Stereoselective synthesis of 2-deoxy-beta-galactosides via 2-deoxy-2-bromo- and 2-deoxy-2-iodo-galactopyranosyl donors.

Article abstract of DOI:10.1021/ol034393t

[reaction: see text] A series of 2-bromo- and 2-iodo-galactopyranosyl acetates and trichloroacetimidates were evaluated as glycosyl donors for the synthesis of 2-deoxygalactopyranosides. The best selectivity for the beta-glycosidic linkage was achieved by

Full text of DOI:10.1021/ol034393t

ORGANIC
LETTERS
2003
Vol. 5, No. 11
1871-1874
Stereoselective Synthesis of
2-Deoxy-â-Galactosides via
2-Deoxy-2-bromo- and
2-Deoxy-2-iodo-galactopyranosyl Donors
Timothy B. Durham and William R. Roush*
Department of Chemistry, UniVersity of Michigan, Ann Arbor, Michigan 48109
roush@umich.edu
Received March 6, 2003
ABSTRACT
A series of 2-bromo- and 2-iodo-galactopyranosyl acetates and trichloroacetimidates were evaluated as glycosyl donors for the synthesis of
2-deoxygalactopyranosides. The best selectivity for the â-glycosidic linkage was achieved by using 6-deoxy-3,4-carbonate-protected galactosyl
donors.
2-Deoxy carbohydrates are structurally important components
of numerous biologically active natural products.1 The ability
to control the stereochemistry of 2-deoxy glycosidic linkages
represents an important and challenging synthetic problem.2,3
Recent reports from these laboratories and others have
demonstrated the use of 2-deoxy-2-iodo- and 2-deoxy-2-
bromo-glucopyranosyl acetates,4-8 trichloroacetimidates,9,10
and fluorides11 as highly diastereoselective glycosidating
agents for the synthesis of 2-deoxy-â-glucosides (and also
of 2-deoxy-R-glucosides).4,6 The utility of these procedures
has been demonstrated in syntheses of the landomycin A
hexasaccharide12 and olivomycin A.13
Given our continued interest in 2-deoxy glycosides, we
wanted to extend our efforts to the synthesis of 2-deoxy-â-
galactosides. Prior to this study, we were aware of one
published example of a â-selective glycosidation reaction
of a 2-deoxy-2-bromo-galactopyranosyl acetate, which pro-
ceeded with excellent selectivity.8 Other work, however,
indicated that the â-selectivity of glycosidation reactions of
2-thioaryl-2-deoxygalacto-pyranosyl donors is lower than that
of corresponding donors in the glucopyranosyl series.14
Therefore, we decided to evaluate the utility of 2-deoxy-2-
halogalactopyranosyl acetates and trichloroacetimidates for
the synthesis of 2-deoxy-â-galactosides.
(1) Kirschning, A.; Bechthold, A. F.-W.; Rohr, J. Top. Curr. Chem. 1997,
188, 1-84.
(2) Veyrie`res, A. In Carbohydrates in Chemistry and Biology; Ernst,
B., Hart, G. W., Sinay¨, P., Eds.; Wiley-VCH: Weinheim, Germany, 2000;
Vol. 1, pp 368-405.
(3) Marzabadi, C. H.; Franck, R. W. Tetrahedron 2000, 56, 8385-8417.
(4) Roush, W. R.; Briner, K.; Sebesta, D. P. Synlett 1993, 264-266.
(5) Roush, W. R.; Bennett, C. E. J. Am. Chem. Soc. 1999, 121, 3541-
3542.
2,6-Dideoxy-â-galactoside units are found in many of the
aureolic acid antibiotics, including mithramycin15 and
UCH9,16,17 as well as the recently isolated durhamycins.
(6) Roush, W. R.; Narayan, S. Org. Lett. 1999, 1, 899-902.
(7) Kirschning, A.; Plumeier, C.; Rose, L. Chem. Commun. 1998, 33-
34.
(8) Hashem, M. A.; Jung, A.; Ries, M.; Kirschning, A. Synlett 1998,
195-197.
(9) Roush, W. R.; Gung, B. W.; Bennett, C. E. Org. Lett. 1999, 1, 891-
893.
(12) Roush, W. R.; Bennett, C. E. J. Am. Chem. Soc. 2000, 122, 6124-
6125.
(13) Roush, W. R.; Hartz, R. A.; Gustin, D. J. J. Am. Chem. Soc. 1999,
121, 1990-1991.
(14) Hashimoto, S.-i.; Yanagiya, Y.; Honda, T.; Ikegami, S. Chem. Lett.
1992, 1511-1514.
(10) Chong, P. Y.; Roush, W. R. Org. Lett. 2002, 4, 4523-4526.
(11) Blanchard, N.; Roush, W. R. Org. Lett. 2003, 5, 81-84.
(15) Wohlert, S. E.; Ku¨nzel, E.; Machinek, R.; Me´ndez, C.; Salas, J. A.;
Rohr, J. J. Nat. Prod. 1999, 62, 119-121.
10.1021/ol034393t CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/29/2003

the resultant hemiacetal to imidate 2 was achieved by
treatment with Cl3CCN and DBU.21,22
Scheme 1
Imidate 2 was then subjected to TBSOTf-promoted gly-
cosidation with model acceptor 7 (Scheme 2). Surprisingly,
this reaction was only modestly â-selective (60:40 â:R).23
Further, we observed that imidate 2 is highly unstable. This
led us to change the C(2) directing group to a more
electronegative bromide substituent, which we hoped would
increase the stability of the donor. We also decided to change
the C(4)-protecting group to a benzyl ether, since it seemed
possible that the axial C(4) acetate unit in 2 might participate
in the glycosidation reaction by interacting with any C(1)-
cationic intermediates, thereby decreasing the reaction ste-
reoselectivity compared to previously studied examples in
the 2-deoxy-2-halo-glucopyranosyl series.
Accordingly, donors 9 and 10 were examined in glycosi-
dation reactions with 7, as well as with the less hindered
acceptor 13 (Scheme 3). Unfortunately, neither 9 or 10
Durhamycin A (1, Scheme 1) is a potent inhibitor of HIV-1
Tat transactivation.18 The unusual biological activity of 1
and the presence of a challenging 2,6-dideoxy-â-galactosidic
linkage make it an excellent synthetic target. In this context,
the 2-deoxy-2-iodogalactopyranosyl acetate 2 (Scheme 1)
initially appeared to be an appropriate precursor of the
2-deoxy-â-galactoside unit (“D”) in 1.
Scheme 3a
Donor 2 was prepared starting from protected galactal 4
(Scheme 2).19 Displacement of the tosylate with Bu4NBr
Scheme 2
a Conditions: donor (2 equiv), acceptor (1 equiv), TBSOTf (0.1-
0.3 equiv), CH2Cl2, -78 °C.
displayed synthetically useful â-selectivity in these reactions.
Differences in the directing ability of the C(2)-Br and C(2)-I
substituents were observed in the reactions with 7, but not
with 13.
Changes to the C(6)-substituent also had a relatively minor
effect on the selectivity of the glycosidation reactions, as
the results summarized in Scheme 3 (9 and 10 with C(6)-
tosyloxy substituents) and Scheme 4 (C(6)-benzyloxy sub-
stituted donors 16 and 17) indicate. Interestingly, however,
our results for the glycosidation reaction of donor 17 and
galactoside acceptor 22 (Scheme 4) are not in agreement with
the literature report for this reaction,8 which indicated that
â-galactoside 23 was formed selectively. In our our hands,
this reaction was only moderately â-selective, in agreement
followed by treatment with NIS and AcOH provided the
desired iodo acetate 6 as a 1:1 mixture of anomers along
with talo isomer 5. After selective deprotection of the
anomeric acetate unit of 6 with hydrazine,20 conversion of
(20) Excoffier, G.; Gagnaire, D.; Utille, J.-P. Carbohydr. Res. 1975, 39,
368-373.
(16) Ogawa, H.; Yamashita, Y.; Katahira, R.; Chiba, S.; Iwasaki, T.;
Ashizawa, T.; Nakano, H. J. Antibiotics 1998, 51, 261-266.
(17) Katahira, R.; Uosaki, Y.; Ogawa, H.; Yamashita, Y.; Nakano, H.;
Yoshida, M. J. Antibiotics 1998, 51, 267-274.
(18) Jayasuriya, H.; Lingham, R. B.; Graham, P.; Quamina, D.; Herranz,
L.; Genilloud, O.; Gagliardi, M.; Danzeisen, R.; Tomassini, J. E.; Zink, D.
L.; Guan, Z.; Singh, S. B. J. Nat. Prod. 2002, 65, 1091-1095.
(19) Roush, W. R.; Lin, X.-F. J. Am. Chem. Soc. 1995, 117, 2236-
2250.
(21) Schmidt, R. R. AdV. Carbohydr. Chem. Biochem. 1994, 33, 21-
123.
(22) Schmidt, R. R.; Jung, K.-H. In Carbohydrates in Chemistry and
Biology; Ernst, B., Hart, G. W., Sinay¨, P., Eds.; Wiley-VCH: Weinheim,
Germany, 2000; Vol. 1, pp 5-59.
(23) All reported ratios were determined by 1H NMR analysis of the
crude reaction mixtures. Stereochemical assignments of all isolated products
1
were made using H, 13C, and COSY NMR analysis.
1872
Org. Lett., Vol. 5, No. 11, 2003

cosidation reactions of 2-iodoglucosyl donors.10,27 Several
groups have calculated the conformational preferences of
substituted pyranosyl oxocarbenium ions and have reported
that C(3) and C(4) oxygen substituents have a strong
preference for pseudoaxial orientation due to electronic
stabilization of the cationic center (e.g., conformation 27,
Figure 1).28-30 However, in studies with donors containing
Scheme 4a
a Conditions: TMSOTf (0.5 equiv) was added to a mixture of
donor (2 equiv) and acceptor (1 equiv), CH2Cl2, -30 to -23 °C.
with the other reactions of donors 16 and 17 (Scheme 4).
Deoxygenation of C(6) altogether, as in the case of donor
24, resulted in glycosidation reactions that were modestly
R-selective (Scheme 5). While these results indicated that
Figure 1. Stereochemical considerations.
4,6-benzylidiene protecting groups,10 high â-selectivity was
obtained even though ions such as 27 are inaccessible.31 An
implication of the latter study is that the â-selectivity might
be governed by substitution reactions of iodonium ion
intermediates rather than oxocarbenium ions. In contrast, the
results summarized in Schemes 2-5 suggest that halonium
ion intermediates are not significantly involved in the
glycosidation reactions of 2-deoxy-2-halo-galactopyranose
donors, since the large amounts of R-glycosides produced
in these reactions require that oxocarbenium ions play a
substantial role.28,32
Assuming that cationic intermediates play a greater role
in the reactions of the 2-deoxy-2-halogalactopyranosyl donors
than in the glucopyranosyl series, selectivity is then deter-
mined either by the conformational preferences of ions 30
and 31 or by the relative reactivity of these two similar
species. In contemplating strategies to increase the preference
for reaction by way of 30, we were reminded of studies by
Danishefsky, who demonstrated that the â-selectivity of
glycosidation reactions of 2,3-epoxy sugars in the galactose
series were considerably enhanced when the 3,4-hydroxyl
Scheme 5a
a Conditions: TMSOTf (0.2 equiv) was added to a mixture of
donor (2 equiv) and acceptor (1 equiv) in CH2Cl2 at 0 °C.
the C(6)-substituent on the galactopyranosyl donor can
influence the diastereoselectivity of the glycosidation reac-
tions, access to a highly â-selective 2-deoxy-2-halogalacto-
pyranosyl donor still eluded us.
In previous studies of the glycosidation reactions of
2-deoxy-2-iodoglucopyranosyl acetates and trichloroacetimi-
dates, we speculated that stereoselectivity might be governed
by the conformational preferences of intermediate pyranosyl
oxocarbenium ions.24-26 Evidence was also presented indi-
cating that glycosyl triflates are not intermediates in gly-
(27) Crich, D.; Cai, W. J. Org. Chem. 1999, 64, 4926-4930.
(28) Romero, J. A. C.; Tabacco, S. A.; Woerpel, K. A. J. Am. Chem.
Soc. 2000, 122, 168-169.
(29) Miljkovic, M.; Yeagley, D.; Deslongchamps, P.; Dory, Y. L. J. Org.
Chem. 1997, 62, 7597-7604.
(24) Roush, W. R.; Sebesta, D. P.; Bennett, C. E. Tetrahedron 1997,
53, 8825-8836.
(30) Woods, R. J.; Andrews, C. W.; Bowen, J. P. J. Am. Chem. Soc.
1992, 114, 859-864.
(25) Roush, W. R.; Sebesta, D. P.; James, R. A. Tetrahedron 1997, 53,
8837-8852.
(31) Abdel-Rahman, A. A.-H.; Jonke, S.; El Ashry, E. S. H.; Schmidt,
R. R. Angew. Chem., Int. Ed. 2002, 41, 2972-2974.
(32) Deslongchamps, P. In Stereoelectronic Effects in Organic Chemistry;
Pergamon: New York, 1983; pp 209-290.
(26) Tamura, S.; Abe, H.; Matsuda, A.; Shuto, S. Angew. Chem., Int.
Ed. 2003, 42, 1021.
Org. Lett., Vol. 5, No. 11, 2003
1873

groups were protected as a 3,4-carbonate group.33 Accord-
ingly, we targeted 32 as a key reactive intermediate. The
cis-fusion of the cyclic 3,4-protecting group encourages 32
to adopt the indicated boatlike conformation with the C(2)-X
substituent in a pseudoaxial position, which should direct
the glycosidation in a â-selective manner.31
Scheme 7a
To test this hypothesis, we studied the glycosidation
reactions of the 3,4-acetonide-protected donor 34 and also
the 3,4-carbonate-protected donors 37, 39, and 40 (Schemes
6 and 7). Donors 34 and 40 gave, in most cases, only modest
Scheme 6
selectivity in reactions with 7 and 13. Gratifyingly, 6-deoxy
donors 37 and 39 displayed good to excellent â-selectivities
with a variety of carbohydrate acceptors (7, 13, and 49). It
is noteworthy that these two donors complement each other
in their reactivity profiles such that, under identical condi-
tions, donor 39 undergoes glycosidation at -78 °C, while
donor 37 requires a significantly higher temperature (0 °C)
for reaction to occur. However, it remains unclear at present
why donors 37 and 39 display such very different selectivity
patterns in reactions with similar acceptors.
In conclusion, we have developed two 3,4-carbonate-
protected 2,6-dideoxy-2-halo-galactosyl donors (37 and 39)
that provide access to 2,6-dideoxy-â-galactosides with high
diastereoselectivity. Selectivity decreases, however, with rigid
donors 34 and 40 containing C(6)-oxygenated substituents
or when the donor lacks a cyclic 3,4-protecting group (2, 9,
10, 16, 17, and 24). Further studies into the origin of
selectivity with donors 37 and 39 are in progress. Application
of this methodology to the synthesis of the CDEF tetrasac-
charide unit of Durhamycin A (1) is reported in the following
paper in this issue.
a Conditions: (a) TMSOTf (0.2 equiv) was added to donor (2
equiv) and acceptor (1 equiv) at 0 °C, CH2Cl2; (b) TBSOTf (0.3
equiv) was added to donor (2 equiv) and acceptor (1 equiv) at -78
°C, CH2Cl2.
Acknowledgment. We gratefully acknowledge the Na-
tional Institutes of Health (GM38907) for financial support
of this research.
Supporting Information Available: Experimental pro-
tocols and characterization data for all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(33) Randolph, J. T.; Danishefsky, S. J. J. Am. Chem. Soc. 1995, 117,
5693-5700.
OL034393T
1874
Org. Lett., Vol. 5, No. 11, 2003