Synthesis of novel oxepine-containing polycyclic-fused quinoline systems

Introduction

Polycyclic-fused quinolines are an important group of compounds in medicinal chemistry [1–3] and are ubiquitous substructures associated with biologically active natural products such as Luotomins (1, Figure 1) [4] and Calothrixin B (2, Figure 1) [5]. Studies have shown that planar polycyclic-fused quinoline systems can cleave DNA [6] and exhibit significant biological properties such as antitumoral [7], anti-cancer [8], anti-inflammatory [9], antibacterial, and antituberculosis activities [10]. For example, owing to their aromatic character and planar structure, indenoquinoline derivatives (3, Figure 1) show potent antiproliferative activities against breast (MCF-7), lung epithelial (A-549), and cervical (HeLa) adenocarcinoma cells [3]. Accordingly, attention has been increasingly focused on the construction of various polycyclic-fused quinoline systems. Most reports in the literature describe a common heteroaromatic system, such as indene [3], pyrazole [11], pyran [12], indole [13], benzofuran [14], and benzothienoquinoline [15], fused to a quinoline system, and synthetic analogs thereof. However, to the best of our knowledge, there are very few reports on a medium-sized seven-membered oxepin ring fused to a quinoline unit [16].

Figure 1

Structures of natural products and synthetic compounds 1–6.

The seven-membered oxepine system is a pharmaceutically important structural unit of some natural products such as tournefolic acid B (4, Figure 1) [17] and bauhiniastatin 1 (5, Figure 1) [18], and has found wide application in the design and discovery of novel bioactive molecules and drugs. Some compounds with the oxepine ring have been reported to exhibit potent biological activities such as antifungal [19], anxiolytic [20], cyclooxygenase inhibitory [21], and aminopeptidase N (APN) inhibitory activities [22]. Given the importance of the oxepine core as a pharmacophore, extensive synthetic efforts have been devoted to the design and synthesis of oxepine-containing heterocycles [23–26]. A recent review concerning the synthesis of oxepines has been reported [27].

It is well known that the development of hybrid compounds through the combination of different chromophores or pharmacophores in a molecular framework is an important goal of synthetic organic chemistry and plays a prominent role in modern drug discovery [28]. In light of these findings, it would be of synthetic importance to incorporate the oxepine nucleus into planar polycyclic-fused quinoline systems leading to new prototypes, as possible drug candidates for pharmacological studies. Thus, in the context of our ongoing work on the synthesis of fused quinolines [29–32], we would like to report herein the synthesis of novel oxepine-containing polycyclic-fused quinoline systems, wherein the oxepine ring is fused at its 2,3-position to the b-position of the quinoline system to give a compact structure exemplified by compounds 6 shown in Figure 1. To the best of our knowledge, this simple polycyclic-fused system has never been previously described in the literature.

Results and discussion

Recently, we have reported a direct and efficient synthesis of ethyl 2-(chloromethyl)benzo[h]quinoline-3-carboxylate (8 in Scheme 1) and its application in the one-pot synthesis of 2-(benzofuran-2-yl)benzo[h]quinoline-3-carboxylic acid derivatives [29]. The inspiring results obtained prompted us to further study the synthetic application of 8 associated with the construction of a polycyclic-fused oxepine-containing quinoline system. Our synthetic plan, as shown in Scheme 1, was to use the 2-chloromethyl group of 8 to generate the aryloxymethyl moiety via the Williamson reaction with substituted phenols followed by the transformation of the ethyl ester at its 3-position into corresponding quinoline-3-carboxylic acids to afford intermediates 7. Subsequently, the intramolecular Friedel-Crafts acylation reaction of 7 would result in the formation of the oxepine ring by the generation of a new C-C bond.

Scheme 1

Synthetic route to benzo[h]areno[6,7]oxepino[3,4-b]quinolin-8(14H)-ones 6a–h.

Ethyl 2-(chloromethyl)benzo[h]quinoline-3-carboxylate (8) was first subjected to the Williamson reaction with substituted phenols in the presence of NaOEt as base in refluxing anhydrous EtOH. In this reaction, the NaOEt/EtOH base system was selected because the formation of a precipitate provided a convenient visual indicator of the reaction progress. Because the formed aryloxymethyl ether from the Williamson reaction did not interfere with further ester hydrolysis reaction, purification at this stage was unnecessary. Thus, we simply added 80% ethanolic potassium hydroxide solution to the reaction mixture and continued the reflux for an additional 2 h. After the reaction was complete followed by an acidic workup, the corresponding free quinoline-3-carboxylic acids 7a–f were obtained in good overall yields of 78–85%. Owing to the simplicity and efficiency of the one-pot synthesis, we further attempted the reaction of 8 with 1-naphthol and 2-naphthol. Interestingly, these substrates were equally amenable to the one-pot reaction sequence and gave rise to the corresponding 2-[(naphthalen-1-yloxy)methyl]benzo[h]quinoline-3-carboxylic acid (7g) and 2-[(naphthalen-2-yloxy)methyl]benzo[h]quinoline-3-carboxylic acid (7h) in 75% and 77% yields, respectively. Products 7a–h are new compounds and their structures were confirmed by both spectral data and elemental analyses.

Having in hand a series of the newly synthesized intermediates 7a–h, our attention was transferred to their intramolecular Friedel-Crafts acylation reaction for building the desired oxepino-fused quinoline systems. At this stage, we investigated the reaction using POCl₃ according to the method recently reported by Meesala and Nagarajan [33]. However, in the present case, these conditions failed to efficiently promote the reaction and none of the expected products were detected. We also attempted to conduct the reaction using polyphosphoric acid (PPA) as the cyclization agent [30, 31]. Unfortunately, the reaction did not occur at temperatures lower than 150°C even after 12 h. At the temperature above 150°C, black tarry byproducts were formed, which significantly reduced the yield of the desired products. Despite numerous trials under different conditions, the yield of this reaction could not be optimized. After these fruitless attempts, it was decided to focus on the use of Eaton’s reagent (phosphorus pentoxide in methanesulfonic acid) [31, 32]. To our delight, with Eaton’s reagent the intramolecular cyclization reaction of 7a proceeded smoothly at 80°C with advantages in ease of handling, low reaction temperature, and good yield (76%). The ease of isolation of the product was notable. After simple aqueous workup and crystallization from ethanol, the corresponding product was isolated in an analytically pure form.

Encouraged by the successful synthesis of 7a, we tested this protocol to the preparation of other substrates 7b–h. With the exception of 7f, good yields of the remaining products in the range of 69–80% were obtained. The intramolecular Friedel-Crafts acylation reaction of 7f failed to give the desired cyclized product, even after raising the reaction temperature and prolonging the reaction time. This negative result can be explained in terms of the presence of the strongly electron-withdrawing NO₂ group, which renders the benzene ring highly electron-deficient, thereby retarding the cyclization process.

We would also like to comment on the transformation of 7h to the cyclized product 6h. In principle, the intramolecular cyclization reaction of 7h might produce two possible structural isomers 6h and 6h′ because of the existence of two vacant ortho sites, the positions 1- and 3-, as indicated in Scheme 2. However, the cyclization is regioselective occurring at the more electron-rich position 1 to give 6h as the sole product as evidenced by analysis of the ¹H NMR spectrum. In particular, the signal of the H-4 proton is a doublet at 8.52 ppm with a coupling constant J = 8.4 Hz, which is a typical value for ortho-protons coupling. If the cyclization reaction occurred at the 3-position affording the corresponding product 6h′, the H-4 proton signal in the ¹H NMR spectrum should have been a singlet. A similar observation was also reported by Wang et al. for the intramolecular cyclization reaction of 2-naphthyl-substituted benzoic acid [34].

Scheme 2

Possible intramolecular cyclization reactions of 7h.

Conclusions

A series of structurally new oxepine-containing polycyclic-fused quinoline systems, benzo[h]areno[6,7]oxepino[3,4-b] quinolin-8(14H)-ones, via the intramolecular Friedel-Crafts acylation reaction of 2-(aryloxymethyl)benzo[h]quinoline-3-carboxylic acids was synthesized. Biological activities of the newly synthesized compounds are currently being studied.

Experimental

The chemicals used in this work were obtained from Fluka and were used without purification. Melting points (uncorrected) were determined by using WRS-1B melting points apparatus. ¹H NMR (600 MHz) and ¹³C NMR (150 MHz) spectra were recorded on a Brucker AVANCE NMR spectrometer. Elemental analyses were performed for C, H, and N using an Elementar Vario EL-III element analyzer. The progress of reactions was monitored by thin layer chromatography (TLC) on silica gel GF254 using ethyl acetate/petroleum ether (1:6) as eluent. Anhydrous ethanol was used in the synthesis work.