The trapping of different conformations of the Escherichia coli F₁ ATPase by disulfide bond formation: Effect on nucleotide binding affinities of the catalytic sites

Two mutants of the Escherichia coli F₁ ATPase, βY331W:E381C/εS108C and αS411C/βY331W/εS108C, have been used to relate nucleotide binding in catalytic sites with different interactions of the stalk-forming subunits γ and ε at the α₃β₃subunit domain. Essentially full yield cross-linking between β+ γ and β+ ε, or between α + γ and α + ε, was obtained in these mutants by Cu²⁺-induced disulfide bond formation, thereby trapping the enzyme in states with the small subunits interacting either with β or α subunits. The presence of the Trp for β Tyr-331 in both mutants allowed direct measurement of nucleotide occupancy of catalytic sites. Before cross-linking, Mg²⁺ATP could be bound in all three catalytic sites in both mutants with a K_d of around 0.1 μM for the highest affinity site and K_d values of approximately 2 μM and 30-40 μM for the second and third sites, respectively. In the absence of Mg²⁺, ATP also bound in all three catalytic sites but with a single low affinity (above 100 μM) in both mutants. Cu²⁺-induced cross-linking of ECF₁ from the mutant βY331W:E381C/ εS108C had very little effect on nucleotide binding. The binding affinities of the three catalytic sites for Mg²⁺ATP were not significantly altered from those obtained before cross-linking, and the enzyme still switched between cooperative binding and equal binding affinities of the three catalytic sites (when Mg²⁺was absent). When the γ and ε subunits were cross-linked to α subunits, ATP binding in the highest affinity catalytic site was dramatically altered. This site became closed so that nucleotide (ATP or ADP) that had been bound into it prior to cross-linking was trapped and could not exchange out. Also, ATP or ADP could not enter this site, although empty, once the enzyme had been cross-linked. Finally, cross-linking of the γ and ε to the α subunits prevented the switching between cooperative binding and the state where the three catalytic sites are equivalent. We argue that the conformation of the enzyme in which the small subunits are at α subunits occurs during functioning of the enzyme in the course of the rotation of γ and ε subunits within the α₃β hexamer and that this may be the activated state for ATP synthesis.