Modulation of the GTPase activity of transducin. Kinetic studies of reconstituted systems.

We seek to define the influence of retinal cGMP phosphodiesterase (PDE) on the GTPase activity of transducin (T). A novel stopped-flow/fast filtration apparatus [Antonny, B., et al. (1993) Biochemistry 32, 8646-8653] is used to deliver T alpha GTP free of rod outer segment (ROS) membranes to a suspension of phospholipid vesicles bearing holoPDE. As measured by a pH electrode, the decay of cGMP hydrolysis from these samples, which contain no other proteins but T alpha and holoPDE, requires GTP hydrolysis and occurs in 40 s. The addition of T beta gamma to the vesicles does not accelerate this deactivation. When ROS membranes are urea-stripped, reconstituted with transducin + holoPDE, and illuminated, the injection of an amount of GTP that is substoichiometric to holoPDE gives a cGMP hydrolysis pulse that lasts for 30 s. However, the same reconstitution performed with ROS stripped by extensive dilution in isotonic buffer results in a deactivation time of only 8 s, which resembles the 7 s observed with native ROSs. With these isotonically stripped ROSs, when GTP injection comes after a first injection with GTP gamma S, the cGMP hydrolysis pulse is lengthened and lasts for 17 s; with urea-washed ROS, no such lengthening is observed. These results clearly demonstrate that holoPDE by itself cannot enhance the GTPase activity of transducin, even when the two proteins are localized on a membrane surface. Instead, they point to the existence of a membrane-bound, urea-sensitive protein factor that activates the GTPase of T alpha in the transducin-holoPDE complex.