Membrane-bound ATP synthases (F1F0) catalyze the synthesis of ATP via a rot
ary catalytic mechanism utilizing the energy of an electrochemical ion grad
ient. The transmembrane potential is supposed to propel rotation of a subun
it c ring of F-0 together with subunits gamma and epsilon of F-1, thereby f
orming the rotor part of the enzyme, whereas the remainder of the F1F0 comp
lex functions as a stator for compensation of the torque generated during r
otation. This review focuses on our recent work on the stator part of the F
-o complex, e.g., subunits a and b. Using epitope insertion and antibody bi
nding, subunit a was shown to comprise six transmembrane helixes with both
the N- and C-terminus oriented toward the cytoplasm. By use of circular dic
hroism (CD) spectroscopy, the secondary structure of subunit b incorporated
into proteoliposomes was determined to be 80% alpha -helical together with
14% beta turn conformation, providing flexibility to the second stalk. Rec
onstituted subunit b together with isolated ac subcomplex was shown to be a
ctive in proton translocation and functional F-1 binding revealing the nati
ve conformation of the polypeptide chain. Chemical crosslinking in everted
membrane vesicles led to the formation of subunit b homodimers around resid
ues bQ37 to bL65, whereas bA32C could be crosslinked to subunit a, indicati
ng a close proximity of subunits a and b near the membrane. Further evidenc
e for the proposed direct interaction between subunits a and b was obtained
by purification of a stable ab(2) subcomplex via affinity chromatography u
sing His tags fused to subunit a or b. This ab(2) subcomplex was shown to b
e active in proton translocation and F-1 binding, when coreconstituted with
subunit c. Consequences of crosslink formation and subunit interaction wit
hin the F1F0 complex are discussed.