Synlett 2014; 25(10): 1473-1477
DOI: 10.1055/s-0033-1341241
letter
© Georg Thieme Verlag Stuttgart · New York

Synthesis of 3,3-Disubstituted 2-Aminoindolenines by Palladium-Catalyzed Allylic Amidination with Isocyanide

Takeshi Nanjo
Graduate School of Pharmaceutical Sciences, Kyoto University, Yoshida, Sakyo-ku, Kyoto 606-8501, Japan   Fax: +81(75)7534569   Email: takemoto@pharm.kyoto-u.ac.jp
,
Chihiro Tsukano
Graduate School of Pharmaceutical Sciences, Kyoto University, Yoshida, Sakyo-ku, Kyoto 606-8501, Japan   Fax: +81(75)7534569   Email: takemoto@pharm.kyoto-u.ac.jp
,
Yoshiji Takemoto*
Graduate School of Pharmaceutical Sciences, Kyoto University, Yoshida, Sakyo-ku, Kyoto 606-8501, Japan   Fax: +81(75)7534569   Email: takemoto@pharm.kyoto-u.ac.jp
› Author Affiliations
Further Information

Publication History

Received: 19 February 2014

Accepted after revision: 24 March 2014

Publication Date:
30 April 2014 (online)

 


Abstract

Synthesis of 3,3-disubstituted 2-aminoindolenines was achieved by palladium-catalyzed allylic amidination with an isocyanide. It was found that isocyanides are effective building blocks in palladium-catalyzed allylic functionalizations, analogous to carbon monoxide. This approach enables the direct construction of the indolenine ring along with the formation of a quaternary carbon and the introduction of an amino substituent in one step under mild conditions.


#

3,3-Disubstituted indole skeletons are one of the most important structures in alkaloid chemistry and are present in a wide range of natural products and pharmaceuticals.[1] Among these derivatives, 3,3-disubstituted 2-aminoindolenines and 2-aminoindolines are substructures found in biologically active natural products such as flustramine C,[2] perophoramidine,[3] and quinadoline B[4] (Figure [1]). These compounds are structurally complex, and much effort has been put into the synthetic study toward them.[5] There are three key aspects in the synthesis of 3,3-disubstituted 2-aminoindolenines; (i) construction of the indole ring, (ii) formation of the quaternary carbon at the 3-position, and (iii) introduction of the amino substituent at the 2-position (Scheme [1]). A direct, one-step method would be an efficient approach to these structures, but to date there have been no published reports on the direct construction of 3,3-disubstituted 2-aminoindolenines.

Zoom Image
Figure 1 Bioactive 3,3-disubstituted 2-aminoindole derivatives
Zoom Image
Scheme 1 Allylic amidination for 3,3-disubstituted 2-aminoindolenine derivatives

Our group has previously developed a synthetic method for indole derivatives which involves the formation of a C2–C3 bond.[6] Based on this characteristic retrosynthetic analysis, the efficient construction of various 3,3-disubstituted indole derivatives was achieved. We also applied this strategy to the synthesis of 2-iminoindolines via SmI2-mediated reductive cyclization of carbodiimides.[6b] Although this method is an effective approach for 3,3-disubstituted 2-aminoindolenine derivatives, stepwise introduction of the nitrogen unit and construction of the indolenine skeleton was required, meaning a one-step method was still needed. Herein we describe a direct and general approach for the construction of 3,3-disubstituted 2-aminoindolenines by palladium-catalyzed allylic amidination with an isocyanide (Scheme [1]).

Table 1 Investigation of Reaction Conditions

Entry

R

Ligand

Base

Solvent

Yield (%)a

 1b

CO2Me

none

none

toluene

18

 2

CO2Me

Ph3P

none

toluene

35

 3

CO2Me

DPPEc

none

toluene

15

 4

CO2Me

BINAPc

none

toluene

20

 5

CO2Me

(o-tolyl)3P

none

toluene

36

 6

CO2Me

(2-furyl)3P

none

toluene

42

 7

Ac

(2-furyl)3P

none

toluene

55

 8

Ac

(2-furyl)3P

Et3N

toluene

65

 9

Ac

(2-furyl)3P

Et3N

MeCN

19

10

Ac

(2-furyl)3P

Et3N

DCE

16

11

Ac

(2-furyl)3P

Et3N

DMF

30

12

Ac

(2-furyl)3P

Et3N

1,4-dioxane

71

13

Ac

(2-furyl)3P

Et3N

THF

73

14d,e

Ac

(2-furyl)3P

Et3N

THF

64

15d,f

Ac

(2-furyl)3P

Et3N

THF

63

a Yield of isolated product.

b Pd(PPh3)4 was used instead of Pd(dba)2.

c 10 mol % of ligand were used.

d The reaction was performed at 50 °C.

e Conditions: 5 mol% of Pd(dba)2 and 10 mol% of (2-furyl)3P were used.

f Conditions: 2 mol% of Pd(dba)2 and 4 mol% of (2-furyl)3P were used.

The reaction shown in Scheme [1] was designed to develop the direct method. Reaction of isocyanide 1 bearing an allyl ester moiety with an amine in the presence of a palladium catalyst would give 3,3-disubstituted 2-aminoindolenine 2 via oxidative addition and allylic amidination. Recent publications disclosed that isocyanides are useful building blocks for multicomponent reactions (e.g., Passerini and Ugi reactions) as well as for palladium-catalyzed reactions.[7] [8] [9] [10] However, there are only a few reports of palladium-catalyzed allylic functionalization reactions with isocyanides, unlike the well-­developed chemistry of carbon monoxide, which is isoelectronic with an isocyanide.[11] Usually, palladium-catalyzed allylic functionalization reactions occur at the less sterically hindered site.[11a] [d] Our approach, however, would enable the formation of a quaternary carbon at the more sterically hindered site under mild conditions, owing to intramolecular cyclization. This method would also enable the introduction of various amino substituents at the 2-position by using a range of amines.

To investigate the allylic amidination, we synthesized substrate 1a bearing an isocyanide and an allyl carbonate moiety.[12] Treatment of 1a with piperidine and 10 mol% of Pd(PPh3)4 in toluene at room temperature afforded 2-aminoindolenine 2a in 18% yield (Table [1], entry 1). Using Pd(dba)2 and Ph3P instead of Pd(PPh3)4 increased the yield of 2a to 35% (Table [1], entry 2). Next, several ligands were screened. It was found that monodentate triarylphosphines were effective, and (2-furyl)3P gave the best results in this reaction (Table [1], entries 3–6), however, the yield of 2a was still below 50%. We assumed that the low yields were caused by high reactivity of the allyl carbonate moiety in substrate 1a, thus substrate 1b bearing an allyl acetate moiety was used instead. As a result, the yield of the desired product 2a increased to 55% (Table [1], entry 7). When substrate 1a was used, additional base was not necessary for this reaction, but the addition of two equivalents of Et3N when using 1b increased the yield of 2a (Table [1], entries 7, 8). Further optimization revealed that THF was the best solvent (Table [1], entries 9–13). Although lowering the amount of catalyst slightly decreased the yield of 2a, catalyst loadings as low as 2 mol% were sufficient for complete conversion (Table [1], entries 14, 15).

Zoom Image
Scheme 2 Investigation of substrate scope
Zoom Image
Scheme 3 Investigation using l-proline methyl ester and an allyl acetate 1g

Next we investigated the substrate scope of the reaction under the optimal conditions (Scheme [2]).[13] [14] Initially, a range of amines were examined as the nucleophile. The reactions of simple cyclic amines gave the desired products 2b and 2c in 58% and 50% yields, respectively. Morpholine could also be used in this reaction (2d). The reaction of acyclic amines such as diethylamine and benzylmethylamine also gave good results (2e and 2f). However, using a primary amine gave a low yield of the desired 2-aminoindolenine 2g. When N-methylaniline was used, the desired product 2h was not obtained, probably because of the weaker nucleophilicity. Next the reaction was performed using several isocyanides. The reactions of substrates bearing trifluoromethyl and methoxy groups at the para position of the aromatic ring gave the corresponding products 2i and 2j in 45% and 69% yields, respectively. An alkyl substituent on the olefin did not significantly influence the yield of the products (2k and 2l).

Next we performed the reactions using a chiral amine and an allyl acetate 1g derived from a secondary alcohol (Scheme [3]). The optimal conditions were applied to the reaction using l-proline methyl ester as the nucleophile and interestingly, the desired product 2m was obtained as a 3:1 mixture of diastereomers at the 3-position of the indolenine ring (Scheme [3], eq 1). This result indicates that the configuration of the quaternary carbon was influenced by the steric effect of the nucleophile. When allyl acetate 1g was used, the desired product 2n was obtained as a 2:1 mixture of olefin geometric isomers (Scheme [3], eq 2).

Zoom Image
Scheme 4 Plausible mechanism of allylic amidination

A plausible mechanism is shown in Scheme [4]. Firstly, oxidative addition of allyl acetate 1b to palladium(0) generates allylpalladium complex A. There are two possibilities for the next step. One is that intramolecular isocyanide insertion proceeds to form intermediate B, and then C–N reductive elimination regenerates the palladium(0) species and affords the desired 2-aminoindolenine 2a (path A). The other possibility is that nucleophilic addition to the isocyanide, activated by the palladium(II), occurs to give palladacycle C followed by C–C reductive elimination (path B). Path B is analogous to the proposed catalytic cycle of the palladium-catalyzed decarboxylative cyclization reaction reported by Hayashi and co-workers.[11a] Considering the scope of this reaction and the nucleophilicity of the amine to isocyanide of intermediate A, we believe that path B is dominant.[15]

In summary, we have developed the synthesis of 3,3-disubstituted 2-aminoindolenine derivatives by palladium-catalyzed allylic amidination of isocyanides. This approach enables the direct construction of an indolenine ring along with the formation of a quaternary carbon and introduction of an amino substituent under mild conditions. We are currently investigating the mechanistic detail of the reaction and extending the strategy to an asymmetric reaction.


#

Acknowledgment

This work was supported by a Grant-in-Aid for Scientific Research on the Innovation Area ‘Molecular Activation Directed toward Straightforward Synthesis’ from The Ministry of Education, Culture, Sports, Science and Technology, Japan (C.T.), and JSPS Research Fellowships for Young Scientists (T.N.).

Supporting Information



Zoom Image
Figure 1 Bioactive 3,3-disubstituted 2-aminoindole derivatives
Zoom Image
Scheme 1 Allylic amidination for 3,3-disubstituted 2-aminoindolenine derivatives
Zoom Image
Scheme 2 Investigation of substrate scope
Zoom Image
Scheme 3 Investigation using l-proline methyl ester and an allyl acetate 1g
Zoom Image
Scheme 4 Plausible mechanism of allylic amidination