Redox dependent interactions of the metal sites in carbon monoxide-bound cytochrome c oxidase monitored by infrared and UV/visible spectroelectrochemical methods.

Spectroelectrochemical titration studies involving the binding of the infrared-active probe ligand carbon monoxide (CO) to the heme alpha 3/CuB site of bovine heart cytochrome c oxidase (CcO) have been reexamined. The spectroelectrochemical cell employed was constructed to monitor both the infrared (IR) and visible/Soret spectra of the CcO-CO complex as a function of the overall oxidation state of the enzyme. A number of commonly used electron transfer mediators were employed to shuttle electrons between the redox active sites within the enzyme and the electrode surface. The well-documented shift in the CO infrared stretch band maximum from 1963.3 cm-1 (CcO fully reduced) to 1965.5 cm-1 (CcO partially oxidized) was carefully titrated electrochemically. Deconvolution of the asymmetric CO stretches indicates the existence of two different states of CO vibrators within the enzyme, presumably due to two conformers which are present in a ratio of approximately 5:1. Upon incrementally stepping the potential from the fully reduced state to the partially oxidized state, we found it possible to follow the decrease in the intensity of the original pair of these conformers and the concomitant increase of a resultant pair while maintaining this 5:1 ratio between the conformers. By plotting the change in the deconvoluted CO peak intensities vs the redox potential, as well as the absorbance changes in the visible/Soret spectra vs the redox potential, we found not only that both fit an n = 1 electron process but also that the spectral changes tracked each other identically with experimental error. Furthermore, analysis of the second derivative of the Soret spectra allowed for the qualitative monitoring of the oxidation state of the Fe alpha site which again tracked identically to that of the CO shift in the IR region. These results would seem to confirm earlier suggestions that perturbing the oxidation state of Fe alpha causes a conformational change in the enzyme which affects the binding site for CO, namely heme alpha 3. As a consequence of the CO IR stretching frequencies changing by only 2 cm-1 during this redox titration, with no accompanying changes in half band width, we suggest that it is impossible that this small but significant change seen in the CO stretching frequencies could be due to an oxidation state change in CuB, given the known sensitivity of the CO stretching frequency to perturbations and the close proximity of Cu(B) to the CO binding site at heme alpha 3 (4.5 A). Therefore, it would appear that Cu(B) must remain reduced as long as CO is bound to the heme alpha 3 site. This is consistent with earlier proposals that Fe alpha 3 and Cu(B) are acting together as a two-electron donor to dioxygen.