Revision of the structure of fagaridine based on the comparison of UV and NMR data of synthetic compounds

Article abstract of DOI:10.1021/np980193s

Fagaridine is a quaternary benzo[c]phenanthridine alkaloid, originally isolated from Fagara xanthoxyloides in 1973. The assigned structure of this alkaloid was 7-hydroxy-8-methoxy-5-methyl-2,3-(methylenedioxy)- benzo[c]phenanthridinium (1). We have synthe

Full text of DOI:10.1021/np980193s

J . Nat. Prod. 1998, 61, 1263-1267
1263
Revision of th e Str u ctu r e of F a ga r id in e Ba sed on th e Com p a r ison of UV a n d
NMR Da ta of Syn th etic Com p ou n d s
Takeshi Nakanishi* and Masanobu Suzuki
Pharmaceuticals Group, Nippon Kayaku Company, Ltd., 31-12, Shimo 3-Chome, Kita-ku, Tokyo 115-8588, J apan
Received May 12, 1998
Fagaridine is a quaternary benzo[c]phenanthridine alkaloid, originally isolated from Fagara xanthoxyloides
in 1973. The assigned structure of this alkaloid was 7-hydroxy-8-methoxy-5-methyl-2,3-(methylenedioxy)-
benzo[c]phenanthridinium (1). We have synthesized this compound, coded NK109, aiming at a practical
antitumor drug, and during synthetic studies we questioned the original assigned structure. Thus, we
synthesized 8-hydroxy-7-methoxy-5-methyl-2,3-(methylenedioxy)benzo[c]phenanthridinium (2), isomer of
the assigned structure, and compared the spectroscopic data of both 1 and 2. The NMR data of 1 and 2
were very similar, but the UV spectra were completely different. The UV data for fagaridine agreed
with these for 2; consequently, the true structure of fagaridine is 2, not 1.
Sch em e 1. Equilibrium of Compound 1
Fagaridine, a benzo[c]phenanthridine alkaloid, was iso-
lated from Fagara xanthoxyloides by Torto et al. in 1973
and assigned structure 1.1 We reported that compound 1
possessed significant antitumor activities in 1989.2 Also,
we had developed a practical synthesis of 1 for a clinical
study.3 During these synthetic studies, we questioned the
proposed structure for fagaridine. Torto et al. isolated
fagaridine as the hydroxide (X ) OH-). Our compound 1
does not convert into the hydroxide under alkaline condi-
tions, but it converts to a ketoamine form (Scheme 1).2 The
keto amine is a characteristic purple solid, but Torto et al.
did not mention this property. We also confirmed that the
analytical data for our synthetic fagaridine did not agree
with those for Torto’s fagaridine. They assigned fagaridine
on the basis of a comparison of its NMR data with those
for similar benzo[c]phenanthridiniums, nitidine and chel-
erythrine, and on the fact that it was methylated with
diazomethane to give chelerythrine. Consequently, we
considered that the true structure of fagaridine was
presumably its isomer 2. To confirm our assumption, we
independently synthesized the compound 2, called isofa-
garidine in several papers.4 In this paper, we present the
analytical data for both 1 and 2 and reveal that the true
structure of naturally occurring fagaridine is 2, not 1. The
quaternarybenzo[c]phenanthridinium 1 (X- )HSO4-‚2H2O),
coded NK109, is currently undergoing evaluation in clinical
trials in J apan.
prepared from 2,3-dihydroxybenzaldehyde 3. First, 3 was
treated with methyl iodide in the presence of lithium
carbonate.5 3 was then selectively methylated to 3-hy-
droxy-2-methoxybenzaldehyde (4). Subsequent bromina-
tion6 and benzylation led to 3-(benzyloxy)-6-bromo-2-
methoxybenzaldehyde (6). The aldehyde unit 6 was
identified by NOE between OCH3 and CHO and OCH2Ph
and H-4. On the other hand, 6,7-(methylenedioxy)-1-
naphthylamine (7) was derived from 2,3-dihydroxynaph-
thalene.7 The aldehyde 6 and the amine 7 were condensed
to the Schiff base.8 The imine portion of the Schiff base
was reduced to amine with dimethylamineborane to yield
the key intermediate 8.9 The benzo[c]phenanthridine ring
system was constructed by radical cyclization, using tri-
n-octyltin hydride and 2,2′-azobis(2-methylbutyronitrile),
and subsequent oxidative aromatization with manganese
dioxide. These procedures gave the benzo[c]phenanthri-
dine 9. N-Methylation of 9 was successfully performed by
heating with methyl 2-nitrobenzenesulfonate10 in toluene
to give the benzo[c]phenanthridinium 10. Finally 2 was
obtained by debenzylation of 10 with hydrochloric acid.
Compound 1, NK109, had been synthesized as the hydro-
gen sulfate because of long-term stability and handling
efficacy for clinical use.3 In this study, 1 was converted
into the chloride for appropriate comparison to the isomer
2.
Our synthetic 1 and 2 were identified by NMR experi-
1
ments. The 13C and H chemical shifts are listed in Table
1, and HMBC and NOE correlations are shown in Figure
1. HMBC cross-peaks of 2, observed between NCH3 and
C-6, H-6 and C-7, and OCH3 and C-7, supported structure
2. Also the NOE of 1 between OCH3 and H-9 supported
structure 1.
Next, we compared the additional spectroscopic data for
our compounds 1 and 2 with those for Torto’s fagaridine.1
NMR chemical shifts of 1 and 2 in trifluoroacetic acid are
given in Table 2. The data similarity indicated closely
related structures. The significantly different point was
Resu lts a n d Discu ssion
We had previously established the synthetic procedure
for 1.3 Isomer 2 was synthesized according to a similar
procedure (Scheme 2). The appropriate aldehyde unit was
* To whom correspondence should be addressed. Tel.: +81-3-3598-5241.
Fax: +81-3-3598-5422. E-mail: kendama@kk.iij4u.or.jp.
10.1021/np980193s CCC: $15.00
© 1998 American Chemical Society and American Society of Pharmacognosy
Published on Web 08/27/1998

1264 J ournal of Natural Products, 1998, Vol. 61, No. 10
Sch em e 2. Synthetic Procedure of Compound 2a
Nakanishi and Suzuki
a
Key: (a) MeI, Li2CO3, DMF, 60 °C, 20 h; (b) NBS, DMF, rt, 2.5 h; (c) BnBr, K2CO3, DMF, 50 °C, 1.5 h; (d) 6 + 7, toluene (- H2O), reflux, 1 h; (e)
Me2NHBH3, AcOH, toluene, rt, 1 h; (f) n-Oct3SnH, 2,2′-azobis(2-methylbutyronitrile), toluene, 110 °C, 30 min; (g) MnO2, toluene, rt, 1 h; (h) methyl
2-nitrobenzenesulfonate, toluene, 110 °C, 86 h; (i) NaOH aq, EtOH; (j) HCl, aq; (k) concd HCl, AcOH, 60 °C, 5 h.
Ta ble 1. NMR Dataa for Our Compounds 1 and 2
position
δC
δH (no., mult, J (Hz))
HMBC correlations
C-3, C-4a, C-12
compound 1
1
106.59
7.75 (1H, s)
8.26 (1H, s)
2, 3
149.43, 149.46
105.01
121.13
132.18
151.93
116.10
146.83
146.90
125.48
114.71
128.33
126.06
119.69
131.60
133.02
103.53
52.65
4
C-2, C-4b, C-12a
4a
4b
6
10.09 (1H, s)
C-4b, C-6a, C-7, C-10a, NCH3
6a
7
8
9
8.15 (1H, d, 9.0)
8.50 (1H, d, 9.0)
C-7, C-10a
C-6a, C-8, C-10b
10
10a
10b
11
12
12a
2,3-OCH2O-
5-NCH3
7-OH
8-OCH3
8.77 (1H, d, 9.0)
8.26 (1H, d, 9.0)
C-4b, C-10a, C-10b, C-12a
C-1, C-4a, C-10b
6.34 (2H, s)
4.93 (3H, s)
11.40-11.80 (1H, br s)
4.08 (3H, s)
C-2,3
C-4, C-4b, C-6
57.87
C-8
compound 2
1
2, 3
4
4a
4b
6
6a
7
8
9
10
10a
10b
11
12
12a
105.82
148.61, 148.69
104.33
120.12
131.48
150.09
119.85
143.32
149.14
130.20
119.31
127.60
125.45
118.59
130.95
132.08
102.71
52.13
7.79 (1H, s)
8.32 (1H, s)
C-3, C-4a, C-12
C-2, C-4b, C-12a
10.04 (1H, s)
C-4b, C-6a, C-7, C-10a, NCH3
8.01 (1H, d, 9.0)
8.73 (1H, d, 9.0)
C-7, C-10a
C-6a, C-8, C-10b
8.76 (1H, d, 9.0)
8.30 (1H, d, 9.0)
C-4b, C-10a, C-10b, C-12a
C-1, C-4a, C-10b
2,3-OCH2O-
5-NCH3
7-OCH3
8-OH
6.35 (2H, s)
4.99 (3H, s)
4.16 (3H, s)
11.08 (1H, s)
C-2,3
C-4, C-4b, C-6
C-7
61.69
C-7, C-8, C-9
a
NMR spectra were recorded in DMSO-d6.
that of H-10, which is influenced by the methoxy or hydroxy
group of the 7- and 8-positions. Nevertheless, it was dif-
ficult to determine which was the true structure of Torto’s
compound using only chemical shift comparison. The
difference was not sufficient to identify the true structure.
1.11 Its pKa value was 5.3; therefore, the keto amine 1′ is
a preferable form in dilute solution for UV measurement.
The UV absorption of 539 nm reflected from the structure
1′ was strikingly characteristic of compound 1. The UV
data for Torto’s fagaridine12 were obviously different from
the data for 1 and were very close to those for 2 except
that the peak of 321 nm was the shoulder peak. This minor
difference was considered due to the influence of its
However, the UV spectra of 1 and 2, shown in Figure 2,
were completely different. Compound 1 shows equilibrium
based on the 7-position of the hydroxy group as in Scheme

Revision of the Structure of Fagaridine
J ournal of Natural Products, 1998, Vol. 61, No. 10 1265
Ta ble 2. NMR Dataa for Compounds 1 and 2 and Torto’s Fagaridine
compd 1
compd 2
Torto’s fagaridineb
4.23 (s, 8-OCH3)
5.05 (s, 5-NCH3)
4.35 (s, 7-OCH3)
5.11 (s, 5-NCH3)
4.29 (s, OCH3)
5.05 (s, NCH3)
6.26 (s, 2,3-OCH2O-)
6.28 (s, 2,3-OCH2O-)
6.21 (s, OCH2O)
7.52 (s, 1-H)
7.55 (s, 1-H)
ndc
8.06 (s, 4-H)
8.11 (s, 4-H)
ndc
8.08 (d, J ) 9.0 Hz, 9-H)
8.21 (d, J ) 9.0 Hz, 12-H)
8.44 (d, J ) 9.0 Hz, 10-H)
8.58 (d, J ) 9.0 Hz, 11-H)
9.85 (s, 6-H)
8.11 (d, J ) 9.0 Hz, 9-H)
8.24 (d, J ) 9.0 Hz, 12-H)
8.60 (d, J ) 9.0 Hz, 10-H)
8.63 (d, J ) 9.0 Hz, 11-H)
9.78 (s, 6-H)
8.08 (d, J ) 9 Hz)
8.21 (d, J ) 9 Hz)
8.57 (d, J ) 9 Hz)
8.59 (d, J ) 9 Hz)
9.77 (s)
a
b
NMR spectra were recorded in CF3COOD solvent. Described in ref 1. c Not described.
those of our compound 2. The presented structure of 2 was
correct, but their compound was not exactly a new benzo-
[c]phenanthridinium. They isolated the natural fagaridine
and assigned it correctly for the first time.
After Torto and co-workers had reported the isolation of
fagaridine from F. xanthoxyloides, many researchers iso-
lated fagaridine from several Rutaceae plants.14 All of
them were identical with Torto’s fagaridine; therefore, they
are compound 2 as well. As far as we know, compound 1
was never isolated from plants; only synthetic approaches
were achieved by Hanaoka in 198515 and our laboratory
in 1990.16 Accordingly, compound 1 is not a natural
product but a completely synthetic one.
In conclusion, natural fagaridine was isolated by Torto
and co-workers for the first time; however, its true struc-
ture has not been recognized until now. Both fagaridine
and isofagaridine isolated from plants are the same prod-
uct, that is, compound 2. From now on, compound 2 should
be called “fagaridine”. Also, compound 1, not yet known
as a natural product, should be called “isofagaridine”.
F igu r e 1. HMBC (arrow) and NOE (bold arrow) correlations of
compounds 1 and 2.
Exp er im en ta l Section
Gen er a l Exp er im en ta l P r oced u r es. 2,2′-Azobis(2-me-
thylbutyronitrile) and tri-n-octyltin hydride were purchased
from J apan Hydrazine Co., Inc., and Sankyo Organic Chemi-
cals Co., Ltd., respectively. Other materials were obtained
from commercial suppliers. DMF was dehydrated by molec-
ular sieves 4A, and others were used without further purifica-
tions. Melting points were determined on a Bu¨chi 535 melting
point apparatus. UV and IR spectra were recorded on a
Shimadzu UV-2200 and a Hitachi 260-10 spectrometer, re-
spectively. NMR spectra were obtained on a Varian Gemini-
200 and a J EOL J NM-GX400 spectrometer. All chemical
shifts are reported in ppm relative to the internal standard
tetramethylsilane (TMS, δ 0.00). Coupling constants are
reported in Hz. HMBC and GOESY spectra of compounds 1
and 2 were obtained on a BRUKER AVANCE DRX 400. Mass
spectra were recorded on a Micromass Limited Auto Spec-Q
spectrometer. Elemental analyses were performed on a Yana-
ko CHN Corder (C, H, N analysis) and a Yokogawa IC-7000D
(Br analysis). TLC for 1 and 2 were carried out on silica gel
60 F254 (0.25 mm-thick plates) using CH2Cl2-MeOH (9:1) as
solvent.
3-Hyd r oxy-2-m eth oxyben za ld eh yd e (4). To a mixture
of 2,3-dihydroxybenzaldehyde (3) (3.84 g, 27.8 mmol) and
lithium carbonate (5.14 g, 69.5 mmol) in DMF (70 mL) was
added 2.60 mL of methyl iodide (41.7 mmol) and the mixture
heated at 60 °C for 20 h. The mixture was condensed in vacuo,
and the resulting residue was diluted with water, acidified
with concentrated HCl, and then extracted with diisopropyl
ether. The organic layer was washed with water, dried over
Na2SO4, filtered, and evaporated. The crude residue was
chromatographed on silica gel (hexanes-ethyl acetate 4:1) to
give 4 (2.23 g, 53%) as a white solid: mp 115 °C; 1H NMR
(CDCl3, 200 MHz) δ 3.98 (3H, s), 5.96 (1H, s), 7.15 (1H, dd, J
) 8.2, 7.4), 7.24 (1H, dd, J ) 8.2, 2.0), 7.38 (1H, dd, J ) 7.4,
F igu r e 2. UV spectra of compounds 1 and 2 in ethanol solution (20
µmol/L).
concentration. Thus, we conclude that 2 is the true
structure of fagaridine isolated by Torto.
Among other analytical data, TLC was notable in
identifying the structures because it easily displayed the
equilibrium property of compound 1. After the develop-
ment of both compounds in CH2Cl2-MeOH (9:1) on TLC
plates, 1 and 2 were observed as purple and yellow spots,
respectively. As in Scheme 1, the color of the quarternary
benzo[c]phenanthridinium 1 is yellow and the other keto
amine 1′ is purple. The compound 1 lost its acid residue
on the TLC plate and was developed as structure 1′, but
compound 2 did not show any similar evidence. This is a
simple method that distinguishes between them.
As mentioned above, the natural fagaridine should be
described as structure 2, which was reported previously
by Fang and co-workers in 1993.4a They isolated compound
2 as an inhibitor of DNA topoisomerase I from Zanthoxy-
lum nitidum and named it isofagaridine. They determined
the structure on the basis of NMR experiments and several
spectral methods. Also, their UV data13 were identical with

1266 J ournal of Natural Products, 1998, Vol. 61, No. 10
Nakanishi and Suzuki
2.0), 10.27 (1H, s); ESIMS m/z 151 [M- - H]; anal. C 63.02%,
H 5.29%, calcd for C8H8O3, C 63.15%, H 5.30%.
148.4, 148.5; FAB-MS m/z 410 [M+ + H]; anal. C 75.99%, H
4.59%, N 3.16%, calcd for C26H19NO4, C 76.27%, H 4.68%, N
3.42%.
6-Br om o-3-h yd r oxy-2-m eth oxyben za ld eh yd e (5). To a
solution of 4 (2.23 g, 14.7 mmol) in DMF (73 mL) was added
dropwise N-bromosuccinimide (3.92 g, 22.0 mmol) in DMF (100
mL). After 4 disappeared, the solution was concentrated in
vacuo. The resulting residue was partitioned between diiso-
propyl ether and water. The organic layer was washed with
water, dried over Na2SO4, filtered, and evaporated. The crude
residue was chromatographed on silica gel (CH2Cl2) to give 5
8-(Ben zyloxy)-7-m et h oxy-5-m et h yl-2,3-(m et h ylen ed i-
oxy)ben zo[c]p h en a n th r id in iu m Ch lor id e (10). The solu-
tion of 9 (282 mg, 0.69 mmol) and methyl 2-nitrobenzene-
sulfonate (299 mg, 1.38 mmol) in toluene (5.6 mL) was heated
at 110 °C for 86 h to yield a yellow suspension. It was filtered,
and the residue was neutralized with ethanol (13.8 mL) and
0.1 N NaOH (13.8 mL). The resulting white suspension was
partitioned between CH2Cl2 and water. The organic layer was
washed with saturated NaCl, dried over Na2SO4, filtered, and
evaporated. The crude residue was chromatographed on silica
gel (toluene-ethyl acetate) to give the pure pseudobase.3,17 It
was dissolved in acetone (10 mL), and then 1 N HCl (1.38 mL)
was added. The resulting precipitate was filtered, washed
with acetone, and dried in vacuo to give 10 (250 mg, 79%) as
a yellow powder: mp 174-175 °C dec; 1H NMR (DMSO-d6,
200 MHz) δ 4.21 (3H, s), 5.00 (3H, s), 5.50 (2H, s), 6.35 (2H,
s), 7.36-7.51 (3H, m), 7.54-7.62 (2H, m), 7.78 (1H, s), 8.30
(1H, d, J ) 8.8 H), 8.31 (1H, s), 8.37 (1H, d, J ) 9.4 H), 8.82
(1H, d, J ) 8.8 H), 8.82 (1H, d, J ) 9.4 H), 10.12 (1H, s); 13C
NMR (DMSO-d6, 50 MHz) δ 52.1, 62.2, 70.9, 102.7, 104.2,
105.7, 118.7, 119.1, 119.3, 120.0, 125.1, 127.2, 127.8 × 2, 128.2,
128.6 × 2, 130.9, 131.7, 132.2, 136.1, 145.7, 148.6, 148.7, 149.4,
150.7; FABMS m/z 424 [M+ + H]; anal. C 62.59%, H 5.20%,
N 2.58%, calcd for C27H22ClNO4‚1.5HCl‚0.15H2O, C 62.40%,
H 5.08%, N 2.70%.
8-Hyd r oxy-7-m eth oxy-5-m eth yl-2,3-(m eth ylen ed ioxy)-
ben zo[c]p h en a n th r id in iu m Ch lor id e (2). The mixture of
10 (217 mg, 0.47 mmol), acetic acid (4.34 mL), and concen-
trated HCl (2.17 mL) was warmed at 60 °C for 5 h. When the
reaction was completed, the solution was cooled to room
temperature and diluted with acetone (26 mL). The resulting
orange suspension was filtered, washed with acetone, and
dried in vacuo to give 2 (154 mg, 88%) as an orange powder:
mp 231-233 °C dec; UV (EtOH) λmax (log ꢀ) 227 (4.52), 283
(4.63), 321 (4.18), λmin (log ꢀ) 251 (4.19), 311 (4.16); IR (KBr)
νmax 3440, 3072, 1606, 1547, 1483, 1475, 1404, 1385, 1321,
1284, 1257, 1203, 1161, 1126, 1082, 1039, 976, 935, 879, 827
cm-1; FABMS m/z 334 [M+]; TLC Rf 0.40 (as a yellow spot);
1H and 13C NMR see Tables 1 and 2.
1
(2.07 g, 61%) as a pale yellow powder: mp 135 °C; H NMR
(CDCl3, 200 MHz) δ 3.93 (3H, s), 6.03 (1H, s), 7.07 (1H, d, J )
8.7), 7.34 (1H, d, J ) 8.7), 10.35 (1H, s); 13C NMR (CDCl3, 50
MHz) δ 63.4, 115.9, 121.4, 126.7, 130.0, 148.3, 149.4, 191.1;
FABMS m/z 230, 232 [M•+], 231, 233 [M+ + H]; anal.
C
41.37%, H 3.05%, Br 34.52%, calcd for C8H7BrO3, C 41.59%,
H 3.05%, Br 34.58%.
3-(Ben zyloxy)-6-br om o-2-m eth oxyben zaldeh yde (6). To
a mixture of 5 (1.94 g, 8.39 mmol) and potassium carbonate
(1.16 g, 12.6 mmol) in DMF (40 mL) was added 1.2 mL of
benzyl bromide (10.1 mmol), and the mixture was heated at
50 °C for 1.5 h. The mixture was concentrated in vacuo and
partitioned between toluene and water. The organic layer was
washed with water, dried over Na2SO4, filtered, and evapo-
rated. The crude residue was chromatographed on silica gel
(hexanes-ethyl acetate 92:8) to give 6 (2.44 g, 90%) as a white
1
solid: mp 76 °C; H NMR (CDCl3, 200 MHz) δ 3.97 (3H, s),
5.14 (2H, s), 6.99 (1H, d, J ) 8.8), 7.30 (1H, d, J ) 8.8), 7.33-
7.46 (5H, m), 10.35 (1H, s); 13C NMR (CDCl3, 50 MHz) δ 62.4,
71.3, 113.5, 119.6, 127.3 × 2, 128.3, 128.7 × 2, 128.9, 129.3,
136.0, 151.8, 152.7, 190.5; FABMS m/z 320, 322 [M•+], 321,
323 [M+ + H]; anal. C 56.01%, H 3.94%, Br 24.48%, calcd for
C
15H13BrO3, C 56.10%, H 4.08%, Br 24.88%.
N-[3′-(Ben zyloxy)-6′-br om o-2′-m eth oxyben zyl]-6,7-(m e-
th ylen ed ioxy)-1-n a p h th yla m in e (8). The solution of 6 (2.44
g, 7.59 mmol) and 6,7-(methylenedioxy)-1-naphthylamine 7
(1.42 g, 7.59 mmol) in toluene (77 mL) was refluxed for 1 h,
and the toluene was distilled in atmospheric pressure to give
a Schiff base as an oil. This was dissolved in toluene (77 mL),
447 mg of dimethylamineborane (7.59 mmol) was added, and
then acetic acid (10.3 mL, 179.9 mmol) was added to the
solution and the mixture stirred for 1 h. When the reduction
was completed, the reaction mixture was quenched with 1 N
HCl (38.5 mL). After 1 h, the mixture was neutralized with 6
N NaOH. The organic layer was separated, washed with
water, dried over Na2SO4, filtered, and evaporated. The
resulting residue was recrystallized with ethyl acetate-
ethanol to give 8 (3.28 g, 88%) as a white powder: mp 160 °C;
1H NMR (CDCl3, 200 MHz) δ 3.89 (3H, s), 4.48 (1H, br s), 4.57
(2H, s), 5.11 (2H, s), 5.99 (2H, s), 6.82 (1H, d, J ) 8.9), 6.83
(1H, dd, J ) 7.0, 1.1), 7.07 (1H, s), 7.12 (1H, dd, J ) 7.6, 1.1),
7.14 (1H, s), 8.83 (1H, d, J ) 8.9), 7.21-7.47 (6H, m); 13C NMR
(CDCl3, 50 MHz) δ 43.8, 61.8, 71.0, 97.5, 100.9, 104.7, 105.5,
115.0, 116.0, 117.8, 120.3, 125.1, 127.3 × 2, 128.0, 128.1, 128.7
× 2, 131.3, 132.7, 136.5, 143.1, 147.1, 147.3, 149.5, 151.4;
7-Hyd r oxy-8-m eth oxy-5-m eth yl-2,3-(m eth ylen ed ioxy)-
ben zo[c]p h en a n th r id in iu m Ch lor id e (1). 7-Hydroxy-8-
m et h oxy-5-m et h yl-2,3-(m et h ylen edioxy)ben zo[c]ph en a n -
thridinium hydrogensulfate dihydrate3 (467 mg, 1.00 mmol)
was dissolved with water, neutralized with 0.1 N NaOH, and
extracted with CH2Cl2. The organic layer was washed with
water, dried over Na2SO4, filtered, and concentrated in vacuo.
The resulting violet solid was suspended with acetone (8 mL),
and 1 N HCl (2 mL) was added to yield an orange suspension.
It was filtered, washed with acetone, and dried in vacuo to
give 1 (347 mg, 94%) as an orange powder: mp 224-226 °C
dec; UV (EtOH) λmax (log ꢀ) 245 (4.63), 286 (4.55), 328 (4.22),
539 (3.81), λmin (log ꢀ) 261 (4.30), 312 (4.16), 433 (3.16); IR (KBr)
νmax 3400, 1603, 1550, 1492, 1481, 1379, 1352, 1301, 1277,
FABMS m/z 491, 493 [M•+], 492, 494 [M+ + H]; anal.
C
63.00%, H 4.35%, N 2.33%, Br 15.22%, calcd for C26H22BrNO4‚
1260, 1211, 1157, 1117, 1099, 1037, 969, 934, 863, 823 cm-1
;
0.3AcOEt, C 62.97%, H 4.74%, N 2.70%, Br 15.40%.
FABMS m/z 334 [M+]; TLC Rf 0.32 (as a purple spot); 1H and
8-(Ben zyloxy)-7-m eth oxy-2,3-(m eth ylen ed ioxy)ben zo-
[c]p h en a n th r id in e (9). The solution of 8 (2.78 g, 5.64 mmol)
and tri-n-octyltin hydride (6.48 g, 14.1 mmol) in toluene (278
mL) was heated at 110 °C and then azobis(2-methylbutyroni-
trile) (1.63 g, 8.46 mol) in toluene (5 mL) was added. After 30
min, the solution was cooled to room temperature, and
manganese dioxide (2.78 g) was added. After being stirred for
1 h, the reaction mixture was filtered through Celite, and the
solvent was evaporated. The residue was recrystallized with
CHCl3-hexane to give 9 (567 mg, 25%) as a white powder:
mp 212 °C; 1H NMR (CDCl3, 200 MHz) δ 4.17 (3H, s), 5.32
(2H, s), 6.13 (2H, s), 7.26 (1H, s), 7.34-7.47 (3H, m), 7.48-
7.56 (2H, m), 7.59 (1H d, J ) 9.0), 7.83 (1H, d, J ) 9.0), 8.30
(1H, d, J ) 9.0), 8.32 (1H, d, J ) 9.0), 8.72 (1H, s), 9.76 (1H,
s); 13C NMR (CDCl3, 50 MHz) δ 62.0, 72.0, 101.3, 102.2, 104.4,
118.16, 118.22, 120.0, 120.9, 122.0, 127.0, 127.5 × 2, 128.2,
128.6, 128.7 × 2, 129.1, 129.8, 136.8, 140.1, 146.0, 146.5, 148.3,
13C NMR see Tables 1 and 2.
Ack n ow led gm en t. We are very grateful to Prof. Miyoji
Hanaoka (Kanazawa University, J apan) for providing several
benzo[c]phenanthridine alkaloids.
Su p p or tin g In for m a tion Ava ila ble: Copies of the 1H NMR, 13C
NMR, and HMBC spectra of compounds 1 and 2 (6 pages). Ordering
information is given on any current masthead page.
Refer en ces a n d Notes
(1) Torto, F. G.; Mensah, I. A.; Baxter, I. Phytochemistry 1973, 12, 2315-
2317.
(2) Hanaoka, M.; Ekimoto H.; Kobayashi, M. F.; Irie, Y.; Takahashi, K.
Chem. Abstr. 1989, 116, 718.
(3) Nakanishi, T.; Suzuki, M.; Mashiba, A.; Ishikawa, K.; Yokotsuka, T.
J . Org. Chem. 1998, 63, 4235-4239.

Revision of the Structure of Fagaridine
J ournal of Natural Products, 1998, Vol. 61, No. 10 1267
(4) (a) Fang, S.-D.; Wang, L.-K.; Hecht, S. M. J . Org. Chem. 1993, 58,
5025-5027. (b) Makhey, D.; Gatto, B.; Yu, C.; Liu, A.; Liu, L. F.;
LaVoie, E. J . Med. Chem. Res. 1995, 5, 1-12. (c) Cho, W.-J .; Hanaoka,
M. Arch. Pharm. Res. 1996, 19, 240-242.
(13) Described in ref 4a: UV (MeOH) λmax (log ꢀ) 228 (4.20), 284 (4.29),
322 (sh) (3.82), λmin (log ꢀ) 252 (3.85) nm.
(14) (a) Addae-Mensah, I.; Sofowora, E. A. Planta Med. 1979, 35, 94-96
(Fagara tessmannii). (b) Adesina, S. K.; Akinwusi, D. D. J . High.
Resolut. Chromatogr. Chromatogr. Commun. 1986, 9, 412-414
(Zanthoxylum zanthoxyloides). (c) Adesina, S. K. J . Nat. Prod. 1986,
49, 715-716 (Zanthoxylum zanthoxyloides). (d) Adesina, S. K.;
Olatunji, O. A.; Akinwusi, D. D. Pharmazie 1986, 41, 747 (Zanthoxy-
lum rigidifolium). (e) Adesina, S. K. Fitoterapia 1987, 58, 123-126
(Zanthoxylum leprieurii).
(15) Hanaoka, M.; Yamagishi, H.; Mukai, C. Chem. Pharm. Bull. 1985,
33, 1763-1765.
(16) Suzuki, M.; Nakanishi, T.; Kogawa, O.; Ishikawa, K.; Kobayashi, F.;
Ekimoto, H. Chem. Abstr. 1990, 117, 191706.
(5) Wymann, W. E.; Davis, R.; Patterson, J . W., J r.; Pfister, J . R. Synth.
Commun. 1988, 18, 1379-1384.
(6) Mitchell, R. H.; Lai, Y.-H.; Williams, R. V. J . Org. Chem. 1979, 44,
4733-4735.
(7) (a) Stermitz, F. R.; Gillespie, J . P.; Amoros, L. G.; Romero, R.;
Stermitz, T. A. J . Med. Chem. 1975, 18, 708-713. (b) Clark, J . H.;
Holland, H. L.; Miller, J . M. Tetrahedron Lett. 1976, 41, 3361-3364.
(8) Castellano, J . A.; Goldmacher, J . E.; Barton, L. A.; Kane, J . S. J . Org.
Chem. 1968, 33, 3501-3504.
(9) Billman, J . H.; McDowell, J . W. J . Org. Chem. 1961, 26, 1437-1440.
(10) Kiprianov, A. I.; Tolmachev, A. I. Zh. Obsh. Khim. 1959, 29, 2868-
2874; Chem. Abstr. 1960, 54, 12126.
(17) Walterova´, D.; Preininger, V.; Grambal, F.; Sˇima´nek, V.; Sˇantavy´,
(11) Manuscript in preparation.
F. Heterocycles 1980, 14, 597-600.
(12) Described in ref 1: UV (EtOH) λmax (log ꢀ) 228 (4.55), 284 (4.65), 322
(sh) (4.14), λmin (log ꢀ) 252 (4.16) nm.
NP980193S