, the peptide sequence was completed by Fmoc
deprotection and EEDQ mediated coupling to the aspartate-linked donor 12. In this way the series of
peptides 14a–g was constructed for use in subsequent glycosylation reactions.
Scheme 2. Reagents: (i) 9, DCC, DMAP, CH2Cl2, 75%; (ii) CF3CO2H, CH2Cl2, quantitative; (iii) 9, DCC, DMAP, CH2Cl2,
72%; (iv) ceric ammonium nitrate, CH3CN:H2O, 9:1, 80%
2.3. Glycosylation reactions
As a control reaction, the protected thioglycosyl donor 7 was glycosylated with the known triol
15,10 under standard activation conditions (N-iodosuccinimide and triflic acid in dichloromethane). This
proceeded in moderate yield but with extremely poor selectivity (Scheme 4); of the seven products only
two were found to be disaccharides, which were identified as the α(1–3)- and β(1–3)-linked materials 16
and 17, respectively. Four of the other products were found to be undesired trisaccharides, and the final
material was simply the product of hydrolysis.

234
R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
Scheme 3.
Intramolecular glycosylation of the peptides 14a–f was then undertaken, again mediated by N-
iodosuccinimide and triflic acid in dichloromethane. The resulting disaccharides were separated by
chromatography and cleaved from the peptide template by treatment with potassium carbonate in
methanol, before complete acetylation and characterisation. Structural assignment was achieved by
detailed NMR analysis.11 Table 1 details the outcome of the glycosylation reactions of peptides 14a–f.
Scheme 4.
The first positive result was that glycosylation did, indeed, occur between the templated donor and
acceptor and that the template was compatible with the NIS/triflic acid activation conditions. In fact,
in all cases the overall yields for these reactions were similar to that observed for the control reaction.
Secondly, the absence of any trisaccharide products, in contrast to the control, was a clear indicator
that glycosylation was effected intramolecularly. Thirdly, it was encouraging to find that moderate levels
of regio- and stereoselectivity were achieved, and that the product distribution varied with the peptide
sequence. Considering the results taken as a whole it is clear that the α(1–3) 16 and β(1–3) 17 products
predominate throughout, although the relative amount of these two isomers is variable. In the case of no
intermediate amino acid (AspAsp, 14a) some of the β(1–2) product 18 (13%) was observed along with
the α(1–3) (16, 11%) and β(1–3) (17, 23%) products. Insertion of an intermediate glycine (AspGlyAsp,
14b) lead to the exclusion of the β(1–2) product 18, and the formation of the α(1–3) (16, 21%) and β(1–3)
(17, 20%) materials. Subsequent variation of the intermediate amino acid in the tripeptides from glycine
to alanine (Ala) 14c, to phenylalanine (Phe) 14d actually influenced the outcome of the glycosylation
reaction only slightly, the same α(1–3) and β(1–3) products being the major products in all cases. The

R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
235
Table 1
Results of peptide templated glycosylation reactions
case of asparagine (Asn) 14e was slightly more complicated. Here the β(1–3) product 17 was isolated
in 20% yield, whilst the remaining 23% was an inseparable mixture of materials. Most notably the use
of proline (Pro) as the linking residue 14f resulted in a much higher degree of β selectivity; the β(1–2)
18-linked disaccharide was now a major product, together with the β(1–3) 17 disaccharide, and now
none of the previously predominant α(1–3)-linked product 16 was observed. Increasing the length of the
peptide by the addition of another Gly residue 14g did not greatly alter the outcome of the glycosylation
reaction; again the α(1–3) 16 and β(1–3) 17 products were the sole products isolated as a mixture in a
combined yield of 56%.
3. Molecular modelling
Some simple molecular modelling was undertaken to investigate the relationship between relative
proportions of each disaccharide formed and the amino acid sequence of the peptide template. Potential
energy minima locations for a selection of glycopeptides were probed by variation of the starting point,
this being achieved by variation of the torsional angles in the linker. This process allowed a number
of representative points at different relative orientations of the glycosyl acceptor ranging from the β-
face to the α-face of the glycosyl donor to be probed. The glycosyl donor was itself approximated by a
tetrahydropyran in which the thioethyl leaving group was replaced by a hydrogen atom. This was done
in order to avoid any strong conformational influences of the anomeric leaving group and, therefore,
provide a more applicable model for the glycosyl cation iself.

236
R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
It was found that all of the modelled peptides AspGlyAsp 14b, AspAlaAsp 14c and AspPheAsp 14d
have energy minima in positions above both the α- and β-faces, i.e. there is little facial distinction,
although the α-face minima are typically slightly lower in energy. For example, Fig. 1(a) shows
(hydrogen atoms omitted for clarity) the AspGlyAsp peptide system 14b in which the glycosyl acceptor
occupies a minimum above the α-face of the glycosyl acceptor. In contrast, the half chair conformation
of the proline residue shown in Fig. 1(b) biases the AspProAsp model towards a β-face energy minimum.
This, therefore, suggests that the increased β selectivity observed for the AspProAsp peptide 14f is
a consequence of the conformation of this tether. In addition, changing identity of the intermediate
amino acid from Gly to Ala or Phe changed the conformations of these minima only very slightly. This
is consistent with the similar product distributions that were observed for these three peptide linkers.
Finally, these simple models also provided a plausible explanation for the preponderance of (1–3)-linked
disaccharide products; in all of the minimised models the 3-hydroxyl of the acceptor is readily disposed
to react with the anomeric centre of the donor.
Fig. 1. Minimised structures of tripeptide-tethered systems illustrating the positions of O-2 and O-3 of the glycosyl acceptor
relative to the anomeric centre of the glycosyl donor (hydrogens omitted for clarity)
4. Conclusions
In summary, we have successfully demonstrated for the first time that glycosylation reactions may
be achieved between glycosyl donors and acceptors that are bound to a peptide template. In addition,
these reactions show enhanced, although at present modest, regio- and stereoselectivities, the product
distribution of which can be rationalised by simple molecular modelling. Since the current study has
so far only examined an extremely limited number of peptide templates these moderate selectivities
were not wholly unexpected. It is now necessary to screen rapidly large numbers of combinatorially
assembled tethers in order to search for particular peptide sequences that promote completely regio- and
stereoselective reactions, via a solid-phase approach. However, although the conformational preferences
of this type of very short oligopeptide may not be great enough to effect really high levels of regio-
and stereocontrol. For this reason it may well be that either peptides built from β amino acids, or other
foldamers,12 will be better choices as templates. Further investigations in these areas are currently in
progress and the results will be published in due course.

R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
237
5. Experimental
5.1. General experimental
Melting points were recorded on a Kofler hot block. Proton nuclear magnetic resonance (δH) spectra
were recorded on Varian Gemini 200 (200 MHz), Bruker AC 200 (200 MHz), Bruker DPX 400 (400
MHz) or Bruker AMX 500 (500 MHz) spectrometers. Carbon nuclear magnetic resonance (δC) spectra
were recorded on a Varian Gemini 200 (50.3 MHz). Multiplicities were assigned using DEPT sequence.
All chemical shifts are quoted on the δ-scale. Infrared spectra were recorded on a Perkin–Elmer
150 Fourier Transform spectrophotometer. Mass spectra were recorded on VG Micromass 30F, ZAB
1F, Masslab 20-250, Micromass Platform 1 APCI, or Trio-1 GCMS (DB-5 column) spectrometers,
using desorption chemical ionisation (NH3 DCI), electron impact (EI), chemical ionisation (NH3 CI),
atmospheric pressure chemical ionisation (APCI), and fast atom bombardment (FAB) techniques as
stated. Optical rotations were measured on a Perkin–Elmer 241 polarimeter with a path length of 1 dm.
Concentrations are given in g/100 ml. Hydrogenations were run under an atmosphere of hydrogen gas
maintained by an inflated balloon. Microanalyses were performed by the microanalytical services of the
Inorganic Chemistry Laboratory, Oxford. Thin layer chromatography (t.l.c.) was carried out on Merck
glass-backed sheets, pre-coated with 60F254 silica. Plates were developed using 0.2% w/v cerium(IV)
sulfate and 5% ammonium molybdate in 2 M sulphuric acid. Flash chromatography was carried out using
Sorbsil C60 40/60 silica. Solvents and available reagents were dried and purified before use according to
standard procedures; methanol was distilled from magnesium methoxide, dichloromethane was distilled
from calcium hydride, pyridine was distilled from calcium hydride and stored over potassium hydroxide,
and tetrahydrofuran was distilled from a solution of sodium benzophenone ketyl immediately before use.
Hexane was distilled between 40 and 60°C before use to remove involatile fractions.
5.2. Molecular modelling
Structural set up was performed with Insight II, version 95.0 (Biosym/MSI Technologies, Inc. San
Diego, CA, USA). Structures were generated using the Builder module of Insight; the glycosyl cation
intermediate was approximated by a tetrahydropyran in which the thioalkyl leaving group was replaced
by a hydrogen atom. Initial coordinates for the minimisation were adjusted through torsional angle
increments of approx. 30° to provide a representative number of differently oriented positions of donor
relative to acceptor such that the pyran rings of donor and acceptor were within approx. 10 Å. Energy
simulations were performed with the DISCOVER program, Version 2.9.7 on a Silicon Graphics Indigo
2 computer, using the consistent valence force field (CVFF) function. A non-bonded cut-off distance
of 18 Å with a switching distance of 2 Å was employed. The non-bonded pair list was updated every
20 cycles and a dielectric constant of 1 was used in all calculations. Each stage of energy minimisation
was conducted by means of the method of steepest descents without Morse or cross terms, until the
derivative of energy with respect to structural perturbation was less than 1.0 kcal/Å; then the method of
conjugate gradients, without Morse or cross terms until the derivative of energy with respect to structural
perturbation was less than 0.1 kcal/Å; and, finally, the method of conjugate gradients, with Morse and
cross terms until the final derivative of energy with respect to structural perturbation was less than 0.1
kcal/Å.

238
R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
5.3. Methyl 6-O-triphenylmethyl-α-D-mannopyranoside 2
Triphenylmethylchloride (21.56 g, 77.3 mmol) and DMAP (250 mg) were added to a stirred solution of
methyl α-D-mannopyranoside 1 (10 g, 51.5 mmol) in pyridine (150 ml) and the reaction mixture heated
to 80°C. After 2.5 h, t.l.c. (ethyl acetate:methanol, 9:1) indicated complete conversion of the starting
material (Rf 0.1) to a single product (Rf 0.6). The reaction mixture was diluted with distilled water (1
ml) and concentrated in vacuo. The resulting syrup was dissolved in DCM (500 ml) and washed with a
saturated ammonium chloride solution (2×200 ml). The organic phase was dried (MgSO4), filtered and
concentrated in vacuo. The residue was purified by flash column chromatography (hexane:ethyl acetate,
1:1) to give the trityl ether 2 (16.89 g, 75%) as a white foam; [α]D +20.8 (c 0.91 in CHCl3) [lit.13
22
[α]D20 +24 (c, 1.1 in CHCl3)]; δH (200 MHz, CDCl3): 3.39 (3H, s, MeO), 3.44 (2H, m), 3.69–3.79 (3H,
m), 3.91 (1H, s), 4.73 (1H, d, J1,2=1.5 Hz, H-1), 7.21–7.49 (15H, m, Ph3C).
5.4. Methyl 2,3,4-tri-O-(4-methoxybenzyl)-6-O-triphenylmethyl-α-D-mannopyranoside 3
4-Methoxybenzyl chloride (3.74 ml, 27.5 mmol) was added to a stirred solution of the triol 2 (3.0
g, 6.8 mmol) in DMF (15 ml). The reaction mixture was cooled to 0°C and sodium hydride (826 mg,
34.4 mmol) added portionwise. The mixture was allowed to warm to room temperature. After 16 h, t.l.c.
(hexane:ethyl acetate, 3:1) indicated complete conversion of starting material (Rf 0.1) to a single product
(Rf 0.3). The reaction mixture was quenched with methanol, diluted with DCM (200 ml) and washed
with distilled water (3×100 ml). The organic phase was dried (MgSO4), filtered and concentrated in
vacuo. The residue was purified by flash column chromatography (hexane:ethyl acetate, 3:1) to give 3
(4.136 g, 76%) as a white foam; (found: C, 75.24; H, 6.64. C50H52O9 requires: C, 75.40; H, 6.58%);
[α]D26 +15.2 (c 1.0 in CHCl3); νmax (KBr disk): no significant peaks; δH (400 MHz, CDCl3): 3.23 (1H,
dd, J3,2=5.5 Hz, J3,4=9.8 Hz, H-3), 3.38 (3H, s, MeO), 3.47 (1H, dd, J4,5=1.5 Hz, H-4), 3.74–3.77 (2H,
m, H-2, 1H), 3.78, 3.79, 3.81 (9H, 3×s, 3×MeOBn), 3.83 (1H, m), 3.91 (1H, a-t, J=9.5 Hz), 4.19 (1H,
d, J=12.1 Hz, PhCH2), 4.56 (2H, s, PhCH2), 4.63, 4.66 (2H, 2×s, PhCH2), 4.76 (2H, m, H-1, PhCH2),
6.69–6.71, 6.78–6.87, 7.21–7.36, 7.49–7.52 (27H, 4×m, Ar); δC (50.3 MHz, CDCl3): 54.5 (q, MeO),
55.3 (q, 3×MeOBn), 63.5, 71.9, 72.0, 74.7 (4×t, C-6, 3×CH2Ph), 71.9, 74.7, 74.8, 80.1 (4×d, C-2, C-3,
C-4, C-5), 86.6 (s, CPh3), 98.9 (d, C-1), 119.7, 119.9, 127.1, 127.5, 128.0, 128.2, 128.5, 128.5, 129.9,
+
129.2, 129.5, 130.1 (12×d, Ar), 144.5, 144.8 (2×s, Ar); m/z (electrospray): 814 (M+NH4 , 100%).
5.5. Methyl 2,3,4-tri-O-(4-methoxybenzyl)-α-D-mannopyranoside 4
A solution of the trityl ether 3 (17.87 g, 22.4 mmol) was stirred in ethanol:acetic acid (1:1, 100 ml) at
80°C. After 4 h, t.l.c. (hexane:ethyl acetate, 3:1) indicated complete conversion of the starting material
(Rf 0.7) to a single product (Rf 0.3). The reaction mixture was concentrated in vacuo (co-evaporation with
toluene) and the resulting residue purified by flash column chromatography (hexane:ethyl acetate, 1:1)
to give the alcohol 4 (8.804 g, 71%) as a white foam; (found: C, 67.30; H, 6.96. C31H38O9 requires: C,
67.14; H, 6.91%); [α]D26 +30.4 (c 1.0 in CHCl3); νmax (KBr disk): 3480 (br, OH) cm−1; δH (400 MHz,
CDCl3): 2.08 (1H, s, OH), 3.29 (3H, s, MeO), 3.59 (1H, m, H-5), 3.75 (2H, m), 3.80, 3.81, 3.82 (9H,
3×s, 3×MeOBn), 3.84 (1H, m), 3.87 (2H, m), 4.56 (2H, s, PhCH2), 4.60 (1H, d, J=11.1 Hz, PhCH2),
4.65 (2H, m, H-1, PhCH2), 4.71 (1H, d, J=11.9 Hz, PhCH2), 4.86 (1H, d, J=10.6 Hz, PhCH2), 6.85–6.89,
7.23–7.29 (12H, 2×m, Ar); δC (50.3 MHz, CDCl3): 54.7 (q, MeO), 55.3 (q, 3×MeOBn), 62.4, 71.9, 72.6,
74.9 (4×t, C-6, 3×CH2Ph), 72.1, 74.3, 74.7, 79.9 (4×d, C-2, C-3, C-4, C-5), 99.5 (d, C-1), 113.8, 129.4,

R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
239
129.8, 129.9 (4×d, Ar), 130.5, 130.8 (2×s, Ar); m/z (electrospray): 577 (M+Na+, 100), 593 (M+K+,
11%).
5.6. Methyl 2,3,4,6-tetra-O-benzyl-α-D-mannopyranoside 5
Benzyl bromide (9.19 ml, 77 mmol) was added to a stirred solution of methyl α-D-mannopyranoside
1 (3.0 g, 15 mmol) in DMF (15 ml). The reaction mixture was cooled to 0°C and sodium hydride (2.23
g, 93 mmol) was then added portionwise. The mixture was then allowed to warm to room temperature.
After 17 h, t.l.c. (hexane:ethyl acetate, 3:1) indicated complete conversion of starting material (Rf 0.1)
to a major product (Rf 0.6). The reaction mixture was quenched with methanol, diluted with diethyl
ether (200 ml) and washed with distilled water (150 ml). The aqueous phase was extracted with diethyl
ether (200 ml) and the combined organic extracts dried (MgSO4), filtered and concentrated in vacuo.
The residue was purified by flash column chromatography (hexane:ethyl acetate, 4:1) to give 5 (7.138 g,
83%) as a pale yellow oil; [α]D22 +29.2 (c 1.59 in CHCl3) [lit.14 [α]D +28.0 (c 1.0 in CHCl3)]; δH (500
MHz, CDCl3): 3.34 (3H, s, MeO), 3.75–3.81 (4H, m), 3.89 (1H, dd, J=3.1, 9.6 Hz), 3.99 (1H, a-t, J=9.0
Hz), 4.79 (1H, d, J1,2=1.6 Hz, H-1), 4.52 (1H, d, J=10.8 Hz, CH2Ph), 4.57 (1H, d, J=12.1 Hz, CH2Ph),
4.62 (2H, s, CH2Ph), 4.68 (1H, d, J=12.1 Hz, CH2Ph), 4.75 (2H, d, 3.8 Hz, CH2Ph), 4.89 (1H, d, J=10.8
Hz, CH2Ph), 7.17–7.40 (20H, m, Ar).
5.7. 1,6-Di-O-acetyl-2,3,4-tri-O-benzyl-α-D-mannopyranoside 6
Concentrated sulphuric acid (2 ml) was added dropwise to a stirred solution of 5 (10.78 g, 19.5
mmol) in acetic acid:acetic anhydride (1:1, 100 ml) at 0°C. After 30 min, t.l.c. (hexane:ethyl acetate,
3:1) indicated complete conversion of the starting material (Rf 0.5) to a major product (Rf 0.3). The
reaction mixture was washed with saturated sodium bicarbonate solution (100 ml) and ice cold distilled
water (50 ml). The aqueous phase was extracted with DCM (2×150 ml) and the combined organics dried
(MgSO4), filtered and concentrated in vacuo. The residue was purified by flash column chromatography
(hexane:ethyl acetate, 2:1) to give the diacetate 6 (9.361 g, 90%) as a white solid; m.p. 92–96°C
(hexane/ethyl acetate); (found: C, 69.53; H, 6.41. C31H34O8 requires: C, 69.65; H, 6.41%); [α]D22 +31.3
(c 1.0 in CHCl3); νmax (CHCl3, thin-film): 1740 (2×C_O) cm−1; δH (500 MHz, CDCl3): 2.04, 2.07 (6H,
2×s, 2×MeCO), 3.75 (1H, dd, J2,3=2.9 Hz, H-2), 3.88 (2H, m, H-3, H-5), 3.99 (1H, a-t, J4,5=J4,3=9.5
Hz, H-4), 4.33 (2H, m, H-6, H-60), 4.60 (3H, m, CH2Ph), 4.73 (1H, d, J=12.1 Hz, CH2Ph), 4.78 (1H, d,
J=12.1 Hz, CH2Ph), 4.96 (1H, d, J=10.6 Hz, CH2Ph), 6.19 (1H, d, J1,2=2.0 Hz, H-1), 7.29–7.42 (15H,
m, Ar); δC (50.3 MHz, CDCl3): 20.9, 21.0 (2×q, 2×MeCO), 63.2, 72.0, 72.6, 75.4 (4×t, 3×CH2Ph,
C-6), 72.4, 73.2, 73.8, 79.1, 91.6 (5×d, C-1, C-2, C-3, C-4, C-5), 127.7–128.5 (15×d, Ar), 137.7, 137.9
+
(3×s, Ar), 168.9, 170.9 (2×s, 2×C_O); m/z (Cl, NH3): 552 (M+NH4 , 35%).
5.8. Ethyl 6-O-acetyl-2,3,4-tri-O-benzyl-1-thio-α-D-mannopyranoside 7
Ethanethiol (5.38 ml, 48.4 mmol) was added to a stirred solution of diacetate 6 (8.614 g, 16.1 mmol)
in DCM (80 ml). The reaction mixture was cooled to 0°C and boron trifluoride etherate (3.07 ml, 24.2
mmol) added dropwise. After 3 h at 0°C, t.l.c. (hexane:ethyl acetate, 3:1) indicated complete conversion
of starting material (Rf 0.4) to a major product (Rf 0.6). The reaction mixture was diluted with DCM (100
ml) and washed with saturated sodium bicarbonate solution (200 ml) and brine (200 ml). The aqueous
phase was extracted with DCM (2×200 ml) and the combined organic extracts dried (MgSO4), filtered
and concentrated in vacuo. The residue was purified by flash column chromatography (hexane:ethyl

240
R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
acetate, 5:1) to give the thioglycoside 7 (5.579 g, 65%) as a clear oil; (found: C, 69.51; H, 6.66.
22
C31H36O6S requires C, 69.38; H, 6.76%); [α]D +73.8 (c 0.91 in CHCl3); νmax (CHCl3, thin-film):
1740 (C_O) cm−1; δH (500 MHz, CDCl3): 1.26 (3H, t, MeCH2S), 2.06 (3H, s, MeCO), 2.59 (2H, m,
MeCH2S), 3.85 (2H, m, H-2, H-3), 3.95 (1H, dd, J4,3=9.2 Hz, J4,5=9.6 Hz, H-4), 4.18 (1H, ddd, J5,6=2.2
0
0
0
Hz, J5,6 =5.0 Hz, H-5), 4.31 (1H, dd, J6,6 =11.8 Hz, H-6), 4.36 (1H, dd, H-6 ), 4.58 (3H, m, CH2Ph),
4.68 (2H, m, CH2Ph), 4.94 (1H, d, J=10.8 Hz, CH2Ph), 5.37 (1H, d, J1,2=1.0 Hz, H-1), 7.27–7.41 (15H,
m, Ar); δC (50.3 MHz, CDCl3): 15.1, 21.7 (2×q, SCH2Me, MeCO), 25.5 (t, SCH2Me), 63.7, 72.0, 75.4
(3×t, 3×CH2Ph, C-6), 70.3, 74.8, 76.3, 80.6, 82.2 (5×d, C-1, C-2, C-3, C-4, C-5), 128.0–129.1 (15×d,
Ar); m/z (APCI+): 558.9 (M+H+, 100%).
5.9. Ethyl 2,3,4-tri-O-benzyl-1-thio-α-D-mannopyranoside 8
A solution of sodium (115 mg) in methanol (25 ml) was added to a stirred solution of the acetate
7 (4.218 g) in methanol (25 ml) at 0°C, under argon. After 10 min, t.l.c. (hexane:ethyl acetate, 3:1)
indicated complete conversion of the starting material (Rf 0.4) to a single product (Rf 0.2). 1 M HCl (5
ml) and brine (100 ml) were then added. The aqueous phase was extracted with DCM (2×150 ml) and
the combined organic extracts were dried (MgSO4), filtered and concentrated in vacuo. The residue was
purified by flash column chromatography (hexane:ethyl acetate, 2:1) to give the alcohol 8 (3.504 g, 90%)
as a clear oil; [α]D +88.4 (c 1.15 in CHCl3) [lit.15 [α]D +88.0 (c 1.0 in CHCl3)]; δH (500 MHz,
CDCl3): 1.29, (3H, t, SCH2Me), 1.99 (1H, s, OH), 2.61 (2H, m, SCH2Me), 3.87 (2H, m, H-4, H-6), 3.89
(2H, m, H-2, H-60), 4.05, (2H, m, H-3, H-5), 4.66, 4.76 (4H, 2×m, 2×CH2Ph), 4.71, (1H, d, J=10.9 Hz,
CH2Ph), 4.99 (1H, d, J=10.9 Hz, CH2Ph), 5.36 (1H, s, H-1), 7.31–7.44 (15H, m, Ar).
22
22
5.10. General procedure for peptide couplings
Piperidine (0.5 ml) was added dropwise to a stirred solution of the Fmoc protected amino acid (0.1
mmol) in DMF (2.5 ml). After 30 min, t.l.c. (hexane:ethyl acetate, 1:9) typically indicated complete
conversion of starting material (Rf 0.3) to a single product (Rf 0.1). The solvents were removed in
vacuo (co-evaporation with toluene) and the residue dissolved in benzene:ethanol (1:1, 2 ml). EEDQ
(31 mg, 0.13 mmol) and the other amino acid (free carboxylic acid) (0.15 mmol, 1.5 equiv.) were added
and the solution stirred under argon. After 15 h, t.l.c. (hexane:ethyl acetate, 1:9) typically indicated
conversion of starting material (Rf 0.1) to a major product (Rf 0.2). The solvents were removed in vacuo
and the residue was purified by flash column chromatography (hexane:ethyl acetate, 1:9) to give the
Fmoc protected oligopeptide, typically as a white foam (yields ∼80%). Simple iteration of this sequence
allowed construction of all glycopeptides 14a–f.
5.11. General procedure for glycosylation reactions
The peptide (14a–f, 0.44 mmol) was stirred with molecular sieves in DCM (5 ml) for 1 h at room
temperature. The reaction mixture was then cooled to –40°C and NIS (94 mg, 0.42 mmol) and triflic acid
(2.4×10−3 ml, 0.03 mmol) were added. After 5 min, t.l.c. (hexane:ethyl acetate, 1:5) typically indicated
conversion of starting material (Rf 0.1) to two or three major products (Rf 0.6, 0.8 and 0.9). A single
drop of collidine was added and the mixture was filtered through Celite®, diluted with DCM (200 ml)
and washed with 10% aqueous sodium thiosulphate solution (100 ml). The aqueous phase was extracted
with DCM (200 ml) and the combined organic phases dried (MgSO4), filtered and concentrated in vacuo.
The residue was purified by flash column chromatography (DCM:diethyl ether, 4:1 to 2:1) to typically

R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
241
give two or three products as clear oils which were used without characterisation. Each of the separated
products were individually stirred in methanol (3 ml) in the presence of K2CO3 (5 mg) for 30 min. They
were then concentrated in vacuo and dissolved in a mixture of pyridine:acetic anhydride (1:1, 2 ml).
After 4 h, t.l.c. (hexane:ethyl acetate, 2:1) indicated the formation of a single product in each case. The
solvents were removed in vacuo and the individual disaccharides purified by flash chromatography.
5.12. Methyl (6-O-acetyl-2,3,4-tri-O-benzyl-α-D-mannopyranosyl)-(1→3)-2,4,6-tri-O-acetyl-α-D-man-
nopyranoside 16
A clear oil; [α]D +35.5 (c 0.61 in CHCl3); νmax (CHCl3, thin-film): 1745 (s, 4×C_O) cm−1; δH
22
(400 MHz, CDCl3): 1.86, 2.06, 2.09, 2.11 (12H, 4×s, 4×MeCO), 3.38 (3H, s, MeO), 3.67 (1H, a-t,
0
J2b,3b=J2b,1b=2.3 Hz, H-2b), 3.76–3.87 (4H, m, H-3b, H-4b, H-5b, H-5a), 4.08 (1H, dd, J6a,6a =12.2 Hz,
J6a,5a=2.4 Hz, H-6a), 4.13 (1H, dd, J3a,2a=3.4 Hz, J3a,4a=9.8 Hz, H-3a), 4.20–4.30 (3H, m, H-6a0, H-6b,
H-6b0), 4.55 (1H, d, J=11.1 Hz, CH2Ph), 4.61 (2H, m, CH2Ph), 4.67 (2H, ABq, J=12.0 Hz, CH2Ph),
4.74 (1H, d, J1a,2a=1.5 Hz, H-1a), 4.84 (1H, d, J=11.1 Hz, CH2Ph), 4.95 (1H, d, H-1b), 5.15 (1H, dd,
H-2a), 5.21 (1H, dd, J4a,5a=10.0 Hz, H-4a), 7.28–7.35 (15H, m, Ar); δC (50.3 MHz, CDCl3): 20.6, 20.7,
20.7, 20.9 (4×q, MeCO), 55.2 (q, MeO), 62.1 (t, C-6a), 63.5 (t, C-6b), 67.9 (d, C-4a), 68.4 (d, C-5a),
70.9 (d, C-5b), 71.5 (d, C-2a), 72.3 (t, CH2Ph), 72.7 (t, CH2Ph), 74.2 (d, C-3a), 74.4 (d, C-4b), 74.6 (t,
CH2Ph), 74.8 (d, C-2b), 79.7 (d, C-3b), 98.3 (d, C-1a), 100.5 (d, C-1b), 127.6, 127.6, 127.7, 127.8, 128.0,
128.3, 128.4 (7×d, Ar), 138.0 (s, Ar), 169.5, 170.1, 170.7, 170.9 (4×s, 4×MeCO); m/z (electrospray):
812.48 (M+NH4 , 100%). HRMS (electrospray) calcd. for C42H50O15Na (M+Na+): 817.3047. Found:
+
817.3047.
5.13. Methyl (6-O-acetyl-2,3,4-tri-O-benzyl-β-D-mannopyranosyl)-(1→3)-2,4,6-tri-O-acetyl-α-D-man-
nopyranoside 17
A clear oil; [α]D –20.7 (c 1.23 in CHCl3); νmax (CHCl3, thin-film): 1744 (s, 4×C_O) cm−1; δH
22
(400 MHz, CDCl3): 1.99, 2.02, 2.11, 2.12 (12H, 4×s, 4×MeCO), 3.41 (3H, s, MeO), 3.44 (1H, ddd,
0
J5b,4b=9.6 Hz, J5b,6b=4.4 Hz, J5b,6b =2.8 Hz, H-5b), 3.52 (1H, dd, J3b,4b=9.5 Hz, J3b,2b=2.9 Hz, H-3b),
0
3.75 (1H, d, H-2b), 3.86 (1H, dd, H-4b), 3.94 (1H, ddd, J5a,4a=10.0 Hz, J5a,6a=2.4 Hz, J5a,6a =5.6 Hz,
H-5a), 4.16–4.19 (2H, m, H-3a, H-6a), 4.28 (1H, dd, J6a ,6a=12.2 Hz, H-6a0), 4.34 (2H, m, H-6b, H-6b0),
0
4.46 (1H, s, H-1b), 4.53–4.62 (4H, m, CH2Ph), 4.74 (1H, d, J1a,2a=1.5 Hz, H-1a), 4.82 (1H, d, J=12.2 Hz,
CH2Ph), 4.93 (1H, d, J=10.9 Hz, CH2Ph), 5.22–5.29 (2H, m, H-2a, H-4a), 7.25–7.43 (15H, m, Ar); δC
(50.3 MHz, CDCl3): 20.6, 20.7, 20.7, 20.9 (4×q, 4×MeCO), 55.1 (q, MeO), 62.8 (t, C-6a), 63.1 (t, C-6b),
66.2 (d, C-4a), 68.0 (d, C-2a), 68.5 (d, C-5a), 71.3 (t, CH2Ph), 72.9 (d, C-3a), 73.6 (d, t, C-5b, CH2Ph),
73.9 (d, C-2b), 74.1 (d, C-4b), 75.1 (t, CH2Ph), 82.2 (d, C-3b), 98.5 (d, C-1b), 98.9 (d, C-1a), 127.3,
127.6, 127.8, 127.9, 128.0, 128.1, 128.3, 128.4 (8×d, Ar), 138.0, 138.1, 148.9 (3×s, Ar), 170.1, 170.3,
+
170.6, 170.7 (4×s, 4×MeCO); m/z (electrospray): 812.84 (M+NH4 , 100%). HRMS (electrospray) calcd
for C42H50O15Na (M+Na+): 817.3047. Found: 817.3044.
5.14. Methyl (6-O-acetyl-2,3,4-tri-O-benzyl-β-D-mannopyranosyl)-(1→2)-3,4,6-tri-O-acetyl-α-D-man-
nopyranoside 18
A clear oil; [α]D –30.1 (c 0.58 in CHCl3); νmax (CHCl3, thin-film): 1744 (s, 4×C_O) cm−1; δH
22
(400 MHz, CDCl3): 1.97, 1.99, 2.06, 2.07 (12H, 4×s, 4×MeCO), 3.43 (3H, s, MeO), 3.40–3.45 (1H, m,
H-5b), 3.50 (1H, dd, J3b,2b=3.0 Hz, J3b,4b=9.2 Hz, H-3b), 3.81 (1H, dd, J4b,5b=9.7 Hz, H-4b), 3.94 (1H,

242
R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
0
0
ddd, J5a,4a=10.1 Hz, J5a,6a=2.1 Hz, J5a, =4.3 Hz, H-5a), 3.99 (1H, d, H-2b), 4.10 (1H, dd, J6a,6a =12.2
6a
Hz, H-6a), 4.26–4.34 (2H, m, H-6b, H-6b0), 4.40–4.43 (2H, m, H-2a, H-6a0), 4.44 (2H, s, H-1b, CH2Ph),
4.54 (2H, m, CH2Ph), 4.75 (1H, d, J1a,2a=1.3 Hz, H-1a), 4.92 (2H, m, CH2Ph), 5.09–5.14 (2H, m, H-3a,
CH2Ph), 5.48 (1H, dd, J4a,3a=10.2 Hz, H-4a), 7.26–7.58 (15H, m, Ar); δC (50.3 MHz, CDCl3): 20.7,
20.8, 20.8 (3×q, 4×MeCO), 55.2 (q, MeO), 62.4 (t, C-6a), 63.7 (t, C-6b), 65.5 (d, C-4a), 68.5 (d, C-
5a), 70.0 (d, C-3a), 70.9 (t, CH2Ph), 72.7 (d, C-2a), 73.2 (d, C-2b), 73.7 (d, t, C-5b, CH2Ph), 74.5 (d,
C-4b), 75.2 (t, CH2Ph), 81.6 (d, C-3b), 98.6 (d, C-1a), 99.9 (d, C-1b), 127.5, 127.6, 127.7, 127.8, 127.8,
128.1, 128.2, 128.4, 128.4 (9×d, Ar), 137.8, 137.9, 138.7 (3×s, Ar), 169.4, 170.1, 170.2, 170.3 (4×s,
+
4×MeCO); m/z (electrospray): 812.42 (M+NH4 , 100%). HRMS (electrospray) calcd for C42H50O15Na
(M+Na+): 817.3047. Found: 817.3046.
5.15. Methyl (6-O-acetyl-2,3,4-tri-O-benzyl-α-D-mannopyranosyl)-(1→2)-3,4,6-tri-O-acetyl-α-D-man-
nopyranoside 19
A clear oil; [α]D +22.6 (c 0.24 in CHCl3); νmax (KBr disk): 1747 (s, 4×C_O) cm−1; δH (400
24
MHz, C6D6): 1.63, 1.66, 1.68, 1.75 (12H, 4×s, 4×MeCO), 3.78 (1H, ddd, J5a,4a=9.9 Hz, J5a,6a=2.5 Hz,
0
J5a,6a =4.4 Hz, H-5a), 3.93 (1H, a-t, J2b,3b=J2b,1b=2.5 Hz, H-2b), 4.03 (1H, a-t, J4b,3b=J4b,5b=9.2 Hz, H-
0
4b), 4.08 (1H, a-t, J2a,1a=J2a,3a=2.4 Hz, H-2a), 4.11 (1H, dd, H-3b), 4.14 (1H, dd, J6a,6a =12.2 Hz, H-6a),
0
0
0
4.21 (1H, ddd, J5b,6b=6.2 Hz, J5b,6b =1.8 Hz, H-5b), 4.26 (1H, dd, H-6a ), 4.31 (1H, dd, J6b,6b =11.8 Hz,
H-6b), 4.40–4.47 (3H, m, CH2Ph), 4.56 (2H, m, H-6b0, CH2Ph), 4.67 (1H, d, J=12.2 Hz, CH2Ph), 4.85
(1H, d, J=11.0 Hz, CH2Ph), 5.02 (2H, m, H-1a, H-1b), 5.61 (1H, dd, J3a,4a=9.9 Hz, H-3a), 5.67 (1H,
d, H-4a), 6.98–7.37 (15H, m, Ar); δC (50.3 MHz, CDCl3): 20.3, 20.4, 20.4 (3×q, 4×MeCO), 54.9 (q,
MeO), 61.6 (t, C-6a), 61.9 (d, C-3b), 63.4 (t, C-6b), 66.3 (d, C-4a), 68.8 (d, C-5a), 70.9 (d, C-3a), 71.3
(d, C-5b), 72.3, 72.9, 75.0 (3×t, 3×CH2Ph), 75.7 (d, C-2b), 76.6 (d, C-4b), 80.2 (d, C-2a), 99.3 (d, C-
1a), 100.5 (d, C-1b), 127.8, 128.0, 128.2, 128.3, 128.5, 128.6, 128.7 (7×d, Ar), 138.8, 138.9 (2×s,Ar),
169.4, 169.9, 170.2 (3×s, 4×CO); m/z (FAB+): 817.4 (M+Na+, 47%). HRMS (electrospray) calcd for
C42H50O15Na (M+Na+): 817.3047. Found: 817.3048.
Acknowledgements
We gratefully acknowledge the EPSRC (Quota award to RTE) and the University of Durham for
financial support. We would also like to acknowledge the use of the EPSRC’s Mass Spectrometry Service
at Swansea, UK, and of the EPSRC’s Chemical Database Service at Daresbury.
References
1. Varki, A. Glycobiology 1993, 3, 97.
2. For some recent reviews and books, see: Toshima, K.; Tatsuta, K. Chem. Rev. 1993, 93, 1503; Boons, G.-J. Tetrahedron 1996,
52, 1095; Modern Methods in Carbohydrate Synthesis; Khan, S. H.; O’Neill, R. A., Eds.; Harwood Academic: Amsterdam,
1996; Preparative Carbohydrate Chemistry; Hanessian, S., Ed.; Marcel Dekker: New York, 1997.
3. For a preliminary communication of some of this work, see: Tennant-Eyles, R. J.; Davis, B. G.; Fairbanks, A. J. Chem.
Commun. 1999, 1037.
4. Stork, G.; Kim, G. J. Am. Chem. Soc. 1992, 114, 1087.
5. Barresi, F.; Hindsgaul, O. Can. J. Chem. 1994, 72, 1447; Barresi, F.; Hindsgaul, O. Synlett 1992, 759; Barresi, F.; Hindsgaul,
O. J. Am. Chem. Soc. 1991, 113, 9376.
6. Ito, Y.; Ohnishi, Y.; Ogawa, T.; Nakahara, Y. Synlett 1998, 1102; Dan, A.; Ito, Y.; Ogawa, T. J. Org. Chem. 1995, 60, 4680;
Ito, Y.; Ogawa, T. Angew. Chem., Int. Ed. Engl. 1994, 33, 1765.

R. J. Tennant-Eyles et al. / Tetrahedron: Asymmetry 11 (2000) 231–243
243
7. Zeigler, T.; Lemanski, G.; Rakoczy, A. Tetrahedron Lett. 1995, 36, 8973; Zeigler, T.; Lau, R. Tetrahedron Lett. 1995, 36,
1417; Lau, R.; Schülle, G.; Schwaneberg, U.; Zeigler, T. Liebigs Ann. 1995, 1475; Yamada, H.; Imamura, K.; Takahashi,
T. Tetrahedron Lett. 1997, 36, 391; Valverde, S.; Gómez, A. M.; Hernández, A.; Herradón, B.; López, J. C. J. Chem. Soc.,
Chem. Commun. 1995, 2005.
8. Ponpipom, M. M. Carohydr. Res. 1977, 59, 311.
9. Tennant-Eyles, R. J.; Fairbanks, A. J. Tetrahedron: Asymmetry 1999, 10, 391.
10. Borén, H. B.; Garegg, P. J.; Kenne, L.; Pilotti, A.; Svenson, S.; Swahn, C.-G. Acta. Chem. Scand. 1973, 27, 2740; Bianco,
A.; Brufani, M.; Melchioni, C.; Romagnoli, P. Tetrahedron Lett. 1997, 38, 651.
11. HMBC experiments indicated cross coupling between C-1b/H-1b and the corresponding H-a/C-a, respectively (a/b referring
to the acceptor/donor, respectively). The chemical shift of the correspondingly non-acetylated H-a position was also
diagnostic (shielded by approximately 1 ppm). The stereochemistry of the newly formed anomeric linkage was also
confirmed by measurement of the JC-1,H-1 coupling constants (Jα=165–175 Hz, Jβ=154–157 Hz).
12. For a manifesto, see: Gellman, S. H. Acc. Chem. Res. 1998, 31, 173.
13. Winkler, A. D.; Holan, G. J. Med. Chem. 1989, 32, 2084.
14. Fugedi, P.; Liptak, A.; Namasi, P.; Neszmelyi, A. Carbohydr. Res. 1982, 107, C5.
15. Ottoson, H. Carbohydr. Res. 1990, 197, 101.