A chemical study of 4-aminobutyrate aminotransferase active sites

A method for the purification of the enzyme was developed that yielded a highly purified enzyme as determined by the criteria of polyarylamide gel electrophoresis at pH 8.3 and 4.5. The pure enzyme had a specific activity 20 units/mg protein. Its specific activity was unaffected by exogenous PLP. A sedimentation coefficient of 5.43 was determined for the enzyme. An optimum condition for the aminotransferase-succinic semialdehyde dehydrogenase coupled assay system was established at a five fold dehydrogenase excess. It was also established that the two consecutive enzymes had no physical interaction at protein concentrations and experimental conditions of our investigations.

The enzyme contains two binding sites per mole of the dimeric protein. The enzyme as purified contained one molecule of P-pyridoxal and can bind an additional molecule of P-pyridoxal per mole with dissociation constants of 4.2 x 10^-9 M and 3.0 x 10^-6 M respectively. However, the enzyme containing one molecule and that containing two molecules of coenzyme exhibited the same ^k_catand ^k_m for 4-aminobutyrate. The nonequivalent binding and functionality of P-pyridoxal bound to the 111 IV enzyme was then studied. Derivatization of one P-pyridoxal molecule per dimer by reaction with DL-gabaculine or reduction of the enzyme containing one molecule of P-pyridoxal per dimer with NaBH^ did not turn off the catalytic function of the vacant site, since the enzyme containing one molecule of the gabaculine derivatized or reduced P-pyridoxal residue was reconstituted by addition of P-pyridoxal to the vacant site, and the exhibited ^k_cat was the same as the unreacted enzyme.

The resolution of the holoenzyme into the apoprotein and the coenzyme was accomplished. The dissociation of the cofactor from the active site of the aminotransferase brought about a substantial increase in the protein fluorescence yield without any change in the band position of the emission spectra. The fluorescence yield of the enzyme is restored upon binding of P-pyridoxal to the apoprotein.

Coenzyme analogues such as pyridoxal-5-sulfate and pyridoxal reversibly inhibited the restoration of enzymatic activity upon addition of P-pyridoxal to the apoprotein with inhibition constants of 3.0 x 10 ^-8 and 2.1 x 10^-3 respectively. The intrinsic binding energy of the phosphate group of the coenzyme was obtained from a comparison of the free energies of binding of P-pyridoxamine and pyridoxamine, since the difference in the free energies approximates the free binding energy of the phosphate group.

We sought to identify the essential amino acid residues on the protein which are critically connected either with cofactors binding or with enzymatic activity. Two such residues which were identified are cysteine and arginine. The modification of one thiol group on the protein caused 90% loss in enzymatic activity. However, the absorption spectra of the bound P-pyridoxal was not perturbed by the chemical modification, nor was the thiol residue protected against chemical modification by 2-oxoglutarate. We proposed that the modification of the thiol group might have caused conformational changes which influences catalytic function.

Reaction of the enzyme with 2,3-butanedione results in the loss of enzymatic activity. Spectral analysis of the enzyme revealed that the absorption bands attributable to the bound coenzyme has been obliterated, supporting the concept that arginyl residues are involved in the binding of phosphate anions of the coenzyme via electrostatis interactions. Amino acid analysis of the modified protein indicated that four arginyl residues were modified per protein subunit. Amino acid composition of the purified aminotransferase was subsequently established.