A select group of herbicides that inhibit photosystem II also act at the acceptor side of the reaction center (RC) from the photosynthetic bacterium Rhodopseudomonas sphaeroides, with much the same relative specificity as in plants. These include the triazines and some phenolic compounds. The proposal that herbicides inhibit the electron transfer from the primary quinone (QA) to the secondary quinone (QB) by competing for the secondary quinone binding site--the B-site--[5], is tested here with terbutryn, the most potent of the triazines. Competition between terbutryn and ubiquinone (Q-10) was observed using the kinetics of the back-reaction as a measure of inhibition. The model includes binding equilibria before and after flash activation. The binding constants for the preflash (dark) equilibria, for reaction centers in 0.14% lauryl dimethylamine-N-oxide (LDAO), were KDi = 0.8 microM terbutryn, KDq = 2 microM Q-10; both are detergent-concentration dependent. After flash activation, binding equilibrium is not fully restored on the time scale of the back-reaction because terbutryn unbinds slowly. This gives rise to biphasic decay kinetics from which koff for terbutryn was estimated to be 3 sec-1. Titrations of the rate of the slow back reaction indicated that the post-flash equilibrium is less sensitive to inhibitor, in a manner that is independent of the much stronger binding of the semiquinone, Q-B, and indicative of a direct effect of the redox state of QA on the affinity of the B-site for ligands. However, the effects on KLi and KDq could not be separated: either KLi greater than KDi or KLq less than KDq. Some triazine-resistant mutants have been isolated and are described. All appear to be herbicide binding site mutants. Whole cells and photosynthetic membrane vesicles (chromatophores) exhibit a 10-50-fold increase in resistance to triazines due, in large part, to an increase in the rate of unbinding (koff). The modifications of the binding site appear to diminish the affinity of the B-site for ubiquinone as well as terbutryn. It is concluded that bacterial RCs are a useful model for the study of herbicide activity and specificity.
