Regioselectivity of Pictet-Spengler cyclization reactions to construct the pentacyclic frameworks of the ecteinascidin-saframycin class of tetrahydroisoquinoline antitumor antibiotics

Article abstract of DOI:10.1016/j.tetlet.2007.03.113

The regiochemical outcome of Pictet-Spengler cyclization reactions directed toward the preparation of the pentacyclic core of the ecteinascidin class of antitumor antibiotics has been investigated on two different phenolic substrates. In one substrate, th

Full text of DOI:10.1016/j.tetlet.2007.03.113

Tetrahedron Letters 48 (2007) 3719–3722
Regioselectivity of Pictet–Spengler cyclization reactions
to construct the pentacyclic frameworks of the
ecteinascidin–saframycin class of tetrahydroisoquinoline
antitumor antibiotics
Guillaume Vincent, Jonathan W. Lane and Robert M. Williams*
Department of Chemistry, Colorado State University, Fort Collins, CO 80523, USA
University of Colorado Cancer Center, Aurora, CO 80045, USA
Received 2 January 2007; revised 19 March 2007; accepted 20 March 2007
Available online 24 March 2007
Abstract—The regiochemical outcome of Pictet–Spengler cyclization reactions directed toward the preparation of the pentacyclic
core of the ecteinascidin class of antitumor antibiotics has been investigated on two different phenolic substrates. In one substrate,
the assistance of an incipient benzylamine group at C-4 is postulated to direct the cyclization in favor of the pentacyclic framework
of ET-743, which bears a hydroxyl group at C-18. Conversely, cyclization of an alternative substrate lacking a heteroatom at C-4
favors the opposite regiochemical outcome, primarily affording an unnatural pentacyclic core bearing a hydroxyl group at C-16.
Ó 2007 Elsevier Ltd. All rights reserved.
MeO
In 1990, the isolation of the ecteinascidins, a new family
of tetrahydroisoquinoline alkaloids1 was reported from
the marine tunicate Ecteinascidin tubinata.2 Among
these compounds, ecteinascidin 743 (Et-743, YondelisTM,
1) is presently in phase II/III clinical trials due to its
exceedingly low nanomolar activity against different
tumor cell lines (Fig. 1).3,4 It has been demonstrated that
Et-743 (1) binds to guanine in the minor groove of
DNA at the reactive C-21 carbinolamine position. The
DNA–drug adduct exhibits a bend or widening in the
minor groove of the DNA component, presumably
due to the steric requirements of the GH-subunit of
Et-743 (1). It has been postulated that this bend in
DNA disrupts DNA–protein binding and may be, in
part, a foundation for the enhanced activities of Et-
743 over related agents. Recently, Et-743 was demon-
strated to halt the DNA nucleotide excision repair
(NER) mechanism in cells.4 The restricted natural avail-
ability of Et-743 (1) from the tunicate has been over-
come through semi-synthesis of multigram quantities
from the more readily available metabolite cyanosafra-
cin B (2a). Cyanosafracin B (2a) is available in kilogram
Me
17
16
15
HO
H 18
HO
O
G
H
H
N
4
NH
5
MeO
Me
MeO
Me
11
21
NMe
6
H3
O
14
F
O
HO
H
8
E
7
13
MeO
H
S
1
OAc
A
H
N
O
22
R
Me
NMe
H
C
NH
NH2
O
O
O
H
OH
Me
2a, Cyanosafracin B; R=CN
2b, Safracin A; R=H
1, Ecteinascidin 743
MeO
2c, Safracin B; R=OH
Me
HO
H
OAc
H
N
Me
O
NMe
H
H
O
CN
Nphth
3, Phtalascidin 650
Figure 1. Ecteinascidin 743, the safracins and phtalascidin 650.
quantities through bacterial fermentation. The route,
developed by Cuevas and Manzanares at PharmaMar,
supports current clinical trials.5 Closely related to the
ecteinascidins, safracins 2a–c were isolated in 1983 from
Pseudomonas fluorescens and also display potent antimi-
crobial and antitumor activities.6
*
Corresponding author. Tel.: +1 970 491 6747; fax: +1 970 491
3944; e-mail addresses: rmw@chem.colostate.edu; rmw@lamar.
colostate.edu
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.03.113

3720
G. Vincent et al. / Tetrahedron Letters 48 (2007) 3719–3722
OTBS
The structural challenges associated with Et-743 (1)
have attracted significant attention from the synthetic
community. Corey reported the first total synthesis of
1 in 1996.7 Subsequently, Corey and Schreiber reported
a simplified analogue of 1, phthalascidin 650 (Pt-650,
3),8 which exhibited comparable biological activity.
Fukuyama9 and Zhu10 have also completed the total
synthesis of Et-743 (1), and Danishefsky has described
a formal synthesis.11 Additionally, Zhu has prepared
the putative biosynthetic precursors of 1, Et-597, and
Et-583,12 and Kubo has published synthetic studies
toward safracins 2.13
OMe
Me
OTBS
OMe
(COCl)2, DMF
CH2Cl2;
Fmoc
H
NMe
NH
+
MeO
H
2,6-lutidine
HO2C
OAllyl
CH2Cl2, 71%
Me
8
9
OAllyl
OTBS
TBSO
OMe
1) HCO2H, H2O, THF
OMe
Me
Me
Fmoc
77%
NMe
H
N
H
2) DMSO, (COCl)2
CH2Cl2 then Et3N
91%
MeO
OAllyl
O
OAllyl
10
OMe
TBSO
TBAF, CH2Cl2
then MeSO3H
CHO
OMe
Me
Fmoc
We previously described a reductive opening/Pictet–
Spengler cyclization cascade sequence from b-lactam 4
to form the C3–C4 unsaturated pentacycle 5 (Scheme
1, route A).14 Recently, this strategy was successfully
employed in our total synthesis of cribrostatin 4.15 We
conjectured that the C3–C4 alkene of 5 was also flexibly
poised to deliver either Et-743 (1), safracins 2, or Pt-650
(3). Specifically, functionalization of the alkene group in
5 would ultimately permit formation of the macrocyclic
ring of Et-743 through closure at C-4. Alternatively, sat-
uration of the alkene group in 5 could deliver the core of
the safracins and Pt-650. The C3–C4 saturated pentacy-
clic skeleton of the safracins 6 could also be obtained
through a Pictet–Spengler cyclization of the C3–C4 sat-
urated amino aldehyde 7 (route B).16
NMe
H
N
H
Me
MeO
OAllyl
O
OAllyl
11
MeO
18
MeO
OH
+
Me
NOESY
OMe
Me
H
16
HO
H
NOESY
11
OMe
H
15
H
N
H
N
Me
Me
NMe
H
3
NMe
H
H
H
MeO
MeO
22
OAllyl
O
OAllyl
O
OAllyl
OAllyl
Fremy's salt
KH2PO4
12 (62%)
13 (19%)
major
minor
H2O, CH3CN
MeO
MeO
Me
Me
O
HO
H
O
H
OMe
OMe
DDQ
OMe
H
N
H
Me
MeO
Me
NMe
NMe
H
H
H
In an effort to investigate the preparation of 6 from 7
(Scheme 1, route B), tetrahydroisoquinoline 8, repre-
senting the western half of 7, was coupled with the acid
chloride derived from amino acid 9, representing the
eastern half of 7, to deliver 10 (Scheme 2). Selective
deprotection of the primary alcohol group in 10 with
formic acid and water was followed by a Swern oxida-
tion to afford the desired aldehyde 11. With 11 in hand,
the stage was set to form the pentacyclic core of 2 and 3
through the key Pictet–Spengler reaction. Simultaneous
removal of the N-Fmoc and O-TBS groups in 11 was
H2O, CH3CN
N
H
MeO
OAllyl
O
OAllyl
OAllyl
O
15
OAllyl
14
Scheme 2. Pictet–Spengler cyclization of C3–C4 saturated aldehyde
11.
accomplished with TBAF resulting in an iminium ion
intermediate, after which methanesulfonic acid was then
added in situ to promote subsequent ring-closure and
pentacycle formation. Interestingly, this transformation
resulted in the formation of two C-3 epi pentacyclic
regioisomers, 12 and 13, in a 3.3:1 ratio, respectively,
Bn
OMe
N
HO
NHMe
H
O
OMe
Me
Me
1
based on H NMR analysis. The unusual epimerization
H
N
MeO
at C-3 was consistent with observations from our prior
total syntheses of 3-epi-renieramycin G, renieramycin
G, 3-epi-jorumycin and jorumycin.16 Furthermore, for-
mation of E-ring regioisomer 12 as the major product
indicated that the cyclization had preferentially occurred
para to the phenolic group in 11. Conversely, the minor
product 13 was derived from cyclization ortho to the
phenolic group in 11. The regiochemistries of 12 and
13 were secured by NOESY correlations between H-11
and the C-18 Me-group and between H-15 and the
C-16 Me-group, respectively. The trans-relationship
between H-3 and H-11 in 12 and 13 was verified by
OP
O
OP
4
MeO
A
Me
HO
H
OMe
Me
NMe
H
2a-c, 3
1
H
N
MeO
OP
O
OP
MeO
5
Me
HO
H
OMe
OMe
HO
H
N
CHO
OMe
Me
Me
Me
NHMe
1H NMR through coupling constants (JH3–H11
=
NMe
H
H
H
N
H
0 Hz). Moreover, the structure of the minor product
12 was further verified through oxidation with Fremy’s
salt to quinone 14, which was also successfully generated
from compound 15, an intermediate in the synthesis of
3-epi-jorumycin.16
MeO
MeO
B
OP
O
OP
O
OP
OP
6
7
Scheme 1. Retrosynthetic analysis.

G. Vincent et al. / Tetrahedron Letters 48 (2007) 3719–3722
3721
The observed epimerization at C3 in 11 could poten-
tially be avoided by pre-forming a hemiaminal unit
through the aldehyde and amine groups, prior to con-
ducting the Pictet–Spengler cyclization. This strategy
was successfully demonstrated in our syntheses of joru-
mycin and renieramycin G.16 Unfortunately, however,
the regiochemical outcome afforded through the cycliza-
tion of 11 remains unsuitable for the synthesis of the
safracins or Pt-650. Historically, methods have been suc-
cessfully devised to selectively access pentacyclic inter-
mediates bearing the appropriate E-ring regiochemical
relationship for the ecteinascidin family of natural prod-
ucts. In their studies toward Et-743 (1) Corey and
co-workers employed a symmetric 3,5-dihydroxy aryl
residue, representing the E-ring, to circumvent regiose-
lectivity issues intrinsic to mono-hydroxy phenol cycliza-
tion substrates. Subsequent differentiation of the less
hindered hydroxy position through activation as the
corresponding triflate group followed by cross-coupling
with SnMe4 afforded the correctly functionalized E
ring.7 Alternatively, Kubo and Fukuyama utilized an
E-ring precursor bearing a bromide group para to the
phenolic hydroxyl group, which successfully blocked
this position and promoted exclusive ortho cyclization.17
However, despite the elegance and utility of these strat-
egies, an enhanced understanding of the factors influ-
encing the regiochemical outcome of cyclizations of
substrates capable of affording both ortho- and para-
products would certainly be useful and would perhaps
obviate the need for blocking groups or post-cyclization
manipulations at the E-ring unit.
Therefore, an effort was made to rationalize these find-
ings. Initially, a more thorough analysis of the reductive
opening/Pictet–Spengler cyclization sequence confirmed
the production of two regioisomers, 17 and 18, in a 4:1
ratio, respectively, based on 1H NMR analysis.
Although the structure of compound 17 was originally
1
secured through H NMR NOE studies,14 additional
1H NMR NOESY experiments further verified the cor-
relation between H-15 and the C-16 Me-group. The
structure of compound 17 was unequivocally confirmed
through oxidation with Fremy’s salt to quinone 19,
which was identical to the quinone obtained by PIFA
oxidation of 20, an intermediate in our synthesis of crib-
rostatin 4.15 The structure of the minor regioisomer 18
was also secured by NOESY experiments (correlation
between H-11 and the C-18 Me-group).
Successful and rigorous verification of our previously
reported results,14 lead us to propose the following
pathway to explain the two distinct cyclization out-
comes (Scheme 4). Partial reduction of the b-lactam unit
in 16 would lead to the coordinated lithium complex 21,
which also serves to prevent over-reduction of the corre-
sponding aldehyde group. Following aqueous quench
with NH4Cl, condensation of the secondary amine
group onto the hemi-aminal unit in 21 would generate
the key iminium ion intermediate 22. Within intermedi-
ate 22, a postulated hydrogen bonding interaction be-
tween the C-4 benzylic amine and the E-ring phenolic
residue would direct the Pictet–Spengler cyclization
ortho- to the phenol. Finally, spontaneous elimination
of the C-4 benzylamine group occurs through the inter-
mediacy of ortho-quinone methide 23 to afford the C3–
C4 unsaturated pentacycle 17 as the major product. The
minor pentacycle regioisomer 18 (Scheme 3) may arise
from a pathway whereby C-4 benzylamine elimination,
potentially through intermediate 22, occurred prior to
ring-closure onto the incipient iminium ion.
The preference for the formation of compound 12 from
the Pictet–Spengler cyclization of 11 was not consistent
with the regiochemical outcome observed in our previ-
ously reported reductive opening/Pictet–Spengler cycli-
zation sequence through b-lactam 16 (Scheme 3).14,18
In summary, we have postulated the assistance of an
incipient benzylamine group at C-4, which was derived
from b-lactam 16 in a reductive opening/Pictet–Spengler
reaction sequence, to direct the formation of the
pentacyclic framework of the ecteinascidin family of
Bn
OMe
N
HO
NHMe
O
OMe
Me
Me
LiEt3BH, THF
H
N
H
then NH4Cl aq.
MeO
OBn
O
OBn
16
NOESY
MeO
MeO
Me 18
H
Me
OH
16
NOESY
HO
H
Bn
OMe
Li
OMe
H
OMe
15
N
O
11
HO
Me
LiEt3BH, THF
Me
NMe
H
+
OMe
Me
Me
NH4Cl aq,
NMe
16
H
NHMe
H
N
H
then NH4Cl aq.
N
MeO
N
H
MeO
H
MeO
OBn
17
O
OBn
O
OBn
O
OBn
OBn
major (40%)
OBn
18
minor (9%)
21
Fremy's salt
KH2PO4
MeO
H O
MeO
Me
Me
MeO
MeO
HO
H
H2O, CH3CN
Bn
OMe N
HO
H
Me
Me
OMe
H
OMe
O
H
H
Me
MeO
OMe
Me
Me
MeO
17
OMe
O
NMe
H
NMe
H
H
H
Me
PIFA
H2O
N
NMe
H
N
NMe
H
H
H
N
N
OBn
O
OBn
MeO
OBn
OBn
23
O
OBn
MeO
CH3CN
O
OBn
O
22
OBn
OBn
19
20
Scheme 3. Reductive opening/Pictet–Spengler sequence from b-lactam
16.
Scheme 4. Postulated pathway for the reductive opening/Pictet–
Spengler sequence of b-lactam 16.

3722
G. Vincent et al. / Tetrahedron Letters 48 (2007) 3719–3722
´
Rodrıguez, I.; Manzanares, I. Org. Lett. 2000, 2, 2545–
antitumor antibiotics. In the absence of a heteroatom at
C-4 (compound 11), Pictet–Spengler cyclization resulted
in the preferential formation of an unnatural regio-
isomer at the E-ring position. The transformation of
17 into the safracins and the ecteinascidins is currently
under investigation in our group.19 We are also pursuing
the introduction of a C-4 hydroxyl group into our Pic-
tet–Spengler cyclization substrate prior to forming the
pentacyclic core of Et-743.20
´
2548; (b) Menchaca, R.; Martinez, V.; Rodrıguez, A.;
´
Rodrıguez, N.; Flores, M.; Gallego, P.; Manzanares, I.;
Cuevas, C. J. Org. Chem. 2003, 68, 8859–8866.
6. (a) Ikeda, Y.; Idemoto, H.; Hirayama, F.; Yamamoto, K.;
Iwao, K.; Asao, T.; Munakata, T. J. Antibiot. 1983, 36,
1279–1283; (b) Ikeda, Y.; Matsuki, H.; Ogawa, T.;
Munakata, T. J. Antibiot. 1983, 36, 1284–1289; (c) Ikeda,
Y.; Shimada, Y.; Yoshiko, H.; Katsuhiko, O.; Munakata,
T. J. Antibiot. 1983, 36, 1290–1294.
7. (a) Corey, E. J.; Gin, D. Y.; Kania, R. S. J. Am. Chem.
Soc. 1996, 118, 9202–9203; (b) Martinez, E. J.; Corey, E. J.
Org. Lett. 2000, 2, 993–996.
Acknowledgments
8. Martinez, E. J.; Owa, T.; Schreiber, S. L.; Corey, E. J.
Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 3496–3501.
9. Endo, A.; Yanagisawa, A.; Abe, M.; Tohma, S.; Kan, T.;
Fukuyama, T. J. Am. Chem. Soc. 2002, 124, 6552–
6554.
We gratefully acknowledge financial support from the
National Institutes of Health (Grant CA85419) and
the NIH National Cancer Institute for postdoctoral fel-
lowship support for J.W.L. (Grant CA105898). Mass
spectra were obtained on instruments supported by the
National Institutes of Health Shared Instrumentation
Grant No. GM49631.
10. Chen, J.; Chen, X.; Bois-Choussy, M.; Zhu, J. J. Am.
Chem. Soc. 2005, 128, 87–89.
11. (a) Zheng, S.; Chan, C.; Furuuchi, T.; Wright, B. J. D.;
Zhou, B.; Guo, J.; Danishefsky, S. J. Angew. Chem., Int.
Ed. 2006, 45, 1754–1759; (b) Chan, C.; Zheng, S.; Zhou,
B.; Guo, J.; Heid, R. M.; Wright, B. J. D.; Danishefsky, S.
J. Angew. Chem., Int. Ed. 2006, 45, 1749–1754.
12. Chen, J.; Chen, X.; Willot, M.; Zhu, J. Angew. Chem., Int.
Ed. 2006, 45, 8028–8032.
13. (a) Saito, N.; Obara, Y.; Aihara, T.; Harada, S.; Shida, Y.;
Kubo, A. Tetrahedron 1994, 50, 3915–3928; (b) Saito, N.;
Harada, S.; Yamashita, M.; Saito, T.; Yamaguchi, K.;
Kubo, A. Tetrahedron 1995, 51, 8213–8230.
Supplementary data
Supplementary data (complete experimental procedures
and NMR spectral data) associated with this article can
be found, in the online version, at doi:10.1016/
j.tetlet.2007.03.113.
14. Jin, W.; Metobo, S.; Williams, R. M. Org. Lett. 2003, 5,
2095–2098.
References and notes
15. Vincent, G.; Williams, R. M. Angew. Chem., Int. Ed. 2007,
46, 1517–1520.
16. Lane, J. W.; Chen, Y.; Williams, R. M. J. Am. Chem. Soc.
2005, 127, 12684–12690.
1. Scott, J. D.; Williams, R. M. Chem. Rev. 2002, 102, 1669–
1730.
17. (a) Saito, N.; Tachi, M.; Seki, R.; Kamayachi, H.; Tachi,
M.; Kubo, A. Chem. Pharm. Bull. 2000, 48, 1557; (b)
Saito, N.; Kamayachi, H.; Tachi, M.; Kubo, A. Hetero-
cycles 1999, 51, 9–12; (c) Endo, A.; Kan, T.; Fukuyama, T.
Synlett 1999, 1103–1105.
18. Zhu and co-workers also observed a similar difference of
regiochemistry during their studies toward Et-743: see Ref.
10 and Chen, X.; Chen, J.; De Paolis, M.; Zhu, J. J. Org.
Chem. 2005, 70, 4397–4408.
19. Danishefsky and co-workers introduced a hydroxyl group
at C-4 on a similar benzylic alkene by treatment with
DMDO followed by NaBH3CN: see Ref. 11a.
20. Fishlock, D.; Williams, R. M. Org. Lett. 2006, 8, 3299–
3301.
2. 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.
3. (a) Aune, G. J.; Furuta, T.; Pommier, Y. Anti-Cancer
Drug. 2002, 13, 545–555; (b) Drugs R. D. 2006, 7, 317–328.
4. (a) Zewail-Foote, M.; Li, V.; Kohn, H.; Bearass, D.;
Guzman, M.; Hurley, L. H. Chem. Biol. 2001, 135, 1–17;
(b) Takebayashi, Y.; Pourquier, P.; Zimonjic, D. B.;
Nakayama, K.; Emmert, S.; Ueda, T.; Urasaki, Y.;
Kanzaki, A.; Akiyama, S.; Popescu, N.; Kraemer, K. H.;
Pommier, Y. Nat. Med. 2001, 7, 961–966.
´
´
5. (a) Cuevas, C.; Perez, M.; Martın, M. J.; Chicharro, J. L.;
´
Fernandez-Rivas, C.; Flores, M.; Francesch, A.; Gallego,
´
P.; Zarzuelo, M.; de la Calle, F.; Garcıa, J.; Polanco, C.;