Satish P. Nikumbhab,
Akula Raghunadha,
V. Narayana Murthya,
Rajesh Jinkalaa,
Suju C. Josephab,
Y. L. N. Murthyb,
Bagineni Prasadc and
Manojit Pal*c
aTechnology Development Centre, Custom Pharmaceutical Services, Dr Reddy's Laboratories Ltd, Hyderabad 500049, India
bDepartment of Organic Chemistry, Andhra University, Visakhapatnam 530003, India
cDr Reddy's Institute of Life Sciences, University of Hyderabad Campus, Gachibowli, Hyderabad 500 046, India. E-mail: manojitpal@rediffmail.com; Tel: +91 40 6657 1500
First published on 24th August 2015
A catalyst, ligand and solvent free method for double heteroarylation of N, O and S nucleophiles has been developed for the first time leading towards the synthesis of compounds containing an indole ring fused with pyrrolo-, furo- and thieno[2,3-b]quinoxaline moieties. This general and greener approach afforded novel compounds of medicinal importance.
While various polynuclear fused N-heteroarenes are known in the literature4 molecules represented by A are unknown. It was therefore necessary to develop a suitable and general synthetic route for accessing molecules designed based on A. A retrosynthetic analysis of compound A revealed that its synthesis can be achieved via a 2-(1H-indol-3-yl)quinoxaline species B which in turn could be accessed from an indole derivative C (Fig. 2). While this route appeared to be feasible the major challenging issues however emerged as if conversion of B to A could be performed (i) with high efficiency (considering unknown reactivity of B towards the reactants employed), (ii) under metal catalyst free conditions, (iii) via a single-step method that is common and general for X = NR2, S, O etc. (Fig. 1). We envisaged that both the leaving groups of B could be displaced by a single nucleophile in a single step under appropriate reaction conditions to afford A. Herein we report our preliminary results on a catalyst, ligand and solvent free greener synthesis of A or 3 from 1 (Scheme 1).
The key starting material 1 was prepared from indolin-2-one (4) that was converted to 5 (Scheme 2).5 The quinoxaline ring was constructed using the oxo-ester side chain of 5 following a similar procedure reported earlier6 to give 6. The chlorination of 6 afforded 1 that was either used directly (or after N-methylation) for the next step. The Pd-catalyzed double N-arylation of primary amines with 2,2′-dihalobiphenyl has become a powerful method for the construction of central 5-membered pyrrole ring of various carbazole derivatives.7 The methodology has been used for the elegant synthesis of several carbazole based alkaloids or natural products. We took inspiration from these works and decided to use a similar approach i.e. a double heteroarylation strategy for our synthesis. Accordingly, the compound 1a was reacted with amine 2a under various conditions (Table 1). Initially, several combinations of a Pd catalyst [e.g. Pd(OAc)2, Pd2(dba)3 and PdCl2] and a ligand (e.g. xantphos or BINAP) were used for the coupling of 1a with 2a at 80–100 °C depending on the solvent used e.g. 1,4-dioxane, DMF, toluene and MeCN (entries 1–8, Table 1). Generally, Cs2CO3 was used as a base in most of these cases. While the reaction proceeded in all these cases the yield of desired product 3a was not satisfactory except in two cases (entries 2 & 3, Table 1) and was poor when Et3N was used as a base (entry 5, Table 1). We then performed the reaction in the absence of any Pd-catalyst and ligands (entries 6–13, Table 1). To our surprise, the product yield was improved significantly in these cases. Change of base from Cs2CO3 to K2CO3 or DBU did not affect the yield of 3a dramatically (entries 14 & 15, Table 1) whereas Et3N once again found to be an inferior base in the present reaction (entry 16, Table 1). Interestingly, the product yield was continued to increase when the reaction was performed in the absence of any base (entries 17 & 18, Table 1) and finally in the absence of any solvent (entry 19, Table 1). Indeed the reaction was completed within 6 h in these cases affording 3a in 83–87% yield. Though it required marginally higher reaction temperature, we were delighted particularly with the catalyst, ligand and solvent free method (entry 19, Table 1) as this approach not only avoids the environmental hazard but also reduce the cost. Moreover, the product 3a was isolated in pure form without performing any chromatographic purification process [i.e. after completion of the reaction, the mixture was cooled to room temperature, diluted with cold water, filtered, and the solid obtained was titrated with methyl t-butyl ether (MTBE), see ESI†]. Thus the condition of entry 19 of Table 1 was appeared to be optimal and used for further study.
Entry | Catalyst/ligand | Base | Solvent | %Yieldb |
---|---|---|---|---|
a Reaction conditions: 1a (1.0 mmol), amine 2a (1.1 mmol) and a base (3 mmol) at 80–100 °C for 8 h.b Isolated yield.c The reaction was performed for 6 h.d The reaction was performed at 135 ± 5 °C. | ||||
1 | Pd(OAc)2/xantphos | Cs2CO3 | 1,4-Dioxane | 40 |
2 | Pd(OAc)2/BINAP | Cs2CO3 | 1,4-Dioxane | 62 |
3 | Pd(OAc)2/BINAP | Cs2CO3 | 1,4-Dioxane | 60 |
4 | Pd2(dba)3/BINAP | Cs2CO3 | 1,4-Dioxane | 50 |
5 | Pd(OAc)2/BINAP | Et3N | 1,4-Dioxane | 5 |
6 | PdCl2/BINAP | Cs2CO3 | DMF | 10 |
7 | Pd(OAc)2/BINAP | Cs2CO3 | Toluene | 30 |
8 | Pd(OAc)2/BINAP | Cs2CO3 | MeCN | 10 |
9 | — | Cs2CO3 | 1,4-Dioxane | 50 |
10 | — | Cs2CO3 | DMF | 75 |
11 | — | Cs2CO3 | DMSO | 70 |
12 | — | Cs2CO3 | o-Xylene | 65 |
13 | — | Cs2CO3 | NMP | 76 |
14 | — | K2CO3 | DMF | 68 |
15 | — | DBU | Toluene | 69 |
16 | — | Et3N | DMF | 23 |
17 | — | — | DMF | 83c |
18 | — | — | NMP | 82c |
19 | — | — | — | 87c,d |
To test the generality and scope of this method the optimized reaction conditions were applied to a variety of substrates e.g. 1a, b and 2a–o (Table 2). Thus amines containing aliphatic and aromatic side chain or functional groups like ether, hydroxyl etc. or a chiral center in the side chain were examined. The reaction proceeded well in all these cases affording the desired products in good to excellent yields (entries 1–12, Table 2). The reaction also proceeded well with other N, O and S nucleophiles when the reaction time was shorter (entries 13–17, Table 2). Thus, indole fused with furo- and thieno[2,3-b]quinoxalines (3m, 3n, 3p and 3q) were prepared by reacting 1a, b with NaOH and Na2S separately. All these reactions were performed under open air as the process was not sensitive towards aerial oxygen or moisture. Moreover, NaCl or HCl (that can be neutralized by NaOH to harmless NaCl) being the byproduct in these reactions all the products were isolated in pure form after treating with water followed by titration with MTBE (see ESI†). Notably, the Pd-based strategy for double arylation7 (leading to carbazoles) was successful only with amines and the use of O- or S- reactants is not common.8a Though a different strategy has been reported8b,c for the double arylation of S- reactants (leading to dibenzo[b,d]thiophene derivatives) involving the sequential use of n-BuLi followed by S2Cl2 (as sulfur source) none is known for double arylation of oxygen reactants leading to dibenzo[b,d]furan derivatives. Moreover, all these reactions required the use of an inert or anhydrous atmosphere. Thus our strategy presented here appeared to be a general one as all N, O and S nucleophiles could be reacted with 1 under a common reaction condition. To test the scale-up potential of this method the reaction of 1a with 2a (cf. entry 19, Table 1) was performed in g scale [i.e. 1.57 g (∼5 mmol) of 1a and 5.1 mmol of 2a] when 3a was isolated in 93% yield.
Entry | Compd (1), R1 = | Nucleophile (2), X, Y, Z = | T (h) | Product (3) | Yieldb (%) |
---|---|---|---|---|---|
a The reaction was carried out using 1 (1.0 mmol), and 2 (1.1 mmol) at 135 ± 5 °C.b Isolated yield. | |||||
1 | 1a; H | 2a; PhCH2N, H, H | 6 | 3a | 87 |
2 | 1a | 2b; Ph(CH2)2N, H, H | 5 | 3b | 89 |
3 | 1a | 2c; (pyridin-2-yl)CH2N, H, H | 6.5 | 3c | 85 |
4 | 1a | 2d; EtO(CH2)3N, H, H | 5 | 3d | 86 |
5 | 1a | 2e; p-MeOC6H4CH2N, H, H | 5 | 3e | 87 |
6 | 1a | 2f; MeO(CH2)3N, H, H | 5 | 3f | 87 |
7 | 1a | 2g; Me(CH2)5N, H, H | 5 | 3g | 85 |
8 | 1a | 9 | 3h | 78 | |
9 | 1a | 2i; n-BuN, H, H | 5 | 3i | 80 |
10 | 1a | 2j; 3,4-di-MeOC6H3(CH2)2N, H, H | 5 | 3j | 82 |
11 | 1a | 2k; HO(CH2)2N, H, H | 4 | 3k | 78 |
12 | 1a | 2l; 3,5-di-MeC6H3CH2N, H, H | 7.5 | 3l | 79 |
13 | 1a | 2m; S, Na, Na | 1 | 3m | 91 |
14 | 1a | 2n; O, Na, H | 2 | 3n | 67 |
15 | 1a | 2o; NH2N, H, H | 1.5 | 3o | 85 |
16 | 1b; Me | 2n | 1.5 | 3p | 70 |
17 | 1b | 2m | 0.5 | 3q | 91 |
Mechanistically, the reaction may follow either “path a” or “path b” (Scheme 3). The path a involve a nucleophilic attack on the chloro group bearing C-2 of the quinoxaline ring to give E-1 whereas path b involve a similar attack on the chloro group bearing C-2 of the indole ring leading to E-2. Both E-1 and E-2 can undergo a second nucleophilic attack on its C–Cl moiety in an intramolecular fashion to afford the product 3.9a To gain evidence on which path was actually followed we revisited the reaction of 1a with 2a under various conditions. This allowed us to isolate an intermediate (4) in 53% yield along with 3a when the reaction was performed at 120 °C and stopped after 3 h (Scheme 4). While the initial spectral/analytical data indicated compound 4 as a monochloro derivative [(HRMS: m/z [M + 1] calcd for C23H18N4Cl (M + H): 385.1220; found: 385.1207) that was formed after displacing one of the two chloro groups of 1a by 2a] two alternative structures i.e. 4A and 4B (Fig. 3) were possible in this case. However, the structure of 4 was confirmed as 4A that showed single NOE interaction between NH and C(7)–H of indole ring (Fig. 3). Notably, two NOE interactions were expected in case of 4B. Finally, the intermediate 4 afforded 3a when heated at 130 °C for 3 h indicating its intermediacy in the present reaction. It is to be mentioned that compound 4 did not react with 2a when treated with 2a in a separate experiment but afforded 3a indicating intramolecular cyclization of 4 leading to 3a was preferred over other reaction.9b Thus, all these observations suggested that the reaction followed path a rather than path b.
In view of known anti-tumor/anti-cancer properties of related indolo[2,3-b]quinoxaline derivatives,2 all the synthesized compounds were evaluated for their ability to inhibit the growth of cancer cells. These compounds were tested at 10 μM against three cancer cells e.g. A549 (lung), MCF-7 (breast) and TZM-BL (cervical) using the sulphorhodamine B (SRB) assay10a,b with gemcitabine10c as a reference compound. Among these compounds, 3a–l showed >90% inhibition against lung cancer cells (comparable to gemcitabine's 90% inhibition), >75% inhibition against breast cancer cells (gemcitabine 49%), and >60% against cervical cancer cells (gemcitabine 90%). Moreover, these compounds did not showed significant effect on normal HEK 293T cells (10–15% inhibition vs. gemcitabine's 25%) indicating their selectivity toward cancer cells.
In conclusion, a straightforward yet innovative method has been developed for the double heteroarylation of N, O and S nucleophiles leading towards the synthesis of polynuclear N-heteroarenes. This operationally simple, general, greener and cost-effective method can be performed under open air and is amenable for scale-up. The methodology afforded a library of novel compounds containing indole ring fused with pyrrolo-, furo- and thieno[2,3-b]quinoxaline moiety without the need of chromatographic purification. Several of these compounds showed promising and selective cytotoxicities against cancer cells. Overall, being unprecedented the present strategy of catalyst/ligand/solvent free double heteroarylation of N, O and S nucleophiles could be useful in accessing a wide range of novel bioactive molecules of medicinal importance.
Footnote |
† Electronic supplementary information (ESI) available: Experimental procedures, spectral data for all new compounds, and copies of spectra. See DOI: 10.1039/c5ra16727b |
This journal is © The Royal Society of Chemistry 2015 |