Diastereoselective addition of sugar radicals to camphorsultam glyoxilic oxime ether: A route toward C-glycosylthreonine and allothreonine

Article abstract of DOI:10.1039/b910050d

C-glucosyl, mannosyl and galactosyl 2-iodopropane, readily obtained from the corresponding C-glycosyl ketones, were coupled with (+)- or (-)-camphorsultam glyoxylic oxime ether with diastereoselectivity ranging from 70:30 to 80:20. C-glucosyl allothreonin

Full text of DOI:10.1039/b910050d

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Diastereoselective addition of sugar radicals to camphorsultam glyoxilic
oxime ether: a route toward C-glycosylthreonine and allothreonine†
Nicolas Bragnier, Regis Guillot and Marie-Christine Scherrmann*
Received 26th May 2009, Accepted 20th July 2009
First published as an Advance Article on the web 30th July 2009
DOI: 10.1039/b910050d
C-glucosyl, mannosyl and galactosyl 2-iodopropane, readily
obtained from the corresponding C-glycosyl ketones, were
coupled with (+)- or (-)-camphorsultam glyoxylic oxime
ether with diastereoselectivity ranging from 70:30 to 80:20.
C-glucosyl allothreonine was obtained by cleavage of the
camphorsultam moiety.
building blocks for the synthesis of artificial, hydrolytically stable
C-glycosylated proteins.
Although many efforts have been done to prepare C-glycosyl
amino acids, the studies have focussed on cases where the amino
acid moiety contains only one stereogenic center, the most
common being alanine, serine, and asparagine.7 The only example
of a C-glycoside analogue of a b-D-galactosylthreonine has been
reported by the group of Kihlberg.8 The benzylated 2-azido-
3-methyl acid derivative 3, precursor of the corresponding a-
amino acid, was prepared in 14 steps from D-galactonolactone 2
(Scheme 2).9 Therefore, a general method toward the construction
of C-glycosyl amino acids containing two stereocenters onto the
amino acid moiety, i.e. featuring a glycosyl-threonine skeleton,
should be of great interest.
Since the introduction of camphorsultam as a chiral auxiliary by
Oppolzer and coworkers,1 many papers have outlined the potential
of this moiety in asymmetric induction, in particular for radical
additions.2,3 Naito and co-workers have extensively studied the
radical addition to Oppolzer’s camphorsultam derivatives of gly-
oxylic oxime (+)-1 (Fig. 1).3 High yields and diastereocontrol were
obtained in the stannyl radical-mediated reaction of various alkyl
radicals allowing the preparation of enantiomerically enriched a-
amino acids.
Scheme 2
The ketones 4–6 were prepared from unprotected sugars and
pentanedione in water by means of a very convenient method
developed in our group.4 They were treated with acetic anhydride
and pyridine to afford quantitatively the acetylated gluco- 7,10
manno- 811 and galacto-ketone 94c that were allowed to react with
NaBH4 (MeOH, -10 ◦C) to give the alcohols 10–12 in quantitative
yield as diastereoisomeric mixtures (Scheme 3).
Fig. 1
Recently, our group developed a very efficient one-step synthesis
of b-D-C-glycosyl ketones from unprotected sugars.4 We decided
to investigate the use of C-glycosyl iodopropane, readily obtained
from the above ketones, in the carbon radical addition to glyoxilic
oxime ether (+)-13b and (-)-15 in order to provide an access to
C-glycosyl threonine and allothreonine (Scheme 1). The latter
compounds belong to the family of unnatural amino acids in which
the glycinyl moiety is connected directly or through a carbon tether
to the anomeric carbon of the sugar by a carbon–carbon bond.
Since natural O- and N-glycoproteins play key roles in a large
number of biological processes,6 they constitute very interesting
Scheme 1
Universite´ de Paris-Sud 11, Institut de Chimie Mole´culaire et des Mate´riaux
d’Orsay, Baˆtiment 410, 91405, Orsay, France. E-mail: mcscherr@icmo.u-
psud.fr; Fax: +33(0)169154679; Tel: +33(0)169157256
† Electronic supplementary information (ESI) available: Experimental
procedures and analytical data for all new compounds. CCDC reference
numbers 731615–731618. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/b910050d
Scheme 3 Gluco series: 4, 7, 10. Manno series: 5, 8, 11a,b. Galacto series:
6, 9, 12a,b.
The transformation of the alcohols 10–12 into the iodo
derivatives 13–15 was initially accomplished in one step using
3918 | Org. Biomol. Chem., 2009, 7, 3918–3921
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
iodine, triphenylphosphine and imidazole in toluene. In this case,
although the conversion was total (TLC analysis), the isolated
yield was only 80% due to the difficult purification. Thus, this
transformation was carried out via a two-step procedure involving
tosylation (tosyl chloride, pyridine) followed by a nucleophilic sub-
stitution with NaI in acetone affording the iodinated compounds
in quantitative yields. Gluco and galacto derivatives 13a, 13b
and 15a, 15b (Scheme 4) were obtained as 2:1 diastereoisomeric
mixtures whereas the manno isomers 14a and 14b were obtained
in nearly equimolar amount. Pure samples for all diastereomeric
iodide derivatives 13–15 were obtained by preparative HPLC and
fully characterized (see ESI).†
compounds 16 and 18 were obtained as 80:20 diastereoisomeric
mixtures, whereas a 85:15 selectivity was observed with the manno-
configured sugar. The diastereoisomers were separated by HPLC
or crystallisation. NMR studies were unsuccessful in ascertaining
the stereochemistry of the newly formed stereocenters, however,
suitable crystals for X-ray measurements were obtained for the
major gluco isomer 16a ‡ (Fig. 2) and the major manno isomer
17a § (Fig. 3). As expected from previous work employing this
chiral auxiliary,5,12 the S configuration was observed at the amino
acid carbon (C2) whereas the stereocenter deriving from the
sugar moiety (C3) was R. As the selectivity obtained with the
Oppolzer’s camphorsultam derivatives are always very high in
these experimental conditions,3 we assumed that the minor isomers
were (S,S) configured.13
Scheme 4 Gluco series: 13, 16. Manno series: 14, 17. Galacto series: 15,
18.
The Oppolzer’s camphorsultam derivatives of glyoxilic oxime
ether (-)-1 and (+)-1 were readily prepared as described3b and
used in the carbon radical addition under the experimental
conditions described by Naito and coworkers.3 First we exam-
ined the stereochemical outcome of the reaction involving (-)-
camphorsultam as a chiral auxiliary. The reactions between (-)-1
and iodo 13–15 were run in CH2Cl2 at -78 ◦C using Bu3SnH,
Et3B and Zn(OTf)2 as a Lewis acid (Scheme 4). As expected, the
radical addition took place regioselectively at the iminocarbon
to give in all cases only 2 diastereoisomers in 60% isolated yield
for all the sugar derivatives. It appears that the stereochemical
outcome did not depend on the configuration of the starting
iodide since the reaction of (-)-1 with diastereomeric pure 13a
gave the same 4:1 mixture of adducts 16a,b. The selectivity of
the radical coupling was determined by 1H NMR analysis by
integration of the NH signal that always resonated at lower
field for the major diastereoisomer (Table 1). Gluco and galacto
Fig. 2 ORTEP drawing of 16a. Displacement ellipsoids are drawn at the
50% probability level and H atoms are shown as small spheres of arbitrary
radii.
Table 1 Results of the radical coupling with (-)-1
dNH (ppm,
CDCl3)
C2-C3
configurations
Substrate Products Yield
a
b
dr (a/b)
a
b
Fig. 3 ORTEP drawing of 17a. Displacement ellipsoids are drawn at the
50% probability level and H atoms are shown as small spheres of arbitrary
radii.
13a,b
14a,b
15a,b
16a,b
17a,b
18a,b
60% 6.38
60% 6.37
60% 6.40
6.18 80:20
6.15 85:15
6.18 80:20
S-Ra
S-Ra
S-Rb
S-Sb
S-Sb
S-Sb
Despite our efforts, we were not able to obtain from the galacto
compound 18a suitable crystals for the X-ray diffraction studies.
However, by similarity with the cases of the gluco and manno
The major diastereomers were noted as XXa and the minor ones as
XXb.a Assigned by X-ray analysis. b Tentatively assigned configuration.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3918–3921 | 3919
©

View Article Online
Table 2 Results of the radical coupling with (+)-1
We decided to check the viability of this approach for the prepa-
ration of C-glycosyl amino acids by deprotecting 16a. Removal
of the camphorsultam moiety was first tried using conventional
methods.1a Unfortunately, the hydrolysis of 16a with lithium
hydroxide in aqueous THF gave a complex mixture of products.
Attempts to use other bases such as lithium hydroperoxide or
tetrabutylammonium hydrogen peroxide, which proved to be
effective in difficult cases,14 also failed. On the other hand, the
use of sodium methoxide in methanol gave quantitatively 22a
arising from the N-S bond cleavage as already observed during
base-assisted hydrolysis of sterically hindered camphorsultam
derivatives.14,15 Removal of the acetates was thus carried out in
acidic conditions (HCl in MeOH) affording 23a in quantitative
yield. A very mild method, based on formaldehyde catalysis,16 has
been described for the hydrolysis of amides derived from amino-
acids, but when applied to 23a, this method led to the formation
of 24a (Fig. 6). The structure of this unexpected product was
ascertained by the X-ray analysis of its peracetylated methyl ester
derivative 25a ꢀ (Fig. 7) obtained in 75% yield (3 steps).17
dNH (ppm,
CDCl3)
C2-C3
configurations
Substrate Products Yield
a
b
dr (a/b)
a
b
13a,b
14a,b
15a,b
19a,b
20a,b
21a,b
60% 5.96
60% 5.98
60% 5.95
6.24 75:25
6.23 75:25
6.23 67:33
R-Ra
R-Rb
R-Rb
R-Sb
R-Sb
R-Sb
The major diastereomers were noted as XXa and the minor ones as
XXb.a Assigned by X-ray analysis. b Tentatively assigned configuration.
derivatives, the (S,R) configuration was tentatively attributed to
the major diastereoisomer 18a and the (S,S) configuration to 18b.
When the above coupling experiments were repeated using the
(+)-camphorsultam derivative (+)-1, the corresponding products
19–21 were isolated in 60% yield (Fig. 4, Table 2).
Fig. 4 Gluco series: 19. Manno series: 20. Galacto series: 21.
Also in this case the selectivity was determined by integration
1
of the NH signal in their H NMR spectra, the major isomer
showing always a signal at higher field, contrary to what was
observed using the S auxiliary (Table 2). In this series, the X-ray
structure obtained from the gluco major isomer 19a ¶ allowed us
to establish that the configuration at the amino-acid carbon C2
was R as expected by the use of (+)-camphorsultam oxime ether
(+)-1, and the configuration at C3 was R (Fig. 5).
Fig. 6
We thus turn to the cleavage of the N–O bond of 22a
employing Mo(CO)6, and the resulting amine was protected as
NCbz derivative 26a by standard methods.3 This compound was
subjected to basic hydrolysis under optimized reaction conditions,
i.e. tetrabutylammonium hydrogen peroxide in aqueous THF. Al-
though the N–S bond cleavage could not be totally suppressed
(NMR analysis of the crude mixture), we were able to isolate pure
C-glucosyl allothreonine 27a in 40% yield (3 steps).
In order to check that no epimerisation occurred during the
hydrolysis steps, we prepared all four diastereoisomers 27a,b,c,d
by reacting 13a with glyoxilic oxime ether 28 (Scheme 5). The
equimolar mixture of the four diasteroisomers 29a,b,c,d obtained
in 74% yield was then subjected to reductive cleavage of the N–O
bond followed by the N-Cbz protection of the amine (Scheme 5).
The NMR spectra of this mixture showed 4 distinguishable
signals for the NH proton (1H NMR) as well as different signals
for most of the carbon atoms of the diastereoisomers (13C NMR).
No traces of the signals attributed to 27b,c,d were observed in
the spectra of 27a, proving without ambiguity that the removal of
the sultam auxiliary occurred without any loss of stereochemical
purity.
Fig. 5 ORTEP drawing of 19a. Displacement ellipsoids are drawn at the
50% probability level and H atoms are shown as small spheres of arbitrary
radii.
3920 | Org. Biomol. Chem., 2009, 7, 3918–3921
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
100(1) K, space group P 21 21 21 (no. 19), Z = 4, 22090 reflections measured,
9135 independent reflections (Rint = 0.0239) which were used in all calcu-
lations. The final R(F2) was 0.0737 (all data), Flack parameter = 0.3(2).
1 (a) W. Oppolzer, C. Chapuis and G. Bernardinelli, Helv. Chim. Acta,
1984, 67, 1397–1401; (b) W. Oppolzer, Pure Appl. Chem., 1988, 60,
39–48.
2 (a) D. P. Curran, W. Shen, J. Zhang and T. A. Heffner, J. Am. Chem.
Soc., 1990, 112, 6738–6740; (b) P. A. Zoretic, X. Y. Weng, C. K. Biggers,
M. S. Biggers, M. L. Caspar and D. G. Davis, Tetrahedron Lett., 1992,
33, 2637–2640; (c) B. H. Kim and D. P. Curran, Tetrahedron, 1993,
49, 293–318; (d) E. Baciocchi, E. Muraglia and C. Villani, Synlett,
1994, 821–822; (e) D. P. Curran, S. J. Geib and C.-H. Lin, Tetrahedron:
Asymmetry, 1994, 5, 199–202; (f) D. P. Curran, W. Shen, J. Zhang, S. J.
Gieb and C.-H. Lin, Heterocycles, 1994, 37, 1773–1788.
3 (a) H. Miyabe, C. Ushiro and T. Naito, Chem. Commun., 1997, 1789–
1790; (b) H. Miyabe, C. Ushiro, M. Ueda, K. Yamakawa and T. Naito,
J. Org. Chem., 2000, 65, 176–185; (c) H. Miyabe, C. Konishi and T.
Naito, Org. Lett., 2000, 2, 1443–1445; (d) M. Ueda, H. Miyabe, A.
Nishimura, H. Sugino and T. Naito, Tetrahedron: Asymmetry, 2003,
14, 2857–2859; (e) H. Myiabe, M. Ueda and T. Naito, Synlett, 2004,
7, 1140–1157; (f) M. Ueda, H. Miyabe, H. Sugino and T. Naito, Org.
Biomol. Chem., 2005, 3, 1124–1128; (g) T. Naito, Chem. Pharm. Bull.,
2008, 56, 1367–1383.
4 (a) F. Rodrigues, Y. Canac and A. Lubineau, Chem. Commun., 2000,
2049–2050; (b) Y. Hersant, R. Abou-Jneid, Y. Canac, A. Lubineau, M.
Philippe, D. Semeria, X. Radisson and M.-C. Scherrmann, Carbohydr.
Res., 2004, 339, 741–745; (c) N. Bragnier and M.-C. Scherrmann,
Synthesis, 2005, 5, 814–818.
Fig. 7 ORTEP drawing of 25a. Displacement ellipsoids are drawn at the
50% probability level and H atoms are shown as small spheres of arbitrary
radii.
5 S. Hanessian and R. Y. Yang, Tetrahedron Lett., 1996, 37, 5273–5276.
6 (a) D. H. Dube and C. R. Bertozzi, Nat. Rev. Drug Discovery, 2005, 4,
477–488; (b) R. A. Dwek, Chem. Rev., 1996, 96, 683–730.
7 (a) A. Dondoni and A. Marra, Chem. Rev., 2000, 100, 4395–4421 and
references therein; (b) A. Dondoni, P. P. Giovannini and A. Marra,
J. Chem. Soc., Perkin Trans. 1, 2001, 2380–2388; (c) A. Dondoni, G.
Mariotti, A. Marra and A. Massi, Synthesis, 2001, 14, 2129–2137; (d) T.
Nishikawa, M. Ishikawa, K. Wada and M. Isobe, Synlett, 2001, 945–
947; (e) L. Combo, M. di Giacomo and P. Ciceri, Tetrahedron, 2002,
58, 9381–9386; (f) B. Lygo, B. I. Andrews and D. Slack, Tetrahedron
Lett., 2003, 44, 9039–9041.
8 T. Gustafsson, M. Saxin and J. Kihlberg, J. Org. Chem., 2003, 68,
2506–2509.
Scheme 5
9 (a) Hydroxyethylene isosteres of glycosyl asparagines have been de-
scribed, however, these compounds are not isosteres of natural O- or
N-glycosyl amino acids. C. M. Huwe, T. J. Woltering, J. Jiricek, G.
Weitz-Schmidt and C.-H. Wong, Bioorg. Med. Chem., 1999, 7, 773–
788; (b) B. Westermann, A. Walter, U. Flo¨rke and H.-J. Altenbach, Org
Lett., 2001, 3, 1375–1378.
10 I. Riemann, M. A. Papadopoulos, M. Knorst and W.-D. Fessner,
Aust. J. Chem., 2002, 55, 147–154.
11 S. Norsikian, J. Zeitouni, S. Rat, S. Ge´rard and A. Lubineau, Carbohydr.
Res., 2007, 342, 2716–2728.
Acknowledgements
N. B. thanks the Ministe`re de l’Education Nationale, de
l’Enseignement Supe´rieur et de la Recherche for a research
fellowship. This work was supported by the Universite´ Paris-Sud
11 and the CNRS. We thank J.-P. Baltaze for some help in NMR
measurements.
12 (a) S. Hanessian, N. Bernstein, R.-Y. Yang and R. Maguire, Bioorg.
Med. Chem. Lett., 1999, 9, 1437–1442; (b) S. Hanessian, P.-P. Lu, J.-Y.
Sance´au, P. Chemla, K. Gohda, R. Fonne-Pfister, L. Prade and S. W.
Cowan-Jacob, Angew. Chem., Int. Ed., 1999, 38, 3160–3162.
13 The (R,S) configuration for compound 16b could also been envisaged
whereas the (R,R) configuration can be excluded since it was found in
product 19a i.e. that obtained from (+)-1 (X-ray analysis). However, if
the (R,S) configuration was assumed, it should be concluded that the
stereoselectivity at C3 was total. From the results obtained previously
(ref. 2, 3, 5, 12) using camphorsultam derivatives, we think it’s more
reasonable to make the hypothesis that the choice of the camphorsultam
auxiliary allowed a total control of the selectivity at C2, whereas the
stereocenter arising from the sugar moiety (C3) is mainly R.
14 T. Hasegawa and H. Yamamoto, Synlett, 1998, 882–884.
15 A. Khrimian, A. K. Margaryan and W. F. Schmidt, Tetrahedron, 2003,
59, 5475–5480.
16 R. Pascal, M. Laspe´ras, J. Taillades and A. Commeyras, New J. Chem.,
1987, 11, 235–244.
17 A proposed mechanism involves the formation of a carbinolamine
whose cyclisation gave an oxazolidin-5-one. Hydrolysis of this inter-
mediate should afford the amino acid (ref. 15). Participation of the
hydroxyl at C2 of the glucose moiety and elimination of O-benzyl N-
methyl hydroxylamine gave rise to 23a.
Notes and references
‡ Crystal data for compound 16a. C36H50N2O13S, M = 750.85, orthorhom-
3
˚
˚
bic, a = 12.629(5), b = 14.954(5), c = 20.739(5) A, U = 3917(2) A , T =
100(1) K, space group P 21 21 21 (no. 19), Z = 4, 35511 reflections measured,
11061 independent reflections (Rint = 0.0592) which were used in all calcu-
lations. The final R(F2) was 0.0687 (all data), Flack parameter = 0.01(5).
§ Crystal data for compound 17a. C36H50N2O13S, M = 750.85, monoclinic,
◦
˚
a = 8.9756(5), b = 17.3489(10), c = 12.2509(7) A, b = 94.036(2) , U =
3
˚
1902.94(19) A , T = 100(1) K, space group P 21 (no. 4), Z = 2, 10784
reflections measured, 7481 independant reflections (Rint = 0.0204) which
were used in all calculations. The final R(F2) was 0.0386 (all data), Flack
parameter = 0.04(5).
¶ Crystal data for compound 19a. C37H52Cl2N2O13S, M = 835.78,
˚
orthorhombic, a = 7.4542(15), b = 23.187(5), c = 23.706(5) A, U =
3
˚
4097.2(14) A , T = 100(1) K, space group P 21 21 21 (no. 19), Z = 4, 22600
reflections measured, 7989 independent reflections (Rint = 0.0437) which
were used in all calculations. The final R(F2) was 0.0927 (all data), Flack
parameter = 0.11(13).
ꢀ Crystal data for compound 25a. C18H26O11, M = 418.39, orthorhombic,
3
˚
˚
a = 9.9034(12), b = 10.2924(13), c = 20.413(2) A, U = 2080.7(4) A , T =
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 3918–3921 | 3921
©