6258
I. Cumpstey / Tetrahedron Letters 46 (2005) 6257–6259
HO OH
HO OH
O
OH
7
1
BnO
BnO
6
2
8
BnO
BnO
BnO
BnO
(i)
5
3
+
OBn
4
OBn
OBn
OBn
OBn
OBn
6
8
7
Scheme 1. Reagents and conditions: (i)CH 2@CHMgBr, THF, 0 °C; 8, 29%; 7, 65%.
as a 2.2:1 mixture in favour of the desired R diastereo-
mer 7 (vide infra)( Scheme 1).9
Pivaloylation of the alcohols 11 and removal of the
DMB protecting groups gave the alcohols 14a and
14b, still as an inseparable mixture. Swern oxidation of
the alcohols 14a and 14b gave the ketones 15 and 16,
which were easily separated by flash column chromato-
graphy, in 81% and 7% yields, respectively. Wittig meth-
ylenation of the ketone 15 gave the diene 17, which
smoothly underwent ring-closing metathesis mediated
by GrubbsÕ second generation catalyst to give the carba-
sugar 18.5 The stereochemistry at the pseudoanomeric
centre could then be assigned as R, based on the cou-
It was necessary to differentiate the two alcohol func-
tionalities of the resulting diol 7. Success in similar sys-
tems has been reported previously: Martin selectively
protected the allylic alcohol of 9 using benzoyl chloride
under phase-transfer conditions,10 whereas Al-Abed
selectively protected the allylic alcohol of the epimeric
mixture 10 using para-methoxybenzyl chloride and
sodium hydride (Fig. 2).6
3
pling constant J1,2 of 7.9 Hz.12 Straightforward deacyl-
Unfortunately, these conditions could not be generally
applied, as in my system, treatment of 7 with benzoyl
chloride under the reported conditions gave a 2.8:1 mix-
ture of the regioisomeric monobenzoates in favour of
the 2-O-benzoate; attempted para-methoxybenzylation
of 8 under the reported conditions gave a 1:1 mixture
of the regioisomeric products. However, I found
that excellent regioselectivity could be achieved using
3,4-dimethoxybenzyl chloride (DMBCl)11 with sodium
hydride at 0 °C in DMF, and 11a and 11b were obtained
as an inseparable 10:1 mixture of regioisomers in favour
of the 2-O-dimethoxybenzyl derivative 11a, along with
the diprotected compound 12 (Scheme 2). As an alterna-
tive solution to this problem, the dithioacetal route
described by Jeon and Kim avoids the need for differen-
tiation of the two secondary alcohols, but has the
disadvantages of using ethanethiol as solvent during
dithioacetal formation, and mercury salts for its
removal.8
ation of the pseudoanomeric protecting group gave the
alcohol 5, which may be transformed into valienamine
1 in three steps using FukaseÕs procedure.7
In summary, I have demonstrated a new synthesis of the
precursor 5 to valienamine 1 in eight steps from com-
mercially available tetrabenzyl glucose 6 and in 7.7%
overall yield. The key steps were the selective protection
of one of two secondary alcohols in the acyclic deriva-
tive 7 and the formation of the cyclohexene ring using
ring-closing metathesis mediated by GrubbsÕ second
generation catalyst, with the double bond being in the
correct position for valienamine. Further investigations,
including the synthesis of mannose- and galactose-
derived valienamine analogues using this methodology,
the introduction of nitrogen at the pseudoanomeric
position before ring closure and the conjugation of
valienamine to other sugars, is in progress, and the
results will be reported in due course.
RO OR'
BnO
HO OH
PivO
O
O
OPiv
BnO
OBn
(iv)
(i)
BnO
BnO
BnO
BnO
BnO
OBn
+
OBn
OBn
BnO
OBn
11a R=DMB, R'=H
11b R=H, R'=DMB
+
OBn
OBn
OBn
15
16
7
(ii)
12 R=R'=DMB
13a R=DMB, R'=Piv
13b R=Piv, R'=DMB
(v)
(iii)
14a R=H, R'=Piv
14b R=Piv, R'=H
OPiv
OPiv
OBn
OH
OBn
BnO
BnO
BnO
BnO
(vi)
(vii)
BnO
BnO
OBn
OBn
OBn
OBn
17
5
18
Scheme 2. Reagents and conditions: (i)DMBCl, DMF, NaH, 0 °C; 57%; 11a/11b, 10:1; (ii)PivCl, pyridine, DMAP, 93%; (iii)CAN, MeCN, H 2O,
0 °C ! rt; 74%; (iv)oxalyl chloride, DMSO, Et N, DCM, ꢀ60 °C; 16, 7%; 15, 81%; (v)PPh CH3Br, THF, NaHMDS, ꢀ78 °C; 63%; (vi)Grubbs Õ
3
3
2nd gen. cat., toluene, 60 °C, 65%; (vii)NaOMe, MeOH, 40 °C, >99% (DMB = 3,4-dimethoxybenzyl).