Stereoselective synthesis of (2S,3S,4R,5S)-5-methylpyrrolidine-3,4-diol derivatives that are highly selective α-L-fucosidase inhibitors

Article abstract of DOI:10.1039/b508855k

N-Phenylaminomethyl benzimidazolyl moieties attached at C-2 of (2S,3S,4R,5S)-5-methylpyrrolidine-3,4-diol increase the potency and selectivity of the inhibitory activity of these systems towards α-L-fucosidases. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b508855k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Stereoselective synthesis of (2S,3S,4R,5S)-5-methylpyrrolidine-3,4-diol
derivatives that are highly selective a-L-fucosidase inhibitors{
Antonio J. Moreno-Vargas,a Ana T. Carmona,a Federico Mora,a Pierre Vogelb and Inmaculada Robina*a
Received (in Cambridge, MA, USA) 24th June 2005, Accepted 21st July 2005
First published as an Advance Article on the web 8th September 2005
DOI: 10.1039/b508855k
N-Phenylaminomethyl benzimidazolyl moieties attached at
C-2 of (2S,3S,4R,5S)-5-methylpyrrolidine-3,4-diol increase the
potency and selectivity of the inhibitory activity of these systems
towards a-L-fucosidases.
Glycosidases are enzymes playing a crucial role in the biosynthesis
of glycoproteins.1 They modulate the cellular processes including
cell/cell, cell/invader recognition and inflammation.2 Inhibitors of
glycosidases are potential drugs against diseases that imply cellular
dysfunctions or alter cell/invader communication.3,4
Chart 1
1,5-Dideoxy-iminoalditols or 1,4-dideoxy-1,4-iminoalditols (aza-
sugars) resembling natural sugar substrates and/or mimicking the
nanomolar range. Structural modifications of 7 generated less
potent inhibitors.11,12
corresponding oxycarbenium ion intermediate are classical glyco-
sidase inhibitors.5 Unfortunately these compounds quite often lack
Interestingly, however, Wong and co-workers have found that
enzyme specificity.
(1R)-aminomethyl-1-deoxy-L-fucononojirimycin (8) is a useful
We have found that simple meso-pyrrolidine-3,4-diol (1) is a
weak and non-selective glycosidase inhibitor.6 Enzyme selectivity
template to construct libraries of amides that can be very potent
a-L-fucosidase inhibitors. For instance, amide 9 is a picomolar
slow, tight binding inhibitor of a-L-fucosidases.13
can be improved if the iminosugar, which mimics the pyranosyl
cation, includes some information on the aglycon undergoing the
Derivatives of 1,4-dideoxy-1,4-iminoalditol can also be good
hydrolytic process.7 We have reported that activity and selectivity
can be improved on joining aromatic and heteroaromatic moieties
to 1. Thus, derivatives 2 with (2R)-aminomethyl side chains can be
highly selective and competitive inhibitors of a-mannosidases.8
Derivatives with (2R)- or (2S)-aminoethyl side chains 3 or 4 are
less selective. Enantiomers ent-3 and ent-4 are moderate and
specific inhibitors of b-glucosidases.6 We have also found that
[(2S,3S,4R)-3,4-dihydroxypyrrolidin-2-yl]furan derivatives 5 are
good b-galactosidase inhibitors whereas their C-2 epimers 6 are
good and selective a-L-fucosidase inhibitors.9
a-L-fucosidase inhibitors,11,14–17 whereas 1,6-dideoxy-1,6-iminoal-
ditols have shown up to now only moderate inhibitory activities.18
Recent work in the search for anti-cancer agents has shown that
a-mannosidase inhibitors such as 2 have low cell membrane
permeability and must be esterified to generate compounds with
anti-cancer activity.19 This suggests that less polar compounds
than fuconojirimycin analogues 9 might be required to construct
a-L-fucosidase inhibitors able to penetrate cells. We thus have
decided to explore the use of (2S,3S,4R,5S)-5-methylpyrrolidine-
3,4-diol derivatives 10 as templates for the creation of potential
drugs. It is believed that analogs of 10 with fewer hydroxyl groups
represent a novel scaffold for a-L-fucosidase inhibitors with
potentially improved properties. Imitating the strategy we used
to develop selective a-mannosidase inhibitors,6,8 we are now
searching for derivatives of type 10 with high a-L-fucosidase
inhibitory activity and high selectivity toward this type of
enzymes. Pyrrolidine-diol derivative 10 can be made applying a
chemoenzymatic cross-aldol reaction of (2S)-azido propanal
and dihydroxyacetone monophosphate.16 Stereoisomers of 10
were obtained via an asymmetric Diels–Alder reaction of
(E,E)-sorbaldehyde dimethyl acetal with an a-chloronitroso-D-
mannose derivative17 through a 13-step synthetic route.
a-L-Fucosidases participate in the last stages of glycoprotein
biosynthesis and their inhibitors have been found to inhibit the
cytophatic effect of HIV and reduce infection.4 a-L-Fucosidases
also facilitate sperm transport and sperm–egg interactions.
Inhibitors of these enzymes could have contraceptive properties.10
Consequently, a great deal of effort has been made towards the
synthesis of new fucosidase inhibitors.
Up to now, the most potent fucosidase inhibitors are derivatives
of 1,5-dideoxy-1,5-iminoalditols, such as L-fuconojirimycin (7)
which inhibits a-L-fucosidases with Ki values in the low
aDepartment of Organic Chemistry, Faculty of Chemistry, University of
Seville, P. O. Box 553, E-41071 Seville, Spain. E-mail: robina@us.es;
Fax: +34954624960; Tel: +34954557151
bLaboratoire de glycochimie et synthe`se asyme´trique, Swiss Federal
Institute of Technology (EPFL), BCH-EPFL, CH-1015 Lausanne,
Switzerland. E-mail: pierre.vogel@epfl.ch
{ Electronic supplementary information (ESI) available: Experimental and
characterization data, including NMR spectra, for compounds 12–18, 23,
24, 26, 27. See http://dx.doi.org/10.1039/b508855k
Chart 2
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 4949–4951 | 4949

View Article Online
We disclose here an alternative shorter and more efficient
route to systems 10 and their conversion into new selective a-L-
fucosidase inhibitors.
of the mesyloxy group at C-5 giving a 3 : 1 mixture of (2S)-15 and
(2R)-15 in quantitative yield. The ratio of pyrrolidines 15 indicates
the stereoselectivity of the displacement reaction at C-2 in the open
chain compounds 13 that gives the azido derivatives 14. Amine
protection as benzylcarbamate followed by selective hydrolysis of
the least sterically hindered acetonides (H2O/MeCN, Zn(NO3)2)22
provided a 3 : 1 mixture of diols 16 and 17 that were readily
separated by flash chromatography on silica gel (58% overall on
three steps). Standard oxidative cleavage of the diols 16 and 17 with
NaIO4 furnished aldehydes 18 (92%) and 19 (100%), respectively.
Attempts to generate N-benzylpyrrolidine analogues of 15 by
direct reaction of bis-mesylates 13 with benzylamine were not
successful, even on using better leaving groups such as iodide or
triflates. Similarly, oxidation of diols 12 into the corresponding
1,4-diketone and subsequent reductive amination using ammo-
nium formate and NaBH3CN23 failed to produce any pyrrolidine
derivative, but led to the formation of mixtures of diastereoiso-
meric tetrahydrofuran derivatives.
According to Wightman and co-workers,20 the reaction
of methylmagnesium chloride with commercially available
D-mannose diacetonide (11) should occur already at 278 uC and
give a major diol. In our hands, no reaction occurs below 20 uC.
At room temperature, a 1 : 1 mixture of diastereoisomeric diols
(2R)-12 and (2S)-12 was formed and isolated in 95% yield.
Standard esterification of the diols with an excess of methane-
sulfonyl chloride and DMAP provided a 1 : 1 mixture of
dimesylates (2R)-13 and (2S)-13 in 74% yield (Scheme 1). Using
tetrabutylammonium azide (generated in situ by reaction of
Me3SiN3 and Bu4NF in DMF) as nucleophile, the chemoselective
SN2 displacement of the mesyloxy group at C-2 of 13 is by some
means faster with (2R)-13 than with (2S)-13. Indeed, after 75%
conversion (DMF, 90 uC) a 3 : 1 mixture of azides (2S)-14 and
(2R)-14, respectively, was formed and isolated in 55% yield.
The selectivities observed are attributed to steric factors.
Mesylate (2S)-13 offers a population of rotamers less suitable for
the SN2 displacement than (2R)-13.21
Structures of the 2,5-dideoxy-2,5-iminoalditol derivatives 16 and
17 were established by their spectral data. A strong NOE was
observed between the signals of 16 assigned to pair of protons H2
(d 5 3.85)/H3 (d 5 4.59), thus confirming the (2S)-configuration.
In the case of 17, a NOE was observed between the signals
assigned to H2 (d 5 3.96) and H5 (d 5 4.15) and and also between
signals assigned to H3 (d 5 4.37) and Me (d 5 1.28), thus
confirming the (2R,5S)-configuration. No NOE between signals
assigned to protons H4/H5 and a coupling constant J4,5 5 0 in
both compounds is observed, thus confirming the trans relative
configuration of these protons.
Catalytic hydrogenation of azides 14 gave a 3 : 1 mixture of
primary amines that were not isolated but treated directly with
DBU in MeOH. This promoted the intramolecular displacement
In order to obtain 1,2-diamines to be used for the rapid
discovery of new glycosidase inhibitors through a combinatorial
approach,8 we have synthesized diamines 2014 and 21 by reductive
amination of aldehydes 18 and 19 with BnNH2 followed by
deprotection (Scheme 1). On the other hand, reductive amination
of aldehyde 18 with aniline provided 22 (75%). After deprotection
of 22 under standard conditions, diamine 24 was obtained in
quantitative yield (Scheme 2).
Oxidation of aldehyde 18 gave the corresponding carboxylic
acid 23 (82% yield). It reacted with o-phenylenediamine in the
presence of PyBOP and DIPEA to give amide 25 in 65% yield. On
heating in AcOH at 50 uC,24 benzimidazole 26 was formed
and isolated in high yield. Deprotection of 26 under standard
conditions gave benzimidazole 27 (Scheme 2). The same reaction
sequence applied to aldehyde 19 provided benzimidazole 28. The
structures of 27 and 28 were based on their spectral data and
confirmed by NOE experiments.
Diamines 20, 21, 24 and benzimidazoles 27 and 28 have been
analyzed for their inhibitory activities towards fifteen commercially
available glycosidases.25 Apart from a weak inhibition (34% at
1 mM concentration) towards b-galactosidases from bovine liver,
diamine 20 was a selective and competitive inhibitor of a-L-
fucosidases from bovine epididymis (Ki 5 1.8 mM)26 and did not
inhibit a-galactosidases from coffee beans and from E. coli,
b-galactosidases from coffee beans and from E. coli, Aspergillus
niger and Aspergillus orizae, a-glucosidases from yeast and from
rice, amyloglucosidases from Aspergillus niger and from Aspergillus
mold, b-glucosidases from almonds and from Saccharomyces
cerevisae, a-mannosidases from Jack beans and from almonds,
b-mannosidases from Helix pomatia, b-xylosidases from Aspergillus
Scheme 1 {Reagents and conditions: a) MeMgCl, THF, r.t., 95%; b) MsCl,
pyridine, DMAP, 74%; c) TMSN3 (4 eq.)/TBAF (4 eq.), DMF, 90 uC,
4.5 h, 55%; d) (i) H2, Pd/C, MeOH; (ii) MeOH, DBU, 99% (i + ii); e) (i)
CbzCl, NaHCO3, EtOH/H2O (1/1); (ii) Zn(NO3)2?6H2O, MeCN, 50 uC;
(iii) Flash chromatography; 58% (i + ii + iii); f) NaIO4, THF–H2O; g) (i)
BnNH2, NaBH(OAc)3, ClCH2CH2Cl; (ii) H2, Pd/C, MeOH;(iii)HCl(1M).
4950 | Chem. Commun., 2005, 4949–4951
This journal is ß The Royal Society of Chemistry 2005

View Article Online
We thank MEC, Spain (CTQ2004-00649/BQU), EC (FP6,
TRIoH, LSHG-CT-2003-503480), COST D25/0001/01, the Office
Fe´de´ral de l’Education et de la Science (TRIoH) and the Swiss
NSF.
Notes and references
{ For the sake of clarity, the numbering indicated in the scheme was used.
1 D. H. DubeandC. R. Bertozzi,Curr. Opin. Chem. Biol., 2003, 7, 616–625.
2 A. Varki, Glycobiology, 1993, 3, 97–130.
3 R. A. Dwek, Chem. Rev., 1996, 96, 683–720; A. D. Elbein and
R. J. Molyneux, in Imino Sugars as Glycosidase Inhibitors:
Nojirimycin and Beyond (Ed.: A. E. Stu¨tz), Wiley-VCH, Weinheim,
1999, pp. 216–252.
4 I. Robina, A. J. Moreno-Vargas, A. T. Carmona and P. Vogel, Curr.
Drug Metab., 2004, 5, 329–361.
5 T. A. Houston and J. T. Blanchfield, MiniRev. Med. Chem., 2003, 3,
669–678.
6 F. Popowycz, S. Gerber-Lemaire, R. Demange, E. Rodr´ıguez-Garc´ıa,
A. T. Carmona Asenjo, I. Robina and P. Vogel, Bioorg. Med. Chem.
Lett., 2001, 11, 2489–2493; A. T. Carmona, F. Popowycz, S. Gerber-
Lemaire, E. Rodr´ıguez-Garc´ıa, C. Schu¨tz, P. Vogel and I. Robina,
Bioorg. Med. Chem., 2003, 11, 4897–4911.
7 B. A. Johns, Y. T. Pan, A. D. Elbein and C. R. Johnson, J. Am. Chem.
Soc., 1997, 119, 4856–4865.
8 S. Gerber-Lemaire, F. Popowycz, E. Rodrı´guez-Garcı´a,
A. T. Carmona Asenjo, I. Robina and P. Vogel, ChemBioChem,
2002, 3, 466–470.
9 I. Robina, A. J. Moreno-Vargas, J. G. Ferna´ndez-Bolan˜os, J. Fuentes,
R. Demange and P. Vogel, Bioorg. Med. Chem. Lett., 2001, 11,
2555–2559; A. J. Moreno-Vargas, R. Demange, J. Fuentes, I. Robina
and P. Vogel, Bioorg. Med. Chem. Lett., 2002, 12, 2335–2339;
A. J. Moreno-Vargas, I. Robina, R. Demange and P. Vogel, Helv.
Chim. Acta, 2003, 86, 1894–1913.
Scheme 2 Reagents and conditions: a) PhNH2, NaBH(OAc)3, 1,2-
dichloroethane; 75%; b) (i) HCl (1 M); (ii) H2, Pd/C, MeOH; 100% (i +
ii); c) NaClO2, KH2PO4, ButOH–H2O, 2-methyl-2-butene, 82%; d)
o-phenylenediamine, PyBOP, DIPEA, DMF, 65%; e) AcOH, 50 uC,
100%; f) (i) THF : HCl (1 M) 1 : 1, (ii) H2, Pd/C, MeOH; 87% (i + ii).
niger, N-acetylgalactosaminidase from chicken liver and b-N-acetyl
glucosaminidases from Jack beans and from bovine kidney.
Diastereoisomeric diamine 21 is a much weaker inhibitor of a-L-
fucosidases from bovine epididymis (52% at 1 mM concentration)
and inhibits a-glucosidases from yeast (37% at 1 mM) and
a-mannosidases from Jack beans (36% at 1 mM). The aniline
derivative 24 is a selective inhibitor of a-L-fucosidase from bovine
kidney and from bovine epididymis (Ki 5 0.24 mM, competitive).
Thus, as for a-mannosidase inhibitors of type 2, the aromatic
moiety enhanced the inhibitory activity of diamine 20 by a factor
of about 10. It contributes also to the high selectivity of the
inhibitor toward a single type of glycosidases. This phenomenon
appears to be further enhanced with benzimidazole 27 which is a
competitive inhibitor of a-L-fucosidase from bovine kidney with
Ki 5 80 nM. Importantly, this compound did not inhibit any of
the other enzymes assayed. As expected, benzimidazole 28 is a
much weaker inhibitor of a-L-fucosidase from bovine epididymis
(94% inhibition at 1 mM, Ki 5 240 mM) than 27 (100% inhibition
at 1 mM). Furthermore, 28 is not a selective inhibitor as it inhibits
also b-galactosidase from bovine liver (48%), a-glucosidase from
yeast (94%, Ki 5 46 mM, uncompetitive) and a-mannosidases
from Jack beans (80%) and from almonds (60% inhibition at 1 mM
concentration, optimal pH).
10 S. Khunsook, B. Bean, S. R. McGowan and J. A. Alhadeff, Biol.
Reprod., 2003, 68, 709–716.
11 C. W. Ekhart, M. H. Fletcher, P. Hadwiger, E. Mlaker, A. E. Stu¨tz,
A. A. Tauss and T. M. Wrodnigg, in Iminosugars as Glycosidases
inhibitors, Nojirimycin and Beyond; Wiley-VCH: Weinheim, 1999; p. 254.
12 M. Ogawa, S.-J. Ma. Yamamoto and K. Sakata, Bioorg. Med. Chem.
Lett., 2001, 11, 467–470; A. Bordier, P. Compain, O. R. Martin,
K. Ikeda and N. Asano, Tetrahedron: Asymmetry, 2003, 14, 47–51.
13 C.-Y. Wu, C.-F. Chang, S.-Y. J. Chen, C.-H. Wong and C.-H. Lin,
Angew. Chem., Int. Ed., 2003, 42, 4661–4664; C.-F. Chang, C. W. Ho,
C.-Y. Wu, T.-A. Chao, C.-H. Wong and C.-H. Lin, Chem. Biol., 2004,
11, 1301–1306.
14 C.-H. Wong, L. Provencher, J. A. Porco, Jr, S.-H. Jung, Y.-F. Wang,
L. Chen, R. Wang and D. H. Steensma, J. Org. Chem., 1995, 60,
1492–1501.
15 C. Hevrier, D. LeNouen, M. Neuburger, A. Defoin and C. Tarnus,
Tetrahedron Lett., 2004, 45, 5363–5366.
16 Y.-F. Wang, D. P. Dumas and C.-H. Wong, Tetrahedron Lett., 1993,
34, 403–406.
17 T. Sifferlen, A. Defoin, J. Streith, D. Le Nou¨en, C. Tarnus, I. Dosbaˆo
and M.-J. Foglietti, Tetrahedron, 2000, 56, 971–978.
18 H. Li, Y. Ble´riot, J.-M. Mallet, E. Rodr´ıguez-Garc´ıa, P. Vogel, Y. Zhang
and P. Sinay¨, Tetrahedron: Asymmetry, 2005, 16, 313–319.
19 H. Fiaux, F. Popowycz, S. Faure, C. Schu¨tz, P. Vogel, S. Gerber-
Lemaire and L. Juillerat-Jeanneret, J. Med. Chem., 2005, 48, 4237–4246.
20 B. Mekki, G. Singh and R. H. Wightman, Tetrahedron Lett., 1991, 32,
5143–5146.
21 For similar examples, see: J. G. Buchanan, K. A. MacLean and
R. H. Wightman, J. Chem. Soc., Perkin Trans. 1, 1985, 1463–1470.
22 S. Vijayasaradhi, J. Singh and I. Singh Aidhem, Synlett, 2000, 110–112.
23 A. Momotake, J. Mito, K. Yamaguchi, H. Togo and M. Yokohama,
J. Org. Chem., 1998, 63, 7207–7212.
24 M. Jazouli, D. Guianvarch, M. Soufiaoui, K. Bougrin, P. Vierling and
R. Benhida, Tetrahedron Lett., 2003, 44, 5807–5810.
25 A. Brandi, S. Cicchi, F. M. Cordero, R. Frignoli, A. Goti, S. Picasso
and P. Vogel, J. Org. Chem., 1995, 60, 6806–6812.
26 Reported Ki 5 1.9 mM a-L-fucosidase (bovine kidney) for 20, see Ref. 14.
In conclusion, a new method has been developed for the prepara-
tion of (2S,3S,4R,5S)-5-methylpyrrolidine-3,4-diols bearing either
aminomethyl or heterocyclic moieties at C-2. New types of highly
selective and competitive inhibitors of a-L-fucosidases have been
discovered. Although they are less active than 1-deoxyfuconojir-
imycin analogs reported,13 they are less polar than the latter and
thus represent valuable leads as in vivo a-L-fucosidase inhibitors.
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 4949–4951 | 4951