Microwave-Assisted Synthesis of Potential Bioactive Benzo-, Pyrido- or Pyrazino-thieno[3,2-d]pyrimidin-4-amine Analogs of MPC-6827

Efficient microwave-assisted chemical processes were applied to the synthesis of an array of novel N-(4-methoxyphenylamino)-2-methyl benzo-, pyrido- or pyrazino-thieno[3,2-d]pyrimidin-4-amine derivatives. These heteroaromatic systems were envisioned as potent bioisosteric analogues of MPC-6827, an anticancer agent previously developed until phase II clinical studies. A brief evaluation and comparison of their antiproliferative activity on HT-29 and Caco-2, two human colorectal cancer cell lines, were also reported. At the tested concentrations (5 and 10 µM), thieno[3,2-d]pyrimidin-4-amines 4a and 4c exhibited an inhibitory effect similar to MPC-6827 on human colorectal cancer cell proliferation.


Introduction
Due to their presence in numerous biologically active compounds, quinazoline derivatives are of particular interest for medicinal chemistry and remain a major research area in organic chemistry [1][2][3][4][5]. Among the numerous bioactive quinazolines, MPC-6827 (N-(4-methoxyphenylamino)-N,2dimethylquinazoline) has been extensively studied for its therapeutic use against cancer [6,7]. MPC-6827, also named Azixa or verubulin, is a microtubule-destabilizing agent exhibiting a dual mode of action, leading to apoptosis by blocking cell cycle and to growth inhibition on several types of cancer such as breast, colon and ovarian cancers [8][9][10]. MPC-6827 is also known to reduce blood supply to the tumors [11]. Based on these data, this benzo[e]pyrimidine emerged as a good candidate for phase I and phase II clinical trials in patients with metastatic melanoma and glioblastoma multiforme [12][13][14]. Despite these investigations revealing some cardiotoxicity and leading to the suspension of clinical development in phase II [15], MPC-6827 remains an excellent model for the design of potential cytotoxic agents [16].
There are a few synthetic routes of MPC-6827 reported in the literature. The initial work of Sirisoma and his co-workers was carried out in three steps from anthranilic acid methyl ester [9]. In the last step, 4-chloro-2-methylquinazoline was reacted with N-methyl-4-methoxyaniline in anhydrous propanol to give the target product in an overall yield of 55% (Scheme 1) [9]. These methods usually require forcing conditions with long reaction times and, sometimes, conditions using toxic reagents (e.g., POCl 3 ). To develop sustainable and convenient multicomponent processes for the synthesis of quinazoline and quinazolinone derivatives [17][18][19], our group investigated a novel and efficient two-step synthesis of MPC-6827 [20]. Indeed, reaction of anthranilonitrile with N,N-dimethylacetamide dimethyl acetal (DMA-DMA) at 115 • C for 2 min gave acetimidamide intermediate in excellent yield (90%). It was intensely heated (200 • C for 2 h) with N-methyl-p-anisidine (1.5 equiv.), in N-methylpyrrolidone (NMP), in the presence of aluminium chloride (AlCl 3 , 1.5 equiv.). MPC-6827 was obtained in 71% yield, i.e., 64% using the two-step synthesis method ( Figure 1). To develop sustainable and convenient multicomponent processes for the synthesis of quinazoline and quinazolinone derivatives [17][18][19], our group investigated a novel and efficient twostep synthesis of MPC-6827 [20]. Indeed, reaction of anthranilonitrile with N,N-dimethylacetamide dimethyl acetal (DMA-DMA) at 115 °C for 2 min gave acetimidamide intermediate in excellent yield (90%). It was intensely heated (200 °C for 2 h) with N-methyl-p-anisidine (1.5 equiv.), in Nmethylpyrrolidone (NMP), in the presence of aluminium chloride (AlCl3, 1.5 equiv.). MPC-6827 was obtained in 71% yield, i.e., 64% using the two-step synthesis method ( Figure 1). Combining our chemistry work with a scaffold hopping strategy, we envisioned to extend and replace the benzenic part of this small molecule into an aryl thiophene ring. Herein, we report the convenient synthesis of an array of novel N-(4-methoxyphenylamino)-2-methyl benzo-, pyrido-or pyrazino-thieno [3,2-d]pyrimidine derivatives, envisioned as bioisosteric analogs of MPC-6827 ( Figure 2). A brief evaluation of their antiproliferative activity on two human colorectal cancer cell lines (human HT-29 and Caco-2) is also reported and compared with data obtained for the source of inspiration (MPC-6827).  Combining our chemistry work with a scaffold hopping strategy, we envisioned to extend and replace the benzenic part of this small molecule into an aryl thiophene ring. Herein, we report the convenient synthesis of an array of novel N-(4-methoxyphenylamino)-2-methyl benzo-, pyridoor pyrazino-thieno [3,2-d]pyrimidine derivatives, envisioned as bioisosteric analogs of MPC-6827 ( Figure 2). A brief evaluation of their antiproliferative activity on two human colorectal cancer cell lines (human HT-29 and Caco-2) is also reported and compared with data obtained for the source of inspiration (MPC-6827).
Microwave-assisted heating is an efficient technology that allows reproducible and safe operating conditions for convenient access to various molecules when traditional multistep processes would need long reaction times and unstable and toxic reagents (e.g., formamide and POCl 3 ) [21][22][23][24]. The innovative conditions previously described for the synthesis of MPC-6827 failed to provide compounds 4a-d and a more traditional approach was investigated.
The crucial part of our synthetic pathway was the cyclization step in which cyanoenamines 2 were converted into tricyclic compounds 3. Despite our efforts, the synthesis of thienopyridines 3b and 3c remained difficult, as demonstrated by the low yields described in Table 1. The suggested mechanism is described in Scheme 3. Based on the electrophilic character of the cyano group, the reaction starts by nucleophilic attack of the primary aromatic amine on the carbon of the cyano group, which is activated by AlCl 3 [25]. The intermediate p-methoxyphenylamidine can then undergo intramolecular cyclization via the attack of the more nucleophilic amidino secondary amine to enamine function, generating the pyrimidine ring present in the expected N-(4-methoxyphenyl)-2-methyl-benzo [4,5]thieno [3,2-d]pyrimidin-4-amines (3). Interestingly, the release of the dimethylamino group, during the cyclisation step, would favor afterwards the aromatization step.

Antiproliferative Activity on Colorectal Cancer Cell Lines (Caco-2 and HT-29)
In preliminary experiments, the antiproliferative effect of compounds (3b-d) and (4a-d) was evaluated on Caco-2 cells and compared with that of MPC-6827. Each molecule was tested at two concentrations (5 and 10 µM) for 1, 24, 48 and 72 h. The benzo [4,5]thieno[3,2-d]pyrimidine 3a was not tested because of its insolubility in the conditions used. The proliferation of Caco-2 cells was not altered with molecules 3b, 3c, 3d, 4b and 4d (data not shown). The two most interesting compounds (4a and 4c) were also tested on HT-29 cells. Results obtained on Caco-2 and HT-29 cells for 4a, 4c and MPC-6827 are presented in Figures 3 and 4, respectively.  The results of the biological evaluation highlighted that the two compounds (4a and 4c) that possessed antiproliferative properties on Caco-2 and HT-29 were as interesting as MPC-6827 (Azixa or verubulin).
Overall, the number of surviving Caco-2 cancer cells decreased in a time-dependent manner. More precisely, inhibition of the proliferation of Caco-2 colon cancer cells was 25% after 24 h of treatment with 10 µM concentrations of 4c. Higher inhibitory effects measured after 48 h of incubation were close to the final results observed after 72 h of experiment. In this case, proliferation of Caco-2 cells was significantly inhibited in the presence of 4a and 4c, in a manner equivalent to that of MPC-6827, with inhibitory effects of 30% and 45% at 5 µM and 10 µM, respectively (Figure 3).
Similar to Caco-2 cells, HT-29 cancer cell proliferation was significantly inhibited in the presence of MPC-6827 and compounds 4a and 4c, after 72 h of treatment ( Figure 4). The growth inhibition of the HT-29 colorectal cancer cell line also appeared to be time-dependent. In addition, more important effects were observed compared to those described for Caco-2 cells (around 50% and 60% inhibition at 5 µM and 10 µM concentration, respectively) (Figures 3 and 4).

Conclusions
This work demonstrated that the novel thieno[3,2-d]pyrimidin-4-amines 4a and 4c exhibited a similar inhibitory effect on colon cancer cell proliferation as MPC-6827. The exchange of the carbon at position 6 of 4a by a nitrogen atom, as in 4c, appeared to maintain the biological activity studied. Furthermore, a comparison of results described for 4a and 4c with those obtained for 4b and 4d suggested that modifying the atom at position 9 of the heteroaromatic scaffold strongly decreased the biological effect. These preliminary results encourage us to carry on the development of such compounds in the hope of identifying new leads.

General Information
All reagents were purchased from commercial suppliers and were used without further purification. All reactions were monitored by thin-layer chromatography with aluminium plates (0.25 mm) precoated with silica gel 60 F254 (Merck KGaA, Darmstadt, Germany). Visualization was performed with UV light at a wavelength of 254 nm. Purifications were conducted with a flash column chromatography system (PuriFlash, Interchim, Montluçon, France) using stepwise gradients of petroleum ether (PE)/dichloromethane (DCM) or ethyl acetate (EtOAc) as the eluent. Melting points were measured with an SMP3 Melting Point instrument (STUART, Bibby Scientific Ltd., Roissy, France) with a precision of 1.5 • C. IR spectra were recorded with a Spectrum 100 Series FTIR spectrometer (PerkinElmer, Villebon S/Yvette, France). NMR spectra ( 1 H and 13 C) were acquired at 295 K using an AVANCE 300 MHz spectrometer (Bruker, Wissembourg, France) at 300 and 75.4 MHz. Coupling constant J was in Hz and chemical shifts were given in ppm. Mass (ESI, EI and field desorption (FD) were recorded with an LCP 1er XR spectrometer (WATERS, Guyancourt, France). Microwave experiments at atmospheric pressure were carried out in RotoSYNTH (0-1200 W) (Milestone Srl, Italy). Microwave reactions in sealed tubes (10 mL) were performed with an Initiator microwave synthesis instrument (0-400W) (Biotage, Uppsala, Sweden). The percentage of purity of all tested products was more than 95% (determined by HPLC analysis). 1 H-NMR and 13 C-NMR spectra of new compounds are available in Supplementary Materials (Section Figures S1-S12).

General Procedure for the Synthesis of N,N-dimethylacetimidamide Derivatives (2a-d)
A mixture of the appropriate cyanoenamine (3.0 mmol) and DMA-DMA (4 mL, 30 mmol) was heated under microwave irradiation (800 W). On completion, the reaction was cooled to room temperature and crude products were extracted 3 times with EtOAc (5 mL). The organic layers were washed with cold water, dried over Na 2 SO 4 , filtered and evaporated in vacuo. The crude product was purified by silica gel column chromatography using PE/DCM (100:0-0:100, v/v) as the eluent to give the desired products. . On completion, the reaction was cooled to room temperature and water was added. The solid was filtered off, washed twice with water and dried. The crude solid was purified by silica gel column chromatography using PE/EtOAc (100:0-0:100, v/v) as the eluent to give the desired products.
MTT proliferation assay: Caco-2 and HT-29 cells were seeded at a density of 8000 cells/well in 96-well microplates. Cells were treated after 24 h and allowed to proliferate with or without molecules, for 4 days. Molecules were tested at indicated concentrations after extemporaneous dilution in a medium of a starting solution (100 mmol/L in DMSO). The final concentration of DMSO in culture medium was maintained at 0.1%. Molecule 3a was not tested because of its poor solubility in the conditions used. The MTT test was carried out daily after treatment (1,24,48 or 72 h).
For the tested molecules, data of statistical analysis were presented as mean ± SD from at least three independent experiments. At least six different replicates were conducted for each compound. A statistical analysis was performed with non-parametric test (Mann-Whitney U test between two groups) using the SigmaStat software. Differences were considered to be statistically significant at a P value < 0.05. No effect of the solvent was observed. The results obtained for the treated cells were compared with those obtained with cells cultured in the presence of DMSO, which represented 100% of proliferation (control).