Approaches to the synthesis of CF2-analogues of 2-deoxy-2-aminoglycosides

Article abstract of DOI:10.1016/j.tetlet.2009.01.155

The preparation of a fluorinated C-glycosidic analogue of a 2-deoxy-2-acetamido-d-altrose is described. The synthetic sequence involves the addition of a difluoroenoxysilane to a d-glucal, an epoxidation of the resulting unsaturated CF₂-glycoside and a ring-opening reaction with TMSN₃. An Overman rearrangement of the unsaturated intermediate is also described.

Full text of DOI:10.1016/j.tetlet.2009.01.155

Tetrahedron Letters 50 (2009) 1803–1805
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Approaches to the synthesis of CF2-analogues of 2-deoxy-2-aminoglycosides
*
*
Florent Poulain, Eric Leclerc , Jean-Charles Quirion
Université et INSA de Rouen, CNRS, UMR 6014 C.O.B.R.A.–IRCOF, 1 rue Tesnière, 76821 Mont Saint-Aignan Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
The preparation of a fluorinated C-glycosidic analogue of a 2-deoxy-2-acetamido-D-altrose is described.
Received 5 January 2009
Revised 27 January 2009
Accepted 30 January 2009
Available online 5 February 2009
The synthetic sequence involves the addition of a difluoroenoxysilane to a D-glucal, an epoxidation of
the resulting unsaturated CF2-glycoside and a ring-opening reaction with TMSN3. An Overman rearrange-
ment of the unsaturated intermediate is also described.
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Aminoglycosides
Glycomimetics
Fluorine
Epoxidation
A high number of natural glycoconjugates with significant bio-
logical functions feature a 2-deoxy-2-aminoglycoside moiety in
their glycone group.1,2 More generally, the majority of natural
cation of this sensitive material.6c The double bond displayed in the
addition products -4 and b-4, which was first dihydroxylated for
a
the preparation of CF2-glycosides, could also serve for the introduc-
tion of an amino group (Scheme 1). Two sequences were planned
for such a purpose: an epoxidation of the double bond followed
by a ring-opening reaction with nitrogen nucleophiles, or a sigma-
tropic [3,3]-rearrangement of the corresponding trichloroacetimi-
date followed by a dihydroxylation reaction. Our investigation of
these two pathways is described as follows.
O-glycopeptides display a a-GalNAc/Ser or a-GalNAc/Thr core frag-
ment.2 The glycone and aglycone parts of all the tumour-associated
antigens, which are used in the elaboration of anti-cancer vaccines,
are, for example, linked by an
-GalNAc) residue.1a,b A synthetic hexasaccharide malarial toxin,
used for the development of anti-malarial vaccines, features also
an
-glucosamine residue.3 Glycosylation of proteins with b-2-
deoxy-2-acetamido- -glucose (b-GlcNAc) seems on the other hand
a-2-deoxy-2-acetamido-D-galactose
(a
a
The epoxidation of the addition products followed the same fea-
D
tures as the previously described dihydroxylation reaction.6c The
to be involved in transcriptional regulation.4 Finally, glycosamino-
glycans, which are polysaccharide compounds such as the well-
known heparin, exhibit a variety of biological properties and have
been exploited for the preparation of therapeutic drugs.5 Our group
has been involved for many years in the development of method-
ologies for the synthesis of CF2-glycosides, a subclass of hydrolyti-
cally stable glycomimetics.6,7 As a consequence, these numerous
examples illustrating the importance of the 2-deoxy-2-aminogly-
cosidic link among natural glycopeptides prompted us to apply
some of these methods to the synthesis of their CF2-analogues
(Fig. 1).
a/b mixture (6:4) of acetylated compounds 2 was totally unreac-
tive under all the epoxidation conditions which were tested
(m-CPBA, V(acac)3/t-BuOOH, MnSalen/m-CPBA, Oxone, Oxone/
Trifluoroacetone). On the other hand, deacetylation provided
compounds which underwent the desired reaction using methyl-
(trifluoromethyl)dioxirane (generated in situ from trifluoroacetone
and oxone).8 The
a/b mixture of anomers 3 was indeed successfully
epoxidized to provide 5 (64% yield) as a mixture of two diastereo-
mers in the same ratio, thus indicating that the epoxidation was
probably totally diastereoselective for each anomer (Scheme 1).
We recently reported the addition of the silylated difluoroke-
tene acetal 1 to
D
-glucal and its application to the synthesis of
OH
O
HO
HO
O
HO
a-CF2-mannosides and pseudo-glycopeptides. An improved proce-
OH
O
-
OSO3
O
-O2C
dure indeed allowed us to prepare a salt-free solution of 1 which
was directly used in glycosylation reactions without further purifi-
HN
O
AcHN
O
HO
-O3SHN
O
O
α-GalNAc/Ser (TN antigen)
Major disaccharide sequence of heparin
* Corresponding authors. Tel.: +33 235522901; fax: +33 235522959 (E.L.).
E-mail addresses: eric.leclerc@insa-rouen.fr (E. Leclerc), quirion@insa-rouen.fr
(J.-C. Quirion).
Figure 1.
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.01.155

1804
F. Poulain et al. / Tetrahedron Letters 50 (2009) 1803–1805
OAc
O
OP
OTBDPS
O
OTBDPS
AcO
AcO
TBDPSCl
F
F
F
F
OSiMe3
OEt
O
PO
O
F
NEt3, DMAP
HO
HO
+
F
CO2Et
BF3.Et2O
65%
CO2Et
F
CH2Cl2, -40°C
1
CO2Et
F
72%, α:β = 6:4
2 (P = Ac)
3 (P =H)
HCl, EtOH
91%
α-4
β-4
Oxone, CF3C(O)CH3
NaHCO3, MeCN/H2O
Oxone, CF3C(O)CH3
NaHCO3, MeCN/H2O
77%
86%
64%
OTBDPS
OH
TBDPSO
HO
F
F
O
O
O
O
HO
HO
F
CO2Et
F
O
O
F
CO2Et
CO2Et
F
5
7
6
Scheme 1.
This was confirmed when submitting each anomer
a
-4 or b-4,
the best of our knowledge, the first reported example of a 2-
deoxy-2-acetamido-CF2- -glycoside (Scheme 2).
obtained and separated after selective silylation of the C-6 alcohol,
to the same epoxidation conditions. Epoxides 6 and 7 were indeed
isolated as a single diastereomer for each reaction in 77% and 86%
yield, respectively (Scheme 1).9 The lack of reactivity of the acety-
lated product 2 in the dihydroxylation and epoxidation reactions
was attributed to the combined electron-withdrawing effects of
the difluoroacetate and acetoxy groups.
The relative configurations of epoxides 6 and 7 were not deter-
mined at that stage due to the overlaps of crucial signals in the 1H
NMR spectra, which prevented us from collecting useful NOESY
data. Ring opening reactions using nitrogen nucleophiles were
then investigated, beginning with the b derivative 7. Benzylamine
and azides (TMSN3, NaN3) were tested as nucleophiles using Brön-
sted or Lewis acids (NH4Cl, BF3ÁEt2O, LiClO4, Yb(OTf)3) as additives.
The best result was obtained using trimethylsilyl azide in the pres-
ence of a stoichiometric amount of BF3ÁEt2O at low temperature.
These conditions yielded the 2-deoxy-2-azido-CF2-glycoside 8 in
52% yield as a single regio- and stereoisomer (Scheme 2).10 On
D
The different NMR data which were collected unambiguously
proved the b-altrose configuration of compound 8 as well as the
regioselectivity of the ring-opening reaction. The b-configuration
was confirmed by a NOESY correlation between H-1 and H-5,
and the different coupling constants were in favour of this relative
configuration.11 The regioselectivity of the epoxide opening was
proved by a scalar coupling between H-3 and an OH group, as indi-
cated by a COSY correlation between these two protons. The rela-
tive configuration of epoxide 7, as displayed in Scheme 1, has been
deduced from this study. The regio- and stereoselectivity of the
reaction leading to 8 is consistent with a trans diaxial opening of
epoxide 7. If the electron-withdrawing effect of the difluorinated
substituent should destabilize the apparition of a positive charge
on C-2, this well-known ring-opening pathway for cyclohexene
oxides seems nevertheless favoured.12 The stereoselectivity of
the reaction leading to 7 is consistent with an epoxidation anti
to the difluoroacetate group, the largest substituent
a to the dou-
the other hand, the
a
compound 6 was totally unreactive under
ble bond. A hydrogen bonding of the dimethyldioxirane reagent
with a free hydroxy group has been invoked in the literature to ex-
plain a syn epoxidation of a cyclohex-2-en-1-ol.13 Although, in our
case, this model would of course lead to the same product 7, the H-
bonding between the OH group and methyl(trifluoromethyl)dioxi-
rane in a protic medium is not likely. As epoxide 6 remained
poorly reactive towards the addition of nitrogen nucleophiles,
the relative configuration displayed in Scheme 1 could not be de-
duced from a ring-opening product and has only been postulated.
An epoxidation anti to the CF2CO2Et and OH groups appeared as
these conditions and all our efforts devoted to the ring-opening
of this epoxide were unsuccessful, whatever the nucleophile
(TMSN3, NaN3, BnNH2) and the acid (BF3ÁEt2O, TMSOTf, LiClO4,
AlCl3, Yb(OTf)3, etc.) involved in the process. All these conditions
either left the starting material unchanged or led to complex mix-
tures from which a ring-opening compound was isolated only
once, although in very low yield, using BnNH2 as the nucleophile
in the presence of Yb(OTf)3. The azido group of compound 8 was
hydrogenated afterward and acetylated to yield 9, which is, to
OTBDPS
O
OTBDPS
O
OTBDPS
N3
TMSN3
BF3.Et2O
OTBDPS
O
HO
Cl3CCN, DBU
86%
F
F
F
F
CO2Et
O
O
CH2Cl2, -30°C
HO
HO
Cl3C
N
H
F
F
CO2Et
CO2Et
CO2Et
F
F
52%
OH
10
O
α-4
8
7
1- H2, Pd(OH)2/C
2- Ac2O
1,2-Dichlorobenzene, 180°C
No efficient epoxide-opening
46%
reaction with α compound 6.
OTBDPS
O
O
OTBDPS
O
OTBDPS
NHAc
CO2Et
N
H
CCl3
F
F
CO2Et
O
F
AcO
F
32%
O
NH
F
F
CO2Et
OAc
2-deoxy-2-acetamido-CF2-D-altropyranoside
CCl3
11
9
Scheme 2.
Scheme 3.

F. Poulain et al. / Tetrahedron Letters 50 (2009) 1803–1805
1805
Germany, 2003; pp 381–406; (b) Danishefsky, S. J.; Allen, J. R. Angew. Chem., Int.
Ed. 2000, 39, 836–863.
2. Seitz, O. ChemBioChem. 2000, 1, 214–246.
3. Schofield, L.; Hewitt, M. C.; Evans, K.; Siomos, M. A.; Seeberger, P. H. Nature
2002, 418, 785–789.
4. Zachara, N. E.; Hart, G. W. Chem. Rev. 2002, 102, 431–438.
5. Islam, T.; Linhardt, R. J. Chemistry, Biochemistry, and Pharmaceutical Potentials
of Glycosaminoglycans and Related Saccharides. In Carbohydrate-Based Drug
Discovery; Wong, C.-H., Ed.; Wiley-VCH: Weinheim, Germany, 2003; Vol. 1, pp
407–439.
the most reasonable pathway since, as explained above, a steric
discrimination is more likely under these conditions than the
H-bonding approach.
A second strategy was developed which is based on the Over-
man sigmatropic [3,3]-rearrangement of a trichloroacetimidate de-
rived from 4. This reaction has been quite developed for the
synthesis of classical 2-deoxy-2-amino-D
-glycosides.14 Trichloro-
acetimidate 10 was thus prepared using standard conditions, and
several attempts were made to promote the sigmatropic rear-
rangement (Scheme 3). It was already known from previous
reports that compounds with the glucose configuration required
higher temperatures than their galactose counterparts for the rear-
rangement to occur. For stereoelectronic reasons, a pseudo-axial
arrangement of the trichloroacetimidate moiety in the transition
state is indeed necessary for this reaction. In the glucose series, this
can be achieved only by reaching a pseudo-chair conformation of
higher-energy (Scheme 3). High temperatures are thus required
and the best results were obtained in refluxing 1,2-dichloroben-
zene with a catalytic amount of potassium carbonate, yielding
trichloroacetamide 11 in 32% yield (Scheme 3).15
6. (a) Karche, N. P.; Pierry, C.; Poulain, F.; Oulyadi, H.; Leclerc, E.; Pannecoucke, X.;
Quirion, J.-C. Synlett 2007, 123–126; (b) Moreno, B.; Quehen, C.; Rose-Hélène,
M.; Leclerc, E.; Quirion, J.-C. Org. Lett. 2007, 9, 2477–2480; (c) Poulain, F.; Serre,
A.-L.; Lalot, J.; Leclerc, E.; Quirion, J.-C. J. Org. Chem. 2008, 73, 2435–2438.
7. For other CF2-glycoside syntheses, see: (a) Houlton, J. S.; Motherwell, W. B.;
Ross, B. C.; Tozer, M. J.; Williams, D. J.; Slawin, A. M. Z. Tetrahedron 1993, 49,
8087–8106; (b) Herpin, T. F.; Motherwell, W. B.; Weibel, J.-M. Chem. Commun.
1997, 923–924; (c) Brigaud, T.; Lefebvre, O.; Plantier-Royon, R.; Portella, C.
Tetrahedron Lett. 1996, 37, 6115–6116; (d) Berber, H.; Brigaud, T.; Lefebvre, O.;
Plantier-Royon, R.; Portella, C. Chem. Eur. J. 2001, 7, 903–909; (e) Wegert, A.;
Miethchen, R.; Hein, M.; Reinke, H. Synthesis 2005, 1850–1858; (f) Wegert, A.;
Hein, M.; Reinke, H.; Hoffmann, N.; Miethchen, R. Carbohydr. Res. 2006, 341,
2641–2652; (g) Hirai, G.; Watanabe, T.; Yamaguchi, K.; Miyagi, T.; Sodeoka, M.
J. Am. Chem. Soc. 2007, 129, 15420–15421; (h) Tony, K. A.; Denton, R. W.; Dilhas,
A.; Jiménez-Barbero, J.; Mootoo, D. R. Org. Lett. 2007, 9, 1441–1444; (i) Denton,
R. W.; Tony, K. A.; Hernandez-Gay, J. J.; Canada, F. J.; Jimenez-Barbero, J.;
Mootoo, D. R. Carbohyd. Res. 2007, 342, 1624–1635.
The next step was thus to functionalize the remaining double
bond in order to access to the desired 2-amino-CF2-glycoside.
The dihydroxylation of related compounds was well documented
and was expected to occur anti to the trichloroacetamide group
8. (a) Mello, R.; Fiorentino, M.; Sciacovelli, 0.; Curci, R. J. Org. Chem. 1988, 53, 3890–
3891; (b) Adam, W.; Curci, R.; Edwards, J. O. Acc. Chem. Res. 1989, 22, 205–211; (c)
Yang, D.; Wong, M.-K.; Yip, Y.-C. J. Org. Chem. 1995, 60, 3887–3889.
9. To a solution of b-4 (245 mg, 0.50 mmol) in acetonitrile (8 mL) cooled at 0 °C
were added a solution of EDTA (0.0004 M in water, 4 mL) and trifluoroacetone
to deliver preferentially the 2-deoxy-2-amino-D-galactose ana-
(3.5 mmol).
A mixture of NaHCO3 (0.8 g, 9.5 mmol) and oxone (1.8 g,
logue.14b,c To our disappointment, compound 11 remained totally
unreactive under all the conditions which were tested (cat. OsO4/
NMO or stoichiometric OsO4 in different solvent systems, at room
temperature or at 60 °C). This lack of reactivity of the double bond
was attributed to the combined strong electron-withdrawing ef-
fects of the trichloroacetamide group and, to a lesser extent, of
the difluoroacetate group. Unfortunately, our attempts to depro-
tect the trichloroacetamide moiety affected also the highly reactive
difluoroacetate and raised many compatibility problems. The
epoxidation reaction of 11 was not examined but the synthesis
of the desired 2-deoxy-2-amino-CF2-glycoside would anyway in-
volve a difficult opening of the resulting epoxide with oxygen
nucleophiles. Combined with the low yield obtained for the rear-
rangement, this approach was thus too compromised to be further
investigated.
2.93 mmol) was added at 0 °C in small portions over 4 h. The slurry was
vigorously stirred at 0 °C until complete consumption of the starting material,
and then concentrated under reduced pressure to remove acetonitrile. The
concentrate was dissolved in a minimal amount of water and AcOEt (10 mL).
The aqueous layer was extracted with AcOEt (2 Â 10 mL). The organic layers
were combined and washed with brine (15 mL), dried over magnesium sulfate
and concentrated under reduced pressure. Purification by column
chromatography over silica gel (cyclohexane/AcOEt 85:15) afforded 7 as a
colourless oil (220 mg, 86% yield): Rf = 0.30 (cyclohexane/AcOEt 8:2); [a]
D
+1.59 (c 0.39, CHCl3); 1H NMR (300 MHz; CDCl3) d 7.68–7.62 (m, 4H), 7.49–7.36
(m, 6H), 4.38–4.20 (m, 3H), 4.15–4.08 (m, 1H), 3.87–3.77 (m, 2H), 3.76 (d,
J = 4.1 Hz, 1H), 3.59 (dd, J = 4.1 Hz, J = 1.3 Hz, 1H), 3.50–3.45 (m, 1H), 2.58 (d,
J = 6.4 Hz, 1H), 1.26 (t, J = 7.2 Hz, 3H), 1.04 (s, 9H); 19F NMR (282.5 MHz; CDCl3)
d À113.3 (dd, J = 271.5 Hz, J = 7.5 Hz, 1F), À120.3 (dd, J = 271.5 Hz, J = 14.0 Hz,
1F); 13C NMR (75.5 MHz; CDCl3) d 135.9, 130.3, 128.2, 112.9 (t, J = 254.7 Hz),
74.2, 73.8 (t, J = 28.8 Hz), 66.8, 64.3, 60.9, 54.5 (2C), 27.0, 14.5; IR (neat) mmax
3436.1, 2930.8, 2857, 1762.2 cmÀ1; MS (ESI+) m/z = 529 ([M+Na]+); Anal. Calcd
for C26H32F2O6Si: C, 61.64; H, 6.37. Found: C, 61.59; H, 6.43.
10. To a solution of 7 (130 mg, 0.26 mmol) in dichloromethane (3 mL) cooled at
À30 °C were added TMSN3 (0.8 mL, 6 mmol) and BF3ÁEt2O (36
lL, 0.28 mmol).
In conclusion, we applied our recently developed difluoroenox-
ysilane addition reaction to the synthesis of 2-deoxy-2-aminogly-
The solution was allowed to warm up to 10 °C over 4 h. After complete
conversion of the starting material (TLC monitoring), a saturated aqueous
solution of NaHCO3 (5 mL) was added and the mixture was extracted with
AcOEt (3 Â 5 mL). The combined organic extracts were washed with brine
(8 mL), dried over magnesium sulfate and concentrated under reduced
pressure. Purification by column chromatography over silica gel
coside analogues.
A CF2-analogue of 2-deoxy-2-acetamido-D-
altrose, first example of a CF2-aminopyranoside, has been synthe-
sized through an epoxidation/azidation sequence performed on
the addition product b-4. Another approach, based on the Overman
(cyclohexane/ethyl acetate 80:20, 1% triethylamine) afforded
8 as a
colourless oil (73 mg, 0.135 mmol, 52% yield): Rf = 0.20 (cyclohexane/AcOEt
rearrangement of the trichloroacetimidate derived from
a-4, was
8:2); [
a
]
D À8.25 (c 0.80, CHCl3); 1H NMR (300 MHz; CDCl3) d 7.68–7.63 (m, 4H),
also investigated. The expected product 11 was obtained but
unfortunately could not be converted to the corresponding
CF2-aminopyranoside using a dihydroxylation reaction. However,
these strategies remain reasonable approaches for the synthesis
of the CF2-analogues of these valuable carbohydrate derivatives.
Our efforts will now focus on the preparation of surrogates to
2-deoxy-2-aminoglycosides of biological importance, such as
7.50–7.38 (m, 6H), 4.38 (br s, 1H), 4.28 (dd, J = 2.7 Hz, J = 8.3, 1H), 4.22 (br s,
1H), 4.02–3.95 (m, 2H), 3.88 (s, 1H), 3.82–3.75 (m, 4H), 2.78 (d, J = 1.3 Hz, 1H),
1.12 (t, J = 7.2 Hz, 3H), 1.09 (s, 9H); 19F NMR (282.5 MHz; CDCl3) d À114.52
(ddd, J = 245.7 Hz, J = 7.5 Hz, J = 2.1 Hz, 1F), -130.34 (d, J = 245.7 Hz, 1F); 13C
NMR (75.5 MHz; CDCl3) d 135.7, 132.2, 130.3, 128.1, 120.9 (t, J = 250.1 Hz),
77.6, 72.7 (dd, J = 17.7 Hz, J = 36.6 Hz), 70.7, 69.1, 66.8, 66.3, 60.1, 26.9, 19.2,
15.2; IR (neat) mmax 3448.2, 2931.6, 2858.5, 2130.9, 1762 cmÀ1; MS (IC+) m/
z = 567 ([M+NH4]+); Anal. Calcd for C26H33F2N3O6Si: C, 56.82; H, 6.05; N, 7.64.
Found: C, 56.86; H, 6.05; N, 7.58.
11. H-2 and H-3 appear as broad singlets on the 1H NMR spectrum, indicating the
absence of large diaxial coupling constants and thus that these protons are in
equatorial position (see Ref. 10).
a-GalNAc and b-GlcNAc.
Acknowledgements
12. Rickborn, B. Acid-catalyzed Rearrangements of Epoxides. In Comprehensive
Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon: New York, USA,
1991; Vol. 3, pp 733–775.
13. Murray, R. W.; Singh, M.; Williams, B. L.; Moncrieff, H. M. J. Org. Chem. 1996, 61,
1830–1841.
We thank Perigene Inc. for a PhD grant to F.P. and the Associa-
tion pour la Recherche contre le Cancer for financial support.
14. (a) Dyong, I.; Weigand, J.; Merten, H. Tetrahedron Lett. 1981, 22, 2965–2968; (b)
Takeda, K.; Kaji, E.; Konda, Y.; Sato, N.; Nakamura, H.; Miya, N.; Morizane, A.;
Yanagisawa, Y.; Akiyama, A.; Zen, S.; Harigaya, Y. Tetrahedron Lett. 1992, 33,
7145–7148; (c) Donohoe, T. J.; Blades, K.; Helliwell, M. Chem. Commun. 1999,
1733–1734.
References and notes
1. (a) Keding, S. J.; Danishefsky, S. J. Synthetic Carbohydrate Vaccines. In
Carbohydrate-Based Drug Discovery; Wong, C.-H., Ed.; Wiley-VCH: Weinheim,
15. Nishikawa, T.; Asai, M.; Ohyabu, N.; Isobe, M. J. Org. Chem. 1998, 63, 188–192.