The metal binding sites of isolated F1 ATPase from spinach chloroplasts (CF1) and from the thermophilic bacterium Bacillus PS3 (TF1) have been studied by EPR and pulsed EPR spectroscopy using Mn(II) as a paramagnetic probe. After dialysis in the presence of EDTA, purified CF1 retains 0.14 +/- 0.07 Mg(II) and approximately 0.75 +/- 0.25 ADP. TF1 retains 0.31 +/- 0.03 Mg(II) and 0.08 +/- 0.01 nucleotide (ADP + ATP) after the same treatment. Supplementing known quantities of Mn(II) to metal-depleted CF1 allowed a spectroscopic characterization of the bound Mn(II) cations, for which the EPR spectra at X- and Q-band are reported. The zero field splitting parameters of Mn(II) are derived from the simulation of the EPR signal recorded at Q-band for a sample supplemented with 0.3 Mn/CF1. The values, magnitude of D approximately 200 x 10(-4) cm-1 and magnitude of E approximately 40 x 10(-4) cm-1 suggest that the Mn(II) binds to CF1 in a slightly distorted environment. The ESEEM spectra of complexes of Mn(II) with CF1 were also recorded for different Mn/CF1 ratios. For a complex with 0.8 Mn/CF1, the ESEEM spectrum shows two frequencies at 3.7 and 8.6 MHz that are attributed to the magnetic coupling with 31P with a hyperfine coupling constant of magnitude of A approximately 5.3 MHz, reflecting the interaction with a phosphate group from the endogenous ADP molecule. This demonstrates close proximity of the strong affinity metal site M1 and the endogenous ADP binding site N1, and binding of the ADP beta-phosphate to the divalent metal cation. For Mn(II) complexes with higher Mn/CF1 ratios, new frequency components below approximately 5 MHz are resolved in the spectra in addition to the peaks from 31P. From a comparison of the CF1 spectra and their magnetic field dependence across the Mn(II) EPR line shape with those of Mn(II) complexes with imidazole, glycine, poly-L-lysine, and nucleotide ligands, it is concluded that additional metal binding sites are filled at higher Mn contents and that these involve 14N donors. It is suggested that the most probable set of ligands of the divalent metal(s) for these additional metal sites in CF1 includes a lysine residue, in line with a previous proposal [Houseman, A. L. P., Morgan, L., LoBrutto, R., & Frasch, W. D. (1994) Biochemistry 33, 4910-4917]. Similar experiments for a Mn(II) complex with TF1 (0.4 Mn/TF1) showed no interaction with 31P; instead modulations are detected in the ESEEM below approximately 5 MHz that are attributed to a 14N ligand. This is tentatively attributed to the deprotonated amine of Lys-162 from a beta subunit, on the basis of the structural data available for the mitochondrial F1 complex. Addition of the substrate ATP to this Mn.TF1 complex leads to the formation of a ternary Mn.TF1.ATP complex with coordination of the Mn(II) by a phosphate group from the ATP as judged from the ESEEM results (magnitude of A(31P) approximately 4.5 MHz). An increase in the hyperfine coupling constant of 31P of the phosphate bound to Mn(II) to magnitude of A(31P) approximately 5.1 MHz is observed after incubation of the ternary complex at room temperature. This is interpreted as a significant rearrangement of the coordination sphere of the Mn(II) in the M1 site of the Mn.TF1.ATP complex and may reflect conformational changes of catalytic significance that occur in the nucleotide binding site during unisite hydrolysis of ATP to ADP by this complex.