2610
M. Michieletti et al.
LETTER
probably due to a restricted conformational mobility of drous dichloromethane in the presence of 0.4 nm
the Boc protecting group. This behavior was partly con- molecular sieves.
served also after deprotection of the benzoate and glyco-
sylation. Basic hydrolysis of the benzoyl ester gave pure
doubly protected sphingosine (5) in 57% overall yield.
a-Mannosylceramide (a-ManCer)14 was efficiently syn-
thesized in this way starting from tetra-O-benzoyl-a-D-
mannopyranosyl bromide. The glycosylation gave com-
We chose to apply the classical Helferich glycosylation, pound 7 in 71% yield and standard deprotection and acy-
namely the glycosylation using glycosyl bromides as do- lation reactions gave pure a-ManCer (9, Scheme 2).
nors promoted by Hg(II) salts.11 Indeed, mercury salts are
Glycosylation of galactosyl bromide was performed both
very mild Lewis acids, and presumably compatible with
on acetylated and on benzoylated donors. Higher yield
our protecting groups. We performed the reaction using
and cleaner reaction were obtained using the second donor
different peracylated glycosyl bromides which were acti-
so we did not perform the reaction on other acetylated do-
vated using Hg(CN)2 [Caution: Hg(CN)2 is very toxic and
nors. It is noteworthy that, during the reaction, the initial
has to be manipulated with the necessary precautions] as
mild promoter (for experimental details see the Support-
ing Information). The use of esters as protecting groups
allows to obtain a b-glycosydic linkage, exploiting the an-
chimeric assistance of the carbonyl group in position 2 of
the donor. Moreover, acylated glycosyl bromides are very
straightforward to be obtained.
formation of an intermediate was observed, which
evolved to 12 for longer reaction time. Compound 12 al-
lowed to obtain pure b-galactosylceramide (b-GalCer,
16), analogously to a-ManCer, and its selective sulfation
in position 3 gave pure sulfatide15 17, a compound that is
active on all members of the CD1 family (Scheme 3).
In order to have a more complete picture of the scope of
Different sugar donors were chosen to explore the scope
the method, we tried the reaction using perbenzoylated
of the reaction and to obtain different glycosphingolipids
glucopyranosyl bromide as donor, but, as already men-
tioned, we were not able to recover the expected glucoside
[e.g., a-mannosylceramide (a-ManCer) and sulfatide].
The glycosylation protocol required the use of two differ- derivative, but only undesired orthoester was obtained
ently protected glycosyl bromides; with more reactive (Scheme 4).
sugars (i.e., galactosyl bromide and mannosyl bromide)
Moreover, attempts to rearrange the orthoester under acid-
the donors were peracetylated or perbenzoylated while
ic catalysis or increasing the temperature of glycosylation
with glucose it was necessary to use a pivaloylated bro-
reaction gave product decomposition. On the other hand,
mide.
using pivaloylated donor the desired product was obtained
Indeed, the reaction with galactose or mannose probably in an acceptable 58% yield. All glycosylations are sum-
proceeded through the temporary formation of an ortho- marized in Table 1.
ester which rearranged to the desired compound, whereas
Finally, some attempts to perform glycosylation with oth-
the perbenzoylated glucose gave a more stable orthoester
er donors or in different conditions were explored. Unfor-
which decomposed instead to proceed toward the desired
tunately, neither the use of trichloroacetimidates or
glycosylations promoted with silver triflate gave the ex-
compound.12
Therefore, in order to avoid the formation of the ortho- pected products but only complex mixtures, probably be-
ester, glucose benzoyl protecting groups were substituted cause of the high sensitivity of the protecting groups of the
by pivaloates.13
acceptor to acidic conditions.
All the glycosyl bromides were conventionally obtained With respect to other methods, our approach is quite sim-
by reaction of the peresterified compound with hydro- ple and straightforward, and the glycosylation yields are
bromic acid (33% solution in glacial acetic acid) in anhy- more than acceptable. On the other hand, the main draw-
OBz
OBz
O
OBz
OBz
O
BzO
BzO
5, Hg(CN)2, MS, 40 °C
Boc
N
a. TFA–H2O (9:1)
BzO
BzO
in MeNO2, toluene
(71%)
O
b. stearic acid, EDC, HOBt,
DIPEA, in THF–EtOH (8:2)
(overall yield 68%)
C12H25
Br
O
6
7
OBz
OBz
O
OH
OH
O
O
O
BzO
BzO
HO
HO
NaOMe in MeOH
(80%)
HN
C17H35
HN
C17H35
O
O
C12H25
C12H25
OH
OH
8
9
Scheme 2
Synlett 2009, No. 16, 2609–2612 © Thieme Stuttgart · New York
Michieletti, Mario
