Developing high affinity oligosaccharide inhibitors: Conformational pre-organization paired with functional group modification

Article abstract of DOI:10.1039/b416106h

Intramolecular tethering combined with functional group modification has been investigated as an approach to design high affinity oligosaccharide ligands. The preceding paper reported successful tethering to constrain a trisaccharide in the conformation o

Full text of DOI:10.1039/b416106h

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Developing high affinity oligosaccharide inhibitors: conformational
pre-organization paired with functional group modification
Robert S. McGavin and David R. Bundle*
Alberta Ingenuity Centre for Carbohydrate Science, Department of Chemistry, University of
Alberta, Edmonton, Alberta T6G 2G2. E-mail: Dave.Bundle@ualberta.ca
Received 18th October 2004, Accepted 9th May 2005
First published as an Advance Article on the web 10th June 2005
Intramolecular tethering combined with functional group modification has been investigated as an approach to
design high affinity oligosaccharide ligands. The preceding paper reported successful tethering to constrain a
trisaccharide in the conformation of its bound state with an antibody and thereby achieved a 15-fold increase in
association constant. Here we report the synthesis of two b-alanyl tethered derivatives that employ monochlorination
and monodeoxygenation strategies to create inhibitors that should enhance the binding affinity of the target
molecules by an additional 10–25-fold, provided that free energy changes are additive when tethering is paired with
functional group changes. The binding parameters of the new ligands 5 and 6 were measured by isothermal titration
calorimetry and the results rationalized with molecular dynamics calculations and a simple docking analysis. The
data indicate that while the alanine tether is a reasonable method to constrain trisaccharide 1, free energy gains
obtained by pairing it with functional group modification are not additive and in one case counter-productive.
Introduction
If carbohydrates are to find uses as therapeutics and practical
affinity reagents, the weak attraction that dominates most
oligosaccharide–protein interactions must be overcome. Efforts
to design high-affinity carbohydrate epitopes have so far met
with relatively few successes and, of these, free energy gains are
frequently modest.
In the preceding paper we reported our attempts to tether a
trisaccharide in its bound conformational state with an antibody.
The active tethered trisaccharide exhibited a 15-fold affinity
gain when comapred with the native untethered trisaccharide
1. The complex of 1 with the monoclonal antibody SYA/J6
has been extensively mapped by functional group modification
with the introduction of either monodeoxy or monochlorodeoxy
Fig. 1 Native trisaccharide 1 and previously synthesized derivatives
thereof, including 2 and 3 containing modifications at the rhamnose C
functions, creating ligands that exhibited 10- and 20-fold gains
in affinity.1 It was therefore of interest to investigate tethering
of these derivatives and attempt to create significantly higher
avidity ligands (possibly as high as 300-fold) through a combi-
nation of tethering and functional group changes, each of which
individually yield affinity gains.
The monoclonal IgG antibody (mAb) SYA/J6 binds the
linear tetrasaccharide repeating unit of the O-polysaccharide
of Shigella flexneri variant Y:2
residue, along with tethered 4.
also confirmed that indeed the BCD residues provide the major
contribution to binding energy through polar and hydrophobic
contacts. The Rha C residue was almost entirely buried in
the deepest area of the combining site. The modes of binding
between the ABCDAꢀ pentasaccharide and 3 are similar, but
differ in a few key aspects. For the 3-SYA/J6 complex, it
was noted that the 2-deoxy-Rha C residue sank even deeper
in the binding groove and was accompanied by a change in
the ␾ and w angles of the CD glycosidic bond that allowed
the GlcNAc-D ring of 3 to make more polar contacts than
pentasaccharide (1). The deoxygenated center also extends a
hydrophobic area originating at the methyl group of the Rha B
residue and the methyl moiety of the 2-acetamido group of the D
residue, including Tyr L32 and His L27 of the protein. Although
trisaccharide 2 has not been crystallized with SYA/J6, it has
been proposed that its conformation in the bound state should
be the same or similar to the conformation of the BCD portion
of the ABCDAꢀ pentasaccharide and would not penetrate the
binding site as deeply as the monodeoxy congener 3 since the
size and electronegativity of chlorine and hydroxyl moieties are
similar.7
[→ 2)-a-L-Rhap(1 → 2)-a-L-Rhap(1 → 3)-a-L-
A
B
Rhap(1 → 3)-b-D-GlcNAcp(1 →]n
C
D
Chemical mapping determined the reading frame of the binding
groove and identified a trisaccharide BCD as the optimal
binding trisaccharide epitope (1). During these studies, it was
found that deoxy-chloro (2)3 and deoxy-(3)4 congeners (Fig. 1)
at the 2-position of the rhamnose C residue produced tighter
binding trisaccharide derivatives with free energy gains (DDG
that ranged between approximately −1.4 to −1.7 kcal mol−1),
relative to the native trisaccahride 1.
Detailed structures for the natural ABCDAꢀ pentasaccharide
and trisaccharide 3 (Fig. 1) in the binding site have been obtained
from solved crystal structures and complement the data from
epitope mapping by functional group replacement.5,6 The crystal
structures revealed that the binding site is a deep groove, capable
of binding internal residues of the natural O-polysaccharide. It
Recent efforts to tailor the BCD epitope for high affinity
binding have focused on intramolecular tethering to pre-
organize the ligand in its bioactive conformation.1 It has been
postulated that immobilizing rotation upon binding to protein
can incur a penalty of 0.5 to 2.0 kcal mol−1 for each glycosidic
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 7 3 3 – 2 7 4 0
2 7 3 3
©

View Article Online
bond.8,9 Therefore, conformational restriction could lead to
substantial favourable increases in entropic binding energy. This
methodology has been used previously on several carbohydrate
systems and resulted in both moderate gains and losses in in
binding energy.10 Crystal structure data for the SYA/J6 complex
with 1 shows that the two methyl groups of residue D and residue
B face away from the protein surface toward bulk solvent. This
suggested that both sites are suitable for tethering that could
provide a favourable entropic contribution for binding within
the range 1–4 kcal mol−1, provided that the tether preorganizes
the trisaccharide in the bound conformation while it avoids
unfavourable contacts with protein. We showed in the preceding
paper that BCD trisaccharide derivative (4) containing a b-
alanine (3-aminopropanoic acid) tether spanning these moieties
effectively achieves this pre-organization, with an approximate
1.5 kcal mol−1 gain in free energy compared to 1 (Fig. 1).8
However, we also concluded that the free energy gains from
solvent reorganization at the periphery of the complex were
of equal or greater significance to the gains in conformational
entropy.
glycosylation or tethering reactions. Given the relative ease
of constructing 1,2-trans-linkages, the L-mannosyl donor was
protected as a 2,3,4-tri-O-benzyl-6-O-silyl derivative 12. The
small amount of b-anomer that was likley to be formed as
a consequence of the use of a non-partcipating group at C-2
was deemed acceptable so that the orthogonality of the benzyl
protecting groups could be preserved and reduce the number
of protecting group manipulations at advanced stages of the
synthesis.
Installation of the chlorine atom at the 2-position of 1,5-
anhydro-3-O-benzoyl-4-O-benzyl-2,6-dideoxy-hex-1-enitol (7)3
to afford the rhamno-configuration has previously been ac-
complished by chloromethoxylation, but suffers from poor
selectivities, difficult purification and, consequently, low yield
(30–35%).3,11 A modification of the Igarashi et al. protocol12 led
to highly diastereofacial cis-addition of chlorine gas to 9 in CCl4,
followed by the Koenig–Knorr glycosylation of acetic acid13 to
afford 8 in 88% overall yield in a ratio of 4 : 1 rhamno : quinovo
(6-deoxy-L-gluco) configurations (Scheme 1). It should be noted
that the use of 1,2-propanediol cyclic carbonate as a solvent
for the chlorination12 leads to the opposite diastereofacial
selectivity; a 6 : 1 mixture favoring the b-quinovosyl diastereomer
of 8 in good yield. The stereochemistry of the 2-chloro-2-
deoxy derivatives 8 and 9 were established by observation of
3JH–H coupling constants. Lewis acid catalyzed thioglycosidation
with ethanethiol produced donor 9 in excellent yield. The L-
mannosyl thioglycoside donor 12 was synthesized from the
known thiophenyl mannoside 1014 via regioselective silylation
and exhaustive benzylation, both in high yield (Scheme 1).
Here we test the hypothesis that the combination of teth-
ering with specific functional group replacement could yield
substantial affinity enhencement (Fig. 2). If free energy gains for
pairing of individual modifications (Table 1, entries 2, 3 and 4)
proved to be additive, the concept of combining intramolecular
pre-organization with functional group modifications might
become a useful strategy for the preparation of high-affinity
carbohydrate inhibitors.
Fig. 2 Targets of this work, depicting the b-alanyl tether of 4 and
functional group modifications at the rhamnose C residue.
Scheme 1 Reagents and Conditions: (a) Cl2, CCl4, −10 ◦C; (b) AgOAc,
AcOH, 70%, two steps; (c) EtSH, BF3–OEt2, CH2Cl2, 92%;
(d) TBDMS–Cl, pyridine, Et3N, 91%; (e) BnBr, NaH, DMF, 92%.
Results and discussion
Synthesis of cyclic trisaccharides 5 and 6
The synthesis of 5 employed the same strategy as reported
for 4,1 using standard thioglycoside methodology. It has been
shown that 2-chloro-2-deoxy thioglycosides function well as
thioglycoside donors and also as acceptors, since glycosylation
of the 3-position of a terminal rhamnose residue containing
a 2-chloro-2-deoxy functionality proceeded in good yield.3
Base labile protecting groups were employed at those hydroxyl
and amino groups which were the sites of glycosylation and
modification. Accordingly this required the deployment of
orthogonal protecting groups, such as benzyl ethers and a
benzylidene acetal, for those hydroxyl groups not involved in
Glycosylation of acceptor 1315 using 9, NIS and AgOTf
proceeded smoothly and rapidly to afford protected disaccharide
14 (92%) with no detectable trace of the b-anomer (Scheme 2).
The anomeric configuration of the newly formed rhamnosyl
linkage was determined by heteronuclear correlation spectr-
socopy, HMQC (1JC1–H1 = 173.2 Hz).16 Transesterification of
the 3ꢀ-O-benzoate gave disaccharide acceptor 15, which was
glycosylated by 12 in good yield. Initial attempts using NIS–
AgOTf to promote glycosylation gave very little trisaccharide
product, but use of NIS–TfOH as promotor furnished 16 in
86% yield. HMQC experiments confirmed the assignment of
Table 1 Isothermal microcalorimetric data for native trisaccharide 1 and derivatives 2–6
DG/Kcal mol−1
DH/Kcal mol−1
−TDS/Kcal mol−1
−1
Compound
KA/mol
110
2
1.1 × 105
−6.8 0.2
−8.1
−3.9 0.1
−6.3
−2.9 0.1
−1.8
1.1 × 106
3
2.5 × 106
−8.5
−8.1
−0.5
4
1.5 0.05 × 106
−8.3 0.1
−8.6 0.15
−7.4
−4.2 0.05
−4.4 0.1
−5.6
−4.1 0.05
−4.2 0.05
−1.8
5
2.6 × 106
6
4.3 × 105
2 7 3 4
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 7 3 3 – 2 7 4 0

View Article Online
˚
Scheme 2 (a) 6, NIS, AgOTf, CH2Cl2, 3 A MS, 92%; (b) Na, MeOH, CH2Cl2, 81%; (c) 8, NIS, TfOH (cat.), CH2Cl2, 86%; (d) (i) NH2NH2–H2O,
EtOH; (ii) TBAF, THF; (iii) Fmoc-b-ala-OH, TBTU, HOBt, NEM, DMF, 68%, 3 steps; (e) (i) TEMPO, KBr, NaHCO3, NaOCl, MeCN (pH 10.5
with NaOH); (ii) 20% piperidine–DMF; (iii) PYBOP, DMF, 61%, 3 steps; (f) H2, Pd(OH)2, MeOH, H2O, 79%; (g) n-Bu3SnH, AIBN, THF, dioxane,
90%.
the linkage as a (1JC1–H1 = 169.8 Hz) and, as expected for the 2-
O-benzyl-mannosyl donor, approximately 5% of the undesired
b-anomer was formed.
residue coupling for both H-1ꢀ and H-1ꢀꢀ across the glycosidic
bonds. While these extended coupling pathways have very small
values and are only observed in two dimensional spectra of a-
linked sugars, they were not observed for the less restricted open
chain molecules described here or in the preceding paper.
Initially, it was intended to acetylate 5, effect reduction of the
chloride with tri-n-butyltin hydride (TBTH)20 and deacetylate to
get cyclic deoxy 6. Standard acetylation conditions (Ac2O, Pyr.)
produced a mixture of hexa- and hepta-acylated compounds.
Although it is rarely observed that a lactam nitrogen is acetylated
under these conditions it is the only available site for over-
acetylation.21 Reduction of this mixture with TBTH–AIBN in
toluene, followed by Ze´mplen transesterification produced poor
yields of 6 along with recovered 5 after HPLC. Direct reduction
of the 2ꢀ-chloro substituent of 5 was facilitated with TBTH–
AIBN in 3 : 1 dioxane : THF and produced 6 in good yield with
little or no detectable 5 (Scheme 2).
The transformation of protected trisaccharide 16 to 17 was
accomplished by using NH2NH2–H2O to remove the phthal-
imido group, followed by TBAF to remove the 6ꢀꢀ-O-TBDMS
ether. The amino-alcohol was treated with N-Fmoc-b-alanine-
OH under peptide coupling conditions (TBTU–HOBt) to afford
17 in 68% over 3 steps. This alcohol was oxidized using TEMPO–
NaOCl17 and then filtered through a short pad of silica-gel and
treated under standard Fmoc deprotection conditions. Initial
attempts at macro-cyclization using the above stated peptide-
coupling conditions resulted in a large amount of unreacted
activated ester. Employing a single component coupling system
of PyBOP in DMF18 at high dilution (1 mg mL−1) afforded
the protected lactam 18 (91% for the cyclization from the acid,
61% from 17), which was hydrogenated over Pd(OH)2 to afford
deprotected 5 (79% yield).
The one and two dimensional proton NMR spectra of
protected cyclic trisaccharide 18 indicated that tethering had
restricted the solution conformation of the molecule. Reso-
nances corresponding to the methylene protons of the tether
had sharpened into diastereotopic signals with defined line
structure. As expected, this fine structure was not present for
acyclic precursor 17, as rotamer conformations about the amide
linkage averaged and broadened a number of signals in the
Native trisaccharide 1 was synthesized based on a published
procedure.22
Binding with SYA/J6 and conformation
Microcalorimetry23,24 was carried out using HPLC-purified
samples of trisaccharides 1 to 6. Solutions of ligands (0.50 mM)
were titrated against purified mAb SYA/J6 (30 lM, all of which
were obtained from a single preparation of protein). Entries that
report standard deviations were run in duplicate and on separate
batches of purified protein.
spectrum. The gradient-enhanced COSY spectra (GCOSY) of
5
18 exhibited long-range intra-residue coupling (up to JH–H
)
The cyclic 2ꢀ-chloro tethered trisaccharide 5 exhibited the
highest affinity of all trisaccharide ligands (KA = 2.6 × 106),
a value that is esentially indistinguisable from the activity of
within the rhamno- and manno-ring systems. The deprotected
compounds 5 and 6 also exhibited this long-range scalar
coupling.19 The protected derivative 18 also exhibited 4JH–H inter-
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 7 3 3 – 2 7 4 0
2 7 3 5

View Article Online
3 and 4. However the orgins of the free energy are quite different.
Enthalpy and entropy are both favourable and contribute
equally to the binding free energy of 5 and 4. This suggests that
functional group modification and tethering both contribute
to the observed activity but the effects are in no way additive.
Definitive proof for the conclusion that intramolecular pre-
organization was not the only cause for affinity increases cannot
be derived from thermodynamics since entropy–enthalpy com-
pensation effects are always observed and mask the molecular
basis for the observed activities.25,26 Surprisingly, cyclic 2ꢀ-deoxy
6 exhibited decreased affinity and bound with lower free energy
than either cyclic 4 or acyclic deoxy 3. When the data for 6 and
3 are compared a substantially reduced enthalpy of interaction
occurs when 6 is bound. When compared to the binding of
the tethered trisaccharide 4 comparable enthalpy it is seen but
binding occurs with less favourable entropy. This is consistent
with fewer attractive interactions between the protein and the
tethered 2-deoxy compound 6, which could arise if the tether
prevents the deep penetration proposed to account for the
activity of 3 (Table 1).
We sought to explain the binding data, especially with respect
to 6, by focusing our attention on modeling the solution
conformation, dynamics and performing a docking analysis.
Calculations were performed using Insight II with the AMBER
Plus force field27 containing an imported parameter designed
to account for L-sugars in the 1C4 conformation.28 Parallel
calculations were performed on the disaccharides a-L-Rha-p-
(1→3)-a-L-Rha-p-OMe and a-L-Rha-p-(1→3)-b-D-GlcNAc-p-
OMe using the HSEA-based GeGOP29 and Insight II. Both
programs gave virtually identical results, encouraging us to use
the Insight software for the remainder of our dynamics and
conformational calculations. The only limitation of this force
field is the inability to properly assign atom potentials for a
covalent carbon–chlorine bond. Therefore, calculations could
not be performed for compound 5.
Fig. 4 Superimposition of the global energy minimum of 6 upon the
bound conformation of 3 (taken from the co-complexed crystal structure
of 3 and mAb SYA/J6).
more flexible bond was defined by the dihedral H-5ꢀꢀ–C-5ꢀꢀ–C-
6ꢀꢀ–O-6ꢀꢀ. Calculations indicated that this bond sampled two
conformations with the minimum energy structure being the
preferred conformation approximately 90% of the time.
Docking analysis of SYA/J6 was performed by superimpos-
ing the global minimum conformer of 6 on the 3-SYA/J6 co-
complex (Fig 5). If tethered 6 binds in the same mode as 3,
there does not appear to be any evident steric clash between the
tether and the protein surface, although closer inspection shows
that the nitrogen from the amide used for tethering at the L-
mannosyl residue points directly towards a hydophpobic area of
the protein surface, which is extended by deoxygenation of the
2ꢀ-position of the trisaccharide. It may also be the case that our
paired modifications have induced a shift of the trisaccharide in
the binding site that induces a significant clash of the tether and
the protein.
The dynamics calculations of 6 indicate highly restricted
conformations around the glycosidic bonds that are almost
identical to those of the solution state of the native trisaccharide
epitope (Fig. 3). Superimposition of the calculated global
minimum energy conformation of 6 on the structure of 3,
extracted from the co-crystallized complex of SYA/J6 with 3,
shows that the pre-organized conformation of 6 mimics the
bound conformation of 3 (Fig. 4). By way of contrast, a slightly
Fig. 5 Depiction of 6 docked on the bound structure of 3 in the
antibody-binding site. No apparent clash of the ligand is visible with
the protein surface.
Since we could not assign the atom potentials for 5 due to the
lack of support for the carbon–chlorine covalent bond within the
AMBER Plus force field, we can only hypothesize why it bound
with the highest affinity. We have postulated that the binding
of 2 would likely be the same as 1, since the substitution of a
hydroxyl for a chlorine is a conservative substitution in terms of
size. Modeling indicates that tethered trisaccharide 4 can bind
in the same mode as 1. However, 5 binds with an affinity almost
identical to that of 4, which shows that the free energy changes
are not additive. While we have succesfully used intramolecular
tethering to gain affinity, when this modification is combined
with functional group replacement only a very small increase in
affinity is observed.
Conclusions
Fig. 3 ␾ vs. w plots for 1 (top) and 6 (bottom) derived from molecular
dynamics simulations. Subscripts BC and CD refer to the glycosidic
bonds between the B, C and D rings of the trisaccharides.
Analysis of the MD calculations, docking and microcalorimetry
data lead us to conclude that a more thorough analysis and
2 7 3 6
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 7 3 3 – 2 7 4 0

View Article Online
a-rhamno- and b-quinovo-isomers): [a]2D2 +1.98 (c 1.0, CHCl3);
1H NMR (600 MHz, CDCl3) d 8.09–7.19 (m, 10H, Ar–H), 6.19
(d, 1H, J = 2.0 Hz, H-1), 5.61 (dd, 1H, J = 3.8, 9.1 Hz, H-3),
4.62 (ABq, 2H, J = 11 Hz, PhCH2O–), 4.52 (dd, 1H, J = 2.0,
3.8 Hz, H-2), 3.99 (dq, 1H, J = 6.2, 9.5 Hz, H-5), 3.98 (dd ≈ t,
1H, J = 9.3 Hz, H-4), 2.14 (s, 3H, –C(O)CH3), 1.37 (d, 3H, J =
6.2 Hz, H-6); 13C NMR (125 MHz, CDCl3) d 168.6, 165.6, 137.4,
133.5, 129.8, 129.4, 128.6, 128.0, 93.1 (1JC1–H1 = 180.19 Hz), 77.9,
75.5, 72.3, 70.5, 57.5, 20.9, 18.0; anal. calcd for C22H23ClO6: C,
63.08; H, 5.53%, found C, 63.06; H, 5.57%. ES HRMS calcd for
C22H23ClO6Na (M + Na): 441.1080. Found 441.1072.
detailed computational model are required to investigate the
optimum positions of the atoms of the tether for compounds
4 and 5 and their compatibility with the adjacent protein
surface. We are investigating the binding of compounds 5 and
6 using the technique of STD NMR.30 This technique reveals
those ligand groups that make the most intimate contact with
protein. This approach, together with MD simulations for the
different trisaccharide complexes with SYA/J6, should be able to
correlate modes of binding between our ligands. If this technique
permits us to identify the ligand–protein contacts for the pairs
2 with 5 and 3 with 6, we should be able to establish the orgins
for the absence of additive free energies and the disappointing
affinites of 5 and 6.
At present, the most likely explanation for the absence of
strong additive free energy gains is the inherent flexibility of
glycosidic linkages. In untethered oligosaccharides this could
permit energy gains for single site modifications to be obtained
by segmental movements that are permissible and create higher
complimentary, while in the tethered derivatives such mutually
complementary movements are constrained by the presence
of a tether. If this is the case, computional approaches that
simulate and identify ligand intra-residue movements that result
in enhanced activity offer the best hope to achieve high affinity
carbohydrate ligands.
Ethyl 3-O-benzoyl-4-O-benzyl-2-chloro-2-deoxy-1-thio-a-L-
rhamnopyranoside (9). To a solution of 8 (3.31g, 7.90 mmol),
˚
3 A molecular sieves in EtSH (0.76 mL, 10.3 mmol) and distilled
CH2Cl2 (50 mL) under Ar at 0 ◦C was added BF3–OEt2 (1.29 mL,
10.3 mmol). The reaction stirred for 2 h before the addition
of Et3N (4 mL). The solution was filtered and washed with
NaHCO3 and water. The organic layer was dried with Na2SO4
and concentrated under a vacuum. Column chromatography (9 :
1 hexanes : ethyl acetate) gave a clear syrup (3.04 g, 91%): [a]D22
−62.3 (c 2.56, CHCl3); 1H NMR (600 MHz, CDCl3) d 8.06–7.20
(m, 10 H, Ar–H), 5.53 (dd, 1H, J = 3.7, 9.2 Hz, H-3), 5.37 (s,
1H, H-1), 4.71 (dd, 1H, J = 1.2, 3.6 Hz, H-2), 4.69 (Abq, 2H,
J = 10.9 Hz, PhCH2O–), 4.26 (dq, 1H, J = 6.3, 9.4 Hz, H-5),
3.80 (dd ≈ t, 1H, J = 9.3 Hz, H-4), 2.64 (m, 2H, SCH2CH3),
1.37 (d, 3H, J = 6.3 Hz, H-6), 1.27 (t, 3H, SCH2CH3); 13C
NMR (125 MHz, CDCl3) d 165.3, 137.5, 133.3, 129.7, 129.5,
128.5, 128.3, 127.9, 127.8, 84.3 (JC1–H1 = 168.8 Hz), 78.6, 75.3,
73.2, 68.9, 61.1, 26.0, 18.0, 15.0; anal. calcd for C22H25ClO4S: C,
62.77; H, 5.99%, found: C, 62.74; H, 5.89%. ES HRMS calcd
for C22H25ClO4SNa (M + Na): 443.1059, found 443.1056.
Experimental
General methods
Analytical thin-layer chromatography (TLC) was performed
on silica gel 60 F254 (Merck) and detected by UV light at
254 nm and charring with 5% sulfuric acid in ethanol. All
reagents were used as supplied. Chlorine gas was purchased
and used without any further purification. Caution: chlorine gas
is extremely poisonous, corrosive and is strongly oxidizing—
use proper backflow prevention devices in apparatus piping.
When noted, solvents were distilled and dried according to
common literature protocol. Column chromatography was per-
Phenyl 6-O-(tert-butyldimethylsilyl)-1-thio-a-L-mannopyra-
noside (11). To a stirred solution of 1016 (1.72 g, 6.34 mmol) in
pyridine (15 mL) was added TBDMS–Cl (1.43 g, 9.513 mmol).
Et3N (10 mL) was added in 1 mL portions over 10 min,
whereupon the solvents were evaporated under a vacuum. The
residue was dissolved in CH2Cl2, washed with water and dried
with Na2SO4. Column chromatography (3 : 1 ethyl acetate :
toluene) gave the title compound as a white foam (2.20 g, 90%):
˚
formed with silica gel (230–400 mesh, 60 A). High-performance
liquid chromatography (HPLC) was preformed using a Waters
HPLC system using an absorbance detector. Separations were
done using a C18 semi-preparative reverse-phase column, with
combinations of water and methanol as eluents. 1H NMR
spectra were recorded at 500 or 600 MHz and 13C spectra
were recorded at 125 MHz. They were calibrated to residual
solvent signals: dH 7.24 ppm and dC 77.0 ppm for samples
in CDCl3. Samples recorded in D2O were referenced to 0.1%
external acetone (dH 2.225 ppm and dC 31.07 ppm). Isothermal
titration microcalorimetry was performed on a Microcal VP ITC
calorimeter. The optical rotations were measured with a Perkin-
Elmer 241 polarimeter for samples in a 10 cm cell at 22 2 ◦C.
[a]D values are given in units of 10−1 deg cm2 g−1. Electro-spray
mass spectrometry and microanalyses were performed by the
analytical services of this department.
1
[a]2D2 −1.91 (c 0.76, CHCl3); H NMR (600 MHz, CD3OD) d
7.54–7.23 (m, 5H, Ar–H), 5.41 (d, 1H, J = 1.2 Hz, H-1), 4.08–
4.04 (m, 2H, H-2, H-5), 3.99 (dd, 1H, Jgem = 11.2 Hz, Jvic
=
2.0 Hz, H-6a), 3.82 (dd, 1H, Jgem = 11.2 Hz, Jvic = 6.7 Hz,
H-6b), 3.66 (dd, 1H, J = 2.8, 9.3 Hz, H-3), 3.63 (dd ≈ t, 1H,
J = 9.3 Hz), 0.89 (s, 9H, –C(CH3)3), 0.07 (s, 3H, –Si(CH3)a),
0.05 (s, 3H, –Si(CH3)b); 13C NMR (125 MHz, CD3OD) d 136.2,
132.7, 129.9, 128.3, 90.3 (1JC1–H1 = 167.1 Hz), 76.1, 73.6, 73.2,
69.0, 64.5, 26.4, 19.3, −5.0, −5.1; anal. calcd for C18H30O5SSi:
C, 55.92; H, 7.82%, found: C, 55.55; H, 8.00%. ES HRMS calcd
for C18H30O5SSiNa (M + Na): 409.1475, found: 409.1474.
Phenyl 2,3,4-tri-O-benzyl-6-O-(tert-butyldimethylsilyl)-1-thio-
a-L-mannopyranoside (12). A stirring solution of 11 (3.34 g,
8.65 mmol) in dry DMF (70 mL) was cooled to 0 ◦C under
Ar and NaH (95% dry, 1.24 g, 51.9 mmol) was added. The
reaction was allowed to stir for 30 min before the addition of
BnBr (5.15 mL, 43.3 mmol). The mixture was stirred for 16 h
before the slow addition of methanol (25 mL). The solvents
were removed and the crude mixture was dissolved in CH2Cl2
and washed with water. The organic layer was dried with Na2SO4
and evaporated under a vacuum. Column chromatography (12 :
1 hexanes : ethyl acetate) gave a thin clear syrup (5.34 g, 94%):
[a]2D2 −72.8 (c 1.4, CHCl3); 1H NMR (600 MHz, CDCl3); d 7.47–
7.23 (m, 20H, Ar–H), 5.57 (dd, 1H, J = 1.7 Hz, H-1), 4.96 (d, 1H,
Jgem = 11.0 Hz, PhCH2O–), 4.71–4.62 (m, 5H, PhCH2O–), 4.12
(ddd, 1H, J = 1.7, 4.9, 9.6 Hz, H-5), 4.03 (dd ≈ t, 1H, J = 9.4 Hz,
Acetyl 3-O-benzoyl-4-O-benzyl-2-chloro-2-deoxy-a-L-rham-
nopyranose (8). Chlorine gas was slowly bubbled through a
stirring solution of 74 (◦2.18 g, 6.74 mmol) in CCl4 (6 mL) in a
darkened flask at −10 C, until a faint yellow color persisted.
Excess chlorine gas was removed by bubbling argon through
the solution for 1 h, the solvent was removed under a reduced
pressure and evacuated under a high vacuum for 20 min. The
opaque residue was dissolved in glacial acetic acid (5 mL) under
an argon atmosphere, silver acetate (0.702 g, 4.20 mmol.) was
added and the solution was stirred for 12 h at rt. The solution
was filtered through celite and concentrated. The light yellow
syrup was dissolved in CH2Cl2, washed with saturated NaHCO3
and water, then dried with Na2SO4 and evaporated to dryness.
The product was purified by column chromatography (6 : 1
hexanes : ethyl acetate) to afford the target compound as a
clear syrup (1.99 g, 88% overall yield; 3.9 : 1 distribution of
H-4), 4.00 (dd, 1H, J = 1.7, 3.0 Hz, H-2), 3.93 (dd, 1H, Jgem
=
11.5 Hz, Jvic = 5.1 Hz, H-6a), 3.89–3.85 (m, 2H, H-3, H-6b),
0.91 (s, 9H, –SiC(CH3)3), 0.08 (s, 3H, –Si(CH3)a), 0.06 (s, 3H,
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 7 3 3 – 2 7 4 0
2 7 3 7

View Article Online
–Si(CH3)b); 13C NMR (125 MHz, CDCl3) d 138.7, 138.3, 138.0,
134.8, 131.5, 128.9, 128.4, 138.37, 138.31, 128.0, 127.9, 127.8,
127.7, 127.6, 127.2, 85.6, 80.2, 76.6, 75.2, 74.9, 74.3, 72.2, 71.8,
62.7, 26.0, 18.4, −5.1, −5.3; anal. calcd for C39H48O5SSi: C,
71.30; H, 7.36%, found: C, 71.45; H, 7.33%. ES HRMS calcd
for C39H48O5SSiNa (M + Na): 679.2884. Found: 679.2885.
were dissolved in dry CH2Cl2 (40 mL), protected from light and
cooled to −30 ◦C under Ar. NIS (0.81 g, 3.60 mmol) and a
solution of CH2Cl2 (200 lL) saturated with TfOH were added all
at once. The reaction was warmed to −20 ◦C and allowed to stir
for 2 h before neutralization with Et3N (1.5 mL). The reaction
mixture was washed with Na2S2O3 (10%), NaHCO3 and water.
The organic layer was dried with Na2SO4 and concentrated
under a vacuum. Stepped gradient column chromatography
(7 : 1→6 : 1→4 : 1 hexanes : ethyl acetate) afforded a white
Methyl (3-O-benzoyl-4-O-benzyl-2-chloro-2-deoxy-a-L-
rhamnopyranosyl)-(1→3)-4,6-O-benzylidene-2-phthalimido-b-D-
glucopyranoside (14). A solution of glucosamine acceptor 1315
(2.31 g, 5.62 mmol) and rhamnosyl donor 9 (2.86 g, 6.79 mmol)
1
foam (2.10 g, 86%): [a]2D2 −15.65 (c 0.69, CDCl3); H NMR
(600 MHz, CDCl3) d 7.83–7.11 (m, 29H, Ar–H), 5.56 (s, 1H,
PhCHO2–), 5.20 (d, 1H, J = 8.4 Hz, H-1), 4.97 (d, 1H, J =
1.8 Hz, H-1ꢀꢀ), 4.88 (d, 1H, Jgem = 10.8 Hz, PhCH2O–), 4.67 (d,
1H, Jgem = 10.8 Hz, PhCH2O–), 4.61 (d, 1H, Jgem = 11.4 Hz,
PhCH2O–), 4.58 (d, 1H, J = 1.8 Hz, H-1ꢀ), 4.56 (dd, 1H, J =
8.4, 10.2 Hz, H-3), 4.53–4.46 (m, 3H, PhCH2O–), 4.43–4.35 (m,
3H, PhCH2O-, H-6a), 4.25 (dd, 1H, J = 8.5, 10.1 Hz, H-2), 4.08
(dd ≈ t, 1H, J = 9.7 Hz, H-4ꢀꢀ), 4.03 (dd, 1H, J = 3.7, 9.0 Hz,
˚
were dissolved in dry CH2Cl2 (40 mL) and stirred with 3 A
molecular sieves for 1 h under Ar. The flask was darkened
and NIS (1.53 g, 6.79 mmol) and AgOTf (1.395 g, 5.43 mmol)
were added all at once. The reaction mixture was stirred for
15 min and neutralized with Et3N (∼1 mL) before filtration
through celite. The CH2Cl2 solution was washed with Na2S2O3
(10%), NaHCO3 and water. The organic layer was dried with
NaSO4 and concentrated under a reduced pressure. Stepped
gradient column chromatography (5 : 1→4 : 1→2 : 1 hexanes :
ethyl acetate) gave a white foam (3.98 g, 92%): [a]2D2 +9.8 (c 1.1,
H-3ꢀ), 3.95–3.91 (m, 2H, H-5ꢀ, H-6a ), 3.90 (dd, 1H, J = 1.6,
ꢀꢀ
3.7 Hz, H-2ꢀ), 3.87–3.81 (m, 3H, H-6b, H-6b , H-3ꢀꢀ), 3.69–3.63
ꢀꢀ
(m, 3H, H-4, H-5, H-5ꢀꢀ), 3.59 (dd, 1H, J = 1.9, 2.9 Hz), H-2ꢀꢀ),
3.42 (s, 3H, OCH3), 3.19 (dd ≈ t, 1H, J = 9.3 Hz, H-4ꢀ), 0.91
(s, 9H, (CH3)3CSi–), 0.77 (d, 3H, J = 6.1 Hz, H-6ꢀ), 0.12 (s, 6H,
(CH3)2Si–); 13C NMR (125 MHz, CDCl3) d 167.9, 138.9, 138.7,
138.5, 138.3, 137.0, 134.4, 131.2, 129.0, 128.3, 128.2, 128.1,
127.8, 127.6, 127.5, 127.4, 127.2, 127.1, 126.3, 123.6, 101.8,
100.2 (1JC1-H1 = 169.8 Hz, C-1ꢀꢀ), 99.9 (1JC1-H1 = 172.7 Hz, C-1ꢀ),
99.7 (1JC1-H1 = 164.9 Hz, C-1), 80.5, 79.9, 79.3, 77.3, 76.2, 75.1,
74.6, 74.4, 73.9, 72.4, 72.3, 68.7, 68.6, 66.4, 61.2, 60.7, 57.1,
56.5, 26.0, 17.2, −5.0, −5.3; anal. calcd for C68H78ClNO15Si: C,
67.34; H, 6.48; N, 1.15%, found: C, 67.46; H, 6.53; N, 1.13%.
ESMS calcd for C68H78ClNO15SiNa (M + Na): 1234.5, found:
1234.5.
1
CHCl3); H NMR (600 MHz, CDCl3) d 7.98–7.06 (m, 19H,
Ar–H), 5.56 (s, 1H, PhCHO2–), 5.48 (dd, 1H, J = 3.7, 8.7 Hz,
H-3ꢀ), 5.18 (d, 1H, J = 8.5 Hz, H-1), 4.68 (d, 1H, J = 1.9 Hz,
H-1ꢀ), 4.65 (dd, 1H, J = 8.9, 10.2 Hz, H-3), 4.51 (Abq, 2H, J =
11.1, 36.0 Hz, Ph CH2O–), 4.41 (dd, 1H, Jgem = 10.5 Hz, Jvic
=
4.8 Hz, H-6a), 4.30 (dd, 1H, J = 8.6, 10.3 Hz, H-2), 4.13 (dd,
1H, J = 1.9, 3.7 Hz, H-2ꢀ), 4.08 (dq, 1H, J = 6.2, 8.85 Hz, H-5ꢀ),
3.83 (dd ≈ t, 1H, J5,6 ≈ Jvic = 10.1 Hz, H-6b), 3.72 (dd ≈ t, 1H,
J = 9.3 Hz, H-4), 3.67 (ddd, 1H, J5,6 ≈ Jgem = 9.4 Hz, Jvic
=
4.8 Hz, H-5), 3.49 (dd ≈ t, 1H, J = 9.1 Hz, H-4ꢀ), 3.41 (s, 3H,
OCH3), 0.82 (d, 3H, J = 6.2 Hz, H-6ꢀ); 13C NMR (125 MHz,
CDCl3) d 179.8, 165.5, 137.3, 136.9, 134.3, 133.2, 131.3, 129.7,
129.6, 129.0, 128.4, 128.2, 128.1, 127.7, 127.6, 126.3, 101.9,
100.2 (1JC1–H1 = 165.9 Hz), 99.7 (1JC1–H1 = 173.2 Hz), 80.6, 78.5,
74.6, 74.3, 71.9, 68.7, 68.4, 58.7, 57.0, 56.3, 17.3; anal. calcd for
C42ClH40NO11: C, 65.49; H, 5.23; N, 1.82%, found: C, 65.18; H,
5.24; N, 1.69%. ES HRMS calcd for C42ClH40NO11Na (M +
Na): 792.2187, found: 792.2180.
Methyl (2,3,4-tri-O-benzyl-a-L-mannopyranosyl)-(1→3)(4-O-
benzyl-2-chloro-2-deoxy-a-L-rhamnopyranosyl)-(1→3)-4,6-O-
benzylidene-2-deoxy-2-(N-fluorenylmethoxycarbonyl-b-alanyl)-
amido-b-D-glucopyranoside (17). To
a solution of fully
protected trisaccharide 16 (0.721g, 0.595 mmol) in EtOH
(15 mL) was added NH2NH2–H2O (1.0mL, 12.9 mmol) and
the solution was heated to reflux under Ar for 24 h. Upon
cooling the reaction was diluted with CHCl3 and washed
with water. The organic layer was evaporated to dryness and
the resulting solid was suspended in boiling EtOH (100%),
cooled, filtered and washed with cold EtOH to afford the free
amine that needed no further purification. The white solid
was dissolved in THF and TBAF (1.5 mL, 1.0 M in THF)
was added. The solution stirred for 24 h and evaporated to
dryness. The brown syrup was filtered through silica using
toluene as the eluent and evaporated to dryness. The resultant
white solid (0.467 g, 0.482 mmol) was dissolved in dry DMF
(12 mL) and N-Fmoc-b-alanine (0.180, 0.578 mmol), TBTU
(0.309 g, 0.964 mmol), HOBt (0.147 g, 0.964 mmol) and
N-ethylmorpholine (0.24 mL, 1.880 mmol) were added. The
reaction was allowed to stir for 2 h, the solvents were removed
under a vacuum and the residue was dissolved in CH2Cl2 and
washed with water. The organic layer was dried with Na2SO4,
filtered and evaporated to dryness. Column chromatography (2 :
3 : 1 cyclohexane : ethyl acetate : acetone) afforded the target
compound as a white powder (0.510 g, 68%): [a]2D2 −40.69 (c
Methyl (4-O-benzyl-2-chloro-2-deoxy-a-L-rhamnopyranosyl)-
(1→3)-4,6-O-benzylidene-2-phthalimido-b-D-glucopyranoside
(15). Protected disaccharide 14 (2.125 g, 2.76 mmol) was
dissolved in 20 mL of dry CH2Cl2 under Ar at 0 ◦C. Methanol
(40 mL) and sodium metal (88 mg) were added, the reaction
warmed to rt and stirred for 12 h before neutralization with
Rexyn 101 (H+) resin. The resin was filtered and the solution
was concentrated under a vacuum. Column chromatography
(2% acetone in CHCl3) afforded a white powder (1.49 g, 81%):
1
[a]2D2 −55.12 (c 2.1, CHCl3); H NMR (600 MHz, CD2Cl2) d
7.91–7.23 (m, 14H, Ar–H), 5.56 (s, 1H, PhCHO2–), 5.17 (d,
1H, J = 8.6 Hz, H-1), 4.64 (d, 1H, J = 11.3 Hz, PhCH2O–),
4.62 (s, 1H, H-1ꢀ), 4.59 (dd, 1H, J = 8.9, 9.9 Hz, H-3), 4.52
(d, 1H, J = 11.3 Hz, PhCH2O–), 4.40 (d, 1H, J = 4.3, 7.1 Hz,
H-6a), 4.19 (dd, 1H, J = 8.6, 10.1 Hz, H-2), 4.00 (ddd, 1H, 3.8,
7.2, 9.0 Hz, H-3ꢀ), 3.94–3.87 (m, H-2ꢀ, H-5ꢀ), 3.85–3.79 (m, 1H,
H-6b), 3.70–3.63 (m, 2H, H-4, H-5), 3.41 (s, 3H, OCH3), 3.14
(dd ≈ t, 1H, J = 9.2 Hz, H-4ꢀ), 2.13 (d, 1H, J = 7.1 Hz, 3ꢀ-OH),
0.79 (d, 3H, J = 6.2 Hz, H-6ꢀ); 13C NMR (125 MHz, CD2Cl2)
d 168.3, 138.8, 137.7, 135.0, 131.6, 129.4, 128.7, 128.5, 128.09,
128.07, 126.8, 124.0, 102.3, 100.5, 100.1, 81.6, 81.0, 75.1, 75.0,
70.1, 69.1, 68.8, 66.9, 62.8, 57.5, 57.0, 17.7; anal. calcd for
C35H36ClNO10: C, 63.11; H, 5.45; N, 2.10%, found: C, 62.75; H,
5.34; N, 2.09%. ES HRMS calcd for C35H36ClNO10Na (M +
Na): 688.1925, found: 688.1924.
1
1.02, CHCl3); H NMR (600 MHz, CDCl3) d 7.75–7.15 (m,
33H, Ar–H), 5.50 (s, 1H, PhCHO2–), 5.11 (s, 1H, H-1ꢀꢀ), 4.88
(s, 1H, H-1ꢀ), 4.84 (d, 1H, Jgem = 10.9 Hz, PhCH2O–), 4.61
(d, 1H, Jgem = 11.7 Hz, PhCH2O–), 4.59–4.40 (m, 8H, H-1,
H-2ꢀ, PhCH2O–), 4.39–4.32 (m, 3H, H-6a, –O–CH2CHAr),
4.20–4.16 (m, 2H, H-3ꢀ, –OCH2CHAr), 4.01–4.09 (bm, 1H,
Methyl (2,3,4-tri-O-benzyl-6-O-tert-butyldimethylsilyl-a-L-
mannopyranosyl)-(1→3)(4-O-benzyl-2-chloro-2-deoxy-a-L-
rhamnopyranosyl)-(1→3)-4,6-O-benzylidene-2-phthalimido-b-
D-glucopyranoside (16). Disaccharide acceptor 15 (1.33 g,
2.00 mmol) and thiomannosyl donor 12 (2.37 g, 3.60 mmol)
ꢀꢀ
H-3), 3.97–3.88 (m, 4H, H-5ꢀ, H-3ꢀꢀ, H-5ꢀꢀ, H-6aꢀꢀ ), 3.77–3.72 (m,
2H, H-6b, H-4ꢀꢀ), 3.69–3.64 (m, 2H, H-2ꢀꢀ, H-6b ), 3.61–3.53 (m,
1H, H-2), 3.52 (dd ≈ t, J = 9.3 Hz, H-4), 3.48–3.38 (m, 3H,
H-5,tether-ba/b), 3.38–3.33 (m, 4H, OCH3, H-4ꢀ), 2.48–2.36 (m,
2 7 3 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 7 3 3 – 2 7 4 0

View Article Online
2H, tether-aHa/b), 0.81 (d, 3H, J = 6.0 Hz, H-6ꢀ); 13C NMR
(125 MHz, CDCl3) d 172.1, 157.0, 144.1, 143.8, 138.4, 138.3,
138.2, 138.1, 137.0, 129.0, 128.38, 128.35, 128.31, 128.26, 128.2,
128.1, 127.8, 127.7, 127.6, 127.55, 127.53, 127.51, 127.2, 127.1,
126.3, 125.2, 101.7, 101.4, 100.9, 100.8, 79.94, 79.87, 77.9, 77.7,
75.9, 75.4, 75.1, 75.0, 74.0, 72.7, 72.4, 68.9, 68.7, 66.8, 66.3, 63.0,
61.2, 57.2, 56.8, 37.1, 36.7, 17.3; anal. calcd for C72H77ClN2O16:
C, 68.53; H, 6.15; N, 2.22%, found: C, 68.49; H, 6.20; N, 2.37%.
ESMS calcd for C72H77ClN2O16Na (M + Na): 1283.4, found:
1283.4.
1H, J = 6.1 Hz, Jgem = 12.4 Hz, H-6b), 3.67–3.60 (m, 2H, H-4ꢀ,
tether-bHa), 3.57–3.51 (m, 3H, H-3, H-4, tether-bHb), 3.48 (s,
3H, OCH3), 3.42 (ddd, 1H, J = 2.0, 5.9, 8.5 Hz, H-5), 2.64–2.51
(m, 2H, tether-aHa/b), 1.28 (d, 3H, J, 6.2 Hz, H-6ꢀ). 13C NMR
(125 MHz, D2O) d 174.5, 170.8, 104.3, 103.1, 102.7, 86.2, 78.7,
76.8, 73.2, 71.9, 71.2, 70.7, 70.4, 69.8, 68.9, 62.0, 61.6, 57.9,
55.1, 35.4, 34.7, 17.5. ES HRMS calcd for C22H35ClN2O14Na
(M + Na): 609.1674, found 609.1673.
Methyl (a-L-mannopyranosyluronyl)-(1→3)-(2-deoxy-a-L-
rhamnopyranosyl)-(1→3)-2-(3-aminopropionamido)-2-deoxy-
b-D-glucopyranoside lactam (6). To a refluxing solution of 5
(19.4 mg, 33 lmol) in a 1,4-dioxane–THF mixture (3 : 1, 2 mL)
was added TBTH (18 lL, 66 lmol) and AIBN (6 mg, 33 lmol)
and was allowed to react for 12 h. The reaction was cooled to
40 ◦C, dichloromethane was added and refluxed for 30 min. The
mixture was then concentrated under a vacuum. The reaction
was pre-purified with a Waters C-18 Sep-Pak, followed by
reverse-phase (C-18) HPLC utilizing a gradient from 0% to
15% methanol in water to give a clear glass that was lyophilized
to a white solid (14.1 mg, 77%). [a]2D2 = −36.15 (c = 0.52, H2O).
1H NMR (600 MHz, D2O) d 5.07 (s, 1H, H-1ꢀꢀ), 4.94 (d, 1H,
J = 3.5 Hz, H-1ꢀ), 4.40 (d, 1H, J = 8.7 Hz, H-1), 4.04 (dd,
1H, J = 1.6, 3.5 Hz, H-2ꢀꢀ), 3.98–3.86 (m, 6H, H-2, H-6a, H-3ꢀ,
H-5ꢀ, H-4ꢀꢀ, H-5ꢀꢀ), 3.77–3.67 (m, 3 H, H-3ꢀꢀ, H-6b, tether-bHa),
3.52–3.42 (m, 7H, tether-bHb, H-3, H-4, H-5, OCH3), 3.25 (dd,
Methyl (2,3,4-tri-O-benzyl-a-L-mannopyranosyluronyl)-(1→3)-
(4-O-benzyl-2-chloro-2-deoxy-a-L-rhamnopyranosyl)-(1→3)-2-
(3-aminopropionamido)-4,6-O-benzylidene-2-deoxy-b-D-gluco-
pyranoside lactam (18). To a solution of primary alcohol 17
(0.46 g, 0.36 mmol) in MeCN (3 mL) was added TEMPO
(0.057 g, 0.36 mmol), 1 M KBr (0.365 mL), 0.50 M NaHCO3
(1.83 mL) and NaOCl (5–6%, 2.20 mL) at −5 ◦C. The reaction
was allowed to stir for 1 hr before the addition of methanol
(0.5 mL). After 15 min 1 M HCl (15 mL) and CH2Cl2 (25 mL)
were added and the organic layer was removed, washed with
water and dried with Na2SO4. The crude syrup was passed
through a small plug of silica using CH2Cl2 : MeOH (50 :
1) as eluent. The resultant off-white solid was stirred in
20% piperidine–DMF (5 mL) for 30 min. The solvents were
evaporated under a vacuum and co-evaporated with toluene
(3 × 10 mL). The resulting light yellow syrup was dissolved in
DMF (200 mL). PyBOP (0.360 g, 0.69 mmol) was added and
the reaction stirred for 1 h before evaporation under a vacuum.
The residue was chromatographed (4 : 2 : 1 hexane : ethyl
acetate : acetone) to afford an amorphous white solid (0.22 g,
61%): [a]2D2 −39.55 (c = 0.89, CHCl3); 1HNMR (600 MHz,
CD3C(O)CD3) d 7.51–7.21 (m, 27 H, Ar–H, 2 × NH), 5.63
(s, 1H, PhCHO2–), 5.25 (d, 1H, J = 3.5 Hz, H-1ꢀꢀ), 4.94 (d,
1H J = 1.3 Hz, H-1ꢀ), 4.78 (d, 1H, J = 10.3 Hz, PhCH2O–),
4.77 (d, 1H, J = 11.0 Hz, PhCH2O–), 4.64–4.58 (m, 5H,
PhCH2O–), 4.52 (d, 1H, J = 11.3 Hz, PhCH2O–), 4.46 (d, 1H,
J = 8.4 Hz, H-1), 4.34 (dd, 1H, J = 1.7, 3.7 Hz, H-2ꢀ), 4.29
(d, 1H, J = 8.8 Hz, H-5ꢀꢀ), 4.27–4.21 (m, 2H, H-6a, H-2), 4.15
(dd, 1H, J = 3.7, 9.3 Hz, H-3ꢀ), 3.91–3.75 (m, 7H, tether-bHa,
1H J3 ,4 ≈ J4 ,5 = 9.5 Hz, H-4ꢀ), 2.64–2.50 (m, 2H, tether-aa/b),
ꢀ
ꢀ
ꢀ ꢀ
ꢀ
2.05 (dd, 1H, Jvic = 12.5 Hz, Jgem = 4.9 Hz, H-2a ), 1.92 (ddd,
ꢀ
1H, Jvic = 12.5 Hz, Jgem = 12.5, 4.9 Hz, H-2b ), 1.24 (d, 3H,
J = 6.2 Hz, H-6ꢀ). 13C NMR (125 MHz, D2O) d 172.9, 169.6,
102.5, 101.3, 99.9, 84.5, 77.2, 75.4, 74.1, 70.9, 70.5, 69.3, 68.5,
67.5, 60.3, 56.4, 53.7, 37.2, 33.4, 16.4. ES HRMS calcd for
C22H36N2O14Na (M + Na): 575.2064, found 575.2065.
Isothermal titration calorimetry
Titration microcalorimetry was performed for compounds 1–6
according to the procedure described in the preceding paper.1
Acknowledgements
H-3, H-6b, H-5ꢀ, H-2ꢀꢀ, H-3ꢀꢀ, H-4ꢀꢀ), 2.67 (dd, 1H, J3,4
≈
Financial support was provided by research grants to D. R. Bun-
dle from the Natural Science and Engineering Research Council
(NSERC) and the University of Alberta. R. S. McGavin received
NSERC, Alberta Heritage Foundation for Medical Research
and Alberta Research Council postgraduate scholarships. The
authors thank Mrs Joanna Sadowska for the preparation and
purification of monoclonal antibody, and Dr C.-C. Ling for
assistance with MD calculations.
J4,5 = 9.5 Hz, H-4), 3.44 (ddd, 1H, J = 5.1, 9.7, 9.7 Hz, H-5),
3.34 (dd, J3,4 ≈ J4,5 = 9.4 Hz, H-4ꢀ), 3.24 (m, 1H, tether-bHb),
2.48 (ddd, 1H, Jgem = 16.5 Hz, Jvic = 3.0 Hz, tether-aHa), 2.40
(ddd, 1H, Jgem = 16.5 Hz, Jvic = 12.0, 4.0 Hz, tether-aHb), 0.74
(d, 3H, J = 6.2 Hz, H-6ꢀ). 13C NMR (125 MHz, CDCl3) d 171.2,
169.2, 138.3, 138.1, 137.8, 137.1, 129.1, 128.32, 128.28, 128.26,
128.24, 127.9, 127.7, 127.63, 127.58, 127.55, 126.4, 102.6, 102.3,
101.7, 101.6, 82.2, 80.3, 19.1, 78.3, 77.5, 75.1, 73.6, 73.0, 72.8,
72.0, 68.8, 68.6, 66.5, 61.9, 56.7, 54.3, 35.3, 33.5, 17.6; anal.
calcd for C57H63ClN2O14: C, 66.11; H, 6.13; N, 2.71%, found: C,
65.86; H, 6.03; N, 2.69%. ES MS calcd for C57H63ClN2O14Na
(M + Na): 1057.4, found: 1057.4.
References
1 R. A. Gagne, R. S. McGavin and D. R. Bundle, Org. Biomol. Chem.,
DOI: 10.1039/b416105j.
2 N. I. A. Carlin, A. A. Lindberg, K. Bock and D. R. Bundle,
Eur. J. Biochem., 1984, 139, 189.
Methyl (a-L-mannopyranosyluronyl)-(1→3)-(2-chloro-2-deoxy-
a-L-rhamnopyranosyl)-(1→3)-2-(3-aminopropionamido)-2-deoxy-
b-D-glucopyranoside lactam (5). Protected lactam 18 (105 mg,
0.101 mmol) was suspended in methanol (10 mL) and water
(1 mL). Pd(OH)2 (Degussa type, 85 mg) was added and the
reaction stirred under a hydrogen atmosphere (balloon) for
24 h. The reaction was filtered, evaporated and pre-purified with
a Waters C-18 Sep-Pak. Further purification was performed
using reverse-phase (C-18) HPLC utilizing a gradient from
0% to 15% methanol in water. The resultant clear glass was
lyophilized to a white powder (47.2 mg, 79%). [a]2D2 −27.31
J = 1.1 Hz, H-1ꢀꢀ), 5.02 (d, 1H, J = 1.1 Hz, H-1ꢀ), 4.39 (d, 1H,
J = 8.5 Hz, H-1), 4.12–4.06 (m, 4H, H-2ꢀ, H-3ꢀ, H-2ꢀꢀ, H-5ꢀꢀ),
4.03 (dq, 1H, J = 9.7, 6.2 Hz, H-5ꢀ), 3.95–3.85 (m, 3H, H-2,
H-6a, H-4ꢀꢀ), 3.79 (dd, 1H, J = 3.4, 9.3 Hz, H-3ꢀꢀ), 3.74 (dd,
3 F.-I. Auzanneau and D. R. Bundle, Carbohydr. Res., 1993, 247, 195.
4 H. R. Hanna and D. R. Bundle, Can. J. Chem., 1993, 71, 125.
5 M. N. Vyas, N. K. Vyas, P. J. Meikle, B. Sinnot, B. M. Pinto, D. R.
Bundle and F. A. Quiocho, J. Mol. Biol., 1993, 231, 133.
6 N. K. Vyas, M. N. Vyas, M. C. Chervenak, M. A. Johnson, B. M.
Pinto, D. R. Bundle and F. A. Quiocho, Biochemistry, 2002, 41,
13575.
7 A. Maradufu and A. S. Perlin, Carbohydr. Res., 1974, 32, 93.
8 J. P. Carver, Pure Appl. Chem., 1993, 65, 763.
9 M. S. Searle and D. H. Williams, J. Am. Chem. Soc., 1992, 114, 10690.
10 (a) N. Navarre, A. H. van Oijen and G. J. Boons, Tetrahedron Lett.,
1997, 38, 2023; (b) N. Navarre, N. Amiot, A. van Oijen, A. Imberty,
A. Poveda, J. Jimenez-Barbero, A. Cooper, M. A. Nutley and G. J.
Boons, Chem. Eur. J., 1999, 5, 2281; (c) M. C. Galan, A. P. Venot,
J. Glushka, A. Imberty and G. J. Boons, J. Am. Chem. Soc., 2002,
124, 5964; (d) H. C. Kolb, Bioorg. Med. Chem. Lett., 1997, 7, 2629; (e)
M. Wilstermann, J. Balogh and G. Magnusson, J. Org. Chem.,
1997, 62, 3659; (f) D. R. Bundle, R. Alibes, S. Nilar, A. Otter,
1
(c = 0.67, H2O), H NMR (600 MHz, D2O) d 5.13 (d, 1H,
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 7 3 3 – 2 7 4 0
2 7 3 9

View Article Online
M. Warwas and P. Zhang, J. Am. Chem. Soc., 1998, 120, 5317; (g) R.
Alibes and D. R. Bundle, J. Org. Chem., 1998, 63, 6288; (h) S. A.
Wacowich-Sgarbi and D. R. Bundle, J. Org. Chem., 1999, 64, 9080;
(i) P. Zhang and D. R. Bundle, Isr. J. Chem., 200, 40, 189.
11 (a) B. Fraser-Reid and R. U. Lemieux, Can. J. Chem., 1964, 42, 532;
(b) B. Fraser-Reid and R. U. Lemieux, Can. J. Chem., 1965, 43, 1460.
12 K. Igarashi, T. Honma and T. Imagawa, J. Org. Chem., 1970, 35, 610.
13 L. Vargha and J. Kuszmann, Chem. Ber., 1963, 96, 411.
14 T. Zeigler and G. Lemanski, Eur. J. Org. Chem., 1998, 63, 163.
15 N. K. Kochetkov, N. E. Byramova, Y. E. Tsvetkov and L. V.
Backinowsky, Tetrahedron, 1985, 41, 3363.
16 K. Bock, I. Lundt and C. Pedersen, Tetrahedron Lett., 1973, 14, 1037.
17 N. J. Davis and S. L. Flitsch, Tetrahedron Lett., 1993, 34, 1181.
18 J. Coste, D. Lenguyen and B. Castro, Tetrahedron Lett., 1990, 31,
205.
19 A. Otter and D. R. Bundle, J.Magn. Reson., 1995, 109, 194.
20 W. P. Neumann, Synthesis, 1987, 8, 665.
Occhiato, G. Danza, M. Serio and A. Guarnaa, Bioorg. Med. Chem.,
2002, 10, 3455; (c) V. St Georgiev, R. G. van Inwegen and P. Carlson,
Eur. J. Med. Chem., 1990, 25, 375.
22 F.-I. Auzanneau and D. R. Bundle, Can. J. Chem., 1993, 71, 534.
23 B. W. Sigurskjold, E. Altman and D. R. Bundle, Eur. J. Biochem.,
1991, 195, 239.
24 D. R. Bundle and B. W. Sigurskjold, Methods Enzymol., 1994, 247,
288.
25 R. U. Lemieux, M.-H. Du and U. Spohr, J. Am. Chem. Soc., 1994,
116, 9803.
26 (a) R. U. Lemieux, Acc. Chem. Res., 1996, 29, 373; (b) R. U. Lemieux,
L. T. J. Delbaere, H. Beierbeck and U. Spohr, Host–Guest Molecular
Interactions: From Chemistry to Biology, Wiley, London, UK, 1991,
p. 231.
27 S. W. Homans, Biochemistry, 1990, 29, 9110.
28 D. Cameron and D. R. Bundle, unpublished results.
29 R. Stuike-Prill and B. Meyer, Eur. J. Biochem., 1990, 194, 903.
30 (a) M. Mayer and B. Meyer, J. Am. Chem. Soc., 2001, 123, 6108;
(b) B. Meyer and T. Peters, Angew. Chem., Int. Ed., 2003, 42, 864.
21 (a) N. Hossain, A. Zapata, M. Wilstermann, U. J. Nilsson and
G. Magnusson, Carbohydr. Res., 2002, 337, 569; (b) D. Scarpi, E. G.
2 7 4 0
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 2 7 3 3 – 2 7 4 0