An anionic inositol phosphate glycan pseudotetrasaccharide exhibits high insulin-mimetic activity in rat adipocytes

Article abstract of DOI:10.1016/j.bmc.2005.07.020

Inositol phosphate glycan pseudotetrasaccharides consisting of man-(α1-6)-man-(α1-4)-glcN-(α,β1-6)-myo-inositol-1, 2-cyclic phosphate possessing a sulfate group at either O-6 (compounds 3α,β) or O-2 (compounds 4α,β) of the terminal mannose have been prepa

Full text of DOI:10.1016/j.bmc.2005.07.020

Bioorganic & Medicinal Chemistry 13 (2005) 6732–6741
An anionic inositol phosphate glycan pseudotetrasaccharide
exhibits high insulin-mimetic activity in rat adipocytes
Nilanjana Chakraborty and Marc dꢀAlarcao*
Michael Chemistry Laboratory, Department of Chemistry, Tufts University, 62 Talbot Avenue, Medford, MA 02155, USA
Received 26 May 2005; revised 20 July 2005; accepted 20 July 2005
Available online 22 August 2005
Abstract—Inositol phosphate glycan pseudotetrasaccharides consisting of man-(a1-6)-man-(a1-4)-glcN-(a,b1-6)-myo-inositol-1,2-
cyclic phosphate possessinga sulfate group at either O-6 (compounds 3a,b) or O-2 (compounds 4a,b) of the terminal mannose have
been prepared. Compound 4a was able to stimulate lipogenesis in native rat adipocytes to 78% of the maximal insulin response
(MIR) with an EC50 of 1.1 lM. The other compounds exhibited lower maximal stimulations (47–63% MIR) and higher EC50 values
(9.5–10.6 lM).
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
similar to the glycosylphosphatidylinositol (GPI) mem-
brane anchors. Indeed, IPGs capable of activatingthe
insulin signaling pathway have been generated in vitro
from T. bruceii6,7 and S. cerevisiae8–10 by lipolytic and
proteolytic degradation of their GPI-anchored proteins.
Insulin participates in glucose homeostasis by promot-
ing uptake of glucose and synthesis of lipids and/or gly-
cogen by insulin-sensitive cells. The mechanism by
which insulin concentration in the blood is transduced
into the metabolic effects within these cells is complex
and still under active investigation. It is well established
that bindingof insulin to the extracellular domain of the
insulin receptor, a transmembrane tyrosine kinase, re-
sults in phosphorylation of a number of intracellular
proteins and lipids. Ultimately, this phosphorylation
cascade results in the activation of the enzymes control-
ling the various metabolic effects, including glycogen
and lipid synthesis and glucose transport. This part
of the signal transduction pathway has been extensively
reviewed.1–5
Although it has been established that endogenously gen-
erated IPGs mimic many metabolic actions of insulin,
their exact chemical structures and the mechanism(s)
of action remain largely unknown. Labeling and degra-
dation studies of the natural IPGs have revealed that
there are (at least) two classes of molecules with the abil-
ity to mimic some of insulinꢀs activities. One class, re-
ferred to by Larner11 as the pH 2 mediators and by
Rademacher12 as Type P IPGs, contains chiro-inositol
and galactosamine, while the other class, known as the
pH 1.3 mediators or Type A IPGs, contains myo-inosi-
tol and glucosamine. The chemistry and biology of the
endogenous IPGs have been reviewed.13–16
The insulin signal transduction pathway, however, ap-
pears to be even more complex since several cross-talk
mechanisms have been identified that can activate the
signaling cascade without requiring insulin receptor
occupancy. Members of one class of compounds, the
inositol phosphate glycans (IPGs), are capable of stimu-
latingthe insulin-singalingcascade in the absence of
insulin. The IPGs, produced endogenously by insulin-
sensitive cells upon insulin stimulation, are structurally
To understand better the relationship between IPG
structure and biological activity, a large number of
IPG variants with known structures have been chemical-
ly synthesized.17–33 Early work established that even a
pseudodisaccharide containinga myo-inositol cyclic
phosphate moiety glycosidically linked to a glucosamine
(1, Fig. 1) was able to stimulate some insulin-like effects
in insulin-sensitive cells, although the activity is weak
and the concentration required for maximal stimulation
of lipogenesis in rat adipocytes is high (40 lM).20 In a
Keywords: Inositol phosphate glycan; Insulin; Lipogenesis; Diabetes.
*
very impressive study, Muller and his co-workers
¨
Correspondingauthor. Tel.: +1 617 6273686; e-mail: marc.
dalarcao@tufts.edu
reported27 the syntheses and insulin-mimetic activity of
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.07.020

N. Chakraborty, M. dꢀAlarcao / Bioorg. Med. Chem. 13 (2005) 6732–6741
6733
HO
OH
O
HO
HO
O
HO
HO
HO
HO
O
+H3N
HO
O
O
OH
OH
O
HO
O
HO
P
O
-O
O
OH
O
HO
HO
OH
1
O
O
HO
HO
O
O
O
HO
+H3N
HO
HO
OH
OH
OH
O
O
O
NH2
O
2
HO
P
H
H
N
O
-O
O
N
O
O
H2N
N
O
P
O
-O HO
H
O
HO
HO
SH
OH
O
O
HO
O
HO
HO
HO
-O3SO
O
O
O
OH
O
HO
HO
-O3SO
HO
HO
OH
O
O
HO
O
O
O
HO
OH
OH
O
HO
HO
O
+H3N
O
HO
OH
O
P
-O
O
HO
OH
"YCN-PIG"
O
HO
O
O
OH
OH
O
HO
O
+H3N
P
-O
O
"PIG 41"
Figure 1. A selection of synthetic IPGs.
forty-six IPG variants. The IPG referred to by these
workers as PIG 41 (Fig. 1), a pseudohexasaccharide
containinga myo-inositol-1,2-cyclic phosphate, a gluco-
samine, and four mannose residues with sulfate groups
on two of the mannoses, was identified as the most
actively insulin-mimetic synthetic IPG variant. Impor-
tantly, this compound was able to stimulate cells to al-
most the same level of activity as insulin, i.e., PIG 41
showed >90% of the maximal insulin response (MIR)
and an EC50 of 2.5 lM in stimulation of lipogenesis in
that attachinga tripeptide to the distal mannose via a
phosphoethanolamine linkage, as in YCN-PIG
(Fig. 1), further increased the activity of the compound.
While the syntheses of the complex oligosaccharides
PIG 41 and YCN-PIG are impressive, they are also rath-
er lengthy. To simplify future work on the mechanism of
IPG action, it would be desirable to establish the struc-
turally simplest IPG variant that still possesses a high
proportion of the MIR, thereby decreasingthe difficulty
of the synthesis and increasingthe availability of the
rat adipocytes. Mullerꢀs work established that at least
¨
one anionic group on the mannose residues distal from
the inositol moiety is required for achievinga high pro-
portion of the MIR. This conclusion was confirmed by
the observation that IPG 230,32 (Fig. 1) has been shown
compound. Mullerꢀs most active IPGs were those
¨
containingfour mannoses and two anionic groups on
consecutive mannoses. Our goal was to investigate
whether simpler IPGs containingonly two mannose res-
idues but retainingthe requisite anionic group on the
distal mannose, the inositol cyclic phosphate, and gluco-
samine retain significant insulin-mimetic activity.
to be almost inactive.30 Mullerꢀs study, taken together
¨
with previously published observations,6,7 established
that the anomeric configuration between the glucosa-
mine and the terminal inositol residue is relatively unim-
portant to the activity, since compounds with either
anomeric configuration were found to be highly active.
In this correspondence we report the syntheses and bio-
logical activities of IPG variants 3a, 3b, 4a, and 4b
(Fig. 2) that are considerably easier to synthesize than
More recent work by Mullerꢀs group34 demonstrated
¨
-O3SO
OH
HO
-
OSO3
O
O
HO
HO
HO
HO
O
O
OH
OH
O
O
HO
HO
HO
HO
HO
HO
O
O
O
HO
O
HO
O
O
HO
HO
+H3N
+H3N
OH
OH
OH
OH
O
O
O
O
3
α,β
4α,β
P
P
-O
-O
O
O
Figure 2. IPG pseudotetrasaccharides with high insulin-like activity.

6734
N. Chakraborty, M. dꢀAlarcao / Bioorg. Med. Chem. 13 (2005) 6732–6741
PIG 41 and YCN-PIG but that showed high percentages
of the MIR in stimulatinglipoegnesis in adipocytes
in vitro.
usingthe glycosyl fluorides as donors and silver triflate
and bis(cyclopentadienyl)zirconium dichloride as the
promoters.35 This couplingmethod yielded the a-ano-
mer almost exclusively in all cases where a mannosyl
fluoride was the donor but produced a mixture of a
and b anomers (the former predominating) when gluco-
syl fluorides were used in the final couplingreaction.
Therefore, this couplingmethod was ideal for producing
both anomers at the glucosamine-inositol linkage, as
desired based on the earlier determination that both
anomers at this position of IPGs are generally active.
2. Results and discussion
2.1. Syntheses of IPGs 3 and 4
The syntheses of 3 and 4 involved glycosidic coupling of
the differentially protected mannoses 6 or 10 (Fig. 3),
respectively, with disaccharide 11 (Fig. 4), and then gly-
cosylation of myo-inositol 1832 with each of the resulting
trisaccharides. Subsequent modification of the protec-
tive groups and addition of the sulfate and cyclic phos-
phate moieties, followed by complete deprotection in
one step by dissolvingmetal reduction, yielded the final
products (Fig. 4). The modular nature of the syntheses
allowed room for modifications in the final structures,
if necessary. The glycosidic couplings were carried out
Synthesis of the differentially protected mannosyl fluo-
ride donor 6 was carried out in a two-step process from
the known36 mannosyl fluoride 5. This transformation
involved removal of the O-acetyl group with ammonia
in methanol and then formation of the allyl ether37
in its place usingallyl bromide and sodium hydride in
dry DMF. The mannosyl fluoride 6 was obtained in
77% yield over the two steps.
AcO
BnO
BnO
OBn
O
AllO
HO
BnO
OBn
O
OBn
O
a
b
BnO
BnO
BnO
F
F
F
5
6
OMe
BnO
BnO
BnO
BnO
OAll
O
OAc
O
BnO
BnO
BnO
O
O
c
d, a, b
O
BnO
BnO
F
OH
10
8
7
Figure 3. Synthesis of differentially protected mannose units. (a) NH3, CH3OH, 20 ꢁC; (b) allyl bromide, NaH, DMF, 20 ꢁC; (c) NaHSO4, DME/
H2O (4:1), 20 ꢁC; (d) DAST, THF, 20 ꢁC.
RO
BnO
OR'
O
HO
BnO
OBn
O
BnO
a
BnO
OBn
O
O
OBn
O
O
BnO
BnO
BnO
OTBS
OBn
O
N3
O
BnO
11
OTBS
N3
12: R=All,R'=Bn
13: R=Bn,R'=All
b,c
RO
BnO
OR'
O
RO
BnO
OR'
O
OBn
BnO
BnO
BnO
BnO
O
O
O
O
O
OBn
O
OBn
O
BnO
BnO
BnO
BnO
OBn
OH
OBn
O
18
O
O
BnO
O
BnO
F
d
O
N3
BnO
N3
OBn
O
OBn
O
16: R=All,R'=Bn
17: R=Bn,R'=All
19: R=All,R'=Bn
20: R=Bn,R'=All
O
e-i
3
4
α,β
α,β
Figure 4. Completion of the synthesis of 3 and 4. (a) 6 or 10, Cp2ZrCl2, AgOTf, toluene, 20 ꢁC; (b) n-Bu4NF, HOAc, THF, 20 ꢁC; (c) DAST, THF,
20 ꢁC; (d) 18, Cp2ZrCl2, AgOTf, toluene, 20 ꢁC; (e) (Ph3P)3RhCl, DABCO, EtOH/H2O (9:1); 80% HOAc; (f) SO3ÆPy, 20 ꢁC; (g) 1 M aq LiOH, THF,
0 ꢁC; (h) MeOPOCl2, Py, 20 ꢁC; (i) Na, NH3, ꢀ78 ꢁC.

N. Chakraborty, M. dꢀAlarcao / Bioorg. Med. Chem. 13 (2005) 6732–6741
6735
Synthesis of the mannosyl fluoride donor 10 was slightly
more complex, but was achieved in four steps from the
known38 compound 7. The methyl orthoester protecting
group in 7 was removed by mild acid hydrolysis,39 yield-
inga mixture of products. Mannose 8 was easy to sepa-
rate from the 1-acetate that was also formed in the
hydrolysis, and was obtained in 74% yield. It was further
transformed by anomeric fluorination usingDAST in
dry THF in 88% yield,40 after which removal of the O-
acetyl group and allylation provided 10 in a 72% yield
over the final two steps.
yielded the final IPGs 3a,b and 4a,b, which were desalt-
ed usinga Sephadex G-10 size exclusion column to
afford pure products.
2.2. Lipogenesis in native rat adipocytes
Synthetic IPGs 3a, 3b, 4a, and 4b were tested for their
ability to stimulate lipogenesis in isolated native rat adi-
pocytes. For comparison, IPG 1 and insulin were also
tested. The assays were carried out with slight modifica-
tions of the published procedures45,46 by incubating
[6-3H]glucose with a cell suspension and the correspond-
inganalyte (insulin: positive control, buffer: neagtive
control, or the synthetic IPG) for 1.5 h at 37 ꢁC and then
measuringincorporation of tritium into toluene-extract-
able lipids (triglycerides).
The disaccharide acceptor 11 (Fig. 4), prepared by
27
deacetylation of the known correspondingacetate,
was glycosidically linked with either 6 or 10 using
bis(cyclopentadienyl) zirconium dichloride and silver tri-
flate in dry toluene to produce 12 and 13 in 70% and
76% yields, respectively. The a-configuration of the
new glycosidic linkage in the products was confirmed
by the 13C–1H couplingconstant at the anomeric posi-
tion of the mannose accordingto the observation of
Bock and Pedersen41 that a-mannosides exhibit anomer-
ic JC,H = ꢁ 170 Hz while b-mannosides exhibit anomeric
JC,H = ꢁ 160 Hz. In every case in our work where a gly-
cosidic bond was formed from mannose, the anomeric
configuration of the product(s) was assigned by this
method.
The data obtained for the stimulation of lipogenesis by
IPGs (Fig. 5) are reported as the percentage of the max-
imal stimulation (MIR) of lipogenesis above the basal
activity (negative control) produced by insulin (positive
control). In a typical assay, insulin (5 nM) resulted in a
2- to 3-fold increase over the negative control of toluene
soluble tritium, under these conditions. The data relat-
ingIPG concentration to activity could be fit to a stan-
dard receptor model (Eq. 1, n = 1) with R values ranging
from 0.81 to 0.98, but the curve fits were significantly
better if the Hill constant n was allowed to be greater
than 1, suggesting cooperativity in IPG binding. The
parameters obtained from the best curve fits are present-
ed in Table 1.
Removal of the tert-butyldimethylsilyl protectinggroup
was achieved by usingtetrabutylammonium fluoride in
THF under acidic conditions, yielding 14 and 15 in
95% and 92% yields, respectively. Anomeric fluorination
with DAST in dry THF afforded the trisaccharide do-
nors 16 and 17 in 71% and 73% yields, respectively. Gly-
cosidic couplingof the trisaccharide donor with the
protected myo-inositol acceptor 1832 usingthe same cou-
plingmethod yielded a mixture of a and b anomers.
Pseudotetrasaccharide 19a was obtained in 38% yield
together with 15% 19b, while 20a was obtained in 74%
yield alongwith 18% 20b. The anomers were separated
at this stage.
100
80
60
40
20
0
Selective removal of the O-allyl protectingrgoups of
19a, 19b, 20a, and 20b was achieved with Wilkinsonꢀs
catalyst in a 9:1 ethanol/water solution, followed by
42
warmingat reflux in 80% acetic acid. Products 21a,b
and 22a,b were obtained from this reaction in 75%,
46%, 70%, and 31% yields, respectively. The lower con-
version in the case of the b-anomers was likely due to the
lower solubility of these compounds in the ethanol/water
solution. Sulfur trioxide–pyridine complex in dry pyri-
dine solution was used to generate the sulfate ester of
the free hydroxyl group,43 yielding 23a,b and 24a,b in
55–61% yields. Removal of the cyclic carbonate group
on the myo-inositol moiety and replacingit with a cyclic
phosphate were crucial steps, but they were achieved by
reactions that were successfully used previously in our
laboratory.32 The cyclic carbonate was hydrolyzed using
1 M LiOH (aq) in THF at 0 ꢁC for 2 h, affordingthe
pseudotetrasaccharide diols in 55–60% yields for the
four compounds. Cyclic phosphorylation with MeO-
POCl2 in pyridine44 and deprotection by dissolving
0
10
20
30
40
50
60
Concentration (µM)
Figure 5. Stimulation of lipogenesis by synthetic IPGs 1, 3a, and 4a.
Isolated native rat adipocytes were incubated with 6-[3H]glucose
(0.55 mM) and various concentrations of the analyte for 1.5 h. Incor-
poration of tritium into lipids was measured and is expressed as a percent
of the maximal insulin response (%MIR). Black circles (d): compound
1, (0–60 lM); blue diamonds (Ç): compound 3a(0–60 lM); red squares
(j): compound 4a (0–50 lM). Each data point is the average of five
replicates. Solid lines are the result of curve fittingas described in the
text. Error bars represent 1 SD.
32
metal reduction usingthe established procedure

6736
N. Chakraborty, M. dꢀAlarcao / Bioorg. Med. Chem. 13 (2005) 6732–6741
Table 1. Curve-fittingparameters for stimulation of lipogenesis by synthetic IPGs
IPG
Emax (% MIR)
EC50 (lM)
[Glucose] (mM)
n (Hill Constant)
R value for curve fit
1
18
19
63
51
61
78
47
10.5
7.0
0.55
5.50
0.55
5.50
0.55
0.55
0.55
4.2
2.5
2.2
5.6
3.8
1.8
7.7
0.8850
0.9909
0.9897
0.9974
0.9981
0.9630
0.9906
1
3a
3a
3b
4a
4b
10.6
9.8
13.9
1.1
9.5
n
n
n
and 4, or their derivatives, will be useful tools for inves-
tigating the role of IPGs in the insulin signal transduc-
tion process.
Activityð%MIRÞ ¼ ðEmax à ½IPGŠ Þ=ððEC50Þ þ ½IPGŠ Þ.
ð1Þ
At low glucose concentration (0.55 mM), IPG 1 exhibit-
ed only modest ability to stimulate lipogenesis with a
maximal activity (Emax) of 18% that of the maximal
insulin response (MIR) and an EC50 of 10.5 lM, con-
firmingthe earlier report. 20 In contrast, the pseudotetra-
saccharide 4a was extremely active, exhibitingan
Emax = 78% MIR and an EC50 = 1.1 lM. The other
synthetic pseudotetrasaccharides were less active (both
lower Emax and higher EC50) than 4a, but still elicited
significantly higher Emax than 1.
4. Experimental
4.1. Lipogenesis in rat adipocytes
4.1.1. Adipocyte preparation. Epididymal fat pads were
dissected from freshly sacrificed male Long–Evans rats
and kept in 0.15 M NaCl solution at 37 ꢁC until use;
not longer than 0.5 h. The fat pads were cut into small
pieces with scissors and incubated with collagenase
(Sigma–Aldrich; Type 2 from Clostridium histolyticum;
1 mg/mL) in a KRB buffer (10 mM Hepes, 118 mM
NaCl, 4.75 mM KCl, 1.87 mM MgSO4, 1.9 mM CaCl2,
1.2 mM KHSO4, and 25 mM NaHCO3, pH 7.4, saturat-
ed with 95% O2/5% CO2) containing4% BSA and
5.5 mM glucose at 37 ꢁC for 45 min. The digested fat
pads were filtered through silk in order to separate the
adipocytes from the cellular debris. The filtrate contain-
ingthe cell suspension was centrifuegd (Clay Adams
model 151 bench-top centrifuge), the solution below
the floatingcells was discarded, and the cells were
washed three times with KRB buffer containing1%
BSA and 0.55 mM glucose. After each wash, any oil
floatingabove the cells on the surface of the suspension
was removed by suction. The washed adipocytes were
diluted to a final concentration of approximately
5 · 105 cells/mL with KRB buffer (resaturated with
95%O2/5%CO2) containing1% BSA and 0.55 mM glu-
cose, and incubated with 90 nM isoproterenol at 37 ꢁC
for 45 min before use in the lipogenesis assay.
It has been reported that at low glucose concentration
the rate-limiting step in lipogenesis is glucose transport,
while at [glucose] > 1 mM, other processes limit the
rate.47 Accordingly, we examined the ability of two rep-
resentative IPGs, 1 and 3a, to stimulate lipogenesis at
5.5 mM glucose. The resulting Emax and EC50 values
(Table 1) are very similar to those obtained at
0.55 mM glucose, but the Hill constants differ. The
molecular basis of this phenomenon, and indeed of the
observation that the Hill coefficients differ amongst the
compounds, is not clear and probably will have to await
isolation of the IPG receptor for further study. Work
toward this goal is underway.
3. Conclusions
Synthetic IPGs can mimic many of the metabolic actions
of insulin, but to varyingdergees, dependingon the
IPGsꢀ structures. Mullerꢀs work27 demonstrated the
¨
necessity of an anionic group on at least one of the distal
mannose residues, but left open the question of how
many mannose residues are necessary for high activity.
In the current work, we have demonstrated that the
pseudotetrasaccharides 3 and 4, containingtwo man-
nose residues, can stimulate lipogenesis in rat adipocytes
with high percentages of MIR. Furthermore, IPG 4a
has an EC50 value an order of magnitude lower than
those of the other synthetic IPGs reported here and
4.1.2. Lipogenesis assay. To assay stimulation of lipo-
genesis at low glucose concentration, 40 lL of the cell
suspension from above was added to a solution contain-
ing[6- 3H]glucose (0.12 lCi, 0.55 mM) and the analyte in
10 lL KRB buffer containing1% BSA at 37 ꢁC for 1.5 h.
The assay was stopped by adding450 lL water immedi-
ately followed by 5.0 mL of toluene-based scintillation
cocktail (2.5 gPPO and 0.15 gPOPOP per 500 mL tol-
uene) and vortexing. The tubes were centrifuged to expe-
dite phase separation, 3.0 mL of the toluene phase was
removed, and [3H] was measured by scintillation. For
high glucose concentration assays, the procedure was
identical, except that the wash buffer, isoproterenol
incubation buffer, and the assay mixture contained
5.5 mM glucose, and 0.25 lCi of [6-3H]glucose was used
in the assay.
slightly lower than Mullerꢀs ꢁPIG 41ꢀ, although the latter
¨
compound has a slightly higher Emax
27
.
In fact, to the
best of our knowledge, IPG 4a has the lowest EC50
for stimulation of lipogenesis in adipocytes of any syn-
thetic IPG so far reported. Thus, we have successfully
designed and synthesized highly active IPGs that are
considerably simpler to synthesize than previously
reported highly active IPGs. We anticipate that IPGs 3

N. Chakraborty, M. dꢀAlarcao / Bioorg. Med. Chem. 13 (2005) 6732–6741
6737
4.2. Synthesis of 6-O-allyl-2,3,4-tri-O-benzyl-a-D-manno-
pyranosyl fluoride (6)
OCOCH3), 3.63–3.99 (m, 5H), 4.43–4.89 (m, 6H,
CH2Ph), 5.47 (dd, 1H, J = 2 Hz, H-2), 5.61 (dd, 1H,
J = 2, 50 Hz, H-1), 7.12-7.36 (m, 15H, ArH).
Ammonia gas was bubbled through a solution of 5
(80 mg, 0.16 mmol) in methanol at 0 ꢁC for 5 min. The
resultingsolution was stirred at 20 ꢁC for 12 h in a
heavy-walled vessel. The reaction vessel was re-cooled
to 0 ꢁC, opened, and the solution was concentrated.
The crude product was dissolved in DMF (1 mL), and
allyl bromide (20 l L, 0.24 mmol) was added, followed
by sodium hydride (24 mgof a 50% suspension in min-
eral oil, 0.5 mmol). The mixture was stirred at 20 ꢁC un-
til TLC (silica, 1:2 EtOAc/hexanes) showed completion
of reaction, about 2 h. Methanol was then added slowly
to quench the excess sodium hydride, and the solution
was diluted with water (5 mL) and extracted with chlo-
roform (3· 5 mL). The combined organic layers were
washed with brine, dried (Na2SO4), and concentrated
in vacuo. Chromatography of the residue (silica, 1:3
EtOAc/hexanes) gave 6 (61 mg, 77%). Rf = 0.70 (1:2
EtOAc/hexanes). 1H NMR (CDCl3): d 3.59–4.08 (m,
8H), 4.61–5.03 (m, 6H, CH2Ph), 5.17 (dd, 1H, J = 1.3,
10.4 Hz, allyl H-3a), 5.28 (dd, 1H, J = 1.3, 15.3 Hz, allyl
H-3b), 5.57 (dd, 1H, J = 1.8, 50.6 Hz, H-1), 5.87–5.96
(ddt, 1H, J = 10.4, 10.4 and 15.3 Hz, allyl H-2), 7.06–
7.40 (m, 15H, ArH).
4.5. Synthesis of 2-O-allyl-3,4,6-tri-O-benzyl-a-D-manno-
pyranosyl fluoride (10)
Deacetylation and then allylation of
9 (67 mg,
0.13 mmol) were carried out, as described above for
the synthesis of 6. The crude product from the reaction
was purified by chromatography (silica, 1:3 EtOAc/hex-
anes), yielding 10 (48 mg, 72%). Rf = 0.72 (1:2 EtOAc/
hexanes). 1H NMR (CDCl3): d 3.68–4.29 (m, 8H),
4.48–4.90 (m, 6H, CH2Ph), 5.23 (dd, 1H, J = 1.3,
10.4 Hz, allyl H-3a), 5.30 (dd, 1H, J = 1.3, 15.3 Hz, allyl
H-3b), 5.56 (dd, 1H, J = 1.8, 50.6 Hz, H-1), 5.90 (ddt,
1H, J = 10.4, 10.4, 15.3 Hz, allyl H-2), 7.10–7.39 (m,
15H, ArH).
4.6. Synthesis of 2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-1-O-tert-butyl-
dimethylsilyl-a-D-glucopyranoside (11).
Deacetylation of the known27 compound 6-O-acetyl-2,
3,4-tri-O-benzyl-D-mannopyranosyl-(a1 ! 4)-2-azido-2-
deoxy-3,6-di-O-benzyl-1-O-tert-butyldimethylsilyl-a-D-
glucopyranoside (0.47 g, 0.48 mmol) was carried out
usingthe same procedure as described for the synthesis
of 6. Purification by flash column chromatography (silica,
1:3 EtOAc/hexanes) yielded 11 (0.33 g, 74%). Rf = 0.5
4.3. Synthesis of 2-O-acetyl-3,4,6-tri-O-benzyl-D-manno-
pyranose (8)
1
To a solution of 7 (105 mg, 0.21 mmol) in 4:1 DME/
H2O (2 mL) was added NaHSO4 (32 mg, 0.27 mmol)
and the mixture was stirred at 20 ꢁC until TLC (silica,
1:1 EtOAc/hexanes) showed completion of reaction,
approximately 30 min. The excess acid was quenched
with LiOH (1 M, 0.5 mL), the reaction mixture was
diluted with water (5 mL) and extracted with chloro-
form (3· 5 mL). The combined organic layers were dried
(MgSO4) and concentrated. The crude mixture of
products was purified by flash column chromatography
(silica, 1:2 EtOAc/hexanes) to yield 8 (76 mg, 74%).
Rf = 0.55 (1:1 EtOAc/hexanes). 1H NMR (CDCl3): d
2.14 (s, 3H, OCOCH3), 3.58–3.75 (m, 3H, H-4, H-6a,
H-6b), 4.03–4.10 (m, 2H, H-3, H-5), 4.44–4.87 (m, 6H,
CH2Ph), 5.20 (d, 1H, J = 3 Hz, H-1), 5.38 (dd, 1H,
J = 3 Hz, H-2), 7.14–7.32 (m, 15H, ArH).
(1:3 EtOAc/hexanes). H NMR (CDCl3): d 0.15 (s, 6H,
OSiCH3), 0.95 (s, 9H, OSiC(CH3)3), 2.26 (br s, 1H,
OH), 3.21–3.39 (m, 3H), 3.58–3.71 (m, 7H), 3.73–3.85
(m, 3H), 4.21 (d, 1H, J = 12 Hz, HCHPh), 4.39 (d, 1H,
J = 12 Hz, HCHPh), 4.48–4.64 (m, 6H), 4.85 (d, 1H,
J = 10.5 Hz, HCHPh), 4.98 (d, 1H, J = 10.5 Hz,
HCHPh), 5.25 (d, 1H, J = 1.5 Hz, H-1 man), 7.09–7.32
(m, 25H, ArH).
4.7. Synthesis of 6-O-allyl-2,3,4-tri-O-benzyl-D-manno-
pyranosyl-(a1 ! 6)-2,3,4-tri-O-benzyl-D-manno pyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-1-O-tert-butyl-
dimethylsilyl-a-D-glucopyranoside (12) or 2-O-allyl-3,4,
6-tri-O-benzyl-D-mannopyranosyl-(a1 ! 6)-2,3,4-tri-O-ben-
zyl-D-mannopyranosyl-(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-
benzyl-1-O-tert-butyldimethylsilyl-a-D-glucopyranoside (13)
4.4. Synthesis of 2-O-acetyl-3,4,6-tri-O-benzyl-a-D-manno-
pyranosyl fluoride (9)
To a solution of the disaccharide 11 (63 mg, 0.07 mmol
or 20 mg, 0.02 mmol, respectively) and either 6 (61 mg,
0.13 mmol) or 10 (21 mg, 0.04 mmol) in dry toluene
˚
DAST (40 lL, 0.23 mmol) was added to a solution of 8
(76 mg, 0.15 mmol) in dry THF (0.4 mL) at ꢀ42 ꢁC. The
reaction was stirred at 20 ꢁC for approximately 30 min,
until TLC (silica, 1:1 EtOAc/hexanes) showed comple-
tion of reaction. The mixture was cooled to ꢀ42 ꢁC
and the excess DAST was quenched with methanol
(0.5 mL). The mixture was warmed up to 20 ꢁC, fol-
lowed by addition of NaHCO3 (1 M, 2 mL), and
extracted with ether (3· 3 mL). The combined organic
layers were dried (MgSO4) and concentrated. Purifica-
tion by flash column chromatography (silica, 1:3
EtOAc/hexanes) gave 9 (67 mg, 88%). Rf = 0.85 (1:1
EtOAc/hexanes). 1H NMR (CDCl3): d 2.12 (s, 3H,
(2 mL or 0.6 mL, respectively) was added 4A molecular
sieves,
bis(cyclopentadienyl)zirconium
dichloride
(99 mg, 0.34 mmol or 31 mg, 0.11 mmol, respectively),
and then silver trifloromethanesulfonate (174 mg,
0.68 mmol or 54 mg, 0.21 mmol, respectively) at
ꢀ42 ꢁC. The reaction mixture was stirred at 20 ꢁC in
the dark until TLC (silica, 1:2 EtOAc/hexanes) showed
complete disappearance of the limitingstartingmaterial,
approximately 12 h. The reaction was quenched with
NaHCO3 (1 M, 0.5 mL or 0.2 mL, respectively), diluted
with methylene chloride (5 mL or 2 mL, respectively),
and filtered through Celite. The layers were separated
and the aqueous layer was washed with methylene

6738
N. Chakraborty, M. dꢀAlarcao / Bioorg. Med. Chem. 13 (2005) 6732–6741
chloride (3· 5 mL or 3· 2 mL, respectively). The com-
bined organic phase was washed with brine, dried
(MgSO4), and concentrated. Chromatographic purifica-
tion (silica gel, 1:4 EtOAc/hexanes) yielded 12 (67 mg,
70%) or 13 (23 mg, 76%), from reactions performed with
6 or 10, respectively. For 12, Rf = 0.60 (1:2 EtOAc/hex-
Purification by flash column chromatography (silica,
1:4 EtOAc/hexanes) yielded 16 (43 mg, 71%) or 17
(15 mg, 73%). For 16, Rf = 0.52 (1:3 EtOAc/hexanes).
1H NMR (CDCl3): d 3.55–3.73 (m, 19H), 3.80–4.09
(m, 12H), 4.46–5.29 (m, 9H), 5.50 (dd, 1H, J = 2,
50 Hz, H-1 glc-a), 5.79–5.95 (ddt, 1H, J = 10.4, 10.4,
15.3 Hz, allyl H-2), 7.18–7.48 (m, 40H, ArH). HRMS
(ESI): m/z calcd for C77H82FN3O14: 1291.5781.
Found: 1291.5775. For 17, Rf = 0.55 (1:3 EtOAc/hex-
1
anes). H NMR (CDCl3): d 0.15 (s, 6H, OSiCH3), 0.95
(s, 9H, OSiC(CH3)3), 3.20–4.20 (m, 24H), 4.31–4.64
(m, 9H), 4.80–5.29 (m, 8H), 5.78–5.91 (ddt, 1H,
J = 10.4, 10.4, 15.3 Hz, allyl H-2), 7.03–7.39 (m, 40H,
ArH). 13C NMR (CDCl3): d 97.35 (d, 1H, J = 161 Hz,
C-1 glc), 98.54 (d, 1H, J = 170 Hz, C-1 man-2), 100.04
(d, 1H, J = 167 Hz, C-1 man-1). For 13, Rf = 0.65 (1:2
EtOAc/hexanes). 1H NMR (CDCl3): d 0.17 (s, 6H,
OSiCH3), 0.95 (s, 9 H, OSiC(CH3)3), 3.20–3.40 (m,
4H), 3.47–4.02 (m, 16H), 4.09–4.61 (m, 13H), 4.79–
5.32 (m, 8H), 5.77–5.90 (m, allyl H-2), 7.08–7.39 (m,
40H, ArH). 13C NMR (CDCl3): d 98.14 (d, 1H,
J = 159 Hz, C-1 glc), 99.32 (d, 1H, J = 170 Hz, C-1
man-2), 100.89 (d, 1H, J = 171 Hz, C-1 man-1).
anes). 1H NMR (CDCl3):
d
3.37 (wt, 0.6H,
J = 7.5 Hz, H-2 glc-b), 3.45 (dd, 0.4H, J = 3, 10 Hz,
H-2 glc-a), 3.50–3.95 (m, 17H), 4.00–4.05 (m, 2H, allyl
H-1), 4.28 (dd, 1H, J = 2, 15 Hz), 4.38–4.68 (m, 13H),
4.79–5.19 (m, 7H), 5.61 (dd, 0.4H, J = 2, 60 Hz, H-1
glc-a), 5.77–5.92 (ddt, 1H, J = 10.4, 10.4, 15.3 Hz, allyl
H-2), 7.08–7.30 (m, 40H, ArH). HRMS (ESI): m/z
calcd for
1291.5772.
C77H82FN3O14:
1291.5781.
Found:
4.10. Synthesis of 6-O-allyl-2,3,4-tri-O-benzyl-D-manno-
pyranosyl-(a1 ! 6)-2,3,4-tri-O-benzylD-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyrano-
syl-(a and b1 ! 6)-3,4,5-tri-O-benzyl-1-D-myo-inositol-1,
2-cyclic carbonate (19a and 19b)
4.8. Synthesis of 6-O-allyl-2,3,4-tri-O-benzyl-D-mannopyr-
anosyl-(a1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyr-
anose (14) or 2-O-allyl-3,4,6-tri-O-benzyl-D-mannopyrano-
syl-(a1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-(a1 !
4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyranose (15)
Trisaccharide 16 (43 mg, 0.03 mmol) was coupled with
inositol 18 (29 mg, 0.06 mmol), following the proce-
dures described for the synthesis of 12 and 13. The
products were purified by flash column chromatography
(silica, 1:3 EtOAc/hexanes), providing 19a (22 mg, 38%)
and 19b (9 mg, 15%). For 19a, Rf = 0.55 (3:7 EtOAc/
hexanes). 1H NMR (CDCl3): d 3.38 (dd, 1H, J = 3,
9 Hz, H-2 glc), 3.47–3.72 (m, 9H), 3.74–4.03 (m, 13H),
4.13–4.66 (m, 20H), 4.78–4.92 (m, 5H), 4.98–5.33 (m,
5H), 5.77–5.89 (ddt, 1H, J = 10.4, 10.4, 15.3 Hz, allyl
H-2), 7.08–7.32 (m, 55H, ArH). For 19b, Rf = 0.58 in
3:7 EtOAc/hexanes. 1H NMR (CDCl3): d 3.25 (wt,
1H, J = 8 Hz, H-2 glc), 3.41–4.05 (m, 22H), 4.11–4.91
(m, 25H), 5.01–5.40 (m, 6H), 5.78–5.94 (ddt, 1H,
J = 10.4, 10.4, 15.3 Hz, allyl H-2), 7.04–7.40 (m, 55H,
ArH).
To a solution of 12 (67 mg, 0.05 mmol) or 13 (23 mg,
0.02 mmol) in dry THF (2 mL or 0.7 mL, respectively)
was added glacial acetic acid (100 lL, 1.82 mmol or
34 lL, 0.73 mmol, respectively), followed by tetra-
butylammonium fluoride (1 M solution in THF,
0.8 mL, 0.8 mmol or 260 lL, 0.26 mmol, respectively).
The reaction mixture was stirred overnight at 20 ꢁC
and quenched with NaHCO3 (1 M, 2 mL or 1 mL,
respectively). The aqueous layer was extracted with
chloroform (3· 5 mL or 3· 2 mL, respectively), rinsed
with water (3· 5 mL or 3· 2 mL, respectively), brine,
dried with MgSO4 and concentrated. Chromatographic
purifications (silica, 1:2 EtOAc/hexanes) produced 14
(61 mg, 95%) or 15 (20 mg, 92%), from reactions of
12 or 13, respectively. For 14, Rf = 0.3 (1:2 EtOAc/
4.11. Synthesis of 2-O-allyl-3,4,6-tri-O-benzyl-D-manno-
pyranosyl-(1 ! 6)-2,3,4-tri-O-benzyl-D- mannopyranosyl-
(1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyrano-
syl-(a and b1 ! 6)-3,4,5-tri-O-benzyl-1-D-myo-inositol-
1,2-cyclic carbonate (20a and 20b)
1
hexanes). H NMR (CDCl3): d 3.18 (br s, 1H, OH),
3.55–4.09 (m, 19H), 4.44–5.31 (m, 22H), 5.79–5.95
(ddt, 1H, J = 10.4, 10.4, 15.3 Hz, allyl H-2), 7.09–
7.38 (m, 40H, ArH). For 15, Rf = 0.35 (1:2 EtOAc/
hexanes). 1H NMR (CDCl3): d 3.18–4.04 (m, 20H),
4.15–4.66 (m, 14H), 4.79–5.30 (m, 7H), 5.72–5.90
(ddt, 1H, J = 10.4, 10.4, 15.3 Hz, allyl H-2), 7.03–
7.32 (m, 40H, ArH).
Trisaccharide 17 (15 mg, 0.01 mmol) was coupled with
inositol 18 (10 mg, 0.02 mmol, respectively), following
the procedures described for the synthesis of 12 and
13. The products were purified by flash column chroma-
tography (silica, 1:3 EtOAc/hexanes), providing 20a
(15 mg, 74%) and 20b (4 mg, 18%). For 20a, Rf = 0.63
(3:7 EtOAc/hexanes). 1H NMR (CDCl3): d 3.37 (dd,
1H, J = 3.4, 9.8 Hz, H-2 glc), 3.48–4.02 (m, 22H),
4.16–4.65 (m, 20H), 4.78–5.12 (m, 10H), 5.61–5.86
(ddt, 1H, J = 10.4, 10.4, 15.3 Hz, allyl H-2), 7.08–7.31
(m, 55H, ArH). For 20b, Rf = 0.65 in 3:7 EtOAc/hex-
4.9. Synthesis of 6-O-allyl-2,3,4-tri-O-benzyl-D-mannopyr-
anosyl-(a1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-a-D-glucopyr-
anosyl fluoride (16) or 2-O-allyl-3,4,6-tri-O-benzyl-D-manno-
pyranosyl-(a1 ! 6)- 2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-a-D-glucopyr-
anosyl fluoride (17)
1
anes. H NMR (CDCl3): d .21 (wt, 1H, J = 8.5 Hz, H-
Anomeric fluorination of 14 (61 mg, 0.05 mmol) or 15
(20 mg, 0.02 mmol) was carried out following the same
procedure as described above for the synthesis of 9.
2 glc), 3.45–4.05 (m, 20H), 4.16–4.95 (m, 25H), 5.05–
5.29 (m, 6H), 5.75–5.89 (ddt, 1H, J = 10.4, 10.4,
15.3 Hz, allyl H-2), 7.01–7.35 (m, 55H, ArH).

N. Chakraborty, M. dꢀAlarcao / Bioorg. Med. Chem. 13 (2005) 6732–6741
6739
4.12. Synthesis of 2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 6)-2,3,4-tri- O-benzyl-D-mannopyranosyl-(a1 ! 4)-
2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyranosyl-(a1 ! 6)-
3,4,5-tri-O-benzyl-1-D-myo-inositol-1,2-cyclic carbonate (21a)
or 3,4,6-tri-O-benzyl-D-mannopyranosyl-(a1 ! 6)-2,3,4-tri-
O-benzyl-D-mannopyranosyl-(a1 ! 4)-2-azido-2-deoxy-3,6-
di-O-benzyl-D-glucopyranosyl-(a1 ! 6)-3,4,5-tri-O-benzyl-
1-D-myo-inositol-1,2-cyclic carbonate (22a)
5.25 (d, 1H, J = 1.5 Hz, H-1, man-1), 7.05–7.45 (m,
55H, ArH). For 22b, Rf = 0.35 (3:7 EtOAc/hexanes). 1H
NMR (CDCl3): d 3.21 (wt, 1H, J = 8.5 Hz, H-2 glc),
3.48–4.05 (m, 16H), 4.29–4.65 (m, 22H), 4.78–4.95 (m,
10H), 5.10 (ws, 1H, H-1, man-2), 5.25 (d, 1H,
J = 1.5 Hz, H-1, man-1), 7.11-7.49 (m, 55H, ArH).
4.14. Synthesis of 6-O-sulfato-2,3,4-tri-O-benzyl-D-manno-
pyranosyl-(a1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyrano-
syl-(a1 ! 6)-3,4,5-tri-O-benzyl-1-D-myo-inositol-1,2-cyclic
carbonate (23a) or 2-O-sulfato-3,4,6-tri-O-benzyl-D-manno-
pyranosyl-(a1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyrano-
syl-(a1 ! 6)-3,4,5-tri-O-benzyl-1-D-myo-inositol-1,2-cyclic
carbonate (24a)
To tetrasaccharide 19a (5 mg, 0.003 mmol) or 20a (5 mg,
0.003 mmol) and DABCO (1.3 mg, 0.012 mmol) in 9:1
EtOH:H2O (0.3 mL) was added chlorotris(triphenyl-
phosphine) rhodium(I) (0.4 mg) and the mixture was
stirred at 20 ꢁC for 20 min, followed by warmingat re-
flux for 25 min. The reaction was allowed to cool to
20 ꢁC, water (0.2 mL) and chloroform (0.5 mL) were
added, and the layers were separated. The aqueous layer
was extracted with chloroform (3· 0.5 mL). The com-
bined organic layers were washed with brine, dried
(MgSO4), and concentrated. Chromatographic purifica-
tions (silica gel, 1:3 EtOAc/hexanes) yielded the interme-
diate enol ethers, which were dissolved in 80% acetic acid
(100 lL) and heated at 95 ꢁC for 30 min. The reaction
was allowed to cool to 20 ꢁC, diluted with water
(100 lL), chloroform (100 lL), and neutralized by drop-
wise addition of sat. NaHCO3. The aqueous layer was
extracted with chloroform (4· 0.2 mL) and the combined
organic layers were dried (MgSO4) and concentrated.
Purification with column chromatography (silica gel,
1:3 EtOAc/hexanes) yielded 21a (3.7 mg, 75%) or 22a
(3.4 mg, 70%). For 21a, Rf = 0.35 (3:7 EtOAc/hexanes).
1H NMR (CDCl3): d 3.38 (dd, 1H, J = 3, 9 Hz, H-2
glc), 3.47–4.05 (m, 16H), 4.31–4.68 (m, 22H), 4.81–4.96
(m, 9H), 5.05 (ws, 1H, H-1, man-2), 5.21 (d, 1H,
J = 1.5 Hz, H-1 man-1), 5.35 (d, 1H, J = 3.5 Hz, H-1
glc), 7.11–7.49 (m, 55H, ArH). For 22a, Rf = 0.38 (3:7
EtOAc/hexanes). 1H NMR (CDCl3): d 3.37 (dd, 1H,
J = 3.4, 9.8 Hz, H-2 glc), 3.49–4.02 (m, 16H), 4.35–4.71
(m, 22H), 4.82–4.99 (m, 9H), 5.06 (ws, 1H, H-1, man-
2), 5.23 (d, 1H, J = 1.5 Hz, H-1 man-1), 5.38 (d, 1H,
J = 3.5 Hz, H-1 glc), 7.15–7.51 (m, 55H, ArH).
To the solution of 21a (3.7 mg, 0.002 mmol) or 22a
(3.4 mg, 0.002 mmol) in dry pyridine (50 lL) was added
the sulfur trioxide–pyridine complex (9.5 mg, 0.06 mmol)
and the reaction mixture was stirred at 20 ꢁC until TLC
(silica gel, 1:2 EtOAc/hexanes) showed completion of
reaction, approximately 2 h. The reaction was quenched
with NaHCO3 (1 M, 0.1 mL), diluted with water
(1 mL), extracted with ethyl acetate (3· 2 mL), dried
(MgSO4), and concentrated to yield 23a (2.4 mg, 61%)
or 24a (2 mg, 55%), respectively. For 23a, Rf = 0.45 (silica
1
gel, 3:1:0.05 toluene/EtOH/Et3N). H NMR (CDCl3): d
3.38 (dd, 1H, J = 3, 9 Hz, H-2 glc), 3.45–4.32 (m, 22H),
4.41–4.95 (m, 24H), 5.05 (ws, 1H, H-1, man-2), 5.21 (d,
1H, J = 1.5 Hz, H-1 man-1), 5.35 (d, 1H, J = 3.5 Hz, H-
1 glc), 7.08–7.49 (m, 55H, ArH). For 24a, Rf = 0.48 (silica
1
gel, 3:1:0.05 toluene/EtOH/Et3N). H NMR (CDCl3): d
3.37 (dd, 1H, J = 3.4, 9.8 Hz, H-2 glc), 3.52–4.02 (m,
16H), 4.11–4.95 (m, 31H), 5.06 (ws, 1H, H-1, man-2),
5.23 (d, 1H, J = 1.5 Hz, H-1 man-1), 5.38 (d, 1H,
J = 3.5 Hz, H-1 glc), 7.15–7.45 (m, 55H, ArH).
4.15. Synthesis of 6-O-sulfato-2,3,4-tri-O-benzyl-D-manno-
pyranosyl-(a1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyrano-
syl-(b1 ! 6)-3,4,5-tri-O-benzyl-1-D-myo-inositol-1,2-cyclic
carbonate (23b) or 2-O-sulfato-3,4,6-tri-O-benzyl-D-manno-
pyranosyl-(a 1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyrano-
syl-(b1 ! 6)-3,4,5-tri-O-benzyl-1-D-myo-inositol-1,2-cyclic
carbonate (24b)
4.13. Synthesis of 2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-(a1 ! 4)-
2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyranosyl-(b1 ! 6)-
3,4,5-tri-O-benzyl-1-D-myo-inositol-1,2-cyclic carbonate (21b)
or 3,4,6-tri-O-benzyl-D-mannopyranosyl-(a 1 ! 6)-2,3,4-
tri-O-benzyl-D-mannopyranosyl-(a1 ! 4)-2-azido-2-deoxy-
3,6-di-O-benzyl-D-glucopyranosyl-(b1 ! 6)-3,4,5-tri-O-ben-
zyl-1-D-myo-inositol-1,2-cyclic carbonate (22b)
The procedure for formingthe sulfate esters of 21b
(1.5 mg, 0.88 lmol) and 22b (1.3 mg, 0.76 lmol) was the
same as for 23a and 24a, except that 4 mg(0.026 mmol)
of sulfur trioxide–pyridine complex was used in 25 lL
dry pyridine in each case. 23b and 24b were obtained in
60% (0.8 mg) and 58% (0.7 mg) yields, respectively. For
23b, Rf = 0.45 (silica gel, 3:1:0.05 toluene/EtOH/Et3N).
1H NMR (CDCl3): d 3.25 (wt, 1H, J = 8 Hz, H-2 glc),
3.45–4.15 (m, 20H), 4.28–4.91 (m, 27H), 5.07 (ws, 1H,
H-1, man-2), 5.25 (d, 1H, J = 1.5 Hz, H-1, man-1),
7.05–7.55 (m, 55H, ArH). For 24b, Rf = 0.47 (silica gel,
3:1:0.05 toluene/EtOH/Et3N). 1H NMR (CDCl3): d 3.21
(wt, 1H, J = 8.5 Hz, H-2 glc), 3.45–4.18 (m, 18H), 4.25–
4.96 (m, 29H), 5.10 (ws, 1H, H-1, man-2), 5.25 (d, 1H,
J = 1.5 Hz, H-1, man-1), 7.10–7.42 (m, 55H, ArH).
Removal of the allyl groups from 19b (3.2 mg,
0.0015 mmol) and 20b (4.2 mg, 0.0024 mmol) was per-
formed exactly as described for 21a and 22a, except that
0.65 mg(0.006 mmol) or 0.95 mg(0.009 mmol) of DAB-
CO and 0.2 mgor 0.32 mgof chlorotris(triphenylphos-
phine) rhodium(I) were used in 0.3 mL of 9:1 EtOH/
H2O, respectively. Purification with column chromatog-
raphy (silica gel, 1:3 EtOAc/hexanes) yielded 21b
(1.5 mg, 46%) or 22b (1.3 mg, 31%). For 21b, Rf = 0.35
(3:7 EtOAc/hexanes). 1H NMR (CDCl3): d 3.25 (wt, 1H,
J = 8 Hz, H-2 glc), 3.41–3.96 (m, 16H), 4.25–4.65 (m,
22H), 4.76–4.91 (m, 10H), 5.07 (ws, 1H, H-1, man-2),

6740
N. Chakraborty, M. dꢀAlarcao / Bioorg. Med. Chem. 13 (2005) 6732–6741
4.16. Synthesis of 6-O-sulfato-2,3,4-tri-O-benzyl-D-manno-
pyranosyl-(a1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyrano-
syl-(a1 ! 6)-3,4,5-tri-O-benzyl-1-D-myo-inositol (25a) or
2-O-sulfato-3,4,6-tri-O-benzyl-D-mannopyranosyl-(a1 ! 6)-
2,3,4-tri-O-benzyl-D-mannopyrano-syl-(a1 ! 4)-2-azido-2-
deoxy-3,6-di-O-benzyl-D-glucopyranosyl- (a1 ! 6)-3,4,5-tri-
O-benzyl-1-D-myo-inositol (26a)
dine (20lL) and stirringwas continued at 20 ꢁC until
TLC (silica, 1:1:1 CHCl3–diethyl ether–MeOH) showed
completion of reaction. The reaction was quenched by
addition of sat. NaHCO3 (0.2 mL) and co-evaporation
with heptane. The resultingsolid was dissolved in water
(0.75 mL) and adjusted to pH 1 by dropwise addition of
2 M HCl. The solution was extracted with EtOAc
(5 · 0.5 mL) and the combined organic extracts were
dried (Na2SO4), concentrated, dried by coevaporation
with toluene, and dissolved in dry THF (0.5 mL). In a sep-
arate flask, NH3 (1.5 mL) was condensed at ꢀ78 ꢁC and
sodium (4 mg, 0.17 mmol) was added. Once the blue color
persisted, the THF solution of crude product from above
was added dropwise, stirred at ꢀ78 ꢁC for 15 min, and
quenched with solid NH4Cl (9 mg, 0.17 mmol) at
ꢀ78 ꢁC. Once the blue color had disappeared, methanol
(1.25 mL) was added. The reaction mixture was allowed
to warm to 20 ꢁC and evaporated for 12 h. The resulting
white powder was dissolved in deionized water
(0.75 mL) and was desalted by passingthrough Sephadex
G-10 (washed with water), yieldingafter evaporation 3a
(0.3 mg, 50%) or 4a (0.3 mg, 50%). For 3a,1H NMR
(D2O): d 2.55 (wd, 1H, J = 7.5 Hz, H-2 glc), 3.20–4.15
(m, 15H), 4.35 (ddd, 1H, H-1 inos), 4.55 (wt, 1H, H-2
inos), 4.85 (ws, 1H, H-1 man), 5.02 (ws, 1H, H-1 man),
5.15 (d, 1H, J = 3 Hz, H-1 glc). 31P NMR (D2O) d 17.5.
HRMS (ESI): m/z calcd for C24H43NO25PS: 808.1577.
To a stirred solution of 23a (2.4 mg, 0.001 mmol) or 24a
(2 mg, 0.001 mmol) in THF (100 lL) at 0 ꢁC was added
LiOH (1 M, 20 lL). The reaction mixture was stirred at
0 ꢁC for 1.5 h and quenched with NH4Cl (1 M, 0.3 mL).
The aqueous layer was extracted with ethyl acetate (4·
0.5 mL), and the combined organic layers were rinsed
with brine, dried (MgSO4), and concentrated. Purifica-
tion via preparative TLC (silica gel, 2:1:0.05 toluene/
EtOH/Et3N) yielded 25a (1.4 mg, 60%) or 26a (1.2 mg,
60%). For 25a, Rf = 0.3 (2:1:0.05 toluene/EtOH/Et3N).
1H NMR (CDCl3): d 2.52 (br s, 1H, OH), 3.21–4.05
(m, 22 H), 4.11–4.98 (m, 26H), 5.05 (ws, 1H, H-1,
man-2), 5.21 (d, 1H, J = 1.5 Hz, H-1 man-1), 5.35 (d,
1H, J = 3.5 Hz, H-1 glc), 7.11–7.48 (m, 55H, ArH).
For 26a, Rf = 0.35 (2:1:0.05 toluene/EtOH/Et3N). 1H
NMR (CDCl3): d 2.50 (br s, 1H, OH), 3.25–4.08 (m,
22H), 4.15–4.95 (m, 26H), 5.06 (ws, 1H, H-1, man-2),
5.23 (d, 1H, J = 1.5 Hz, H-1 man-1), 5.38 (d, 1H,
J = 3.5 Hz, H-1 glc), 7.11–7.50 (m, 55H, ArH).
1
Found: 808.1567. For 4a, H NMR (D2O): d 2.60 (wd,
1H, J = 7.5 Hz, H-2 glc), 3.22–3.95 (m, 15H), 4.38 (ddd,
1H, H-1 inos), 4.52 (wt, 1H, H-2 inos), 4.85 (ws, 1H, H-
1 man), 4.95 (ws, 1H, H-1 man), 5.15 (d, 1H, J = 3 Hz,
H-1 glc). 31P NMR (D2O) d 16.85. HRMS (ESI): m/z calcd
for C24H43NO25PS: 808.1577. Found: 808.1571.
4.17. Synthesis of 6-O-sulfato-2,3,4-tri-O-benzyl-D-manno-
pyranosyl-(a1 ! 6)-2,3,4-tri-O-benzyl-D-mannopyranosyl-
(a1 ! 4)-2-azido-2-deoxy-3,6-di-O-benzyl-D-glucopyrano-
syl-(b1 ! 6)-3,4,5-tri-O-benzyl-1-D-myo-inositol (25b) or
2-O-sulfato-3,4,6-tri-O-benzyl-D-mannopyranosyl-(a1 ! 6)-
2,3,4-tri-O-benzyl-D-mannopyrano-syl-(a1 ! 4)-2-azido-2-
deoxy-3,6-di-O-benzyl-D-glucopyranosyl-(b1 ! 6)-3,4,5-tri-
O-benzyl-1-D-myo-inositol (26b)
4.19. Synthesis of 6-O-sulfato-D-mannopyranosyl-(a1 ! 6)-
D-mannopyranosyl-(a1 ! 4)-2-amino-2-deoxy-D- glucopyr-
anosyl-(b1 ! 6)-D-myo-inositol-1,2-cyclic phosphate (3b) or
2-O-sulfato-D-mannopyranosyl-(a1 ! 6)-D-mannopyrano-
syl-(a1 ! 4)-2-amino-2-deoxy-D-glucopyranosyl-(b1 ! 6)-
D-myo-inositol-1,2-cyclic phosphate (4b)
Syntheses were carried out from 23b (0.8 mg, 0.44 lmol)
and 24b (0.7 mg, 0.38 lmol) exactly as described for 25a
and 26a. 25b and 26b were obtained in 55% and 58%
yields (0.4 mgof each), respectively. For 25b, Rf = 0.3
(2:1:0.05 toluene/EtOH/Et3N). 1H NMR (CDCl3): d
2.55 (br s, 1H, OH), 3.25–4.15 (m, 22H), 4.25–4.93 (m,
27H), 5.07 (ws, 1H, H-1, man-2), 5.25 (d, 1H,
J = 1.5 Hz, H-1, man-1), 7.08–7.51 (m, 55H, ArH).
For 26b, Rf = 0.35 (silica gel, 2:1:0.05 toluene/EtOH/
Et3N). 1H NMR (CDCl3): d 2.55 (br s, 1H, OH),
3.20–4.11 (m, 22H), 4.20–4.98 (m, 27H), 5.10 (ws, 1H,
H-1, man-2), 5.25 (d, 1H, J = 1.5 Hz, H-1, man-1),
7.05–7.49 (m, 55H, ArH).
Syntheses of these compounds were carried out from
25b and 26b (0.4 mg, 0.22 lmol each) exactly as de-
scribed for 3a and 4a, yielding 3b and 4b (40%,
1
0.07 mgeach), respectively. For 3b, H NMR (D2O): d
3.01–4.05 (m, 22H), 4.31 (ddd, 1H, H-1 inos), 4.56–
5.01 (m, 4H). 31P NMR (D2O) d 16.65. HRMS (ESI):
m/z calcd for C24H43NO25PS: 808.1577. Found:
808.1563. For 4b,1H NMR (D2O): d 3.05–4.20 (m,
22H), 4.29 (ddd, 1H, H-1 inos), 4.50–5.05 (m, 4H). 31P
NMR (D2O) d 17.01. HRMS (ESI): m/z calcd for
C24H43NO25PS: 808.1577. Found: 808.1561.
4.18. Synthesis of 6-O-sulfato-D-mannopyranosyl-(a1 ! 6)-
D -mannopyranosyl-(a1 ! 4)-2-amino-2-deoxy-D-glucopyr-
anosyl-(a 1 ! 6)-D-myo-inositol-1,2-cyclic phosphate (3a)
or 2-O-sulfato-D-mannopyranosyl-(a1 ! 6)-D-mannopyr-
anosyl-(a1 ! 4)-2-amino-2-deoxy-D-glucopyranosyl-(a1 !
6)-D-myo-inositol-1,2-cyclic phosphate (4a)
Acknowledgments
We are grateful to Dean Robin Kanarek, Dr. Wendy
Foulds Mathes, and Ms. Monica Leibovici for their pa-
tient and invaluable assistance with the adipocyte prep-
arations and to Mr. Viatcheslav Azev for important
assistance with the mass spectrometry. Fundingfor this
work was provided by Tufts University. The Tufts
University NMR and mass spectrometry facilities are
PCl2O2Me (500 lL) was slowly added to dry pyridine
(5 mL). The reaction was stirred at 20 ꢁC for 30 min. This
solution (30 lL) was added to a stirred solution of 25a
(1.4 mg, 0.7 lmol) or 26a (1.2 mg, 0.7 lmol) in dry pyri-

N. Chakraborty, M. dꢀAlarcao / Bioorg. Med. Chem. 13 (2005) 6732–6741
6741
24. Zapata, A.; Leon, Y.; Mato, J. M.; Varela-Nieto, I.;
Penades, S., et al. Carbohydr. Res. 1994, 264, 21–31.
25. Garegg, P. J.; Konradsson, P.; Oscarson, S.; Ruda, K.
Tetrahedron 1997, 53, 17727–17734.
supported in part by National Science Foundation
Grants CHE 0320783 and CHE 9723772, respectively.
26. Garegg, P. J.; Konradsson, P.; Lezdins, D.; Oscarson, S.;
Ruda, K., et al. Pure Appl. Chem. 1998, 70, 293–298.
27. Frick, W.; Bauer, A.; Bauer, J.; Wied, S.; Muller, G.
¨
References and notes
1. Saltiel, A. R. Am. J. Physiol. 1996, 270, E375–E385.
2. White, M. F. Diabetologia 1997, 40, S2–S17.
3. Avruch, J. Mol. Cell. Biochem. 1998, 182, 31–48.
4. Watson, R. T.; Pessin, J. E. Recent Prog. Horm. Res. 2001,
56, 175–193.
5. Cheng, A.; Dube, N.; Gu, F.; Tremblay, M. L. Eur. J.
Biochem. 2002, 269, 1050–1059.
6. Misek, D. E.; Saltiel, A. R. J. Biol. Chem. 1992, 267,
16266–16273.
Biochemistry 1998, 37, 13421–13436.
28. Jaworek, C. H.; Calias, P.; Iacobucci, S.; dꢀAlarcao, M.
Tetrahedron Lett. 1999, 40, 667–670.
29. Dietrich, H.; Espinosa, J. F.; Chiara, J. L.; Jimenez-
Barbero, J.; Leon, Y., et al. Chem. Eur. J. 1999, 5, 320–
336.
30. Martin-Lomas, M.; Khiar, N.; Garcia, S.; Koessler, J. L.;
Nieto, P. M., et al. Chem. Eur. J. 2000, 6, 3608–3621.
31. Martin-Lomas, M.; Flores-Mosquera, M.; Khiar, N. Eur.
J. Org. Chem. 2000, 1539–1545.
32. Jaworek, C. H.; Iacobucci, S.; Calias, P.; dꢀAlarcao, M.
Carbohydr. Res. 2001, 331, 375–391.
7. Misek, D. E.; Saltiel, A. R. Endocrinology 1994, 135,
1869–1876.
8. Muller, G.; Wied, S.; Crecelius, A.; Kessler, A.; Eckel, J.
¨
Endocrinology 1997, 138, 3459–3475.
9. Kessler, A.; Muller, G.; Wied, S.; Crecelius, A.; Eckel, J.
¨
Biochem. J. 1998, 330, 277–286.
10. Muller, G.; Wied, S.; Piossek, C.; Bauer, A.; Bauer, J.,
¨
33. Bonilla, J. B.; Munoz-Ponce, J. L.; Nieto, P. M.; Cid, M.
B.; Khiar, N., et al. Eur. J. Org. Chem. 2002, 889–898.
34. Muller, G.; Jung, C.; Frick, W.; Bandlow, W.; Kramer, W.
¨
Arch. Biochem. Biophys. 2002, 408, 7–16.
35. Suzuki, K. Pure Appl. Chem. 1994, 66, 1557–1564.
36. Murakata, C.; Ogawa, T. Carbohydr. Res. 1992, 235,
95–114.
37. Gigg, R.; Payne, S.; Conant, R. J. Carbohydr. Chem. 1983,
2, 207–223.
38. Iacobucci, S.; Ph.D. Thesis; Department of Chemistry,
Tufts University, Medford, MA, 1996.
et al. Mol. Med. 1998, 4, 299–323.
11. Romero, G.; Larner, J. Adv. Pharmacol. 1993, 24, 21–50.
12. Caro, H. N.; Kunjara, S.; Rademacher, T. W.; Leon, Y.;
Jones, D. R., et al. Biochem. Mol. Med. 1997, 61, 214–228.
13. Saltiel, A. R. Diabetes Care 1990, 13, 244–256.
14. Varela–Nieto, I.; Leon, Y.; Caro, H. N. Comp. Biochem.
Physiol. B: Biochem. Mol. Biol. 1996, 115, 223–241.
15. Jones, D. R.; Varela-Nieto, I. Int. J. Biochem. Cell Biol.
1998, 30, 313–326.
39. Corey, E. J.; dꢀAlarcao, M.; Kyler, K. S. Tetrahedron Lett.
1985, 26, 3919–3922.
16. Larner, J.; Huang, L. C. Diabetes Rev. 1999, 7, 217–231.
17. Plourde, R.; dꢀAlarcao, M. Tetrahedron Lett. 1990, 31,
2693–2696.
18. Berlin, W. K.; Zhang, W. S.; Shen, T. Y. Tetrahedron
1991, 47, 1–20.
19. Cobb, J. E.; Johnson, M. R. Tetrahedron 1991, 47, 21–30.
20. Plourde, R.; dꢀAlarcao, M.; Saltiel, A. R. J. Org. Chem.
1992, 57, 2606–2610.
21. Ley, S. V.; Yeung, L. L. Synlett 1992, 997–998.
22. Reddy, K. K.; Falck, J. R.; Capdevila, J. Tetrahedron Lett.
1993, 34, 7869–7872.
40. Posner, G. H.; Haines, S. R. Tetrahedron Lett. 1985, 26, 5–
8.
41. Bock, K.; Pedersen, C. J. Chem. Soc., Perkin Trans. 2
1974, 293–299.
42. Corey, E. J.; Suggs, J. W. J. Org. Chem. 1973, 38, 3224.
43. Xia, J.; Piskorz, C. F.; Alderfer, J. L.; Locke, R. D.;
Matta, K. L. Tetrahedron Lett. 2000, 41, 2773–2776.
44. Smrt, J.; Catlin, J. Tetrahedron Lett. 1970, 11, 5081–5082.
45. Rodbell, M. J. Biol. Chem. 1964, 239, 375–384.
46. Moody, A. J.; Stan, M. A.; Stan, M.; Gliemann, J. Horm.
Metab. Res. 1974, 6, 12–16.
23. Jaramillo, C.; Chiara, J. L.; Martin-Lomas, M. J. Org.
Chem. 1994, 59, 3135–3141.
47. Saltiel, A. R.; Sorbara-Cazan, L. R. Biochem. Biophys.
Res. Commun. 1987, 149, 1084–1092.