Modeling of synthetic phosphono and carba analogues of N-acetyl-α-d- mannosamine 1-phosphate, the repeating unit of the capsular polysaccharide from Neisseria meningitidis serovar A

Article abstract of DOI:10.1039/b907000a

The conformational behavior of methyl (2-acetamido-2-deoxy-α-d- mannopyranosyl)phosphate 1, and its analogues, methyl C-(2-acetamido-2-deoxy- α-d-mannopyranosyl)methanephosphonate 2 and methyl O-(2-acetamido-2-deoxy- 5a-carba-α-d-mannopyranosyl)phosphate

Full text of DOI:10.1039/b907000a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Modeling of synthetic phosphono and carba analogues of
N-acetyl-a-D-mannosamine 1-phosphate, the repeating unit of the capsular
polysaccharide from Neisseria meningitidis serovar A†
Lucio Toma,*a Laura Legnani,a Anna Rencurosi,‡b Laura Poletti,b Luigi Lay*b and Giovanni Russob
Received 7th April 2009, Accepted 15th June 2009
First published as an Advance Article on the web 10th July 2009
DOI: 10.1039/b907000a
The conformational behavior of methyl (2-acetamido-2-deoxy-a-D-mannopyranosyl)phosphate 1, and
its analogues, methyl C-(2-acetamido-2-deoxy-a-D-mannopyranosyl)methanephosphonate 2 and
methyl O-(2-acetamido-2-deoxy-5a-carba-a-D-mannopyranosyl)phosphate 3, where a methylene group
replaces, respectively, the anomeric and the pyranose oxygen atom, was investigated at the
B3LYP/6-311+G(d,p) level [6-311+G(2df,p) for the phosphorus atom]. The energy of the optimized
structures was recalculated using the continuum solvent model C-PCM choosing water as the solvent.
The compounds exhibited several populated conformations, but they all showed a marked preference
for the 4C1 geometry of the pyranose ring; this preference was almost complete for 1, very large for the
phosphono analogue 2, and large for the carba analogue 3. To give experimental support to these
results, compounds 2 and 3 were synthesized and characterized by NMR spectroscopy. The comparison
of the theoretical and experimental vicinal coupling constants confirmed the marked preference for the
4C1 geometry in the case of 2 and suggested that the same holds true for compound 3.
Introduction
Neisseria meningitidis1 belongs to the class of encapsulated bacte-
ria, whose extremely powerful virulence is due to the presence of
a saccharide capsule, embodying the bacterium cell, which delays
the host’s immune response. A possible defense against infections
from encapsulated bacteria can be mediated by the presence of an-
tibodies directed towards their capsular polysaccharides (CPSs).2
Accordingly, a vaccine including a mixture of the CPSs associated
to the most virulent serogroups induce a long term protection.3
Among thirteen clinically significant capsular serogroups of N.
meningitidis, serotype A (MenA), the main responsible for African
outbreaks and epidemics of meningitis, is one of the most virulent.4
The MenA capsule is constituted of N-acetyl-a-D-mannosamine
residues (1→6)-linked through phosphodiester bridges (Chart 1).5
Chart 1
Recent studies have shown the potential of effective synthetic
carbohydrate vaccines against bacteria-induced diseases. Success-
the saccharide fragments in the formulation of the conjugated
ful examples have been reported in the literature for Shigella dysen-
vaccines.9 To overcome this problem, we became interested in
teriae type 1,6 Haemophilus influenzae,7 and Bacillus anthracis.8
designing stable synthetic analogues of the repeating unit of MenA
The design of an effective vaccine against MenA is hampered by
CPS, which could be incorporated into a saccharide chain in order
the major problem of the chemical lability of anomeric glycosyl
to obtain more stable oligomers. These, after conjugation with a
phosphates, which makes it difficult to store and manipulate
proper immunogenic protein carrier, should elicit the proliferation
of protective antibodies cross-reacting with the natural CPS.3a,10
aDipartimento di Chimica Organica, Universita` di Pavia, Via Taramelli 10,
27100, Pavia, Italy. E-mail: lucio.toma@unipv.it; Fax: +39 0382 987323;
Tel: +39 0382 987843
We envisaged two main approaches in order to increase
the stability of the phosphodiester-linked oligomers of MenA
CPS. The first kind of analogue can be obtained by replacing
the anomeric oxygen of the phosphodiester with a methylene
group.11 Following this approach, we recently reported the syn-
thesis of phosphonoester-linked oligomers of MenA CPS.12 Their
biological evaluation showed that they are recognized by a human
polyclonal anti-MenA serum, confirming that the replacement of
the anomeric oxygen atom with a methylene group still permits
antibody binding.12a However, this result does not imply that
bDipartimento di Chimica Organica e Industriale and CISI, Universita` di
Milano, Via Venezian 21, 20133, Milano, Italy. E-mail: luigi.lay@unimi.it;
Fax: +39 02 50314072; Tel: +39 02 50314062
† Electronic supplementary information (ESI) available: 1H, 13C, and
31P NMR spectra of all new compounds. Total energies and carte-
sian coordinates of the conformations of compounds 1–3. See DOI:
10.1039/b907000a
‡ Present address: NiKem Research Srl, Via Zambeletti, 25, 20021
Baranzate (MI), Italy.
3734 | Org. Biomol. Chem., 2009, 7, 3734–3740
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
the overall geometrical features of the natural compound are
preserved in the synthetic oligomers. Due to the high complexity of
their 1H NMR spectra, the experimental NMR data of a suitable
monomeric N-acetylmannosamine 1-phosphonate should allow
easier determination of the conformational preferences of this
kind of compound.
An alternative approach is the replacement of the pyranose
oxygen atom with a methylene group to give a carbocyclic
analogue.13 In this way, the acetal character of the phosphodiester
is lost, and it is expected to be of comparable stability to that of
oligonucleotides.
Since we planned the synthesis of short MenA CPS oligomers
incorporating carba-N-acetylmannosamine units, we decided to
perform preliminary modeling studies on a simpler monomeric
structure to verify on theoretical grounds the preservation of
the overall geometrical properties in this class of analogues.
Thus, looking at the structure of methyl(2-acetamido-2-deoxy-a-
D-mannopyranosyl)phosphate anion 1 (Chart 1) as the minimum
substructure representing the N-acetylmannosamine moiety in
MenA CPS, we report herein a comparative modeling study of
1 with its phosphono and carba analogues 2 and 3, respectively.
Moreover, we also describe the synthesis of compounds 2 and 3,
whose spectral data were employed to give experimental support
to the modeling study.
stability of the 1C4 conformations in 2 and 3 to become comparable
with that of 4C1.
Previous theoretical studies,15 as well as experimental evidence
of the effects of the oxygen/methylene substitution in a-L-
rhamnopyranosyl derivatives,16 showed that the preferred ring
conformation in the analogues is the same as in rhamnopyra-
nosyl phosphate, though the analogues present smaller energy
differences between the 1C4 and 4C1 conformations. 2-Acetamido-
2-deoxy-D-mannose and L-rhamnose belong to opposite steric
series but share the same relative arrangement of the ring
substituents, the main difference being the bulkier substituent
at C2 in mannosamine. The higher steric requirements of the
acetamido group might contribute to a further reduction of the
1
energy difference between the conformers, making the C4–4C1
equilibrium significant.
The modeling study of compounds 1–3 was performed on the
anions, without considering the sodium cation, through opti-
mizations within the density functional approach at the B3LYP
level.17 Large basis sets were used: 6-311+G(d,p) for all the atoms
and 6-311+G(2df,p) for the third-row atom phosphorus. Diffuse
functions were required by the anionic nature of the structures.
All the degrees of conformational freedom were considered,
including rotation around single bonds of the substituents and the
different rotamers of the hydroxymethyl group. The formation of
intramolecular hydrogen bonds among the three hydroxyl groups
and with the several H-bond acceptors in the molecules as well
as the orientation of the phosphate or phosphonate groups were
taken into consideration. The conformational mobility of the
pyranose or carbocyclic ring was investigated considering the 4C1
and 1C4 conformations as well as the twisted-boat geometries. The
energy of the conformations optimized in vacuum was recalculated
using a continuum solvent model, C-PCM,18 choosing water as the
solvent, to obtain values comparable with water solutions. The
results are gathered in Table 1, which reports the energy values
and some geometrical features of the most significant conforma-
tions, together with their percentage contributions to the overall
population determined through the Boltzmann equation. Each
conformation represents an ensemble of conformers differing in
the orientation of the hydroxymethyl group and of all the hydroxyl
groups; only the conformer energetically favored in each ensemble
is reported. The ring geometry and the three torsional angles f, y1,
and y2, describe the differences among the conformers in Table 1
and represent the main conformational features of the compounds
under investigation.
Results and discussion
The pyranose ring of D-mannopyranose and its derivatives usually
assumes the 4C1 conformation, even when the configuration at C1
is a, the anomeric effect being one of the factors that stabilizes
this geometry with respect to the inverted 1C4 chair conformation.
The 2-acetamido-2-deoxy-a-D-mannopyranosyl moiety conforms
to this general behavior also when a phosphate group is linked
at the anomeric position. In fact, complete NMR assignment of
MenA fragments synthesized by Poszgay et al.14 allowed most of
the vicinal 1H–1H coupling constants to be measured, confirming
that the pyranose rings exclusively assume the 4C1 geometry
(J1,2 ª 1.6 Hz, J2,3 ª 4.7 Hz, J3,4 ª 10.0 Hz, J4,5 ª 9.7 Hz).
In compounds 2 and 3, conversely, the equilibrium depicted in
Chart 2 cannot be excluded. Although methylene and oxygen can
be considered isosteric groups, their stereoelectronic properties
are quite different, and the replacement of one of the acetalic
oxygen atoms with a methylene could, in principle, cause major
changes in the conformational behavior. The loss of the anomeric
effect, which in the case of 1 favorably cooperates with the axial
orientation of the substituent at C1, together with the steric
hindrance of both the substituents at C1 and C2, might cause the
As expected, the reference compound 1 showed an almost
4
complete preference for the C1 geometry, populated to a level
>99% (vs. 0.2% of the inverted 1C4 chair), whereas the twisted-boat
conformations gave no contribution to the overall population.
Moreover, the value of the torsional angles f shown by all the
conformers of compound 1 corresponds to that of the exo-
anomeric effect. The lowest energy 1C4 conformation (not reported
in Table 1) showed an energy >3 kcal/mol higher than the global
minimum 1A.
As far as the data of the analogues 2 and 3 are concerned,
Table 1 shows that their minimum energy conformers 2A and 3A,
respectively, still assume the 4C1 conformation and present the exo-
anomeric effect. However, the relative energy of the most stable 1C4
conformations 2D and 3F, 1.94 and 1.72 kcal/mol, respectively,
is lower than the corresponding conformation of compound 1
Chart 2
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3734–3740 | 3735
©

View Article Online
Table 1 Relative energy, recalculated using the continuum solvent model
C-PCM, equilibrium percentages at 298 K, and selected geometrical data
of the B3LYP/6-311+G(d,p)a calculated minimum energy conformations
of compounds 1–3
Erel
(kcal/mol)
Equilibrium Ring
%
◦
◦
conf. f ( ) y1 (◦) y2 ( )
b
c
d
Scheme 1
6 was protected at the primary position as
1A
0.00
0.09
0.70
1.11
1.31
1.49
1.82
38.3
32.9
11.7
5.9
4.2
3.1
4C1
4C1
4C1
4C1
4C1
4C1
4C1
62
60
62
84
88
58
82
167
53
163
-94
-85
43
-75
80
75
-84
101
-122
-168
1B
Compound
1C
thexyldimethylsilyl ether, and the axial hydroxyl was regioselec-
tively acetylated, providing intermediate 7 in 72% overall yield
(Scheme 2). The C2 hydroxyl of 7 was then activated to SN2
reaction by transformation into a triflate ester; this was reacted
with freshly prepared tetrabutylammonium azide, providing the
carba-mannosamine derivative 8 in 69% overall yield, in which
the presence of the azido group was confirmed by IR spectroscopy.
Conversion of the azide into the acetamido group was achieved by
Staudinger reaction, followed by N-acetylation to give compound
9 in 95% yield. The synthesis of methylphosphate 3 was then com-
pleted by deacetylation of compound 9 under Ze´mplen conditions,
followed by the introduction of the monomethyl phosphate group
at the free hydroxyl employing the H-phosphonate protocol,21
to afford phosphodiester 10 in 65% overall yield. Removal
of thexyldimethylsilyl (tetrabutylammonium fluoride in THF)
and benzyl groups (hydrogenolysis over palladium-on-charcoal),
followed by final elution through a column of Dowex-50W X8
resin (Na+ form) provided the carba-analogue 3 in 75% overall
yield. To the best of our knowledge, this is the first synthesis of the
carba-analogue of N-acetylmannosamine 1-phosphate reported
in the literature.
1D
1E
1F
1G
1.8
2.1
-90
Others
2A
0.00
0.42
1.01
1.94
56.8
28.1
10.3
2.1
4C1
4C1
4C1
1C4
49
46
44
174
50
49
-88
89
-119
-85
2B
2C
2D
145
-65
Others
2.7
3A
0.00
0.33
0.95
1.11
1.51
1.72
1.87
2.16
46.5
26.7
9.3
7.1
3.6
2.6
2.0
1.2
1.0
4C1
4C1
4C1
4C1
4C1
1C4
1C4
1C4
68
72
57
79
69
133
113
140
-97
62
155
73
-94
-67
73
-81
74
3B
3C
-72
-169
-168
-79
-172
77
3D
3E
3F
3G
3H
Others
-90
a B3LYP/6-311+G(2df,p) level for the phosphorus atom. b f: O5-C1-O-P
c
for 1, O5-C1-C7-P for 2, C5a-C1-O-P for 3. y1: C1-O-P-O for 1 and 3,
C1-C7-P-O for 2. d y2: O-P-O-CH3 for 1 and 3, C7-P-O-CH3 for 2.
(Erel >3 kcal/mol). Actually, the percentage contribution to the
overall population of the 1C4 conformations was 3% for 2 and 7%
4
for 3, showing the possibility of an equilibrium between the C1
and 1C4 conformations.
It is worth pointing out that the results are strongly influenced
by the computational approach and by the basis set chosen for
the calculations. In fact, if the zero-point energy corrections are
not taken into consideration (data not shown), the percentage
1
contributions of the C4 conformations to the populations of 2
and 3 are predicted to be much greater than those reported in
Table 1.19 Moreover, a similar overestimation of the 1C4 percentage
is found if the medium basis set 6-31G(d) is used (data not
shown). This should suggest the use of a basis set even larger than
6-311+G(d,p); however, the molecular flexibility would make the
computational cost too high, since the optimization of a very
large number of conformations is required. Moreover, it should be
also emphasized that modeling of carbohydrates bearing charged
groups contains a degree of uncertainty higher than modeling
of non polar molecules. For all these reasons, we performed the
synthesis of compounds 2 and 3 to obtain the validation of the
theoretical results by comparing the experimental scalar coupling
constants with those calculated for the predicted conformations.
The synthesis of compound 2 was carried out starting from
known monomethyl phosphonate derivative 4.12a Catalytic hy-
drogenolysis of the benzyl groups of 4 and subsequent elution
through ion exchange resin Dowex-50W X8 (Na+ form), provided
the final phosphonate analogue 2 in 90% overall yield (Scheme 1).
The synthesis of the carba-analogue 3 started from triol 6,
obtained from triacetyl glucal 5 as described in the literature.20
Scheme 2
1
Compounds 2 and 3 were characterized by H NMR, and the
experimental values of the vicinal coupling constants of the ring
hydrogen atoms, diagnostic for the conformational preferences of
the compounds, were determined (Table 2). For every populated
conformation of 2 and 3, the theoretical values of the coupling
3736 | Org. Biomol. Chem., 2009, 7, 3734–3740
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 2 Experimental 1H NMR coupling constants (Hz) of compounds
2 and 3 in comparison with the values calculated with the Haasnoot et al.
equationa for their 4C1 and 1C4 conformations and for the entire set of
populated conformations
the pyranose oxygen atom, was investigated through a modeling
study with a DFT approach. The energy of the optimized
structures was recalculated using a continuum solvent model,
C-PCM, choosing water as the solvent. The compounds exhibited
several populated conformations so that the overall properties of
flexibility and mobility were evaluated with particular attention to
the possibility of the pyranose ring inversion and to the orientation
of the phosphate or phosphonate aglycone, as well as of the
hydroxymethyl group. The 4C1 geometry was preferred by all the
compounds; this preference was almost complete in the natural
reference compound 1, very large in the phosphono analogue 2,
and large for the carba analogue 3.
Calculated values
Exp. values
4C1
1C4
All
Compound 2
J1,2
J2,3
J3,4
J4,5
2.4
4.3
8.8
—
1.3
9.2
1.5
3.3
8.5
9.1
3.1
2.2
1.0
3.3
8.4
8.9
Compound 3
J1,2
3.2
4.8
9.8
10.0
—
—
2.8
3.5
3.0
3.7
9.2
11.0
12.4
3.4
9.3
3.0
2.6
1.7
4.6
1.9
11.3
4.1
3.4
3.6
8.7
10.3
11.8
3.3
To give experimental support to these results, compounds 2 and
3 were synthesized. Their NMR spectroscopic characterization
and the comparison of the theoretical and experimental vicinal
J2,3
J3,4
J4,5
J5,5aax
J5,5aeq
J1,5aax
J1,5aeq
4
coupling constants confirmed the marked preference for the C1
geometry in the case of 2 and suggested that the same holds true
in compound 3.
2.5
3.7
3.1
3.7
These experimental results indicate that replacement of either
the anomeric or the pyranose oxygen atom with a methylene group
allows the preservation of the overall geometrical properties in
both classes of analogues. Moreover, the comparison between the
experimental and the theoretical data highlights the importance
of a proper choice of the calculation level to obtain reliable
computational results in the modeling of these compounds.
a Ref. 22.
constants were calculated by using the equation of Haasnoot
et al.,22 an electronegativity-modified Karplus relationship, appli-
cable to a wide variety of compounds, that reproduces the effects
of the presence and orientation of electronegative atoms on the
coupling constants. The theoretical values were weighted averaged
4
on the basis of the population percentages separately for the C1
and the C4 groups of conformations, as well as for the entire
Experimental
1
General
set of populated conformations. In Table 2 these three series of
computed vicinal coupling constants are reported in comparison
with the experimental values.
NMR spectra were recorded on a Bruker Avance 400 spectrometer
at 298 K, unless otherwise reported. In 13C NMR spectra, signals
corresponding to aromatic carbons are omitted. Chemical shifts
are reported on the d (ppm) scale and in 31P spectra they are
relative to H3PO4. J values are given in Hz. Peak assignments
were based on analysis of 2D spectra (H,H-COSY and HSQC
or HMQC spectra). IR spectra were recorded on a Jasco 4100
FT-IR instrument and wave number values are reported in cm-1.
Optical rotations were measured at rt with a Perkin–Elmer 241
polarimeter. Elemental analyses were performed using the Carlo
Erba elemental analyser 1108. TLC and HPTLC were carried
out on Merck silica gel 60 F-254 plates (0.25 mm and 0.2 mm
thickness, respectively), and spots were viewed by spraying with a
solution containing H2SO4 (31 mL), ammonium molybdate (21 g)
and Ce(SO4)2 (1 g) in 500 mL water, followed by heating at 110 ◦C
for 5 min. Column chromatography was performed by the flash
procedure on Merck silica gel 60 (230–400 mesh). Solvents were
dried by standard procedures.
Measurement of the experimental coupling constant values at
400 MHz are reasonably accurate and precise, especially with
regard to the high values of axial–axial pairs of protons. Thus,
in very good agreement with the theoretical predictions, 1H NMR
4
data of compound 2 confirmed its strong preference for the C1
conformations. Actually, the coupling constant values calculated
both for the entire conformational set and for the sole 4C1
conformations (Table 2) agree with the experimental data, as the
1
small computed contribution of the C4 conformations (3%) is
below the threshold that the approximations connected with the
Haasnoot equation allow one to detect.
The ring coupling constants of 3 calculated for the entire
conformational set have some differences from those determined
4
for the only C1 conformations (Table 2), but in both the cases
they are in fair agreement with the experimental data. The values
4
of J3,4 and J4,5 close to 10 Hz ensure that the C1 conformations
largely prevail, as they would be significantly reduced in the case
of an appreciable contribution of the 1C4 conformations.23 These
pieces of evidence suggest that, for compound 3 in water, the 4C1
conformations are also preferred to a considerable extent, and
Methyl C-(2-acetamido-2-deoxy-a-D-mannopyranosyl)metha-
nephosphonate sodium salt (2). Compound
4 (130 mg,
0.19 mmol) was dissolved in a 3:2 MeOH:H2O mixture (5 mL)
at rt, together with 130 mg of Pd/C catalyst. The reaction was
stirred overnight under hydrogen atmosphere, then it was filtered
through a Celite pad and concentrated. The residue was dissolved
in water and passed through a column of Dowex-50W X8 (Na+
form) ion exchange resin. The solvent was evaporated, giving 2
1
that the contribution of the C4 conformations corresponds to
that computed, or is even lower.
Conclusions
25
The conformational behavior of methyl(2-acetamido-2-deoxy-a-
D-mannopyranosyl)phosphate 1, together with its analogues 2 and
3 in which a methylene group replaces either the anomeric or
(57 mg, 90%); [a]D +3.2 (c 1 in H2O); (Found: C, 35.95; H, 5.11;
N, 4.18. Calc. for C10H19NNaO8P: C, 35.83; H, 5.71; N, 4.18);
dH(400 MHz; D2O) 1.98 (3H, s, NHAc), 2.00 (1H, ddd, J7,1 7.0,
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3734–3740 | 3737
©

View Article Online
J7,7¢ 15.4, J7,P 18.8, 7-H), 2.06 (1H, ddd, J7¢,1 7.7, J7¢,P 18.2, 7¢-H),
3.49 (3H, d, J 10.8, CH3OP), 3.50–3.58 (2H, m, 4-H, 5-H), 3.72
(1H, dd, J6,5 1.9, J6,6¢ 12.2, 6-H), 3.80 (1H, dd, J6¢,5 4.6, 6¢-H), 3.93
(1H, dd, J3,4 8.8, J2,3 4.3, 3-H), 4.1 (1H, dddd, J1,2 2.4, J1,P 9.6,
1-H), 4.28 (1H, dd, 2-H); dC(100.6 MHz; H2O) 22.0 (NHAc); 26.7
(d, J7,P 133, 7-C), 51.2 (d, JMe,P 5, OMe), 52.8 (d, J2,P 9, 2-C), 60.4
The residue was co-evaporated with toluene (3 ¥ 1 mL) and
dried in high-vacuum pump. Then it was dissolved in dry
toluene (2 mL) and dropped into a solution of freshly prepared
tetrabutylammonium azide in dry toluene (2 mL) under argon
atmosphere. The reaction mixture was stirred overnight at 80 ◦C,
then the solvent was evaporated and the residue was purified
by flash chromatography (eluent toluene:EtOAc 98:2) giving 8
=
(6-C), 67.3, 72.9 (4-C, 5-C), 69.3 (3-C), 73.9 (1-C), 174.4 (C O);
25
dP(162 MHz; D2O) 24.5.
(78 mg, 69%); [a]D +22.9 (c 2 in CHCl3); (Found: C, 65.61; H,
8.05; N, 7.30. Calc. for C31H45N3O5Si: C, 65.58; H, 7.99; N, 7.40);
n
3,4-Di-O-benzyl-5a-carba-a-D-glucopyranose (6). Compound
6 was prepared according to the procedure reported in ref. 20,
and its identity was checked by NMR and elemental analysis.
[a]D +85.5 (c 1 in CHCl3); (Found: C, 70.45; H, 7.38. Calc. for
C21H26O5: C, 70.37; H, 7.31); dH(400 MHz; CDCl3) 1.40 (1H, ddd,
max/cm-1 1737.07, 2108.30 and 2399.98; dH(400 MHz, CDCl3)
0.09, 0.10 (6H, 2 s, 2 CH3Si), 0.88 (6H, s, 2 CH3-thexyl), 0.92 (6H,
d, J 6.8, 2 CH3CHthexyl), 1.62–1.68 (2H, m, 5aax-H, CHthexyl), 1.87–
1.95 (2H, m, 5-H, 5aeq-H), 2.02 (3H, s, OAc), 3.56 (1H, dd, J6,5
1.9, J6,6¢ 9.8, 6-H), 3.83–3.92 (4H, m, 2-H, 3-H, 4-H, 6¢-H), 4.64,
4.91 (2H, AB system, J 10.7, CH2Ph), 4.71–4.79 (2H, AB system,
J 11.0, CH2Ph), 5.00 (1H, m, 1-H), 7.28–7.41 (10H, m, HAr);
dC(100.6 MHz, CDCl3) -3.0 (2 CH3Si), 18.6 (2 CH3-thexyl), 20.0 (2
CH3CHthexyl), 21.1 (OAc), 27.1 (5a-C), 34.3 (CHthexyl), 39.8 (5-C),
61.4 (2-C), 62.1 (6-C), 70.5 (1-C), 73.2, 75.6 (2 CH2Ph); 76.8 (4-C),
25
J
5a,1 2.6, J5a,5a¢ 15.1, J5a,5 13.1, 5aax-H), 1.86 (1H, ddd, J5a¢,1 3.7, J5a¢,5
3.7, 5aeq-H), 2.15 (1H, m, 5-H), 3.42 (1H, dd, J4,3 9.2, J4,5 10.4,
4-H), 3.49 (1H, dd, J2,1 3.1, J2,3 9.3, 2-H), 3.58 (1H, dd, J6,5 4.5,
J
6,6¢ 10.8, 6-H), 3.70 (1H, dd, J6,5 4.0, 6¢-H), 3.78 (1H, t, 3-H), 4.06
(1H, m, 1-H), 4.71, 4.95 (2H, AB system, J 11.1, CH2Ph), 4.75,
5.00 (2H, AB system, J 11.4, CH2Ph), 7.32–7.39 (10H, m, HAr);
dC(100.6 MHz; CDCl3) 29.9 (5a-C), 38.5 (5-C), 63.9 (6-C), 68.1
(1-C), 74.4 (2-C), 74.7, 76.7 (2 CH2Ph), 81.9 (4-C), 83.6 (3-C).
=
81.3 (3-C), 169.7 (C O).
2-Acetamido-1-O-acetyl-3,4-di-O-benzyl-2-deoxy-6-O-thexyldi-
methylsilyl-5a-carba-a-D-mannopyranose (9). A mixture of 8
(115 mg, 0.20 mmol) and PPh3 (160 mg, 0.60 mmol) in dry THF
1-O-Acetyl-3,4-di-O-benzyl-6-O-thexyldimethylsilyl-5a-carba-
a-D-glucopyranose (7). To a solution of 6 (204 mg, 0.57 mmol) in
THF (3 mL), imidazole (58 mg, 0.85 mmol) and thexyldimethylsi-
lyl chloride (123 mL, 0.63 mmol) were added at rt. After 24 h, satd
NaHCO3 (10 mL) was added, followed by extraction with EtOAc
(3 ¥ 10 mL). The organic phases were dried (Na2SO4), filtered, and
concentrated under reduced pressure. The crude product was pu-
rified by a short filtration on flash silica gel (eluent hexane:EtOAc
7:3). It was successively dissolved in dry acetonitrile under nitrogen
atmosphere, then trimethyl orthoacetate (154 mL, 1.23 mmol) and
a catalytic amount of p-toluenesulfonic acid were added at rt.
After 10 min, 80% acetic acid (7.5 mL) was added and stirring was
continued for 15 min. Dichloromethane was added and the organic
layer was washed with satd NaHCO3, dried with Na2SO4, filtered,
and concentrated. The crude compound was purified by flash
chromatography (eluent hexane:EtOAc 6:4), providing 7 (217 mg,
◦
was stirred overnight at 60 C under nitrogen atmosphere. After
addition of water (0.5 mL), the reaction was stirred for 24 h
at the same temperature, then the solvent was evaporated. The
residue was dissolved in methanol and acetic anhydride (380 mL,
4.04 mmol) was added. After 24 h the solvent was evaporated
and the crude compound was purified by flash chromatography
(eluent hexane:EtOAc 7:3), providing 9 (111 mg, 95%); [a]D25 +25.2
(c 1.15 in CHCl3); (Found: C, 68.00; H, 8.42; N, 2.35. Calc. for
C33H49NO6Si: C, 67.89; H, 8.46; N, 2.40); dH(400 MHz, CDCl3)
0.09, 0.12 (6H, 2 s, 2 CH3Si), 0.88 (6H, s, 2 CH3-thexyl), 0.93 (6H,
d, J 6.8, 2 CH3CHthexyl), 1.66 (1H, m, CHthexyl), 1.81–2.16 (3H, m,
5-H, 5a-H, 5a¢-H), 1.92 (3H, s, NHAc), 2.04 (3H, s, OAc), 3.67
(1H, dd, J6,5 5.7, J6,6¢ 9.5, 6-H), 5.71 (1H, dd, J4,3 5.8, J4,5 5.8, 4-H),
3.83–3.90 (2H, m, 3-H, 6¢-H), 4.41–4.49 (2H, m, 2-H, CHHPh);
4.58–4.65 (2H, AB system, CH2Ph), 4.74 (1H, d, J 11.3, CHHPh),
5.17 (1H, m, 1-H), 5.63 (1H, d, JNH,2 7.8, NH), 7.30–7.43 (10H, m,
HAr); dC(100.6 MHz, CDCl3) -3.6, -3.5 (2 CH3Si), 18.7 (2 CH3-thexyl),
20.4 (2 CH3CHthexyl), 21.1 (OAc), 23.4 (NHAc), 27.1 (5a-C), 34.3
(CHthexyl), 39.8 (5-C), 50.6 (2-C), 62.5 (6-C), 69.5 (1-C), 72.2, 73.4
25
72% over 2 steps); [a]D +22.4 (c 1 in CHCl3); (Found: C, 68.70;
H, 8.49. Calc. for C31H46O6Si: C, 68.60; H, 8.54); dH(400 MHz;
CDCl3) 0.09, 0.10 (6H, 2 s, 2 CH3Si), 0.87 (6H, s, 2 CH3-thexyl),
0.90 (6H, d, J 6.9, 2 CH3CHthexyl), 1.61–1.68 (2H, m, 5aax-H,
CHthexyl), 1.88 (1H, ddd, J5a,5a¢ 14.8, J5a¢,1 3.7, J5a¢,5 3.7, 5aeq-H),
1.95 (1H, m, 5-H), 2.10 (3H, s, OAc), 2.36 (1H, br s, OH), 3.54
(1H, dd, J6,5 2.0, J6,6¢ 9.7, 6-H), 3.57 (1H, t, J4,3 9.4, J4,5 9.4, 4-H),
3.64 (1H, dd, J2,1 2.9, J2,3 9.5, 2-H), 3.79 (1H, t, 3-H), 3.97 (1H, dd,
J6¢,5 3.4, 6¢-H), 4.70, 4.92 (2H, AB system, J 10.6, CH2Ph), 4.81,
4.99 (2H, AB system, J 11.1, CH2Ph), 5.29 (1H, m, 1-H), 7.31–7.39
(10H, m, HAr); dC(100.6 MHz, CDCl3, signals from HSQC NMR
spectrum) -3.0 (2 CH3Si), 18.6 (2 CH3-thexyl), 20.3 (2 CH3CHthexyl),
21.2 (OAc); 28.8 (5a-C), 34.3 (CHthexyl), 39.7 (5-C), 61.7 (6-C), 71.6
(1-C), 73.3 (2-C), 75.1, 75.6 (2 CH2Ph), 80.3 (4-C), 84.1 (3-C).
=
(2 CH2Ph), 74.3 (4-C), 79.0 (3-C), 170.0, 170.6 (2 C O).
Methyl O-(2-acetamido-3,4-di-O-benzyl-2-deoxy-6-O-thexyl-
dimethylsilyl-5a-carba-a-D-mannopyranosyl)phosphate
triethyl-
ammonium salt (10). Compound 9 (107 mg, 0.18 mmol) was
dissolved into a 1M mixture of MeONa in methanol (3 mL) and
stirred at rt for 2 h, then the reaction mixture was neutralized
to pH 7 with IR-120 resin (H+ form), filtered, and concentrated.
The residue was dissolved into dry CH3CN (1.8 mL), then
pyridine (0.80 mL) and a 1M solution of 2-chloro-4H-1,3,2-
benzodioxaphosphinin-4-one in CH3CN (550 mL, 0.22 mmol)
were added at rt. The mixture was stirred under nitrogen for
45 min, then a pyridine:H2O 1:1 mixture (1.8 mL) was added.
After addition of dichloromethane, the reaction was washed
with water (2 ¥ 10 mL) and a 1M solution of triethylammonium
hydrogen carbonate (TEAB, 2 ¥ 10 mL), dried (Na2SO4), filtered,
and concentrated. The residue was filtered through a short silica
1-O-Acetyl-2-azido-3,4-di-O-benzyl-2-deoxy-6-O-thexyldi-
methylsilyl-5a-carba-a-D-mannopyranose (8). To a solution of 7
(108 mg, 0.20 mmol) in dry dichloromethane (2 mL), pyridine
(64 mL, 0.80 mmol) and triflic anhydride (98 mL, 0.60 mmol) were
added. After 45 min, the reaction mixture was extracted with
dichloromethane (2 ¥ 5 mL), dried (Na2SO4), and concentrated.
3738 | Org. Biomol. Chem., 2009, 7, 3734–3740
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
gel column (eluent CH2Cl2:MeOH 9:1 + 1% triethylamine) to
afford a yellow oil. The crude compound was dissolved in dry
pyridine (1 mL) under argon atmosphere, then methanol (20 mL,
0.50 mmol) and pivaloyl chloride (52 mL, 0.43 mmol) were added
at rt. After 40 min. a freshly prepared solution of iodine in
pyridine:H2O 19:1 was added, then after 15 min. the reaction
was diluted with chloroform and washed with a 1 M solution of
Na2S2O4 (2 ¥ 10 mL) and a 1 M solution of TEAB (2 ¥ 10 mL),
dried (Na2SO4), filtered, and concentrated. The crude compound
was purified by flash chromatography (eluent CH2Cl2:MeOH
of the conformations was recalculated in water at the same
level through single-point calculations using a continuum solvent
model (C-PCM).18 A zero-point correction was carried out.
For the most populated conformers the proton vicinal coupling
constants were calculated with the electronegativity-modified
Karplus relationship.22
References
1 (a) E. Tikhomirov, M. Santamaria and K. Esteves, World Health
Statistics Quarterly, 1997, 50, 170–177; (b) C. Jones, Carbohydrates
in Europe, 1998, 21, 10–16.
25
9:1 + 1% triethylamine) to afford 10 as a glass (84 mg, 62%); [a]D
+8.0 (c 1 in CHCl3); (Found: C, 62.01; H, 8.85; N, 3.74. Calc. for
C38H65N2O8PSi: C, 61.93; H, 8.89; N, 3.80); dH(400 MHz, CDCl3)
0.05, 0.06 (6H, 2 s, CH3Si), 0.83, 0.84 (6H, 2 s, 2 CH3-thexyl), 0.89
(6H, d, J 6.8, 2 CH3CHthexyl), 1.30 (9H, t, J 7.2, CH3CH2NH),
1.62 (1H, m, CHthexyl), 1.71–182 (2H, m, 5a-H, 5a¢-H), 1.92 (3H, s,
NHAc), 2.07 (1H, m, 5-H), 3.02 (6H, q, CH3CH2NH), 3.48–3.69
(2H, m, 4-H, 6-H), 3.54 (d, JH,P 10.4, CH3OP), 3.86 (1H, m,
6¢-H), 4.07 (1H, dd, J3,4 7.6, J3,2 4.4, 3-H), 4.48–4.61 (5H, m, 1-H,
2-H, CH2Ph, CHHPh), 4.81 (1H, d, J 11.2, CHHPh), 7.22–7.36
(10H, m, HAr); dC(100.6 MHz; CDCl3) -3.7, -3.5 (2 CH3Si), 8.7 (3
CH3CH2NH), 18.6, 18.7 (2 CH3-thexyl), 20.4, 20.5 (2 CH3CHthexyl),
23.3 (NHAc), 28.7 (5a-C), 34.3 (CHthexyl), 39.0 (5-C), 45.3 (3
CH3CH2NH), 52.1 (2-C), 52.8 (CH3OP), 62.4 (6-C), 70.4 (1-C),
71.8 (2 CH2Ph), 76.0 (4-C, from HSQC NMR spectrum), 78.5
2 S. Segal and A. J. Pollard, J. Br. Med. Bull., 2005, 72, 65–81 and
references therein.
3 (a) C. T. Bishop and H. J. Jennings, in The Polysaccharides (ed.:
G. O. Aspinall), Academic Press, New York, 1982; (b) M. M. Giuliani,
J. Adu-Bobie, M. Comanducci, B. Arico`, S. Savino, L. Santini, B.
Brunelli, S. Bambini, A. Biolchi, B. Capecchi, E. Cartocci, L. Ciucchi,
F. Di Marcello, F. Ferlicca, B. Galli, E. Luzzi, V. Masignani, D. Serruto,
D. Veggi, M. Contorni, M. Morandi, A. Bartalesi, V. Cinotti, D.
Mannucci, F. Titta, E. Ovidi, J. A. Welsch, D. Granoff, R. Rappuoli
and M. Pizza, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 10834–10839;
(c) A. Gonzalez-Ferna´ndez, J. Faro and C. Ferna´ndez, Vaccine, 2008,
26, 292–300.
4 (a) A. J. Pollard and C. E. Frasch, Vaccine, 2001, 19, 1327–1346; (b) S. L.
Morley and A. J. Pollard, Vaccine, 2001, 20, 666–687.
5 (a) D. R. Bundle, I. C. P. Smith and H. J. Jennings, J. Biol. Chem., 1974,
249, 2275–2281; (b) X. Lemercinier and C. Jones, Biologicals, 2000, 28,
175–183.
6 (a) V. Pozsgay, C. Chu, L. Pannell, J. Wolfe, J. B. Robbins and R.
Schneerson, Proc. Natl. Acad. Sci. U. S. A., 1999, 96, 5194–5197; (b) V.
Pozsgay, J. Kubler-Kielb, R. Schneerson and J. B. Robbins, Proc. Natl.
Acad. Sci. U. S. A., 2007, 104, 14478–14482.
=
(3-C), 171.5 (C O); dP(162 MHz; CDCl3) 2.91.
Methyl
O-(2-acetamido-2-deoxy-5a-carba-a-D-mannopyra-
nosyl)phosphate sodium salt (3). Compound 10 (110 mg,
0.15 mmol) was dissolved in THF (5 mL), then TBAF (1M in
THF + 5% water, 630 mL) was added and the reaction was stirred
overnight at rt. The reaction was extracted with chloroform
(5 ¥ 10 mL), dried (Na2SO4), filtered, and concentrated. The
residue was dissolved in a 1:1 MeOH:H2O mixture (5 mL) at
rt, together with 100 mg of Pd/C catalyst. The reaction was
stirred overnight under hydrogen atmosphere, then it was filtered
through a Celite pad and concentrated. The residue was dissolved
in water and passed through a column of Dowex-50W X8 ion
exchange resin (Na+ form). The solvent was evaporated, giving
7 V. Verez-Bencomo, V. Ferna´ndez-Santana, E. Hardy, M. E. Toledo,
M. C. Rodr´ıguez, L. Heynngnezz, A. Rodriguez, A. Baly, L. Herrera,
M. Izquierdo, A. Villar, Y. Valde´s, K. Cosme, M. L. Deler, M. Montane,
E. Garcia, A. Ramos, A. Aguilar, E. Medina, G. Toran˜o, I. Sosa, I.
Hernandez, R. Mart´ınez, A. Muzachio, A. Carmenates, L. Costa, F.
Cardoso, C. Campa, M. Diaz and R. Roy, Science, 2004, 305, 522–525.
8 (a) D. B. Werz and P. H. Seeberger, Angew. Chem., Int. Ed., 2005, 44,
6315–6318; (b) M. Tamborrini, D. B. Werz, J. Frey, G. Pluschke and P.r
H. Seeberger, Angew. Chem., Int. Ed., 2006, 45, 6581–6582; (c) A. S.
Mehta, E. Saile, W. Zhong, T. Buskas, R. Carlson, E. Kannenberg, Y.
Reed, C. P. Quinn and G.-J. Boons, Chem. Eur. J., 2006, 12, 9136–9149.
9 C. E. Frasch, Adv. Biotechnol. Processes, 1990, 13, 123–145.
10 (a) M. D. Snape and A. J. Pollard, The Lancet, 2005, 5, 21–30 and
references therein; (b) V. Pozsgay, Adv. Carbohydr. Chem. Biochem.,
2000, 56, 153–199.
11 (a) P. Compain and O. R. Martin, Bioorg. Med. Chem., 2001, 9, 3077–
3092; (b) A. Dondoni and A. Marra, Chem. Rev., 2000, 100, 4395–4421;
(c) R. Engel, Chem. Rev., 1977, 77, 349–367; (d) K.-H. Jung and R. R.
Schmidt, in Carbohydrate-based Drug Discovery (ed.: C.-H. Wong),
Wiley-VCH, Weinheim, 2003, vol. 2, pp. 609–659; (e) W. Zou, Curr.
Top. Med. Chem., 2005, 5, 1363–1391.
12 (a) M. I. Torres-Sanchez, C. Zaccaria, B. Buzzi, G. Miglio, G.
Lombardi, L. Polito, G. Russo and L. Lay, Chem. Eur. J., 2007, 13,
6623–6635; (b) M. I. Torres-Sanchez, V. Draghetti, L. Panza, L. Lay
and G. Russo, Synlett, 2005, 7, 1147–1151.
13 O. Arjona, A. M. Go´mez, J. C. Lo´pez and J. Plumet, Chem. Rev., 2007,
107, 1919–2036.
14 A. Berkin, B. Coxon and V. Pozsgay, Chem. Eur. J., 2002, 8, 4424–4433.
15 F. Compostella, F. Marinone Albini, F. Ronchetti and L. Toma,
Carbohydr. Res., 2004, 339, 1323–1330.
25
3 (38 mg, 75%); [a]D +1.4 (c 0.75 in H2O); (Found: C, 35.91;
H, 5.68; N, 4.10. Calc. for C10H19NNaO8P: C, 35.83; H, 5.71; N,
4.18); dH(400 MHz, D2O) 1.60 (1H, m, 5a-H), 1.87–1.93 (2H, m,
5a¢-H, 5-H), 2.00 (3H, s, NHAc), 3.52 (1H, dd, J3,4 9.8, J4,5 10.0,
4-H), 3.56 (3H, d, JH,P 10.8, CH3OP), 3.68 (1H, dd, J6,6¢ 16.7,
J6,5 5.5, 6-H), 3.71 (1H, dd, J6¢,5 3.6, 6¢-H), 3.91 (1H, dd, J2,3 4.8,
3-H) 4.27 (1H, ddd, J1,2 3.2, J1,5a 2.8, J1,5a¢ 3.5, 1-H), 4.36 (1H, dd,
2-H); dC(100.6 MHz, H2O) 22.1 (NHAc), 28.0 (5a-C), 38.7 (5-C),
53.0 (2-C), 53.6 (CH3OP), 62.1 (6-C), 70.1 (4-C), 70.4 (3-C), 72.1
=
(1-C), 174.8 (C O); dP(162 MHz, D2O) 1.22.
Computational methods
16 S. Ronchi, F. Compostella, L. Lay, F. Ronchetti and L. Toma,
Eur. J. Org. Chem., 2005, 4459–4463.
17 (a) A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652; (b) C. Lee, W.
Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
18 V. Barone, M. Cossi and J. Tomasi, J. Comput. Chem., 1998, 19, 404–
417.
19 We are very grateful to one of the referees for suggesting the use of the
zero-point energy corrections.
Calculations were carried out using the Gaussian03 program
package24 through optimizations in the gas phase at the B3LYP/
6-311+G(2df,p) level for phosphorus and 6-311+G(d,p) level for
the other atoms. The high level of calculation was necessary to
correctly describe compounds 1–3 that contain a phosphorus
atom, and the inclusion of diffuse functions was required by
their anionic nature. Vibrational frequencies were computed to
verify that the optimized structures were minima. The energy
20 A. V. R. L. Sudha and M. Nagarajan, Chem. Commun., 1998, 925–
926.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3734–3740 | 3739
©

View Article Online
21 A. V. Nikolaev, I. V. Botvinko and A. J. Ross, Carbohydr. Res., 2007,
342, 297–344.
22 C. A. G. Haasnoot, F. A. A. M. De Leeuw and C. Altona, Tetrahedron,
1980, 36, 2783–2792.
23 See, for example, protected intermediates 9 and 10 which show,
respectively, J4,5 5.8 and J3,4 7.6 (see Experimental section), indicating
that the contribution of the 1C4 conformation in these derivatives is
significant.
24 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N.
Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H.
Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X.
Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth,
P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul,
S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-
Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W.
Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez and J. A.
Pople, Gaussian 03, Revision B.04, Gaussian, Inc., Pittsburgh PA,
2003.
3740 | Org. Biomol. Chem., 2009, 7, 3734–3740
This journal is
The Royal Society of Chemistry 2009
©