Novel synthesis of the glycosidase inhibitor deoxymannojirimycin and of a synthetic precursor D-lyxo-hexos-5-ulose

Article abstract of DOI:10.1021/ol016596s

Equation presented The synthesis of D-lyxo-hexos-5-ulose (5-ketomannose, 1,5-dicarbonyl sugar), a synthetic precursor to the glycoprocessing inhibitor deoxymannojirimycin, was carried out by an in situ epoxidation and hydrolysis of a trimethylsilyl-protected 6-deoxyhex-5-enopyranoside followed by facile removal of the protecting groups. A novel nine-step synthesis of deoxymannojirimycin has also been achieved from methyl α-D-mannopyranoside; this involved methanolysis of epoxides derived from an acetylated 1-azido-6-deoxyhex-5-enopyranoside followed by deprotection and catalytic hydrogenation.

Full text of DOI:10.1021/ol016596s

ORGANIC
LETTERS
2001
Vol. 3, No. 21
3353-3356
Novel Synthesis of the Glycosidase
Inhibitor Deoxymannojirimycin and of a
Synthetic Precursor
D-lyxo-Hexos-5-ulose
Julie L. O’Brien, Manuela Tosin, and Paul V. Murphy*
Chemistry Department, Conway Institute of Biomolecular and Biomedical Research,
UniVersity College Dublin, Belfield, Dublin 4, Ireland
paul.V.murphy@ucd.ie
Received August 16, 2001
ABSTRACT
The synthesis of D-lyxo-hexos-5-ulose (5-ketomannose, 1,5-dicarbonyl sugar), a synthetic precursor to the glycoprocessing inhibitor
deoxymannojirimycin, was carried out by an in situ epoxidation and hydrolysis of a trimethylsilyl-protected 6-deoxyhex-5-enopyranoside followed
by facile removal of the protecting groups. A novel nine-step synthesis of deoxymannojirimycin has also been achieved from methyl r-D-
mannopyranoside; this involved methanolysis of epoxides derived from an acetylated 1-azido-6-deoxyhex-5-enopyranoside followed by
deprotection and catalytic hydrogenation.
Deoxynojirimycin (1) and deoxymannojirimycin (2) are
natural products that have been isolated from a variety of
sources and are among the most promising lead compounds
for treatment of HIV infection, diabetes, and other metabolic
disorders.1 These and related compounds have attracted
considerable attention from synthetic2 and medicinal chem-
ists, biologists, and clinical researchers in recent years as a
result of their potent inhibition of glycoprotein- and glyco-
lipid-processing enzymes such as glycosidases and glyco-
syltransferases. N-Butyldeoxynojirimycin, for example, cur-
rently in clinical trials as a potential therapy for Gaucher
disease, is an inhibitor of glucosidase I3 and ceramide
glucosyltransferase.4 The mechanism of action of these aza-
sugars seems to arise from their resemblance with the
transition state structure of the glycosidic cleavage or
glycosyltransferase reaction. There is also evidence that the
anti-HIV-1 activity displayed by some of these molecules
may be due to inhibition by the aza-sugars at novel and as
(2) For some recent syntheses of 1 and 2, see: (a) Kiguchi, T.; Tajiri,
K.; Ninomiya, I.; Naito, T. Tetrahedron 2000, 56, 5819. (b) Pistia, G.;
Hollingsworth, R. I. Carbohydr. Res. 2000, 328, 467. (c) Lindstro¨m, U.
M.; Somfai, P. Tetrahedron Lett. 1998, 39, 7173. (d) Asano, K.; Hagoki,
T.; Iwama, S.; Katsumura, S. Chem. Commun. 1999, 41. (e) Haukaas, M.
H.; O’Doherty, G. A. Org. Lett. 2001, 3, 401. (f) Wu, X.-D.; Khim, S.-K.;
Zhang, X.; Cederstrom, E. M.; Mariano, P. S. J. Org. Chem. 1998, 63,
841. (g) Schaller, C.; Vogel, P.; Ja¨ger, V. Carbohydr. Res. 1998, 314, 25.
(h) Meyers, A. I.; Andres, C. J.; Resek, J. E.; Woodall, C. C.; McLaughlin
M. A.; Lee, P. H.; Price, D. A. Tetrahedron 1999, 55, 8931. (i) Xu, Y.-M.;
Zhou, W.-S. J. Chem. Soc., Perkin Trans. 1 1997, 741. (j) Mart´ın, R.;
Moyano, A.; Pericaˆs, Riera, A. Org. Lett. 2000, 2, 93. (k) Yokoyama, H.;
Otaya, K.; Kobayashi, H.; Miyazawa, M.; Yamaguchi, S.; Hirai, Y. Org.
Lett. 2000, 2, 2427.
(3) Schweden, J.; Borgmann, C.; Legler, G.; Bause, E. Arch. Biochim.
Biophys. 1986, 248, 335.
(1) (a) Hughes, A. B.; Rudge, A. J. Nat. Prod. Rep. 1994, 135. (b) Fleet,
G. W. J. Glycobiology 1992, 2, 199 and references therein.
(4) Platt, F. M.; Neises, G. R.; Karlsson, G. B.; Dwek, R. A.; Butters,
T. D. J. Biol. Chem. 1994, 269, 27108.
10.1021/ol016596s CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/27/2001

yet undetermined sites of action.5 Herein we describe a novel
route to deoxymannojirmycin and a new synthesis of one of
its precursors, D-lyxo-hexos-5-ulose.
straightforward than removal of benzyl groups from 4. The
preparation of 8 was achieved from commercially available
methyl R-D-mannopyranoside by exchange of iodine for the
hydroxyl group at C-6,12 acetylation, elimination of hydrogen
iodide using DBU in anhydrous toluene, and exchange of
the protecting groups from acetate to TMS (Scheme 2, 46%
over five steps).
Scheme 2a
We have been interested in establishing a synthesis of
multigram quantities of deoxymannojirimycin 2 as a starting
material for preparing analogues for biological evaluation.
1,5-Dicarbonylsugars,6-8 such as 5-ketoglucose 6 and 5-keto-
mannose 10,9 can be converted to 1 and 2, respectively, by
double reductive amination reactions as described by Baxter
and Reitz.10 Workers in our laboratory have been studying
the synthetic potential of epoxides derived from 6-deoxyhex-
5-enopyranosides 3 and found that they are easily hydrolyzed
and ultimately give the disguised 1,5-dicarbonyl derivative
4.11 Attempts to directly remove the benzyl protecting groups
from 4 to give 6, a precursor to deoxynojirimycin, were not
successful. However, 6 was ultimately prepared after con-
verting 4 to 5 followed by removal of the silyl protecting
group (Scheme 1).11
a Reagents and conditions: (i) PPh3, Im, I2, toluene, 110 °C; (ii)
Ac2O, Py; (iii) DBU, toluene (anhydr), 110 °C, 50% over 3 steps;
(iv) NaOMe, MeOH; (v) TMSCl, Py, 93% over two steps; (vi)
1,1,1-trifluoroacetone, Oxone, NaHCO3, Na2EDTA, CH3CN, H2O,
1 h, 71%; (vii) MeOH, 94%.
Scheme 1
Interestingly, the hemiketal derivative 9 was isolated in
71% yield13 after treatment of 8 with methyl(trifluoromethyl)-
dioxirane generated in situ.14 The TMS groups can be
removed simply from 8 by stirring in methanol, and the
desired 1,5-dicarbonyl product 10 can be isolated in good
yield (94%) without a need for chromatography.15
Baxter and Reitz have previously reported a yield of 29%
for conversion of 10 to 2, using their double reductive
amination strategy, suggesting alternative routes to the aza
sugar 2 could be explored that might be more efficient in
this case. It seemed reasonable that the imine 12 would be
generated in situ during catalytic hydrogenation of epoxide
13 (or of a synthetic equivalent) and that further reduction
would give 2 via the cyclic imine 11 (Scheme 3).16
We wanted to reduce the number of protecting group
manipulations used in this synthesis of ketosugars and thus
investigated epoxidation-hydrolysis of the TMS protected
hex-5-enopyranoside 8 with a view to establishing a con-
venient synthesis of 10. It seemed reasonable to assume that
the ultimate removal of the TMS groups would be more
Scheme 3
(5) Asano, N.; Nishida, M.; Kato, A.; Kizu, H.; Matsui, K.; Shimida,
Y.; Itoh, T.; Baba, M.; Watson, A. A.; Nash, R. J.; de Q. Lilley, P. M.;
Watkin, D. J.; Fleet, G. W. J. J. Med. Chem. 1998, 41, 2565.
(6) These compounds have also been postulated as intermediates in
inositol biosynthesis; see: (a) Wong, Y.-H. H.; Sherman, W. R. J. Biol.
Chem. 1981, 256, 7077. (b) Eisenberg, F., Jr.; Maeda, T.; In Inositols and
Phosphoinositides; Bleasdale, J. E., Eichberg, J., Hauser, G.; Eds.; Hu-
mana: New Jersey, 1985; p 3.
(7) For a proposed biosynthesis of aza-sugars via 5-ketoglucose, see:
(c) Hardick, D. J.; Hutchinson, D. W.; Trew, S. J.; Wellington, E. M. H.
Chem. Commun. 1991, 729.
(8) 1,5-Dicarbonyl sugars have recently been used for a synthesis of
glycosylated deoxynojirimycin derivatives; see: D’Andrea F.; Catelani, G.;
Mariana, M.; Vecchi, B. Tetrahedron Lett. 2001, 42, 1139.
(9) 5-Ketoglucose is also known as D-xylo-hexos-5-ulose and 5-keto-
mannose as D-lyxo-hexos-5-ulose.
(10) Baxter, E.; Reitz, A. B. J. Org. Chem. 1994, 59, 3175.
(11) Enright, P. M.; O’Boyle, K. M.; Murphy, P. V. Org. Lett. 2000, 2,
3929.
Thus we commenced with the synthesis of 15 (Scheme
4). The iodo-derivative 7 can be converted into 14 (83%)
3354
Org. Lett., Vol. 3, No. 21, 2001

obtained in 23% yield.19 All the analytical data for the salt
were identical with that of a sample purchased from Sigma
and with literature data reported by Fleet.20 In summary, we
Scheme 4a
Scheme 5a
a Reagents and conditions: (i) H2SO4 (cat.), Ac2O; (ii) TMSN3,
SnCl4, CH2Cl2; (iii) DBU, toluene, 110 °C, 46% for 5 steps from
methyl R-D-mannopyranoside; (iv) 1,1,1-trifluoroacetone, Oxone,
NaHCO3, Na2EDTA, CH3CN, H2O, 97%; (v) silica gel chroma-
tography.
a Reagents and conditions: (i) MeOH; silica gel chromatography;
(ii) silica gel chromatography, 98% from 16; (iii) NaOMe, MeOH;
(iv) Pd-C, H2, MeOH then HCl, Et2O, yield of 20 estimated to be
∼60% in crude product, yield is 23% after purification by
chromatography.
by an acetolysis reaction followed by reaction with tri-
methylsilyl azide catalyzed by tin(IV) chloride in dichloro-
methane. Elimination of hydrogen iodide was effected as
before to give the desired 1-azido-6-deoxyhex-5-enopyrano-
side 15 in 50% yield. It was possible in this case to isolate
a mixture of epoxides 1617 from the reaction of 15 with
methyl(trifluoromethyl)dioxirane, generated in situ, despite
the presence of water in the reaction mixture. However,
prolonged reaction time or attempts to separate the epoxides
by chromatography lead to formation of the hexos-5-ulose
derivative 17.
have developed a new synthesis of 5-ketomannose from
TMS-protected 6-deoxyhex-5-enopyranosides; this method-
ology should be applicable to the synthesis of a range of
other 1,5-dicarbonyl derivatives. It has also been shown that
epoxides prepared from 1-azido-6-deoxyhex-5-enopyrano-
sides can be used in the synthesis of deoxymannojirimycin.
The sequence is being further optimized so that it will be
useful for preparation of multigram quantities of the desired
product and its derivatives. The efficiency of this route can
be compared with the overall yields possible by the Baxter
and Reitz method. They used methyl 2,3-di-O-isopropyl-
idene-R-D-mannofuranoside as a starting material and con-
verted this to 10 by selective oxidation at C-5 followed by
hydrolysis of the isopropylidene group and methyl glycoside
in 75% overall yield; Barton and co-workers have more
recently described a synthesis of the mannofuranoside in one
pot from D-mannose in 83% yield.21 This route would thus
provide 10 in 63% yield over three steps and deoxymanno-
jirimycin in 18% yield over five steps from D-mannose. The
routes described herein provides 10 in 31% yield over seven
steps and 2 in 9% yield22 over nine steps. Although the
strategy is not yet as potent as that of Baxter and Reitz, there
could be advantages in some cases of using 1-azido-6-
deoxyhex-5-enopyranosides as intermediates. This may
include synthesis of oligosaccharides that incorporate an aza-
sugar component or for the synthesis of oligosaccharides
incorporating aza-C-disaccharides. The route is thus being
We decided to explore the conversion of the epoxides into
a more stable intermediate suitable for completion of the
synthesis before proceeding further. Thus reaction of 16 with
methanol proceeded smoothly to give a single product 18.
Acetate migration occurred from the 4-OH group to the 6-OH
group during attempted purification of 18 by chromatography
giving 19 (98%). The acetate protecting groups were
removed to give 20.18 Catalytic hydrogenation of 20 gave
deoxymannojirimycin 2, which can be isolated as the amine
or as its hydrochloride salt, which after purification was
(12) Garegg, P. J.; Samuelsson, B. J. Chem. Soc., Perkin Trans. 1 1980,
2866.
(13) Prolonged reaction times lead to depleted yields of product. We
had not observed these hemiketals previously from these reactions; see ref
11. Related hemiketals have been observed by other workers; see:
Taillefumier, C.; Lakhrissi, M.; Chapleur, Y. Synlett 1999, 697.
(14) Yang, D.; Wong, M.-K.; Yip, Y.-C. J. Org. Chem. 1995, 60,
3887.
(15) The 1H and 13C NMR data are identical with those previously
reported. These spectra are complex as a result of the presence of a number
of interconverting isomers that have been studied in detail previously; see:
Kiely, D. E.; Harry-O’Kuru, R. E.; Morris, P. E., Jr.; Morton, D. W.;
Riordan, J. M. J. Carbohydr. Chem. 1997, 16, 1159.
(16) A cyclic imine was first used by Paulsen in a synthesis of 1; see:
Paulsen, H.; Sangster, I.; Heyns, K. Chem. Ber. 1967, 100, 802.
(17) A 1.4:1 mixture of epoxides was obtained. The oxygen atom of the
oxirane ring is believed to be trans to the pyranose oxygen in the major
stereoisomer; this is based on chemical shift and NOE data.
(19) NMR analysis of the crude product did not indicate that the C-5
epimer was present in the reaction mixture.
(20) Fleet, G. W. J.; Ramsden, N. G.; Witty, D. R. Tetrahedron Lett.
1989, 30, 327.
(21) Barton, D. H. R.; Gero, S. D.; Quiclet-Sire, B.; Samadi, M.
Tetrahedron: Asymmetry 1994, 5, 2123-2136.
(22) The yield is given after purification of 2 using chromatography;
this requires further optimization as NMR indicates that the major product
in the crude reaction mixture is 2, estimated to be 60%. We are also
investigating the possibility of carrying out the final three steps in a single
pot.
1
(18) From analysis of coupling constants in the H NMR spectrum the
1
pyranose ring has a C4 conformation. Some epimerization of the R-azide
(azide in equatorial orientation) to the â-azide can be observed during
prolonged reaction times, which occur if more dilute concentrations of
sodium methoxide are used. The epimerization can be avoided by reducing
the reaction time by increasing methoxide concentration. See Supplementary
Information for details. It appears advantageous to use the pure R-azide in
the final step rather than an R/â mixture as reduction of â-isomer (azide in
axial orientation) is much slower.
Org. Lett., Vol. 3, No. 21, 2001
3355

investigated for its potential in the synthesis of deoxynojiri-
mycin and other important aza-sugars. This work and the
biological activity of analogues will be reported in due
course.
Supporting Information Available: Descriptions of
experimental procedures, analytical and spectroscopic data,
1
and selected H and 13C spectra. This material is available
free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. The authors thank Enterprise Ireland
for funding for M.T. (SC/2000/182) and for a scholarship
to J.O.B.
OL016596S
3356
Org. Lett., Vol. 3, No. 21, 2001