Several modifications of the manganese coordination environment and oxidation states of a family of synthetic dimanganese complexes have been introduced in search of the structural features that promote high rates of hydrogen peroxide dismutation (catalase activity). The X-ray structure of reduced catalase (T, thermophilus) reveals a dimanganese(II,II) site linked by three bridges: mu(13)-glutamate(-), mu-OH-, and mu-OH2. The roles of a bridging hydroxide vs mu-aqua and the carboxylate have been examined in the reduced Mn-2(II,II) complexes, [(L-1,L-2)Mn-2(mu-O2CCH3)(mu-X)](2+) for X- = OH-(7A) or X = H2O (1-4), and their oxidized Mn-2(II,III) analogues, [(L-1,L-2)Mn 2(mu-O)(O2CCH3)(OH)](+) (6) (L-1 is N,N,N',N'-tetrakis(2-methylenebenzamidazolyl)-1,3-diaminopropan-2-ol, and L-2 is the tetrakis-N-ethylated analogue of L-1, which has all amine protons replaced by ethyl groups). The steady-state catalase rate is first-order in concentration of both substrate and reduced catalyst and saturates at high peroxide concentrations in all cases, confirming peroxide/catalyst complex formation. No catalyst decomposition is seen after > 2000 turnovers. Catalysis proceeds via a ping-pong mechanism between the Mn-2(II,II/III,III) redox states, involving complexes 6 and 7A/7A'. The Mn-2(II,IV) oxidation state was not active in catalase activity. Replacement of the mu-aqua bridge by mu-hydroxide eliminates a kinetic lag phase in production of the Oz product, increases the affinity for substrate peroxide in the rate-limiting step as seen by a 5-fold decrease in the Michaelis constant (K-M), and accelerates the maximum rate (k(cat)) by 6.5-fold. The kinetic and spectroscopic data are consistent with substrate deprotonation by the hydroxide bridge, yielding a hydroperoxyl bridge coordinated between the Mn ions ( mu,eta(2) geometry, "end-on") as the basis for catalysis: mu-OH- + H2O2 --> mu-O2H- + H2O. Binding of a second hydroxide ion to 7A causes a further increase in k(cat) by 4-fold with no further change in substrate affinity (K-M). By contrast, free (noncoordinating) bases in solution have no effect on catalysis, thus establishing intramolecular sites for both functional hydroxide anions. Solution structural studies indicate that the presence of 2-5 equiv of hydroxide in solution leads to formation of a bishydroxide species, [(L-1,L-2)Mn-2(mu(13)-O2CCH3)(OH)(2)], which in the presence of air or oxygen auto-oxidizes to yield complex 6, a Mn-2(III,III)(mu-O) species. Complex 6 oxidizes H2O2 to Oz without a kinetic lag phase and is implicated as the active form of the oxidized catalyst. A maximum increase by 240-fold in catalytic efficiency (k(cat)/K-M = 700 s(-1) M-1) is observed with the bishydroxide species versus the aquo complex 1, or only 800-fold less efficient than the enzyme. Deprotonation of the amine groups of the chelate ligand L was shown not to be involved in the hydroxide effects because identical results were obtained using the catalyst with tetrakis(N-ethylated)-L. Uncoupling of the Mn(II) spins by protonation of the alkoxyl bridge (LH) was observed to lower the catalase activity. Comparisons to other dimanganese complexes reveals that the Mn-2(II,II)/Mn-2(III,III) redox potential is not the determining factor in the catalase rate of these complexes. Rather, rate acceleration correlates with the availability of an intramolecular hydroxide for substrate deprotonation and with binding of the substrate at the bridging site between Mn ions in the reductive O-O bond cleavage step that forms water and complex 6.