Improved iterative olefination approach to oligosaccharide mimics: Stereoselective synthesis of β-(1→6)-D-galactopentaose methylene isostere

Article abstract of DOI:10.1016/S0040-4039(00)01112-6

The title C-pentagalactoside featuring β-(1→6) ethylene bridges between the D-galactopyranose units has been prepared through four consecutive cycles comprising the Wittig reaction of a formyl C-galactoside building block with sugar phosphorane substrate

Full text of DOI:10.1016/S0040-4039(00)01112-6

Tetrahedron Letters 41 (2000) 6657±6660
Improved iterative ole®nation approach to oligosaccharide
mimics: stereoselective synthesis of
b-(1!6)-d-galactopentaose methylene isostere
Alessandro Dondoni,* Mamoru Mizunoy and Alberto Marra
Á
Dipartimento di Chimica, Laboratorio di Chimica Organica, Universita di Ferrara, Via L. Borsari 46,
I-44100 Ferrara, Italy
Received 14 June 2000; accepted 29 June 2000
Abstract
The title C-pentagalactoside featuring b-(1!6) ethylene bridges between the d-galactopyranose units
has been prepared through four consecutive cycles comprising the Wittig reaction of a formyl C-galactoside
building block with sugar phosphorane substrate followed by the introduction of a methylphosphonium
group at the nonreducing end; the target product was obtained by hydrogenolysis of the tetraene penta-
saccharide. # 2000 Elsevier Science Ltd. All rights reserved.
Keywords: glycomimetics; glycosidase inhibitors; oligosaccharide isosteres; Wittig reaction.
Among the various types of C-glycosides which have been reported in recent years,1 C-oligo-
saccharides are very important as glycosidase-resistant probes for the study, at the molecular
level, of the role that carbohydrate moieties of glycoproteins and glycolipids play in biological
processes.2 Consequently, various C-disaccharides have been prepared by di€erent methods,3
although the most important types of these carbohydrate mimics are methylene isosteres of
natural products. A few examples of methylene-bridged C-trisaccharides4 and one example of
mixed O,C-trisaccharide5 have been reported. Quite recently, we,6 and Sinay and co-workers,7
developed iterative synthetic protocols based on Wittig ole®nation and lactone±alkyne coupling,
respectively, which a€orded b-(1!6)-C-oligogalactosides up to the tetrameric stage. Hence, quite
fortuitously the synthesis of the same C-tetrasaccharide, i.e. a b-(1!6)-C-d-galactotetraose, was
reported independently. Each cycle of our method constituted of the coupling of a galactosyl-
methylene phosphorane with a sugar aldehyde and the introduction of the formyl group at the
nonreducing end of the growing sugar oligomer. Quite deceptively, the resulting sugar aldehyde
subjected to the subsequent Wittig reaction partly underwent elimination of the benzyloxy group
* Corresponding author. Tel: 532-291176; fax: 532-291167; e-mail: adn@dns.unife.it
y
On temporary leave from the Noguchi Institute, Tokyo, Japan.
0040-4039/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)01112-6

6658
at the b-position. This side reaction caused a partial loss of the elongated constructs, with a
consequential decrease of the eciency of the method. The ecient protocol in the Sinay method,
as judged from the quoted yields in three consecutive cycles, prompted us to report here an
alternative Wittig-based synthesis of the same b-(1!6)-C-d-galactans, wherein the drawback in
our previous method is suppressed by the installation of the methylphosphonium group at the
nonreducing end of the growing oligosaccharide. The superior eciency of this new protocol is
substantiated by the high yields registered over four consecutive iterative cycles, culminating with
the ®rst synthesis of the title C,C,C,C-pentasaccharide.
The orthogonally hydroxy-protected formyl C-galactoside 2,8 the building block in the new
iterative protocol (Scheme 1), was prepared in ca. 5 g batches from tetra-O-benzyl-galactono-
lactone (®ve steps, 41%) by an improvement of the earlier method6 using benzothiazole as
formylating agent.9 The starting galactopyranose phosphonium iodide 1,8 easily available in gram
quantities, was obtained in 72% yield from methyl 2,3,4-tri-O-benzyl-a-d-galactopyranoside via
iodination and phosphanation. The initial Wittig reaction (coupling) of the ylide generated in situ
from 1 with the aldehyde 2 (1.5 equiv.) produced the Z-alkene 3 (J6,7=11.3 Hz) in very good
isolated yield (81%) after column chromatography (15:1 cyclohexane:AcOEt). A reversed ratio of
the reagents a€orded the alkene 3 in lower yield (61%), very likely because of partial
decomposition of the aldehyde 2. The preservation of the original con®guration at C5 and C8 in 3
was con®rmed by the J8,9 value of 9.3 Hz, in agreement with a b-d-linkage at the anomeric center
of one sugar moiety and the J4,5 value of 0.8 Hz consistent with the d-galacto con®guration in the
other moiety.10 The cycle was completed by the introduction of the phosphonium group in 3
(arming) by removal of the t-butyldiphenylsilyl (TBDPS) group, followed by iodination and
phosphanation. The phosphonium salt 4 was isolated in 88% yield after crystallization (CH2Cl2±
Et2O); the overall yield of 4 from 1 was 71%.
Scheme 1. Reagents and conditions: (a) BuLi, 3:1 THF:HMPA, ^20ꢀC, 4 h; (b) TBAF, THF, re¯ux, 1.5 h; (c) I2, PPh3,
imidazole, toluene, re¯ux, 1 h; (d) PPh3, 110ꢀC, 4 h

6659
The reiteration of the coupling±arming sequence over two cycles11 converted the alkene 4 into
the diene 5 (71% yield) and the latter into the triene 6 (67%) in very good overall yields.
Evidently, the nonstabilized ylides derived from the phosphonium salts 4 and 5 were reactive
enough, despite the increased size of these molecules. A clean ole®nation also took place from the
reaction of the ylide derived from 6 and the aldehyde 2 to give the tetraene 7 in excellent yield
(92%). Having demonstrated the eciency of the chain-growing protocol, the homologative
sequence was stopped at this stage. The sugar-alkene 7 was subjected to a suitable elaboration
toward the ®rst synthesis of the methylene isostere of a (1!6) pentagalactoside. This
transformation was carried out in two steps via reaction with tetrabutylammonium ¯uoride
(TBAF) to remove the silyl protective group and hydrogenolysis with Pd/C to reduce the double
bonds and remove the benzyl protective groups (Scheme 2). The methyl b-(1!6)-C-d-galacto-
pentapyranoside 8 was isolated and fully characterized as the corresponding O-acetyl derivative
9. This compound showed consistent MALDI-TOF mass spectrum and NMR spectra.12 The
complete assignment of the proton signals by NMR analysis at 500 MHz con®rmed the b-linkage
at the anomeric center of the carbohydrate units B±E, since a J1,2 value of 9.3 Hz was observed in
each case. Moreover, ROESY experiments indicated a cis-relationship between the H-5 and H-3
4
protons of all the monosaccharide moieties, as expected for d-galactopyranoses in the C1
conformation. Evidently, the structure of 9 demonstrates that the Wittig reactions in all
consecutive cycles did not a€ect the con®guration of the carbon atoms of the pyranose rings
bearing the reactive functional groups.
Scheme 2. Reagents and conditions: (a) TBAF, THF, re¯ux, 1 h; (b) H2, Pd/C, 7 bar, AcOEt±MeOH, rt; (c) Ac2O, Py,
rt, 16 h
In conclusion, the new iterative protocol comprised of operationally simple and high yield
reactions which allow us to obtain (1!6)-carbon-linked oligogalactosides on a preparatively
meaningful scale. The method has been tested for the synthesis of compounds up to a pentamer.
However, this appeared not to be the limit of application of the iterative protocol, since high
levels of eciency in the coupling±arming sequence were maintained over the four consecutive
cycles. The extension of this method to the preparation of various (1!6)-C-oligosaccharides now
becomes of great interest in view of the biological appeal of these compounds as mimics of nat-
ural O-oligosaccharide immunodeterminants that bind to monoclonal immunoglobulins.13
Acknowledgements
Financial support from MURST COFIN-98 (Italy) is gratefully acknowledged. We thank Mr.
Paolo Formaglio (University of Ferrara) for assistance in NMR analysis and Dr. Carla Marchiorro
(GlaxoWellcome Medicines Research Laboratory, Verona, Italy) for the 500 MHz NMR
experiments.

6660
References
1. (a) Postema, M. H. D. C-Glycoside Synthesis; CRC Press: Boca Raton, 1995. (b) Levy, D. E.; Tang, C. The
Chemistry of C-Glycosides. Elsevier Science: Oxford, 1995.
2. (a) Dwek, A. R. Chem. Rev. 1996, 96, 683. (b) Essentials of Glycobiology, Varki, A.; Cummings, R.; Esko, J.;
Freeze, H.; Hart, G.; Marth, J.; Eds.; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, New York,
1999.
3. For selected recent papers with leading references, see: (a) Dondoni, A.; Zuurmond, H. M.; Boscarato, A. J. Org.
Chem. 1997, 62, 8114. (b) Rubinstenn, G.; Esnault, J.; Mallet, J.-M.; Sinay, P. Tetrahedron: Asymmetry 1997, 8,
1327. (c) Andersen, L.; Munch Mikkelsen, L.; Beau, J.-M.; Skrydstrup, T. Synlett 1998, 1393. (d) Ravishankar,
R.; Surolia, A.; Vijayan, M.; Lim, S.; Kishi, Y. J. Am. Chem. Soc. 1998, 120, 11297. (e) Bazin, H. G.; Du, Y.;
Polat, T.; Linhardt, R. J. J. Org. Chem. 1999, 64, 7254. (f) Jarreton, O.; Skrydstrup, T.; Espinosa, J.-F.; Jimenez-
Barbero, J.; Beau, J.-M. Chem. Eur. J. 1999, 5, 430. (g) Postema, M. H. D.; Calimente, D. Tetrahedron Lett. 1999,
40, 4755. (h) Grin, F. K.; Paterson, D. E.; Taylor, R. J. K. Angew. Chem., Int. Ed. 1999, 38, 2939. (i) Roy, R.;
Dominique, R.; Das, S. K. J. Org. Chem. 1999, 64, 5408. (j) Zhu, Y.-H.; Demange, R.; Vogel, P. Tetrahedron:
Asymmetry 2000, 11, 263.
4. (a) Haneda, T.; Goekjian, P. G.; Kim, S. H.; Kishi, Y. J. Org. Chem. 1992, 57, 490. (b) Wei, A.; Haudrechy, A.;
Audin, C.; Jun, H.-S.; Haudrechy-Bretel, N.; Kishi, Y. J. Org. Chem. 1995, 60, 2160. (c) Sutherlin, D. P.;
Armstrong, R. W. J. Org. Chem. 1997, 62, 5267.
5. Spak, S. J.; Martin, O. R. Tetrahedron 2000, 56, 217.
6. Dondoni, A.; Kleban, M.; Zuurmond, H.; Marra, A. Tetrahedron Lett. 1998, 39, 7991.
7. Xin, Y.-C.; Zhang, Y.-M.; Mallet, J.-M.; Glaudemans, C. P. J.; Sinay, P. Eur. J. Org. Chem. 1999, 471.
20
8. Compound 1. M.p. 93±94ꢀC (CH2Cl2±Et2O); ‰ꢀŠ +50.5 (c 0.7; CHCl3); 1H NMR (300 MHz, CDCl3): ꢁ 7.81±7.23
D
(m, 30H, 6Ph), 5.14 and 4.89 (2d, 2H, J=11.2 Hz, PhCH2), 5.09 (ddd, 1H, J5,6a=3.0, J6a,6b=J6a,P=16.0 Hz, H-
6a), 4.96 (ddd, 1H, J3,4=2.6, J4,5=J4,P=1.0 Hz, H-4), 4.92 and 4.82 (2d, 2H, J=11.5 Hz, PhCH2), 4.85 and 4.67
(2d, 2H, J=12.3 Hz, PhCH2), 4.45 (dddd, 1H, J5,6b=10.7, J5,P=10.0 Hz, H-5), 4.39 (d, 1H, J1,2=3.6 Hz, H-1),
4.07 (dd, 1H, J2,3=10.3 Hz, H-3), 3.98 (dd, 1H, H-2), 3.50 (ddd, 1H, J6b,P=10.3 Hz, H-6b), 2.48 (s, 3H, OMe). 31
P
1
NMR (121 MHz, CDCl3): ꢁ 24.9. Compound 2. H NMR (300 MHz, CDCl3): ꢁ 9.59 (d, 1H, J1,2=1.2 Hz, H-1),
7.66±7.60 and 7.50±7.25 (2m, 25H, 5Ph), 5.00 and 4.67 (2d, 2H, J=11.2 Hz, PhCH2), 4.90 and 4.69 (2d, 2H,
J=10.6 Hz, PhCH2), 4.81 (s, 2H, PhCH2), 4.07 (dd, 1H, J4,5=2.6, J5,6=0.7 Hz, H-5), 4.06 (dd, 1H, J2,3=10.0,
J3,4=9.0 Hz, H-3), 3.88 (dd, 1H, J6,7a=8.2, J7a,7b=10.3 Hz, H-7a), 3.83 (dd, 1H, J6,7b=5.7 Hz, H-7b), 3.71 (dd,
1H, H-2), 3.68 (dd, 1H, H-4), 3.48 (ddd, 1H, H-6), 1.08 (s, 9H, t-Bu).
9. The replacement of thiazole with benzothiazole in the synthesis of formyl C-glycosides gives rise to considerable
economical advantages and produces, in many cases, crystalline compounds which can be more easily puri®ed and
handled. However, the liberation of the formyl group from benzothiazole requires the assistance of silver ion in the
hydrolysis of the benzothiazoline.
10. Further proof of the assigned structure came from NOE experiments performed on the peracetylated (1!6)-C-
disaccharide obtained from 3 through desilylation, hydrogenation and acetylation.
11. The Wittig reactions were carried out under the same conditions employed in the ®rst cycle but using 1.2 equiv. of
the formyl C-galactoside 2.
20
12. Compound 9. ‰ꢀŠ +32.8 (c 0.6; CHCl3); 1H NMR (500 MHz, CDCl3): ꢁ 5.40 (dd, 1H, J3,4=3.5, J4,5=1.1 Hz, H-
D
4E), 5.34 (dd, 1H, J3,4=3.5, J4,5=0.7 Hz, H-4A), 5.31 (dd, 1H, J2,3=10.8 Hz, H-3A), 5.28 (dd, 3H, J3,4=3.5,
J4,5=0.7 Hz, H-4B, H-4C, H-4D), 5.13 (dd, 1H, J1,2=3.7, J2,3=10.6 Hz, H-2A), 5.06, 5.04, and 5.03 (3dd, 4H,
J1,2=9.3, J2,3=10.3 Hz, H-2B, H-2C, H-2D, H-2E), 4.99 (dd, 1H, H-3E), 4.96, 4.95 and 4.94 (3dd, 3H, H-3B, H-
3C, H-3D), 4.96 (d, 1H, H-1A), 4.11 (dd, 1 H, J5,6a=6.8, J6a,6b=11.4 Hz, H-6aE), 4.06 (dd, 1H, J5,6b=6.5 Hz, H-
6bE), 3.91 (ddd, 1H, J5,6a=6.0, J5,6b=8.0 Hz, H-5A), 3.84 (ddd, 1H, J5,6a=J5,6b=6.5 Hz, H-5E), 3.52±3.48 (m,
3H, H-5B, H-5C, H-5D), 3.38 (s, 3H, OMe), 3.38 (ddd, 1H, J=2.0, 8.2, 9.3 Hz, H-1E), 3.32 and 3.30 (2ddd, 3H,
J=2.0, 8.2, 9.3 Hz, H-1B, H-1C, H-1D), 2.18±1.96 (9s, 48H, 16Ac), 1.72±1.58 and 1.48±1.40 (2m, 16H,
4CH2CH2). MALDI-TOF MS: 1530.4 (M+Na), 1546.5 (M+K).
13. Glaudemans, C. P. J. Chem. Rev. 1991, 91, 25.