KATG as a Defense Against Hydrogen Peroxide Toxicity: From a Redundant C-Terminal Domain to the Paradoxical Synergy of Two Mutually Antagonistic Activities

Abstract

Catalase-peroxidases (KatGs), first discovered in 1979, are heme-dependent enzymes mainly found in bacteria, archaea and fungi. They are capable of decomposing H2O2 by catalatic and peroxidatic pathways. KatG has garnered considerable attention due to its role in activation of the front-line antitubercular drug isoniazid. Mutations to the katG gene in Mycobacterium tuberculosis are frequently associated with isoniazid resistance. Moreover, periplasmic catalase-peroxidases from highly pathogenic bacteria such as Escherichia coli O157:H7, Yersinia pestis and Legionella pneumophila have been implicated as virulence factors. The robust catalase activity of KatG that even rivals the canonical catalase is astounding given that the enzyme bears a peroxidase-like protein core with little if any resemblance to the canonical catalases. Indeed, the active site of KatG is almost superimposable to peroxidases such as cytochrome c peroxidase and ascorbate peroxidase. Nevertheless, KatG is the only enzyme in the entire superfamily that has appreciable catalase activity. One of the most important structural features that distinguish KatG from its relative peroxidases is an extra C-terminal domain which is proposed to have originated from gene duplication and fusion. It has been observed that the C-terminal domain maintains the active site architecture of KatG that is essential for both catalase and peroxidase activity. It appears to prevent the otherwise spontaneous coordination of the distal histidine to the heme iron. However, the mechanisms by which the C-terminal domain does this from a distance no closer than 30 Å from the active site remains to be elucidated. The research reported in this dissertation evaluated the roles of C-terminal domain in KatG function. In particular, the interface between N-terminal domain harboring the active site and C-terminal domain within each subunit reveals two strictly conserved H-bond networks, namely distant- and near-network based on their distance from the active site. With C-terminal domain’s ability to reactivate the N-terminal domain being employed, stand-alone domain proteins targeting these conserved H-bond interactions in both networks were generated by site-directed mutagenesis and characterized by spectroscopic and kinetic studies. It was shown while the reactivation process was largely unaffected by substitutions, the resulting domain variants exhibited distinct properties from the unmodified domain proteins. Substitution of Arg 117 and Arg 479 substantially shifted the heme species from predominantly narrow rhombic to broad rhombic high-spin. Although catalase activity was diminished to different extent ranging from 24% to 99%, it appeared to have no significant correlation with low-spin contribution observed in the reactivated domain variants. Moreover, substantially impaired peroxidase activity was observed in near-network variants with either Asp 482 or Arg 479 substitution. These results suggest that the C-terminal domain modulated catalase and peroxidase activity of KatG by low-spin-independent mechanisms. In light of peroxidatic electron donors (PxEDs) of peroxidase activity in KatG, it was strikingly observed that instead of inhibiting catalase activity as expected based on the current proposed mechanism, PxEDs stimulated catalatic turnover of the periplasmic catalase-peroxidase from E. coli O157:H7 (KatP) by reduced apparent KM and enhanced apparent kcat. The acidic conditions under which PxED-assisted catalatic turnover was optimal are well mirrored in host immune responses where bacteria are challenged with copious H2O2 in the acidic environment of phagolysosomes. With the model organism E. coli, we constructed single, double and triple mutants with ahpCF, katE and KatG null mutation and demonstrated that KatG played the pivotal role in aerobic growth of E. coli. To evaluate the stimulatory effect observed in vitro, we complemented katE/katG and ahpCF/katE/katG mutants with fully active wild-type katG gene and Y229F katG gene from M. tuberculosis. Our results suggest that in the presence of PxEDs, catalase-inactive Y229F KatG with an even enhanced peroxidase activity was not sufficient to alleviate H2O2-induced growth inhibition. Conversely, PxEDs stimulated the catalase activity of KatG as evidenced by that wild-type katG complemented katE/katG mutant recovered from H2O2 challenge within 2 hours in the presence of TMB compared to a lag period of 12 hours when TMB was absent. Our work provides the first evidence that PxEDs remarkably enhanced bacterial defenses against H2O2 by specifically stimulating the catalase activity of KatG thereby contributed to bacterial survival under H2O2 assaults.