Automated Synthesis of Oligosaccharides

Article abstract of DOI:10.1002/hlca.200390331

The chemical synthesis of oligosaccharides with an automated solid-phase synthesizer is described. An octenediol linker served to attach the growing oligosaccharide chain to the solid support, and the desired structures were cleaved from the support via o

Full text of DOI:10.1002/hlca.200390331

Helvetica Chimica Acta Vol. 86 (2003)
3975
Automated Synthesis of Oligosaccharides
by Emma R. Palmacci, Obadiah J. Plante, Michael C. Hewitt, and Peter H. Seeberger*1)
Massachusetts Institute of Technology (MIT), Department of Chemistry, 77 Massachusetts Avenue, Cambridge,
MA 02139, USA
Dedicated to Professor Duilio Arigoni on the occasion of his 75th birthday
The chemical synthesis of oligosaccharides with an automated solid-phase synthesizer is described. An
octenediol linker served to attach the growing oligosaccharide chain to the solid support, and the desired
structures were cleaved from the support via olefin metathesis to afford a pentenyl glycoside. The automated
syntheses of several important carbohydrates, including a pentarhamnoside, a proteoglycan linkage-region
tetrasaccharide, a phytoalexin elicitor dodecasaccharide, and a branched Leishmania lipophosphoglycan
tetrasaccharide, were accomplished in higher overall yield and ca. 20times faster than with solution-phase
methods.
Introduction. Nonspecialists can routinely access peptides and oligonucleotides
for biochemical studies with commercially available machines, while only highly
specialized laboratories can synthesize complex carbohydrates. Our effort towards the
automated synthesis of oligosaccharides was inspired by automated methods for
peptide and DNA synthesis, and the impact these techniques have had on proteomics
and genomics. Here, we describe the development and use of the first automated
oligosaccharide synthesizer.
The traditional solution-phase chemistry of complex carbohydrates remains a time-
consuming art. Although a plethora of new methods for the construction of
oligosaccharides, such as enzymatic [1] and one-pot procedures, [2][3] have been
described, few are generally applicable. The OptiMer One-Pot approach [4] aims to
automate synthesis planning by selecting efficient couplings based on a computer
program and has been successfully applied to the construction of several complex
oligosaccharides [5]. As an alternative to solution-phase assembly, the solid-phase
synthesis of oligosaccharides has received increased attention [6]. The elimination of
intermediate purification steps has led to the accelerated synthesis of several complex
structures. We sought to exploit the benefits of the solid-phase paradigm, without the
need for manual manipulations.
Influenced by the peptide-synthesizer platform [7][8], we envisioned a similar
machine that could easily accomplish glycosylation and deprotection reactions in an
analogous fashion. In this way, the synthesis of complex carbohydrates could be
reduced to the acquisition of suitably protected mono- or disaccharide glycosyl donors.
The building blocks and reaction cycles required for automated oligosaccharide
1
)
Current address: Laboratorium f¸r Organische Chemie, ETH-Hˆnggerberg, CH-8093 Z¸rich (Phone:
‡‡ 41-1-6332103, fax: ‡‡ 41-1-6331235, e-mail: seeberger@org.chem.ethz.ch).

3976
Helvetica Chimica Acta Vol. 86 (2003)
construction were determined through the synthesis of representative structures. The
automated synthesis of oligosaccharides must occur in a linear fashion, thereby
departing from the traditional goal of convergency. Initial solution-phase syntheses
validated the proposed linear synthetic strategy and provided standards for the
confirmation of products obtained from the synthesizer. The design of the automated
oligosaccharide synthesizer took several key issues into account: appropriate linker
systems and polymeric resins; competent glycosylating agents and protecting groups;
and efficient cleavage and purification methods.
Results and Discussion.
The solid-phase synthesis of carbohydrates can be
accomplished by attaching the first sugar either through the reducing end (−acceptor-
bound×) or the non-reducing end (−donor-bound×). Several large structures have been
synthesized by the donor-bound approach [9], but side reactions involving the resin-
bound glycosyl donor result in a direct loss in coupling yield. To increase the overall
yield, the acceptor-bound assembly method was chosen for the automated synthesizer.
Glycosyl trichloroacetimidates [10][11] and glycosyl phosphates [12] were selected as
the building blocks due to their reliability and prior success in solid-phase
oligosaccharide synthesis [13].
Oct-4-ene-1,8-diol linker 3 [13], stable to acidolysis and saponification unlike most
other linkers [14], provided the basis for our protecting-group concept. Linker 3 is
stable under a host of conditions but is cleanly cleaved by olefin metathesis (Scheme 1)
[15]. Resin loading is determined by detritylation of 2. Acetate, levulinate, and silyl
ether protecting groups are fully compatible with the oct-4-ene-1,8-diol linker, and their
removal readily reveals the OH group ready for coupling with the glycosyl donor
delivered in solution. The flexibility of this linker system was essential to the success of
the automated synthesizer.
Scheme 1. Functionalization and Use of Oct-4-ene-1,8-diol Linker. Shaded circle ˆ polystyrene resin.
Initial work on the manual solid-phase synthesis of carbohydrates was accom-
plished with Merrifield×s resin (1% crosslinked polystyrene) functionalized with the
oct-4-ene-1,8-diol linker. Attempts to functionalize polyethylene glycol (PEG) grafted
¾
Tentagel polystyrene support with the linker system were unsuccessful. While
¾
¾
Argogel , a similar PEG-grafted resin, and Argopore , a highly crosslinked polystyrene
resin, were successfully functionalized, and initial results indicated that Merrifield×s

Helvetica Chimica Acta Vol. 86 (2003)
3977
resin was the best choice for the automated synthesis. Resin loadings in the range of
0.1 0.3 mmol/g were routinely accomplished with the polystyrene resin.
A commercially available Applied Biosystems model 433A peptide synthesizer was
reconfigured to carry out the automated solid-phase oligosaccharide synthesis on a
0.1 1.0-mmol scale. Reagent bottles contained the chemicals necessary for oligosac-
charide synthesis. Cartridges that stored the glycosyl donor building blocks were
advanced in the order of incorporation to a needle that delivered the monomers to the
reaction vessel. The computer program SynthAssist¾, provided by Applied Biosystems
for peptide construction, was employed to design the automated cycles for oligosac-
charide synthesis.
While all the manipulations in peptide chemistry are carried out at room
temperature, many transformations in oligosaccharide synthesis occur below ambient
temperature. To address the need for variable reaction temperatures, a double-walled
glass reaction vessel (8-ml internal volume) was designed with inlet and outlet hose
fittings that were connected to an external circulating bath with heating and cooling
capabilities. A glass-frit-closed bottom of the reaction vessel, and the resin was loaded
and removed from the top of the vessel.
The construction of a-(1 ! 2)-oligomannosides had been accomplished by manual
solid-phase synthesis, and these structures provided a starting point to explore the
efficacy of automation [13][16]. Trichloroacetimidate donor 4 utilized a temporary
C(2)-acetate protecting group that provides anchimeric assistance. The activation of
donor 4 was accomplished by the addition of a catalytic amount of trimethylsilyl
trifluoromethanesulfonate (TMSOTf) in CH2Cl2. Deprotection was achieved through
the use of MeONa in a MeOH/CH2Cl2 solution. A synthesis cycle consisting of coupling
and deprotection steps was developed by varying the concentration and quantity of
reagents as well as the reaction times. Repetition of coupling and deprotection in each
cycle ensured complete reactivity, and washing cycles effectively removed the excess
reagent and solution-phase impurities.
Using oct-4-ene-1,8-diol functionalized Merrifield×s resin 3, double couplings, and
double deprotections (Table 1), the synthesis of pent-4-enyl pentamannoside 7 was
accomplished in 14 h in 74% isolated yield after cleavage with Grubbs catalyst in an
atmosphere of ethylene (Scheme 2). Heptamannoside 8 and decamannoside 9 were
prepared by the same automation cycle in average yields of 90 95% per step. The
Table 1. Automated Cycle for the Synthesis of Oligomannosides
Step
Function
Reagent
Time [min]
30
1
2
Couple
Wash
10equiv. donor and 0.5 equiv. TMSOTf
CH2Cl2
6
3
4
Couple
Wash
10equiv. donor and 0.5 equiv. TMSOTf
CH2Cl2
30
6
6
5
Wash
MeOH/CH2Cl2 1 : 9
6
7
8
9
Deprotection
Wash
Wash
2 Â 10equiv. MeONa (MeOH/CH 2Cl2 1 : 9)
80
4
4
MeOH/CH2Cl2 1 : 9
0.2m AcOH in THF
THF
Wash
6
10
Wash
CH2Cl2
6

3978
Helvetica Chimica Acta Vol. 86 (2003)
Scheme 2. Automated Synthesis of Polymannosides
short reaction times 3 h per monomer unit allowed for the synthesis of 9 in 20h. For
comparison, heptamannoside 8 was manually synthesized in our laboratory on the solid
support in 14 d and 9% overall yield [13]. The cycle developed for the synthesis of
polymannosides 7 9 provided a basis for general acetate deprotection and trichlor-
oacetimidate donor activation.
Besides glycosyl trichloroacetimidates, which are a staple of many oligosaccharide
syntheses, glycosyl phosphates have proven to be readily accessible and versatile
glycosylating agents [12]. The use of glycosyl phosphates in an automated protocol
demonstrated that the automated machine was compatible with different glycosylation
chemistries. The phytoalexin elicitor (PE) family of b-glucans [17] (Fig. 1) was selected
as the target to evaluate the use of glycosyl phosphates (for a preliminary
communication of this work, see [18]).
Fig. 1. Phytoalexin elicitor structure
Two different glycosyl phosphate donors that were accessible from glycal starting
materials were used for the synthesis of the branched b-(1 ! 3)/b-(1 ! 6) phytoalexin
elicitor structure (Scheme 3) [19 23]. A levulinoyl (Lev) ester served as the 6-O-
temporary protecting group while the 2-O-pivaloyl (Piv) group ensured b-selectivity in
the glycosylation reaction. Monosaccharide glycal 12 was prepared from the
corresponding 3,4-di-O-benzyl-6-O-TIPS (TIPS ˆ (i-Pr)3Si) glucal 11 by treatment
with Bu4NF (TBAF) and subsequent acylation. By our one-pot procedure [24], 12 was
converted into phosphate 13 in 95% yield. Disaccharide phosphate 17 was prepared via

Helvetica Chimica Acta Vol. 86 (2003)
3979
a similar route. The epoxidation of glucal 14 was followed by regioselective opening
with 6-O-TIPS glucal 10, and the crude reaction mixture was benzylated to afford the b-
(1 ! 3)-linked disaccharide glycal 15. Removal of the C(6)-TIPS and subsequent
installation of the C(6)-levulinoyl protecting group furnished glycal 16. Glycosyl
phosphate 17 was synthesized from 16 again by our one-pot procedure.
Scheme 3. Synthesis of Phosphate Building Blocks
The coupling and deprotection conditions for glycosyl phosphates and levulinates
were established (Table 2) by the synthesis of b-(1 ! 6) glucose trisaccharide 18 as a
model. Optimal yields with glycosyl phosphate 13 were accomplished with coupling
temperatures below ambient temperature. The HPLC traces of the crude reaction
products with glycosylations occurring at ‡258 (run 1) and at À158 (run 2) showed the
benefit of utilizing the vessel×s temperature-control abilities (Fig. 2). The activation of
13 with TMSOTf at À158 was complete within 15 min. The results of these
glycosylation experiments mirrored solution-phase oligosaccharide synthesis, where
low temperatures are commonly used. The levulinoyl ester was rapidly removed at
Table 2. Automated Cycle for the Synthesis of Phytoalexin Elicitors 18 20
Step
Function
Reagent
Time [min]
1
2
Couple
Wash
5 equiv. donor and 5 equiv. TMSOTf
CH2Cl2
30
6
3
4
5
Couple
Wash
Wash
5 equiv. donor and 5 equiv. TMSOTf
MeOH/CH2Cl2 1 : 9
THF
30
4
4
6
7
8
9
10
11
12
Wash
Deprotection
Wash
Wash
Wash
pyridine/AcOH 3 : 2
2 Â 20equiv. Hydrazine (pyridine/AcOH 3 : 2)
pyridine/AcOH 3 : 2
MeOH/CH2Cl2 1 : 9
0.2m AcOH in THF
THF
3
80
3
4
4
4
6
Wash
Wash
CH2Cl2

3980
Helvetica Chimica Acta Vol. 86 (2003)
Fig. 2. a) Analytical HPLC chromatogram of room-temperature triglucoside synthesis following resin cleavage.
b) Analytical HPLC chromatogram of À 158 triglucoside synthesis following resin cleavage. Flow rate: 1 ml/min,
20 ! 30% AcOEt/Hexanes (20 min): tR(18) ˆ 5 min.
‡158 with hydrazine in a solution of pyridine/AcOH. Double couplings and double
deprotections were utilized in analogy to the synthesis of the polymannosides.
The automated cycle developed for glycosyl phosphates and levulinoyl esters was
applied to the synthesis of the phytoalexin elicitor oligomers (Scheme 4). Hexasac-
charide 19 was constructed by using glycosyl phosphates 13 and 17 in alternating cycles
in 10h. Hexasaccharide 19 was the major product as determined by HPLC analysis.
With the same glycosyl donors and automated cycles, we prepared dodecasaccharide 20
in 17 h. Notably, our synthesis of 20 required the solution-phase synthesis of only two
glycosyl-phosphate building blocks, greatly reducing the manual labor usually
necessary for the assembly of large structures. The coupling cycles for glycosyl
phosphates and levulinoyl esters established in the synthesis of the phytoalexin elicitor

Helvetica Chimica Acta Vol. 86 (2003)
3981
Scheme 4. Automated Synthesis of Phytoalexin Elicitor Structures
structures may serve as the foundation for other automated syntheses utilizing these
groups.
Synthesis of the Proteoglycan Linkage Region. The proteoglycan-linkage region is
the attachment point for heparin and related glycosaminoglycan oligomers to the
proteoglycan core protein [25][26]. The linkage region is crucial to the biosynthesis of
these molecules, and studies to elucidate this biochemical pathway are on-going
[27][28]. The tetrasaccharide is attached to the core protein via a xylose-threonine/
serine bond and consists of xylose, two galactose units, and a terminal glucuronic acid
moiety (Fig. 3). In the biosynthetic pathway, a galactosamine or glucosamine unit
attached to the linkage region tetrasaccharide determines whether the glycosamino-
glycan structure will be chondroitin sulfate or heparin/heparan sulfate. Several previous
solution-phase syntheses of the tetrasaccharide have been accomplished; however,
none were amenable to automation [29 33]. The synthesis of 30 would necessitate
unique monosaccharide donors of xylose, galactose, and glucuronic acid, thereby
expanding the number of building blocks successfully incorporated in our automated
scheme.
Fig. 3. Proteoglycan linkage region tetrasaccharide
(R1 ˆ protein, R2 ˆ glycosaminoglycan chain)
The automated synthesis of the linkage-region tetrasaccharide required an efficient
xylose building block. Xylose phosphates had not been explored previously, and
synthesis of the glycosyl phosphate from the corresponding glycal starting material was
investigated. Starting from xylal 21 [34], the necessary differentiation of the OH groups
at C(3) and C(4) proved difficult [35][36]. Unlike monosaccharides containing the
C(5)-substituent, the two secondary OH groups of xylal are nearly equivalent in
reactivity. Utilizing tin-ether chemistry [37], a 3 :1 ratio of the 3-O-benzyl/4-O-benzyl-
xylal was obtained in 31% yield (Scheme 5). Installation of the 4-O-levulinoyl ester,
followed by creation of the glycosyl phosphate, afforded xylose donor 23.

3982
Helvetica Chimica Acta Vol. 86 (2003)
Scheme 5. Synthesis of Xylose Phosphate 23
Solution-phase validation of the proposed synthetic route commenced with the
coupling of the xylose phosphate 23 to pent-4-enyl alcohol, a useful model for the
functionalized resin. Removal of the temporary levulinoate protecting group afforded
monosaccharide acceptor 24. Glycosylation with galactose phosphate 25 [38], and
removal of the levulinoyl protecting group at C(3) gave disaccharide 27 (Scheme 6).
Attempts to install the second galactose unit with 25 failed, and unreacted acceptor 27
was recovered, thereby necessitating a new route to the Gal-b-(1 ! 3)-Gal linkage.
Scheme 6. Solution-Phase Synthesis of Proteoglycan-Linkage Region Tetrasaccharide 30
Glycosylation of 24 with C(2)-acetate galactose phosphate 26, followed by
deprotection, afforded disaccharide 28 in good yield. Reaction with glycosyl phosphate
26, in this case, gave good yields of the corresponding trisaccharide. Cleavage of the
temporary protecting group and glycosylation with glucuronic acid trichloroacetimi-
date 30 [39] afforded the final tetrasaccharide 31 in solution.
The automated synthesis of the proteoglycan linkage region was achieved by using
glycosyl donors 23, 26, and 30. Double glycosylations and double deprotections were
employed to ensure high conversion of all steps. Cleavage from the support with the
Grubbs catalyst under an atmosphere of ethylene afforded the crude tetrasaccharide
31. Analysis of the reaction mixture revealed that 31 was obtained as the major product,
demonstrating the successful incorporation of xylose, galactose, and glucuronic acid
into our automated synthesis scheme.

Helvetica Chimica Acta Vol. 86 (2003)
3983
Synthesis of Branched Structures. Different sections of the Leishmania lipophos-
phoglycan have served as targets for previous synthetic studies [30 42]. During our
efforts to identify a potential carbohydrate vaccine candidate for leishmaniasis [43], the
automated synthesis of the cap tetrasaccharide 37 was investigated. The automated
assembly of 37 presented a synthetic challenge since it required the incorporation of a
branching pattern, which had not been attempted previously in an automated context.
While acetates and levulinates had been used successfully as temporary protecting
groups, we needed to illustrate that these groups, integrated into one monosaccharide,
would allow for the synthesis of a branched structure.
A key challenge of the automated synthesis of 37 was the ability to selectively
remove the protecting groups on the central mannose unit. To examine this issue, a
solution-phase model study was conducted (Scheme 7). Reaction of glycosyl donor 32
with pent-4-enyl alcohol, upon activation with catalytic TMSOTf, generated 33.
Removal of the levulinoyl ester in the presence of the C(2)-acetate by treatment with
hydrazine furnished acceptor 34 in 71% yield. Installation of the difficult b-(1 ! 4)
linkage between galactose and mannose was readily accomplished by reaction with
galactose phosphate 35. Treatment with MeONa provided disaccharide acceptor 36.
These results confirmed the validity of our protecting group strategy en route to the
automated synthesis of 37.
Scheme 7. Solution-Phase Model for Automation of 37
With the four monosaccharide building blocks in hand, the automated synthesis of
the branched Leishmania tetrasaccharide was carried out (Scheme 8). Glycosylation of
the resin with 32, followed by levulinate cleavage and glycosylation with 35, afforded
Scheme 8. Automation of Leishmania Tetrasaccharide 37

3984
Helvetica Chimica Acta Vol. 86 (2003)
resin-bound disaccharide. Deprotection of the axial acetate and coupling with
mannosyl donor 4, followed by acetate saponification and reaction with 4, yielded
the resin bound Leishmania tetrasaccharide. Cleavage of the oct-4-ene-1,8-diol linker
using the Grubbs catalyst in an atmosphere of ethylene provided crude pent-4-enyl
glycoside tetrasaccharide 37. Analysis of the reaction products showed that the
tetrasaccharide was the major product along with the trisaccharide [44].
Synthesis of Difficult Linkages. While mannose donor 4 has been used extensively
for the construction of a-(1 ! 2)-mannosidic linkages, rhamnose donor 38 had been
employed only once previously [45]. Modest yields (53%) were obtained for coupling
with the C(2)-OH group of glucose under the activation conditions (BF3OEt2, CH2Cl2/
hexanes À208) reported. The synthesis of poly-b-(1 ! 2)-rhamnosides would require a
significant improvement over previous methods and prompted us to focus attention on
this challenge [46].
The initial coupling cycle employed the temperature-control feature of the jacket
vessel to synthesize pentarhamnoside 43 (Scheme 9). Double glycosylations were
performed at À158 for 15 min and the deprotection of the C(2)-acetate was carried out
Scheme 9. Automated Synthesis of Polyrhamnosides
twice at ‡158 (Table 3). After cleavage from the resin by olefin metathesis, HPLC
analysis revealed that the desired pentasaccharide 43 was formed in 35% yield.
However, the undesired tetramer 42 (40%) and trimer 41 (18%) were also observed.
Since no unreacted C(2)-OH compounds were observed in the HPLC chromato-
gram, it was postulated that the n À 1 oligomers were due, most likely, to incomplete
deprotection of the C(2)-acetate and did not stem from incomplete glycosylation.
While extending the reaction time to 1 h did improve the overall yield of pentamer 43,
we were interested in developing a shorter coupling cycle. Changing the resin from the
¾
swellable Merrifield×s resin to the functionalized Argopore support allowed the use of
a higher concentration of MeONa (40equiv.). To ensure full deprotection, the reaction
temperature was elevated to ‡308. This optimized reaction cycle employing the oct-4-

Helvetica Chimica Acta Vol. 86 (2003)
3985
Table 3. Automated Cycle Utilizing Rhamnosyl trichloracetimidate 38
Step
Function
Reagent
Time [min]
1
2
Couple
Wash
5 equiv. donor and 0.2 equiv. TMSOTf (À 158)
15
6
CH2Cl2
3
4
Couple
Wash
2.5 equiv. donor and 0.1 equiv. TMSOTf (158)
CH2Cl2
15
6
5
Wash
MeOH
4
6
7
8
9
Deprotection
Wash
Wash
Wash
Wash
2  40equiv. NaOMe, 0.75 m in MeOH (‡ 308)
20each
MeOH
4
4
6
6
0.2m AcOH in THF
THF
CH2Cl2
10
¾
ene-1,8-diol-functionalized Argopore resin resulted in the synthesis of pentarhamnose
43 in 77% yield, a considerable improvement over the previously reported synthesis.
Scale-Up and Deprotection of Automated Syntheses. The ability to scale-up the
automated process in order to obtain useful quantities of deprotected oligosaccharides
is essential. To address this challenge, we considered the assembly of a Gal-Man-
GlcNAc trisaccharide motif found in N-linked complex-type glycoproteins [47].
Trisaccharide 45 contains two challenging linkages: glycosylation onto the C(2)-OH of
mannose with glucosamine donors and glycosylation onto the C(4)-OH of glucosamine
[48].
The automated synthesis of trisaccharide 45 required the preparation of the
monosaccharide building blocks 4, 44, and 35. The reaction scale was increased from
25 mmol in the previous syntheses to 100 mmol, and double couplings and double
deprotections were utilized. The desired trisaccharide 45 was isolated in 43% yield, and
subsequent removal of all protecting groups was carried out in solution to afford 46 in
62% yield (Scheme 10). In comparison, trisaccharide 45 was isolated in 17% yield by
Scheme 10. Automated Synthesis of the Complex-Type Trisaccharide 46

3986
Helvetica Chimica Acta Vol. 86 (2003)
manual solution-phase synthesis. The synthesis and deprotection of trisaccharide 46
demonstrated that automated oligosaccharide synthesis is fully scalable, and the
products obtained were readily deprotected by established procedures [49].
Conclusions. Described is the development of the first automated solid-phase
oligosaccharide synthesizer and its application to the assembly of several biologically
relevant structures of oligosaccharides. Glycosyl phosphates and glycosyl trichloroa-
cetimidates serve well as building blocks for this automated synthesizer and esters and
silyl ethers can be used as temporary protecting groups. Linear and branched structures
were prepared along with several structures containing difficult linkages. The
automated method is scalable, and the jacketed reaction vessel allows reactions to
occur below or above ambient temperature.
The automated synthesizer greatly reduces the necessary labor required to access
large oligosaccharide structures. Once the set of building blocks required for the
synthesis of a class of oligosaccharides has been established, access to desired structures
may be readily achieved by automated synthesis.
This report details the initial efforts directed at the automated synthesis of
carbohydrates. Further improvements to the automated method, such as colorimetric
determination of coupling efficiencies and alternate methods to cleave from the
support not requiring a metal catalyst, are being investigated. Further extension of this
technology may allow the automated synthesis of complex glycoproteins, proteogly-
cans, and glycolipids on one instrument.
Experimental Part
General. All chemicals used were reagent grade and used as supplied except where noted. THF and CH2Cl2
used for washing cycles was purchased from Mallinckrodt (HPLC-Grade) and used without further purification.
CH2Cl2 used for reagent preparation was purchased from J. T. Baker (CycletainerTM) and passed through a
neutral alumina column prior to use. THF was purchased from J. T. Baker (CycletainerTM) and passed through a
neutral alumina column prior to use. Pyridine was refluxed over CaH2 and distilled prior to use. Trimethylsilyl
trifluoromethanesulfonate (TMSOTf) was purchased from Acros Chemicals. MeONa (25% (w/v) in MeOH),
glacial AcOH, hydrazine acetate, and hydrazine monohydrate (98%) were purchased from Aldrich Chemicals.
Anal. TLC performed on E. Merck silica-gel 60 F254 plates (0.25 mm). Compounds were visualized by dipping
the plates in a cerium sulfate-ammonium molybdate soln., followed by heating. Liquid column chromatography
(CC) was performed by forced flow of the indicated solvent on Silicycle 230 400 mesh (60-ä pore diameter)
silica gel. HPLC Analysis was performed on a Waters model 600E multisolvent delivery system with anal.
¾
¾
(Nova-Pak , 60ä, 4 mm, 3.6 Â 150mm) and prep. ( Nova-Pak , 60ä, 6 mm, 7.8 Â 300 mm) silica columns.
1H-NMR spectra were obtained on a Varian VXR-500 spectrometer (500 MHz), Varian VXR-300 (300 MHz),
or Bruker (400 MHz), and are reported in ppm (d) relative to CHCl3 (7.27 ppm). Coupling constants (J) are
reported in H. 13C-NMR spectra were obtained on a Varian VXR-500 spectrometer (125 MHz), Varian VXR-
300 (75 MHz), or Bruker (100 MHz), and are reported in d relative to CDCl3 (77.23 ppm). MALDI-TOF mass
spectrometry was performed on a PE Biosystems Voyager System 102 as follows: A 1-ml aliquot of matrix soln.
(10-mg/ml 2,5-dihydroxybenzoic acid (DHB) in THF) was spotted on the sample holder and allowed to dry.
Addition of a 1-ml aliquot of oligosaccharide soln. (5 mg/ml AcOEt) was co-spotted on the matrix, dried, and
analyzed in the positive-ion mode.
Dibutyl 3-O-Benzyl-4-O-levulinoyl-2-O-pivaloyl-b-d-xylopyranosyl Phosphate (23). 3-O-Benzyl-4-O-lev-
ulinoyl-1,5-anhydro-2-deoxy-d-threo-pent-1-enitol (22; 512 mg, 1.67 mmol) was co-evaporated with toluene
(3 Â 10ml), dissolved in CH 2Cl2 (10ml), and cooled to 0 8. Dimethyldioxirane (DMDO; 30 ml, 0.08m in
acetone) was added via cannula, and the mixture was stirred for 30min. Solvents were removed in vacuo, and
the residue was dissolved in CH2Cl2 (20ml). The resulting soln. was cooled to À788, dibutyl phosphate (320 ml,

Helvetica Chimica Acta Vol. 86 (2003)
3987
1.67 mmol) was added dropwise. After 15 min, the mixture was warmed to amb. temp., and pivaloyl chloride
(410 ml, 3.34 mmol) and 4-(dimethylamino)pyridine (DMAP; 815 mg, 6.68 mmol) were added. The resulting
mixture was stirred for 2 h, hexanes/AcOEt 3 :1 (100 ml) were added, and the white precipitate was filtered off
through a silica-gel plug. The filtrate solvents were removed in vacuo and by CC (hexanes/AcOEt 3 :1) afforded
23 (593 mg, 58%). [a]2D4 ˆ À2.68 (c ˆ 3.84, CH2Cl2); IR (thin film, NaCl): 2962, 2934, 1742, 1151. 1H-NMR
(300 MHz, CDCl3): 7.34 7.27 (m, 5 H); 5.42 (dd, J ˆ 4.4, 6.2, 1 H); 5.08 (app. t, J ˆ 6.2, 1 H); 4.95 4.93 (m,
1 H); 4.74 (d, J ˆ 11.6, 1 H); 4.67 (d, J ˆ 11.6, 1 H); 4.26 (dd, J ˆ 3.6, 12.4, 1 H); 4.09 3.99 (m, 4 H); 3.72 (app. t,
J ˆ 6.0, 1 H); 3.58 (dd, J ˆ 5.5, 12.4, 1 H); 2.77 2.72 (m, 2 H); 2.58 2.53 (m, 2 H); 2.20( s, 3 H); 1.68 1.60( m,
4 H); 1.44 1.26 (m, 4 H); 1.25 (s, 9 H); 0.98 0.90 (m, 6 H); 13C-NMR (125 MHz, CDCl3): 206.4; 177.1; 172.1;
137.8. 128.6; 128.5; 128.5; 127.9; 127.7; 127.6; 127.6; 96.2; 96.2; 75.9; 75.0; 73.3; 69.4; 69.3; 69.3; 68.1; 68.1; 68.0;
68.0; 61.3; 39.0; 38.0; 37.9; 37.9; 32.4; 32.4; 32.3; 32.3; 32.3; 30.0; 30.0; 28.0; 27.9; 27.3; 27.3; 27.3; 18.8; 18.8; 18.8;
‡
13.8; 13.8; 13.8. 31P-NMR (120MHz, CDCl 3): À2.5; ESI-MS: 637.2752 ([M ‡ Na] ; calc. 637.1856).
Dibutyl 2-O-Acetyl-4,6-di-O-benzyl-3-O-levulinoyl-b-d-galactopyranosyl Phosphate (26). 4,6-Di-O-benzyl-
3-O-levulinoyl-d-galactal (430mg, 1.01 mmol) was co-evaporated with toluene (3 Â 10ml), dissolved in
CH2Cl2, and the soln. was cooled to 08. DMDO (40 ml, 0.08m in acetone) was added, and the resulting mixture
was stirred for 30min. The solvents were removed in vacuo, the residue was dissolved in CH2Cl2 (20ml), and the
soln. was cooled to À788. Dibutyl phosphate (200 ml, 1.05 mmol) was added dropwise. After 30 min, the soln.
was warmed to amb. temp., and Ac2O (190mg, 2.0mmol) and DMAP (493 mg, 4.0mmol) were added, and the
soln. was stirred for 1 h. Hexanes/AcOEt 3 :1 (100 ml) and the mixture were passed through a silica-gel plug,
and the solvents were removed in vacuo. CC (hexanes/AcOEt 3 :2 ! 1:1) afforded 26 (375 mg, 65%). Clear oil.
[a]2D4 ˆ ‡18.38 (c ˆ 1.15, CH2Cl2). IR (thin film, NaCl): 2961, 2934, 1750, 1028. 1H-NMR (300 MHz, CDCl3):
7.30 7.26 ( m, 10H); 5.43 ( dd, J ˆ 8.0, 10.2, 1 H); 5.21 (app. t, J ˆ 7.4, 1 H); 4.98 (dd, J ˆ 3.0, 10.2, 1 H); 4.72 (d,
J ˆ 11.7, 1 H); 4.55 (d, J ˆ 11.7, 1 H); 4.15 3.96 (m, 5 H); 3.83 (app. t, J ˆ 6.6, 1 H); 3.63 3.58 (m, 2 H); 2.69
2.65 (m, 2 H); 2.46 2.65 (m, 2 H); 2.15 (s, 3 H); 2.09 (s, 3 H); 1.67 1.57 (m, 4 H); 1.41 1.27 (m, 4 H); 0 .94
0.86 (m, 6 H). 13C-NMR (75 MHz, CDCl3): 206.3; 172.0; 169.8; 138.2; 137.7; 128.6; 128.5; 128.2; 128.0; 128.0;
127.9; 97.0; 96.9; 75.3; 74.0; 73.9; 73.6; 69.8; 69.7; 68.2; 68.1; 68.1; 68.1; 67.6; 37.7; 32.3; 32.2; 32.2; 29.9; 28.0; 21.0;
‡
18.8; 13.8; 13.7. 31P-NMR (120MHz, CDCl 3): À2.4. ESI-MS: 715.2861 ([M ‡ Na] ; calc. 715.1962).
Pent-4-enyl (Methyl 3,4-di-O-benzyl-2-O-pivaloyl-b-d-glucopyranosiduronate)-(1 ! 3)-2-O-acetyl-4-6-di-
O-benzyl-b-d-galactopyranosyl-(1 ! 3)-2-O-acetyl-4-6-di-O-benzyl-b-d-galactopyranosyl-(1 ! 4)-3-O-benzyl-
2-O-pivaloyl-b-d-xylopyranoside (31). Oct-4-ene-1,8-diol-functionalized resin 3 (25 mmol, 61 mg; 0.40 mmol/g
loading) was manually loaded into a reaction vessel, and the vessel was inserted into a modified ABI-433A
peptide synthesizer. The coupling, deprotection, and washing steps were controlled by the synthesizer. The resin
was glycosylated with donor 23 (5 equiv., 77mg) delivered in CH2Cl2 (4 ml) and TMSOTf (5 equiv., 1 ml; 0.125m
in CH2Cl2) at À208. Mixing of the suspension was performed (10s vortex, 50s rest) for 15 min. The resin was
then washed with CH2Cl2 (4 Â 4 ml each) and glycosylated a second time. Removal of any soluble impurities was
accomplished by washing the resin with CH2Cl2 (4 Â 4 ml each) and THF (4 Â 4 ml each). Deprotection of the
levulinoyl ester was carried out by treating the resin with hydrazine acetate (40equiv., 4 ml, 0.25 m N2H4 ¥ H2O in
pyridine/AcOH 3 :2) for 15 min at 158. The resin was subjected to the deprotection conditions a second time for
15 min. Removal of any soluble impurities was accomplished by washing the resin with DMF (4 Â 4 ml), 0 .2m
AcOH in THF (4 Â 4 ml each), THF (4 Â 4 ml each), and CH2Cl2 (4 Â 4 ml each). Identical glycosylation/
deprotection cycles were used with glycosyl donors 26 (2 Â , 5 equiv., 86 mg, 5 equiv. TMSOTf) and 30 (5 equiv.,
83 mg, 0.5 equiv. TMSOTf). The glycosylated resin (25 mmol) was dried in vacuo over P2O5 for 12 h and
transferred to a 10ml flask. The flask was purged with ethylene and Grubbs× catalyst (bis(tricyclohexylphos-
phine)benzylideneruthenium(IV) dichloride; 4.1 mg, 20mol%) was added. The mixture was diluted with
CH2Cl2 (3 ml) and stirred under 1 atm ethylene for 36 h. The mixture was diluted with CH2Cl2 (5 ml) and
washed with H2O (3 Â 5 ml) and saturated aqueous NaHCO3 (3 Â 5 ml). The aq. phase was extracted with
CH2Cl2 (3 Â 5 ml), and the combined org. extracts were dried (Na2SO4), filtered through a plug of silica gel, and
the solvents were removed to afford crude 31 (6.1 mg). The resulting crude mixture was analyzed by HPLC
(Fig. 4). Mass spectrometry also confirmed the presence of 31 in the crude mixture. ESI-MS: 1682.0([ M ‡
NH4] ; calc. 1681.4). [a]2D4 ˆ ‡26.98 (c ˆ 0.16, CH2Cl2). 1H-NMR (500 MHz, CDCl3): 7.95 (d, J ˆ 7.2, 2 H); 7.92
‡
(d, J ˆ 7.3, 2 H); 7.83 (d, J ˆ 7.2, 2 H); 7.56 7.13 (m, 34 H); 5.94 (app. t, J ˆ 9.6, 1 H); 5.81 5.73 (m, 1 J); 5.68
(app. t, J ˆ 9.8, 1 H); 5.53 (dd, J ˆ 7.5, 9.3, 1 H); 5.31 (dd, J ˆ 8.1, 10.1, 1 H); 5.20 (dd, J ˆ 7.8, 10.2, 1 H); 5.16 (d,
J ˆ 11.9, 1 H); 5.01 4.85 (m, 7 H); 4.71 (d, J ˆ 11.8, 1 H); 4.55 4.14 (m, 10H); 3.91 3.53 ( m, 18 H); 3.17 3.14
‡
(m, 2 H); 2.11 2.03 (m, 2 H); 2.02 (s, 3 H); 1.48 (s, 3 H); 1.16 (s, 9 H). ESI-MS 1685.6492 ([M ‡ Na] ; calc.
1685.6506).

3988
Helvetica Chimica Acta Vol. 86 (2003)
Fig. 4. Analytical HPLC chromatogram of crude 31. Flow rate: 1 ml/min, 25 ! 40% AcOEt/Hexanes (30 min);
tR(31) 11.5 min.
¾
Automated Synthesis of Poly-a-(1 ! 2)-l-rhamnosides. Oct-4-ene-1,8-diol-functionalized Argopore resin 3
(0.100 mmol, 256 mg, 0.39 mmol/g) was manually loaded into a reaction vessel and inserted into a modified ABI-
433A peptide synthesizer. The coupling, deprotection, and washing steps were controlled by the synthesizer.
Double glycosylations were performed for 15 min each at À158, and double deprotections (40equiv. MeONa,
0.75m MeONa/MeOH) were carried out at 308 for 20min each. Elongation to the pentamer stage was followed
by cleavage from the resin. HPLC Analysis gave pentamer 43 (77%), tetramer 42 (18%), and trimer 41 (4%).
Pent-4-enyl 2-O-Acetyl-3,4-di-O-benzyl-a-l-rhamnopyranosyl-(1 ! 2)-3,4-di-O-benzyl-a-l-rhamnopyra-
nosyl-(1 ! 2)-3,4-di-O-benzyl-a-l-rhamnopyranoside (41). [a]2D4 ˆ À10.58 (c ˆ 0.81, CH2Cl2). IR (thin film):
2927, 1714, 1365, 1236, 1088. 1H-NMR (500 MHz, CDCl3): 7.52 7.12 (m, 30H); 5.81 5.76 ( m, 1 H); 5.56 5.54
(m, 1 H); 5.07 (app. s, 1 H); 5.02 4.95 (m, 3 H); 4.90( d, J ˆ 10.7, 1 H); 4.89 (d, J ˆ 11.0, 1 H); 4.86 (d, J ˆ 11.3,
1 H); 4.72 4.58 (m, 11 H); 4.54 (d, J ˆ 11.0, 1 H); 4.11 (app. s, 1 H); 3.98 (dd, J ˆ 3.4, 9.5, 1 H); 3.92 (app. s,
1 H); 3.89 (dd, J ˆ 3.2, 9.5, 1 H); 3.84 3.81 (m, 2 H); 3.77 3.74 (m, 1 H); 3.65 3.59 (m, 3 H); 3.47 3.31 (m,
5 H); 2.14 (s, 3 H); 2.11, 2.07 (m, 2 H); 1.65 1.59 (m, 2 H); 1.28 (d, J ˆ 6.7, 3 H); 1.27 (d, J ˆ 7.0, 3 H); 1.22 (d,
J ˆ 6.4, 3 H). 13C-NMR (125 MHz, CDCl3): 170.3; 138.7; 138.6; 138.5; 138.3; 138.2; 128.6; 128.5; 128.4; 128.3;
128.2; 128.0; 127.9; 127.8; 127.7; 115.1; 100.5; 99.3; 99.0; 80.5; 80.3; 79.8; 78.0; 75.6; 75.5; 74.8; 74.7; 72.3; 72.0;
‡
69.1; 68.6; 68.5; 68.0; 66.9; 30.5; 28.8; 21.4; 18.2; 18.1; ESI-MS: 1129.5231 ([ M ‡ Na] ; calc. 1129.5284).
Pent-4-enyl 2-O-Acetyl-3,4-di-O-benzyl-a-l-rhamnopyranosyl-(1 ! 2)-3,4-di-O-benzyl-a-l-rhamnopyra-
nosyl-(1 ! 2)-3,4-di-O-benzyl-a-l-rhamnopyranosyl-(1 ! 2)-3,4-di-O-benzyl-a-l-rhamnopyranoside (42).
[a]2D4 ˆ À9.08 (c ˆ 0.20, CH2Cl2). IR (thin film): 2927, 1714, 1366, 1234, 1088. 1H-NMR (400 MHz, CDCl3):
7.38 7.20( m, 40H); 5.89 5.77 ( m, 1 H); 5.61 (dd, J ˆ 1.8, 3.1, 1 H); 5.15 (app. s, 1 H); 5.07 4.88 (m, 8 H); 4.80
(d, J ˆ 11.0, 1 H); 4.76 4.58 (m, 12 H); 4.16 (d, J ˆ 2.0, 1 H); 4.10 (d, J ˆ 2.0, 1 H); 4.04 (dd, J ˆ 3.0, 9.4, 1 H);
3.96 3.74 (m, 7 H); 3.68 3.59 (m, 2 H); 3.51 3.33 (m, 6 H); 2.19 (s, 3 H); 2.16 2.08 (m, 2 H); 1.68 1.62 (m,
2 H); 1.32 1.23 (m, 12 H). 13C-NMR (100 MHz, CDCl3): 170.5; 138.9; 138.7; 138.6; 138.5; 128.8; 128.7; 128.6;
128.5; 128.4; 128.3; 128.1; 115.3; 101.0; 100.7; 99.5; 99.2; 79.6; 79.3; 78.2; 77.8; 75.9; 75.8; 75.2; 75.1; 75.0; 72.7;
‡
72.5; 72.4; 72.2; 69.3; 68.9; 68.7; 68.2; 67.2; 30.7; 29.0; 21.6; 18.4; 18.3. ESI MS: 1455.6810 ([M ‡ Na] ; calc.
1455.6802).
Pent-4-enyl 2-O-Acetyl-3,4-di-O-benzyl-a-l-rhamnopyranosyl-(1 ! 2)-3,4-di-O-benzyl-a-l-rhamnopyra-
nosyl-(1 ! 2)-3,4-di-O-benzyl-a-l-rhamnopyranosyl-(1 ! 2)-3,4-di-O-benzyl-a-l-rhamnopyranosyl-(1 ! 2)-
3,4-di-O-benzyl-a-l-rhamnopyranoside (43). [a]2D4 ˆ À11.78 (c ˆ 0.41, CH2Cl2). IR (thin film): 2931, 1744, 1453,
1364, 1087. 1H-NMR (400 MHz, CDCl3): 7.36 7.23 (m, 46 H); 7.21 7.15 (m, 4 H); 5.86 5.75 (m, 1 H); 5.61 (dd,

Helvetica Chimica Acta Vol. 86 (2003)
3989
J ˆ 1.8, 3.1, 1 H); 5.16 (d, J ˆ 1.3, 1 H); 5.09 (d, J ˆ 1.3, 1 H); 5.06 5.04 (m, 2 H); 5.01 4.86 (m, 7 H); 4.81
4.58 (m, 16 H); 4.17 (t, J ˆ 2.3, 1 H); 4.11 (t, J ˆ 2.1, 1 H); 4.06 4.02 (m, 2 H); 3.96 (dd, J ˆ 2.8, 9.3, 1 H); 3.93
3.72 (m, 9 H); 3.66 3.59 (m, 2 H); 3.51 3.30( m, 6 H); 2.18 (s, 3 H); 2.17 2.06 (m, 2 H); 1.67 1.60( m, 2 H);
1.34 1.21 (m, 15 H). 13C-NMR (100 MHz, CDCl3): 170.5; 138.9; 138.7; 138.5; 128.9; 128.8; 128.6; 128.5; 128.4;
128.3; 128.1; 128.0; 127.9; 115.3; 101.0; 100.9; 100.7; 99.5; 99.2; 80.9; 80.8; 80.7; 80.5; 79.8; 79.7; 79.4; 79.2; 78.2;
75.9; 75.8; 75.3; 75.2; 75.1; 75.0; 72.7; 72.5; 72.4; 72.2; 69.3; 68.9; 68.8; 68.7; 68.2; 67.1; 30.7; 29.0; 21.6; 18.4. ESI
‡
MS: 1781.8241 ([M ‡ Na] ; calc. 1781.8320).
Pent-4-enyl 3,4,6-Tri-O-benzyl-b-d-galactopyranosyl-(1 ! 4)-3,6-di-O-benzyl-2-deoxy-2-phthalimido-b-d-
glucopyranosyl-(1 ! 2)-3,4,6-tri-O-benzyl-a-d-mannopyranoside (45). Oct-4-ene-1,8-diol functionalized resin
3 (100 mmol, 333 mg, 0.30 mmol/g loading) was manually loaded into a reaction vessel and inserted into a
modified ABI-433A peptide synthesizer. The coupling, deprotection, and washing steps were controlled by the
synthesizer. The resin was glycosylated with donor 4 (4 equiv.) delivered in CH2Cl2 (4 ml) and TMSOTf
(0.4 equiv., in CH2Cl2). Mixing of the suspension was performed (10s vortex, 50s rest) for 30min. The resin was
then washed with CH2Cl2 (6 Â 4 ml each) and glycosylated a second time. Upon completion of the double
glycosylation, the resin was washed with CH2Cl2 (6 Â 4 ml each) MeOH/CH2Cl2 1:9 (4 Â 4 ml each).
Deprotection of the Ac ester was carried out by treatment of the resin with MeONa (5 equiv.; 0.5m MeONa
in MeOH) in CH2Cl2 (5 ml) for 30min. The resin was then washed with MeOH/CH 2Cl2 1:9 (1 Â 4 ml) and
subjected to the deprotection conditions a second time for 30min. Removal of any soluble impurities was
accomplished by washing the resin with MeOH/CH2Cl2 1:9 (4 Â 4 ml each), 0.2m AcOH in THF (4 Â 4 ml
each), THF (4 Â 4 ml each), and CH2Cl2 (6 Â 4 ml each). The resulting resin-bound C(2)-OH was glycosylated
with donor 44 (4 equiv.) delivered in CH2Cl2 (4 ml) and TMSOTf (0.4 equiv., in CH2Cl2) at À158. Mixing of the
suspension was performed (10s vortex, 50s rest) for 30min. The resin was then washed with CH 2Cl2 (6 Â 4 ml
each) and glycosylated a second time. Deprotection of the levulinoyl ester was carried out by treating the resin
with hydrazine acetate (40equiv., 4 ml; 0.25M N 2H4 ¥ HOAc in pyridine/AcOH 3 :2) for 15 min at 158. The resin
was subjected to the deprotection conditions a second time for 15 min. Removal of any soluble impurities was
accomplished by washing the resin with pyridine/AcOH 3 :2 (3 Â 4 ml), 0 .2m AcOH in THF (4 Â 4 ml each),
THF (4 Â 4 ml each), and CH2Cl2 (6 Â 4 ml each). The C(4)-OH was glycosylated using donor 35 (4 equiv.)
delivered in CH2Cl2 (4 ml) and TMSOTf (4 equiv., in CH2Cl2) at À158. Mixing of the suspension was perfomed
(10s vortex, 50s rest) for 15 min. The resin was then washed with CH 2Cl2 (6 Â 4 ml each) and glycosylated a
second time. Upon completion of the double glycosylation the resin was washed with CH2Cl2 (4 Â 4 ml each) and
THF (4 Â 4 ml). The glycosylated resin was dried in vacuo over P2O5 for 12 h and transferred to a 10-ml flask.
The flask was purged with ethylene and the Grubbs catalyst (bis(tricyclohexylphosphine)benzylidene
ruthenium(IV) dichloride; 16.4 mg, 20mol-%) was added. The mixture was diluted with CH 2Cl2 (6 ml) and
stirred under 1 atm ethylene for 36 h. Et3N (440ml, 3.20mmol, 80equiv.) and tris(hydroxymethyl)phosphine
(200 mg, 1.60 mmol, 80 equiv.) were added, and the resulting soln. was stirred at r.t. for 1 h. The pale yellow
mixture was diluted with CH2Cl2 (25 ml) and washed with H2O (3 Â 25 ml), sat. aq. NaHCO3 soln. (3 Â 25 ml),
and brine (3 Â 25 ml). The aq. phase was extracted with CH2Cl2 (3 Â 25 ml), and the combined org. extracts
were dried (Na2SO4), filtered, and concentrated. Purification by FC (20% AcOEt/hexanes) gave 13 (65 mg,
43% isolated yield). IR (thin film, NaCl): 2918, 2869, 1776, 1739, 1714, 1087. 1H-NMR (500 MHz, CDCl3): 7.65
(br. s, 1 H); 7.57 7.54 (m, 1 H); 7.37 7.10( m, 40H); 6.94 ( d, J ˆ 7.8, 1 H); 6.86 (t, J ˆ 7.3, 1 H); 6.78 (t, J ˆ 6.9,
1 H); 5.78 5.63 (m, 1 H); 4.97 4.64 (m, 6 H); 4.54 4.23 (m, 8 H); 4.13 3.93 (m, 4 H); 3.84 3.75 (m, 3 H);
3.57 3.39 (m, 6 H); 3.21 (app. dd, J ˆ 7.0, 16.3, 1 H); 2.96 (dd, J ˆ 6.4, 11.3, 1 H); 2.01 1.93 (m, 2 H); 1.56 1.50
(m, 2 H); 1.19 (s, 9 H). 13C-NMR (125 MHz, CDCl3): 176.9; 139.0; 138.9; 138.7; 138.7; 138.3; 138.0; 133.7; 132.0;
128.7; 128.6; 128.6; 128.4; 128.4; 128.4; 128.3; 128.3; 128.2; 128.1; 128.0; 127.9; 127.8; 127.6; 127.6; 127.5; 127.5;
127.3; 126.9; 123.3; 114.9; 100.1; 97.2; 97.1; 81.1; 77.9; 77.5; 76.7; 76.6; 75.4; 75.0; 74.9; 74.5; 73.9; 73.7; 73.6; 73.5;
‡
73.1; 72.8; 72.1; 71.9; 71.9; 70.6; 70.2; 69.0; 68.5; 67.2; 55.8; 39.0; 30.4; 28.8; 27.5. ESI-MS: 1528.6715 ([M ‡ Na] ;
calc. 1529.7562).
REFERENCES
[1] K. M. Koeller, C.-H. Wong, Chem. Rev. 2000, 100, 4465.
[2] S. Raghavan, D. S. Kahne, J. Am. Chem. Soc. 1993, 115, 1580.
[3] T. Zhu, G.-J. Boons, Angew. Chem., Int. Ed. 1998, 37, 1898.
[4] Z. Zhang, I. R. Ollman, X.-S. Ye, R. Wischnat, T. Baasov, C.-H. Wong, J. Org. Chem. 2000, 65, 2410.
[5] F. Burkhart, Z. Zhang, S. Wacowich-Sgarbi, C.-H. Wong, Angew. Chem., Int. Ed. 2001, 40, 1274.

3990
Helvetica Chimica Acta Vol. 86 (2003)
[6] P. H. Seeberger, −Solid Support Oligosaccharide Synthesis and Combinatorial Libraries×, John Wiley &
Sons, New York, 2001.
[7] B. Merrifield, Meth. Enzymol. 1997, 289, 3.
[8] B. Merrifield, J. Am. Chem. Soc. 1963, 85, 2149.
[9] P. H. Seeberger, S. J. Danishefsky, Acc. Chem. Res. 1998, 31, 685.
[10] J. Rademann, R. R. Schmidt, Carbohydr. Res. 1995, 269, 217.
[11] F. Roussel, M. Takhi, R. R. Schmidt, J. Org. Chem. 2001, 66, 8540.
[12] E. R. Palmacci, O. J. Plante, P. H. Seeberger, Eur. J. Org. Chem. 2002, 595.
[13] R. B. Andrade, O. J. Plante, L. G. Melean, P. H. Seeberger, Org. Lett. 1999, 1, 1811.
[14] W. C. Haase, P. H. Seeberger, Curr. Org. Chem. 2000, 4, 481.
[15] R. H. Grubbs, Adv. Synth. Catal. 2002, 344, 569.
[16] S. P. Douglas, D. M. Whitefield, J. J. Krepinsky, J. Am. Chem. Soc. 1995, 117, 2116.
[17] M. G. Hahn, A. G. Darvill, P. Albersheim, Plant Physiol. 1981, 68, 1161.
[18] O. J. Plante, E. R. Palmacci, P. H. Seeberger, Science 2001, 291, 1523.
[19] T. Takahashi, A. Okano, T. Amaya, H. Tanaka, T. Doi, Synlett 2002, 911.
[20] H. Tanaka, M. Adachi, H. Tsukamoto, T. Ikeda, H. Yamada, T. Takahashi, Org. Lett. 2002, 4, 4213.
[21] K. C. Nicolaou, N. Winssinger, J. Pastor, F. DeRoose, J. Am. Chem. Soc. 1997, 119, 449.
[22] T. Mukaiyama, K. Ikegai, T. Hashihayata, K. Kiyota, H. Jona, Chem. Lett. 2002, 730.
[23] R. Geurtsen, F. Cote, M. G. Hahn, G. J. Boons, J. Org. Chem. 1999, 64, 7828.
[24] O. J. Plante, R. B. Andrade, P. H. Seeberger, Org. Lett. 1999, 1, 211.
[25] L. C. Pedersen, K. Tsuchida, H. Kitagawa, K. Sugahara, T. A. Darden, M. Negishi, J. Biol. Chem. 2000, 275,
34580.
[26] K. Prydz, K. T. Dalen, J. Cell Sci. 2000, 113, 193.
[27] K. Sugahara, H. Kitagawa, Curr. Opin. Struct. Biol. 2000, 10, 518.
[28] J. Tamura, A. Yamaguchi, J. Tanaka, Bioorg. Med. Chem. Lett. 2002, 12, 1901.
[29] J. Tamura, Y. Miura, H. H. Freeze, J. Carbohydr. Chem. 1999, 18, 1.
[30] J. G. Allen, B. Fraser-Reid, J. Am. Chem. Soc. 1999, 121, 468.
[31] J. Tamura, K. W. Neumann, T. Ogawa, Liebigs Ann. Chem. 1996, 1239.
[32] L. Q. Chen, F. Z. Kong, Carbohydr. Res. 2002, 337, 1373.
[33] A. K. Sarkar, J. D. Esko, Carbohydr. Res. 1995, 279, 161.
[34] F. Weygand, −Methods in Carbohydrate Chemistry×, Academic Press: New York, 1962; Vol. 1.
[35] H. B. Mereyala, V. R. Kulkarni, Carbohydr. Res. 1989, 187, 154.
[36] S. David, S. Hanessian, Tetrahedron 1985, 41, 643.
[37] T. B. Grindley, −Applications of Tin-Containing Intermediates to Carbohydrate Chemistry×, Academic
Press, New York, 1998, Vol. 53.
[38] K. R. Love, R. B. Andrade, P. H. Seeberger, J. Org. Chem. 2001, 66, 8165.
[39] J. G. Allen, B. Fraser-Reid, J. Am. Chem. Soc. 1999, 121, 468.
[40] K. Ruda, J. Lindberg, P. J. Garegg, S. Oscarson, P. Kondradsson, J. Am. Chem. Soc. 2000, 122, 11067.
[41] A. Arasappan, B. Fraser-Reid, J. Org. Chem. 1996, 61, 2401.
[42] V. A. Nikolaev, T. J. Rutherford, M. A. Ferguson, J. S. Brimacombe, J. Chem. Soc. Perkin Trans. 1 1996,
1559.
[43] B. L. Herwaldt, Lancet 1999, 354, 1191.
[44] M. C. Hewitt, P. H. Seeberger, J. Org. Chem. 2001, 66, 4233.
[45] T. Muller, A. F¸rstner, J. Am. Chem. Soc. 1999, 121, 7814.
[46] D. Crich, J. Piccione, Org. Lett. 2003, 5, 781.
[47] A. Shilatifard, R. K. Merkle, D. E. Helland, J. L. Welles, W. A. Haseltine, R. D. Cummings, J. Virol. 1993,
67, 943.
[48] H. Lˆnn, Carbohydr. Res. 1985, 139, 105.
[49] P. P. Desphande, S. J. Danishefsky, Nature 1997, 387, 164.
Received September 2, 2003