The first syntheses of 6,7-dihydroxylated calystegines and homocalystegines

Article abstract of DOI:10.1016/j.tetlet.2006.10.022

The synthesis of novel calystegine analogues containing hydroxyl substituents on both C6 and C7 positions has been achieved. The synthetic strategy employed is based on stereoselective dihydroxylation reactions and has also been utilised for the preparati

Full text of DOI:10.1016/j.tetlet.2006.10.022

Tetrahedron Letters 47 (2006) 9147–9150
The first syntheses of 6,7-dihydroxylated calystegines
and homocalystegines
Birgit Groetzl, Sandeep Handa* and John R. Malpass
Department of Chemistry, University of Leicester, Leicester, LE1 7RH, UK
Received 14 August 2006; revised 27 September 2006; accepted 5 October 2006
Available online 3 November 2006
Abstract—The synthesis of novel calystegine analogues containing hydroxyl substituents on both C6 and C7 positions has been
achieved. The synthetic strategy employed is based on stereoselective dihydroxylation reactions and has also been utilised for the
preparation of homocalystegines.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Calystegines 1 are polyhydroxylated nortropane alka-
loids that were first isolated by Tepfer et al. from the
roots and root extrudates of Calystegia sepium.1 They
are rigid azasugar derivatives and have bicyclic struc-
tures that are similar to swainsonine 2, castanospermine
3 and australine 4. Azasugars are often strong inhibitors
of glycosidase enzymes and have consequently been
investigated as potential therapies in the treatment of
diabetes, HIV and cancer.2 In common with these bio-
logical properties, calystegines have also been shown
to be selective glycosidase inhibitors. For example, caly-
stegine B2, which has a hydroxylation pattern resem-
bling that of D-glucose, shows 100-fold higher
inhibition of b-glycosidases from almonds and Caldocel-
lum saccharolyticum than deoxynojirimycin 5.3
cerning the synthesis of novel 6,7-dihydroxylated caly-
stegines and the corresponding homocalystegine
analogues.
OH
OH
OH
OH
HN
5
N
N
H
1
4
HO
HO
HO
2
6
3
H
OH
HO
HO
OH
OH
7
1
2
3
OH
OH
HN
N
H
OH
OH
HO
OH
OH
HO
4
5
A number of naturally occurring calystegines have been
isolated (and synthesised) to date.4 These compounds
have been found to possess between three to five hydr-
oxyl groups around the bicyclic core. However, in spite
of this variable hydroxylation pattern, none of them
contain hydroxyl groups on both C6 and C7. In addition,
only a small number of non-natural calystegines have
been synthesised and few have been tested with respect
to their biological activity.5 This prompted us to explore
an easily adaptable synthetic route to non-natural caly-
stegines in order to further elucidate the structure–activ-
ity profile of this class of compounds as glycosidase
inhibitors. This letter presents our initial results con-
Our synthetic strategy towards non-natural calystegines
and homocalystegines is outlined in Scheme 1. The
known bicyclic starting materials 6a6 and 6b7 can be
readily accessed via cycloaddition of a nitroso-com-
pound (generated in situ from the reaction of benzyl
N-hydroxycarbamate with tetra-n-butyl-ammonium
periodate)8 with cycloheptadiene or cyclooctadiene,
respectively. We envisaged that the known conversion6,7
of 6a–b to the corresponding alcohols 7a–b and ketones
8a–b would allow us to investigate the stereoselective
dihydroxylation of each of these three species with a
view to access both possible syn-diols at C6 and C7.
Although the syntheses described below are racemic, it
should be noted that an enantioselective version of the
cycloaddition reaction to produce 6a–b is possible thus
allowing the methods reported here to be used for the
production of homochiral material.9
*
Corresponding author. Tel.: +44 116 2522128; fax: +44 116
2523789; e-mail: s.handa@le.ac.uk
0040-4039/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.10.022

9148
B. Groetzl et al. / Tetrahedron Letters 47 (2006) 9147–9150
NHCbz
NHCbz
NHCbz
NHCbz
NHCbz
NHCbz
NCbz
O
HO
HO
HO
a
b
+
(
)
n
(
)
n
HO
(
)
n
O
HO
O
O
HO
O
6a: n=1
: n=2
7a: n=1
7b
8a: n=1
8b
: n=2
7a
8a
13
14
(5:2)
6b
: n=2
c
c
+Cl-
NHCbz
( )n
+
NHCbz
NH3 Cl-
NH3+ Cl-
NH2
NCbz
O
HO
HO
HO
HO
HO
HO
HO
HO
HO
HO
HO
HO
(
)
n
HO
(
)
n
HO
O
O
O
15-bicycl
15-mono
16-mono
Scheme 3. Reagents and conditions: (a) Jones reagent, acetone, 0 ꢁC,
5 min (62%); (b) OsO4, NMO, acetone/H2O, rt, 24 h (13–51%; 14–
20%) and (c) Pd(OAc)2, H2, EtOH:HOAc 18:1, rt; then H2O/HCl (15-
100%; 16–68%).
H
N
HO
HO
( )n
HO
Scheme 1.
7a and 8a (Scheme 3). The alcohol 7a was readily ac-
cessed from 6a by N–O bond cleavage using Mo(CO)6.16
However, syn-dihydroxylation of 7a under Upjohn con-
ditions was non-selective and gave an inseparable (1:1)
mixture of diastereoisomeric triols. A similar lack of
facial selectivity has previously been observed in the
epoxidation of 7a.17
We started our investigations with the seven-membered
calystegine series. Dihydroxylation of 6a under Upjohn
conditions gave the b-diol 9 exclusively in good yield
and with no evidence of attack from the a-face (Scheme
2).10 The stereostructure of 9 was determined by
NOESY spectroscopy which indicated strong cross
peaks between the C6 and C7 hydrogen atoms and the
C3 a-hydrogen.11,12 Whilst the stereoselectivity for the
dihydroxylation of 6a is in agreement with the exo-at-
tack observed for bicyclo[2.2.1]heptenes,13 it should be
noted that an analogous bicyclo[2.2.2]oxazine substrate
reportedly undergoes endo-selective dihydroxylation.14
Protection of the diol 9 followed by N–O bond cleavage
using samarium diiode15 gave the alcohol 10. Dess–Mar-
tin oxidation of 10 then generated the corresponding ke-
tone which existed predominantly as the hemiaminal 11
(>90% by NMR). Removal of the Cbz group to give 12
proved straightforward, but unfortunately all efforts to
remove the isopropylidene protecting group from 12
using acidic catalysis met with failure due to significant
decomposition of the starting material and/or product
under the reaction conditions.
Jones oxidation of 7a produced the ketone 8a18 in good
yield. Pleasingly, dihydroxylation of 8a proved to be
more facially selective than the reaction of 7a and affor-
ded a mixture of 13 and 14 (ratio ca. 5:2) which could be
separated by careful column chromatography and
recrystallisation.19 Subsequent deprotection of 13
cleanly generated the novel calystegine analogue 15.
Interestingly, NMR analysis indicated that 15 exists pre-
dominantly in its monocyclic form rather than as the
hemiaminal (15-mono:15-bicycl = 10:1).20 This is in
contrast to the majority of naturally occurring calyste-
gines that have been isolated.4 In an analogous manner,
deprotection of 14 gave the novel calystegine analogue
16 which again exists mainly as the monocyclic isomer
(16-mono:16-bicycl = 2:1).21
Having successfully accessed the novel calystegine ana-
logues 15 and 16 we next examined the dihydroxylation
reactions of 6b–8b with a view to produce the corre-
sponding homocalystegines. However, in contrast to
the reaction of 6a, the dihydroxylation of the bicyclic
oxazine 6b led only to a mixture of unidentifiable prod-
ucts. The reaction of alcohol 7b (produced from 6b)7
was more successful, affording 17 as a single stereoiso-
mer (Scheme 4).22 The increased facial selectivity in
the dihydroxylation reaction of 7b compared to 6b has
been noted before in the corresponding epoxidation of
these substrates.17,23 Disappointingly, all attempts to
protect the syn-diol in 17 led only to selective protection
of the anti-1,2-diol functionality,24 thus limiting the
effectiveness of this route. Oxidation of the alcohol 7b
gave the corresponding ketone as a separable mixture
of keto-8b and hemiaminal-8c tautomers (8b:8c =
1:2).7,25 Gratifyingly, dihydroxylation of the bicyclic
tautomer 8c proceeded in a stereoselective manner to af-
ford the triol 18 in good yield.26 Subsequent deprotec-
In view of the failure to effect the deprotection of 12 we
turned our attention to the dihydroxylation reactions of
NCbz
O
NHCbz
NCbz
O
O
HO
HO
b, c
a
O
HO
6a
9
10
d
H
N
NCbz
O
O
e
O
O
HO
HO
12
11
Scheme 2. Reagents and conditions: (a) OsO4, NMO, acetone/H2O, rt,
48 h, (76%); (b) dimethoxypropane, p-TSA, acetone, rt, 72 h, (67%); (c)
SmI2, THF, rt, 1 h, (88%); (d) Dess–Martin periodinane, CH2Cl2, rt,
18 h, (48%) and (e) Pd(OH)2, H2, EtOAc, 18 h, (99%).

B. Groetzl et al. / Tetrahedron Letters 47 (2006) 9147–9150
9149
9. Faitg, T.; Soulie, J.; Lallemand, J-Y.; Ricard, L. Tetra-
hedron: Asymmetry 1999, 10, 2165–2174.
10. The a- and b-faces are defined as used in tropane systems.
11. All new compounds gave satisfactory spectroscopic and
microanalytical and/or HRMS analysis.
NHCbz
NHCbz
NCbz
NCbz
HO
HO
b
c
HO
HO
O
HO
7b
8b
8c
18
12. For 9, dH (300 MHz, CDCl3): 1.30–1.47 (1H, m, a-H3),
1.67–1.74 (1H, m, b-H3), 1.85–1.97 (3H, m, 2 · H4 &
1 · H2), 2.01–2.10 (1H, m, 1 · H2), 3.91 (1H, d,
J = 8.3 Hz, H7), 4.03 (1H, dd, J = 8.3, 1.8 Hz, H6), 4.48
(1H, dd, J = 3.8, 1.8 Hz, H1), 4.58 (1H, br s, H5), 5.21
(2H, s, CH2–Ph), 7.29–7.40 (5H, m, Ph). dC (75 MHz,
CDCl3) 19.1 (CH2), 28.9 (CH2), 29.8 (CH2), 57.5 (CH),
65.1 (CH), 65.8 (CH), 67.7 (CH2), 82.0 (CH), 128.1 (CH),
128.2 (CH), 128.5 (CH), 136.1 (C), 155.5 (C). Accurate
mass (FAB): found: 294.1342 (M+H+); calcd. for
C15H20NO5: 294.1342.
13. Donohoe, T. J.; Mitchell, L.; Waring, M. J.; Helliwell,
M.; Bell, A.; Newcombe, N. J. Org. Biol. Chem. 2003, 1,
2173–2186.
14. Noguchi, H.; Aoyama, T.; Shioiri, T. Heterocycles 2002,
58, 471–504.
d
a
+ Cl-
+ Cl-
NHCbz
NH3
O
NH2
HO
HO
HO
HO
HO
HO
HO
HO
17
19-mono
19-bicycl
Scheme 4. Reagents and conditions: (a) OsO4, NMO, acetone/H2O, rt,
72 h (72%); (b) Jones reagent, acetone, 0 ꢁC, 5 min (77%–8b:8c; 1:2); (c)
OsO4, NMO, acetone/H2O, rt, 72 h (68%); (d) Pd(OAc)2, H2,
EtOH:HOAc 18:1, rt; then H2O/HCl (68%).
15. Pulz, R.; Al-Harassi, A.; Reissig, H.-U. Org. Lett. 2002, 4,
2353–2355.
16. Ritter, A. R.; Miller, M. J. J. Org. Chem. 1994, 59, 4602–
4611.
tion gave the novel homocalystegine 19 which again
exists mainly as the monocyclic isomer (19-mono:19-
bicycl = 10:3).27
17. Justice, D. E.; Malpass, J. R. Tetrahedron 1996, 52, 11963–
11976.
18. Ketone 8a exists solely as the monocyclic tautomer
shown—see Ref. 6.
19. Compound 13 exists solely as the monocyclic keto-
tautomer shown. The stereostructure of 13 was confirmed
by protection of the diol as the isopropylidene acetal to
give 11 and comparison to the material synthesised by the
route shown in Scheme 2.
20. For 15-mono (assignments based on calystegine number-
ing); dH (300 MHz, D2O): 1.52–1.75 (2H, m, 1 · H3 &
1 · H4), 1.80–1.95 (2H, m, 1 · H3 & 1 · H4), 2.30 (1H,
dddd, J = 2.1, 6.6, 10.0, 12.8 Hz, 1 · H2), 2.57 (1H, dt,
J = 12.8, 3.5 Hz, 1 · H2), 3.50–3.59 (1H, m, H5), 4.04 (br
s, 1H, H6), 4.70 (1H, H7 overlap with D2O); dC (75 MHz,
D2O) 17.7 (CH2), 26.4 (CH2), 40.2 (CH2), 53.9 (CH), 72.8
(CH), 77.1 (CH), 212.5 (C); accurate mass (FAB): found:
160.0973 (MÀClÀ); calcd. for C7H14NO3: l60.0974.
Selected signals for 15-bicycl; dH (300 MHz, D2O): 3.99
(1H, d, J = 6.9 Hz, H6 or H7), 4.19 (1H, d, J = 6.9 Hz, H6
or H7), all other 1H signals overlap with 15-mono; dC
(75 MHz, D2O) 16.2 (CH2), 24.1 (CH2), 32.4 (CH2), 61.9
(C), 69.8 (CH), 70.9 (CH), 92.8 (C).
In summary, we have reported the first synthesis of non-
natural calystegines 15 and 16 and the first homocalyste-
gine 19 in racemic form. Surprisingly, we have found
that these natural product analogues exist mainly as
the monocyclic keto-tautomer which is in contrast to
the bicyclic hemiaminal-form reported for the majority
of the calystegines isolated to date. Biological testing
of 15, 16 and 19 produced in the course of this work
is currently underway and these results together with
the synthesis of further calystegine analogues will be
reported in due course.
Acknowledgement
The authors thank the University of Leicester for finan-
cial support of this research.
References and notes
21. For 16-mono (assignments based on calystegine number-
ing); dH (300 MHz, D2O): 1.50–2.06 (4H, m, 2 · H3 &
2 · H4), 2.46 (1H, dddd, J = 16.5, 15.2, 8.1, 7.2 Hz,
1 · H2), 2.65 (1H, dt, J = 16.5, 5.6 Hz, 1 · H2), 3.45
(1H, dt, J = 3.2, 8.7 Hz, H5), 3.88 (1H, dd, J = 8.7,
2.4 Hz, H6), 4.47 (1H, d, J = 2.4 Hz, H7); dC (75 MHz,
D2O) 18.0 (CH2), 27.6 (CH2), 39.3 (CH2), 54.7 (CH), 71.9
(CH), 78.5 (CH), 214.9 (C); accurate mass (FAB): found:
160.0975 (MÀClÀ); calcd. for C7H14NO3: 160.0974.
22. The stereostructure of 17 was confirmed by X-ray
crystallographic analysis of the derivative 20.
1. Tepfer, D.; Goldmann, A.; Pamboukdjian, N.; Maille, M.;
´
´
Lepingle, A.; Chevalier, D.; Denarie, J.; Rosenberg, C.
J. Bacteriol. 1988, 170, 1153–1161.
2. Asano, N. Glycobiology 2003, 13, 93–104.
3. Iminosugars as Glycosidase Inhibitors: Nojirimycin and
Beyond; Stutz, A. E., Ed.; Wiley-VCH: Weinheim, 1999.
¨
4. Dra¨ger, B. Nat. Prod. Rep. 2004, 21, 211–223, and
references cited therein.
5. For examples, of calystegine analogues see—Garcia-
Moreno, M. I.; Mellet, C. O.; Fernandez, J. M. G. Eur.
J. Org. Chem. 2004, 1803–1819.
6. Justice, D. E.; Malpass, J. R. J. Chem. Soc., Perkin Trans.
1 1994, 2559–2564.
7. Smith, C. R.; Justice, D. E.; Malpass, J. R. Tetrahedron
1993, 49, 11037–11054.
8. For safety reasons we replaced the Me4NIO4 used in the
original preparations of 6a and 6b with n-Bu4NIO4—see
Malpass, J. R.; Hemmings, D. A.; Wallis, A. L.; Fletcher,
S. R.; Patel, S. J. Chem. Soc., Perkin Trans. 1 2001, 1044–
1050.
O
NH
NHCbz
O
p-nitrobenzoyl chloride,
NEt3, DMAP
HO
HO
HO
20
O
O
HO
17
O2N

9150
B. Groetzl et al. / Tetrahedron Letters 47 (2006) 9147–9150
Crystal data for compound 20: C16H18N2O7, M = 350.32,
triclinic, space group P-1, a = 7.082(2), b = 7.129(2),
25. Once separated 8c appears to be stable and does not re-
equilibrate with 8b to any appreciable extent after stand-
ing for 24 h at room temperature in CDCl3. Storage of
ketone 8b under these conditions, however, re-establishes
the 1:2 ratio of 8b:8c.
3
˚
˚
c = 17.347(5) A, V = 781.4(4) A , T = 150(2) K, Z = 2,
Dc = 1.489 g cmÀ3 (Mo-Ka) = 0.118 mmÀ1
,
l
.
Final,
R1 = 0.0518 (for 5693 reflections with I > 2r(I)) and
wR2 = 0.0642 for all data. Crystallographic data has been
deposited with the Cambridge Crystallographic Data
Centre as supplementary publication number CCDC
617411. Copies of the data can be obtained, free of
charge, on application to CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK (fax: +44 1223 336033 or
e-mail: deposit@ccdc.cam.ac.uk).
26. Triol 18 exists mainly as the hemiaminal tautomer shown.
27. For 19-mono (assignments based on calystegine number-
ing); dH (300 MHz, D2O): 1.31–2.04 (6H, m, 2 · H3,
2 · H4 and 2 · H5), 2.26–2.34 (1H, m, 1 · H2), 2.73 (1H,
dt, J = 3.5, 12.8 Hz, 1 · H2), 3.44–3.51 (1H, m, H6), 4.55
(1H, d, J = 2.5 Hz, H8), 4.69 (1H, H7, overlap with D2O);
dC (75 MHz, D2O) 22.6 (CH2), 26.6 (CH2), 35.8 (CH2),
38.2 (CH2), 54.9 (CH), 72.5 (CH), 78.7 (CH), 215.7 (C);
accurate mass (FAB): found: 174.1130 (MÀClÀ); calcd.
for C8H16H16NO3: 174.1130. Selected signals for 19-
bicycl; dH (300 MHz, D2O): 3.65–3.67 (1H, m, H6), 4.12
(1H, d, J = 5.3 Hz, H8), 4.24 (1H, dd, J = 2.5, 5.3 Hz,
H7), all other 1H signals overlap with 19-mono; dC
(75 MHz, D2O) 20.4 (CH2), 21.7 (CH2), 26.8 (CH2), 27.5
(CH2), 60.4 (C), 66.6 (CH), 73.9 (CH), 95.5 (C).
23. Justice, D. E.; Malpass, J. R. Tetrahedron 1996, 52, 11947–
11962.
24. For example:
NHCbz
NHCbz
HO
HO
HO
O
dimethoxypropane,
74%
p-TSA
HO
O
17