This Is AuburnElectronic Theses and Dissertations

Catalase-Peroxidase: A Structure that Facilitates Electron Transfer, Protein-based Cofactor Formation, and Antibiotic Activation




Barton, Callie

Type of Degree

PhD Dissertation


Chemistry and Biochemistry

Restriction Status


Restriction Type

Auburn University Users

Date Available



Catalase-peroxidase (KatG) is a heme-dependent enzyme that uses a single active for the degradation of H2O2 by two distinct mechanisms. One is a catalase mechanism where two equivalents of H2O2 are disproportionated with each catalytic cycle, one being reduced to two H2O and the other being oxidized to O2. The second is a peroxidase mechanism where H2O2 is reduced to two H2O and an exogenous electron donor or peroxidatic electron donor (PxED) is oxidized. There is a broad range of possible PxEDs and which of these are utilized depends on the peroxidase in question as well as the electron donors, that are available in the physiological surroundings of the enzyme. Interestingly, KatG is widely (though not universally) distributed among prokaryotes, including some archaeal species; it is also found in fungi. Interestingly, KatG is very commonly found among pathogenic bacteria and fungi. These include some of the most notorious threats to human health and food security. They include but are not limited to, Mycobacterium tuberculosis, Escherichia coli O157:H7, Yersinia pestis, Acinetobacter baumanii, Aeromonas hydrophila, Magnaporthe grisea and M. oryzae, etc. The enzyme appears to impart an advantage to these organisms in light of the near universal reliance by higher eukaryotes on H2O2 as a defense against pathogen invasion. In the case of M. tuberculosis KatG, it is responsible for the activation of the antitubercular prodrug isoniazid and it is essential for the drug’s antibiotic activity. Accordingly, mutations to the katG gene and the corresponding changes to the KatG enzyme structure are at the heart of 70% of isoniazid resistance. The overall structure of the enzyme as well as specific arrangements of the KatG active site are similar to those of a heme-dependent peroxidase from the peroxidase-catalase enzyme superfamily (formerly known as the plant peroxidase superfamily). Interestingly, it is the only enzyme in the entire superfamily with robust catalase activity—the rest of the superfamily’s members are monofunctional peroxidases. KatG’s unique catalase activity operates by a mechanism which is distinct from all other heme-based catalases, and this mechanism depends on a protein-based cofactor found nowhere else in nature, a Met-Tyr-Trp (MYW) covalent adduct which is post-translationally and autocatalytically generated the active site of KatG. Redox cycling of this MYW cofactor between its fully covalent and free-radical states is integral for robust catalase activity from an otherwise catalase-inactive, heme-peroxidase active site. Interestingly and counterintuitively, KatG’s catalase and peroxidase activities are synergistic with one another. The peroxidase activity serves to prevent the inactivation of KatG’s catalase activity. This function appears to depend on a protein structure, which is uniquely rich in oxidizable amino acids, especially Trp and Tyr. It has been proposed that these residues serve as electron-hole transfer conduits to move misdirected oxidizing equivalents to the enzyme’s surface. Here, PxEDs can reduce and resolve these potentially destructive protein oxidation events without interfering in KatG’s catalase function. Notably, it has also been put forward that the mechanism for antibiotic activation depends on properly directed intraprotein radical transfer reactions. KatG has a structure that is particularly well suited for through-protein electron transfer. For example, the enzyme’s N-terminal domain (which contains the active site and heme cofactor) has four times the expected contribution of Trp residues to its amino acid composition. Structurally, the oxidizable amino acids in the KatG N-terminal domain appear to divide into four sectors or networks, each of which has a Trp residue which is closest point of the network to active site and the heme cofactor. One network is dominated by the MYW cofactor (headed by Trp 107); its primary function is to support catalase activity. Three other networks are each headed by a strictly conserved Trp, including KatG’s so-called proximal tryptophan (W321), as well as W412 and W91. The purpose of the research described in this dissertation is to evaluate the role of these networks of oxidizable amino acids in KatG function within two contexts: 1) KatG prior to the autocatalytic establishment of the MYW cofactor, and 2) the fully functional MYW-bearing KatG enzyme. To the former, does the highly oxidizable framework of the KatG structure facilitate or interfere with formation of the MYW cofactor? To the latter, what is the relative contribution of each sector of KatG’s oxidizable scaffold to the preservation of KatG’s catalase activity? In order to evaluate the autocatalytic formation of the MYW cofactor, we first had to establish an expression and purification protocol that would result in protein without the fully formed MYW cofactor. To form the MYW cofactor, heme and peroxide are needed. Knowing this, we expressed KatG without heme and developed a reconstitution procedure that allowed for the proper insertion of the heme into the active site without forming the MYW cofactor. We called with protein, reconstituted KatG (rKatG). We also expressed and purified the protein in the presence of heme for the purpose of having a protein with the fully formed MYW cofactor. We called this protein mature KatG (mKatG). First, it was important to ensure the heme was in the correct pocket with the correct spin states in rKatG, which was confirmed with UV-Visible spectra and EPR. Second, it was essential to confirm the MYW cofactor had not formed in rKatG and had formed in mKatG, which was done through peptide mapping of tryptic digested protein with LC-MS/MS. Once the protocol was established and the presence or absence of the MYW cofactor had been confirmed, the protocol was used in KatG variants where oxidizable residues had been replaced with non-oxidizable phenylalanine. This allowed for the evaluation of the role of the network of oxidizable amino acids on the facilitation or interference of the autocatalytic formation of the MYW cofactor. It also allowed for the evaluation of the role of the network of oxidizable amino acids on the preservation of the active site through electron transfer. First, the investigation of the formation of the MYW adduct began with the observations of heme transition states in reactions with limited-turnover H2O2 concentrations compared to reactions with H2O2 concentrations supportive of steady-state catalytic turnover. This was followed up with electron paramagnetic resonance (EPR) to track the protein radical formation of the MYW cofactor. Second, the heme intermediates generated by these KatG variants heme were evaluated when reacted with other peroxides, like peracetic acid (PAA), which would allow the enzyme to form initial heme intermediates without allowing the enzyme to perform catalase turnover. Other heme intermediates were observed with reactions with PxEDs like ABTS and ascorbic acid to investigate through-protein electron transfer. Another technique used to investigate the effects of the network of oxidizable amino acids was the extent of O2 production. In this, how much O2 each KatG variant produced at different pH’s, in the presence and absence of PxED, served to connect changes in each variant’s oxidizable amino-acid network to any diminished ability to sustain with O2 production. This experiment was also used for determining if each KatG variant was still active or inactive at the end of the initial reaction with H2O2 indicating the networks ability to prevent inactivation of the enzyme. To this extent we also performed inactivation assays at varying pH’s to even more directly evaluate each KatG variants’ ability to prevent inactivation. In this work we also investigated how electron transfer throughout the enzyme can effect antibiotic inactivation. More specifically, how the prodrug isoniazid’s (INH) most resistant strains that are caused by the S315T mutation to the katG gene effect electron transfer within KatG enzyme. We began this investigation by producing the S315T KatG variant in both the rKatG and mKatG forms. We used stopped-flow spectroscopy to observe the change in heme intermediates when reacted with varying concentrations of H2O2 and other peroxides to investigate this. We also used the extent of O2 production and evaluations of inactivation as different pH’s in the presence and absence of ABTS. Finally, we used EPR to evaluate protein-radicals throughout the reactions with both forms of the enzyme. Knowing how this network of oxidizable amino acids interferes with, or contributes to, the formation of the MYW adduct can elucidate not only mechanistically how the formation occurs but also more specific roles for each of the amino acids investigated. Knowing the pathway of electron transfer throughout the enzyme can help to elucidate how KatG performs its catalase and peroxidase function more efficiently and how KatG prevents itself from becoming inactivated. These studies also revealed how often one of these pathways is used in relation to the other pathways and how that may be a function of distance from the active site, further elucidating how electron transfer can occur within the active site. All of this knowledge, and more specifically the investigation of the S315T mutant can all contribute to the study on antibiotic resistance. Knowing mechanistically how this mutant functions can help to develop future drugs to combat these specific INH resistance strains. This will provide for many future studies in drug development. Future mechanistic studies should be pursued which more extensively characterize transient protein-based radicals through RFQ-EPR. This would include isotopic and other modes of labeling of KatG and its variants, including but not limited to deuterated Tyr, deuterated Trp, Fe57, selenomethionine, etc. High-frequency EPR investigations in combination with the capture of transient intermediates and labeled KatG and labeled KatG variants is anticipated to yield much greater definition of MYW adduct formation events, off-catalase electron transfer events, in addition to those events central to the enzyme’s catalase activity. Future LC and LC-MS/MS studies should be engaged to further characterize the peroxide stoichiometry for MYW, YW, and YW quinoid formation against the backdrop of WT KatG in contrast to variants, which limit alternative pathways for through-protein radical transfer. Other studies to confirm the electronic structure of heme intermediates involved in the MYW adduct formation and the kinetics of each step in the mechanism will also need to be completed. All of these things can be accomplished with a combination of optical stopped-flow, RFQ-EPR, and RFQ-Mössbauer spectroscopies along with LC-MS and LC-MS/MS analyses.