Recent studies on cytochrome oxidase have indicated that the putative "peroxy" intermediatein the catalytic cycle (
PR) is a mixture of intermediates, including
P and
F [Sucheta, A., et al. (1998)
Biochemistry 37, 17905-17914], and the bench-made
P and
F forms appear to have the same redox state(Fe
a34+=O
2-), but a different protonation state [
Fabian, M., and Palmer, G. (2001)
Biochemistry 40, 1867-1874]. To explore the possibility that the putative
PR state is a pH-dependent mixture of intermediates,we investigated the reduction of dioxygen to water by the fully reduced cytochrome oxidase at pH 6.2,7.5, and 8.5 in the visible and Soret regions (350-800 nm) using the CO flow-flash technique. Singularvalue decomposition and global exponential fitting of the time-resolved absorption difference spectraresolved five apparent lifetimes. The fastest three (1.5, 13, and 34
s) were independent of pH, while thetwo slowest rates (80-240
s and 1.1-2.4 ms) decreased by a factor of 2-3 as the pH increased. Whenthe time-resolved spectra were analyzed using a unidirectional sequential model, the spectra of the reducedenzyme and the dioxygen-bound intermediate, compound
A, were found to be pH-independent. However,the putative
PR intermediate was best represented by a pH-dependent mixture of compound
A,
P, and
F.The ferryl form was favored at low pH. The subsequent intermediate is a ferryl with a pH-dependentelectron transfer equilibrium between heme
a and Cu
A, the reduced heme
a being favored at low pH.These results suggest a pH-dependent reaction mechanism of the reduction of dioxygen to water by thefully reduced enzyme that is more complex than previously proposed.