Desymmetrization of 7-dimethylphenylsilylcycloheptatriene. Towards the synthesis of new aminocycloheptitols

Article abstract of DOI:10.1039/c0ob00521e

Desymmetrization of 7-silylcycloheptatriene through consecutive dihydroxylation and acyl-nitroso cycloaddition of the resulting diene moiety is described. Dihydroxylation occurred anti relative to the resident silicon group in line with previous observati

Full text of DOI:10.1039/c0ob00521e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Desymmetrization of 7-dimethylphenylsilylcycloheptatriene. Towards the
synthesis of new aminocycloheptitols†
Emeline Girard,a Vale´rie Desvergnes,a Ce´line Tarnusb and Yannick Landais*a
Received 30th July 2010, Accepted 7th September 2010
DOI: 10.1039/c0ob00521e
Desymmetrization of 7-silylcycloheptatriene through consecutive dihydroxylation and acyl-nitroso
cycloaddition of the resulting diene moiety is described. Dihydroxylation occurred anti relative to the
resident silicon group in line with previous observations made in the cyclohexadiene series. In contrast,
the subsequent acyl-nitroso cycloaddition occurred with poor regiocontrol but good level of
diastereocontrol syn to the bulky silyl substituent. The resulting cycloadducts were then elaborated
further to provide a straightforward entry toward aminocycloheptitols in ten steps from commercially
available tropylium salts.
Introduction
Aminoheptitols are seven membered-ring aminocyclitols, a class
of significantly relevant natural products in medicinal chemistry.1
Structurally, aminocyclitols are regarded as amino-substituted
carbasugars.2 Among these compounds, 1,3-diaminocyclitols are
units of various aminoglycosides (e.g. streptomycin), which are
particularly well known as antibacterial agents.3 Aminocyclitols
exhibit a wide structural diversity and consequently display
various biological activities. In particular, many of them are
glycosidase inhibitors due to their ability to act as carbohydrate
mimics.4 As glycosidases are involved in various fundamental
biological pathways, glycosidase inhibitors are potent targets in
the context of research of new therapeutic agents against miscel-
laneous affections (including viral infections, cancer, lysosomal
diseases, diabetes, inflammatory processes etc.) extending the
Fig. 1 Aminocyclitols of biological interest.
interest for such structures. Very recently, promising results have
been reported concerning the role of aminocyclitols 1 and 2
as chaperone on glucocerebrosidases responsible for lysosomic
Gaucher disease (Fig. 1).5 C7N-aminocyclitol derivatives, bearing
a characteristic hydroxymethyl side chain6 represent an important
family of aminocarbasugars. For instance, the a-glycosidase
inhibitor voglibose 3 has been shown to lower blood glucose level
and is used in type 2 diabetes treatments.
Fig. 2 Natural aminoheptitol 4 as potential calystegine A5 5 intermediate.
The natural aminoheptitol 4 (Fig. 2) was isolated from the
roots of P. alkekengi var. francheti in 1996 and was assumed to
be a precursor or a degradation product of Calystegine A5,7 a
bicyclic natural iminosugar. Similarly to many aminocarbasugars,
aminoheptitols were quite recently reported to possess glycosidase
inhibitory activities.8
Several natural 5- and 6-membered ring aminocyclitols of ther-
apeutic interest are known and the synthesis of such carbohydrates
analogues has been broadly illustrated.9 Surprisingly, if the synthe-
sis of 7-membered ring analogues of these aminocarbasugars was
initially reported in 1973,10 it has attracted much less attention. To
our knowledge, all the reported syntheses of aminocycloheptitols
use the carbohydrate chiral pool as starting material, with the
introduction of most hydroxyl substituents at the very beginning
of the synthesis.11 Three of them were based on an intramolecular
nitrone-alkene cycloaddition of sugars.11b–d
In the course of our continuing interest in desymmetrization
processes and its application to the synthesis of natural products,12
we have explored the desymmetrization of cyclohexadienylsilanes
and its use in the synthesis of cyclitols and aminopolyols.13 We re-
cently extended these investigations, by devising a novel approach
to carbohydrate analogues relying on the desymmetrization
aUniversite´ de Bordeaux, Institut des Sciences Mole´culaires, UMR-CNRS
5255, 351, Cours de la Libe´ration, F-33405 Talence cedex, France.
E-mail: y.landais@ism.u-bordeaux1.fr; Fax: +33 05 4000 6286; Tel: +33
05 4000 2289
bENSCMu, Universite´ de Haute-Alsace, UMR 7015, 3 rue Alfred Werner,
F-68200 Mulhouse, France
† Electronic supplementary information (ESI) available: Intermediate
compounds 6, 7, 13, 14, 15 and 16 characterization, spectra and sup-
plementary information. CCDC reference numbers 787046–787049. For
ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/c0ob00521e
5628 | Org. Biomol. Chem., 2010, 8, 5628–5634
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
of silyl-cycloheptatrienes, available from commercial tropylium
salts.14
We report here the extension of the above approach to pre-
pare original aminocycloheptitols of type II, using consecutive
dihydroxylation and acyl-nitroso cycloaddition onto the silylcy-
cloheptatriene 6 to introduce OH and NH2 substituents (Fig. 3).
sensitive and tends to degrade even if stored at low temperature.
Concerning the dihydroxylation step, the silicon group was found
to be efficient at differentiating the two diastereotopic faces,
leading to a complete diastereocontrol. As previously observed
in the 6-membered ring series,17 the osmium reagent approached
anti relative to the silyl group mainly for steric reasons. Attempts
were also made to protect the hydroxyl groups as benzyl ethers, but
all the reaction conditions (acidic, basic or neutral using Dudley’s
reagent18), led to the degradation of the substrate, especially to the
loss of the silyl group. Having in hand the bisacetate compound 9,
the other amino and hydroxyl groups were introduced through
a cycloaddition process involving the diene moiety and an
acyl-nitroso reagent, generated in situ from the corresponding
hydroxylamine.19 While the level of diastereocontrol was found
to be high, the observed regioselectivity was low, as three of the
four possible isomers were obtained, which could be separated by
column chromatography (Scheme 2). When acetamide was used as
a protective group, the four possible stereoisomers were obtained
with very low diastereo- and regiocontrol and were found to be
very difficult to separate.
Fig. 3 Aminoheptitol disconnection from tropylium salt.
Results and discussion
We started our study with the gram-scale synthesis of
phenyldimethylsilylcycloheptatriene 6, obtained in satisfying
yield, as a single isomer, through treatment of the tropylium
tetrafluoroborate with the suitable zinc reagent (Scheme 1).15
Dihydroxylation of 6 was then performed following standard
Sharpless conditions,16 affording the diol 7 and some tetrol 8
depending on the amount of potassium ferricyanide reoxidant
used in excess. When the reaction was performed with only 2
equivalents of reoxidant, instead of three equivalents as usually
indicated in standard protocols, tetrol 8 formation was limited.
Scheme 2 Acyl-nitroso cycloaddition onto the silylcycloheptadiene 9.
The relative configuration of each isomer was determined
unambiguously through X-ray diffraction studies of compound
11 and of derivatives for 10 and 12 (vide infra). Interestingly,
the approach of the acyl-nitroso reagent mainly occurred syn to
the silyl substituent, in contrast with what is usually observed
in the 6-silylcyclohexa-1,4-diene series,17b,c,20 where approach of
reagents on both double bonds occurs anti relative to the silicon
group for steric reasons (for instance when osmylation is followed
by cyclopropanation). Structural diversity was then obtained by
elaborating further the two major isomers (Scheme 3). We first per-
formed the C–Si oxidation, under classical Fleming conditions,21
after hydrogenation of the double bond. The resulting alcohols 15
and 16 were then protected as acetate to provide the compounds
17 and 18 in good yields. A three-step sequence without further
purification from compounds 10 and 11 was found to be very
efficient, providing compounds 17 and 18 in respectively 56% and
40% overall yield.
Scheme 1 Synthesis and dihydroxylation of 7-silylcycloheptatriene 6.
The diol intermediate 7 is over-oxidised in the medium, as it
likely reacts faster than the triene substrate for conformational
reasons. It is worth noticing that tetrol 8, not identified, was only
recovered after the protection step as the tetraacetate derivative,
8 being partly soluble in the aqueous layer and difficult to extract
from this medium. The overall yield of the sequence affording the
bis-acetate compound 9 was found to be better when reactions
were performed following a three-step one-pot sequential proce-
dure (34% over three steps versus 13% if purification is performed
at each step). We effectively observed that compound 6 is quite
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5628–5634 | 5629
©

View Article Online
Fig. 4 Acetate migration structural probe.
The inhibitory activity of compounds 22 and 24 against a
range of commercially available glycosidases (a-glucosidase,
b-glucosidase, a-galactosidase, b-galactosidase, a-mannosidase,
b-mannosidase, a-L-fucosidase, b-xylosidase) was then
evaluated.23 Unfortunately, in both cases, no inhibition was
observed at 100 mM.
Scheme 3 Oxidation of the C–Si bond.
Among standard conditions, Mo(CO)6 proved to be the most
reliable reagent to perform the N–O bond reductive cleavage.19b
Compounds 19 and 20 were thus obtained in 80% and 95% yield
respectively, both as isomeric mixtures due to the ability of acetate
groups to migrate onto the neighboring free hydroxyl group.
Such migrations were unambiguously established through X-ray
crystallographic studies of isomers in both series (Fig. 4). Final
deprotection of the acetate protective groups to get aminoheptitols
then proved to be troublesome. After extensive efforts, mild and
efficient conditions were eventually found, using a basic carbonate
resin (DOWEX) to deprotect acetate groups, and an acidic resin
to remove the carbamate moiety.22 The original aminoheptitols 22
and 24 (Scheme 4) were finally obtained cleanly in moderate, but
non optimized yields, eluting the acidic resin with a 15% ammonia
solution. It is worth noticing that the use of standard conditions,
i.e. a large excess of TFA, was not efficient in our case, leading
to incomplete deprotection of the carbamate. Moreover, when the
deprotection was performed on a substrate bearing a free amino
group, acetate migrations occurred onto nitrogen. In our hands,
the resulting amide was then found impossible to cleave.
Conclusions
As a conclusion, we have described along these lines a short and
efficient access to aminocycloheptitols, through desymmetrization
and functionalization of 7-silylcycloheptatrienes. These aminocy-
clitols are obtained in 10 steps and only four purification op-
erations from a commercially available tropylium salt. Complete
diastereocontrol was observed during the dihydroxylation (desym-
metrization). Interestingly, while the former functionalization
occurred anti relative to the silicon substituent, the acyl-nitroso
reagent approached syn to the silicon, in contrast with previous
studies with analogous silyl-cyclohexadiene precursors. This study
thus provides further insight into the reactivity of silylated
polyenes toward electrophilic and dienophile reagents.13,17,20 The
functionalization of these 7-silylcycloheptatrienes has been per-
formed in racemic series but might be rapidly extended into an
enantioselective series following recent reports by Schreiner et al.24
on organocatalytic desymmetrization and enantioselective kinetic
resolution of 1,2-diols.
Scheme 4 Elaboration of original aminocycloheptitols 22 and 24.
5630 | Org. Biomol. Chem., 2010, 8, 5628–5634 This journal is
The Royal Society of Chemistry 2010
©

View Article Online
by column chromatography on silica gel (90 : 10, Petroleum
ether/EtOAc) to provide the acetylated product 9 (1.3 g, 34%)
as a yellow oil. Rf 0.61 (80 : 20 Petroleum ether/EtOAc). FTIR
(film, NaCl): 3070, 3022, 2960, 1737, 1607, 1428, 1372, 1248, 1028
cm-1. 1H NMR (CDCl3, 300 MHz): d (ppm) = 7.59–7.51 (m, 2H,
CHar), 7.39–7.30 (m, 3H, CHar), 5.92–5.82 (m, 1H, CHolefinic), 5.81–
5.72 (m, 1H, CHolefinic), 5.65–5.52 (m, 2H, 1CHolefinic + 1CH-O),
5.45–5.40 (m, 1H, CH–O), 5.36–5.28 (m, 1H, CHolefinic), 2.68 (t, J =
5.3 Hz, 1H, CH–Si), 2.03 (s, 3H, CH3 of acetate), 1.98 (s, 3H, CH3
of acetate), 0.45 (s, 6H, CH3–Si). 13C NMR (CDCl3, 75.5 MHz): d
(ppm) = 170.8 (Cq, C O of acetate), 170.2 (Cq, C O of acetate),
135.9 (Cq, ar), 134.2 (CH, ar), 131.0 (CH, olefinic), 129.6 (CH, ar),
128.0 (CH, ar), 126.4 (CH, olefinic), 125.4 (CH, olefinic), 122.2
(CH, olefinic), 72.9 (CH, CH–O), 70.3 (CH, CH–O), 37.3 (CH,
CH–Si), 21.3 (CH3, acetate), 21.1 (CH3, acetate), -4.2 (CH3, CH3–
Si), -4.4 (CH3, CH3–Si). HRMS (ESI): calc. for C19H24O4SiNa
[M+Na]+ : 367.1342, found: 367.1339.
Experimental part
7-(Dimethyl(phenyl)silyl)cyclohepta-3,5-diene-1,2-diyl diacetate
(9)
A three step sequence from the tropylium tetrafluoroborate can
be carried out without purification. Each step was considered as
complete and a yield of 34% was obtained over three steps.
Hammered lithium wire (1 g, 146 mmol, 13 eq) was added
to THF (50 mL). The mixture was cooled to 0 ◦C and
chlorodimethyl(phenyl)silane (5 mL, 30 mmol, 2.7 eq) was added.
The mixture turned dark red within 30 min. The reaction was
allowed to warm to room temperature overnight and the resulting
lithiated species was titrated using phenolphthalein as color
indicator and a 0.120 M solution of HCl. The concentration
was found to be 0.52 M (27 mmol, 2.4 eq)). In parallel, ZnCl2
(1.8 g, 13.49 mmol, 1.2 eq) was gun-heated under vacuum until
complete melting. After complete cooling the flask was placed
under argon and THF (37.5 mL) was added, sonication helped
the dissolution of ZnCl2. The mixture was cooled to 0 ◦C and the
lithiated species was transferred via a cannula, the reaction mixture
turned immediately green. The mixture was stirred during half an
hour and a suspension of tropylium tetrafluoroborate I (ALFA
AESAR 2 g, 11.2 mmol, 1 eq) in THF (37.5 mL) was prepared.
It was stirred during half an hour more and then the zinc reagent
was transferred into it via a cannula. The resulting solution was
dark green and was left stirring until the coloration turned bright
yellow (from 4 to 14 days).
General procedure for cycloaddition
NaIO4 (4.2 g, 31.3 mmol, 10 eq) was added to a solution of the
diacetate 9 (1 g, 3.13 mmol, 1 eq) in a mixture of methanol–water
(100 mL/39 mL). A solution of hydroxamic acid (6.7 g, 31.3 mmol,
10 eq) in methanol (17 mL) was then added slowly over 6 h using a
syringe pump leading to a very thick orange mixture. The medium
was stirred overnight and then the reaction was buffered using a
saturated aqueous solution of NaHCO3 (40 mL) and quenched
using a saturated aqueous Na2SO3 solution (40 mL). Extraction
was carried out using EtOAc (3 ¥ 40 mL). The combined organic
layers were washed with brine and dried over Na2SO4. Evaporation
of the organic solvents led to a biphasic mixture, insoluble salts
were removed by a quick filtration over silica gel pad using EtOAc
The reaction was quenched using saturated aqueous NH4Cl
solution and extracted with EtOAc (3 ¥ 50 mL). Combined organic
layers were washed with brine and dried over MgSO4. The solvent
was evaporated under reduced pressure to provide an orange oily
product which could be purified by chromatography on silica gel
(petroleum ether). See ref. 14 for description.
1
as eluent. After evaporation, H NMR spectra of the resulting
residue was performed in C2D6CO to split the signals and allowed
the ratio measurement by integration of the relevant signals. The
crude products were then purified by column chromatography
on silica gel (85 : 15 Petroleum ether/EtOAc) and 3 compounds
(10/11/12) were isolated.
“AD-mix like” mixture was prepared by dry mixing K2CO3
(2.9 g, 22.5 mmol, 2 eq), K2OsO4·2H2O (ALDRICH, 166 mg,
0.45 mmol, 0.04 eq), K3Fe(CN)6 (7.4 g, 22.5 mmol, 3 eq),
quinuclidine (50 mg, 0.45 mmol, 0.04 eq) in a round bottomed
flask for 10 min. A mixture of t-BuOH/H2O (1 : 1, 56 mL/56 mL)
and methanesulfonamide (1 g, 11.2 mmol, 1 eq) were added. After
10 min the resulting orange mixture was added in one portion
onto the cycloheptatriene 6 (11.2 mmol, 1 eq). The reaction
mixture quickly thickened and turned brown. After 3 h, no starting
material remained and the reaction was quenched with solid
sodium sulfite (Na2SO3). Extraction was carried out with EtOAc
(3 ¥ 50 mL). The combined organic layers were washed with brine
and dried over Na2SO4. The solvent was evaporated under reduced
pressure (tBuOH was co-evaporated with hexane) to provide a
brown pasty residue mixture of 7 and 8.
The crude product (11.2 mmol, 1 eq) was dissolved CH2Cl2
(112 mL). Pyridine (7.3 mL, 90 mmol, 8 eq), acetic anhydride
(6.5 mL, 90 mmol, 8 eq) and DMAP (catalytic amount) were then
added. No starting material remained in the orange solution after
9 h (TLC control), and quench was performed using saturated
aqueous NH4Cl solution. Extraction was carried out using CH2Cl2
(3 ¥ 50 mL), the combined organic layers were washed with brine
and dried over MgSO4. The solvent was evaporated under reduced
pressure (pyridine and acetic anhydride were co-evaporated with
toluene) to provide a yellow oil. The residue was then purified
7-(tert-Butoxycarbonyl)-4-(dimethyl(phenyl)silyl)-6-oxa-7-
azabicyclo[3.2.2]non-8-ene-2,3-diyl diacetate (10)
Major cycloadduct. The product was obtained as a yellow
powder (43◦1 mg, 29%). Rf 0.29 (80 : 20 Petroleum ether/EtOAc)
mp: 92–93 C. FTIR (film, NaCl): 3068, 2977, 1744, 1427, 1368,
1
1248, 1029 cm-1. H NMR (CDCl3, 250 MHz): d (ppm) = 7.66–
7.53 (m, 2H, CHar), 7.42–7.30 (m, 3H, CHar), 6.44–6.23 (m, 2H,
CHolefinic), 5.52 (t, J = 4.2 Hz, 1H, CH–O), 5.25 (tappearing, J1 = 3.8 Hz,
J2 = 3.6 Hz, 1H, CH–O), 4.84–4.73 (m, 1H, CH–N), 4.70–4.59 (m,
1H, CH–O), 2.02 (s, 3H, CH3 of acetate), 1.86 (s, 3H, CH3 of
acetate), 1.72–1.61 (m, 1H, CH–Si), 1.48 (s, 9H, CH3 of Boc),
0.50 (s, 3H, CH3–Si), 0.47 (s, 3H, CH3–Si). 13C NMR (CDCl3,
75.5 MHz): d (ppm) = 169.9 (Cq, C O of acetate), 169.3 (Cq,
C
O of acetate), 155.5 (Cq, C O of Boc), 136.5 (Cq, ar), 134.2
(CH, ar), 132.2 (CH, olefinic), 129.5 (CH, ar), 128.0 (CH, ar or
olefinic), 127.9 (CH, ar or olefinic), 82.2 (Cq, Boc), 73.6 (CH, CH–
O), 72.7 (CH, CH–O), 69.4 (CH, CH–O), 55.4 (CH, CH–N), 36.5
(CH, CH–Si), 28.3 (CH3, Boc), 21.0 (CH3, acetate), 20.8 (CH3,
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5628–5634 | 5631
©

View Article Online
acetate), -3.2 (CH3, CH3–Si), -3.6 (CH3, CH3–Si). HRMS (ESI):
calc. for C24H33NO7SiNa [M+Na]+ = 498.1924, found: 498.1923.
atmosphere. Palladium was removed by filtration on celite pad,
using EtOAc as eluent. 13 was obtained as a white sticky foam.
KBr (640 mg, 5.38 mmol, 2 eq) and NaOAc (927 mg, 11.3 mmol,
4.2 eq) at 0 ◦C, acetic acid (6.1 mL, 0.44 M) was added to
13 (2.69 mmol, 1 eq). The mixture solidified and peracetic acid
(9.78 mL, 0.275 M, 32 wt% in acetic acid) was added over 10
min, the liquid mixture then bubbled and turned orange. 5 min
after the addition, the ice bath was removed, and the mixture
was stirred overnight. The reaction was quenched using a 25%
7-(tert-Butoxycarbonyl)-2-(dimethyl(phenyl)silyl)-6-oxa-7-
azabicyclo[3.2.2]non-8-ene-3,4-diyl diacetate (11)
Second major cycloadduct. The product was obtained as
a white solid (312 mg, 21%). Rf 0.22 (80 : 20 Petroleum
ether/EtOAc) Anal. calcd. for C24H33NO7Si, C, 60.61; H, 6.99; N,
2.94 found C, 60.35; H, 6.92; N, 2.88. mp: 118–119 ◦C. FTIR (film,
NaCl): 2978, 1742, 1428, 1368, 1248, 1051 cm-1. 1H NMR (CDCl3,
300 MHz): d (ppm) = 7.64–7.48 (m, 2H, CHar), 7.41–7.30 (m, 3H,
CHar), 6.52 (t, J = 7.9 Hz, 1H, CHolefinic), 6.16 (tappearing, J1 = 8.7 Hz,
J2 = 7.0 Hz, 1H, CHolefinic), 5.58 (tappearing, J1 = 4.7, J2 = 4.5 Hz, 1H,
CH–O), 5.20 (tappearing, J1 = 4.3 Hz, J2 = 4.1 Hz, 1H, CH–O), 4.83–
4.67 (m, 2H, CH–O and CH–N), 2.00 (s, 3H, CH3 of acetate), 1.82
(s, 3H, CH3 of acetate), 1.76 (tappearing, J = 4.5 Hz, 1H, CH–Si), 1.38
◦
aqueous solution of Na2S2O3 at 0 C, followed by saturation of
the aqueous phase with Na2S2O3. The two phases were separated
and the aqueous phase was extracted with EtOAc (3 ¥ 20 mL). The
combined organic layers were then washed first with a saturated
aqueous solution of NaHCO3, secondly with brine and were then
dried over sodium sulfate. An orange crude mixture containing 15
was obtained.
To
a solution of this residue (2.69 mmol, 1 eq) in
(s, 9H, CH3 of Boc), 0.49 (s, 3H, CH3–Si), 0.44 (s, 3H, CH3–Si). 13
C
dichloromethane (27 mL, 0.1 M), pyridine (0.9 mL, 10.76 mmol,
4 eq), acetic anhydride (0.8 mL, 10.76 mmol, 4 eq) and DMAP
(catalytic amount) were added. The reaction mixture was stirred
overnight. It was then quenched using NH4Cl saturated aqueous
solution and extracted with CH2Cl2 (2 ¥ 10 mL). The combined
organic fractions were washed successively with H2O and brine and
were dried over Na2SO4. Solvents were evaporated under reduced
pressure (pyridine and acetic anhydride were co-evaporated with
toluene). The crude was purified on silica gel column chromatog-
raphy (60 : 40 Pentane/EtOAc) providing 17 as a vitrified colorless
oil (604 mg, 56%).
NMR (CDCl3, 75.5 MHz): d (ppm) = 169.8 (Cq, C O of acetate),
169.6 (Cq, C O of acetate), 157.2 (Cq, C O of Boc), 136.9 (Cq,
ar), 135.0 (CH, ar), 134.1 (CH, olefinic), 129.4 (CH, ar), 128.0
(CH, olefinic), 125.7 (CH, ar), 82.2 (Cq, Boc), 72.9 (CH, CH–O),
72.8 (CH, CH–O), 69.3 (CH, CH–O), 54.2 (CH, CH–N), 32.1
(CH, CH–Si), 28.2 (CH3, Boc), 21.0 (CH3, acetate), 20.8 (CH3,
acetate), -3.2 (CH3, CH3–Si), -3.6 (CH3, CH3–Si). HRMS (ESI):
calc. for C24H33NO7SiNa [M+Na]+ = 498.1924, found: 498.1927.
7-(tert-Butoxycarbonyl)-2-(dimethyl(phenyl)silyl)-6-oxa-7-
azabicyclo[3.2.2]non-8-ene-3,4-diyl diacetate (12)
FTIR (film, NaCl): 2979, 1751, 1460, 1370, 1228, 1048 cm-1. 1H
NMR (CDCl3, 250 MHz): d (ppm) = 5.57–5.40 (m, 2H, CH–O
or/and CH–N), 4.95 (d, J = 8.1 Hz, 1H, CH–O or CH–N), 4.52
(tappearing, J1 = 5.3 Hz, J2 = 5.8 Hz, 1H, CH–O or CH–N), 4.33
(dappearing, J = 5.2 Hz, 1H, CH–O or CH–N), 2.14–1.86 (m, 13H,
2CH2 + CH3 of 3 acetate), 1.43 (s, 9H, CH3 of Boc). 13C NMR
(CDCl3, 62.9 MHz): d (ppm) = 170.3 (Cq, C O of acetate), 169.3
(Cq, C O of acetate), 154.2 (Cq, C O of Boc), 82.3 (Cq, Boc),
76.8 (CH, CH–O), 76.2 (CH, CH–O), 69.3 (CH, CH–O), 68.3 (CH,
CH–O), 51.5 (CH, CH–N), 28.2 (CH3, Boc), 20.8 (CH3 or CH2),
20.71 (CH3 or CH2), 20.68 (CH3 or CH2), 20.6 (CH3 or CH2),
Minor cycloadduct. The product was obtained as a white
powder (119 mg, 8%). Rf 0.16 (80 : 20 Petroleum ether/EtOAc)
mp: 99–102 ◦C. FTIR (film, NaCl): 2977, 1744, 1427, 1368, 1246,
1056 cm-1. 1H NMR (CDCl3, 250 MHz): d (ppm) = 7.58–7.41 (m,
2H, CHar), 7.41–7.29 (m, 3H, CHar), 6.24 (tappearing, J1 = 8.7 Hz, J2 =
6.5 Hz, 1H, CHolefinic), 6.01 (tappearing, J1 = 7.5 Hz, J2 = 8.5 Hz, 1H,
CHolefinic), 5.22 (tappearing, J1 = 4.8 Hz, J2 = 5.0 Hz, 1H, CH–O), 5.04
(dd, J1 = 4.3 Hz, J2 = 11.9 Hz, 1H, CH–O), 4.90–4.74 (m, 2H,
CH–O and CH–N), 2.13 (s, 3H, CH3 of acetate), 1.71 (s, 3H, CH3
of acetate), 1.41 (s, 9H, CH3 of Boc), 1.26–1.10 (m, 1H, CH–Si),
0.38 (s, 3H, CH3–Si), 0.34 (s, 3H, CH3–Si). 13C NMR (CDCl3,
100.6 MHz): d (ppm) = 170.9 (Cq, C O of acetate), 169.9 (Cq,
16.2 (CH2). HRMS (ESI): calc. for C18H27NO9Na [M+Na]+
=
424.15835 found: 424.1582.
C
O of acetate), 155.0 (Cq, C O of Boc), 136.5 (Cq, ar), 133.8
7-(tert-Butoxycarbonyl)-6-oxa-7-azabicyclo[3.2.2]nonane-2,3,4-
triyl triacetate (18)
(CH, ar), 131.1 (CH, olefinic), 129.7 (CH, ar), 128.2 (CH, ar),
127.6 (CH, olefinic), 81.8 (Cq, Boc), 71.6 (CH, CH–O), 70.4 (CH,
CH–O), 70.0 (CH, CH–O), 53.6 (CH, CH–N), 29.1 (CH, CH–Si),
28.4 (CH3, CH3 of Boc), 21.0 (CH3, acetate), 20.8 (CH3, acetate),
-3.2 (CH3, CH3–Si), -3.5 (CH3, CH3–Si). HRMS (ESI): calc. for
C24H33NO7SiNa [M+Na]+ = 498.1924, found: 498.1928.
Following the same experimental protocol as described above
for 17 synthesis, 11 (581 mg, 1.22 mmol, 1 eq) was used as
starting material providing 18 as a white crystalline solid (196 mg,
40%) after purification on silica gel chromatography (80 : 20
Pentane/EtOAc).
Rf 0.08 (60 : 40, Petroleum ether/EtOAc) Anal. calcd. for
C18H27NO9, C, 53.86; H, 6.78; N, 3.49, found C, 53.81; H, 6.81;
N, 3.30. mp: 133–134 ◦C. FTIR (film, NaCl): 2979, 1746, 1699,
1430, 1370, 1226, 1047, 918 cm-1. 1H NMR (CDCl3, 250 MHz): d
(ppm) = 5.66–5.51 (m, 2H, 2 CH–O), 5.07–4.97 (m, 1H, CH–O),
4.64–4.50 (m, 2H, CH–O and CH–N), 2.27–2.12 (m, 1H, 1H of
CH2), 2.05 (s, 1.5H, CH3 of acetate), 2.04 (s, 1.5H, CH3 of acetate),
2.00 (s, 1.5H, CH3 of acetate), 1.99 (s, 1.5H, CH3 of acetate), 1.96
(s, 1.5H, CH3 of acetate), 1.95 (s, 1.5H, CH3 of acetate), 1.92–1.79
7-(tert-Butoxycarbonyl)-6-oxa-7-azabicyclo[3.2.2]nonane-2,3,4-
triyl triacetate (17)
To a solution of the olefinic compound 10 (1281 mg, 2.69 mmol,
1 eq) in a 2 : 1 mixture of EtOAc–MeOH (0.12 M, 15 mL+7.5 mL),
10% Pd/C (287 mg of the mixture, 0.27 mmol of palladium, 0.1 eq)
was added. Vacuum followed by nitrogen refill was performed 3
times. Then vacuum followed by dihydrogen refill was performed
twice and the mixture was stirred overnight under dihydrogen
5632 | Org. Biomol. Chem., 2010, 8, 5628–5634
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
(m, 3H,CH2), 1.43 (s, 4.5H, CH3 of Boc), 1.42 (s, 4.5H, CH3 of
Boc). 13C NMR (CDCl3, 75.5 MHz): d (ppm) = 170.3 (Cq, C O of
acetate), 169.6 (Cq, C O of acetate), 169.5 (Cq, C O of acetate),
155.1 (Cq, C O of Boc), 82.0 (Cq, Boc), 76.6 (CH, CH–O), 73.4
(CH, CH–O), 70.9 (CH, CH–O), 70.6 (CH, CH–O), 53.7 (CH,
CH–N), 28.2 (CH3, Boc), 20.9 (CH3, acetate), 20.7 (CH3, acetate),
19.9 (CH2), 19.3 (CH2). HRMS (ESI): calc. for C18H27NO9Na
[M+Na]+ = 424.15835 found: 424.1577.
(CH2), 28.3 (CH2). HRMS (ESI): calc. for C7H16NO4 [M+H]+ =
178.10793 found: 178.1076.
4-(tert-Butoxycarbonylamino)-7-hydroxycycloheptane-1,2,3-triyl
triacetate (20) and regioisomers
Following the same experimental protocol as described above for
19 synthesis, 18 (152 mg, 0.38 mmol, 1 eq) was used as starting
material providing 20 as a white solid (152 mg, 99%) purified
by silica gel column chromatography (60 : 40 Pentane/EtOAc).
Spectra of the isomers mixture are available in the ESI.
4-(tert-Butoxycarbonylamino)-7-hydroxycycloheptane-1,2,3-triyl
triacetate and regioisomers (19)
To a solution of hydroxylamine 17 (199 mg, 0.5 mmol, 1 eq)
in a 9 : 1 CH3CN–H2O (0.1 M, 4.5/0.5 mL) mixture, Mo(CO)6
(ALDRICH-98%, 162 mg, 0.6 mmol, 1.3 eq) was added. The white
mixture was refluxed overnight and rapidly turned black when
temperature increased. Reflux was then stopped and silica was
used to quench the reaction. The mixture was then filtered and the
products were eluted with EtOAc. Solvent was evaporated under
reduced pressure. The residue was purified on silica gel column
chromatography (60 : 40 Pentane/EtOAc) affording a mixture of
19 and regioisomers as a white solid (161 mg, 80%). Spectra of the
isomers mixture are available in ESI.†
tert-Butyl 2,3,4,5-tetrahydroxycycloheptylcarbamate (23)
Following the same experimental protocol as described above for
21 synthesis, 20 (98 mg, 0.24 mmol) was used as starting material
providing 23 as a vitrified colorless solid (42 mg, 63%) after
evaporation. No further purification was required.
◦
Rf 0.14 (90 : 10 CH2Cl2–MeOH) mp: 154–155 C. FTIR (film,
NaCl): 3351, 3003, 2920, 1686, 1507, 1452, 1363, 1172, 1087
cm-1.1H NMR (CD3OD, 400 MHz): d (ppm) = 3.99–3.90 (m, 2H,
CH–O), 3.88–3.77 (m, 3H, CH–O and CH–N), 2.11–1.95 (m, 1H,
1H of CH2), 1.95–1.83 (m, 1H, 1H of CH2), 1.83–1.68 (m, 1H,
1H of CH2), 1.45 (s, 10H, 9H of Boc + 1H of CH2)13C NMR
(CD3OD, 100.6 MHz): d (ppm) = 157.5 (Cq, C O of Boc), 80.1
(Cq, Boc), 76.2 (CH, CH–O), 75.7 (CH, CH–O), 73.8 (CH, CH–
O), 72.6 (CH, CH–O), 53.0 (CH, CH–N), 32.0 (CH2), 28.7 (CH3,
CH3 of Boc), 25.2 (CH2). HRMS (ESI): calc. for C12H23NO6Na
[M+Na]+ = 300.14231 found: 300.1420.
tert-Butyl-2,3,4,5-tetrahydroxycycloheptylcarbamate (21)
To a solution of the acetate compounds mixture (19 and regioi-
2-
somers) (482 mg, 1.2 mmol) in methanol, DOWEX CO3 (2
spoons) was added. The mixture was stirred overnight. It was
then filtered and the resin was rinsed thoroughly with methanol.
The clean material 21 was obtained after evaporation as a white
solid (320 mg, 96%). No further purification was performed.
Rf 0.23 (90 : 10 CH2Cl2–MeOH). mp: 149–151 ◦C. FTIR (neat)
: 3334, 1725, 1519, 1455, 1366, 1157 cm-1.1H NMR (CD3OD,
300 MHz): d (ppm) = 4.05–3.91 (m, 3H, 3 CH–O), 3.82–3.78 (m,
1H, CH–O), 3.78–3.65 (m, 1H, CH–N), 1.96–1.79 (m, 2H, CH2),
1.79–1.61 (m, 2H, CH2), 1.47 (s, 9H, CH3 of Boc). 13C NMR
(CD3OD, 75.5 MHz): d (ppm) = 158.0 (Cq, C O of Boc), 80.0
(Cq, Boc), 74.7 (CH, CH–O), 73.9 (CH, CH–O), 73.2 (CH, CH–
O), 71.8 (CH, CH–O), 54.1 (CH, CH–N), 28.8 (CH3, CH3 of Boc),
28.4 (CH2), 27.2 (CH2). HRMS (ESI): calc. for C12H23NO6Na
[M+Na]+ = 300.14231 found: 300.1419
5-Aminocycloheptane-1,2,3,4-tetrol (24)
Following the same experimental protocol as described above for
22 synthesis, 23 (40 mg, 0.14 mmol) was used as starting material
providing 24 (8 mg, 32%) as a colorless oil.
1H NMR (CD3OD, 300 MHz): d (ppm) = 4.01–3.90 (m, 2H, 2
CH–O), 3.86–3.76 (m, 2H, 2 CH–O), 3.13 (dt, J1 = 9.5 Hz, J2 =
3.2 Hz, 1H, CH–N), 2.05–1.68 (m, 3H, CH2), 1.58–1.47 (m, 1H,
1H of CH2). 13C NMR (CD3OD, 75.5 MHz): d (ppm) = 75.6 (CH,
CH–O), 74.8 (CH, CH–O), 73.9 (CH, CH–O), 72.1 (CH, CH–O),
52.9 (CH, CH–N), 30.7 (CH2), 26.3 (CH2). HRMS (ESI): calc. for
C7H16NO4 [M+H]+ = 178.10793 found: 178.1082.
5-Aminocycloheptane-1,2,3,4-tetrol (22)
Enzyme inhibitions
To a solution of carbamate 21 (180 mg, 0.65 mmol) in methanol
(8 mL) Amberlite IRA 120 (1 spoon) was added. The mixture was
refluxed during 2 h, stirred overnight at room temperature and
refluxed again for additional 2 h. The resin was then separated
by filtration and placed in a round bottomed flask. The flask was
cooled to 0 ◦C and a 15% NH3 aqueous solution was added.
After 15 h of stirring, the resin was eliminated by filtration and
the solution was evaporated under reduced pressure providing the
colorless oil (63 mg, 56%).
1H NMR (CD3OD, 300 MHz): d (ppm) = 4.15–3.91 (m, 2H,
2 CH–O), 3.91–3.76 (m, 2H, 2 CH–O), 3.14–2.96 (m, 1H, CH–
N), 2.05–1.77 (m, 2H, CH2), 1.77–1.55 (m, 2H, CH2). 13C NMR
(CD3OD, 75.5 MHz): d (ppm) = 74.9 (CH, CH–O), 74.8 (CH, CH–
O), 74.4 (CH, CH–O), 71.7 (CH, CH–O), 53.2 (CH, CH–N), 28.4
Glycosidase activities were determined at 25 ◦C at the optimal
pH of each enzyme25 with the corresponding p-nitrophenyl
glycopyranoside as substrate against a-D glucosidase (EC 3.2.1.20)
from baker’s yeast (9 units per mg of protein, Km 0.3 mM, pH 7),
b-D-glucosidase (EC3.2.1.21) from almonds (20–40 units per mg
of protein, Km 1.3 mM, pH 4.5), a-D galactosidase (EC 3.2.1.22)
from green coffee beans (Km 0.25 mM, pH 6.5), b-D galactosidase
(EC 3.2.1.23) from Escherichia coli (Km 0.4 mM, pH 7), a-D-
mannosidase (EC 3.2.1.24) from Jack beans (ca. 20 units per mg
of proteins, Km 2 mM, pH 4.5), b-D-mannosidase (EC 3.2.1.25)
from Helix pomatia (Km 0.5 mM, pH 4.5), a-L-fucosidase (EC
3.2.1.51) from bovine kidney (5–15 units per mg, Km 0.3 mM, pH
5.8) and b-D-xylosidase (EC 3.2.1.37) from Aspergilus niger (Km
0.2 mM, pH 4.5).
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5628–5634 | 5633
©

View Article Online
Glycosidases and corresponding p-nitrophenyl glycopyrano-
sides were obtained from Sigma Chemical Co. The release of p-
nitrophenol was measured continuously at 400 nm to determine
initial velocities.26 All kinetics were started by enzyme addition
in a 1 mL assay medium using substrate concentrations around
the Km value of each enzyme. IC50 values were determined for
weak inhibitions (at a substrate concentration equal to Km value)
and correspond to the inhibitor concentration required for 50%
inhibition of the enzyme in our experimental conditions.
9 A. Delgado, Eur. J. Org. Chem., 2008, 3893–3906.
10 I. Dyong and R. Bonn, Chem. Ber., 1973, 106, 944–955.
11 (a) R. Patra, N. C. Bar, A. Roy, B. Achari, N. Ghoshal and S. B. Mandal,
Tetrahedron, 1996, 52, 11265–11272; (b) A. Roy, K. Chakrabarty, P. K.
Dutta, N. C. Bar, N. Basu, B. Achari and S. B. Mandal, J. Org. Chem.,
1999, 64, 2304–2309; (c) C. Curti, F. Zanardi, L. Battistini, A. Sartori,
G. Rassu, L. Auzzas, A. Roggio, L. Pinna and G. Casiraghi, J. Org.
Chem., 2006, 71, 225–230; (d) T. K. M. Shing, W. F. Wong, T. Ikeno
and T. Yamada, Org. Lett., 2007, 9, 207–209.
12 N. Abd Rahman and Y. Landais, Curr. Org. Chem., 2002, 6, 1369–1395.
13 (a) R. Angelaud, Y. Landais and K. Schenk, Tetrahedron Lett., 1997,
38, 1407–1410; (b) Y. Landais, Chimia, 1998, 52, 104–111.
14 R. Beniazza, V. Desvergnes and Y. Landais, Org. Lett., 2008, 10, 4195–
4198.
15 (a) Y. Morizawa, H. Oda, K. Oshima and H. Nozaki, Tetrahedron
Lett., 1984, 25, 1163–1166; (b) M. Oestreich and B. Weiner, Synlett,
2004, 2139–2142; (c) PhMe2SiLi in THF did not react with I.
16 H. C. Kolb, M. S. vanNieuwenhze and K. B. Sharpless, Chem. Rev.,
1994, 94, 2483–2547.
17 (a) R. Angelaud and Y. Landais, J. Org. Chem., 1996, 61, 5202–5203;
(b) R. Angelaud and Y. Landais, Tetrahedron Lett., 1997, 38, 8841–
8844; (c) R. Angelaud, Y. Landais and L. Parra-Rapado, Tetrahedron
Lett., 1997, 38, 8845–8848; (d) R. Angelaud, O. Babot, T. Charvat and
Y. Landais, J. Org. Chem., 1999, 64, 9613–9624.
Acknowledgements
CEFIPRA program (N◦ 3605-1) and CNRS are gratefully
acknowledged for the financial support. EG also thanks the
Ministe`re de l’Enseignement Supe´rieur et de la Recherche for a
PhD fellowship. G. Lautrette is acknowledged for preliminary
work. Authors also thank Dr B. Kauffmann for X-ray structure
determinations.
18 K. W. C. Poon, S. E. House and G. B. Dudley, Synlett, 2005, 3142–3144.
19 (a) J. Soulie´, T. Faitg, J. F. Betzer, B. Muller and J. Y. Lallemand,
Tetrahedron Lett., 1996, 52, 9485–9488; (b) J. Soulie´, T. Faitg,
J. F. Betzer and J. Y. Lallemand, Tetrahedron, 1996, 52, 15137–
15146.
Notes and references
1 D. Tepfer, A. Goldmann, N. Pamboukdjian, M. Maille, A. Lepingle,
D. Chevalier, J. De´narie´ and C. Rosenberg, J. Bacteriol., 1988, 170,
1153–1161.
20 Y. Landais and L. Parra-Rapado, Eur. J. Org. Chem., 2000, 401–418.
21 (a) For reviews on the oxidation of the C–Si bond, see: I. Fleming,
Chemtracts Org. Chem., 1996, 9, 1–64; (b) G. R. Jones and Y. Landais,
Tetrahedron, 1996, 52, 7599–7662; (c) K. Tamao, Adv. Silicon Chem.,
1996, 3, 1.
22 Use of Amberlite IRA 120 for deprotection of Boc, see: T. Rawalpally,
Y. Ji, T. Cleary and B. Edwards, Org. Process Res. Dev., 2009, 13, 478–
482.
23 T. Sifferlen, A. Defoin, J. Streith, D. Le Noue¨n, C. Tarnus, I. Dosbaaˆ
and M-J. Foglietti, Tetrahedron, 2000, 56, 971–978.
24 (a) C. E. Mu¨ller, L. Wanka, K. Jewell and P. R. Schreiner, Angew.
Chem., Int. Ed., 2008, 47, 6180–6183; (b) C. E. Mu¨ller, D. Zell and P.
R. Schreiner, Chem.–Eur. J., 2009, 15, 9647–9650.
2 O. Arjona, A. M. Go´mez, J. C. Lo´pez and J. Plumet, Chem. Rev., 2007,
107, 1919–2036.
3 (a) P. Veyssier and A. Bryskier, Antimicrob. Agents & Chemother., 2005,
453–469; (b) T. Hermann, Cell. Mol. Life Sci., 2007, 64, 1841–1852;
(c) M. Shahid, Anti-Infect. Agents Med. Chem., 2007, 6, 107–117.
4 V. H. Lillelund, H. H. Jensen, X. Liang and M. Bols, Chem. Rev., 2002,
102, 515–553.
5 (a) Z. Yu, A. R. Sawkar and J. W. Kelly, FEBS Lett., 2007, 274, 4944–
4950; (b) G. Sanchez-Olle´, J. Duque, M. Egido-Gabas, J. Casas, M.
Lluch, A. Chabas, D. Grinberg and L. Vilageliu, Blood Cells, Mol. Dis.,
2009, 42, 159–166.
6 T. Mahmud, Nat. Prod. Rep., 2003, 20, 137–166.
7 N. Asano, A. Kato, H. Kizu and K. Matsui, Phytochemistry, 1996, 42,
719–721.
8 O. Andriuzzi, C. Gravier-Pelletier, G. Bertho, T. Prange´ and Y. Le
Merrer, Beilstein J. Org. Chem., 2005, 1, 12–8720.
25 G. Lundblad, U. Bjare, S. Lybing and A. Ma˚hle´n, Int. J. Biochem.,
1983, 15, 835–840.
26 Y.-T. La, J. Biol. Chem., 1967, 242, 5474–5480.
5634 | Org. Biomol. Chem., 2010, 8, 5628–5634
This journal is
The Royal Society of Chemistry 2010
©