Stereodivergent access to polyhydroxylated 10-azabicyclo[4.3.1]decanes as new calystegine analogues

Article abstract of DOI:10.1021/jo900235d

A rapid and stereodivergent access to polyhydroxylated 10- azabicyclo[4.3.1]decanes as new calystegine analogues by way of a double benzotriazolyl/carbon nucleophile exchange followed by a ring-closing metathesis was achieved. Preliminary evaluation of the new compounds as glucocerebrosidase inhibitors was also performed.

Full text of DOI:10.1021/jo900235d

riasis,7 and rare genetic diseases (lysosomal storage disorders11
and cystic fibrosis12).1 The recent approval of Zavesca as the
first oral treatment for Gaucher disease, a rare genetic disease,
is a spectacular demonstration of the importance of iminosugars
as medicines for unmet medical needs.11 In this context, the
development of rapid and general access to original iminosugars
is more than ever highly needed.
Stereodivergent Access to Polyhydroxylated
10-Azabicyclo[4.3.1]decanes as New Calystegine
Analogues
Vincent Chagnault,† Philippe Compain,*,†,¶
Krzysztof Lewinski,‡ Kyoko Ikeda,§ Naoki Asano,§ and
Olivier R. Martin†
In connection with our studies on pharmacological chaperone
therapy for Gaucher disease,13,14 we became interested in
bicyclic structures related to calystegines,15 a new class of
iminosugars discovered in the 1990s. Our first aim was the
preparation of constrained analogues of R-1-C-nonyl-DIX (1),13
a potent and highly selective inhibitor of ꢀ-glucocerebrosidase,
the enzyme involved in Gaucher disease (Figure 1).16 In addition
to that specific objective, our aim was to access rapidly a
diversity of original non-natural calystegine analogues such as
217 in order to evaluate their biological activities (Figure 1).
Azabicyclo[n.3.1]alkanes containing a nitrogen atom in the one-
atom bridge are indeed an important class of alkaloids with
useful biological properties.18
Institut de Chimie Organique et Analytique, UniVersite´
d’Orle´ans-CNRS UMR 6005, BP 6759,
45067 Orle´ans, France, Faculty of Chemistry, Jagiellonian
UniVersity, ul. R. Ingardena 3, 30-060 Krakow, Poland, and
Faculty of Pharmaceutical Sciences, Hokuriku UniVersity,
Kanazawa 920-1181, Japan
philippe.compain@ecpm.u-strasbg.fr
ReceiVed February 4, 2009
(3) (a) Compain, P.; Martin, O. R. Bioorg. Med. Chem. 2001, 9, 3077. (b)
Compain, P.; Martin, O. R. Curr. Top. Med. Chem. 2003, 3, 541. (c) Whalen,
L. J.; Greenberg, W. A.; Mitchell, M. L.; Wong, C.-H. In Iminosugars, From
Synthesis to Therapeutic Applications; Compain, P., Martin, O. R., Eds.; Wiley-
VCH: Weinheim, Germany, 2007; pp 153-176.
(4) (a) Bols, M.; Hazelle, R.; Thomsen, I. B. Chem.sEur. J. 1997, 3, 940.
(b) Heightman, T. D.; Vasella, A.; Tsitsanou, K. E.; Zographos, S. E.; Skamnaki,
V. T.; Oikonomakos, N. G. HelV. Chim. Acta 1998, 81, 853.
(5) Schramm, V. L.; Tyler, P. C. Curr. Top. Med. Chem. 2003, 3, 525.
(6) (a) Lee, R. E.; Smith, M. D.; Nash, R. J.; Griffiths, R. C.; McNeil, M.;
Grewal, R. K.; Yan, W.; Besra, G. S.; Brennan, P. J.; Fleet, G. W. J. Tetrahedron
Lett. 1997, 38, 6733. (b) Lee, R. E.; Smith, M. D.; Pickering, L; Fleet, G. W. J.
Tetrahedron Lett. 1999, 40, 8689.
(7) Moriyama, H.; Tsukida, T.; Inoue, Y.; Yokota, K.; Yoshino, K.; Kondo,
H.; Miura, N.; Nishimura, S.-I. J. Med. Chem. 2004, 47, 1930.
(8) Somsak, L.; Nagy, V.; Hadazy, Z.; Docsa, T.; Gergely, P. Curr. Pharm.
Des. 2003, 9, 1177.
(9) Nishimura, Y. In Iminosugars, From Synthesis to Therapeutic Applica-
tions; Compain, P., Martin, O. R., Eds.; Wiley-VCH: Weinheim, Germany, 2007;
pp 269-294.
(10) (a) Greimel, P.; Spreitz, J.; Stu¨tz, A. E.; Wrodnigg, T. M. Curr. Top.
Med. Chem. 2003, 3, 513. (b) Robina, I.; Moreno-Vargas, A. J.; Carmona,
A. T.; Vogel, P. Curr. Drug. Met. 2004, 5, 329. (c) Norton, P. A.; Baohua,
G.; Block, T. M. In Iminosugars, From Synthesis to Therapeutic Applications;
Compain, P., Martin, O. R., Eds.; Wiley-VCH: Weinheim, Germany, 2007;
pp 209-224.
(11) (a) Butters, T. D.; Dwek, R. A.; Platt, F. M. Chem. ReV. 2000, 100,
4683. (b) Butters, T. D.; Dwek, R. A.; Platt, F. M. Glycobiology 2005, 15, 43R.
(12) Norez, C.; Noel, S.; Wilke, M.; Bijvelds, M.; Jorna, H.; Melin, P.;
DeJonge, H.; Becq, F. FEBS Lett. 2006, 580, 2081.
A rapid and stereodivergent access to polyhydroxylated 10-
azabicyclo[4.3.1]decanes as new calystegine analogues by
way of a double benzotriazolyl/carbon nucleophile exchange
followed by a ring-closing metathesis was achieved. Pre-
liminary evaluation of the new compounds as glucocer-
ebrosidase inhibitors was also performed.
Over the past decade, the pace of discoveries in the field of
iminosugars has been breathtaking.1 Historically known as
glycosidase inhibitors,2 the scope of their biological activity has
been extended to the inhibition of numerous enzymes such as
glycosyltransferases,3 glycogen phosphorylases,4 nucleoside-
processing enzymes,5 UDP-Galp mutase,6 and more recently
metalloproteinases.7 As a consequence, iminosugars are now
lead compounds for the treatment of an impressive variety of
diseases including diabetes,8 cancers,9 viral infections,10 pso-
(13) Compain, P.; Martin, O. R.; Boucheron, C.; Godin, G.; Yu, L.; Ikeda,
K.; Asano, N. ChemBioChem 2006, 7, 1356.
(14) (a) Boucheron, C.; Desvergnes, V.; Compain, P.; Martin, O. R.; Lavi,
A.; Mackeen, M.; Wormald, M. R.; Dwek, R. A.; Butters, T. D. Tetrahedron:
Asymmetry 2005, 16, 1747. (b) Yu, L; Ikeda, K.; Kato, A.; Adachi, I; Godin,
G.; Compain, P.; Martin, O. R.; Asano, N. Bioorg. Med. Chem. 2006, 14, 7736.
(c) Boucheron, C.; Toumieux, S.; Compain, P.; Martin, O. R.; Ikeda, K.; Asano,
N. Carbohydr. Res. 2007, 342, 1960.
(15) For reviews, see: (a) Biastoff, S.; Dra¨ger, B. Alkaloids 2007, 64, 49.
(b) Dra¨ger, B. Nat. Prod. Rep. 2004, 21, 211.
(16) For examples of pharmacological calystegine-based chaperones for
Gaucher disease, see: Chang, H.-H.; Asano, N.; Ishii, S.; Ichikawa, Y.; Fan,
J.-Q. FEBS J. 2006, 273, 4082.
† University of Orle´ans-CNRS.
‡ Jagiellonian University.
§ Hokuriku University.
¶ Current address: Laboratoire de Synthe`se Organique et Mole´cules Bioactives,
Universite´ de Strasbourg-CNRS UMR 7509, ECPM, 25 rue Becquerel, 67087
Strasbourg, France.
(17) For examples of original calystegine analogues and recent synthesis,
see: (a) Groetzl, B.; Handa, S.; Malpass, J. R. Tetrahedron Lett. 2006, 47, 9147.
(b) Garc´ıa-Moreno, M. I.; Ortiz Mellet, C.; Garc´ıa Ferna´ndez, J. M. Tetrahedron
2007, 63, 7879. (c) Shing, T. K. M.; Wong, W. F.; Ikeno, T.; Yamada, T. Org.
Lett. 2007, 9, 207.
(1) Iminosugars: From Synthesis to Therapeutic Applications; Compain, P.,
Martin, O. R., Eds.; Wiley-VCH: Weinheim, Germany, 2007.
(2) Iminosugars as Glycosidase Inhibitors: Nojirimycin and Beyond; Stu¨tz,
A. E., Ed.; Wiley-VCH: New York, 1999.
(18) Lounasmaa, M.; Tamminen, T. The Alkaloids 1993, 44, 1.
10.1021/jo900235d CCC: $40.75
Published on Web 03/19/2009
2009 American Chemical Society
J. Org. Chem. 2009, 74, 3179–3182 3179

SCHEME 1. Access to the 10-Azabicyclo[4.3.1]decane Ring
System
FIGURE 1. Some iminosugars and calystegine derivatives.
Our stereodivergent strategy hinges on the polyhydroxylated
2,6-bis(benzotriazolyl)piperidine 3 as a key substrate for double
benzotriazolyl/carbon nucleophile exchange19 allowing the one-
step introduction of two alkenyl chains at C-2 and C-6. The
next key step is then the formation of the desired bicyclic
framework by way of a ring-closing metathesis (RCM). One
of the main challenges in this straightforward approach was to
find a strategy to separate the three possible stereoisomers
obtained from the nucleophilic displacement reaction (i.e., the
trans isomer, and the two meso compounds, the 2,3-trans-2,6-
cis- and the 2,3-cis-2,6-cis-isomers). Bis(benzotriazolyl)piperi-
dine 3 was synthesized in four steps from inexpensive diacetone-
D-glucose according to the procedure reported by Shankar
(Scheme 1).20 In our hands, compounds 3 could be obtained
on a multigram scale (up to 10 g) in 45% overall yield from
diacetone-D-glucose after optimization.21 Only one purification
is required in this four-step process, 3 being purified by
crystallization in AcOEt. The double benzotriazolyl/carbon
nucleophile exchange was first investigated with alkyl-,22 allyl-,
or vinylmagnesium bromide with little success. With or without
additives, such as ZnBr2 or MgBr2, the expected 2,6-dialkyl
piperidines 4 were obtained in poor and unreproducible yields.
The best results were finally obtained with 5 equiv of allyl zinc
bromide, generated by treatment of allylbromide with activated
Zn dust according to Knochel’s procedure.23 Following these
conditions, 2,6-diallyl piperidines 4 were obtained as a mixture
of stereoisomers in 57% yield after protection of the hydroxyl
groups. Subsequent acetylation of the crude product was
necessary for the next step of the synthesis and for the separation
of the desired products from benzotriazole. The purification of
the two RCM substrate precursors (i.e., the 2,6-cis-isomers 5a
and 5b) and the chemoselective removal of the N-benzyl group24
were performed in the single step (Scheme 1).
endocyclic amino function that could potentially chelate the
RCM catalyst metal center and thus form unproductive com-
plexes. The replacement of the endocyclic amine by a less
coordinating function thus required the deprotection of the amino
group. Even though an increasing number of examples of
sterically crowded amines that are substrates for metathesis are
reported in the literature,25 reaction with Grubbs catalyst (I or
II) using N-benzyl-2,6-diallyl piperidines 4 failed under various
experimental conditions. As a prelude to RCM, the N-benzyl
group in 4 was selectively removed by using 4 equiv of CAN24
in a cosolvent system (THF/H2O) to obtain a mixture of the
three secondary amine stereoisomers 5a,b,c (Scheme 1). Careful
purification by flash chromatography on silica gel afforded the
pure 2,3-trans-2,6-cis-isomer 5a and 2,3-cis-2,6-cis-isomer 5b
in 29 and 13% yield, respectively. The chiral trans-isomer 5c
was obtained in a 1:1 mixture with the 2,3-trans-2,6-cis-isomer
5a in 13% yield (5a:5b:5c ) 23:17:10 determined by 1H NMR
on the crude reaction mixture). The relative configurations of
the substituents in the piperidine rings were unambiguously
established by NMR spectra (COSY and NOESY). These
configurations were further confirmed by X-ray crystallography
of compound 11b at a later stage in the synthesis (Figure 2).
Unfortunately, the same synthetic sequence could not be
reproduced to obtain the divinyl analogues of 5, as precursors
of the 8-azabicyclo[3.2.1] ring system, because the alkylation
step led to low yields and the deprotection step to degradation
products. Having in hand the two diastereomerically pure 2,6-
cis-diallyl piperidines 6a and 6b obtained after protection with
In addition to stereoisomer separation, one of the issues
associated with our strategy was indeed the presence of an
(19) (a) Katritzky, A. R.; Fan, W.-Q. J. Org. Chem. 1990, 55, 3205. (b)
Katritzky, A. R.; Rogovoy, B. V. Chem.sEur. J. 2003, 9, 4586. (c) Katritzky,
A. R.; Lan, X.; Yang, J. Z.; Denisko, O. V. Chem. ReV. 1998, 98, 409.
(20) Shankar, B. B.; Kirkup, M. P.; McCombie, S. W.; Ganguly, A. K.
Tetrahedron Lett. 1993, 34, 7171.
(21) The following modifications were found to increase significantly the
overall yield of the process; no NaHCO3 is needed during the second step, and
the reaction time in steps 2 and 3 has to be increased to 16 h (see Supporting
Information for more details). In our hands, additional crystallization from the
mother liquor, as reported by Shankar, afforded 3 polluted by unreacted
benzotriazole. Compound 3 was found to decompose under purification by flash
chromatography on silica gel.
(22) In our hands, the reaction performed with EtMgBr gave the desired
pure product in only 20% yield.
(23) Knochel, P.; Yeh, M. C. P.; Beck, S. C.; Talbert, J. J. Org. Chem. 1988,
53, 2390.
(24) (a) Bull, S. D.; Davies, S. G.; Fenton, G.; Mulvaney, A. W.; Prasad,
R. S.; Smith, A. D. Chem. Commun. 2000, 337. (b) Cipolla, L.; Palma, A.; La
Ferla, B.; Nicotra, F. J. Chem. Soc., Perkin Trans. 1 2002, 2161.
(25) For a review on olefin metathesis of amine-containing systems, see:
Compain, P. AdV. Synth. Catal. 2007, 349, 1829.
3180 J. Org. Chem. Vol. 74, No. 8, 2009

FIGURE 3. Rationale for observed selectivity in the dihydroxylation
reaction.
pound 14a, the more flexible saturated analogue of 9a, was also
obtained from 8a by way of hydrogenation. Homocalystegines
analogues were then synthesized by dihydroxylation of the
endocyclic double bond of 7 under Upjohn conditions. The
reaction performed with the 2,3-cis-2,6-cis-isomer 7b provided
the exo-diol 11b and the endo-diol 10b in 43 and 15% yield,
respectively, after deprotection of the N-Troc group and
purification on silica gel (de 48%). Structural evidence for
compounds 10b and 11b was obtained unambiguously by NMR
spectra (COSY and NOESY) and X-ray crystallographic analysis
(Figure 2).29 Interestingly, 11b adopts a chair-twist chair
conformation. The twist chair conformation of the azepane
moiety is favored by a hydrogen bond between one OH and
the nitrogen atom and is known to be generally preferred in
seven-membered rings.30
FIGURE 2. Perspective ORTEP view of compounds 11b.29
SCHEME 2. Access to Polyhydroxylated
10-Azabicyclo[4.3.1]decane Derivatives
The exo-attack observed is in agreement with dihydroxylation
reactions performed on related bicyclic systems such as
8-azabicyclo[3.2.1]octenes31 and 9-azabicyclo[4.2.1]nonenes.17a
We assume that the 10-azabicyclo[4.3.1]decene ring system of
7b adopts a chair-chair conformation in which all the acetate
1
groups are in an equatorial position as indicated by H NMR
and suggested by the X-ray crystallographic analysis of 11b
(Figure 2). In such a system, the exo-attack is much more
favored than the endo one because of steric hindrance (Figure
3). In sharp contrast, dihydroxylation of 7a, the 2,3-trans-2,6-
cis-isomer of 7b, afforded the endo-diol 12a as the major
diastereoisomer after deprotection of the N-Troc group (de 34%).
The reversal of facial selectivity may be explained by a
conformational change of the azabicyclo[4.3.1]decene ring
system from a chair-chair to a chair-boat conformation. The
piperidine ring moiety adopts a boat conformation in which all
the acetoxy groups are in a pseudoequatorial position, as
1
indicated by H NMR (Figure 3). The endo face of the double
a Troc group, we submitted them to 10 mol % of Grubbs I in
refluxing dichloromethane.26 We were pleased to find that the
RCM reaction performed on each of the two cis stereoisomers
afforded the expected 10-azabicyclo[4.3.1]decane derivatives
7 in almost quantitative yields (Scheme 1).
A diversity of calystegine analogues was then obtained from
advanced precursors 7. The N-Troc protecting group of stere-
oisomers 7 was selectively removed using Zn in AcOH/AcOEt
to yield the secondary endocyclic amines 8 (Scheme 2).27
Subsequent deprotection of the acetyl groups using allylmag-
nesium bromide28 provided the fully deprotected 10-
azabicyclo[4.3.1]dec-3-ene derivatives 9 in high yields. Com-
bond is then much more accessible than in a chair-chair
conformation. The greater flexibility of the 10-
azabicyclo[4.3.1]decene ring system compared to that of the
8-azabicyclo[3.2.1]octene31 and 9-azabicyclo[4.2.1]nonene17a
ring systems may explain the modest diastereoselectivity
observed.
Preliminary biological evaluation of azabicyclo derivatives
as inhibitors of ꢀ-glucocerebrosidase gave very promising results
(Table 1). The best inhibition was observed for the pentahydroxy
compound 15a with an IC50 value in the micromolar range
similar to the one observed for a reference compound, 1,5-
dideoxy-1,5-iminoxylitol.13 Comparison of the IC50 values
between 9a, 9b, and 14a indicated that best inhibitions were
obtained for 2,3-trans-2,6-cis-isomers with an endocyclic double
bond. In addition, compound 15a was found to be a quite potent
inhibitor of almond ꢀ-glucosidase.2
(26) Neipp, C. E.; Martin, S. F. Tetrahedron Lett. 2002, 43, 1779.
(27) Endo, A.; Yanagisawa, A.; Abe, M.; Tohma, S.; Kan, T.; Fukuyama,
T. J. Am. Chem. Soc. 2002, 124, 6552.
(28) The use of allylmagnesium bromide instead of sodium methoxide was
found to facilitate the workup and the purification step.
(29) Crystallographic data for structure 11b have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publication No. CCDC
713956.
(30) Hendrickson, J. B. J. Am. Chem. Soc. 1961, 83, 4537.
(31) Riche´, F.; Masri, F.; Lopez, M. Tetrahedron Lett. 2007, 48, 1609.
J. Org. Chem. Vol. 74, No. 8, 2009 3181

TABLE 1. IC50 Values for 9, 14a, and 15a toward ꢀ-Glucosidase
pH was reached. The aqueous phase was extracted using CH2Cl2
(3×), and the organic phases were combined, dried, and concen-
trated under vacuum. The desired products were obtained by
purification on silica gel chromatography using EtOAc/PE (20/80)
to afford compounds 4 (1.611 g, 3.75 mmol, 57% yield).
Almond
DIXa 13
9a
9b
14a
15a
,
ꢀ-glucosidases almonds
ꢀ-glucocerebrosidase
180
2.3
240
77
NIb
NI
NI
750
2.0
1.2
N-Deprotection. To compounds 4 (1.446 g, 3.37 mmol),
dissolved in a 5:1 mixture of THF (190 mL) and water (37 mL),
was added CAN (7.38 g, 13.47 mmol) in portions. When the
reaction was complete (5 h), the mixture was treated with saturated
aqueous NaHCO3 until a basic pH was reached and extracted with
EtOAc. The organic phase was dried (MgSO4), filtered, and
concentrated. The different products were isolated by flash chro-
matography using EtOAc/PE (20/80): 5a (29%, 331 mg, Rf ) 0.25),
5b (13%, 149 mg, Rf ) 0.05). 5a: IR (cm-1) (NaCl, FTIR) 1749.0,
a 1,5-Dideoxy-1,5-iminoxylitol. b Less than 50% inhibition at 1 mM.
In conclusion, we have reported a rapid and stereodivergent
access to polyhydroxylated azabicyclo[4.3.1]decenes by way of
a double benzotriazolyl/carbon nucleophile exchange followed
by a RCM. Preliminary investigations on the activity of these
original non-natural calystegine analogues as glucocerebrosidase
inhibitors indicated that these compounds constitute a new,
promising class of iminosugars of therapeutic interest.
1
1247.6, 1225.7, 1030.0; H NMR (CDCl3, 400 MHz) δ 5.71 (tdd,
J ) 6.0, 8.4, 14.5 Hz, 2H, H-8, H-11), 5.15-5.08 (m, 4H, H-9,
H-12), 5.04 (t, J ) 9.4 Hz, 1H, H-4), 4.78 (t, J ) 9.6 Hz, 2H, H-3,
H-5), 2.71 (td, J ) 3.4, 9.2 Hz, 2H, H-2, H-6), 2.32 (dd, J ) 5.2,
10.0 Hz, 2H, H-10, H-7), 2.07-2.00 (m, 2H, H-10, H-7), 2.02 (s,
6H, CH3CO), 1.99 (s, 3H, CH3CO), 1.68 (s, NH); 13C NMR (CDCl3,
101 MHz) δ 170.7 (CO), 170.1 (CO), 134.0 (C-11, C-8), 119.0
(C-12, C-9), 75.7 (C-4), 74.0 (C-3, C-5), 56.5 (C-2, C-6), 36.3 (C-
10, C-7), 21.0 (2 × CH3CO), 20.9 (CH3CO); HRMS [M + H]+
340.1753 (calcd for C17H26NO6 340.1760).
Experimental Section
Preparation of Allyl Derivatives 4. Allyl bromide (2.8 mL, 32.8
mmol) in anhydrous THF (30 mL) was added slowly to a THF
suspension of activated Zn* (cf. Supporting Information), and the
reaction was stirred for 15-30 min. Compounds 3 (mixture of regio-
and stereoisomers, 3 g, 6.56 mmol) in anhydrous THF (450 mL)
were then added slowly at room temperature, and the reaction
mixture was stirred overnight (20 h). Water was then added, and
the solids were removed by filtration over Celite. Saturated aqueous
Na2CO3 (100 mL) was added, and filtration over Celite was carried
out. The filtrate was extracted with AcOEt (3×), and the organic
phases were dried and concentrated under vacuum. DMAP (0.401
g, 3.28 mmol) was added to a solution of this crude mixture (1.99
g, 6.56 mmol) in CH2Cl2 (100 mL). Then acetic anhydride (6.19
mL, 65.6 mmol) was added dropwise, and the reaction mixture was
stirred overnight at room temperature. Water (20 mL) was added
while keeping stirring for 30 min. Na2CO3 was added until a basic
Acknowledgment. Financial support of this study by grants
from ANR, VLM, and the French department of research is
gratefully acknowledged.
Supporting Information Available: Additional procedures
and characterization data for new compounds are included. This
material is available free of charge via the Internet at
http://pubs.acs.org.
JO900235D
3182 J. Org. Chem. Vol. 74, No. 8, 2009