Facile syntheses of 1-deoxynojirimycin (DNJ) and 1-deoxymannojirimycin (DMJ)

Article abstract of DOI:10.1016/j.tetlet.2007.01.110

1-Deoxynojirimycin (DNJ) and 1-deoxymannojirimycin (DMJ) were synthesized through a concise and practicable pathway. A strategy of using carbohydrate derivatives bearing two leaving groups in the preparation of azasugars by di-N-alkylation of amines was developed. The strategy involved the selective partial protection of dibromo alditols.

Full text of DOI:10.1016/j.tetlet.2007.01.110

Tetrahedron Letters 48 (2007) 3115–3118
Facile syntheses of 1-deoxynojirimycin (DNJ)
and 1-deoxymannojirimycin (DMJ)
a
a,b,
*
Xuezheng Song and Rawle I. Hollingsworth
a
Department of Chemistry, Michigan State University, East Lansing, MI 48824, USA
Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI 48824, USA
b
Received 15 February 2006; revised 18 January 2007; accepted 19 January 2007
Available online 25 January 2007
Abstract—1-Deoxynojirimycin (DNJ) and 1-deoxymannojirimycin (DMJ) were synthesized through a concise and practicable path-
way. A strategy of using carbohydrate derivatives bearing two leaving groups in the preparation of azasugars by di-N-alkylation of
amines was developed. The strategy involved the selective partial protection of dibromo alditols.
Ó 2007 Published by Elsevier Ltd.
Because of their inhibition property toward glycosidase,
azasugars have long been suggested as drug candidates
to different kinds of carbohydrate-mediated diseases.
groups and neighboring –OH groups was found to be
important in these reactions. The epoxide appeared to
be a better alkylating group than the bromino-alkyl
group.
1
1
-Deoxynojirimycin (DNJ) 1 and 1-deoxymannojirimy-
cin (DMJ) 2 are among the most important (Fig. 1).
Their syntheses are of great importance. Although many
elegant synthetic methodologies toward DNJ and DMJ
To synthesize 6-membered ring instead of 5-membered
ring azasugars, the involvement of protecting groups is
inevitable. Our preliminary investigation showed that
full protection of hydroxyl groups of dibromo sugars
prohibits the epoxide formation between bromo group
and neighboring –OH group and greatly decreases the
amine N-alkylation reactivity toward bromo leaving
groups. We therefore developed a synthetic strategy to-
ward DNJ 1 and DMJ 2 from sugar lactones (Scheme 1)
based on the above findings. Selective and partial
protection of dibromoalditols could generate good
substrates for double N-alkylation of amines. The
dibromoalditols, on the other hand, could be readily
obtained from sugar lactones. 1,2-Diacetal protecting
2
have been developed, only a few of them are industri-
ally practicable. The nitrogen introduction often
involved azide substitution. Heavy-metal catalysis was
also widely used to install ring systems or hydroxyl
groups. Based on the viewpoint to avoid these undesired
processes in their syntheses, here we introduce a short
and practicable synthetic pathway of DNJ and DMJ
from potentially abundant carbohydrate material.
The 2,6-dibromination of sugar lactones has become a
3
very important derivatization of sugar lactones, intro-
ducing leaving groups into carbohydrate structures. The
5
reduction of 2,6-dibromoaldonolactones generally gives
group was initially chosen to avoid high ring strain
4
2
,6-dibromoalditols. Straightforwardly, di-N-alkylation
of the target intermediate structure by forming a
[4,4,0] trans-ring system, which could also facilitate the
azasugar ring cyclization.
of amines with 2,6-dibromoalditols by displacement of
the two bromo groups should afford 2,6-iminoalditols,
which are 6-membered ring azasugars. The research of
Lundt’s group showed that various naked 2,6-dibromo-
alditols react with ammonia to give 5-membered ring
From commercially available L-gulono-1,4-lactone 5,
2,6-dibromo-2,6-dideoxy-L-iditol 6 was easily prepared
4
4
iminoalditiols. Epoxide formation between bromo
by dibromination followed by reduction. When 6
was treated with 2,2,3,3-tetramethoxybutane under
catalysis of boron trifluoride etherate (BF ÆOEt ), a
3
2
0 0
1:1 inseparable mixture of 3,4-O- and 4,5-O-(2 ,3 -
dimethoxybutane-2 ,3 -diyl) protected dibromo-L-iditol
was obtained (Scheme 2). This mixture was treated with
ꢀ
Keywords: Deoxynojirimycin; Deoxymannojirimycin; N-Alkylation.
0
0
*
Corresponding author. Tel.: +1 517 353 0613; fax: +1 517 432
113; e-mail: rih@cem.msu.edu
1
0040-4039/$ - see front matter Ó 2007 Published by Elsevier Ltd.
doi:10.1016/j.tetlet.2007.01.110

3116
X. Song, R. I. Hollingsworth / Tetrahedron Letters 48 (2007) 3115–3118
OH
NH
OH
OH
NH
HO
HO
HO
HO
HO
HO
N
OH
OH
HO
HO
N
OH
OH
1
2
3
4
1
-Deoxynojirimycin 1-Deoxymannojirimycin
Castanospermine
Swainsonine
(
DNJ)
(DMJ)
Figure 1. Several important azasugars.
OH
NH
OH
Br
OP
OMe
O
HO
HO
HO
HO
Br
NH
O
OH
OH
OH
OMe
DNJ or DMJ
2
,6-Dibromo-L-iditol
or D-glucitol
Scheme 1. The retro-synthetic analysis of 1-deoxynojirimycin (DNJ) and 1-deoxymannojirimycin (DMJ).
OH
OH
O
OH
Br
c, d
O
a,b
HO
HO
H
Br
OH
HO
OH
5
6
OTBS
NH
OMe
O
OTBS
OMe
OTBS
NH2
OMe
Br
Br
f
O
O
Br
O
O
O
OH
OH
OH
OMe
OMe
OMe
7
e
9
11
+
g
+
Br
Br
OMe
OMe
OH
O
OH
TBSO
~
TBSO
O
O
O
O
NH
OH
HO
HO
Br
OMe
OMe
8
10
~1:1, easily separated
1
1:1, not separated
Scheme 2. The synthesis of 1-deoxynojirimycin (DNJ). Reagents and conditions: (a) 30% HBr in acetic acid, then MeOH; (b) NaBH
4
, MeOH then
, MeOH, rt; (d) TBDMSCl, Py, rt; (e) aqueous NH in MeOH,
rt; 9: 28%; 10: 26% over three steps; (f) DBU, toluene, reflux, 74%; and (g) 4 N HCl, MeOH, 95%.
+
Amberlite 120 H resin; 55% over two steps; (c) 2,2,3,3-tetramethoxybutane, BF
3
ÆOEt
2
3
t-butyldimethylsilyl chloride (TBDMSCl) to selectively
protect the primary hydroxyl groups to give an insepara-
ble mixture of 7 and 8. After the mixture of 7 and 8
was treated with aqueous ammonia in methanol, two
ammonia to attack under these reaction conditions,
therefore it stayed untouched. This interesting differenti-
ation greatly facilitated the separation process. Com-
pound 9 was then refluxed with DBU in toluene to
furnish 2,6-iminoalditol 11, which was deprotected by
hydrogen chloride in methanol to afford DNJ 1 as its
hydrochloride salt.
0
0
major products, (2 S,3 S)-6-Amino-2-bromo-1-O-t-
0
0
0
butyldimethylsilyl-2,6-dideoxy-3,4-O-(2 ,3 -dimethoxy-
0
0
0
butane-2 ,3 -diyl)-L-iditol 9 and (2 R,3 R)-2,3-anhy-
drous-6-bromo-1-O-t-butyldimethyl-silyl-6-deoxy-4,5-O-
2 ,3 -dimethoxybutane-2 ,3 -diyl)-L-gulitol 10, were
0
0
0
0
(
In the above synthesis, the partial protection of di-
bromo-L-iditol made epoxide formation possible under
mild reaction conditions, which facilitated the inter-
molecular introduction of amine functionality. Subsequent
intramolecular N-alkylation cyclized the desired 6-mem-
bered azasugar ring. Because both 3,4- and 4,5- posi-
tions of 6 have threo- configurations, there is no
apparent selectivity between 3,4-O-diacetalation and
4,5-O-diacetalation. Fortunately, in the later steps, the
inseparable isomers were further transformed and easily
separated based on the different reactivity of sterically
different epoxides toward nucleophilic attacks. This
easily separated by column chromatography. Under
basic conditions, epoxides form from bromo- groups
and neighboring free hydroxyl groups. A sterically
unhindered epoxide is susceptible to nucleophilic attack
by amines to give amino alcohols while a sterically
hindered epoxide stays unchanged under mild reaction
conditions. When treated with ammonia at room
temperature, compound 7 transformed to a 5,6-anhy-
drosugar, which subsequently reacted with ammonia
to yield 9. In contrast, compound 8 transformed to
2
,3-anhydrosugar 10, which was too hindered for

X. Song, R. I. Hollingsworth / Tetrahedron Letters 48 (2007) 3115–3118
3117
process, however, decreased the efficiency of the synthe-
sis. Isopropylidene protecting group, on the other hand,
is easier to handle and frequently shows selectivity to-
ward different vicinal –OH’s. We therefore modified
the synthesis with the isopropylidene protecting group.
smoothly to afford compound 15, which after hydrolysis
yielded DNJ 1.
We then applied this methodology to the synthesis of
DMJ 2 (Scheme 4). Bromine oxidation of D-mannose
gave D-manno-1,4-lactone, which was then treated with
hydrogen bromide in acetic acid. The resulted 2,6-di-
bromo-2,6-dideoxy-D-glucono-1,4-lactone was reduced
by sodium borohydride to yield 2,6-dibromo-2,6-di-
deoxy-D-glucitol 19. Compound 19 was treated with
acetone to afford 2,6-dibromo-2,6-dideoxy-3,4-O-iso-
propylidene-D-glucitol 20. Unlike in the case of DNJ
synthesis, extension of reaction time favored the desired
product 20 due to the D-gluco configuration (3,4-threo
and 4,5-erythro). The following steps were essentially
the same as above for the DNJ synthesis with similar
yields.
Dibromo-L-iditol 6 was treated with acetone under
catalysis of p-toluenesulfonic acid (PTSA) for 5 min.
Fortunately, the desired isomer 2,6-dibromo-2,6-dide-
oxy-3,4-O-isopropylidene-L-iditol 12 was found out to
be the dominant product (Scheme 3). The ratio between
1
2 and 2,6-dibromo-2,6-dideoxy-4,5-O-isopropylidene-
L-iditol was 5:1. When the reaction time was extended
to 2 h, a ꢀ1:1 ratio of these two major isomers was even-
tually reached. This is also due to the L-ido configuration
of 6. The following steps were similar to those when 1,2-
diacetal was used as the protecting group except for the
final cyclization step. In this case, when DBU were used
as the acid scavenger, elimination reaction dominated.
This suggested that the cyclization process was slower
when isopropylidene protecting group was used, pre-
sumably due to the higher strain of resulted 5,6-trans
fused ring system 15. The problem was successfully
solved by using a polar aprotic solvent with a weak base
such as sodium acetate. The cyclization occurred
In summary, DNJ was synthesized in five steps from 2,6-
dibromo-2,6-dideoxy-L-iditol 2. Both 1,2-diacetal and
isopropylidene were employed as protecting groups.
Although 1,2-diacetal protection showed to facilitate
transfused ring system formation more, isopropylidene
was superior for the actual synthesis. DMJ was also syn-
thesized similarly to demonstrate the methodology
OH
OTBS
OH
OH
Br
Br
Br
a
b
c
O
O
Br
HO
HO
Br
Br
O
O
OH
OH
13
6
12
OTBS
NH
OTBS
OH
Br
e
O
d
O
NH
OH
NH2
HO
HO
O
O
OH
OH
14
15
1
Scheme 3. Synthesis of 1-deoxynojirimycin (DNJ) using isopropylidene as the protecting group. Reagents and conditions: (a) acetone, p-TsOH, 72%;
(
b) TBDMSCl, Py; (c) aqueous NH
3
in MeOH, rt; 81% over two steps; (d) NaOAc, MeNO
2
, reflux, 68%; and (e) 4 N HCl, MeOH, 96%.
OH
OH
OH
O
HO
HO
HO
O
O
HO
Br
c
a
O
b
O
H
H
OH
HO
OH
HO
Br
16
17
18
OH
OH
OH
Br
OTBS
Br
OH
Br
Br
OH
e
f
d
O
O
HO
HO
O
Br
O
Br
21
19
20
OH
Br
OTBS
OH
NH
OH
NH
TBSO
HO
HO
HO
g
h
O
NH2
O
O
O
22
23
2
Scheme 4. Synthesis of 1-deoxymannojirimycin (DMJ). Reagents and conditions: (a) Br
2
, NaHCO
3
, water; (b) 30% HBr in acetic acid; (c) NaBH
4
,
+
H
resin, water; (d) acetone; 52% over four steps; (e) TBDMSCl, Py; (f) aqueous NH
3
in MeOH, rt; 80% over two steps; (g) NaOAc, MeNO
2
, reflux,
74%; and (h) 4 N HCl, MeOH, 95%.

3118
X. Song, R. I. Hollingsworth / Tetrahedron Letters 48 (2007) 3115–3118
scope. None of the above processes involved any azide
or heavy-metal catalysis. Only cheap and easily avail-
able reagents were used. All these features made the syn-
thesis not only of academic interest, but also practicable.
This methodology could be expanded to the synthesis of
other N-alkylated DNJ and DMJ by changing the
amine source. Furthermore, intermediates 11, 15, and
2. Examples of syntheses of azasugars: (a) Rudge, A.; Collins,
I.; Holmes, A.; Baker, R. Angewandte Chemie 1994,
1
06, 2416–2418; (b) Matos, C.; Lopes, R.; Lopes, C.
Synthesis 1999, 4, 571–573; (c) Comins, D.; Fulp, A.
Tetrahedron Lett. 2001, 42, 6839–6841; (d) Knight, J. G.;
Tchabanenko, K. Tetrahedron 2003, 59, 281–286; (e)
Somfai, P.; Marchand, P.; Torsell, S.; Lindstrom, U.
Tetrahedron 2003, 59, 1293–1299; (f) Chery, F.; Murphy,
P. V. Tetrahedron Lett. 2004, 45, 2067–2069; (g) Takahata,
H.; Banba, Y.; Sasatani, M.; Nemoto, H.; Kato, A.;
Adachi, I. Tetrahedron 2004, 60, 8199–8205; (h) Martin, R.;
Murruzzu, C.; Pericas, M. A.; Riera, A. J. Org. Chem.
2
3 are orthogonally and partially protected. They could
be used as versatile building blocks for syntheses of
other more complex azasugars such as castanospermine.
2
005, 70, 2325–2328.
. (a) Lundt, I. Topics in Current Chemistry 1997, 187, 117–
56; (b) Pedersen, C.; Bock, K.; Lundt, I. Pure and Applied
3
Acknowledgements
1
Chemistry 1978, 50, 1385–1400; (c) Bock, K.; Lundt, I.;
Pedersen, C. Carbohydr. Res. 1979, 68, 313–319; (d) Bock,
K.; Lundt, I.; Pedersen, C. Carbohydr. Res. 1981, 90, 7–16,
and 17–26; (e) Bock, K.; Lundt, I.; Pedersen, C. Carbohydr.
Res. 1982, 104, 79–85; (f) Bock, K.; Lundt, I.; Pedersen, C.;
Refn, S. Acta. Chem. Scand. 1984, B38, 555–561; (g) Bock,
K.; Lundt, I.; Pedersen, C.; Refn, S. Acta. Chem. Scand.
This is research was supported by a Strategic Grant
from the Michigan State University Foundation and by
the Michigan State University Research Excellence
Fund.
1
986, B40, 740.
Supplementary data
4. (a) Lundt, I.; Madsen, R. Synthesis 1993, 714–719; (b)
Lundt, I.; Madsen, R. Synthesis 1993, 720–724.
Experimental procedure and data are included. NMR
spectra of selected compounds are included. Supplemen-
tary data associated with this article can be found, in the
online version, at doi:10.1016/j.tetlet.2007.01.110.
5. (a) Ley, S. V.; Baeschlin, D. K.; Dixon, D. J.; Foster, A. C.;
Ince, S. J.; Priepke, H. W.; Reynolds, D. J. Chem. Rev.
2
001, 101, 53–80; (b) Ley, S. V.; Priepke, H. W. M.;
Warriner, S. L. Angew. Chem., Int. Ed. Engl. 1994, 33, 2290;
c) Montchamp, J. L.; Tian, F.; Hart, M. E.; Frost, J. W. J.
(
Org. Chem. 1996, 61, 3897; (d) Douglas, N. L.; Ley, S. V.;
Osborn, H. M. I.; Owen, D. R.; Priepke, H. W. M.;
Warriner, S. L. Synlett 1996, 793; (e) Berens, U.; Leckel, D.;
Oepen, S. C. J. Org. Chem. 1995, 60, 8204; (f) Hense, A.;
Ley, S. V.; Osborn, H. M. I.; Owen, D. R.; Poisson, J. F.;
Warriner, S. L.; Wesson, K. E. J. Chem. Soc., Perkin Trans.
References and notes
1
. Iminosugars as Glycosidase Inhibitors: Nojirimycin and
Beyond; St u¨ tz, A. E., Ed.; Wiley-VCH: Weinheim, Ger-
many, 1999.
1
1997, 2023.