Preparation of neuroprotective condensed 1,4-benzoxazepines by regio- and diastereoselective domino Knoevenagel–[1,5]-hydride shift cyclization reaction

Summary Condensed O,N-heterocycles containing tetrahydro-1,4-benzoxazepine and tetrahydroquinoline moieties were prepared by a regio- and diastereoselective domino Knoevenagel–[1,5]-hydride shift cyclization reaction of a 4-aryl-2-phenyl-1,4-benzoxazepine derivative obtained from flavanone. The relative configuration of products were determined by the correlation of 3 J H,H coupling data with the geometry of major conformers accessed by DFT conformational analysis. Separated enantiomers of the products were characterized by HPLC-ECD data, which allowed their configurational assignment on the basis of TDDFT-ECD calculation of the solution conformers. Two compounds showed neuroprotective activities against hydrogen peroxide (H2O2) or β-amyloid25–35 (Aβ25–35)-induced cellular injuries in human neuroblastoma SH-SY5Y cells in the range of those of positive controls.

The Schmidt reaction of rac-8 was carried out according to the procedure of Litkey and Patonay [18] affording the racemic 3,4dihydro-2-phenyl-1,4-benzoxazepine-5-one (rac-9) with high regioselectivity, which was reduced to (rac)-10 with LAH in dry THF.In the following step, the N-arylation of rac-10 was performed through a nucleophilic aromatic substitution by using 2-fluoro-5-nitrobenzaldehyde (11b) to give rac-5 containing a tertiary arylamine nitrogen in 71% yield.With the 2-chloro-5nitrobenzaldehyde reagent, only 3% yield for rac-5 could be achieved.Subsequently rac-5 was reacted with 1,3-dimethylbarbituric acid (12), Meldrum's acid (13) and malononitrile (14) all containing active methylene groups, which initiated a domino reaction (Scheme 2).The Knoevenagel reaction of the formyl group in rac-5 and the active methylene groups of 12-14 afforded the intermediates rac-6a-c (Scheme 1), which underwent regioselective [1,5]-hydride shift with participation of the benzylic hydrogen to result in the zwitterionic iminium ion intermediates A. The 6-endo cyclization of this intermediate gave rac-trans-7a,b with high diastereoselectivity, which was governed by the C-2 chirality center of intermediate A. In contrast to the domino reaction with 1,3-dimethylbarbituric acid (12) and Meldrum's acid (13), the reaction with malonitrile (14) stopped at the stage of the Knoevenagel product rac-7c with MgSO 4 in CHCl 3 and further cyclization did not occur when heated in DMSO at 150 °C or refluxed in n-butanol.
The NMR data of 7a,b clearly indicated that they were single diastereomers and the determination of the relative configuration was carried out by the correlation of NMR data and DFT conformational analysis of cisand trans-7a,b (Figure 2 and Figure 3).Although the hydrogens 2-H and 15a-H of cis-7a,b are expected to point toward the same side of the molecules, the analysis of the computed solution conformers of cis-7a,b revealed that the orientation and distance (>4.6 Å) of these methine protons (Figure 3) does not allow for the detection of their NOE contact, rendering the NOE-based assignment of the relative configuration ambiguous.Similarly, no NOE correlation is expected between 2-H and 15a-H of trans-7a,b, which was also evident from the structures of their computed lowest-energy conformers (Figure 2).Accordingly, no characteristic NOE was observed between these methine protons for 7a,b, which did not help for the assignment of the relative configuration.However, in the computed solution conformers of cis-7a,b, the oxazepine ring The chiral HPLC analysis of the products 7a,b also confirmed that the reaction took place diastereoselectively and enantiomers of trans-7a,b were separated on a Chiralpak IA column using hexane/dichloromethane as eluent and online HPLC-ECD spectra of the separated enantiomers were recorded.Due to their similar chromophoric system, the HPLC-ECD spectra of 7a and 7b were quite similar.The first-eluting enantiomer of 7a had an intense broad positive Cotton effect (CE) at 378 nm, negative ones at 314, 305, 272 nm and positive ones at 284, 279 and 224 nm (Figure 4a).
The HPLC-ECD spectrum of the first-eluting enantiomer of 7b showed similar ECD pattern with somewhat different shape and intensities in the 290-240 nm range (Figure 4b).TDDFT-ECD calculation was proved an efficient method to determine the absolute configuration of separated stereoisomers of bioactive synthetic [19] and natural derivatives [20,21] on the basis of their HPLC-ECD spectra.
For the configurational assignment of the separated enantiomers, TDDFT-ECD calculations were carried out on the solution conformers of (2S,15aS)-7d (N-methyl groups were replaced by hydrogens for the calculation) and (2S,15aS)-7b and computed ECD curves were compared with the measured HPLC-ECD ones of the separated enantiomers of trans-7a,b (Figure 4a and Figure 4b).The computed TDDFT-ECD spectra of (2S,15aS)-7d and (2S,15aS)-7b gave good agreement with the experimental HPLC-ECD spectra of the second-eluting enantiomers (Figure 4a,b), which allowed determining the absolute configuration of their second-eluting enantiomers (negative CEs at 370 and 371 nm, respectively) as (2S,15aS).The good agreement between the experimental HPLC-ECD and computed ECD spectra of 7a,b not only allowed the configurational assignment of their separated enantiomers but also confirmed independently the trans-diastereoselectivity of the cyclization.

Experimental
Melting points were determined on a Kofler hot-stage apparatus and are uncorrected.The NMR spectra were recorded on Bruker-AMX 500 ( 1 H: 500 MHz; 13 C: 125 MHz) and Bruker Avance II 400 ( 1 H: 400 MHz; 13 C: 100 MHz) spectrometers using TMS as internal standard.Chemical shifts were reported as δ in ppm and 3 J H,H coupling constants in Hz.Chiral HPLC separation of 7a,b were performed on a Jasco HPLC system with Chiralpak IA column (5 μm, 150 × 4.6 mm, hexane/ dichloromethane 80:20 or 70:30 eluents, 1 mL min −1 flow rate) and HPLC-ECD spectra were recorded in stopped-flow mode on a JASCO J-810 electronic circular dichroism spectropolarimeter equipped with a 10 mm HPLC flow cell.ECD ellipticity (Φ) values were not corrected for concentration.For an HPLC-ECD spectrum, three consecutive scans were recorded and averaged with 2 nm bandwidth, 1 s response, and standard sensitivity.The HPLC-ECD spectrum of the eluent recorded in the same way was used as background.The concentration of the injected sample was set so that the HT value did not exceed 500 V in the HT channel down to 230 nm.IR spectra were recorded on a JASCO FTIR-4100 spectrometer and absorption bands are presented as wavenumber in cm −1 .Electrospay Quadrupole Time-of-Flight HRMS measurements were performed with a MicroTOF-Q type QqTOF MS instrument equipped with an ESI source from Bruker (Bruker Daltoniks, Bremen, Germany).Elementary analysis was carried with a Vario Micro V1.9.6 instrument.
Computational section: Mixed torsional/low mode conformational searches were carried out by means of the Macromodel 9.9.223 software [25] using OPLS-2005 force field with implicit solvent model for chloroform applying a 42 kJ/mol energy window.Geometry reoptimizations (B3LYP/6-31G(d) level in gas phase and B3LYP/TZVP level with PCM solvent model for CHCl 3 ) and TDDFT-ECD calculations were performed with Gaussian 09 [26] using various functionals (B3LYP, BH&HLYP, PBE0) and TZVP basis set for ECD calculations.ECD spectra were generated as the sum of Gaussians [27] with 2100, 2400 and 4200 cm −1 half-height width (corresponding to ca. 19, 22 and 38 nm at 300 nm) using dipolevelocity computed rotational strengths.Boltzmann distributions were estimated from the ZPVE-corrected B3LYP/6-31G(d) energies in the gas-phase calculations and from the B3LYP/ TZVP energies in the PCM ones.The MOLEKEL [28] software package was used for visualization of the results.
Bioassay on neuroprotective activity: SH-SY5Y cells were high passages from the American Type Culture Collection and were maintained at 37 °C in a humidified atmosphere containing 5% CO 2 .Cells were pretreated with compounds for 2 h and then suffered cell injury by treatment with 10 µM Aβ 25-35 or 100 µM H 2 O 2 for another 24 h; to SH-SY5Y cells, pretreated with compounds before exposed to OGD for 1 h, was added 1 mg/mL glucose and 10% serum and cultured for another 24 h under normal condition.SH-SY5Y cells cultured with glucose under normal condition served as control.Cell viability was evaluated by incubating cells with 0.5 mg/mL 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) for 3 h under 5% CO 2 /95% air at 37 °C.After replacement of the medium with 100 µL DMSO, absorbance was read at 490 nm.Data were analyzed by one-way analysis of variance (ANOVA) and expressed as means ± SD with P < 0.05 as significance.

2-Phenyl-3,4-dihydro-1,4-benzoxazepin-5(2H)-one (rac-9):
To a stirred solution of racemic flavanone (rac-8, 10.00 g, 44.63 mmol) in TFA (55 mL), sodium azide (5.80 g, 89.22 mmol) was added in two parts and the mixture was stirred for 6 h.Cold diethyl ether (200 mL) was added to the reaction mixture and stirring was continued for 1 h.The precipitated white solid was filtered and refluxed in 180 mL water for 3 h.After cooling to room temperature, it was refrigerated for 1 h, the resultant precipitate was filtered and washed with cold water.The product rac-9 was dried and isolated as white powder (8.23 g, 77%) with mp 127-129 °C. 1