Synthesis of phosphate derivatives related to the glycosidase inhibitor salacinol

Article abstract of DOI:10.1016/j.carres.2007.05.030

The syntheses of polyhydroxylated imino- and anhydro thio-alditol compounds related to the naturally occurring glycosidase inhibitor, salacinol, containing a phosphate group in the side chain are described. The compounds lack hydroxyl groups on the acycli

Full text of DOI:10.1016/j.carres.2007.05.030

Carbohydrate Research 342 (2007) 1934–1942
Note
Synthesis of phosphate derivatives related to the
glycosidase inhibitor salacinol
Ramakrishna G. Bhat, Nag S. Kumar and B. Mario Pinto*
Department of Chemistry, Simon Fraser University, Burnaby, British Columbia, Canada V5A 1S6
Received 8 May 2007; received in revised form 23 May 2007; accepted 25 May 2007
Available online 2 June 2007
Abstract—The syntheses of polyhydroxylated imino- and anhydro thio-alditol compounds related to the naturally occurring glyco-
sidase inhibitor, salacinol, containing a phosphate group in the side chain are described. The compounds lack hydroxyl groups on
the acyclic side chain and are prototypes of the exact salacinol analogue. The synthetic strategy relies on the Mitsunobu reaction of
N- and S-hydroxyalkyl derivatives of 2,3,5-tri-O-benzyl-1,4-dideoxy-1,4-imino-D-arabinitol and 1,4-anhydro-2,3,5-tri-O-benzyl-1-
thio-D-arabinitol with dibenzyl phosphate to yield the corresponding protected heteroalditol phosphates. Screening of these com-
pounds against recombinant human maltase glucoamylase (MGA), a critical intestinal glucosidase involved in the processing of
oligosaccharides of glucose into glucose itself, shows that they are not effective inhibitors of MGA and demonstrates the importance
of the hydroxyl and/or sulfate substituents present on the side chain for effective inhibition. The attempted synthesis of the exact
analogue of salacinol by opening of cyclic phosphates is also described.
ꢀ
2007 Elsevier Ltd. All rights reserved.
Keywords: Glycosidase inhibitors; Salacinol-related compounds; Synthesis; Phosphate derivatives; Maltase glucoamylase
Glycosidases play an important role in the metabolism
of carbohydrates and in the processing of specific oligo-
saccharide structures on glycoproteins; the latter play
important roles in intercellular recognition processes
and their modification has been implicated in disease
the positively charged, protonated nitrogen atom with
1
1
carboxylate residues in the enzyme active-site. The
lead glycosidase inhibitors bearing a permanent positive
charge in the form of cyclic sulfonium ions, and presum-
ably also interacting with active-site carboxylate resi-
dues, are perhaps the naturally occurring compounds
1
–7
8
states such as cancer. Inhibitors of glycosidases have
been attractive target compounds for synthetic chemists
and biochemists, not only because they serve as useful
biological tools for studying the biological functions of
1
2,13
salacinol (1) and kotalanol (2).
OH
OH OH OH
9
oligosaccharides, but also because they have great po-
OH
OH
OSO3
tential as drugs to treat a variety of carbohydrate-med-
S
S
OH
1
0
OSO3
iated diseases.
HO
HO
Alkaloids mimicking the structures of monosaccha-
rides are now believed to be widespread in plants and
microorganisms, and these sugar mimics inhibit glycosi-
dases because of their structural resemblance to the
sugar moiety. It has been hypothesized that this strong
binding is the outcome of electrostatic interactions of
HO
OH
HO
OH
1
2
The replacement of the carboxylic acid functional
group in biologically important molecules by phospho-
ric acid continues to attract much interest in bioorganic
and medicinal chemistry. Much of the progress in this
field has been associated with the phosphorus analogues
of amino acids. The tetrahedral configuration, owing to
1
4
*
Corresponding author. Tel.: +1 604 291 4152; fax: +1 604 291
860; e-mail: bpinto@sfu.ca
4
0
008-6215/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2007.05.030

R. G. Bhat et al. / Carbohydrate Research 342 (2007) 1934–1942
1935
the presence of the phosphorus atom, allows these com-
pounds to serve as stable analogues of the unstable tetra-
hedral intermediates formed in enzymatic processes.
cinol (1) as potential glycosidase inhibitors. These com-
pounds will also serve to test the significance of sulfate
and/or hydroxyl groups on inhibition of glycosidases.
CO H
2
HO C
O
O
P
2
OH
O
OH
OH
O
OH
OH
HO C
H N
P
2
2
Me N
N
H
3
O
O
3
4
5
OH
O
OH
O
OH O
H N
Me N
3
P
O
H3N
OH
P
O
3
OH
OH
O
6
7
8
O
P
O
OH
HO
HO
OH
O
N
N
OH
S
OPO H
3
OH OH
HO
O
NH
HN
HO
OH
O
9
10
O
OH
P
P
O
OH
OH
OH
O
N
S
OTf
O
HO
HO
HO
OH
HO
OH
1
1
12
Many of these compounds act as enzyme inhibitors.
Others are very interesting because they can help eluci-
date the functions of biologically active compounds in
living organisms. For example, N-(phosphonoacetyl)-
L-aspartate (3) and O-phosphate serine (4) were shown
Compounds 10–12 could be synthesized by alkylating
the anhydro-alditol derivatives at the ring heteroatom.
The alkylation of a protected 1,4-anhydro-1-thio-D-ara-
binitol with the cyclic phosphate (13) derived from L-
erythritol would afford compound 10 (Scheme 1). This
potential reaction is patterned after our earlier syntheses
of salacinol (1) and its analogues by opening of cyclic
1
5
to be inhibitors of the carbonic anhydrase enzyme.
The involvement of carnitine (5) and c-amino-b-
hydroxybutyric acid (6) in the biology of mammalian
cells and some important aspects of medicinal treatment
has led to the development of their pharmacologically
1
8–30
sulfates.
In order to check the general reactivity of cyclic phos-
phates, the cyclic phosphates (14, 15) derived from D-
erythritol were synthesized, in turn, from the less expen-
sive D-glucose using a similar protocol as for the synthe-
1
6
potent phosphate analogues (7, 8). The purinetrione
bearing an alkyl phosphate (9) was tested and found
1
7
20
to be an inhibitor of lumazine synthase.
sis of the corresponding cyclic sulfate (Scheme 2). The
2
0
The structural modification of known inhibitors rep-
resents a promising approach in the search for new gly-
cosidase inhibitors. It is of interest to synthesize
phosphate derivatives of known glycosidase inhibitors
containing other internal anions such as sulfates and
carboxylates and study their inhibitory activities. Hence,
we designed the phosphate analogues (10–12) of sala-
diol (16) was prepared in three steps from D-glucose.
Subsequent treatment of the diol (16) with either p-
nitrophenyl phosphorodichloridate or phenyl dichloro-
phosphate gave the D-cyclic phosphates 14 and 15,
respectively.
1,4-Anhydro-2,3,5-tri-O-benzyl-1-thio-D-arabinitol (17)
was prepared from commercially available L-xylose in

1936
R. G. Bhat et al. / Carbohydrate Research 342 (2007) 1934–1942
OH
OH
OPO H
O
S
S
3
P
HO
RO
+
O
RO
O
O
Ph
O
HO
OH
RO
OR
Cyclic Phosphate (13)
R = Protecting group
1
0
Scheme 1.
D-glucose
Ref. 20
(
3 steps)
Ph
O
OH
OH
O
P
O
O
Cl
Cl
P
O N
2
O
Cl
PhO
Cl
1
6
Pyr.
Et N
3
Ph
O
Ph
O
O
O
O
OPh
O
O
NO2
O
O
P
P
O
O
6
1% (mixture of diasteriomers)
50% (mixture of diasteriomers)
1
5
14
Scheme 2.
3
1
six steps using a procedure developed by Satoh et al.
not as facile as compared to their sulfate counterparts;
thus, the syntheses using cyclic phosphates were
abandoned.
Various conditions were tried to couple the D-cyclic
phosphates (14, 15) with either compound 17 or tetra-
hydrothiophene, but the reactions did not proceed as
planned (Table 1). However, the reaction did proceed
with thiourea as a nucleophile. The principal difference
between thiourea and the cyclic thioether appears to be
the greater steric hindrance in the approach of the cyclic
thioether. In addition, the sulfur atom in thiourea has a
partial negative charge which makes it an even better
nucleophile. Hence, it appears that the opening of the
cyclic phosphates with the less nucleophilic thioether is
We focused, therefore, on the synthesis of simpler
prototype compounds related to salacinol (1) but con-
taining a phosphate anion. Compounds 11 and 12 were
synthesized by the Mitsunobu reaction of N- and S-
hydroxyalkyl derivatives of heteroalditols with dibenzyl
phosphate, followed by catalytic hydrogenolysis. The
required dimesylated compound (18) was prepared from
commercially available L-xylose in five steps following
3
2
3
1
a literature procedure. N-(2-Hydroxyethyl)-2,3,5-tri-
Table 1. Attempts at coupling 17 or tetrahydrothiophene with the D-cyclic phosphates (14, 15)
Reactant 1
Reactant 2
Reaction conditions
Result
a
1
7
7
15
15
15
15
14
14
HFIP /K
2
CO
3
, 2 days, 75 ꢁC
Reaction did not proceed (starting materials were recovered)
1
Acetone, 2 days, 75 ꢁC
a
Tetrahydrothiophene
Tetrahydrothiophene
1
HFIP /K
2 3
CO , 4 days, 80 ꢁC
K
2
CO , 5 days, 80 ꢁC
3
a
7
7
HFIP /K
DMSO–CH
CO , 5 days, 100 ꢁC
CH CN/K CO , 7 days, 105 ꢁC
HFIP /K CO , 4 days, 80 ꢁC
DMF, 3 days, 98 ꢁC
2
CO
3
, 4 days, 85 ꢁC
1
3
CN (2:3)
K
2
3
1
7
14
14
14
3
2
3
a
Tetrahydrothiophene
Tetrahydrothiophene
2
3
a
HFIP refers to 1,1,1,3,3,3-hexafluoroisopropanol.

R. G. Bhat et al. / Carbohydrate Research 342 (2007) 1934–1942
1937
OH
OMs
OMs
OH
N
Ph P/DIAD
3
BnO
BnO
H2N
BnO
(
BnO) P(O)OH
2
CH CN, reflux
OBn
3
BnO
OBn
THF, 0 ºC - rt
8
3%
1
8
19
76%
O
P
O
P
O
OH
O
OBn
OH
OBn
N
N
1
0% Pd/C, H2
MeOH, 80 psi
9%
HO
BnO
BnO
HO
OH
OBn
7
1
1
20
Scheme 3.
O-benzyl-1,4-dideoxy-1,4-imino-D-arabinitol (19) was
The desired sulfonium salt 12 was synthesized from
synthesized in 83% yield from the reaction of dimesylate
compound 17 as shown in Scheme 5. Reaction of 17
with 3-bromo-1-propanol in 1,1,1,3,3,3-hexafluoroiso-
propanol (HFIP), followed by treatment of the reaction
mixture with silver triflate, afforded compound 24 in
60% yield. The stereochemistry at the sulfonium-ion
center was assigned with the aid of a NOESY experi-
ment which showed a correlation between H-4 and H-
(
18) with ethanolamine (Scheme 3). Treatment of com-
pound 19 with dibenzyl phosphate under Mitsunobu
reaction conditions furnished the corresponding phos-
phorylated derivative of N-(2-hydroxyethyl)-1,4-dide-
oxy-1,4-imino-D-arabinitol (20) in 76% yield. Finally,
hydrogenolysis of 20 using 10% Pd on carbon afforded
the desired product (11) as a white solid in 79% yield.
Initially, we anticipated that sulfonium salt 22 could
be synthesized directly by coupling compound 17 and
0
1 , suggesting that the propanol side chain and the C-4
substituent were trans to each other. Preparation of
phosphate 25 from 24 was critical from a synthetic point
of view as the mode of addition of reactants was found
to be a determining factor for the efficient synthesis of
25. Initially, a mixture of triphenylphosphine, com-
pound 24, and dibenzyl phosphate in anhydrous THF
at 0 ꢁC was stirred for 5–10 min, and then diisopropyl-
azodicarboxylate (DIAD) was added dropwise, in accor-
3
-bromopropyldibenzyl phosphate (21). After numerous
attempts under a variety of reaction conditions, we
failed to synthesize compound 22 directly since the
nucleophilic sulfur atom attacked the benzyl group on
the phosphate moiety to furnish the salt, 1,4-anhydro-
2
,3,5-tri-O-benzyl-1-[benzyl-episulfoniumylidene]-D-ara-
3
3
binitol triflate (23), as shown in Scheme 4.
dance with the literature procedure. However, the
OBn
P
O
OBn
TfO
S
O
BnO
BnO
OBn
2
2
S
OBn
OBn
BnO
+
Br
O
P
BnO
OBn
O
1
7
21
S
OTf
BnO
BnO
OBn
2
3
Scheme 4.

1938
R. G. Bhat et al. / Carbohydrate Research 342 (2007) 1934–1942
OH
TfO
Ph P/DIAD
3
S
i. Br(CH2)3OH
HFIP
S
(
BnO) P(O)OH
2
BnO
BnO
ii. AgOTf, CH Cl ,
THF, 0 °C - rt,
2
2
BnO
OBn
BnO
OBn
6
0%
8
1%
1
7
24
OH
OH
OBn
OBn
1
'
3' O
O
2
'
P
P
TfO
1
0% Pd/C
TfO
S
O
S
O
5
4
1
2
H , 80 psi
HO
2
3
BnO
MeOH
HO
OH
BnO
OBn
65%
12
25
Scheme 5.
3
6
desired product was not obtained. We suspected that tri-
phenylphosphine must have reacted with the sulfonium
salt. A second set of conditions that involved initial stir-
ring of triphenylphosphine and DIAD in dry THF for
was replaced by a phosphate group. The same group
later looked at a similar replacement in inhibitors of
myo-inositol-1-phosphatase and suggested that the key
differences between sulfate and phosphate groups are
that they are present as mono-anionic and di-anionic
1
0 min followed by the gradual addition of dibenzyl
3
7
phosphate, then compound 24, resulted in the consump-
tion of 50% of the substrate. Finally, stirring triphenyl-
phosphine and DIAD in dry THF at 0 ꢁC for a short
time (2–3 min), followed by the immediate addition of
dibenzyl phosphate, followed, in turn, by the addition
of a solution of 24 in anhydrous THF dropwise at
entities, respectively. The difference in the anionic
charges of sulfate and phosphate groups might be one
of the factors accounting for the low activities of com-
pounds 11 and 12 against MGA.
Recently, Muraoka et al. reported the synthesis and
inhibitory activity of compound 26, which lacks a
hydroxyl group at C-2 and a hydroxymethyl moiety
at C-3 , against intestinal a-glucosidases. Compound
26 was not an effective inhibitor of this enzyme. This
0
0
6
ꢁC, with subsequent stirring at room temperature for
h furnished the desired product 25 in 81% yield. De-
0
38
benzylation was carried out by hydrogenolysis using
0% Pd on carbon to obtain the thioalditol phosphate
12) as a viscous oil in 65% yield. The stereochemistry
0
1
(
suggests that both the hydroxyl group at C-2 and the
hydroxymethyl group at C-3 are essential for a-glucosi-
0
at the sulfonium center was assigned with the aid of a
NOESY experiment, which showed a correlation
between H-4 and H-1 , suggesting that the side chain
dase inhibitory activity. In addition, Tanabe et al. con-
cluded that the internal sulfate counterion is not
necessary for activity since the de-O-sulfonated ana-
logues 27 and 28 showed almost equal inhibitory acti-
vities against intestinal a-glucosidases when compared
to salacinol (1). Thus, it is likely that an exact phos-
phate analogue of salacinol (1) will be required in order
to provide a definite answer as to whether the presence
of a phosphate moiety or the lack of the hydroxyl group
0
and the C-4 substituent were trans to each other.
As a final point of interest, we comment on the screen-
ing of the compounds synthesized in this study against
recombinant human maltase glucoamylase (MGA), a
critical intestinal glucosidase involved in the processing
of oligosaccharides of glucose into glucose itself. Com-
pounds 11 and 12 were not effective inhibitors of
MGA, whereas salacinol 1 inhibited this enzyme with
3
9
0
0
at C-2 and hydroxymethyl moiety at C-3 is the cause of
the low activities of compounds 11 and 12 against
MGA.
3
4
a K value of 0.19 lM. Hence, it appears that hydroxyl
i
and/or sulfate substituents present on the side chain are
critical for activity.
OH
It is of interest to note that when a phosphate group
was replaced by a sulfate in myo-inositol 1,4,5-trisphos-
phate, the resulting compound had no affinity for the
OH
X
S
OSO3
S
OH
HO
HO
3
5
IP receptor or could not induce platelet aggregation.
3
A similar loss of biological activity was observed when
an essential sulfate group in the antithrombin III bind-
ing motif of heparin at the non-reducing glucosamine
HO
OH
HO
OH
27: X=CH OSO
3
3
2
6
2
8: X=Cl

R. G. Bhat et al. / Carbohydrate Research 342 (2007) 1934–1942
1939
1. Experimental
J1,P = 8.7 Hz, C-1), 68.40 (d, 1C, J4,P = 12.2 Hz, C-4).
Anal. Calcd for C H NO P: C, 51.92; H, 4.10. Found:
1
7
16
8
1
.1. General methods
C, 51.84; H, 4.30.
Compound 14b: H NMR (CDCl ): d 8.29–8.26 (2H,
1
3
Optical rotations were measured at 23 ꢁC and reported
m, NO -Armeta), 7.46–7.36 (7H, m, Ar), 5.60 (1H, s,
2
ꢀ
1
ꢀ1
3
1
13
in deg dm
g
cm . H and C NMR spectra were
CHPh), 4.65 (1H, dddd, J3,P = 0.9 Hz, J3,4eq = 5.1 Hz,
J3,4ax = J2,3 = 10.1 Hz, H-3), 4.57–4.47 (2H, m, H-1ax,
H-1eq), 4.51 (1H, dd, J4ax,4eq = 9.9 Hz, H-4eq), 4.20
recorded with frequencies of 500 and 125 MHz, respec-
tively. All assignments were confirmed with the aid of
1
1
1
13
two-dimensional H,
H
(gCOSY) and H,
C
(1H, ddd, J1ax,2 = 9.7 Hz, J
(1H, dd, H-4ax); C NMR (CDCl ): d 129.71–120.10
= 2.1 Hz H-2), 3.86
1eq,2
1
3
(
gHMQC) experiments using standard Varian pulse
3
programs. Processing of the spectra was performed with
MestRec software. 1D-NOESY experiments were re-
corded at 295 K on a 500 MHz spectrometer. For each
(12C, Ar), 102.22 (CHPh), 72.65 (d, 1C, J2,P = 7.9 Hz,
C-2), 70.51 (d, 1C, J3,P = 4.8 Hz, C-3), 69.34 (d, 1C,
J1,P = 7.7 Hz, C-1), 68.59 (d, 1C, J4,P = 11.8 Hz, C-4).
For the mixture of diastereomers: Anal. Calcd for
C H NO P: C, 51.92; H, 4.10. Found: C, 51.85; H,
1
D-NOESY spectrum, 512 scans were acquired with a
Q3 Gaussian Cascade pulse. A mixing time of 500 ms
was used in the 1D-NOESY experiments. Matrix-
assisted laser desorption ionization time-of-flight
1
7
16
8
4.25.
(
MALDI-TOF) mass spectra were obtained using 2,5-
1.3. Phenyl-2,4-O-benzylidene-D-erythritol-1,3-cyclic
phosphate (15)
dihydroxybenzoic acid as a matrix. Analytical thin-layer
chromatography (TLC) was performed on aluminum
plates precoated with silica gel 60F-254 as the adsor-
bent. The developed plates were air-dried, exposed to
UV light and/or sprayed with a solution containing
Phenyl dichlorophosphate (0.4 mL, 2.7 mmol) was
added to a stirred solution of diol 16 (0.5 g, 2.4 mmol)
in pyridine (5 mL). The mixture was stirred at ambient
temperature for 18 h. The reaction was quenched with
1
% Ce(SO ) and 1.5% molybdic acid in 10% aqueous
4 2
H SO , and heated. Column chromatography was per-
H O (1.5 mL) and the solvent was removed under high
2
2
4
formed with silica gel 60 (230–400 mesh).
vacuum. The product was purified by flash chromato-
graphy (hexanes–EtOAc, 2:1) to give 15 as a mixture
1.2. p-Nitrophenyl-2,4-O-benzylidene-D-erythritol-1,3-
cyclic phosphate (14)
of diastereomers (0.4 g, 50%).
Compound 15a: H NMR (CDCl ): d 7.55–7.20 (10H,
1
3
m, Ph), 5.65 (1H, s, CHPh), 4.58–4.42 (3H, m, H-1eq,
H-3, H-4eq), 4.43 (1H, dd, J1ax,2 = J1ax,1eq = 10.4 Hz,
H-1ax), 4.17 (1H, ddd, J1eq,2 = 5.1 Hz, J2,3 = 9.8 Hz,
A solution of diol (16) (1.0 g, 4.8 mmol) and Et N
3
(
2.5 mL, 4 equiv) in dry CH Cl (15 mL) was added
2 2
dropwise to a solution of p-nitrophenyl phosphorodi-
H-2), 3.93 (1H, dd, J
C NMR (CDCl ): d 135.98–128.75 (12C, Ph), 102.28
3
= J3,4ax = 10.4 Hz, H-4ax);
ax,4eq
4
1
3
chloridate (1.3 g, 1.1 equiv) in dry CH Cl (15 mL) at
2
2
0
9
ꢁC under N . The mixture was stirred at 0 ꢁC for
(CHPh), 72.90 (d, 1C, J2,P = 7.2 Hz, C-2), 72.15 (d,
1C, J3,P = 6.4 Hz, C-3), 69.38 (d, 1C, J1,P = 8.8 Hz,
2
0 min and then diluted with CH Cl (20 mL). The
2
2
mixture was washed with H O (3 · 10 mL), dried over
C-1), 68.58 (d, 1C, J = 12.1 Hz, C-4).
2
4,P
1
anhydrous Na SO , and concentrated. The product
Compound 15b: H NMR (CDCl ): d 7.55–7.20 (10H,
2
4
3
was purified by flash chromatography (CH Cl –MeOH,
m, Ph), 5.60 (1H, s, CHPh), 4.60 (1H, dddd,
J3,P = 0.9 Hz, J3,4eq = 5.1 Hz, J3,4ax = J2,3 = 9.7 Hz, H-
3) 4.58–4.47 (3H, m, H-1ax, H-1eq, H-4eq), 4.08 (1H,
ddd, J1eq,2 = 6.1 Hz, J1ax,2 = 9.8 Hz, H-2), 3.80 (1H,
2
2
3
0:1 + 0.1% Et N) to give 14 as a mixture of diastereo-
3
mers. Most of the more polar isomer 14a was selectively
recrystallized from hexanes and EtOAc giving faint pink
crystals (0.63 g, 34%). The filtrate was recrystallized
using hexanes and EtOAc to afford a mixture of 14a
1
3
dd, J4ax,4eq = 10.5 Hz, H-4ax); C NMR (CDCl ): d
3
130.01–128.75 (12C, Ph), 102.14 (CHPh), 73.03 (d, 1C,
J2,P = 6.4 Hz, C-2), 70.01 (d, 1C, J3,P = 5.6 Hz, C-3),
68.92 (d, 1C, J1,P = 7.2 Hz, C-1), 68.70 (d, 1C,
J4,P = 11.3 Hz, C-4).
and 14b as a white solid (0.51 g, 27%).
Compound 14a: [a]D +60.0 (c 0.5, CHCl ); H NMR
1
3
(
(
CDCl ): d 8.30–8.26 (2H, m, NO -Armeta), 7.35–7.48
7H, m, Ar), 5.64 (1H, s, CHPh), 4.58 (1H, ddd,
3
2
For the mixture of diastereomers: Anal. Calcd
for C H O P: C, 58.63; H, 4.92. Found: C, 58.31; H,
J1eq,2 = 5.0 Hz, J1ax,1eq = 10.3 Hz, J1eq,P = 23.7 Hz, H-
eq), 4.52 (1H, dddd, J3,P = 0.9 Hz, J3,4eq = 5.1 Hz,
J3,4ax = J2,3 = 9.2 Hz, H-3), 4.45 (1H, dd, J4ax,4eq
0.7 Hz, H-4eq), 4.43 (1H, ddd, J1ax,P = 1.0 Hz,
J1ax,2 = 10.7 Hz H-1ax), 4.21 (1H, ddd, H-2), 3.56 (1H,
17
17
6
1
5.01.
=
1
1.4. N-(2-Hydroxyethyl)-2,3,5-tri-O-benzyl-1,4-dideoxy-
1,4-imino-D-arabinitol (19)
1
3
dd, H-4ax); C NMR (CDCl ): d 135.71–120.10 (12C,
3
Ar), 102.35 (CHPh), 72.68 (d, 1C, J2,P = 6.3 Hz, C-2),
Ethanolamine (2.1 mL, 34.6 mmol) was added to a stir-
7
2.46 (d, 1C, J3,P = 6.9 Hz, C-3), 69.80 (d, 1C,
red solution of dimesylate 18 (2.0 g, 3.5 mmol) in aceto-

1940
R. G. Bhat et al. / Carbohydrate Research 342 (2007) 1934–1942
nitrile (15 mL) and the mixture was heated to reflux for
0 h under N . The reaction mixture was concentrated
1.6. N-(2-Phosphorylethyl)-1,4-dideoxy-1,4-imino-D-
arabinitol (11)
2
2
in vacuum and the residue was dissolved in EtOAc
30 mL) and washed with water (3 · 15 mL). The
organic layer was dried over anhydrous Na SO and
(
A solution of compound 20 (1.2 g, 1.7 mmol) in MeOH
(5 mL) containing 10% palladium on carbon (0.5 g) was
stirred under 80 psi of hydrogen at room temperature
for 10 h. The mixture was diluted with 15% aqueous
MeOH, filtered, and the solvent was removed in vacuum
to give a white solid. The compound was purified further
by column chromatography (H O–MeOH–H O, 10:3:1)
2
4
concentrated. Purification by column chromatography
EtOAc–hexanes, 7:3) gave compound 19 as a colorless
(
2
D
3
1
oil (1.28 g, 83%). ½aꢁ +1.0 (c 1.1, MeOH); H NMR
(
CDCl ): d 7.31 (15H, m, 3 · Ph), 4.50 (6H, ddd,
3
J = 12.0 Hz, 3 · CH Ph), 3.99 (1H, dd, J = 1.6 Hz,
J
2
2,3
2
2
= 3.7 Hz, H-2), 3.89 (1H, br d, J = 2.5 Hz, H-3),
.65–3.52 (2 · 2H, m, H-2 and 5-H), 3.28 (1H, br d,
to give 11 as a white solid (0.35 g, 79%). Mp 178–180 ꢁC;
2
,1
3,4
0
23
D
1
3
½aꢁ +2.1 (c 1.1, H O); H NMR (D O): d 4.36 (1H, td,
2
2
0
J1,2 = 3.6 Hz, H-1a), 3.10–3.06 (1H, m, H-1 a), 2.91
J
= 2.0 Hz, J2,1 = 4.3 Hz, H-2), 4.17 (2H, td,
¼ 4:8 Hz, H-2 ), 4.12 (1H, t, J = 2.7 Hz, H-3),
3,4
2,3
0
(
1H, br d, J4,5 = 3.6 Hz, H-4), 2.69 (1H, dd,
J
0
0
2 ;1
J1,2 = 5.3 Hz, J1a,1b = 10.6 Hz, H-1b), 2.61 (1H, d,
3.99 (2H, dq, J = 6.1 Hz, J
(1H, dd,
J1,2 = 5.0 Hz, H-1a) 3.62 (1H, dd, J = 5.0 Hz, H-1b),
= 12.6 Hz, H-5), 3.82
5a,5b
5,4
0
13
0
J1
1
3
(
(
0
0
¼ 12:6 Hz, H-1 b);
C
NMR (CDCl3):
d
J
0 0
1 ;2
¼ 4:6 Hz, H-1 a), 3.78 (1H, dd,
a;1 b
38.39, 138.32, 138.25 (3 · Cipso), 128.6–127.8 (15C,
1,2
· 5, Ph), 85.2 (C-3), 82.0 (C-2), 71.68, 71.39, 71.04
3.60 (1H, m, H-4), 3.51 (td, 1H, J
0
0
;2
¼ 4:5 Hz,
1
0
0
13
3 · CH Ph), 69.3 (C-4), 60.0 (C-2 ), 57.5 (C-1), 57.3
J
0
0
¼ 13:8 Hz, H-1 b); C NMR (D O): d 75.8 (C-
2
1 a;1 b
2
0
+
2
0
C-1 ). MALDI-TOF-MS: m/z 448.04 [M+H] . Anal.
3), 75.6 (C-4), 73.8 (C-2), 60.2 (d, J
2 ), 59.4 (C-1), 58.3 (C-5), 57.01 (d, J
,P = 4.7 Hz, C-
Cꢀ2
0
3
0
Calcd for C H NO : C, 75.14; H, 7.43; N, 3.13.
,P = 3.9 Hz,
Cꢀ1
C-1 ). MALDI-TOF-MS: m/z 258.2 [M+H] . Anal.
2
8
33
4
0
+
Found: C, 75.44; H, 7.10; N, 3.40.
Calcd for C H NO P: C, 32.69; H, 6.27; N, 5.45.
7
16
7
Found: C, 32.35; H, 6.16; N, 5.20.
1.5. N-(2-Dibenzylphosphorylethyl)-2,3,5-tri-O-benzyl-
1,4-dideoxy-1,4-imino-D-arabinitol (20)
1.7. 1,4-Anhydro-2,3,5-tri-O-benzyl-1-[3-hydroxypropyl-
(R)-episulfoniumylidene]-D-arabinitol triflate (24)
To a well-stirred mixture of triphenylphosphine (0.9 g,
.4 mmol) and diisopropyl azodicarboxylate (DIAD)
0.65 mL, 3.4 mmol) in anhydrous THF (5 mL) at
ꢁC, dibenzyl phosphate (0.94 g, 3.4 mmol) was added.
After stirring the reaction mixture for 5 min at 0 ꢁC, a
solution of N-hydroxyethyl-2,3,5-tri-O-benzyl-1,4-di-
deoxy-1,4-imino-D-arabinitol (19) (1.0 g, 2.2 mmol) in
3
(
0
The benzyl protected 4-thio-D-arabinitol (17) (2.0 g,
4.8 mmol)
and
3-bromo-1-propanol
(0.43 mL,
4.8 mmol) were dissolved in 1,1,1,3,3,3-hexafluoroiso-
propanol (HFIP) (2 mL) and the mixture was stirred
in a sealed tube at 92 ꢁC for 26 h (prolonged heating
of the reaction mixture resulted in the formation of side
products). HFIP was removed and the residue was dis-
solved in CH Cl . Silver triflate (1.2 g, 4.8 mmol) was
THF (5 mL) was added dropwise at 0 ꢁC under N
2
and the reaction mixture was stirred for 3 h at room
temperature. After completion of the reaction, THF
was removed in vacuum and the residue was purified
by column chromatography (EtOAc–hexanes, 4:6) to
2
2
added and the mixture was stirred at ambient tempera-
ture for 2 h. The solvent was removed and the resulting
residue was purified by column chromatography
2
D
3
give compound 20 as a colorless oil (1.2 g, 76%). ½aꢁ
1
+
26.0 (c 1.0, MeOH); H NMR (CDCl ): d 7.25–7.17
(CHCl –MeOH, 20:1) to give compound 24 as a color-
3
3
2
D
3
1
(
4
2
25H, m, 5 · Ph), 5.07–4.92 (m, 4H, 2 · POCH Ph),
less oil (1.82 g, 60%). ½aꢁ ꢀ6.0 (c 1.0, MeOH); H
2
.42–4.29 (6H, m, 3 · OCH Ph), 4.10–3.96 (2H, m, H-
NMR (CDCl ): d 7.36–7.17 (15 H, m, 3 · Ph), 4.60–
2
3
0
), 3.82 (1H, d, J = 5.0 Hz, H-2), 3.73 (1H, d,
4.47 (6H, m, 2 · CH Ph, HaCHPh, H-2), 4.40 (1H, d,
2
,3
2
J3,4 = 3.7 Hz, H-3), 3.42 (2H, ddd, J5,4 = 6.0 Hz,
J5a,5b = 9.7 Hz, H-5), 3.16 (1H, d, J1a,1b = 10.5 Hz,
J
= 11.5 Hz, H CHPh 5.2 Hz, H-2), 4.22–4.18 (2H,
A,B b
m, H-2, H-1a), 3.97–3.90 (2H, ddd, J5,4 = 5.0 Hz,
H-1a), 3.07 (1H, td, J1
0
0
¼ 6:3 Hz, J
0
0
¼ 12:9 Hz,
J
= 10 Hz 5-H), 3.77 (1H, dd, J
0
0
0
¼ 6:0 Hz, H-
;2
1 a;1 b
5a,5b
0
3 ;2
0
0
H-1 a), 2.72 (1H, dd, J = 5.7 Hz, H-4), 2.57 (2H, m,
3a ), 3.72–3.65 (3H, m, H-3 b, H-4, H-1 a), 3.622 (1H,
dd, J1,2 = 3.5 Hz, J1a,1b = 13.0 Hz, 1-Hb) 3.48 (1H, td,
4
,5
0
13
H-1b, H-1 b); C NMR (CDCl ): d 138.53, 138.36,
38.31, 136.22, 138.18 (5 · Cipso), 128.7–127.7 (5 · Ph),
5.3 (C-3), 81.8 (C-2), 73.4–71.5 (CH Ph), 69.44 (d,
= 5.4 Hz, POCH Ph), 68.1 (d, J = 5.7 Hz, C-
), 58.0 (C-1), 54.84 (d, J = 6.6 Hz, C-1 ). MAL-
DI-TOF-MS: m/z 708.3 [M+H] . Anal. Calcd for
C H NO P: C, 71.27; H, 6.55; N, 1.98. Found: C,
3
0
1
8
J
2
J
0
0
¼ 6:5 Hz, J
0 0
¼ 12:5 Hz, H-1 b), 2.12–1.99 (2H,
1 2
1 a;1 b
0
13
m, H-2 ); C NMR (CDCl ): d 136.94, 136.24, 136.11
2
3
2
(3 · Cipso), 129.0–128.1 (3 · Ph), 128.3-119.6 (CF tri-
flate), 83.0 (C-3), 82.5 (C-2), 73.9 (CH Ph), 72.6
(CH Ph), 72.1 (CH Ph), 67.1 (C-3 ), 66.5 (C-5), 59.5
(C-4), 47.5 (C-1), 44.3 (C-1 ), 28.7 (C-2 ). MALDI-
C,P
0
2
C,P
3
3
0
C,P
2
+
0
2
2
0
0
4
2
46
7
+
7
1.09; H, 6.50; N, 1.92.
TOF-MS: m/z 479.03 [MꢀOtf] . Anal. Calcd for

R. G. Bhat et al. / Carbohydrate Research 342 (2007) 1934–1942
1941
0
3
0
C H F O S : C, 57.31; H, 5.61. Found: C, 57.23; H,
26.4 (d, C-2 ,
J
,P = 4.5 Hz). MALDI-TOF-MS:
Cꢀ2
+
3
0
35
3
7
2
5
.59.
m/z 289.2 [MꢀOtf] . Anal. Calcd for C H F O PS :
9
18
3
10
2
C, 24.66; H, 4.14. Found: C, 24.32; H, 3.98.
1.8. 1,4-Anhydro-2,3,5-tri-O-benzyl-1-[3-dibenzyl-
phosphorylpropyl-(R)-episulfoniumylidene]-D-arabinitol
triflate (25)
Acknowledgements
A mixture of triphenylphosphine (1.5 g, 2.4 mmol)
and diisopropyl azodicarboxylate (DIAD, 0.46 mL,
We thank Dr. B. D. Johnston for helpful discussions.
We also thank the Natural Sciences and Engineering
Research Council of Canada for financial support and
the Michael Smith Foundation for Health Research
for a trainee fellowship (to N.S.K.). We are also grateful
to one of the reviewers for helpful comments.
5.2 mmol) in anhydrous THF (5 mL) was stirred at
0 ꢁC for 3 min and then dibenzyl phosphate (1.46 g,
5.2 mmol) was added. After stirring the reaction mixture
for an additional 4 min at 0 ꢁC, a solution of 24 (1.48 g,
2
.4 mmol) in THF (5 mL) was added dropwise at 0 ꢁC
under N , and the reaction mixture was stirred for 3 h
2
at room temperature. After completion of the reaction,
THF was removed in vacuum and the residue was puri-
References
fied by column chromatography (CHCl –MeOH, 50:1)
3
to give compound 25 as a colorless oil (1.7 g, 81%).
1
. Dwek, R. A. Chem. Rev. 1996, 96, 683–720.
2
3
1
½
aꢁ +3.0 (c 1.0, MeOH); H NMR (CDCl ): d 7.36–
D
3
2. Varki, A.; Cummings, R.; Esko, J.; Freeze, H.; Hart, G.;
Marth, J. Essentials of Glycobiology; Cold Spring Harbor
Laboratory Press: NY, 1999.
7
4
.17 (15H, m, 3 · Ph), 4.94–4.89 (4H, m, 2 · POCH Ph),
2
.60–4.47 (6H, m, 2 · CH Ph, 1 · H CHPh, H-2), 4.31
2
a
3
4
. Jacob, G. S. Curr. Opin. Struct. Biol. 1995, 5, 605–611.
. Holman, R. R.; Cull, C. A.; Turner, R. C. Diabetes Care
(
1H, d, JA,B = 12.0 Hz, H CHPh), 4.18 (1H, d,
b
J1a,1b = 16.0 Hz, H-1a), 4.06 (1H, br s, H-3), 4.03–3.98
1
999, 22, 960–964.
0
0
(
3
1H, m, H-3 a), 3.95 (3H, m, H3 b, H-5a, H-4), 3.65–
5
. Asano, N. Glycobiology 2003, 13, 93–104.
0
0
.50 (2H, m, H-5b, H-1 a), 3.48–3.30 (2H, m, H-1 b,
6. Vasella, A.; Davies, G. J.; Bohm, M. Curr. Opin. Chem.
Biol. 2002, 6, 619–629.
0
13
H-1b), 2.12–1.99 (2H, m, H-2 ); C NMR (CDCl ): d
3
7
. Heightman, T. D.; Vasella, A. T. Angew. Chem., Int. Ed.
999, 38, 750–770.
1
1
36.94, 136.24, 136.11 (3 · Cipso), 129.0–128.1 (3 · Ph),
1
28.3–119.6 (CF triflate), 83.0 (C-3), 82.5 (C-2), 73.9
3
8
. de Melo, E. B.; Gomes, A. D.; Carvalho, I. Tetrahedron
2006, 62, 10277–10302.
(
CH Ph), 72.6 (CH Ph), 72.1 (CH Ph), 69.93 (d,
2 2 2
JC,P = 5.75 Hz, POCH Ph) 66.9 (C-4), 66.5 (C-5),
9. Elbein, A. D. Annu. Rev. Biochem. 1987, 56, 497–534.
0. Hughes, A. B.; Rudge, A. J. Nat. Prod. Rep. 1994, 11, 135–
162.
1. Stutz, A. E. Iminosugars as Glycosidase Inhibitors:Nojiri-
mycin and Beyond; Wiley-VCH, Weinheim: NY, 1999.
2. Yoshikawa, M.; Murakami, T.; Shimada, H.; Matsuda,
H.; Yamahara, J.; Tanabe, G.; Muraoka, O. Tetrahedron
Lett. 1997, 38, 8367–8370.
2
2
0 0
5.07 (d, J = 5.8 Hz, C-3 ), 47.5 (C-1) 42.3 (C-1 ),
C,P
1
1
1
6
2
7
5
3
0
6.9 (d, J = 6.1 Hz, C-2 ). MALDI-TOF-MS: m/z
C,P
+
39.08 [MꢀOtf] . Anal. Calcd for C H F O PS : C,
4
4
48
3
10
2
9.45; H, 5.44. Found: C, 59.11; H, 5.43.
1.9. 1,4-Anhydro-1-[3-phosphorylpropyl-(R)-episulfon-
iumylidene]-D-arabinitol triflate (12)
1
1
3. Yoshikawa, M.; Murakami, T.; Yashiro, K.; Matsuda, H.
Chem. Pharm. Bull. 1998, 46, 1339–1340.
4. Kafarski, P.; Lejczak, B. Phosphorus, Sulfur, Silicon Relat.
Elem. 1991, 63, 193–215.
A solution of compound 25 (1.0 g, 1.1 mmol) in MeOH
(
5 mL) containing 10% palladium on carbon (0.5 g) was
15. Winum, J. Y.; Innocenti, A.; Gagnard, V.; Montero, J. L.;
Scozzafava, A.; Vullo, D.; Supuran, C. T. Bioorg. Med.
Chem. Lett. 2005, 15, 1683–1686.
stirred under 80 psi of hydrogen at room temperature
for 16 h. The mixture was filtered, and the solvent was
removed in vacuum to give the thio-alditol phosphate
derivative as a viscous oil. The compound was purified
1
1
6. Tadeusiak, E. J. Bioorg. Chem. 2004, 32, 473–482.
7. Cushman, M.; Sambaiah, T.; Jin, G. Y.; Illarionov, B.;
Fischer, M.; Bacher, A. J. Org. Chem. 2004, 69, 601–612.
further by column chromatography (H O–MeOH–
18. Ghavami, A.; Johnston, B. D.; Jensen, M. T.; Svensson,
B.; Pinto, B. M. J. Am. Chem. Soc. 2001, 123, 6268–6271.
19. Ghavami, A.; Johnston, B. D.; Maddess, M. D.; China-
poo, S. M.; Jensen, M. T.; Svensson, B.; Pinto, B. M. Can.
J. Chem. 2002, 80, 937–942.
2
23
H O, 10:3:1) to give 12 (0.25 g, 65%). ½aꢁ ꢀ2.0 (c 1.1,
2
D
1
MeOH);
H
NMR (D O):
d
4.70 (1H, ddd,
2
J2,3 = 3.5 Hz, J2,1b = 7.0 Hz, H-2), 4.39 (1H, dd,
J3,4 = 3.0 Hz, H-3), 4.09 (1H, dd, J = 4.5 Hz, H-5a),
5,4
20. Ghavami, A.; Johnston, B. D.; Pinto, B. M. J. Org. Chem.
2001, 66, 2312–2317.
0
4
.03 (2H, ddd, J3
0
0
¼ 6:0 Hz, H-3 ), 3.97–3.94 (1H, m,
;2
21. Ghavami, A.; Sadalapure, K. S.; Johnston, B. D.; Lobera,
M.; Snider, B. B.; Pinto, B. M. Synlett 2003, 1259–1262.
22. Johnston, B. D.; Ghavami, A.; Jensen, M. T.; Svensson,
B.; Pinto, B. M. J. Am. Chem. Soc. 2002, 124, 8245–8250.
23. Kumar, N. S.; Pinto, B. M. Carbohydr. Res. 2005, 340,
2612–2619.
H-4), 3.87 (1H, dd, J = 4.4 Hz, H-5b), 3.89–3.76
2H, m, H-1), 3.67–3.56 (2H, m, H-3 ), 2.25–2.16 (2H,
m, H-2 ); C NMR (D O): d 119.7 (q, J = OCF ),
7.7 (C-3), 77.0 (C-2), 69.8 (C-4), 63.5 (d, C-3
5
,4
0
(
0
13
2
CF
3
0
7
2
0
0
J
,P = 5.0 Hz), 59.3 (C-5), 46.1 (C-1), 42.3 (C-1 ),
Cꢀ3

1942
R. G. Bhat et al. / Carbohydrate Research 342 (2007) 1934–1942
2
4. Szczepina, M. G.; Johnston, B. D.; Yuan, Y.; Svensson,
B.; Pinto, B. M. J. Am. Chem. Soc. 2004, 126, 12458–
2469.
5. Nasi, R.; Pinto, B. M. Carbohydr. Res. 2006, 341, 2305–
311.
6. Nasi, R.; Sim, L.; Rose, D. R.; Pinto, B. M. J. Org. Chem.
007, 72, 180–186.
7. Johnston, B. D.; Jensen, H. H.; Pinto, B. M. J. Org. Chem.
006, 71, 1111–1118.
H.; Lam, M. H.; Luell, S.; Macintyre, D. E.; Meurer, R.;
Roberts, C. D.; Sahoo, S. P.; Wu, M. S. J. Med. Chem.
1992, 35, 3474–3482.
1
2
2
2
34. Rossi, E. J.; Sim, L.; Kuntz, D. A.; Hahn, D.; Johnston, B.
D.; Ghavami, A.; Szczepina, M. G.; Kumar, N. S.;
Sterchi, E. E.; Nichols, B. L.; Pinto, B. M.; Rose, D. R.
FEBS J. 2006, 273, 2673–2683.
35. Westerduin, P.; Willems, H. M. A.; van Boeckel, C. A. A.
Carbohydr. Res. 1992, 234, 131–140.
2
2
2
2
2
3
8. Liu, H.; Pinto, B. M. J. Org. Chem. 2005, 70, 753–755.
9. Liu, H.; Pinto, B. M. Can. J. Chem. 2006, 84, 497–505.
0. Liu, H.; Sim, L.; Rose, D. R.; Pinto, B. M. J. Org. Chem.
36. Vos, J. N.; Westerduin, P.; van Boeckel, C. A. A. Bioorg.
Med. Chem. Lett. 1991, 1, 143–146.
37. van Steijn, A. M. P.; Willems, H. A. M.; de Boer, T.;
Geurt, J. L. T.; van Boeckel, C. A. A. Bioorg. Med. Chem.
Lett. 1995, 5, 469–474.
2006, 71, 3007–3013.
3
1. Satoh, H.; Yoshimura, Y.; Sakata, S.; Miura, S.; Machida,
H.; Matsuda, A. Bioorg. Med. Chem. Lett. 1998, 8, 989–
38. Muraoka, O.; Yoshikai, K.; Takahashi, H.; Minematsu,
T.; Lu, G.; Tanabe, G.; Wang, T.; Matsuda, H.; Yoshi-
kawa, M. Bioorg. Med. Chem. 2006, 14, 500–509.
39. Tanabe, G.; Yoshikai, K.; Hatanaka, T.; Yamamoto, M.;
Shao, Y.; Minematsu, T.; Muraoka, O.; Wang, T.;
Matsuda, H.; Yoshikawa, M. Bioorg. Med. Chem. 2007,
15, 3926–3937.
992.
3
3
2. Kumar, N. S.; Pinto, B. M., unpublished.
3. Girotra, N. N.; Biftu, T.; Ponpipom, M. M.; Acton, J. J.;
Alberts, A. W.; Bach, T. N.; Ball, R. G.; Bugianesi, R. L.;
Parsons, W. H.; Chabala, J. C.; Davies, P.; Doebber, T.
W.; Doherty, J.; Graham, D. W.; Hwang, S. B.; Kuo, C.