Communications
DOI: 10.1002/anie.201006035
Oligosaccharides
The Total Synthesis of the Neurogenic Ganglioside LLG-3 Isolated
from the Starfish Linckia laevigata**
Hideki Tamai, Hiromune Ando,* Hide-Nori Tanaka, Ritsuko Hosoda-Yabe, Tomio Yabe,
Hideharu Ishida, and Makoto Kiso*
Recently, echinodermatous gangliosides have attracted much
attention because of their characteristic structure and their
potent neurogenic activity towards neuron-like rat adrenal
pheochromocytoma (PC-12) cells in the presence of the nerve
growth factor (NGF); this activity can greatly exceed that of
the mammalian ganglioside GM1.[1] Therefore, it is of great
importance to cultivate their potential as drug leads for
combating neurological disorders at the molecular level.
However, the extensive structural diversity in echinoderma-
tous gangliosides has impeded studies into their structure–
activity relationships and the mechanism by which they
promote neurogenesis. Their structure characteristically con-
tains a range of modified sialic acid units connected by a
variety of linker groups, and such units are never seen in
mammalian gangliosides.[2] Previously, we carried out the first
synthesis of two of these structures, namely, Neu5Gca-
(2,4)Neu5Ac and 8-O-SO3H-Neu5Aca(2,8)Neu5Ac, by
developing the N-Troc sialyl (Neu5Troc) donor as a key
intermediate to modified sialic acid residues.[3] Recently, we
were the first to synthesize the neurogenic ganglioside HLG-
2, which is found in the sea cucumber.[4] Herein we report the
first total synthesis of ganglioside LLG-3 (1; Scheme 1),[5]
which contains the 8-O-Me-Neu5Aca(2,11)Neu5Ac struc-
ture, by using Neu5Troc chemistry combined with the glucosyl
ceramide (Glc-Cer) cassette approach.
NGF) (20.6 Æ 2.2)% of PC-12 cells were affected. Ganglio-
side 1 is the second most potent enchinoderm ganglioside
amongst the fifteen that have been examined. The most
potent activity [(64.8 Æ 7.6)%] was exhibited by the gangli-
oside SJG-2,[1,7] which is a heptasaccharide and contains a
trisialyl tetrasaccharide unit. Therefore, even though 1 is a
tetrasaccharide, it is considered to be more effective than
SJG-2 and has attracted much attention in the field of
medicinal chemistry.
The characteristic structure of modified sialic acids (8-O-
Me-Neu5Aca(2,11)Neu5Gc) that is presumably responsible
for the neurogenic activity, was expected to be efficiently
synthesized from the common Neu5Troc donor
4
(Scheme 1).[8] Thus, it was envisioned that the synthesis of
the 5-NH2-Neu intermediate and the 8-O-methyl-Neu glyco-
late derivative could be achieved from 4 through the cleavage
of the Troc group with subsequent migration of the acetyl
group from O8 to N5. To avoid the loss of the glycan portion
as a result of inefficient coupling with the lipid portion we
anticipated using the glycosyl ceramide cassette approach,
which was recently established in our laboratory.[9] Therefore,
1 was disconnected into the trisaccharide segment 2 and the
Glc-Cer cassette 3. Then, 3 was fragmented into the glucosyl
donor 5 and the phytoceramide 6 as an aglycon.
As depicted in Scheme 2, the synthesis of the 8-O-methyl
sialyl unit from 4 commenced with the stereoselective
incorporation of benzylglycolate at the anomeric position to
give 8[5b] in a 91% yield (a/b = 7.5:1). Next, 8 was treated with
zinc under acidic conditions to produce the 8-hydroxy-N-
acetyl sialyl derivative 9, through the migration of the acetyl
group from O8 to N5, in a high yield. The 8-O-methylation of
9 was attempted using two different methylating reagents
(Me3OBF4 and MeOTf) but each attempt was unsuccessful,
and only generated a complex mixture of 5-methylimidate
derivatives and N-acetyl-N-methyl derivatives. Therefore, we
masked the C8 hydroxy group with a chloroacetyl group
[(ClCH2CO)2O, cat. DMAP, THF],[10] and then protected the
C5 acetamide with an acetyl group (isopropenyl acetate,
TsOH),[11] thereby obtaining 11 in a high yield (93% over two
steps). Compound 11 was then subjected to a two-step
sequence for 8-O-methylation. 1-Selenocarbamoylpiperidine
(12), which was recently developed by our research group,[12]
was found to be the best reagent for the selective dechloro-
acetylation at C8. The chloroacetyl group could also be
removed by using the widely used DABCO method[13]
(DABCO, EtOH), however this strategy promoted the
migration of the acetyl group from N5 to O8, giving the
fully acetylated derivative as the major product. Removal of
the chloroacetyl group using 12 and 2,6-lutidine as an acid
Ganglioside LLG-3 (1) was identified in the starfish
Linckia laevigata by Higuchi and co-workers (Scheme 1).[6]
They revealed that at a concentration of 10 mm LLG-3 caused
neurogenesis of (63.1 Æ 6.3)% of PC-12 cells in the presence
of NGF (5 ngmLÀ1), whereas in the control experiment (only
[*] H. Tamai, Dr. H. Ando, Dr. H.-N. Tanaka, R. Hosoda-Yabe, Dr. T. Yabe,
Dr. H. Ishida, Dr. M. Kiso
Department of Applied Bioorganic Chemistry, Gifu University
1-1 Yanagido, Gifu-shi, Gifu 501-1193 (Japan)
Fax: (+81)58-293-3452
hando@ gifu-u.ac.jp
H. Tamai, Dr. H. Ando, Dr. H.-N. Tanaka, R. Hosoda-Yabe,
Dr. M. Kiso
Department of Applied Bioorganic Chemistry
Institute for Integrated Cell-Material Sciences (WPI program)
Kyoto University
Yoshida Ushinomiya-cho, Sakyo-ku, Kyoto 606-8501 (Japan)
[**] Part 152 of the series: Synthetic studies on sialoglycoconjugates.
This work was financially supported by MEXT of Japan (WPI
program and grant-in-aid for Scientific Research to M.K., No.
1701007 and No. 22380067). We thank Kiyoko Ito for technical
assistance.
Supporting information for this article is available on the WWW
2330
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 2330 –2333