An efficient total synthesis of 9-methoxycarbazole-3-carbaldehyde based on a novel methodology for the preparation of methoxyindoles

Article abstract of DOI:10.1016/S0040-4039(03)01745-3

A direct synthesis of naturally occurring 9-methoxycarbazole-3-carbaldehyde 1, based on our methodology for the synthesis of 1-methoxyindoles, is reported. A novel benzannulation strategy was employed using ring closing metathesis as the key step in this total synthesis. The synthesis of the natural product 1 has been achieved in seven steps in 14% overall yield from commercial materials and in only four steps from a methoxyindole compound obtained using the new methodology.

Full text of DOI:10.1016/S0040-4039(03)01745-3

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 7071–7074
An efficient total synthesis of 9-methoxycarbazole-3-carbaldehyde
based on a novel methodology for the preparation
of methoxyindolesꢀ
N. Selvakumar,* Manoj Kumar Khera, B. Yadi Reddy, D. Srinivas, A. Malar Azhagan and
Javed Iqbal
Department of Discovery Chemistry, Discovery Research, Dr. Reddy’s Laboratories Ltd., Miyapur, Hyderabad 500 050, India
Received 12 April 2003; revised 2 July 2003; accepted 17 July 2003
Abstract—A direct synthesis of naturally occurring 9-methoxycarbazole-3-carbaldehyde 1, based on our methodology for the
synthesis of 1-methoxyindoles, is reported. A novel benzannulation strategy was employed using ring closing metathesis as the key
step in this total synthesis. The synthesis of the natural product 1 has been achieved in seven steps in 14% overall yield from
commercial materials and in only four steps from a methoxyindole compound obtained using the new methodology.
© 2003 Elsevier Ltd. All rights reserved.
A number of alkaloids possessing a 1-methoxyindole
structural framework have been isolated and reported
in the literature.1 These include 9-methoxycarbazole-3-
carbaldehyde 1 isolated from Murraya euchrestifolia
HAYATA (Rutaceae) and Paniculidine B 2 isolated
from Murraya paniculata (Linn.) Jack.2,3 Various parts
of the latter plant are used as a folk medicine for the
treatment of stomachache and toothache and also as a
stimulant throughout areas ranging from India, South-
east Asia, southern China, Taiwan etc. In the preceding
letter, we disclosed a new synthesis of substituted 1-
methoxyindoles based on a novel methodology (Scheme
1).4 The methodology involves a rearrangement of a
nitro compound of type 3 under neutral conditions to a
1-methoxyindole derivative such as 4. In an effort to
substantiate the usefulness of the methodology, we
initiated a program towards the total synthesis of the
title natural product. To this end, we now report a
highly efficient total synthesis of 9-methoxycarbazole-3-
carbaldehyde 1 in this communication.5
The disconnection of the natural product 1, illustrated
in Scheme 2, led to the indole aldehydes 7 or 8 based on
two conceptually different retrosynthetic approaches
Scheme 1.
Keywords: rearrangement; metathesis; alkaloid; decarboxylation; nucleophilic catalysis.
ꢀ DRL Publication No. 311
* Corresponding author. Tel.: 91-40-23045439; fax: 91-40-23045438; e-mail: selvakumarn@drreddys.com
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)01745-3

7072
N. Sel6akumar et al. / Tetrahedron Letters 44 (2003) 7071–7074
Scheme 3. Reagents and conditions: (a) Methyl cyanoacetate,
NaH, THF, 0“60°C, 75%; (b) NaH, DMF, allyl bromide, rt,
68%; (c) NaCl, DMSO, 155°C, 72%.
Scheme 2. Retrosynthetic plan for the natural product 1.
and the aldehydes in turn would be available from our
methodology reported in the preceding letter.4 The first
route involved an electrocyclic ring closure of an appro-
priate unsaturated ester 5 that is accessible from the
aldehyde 7 by a Wittig–Horner reaction. The second
protocol required a conjugated olefin of type 6 or its
equivalent for a ring closing metathesis (RCM) reaction,
compound 6 could be easily prepared from the aldehyde
8 by a three carbon homologation procedure. Having
conceived two attractive routes for the title compound
1, we turned our attention towards the preparation of the
aldehydes 7 and 8.
We envisioned that the requisite aldehydes 7 and 8 would
be available from nitriles such as 11 by standard reduc-
tion procedures (Scheme 3). However, our methodology
renders access to only 2-substituted 1-methoxyindole-3-
carboxylates but not to 3-substituted nitriles. Thus, we
used methyl cyanoacetate instead of dimethyl malonate
in the aromatic nucleophilic substitution on 2-fluoro-
nitrobenzene anticipating that the carbomethoxy group
would take part in the rearrangement leaving the C-3
nitrile of the resultant indole. Thus, the cyanoester 10,
prepared by the usual reaction sequence,6 underwent the
rearrangement to the required 1-methoxyindole 11 pos-
sessing the nitrile substituent at C-3.
Scheme 4. Reagents and conditions: (a) DIBAL-H, dry PhMe,
−78°C, 73%; (b) trimethyl phosphonoacetate, NaH, THF, 62%;
(c) dry PhMe, reflux.
As the electrocyclic ring closure method failed to deliver
the requisite 9-methoxy carbazole skeleton, we turned
our attention towards the second route involving RCM
as the key step. Because of the ready availability of the
aldehyde 7 from Scheme 4, we studied the feasibility of
RCM methodology in synthesizing the 9-methoxycarba-
zole skeleton on this aldehyde as a model system (Scheme
5). Attempted three-carbon homologation of the alde-
hyde 7 using allyl magnesium bromide failed to yield the
desired results. However, the transmetallation of allyl-
tributyltin followed by treatment with aldehyde 7 cleanly
afforded the alcohol 15 as a stable oil. It was expected
at this stage that the product obtained by the RCM
reaction of alcohol 15 would eventually undergo dehy-
dration leading to the corresponding aromatised com-
pound. It was indeed true that treatment of the alcohol
15 with Grubbs’ catalyst resulted in smooth cyclisation
with concomitant dehydration leading to 9-methoxycar-
bazole 16 in very good yields.8
With the nitrile 11 in hand, the synthesis of the natural
compound 1 was initially attempted by the first route
involving electrocyclic ring closure as the key step
(Scheme 4). Thus, subjecting aldehyde 7, obtained by the
reduction of nitrile 11, to Wittig–Horner reaction condi-
tions resulted in the substrate 5 setting the stage for the
crucial electrocyclic ring closure reaction. However, all
our attempts to effect the ring closure resulted only in
the demethoxylated compound 12 along with an appre-
ciable quantity of the corresponding aromatised com-
pound 13. A mechanism of demethoxylation is illustrated
in structure 5 and is believed to go through the interme-
diacy of structure 14. It is important to note that
precedents exist for such demethoxylations in the litera-
ture in the area of methoxyindoles.7

N. Sel6akumar et al. / Tetrahedron Letters 44 (2003) 7071–7074
7073
Scheme 5. Reagents and conditions: (a) Allyltributyltin, BuLi, THF, −78°C, 74%; (b) Grubbs’ catalyst, CH2Cl2, rt“reflux, 66%.
Having established the feasibility of the RCM reaction
in effecting the ring closure on the model system, we
turned our attention in applying this methodology to
the synthesis of the natural compound 1 (Scheme 6).
Applying our methodology as in Scheme 3 on commer-
cially available 3-fluoro-4-nitrotoluene resulted in the
methoxyindole 19 in much improved yields. Treatment
of the nitrile 19 with DIBAL-H cleanly afforded the
aldehyde 8, which was allylated as described in Scheme
5 resulting in the alcohol 20. The alcohol 20 underwent
a smooth RCM reaction as before leading to the carba-
zole 21 in excellent yield.9 Among the many methods
attempted to convert the aromatic methyl group in 21
to the aldehyde as in 1, DDQ in acetic acid was
particularly impressive affording the target compound
in moderate yields.10 The spectral data of the synthetic
compound 1 were in agreement in all aspects with those
of the natural material.2,11 The synthesis of the natural
product 1 has thus been achieved in seven steps in 14%
overall yield from commercial materials and in only
four steps from a methoxyindole compound obtained
using our new methodology.
In conclusion, a direct total synthesis of 9-methoxycar-
bazole-3-carbaldehyde 1, based on our methodology
reported in the preceding letter,4 has been achieved
using RCM as the key step. The overall process of
converting the aldehyde 8 into carbazole 21 constitutes
a novel benzannulation strategy. The synthesis of a
methoxyindole compound having no substituent in the
2-position and converting such material to the other
natural products having a 1-methoxyindole skeleton,
particularly paniculidine B 2, is currently underway,
and will be the content of future publications of this
laboratory.
Acknowledgements
We thank Dr. K. Anji Reddy for his continued encour-
agement in this work. We would like to record our
special thanks to Dr. Sanjay Trehan for some useful
discussions. The help extended by Dr. R. Rajagopalan
is greatly acknowledged. We appreciate the services
extended by the Analytical Research Department of
Discovery Research, Dr. Reddy’s Laboratories Ltd. for
this communication.
References
1. (a) Somei, M. Heterocycles 1999, 50, 1157; (b) Acheson,
R. M. In Advances in Heterocyclic Chemistry; Katritzky,
A. R., Ed.; Academic Press: San Diego, 1990; Vol. 51,
pp. 105–175.
2. Ito, C.; Wu, T.-S.; Furukawa, H. Chem. Pharm. Bull.
1988, 36, 2377.
3. Kinoshita, T.; Tatara, S.; Sankawa, U. Chem. Pharm.
Bull. 1985, 33, 1770.
4. Selvakumar, N.; Reddy, B. Y.; Azhagan, M. A.; Khera,
M. K.; Babu, J. M.; Iqbal, J. Tetrahedron Lett. 2003, 44,
7065.
5. For a synthesis of this natural product, see: Kawasaki,
T.; Somei, M. Heterocycles 1990, 31, 1605.
6. Selvakumar, N.; Yadi Reddy, B.; Sunil Kumar, G.; Iqbal,
J. Tetrahedron Lett. 2001, 42, 8395.
Scheme 6. Reagents and conditions: (a) Methyl cyanoacetate,
NaH, THF, 0“60°C, 94%; (b) NaH, DMF, allyl bromide,
0°C“rt, 77%; (c) NaCl, DMSO, 155°C, 85%; (d) DIBAL-H,
dry PhMe, −78°C, 85%; (e) allyltributyltin, BuLi, THF,
−78°C, 83%; (f) Grubbs’ catalyst, CH2Cl2, rt“reflux, 72%; (g)
DDQ, AcOH, rt, 45%.

7074
N. Sel6akumar et al. / Tetrahedron Letters 44 (2003) 7071–7074
7. For a similar demethoxylation, see: Yamada, F.; Fukui,
8.00 (d, J=7.3 Hz, 1H), 7.83 (s, 1H), 7.60–7.15 (m, 5H),
4.10 (s, 3H), 2.53 (s, 3H); lC (50 MHz, CDCl3): 21.4,
63.2, 108.2, 108.4, 119.9, 120.3, 120.4, 125.9, 127.4, 129.5,
136.2, 138.2 (signals for two quaternary carbons possibly
overlap with other peaks); Mass (EI): 211, 196, 180;
HRMS (EI) calcd for C14H13NO, 211.0997; found,
211.0997.
Y.; Shinmyo, D.; Somei, M. Heterocycles 1993, 35, 99.
8. Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem.
Soc. 1996, 118, 100.
9. Compound 21: A suspension of Grubbs’ catalyst (306
mg, 0.37 mmol) in dry CH2Cl2 (50 mL) was stirred for 15
min at rt. To the above suspension was added a solution
of the alcohol 20 (910 mg, 3.7 mmol) in dry CH2Cl2 (20
mL) dropwise over a period of 10 min under argon. After
stirring for a further 30 min at the same temperature, the
reaction mixture was refluxed for 16 h and was then
allowed to cool to rt. A few drops of water were added to
the resultant mixture and the contents were filtered over
sodium sulfate. The residue obtained upon concentration
of the filtrate was passed through a column of silica gel to
afford the compound 21 as a colorless oil (567 mg, 72%).
IR (neat, w cm−1): 1452, 1232 743; lH (200 MHz, CDCl3):
10. Lee, H.; Harvey, R. G. J. Org. Chem. 1988, 53, 4253.
11. Spectral data of synthetic compound 1: IR (neat, w cm−1):
1687, 1598, 1234, 745; lH (400 MHz, CDCl3): 10.11 (s,
1H), 8.61 (d, J=1 Hz, 1H), 8.15 (dd, J=7.8 and 1 Hz,
1H), 8.15 (dd, J=8.3 and 1.5 Hz, 1H), 7.64–7.56 (m, 3H),
7.40–7.35 (m, 1H), 4.23 (s, 3H); lC (50 MHz, CDCl3):
191.5, 140.0, 137.3, 129.3, 127.6, 127.1, 123.8, 121.0,
120.8, 119.5, 108.3, 107.9, 64.1; Mass (EI): 225, 210, 194,
164; HRMS (EI) calcd for C14H11NO2, 225.0790; found,
225.0788.