Mechanism of an unusual decarboxylative cyclization

Article abstract of DOI:10.1016/j.tetlet.2004.05.161

The mechanism of an unusual decarboxylative cyclization from 5-methoxy-1-(2-carboxyphenyl)-1,4-dihydro-4-oxopyridine-2-carboxylic acid (diacid) to 3-methoxypyrido[1,2-a]indole-2,10-dione (ketone) has been investigated. ¹³C-labeling has demonstrated that the carbonyl carbon of the ketone arises exclusively from the anthranilic acid carboxyl of the diacid. A zwitterionic mechanism has been proposed.

Full text of DOI:10.1016/j.tetlet.2004.05.161

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 7809–7812
Mechanism of an unusual decarboxylative cyclization
Andrew S. Kende,* Olivier Henry and Zecheng Chen
Department of Chemistry, University of Rochester, Rochester, NY 14627, USA
Received 8 April 2004; accepted 8 May 2004
Abstract—The mechanism of an unusual decarboxylative cyclization from 5-methoxy-1-(2-carboxyphenyl)-1,4-dihydro-4-oxopyr-
idine-2-carboxylic acid (diacid) to 3-methoxypyrido[1,2-a]indole-2,10-dione (ketone) has been investigated. 13C-labeling has demon-
strated that the carbonyl carbon of the ketone arises exclusively from the anthranilic acid carboxyl of the diacid. A zwitterionic
mechanism has been proposed.
Ó 2004 Elsevier Ltd. All rights reserved.
Pterocellins A (1) and B (2) are antineoplastic alkaloids
(Fig. 1) recently isolated from a New Zealand marine
Bryozoan.1 NMR studies on these substances, sup-
ported by a single crystal X-ray structure determination
on the deep red Pterocellin A, reveal that these com-
pounds possess the structurally novel pyrido[4,3-b]indol-
izine-5,7-dione core, a deceptively simple yet challenging
synthetic target.
O
N
O
N
MeO
MeO
Ac O, reflux
2
COOH
COOH
O
88%
4
3
Scheme 1.
In surveying potential routes to the Pterocellins we dis-
covered that the benzenoid counterpart of their core,
namely 3-methoxypyrido[1,2-a]indole-2,10-dione (4)
had been described in 1997 by Korenova et al.2 In that
report, the model core 4 had been produced in high
yield as a dark red solid, mp 204–206°C, by heating 5-
methoxy-1-(2-carboxyphenyl)-1,4-dihydro-4-oxopyridine-
2-carboxylic acid (3) in refluxing acetic anhydride
(Scheme 1).
Since the relatively mild decarboxylative cyclization of
diacid 3 to ketone 4 was unusual, we undertook to con-
firm KorenovaÕs observations. In fact, when diacid 3 is
held at 100°C in acetic anhydride for several hours
(optimum conditions), a dark red crystalline ketone
melting at 251–252°C was obtained in nearly quantita-
tive yield. Despite the melting point discrepancy, the
1H NM R,13C NMR, IR, MS, and HRMS of our prod-
uct were in agreement with the data of Korenova, and in
full accord with the proposed tricyclic structure 4.3 With
the Korenova cyclization thus confirmed, we considered
that this unusual ring closure may represent an intramo-
lecular acylation of the pendant anthranilic acid ring by
an activated pyridone-2-carboxyl group (Scheme 2).
O
N
O
N
MeO
MeO
O
O
Ph
N
N
Two new experimental observations cast doubt on the
above mechanism. When 5-methoxy-1-phenyl-1,4-dihy-
dro-4-oxopyridine-2-carboxylic acid (7) was heated in
acetic anhydride or in dimethyl sulfoxide, no cyclization
was observed, and only the decarboxylation product 8
was isolated. In contrast, pyridone ester 9 in DMSO at
120°C was smoothly converted to the tricyclic ketone
4 in 91% yield (Scheme 3).
2 : PTEROCELLIN B
1 : PTEROCELLIN A
Figure 1.
Keywords: Mechanism; Decarboxylative cyclization.
*
Corresponding author. Tel.: +1 585 275 4236; fax: +1 585 276
0205; e-mail: kende@chem.rochester.edu
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.05.161

7810
A. S. Kende et al. / Tetrahedron Letters 45 (2004) 7809–7812
O
N
O
N
O
N
MeO
MeO
MeO
O
C
13
O
COOH
NH
O
OH
CO H
2
Ac O
AcOH-H O (1:1)
2
-AcO
2
O
O
MeO
+
COOH
COOH
C
2
C
13
o
CO H
CH
110
COOH
C
2
3
O
10
11
12
3
5
O
O
O
N
O
N
MeO
MeO
MeO
MeO
o
Ac O, 100
C
2
- CO , -H
2
N
N
COOH
O
O
13
O
C
O
13
COOH
H
O
6
13
4
12
Scheme 2.
Scheme 4.
O
N
O
N
the red crystalline ketone product 13 was identified as
having retained all of the original 13C-label at the new
carbonyl group. Thus, ketone 13 showed in the 13C
NMR a huge peak at 183.8ppm for the 13C-carbonyl
carbon, and in the mass spectrum the [MH]+ ion was
at m/z = 229, with the m/z = 228 peak for the unlabeled
MeO
MeO
o
O
OH
Ac O, 100 C or
2
H
C
o
DMSO, 120
C
1
7
O
compound 4 absent. Moreover, in the H NMR, long-
8
range coupling of the 13C-carbonyl carbon to both adja-
cent ortho protons could be clearly seen, with
J = 2.8Hz.5 The above data establish that the ketone
carbonyl in the cyclization of 3 arises from the anthrani-
lic carboxyl, and not from the pyridone carboxyl.
O
N
MeO
MeO
o
DMSO, 120
91%
C
N
CO H
2
O
CO Me
2
The preceding results led to a new mechanism (Scheme
5) in which the C(2) pyridone carbon loses H+ and
CO2 to become an anionic nucleophilic center, which
subsequently attacks the anthranilic carboxyl. There is
some literature analogy for direct C(2) anion formation
from N-alkyl-4-oxo-pyridines using n-BuLi at À78°C,6
and from deuterium exchange at 100°C.7
9
4
Scheme 3.
In order to firmly establish the origin of the ketone car-
bonyl in product 4, a 13C-labeled precursor 12 was syn-
thesized by condensation of 13C-anthranilic acid 10
(prepared in three steps from 2-bromonitrobenzene
and Cu13CN)4 with the methoxypyranone acid 11 as
shown in Scheme 4. When this 13C-labeled 12 was
heated in acetic anhydride at 100°C for several hours,
We obtained evidence consistent with C(2) nucleophilic-
ity by heating pyridone acid 7 in DMSO at 120°C in the
presence of weak electrophiles, leading to 14,8 15,9 and
16,10 as summarized in Scheme 6.
O
N
O
O
N
MeO
MeO
O
MeO
O
H
O
Ac O
2
COOH
N
COOH
COOH
O
O
O
O
CH
3
O
CH
3
3
O
N
O
O
N
MeO
MeO
MeO
- H
- CO
O
O
C
O
N
O
O
CH
3
2
O
CH
3
O
4
Scheme 5.

A. S. Kende et al. / Tetrahedron Letters 45 (2004) 7809–7812
7811
decarboxylation occurred upon heating the salt 17 in
DMSO at 120°C for many hours. Moreover, the sodium
salt 18 was similarly stable under these conditions
(Scheme 7).
O
N
O
N
MeO
MeO
I
2
o
I
CO H
DMSO, 120
C
2
37 %
These last results led us to a third mechanism (Scheme 8)
which requires the participation of the carboxyl proton
to drive a zwitterionic reaction sequence.
8
Cl
14
OHC
D O
2
o
DMSO
o
DMSO, 120 C
120
C
This last mechanism appears to fit all of our data and is
entirely consistent with the 1965 observation of Beak
and Bonham7 on the rapid decarboxylation of 4-meth-
oxypyridinium-2-carboxylates, as well as with the early
work of Hammick and co-workers11 on the decarboxyl-
ation of a-picolinic acid. To our knowledge this is the
first example of the use of such decarboxylations to form
fused pyridone ring systems.
47 %
O
O
MeO
MeO
Cl
N
D
N
OH
15
16
Scheme 6.
References and notes
1. Yao, B.; Prinsep, M. K.; Gordon, D. P. J. Nat. Prod.
2003, 66, 1074–1077.
2. Korenova, A.; Povazanec, F.; Uher, M.; Hajas, M. Chem.
Papers 1997, 51, 383–389.
O
O
MeO
MeO
o
DMSO, 120
C
3. Compound 4: Mp: 251–252°C. MS (API-ES, positive):
1
m/z 228.2 [MH]+. H NMR (CDCl3, 400MHz): d 7.84 (d,
N
CO Na
2
N
CO Na
2
no reaction!
J = 7.5Hz, 1H), 7.72 (t, J = 7.8Hz, 1H), 7.46 (s, 1H), 7.35
(d, J = 8.0Hz, 1H), 7.33 (t, J = 7.6Hz, 1H), 7.03 (s, 1H),
3.96 (s, 3H). 13C NMR (CDCl3, 100MHz): d 183.8, 174.1,
169.1, 151.7, 147.4, 136.9, 126.0 (2C), 123.1, 113.3, 111.7,
110.0, 56.7. HRMS (DCI): calcd for C13H10NO3 [MH]+
m/z 228.0661, found 228.0660.
17
O
O
N
MeO
MeO
o
DMSO, 120
C
4. Chaudhuri, N. K.; Potdar, S. V.; Ball, T. J. J. Labelled
Compd. RadioPharm. 1982, 19, 75–82.
N
CO Na
CO Na
2
2
CO Me
CO Me
2
2
no reaction!
5. Compound 13: Mp: 251–252°C. MS (API-ES, positive):
m/z 229.1 [MH]+. 1H NMR (CDCl3, 400MHz): d 7.80 (dd,
J = 7.3, 2.8Hz, 1H), 7.70 (t, J = 7.8Hz, 1H), 7.50 (s, 1H),
7.37 (d, J = 8.0Hz, 1H), 7.31 (t, J = 7.8Hz, 1H), 6.99 (d,
J = 2.8Hz, 1H), 3.93 (s, 3H). HRMS (DCI): calcd for C12
13CH10NO3 [MH]+ m/z 229.0694, found 229.0699. The
observed 3JC–H of 2.8Hz is similar to the 4.08Hz coupling
constant found for analogous ortho 3JC–H coupling in 13C-
carboxyl labeled methyl benzoate by Hutchins, R. O.;
Maryanoff, B. E. J. Am. Chem. Soc. 1972, 94, 3268–3269.
18
Scheme 7.
Two subsequent experiments employing the sodium salt
of acid 7 and ester acid 9 required us to further revise the
second mechanism (Scheme 5). We found that no
PROPOSED ZWITTERIONIC MECHANISM
O
N
H
H
O
N
O
N
MeO
MeO
MeO
~ H
O
O
COOH
OMe
C
C
O
O
OMe
OMe
O
O
O
9
H
O
H
O
N
O
MeO
MeO
MeO
- CO
2
N
O
N
OMe
O
O
4
Scheme 8.

7812
A. S. Kende et al. / Tetrahedron Letters 45 (2004) 7809–7812
6. Meghani, P.; Joule, J. A. J. Chem. Soc., Perkin Trans. 1
1988, 1, 1–8.
(m, 5H), 6.59 (s, 1H), 3.89 (s, 3H). 13C NMR (CDCl3,
100MHz): d 172.5, 150.5, 143.5, 136.3, 130.2 (2C), 128.4,
122.8 (2C), 120.6, 116.0, 56.2. HRMS (DCI): calcd for
C12H112HNO2 [MH]+ m/z 203.0931, found 203.0928.
10. Compound 16: MS (API-ES, positive): m/z 342.3 [MH]+.
Mp: 136–138°C. 1H NMR (CD3OD, 400MHz): d 7.60–
7.50 (m, 3H), 7.46 (s, 1H), 7.35 (t, J = 6.6Hz, 1H), 7.23 (d,
J = 8.4Hz, 2H), 6.95 (s, 1H), 6.91–6.86 (m, 3H), 5.47 (s,
1H), 3.77 (s, 3H). 13C NMR (CDCl3, 100MHz): d 172.5,
152.2, 148.5, 140.9, 138.7, 133.6, 129.8, 129.7, 129.4, 128.5
(2C), 128.3 (2C), 127.5, 127.3, 123.7, 115.0, 70.3, 56.2.
HRMS (FAB): calcd for C19H17NO3Cl [MH]+ m/z
342.0899, found 342.0910.
7. Beak, P.; Bonham, J. J. Am. Chem. Soc. 1965, 87, 3365–
3371; For related chemistry, see: Katritzky, A. R.;
Murugan, R.; Sakizadeh, K. J. Heterocycl. Chem. 1984,
21, 1465–1467; Katritzky, A. R.; Cozens, A. J.; Ossana,
A.; Rubio, O.; Dabbas, N. J. Chem. Soc., Perkin Trans. 1
1985, 2167–2172.
8. Compound 14: Mp: 128–130°C. MS (API-ES, positive):
m/z 327.9 [MH]+. 1H NMR (CD3OD, 400MHz): d 7.78 (s,
1H), 7.62–7.60 (m, 3H), 7.45–7.42 (m, 2H), 7.15 (s, 1H),
3.78 (s, 3H). HRMS (DEI): calcd for C12H10NO2I [M +]
m/z 326.9756, found 326.9750.
9. Compound 15: MS (API-ES, positive): m/z 203.1 [MH]+.
1H NMR (CD3OD, 400MHz): d 7.82 (s, 1H), 7.65–7.53
11. Ashworth, M. R. F.; Daffern, R. P.; Hammick, D. L. J.
Chem. Soc. 1939, 809–812.