A new efficient synthetic process for the construction of the pentacyclic core of marine alkaloid ecteinascidins

Article abstract of DOI:10.1016/S0040-4039(03)01785-4

The pentacyclic core of ecteinascidins was constructed from two fundamental building blocks, the 1,2,3,4-tetrahydroisoquinoline derivative and the substituted phenylalanine derivative, via 8 steps using readily available L-Dopa as starting material. The k

Full text of DOI:10.1016/S0040-4039(03)01785-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 7091–7094
A new efficient synthetic process for the construction of the
pentacyclic core of marine alkaloid ecteinascidins
Ye-Feng Tang, Zhan-Zhu Liu* and Shi-Zhi Chen
Institute of Materia Medica, Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing 100050,
PR China
Received 13 May 2003; revised 27 June 2003; accepted 14 July 2003
Abstract—The pentacyclic core of ecteinascidins was constructed from two fundamental building blocks, the 1,2,3,4-tetra-
hydroisoquinoline derivative and the substituted phenylalanine derivative, via 8 steps using readily available L-Dopa as starting
material. The key steps involve coupling of the two aforementioned building blocks, regioselective reduction of the 11-carbonyl
group of the key intermediate piperazine-1,4-dione derivative, and intramolecular Pictet–Spengler cyclization.
© 2003 Elsevier Ltd. All rights reserved.
Ecteinascidins are a family of marine-derived tetra-
hydroisoquinoline alkaloids, which possess extremely
potent antitumor activity.1 Ecteinascidin 743 (Et 743,
1), the most abundant and bioactive of this family, is
being studied in phase I/II clinical trials.2 Much syn-
thetic effort has been directed towards the synthesis of
Et 743,3 but due to its formidable structural complexity,
there have been only three successful results reported so
far by Corey, Cuevas and Fukuyama groups, respec-
tively.4 Fortunately, phthalascidin (Pt 650, 2), a struc-
turally simplified version of 1, was found to exhibit
comparable antitumor activity to that of Et 743 and
may therefore be a more practical therapeutic agent.5 In
view of the structural feature, the cores of both 1 and 2
are characterized by a bis(tetrahydroisoquinoline) pen-
tacyclic system (ABCDE), which is believed to be the
primary pharmacophore through the preliminary struc-
ture–activity study of Et 743 and its analogs (Fig. 1).5
This inspired us to start a research project on the
discovery of structurally simplified, synthetically acces-
sible and biologically active ecteinascidin analogs as
antitumor agents. In this communication, we report the
synthesis of 3 that bears the same pentacyclic frame-
work as that of 1 and 2 via a new efficient synthetic
route, which we believe is of considerable value in the
synthesis of simplified ecteinascidin analogs.
Figure 1.
Keywords: ecteinascidins; pentacyclic core; L-Dopa; Pictet–Spengler cyclization.
* Corresponding author. Tel.: 86-10-63165253; fax: 86-10-63017757; e-mail: liuzhanzhu@imm.ac.cn
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)01785-4

7092
Y.-F. Tang et al. / Tetrahedron Letters 44 (2003) 7091–7094
Retrosynthetic analysis of the pentacyclic core of
ecteinascidins leads us to employ an original synthetic
strategy different from those reported.3 According to
our design, the BCD-ring system could be constructed
sequentially (Scheme 1). Thus, 4 could be derived from
5 through reduction of the 21-carbonyl group followed
by replacement of the hydroxyl group with the cyanide
group. Compound 5 could be obtained from 6 via an
0.5N H2SO4 to yield the 6,7-dihydroxy-1,2,3,4-tetra-
hydroisoquinoline-3-carboxylic acid 14. Of particular
note, formaldehyde was chosen here for convenience.
We have investigated a series of aliphatic and aromatic
aldehydes in this reaction to produce the corresponding
1-substituded tetrahydroisoquinoline derivatives. In
each case, the 1,3-cis isomer was obtained as the major
product.9 Thus, it is feasible to introduce a desired
stereocenter on the 1-postion of the pentacyclic core via
an asymmetric Pictet–Spengler reaction when needed.
Esterification of 14 in HCl/CH3OH at reflux provided
15. Protection of the nitrogen of 15 by a formyl group
afforded 16, which was transformed into the methyla-
tion product 17 upon treatment with Me2SO4/K2CO3 in
acetone at reflux. Cleavage of the formyl group in
HCl/CH3OH provided another building block 18.10 The
overall yield of 18 from L-Dopa is ca. 50% (Scheme 3).
intramolecular Pictet–Spengler cyclization, and
6
maybe accessed by regioselective reduction of the 11-
carbonyl group of the piperazine-1,4-dione derivative 7,
a key intermediate in our synthesis. Compound 7, as we
assumed, could be derived from two fundamental build-
ing blocks, the 1,2,3,4-tetrahydroisoquinoline derivative
8 and the substituted phenylalanine derivative 9, via
formation of two amide bonds. Both 8 and 9 could be
easily obtained from the substituted phenylalanine 10
by several functional group manipulations.
Coupling of 13 and 18 to form 19 was accomplished
through the action of bis(2-oxo-3-oxazolidinyl) phos-
phinic chloride (BOP-Cl) in 80% yield. Then 19 was
treated with CF3CO2H in CH2Cl2 to afford the key
intermediate piperazine-1,4-dione derivative 20, which
was converted to the carbamate 21 in 80% yield accord-
ing to a known procedure.11 Regioselective reduction of
11-carbonyl group of 21 with an excess amount of
Li(t-BuO)3AlH in THF provided the diastereomeric
mixture of the alcohol 22, which, on treatment with
HCO2H at 70°C, yielded exclusively the expected penta-
cyclic framework of ecteinascidins 23. It is worth noting
that a compound similar to 23 was committed to the
same cyclization condition by another research team,
Partly enlightened by the biosynthetic study on
ecteinascitins,6 which supposed that the pentacyclic core
of ecteinascidins was derived from two molecules of
L-Dopa or L-tyrosine, we chose the commercially avail-
able L-Dopa as the starting material. Thus, protection
of the amino group of L-Dopa by BOC provided 11,
which was treated with an excess of Me2SO4/K2CO3 in
acetone at reflux to give 12.7 Then hydrolysis of 12 in
1N NaOH afforded 13 as one of the two building
blocks. The overall yield of 13 from L-Dopa was 70%
(Scheme 2).
According to a known method,8 L-Dopa was submitted
to Pictet–Spengler reaction using 37% HCHO in
Scheme 1.
Scheme 2. Reagents and conditions: (a) (Boc)2O, Et3N, DMF, 60°C, 89%; (b) Me2SO4, K2CO3, CH3COCH3, reflux, 92%; (c) 1N
KOH/CH3OH, 90%.

Y.-F. Tang et al. / Tetrahedron Letters 44 (2003) 7091–7094
7093
Scheme 3. Reagents and conditions: (a) HCHO, 0.5N H2SO4, 24 h, 75%; (b) HCl/CH3OH, reflux, 6 h, 99%; (c) Ac2O/HCO2H, 4
h; then H2O/CH3OH, 87%; (d) Me2SO4, K2CO3, CH3COCH3, reflux, 90%; (e) HCl/CH3OH, reflux, 4 h, 85%.
Scheme 4. Reagents and conditions: (a) BOP-Cl, Et3N, CH2Cl2, 4 h, 80%; (b) CF3COOH/CH2Cl2, 2 h, 86%; (c) ClCO2CH(CH3)2,
DMAP, Et3N, CH2Cl2, 6 h, 80%; (d) Li(t-BuO)3AlH, THF, 0°C, 2 h, then H2O, 90%; (e) HCO2H, 70°C, 3 h, 65%; (f)
H2SO4/CF3CO2H, 15 h, 98%; (g) HCHO/HCO2H, 70°C, 2 h, 95%; (h) LiAlH4, THF, −17°C, 0.5 h, then 0°C, 1 h; KCN,
phosphate buffer (pH=7), 2 h, 82%.
but an unsuccessful result was reported.12 Next, depro-
tection of 23 with a mixture of CF3CO2H and H2SO4
gave the secondary amine 24 in quantitative yield.
Reductive methylation of 24 with HCHO/HCO2H at
70°C for 2 h provided 25 in 95% yield. Finally, the
lactam ring of 25 could be easily reduced through
treatment with an excess of LiAlH4 in THF at 0°C for
l h to afford the corresponding cyclic hemiaminal,
which upon exposure to KCN in phosphate buffer
(pH=7) afforded the pentacyclic amino nitrile 3 as an
enantiomerically pure product in 82% overall yield
from 25 (Scheme 4).
was observed between H-3 and H-11, thus the syn
C3–C11 backbone stereochemical relationship was
established. Similarly, irradiation of the H-21 produced
noticeable NOE enhancement of Hb-1,but not of Ha-
1,which indicated that H-21 was of b-configuration
Gratifyingly, 3 does not only possess the desired penta-
cyclic core of ecteinascidins, but also bears four com-
pletely correct stereocenters (3S,11S,13S,21R) as those
of the natural products. This conclusion was verified on
the basis of its spectroscopic data,13 especially the NOE
difference spectroscopy. Obvious NOE enhancement
Figure 2.

7094
Y.-F. Tang et al. / Tetrahedron Letters 44 (2003) 7091–7094
(Fig. 2). These data are consistent with the stereochem-
ical relationship of the four stereocenters.
4. (a) Corey, E. J.; Gin, D. Y.; Kania, R. S. J. Am. Chem.
Soc. 1996, 118, 9202–9203; (b) Cuevas, C.; Pe´rez, M.;
Mart´ın, M. J.; Chicharro, J. L.; Ferna´dez-Rivas, C.;
Flores, M.; Francesch, A.; Gallego, P.; Zarzuelo, M.; de
la Calle, F.; Garc´ıa, J.; Polanco, C.; Rodr´ıguez, I.; Man-
zanares, I. Org. Lett. 2000, 2, 2545–2548; (c) Endo, A.;
Yanagisawa, A.; Abe, M.; Tohma, S.; Kan, T.;
Fukuyama, T. J. Am. Chem. Soc. 2002, 124, 6552–6554.
5. Martinez, E. J.; Owa, T.; Schreiber, S. L.; Corey, E. J.
Proc. Natl. Acad. Sci. 1999, 96, 3496–3501.
In summary, we have developed a new efficient syn-
thetic route for the construction of the pentacyclic core
of ecteinascidins starting from L-Dopa. Further study
on the synthesis of structurally simplified and bioactive
ecteinascidin analogs based on this methodology is
ongoing in our laboratory.
6. (a) Kerr, R. G.; Miranda, N. F. J. Nat. Prod. 1995, 58,
1618–1621; (b) Sakai, R.; Jares-Erijman, E. A.; Manza-
nares, I.; Elipe, M. V. S.; Rinehart, K. L. J. Am. Chem.
Soc. 1996, 118, 9017–9023.
References
7. Garcia´, E.; Arrasate, S.; Ardeo, A.; Lete, E.; Sotomayor,
1. For a review, see: Scott, J. D.; Williams, R. M. Chem.
Rev. 2002, 102, 1669–1730.
N. Tetrahedron Lett. 2001, 42, 1511–1513.
8. Brossi, A.; Focella, A.; Teitel, S. Helv. Chim. Acta 1972,
55, 15–21.
9. Wang, Y.; Liu, Z. Z.; Chen, S. Z.; Liang, X. T. Chin.
Chem. Lett. 2003, in press.
10. Dean, R. T.; Rapoport, H. J. Org. Chem. 1978, 43,
4183–4189.
2. For clinical data, see: (a) Ryan, D. P.; Supko, J. G.;
Eder, J. P.; Seiden, M. V.; Demetri, G.; Lynch, T. J.;
Fischman, A. J.; Davis, J.; Jimeno, J.; Clark, J. W. Clin.
Cancer Res. 2001, 7, 231–242; (b) Zelek, L.; Yovine, A.;
Etienne, B.; Jimeno, J.; Taamma, A.; Martin, C.; Spiel-
mann, M.; Cvitkovic, E.; Misset, J. L. The American
Society of Clinical Oncology, 36th Annual Meeting, New
Orleans, May 20–23, 2000; Abstract number 592; (c)
Delaloge, S.; Yovine, A.; Taamma, A.; Cottu, P.; Riofrio,
M.; Raymond, E.; Brain, E.; Marty, M.; Jimeno, J.;
Cvitkovic, E.; Misset, J. L. The American Society of
Clinical Oncology, 36th Annual Meeting, New Orleans,
May 20–23, 2000; Abstract number 2181; (d) Le Cesne,
A.; Judson, I.; Blay, J. Y.; Radford, J.; van Oosterom,
A.; Lorigan, P.; Rodenhuis, E.; Donato Di Paoula, E.;
Van Glabbeke, M.; Jimeno, J.; Verweij, J. The American
Society of Clinical Oncology, 36th Annual Meeting, New
Orleans, May 20–23, 2000; Abstract number 2182; (e)
Aune G. J.; Furuta T.; Pommier Y. Anti-cancer Drugs
2002, 13, 545–555.
11. Kubo, A.; Saito, N.; Yamato, H.; Masubuchi, K.; Naka-
mura, M. J. Org. Chem. 1988, 53, 4295–4310.
12. Fukuyama, T.; Sacheben, R. A. J. Am. Chem. Soc. 1982,
104, 4957–4958.
13. Analytical and spectroscopic data for 3: light yellow
solid; mp 124.5–127°C; [h]2D0=+19.3 (c 0.4, CHCl3); 1H
NMR (300 Hz, CDCl3): l 2.35 (s, 3H, -NCH3), 2.64–2.55
(m, 3H, H-4+H-14), 3.14 (dd, 1H, J=8.4, 18.0 Hz, H-4),
3.24–3.18 (ddd, 1H, J=3.0, 4.8, 8.4 Hz, H-3), 3.47 (d,
1H, J=8.1 Hz, H-13), 3.59 (d, 1H, J=3.0 Hz, H-11),
3.71 (d, 1H, J=15.0 Hz, Hb-1), 3.78–3.81 (m, 9H, 3×
CH3O), 3.87 (d, 1H, J=15.0Hz, Ha-1), 3.90 (s, 3H,
CH3O), 3.94 (d, 1H, J=2.1 Hz, H-21), 6.43 (s, 1H,
Ar-H), 6.48 (s, 1H, Ar-H), 6.52 (s, 1H, Ar-H), 6.57 (s,
1H, Ar-H); 13C NMR (300 Hz, CDCl3): l 25.7, 31.6,
42.0, 54.2, 55.4, 55.8, 55.9, 55.9, 56.3, 56.3, 62.2, 62.4,
108.6, 110.8, 110.9, 113.5, 117.2, 123.5, 123.8, 124.3,
126.1, 146.5, 147.4, 147.7, 148.4; IR (KBr, cm−1): 2220
(CN, w), 1660 (CꢀO, s); FAB-MS (m/z): 436 (M+1,
27%), 435 (M, 10%), 409 (22%), 204 (100%); HRMS
(FAB): calcd for C25H30N3O4 (M+1) 436.2236; found
435.2214.
3. For some recent examples, see: (a) Saito, N.; Tachi, M.;
Seki, R.; Kamayachi, H.; Kubo, A. Chem. Pharm. Bull.
2000, 48,1549–1557; (b) Endo, A.; Kann, T.; Fukuyama,
T. Synlett 1999, 1103–1105; (c) Zhou, B.; Guo J.; Dan-
ishefsky, S. J. Tetrahedron Lett. 2000, 41, 2043–2046; (d)
Martinez, E. J.; Corey, E. J. Org. Lett. 2000, 2, 993–996;
(e) Jin, W.; Metobo, S.; Williams, R. M. Org. Lett. 2003,
in press.