This Is AuburnElectronic Theses and Dissertations

How an Arginine Switch Promotes the Self-preservation of an Hydrogen Peroxide-degrading Enzyme KatG: the Strategic Use of an Active-Site Tryptophan




Xu, Hui

Type of Degree

PhD Dissertation


Chemistry and Biochemistry


Catalase-peroxidase (KatG) is a heme-dependent enzyme and an effective H2O2 disposal catalyst, sustaining life for many archaea, bacteria, and lower eukaryotes. Many pathogens employ KatG catalase activity (2 H2O2 -> 2 H2O + O2) to manage oxidative stress. Notably, the intracellular human pathogen Mycobacterium tuberculosis survives within the harsh acidic and highly oxidizing phagolysosomes of host neutrophils and macrophages with KatG as one of its two catalase-active enzymes and research suggests M. tuberculosis with katG gene knocked out is more susceptible to oxidative stress induced by host cells. With a unique active site incorporated into a peroxidase fold, KatG has robust catalase activity that no other member of its superfamily possesses. KatG carries out its catalase activity by a novel radical-based mechanism involving a protein-derived cofactor, the Met-Tyr-Trp (MYW) adduct. In connection with this mechanism, KatG also undergoes inactivation resulting from off-catalase radical transfer events. A central participant in this process is the active site’s proximal tryptophan, (Trp 321 according to M. tuberculosis KatG (MtKatG) numbering). Interestingly, KatG also has an arginine “switch” (Arg 418 in MtKatG), the conformational position of which is pH dependent. Above pH 6.5, its side chain points predominantly toward the phenolic oxygen of the MYW tyrosine (i.e., “in” conformation), and below pH 6.5 it points away from the MYW cofactor and toward the protein surface (i.e., “out” conformation). It has been observed that susceptibility of KatG catalase to inactivation increases below pH 6.5. Prior to the research described in this dissertation, any connection between inactivation and arginine switch conformation was unknown and uninvestigated. Similarly, the strict conservation of the proximal Trp across all KatGs even though it appears to contribute to KatG inactivation has raised questions about its role that have remained unexplored. To investigate the connection between the arginine switch, off-pathway radical transfer, and KatG inactivation, as well as to shed light on the role of Trp 321 oxidation in enzyme inactivation, we evaluated four KatG variants: 1) The fully functional wild-type KatG, 2) enzyme missing a functional arginine switch (R418N KatG), 3) enzyme lacking an oxidizable proximal Trp (W321F KatG), and enzyme lacking both features (W321F/R418N KatG). Each variant was investigated for its reactions with H2O2 to determine not only standard steady-state kinetic parameters, but also to evaluate the kinetics of enzyme inactivation. Likewise, enzyme intermediates involved in active catalatic turnover as well as those observed during inactivation were monitored by optical stopped-flow and rapid freeze-quench (RFQ) EPR. Across all of these studies, the impact of pH and the presence or absence of peroxidatic electron donor (PxED), were also evaluated. Compared to the wild-type enzyme, our R418N KatG variant showed a 38-fold decrease in catalase catalytic efficiency (i.e., kcat/KM), but a 60-fold increase in peroxidase efficiency (i.e., kcat/KM with respect to H2O2). The catalase activity of R418N KatG was susceptible to inactivation, a phenomenon that varied little with pH. In contrast, the wild-type enzyme was vulnerable to inactivation, but only at acidic pH. These data suggested that withdrawal of the arginine switch, either by pH-dependent conformational change (e.g., wild-type KatG at pH 5) or by mutagenesis (e.g., R418N substitution), created an enzyme more susceptible to inactivation. Inclusion of PxEDs prevented inactivation of KatG in all cases where its susceptibility to inactivation was observed. Interestingly, much greater PxED oxidation per H2O2 consumed was observed for R418N KatG than for wild-type during the catalatic depletion of H2O2. In light of the fact that most PxEDs are too large to transit the narrow substrate access channel leading to the enzyme active site, their peroxidatic oxidation relies on through-protein radical transfer oxidation. As such, peroxidase activity would seem to constitute a record of off-catalase electron transfer events that occur during KatG catalase turnover. The dramatically elevated peroxidase activity of R418N KatG suggests that more PxED intervention is needed by the variant to prevent the accumulation of inactive states. Stated another way, these data indicate a much greater participation of off-pathway electron transfer during H2O2 processing by R418N KatG compared to the wild-type enzyme. Though the kinetics were distinct, enzyme-monitored stopped-flow experiments indicated that upon reaction with H2O2, wild-type (e.g., pH 5) and R418N KatG (e.g., pH 5 and 7) both accumulated catalase-inactive intermediates characterized by spectra with compound II-like (i.e., FeIV=O) features. Consistent with this conclusion, the rates of conversion of these intermediates to the FeIII (i.e., resting) state were not catalytically competent to account for the catalase activity of either protein. For both wild-type and R418N KatG, inclusion of a PxED stimulated catalatic H2O2 consumption, prevented accumulation of FeIV=O-like states, and produced the rapid return of the FeIII enzyme after H2O2 depletion. These data suggest that PxEDs act to prevent accumulation of catalase-inactive species. In our RFQ-EPR studies upon reaction of either wild-type or R418N KatG with H2O2 at pH 5.0, an intense narrow-doublet EPR spectrum was observed at the earliest reaction quench times. This signal has been assigned as an MYW adduct radical. Likewise, for both variants, at a time estimated to correspond to the depletion of H2O2, an exchange-broadened singlet dominated the spectrum. This species has been assigned as a radical centered on the proximal Trp (Trp 321). An identical species has been observed upon reaction of wild-type enzyme with peracetic acid, a peroxide that can oxidize FeIII peroxidases to their compound I (i.e., FeIV=O[porphyrin]•+) state, but does not support full catalase turnover. Spectral features indicative of exchange broadening diminished over time, suggesting that the radicals migrate away from the active site heme iron toward the protein surface. Inclusion of a PxED did not interfere with the MYW•+, but fully prevented the accumulation of other radicals, and produced the fast return of the ferric enzyme. Though the kinetics were different, reaction of R418N KatG at pH 7 showed the same steady-state radical migration from MYW to Trp 321, and to remote oxidizable amino acids. Likewise, only the catalase-essential MYW•+ species were detected upon inclusion of a PxED. These results suggest that Trp-321 is the primary participant for off-pathway electron transfer when the Arg switch is withdrawn by mutagenesis or by the conformational adjustments induced under low-pH conditions. Further, PxEDs serve to prevent accumulation of catalase-inactive species. To further investigate the participation of the proximal tryptophan, a double variant was prepared wherein the Trp was replaced by Phe, a residue that is not oxidizable under typical biological conditions (i.e., W321F/R418N KatG). The double variant produced catalase steady-state kinetic parameters (i.e, kcat, kcat/KM) highly similar to R418N under standard assay conditions. However, the double variant showed almost no resistance to H2O2-dependent catalase inactivation, regardless of pH or the presence or absence of a PxED. On the other hand, the peroxidase activities and the extent of PxED oxidation for H2O2 consumed by the double variant were nearly identical or even higher, compared to R418N KatG, suggesting facile off-catalase through-protein electron transfer via conduit(s) bypassing Trp 321 occurred in the double variant. The disparity between effective peroxidatic oxidation and failure to restore catalase activity by inclusion of a PxED suggested that the off-pathway electron transfer by route(s) bypassing Trp 321 are not effective for PxED-based prevention of KatG inactivation. Further, off-catalase radicals handled by another route(s) appeared to produce the more rapid demise of the enzyme as compared to variant where the proximal Trp was present. KatG variants with an arginine switch (e.g., wt and W321F) are relatively insensitive to H2O2-dependent inactivation at pH 7 but become increasingly so as pH decreases below 6.5. KatG variants without an arginine switch (e.g., R418N and W321F/R418N) showed relatively high sensitivity to H2O2-dependent inactivation across the pH range. This suggests that the absence of the arginine switch either induced by low pH or by mutagenesis produces a KatG more susceptible to inactivation. The peroxidase activities of these enzymes and the extent of PxED oxidation for H2O2 consumed all point to a higher propensity toward off-catalase electron transfer when the arginine switch is absent. At pH 7, where the off-catalase electron transfer is not an issue, the tolerance to H2O2-dependent inactivation followed wtKatG W321F >> R418N > W321F/R418N. In contrast, at pH 5, where there is much greater frequency of off-pathway electron transfer, the order followed wtKatG > R418N > W321F > W321F/R418N. Taken together, our data suggest that Trp 321 enables KatG to most effectively cope with the off-catalase electron transfer events when they occur. In conclusion, the results of the research described in this dissertation indicate the arginine switch is the central player in managing the frequency of off-pathway radical transfer, whereas the proximal tryptophan is an essential residue for handling these events when they occur. These data have potentially far reaching implications. First, they inform the mechanism by which KatG preserves its catalase activity. KatG, on one hand, is the frontline defense against H2O2 in many pathogens, important for pathogen colonization. However, on the other hand, KatG is prone to inactivation at pH 5, the condition coincides with host defense mechanism like neutrophil oxidative stress (i.e., pH 5, large [H2O2]). Understanding the mechanism by which KatG preserves its central catalase activity will lead to better understanding of the overall host-pathogen interaction. This knowledge may allow intervention into the host-pathogen interaction for developing a treatment which enhances a host’s ability to combat pathogenic infections in general. Second, KatG is an ideal system to investigate how oxidoreductases cope with the potentially destructive chemistry they catalyze. This work was supported by a grant from the National Science Foundation (MCB 1616059).