Addition of Lithiated C-Nucleophiles to 2,3-O-Isopropylidene-D-erythronolactone: Stereoselective Formation of a Furanose C-Disaccharide

Article abstract of DOI:10.1021/jo0350256

Addition of PhLi and lithiated dithianes to 2,3-O-isopropylidene-D-erythronolactone affords lactols, which are reduced with Et₃SiH to the corresponding C-glycosides, the structures of two of which have been solved by X-ray diffraction. The use

Full text of DOI:10.1021/jo0350256

Many routes to C-glycosides have been developed2 and
the applications to C-disaccharides and C-oligosaccha-
rides are important extensions.4 We are interested in the
application of dithioacetal chemistry to the creation of
the crucial C-C bond between the “nonreducing” end of
a C-disaccharide and the “aglycon” portion. With the
growing interest in furanose-based oligosaccharides as
components of, for example, bacterial cell walls5 we chose
to investigate the synthesis of a furanose C-disaccharide.
As shown in Figure 1, the coupling of a dithiane nucleo-
phile with an electrophilic sugar lactone, followed by
removal of -SPh groups and reduction of the resultant
lactol, should provide an efficient protocol for C-disac-
charide synthesis. Addition of organometallic reagents,6
including lithiated dithiane nucleophiles,7 to carbohydrate-
derived lactones is a well-known route to C-glycosidic
compounds and the reduction of lactols is usually ste-
reoselective. The work of Dondoni and colleagues2c,d is of
particular value here in which sugar lactones were
coupled with lithiated heterocycles.8 The lactol interme-
diates from these reactions may be reduced stereoselec-
tively, for example via their acetates, en route to
C-glycosides.8b Questions of stereocontrol are of interest
in the addition step and of paramount importance in the
reduction step for the present synthesis to be useful. Here
we detail the application of this approach to the simple
furanose C-disaccharide 1, as well as evidence from X-ray
diffraction for the stereochemical course of the synthetic
steps.
Ad d ition of Lith ia ted C-Nu cleop h iles to
2,3-O-Isop r op ylid en e-D-er yth r on ola cton e:
Ster eoselective F or m a tion of a F u r a n ose
C-Disa cch a r id e
J ason L. McCartney, Christopher T. Meta,
Robert M. Cicchillo, Matthew D. Bernardina,†
Timothy R. Wagner,‡ and Peter Norris*
Department of Chemistry, Youngstown State University,
1 University Plaza, Youngstown, Ohio 44555-3663
pnorris@ysu.edu
Received J uly 16, 2003
Abstr a ct: Addition of PhLi and lithiated dithianes to 2,3-
O-isopropylidene-D-erythronolactone affords lactols, which
are reduced with Et3SiH to the corresponding C-glycosides,
the structures of two of which have been solved by X-ray
diffraction. The use of a D-ribose-derived lithiated dithiane
nucleophile in this chemistry allows for the convenient
construction of a furanose C-disaccharide.
The chemical synthesis of carbohydrate mimics (gly-
comimetics) is driven by the needs of glycobiology since
compounds such as iminosugars and C-glycosides are
valuable tools for the study of carbohydrate-associated
proteins.1 The C-glycosides are compounds in which a
methylene group replaces the exocyclic oxygen of the
O-glycoside to be mimicked, and this substitution endows
the mimic with the ability to withstand enzymatic
hydrolysis and thus serve as a stable substitute for the
O-glycoside.2 There is continued debate about the actual
efficacy of C-glycoside mimetics since there are no ano-
meric effects (either endo or exo) in play and there may
be significant differences in their preferred conformations
versus the O-glycosides;3 however, there is evidence that
C-glycosides are capable of binding to proteins such as
glycosyl hydrolases, glycosyl transferases, and lectins.4
(3) (a) Kishi, Y. Tetrahedron 2002, 58, 6239-6258. (b) Hideya, Y.;
Hironobu, H. Trends Glyosci. Glycotechnol. 2001, 13, 31-55. (c)
J imenez-Barbero, J .; Espinosa, J . F.; Asensio, J . L.; Canada, F. J .;
Poveda, A. Adv. Carbohydr. Chem. Biochem. 2001, 56, 235-284.
(4) For recent examples see: (a) McAllister, G. D.; Paterson, D. E.;
Taylor, R. J . K. Angew. Chem., Int. Ed. 2003, 42, 1387-1391. (b)
Gemmell, N.; Meo, P.; Osborn, H. M. I. Org. Lett. 2003, 5, 1649-1652.
(c) Viode, C.; Vogel, P. J . Carbohydr. Chem. 2001, 20, 733-746. (d)
Liu, L.; Postema, M. H. D. J . Am. Chem. Soc. 2001, 123, 8602-8603.
(e) Compain, P.; Martin, O. R. Bioorg. Med. Chem. 2001, 9, 3077-
3092. (f) Vogel, P. Chimia 2001, 55, 359-365. (g) Nicotra, F. Top. Curr.
Chem. 1997, 187, 55-83. (h) Legler, G. Naturwissenschaften 1993, 80,
397-409.
(5) (a) Han, J .; Gadikota, R. R.; McCarren, P. R.; Lowary, T. L.
Carbohydr. Res. 2003, 338, 581-588. (b) Lowary, T. L. J . Carbohydr.
Chem. 2002, 21, 691-722.
† National Science Foundation REU program participant from Ohio
Wesleyan University, Summer 2002.
‡ To whom enquiries about the X-ray data should be addressed
(trwagner@ysu.edu).
(6) Addition of organometallics: (a) Yang, W.-B.; Yang, Y.-Y.; Gu,
Y.-F.; Wang, S.-H.; Chang, C.-C.; Lin, C.-H. J . Org. Chem. 2002, 67,
3773-3782. (b) Dondoni, A.; Mariotti, G.; Marra, A.; Massi, A.
Synthesis 2001, 2129-2137. (c) Wellner, E.; Gustafsson, T.; Backlund,
J .; Holmdahl, R.; Kihlberg, J . ChemBioChem. 2000, 1, 272-280. (d)
Best, W. M.; Ferro, V.; Harle, J .; Stick, R. V.; Tilbrook, D. M. G. Aust.
J . Chem. 1997, 50, 463-472. (e) Csuk, R.; Kuehn, M.; Stroehl, D.
Tetrahedron 1997, 53, 1311-1322. (f) Ayadi, E.; Czernecki, S.; Xie, J .
Chem. Commun. 1996, 347-348. (g) Boyd, V. A.; Drake, B. E.;
Sulikowski, G. A. J . Org. Chem. 1993, 58, 3191-3193. (h) Preuss, R.;
Schmidt, R. R. J . Carbohydr. Chem. 1991, 10, 887-900. (i) Kraus, G.
A.; Molina, M. T. J . Org. Chem. 1988, 53, 752-753. (j) Lancelin, J .
M.; Zollo, P. H. A.; Sinåy, P. Tetrahedron Lett. 1983, 24, 4833-4836.
(k) Lewis, M. D.; Cha, J . K.; Kishi, Y. J . Am. Chem. Soc. 1982, 104,
4976-4978.
(1) For general reviews see: (a) Carbohydrate-based Drug Discovery;
Wong, C.-H., Ed.; Wiley-VCH: Weinheim, Germany, 2003. (b) Bach,
P.; Bols, M.; Damsgard, A.; Hansen, S. U.; Lohse, A. Introduction to
Glycomimetics. In Glycoscience; Fraser-Reid, B. O., Tatsuta, K., Thiem,
J ., Eds.; Springer-Verlag: Heidelberg, Germany, 2001; Vol. 3, pp 2533-
2540.
(2) For recent general reviews on C-glycosides see: (a) Liu, L.;
McKee, M.; Postema, M. H. D. Curr. Org. Chem. 2001, 5, 1133-1167.
(b) Gyorgydeak, Z.; Pelyvas, I. F. C-Glycosylation. In Glycoscience;
Fraser-Reid, B. O., Tatsuta, K., Thiem, J ., Eds.; Springer-Verlag:
Heidelberg, Germany, 2001; Vol. 1, pp 691-747. (c) Dondoni, A. Pure
Appl. Chem. 2000, 72, 1577-1588. (d) Dondoni, A.; Marra, A. Chem.
Rev. 2000, 100, 4395-4421. (e) Smoliakova, I. P. Curr. Org. Chem.
2000, 4, 598-608. (f) Spencer, R. P.; Schwartz, J . Tetrahedron, 2000,
56, 2103-2112. (g) Du, Y.; Linhardt, R. J .; Vlahov, I. R. Tetrahedron,
1998, 54, 9913-9959. (h) Bertozzi, C.; Bednarski, M. Synthesis of
C-Glycosides; Stable Mimics of O-Glycosidic Linkages. In Frontiers in
Natural Product Research (Modern Methods in Carbohydrate Synthe-
sis); Harwood Academic, 1996; Vol. 1, pp 316-351.
(7) Horton, D.; Priebe, W. Carbohydr. Res. 1981, 94, 27-41.
(8) (a) Dondoni, A.; Giovannini, P. P. Synthesis 2002, 1701-1706.
(b) Dondoni, A.; Scherrmann, M.-C. J . Org. Chem. 1994, 59, 6404-
6412. (c) Dondoni, A.; Narra, A.; Scherrmann, M.-C. Tetrahedron Lett.
1993, 34, 7323-7326. See also: Zhdanov, Y. A.; Alekseeva, V. G.;
Fomina, V. N. Dokl. Akad. Nauk SSR 1973, 212, 99-100.
10.1021/jo0350256 CCC: $25.00 © 2003 American Chemical Society
Published on Web 11/12/2003
10152
J . Org. Chem. 2003, 68, 10152-10155

F IGURE 1. Possible disconnection for C-disaccharide synthesis and the target compound (1).
SCHEME 1 a
SCHEME 2
by chromatography and drying under vacuum the resi-
due, which originally contained R- and â-lactols, solidified
and was found to now only be one anomer. Although the
orientation of the dithiane group relative to the furanose
ring could not be determined directly from NMR data it
is reasonable to assume that the major lactol is the
anomer with the OH and the ring O-2 trans to each other.
With the less expensive (PhS)2CHLi as the nucleophile,
lactone 2 reacted cleanly to afford one lactol product,
isolated as a colorless syrup in 83% yield, thought to be
the â-anomer 5. Again the stereochemical identity of 5
is not certain.
a
Reagents and conditions: (a) PhLi, THF, -78 °C, then aq
NH4Cl. (b) 2-Lithio-1,3-dithiane, THF, -78 °C, then aq NH4Cl.
(c) (PhS)2CHLi, THF, -78 °C, then aq NH4Cl.
It is the reasonable to assume that the â-lactols 3, 4,
and 5 are the thermodynamic products of addition to
lactone 2; however, we investigated the addition of
lithiated dithianes to lactone 2 further by attempting to
form acetates of the lactol products. Quenching the
reaction mixtures with Ac2O would possibly allow for the
trapping of R- and â-lactols before the mixture reached
equilibrium and then the use of chemical shifts from the
acetate 13C spectra to aid in designating anomeric con-
figuration as has been used previously.9 First, when
lactols 4 and 5 were isolated by aqueous workup, and
then subjected to standard acetylation conditions with
Ac2O and pyridine, the only products isolated proved to
be the acetylated ketene dithioacetals 7 and 8, formed
in 98% and 85% yields, respectively (Scheme 2). These
compounds must arise from ring opening of the lactols
under the reaction conditions and enolization at C-1 and
C-2 followed by acetylation. The isolation of 8 proved
useful in subsequent experiments in which the reaction
between (PhS)2CHLi and lactone 2 was quenched at
different time intervals with Ac2O in an attempt to trap
any intermediate lactol alkoxides. When the reaction
mixtures were quenched at low temperature after 5 min,
F IGURE 2. D-Mannofuranose-derived lactol acetate 6.
We chose to begin our studies toward C-disaccharide
1 by adding phenyllithium to commercially available 2,3-
O-isopropylidene-D-erythronolactone (2) to learn about
the stereochemical preferences of a ketol formed from this
system and the possibility of stereoselectivity in the
subsequent ketol reduction. Thus, treatment of 2 with
PhLi at -78 °C resulted in the clean formation of a single
product, 3 (Scheme 1), as a solid in 86% yield. The stereo-
chemical identity of 3 is not obvious from its NMR spectra
and is assumed to be the â-anomer by comparison with
a similar compound produced by Dondoni and colleagues
from 2,3;5,6-di-O-isopropylidene-D-mannofuranolactone8b
that was found to equilibrate with its R-anomer in
solution. The anomeric D-mannofuranosyl lactols in that
case were trapped as the respective acetates and the
identity of the thermodynamic R-anomer (6) was proven
by X-ray crystallography.8b In the present work com-
pound 3 does not apparently undergo measurable ano-
merization and is assumed to be the thermodynamic
lactol with the anomeric OH and O-2 of the furanose ring
trans as in the closely related 6 (Figure 2).
1
30 min, and 1 h, H NMR spectra of the crude material
showed at least three components in each, one of which
could be positively identified as the enol diacetate 8.
When the reaction mixture was allowed to warm to room
temperature over 2 h and then quenched with Ac2O, 8
proved to be the main component of the mixture. We were
unable to resolve these mixtures adequately for absolute
confirmation of structure of each component and so the
acetylation experiments were not useful in ascertaining
We next moved to lithiated dithianes as the nucleophile
and found that reaction of lactone 2 with 2-lithio-1,3-
dithiane gave a 5:1 inseparable mixture of two com-
pounds as a syrup in 77% yield based on 1,3-dithiane. It
was found that using an excess of lactone led to the
cleanest product mixtures and since the 13C NMR spec-
trum of the syrup contained no carbonyl signal the ring-
closed lactols 4 are most likely formed. Upon purification
(9) Ohrui, H.; J ones, G. H.; Moffatt, J . G.; Maddox, M. L.; Chris-
tensen, A. T.; Byram, S. K. J . Am. Chem. Soc. 1975, 97, 4602-4613.
J . Org. Chem, Vol. 68, No. 26, 2003 10153

SCHEME 3 a
SCHEME 4 a
a
Reagents and conditions: (a) (CF3SO2)2O, pyridine, CH2Cl2.
(b) (PhS)2CHLi, THF, -78 °C.
SCHEME 5 a
a
Reagents and conditions: (a) Et3SiH, BF3‚OEt2, CH2Cl2, -78
°C.
the exact structure of the lactols. That the acetylation of
the 1,3-dithianyl lactols 4 gave only one diacetate (7) is
evidence that the two components of that lactol mixture
are most likely epimers.
To learn if reduction of lactols 3-5 would be stereo-
selective, each was treated with Et3SiH and BF3‚OEt2
at low temperature. Reduction of 3 resulted in the
formation of two compounds in approximately 3:1 ratio,
the major one of which (9, 33% yield, Scheme 3) was
isolated cleanly by chromatography and then crystallized
in a form suitable for analysis by X-ray. The structure10
clearly shows the major reduction compound to be the
stereoisomer that arises from delivery of the hydride from
above the furanose ring, presumably to avoid steric
interaction of Et3SiH with the bulky isopropylidene group
below the ring. The 1H signal assignments for 9 were
proven from a COSY spectrum and the coupling constant
for H-1 to H-2 (numbering shown in 9, Scheme 3) was
found to be ∼3.7 Hz. Comparing the rest of the spectrum
to that of the precursor 3 revealed a general upfield shift
of the signals for H-2, H-3, H-4, and H-4′, as would be
expected for replacement of OH with H.
Reduction of lactol 4 under the same conditions gener-
ated one major C-glycoside (10) in 84% yield, which was
isolated cleanly by chromatography and crystallized for
X-ray analysis. As with phenyl C-glycoside 9, compound
10 was found to have the exocyclic group orientated cis
to O-2 of the furanose, a consequence of hydride delivery
from above the sugar ring. The X-ray structure of 10
clearly shows this orientation.10 The reduction of lactol
5 with Et3SiH and BF3‚OEt2 produced an inseparable
mixture of two compounds in a 2 to 1 ratio and in 82%
yield. The structure of the major component was thought
to be 11. Since the lactols reduced here gave mainly “cis”
C-glycosides, in which the exocylic substituent (Ph in 9;
-2-(1,3-dithianyl) in 10; -CH(SPh)2 in 11) and O-2 of
the furanose ring are cis to each other, it was expected
that application of this chemistry in the synthesis of
C-disaccharide 1 would result in a similar stereochemical
outcome.
a
Reagents and conditions: (a) n-BuLi, THF, -78 °C, then 3
equiv of 2. (b) Ac2O, pyridine. (c) Raney Ni, EtOH, H2O, rt. (d)
Et3SiH, BF3‚OEt2, -78 °C.
derived dithiane nucleophile since this compound is
available in only one step from D-ribose on large scale11
and has only O-5 unprotected and thus available for
manipulation. Activating 12 as the triflate 13 (92% yield)
allowed for smooth displacement with LiCH(SPh)2 to
afford dithioacetal 14 in 63% yield as a syrup (Scheme
4). When 14 was deprotonated with n-BuLi and treated
with lactone 2 at low temperature (Scheme 5) it was
found that the yield of the coupling product 15 could be
maximized by using an excess of the lactone with the
ketol subsequently being isolated in 77% yield by chro-
matography. H and 13C NMR showed the product to be
1
a mixture of at least two compounds with a carbonyl
signal at 195 ppm revealing that some of the open chain
ketone was present. When the crude reaction mixture
was acetylated after workup with Ac2O in pyridine, the
only acetate formed proved to be the acyclic ketone 16,
which was isolated as a colorless syrup in 87% yield.
As with the simple lactols 3-5 above, we were unable
to trap 15 as a closed-ring acetate that would have been
suitable for reduction and we had difficulty trying to
reduce 15 directly with Et3SiH and BF3‚OEt2, ending up
instead with a complex mixture of products. We therefore
chose to remove the SPh groups first using Raney nickel,
and lactol 17 was isolated as a syrup that contained two
components in ∼4:1 ratio. Since there was no carbonyl
signal present in the 13C spectrum, the two compounds
are thought to be R- and â-lactols with the latter being
major for reasons detailed earlier. With 17 in hand it was
straightforward to reduce the lactol, which was gratify-
ingly found to occur with a high degree of stereocontrol
Methyl 2,3-O-isopropylidene â-D-ribofuranoside (12,
Scheme 4) was chosen as the precursor for the sugar-
(10) X-ray structures of compounds 9 and 10 are found in the
Supporting Information.
(11) Leonard, N. J .; Carraway, K. L. J . Heterocycl. Chem. 1966, 3,
485-489.
10154 J . Org. Chem., Vol. 68, No. 26, 2003

State University is acknowledged for funding through
the PACER initiative. The National Science Foundation
is thanked for funding the Research Experience for
Undergraduates site at Youngstown State (NSF REU-
CHE 0097682) and for funds used to purchase the YSU
diffractometer used in this work (NSF DUE-CCLI-A&I
0087210), which was supplemented by the College of
Arts and Sciences at YSU. Penny Miner, Ray Hoff, and
Dr. Bruce Levinson are thanked for collecting mass
spectra.
F IGURE 3. Numbering scheme for C-glycoside 1.
to afford the C-disaccharide 1 in 51% isolated yield. A
NOESY spectrum of 1 (see Supporting Information) was
complicated by the signals for H1a and H4a (see Figure 3
for numbering scheme) overlapping in the 1-D proton
spectrum. The structure of 1 as a “cis” 1,2-C-glycoside
follows from the stereochemical course of the earlier
reductions to form C-glycosides 9 and 10.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for all new compounds; 1H and 13C NMR spectra for
compounds 1, 3, 4, 5, 7, 8, 9, 10, 11, 14, 15, 16, 17; COSY and
NOESY spectra for compound 1; X-ray structures of com-
pounds 9 and 10. This material is available free of charge via
the Internet at http://pubs.acs.org.
Ackn owledgm en t. Research Corporation is thanked
for a Cottrell College Science Award and Youngstown
J O0350256
J . Org. Chem, Vol. 68, No. 26, 2003 10155