Abstract
The spectral shift from 420 to 338 nm when pure bacterial D-amino acid transaminase binds D-amino acid substrates is also exhibited in part by high concentrations of L-amino acids (L-alanine and L-glutamate) but not by simple dicarboxylic acids or monoamines. Slow processing of L-alanine to D-alanine was observed both by coupled enzymatic assays using D-amino acid oxidase and by high pressure liquid chromatography analysis employing an optically active chromophore (Marfey's reagent). When the acceptor for L-alanine was α-ketoglutarate, D-glutamate was also formed. This minor activity of the transaminase involved both homologous (L-alanine and D-alanine) and heterologous (L-alanine and D-glutamate) substrate pairs and was a function of the nature of the keto acid acceptor. In the presence of α-ketoisovalerate, DL-alanine was almost completely processed to D-valine; within the limits of the assay no L-valine was detected. With α-ketoisocaproate, 90% of the DL-alanine was converted to D-leucine. In the mechanism of this transaminase reaction, there may be more stereoselective constraints for the protonation of the quinonoid intermediate during the second half-reaction of the transamination reaction, i.e. the donation of the amino group from the pyridoxamine 5′-phosphate coenzyme to a second keto acid acceptor, than during removal of the α proton in the initial steps of the reaction pathway. Thus, with this D-amino acid transaminase, the discrete steps of transamination ensure fidelity of the stereospecificity of reaction pathway.