Stereochemical aspects in the synthesis of novel N-(purin-6-yl)dipeptides as potential antimycobacterial agents.

The synthesis of purine conjugates with natural amino acids is one of the promising directions in search for novel therapeutic agents, including antimycobacterial agents. The purpose of this study was to synthesize N-(purin-6-yl)dipeptides containing the terminal fragment of (S)-glutamic acid. To obtain the target compounds, two synthetic routes were tested. The first of them is based on coupling of N-(purin-6-yl)-(S)-amino acids to dimethyl (S)-glutamate in the presence of carbodiimide coupling agent followed by the removal of ester groups. However, it turned out that this coupling process was accompanied by racemization of the chiral center of N-(purin-6-yl)-α-amino acids and in all cases led to mixtures of (S,S)- and (R,S)-diastereomers (6:4). Individual (S,S)-diastereomers were obtained using an alternative approach based on the nucleophilic substitution of chlorine in 6-chloropurine or 2-amino-6-chloropurine with corresponding dipeptides as nucleophiles. The enantiomeric purity of the target compounds was confirmed by chiral HPLC. To test the assumption that racemization of the chiral center of N-(purin-6-yl)-α-amino acids occurs with the participation of nitrogen atoms of the imidazole ring via the stage of formation of a chirally labile intermediate, we obtained such structural analogs of N-(purin-6-yl)-(S)-alanine as N-(9-benzylpurin-6-yl)-(S)-alanine and N-(7-deazapurin-6-yl)-(S)-alanine. It was found that coupling of these compounds to dimethyl (S)-glutamate was also accompanied by racemization. This indicates that the imidazole fragment does not play a crucial role in this process. When testing the antimycobacterial activity of some of the obtained compounds, conjugates with moderate activity against the laboratory Mycobacterium tuberculosis H37Rv strain (MIC 3.1-6.25 μg/mL) were identified.