Design and synthesis of iminosugar-based inhibitors of glucosylceramide synthase: The search for new therapeutic agents against Gaucher disease

Article abstract of DOI:10.1016/j.tetasy.2005.03.015

A series of iminosugars bearing two or three alkyl chains ('iminoglycolipids') were designed as ceramide mimics and analogues of N-butyl 1-deoxynojirimycin (N-Bu DNJ, Zavesca). This orally active iminosugar inhibits the biosynthesis of glucosyl

Full text of DOI:10.1016/j.tetasy.2005.03.015

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 1747–1756
Design and synthesis of iminosugar-based inhibitors
of glucosylceramide synthase: the search for new therapeutic
agents against Gaucher disease
a
Charlotte Boucheron, Valerie Desvergnes, Philippe Compain, Olivier R. Martin,
a
a,
a,
*
*
´
Alan Lavi,b Muckram Mackeen,b Mark Wormald,b Raymond Dwekb and
Terry D. Buttersb
a
´ ´
Institut de Chimie Organique et Analytique, UMR 6005 CNRS-Universite d’Orleans, rue de Chartres,
´
BP 6759, 45067 Orleans, France
bOxford Glycobiology Institute, Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, UK
Received 25 January 2005; accepted 14 March 2005
Abstract—A series of iminosugars bearing two or three alkyl chains (ÔiminoglycolipidsÕ) were designed as ceramide mimics and
analogues of N-butyl 1-deoxynojirimycin (N-Bu DNJ, Zavescaꢀ). This orally active iminosugar inhibits the biosynthesis of gluco-
sylceramides, which accumulate pathologically in macrophages of patients with Gaucher disease (substrate reduction therapy,
SRT). Molecular modeling and kinetic experiments have suggested that N-Bu DNJ is a competitive inhibitor that mimics the cer-
amide acceptor but not the donor substrate (UDP-glucose) in the glucosylceramide synthase-catalyzed process. Kinetic measure-
ments were made with the glucosyltransferase to assess the selectivity of the new iminoglycolipids with respect to the length (C4
or C8) and the position of the second alkyl chain (C-1, O-2 and/or O-4). This structure–activity relationship study showed that
the addition of a second alkyl chain, to obtain better ceramide mimics, led to less potent inhibitors. Moreover, the synthase active
site did not discriminate inhibitors differing by the position of the second alkyl chain (C-1, O-2 or O-4). Best inhibition was found for
1,5-dideoxy-1,5-imino-N-octyl-4-O-octyl-D-glucitol (IC50 134 lM).
ꢁ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
There are currently two therapies available for Gaucher
disease, which are both aimed at reducing GlcCer
storage.4 The first-line treatment is based on the
administration of Cerezymeꢀ, a recombinant form of
b-glucocerebrosidase, to supplement the defective
hydrolytic enzyme (enzyme replacement therapy,
ERT). A second strategy uses a small molecule, N-butyl
1-deoxynojirimycin 1 (N-Bu DNJ, Zavescaꢀ, Scheme 1),
as a potent inhibitor of glucosylceramide synthase, the
glucosyltransferase involved in the biosynthesis of glyco-
sphingolipids and responsible for the glucosylation of
ceramide (substrate reduction therapy, SRT).5 More
than 3000 patients have been treated worldwide with
Cerezymeꢀ since the 90s and Zavescaꢀ is the first orally
administered treatment for lysosomal diseases. Despite
these therapeutic breakthroughs, both approaches have
various drawbacks.4 ERT is possible only for non-neur-
onopathic Gaucher patients since the enzyme does not
cross the blood–brain barrier. Moreover, Cerezymeꢀ
named the ÔworldÕs most expensive drugÕ is extremely
costly ($150,000 per year). Zavescaꢀ has been
Gaucher disease,1 the most common glycolipid storage
disease, is a relatively rare hereditary disorder due to a
deficiency in a specific b-glucosidase (b-glucocerebrosid-
ase) involved in the catabolism of glycosphingolipids in
lysosomes.2 Defects in the catalytic activity of this
enzyme leads to the accumulation of undegraded gluco-
sylceramide (GlcCer) in macrophages and to severe
symptoms. In type 1, the mildest and most common of
the three clinical forms of Gaucher disease, patients suf-
fer from bone pains, skeletal lesions, anaemia, thrombo-
cytopenia and liver or spleen damage. Type 1 Gaucher
disease occurs worldwide in all population (1 in
40,000–60,000) but it is the most common genetically-
based disease affecting Jews (1 in 500–1000 among Ash-
kenazi Jews).3
*
Corresponding authors. Fax: +33 (0)2 38 41 72 81; e-mail addresses:
philippe.compain@univ-orleans.fr; olivier.martin@univ-orleans.fr
0957-4166/$ - see front matter ꢁ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2005.03.015

1748
C. Boucheron et al. / Tetrahedron: Asymmetry 16 (2005) 1747–1756
O
Glucosyl
OH
ceramide
OH
HN
C15H31
C13H27
synthase
O
O
HO
HO
HO
O
ceramide-OH HO
OUDP
OH
OH
OH
GlcCer
O
N
HN
2'
C15H31
HO
5
2
OH
HO
C13H27
3'
OH
3
HO
1'
OH
N-Bu DNJ 1
IC50 20 µM
ceramide-OH 2
Scheme 1.
recommended only for adults with mild-to-moderate
type 1 Gaucher disease for whom ERT is not an option.
Large doses are required (ꢀ300 mg daily) but can lead
to serious side-effects including abdominal pains and
loss of weight arising mainly from the inhibition of
digestive glucosidases. Moreover, the fact that this com-
pound is also a strong inhibitor of glucosidase I, a cellu-
lar glycosidase involved in the elaboration of the glycan
structure of glycoproteins, is a major concern, particu-
larly for long-term treatments.6 Thus, whereas the valid-
ity of the substrate deprivation approach has been
clearly demonstrated, there is a need for more selective
new therapeutic agents.
UDP-glucose:ceramide glucosyltransferase also known
as glucosylceramide synthase catalyzes the transfer of
a glucose moiety from UDP-glucose to the primary hy-
droxyl group of ceramide to yield GlcCer with inversion
of the anomeric configuration (Scheme 1).
Figure 1. Possible overlay of ceramide and N-butyl DNJ highlighting
structural mimicry.7
tical synthesis of nojirimycin C-glycosides.9 Experi-
ments with glucosylceramide synthase were made to
assess their specificity with respect to the length (C4 or
C8) and the position of the second alkyl chain (C-1,
O-2 and/or O-4).
The study of N-butyl DNJ 1 and its derivatives has indi-
cated that this compound is, somewhat unexpectedly, a
mimic of ceramide 2 and not of the glucose moiety;7,8
this is shown in particular by the fact that the inhibition
of glucosylceramide synthase by this compound is com-
petitive with respect to ceramide and not to the UDP-
glucose. Molecular modeling has revealed a strong
structural homology between 1 and the ceramide struc-
ture. The N-alkyl chain and three stereogenic centres
(C-2, C-3 and C-5) of iminosugar 1 show structural sim-
ilarities with the N-acyl chain and the C-10–C-30 back-
bone of ceramide, respectively. Remarkably, the O-3
hydroxyl group of 1 overlays with the O-30 hydroxyl
of ceramide 2 (Fig. 1).7
2. Results and discussion
To prepare iminosugar C-glycosides 7, we applied
directly our initial strategy to N-alkylated imines 4a and
4b instead of the corresponding N-benzyl derivatives
(Scheme 2). We were pleased to find that chain extension
of these imines with butylmagnesium chloride or octyl-
magnesium bromide proceeded in high diastereoselectiv-
ity and in almost quantitative yield as judged by proton
NMR spectroscopy. Amines 5a and 5b were directly en-
gaged in the next step without purification since partial
decomposition occurred on silica gel.10 The deprotection
of the acetal function in aqueous trifluoroacetic acid,
followed by intramolecular reductive amination afford-
ed the expected piperidines 6a and 6b in high diastereo-
selectivity but poor yield. In sharp contrast with
findings in the N-benzyl series, degradation occurred
during the reductive amination step. Removal of the
benzyl protecting groups by hydrogenolysis provided
the desired iminosugar C-glycosides 7a and 7b.11
This rational model suggests very clearly that the imino-
sugar would be a better ceramide mimic if a second
alkyl chain was present at O-2 or C-1 to simulate the
second hydrophobic chain of the ceramide. The imino-
glycolipids thus designed were expected to be more po-
tent and more selective inhibitors than N-Bu DNJ 1. To
assess this working hypothesis, various di- or tri-alkyl-
ated iminosugars were prepared by taking advantage
of a general strategy we recently developed for the prac-

C. Boucheron et al. / Tetrahedron: Asymmetry 16 (2005) 1747–1756
1749
O
( )n
a,b
O
5 steps
O
O
L-sorbose
OBn
b,c,d
O
a
O
N
70%
Ref. 9
n = 2 83%
n = 6 71%
HO
OBn
3
HO
BnO
OBn
n = 2 52%
n = 6 54%
MsO
OBn
94%
OH
OBn
3
8
n = 2 9a
O
O
( )n
N
( )
O
O
O
O
n = 6 9b
n H
d,e
c
OBn
( )n
( )n
OBn
N
H
( )n
( )n
n = 2 95%
n = 6 95%
n = 2 38%
n = 6 18%
( )n
OBn
( )n
OBn
OBn
OBn
N
OBn
N
e
N
n = 2 4a
n = 6 4b
n = 2 5a
n = 6 5b
de>98%
+
HO
+
O
O
BnO
BnO
BnO
O
O
OH
( )n
( )n
( )n
( )n
OBn
N
OH
N
f
n = 2 34% 10a
n = 6 23% 10b
n = 2 27% 11a
n = 6 15% 11b
n = 2 33% 12a
n = 6 55% 12b
HO
HO
HO
BnO
n = 2 37%
n = 6 25%
OH
OH
( )n
( )n
f
f
f
n = 2 6a
n = 6 6b
de>98%
n = 2 7a
n = 6 7b
( )n
( )n
( )n
( )n
( )n
OH
N
OH
N
OH
N
HO
HO
O
O
Scheme 2. Reagents and conditions: (a) PCC (1.8 equiv), CH2Cl2,
˚
HO
HO
16 h; (b) butyl- or octylamine (1.1 equiv), molecular sieves 4 A, 4 ꢂC,
O
O
OH
( )n
16 h; (c) butylMgCl (4 equiv) or octylMgBr (3 equiv), Et2O, 0–20 ꢂC,
16 h; (d) TFA/H2O (9:1), 36 h; (e) NaBH3CN (3 equiv), AcOH
(1 equiv), MeOH, 16 h; (f) H2, Pd/C, HCl 4 M cat., MeOH (n = 2)
or THF/H2O (n = 6), 24 h.
( )n
n = 2 81% 13a
n = 6 83% 13b
n = 2 96% 14a
n = 6 72% 14b
n = 2 59% 15a
n = 6 55% 15b
Scheme 3. Reagents and conditions: (a) MsCl (1.2 equiv), Et3N
(1.2 equiv), CH2Cl2, 4 h; (b) neat butyl- or octylamine, 85 ꢂC, 16 h;
(c) TFA/H2O (9:1), 0–20 ꢂC, 28 h; (d) NaBH3CN (4.1 equiv for n = 2
or 5.2 equiv for n = 6), AcOH (1 equiv), MeOH, 48 h; (e) NaH
(6 equiv), butyl iodide (7 equiv) or octyl iodide (8 equiv), n-Bu4NI
(25 mol %), THF, D, 72 h; (f) H2, Pd/C, MeOH/HCl 5 N (10:1), 48 h
(n = 2) or 70 h (n = 6).
To prepare the analogues of 1 bearing alkyl chains at O-2
and/or O-4, we took advantage of the key step of our ini-
tial strategy.9 The one-pot sequence of acetal deprotec-
tion and intramolecular reductive amination was of
particular interest since this process provides a properly
protected piperidine with two free hydroxyl groups tacti-
cally positioned at C-2 and C-4. Mesylate derivative 8
was obtained efficiently under classical conditions from
alcohol 3. Nucleophilic substitution by alkyl amines
under neat conditions12 afforded the corresponding
aminosorbofuranose derivatives, which were directly
engaged in the next step without purification (Scheme
3). The intramolecular reductive amination of the un-
masked aminosorbose hemiketal proceeded in good
yields to provide the expected key intermediates 9a and
9b. At this stage, our aim was to functionalize or protect
regioselectively the OH-2 since it is less hindered when
compared to OH-4. After various attempts, no satisfac-
tory solutions could be found in terms of reproducibility
or practicability.13 From the finding that the two regio-
isomers 10a and 10b and 11a and 11b could be easily sep-
arated on silica gel, we intentionally performed a non-
regioselective alkylation of diols 9a and 9b using large
amounts of NaH and alkyl iodide. This method afforded
after purification three iminoglycolipids of interest
including the precursors of target compounds 13a and
13b. Finally, removal of benzyl groups by hydrogenolysis
gave, after purification on silica gel, the expected piperid-
inols in 11 steps with 4–10% overall yield from L-sorbose.
small (2.6 Hz), indicating that there is little or no rota-
tion around the C5–C6 bond: this pattern of coupling
constants is only consistent with an N5–C5–C6–O6
torsion angle of either ꢁ60ꢂ (staggered, gg) or +120ꢂ
(fully eclipsed), and only the former is sterically allowed.
The equatorial position of the octyl group at nitrogen is
confirmed by the pattern of NOE interactions between
the alkyl chain protons and H5, H6 and both H1
protons.14 The N- and O-octyl groups adopt a confor-
mation that is typical of extended alkyl chains.
With this series of iminoglycolipids, we investigated the
influence of the position and length of alkyl chains on
glucosylceramide synthase inhibition (Fig. 2). For the
purpose of comparison, the IC50 values for N-butyl
DNJ 1, and its N-nonyl analogue 16 are also included
in Figure 2. This biological evaluation indicated clearly
that the addition of a second or third lipophilic chain
does not provide better affinity for the enzyme. The al-
kyl chain-length of these iminosugars plays a decisive
role in their inhibition profile. The best inhibitions were
found for iminosugars bearing two octyl chains, 13b,
14b and 7b whereas those bearing butyl chains displayed
no significant inhibition towards glucosylceramide
synthase (13a, 14a and 15a) with the exception of com-
pound 7a (IC50 609 lM). The addition of a third lipo-
philic chain is detrimental to the inhibitory activity of
NMR analysis of iminoglycolipids 13b (Table 1) and 14b
(Table 2), indicated that the piperidine ring in CD3OD
4
adopts a classical non-distorted C1 (D) conformation.
The 3JH5,H6(R) and 3JH5,H6(S) coupling constants are very

1750
C. Boucheron et al. / Tetrahedron: Asymmetry 16 (2005) 1747–1756
OH
OH
N
OH
( )n
N
( )n
O
( )n
( )n
HO
HO
HO
O
n = 2 13a IC50 no inhibition
n = 6 13b IC50 164 µM
n = 2 14a no inhibition
n = 6 14b IC50 134 µM
OH
N
( )n
HO
HO
OH
n = 2 1 IC50 20 µM5c
n = 7 16 IC50 1.6 µM5c
OH
OH
N
N
O
( )n
( )n
( )n
HO
HO
O
HO
OH
( )n
( )n
n = 2 7a IC50 609 µM
n = 6 7b IC50 174 µM
n = 2 15a 29% inhibition at 1 mM
n = 6 15b no inhibition
Figure 2. Inhibition of glucosylceramide synthase.
iminosugars 13b and 14b but has a slightly positive effect
for the corresponding analogues bearing shorter butyl
chains. Interestingly, comparison of the IC50 values be-
tween di-alkylated iminosugars 13b, 14b and 7b showed
that the position of the second alkyl chain (a-C-1, O-2 or
O-4) has almost no influence on their inhibitory profile.
mide in the solid state (Fig. 1). It appears therefore that
in spite of the flexibility of the alkyl substituents, these
chains do not fold easily to adopt a conformation resem-
bling that of ceramide.
Further studies are still yet to be performed toevaluate the
specificity of di-alkyl iminoglycolipids 13b and 14b as
GlcCer synthase inhibitors, since these compounds are
expected to be weaker inhibitors of glucosidases than 1.
According to our initial hypothesis, compounds having
a closer structural homology to ceramide than N-butyl
DNJ 1 were expected to be better inhibitors of the
synthase. Inhibition results obtained for iminoglycoli-
pids 13a and 13b, 14a and 14b and 7a and 7b indicated
however that this hypothesis does not hold since the best
ceramide mimics were sevenfold less potent than N-but-
yl DNJ. This can be understood in terms of the relative
orientation of the alkyl chains on the piperidine scaffold:
in their most favourable conformation in solution, as
established by NMR for 13b (Table 1) and 14b (Table
2), the alkyl chains are oriented in remote directions
(Fig. 3), unlike the nearly parallel alkyl chains of cera-
3. Conclusion
In conclusion, we have achieved the synthesis of original
iminosugars carrying two or more lipophilic chains
(ÔiminoglycolipidsÕ) from L-sorbose. These analogues of
N-butyl DNJ 1 were designed as ceramide mimetics
and evaluated as inhibitors of glucosylceramide syn-
thase. This study showed however that the addition of
a second alkyl chain does not improve their activity.
Moreover, the glucosylceramide synthase active site
does not discriminate inhibitors differing by the position
of the second alkyl chain (a-C-1, O-2 or O-4). Further
work will focus on the confirmation and consequence
of these findings for the design of new iminosugar-based
glucosylceramide synthase inhibitors as potential thera-
peutic agents for Gaucher disease.15
4. Experimental
4.1. General
Unless otherwise stated, all reactions requiring anhy-
drous conditions were carried out under Argon. THF
Figure 3. Solution conformation of 13b.

C. Boucheron et al. / Tetrahedron: Asymmetry 16 (2005) 1747–1756
1751
and diethyl ether were freshly distilled from sodium/ben-
zophenone under Argon prior to use. Dichloromethane
was distilled from calcium hydride. Infrared spectra were
recorded using films on NaCl windows or KBr pellets.
Low-resolution mass spectra (MS) were recorded with
a Perkin–Elmer Sciex API 3000 in the ion spray (IS)
mode. High-resolution mass spectra (HRMS) were re-
corded with a Micromass ZABSpec TOF in the electro-
spray ionization (ESI) mode and with a Finnigan MAT
95 XL in the chemical ionization (CI) mode. Optical
rotations were measured at room temperature (20 ꢂC)
in a 1 dm cell with a Perkin–Elmer 241 polarimeter.
Analytical thin layer chromatography was performed
using silica gel 60F254 precoated plates (Merck) with
visualization by ultraviolet light, phosphomolybdate
solution (2% in H2SO4/EtOH 1:7) and/or exposure to
I2 vapour. Flash chromatography was performed on sil-
ica gel 60 (230–400 mesh) with ethyl acetate (AcOEt) and
petroleum ether (PE) as eluants unless indicated other-
(d, 1H, J = 5.4 Hz); 13C NMR (CDCl3): d 14.1, 26.6,
26.9, 27.5, 29.3, 29.5, 30.5, 61.2, 70.1, 72.0, 73.7, 82.2,
82.3, 84.6, 97.4, 112.6, 114.6, 127.3, 127.6, 127.8,
128.3, 128.4, 137.4, 138.1, 161.7; MS-IS m/z: 510.0
[MH]+.
4.4. (6R)-1,4-Di-O-benzyl-6-butylamino-6-C-butyl-6-
deoxy-2,3-O-isopropylidene-a-L-sorbofuranose 5a
To a solution of 2.0 M butylmagnesium chloride
(661 lL, 1.32 mmol) in anhydrous Et2O at 0 ꢂC, was
added slowly
a 0.15 M solution of 4a (150 mg,
0.33 mmol) in anhydrous Et2O. The mixture was slowly
warmed up to room temperature and stirred overnight
(16 h). The reaction was quenched by the slow addition
of saturated aqueous NH4Cl at room temperature. The
layers were separated, the aqueous phase extracted with
Et2O and the organic layer dried over Na2SO4 and con-
centrated under reduced pressure. The colourless oily
1
wise. H and 13C NMR spectra were recorded at 25 ꢂC
product 5a (160 mg, 95%) was used without further puri-
on a Bruker DPX 250 Advance (250 MHz) spectrometer
with Me4Si as internal reference unless otherwise stated,
and J values are quoted in hertz. Carbon multiplicities
were assigned by distortionless enhancement by polariza-
tion transfer (DEPT) experiments.
20
fication. ½aꢂ ¼ þ76:0 (c 1.0, CHCl3); IR (NaCl, neat)
D
1
1610, 1205, 1085, 733, 697 cmꢁ1; H NMR (CDCl3): d
0.89 (m, 7H), 1.33 (m, 7H), 1.43 (s, 3H), 1.52 (s, 3H),
2.58 (m, 2H), 2.97 (m, 1H), 3.37 (br s, 1H), 3.63 (d,
1H, J = 11.0 Hz), 3.76 (d, 1H, J = 11.0 Hz), 3.88 (m,
1H), 4.18 (dd, 1H, J = 9.0, 2.7 Hz), 4.37 (d, 1H,
J = 12.0 Hz), 4.62 (m, 4H), 7.26–7.31 (m, 10H); 13C
NMR (CDCl3): d 14.09, 14.13, 20.6, 23.2, 26.8, 27.7,
27.8, 29.8, 32.5, 46.4, 56.2, 70.5, 71.6, 73.7, 81.5, 81.9,
82.9, 112.2, 113.5, 127.7, 128.0, 128.1, 128.4, 128.5,
137.4, 138.4; MS-IS m/z: 512.0 [MH]+.
4.2. 3,6-Di-O-benzyl-4,5-O-isopropylidene-D-xylo-alde-
hydo-hexos-5-ulo-2,5-furanose N-butylimine 4a
To a stirred solution of 3,6-di-O-benzyl-4,5-O-isopropyl-
idene-D-xylo-aldehydo-hexos-5-ulo-2,5-furanose9 (200 mg,
0.5 mmol) in CH2Cl2 (4.5 mL) were added powdered
˚
4 A molecular sieves (50 mg) and butylamine (64.5 lL,
4.5. (6R)-1,4-Di-O-benzyl-6-deoxy-2,3-O-isopropylidene-
6-C-octyl-6-octylamino-a-L-sorbofuranose 5b
0.65 mmol, 1.3 equiv) at 0 ꢂC. Stirring was stopped
and the mixture allowed to stand for 16 h at 0 ꢂC. The
solids were removed by filtration and washed with anhy-
drous CH2Cl2 (10 mL). The filtrate was concentrated in
vacuo to afford homogeneous 4a (233 mg, quant.) as a
yellow oil. This compound being very moisture sensitive,
was used without further purification. IR (NaCl, neat)
To a solution of 2.0 M octylmagnesium bromide
(750 lL, 1.47 mmol) in anhydrous Et2O at 0 ꢂC was
slowly added
a 0.15 M solution of 4b (249 mg,
0.49 mmol) in anhydrous Et2O. The mixture was then
warmed up slowly to room temperature, and stirred
overnight (16 h). Work-up as described above afforded
5b as a colourless oil (290 mg, 95%). IR (NaCl, neat)
1673 cmꢁ1 1H NMR (CDCl3): d 0.92 (m, 4H), 1.35
;
(m, 3H), 1.44 (s, 3H), 1.53 (s, 3H), 1.58 (m, 2H), 3.45
(t, 1H, J = 6.9 Hz), 3.75 (d, 1H, J = 11.0 Hz), 3.84 (d,
1H, J = 11.0 Hz), 4.18 (d, 1H, J = 3.4 Hz), 4.43 (d, 1H,
J = 12.0 Hz), 4.61 (d, 2H, J = 12.0 Hz), 4.74 (d, 1H,
J = 12.0 Hz), 4.81 (m, 1H), 7.21–7.37 (m, 10H), 7.68
(d, 1H, J = 5.4 Hz); MS-IS m/z: 454.0 [MH]+.
1
1612, 1209, 1085, 738, 692 cmꢁ1; H NMR (CDCl3): d
0.88 (m, 10H), 1.26 (m, 20H), 1.42 (s, 3H), 1.51 (s,
3H), 2.54 (m, 2H), 2.93 (m, 1H), 3.63 (m, 2H), 3.74 (d,
1H, J = 11.0 Hz), 3.86 (d, 1H, J = 3.0 Hz), 4.15 (dd,
1H, J = 9.0, 2.9 Hz), 4.35 (d, 1H, J = 12.0 Hz), 4.61
(m, 4H), 7.26–7.32 (m, 10H); 13C NMR (CDCl3): d
14.1, 22.6, 25.4, 25.7, 26.7, 27.4, 27.6, 29.2, 29.4,
29.7, 30.1, 30.5, 31.8, 32.8, 46.7, 56.0, 62.9, 70.4,
71.5, 73.6, 81.4, 81.8, 83.0, 112.0, 113.4, 127.5, 127.6,
127.8, 128.3, 128.4, 137.3, 138.3; MS-IS m/z: 625.0
[MH]+.
4.3. 3,6-Di-O-benzyl-4,5-O-isopropylidene-D-xylo-alde-
hydo-hexos-5-ulo-2,5-furanose N-octylimine 4b
3,6-Di-O-benzyl-4,5-O-isopropylidene-D-xylo-aldehydo-
hexos-5-ulo-2,5-furanose9 (200 mg, 0.5 mmol) was treat-
ed with octylamine (107 lL, 0.64 mmol) as described
for 4a. The crude product was obtained as a yellow oil
and used without further purification. Compound 4b
4.6. (1R)-3,6-Di-O-benzyl-1-C-butyl-N-butyl-1,5-dide-
oxy-1,5-imino-D-glucitol 6a
(268 mg, quant.): IR (NaCl, neat) 1674 cmꢁ1
;
1H
NMR (CDCl3): d 0.87 (m, 4H), 1.26 (m, 10H), 1.42 (s,
3H), 1.51 (s, 3H), 1.58 (m, 2H), 3.41 (t, 2H,
J = 6.9 Hz), 3.73 (d, 1H, J = 11.0 Hz), 3.82 (d, 1H,
J = 11.0 Hz), 4.15 (d, 1H, J = 3.4 Hz), 4.42 (d, 1H, J =
12.0 Hz), 4.57 (d, 2H, J = 12.0 Hz), 4.71 (d, 1H,
J = 12.0 Hz), 4.79 (m, 1H), 7.18–7.35 (m, 10H), 7.68
A ꢀ0.15 M solution of precursor 5a (47 mg, 0.09 mmol)
in trifluoroacetic acid was prepared at 0 ꢂC; H2O was
then added to form a 9:1 (v/v) TFA/H2O mixture. The
reaction mixture was warmed up to room temperature
and stirred for 36 h. The solvents were removed by three

1752
C. Boucheron et al. / Tetrahedron: Asymmetry 16 (2005) 1747–1756
co-evaporations with toluene and the residual product
taken up in MeOH to form a ꢀ0.05 M solution. Acetic
acid (1 equiv) and NaBH3CN (3 equiv) were added at
0 ꢂC. The reaction mixture was stirred for 24 h at room
temperature. The solvents were evaporated under re-
duced pressure. The crude product was taken up in
CH2Cl2 and the solution was washed with saturated
aqueous NaHCO3 then with water. The organic layer
was dried over Na2SO4 and concentrated under reduced
pressure. The resulting crude product was purified by sil-
4.9. (1R)-1-C-Octyl-N-octyl-1,5-dideoxy-1,5-imino-D-
glucitol 7b
To a solution of 6b (40 mg, 0.07 mmol) in THF/H2O 4:1
(2 mL) was added 10% Pd/C (0.2 equiv) and four drops
of 4 M aqueous HCl. Work-up and purification proce-
dure as described for 7a provided 7b (7 mg, 25%) as a
colourless oil. 1H NMR (CD3OD): d 0.89 (m, 5H),
1.31–1.42 (m, 28H), 2.63 (m, 3H), 2.95 (m, 1H), 3.39
(m, 4H), 3.75 (m, 4H); 13C NMR (CD3OD): d 20.7,
29.9, 30.9, 31.0, 34.6, 35.6, 36.7, 36.9, 37.0, 37.2, 37.9,
39.2, 39.3, 55.0, 55.2, 66.9, 67.5, 68.0, 78.0, 79.0, 83.2;
HRMS-CI m/z 388.3424 [M+H]+ (C22H45NO4 required
388.3427).
ica gel chromatography (PE/AcOEt 8:2–7:3) to provide
20
D
6a (16 mg, 38%) as a colourless oil. ½aꢂ ¼ þ38:5 (c
1
1.0, CHCl3); H NMR (CDCl3): d 0.91 (m, 7H), 1.42
(m, 11H), 2.54 (m, 2H), 2.94 (m, 2H), 3.47 (t, 1H,
J = 8.5 Hz), 3.79 (m, 4H), 4.54 (s, 2H), 4.70 (d, 1H,
J = 11.5 Hz), 4.98 (d, 1H, J = 11.5 Hz), 7.27–7.37 (m,
10H); 13C NMR (CDCl3): d 14.0, 14.1, 21.1, 22.8,
24.3, 29.5, 32.4, 48.5, 57.1, 59.3, 69.2, 71.5, 73.3, 73.5,
74.6, 84.0, 127.70, 127.73, 127.80, 127.84, 128.4, 128.5,
137.5, 138.8; HRMS-CI m/z 456.31102 [M+H]+
(C28H42NO4 required 456.3113).
4.10. 1,4-Di-O-benzyl-2,3-O-isopropylidene-6-O-methane-
sulfonyl-a-L-sorbofuranose 8
To a solution of 3 (4.32 g, 10.8 mmol) in CH2Cl2
(32 mL) was added Et3N (1.8 mL, 12.9 mmol) and MsCl
(1 mL, 12.9 mmol). The mixture was then stirred at
room temperature. After 4 h, the reaction mixture was
washed with water (40 mL). The organic layer was dried
over MgSO4 and concentrated under reduced pressure.
The resulting crude product was purified by silica gel
4.7. (1R)-3,6-Di-O-benzyl-1-C-octyl-N-octyl-1,5-dide-
oxy-1,5-imino-D-glucitol 6b
Compound 5b (60 mg, 0.08 mmol) was treated as de-
scribed for 5a. The resulting crude product was purified
by silica gel chromatography (PE/AcOEt 8:2) to provide
chromatography (PE/AcOEt 3:1) to provide 8 (4.87 g,
20
94%) as a colourless oil. ½aꢂ ¼ þ10:0 (c 1.2, CHCl3);
D
1H NMR (250 MHz, CDCl3): d 1.42 (s, 3H), 1.51 (s,
3H), 2.95 (s, 3H), 3.62 (d, 1H, J = 11.0 Hz), 3.74 (d,
1H, J = 11.0 Hz), 4.01 (d, 1H, J = 3.2 Hz), 4.35–4.41
(m, 3H), 4.51–4.58 (m, 3H), 4.62–4.70 (m, 2H), 7.22–
7.30 (m, 10H); 13C NMR (62.9 MHz, CDCl3): d 26.6,
27.6, 37.5, 67.8, 70.0, 71.9, 73.7, 78.5, 81.4, 81.5, 112.7,
114.3, 127.6, 127.7, 127.8, 128.2, 128.4, 128.6, 136.9,
138.0; MS-IS m/z: 496.5 [M+NH4]+.
20
6b (8 mg, 18%) as a colourless oil. ½aꢂ ¼ þ23:5 (c 1.0,
D
1
CHCl3); H NMR (CDCl3): d 0.86 (m, 7H), 1.28 (m,
27H), 2.53 (t, 2H, J ꢀ 7.0 Hz), 2.93 (m, 2H), 3.45 (t,
1H, J = 8.5 Hz), 3.76 (m, 4H), 4.51 (s, 2H), 4.67 (d,
1H, J = 11.5 Hz), 4.98 (d, 1H, J = 11.5 Hz), 7.27–7.35
(m, 10H); 13C NMR (CDCl3): d 14.1, 22.6, 24.5, 27.0,
27.3, 29.2, 29.3, 29.4, 29.5, 29.6, 29.8, 30.0, 31.8, 31.9,
48.9, 57.1, 59.2, 69.1, 71.5, 73.2, 73.5, 74.6, 83.9, 127.7,
127.8, 127.9, 128.4, 128.5, 137.5, 138.8; HRMS-CI m/z
568.43663 [M+H]+ (C36H58NO4 required 568.4365).
4.11. 3,6-Di-O-benzyl-N-butyl-1,5-dideoxy-1,5-imino-D-
glucitol 9a
4.8. (1R)-N-Butyl-1-C-butyl-1,5-dideoxy-1,5-imino-D-
glucitol 7a
A solution of 8 (4.7 g, 9.82 mmol) in n-butylamine
(50 mL, 506 mmol) was stirred overnight (16 h) at
85 ꢂC. The reaction mixture was concentrated in vacuo
taken up in AcOEt (150 mL) and the solution washed
with water (100 mL) and saturated aqueous NaCl
(100 mL). The organic layer was dried over MgSO4
and concentrated under reduced pressure. The resulting
product was directly used in the next step since partial
decomposition of the intermediate amine occurred on
silica gel. A ꢀ0.15 M solution of the aminosorbofura-
nose obtained (4.3 g, 9.5 mmol) in trifluoroacetic acid
was prepared at 0 ꢂC; H2O was then added to form a
9:1 (v/v) TFA/H2O mixture. The reaction mixture was
warmed up to room temperature and stirred for 28 h.
The solvents were removed by co-evaporation with tol-
uene and the residual product taken up in MeOH to
To a solution of 6a (37 mg, 0.08 mmol) in MeOH
(4 mL) was added 10% Pd/C (0.2 equiv) and four drops
of 4 M aqueous HCl. The flask was purged three times
with Ar and then filled with H2. After 24 h, the solids
were removed by filtration and the filtrate concentrated
under reduced pressure. The crude product was filtered
through Amberlyst ion-exchange resin IRA-400(OHꢁ)
(elution with H2O) and the filtrate concentrated under
reduced pressure. The residual product was purified
on an Amberlyst ion-exchange resin IR-120(H+) (wash-
ing with H2O, elution with 5% aqueous ammonia); the
fractions containing the product were pooled and con-
centrated under reduced pressure to afford 7a (8 mg,
20
37%) as a colourless oil. ½aꢂ ¼ þ27:5 (c 0.7, CHCl3);
form
a
ꢀ0.05 M solution. Acetic acid (0.55 mL,
D
1H NMR (CD3OD): d 0.96 (m, 6H), 1.41 (m, 10H),
1.71 (m, 1H), 2.90 (m, 3H), 3.18 (m, 1H), 3.32 (m,
3H), 3.49 (m, 2H), 3.73 (m, 1H), 3.88 (d, 2H,
J = 4.4 Hz); 13C NMR (CD3OD): d 14.5, 21.5, 24.0,
24.7, 31.7, 33.1, 48.9, 60.7, 61.4, 61.8, 71.8, 72.9, 77.0;
HRMS-CI m/z 276.2178 [M+H]+ (C14H29NO4 required
276.2176).
9.6 mmol) and NaBH3CN (2.45 g, 39.1 mmol) were then
added at 0 ꢂC. The reaction mixture was stirred for 48 h
at room temperature. The mixture was then concen-
trated under reduced pressure. The crude product was
taken up in CH2Cl2 (500 mL) and the solution was
washed with saturated aqueous NaHCO3 (2 · 250 mL)
then with water (250 mL). The organic layer was dried

C. Boucheron et al. / Tetrahedron: Asymmetry 16 (2005) 1747–1756
1753
over MgSO4 and concentrated under reduced pressure.
The resulting crude product was purified by silica gel
chromatography (PE/AcOEt 1:4) to provide 9a (2.23 g,
4.13.1. 3,6-Di-O-benzyl-N-butyl-2-O-butyl-1,5-dideoxy-
1,5-imino-D-glucitol 10a. Rf 0.3 (PE/AcOEt 4:1);
20
D
½aꢂ ¼ ꢁ15:0 (c 0.9, CHCl3); 1H NMR (250 MHz,
20
52%) as a yellowish solid. ½aꢂ ¼ ꢁ8:0 (c 3.3, CHCl3);
CDCl3): d 0.89 (m, 6H), 1.15–1.28 (m, 2H), 1.33–1.42
(m, 4H), 1.50–1.61 (m, 2H), 2.20 (t, 1H, J = 10.7 Hz),
2.31 (m, 1H), 2.56–2.74 (m, 3H), 3.09 (dd, 1H, J = 4.4,
11.3 Hz), 3.20 (t, 1H, J = 8.8 Hz), 3.43–3.69 (m, 6H),
4.52 (s, 2H), 4.69 (d, 1H, J = 11.6 Hz), 4.97 (d, 1H,
J = 11.6 Hz), 7.27–7.36 (m, 10H); 13C NMR
(62.9 MHz, CDCl3): d 14.0, 14.1, 19.4, 20.7, 26.2, 32.5,
52.5, 54.1, 63.5, 66.7, 70.1, 70.5, 73.5, 74.6, 78.6, 85.7,
127.7, 127.8, 128.0, 128.1, 128.4, 128.6, 138.2, 139.1;
HRMS-ESI m/z 456.3114 [M+H]+ (C28H42NO4 required
456.3114), m/z 478.2931 [M+Na]+ (C28H41NO4Na re-
quired 478.2933).
D
1H NMR (250 MHz, CDCl3):
d
0.86 (t, 3H,
J = 7.2 Hz), 1.20 (m, 2H), 1.36 (m, 2H), 2.24 (t, 1H,
J = 10.3 Hz), 2.37 (m, 1H), 2.50 (m, 1H), 2.65 (m, 1H),
3.00 (dd, 1H, J = 4.7, 11.3 Hz), 3.16 (t, 1H,
J = 8.5 Hz), 3.60–3.80 (m, 4H), 4.53 (s, 2H), 4.79 (d,
1H, J = 11.6 Hz), 4.88 (d, 1H, J = 11.6 Hz), 7.26–7.38
(m, 10H); 13C NMR (62.9 MHz, CDCl3): d 14.1, 20.7,
26.6, 52.5, 56.1, 63.8, 67.7, 69.6, 72.1, 73.6, 74.5, 87.2,
128.0, 128.1, 128.5, 128.8, 137.9, 138.9; HRMS-ESI
m/z 400.2480 [M+H]+ (C24H34NO4 required 400.2488),
m/z 422.2293 [M+Na]+ (C24H33NO4Na required
422.2307).
4.13.2. 3,6-Di-O-benzyl-N-butyl-4-O-butyl-1,5-dideoxy-
1,5-imino-D-glucitol 11a. Rf 0.4 (PE/AcOEt 1:1);
4.12. 3,6-Di-O-benzyl-N-octyl-1,5-dideoxy-1,5-imino-D-
glucitol 9b
20
D
½aꢂ ¼ þ18:0 (c 1.1, CHCl3); 1H NMR (250 MHz,
CDCl3): d 0.88 (m, 6H), 1.21–1.50 (m, 8H), 2.23 (t,
1H, J = 10.9 Hz), 2.39 (m, 1H), 2.48–2.71 (m, 3H),
3.00 (dd, 1H, J = 4.4, 11.3 Hz), 3.24 (t, 1H,
J = 8.1 Hz), 3.37–3.46 (m, 2H), 3.55–3.77 (m, 4H), 4.52
(s, 2H), 4.66 (d, 1H, J = 11.3 Hz), 4.86 (d, 1H,
J = 11.6 Hz), 7.29–7.34 (m, 10H); 13C NMR
(62.9 MHz, CDCl3): d 14.1, 14.2, 19.6, 20.7, 26.8, 32.6,
52.5, 54.9, 63.6, 65.2, 69.3, 72.5, 73.6, 74.5, 78.5, 86.2,
127.8, 127.9, 128.4, 128.7, 138.0, 138.9; HRMS-ESI
m/z 456.3104 [M+H]+ (C28H42NO4 required 456.3114),
m/z 478.2924 [M+Na]+ (C28H41NO4Na required
478.2933).
The n-octyl analogue of 9a, compound 9b, was obtained
from 8 (2.7 g, 5.64 mmol) by the same sequence of reac-
tion as described above, n-butylamine being replaced by
n-octylamine (20 mL, 121 mmol), with the difference that
octylamine could only be partially removed from the
intermediate aminosorbofuranose by co-evaporation
with toluene. The reductive amination process was per-
formed in the presence of residual octylamine, and the
resulting product was purified by silica gel chromatogra-
phy (PE/AcOEt 2:1–1:1) to provide 9b (1.38 g, 54%) as a
20
1
yellowish solid. ½aꢂ ¼ ꢁ4:5 (c 1.4, CHCl3); H NMR
D
(250 MHz, CDCl3): d 0.88 (t, 3H, J = 6.9 Hz), 1.23–
1.39 (m, 12H), 2.24 (t, 1H, J = 10.3 Hz), 2.37 (m, 1H),
2.49 (m, 1H), 2.63 (m, 1H), 3.01 (dd, 1H, J = 4.4,
11.0 Hz), 3.17 (t, 1H, J = 8.5 Hz), 3.59–3.77 (m, 4H),
4.53 (s, 2H), 4.79 (d, 1H, J = 11.9 Hz), 4.88 (d, 1H,
J = 11.6 Hz), 7.25–7.37 (m, 10H); 13C NMR
(62.9 MHz, CDCl3): d 14.2, 22.8, 24.5, 27.6, 29.4, 29.6,
31.9, 52.8, 56.1, 63.8, 67.8, 69.7, 72.2, 73.7, 74.6, 87.2,
128.0, 128.1, 128.5, 128.8, 137.9, 138.9; HRMS-ESI m/z
456.3111 [M+H]+ (C28H42NO4 required 456.3114), m/z
478.2937 [M+Na]+ (C28H41NO4Na required 478.2933).
4.13.3. 3,6-Di-O-benzyl-N-butyl-2,4-di-O-butyl-1,5-dide-
oxy-1,5-imino-D-glucitol 12a. Rf 0.75 (PE/AcOEt 4:1);
20
D
½aꢂ ¼ þ2:0 (c 2.4, CHCl3); 1H NMR (250 MHz,
CDCl3): d 0.88 (m, 9H), 1.19–1.57 (m, 12H), 2.14 (t,
1H, J = 10.9 Hz), 2.20 (m, 1H), 2.52–2.70 (m, 2H),
3.06 (dd, 1H, J = 4.7, 11.0 Hz), 3.22–3.38 (m, 3H),
3.39–3.47 (m, 1H), 3.51–3.60 (m, 4H), 3.76–3.85 (m,
1H), 4.52 (s, 2H), 4.74 (d, 1H, J = 11.3 Hz), 4.89 (d,
1H, J = 11.0 Hz), 7.25–7.35 (m, 10H); 13C NMR
(62.9 MHz, CDCl3): d 14.0, 14.1, 14.2, 19.4, 19.5, 20.8,
25.7, 32.5, 32.7, 52.3, 54.6, 63.9, 65.5, 70.6, 73.2, 73.5,
75.2, 78.6, 78.9, 87.3, 127.4, 127.8, 127.9, 128.3, 128.4,
128.5, 138.0, 139.4; HRMS-ESI m/z 512.3740 [M+H]+
(C32H50NO4 required 512.3740).
4.13. Preparation of protected di- and tri-butyl imino-
sugars 10a, 11a and 12a
To a solution of 9a (3 g, 7.5 mmol) in THF (135 mL) at
0 ꢂC was added slowly NaH (60% dispersion in oil,
1.85 g, 46.2 mmol). After 30 min at room temperature,
1-iodobutane (6 mL, 52.7 mmol) and tetrabutylammo-
nium iodide (0.59 g, 1.59 mmol) were added and the
reaction mixture heated under reflux for 3 days. The
reaction was quenched by the slow addition of MeOH
at 0 ꢂC and the mixture concentrated in vacuo. The
crude mixture was taken up in CH2Cl2 (300 mL) and
the solution washed with water (2 · 150 mL) and satu-
rated aqueous NaCl (150 mL). The organic layer was
dried over MgSO4 and concentrated under reduced pres-
sure. The resulting products were separated by silica gel
chromatography (PE/AcOEt 5:1–1:1) to provide in the
following order of elution: 12a (1.25 g, 33%), 10a
(1.16 g, 34%) and 11a (0.91 g, 27%) as slightly yellow
oils.
4.14. Preparation of protected di- and tri-octyl iminosu-
gars 10b, 11b and 12b
Compound 9b (1.72 g, 3.78 mmol) was treated as de-
scribed for 9a. The resulting products were separated
by silica gel chromatography (PE/AcOEt 8:1–6:1) to
provide in the following order of elution: 12b (1.40 g,
55%), 10b (0.48 g, 23%) and 11b (0.32 g, 15%) as yellow-
ish oils.
4.14.1. 3,6-Di-O-benzyl-N-octyl-2-O-octyl-1,5-dideoxy-
1,5-imino-D-glucitol 10b. Rf 0.2 (PE/AcOEt 6:1);
20
½aꢂ ¼ ꢁ8:5 (c 0.6, CHCl3); 1H NMR (250 MHz,
D
CDCl3): d 0.87 (m, 6H), 1.15–1.39 (m, 22H), 1.51–1.59

1754
C. Boucheron et al. / Tetrahedron: Asymmetry 16 (2005) 1747–1756
20
D
(m, 2H), 2.21 (t, 1H, J = 10.7 Hz), 2.31 (m, 1H), 2.53–
2.68 (m, 3H), 3.09 (dd, 1H, J = 4.4, 11.3 Hz), 3.20 (t,
1H, J = 8.7 Hz), 3.43–3.69 (m, 6H), 4.53 (s, 2H), 4.69
(d, 1H, J = 11.6 Hz), 4.97 (d, 1H, J = 11.6 Hz), 7.25–
7.33 (m, 10H); 13C NMR (62.9 MHz, CDCl3): d 14.2,
22.8, 24.1, 26.3, 27.6, 29.3, 29.4, 29.5, 29.6, 30.4, 31.9,
52.8, 54.1, 63.6, 66.7, 70.5, 70.56, 73.6, 74.7, 78.6, 85.7,
127.7, 127.8, 128.0, 128.1, 128.4, 128.6, 138.2, 139.1;
MS-IS m/z 568.5 [M+H]+.
tography (AcOEt/MeOH 5:1). ½aꢂ ¼ þ12:0 (c 0.8,
MeOH); 1H NMR (250 MHz, CD3OD): d 0.92 (m,
6H), 1.23–1.60 (m, 8H), 2.12–2.18 (m, 2H), 2.66 (m,
1H), 2.80 (m, 1H), 3.12 (dd, 1H, J = 4.4, 11.2 Hz),
3.22 (m, 2H), 3.37 (t, 1H, J = 8.8 Hz), 3.59 (t, 2H,
J = 6.8 Hz), 3.84 (d, 2H, J = 2.2 Hz); 13C NMR
(62.9 MHz, CD3OD): d 14.2, 14.3, 20.2, 21.6, 27.0,
33.3, 53.5, 54.9, 58.9, 66.9, 71.4, 71.7, 78.7, 79.3;
HRMS-ESI m/z 276.2172 [M+H]+ (C14H30NO4 required
276.2175), m/z 298.1998 [M+Na]+ (C14H29NO4Na re-
quired 298.1994).
4.14.2. 3,6-Di-O-benzyl-N-octyl-4-O-octyl-1,5-dideoxy-
1,5-imino-D-glucitol 11b. Rf 0.1 (PE/AcOEt 6:1);
20
D
½aꢂ ¼ þ12:0 (c 1.1, CHCl3); 1H NMR (250 MHz,
4.17. N-Butyl-4-O-butyl-1,5-dideoxy-1,5-imino-D-glucitol
14a
CDCl3): d 0.88 (t, 6H, J = 6.9 Hz), 1.25–1.51 (m, 24H),
2.23 (dd, 1H, J = 9.4, 11.0 Hz), 2.39 (m, 1H), 2.47–
2.67 (m, 2H), 3.01 (dd, 1H, J = 4.4, 11.1 Hz), 3.24 (t,
1H, J = 8.2 Hz), 3.35–3.46 (m, 2H), 3.53–3.73 (m, 4H),
4.52 (s, 2H), 4.66 (d, 1H, J = 11.6 Hz), 4.87 (d, 1H,
J = 11.6 Hz), 7.25–7.34 (m, 10H); 13C NMR (62.9
MHz, CDCl3): d 14.2, 22.8, 24.6, 26.4, 27.6, 29.4, 29.6,
29.7, 30.5, 31.9, 52.8, 54.9, 63.6, 65.2, 69.4, 72.8, 73.6,
74.5, 78.6, 86.2, 127.8, 127.9, 128.4, 128.7, 138.1,
138.9; MS-IS m/z 568.5 [M+H]+.
Compound 11a (232 mg, 0.51 mmol) was submitted to
general procedure A, which provided 14a (134.5 mg,
96%) as a colourless oil after purification by silica gel
20
D
chromatography (AcOEt/MeOH 5:1). ½aꢂ ¼ ꢁ4:5 (c
0.9, MeOH); 1H NMR (250 MHz, CD3OD): d 0.89
(m, 6H), 1.24–1.40 (m, 4H), 1.41–1.57 (m, 4H), 2.27
(m, 2H), 2.67 (m, 1H), 2.87 (m, 1H), 3.03 (dd, 1H,
J = 4.7, 11.3 Hz), 3.22 (m, 2H), 3.46 (m, 1H), 3.58
(m, 1H), 3.73 (dd, 1H, J = 1.9, ꢀ11.9 Hz), 3.82 (over-
lapping signals, m, 1H), 3.86 (m, 1H); 13C NMR
(62.9 MHz, CD3OD): d 14.2, 14.3, 20.3, 21.5, 27.0,
33.5, 53.4, 56.7, 57.7, 66.6, 70.2, 73.9, 79.1, 80.1;
HRMS-ESI m/z 276.2174 [M+H]+ (C14H30NO4 re-
quired 276.2175).
4.14.3. 3,6-Di-O-benzyl-N-octyl-2,4-di-O-octyl-1,5-dide-
oxy-1,5-imino-D-glucitol 12b. Rf 0.6 (PE/AcOEt 7:1);
20
D
½aꢂ ¼ þ4:0 (c 0.8, CHCl3); 1H NMR (250 MHz,
CDCl3): d 0.88 (m, 9H), 1.24–1.57 (m, 36H), 2.13
(t, 1H, J = 10.7 Hz), 2.20 (m, 1H), 2.50–2.73 (m,
2H), 3.05 (dd, 1H, J = 4.7, 11.3 Hz), 3.25 (t, 1H,
J = 9.1 Hz), 3.27–3.37 (m, 2H), 3.39–3.46 (m, 1H),
3.51–3.63 (m, 4H), 3.75–3.84 (m, 1H), 4.52 (s, 2H),
4.74 (d, 1H, J = 11.0 Hz), 4.88 (d, 1H, J = 11.3 Hz),
7.25–7.38 (m, 10H); 13C NMR (62.9 MHz, CDCl3):
d 14.2, 22.8, 23.6, 26.3, 26.4, 27.7, 29.3, 29.4, 29.6,
29.7, 30.5, 30.6, 31.9, 32.0, 52.6, 54.7, 63.9, 65.6,
71.0, 73.6, 75.2, 78.7, 78.9, 87.3, 127.4, 127.8, 127.9,
128.3, 128.4, 128.6, 138.1, 139.4; MS-IS m/z: 681.0
[M+H]+.
4.18. N-Butyl-2,4-di-O-butyl-1,5-dideoxy-1,5-imino-D-
glucitol 15a
Compound 12a (292 mg, 0.57 mmol) was submitted to
general procedure A, which provided 15a (111 mg,
59%) as a yellowish solid after purification by silica
20
D
gel chromatography (AcOEt/MeOH 9:1). ½aꢂ ¼ þ15:0
(c 1.1, MeOH); 1H NMR (250 MHz, CD3OD): d
0.96 (m, 9H), 1.29–1.64 (m, 12H), 2.04–2.12 (m,
2H), 2.63 (m, 1H), 2.81 (m, 1H), 3.10 (dd, 1H,
J = 4.1, 10.7 Hz), 3.17–3.35 (m, 3H), 3.63 (t, 3H,
J = 6.6 Hz), 3.76 (dd, 1H, J = 1.9, 11.6 Hz), 3.86
(dd, 1H, J = 1.9, 11.9 Hz), 3.95 (m, 1H); 13C NMR
(62.9 MHz, CD3OD): d 14.3, 14.4, 20.2, 20.3, 21.7,
27.1, 33.4, 33.6, 53.4, 55.0, 58.5, 66.4, 71.5, 73.9, 79.4,
79.8, 79.9; HRMS-ESI m/z 332.2797 [M+H]+
(C18H38NO4 required 332.2801), m/z 354.2609
[M+Na]+ (C18H37NO4Na required 354.2620).
4.15. General procedure A for the synthesis of N-alkyl-
1,5-dideoxy-1,5-imino-D-glucitol derivatives 13, 14 and 15
To a ꢀ0.04 M (butyl chains) or ꢀ0.02 M (octyl chains)
solution of precursor 10, 11 or 12 in a 10:1 (v/v)
MeOH/HCl 5 M mixture was added 10% Pd/C
(ꢀ0.2 equiv). The flask was purged three times with Ar
then filled with H2. The reaction mixture was stirred at
room temperature. After 48 h (butyl chains) or 70 h (oct-
yl chains), the solids were removed by filtration and the
filtrate concentrated under reduced pressure. The result-
ing crude product was purified by silica gel
chromatography.
4.19. N-Octyl-2-O-octyl-1,5-dideoxy-1,5-imino-D-glucitol
13b
Compound 10b (64 mg, 0.112 mmol) was submitted to
general procedure A, which provided 13b (36 mg, 83%)
as a white solid after purification by silica gel chroma-
20
4.16. N-Butyl-2-O-butyl-1,5-dideoxy-1,5-imino-D-glucitol
13a
tography (AcOEt/MeOH/NH4OH 10:2:1.5). ½aꢂ ¼
D
þ 5:5 (c 0.6, MeOH); H NMR see Table 1; 13C NMR
(62.9 MHz, CD3OD): d 14.4, 23.7, 25.0, 27.1, 28.6,
30.4, 30.5, 30.6, 31.2, 32.9, 33.0, 53.8, 55.1, 59.3, 67.1,
71.7, 72.0, 79.1, 79.5; HRMS-ESI m/z 388.3425
[M+H]+ (C22H46NO4 required 388.3427).
1
Compound 10a (284 mg, 0.624 mmol) was submitted to
general procedure A, which provided 13a (140 mg, 81%)
as a colourless oil after purification by silica gel chroma-

C. Boucheron et al. / Tetrahedron: Asymmetry 16 (2005) 1747–1756
Table 2. NMR assignments for 14b (see footnotes to Table 1)
1755
Table 1. NMR assignments for 13ba
Position
14b
OD
d (ppm)
3JHH (Hz)
4
2
6'
6
C16-22
1eq
1ax
2.881
2.042
ꢁ11.2/5.0
N
DO
ꢁ11.2/10.4
10.4/5.0/8.6
1eq
15'
DO
15
O
2
3
4
5
3.355
3.127
3.098
2.022
5
1ax
7'
3
C8-15
8.6/8.9
7
8.9/8.9
Positionb
13b
8.9/2.3/2.4
d (ppm)c
3JHH (Hz)
6
3.744
3.646
ꢁ11.8/2.3
ꢁ11.8/2.4
8.9/6.5
8.9/7.0
60
1eq
1ax
3.001
2.002
ꢁ11.0/4.3
ꢁ11.0/9.9
4.3/9.9/8.9
7
70
3.810
3.514
2
3
4
5
3.117
3.085
3.258
2.002
8
1.494
1.22
o/l
o/l
7.0
8.9/9.0
9–13
14
9.0/9.3
0.803
9.3/2.6/2.6
15
150
2.689
2.478
13.4/8.1
13.4/7.6
6
3.748
3.734
ꢁ12.0/2.6
ꢁ12.0/2.6
60
7
70
3.52
3.49
o/ld
o/l
16
1.382
1.22
o/l
o/l
7.0
17–21
22
8
1.476
1.22
o/l
o/l
7.1
0.803
9–13
14
20
0.805
tography (PE/AcOEt/NH4OH 4:1:0.5). ½aꢂ ¼ þ7:5 (c
D
0.6, MeOH); 1H NMR (250 MHz, CD3OD): d 0.90
(m, 9H), 1.31–1.59 (m, 36H), 2.01–2.09 (m, 2H), 2.61
(m, 1H), 2.76 (m, 1H), 3.07 (dd, 1H, J = 4.1, 10.7 Hz),
3.15–3.26 (m, 3H), 3.60 (t, 3H, J = 6.6 Hz), 3.74 (dd,
1H, J = 1.9, 11.6 Hz), 3.84 (dd, 1H, J = 1.9, 11.9 Hz),
3.91 (m, 1H); 13C NMR (62.9 MHz, CD3OD): d 14.5,
23.7, 24.9, 27.2, 27.3, 28.7, 30.4, 30.5, 30.6, 30.7, 31.2,
31.5, 33.0, 53.7, 55.1, 58.5, 66.5, 71.8, 74.3, 79.4, 79.9,
80.0; HRMS-ESI m/z 500.4678 [M+H]+ (C30H62NO4 re-
quired 500.4679), m/z 522.4504 [M+Na]+ (C30H61NO4-
Na required 522.4498).
15
150
2.692
2.494
13.5/9.0/7.3
13.5/8.9/6.5
16
1.388
1.22
o/l
o/l
7.1
17–21
22
0.805
a 1H NMR spectra were recorded in CD3OD at 30 ꢂC (500 MHz).
b C1–C6 is DNJ; C7–C14 is the octyl group attached to oxygen; C15–
C22 is the octyl group attached to the ring nitrogen.
c Chemical shifts are measured relative to the methyl peak of tri-
methylsilylpropanesulfonic acid at 0.0 ppm.
d o/l indicates difficulty in assignment due to overlapping peaks.
4.22. Conformational analysis and modelization
4.20. N-Octyl-4-O-octyl-1,5-dideoxy-1,5-imino-D-glucitol
14b
Samples for NMR analysis were prepared by dissolving
approximately 6 mg of solid compound in CD3OD. One
and two-dimensional 1H NMR spectra were run at
500 MHz on a Varian Unity Inova spectrometer and
an additional two-dimensional NOESY spectrum at
750 MHz on a home-built/GE Omega spectrometer.
All spectra were recorded at 30 ꢂC. The mixing time
used for the 2D NOESY spectra was 200 ms. Molecular
modeling was performed on a Silicon Graphics Fuel
workstation, using the programs Insight II and Discover
Compound 11b (49 mg, 0.086 mmol) was submitted to
general procedure A, which provided 14b (24 mg, 72%)
as a yellowish solid after purification by silica gel chro-
20
D
matography (AcOEt/MeOH/NH4OH 10:2:1.5). ½aꢂ
¼
ꢁ 3:5 (c 0.4, MeOH); 1H NMR, see Table 2;
13C NMR (62.9 MHz, CD3OD): d 14.4, 23.7, 25.1,
27.3, 28.6, 30.4, 30.6, 30.7, 31.4, 33.0, 53.6, 57.5, 58.4,
66.7, 70.9, 74.2, 79.9, 80.7; HRMS-ESI m/z 388.3427
[M+H]+ (C22H46NO4 required 388.3427).
3
(Accelrys Inc., San Diego, USA). Accurate JHH cou-
pling constants were obtained by fitting a simulated
1
1D H NMR spectrum to the experimental spectrum
4.21. N-Octyl-2,4-di-O-octyl-1,5-dideoxy-1,5-imino-D-
glucitol 15b
using the program gNMR (Cherwell Scientific, Oxford,
UK). Figures were produced using the program Mol-
script.16 The X-ray crystal structure of glucosylcera-
mide17 (Fig. 1) was obtained from the Cambridge
Crystallographic Database18 at the Chemical Database
Service at Daresbury.19
Compound 12b (62 mg, 0.0908 mmol) was submitted to
general procedure A, which provided 15b (25 mg, 55%)
as a colourless oil after purification by silica gel chroma-

1756
C. Boucheron et al. / Tetrahedron: Asymmetry 16 (2005) 1747–1756
7. Butters, T. D.; van den Broek, L. A. G. M.; Fleet, G. W.
4.23. Biochemical assays
J.; Krulle, T. M.; Wormald, M. R.; Dwek, R. A.; Platt, F.
M. Tetrahedron: Asymmetry 2000, 11, 113.
Biochemical assays were conducted according to the
procedure previously described.7
8. For reviews on carbohydrate mimetics-based glycosyl-
transferase inhibitors, see: (a) Compain, P.; Martin, O. R.
Bioorg. Med. Chem. 2001, 9, 3077; (b) Compain, P.;
Martin, O. R. Curr. Top. Med. Chem. 2003, 3, 541.
9. Godin, G.; Compain, P.; Masson, G.; Martin, O. R. J.
Org. Chem. 2002, 67, 6960.
Acknowledgements
Financial support of this study by grants from CNRS
and the association ÔVaincre les Maladies LysosomalesÕ
is gratefully acknowledged. C.B. thanks the French
ministry of research for a fellowship and the COST
program (D28 action) for a Short-Term Scientific Mis-
sion grant. We thank Michael Deschamps, Dept. of Bio-
chemistry, Oxford University, for help in running the
750 MHz NMR experiments. Support from the Oxford
Glycobiology Institute is gratefully acknowledged.
10. The corresponding N-benzylated analogues of 5 are stable
on silica gel; see Ref. 9.
11. Low isolated yields were due to solubility problems during
the purification step using ion-exchange resins.
12. Elliott, R. D.; Brockman, R. W.; Montgomery, J. A. J.
Med. Chem. 1986, 29, 1052.
13. The diphenylacetyl group could be introduced highly
selectively at O-2 in modest yield (ꢀ40%). Further
elaboration of this product towards the desired 2-O-
alkylated derivative proved difficult.
14. Main NOE interactions observed in 14b:
References
1. Zimran, A. GaucherÕs Disease; Balliere Tindall, 1997.
2. A special issue of Philos. Trans.: Biol. Sci. has been
devoted to glycolipids in cell biology and medicine: Philos.
Trans. R. Soc. London, Ser. B 2003, 358, 845–983.
3. Horowitz, M.; Zimram, A. Hum. Mutat. 1994, 3, 1.
4. Futerman, A. H.; Sussman, J. L.; Horowitz, M.; Silman,
I.; Zimran, A. Trends Pharm. Sci. 2004, 25, 147.
5. (a) Cox, T.; Lachmann, R.; Hollak, C.; Aerts, J.; van
Weely, S.; Hrebicek, M.; Platt, F.; Butters, T.; Dwek, R.;
Moyses, C.; Gow, I.; Elstein, D.; Zimran, A. Lancet 2000,
355, 1481; (b) Butters, T. D.; Dwek, R. A.; Platt, F. M.
Chem. Rev. 2000, 100, 4683; (c) Butters, T. D.; Dwek, R.
A.; Platt, F. M. Curr. Top. Med. Chem. 2003, 3, 561; (d)
Cox, T. M.; Aerts, J. M. F. G.; Andria, G.; Beck, M.;
Belmatoug, N.; Bembi, B.; Chertkoff, R.; Vom Dahl, S.;
Elstein, D.; Erikson, A.; Giralt, M.; Heitner, R.; Hollak,
C.; Hrebicek, M.; Lewis, S.; Mehta, A.; Pastores, G. M.;
Rolfs, A.; Sa Miranda, M. C.; Zimran, A. J. Inherit.
Metab. Dis. 2003, 26, 513.
15. A special issue of Current Topics in Medicinal Chemistry
has been devoted to iminosugars of therapeutic relevance:
Iminosugars: Recent Insights Into Their Bioactivity and
Potential as Therapeutic Agents. In Curr. Top. Med.
Chem.; Martin, O. R., Compain, P., Eds.; Bentham:
Netherlands, 2003; Vol. 3, issue 5.
16. Kraulis, P. J. J. Appl. Crystallogr. 1991, 24, 945.
17. Pascher, I.; Sundell, S. Chem. Phys. Lipids 1977, 20,
175.
6. (a) Mellor, H. R.; Platt, F. M.; Dwek, R. A.; Butters, T.
D. Biochem. J. 2003, 374, 307; (b) Mellor, H. R.; Nolan,
J.; Pickering, L.; Wormald, M. R.; Platt, F. M.; Dwek, R.
A.; Fleet, G. W.; Butters, T. D. Biochem. J. 2002, 366, 225.
18. Allen, F. H.; Kennard, O. Chem. Design Autom. News
1993, 8, 31.
19. Fletcher, D. A.; McMeeking, R. F.; Parkin, D. J. Chem.
Inf. Comput. Sci. 1996, 36, 746.