Synthesis and evaluation of fructose analogues as inhibitors of the D-fructose transporter GLUT5

Article abstract of DOI:10.1016/S0968-0896(00)00108-5

We have examined the specificity and binding-site spatial requirements of the fructose transporter GLUT5. Interaction with a series of fructofuranosides and fructopyranosides suggests that both furanose and pyranose ring forms of D-fructose combine with G

Full text of DOI:10.1016/S0968-0896(00)00108-5

Bioorganic & Medicinal Chemistry 8 (2000) 1825±1833
Synthesis and Evaluation of Fructose Analogues as Inhibitors of
the D-Fructose Transporter GLUT5
Arnaud Tatibouet,a Jing Yang,a Christophe Morinb and Geo€rey D. Holmana,*
aDepartment of Biology and Biochemistry, University of Bath, Bath, BA2 7AY, UK
b
Â
Equipe des Marqueurs Biomedicaux, LEDSS-UMR CNRS 5616, Universite de Grenoble, 38402, St Martin d'Heres, France
Â
Á
Received 24 February 2000; accepted 31 March 2000
AbstractÐWe have examined the speci®city and binding-site spatial requirements of the fructose transporter GLUT5. Interaction
with a series of fructofuranosides and fructopyranosides suggests that both furanose and pyranose ring forms of d-fructose com-
bine with GLUT5. The epimers of d-fructose all have low anity for GLUT5 suggesting that the transporter requires all hydroxyls
to be in the fructo-con®guration. Similarly there is poor tolerance of all allyl derivatives of d-fructose except 6-O-allyl-d-fructo-
furanose. Therefore, the C-6 position o€ers the most suitable position for development of anity probes and labels for exploring
GLUT5 biochemistry. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
Despite its large-scale accessibility and the elucidation
of its structure over a century ago,9 the chemistry of
fructose10 makes the preparation of simple derivatives in
tautomeric ®xed structures, particularly in furanoid forms,
dicult. Some simple derivatives of d-fructose exist with
well de®ned ring shapes including in the pyranoid form:
the tetraacetate,11,12 the tetrabenzoate,13 the 1,2:4,5 di-O-
isopropylidene,14 and the 2,3:4,5-di-O-isopropylidene15
and in the furanoid form: the tetrabenzoate,16 and in open
chain: the tetracetates of 6-halo-6-deoxy-keto-d-fruc-
toses.17,18 The aim of our study was therefore to develop
analogues of d-fructose in which each position was
substituted and thereby determine the importance of
each of the hydroxyl groups in the binding and interac-
tion with GLUT5. In particular, we have studied the
complete series of allyl substituted d-fructoses as these
provide information on the steric constraints around the
GLUT5 binding site. In addition, the allyl compounds
described here can easily be further functionalized and
will be useful for studying other proteins that speci®cally
combine with d-fructose.
d-fructose is the second most abundant simple sugar in
nature and so is a signi®cant component of human diet-
ary sugar intake. However, unlike d-glucose, which is
absorbed across the intestine by the sodium dependent
glucose transporter (SGLT1),1 d-fructose uptake into
mammals occurs via a specialized, sodium independent,
facilitative transporter, GLUT5.2,3 This protein is also
present at high levels in the intestine, particularly the
apical surface.4 The GLUT5 protein shows sequence
homology with a family of mammalian facilitative hex-
ose transporters (GLUT isoforms 1±4) each of which
shows a distinct pattern of subcellular distribution.5,6
Isoforms 1-4 have high anity for d-glucose but low
anity for d-fructose. By contrast, GLUT5 has low a-
nity for d-glucose and high anity for d-fructose. The
distribution of GLUT5 is quite widespread and also
occurs at low levels in many tissues including the meta-
bolically important peripheral tissues of fat and
muscle.6 8 The abundance and speci®city of GLUT5
suggest that it performs a metabolically important role.
However, to date the speci®city and substrate structure
requirements of GLUT5 have not been studied with iso-
form-speci®c chemical reagents and fructose analogues.
We have therefore addressed this problem in the study
described here.
Results
Chemical synthesis
Several compounds used in the characterisation of the
speci®city of GLUT5 were synthesised by adaptation of
previously described methods. 2,5-Anhydro-d-mannitol
*Corresponding author. Tel.: +44-1255-826874; fax: +44-1225-826779;
e-mail: g.d.holman@bath.ac.uk
0968-0896/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00108-5

1826
A. Tatibouet et al. / Bioorg. Med. Chem. 8 (2000) 1825±1833
È
was prepared as described.19 1,2-O-Isopropylidene-b-d-
fructopyranose was prepared as described.20,21 Methyl
a,b-d-fructofuranosides22 were prepared according to
literature methods,23 or from 1,3,4,6-tetra-O-benzoyl-d-
fructofuranose which was methylated using methyl
iodide and sodium hydride or silver oxide to give methyl
1,3,4,6-tetra-O-benzoyl-a,b-d-fructofuranosides which
were deprotected using sodium methoxide. Methyl b-d-
fructopyranoside was prepared by methylation of 1,3,4,5-
tetra-O-benzoyl-d-fructopyranose using methyl iodide
and sodium hydride or silver oxide which gave pre-
dominately (67% yield) methyl 1,3,4,5-tetra-O-benzoyl-b-
d-fructopyranoside. Deprotection gave methyl b-d-fruc-
topyranoside with data consistent with the literature.22
Allylation of the benzoylated fructofuranoses and fruc-
topyranoses (Scheme 1) with sodium hydride and allyl
bromide gave low yields. Better results were obtained for
compounds 1 and 3 through a palladium (0) catalyzed
allylation24 with 93% and 88% yields, respectively. The
anomeric ratio obtained for the allyl-tetrabenzoyl-fruc-
tofuranoside 1 was 7:3 in favor of the a anomer.
Deprotections25 of these benzoyl fructosides to give
compounds 2 and 4 were performed with sodium meth-
oxide in dry methanol with nearly 80% yields. The a
and b alkyl fructofuranosides anomers were separated
by chromatography.
Scheme 1. Reagents and conditions: (i) allylmethylcarbonate, Pd2(dba)3,
Ph3P; (ii) MeONa, MeOH.
1-OH and 3-OH positions of d-fructose were O-allylated
starting with and 2,3:4,5 di-O-isopropylidene-b-d- fruc-
topyranose and 1,2:4,5-di-O-isopropylidene-b-d-fructo-
pyranose15 (Scheme 2). These hydroxyl groups were
allylated with allyl bromide and sodium hydride26 yield-
ing 5 and 7. Functionalization of 4-OH and 5-OH was
realised through a multi-step process (Scheme 3). A pre-
viously described procedure for benzoylation of 1,2:4,5-
di-O-isopropylidene-b-d-fructopyranose27 was used fol-
lowed by a selective hydrolysis of one O-isopropylidene
group.28 Mono-allylation conditions using palladium
catalysis were then used to substitute both the 4-OH and
5-OH positions leading to 9 with 39%, and 10 with 32%
yield. In addition, a selective additional benzoylation on
the 4-OH position of 3-O-benzoyl-1,2-O-isopropylidene-
b-d-fructopyranose was followed by the speci®c allylation
Scheme 2. Reagents and conditions: (i) allylbromide, NaH, DMF; (ii)
TFA±H2O.
Scheme 3. Reagents and conditions: (i) BzCl, pyridine; (ii) allylmethylcarbonate, Pd2(dba)3, Ph3P; (iii) MeONa, MeOH; (iv) TFA±H2O (4:1).

A. Tatibouet et al. / Bioorg. Med. Chem. 8 (2000) 1825±1833
È
1827
of the 5-OH position to give 13. Deprotection of all these
protected fructose derivatives were realised through
standard methods:benzoyl esters by transesteri®cation
with sodium methanolate and O-isopropylidene groups
were removed with tri¯uoroacetic acid±water mixtures.
At this stage, each position of d-fructose was functiona-
lized with an allyl group with overall yields starting from
d-fructose: 1-OH position 6 30%, 2-OH position 2 and 4
50%, 3-OH position 8 30%, 4-OH position 12 11%, and
the 5-OH position 15 17%.
been used to study the e€ects of insulin on the subcellular
localisation and tracking of the insulin-responsive
GLUT4 isoform.37,38 In the present study we have probed
the structure and steric factors that govern the interac-
tion of d-fructose with its speci®c transporter GLUT5
so that similar photoanity probes can ultimately be
developed for this system.
To evaluate the interaction of d-fructose analogues with
GLUT5 we have chosen a CHO cell line in which the
GLUT5 transporter is expressed at high levels.39 The
substrate for the assays is a tracer concentration (1 mM)
of [14C-U]-d-fructose. The ratio of GLUT5 mediated
uptake to the endogenous uptake was <9:1. Preliminary
studies established that the tracer uptake was linear for
10 min and that this radioactive substrate concentration
was low compared with the half maximal anity con-
stant of GLUT5 for d-fructose. Therefore, the uptake
in the presence of an analogue inhibitor is determined:
V0=V ˆ 1 ‡ I=Ki, where V0 and V are the tracer uptake
rates in the absence and presence of inhibitor, I is the
inhibitor concentration and Ki is the half maximal inhi-
bition constant or anity constant.31
Allylation of the 6-OH position took into account the
preferential reactivity of the 6-hydroxyl in fructofur-
anosides23 (Scheme 4). Reaction of methyl a-d-fructo-
furanose with monomethoxytrityl chloride a€orded 16;
the free hydroxyl groups were then acetylated, which was
followed by deprotection of the trityl ether to give 17. To
introduce an allyl group a palladium catalysed reaction24,29
was used. De-acylation using sodium methoxide then gave
18, which was followed by acidic cleavage of the glycosidic
bond to a€ord 6-O-allyl-d-fructose, 19 (>20% overall
yield from methyl a-d-fructofuranoside).
Biological evaluation
In Fig. 1A the interaction of GLUT5 with the furan
compound 2,5-anhydro-d-mannitol is compared with
the pyranose ring analogues methyl b-d-fructopyrano-
side and 1,2-O-isopropylidene b-d-fructopyranose. Each
of these compounds inhibits the uptake of tracer d-fruc-
tose with similar potency (Table 1). Comparisons of the
common structural features of 2,5-anhydro-d-mannitol
with b-d-fructofuranose (Fig. 1B) indicate that the b-
hydroxy is absent from the former compound but the
molecules are elsewhere structurally identical. Additional
data in Table 1 indicate that the methyl b-d-fructofurano-
side compound also has high anity (Ki=15.5 mM) but
that in comparison, the methyl a-d-fructofuranoside has
reduced interaction (Ki=32.4 mM). Therefore, the fructo-
furanose-C2 R con®guration appears to be preferred.
We have previously studied the interaction of d-glucose
analogues with the binding site of the erythrocyte and
adipocyte glucose transporters, GLUT1 and GLUT4,
respectively.30,31 For example, we have found that methyl
and propyl glucosides are poorly tolerated at the exofacial
binding site.32,33 It was concluded that the d-glucose (in
its pyranose ring form) enters the GLUT binding site in a
speci®c orientation with C1 (the front of the molecule
relative to the binding site) entering ®rst. Conversely,
substitutions into the C4 and C6 positions of d-glucose
are well tolerated at the exofacial site.32 -34 Considera-
tion of these factors led to the development of bis-hexose
photoanity labels in which C4 positions are linked35
and in which the linker is used as a platform for attach-
ment of the bulky tri¯uoromethylazibenzoate moiety.36
These compounds have been used extensively for
studying GLUT1 biochemistry.36 More importantly
from the metabolic perspective, the photolabels have
The b-d-fructofuranose form is 20% of the total d-
fructose added40 and yet the anity for the d-fructose
mixture is similar to 2,5-anhydro-d-mannitol (Table 1).
Scheme 4. Reagents and conditions: (i) (MeOPh)Ph2CCl, pyridine, 4 ꢀC; (ii) Ac2O, pyridine, 4 ꢀC; (iii) AcOH; (iv) allylethylcarbonate, Ph2P-C4H8-
PPh2; (v) MeONa, MeOH; (vi) dioxane, 0.5M HCl.

1828
A. Tatibouet et al. / Bioorg. Med. Chem. 8 (2000) 1825±1833
È
Therefore, either the b-hydroxyl of d-fructofuranose
(which is absent in 2,5-anhydro-d-mannitol) is required
for high anity binding or alternatively b-d-fructopyr-
anose also interacts with GLUT5. The latter possibility
is consistent with the observed interaction of the b-pyr-
anose ring analogues. Structural considerations of the
common features of d-fructofuranose and d-fructopyr-
anose suggest that the arrangement of hydroxyls at C1,
C2, C3 and C4 and the ring oxygen are essentially the
same in these two forms which only di€er around the
C5 and C6 positions. The speci®city of GLUT5 there-
fore has evolved in such a way as to make use of d-
fructose in all its ring forms.
We have examined the tolerance of d-fructose epimers
by GLUT5 (Table 1). The C3, C4 and C5 epimers are
all poorly tolerated. The C4 epimer, d-tagatose has
slightly higher anity than the C3 and C5 epimers d-
psicose and l-sorbose, respectively. These data suggests
that each of these hydroxyls in the fructo-con®guration
may be involved in H-bonding to GLUT5. However, it
is also possible that the epimerized hydroxyls sterically
reduce binding.
Figure 1. Comparison of the interaction of furanose and pyranose
ring forms of fructose with GLUT5. In A, the uptake of 1 mM [14C]-d-
fructose into CHO cells expressing GLUT5 was measured in the pre-
sence of the indicated concentrations of 2,5-anhydro-d-mannitol (*),
methyl-b-d-fructopyranoside (~) and 1,2-O-isopropylidene-b-d-fruc-
topyranose (~). Ki values were determined by linear regression ®tting
In order to explore the spatial requirements around the
d-fructose binding site in GLUT5, we have chosen to
substitute allyl groups at all the hydroxyl positions. In
the case of the anomeric centre, we have compared
methyl and allyl substitution. The allyl group was chosen
as it is quite planar and therefore occupies a relatively
®xed distance from the fructose ring. Second, it is a
useful group for further modi®cation. Comparison of
to Eqn. 1. The data points are meanÆS.E.M. from
3 separate
experiments. In B, the structures of furanose and pyranose deriva-
tives of d-fructose are contrasted. The structures are similar around
the ring oxygen and C1, C2, C3 and C4.
Table 1. Interaction of d-fructose analogues with the GLUT5 transportera
Position substituted and derivative
Structure and
compound number
Pyranose/furanose ratio
(where known)
KiÆS.E.M.
(mM)
aP,bP
aF,bF
d-fructose
C1,C2
Fig. 1
Fig. 1
6
trace,75
14,67
4,21
15.5Æ2.9
20.9Æ4.3
1,2-O-isopropylidene-b-d-fructopyranose
C1
1-O-allyl-d-fructose
C2
100 bP
trace,19
105.6Æ19.0
Deoxy: 2,5-anhydro-d-mannitol
methyl a-d-fructofuranoside
allyl-a-d-fructofuranoside
methyl-b-d-fructofuranoside
methyl-b-d-fructopyranoside
allyl-b-d-fructofuranoside
allyl-b-d-fructopyranoside
C3
Fig. 1
Fig. 1
2a
Fig. 1
Fig. 1
2b
100bF
100aF
100aF
100bF
100bP
100bF
100bP
12.6Æ1.8
32.4Æ4.1
79.6Æ3.9
15.5Æ1.8
15.0Æ0.3
65.1Æ5.8
28.5Æ3.1
4
Epimer: d-psicose
3-O-allyl-d-fructose
C4
22,24
15,38
39,15
12,35
134.2Æ18.0
89.6Æ31.1
8
Epimer: d-tagatose
4-O-allyl-d-fructose
C5
79,16
55,18
1,4
18,9
59.4Æ8.2
12
105.6Æ20.1
Epimer: l-sorbose
5-O-allyl-d-fructose
C6
98aP
142.7Æ10.9
153.0Æ36.6
15
19
73,27
noF
6-O-allyl-d-fructose
100 a+bF, no open chain
20.3Æ0.8
aUptake of 1 mM [14C]-d-fructose into CHO cells expressing GLUT5 was measured at a range of concentrations of the d-fructose analogues (usually serial
increases in concentrations up to 40 mM). The Ki values were calculated by least squares ®tting to Eqn. 1. The tabulated values are the mean and SEM of 3±5
independent experiments. The pyranose/furanose ratios in solution are from literature values40 or are estimated from 13C NMR data for C2 of the compounds
in D2O. As discussed40 values for the proportion of anomeric mixtures are approximations only because of inherent inaccuracies in peak intensities and peak
area estimation by NMR.

A. Tatibouet et al. / Bioorg. Med. Chem. 8 (2000) 1825±1833
È
1829
the anity (Table 1) for methyl b-d-fructofuranoside
(Ki=15.5 mM) and allyl b-d-fructofuranoside 2b
(Ki=65.1 mM) suggests that the additional size of the
allyl group reduces interaction with GLUT5 around the
C2 position. Similarly, methyl b-d-fructopyranoside
(Ki=15.0 mM) has higher anity than allyl b-d-fructo-
pyranoside 4 (Ki=28.5 mM). These data are consistent
with the postulates that GLUT5 accepts both furanose
and pyranose ring compounds, prefers the b con®gura-
tion and has limited capacity to accept bulky groups at
C2 (particularly in the a-con®guration).
glucose would exclude d-fructose from signi®cant
transport via GLUTs 1±4 and consequently its access to
intracellular metabolic processes.
Conclusion
The GLUT5 protein is a versatile transporter in that it
has evolved to accept d-fructose in both furanose and
pyranose ring shapes with high anity and yet can
exclude d-glucopyranoses. The 2,5-anhydro-d-mannitol,
methyl b-d-fructopyranoside and 1,2-O-isopropylidene-
b-d-fructopyranose are tolerated while 3-O-methyl-d-
glucose is totally excluded from interaction. Structural
considerations of the common features of d-fructofur-
anose and d-fructopyranose suggest that the arrange-
ment of hydroxyls at C1, C2, C3 and C4 and the ring
oxygen are essentially the same in these two forms. They
only di€er around the C5 and C6 positions. Data pre-
sented here suggest that the GLUT5 transporter is highly
speci®c for the fructo-hydroxyls and will not accept d-
fructose epimers. It is concluded that there is a close
approach of the GLUT5 binding site to these hydroxyl
positions. Consistent with this, we ®nd that allyl sub-
stitutions are not well tolerated at positions C1, C2, C3,
C4 and C5. However, the bulky allyl group is tolerated
at the C6 position of d-fructofuranose. C6 substituted
fructofuranoses will therefore be useful as lead com-
pounds for generation of new anity probes to speci®-
cally tag the GLUT5 protein. Such reagents can be used to
study its expression in cells and tissues and to investigate
the regulation of its subcellular localization.
1-O-allyl-d-fructose 6 (a mixture of ꢁ20% furanose and
80% pyranose ring forms, Table 1) is poorly tolerated
(Fig. 2). We have suggested above that the anomeric
half of fructose is preferentially recognised as it is this
half of the molecule that is similar in both furanose and
pyranose forms of d-fructose. The anomeric center can
therefore be considered as the ``front'' of the molecule
relative to the binding site. The poor anity of 1-O-
allyl-d-fructose 6 may be due to a tight approach of the
GLUT5 protein around C1.
3-O-, 4-O-, and 5-O-allyl-d-fructoses (compounds 8, 12
and 15 all have low anity (Kisꢁ80±100 mM; Table 1)
for GLUT5. The GLUT5 protein must therefore quite
tightly approach these positions when interacting with
the associated hydroxyls. By contrast, the data on 6-O-
allyl-d-fructose 19 (Ki=20 mM, Fig. 2) suggest that this
position occupies an open portion of the binding site.
Data described here therefore suggests that the orienta-
tion of d-fructose as it enters GLUT5 is similar to the
orientation of d-glucose in its interactions with GLUT1
and GLUT4. The ``front'' of d-glucose and d-fructose
are most important for recognition. It is probably the
accomodation, by GLUT5, of both the d-fructose
hydroxyl and hydroxymethyl groups at the anomeric
center that allows it to distinguish this substrate from d-
glucose. 3-O-methyl-d-glucose has no detectable anity
for GLUT5 (unpublished data). Remarkably therefore,
GLUT5 has evolved with the ability to exclude d-glu-
cose while retaining recognition of both furanose and
pyranose ring forms of d-fructose. This is a metaboli-
cally important property as d-fructose has low anity
for the glucose transporters (GLUTs 1±4). Therefore, if
GLUT5 were absent, competition with circulating d-
Experimental
General. Methanol was distilled over magnesium, pyri-
dine was dried over KOH, dioxane and tetrahydrofuran
were distilled over LiAlH4, anhydrous DMF was pur-
chased from Aldrich in Sure-Seal bottles. Column chro-
matography was performed as described.41,42 NMR
spectra were recorded for solutions in D2O (internal 3-
(trimethylsilyl)-propionic acid) in CD3OD (residual
absorption of solvent) and in CDCl3 (internal tetramethyl
silane). FAB-mass spectra were measured in the positive
or negative mode with a VG Autospec Fisons instrument
(the samples were dissolved in m-nitrobenzyl alcohol).
An Optical Activity Ltd polarimeter was used for the
determination of optical rotations at ꢁ20 ꢀC. Elemental
analysis was performed by the micro-analysis service at
the University of Bath. Attributions of anomeric car-
bons of fructose derivatives were based on 13C NMR of
fructose in D2O and CD3OD40 (F: Furanose ring, P:
Pyranose ring). All evaporations of volatiles were per-
formed at t <30 ꢀC under reduced pressure.
Allyl 1,3,4,6-tetra-O-benzoyl-ꢀ,ꢁ-D-fructofuranoside (1).
1,3,4,6-tetra-O-benzoyl-b-d-fructofuranose (1 g, 1.678
mmol), Pd2(dba)3 [trisdibenzylideneacetone palladium
(0)] (38.6 mg, 0.042 mmol) and Ph3P [triphenylphosphine]
(44 mg, 0.167 mmol) was dissolved in freshly distilled
dioxane (10 mL). Then allyl methyl carbonate (572 mL,
5.03 mmol) was added. The solution was stirred at 50 ꢀC
Figure 2. Interaction of GLUT5 with 1-O-allyl-d-fructose and 6-O-
allyl-d-fructose. In A, the uptake of 1 mM [14C]-d-fructose into CHO
cells expressing GLUT5 was measured in the presence of the indicated
concentrations of 1-O-allyl-d-fructose (~) and 6-O-allyl-d-fructose
(&). Ki values were determined by linear regression ®tting to Eqn. 1.
The data points are mean ÆSEM from 3 separate experiments.

1830
A. Tatibouet et al. / Bioorg. Med. Chem. 8 (2000) 1825±1833
È
for 1 h. The solvent was evaporated and the syrup puri®ed
on silica gel by ¯ash chromatography with 4:1 petroleum
ether±EtOAc. Compound 1 (0.997g, 1.567 mmol) was
(0.1 g, 1.85 mmol) and stirred at room temperature for
14 h. The reaction was neutralized with Dowex 50W-
8X. The resin was removed by ®ltration and washed
with MeOH. The ®ltrate was evaporated and the syrup
obtained, puri®ed on silica gel by ¯ash chromatography
with 4:1 EtOAc±(MeOH±H2O 2:1). Allyl b-d-fructopyr-
anoside 4 (0.182 g, 0.245 mmol) was isolated after crys-
tallization in water with 75% yield: [a]2d0 137 (c 0.71,
CH3OH). 1H NMR (CD3OD, 400 MHz): d 5.98±5.89 (m,
1H, H-20), 5.33±5.27 (m, 1H, H-30a), 5.13±5.09 (m, 1H,
H-30b), 4.13±4.02 (m, 1H, H-10), 3.93 (d, 1H, J3,4 9.77 Hz,
H-3), 3.85±3.83 (m, 1H, H-5), 3.81 (dd, 1H, J4,5 3.36 Hz,
H-4), 3.76 (d, 1H, J1a,1b 11.29 Hz, H-1a), 3.75 (dd, 1H,
J6a,6b 12.2 Hz, H-6a), 3.69 (d, 1H, H-1b), 3.66 (dd, 1H,
J5,6b 1.83 Hz, H-6b). 13C NMR (CD3OD, 100 MHz): d
136.4 (allyl CH), 116.0 (allyl CH2), 102.0 (C-2), 71.5 (C-
4), 71.0 (C-5), 70.4 (C-3), 65.3 (C-6), 63.4 (C-1), 62.8
(allyl CH2). Anal. calcd (C9H16O6): C, 49.09; H, 7.32.
Found: C, 49.2; H, 7.33.
isolated with 93% yield: [a]2d0 +5 (c 1.15, CHCl3). H
1
NMR (CDCl3, 400 MHz): d 8.11±7.93 (m, 8H, Ph-H),
7.62±7.21 (m, 12H, Ph-H), 6.12 (d, 0.3H, J3,4 6.71Hz, H-
3b), 6.13±5.82 (m, 2H, J3,4 1.23Hz, H-3a, H-4b, H-20),
5.58 (dd, 0.7 Hz, J4,5 4.58Hz, H-4a), 5.41±5.35 (m, 0.7 H,
H-30aa), 5.30±5.25 (m, 0.3H, H-30ab), 5.15±5.11 (m, 0.7H,
H-30ba), 5.10±5.07 (m, 0.3H, H-30bb), 4.95 (d, 0.7H, J1a,1b
12.51 Hz, H-1aa), 4.85 (dd, 0.7H, J6a,6b 11.9Hz and J5,6a
3.36Hz, H-6aa), 4.78 (dd, 0.3H, J6a,6b 11.0 Hz and J5,6a
3.66Hz, H-6ab), 4.71 (dd, 0.7H, J5,6b 3.89Hz, H-6ba),
4.67±4.55 (m, 1.6H, H-1b, H-6bb, H-5), 4.45 (d, 0.7H, H-
1ba). 13C NMR (CDCl3 ; 75 MHz): d 166.2±164.7 (CO),
134.1±128.3 (Ph-H, allyl C-Ha,b), 116.4 (allyl C-20a),
116.3 (allylC-20b), 107.4 (C-2a), 103.5 (C-2b), 81.9, 81.0,
78.9, 78.8, 78.0, 77.3, 63.7 (C-1b), 63.6 (C-1a), 63.2 (C-
6b), 62.3 (C-6a), 59.7 (allyl C-10a), 58.3 (allyl C-10b).
.
Anal. calcd for (C37H32O10 CH3CH2OH): C, 68.61; H,
5.61. Found: C, 68.60: H, 5.58.
1-O-Allyl-D-fructose (6). 526 (780 mg, 2.6 mmol) was dis-
solved in 4:1 TFA±H2O and stirred at room temperature
for 4 h. TFA was removed by co-evaporation with
MeOH until no smell was noticeable. The pale yellow
syrup obtained was puri®ed on silica gel by ¯ash chro-
matography with EtOAc then 4:1 EtOAc±(MeOH±H2O
2:1). Compound 6 (435 mg, 1.97 mmol) was isolated with
76% yield: [a]2d0 29 (c 1.865, CH3OH). 1H NMR
(CD3OD, 400 MHz): d 5.98±5.87 (m, 1H, H-20), 5.33±
5.26 (m, 1H, H-30a), 5.18±5.15 (m, 1H, H-30b), 4.12±3.39
(m, 9H). 13C NMR (CD3OD, 100 MHz): d 136 (allyl-
CH), 135.9 (allyl-CH), 135.8 (allyl-CH), 102.6 (C-2b-F),
101.5 (C-2a-P), 99.0 (C-2b-P), 83.4 (CH), 83.1 (CH),
78.4 (CH), 78.0 (CH), 76.9 (CH), 76.5 (CH), 73.7 (CH2),
73.6 (CH2), 73.5 (CH2), 73.4 (CH2), 72.8 (CH2), 72.3
(CH2), 71.6 (CH), 71.3 (CH), 71.0 (CH), 70.8 (CH), 70.6
(CH2), 70.5 (CH), 70.0 (CH), 69.7 (CH), 65.0 (CH2), 64.8
(CH2), 64.6 (CH2), 64.4 (CH2), 63.9 (CH2), 63.2 (CH2).
MS (FAB(+), NBA): m/z 243 (M++ Na). Anal. calcd
(C9H16O6) C, 49.09; H, 7.32. Found: C, 49.5; H, 7.34.
Allyl ꢀ,ꢁ-D-fructofuranosides (2ꢀ, 2ꢁ). Allyl-1,3,4,6-tetra-
O-benzoyl-d-fructofuranoside 1 (2.65 g, 4.17 mmol) was
suspended in dry MeOH (10 mL) and NaOMe (0.4 g,
7.4 mmol) was added. The solution was stirred at room
temperature for 4 h, then neutralized with Dowex 50W-
8X, evaporated and puri®ed on silica gel by ¯ash chro-
matography with 9:1 EtOAc±(MeOH-H2O 2:1). Allyl-a-
d-fructofuranoside 2a (512 mg, 2.33mmol) and allyl-b-d-
fructofuranoside 2b (235mg, 1.07mmol) were isolated
with 56% and 25% yields, respectively: 2a [a]2d0 +20.3
(c 0.985, CH3OH). 1H NMR (CDCl3, 400 MHz): d 5.99±
5.89 (m, 1H, H-20), 5.31±5.26 (m, 1H, H-30a), 5.11±5.08
(m, 1H, H-30b), 4.19±4.13 (m, 1H, H-10a), 4.19±4.13 (m,
1H, H-10a), 4.10±4.04 (m, 2H, H-10b and H-3), 3.90 (dd,
1H, J3,4 4.88Hz and J4,5 7.02Hz, H-4), 3.85 (m, 1H, H-5),
3.76 (dd, 1H, J5,6a 2.74Hz and J6a,6b 11.9 Hz, H-6a), 3.71
(d, 1H, J1a,1b 11.9 Hz, H-1a), 3.64 (d, 1H, H-6b), 3.62 (d,
1H, H-1b), 13C NMR (CDCl3, 100 MHz): d 136.7 (allyl
CH), 116.2 (allyl CH2), 109.0 (C-2), 84.0 (C-5), 83.3 (C-3),
78.5 (C-4), 63.5 (C-6), 62.6 (C-1), 61.6 (allyl CH2). MS
(FAB(+), NBA): m/z=243 (M++Na). MS (FAB
3-O-Allyl-1,2,4,5-di-O-isopropylidene-ꢁ-D-fructopyranose
(7). 1,2,4,5-di-O-isopropylidene fructopyranose14 (1.5 g,
5.77mmol) was dissolved in 10mL of dry THF and then
chilled to 20 ꢀC in an ethanol±dry ice bath. NaH (0.353g,
9.3 mmol) and allylbromide (1.8 mL, 3.6 eq) were added.
The solution was allowed to return to room temperature
and stirred for 21h. H2O (50 mL) was added, and the
solution was extracted with EtOAc (3Â25mL), The
organic layers were collected, washed with H2O
(2Â50mL), dried over Na2SO4 and evaporated. The pale
yellow liquid was subjected to ¯ash chromatography with
3:1 petroleum ether±ethyl acetate. The product 7 (1.36 g,
4.53mmol) was isolated as a pale yellow liquid with
( ),NBA): m/z 219 (M+ H). Anal. calcd (C9H16O6
.
H2O) C, 48.10 H, 7.40. Found: C, 48.10; H, 7.39. 2b [a]d20
27.1 (c 1.66, CH3OH). 1H NMR (CDCl3, 400 MHz): d
5.90±5.84 (m, 1H, H-20), 5.25±5.21 (m, 1H, H-30a), 5.06±
5.03 (m, 1H, H-30b), 4.21±3.90 (m, 4H), 3.74±3.48 (m,
5H). 13C NMR (CDCl3; 100 MHz): d 137.1 (allyl CH),
116.2 (allyl CH2), 105.7 (C-2), 83.7 (C-5), 78.6 (C-3), 77.3
(C-4), 65.1 (C-6), 63.7 (C-1), 62.0 (allyl CH2) ). MS
(FAB(+), NBA): m/z=243 (M++Na). MS (FAB ( ),
NBA): m/z 219 (M+ H). Anal. calcd (C9H16O6.0.5
H2O): C, 47.16; H, 7.48, Found: C, 47.10; H, 7.39.
78.5%. [a]2d0 100 (c 1.5, CHCl3). H NMR (CDCl3,
1
Allyl 1,3,4,5-tetra-O-benzoyl-ꢁ-D-fructopyranoside (3).
Same conditions as 2. The product 3 (0.937 g, 1.473mmol)
was isolated after crystallization in absolute ethanol with
88% yield: data in accordance with the literature.43
400 MHz): d 5.95±5.85 (m, 1H, H0-2), 5.32-5.15 (m, 2H,
H0-3), 4.44±4.39 (m, 1H, H0-1a), 4.30 (dd, 1H, J4,3 7.33Hz
and J4,5 5.5 Hz H-4), 4.21±4.19 (m, 1H, H-5), 4.15±3.92
(m, 5H, H0-1b, H-6, H-1), 3.44 (dd, 1H, J3,5 2.14Hz and
J3,4, H-3), 1.54 (s, 3H, CH3), 1.49 (s, 3H, CH3), 1.41 (s,
3H, CH3), 1.36 (s, 3H, CH3). 13C NMR (CDCl3, 100
MHz): d 134.5 (C0-2), 116.7 (C0-3), 112 (C(Me)2); 108.7
(C(Me)2), 104.3 (C-2), 77.5 (CH), 75.8 (CH), 73.7 (CH),
Allyl ꢁ-D-fructopyranoside (4). Allyl-1,3,4,5-tetra-O-ben-
zoyl-b-d-fructopyranoside 3 (0.7 g, 1.1 mmol) was sus-
pended in freshly distilled MeOH (10 mL) with NaOMe

A. Tatibouet et al. / Bioorg. Med. Chem. 8 (2000) 1825±1833
È
1831
72.1 (CH2), 71.6 (CH2), 59.9 (CH2), 28.0 (CH3), 26.7
(CH3), 26.1 (CH3), 25.8 (CH3). MS (FAB(+), NBA): m/z
301 (M++H). Anal. calcd (C15H24O6): C, 59.98; H, 8.05.
Found: C, 60.2; H, 8.15.
2H, H-10b, H-1a), 4.03±4.02 (m, 1H, H-5), 3.90±3.87 (m,
2H, H-4 and H-1b), 3.76 (d, 1H, J6a,6b 10.07, H-6a),
3.64 (dd, 1H, J3,5 2.14 Hz and J3,4 12.51 Hz, H-3), 3.52
(dd, 1H, J5,6b 3.35 Hz, H-6b), 1.45 (s, 3H, CH3), 1.40 (s,
3H, CH3). 13C NMR (CD3OD, 100 MHz): d 136.6 (allyl
CH), 117.3 (allyl CH2), 113.1 (CMe2), 107.7 (C-2), 79.8
(CH), 72.8 (CH2), 71.7 (CH2), 68.2 (CH), 67.9 (CH),
65.3 (CH2), 27.3 (CH3), 26.4 (CH3). MS (FAB(+),
NBA): m/z 261 (M++H). Anal. calcd (C12H20O6): C,
55.37; H, 7.74. Found: C, 55.25; H, 7.95.
3-O-Allyl-D-fructose (8). Compound 7 (810 mg, 2.7 mmol)
was dissolved in 4:1 TFA±H2O (2 mL) and stirred at
room temperature for 4 h. TFA was removed by co-
evaporation with MeOH. The syrup was puri®ed on
silica gel by ¯ash chromatography with EtOAc then 4:1
EtOAc±(MeOH±H2O 2:1) giving 8 (497 mg, 2.26 mmol)
as an oil with 83% yield. [a]2d0 +7.7 (c 1, CH3OH). H
4-O-Allyl-D-fructose (12). Same conditions as 6. Com-
pound 12 (206 mg, 0.936 mmol) was isolated as an oil
1
NMR (D2O, 400 MHz): d 6.00±5.70 (m, 1H, H-20),
5.20±5.10 (m, 2H, H-30), 4.12±3.27 (m, 9H). 13C NMR
(D2O, 100 MHz): d 134.7±134.1 (allyl CH), 120.1±119.5
(allyl CH2), 103.0 (C-2a-F), 102.8 (C-2b-F), 99.9 (C-2a-
P), 96.8 (C-2b-P), 90.0, 89.9, 84.3, 84.2, 82.5, 77.4, 76.1,
75.7, 75.4, 73.1, 72.1, 72.0, 70.2, 70.1, 64.5, 63.8, 63.4,
62.7, 62.5 (2C), 62.2, 61.9, 61.8. MS (FAB(+), NBA):
m/z 243 (M++Na). Anal. calcd (C9H16O6): C, 49.09;
H, 7.32. Found: C, 49.0; H, 6.95.
with 90% yield. [a]d20 21.3 (c 0.61, CH3OH). H NMR
1
(CD3OD, 400 MHz): d 6.04±5.85 (m, 1H, H-20), 5.35±
5.25 (m, 1H, H-30a), 5.16±5.13 (m, 1H, H-30b), 4.23±3.45
(m, 9H). 13C NMR (CD3OD, 100 MHz): d 136.7, 136.2,
136.1, 117.4, 117.3, 117.0, 106.3 (C-2a-F), 103.5 (C-2b-
F), 99.3 (C-2b-P), 84.8, 82.5, 79.3, 77.3, 72.2, 72.1, 71.8,
68.5, 68.2, 65.9, 64.6, 64.4, 64.3, 63.1, 62.9. MS (FAB(+),
NBA): m/z 243 (M++Na). Anal. calcd (C9H16O6): C,
49.09; H, 7.32> Found: C, 49.40; H, 7.21.
4-O-Allyl-3-O-benzoyl-1,2-O-isopropylidene-ꢁ-D-fructo-
pyranose (9) and 5-O-allyl-3-O-benzoyl-1,2-O-isopropyli-
dene-ꢁ-D-fructopyranose (10). Same conditions as 1 but
using 3-O-benzoyl-1,2-O-isopropylidene-fructopyranose28
(1.625 g, 5 mmol) and allylmethylcarbonate (0.81 mL,
7.13 mmol). Compounds 9 (580 mg, 1.59 mmol) and 10
(710 mg, 1.95 mmol) were isolated as oils with 39% and
32% yields respectively. 9 [a]2d0 113.5 (c 1.26, CHCl3).
1H NMR (CDCl3, 400 MHz): d 8.11±8.09 (2H, m, Ph-
H), 7.61±7.57 (m, 1H, Ph-H), 7.49±7.45 (m, 2H, Ph-H),
5.79±5.69 (m, 1H, H-20), 5.58 (d, 1H, J3,4 9.77 Hz, H-3),
5.21±5.13 (m, 1H, H-30a), 5.12±5.04 (m, 1H, H-30b),
4.14±4.09 (m, 2H, H-10), 4.05±3.99 (m, 3H), 3.95±3.91
(m, 3H), 2.77 (s, 1H, OH), 1.51 (s, 3H, CH3), 1.41 (s,
3H, CH3). 13C NMR (CDCl3; 100 MHz): d 165.9 (CO),
134.1±128.5 (PhCH, allyl CH), 117.7 (allyl CH2), 111.9
(CMe2), 104.7 (C-2), 76.5 (CH), 71.9 (CH2), 71.2 (CH2),
68.6 (CH), 67.3 (CH), 63.3 (CH2), 26.6 (CH3), 26.3
(CH3). MS (FAB(+), NBA): m/z 365 (M++H). Anal.
calcd (C19H24O7): C, 62.63; H, 6.64. Found: C, 62.70; H,
5-O-Allyl-3,4-di-O-benzoyl-1,2-O-isopropylidene-ꢁ-D-
fructopyranose (13). Same conditions as 1. Compound
13 (0.478mg, 1.021 mmol) was isolated with 87% yield.
[a]2d0 177 (c 1.815, CHCl3). 1H NMR (CDCl3, 400 MHz):
d 8.12±8.10 (m, 2H, Ph-H), 8.02±7.99 (m, 1H, Ph-H),
7.59±7.55 (m, 1H, Ph-H), 7.51±7.44 (m, 2H, Ph-H),
7.37±7.32 (m, 3H, Ph-H), 5.95 (d, 0.64H, J3,4 10.7 Hz,
H-3), 5.87±5.77 (m, 0.64H, allyl H-20), 5.73±5.58 (m,
1.72H, H-3, allyl H-20 and H-4), 5.23±5.18 (m, 0.64H,
H-30a), 5.17±5.11 (m, 0.36H, H-30a), 5.10±5.06 (m,
0.64H, H-30), 5.03±5.00 (m, 0.36H, H-30b), 4.19±3.90 (m,
7H, H-1, H-5, H-6 and H-10), 1.53 (s, 3H, CH3), 1.44 (s,
1.2H, CH3), 1.43 (s, 1.8H, CH3). 13C NMR (CDCl3,
100 MHz): d 166.1±165.9 (CO), 134.4±128.1 (PH-H),
117.5 (C-30), 117.3 (C-30), 112.1 (C-20), 104.9 (C-2), 104.8
(C-2), 74.5 (CH), 74.3 (CH), 72.3 (CH), 72 (CH2), 71.7
(CH2), 70.8 (CH2), 69.1 (CH), 68.9 (CH), 67.5 (CH),
67.1 (CH2), 65.8 (CH2), 62.5 (CH2), 26.6 (CH3), 26.5
(CH3), 26.3 (CH3), 26.1 (CH3). MS (FAB(+), NBA):
m/z 491(M++Na), 469 (M++H). Anal. calcd
(C26H27O8): C, 66.66; H, 6.02. Found: C, 66.70; H, 6.15.
6.83. 10 [a]2d0 141 (c 1.68, CHCl3). H NMR (CDCl3,
1
400 MHz): d 8.12±8.09 (m, 2H, Ph-H), 7.61±7.55 (m,
1H, Ph-H), 7.47±7.41 (m, 2H, Ph-H), 6.00±5.90 (m, 1H,
H-20), 5.43 (d, 1H, J3,4 10.07 Hz, H-3), 5.35±5.30 (m, 1H,
H-30a), 5.24±5.21 (m, 1H, H-30b), 4.26±4.22 (m, 1H, H-
10a), 4.14±4.00 (m, 3H, H-10b, H-4, H-1a), 3.96±3.88 (m,
3H, H-6, H-1b), 3.77 (d, 1H, J5,6 1.83 Hz, H-5), 2.56 (d,
1H, J4,OH 11.29 Hz, OH), 1.51 (s, 3H, CH3), 1.45 (s, 3H,
CH3). 13C NMR (CDCl3, 100 MHz): d 166.7 (CO),
134.2±128.4 (Ph-CH), 117.9 (CH2), 112.0 (CMe2), 104.6
(C-2), 76.6 (CH), 71.9 (CH2), 70.7 (CH2), 70.6 (CH),
69.6 (CH), 60.7 (CH2), 26.6 (CH3), 26.3 (CH3). MS
(FAB(+), NBA): m/z 365 (M++H). Anal. calcd
(C19H24O7): C, 62.63; H, 6.64. Found: C, 62.40; H, 6.65.
5-O-Allyl-1,2-O-isopropylidene-D-fructopyranose (14).
Same conditions as 4. Compound 14 (0.15g, 0.577 mmol)
was isolated as a white solid with 68% yield. [a]2d0 150 (c
1
0.4, CH3OH). H NMR (CDCl3, 400MHz): d 5.99±5.88
(m, 1H, H-20), 5.35±5.30 (m, 1H, H-3a0), 5.23±5.19 (m, 1H,
H-3b0), 4.34±4.29 (m, 1H, H-1a0), 4.21±4.17 (m, 1H, H-
1b0), 4.1±3.94 (m, 4H); 3.78±3.74 (m, 1H), 3.59 (d, 1H), 2.89
(s, 2H, OH), 1.49 (s, 3H, CH3), 1.42 (s, 3H, CH3). 13C
NMR (CDCl3, 100 MHz): d 134.3 (C-20), 117.5 (C-30),
111.9 (C-1), 105.5 (C-2), 76.3 (CH), 74.2 (CH2), 71.6 (CH2),
71.2 (CH), 69.7 (CH), 63.5 (CH2), 26.9 (CH3), 26.0 (CH3).
MS (FAB(+), NBA): m/z 259 (M+ H). Anal. calcd
(C12H20O6): C, 55.37; H, 7.74. Found: C, 55.50; H, 7.80.
4-O-Allyl-1,2-O-isopropylidene-ꢁ-D-fructopyranose (11).
Same conditions as 4. Compound 9 (287 mg, 1.11 mmol)
was isolated as an oil with 78% yield. [a]2d0 137.5 (c
5-O-Allyl-D-fructopyranose (15). Same conditions as 6.
Compound 15 (0.152 g, 0.69 mmol) was isolated as an oil
1
0.96, CH3OH). H NMR (CD3OD, 400 MHz): d 6.05±
5.95 (m, 1H, H-20), 5.35±5.29 (m, 1H, H-30a), 5.17±5.13
(m, 1H, H-30b), 4.24±4.19 (m, 1H, H-10a), 4.15±4.09 (m,
with 93% yield. [a]2d0 89 (c 1.28, CH3OH). H NMR
1
(CD3OD, 400 MHz): d 6.02±5.91 (m, 1H, H-20), 5.34±

1832
A. Tatibouet et al. / Bioorg. Med. Chem. 8 (2000) 1825±1833
È
5.28 (m, 1H, H-30a), 5.17±5.13 (m, 1H, H-30b), 4.19±4.08
(m, 2H, H-10), 3.95±3.5 (m, 7H, H-1, H-3, H-4, H-5, H-
6). 13C NMR (CD3OD, 100 MHz): d 136.5 (allyl CH),
136.4 (allyl CH), 117.2 (allyl CH2), 101.3 (C-2a), 99.0
(C-2b), 78.5 (CH), 78.3 (CH), 72.1 (CH2), 72.0 (CH),
71.6 (CH), 71.4 (CH), 71.3 (CH), 70.2 (CH2), 66.1
(CH2), 63.0 (CH2), 62.0 (CH2), 61.4 (CH2). MS
(FAB(+), NBA): m/z 243 (M++Na). Anal. calcd
(C9H16O6): C, 49.09; H, 7.32. Found: C, 48.7; H, 7.70.
and the solution was stirred at 50 ꢀC overnight. After
cooling and evaporation of the solvent, column chroma-
tography of the residue (toluene:AcOEt (4:1)), a€orded
methyl 6-O-allyl-1,3,4-tri-O-acetyl-a-d-fructofuranoside
(390 mg) and unreacted 17 (155 mg). Methyl 6-O-allyl-
1,3,4-tri-O-acetyl-a-d-fructofuranoside was then dis-
solved in dry methanol (10 mL) and sodium methoxide
(250 mL of a 1.5 M in methanol) was then added. After
stirring overnight, water (10 mL) was added and the
solution concentrated to one third under reduced pres-
sure. The pH was adjusted to 6±7 (0.1 N HCl) and the
solution washed with ethyl acetate. The organic layer
was extracted with water and the combined aqueous
layers evaporated to dryness. The residue was taken up
in acetone and after ®ltration, evaporation yielded
209 mg (50%) of 18 as a colorless oil. [a]2d0 +80 (c 0.48,
Methyl 6-O-(diphenyl-(4-methoxyphenyl)-methyl)-ꢀ-D-
fructofuranoside (16). To a cooled (4 ꢀC) solution of
methyl a-d-fructofuranoside23 (1.0 g, 5.15 mmol) in
10 mL pyridine, 4-methoxyphenyldiphenylmethyl chlo-
ride (1.7 g, 5.5 mmol) was added and the solution was
left overnight at 4 ꢀC. After addition of water, extrac-
tion was performed (CH2Cl2), the organic layer washed
with water, dried and concentrated with toluene. The
residue was puri®ed by column chromatography on
silica gel (CH2Cl2±CH3OH (95-5)) a€ording 16 (1.35 g,
56%) as a colorless glass. [a]2d0+51ꢀ (c 0.73, CH2Cl2).
1H NMR (200 MHz, CDCl3): d 8.0±7.2 (m, 12 H, C6H5
and H-Ar meta to OCH3), 6.9 (d, J 8 Hz, 2 H, H-Ar
ortho to OCH3), 4.6±3.2 (m, 14 H). 13C NMR (50 MHz,
CDCl3) d 158.2 (C ipso of -Ar-OCH3), 143.7 (C ipso
of -C6H5), 134.8 (C para of -Ar-OCH3, 130.1 (C para
of -C6H5), 128.1, 127.5 (C ortho and meta of -C6H5),
126.7 (C ortho of-Ar-OCH3), 112.8 (C ortho of -Ar-
OCH3), 108.2 (C-2), 86.6 (C(Ar)3), 82.6, 80.6, 78.2 (C-3,
-4, -5), 63.5, 58.5 (C-1, -6), 54,7 (ArOCH3), 48.2
(OCH3). Anal. calcd (C27H30O7): C, 69.51; H, 6.48.
Found: C, 69.53; H, 6.50.
1
CH3OH). H NMR (300 MHz, (CD3)2CO); d 5.9 (m, 1
H, H-20), 5.1 and 5.25 (m, 2 H, H-30), 3.5±4.2 (m, 12 H),
3.27 (s, 3 H, OCH3).13C NMR (75 MHz, (CD3)2CO); d
135.7 (C-20), 116.6 (C-30), 108.8 (C-2), 82.2, 81.5, 79.0
(C-3, -4, -5), 58.2, 72.5, 70.9 (C-10, C-1, -6), 48.7
(OCH3). HRFABMS (NBA) calcd for C10H18O6Na:
257.1001 (M++Na). Found: 257.1000.
6-O-Allyl-D-fructose (19). To a solution of 18 (105 mg,
0.45 mmol) in 1,4-dioxane (1 mL) was added 0.5 M
hydrochloric acid (1 mL) and the solution was stirred
overnight at 60 ꢀC. The solvent was removed and the
residue co-evaporated twice with water before adjusting
the pH to 7 with Amberlite IRA-93 (OH ). After ®ltra-
tion, removal of water (evaporation or lyophilisation)
oily 19 was isolated (79 mg, 80%). [a]2d0 +13ꢀ (10 min)!
+18ꢀ (15 h) (c 0.28, CH3OH), 1H NMR (300 MHz,
(D2O), d 6.1 (m, 1 H, H-20), 5.6±5.4 (m, 2 H, H-30), 3.6±
4.4 (m, 9 H). 13C NMR (75 MHz, D2O): d 136.9 (C-20),
121.1 (C-30), 107.3 (C-2b), 104.3 (C-2a), 84.4, 82.7, 79.4
(C-3b, -4b, -5b), 81.7, 78.0, 77.6 (C-3a,-4a,-5a), 74.7,
73.7, 72.3, 65.5, 65.4 (C-6, C-1 and C-10). HRFABMS
(NBA) calcd for C9H16O6Na: 243.0845 (M++Na).
Found: 243.0847.
Methyl 1,3,4-tri-O-acetyl-ꢀ-D-fructofuranoside (17). To
a stirred solution of 16 (1.05 g, 2.25 mmol) in pyridine
(10 mL) at 4 ꢀC, acetic anhydride (4 mL) was added in
portions and the reaction mixture stirred at room tem-
perature for 24 h. Ice was then added and the reaction
mixture was extracted with CH2Cl2. The organic layer
was washed with water, dried and concentrated with
toluene in vacuo. The residue was added to acetic acid
(5 mL). Water was added in portions until turbidity and
then the solution was stirred at room temperature for 6
h. Solvent was removed with toluene before being pur-
i®ed by column chromatography on silica gel (CH2Cl2/
CH3OH (95:5)) a€ording 17 (695 mg, 96%) as a color-
Fructose transport assays. CHO cells expressing
GLUT539 were grown to con¯uence in 19 mm culture
dishes. Growth medium was removed by washing 3
times with KRPB (Krebs-Ringer Phosphate bu€er;
130 mM NaCl, 5 mM KCl, 1.3 mM CaCl2, 1.3 mM
MgSO4 and 10 mM Na2HPO4, pH 7.4). Cells were then
maintained in this bu€er at 37 ꢀC for 10 min. Uptake was
initiated by adding d-[14C-U]-fructose to give a ®nal
concentration of 1 mM, 0.3 mCi in 0.25 mL. After 10 min
at 37 ꢀC, 1 mL of ice cold bu€er containing 0.3 mM
phloretin was added to terminate the reaction. The
dishes were washed rapidly 3 times and then cell radio-
activity was extracted into 0.5 mL of 0.1 M NaOH and
estimated by scintillation counting.
less oil. [a]2d0 +81 (c 1.2, AcOEt). H NMR (300 MHz,
1
CDCl3): d 5.3 (d, 1 H, J 2 Hz, H-3), 4.95 (dd, 1 H, J
2 Hz and J 5 Hz, H-4), 4.1 and 4.1 (AB system, 2 H, J
12 Hz, H-1), 4.0 (m, 1 H, H-5), 3.8 and 3.7 (AB part of
an ABMX system J 12 Hz, 2 H, H-6), 3.28 (s, 3 H,
OCH3), 2.06, 2.04 and 2.01 (3 s, 9 H, COCH3).13C
NMR (75 MHz, CDCl3): d 170.2, 169.9, 168.9 (CO),
108.2 (C-2), 82.6, 79.7, 78.0 (C-3, -4, -5), 61.5, 58.2 (C-1,
-6), 48.3 (OCH3), 20.3 (COCH3). Anal. calcd
(C13H22O9): C, 48.44; H, 6.88. Found: C, 48.42; H, 6.89.
Methyl 6-O-allyl-ꢀ-D-fructofuranoside (18). To a stirred
solution of 17 (580 mg, 1.81 mmol) under argon in dry
THF (12 mL) was added a solution of tris-(dibenzylide-
neacetone)-dipalladium(0) (30 mg) and 1,4-bis-(diphe-
nylphosphino)butane (105 mg) in dry THF (10 mL).
Allyl ethyl carbonate (1.1 g, 8.46 mmol) was then added
Acknowledgements
We are grateful to the BBSRC (UK), Wellcome Trust
and the MRC (UK) for grant support. We also thank
Dr. Y. Oka (Yamaguichi University, Japan) for a gift of
the clone of CHO cells expressing GLUT5.

A. Tatibouet et al. / Bioorg. Med. Chem. 8 (2000) 1825±1833
È
1833
References
23. Furneaux, R. H.; Tyler, P. C.; Whitehouse, L. A. Tetra-
hedron Lett. 1993, 34, 3613.
24. Lakhmiri, R.; Lhoste, P.; Sinou, D. Tetrahedron Lett.
1989, 30, 4669.
1. Wright, E. M. (special issue). Am. J. Physiol. 1998, 38,
G879-G882.
25. Binder, T. P.; Robyt, J. F. Carbohydr. Res. 1984, 132, 173.
26. Diaz, R. R.; Melgarejo, C. R.; Lopezespinosa, M. T. P.;
Cubero, I. I. J. Org. Chem. 1994, 59, 7928.
27. Szeja, W. Pol. J.Chem. 1980, 54, 1301.
28. Fischer, E.; Noth, H. Ber. Dtsch. Chem. Ges. 1918, 51,
321.
29. Lakhmiri, R.; Lhoste, P.; Sinou, D. Synth. Commun. 1990,
20, 1551.
30. Barnett, J. E. G.; Holman, G. D.; Munday, K. A. Bio-
chem. J. 1973, 131, 211.
2. Kayano, T.; Burant, C. F.; Fukumoto, H.; Gould, G. W.;
Fan, Y. S.; Eddy, R. L.; Byers, M. G.; Shows, T. B.; Seino, S.;
Bell, G. I. J. Biol. Chem. 1990, 265, 13276.
3. Rand, E. B.; Depaoli, A. M.; Davidson, N. O.; Bell, G. I.;
Burant, C. F. Am. J. Physiol. 1993, 264, G1169-G1176.
4. Hundal, H. S.; Ahmed, A.; Guma, A.; Mitsumoto, Y.;
Marette, A.; Rennie, M. J.; Klip, A. Biochem. J. 1992, 286, 339.
5. Mueckler, M. Eur. J. Biochem. 1994, 219, 713.
6. Gould, G. W.; Holman, G. D. Biochem. J. 1993, 295, 329.
7. Hajduch, E.; Darakhshan, F.; Hundal, H. S. Diabetologia
1998, 41, 821.
31. Rees, W. D.; Holman, G. D. Biochim. Biophys. Acta.
1981, 642, 251.
32. Barnett, J. E. G.; Holman, G. D.; Chalkley, R. A.; Mun-
day, K. A. Biochem. J. 1975, 145, 417.
8. Zierath, J. R.; Nolte, L. A.; Wahlstrom, E.; Galuska, D.;
Shepherd, P. R.; Kahn, B. B.; Wallberg-Henriksson, H. Bio-
chem. J. 1995, 517.
33. Holman, G. D.; Rees, W. D. Biochim. Biophys. Acta.
1982, 685, 78.
34. Abbadi, M.; Holman, G. D.; Morin, C.; Rees, W. D.;
Yang, J. Tetrahedron Lett. 1999, 40, 5861.
35. Holman, G. D.; Midgley, P. J. W. Carbohydr. Res. 1985,
135, 337.
36. Clark, A. E.; Holman, G. D. Biochem. J. 1990, 269, 615.
37. Yang, J.; Holman, G. D. J. Biol. Chem. 1993, 268, 4600.
38. Koumanov, F.; Yang, J.; Jones, A. E.; Hatanaka, Y.;
Holman, G. D. Biochem. J. 1998, 330, 1209.
39. Inukai, K.; Katagiri, H.; Takata, K.; Asano, T.; Anai, M.;
Ishihara, H.; Nakazaki, M.; Kikuchi, M.; Yazaki, Y.; Oka, Y.
Endocrinology 1995, 136, 4850.
40. Angyal, S. J.; Bethell, G. S. Aust. J. Chem. 1976, 29, 1249.
41. Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43,
2923.
9. Fischer, E. Ber. Dtsch. Chem. Ges 1884, 17, 574.
10. Lichtenthaler, F. W. Carbohydr. Res. 1998, 313, 69.
11. Hudson, C. S.; Brauns, D. H. J. Am. Chem. Soc. 1915, 37,
2736.
12. Pascu, E.; Rich, F. V. J. Am. Chem. Soc. 1933, 55, 3018.
13. Lichtenthaler, F. W.; Klotz, J.; Flath, F. J. Liebigs
Annalen 1995, 2069.
14. Mio, S.; Ueda, M.; Hamura, M.; Kitagawa, J.; Sugai, S.
Tetrahedron 1991, 47, 2145.
15. Brady, R. F. J. Carbohydr. Res. 1970, 15, 35.
16. Van Cleve, J. W. Methods Carbohydr. Chem. 1963, 2, 237.
17. Brauns, D. H. J. Am. Chem. Soc 1920, 1850.
18. Fellahi, M.; Morin, C. Carbohydr. Res. 1999, 322, 142.
19. Horton, D.; Philips, K. Carbohydr. Res. 1973, 30, 367.
20. Szarek, W. A.; Zamojski, A.; Tiwari, K. N.; Ison, E. R.
Tetrahedron Lett. 1986, 27, 3827.
42. Morin, C. J. Chem. Educ. 1988, 65, 903.
43. Raaijmakers, H. W. C.; Arnouts, E. G.; Zwanenburg, B.;
Chittenden, G. J. F. Carbohydr. Res. 1994, 257, 293.
21. Lee, C. K. Carbohydr. Res. 1987, 170, 255.
22. Johnson, L.; Verraest, D. L.; Vanhaveren, J.; Hakala, K.;
Peters, J. A.; Vanbekkum, H. Tetrahedron:Asymmetry 1994, 5,
2475.