Dissociation and Assembly of Pyridine Nucleotide Transhydrogenase from Azotobacter vinelandii

Thermal inactivation studies of pyridine nucleotide transhydrogenase (55°C, pH 7.5, ionic strength 0.1 M) show a biphasic loss of enzyme activity with time. Approximately half of the activity is lost in both the rapid and the slow phase of the inactivation reaction. With electron microscopy it is found that the quaternary structure of transhydrogenase does not change during thermal inactivation. Using Naja naja venom phosphodiesterase to monitor the release of FAD it is shown that dissociation of FAD and activity loss are proportional at every degree of inactivation. The biphasic loss of enzyme activity thus indicates non‐equivalent positions for the subunits ( M r 54000, 1 FAD per subunit) in the transhydrogenase polymer. The process of thermal inactivation is greatly accelerated when the enzyme‐bound flavin is reduced by addition of NADH. The apoenzyme, obtained by gel filtration of transhydrogenase in the presence of NADH at 37°C, can be partly (10–15%) reactivated by addition of excess FAD. The dissociation of FAD under non‐reducing conditions at 55°C is irreversible. In a guanidine hydrochloride denaturing solvent ( c GuHCl > 2 M, 25°C, pH 7.5) transhydrogenase dissociates into subunits. These are highly unfolded as judged by hydrodynamic criteria. Retransfer of the enzyme to conditions stabilizing the native conformation ( c GuHcl < 0.3 M) leads to reactivation provided that FAD is also present in the reactivation mixture. The yield of the reactivation is low, 3–8% of the theoretical maximum, and the process is slow. Delaying the time of addition of FAD to the reactivation mixture causes a decrease in reactivation yield. The results indicate that FAD is inserted in an intermediate stage in the assembly path leading to active transhydrogenase. Polyacrylamide gel electrophoresis in combination with the Western blotting technique shows the presence of a covalent dimer in purified transhydrogenase preparations, which is not dissociated by treatment with dodecylsulphate and 2‐mercaptoethanol. Since a dimer or other form of high relative molecular mass is not detected in cell extracts it follows that such forms do not play a role in the assembly of the complex quaternary structure of the enzyme in vivo.