Efficient, one-pot syntheses of biologically active α-linked glycolipids

Article abstract of DOI:10.1039/b702551c

Per-O-silylated galactosyl iodides undergo α-glycosidation with fully functionalized glycolipids producing biologically relevant conjugates. The Royal Society of Chemistry.

Full text of DOI:10.1039/b702551c

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Efficient, one-pot syntheses of biologically active a-linked
glycolipids{{
Wenjun Du, Suvarn S. Kulkarni and Jacquelyn Gervay-Hague*
Received (in Bloomington, IN, USA) 19th February 2007, Accepted 26th April 2007
First published as an Advance Article on the web 17th May 2007
DOI: 10.1039/b702551c
synthetic efficiency. Moreover, commonly employed glycosyl
Per-O-silylated galactosyl iodides undergo a-glycosidation with
fully functionalized glycolipids producing biologically relevant
conjugates.
fluorides14–17 and trichloroacetimidates16,18–20 typically afford
moderate yields (30–60%) and purification of the products can
be complicated by the formation of a/b mixtures. The use of other
glycosyl donors such as bromides,21 thiogalactosides,19 and
phosphites22 has also been attempted but without significant
improvement.
In 1993, six novel galactosyl ceramides with unique a-glycosidic
linkages were isolated from the marine sponge Agelas mauritianus
near Okinawa, Japan.1,2 These compounds showed highly potent
anti-tumor activities, which prompted various synthetic studies.3
Among the many analogues synthesized, KRN7000 (Fig. 1) was
found to be the most potent4 and extensive mechanistic studies
indicated that the anti-tumor activity resulted from CD1d-
dependent natural killer T-cell (NKT) stimulation.4,5
Recent studies in glycosyl iodide chemistry have led to
important advances in achieving high stereoselectivity, but there
is still room for improvement.23 For example, reactions of per-
O-benzylated galactosyl iodide with an azido sphingosine under
in situ anomerization conditions exclusively produce the a-anomer
in over 90% yield (Fig. 2). An azido group is used in place of the
amide because, if left intact during the glycosylation, the amide
deactivates the primary hydroxyl of the acceptor through
unfavorable hydrogen bonding interactions.24,25 While azide
replacement offers a viable solution to this problem, it also
requires four additional steps to convert the amine of sphingosine
into an azide and, after glycosidation is complete, the azide must
then be reduced and conjugated to the fatty acid of interest.
An attractive alternative strategy employs fully functionalized
ceramide acceptors, avoiding tedious protection and deprotection
steps. Here, we report a streamlined synthesis of bioactive
glycolipids using a transiently protected galactosyl iodide donor.
The use of per-O-silylated galactosyl iodide allows direct
glycosidation of fully functionalized lipid components without
the need for any protection, providing what is essentially a one-pot
synthesis of a-GalCer analogs.
CD1d is a member of the CD1 family of proteins, which present
lipid antigens to NKT cells to activate the immune response. The
current model of activation suggests that CD1d recognizes
KRN7000 and the binding complex interacts with the T-cell
receptor (TCR) of NKT cells, stimulating the release of two major
cytokines known as IFN-c and IL-4.6–8 These cytokines are
members of the T helper 1 (Th1) and T helper 2 (Th2) cytokine
families, respectively. Studies suggest that cytokine production can
be tuned by altering the ceramide structure, offering the
opportunity to elicit selective Th1 and Th2 secretion.9–12 For
example, Wong and co-workers recently identified an aryl
containing fatty acyl chain analog of a-galactosyl ceramide that
increases IFN-c secretion relative to IL-4, and this compound has
become a prototype for adjuvant development.13
Numerous efforts have been invested in the syntheses of
a-GalCer analogs to gain access to these biologically important
compounds in pure form for biological and biomedical studies.
Arguably, the biggest hurdle in the synthesis is the glycosylation
reaction. Achieving chemo- and stereoselectivity requires multi-
step protections and deprotections, which consequently lower
Several attempts to directly incorporate ceramide acceptors have
been reported. The nature of the protecting group on the ceramide
moiety has significant impact upon the outcome. For example,
Wang and co-workers reported that a-selectivity is achieved with
O-benzoyl protecting groups, whereas b-selectivity is observed with
O-benzyl protecting groups.18 Similarly, TBS-protected ceramide
Fig. 1 Analogs of naturally occurring a-GalCer from Agelas mauritianus.
Department of Chemistry, University of California Davis, One Shields
Avenue, Davis, CA, 95616, USA. E-mail: gervay@chem.ucdavis.edu;
Fax: +1 530 754 6915; Tel: +1 530 754 9557
{ Dedicated to Professor Michael E. Jung on the occasion of his 60th
birthday.
{ Electronic supplementary information (ESI) available: Experimental
details and characterization data of all the new compounds. See DOI:
10.1039/b702551c
Fig. 2 First generation synthesis requires multiple protection–
deprotection steps.
2336 | Chem. Commun., 2007, 2336–2338
This journal is ß The Royal Society of Chemistry 2007

View Article Online
acceptors afford coupled products in moderate yields and
selectivities depending upon the donor.16 We were primarily
interested in utilizing unprotected acceptors but the only example
reported in the literature involved tin activation of a glycosyl
fluoride, yielding a complex mixture and poor yields of the desired
compound.14
Glycolipids 1–3 (Fig. 3) were targeted in order to test the
generality of our strategy. Compound 1a (BbGL-II) was recently
isolated from a bacterial source and has led to speculation that the
glycolipids isolated from marine sponge may originate from
bacteria.26 Compound 1c is a saturated analog of 1a. Recent SAR
studies indicate that 1b, with a reversed arrangement of fatty acid
chains, exhibits the most potent immunogenic activity amongst the
BbGL-II analogs.27 Compounds 2 and 3 are analogs of KRN7000
with shorter ceramide lipids and differences in unsaturation. As
mentioned earlier, lipid length can bias cytokine production,
leading to improved clinical results with lower toxicity in certain
disease treatments.9,11 The most direct synthetic route to these
targets would employ fully functionalized lipid components.
Although the ceramide acceptors would be deactivated, we were
hopeful that a highly reactive glycosyl iodide could overcome this
barrier and that a combination of sterics and donor reactivity
would make selective protection of the acceptor alcohols
unnecessary.
Scheme 1 Synthesis of target compounds.
three-step, one-pot procedure are shown in Table 1. It is notable
that this methodology tolerates unsaturation in the fatty acid side-
chains; in contrast to benzyl-protected donors, which require
hydrogenation for deprotection. To illustrate the point, compound
1a was readily converted to
a saturated analog (1c) by
hydrogenation of the double bond.
Reactions with the glyceride acceptors 6 and 730 were more
facile than with the ceramide acceptors. This is likely due to the
fact that 6 and 7 do not possess the unfavorable hydrogen bonding
interactions that plague 8 and 9. Although all the reactions were
a-selective, some b-product was observed with acceptor 6 when
using benzene as the solvent. A simple solution to this problem
was provided by conducting the reaction at room temperature and
switching the solvent to CH2Cl2. In a similar fashion, glyceride 7
reacted to give 1b in 72% yield (entry 3) and ceramide 8 afforded 2
(77%) uneventfully (entry 4).
From our previous studies, we knew that per-O-benzyl
galactosyl iodide would not undergo glycosylation with ceramide
acceptors, but recent investigations in our lab suggested that per-
O-silylated glycosyl iodides are orders of magnitude more
reactive.23 There are several other advantages to engaging per-
O-silylated donors in synthetic protocols. First, they are readily
prepared on a large scale, and the iodide can be generated
quantitatively upon reaction with trimethylsilyl iodide (TMSI).28
Finally, deprotection of the trimethylsilyl group requires only mild
acidic conditions, which does not affect the glycosidic linkage.
Taken together, the prospects for being able to achieve a highly
convergent synthesis of a-Gal analogs appeared favorable.
The synthesis began with 4, which was reacted with TMSI to
give galactosyl iodide 5. The donor was then added to the
acceptor, which was premixed with tetrabutylammonium iodide, a
promoter that accelerates the reaction and determines a-stereo-
selectivity through in situ anomerization.29 After the indicated
time, the solvent was evaporated and the remaining residue was
subjected to hydrolysis using acidic resin in methanol to give
the final products 1a, 1b, 2 and 3 (Scheme 1). The results of the
More problematic was acceptor 9,16 which initially gave only
low yields (20–30%, entries 5 and 6). We attempted various
reaction modifications including changing solvent, reaction
temperature and ratios of donor to acceptor. Despite all these
attempts, significant improvements were not forthcoming. We did
observe solubility differences between 8 and 9, which led us to
consider micelle formation as a contributing factor. Since the side
chain of 9 is saturated, it is able to pack better than 8. At this
point, we attempted microwave assisted glycosylation between 5
and 9 (Table 2). Under low power (25–35 W), the yield was not
significantly improved (entries 1–4), although the acceptor
appeared to be soluble under these conditions. Increasing the
reaction temperature did not improve the yield either. However,
much to our delight, a 67% yield of 3 was obtained when
performing the reaction at 120 uC for 1.5 h at 225 W (entry 5). We
do not fully understand the reasons for the improvement but it is
reasonable to suggest that microwave energy may disturb lipid
packing making the acceptor more accessible for glycosidation.
Table 1 Results of one-pot synthesis
Entry Acceptor Product Solvent Conditions a : b ratio (yield)
1
2
3
4
5
6
6
6
7
8
9
9
1a
1a
1b
2
3
3
Benzene 65 uC, 24 h 10 : 1 (89%)
CH2Cl2 rt, 24 h
CH2Cl2 rt, 36 h
CH2Cl2 rt, 48 h
a only (81%)
a only (72%)
a only (77%)
Benzene 65 uC, 48 h a only (20%)
CH2Cl2 rt, 48 h a only (30%)
Fig. 3 Targeted bioactive glycolipids.
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 2336–2338 | 2337

View Article Online
7 S. A. Porcelli and R. L. Modlin, Annu. Rev. Immunol., 1999, 17, 297.
Table 2 Microwave results for the reaction between 5 and 9
8 Y. Kinjo, D. Wu, G. Kim, G. W. Xing, M. A. Poles, D. D. Ho,
M. Tsuji, K. Kawahara, C.-H. Wong and M. Kronenberg, Nature,
2005, 434, 520.
Entry Donor Solvent Reaction conditions Yield (%) a : b ratio
1
2
3
4
5
3 eq. Benzene 80 uC, 1.5 h, 25 W
3 eq. CH2Cl2 80 uC, 1.5 h, 30 W
1 eq. CH2Cl2 100 uC, 1.5 h, 35 W 30
1 eq. CH2Cl2 120 uC, 1.5 h, 37 W 37
1 eq. CH2Cl2 120 uC, 1.5 h, 225 W 67
20
30
a only
a only
a only
a only
a only
9 E. Pal, T. Tabira, T. Kawano, M. Taniguchi, S. Miyake and
T. Yamamura, J. Immunol., 2001, 166, 662.
10 C. R. Berkers and H. Ovaa, Trends Pharmacol. Sci., 2005, 26, 252.
11 K. Miyamoto, S. Miyake and T. Yamamura, Nature, 2001, 413, 531.
12 G. Yang, J. Schmieg, M. Tsuji and R. W. Franck, Angew. Chem., Int.
Ed., 2004, 43, 3818.
13 M. Fujio, D. Wu, R. Garcia-Navarro, D. D. Ho, M. Tsuji and
C.-H. Wong, J. Am. Chem. Soc., 2006, 128, 9022.
In summary, we report highly efficient, one-pot syntheses of
14 T. Sakai, H. Ueno, T. Natori, A. Uchimura, K. Motoki and
Y. Koezuka, J. Med. Chem., 1998, 41, 650.
15 R. M. Ndonye, D. P. Izmirian, M. F. Dunn, K. O. Yu, S. A. Porcelli,
A. Khurana, M. Kronenberg, S. K. Richardson and A. R. Howell,
J. Org. Chem., 2005, 70, 10260.
16 S. Kim, S. Song, T. Lee, S. Jung and D. Kim, Synthesis, 2004, 847.
17 M. Morita, K. Motoki, K. Akimoto, T. Natori, T. Sakai, E. Sawa,
K. Yamaji, Y. Koezuka, E. Kobayashi and H. Fukushima, J. Med.
Chem., 1995, 38, 2176.
18 C. Xia, Q. Yao, J. Schumann, E. Rossy, W. Chen, L. Zhu, W. Zhang,
G. De Libero and P. G. Wang, Bioorg. Med. Chem. Lett., 2006, 16,
2195.
biologically active a-GalCer and BbGL-II analogs using a
combination of a TMS-protected glycosyl iodide donor and an
unprotected ceramide or glyceride acceptor. The method is highly
efficient, yielding exclusively the a-isomer and providing rapid
access to biologically relevant compounds of wide interest. We also
demonstrate that microwave energy can significantly improve
reaction yields and shorten the reaction time when saturated
ceramides are utilized. Further application of these methods to
branched oligosaccharides is currently under study in our
laboratory.
19 O. Plettenburg, V. Bodmer-Narkevitch and C.-H. Wong, J. Org. Chem.,
2002, 67, 4559.
This work is supported by the National Science Foundation
CHE-0210807. NSF CRIF program (CHE-9808183), NSF Grant
OSTI 97-24412, and NIH Grant RR11973 provided funding for
the NMR spectrometers used on this project. W. D. would like to
acknowledge a fellowship from the Miller Foundation, which
partially funded this work.
20 S. Figueroa-Perez and R. R. Schmidt, Carbohydr. Res., 2000, 328, 95.
21 R. D. Goff, Y. Gao, J. Mattner, D. Zhou, N. Yin, C. Cantu, L. Teyton,
A. Bendelac and P. B. Savage, J. Am. Chem. Soc., 2004, 126, 13602.
22 S. Y. Luo, S. S. Kulkarni, C. H. Chou, W. M. Liao and S. C. Hung,
J. Org. Chem., 2006, 71, 1226.
23 W. Du and J. Gervay-Hague, Org. Lett., 2005, 7, 2063.
24 R. Polt, L. Szabo, J. Treiberg, Y. Li and V. J. Hruby, J. Am. Chem.
Soc., 1992, 114, 10249.
Notes and references
25 R. R. Schmidt and P. Zimmermann, Angew. Chem., Int. Ed. Engl., 1986,
25, 725.
1 T. Natori, Y. Koezuka and T. Higa, Tetrahedron Lett., 1993, 34, 5591.
2 T. Natori, M. Morita, K. Akimoto and Y. Koezuka, Tetrahedron, 1994,
50, 2771.
3 K. Motoki, E. Kobayashi, T. Uchida, H. Fukushima and Y. Koezuka,
Bioorg. Med. Chem. Lett., 1995, 5, 705.
4 E. Kobayashi, K. Motoki, Y. Yamaguchi, T. Uchida, H. Fukushima
and Y. Koezuka, Bioorg. Med. Chem., 1996, 4, 615.
5 T. Kawano, J. Cui, Y. Koezuka, I. Toura, Y. Kaneko, K. Motoki,
H. Ueno, R. Nakagawa, H. Sato, E. Kondo, H. Koseki and
M. Taniguchi, Science, 1997, 278, 1626.
6 D. Wu, G. W. Xing, M. A. Poles, A. Horowitz, Y. Kinjo, B. Sullivan,
V. Bodmer-Narkevitch, O. Plettenburg, M. Kronenberg, M. Tsuji,
D. D. Ho and C.-H. Wong, Proc. Natl. Acad. Sci. U. S. A., 2005, 102,
1351.
26 G. Ben-Menachem, J. Kubler-Kielb, B. Coxon, A. Yergey and
R. Schneerson, Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 7913.
27 Y. Kinjo, E. Tupin, D. Wu, M. Fujio, R. Garcia-Navarro,
M. R. Benhnia, D. M. Zajonc, G. Ben-Menachem, G. D. Ainge,
G. F. Painter, A. Khurana, K. Hoebe, S. M. Behar, B. Beutler,
I. A. Wilson, M. Tsuji, T. J. Sellati, C.-H. Wong and M. Kronenberg,
Nat. Immunol., 2006, 7, 978.
28 A. S. Bhat and J. Gervay-Hague, Org. Lett., 2001, 3, 2081.
29 S. N. Lam and J. Gervay-Hague, Carbohydr. Res., 2002, 337, 1953.
30 Acceptor 6 is commercially available, 7 is prepared from (S)-(+)-2,2-
dimethyl-1,3-dioxolane-4-methanol in five steps, and 8 and 9 are
prepared from sphingosine and phytosphingosine, respectively, through
amide bond formation with stearic acid (see supporting information{).
2338 | Chem. Commun., 2007, 2336–2338
This journal is ß The Royal Society of Chemistry 2007