1232
M. Kubo et al. / Tetrahedron Letters 53 (2012) 1231–1235
Table 1
1H (600 MHz) and 13C (150 MHz) NMR data for 1 and 2 in CD3OD
Position
1
2
dH
dC
dH
dC
1
2
2.60 (dq, 7.8, 7.6)
4.31 (dt, 7.8, 4.9)
2.44 (dd, 14.7, 7.8)
1.45 (dd, 14.7, 4.9)
38.8
73.0
43.7
138.6
210.2
36.5
3a
2.53 (dd, 18.9, 6.7)
2.03 (ddq, 18.9, 2.7, 0.6)
2.88 (dd, 6.7, 2.7)
3b
4
5
6
80.2
47.5
76.2
82.0
30.2
44.4
47.9
79.5
81.1
32.9
7
4.66 (d, 5.6)
2.75 (d, 12.2)
2.28 (dd, 12.2, 5.6)
4.15 (dd, 3.0, 3.0)
2.70 (ddq, 13.7, 3.0, 1.4)
2.74 (dd, 13.7, 3.0)
8a
8b
9
60.7
171.9
180.5
23.2
10
11
12
181.1
178.8
19.9
1.22 (3H, d, 0.8)
3.80 (d, 8.0)
a
1.22 (3H, s)
72.2
12b
4.10 (dq, 8.0, 0.8)
1.71 (3H, dd, 1.4, 0.6)
13
a
3.92 (d, 10.0)
4.09 (d, 10.0)
1.04 (3H, d, 7.6)
75.4
74.6
7.7
7.7
13b
14
Figure 3. NOESY correlations of 1.
which is the first example of seco-prezizaane-type norsesquiterpe-
noid, and thus 1 was named as (2R)-hydroxy-norneomajucin.
Next, we have decided to confirm the absolute structure of 1 by
synthesizing from (2R⁄)-hydroxyneomajucin (4). Previously, we
have demonstrated that jiadifenolide can be obtained from neo-
majucine by oxidizing the C-10 hydroxy group.6 Moreover, Dani-
shefsky’s group had succeeded in the total synthesis of ( )-
jiadifenin by applying our oxidative method in the final step.11 Fol-
lowing the previous protocol, the secondary hydroxy group at C-2
in 4 was first acetylated, but the reaction did not proceed presum-
ably because of a steric hindrance. Therefore, (2S⁄)-hydroxyneo-
majucin (3) was used for further elaboration. In the case of
compound 3, the secondary hydroxyl group at C-2 was readily
acetylated, and then the subsequent Jone’s oxidation gave rise to
6 as anticipated (Scheme 1). Hydrolysis of 6, followed by the oxida-
tion of the generated secondary alcohol at C-2 with IBX, gave rise
to ketone 7. Finally, the NaBH4 reduction of 7 solely led to the alco-
hol, all the spectroscopic data of which were identical with those of
compound 1. This chemical conversion confirmed that the absolute
configuration of 1 is the same as that of 3. Next, the Kusumi’s
method12 was applied to establish the absolute configuration of
J = 14.7, 4.9 Hz), 2.44 (dd, J = 14.7, 7.8 Hz); dC 43.7 (C-3)]. The afore-
mentioned spectroscopic data indicated that 1 belongs to majucin-
like seco-prezizaane-type sesquiterpenoids except for the absence
of a d-lactone ring, which is a characteristic of these sesquiterpe-
noids. Furthermore, the molecular formula indicated that the car-
bon number of 1 is one less than that of normal sesquiterpenoids,
implying that 1 is a norsesquiterpenoid. Therefore, the spectro-
scopic data of 1 failed to refer to those of the previously known
seco-prezizaane-type sesquiterpenoids. Extensive analyses of
1H–1H COSY, HMQC, and HMBC of 1 (Fig. 2) showed that 1 has
the same A–C ring system as (2R⁄)-hydroxyneomajucin (4). In addi-
tion, the ester carbon signal (dC 181.1) showed HMBC correlations
with each H-7, H-8, and H-1, thereby allowing us to form a c-lac-
tone ring D between C-7 and C-9. Since the NMR signal correspond-
ing to the C-11 position which commonly exists in all the majucin
derivatives was found to be missing, 1 was assumed to be a nor-
type of (2R⁄)-hydroxyneomajucin (4).
The relative stereochemistry of 1 was elucidated on the basis of
NOESY as shown in Figure 3. Namely, H-1/H-2, H-2/H-3
/H3-12 indicated that the methyl group at C-14 and the hydroxy
group at C-2 take b-configurations and the methyl group at C-12 is
in -configuration. In addition, both the methyl group at C-12 and
hydroxy group at C-6 were assigned as -configurations from the
NOE correlation of H3-14/H-8, and H-13b/H-3b and H-13 /H3-12.
a, and H-
3a
3, which has not yet been determined. The
Dd values as shown
a
in Figure 4, enabled us to unambiguously assign the C-2 configura-
tion in 3 and 4 as S and R, respectively. Accordingly, it was noted
that the absolute configuration of 1 is the same as that of (2S)-
hydroxyneomajucin (3).
a
a
On the basis of the aforementioned data, the structure of 1 was de-
duced to be represented as 11-nor-(2R)-hydroxyneomajucin,
Compound 2 has the molecular formula C13H16O5, as deduced in
the HR-EI-MS at m/z 252.1012 [M]+, indicating 6 degrees of unsat-
uration. Its IR spectrum revealed the presence of a hydroxy group
at 3417 cmÀ1, an
and a
a ,
,b-unsaturated carbonyl group at 1693 cmÀ1
c
-lactone moiety at 1770 cmÀ1. The 1H NMR data of 2 (Ta-
ble 1) showed a low-shifted signal at dH 1.71 (dd, J = 1.4, 0.6 Hz)
due to an olefinic methyl group, which long-range coupled to the
H-3b [dH 2.03 (ddq, J = 18.9, 2.7, 0.6 Hz)] and H-8a [dH 2.70 (ddq,
J = 13.7, 3.0, 1.4 Hz)]. The NMR spectrum of 2 was similar to that
of 2-oxo-3,4-dehydroxyneomajucin except for the absence of a d-
lactone ring.
Furthermore, the molecular formula of 2 indicated that the car-
bon number of 2 was two less than that of normal seco-prezizaane-
type sesquiterpenoids. Analyses of 1H–1H COSY, HMQC, and HMBC
experiments were carried out (Fig. 5). The HMBC correlations of H-
3 and H-4/C-2 and C-9, as well as H3-13/C-1 and C-2, revealed the
presence of 2-methylcyclopent-2-enone moiety A. The other HMBC
correlations showed that 2 has the same B–C ring system as 2-oxo-
3,4-dehydroxyneomajucin. The additional HMBC correlations, as
Figure 2. 13C NMR data and HMBC correlations of 1.