Intramitochondrial positions of cytochrome haem groups determined by dipolar interactions with paramagnetic cations.

E.p.r.(electron-paramagnetic-resonance) spectra of the ferricytochromes were studied in normal and 'nickel-plated' pigeon heart mitochondria and pigeon heart submitochondrial particles. NiCL2 added to either mitochondria or particles was bound completely to the membranes, but none was transported across the vesicles. Hence, any perturbations of the haem e.p.r. spectra by Ni(II) should occur only for those cytochromes in close proximity to the exterior surface. Whenever Ni(II) can approach to within 1 nm of cytochrome haem. the consequent acceleration of the haem e.p.r. relaxation kinetics should elicit dipolar line broadening. Relaxation acceleration should also increase the incident power level required to saturate the haem e.p.r. signal. In pigeon heart mitochondria, at least three e.p.r. resonances, attributable in part to cytochromes c1, bK and br, are observed at gz=3.3 resonance. In these submitochondrial particles, the peak at gz=3.5 is missing, and the resonance at gz=3.6 resolves into two components, neither of which is sensitive to added Ni(ii). Addition of free haemin (ferric, a paramagnetic anion) to intact mitochondria elicits the same e.p.r. signal changes as does a preparation of submitochondrial particles. Saturation curves for cytochrome oxidase obtained for e.p.r. spectra of the high-spin form (g = 6) and the low-spin form (gz=3.1) also reveal no effect of Ni(II) on the haem e.p.r. relaxation in either mitochondria or inverted submitochondrial particles. Further, Ni(II) fails to alter the spectra or saturation properties of cytochrome c in either mitochondria or submitochondrial particles therefrom. Only with a 50-fold molar excess of Ni(II) can one accelerate the e.p.r. relaxation of cytochrome c in aqueous solution, although other more subtle types of magnetic interactions may occur between the cytochrome and either Ni(II) or ferricyanide. Addition of haemin to mitochondria likewise failed to alter the e.p.r. characteristics of either cytochrome c or cytochrome oxidase. The present observations strongly suggest that cytochromes bK, br and c1 reside on the exterior surface of the inner mitochondrial membrane. On the other hand, we find no positive evidence for the location of cytochrome c or cytochrome oxidase haem groups within 1 nm of either membrane surface. Because of possible shielding effects from the protein moieties, however, we cannot unequivocally assign the location of the haem groups to the membrane interior. The present results are not inconsistent with the observations of other investigators who used different techniques. However, it is clear that any model of energy coupling in mitochondrial oxidative phosphorylation must account for the positioning of all the b-c cytochrome haem groups on the outside.