Large-scale purification and characterization of the five subunits of F1-ATPase from pig heart mitochondria

The nucleotide-independent Fe(III)-binding site is located on β subunit of the mitochondrial F1-ATPase

Upon separation of a crude preparation of beta subunit ("beta fraction") from mitochondrial F(1)-ATPase containing one equivalent of Fe(III) in the nucleotide-independent site (1Fe(III)-loaded MF(1)), Fe(III) is almost completely recovered. CD spectra show that "beta fraction" maintains the structural changes induced by Fe(III) in the whole enzyme. In accordance, EPR reveals that the Fe(III) site geometry is conserved in "beta fraction." Moreover, the EPR spectra of 1Fe(III)-loaded MF(1) and its "beta fraction" undergo similar changes of the line-shape upon Pi binding at the catalytic site, indicating that the Pi and Fe(III) are proximal on beta. Highly purified beta in nucleotide-free form binds 1mol of Fe(III)/mol of protein. MF(1) "freezed" by inhibitors with two beta in closed conformation and one beta in open or half-closed conformation binds 1mol of Fe(III)/mol of enzyme. Therefore, the Fe(III) site location in the unique beta subunit not adopting the closed conformation is proposed.

Docking the mitochondrial inhibitor protein IF1 to a membrane receptor different from the F1-ATPase beta subunit

Monoclonal antibodies reacting with the inhibitor protein (IF1) of the mitochondrial ATPase/ATP synthase complex did not modify the IF1-induced inhibition of soluble F1 ATPase activity. On the contrary, they increased the ATPase activity of inverted electron-transport particles without inducing a significant release of IF1 from these particles. This suggested that IF1 could be linked to a membrane protein when it was not inhibiting the ATPase activity. IF1 antibodies have been used to show that IF1 can bind not only to the beta subunit of F1-ATPase [Klein, G., Satre, M., Dianoux, A. C. & Vignais, P. V. (1981) Biochemistry 20, 1339-1344] but also to a protein present in the inner-mitochondrial membrane. The cross-linking of IF1 to this membrane protein gave a product of M(r) 15000-16000 that migrated differently from IF1 and from the dimer of IF1 using SDS/PAGE. When the cross-linked product was obtained by using a cleavable cross-linking reagent, the complex between IF1 and the docking protein was partly dissociated and free IF1 was recovered. Considering the molecular mass of IF1, this docking protein for IF1 has apparent M(r) 5000-6000. The complex between IF1 and this receptor protein can be detected in low amounts by antibodies against IF1 in the absence of cross-linking reagent. Since this complex remained in the pellet after treatment of the membrane with Triton X-100, it should be a membrane protein. Therefore, IF1 can bind not only to its inhibitory-binding site at the beta subunit of F1, but also to a non inhibitory site which is a membrane protein of approximate M(r) 5000-6000.

Tryptophan phosphorescence as a structural probe of mitochondrial F1-ATPase epsilon-subunit

We report the detection of tryptophan phosphorescence emission from the sole residue in the epsilon-subunit of the bovine heart mitochondrial F1-ATPase complex. The phosphorescence spectrum, intensity and decay kinetics have been measured over the temperature range 160-273 K. The fine structure in the phosphorescence spectrum at low temperature, with the 0-0 vibrational band centered at 411 nm, reveals the hydrophobic nature of the chromophore's environment. Both the large width of the 0-0 vibrational band and the heterogeneous decay kinetics in fluid solution emphasize the existence of multiple conformations of the epsilon-subunit, structures which are rather stable as they do not interconvert in the millisecond time scale. Further, from the relatively long triplet lifetime at 273 K, it is possible to infer the existence of a tight, rigid core in the structure of the epsilon-subunit. Under subunit-dissociating conditions (6 M urea), the spectrum at 160 K undergoes a slight blue shift but since the phosphorescence lifetime, at all temperatures, is similar or longer than in the absence of dissociant, we conclude that dissociation does not lead to solvent exposure of the tryptophanyl side-chain. This conclusion is supported by the results obtained at 273 K by dissociating F1 in the presence of 0.3 M guanidine hydrochloride. Phosphorescence lifetimes indicate that 6 M urea leads to a more compact structure of the epsilon-subunit, whereas the opposite occurs when Mg-ATP is added to nucleotide-depleted F1. These spectroscopic changes establish unequivocally that the binding of the adenine nucleotide to the enzyme is accompanied by conformational changes involving the epsilon-subunit.