1462
A. Krief et al. / Tetrahedron Letters 45 (2004) 1461–1463
physical properties favorably compare with those of
2
authentic samples of rhizopine isolatedfrom alfalfa.
group with the requiredaxial orientation. This was
effectively achievedin two successive steps involving
oxidation of the equatorial hydroxyl group to a ketone
followedby its reductive amination.
The synthetic strategy developed is reminiscent to that
6
of Kirby for a relatedcompound. It uses commercially
available myo-inositol 2 as starting material andtakes
advantage of the conformational rigidity of the corre-
sponding orthoformate 3 in which two hydroxyl groups
are axially orientedwhereas the remaining one is equa-
torial [2 mol equiv HC(OEt)3, Amberlyst 15, DMF,
100 °C, 3 h, 90% yield, Scheme 1].7
The first step, performedwith tetrapropylammonium
perruthenate (TPAP) in the presence of N-methyl mor-
pholine N-oxide (NMO) surprisingly gives, beside the
desired ketone 7, the epimeric alcohol 6 with the
hydroxyl group in an axial position (80/20 ratio).8
Reaction of the crude mixture with ammonium acetate
and sodium cyanoborohydride6;9 leads to the required
amine 8, which was separatedfrom the accompanying
alcohol 6 by taking advantage of the insolubility of its
ammonium salt in ether (Scheme 1, step viii). Pure
amine 8 was finally obtainedin 45% yieldafter a basic
workup.
Alkylations, sequentially carriedout with soiudm
hydride in DMF and methyl iodide then with benzyl
bromide led to the monomethyl–mono benzyl derivative
5 as well as some dibenzylated compound from which it
couldbe easily separated(Scheme 1). The selective
methylation of one of the axially orientedhyrdoxyl
group has been attributedto the easier formation of the
corresponding alkoxide due to intramolecular stabil-
ization from the remaining alcohol.6 The choice of the
sequence is crucial since reversing the order of alkylation
dramatically lowers the overall yield due to the com-
peting formation of the dibenzyl ether in addition to the
desired monobenzyl derivative.
The presence of the axial alcohol 6 as well as the
ketone 7 was unexpected. Its formation was also
observed when the oxidation was carried out under
Swern conditions.5 Furthermore we have been unable
to oxidize 6 under conditions, which allow the oxi-
dation of its epimer 5 andnot even when using PDC
or PCC instead.
It was decided to take advantage of the better accessi-
bility of reagents from the b-face to introduce the amino
The structure andstereochemistry of 8 has been ascer-
tainedon the basis of X-ray crystallography of its 3,5-
dinitrobenzoate 12.10
OH
O
O
3-O-Methyl-scyllo-inosamine hydrochloride 10 was
synthesizedfrom 8 by (i) debenzylation using sodium in
liquidammonia [Scheme 1, step x, 86% yieldof 9]1 and
(ii) acid hydrolysis of the tripode [Scheme 1, step xi, 90%
yieldof 10].
O
O
O
O
HO
HO
(i)
(iv, v)
48%
(ii, iii)
63%
HO
OH
OH
HO
90%
OH
O
Me
OH
OH
OH
2
3
4
O
O
O
O
O
6
O
6
The electrophoretic mobilities of natural rhizopine 1
extracted from nodules induced by S. meliloti strain L5-
30 andour sample of racemic-3- O-methyl-scyllo-inos-
amine are similar.
O
Me
O
O
O
Me
Ph
O
Ph
OH
O
O
(vii)
OH
(vi)
+
HO
+
6/7: 4/1
45%
O
O
O
O
O
Me
Ph
O
O
O
7
8
5
O
Me
O
Me
O
Ph
Ph
Both samples remain unaffectedwhen incubatedwith
S. meliloti Rm1021 (moc)) strain missing the gene
coding for rhizopine catabolism and are catabolized
when incubatedwith S. meliloti L5-30 (moc+) strains. It
is surprising that our racemic sample was fully catabo-
lizedby ( moc+) type Rhizobium, which wouldbe
expectedto react with only one of the two antipodes.
O
NH2
O
OAc OAc
OAc
OH
OH
OH
O
O
O
(xii)
45%
(x)
(xi)
90%
Cl
O
Me
86%
O
Me
O
Me
AcHN
OAc
NH3 OH
NH2 OH
O
O
O
(viii)
(ix)
9
10
11
O
Me
O
O
Ph
NH2
O
Furthermore, not only are the IR spectra of both the
synthetic product and the natural rhizopine superim-
posable (Fig. 2) but in addition the GC mass spectra of
the fully acetylated3- O-methyl-scyllo-inosamine 11
(Fig. 3) exhibits strong similarities with that of an
O
O2N
8
12
O
Me
Ph
NH
O
O
O2N
2
acetylatedauthentic sample recently published.
Scheme 1. Synthesis of racemic 3-O-methyl-scyllo-inosamine from
myo-inositol. (i) 2 mol equiv HC(OEt)3, Amberlyst 15, DMF, 100 °C,
3 h; (ii) NaH, DMF, 25 °C, 1 h; (iii) MeI, 25 °C, 15 h; (iv) NaH, DMF,
25 °C, 2 h; (v) BnBr, 25 °C, 15 h; (vi) 1.5 equiv NMO, 0.2 equiv TPAP,
CH2Cl2, molecular sieves; (vii) 10 equiv NH4OAc, 1.05 equiv
NaBH3CN, 20 °C, 72 h; (viii) HCl, ether; (ix) 1 M NaOH; (x) 4 equiv
Na, liq. NH3, t-BuOH, THF, )78 to 20 °C, 18 h; (xi) HCl (g), meth-
anol, 20 °C, 3 h; (xii) Ac2O, AcONa, 20 °C, 1.5 h.
The data described above strongly support the struc-
tural identification of 3-O-methyl-scyllo-inosamine 1 as
rhizopine. We are planning the synthesis of scalemic-3-
O-methyl-scyllo-inosamine to understand its biological
activity andto determine the absolute stereochemistry of
rhizopine.