A combinatorial approach towards 2-acyl-3-amino-indole derivatives

Article abstract of DOI:10.1016/S0040-4039(00)01467-2

A widely applicable reaction sequence towards functionalised 2-acyl-3-amino-indole derivatives is presented. A set of 205 indole derivatives were prepared, in a solution phase combinatorial fashion, by coupling a reliable reaction protocol to a rapid puri

Full text of DOI:10.1016/S0040-4039(00)01467-2

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 8251–8254
A combinatorial approach towards 2-acyl-3-amino-indole
derivatives
Matthias Nettekoven*
F. Hoffmann-La Roche Ltd, Pharmaceutical Research, Chemical Technologies, Automated Chemistry,
CH-4070 Basel, Switzerland
Received 15 August 2000; accepted 29 August 2000
Abstract
A widely applicable reaction sequence towards functionalised 2-acyl-3-amino-indole derivatives is
presented. A set of 205 indole derivatives were prepared, in a solution phase combinatorial fashion, by
coupling a reliable reaction protocol to a rapid purification system employing parallel filtration and
application of automated reversed phase HPLC. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: combinatorial chemistry; solution phase synthesis; indole derivative library; HPLC purification.
1
Combinatorial chemistry is currently under intense investigation and the search for new
pharmacophore containing scaffolds is an ongoing process. Indole ring systems are ubiquitously
2
found within pharmaceutical agents as well as in natural products and therefore access to novel
functionalised indole derivatives in a combinatorial fashion would be of great interest.
3
A facile approach to 3-amino-indoles was first disclosed in 1991 by Viti and coworkers, who
described the synthesis of such compounds as intermediates of a multistep synthesis leading to
indolobenzoazepin-6-ones. Complementary to this route, our efforts were concerned with the
evaluation of the potential of 3-amino-indole derivatives in combinatorial library synthesis.
The synthetic approach is outlined in Scheme 1 and comprised the coupling of eight different
aminobenzonitriles 1 with a selected set of acid chlorides, which yielded a matrix of 90 amides
Scheme 1.
*
Corresponding author. Tel: +41-61-6886227; fax: +41-61-6886459; e-mail: matthias.nettekoven@roche.com
0
040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)01467-2

8252
2
. A total of 83 were successfully isolated in yields from 70 to 99% with purities in the range
of 66 to 99%. The aminobenzonitriles were variously substituted with alkyl or halogen groups
and the acid chlorides employed were either cycloalkyl, substituted benzoyl or heteroaromatic
acid chlorides. The acid chlorides all reacted smoothly whereas the nature of the aminoben-
zonitrile dramatically affected the yield and the purity of the amide products. The unsubstit-
uted and monosubstituted aminobenzonitrile reacted cleanly whereas polysubstituted
aminobenzonitriles generally gave poorer yields and purities of the desired amides 2.
From the 83 amides 2 isolated, the 48 with the highest purity were selected and used in the
4
consecutive step without further purification.
a-Bromoketones are known to react under various reaction conditions with, for example
5
6
7
8
sulfides, alcohols, amines and amides. Interestingly, the reaction of amide 2 with a-bro-
moketones 3 could not be affected in CH Cl or DMF with pyridine as a base or even in neat
2
2
pyridine. Use of other bases such as KOH, NaHCO , KOtBu, NEt or DIPEA did not change
3
3
the outcome. It has been previously described that the deprotonation can be achieved in DMF
3
1
2
3
using NaH as the base. In a model, the reaction of 2 (R =H and R =Ph) with 3 (R =Ph) in
1
2
3
DMF with 1 equiv. NaH as the base indole 5 (R =H, R =R =Ph) was obtained in 53%
yield. Employing NaH in this synthetic array was considered to be impractical due to the
moisture sensitive nature of the reagent. Therefore, a subsequent search for alternative condi-
tions revealed Cs CO as a base in DMF at room temperature to be a suitable protocol in this
reaction sequence, which has not been reported earlier. The use of DMF as a solvent was
advantageous since it allowed the preparation of stock solutions of the starting material for
2
3
9
1
2
3
both poorly soluble amides as well as a-bromoketones. The yield of 5 (R =H, R =R =Ph)
employing the Cs CO protocol was comparable (51%) to that of using NaH as the base,
10
2
3
which indicated that this new procedure could be readily applied in parallel array synthesis.
A library was prepared from 280 reactions, consisting of seven separate arrays, each
containing eight amides 2 reacting in parallel with five a-bromoketones 3 in DMF with
Cs CO as a base at room temperature overnight (12 h). Tedious work-up was avoided by
2
3
simple parallel filtration of the DMF solutions to remove the Cs CO , which were then
2
3
submitted to preparative reversed phase HPLC purification. UV controlled collection allowed
the rapid purification of the products, which were UV active at 400 nm. Subsequent removal
of the elution solvents yielded a total of 205 isolated indole derivatives 5 in multi-milligram
quantities (2–30 mg). Predictably, no intermediate 4 was isolated in any case. The yields varied
between 10–92%, depending on the nature of the substituents. Representative examples are
shown in Scheme 2.
1
2
It was found that the nature of R affected the reaction only to a minor degree, whereas R
3
and R had a large impact. The reactions of aryl- or heteroaryl substituted amides 2 with aryl-
or heteroaryl substituted a-bromoketones which proceeded smoothly in 80% of the cases,
yielded the desired indoles 5 in a range of 4 to 92% (average yield: 47%) (Table 1, entry 1).
Amides that were substituted with electron withdrawing groups, however, afforded the prod-
ucts in lower average yields. Aryl- and heteroaryl substituted amides 2 also reacted smoothly
with cycloalkyl- and alkyl substituted a-bromoketones (13–69%, average yield: 35%), and
generally showed a good reactivity in 57% of the reactions producing indoles 5 (Table 1, entry
2
). Cycloalkyl substituted amides also reacted nicely in 67% respectively 73% of the examples
with a variety of a-bromoketones (Table 1, entries 3 and 4) affording the products in 4–66%
yield. It is believed that this scope demonstrates the broad applicability of this reaction
sequence.

8253
Scheme 2. Representative examples of isolated indoles 5
Table 1
Library reaction statistics
2
3
a
Entry
R
R
No. of performed reactions
Yield (%)
1
2
3
4
Aryl-/heteroaryl-
Aryl-/heteroaryl-
Cycloalkyl-
Aryl-/heteroaryl-
Cycloalkyl-/alkyl-
Aryl-/heteroaryl-
Cycloalkyl-/alkyl-
148 (118) 80%
28 (16) 57%
89 (60) 67%
15 (11) 73%
4–92 average: 47
13–69 average: 35
4–66 average: 39
16–66 average: 45
Cycloalkyl-
a
No. in parentheses indicates the number of reactions that yielded detectable amounts of product.
In conclusion, it was shown that out of 280 reactions of various amides with a-bromoketones
in a combinatorial fashion, the desired product was isolated in multi-milligram quantities, in
acceptable yields in 205 (73%) of the reactions. The novel reaction conditions (Cs CO in DMF)
2
3
employed proved to be the key for successful synthesis of the 2-acyl-3-amino-indole library 5.
The Cs CO was easily handled and DMF proved to be the solvent of choice due to its high
2
3
solublising properties. The crude products from this library were efficiently and rapidly purified
by automated reversed phase HPLC. To the best of our knowledge this is the first report of
library synthesis of 3-amino-2-acyl-indole derivatives. Based on these encouraging results the
chemistry efforts towards novel substituted and functionalised indole derivatives are currently
under investigation.

8254
Acknowledgements
Thanks to B. Mathys and C. M u¨ ller for technical assistance and Drs. A. Chucholowski, A.
Alanine and A. Thomas for helpful discussions.
References
1. (a) Balkenhohl, F.; von den Bussche-H u¨ nefeld, C.; Lansky, A.; Zechel, C. Angew. Chem., Int. Ed. Engl. 1996, 35,
2
1
288–2337. (b) Booth, S.; Hermkens, P. H. H.; Ottenheijm, H. C. J.; Rees, D. C. Tetrahedron 1998, 54,
5385–15443. (c) Thompson, L. A.; Ellmann, J. A. Chem. Rev. 1996, 96, 555–600.
2
3
4
. Sundberg, R. Indoles (Best synthetic methods); Academic, 1996.
. Viti, G.; Gianotto, D.; Nannicini, R.; Ricci, R.; Pestellini, V. J. Heterocycl. Chem. 1991, 28, 379–384.
. Typical procedure for the synthesis 2: A mixture of 10 mmol of aminobenzonitrile in 15 ml of DCM and 15 mmol
of pyridine was treated with 11 mmol of acid chloride and reacted over night (12 h) at room temperature. In the
cases where the product was soluble in 15 ml of DCM, water was added and the layers were separated by
filtration through hydrophobic filter devices (Macherey & Nagel) and the volatiles removed in vacuo. In the cases
where the product was not soluble in 15 ml of DCM, hexane was added and the precipitate collected, washed
with hexane and dried in vacuo. MS and HPLC at 230 nm corroborated the structures and purities.
. (a) Andrew, H. F.; Bradsher, C. K. J. Heterocycl. Chem. 1967, 4, 577–581. (b) Alper, H.; Stout, R. W. J.
Heterocycl. Chem. 1973, 10, 5–10. (c) Scrowston, R. M.; Shaw, D. C. J. Chem. Soc., Perkin Trans 1 1976,
5
6
7
7
49–754. (d) Kishi, I.; Imafuku, K.; Ogawa, K.; Matsushita, Y. Heterocycles 1990, 31, 677–681.
. (a) Wang, J.-X.; Zhang, M.; Xing, Z.; Hu, Y. Synth. Commun. 1996, 26, 301–305. (b) Bartsch, H.; Kropp, W.;
Pailer, M. Monatsh. Chem. 1979, 110, 257–265. (c) Xu, W.; Mohan, R.; Morissey, M. M. Tetrahedron Lett. 1997,
3
8, 7337–7340.
. (a) Sorrell, T. N.; Allen, W. E. J. Org. Chem. 1994, 59, 1589–1590. (b) Black, D. S.; Bowyer, M. C.; Bowyer, P.
K.; Ivory, A. J.; Kim, M.; Kumar, N.; McConnell, D. B.; Popiolek, M. Aust. J. Chem. 1994, 47, 1741–1750. (c)
Colonna, M. Gazz. Chim. Ital. 1948, 78, 502–507.
8
. (a) Ranganathan, D.; Farooqui, F. Tetrahedron Lett. 1984, 25, 5701–5704. (b) Semple, G.; Ryder, H.; Ohta, M.;
Satoh, M. Synth. Commun. 1996, 26, 721–727. (c) Marfat, A.; Carta, M. P. Synthesis 1987, 515–517.
. General procedure: To a solution of 0.3 mmol substituted benzonitrile 2 in 0.5 ml DMF was added 0.45 mmol
of a-bromoketones 3 in 0.5 ml of DMF, approximately 150 mg (0.5 mmol) of Cs CO and stirred over night (12
9
2
3
h) at rt. Filtration yielded a DMF solution, which was directly applied to reversed phase column chromatography
(Dynamax pumps and UV detector in combination with a Gilson 215 liquid handler on a YMC ODS-A column
(50×20mm)) eluting with a gradient of acetonitrile and water (20–95%). Evaporation of the elution solvents
yielded the desired indoles 5. MS corroborated the structures and the purities were determined by HPLC at 230
nm.
1
2
3
1
1
0. 5a (R =H, R =R =Ph) Yield: 51%; 250 MHz H NMR (CDCl ): l=8.28 (d, J=7.2 Hz, 1H, Ind-H7), 7.70 (d,
3
J=7.2 Hz, 1H, Ind-H4), 7.60 (t, J=7.2 Hz, 1H, Ind-H6), 7.40 (t, J=7.2 Hz, 1H, Ind-H5), 7.35 and 7.15 (m, 10H,
2
×Ph), 5.88 (s, br, 2H, NH2).
.
.