Role of subunit III in proton translocation

There are three main reasons to suggest a specific function of subunit III in proton translocation. First, Casey et al. [171] showed that modification of this subunit with dicyclohexylcarbodiimide (DCCD) blocks proton translocation, but has little effect on electron transfer. Similar results have been obtained with the reconstituted oxidase from the thermophilic bacterium PS3 [164]. Prochaska et al. [160] showed that DCCD binds mainly to Glu-90 of the bovine subunit III, which is predicted to lie within the membrane domain and hence to be a site analogous to the DCCD binding site in the membranous fj, sector of the ATP-synthase (Fig. 3.8 see also Ref. 85). Since the latter is a part of a proton-conducting channel in ATP synthase, subunit III was thought to have the same function. However, there is one essential difference between the two phenomena. Modification of the membranous glutamic residue in by DCCD leads also to inhibition of ATP hydrolysis in the complex, as expected for two linked reactions. In contrast, DCCD has little or no effect on electron transfer in cytochrome oxidase under conditions where H translocation is abolished. Hence, DCCD cannot simply be judged to block a proton channel in the oxidase. More appropriately, it decouples proton translocation from electron transfer. 

Recently, a third piece of evidence was added by Chan and Freedman [161], who showed that an antibody towards subunit III specifically blocked proton translocation in cytochrome oxidase vesicles. Resting or coupled respiration was stimulated so that the respiratory control index fell by a factor of two. These findings exactly parallel those obtained by removal of subunit III [55,172]. 

The role of subunit III in proton translocation became more enigmatic after the demonstration [163] that the reconstituted cytochrome oxidase from Paracoccus denitrificans (which lacks the equivalent of subunit III [173]) translocates protons, albeit with an apparently lower efficiency. The proton translocation is insensitive to DCCD, which does not bind covalently to this enzyme [174]. 

Two main possibilities are apparent. Subunit III might be an essential functional part of the proton pump, in which case its most essential structures may be expected to be built into the two subunits of Paracoccus. Unfortunately, the primary structures of the latter are not yet known. The absence of DCCD-sensitivity does not discount this possibility. Bacterial mutants are known, for example, in which the H -ATPase functions properly but has lost the sensitivity towards DCCD, although the potential DCCD-binding residue is retained in the primary structure [175]. The 

---