Resonance Raman spectra of native, overexpressed M. tuberculosis catalase-peroxidase (KatG), the enzyme responsible for activation of the antituberculosis antibiotic isoniazid (isonicotinic acid hydrazide), have confirmed that the heme iron in the resting (ferric) enzyme is high-spin five-coordinate. Difference Raman spectra did not reveal a change in coordination number upon binding of isoniazid to KatG. Stopped-flow spectrophotometric studies of the reaction of KatG with stoichiometric equivalents or small excesses of hydrogen peroxide revealed only the optical spectrum of the ferric enzyme with no hypervalent iron intermediates detected. Large excesses of hydrogen peroxide generated oxyferrous KatG, which was unstable and rapidly decayed to the ferric enzyme. Formation of a pseudo-stable intermediate sharing optical characteristics with the porphyrin pi-cation radical-ferryl iron species (Compound I) of horseradish peroxidase was observed upon reaction of KatG with excess 3-chloroperoxybenzoic acid, peroxyacetic acid, or tert-butylhydroperoxide (apparent second-order rate constants of 3.1 x 10(4), 1.2 x 10(4), and 25 M(-1) s(-1), respectively). Identification of the intermediate as KatG Compound I was confirmed using low-temperature electron paramagnetic resonance spectroscopy. Isoniazid, as well as ascorbate and potassium ferrocyanide, reduced KatG Compound I to the ferric enzyme without detectable formation of Compound II in stopped-flow measurements. This result differed from the reaction of horseradish peroxidase Compound I with isoniazid, during which Compound II was stably generated. These results demonstrate important mechanistic differences between a bacterial catalase-peroxidase and the homologous plant peroxidases and yeast cytochrome c peroxidase, in its reactions with peroxides as well as substrates.

Stable films of dimyristoylphosphatidylcholine and M. tuberculosis catalase-peroxidase (KatG), several peroxidases, myoglobin, and catalase showed reversible FeIII/FeII voltammetry on pyrolytic graphite electrodes and catalytic current for hydrogen peroxide and oxygen. Amperometric responses for these films to H2O2 at 0 V are likely to contain significant contributions from catalytic reduction of oxygen produced during the catalytic cycles. Relative apparent turnover rates at pH 6 based on steady-state currents at 0 V versus SCE in the presence of H2O2 were in the order horseradish peroxidase > cytochrome c peroxidase (CcP) > soybean peroxidase > myoglobin > KatG > catalase. Lower currents for the very efficient peroxide scavengers KatG and catalase may be related to the instability of their compounds I in the presence of H2O2. KatG catalyzed the electrochemical reduction of oxygen more efficiently than catalase and CcP but less efficiently than the other peroxidases. DMPC films incorporating glucose oxidase and peroxidases gave good analytical responses to glucose, demonstrating the feasibility of dual enzyme-lipid films for biosensor fabrication.