Electrostatic and hydrophobic interactions during complex formation and electron transfer in the ferredoxin/ferredoxin:NADP(+) reductase system from Anabaena

Transient kinetics and protein-protein binding measurements over a wide range of ionic strength (I) have been used to characterize the interactions occurring during complex formation and electron transfer (et) between recombinant ferredoxin (Fd) and both native and recombinant ferredoxin:NADP(+) reductase (FNR) from the cyanobacterium Anabaena. Between I = 12 mM and I = 100 mM, the dissociation constant (K-d) for the complex formed between oxidized Fd and oxidized native FNR increases from 4.5 to 8.1 mu M, whereas K-d for the d complex with recombinant FNR increases from 0.3 to 3.3 mu M. For both pairs of proteins, the ionic strength dependence of k(obs) for the et reaction is biphasic. With native FNR, k(obs), increases only slightly between I = 12 mM and I = 100 mM, whereas for recombinant FNR k(obs) increases by about 4-fold over this ionic strength range. For both proteins, k(obs) decreases monotonically above I = 100 mM. The dependence of k(obs) on FNR concentration is linear for both pairs of proteins at I = 12 mM, with the second-order rate constant for recombinant FNR being about 3 times smaller than that for the native protein. In contrast, at I = 100 mM the k(obs) values are the same for both protein pairs, and show saturation behavior with respect to the FNR concentration, indicating that et becomes rate-limiting at high FNR concentrations. Electrostatic analysis of the kinetic data above I = 100 mM allows a prediction of the ionic strength dependence of the K-d values, if electrostatic interactions are the only determinant of complex stability. The predicted dependence is dramatically larger than the observed one, indicating that hydrophobic interactions make an important contribution to complex stability. The differences in binding between native and recombinant FNR are ascribed to proteolytic cleavage at the N-terminus, which occurs during preparation of the native enzyme and which removes two positively charged residues, thereby decreasing the electrostatic interactions with Fd. The kinetic results are explained by assuming that formation of the oxidized protein-protein complex blocks the et site, and thus reaction only occurs between reduced Fd and free FNR. However, even after correction for the presence of the preexisting complex, the reactivity of FNR at I = 12 mM is significantly lower than that at I = 100 mM. This is ascribed to electrostatic effects which force the complex with reduced Fd to be less optimal, implying that hydrophobic interactions favor a more productive interaction between the two redox proteins.