Structure of saframycin R

Article abstract of DOI:10.1016/S0040-4020(00)00972-8

The structure of saframycin R was determined to be 1 (form I) by the two-dimensional ¹H detected heteronuclear correlation experiments (HMQC and HMBC) of its acylated compounds 4a and 4b. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00972-8

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 9937±9944
Structure of Saframycin R
*
Naoki Saito, Noriko Kameyama and Akinori Kubo
*
Meiji Pharmaceutical University, 2-522-1 Noshio, Kiyose, Tokyo 204-8588, Japan
Received 22 September 2000; accepted 16 October 2000
1
AbstractÐThe structure of saframycin R was determined to be 1 (form I) by the two-dimensional H detected heteronuclear correlation
experiments (HMQC and HMBC) of its acylated compounds 4a and 4b. q 2000 Elsevier Science Ltd. All rights reserved.
Saframycin is a class of antibiotics with activity against
gram-positive bacteria and also against several kinds of
tumor cells.1 A similar group of metabolites that includes
renieramycins and ecetinascidins was isolated from marine
sources.2,3 Although saframycin A (2) is one of the most
biologically active components of the groups, it is highly
Figure 1.
Keywords: structure; saframycin R; saframycin A; transformation.
* Corresponding authors. Tel.: 181-424-95-8794; fax: 181-424-95-8792; e-mail: naoki@my-pharm.ac.jp
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00972-8

9938
N. Saito et al. / Tetrahedron 56 (2000) 9937±9944
Table 1. NMR Data for saframycin R (1) (all data were recorded in CDCl3)
Atom no.
13C NMR dmult
1H NMR (mult. Integral, J (Hz))
Correlations from C no.
1
2
3
4
4a
135.7 s
148.3 s
118.1 s
149.1 s
116.8 s
15-H
2-OMe, 3-Me
3-Me
3-Me, 5-Hb, 5-Ha
5-Hb, 5-Ha, 6-H
15-H
15a
5
121.9 s
20.8 t
5-Hb, 5-Ha, 14a-H
15-H
6-H, 7-H
2.43 (d, 1H, 17.7)
2.93 (dd, 1H, 17.7, 8.1)
3.48 (m, 1H)
6
54.5 d
58.9 d
5-Hb, 5-Ha, 7-H
N-Me, 15-H
5-Hb, 5-Ha, 6-H
9-H, 14a-H
7
4.07 (d, 1H, 2.4)
9
9a
56.9 d
135.3 s
3.93 (ddd, 1H, 2.7, 2.7, 2.0)
14a-H, 17-H2
9-H, 17-H2, 14-Hb
14-Ha
10
11
12
13
13a
14
180.8 s
155.9 s
128.4 s
185.6 s
141.8 s
24.3 t
11-OMe, 12-Me
12-Me
12-Me, 14-Ha
14-Hb, 14-Ha, 9-H
15-H
1.54 (ddd, 1H, 17.7, 11.3, 2.7)
2.85 (dd, 1H, 17.7, 2.5)
3.11 (ddd, 1H, 11.3, 2.7, 2.5)
3.69 (dd, 1H, 2.7, 0.5)
3.48 (m, 2H)
14a
15
17
19
20
21
2-OMe
11-OMe
3-Me
12-Me
16-Me
CN
54.3 d
56.7 d
40.4 t
160.5 s
195.7 s
24.3 q
61.0 q
61.1 q
9.2 q
7-H, 14-Ha, 15-H
6-H, N-Me
N±H
N±H, 17-H2, COMe
COMe
2.20 (s, 3H)
3.71 (s, 3H)
4.02 (s, 3H)
2.20 (s, 3H)
1.93 (s, 3H)
2.28 (s, 3H)
8.7 q
41.5 q
117.0 s
60.7 t
6-H, 15-H
7-H
COCH2OH
COCH2OH
NH
4.56 (s, 2H)
171.7 s
CH2OH
6.33 (t, 1H, 6.7)
5.28 (br s, 1H)
OHa
a One OH proton was not detected.
toxic, which prevents it from gaining wider acceptance for
cancer chemotherapy.4,5² Arai and co-workers have investi-
gated minor components of 2 cultures of Streptomyces
lavendulae No. 314, and discovered saframycin R (1),
which is active against several experimental tumor cells
and the acute toxicity in mice is one tenth that of 2.6 The
present paper describes the structure of saframycin R
determined by two-dimensional H detected heteronuclear
correlation experiments, thus precluding efforts to obtain
suitable crystals for an expensive X-ray analysis.
was determined by spin-echo 13C effects. Lown et al. further
elucidated the structure of saframycin R by comparing the
1H NMR spectral data of saframycin A (1a).6b They ruled
out I and II by comparing the chemical shift differences
between the groups of protons adjacent to the E ring
(such as 9-H, 14-Hb, 14-Ha) and those immediately
adjacent to ring A (such as 5-Hb, 5-Ha, 15-H) in safra-
mycins R and A. Saframycin R experiences shifts of 0.20
to 0.10 ppm whereas saframycin A experiences smaller
shifts of 0.08±0.00 ppm. Furthermore, the average con-
formation of the side chain is similar in saframycin R and
in saframycin A, which argues against any steric effects due
to a large group at C-13 and favors the orientation of form
III.
1
In 1982, Arai and co-workers presented four possible
structures (I±IV) for saframycin R based on its chemical
and spectroscopic data (Fig. 1).6a The proposed structure
was corroborated by the 13C NMR spectrum showing 31
distinct resonances, 15 sp3, 15 sp2, and 1 sp carbons,
whose multiplicity (6£CH3, 4£CH2, 5£CH, and 16£C)
For the purposes of discussion of the NMR spectra of
saframycin R, the 14 protons (excluding the methyl groups
and 2 OH protons) have been reassigned as shown in
Table 1. The diagnostic homoallylic coupling (2.7 Hz)
between 9-H and 14-Hb through ®ve bonds was con®rmed.
In our previous work, this coupling was negligible when
the compound did not have quinone functionality at
the E ring.7 Thus, the data allow III and IV to be ruled
out unequivocally. Additional evidence is provided by
²
Ecteinascidin 743 is an exceedingly potent antitumor agent obtained
from marine ascidian that is currently undergoing phase II clinical trials
as a result of its promising ef®cacy in preclinical antitumor tests. Recently,
the Harvard group has found a structural analogue of ecteinascidin 743,
phthalascidin, which exhibits antitumor activity essentially indistinguish-
able from that of the natural product.5a

N. Saito et al. / Tetrahedron 56 (2000) 9937±9944
9939
Chart 1.
long-range 1H±13C connectivity, which was determined
through a series of H detected two-dimensional hetero-
The substituents of ring E were assigned as follows: While
two quinone carbonyls appear in the spectrum at d 185.6
and d 180.8, one quinone carbonyl (C-13) was assigned
based on long-range 1H±13C correlations between C-13
and 12-CH3 protons and 14-Ha, and in the other quinone
1
nuclear multiple-bond correlation (HMBC) experiments.8
The aromatic substituents of the A ring were assigned as
follows: The signal at d 149.1 was assigned to C-4 based on
1
carbonyl (C-10), there were negligible long-range H±13C
1
long-range H±13C correlations observed between C-4 and
correlations. The signal at d 135.3 was assigned to C-9a
based on long-range 1H±13C correlations between C-9a
and the following protons; 9-H, 17-CH2, 14-Hb, and 14-
Ha. The signal at d 141.8 was assigned to C-13a based on
long-range 1H±13C correlations between C-13a and the
three protons (14-Hb, 14-Ha, and 9-H).
the 3-CH3 protons and 5-Hb and 5-Ha. A methyl group (d
9.2) was located on C-3 (d 118.1) based on long-range
1H±13C correlations between the 3-CH3 protons and the
three carbons C-2, C-3, and C-4. A methoxyl group (d
61.0) was attached to C-2 (d 148.3) based on a long-range
1H±13C correlation between the 2-OCH3 protons and C-2.
The signal at d 135.7 was assigned to C-1 based on a long-
range 1H±13C correlation between 15-H and C-1. The signal
at d 121.9 was assigned to C-15a based on long-range
1H±13C correlations between C-15a and the following
protons; 5-Hb, 5-Ha, 15-H, and 14a-H. The signal at d
Thus, there are two possible orientations of the glycolic
ester substituents at C-1 or C-4. Unfortunately, there were
no con®rmatory nuclear Overhauser enhancement (NOE)
effects between the phenolic proton and the protons of 2-
OCH3 or the phenolic proton and the protons of 3-CH3,
because the phenolic proton signal was overlapped with
other signals and could not be assigned. We then attempted
to obtain derivatives with suitable crystals for X-ray analysis.
1
116.8 was assigned to C-4a based on long-range H±13C
correlations between C-4a and the four protons 5-Hb,
5-Ha, 6-H, and 15-H.

9940
N. Saito et al. / Tetrahedron 56 (2000) 9937±9944
Figure 2.
Chart 2.
³
Numerous efforts to convert 1 to the corresponding methyl
derivative were unsuccessful; only polar polymeric
materials were obtained. Treatment of 1 with a large excess
of acetic anhydride in dry pyridine, however, gave the
diacetate (3a) in 68% yield (Chart 1). On the other hand,
acetylation of 1 with acetic anhydride (1.8 equiv.) in the
presence of 4-dimethylaminopyridine (DMAP) in dry
pyridine afforded the monoacetate (4a) in 54.9% yield
along with 3a (38.0%). We had hoped that the bulky reagent
would exert enough steric in¯uence on the course of acyl-
ation at the C-4 position. Indeed, reaction of 1 with pivaloyl
chloride (2.2 equiv.) and base in CH2Cl2 afforded 4b
(46.1%) and 3b (9.0%). There were no crystals of the four
products, however characterization of 4a and 4b by exten-
sive NMR measurements (including COSY, HMQC, and
HMBC techniques) established unexpected results. Analy-
sis of the 1H and 13C NMR and high-resolution MS data of
4a and 4b suggested the formulas of C31H34N4O9 and
C34H40N4O9 for compounds 4a and 4b, respectively. The
1H- and 13C NMR spectral data of 4a and 4b were very
similar, with the major difference being the presence of
signals attributable to an acetyl group in 4a [1H: d 2.32
(3H, s); 13C: d 20.5 (q) and 169.8 (s)] and to a pivaloyl
group in 4b [1H: d 1.38 (9H, s); 13C: d 27.3 (q), 39.4 (s),
and 176.8 (s)]. Both compounds have a sharp phenolic OH
signal (4a: d 5.73, 4b: d 5.69) with an NOE enhancement of
³
Acylation of 1 with 4-bromobenzyl chloride and base gave the diacylated
compound (3: Rˆp-Br±C6H4CH2: 89%), however we have not been able to
obtain any crystals for X-ray analysis.

N. Saito et al. / Tetrahedron 56 (2000) 9937±9944
9941
Chart 3.
the 2-OCH3 methyl protons (4a: d 3.81, 4b: d 3.81). Thus,
the OH group must be orientated at C-1 in both 4a and 4b.
Experimental
1
This is consistent with the long-range H±13C correlations
IR spectra were measured in CHCl3 with a Hitachi 260
spectrophotometer. H spectra were measured at 270 MHz
1
between C-1 (4a: d 144.7, 4b: d 144.5) and H-15 (4a: d
4.18, 4b: d 4.19). Selected HMBC correlations data in
compound 4a were shown in Fig. 2 and a probable mechan-
istic pathway for the formation of the acetates 4a and 4b is
shown in Chart 2. The acylation was especially slow for the
sterically crowded alcohol, and the initial attack of alkoxide
to the carbonyl in glycolic ester afforded 4a and 4b. The
possibility of preparing 3b and 4b from 3a and 4a by regio-
selective hydrolysis was excluded because the diacylated
compounds 3a and 4a were very stable in organic base
(such as triethylamine, DMAP).§ Treatment of 4a with
acetic anhydride in pyridine afforded the diacetate (5) in
83.6% yield.
with a JEOL JNM-EX 270 spectrometer. 13C NMR were
recorded at 67.5 MHz (JEOL JNM-EX 270) and 125 MHz
(JEOL JNM-LA 500) (multiplicity determined from off-
resonance decoupling or distortionless enhancement by
polarization transfer (DEPT) spectra). NMR spectra were
measured in CDCl3, and chemical shifts were recorded in
dH values relative to internal (CH3)4Si as a standard. Mass
spectra were recorded on a JMS-DX 302 mass spectrometer.
Optical rotation [a]D measurements were made on a
Horiba-SEPA-200 automatic digital polarimeter at 238C.
All reactions were conducted under an argon atmosphere.
Dry solvents and reagents were obtained using standard
procedures. Anhydrous sodium sulfate was used for drying
organic solvent extracts. Removal of the solvent was done
with a rotary evaporator and, ®nally, under high vacuum.
Column chromatography was performed with E. Merck
Silica gel 60 (70±230 mesh).
Finally, we examined the transformation of saframycin R
(1) to saframycin A (2) (Chart 3). Several common methods
of effecting hydrolysis were eliminated due to their
ineffectiveness. Numerous attempts at this conversion
under aqueous acidic or basic conditions were totally
unsuccessful because of the labile nature of the quinone.
Treatment of 1 with concentrated H2SO4 in methanol at
608C for 24 h afforded saframycin A (2) in 19.9% yield,
and the major product was the ketal (6) in 71.7% yield.
The structure proposed for 6 was supported by the 13C
NMR spectrum, which showed a peak at d 100.3 assigned
Preparation of saframycin R (1). An stocked saframycin R
(1) was slightly impure and contained a few degradation
products. It was puri®ed by preparative thin layer chromato-
graphy on silica gel plates (E. Merck, No. 5715: solvent 1:2
benzene±ethyl acetate) immediately before use. The pure 1
as a pale yellow amorphous powder, showed [a]Dˆ279.28
(c 0.6, CHCl3) and its IR, MS data were identical with
1
to the ketal carbon. In addition, the H NMR spectrum
reported values. H- and 13C NMR data were shown in
1
showed four methoxyl methyl signals at d 2.88, 3.02,
4.06, and 4.08.9 Accordingly, this problem was solved as
follows: The reaction of 1 with a high excess triethylamine
and DMAP in CH2Cl2 at room temperature for 40 h gave 2
in 44.7% yield. The synthetic 2 was identical in all respects
with an authentic sample.
Table 1. IR (CHCl3) 3600w, 3400, 1770, 1725w, 1685,
1660, 1620 cm21; MS m/z (relative intensity) 622 (M1, 5),
524 (10), 522 (M12100, 12), 318 (44), 279 (53), 278 (100),
220 (64), 205 (17), 204 (16), 78 (98), 77 (17), 57 (12), 44
(58); high-resolution EIMS calcd for C31H34N4O10
622.2275, found 622.2280; Positive ion FAB-MS (magic
Bullet) m/z 623 (M1 1 H), 596, 278, 220.
Saframycin R (1) is the ®rst example of a latent hydro-
quinone at the A ring. The biological properties of 4a and
4b will be reported elsewhere.
Acetylation of 1. Method A: Acetic anhydride (0.4 mL) was
added to a solution of (2)-1 (31.1 mg, 0.5 mmol) in dry
pyridine (1.0 mL), and the mixture was stand at room
temperature for 1 h. The reaction mixture was concentrated
in vacuo. The residue was diluted with water (10 mL), and
the mixture was extracted with chloroform (10 mL£3). The
§
During the reaction, we detected a highly polar initial product along with
a less polar the diacylated compound (3a and 4a) using TLC. After work up,
however, the highly polar compound was transformed to the ®nal product
(3b and 4b).

9942
N. Saito et al. / Tetrahedron 56 (2000) 9937±9944
combined extracts were washed with brine (10 mL), dried,
and concentrated in vacuo to give the residue (38.1 mg).
Chromatography on a silica gel (9 g) column with 2:1
benzene±ethyl acetate afforded the diacetate 3a (24.0 mg,
68.0%) as pale yellow amorphous powder: [a]Dˆ277.18 (c
0.5, CHCl3); IR (CHCl3) 3370, 1745, 1675, 1655 cm21; 1H
NMR d (CDCl3) 1.42 (1H, ddd, Jˆ17.2, 11.2, 2.6 Hz, 14-
Hb), 1.91 (3H, s, 12-CH3), 2.10 (3H, s, COCOCH3), 2.10
(1H, d, Jˆ18.1 Hz, 5-Hb), 2.18 (3H, s, 3-CH3), 2.29 (3H, s,
NCH3), 2.32, 2.35 (each 3H, s, COCH3), 2.88 (1H, dd,
Jˆ17.2, 3.0 Hz, 14-Ha), 2.94 (1H, dd, Jˆ18.1, 7.3 Hz,
5-Ha), 2.94 (1H, signals overlap with 5-Ha, 17-H), 3.16
(1H, ddd, Jˆ11.2, 3.0, 2.6 Hz, 14a-H), 3.41 (1H, d like,
6-H), 3.72 (1H, ddd, Jˆ13.9, 9.2, 1.7 Hz, 17-H), 3.77
(3H, s, 2-OCH3), 3.79 (1H, dd, Jˆ2.6, 0.5 Hz, 15-H), 3.92
(2H, s like, 7-H and 9-H), 4.07 (3H, s, 11-OCH3), 4.87, 4.96
(each 1H, d, Jˆ16.2 Hz, OCOCH), 6.88 (1H, br s, NH); 13C
NMR d (CDCl3) 8.6 (q, 12-CH3), 9.8 (q, 3-CH3), 20.4, 20.6
(each q, OCOCH3), 21.4 (t, C-5), 23.8 (t, C-14), 24.4 (q,
COCOCH3), 41.5 (q, NCH3), 42.0 (t, 17-C), 54.5 (d, C-6),
54.5 (d, C-14a), 56.6 (d, C-9), 56.6 (d, C-15), 59.5 (d, C-7),
60.8 (t, OCOCH2), 61.0 (q, 11-OCH3), 61.1 (q, 2-OCH3),
117.0 (s, CN), 122.9 (s), 123.1 (s), 123.6 (s), 125.3 (s), 127.3
(s, C-12), 136.7 (s, C-9a), 139.7 (s, C-13a), 144.7 (s), 148.7
(s), 156.4 (s, C-11), 161.3 (s, COCOCH3), 166.2 (s,
OCOCH2), 169.5, 170.7 (each s, OCOCH3), 180.5 (s,
C-10), 185.8 (s, C-13), 195.6 (s, COCOCH3); MS m/z
(relative intensity) 706 (M1, 2), 608 (20), 607 (28), 606
(M12100, 38), 581 (20), 403 (11), 402 (46), 363 (32),
362 (100), 348 (16), 320 (11), 262 (28), 229 (22), 218
(13), 204 (14), 43 (18); high-resolution EIMS calcd for
C35H38N4O12 706.2486, found 706.2484.
3-CH3), 20.5 (q, OCOCH3), 21.4 (t, C-5), 23.9 (t, C-14),
24.4 (q, COCOCH3), 41.7 (q, NCH3), 41.9 (t, C-17), 54.6
(d, C-6), 54.9 (d, C-14a), 55.7 (d, C-15), 56.5 (d, C-9), 59.4
(d, C-7), 61.0 (q, 11-OCH3), 61.1 (q, 2-OCH3), 115.6 (s,
C-15a), 117.2 (s, CN), 122.7 (s, C-3), 122.7 (s, C-4a),
128.5 (s, C-12), 136.0 (s, C-9a), 139.1 (s, C-4), 140.6 (s,
C-13a), 143.5 (s, C-2), 144.7 (s, C-1), 156.3 (s, C-11), 161.6
(s, COCOCH3), 169.8 (s, OCOCH3), 180.6 (s, C-10), 186.0
(s, C-13), 195.5 (s, COCOCH3); MS m/z (relative intensity)
606 (M1, 2), 509 (11), 508 (34), 507 (10), 506 (M12100,
25), 483 (22), 481 (19), 302 (27), 264 (11), 263 (27), 262
(100), 248 (12), 245 (14), 229 (14), 220 (30), 206 (12), 205
(17), 204 (20), 171 (15), 171 (15), 149 (25), 107 (28), 97
(12), 91 (37), 83 (10), 71 (17), 69 (20), 59 (16), 57 (29), 56
(10), 55 (20), 43 (15), 41 (18); high-resolution EIMS calcd
for C31H34N4O9 606.2326, found 606.2330.
Acylation of 1 with pivaloyl chloride. Trimethylacetyl
chloride (pivaloyl chloride, 15.4 mL, 0.125 mmol) was
added to a solution of (2)-1 (16.7 mg, 0.027 mmol),
triethylamine (34.8 mL, 0.250 mmol), and DMAP (1.2 mg,
0.010 mmol) in dry dichloromethane (5.0 mL), and the
mixture was stand at room temperature for 24 h. The reac-
tion mixture was diluted with 5% NaHCO3 (10 mL), and
then extracted with chloroform (10 mL£3). The combined
extracts were washed with brine (10 mL), dried, and
concentrated in vacuo to give the residue (24.9 mg).
Chromatography on a silica gel (9 g) column with 3:1
hexane±ethyl acetate afforded 3b (15.6 mg, 73.7%) as
pale yellow amorphous powder. Further elution with 2:1
hexane±ethyl acetate afforded 4b (4.0 mg, 23.0%) as pale
yellow amorphous powder.
Method B: Acetic anhydride (4.7 mL, 0.050 mmol) was
stirred solution of (2)-1 (17.4 mg,
Compound 3b. [a]Dˆ256.98 (c 0.7, CHCl3); IR (CHCl3)
1
added to
a
3550, 3390, 1740, 1680, 1665 cm21; H NMR d (CDCl3)
0.028 mmol), triethylamine (78.0 mL, 0.560 mmol), and
DMAP (6.8 mg, 0.056 mmol) in dry dichloromethane
(14 mL), and the stirring was continued at room temperature
for 48 h. The reaction mixture was diluted with 5%
NaHCO3 (10 mL), and then extracted with chloroform
(10 mL£3). The combined extracts were washed with
brine (10 mL), dried, and concentrated in vacuo to give
the residue (22.7 mg). This residue was subjected to
chromatography on preparative layer silica gel plates
(solvent 1:2 hexane±ethyl acetate) to afford 3a (7.5 mg,
38.0%) and 4a (9.3 mg, 54.9%) as pale yellow amorphous
powder: N-[4-acetyl-7-cyano-6,7,9,10,13,14,14a,15-octa-
hydro-1-hydroxy-2,11-dimethoxy-3,12,16-trimethyl-10,13-
dioxo-6a,7a,9a,14aa,15a-6,15-imino-5H-isoquino[3,2-
b][3]benzazocin-9-yl]-methyl]-2-oxopropanamide (4a):
[a]Dˆ274.58 (c 0.5, CHCl3); IR (CHCl3) 3570, 3410,
1.27 and 1.38 (each 9H, s, C(CH3)3), 1.45 (1H, ddd, Jˆ17.5,
10.0, 2.0 Hz, 14-Hb), 1.91 (3H, s, 12-CH3), 2.01 (1H, d,
Jˆ18.1 Hz, 5-Hb), 2.09 (3H, s, COCOCH3), 2.16 (3H, s,
3-CH3), 2.31 (3H, s, NCH3), 2.89±2.96 (2H, m, signals
overlap with 5-Ha, and 17-H), 3.01 (1H, dd, Jˆ17.5,
2.6 Hz, 14-Ha), 3.18 (1H, ddd, Jˆ11.2, 2.6, 2.6 Hz, 14a-
H), 3.40 (1H, d like, 6-H), 3.72 (1H, ddd, Jˆ13.9, 9.2,
1.7 Hz, 9-CH), 3.77 (3H, s, 2-OCH3), 3.86 (1H, dd,
Jˆ2.6, 0.5 Hz, 15-H), 3.92 (2H, s like, 7-H and 17-H),
4.07 (3H, s, 11-OCH3), 6.91 (1H, br s, NH); 13C NMR d
(CDCl3) 8.6 (q, 12-CH3), 9.7 (q, 3-CH3), 21.4 (t, C-5), 22.8
(t, C-14), 24.4 (q, COCOCH3), 27.0 and 27.3 (each s,
C(CH3)3), 38.7, 39.5 (each s, C(CH3)3), 41.4 (q, NCH3),
42.2 (t, C-17), 54.5 (d, C-6), 54.5 (d, C-14a), 56.6 (d,
C-9), 56.6 (d, C-15), 59.5 (d, C-7), 60.6 (t, OCOCH2),
61.0 (q, 11-OCH3), 61.1 (q, 2-OCH3), 117.0(s, CN), 122.8
(s), 122.9 (s), 123.0 (s), 125.3 (s), 127.3 (s, C-12), 136.3 (s,
C-9a), 139.9 (C-13a), 144.6 (s), 148.8 (s), 156.4 (s, C-11),
161.5 (s, COCOCH3), 166.2 (s, OCOCH2), 176.6, 178.0
(each s, OCOC(CH3)3), 180.5 (s, C-10), 185.8 (s, C-13),
195.4 (s, COCOCH3); MS m/z (relative intensity) 790
(M1, 2), 693 (45), 692 (45), 691 (35), 690 (M12100, 35),
674 (13), 667 (22), 666 (11), 665 (26), 487 (12), 486 (38),
448 (12), 447 (36), 446 (100), 362 (15), 361 (11), 344 (17),
305 (15), 304 (59), 220 (26), 219 (12), 218 (34), 205 (11),
204 (18), 149 (13), 97 (10), 85 (10), 83 (10), 71 (14), 69
(15), 59 (11); high-resolution EIMS calcd for C41H50N4O12
790.3425, found 790.3427.
1
1760, 1710, 1690, 1670 cm21; H NMR d (CDCl3) 1.48
(1H, ddd, Jˆ17.5, 11.2, 2.3 Hz, 14-Hb), 1.90 (3H, s,
12-CH3), 2.10 (3H, s, COCOCH3), 2.13 (1H, d,
Jˆ18.5 Hz, 5-Hb), 2.17 (3H, s, 3-CH3), 2.32 (3H, s,
COCH3), 2.37 (3H, s, NCH3), 2.91 (1H, dd, Jˆ18.5,
6.9 Hz, 5-Ha), 2.91 (1H, signals overlap with 5-Ha, 17-
H), 3.01 (1H, dd, Jˆ17.5, 2.6 Hz, 14-Ha), 3.16 (1H, ddd,
Jˆ11.2, 2.6, 2.6 Hz, 14a-H), 3.39 (1H, d like, 6-H), 3.73
(1H, ddd, Jˆ13.5, 8.9, 1.7 Hz, 17-H), 3.82 (3H, s, 2-OCH3),
3.93 (2H, s like, 7-H and 9-H), 4.06 (3H, s, 11-OCH3), 4.18
(1H, dd, Jˆ2.6, 0.5 Hz, 15-H), 5.73 (1H, s, OH), 6.88 (1H,
br s, NH); 13C NMR d (CDCl3) 8.6 (q, 12-CH3), 9.8 (q,

N. Saito et al. / Tetrahedron 56 (2000) 9937±9944
9943
Compound 4b. [a]Dˆ255.38 (c 0.35, CHCl3); IR (CHCl3)
9.8 (q, 3-CH3), 20.6, 20.7 (each q, OCOCH3), 21.5 (t, C-5),
23.8 (t, C-14), 24.5 (q, COCOCH3), 40.9 (q, NCH3), 41.5 (t,
9-CH2), 54.5 (d, C-6), 57.0 (d, C-9), 57.2 (d, C-15), 57.5 (d,
C-14a), 59.4 (d, C-7), 60.9 (q, 11-OCH3), 61.0 (q, 2-OCH3),
MS m/z (relative intensity) 648 (M1, 3), 551 (13), 550 (40),
549 (30), 548 (M12100, 33), 525 (11), 523 (21), 344 (44),
305 (32), 304 (100). 302 (11), 290 (19), 262 (35), 246 (16),
229 (10), 220 (26), 205 (11), 204 (13), 43 (12); high-resolu-
tion EIMS calcd for C33H36N4O10 648.2432, found
648.2424.
1
3570, 3410, 1760, 1710, 1690, 1670 cm21; H NMR d
(CDCl3) 1.38 (9H, s, C(CH3)3), 1.48 (1H, ddd, Jˆ17.5,
11.2, 2.3 Hz, 14-Hb), 1.91 (3H, s, 12-CH3), 2.08 (3H, s,
COCOCH3), 2.13 (1H, d, Jˆ18.5 Hz, 5-Hb), 2.14 (3H, s,
3-CH3), 2.37 (3H, s, NCH3), 2.91 (1H, dd, Jˆ18.5, 6.9 Hz,
5-Ha), 2.91 (1H, signals overlap with 5-Ha, 17-H), 3.01
(1H, dd, Jˆ17.5, 2.6 Hz, 14-Ha), 3.18 (1H, ddd, Jˆ11.2,
2.6, 2.6 Hz, 14a-H), 3.40 (1H, d like, 6-H), 3.73 (1H, ddd,
Jˆ13.5, 8.9, 1.7 Hz, 17-H), 3.81 (3H, s, 2-OCH3), 3.93 (2H,
s like, 7-H and 9-H), 4.06 (3H, s, 11-OCH3), 4.19 (1H, dd,
Jˆ2.6, 0.5 Hz, 15-H), 5.69 (1H, s, OH), 6.91 (1H, br s, NH);
13C NMR d (CDCl3) 8.6 (q, 12-CH3), 9.8 (q, 3-CH3), 21.5 (t,
C-5), 23.9 (t, C-14), 24.5 (q, COCOCH3), 27.3 (q, C(CH3)3),
39.4 (s, C(CH3)3), 41.7 (q, NCH3), 42.2 (t, C-17), 54.7 (d,
C-6), 54.9 (d, C-14a), 55.9 (d, C-15), 56.5 (d, C-9), 59.6 (d,
C-7), 61.0 (q, 11-OCH3), 61.1 (q, 2-OCH3), 115.6 (s,
C-15a), 117.2 (s, CN), 122.7 (s, C-3), 122.8 (s, C-4a),
127.3 (s, C-12), 136.2 (s, C-9a), 139.1 (s, C-4), 140.6 (s,
C-13a), 143.5 (s, C-2), 144.5 (s, C-1), 156.4 (s, C-11), 161.6
(s, COCOCH3), 176.8 (s, OCO(CH3)3), 180.6 (s, C-10),
186.0 (s, C-13), 195.4 (s, COCOCH3); MS m/z (relative
intensity) 648 (M1, 6), 550 (17), 549 (30), 548 (M12100,
77), 524 (11), 523 (30), 345 (15), 344 (64), 305 (28), 304
(100), 290 (17), 259 (15), 243 (10), 220 (30), 218 (14), 205
(17), 204 (23), 149 (12), 69 (12), 57 (41), 55 (11), 43 (11);
high-resolution EIMS calcd for C34H40N4O9 648.2795,
found 648.2800.
Transformation of saframycin R (1) to saframycin A (2).
Method A: A solution of (2)-1 (9.9 mg, 0.0159 mmol) was
stirred with triethylamine (44.4 mL, 0.319 mmol) and
DMAP (3.9 mg, 0.0319 mmol) in dry dichloromethane
(8 mL) at room temperature for 40 h. The reaction mixture
was diluted with 1N HCl (10 mL), and extracted with
chloroform (10 mL£3). The combined extracts were
washed with water (10 mL), dried, and concentrated in
vacuo to give the neutral fraction (3.2 mg). This fraction
was subjected to chromatography on preparative layer silica
gel (Merck 5715: solvent 1:2 hexane±ethyl acetate) to give
the starting material (2.6 mg, 26.3% recovery). The acidic
aqueous layer was made alkaline with saturated NaHCO3
solution and extracted with chloroform (10 mL£3). The
combined extracts were washed with brine (10 mL), dried,
and concentrated in vacuo. The residue (6.1 mg) was
subjected to chromatography on preparative layer silica
gel (Merck 5715: solvent 1:1 hexane±ethyl acetate) to
give saframycin A (2: 4.0 mg, 44.7%) as dark yellow amor-
phous powder, which was identical in all respects with an
authentic sample.
The same procedure as described above, but using (2)-1
(16.7 mg, 0.0268 mmol) with pivaloyl chloride (7.4 mL,
0.06 mmol), triethylamine (69.6 mL, 0.50 mmol), and
DMAP (6.1 mg, 0.05 mmol) in dry dichloromethane
(14 mL) at room temperature for 24 h gave 3b (1.9 mg,
9.0%) and 4b (8.0 mg, 46.1%).
Method B: Concentrated H2SO4 (0.1 mL) was added to a
stirred solution of (2)-1 (16.7 mg, 0.0268 mmol) in metha-
nol (5.0 mL), and the reaction mixture was heated at 708C
for 5 h. The reaction mixture was diluted with 5% NaHCO3
(15 mL), and extracted with chloroform (10 mL£3). The
combined extracts were washed with brine (10 mL), dired,
and concentrated in vacuo to give the residue (22.4 mg).
Chromatography on a silica gel (8 g) column with 2:1
hexane±ethyl acetate afforded saframycin A (3.0 mg,
19.9%). Further elution with 1:1 hexane±ethyl acetate
afforded 6 (11.7 mg, 71.7%) as pale yellow amorphous
powder.
Acetylation of 4a. Acetic anhydride (0.2 mL) was added to
a solution of 4a (8.5 mg, 0.014 mmol) in dry pyridine
(0.5 mL), and the mixture was stand at room temperature
for 14 h. The reaction mixture was concentrated in vacuo.
The residue was diluted with water (10 mL), and the
mixture was extracted with chloroform (10 mL£3). The
combined extracts were washed with brine (10 mL), dried,
and concentrated in vacuo to give the residue (10.3 mg).
This residue was subjected to chromatography on prepara-
tive layer silica gel plates (solvent 1:1 hexane±ethyl acetate)
to afford 5 (7.6 mg, 83.6%) as pale yellow amorphous
(2)-Saframycin A dimethylketal (6). [a]Dˆ224.98 (c 1.0,
CHCl3); IR (CHCl3) 3450, 1690, 1660, 1620 cm21 1H
;
powder:
N-[1,4-diacetyl-7-cyano-6,7,9,10,13,14,14a,15-
octahydro -2,11-dimethoxy-3,12,16-trimethyl-10,13-dioxo-
6a,7a,9a,14aa,15a-6,15-imino-5H-isoquino[3,2-b][3]benza-
zocin-9-yl]-methyl]-2-oxopropanamide (5): IR (CHCl3)
NMR d (CDCl3) 1.18 (3H, s, C±CH3), 1.31 (1H, ddd,
Jˆ17.5, 11.2, 2.6 Hz, 14-Hb), 1.87, 1.98 (each 3H, s, 3-
and 12-CH3), 2.23 (1H, d, Jˆ21.1 Hz, 5-Hb), 2.32 (3H, s,
NCH3), 2.86 (1H, dd, Jˆ21.1, 8.9 Hz, 5-Ha), 2.88 (3H, s,
OCH3), 2.89 (1H, dd, Jˆ17.5, 2.6 Hz, 14-Ha), 2.92 (1H,
ddd, Jˆ13.0, 4.0, 3.3 Hz, 17-H), 3.02 (3H, s, OCH3), 3.12
(1H, ddd, Jˆ11.2, 3.0, 2.6 Hz, 14a-H), 3.46 (1H, ddd,
Jˆ8.9, 2.3, 0.5 Hz, 6-H), 3.95 (1H, ddd, Jˆ13.0, 10.2,
1.7 Hz, 17-H), 3.96 (1H, s like, 9-H), 4.04 (1H, d,
Jˆ2.3 Hz, 7-H), 4.05 (1H, dd, Jˆ3.0, 0.5 Hz, 15-H), 4.06,
4.08 (each 3H, s, 2- and 11-OCH3), 6.57 (1H, dd, Jˆ10.2,
3.3 Hz, NH); 13C NMR d (CDCl3) 8.5 (q, quinone-CH3), 8.8
(q, quinone-CH3), 21.1 (q, C±CH3), 21.6 (t, C-5), 25.3 (t,
C-14), 40.8 (t, C-17), 41.6 (q, NCH3), 49.3 (q, ketal-OCH3),
49.6 (q, ketal-OCH3), 53.9 (d, C-14a), 54.2 (d, C-9), 54.4 (d,
1
3400, 1750, 1690, 1670 cm21; H NMR d (CDCl3) 1.46
(1H, ddd, Jˆ17.5, 11.2, 2.3 Hz, 14-Hb), 1.91 (3H, s, 12-
CH3), 2.10 (3H, s, COCOCH3), 2.13 (1H, d, Jˆ18.5 Hz,
5-Hb), 2.18 (3H, s, 3-CH3), 2.34, 2.35 (each 3H, s,
COCH3), 2.42 (3H, s, NCH3), 2.87±3.01 (3H, m, 5-Ha,
17-H, and 14-Ha), 3.16 (1H, ddd, Jˆ11.2, 3.0, 2.6 Hz,
14a-H), 3.43 (1H, d like, 6-H), 3.71 (1H, ddd, Jˆ13.2,
9.8, 2.0 Hz, 17-H), 3.73 (1H, dd, Jˆ2.0, 0.5 Hz, 15-H),
3.78 (3H, s, 2-OCH3), 3.93 (2H, s like, 7-H and 9-H), 4.07
(3H, s, 11-OCH3), 6.87 (1H, br s, NH); 13C NMR d (CDCl3:
all unsaturated carbon peaks could not determined because
of the limited amount of sample available) 8.6 (q, 12-CH3),

9944
N. Saito et al. / Tetrahedron 56 (2000) 9937±9944
York, 1987; pp 633±666. (c) Kubo, A.; Saito, N. Synthesis of
Isoquinolinequinone Antibiotics. In Studies in Natural Products
Chemistry; Atta-ur-Rahman, Ed.; Elsevier: Amsterdam, 1992;
Vol. 10, pp 77±145. (d) Ozturk, T. In The Alkaloids; Cordell,
G. A., Ed.; Academic: New York, 2000; Vol. 53, pp 119±238.
2. (a) Frincke, J.; Faulkner, D. J. J. Am. Chem. Soc. 1982, 104,
265±269. (b) He, H.; Faulkner, D. J. J. Org. Chem. 1989, 54,
5822±5826. (c) Davidson, B. S. Tetrahedron Lett. 1992, 33,
3721±3724. (d) Parameswaran, P. S.; Naik, C. G.; Kamat, S. Y.;
Pramanik, B. N. Indian J. Chem. 1998, 37B, 1258±1263.
3. (a) Wright, A. E.; Forleo, D. A.; Gunawardana, G. P.;
Gunasekera, S. P.; Koehn, F. K.; McConnel, J. M. J. Org. Chem.
1990, 55, 4508±4512. (b) Rinehart, K. L.; Holt, T. G.; Fregeau,
N. L.; Stroh, J. G.; Keifer, P. A.; Sun, F.; Li, L. H.; Martin, D. G.
J. Org. Chem. 1990, 55, 4512±4515. (c) Sakai R.; Rinehart, K. L.;
Guan, Y.; Wang, A. H. -J. Proc. Natl. Acad. Sci. USA 1992, 89,
11456±11460. (d) Guan, Y.; Sakai, R.; Rinehart, K. L.; Wang,
A. H.-J. J. Biomol. Struc. Dyn. 1993, 10, 793±818. (e) Sakai, R.;
Jares-Erijman, E. A.; Manzanares, I.; Elipe, M. V. S.; Rinehart,
K. L. J. Am. Chem. Soc. 1996, 118, 9017±9023.
C-6), 57.3 (d, C-15), 58.3 (d, C-7), 61.1 (q, quinone-OCH3),
61.1 (q, quinone-OCH3), 100.3 (s, ketal-C), 116.6 (s, CN),
127.0 (s), 128.1 (s), 135.8 (s), 136.4 (s), 139.9 (s, C-4),
141.5 (s), 155.3 (s), 156.6 (s), 170.2 (s, COC(OCH3)2CH3),
180.6 (s), 182.3 (s), 185.6 (s), 186.3 (s); MS m/z (relative
intensity) 608 (M1, 1), 464 (21), 437 (11), 245 (18), 244
(12), 243 (12), 221 (22), 220 (63), 219 (11), 218 (16), 204
(10), 203 (11), 89 (100); high-resolution EIMS calcd for
C31H36N4O9 608.2482., found 608.2497.
Synthesis of saframycin A dimethylketal (6). Concen-
trated H2SO4 (0.1 mL) was added to a stirred solution of
(2)-2 (28.1 mg, 0.50 mmol) in methanol (5.0 mL), and
the reaction mixture was heated at 708C for 5 h. The reac-
tion mixture was diluted with 5% NaHCO3 (15 mL), and
extracted with chloroform (10 mL£3). The combined
extracts were washed with brine (10 mL), dried, and
concentrated in vacuo to give the residue (34.5 mg).
Chromatography on a silica gel (8 g) column with 2:1
hexane±ethyl acetate afforded saframycin A (4.2 mg,
14.9% recovery). Further elution with 1:1 hexane±ethyl
acetate afforded 6 (17.5 mg, 57.6%) as pale yellow
amorphous powder, whose spectra were identical with
those of an authentic sample described above.
4. Total synthesis of 2, see: (a) Fukuyama, T.; Sachleben, R. A.
J. Am. Chem. Soc. 1982, 104, 4957±4958. (b) Myers, A. G.; Kung,
D. W. J. Am. Chem. Soc. 1999, 121, 10828±10829. (c) Martins,
E. J.; Corey, E. J. Org. Lett. 1999, 1, 75±77.
5. (a) Martinez, E. J.; Owa, T.; Schreiber, S. L.; Corey, E. J. Proc.
Natl. Acad. Sci., USA 1999, 96, 3496±3501. (b) Martinez, E. J.;
Corey, E. J. Org. Lett. 2000, 2, 993±996. also, see; (c) Corey, E. J.;
Gin, D. Y.; Kania, R. S. J. Am. Chem. Soc. 1996, 118, 9202±9203.
(d) Cuevas, C.; Perez, M.; Martin, M. J.; Chicharro, J. L.;
Fernandez-Rivas, C.; Flores, M.; Francesch, A.; Gallego, P.;
Zarzuelo, M.; Calle, F. de la; Garcia, J.; Polanco, C.; Rodriguez,
I.; Manzanares, I. Org. Lett. 2000, 2545±2548.
Acknowledgements
This work was supported in part by a Grant-in-Aid for
Scienti®c Research from the Ministry of Education,
Science, Sports, and Culture, Japan. We are grateful to
Professor Emeritus T. Arai of Chiba University for kindly
providing a sample of natural saframycin R. We acknowl-
edge Dr K. Koyama of Meiji Pharrmaceutical University for
valuable discussions. We would like thank Ms T. Kozeki
and Ms A. Ohmae in the Analytical Center of this
University for measurements of mass spectral data and
NMR data.
6. (a) Asaoka, T.; Yazawa, K.; Mikami, Y.; Arai, T.; Takahashi,
K. J. Antibiot. 1982, 35, 1708±1710. (b) Lown, J. W.; Hanstock,
C. C.; Joshua, A. V.; Arai, T.; Takahashi, K. J. Antibiot. 1983, 36,
1184±1710.
7. (a) Kubo, A.; Saito, N.; Kitahara, Y.; Takahashi, K.; Yazawa,
K.; Arai, T. Chem. Pharm. Bull. 1987, 35, 440±442. (b) Saito, N.;
Harada, S.; Nishida, M.; Inouye, I.; Kubo, A. Chem. Pharm. Bull.
1995, 43, 777±782.
References
8. Bax, A.; Summers, M. F. J. Am. Chem. Soc. 1986, 108, 2093±
2094.
1. Reviews: (a) Arai, T.; Kubo, A. In The Alkaloids; Brossi, A.,
Ed.; Academic: New York, 1983; Vol. 21, pp 55±100. (b) Tomson,
R. H. Naturally Occurring Quinone III; Chapman and Hall: New
9. Saito, N.; Ohira, Y.; Wada, N.; Kubo, A. Tetrahedron 1990, 46,
7711±7728.