Further comparison of ubiquinol and cytochrome c terminal oxidases

Abstract

In the preceeding commentary, Haltia reflects on our view presented in [l] that the ubiquinol and cytochrome c terminal oxidases are likely to utilize different electron transfer and proton translocation mechanisms. We would like to take this opportunity to clarify our position on some of the issues that he discusses. Haltia compares the structure and function of subunit II of the two families of terminal oxidases, the ubiquinol oxidases and the cytochrome c oxidases. While structure prediction in general is hardly a mature field, hydrophobicity calculations are relatively accurate for predicting the number and approximate location of transmembrane helices in a membrane protein. However, such predictions are inaccurate at times [2,3] and, moreover, say virtually nothing about the extramembraneous domains. To predict that two polypeptides have a similar structure on the basis of hydrophobicity calculations is pushing the technique beyond the limits of its capabilities. As pointed out in our minireview [l], the Cu, site and part or all of the residues responsible for the binding of cytochrome c are thought to be located in the extra-membraneous domain of subunit II of the cytochrome c oxidases. In contrast, due to the extreme hydrophobic nature of the physiological substrate ubiquinol-8, the ubiquinol binding site(s) is (are) expected to exist in the transmembrane region of the ubiquinol oxidase complexes as is analogously observed in the structures of bacterial photosynthetic reaction centers. If a ubiquinol binding domain exists on subunit II of the cytochrome bo, complex, as has been suggested by a number of studies [4,5], the substrate oxidizing domains of the two families of terminal oxidases would exist in different three-dimensional locations and protein environments of subunit II. Thus, the subunit II’s of the two families of oxidases could have either very different folding patterns to accommodate the different substrates or they could have very similar overall folding patterns for evolutionary reasons. We note that the 10% sequence identity of subunit II of the Escherichia coli cytochrome bo, complex and the Paracoccus denitrz@ans cytochrome au, complex is well below what Doolittle terms the ‘twilight zone’ (15-25% sequence identity), and accordingly, these polypeptides are unlikely to be evolutionarily related based on sequence comparison alone [6].