Chem. Pharm. Bull. 53(10) 1359—1361 (2005)

tinues to foster immense interest in their design and synthesis. DHPs are produced primarily from sugar in vitro. Predictably, the formation of DHPs in vitro also occurs in vivo under non-enzymatic or enzymatic conditions. A number of pyrazine derivatives which are readily created from the DHP ring structure via oxidation have been detected in human urine and foods. Thus, it is thought that DHPs are formed and found in vivo as precursors of pyrazine derivatives. For these reasons, we are interested in the effects of DHPs in vivo. In our previous papers, we discussed the synthesis and DNA strand-breakage activities of some DHPs, such as 2,3dihydro-5,6-dimethylpyrazines (DHP-1, DHP-2 and DHP-3) (Fig. 1). Recently, we also reported that DHPs cause apoptosis and mutagenesis in vivo. In this context, the synthesis of new DHPs provides an interesting challenge. Despite considerable interest in the DHP core among the medicinal and synthetic chemistry communities, we are not aware of any available routes that give the various substituted patterns. Indeed, very few examples of the synthesis of substituted DHPs have been reported. In addition, few reports have been published on the biological and physiological roles of DHPs. In keeping with our interest in the synthetic chemistry and biological activity of DHPs, to investigate the hypothesis that intermediate (exo-type) DHPs might reveal higher DNA strand-breakage activity than other types, we attempted to synthesize new DHPs and to elucidate their biological activity, including DNA strand-breakage activity, in vitro. Preparation of Dihydropyrazines 1—6 The DHPs used in our investigation were synthesized via cyclocondensation of a-diketone and 1,2-diamines, according to our earlier work. All synthesized products were too labile to exist at room temperature, but were fairly stable in the freezer. It is known that DHPs, such as DHP-1, DHP-2 and DHP-3 (Fig. 1), are unstable at room temperature. When an equimolar mixture of 1,2-cyclohexanedione and ethylenediamine (ED) in chloroform was stirred at room temperature for 24 h, the expected hexahydroquinoxaline derivative 1 was obtained in 63% yield (Chart 1). Compound 1 was purified by distillation and solidified in the freezer. The elemental analysis and spectral data of 1 are consistent with the proposed structures (see references and notes 16). Interestingly, the HNMR spectrum of 1 showed the spectral patterns of two isomers, 1A and 1B. In CD3CN at room temperature, compound 1 existed almost exclusively as 1A, with the following ratios observed: 1A : 1B 38 : 1 in CD3CN, 28 : 1 in C5D5N and 3 : 1 in CDCl3. The H-NMR spectrum of 1 in CDCl3 at room temperature shows three signals at d 3.03 assignable to the H-3 methylene protons of 1A, near d 3.3 due to the H-2 and -3 methylene protons of 1B and near d 3.7 due to the H-2 methylene protons of 1A. In addition, the NH and H-8 protons of 1A appear at d 3.40 and 5.15, respectively. The CNMR spectrum of 1 exhibits seven signals at d 40.2 assignable to C-3 of 1A, at d 44.9 due to the C-2 and -3 carbons of 1B, at d 49.5 due to the C-2 carbon of 1A, at d 109.4 due to the C-8 carbon of 1A, at d 134.8 due to the C-8a carbon of 1A, at d 161.1 due to the C-4a and -8a carbons of 1B and at d 162.1 due to the C-4a carbon of 1A. On the basis of these observations, it seems likely that the ratio of the two isomers depends on the solvent. The above spectral studies confirmed the formation of the desired DHP derivative 1, which was characterized as 1,2,3,5,6,7-hexahydroquinoxaline (1A) or 2,3,5,6,7,8-hexahydroquinoxaline (1B). The treatment of 1,2cyclohexanedione (1.0 equiv.) with ED (2.0 equiv.) in chloroform at room temperature for 24 h afforded 4a,8a-butanodecahydropyrano[2,3-b]pyrazine (2) as a pale yellow solid in 54% yield (Chart 1). In H-NMR solvents, e.g. CDCl3 and CD3CN, it was found that compound 2 dissociated into 1B and ED at room temperature. In CD3CN, however, reassociation to 2 was observed at certain temperatures, i.e. 2 : 1B : ED 1 : 2 : 2 at 40 °C. The H-NMR spectrum of 2 October 2005 Notes Chem. Pharm. Bull. 53(10) 1359—1361 (2005) 1359