Radical-mediated synthesis of α-C-glycosides based on N-acyl galactosamine

Article abstract of DOI:10.1021/ol006682c

(Matrix presented) C-glycoside analogue of O-benzyl α-D-GalNAc C-Glycosides of N-acyl 2-amino-2-deoxygalactose (acyl = MeCO, CF₃CO, t-BuOCO) are available in a stereoselective manner by trapping of an anomeric radical with an activated alkene.

Full text of DOI:10.1021/ol006682c

ORGANIC
LETTERS
2000
Vol. 2, No. 25
4051-4054
Radical-Mediated Synthesis of
r-C-Glycosides Based on N-Acyl
Galactosamine
Raul SanMartin,† Bahareh Tavassoli,† Kenneth E. Walsh,† Daryl S. Walter,‡ and
,†
Timothy Gallagher*
School of Chemistry, UniVersity of Bristol, Bristol BS8 1TS, U.K., and Roche
DiscoVery Welwyn, Broadwater Road, Welwyn Garden City AL7 3AY, U.K.
t.gallagher@bristol.ac.uk
Received October 3, 2000
ABSTRACT
C-Glycosides of N-acyl 2-amino-2-deoxygalactose (acyl ) MeCO, CF3CO, t-BuOCO) are available in a stereoselective manner by trapping of
an anomeric radical with an activated alkene. Using anomeric selenides, radical generation and trapping is carried out under conditions that
avoid competitive reduction, and this chemistry has been applied to the synthesis of the novel C-glycoside analogue of O-benzyl r-D-GalNAc.
Molecules that are effective inhibitors of glycosylation are
valuable as tools for probing the significance of conjugating
carbohydrates to proteins and lipids,1 and recent results2,3
based on the inhibitory properties of O-benzyl R-D-GalNAc
1 have furthered current understanding of the regulation of
O-glycosylation and mucin metabolism.
of O-benzyl GalNAc inhibition specifically blocks sialic acid
transfer, which is an additional regulatory process of physi-
ological significance.3
In the course of our studies, uncertainties arose regarding
the stability of 1, which appears to be processed in some
cell lines. As a result, we have initiated a parallel program
to explore the corresponding C-glycosyl variants represented
by general structure 2, with a key target being C-glycoside
3 as a stable analogue of 1.
This simple monosaccharide exhibits a differential action2
on a family of GalNAc transferases, and a cellular product
‡ Roche Discovery Welwyn.
† University of Bristol.
(1) Barchi, J. J. Curr. Pharm. Des. 2000, 6, 485-501.
(2) Kuan, S. F.; Byrd, J. C.; Basbaum, C.; Kim, Y. S. J. Biol. Chem.
1989, 264, 19271-19277. Byrd, J. C.; Dahiya, R.; Huang, J.; Kim, Y. S.
Eur. J. Cancer 1995, 31A, 1498-1505. Hennebicq-Reig, S.; Lesuffleur,
T.; Capon, C.; De Bolos, C.; Kim, I.; Moreau, O.; Richet, C.; He´mon, B.;
Recchi, M. A.; Mae¨s, E.; Aubert, J. P.; Real, F. X.; Zweibaum, A.; Delannoy,
P.; Degand, P.; Huet, G. Biochem. J. 1998, 334, 283-295. Zanetta, J. P.;
Gouyer, V.; Mae¨s, E.; Pons, A.; He´mon, B.; Zweibaum, A.; Delannoy, P.;
Huet, G. Glycobiology 2000, 10, 565-575.
(3) Delannoy, P.; Kim, I.; Emery, N.; DeBolos, C.; Verbert, A.; Degand,
P.; Huet, G. Glycoconjugate J. 1996, 13, 717-726. Nakano, T.; Matsui,
T.; Ota, T. Anticancer Res. 1996, 16, 3577-3584. Huet, G.; Hennebicq-
Reig, S.; de Bolos, C.; Ulloa, F.; Lesuffleur, T.; Barbat, A.; Carrie`re, V.;
Kim, I.; Real, F. X.; Delannoy, P.; Zweibaum, A. J. Cell Biol. 1998, 141,
1311-1322. Ulloa, F.; Franci, C.; Real, F. X. J. Biol. Chem. 2000, 275,
18785-18793
The chemical stability and favorable conformational
properties associated with C-glycosides have encouraged the
development of a wide range of synthetic strategies,4 within
which radical-based methods have found broad application.
While nucleophilic radical character at the anomeric site has
10.1021/ol006682c CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/11/2000

been applied successfully to elaborate 2-alkoxy, 2-fluoro, and
2-deoxysaccharides, as well as higher sugars, this process
has yet to encompass fully 2-amino and 2-acetamido sugars.5
Keck-type allylations (using allyl tributylstannane) have been
described,6 but this method provides only the C-allyl variant.
Czernecki5k and Sinay¨5p have shown that anomeric selenides
do provide radicals, but these have only been trapped in
intramolecular processes.
1-(Phenylselenyl)-2-azido-2-deoxy-R-D-galactose (4) pro-
vides an attractive starting point and is readily available via
azide radical addition to D-galactal, as described by both
Czernecki9 and Santoyo-Gonza´lez10 (Scheme 1). Selenide 4
Scheme 1a
More recently, Fessner7 has captured the radical derived
from 1-bromo-N-(trifluoroacetyl)-R-D-glucosamine with vinyl
phosphonates to give the corresponding C-glycosides.
While anomeric halides (see above) are generally a
valuable source of anomeric radicals, there are limitations
associated with using glycosyl bromides of 2-amino sugars.8
For this reason, we have examined alternative radical
precursors with the additional aim of providing access to a
broad range of other 2-amino C-glycosyl derivatives, and
our initial results, which focus on the galacto series, are
described in this paper.
a Reagents and conditions: (a) PhI(OAc)2, NaN3, PhSeSePh
(82%); (b) 5a: MeCOSH (95%); 5b: Et3N, HS(CH2)3SH, then
(F3CCO)2O (75%); 5c: Et3N, HS(CH2)3SH, then Boc2O (80%).
is readily handled, and homolytic C-Se bond cleavage is a
well established and efficient method for radical genera-
tion.5k,p,11 Further, the primary amine (arising from azide
reduction) is easily converted to the N-acetyl derivative 5a,
as well as the N-trifluoroacetyl and N-Boc variants 5b and
5c, respectively, all of which exhibit good stability.
Using “standard” as well as a series of modified conditions
for radical generation, reaction of 5a in the presence of either
methyl acrylate or styrene failed to give any C-glycoside.12
C-Se cleavage occurred, but only reduction product 6a was
observed, which was generally obtained in quantitative yield.
The key to the successful trapping of an anomeric radical
derived from 5 (other than by reduction) involves use of Et3B
as initiator at room temperature. Under these conditions, we
were able to generate the desired radical and capture this
species with a series of activated alkenes to give the
corresponding R-C-glycosides 7 in 17-93% yield (Scheme
2 and Table 1).
(4) Postema, M. H. D. Tetrahedron 1992, 48, 8545-8599. Levy, D. E.;
Tang, C. The Chemistry of C-Glycosides; Elsevier Science Publishers:
Amsterdam, 1995. Postema, M. H. D. C-Glycoside Synthesis; CRC Press:
USA, 1995. Togo, H.; He, W.; Waki, Y.; Yokoyama, M. Synlett 1998, 700-
717.
(5) For other synthetic approaches to C-glycosides of 2-amino sugars,
see (a) Nicotra, F.; Russo, G.; Ronchetti, F.; Toma, L. Carbohydr. Res.
1983, 124, C5-C7. (b) Hoffmann, M. G.; Schmidt, R. R. Liebigs Ann.
Chem. 1985, 2403-2419. (c) Giannis, A.; Mu¨nster, P.; Sandhoff, K.;
Steglich, W. Tetrahedron 1988, 44, 7177-7180. (d) Carcano, M.; Nicotra,
F.; Panza, L.; Russo, G. J. Chem. Soc., Chem. Commun. 1989, 297-298.
(e) Abel, S.; Linker, T.; Giese, B. Synlett 1991, 171-172. (f) Grondin, R.;
Leblanc, Y.; Hoogsteen, K. Tetrahedron Lett. 1991, 32, 5021-5024. (g)
Bertozzi, C. R.; Bednarski, M. D. Tetrahedron Lett. 1992, 33, 3109-3112.
(h) Kim, K. I.; Hollingsworth, R. I. Tetrahedron Lett. 1994, 35, 1031-
1032. (i) Leteux, C.; Veyrie`res, A. J. Chem. Soc., Perkin Trans. 1 1994,
2647-2655. (j) Ayadi, E.; Czernecki, S.; Xie, J. A. J. Chem. Soc., Chem.
Commun. 1996, 347-348. (k) Czernecki, S.; Ayadi, E.; Xie, J. Tetrahedron
Lett. 1996, 37, 9193-9194. (l) Petrusˇova´, M.; BeMiller, J. N.; Petrusˇ, L.
Tetrahedron Lett. 1996, 37, 2341-2344. (m) Roe, B. A.; Boojamra, C. G.;
Griggs, J. L.; Bertozzi, C. R. J. Org. Chem. 1996, 61, 6442-6445. (n)
Wang, L. X.; Fan, J. Q.; Lee, Y. C. Tetrahedron Lett. 1996, 37, 1975-
1978. (o) Cipolla, L.; Lay, L.; Nicotra, F. J. Org. Chem. 1997, 62, 6678-
6681. (p) Rubinstenn, G.; Esnault, J.; Mallet, J. M.; Sinay¨, P. Tetrahedron:
Asymmetry 1997, 8, 1327-1336. (q) Cui, J. R.; Horton, D. Carbohydr.
Res. 1998, 309, 319-330. (r) Urban, D.; Skrydstrup, T.; Beau, J. M. J.
Org. Chem. 1998, 63, 2507-2516. (s) Urban, D.; Skrydstrup, T.; Beau, J.
M. Chem. Commun. 1998, 955-956. (t) Westermann, B.; Walter, A.;
Diedrichs, N. Angew. Chem., Int. Ed. 1999, 38, 3384-3386. (u) Cipolla,
L.; La Ferla, B.; Lay, L.; Peri, F.; Nicotra, F. Tetrahedron: Asymmetry
2000, 11, 295-303. (v) Vidal, T.; Haudrechy, A.; Langlois, Y. Tetrahedron
Lett. 1999, 40, 5677-5680. (w) Dondoni, A.; Mariotti, G.; Marra, A.
Tetrahedron Lett. 2000, 41, 3483-3487.
Scheme 2a
(6) Keck, G. E.; Yates, J. B. J. Am. Chem. Soc. 1982, 104, 5829-5831.
Keck, G. E.; Enholm, E. J.; Yates, J. B.; Wiley, M. R. Tetrahedron 1985,
41, 4079-4094. For carbohydrate-based applications of this methodology,
see: Giese, B.; Linker, T.; Muhn, R. Tetrahedron 1989, 45, 935-940.
Paulsen, H.; Matschulat, P. Liebigs Ann. Chem. 1991, 487-495. Waglund,
T.; Claesson, A. Acta Chem. Scand. 1992, 46, 73-76. See also ref 5m.
(7) Junker, H. D.; Fessner, W. D. Tetrahedron Lett. 1998, 39, 269-
272. Junker, H. D.; Phung, N.; Fessner, W. D. Tetrahedron Lett. 1999, 40,
7063-7066.
a Reagents and conditions: (a) see footnote 12; (b) Et3B,
n-Bu3SnH, PhMe, hυ, ultrasound, rt.
(8) Horton5q has reported the successful Keck-type allylation process
using anomeric chlorides and xanthates of 2-amido sugars. Attempts to use
2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-R-D-glycopyranosyl bromide as a
radical precursor led only to the corresponding oxazoline. N-Trifluoroacetyl
derivatives, as used by Fessner,7 are less prone to oxazoline formation.
(9) Czernecki, S.; Randriamandimby, D. Tetrahedron Lett. 1993, 34,
7915-7916. Czernecki, S.; Ayadi, E.; Randriamandimby, D. J. Org. Chem.
1994, 59, 8256-8260.
Importantly, the R-anomer 7 was observed, and the
corresponding â-isomer has not been detected. We have
found that a combination of irradiation (200 or 400 W) and
use of an ultrasonic bath also improved yields and reaction
times. Reduction of the intermediate radical resulting in 6a-c
does still compete and remains the principle side reaction
observed.
(10) Santoyo-Gonza´lez, F.; Calvo-Flores, F. G.; Garc´ıa-Mendoza, P.;
Herna´ndez-Mateo, F.; Isac-Garc´ıa, J.; Robles-D´ıaz, R. J. Org. Chem. 1993,
58, 6122-6125. See also: Giuliano, R. M.; Davis, R. S.; Boyko, W. J. J.
Carbohyd. Chem. 1994, 13, 1135-1143.
4052
Org. Lett., Vol. 2, No. 25, 2000

Table 1. R-C-Glycosides 7 Derived from Selenides 5a-c
a Isolated yields are after chromatography. b A 2:1 inseparable mixture of isomers was observed, provisionally assigned as diastereoisomers at the R-amino
acid center, rather than epimers at C(1).
1
The structure of adducts 7 was based primarily on H
analysis, and complete proton assignments were made using
1H-1H and 1H-13C correlations. Anomeric stereochemistry
(at C(1), using conventional sugar numbering) was based
on the signal attributed to H(2), since H(1) was not well
resolved.13 A sound basis for this protocol was provided by
the H NMR spectrum obtained for the reduction product
1
6a: JH(1R)-H(2) ) 11.0; JH(1â)-H(2) ) 5.1; JH(2)-NH ) 7.7;
JH(2)-H(3) ) 11.2 Hz. Comparing these data to the half-line
4
(13) C-Glycosides 7 exist predominantly in the C1 chair conformation
(11) Stork, G.; Suh, H. S.; Kim, G. J. Am. Chem. Soc. 1991, 113, 7054-
7056. Gupta, V.; Besev, M.; Engman, L. Tetrahedron Lett. 1998, 39, 2429-
2432. Abe, H.; Shuto, S.; Matsuda, A. J. Org. Chem. 2000, 65, 4315-
4325. Shuto, S.; Terauchi, M.; Yahiro, Y.; Abe, H.; Ichikawa, S.; Matsuda,
A. Tetrahedron Lett. 2000, 41, 4151-4155.
(12) A variety of reaction conditions and reagents were studied: Bu3-
SnH, Et3B (at temperatures other than at rt and also in the absence of a
hydride source) or AIBN, PhH or PhMe, at room temperature, 40 °C, 60
°C, and at reflux. Bu3SnSnBu3 and (TMS)3Si-H were also examined under
various conditions, without success. When reaction of 5a with Bu3SnH was
carried out without an alkene trap, quantitative reduction to give 6a was
observed.
[data for 7a: JH(2)-H(3) ) 9.5, JH(3)-H(4) ) 3.1, J
) 3.1 Hz]. As a
H(4)-H(5)
result, and using data available from reduction product 6a, H(2) was
predicted to exhibit a bandwidth (W1/2) of 23.8 or 29.7 Hz for the R- and
â-C-glycoside configurations, respectively. For 7a, we observed 23.5 Hz,
and for all other cases where H(2) was resolved, a consistent pattern was
4
followed. It has been reported5s that the C1 conformation is preferred for
both peracetylated C-glycosides (related to 7) and deprotected variants (such
as 3). It is interesting to note that alternative conformations are associated
with variants carrying O-benzyl protecting groups.5p
(14) Tingoli, M.; Tiecco, M.; Chianelli, D.; Balducci, R.; Temperini, A.
J. Org. Chem. 1991, 56, 6809-6813. Tingoli, M.; Tiecco, M.; Testaferri,
L.; Temperini, A. J. Chem. Soc., Chem. Commun. 1994, 1883-1884.
Org. Lett., Vol. 2, No. 25, 2000
4053

width (W1/2) of the H(2) signal for the C-glycosides 7 (e.g.,
7a W1/2 ) 24 Hz) allowed assignment of the R-stereochem-
istry shown in Table 1.
intermediate with methyl acrylate have failed. Only D-galactal
was recovered.15
Finally, the acetylated C-glycoside 7b was deprotected
(cat. NaOMe, MeOH, then Dowex 50 H+ resin) in quantita-
tive yield to give 3, the C-glycoside analogue of benzyl R-D-
GalNAc 1. The biological evaluation of 3, as well as a series
of other novel O- and C-glycosyl variants, is now underway.
In summary, the radical-based strategy that has already
found widespread application within 2-alkoxy and 2-deoxy
sugars can be applied generally to the synthesis of C-glycosyl
derivatives of 2-amino sugars. The success of this process
derives from (i) use of anomeric selenides as the source of
radical reactivity, which enables a range of N-protecting
groups to be employed, and (ii) suppression of the competi-
tive reduction of the intermediate radical by use of mild
reaction conditions.
The C-glycosides shown in Table 1 represent a versatile
group of functionalized galacto derivatives, and the suc-
cessful trapping of styrene to give, e.g., 7b is noteworthy.
More generically, the use of an anomeric selenide (as
opposed to a bromide) provides C-glycosides with differential
and orthogonal protection of the 2-amino moiety (NAc vs
NCOCF3 vs NBoc). Further, an ability to incorporate both
carboxylates, e.g., 7a and 7f, and sulfonates, e.g., 7c, offers
added flexibility in terms of future chemical modification
of both the core galacto and the side chain C-glycosyl
components.
A potentially more direct route toward the target C-
glycosides has also been evaluated (Scheme 3). Addition of
Azidoselenation is applicable to other glycals,9,16 and this
offers an opportunity to extend the radical-based methodol-
ogy and make available a range of other 2-amino C-glycosyl
derivatives.
Scheme 3a
Acknowledgment. We thank Dr. Anthony P. Corfield
(Division of Medicine, University of Bristol) for his advice.
R.S.M. thanks the Spanish Government and the Royal
Society for a fellowship, and K.E.W. thanks Roche Discovery
Welwyn and EPSRC for a CASE award.
Supporting Information Available: Experimental pro-
cedures and 1H and 13C NMR and other characterization data
for key compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
OL006682C
a Reagents and conditions: (a) PhI(OAc)2, NaN3; (b) (PhSe)2;
(c) H2CdCHCO2Me and conditions (a); (c) Et3B n-Bu3SnH, PhMe,
hυ, ultrasound, room temperature.
(15) Selenide 4 undergoes cleavage (AIBN, PhH at reflux, and also with
Et3B, Scheme 3) followed by rapid elimination of N3 to give D-galactal.
Santoyo-Gonzalez, F.; Calvo-Flores, F. G.; Hernandez-Mateo, F.; Garcia-
Mendoza, P.; Isac-Garcia, J.; Perez-Alvarez, M. D. Synlett 1994, 454-
456.
•
(16) For stereoselective approaches to gluco variants based on azidoni-
tration, see: Seeberger, P. H.; Roehrig, S.; Schell, P.; Wang, Y.; Christ,
W. J. Carbohydr. Res. 2000, 328, 61-69.
•
N3 to D-galactal is postulated14 to proceed via the 2-azido
anomeric radical 8, but to date all attempts to trap this
4054
Org. Lett., Vol. 2, No. 25, 2000