3
catalytic hydrogenation of azide 21 gave a complex mixture, while
the Staudinger reaction of 21 gave a stable N-ylide compound
which could not be hydrolyzed to the corresponding amine under
any conditions examined. After screening various reduction
conditions, we found that the reduction of 21 was successful when
using zinc in an acidic medium. The optimized condition, using
zinc powder (2 equiv.) in the presence of HCO2NH4 (6 equiv.) as
a hydrogen donor [22]in CH2Cl2-MeOH (3:1), provided the amine
22 with good reactivity and selectivity, in a moderate yield on the
gram scale. The primary amine of 22 was then acylated with
phenylacetyl chloride to give the amide 23. Subsequent
deprotection with TFA provided the amino lactam 24 in 46% yield
in three steps. Next, the reduction of the two amide moieties of 24
and the oxidative aromatization to provide the desired compound
1 were examined. Reduction of the diamide proceeded smoothly
using BH3•THF complex (10 equiv) under heating condition (45
°C) to give the amine-borane complex, which was hydrolyzed in a
one-pot reaction by 6N HCl at 80 °C under air to obtain the crude
triamine (Scheme 3). Interestingly, we found that another product
with red color, the desired target compound 1, was formed during
the hydrolysis reaction, albeit in a low yield (~20%). We assumed
that this unexpected oxidation might be induced by the presence
of oxygen and acid. As expected, a higher yield (~30%) of 1 was
obtained when carrying out the hydrolysis-oxidation reaction
under an O2 atmosphere, and no trace of 1 was observed under an
argon atmosphere. In addition, longer reaction time (>15 h) or
higher temperature (>90 °C) gave 1 in poor yield, and the
oxidation did not proceed under lower temperatures (< 60 °C) or
under basic condition (under O2, 80 °C).
effective as an acid and aqueous medium could significantly
promote the oxidation reaction. Finally, the optimized condition
(20% TFA, 85 °C) provided 1 in 69% yield from 25. Furthermore,
this reaction could be scaled up to 100 mg to give 1 with slightly
lower yield (45%). All physical properties (1H and 13C NMR
spectra and MS) of the synthesized 1 were in accordance to those
reported for the natural substance. In addition, the prepared 1
exhibited a comparable anti-mycobacterial activity to the natural 1
against M. smegmatis and M. bovis BCG under both aerobic and
hypoxic conditions.
To evaluate the validity of compound 1 as a candidate new
lead for an anti-TB drug, we further investigated its anti-microbial
activity against various pathogenic strains of M. tuberculosis
(Table 2). As a result, compound 1 was found to exhibit a potent
anti-mycobacterial activity (MIC values ranging from 0.5 to 2.0
µg/mL) against the drug sensitive M. tuberculosis H37Rv, Erdman
and Beijing strains grown under aerobic conditions. In addition,
compound 1 exhibited similar anti-microbial activity against drug-
resistant, multidrug-resistant and extensively drug-resistant M.
tuberculosis strains, with MIC values of 0.5–2.0 µg/mL. These
results imply that the mechanism of action of compound 1 differs
from those of existing anti-TB drugs. To date, a few semi-synthetic
derivatives of aaptamine (5) and isoaaptamine (10) [16] have been
reported to exhibit anti-microbial activity against M. tuberculosis
H37Rv or M. intracellulare under actively growing conditions,
while no anti-microbial activity against pathogenic mycobacterial
species has been reported for aaptamine (5) and other natural
related compounds [17,18]. In addition, the structure-activity
relationships and the target molecule of this class of alkaloids as
an anti-mycobacterial remain to be investigated.
MeO
BnO
Boc
MeO
BnO
Boc
N
R
a
b
c,d,e
Table 2. MICs of compound 1 against M. tuberculosis strains
N
Boc
19
Boc
N
N
Strains
Drug Resistance
MIC (µg/mL)
O
O
H37Rv
Erdman
Beijing
mc24977
mc24986
mc25886
mc25858
CI5071
CI5483
CI12081
KZN11
TF275
-
-
-
0.5~1.0
1.0
2.0
1.0
1.0
0.5
1.0
2.0
0.5~1.0
0.5
21: R = N3
22: R = NH2
20
INHa
MeO
BnO
MeO
HO
MeO
O
RIFb
N
NH
N
f
g
R
O
OFc
N
HN
Ph
N
Ph
R
NH
N
N
INH, RIF
H
H
INH, SMd
O
Ph
O
O
25
1
EMBe, SM
23: R = Boc
24: R = H
INH, RIF, SM, EMB, ETHf
INH, RIF, SM, EMB
INH, RIF, SM, EMB, ETH, KMg, PZAh
0.5
0.5
Scheme 2. Reagents and conditions: (a) Boc2O, Et3N, DMAP, CHCl3, reflux,
(89%); (b) KHMDS, trisylazide, –78 °C, then AcOH, 0 °C, 70%; (c) Zn,
NH4HCO2, CH2Cl2-MeOH; (d) PhCH2COCl, pyridine; (e) TFA, CH2Cl2, 46%
(3 steps); (f) H2, Pd-C, THF-MeOH, 80%; (g) (i) BH3•THF, THF, 45 °C, (ii)
5% HCl, THF; (iii) O2, 20% TFA, 85 °C, 45%.
a: isoniazid, b: rifampicin, c: ofloxacin, d: streptomycin,
e: ethambutol f: ethionamide, g: kanamycin, h: pyrazinamide.
Considering
that
the
synthesis
of
3-(N-
substituted)demethyl(oxy)aaptamine analogues could lead to
potential novel antimycobacterial compounds, we hypothesized
that 3-aminodemethyl(oxy)aaptamine (2) may be a useful building
block and therefore attempted its synthesis. We envisioned that the
BH3-reduction–hydrolysis–aromatization reaction of a 3-azide
compound would give 2. Thus, removal of bis Boc groups of the
3-azide 21 by treatment with TFA in CH2Cl2 was carried out to
give the 3-azide-amino lactam 26 in a quantitative yield (Scheme
4). We then examined the one-pot BH3 reduction–hydrolysis–
oxidation reaction. After treatment of the azide amino lactam 26
by BH3, the reaction mixture was hydrolyzed with 6N HCl and
heated at 85 °C under oxygen atmosphere. As expected, removal
of Bn group, hydrolysis reaction and aromatization afforded the
desired product 2 in a low yield (Scheme 4). All of the spectral
data of the synthetic 2 were identical with those of the naturally
occurring 2.
MeO
RO
H3B
MeO
RO
MeO
NH
NH
N
O
O
N
Ph
HN
Ph
N
Ph
N
N
N
H
H
BH3
O
1
24: R = Bn
25: R = H
Scheme 3. Conversion of 24 or 25 to 1.
From these results, we estimated that the phenol 25 might be
more reactive in the hydrolysis–oxidation reaction since removal
of the Bn group is necessary to the oxidation reaction providing
the oxyaaptamine skeleton. Thus, the Bn group of diamide 24 was
removed by hydrogenation to give 25 in 80% yield, and the
reduction–hydrolysis–oxidation reaction sequence against the
phenol 25 was attempted. As expected, phenol 25 was smoothly
converted to 1 in moderate yield (42%) under the same reaction
condition ((i) BH3, 45 °C, (ii) O2, 6N HCl, 85 °C). Upon further
screening of the reaction conditions, we found that using TFA was