complex is still unknown, although a number of mechanisms have been proposed (6, 7). Elucidation of the mechanism of action of this enzyme complex requires a detailed description of its catalytic site. This includes identification of the catalytic subunit and characterization of substrate binding sites and essential amino acid residues on this subunit.A large number of studies using different approaches point to the p subunit as the one that contains the catalytic site and is involved in substrate binding (3,4,8). Thus, various studies have shown that the F1 sector has several nucleotide binding sites that reside in its two larger subunits, a and /3(4, 7, 9-11). At least one Pi binding site has also been revealed in the F1 sector (12, 13), and from studies with a Pi analog it seems to be located on the p subunit (14-16). Attempts to identify functional amino acids in the catalytic site have been carried out by using labeled chemical modifiers, known to interact with specific amino acid residues, that bind to the F1 ATPase and inactivate it. After dissociating the labeled enzyme complex to its individual subunits, the label was detected mainly on the p subunit (7,8).A detailed characterization of individual substrate binding sites and their possible identification with the catalytic site in the Fl-enzyme complex is difficult because of the complexity of its structure and function. It contains two or three copies of aB pairs, which could be in different conformational states in the catalytically active complex (17,18), and its activity leads to interconversion of the substrates. These difficulties may be overcome if, instead of testing the whole F1 sector, isolated, purified, and reconstitutively active a and p subunits are used. Such subunits have been obtained so far only from three bacterial sources: a thermophilic bacterium (19), Escherichia coli (20), and Rhodospirillum rubrum (21). They are indeed a much simpler system for study as they show no subunit-subunit interactions and no catalytic activity by themselves, although they can restore ATP synthesis or hydrolysis (or both) to preparations that lack them (19-21).Direct binding studies with labeled substrates have been carried out only with the a subunit of E. coli (22) and the 8 subunit of R. rubrum (23,24,41). These studies have identified one nucleotide binding site on the a subunit (22), whereas the p subunit contained two binding sites for ATP (23) or ADP (24) and one binding site for Pi (41). One of the nucleotide binding sites on the p subunit is a Mg-dependent high-affinity site, which is very similar to the binding site located on the a subunit. The second is a Mg-dependent low-affinity site, which has been suggested to contain also the binding site for Pi, because both nucleotides have been found to be competitive inhibitors of Pi binding (41). Furthermore, Pi seems to bind at the site occupied by the yphosphate group of ATP, since ATP is a much more potent inhibitor of Pi binding to the p subunit than is ADP.In this study we attempt to identify ...