Design, Synthesis, and In Vitro Evaluation of Benzofuro[3,2-c]Quinoline Derivatives as Potential Antileukemia Agents

Herein, we design and synthesize an array of benzofuro[3,2-c]quinolines starting from 3-(2-methoxyphenyl)quinolin-4(1H)ones via a sequential chlorination/demethylation, intramolecular cyclization pathway. This sequential transformation was efficient, conducted under metal-free and mild reaction conditions, and yielded corresponding benzofuro[3,2-c]quinolines in high yields. In vitro biological evaluation indicated that such type of compounds showed excellent antileukemia activity and selectivity, and therefore may offer a promising hit compound for developing antileukemia compounds.


Introduction
Furo [3,2-c]pyridines belong to a novel class of heterocycles that exhibit a wide range of biological activities [1][2][3]. A large number of heterocyclic compounds bearing a furo [3,2-c]pyridine core have been reported as anticancer ( Figure 1A, compounds A and B) [4,5] or antibacterial agents ( Figure 1A, compound C) [6]. Among them, the tricyclic furo [3,2-c]quinoline skeleton possessing a fused aryl ring on the pyridine part has also shown unique pharmaceutical and biological activities. For example, compound D has been found to be a neuroprotective agent [7]. Compound E was reported to be a potent antiproliferative molecule against renal UO-31, melanoma UACC-257, and UACC-62 cell lines [8]. Agrawal et al. reported that compound F may be a potential topoisomerase-II inhibitor based on related computational studies [9]. In view of the diverse biological activities of such molecules, the design and synthesis of novel heterocyclic structures possessing the furo [3,2-c]pyridine skeleton would be highly valuable for biological studies. As our continuous research interest, herein we design and synthesize a new type of tetracyclic benzofuro [3,2-c]quinoline ( Figure 1B, compound G) that contains benzofuro [3,2-c]quinoline ( Figure 1B, compound G) that contains the key furo [3,2-c]pyridine skeleton by introducing a fused aryl ring on the furan side. Although such types of tetracyclic heterocycles have been reported by several research groups, related biological studies are still very limited [10]. Based on our newly synthesized compounds, an in vitro biological evaluation was then conducted.
Recently, our group has developed an efficient route for the synthesis of 3-(2-methoxyphenyl)quinolin-4(1H)-ones 1 via an in situ Meinwald rearrangement/intramolecular reductive cyclization of chalcone epoxides ( Figure 1B) [11,12]. This transformation was easily scaled up, showed a broad substrate scope, and tolerated a variety of functional groups. With this method developed, we envisioned that an efficient synthesis of benzofuro [3,2-c]quinoline G could be designed via demethylation followed by either one-pot or stepwise cyclization starting from 3-(2-methoxyphenyl)quinolin-4(1H)-ones 1 ( Figure 1B). A batch of methods for the synthesis of such compounds have been reported, such as the palladium-catalyzed aromatization reaction [10], O-arylation/dehydrogenative cross-coupling reaction [13], C-O coupling reaction [14], demethyl-cyclization reaction [15], and many others [16][17][18]. The current method is featured due to it using a readily accessible starting material and involving the easy installation of different functional groups and the mild and metal-free reaction conditions to access a number of substituted benzofuro [3,2-c]quinoline derivatives for in vitro biological evaluations. Furthermore, it was found that this type of compound showed promising antileukemia activities.

Chemistry
Initially, 3-(2-methoxyphenyl)quinolin-4(1H)-one 1a, which was conveniently obtained from the corresponding chalcone epoxide, was chosen as the model substrate for the synthesis of benzofuro [3,2-c]quinoline 2a via a demethylation/cyclization cascade. However, upon a thorough condition screening, the formation of 2a could not be observed at all [Supplementary materials]. In order to access the desired product 2a from the readily available starting substrate 1a, a stepwise reaction was then developed (Scheme 1). First, by treating with suitable chlorination reagents, 1a was converted into the corresponding 4-chloroquinine 3a. Upon carefully investigating condition optimizations, SOCl 2 was found to be the most effective, yielding 3a with a 93% yield using CH 2 Cl 2 as the solvent. Treating 3a with 48% HBr under reflux gave the demethylation product 4a with a 98% yield. The chlorination and demethylation steps were conducted in a one-pot manner without the isolation of 3a, and the overall yield of 4a was maintained at a high level (96%). Finally, a simple intramolecular annulation of 4a with KOt-Bu led to the formation of the expected benzofuro [3,2- With the optimized reaction conditions identified, the scope of this stepwise chlorination/demethylation/cyclization transformation was investigated by preparing different substituted benzofuro[3,2-c]quinolines (Table 1). Several substrates bearing different functional groups on both the benzofuran and the quinoline were prepared and subjected to the standard reaction conditions. Halogen substituents at different positions on the benzofuran ring were all well tolerated, giving the corresponding tetracyclic products in high overall yields (2c, 2f-2i). A substrate bearing a trifluoromethoxyl substituent on the benzofuran ring also worked well (2k). With the methyl substituent at different positions on the benzofuran ring, this two-step transformation ran efficiently, providing the corresponding cyclization products in good overall yields (2b, 2d, and 2j). The substrate scope with different substituents on the quinoline moiety was then investigated. Both the electron-donating methyl substituent (2n, 2p, 2t, and 2v) and electron-withdrawing halogen substituents (2o and 2r-2s) at different positions all gave the corresponding cyclization products in good overall yields. A free hydroxyl substituent was introduced to different positions on both the benzofuran and quinoline rings, with the aim of increasing the potential biological activities of these molecules [19][20][21]. While the one-pot chlorination/demethylation step was still very efficient, the subsequent base-promoted cyclization for these hydroxyl-substituted substrates generally gave low yields of the corresponding cyclization products. The low yield of the second step may have come from the competitive reactions between the free hydroxyl groups with the base (2e, 2l, 2q, 2u, and 2w-2x), and we speculate that the low yields of the second steps may be improved via selective protection of the hydroxyl group, which is not involved in the cyclization reaction.
With the optimized reaction conditions identified, the scope of this stepwise chlorination/demethylation/cyclization transformation was investigated by preparing different substituted benzofuro[3,2-c]quinolines (Table 1). Several substrates bearing different functional groups on both the benzofuran and the quinoline were prepared and subjected to the standard reaction conditions. Halogen substituents at different positions on the benzofuran ring were all well tolerated, giving the corresponding tetracyclic products in high overall yields (2c, 2f-2i). A substrate bearing a trifluoromethoxyl substituent on the benzofuran ring also worked well (2k). With the methyl substituent at different positions on the benzofuran ring, this two-step transformation ran efficiently, providing the corresponding cyclization products in good overall yields (2b, 2d, and 2j). The substrate scope with different substituents on the quinoline moiety was then investigated. Both the electron-donating methyl substituent (2n, 2p, 2t, and 2v) and electron-withdrawing halogen substituents (2o and 2r-2s) at different positions all gave the corresponding cyclization products in good overall yields. A free hydroxyl substituent was introduced to different positions on both the benzofuran and quinoline rings, with the aim of increasing the potential biological activities of these molecules [19][20][21]. While the one-pot chlorination/demethylation step was still very efficient, the subsequent base-promoted cyclization for these hydroxyl-substituted substrates generally gave low yields of the corresponding cyclization products. The low yield of the second step may have come from the competitive reactions between the free hydroxyl groups with the base (2e, 2l, 2q, 2u, and 2w-2x), and we speculate that the low yields of the second steps may be improved via selective protection of the hydroxyl group, which is not involved in the cyclization reaction.  1 Reaction conditions: 1 (1 mmol), SOCl2 (4 equiv.), CH2Cl2 (5 mL), 48% HBr (5 mL), 4 (1 equiv.), KOt-Bu (2 equiv.), N,N-dimethylformamide (DMF) (5 mL).

Benzofuro[3,2-c]Quinoline Inhibition Studies
With these newly synthesized benzofuro [3,2-c]quinoline derivatives available, their in vitro antileukemia activities were evaluated against the MV-4-11 cell line, which is one of the acute myelocytic leukemia (AML) cell lines (Table 2) [22,23]. As shown in Table 2, most of these compounds, except for 2h, 2k-2m, and 2v, exhibited good antileukemia activity against MV-4-11 cells. Among them, compound 2e, possessing a free hydroxyl group at the C5 position on the benzofuran ring, showed the highest antileukemia activity (IC 50 = 0.12 µM). When the free hydroxyl group was replaced with Cl or CH 3 , the activities decreased 10-fold (2d and 2g). Compound 2q, bearing a free hydroxyl group at C8 on the quinoline ring, also showed promising antileukemia activity (IC 50 = 0.24 µM). Similarly, when this hydroxyl substituent was replaced with Cl, the activity was reduce 10-fold (2o). On the other hand, replacing this hydroxyl group with methyl resulted in a slightly decreased activity (2p). Methyl, halogen, or hydroxyl substitutions at C3 showed only a slight difference on antileukemia activity (2r-2u). As concluded from these results, compounds with hydroxyl substitution (2e, 2q, and 2u) generally exhibited good activities; however, compounds bearing a free hydroxyl group at C4 of the benzofuran ring (2l and 2x) did not lead to increased activities. It is vital to examine the cytotoxicity on normal cells during the development of an anticancer drug. Accordingly, four compounds that showed high antileukemia activities-2e, 2p, 2q, and 2u-were selected for the selectivity test on peripheral blood mononuclear cells (PBMCs) using an MTS assay ( Table 3). The selectivity indexes (SIs) were counted by dividing the IC 50 values in a PBMC by the IC 50 values in MV-4-11 cell lines [25]. It was obvious that benzofuro[3,2-c]quinolines were less toxic on human peripheral blood mononuclear cell compared to human leukemia MV-4-11 cells. The most active compound, 2e (SI = 79.5), showed a 79.5 times higher selectivity toward cancer cells than toward normal cells. From the MTS assay after 72 h of treatment 1 IC 50 data is an average of at least three independent experiments. 2 SI: selectivity index. 3 >IC 50 value was above the highest concentration used in the assay.

General
All reagents were commercially available and used without purification. The progress of the reactions was monitored using silica gel thin-layer chromatography (TLC). Flash column chromatography was conducted using silica gel (Yantai Kangbinuo, ShanDong, Yantai, China, 200-300). 1 H-and 13 C-NMR spectra (Bruker Ascend, Bremen, Germany) were recorded at 400 and 100 MHz, respectively, using TMS (trimethyl chlorosilane) as the internal standard. Melting points were measured using a WRR-Y drug melting point measurement (Shanghai INESA Physico-Optical Instrument Co. Ltd, Shanghai, China). The high-resolution mass spectra were recorded on a Bruker micOTOF II spectrometer (Electrospray Ionization).

General Procedures for the Synthesis of Benzofuro[3,2-c]Quionlines 2a-2x
To a solution of the 2-(4-chloroquinolin-3-yl)phenol (1 equiv.) in DMF (5 mL), KOt-Bu (2 equiv.) was added in one portion. The reaction mixture was first stirred at 25 • C for 0.5 h, then stirred at 60 • C for 2 h. The mixture was poured into ice water. The organic phase was separated and the water phase was extracted with ethyl acetate three times (3 × 20 mL). The combined organic phases were dried over anhydrous Na 2 SO 4 . Removal of the solvent was done in vacuo, and the crude mixture was further purified using column chromatography over silica gel using petroleum ether/ethyl acetate (15:1) as eluent.  Table S1: Condition optimization for the dimethyl-cyclization (one-pot). Table S2: Condition optimization for the cyclization (stepwise). Table S3 Funding: Financial support from the national Natural Science Foundation of China (21772043) is greatly acknowledged.

Conflicts of Interest:
The authors declare no conflict of interest.