Taken together, these observations suggested that the amino-terminal domain is

predominantly involved in the binding of the 6 subunit to Fl and the carboxyl-terminal

domain is involved in binding to Fo.

 The location of the 6 subunit has had a history of being very controversial. Prior to

the high resolution structure obtained by Abrahams et al. (1994), the b and 6 subunits

were expected to form part of the central stalk of the F1Fo ATP synthase enzyme, which

is now known to consist of the y and a subunits (20). Due to the dimensions, it seemed

unlikely that the b and y subunits could fit as part of the central stalk, which implied that

they must form a separate connection between Fl and Fo. Improving EM technology did

not allow visualization of the second stalk structure at the periphery of the F1Fo ATP

synthase complex until many years later (27, 28). Prior to visualization by EM, several

early crosslinking studies had been reported in the quest to find the location of the

6 subunit binding on Fl, finding it to be on the a subunit (89, 181-184). Notably,

crosslinking the 6 subunit to the a subunit did not have a great impact on F1Fo ATP

synthase function, as would be expected if the 6 subunit formed part of the stator stalk

(185). High-resolution structure of the Fl u3 3 hexamer with a partial structure of the y

subunit had revealed a dimple in the top of the hexamer approximately 15 A+ deep that

was adj acent to the core space where the amino- and carboxyl-terminal a-helices of the y

subunit resided (20). EM studies had revealed a "cap" structure at the very top of F1 in

both E. coli and mitochondrial complexes (27, 29, 186). It was thus believed that the 6

subunit resided in the dimple of F1 as the cap seen in the EM structures (187). This

possibility was refuted when Prescott et al. demonstrated the ability to stably incorporate

the green fluorescent protein (GFP), via varying length peptide linkers (0, 4 or 27 amino