Iminosugars as glycosyltransferase inhibitors: Synthesis of polyhydroxypyrrolidines and their evaluation on chitin synthase activity

Article abstract of DOI:10.1016/j.ejmech.2005.07.001

Chitin synthase is an enzyme involved in the biosynthesis of chitin, a major structural component of the cell wall of many fungi. Since chitin is absent in vertebrates, chitin synthase has been envisaged as a valuable target in the search for new antifung

Full text of DOI:10.1016/j.ejmech.2005.07.001

European Journal of Medicinal Chemistry 40 (2005) 1255–1261
www.elsevier.com/locate/ejmech
Original article
Iminosugars as glycosyltransferase inhibitors: synthesis of
polyhydroxypyrrolidines and their evaluation on chitin synthase activity
Isabelle Gautier-Lefebvre, Jean-Bernard Behr *, Georges Guillerm, Murielle Muzard
Laboratoire Réactions Sélectives et Applications UMR 6519, UFR Sciences - CNRS, BP 1039, 51687 Reims cedex 2, France
Received 6 November 2004; received in revised form 23 June 2005; accepted 1 July 2005
Available online 10 August 2005
Abstract
Chitin synthase is an enzyme involved in the biosynthesis of chitin, a major structural component of the cell wall of many fungi. Since
chitin is absent in vertebrates, chitin synthase has been envisaged as a valuable target in the search for new antifungal agents. In this report, a
series of C-2 substituted polyhydroxypyrrolidines were designed and synthesized with the aim of mimicking the glycosylcation involved at
the transition state of the enzymatic reaction governed by chitin synthase. Some of these models displayed chitin synthase inhibition in the
millimolar range. However, no significant antifungal activity was noted on a panel of fungal strains.
© 2005 Elsevier SAS. All rights reserved.
Keywords: Chitin synthase; Glycosyltransferase; Inhibition; Iminosugars; Difluoromethylphosphonoazasugars
1. Introduction
nucleotide-diphosphate product in the enzymatic assay
resulted in potent synergistic effect, which reflects a better
recognition of both the inhibitors taken together than each
inhibitor itself [8]. In addition, some iminosugars, recently
described, displayed affinities for glycosyltransferases in the
micromolar range [7]. Among these, pyrrolidinol 1 (Fig. 1)
had Ki = 61 µM against b-1,4-galactosyltransferase from
bovine milk [9] and the natural 6-deoxy-homoDMDP 2 dis-
played strong inhibition of chitin synthase (Ki = 38 µM) [10].
Surprisingly, these compounds did not compete with the natu-
ral substrate but were shown to act as uncompetitive inhibi-
tors. However, these pioneering results prompted us to design
and develop new series of iminosugar-based glycosyltrans-
ferase inhibitors.
The rational design of potent glycosyltransferase inhibi-
tors has attracted increasing interest during the last few years
due to the pivotal role of these enzymes in the biosynthesis of
polysaccharides and glycoconjugates [1]. Thus, inhibition of
glycosyltransferases is a new approach for controlling inva-
sive fungal infections [2], xenotransplantation rejection [3],
inflammatory processes [4] or glycosphingolipid storage dis-
orders like the Gaucher disease [5]. Since glycosyltransfer
reactions are thought to proceed through transition states simi-
lar to those of glycosidases [6], iminosugars have been envis-
aged as potential inhibitors. These small molecules must be
able to mimic, at the transition state, the charge and shape of
the carbohydrate moiety of the sugar-nucleotide generally
involved in these biological processes.
Accordingly, the rational design of new chitin synthase
(EC 2.4.1.16) inhibitors [2] as antifungal agents led us to
Among the iminosugars and related compounds which
have been assayed against several glycosyltransferases [7],
simple aza-pyranoses and furanoses displayed generally com-
petitive inhibition towards the sugar-nucleotide donor with
Ki’s in the 10 mM range. Interestingly, Quiao et al. [8] dem-
onstrated that the combination of an azasugar with the
* Corresponding author. Tel.: +33 3 26 91 32 38; fax: +33 3 26 91 31 66.
E-mail address: jb.behr@univ-reims.fr (J.-B. Behr).
Fig. 1. Potent azasugar-type glycosyltransferase inhibitors.
0223-5234/$ - see front matter © 2005 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2005.07.001

1256
I. Gautier-Lefebvre et al. / European Journal of Medicinal Chemistry 40 (2005) 1255–1261
and was shown to exhibit potent inhibition of a-glucosidases
[11]. We describe here a straightforward synthetic route
to 4, starting from N-benzyl-2,3,5-tri-O-benzyl-L-
xylofuranosylamine 10 [12]. As reported in Fig. 3, addition
of vinylmagnesium bromide to 10 gave the syn aminoalcool
11 in a very stereoselective manner (no trace of the anti iso-
mer could be detected in the 1H and 13C NMR spectra) in an
excellent yield (91%). The configuration at the newly formed
stereocenter could be firmly assigned by NOE experiments
on the corresponding vinylpyrrolidine 12 obtained by treat-
ment of 11 with methanesulfonyl chloride. Ozonolysis was
then performed on the sulfate salt of 12 to avoid oxidation of
the free amine. Reduction of the crude aldehyde with NaBH4
gave alcohol 13 in 56% yield (Fig. 3).Alternatively, the chemi-
cal transformation of 12–13 was also performed by bishy-
droxylation of 12 with osmium tetroxide in presence of
N-methylmorpholine N-oxide, followed by NaIO4 treatment
in an EtOH/H2O solution and subsequent NaBH4 reduction.
Though the latter procedure was more time consuming, bet-
ter yields (83%) could be obtained. The benzyl protecting
groups were removed by extensive hydrogenolysis over Pd–C
to afford 4 after purification by silica gel chromatography
(CHCl3/MeOH/NH4OH, 60:30:5, v/v).
In addition to their potential as potent and specific glycosi-
dase or glycosyltransferase inhibitors there is also consider-
able interest in using polyhydroxylated pyrrolidines, in par-
ticular those featuring C2 symmetry, as chiral auxiliaries in
asymmetric synthesis [13].A former synthesis of 2,5-dideoxy-
2,5-imino-L-iditol 7 was performed with this aim in view [14].
In our study, azasugar 7 was synthesized in good yield start-
ing from the known vinylpyrrolidine 14 [15], prepared in two
steps from commercial N-benzyl-2,3,5-tri-O-benzyl-D-
arabinofuranosylamine (Fig. 3). The synthetic procedure
described above for the preparation of 4 was applied to 14
and permitted to isolate pure 7. The 1H and 13C NMR spec-
tral data of compounds 4 and 7 as well as their optical rota-
tions were similar to the previously reported ones [11,14].
Fig. 2. Design of pyrrolidinols 4–9 as chitin synthase inhibitors.
envisage polyhydroxypyrrolidines 4–9 as possible mimics of
the glycosylcation involved at the transition state of this
N-acetylglucosamine-transfer reaction (Fig. 2). Several ste-
reoisomers were targeted to investigate the influence of the
stereochemistry at position C-2 or C-5. Thus, the D-arabino
pyrrolidinols 4–6 might fit the configuration and the half-
chair conformation of the D-gluco-configurated transition
state. On the other hand, compounds 7–9 apparented to
L-xylose, could induce an additional electrostatic interaction,
the C-5 hydroxymethyl substituent replacing the NHAc group
in a putative NH/enzyme hydrogen bond (Fig. 2). In addi-
tion, incorporation of a difluoromethylphosphonate group,
mimicking the first portion of the pyrophosphate moiety of
the donor, might exert binding forces increasing the inhibi-
tory potencies. We report here the synthesis and kinetic evalu-
ation of compounds 4–9 as inhibitors of chitin synthase as
well as their bioevaluation against a variety of pathogenic
fungi for antifungal activities.
2.2. Synthesis of difluoromethylphosphono-azasugars 5, 6,
8, 9 (Fig. 4)
In a previous paper we described the general synthetic
approach to protected difluoromethylphosphono pyrro-
lidines [16]. Compounds 16–19 were prepared according to
this procedure and we investigated here the experimental con-
ditions leading to deprotected phosphonoazasugars 5, 6, 8, 9
(Fig. 4). On the basis of previous work emphasizing the more
efficient intracellular transport of mono-phosphonate esters
when compared to their phosphonic acid counterparts [17],
we intended to remove only one ethyl group in order to retain
the lipophilic character of these antifungal candidates. Thus,
the targeted monophosphonates might be obtained in two
steps, i.e. hydrogenolysis of benzyl groups and mono-
saponification of the so-obtained diethylphosphonates. How-
ever, the debenzylation step proved to be fairly difficult. Only
the N-benzyl protecting group was cleaved by standard treat-
2. Chemistry
2.1. Synthesis of iminosugars 4, 7
Pyrrolidinol 4 (2,5-dideoxy-2,5-imino-D-glucitol) had been
synthesized at first in 1991 by a chemoenzymatic approach

I. Gautier-Lefebvre et al. / European Journal of Medicinal Chemistry 40 (2005) 1255–1261
1257
Fig. 3
. Synthesis of azasugars 4, 7.
Reagents and conditions: (i) H2C=CH–MgBr, THF; (ii) MsCl, pyr.; (iii) a) H2SO4, O3, b) NaBH4; (iv) H2, Pd–C, HCl.
lytic step was not required in the case of isomers 5 and 8
which were directly isolated after the debenzylation proce-
dure. In this latter case, it is assumed that an in situ intramo-
lecular transesterification occurred with assistance of the
C-3 free hydroxyl group leading to a cyclic phosphonate ester
which was readily hydrolized to the isolated monoesters.
3. Biological evaluation
3.1. Enzymology
Saccharomyces cerevisiae (X2180 strains) chitin syn-
thase 1 activity was assayed on permeabilized cells by mea-
suring the rate of formation of [14C]-chitin from UDP-N-
acetyl-[14C]-glucosamine. The initial permeabilization of
freshly grown cells was performed by osmotic shock, accord-
ing to the procedure described by Crotti et al. [19].
Since chitin synthase 1 is present as its zymogen in the
permeabilized cells, it must be activated prior to assay by
partial proteolysis with trypsin and treatment with digitonin.
Thus, an aliquot of permeabilized cell suspension (to a final
concentration of 78 mg/ml) was incubated for 15 min in
30 mM Tris–HCl (pH 6.5) containing 55 mM GlcNAc, digi-
tonin (5.2 mg/ml) and trypsin (1.0 mg/ml). Activation was
stopped by adding soybean trypsin inhibitor (1.5 mg/ml). The
resulting preparation was kept at 0 °C for assay.
Fig. 4. Synthesis of difluoromethylphosphono-azasugars 5, 6, 8, 9.
Reagents and conditions: (i) HCO2NH4, 10% Pd/C, MeOH, 60 °C; (ii) 10%
NH4OH.
ment with Pd-C under hydrogen atmosphere. The use of black
Pd as catalyst, AcOH as solvent or high pressure of hydrogen
gave only partial deprotection. It is assumed that the second-
ary amine, formed during the process, might inhibit the cata-
lytic properties of Pd [18]. Nevertheless, Pd(0)-catalyzed
transfer hydrogenation in refluxing methanol, using ammo-
nium formate as the hydrogen donor, allowed the cleavage of
all the benzyl groups. The a-configurated mono-phosphonate
esters 6 and 9 were obtained after subsequent mono-
saponification in aqueous ammonia. Surprisingly, this hydro-
The enzymatic assays were then carried out as reported in
Section 6.
3.2. Antifungal evaluation
Antifungal activity of compounds 4–9 was tested on patho-
genic fungi (Cryptococcus neoformans, Trichophyton men-
tagrophytes, Microsporum canis, Aspergillus fumigatus, Can-
dida albicans, S. cerevisiae). MIC’s were determined in broth

1258
I. Gautier-Lefebvre et al. / European Journal of Medicinal Chemistry 40 (2005) 1255–1261
macrodilution assays according to guidelines of the NCCLS
M27-A protocol [20], with slight modifications [21]. Inocula
for assays were prepared from stocks frozen at –80 °C by
dilution in growth medium to give a final viable cell count of
2.5 × 103 CFU/ml. Each assay was performed with a dupli-
cate series of drug dilutions. For each fungal strain, the assays
were done in RPMI 1640 medium buffered to pH 7.0 with
MOPS buffer, incubated at 35 °C for 48 h according to the
described procedure [21]. C. neoformans and A. fumigatus
were tested in the same assay, except that incubations were
for 72 h. The dermatophytes T. mentagrophytes and M. canis
were incubated at 30 °C for 7 days.
whereas isomers 5 and 7 were completely inactive at the high-
est concentration tested. These data also point out important
differences between parent iminosugars. Thus, compound 7
was totally inactive at the highest concentration tested (5 mM)
whereas the epimer 4 inhibited enzyme activity by 50% at
the same concentration. This effect was also observed for the
phosphonoazasugars since compound 9 (IC50 = 1.6 mM) was
significantly more active than its C-5 epimer 6 (IC50 = 38 mM)
or its C-2 epimer 8 (IC50 = 4 mM).
Unfortunately, incorporation of a difluoromethylphospho-
nate moiety in the structure did not markedly increase the
inhibitory potencies. Indeed, conflicting results might be
noticed when comparing the affinities of pyrrolidinol 7 and
its phosphonated analogue 8 (increase in inhibition) or com-
pounds 4 and 5 (decrease in inhibition). Furthermore, the
results obtained with phosphonoazasugar 6 were particularly
disappointing since, in our minds, the structural feature of 6
was thought to reproduce at best the first portion of the pyro-
phosphate moiety as well as the configurations at C-1, C-3,
C-4 and C-5 of UDP-GlcNAc. Clearly, among the com-
pounds tested, best results were obtained with the L-xylo con-
figurated difluoromethylphosphono azasugar 9.
4. Results and discussion
The reaction rate of chitin synthase showed normal
Michaelis–Menten kinetics and a Km value of 0.62 mM was
determined for UDP-GlcNAc (Fig. 5), which is in good agree-
ment with previously published data [22]. In addition, the
chitin synthase 1 activity was confirmed by positive control
with nikkomycin Z.
The inhibitory activities of iminosugars 4–9 were exam-
ined at 1 mM of substrate UDP-GlcNAc and the results are
reported in Table 1. For all the pyrrolidinols tested, an IC50
value (i.e. the concentration leading to half the original rate)
was determined using Dixon plots, by assaying at least five
drug dilutions of each compound.
From the data obtained in Table 1, the following conclu-
sions can be drawn: azasugars 4, 6, 8, 9 lowered significantly
chitin synthase activity with IC50 ranging from 1.6 to 38 mM,
As noticed above, inhibition of glycosidases or glycosyl-
transferases by iminosugars is commonly attributed to the
ability of mimicking both the conformational and the elec-
tronic features of the oxocarbenium ion involved at the tran-
sition state, in particular through protonation of the amine at
physiological pH. However, the potent electronic effect of
the difluoromethylene substituent might significantly influ-
ence the basicity of polyhydroxypyrrolidines 5, 6, 8, 9 and
thus prevent the requisite protonation. For this purpose we
determined the dissociation constants for the fully depro-
tected difluoromethylphosphonic acid 20 prepared from 18
by treatment with bromotrimethylsilane and subsequent cleav-
age of the benzyl protecting groups (Fig. 6).
The pKa values were evaluated with an automatic titrator
using a 0.1 N sodium hydroxide solution. Among the three
constants thus determined, two might be attributed to the phos-
phonic acid (1.73–3.71) and one to the ammonium salt (6.64).
In
consequence,
the
state
of
ionization
of
difluoromethylphosphono-azasugars 5, 6, 8, 9 at pH 6.5 (stan-
dard pH of the enzymatic assay) differs significantly from
that of “simple” iminosugars of type 4 or 7 (pKa ca. 7.2). The
lower basicity of azasugars 5, 6, 8, 9 might thus explain, at
least in part, the moderate affinities of these models for chitin
synthase.
Fig. 5. Km determination for UDP-GlcNAc.
Table 1
Inhibition of chitin synthase with iminosugars 4–9
Compounds
CS1 inhibitionIC50 (mM)
4
5
6
7
8
9
5.7 0.8
NIa at 8 mM
38
4
NIa at 5 mM
4.0 0.5
1.6 0.2
a No inhibition.
Fig. 6. Dissociation constants of difluoromethylphosphonic acid 20.

I. Gautier-Lefebvre et al. / European Journal of Medicinal Chemistry 40 (2005) 1255–1261
1259
In addition to the enzymatic evaluation, antifungal activ-
ity of compounds 4–9 was also tested on a panel of several
pathogenic fungi (C. neoformans, T. mentagrophytes, M.
canis, A. fumigatus, C. albicans, S. cerevisiae). Unfortu-
nately, azasugars 4–9 showed no activity up to the maximum
tested concentration of 128 µg/ml, whereas the standard com-
pound terbinafine showed the expected activity [21]. Modest
chitin synthase inhibition potencies as well as poor lipophilic
character of the tested compounds might explain the absence
of antifungal activity.
ing mixture was left to react at rt for 7 h. Saturated NH4Cl
was then added and the solution was extracted three times
with Et2O. The combined organic layers were dried (MgSO4),
evaporated and the crude aminoalcool was purified by FC
(EtOAc/petroleum ether: 4:6; Rf = 0.5) to give pure 11 as a
1
yellow oil (2.83 g, 91%). H NMR (CDCl3) 7.30–7.20 (m,
20H, Ar-H), 5.70 (ddd, J 18.8 2 × 9.4, CH=CH2), 5.20 (dd, J
9.4 1.0, 1H, CH=CH2), 5.10 (dd, J 18.8 1.0, 1H, CH=CH2),
4.70–4.40 (m, 6H, 3xCH2Ph), 4.10 (dd, J 6.2 5.0, 1H, H-6),
3.75 (d, J 6.2, 1H, H-5), 3.80–3.50 (m, 6H). – 13C NMR
(CDCl3) 139.0–138.0 (Ar-C), 138.2 (CH=CH2), 130.0–
127.0 (Ar-C), 117.4 (CH=CH2), 81.0 and 75.8 (C-4 and C-5),
78.9 , 73.6 and 73.1 (3 × CH2Ph), 70.5 (C-7), 69.2 (C-6), 58.5
(C-3), 54.2 (NCH2Ph).
5. Conclusion
We
designed
and
synthesized
new
6.1.2. (2S,3R,4R,5R)-N-benzyl-3,4-dibenzyloxy-5-[(benzy-
loxy)methyl]-2-vinyl-pyrrolidine 12
difluoromethylphosphono-azasugars 5, 6, 8, 9 as well as
known pyrrolidinols 4, 7 as chitin synthase inhibitors. Some
of these models displayed encouraging enzyme inhibition with
IC50 in the millimolar range. However, no antifungal activity
was observed against a panel of pathogenic fungal strains.
Furthermore, introduction of a difluoromethylphosphonate
moiety in the structure, mimicking the first portion of the pyro-
phosphate moiety of the donor, did not improve the binding
affinities. Due to the electronic effect of this substituent, only
partial protonation of the pyrrolidine ring might occur in situ,
which precludes efficient recognition of compounds 5, 6, 8, 9
as transition state analogues. However, these pyrrolidinols
with unprecedented substitution pattern, might be exploited
as valuable tools for inhibition of parent glycosyltransferases
utilizing a sugar-nucleotide donor.
To a stirred solution of aminoalcool 11 (4.76 g, 8.86 mmol)
in pyridine (13 ml) at 0 °C, was added methanesulfonyl chlo-
ride (1.3 ml, 1.1 eq) and the resulting mixture was left to
react at rt for 16 h. The reaction was quenched with water
and the solution was extracted twice with CH2Cl2. The com-
bined organic layers were dried (MgSO4), evaporated and the
crude pyrrolidine was purified by FC (EtOAc/petroleum ether:
2:8; Rf = 0.7) to give pure 12 as a yellow oil (3.80 g, 80%).
1
[␣]D = +30.1 (c 2.14, CHCl3). – H NMR (CDCl3) 7.40–
7.10 (m, 20H, Ar-H), 6.00 (ddd, J 18.8 2 × 9.4, CH=CH2),
5.30 (dd, 1H, J 9.4 1.0, CH=CH2), 5.25 (dd, J 18.8 1.0,
CH=CH2), 4.60–4.20 (m, 6H, 3 × CH2Ph), 4.00 (d, 1H, JAB
12.5, NCH2Ph), 3.88 (d, 1H, J 2.2, H-4), 3.79 (d, 1H, J 4.6,
H-3), 3.6 (d, 1H, JAB 12.5, NCH2Ph), 3.46 (m, 1H, H-2), 3.37
(dd, 1H, J 9.2 8.4, CH2OBn), 3.12 (dd, 1H, J 8.4 4.5,
CH2OBn), 3.05 (m, 1H, J 8.4 4.5, H-5). – 13C NMR (CDCl3)
139.0–137.0 (Ar-C), 136.0 (CH=CH2), 129.6–126.9 (Ar-C),
118.4 (CH=CH2), 84.6 (C-3), 83.6 (C-4), 72.9 and 71.7
(2 × CH2Ph), 71.6 (C-6), 70.8 (CH2Ph), 69.9 (C-2), 67.8
(C-5), 56.7 (NCH2Ph). – ESI-MS (NH3) m/z 520 [(M + H)+].
6. Experimental protocols
6.1. Chemistry
Merck silica gel F254 (0.2 mm) was used for TLC plates,
detection being carried out by spraying with an alcoholic solu-
tion of phosphomolybdic acid or an aqueous solution of
KMnO4 (2%)/Na2CO3 (4%), followed by heating. Flash col-
umn chromatography (FC) was performed on silica gel Merck
9385 (40–63 µM) Kieselgel 60. NMR spectra were recorded
on a Brücker AC 250 spectrometer (250 MHz for 1H,
62.5 MHz for 13C). Chemical shifts are expressed in parts per
million from TMS (1H and 13C) as internal standard. Cou-
pling constants are in Hz and splitting pattern abbreviations
are: b, broad; s, singlet; d, doublet; t, triplet; q, quartet; m,
multiplet. Optical rotations were determined with a Perkin–
Elmer Model 241 polarimeter. Mass spectra were recorded
on a Jeol D 30 spectrometer, at 70 eV.
6.1.3. (2S,3R,4R,5R)-N-benzyl-3,4-dibenzyloxy-5-[(benzy-
loxy)methyl]-2-hydroxymethyl pyrrolidine 13
Concentrated H2SO4 was added drop by drop to a solution
of alkene 12 in Et2O at 0 °C to form the insoluble sulfate salt.
The supernatant was then carefully removed and the residual
orange oil was washed twice with 5 ml Et2O, dissolved in
50 ml CH2Cl2 and subjected to ozonolysis at –50 °C. After
25 min, the reaction was quenched with Me2S and the solu-
tion neutralized with 20 ml of sat. NaHCO3. The organic layer
was dried (MgSO4), concentrated and dissolved in EtOH
(10 ml).At 0 °C, NaBH4 (225 mg) was added and to the result-
ing solution was stirred 1 h at rt.A saturated solution of NH4Cl
was then added and the organic layer was separated, dried
and concentrated to give crude 13 which was purified by FC
(EtOAc/petroleum ether: 2:8; Rf = 0.5) and isolated as a yel-
lowish oil (450 mg, 56%, yellow oil). [␣]D = –21.8 (c 0.68,
CHCl3). – 1H NMR (CDCl3) 7.40–7.10 (m, 20H, Ar-H), 4.66
(d, 1H, JAB 12.0, CH2Ph), 4.58 (d, 1H, JAB 12.0, CH2Ph),
6.1.1. (3S,4R,5S,6S)-3-benzylamino-4,5,7-tribenzyloxy-6-
hydroxyhept-1-ene 11
To a stirred solution of glycosylamine 10 (3 g, 5.9 mmol)
in THF (20 ml) at 0 °C, was added vinylmagnesium bromide
(35 ml of a commercial 1 M solution, 35 mmol) and the result-

1260
I. Gautier-Lefebvre et al. / European Journal of Medicinal Chemistry 40 (2005) 1255–1261
4.56 (d, 1H, JAB 12.0, CH2Ph), 4.50 (d, 1H, JAB 12.0, CH2Ph),
4.39 (s, 2H, CH2Ph), 4.18 (t, 1H, J 2 × 5.0, H-4), 4.10 (dd,
1H, J 5.0 6.7, H-3), 3.86 (s, 2H, NCH2Ph), 3.60 (dd, 1H, J
2.7 11.2, 1 × CH2OH), 3.40 (m, 2H, 1 × CH2OH and
1 × CH2OBn), 3.32 (dd, 1H, J 4.0 10.0, 1 × CH2OBn), 3.18
(ddd, 1H, J 2.7 4.7 6.7, H-2), 3.05 (ddd, 1H, J 4.0 4.5 5.0,
H-5). – 13C NMR (CDCl3) 139.0–137.0 (Ar-C), 129.3–127.6
(Ar-C), 83.9 and 82.7 (C-3 and C-4), 73.5 (OCH2Ph), 72.6
(OCH2Ph), 72.5 (CH2OBn), 70.4 (OCH2Ph), 67.1 (C-5), 65.0
(C-2), 60.7 (CH2OH), 59.2 (NCH2Ph). – ESI-MS (NH3) m/z
524 [(M + H)+].
as a colorless precipitate. [␣]D = +16.8 (c 0.5, H2O)[14]
[␣]D = +14.3 (c 0.93, H2O). – 1H NMR (D2O) 4.10 (d, 2H, J
3.7, H-3, H-4), 3.68 (dd, 2H, J 6.7 10.3, 2 × CH2OH), 3.57
(dd, 2H, J 6.7 10.3, 2 × CH2OH), 3.36 (dt, 2H, J 3.7 6.7 6.7,
H-2, H-5). – 13C NMR (D2O) 77.1, 65.4, 60.0. – ESI-MS
(NH3) m/z 186 [(M + Na)+], 164 [(M + H)+].
6.1.7.1. General procedure for the debenzylation of phospho-
noazasugars. A solution of each protected phosphonoaza-
sugar 16–19 (0.44 g, 0.65 mmol) in MeOH (9 ml) was refluxed
for 1 h in presence of 10% Pd–C (0.891 mg) and HCO2NH4
(2 × 0.286 g). The resulting mixture was filtered over a celite
pad and the solvent was evaporated. Crude compounds 5 and
8 were obtained and purified at this stage by chromato-
graphic separation on a hydrophobic HP20SS support (elu-
ent, MeOH/H2O: 3:7). Compounds 6, 9 were obtained after
subsequent treatment with 2 ml of 10% NH4OH for 16 h,
lyophilization of the solution and purification by chromatog-
raphy on HP20SS (Prolabo).
6.1.4. 2,5-Dideoxy-2,5-imino-D-glucitol 4
A solution of benzylated pyrrolidine 13 (436 mg,
1.0 mmol), 10% Pd–C (800 mg) and aq. HCl (0.1 ml of a 1 M
solution) in MeOH (4 ml) was stirred under H2 at rt. After
48 h, the reaction mixture was filtered over a celite pad and
the solvent was evaporated. The crude material was purified
by silica gel FC (CHCl3/MeOH/NH4OH, 60:30:5) to give imi-
nosugar 4 (75 mg, 46%) as a yellowish oil. [␣]D = + 22.7 (c
0.77, H2O) [11]: [␣]D = +27.6 (c 1.3, MeOH),[23]:
[␣]D = +25.1 (c 1.5, H2O). – 1H NMR (D2O) 4.06 (dd, 1H, J
2.1 4.4), 3.82 (dd, 1H, J 2.1 3.6), 3.80–3.50 (m, 4H, H-2,
H-6), 3.32 (q, 1H, J 5.0), 3.02 (dd, 1H). – 13C NMR (D2O)
81.2, 79.4, 67.6, 64.3, 63.6, 62.2. – ESI-MS (NH3) m/z 186
[(M + Na)+], 164 [(M + H)+].
6.1.7.2. (2S,3R,4R,5S)-2-[(1-ethylphosphonyl)-1,1-difluorom-
ethyl]-3,4-dihydroxy-5-hydroxymethyl-pyrrolidine 5. Gen-
eral procedure from 18 gave 5 as a white foam after purifica-
tion (74%). [␣]D = +23.3 (c 0.9, H2O). – 1H NMR (D2O) 4.35
(m, 1H, H-3), 4.10 (m, 2H, CH2CH3), 3.95 (m, 1H, H-4),
3.65–3.90 (m, 3H, H-2, H-6a, b), 3.20 (q, 1H, J 5.4, H-5),
1.30 (t, 3H, CH3CH2). – 13C NMR (D2O) 122,5 (m, CF2),
80.4 (C-3), 79.1 (d, J 4.5, C-4), 67.8 (C-5), 66.3 (m, CH2CH3),
64.3 (qd, C-2), 64.1 (C-6), 18.7 (m, CH3CH2). – 19F NMR
6.1.5. (2S,3R,4R,5S)-N-benzyl-3,4-dibenzyloxy-5-[(benzy-
loxy)methyl]-2-vinyl pyrrolidine 14
Pyrrolidine 14 was prepared as described [15] with slight
modifications (replacement of Tf2O by MsCl for cycliza-
tion). [␣]D = -39.4 (c 2.2, CHCl3) (Ref. [15]: [␣]D = –28.6
(c 1; CHCl3). 1H NMR and 13C NMR data were identical to
those reported.
2
2
3
(D2O, 235 MHz) –107.6 (ddd, 1F, JF,F 300, JF,P 90, JF,H
11.5), –115.6 (ddd, 1F, 2JF,F 300, 2JF,P 90, 3JF,H 18.7). – 31
P
NMR (D2O, 101 MHz) 6.6 (bt). – ESI-MS (NH3) m/z 313.9
[(M + Na)+], 292 [(M + H)+].
6.1.7.3. (2R,3R,4R,5S)-2-[(1-ethylphosphonyl)-1,1-difluo-
romethyl]-3,4-dihydroxy-5-hydroxymethyl-pyrrolidine 6. Gen-
eral procedure from 16 gave 6 as a white foam after purifica-
tion (45%). [␣]D = +21.5 (c 0.8, MeOH). – 1H NMR (D2O)
4.24 (t, 1H, J 7.0, H-3), 4.03 (m, 2H, CH2CH3), 3.86 (t, 1H, J
7.0, H-4), 3.8-3.4 (m, 3H, H-2, H-6a, b), 3.13 (m, 1H, H-5),
1.25 (t, 3H, CH3CH2). – 13C NMR (D2O) 122.6 (m, CF2),
78.8 (C-3), 78.2 (C-4), 66.3 (m, CH2CH3), 65.4 (qd, C-2),
64.6 (C-5), 62.9 (C-6), 18.7 (m, CH3CH2). – 19F NMR (D2O,
6.1.6. (2S,3R,4R,5S)-N-benzyl-2-hydroxymethyl-3,4-diben-
zyloxy-5-[(benzyloxy)methyl]-2-vinyl-pyrrolidine 15
Ozonolysis of the sulfate salt of pyrrolidine 14 (1.097 g,
2.11 mmol) in CH2Cl2 (70 ml) was performed as described
for the preparation of 13. Subsequent reduction of the result-
ing aldehyde with NaBH4 (280 mg) gave, after FC
(EtOAc/petroleum ether: 2:8; Rf = 0.5), pure 15 (1.006 g,
91%) as a yellowish oil. [␣]D = –18.0 (c 5.28, CHCl3). – 1H
NMR (CDCl3) 7.40–7.20 (m, 20H, Ar-H), 4.80 (d, 1H, JAB
12.0, CH2Ph), 4.68 (d, 1H, JAB 12.0, CH2Ph), 4.51 (dd, 1H, J
3.7 7.0, H-3), 4.60 (m, 4H, 4 × CH2Ph), 4.23 (t, 1H, J 7.0,
H-4), 4.00 (s, 2H, NCH2Ph), 3.65 (m, 3H, 3 × CH2O–), 3.60
(dd, 1H, J 2.9 9.9, CH2O), 3.40 (ddd, 1H, J 3.7 3.7 8.3, H-2),
3.35 (ddd, 1H, J 2.9 2.9 9.9, H-5). – 13C NMR (CDCl3) 139.3,
138.4, 138.2 (Ar-C), 128.3–126.9 (Ar-C), 84.4 (C-4), 83.2
(C-3), 73.4, 72.9, 72.6 (3 × OCH2Ph), 66.2 (CH2OBn), 63.3
(C-2), 60.3 (CH2OH), 58.3 (C-5), 52.6 (NCH2Bn). – ESI-MS
(NH3) m/z 524 [(M + H)+].
2
2
3
235 MHz) –115.1 (ddd, 1F, JF,F 298, JF,P 90, JF,H 12.4),
2
2
3
–120.2 (ddd, 1F, JF,F 298, JF,P 90, JF,H 3.3). – 31P NMR
(D2O, 101 MHz) 6.5 (bt). – ESI-MS (NH3) m/z 314 [(M +
Na)+], 292 [(M + H)+].
6.1.7.4. (2S,3R,4R,5R)-2-[(1-ethylphosphonyl)-1,1-difluo-
romethyl]-3,4-dihydroxy-5-hydroxymethyl-pyrrolidine 8. Gen-
eral procedure from 19 gave 8 as a white foam after purifica-
tion (62%). [␣]D = +3.8 (c 1.36, H2O). – 1H NMR (D2O) 4.50
(m, 1H, H-3), 4.31 (m, 1H, H-4), 4.10 (m, 1H, H-2), 3.97 (m,
2H, CH2CH3), 3.90–3.70 (m, 3H, H-5, H-6a, b), 1.25 (t, 3H,
CH3CH2). – 13C NMR (D2O) 121.4 (m, CF2), 77.8 (C-3),
77.2 (C-4), 66.6 (m, CH2CH3), 65.8 (C-5), 64.8 (qd, C-2),
6.1.7. 2,5-Dideoxy-2,5-imino-L-iditol 7
Debenzylation of 15 (0.25 g, 0.48 mmol) in the conditions
stated above for the synthesis of 4 gave pure 7 (25 mg, 35%)

I. Gautier-Lefebvre et al. / European Journal of Medicinal Chemistry 40 (2005) 1255–1261
1261
60.5 (C-6), 18.8 (m, CH3CH2). – 19F NMR (D2O, 235 MHz)
–103.8 (ddd, 1F, 2JF,F 302, 2JF,P 84, 3JF,H 8); –116.6 (ddd, 1F,
2JF,F 302, 2JF,P 86, 3JF,H 21). – 31P NMR (D2O, 101 MHz) 4.8
(bt). – FAB-MS (glycerol) m/z 314 [(M + Na)+], 292
[(M + H)+].
suspension were filtered through a glass-fiber filter (Whatt-
man GF-C). The filter was washed 3 times with 1 ml 60%
aqueous EtOH and dried. The 14C-chitin formed was quanti-
tated by liquid scintillation counting in 4 ml of ultima flow
AP (Packard) scintillation fluid.
6.1.7.5. (2R,3R,4R,5R)-2-[(1-ethylphosphonyl)-1,1-difluo-
romethyl]-3,4-dihydroxy-5-hydroxymethyl-pyrrolidine 9. Gen-
eral procedure from 17 gave 9 as a white foam after purifica-
tion (61%). [␣]D = –26.7 (c 0.3, MeOH). – 1H NMR (D2O)
4.27 (dd, J 3.7 4.8, H-3), 4.20–4,00 (m, 3H, H-4, CH2CH3),
3.79 (dd, 1H, J 5.5 11.7, H-6a), 3.64 (dd, 1H, J 6.6 11.7,
H6b), 3.53–3.34 (m, 2H, H-2, H-5), 1.28 (t, 3H, CH3CH2). –
13C NMR (D2O) 120.6 (m, CF2), 79.1 (C-3, C-4), 68.4 (q,
C-2), 66.1 (m, CH2CH3), 63.5 (C-5), 62.1 (C-6), 18.7 (m,
CH3CH2). – 19F NMR (D2O, 235 MHz) –115.5 (ddd, 1F, 2JF,F
297, 2JF,P 92, 3JF,H 12.5), –118.7 (ddd, 1F, 2JF,F 297, 2JF,P 92,
3JF,H 17.2). – 31P NMR (D2O, 101 MHz) 6.6 (bt). – ESI-MS
(NH3) m/z 314 [(M + Na)+], 292 [(M + H)+].
Acknowledgments
The authors would like to thank ProfessorA. Bellarby and
Dr. A. Gainvors for very fruitful discussions, Dr. I. Des-
champs for her help in the pKa determination and Dr. N. Ryder
for the antifungal evaluation. This work was supported by
Europol’Agro and the “Conseil Général de la Marne”.
References
[1] (a) A.E. Stütz, Angew. Chem. Int. Ed. Engl. 35 (1996) 1926. (b) R.A.
Dwek, Chem. Rev. 96 (1996) 683-720. (c), J. Alper, Science 291
(2001) 2338–2343.
[2] J.-B. Behr, Curr. Med. Chem. 2 (2003) 173–189 (Anti Infective
Agents).
6.1.8. (2S,3R,4R,5S)-2-[(1,1-difluoromethylphosphonic
acid]-3,4-dihydroxy-5-hydroxymethyl-pyrrolidine 20
A solution of protected phosphonate 18 (0.64 g,
1.15 mmol) and Me3SiBr (0.7 ml, 5.3 mmol) in CH3CN was
stirred at rt for 7 days. The solvents were evaporated and the
residual oil was dissolved in EtOH (15 ml) and stirred for
3 days under H2 in presence of 10% Pd–C (0.15 g). Filtration
over a celite pad and evaporation of the solvents gave a brown
oil from which 5 crystallized in the presence of EtOH (0.125 g,
53%). Analytically pure 5 was obtained by preparative
[3] (a) D.H. Joziasse, R. Oriol, Biochim. Biophys. Acta 1455 (1999)
403–418. (b) C. Saotome, C.-H. Wong, O. Kanie, Chem. Biol. 8
(2001) 1061–1070.
[4] (a) I. Brockhausen, Biochim. Biophys. Acta 1473 (1999) 67–95 (b)
T.A. Springer, L.A. Lasky, Nature (London) 349 (1991) 196–197. (c)
L, Osborn, Cell 62 (1990) 3–6.
[5] T.D. Butters,A. Dwek, F.M. Platt, Chem. Rev. 100 (2000) 4683–4696.
[6] M.L. Sinnott, Chem. Rev. 90 (1990) 1171–1202.
[7] P. Compain, O.R. Martin, Curr. Top. Med. Chem. 3 (2003) 541–560.
[8] L. Qiao, B.W. Murray, M. Shimazaki, J. Schultz, C.-H. Wong, J. Am.
Chem. Soc. 118 (1996) 7653–7662.
1
RP-HPLC (eluent, H2O). [␣]D = +47 (c 0.36, H2O). – H
NMR (D2O) 4.50 (dd, 1H, J 1.3 3.0, H-3), 4.20 (m, 1H, H-2),
4.10 (m, 1H, H-4), 3.95 (dd, 1H, J 5.5 12.2, H6a), 3.85 (dd,
1H, J 8.5 12.2, H6b), 3.60 (m, 1H, H-5). – 13C NMR (D2O)
120.2 (m, CF2), 78.2 (C-4), 77.0 (d, J 4.5, C-3), 70.4 (C-5),
65.2 (m, C-2), 61.8 (C-6). – 19F NMR (D2O, 235 MHz) –105.2
(ddd, 1F, 2JF,F 304, 2JF,P 84, 3JF,H 8.5), –115.8 (ddd, 1F, 2JF,F
304, 2JF,P 84, 3JF,H 20). – 31P NMR (D2O, 101 MHz) 3.8 (t).
– FAB-MS (glycerol) m/z 286 [(M + Na)+], 264 [(M + H)+].
[9] C. Saotome, Y. Kanie, O. Kanie, C.-H. Wong, Bioorg. Med. Chem. 8
(2000) 2249–2261.
[10] M. Djebaili, J.-B. Behr, J. Enzyme Inhib. Med. Chem. 20 (2005)
123–127.
[11] K.K.-C. Liu, T. Kajimoto, L. Chen, Y. Ichikawa, C.-H. Wong, J. Org.
Chem. 56 (1991) 6280–6289.
[12] J.-B. Behr, A. Erard, G. Guillerm, Eur. J. Org. Chem. (2002) 1256–
1262.
[13] Y. Masaki, H. Oda, K.K. Kazuta, A. Usui, A. Itoh, F. Xu, Tetrahedron
Lett. 33 (1992) 5089–5092.
[14] T.K.M. Shing, Tetrahedron 44 (1988) 7261–7264.
[15] L. Cipolla, L. Lay, F. Nicotra, C. Pangrazio, L. Panza, Tetrahedron 51
(1995) 4679–4690.
[16] J.-B. Behr, C. Mvondo Evina, N. Phung, G. Guillerm, J. Chem. Soc.
Perkin. Trans. I (1997) 1597–1599.
6.2. Chitin synthase assay
Digitonin, trypsin, soybean trypsin inhibitor, UDP-GlcNAc
were obtained from Sigma. UDP trisodium salt hydrate was
from Aldrich and N-acetyl-D-glucosamine from Acros.
[UDP14C]-GlcNAc (289 mCi/mmol) came from Perkin–
Elmer Life Sciences.
Assays were carried out at 30 °C in a volume of 60 µl
which contained, in addition to the activated permeabilized
cells preparation (40 µl), the following final concentrations
of components: 30 mM Tris–HCl (pH 6.5), 35 mM GlcNAc,
3 mM Mg(Ac)2, digitonin (3.5 mg/ml), cell suspension
(52 mg/ml) and 1 mM UDP-[14C]-GlcNAc (20,000 cpm).
Reaction was initiated by addition of the cell suspension.
After 30 min incubation the reaction was stopped by adding
1 ml of 10% trichloroacetic acid and 950 µl of the resulting
[17] R. Fahti, W. Delaney, Q. Huang, Nucleosides Nucleotides 14 (1995)
1725–1735.
[18] a) P. Bronislaw, C. Bartsch, J. Org. Chem. 49 (1984) 4076–4078. b) H.
Sajiki, K, Hirota, Tetrahedron 54 (1998) 13981–13996.
[19] L.B. Crotti, T. Drgon, E. Cabib, Anal. Biochem. 298 (2001) 8–15.
[20] National Committe for Clinical Laboratory Standards, Reference
method for broth dilution antifungal susceptibility testing of yeasts;
approved standard. NCCLS Document M27-A. NCCLS, Wayne, PA,
USA, 1997.
[21] N.S. Ryder, S. Wagner, I. Leitner, Antimicrob. Agents Chemother. 42
(1998) 1057–1059.
[22] W.-J. Choi, E. Cabib, Anal. Biochem. 219 (1994) 368–372.
[23] A. Dondoni, P.P. Giovannini, O. Perrone, J. Org. Chem. 67 (2002)
7203–7214.