A manganese-dependent 3,4-dihydroxyphenylacetate 2,3-dioxygenase from Arthrobacter globiformis strain CM-2 (MndD) cloned in Escherichia coli has been purified to homogeneity. Sedimentation equilibrium analysis indicates an α4 homotetrameric holoenzyme structure (4 x 38 861 Da). Steady-state kinetic analysis of MndD with a variety of substrates and inhibitors yields very similar relative rates to the known Fe(II)- and Mn(II)-dependent 3,4- dihydroxyphenylacetate 2,3-dioxygenases from Pseudomonas ovalis and Bacillus brevis, respectively. Yet, unlike the Fe(II)-dependent enzyme, MndD retains almost all activity in the presence of H2O2 and CN- and is inactivated by Fe(II). ICP emission analysis confirms the presence of 3.0 ± 0.2 g-atoms Mn (and only 0.7 ± 0.2 g-atoms Fe) per tetrameric holoenzyme molecule. Comparison of MndD samples with varying metal content, including an apo and partial-apo enzyme preparation, shows u strong positive correlation between specific activity and Mn content. EPR spectra of MndD as isolated exhibit u nearly isotropic g = 2.0 signal having 6-fold hyperfine splitting (A = 95 G) typical of octahedrally coordinated Mn(II) in a protein. Quantitation of the EPR spin yields 3.4 ± 0.3 g-atoms of Mn(II) per holoenzyme. When exposed anaerobically to its natural substrate, 3,4-dihydroxyphenylacetate (3,4- DHPA), the EPR spectrum undergoes a dramatic change characterized by the attenuation of the g = 2 signal and the appearance of new signals at g = 1.2, 2.9, 4.3, and 16. The g = 4.3 signal displays 6-fold hyperfine splitting (A = 95 G) that unambiguously assigns it to the Mn(II) center. The appearance of these new signals indicates u large increase in zero-field splitting suggestive of a change in ligand coordination to the Mn(II) center. Similarly perturbed signals are seen in the EPR spectra of MndD complexed with the comparably active substrate analog, D,L-3,4-dihydroxymandelate, or the tight- binding inhibitor, p-nitrocatechol, but not in the complexes with weaker binding substrates and inhibitors. The fact that only strong-binding substrates and inhibitors significantly perturb the Mn(II) EPR signal strongly suggests that the substrate coordinates to the Mn(II) center in the catalytic pathway.