Synthesis of new (difluoromethylphosphono)azadisaccharides designed as bisubstrate analogue inhibitors for GlcNAc:β-1,4 glycosyltransferases

Article abstract of DOI:10.1016/S0960-894X(00)00263-8

We report here the design, synthesis and antifungal evaluation of a new model of bisubstrate analogue inhibitor for glycosyltransferases. The synthetic strategy relies on the reductive amination between the aldehyde derived from an N-allylphosphono-pyrrolidine and an aminosugar. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00263-8

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1483±1486
Synthesis of New (Di¯uoromethylphosphono)Azadisaccharides
Designed as Bisubstrate Analogue Inhibitors for GlcNAc:ꢀ-1,4
Glycosyltransferases
Isabelle Gautier-Lefebvre,a Jean-Bernard Behr,a Georges Guillerma,* and Neil S. Ryderb
aLaboratoire de Chimie Bioorganique UMR 6519, UFR Sciences BP 1039, 51687 Reims Cedex 2, France
bNovartis Research Institute, Brunner Strasse 59, A-1235 Vienna, Austria
Received 26 November 1999; accepted 5 May 2000
AbstractÐWe report here the design, synthesis and antifungal evaluation of a new model of bisubstrate analogue inhibitor for
glycosyltransferases. The synthetic strategy relies on the reductive amination between the aldehyde derived from an N-allylphos-
phono-pyrrolidine and an aminosugar. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
diphosphate sugars as donors. To mimic the proposed
transition state possibly involved in this class of glycosyl
transferases (Fig. 1), we designed I as a bisubstrate ana-
logue prototype which includes the following structural
features:
Glycosyl transferase inhibitors have recently become the
focus of particular interest because of their value in basic
biochemical research1 and their therapeutic potential as
antifungal, antibacterial, antiviral, antimetastatic or anti-
adhesive agents.2
1. The ®ve-membered ring azasugar, stereochemically
apparented to l-xylose, would mimic the half chair
conformation of the glycosyl cation and a possible
hydrogen bonded contact of the N-acetyl group
with enzymes.3a
The knowledge of the active site of glycosyl transferases
is rather limited.3 Putative mechanisms for the enzymatic
process involved in glycosyl transfer reaction using sugar
nucleotides, UDP-galactose, UDP-N-acetylglucosamine
and other so-called Leloir donors have been put forward
and subsequently used in the design of potent transition
state and bisubstrate analogues of di€erent enzymes of
this class.4 To improve the selectivity and inhibitory
potency of such inhibitors, tricomponent bisubstrates
containing both the donor and the acceptor molecules
involved in the enzymatic glycosyl transfer reaction of the
target enzyme, have recently been developed.5
2. Incorporation of
a di¯uoromethylphosphonate
group, mimicking the ®rst portion of the pyrophos-
phate moiety of the donor, might provide favourable
contribution to recognition by glycosyltransferases.
3. Insertion of an amino group in the spacer linking
the azasugar and GlcNAc glycon could introduce
additional electrostatic interaction between the 4-
position of the acceptor and enzymes.6
Synthesis
As a part of a research programme aimed at designing
novel inhibitors of chitin synthetase (an enzyme which
catalyses polymerisation of N-acetyl-d-glucosamine start-
ing from UDP-GlcNAc), we have become interested in
®nding a simple, versatile and ecient synthetic route
for novel bisubstrate analogues for glycosyl transferases
which use GlcNAc as acceptor and di€erent nucleoside
The synthetic strategy to reach bisubstrate analogues of
type I relies on the reductive amination between the alde-
hyde 11 derived from N-allylphosphonopyrrolidines 4 and
the 2-acetamido-4-amino-2,4-dideoxy-d-glucopyranoside
derivative 9 (Scheme 3). We have already shown the pos-
sibility of reaching N-benzyl-di¯uoromethylphosphono-
pyrrolidines by condensation of (diethylphosphinoyl)-
di¯uoromethyllithium with N-benzylglycosylamines.7
We used the same methodology (Scheme 1) to prepare
the corresponding N-allyl derivatives 4, precursors for
*Corresponding author. Fax: +3-2605-3166; e-mail: georges.guillerm@
univ-reims.fr
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00263-8

1484
I. Gautier-Lefebvre et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1483±1486
the target aldehydes: N-allylfuranosylamines8 2a,b (pre-
pared from commercial 2,3,5-Tri-O-benzyl-d-arabinose
and from 2,3,5-Tri-O-benzyl-l-xylose9 by reaction with
allylamine) were reacted with (diethylphosphinoyl)
di¯uoromethyllithium.
In each case, both possible diastereoisomers 3 were
formed in good yields: the reaction favors the formation
of the threo adduct in a 50% d.e., con®rming the ste-
reochemical control by the neighbouring C-2 carbon, as
previously postulated.10 The con®guration at the newly
created stereocentre was ®rmly assigned after conversion
of 3 to pyrrolidines 4 with methanesulfonylchloride: a
substantial nuclear Overhauser e€ect was observed on
19F NMR signals by saturation of H-3 for minor a iso-
mers ꢁ-4a,b, suggesting the cis relationship between
H-3 and the di¯uoromethylphosphonate moiety in
these compounds. Much weaker NOE were measured
for major 1-b con®gurated epimers (not shown on
Scheme 1).
Figure 1.
The 2-acetamido-4-amino-2,4-dideoxy-d-glucopyranose
derivative 9 can be prepared starting from GalNAc as
described by Field.6 Our synthesis used N-acetylglucos-
amine 5 as starting material (a cheap commercial com-
pound) and required selective protection of 1-OH, 3-OH
and 6-OH (6, Scheme 2) which could be achieved in one
step with N-pivaloyl imidazole as the acylating agent.11
After inversion of the 4-hydroxyl function, nucleophilic
displacement of the corresponding galactotri¯ate with
NaN3 followed by catalytic hydrogenation of the azide,
led to amine 8. The gluco con®guration of 8 was asses-
sed by the 10 Hz coupling constant between H-3 and H-
4, con®rming the trans-diaxial position of these atoms.
Formation of the methyl glycoside with in situ-generated
HCl in MeOH, followed by classical deprotection of
pivaloyl functions with MeOH/MeONa, led to fully
deprotected amine 9.
Scheme 1. (i) HCF2P(O)(OEt)2, LDA, THF±hexane, 78 ꢀC; (ii)
Methanesulfonyl chloride, pyridine.
Coupling of 9 with the N-allylphosphonopyrrolidine 4a
relies on a one-pot sequence including: conversion of N-
allylpyrrolidine to the corresponding aldehyde, addition
of 9 and reduction of the resulting imine (Scheme 3).
Unexpected ozonolysis of 4a (as its sulfate salt) gave
unsatisfactory results. The aldehyde 11a was obtained
after osmylation followed by the oxidative cleavage of
glycol 10a in CH2Cl2 with silica gel-supported NaIO4 as
the oxidant.12 Unstable aldehyde 11a was condensed
without further puri®cation with the amine 9. Reduc-
tion of the in situ generated imine with NaBH3CN led
Scheme 2. (i) (a) Tf2O, pyridine; (b) NaN3, DMF; (c) H2, Pd-C,
MeOH; (ii) (a) CH3COCl, MeOH; (b) MeONa, MeOH.
Scheme 3. (i) OsO4, NMO, acetone±water; (ii) NaIO4/SiO2, CH2Cl2; (iii) 9, MgSO4, EtOH, then NaBH3CN; (iv) HCO2NH4, Pd/C, re¯uxing EtOH.

I. Gautier-Lefebvre et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1483±1486
1485
to the protected bisubstrate analogue 12a in a 50% overall
yield from 4a. Hydrogenolysis of the O-benzyl protecting
groups was fairly dicult, and their removal was best
achieved by Pd(0)-catalysed transfer hydrogenation
a€ording the hydrosoluble compound 13a.13
Survey 1988, 7, 573. Springer, T. A.; Lasky, L. A. Nature 1991,
349, 193. Osborn, L. Cell 1990, 62, 3. Olden, K.; Breton, P.;
Grzegorzewski, K.; Yasuda, Y.; Gause, B. L.; Oredipe, O. A.;
Newton, S. A.; White, S. L. Pharmacol. Ther. 1991, 50, 285;
Goss, P. E.; Baptiste, J.; Fernandes, B.; Baker, M.; Dennis, J.
W. Cancer Res. 1994, 54, 1450.
3. (a) Sinnott, M. L. J. Chem. Rev. 1990, 90, 1171 and ref
therein. (b) Timmers, C. M.; Dekker, M.; Buijsman, R. C.;
Van der Marel, G. A.; Ethell, B.; Anderson, G.; Burchell, B.;
Mulder, G. J.; Van Boom, J. H. Bioorg. Med. Chem. Lett.
1997, 7, 1501. (c) Palcic, M. M.; Heerse, L. D.; Srivastava, O.
P.; Hindsgaul, O. J. Biol. Chem. 1989, 264, 17174. (d) Noort,
D.; Van der Marel, G. A.; Van der Gen, A.; Mulder, G. J.;
Van Boom, J. H. Red. Trav. Chim. Pays-Bas 1991, 110, 53 and
ref therein.
4. (a) Wischnat, R.; Martin, R.; Wong, C. H. J. Org. Chem.
1998, 63, 8361. (b) Takaoka, Y.; Kajimoto, T.; Wong, C. H. J.
Org. Chem. 1993, 58, 4809. (c) Palcic, M. M.; Ripka, J.; Kaur,
K. J.; Shoreibah, M.; Hindsgaul, O.; Pierce, M. J. Biol. Chem.
1990, 265, 6759. (d) Khan, S. H.; Crawley, S. C.; Kane, O.;
Hindsgaul, O. J. Biol. Chem. 1993, 268, 2468.
5. Hashimoto, H.; Endo, T.; Kajihara, Y. J. Org. Chem. 1997,
62, 1914.
6. Field, R. A.; Neville, D. C. A.; Smith, R. W.; Ferguson, M.
Bioorg. Med. Chem. Lett. 1994, 4, 391.
7. Behr, J.-B.; Mvondo Evina, C.; Phung, N.; Guillerm, G. J.
Chem. Soc. Perkin Trans. 1 1997, 1599.
8. N-Allyl glycosylamines are easily hydrolized back to the
parent sugar (Spevak, W.; Dasgupta, F.; Hobbs, C. J.; Nagy,
J. O. J. Org. Chem. 1996, 61, 3417) so that 2a,b could not be
puri®ed and were reacted as crude material.
This synthetic route has been applied to ꢁ-4b and ꢀ-
4a,b isomers to reach the corresponding bisubstrate
analogues.
Biological Evaluation
Antifungal assays: as 13a can be considered as a valuable
model of bisubstrate analogue inhibitor for chitin synthe-
tase (an important target for antifungal agents)14 its
activity was tested on a panel of several pathogenic fungi15
(Cryptococcus neoformans, Trichophyton mentagrophytes,
Microsporum canis, Aspergillus fumigatus, Candida albi-
cans, Saccharomyces cerevisiae). Compound 13a showed
no activity up to the maximum tested concentration of
128 mg/mL, whereas the standard compound terbina®ne
showed the normal activity expected.17 To explain this
lack of cellular activity it may be advanced that 13a does
not penetrate the fungal cell envelope. In order to check
this hypothesis, we evaluated the enzymatic inhibitory
potency of 13a on the activity of chitin synthetase from
Saccharomyces cerevisiae.
9. Barker, R.; Fletcher, H. G. J. Org. Chem. 1961, 26, 4605.
10. Lay, L.; Nicotra, F.; Paganini, A.; Pangrazio, C.; Panza,
L. Tetrahedron Lett. 1993, 34, 4555.
11. Santoyo-Gonzalez, F.; Uriel, C.; Calvo-Asin, J. A. Synth-
esis 1998, 1787.
12. Zhong, Y.-L.; Shing, T.K.M. J. Org. Chem. 1997, 62,
2622.
13. 1H NMR (D2O, 500 MHz, d ppm, J Hz) 4.62 (d, 1H, J1,2
For the inhibition studies, enzymatic assays were per-
formed using Saccharomyces cerevisiae cells permeabilized
with toluene±ethanol according to the procedure descri-
bed by Masson et al.18 The activation of enzyme present
as zymogen was performed essentially as described by
Cabib.19 Chitin synthetase was assayed by measuring the
rate of formation of (14C)-chitin from [UDP14C]-GlcNAc
according to the general method reported previously.20
0
0
0
0
3.5, H-1); 4.24 (m, 4H, 2CH2CH3); 4.17 (dd, 1H, J3 ,4 7.0, J3 ,2
0
6.0, H-30); 4.08 (dd, 1H, J4 ,3 =J4 ,5 7.0, H-40); 3.79 (dd, 1H,
J2,3 10.4, H-2); 3.73 (dd, 1H, J6a,6b 12.1, J6a,5 2.4, H-6a); 3.66±
3.55 (m, 4H, H-6b, H-5, H-3, H-60a); 3.36 (dd, 1H, J6 a,6 b 11.4,
0
0
0
Using our enzymatic preparation, Km value for UDP-N-
acetylglucosamine was 0.78 mM. Nikkomycin Z, used as
reference in our inhibition studies, strongly inhibited
chitin synthetase preparation (Ki=0.6 mM), whereas 13a,
tested in the same conditions, and for concentrations up
to 500 mM did not inhibit chitin synthetase activity.
0
0
J6 b,5 7.5, H-60b); 3.28±3.19 (m, 4H, H-20, CH3O); 3.02 (m,
1H, H-50); 2.90 (m, 1H, N-CH2-); 2.78±2.65 (m, 3H, N-CH2);
2.55 (bt, 1H, J4,5=J4,3 10.0, H-4); 1.9 (s, 3H, C(O)CH3); 1.26
(t, 6H, J 7.1, 2CH3CH2). 13C NMR (D2O, 125 MHz) 174.4
0
0
1
(CˆO); 122.2 (dt, JC-F=262, JC-P=210, CF2); 98.0 (C-1);
75.1 (C-40); 75.0 (C-30); 70.5 (C-20); 70.3 (C-5); 66.4 (m,
CH2CH3); 65.3 (C-50); 65.2 (C-3); 61.0 and 60.6 (C-60, C-6);
59.6 (C-4); 56.3 (N-CH2); 55.0 (CH3O); 54.1 (C-2); 45.5 (N-
CH2); 21.7 (COCH3); 15.6 (m, CH2CH3). 19F NMR (D2O, 235
Acknowledgements
1
2
3
MHz) 111.9 ppm (ddd, 1F, JFa,Fb 309, JF,P 110, JF,H 7);
1
114.8 ppm (ddd, 1F, JFa,Fb 309, JF,P 109, JF,H 21). 31P
NMR (D2O 101 MHz) 10.4 ppm (t).
2
3
The authors would like to thank Europol'Agro and the
`Conseil General de la Marne' for a PhD grant to one of
us (I. G.-L.).
14. Cabib, E.; Shaw, J.A.; Mol, P.C.; Bowers, B.; Choi, W.-J.
in Biochemistry and Molecular Biology; Springer Verlag, Ed.;
1996, pp 243.
15. MIC's were determined in broth macrodilution assays
according to guidelines of the NCCLS M27-A protocol,16 with
slight modi®cations.17 Inocula for assays were prepared from
stocks frozen at 80 ꢀC by dilution in growth medium to give
a ®nal viable cell count of 2.5Â103 CFU/mL. Each assay was
performed with a duplicate series of drug dilutions. In brief,
the assays were done in RPMI 1640 medium bu€ered to pH
7.0 with MOPS bu€er, incubated at 35 ꢀC for 48 h. C. neofor-
mans and A. fumigatus were tested in the same assay, except
that incubations were for 72 h. The dermatophytes T. menta-
grophytes and M. canis were incubated at 30 ꢀC for 7 days.
References and Notes
1. (a) Stutz, A.E. Angew Chem., Int. Ed. Engl. 1996, 35, 1926.
(b) Dwek, R. A. Chem. Rev. 1996, 96, 683. (c) Smith, C. A.;
Davis, T.; Anderson, D.; Solam, L.; Beekman, M. P.; Jerzy,
R.; Tersu, R.; Dower, D. K.; Cosman, D.; Goodwin, R. G.
Science 1990, 248, 1019. (d) Warki, A. Glycobiology 1993, 3,
97. (e) Isono, K.; Suzuki, S. Heterocycles 1979, 13, 333.
2. Dennis, K. W.; Laferte, S.; Waghorne, C.; Breitman, M. L.;
Kerbel, R. S. Science 1987, 236, 582. Dennis, J. W. Cancer

1486
I. Gautier-Lefebvre et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1483±1486
The MIC was de®ned as the lowest concentration of the drug
causing 80% inhibition of fungal growth in comparison with
untreated controls.
16. National Committee 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.
17. Ryder, N. S.; Wagner, S. I.; Leitner, I. Antimicrob. Agents
Chemother. 1998, 42, 1057.
18. Masson, J. M.; Guillermet, V.; Marinos, E.; Le Goc, F.
Eur. J. Med. Chem. 1987, 22, 377.
19. Cabib, E. Methods Enzymol. 1972, 28, 572.
20. Ruiz-Herrera, J.; Bartinicki-Garcia, S. J. Gen. Microbiol
1976, 97, 241.