Synthesis and mannosidase inhibitory activity of 6- and 7-substituted analogs of swainsonine

Article abstract of DOI:10.1016/S0040-4039(01)01778-6

Swainsonine (1), an inhibitor of the important glycoprotein-processing enzyme Golgi α-mannosidase II, is a clinical candidate for cancer treatment. Analogs bearing substituents at C-6 and C-7 have been prepared and evaluated as inhibitors of α-mannosidase

Full text of DOI:10.1016/S0040-4039(01)01778-6

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 8273–8276
Synthesis and mannosidase inhibitory activity of 6- and
7-substituted analogs of swainsonine
William H. Pearson* and Erik J. Hembre†
Department of Chemistry, University of Michigan, Ann Arbor, MI 48109-1055, USA
Received 27 August 2001; revised 19 September 2001; accepted 21 September 2001
Abstract—Swainsonine (1), an inhibitor of the important glycoprotein-processing enzyme Golgi a-mannosidase II, is a clinical
candidate for cancer treatment. Analogs bearing substituents at C-6 and C-7 have been prepared and evaluated as inhibitors of
a-mannosidase (jack bean), a closely related enzyme. © 2001 Published by Elsevier Science Ltd.
1. Introduction
would allow linking to an affinity matrix in order to
facilitate the isolation of human GMII.15
Swainsonine (1) is a potent inhibitor of certain a-man-
nosidases, and has proven useful as a biochemical tool
for the study of glycoprotein processing, since it inhibits
a key late-stage enzyme in the biosynthesis of glyco-
proteins.1–6 That enzyme, Golgi a-mannosidase II
(GMII), is necessary for the formation of so-called
‘complex glycoproteins’. The altered distribution of
such glycoproteins on the surface of cancer cells is
associated with metastasis and disease progression,
hence inhibitors of GMII are potentially useful for
cancer treatment.7–10 Unfortunately, human GMII has
proven difficult to isolate and characterize.3,11,12 More
selective inhibition of GMII over other mannosidases13
is a desirable goal for cancer drug development, and
makes the synthesis of analogs of swainsonine a signifi-
cant undertaking (see accompanying paper). Many
analogs of swainsonine have been reported, e.g. those
where the oxygenation pattern, ring size, or configura-
tion has been modified,14 but these changes usually
result in a diminution of potency. We report herein the
synthesis and initial biological evaluation of 6- and
7-substituted analogs of swainsonine (2 and 3, respec-
tively) (Fig. 1). We hoped that such substituents would
afford more selectivity and potency in the inhibition of
GMII, and we were also interested in collecting SAR
data for this poorly studied system. In addition, we
hoped that the relative remoteness of the planned sub-
stituents from the key functionality of swainsonine
2. Results and discussion
The major goal of our initial work was to develop a
route to both 6- and 7-substituted analogs of swain-
sonine, hopefully from a common intermediate. We
were also interested in obtaining both diastereomers of
each of these types of analogs. The scarcity of swain-
sonine and difficulties associated with the manipulation
of its functional groups made it a poor choice of
starting materials; thus, a totally synthetic route was
developed.
Our initial efforts were aimed at producing 6-substi-
tuted swainsonine analogs, and resulted in the synthesis
of both epimers of 6-ethylswainsonine (18 and 20) and
* Corresponding author. E-mail: wpearson@umich.edu
† Present address: Lilly Research Laboratories, Lilly Corporate Cen-
ter, Indianapolis, IN 46285, USA.
Figure 1. Swainsonine (1) and potential analogs (2) and (3).
0040-4039/01/$ - see front matter © 2001 Published by Elsevier Science Ltd.
PII: S0040-4039(01)01778-6

8274
W. H. Pearson, E. J. Hembre / Tetrahedron Letters 42 (2001) 8273–8276
6-(2-benzyloxy)ethylswainsonine (19 and 21), as shown
in Scheme 1. The allylic alcohol 5 was prepared from 4
as reported in our earlier work on the synthesis of
swainsonine itself.16 A Johnson orthoester Claisen
rearrangement17 using either trimethyl orthobutyrate or
4-benzyloxy-1,1,1-trimethoxybutane18 gave the esters 6
and 7, respectively, after desilylation, each as a 1:1
mixture of diastereomers. We chose this nonstereoselec-
tive method because we wished to prepare both epimers
of the C-6 analogs. The Ireland version of the Claisen
rearrangement19 could presumably be used for the
stereoselective formation of either of these diastereo-
mers. Conversion of 6 and 7 to the azides 8 and 9 was
followed by epoxidation and diastereomer separation to
give 10–13.20 Reductive cyclization then afforded the
indolizidin-5-ones 14–17. Substantial epimerization was
observed in the formation of 16 and 17, presumably
because of the axial orientation of the C-6 substituents
in these compounds. Finally, lactam reduction and
ketal hydrolysis afforded the desired 6-substituted
swainsonine analogs 18–21.
with the Johnson orthoester Claisen rearrangement,
should allow control of the configuration of the R
substituent. Thus, Claisen rearrangements on 24, 25,
40, and 41 followed by desilylation gave the esters 26,
27, 42, and 43, where the alkene geometry of the allylic
alcohols had been completely translated into stereocon-
trol at the R substituents. Azide formation, epoxida-
tion, and reductive cyclization was employed as
described above to produce the indolizidinones 32–35
and 48–51. In this case, the undesired epoxide
diastereomers were carried on to the indolizidinones
(i.e. 34, 35, 50, and 51), despite having the wrong
configurations at C-8 and C-8a, since we were curious
about the mannosidase inhibitory power of these bis-
epimers. Lactam reduction and acetonide removal gave
the 7-substituted swainsonine analogs 36, 37, 52, and
53, as well as the 7-substituted-8,8-diepiswainsonine
analogs 38, 39, 54, and 55.
Screening of the new analogs 18–21, 36–39, and 52–55
against jack bean a-mannosidase were carried out using
standard methods.21 This enzyme is a commercially
available enzyme that is a useful model for mammalian
a-mannosidases such as GMII.22 The IC50s measured
are reported in Table 1, where they are compared to
those of swainsonine (1). Alkylation of the swainsonine
backbone results in compounds with varying degrees of
a-mannosidase inhibitory activity. The analogs 18 and
19 that have equatorially-disposed substituents at the
C-6 position have the highest degree of inhibitory activ-
A similar route was chosen for the preparation of
7-substituted analogs of swainsonine (Scheme 2).
Reduction of 4 followed by the addition of two differ-
ent alkynyl Grignard reagents and primary alcohol
protection gave the propargylic alcohols 22 and 23,
which were transformed in a stereoselective fashion into
the Z or E allylic alcohols 24/25 and 40/41, respectively,
by selection of the appropriate reduction methods.
Control of the allylic alcohol geometry, when combined
1:1
1:1
O
O
O
O
O
a
97:3
b
O
O
O
O
O
O
O
OH
Ref. 16
R
R
TBSO
N3
HO
OMe
OMe
4
5
6 R=Et (83%)
8 R=Et (85%)
7 R=(CH2)2OBn (67%)
9 R=(CH2)2OBn (66%)
OH
OH
H
HO
H
N
O
O
O
d
e
O
HO
6
O
2
O
N
R
R
R
(6α)-substituted
swainsonine analogs
18 R=ethyl (99%)
N3
OMe
O
10 R=ethyl (35%)
11 R=(CH2)2OBn (28%)
14 R=ethyl (66%)
15 R=(CH2)2OBn (79%)
19 R=(CH2)2OBn (97%)
c
OH
OH
HO
H
N
O
O
H
O
O
d
e
1
O
HO
6
O
2
6
3
N
R
R
R
N3
(6β)-substituted
swainsonine analogs
20 R=ethyl (96%)
OMe
O
12 R=ethyl (34%)
13 R=(CH2)2OBn (28%)
16 R=ethyl (41%)
+ 20% of 14
17 R=((CH2)2OBn (13%)
+ 66% of 15
21 R=(CH2)2OBn (96%)
Scheme 1. Synthesis of 6-substituted swainsonine analogs. Reagents and conditions: (a) PrC(OMe)3 or BnO(CH2)3C(OMe)3, cat.
EtCO2H, toluene, reflux; n-Bu4NF, THF; (b) HN3, PPh3, EtO2CNꢀNCO2Et, PhH; (c) m-CPBA, CH2Cl2; separate diastereomers.
Each reaction also produced the other epoxide diastereomer (18% of 1:1 at C-2 for R=Et; 22% of 1:1 at C-2 for R=(CH2)2OBn);
(d) H2, Pd(OH)2/C, MeOH, EtOAc; NaOMe, MeOH reflux; (e) BH3·SMe2; 6N HCl/THF.

W. H. Pearson, E. J. Hembre / Tetrahedron Letters 42 (2001) 8273–8276
8275
R
O
O
O
O
R
a,b
c
d
O
O
O
O
O
O
OH
OH
R
O
HO
HO
TBSO
24 R=n-Bu (81%)
OMe
4
22 R=n-Bu (84%)
26 R=n-Bu (75%)
23 R=(CH2)2OBn (88%)
25 R=(CH2)2OBn (70%)
27 R=(CH2)2OBn (70%)
OH
OH
O
H
O
O
O
O
H
R
R
O
f
R
R
O
O
N
O
g
e
+
O
O
N
N3
N3
OMe
OMe
O
O
28 R=n-Bu (77%)
29 R=(CH2)2OBn (80%)
30 R=n-Bu (95%, 2:1
α:β epoxide)
31 R=(CH2)2OBn (75%,
32 R=n-Bu
33 R=(CH2)2OBn
34 R=n-Bu
35 R=(CH2)2OBn
h, i
1.7:1 α:β epoxide)
OH
OH
HO
H
HO
H
R
R
O
O
8
8a
7
7
R
d
R
22
or
HO
j
HO
O
O
N
N
O
OH
23
(7β)-substituted
swainsonine analogs
36 R=n-Bu (99%)
(7β)-substituted, 8,8a-diepi-
swainsonine analogs
38 R=n-Bu (99%)
TBSO
HO
OMe
40 R=n-Bu (88%)
41 R=(CH2)2OBn (73%)
42 R=n-Bu (74%)
43 R=(CH2)2OBn
(79%)
37 R=(CH2)2OBn (97%)
39 R=(CH2)2OBn (97%)
e
OH
OH
O
O
O
H
O
H
O
O
R
R
R
R
O
O
O
N
O
N
O
+
f
k
N3
44 R=n-Bu (84%)
45 R=(CH2)2OBn (87%)
N3
OMe
OMe
O
O
46 R=n-Bu (95%, 3.5:1
α:β epoxide)
48 R=n-Bu (60%)
49 R=(CH2)2OBn (68%)
50 R=n-Bu (18%)
51 R=(CH2)2OBn (17%)
47 R=(CH2)2OBn (84%,
3:1 α:β epoxide)
l, i
l, i
OH
OH
HO
H
HO
H
R
R
8
8a
7
7
HO
HO
N
N
(7α)-substituted
swainsonine analogs
52 R=n-Bu (98, 81%;
(7α)-substituted, 8,8a-diepi-
swainsonine analogs
54 R=n-Bu (92, 83%;
2 steps)
2 steps)
53 R=(CH2)2OBn (76,
97%; 2 steps)
55 R=(CH2)2OBn
(100, 86%; 2 steps)
Scheme 2. Reagents and conditions: Synthesis of 7-substituted swainsonine analogs. (a) DIBAL-H; (b) RC CMgBr; (c) H2,
Pd/BaSO4; tBuMe2SiCl, imidazole; (d) MeC(OMe)3, cat. EtCO2H, toluene, reflux; n-Bu4NF, THF; (e) HN3, PPh3,
EtO2CNꢀNCO2Et, PhH; (f) m-CPBA, CH2Cl2. Diastereomeric epoxides were not separated. (g) H2, Pd(OH)2/C, MeOH, EtOAc;
NaOMe, MeOH reflux. Diastereomers at 8,8a were not separated. Yields (ratio of swainsonine configuration:diepi-swainsonine
configuration): R=n-Bu:60% (2.4:1); R=(CH2)2OBn:89% (1.7:1). (h) BH3·SMe2; separate diastereomers; (i) 6N HCl/THF; (j)
Red-Al®; tBuMe2SiCl, imidazole; (k) H2, Pd(OH)2/C, MeOH, EtOAc; NaOMe, MeOH reflux; separate diastereomers; (l)
BH3·SMe2.
ity. Increasing the substituent size from ethyl (18) to
2-(benzyloxy)ethyl (19) results in a slight decrease in
inhibitory activity, while switching from equatorial to
axial substitution at C-6 (20 and 21) leads to a signifi-
cant decrease in activity. In contrast, most of the C-7
n-butyl and 2-(benzyloxy)ethyl-substituted compounds
show complete loss of inhibitory activity, except for 53,
which is a weak inhibitor. As expected the 8,8a-
diepiswainsonine analogs (38, 39, 54, 55) are devoid of
activity.

8276
W. H. Pearson, E. J. Hembre / Tetrahedron Letters 42 (2001) 8273–8276
Table 1. Concentration of swainsonine and analogs
required to jack bean a-mannosidase by 50% (IC50 inM)
2. Kaushal, G. P.; Elbein, A. D. Methods Enzymol. 1994,
230, 316–329.
3. Moremen, K. W.; Trimble, R. B.; Herscovics, A. Glycobi-
ology 1994, 4, 113–125.
Compound
IC50 (mM)
4. Ganem, B. Acc. Chem. Res. 1996, 29, 340–347.
5. Iminosugars as Glycosidase Inhibitors; Stu¨tz, A. E., Ed.;
Wiley-VCH: Weinheim, 1999.
6. Asano, N.; Nash, R. J.; Molyneux, R. J.; Fleet, G. W. J.
Tetrahedron: Asymmetry 2000, 11, 1645–1680.
7. Goss, P. E.; Baker, M. A.; Carver, J. P.; Dennis, J. W.
Clin. Cancer Res. 1995, 1, 935–944.
Swainsonine (1)
(6a)-Substituted analogs
18 (6a-Ethyl)
19 (6a-CH2CH2OBn)
(6b)-Substituted analogs
20 (6b-Ethyl)
21 (6b-CH2CH2OBn)
(7b)-Substituted analogs
36 (6b-n-Butyl)
37 (6b-CH2CH2OBn)
(7a)-Substituted analogs
52 (6a-n-Butyl)
53 (6a-CH2CH2OBn)
7-Substituted -8,8-diepi analogs
38 (6b-n-butyl)
39 (6b-CH2CH2OBn)
54 (6a-n-butyl)
55 (6a-CH2CH2OBn)
0.4–0.1,23,24 0.1 (our labs)25
30
230
70
275
8. Goss, P. E.; Reid, C. L.; Bailey, D.; Dennis, J. W. Clin.
NI
NI
Cancer Res. 1997, 3, 1077–1086.
9. Dennis, J. W.; Granovsky, M.; Warren, C. E. BioEssays
1999, 21, 412–421.
NI
890
10. Dennis, J. W.; Granovsky, M.; Warren, C. E. Biochim.
Biophys. Acta 1999, 1473, 21–34.
11. Moremen, K. W.; Touster, O.; Robbins, P. W. J. Biol.
Chem. 1991, 266, 16876–16885.
12. Misago, M.; Liao, Y.-F.; Kudo, S.; Eto, S.; Mattei,
M.-G.; Moremen, K. W.; Fukuda, M. N. Proc. Natl.
Acad. Sci. 1995, 92, 11766–11770.
13. Pearson, W. H.; Guo, L. Tetrahedron Lett. 2001, 42,
8267–8271.
NI
NI
NI
NI
NI=no inhibition.
Our preliminary results indicate that there is a very low
steric tolerance in the mannosidase binding pocket for
substituents at C-7 of swainsonine, while substituents at
C-6 are better accommodated. Further, the top (beta)
face of swainsonine appears more sensitive to substitu-
tion than the bottom (alpha) face. This general pattern
is consistent with the idea that the natural a-linked
oligomannoside substrate has a sterically bulky a-face
compared to its b-face. Note that the addition of the
more bulky 2-(benzyloxy)ethylsubstituent in 19 has a
relatively small effect on inhibitory activity compared
to the ethyl-substituted 18. It appears then that interac-
tions between the b-face of swainsonine and the enzyme
active site play the most important roles in binding
ability. The a-face of swainsonine, on the other hand, is
less essential for important binding interactions and
thus synthetic modifications in this region may provide
a means to differentiate between the various mannosi-
dases while maintaining good inhibitory activity.
Finally, the tolerance for relatively large substituents at
C-6 bodes well for the construction of affinity labeling
compounds and ligands for affinity chromatography.
Results of these studies, as well as the results of inhibi-
tion studies with GMII, will be reported elsewhere.
14. For examples of synthetic analogs of swainsonine, see: (a)
Cenci Di Bello, I.; Fleet, G.; Namgoong, S. K.; Tadano,
K.; Winchester, B. Biochem. J. 1989, 259, 855–861; (b)
Dennis, J. W.; White, S. L.; Freer, A. M.; Dime, D.
Biochem. Pharmacol. 1993, 46, 1459–66; (c) Pearson, W.
H.; Hembre, E. J. J. Org. Chem. 1996, 61, 5537–5545.
15. A brief report of a 7-substituted swainsonine analog has
appeared: Holmes, A. B.; Bourdin, B.; Collins, I.; Davi-
son, E. C.; Rudge, A. J.; Stork, T. C.; Warner, J. A. Pure
Appl. Chem. 1997, 69, 531–536.
16. Pearson, W. H.; Hembre, E. J. J. Org. Chem. 1996, 61,
7217–7221.
17. Johnson, W. S.; Werthemann, L.; Bartlett, W. R.; Brock-
som, T. J.; Li, T.; Faulkner, D. J.; Petersen, M. R. J. Am.
Chem. Soc. 1970, 92, 741–743.
18. Prepared from 4-benzyloxybutyronitrile according to the
general procedure of Ueno et al.; used without purifica-
tion. Ueno, H.; Maruyama, A.; Miyake, M.; Nakao, E.;
Nakao, E.; Umezu, K.; Nitta, I. J. Med. Chem. 1991, 34,
2468–2473.
19. Ireland, R. E.; Mueller, R. H.; Willard, A. K. J. Am.
Chem. Soc. 1976, 98, 2868–2877.
20. A Sharpless asymmetric dihydroxylation route has been
developed in our group (see Ref. 16), providing a more
stereoselective route than the epoxidation method.
Although we have not yet adapted the dihydroxylation
method to the current work, it should offer a more
selective route to the 6- and 7-substituted swainsonine
analogs.
Acknowledgements
21. Tropea, J. E.; Molyneux, R. J.; Kaushal, G. P.; Pan, Y.
T.; Mitchell, M.; Elbein, A. D. Biochemistry 1989, 28,
2027–2034.
22. Howard, S.; He, S.; Withers, S. G. J. Biol. Chem. 1998,
273, 2067–2072.
We thank the National Institues of Health (GM35572
and CA77365) for support of this work. E.J.H. was
supported in part by National Research Service Award
T32 GM007767.
23. Tulsiani, D. R. P.; Broquist, H. P.; Touster, O. Arch.
Biochem. Biophys. 1985, 236, 427–434.
References
24. Elbein, A. D. Ann. Rev. Biochem 1987, 56, 497–534.
25. Hembre, E. J.; Pearson, W. H. Tetrahedron 1997, 53,
11021–11032.
1. Elbein, A. D. FASEB J. 1991, 5, 3055–3063.