The adenosine triphosphate (ATP) entrance into the nucleotide-binding subunits of ATP synthases is a puzzle. In the previously determined structure of subunit B mutant R416W of the Methanosarcina mazei Go1 A-ATP synthase one ATP could be trapped at a transition position, close to the phosphate-binding loop. Using defined parameters for co-crystallization of an ATP-bound B-subunit, a unique transition position of ATP could be found in the crystallographic structure of this complex, solved at 3.4 angstrom resolution. The nucleotide is found near the helix-turn-helix motif in the C-terminal domain of the protein: the location occupied by the gamma-subunit to interact with the empty p-subunit in the thermoalkaliphilic Bacillus sp. TA2.A1 of the related F-ATP synthase. When compared with the determined structure of the ATP-transition position, close to the P-loop, and the nucleotide-free form of subunit B, the C-terminal domain of the B mutant is rotated by around 6 degrees, implicating an ATP moving pathway. We propose that, in the nucleotide empty state the central stalk subunit D is in close contact with subunit B and when the ATP molecule enters, D moves slightly, paving way for it to interact with the subunit B, which makes the C-terminal domain rotate by 6 degrees. (C) 2008 Elsevier Inc. All rights reserved.