We have found correlations between rate constants and the difference in redox potential of the reactants for electron-transfer reactions between oxidized cytochromes and either photoproduced riboflavin or flavin mononucleotide (FMN) semiquinones (the latter rate constants extrapolated to infinite ionic strength). The riboflavin-cytochrome rate constants are about 70% of those for reduction by lumiflavin, probably because of steric interference by the ribityl side chain. Reduction of cytochromes by FMN semiquinone was ionic strength dependent in all cases, due to electrostatic interactions. Extrapolation of rate constants to infinite ionic strength shows that the phosphate exerts a significant steric effect as well (rate constants average about 27% of those for lumiflavin, although part of this decrease is due to a difference in the semiquinone pK value). Differences in the magnitude of the FMN steric effect correlate well with surface topology differences for those cytochromes whose three-dimensional structures are known. Mitochondrial cytochromes c and the cytochromes c2 all showed attractive (plus-minus) interaction with FMN in spite of the fact that some of these proteins have large net negative charges. Four small c-type cytochromes (including Pseudomonas cytochrome c-551) show a weak repulsive interaction with FMN semiquinone. We conclude that flavosemiquinones interact at a site on the cytochromes that is near the exposed heme edge. There is a large positive electrostatic field at this site in mitochondrial cytochrome c and the cytochromes c2, but this region is primarily hydrophobic in Pseudomonas cytochrome c-551 and in the other small bacterial cytochromes.(ABSTRACT TRUNCATED AT 250 WORDS)

Rate constants for the reduction of approximately 40 electron transfer proteins by photereduced flavins have been determined by laser flash photolysis techniques. The data for a series of 12-homologous cytochromes and 10 homologous high redox potential ferredoxins (HiPIPs) are in excellent agreement with semi-empirical equations relating rate constant and thermodynamic redox potential that have proven applieable to nonbiological electron transfer systems; These correlations allow the establishment of relative reactivities within structurally homolo-.gous classes of biological oxidation-reduction proteins, including cytoehromes and HiPIPs, and a variety of nonhomologous heme-, iron-sulfur-, copper-, and flavin-containing proteins. A qualitative correspondence is shown to exist between such relative reactivity and the extent of solvent exposure of the redox centers in a particular structural class. The implications of these results are considered, and it is concluded that free energy relationships provide a sound basis for systematic analysis of+ reaction mechanisms of electron transfer proteins.Theoretical analyses of electron transfer reactions, based on outer-sphere mechanisms (1, 2) with or without inefficient electron tunneling (nonadiabaticity) (3-6), predict that a relationship should exist between the rate constant for electron transfer and the redox potential difference between donor and acceptor. The most easily used and frequently cited of these theoretical treatments, due to Marcus (1, 2), leads to the following equation for the free energy of activation for electron transfer within an association complex: (0) Several empirical relations, analogous to Eq. 1, have been proposed in an attempt to overcome this discrepancy. The two most successful in. their ability to fit the experimental data are due to Rehm and Weller (9, 18) (Eq. -2), and to Marcus (19) and Agmon and Levine (20) (Eq. 3).[2] AG*t(0)[AGOln 21The parameters in these equations have the same meanings as in Eq. 1. Eqs.-2 and 3 have never been applied to biochemical redox proteins, and thus it is not known whether or not a free energy relationship that correlates rate constants.and redox potentials exists for such systems. It is certainly conceivable that the protein moiety can change the pathway of electron transfer sufficiently to preclude such a relationship. The only redox protein for which data are available that suggests that this type of correlation might occur is flavodoxin (21). The c-type cytochromes, which have been studied extensively, show no correlation between redox potential and rate constant when oxidized or reduced by ionic reactants (22). However, the reactions studied to date are strongly dependent on ionic strength, and thus electrostatics are playing-an important role. In the present work we have examined the. reduction by neutral flavin semiquinones of a wide variety of electron transfer proteins in an attempt to test the applicability-of Eqs. 2 and 3.Two possible mechanisms can be considered for the types of Ab...

We have measured the ionic strength dependence of the rate constants for electron transfer from the semiquinone of Clostridium pasteurianum flavodoxin to 12 c-type cytochromes and several inorganic oxidants using stopped-flow methodology. The experimental data were fit quite well by an electrostatic model that represents the interaction domains as parallel disks with a point charge equal to the charge within this region of the protein. The analysis provides an evaluation of the electrostatic interaction energy and the rate constant at infinite ionic strength (k affinity). The electrostatic charge on the oxidant within the interaction site can be obtained from the electrostatic energy, and for most of those reactants for which structures are available, the results are in good agreement with expectation. The k affinity values were found to correlate with redox potential differences, as expected from the theory of adiabatic (or nonadiabatic) outer-sphere electron-transfer reactions. Deviations from the theoretical curves are interpreted in terms of the influence of surface topology on reaction rate constants. In general, we find that electrostatic effects, steric influences, and redox potential all exert a much larger effect on reaction rate constants for the flavodoxin-cytochrome system than has been previously observed for free flavin-cytochrome interactions. The implications of this for determining biological specificity are discussed.

A "parallel plate" model describing the electrostatic potential energy of protein-protein interactions is presented that provides an analytical representation of the effect of ionic strength on a bimolecular rate constant. The model takes into account the asymmetric distribution of charge on the surface of the protein and localized charges at the site of electron transfer that are modeled as elements of a parallel plate condenser. Both monopolar and dipolar interactions are included. Examples of simple (monophasic) and complex (biphasic) ionic strength dependencies obtained from experiments with several electron transfer protein systems are presented, all of which can be accommodated by the model. The simple cases do not require the use of both monopolar and dipolar terms (i.e., they can be fit well by either alone). The biphasic dependencies can be fit only by using dipolar and monopolar terms of opposite sign, which is physically unreasonable for the molecules considered. Alternatively, the high ionic strength portion of the complex dependencies can be fit using either the monopolar term alone or the complete equation; this assumes a model in which such behavior is a consequence of electron transfer mechanisms involving changes in orientation or site of reaction as the ionic strength is varied. Based on these analyses, we conclude that the principal applications of the model presented here are to provide information about the structural properties of intermediate electron transfer complexes and to quantify comparisons between related proteins or site-specific mutants. We also conclude that the relative contributions of monopolar and dipolar effects to protein electron transfer kinetics cannot be evaluated from experimental data by present approximations.

No abstract