A domino N-amidoacylation/aldol-type condensation approach to the synthesis of the topo-I inhibitor rosettacin and derivatives

Article abstract of DOI:10.1021/jo702387q

(Chemical Equation Presented) The pot, atom, and step-economic synthesis of Rosettacin topo-I poison and its derivatives has been achieved using a novel domino N-amidoacylation/aldol-type condensation, followed by decarboxylation of the ester function. Th

Full text of DOI:10.1021/jo702387q

A Domino N-Amidoacylation/Aldol-Type
Condensation Approach to the Synthesis of the
Topo-I Inhibitor Rosettacin and Derivatives†
Fre´de´ric Pin,‡ Se´bastien Comesse,‡ Morgane Sanselme,§ and
Adam Da¨ıch*,‡
URCOM, EA 3221, UFR des Sciences & Techniques, UniVersite´
du HaVre, 25 rue Philippe Lebon, BP: 540, F-76058
Le HaVre Cedex, France, and UPRES-EA 3233-IRCOF,
UniVersity of Rouen, 1 rue Tesnie`re F-76821,
Mont-Saint-Aignan Cedex, France
FIGURE 1. Camptothecins 1a-c and aromathecins 2a,b,g.
respectively.3 Other derivatives are in preclinical development
or clinical trial as exemplified by exatecan, 9-nitrocamptothecin,
and BAY 38-3441.4,5 Because CPTs have some limitations,6,7
non-camptothecin topoisomerase-I (topo-I) inhibitors such as
aromathecins 2a,b,g (Figure 1) are being pursued for therapeutic
development. This is exemplified by 22-hydroxyacuminatine
(2a) as a novel and rare quinoline alkaloid isolated along with
CPT from C. accuminata in very low yield (0.000006%).8 In
spite of its significant cytotoxicity against the murine leukaemia
P-388 and KB cell lines in vitro,9 only four total syntheses have
been achieved (Scheme 1). By analogy to the synthesis of CPT,10
product 2a was obtained by strategies based on intramolecular
Heck ring closure of I (7 steps),11 aza Diels-Alder of II (8
steps)12 and the coupling of 3-cyanophthalide derivatives and
pyrroloquinoline III (7 steps).13 In a parallel and recent
contribution, Greene et al.14 have illustrated a modular approach
to camptothecinoids from the Padwa hydroxypyridone15 that
involves successively a Claisen rearrangement, a Heck coupling,
and a classical Friedla¨nder condensation in an ultimate stage
(7 steps).
adam.daich@uniV-lehaVre.fr
ReceiVed NoVember 7, 2007
The pot, atom, and step-economic synthesis of Rosettacin
topo-I poison and its derivatives has been achieved using a
novel domino N-amidoacylation/aldol-type condensation,
followed by decarboxylation of the ester function. The key
domino procedure simply involves mixing HOBt ester as
new reagent with lactam and NaH together in THF or THF/
DMF. The reaction seems to be general and led to suitable
N-heterocyclic products in moderate to good yields.
Most importantly, rosettacins 2b,g belonging to the aromath-
ecin family are scarce and have been used16,17 as CPT/
(3) (a) Garcia-Carbonero, R.; Supko, J. G. Clin. Cancer Res. 2002, 8,
641-661. (b) Meng, L.-H.; Liao, Z.-H.; Pommier, Y. Curr. Top. Med. Chem.
2003, 3, 305-320.
(4) Pizzolato, J. F.; Saltz, L. B. Lancet 2003, 361, 2235-2242.
(5) Thomas, C. J.; Rahier, N. J.; Hecht, S. M. Bioorg. Med. Chem. 2004,
12, 1585-1604.
Camptothecin (CPT) and aromathecin alkaloids are a family
of natural products containing an indolizino[1,2-b]quinolin-
9(11H)-one nucleus fused to diverse hetero- and carbocyclic
rings at the C7- and C8-positions. Among them, 20(S)-camp-
tothecin (CPT, 1a, Figure 1), first isolated from the Chinese
tree Camptotheca accuminata in 1966 by Wall et al.,1 still serves
as a very attractive lead structure for the development of new
and potent anticancer drugs due to its antiproliferative activity.2
Two compounds in this class (Figure 1), topotecan (1b) and
irinotecan (1c), have been used clinically in the United States
for advanced ovarian and lung cancers and for colon carcinomas,
(6) Burke, T. G.; Mi, Z. M. J. Med. Chem. 1994, 37, 40-46.
(7) (a) Brangi, M.; Litman, T.; Ciotti, M.; Nishiyama, K.; Kohlhagen,
G.; Takimoto, C.; Robey, R.; Pommier, Y.; Fojo, T.; Bates, S. E. Cancer
Res. 1999, 59, 5938-5946.
(8) Marchand, C.; Antony, S.; Kohn, K. W.; Cushman, M.; Ioanoviciu,
A.; Staker, B. L.; Burgin, A. B.; Stewart, L.; Pommier, Y. Mol. Cancer
Ther. 2006, 5, 287-295 and references therein.
(9) Lin, L. Z.; Cordell, G. A. Phytochemistry 1989, 28, 1295-1297.
(10) Li, Q.-Y.; Zu, Y.-G.; Shi, R.-Z.; Yao, L.-P. Curr. Med. Chem. 2006,
13, 2021-2039.
(11) (a) Ma, Z.; Lee, D. Y. W. Tetrahedron Lett. 2004, 45, 6721-6723.
(b) Comins, D. L.; Baevsky, M. F.; Hong, H. J. Am. Chem. Soc. 1992,
114, 10971-10972. (c) Comins, D. L.; Nolan, J. M. Org. Lett. 2001, 3,
4255-4257.
(12) Zhou, H.-B.; Liu, G.-S.; Yao, Z.-J. J. Org. Chem. 2007, 72, 6270-
6272.
† Dedicated to our colleague, Prof. B. Decroix, upon his retirement.
‡ Universite´ du Havre.
(13) Xiao, X.; Antony, S.; Pommier, Y.; Cushman, M. J. Med. Chem.
2006, 49, 1408-1412.
§ University of Rouen.
(1) Wall, M. E.; Wani, M. C.; Cook, C. E.; Palmer, K. H.; McPhail, A.
T.; G. A. Sim, J. Am. Chem. Soc. 1966, 88, 3888-3890.
(2) (a) Wall, M. E. In Chronicles of Drug DiscoVery; Lednicer, D., Ed.;
American Chemical Society: Washington D.C., 1993; Vol. 3, p 327. (b)
Neoptolemos, J. P.; Cunningham, D.; Friess, H.; Bassi, C.; Stocken, D. D.;
Tait, D. M.; Dunn, J. A.; Dervenis, C.; Lacaine, F.; Hickey, H.; Raraty, M.
G. T.; Ghaneh, P.; Bu¨chler, M. W. Ann. Oncol. 2003, 14, 675-692.
(14) Babjak, M.; Kanazawa, A.; Anderson, R. J.; Greene, A. E. Org.
Biomol. Chem. 2006, 4, 407-409.
(15) Padwa, A.; Sheehan, S. M.; Straub, C. S. J. Org. Chem. 1999, 64,
8648-8659.
(16) Fox, B. M.; Xiao, X.; Antony, S.; Kohlhagen, G.; Pommier, Y.;
Staker, B. L.; Stewart, L.; Cushman, M. J. Med. Chem. 2003, 46, 3275-
3282.
10.1021/jo702387q CCC: $40.75 © 2008 American Chemical Society
Published on Web 02/07/2008
J. Org. Chem. 2008, 73, 1975-1978
1975

SCHEME 1. Retrosynthesis of Aromathecin Derivatives 2
SCHEME 2. Reactivity of 4 and ORTEP Plot of Derivative
7a
a Reagents and conditions: (i) MeOH, NiCl2‚H2O 10 mol %, reflux, 6
h, 95%; (ii) (COCl)2, DCM, 0 °C up to rt, 2 h, quant; (iii) pyrrolidin-2-one
(9a), 2.5 equiv of NaH, THF, reflux, 12 h, 70% (8/7 ) 13:1); (iv) pyrrolidin-
2-one (9a), toluene, reflux, 12 h, 68% (8/7 ) 8.5:1).
luotonin-A hybrids for binding to the topo-I/DNA covalent
binary complex. Both products 2b,g were obtained by improving
the strategy in path A (Scheme 1) starting from adequately
substituted phthalide 3.16
We began the development of this domino by defining its
basic requirements with respect to the electrophiles 4 and the
reaction conditions. Initially, coupling of 9a with acid chloride
4b was attempted using NaH (2.5 equiv) in refluxing THF
(conditions iii, Scheme 2). This method was not effective since
it provides an unexpected mixture of products 7/8 in a 1/13
ratio (70%), which differ only minimally (1/8.5 ratio) when a
toluene solution of 4b was simply refluxed without additional
additive (conditions iv, Scheme 2, 68% yield). The formation
of 8 could be attributed to the dimerization of 4a22 demonstrating
the instability of substrate 4b. Importantly, product 7 has an
original structure confirmed further by X-ray analysis.23 This
suggests that the reaction proceeds through an amidoacylation
of 4b at the acid chloride site either with lactam 9a or with its
salt followed ultimately by an intramolecular olefination via the
stable enol or enolate species of the acetate fragment. The
reaction seems to take place in a one-pot procedure, and
during the reaction mechanism the integrity of 4b was main-
tained.
In connection with our ongoing project aimed at the synthesis
of aromathecins, we became interested in the reactivity profile
of hitherto unknown 4b,c derived from the ester-acid 6
(Scheme 2).16 The differentiation we put between both trivalent
functions in 4a-c is strategically used, so that the aromatic ester
can be transformed selectively into imides by different lactams
under basic conditions (path B, Scheme 1). It was felt that the
formation of the D-ring of these platforms might be ac-
complished by nucleophilic attack of the “enolate” species at
one imide carbonyl. This attractive option was based on reports
showing that 2-succinimidylbenzylphosphonium ylides under-
went smooth Wittig olefination intramolecularly to give the
mitosane18 and quinocarcin19 ring systems, respectively. Previ-
ous investigations20,21 in isoquinolinone series also validated the
practicality of this approach. In these olefination reactions, an
imide serves as the electrophile and benzyltrimethylsilane20 or
benzylphosphonium bromide21 as the nucleophile under the
influence of different bases.
In this paper, we report the development of a new and rapid
protocol consisting of the domino N-amidoacylation/aldol-type
condensation on the basis of the unique reactivity of the
unknown HOBt ester reagent 4c (Scheme 2). To the best of
our knowledge, no examples from this domino have been
reported. Encouraged by recent advances in cascade processes,
we decided to explore such sequence notably to elaborate highly
functionalized modular scaffolds including the rosettacin topo-I
poison, isorosettacin, and various heterocyclic analogues with
promising biological profiles.
From these results, we speculated that diester 4c would
provide the intramolecular olefination of the intermediate A for
a newly designed domino N-amidoacylation/aldol-type conden-
sation (Scheme 2). So, reaction of ester-acid 6 with HOBt and
DCC in THF provided the ester 4c (∼100%).24 Our latter
proposal was then tested by reacting HOBt ester 4c with lactam
9a (Table 1) as a model lactam in the presence of a base under
different conditions. For instance, after complete consumption
of 4c, to our satisfaction the product 10a was isolated as the
sole reaction product in 82% yield when 2.5 equiv of NaH was
used (entry 6, Table 1). The effectiveness and the yield of this
process varied greatly depending on both the nature and the
quantity of the base used (Table 1).
(17) Cheng, K.; Rahier, N. J.; Eisenhauer, B. M.; Gao, R.; Thomas, S.
J.; Hecht, S. M.; J. Am. Chem. Soc. 2005, 127, 838-839.
(22) For synthesis of acid chloride 4a, see: Padwa, A.; Beal, L. S.; Eidell,
C. K.; Worsencroft, K. J. J. Org. Chem. 2001, 66, 2414-2421.
(23) See supporting information part or copies of the data of the product
7 (CCDC no. 661267) at http://www.ccdc.cam.ac.uk.
(24) HOBt esters are usually formed in situ and not isolated due to their
instability except in rare cases. For one example, see: Bartley, D. M.;
Coward, J. K. J. Org. Chem. 2005, 70, 6757-6774.
(18) Flitsch, W.; Langer, W. Liebigs Ann. Chem. 1988, 391-395.
(19) Garner, P.; Ho, W. B.; Shin, H. J. Am. Chem. Soc. 1993, 115,
10742-10753 and references therein.
(20) Cuevas, J.-C.; Snieckus, V. Tetrahedron Lett. 1989, 30, 5837-5840.
(21) Couture, A.; Cornet, H.; Deniau, E.; Grandclaudon, P.; Lebrun, S.
J. Chem. Soc., Perkin Trans. 1 1997, 469-476 and references therein.
1976 J. Org. Chem., Vol. 73, No. 5, 2008

TABLE 1. Optimization of the Domino Process
TABLE 2. Examining the Scope of the Domino Processa
entry base (equiv)a solvent T (°C)/time (h) product yieldb (%)
1
2
3
4
5
6
NEt3 (2.5)
tBuOK (2.5)
K2CO3 (2.5)
NaH (1)
NaH (2.5)
NaH (2.5)
CH3CN
THF
CH3CN
THF
THF
THF
reflux/12
reflux/12
reflux/12
rt/12
rt/12
rt/12
no reaction
31 + 60 (SM)c
60 + 23 (SM)c
no reaction
83
83
a The pyrrolidin-2-one (9a) was purified by distillation before its use.
b Isolated yield after silica gel chromatography. c Starting material.
Having established the effectiveness of the domino N-
amidoacylation/aldol-type condensation for construction of a
pyrroloisoquinolinone template, we then explored the scope of
the reaction. The results are tabulated in Table 2. For these
reactions, the optimized conditions were used (entry 6, Table
1). It can be seen that changing the size of the lactam (9b, entry
2) and its substitution (9e, entry 5) gave suitable products 10b,e.
However the yields decreased, respectively, to 50% (10b) and
30% (10e) depending in the latter case probably on the low
solubility of the dihydro-â-carboline salt intermediate. Reaction
efficiency with fused lactams as isoindolones (entries 3 and 4)
was proved, and the yields remained the same for 10c (72%)
and lower for 10d (37%), but in the last case, the reaction
heating was prolonged for 12 h. Importantly, a complex substrate
9f,25 bearing other functionalities (N,S-acetal and urea), was also
a competent substrate in this process (entry 6, product 10f, 47%
yield). The reactivity of 9g26 was effective with acceptable 50%
yield under the same conditions. The operability of this domino
process was evidenced again when 9h-j26 (entries 8-10), as
the regioisomers of 9g were submitted to our established
protocol. This happened in the presence of DMF as cosolvent
in a 1:1 ratio at reflux. Under these conditions, the expected
products 10h-j were isolated in yields of 47%, 44%, and 34%,
respectively. As expected, a substituent at the C4-position of
the quinoline nucleus proved to have a dramatic effect on the
reaction yields, which may be attributed to the hindrance of
the methyl and phenyl groups due to their proximity to the ester
group (see products 10i,j). It is noteworthy that under standard
conditions, 10h was isolated in low yield (23%) and no reaction
occurred starting from 9i but only traces of 10j were detected.
The result from entry 8 also confirmed our hypothesis regarding
the poor solubility during the process of the anionic species
formed as intermediates.
a 9a,b were commercial, and 9c-j were obtained according to known
protocols.25,26 b Isolated yield after chromatography. c In THF at reflux for
12 h. d In THF/DMF (1/1) at reflux for 12 h. e In refluxing THF without
other cosolvent. f Only traces of 10j were detected.
one-pot procedure. After certain optimization work (DS, PTSA
10 mol %, toluene, reflux, 12 h), the latter 12 underwent smooth
coupling condensation with 1 equiv of 2-aminobenzaldehyde
(13g) according to the Friedla¨nder reaction.28 The isolated
product 10g (51%) is identical to that obtained by the direct
synthesis via the domino process (entry 7, Table 2). Furthermore,
this reaction appeared to be quite general to both aromatic keto-
amines 13k,l cited above and provided suitable pentacyclic
products 10k,l in 71% and 86% yield, respectively. It is
noteworthy that these products are not reported by the direct
method due to the long and laborious access to the starting
materials such as 1,2-dihydropyrrolo[3,4-b]quinolin-3-one sub-
stituted with methyl and phenyl groups at the C9-position.
Interestingly, when the reaction was conducted in a one-pot
procedure starting from 10a, evaporation of the solvents after
4 h of the reaction and their replacement with toluene and PTSA/
aromatic amine 13g,k,l gave the expected products 10g, 10k,
Structure confirmation of these polyheterocyclic systems was
performed chemically by an approach using a Friedla¨nder
reaction as a key step starting from the lactam 10a (Scheme 3).
So, subjected to a Bredereck’s reagent (11)27 followed by
treatment with NaIO4 in THF/H2O (1/1), 10a provided the keto
derivative 12 (60%) that was obtained through successive
benzylic oxidation and cleavage of the Csp2-Csp2 linkage in a
(25) Hamid, A.; Elomri, A.; Da¨ıch, A. Tetrahedron Lett. 2006, 47, 1777-
1781 and references therein.
(26) For 9g,h, see: (a) Wang, H.; Ganesan, A. Tetrahedron Lett. 1998,
39, 9097-9098. (b) Tanaka, T.; Mashimo, K.; Wagatsuma, M. Tetrahedron
Lett. 1971, 12, 2803-2806.
(28) Jones, G. In ComprehensiVe Heterocyclic Chemistry II; Ramsden,
C. A., Ed.; Pergamon Press: Tarrytown, 1996; Vol. 5, Chapter 5.
(27) Wasserman, H. H.; Ives, J. L. J. Org. Chem. 1985, 50, 3573-3580.
J. Org. Chem, Vol. 73, No. 5, 2008 1977

SCHEME 3. Rosettacins 2g-l and Isorosettacins 2h-ja
of quinolinones fused to numerous N-heterocycles were prepared
in this way, which was illustrated by a rapid, two-step total
synthesis of the rosettacin topo I poison and derivatives. Further
research from our group will focus on the use of this strategy
to build efficiently a variety of compounds including alkaloids
with promising therapeutic profiles.
Experimental Section
Typical Procedure for the Synthesis of Rosettacin (2g).
To a 0.1 M THF solution of HOBt ester (4c, 311 mg, 1 mmol)
and lactam (9g, 184 mg, 1 equiv) was added in one portion 2.5
equiv of NaH. After 12 h at reflux, the reaction was hydrolyzed
with saturated NH4Cl solution and worked up by extracting
either with Et2O or with DCM. Pure 10g (280 mg) was isolated
by chromatography on a silica gel column (AcOEt/cyclohexane
as eluent) in 82% yield. Compound 10g (342 mg, 1 mmol) in
7 mL of 48% HBr was then heated under stirring at 135 °C for
12 h. After cooling, the solution was alkalinized and extracted
twice with DCM (10 mL). A classical workup provided pure
rosettacin (2g) as a light-yellow solid in 61% yield (171 mg):
Rf 0.19 (cyclohexane/AcOEt 1:1); mp 288 °C (lit.30 mp 289-
291 °C; lit.31 mp 286 °C); IR (KBr) ν 1663, 1628, 1602, 1504,
1479, 1398, 1384, 1239, 1216, 1166 cm-1; 1H NMR (200 MHz,
CDCl3) δ 5.31 (s, 2H, CH2N), 7.49-7.65 (m, 3H, Haro), 7.68-
7.86 (m, 4H, Haro), 8.17 (d, 1H, Haro, J ) 7.8 Hz), 8.26 (s, 1H,
a Reagents and conditions: (i) Bredereck’s reagent (11, 1.4 equiv), 110
°C, 2 h; (ii) NaIO4 (3 equiv), THF/H2O (1/1), rt, 0.5 h, 60% in two steps;
(iii) Aromatic amine (13g-l, 1 equiv), 10% PTSA, toluene, reflux, 12 h,
51 up to 86%; (iv) 48% HBr, 135 °C, 12 h, 51 up to 86%.
and 10l in yields comprable to those reported for the two
separate steps.
H
aro), 8.49 (d, 1H, Haro, J ) 7.8 Hz); 13C NMR (50 MHz,
CDCl3) δ 49.7 (CH2), 101.6 (CHaro), 126.3 (Cq), 127.6 (2CHaro),
127.7 (CHaro), 127.8 (CHaro), 128.2 (Cq), 128.3 (CHaro), 129.0
(Cq), 129.6 (CHaro), 130.5 (CHaro), 131.1 (CHaro), 132.7 (CHaro),
137.7 (Cq), 140.1 (Cq), 148.9 (Cq), 153.7 (Cq), 161.2 (CdO).
Anal. Calcd for C19H12N2O (284.09): C, 80.27; H, 4.25; N,
9.85. Found: C, 80.11; H, 4.09; N, 9.77.
The concluding step for reaching the targets rosettacin (2g),
isorosettacin (2h), and other aromathecins 2k,l and 2i,j from
the corresponding esters 10g-l proved, however, to be unex-
pectedly challenging. After several unrewarding attempts, it was
discovered that the conditions described by Rigo and co-
workers29 (48% HBr, 135 °C, 12 h) led to the desired products
2g-l (Scheme 3) in yields ranging from 51% to 86%. The
application of the same protocol to the regioisomers esters
10h-j (entries 8-10, Table 2) also resulted in the formation
of the expected and novel isorosettacins 2h-j in yields around
65% in all cases.
In summary, we have developed a highly efficient, extremely
simple, and novel methodology involving a domino N-amidoa-
cylation/aldol-type condensation to provide poly-N-heterocycles.
This is based on the use of HOBt ester 4c as a new reagent and
quite effective source of the quinoline nuclei. A broad variety
Acknowledgment. We are grateful to the Region of “Haute
Normandie-76, France” for a Regional Graduate Fellowship
(2005-2007), attributed to Fre´de´ric Pin and to Dawn Hallidy
for technical assistance.
Supporting Information Available: Experimental procedures,
product characterization for all new compounds synthesized, and
an ORTEP plot of product 7 as well as its CIF file. This material
is available free of charge via the Internet at http://pubs.acs.org.
JO702387Q
(30) Warneke, J.; Winterfeldt, E. Chem. Ber. 1972, 105, 2120-2124.
(31) Walraven, H. G. M.; Pandit, U. K. Tetrahedron 1980, 36, 321-
327.
(29) Gavara, L.; Rigo, B.; Couturier, D.; Goossens, L.; He´nichart, J.-P.
Tetrahedron 2007, 63, 9456-9464 and references therein.
1978 J. Org. Chem., Vol. 73, No. 5, 2008